From e52482dec8366a98ac380b3bdc1a4abb8a390914 Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Fri, 23 Mar 2018 15:32:28 -0400
Subject: drm/amdgpu: Add MMU notifier type for KFD userptr

This commit adds the notion of MMU notifier types GFX and HSA. GFX
continues to work like MMU notifiers did before. HSA adds support for
KFD userptr BOs. The implementation of KFD userptr eviction is a stub
for now.

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h       |  1 +
 drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c |  7 ++
 drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c           |  2 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c           | 94 ++++++++++++++++++++----
 drivers/gpu/drm/amd/amdgpu/amdgpu_mn.h           | 11 ++-
 5 files changed, 97 insertions(+), 18 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
index c2c2bea..83e0c5c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
@@ -104,6 +104,7 @@ void amdgpu_amdkfd_device_probe(struct amdgpu_device *adev);
 void amdgpu_amdkfd_device_init(struct amdgpu_device *adev);
 void amdgpu_amdkfd_device_fini(struct amdgpu_device *adev);
 
+int amdgpu_amdkfd_evict_userptr(struct kgd_mem *mem, struct mm_struct *mm);
 int amdgpu_amdkfd_submit_ib(struct kgd_dev *kgd, enum kgd_engine_type engine,
 				uint32_t vmid, uint64_t gpu_addr,
 				uint32_t *ib_cmd, uint32_t ib_len);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
index 1d6e147..2463ff6 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
@@ -1418,6 +1418,13 @@ bo_reserve_failed:
 	return ret;
 }
 
+int amdgpu_amdkfd_evict_userptr(struct kgd_mem *mem,
+				struct mm_struct *mm)
+{
+	/* TODO */
+	return 0;
+}
+
 /** amdgpu_amdkfd_gpuvm_restore_process_bos - Restore all BOs for the given
  *   KFD process identified by process_info
  *
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
index dc34b50..8e66f37 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
@@ -536,7 +536,7 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p,
 	if (p->bo_list) {
 		amdgpu_bo_list_get_list(p->bo_list, &p->validated);
 		if (p->bo_list->first_userptr != p->bo_list->num_entries)
-			p->mn = amdgpu_mn_get(p->adev);
+			p->mn = amdgpu_mn_get(p->adev, AMDGPU_MN_TYPE_GFX);
 	}
 
 	INIT_LIST_HEAD(&duplicates);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c
index bd67f4c..f2ed18e 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c
@@ -36,12 +36,14 @@
 #include <drm/drm.h>
 
 #include "amdgpu.h"
+#include "amdgpu_amdkfd.h"
 
 struct amdgpu_mn {
 	/* constant after initialisation */
 	struct amdgpu_device	*adev;
 	struct mm_struct	*mm;
 	struct mmu_notifier	mn;
+	enum amdgpu_mn_type	type;
 
 	/* only used on destruction */
 	struct work_struct	work;
@@ -185,7 +187,7 @@ static void amdgpu_mn_invalidate_node(struct amdgpu_mn_node *node,
 }
 
 /**
- * amdgpu_mn_invalidate_range_start - callback to notify about mm change
+ * amdgpu_mn_invalidate_range_start_gfx - callback to notify about mm change
  *
  * @mn: our notifier
  * @mn: the mm this callback is about
@@ -195,10 +197,10 @@ static void amdgpu_mn_invalidate_node(struct amdgpu_mn_node *node,
  * We block for all BOs between start and end to be idle and
  * unmap them by move them into system domain again.
  */
-static void amdgpu_mn_invalidate_range_start(struct mmu_notifier *mn,
-					     struct mm_struct *mm,
-					     unsigned long start,
-					     unsigned long end)
+static void amdgpu_mn_invalidate_range_start_gfx(struct mmu_notifier *mn,
+						 struct mm_struct *mm,
+						 unsigned long start,
+						 unsigned long end)
 {
 	struct amdgpu_mn *rmn = container_of(mn, struct amdgpu_mn, mn);
 	struct interval_tree_node *it;
@@ -220,6 +222,49 @@ static void amdgpu_mn_invalidate_range_start(struct mmu_notifier *mn,
 }
 
 /**
+ * amdgpu_mn_invalidate_range_start_hsa - callback to notify about mm change
+ *
+ * @mn: our notifier
+ * @mn: the mm this callback is about
+ * @start: start of updated range
+ * @end: end of updated range
+ *
+ * We temporarily evict all BOs between start and end. This
+ * necessitates evicting all user-mode queues of the process. The BOs
+ * are restorted in amdgpu_mn_invalidate_range_end_hsa.
+ */
+static void amdgpu_mn_invalidate_range_start_hsa(struct mmu_notifier *mn,
+						 struct mm_struct *mm,
+						 unsigned long start,
+						 unsigned long end)
+{
+	struct amdgpu_mn *rmn = container_of(mn, struct amdgpu_mn, mn);
+	struct interval_tree_node *it;
+
+	/* notification is exclusive, but interval is inclusive */
+	end -= 1;
+
+	amdgpu_mn_read_lock(rmn);
+
+	it = interval_tree_iter_first(&rmn->objects, start, end);
+	while (it) {
+		struct amdgpu_mn_node *node;
+		struct amdgpu_bo *bo;
+
+		node = container_of(it, struct amdgpu_mn_node, it);
+		it = interval_tree_iter_next(it, start, end);
+
+		list_for_each_entry(bo, &node->bos, mn_list) {
+			struct kgd_mem *mem = bo->kfd_bo;
+
+			if (amdgpu_ttm_tt_affect_userptr(bo->tbo.ttm,
+							 start, end))
+				amdgpu_amdkfd_evict_userptr(mem, mm);
+		}
+	}
+}
+
+/**
  * amdgpu_mn_invalidate_range_end - callback to notify about mm change
  *
  * @mn: our notifier
@@ -239,23 +284,39 @@ static void amdgpu_mn_invalidate_range_end(struct mmu_notifier *mn,
 	amdgpu_mn_read_unlock(rmn);
 }
 
-static const struct mmu_notifier_ops amdgpu_mn_ops = {
-	.release = amdgpu_mn_release,
-	.invalidate_range_start = amdgpu_mn_invalidate_range_start,
-	.invalidate_range_end = amdgpu_mn_invalidate_range_end,
+static const struct mmu_notifier_ops amdgpu_mn_ops[] = {
+	[AMDGPU_MN_TYPE_GFX] = {
+		.release = amdgpu_mn_release,
+		.invalidate_range_start = amdgpu_mn_invalidate_range_start_gfx,
+		.invalidate_range_end = amdgpu_mn_invalidate_range_end,
+	},
+	[AMDGPU_MN_TYPE_HSA] = {
+		.release = amdgpu_mn_release,
+		.invalidate_range_start = amdgpu_mn_invalidate_range_start_hsa,
+		.invalidate_range_end = amdgpu_mn_invalidate_range_end,
+	},
 };
 
+/* Low bits of any reasonable mm pointer will be unused due to struct
+ * alignment. Use these bits to make a unique key from the mm pointer
+ * and notifier type.
+ */
+#define AMDGPU_MN_KEY(mm, type) ((unsigned long)(mm) + (type))
+
 /**
  * amdgpu_mn_get - create notifier context
  *
  * @adev: amdgpu device pointer
+ * @type: type of MMU notifier context
  *
  * Creates a notifier context for current->mm.
  */
-struct amdgpu_mn *amdgpu_mn_get(struct amdgpu_device *adev)
+struct amdgpu_mn *amdgpu_mn_get(struct amdgpu_device *adev,
+				enum amdgpu_mn_type type)
 {
 	struct mm_struct *mm = current->mm;
 	struct amdgpu_mn *rmn;
+	unsigned long key = AMDGPU_MN_KEY(mm, type);
 	int r;
 
 	mutex_lock(&adev->mn_lock);
@@ -264,8 +325,8 @@ struct amdgpu_mn *amdgpu_mn_get(struct amdgpu_device *adev)
 		return ERR_PTR(-EINTR);
 	}
 
-	hash_for_each_possible(adev->mn_hash, rmn, node, (unsigned long)mm)
-		if (rmn->mm == mm)
+	hash_for_each_possible(adev->mn_hash, rmn, node, key)
+		if (AMDGPU_MN_KEY(rmn->mm, rmn->type) == key)
 			goto release_locks;
 
 	rmn = kzalloc(sizeof(*rmn), GFP_KERNEL);
@@ -276,8 +337,9 @@ struct amdgpu_mn *amdgpu_mn_get(struct amdgpu_device *adev)
 
 	rmn->adev = adev;
 	rmn->mm = mm;
-	rmn->mn.ops = &amdgpu_mn_ops;
 	init_rwsem(&rmn->lock);
+	rmn->type = type;
+	rmn->mn.ops = &amdgpu_mn_ops[type];
 	rmn->objects = RB_ROOT_CACHED;
 	mutex_init(&rmn->read_lock);
 	atomic_set(&rmn->recursion, 0);
@@ -286,7 +348,7 @@ struct amdgpu_mn *amdgpu_mn_get(struct amdgpu_device *adev)
 	if (r)
 		goto free_rmn;
 
-	hash_add(adev->mn_hash, &rmn->node, (unsigned long)mm);
+	hash_add(adev->mn_hash, &rmn->node, AMDGPU_MN_KEY(mm, type));
 
 release_locks:
 	up_write(&mm->mmap_sem);
@@ -315,12 +377,14 @@ int amdgpu_mn_register(struct amdgpu_bo *bo, unsigned long addr)
 {
 	unsigned long end = addr + amdgpu_bo_size(bo) - 1;
 	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
+	enum amdgpu_mn_type type =
+		bo->kfd_bo ? AMDGPU_MN_TYPE_HSA : AMDGPU_MN_TYPE_GFX;
 	struct amdgpu_mn *rmn;
 	struct amdgpu_mn_node *node = NULL;
 	struct list_head bos;
 	struct interval_tree_node *it;
 
-	rmn = amdgpu_mn_get(adev);
+	rmn = amdgpu_mn_get(adev, type);
 	if (IS_ERR(rmn))
 		return PTR_ERR(rmn);
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.h
index d0095a3..eb0f432 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.h
@@ -29,16 +29,23 @@
  */
 struct amdgpu_mn;
 
+enum amdgpu_mn_type {
+	AMDGPU_MN_TYPE_GFX,
+	AMDGPU_MN_TYPE_HSA,
+};
+
 #if defined(CONFIG_MMU_NOTIFIER)
 void amdgpu_mn_lock(struct amdgpu_mn *mn);
 void amdgpu_mn_unlock(struct amdgpu_mn *mn);
-struct amdgpu_mn *amdgpu_mn_get(struct amdgpu_device *adev);
+struct amdgpu_mn *amdgpu_mn_get(struct amdgpu_device *adev,
+				enum amdgpu_mn_type type);
 int amdgpu_mn_register(struct amdgpu_bo *bo, unsigned long addr);
 void amdgpu_mn_unregister(struct amdgpu_bo *bo);
 #else
 static inline void amdgpu_mn_lock(struct amdgpu_mn *mn) {}
 static inline void amdgpu_mn_unlock(struct amdgpu_mn *mn) {}
-static inline struct amdgpu_mn *amdgpu_mn_get(struct amdgpu_device *adev)
+static inline struct amdgpu_mn *amdgpu_mn_get(struct amdgpu_device *adev,
+					      enum amdgpu_mn_type type)
 {
 	return NULL;
 }
-- 
cgit v1.1


From 0919195f2b0d7437cb0de49b8975fdd7b5575490 Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Fri, 23 Mar 2018 15:32:29 -0400
Subject: drm/amdgpu: Enable amdgpu_ttm_tt_get_user_pages in worker threads

This commit allows amdgpu_ttm_tt_get_user_pages to work in a worker
thread rather than regular process context. This will be used when
KFD userptr BOs are restored after an MMU-notifier eviction.

v2: Manage task reference with get_task_struct/put_task_struct

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Oded Gabbay <oded.gabbay@gmail.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 38 +++++++++++++++++++++++++--------
 1 file changed, 29 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index 205da3f..c713d30 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -695,7 +695,7 @@ struct amdgpu_ttm_tt {
 	struct ttm_dma_tt	ttm;
 	u64			offset;
 	uint64_t		userptr;
-	struct mm_struct	*usermm;
+	struct task_struct	*usertask;
 	uint32_t		userflags;
 	spinlock_t              guptasklock;
 	struct list_head        guptasks;
@@ -706,14 +706,18 @@ struct amdgpu_ttm_tt {
 int amdgpu_ttm_tt_get_user_pages(struct ttm_tt *ttm, struct page **pages)
 {
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
+	struct mm_struct *mm = gtt->usertask->mm;
 	unsigned int flags = 0;
 	unsigned pinned = 0;
 	int r;
 
+	if (!mm) /* Happens during process shutdown */
+		return -ESRCH;
+
 	if (!(gtt->userflags & AMDGPU_GEM_USERPTR_READONLY))
 		flags |= FOLL_WRITE;
 
-	down_read(&current->mm->mmap_sem);
+	down_read(&mm->mmap_sem);
 
 	if (gtt->userflags & AMDGPU_GEM_USERPTR_ANONONLY) {
 		/* check that we only use anonymous memory
@@ -721,9 +725,9 @@ int amdgpu_ttm_tt_get_user_pages(struct ttm_tt *ttm, struct page **pages)
 		unsigned long end = gtt->userptr + ttm->num_pages * PAGE_SIZE;
 		struct vm_area_struct *vma;
 
-		vma = find_vma(gtt->usermm, gtt->userptr);
+		vma = find_vma(mm, gtt->userptr);
 		if (!vma || vma->vm_file || vma->vm_end < end) {
-			up_read(&current->mm->mmap_sem);
+			up_read(&mm->mmap_sem);
 			return -EPERM;
 		}
 	}
@@ -739,7 +743,12 @@ int amdgpu_ttm_tt_get_user_pages(struct ttm_tt *ttm, struct page **pages)
 		list_add(&guptask.list, &gtt->guptasks);
 		spin_unlock(&gtt->guptasklock);
 
-		r = get_user_pages(userptr, num_pages, flags, p, NULL);
+		if (mm == current->mm)
+			r = get_user_pages(userptr, num_pages, flags, p, NULL);
+		else
+			r = get_user_pages_remote(gtt->usertask,
+					mm, userptr, num_pages,
+					flags, p, NULL, NULL);
 
 		spin_lock(&gtt->guptasklock);
 		list_del(&guptask.list);
@@ -752,12 +761,12 @@ int amdgpu_ttm_tt_get_user_pages(struct ttm_tt *ttm, struct page **pages)
 
 	} while (pinned < ttm->num_pages);
 
-	up_read(&current->mm->mmap_sem);
+	up_read(&mm->mmap_sem);
 	return 0;
 
 release_pages:
 	release_pages(pages, pinned);
-	up_read(&current->mm->mmap_sem);
+	up_read(&mm->mmap_sem);
 	return r;
 }
 
@@ -978,6 +987,9 @@ static void amdgpu_ttm_backend_destroy(struct ttm_tt *ttm)
 {
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
 
+	if (gtt->usertask)
+		put_task_struct(gtt->usertask);
+
 	ttm_dma_tt_fini(&gtt->ttm);
 	kfree(gtt);
 }
@@ -1079,8 +1091,13 @@ int amdgpu_ttm_tt_set_userptr(struct ttm_tt *ttm, uint64_t addr,
 		return -EINVAL;
 
 	gtt->userptr = addr;
-	gtt->usermm = current->mm;
 	gtt->userflags = flags;
+
+	if (gtt->usertask)
+		put_task_struct(gtt->usertask);
+	gtt->usertask = current->group_leader;
+	get_task_struct(gtt->usertask);
+
 	spin_lock_init(&gtt->guptasklock);
 	INIT_LIST_HEAD(&gtt->guptasks);
 	atomic_set(&gtt->mmu_invalidations, 0);
@@ -1096,7 +1113,10 @@ struct mm_struct *amdgpu_ttm_tt_get_usermm(struct ttm_tt *ttm)
 	if (gtt == NULL)
 		return NULL;
 
-	return gtt->usermm;
+	if (gtt->usertask == NULL)
+		return NULL;
+
+	return gtt->usertask->mm;
 }
 
 bool amdgpu_ttm_tt_affect_userptr(struct ttm_tt *ttm, unsigned long start,
-- 
cgit v1.1


From 6e08e0995b8f339fd2a7ee4fa11f17396405ef60 Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Fri, 23 Mar 2018 15:32:30 -0400
Subject: drm/amdgpu: Avoid reclaim while holding locks taken in MMU notifier
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

When an MMU notifier runs in memory reclaim context, it can deadlock
trying to take locks that are already held in the thread causing the
memory reclaim. The solution is to avoid memory reclaim while holding
locks that are taken in MMU notifiers.

This commit fixes kmalloc while holding rmn->lock by moving the call
outside the lock. The GFX MMU notifier also locks reservation objects.
I have no good solution for avoiding reclaim while holding reservation
objects. The HSA MMU notifier will not lock any reservation objects.

v2: Moved allocation outside lock instead of using GFP_NOIO

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Oded Gabbay <oded.gabbay@gmail.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c | 17 +++++++++--------
 1 file changed, 9 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c
index f2ed18e..83e344f 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c
@@ -380,7 +380,7 @@ int amdgpu_mn_register(struct amdgpu_bo *bo, unsigned long addr)
 	enum amdgpu_mn_type type =
 		bo->kfd_bo ? AMDGPU_MN_TYPE_HSA : AMDGPU_MN_TYPE_GFX;
 	struct amdgpu_mn *rmn;
-	struct amdgpu_mn_node *node = NULL;
+	struct amdgpu_mn_node *node = NULL, *new_node;
 	struct list_head bos;
 	struct interval_tree_node *it;
 
@@ -388,6 +388,10 @@ int amdgpu_mn_register(struct amdgpu_bo *bo, unsigned long addr)
 	if (IS_ERR(rmn))
 		return PTR_ERR(rmn);
 
+	new_node = kmalloc(sizeof(*new_node), GFP_KERNEL);
+	if (!new_node)
+		return -ENOMEM;
+
 	INIT_LIST_HEAD(&bos);
 
 	down_write(&rmn->lock);
@@ -401,13 +405,10 @@ int amdgpu_mn_register(struct amdgpu_bo *bo, unsigned long addr)
 		list_splice(&node->bos, &bos);
 	}
 
-	if (!node) {
-		node = kmalloc(sizeof(struct amdgpu_mn_node), GFP_KERNEL);
-		if (!node) {
-			up_write(&rmn->lock);
-			return -ENOMEM;
-		}
-	}
+	if (!node)
+		node = new_node;
+	else
+		kfree(new_node);
 
 	bo->mn = rmn;
 
-- 
cgit v1.1


From d1853f42b63da94fa0147091d22bf5675b0ff89b Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Fri, 23 Mar 2018 15:32:31 -0400
Subject: drm/amdkfd: GFP_NOIO while holding locks taken in MMU notifier

When an MMU notifier runs in memory reclaim context, it can deadlock
trying to take locks that are already held in the thread causing the
memory reclaim. The solution is to avoid memory reclaim while holding
locks that are taken in MMU notifiers by using GFP_NOIO.

This commit fixes memory allocations done while holding the dqm->lock
which is needed in the MMU notifier (dqm->ops.evict_process_queues).

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/kfd_device.c          | 2 +-
 drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c | 2 +-
 drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c  | 2 +-
 3 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
index 334669996..0434f65 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
@@ -652,7 +652,7 @@ int kfd_gtt_sa_allocate(struct kfd_dev *kfd, unsigned int size,
 	if (size > kfd->gtt_sa_num_of_chunks * kfd->gtt_sa_chunk_size)
 		return -ENOMEM;
 
-	*mem_obj = kmalloc(sizeof(struct kfd_mem_obj), GFP_KERNEL);
+	*mem_obj = kmalloc(sizeof(struct kfd_mem_obj), GFP_NOIO);
 	if ((*mem_obj) == NULL)
 		return -ENOMEM;
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c
index c00c325..2bc49c6 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c
@@ -412,7 +412,7 @@ struct mqd_manager *mqd_manager_init_cik(enum KFD_MQD_TYPE type,
 	if (WARN_ON(type >= KFD_MQD_TYPE_MAX))
 		return NULL;
 
-	mqd = kzalloc(sizeof(*mqd), GFP_KERNEL);
+	mqd = kzalloc(sizeof(*mqd), GFP_NOIO);
 	if (!mqd)
 		return NULL;
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c
index 89e4242..481307b 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c
@@ -394,7 +394,7 @@ struct mqd_manager *mqd_manager_init_vi(enum KFD_MQD_TYPE type,
 	if (WARN_ON(type >= KFD_MQD_TYPE_MAX))
 		return NULL;
 
-	mqd = kzalloc(sizeof(*mqd), GFP_KERNEL);
+	mqd = kzalloc(sizeof(*mqd), GFP_NOIO);
 	if (!mqd)
 		return NULL;
 
-- 
cgit v1.1


From 6b95e7973a136181e37446bd29b0b2e2f0d2d653 Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Fri, 23 Mar 2018 15:32:32 -0400
Subject: drm/amdkfd: Add quiesce_mm and resume_mm to kgd2kfd_calls

These interfaces allow KGD to stop and resume all GPU user mode queue
access to a process address space. This is needed for handling MMU
notifiers of userptrs mapped for GPU access in KFD VMs.

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/kfd_device.c         | 38 +++++++++++++++++++++++++
 drivers/gpu/drm/amd/amdkfd/kfd_module.c         |  2 ++
 drivers/gpu/drm/amd/amdkfd/kfd_priv.h           |  4 +++
 drivers/gpu/drm/amd/amdkfd/kfd_process.c        | 10 +++----
 drivers/gpu/drm/amd/include/kgd_kfd_interface.h |  6 ++++
 5 files changed, 55 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
index 0434f65..7b57995 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
@@ -541,6 +541,44 @@ void kgd2kfd_interrupt(struct kfd_dev *kfd, const void *ih_ring_entry)
 	spin_unlock(&kfd->interrupt_lock);
 }
 
+int kgd2kfd_quiesce_mm(struct mm_struct *mm)
+{
+	struct kfd_process *p;
+	int r;
+
+	/* Because we are called from arbitrary context (workqueue) as opposed
+	 * to process context, kfd_process could attempt to exit while we are
+	 * running so the lookup function increments the process ref count.
+	 */
+	p = kfd_lookup_process_by_mm(mm);
+	if (!p)
+		return -ESRCH;
+
+	r = kfd_process_evict_queues(p);
+
+	kfd_unref_process(p);
+	return r;
+}
+
+int kgd2kfd_resume_mm(struct mm_struct *mm)
+{
+	struct kfd_process *p;
+	int r;
+
+	/* Because we are called from arbitrary context (workqueue) as opposed
+	 * to process context, kfd_process could attempt to exit while we are
+	 * running so the lookup function increments the process ref count.
+	 */
+	p = kfd_lookup_process_by_mm(mm);
+	if (!p)
+		return -ESRCH;
+
+	r = kfd_process_restore_queues(p);
+
+	kfd_unref_process(p);
+	return r;
+}
+
 /** kgd2kfd_schedule_evict_and_restore_process - Schedules work queue that will
  *   prepare for safe eviction of KFD BOs that belong to the specified
  *   process.
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_module.c b/drivers/gpu/drm/amd/amdkfd/kfd_module.c
index e0c07d2..45bc458 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_module.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_module.c
@@ -43,6 +43,8 @@ static const struct kgd2kfd_calls kgd2kfd = {
 	.interrupt	= kgd2kfd_interrupt,
 	.suspend	= kgd2kfd_suspend,
 	.resume		= kgd2kfd_resume,
+	.quiesce_mm	= kgd2kfd_quiesce_mm,
+	.resume_mm	= kgd2kfd_resume_mm,
 	.schedule_evict_and_restore_process =
 			  kgd2kfd_schedule_evict_and_restore_process,
 };
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
index 96a9cc0..4d5c49e 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
@@ -512,6 +512,8 @@ struct qcm_process_device {
 /* Approx. time before evicting the process again */
 #define PROCESS_ACTIVE_TIME_MS 10
 
+int kgd2kfd_quiesce_mm(struct mm_struct *mm);
+int kgd2kfd_resume_mm(struct mm_struct *mm);
 int kgd2kfd_schedule_evict_and_restore_process(struct mm_struct *mm,
 					       struct dma_fence *fence);
 
@@ -681,6 +683,8 @@ struct kfd_process *kfd_get_process(const struct task_struct *);
 struct kfd_process *kfd_lookup_process_by_pasid(unsigned int pasid);
 struct kfd_process *kfd_lookup_process_by_mm(const struct mm_struct *mm);
 void kfd_unref_process(struct kfd_process *p);
+int kfd_process_evict_queues(struct kfd_process *p);
+int kfd_process_restore_queues(struct kfd_process *p);
 void kfd_suspend_all_processes(void);
 int kfd_resume_all_processes(void);
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process.c b/drivers/gpu/drm/amd/amdkfd/kfd_process.c
index 1711ad0..2791e72 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_process.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c
@@ -808,7 +808,7 @@ struct kfd_process *kfd_lookup_process_by_mm(const struct mm_struct *mm)
  * Eviction is reference-counted per process-device. This means multiple
  * evictions from different sources can be nested safely.
  */
-static int process_evict_queues(struct kfd_process *p)
+int kfd_process_evict_queues(struct kfd_process *p)
 {
 	struct kfd_process_device *pdd;
 	int r = 0;
@@ -844,7 +844,7 @@ fail:
 }
 
 /* process_restore_queues - Restore all user queues of a process */
-static  int process_restore_queues(struct kfd_process *p)
+int kfd_process_restore_queues(struct kfd_process *p)
 {
 	struct kfd_process_device *pdd;
 	int r, ret = 0;
@@ -886,7 +886,7 @@ static void evict_process_worker(struct work_struct *work)
 	flush_delayed_work(&p->restore_work);
 
 	pr_debug("Started evicting pasid %d\n", p->pasid);
-	ret = process_evict_queues(p);
+	ret = kfd_process_evict_queues(p);
 	if (!ret) {
 		dma_fence_signal(p->ef);
 		dma_fence_put(p->ef);
@@ -946,7 +946,7 @@ static void restore_process_worker(struct work_struct *work)
 		return;
 	}
 
-	ret = process_restore_queues(p);
+	ret = kfd_process_restore_queues(p);
 	if (!ret)
 		pr_debug("Finished restoring pasid %d\n", p->pasid);
 	else
@@ -963,7 +963,7 @@ void kfd_suspend_all_processes(void)
 		cancel_delayed_work_sync(&p->eviction_work);
 		cancel_delayed_work_sync(&p->restore_work);
 
-		if (process_evict_queues(p))
+		if (kfd_process_evict_queues(p))
 			pr_err("Failed to suspend process %d\n", p->pasid);
 		dma_fence_signal(p->ef);
 		dma_fence_put(p->ef);
diff --git a/drivers/gpu/drm/amd/include/kgd_kfd_interface.h b/drivers/gpu/drm/amd/include/kgd_kfd_interface.h
index 237289a..286cfe7 100644
--- a/drivers/gpu/drm/amd/include/kgd_kfd_interface.h
+++ b/drivers/gpu/drm/amd/include/kgd_kfd_interface.h
@@ -382,6 +382,10 @@ struct kfd2kgd_calls {
  *
  * @resume: Notifies amdkfd about a resume action done to a kgd device
  *
+ * @quiesce_mm: Quiesce all user queue access to specified MM address space
+ *
+ * @resume_mm: Resume user queue access to specified MM address space
+ *
  * @schedule_evict_and_restore_process: Schedules work queue that will prepare
  * for safe eviction of KFD BOs that belong to the specified process.
  *
@@ -399,6 +403,8 @@ struct kgd2kfd_calls {
 	void (*interrupt)(struct kfd_dev *kfd, const void *ih_ring_entry);
 	void (*suspend)(struct kfd_dev *kfd);
 	int (*resume)(struct kfd_dev *kfd);
+	int (*quiesce_mm)(struct mm_struct *mm);
+	int (*resume_mm)(struct mm_struct *mm);
 	int (*schedule_evict_and_restore_process)(struct mm_struct *mm,
 			struct dma_fence *fence);
 };
-- 
cgit v1.1


From 5ae0283e831a94c714fce61063e4724baf364ef3 Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Fri, 23 Mar 2018 15:32:33 -0400
Subject: drm/amdgpu: Add userptr support for KFD

This adds support for allocating, mapping, unmapping and freeing
userptr BOs, and for handling MMU notifiers.

v2: updated a comment

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h       |  11 +
 drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c | 567 ++++++++++++++++++++++-
 2 files changed, 554 insertions(+), 24 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
index 83e0c5c..c3024b1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
@@ -28,6 +28,7 @@
 #include <linux/types.h>
 #include <linux/mm.h>
 #include <linux/mmu_context.h>
+#include <linux/workqueue.h>
 #include <kgd_kfd_interface.h>
 #include <drm/ttm/ttm_execbuf_util.h>
 #include "amdgpu_sync.h"
@@ -59,7 +60,9 @@ struct kgd_mem {
 
 	uint32_t mapping_flags;
 
+	atomic_t invalid;
 	struct amdkfd_process_info *process_info;
+	struct page **user_pages;
 
 	struct amdgpu_sync sync;
 
@@ -84,6 +87,9 @@ struct amdkfd_process_info {
 	struct list_head vm_list_head;
 	/* List head for all KFD BOs that belong to a KFD process. */
 	struct list_head kfd_bo_list;
+	/* List of userptr BOs that are valid or invalid */
+	struct list_head userptr_valid_list;
+	struct list_head userptr_inval_list;
 	/* Lock to protect kfd_bo_list */
 	struct mutex lock;
 
@@ -91,6 +97,11 @@ struct amdkfd_process_info {
 	unsigned int n_vms;
 	/* Eviction Fence */
 	struct amdgpu_amdkfd_fence *eviction_fence;
+
+	/* MMU-notifier related fields */
+	atomic_t evicted_bos;
+	struct delayed_work restore_userptr_work;
+	struct pid *pid;
 };
 
 int amdgpu_amdkfd_init(void);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
index 2463ff6..5296e24 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
@@ -23,6 +23,7 @@
 #define pr_fmt(fmt) "kfd2kgd: " fmt
 
 #include <linux/list.h>
+#include <linux/sched/mm.h>
 #include <drm/drmP.h>
 #include "amdgpu_object.h"
 #include "amdgpu_vm.h"
@@ -33,10 +34,20 @@
  */
 #define VI_BO_SIZE_ALIGN (0x8000)
 
+/* BO flag to indicate a KFD userptr BO */
+#define AMDGPU_AMDKFD_USERPTR_BO (1ULL << 63)
+
+/* Userptr restore delay, just long enough to allow consecutive VM
+ * changes to accumulate
+ */
+#define AMDGPU_USERPTR_RESTORE_DELAY_MS 1
+
 /* Impose limit on how much memory KFD can use */
 static struct {
 	uint64_t max_system_mem_limit;
+	uint64_t max_userptr_mem_limit;
 	int64_t system_mem_used;
+	int64_t userptr_mem_used;
 	spinlock_t mem_limit_lock;
 } kfd_mem_limit;
 
@@ -57,6 +68,7 @@ static const char * const domain_bit_to_string[] = {
 
 #define domain_string(domain) domain_bit_to_string[ffs(domain)-1]
 
+static void amdgpu_amdkfd_restore_userptr_worker(struct work_struct *work);
 
 
 static inline struct amdgpu_device *get_amdgpu_device(struct kgd_dev *kgd)
@@ -78,6 +90,7 @@ static bool check_if_add_bo_to_vm(struct amdgpu_vm *avm,
 
 /* Set memory usage limits. Current, limits are
  *  System (kernel) memory - 3/8th System RAM
+ *  Userptr memory - 3/4th System RAM
  */
 void amdgpu_amdkfd_gpuvm_init_mem_limits(void)
 {
@@ -90,8 +103,10 @@ void amdgpu_amdkfd_gpuvm_init_mem_limits(void)
 
 	spin_lock_init(&kfd_mem_limit.mem_limit_lock);
 	kfd_mem_limit.max_system_mem_limit = (mem >> 1) - (mem >> 3);
-	pr_debug("Kernel memory limit %lluM\n",
-		(kfd_mem_limit.max_system_mem_limit >> 20));
+	kfd_mem_limit.max_userptr_mem_limit = mem - (mem >> 2);
+	pr_debug("Kernel memory limit %lluM, userptr limit %lluM\n",
+		(kfd_mem_limit.max_system_mem_limit >> 20),
+		(kfd_mem_limit.max_userptr_mem_limit >> 20));
 }
 
 static int amdgpu_amdkfd_reserve_system_mem_limit(struct amdgpu_device *adev,
@@ -111,6 +126,16 @@ static int amdgpu_amdkfd_reserve_system_mem_limit(struct amdgpu_device *adev,
 			goto err_no_mem;
 		}
 		kfd_mem_limit.system_mem_used += (acc_size + size);
+	} else if (domain == AMDGPU_GEM_DOMAIN_CPU) {
+		if ((kfd_mem_limit.system_mem_used + acc_size >
+			kfd_mem_limit.max_system_mem_limit) ||
+			(kfd_mem_limit.userptr_mem_used + (size + acc_size) >
+			kfd_mem_limit.max_userptr_mem_limit)) {
+			ret = -ENOMEM;
+			goto err_no_mem;
+		}
+		kfd_mem_limit.system_mem_used += acc_size;
+		kfd_mem_limit.userptr_mem_used += size;
 	}
 err_no_mem:
 	spin_unlock(&kfd_mem_limit.mem_limit_lock);
@@ -126,10 +151,16 @@ static void unreserve_system_mem_limit(struct amdgpu_device *adev,
 				       sizeof(struct amdgpu_bo));
 
 	spin_lock(&kfd_mem_limit.mem_limit_lock);
-	if (domain == AMDGPU_GEM_DOMAIN_GTT)
+	if (domain == AMDGPU_GEM_DOMAIN_GTT) {
 		kfd_mem_limit.system_mem_used -= (acc_size + size);
+	} else if (domain == AMDGPU_GEM_DOMAIN_CPU) {
+		kfd_mem_limit.system_mem_used -= acc_size;
+		kfd_mem_limit.userptr_mem_used -= size;
+	}
 	WARN_ONCE(kfd_mem_limit.system_mem_used < 0,
 		  "kfd system memory accounting unbalanced");
+	WARN_ONCE(kfd_mem_limit.userptr_mem_used < 0,
+		  "kfd userptr memory accounting unbalanced");
 
 	spin_unlock(&kfd_mem_limit.mem_limit_lock);
 }
@@ -138,12 +169,17 @@ void amdgpu_amdkfd_unreserve_system_memory_limit(struct amdgpu_bo *bo)
 {
 	spin_lock(&kfd_mem_limit.mem_limit_lock);
 
-	if (bo->preferred_domains == AMDGPU_GEM_DOMAIN_GTT) {
+	if (bo->flags & AMDGPU_AMDKFD_USERPTR_BO) {
+		kfd_mem_limit.system_mem_used -= bo->tbo.acc_size;
+		kfd_mem_limit.userptr_mem_used -= amdgpu_bo_size(bo);
+	} else if (bo->preferred_domains == AMDGPU_GEM_DOMAIN_GTT) {
 		kfd_mem_limit.system_mem_used -=
 			(bo->tbo.acc_size + amdgpu_bo_size(bo));
 	}
 	WARN_ONCE(kfd_mem_limit.system_mem_used < 0,
 		  "kfd system memory accounting unbalanced");
+	WARN_ONCE(kfd_mem_limit.userptr_mem_used < 0,
+		  "kfd userptr memory accounting unbalanced");
 
 	spin_unlock(&kfd_mem_limit.mem_limit_lock);
 }
@@ -506,7 +542,8 @@ static void remove_bo_from_vm(struct amdgpu_device *adev,
 }
 
 static void add_kgd_mem_to_kfd_bo_list(struct kgd_mem *mem,
-				struct amdkfd_process_info *process_info)
+				struct amdkfd_process_info *process_info,
+				bool userptr)
 {
 	struct ttm_validate_buffer *entry = &mem->validate_list;
 	struct amdgpu_bo *bo = mem->bo;
@@ -515,8 +552,93 @@ static void add_kgd_mem_to_kfd_bo_list(struct kgd_mem *mem,
 	entry->shared = true;
 	entry->bo = &bo->tbo;
 	mutex_lock(&process_info->lock);
-	list_add_tail(&entry->head, &process_info->kfd_bo_list);
+	if (userptr)
+		list_add_tail(&entry->head, &process_info->userptr_valid_list);
+	else
+		list_add_tail(&entry->head, &process_info->kfd_bo_list);
+	mutex_unlock(&process_info->lock);
+}
+
+/* Initializes user pages. It registers the MMU notifier and validates
+ * the userptr BO in the GTT domain.
+ *
+ * The BO must already be on the userptr_valid_list. Otherwise an
+ * eviction and restore may happen that leaves the new BO unmapped
+ * with the user mode queues running.
+ *
+ * Takes the process_info->lock to protect against concurrent restore
+ * workers.
+ *
+ * Returns 0 for success, negative errno for errors.
+ */
+static int init_user_pages(struct kgd_mem *mem, struct mm_struct *mm,
+			   uint64_t user_addr)
+{
+	struct amdkfd_process_info *process_info = mem->process_info;
+	struct amdgpu_bo *bo = mem->bo;
+	struct ttm_operation_ctx ctx = { true, false };
+	int ret = 0;
+
+	mutex_lock(&process_info->lock);
+
+	ret = amdgpu_ttm_tt_set_userptr(bo->tbo.ttm, user_addr, 0);
+	if (ret) {
+		pr_err("%s: Failed to set userptr: %d\n", __func__, ret);
+		goto out;
+	}
+
+	ret = amdgpu_mn_register(bo, user_addr);
+	if (ret) {
+		pr_err("%s: Failed to register MMU notifier: %d\n",
+		       __func__, ret);
+		goto out;
+	}
+
+	/* If no restore worker is running concurrently, user_pages
+	 * should not be allocated
+	 */
+	WARN(mem->user_pages, "Leaking user_pages array");
+
+	mem->user_pages = kvmalloc_array(bo->tbo.ttm->num_pages,
+					   sizeof(struct page *),
+					   GFP_KERNEL | __GFP_ZERO);
+	if (!mem->user_pages) {
+		pr_err("%s: Failed to allocate pages array\n", __func__);
+		ret = -ENOMEM;
+		goto unregister_out;
+	}
+
+	ret = amdgpu_ttm_tt_get_user_pages(bo->tbo.ttm, mem->user_pages);
+	if (ret) {
+		pr_err("%s: Failed to get user pages: %d\n", __func__, ret);
+		goto free_out;
+	}
+
+	amdgpu_ttm_tt_set_user_pages(bo->tbo.ttm, mem->user_pages);
+
+	ret = amdgpu_bo_reserve(bo, true);
+	if (ret) {
+		pr_err("%s: Failed to reserve BO\n", __func__);
+		goto release_out;
+	}
+	amdgpu_ttm_placement_from_domain(bo, mem->domain);
+	ret = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
+	if (ret)
+		pr_err("%s: failed to validate BO\n", __func__);
+	amdgpu_bo_unreserve(bo);
+
+release_out:
+	if (ret)
+		release_pages(mem->user_pages, bo->tbo.ttm->num_pages);
+free_out:
+	kvfree(mem->user_pages);
+	mem->user_pages = NULL;
+unregister_out:
+	if (ret)
+		amdgpu_mn_unregister(bo);
+out:
 	mutex_unlock(&process_info->lock);
+	return ret;
 }
 
 /* Reserving a BO and its page table BOs must happen atomically to
@@ -748,7 +870,8 @@ static int update_gpuvm_pte(struct amdgpu_device *adev,
 }
 
 static int map_bo_to_gpuvm(struct amdgpu_device *adev,
-		struct kfd_bo_va_list *entry, struct amdgpu_sync *sync)
+		struct kfd_bo_va_list *entry, struct amdgpu_sync *sync,
+		bool no_update_pte)
 {
 	int ret;
 
@@ -762,6 +885,9 @@ static int map_bo_to_gpuvm(struct amdgpu_device *adev,
 		return ret;
 	}
 
+	if (no_update_pte)
+		return 0;
+
 	ret = update_gpuvm_pte(adev, entry, sync);
 	if (ret) {
 		pr_err("update_gpuvm_pte() failed\n");
@@ -820,6 +946,8 @@ static int init_kfd_vm(struct amdgpu_vm *vm, void **process_info,
 		mutex_init(&info->lock);
 		INIT_LIST_HEAD(&info->vm_list_head);
 		INIT_LIST_HEAD(&info->kfd_bo_list);
+		INIT_LIST_HEAD(&info->userptr_valid_list);
+		INIT_LIST_HEAD(&info->userptr_inval_list);
 
 		info->eviction_fence =
 			amdgpu_amdkfd_fence_create(dma_fence_context_alloc(1),
@@ -830,6 +958,11 @@ static int init_kfd_vm(struct amdgpu_vm *vm, void **process_info,
 			goto create_evict_fence_fail;
 		}
 
+		info->pid = get_task_pid(current->group_leader, PIDTYPE_PID);
+		atomic_set(&info->evicted_bos, 0);
+		INIT_DELAYED_WORK(&info->restore_userptr_work,
+				  amdgpu_amdkfd_restore_userptr_worker);
+
 		*process_info = info;
 		*ef = dma_fence_get(&info->eviction_fence->base);
 	}
@@ -872,6 +1005,7 @@ reserve_pd_fail:
 		dma_fence_put(*ef);
 		*ef = NULL;
 		*process_info = NULL;
+		put_pid(info->pid);
 create_evict_fence_fail:
 		mutex_destroy(&info->lock);
 		kfree(info);
@@ -967,8 +1101,12 @@ void amdgpu_amdkfd_gpuvm_destroy_cb(struct amdgpu_device *adev,
 	/* Release per-process resources when last compute VM is destroyed */
 	if (!process_info->n_vms) {
 		WARN_ON(!list_empty(&process_info->kfd_bo_list));
+		WARN_ON(!list_empty(&process_info->userptr_valid_list));
+		WARN_ON(!list_empty(&process_info->userptr_inval_list));
 
 		dma_fence_put(&process_info->eviction_fence->base);
+		cancel_delayed_work_sync(&process_info->restore_userptr_work);
+		put_pid(process_info->pid);
 		mutex_destroy(&process_info->lock);
 		kfree(process_info);
 	}
@@ -1003,9 +1141,10 @@ int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu(
 {
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
 	struct amdgpu_vm *avm = (struct amdgpu_vm *)vm;
+	uint64_t user_addr = 0;
 	struct amdgpu_bo *bo;
 	int byte_align;
-	u32 alloc_domain;
+	u32 domain, alloc_domain;
 	u64 alloc_flags;
 	uint32_t mapping_flags;
 	int ret;
@@ -1014,14 +1153,21 @@ int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu(
 	 * Check on which domain to allocate BO
 	 */
 	if (flags & ALLOC_MEM_FLAGS_VRAM) {
-		alloc_domain = AMDGPU_GEM_DOMAIN_VRAM;
+		domain = alloc_domain = AMDGPU_GEM_DOMAIN_VRAM;
 		alloc_flags = AMDGPU_GEM_CREATE_VRAM_CLEARED;
 		alloc_flags |= (flags & ALLOC_MEM_FLAGS_PUBLIC) ?
 			AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED :
 			AMDGPU_GEM_CREATE_NO_CPU_ACCESS;
 	} else if (flags & ALLOC_MEM_FLAGS_GTT) {
-		alloc_domain = AMDGPU_GEM_DOMAIN_GTT;
+		domain = alloc_domain = AMDGPU_GEM_DOMAIN_GTT;
 		alloc_flags = 0;
+	} else if (flags & ALLOC_MEM_FLAGS_USERPTR) {
+		domain = AMDGPU_GEM_DOMAIN_GTT;
+		alloc_domain = AMDGPU_GEM_DOMAIN_CPU;
+		alloc_flags = 0;
+		if (!offset || !*offset)
+			return -EINVAL;
+		user_addr = *offset;
 	} else {
 		return -EINVAL;
 	}
@@ -1078,18 +1224,34 @@ int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu(
 	}
 	bo->kfd_bo = *mem;
 	(*mem)->bo = bo;
+	if (user_addr)
+		bo->flags |= AMDGPU_AMDKFD_USERPTR_BO;
 
 	(*mem)->va = va;
-	(*mem)->domain = alloc_domain;
+	(*mem)->domain = domain;
 	(*mem)->mapped_to_gpu_memory = 0;
 	(*mem)->process_info = avm->process_info;
-	add_kgd_mem_to_kfd_bo_list(*mem, avm->process_info);
+	add_kgd_mem_to_kfd_bo_list(*mem, avm->process_info, user_addr);
+
+	if (user_addr) {
+		ret = init_user_pages(*mem, current->mm, user_addr);
+		if (ret) {
+			mutex_lock(&avm->process_info->lock);
+			list_del(&(*mem)->validate_list.head);
+			mutex_unlock(&avm->process_info->lock);
+			goto allocate_init_user_pages_failed;
+		}
+	}
 
 	if (offset)
 		*offset = amdgpu_bo_mmap_offset(bo);
 
 	return 0;
 
+allocate_init_user_pages_failed:
+	amdgpu_bo_unref(&bo);
+	/* Don't unreserve system mem limit twice */
+	goto err_reserve_system_mem;
 err_bo_create:
 	unreserve_system_mem_limit(adev, size, alloc_domain);
 err_reserve_system_mem:
@@ -1122,12 +1284,24 @@ int amdgpu_amdkfd_gpuvm_free_memory_of_gpu(
 	 * be freed anyway
 	 */
 
+	/* No more MMU notifiers */
+	amdgpu_mn_unregister(mem->bo);
+
 	/* Make sure restore workers don't access the BO any more */
 	bo_list_entry = &mem->validate_list;
 	mutex_lock(&process_info->lock);
 	list_del(&bo_list_entry->head);
 	mutex_unlock(&process_info->lock);
 
+	/* Free user pages if necessary */
+	if (mem->user_pages) {
+		pr_debug("%s: Freeing user_pages array\n", __func__);
+		if (mem->user_pages[0])
+			release_pages(mem->user_pages,
+					mem->bo->tbo.ttm->num_pages);
+		kvfree(mem->user_pages);
+	}
+
 	ret = reserve_bo_and_cond_vms(mem, NULL, BO_VM_ALL, &ctx);
 	if (unlikely(ret))
 		return ret;
@@ -1173,21 +1347,32 @@ int amdgpu_amdkfd_gpuvm_map_memory_to_gpu(
 	struct kfd_bo_va_list *bo_va_entry = NULL;
 	struct kfd_bo_va_list *bo_va_entry_aql = NULL;
 	unsigned long bo_size;
-
-	/* Make sure restore is not running concurrently.
-	 */
-	mutex_lock(&mem->process_info->lock);
-
-	mutex_lock(&mem->lock);
+	bool is_invalid_userptr = false;
 
 	bo = mem->bo;
-
 	if (!bo) {
 		pr_err("Invalid BO when mapping memory to GPU\n");
-		ret = -EINVAL;
-		goto out;
+		return -EINVAL;
+	}
+
+	/* Make sure restore is not running concurrently. Since we
+	 * don't map invalid userptr BOs, we rely on the next restore
+	 * worker to do the mapping
+	 */
+	mutex_lock(&mem->process_info->lock);
+
+	/* Lock mmap-sem. If we find an invalid userptr BO, we can be
+	 * sure that the MMU notifier is no longer running
+	 * concurrently and the queues are actually stopped
+	 */
+	if (amdgpu_ttm_tt_get_usermm(bo->tbo.ttm)) {
+		down_write(&current->mm->mmap_sem);
+		is_invalid_userptr = atomic_read(&mem->invalid);
+		up_write(&current->mm->mmap_sem);
 	}
 
+	mutex_lock(&mem->lock);
+
 	domain = mem->domain;
 	bo_size = bo->tbo.mem.size;
 
@@ -1200,6 +1385,14 @@ int amdgpu_amdkfd_gpuvm_map_memory_to_gpu(
 	if (unlikely(ret))
 		goto out;
 
+	/* Userptr can be marked as "not invalid", but not actually be
+	 * validated yet (still in the system domain). In that case
+	 * the queues are still stopped and we can leave mapping for
+	 * the next restore worker
+	 */
+	if (bo->tbo.mem.mem_type == TTM_PL_SYSTEM)
+		is_invalid_userptr = true;
+
 	if (check_if_add_bo_to_vm(avm, mem)) {
 		ret = add_bo_to_vm(adev, mem, avm, false,
 				&bo_va_entry);
@@ -1217,7 +1410,8 @@ int amdgpu_amdkfd_gpuvm_map_memory_to_gpu(
 			goto add_bo_to_vm_failed;
 	}
 
-	if (mem->mapped_to_gpu_memory == 0) {
+	if (mem->mapped_to_gpu_memory == 0 &&
+	    !amdgpu_ttm_tt_get_usermm(bo->tbo.ttm)) {
 		/* Validate BO only once. The eviction fence gets added to BO
 		 * the first time it is mapped. Validate will wait for all
 		 * background evictions to complete.
@@ -1235,7 +1429,8 @@ int amdgpu_amdkfd_gpuvm_map_memory_to_gpu(
 					entry->va, entry->va + bo_size,
 					entry);
 
-			ret = map_bo_to_gpuvm(adev, entry, ctx.sync);
+			ret = map_bo_to_gpuvm(adev, entry, ctx.sync,
+					      is_invalid_userptr);
 			if (ret) {
 				pr_err("Failed to map radeon bo to gpuvm\n");
 				goto map_bo_to_gpuvm_failed;
@@ -1418,13 +1613,337 @@ bo_reserve_failed:
 	return ret;
 }
 
+/* Evict a userptr BO by stopping the queues if necessary
+ *
+ * Runs in MMU notifier, may be in RECLAIM_FS context. This means it
+ * cannot do any memory allocations, and cannot take any locks that
+ * are held elsewhere while allocating memory. Therefore this is as
+ * simple as possible, using atomic counters.
+ *
+ * It doesn't do anything to the BO itself. The real work happens in
+ * restore, where we get updated page addresses. This function only
+ * ensures that GPU access to the BO is stopped.
+ */
 int amdgpu_amdkfd_evict_userptr(struct kgd_mem *mem,
 				struct mm_struct *mm)
 {
-	/* TODO */
+	struct amdkfd_process_info *process_info = mem->process_info;
+	int invalid, evicted_bos;
+	int r = 0;
+
+	invalid = atomic_inc_return(&mem->invalid);
+	evicted_bos = atomic_inc_return(&process_info->evicted_bos);
+	if (evicted_bos == 1) {
+		/* First eviction, stop the queues */
+		r = kgd2kfd->quiesce_mm(mm);
+		if (r)
+			pr_err("Failed to quiesce KFD\n");
+		schedule_delayed_work(&process_info->restore_userptr_work,
+			msecs_to_jiffies(AMDGPU_USERPTR_RESTORE_DELAY_MS));
+	}
+
+	return r;
+}
+
+/* Update invalid userptr BOs
+ *
+ * Moves invalidated (evicted) userptr BOs from userptr_valid_list to
+ * userptr_inval_list and updates user pages for all BOs that have
+ * been invalidated since their last update.
+ */
+static int update_invalid_user_pages(struct amdkfd_process_info *process_info,
+				     struct mm_struct *mm)
+{
+	struct kgd_mem *mem, *tmp_mem;
+	struct amdgpu_bo *bo;
+	struct ttm_operation_ctx ctx = { false, false };
+	int invalid, ret;
+
+	/* Move all invalidated BOs to the userptr_inval_list and
+	 * release their user pages by migration to the CPU domain
+	 */
+	list_for_each_entry_safe(mem, tmp_mem,
+				 &process_info->userptr_valid_list,
+				 validate_list.head) {
+		if (!atomic_read(&mem->invalid))
+			continue; /* BO is still valid */
+
+		bo = mem->bo;
+
+		if (amdgpu_bo_reserve(bo, true))
+			return -EAGAIN;
+		amdgpu_ttm_placement_from_domain(bo, AMDGPU_GEM_DOMAIN_CPU);
+		ret = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
+		amdgpu_bo_unreserve(bo);
+		if (ret) {
+			pr_err("%s: Failed to invalidate userptr BO\n",
+			       __func__);
+			return -EAGAIN;
+		}
+
+		list_move_tail(&mem->validate_list.head,
+			       &process_info->userptr_inval_list);
+	}
+
+	if (list_empty(&process_info->userptr_inval_list))
+		return 0; /* All evicted userptr BOs were freed */
+
+	/* Go through userptr_inval_list and update any invalid user_pages */
+	list_for_each_entry(mem, &process_info->userptr_inval_list,
+			    validate_list.head) {
+		invalid = atomic_read(&mem->invalid);
+		if (!invalid)
+			/* BO hasn't been invalidated since the last
+			 * revalidation attempt. Keep its BO list.
+			 */
+			continue;
+
+		bo = mem->bo;
+
+		if (!mem->user_pages) {
+			mem->user_pages =
+				kvmalloc_array(bo->tbo.ttm->num_pages,
+						 sizeof(struct page *),
+						 GFP_KERNEL | __GFP_ZERO);
+			if (!mem->user_pages) {
+				pr_err("%s: Failed to allocate pages array\n",
+				       __func__);
+				return -ENOMEM;
+			}
+		} else if (mem->user_pages[0]) {
+			release_pages(mem->user_pages, bo->tbo.ttm->num_pages);
+		}
+
+		/* Get updated user pages */
+		ret = amdgpu_ttm_tt_get_user_pages(bo->tbo.ttm,
+						   mem->user_pages);
+		if (ret) {
+			mem->user_pages[0] = NULL;
+			pr_info("%s: Failed to get user pages: %d\n",
+				__func__, ret);
+			/* Pretend it succeeded. It will fail later
+			 * with a VM fault if the GPU tries to access
+			 * it. Better than hanging indefinitely with
+			 * stalled user mode queues.
+			 */
+		}
+
+		/* Mark the BO as valid unless it was invalidated
+		 * again concurrently
+		 */
+		if (atomic_cmpxchg(&mem->invalid, invalid, 0) != invalid)
+			return -EAGAIN;
+	}
+
 	return 0;
 }
 
+/* Validate invalid userptr BOs
+ *
+ * Validates BOs on the userptr_inval_list, and moves them back to the
+ * userptr_valid_list. Also updates GPUVM page tables with new page
+ * addresses and waits for the page table updates to complete.
+ */
+static int validate_invalid_user_pages(struct amdkfd_process_info *process_info)
+{
+	struct amdgpu_bo_list_entry *pd_bo_list_entries;
+	struct list_head resv_list, duplicates;
+	struct ww_acquire_ctx ticket;
+	struct amdgpu_sync sync;
+
+	struct amdgpu_vm *peer_vm;
+	struct kgd_mem *mem, *tmp_mem;
+	struct amdgpu_bo *bo;
+	struct ttm_operation_ctx ctx = { false, false };
+	int i, ret;
+
+	pd_bo_list_entries = kcalloc(process_info->n_vms,
+				     sizeof(struct amdgpu_bo_list_entry),
+				     GFP_KERNEL);
+	if (!pd_bo_list_entries) {
+		pr_err("%s: Failed to allocate PD BO list entries\n", __func__);
+		return -ENOMEM;
+	}
+
+	INIT_LIST_HEAD(&resv_list);
+	INIT_LIST_HEAD(&duplicates);
+
+	/* Get all the page directory BOs that need to be reserved */
+	i = 0;
+	list_for_each_entry(peer_vm, &process_info->vm_list_head,
+			    vm_list_node)
+		amdgpu_vm_get_pd_bo(peer_vm, &resv_list,
+				    &pd_bo_list_entries[i++]);
+	/* Add the userptr_inval_list entries to resv_list */
+	list_for_each_entry(mem, &process_info->userptr_inval_list,
+			    validate_list.head) {
+		list_add_tail(&mem->resv_list.head, &resv_list);
+		mem->resv_list.bo = mem->validate_list.bo;
+		mem->resv_list.shared = mem->validate_list.shared;
+	}
+
+	/* Reserve all BOs and page tables for validation */
+	ret = ttm_eu_reserve_buffers(&ticket, &resv_list, false, &duplicates);
+	WARN(!list_empty(&duplicates), "Duplicates should be empty");
+	if (ret)
+		goto out;
+
+	amdgpu_sync_create(&sync);
+
+	/* Avoid triggering eviction fences when unmapping invalid
+	 * userptr BOs (waits for all fences, doesn't use
+	 * FENCE_OWNER_VM)
+	 */
+	list_for_each_entry(peer_vm, &process_info->vm_list_head,
+			    vm_list_node)
+		amdgpu_amdkfd_remove_eviction_fence(peer_vm->root.base.bo,
+						process_info->eviction_fence,
+						NULL, NULL);
+
+	ret = process_validate_vms(process_info);
+	if (ret)
+		goto unreserve_out;
+
+	/* Validate BOs and update GPUVM page tables */
+	list_for_each_entry_safe(mem, tmp_mem,
+				 &process_info->userptr_inval_list,
+				 validate_list.head) {
+		struct kfd_bo_va_list *bo_va_entry;
+
+		bo = mem->bo;
+
+		/* Copy pages array and validate the BO if we got user pages */
+		if (mem->user_pages[0]) {
+			amdgpu_ttm_tt_set_user_pages(bo->tbo.ttm,
+						     mem->user_pages);
+			amdgpu_ttm_placement_from_domain(bo, mem->domain);
+			ret = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
+			if (ret) {
+				pr_err("%s: failed to validate BO\n", __func__);
+				goto unreserve_out;
+			}
+		}
+
+		/* Validate succeeded, now the BO owns the pages, free
+		 * our copy of the pointer array. Put this BO back on
+		 * the userptr_valid_list. If we need to revalidate
+		 * it, we need to start from scratch.
+		 */
+		kvfree(mem->user_pages);
+		mem->user_pages = NULL;
+		list_move_tail(&mem->validate_list.head,
+			       &process_info->userptr_valid_list);
+
+		/* Update mapping. If the BO was not validated
+		 * (because we couldn't get user pages), this will
+		 * clear the page table entries, which will result in
+		 * VM faults if the GPU tries to access the invalid
+		 * memory.
+		 */
+		list_for_each_entry(bo_va_entry, &mem->bo_va_list, bo_list) {
+			if (!bo_va_entry->is_mapped)
+				continue;
+
+			ret = update_gpuvm_pte((struct amdgpu_device *)
+					       bo_va_entry->kgd_dev,
+					       bo_va_entry, &sync);
+			if (ret) {
+				pr_err("%s: update PTE failed\n", __func__);
+				/* make sure this gets validated again */
+				atomic_inc(&mem->invalid);
+				goto unreserve_out;
+			}
+		}
+	}
+
+	/* Update page directories */
+	ret = process_update_pds(process_info, &sync);
+
+unreserve_out:
+	list_for_each_entry(peer_vm, &process_info->vm_list_head,
+			    vm_list_node)
+		amdgpu_bo_fence(peer_vm->root.base.bo,
+				&process_info->eviction_fence->base, true);
+	ttm_eu_backoff_reservation(&ticket, &resv_list);
+	amdgpu_sync_wait(&sync, false);
+	amdgpu_sync_free(&sync);
+out:
+	kfree(pd_bo_list_entries);
+
+	return ret;
+}
+
+/* Worker callback to restore evicted userptr BOs
+ *
+ * Tries to update and validate all userptr BOs. If successful and no
+ * concurrent evictions happened, the queues are restarted. Otherwise,
+ * reschedule for another attempt later.
+ */
+static void amdgpu_amdkfd_restore_userptr_worker(struct work_struct *work)
+{
+	struct delayed_work *dwork = to_delayed_work(work);
+	struct amdkfd_process_info *process_info =
+		container_of(dwork, struct amdkfd_process_info,
+			     restore_userptr_work);
+	struct task_struct *usertask;
+	struct mm_struct *mm;
+	int evicted_bos;
+
+	evicted_bos = atomic_read(&process_info->evicted_bos);
+	if (!evicted_bos)
+		return;
+
+	/* Reference task and mm in case of concurrent process termination */
+	usertask = get_pid_task(process_info->pid, PIDTYPE_PID);
+	if (!usertask)
+		return;
+	mm = get_task_mm(usertask);
+	if (!mm) {
+		put_task_struct(usertask);
+		return;
+	}
+
+	mutex_lock(&process_info->lock);
+
+	if (update_invalid_user_pages(process_info, mm))
+		goto unlock_out;
+	/* userptr_inval_list can be empty if all evicted userptr BOs
+	 * have been freed. In that case there is nothing to validate
+	 * and we can just restart the queues.
+	 */
+	if (!list_empty(&process_info->userptr_inval_list)) {
+		if (atomic_read(&process_info->evicted_bos) != evicted_bos)
+			goto unlock_out; /* Concurrent eviction, try again */
+
+		if (validate_invalid_user_pages(process_info))
+			goto unlock_out;
+	}
+	/* Final check for concurrent evicton and atomic update. If
+	 * another eviction happens after successful update, it will
+	 * be a first eviction that calls quiesce_mm. The eviction
+	 * reference counting inside KFD will handle this case.
+	 */
+	if (atomic_cmpxchg(&process_info->evicted_bos, evicted_bos, 0) !=
+	    evicted_bos)
+		goto unlock_out;
+	evicted_bos = 0;
+	if (kgd2kfd->resume_mm(mm)) {
+		pr_err("%s: Failed to resume KFD\n", __func__);
+		/* No recovery from this failure. Probably the CP is
+		 * hanging. No point trying again.
+		 */
+	}
+unlock_out:
+	mutex_unlock(&process_info->lock);
+	mmput(mm);
+	put_task_struct(usertask);
+
+	/* If validation failed, reschedule another attempt */
+	if (evicted_bos)
+		schedule_delayed_work(&process_info->restore_userptr_work,
+			msecs_to_jiffies(AMDGPU_USERPTR_RESTORE_DELAY_MS));
+}
+
 /** amdgpu_amdkfd_gpuvm_restore_process_bos - Restore all BOs for the given
  *   KFD process identified by process_info
  *
-- 
cgit v1.1


From d93ae190e2c95276caceb3642e6d541d93bba705 Mon Sep 17 00:00:00 2001
From: Russell King <rmk+kernel@armlinux.org.uk>
Date: Thu, 17 Nov 2016 23:38:29 +0000
Subject: drm/i2c: tda998x: move mutex/waitqueue/timer/work init early

Move the mutex, waitqueue, timer and detect work initialisation early
in the driver's initialisation, rather than being after we've registered
the CEC device.

Acked-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
---
 drivers/gpu/drm/i2c/tda998x_drv.c | 11 +++++------
 1 file changed, 5 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/i2c/tda998x_drv.c b/drivers/gpu/drm/i2c/tda998x_drv.c
index cd3f087..8340715 100644
--- a/drivers/gpu/drm/i2c/tda998x_drv.c
+++ b/drivers/gpu/drm/i2c/tda998x_drv.c
@@ -1475,7 +1475,11 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 	u32 video;
 	int rev_lo, rev_hi, ret;
 
-	mutex_init(&priv->audio_mutex); /* Protect access from audio thread */
+	mutex_init(&priv->mutex);	/* protect the page access */
+	mutex_init(&priv->audio_mutex); /* protect access from audio thread */
+	init_waitqueue_head(&priv->edid_delay_waitq);
+	timer_setup(&priv->edid_delay_timer, tda998x_edid_delay_done, 0);
+	INIT_WORK(&priv->detect_work, tda998x_detect_work);
 
 	priv->vip_cntrl_0 = VIP_CNTRL_0_SWAP_A(2) | VIP_CNTRL_0_SWAP_B(3);
 	priv->vip_cntrl_1 = VIP_CNTRL_1_SWAP_C(0) | VIP_CNTRL_1_SWAP_D(1);
@@ -1489,11 +1493,6 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 	if (!priv->cec)
 		return -ENODEV;
 
-	mutex_init(&priv->mutex);	/* protect the page access */
-	init_waitqueue_head(&priv->edid_delay_waitq);
-	timer_setup(&priv->edid_delay_timer, tda998x_edid_delay_done, 0);
-	INIT_WORK(&priv->detect_work, tda998x_detect_work);
-
 	/* wake up the device: */
 	cec_write(priv, REG_CEC_ENAMODS,
 			CEC_ENAMODS_EN_RXSENS | CEC_ENAMODS_EN_HDMI);
-- 
cgit v1.1


From 6a765c3fe5497359c11536dfbdcf7526ccb2a33f Mon Sep 17 00:00:00 2001
From: Russell King <rmk+kernel@armlinux.org.uk>
Date: Thu, 17 Nov 2016 23:49:43 +0000
Subject: drm/i2c: tda998x: fix error cleanup paths

If tda998x_get_audio_ports() fails, and we requested the interrupt, we
fail to free the interrupt before returning failure.  Rework the failure
cleanup code and exit paths so that we always clean up properly after an
error, and always propagate the error code.

Acked-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
---
 drivers/gpu/drm/i2c/tda998x_drv.c | 31 ++++++++++++++++++-------------
 1 file changed, 18 insertions(+), 13 deletions(-)

diff --git a/drivers/gpu/drm/i2c/tda998x_drv.c b/drivers/gpu/drm/i2c/tda998x_drv.c
index 8340715..2a99930 100644
--- a/drivers/gpu/drm/i2c/tda998x_drv.c
+++ b/drivers/gpu/drm/i2c/tda998x_drv.c
@@ -1501,10 +1501,15 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 
 	/* read version: */
 	rev_lo = reg_read(priv, REG_VERSION_LSB);
+	if (rev_lo < 0) {
+		dev_err(&client->dev, "failed to read version: %d\n", rev_lo);
+		return rev_lo;
+	}
+
 	rev_hi = reg_read(priv, REG_VERSION_MSB);
-	if (rev_lo < 0 || rev_hi < 0) {
-		ret = rev_lo < 0 ? rev_lo : rev_hi;
-		goto fail;
+	if (rev_hi < 0) {
+		dev_err(&client->dev, "failed to read version: %d\n", rev_hi);
+		return rev_hi;
 	}
 
 	priv->rev = rev_lo | rev_hi << 8;
@@ -1528,7 +1533,7 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 	default:
 		dev_err(&client->dev, "found unsupported device: %04x\n",
 			priv->rev);
-		goto fail;
+		return -ENXIO;
 	}
 
 	/* after reset, enable DDC: */
@@ -1566,7 +1571,7 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 			dev_err(&client->dev,
 				"failed to request IRQ#%u: %d\n",
 				client->irq, ret);
-			goto fail;
+			goto err_irq;
 		}
 
 		/* enable HPD irq */
@@ -1589,19 +1594,19 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 
 	ret = tda998x_get_audio_ports(priv, np);
 	if (ret)
-		goto fail;
+		goto err_audio;
 
 	if (priv->audio_port[0].format != AFMT_UNUSED)
 		tda998x_audio_codec_init(priv, &client->dev);
 
 	return 0;
-fail:
-	/* if encoder_init fails, the encoder slave is never registered,
-	 * so cleanup here:
-	 */
-	if (priv->cec)
-		i2c_unregister_device(priv->cec);
-	return -ENXIO;
+
+err_audio:
+	if (client->irq)
+		free_irq(client->irq, priv);
+err_irq:
+	i2c_unregister_device(priv->cec);
+	return ret;
 }
 
 static void tda998x_encoder_prepare(struct drm_encoder *encoder)
-- 
cgit v1.1


From 101e996b8d321514c45136bef4dd0b1d3c577036 Mon Sep 17 00:00:00 2001
From: Russell King <rmk+kernel@armlinux.org.uk>
Date: Thu, 17 Nov 2016 23:40:26 +0000
Subject: drm/i2c: tda998x: move CEC device initialisation later

We no longer use the CEC client to access the CEC part itself, so we can
move this later in the initialisation sequence.

Acked-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
---
 drivers/gpu/drm/i2c/tda998x_drv.c | 13 ++++++++-----
 1 file changed, 8 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/i2c/tda998x_drv.c b/drivers/gpu/drm/i2c/tda998x_drv.c
index 2a99930..7f2762f 100644
--- a/drivers/gpu/drm/i2c/tda998x_drv.c
+++ b/drivers/gpu/drm/i2c/tda998x_drv.c
@@ -1489,9 +1489,6 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 	priv->cec_addr = 0x34 + (client->addr & 0x03);
 	priv->current_page = 0xff;
 	priv->hdmi = client;
-	priv->cec = i2c_new_dummy(client->adapter, priv->cec_addr);
-	if (!priv->cec)
-		return -ENODEV;
 
 	/* wake up the device: */
 	cec_write(priv, REG_CEC_ENAMODS,
@@ -1578,6 +1575,12 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 		cec_write(priv, REG_CEC_RXSHPDINTENA, CEC_RXSHPDLEV_HPD);
 	}
 
+	priv->cec = i2c_new_dummy(client->adapter, priv->cec_addr);
+	if (!priv->cec) {
+		ret = -ENODEV;
+		goto fail;
+	}
+
 	/* enable EDID read irq: */
 	reg_set(priv, REG_INT_FLAGS_2, INT_FLAGS_2_EDID_BLK_RD);
 
@@ -1594,14 +1597,14 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 
 	ret = tda998x_get_audio_ports(priv, np);
 	if (ret)
-		goto err_audio;
+		goto fail;
 
 	if (priv->audio_port[0].format != AFMT_UNUSED)
 		tda998x_audio_codec_init(priv, &client->dev);
 
 	return 0;
 
-err_audio:
+fail:
 	if (client->irq)
 		free_irq(client->irq, priv);
 err_irq:
-- 
cgit v1.1


From ba8975f15bb93d7f3ebd995a41c2e4b6945bad78 Mon Sep 17 00:00:00 2001
From: Russell King <rmk+kernel@armlinux.org.uk>
Date: Sat, 11 Mar 2017 11:12:22 +0000
Subject: drm/i2c: tda998x: always disable and clear interrupts at probe

Always disable and clear interrupts at probe time to ensure that the
TDA998x is in a sane state.  This ensures that the interrupt line,
which is also the CEC clock calibration signal, is always deasserted.

Acked-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
---
 drivers/gpu/drm/i2c/tda998x_drv.c | 14 +++++++++-----
 1 file changed, 9 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/i2c/tda998x_drv.c b/drivers/gpu/drm/i2c/tda998x_drv.c
index 7f2762f..16e0439 100644
--- a/drivers/gpu/drm/i2c/tda998x_drv.c
+++ b/drivers/gpu/drm/i2c/tda998x_drv.c
@@ -1546,6 +1546,15 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 	cec_write(priv, REG_CEC_FRO_IM_CLK_CTRL,
 			CEC_FRO_IM_CLK_CTRL_GHOST_DIS | CEC_FRO_IM_CLK_CTRL_IMCLK_SEL);
 
+	/* ensure interrupts are disabled */
+	cec_write(priv, REG_CEC_RXSHPDINTENA, 0);
+
+	/* clear pending interrupts */
+	cec_read(priv, REG_CEC_RXSHPDINT);
+	reg_read(priv, REG_INT_FLAGS_0);
+	reg_read(priv, REG_INT_FLAGS_1);
+	reg_read(priv, REG_INT_FLAGS_2);
+
 	/* initialize the optional IRQ */
 	if (client->irq) {
 		unsigned long irq_flags;
@@ -1553,11 +1562,6 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 		/* init read EDID waitqueue and HDP work */
 		init_waitqueue_head(&priv->wq_edid);
 
-		/* clear pending interrupts */
-		reg_read(priv, REG_INT_FLAGS_0);
-		reg_read(priv, REG_INT_FLAGS_1);
-		reg_read(priv, REG_INT_FLAGS_2);
-
 		irq_flags =
 			irqd_get_trigger_type(irq_get_irq_data(client->irq));
 		irq_flags |= IRQF_SHARED | IRQF_ONESHOT;
-- 
cgit v1.1


From e631ad60d2e8c3d358935fa3eb6c45b24af5d48f Mon Sep 17 00:00:00 2001
From: Gabriel Fernandez <gabriel.fernandez@st.com>
Date: Fri, 6 Apr 2018 08:39:28 +0200
Subject: clk: stm32mp1: add missing static

Add missing static for const parent names and clock ops.

Signed-off-by: Gabriel Fernandez <gabriel.fernandez@st.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
---
 drivers/clk/clk-stm32mp1.c | 30 ++++++++++++++----------------
 1 file changed, 14 insertions(+), 16 deletions(-)

diff --git a/drivers/clk/clk-stm32mp1.c b/drivers/clk/clk-stm32mp1.c
index f1d5967..62f172c 100644
--- a/drivers/clk/clk-stm32mp1.c
+++ b/drivers/clk/clk-stm32mp1.c
@@ -216,7 +216,7 @@ static const char * const usart1_src[] = {
 	"pclk5", "pll3_q", "ck_hsi", "ck_csi", "pll4_q", "ck_hse"
 };
 
-const char * const usart234578_src[] = {
+static const char * const usart234578_src[] = {
 	"pclk1", "pll4_q", "ck_hsi", "ck_csi", "ck_hse"
 };
 
@@ -316,10 +316,8 @@ struct stm32_clk_mgate {
 struct clock_config {
 	u32 id;
 	const char *name;
-	union {
-		const char *parent_name;
-		const char * const *parent_names;
-	};
+	const char *parent_name;
+	const char * const *parent_names;
 	int num_parents;
 	unsigned long flags;
 	void *cfg;
@@ -469,7 +467,7 @@ static void mp1_gate_clk_disable(struct clk_hw *hw)
 	}
 }
 
-const struct clk_ops mp1_gate_clk_ops = {
+static const struct clk_ops mp1_gate_clk_ops = {
 	.enable		= mp1_gate_clk_enable,
 	.disable	= mp1_gate_clk_disable,
 	.is_enabled	= clk_gate_is_enabled,
@@ -698,7 +696,7 @@ static void mp1_mgate_clk_disable(struct clk_hw *hw)
 		mp1_gate_clk_disable(hw);
 }
 
-const struct clk_ops mp1_mgate_clk_ops = {
+static const struct clk_ops mp1_mgate_clk_ops = {
 	.enable		= mp1_mgate_clk_enable,
 	.disable	= mp1_mgate_clk_disable,
 	.is_enabled	= clk_gate_is_enabled,
@@ -732,7 +730,7 @@ static int clk_mmux_set_parent(struct clk_hw *hw, u8 index)
 	return 0;
 }
 
-const struct clk_ops clk_mmux_ops = {
+static const struct clk_ops clk_mmux_ops = {
 	.get_parent	= clk_mmux_get_parent,
 	.set_parent	= clk_mmux_set_parent,
 	.determine_rate	= __clk_mux_determine_rate,
@@ -1048,10 +1046,10 @@ struct stm32_pll_cfg {
 	u32 offset;
 };
 
-struct clk_hw *_clk_register_pll(struct device *dev,
-				 struct clk_hw_onecell_data *clk_data,
-				 void __iomem *base, spinlock_t *lock,
-				 const struct clock_config *cfg)
+static struct clk_hw *_clk_register_pll(struct device *dev,
+					struct clk_hw_onecell_data *clk_data,
+					void __iomem *base, spinlock_t *lock,
+					const struct clock_config *cfg)
 {
 	struct stm32_pll_cfg *stm_pll_cfg = cfg->cfg;
 
@@ -1417,7 +1415,7 @@ enum {
 	G_LAST
 };
 
-struct stm32_mgate mp1_mgate[G_LAST];
+static struct stm32_mgate mp1_mgate[G_LAST];
 
 #define _K_GATE(_id, _gate_offset, _gate_bit_idx, _gate_flags,\
 	       _mgate, _ops)\
@@ -1440,7 +1438,7 @@ struct stm32_mgate mp1_mgate[G_LAST];
 	       &mp1_mgate[_id], &mp1_mgate_clk_ops)
 
 /* Peripheral gates */
-struct stm32_gate_cfg per_gate_cfg[G_LAST] = {
+static struct stm32_gate_cfg per_gate_cfg[G_LAST] = {
 	/* Multi gates */
 	K_GATE(G_MDIO,		RCC_APB1ENSETR, 31, 0),
 	K_MGATE(G_DAC12,	RCC_APB1ENSETR, 29, 0),
@@ -1600,7 +1598,7 @@ enum {
 	M_LAST
 };
 
-struct stm32_mmux ker_mux[M_LAST];
+static struct stm32_mmux ker_mux[M_LAST];
 
 #define _K_MUX(_id, _offset, _shift, _width, _mux_flags, _mmux, _ops)\
 	[_id] = {\
@@ -1623,7 +1621,7 @@ struct stm32_mmux ker_mux[M_LAST];
 	_K_MUX(_id, _offset, _shift, _width, _mux_flags,\
 			&ker_mux[_id], &clk_mmux_ops)
 
-const struct stm32_mux_cfg ker_mux_cfg[M_LAST] = {
+static const struct stm32_mux_cfg ker_mux_cfg[M_LAST] = {
 	/* Kernel multi mux */
 	K_MMUX(M_SDMMC12, RCC_SDMMC12CKSELR, 0, 3, 0),
 	K_MMUX(M_SPI23, RCC_SPI2S23CKSELR, 0, 3, 0),
-- 
cgit v1.1


From aa5fc95fd5e63b31cafa7a5a0c02d0585bbc9263 Mon Sep 17 00:00:00 2001
From: Gabriel Fernandez <gabriel.fernandez@st.com>
Date: Fri, 6 Apr 2018 08:39:29 +0200
Subject: clk: stm32mp1: remove unused dfsdm_src[] const

This patch remove unused constant.

Signed-off-by: Gabriel Fernandez <gabriel.fernandez@st.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
---
 drivers/clk/clk-stm32mp1.c | 4 ----
 1 file changed, 4 deletions(-)

diff --git a/drivers/clk/clk-stm32mp1.c b/drivers/clk/clk-stm32mp1.c
index 62f172c..9a67b1c 100644
--- a/drivers/clk/clk-stm32mp1.c
+++ b/drivers/clk/clk-stm32mp1.c
@@ -224,10 +224,6 @@ static const char * const usart6_src[] = {
 	"pclk2", "pll4_q", "ck_hsi", "ck_csi", "ck_hse"
 };
 
-static const char * const dfsdm_src[] = {
-	"pclk2", "ck_mcu"
-};
-
 static const char * const fdcan_src[] = {
 	"ck_hse", "pll3_q", "pll4_q"
 };
-- 
cgit v1.1


From 4cd2136031f246ae099f18d4568290c64187ddf8 Mon Sep 17 00:00:00 2001
From: Gabriel Fernandez <gabriel.fernandez@st.com>
Date: Fri, 6 Apr 2018 08:39:30 +0200
Subject: clk: stm32mp1: fix SAI3 & SAI4 clocks

fix bad copy / paste.
SAI3 & SAI4 used gate of SAI2 instead SAI3 & SAI4

Signed-off-by: Gabriel Fernandez <gabriel.fernandez@st.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
---
 drivers/clk/clk-stm32mp1.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/clk/clk-stm32mp1.c b/drivers/clk/clk-stm32mp1.c
index 9a67b1c..eefed49 100644
--- a/drivers/clk/clk-stm32mp1.c
+++ b/drivers/clk/clk-stm32mp1.c
@@ -1942,8 +1942,8 @@ static const struct clock_config stm32mp1_clock_cfg[] = {
 	KCLK(FDCAN_K, "fdcan_k", fdcan_src, 0, G_FDCAN, M_FDCAN),
 	KCLK(SAI1_K, "sai1_k", sai_src, 0, G_SAI1, M_SAI1),
 	KCLK(SAI2_K, "sai2_k", sai2_src, 0, G_SAI2, M_SAI2),
-	KCLK(SAI3_K, "sai3_k", sai_src, 0, G_SAI2, M_SAI3),
-	KCLK(SAI4_K, "sai4_k", sai_src, 0, G_SAI2, M_SAI4),
+	KCLK(SAI3_K, "sai3_k", sai_src, 0, G_SAI3, M_SAI3),
+	KCLK(SAI4_K, "sai4_k", sai_src, 0, G_SAI4, M_SAI4),
 	KCLK(ADC12_K, "adc12_k", adc12_src, 0, G_ADC12, M_ADC12),
 	KCLK(DSI_K, "dsi_k", dsi_src, 0, G_DSI, M_DSI),
 	KCLK(ADFSDM_K, "adfsdm_k", sai_src, 0, G_ADFSDM, M_SAI1),
-- 
cgit v1.1


From a1bf646f719f14464be93e163cfa7d003c472e6e Mon Sep 17 00:00:00 2001
From: Gabriel Fernandez <gabriel.fernandez@st.com>
Date: Fri, 6 Apr 2018 08:39:31 +0200
Subject: clk: stm32mp1: add missing tzc2 clock

This patch adds tzc2 clock and rename tzc clock into tzc1

Signed-off-by: Gabriel Fernandez <gabriel.fernandez@st.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
---
 drivers/clk/clk-stm32mp1.c                | 9 ++++++---
 include/dt-bindings/clock/stm32mp1-clks.h | 3 ++-
 2 files changed, 8 insertions(+), 4 deletions(-)

diff --git a/drivers/clk/clk-stm32mp1.c b/drivers/clk/clk-stm32mp1.c
index eefed49..e32fc23 100644
--- a/drivers/clk/clk-stm32mp1.c
+++ b/drivers/clk/clk-stm32mp1.c
@@ -1399,7 +1399,8 @@ enum {
 	G_USBH,
 	G_ETHSTP,
 	G_RTCAPB,
-	G_TZC,
+	G_TZC1,
+	G_TZC2,
 	G_TZPC,
 	G_IWDG1,
 	G_BSEC,
@@ -1500,7 +1501,8 @@ static struct stm32_gate_cfg per_gate_cfg[G_LAST] = {
 	K_GATE(G_BSEC,		RCC_APB5ENSETR, 16, 0),
 	K_GATE(G_IWDG1,		RCC_APB5ENSETR, 15, 0),
 	K_GATE(G_TZPC,		RCC_APB5ENSETR, 13, 0),
-	K_GATE(G_TZC,		RCC_APB5ENSETR, 12, 0),
+	K_GATE(G_TZC2,		RCC_APB5ENSETR, 12, 0),
+	K_GATE(G_TZC1,		RCC_APB5ENSETR, 11, 0),
 	K_GATE(G_RTCAPB,	RCC_APB5ENSETR, 8, 0),
 	K_MGATE(G_USART1,	RCC_APB5ENSETR, 4, 0),
 	K_MGATE(G_I2C6,		RCC_APB5ENSETR, 3, 0),
@@ -1854,7 +1856,8 @@ static const struct clock_config stm32mp1_clock_cfg[] = {
 	PCLK(USART1, "usart1", "pclk5", 0, G_USART1),
 	PCLK(RTCAPB, "rtcapb", "pclk5", CLK_IGNORE_UNUSED |
 	     CLK_IS_CRITICAL, G_RTCAPB),
-	PCLK(TZC, "tzc", "pclk5", CLK_IGNORE_UNUSED, G_TZC),
+	PCLK(TZC1, "tzc1", "ck_axi", CLK_IGNORE_UNUSED, G_TZC1),
+	PCLK(TZC2, "tzc2", "ck_axi", CLK_IGNORE_UNUSED, G_TZC2),
 	PCLK(TZPC, "tzpc", "pclk5", CLK_IGNORE_UNUSED, G_TZPC),
 	PCLK(IWDG1, "iwdg1", "pclk5", 0, G_IWDG1),
 	PCLK(BSEC, "bsec", "pclk5", CLK_IGNORE_UNUSED, G_BSEC),
diff --git a/include/dt-bindings/clock/stm32mp1-clks.h b/include/dt-bindings/clock/stm32mp1-clks.h
index 86e3ec6..6c807fd 100644
--- a/include/dt-bindings/clock/stm32mp1-clks.h
+++ b/include/dt-bindings/clock/stm32mp1-clks.h
@@ -76,7 +76,7 @@
 #define I2C6		63
 #define USART1		64
 #define RTCAPB		65
-#define TZC		66
+#define TZC1		66
 #define TZPC		67
 #define IWDG1		68
 #define BSEC		69
@@ -123,6 +123,7 @@
 #define CRC1		110
 #define USBH		111
 #define ETHSTP		112
+#define TZC2		113
 
 /* Kernel clocks */
 #define SDMMC1_K	118
-- 
cgit v1.1


From 1742aed6e539851ae7cfb12685182aa9044f3b0c Mon Sep 17 00:00:00 2001
From: Gabriel Fernandez <gabriel.fernandez@st.com>
Date: Fri, 6 Apr 2018 08:39:32 +0200
Subject: clk: stm32mp1: set stgen_k clock as critical

stgen_k should be declared as critical to avoid blocking console
when ck_hsi is not used.

Signed-off-by: Gabriel Fernandez <gabriel.fernandez@st.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
---
 drivers/clk/clk-stm32mp1.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/clk/clk-stm32mp1.c b/drivers/clk/clk-stm32mp1.c
index e32fc23..b7b5361 100644
--- a/drivers/clk/clk-stm32mp1.c
+++ b/drivers/clk/clk-stm32mp1.c
@@ -1913,8 +1913,7 @@ static const struct clock_config stm32mp1_clock_cfg[] = {
 	KCLK(RNG1_K, "rng1_k", rng_src, 0, G_RNG1, M_RNG1),
 	KCLK(RNG2_K, "rng2_k", rng_src, 0, G_RNG2, M_RNG2),
 	KCLK(USBPHY_K, "usbphy_k", usbphy_src, 0, G_USBPHY, M_USBPHY),
-	KCLK(STGEN_K, "stgen_k",  stgen_src, CLK_IGNORE_UNUSED,
-	     G_STGEN, M_STGEN),
+	KCLK(STGEN_K, "stgen_k", stgen_src, CLK_IS_CRITICAL, G_STGEN, M_STGEN),
 	KCLK(SPDIF_K, "spdif_k", spdif_src, 0, G_SPDIF, M_SPDIF),
 	KCLK(SPI1_K, "spi1_k", spi123_src, 0, G_SPI1, M_SPI1),
 	KCLK(SPI2_K, "spi2_k", spi123_src, 0, G_SPI2, M_SPI23),
-- 
cgit v1.1


From ccf719b8845d6ca9f24a8708bb39e0df008f4e3f Mon Sep 17 00:00:00 2001
From: Gabriel Fernandez <gabriel.fernandez@st.com>
Date: Fri, 6 Apr 2018 08:39:33 +0200
Subject: clk: stm32mp1: remove ck_apb_dbg clock

It's recommended to use only clk_sys_dbg clock instead to activate
debug IP.

Signed-off-by: Gabriel Fernandez <gabriel.fernandez@st.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
---
 drivers/clk/clk-stm32mp1.c                | 4 ----
 include/dt-bindings/clock/stm32mp1-clks.h | 1 -
 2 files changed, 5 deletions(-)

diff --git a/drivers/clk/clk-stm32mp1.c b/drivers/clk/clk-stm32mp1.c
index b7b5361..edd3cf4 100644
--- a/drivers/clk/clk-stm32mp1.c
+++ b/drivers/clk/clk-stm32mp1.c
@@ -1988,10 +1988,6 @@ static const struct clock_config stm32mp1_clock_cfg[] = {
 		  _DIV(RCC_MCO2CFGR, 4, 4, 0, NULL)),
 
 	/* Debug clocks */
-	FIXED_FACTOR(NO_ID, "ck_axi_div2", "ck_axi", 0, 1, 2),
-
-	GATE(DBG, "ck_apb_dbg", "ck_axi_div2", 0, RCC_DBGCFGR, 8, 0),
-
 	GATE(CK_DBG, "ck_sys_dbg", "ck_axi", 0, RCC_DBGCFGR, 8, 0),
 
 	COMPOSITE(CK_TRACE, "ck_trace", ck_trace_src, CLK_OPS_PARENT_ENABLE,
diff --git a/include/dt-bindings/clock/stm32mp1-clks.h b/include/dt-bindings/clock/stm32mp1-clks.h
index 6c807fd..90ec780 100644
--- a/include/dt-bindings/clock/stm32mp1-clks.h
+++ b/include/dt-bindings/clock/stm32mp1-clks.h
@@ -229,7 +229,6 @@
 #define CK_MCO2		212
 
 /* TRACE & DEBUG clocks */
-#define DBG		213
 #define CK_DBG		214
 #define CK_TRACE	215
 
-- 
cgit v1.1


From 839c42273617787318da7baf6151d553108f5e17 Mon Sep 17 00:00:00 2001
From: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Date: Fri, 6 Apr 2018 15:36:11 -0700
Subject: Input: synaptics-rmi4 - fix an unchecked out of memory error path

When extending the rmi_spi buffers, we must check that no out of memory
error occurs, otherwise we may access data above the currently allocated
memory.

Propagate the error code returned by 'rmi_spi_manage_pools()' instead.

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Reviewed-by: Andrew Duggan <aduggan@synaptics.com>
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
---
 drivers/input/rmi4/rmi_spi.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/drivers/input/rmi4/rmi_spi.c b/drivers/input/rmi4/rmi_spi.c
index 76edbf2..082defc 100644
--- a/drivers/input/rmi4/rmi_spi.c
+++ b/drivers/input/rmi4/rmi_spi.c
@@ -147,8 +147,11 @@ static int rmi_spi_xfer(struct rmi_spi_xport *rmi_spi,
 	if (len > RMI_SPI_XFER_SIZE_LIMIT)
 		return -EINVAL;
 
-	if (rmi_spi->xfer_buf_size < len)
-		rmi_spi_manage_pools(rmi_spi, len);
+	if (rmi_spi->xfer_buf_size < len) {
+		ret = rmi_spi_manage_pools(rmi_spi, len);
+		if (ret < 0)
+			return ret;
+	}
 
 	if (addr == 0)
 		/*
-- 
cgit v1.1


From 5e7086eecc32b95288bc76f2a22aadeb368e25ed Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Tue, 10 Apr 2018 17:32:58 -0400
Subject: drm/amdgpu: Remove unused interface from kfd2kgd interface

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c | 10 ----------
 drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c | 10 ----------
 drivers/gpu/drm/amd/include/kgd_kfd_interface.h   |  5 -----
 3 files changed, 25 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c
index ea54e53..0ff36d4 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c
@@ -98,8 +98,6 @@ static void kgd_program_sh_mem_settings(struct kgd_dev *kgd, uint32_t vmid,
 static int kgd_set_pasid_vmid_mapping(struct kgd_dev *kgd, unsigned int pasid,
 					unsigned int vmid);
 
-static int kgd_init_pipeline(struct kgd_dev *kgd, uint32_t pipe_id,
-				uint32_t hpd_size, uint64_t hpd_gpu_addr);
 static int kgd_init_interrupts(struct kgd_dev *kgd, uint32_t pipe_id);
 static int kgd_hqd_load(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id,
 			uint32_t queue_id, uint32_t __user *wptr,
@@ -183,7 +181,6 @@ static const struct kfd2kgd_calls kfd2kgd = {
 	.free_pasid = amdgpu_pasid_free,
 	.program_sh_mem_settings = kgd_program_sh_mem_settings,
 	.set_pasid_vmid_mapping = kgd_set_pasid_vmid_mapping,
-	.init_pipeline = kgd_init_pipeline,
 	.init_interrupts = kgd_init_interrupts,
 	.hqd_load = kgd_hqd_load,
 	.hqd_sdma_load = kgd_hqd_sdma_load,
@@ -309,13 +306,6 @@ static int kgd_set_pasid_vmid_mapping(struct kgd_dev *kgd, unsigned int pasid,
 	return 0;
 }
 
-static int kgd_init_pipeline(struct kgd_dev *kgd, uint32_t pipe_id,
-				uint32_t hpd_size, uint64_t hpd_gpu_addr)
-{
-	/* amdgpu owns the per-pipe state */
-	return 0;
-}
-
 static int kgd_init_interrupts(struct kgd_dev *kgd, uint32_t pipe_id)
 {
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c
index 89264c9..6ef9762 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c
@@ -57,8 +57,6 @@ static void kgd_program_sh_mem_settings(struct kgd_dev *kgd, uint32_t vmid,
 		uint32_t sh_mem_bases);
 static int kgd_set_pasid_vmid_mapping(struct kgd_dev *kgd, unsigned int pasid,
 		unsigned int vmid);
-static int kgd_init_pipeline(struct kgd_dev *kgd, uint32_t pipe_id,
-		uint32_t hpd_size, uint64_t hpd_gpu_addr);
 static int kgd_init_interrupts(struct kgd_dev *kgd, uint32_t pipe_id);
 static int kgd_hqd_load(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id,
 			uint32_t queue_id, uint32_t __user *wptr,
@@ -141,7 +139,6 @@ static const struct kfd2kgd_calls kfd2kgd = {
 	.free_pasid = amdgpu_pasid_free,
 	.program_sh_mem_settings = kgd_program_sh_mem_settings,
 	.set_pasid_vmid_mapping = kgd_set_pasid_vmid_mapping,
-	.init_pipeline = kgd_init_pipeline,
 	.init_interrupts = kgd_init_interrupts,
 	.hqd_load = kgd_hqd_load,
 	.hqd_sdma_load = kgd_hqd_sdma_load,
@@ -270,13 +267,6 @@ static int kgd_set_pasid_vmid_mapping(struct kgd_dev *kgd, unsigned int pasid,
 	return 0;
 }
 
-static int kgd_init_pipeline(struct kgd_dev *kgd, uint32_t pipe_id,
-				uint32_t hpd_size, uint64_t hpd_gpu_addr)
-{
-	/* amdgpu owns the per-pipe state */
-	return 0;
-}
-
 static int kgd_init_interrupts(struct kgd_dev *kgd, uint32_t pipe_id)
 {
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
diff --git a/drivers/gpu/drm/amd/include/kgd_kfd_interface.h b/drivers/gpu/drm/amd/include/kgd_kfd_interface.h
index 286cfe7..7cf3506 100644
--- a/drivers/gpu/drm/amd/include/kgd_kfd_interface.h
+++ b/drivers/gpu/drm/amd/include/kgd_kfd_interface.h
@@ -173,8 +173,6 @@ struct tile_config {
  * @set_pasid_vmid_mapping: Exposes pasid/vmid pair to the H/W for no cp
  * scheduling mode. Only used for no cp scheduling mode.
  *
- * @init_pipeline: Initialized the compute pipelines.
- *
  * @hqd_load: Loads the mqd structure to a H/W hqd slot. used only for no cp
  * sceduling mode.
  *
@@ -274,9 +272,6 @@ struct kfd2kgd_calls {
 	int (*set_pasid_vmid_mapping)(struct kgd_dev *kgd, unsigned int pasid,
 					unsigned int vmid);
 
-	int (*init_pipeline)(struct kgd_dev *kgd, uint32_t pipe_id,
-				uint32_t hpd_size, uint64_t hpd_gpu_addr);
-
 	int (*init_interrupts)(struct kgd_dev *kgd, uint32_t pipe_id);
 
 	int (*hqd_load)(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id,
-- 
cgit v1.1


From cf05fb8b144dae55d094b0fa7991e985a9b4561e Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Tue, 10 Apr 2018 17:32:59 -0400
Subject: drm/amd: Update GFXv9 SDMA MQD structure

This matches what the HWS firmware expects on GFXv9 chips.

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 MAINTAINERS                              |  1 +
 drivers/gpu/drm/amd/include/v9_structs.h | 48 ++++++++++++++++----------------
 2 files changed, 25 insertions(+), 24 deletions(-)

diff --git a/MAINTAINERS b/MAINTAINERS
index 92be777..dc929dc 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -772,6 +772,7 @@ F:	drivers/gpu/drm/amd/amdkfd/
 F:	drivers/gpu/drm/amd/include/cik_structs.h
 F:	drivers/gpu/drm/amd/include/kgd_kfd_interface.h
 F:	drivers/gpu/drm/amd/include/vi_structs.h
+F:	drivers/gpu/drm/amd/include/v9_structs.h
 F:	include/uapi/linux/kfd_ioctl.h
 
 AMD SEATTLE DEVICE TREE SUPPORT
diff --git a/drivers/gpu/drm/amd/include/v9_structs.h b/drivers/gpu/drm/amd/include/v9_structs.h
index 2fb25ab..ceaf493 100644
--- a/drivers/gpu/drm/amd/include/v9_structs.h
+++ b/drivers/gpu/drm/amd/include/v9_structs.h
@@ -29,10 +29,10 @@ struct v9_sdma_mqd {
 	uint32_t sdmax_rlcx_rb_base;
 	uint32_t sdmax_rlcx_rb_base_hi;
 	uint32_t sdmax_rlcx_rb_rptr;
+	uint32_t sdmax_rlcx_rb_rptr_hi;
 	uint32_t sdmax_rlcx_rb_wptr;
+	uint32_t sdmax_rlcx_rb_wptr_hi;
 	uint32_t sdmax_rlcx_rb_wptr_poll_cntl;
-	uint32_t sdmax_rlcx_rb_wptr_poll_addr_hi;
-	uint32_t sdmax_rlcx_rb_wptr_poll_addr_lo;
 	uint32_t sdmax_rlcx_rb_rptr_addr_hi;
 	uint32_t sdmax_rlcx_rb_rptr_addr_lo;
 	uint32_t sdmax_rlcx_ib_cntl;
@@ -44,29 +44,29 @@ struct v9_sdma_mqd {
 	uint32_t sdmax_rlcx_skip_cntl;
 	uint32_t sdmax_rlcx_context_status;
 	uint32_t sdmax_rlcx_doorbell;
-	uint32_t sdmax_rlcx_virtual_addr;
-	uint32_t sdmax_rlcx_ape1_cntl;
+	uint32_t sdmax_rlcx_status;
 	uint32_t sdmax_rlcx_doorbell_log;
-	uint32_t reserved_22;
-	uint32_t reserved_23;
-	uint32_t reserved_24;
-	uint32_t reserved_25;
-	uint32_t reserved_26;
-	uint32_t reserved_27;
-	uint32_t reserved_28;
-	uint32_t reserved_29;
-	uint32_t reserved_30;
-	uint32_t reserved_31;
-	uint32_t reserved_32;
-	uint32_t reserved_33;
-	uint32_t reserved_34;
-	uint32_t reserved_35;
-	uint32_t reserved_36;
-	uint32_t reserved_37;
-	uint32_t reserved_38;
-	uint32_t reserved_39;
-	uint32_t reserved_40;
-	uint32_t reserved_41;
+	uint32_t sdmax_rlcx_watermark;
+	uint32_t sdmax_rlcx_doorbell_offset;
+	uint32_t sdmax_rlcx_csa_addr_lo;
+	uint32_t sdmax_rlcx_csa_addr_hi;
+	uint32_t sdmax_rlcx_ib_sub_remain;
+	uint32_t sdmax_rlcx_preempt;
+	uint32_t sdmax_rlcx_dummy_reg;
+	uint32_t sdmax_rlcx_rb_wptr_poll_addr_hi;
+	uint32_t sdmax_rlcx_rb_wptr_poll_addr_lo;
+	uint32_t sdmax_rlcx_rb_aql_cntl;
+	uint32_t sdmax_rlcx_minor_ptr_update;
+	uint32_t sdmax_rlcx_midcmd_data0;
+	uint32_t sdmax_rlcx_midcmd_data1;
+	uint32_t sdmax_rlcx_midcmd_data2;
+	uint32_t sdmax_rlcx_midcmd_data3;
+	uint32_t sdmax_rlcx_midcmd_data4;
+	uint32_t sdmax_rlcx_midcmd_data5;
+	uint32_t sdmax_rlcx_midcmd_data6;
+	uint32_t sdmax_rlcx_midcmd_data7;
+	uint32_t sdmax_rlcx_midcmd_data8;
+	uint32_t sdmax_rlcx_midcmd_cntl;
 	uint32_t reserved_42;
 	uint32_t reserved_43;
 	uint32_t reserved_44;
-- 
cgit v1.1


From ab88bded7522dafc1a6beb251092365519a01c4e Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Tue, 10 Apr 2018 17:33:00 -0400
Subject: drm/amdgpu: Add GFXv9 TLB invalidation packet definition

Signed-off-by: Shaoyun Liu <Shaoyun.Liu@amd.com>
Signed-off-by: Jay Cornwall <Jay.Cornwall@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdgpu/soc15d.h | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/soc15d.h b/drivers/gpu/drm/amd/amdgpu/soc15d.h
index 7f408f8..f22f7a8 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15d.h
+++ b/drivers/gpu/drm/amd/amdgpu/soc15d.h
@@ -268,6 +268,11 @@
 			 * x=1: tmz_end
 			 */
 
+#define	PACKET3_INVALIDATE_TLBS				0x98
+#              define PACKET3_INVALIDATE_TLBS_DST_SEL(x)     ((x) << 0)
+#              define PACKET3_INVALIDATE_TLBS_ALL_HUB(x)     ((x) << 4)
+#              define PACKET3_INVALIDATE_TLBS_PASID(x)       ((x) << 5)
+#              define PACKET3_INVALIDATE_TLBS_FLUSH_TYPE(x)  ((x) << 29)
 #define PACKET3_SET_RESOURCES				0xA0
 /* 1. header
  * 2. CONTROL
-- 
cgit v1.1


From d5a114a6c5f7fa41da338e0134fccf3f25723fbd Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Tue, 10 Apr 2018 17:33:01 -0400
Subject: drm/amdgpu: Add GFXv9 kfd2kgd interface functions

Signed-off-by: John Bridgman <john.bridgman@amd.com>
Signed-off-by: Shaoyun Liu <Shaoyun.Liu@amd.com>
Signed-off-by: Jay Cornwall <Jay.Cornwall@amd.com>
Signed-off-by: Yong Zhao <yong.zhao@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 MAINTAINERS                                       |    1 +
 drivers/gpu/drm/amd/amdgpu/Makefile               |    3 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c        |    4 +
 drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h        |    1 +
 drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c | 1043 +++++++++++++++++++++
 drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c             |    1 +
 6 files changed, 1052 insertions(+), 1 deletion(-)
 create mode 100644 drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c

diff --git a/MAINTAINERS b/MAINTAINERS
index dc929dc..051f2fe 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -766,6 +766,7 @@ F:	drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
 F:	drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
 F:	drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c
 F:	drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c
+F:	drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c
 F:	drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_fence.c
 F:	drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
 F:	drivers/gpu/drm/amd/amdkfd/
diff --git a/drivers/gpu/drm/amd/amdgpu/Makefile b/drivers/gpu/drm/amd/amdgpu/Makefile
index 2ca2b51..f300202 100644
--- a/drivers/gpu/drm/amd/amdgpu/Makefile
+++ b/drivers/gpu/drm/amd/amdgpu/Makefile
@@ -130,7 +130,8 @@ amdgpu-y += \
 	 amdgpu_amdkfd.o \
 	 amdgpu_amdkfd_fence.o \
 	 amdgpu_amdkfd_gpuvm.o \
-	 amdgpu_amdkfd_gfx_v8.o
+	 amdgpu_amdkfd_gfx_v8.o \
+	 amdgpu_amdkfd_gfx_v9.o
 
 # add cgs
 amdgpu-y += amdgpu_cgs.o
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
index 4d36203..fcd10db 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
@@ -92,6 +92,10 @@ void amdgpu_amdkfd_device_probe(struct amdgpu_device *adev)
 	case CHIP_POLARIS11:
 		kfd2kgd = amdgpu_amdkfd_gfx_8_0_get_functions();
 		break;
+	case CHIP_VEGA10:
+	case CHIP_RAVEN:
+		kfd2kgd = amdgpu_amdkfd_gfx_9_0_get_functions();
+		break;
 	default:
 		dev_dbg(adev->dev, "kfd not supported on this ASIC\n");
 		return;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
index c3024b1..12367a9 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
@@ -122,6 +122,7 @@ int amdgpu_amdkfd_submit_ib(struct kgd_dev *kgd, enum kgd_engine_type engine,
 
 struct kfd2kgd_calls *amdgpu_amdkfd_gfx_7_get_functions(void);
 struct kfd2kgd_calls *amdgpu_amdkfd_gfx_8_0_get_functions(void);
+struct kfd2kgd_calls *amdgpu_amdkfd_gfx_9_0_get_functions(void);
 
 bool amdgpu_amdkfd_is_kfd_vmid(struct amdgpu_device *adev, u32 vmid);
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c
new file mode 100644
index 0000000..8f37991
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c
@@ -0,0 +1,1043 @@
+/*
+ * Copyright 2014-2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#define pr_fmt(fmt) "kfd2kgd: " fmt
+
+#include <linux/module.h>
+#include <linux/fdtable.h>
+#include <linux/uaccess.h>
+#include <linux/firmware.h>
+#include <drm/drmP.h>
+#include "amdgpu.h"
+#include "amdgpu_amdkfd.h"
+#include "amdgpu_ucode.h"
+#include "soc15_hw_ip.h"
+#include "gc/gc_9_0_offset.h"
+#include "gc/gc_9_0_sh_mask.h"
+#include "vega10_enum.h"
+#include "sdma0/sdma0_4_0_offset.h"
+#include "sdma0/sdma0_4_0_sh_mask.h"
+#include "sdma1/sdma1_4_0_offset.h"
+#include "sdma1/sdma1_4_0_sh_mask.h"
+#include "athub/athub_1_0_offset.h"
+#include "athub/athub_1_0_sh_mask.h"
+#include "oss/osssys_4_0_offset.h"
+#include "oss/osssys_4_0_sh_mask.h"
+#include "soc15_common.h"
+#include "v9_structs.h"
+#include "soc15.h"
+#include "soc15d.h"
+
+/* HACK: MMHUB and GC both have VM-related register with the same
+ * names but different offsets. Define the MMHUB register we need here
+ * with a prefix. A proper solution would be to move the functions
+ * programming these registers into gfx_v9_0.c and mmhub_v1_0.c
+ * respectively.
+ */
+#define mmMMHUB_VM_INVALIDATE_ENG16_REQ				0x06f3
+#define mmMMHUB_VM_INVALIDATE_ENG16_REQ_BASE_IDX		0
+
+#define mmMMHUB_VM_INVALIDATE_ENG16_ACK				0x0705
+#define mmMMHUB_VM_INVALIDATE_ENG16_ACK_BASE_IDX		0
+
+#define mmMMHUB_VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32		0x072b
+#define mmMMHUB_VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32_BASE_IDX	0
+#define mmMMHUB_VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32		0x072c
+#define mmMMHUB_VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32_BASE_IDX	0
+
+#define mmMMHUB_VM_CONTEXT0_PAGE_TABLE_START_ADDR_LO32		0x074b
+#define mmMMHUB_VM_CONTEXT0_PAGE_TABLE_START_ADDR_LO32_BASE_IDX	0
+#define mmMMHUB_VM_CONTEXT0_PAGE_TABLE_START_ADDR_HI32		0x074c
+#define mmMMHUB_VM_CONTEXT0_PAGE_TABLE_START_ADDR_HI32_BASE_IDX	0
+
+#define mmMMHUB_VM_CONTEXT0_PAGE_TABLE_END_ADDR_LO32		0x076b
+#define mmMMHUB_VM_CONTEXT0_PAGE_TABLE_END_ADDR_LO32_BASE_IDX	0
+#define mmMMHUB_VM_CONTEXT0_PAGE_TABLE_END_ADDR_HI32		0x076c
+#define mmMMHUB_VM_CONTEXT0_PAGE_TABLE_END_ADDR_HI32_BASE_IDX	0
+
+#define mmMMHUB_VM_INVALIDATE_ENG16_ADDR_RANGE_LO32		0x0727
+#define mmMMHUB_VM_INVALIDATE_ENG16_ADDR_RANGE_LO32_BASE_IDX	0
+#define mmMMHUB_VM_INVALIDATE_ENG16_ADDR_RANGE_HI32		0x0728
+#define mmMMHUB_VM_INVALIDATE_ENG16_ADDR_RANGE_HI32_BASE_IDX	0
+
+#define V9_PIPE_PER_MEC		(4)
+#define V9_QUEUES_PER_PIPE_MEC	(8)
+
+enum hqd_dequeue_request_type {
+	NO_ACTION = 0,
+	DRAIN_PIPE,
+	RESET_WAVES
+};
+
+/*
+ * Register access functions
+ */
+
+static void kgd_program_sh_mem_settings(struct kgd_dev *kgd, uint32_t vmid,
+		uint32_t sh_mem_config,
+		uint32_t sh_mem_ape1_base, uint32_t sh_mem_ape1_limit,
+		uint32_t sh_mem_bases);
+static int kgd_set_pasid_vmid_mapping(struct kgd_dev *kgd, unsigned int pasid,
+		unsigned int vmid);
+static int kgd_init_interrupts(struct kgd_dev *kgd, uint32_t pipe_id);
+static int kgd_hqd_load(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id,
+			uint32_t queue_id, uint32_t __user *wptr,
+			uint32_t wptr_shift, uint32_t wptr_mask,
+			struct mm_struct *mm);
+static int kgd_hqd_dump(struct kgd_dev *kgd,
+			uint32_t pipe_id, uint32_t queue_id,
+			uint32_t (**dump)[2], uint32_t *n_regs);
+static int kgd_hqd_sdma_load(struct kgd_dev *kgd, void *mqd,
+			     uint32_t __user *wptr, struct mm_struct *mm);
+static int kgd_hqd_sdma_dump(struct kgd_dev *kgd,
+			     uint32_t engine_id, uint32_t queue_id,
+			     uint32_t (**dump)[2], uint32_t *n_regs);
+static bool kgd_hqd_is_occupied(struct kgd_dev *kgd, uint64_t queue_address,
+		uint32_t pipe_id, uint32_t queue_id);
+static bool kgd_hqd_sdma_is_occupied(struct kgd_dev *kgd, void *mqd);
+static int kgd_hqd_destroy(struct kgd_dev *kgd, void *mqd,
+				enum kfd_preempt_type reset_type,
+				unsigned int utimeout, uint32_t pipe_id,
+				uint32_t queue_id);
+static int kgd_hqd_sdma_destroy(struct kgd_dev *kgd, void *mqd,
+				unsigned int utimeout);
+static int kgd_address_watch_disable(struct kgd_dev *kgd);
+static int kgd_address_watch_execute(struct kgd_dev *kgd,
+					unsigned int watch_point_id,
+					uint32_t cntl_val,
+					uint32_t addr_hi,
+					uint32_t addr_lo);
+static int kgd_wave_control_execute(struct kgd_dev *kgd,
+					uint32_t gfx_index_val,
+					uint32_t sq_cmd);
+static uint32_t kgd_address_watch_get_offset(struct kgd_dev *kgd,
+					unsigned int watch_point_id,
+					unsigned int reg_offset);
+
+static bool get_atc_vmid_pasid_mapping_valid(struct kgd_dev *kgd,
+		uint8_t vmid);
+static uint16_t get_atc_vmid_pasid_mapping_pasid(struct kgd_dev *kgd,
+		uint8_t vmid);
+static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid,
+		uint32_t page_table_base);
+static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type);
+static void set_scratch_backing_va(struct kgd_dev *kgd,
+					uint64_t va, uint32_t vmid);
+static int invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid);
+static int invalidate_tlbs_vmid(struct kgd_dev *kgd, uint16_t vmid);
+
+/* Because of REG_GET_FIELD() being used, we put this function in the
+ * asic specific file.
+ */
+static int amdgpu_amdkfd_get_tile_config(struct kgd_dev *kgd,
+		struct tile_config *config)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *)kgd;
+
+	config->gb_addr_config = adev->gfx.config.gb_addr_config;
+
+	config->tile_config_ptr = adev->gfx.config.tile_mode_array;
+	config->num_tile_configs =
+			ARRAY_SIZE(adev->gfx.config.tile_mode_array);
+	config->macro_tile_config_ptr =
+			adev->gfx.config.macrotile_mode_array;
+	config->num_macro_tile_configs =
+			ARRAY_SIZE(adev->gfx.config.macrotile_mode_array);
+
+	return 0;
+}
+
+static const struct kfd2kgd_calls kfd2kgd = {
+	.init_gtt_mem_allocation = alloc_gtt_mem,
+	.free_gtt_mem = free_gtt_mem,
+	.get_local_mem_info = get_local_mem_info,
+	.get_gpu_clock_counter = get_gpu_clock_counter,
+	.get_max_engine_clock_in_mhz = get_max_engine_clock_in_mhz,
+	.alloc_pasid = amdgpu_pasid_alloc,
+	.free_pasid = amdgpu_pasid_free,
+	.program_sh_mem_settings = kgd_program_sh_mem_settings,
+	.set_pasid_vmid_mapping = kgd_set_pasid_vmid_mapping,
+	.init_interrupts = kgd_init_interrupts,
+	.hqd_load = kgd_hqd_load,
+	.hqd_sdma_load = kgd_hqd_sdma_load,
+	.hqd_dump = kgd_hqd_dump,
+	.hqd_sdma_dump = kgd_hqd_sdma_dump,
+	.hqd_is_occupied = kgd_hqd_is_occupied,
+	.hqd_sdma_is_occupied = kgd_hqd_sdma_is_occupied,
+	.hqd_destroy = kgd_hqd_destroy,
+	.hqd_sdma_destroy = kgd_hqd_sdma_destroy,
+	.address_watch_disable = kgd_address_watch_disable,
+	.address_watch_execute = kgd_address_watch_execute,
+	.wave_control_execute = kgd_wave_control_execute,
+	.address_watch_get_offset = kgd_address_watch_get_offset,
+	.get_atc_vmid_pasid_mapping_pasid =
+			get_atc_vmid_pasid_mapping_pasid,
+	.get_atc_vmid_pasid_mapping_valid =
+			get_atc_vmid_pasid_mapping_valid,
+	.get_fw_version = get_fw_version,
+	.set_scratch_backing_va = set_scratch_backing_va,
+	.get_tile_config = amdgpu_amdkfd_get_tile_config,
+	.get_cu_info = get_cu_info,
+	.get_vram_usage = amdgpu_amdkfd_get_vram_usage,
+	.create_process_vm = amdgpu_amdkfd_gpuvm_create_process_vm,
+	.acquire_process_vm = amdgpu_amdkfd_gpuvm_acquire_process_vm,
+	.destroy_process_vm = amdgpu_amdkfd_gpuvm_destroy_process_vm,
+	.get_process_page_dir = amdgpu_amdkfd_gpuvm_get_process_page_dir,
+	.set_vm_context_page_table_base = set_vm_context_page_table_base,
+	.alloc_memory_of_gpu = amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu,
+	.free_memory_of_gpu = amdgpu_amdkfd_gpuvm_free_memory_of_gpu,
+	.map_memory_to_gpu = amdgpu_amdkfd_gpuvm_map_memory_to_gpu,
+	.unmap_memory_to_gpu = amdgpu_amdkfd_gpuvm_unmap_memory_from_gpu,
+	.sync_memory = amdgpu_amdkfd_gpuvm_sync_memory,
+	.map_gtt_bo_to_kernel = amdgpu_amdkfd_gpuvm_map_gtt_bo_to_kernel,
+	.restore_process_bos = amdgpu_amdkfd_gpuvm_restore_process_bos,
+	.invalidate_tlbs = invalidate_tlbs,
+	.invalidate_tlbs_vmid = invalidate_tlbs_vmid,
+	.submit_ib = amdgpu_amdkfd_submit_ib,
+};
+
+struct kfd2kgd_calls *amdgpu_amdkfd_gfx_9_0_get_functions(void)
+{
+	return (struct kfd2kgd_calls *)&kfd2kgd;
+}
+
+static inline struct amdgpu_device *get_amdgpu_device(struct kgd_dev *kgd)
+{
+	return (struct amdgpu_device *)kgd;
+}
+
+static void lock_srbm(struct kgd_dev *kgd, uint32_t mec, uint32_t pipe,
+			uint32_t queue, uint32_t vmid)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+
+	mutex_lock(&adev->srbm_mutex);
+	soc15_grbm_select(adev, mec, pipe, queue, vmid);
+}
+
+static void unlock_srbm(struct kgd_dev *kgd)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+
+	soc15_grbm_select(adev, 0, 0, 0, 0);
+	mutex_unlock(&adev->srbm_mutex);
+}
+
+static void acquire_queue(struct kgd_dev *kgd, uint32_t pipe_id,
+				uint32_t queue_id)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+
+	uint32_t mec = (pipe_id / adev->gfx.mec.num_pipe_per_mec) + 1;
+	uint32_t pipe = (pipe_id % adev->gfx.mec.num_pipe_per_mec);
+
+	lock_srbm(kgd, mec, pipe, queue_id, 0);
+}
+
+static uint32_t get_queue_mask(struct amdgpu_device *adev,
+			       uint32_t pipe_id, uint32_t queue_id)
+{
+	unsigned int bit = (pipe_id * adev->gfx.mec.num_queue_per_pipe +
+			    queue_id) & 31;
+
+	return ((uint32_t)1) << bit;
+}
+
+static void release_queue(struct kgd_dev *kgd)
+{
+	unlock_srbm(kgd);
+}
+
+static void kgd_program_sh_mem_settings(struct kgd_dev *kgd, uint32_t vmid,
+					uint32_t sh_mem_config,
+					uint32_t sh_mem_ape1_base,
+					uint32_t sh_mem_ape1_limit,
+					uint32_t sh_mem_bases)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+
+	lock_srbm(kgd, 0, 0, 0, vmid);
+
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmSH_MEM_CONFIG), sh_mem_config);
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmSH_MEM_BASES), sh_mem_bases);
+	/* APE1 no longer exists on GFX9 */
+
+	unlock_srbm(kgd);
+}
+
+static int kgd_set_pasid_vmid_mapping(struct kgd_dev *kgd, unsigned int pasid,
+					unsigned int vmid)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+
+	/*
+	 * We have to assume that there is no outstanding mapping.
+	 * The ATC_VMID_PASID_MAPPING_UPDATE_STATUS bit could be 0 because
+	 * a mapping is in progress or because a mapping finished
+	 * and the SW cleared it.
+	 * So the protocol is to always wait & clear.
+	 */
+	uint32_t pasid_mapping = (pasid == 0) ? 0 : (uint32_t)pasid |
+			ATC_VMID0_PASID_MAPPING__VALID_MASK;
+
+	/*
+	 * need to do this twice, once for gfx and once for mmhub
+	 * for ATC add 16 to VMID for mmhub, for IH different registers.
+	 * ATC_VMID0..15 registers are separate from ATC_VMID16..31.
+	 */
+
+	WREG32(SOC15_REG_OFFSET(ATHUB, 0, mmATC_VMID0_PASID_MAPPING) + vmid,
+	       pasid_mapping);
+
+	while (!(RREG32(SOC15_REG_OFFSET(
+				ATHUB, 0,
+				mmATC_VMID_PASID_MAPPING_UPDATE_STATUS)) &
+		 (1U << vmid)))
+		cpu_relax();
+
+	WREG32(SOC15_REG_OFFSET(ATHUB, 0,
+				mmATC_VMID_PASID_MAPPING_UPDATE_STATUS),
+	       1U << vmid);
+
+	/* Mapping vmid to pasid also for IH block */
+	WREG32(SOC15_REG_OFFSET(OSSSYS, 0, mmIH_VMID_0_LUT) + vmid,
+	       pasid_mapping);
+
+	WREG32(SOC15_REG_OFFSET(ATHUB, 0, mmATC_VMID16_PASID_MAPPING) + vmid,
+	       pasid_mapping);
+
+	while (!(RREG32(SOC15_REG_OFFSET(
+				ATHUB, 0,
+				mmATC_VMID_PASID_MAPPING_UPDATE_STATUS)) &
+		 (1U << (vmid + 16))))
+		cpu_relax();
+
+	WREG32(SOC15_REG_OFFSET(ATHUB, 0,
+				mmATC_VMID_PASID_MAPPING_UPDATE_STATUS),
+	       1U << (vmid + 16));
+
+	/* Mapping vmid to pasid also for IH block */
+	WREG32(SOC15_REG_OFFSET(OSSSYS, 0, mmIH_VMID_0_LUT_MM) + vmid,
+	       pasid_mapping);
+	return 0;
+}
+
+/* TODO - RING0 form of field is obsolete, seems to date back to SI
+ * but still works
+ */
+
+static int kgd_init_interrupts(struct kgd_dev *kgd, uint32_t pipe_id)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+	uint32_t mec;
+	uint32_t pipe;
+
+	mec = (pipe_id / adev->gfx.mec.num_pipe_per_mec) + 1;
+	pipe = (pipe_id % adev->gfx.mec.num_pipe_per_mec);
+
+	lock_srbm(kgd, mec, pipe, 0, 0);
+
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmCPC_INT_CNTL),
+		CP_INT_CNTL_RING0__TIME_STAMP_INT_ENABLE_MASK |
+		CP_INT_CNTL_RING0__OPCODE_ERROR_INT_ENABLE_MASK);
+
+	unlock_srbm(kgd);
+
+	return 0;
+}
+
+static uint32_t get_sdma_base_addr(struct amdgpu_device *adev,
+				unsigned int engine_id,
+				unsigned int queue_id)
+{
+	uint32_t base[2] = {
+		SOC15_REG_OFFSET(SDMA0, 0,
+				 mmSDMA0_RLC0_RB_CNTL) - mmSDMA0_RLC0_RB_CNTL,
+		SOC15_REG_OFFSET(SDMA1, 0,
+				 mmSDMA1_RLC0_RB_CNTL) - mmSDMA1_RLC0_RB_CNTL
+	};
+	uint32_t retval;
+
+	retval = base[engine_id] + queue_id * (mmSDMA0_RLC1_RB_CNTL -
+					       mmSDMA0_RLC0_RB_CNTL);
+
+	pr_debug("sdma base address: 0x%x\n", retval);
+
+	return retval;
+}
+
+static inline struct v9_mqd *get_mqd(void *mqd)
+{
+	return (struct v9_mqd *)mqd;
+}
+
+static inline struct v9_sdma_mqd *get_sdma_mqd(void *mqd)
+{
+	return (struct v9_sdma_mqd *)mqd;
+}
+
+static int kgd_hqd_load(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id,
+			uint32_t queue_id, uint32_t __user *wptr,
+			uint32_t wptr_shift, uint32_t wptr_mask,
+			struct mm_struct *mm)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+	struct v9_mqd *m;
+	uint32_t *mqd_hqd;
+	uint32_t reg, hqd_base, data;
+
+	m = get_mqd(mqd);
+
+	acquire_queue(kgd, pipe_id, queue_id);
+
+	/* HIQ is set during driver init period with vmid set to 0*/
+	if (m->cp_hqd_vmid == 0) {
+		uint32_t value, mec, pipe;
+
+		mec = (pipe_id / adev->gfx.mec.num_pipe_per_mec) + 1;
+		pipe = (pipe_id % adev->gfx.mec.num_pipe_per_mec);
+
+		pr_debug("kfd: set HIQ, mec:%d, pipe:%d, queue:%d.\n",
+			mec, pipe, queue_id);
+		value = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CP_SCHEDULERS));
+		value = REG_SET_FIELD(value, RLC_CP_SCHEDULERS, scheduler1,
+			((mec << 5) | (pipe << 3) | queue_id | 0x80));
+		WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CP_SCHEDULERS), value);
+	}
+
+	/* HQD registers extend from CP_MQD_BASE_ADDR to CP_HQD_EOP_WPTR_MEM. */
+	mqd_hqd = &m->cp_mqd_base_addr_lo;
+	hqd_base = SOC15_REG_OFFSET(GC, 0, mmCP_MQD_BASE_ADDR);
+
+	for (reg = hqd_base;
+	     reg <= SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_HI); reg++)
+		WREG32(reg, mqd_hqd[reg - hqd_base]);
+
+
+	/* Activate doorbell logic before triggering WPTR poll. */
+	data = REG_SET_FIELD(m->cp_hqd_pq_doorbell_control,
+			     CP_HQD_PQ_DOORBELL_CONTROL, DOORBELL_EN, 1);
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_DOORBELL_CONTROL), data);
+
+	if (wptr) {
+		/* Don't read wptr with get_user because the user
+		 * context may not be accessible (if this function
+		 * runs in a work queue). Instead trigger a one-shot
+		 * polling read from memory in the CP. This assumes
+		 * that wptr is GPU-accessible in the queue's VMID via
+		 * ATC or SVM. WPTR==RPTR before starting the poll so
+		 * the CP starts fetching new commands from the right
+		 * place.
+		 *
+		 * Guessing a 64-bit WPTR from a 32-bit RPTR is a bit
+		 * tricky. Assume that the queue didn't overflow. The
+		 * number of valid bits in the 32-bit RPTR depends on
+		 * the queue size. The remaining bits are taken from
+		 * the saved 64-bit WPTR. If the WPTR wrapped, add the
+		 * queue size.
+		 */
+		uint32_t queue_size =
+			2 << REG_GET_FIELD(m->cp_hqd_pq_control,
+					   CP_HQD_PQ_CONTROL, QUEUE_SIZE);
+		uint64_t guessed_wptr = m->cp_hqd_pq_rptr & (queue_size - 1);
+
+		if ((m->cp_hqd_pq_wptr_lo & (queue_size - 1)) < guessed_wptr)
+			guessed_wptr += queue_size;
+		guessed_wptr += m->cp_hqd_pq_wptr_lo & ~(queue_size - 1);
+		guessed_wptr += (uint64_t)m->cp_hqd_pq_wptr_hi << 32;
+
+		WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_LO),
+		       lower_32_bits(guessed_wptr));
+		WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_HI),
+		       upper_32_bits(guessed_wptr));
+		WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_POLL_ADDR),
+		       lower_32_bits((uint64_t)wptr));
+		WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_POLL_ADDR_HI),
+		       upper_32_bits((uint64_t)wptr));
+		WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_PQ_WPTR_POLL_CNTL1),
+		       get_queue_mask(adev, pipe_id, queue_id));
+	}
+
+	/* Start the EOP fetcher */
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_EOP_RPTR),
+	       REG_SET_FIELD(m->cp_hqd_eop_rptr,
+			     CP_HQD_EOP_RPTR, INIT_FETCHER, 1));
+
+	data = REG_SET_FIELD(m->cp_hqd_active, CP_HQD_ACTIVE, ACTIVE, 1);
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_ACTIVE), data);
+
+	release_queue(kgd);
+
+	return 0;
+}
+
+static int kgd_hqd_dump(struct kgd_dev *kgd,
+			uint32_t pipe_id, uint32_t queue_id,
+			uint32_t (**dump)[2], uint32_t *n_regs)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+	uint32_t i = 0, reg;
+#define HQD_N_REGS 56
+#define DUMP_REG(addr) do {				\
+		if (WARN_ON_ONCE(i >= HQD_N_REGS))	\
+			break;				\
+		(*dump)[i][0] = (addr) << 2;		\
+		(*dump)[i++][1] = RREG32(addr);		\
+	} while (0)
+
+	*dump = kmalloc(HQD_N_REGS*2*sizeof(uint32_t), GFP_KERNEL);
+	if (*dump == NULL)
+		return -ENOMEM;
+
+	acquire_queue(kgd, pipe_id, queue_id);
+
+	for (reg = SOC15_REG_OFFSET(GC, 0, mmCP_MQD_BASE_ADDR);
+	     reg <= SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_HI); reg++)
+		DUMP_REG(reg);
+
+	release_queue(kgd);
+
+	WARN_ON_ONCE(i != HQD_N_REGS);
+	*n_regs = i;
+
+	return 0;
+}
+
+static int kgd_hqd_sdma_load(struct kgd_dev *kgd, void *mqd,
+			     uint32_t __user *wptr, struct mm_struct *mm)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+	struct v9_sdma_mqd *m;
+	uint32_t sdma_base_addr, sdmax_gfx_context_cntl;
+	unsigned long end_jiffies;
+	uint32_t data;
+	uint64_t data64;
+	uint64_t __user *wptr64 = (uint64_t __user *)wptr;
+
+	m = get_sdma_mqd(mqd);
+	sdma_base_addr = get_sdma_base_addr(adev, m->sdma_engine_id,
+					    m->sdma_queue_id);
+	sdmax_gfx_context_cntl = m->sdma_engine_id ?
+		SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_GFX_CONTEXT_CNTL) :
+		SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_GFX_CONTEXT_CNTL);
+
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL,
+		m->sdmax_rlcx_rb_cntl & (~SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK));
+
+	end_jiffies = msecs_to_jiffies(2000) + jiffies;
+	while (true) {
+		data = RREG32(sdma_base_addr + mmSDMA0_RLC0_CONTEXT_STATUS);
+		if (data & SDMA0_RLC0_CONTEXT_STATUS__IDLE_MASK)
+			break;
+		if (time_after(jiffies, end_jiffies))
+			return -ETIME;
+		usleep_range(500, 1000);
+	}
+	data = RREG32(sdmax_gfx_context_cntl);
+	data = REG_SET_FIELD(data, SDMA0_GFX_CONTEXT_CNTL,
+			     RESUME_CTX, 0);
+	WREG32(sdmax_gfx_context_cntl, data);
+
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_DOORBELL_OFFSET,
+	       m->sdmax_rlcx_doorbell_offset);
+
+	data = REG_SET_FIELD(m->sdmax_rlcx_doorbell, SDMA0_RLC0_DOORBELL,
+			     ENABLE, 1);
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_DOORBELL, data);
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR, m->sdmax_rlcx_rb_rptr);
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_HI,
+				m->sdmax_rlcx_rb_rptr_hi);
+
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_MINOR_PTR_UPDATE, 1);
+	if (read_user_wptr(mm, wptr64, data64)) {
+		WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR,
+		       lower_32_bits(data64));
+		WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR_HI,
+		       upper_32_bits(data64));
+	} else {
+		WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR,
+		       m->sdmax_rlcx_rb_rptr);
+		WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR_HI,
+		       m->sdmax_rlcx_rb_rptr_hi);
+	}
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_MINOR_PTR_UPDATE, 0);
+
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_BASE, m->sdmax_rlcx_rb_base);
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_BASE_HI,
+			m->sdmax_rlcx_rb_base_hi);
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_ADDR_LO,
+			m->sdmax_rlcx_rb_rptr_addr_lo);
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_ADDR_HI,
+			m->sdmax_rlcx_rb_rptr_addr_hi);
+
+	data = REG_SET_FIELD(m->sdmax_rlcx_rb_cntl, SDMA0_RLC0_RB_CNTL,
+			     RB_ENABLE, 1);
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL, data);
+
+	return 0;
+}
+
+static int kgd_hqd_sdma_dump(struct kgd_dev *kgd,
+			     uint32_t engine_id, uint32_t queue_id,
+			     uint32_t (**dump)[2], uint32_t *n_regs)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+	uint32_t sdma_base_addr = get_sdma_base_addr(adev, engine_id, queue_id);
+	uint32_t i = 0, reg;
+#undef HQD_N_REGS
+#define HQD_N_REGS (19+6+7+10)
+
+	*dump = kmalloc(HQD_N_REGS*2*sizeof(uint32_t), GFP_KERNEL);
+	if (*dump == NULL)
+		return -ENOMEM;
+
+	for (reg = mmSDMA0_RLC0_RB_CNTL; reg <= mmSDMA0_RLC0_DOORBELL; reg++)
+		DUMP_REG(sdma_base_addr + reg);
+	for (reg = mmSDMA0_RLC0_STATUS; reg <= mmSDMA0_RLC0_CSA_ADDR_HI; reg++)
+		DUMP_REG(sdma_base_addr + reg);
+	for (reg = mmSDMA0_RLC0_IB_SUB_REMAIN;
+	     reg <= mmSDMA0_RLC0_MINOR_PTR_UPDATE; reg++)
+		DUMP_REG(sdma_base_addr + reg);
+	for (reg = mmSDMA0_RLC0_MIDCMD_DATA0;
+	     reg <= mmSDMA0_RLC0_MIDCMD_CNTL; reg++)
+		DUMP_REG(sdma_base_addr + reg);
+
+	WARN_ON_ONCE(i != HQD_N_REGS);
+	*n_regs = i;
+
+	return 0;
+}
+
+static bool kgd_hqd_is_occupied(struct kgd_dev *kgd, uint64_t queue_address,
+				uint32_t pipe_id, uint32_t queue_id)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+	uint32_t act;
+	bool retval = false;
+	uint32_t low, high;
+
+	acquire_queue(kgd, pipe_id, queue_id);
+	act = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_ACTIVE));
+	if (act) {
+		low = lower_32_bits(queue_address >> 8);
+		high = upper_32_bits(queue_address >> 8);
+
+		if (low == RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_BASE)) &&
+		   high == RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_BASE_HI)))
+			retval = true;
+	}
+	release_queue(kgd);
+	return retval;
+}
+
+static bool kgd_hqd_sdma_is_occupied(struct kgd_dev *kgd, void *mqd)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+	struct v9_sdma_mqd *m;
+	uint32_t sdma_base_addr;
+	uint32_t sdma_rlc_rb_cntl;
+
+	m = get_sdma_mqd(mqd);
+	sdma_base_addr = get_sdma_base_addr(adev, m->sdma_engine_id,
+					    m->sdma_queue_id);
+
+	sdma_rlc_rb_cntl = RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL);
+
+	if (sdma_rlc_rb_cntl & SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK)
+		return true;
+
+	return false;
+}
+
+static int kgd_hqd_destroy(struct kgd_dev *kgd, void *mqd,
+				enum kfd_preempt_type reset_type,
+				unsigned int utimeout, uint32_t pipe_id,
+				uint32_t queue_id)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+	enum hqd_dequeue_request_type type;
+	unsigned long end_jiffies;
+	uint32_t temp;
+	struct v9_mqd *m = get_mqd(mqd);
+
+	acquire_queue(kgd, pipe_id, queue_id);
+
+	if (m->cp_hqd_vmid == 0)
+		WREG32_FIELD15(GC, 0, RLC_CP_SCHEDULERS, scheduler1, 0);
+
+	switch (reset_type) {
+	case KFD_PREEMPT_TYPE_WAVEFRONT_DRAIN:
+		type = DRAIN_PIPE;
+		break;
+	case KFD_PREEMPT_TYPE_WAVEFRONT_RESET:
+		type = RESET_WAVES;
+		break;
+	default:
+		type = DRAIN_PIPE;
+		break;
+	}
+
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_DEQUEUE_REQUEST), type);
+
+	end_jiffies = (utimeout * HZ / 1000) + jiffies;
+	while (true) {
+		temp = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_ACTIVE));
+		if (!(temp & CP_HQD_ACTIVE__ACTIVE_MASK))
+			break;
+		if (time_after(jiffies, end_jiffies)) {
+			pr_err("cp queue preemption time out.\n");
+			release_queue(kgd);
+			return -ETIME;
+		}
+		usleep_range(500, 1000);
+	}
+
+	release_queue(kgd);
+	return 0;
+}
+
+static int kgd_hqd_sdma_destroy(struct kgd_dev *kgd, void *mqd,
+				unsigned int utimeout)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+	struct v9_sdma_mqd *m;
+	uint32_t sdma_base_addr;
+	uint32_t temp;
+	unsigned long end_jiffies = (utimeout * HZ / 1000) + jiffies;
+
+	m = get_sdma_mqd(mqd);
+	sdma_base_addr = get_sdma_base_addr(adev, m->sdma_engine_id,
+					    m->sdma_queue_id);
+
+	temp = RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL);
+	temp = temp & ~SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK;
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL, temp);
+
+	while (true) {
+		temp = RREG32(sdma_base_addr + mmSDMA0_RLC0_CONTEXT_STATUS);
+		if (temp & SDMA0_RLC0_CONTEXT_STATUS__IDLE_MASK)
+			break;
+		if (time_after(jiffies, end_jiffies))
+			return -ETIME;
+		usleep_range(500, 1000);
+	}
+
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_DOORBELL, 0);
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL,
+		RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL) |
+		SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK);
+
+	m->sdmax_rlcx_rb_rptr = RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR);
+	m->sdmax_rlcx_rb_rptr_hi =
+		RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_HI);
+
+	return 0;
+}
+
+static bool get_atc_vmid_pasid_mapping_valid(struct kgd_dev *kgd,
+							uint8_t vmid)
+{
+	uint32_t reg;
+	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
+
+	reg = RREG32(SOC15_REG_OFFSET(ATHUB, 0, mmATC_VMID0_PASID_MAPPING)
+		     + vmid);
+	return reg & ATC_VMID0_PASID_MAPPING__VALID_MASK;
+}
+
+static uint16_t get_atc_vmid_pasid_mapping_pasid(struct kgd_dev *kgd,
+								uint8_t vmid)
+{
+	uint32_t reg;
+	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
+
+	reg = RREG32(SOC15_REG_OFFSET(ATHUB, 0, mmATC_VMID0_PASID_MAPPING)
+		     + vmid);
+	return reg & ATC_VMID0_PASID_MAPPING__PASID_MASK;
+}
+
+static void write_vmid_invalidate_request(struct kgd_dev *kgd, uint8_t vmid)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
+	uint32_t req = (1 << vmid) |
+		(0 << VM_INVALIDATE_ENG16_REQ__FLUSH_TYPE__SHIFT) | /* legacy */
+		VM_INVALIDATE_ENG16_REQ__INVALIDATE_L2_PTES_MASK |
+		VM_INVALIDATE_ENG16_REQ__INVALIDATE_L2_PDE0_MASK |
+		VM_INVALIDATE_ENG16_REQ__INVALIDATE_L2_PDE1_MASK |
+		VM_INVALIDATE_ENG16_REQ__INVALIDATE_L2_PDE2_MASK |
+		VM_INVALIDATE_ENG16_REQ__INVALIDATE_L1_PTES_MASK;
+
+	mutex_lock(&adev->srbm_mutex);
+
+	/* Use legacy mode tlb invalidation.
+	 *
+	 * Currently on Raven the code below is broken for anything but
+	 * legacy mode due to a MMHUB power gating problem. A workaround
+	 * is for MMHUB to wait until the condition PER_VMID_INVALIDATE_REQ
+	 * == PER_VMID_INVALIDATE_ACK instead of simply waiting for the ack
+	 * bit.
+	 *
+	 * TODO 1: agree on the right set of invalidation registers for
+	 * KFD use. Use the last one for now. Invalidate both GC and
+	 * MMHUB.
+	 *
+	 * TODO 2: support range-based invalidation, requires kfg2kgd
+	 * interface change
+	 */
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmVM_INVALIDATE_ENG16_ADDR_RANGE_LO32),
+				0xffffffff);
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmVM_INVALIDATE_ENG16_ADDR_RANGE_HI32),
+				0x0000001f);
+
+	WREG32(SOC15_REG_OFFSET(MMHUB, 0,
+				mmMMHUB_VM_INVALIDATE_ENG16_ADDR_RANGE_LO32),
+				0xffffffff);
+	WREG32(SOC15_REG_OFFSET(MMHUB, 0,
+				mmMMHUB_VM_INVALIDATE_ENG16_ADDR_RANGE_HI32),
+				0x0000001f);
+
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmVM_INVALIDATE_ENG16_REQ), req);
+
+	WREG32(SOC15_REG_OFFSET(MMHUB, 0, mmMMHUB_VM_INVALIDATE_ENG16_REQ),
+				req);
+
+	while (!(RREG32(SOC15_REG_OFFSET(GC, 0, mmVM_INVALIDATE_ENG16_ACK)) &
+					(1 << vmid)))
+		cpu_relax();
+
+	while (!(RREG32(SOC15_REG_OFFSET(MMHUB, 0,
+					mmMMHUB_VM_INVALIDATE_ENG16_ACK)) &
+					(1 << vmid)))
+		cpu_relax();
+
+	mutex_unlock(&adev->srbm_mutex);
+
+}
+
+static int invalidate_tlbs_with_kiq(struct amdgpu_device *adev, uint16_t pasid)
+{
+	signed long r;
+	uint32_t seq;
+	struct amdgpu_ring *ring = &adev->gfx.kiq.ring;
+
+	spin_lock(&adev->gfx.kiq.ring_lock);
+	amdgpu_ring_alloc(ring, 12); /* fence + invalidate_tlbs package*/
+	amdgpu_ring_write(ring, PACKET3(PACKET3_INVALIDATE_TLBS, 0));
+	amdgpu_ring_write(ring,
+			PACKET3_INVALIDATE_TLBS_DST_SEL(1) |
+			PACKET3_INVALIDATE_TLBS_ALL_HUB(1) |
+			PACKET3_INVALIDATE_TLBS_PASID(pasid) |
+			PACKET3_INVALIDATE_TLBS_FLUSH_TYPE(0)); /* legacy */
+	amdgpu_fence_emit_polling(ring, &seq);
+	amdgpu_ring_commit(ring);
+	spin_unlock(&adev->gfx.kiq.ring_lock);
+
+	r = amdgpu_fence_wait_polling(ring, seq, adev->usec_timeout);
+	if (r < 1) {
+		DRM_ERROR("wait for kiq fence error: %ld.\n", r);
+		return -ETIME;
+	}
+
+	return 0;
+}
+
+static int invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
+	int vmid;
+	struct amdgpu_ring *ring = &adev->gfx.kiq.ring;
+
+	if (ring->ready)
+		return invalidate_tlbs_with_kiq(adev, pasid);
+
+	for (vmid = 0; vmid < 16; vmid++) {
+		if (!amdgpu_amdkfd_is_kfd_vmid(adev, vmid))
+			continue;
+		if (get_atc_vmid_pasid_mapping_valid(kgd, vmid)) {
+			if (get_atc_vmid_pasid_mapping_pasid(kgd, vmid)
+				== pasid) {
+				write_vmid_invalidate_request(kgd, vmid);
+				break;
+			}
+		}
+	}
+
+	return 0;
+}
+
+static int invalidate_tlbs_vmid(struct kgd_dev *kgd, uint16_t vmid)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
+
+	if (!amdgpu_amdkfd_is_kfd_vmid(adev, vmid)) {
+		pr_err("non kfd vmid %d\n", vmid);
+		return 0;
+	}
+
+	write_vmid_invalidate_request(kgd, vmid);
+	return 0;
+}
+
+static int kgd_address_watch_disable(struct kgd_dev *kgd)
+{
+	return 0;
+}
+
+static int kgd_address_watch_execute(struct kgd_dev *kgd,
+					unsigned int watch_point_id,
+					uint32_t cntl_val,
+					uint32_t addr_hi,
+					uint32_t addr_lo)
+{
+	return 0;
+}
+
+static int kgd_wave_control_execute(struct kgd_dev *kgd,
+					uint32_t gfx_index_val,
+					uint32_t sq_cmd)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+	uint32_t data = 0;
+
+	mutex_lock(&adev->grbm_idx_mutex);
+
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_GFX_INDEX), gfx_index_val);
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmSQ_CMD), sq_cmd);
+
+	data = REG_SET_FIELD(data, GRBM_GFX_INDEX,
+		INSTANCE_BROADCAST_WRITES, 1);
+	data = REG_SET_FIELD(data, GRBM_GFX_INDEX,
+		SH_BROADCAST_WRITES, 1);
+	data = REG_SET_FIELD(data, GRBM_GFX_INDEX,
+		SE_BROADCAST_WRITES, 1);
+
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_GFX_INDEX), data);
+	mutex_unlock(&adev->grbm_idx_mutex);
+
+	return 0;
+}
+
+static uint32_t kgd_address_watch_get_offset(struct kgd_dev *kgd,
+					unsigned int watch_point_id,
+					unsigned int reg_offset)
+{
+	return 0;
+}
+
+static void set_scratch_backing_va(struct kgd_dev *kgd,
+					uint64_t va, uint32_t vmid)
+{
+	/* No longer needed on GFXv9. The scratch base address is
+	 * passed to the shader by the CP. It's the user mode driver's
+	 * responsibility.
+	 */
+}
+
+/* FIXME: Does this need to be ASIC-specific code? */
+static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
+	const union amdgpu_firmware_header *hdr;
+
+	switch (type) {
+	case KGD_ENGINE_PFP:
+		hdr = (const union amdgpu_firmware_header *)adev->gfx.pfp_fw->data;
+		break;
+
+	case KGD_ENGINE_ME:
+		hdr = (const union amdgpu_firmware_header *)adev->gfx.me_fw->data;
+		break;
+
+	case KGD_ENGINE_CE:
+		hdr = (const union amdgpu_firmware_header *)adev->gfx.ce_fw->data;
+		break;
+
+	case KGD_ENGINE_MEC1:
+		hdr = (const union amdgpu_firmware_header *)adev->gfx.mec_fw->data;
+		break;
+
+	case KGD_ENGINE_MEC2:
+		hdr = (const union amdgpu_firmware_header *)adev->gfx.mec2_fw->data;
+		break;
+
+	case KGD_ENGINE_RLC:
+		hdr = (const union amdgpu_firmware_header *)adev->gfx.rlc_fw->data;
+		break;
+
+	case KGD_ENGINE_SDMA1:
+		hdr = (const union amdgpu_firmware_header *)adev->sdma.instance[0].fw->data;
+		break;
+
+	case KGD_ENGINE_SDMA2:
+		hdr = (const union amdgpu_firmware_header *)adev->sdma.instance[1].fw->data;
+		break;
+
+	default:
+		return 0;
+	}
+
+	if (hdr == NULL)
+		return 0;
+
+	/* Only 12 bit in use*/
+	return hdr->common.ucode_version;
+}
+
+static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid,
+		uint32_t page_table_base)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+	uint64_t base = (uint64_t)page_table_base << PAGE_SHIFT |
+		AMDGPU_PTE_VALID;
+
+	if (!amdgpu_amdkfd_is_kfd_vmid(adev, vmid)) {
+		pr_err("trying to set page table base for wrong VMID %u\n",
+		       vmid);
+		return;
+	}
+
+	/* TODO: take advantage of per-process address space size. For
+	 * now, all processes share the same address space size, like
+	 * on GFX8 and older.
+	 */
+	WREG32(SOC15_REG_OFFSET(MMHUB, 0, mmMMHUB_VM_CONTEXT0_PAGE_TABLE_START_ADDR_LO32) + (vmid*2), 0);
+	WREG32(SOC15_REG_OFFSET(MMHUB, 0, mmMMHUB_VM_CONTEXT0_PAGE_TABLE_START_ADDR_HI32) + (vmid*2), 0);
+
+	WREG32(SOC15_REG_OFFSET(MMHUB, 0, mmMMHUB_VM_CONTEXT0_PAGE_TABLE_END_ADDR_LO32) + (vmid*2),
+			lower_32_bits(adev->vm_manager.max_pfn - 1));
+	WREG32(SOC15_REG_OFFSET(MMHUB, 0, mmMMHUB_VM_CONTEXT0_PAGE_TABLE_END_ADDR_HI32) + (vmid*2),
+			upper_32_bits(adev->vm_manager.max_pfn - 1));
+
+	WREG32(SOC15_REG_OFFSET(MMHUB, 0, mmMMHUB_VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32) + (vmid*2), lower_32_bits(base));
+	WREG32(SOC15_REG_OFFSET(MMHUB, 0, mmMMHUB_VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32) + (vmid*2), upper_32_bits(base));
+
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmVM_CONTEXT0_PAGE_TABLE_START_ADDR_LO32) + (vmid*2), 0);
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmVM_CONTEXT0_PAGE_TABLE_START_ADDR_HI32) + (vmid*2), 0);
+
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmVM_CONTEXT0_PAGE_TABLE_END_ADDR_LO32) + (vmid*2),
+			lower_32_bits(adev->vm_manager.max_pfn - 1));
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmVM_CONTEXT0_PAGE_TABLE_END_ADDR_HI32) + (vmid*2),
+			upper_32_bits(adev->vm_manager.max_pfn - 1));
+
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmVM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32) + (vmid*2), lower_32_bits(base));
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmVM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32) + (vmid*2), upper_32_bits(base));
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index 9d39fd5..e5962e6 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -4686,6 +4686,7 @@ static int gfx_v9_0_get_cu_info(struct amdgpu_device *adev,
 
 	cu_info->number = active_cu_number;
 	cu_info->ao_cu_mask = ao_cu_mask;
+	cu_info->simd_per_cu = NUM_SIMD_PER_CU;
 
 	return 0;
 }
-- 
cgit v1.1


From 642a0e80262af8e9d5b8129e2149c670ab3bb4b8 Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Tue, 10 Apr 2018 17:33:02 -0400
Subject: drm/amdgpu: Add doorbell routing info to kgd2kfd_shared_resources

This is needed for Vega10 and later ASICs to let KFD know which
doorbells can be used for SDMA and CP queues respectively.

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c      | 22 ++++++++++++++++++++++
 drivers/gpu/drm/amd/include/kgd_kfd_interface.h | 15 +++++++++++++++
 2 files changed, 37 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
index fcd10db..cd0e8f1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
@@ -179,6 +179,28 @@ void amdgpu_amdkfd_device_init(struct amdgpu_device *adev)
 				&gpu_resources.doorbell_physical_address,
 				&gpu_resources.doorbell_aperture_size,
 				&gpu_resources.doorbell_start_offset);
+		if (adev->asic_type >= CHIP_VEGA10) {
+			/* On SOC15 the BIF is involved in routing
+			 * doorbells using the low 12 bits of the
+			 * address. Communicate the assignments to
+			 * KFD. KFD uses two doorbell pages per
+			 * process in case of 64-bit doorbells so we
+			 * can use each doorbell assignment twice.
+			 */
+			gpu_resources.sdma_doorbell[0][0] =
+				AMDGPU_DOORBELL64_sDMA_ENGINE0;
+			gpu_resources.sdma_doorbell[0][1] =
+				AMDGPU_DOORBELL64_sDMA_ENGINE0 + 0x200;
+			gpu_resources.sdma_doorbell[1][0] =
+				AMDGPU_DOORBELL64_sDMA_ENGINE1;
+			gpu_resources.sdma_doorbell[1][1] =
+				AMDGPU_DOORBELL64_sDMA_ENGINE1 + 0x200;
+			/* Doorbells 0x0f0-0ff and 0x2f0-2ff are reserved for
+			 * SDMA, IH and VCN. So don't use them for the CP.
+			 */
+			gpu_resources.reserved_doorbell_mask = 0x1f0;
+			gpu_resources.reserved_doorbell_val  = 0x0f0;
+		}
 
 		kgd2kfd->device_init(adev->kfd, &gpu_resources);
 	}
diff --git a/drivers/gpu/drm/amd/include/kgd_kfd_interface.h b/drivers/gpu/drm/amd/include/kgd_kfd_interface.h
index 7cf3506..5733fbe 100644
--- a/drivers/gpu/drm/amd/include/kgd_kfd_interface.h
+++ b/drivers/gpu/drm/amd/include/kgd_kfd_interface.h
@@ -100,6 +100,21 @@ struct kgd2kfd_shared_resources {
 	/* Bit n == 1 means Queue n is available for KFD */
 	DECLARE_BITMAP(queue_bitmap, KGD_MAX_QUEUES);
 
+	/* Doorbell assignments (SOC15 and later chips only). Only
+	 * specific doorbells are routed to each SDMA engine. Others
+	 * are routed to IH and VCN. They are not usable by the CP.
+	 *
+	 * Any doorbell number D that satisfies the following condition
+	 * is reserved: (D & reserved_doorbell_mask) == reserved_doorbell_val
+	 *
+	 * KFD currently uses 1024 (= 0x3ff) doorbells per process. If
+	 * doorbells 0x0f0-0x0f7 and 0x2f-0x2f7 are reserved, that means
+	 * mask would be set to 0x1f8 and val set to 0x0f0.
+	 */
+	unsigned int sdma_doorbell[2][2];
+	unsigned int reserved_doorbell_mask;
+	unsigned int reserved_doorbell_val;
+
 	/* Base address of doorbell aperture. */
 	phys_addr_t doorbell_physical_address;
 
-- 
cgit v1.1


From ada2b29c4a79efbdc5bf5eed876bad6b00f43536 Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Tue, 10 Apr 2018 17:33:03 -0400
Subject: drm/amdkfd: Make doorbell size ASIC-dependent

This prepares for GFXv9 (Vega10), which has 64-bit doorbells.

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/kfd_device.c   | 10 +++++++
 drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c | 48 ++++++++++++++++---------------
 drivers/gpu/drm/amd/amdkfd/kfd_priv.h     |  7 +++--
 3 files changed, 39 insertions(+), 26 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
index 7b57995..f563acb 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
@@ -41,6 +41,7 @@ static const struct kfd_device_info kaveri_device_info = {
 	.max_pasid_bits = 16,
 	/* max num of queues for KV.TODO should be a dynamic value */
 	.max_no_of_hqd	= 24,
+	.doorbell_size  = 4,
 	.ih_ring_entry_size = 4 * sizeof(uint32_t),
 	.event_interrupt_class = &event_interrupt_class_cik,
 	.num_of_watch_points = 4,
@@ -55,6 +56,7 @@ static const struct kfd_device_info carrizo_device_info = {
 	.max_pasid_bits = 16,
 	/* max num of queues for CZ.TODO should be a dynamic value */
 	.max_no_of_hqd	= 24,
+	.doorbell_size  = 4,
 	.ih_ring_entry_size = 4 * sizeof(uint32_t),
 	.event_interrupt_class = &event_interrupt_class_cik,
 	.num_of_watch_points = 4,
@@ -70,6 +72,7 @@ static const struct kfd_device_info hawaii_device_info = {
 	.max_pasid_bits = 16,
 	/* max num of queues for KV.TODO should be a dynamic value */
 	.max_no_of_hqd	= 24,
+	.doorbell_size  = 4,
 	.ih_ring_entry_size = 4 * sizeof(uint32_t),
 	.event_interrupt_class = &event_interrupt_class_cik,
 	.num_of_watch_points = 4,
@@ -83,6 +86,7 @@ static const struct kfd_device_info tonga_device_info = {
 	.asic_family = CHIP_TONGA,
 	.max_pasid_bits = 16,
 	.max_no_of_hqd  = 24,
+	.doorbell_size  = 4,
 	.ih_ring_entry_size = 4 * sizeof(uint32_t),
 	.event_interrupt_class = &event_interrupt_class_cik,
 	.num_of_watch_points = 4,
@@ -96,6 +100,7 @@ static const struct kfd_device_info tonga_vf_device_info = {
 	.asic_family = CHIP_TONGA,
 	.max_pasid_bits = 16,
 	.max_no_of_hqd  = 24,
+	.doorbell_size  = 4,
 	.ih_ring_entry_size = 4 * sizeof(uint32_t),
 	.event_interrupt_class = &event_interrupt_class_cik,
 	.num_of_watch_points = 4,
@@ -109,6 +114,7 @@ static const struct kfd_device_info fiji_device_info = {
 	.asic_family = CHIP_FIJI,
 	.max_pasid_bits = 16,
 	.max_no_of_hqd  = 24,
+	.doorbell_size  = 4,
 	.ih_ring_entry_size = 4 * sizeof(uint32_t),
 	.event_interrupt_class = &event_interrupt_class_cik,
 	.num_of_watch_points = 4,
@@ -122,6 +128,7 @@ static const struct kfd_device_info fiji_vf_device_info = {
 	.asic_family = CHIP_FIJI,
 	.max_pasid_bits = 16,
 	.max_no_of_hqd  = 24,
+	.doorbell_size  = 4,
 	.ih_ring_entry_size = 4 * sizeof(uint32_t),
 	.event_interrupt_class = &event_interrupt_class_cik,
 	.num_of_watch_points = 4,
@@ -136,6 +143,7 @@ static const struct kfd_device_info polaris10_device_info = {
 	.asic_family = CHIP_POLARIS10,
 	.max_pasid_bits = 16,
 	.max_no_of_hqd  = 24,
+	.doorbell_size  = 4,
 	.ih_ring_entry_size = 4 * sizeof(uint32_t),
 	.event_interrupt_class = &event_interrupt_class_cik,
 	.num_of_watch_points = 4,
@@ -149,6 +157,7 @@ static const struct kfd_device_info polaris10_vf_device_info = {
 	.asic_family = CHIP_POLARIS10,
 	.max_pasid_bits = 16,
 	.max_no_of_hqd  = 24,
+	.doorbell_size  = 4,
 	.ih_ring_entry_size = 4 * sizeof(uint32_t),
 	.event_interrupt_class = &event_interrupt_class_cik,
 	.num_of_watch_points = 4,
@@ -162,6 +171,7 @@ static const struct kfd_device_info polaris11_device_info = {
 	.asic_family = CHIP_POLARIS11,
 	.max_pasid_bits = 16,
 	.max_no_of_hqd  = 24,
+	.doorbell_size  = 4,
 	.ih_ring_entry_size = 4 * sizeof(uint32_t),
 	.event_interrupt_class = &event_interrupt_class_cik,
 	.num_of_watch_points = 4,
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c b/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c
index ebb4da14..4840314 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c
@@ -33,7 +33,6 @@
 
 static DEFINE_IDA(doorbell_ida);
 static unsigned int max_doorbell_slices;
-#define KFD_SIZE_OF_DOORBELL_IN_BYTES 4
 
 /*
  * Each device exposes a doorbell aperture, a PCI MMIO aperture that
@@ -50,9 +49,9 @@ static unsigned int max_doorbell_slices;
  */
 
 /* # of doorbell bytes allocated for each process. */
-static inline size_t doorbell_process_allocation(void)
+static size_t kfd_doorbell_process_slice(struct kfd_dev *kfd)
 {
-	return roundup(KFD_SIZE_OF_DOORBELL_IN_BYTES *
+	return roundup(kfd->device_info->doorbell_size *
 			KFD_MAX_NUM_OF_QUEUES_PER_PROCESS,
 			PAGE_SIZE);
 }
@@ -72,16 +71,16 @@ int kfd_doorbell_init(struct kfd_dev *kfd)
 
 	doorbell_start_offset =
 			roundup(kfd->shared_resources.doorbell_start_offset,
-					doorbell_process_allocation());
+					kfd_doorbell_process_slice(kfd));
 
 	doorbell_aperture_size =
 			rounddown(kfd->shared_resources.doorbell_aperture_size,
-					doorbell_process_allocation());
+					kfd_doorbell_process_slice(kfd));
 
 	if (doorbell_aperture_size > doorbell_start_offset)
 		doorbell_process_limit =
 			(doorbell_aperture_size - doorbell_start_offset) /
-						doorbell_process_allocation();
+						kfd_doorbell_process_slice(kfd);
 	else
 		return -ENOSPC;
 
@@ -95,7 +94,7 @@ int kfd_doorbell_init(struct kfd_dev *kfd)
 	kfd->doorbell_id_offset = doorbell_start_offset / sizeof(u32);
 
 	kfd->doorbell_kernel_ptr = ioremap(kfd->doorbell_base,
-						doorbell_process_allocation());
+					   kfd_doorbell_process_slice(kfd));
 
 	if (!kfd->doorbell_kernel_ptr)
 		return -ENOMEM;
@@ -132,16 +131,16 @@ int kfd_doorbell_mmap(struct kfd_process *process, struct vm_area_struct *vma)
 	phys_addr_t address;
 	struct kfd_dev *dev;
 
+	/* Find kfd device according to gpu id */
+	dev = kfd_device_by_id(vma->vm_pgoff);
+	if (!dev)
+		return -EINVAL;
+
 	/*
 	 * For simplicitly we only allow mapping of the entire doorbell
 	 * allocation of a single device & process.
 	 */
-	if (vma->vm_end - vma->vm_start != doorbell_process_allocation())
-		return -EINVAL;
-
-	/* Find kfd device according to gpu id */
-	dev = kfd_device_by_id(vma->vm_pgoff);
-	if (!dev)
+	if (vma->vm_end - vma->vm_start != kfd_doorbell_process_slice(dev))
 		return -EINVAL;
 
 	/* Calculate physical address of doorbell */
@@ -158,19 +157,19 @@ int kfd_doorbell_mmap(struct kfd_process *process, struct vm_area_struct *vma)
 		 "     vm_flags            == 0x%04lX\n"
 		 "     size                == 0x%04lX\n",
 		 (unsigned long long) vma->vm_start, address, vma->vm_flags,
-		 doorbell_process_allocation());
+		 kfd_doorbell_process_slice(dev));
 
 
 	return io_remap_pfn_range(vma,
 				vma->vm_start,
 				address >> PAGE_SHIFT,
-				doorbell_process_allocation(),
+				kfd_doorbell_process_slice(dev),
 				vma->vm_page_prot);
 }
 
 
 /* get kernel iomem pointer for a doorbell */
-u32 __iomem *kfd_get_kernel_doorbell(struct kfd_dev *kfd,
+void __iomem *kfd_get_kernel_doorbell(struct kfd_dev *kfd,
 					unsigned int *doorbell_off)
 {
 	u32 inx;
@@ -185,6 +184,8 @@ u32 __iomem *kfd_get_kernel_doorbell(struct kfd_dev *kfd,
 	if (inx >= KFD_MAX_NUM_OF_QUEUES_PER_PROCESS)
 		return NULL;
 
+	inx *= kfd->device_info->doorbell_size / sizeof(u32);
+
 	/*
 	 * Calculating the kernel doorbell offset using the first
 	 * doorbell page.
@@ -210,7 +211,7 @@ void kfd_release_kernel_doorbell(struct kfd_dev *kfd, u32 __iomem *db_addr)
 	mutex_unlock(&kfd->doorbell_mutex);
 }
 
-inline void write_kernel_doorbell(u32 __iomem *db, u32 value)
+void write_kernel_doorbell(void __iomem *db, u32 value)
 {
 	if (db) {
 		writel(value, db);
@@ -228,20 +229,21 @@ unsigned int kfd_queue_id_to_doorbell(struct kfd_dev *kfd,
 {
 	/*
 	 * doorbell_id_offset accounts for doorbells taken by KGD.
-	 * index * doorbell_process_allocation/sizeof(u32) adjusts to
-	 * the process's doorbells.
+	 * index * kfd_doorbell_process_slice/sizeof(u32) adjusts to
+	 * the process's doorbells. The offset returned is in dword
+	 * units regardless of the ASIC-dependent doorbell size.
 	 */
 	return kfd->doorbell_id_offset +
 		process->doorbell_index
-		* doorbell_process_allocation() / sizeof(u32) +
-		queue_id;
+		* kfd_doorbell_process_slice(kfd) / sizeof(u32) +
+		queue_id * kfd->device_info->doorbell_size / sizeof(u32);
 }
 
 uint64_t kfd_get_number_elems(struct kfd_dev *kfd)
 {
 	uint64_t num_of_elems = (kfd->shared_resources.doorbell_aperture_size -
 				kfd->shared_resources.doorbell_start_offset) /
-					doorbell_process_allocation() + 1;
+					kfd_doorbell_process_slice(kfd) + 1;
 
 	return num_of_elems;
 
@@ -251,7 +253,7 @@ phys_addr_t kfd_get_process_doorbells(struct kfd_dev *dev,
 					struct kfd_process *process)
 {
 	return dev->doorbell_base +
-		process->doorbell_index * doorbell_process_allocation();
+		process->doorbell_index * kfd_doorbell_process_slice(dev);
 }
 
 int kfd_alloc_process_doorbells(struct kfd_process *process)
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
index 4d5c49e..d9c0fe12 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
@@ -160,6 +160,7 @@ struct kfd_device_info {
 	const struct kfd_event_interrupt_class *event_interrupt_class;
 	unsigned int max_pasid_bits;
 	unsigned int max_no_of_hqd;
+	unsigned int doorbell_size;
 	size_t ih_ring_entry_size;
 	uint8_t num_of_watch_points;
 	uint16_t mqd_size_aligned;
@@ -364,7 +365,7 @@ struct queue_properties {
 	uint32_t queue_percent;
 	uint32_t *read_ptr;
 	uint32_t *write_ptr;
-	uint32_t __iomem *doorbell_ptr;
+	void __iomem *doorbell_ptr;
 	uint32_t doorbell_off;
 	bool is_interop;
 	bool is_evicted;
@@ -728,11 +729,11 @@ void kfd_pasid_free(unsigned int pasid);
 int kfd_doorbell_init(struct kfd_dev *kfd);
 void kfd_doorbell_fini(struct kfd_dev *kfd);
 int kfd_doorbell_mmap(struct kfd_process *process, struct vm_area_struct *vma);
-u32 __iomem *kfd_get_kernel_doorbell(struct kfd_dev *kfd,
+void __iomem *kfd_get_kernel_doorbell(struct kfd_dev *kfd,
 					unsigned int *doorbell_off);
 void kfd_release_kernel_doorbell(struct kfd_dev *kfd, u32 __iomem *db_addr);
 u32 read_kernel_doorbell(u32 __iomem *db);
-void write_kernel_doorbell(u32 __iomem *db, u32 value);
+void write_kernel_doorbell(void __iomem *db, u32 value);
 unsigned int kfd_queue_id_to_doorbell(struct kfd_dev *kfd,
 					struct kfd_process *process,
 					unsigned int queue_id);
-- 
cgit v1.1


From df03ef9342ce09985210679a734f88a269c19ff5 Mon Sep 17 00:00:00 2001
From: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Date: Tue, 10 Apr 2018 17:33:04 -0400
Subject: drm/amdkfd: Clean up KFD_MMAP_ offset handling

Use bit-rotate for better clarity and remove _MASK from the #defines as
these represent mmap types.

Centralize all the parsing of the mmap offset in kfd_mmap and add device
parameter to doorbell and reserved_mem map functions.

Encode gpu_id into upper bits of vm_pgoff. This frees up the lower bits
for encoding the the doorbell ID on Vega10.

Signed-off-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/kfd_chardev.c  | 35 ++++++++++++++++++----------
 drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c |  9 ++------
 drivers/gpu/drm/amd/amdkfd/kfd_events.c   |  2 +-
 drivers/gpu/drm/amd/amdkfd/kfd_priv.h     | 38 ++++++++++++++++++++++++-------
 drivers/gpu/drm/amd/amdkfd/kfd_process.c  |  8 +++----
 5 files changed, 59 insertions(+), 33 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
index cd679cf..519c7b1 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
@@ -292,7 +292,8 @@ static int kfd_ioctl_create_queue(struct file *filep, struct kfd_process *p,
 
 
 	/* Return gpu_id as doorbell offset for mmap usage */
-	args->doorbell_offset = (KFD_MMAP_DOORBELL_MASK | args->gpu_id);
+	args->doorbell_offset = KFD_MMAP_TYPE_DOORBELL;
+	args->doorbell_offset |= KFD_MMAP_GPU_ID(args->gpu_id);
 	args->doorbell_offset <<= PAGE_SHIFT;
 
 	mutex_unlock(&p->mutex);
@@ -1644,23 +1645,33 @@ err_i1:
 static int kfd_mmap(struct file *filp, struct vm_area_struct *vma)
 {
 	struct kfd_process *process;
+	struct kfd_dev *dev = NULL;
+	unsigned long vm_pgoff;
+	unsigned int gpu_id;
 
 	process = kfd_get_process(current);
 	if (IS_ERR(process))
 		return PTR_ERR(process);
 
-	if ((vma->vm_pgoff & KFD_MMAP_DOORBELL_MASK) ==
-			KFD_MMAP_DOORBELL_MASK) {
-		vma->vm_pgoff = vma->vm_pgoff ^ KFD_MMAP_DOORBELL_MASK;
-		return kfd_doorbell_mmap(process, vma);
-	} else if ((vma->vm_pgoff & KFD_MMAP_EVENTS_MASK) ==
-			KFD_MMAP_EVENTS_MASK) {
-		vma->vm_pgoff = vma->vm_pgoff ^ KFD_MMAP_EVENTS_MASK;
+	vm_pgoff = vma->vm_pgoff;
+	vma->vm_pgoff = KFD_MMAP_OFFSET_VALUE_GET(vm_pgoff);
+	gpu_id = KFD_MMAP_GPU_ID_GET(vm_pgoff);
+	if (gpu_id)
+		dev = kfd_device_by_id(gpu_id);
+
+	switch (vm_pgoff & KFD_MMAP_TYPE_MASK) {
+	case KFD_MMAP_TYPE_DOORBELL:
+		if (!dev)
+			return -ENODEV;
+		return kfd_doorbell_mmap(dev, process, vma);
+
+	case KFD_MMAP_TYPE_EVENTS:
 		return kfd_event_mmap(process, vma);
-	} else if ((vma->vm_pgoff & KFD_MMAP_RESERVED_MEM_MASK) ==
-			KFD_MMAP_RESERVED_MEM_MASK) {
-		vma->vm_pgoff = vma->vm_pgoff ^ KFD_MMAP_RESERVED_MEM_MASK;
-		return kfd_reserved_mem_mmap(process, vma);
+
+	case KFD_MMAP_TYPE_RESERVED_MEM:
+		if (!dev)
+			return -ENODEV;
+		return kfd_reserved_mem_mmap(dev, process, vma);
 	}
 
 	return -EFAULT;
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c b/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c
index 4840314..efc59de 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c
@@ -126,15 +126,10 @@ void kfd_doorbell_fini(struct kfd_dev *kfd)
 		iounmap(kfd->doorbell_kernel_ptr);
 }
 
-int kfd_doorbell_mmap(struct kfd_process *process, struct vm_area_struct *vma)
+int kfd_doorbell_mmap(struct kfd_dev *dev, struct kfd_process *process,
+		      struct vm_area_struct *vma)
 {
 	phys_addr_t address;
-	struct kfd_dev *dev;
-
-	/* Find kfd device according to gpu id */
-	dev = kfd_device_by_id(vma->vm_pgoff);
-	if (!dev)
-		return -EINVAL;
 
 	/*
 	 * For simplicitly we only allow mapping of the entire doorbell
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_events.c b/drivers/gpu/drm/amd/amdkfd/kfd_events.c
index 4890a90..bccf2f7 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_events.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_events.c
@@ -345,7 +345,7 @@ int kfd_event_create(struct file *devkfd, struct kfd_process *p,
 	case KFD_EVENT_TYPE_DEBUG:
 		ret = create_signal_event(devkfd, p, ev);
 		if (!ret) {
-			*event_page_offset = KFD_MMAP_EVENTS_MASK;
+			*event_page_offset = KFD_MMAP_TYPE_EVENTS;
 			*event_page_offset <<= PAGE_SHIFT;
 			*event_slot_index = ev->event_id;
 		}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
index d9c0fe12..2d575c0 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
@@ -41,9 +41,33 @@
 
 #define KFD_SYSFS_FILE_MODE 0444
 
-#define KFD_MMAP_DOORBELL_MASK 0x8000000000000ull
-#define KFD_MMAP_EVENTS_MASK 0x4000000000000ull
-#define KFD_MMAP_RESERVED_MEM_MASK 0x2000000000000ull
+/* GPU ID hash width in bits */
+#define KFD_GPU_ID_HASH_WIDTH 16
+
+/* Use upper bits of mmap offset to store KFD driver specific information.
+ * BITS[63:62] - Encode MMAP type
+ * BITS[61:46] - Encode gpu_id. To identify to which GPU the offset belongs to
+ * BITS[45:0]  - MMAP offset value
+ *
+ * NOTE: struct vm_area_struct.vm_pgoff uses offset in pages. Hence, these
+ *  defines are w.r.t to PAGE_SIZE
+ */
+#define KFD_MMAP_TYPE_SHIFT	(62 - PAGE_SHIFT)
+#define KFD_MMAP_TYPE_MASK	(0x3ULL << KFD_MMAP_TYPE_SHIFT)
+#define KFD_MMAP_TYPE_DOORBELL	(0x3ULL << KFD_MMAP_TYPE_SHIFT)
+#define KFD_MMAP_TYPE_EVENTS	(0x2ULL << KFD_MMAP_TYPE_SHIFT)
+#define KFD_MMAP_TYPE_RESERVED_MEM	(0x1ULL << KFD_MMAP_TYPE_SHIFT)
+
+#define KFD_MMAP_GPU_ID_SHIFT (46 - PAGE_SHIFT)
+#define KFD_MMAP_GPU_ID_MASK (((1ULL << KFD_GPU_ID_HASH_WIDTH) - 1) \
+				<< KFD_MMAP_GPU_ID_SHIFT)
+#define KFD_MMAP_GPU_ID(gpu_id) ((((uint64_t)gpu_id) << KFD_MMAP_GPU_ID_SHIFT)\
+				& KFD_MMAP_GPU_ID_MASK)
+#define KFD_MMAP_GPU_ID_GET(offset)    ((offset & KFD_MMAP_GPU_ID_MASK) \
+				>> KFD_MMAP_GPU_ID_SHIFT)
+
+#define KFD_MMAP_OFFSET_VALUE_MASK	(0x3FFFFFFFFFFFULL >> PAGE_SHIFT)
+#define KFD_MMAP_OFFSET_VALUE_GET(offset) (offset & KFD_MMAP_OFFSET_VALUE_MASK)
 
 /*
  * When working with cp scheduler we should assign the HIQ manually or via
@@ -55,9 +79,6 @@
 #define KFD_CIK_HIQ_PIPE 4
 #define KFD_CIK_HIQ_QUEUE 0
 
-/* GPU ID hash width in bits */
-#define KFD_GPU_ID_HASH_WIDTH 16
-
 /* Macro for allocating structures */
 #define kfd_alloc_struct(ptr_to_struct)	\
 	((typeof(ptr_to_struct)) kzalloc(sizeof(*ptr_to_struct), GFP_KERNEL))
@@ -698,7 +719,7 @@ struct kfd_process_device *kfd_get_process_device_data(struct kfd_dev *dev,
 struct kfd_process_device *kfd_create_process_device_data(struct kfd_dev *dev,
 							struct kfd_process *p);
 
-int kfd_reserved_mem_mmap(struct kfd_process *process,
+int kfd_reserved_mem_mmap(struct kfd_dev *dev, struct kfd_process *process,
 			  struct vm_area_struct *vma);
 
 /* KFD process API for creating and translating handles */
@@ -728,7 +749,8 @@ void kfd_pasid_free(unsigned int pasid);
 /* Doorbells */
 int kfd_doorbell_init(struct kfd_dev *kfd);
 void kfd_doorbell_fini(struct kfd_dev *kfd);
-int kfd_doorbell_mmap(struct kfd_process *process, struct vm_area_struct *vma);
+int kfd_doorbell_mmap(struct kfd_dev *dev, struct kfd_process *process,
+		      struct vm_area_struct *vma);
 void __iomem *kfd_get_kernel_doorbell(struct kfd_dev *kfd,
 					unsigned int *doorbell_off);
 void kfd_release_kernel_doorbell(struct kfd_dev *kfd, u32 __iomem *db_addr);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process.c b/drivers/gpu/drm/amd/amdkfd/kfd_process.c
index 2791e72..131fe2a 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_process.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c
@@ -451,7 +451,8 @@ static int kfd_process_init_cwsr_apu(struct kfd_process *p, struct file *filep)
 		if (!dev->cwsr_enabled || qpd->cwsr_kaddr || qpd->cwsr_base)
 			continue;
 
-		offset = (dev->id | KFD_MMAP_RESERVED_MEM_MASK) << PAGE_SHIFT;
+		offset = (KFD_MMAP_TYPE_RESERVED_MEM | KFD_MMAP_GPU_ID(dev->id))
+			<< PAGE_SHIFT;
 		qpd->tba_addr = (int64_t)vm_mmap(filep, 0,
 			KFD_CWSR_TBA_TMA_SIZE, PROT_READ | PROT_EXEC,
 			MAP_SHARED, offset);
@@ -989,15 +990,12 @@ int kfd_resume_all_processes(void)
 	return ret;
 }
 
-int kfd_reserved_mem_mmap(struct kfd_process *process,
+int kfd_reserved_mem_mmap(struct kfd_dev *dev, struct kfd_process *process,
 			  struct vm_area_struct *vma)
 {
-	struct kfd_dev *dev = kfd_device_by_id(vma->vm_pgoff);
 	struct kfd_process_device *pdd;
 	struct qcm_process_device *qpd;
 
-	if (!dev)
-		return -EINVAL;
 	if ((vma->vm_end - vma->vm_start) != KFD_CWSR_TBA_TMA_SIZE) {
 		pr_err("Incorrect CWSR mapping size.\n");
 		return -EINVAL;
-- 
cgit v1.1


From ef568db792e66216b48fd1567ff4a9d3bf9af866 Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Tue, 10 Apr 2018 17:33:05 -0400
Subject: drm/amdkfd: Implement doorbell allocation for SOC15

Allocate doorbells according to the doorbell routing information on
SOC15 ASICs (Vega10 and later). On older ASICs we continue to use the
queue_id as the doorbell ID to maintain compatibility with the Thunk.

Signed-off-by: Shaoyun Liu <Shaoyun.Liu@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/kfd_chardev.c           |  7 ++
 .../gpu/drm/amd/amdkfd/kfd_device_queue_manager.c  | 82 ++++++++++++++++++++--
 drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c          | 12 ++--
 drivers/gpu/drm/amd/amdkfd/kfd_priv.h              | 11 ++-
 drivers/gpu/drm/amd/amdkfd/kfd_process.c           | 32 +++++++++
 .../gpu/drm/amd/amdkfd/kfd_process_queue_manager.c | 12 +++-
 6 files changed, 139 insertions(+), 17 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
index 519c7b1..5694fbe 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
@@ -295,6 +295,13 @@ static int kfd_ioctl_create_queue(struct file *filep, struct kfd_process *p,
 	args->doorbell_offset = KFD_MMAP_TYPE_DOORBELL;
 	args->doorbell_offset |= KFD_MMAP_GPU_ID(args->gpu_id);
 	args->doorbell_offset <<= PAGE_SHIFT;
+	if (KFD_IS_SOC15(dev->device_info->asic_family))
+		/* On SOC15 ASICs, doorbell allocation must be
+		 * per-device, and independent from the per-process
+		 * queue_id. Return the doorbell offset within the
+		 * doorbell aperture to user mode.
+		 */
+		args->doorbell_offset |= q_properties.doorbell_off;
 
 	mutex_unlock(&p->mutex);
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
index d55d29d..e9c72d8 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
@@ -110,6 +110,57 @@ void program_sh_mem_settings(struct device_queue_manager *dqm,
 						qpd->sh_mem_bases);
 }
 
+static int allocate_doorbell(struct qcm_process_device *qpd, struct queue *q)
+{
+	struct kfd_dev *dev = qpd->dqm->dev;
+
+	if (!KFD_IS_SOC15(dev->device_info->asic_family)) {
+		/* On pre-SOC15 chips we need to use the queue ID to
+		 * preserve the user mode ABI.
+		 */
+		q->doorbell_id = q->properties.queue_id;
+	} else if (q->properties.type == KFD_QUEUE_TYPE_SDMA) {
+		/* For SDMA queues on SOC15, use static doorbell
+		 * assignments based on the engine and queue.
+		 */
+		q->doorbell_id = dev->shared_resources.sdma_doorbell
+			[q->properties.sdma_engine_id]
+			[q->properties.sdma_queue_id];
+	} else {
+		/* For CP queues on SOC15 reserve a free doorbell ID */
+		unsigned int found;
+
+		found = find_first_zero_bit(qpd->doorbell_bitmap,
+					    KFD_MAX_NUM_OF_QUEUES_PER_PROCESS);
+		if (found >= KFD_MAX_NUM_OF_QUEUES_PER_PROCESS) {
+			pr_debug("No doorbells available");
+			return -EBUSY;
+		}
+		set_bit(found, qpd->doorbell_bitmap);
+		q->doorbell_id = found;
+	}
+
+	q->properties.doorbell_off =
+		kfd_doorbell_id_to_offset(dev, q->process,
+					  q->doorbell_id);
+
+	return 0;
+}
+
+static void deallocate_doorbell(struct qcm_process_device *qpd,
+				struct queue *q)
+{
+	unsigned int old;
+	struct kfd_dev *dev = qpd->dqm->dev;
+
+	if (!KFD_IS_SOC15(dev->device_info->asic_family) ||
+	    q->properties.type == KFD_QUEUE_TYPE_SDMA)
+		return;
+
+	old = test_and_clear_bit(q->doorbell_id, qpd->doorbell_bitmap);
+	WARN_ON(!old);
+}
+
 static int allocate_vmid(struct device_queue_manager *dqm,
 			struct qcm_process_device *qpd,
 			struct queue *q)
@@ -301,10 +352,14 @@ static int create_compute_queue_nocpsch(struct device_queue_manager *dqm,
 	if (retval)
 		return retval;
 
+	retval = allocate_doorbell(qpd, q);
+	if (retval)
+		goto out_deallocate_hqd;
+
 	retval = mqd->init_mqd(mqd, &q->mqd, &q->mqd_mem_obj,
 				&q->gart_mqd_addr, &q->properties);
 	if (retval)
-		goto out_deallocate_hqd;
+		goto out_deallocate_doorbell;
 
 	pr_debug("Loading mqd to hqd on pipe %d, queue %d\n",
 			q->pipe, q->queue);
@@ -324,6 +379,8 @@ static int create_compute_queue_nocpsch(struct device_queue_manager *dqm,
 
 out_uninit_mqd:
 	mqd->uninit_mqd(mqd, q->mqd, q->mqd_mem_obj);
+out_deallocate_doorbell:
+	deallocate_doorbell(qpd, q);
 out_deallocate_hqd:
 	deallocate_hqd(dqm, q);
 
@@ -357,6 +414,8 @@ static int destroy_queue_nocpsch_locked(struct device_queue_manager *dqm,
 	}
 	dqm->total_queue_count--;
 
+	deallocate_doorbell(qpd, q);
+
 	retval = mqd->destroy_mqd(mqd, q->mqd,
 				KFD_PREEMPT_TYPE_WAVEFRONT_RESET,
 				KFD_UNMAP_LATENCY_MS,
@@ -861,6 +920,10 @@ static int create_sdma_queue_nocpsch(struct device_queue_manager *dqm,
 	q->properties.sdma_queue_id = q->sdma_id / CIK_SDMA_QUEUES_PER_ENGINE;
 	q->properties.sdma_engine_id = q->sdma_id % CIK_SDMA_QUEUES_PER_ENGINE;
 
+	retval = allocate_doorbell(qpd, q);
+	if (retval)
+		goto out_deallocate_sdma_queue;
+
 	pr_debug("SDMA id is:    %d\n", q->sdma_id);
 	pr_debug("SDMA queue id: %d\n", q->properties.sdma_queue_id);
 	pr_debug("SDMA engine id: %d\n", q->properties.sdma_engine_id);
@@ -869,7 +932,7 @@ static int create_sdma_queue_nocpsch(struct device_queue_manager *dqm,
 	retval = mqd->init_mqd(mqd, &q->mqd, &q->mqd_mem_obj,
 				&q->gart_mqd_addr, &q->properties);
 	if (retval)
-		goto out_deallocate_sdma_queue;
+		goto out_deallocate_doorbell;
 
 	retval = mqd->load_mqd(mqd, q->mqd, 0, 0, &q->properties, NULL);
 	if (retval)
@@ -879,6 +942,8 @@ static int create_sdma_queue_nocpsch(struct device_queue_manager *dqm,
 
 out_uninit_mqd:
 	mqd->uninit_mqd(mqd, q->mqd, q->mqd_mem_obj);
+out_deallocate_doorbell:
+	deallocate_doorbell(qpd, q);
 out_deallocate_sdma_queue:
 	deallocate_sdma_queue(dqm, q->sdma_id);
 
@@ -1070,12 +1135,17 @@ static int create_queue_cpsch(struct device_queue_manager *dqm, struct queue *q,
 		q->properties.sdma_engine_id =
 			q->sdma_id % CIK_SDMA_QUEUES_PER_ENGINE;
 	}
+
+	retval = allocate_doorbell(qpd, q);
+	if (retval)
+		goto out_deallocate_sdma_queue;
+
 	mqd = dqm->ops.get_mqd_manager(dqm,
 			get_mqd_type_from_queue_type(q->properties.type));
 
 	if (!mqd) {
 		retval = -ENOMEM;
-		goto out_deallocate_sdma_queue;
+		goto out_deallocate_doorbell;
 	}
 	/*
 	 * Eviction state logic: we only mark active queues as evicted
@@ -1093,7 +1163,7 @@ static int create_queue_cpsch(struct device_queue_manager *dqm, struct queue *q,
 	retval = mqd->init_mqd(mqd, &q->mqd, &q->mqd_mem_obj,
 				&q->gart_mqd_addr, &q->properties);
 	if (retval)
-		goto out_deallocate_sdma_queue;
+		goto out_deallocate_doorbell;
 
 	list_add(&q->list, &qpd->queues_list);
 	qpd->queue_count++;
@@ -1117,6 +1187,8 @@ static int create_queue_cpsch(struct device_queue_manager *dqm, struct queue *q,
 	mutex_unlock(&dqm->lock);
 	return retval;
 
+out_deallocate_doorbell:
+	deallocate_doorbell(qpd, q);
 out_deallocate_sdma_queue:
 	if (q->properties.type == KFD_QUEUE_TYPE_SDMA)
 		deallocate_sdma_queue(dqm, q->sdma_id);
@@ -1257,6 +1329,8 @@ static int destroy_queue_cpsch(struct device_queue_manager *dqm,
 		goto failed;
 	}
 
+	deallocate_doorbell(qpd, q);
+
 	if (q->properties.type == KFD_QUEUE_TYPE_SDMA) {
 		dqm->sdma_queue_count--;
 		deallocate_sdma_queue(dqm, q->sdma_id);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c b/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c
index efc59de..36c9269e 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c
@@ -49,7 +49,7 @@ static unsigned int max_doorbell_slices;
  */
 
 /* # of doorbell bytes allocated for each process. */
-static size_t kfd_doorbell_process_slice(struct kfd_dev *kfd)
+size_t kfd_doorbell_process_slice(struct kfd_dev *kfd)
 {
 	return roundup(kfd->device_info->doorbell_size *
 			KFD_MAX_NUM_OF_QUEUES_PER_PROCESS,
@@ -214,13 +214,9 @@ void write_kernel_doorbell(void __iomem *db, u32 value)
 	}
 }
 
-/*
- * queue_ids are in the range [0,MAX_PROCESS_QUEUES) and are mapped 1:1
- * to doorbells with the process's doorbell page
- */
-unsigned int kfd_queue_id_to_doorbell(struct kfd_dev *kfd,
+unsigned int kfd_doorbell_id_to_offset(struct kfd_dev *kfd,
 					struct kfd_process *process,
-					unsigned int queue_id)
+					unsigned int doorbell_id)
 {
 	/*
 	 * doorbell_id_offset accounts for doorbells taken by KGD.
@@ -231,7 +227,7 @@ unsigned int kfd_queue_id_to_doorbell(struct kfd_dev *kfd,
 	return kfd->doorbell_id_offset +
 		process->doorbell_index
 		* kfd_doorbell_process_slice(kfd) / sizeof(u32) +
-		queue_id * kfd->device_info->doorbell_size / sizeof(u32);
+		doorbell_id * kfd->device_info->doorbell_size / sizeof(u32);
 }
 
 uint64_t kfd_get_number_elems(struct kfd_dev *kfd)
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
index 2d575c0..ddb3c8c 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
@@ -169,6 +169,8 @@ enum cache_policy {
 	cache_policy_noncoherent
 };
 
+#define KFD_IS_SOC15(chip) ((chip) >= CHIP_VEGA10)
+
 struct kfd_event_interrupt_class {
 	bool (*interrupt_isr)(struct kfd_dev *dev,
 				const uint32_t *ih_ring_entry);
@@ -449,6 +451,7 @@ struct queue {
 	uint32_t queue;
 
 	unsigned int sdma_id;
+	unsigned int doorbell_id;
 
 	struct kfd_process	*process;
 	struct kfd_dev		*device;
@@ -523,6 +526,9 @@ struct qcm_process_device {
 	/* IB memory */
 	uint64_t ib_base;
 	void *ib_kaddr;
+
+	/* doorbell resources per process per device */
+	unsigned long *doorbell_bitmap;
 };
 
 /* KFD Memory Eviction */
@@ -747,6 +753,7 @@ unsigned int kfd_pasid_alloc(void);
 void kfd_pasid_free(unsigned int pasid);
 
 /* Doorbells */
+size_t kfd_doorbell_process_slice(struct kfd_dev *kfd);
 int kfd_doorbell_init(struct kfd_dev *kfd);
 void kfd_doorbell_fini(struct kfd_dev *kfd);
 int kfd_doorbell_mmap(struct kfd_dev *dev, struct kfd_process *process,
@@ -756,9 +763,9 @@ void __iomem *kfd_get_kernel_doorbell(struct kfd_dev *kfd,
 void kfd_release_kernel_doorbell(struct kfd_dev *kfd, u32 __iomem *db_addr);
 u32 read_kernel_doorbell(u32 __iomem *db);
 void write_kernel_doorbell(void __iomem *db, u32 value);
-unsigned int kfd_queue_id_to_doorbell(struct kfd_dev *kfd,
+unsigned int kfd_doorbell_id_to_offset(struct kfd_dev *kfd,
 					struct kfd_process *process,
-					unsigned int queue_id);
+					unsigned int doorbell_id);
 phys_addr_t kfd_get_process_doorbells(struct kfd_dev *dev,
 					struct kfd_process *process);
 int kfd_alloc_process_doorbells(struct kfd_process *process);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process.c b/drivers/gpu/drm/amd/amdkfd/kfd_process.c
index 131fe2a..1d80b4f 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_process.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c
@@ -332,6 +332,7 @@ static void kfd_process_destroy_pdds(struct kfd_process *p)
 			free_pages((unsigned long)pdd->qpd.cwsr_kaddr,
 				get_order(KFD_CWSR_TBA_TMA_SIZE));
 
+		kfree(pdd->qpd.doorbell_bitmap);
 		idr_destroy(&pdd->alloc_idr);
 
 		kfree(pdd);
@@ -586,6 +587,31 @@ err_alloc_process:
 	return ERR_PTR(err);
 }
 
+static int init_doorbell_bitmap(struct qcm_process_device *qpd,
+			struct kfd_dev *dev)
+{
+	unsigned int i;
+
+	if (!KFD_IS_SOC15(dev->device_info->asic_family))
+		return 0;
+
+	qpd->doorbell_bitmap =
+		kzalloc(DIV_ROUND_UP(KFD_MAX_NUM_OF_QUEUES_PER_PROCESS,
+				     BITS_PER_BYTE), GFP_KERNEL);
+	if (!qpd->doorbell_bitmap)
+		return -ENOMEM;
+
+	/* Mask out any reserved doorbells */
+	for (i = 0; i < KFD_MAX_NUM_OF_QUEUES_PER_PROCESS; i++)
+		if ((dev->shared_resources.reserved_doorbell_mask & i) ==
+		    dev->shared_resources.reserved_doorbell_val) {
+			set_bit(i, qpd->doorbell_bitmap);
+			pr_debug("reserved doorbell 0x%03x\n", i);
+		}
+
+	return 0;
+}
+
 struct kfd_process_device *kfd_get_process_device_data(struct kfd_dev *dev,
 							struct kfd_process *p)
 {
@@ -607,6 +633,12 @@ struct kfd_process_device *kfd_create_process_device_data(struct kfd_dev *dev,
 	if (!pdd)
 		return NULL;
 
+	if (init_doorbell_bitmap(&pdd->qpd, dev)) {
+		pr_err("Failed to init doorbell for process\n");
+		kfree(pdd);
+		return NULL;
+	}
+
 	pdd->dev = dev;
 	INIT_LIST_HEAD(&pdd->qpd.queues_list);
 	INIT_LIST_HEAD(&pdd->qpd.priv_queue_list);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
index 7817e32..3045aeb 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
@@ -119,9 +119,6 @@ static int create_cp_queue(struct process_queue_manager *pqm,
 	/* Doorbell initialized in user space*/
 	q_properties->doorbell_ptr = NULL;
 
-	q_properties->doorbell_off =
-			kfd_queue_id_to_doorbell(dev, pqm->process, qid);
-
 	/* let DQM handle it*/
 	q_properties->vmid = 0;
 	q_properties->queue_id = qid;
@@ -248,6 +245,15 @@ int pqm_create_queue(struct process_queue_manager *pqm,
 		goto err_create_queue;
 	}
 
+	if (q)
+		/* Return the doorbell offset within the doorbell page
+		 * to the caller so it can be passed up to user mode
+		 * (in bytes).
+		 */
+		properties->doorbell_off =
+			(q->properties.doorbell_off * sizeof(uint32_t)) &
+			(kfd_doorbell_process_slice(dev) - 1);
+
 	pr_debug("PQM After DQM create queue\n");
 
 	list_add(&pqn->process_queue_list, &pqm->queues);
-- 
cgit v1.1


From f6e27ff19d9db90e55576dea5aef98feb3d0ce5e Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Tue, 10 Apr 2018 17:33:06 -0400
Subject: drm/amdkfd: Move packet writer functions into ASIC-specific file

This is in preparation for GFXv9 (Vega10) which uses incompatible PM4
packet formats from previous ASIC generations.

Signed-off-by: Shaoyun Liu <Shaoyun.Liu@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 .../gpu/drm/amd/amdkfd/kfd_device_queue_manager.c  |  10 +-
 drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_vi.c   | 310 +++++++++++++++++
 drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c    | 381 ++++-----------------
 drivers/gpu/drm/amd/amdkfd/kfd_priv.h              |  35 +-
 4 files changed, 420 insertions(+), 316 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
index e9c72d8..500f022 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
@@ -196,15 +196,19 @@ static int allocate_vmid(struct device_queue_manager *dqm,
 static int flush_texture_cache_nocpsch(struct kfd_dev *kdev,
 				struct qcm_process_device *qpd)
 {
-	uint32_t len;
+	const struct packet_manager_funcs *pmf = qpd->dqm->packets.pmf;
+	int ret;
 
 	if (!qpd->ib_kaddr)
 		return -ENOMEM;
 
-	len = pm_create_release_mem(qpd->ib_base, (uint32_t *)qpd->ib_kaddr);
+	ret = pmf->release_mem(qpd->ib_base, (uint32_t *)qpd->ib_kaddr);
+	if (ret)
+		return ret;
 
 	return kdev->kfd2kgd->submit_ib(kdev->kgd, KGD_ENGINE_MEC1, qpd->vmid,
-				qpd->ib_base, (uint32_t *)qpd->ib_kaddr, len);
+				qpd->ib_base, (uint32_t *)qpd->ib_kaddr,
+				pmf->release_mem_size / sizeof(uint32_t));
 }
 
 static void deallocate_vmid(struct device_queue_manager *dqm,
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_vi.c b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_vi.c
index f1d4828..7ee326f 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_vi.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_vi.c
@@ -22,6 +22,9 @@
  */
 
 #include "kfd_kernel_queue.h"
+#include "kfd_device_queue_manager.h"
+#include "kfd_pm4_headers_vi.h"
+#include "kfd_pm4_opcodes.h"
 
 static bool initialize_vi(struct kernel_queue *kq, struct kfd_dev *dev,
 			enum kfd_queue_type type, unsigned int queue_size);
@@ -54,3 +57,310 @@ static void uninitialize_vi(struct kernel_queue *kq)
 {
 	kfd_gtt_sa_free(kq->dev, kq->eop_mem);
 }
+
+static unsigned int build_pm4_header(unsigned int opcode, size_t packet_size)
+{
+	union PM4_MES_TYPE_3_HEADER header;
+
+	header.u32All = 0;
+	header.opcode = opcode;
+	header.count = packet_size / 4 - 2;
+	header.type = PM4_TYPE_3;
+
+	return header.u32All;
+}
+
+static int pm_map_process_vi(struct packet_manager *pm, uint32_t *buffer,
+				struct qcm_process_device *qpd)
+{
+	struct pm4_mes_map_process *packet;
+
+	packet = (struct pm4_mes_map_process *)buffer;
+
+	memset(buffer, 0, sizeof(struct pm4_mes_map_process));
+
+	packet->header.u32All = build_pm4_header(IT_MAP_PROCESS,
+					sizeof(struct pm4_mes_map_process));
+	packet->bitfields2.diq_enable = (qpd->is_debug) ? 1 : 0;
+	packet->bitfields2.process_quantum = 1;
+	packet->bitfields2.pasid = qpd->pqm->process->pasid;
+	packet->bitfields3.page_table_base = qpd->page_table_base;
+	packet->bitfields10.gds_size = qpd->gds_size;
+	packet->bitfields10.num_gws = qpd->num_gws;
+	packet->bitfields10.num_oac = qpd->num_oac;
+	packet->bitfields10.num_queues = (qpd->is_debug) ? 0 : qpd->queue_count;
+
+	packet->sh_mem_config = qpd->sh_mem_config;
+	packet->sh_mem_bases = qpd->sh_mem_bases;
+	packet->sh_mem_ape1_base = qpd->sh_mem_ape1_base;
+	packet->sh_mem_ape1_limit = qpd->sh_mem_ape1_limit;
+
+	packet->sh_hidden_private_base_vmid = qpd->sh_hidden_private_base;
+
+	packet->gds_addr_lo = lower_32_bits(qpd->gds_context_area);
+	packet->gds_addr_hi = upper_32_bits(qpd->gds_context_area);
+
+	return 0;
+}
+
+static int pm_runlist_vi(struct packet_manager *pm, uint32_t *buffer,
+			uint64_t ib, size_t ib_size_in_dwords, bool chain)
+{
+	struct pm4_mes_runlist *packet;
+	int concurrent_proc_cnt = 0;
+	struct kfd_dev *kfd = pm->dqm->dev;
+
+	if (WARN_ON(!ib))
+		return -EFAULT;
+
+	/* Determine the number of processes to map together to HW:
+	 * it can not exceed the number of VMIDs available to the
+	 * scheduler, and it is determined by the smaller of the number
+	 * of processes in the runlist and kfd module parameter
+	 * hws_max_conc_proc.
+	 * Note: the arbitration between the number of VMIDs and
+	 * hws_max_conc_proc has been done in
+	 * kgd2kfd_device_init().
+	 */
+	concurrent_proc_cnt = min(pm->dqm->processes_count,
+			kfd->max_proc_per_quantum);
+
+	packet = (struct pm4_mes_runlist *)buffer;
+
+	memset(buffer, 0, sizeof(struct pm4_mes_runlist));
+	packet->header.u32All = build_pm4_header(IT_RUN_LIST,
+						sizeof(struct pm4_mes_runlist));
+
+	packet->bitfields4.ib_size = ib_size_in_dwords;
+	packet->bitfields4.chain = chain ? 1 : 0;
+	packet->bitfields4.offload_polling = 0;
+	packet->bitfields4.valid = 1;
+	packet->bitfields4.process_cnt = concurrent_proc_cnt;
+	packet->ordinal2 = lower_32_bits(ib);
+	packet->bitfields3.ib_base_hi = upper_32_bits(ib);
+
+	return 0;
+}
+
+static int pm_set_resources_vi(struct packet_manager *pm, uint32_t *buffer,
+				struct scheduling_resources *res)
+{
+	struct pm4_mes_set_resources *packet;
+
+	packet = (struct pm4_mes_set_resources *)buffer;
+	memset(buffer, 0, sizeof(struct pm4_mes_set_resources));
+
+	packet->header.u32All = build_pm4_header(IT_SET_RESOURCES,
+					sizeof(struct pm4_mes_set_resources));
+
+	packet->bitfields2.queue_type =
+			queue_type__mes_set_resources__hsa_interface_queue_hiq;
+	packet->bitfields2.vmid_mask = res->vmid_mask;
+	packet->bitfields2.unmap_latency = KFD_UNMAP_LATENCY_MS / 100;
+	packet->bitfields7.oac_mask = res->oac_mask;
+	packet->bitfields8.gds_heap_base = res->gds_heap_base;
+	packet->bitfields8.gds_heap_size = res->gds_heap_size;
+
+	packet->gws_mask_lo = lower_32_bits(res->gws_mask);
+	packet->gws_mask_hi = upper_32_bits(res->gws_mask);
+
+	packet->queue_mask_lo = lower_32_bits(res->queue_mask);
+	packet->queue_mask_hi = upper_32_bits(res->queue_mask);
+
+	return 0;
+}
+
+static int pm_map_queues_vi(struct packet_manager *pm, uint32_t *buffer,
+		struct queue *q, bool is_static)
+{
+	struct pm4_mes_map_queues *packet;
+	bool use_static = is_static;
+
+	packet = (struct pm4_mes_map_queues *)buffer;
+	memset(buffer, 0, sizeof(struct pm4_mes_map_queues));
+
+	packet->header.u32All = build_pm4_header(IT_MAP_QUEUES,
+					sizeof(struct pm4_mes_map_queues));
+	packet->bitfields2.alloc_format =
+		alloc_format__mes_map_queues__one_per_pipe_vi;
+	packet->bitfields2.num_queues = 1;
+	packet->bitfields2.queue_sel =
+		queue_sel__mes_map_queues__map_to_hws_determined_queue_slots_vi;
+
+	packet->bitfields2.engine_sel =
+		engine_sel__mes_map_queues__compute_vi;
+	packet->bitfields2.queue_type =
+		queue_type__mes_map_queues__normal_compute_vi;
+
+	switch (q->properties.type) {
+	case KFD_QUEUE_TYPE_COMPUTE:
+		if (use_static)
+			packet->bitfields2.queue_type =
+		queue_type__mes_map_queues__normal_latency_static_queue_vi;
+		break;
+	case KFD_QUEUE_TYPE_DIQ:
+		packet->bitfields2.queue_type =
+			queue_type__mes_map_queues__debug_interface_queue_vi;
+		break;
+	case KFD_QUEUE_TYPE_SDMA:
+		packet->bitfields2.engine_sel = q->properties.sdma_engine_id +
+				engine_sel__mes_map_queues__sdma0_vi;
+		use_static = false; /* no static queues under SDMA */
+		break;
+	default:
+		WARN(1, "queue type %d", q->properties.type);
+		return -EINVAL;
+	}
+	packet->bitfields3.doorbell_offset =
+			q->properties.doorbell_off;
+
+	packet->mqd_addr_lo =
+			lower_32_bits(q->gart_mqd_addr);
+
+	packet->mqd_addr_hi =
+			upper_32_bits(q->gart_mqd_addr);
+
+	packet->wptr_addr_lo =
+			lower_32_bits((uint64_t)q->properties.write_ptr);
+
+	packet->wptr_addr_hi =
+			upper_32_bits((uint64_t)q->properties.write_ptr);
+
+	return 0;
+}
+
+static int pm_unmap_queues_vi(struct packet_manager *pm, uint32_t *buffer,
+			enum kfd_queue_type type,
+			enum kfd_unmap_queues_filter filter,
+			uint32_t filter_param, bool reset,
+			unsigned int sdma_engine)
+{
+	struct pm4_mes_unmap_queues *packet;
+
+	packet = (struct pm4_mes_unmap_queues *)buffer;
+	memset(buffer, 0, sizeof(struct pm4_mes_unmap_queues));
+
+	packet->header.u32All = build_pm4_header(IT_UNMAP_QUEUES,
+					sizeof(struct pm4_mes_unmap_queues));
+	switch (type) {
+	case KFD_QUEUE_TYPE_COMPUTE:
+	case KFD_QUEUE_TYPE_DIQ:
+		packet->bitfields2.engine_sel =
+			engine_sel__mes_unmap_queues__compute;
+		break;
+	case KFD_QUEUE_TYPE_SDMA:
+		packet->bitfields2.engine_sel =
+			engine_sel__mes_unmap_queues__sdma0 + sdma_engine;
+		break;
+	default:
+		WARN(1, "queue type %d", type);
+		return -EINVAL;
+	}
+
+	if (reset)
+		packet->bitfields2.action =
+			action__mes_unmap_queues__reset_queues;
+	else
+		packet->bitfields2.action =
+			action__mes_unmap_queues__preempt_queues;
+
+	switch (filter) {
+	case KFD_UNMAP_QUEUES_FILTER_SINGLE_QUEUE:
+		packet->bitfields2.queue_sel =
+			queue_sel__mes_unmap_queues__perform_request_on_specified_queues;
+		packet->bitfields2.num_queues = 1;
+		packet->bitfields3b.doorbell_offset0 = filter_param;
+		break;
+	case KFD_UNMAP_QUEUES_FILTER_BY_PASID:
+		packet->bitfields2.queue_sel =
+			queue_sel__mes_unmap_queues__perform_request_on_pasid_queues;
+		packet->bitfields3a.pasid = filter_param;
+		break;
+	case KFD_UNMAP_QUEUES_FILTER_ALL_QUEUES:
+		packet->bitfields2.queue_sel =
+			queue_sel__mes_unmap_queues__unmap_all_queues;
+		break;
+	case KFD_UNMAP_QUEUES_FILTER_DYNAMIC_QUEUES:
+		/* in this case, we do not preempt static queues */
+		packet->bitfields2.queue_sel =
+			queue_sel__mes_unmap_queues__unmap_all_non_static_queues;
+		break;
+	default:
+		WARN(1, "filter %d", filter);
+		return -EINVAL;
+	}
+
+	return 0;
+
+}
+
+static int pm_query_status_vi(struct packet_manager *pm, uint32_t *buffer,
+			uint64_t fence_address,	uint32_t fence_value)
+{
+	struct pm4_mes_query_status *packet;
+
+	packet = (struct pm4_mes_query_status *)buffer;
+	memset(buffer, 0, sizeof(struct pm4_mes_query_status));
+
+	packet->header.u32All = build_pm4_header(IT_QUERY_STATUS,
+					sizeof(struct pm4_mes_query_status));
+
+	packet->bitfields2.context_id = 0;
+	packet->bitfields2.interrupt_sel =
+			interrupt_sel__mes_query_status__completion_status;
+	packet->bitfields2.command =
+			command__mes_query_status__fence_only_after_write_ack;
+
+	packet->addr_hi = upper_32_bits((uint64_t)fence_address);
+	packet->addr_lo = lower_32_bits((uint64_t)fence_address);
+	packet->data_hi = upper_32_bits((uint64_t)fence_value);
+	packet->data_lo = lower_32_bits((uint64_t)fence_value);
+
+	return 0;
+}
+
+static int pm_release_mem_vi(uint64_t gpu_addr, uint32_t *buffer)
+{
+	struct pm4_mec_release_mem *packet;
+
+	packet = (struct pm4_mec_release_mem *)buffer;
+	memset(buffer, 0, sizeof(*packet));
+
+	packet->header.u32All = build_pm4_header(IT_RELEASE_MEM,
+						 sizeof(*packet));
+
+	packet->bitfields2.event_type = CACHE_FLUSH_AND_INV_TS_EVENT;
+	packet->bitfields2.event_index = event_index___release_mem__end_of_pipe;
+	packet->bitfields2.tcl1_action_ena = 1;
+	packet->bitfields2.tc_action_ena = 1;
+	packet->bitfields2.cache_policy = cache_policy___release_mem__lru;
+	packet->bitfields2.atc = 0;
+
+	packet->bitfields3.data_sel = data_sel___release_mem__send_32_bit_low;
+	packet->bitfields3.int_sel =
+		int_sel___release_mem__send_interrupt_after_write_confirm;
+
+	packet->bitfields4.address_lo_32b = (gpu_addr & 0xffffffff) >> 2;
+	packet->address_hi = upper_32_bits(gpu_addr);
+
+	packet->data_lo = 0;
+
+	return 0;
+}
+
+const struct packet_manager_funcs kfd_vi_pm_funcs = {
+	.map_process		= pm_map_process_vi,
+	.runlist		= pm_runlist_vi,
+	.set_resources		= pm_set_resources_vi,
+	.map_queues		= pm_map_queues_vi,
+	.unmap_queues		= pm_unmap_queues_vi,
+	.query_status		= pm_query_status_vi,
+	.release_mem		= pm_release_mem_vi,
+	.map_process_size	= sizeof(struct pm4_mes_map_process),
+	.runlist_size		= sizeof(struct pm4_mes_runlist),
+	.set_resources_size	= sizeof(struct pm4_mes_set_resources),
+	.map_queues_size	= sizeof(struct pm4_mes_map_queues),
+	.unmap_queues_size	= sizeof(struct pm4_mes_unmap_queues),
+	.query_status_size	= sizeof(struct pm4_mes_query_status),
+	.release_mem_size	= sizeof(struct pm4_mec_release_mem)
+};
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c
index 89ba4c6..860ff24 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c
@@ -26,8 +26,6 @@
 #include "kfd_device_queue_manager.h"
 #include "kfd_kernel_queue.h"
 #include "kfd_priv.h"
-#include "kfd_pm4_headers_vi.h"
-#include "kfd_pm4_opcodes.h"
 
 static inline void inc_wptr(unsigned int *wptr, unsigned int increment_bytes,
 				unsigned int buffer_size_bytes)
@@ -39,18 +37,6 @@ static inline void inc_wptr(unsigned int *wptr, unsigned int increment_bytes,
 	*wptr = temp;
 }
 
-static unsigned int build_pm4_header(unsigned int opcode, size_t packet_size)
-{
-	union PM4_MES_TYPE_3_HEADER header;
-
-	header.u32All = 0;
-	header.opcode = opcode;
-	header.count = packet_size / 4 - 2;
-	header.type = PM4_TYPE_3;
-
-	return header.u32All;
-}
-
 static void pm_calc_rlib_size(struct packet_manager *pm,
 				unsigned int *rlib_size,
 				bool *over_subscription)
@@ -80,9 +66,9 @@ static void pm_calc_rlib_size(struct packet_manager *pm,
 		pr_debug("Over subscribed runlist\n");
 	}
 
-	map_queue_size = sizeof(struct pm4_mes_map_queues);
+	map_queue_size = pm->pmf->map_queues_size;
 	/* calculate run list ib allocation size */
-	*rlib_size = process_count * sizeof(struct pm4_mes_map_process) +
+	*rlib_size = process_count * pm->pmf->map_process_size +
 		     queue_count * map_queue_size;
 
 	/*
@@ -90,7 +76,7 @@ static void pm_calc_rlib_size(struct packet_manager *pm,
 	 * when over subscription
 	 */
 	if (*over_subscription)
-		*rlib_size += sizeof(struct pm4_mes_runlist);
+		*rlib_size += pm->pmf->runlist_size;
 
 	pr_debug("runlist ib size %d\n", *rlib_size);
 }
@@ -124,137 +110,6 @@ static int pm_allocate_runlist_ib(struct packet_manager *pm,
 	return retval;
 }
 
-static int pm_create_runlist(struct packet_manager *pm, uint32_t *buffer,
-			uint64_t ib, size_t ib_size_in_dwords, bool chain)
-{
-	struct pm4_mes_runlist *packet;
-	int concurrent_proc_cnt = 0;
-	struct kfd_dev *kfd = pm->dqm->dev;
-
-	if (WARN_ON(!ib))
-		return -EFAULT;
-
-	/* Determine the number of processes to map together to HW:
-	 * it can not exceed the number of VMIDs available to the
-	 * scheduler, and it is determined by the smaller of the number
-	 * of processes in the runlist and kfd module parameter
-	 * hws_max_conc_proc.
-	 * Note: the arbitration between the number of VMIDs and
-	 * hws_max_conc_proc has been done in
-	 * kgd2kfd_device_init().
-	 */
-	concurrent_proc_cnt = min(pm->dqm->processes_count,
-			kfd->max_proc_per_quantum);
-
-	packet = (struct pm4_mes_runlist *)buffer;
-
-	memset(buffer, 0, sizeof(struct pm4_mes_runlist));
-	packet->header.u32All = build_pm4_header(IT_RUN_LIST,
-						sizeof(struct pm4_mes_runlist));
-
-	packet->bitfields4.ib_size = ib_size_in_dwords;
-	packet->bitfields4.chain = chain ? 1 : 0;
-	packet->bitfields4.offload_polling = 0;
-	packet->bitfields4.valid = 1;
-	packet->bitfields4.process_cnt = concurrent_proc_cnt;
-	packet->ordinal2 = lower_32_bits(ib);
-	packet->bitfields3.ib_base_hi = upper_32_bits(ib);
-
-	return 0;
-}
-
-static int pm_create_map_process(struct packet_manager *pm, uint32_t *buffer,
-				struct qcm_process_device *qpd)
-{
-	struct pm4_mes_map_process *packet;
-
-	packet = (struct pm4_mes_map_process *)buffer;
-
-	memset(buffer, 0, sizeof(struct pm4_mes_map_process));
-
-	packet->header.u32All = build_pm4_header(IT_MAP_PROCESS,
-					sizeof(struct pm4_mes_map_process));
-	packet->bitfields2.diq_enable = (qpd->is_debug) ? 1 : 0;
-	packet->bitfields2.process_quantum = 1;
-	packet->bitfields2.pasid = qpd->pqm->process->pasid;
-	packet->bitfields3.page_table_base = qpd->page_table_base;
-	packet->bitfields10.gds_size = qpd->gds_size;
-	packet->bitfields10.num_gws = qpd->num_gws;
-	packet->bitfields10.num_oac = qpd->num_oac;
-	packet->bitfields10.num_queues = (qpd->is_debug) ? 0 : qpd->queue_count;
-
-	packet->sh_mem_config = qpd->sh_mem_config;
-	packet->sh_mem_bases = qpd->sh_mem_bases;
-	packet->sh_mem_ape1_base = qpd->sh_mem_ape1_base;
-	packet->sh_mem_ape1_limit = qpd->sh_mem_ape1_limit;
-
-	packet->sh_hidden_private_base_vmid = qpd->sh_hidden_private_base;
-
-	packet->gds_addr_lo = lower_32_bits(qpd->gds_context_area);
-	packet->gds_addr_hi = upper_32_bits(qpd->gds_context_area);
-
-	return 0;
-}
-
-static int pm_create_map_queue(struct packet_manager *pm, uint32_t *buffer,
-		struct queue *q, bool is_static)
-{
-	struct pm4_mes_map_queues *packet;
-	bool use_static = is_static;
-
-	packet = (struct pm4_mes_map_queues *)buffer;
-	memset(buffer, 0, sizeof(struct pm4_mes_map_queues));
-
-	packet->header.u32All = build_pm4_header(IT_MAP_QUEUES,
-						sizeof(struct pm4_mes_map_queues));
-	packet->bitfields2.alloc_format =
-		alloc_format__mes_map_queues__one_per_pipe_vi;
-	packet->bitfields2.num_queues = 1;
-	packet->bitfields2.queue_sel =
-		queue_sel__mes_map_queues__map_to_hws_determined_queue_slots_vi;
-
-	packet->bitfields2.engine_sel =
-		engine_sel__mes_map_queues__compute_vi;
-	packet->bitfields2.queue_type =
-		queue_type__mes_map_queues__normal_compute_vi;
-
-	switch (q->properties.type) {
-	case KFD_QUEUE_TYPE_COMPUTE:
-		if (use_static)
-			packet->bitfields2.queue_type =
-		queue_type__mes_map_queues__normal_latency_static_queue_vi;
-		break;
-	case KFD_QUEUE_TYPE_DIQ:
-		packet->bitfields2.queue_type =
-			queue_type__mes_map_queues__debug_interface_queue_vi;
-		break;
-	case KFD_QUEUE_TYPE_SDMA:
-		packet->bitfields2.engine_sel = q->properties.sdma_engine_id +
-				engine_sel__mes_map_queues__sdma0_vi;
-		use_static = false; /* no static queues under SDMA */
-		break;
-	default:
-		WARN(1, "queue type %d", q->properties.type);
-		return -EINVAL;
-	}
-	packet->bitfields3.doorbell_offset =
-			q->properties.doorbell_off;
-
-	packet->mqd_addr_lo =
-			lower_32_bits(q->gart_mqd_addr);
-
-	packet->mqd_addr_hi =
-			upper_32_bits(q->gart_mqd_addr);
-
-	packet->wptr_addr_lo =
-			lower_32_bits((uint64_t)q->properties.write_ptr);
-
-	packet->wptr_addr_hi =
-			upper_32_bits((uint64_t)q->properties.write_ptr);
-
-	return 0;
-}
-
 static int pm_create_runlist_ib(struct packet_manager *pm,
 				struct list_head *queues,
 				uint64_t *rl_gpu_addr,
@@ -292,12 +147,12 @@ static int pm_create_runlist_ib(struct packet_manager *pm,
 			return -ENOMEM;
 		}
 
-		retval = pm_create_map_process(pm, &rl_buffer[rl_wptr], qpd);
+		retval = pm->pmf->map_process(pm, &rl_buffer[rl_wptr], qpd);
 		if (retval)
 			return retval;
 
 		proccesses_mapped++;
-		inc_wptr(&rl_wptr, sizeof(struct pm4_mes_map_process),
+		inc_wptr(&rl_wptr, pm->pmf->map_process_size,
 				alloc_size_bytes);
 
 		list_for_each_entry(kq, &qpd->priv_queue_list, list) {
@@ -307,7 +162,7 @@ static int pm_create_runlist_ib(struct packet_manager *pm,
 			pr_debug("static_queue, mapping kernel q %d, is debug status %d\n",
 				kq->queue->queue, qpd->is_debug);
 
-			retval = pm_create_map_queue(pm,
+			retval = pm->pmf->map_queues(pm,
 						&rl_buffer[rl_wptr],
 						kq->queue,
 						qpd->is_debug);
@@ -315,7 +170,7 @@ static int pm_create_runlist_ib(struct packet_manager *pm,
 				return retval;
 
 			inc_wptr(&rl_wptr,
-				sizeof(struct pm4_mes_map_queues),
+				pm->pmf->map_queues_size,
 				alloc_size_bytes);
 		}
 
@@ -326,7 +181,7 @@ static int pm_create_runlist_ib(struct packet_manager *pm,
 			pr_debug("static_queue, mapping user queue %d, is debug status %d\n",
 				q->queue, qpd->is_debug);
 
-			retval = pm_create_map_queue(pm,
+			retval = pm->pmf->map_queues(pm,
 						&rl_buffer[rl_wptr],
 						q,
 						qpd->is_debug);
@@ -335,7 +190,7 @@ static int pm_create_runlist_ib(struct packet_manager *pm,
 				return retval;
 
 			inc_wptr(&rl_wptr,
-				sizeof(struct pm4_mes_map_queues),
+				pm->pmf->map_queues_size,
 				alloc_size_bytes);
 		}
 	}
@@ -343,7 +198,7 @@ static int pm_create_runlist_ib(struct packet_manager *pm,
 	pr_debug("Finished map process and queues to runlist\n");
 
 	if (is_over_subscription)
-		retval = pm_create_runlist(pm, &rl_buffer[rl_wptr],
+		retval = pm->pmf->runlist(pm, &rl_buffer[rl_wptr],
 					*rl_gpu_addr,
 					alloc_size_bytes / sizeof(uint32_t),
 					true);
@@ -355,45 +210,25 @@ static int pm_create_runlist_ib(struct packet_manager *pm,
 	return retval;
 }
 
-/* pm_create_release_mem - Create a RELEASE_MEM packet and return the size
- *     of this packet
- *     @gpu_addr - GPU address of the packet. It's a virtual address.
- *     @buffer - buffer to fill up with the packet. It's a CPU kernel pointer
- *     Return - length of the packet
- */
-uint32_t pm_create_release_mem(uint64_t gpu_addr, uint32_t *buffer)
-{
-	struct pm4_mec_release_mem *packet;
-
-	WARN_ON(!buffer);
-
-	packet = (struct pm4_mec_release_mem *)buffer;
-	memset(buffer, 0, sizeof(*packet));
-
-	packet->header.u32All = build_pm4_header(IT_RELEASE_MEM,
-						 sizeof(*packet));
-
-	packet->bitfields2.event_type = CACHE_FLUSH_AND_INV_TS_EVENT;
-	packet->bitfields2.event_index = event_index___release_mem__end_of_pipe;
-	packet->bitfields2.tcl1_action_ena = 1;
-	packet->bitfields2.tc_action_ena = 1;
-	packet->bitfields2.cache_policy = cache_policy___release_mem__lru;
-	packet->bitfields2.atc = 0;
-
-	packet->bitfields3.data_sel = data_sel___release_mem__send_32_bit_low;
-	packet->bitfields3.int_sel =
-		int_sel___release_mem__send_interrupt_after_write_confirm;
-
-	packet->bitfields4.address_lo_32b = (gpu_addr & 0xffffffff) >> 2;
-	packet->address_hi = upper_32_bits(gpu_addr);
-
-	packet->data_lo = 0;
-
-	return sizeof(*packet) / sizeof(unsigned int);
-}
-
 int pm_init(struct packet_manager *pm, struct device_queue_manager *dqm)
 {
+	switch (dqm->dev->device_info->asic_family) {
+	case CHIP_KAVERI:
+	case CHIP_HAWAII:
+		/* PM4 packet structures on CIK are the same as on VI */
+	case CHIP_CARRIZO:
+	case CHIP_TONGA:
+	case CHIP_FIJI:
+	case CHIP_POLARIS10:
+	case CHIP_POLARIS11:
+		pm->pmf = &kfd_vi_pm_funcs;
+		break;
+	default:
+		WARN(1, "Unexpected ASIC family %u",
+		     dqm->dev->device_info->asic_family);
+		return -EINVAL;
+	}
+
 	pm->dqm = dqm;
 	mutex_init(&pm->lock);
 	pm->priv_queue = kernel_queue_init(dqm->dev, KFD_QUEUE_TYPE_HIQ);
@@ -415,38 +250,25 @@ void pm_uninit(struct packet_manager *pm)
 int pm_send_set_resources(struct packet_manager *pm,
 				struct scheduling_resources *res)
 {
-	struct pm4_mes_set_resources *packet;
+	uint32_t *buffer, size;
 	int retval = 0;
 
+	size = pm->pmf->set_resources_size;
 	mutex_lock(&pm->lock);
 	pm->priv_queue->ops.acquire_packet_buffer(pm->priv_queue,
-					sizeof(*packet) / sizeof(uint32_t),
-					(unsigned int **)&packet);
-	if (!packet) {
+					size / sizeof(uint32_t),
+					(unsigned int **)&buffer);
+	if (!buffer) {
 		pr_err("Failed to allocate buffer on kernel queue\n");
 		retval = -ENOMEM;
 		goto out;
 	}
 
-	memset(packet, 0, sizeof(struct pm4_mes_set_resources));
-	packet->header.u32All = build_pm4_header(IT_SET_RESOURCES,
-					sizeof(struct pm4_mes_set_resources));
-
-	packet->bitfields2.queue_type =
-			queue_type__mes_set_resources__hsa_interface_queue_hiq;
-	packet->bitfields2.vmid_mask = res->vmid_mask;
-	packet->bitfields2.unmap_latency = KFD_UNMAP_LATENCY_MS / 100;
-	packet->bitfields7.oac_mask = res->oac_mask;
-	packet->bitfields8.gds_heap_base = res->gds_heap_base;
-	packet->bitfields8.gds_heap_size = res->gds_heap_size;
-
-	packet->gws_mask_lo = lower_32_bits(res->gws_mask);
-	packet->gws_mask_hi = upper_32_bits(res->gws_mask);
-
-	packet->queue_mask_lo = lower_32_bits(res->queue_mask);
-	packet->queue_mask_hi = upper_32_bits(res->queue_mask);
-
-	pm->priv_queue->ops.submit_packet(pm->priv_queue);
+	retval = pm->pmf->set_resources(pm, buffer, res);
+	if (!retval)
+		pm->priv_queue->ops.submit_packet(pm->priv_queue);
+	else
+		pm->priv_queue->ops.rollback_packet(pm->priv_queue);
 
 out:
 	mutex_unlock(&pm->lock);
@@ -468,7 +290,7 @@ int pm_send_runlist(struct packet_manager *pm, struct list_head *dqm_queues)
 
 	pr_debug("runlist IB address: 0x%llX\n", rl_gpu_ib_addr);
 
-	packet_size_dwords = sizeof(struct pm4_mes_runlist) / sizeof(uint32_t);
+	packet_size_dwords = pm->pmf->runlist_size / sizeof(uint32_t);
 	mutex_lock(&pm->lock);
 
 	retval = pm->priv_queue->ops.acquire_packet_buffer(pm->priv_queue,
@@ -476,7 +298,7 @@ int pm_send_runlist(struct packet_manager *pm, struct list_head *dqm_queues)
 	if (retval)
 		goto fail_acquire_packet_buffer;
 
-	retval = pm_create_runlist(pm, rl_buffer, rl_gpu_ib_addr,
+	retval = pm->pmf->runlist(pm, rl_buffer, rl_gpu_ib_addr,
 					rl_ib_size / sizeof(uint32_t), false);
 	if (retval)
 		goto fail_create_runlist;
@@ -499,37 +321,29 @@ fail_create_runlist_ib:
 int pm_send_query_status(struct packet_manager *pm, uint64_t fence_address,
 			uint32_t fence_value)
 {
-	int retval;
-	struct pm4_mes_query_status *packet;
+	uint32_t *buffer, size;
+	int retval = 0;
 
 	if (WARN_ON(!fence_address))
 		return -EFAULT;
 
+	size = pm->pmf->query_status_size;
 	mutex_lock(&pm->lock);
-	retval = pm->priv_queue->ops.acquire_packet_buffer(
-			pm->priv_queue,
-			sizeof(struct pm4_mes_query_status) / sizeof(uint32_t),
-			(unsigned int **)&packet);
-	if (retval)
-		goto fail_acquire_packet_buffer;
-
-	packet->header.u32All = build_pm4_header(IT_QUERY_STATUS,
-					sizeof(struct pm4_mes_query_status));
-
-	packet->bitfields2.context_id = 0;
-	packet->bitfields2.interrupt_sel =
-			interrupt_sel__mes_query_status__completion_status;
-	packet->bitfields2.command =
-			command__mes_query_status__fence_only_after_write_ack;
-
-	packet->addr_hi = upper_32_bits((uint64_t)fence_address);
-	packet->addr_lo = lower_32_bits((uint64_t)fence_address);
-	packet->data_hi = upper_32_bits((uint64_t)fence_value);
-	packet->data_lo = lower_32_bits((uint64_t)fence_value);
+	pm->priv_queue->ops.acquire_packet_buffer(pm->priv_queue,
+			size / sizeof(uint32_t), (unsigned int **)&buffer);
+	if (!buffer) {
+		pr_err("Failed to allocate buffer on kernel queue\n");
+		retval = -ENOMEM;
+		goto out;
+	}
 
-	pm->priv_queue->ops.submit_packet(pm->priv_queue);
+	retval = pm->pmf->query_status(pm, buffer, fence_address, fence_value);
+	if (!retval)
+		pm->priv_queue->ops.submit_packet(pm->priv_queue);
+	else
+		pm->priv_queue->ops.rollback_packet(pm->priv_queue);
 
-fail_acquire_packet_buffer:
+out:
 	mutex_unlock(&pm->lock);
 	return retval;
 }
@@ -539,82 +353,27 @@ int pm_send_unmap_queue(struct packet_manager *pm, enum kfd_queue_type type,
 			uint32_t filter_param, bool reset,
 			unsigned int sdma_engine)
 {
-	int retval;
-	uint32_t *buffer;
-	struct pm4_mes_unmap_queues *packet;
+	uint32_t *buffer, size;
+	int retval = 0;
 
+	size = pm->pmf->unmap_queues_size;
 	mutex_lock(&pm->lock);
-	retval = pm->priv_queue->ops.acquire_packet_buffer(
-			pm->priv_queue,
-			sizeof(struct pm4_mes_unmap_queues) / sizeof(uint32_t),
-			&buffer);
-	if (retval)
-		goto err_acquire_packet_buffer;
-
-	packet = (struct pm4_mes_unmap_queues *)buffer;
-	memset(buffer, 0, sizeof(struct pm4_mes_unmap_queues));
-	pr_debug("static_queue: unmapping queues: filter is %d , reset is %d , type is %d\n",
-		filter, reset, type);
-	packet->header.u32All = build_pm4_header(IT_UNMAP_QUEUES,
-					sizeof(struct pm4_mes_unmap_queues));
-	switch (type) {
-	case KFD_QUEUE_TYPE_COMPUTE:
-	case KFD_QUEUE_TYPE_DIQ:
-		packet->bitfields2.engine_sel =
-			engine_sel__mes_unmap_queues__compute;
-		break;
-	case KFD_QUEUE_TYPE_SDMA:
-		packet->bitfields2.engine_sel =
-			engine_sel__mes_unmap_queues__sdma0 + sdma_engine;
-		break;
-	default:
-		WARN(1, "queue type %d", type);
-		retval = -EINVAL;
-		goto err_invalid;
+	pm->priv_queue->ops.acquire_packet_buffer(pm->priv_queue,
+			size / sizeof(uint32_t), (unsigned int **)&buffer);
+	if (!buffer) {
+		pr_err("Failed to allocate buffer on kernel queue\n");
+		retval = -ENOMEM;
+		goto out;
 	}
 
-	if (reset)
-		packet->bitfields2.action =
-				action__mes_unmap_queues__reset_queues;
+	retval = pm->pmf->unmap_queues(pm, buffer, type, filter, filter_param,
+				       reset, sdma_engine);
+	if (!retval)
+		pm->priv_queue->ops.submit_packet(pm->priv_queue);
 	else
-		packet->bitfields2.action =
-				action__mes_unmap_queues__preempt_queues;
-
-	switch (filter) {
-	case KFD_UNMAP_QUEUES_FILTER_SINGLE_QUEUE:
-		packet->bitfields2.queue_sel =
-				queue_sel__mes_unmap_queues__perform_request_on_specified_queues;
-		packet->bitfields2.num_queues = 1;
-		packet->bitfields3b.doorbell_offset0 = filter_param;
-		break;
-	case KFD_UNMAP_QUEUES_FILTER_BY_PASID:
-		packet->bitfields2.queue_sel =
-				queue_sel__mes_unmap_queues__perform_request_on_pasid_queues;
-		packet->bitfields3a.pasid = filter_param;
-		break;
-	case KFD_UNMAP_QUEUES_FILTER_ALL_QUEUES:
-		packet->bitfields2.queue_sel =
-				queue_sel__mes_unmap_queues__unmap_all_queues;
-		break;
-	case KFD_UNMAP_QUEUES_FILTER_DYNAMIC_QUEUES:
-		/* in this case, we do not preempt static queues */
-		packet->bitfields2.queue_sel =
-				queue_sel__mes_unmap_queues__unmap_all_non_static_queues;
-		break;
-	default:
-		WARN(1, "filter %d", filter);
-		retval = -EINVAL;
-		goto err_invalid;
-	}
+		pm->priv_queue->ops.rollback_packet(pm->priv_queue);
 
-	pm->priv_queue->ops.submit_packet(pm->priv_queue);
-
-	mutex_unlock(&pm->lock);
-	return 0;
-
-err_invalid:
-	pm->priv_queue->ops.rollback_packet(pm->priv_queue);
-err_acquire_packet_buffer:
+out:
 	mutex_unlock(&pm->lock);
 	return retval;
 }
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
index ddb3c8c..873a8fb 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
@@ -866,8 +866,41 @@ struct packet_manager {
 	bool allocated;
 	struct kfd_mem_obj *ib_buffer_obj;
 	unsigned int ib_size_bytes;
+
+	const struct packet_manager_funcs *pmf;
+};
+
+struct packet_manager_funcs {
+	/* Support ASIC-specific packet formats for PM4 packets */
+	int (*map_process)(struct packet_manager *pm, uint32_t *buffer,
+			struct qcm_process_device *qpd);
+	int (*runlist)(struct packet_manager *pm, uint32_t *buffer,
+			uint64_t ib, size_t ib_size_in_dwords, bool chain);
+	int (*set_resources)(struct packet_manager *pm, uint32_t *buffer,
+			struct scheduling_resources *res);
+	int (*map_queues)(struct packet_manager *pm, uint32_t *buffer,
+			struct queue *q, bool is_static);
+	int (*unmap_queues)(struct packet_manager *pm, uint32_t *buffer,
+			enum kfd_queue_type type,
+			enum kfd_unmap_queues_filter mode,
+			uint32_t filter_param, bool reset,
+			unsigned int sdma_engine);
+	int (*query_status)(struct packet_manager *pm, uint32_t *buffer,
+			uint64_t fence_address,	uint32_t fence_value);
+	int (*release_mem)(uint64_t gpu_addr, uint32_t *buffer);
+
+	/* Packet sizes */
+	int map_process_size;
+	int runlist_size;
+	int set_resources_size;
+	int map_queues_size;
+	int unmap_queues_size;
+	int query_status_size;
+	int release_mem_size;
 };
 
+extern const struct packet_manager_funcs kfd_vi_pm_funcs;
+
 int pm_init(struct packet_manager *pm, struct device_queue_manager *dqm);
 void pm_uninit(struct packet_manager *pm);
 int pm_send_set_resources(struct packet_manager *pm,
@@ -883,8 +916,6 @@ int pm_send_unmap_queue(struct packet_manager *pm, enum kfd_queue_type type,
 
 void pm_release_ib(struct packet_manager *pm);
 
-uint32_t pm_create_release_mem(uint64_t gpu_addr, uint32_t *buffer);
-
 uint64_t kfd_get_number_elems(struct kfd_dev *kfd);
 
 /* Events */
-- 
cgit v1.1


From 454150b1f9a6be0a69138a698471bd13424204cc Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Tue, 10 Apr 2018 17:33:07 -0400
Subject: drm/amdkfd: Add GFXv9 PM4 packet writer functions

Signed-off-by: Shaoyun Liu <Shaoyun.Liu@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/Makefile              |   7 +-
 drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_v9.c | 331 +++++++++++++
 drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_vi.c |  18 +-
 drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c  |   4 +
 drivers/gpu/drm/amd/amdkfd/kfd_pm4_headers_ai.h  | 583 +++++++++++++++++++++++
 drivers/gpu/drm/amd/amdkfd/kfd_priv.h            |   6 +
 6 files changed, 937 insertions(+), 12 deletions(-)
 create mode 100644 drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_v9.c
 create mode 100644 drivers/gpu/drm/amd/amdkfd/kfd_pm4_headers_ai.h

diff --git a/drivers/gpu/drm/amd/amdkfd/Makefile b/drivers/gpu/drm/amd/amdkfd/Makefile
index 0d02422..52b3c1b 100644
--- a/drivers/gpu/drm/amd/amdkfd/Makefile
+++ b/drivers/gpu/drm/amd/amdkfd/Makefile
@@ -31,9 +31,10 @@ amdkfd-y	:= kfd_module.o kfd_device.o kfd_chardev.o kfd_topology.o \
 		kfd_process.o kfd_queue.o kfd_mqd_manager.o \
 		kfd_mqd_manager_cik.o kfd_mqd_manager_vi.o \
 		kfd_kernel_queue.o kfd_kernel_queue_cik.o \
-		kfd_kernel_queue_vi.o kfd_packet_manager.o \
-		kfd_process_queue_manager.o kfd_device_queue_manager.o \
-		kfd_device_queue_manager_cik.o kfd_device_queue_manager_vi.o \
+		kfd_kernel_queue_vi.o kfd_kernel_queue_v9.o \
+		kfd_packet_manager.o kfd_process_queue_manager.o \
+		kfd_device_queue_manager.o kfd_device_queue_manager_cik.o \
+		kfd_device_queue_manager_vi.o \
 		kfd_interrupt.o kfd_events.o cik_event_interrupt.o \
 		kfd_dbgdev.o kfd_dbgmgr.o kfd_crat.o
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_v9.c
new file mode 100644
index 0000000..ece7d59
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_v9.c
@@ -0,0 +1,331 @@
+/*
+ * Copyright 2016-2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include "kfd_kernel_queue.h"
+#include "kfd_device_queue_manager.h"
+#include "kfd_pm4_headers_ai.h"
+#include "kfd_pm4_opcodes.h"
+
+static bool initialize_v9(struct kernel_queue *kq, struct kfd_dev *dev,
+			enum kfd_queue_type type, unsigned int queue_size);
+static void uninitialize_v9(struct kernel_queue *kq);
+
+void kernel_queue_init_v9(struct kernel_queue_ops *ops)
+{
+	ops->initialize = initialize_v9;
+	ops->uninitialize = uninitialize_v9;
+}
+
+static bool initialize_v9(struct kernel_queue *kq, struct kfd_dev *dev,
+			enum kfd_queue_type type, unsigned int queue_size)
+{
+	int retval;
+
+	retval = kfd_gtt_sa_allocate(dev, PAGE_SIZE, &kq->eop_mem);
+	if (retval)
+		return false;
+
+	kq->eop_gpu_addr = kq->eop_mem->gpu_addr;
+	kq->eop_kernel_addr = kq->eop_mem->cpu_ptr;
+
+	memset(kq->eop_kernel_addr, 0, PAGE_SIZE);
+
+	return true;
+}
+
+static void uninitialize_v9(struct kernel_queue *kq)
+{
+	kfd_gtt_sa_free(kq->dev, kq->eop_mem);
+}
+
+static int pm_map_process_v9(struct packet_manager *pm,
+		uint32_t *buffer, struct qcm_process_device *qpd)
+{
+	struct pm4_mes_map_process *packet;
+	uint64_t vm_page_table_base_addr =
+		(uint64_t)(qpd->page_table_base) << 12;
+
+	packet = (struct pm4_mes_map_process *)buffer;
+	memset(buffer, 0, sizeof(struct pm4_mes_map_process));
+
+	packet->header.u32All = pm_build_pm4_header(IT_MAP_PROCESS,
+					sizeof(struct pm4_mes_map_process));
+	packet->bitfields2.diq_enable = (qpd->is_debug) ? 1 : 0;
+	packet->bitfields2.process_quantum = 1;
+	packet->bitfields2.pasid = qpd->pqm->process->pasid;
+	packet->bitfields14.gds_size = qpd->gds_size;
+	packet->bitfields14.num_gws = qpd->num_gws;
+	packet->bitfields14.num_oac = qpd->num_oac;
+	packet->bitfields14.sdma_enable = 1;
+	packet->bitfields14.num_queues = (qpd->is_debug) ? 0 : qpd->queue_count;
+
+	packet->sh_mem_config = qpd->sh_mem_config;
+	packet->sh_mem_bases = qpd->sh_mem_bases;
+	packet->sq_shader_tba_lo = lower_32_bits(qpd->tba_addr >> 8);
+	packet->sq_shader_tba_hi = upper_32_bits(qpd->tba_addr >> 8);
+	packet->sq_shader_tma_lo = lower_32_bits(qpd->tma_addr >> 8);
+	packet->sq_shader_tma_hi = upper_32_bits(qpd->tma_addr >> 8);
+
+	packet->gds_addr_lo = lower_32_bits(qpd->gds_context_area);
+	packet->gds_addr_hi = upper_32_bits(qpd->gds_context_area);
+
+	packet->vm_context_page_table_base_addr_lo32 =
+			lower_32_bits(vm_page_table_base_addr);
+	packet->vm_context_page_table_base_addr_hi32 =
+			upper_32_bits(vm_page_table_base_addr);
+
+	return 0;
+}
+
+static int pm_runlist_v9(struct packet_manager *pm, uint32_t *buffer,
+			uint64_t ib, size_t ib_size_in_dwords, bool chain)
+{
+	struct pm4_mes_runlist *packet;
+
+	int concurrent_proc_cnt = 0;
+	struct kfd_dev *kfd = pm->dqm->dev;
+
+	/* Determine the number of processes to map together to HW:
+	 * it can not exceed the number of VMIDs available to the
+	 * scheduler, and it is determined by the smaller of the number
+	 * of processes in the runlist and kfd module parameter
+	 * hws_max_conc_proc.
+	 * Note: the arbitration between the number of VMIDs and
+	 * hws_max_conc_proc has been done in
+	 * kgd2kfd_device_init().
+	 */
+	concurrent_proc_cnt = min(pm->dqm->processes_count,
+			kfd->max_proc_per_quantum);
+
+	packet = (struct pm4_mes_runlist *)buffer;
+
+	memset(buffer, 0, sizeof(struct pm4_mes_runlist));
+	packet->header.u32All = pm_build_pm4_header(IT_RUN_LIST,
+						sizeof(struct pm4_mes_runlist));
+
+	packet->bitfields4.ib_size = ib_size_in_dwords;
+	packet->bitfields4.chain = chain ? 1 : 0;
+	packet->bitfields4.offload_polling = 0;
+	packet->bitfields4.valid = 1;
+	packet->bitfields4.process_cnt = concurrent_proc_cnt;
+	packet->ordinal2 = lower_32_bits(ib);
+	packet->ib_base_hi = upper_32_bits(ib);
+
+	return 0;
+}
+
+static int pm_map_queues_v9(struct packet_manager *pm, uint32_t *buffer,
+		struct queue *q, bool is_static)
+{
+	struct pm4_mes_map_queues *packet;
+	bool use_static = is_static;
+
+	packet = (struct pm4_mes_map_queues *)buffer;
+	memset(buffer, 0, sizeof(struct pm4_mes_map_queues));
+
+	packet->header.u32All = pm_build_pm4_header(IT_MAP_QUEUES,
+					sizeof(struct pm4_mes_map_queues));
+	packet->bitfields2.alloc_format =
+		alloc_format__mes_map_queues__one_per_pipe_vi;
+	packet->bitfields2.num_queues = 1;
+	packet->bitfields2.queue_sel =
+		queue_sel__mes_map_queues__map_to_hws_determined_queue_slots_vi;
+
+	packet->bitfields2.engine_sel =
+		engine_sel__mes_map_queues__compute_vi;
+	packet->bitfields2.queue_type =
+		queue_type__mes_map_queues__normal_compute_vi;
+
+	switch (q->properties.type) {
+	case KFD_QUEUE_TYPE_COMPUTE:
+		if (use_static)
+			packet->bitfields2.queue_type =
+		queue_type__mes_map_queues__normal_latency_static_queue_vi;
+		break;
+	case KFD_QUEUE_TYPE_DIQ:
+		packet->bitfields2.queue_type =
+			queue_type__mes_map_queues__debug_interface_queue_vi;
+		break;
+	case KFD_QUEUE_TYPE_SDMA:
+		packet->bitfields2.engine_sel = q->properties.sdma_engine_id +
+				engine_sel__mes_map_queues__sdma0_vi;
+		use_static = false; /* no static queues under SDMA */
+		break;
+	default:
+		WARN(1, "queue type %d", q->properties.type);
+		return -EINVAL;
+	}
+	packet->bitfields3.doorbell_offset =
+			q->properties.doorbell_off;
+
+	packet->mqd_addr_lo =
+			lower_32_bits(q->gart_mqd_addr);
+
+	packet->mqd_addr_hi =
+			upper_32_bits(q->gart_mqd_addr);
+
+	packet->wptr_addr_lo =
+			lower_32_bits((uint64_t)q->properties.write_ptr);
+
+	packet->wptr_addr_hi =
+			upper_32_bits((uint64_t)q->properties.write_ptr);
+
+	return 0;
+}
+
+static int pm_unmap_queues_v9(struct packet_manager *pm, uint32_t *buffer,
+			enum kfd_queue_type type,
+			enum kfd_unmap_queues_filter filter,
+			uint32_t filter_param, bool reset,
+			unsigned int sdma_engine)
+{
+	struct pm4_mes_unmap_queues *packet;
+
+	packet = (struct pm4_mes_unmap_queues *)buffer;
+	memset(buffer, 0, sizeof(struct pm4_mes_unmap_queues));
+
+	packet->header.u32All = pm_build_pm4_header(IT_UNMAP_QUEUES,
+					sizeof(struct pm4_mes_unmap_queues));
+	switch (type) {
+	case KFD_QUEUE_TYPE_COMPUTE:
+	case KFD_QUEUE_TYPE_DIQ:
+		packet->bitfields2.engine_sel =
+			engine_sel__mes_unmap_queues__compute;
+		break;
+	case KFD_QUEUE_TYPE_SDMA:
+		packet->bitfields2.engine_sel =
+			engine_sel__mes_unmap_queues__sdma0 + sdma_engine;
+		break;
+	default:
+		WARN(1, "queue type %d", type);
+		return -EINVAL;
+	}
+
+	if (reset)
+		packet->bitfields2.action =
+			action__mes_unmap_queues__reset_queues;
+	else
+		packet->bitfields2.action =
+			action__mes_unmap_queues__preempt_queues;
+
+	switch (filter) {
+	case KFD_UNMAP_QUEUES_FILTER_SINGLE_QUEUE:
+		packet->bitfields2.queue_sel =
+			queue_sel__mes_unmap_queues__perform_request_on_specified_queues;
+		packet->bitfields2.num_queues = 1;
+		packet->bitfields3b.doorbell_offset0 = filter_param;
+		break;
+	case KFD_UNMAP_QUEUES_FILTER_BY_PASID:
+		packet->bitfields2.queue_sel =
+			queue_sel__mes_unmap_queues__perform_request_on_pasid_queues;
+		packet->bitfields3a.pasid = filter_param;
+		break;
+	case KFD_UNMAP_QUEUES_FILTER_ALL_QUEUES:
+		packet->bitfields2.queue_sel =
+			queue_sel__mes_unmap_queues__unmap_all_queues;
+		break;
+	case KFD_UNMAP_QUEUES_FILTER_DYNAMIC_QUEUES:
+		/* in this case, we do not preempt static queues */
+		packet->bitfields2.queue_sel =
+			queue_sel__mes_unmap_queues__unmap_all_non_static_queues;
+		break;
+	default:
+		WARN(1, "filter %d", filter);
+		return -EINVAL;
+	}
+
+	return 0;
+
+}
+
+static int pm_query_status_v9(struct packet_manager *pm, uint32_t *buffer,
+			uint64_t fence_address,	uint32_t fence_value)
+{
+	struct pm4_mes_query_status *packet;
+
+	packet = (struct pm4_mes_query_status *)buffer;
+	memset(buffer, 0, sizeof(struct pm4_mes_query_status));
+
+
+	packet->header.u32All = pm_build_pm4_header(IT_QUERY_STATUS,
+					sizeof(struct pm4_mes_query_status));
+
+	packet->bitfields2.context_id = 0;
+	packet->bitfields2.interrupt_sel =
+			interrupt_sel__mes_query_status__completion_status;
+	packet->bitfields2.command =
+			command__mes_query_status__fence_only_after_write_ack;
+
+	packet->addr_hi = upper_32_bits((uint64_t)fence_address);
+	packet->addr_lo = lower_32_bits((uint64_t)fence_address);
+	packet->data_hi = upper_32_bits((uint64_t)fence_value);
+	packet->data_lo = lower_32_bits((uint64_t)fence_value);
+
+	return 0;
+}
+
+
+static int pm_release_mem_v9(uint64_t gpu_addr, uint32_t *buffer)
+{
+	struct pm4_mec_release_mem *packet;
+
+	packet = (struct pm4_mec_release_mem *)buffer;
+	memset(buffer, 0, sizeof(struct pm4_mec_release_mem));
+
+	packet->header.u32All = pm_build_pm4_header(IT_RELEASE_MEM,
+					sizeof(struct pm4_mec_release_mem));
+
+	packet->bitfields2.event_type = CACHE_FLUSH_AND_INV_TS_EVENT;
+	packet->bitfields2.event_index = event_index__mec_release_mem__end_of_pipe;
+	packet->bitfields2.tcl1_action_ena = 1;
+	packet->bitfields2.tc_action_ena = 1;
+	packet->bitfields2.cache_policy = cache_policy__mec_release_mem__lru;
+
+	packet->bitfields3.data_sel = data_sel__mec_release_mem__send_32_bit_low;
+	packet->bitfields3.int_sel =
+		int_sel__mec_release_mem__send_interrupt_after_write_confirm;
+
+	packet->bitfields4.address_lo_32b = (gpu_addr & 0xffffffff) >> 2;
+	packet->address_hi = upper_32_bits(gpu_addr);
+
+	packet->data_lo = 0;
+
+	return 0;
+}
+
+const struct packet_manager_funcs kfd_v9_pm_funcs = {
+	.map_process		= pm_map_process_v9,
+	.runlist		= pm_runlist_v9,
+	.set_resources		= pm_set_resources_vi,
+	.map_queues		= pm_map_queues_v9,
+	.unmap_queues		= pm_unmap_queues_v9,
+	.query_status		= pm_query_status_v9,
+	.release_mem		= pm_release_mem_v9,
+	.map_process_size	= sizeof(struct pm4_mes_map_process),
+	.runlist_size		= sizeof(struct pm4_mes_runlist),
+	.set_resources_size	= sizeof(struct pm4_mes_set_resources),
+	.map_queues_size	= sizeof(struct pm4_mes_map_queues),
+	.unmap_queues_size	= sizeof(struct pm4_mes_unmap_queues),
+	.query_status_size	= sizeof(struct pm4_mes_query_status),
+	.release_mem_size	= sizeof(struct pm4_mec_release_mem)
+};
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_vi.c b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_vi.c
index 7ee326f..f9019ef 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_vi.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_vi.c
@@ -58,7 +58,7 @@ static void uninitialize_vi(struct kernel_queue *kq)
 	kfd_gtt_sa_free(kq->dev, kq->eop_mem);
 }
 
-static unsigned int build_pm4_header(unsigned int opcode, size_t packet_size)
+unsigned int pm_build_pm4_header(unsigned int opcode, size_t packet_size)
 {
 	union PM4_MES_TYPE_3_HEADER header;
 
@@ -79,7 +79,7 @@ static int pm_map_process_vi(struct packet_manager *pm, uint32_t *buffer,
 
 	memset(buffer, 0, sizeof(struct pm4_mes_map_process));
 
-	packet->header.u32All = build_pm4_header(IT_MAP_PROCESS,
+	packet->header.u32All = pm_build_pm4_header(IT_MAP_PROCESS,
 					sizeof(struct pm4_mes_map_process));
 	packet->bitfields2.diq_enable = (qpd->is_debug) ? 1 : 0;
 	packet->bitfields2.process_quantum = 1;
@@ -128,7 +128,7 @@ static int pm_runlist_vi(struct packet_manager *pm, uint32_t *buffer,
 	packet = (struct pm4_mes_runlist *)buffer;
 
 	memset(buffer, 0, sizeof(struct pm4_mes_runlist));
-	packet->header.u32All = build_pm4_header(IT_RUN_LIST,
+	packet->header.u32All = pm_build_pm4_header(IT_RUN_LIST,
 						sizeof(struct pm4_mes_runlist));
 
 	packet->bitfields4.ib_size = ib_size_in_dwords;
@@ -142,7 +142,7 @@ static int pm_runlist_vi(struct packet_manager *pm, uint32_t *buffer,
 	return 0;
 }
 
-static int pm_set_resources_vi(struct packet_manager *pm, uint32_t *buffer,
+int pm_set_resources_vi(struct packet_manager *pm, uint32_t *buffer,
 				struct scheduling_resources *res)
 {
 	struct pm4_mes_set_resources *packet;
@@ -150,7 +150,7 @@ static int pm_set_resources_vi(struct packet_manager *pm, uint32_t *buffer,
 	packet = (struct pm4_mes_set_resources *)buffer;
 	memset(buffer, 0, sizeof(struct pm4_mes_set_resources));
 
-	packet->header.u32All = build_pm4_header(IT_SET_RESOURCES,
+	packet->header.u32All = pm_build_pm4_header(IT_SET_RESOURCES,
 					sizeof(struct pm4_mes_set_resources));
 
 	packet->bitfields2.queue_type =
@@ -179,7 +179,7 @@ static int pm_map_queues_vi(struct packet_manager *pm, uint32_t *buffer,
 	packet = (struct pm4_mes_map_queues *)buffer;
 	memset(buffer, 0, sizeof(struct pm4_mes_map_queues));
 
-	packet->header.u32All = build_pm4_header(IT_MAP_QUEUES,
+	packet->header.u32All = pm_build_pm4_header(IT_MAP_QUEUES,
 					sizeof(struct pm4_mes_map_queues));
 	packet->bitfields2.alloc_format =
 		alloc_format__mes_map_queues__one_per_pipe_vi;
@@ -240,7 +240,7 @@ static int pm_unmap_queues_vi(struct packet_manager *pm, uint32_t *buffer,
 	packet = (struct pm4_mes_unmap_queues *)buffer;
 	memset(buffer, 0, sizeof(struct pm4_mes_unmap_queues));
 
-	packet->header.u32All = build_pm4_header(IT_UNMAP_QUEUES,
+	packet->header.u32All = pm_build_pm4_header(IT_UNMAP_QUEUES,
 					sizeof(struct pm4_mes_unmap_queues));
 	switch (type) {
 	case KFD_QUEUE_TYPE_COMPUTE:
@@ -302,7 +302,7 @@ static int pm_query_status_vi(struct packet_manager *pm, uint32_t *buffer,
 	packet = (struct pm4_mes_query_status *)buffer;
 	memset(buffer, 0, sizeof(struct pm4_mes_query_status));
 
-	packet->header.u32All = build_pm4_header(IT_QUERY_STATUS,
+	packet->header.u32All = pm_build_pm4_header(IT_QUERY_STATUS,
 					sizeof(struct pm4_mes_query_status));
 
 	packet->bitfields2.context_id = 0;
@@ -326,7 +326,7 @@ static int pm_release_mem_vi(uint64_t gpu_addr, uint32_t *buffer)
 	packet = (struct pm4_mec_release_mem *)buffer;
 	memset(buffer, 0, sizeof(*packet));
 
-	packet->header.u32All = build_pm4_header(IT_RELEASE_MEM,
+	packet->header.u32All = pm_build_pm4_header(IT_RELEASE_MEM,
 						 sizeof(*packet));
 
 	packet->bitfields2.event_type = CACHE_FLUSH_AND_INV_TS_EVENT;
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c
index 860ff24..91f0350 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c
@@ -223,6 +223,10 @@ int pm_init(struct packet_manager *pm, struct device_queue_manager *dqm)
 	case CHIP_POLARIS11:
 		pm->pmf = &kfd_vi_pm_funcs;
 		break;
+	case CHIP_VEGA10:
+	case CHIP_RAVEN:
+		pm->pmf = &kfd_v9_pm_funcs;
+		break;
 	default:
 		WARN(1, "Unexpected ASIC family %u",
 		     dqm->dev->device_info->asic_family);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_pm4_headers_ai.h b/drivers/gpu/drm/amd/amdkfd/kfd_pm4_headers_ai.h
new file mode 100644
index 0000000..f2bcf5c
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_pm4_headers_ai.h
@@ -0,0 +1,583 @@
+/*
+ * Copyright 2016 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef F32_MES_PM4_PACKETS_H
+#define F32_MES_PM4_PACKETS_H
+
+#ifndef PM4_MES_HEADER_DEFINED
+#define PM4_MES_HEADER_DEFINED
+union PM4_MES_TYPE_3_HEADER {
+	struct {
+		uint32_t reserved1 : 8; /* < reserved */
+		uint32_t opcode    : 8; /* < IT opcode */
+		uint32_t count     : 14;/* < number of DWORDs - 1 in the
+					 *   information body.
+					 */
+		uint32_t type      : 2; /* < packet identifier.
+					 *   It should be 3 for type 3 packets
+					 */
+	};
+	uint32_t u32All;
+};
+#endif /* PM4_MES_HEADER_DEFINED */
+
+/*--------------------MES_SET_RESOURCES--------------------*/
+
+#ifndef PM4_MES_SET_RESOURCES_DEFINED
+#define PM4_MES_SET_RESOURCES_DEFINED
+enum mes_set_resources_queue_type_enum {
+	queue_type__mes_set_resources__kernel_interface_queue_kiq = 0,
+	queue_type__mes_set_resources__hsa_interface_queue_hiq = 1,
+	queue_type__mes_set_resources__hsa_debug_interface_queue = 4
+};
+
+
+struct pm4_mes_set_resources {
+	union {
+		union PM4_MES_TYPE_3_HEADER	header;		/* header */
+		uint32_t			ordinal1;
+	};
+
+	union {
+		struct {
+			uint32_t vmid_mask:16;
+			uint32_t unmap_latency:8;
+			uint32_t reserved1:5;
+			enum mes_set_resources_queue_type_enum queue_type:3;
+		} bitfields2;
+		uint32_t ordinal2;
+	};
+
+	uint32_t queue_mask_lo;
+	uint32_t queue_mask_hi;
+	uint32_t gws_mask_lo;
+	uint32_t gws_mask_hi;
+
+	union {
+		struct {
+			uint32_t oac_mask:16;
+			uint32_t reserved2:16;
+		} bitfields7;
+		uint32_t ordinal7;
+	};
+
+	union {
+		struct {
+		uint32_t gds_heap_base:6;
+		uint32_t reserved3:5;
+		uint32_t gds_heap_size:6;
+		uint32_t reserved4:15;
+		} bitfields8;
+		uint32_t ordinal8;
+	};
+
+};
+#endif
+
+/*--------------------MES_RUN_LIST--------------------*/
+
+#ifndef PM4_MES_RUN_LIST_DEFINED
+#define PM4_MES_RUN_LIST_DEFINED
+
+struct pm4_mes_runlist {
+	union {
+		union PM4_MES_TYPE_3_HEADER header; /* header */
+		uint32_t ordinal1;
+	};
+
+	union {
+		struct {
+			uint32_t reserved1:2;
+			uint32_t ib_base_lo:30;
+		} bitfields2;
+		uint32_t ordinal2;
+	};
+
+	uint32_t ib_base_hi;
+
+	union {
+		struct {
+			uint32_t ib_size:20;
+			uint32_t chain:1;
+			uint32_t offload_polling:1;
+			uint32_t reserved2:1;
+			uint32_t valid:1;
+			uint32_t process_cnt:4;
+			uint32_t reserved3:4;
+		} bitfields4;
+		uint32_t ordinal4;
+	};
+
+};
+#endif
+
+/*--------------------MES_MAP_PROCESS--------------------*/
+
+#ifndef PM4_MES_MAP_PROCESS_DEFINED
+#define PM4_MES_MAP_PROCESS_DEFINED
+
+struct pm4_mes_map_process {
+	union {
+		union PM4_MES_TYPE_3_HEADER header;	/* header */
+		uint32_t ordinal1;
+	};
+
+	union {
+		struct {
+			uint32_t pasid:16;
+			uint32_t reserved1:8;
+			uint32_t diq_enable:1;
+			uint32_t process_quantum:7;
+		} bitfields2;
+		uint32_t ordinal2;
+	};
+
+	uint32_t vm_context_page_table_base_addr_lo32;
+
+	uint32_t vm_context_page_table_base_addr_hi32;
+
+	uint32_t sh_mem_bases;
+
+	uint32_t sh_mem_config;
+
+	uint32_t sq_shader_tba_lo;
+
+	uint32_t sq_shader_tba_hi;
+
+	uint32_t sq_shader_tma_lo;
+
+	uint32_t sq_shader_tma_hi;
+
+	uint32_t reserved6;
+
+	uint32_t gds_addr_lo;
+
+	uint32_t gds_addr_hi;
+
+	union {
+		struct {
+			uint32_t num_gws:6;
+			uint32_t reserved7:1;
+			uint32_t sdma_enable:1;
+			uint32_t num_oac:4;
+			uint32_t reserved8:4;
+			uint32_t gds_size:6;
+			uint32_t num_queues:10;
+		} bitfields14;
+		uint32_t ordinal14;
+	};
+
+	uint32_t completion_signal_lo;
+
+	uint32_t completion_signal_hi;
+
+};
+
+#endif
+
+/*--------------------MES_MAP_PROCESS_VM--------------------*/
+
+#ifndef PM4_MES_MAP_PROCESS_VM_DEFINED
+#define PM4_MES_MAP_PROCESS_VM_DEFINED
+
+struct PM4_MES_MAP_PROCESS_VM {
+	union {
+		union PM4_MES_TYPE_3_HEADER header;	/* header */
+		uint32_t ordinal1;
+	};
+
+	uint32_t reserved1;
+
+	uint32_t vm_context_cntl;
+
+	uint32_t reserved2;
+
+	uint32_t vm_context_page_table_end_addr_lo32;
+
+	uint32_t vm_context_page_table_end_addr_hi32;
+
+	uint32_t vm_context_page_table_start_addr_lo32;
+
+	uint32_t vm_context_page_table_start_addr_hi32;
+
+	uint32_t reserved3;
+
+	uint32_t reserved4;
+
+	uint32_t reserved5;
+
+	uint32_t reserved6;
+
+	uint32_t reserved7;
+
+	uint32_t reserved8;
+
+	uint32_t completion_signal_lo32;
+
+	uint32_t completion_signal_hi32;
+
+};
+#endif
+
+/*--------------------MES_MAP_QUEUES--------------------*/
+
+#ifndef PM4_MES_MAP_QUEUES_VI_DEFINED
+#define PM4_MES_MAP_QUEUES_VI_DEFINED
+enum mes_map_queues_queue_sel_enum {
+	queue_sel__mes_map_queues__map_to_specified_queue_slots_vi = 0,
+queue_sel__mes_map_queues__map_to_hws_determined_queue_slots_vi = 1
+};
+
+enum mes_map_queues_queue_type_enum {
+	queue_type__mes_map_queues__normal_compute_vi = 0,
+	queue_type__mes_map_queues__debug_interface_queue_vi = 1,
+	queue_type__mes_map_queues__normal_latency_static_queue_vi = 2,
+queue_type__mes_map_queues__low_latency_static_queue_vi = 3
+};
+
+enum mes_map_queues_alloc_format_enum {
+	alloc_format__mes_map_queues__one_per_pipe_vi = 0,
+alloc_format__mes_map_queues__all_on_one_pipe_vi = 1
+};
+
+enum mes_map_queues_engine_sel_enum {
+	engine_sel__mes_map_queues__compute_vi = 0,
+	engine_sel__mes_map_queues__sdma0_vi = 2,
+	engine_sel__mes_map_queues__sdma1_vi = 3
+};
+
+
+struct pm4_mes_map_queues {
+	union {
+		union PM4_MES_TYPE_3_HEADER   header;            /* header */
+		uint32_t            ordinal1;
+	};
+
+	union {
+		struct {
+			uint32_t reserved1:4;
+			enum mes_map_queues_queue_sel_enum queue_sel:2;
+			uint32_t reserved2:15;
+			enum mes_map_queues_queue_type_enum queue_type:3;
+			enum mes_map_queues_alloc_format_enum alloc_format:2;
+			enum mes_map_queues_engine_sel_enum engine_sel:3;
+			uint32_t num_queues:3;
+		} bitfields2;
+		uint32_t ordinal2;
+	};
+
+	union {
+		struct {
+			uint32_t reserved3:1;
+			uint32_t check_disable:1;
+			uint32_t doorbell_offset:26;
+			uint32_t reserved4:4;
+		} bitfields3;
+		uint32_t ordinal3;
+	};
+
+	uint32_t mqd_addr_lo;
+	uint32_t mqd_addr_hi;
+	uint32_t wptr_addr_lo;
+	uint32_t wptr_addr_hi;
+};
+#endif
+
+/*--------------------MES_QUERY_STATUS--------------------*/
+
+#ifndef PM4_MES_QUERY_STATUS_DEFINED
+#define PM4_MES_QUERY_STATUS_DEFINED
+enum mes_query_status_interrupt_sel_enum {
+	interrupt_sel__mes_query_status__completion_status = 0,
+	interrupt_sel__mes_query_status__process_status = 1,
+	interrupt_sel__mes_query_status__queue_status = 2
+};
+
+enum mes_query_status_command_enum {
+	command__mes_query_status__interrupt_only = 0,
+	command__mes_query_status__fence_only_immediate = 1,
+	command__mes_query_status__fence_only_after_write_ack = 2,
+	command__mes_query_status__fence_wait_for_write_ack_send_interrupt = 3
+};
+
+enum mes_query_status_engine_sel_enum {
+	engine_sel__mes_query_status__compute = 0,
+	engine_sel__mes_query_status__sdma0_queue = 2,
+	engine_sel__mes_query_status__sdma1_queue = 3
+};
+
+struct pm4_mes_query_status {
+	union {
+		union PM4_MES_TYPE_3_HEADER   header;            /* header */
+		uint32_t            ordinal1;
+	};
+
+	union {
+		struct {
+			uint32_t context_id:28;
+			enum mes_query_status_interrupt_sel_enum	interrupt_sel:2;
+			enum mes_query_status_command_enum command:2;
+		} bitfields2;
+		uint32_t ordinal2;
+	};
+
+	union {
+		struct {
+			uint32_t pasid:16;
+			uint32_t reserved1:16;
+		} bitfields3a;
+		struct {
+			uint32_t reserved2:2;
+			uint32_t doorbell_offset:26;
+			enum mes_query_status_engine_sel_enum engine_sel:3;
+			uint32_t reserved3:1;
+		} bitfields3b;
+		uint32_t ordinal3;
+	};
+
+	uint32_t addr_lo;
+	uint32_t addr_hi;
+	uint32_t data_lo;
+	uint32_t data_hi;
+};
+#endif
+
+/*--------------------MES_UNMAP_QUEUES--------------------*/
+
+#ifndef PM4_MES_UNMAP_QUEUES_DEFINED
+#define PM4_MES_UNMAP_QUEUES_DEFINED
+enum mes_unmap_queues_action_enum {
+	action__mes_unmap_queues__preempt_queues = 0,
+	action__mes_unmap_queues__reset_queues = 1,
+	action__mes_unmap_queues__disable_process_queues = 2,
+	action__mes_unmap_queues__reserved = 3
+};
+
+enum mes_unmap_queues_queue_sel_enum {
+	queue_sel__mes_unmap_queues__perform_request_on_specified_queues = 0,
+	queue_sel__mes_unmap_queues__perform_request_on_pasid_queues = 1,
+	queue_sel__mes_unmap_queues__unmap_all_queues = 2,
+	queue_sel__mes_unmap_queues__unmap_all_non_static_queues = 3
+};
+
+enum mes_unmap_queues_engine_sel_enum {
+	engine_sel__mes_unmap_queues__compute = 0,
+	engine_sel__mes_unmap_queues__sdma0 = 2,
+	engine_sel__mes_unmap_queues__sdmal = 3
+};
+
+struct pm4_mes_unmap_queues {
+	union {
+		union PM4_MES_TYPE_3_HEADER   header;            /* header */
+		uint32_t            ordinal1;
+	};
+
+	union {
+		struct {
+			enum mes_unmap_queues_action_enum action:2;
+			uint32_t reserved1:2;
+			enum mes_unmap_queues_queue_sel_enum queue_sel:2;
+			uint32_t reserved2:20;
+			enum mes_unmap_queues_engine_sel_enum engine_sel:3;
+			uint32_t num_queues:3;
+		} bitfields2;
+		uint32_t ordinal2;
+	};
+
+	union {
+		struct {
+			uint32_t pasid:16;
+			uint32_t reserved3:16;
+		} bitfields3a;
+		struct {
+			uint32_t reserved4:2;
+			uint32_t doorbell_offset0:26;
+			int32_t reserved5:4;
+		} bitfields3b;
+		uint32_t ordinal3;
+	};
+
+	union {
+	struct {
+			uint32_t reserved6:2;
+			uint32_t doorbell_offset1:26;
+			uint32_t reserved7:4;
+		} bitfields4;
+		uint32_t ordinal4;
+	};
+
+	union {
+		struct {
+			uint32_t reserved8:2;
+			uint32_t doorbell_offset2:26;
+			uint32_t reserved9:4;
+		} bitfields5;
+		uint32_t ordinal5;
+	};
+
+	union {
+		struct {
+			uint32_t reserved10:2;
+			uint32_t doorbell_offset3:26;
+			uint32_t reserved11:4;
+		} bitfields6;
+		uint32_t ordinal6;
+	};
+};
+#endif
+
+#ifndef PM4_MEC_RELEASE_MEM_DEFINED
+#define PM4_MEC_RELEASE_MEM_DEFINED
+
+enum mec_release_mem_event_index_enum {
+	event_index__mec_release_mem__end_of_pipe = 5,
+	event_index__mec_release_mem__shader_done = 6
+};
+
+enum mec_release_mem_cache_policy_enum {
+	cache_policy__mec_release_mem__lru = 0,
+	cache_policy__mec_release_mem__stream = 1
+};
+
+enum mec_release_mem_pq_exe_status_enum {
+	pq_exe_status__mec_release_mem__default = 0,
+	pq_exe_status__mec_release_mem__phase_update = 1
+};
+
+enum mec_release_mem_dst_sel_enum {
+	dst_sel__mec_release_mem__memory_controller = 0,
+	dst_sel__mec_release_mem__tc_l2 = 1,
+	dst_sel__mec_release_mem__queue_write_pointer_register = 2,
+	dst_sel__mec_release_mem__queue_write_pointer_poll_mask_bit = 3
+};
+
+enum mec_release_mem_int_sel_enum {
+	int_sel__mec_release_mem__none = 0,
+	int_sel__mec_release_mem__send_interrupt_only = 1,
+	int_sel__mec_release_mem__send_interrupt_after_write_confirm = 2,
+	int_sel__mec_release_mem__send_data_after_write_confirm = 3,
+	int_sel__mec_release_mem__unconditionally_send_int_ctxid = 4,
+	int_sel__mec_release_mem__conditionally_send_int_ctxid_based_on_32_bit_compare = 5,
+	int_sel__mec_release_mem__conditionally_send_int_ctxid_based_on_64_bit_compare = 6
+};
+
+enum mec_release_mem_data_sel_enum {
+	data_sel__mec_release_mem__none = 0,
+	data_sel__mec_release_mem__send_32_bit_low = 1,
+	data_sel__mec_release_mem__send_64_bit_data = 2,
+	data_sel__mec_release_mem__send_gpu_clock_counter = 3,
+	data_sel__mec_release_mem__send_cp_perfcounter_hi_lo = 4,
+	data_sel__mec_release_mem__store_gds_data_to_memory = 5
+};
+
+struct pm4_mec_release_mem {
+	union {
+		union PM4_MES_TYPE_3_HEADER header;     /*header */
+		unsigned int ordinal1;
+	};
+
+	union {
+		struct {
+			unsigned int event_type:6;
+			unsigned int reserved1:2;
+			enum mec_release_mem_event_index_enum event_index:4;
+			unsigned int tcl1_vol_action_ena:1;
+			unsigned int tc_vol_action_ena:1;
+			unsigned int reserved2:1;
+			unsigned int tc_wb_action_ena:1;
+			unsigned int tcl1_action_ena:1;
+			unsigned int tc_action_ena:1;
+			uint32_t reserved3:1;
+			uint32_t tc_nc_action_ena:1;
+			uint32_t tc_wc_action_ena:1;
+			uint32_t tc_md_action_ena:1;
+			uint32_t reserved4:3;
+			enum mec_release_mem_cache_policy_enum cache_policy:2;
+			uint32_t reserved5:2;
+			enum mec_release_mem_pq_exe_status_enum pq_exe_status:1;
+			uint32_t reserved6:2;
+		} bitfields2;
+		unsigned int ordinal2;
+	};
+
+	union {
+		struct {
+			uint32_t reserved7:16;
+			enum mec_release_mem_dst_sel_enum dst_sel:2;
+			uint32_t reserved8:6;
+			enum mec_release_mem_int_sel_enum int_sel:3;
+			uint32_t reserved9:2;
+			enum mec_release_mem_data_sel_enum data_sel:3;
+		} bitfields3;
+		unsigned int ordinal3;
+	};
+
+	union {
+		struct {
+			uint32_t reserved10:2;
+			unsigned int address_lo_32b:30;
+		} bitfields4;
+		struct {
+			uint32_t reserved11:3;
+			uint32_t address_lo_64b:29;
+		} bitfields4b;
+		uint32_t reserved12;
+		unsigned int ordinal4;
+	};
+
+	union {
+		uint32_t address_hi;
+		uint32_t reserved13;
+		uint32_t ordinal5;
+	};
+
+	union {
+		uint32_t data_lo;
+		uint32_t cmp_data_lo;
+		struct {
+			uint32_t dw_offset:16;
+			uint32_t num_dwords:16;
+		} bitfields6c;
+		uint32_t reserved14;
+		uint32_t ordinal6;
+	};
+
+	union {
+		uint32_t data_hi;
+		uint32_t cmp_data_hi;
+		uint32_t reserved15;
+		uint32_t reserved16;
+		uint32_t ordinal7;
+	};
+
+	uint32_t int_ctxid;
+
+};
+
+#endif
+
+enum {
+	CACHE_FLUSH_AND_INV_TS_EVENT = 0x00000014
+};
+#endif
+
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
index 873a8fb..b68299a 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
@@ -900,6 +900,7 @@ struct packet_manager_funcs {
 };
 
 extern const struct packet_manager_funcs kfd_vi_pm_funcs;
+extern const struct packet_manager_funcs kfd_v9_pm_funcs;
 
 int pm_init(struct packet_manager *pm, struct device_queue_manager *dqm);
 void pm_uninit(struct packet_manager *pm);
@@ -916,6 +917,11 @@ int pm_send_unmap_queue(struct packet_manager *pm, enum kfd_queue_type type,
 
 void pm_release_ib(struct packet_manager *pm);
 
+/* Following PM funcs can be shared among VI and AI */
+unsigned int pm_build_pm4_header(unsigned int opcode, size_t packet_size);
+int pm_set_resources_vi(struct packet_manager *pm, uint32_t *buffer,
+				struct scheduling_resources *res);
+
 uint64_t kfd_get_number_elems(struct kfd_dev *kfd);
 
 /* Events */
-- 
cgit v1.1


From b91d43dd01aadd43b1002160b78d77f8175876a4 Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Tue, 10 Apr 2018 17:33:08 -0400
Subject: drm/amdkfd: Add GFXv9 MQD manager

Signed-off-by: John Bridgman <john.bridgman@amd.com>
Signed-off-by: Jay Cornwall <Jay.Cornwall@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/Makefile             |   1 +
 drivers/gpu/drm/amd/amdkfd/kfd_device.c         |   2 +-
 drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c    |   3 +
 drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c | 443 ++++++++++++++++++++++++
 drivers/gpu/drm/amd/amdkfd/kfd_priv.h           |   3 +
 5 files changed, 451 insertions(+), 1 deletion(-)
 create mode 100644 drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c

diff --git a/drivers/gpu/drm/amd/amdkfd/Makefile b/drivers/gpu/drm/amd/amdkfd/Makefile
index 52b3c1b..094b591 100644
--- a/drivers/gpu/drm/amd/amdkfd/Makefile
+++ b/drivers/gpu/drm/amd/amdkfd/Makefile
@@ -30,6 +30,7 @@ amdkfd-y	:= kfd_module.o kfd_device.o kfd_chardev.o kfd_topology.o \
 		kfd_pasid.o kfd_doorbell.o kfd_flat_memory.o \
 		kfd_process.o kfd_queue.o kfd_mqd_manager.o \
 		kfd_mqd_manager_cik.o kfd_mqd_manager_vi.o \
+		kfd_mqd_manager_v9.o \
 		kfd_kernel_queue.o kfd_kernel_queue_cik.o \
 		kfd_kernel_queue_vi.o kfd_kernel_queue_v9.o \
 		kfd_packet_manager.o kfd_process_queue_manager.o \
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
index f563acb..c368ce3 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
@@ -700,7 +700,7 @@ int kfd_gtt_sa_allocate(struct kfd_dev *kfd, unsigned int size,
 	if (size > kfd->gtt_sa_num_of_chunks * kfd->gtt_sa_chunk_size)
 		return -ENOMEM;
 
-	*mem_obj = kmalloc(sizeof(struct kfd_mem_obj), GFP_NOIO);
+	*mem_obj = kzalloc(sizeof(struct kfd_mem_obj), GFP_NOIO);
 	if ((*mem_obj) == NULL)
 		return -ENOMEM;
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c
index ee7061e..4b8eb50 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c
@@ -38,6 +38,9 @@ struct mqd_manager *mqd_manager_init(enum KFD_MQD_TYPE type,
 	case CHIP_POLARIS10:
 	case CHIP_POLARIS11:
 		return mqd_manager_init_vi_tonga(type, dev);
+	case CHIP_VEGA10:
+	case CHIP_RAVEN:
+		return mqd_manager_init_v9(type, dev);
 	default:
 		WARN(1, "Unexpected ASIC family %u",
 		     dev->device_info->asic_family);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c
new file mode 100644
index 0000000..684054f
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c
@@ -0,0 +1,443 @@
+/*
+ * Copyright 2016-2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include <linux/printk.h>
+#include <linux/slab.h>
+#include <linux/uaccess.h>
+#include "kfd_priv.h"
+#include "kfd_mqd_manager.h"
+#include "v9_structs.h"
+#include "gc/gc_9_0_offset.h"
+#include "gc/gc_9_0_sh_mask.h"
+#include "sdma0/sdma0_4_0_sh_mask.h"
+
+static inline struct v9_mqd *get_mqd(void *mqd)
+{
+	return (struct v9_mqd *)mqd;
+}
+
+static inline struct v9_sdma_mqd *get_sdma_mqd(void *mqd)
+{
+	return (struct v9_sdma_mqd *)mqd;
+}
+
+static int init_mqd(struct mqd_manager *mm, void **mqd,
+			struct kfd_mem_obj **mqd_mem_obj, uint64_t *gart_addr,
+			struct queue_properties *q)
+{
+	int retval;
+	uint64_t addr;
+	struct v9_mqd *m;
+	struct kfd_dev *kfd = mm->dev;
+
+	/* From V9,  for CWSR, the control stack is located on the next page
+	 * boundary after the mqd, we will use the gtt allocation function
+	 * instead of sub-allocation function.
+	 */
+	if (kfd->cwsr_enabled && (q->type == KFD_QUEUE_TYPE_COMPUTE)) {
+		*mqd_mem_obj = kzalloc(sizeof(struct kfd_mem_obj), GFP_NOIO);
+		if (!*mqd_mem_obj)
+			return -ENOMEM;
+		retval = kfd->kfd2kgd->init_gtt_mem_allocation(kfd->kgd,
+			ALIGN(q->ctl_stack_size, PAGE_SIZE) +
+				ALIGN(sizeof(struct v9_mqd), PAGE_SIZE),
+			&((*mqd_mem_obj)->gtt_mem),
+			&((*mqd_mem_obj)->gpu_addr),
+			(void *)&((*mqd_mem_obj)->cpu_ptr));
+	} else
+		retval = kfd_gtt_sa_allocate(mm->dev, sizeof(struct v9_mqd),
+				mqd_mem_obj);
+	if (retval != 0)
+		return -ENOMEM;
+
+	m = (struct v9_mqd *) (*mqd_mem_obj)->cpu_ptr;
+	addr = (*mqd_mem_obj)->gpu_addr;
+
+	memset(m, 0, sizeof(struct v9_mqd));
+
+	m->header = 0xC0310800;
+	m->compute_pipelinestat_enable = 1;
+	m->compute_static_thread_mgmt_se0 = 0xFFFFFFFF;
+	m->compute_static_thread_mgmt_se1 = 0xFFFFFFFF;
+	m->compute_static_thread_mgmt_se2 = 0xFFFFFFFF;
+	m->compute_static_thread_mgmt_se3 = 0xFFFFFFFF;
+
+	m->cp_hqd_persistent_state = CP_HQD_PERSISTENT_STATE__PRELOAD_REQ_MASK |
+			0x53 << CP_HQD_PERSISTENT_STATE__PRELOAD_SIZE__SHIFT;
+
+	m->cp_mqd_control = 1 << CP_MQD_CONTROL__PRIV_STATE__SHIFT;
+
+	m->cp_mqd_base_addr_lo        = lower_32_bits(addr);
+	m->cp_mqd_base_addr_hi        = upper_32_bits(addr);
+
+	m->cp_hqd_quantum = 1 << CP_HQD_QUANTUM__QUANTUM_EN__SHIFT |
+			1 << CP_HQD_QUANTUM__QUANTUM_SCALE__SHIFT |
+			10 << CP_HQD_QUANTUM__QUANTUM_DURATION__SHIFT;
+
+	m->cp_hqd_pipe_priority = 1;
+	m->cp_hqd_queue_priority = 15;
+
+	if (q->format == KFD_QUEUE_FORMAT_AQL) {
+		m->cp_hqd_aql_control =
+			1 << CP_HQD_AQL_CONTROL__CONTROL0__SHIFT;
+	}
+
+	if (q->tba_addr) {
+		m->compute_pgm_rsrc2 |=
+			(1 << COMPUTE_PGM_RSRC2__TRAP_PRESENT__SHIFT);
+	}
+
+	if (mm->dev->cwsr_enabled && q->ctx_save_restore_area_address) {
+		m->cp_hqd_persistent_state |=
+			(1 << CP_HQD_PERSISTENT_STATE__QSWITCH_MODE__SHIFT);
+		m->cp_hqd_ctx_save_base_addr_lo =
+			lower_32_bits(q->ctx_save_restore_area_address);
+		m->cp_hqd_ctx_save_base_addr_hi =
+			upper_32_bits(q->ctx_save_restore_area_address);
+		m->cp_hqd_ctx_save_size = q->ctx_save_restore_area_size;
+		m->cp_hqd_cntl_stack_size = q->ctl_stack_size;
+		m->cp_hqd_cntl_stack_offset = q->ctl_stack_size;
+		m->cp_hqd_wg_state_offset = q->ctl_stack_size;
+	}
+
+	*mqd = m;
+	if (gart_addr)
+		*gart_addr = addr;
+	retval = mm->update_mqd(mm, m, q);
+
+	return retval;
+}
+
+static int load_mqd(struct mqd_manager *mm, void *mqd,
+			uint32_t pipe_id, uint32_t queue_id,
+			struct queue_properties *p, struct mm_struct *mms)
+{
+	/* AQL write pointer counts in 64B packets, PM4/CP counts in dwords. */
+	uint32_t wptr_shift = (p->format == KFD_QUEUE_FORMAT_AQL ? 4 : 0);
+
+	return mm->dev->kfd2kgd->hqd_load(mm->dev->kgd, mqd, pipe_id, queue_id,
+					  (uint32_t __user *)p->write_ptr,
+					  wptr_shift, 0, mms);
+}
+
+static int update_mqd(struct mqd_manager *mm, void *mqd,
+		      struct queue_properties *q)
+{
+	struct v9_mqd *m;
+
+	m = get_mqd(mqd);
+
+	m->cp_hqd_pq_control = 5 << CP_HQD_PQ_CONTROL__RPTR_BLOCK_SIZE__SHIFT;
+	m->cp_hqd_pq_control |= order_base_2(q->queue_size / 4) - 1;
+	pr_debug("cp_hqd_pq_control 0x%x\n", m->cp_hqd_pq_control);
+
+	m->cp_hqd_pq_base_lo = lower_32_bits((uint64_t)q->queue_address >> 8);
+	m->cp_hqd_pq_base_hi = upper_32_bits((uint64_t)q->queue_address >> 8);
+
+	m->cp_hqd_pq_rptr_report_addr_lo = lower_32_bits((uint64_t)q->read_ptr);
+	m->cp_hqd_pq_rptr_report_addr_hi = upper_32_bits((uint64_t)q->read_ptr);
+	m->cp_hqd_pq_wptr_poll_addr_lo = lower_32_bits((uint64_t)q->write_ptr);
+	m->cp_hqd_pq_wptr_poll_addr_hi = upper_32_bits((uint64_t)q->write_ptr);
+
+	m->cp_hqd_pq_doorbell_control =
+		q->doorbell_off <<
+			CP_HQD_PQ_DOORBELL_CONTROL__DOORBELL_OFFSET__SHIFT;
+	pr_debug("cp_hqd_pq_doorbell_control 0x%x\n",
+			m->cp_hqd_pq_doorbell_control);
+
+	m->cp_hqd_ib_control =
+		3 << CP_HQD_IB_CONTROL__MIN_IB_AVAIL_SIZE__SHIFT |
+		1 << CP_HQD_IB_CONTROL__IB_EXE_DISABLE__SHIFT;
+
+	/*
+	 * HW does not clamp this field correctly. Maximum EOP queue size
+	 * is constrained by per-SE EOP done signal count, which is 8-bit.
+	 * Limit is 0xFF EOP entries (= 0x7F8 dwords). CP will not submit
+	 * more than (EOP entry count - 1) so a queue size of 0x800 dwords
+	 * is safe, giving a maximum field value of 0xA.
+	 */
+	m->cp_hqd_eop_control = min(0xA,
+		order_base_2(q->eop_ring_buffer_size / 4) - 1);
+	m->cp_hqd_eop_base_addr_lo =
+			lower_32_bits(q->eop_ring_buffer_address >> 8);
+	m->cp_hqd_eop_base_addr_hi =
+			upper_32_bits(q->eop_ring_buffer_address >> 8);
+
+	m->cp_hqd_iq_timer = 0;
+
+	m->cp_hqd_vmid = q->vmid;
+
+	if (q->format == KFD_QUEUE_FORMAT_AQL) {
+		m->cp_hqd_pq_control |= CP_HQD_PQ_CONTROL__NO_UPDATE_RPTR_MASK |
+				2 << CP_HQD_PQ_CONTROL__SLOT_BASED_WPTR__SHIFT |
+				1 << CP_HQD_PQ_CONTROL__QUEUE_FULL_EN__SHIFT |
+				1 << CP_HQD_PQ_CONTROL__WPP_CLAMP_EN__SHIFT;
+		m->cp_hqd_pq_doorbell_control |= 1 <<
+			CP_HQD_PQ_DOORBELL_CONTROL__DOORBELL_BIF_DROP__SHIFT;
+	}
+	if (mm->dev->cwsr_enabled && q->ctx_save_restore_area_address)
+		m->cp_hqd_ctx_save_control = 0;
+
+	q->is_active = (q->queue_size > 0 &&
+			q->queue_address != 0 &&
+			q->queue_percent > 0 &&
+			!q->is_evicted);
+
+	return 0;
+}
+
+
+static int destroy_mqd(struct mqd_manager *mm, void *mqd,
+			enum kfd_preempt_type type,
+			unsigned int timeout, uint32_t pipe_id,
+			uint32_t queue_id)
+{
+	return mm->dev->kfd2kgd->hqd_destroy
+		(mm->dev->kgd, mqd, type, timeout,
+		pipe_id, queue_id);
+}
+
+static void uninit_mqd(struct mqd_manager *mm, void *mqd,
+			struct kfd_mem_obj *mqd_mem_obj)
+{
+	struct kfd_dev *kfd = mm->dev;
+
+	if (mqd_mem_obj->gtt_mem) {
+		kfd->kfd2kgd->free_gtt_mem(kfd->kgd, mqd_mem_obj->gtt_mem);
+		kfree(mqd_mem_obj);
+	} else {
+		kfd_gtt_sa_free(mm->dev, mqd_mem_obj);
+	}
+}
+
+static bool is_occupied(struct mqd_manager *mm, void *mqd,
+			uint64_t queue_address,	uint32_t pipe_id,
+			uint32_t queue_id)
+{
+	return mm->dev->kfd2kgd->hqd_is_occupied(
+		mm->dev->kgd, queue_address,
+		pipe_id, queue_id);
+}
+
+static int init_mqd_hiq(struct mqd_manager *mm, void **mqd,
+			struct kfd_mem_obj **mqd_mem_obj, uint64_t *gart_addr,
+			struct queue_properties *q)
+{
+	struct v9_mqd *m;
+	int retval = init_mqd(mm, mqd, mqd_mem_obj, gart_addr, q);
+
+	if (retval != 0)
+		return retval;
+
+	m = get_mqd(*mqd);
+
+	m->cp_hqd_pq_control |= 1 << CP_HQD_PQ_CONTROL__PRIV_STATE__SHIFT |
+			1 << CP_HQD_PQ_CONTROL__KMD_QUEUE__SHIFT;
+
+	return retval;
+}
+
+static int update_mqd_hiq(struct mqd_manager *mm, void *mqd,
+			struct queue_properties *q)
+{
+	struct v9_mqd *m;
+	int retval = update_mqd(mm, mqd, q);
+
+	if (retval != 0)
+		return retval;
+
+	/* TODO: what's the point? update_mqd already does this. */
+	m = get_mqd(mqd);
+	m->cp_hqd_vmid = q->vmid;
+	return retval;
+}
+
+static int init_mqd_sdma(struct mqd_manager *mm, void **mqd,
+		struct kfd_mem_obj **mqd_mem_obj, uint64_t *gart_addr,
+		struct queue_properties *q)
+{
+	int retval;
+	struct v9_sdma_mqd *m;
+
+
+	retval = kfd_gtt_sa_allocate(mm->dev,
+			sizeof(struct v9_sdma_mqd),
+			mqd_mem_obj);
+
+	if (retval != 0)
+		return -ENOMEM;
+
+	m = (struct v9_sdma_mqd *) (*mqd_mem_obj)->cpu_ptr;
+
+	memset(m, 0, sizeof(struct v9_sdma_mqd));
+
+	*mqd = m;
+	if (gart_addr)
+		*gart_addr = (*mqd_mem_obj)->gpu_addr;
+
+	retval = mm->update_mqd(mm, m, q);
+
+	return retval;
+}
+
+static void uninit_mqd_sdma(struct mqd_manager *mm, void *mqd,
+		struct kfd_mem_obj *mqd_mem_obj)
+{
+	kfd_gtt_sa_free(mm->dev, mqd_mem_obj);
+}
+
+static int load_mqd_sdma(struct mqd_manager *mm, void *mqd,
+		uint32_t pipe_id, uint32_t queue_id,
+		struct queue_properties *p, struct mm_struct *mms)
+{
+	return mm->dev->kfd2kgd->hqd_sdma_load(mm->dev->kgd, mqd,
+					       (uint32_t __user *)p->write_ptr,
+					       mms);
+}
+
+#define SDMA_RLC_DUMMY_DEFAULT 0xf
+
+static int update_mqd_sdma(struct mqd_manager *mm, void *mqd,
+		struct queue_properties *q)
+{
+	struct v9_sdma_mqd *m;
+
+	m = get_sdma_mqd(mqd);
+	m->sdmax_rlcx_rb_cntl = order_base_2(q->queue_size / 4)
+		<< SDMA0_RLC0_RB_CNTL__RB_SIZE__SHIFT |
+		q->vmid << SDMA0_RLC0_RB_CNTL__RB_VMID__SHIFT |
+		1 << SDMA0_RLC0_RB_CNTL__RPTR_WRITEBACK_ENABLE__SHIFT |
+		6 << SDMA0_RLC0_RB_CNTL__RPTR_WRITEBACK_TIMER__SHIFT;
+
+	m->sdmax_rlcx_rb_base = lower_32_bits(q->queue_address >> 8);
+	m->sdmax_rlcx_rb_base_hi = upper_32_bits(q->queue_address >> 8);
+	m->sdmax_rlcx_rb_rptr_addr_lo = lower_32_bits((uint64_t)q->read_ptr);
+	m->sdmax_rlcx_rb_rptr_addr_hi = upper_32_bits((uint64_t)q->read_ptr);
+	m->sdmax_rlcx_doorbell_offset =
+		q->doorbell_off << SDMA0_RLC0_DOORBELL_OFFSET__OFFSET__SHIFT;
+
+	m->sdma_engine_id = q->sdma_engine_id;
+	m->sdma_queue_id = q->sdma_queue_id;
+	m->sdmax_rlcx_dummy_reg = SDMA_RLC_DUMMY_DEFAULT;
+
+	q->is_active = (q->queue_size > 0 &&
+			q->queue_address != 0 &&
+			q->queue_percent > 0 &&
+			!q->is_evicted);
+
+	return 0;
+}
+
+/*
+ *  * preempt type here is ignored because there is only one way
+ *  * to preempt sdma queue
+ */
+static int destroy_mqd_sdma(struct mqd_manager *mm, void *mqd,
+		enum kfd_preempt_type type,
+		unsigned int timeout, uint32_t pipe_id,
+		uint32_t queue_id)
+{
+	return mm->dev->kfd2kgd->hqd_sdma_destroy(mm->dev->kgd, mqd, timeout);
+}
+
+static bool is_occupied_sdma(struct mqd_manager *mm, void *mqd,
+		uint64_t queue_address, uint32_t pipe_id,
+		uint32_t queue_id)
+{
+	return mm->dev->kfd2kgd->hqd_sdma_is_occupied(mm->dev->kgd, mqd);
+}
+
+#if defined(CONFIG_DEBUG_FS)
+
+static int debugfs_show_mqd(struct seq_file *m, void *data)
+{
+	seq_hex_dump(m, "    ", DUMP_PREFIX_OFFSET, 32, 4,
+		     data, sizeof(struct v9_mqd), false);
+	return 0;
+}
+
+static int debugfs_show_mqd_sdma(struct seq_file *m, void *data)
+{
+	seq_hex_dump(m, "    ", DUMP_PREFIX_OFFSET, 32, 4,
+		     data, sizeof(struct v9_sdma_mqd), false);
+	return 0;
+}
+
+#endif
+
+struct mqd_manager *mqd_manager_init_v9(enum KFD_MQD_TYPE type,
+		struct kfd_dev *dev)
+{
+	struct mqd_manager *mqd;
+
+	if (WARN_ON(type >= KFD_MQD_TYPE_MAX))
+		return NULL;
+
+	mqd = kzalloc(sizeof(*mqd), GFP_NOIO);
+	if (!mqd)
+		return NULL;
+
+	mqd->dev = dev;
+
+	switch (type) {
+	case KFD_MQD_TYPE_CP:
+	case KFD_MQD_TYPE_COMPUTE:
+		mqd->init_mqd = init_mqd;
+		mqd->uninit_mqd = uninit_mqd;
+		mqd->load_mqd = load_mqd;
+		mqd->update_mqd = update_mqd;
+		mqd->destroy_mqd = destroy_mqd;
+		mqd->is_occupied = is_occupied;
+#if defined(CONFIG_DEBUG_FS)
+		mqd->debugfs_show_mqd = debugfs_show_mqd;
+#endif
+		break;
+	case KFD_MQD_TYPE_HIQ:
+		mqd->init_mqd = init_mqd_hiq;
+		mqd->uninit_mqd = uninit_mqd;
+		mqd->load_mqd = load_mqd;
+		mqd->update_mqd = update_mqd_hiq;
+		mqd->destroy_mqd = destroy_mqd;
+		mqd->is_occupied = is_occupied;
+#if defined(CONFIG_DEBUG_FS)
+		mqd->debugfs_show_mqd = debugfs_show_mqd;
+#endif
+		break;
+	case KFD_MQD_TYPE_SDMA:
+		mqd->init_mqd = init_mqd_sdma;
+		mqd->uninit_mqd = uninit_mqd_sdma;
+		mqd->load_mqd = load_mqd_sdma;
+		mqd->update_mqd = update_mqd_sdma;
+		mqd->destroy_mqd = destroy_mqd_sdma;
+		mqd->is_occupied = is_occupied_sdma;
+#if defined(CONFIG_DEBUG_FS)
+		mqd->debugfs_show_mqd = debugfs_show_mqd_sdma;
+#endif
+		break;
+	default:
+		kfree(mqd);
+		return NULL;
+	}
+
+	return mqd;
+}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
index b68299a..fac2882 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
@@ -197,6 +197,7 @@ struct kfd_mem_obj {
 	uint32_t range_end;
 	uint64_t gpu_addr;
 	uint32_t *cpu_ptr;
+	void *gtt_mem;
 };
 
 struct kfd_vmid_info {
@@ -822,6 +823,8 @@ struct mqd_manager *mqd_manager_init_vi(enum KFD_MQD_TYPE type,
 		struct kfd_dev *dev);
 struct mqd_manager *mqd_manager_init_vi_tonga(enum KFD_MQD_TYPE type,
 		struct kfd_dev *dev);
+struct mqd_manager *mqd_manager_init_v9(enum KFD_MQD_TYPE type,
+		struct kfd_dev *dev);
 struct device_queue_manager *device_queue_manager_init(struct kfd_dev *dev);
 void device_queue_manager_uninit(struct device_queue_manager *dqm);
 struct kernel_queue *kernel_queue_init(struct kfd_dev *dev,
-- 
cgit v1.1


From bed4f110251b4f9041e5e797e035bc40c34d60ea Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Tue, 10 Apr 2018 17:33:09 -0400
Subject: drm/amdkfd: Add GFXv9 device queue manager

Signed-off-by: John Bridgman <john.bridgman@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/Makefile                |  2 +-
 .../gpu/drm/amd/amdkfd/kfd_device_queue_manager.c  | 10 ++-
 .../gpu/drm/amd/amdkfd/kfd_device_queue_manager.h  |  2 +
 .../drm/amd/amdkfd/kfd_device_queue_manager_v9.c   | 84 ++++++++++++++++++++++
 drivers/gpu/drm/amd/amdkfd/kfd_module.c            |  5 ++
 drivers/gpu/drm/amd/amdkfd/kfd_priv.h              |  5 ++
 6 files changed, 106 insertions(+), 2 deletions(-)
 create mode 100644 drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_v9.c

diff --git a/drivers/gpu/drm/amd/amdkfd/Makefile b/drivers/gpu/drm/amd/amdkfd/Makefile
index 094b591..ff8b5aa 100644
--- a/drivers/gpu/drm/amd/amdkfd/Makefile
+++ b/drivers/gpu/drm/amd/amdkfd/Makefile
@@ -35,7 +35,7 @@ amdkfd-y	:= kfd_module.o kfd_device.o kfd_chardev.o kfd_topology.o \
 		kfd_kernel_queue_vi.o kfd_kernel_queue_v9.o \
 		kfd_packet_manager.o kfd_process_queue_manager.o \
 		kfd_device_queue_manager.o kfd_device_queue_manager_cik.o \
-		kfd_device_queue_manager_vi.o \
+		kfd_device_queue_manager_vi.o kfd_device_queue_manager_v9.o \
 		kfd_interrupt.o kfd_events.o cik_event_interrupt.o \
 		kfd_dbgdev.o kfd_dbgmgr.o kfd_crat.o
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
index 500f022..9af94b1 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
@@ -1386,7 +1386,10 @@ static bool set_cache_memory_policy(struct device_queue_manager *dqm,
 				   void __user *alternate_aperture_base,
 				   uint64_t alternate_aperture_size)
 {
-	bool retval;
+	bool retval = true;
+
+	if (!dqm->asic_ops.set_cache_memory_policy)
+		return retval;
 
 	mutex_lock(&dqm->lock);
 
@@ -1655,6 +1658,11 @@ struct device_queue_manager *device_queue_manager_init(struct kfd_dev *dev)
 	case CHIP_POLARIS11:
 		device_queue_manager_init_vi_tonga(&dqm->asic_ops);
 		break;
+
+	case CHIP_VEGA10:
+	case CHIP_RAVEN:
+		device_queue_manager_init_v9(&dqm->asic_ops);
+		break;
 	default:
 		WARN(1, "Unexpected ASIC family %u",
 		     dev->device_info->asic_family);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h
index 412beff..59a6b19 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h
@@ -200,6 +200,8 @@ void device_queue_manager_init_vi(
 		struct device_queue_manager_asic_ops *asic_ops);
 void device_queue_manager_init_vi_tonga(
 		struct device_queue_manager_asic_ops *asic_ops);
+void device_queue_manager_init_v9(
+		struct device_queue_manager_asic_ops *asic_ops);
 void program_sh_mem_settings(struct device_queue_manager *dqm,
 					struct qcm_process_device *qpd);
 unsigned int get_queues_num(struct device_queue_manager *dqm);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_v9.c
new file mode 100644
index 0000000..79e5bcf
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_v9.c
@@ -0,0 +1,84 @@
+/*
+ * Copyright 2016-2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include "kfd_device_queue_manager.h"
+#include "vega10_enum.h"
+#include "gc/gc_9_0_offset.h"
+#include "gc/gc_9_0_sh_mask.h"
+#include "sdma0/sdma0_4_0_sh_mask.h"
+
+static int update_qpd_v9(struct device_queue_manager *dqm,
+			 struct qcm_process_device *qpd);
+static void init_sdma_vm_v9(struct device_queue_manager *dqm, struct queue *q,
+			    struct qcm_process_device *qpd);
+
+void device_queue_manager_init_v9(
+	struct device_queue_manager_asic_ops *asic_ops)
+{
+	asic_ops->update_qpd = update_qpd_v9;
+	asic_ops->init_sdma_vm = init_sdma_vm_v9;
+}
+
+static uint32_t compute_sh_mem_bases_64bit(struct kfd_process_device *pdd)
+{
+	uint32_t shared_base = pdd->lds_base >> 48;
+	uint32_t private_base = pdd->scratch_base >> 48;
+
+	return (shared_base << SH_MEM_BASES__SHARED_BASE__SHIFT) |
+		private_base;
+}
+
+static int update_qpd_v9(struct device_queue_manager *dqm,
+			 struct qcm_process_device *qpd)
+{
+	struct kfd_process_device *pdd;
+
+	pdd = qpd_to_pdd(qpd);
+
+	/* check if sh_mem_config register already configured */
+	if (qpd->sh_mem_config == 0) {
+		qpd->sh_mem_config =
+				SH_MEM_ALIGNMENT_MODE_UNALIGNED <<
+					SH_MEM_CONFIG__ALIGNMENT_MODE__SHIFT;
+		if (vega10_noretry &&
+		    !dqm->dev->device_info->needs_iommu_device)
+			qpd->sh_mem_config |=
+				1 << SH_MEM_CONFIG__RETRY_DISABLE__SHIFT;
+
+		qpd->sh_mem_ape1_limit = 0;
+		qpd->sh_mem_ape1_base = 0;
+	}
+
+	qpd->sh_mem_bases = compute_sh_mem_bases_64bit(pdd);
+
+	pr_debug("sh_mem_bases 0x%X\n", qpd->sh_mem_bases);
+
+	return 0;
+}
+
+static void init_sdma_vm_v9(struct device_queue_manager *dqm, struct queue *q,
+			    struct qcm_process_device *qpd)
+{
+	/* Not needed on SDMAv4 any more */
+	q->properties.sdma_vm_addr = 0;
+}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_module.c b/drivers/gpu/drm/amd/amdkfd/kfd_module.c
index 45bc458..76bf2dc 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_module.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_module.c
@@ -83,6 +83,11 @@ module_param(ignore_crat, int, 0444);
 MODULE_PARM_DESC(ignore_crat,
 	"Ignore CRAT table during KFD initialization (0 = use CRAT (default), 1 = ignore CRAT)");
 
+int vega10_noretry;
+module_param_named(noretry, vega10_noretry, int, 0644);
+MODULE_PARM_DESC(noretry,
+	"Set sh_mem_config.retry_disable on Vega10 (0 = retry enabled (default), 1 = retry disabled)");
+
 static int amdkfd_init_completed;
 
 int kgd2kfd_init(unsigned int interface_version,
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
index fac2882..d5cdb5d 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
@@ -137,6 +137,11 @@ extern int debug_largebar;
  */
 extern int ignore_crat;
 
+/*
+ * Set sh_mem_config.retry_disable on Vega10
+ */
+extern int vega10_noretry;
+
 /**
  * enum kfd_sched_policy
  *
-- 
cgit v1.1


From ca750681bc4a897ffa7eed71a1e05762fb1f0a34 Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Tue, 10 Apr 2018 17:33:10 -0400
Subject: drm/amdkfd: Add SOC15 interrupt processing support

Signed-off-by: Shaoyun Liu <Shaoyun.Liu@amd.com>
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/Makefile             |  2 +-
 drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c | 84 +++++++++++++++++++++++++
 drivers/gpu/drm/amd/amdkfd/kfd_priv.h           |  2 +
 drivers/gpu/drm/amd/amdkfd/soc15_int.h          | 47 ++++++++++++++
 4 files changed, 134 insertions(+), 1 deletion(-)
 create mode 100644 drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c
 create mode 100644 drivers/gpu/drm/amd/amdkfd/soc15_int.h

diff --git a/drivers/gpu/drm/amd/amdkfd/Makefile b/drivers/gpu/drm/amd/amdkfd/Makefile
index ff8b5aa..ffd096f 100644
--- a/drivers/gpu/drm/amd/amdkfd/Makefile
+++ b/drivers/gpu/drm/amd/amdkfd/Makefile
@@ -37,7 +37,7 @@ amdkfd-y	:= kfd_module.o kfd_device.o kfd_chardev.o kfd_topology.o \
 		kfd_device_queue_manager.o kfd_device_queue_manager_cik.o \
 		kfd_device_queue_manager_vi.o kfd_device_queue_manager_v9.o \
 		kfd_interrupt.o kfd_events.o cik_event_interrupt.o \
-		kfd_dbgdev.o kfd_dbgmgr.o kfd_crat.o
+		kfd_int_process_v9.o kfd_dbgdev.o kfd_dbgmgr.o kfd_crat.o
 
 ifneq ($(CONFIG_AMD_IOMMU_V2),)
 amdkfd-y += kfd_iommu.o
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c
new file mode 100644
index 0000000..39d4115
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c
@@ -0,0 +1,84 @@
+/*
+ * Copyright 2016-2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#include "kfd_priv.h"
+#include "kfd_events.h"
+#include "soc15_int.h"
+
+
+static bool event_interrupt_isr_v9(struct kfd_dev *dev,
+					const uint32_t *ih_ring_entry)
+{
+	uint16_t source_id, client_id, pasid, vmid;
+
+	source_id = SOC15_SOURCE_ID_FROM_IH_ENTRY(ih_ring_entry);
+	client_id = SOC15_CLIENT_ID_FROM_IH_ENTRY(ih_ring_entry);
+	pasid = SOC15_PASID_FROM_IH_ENTRY(ih_ring_entry);
+	vmid = SOC15_VMID_FROM_IH_ENTRY(ih_ring_entry);
+
+	if (pasid) {
+		const uint32_t *data = ih_ring_entry;
+
+		pr_debug("client id 0x%x, source id %d, pasid 0x%x. raw data:\n",
+			 client_id, source_id, pasid);
+		pr_debug("%8X, %8X, %8X, %8X, %8X, %8X, %8X, %8X.\n",
+			 data[0], data[1], data[2], data[3],
+			 data[4], data[5], data[6], data[7]);
+	}
+
+	return (pasid != 0) &&
+		(source_id == SOC15_INTSRC_CP_END_OF_PIPE ||
+		 source_id == SOC15_INTSRC_SDMA_TRAP ||
+		 source_id == SOC15_INTSRC_SQ_INTERRUPT_MSG ||
+		 source_id == SOC15_INTSRC_CP_BAD_OPCODE);
+}
+
+static void event_interrupt_wq_v9(struct kfd_dev *dev,
+					const uint32_t *ih_ring_entry)
+{
+	uint16_t source_id, client_id, pasid, vmid;
+	uint32_t context_id;
+
+	source_id = SOC15_SOURCE_ID_FROM_IH_ENTRY(ih_ring_entry);
+	client_id = SOC15_CLIENT_ID_FROM_IH_ENTRY(ih_ring_entry);
+	pasid = SOC15_PASID_FROM_IH_ENTRY(ih_ring_entry);
+	vmid = SOC15_VMID_FROM_IH_ENTRY(ih_ring_entry);
+	context_id = SOC15_CONTEXT_ID0_FROM_IH_ENTRY(ih_ring_entry);
+
+	if (source_id == SOC15_INTSRC_CP_END_OF_PIPE)
+		kfd_signal_event_interrupt(pasid, context_id, 32);
+	else if (source_id == SOC15_INTSRC_SDMA_TRAP)
+		kfd_signal_event_interrupt(pasid, context_id & 0xfffffff, 28);
+	else if (source_id == SOC15_INTSRC_SQ_INTERRUPT_MSG)
+		kfd_signal_event_interrupt(pasid, context_id & 0xffffff, 24);
+	else if (source_id == SOC15_INTSRC_CP_BAD_OPCODE)
+		kfd_signal_hw_exception_event(pasid);
+	else if (client_id == SOC15_IH_CLIENTID_VMC ||
+		 client_id == SOC15_IH_CLIENTID_UTCL2) {
+		/* TODO */
+	}
+}
+
+const struct kfd_event_interrupt_class event_interrupt_class_v9 = {
+	.interrupt_isr = event_interrupt_isr_v9,
+	.interrupt_wq = event_interrupt_wq_v9,
+};
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
index d5cdb5d..06b210b 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
@@ -934,6 +934,8 @@ uint64_t kfd_get_number_elems(struct kfd_dev *kfd);
 
 /* Events */
 extern const struct kfd_event_interrupt_class event_interrupt_class_cik;
+extern const struct kfd_event_interrupt_class event_interrupt_class_v9;
+
 extern const struct kfd_device_global_init_class device_global_init_class_cik;
 
 void kfd_event_init_process(struct kfd_process *p);
diff --git a/drivers/gpu/drm/amd/amdkfd/soc15_int.h b/drivers/gpu/drm/amd/amdkfd/soc15_int.h
new file mode 100644
index 0000000..0bc0b25
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/soc15_int.h
@@ -0,0 +1,47 @@
+/*
+ * Copyright 2016-2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#ifndef HSA_SOC15_INT_H_INCLUDED
+#define HSA_SOC15_INT_H_INCLUDED
+
+#include "soc15_ih_clientid.h"
+
+#define SOC15_INTSRC_CP_END_OF_PIPE	181
+#define SOC15_INTSRC_CP_BAD_OPCODE	183
+#define SOC15_INTSRC_SQ_INTERRUPT_MSG	239
+#define SOC15_INTSRC_VMC_FAULT		0
+#define SOC15_INTSRC_SDMA_TRAP		224
+
+
+#define SOC15_CLIENT_ID_FROM_IH_ENTRY(entry) (le32_to_cpu(entry[0]) & 0xff)
+#define SOC15_SOURCE_ID_FROM_IH_ENTRY(entry) (le32_to_cpu(entry[0]) >> 8 & 0xff)
+#define SOC15_RING_ID_FROM_IH_ENTRY(entry) (le32_to_cpu(entry[0]) >> 16 & 0xff)
+#define SOC15_VMID_FROM_IH_ENTRY(entry) (le32_to_cpu(entry[0]) >> 24 & 0xf)
+#define SOC15_VMID_TYPE_FROM_IH_ENTRY(entry) (le32_to_cpu(entry[0]) >> 31 & 0x1)
+#define SOC15_PASID_FROM_IH_ENTRY(entry) (le32_to_cpu(entry[3]) & 0xffff)
+#define SOC15_CONTEXT_ID0_FROM_IH_ENTRY(entry) (le32_to_cpu(entry[4]))
+#define SOC15_CONTEXT_ID1_FROM_IH_ENTRY(entry) (le32_to_cpu(entry[5]))
+#define SOC15_CONTEXT_ID2_FROM_IH_ENTRY(entry) (le32_to_cpu(entry[6]))
+#define SOC15_CONTEXT_ID3_FROM_IH_ENTRY(entry) (le32_to_cpu(entry[7]))
+
+#endif
+
-- 
cgit v1.1


From 2a26fbfe80015faef830bc47c5223b4b31d41791 Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Tue, 10 Apr 2018 17:33:11 -0400
Subject: drm/amdkfd: Fix goto usage

Missed a spot in previous cleanup commit:
Remove gotos that do not feature any common cleanup, and use gotos
instead of repeating cleanup commands.

According to kernel.org: "The goto statement comes in handy when a
function exits from multiple locations and some common work such as
cleanup has to be done. If there is no cleanup needed then just return
directly."

Signed-off-by: Kent Russell <kent.russell@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c
index 69f4964..23e586b 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c
@@ -232,18 +232,16 @@ static int acquire_packet_buffer(struct kernel_queue *kq,
 		 * make sure calling functions know
 		 * acquire_packet_buffer() failed
 		 */
-		*buffer_ptr = NULL;
-		return -ENOMEM;
+		goto err_no_space;
 	}
 
 	if (wptr + packet_size_in_dwords >= queue_size_dwords) {
 		/* make sure after rolling back to position 0, there is
 		 * still enough space.
 		 */
-		if (packet_size_in_dwords >= rptr) {
-			*buffer_ptr = NULL;
-			return -ENOMEM;
-		}
+		if (packet_size_in_dwords >= rptr)
+			goto err_no_space;
+
 		/* fill nops, roll back and start at position 0 */
 		while (wptr > 0) {
 			queue_address[wptr] = kq->nop_packet;
@@ -255,6 +253,10 @@ static int acquire_packet_buffer(struct kernel_queue *kq,
 	kq->pending_wptr = wptr + packet_size_in_dwords;
 
 	return 0;
+
+err_no_space:
+	*buffer_ptr = NULL;
+	return -ENOMEM;
 }
 
 static void submit_packet(struct kernel_queue *kq)
-- 
cgit v1.1


From bebfd2f4126a115420a2b04f44a05552c12e5b46 Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Tue, 10 Apr 2018 17:33:12 -0400
Subject: drm/amdkfd: Fix kernel queue rollback_packet

kq->queue->properties.write_ptr is a GPU address which can'd be
derefenced in the kernel. Use kq->wptr_kernel instead, which is the
kernel CPU address of the same buffer.

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c
index 23e586b..9f38161 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c
@@ -279,7 +279,7 @@ static void submit_packet(struct kernel_queue *kq)
 
 static void rollback_packet(struct kernel_queue *kq)
 {
-	kq->pending_wptr = *kq->queue->properties.write_ptr;
+	kq->pending_wptr = *kq->wptr_kernel;
 }
 
 struct kernel_queue *kernel_queue_init(struct kfd_dev *dev,
-- 
cgit v1.1


From 9d7d024816686f922735f7adccd00e3fc44e2e03 Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Sun, 8 Apr 2018 22:03:51 -0400
Subject: drm/amdkfd: Add 64-bit doorbell and wptr support to kernel queue

v2: Removed redundant 0x before %p.

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c         | 10 +++++++++
 drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c     | 25 +++++++++++++++++------
 drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.h     |  7 ++++++-
 drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_cik.c |  9 ++++++++
 drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_v9.c  |  9 ++++++++
 drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_vi.c  |  9 ++++++++
 drivers/gpu/drm/amd/amdkfd/kfd_priv.h             |  1 +
 7 files changed, 63 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c b/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c
index 36c9269e..c3744d8 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c
@@ -214,6 +214,16 @@ void write_kernel_doorbell(void __iomem *db, u32 value)
 	}
 }
 
+void write_kernel_doorbell64(void __iomem *db, u64 value)
+{
+	if (db) {
+		WARN(((unsigned long)db & 7) != 0,
+		     "Unaligned 64-bit doorbell");
+		writeq(value, (u64 __iomem *)db);
+		pr_debug("writing %llu to doorbell address %p\n", value, db);
+	}
+}
+
 unsigned int kfd_doorbell_id_to_offset(struct kfd_dev *kfd,
 					struct kfd_process *process,
 					unsigned int doorbell_id)
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c
index 9f38161..476951d 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c
@@ -99,7 +99,7 @@ static bool initialize(struct kernel_queue *kq, struct kfd_dev *dev,
 	kq->rptr_kernel = kq->rptr_mem->cpu_ptr;
 	kq->rptr_gpu_addr = kq->rptr_mem->gpu_addr;
 
-	retval = kfd_gtt_sa_allocate(dev, sizeof(*kq->wptr_kernel),
+	retval = kfd_gtt_sa_allocate(dev, dev->device_info->doorbell_size,
 					&kq->wptr_mem);
 
 	if (retval != 0)
@@ -208,6 +208,7 @@ static int acquire_packet_buffer(struct kernel_queue *kq,
 	size_t available_size;
 	size_t queue_size_dwords;
 	uint32_t wptr, rptr;
+	uint64_t wptr64;
 	unsigned int *queue_address;
 
 	/* When rptr == wptr, the buffer is empty.
@@ -216,7 +217,8 @@ static int acquire_packet_buffer(struct kernel_queue *kq,
 	 * the opposite. So we can only use up to queue_size_dwords - 1 dwords.
 	 */
 	rptr = *kq->rptr_kernel;
-	wptr = *kq->wptr_kernel;
+	wptr = kq->pending_wptr;
+	wptr64 = kq->pending_wptr64;
 	queue_address = (unsigned int *)kq->pq_kernel_addr;
 	queue_size_dwords = kq->queue->properties.queue_size / 4;
 
@@ -246,11 +248,13 @@ static int acquire_packet_buffer(struct kernel_queue *kq,
 		while (wptr > 0) {
 			queue_address[wptr] = kq->nop_packet;
 			wptr = (wptr + 1) % queue_size_dwords;
+			wptr64++;
 		}
 	}
 
 	*buffer_ptr = &queue_address[wptr];
 	kq->pending_wptr = wptr + packet_size_in_dwords;
+	kq->pending_wptr64 = wptr64 + packet_size_in_dwords;
 
 	return 0;
 
@@ -272,14 +276,18 @@ static void submit_packet(struct kernel_queue *kq)
 	pr_debug("\n");
 #endif
 
-	*kq->wptr_kernel = kq->pending_wptr;
-	write_kernel_doorbell(kq->queue->properties.doorbell_ptr,
-				kq->pending_wptr);
+	kq->ops_asic_specific.submit_packet(kq);
 }
 
 static void rollback_packet(struct kernel_queue *kq)
 {
-	kq->pending_wptr = *kq->wptr_kernel;
+	if (kq->dev->device_info->doorbell_size == 8) {
+		kq->pending_wptr64 = *kq->wptr64_kernel;
+		kq->pending_wptr = *kq->wptr_kernel %
+			(kq->queue->properties.queue_size / 4);
+	} else {
+		kq->pending_wptr = *kq->wptr_kernel;
+	}
 }
 
 struct kernel_queue *kernel_queue_init(struct kfd_dev *dev,
@@ -310,6 +318,11 @@ struct kernel_queue *kernel_queue_init(struct kfd_dev *dev,
 	case CHIP_HAWAII:
 		kernel_queue_init_cik(&kq->ops_asic_specific);
 		break;
+
+	case CHIP_VEGA10:
+	case CHIP_RAVEN:
+		kernel_queue_init_v9(&kq->ops_asic_specific);
+		break;
 	default:
 		WARN(1, "Unexpected ASIC family %u",
 		     dev->device_info->asic_family);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.h b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.h
index 5940531..97aff20 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.h
@@ -72,6 +72,7 @@ struct kernel_queue {
 	struct kfd_dev		*dev;
 	struct mqd_manager	*mqd;
 	struct queue		*queue;
+	uint64_t		pending_wptr64;
 	uint32_t		pending_wptr;
 	unsigned int		nop_packet;
 
@@ -79,7 +80,10 @@ struct kernel_queue {
 	uint32_t		*rptr_kernel;
 	uint64_t		rptr_gpu_addr;
 	struct kfd_mem_obj	*wptr_mem;
-	uint32_t		*wptr_kernel;
+	union {
+		uint64_t	*wptr64_kernel;
+		uint32_t	*wptr_kernel;
+	};
 	uint64_t		wptr_gpu_addr;
 	struct kfd_mem_obj	*pq;
 	uint64_t		pq_gpu_addr;
@@ -97,5 +101,6 @@ struct kernel_queue {
 
 void kernel_queue_init_cik(struct kernel_queue_ops *ops);
 void kernel_queue_init_vi(struct kernel_queue_ops *ops);
+void kernel_queue_init_v9(struct kernel_queue_ops *ops);
 
 #endif /* KFD_KERNEL_QUEUE_H_ */
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_cik.c b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_cik.c
index a90eb44..19e54ac 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_cik.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_cik.c
@@ -26,11 +26,13 @@
 static bool initialize_cik(struct kernel_queue *kq, struct kfd_dev *dev,
 			enum kfd_queue_type type, unsigned int queue_size);
 static void uninitialize_cik(struct kernel_queue *kq);
+static void submit_packet_cik(struct kernel_queue *kq);
 
 void kernel_queue_init_cik(struct kernel_queue_ops *ops)
 {
 	ops->initialize = initialize_cik;
 	ops->uninitialize = uninitialize_cik;
+	ops->submit_packet = submit_packet_cik;
 }
 
 static bool initialize_cik(struct kernel_queue *kq, struct kfd_dev *dev,
@@ -42,3 +44,10 @@ static bool initialize_cik(struct kernel_queue *kq, struct kfd_dev *dev,
 static void uninitialize_cik(struct kernel_queue *kq)
 {
 }
+
+static void submit_packet_cik(struct kernel_queue *kq)
+{
+	*kq->wptr_kernel = kq->pending_wptr;
+	write_kernel_doorbell(kq->queue->properties.doorbell_ptr,
+				kq->pending_wptr);
+}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_v9.c
index ece7d59..684a3bf 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_v9.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_v9.c
@@ -29,11 +29,13 @@
 static bool initialize_v9(struct kernel_queue *kq, struct kfd_dev *dev,
 			enum kfd_queue_type type, unsigned int queue_size);
 static void uninitialize_v9(struct kernel_queue *kq);
+static void submit_packet_v9(struct kernel_queue *kq);
 
 void kernel_queue_init_v9(struct kernel_queue_ops *ops)
 {
 	ops->initialize = initialize_v9;
 	ops->uninitialize = uninitialize_v9;
+	ops->submit_packet = submit_packet_v9;
 }
 
 static bool initialize_v9(struct kernel_queue *kq, struct kfd_dev *dev,
@@ -58,6 +60,13 @@ static void uninitialize_v9(struct kernel_queue *kq)
 	kfd_gtt_sa_free(kq->dev, kq->eop_mem);
 }
 
+static void submit_packet_v9(struct kernel_queue *kq)
+{
+	*kq->wptr64_kernel = kq->pending_wptr64;
+	write_kernel_doorbell64(kq->queue->properties.doorbell_ptr,
+				kq->pending_wptr64);
+}
+
 static int pm_map_process_v9(struct packet_manager *pm,
 		uint32_t *buffer, struct qcm_process_device *qpd)
 {
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_vi.c b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_vi.c
index f9019ef..bf20c6d 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_vi.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_vi.c
@@ -29,11 +29,13 @@
 static bool initialize_vi(struct kernel_queue *kq, struct kfd_dev *dev,
 			enum kfd_queue_type type, unsigned int queue_size);
 static void uninitialize_vi(struct kernel_queue *kq);
+static void submit_packet_vi(struct kernel_queue *kq);
 
 void kernel_queue_init_vi(struct kernel_queue_ops *ops)
 {
 	ops->initialize = initialize_vi;
 	ops->uninitialize = uninitialize_vi;
+	ops->submit_packet = submit_packet_vi;
 }
 
 static bool initialize_vi(struct kernel_queue *kq, struct kfd_dev *dev,
@@ -58,6 +60,13 @@ static void uninitialize_vi(struct kernel_queue *kq)
 	kfd_gtt_sa_free(kq->dev, kq->eop_mem);
 }
 
+static void submit_packet_vi(struct kernel_queue *kq)
+{
+	*kq->wptr_kernel = kq->pending_wptr;
+	write_kernel_doorbell(kq->queue->properties.doorbell_ptr,
+				kq->pending_wptr);
+}
+
 unsigned int pm_build_pm4_header(unsigned int opcode, size_t packet_size)
 {
 	union PM4_MES_TYPE_3_HEADER header;
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
index 06b210b..10d5b54 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
@@ -769,6 +769,7 @@ void __iomem *kfd_get_kernel_doorbell(struct kfd_dev *kfd,
 void kfd_release_kernel_doorbell(struct kfd_dev *kfd, u32 __iomem *db_addr);
 u32 read_kernel_doorbell(u32 __iomem *db);
 void write_kernel_doorbell(void __iomem *db, u32 value);
+void write_kernel_doorbell64(void __iomem *db, u64 value);
 unsigned int kfd_doorbell_id_to_offset(struct kfd_dev *kfd,
 					struct kfd_process *process,
 					unsigned int doorbell_id);
-- 
cgit v1.1


From 4b66af2d6356a00e94bcdea3e7fea324e8b5c6f4 Mon Sep 17 00:00:00 2001
From: Kevin Easton <kevin@guarana.org>
Date: Sat, 7 Apr 2018 11:40:33 -0400
Subject: af_key: Always verify length of provided sadb_key

Key extensions (struct sadb_key) include a user-specified number of key
bits.  The kernel uses that number to determine how much key data to copy
out of the message in pfkey_msg2xfrm_state().

The length of the sadb_key message must be verified to be long enough,
even in the case of SADB_X_AALG_NULL.  Furthermore, the sadb_key_len value
must be long enough to include both the key data and the struct sadb_key
itself.

Introduce a helper function verify_key_len(), and call it from
parse_exthdrs() where other exthdr types are similarly checked for
correctness.

Signed-off-by: Kevin Easton <kevin@guarana.org>
Reported-by: syzbot+5022a34ca5a3d49b84223653fab632dfb7b4cf37@syzkaller.appspotmail.com
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
---
 net/key/af_key.c | 45 +++++++++++++++++++++++++++++++++++----------
 1 file changed, 35 insertions(+), 10 deletions(-)

diff --git a/net/key/af_key.c b/net/key/af_key.c
index 7e2e718..e62e52e 100644
--- a/net/key/af_key.c
+++ b/net/key/af_key.c
@@ -437,6 +437,24 @@ static int verify_address_len(const void *p)
 	return 0;
 }
 
+static inline int sadb_key_len(const struct sadb_key *key)
+{
+	int key_bytes = DIV_ROUND_UP(key->sadb_key_bits, 8);
+
+	return DIV_ROUND_UP(sizeof(struct sadb_key) + key_bytes,
+			    sizeof(uint64_t));
+}
+
+static int verify_key_len(const void *p)
+{
+	const struct sadb_key *key = p;
+
+	if (sadb_key_len(key) > key->sadb_key_len)
+		return -EINVAL;
+
+	return 0;
+}
+
 static inline int pfkey_sec_ctx_len(const struct sadb_x_sec_ctx *sec_ctx)
 {
 	return DIV_ROUND_UP(sizeof(struct sadb_x_sec_ctx) +
@@ -533,16 +551,25 @@ static int parse_exthdrs(struct sk_buff *skb, const struct sadb_msg *hdr, void *
 				return -EINVAL;
 			if (ext_hdrs[ext_type-1] != NULL)
 				return -EINVAL;
-			if (ext_type == SADB_EXT_ADDRESS_SRC ||
-			    ext_type == SADB_EXT_ADDRESS_DST ||
-			    ext_type == SADB_EXT_ADDRESS_PROXY ||
-			    ext_type == SADB_X_EXT_NAT_T_OA) {
+			switch (ext_type) {
+			case SADB_EXT_ADDRESS_SRC:
+			case SADB_EXT_ADDRESS_DST:
+			case SADB_EXT_ADDRESS_PROXY:
+			case SADB_X_EXT_NAT_T_OA:
 				if (verify_address_len(p))
 					return -EINVAL;
-			}
-			if (ext_type == SADB_X_EXT_SEC_CTX) {
+				break;
+			case SADB_X_EXT_SEC_CTX:
 				if (verify_sec_ctx_len(p))
 					return -EINVAL;
+				break;
+			case SADB_EXT_KEY_AUTH:
+			case SADB_EXT_KEY_ENCRYPT:
+				if (verify_key_len(p))
+					return -EINVAL;
+				break;
+			default:
+				break;
 			}
 			ext_hdrs[ext_type-1] = (void *) p;
 		}
@@ -1104,14 +1131,12 @@ static struct xfrm_state * pfkey_msg2xfrm_state(struct net *net,
 	key = ext_hdrs[SADB_EXT_KEY_AUTH - 1];
 	if (key != NULL &&
 	    sa->sadb_sa_auth != SADB_X_AALG_NULL &&
-	    ((key->sadb_key_bits+7) / 8 == 0 ||
-	     (key->sadb_key_bits+7) / 8 > key->sadb_key_len * sizeof(uint64_t)))
+	    key->sadb_key_bits == 0)
 		return ERR_PTR(-EINVAL);
 	key = ext_hdrs[SADB_EXT_KEY_ENCRYPT-1];
 	if (key != NULL &&
 	    sa->sadb_sa_encrypt != SADB_EALG_NULL &&
-	    ((key->sadb_key_bits+7) / 8 == 0 ||
-	     (key->sadb_key_bits+7) / 8 > key->sadb_key_len * sizeof(uint64_t)))
+	    key->sadb_key_bits == 0)
 		return ERR_PTR(-EINVAL);
 
 	x = xfrm_state_alloc(net);
-- 
cgit v1.1


From 0b5c0305e57ca940713bcb2b202fd2b412c62f31 Mon Sep 17 00:00:00 2001
From: Arend Van Spriel <arend.vanspriel@broadcom.com>
Date: Tue, 3 Apr 2018 10:18:15 +0200
Subject: brcmfmac: fix firmware request processing if nvram load fails

When nvram loading fails a double free occurred. Fix this and reorg the
code a little.

Fixes: d09ae51a4b67 ("brcmfmac: pass struct in brcmf_fw_get_firmwares()")
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Arend van Spriel <arend.vanspriel@broadcom.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
---
 .../broadcom/brcm80211/brcmfmac/firmware.c         | 36 ++++++++++++----------
 1 file changed, 20 insertions(+), 16 deletions(-)

diff --git a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/firmware.c b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/firmware.c
index 9277f4c..94e177d 100644
--- a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/firmware.c
+++ b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/firmware.c
@@ -459,7 +459,7 @@ static void brcmf_fw_free_request(struct brcmf_fw_request *req)
 	kfree(req);
 }
 
-static void brcmf_fw_request_nvram_done(const struct firmware *fw, void *ctx)
+static int brcmf_fw_request_nvram_done(const struct firmware *fw, void *ctx)
 {
 	struct brcmf_fw *fwctx = ctx;
 	struct brcmf_fw_item *cur;
@@ -498,13 +498,10 @@ static void brcmf_fw_request_nvram_done(const struct firmware *fw, void *ctx)
 	brcmf_dbg(TRACE, "nvram %p len %d\n", nvram, nvram_length);
 	cur->nv_data.data = nvram;
 	cur->nv_data.len = nvram_length;
-	return;
+	return 0;
 
 fail:
-	brcmf_dbg(TRACE, "failed: dev=%s\n", dev_name(fwctx->dev));
-	fwctx->done(fwctx->dev, -ENOENT, NULL);
-	brcmf_fw_free_request(fwctx->req);
-	kfree(fwctx);
+	return -ENOENT;
 }
 
 static int brcmf_fw_request_next_item(struct brcmf_fw *fwctx, bool async)
@@ -553,20 +550,27 @@ static void brcmf_fw_request_done(const struct firmware *fw, void *ctx)
 	brcmf_dbg(TRACE, "enter: firmware %s %sfound\n", cur->path,
 		  fw ? "" : "not ");
 
-	if (fw) {
-		if (cur->type == BRCMF_FW_TYPE_BINARY)
-			cur->binary = fw;
-		else if (cur->type == BRCMF_FW_TYPE_NVRAM)
-			brcmf_fw_request_nvram_done(fw, fwctx);
-		else
-			release_firmware(fw);
-	} else if (cur->type == BRCMF_FW_TYPE_NVRAM) {
-		brcmf_fw_request_nvram_done(NULL, fwctx);
-	} else if (!(cur->flags & BRCMF_FW_REQF_OPTIONAL)) {
+	if (!fw)
 		ret = -ENOENT;
+
+	switch (cur->type) {
+	case BRCMF_FW_TYPE_NVRAM:
+		ret = brcmf_fw_request_nvram_done(fw, fwctx);
+		break;
+	case BRCMF_FW_TYPE_BINARY:
+		cur->binary = fw;
+		break;
+	default:
+		/* something fishy here so bail out early */
+		brcmf_err("unknown fw type: %d\n", cur->type);
+		release_firmware(fw);
+		ret = -EINVAL;
 		goto fail;
 	}
 
+	if (ret < 0 && !(cur->flags & BRCMF_FW_REQF_OPTIONAL))
+		goto fail;
+
 	do {
 		if (++fwctx->curpos == fwctx->req->n_items) {
 			ret = 0;
-- 
cgit v1.1


From 77e30e10ee28a53c8af95809866ee8493583e29a Mon Sep 17 00:00:00 2001
From: Haim Dreyfuss <haim.dreyfuss@intel.com>
Date: Wed, 4 Apr 2018 09:23:48 +0300
Subject: iwlwifi: mvm: query regdb for wmm rule if needed

Since our device is regulatory self managed it maintains its regulatory
rules by its own. However the wmm_rules values can't be set by the
device itself but only the indication about the need to set it.
In case the device set wmm indication, proactively query the regulatory
data base to get these values

Signed-off-by: Haim Dreyfuss <haim.dreyfuss@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
---
 drivers/net/wireless/intel/iwlwifi/iwl-nvm-parse.c | 111 ++++++++++++++++++---
 drivers/net/wireless/intel/iwlwifi/iwl-nvm-parse.h |   6 +-
 drivers/net/wireless/intel/iwlwifi/mvm/mac80211.c  |   3 +-
 3 files changed, 101 insertions(+), 19 deletions(-)

diff --git a/drivers/net/wireless/intel/iwlwifi/iwl-nvm-parse.c b/drivers/net/wireless/intel/iwlwifi/iwl-nvm-parse.c
index 8928613..ca01746 100644
--- a/drivers/net/wireless/intel/iwlwifi/iwl-nvm-parse.c
+++ b/drivers/net/wireless/intel/iwlwifi/iwl-nvm-parse.c
@@ -76,6 +76,7 @@
 #include "iwl-io.h"
 #include "iwl-csr.h"
 #include "fw/acpi.h"
+#include "fw/api/nvm-reg.h"
 
 /* NVM offsets (in words) definitions */
 enum nvm_offsets {
@@ -146,8 +147,8 @@ static const u8 iwl_ext_nvm_channels[] = {
 	149, 153, 157, 161, 165, 169, 173, 177, 181
 };
 
-#define IWL_NUM_CHANNELS		ARRAY_SIZE(iwl_nvm_channels)
-#define IWL_NUM_CHANNELS_EXT	ARRAY_SIZE(iwl_ext_nvm_channels)
+#define IWL_NVM_NUM_CHANNELS		ARRAY_SIZE(iwl_nvm_channels)
+#define IWL_NVM_NUM_CHANNELS_EXT	ARRAY_SIZE(iwl_ext_nvm_channels)
 #define NUM_2GHZ_CHANNELS		14
 #define NUM_2GHZ_CHANNELS_EXT	14
 #define FIRST_2GHZ_HT_MINUS		5
@@ -301,11 +302,11 @@ static int iwl_init_channel_map(struct device *dev, const struct iwl_cfg *cfg,
 	const u8 *nvm_chan;
 
 	if (cfg->nvm_type != IWL_NVM_EXT) {
-		num_of_ch = IWL_NUM_CHANNELS;
+		num_of_ch = IWL_NVM_NUM_CHANNELS;
 		nvm_chan = &iwl_nvm_channels[0];
 		num_2ghz_channels = NUM_2GHZ_CHANNELS;
 	} else {
-		num_of_ch = IWL_NUM_CHANNELS_EXT;
+		num_of_ch = IWL_NVM_NUM_CHANNELS_EXT;
 		nvm_chan = &iwl_ext_nvm_channels[0];
 		num_2ghz_channels = NUM_2GHZ_CHANNELS_EXT;
 	}
@@ -720,12 +721,12 @@ iwl_parse_nvm_data(struct iwl_trans *trans, const struct iwl_cfg *cfg,
 	if (cfg->nvm_type != IWL_NVM_EXT)
 		data = kzalloc(sizeof(*data) +
 			       sizeof(struct ieee80211_channel) *
-			       IWL_NUM_CHANNELS,
+			       IWL_NVM_NUM_CHANNELS,
 			       GFP_KERNEL);
 	else
 		data = kzalloc(sizeof(*data) +
 			       sizeof(struct ieee80211_channel) *
-			       IWL_NUM_CHANNELS_EXT,
+			       IWL_NVM_NUM_CHANNELS_EXT,
 			       GFP_KERNEL);
 	if (!data)
 		return NULL;
@@ -842,24 +843,34 @@ static u32 iwl_nvm_get_regdom_bw_flags(const u8 *nvm_chan,
 	return flags;
 }
 
+struct regdb_ptrs {
+	struct ieee80211_wmm_rule *rule;
+	u32 token;
+};
+
 struct ieee80211_regdomain *
 iwl_parse_nvm_mcc_info(struct device *dev, const struct iwl_cfg *cfg,
-		       int num_of_ch, __le32 *channels, u16 fw_mcc)
+		       int num_of_ch, __le32 *channels, u16 fw_mcc,
+		       u16 geo_info)
 {
 	int ch_idx;
 	u16 ch_flags;
 	u32 reg_rule_flags, prev_reg_rule_flags = 0;
 	const u8 *nvm_chan = cfg->nvm_type == IWL_NVM_EXT ?
 			     iwl_ext_nvm_channels : iwl_nvm_channels;
-	struct ieee80211_regdomain *regd;
-	int size_of_regd;
+	struct ieee80211_regdomain *regd, *copy_rd;
+	int size_of_regd, regd_to_copy, wmms_to_copy;
+	int size_of_wmms = 0;
 	struct ieee80211_reg_rule *rule;
+	struct ieee80211_wmm_rule *wmm_rule, *d_wmm, *s_wmm;
+	struct regdb_ptrs *regdb_ptrs;
 	enum nl80211_band band;
 	int center_freq, prev_center_freq = 0;
-	int valid_rules = 0;
+	int valid_rules = 0, n_wmms = 0;
+	int i;
 	bool new_rule;
 	int max_num_ch = cfg->nvm_type == IWL_NVM_EXT ?
-			 IWL_NUM_CHANNELS_EXT : IWL_NUM_CHANNELS;
+			 IWL_NVM_NUM_CHANNELS_EXT : IWL_NVM_NUM_CHANNELS;
 
 	if (WARN_ON_ONCE(num_of_ch > NL80211_MAX_SUPP_REG_RULES))
 		return ERR_PTR(-EINVAL);
@@ -875,10 +886,26 @@ iwl_parse_nvm_mcc_info(struct device *dev, const struct iwl_cfg *cfg,
 		sizeof(struct ieee80211_regdomain) +
 		num_of_ch * sizeof(struct ieee80211_reg_rule);
 
-	regd = kzalloc(size_of_regd, GFP_KERNEL);
+	if (geo_info & GEO_WMM_ETSI_5GHZ_INFO)
+		size_of_wmms =
+			num_of_ch * sizeof(struct ieee80211_wmm_rule);
+
+	regd = kzalloc(size_of_regd + size_of_wmms, GFP_KERNEL);
 	if (!regd)
 		return ERR_PTR(-ENOMEM);
 
+	regdb_ptrs = kcalloc(num_of_ch, sizeof(*regdb_ptrs), GFP_KERNEL);
+	if (!regdb_ptrs) {
+		copy_rd = ERR_PTR(-ENOMEM);
+		goto out;
+	}
+
+	/* set alpha2 from FW. */
+	regd->alpha2[0] = fw_mcc >> 8;
+	regd->alpha2[1] = fw_mcc & 0xff;
+
+	wmm_rule = (struct ieee80211_wmm_rule *)((u8 *)regd + size_of_regd);
+
 	for (ch_idx = 0; ch_idx < num_of_ch; ch_idx++) {
 		ch_flags = (u16)__le32_to_cpup(channels + ch_idx);
 		band = (ch_idx < NUM_2GHZ_CHANNELS) ?
@@ -927,14 +954,66 @@ iwl_parse_nvm_mcc_info(struct device *dev, const struct iwl_cfg *cfg,
 
 		iwl_nvm_print_channel_flags(dev, IWL_DL_LAR,
 					    nvm_chan[ch_idx], ch_flags);
+
+		if (!(geo_info & GEO_WMM_ETSI_5GHZ_INFO) ||
+		    band == NL80211_BAND_2GHZ)
+			continue;
+
+		if (!reg_query_regdb_wmm(regd->alpha2, center_freq,
+					 &regdb_ptrs[n_wmms].token, wmm_rule)) {
+			/* Add only new rules */
+			for (i = 0; i < n_wmms; i++) {
+				if (regdb_ptrs[i].token ==
+				    regdb_ptrs[n_wmms].token) {
+					rule->wmm_rule = regdb_ptrs[i].rule;
+					break;
+				}
+			}
+			if (i == n_wmms) {
+				rule->wmm_rule = wmm_rule;
+				regdb_ptrs[n_wmms++].rule = wmm_rule;
+				wmm_rule++;
+			}
+		}
 	}
 
 	regd->n_reg_rules = valid_rules;
+	regd->n_wmm_rules = n_wmms;
 
-	/* set alpha2 from FW. */
-	regd->alpha2[0] = fw_mcc >> 8;
-	regd->alpha2[1] = fw_mcc & 0xff;
+	/*
+	 * Narrow down regdom for unused regulatory rules to prevent hole
+	 * between reg rules to wmm rules.
+	 */
+	regd_to_copy = sizeof(struct ieee80211_regdomain) +
+		valid_rules * sizeof(struct ieee80211_reg_rule);
+
+	wmms_to_copy = sizeof(struct ieee80211_wmm_rule) * n_wmms;
+
+	copy_rd = kzalloc(regd_to_copy + wmms_to_copy, GFP_KERNEL);
+	if (!copy_rd) {
+		copy_rd = ERR_PTR(-ENOMEM);
+		goto out;
+	}
+
+	memcpy(copy_rd, regd, regd_to_copy);
+	memcpy((u8 *)copy_rd + regd_to_copy, (u8 *)regd + size_of_regd,
+	       wmms_to_copy);
+
+	d_wmm = (struct ieee80211_wmm_rule *)((u8 *)copy_rd + regd_to_copy);
+	s_wmm = (struct ieee80211_wmm_rule *)((u8 *)regd + size_of_regd);
+
+	for (i = 0; i < regd->n_reg_rules; i++) {
+		if (!regd->reg_rules[i].wmm_rule)
+			continue;
+
+		copy_rd->reg_rules[i].wmm_rule = d_wmm +
+			(regd->reg_rules[i].wmm_rule - s_wmm) /
+			sizeof(struct ieee80211_wmm_rule);
+	}
 
-	return regd;
+out:
+	kfree(regdb_ptrs);
+	kfree(regd);
+	return copy_rd;
 }
 IWL_EXPORT_SYMBOL(iwl_parse_nvm_mcc_info);
diff --git a/drivers/net/wireless/intel/iwlwifi/iwl-nvm-parse.h b/drivers/net/wireless/intel/iwlwifi/iwl-nvm-parse.h
index 306736c..3071a23b 100644
--- a/drivers/net/wireless/intel/iwlwifi/iwl-nvm-parse.h
+++ b/drivers/net/wireless/intel/iwlwifi/iwl-nvm-parse.h
@@ -101,12 +101,14 @@ void iwl_init_sbands(struct device *dev, const struct iwl_cfg *cfg,
  *
  * This function parses the regulatory channel data received as a
  * MCC_UPDATE_CMD command. It returns a newly allocation regulatory domain,
- * to be fed into the regulatory core. An ERR_PTR is returned on error.
+ * to be fed into the regulatory core. In case the geo_info is set handle
+ * accordingly. An ERR_PTR is returned on error.
  * If not given to the regulatory core, the user is responsible for freeing
  * the regdomain returned here with kfree.
  */
 struct ieee80211_regdomain *
 iwl_parse_nvm_mcc_info(struct device *dev, const struct iwl_cfg *cfg,
-		       int num_of_ch, __le32 *channels, u16 fw_mcc);
+		       int num_of_ch, __le32 *channels, u16 fw_mcc,
+		       u16 geo_info);
 
 #endif /* __iwl_nvm_parse_h__ */
diff --git a/drivers/net/wireless/intel/iwlwifi/mvm/mac80211.c b/drivers/net/wireless/intel/iwlwifi/mvm/mac80211.c
index 51b3042..90f8c89 100644
--- a/drivers/net/wireless/intel/iwlwifi/mvm/mac80211.c
+++ b/drivers/net/wireless/intel/iwlwifi/mvm/mac80211.c
@@ -311,7 +311,8 @@ struct ieee80211_regdomain *iwl_mvm_get_regdomain(struct wiphy *wiphy,
 	regd = iwl_parse_nvm_mcc_info(mvm->trans->dev, mvm->cfg,
 				      __le32_to_cpu(resp->n_channels),
 				      resp->channels,
-				      __le16_to_cpu(resp->mcc));
+				      __le16_to_cpu(resp->mcc),
+				      __le16_to_cpu(resp->geo_info));
 	/* Store the return source id */
 	src_id = resp->source_id;
 	kfree(resp);
-- 
cgit v1.1


From fd17ed684b6e33312cdcd9270b1fece9df266103 Mon Sep 17 00:00:00 2001
From: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Date: Mon, 9 Apr 2018 10:32:34 +0900
Subject: Revert "ata: ahci-platform: add reset control support"

This reverts commit f0f56716fc3e5d547fd7811eb218a30ed0695605.

According to Thierry's view,
https://www.spinics.net/lists/linux-ide/msg55357.html
some hardware-specific drivers already use their own resets,
and the common reset might make a path to occur double controls of resets.

For now, revert the commit that adds reset control support to ahci-platform,
and hold until the solution is confirmed not be affect all hardware-specific
drivers.

Fixes: f0f56716fc3e ("ata: ahci-platform: add reset control support")
Reported-by: Thierry Reding <thierry.reding@gmail.com>
Suggested-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
---
 .../devicetree/bindings/ata/ahci-platform.txt      |  1 -
 drivers/ata/ahci.h                                 |  1 -
 drivers/ata/libahci_platform.c                     | 24 +++-------------------
 3 files changed, 3 insertions(+), 23 deletions(-)

diff --git a/Documentation/devicetree/bindings/ata/ahci-platform.txt b/Documentation/devicetree/bindings/ata/ahci-platform.txt
index f4006d3..c760ecb 100644
--- a/Documentation/devicetree/bindings/ata/ahci-platform.txt
+++ b/Documentation/devicetree/bindings/ata/ahci-platform.txt
@@ -30,7 +30,6 @@ compatible:
 Optional properties:
 - dma-coherent      : Present if dma operations are coherent
 - clocks            : a list of phandle + clock specifier pairs
-- resets            : a list of phandle + reset specifier pairs
 - target-supply     : regulator for SATA target power
 - phys              : reference to the SATA PHY node
 - phy-names         : must be "sata-phy"
diff --git a/drivers/ata/ahci.h b/drivers/ata/ahci.h
index 4356ef1..a9d996e 100644
--- a/drivers/ata/ahci.h
+++ b/drivers/ata/ahci.h
@@ -350,7 +350,6 @@ struct ahci_host_priv {
 	u32			em_msg_type;	/* EM message type */
 	bool			got_runtime_pm; /* Did we do pm_runtime_get? */
 	struct clk		*clks[AHCI_MAX_CLKS]; /* Optional */
-	struct reset_control	*rsts;		/* Optional */
 	struct regulator	**target_pwrs;	/* Optional */
 	/*
 	 * If platform uses PHYs. There is a 1:1 relation between the port number and
diff --git a/drivers/ata/libahci_platform.c b/drivers/ata/libahci_platform.c
index 46a7624..30cc8f1 100644
--- a/drivers/ata/libahci_platform.c
+++ b/drivers/ata/libahci_platform.c
@@ -25,7 +25,6 @@
 #include <linux/phy/phy.h>
 #include <linux/pm_runtime.h>
 #include <linux/of_platform.h>
-#include <linux/reset.h>
 #include "ahci.h"
 
 static void ahci_host_stop(struct ata_host *host);
@@ -196,8 +195,7 @@ EXPORT_SYMBOL_GPL(ahci_platform_disable_regulators);
  * following order:
  * 1) Regulator
  * 2) Clocks (through ahci_platform_enable_clks)
- * 3) Resets
- * 4) Phys
+ * 3) Phys
  *
  * If resource enabling fails at any point the previous enabled resources
  * are disabled in reverse order.
@@ -217,19 +215,12 @@ int ahci_platform_enable_resources(struct ahci_host_priv *hpriv)
 	if (rc)
 		goto disable_regulator;
 
-	rc = reset_control_deassert(hpriv->rsts);
-	if (rc)
-		goto disable_clks;
-
 	rc = ahci_platform_enable_phys(hpriv);
 	if (rc)
-		goto disable_resets;
+		goto disable_clks;
 
 	return 0;
 
-disable_resets:
-	reset_control_assert(hpriv->rsts);
-
 disable_clks:
 	ahci_platform_disable_clks(hpriv);
 
@@ -248,15 +239,12 @@ EXPORT_SYMBOL_GPL(ahci_platform_enable_resources);
  * following order:
  * 1) Phys
  * 2) Clocks (through ahci_platform_disable_clks)
- * 3) Resets
- * 4) Regulator
+ * 3) Regulator
  */
 void ahci_platform_disable_resources(struct ahci_host_priv *hpriv)
 {
 	ahci_platform_disable_phys(hpriv);
 
-	reset_control_assert(hpriv->rsts);
-
 	ahci_platform_disable_clks(hpriv);
 
 	ahci_platform_disable_regulators(hpriv);
@@ -405,12 +393,6 @@ struct ahci_host_priv *ahci_platform_get_resources(struct platform_device *pdev)
 		hpriv->clks[i] = clk;
 	}
 
-	hpriv->rsts = devm_reset_control_array_get_optional_shared(dev);
-	if (IS_ERR(hpriv->rsts)) {
-		rc = PTR_ERR(hpriv->rsts);
-		goto err_out;
-	}
-
 	hpriv->nports = child_nodes = of_get_child_count(dev->of_node);
 
 	/*
-- 
cgit v1.1


From 6aac0a48b02f5d7ed64e4fdc2aa48843d425905b Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Tue, 10 Apr 2018 17:33:14 -0400
Subject: drm/amdkfd: Remove limit on number of GPUs (follow-up)

This condition was missed in a previous commit with the same title.

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c b/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c
index 66852de..f16ac2b 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c
@@ -307,9 +307,7 @@ int kfd_init_apertures(struct kfd_process *process)
 	struct kfd_process_device *pdd;
 
 	/*Iterating over all devices*/
-	while (kfd_topology_enum_kfd_devices(id, &dev) == 0 &&
-		id < NUM_OF_SUPPORTED_GPUS) {
-
+	while (kfd_topology_enum_kfd_devices(id, &dev) == 0) {
 		if (!dev) {
 			id++; /* Skip non GPU devices */
 			continue;
-- 
cgit v1.1


From 70a31d16ccac518c701b9fbfacce5460a226bfd9 Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Tue, 10 Apr 2018 17:33:15 -0400
Subject: drm/amdkfd: Support flat memory apertures for GFXv9

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c | 115 ++++++++++++++++++++-------
 1 file changed, 87 insertions(+), 28 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c b/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c
index f16ac2b..97d5423 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c
@@ -275,23 +275,35 @@
  * for FLAT_* / S_LOAD operations.
  */
 
-#define MAKE_GPUVM_APP_BASE(gpu_num) \
+#define MAKE_GPUVM_APP_BASE_VI(gpu_num) \
 	(((uint64_t)(gpu_num) << 61) + 0x1000000000000L)
 
 #define MAKE_GPUVM_APP_LIMIT(base, size) \
 	(((uint64_t)(base) & 0xFFFFFF0000000000UL) + (size) - 1)
 
-#define MAKE_SCRATCH_APP_BASE() \
+#define MAKE_SCRATCH_APP_BASE_VI() \
 	(((uint64_t)(0x1UL) << 61) + 0x100000000L)
 
 #define MAKE_SCRATCH_APP_LIMIT(base) \
 	(((uint64_t)base & 0xFFFFFFFF00000000UL) | 0xFFFFFFFF)
 
-#define MAKE_LDS_APP_BASE() \
+#define MAKE_LDS_APP_BASE_VI() \
 	(((uint64_t)(0x1UL) << 61) + 0x0)
 #define MAKE_LDS_APP_LIMIT(base) \
 	(((uint64_t)(base) & 0xFFFFFFFF00000000UL) | 0xFFFFFFFF)
 
+/* On GFXv9 the LDS and scratch apertures are programmed independently
+ * using the high 16 bits of the 64-bit virtual address. They must be
+ * in the hole, which will be the case as long as the high 16 bits are
+ * not 0.
+ *
+ * The aperture sizes are still 4GB implicitly.
+ *
+ * A GPUVM aperture is not applicable on GFXv9.
+ */
+#define MAKE_LDS_APP_BASE_V9() ((uint64_t)(0x1UL) << 48)
+#define MAKE_SCRATCH_APP_BASE_V9() ((uint64_t)(0x2UL) << 48)
+
 /* User mode manages most of the SVM aperture address space. The low
  * 16MB are reserved for kernel use (CWSR trap handler and kernel IB
  * for now).
@@ -300,6 +312,55 @@
 #define SVM_CWSR_BASE (SVM_USER_BASE - KFD_CWSR_TBA_TMA_SIZE)
 #define SVM_IB_BASE   (SVM_CWSR_BASE - PAGE_SIZE)
 
+static void kfd_init_apertures_vi(struct kfd_process_device *pdd, uint8_t id)
+{
+	/*
+	 * node id couldn't be 0 - the three MSB bits of
+	 * aperture shoudn't be 0
+	 */
+	pdd->lds_base = MAKE_LDS_APP_BASE_VI();
+	pdd->lds_limit = MAKE_LDS_APP_LIMIT(pdd->lds_base);
+
+	if (!pdd->dev->device_info->needs_iommu_device) {
+		/* dGPUs: SVM aperture starting at 0
+		 * with small reserved space for kernel.
+		 * Set them to CANONICAL addresses.
+		 */
+		pdd->gpuvm_base = SVM_USER_BASE;
+		pdd->gpuvm_limit =
+			pdd->dev->shared_resources.gpuvm_size - 1;
+	} else {
+		/* set them to non CANONICAL addresses, and no SVM is
+		 * allocated.
+		 */
+		pdd->gpuvm_base = MAKE_GPUVM_APP_BASE_VI(id + 1);
+		pdd->gpuvm_limit = MAKE_GPUVM_APP_LIMIT(pdd->gpuvm_base,
+				pdd->dev->shared_resources.gpuvm_size);
+	}
+
+	pdd->scratch_base = MAKE_SCRATCH_APP_BASE_VI();
+	pdd->scratch_limit = MAKE_SCRATCH_APP_LIMIT(pdd->scratch_base);
+}
+
+static void kfd_init_apertures_v9(struct kfd_process_device *pdd, uint8_t id)
+{
+	pdd->lds_base = MAKE_LDS_APP_BASE_V9();
+	pdd->lds_limit = MAKE_LDS_APP_LIMIT(pdd->lds_base);
+
+	/* Raven needs SVM to support graphic handle, etc. Leave the small
+	 * reserved space before SVM on Raven as well, even though we don't
+	 * have to.
+	 * Set gpuvm_base and gpuvm_limit to CANONICAL addresses so that they
+	 * are used in Thunk to reserve SVM.
+	 */
+	pdd->gpuvm_base = SVM_USER_BASE;
+	pdd->gpuvm_limit =
+		pdd->dev->shared_resources.gpuvm_size - 1;
+
+	pdd->scratch_base = MAKE_SCRATCH_APP_BASE_V9();
+	pdd->scratch_limit = MAKE_SCRATCH_APP_LIMIT(pdd->scratch_base);
+}
+
 int kfd_init_apertures(struct kfd_process *process)
 {
 	uint8_t id  = 0;
@@ -316,7 +377,7 @@ int kfd_init_apertures(struct kfd_process *process)
 		pdd = kfd_create_process_device_data(dev, process);
 		if (!pdd) {
 			pr_err("Failed to create process device data\n");
-			return -1;
+			return -ENOMEM;
 		}
 		/*
 		 * For 64 bit process apertures will be statically reserved in
@@ -328,32 +389,30 @@ int kfd_init_apertures(struct kfd_process *process)
 			pdd->gpuvm_base = pdd->gpuvm_limit = 0;
 			pdd->scratch_base = pdd->scratch_limit = 0;
 		} else {
-			/* Same LDS and scratch apertures can be used
-			 * on all GPUs. This allows using more dGPUs
-			 * than placement options for apertures.
-			 */
-			pdd->lds_base = MAKE_LDS_APP_BASE();
-			pdd->lds_limit = MAKE_LDS_APP_LIMIT(pdd->lds_base);
-
-			pdd->scratch_base = MAKE_SCRATCH_APP_BASE();
-			pdd->scratch_limit =
-				MAKE_SCRATCH_APP_LIMIT(pdd->scratch_base);
+			switch (dev->device_info->asic_family) {
+			case CHIP_KAVERI:
+			case CHIP_HAWAII:
+			case CHIP_CARRIZO:
+			case CHIP_TONGA:
+			case CHIP_FIJI:
+			case CHIP_POLARIS10:
+			case CHIP_POLARIS11:
+				kfd_init_apertures_vi(pdd, id);
+				break;
+			case CHIP_VEGA10:
+			case CHIP_RAVEN:
+				kfd_init_apertures_v9(pdd, id);
+				break;
+			default:
+				WARN(1, "Unexpected ASIC family %u",
+				     dev->device_info->asic_family);
+				return -EINVAL;
+			}
 
-			if (dev->device_info->needs_iommu_device) {
-				/* APUs: GPUVM aperture in
-				 * non-canonical address space
-				 */
-				pdd->gpuvm_base = MAKE_GPUVM_APP_BASE(id + 1);
-				pdd->gpuvm_limit = MAKE_GPUVM_APP_LIMIT(
-					pdd->gpuvm_base,
-					dev->shared_resources.gpuvm_size);
-			} else {
-				/* dGPUs: SVM aperture starting at 0
-				 * with small reserved space for kernel
+			if (!dev->device_info->needs_iommu_device) {
+				/* dGPUs: the reserved space for kernel
+				 * before SVM
 				 */
-				pdd->gpuvm_base = SVM_USER_BASE;
-				pdd->gpuvm_limit =
-					dev->shared_resources.gpuvm_size - 1;
 				pdd->qpd.cwsr_base = SVM_CWSR_BASE;
 				pdd->qpd.ib_base = SVM_IB_BASE;
 			}
-- 
cgit v1.1


From 3e76c2399b55483b1a28499b090f9d6600ab9eff Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Tue, 10 Apr 2018 17:33:16 -0400
Subject: drm/amdkfd: Add GFXv9 CWSR trap handler

Signed-off-by: Shaoyun Liu <Shaoyun.Liu@amd.com>
Signed-off-by: Jay Cornwall <Jay.Cornwall@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 .../gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm  | 1495 ++++++++++++++++++++
 drivers/gpu/drm/amd/amdkfd/kfd_device.c            |   13 +-
 2 files changed, 1505 insertions(+), 3 deletions(-)
 create mode 100644 drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm

diff --git a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm
new file mode 100644
index 0000000..033580c
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm
@@ -0,0 +1,1495 @@
+/*
+ * Copyright 2016 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#if 0
+HW (GFX9) source code for CWSR trap handler
+#Version 18 + multiple trap handler
+
+// this performance-optimal version was originally from Seven Xu at SRDC
+
+// Revison #18	 --...
+/* Rev History
+** #1. Branch from gc dv.   //gfxip/gfx9/main/src/test/suites/block/cs/sr/cs_trap_handler.sp3#1,#50, #51, #52-53(Skip, Already Fixed by PV), #54-56(merged),#57-58(mergerd, skiped-already fixed by PV)
+** #4. SR Memory Layout:
+**			 1. VGPR-SGPR-HWREG-{LDS}
+**			 2. tba_hi.bits.26 - reconfigured as the first wave in tg bits, for defer Save LDS for a threadgroup.. performance concern..
+** #5. Update: 1. Accurate g8sr_ts_save_d timestamp
+** #6. Update: 1. Fix s_barrier usage; 2. VGPR s/r using swizzle buffer?(NoNeed, already matched the swizzle pattern, more investigation)
+** #7. Update: 1. don't barrier if noLDS
+** #8. Branch: 1. Branch to ver#0, which is very similar to gc dv version
+**	       2. Fix SQ issue by s_sleep 2
+** #9. Update: 1. Fix scc restore failed issue, restore wave_status at last
+**	       2. optimize s_buffer save by burst 16sgprs...
+** #10. Update 1. Optimize restore sgpr by busrt 16 sgprs.
+** #11. Update 1. Add 2 more timestamp for debug version
+** #12. Update 1. Add VGPR SR using DWx4, some case improve and some case drop performance
+** #13. Integ  1. Always use MUBUF for PV trap shader...
+** #14. Update 1. s_buffer_store soft clause...
+** #15. Update 1. PERF - sclar write with glc:0/mtype0 to allow L2 combine. perf improvement a lot.
+** #16. Update 1. PRRF - UNROLL LDS_DMA got 2500cycle save in IP tree
+** #17. Update 1. FUNC - LDS_DMA has issues while ATC, replace with ds_read/buffer_store for save part[TODO restore part]
+**	       2. PERF - Save LDS before save VGPR to cover LDS save long latency...
+** #18. Update 1. FUNC - Implicitly estore STATUS.VCCZ, which is not writable by s_setreg_b32
+**	       2. FUNC - Handle non-CWSR traps
+*/
+
+var G8SR_WDMEM_HWREG_OFFSET = 0
+var G8SR_WDMEM_SGPR_OFFSET  = 128  // in bytes
+
+// Keep definition same as the app shader, These 2 time stamps are part of the app shader... Should before any Save and after restore.
+
+var G8SR_DEBUG_TIMESTAMP = 0
+var G8SR_DEBUG_TS_SAVE_D_OFFSET = 40*4	// ts_save_d timestamp offset relative to SGPR_SR_memory_offset
+var s_g8sr_ts_save_s	= s[34:35]   // save start
+var s_g8sr_ts_sq_save_msg  = s[36:37]	// The save shader send SAVEWAVE msg to spi
+var s_g8sr_ts_spi_wrexec   = s[38:39]	// the SPI write the sr address to SQ
+var s_g8sr_ts_save_d	= s[40:41]   // save end
+var s_g8sr_ts_restore_s = s[42:43]   // restore start
+var s_g8sr_ts_restore_d = s[44:45]   // restore end
+
+var G8SR_VGPR_SR_IN_DWX4 = 0
+var G8SR_SAVE_BUF_RSRC_WORD1_STRIDE_DWx4 = 0x00100000	 // DWx4 stride is 4*4Bytes
+var G8SR_RESTORE_BUF_RSRC_WORD1_STRIDE_DWx4  = G8SR_SAVE_BUF_RSRC_WORD1_STRIDE_DWx4
+
+
+/*************************************************************************/
+/*		    control on how to run the shader			 */
+/*************************************************************************/
+//any hack that needs to be made to run this code in EMU (either because various EMU code are not ready or no compute save & restore in EMU run)
+var EMU_RUN_HACK		    =	0
+var EMU_RUN_HACK_RESTORE_NORMAL	    =	0
+var EMU_RUN_HACK_SAVE_NORMAL_EXIT   =	0
+var EMU_RUN_HACK_SAVE_SINGLE_WAVE   =	0
+var EMU_RUN_HACK_SAVE_FIRST_TIME    =	0		    //for interrupted restore in which the first save is through EMU_RUN_HACK
+var SAVE_LDS			    =	1
+var WG_BASE_ADDR_LO		    =	0x9000a000
+var WG_BASE_ADDR_HI		    =	0x0
+var WAVE_SPACE			    =	0x5000		    //memory size that each wave occupies in workgroup state mem
+var CTX_SAVE_CONTROL		    =	0x0
+var CTX_RESTORE_CONTROL		    =	CTX_SAVE_CONTROL
+var SIM_RUN_HACK		    =	0		    //any hack that needs to be made to run this code in SIM (either because various RTL code are not ready or no compute save & restore in RTL run)
+var SGPR_SAVE_USE_SQC		    =	1		    //use SQC D$ to do the write
+var USE_MTBUF_INSTEAD_OF_MUBUF	    =	0		    //because TC EMU currently asserts on 0 of // overload DFMT field to carry 4 more bits of stride for MUBUF opcodes
+var SWIZZLE_EN			    =	0		    //whether we use swizzled buffer addressing
+var ACK_SQC_STORE		    =	1		    //workaround for suspected SQC store bug causing incorrect stores under concurrency
+
+/**************************************************************************/
+/*			variables					  */
+/**************************************************************************/
+var SQ_WAVE_STATUS_INST_ATC_SHIFT  = 23
+var SQ_WAVE_STATUS_INST_ATC_MASK   = 0x00800000
+var SQ_WAVE_STATUS_SPI_PRIO_MASK   = 0x00000006
+var SQ_WAVE_STATUS_HALT_MASK       = 0x2000
+
+var SQ_WAVE_LDS_ALLOC_LDS_SIZE_SHIFT	= 12
+var SQ_WAVE_LDS_ALLOC_LDS_SIZE_SIZE	= 9
+var SQ_WAVE_GPR_ALLOC_VGPR_SIZE_SHIFT	= 8
+var SQ_WAVE_GPR_ALLOC_VGPR_SIZE_SIZE	= 6
+var SQ_WAVE_GPR_ALLOC_SGPR_SIZE_SHIFT	= 24
+var SQ_WAVE_GPR_ALLOC_SGPR_SIZE_SIZE	= 3			//FIXME	 sq.blk still has 4 bits at this time while SQ programming guide has 3 bits
+
+var SQ_WAVE_TRAPSTS_SAVECTX_MASK    =	0x400
+var SQ_WAVE_TRAPSTS_EXCE_MASK	    =	0x1FF			// Exception mask
+var SQ_WAVE_TRAPSTS_SAVECTX_SHIFT   =	10
+var SQ_WAVE_TRAPSTS_MEM_VIOL_MASK   =	0x100
+var SQ_WAVE_TRAPSTS_MEM_VIOL_SHIFT  =	8
+var SQ_WAVE_TRAPSTS_PRE_SAVECTX_MASK	=   0x3FF
+var SQ_WAVE_TRAPSTS_PRE_SAVECTX_SHIFT	=   0x0
+var SQ_WAVE_TRAPSTS_PRE_SAVECTX_SIZE	=   10
+var SQ_WAVE_TRAPSTS_POST_SAVECTX_MASK	=   0xFFFFF800
+var SQ_WAVE_TRAPSTS_POST_SAVECTX_SHIFT	=   11
+var SQ_WAVE_TRAPSTS_POST_SAVECTX_SIZE	=   21
+var SQ_WAVE_TRAPSTS_ILLEGAL_INST_MASK	=   0x800
+
+var SQ_WAVE_IB_STS_RCNT_SHIFT		=   16			//FIXME
+var SQ_WAVE_IB_STS_FIRST_REPLAY_SHIFT	=   15			//FIXME
+var SQ_WAVE_IB_STS_RCNT_FIRST_REPLAY_MASK	= 0x1F8000
+var SQ_WAVE_IB_STS_RCNT_FIRST_REPLAY_MASK_NEG	= 0x00007FFF	//FIXME
+
+var SQ_BUF_RSRC_WORD1_ATC_SHIFT	    =	24
+var SQ_BUF_RSRC_WORD3_MTYPE_SHIFT   =	27
+
+var TTMP11_SAVE_RCNT_FIRST_REPLAY_SHIFT	=   26			// bits [31:26] unused by SPI debug data
+var TTMP11_SAVE_RCNT_FIRST_REPLAY_MASK	=   0xFC000000
+
+/*	Save	    */
+var S_SAVE_BUF_RSRC_WORD1_STRIDE	=   0x00040000		//stride is 4 bytes
+var S_SAVE_BUF_RSRC_WORD3_MISC		=   0x00807FAC		//SQ_SEL_X/Y/Z/W, BUF_NUM_FORMAT_FLOAT, (0 for MUBUF stride[17:14] when ADD_TID_ENABLE and BUF_DATA_FORMAT_32 for MTBUF), ADD_TID_ENABLE
+
+var S_SAVE_SPI_INIT_ATC_MASK		=   0x08000000		//bit[27]: ATC bit
+var S_SAVE_SPI_INIT_ATC_SHIFT		=   27
+var S_SAVE_SPI_INIT_MTYPE_MASK		=   0x70000000		//bit[30:28]: Mtype
+var S_SAVE_SPI_INIT_MTYPE_SHIFT		=   28
+var S_SAVE_SPI_INIT_FIRST_WAVE_MASK	=   0x04000000		//bit[26]: FirstWaveInTG
+var S_SAVE_SPI_INIT_FIRST_WAVE_SHIFT	=   26
+
+var S_SAVE_PC_HI_RCNT_SHIFT		=   28			//FIXME	 check with Brian to ensure all fields other than PC[47:0] can be used
+var S_SAVE_PC_HI_RCNT_MASK		=   0xF0000000		//FIXME
+var S_SAVE_PC_HI_FIRST_REPLAY_SHIFT	=   27			//FIXME
+var S_SAVE_PC_HI_FIRST_REPLAY_MASK	=   0x08000000		//FIXME
+
+var s_save_spi_init_lo		    =	exec_lo
+var s_save_spi_init_hi		    =	exec_hi
+
+var s_save_pc_lo	    =	ttmp0		//{TTMP1, TTMP0} = {3??h0,pc_rewind[3:0], HT[0],trapID[7:0], PC[47:0]}
+var s_save_pc_hi	    =	ttmp1
+var s_save_exec_lo	    =	ttmp2
+var s_save_exec_hi	    =	ttmp3
+var s_save_tmp		    =	ttmp4
+var s_save_trapsts	    =	ttmp5		//not really used until the end of the SAVE routine
+var s_save_xnack_mask_lo    =	ttmp6
+var s_save_xnack_mask_hi    =	ttmp7
+var s_save_buf_rsrc0	    =	ttmp8
+var s_save_buf_rsrc1	    =	ttmp9
+var s_save_buf_rsrc2	    =	ttmp10
+var s_save_buf_rsrc3	    =	ttmp11
+var s_save_status	    =	ttmp12
+var s_save_mem_offset	    =	ttmp14
+var s_save_alloc_size	    =	s_save_trapsts		//conflict
+var s_save_m0		    =	ttmp15
+var s_save_ttmps_lo	    =	s_save_tmp		//no conflict
+var s_save_ttmps_hi	    =	s_save_trapsts		//no conflict
+
+/*	Restore	    */
+var S_RESTORE_BUF_RSRC_WORD1_STRIDE	    =	S_SAVE_BUF_RSRC_WORD1_STRIDE
+var S_RESTORE_BUF_RSRC_WORD3_MISC	    =	S_SAVE_BUF_RSRC_WORD3_MISC
+
+var S_RESTORE_SPI_INIT_ATC_MASK		    =	0x08000000	    //bit[27]: ATC bit
+var S_RESTORE_SPI_INIT_ATC_SHIFT	    =	27
+var S_RESTORE_SPI_INIT_MTYPE_MASK	    =	0x70000000	    //bit[30:28]: Mtype
+var S_RESTORE_SPI_INIT_MTYPE_SHIFT	    =	28
+var S_RESTORE_SPI_INIT_FIRST_WAVE_MASK	    =	0x04000000	    //bit[26]: FirstWaveInTG
+var S_RESTORE_SPI_INIT_FIRST_WAVE_SHIFT	    =	26
+
+var S_RESTORE_PC_HI_RCNT_SHIFT		    =	S_SAVE_PC_HI_RCNT_SHIFT
+var S_RESTORE_PC_HI_RCNT_MASK		    =	S_SAVE_PC_HI_RCNT_MASK
+var S_RESTORE_PC_HI_FIRST_REPLAY_SHIFT	    =	S_SAVE_PC_HI_FIRST_REPLAY_SHIFT
+var S_RESTORE_PC_HI_FIRST_REPLAY_MASK	    =	S_SAVE_PC_HI_FIRST_REPLAY_MASK
+
+var s_restore_spi_init_lo		    =	exec_lo
+var s_restore_spi_init_hi		    =	exec_hi
+
+var s_restore_mem_offset	=   ttmp12
+var s_restore_alloc_size	=   ttmp3
+var s_restore_tmp		=   ttmp2
+var s_restore_mem_offset_save	=   s_restore_tmp	//no conflict
+
+var s_restore_m0	    =	s_restore_alloc_size	//no conflict
+
+var s_restore_mode	    =	ttmp7
+
+var s_restore_pc_lo	    =	ttmp0
+var s_restore_pc_hi	    =	ttmp1
+var s_restore_exec_lo	    =	ttmp14
+var s_restore_exec_hi	    = 	ttmp15
+var s_restore_status	    =	ttmp4
+var s_restore_trapsts	    =	ttmp5
+var s_restore_xnack_mask_lo =	xnack_mask_lo
+var s_restore_xnack_mask_hi =	xnack_mask_hi
+var s_restore_buf_rsrc0	    =	ttmp8
+var s_restore_buf_rsrc1	    =	ttmp9
+var s_restore_buf_rsrc2	    =	ttmp10
+var s_restore_buf_rsrc3	    =	ttmp11
+var s_restore_ttmps_lo	    =	s_restore_tmp		//no conflict
+var s_restore_ttmps_hi	    =	s_restore_alloc_size	//no conflict
+
+/**************************************************************************/
+/*			trap handler entry points			  */
+/**************************************************************************/
+/* Shader Main*/
+
+shader main
+  asic(GFX9)
+  type(CS)
+
+
+    if ((EMU_RUN_HACK) && (!EMU_RUN_HACK_RESTORE_NORMAL))		    //hack to use trap_id for determining save/restore
+	//FIXME VCCZ un-init assertion s_getreg_b32	s_save_status, hwreg(HW_REG_STATUS)	    //save STATUS since we will change SCC
+	s_and_b32 s_save_tmp, s_save_pc_hi, 0xffff0000		    //change SCC
+	s_cmp_eq_u32 s_save_tmp, 0x007e0000			    //Save: trap_id = 0x7e. Restore: trap_id = 0x7f.
+	s_cbranch_scc0 L_JUMP_TO_RESTORE			    //do not need to recover STATUS here  since we are going to RESTORE
+	//FIXME	 s_setreg_b32	hwreg(HW_REG_STATUS),	s_save_status	    //need to recover STATUS since we are going to SAVE
+	s_branch L_SKIP_RESTORE					    //NOT restore, SAVE actually
+    else
+	s_branch L_SKIP_RESTORE					    //NOT restore. might be a regular trap or save
+    end
+
+L_JUMP_TO_RESTORE:
+    s_branch L_RESTORE						    //restore
+
+L_SKIP_RESTORE:
+
+    s_getreg_b32    s_save_status, hwreg(HW_REG_STATUS)				    //save STATUS since we will change SCC
+    s_andn2_b32	    s_save_status, s_save_status, SQ_WAVE_STATUS_SPI_PRIO_MASK	    //check whether this is for save
+    s_getreg_b32    s_save_trapsts, hwreg(HW_REG_TRAPSTS)
+    s_and_b32       ttmp2, s_save_trapsts, SQ_WAVE_TRAPSTS_SAVECTX_MASK    //check whether this is for save
+    s_cbranch_scc1  L_SAVE					//this is the operation for save
+
+    // *********    Handle non-CWSR traps	*******************
+if (!EMU_RUN_HACK)
+    // Illegal instruction is a non-maskable exception which blocks context save.
+    // Halt the wavefront and return from the trap.
+    s_and_b32       ttmp2, s_save_trapsts, SQ_WAVE_TRAPSTS_ILLEGAL_INST_MASK
+    s_cbranch_scc1  L_HALT_WAVE
+
+    // If STATUS.MEM_VIOL is asserted then we cannot fetch from the TMA.
+    // Instead, halt the wavefront and return from the trap.
+    s_and_b32       ttmp2, s_save_trapsts, SQ_WAVE_TRAPSTS_MEM_VIOL_MASK
+    s_cbranch_scc0  L_FETCH_2ND_TRAP
+
+L_HALT_WAVE:
+    // If STATUS.HALT is set then this fault must come from SQC instruction fetch.
+    // We cannot prevent further faults so just terminate the wavefront.
+    s_and_b32       ttmp2, s_save_status, SQ_WAVE_STATUS_HALT_MASK
+    s_cbranch_scc0  L_NOT_ALREADY_HALTED
+    s_endpgm
+L_NOT_ALREADY_HALTED:
+    s_or_b32        s_save_status, s_save_status, SQ_WAVE_STATUS_HALT_MASK
+
+    // If the PC points to S_ENDPGM then context save will fail if STATUS.HALT is set.
+    // Rewind the PC to prevent this from occurring. The debugger compensates for this.
+    s_sub_u32       ttmp0, ttmp0, 0x8
+    s_subb_u32      ttmp1, ttmp1, 0x0
+
+L_FETCH_2ND_TRAP:
+    // Preserve and clear scalar XNACK state before issuing scalar reads.
+    // Save IB_STS.FIRST_REPLAY[15] and IB_STS.RCNT[20:16] into unused space ttmp11[31:26].
+    s_getreg_b32    ttmp2, hwreg(HW_REG_IB_STS)
+    s_and_b32       ttmp3, ttmp2, SQ_WAVE_IB_STS_RCNT_FIRST_REPLAY_MASK
+    s_lshl_b32      ttmp3, ttmp3, (TTMP11_SAVE_RCNT_FIRST_REPLAY_SHIFT - SQ_WAVE_IB_STS_FIRST_REPLAY_SHIFT)
+    s_andn2_b32     ttmp11, ttmp11, TTMP11_SAVE_RCNT_FIRST_REPLAY_MASK
+    s_or_b32        ttmp11, ttmp11, ttmp3
+
+    s_andn2_b32     ttmp2, ttmp2, SQ_WAVE_IB_STS_RCNT_FIRST_REPLAY_MASK
+    s_setreg_b32    hwreg(HW_REG_IB_STS), ttmp2
+
+    // Read second-level TBA/TMA from first-level TMA and jump if available.
+    // ttmp[2:5] and ttmp12 can be used (others hold SPI-initialized debug data)
+    // ttmp12 holds SQ_WAVE_STATUS
+    s_getreg_b32    ttmp4, hwreg(HW_REG_SQ_SHADER_TMA_LO)
+    s_getreg_b32    ttmp5, hwreg(HW_REG_SQ_SHADER_TMA_HI)
+    s_lshl_b64      [ttmp4, ttmp5], [ttmp4, ttmp5], 0x8
+    s_load_dwordx2  [ttmp2, ttmp3], [ttmp4, ttmp5], 0x0 glc:1 // second-level TBA
+    s_waitcnt       lgkmcnt(0)
+    s_load_dwordx2  [ttmp4, ttmp5], [ttmp4, ttmp5], 0x8 glc:1 // second-level TMA
+    s_waitcnt       lgkmcnt(0)
+    s_and_b64       [ttmp2, ttmp3], [ttmp2, ttmp3], [ttmp2, ttmp3]
+    s_cbranch_scc0  L_NO_NEXT_TRAP // second-level trap handler not been set
+    s_setpc_b64     [ttmp2, ttmp3] // jump to second-level trap handler
+
+L_NO_NEXT_TRAP:
+    s_getreg_b32    s_save_trapsts, hwreg(HW_REG_TRAPSTS)
+    s_and_b32	    s_save_trapsts, s_save_trapsts, SQ_WAVE_TRAPSTS_EXCE_MASK // Check whether it is an exception
+    s_cbranch_scc1  L_EXCP_CASE	  // Exception, jump back to the shader program directly.
+    s_add_u32	    ttmp0, ttmp0, 4   // S_TRAP case, add 4 to ttmp0
+    s_addc_u32	ttmp1, ttmp1, 0
+L_EXCP_CASE:
+    s_and_b32	ttmp1, ttmp1, 0xFFFF
+
+    // Restore SQ_WAVE_IB_STS.
+    s_lshr_b32      ttmp2, ttmp11, (TTMP11_SAVE_RCNT_FIRST_REPLAY_SHIFT - SQ_WAVE_IB_STS_FIRST_REPLAY_SHIFT)
+    s_and_b32       ttmp2, ttmp2, SQ_WAVE_IB_STS_RCNT_FIRST_REPLAY_MASK
+    s_setreg_b32    hwreg(HW_REG_IB_STS), ttmp2
+
+    // Restore SQ_WAVE_STATUS.
+    s_and_b64       exec, exec, exec // Restore STATUS.EXECZ, not writable by s_setreg_b32
+    s_and_b64       vcc, vcc, vcc    // Restore STATUS.VCCZ, not writable by s_setreg_b32
+    s_setreg_b32    hwreg(HW_REG_STATUS), s_save_status
+
+    s_rfe_b64       [ttmp0, ttmp1]
+end
+    // *********	End handling of non-CWSR traps	 *******************
+
+/**************************************************************************/
+/*			save routine					  */
+/**************************************************************************/
+
+L_SAVE:
+
+if G8SR_DEBUG_TIMESTAMP
+	s_memrealtime	s_g8sr_ts_save_s
+	s_waitcnt lgkmcnt(0)	     //FIXME, will cause xnack??
+end
+
+    s_and_b32	    s_save_pc_hi, s_save_pc_hi, 0x0000ffff    //pc[47:32]
+
+    s_mov_b32	    s_save_tmp, 0							    //clear saveCtx bit
+    s_setreg_b32    hwreg(HW_REG_TRAPSTS, SQ_WAVE_TRAPSTS_SAVECTX_SHIFT, 1), s_save_tmp	    //clear saveCtx bit
+
+    s_getreg_b32    s_save_tmp, hwreg(HW_REG_IB_STS, SQ_WAVE_IB_STS_RCNT_SHIFT, SQ_WAVE_IB_STS_RCNT_SIZE)		    //save RCNT
+    s_lshl_b32	    s_save_tmp, s_save_tmp, S_SAVE_PC_HI_RCNT_SHIFT
+    s_or_b32	    s_save_pc_hi, s_save_pc_hi, s_save_tmp
+    s_getreg_b32    s_save_tmp, hwreg(HW_REG_IB_STS, SQ_WAVE_IB_STS_FIRST_REPLAY_SHIFT, SQ_WAVE_IB_STS_FIRST_REPLAY_SIZE)   //save FIRST_REPLAY
+    s_lshl_b32	    s_save_tmp, s_save_tmp, S_SAVE_PC_HI_FIRST_REPLAY_SHIFT
+    s_or_b32	    s_save_pc_hi, s_save_pc_hi, s_save_tmp
+    s_getreg_b32    s_save_tmp, hwreg(HW_REG_IB_STS)					    //clear RCNT and FIRST_REPLAY in IB_STS
+    s_and_b32	    s_save_tmp, s_save_tmp, SQ_WAVE_IB_STS_RCNT_FIRST_REPLAY_MASK_NEG
+
+    s_setreg_b32    hwreg(HW_REG_IB_STS), s_save_tmp
+
+    /*	    inform SPI the readiness and wait for SPI's go signal */
+    s_mov_b32	    s_save_exec_lo, exec_lo						    //save EXEC and use EXEC for the go signal from SPI
+    s_mov_b32	    s_save_exec_hi, exec_hi
+    s_mov_b64	    exec,   0x0								    //clear EXEC to get ready to receive
+
+if G8SR_DEBUG_TIMESTAMP
+	s_memrealtime  s_g8sr_ts_sq_save_msg
+	s_waitcnt lgkmcnt(0)
+end
+
+    if (EMU_RUN_HACK)
+
+    else
+	s_sendmsg   sendmsg(MSG_SAVEWAVE)  //send SPI a message and wait for SPI's write to EXEC
+    end
+
+  L_SLEEP:
+    s_sleep 0x2		       // sleep 1 (64clk) is not enough for 8 waves per SIMD, which will cause SQ hang, since the 7,8th wave could not get arbit to exec inst, while other waves are stuck into the sleep-loop and waiting for wrexec!=0
+
+    if (EMU_RUN_HACK)
+
+    else
+	s_cbranch_execz L_SLEEP
+    end
+
+if G8SR_DEBUG_TIMESTAMP
+	s_memrealtime  s_g8sr_ts_spi_wrexec
+	s_waitcnt lgkmcnt(0)
+end
+
+    if ((EMU_RUN_HACK) && (!EMU_RUN_HACK_SAVE_SINGLE_WAVE))
+	//calculate wd_addr using absolute thread id
+	v_readlane_b32 s_save_tmp, v9, 0
+	s_lshr_b32 s_save_tmp, s_save_tmp, 6
+	s_mul_i32 s_save_tmp, s_save_tmp, WAVE_SPACE
+	s_add_i32 s_save_spi_init_lo, s_save_tmp, WG_BASE_ADDR_LO
+	s_mov_b32 s_save_spi_init_hi, WG_BASE_ADDR_HI
+	s_and_b32 s_save_spi_init_hi, s_save_spi_init_hi, CTX_SAVE_CONTROL
+    else
+    end
+    if ((EMU_RUN_HACK) && (EMU_RUN_HACK_SAVE_SINGLE_WAVE))
+	s_add_i32 s_save_spi_init_lo, s_save_tmp, WG_BASE_ADDR_LO
+	s_mov_b32 s_save_spi_init_hi, WG_BASE_ADDR_HI
+	s_and_b32 s_save_spi_init_hi, s_save_spi_init_hi, CTX_SAVE_CONTROL
+    else
+    end
+
+    // Save trap temporaries 6-11, 13-15 initialized by SPI debug dispatch logic
+    // ttmp SR memory offset : size(VGPR)+size(SGPR)+0x40
+    get_vgpr_size_bytes(s_save_ttmps_lo)
+    get_sgpr_size_bytes(s_save_ttmps_hi)
+    s_add_u32	    s_save_ttmps_lo, s_save_ttmps_lo, s_save_ttmps_hi
+    s_add_u32	    s_save_ttmps_lo, s_save_ttmps_lo, s_save_spi_init_lo
+    s_addc_u32	    s_save_ttmps_hi, s_save_spi_init_hi, 0x0
+    s_and_b32	    s_save_ttmps_hi, s_save_ttmps_hi, 0xFFFF
+    s_store_dwordx2 [ttmp6, ttmp7], [s_save_ttmps_lo, s_save_ttmps_hi], 0x40 glc:1
+    ack_sqc_store_workaround()
+    s_store_dwordx4 [ttmp8, ttmp9, ttmp10, ttmp11], [s_save_ttmps_lo, s_save_ttmps_hi], 0x48 glc:1
+    ack_sqc_store_workaround()
+    s_store_dword   ttmp13, [s_save_ttmps_lo, s_save_ttmps_hi], 0x58 glc:1
+    ack_sqc_store_workaround()
+    s_store_dwordx2 [ttmp14, ttmp15], [s_save_ttmps_lo, s_save_ttmps_hi], 0x5C glc:1
+    ack_sqc_store_workaround()
+
+    /*	    setup Resource Contants    */
+    s_mov_b32	    s_save_buf_rsrc0,	s_save_spi_init_lo							//base_addr_lo
+    s_and_b32	    s_save_buf_rsrc1,	s_save_spi_init_hi, 0x0000FFFF						//base_addr_hi
+    s_or_b32	    s_save_buf_rsrc1,	s_save_buf_rsrc1,  S_SAVE_BUF_RSRC_WORD1_STRIDE
+    s_mov_b32	    s_save_buf_rsrc2,	0									//NUM_RECORDS initial value = 0 (in bytes) although not neccessarily inited
+    s_mov_b32	    s_save_buf_rsrc3,	S_SAVE_BUF_RSRC_WORD3_MISC
+    s_and_b32	    s_save_tmp,		s_save_spi_init_hi, S_SAVE_SPI_INIT_ATC_MASK
+    s_lshr_b32	    s_save_tmp,		s_save_tmp, (S_SAVE_SPI_INIT_ATC_SHIFT-SQ_BUF_RSRC_WORD1_ATC_SHIFT)	    //get ATC bit into position
+    s_or_b32	    s_save_buf_rsrc3,	s_save_buf_rsrc3,  s_save_tmp						//or ATC
+    s_and_b32	    s_save_tmp,		s_save_spi_init_hi, S_SAVE_SPI_INIT_MTYPE_MASK
+    s_lshr_b32	    s_save_tmp,		s_save_tmp, (S_SAVE_SPI_INIT_MTYPE_SHIFT-SQ_BUF_RSRC_WORD3_MTYPE_SHIFT)	    //get MTYPE bits into position
+    s_or_b32	    s_save_buf_rsrc3,	s_save_buf_rsrc3,  s_save_tmp						//or MTYPE
+
+    //FIXME  right now s_save_m0/s_save_mem_offset use tma_lo/tma_hi  (might need to save them before using them?)
+    s_mov_b32	    s_save_m0,		m0								    //save M0
+
+    /*	    global mem offset		*/
+    s_mov_b32	    s_save_mem_offset,	0x0									//mem offset initial value = 0
+
+
+
+
+    /*	    save HW registers	*/
+    //////////////////////////////
+
+  L_SAVE_HWREG:
+	// HWREG SR memory offset : size(VGPR)+size(SGPR)
+       get_vgpr_size_bytes(s_save_mem_offset)
+       get_sgpr_size_bytes(s_save_tmp)
+       s_add_u32 s_save_mem_offset, s_save_mem_offset, s_save_tmp
+
+
+    s_mov_b32	    s_save_buf_rsrc2, 0x4				//NUM_RECORDS	in bytes
+    if (SWIZZLE_EN)
+	s_add_u32	s_save_buf_rsrc2, s_save_buf_rsrc2, 0x0			    //FIXME need to use swizzle to enable bounds checking?
+    else
+	s_mov_b32	s_save_buf_rsrc2,  0x1000000				    //NUM_RECORDS in bytes
+    end
+
+
+    write_hwreg_to_mem(s_save_m0, s_save_buf_rsrc0, s_save_mem_offset)			//M0
+
+    if ((EMU_RUN_HACK) && (EMU_RUN_HACK_SAVE_FIRST_TIME))
+	s_add_u32 s_save_pc_lo, s_save_pc_lo, 4		    //pc[31:0]+4
+	s_addc_u32 s_save_pc_hi, s_save_pc_hi, 0x0	    //carry bit over
+    end
+
+    write_hwreg_to_mem(s_save_pc_lo, s_save_buf_rsrc0, s_save_mem_offset)		    //PC
+    write_hwreg_to_mem(s_save_pc_hi, s_save_buf_rsrc0, s_save_mem_offset)
+    write_hwreg_to_mem(s_save_exec_lo, s_save_buf_rsrc0, s_save_mem_offset)		//EXEC
+    write_hwreg_to_mem(s_save_exec_hi, s_save_buf_rsrc0, s_save_mem_offset)
+    write_hwreg_to_mem(s_save_status, s_save_buf_rsrc0, s_save_mem_offset)		//STATUS
+
+    //s_save_trapsts conflicts with s_save_alloc_size
+    s_getreg_b32    s_save_trapsts, hwreg(HW_REG_TRAPSTS)
+    write_hwreg_to_mem(s_save_trapsts, s_save_buf_rsrc0, s_save_mem_offset)		//TRAPSTS
+
+    write_hwreg_to_mem(xnack_mask_lo, s_save_buf_rsrc0, s_save_mem_offset)	    //XNACK_MASK_LO
+    write_hwreg_to_mem(xnack_mask_hi, s_save_buf_rsrc0, s_save_mem_offset)	    //XNACK_MASK_HI
+
+    //use s_save_tmp would introduce conflict here between s_save_tmp and s_save_buf_rsrc2
+    s_getreg_b32    s_save_m0, hwreg(HW_REG_MODE)						    //MODE
+    write_hwreg_to_mem(s_save_m0, s_save_buf_rsrc0, s_save_mem_offset)
+
+
+
+    /*	    the first wave in the threadgroup	 */
+    s_and_b32	    s_save_tmp, s_save_spi_init_hi, S_SAVE_SPI_INIT_FIRST_WAVE_MASK	// extract fisrt wave bit
+    s_mov_b32	     s_save_exec_hi, 0x0
+    s_or_b32	     s_save_exec_hi, s_save_tmp, s_save_exec_hi				 // save first wave bit to s_save_exec_hi.bits[26]
+
+
+    /*		save SGPRs	*/
+	// Save SGPR before LDS save, then the s0 to s4 can be used during LDS save...
+    //////////////////////////////
+
+    // SGPR SR memory offset : size(VGPR)
+    get_vgpr_size_bytes(s_save_mem_offset)
+    // TODO, change RSRC word to rearrange memory layout for SGPRS
+
+    s_getreg_b32    s_save_alloc_size, hwreg(HW_REG_GPR_ALLOC,SQ_WAVE_GPR_ALLOC_SGPR_SIZE_SHIFT,SQ_WAVE_GPR_ALLOC_SGPR_SIZE_SIZE)		//spgr_size
+    s_add_u32	    s_save_alloc_size, s_save_alloc_size, 1
+    s_lshl_b32	    s_save_alloc_size, s_save_alloc_size, 4			    //Number of SGPRs = (sgpr_size + 1) * 16   (non-zero value)
+
+    if (SGPR_SAVE_USE_SQC)
+	s_lshl_b32	s_save_buf_rsrc2,   s_save_alloc_size, 2		    //NUM_RECORDS in bytes
+    else
+	s_lshl_b32	s_save_buf_rsrc2,   s_save_alloc_size, 8		    //NUM_RECORDS in bytes (64 threads)
+    end
+
+    if (SWIZZLE_EN)
+	s_add_u32	s_save_buf_rsrc2, s_save_buf_rsrc2, 0x0			    //FIXME need to use swizzle to enable bounds checking?
+    else
+	s_mov_b32	s_save_buf_rsrc2,  0x1000000				    //NUM_RECORDS in bytes
+    end
+
+
+    // backup s_save_buf_rsrc0,1 to s_save_pc_lo/hi, since write_16sgpr_to_mem function will change the rsrc0
+    //s_mov_b64 s_save_pc_lo, s_save_buf_rsrc0
+    s_mov_b64 s_save_xnack_mask_lo, s_save_buf_rsrc0
+    s_add_u32 s_save_buf_rsrc0, s_save_buf_rsrc0, s_save_mem_offset
+    s_addc_u32 s_save_buf_rsrc1, s_save_buf_rsrc1, 0
+
+    s_mov_b32	    m0, 0x0			    //SGPR initial index value =0
+    s_nop	    0x0				    //Manually inserted wait states
+  L_SAVE_SGPR_LOOP:
+    // SGPR is allocated in 16 SGPR granularity
+    s_movrels_b64   s0, s0     //s0 = s[0+m0], s1 = s[1+m0]
+    s_movrels_b64   s2, s2     //s2 = s[2+m0], s3 = s[3+m0]
+    s_movrels_b64   s4, s4     //s4 = s[4+m0], s5 = s[5+m0]
+    s_movrels_b64   s6, s6     //s6 = s[6+m0], s7 = s[7+m0]
+    s_movrels_b64   s8, s8     //s8 = s[8+m0], s9 = s[9+m0]
+    s_movrels_b64   s10, s10   //s10 = s[10+m0], s11 = s[11+m0]
+    s_movrels_b64   s12, s12   //s12 = s[12+m0], s13 = s[13+m0]
+    s_movrels_b64   s14, s14   //s14 = s[14+m0], s15 = s[15+m0]
+
+    write_16sgpr_to_mem(s0, s_save_buf_rsrc0, s_save_mem_offset) //PV: the best performance should be using s_buffer_store_dwordx4
+    s_add_u32	    m0, m0, 16							    //next sgpr index
+    s_cmp_lt_u32    m0, s_save_alloc_size					    //scc = (m0 < s_save_alloc_size) ? 1 : 0
+    s_cbranch_scc1  L_SAVE_SGPR_LOOP					//SGPR save is complete?
+    // restore s_save_buf_rsrc0,1
+    //s_mov_b64 s_save_buf_rsrc0, s_save_pc_lo
+    s_mov_b64 s_save_buf_rsrc0, s_save_xnack_mask_lo
+
+
+
+
+    /*		save first 4 VGPR, then LDS save could use   */
+	// each wave will alloc 4 vgprs at least...
+    /////////////////////////////////////////////////////////////////////////////////////
+
+    s_mov_b32	    s_save_mem_offset, 0
+    s_mov_b32	    exec_lo, 0xFFFFFFFF						    //need every thread from now on
+    s_mov_b32	    exec_hi, 0xFFFFFFFF
+    s_mov_b32	    xnack_mask_lo, 0x0
+    s_mov_b32	    xnack_mask_hi, 0x0
+
+    if (SWIZZLE_EN)
+	s_add_u32	s_save_buf_rsrc2, s_save_buf_rsrc2, 0x0			    //FIXME need to use swizzle to enable bounds checking?
+    else
+	s_mov_b32	s_save_buf_rsrc2,  0x1000000				    //NUM_RECORDS in bytes
+    end
+
+
+    // VGPR Allocated in 4-GPR granularity
+
+if G8SR_VGPR_SR_IN_DWX4
+	// the const stride for DWx4 is 4*4 bytes
+	s_and_b32 s_save_buf_rsrc1, s_save_buf_rsrc1, 0x0000FFFF   // reset const stride to 0
+	s_or_b32  s_save_buf_rsrc1, s_save_buf_rsrc1, G8SR_SAVE_BUF_RSRC_WORD1_STRIDE_DWx4  // const stride to 4*4 bytes
+
+	buffer_store_dwordx4 v0, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1
+
+	s_and_b32 s_save_buf_rsrc1, s_save_buf_rsrc1, 0x0000FFFF   // reset const stride to 0
+	s_or_b32  s_save_buf_rsrc1, s_save_buf_rsrc1, S_SAVE_BUF_RSRC_WORD1_STRIDE  // reset const stride to 4 bytes
+else
+	buffer_store_dword v0, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1
+	buffer_store_dword v1, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1  offset:256
+	buffer_store_dword v2, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1  offset:256*2
+	buffer_store_dword v3, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1  offset:256*3
+end
+
+
+
+    /*		save LDS	*/
+    //////////////////////////////
+
+  L_SAVE_LDS:
+
+	// Change EXEC to all threads...
+    s_mov_b32	    exec_lo, 0xFFFFFFFF	  //need every thread from now on
+    s_mov_b32	    exec_hi, 0xFFFFFFFF
+
+    s_getreg_b32    s_save_alloc_size, hwreg(HW_REG_LDS_ALLOC,SQ_WAVE_LDS_ALLOC_LDS_SIZE_SHIFT,SQ_WAVE_LDS_ALLOC_LDS_SIZE_SIZE)		    //lds_size
+    s_and_b32	    s_save_alloc_size, s_save_alloc_size, 0xFFFFFFFF		    //lds_size is zero?
+    s_cbranch_scc0  L_SAVE_LDS_DONE									       //no lds used? jump to L_SAVE_DONE
+
+    s_barrier		    //LDS is used? wait for other waves in the same TG
+    s_and_b32	    s_save_tmp, s_save_exec_hi, S_SAVE_SPI_INIT_FIRST_WAVE_MASK		       //exec is still used here
+    s_cbranch_scc0  L_SAVE_LDS_DONE
+
+	// first wave do LDS save;
+
+    s_lshl_b32	    s_save_alloc_size, s_save_alloc_size, 6			    //LDS size in dwords = lds_size * 64dw
+    s_lshl_b32	    s_save_alloc_size, s_save_alloc_size, 2			    //LDS size in bytes
+    s_mov_b32	    s_save_buf_rsrc2,  s_save_alloc_size			    //NUM_RECORDS in bytes
+
+    // LDS at offset: size(VGPR)+SIZE(SGPR)+SIZE(HWREG)
+    //
+    get_vgpr_size_bytes(s_save_mem_offset)
+    get_sgpr_size_bytes(s_save_tmp)
+    s_add_u32  s_save_mem_offset, s_save_mem_offset, s_save_tmp
+    s_add_u32 s_save_mem_offset, s_save_mem_offset, get_hwreg_size_bytes()
+
+
+    if (SWIZZLE_EN)
+	s_add_u32	s_save_buf_rsrc2, s_save_buf_rsrc2, 0x0	      //FIXME need to use swizzle to enable bounds checking?
+    else
+	s_mov_b32	s_save_buf_rsrc2,  0x1000000		      //NUM_RECORDS in bytes
+    end
+
+    s_mov_b32	    m0, 0x0						  //lds_offset initial value = 0
+
+
+var LDS_DMA_ENABLE = 0
+var UNROLL = 0
+if UNROLL==0 && LDS_DMA_ENABLE==1
+	s_mov_b32  s3, 256*2
+	s_nop 0
+	s_nop 0
+	s_nop 0
+  L_SAVE_LDS_LOOP:
+	//TODO: looks the 2 buffer_store/load clause for s/r will hurt performance.???
+    if (SAVE_LDS)     //SPI always alloc LDS space in 128DW granularity
+	    buffer_store_lds_dword s_save_buf_rsrc0, s_save_mem_offset lds:1		// first 64DW
+	    buffer_store_lds_dword s_save_buf_rsrc0, s_save_mem_offset lds:1 offset:256 // second 64DW
+    end
+
+    s_add_u32	    m0, m0, s3						//every buffer_store_lds does 256 bytes
+    s_add_u32	    s_save_mem_offset, s_save_mem_offset, s3				//mem offset increased by 256 bytes
+    s_cmp_lt_u32    m0, s_save_alloc_size						//scc=(m0 < s_save_alloc_size) ? 1 : 0
+    s_cbranch_scc1  L_SAVE_LDS_LOOP							//LDS save is complete?
+
+elsif LDS_DMA_ENABLE==1 && UNROLL==1 // UNROOL	, has ichace miss
+      // store from higest LDS address to lowest
+      s_mov_b32	 s3, 256*2
+      s_sub_u32	 m0, s_save_alloc_size, s3
+      s_add_u32 s_save_mem_offset, s_save_mem_offset, m0
+      s_lshr_b32 s_save_alloc_size, s_save_alloc_size, 9   // how many 128 trunks...
+      s_sub_u32 s_save_alloc_size, 128, s_save_alloc_size   // store from higheset addr to lowest
+      s_mul_i32 s_save_alloc_size, s_save_alloc_size, 6*4   // PC offset increment,  each LDS save block cost 6*4 Bytes instruction
+      s_add_u32 s_save_alloc_size, s_save_alloc_size, 3*4   //2is the below 2 inst...//s_addc and s_setpc
+      s_nop 0
+      s_nop 0
+      s_nop 0	//pad 3 dw to let LDS_DMA align with 64Bytes
+      s_getpc_b64 s[0:1]			      // reuse s[0:1], since s[0:1] already saved
+      s_add_u32	  s0, s0,s_save_alloc_size
+      s_addc_u32  s1, s1, 0
+      s_setpc_b64 s[0:1]
+
+
+       for var i =0; i< 128; i++
+	    // be careful to make here a 64Byte aligned address, which could improve performance...
+	    buffer_store_lds_dword s_save_buf_rsrc0, s_save_mem_offset lds:1 offset:0		// first 64DW
+	    buffer_store_lds_dword s_save_buf_rsrc0, s_save_mem_offset lds:1 offset:256		  // second 64DW
+
+	if i!=127
+	s_sub_u32  m0, m0, s3	   // use a sgpr to shrink 2DW-inst to 1DW inst to improve performance , i.e.  pack more LDS_DMA inst to one Cacheline
+	    s_sub_u32  s_save_mem_offset, s_save_mem_offset,  s3
+	    end
+       end
+
+else   // BUFFER_STORE
+      v_mbcnt_lo_u32_b32 v2, 0xffffffff, 0x0
+      v_mbcnt_hi_u32_b32 v3, 0xffffffff, v2	// tid
+      v_mul_i32_i24 v2, v3, 8	// tid*8
+      v_mov_b32 v3, 256*2
+      s_mov_b32 m0, 0x10000
+      s_mov_b32 s0, s_save_buf_rsrc3
+      s_and_b32 s_save_buf_rsrc3, s_save_buf_rsrc3, 0xFF7FFFFF	  // disable add_tid
+      s_or_b32 s_save_buf_rsrc3, s_save_buf_rsrc3, 0x58000   //DFMT
+
+L_SAVE_LDS_LOOP_VECTOR:
+      ds_read_b64 v[0:1], v2	//x =LDS[a], byte address
+      s_waitcnt lgkmcnt(0)
+      buffer_store_dwordx2  v[0:1], v2, s_save_buf_rsrc0, s_save_mem_offset offen:1  glc:1  slc:1
+//	s_waitcnt vmcnt(0)
+//	v_add_u32 v2, vcc[0:1], v2, v3
+      v_add_u32 v2, v2, v3
+      v_cmp_lt_u32 vcc[0:1], v2, s_save_alloc_size
+      s_cbranch_vccnz L_SAVE_LDS_LOOP_VECTOR
+
+      // restore rsrc3
+      s_mov_b32 s_save_buf_rsrc3, s0
+
+end
+
+L_SAVE_LDS_DONE:
+
+
+    /*		save VGPRs  - set the Rest VGPRs	*/
+    //////////////////////////////////////////////////////////////////////////////////////
+  L_SAVE_VGPR:
+    // VGPR SR memory offset: 0
+    // TODO rearrange the RSRC words to use swizzle for VGPR save...
+
+    s_mov_b32	    s_save_mem_offset, (0+256*4)				    // for the rest VGPRs
+    s_mov_b32	    exec_lo, 0xFFFFFFFF						    //need every thread from now on
+    s_mov_b32	    exec_hi, 0xFFFFFFFF
+
+    s_getreg_b32    s_save_alloc_size, hwreg(HW_REG_GPR_ALLOC,SQ_WAVE_GPR_ALLOC_VGPR_SIZE_SHIFT,SQ_WAVE_GPR_ALLOC_VGPR_SIZE_SIZE)		    //vpgr_size
+    s_add_u32	    s_save_alloc_size, s_save_alloc_size, 1
+    s_lshl_b32	    s_save_alloc_size, s_save_alloc_size, 2			    //Number of VGPRs = (vgpr_size + 1) * 4    (non-zero value)	  //FIXME for GFX, zero is possible
+    s_lshl_b32	    s_save_buf_rsrc2,  s_save_alloc_size, 8			    //NUM_RECORDS in bytes (64 threads*4)
+    if (SWIZZLE_EN)
+	s_add_u32	s_save_buf_rsrc2, s_save_buf_rsrc2, 0x0			    //FIXME need to use swizzle to enable bounds checking?
+    else
+	s_mov_b32	s_save_buf_rsrc2,  0x1000000				    //NUM_RECORDS in bytes
+    end
+
+
+    // VGPR Allocated in 4-GPR granularity
+
+if G8SR_VGPR_SR_IN_DWX4
+	// the const stride for DWx4 is 4*4 bytes
+	s_and_b32 s_save_buf_rsrc1, s_save_buf_rsrc1, 0x0000FFFF   // reset const stride to 0
+	s_or_b32  s_save_buf_rsrc1, s_save_buf_rsrc1, G8SR_SAVE_BUF_RSRC_WORD1_STRIDE_DWx4  // const stride to 4*4 bytes
+
+	s_mov_b32	  m0, 4	    // skip first 4 VGPRs
+	s_cmp_lt_u32	  m0, s_save_alloc_size
+	s_cbranch_scc0	  L_SAVE_VGPR_LOOP_END	    // no more vgprs
+
+	s_set_gpr_idx_on  m0, 0x1   // This will change M0
+	s_add_u32	  s_save_alloc_size, s_save_alloc_size, 0x1000	// because above inst change m0
+L_SAVE_VGPR_LOOP:
+	v_mov_b32	  v0, v0   // v0 = v[0+m0]
+	v_mov_b32	  v1, v1
+	v_mov_b32	  v2, v2
+	v_mov_b32	  v3, v3
+
+
+	buffer_store_dwordx4 v0, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1
+	s_add_u32	  m0, m0, 4
+	s_add_u32	  s_save_mem_offset, s_save_mem_offset, 256*4
+	s_cmp_lt_u32	  m0, s_save_alloc_size
+    s_cbranch_scc1  L_SAVE_VGPR_LOOP						    //VGPR save is complete?
+    s_set_gpr_idx_off
+L_SAVE_VGPR_LOOP_END:
+
+	s_and_b32 s_save_buf_rsrc1, s_save_buf_rsrc1, 0x0000FFFF   // reset const stride to 0
+	s_or_b32  s_save_buf_rsrc1, s_save_buf_rsrc1, S_SAVE_BUF_RSRC_WORD1_STRIDE  // reset const stride to 4 bytes
+else
+    // VGPR store using dw burst
+    s_mov_b32	      m0, 0x4	//VGPR initial index value =0
+    s_cmp_lt_u32      m0, s_save_alloc_size
+    s_cbranch_scc0    L_SAVE_VGPR_END
+
+
+    s_set_gpr_idx_on	m0, 0x1 //M0[7:0] = M0[7:0] and M0[15:12] = 0x1
+    s_add_u32	    s_save_alloc_size, s_save_alloc_size, 0x1000		    //add 0x1000 since we compare m0 against it later
+
+  L_SAVE_VGPR_LOOP:
+    v_mov_b32	    v0, v0		//v0 = v[0+m0]
+    v_mov_b32	    v1, v1		//v0 = v[0+m0]
+    v_mov_b32	    v2, v2		//v0 = v[0+m0]
+    v_mov_b32	    v3, v3		//v0 = v[0+m0]
+
+    if(USE_MTBUF_INSTEAD_OF_MUBUF)
+	tbuffer_store_format_x v0, v0, s_save_buf_rsrc0, s_save_mem_offset format:BUF_NUM_FORMAT_FLOAT format: BUF_DATA_FORMAT_32 slc:1 glc:1
+    else
+	buffer_store_dword v0, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1
+	buffer_store_dword v1, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1  offset:256
+	buffer_store_dword v2, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1  offset:256*2
+	buffer_store_dword v3, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1  offset:256*3
+    end
+
+    s_add_u32	    m0, m0, 4							    //next vgpr index
+    s_add_u32	    s_save_mem_offset, s_save_mem_offset, 256*4			    //every buffer_store_dword does 256 bytes
+    s_cmp_lt_u32    m0, s_save_alloc_size					    //scc = (m0 < s_save_alloc_size) ? 1 : 0
+    s_cbranch_scc1  L_SAVE_VGPR_LOOP						    //VGPR save is complete?
+    s_set_gpr_idx_off
+end
+
+L_SAVE_VGPR_END:
+
+
+
+
+
+
+    /*	   S_PGM_END_SAVED  */				    //FIXME  graphics ONLY
+    if ((EMU_RUN_HACK) && (!EMU_RUN_HACK_SAVE_NORMAL_EXIT))
+	s_and_b32 s_save_pc_hi, s_save_pc_hi, 0x0000ffff    //pc[47:32]
+	s_add_u32 s_save_pc_lo, s_save_pc_lo, 4		    //pc[31:0]+4
+	s_addc_u32 s_save_pc_hi, s_save_pc_hi, 0x0	    //carry bit over
+	s_rfe_b64 s_save_pc_lo				    //Return to the main shader program
+    else
+    end
+
+// Save Done timestamp
+if G8SR_DEBUG_TIMESTAMP
+	s_memrealtime	s_g8sr_ts_save_d
+	// SGPR SR memory offset : size(VGPR)
+	get_vgpr_size_bytes(s_save_mem_offset)
+	s_add_u32 s_save_mem_offset, s_save_mem_offset, G8SR_DEBUG_TS_SAVE_D_OFFSET
+	s_waitcnt lgkmcnt(0)	     //FIXME, will cause xnack??
+	// Need reset rsrc2??
+	s_mov_b32 m0, s_save_mem_offset
+	s_mov_b32 s_save_buf_rsrc2,  0x1000000					//NUM_RECORDS in bytes
+	s_buffer_store_dwordx2 s_g8sr_ts_save_d, s_save_buf_rsrc0, m0	    glc:1
+end
+
+
+    s_branch	L_END_PGM
+
+
+
+/**************************************************************************/
+/*			restore routine					  */
+/**************************************************************************/
+
+L_RESTORE:
+    /*	    Setup Resource Contants    */
+    if ((EMU_RUN_HACK) && (!EMU_RUN_HACK_RESTORE_NORMAL))
+	//calculate wd_addr using absolute thread id
+	v_readlane_b32 s_restore_tmp, v9, 0
+	s_lshr_b32 s_restore_tmp, s_restore_tmp, 6
+	s_mul_i32 s_restore_tmp, s_restore_tmp, WAVE_SPACE
+	s_add_i32 s_restore_spi_init_lo, s_restore_tmp, WG_BASE_ADDR_LO
+	s_mov_b32 s_restore_spi_init_hi, WG_BASE_ADDR_HI
+	s_and_b32 s_restore_spi_init_hi, s_restore_spi_init_hi, CTX_RESTORE_CONTROL
+    else
+    end
+
+if G8SR_DEBUG_TIMESTAMP
+	s_memrealtime	s_g8sr_ts_restore_s
+	s_waitcnt lgkmcnt(0)	     //FIXME, will cause xnack??
+	// tma_lo/hi are sgpr 110, 111, which will not used for 112 SGPR allocated case...
+	s_mov_b32 s_restore_pc_lo, s_g8sr_ts_restore_s[0]
+	s_mov_b32 s_restore_pc_hi, s_g8sr_ts_restore_s[1]   //backup ts to ttmp0/1, sicne exec will be finally restored..
+end
+
+
+
+    s_mov_b32	    s_restore_buf_rsrc0,    s_restore_spi_init_lo							    //base_addr_lo
+    s_and_b32	    s_restore_buf_rsrc1,    s_restore_spi_init_hi, 0x0000FFFF						    //base_addr_hi
+    s_or_b32	    s_restore_buf_rsrc1,    s_restore_buf_rsrc1,  S_RESTORE_BUF_RSRC_WORD1_STRIDE
+    s_mov_b32	    s_restore_buf_rsrc2,    0										    //NUM_RECORDS initial value = 0 (in bytes)
+    s_mov_b32	    s_restore_buf_rsrc3,    S_RESTORE_BUF_RSRC_WORD3_MISC
+    s_and_b32	    s_restore_tmp,	    s_restore_spi_init_hi, S_RESTORE_SPI_INIT_ATC_MASK
+    s_lshr_b32	    s_restore_tmp,	    s_restore_tmp, (S_RESTORE_SPI_INIT_ATC_SHIFT-SQ_BUF_RSRC_WORD1_ATC_SHIFT)	    //get ATC bit into position
+    s_or_b32	    s_restore_buf_rsrc3,    s_restore_buf_rsrc3,  s_restore_tmp						    //or ATC
+    s_and_b32	    s_restore_tmp,	    s_restore_spi_init_hi, S_RESTORE_SPI_INIT_MTYPE_MASK
+    s_lshr_b32	    s_restore_tmp,	    s_restore_tmp, (S_RESTORE_SPI_INIT_MTYPE_SHIFT-SQ_BUF_RSRC_WORD3_MTYPE_SHIFT)   //get MTYPE bits into position
+    s_or_b32	    s_restore_buf_rsrc3,    s_restore_buf_rsrc3,  s_restore_tmp						    //or MTYPE
+
+    /*	    global mem offset		*/
+//  s_mov_b32	    s_restore_mem_offset, 0x0				    //mem offset initial value = 0
+
+    /*	    the first wave in the threadgroup	 */
+    s_and_b32	    s_restore_tmp, s_restore_spi_init_hi, S_RESTORE_SPI_INIT_FIRST_WAVE_MASK
+    s_cbranch_scc0  L_RESTORE_VGPR
+
+    /*		restore LDS	*/
+    //////////////////////////////
+  L_RESTORE_LDS:
+
+    s_mov_b32	    exec_lo, 0xFFFFFFFF							    //need every thread from now on   //be consistent with SAVE although can be moved ahead
+    s_mov_b32	    exec_hi, 0xFFFFFFFF
+
+    s_getreg_b32    s_restore_alloc_size, hwreg(HW_REG_LDS_ALLOC,SQ_WAVE_LDS_ALLOC_LDS_SIZE_SHIFT,SQ_WAVE_LDS_ALLOC_LDS_SIZE_SIZE)		//lds_size
+    s_and_b32	    s_restore_alloc_size, s_restore_alloc_size, 0xFFFFFFFF		    //lds_size is zero?
+    s_cbranch_scc0  L_RESTORE_VGPR							    //no lds used? jump to L_RESTORE_VGPR
+    s_lshl_b32	    s_restore_alloc_size, s_restore_alloc_size, 6			    //LDS size in dwords = lds_size * 64dw
+    s_lshl_b32	    s_restore_alloc_size, s_restore_alloc_size, 2			    //LDS size in bytes
+    s_mov_b32	    s_restore_buf_rsrc2,    s_restore_alloc_size			    //NUM_RECORDS in bytes
+
+    // LDS at offset: size(VGPR)+SIZE(SGPR)+SIZE(HWREG)
+    //
+    get_vgpr_size_bytes(s_restore_mem_offset)
+    get_sgpr_size_bytes(s_restore_tmp)
+    s_add_u32  s_restore_mem_offset, s_restore_mem_offset, s_restore_tmp
+    s_add_u32  s_restore_mem_offset, s_restore_mem_offset, get_hwreg_size_bytes()	     //FIXME, Check if offset overflow???
+
+
+    if (SWIZZLE_EN)
+	s_add_u32	s_restore_buf_rsrc2, s_restore_buf_rsrc2, 0x0			    //FIXME need to use swizzle to enable bounds checking?
+    else
+	s_mov_b32	s_restore_buf_rsrc2,  0x1000000					    //NUM_RECORDS in bytes
+    end
+    s_mov_b32	    m0, 0x0								    //lds_offset initial value = 0
+
+  L_RESTORE_LDS_LOOP:
+    if (SAVE_LDS)
+	buffer_load_dword   v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset lds:1		       // first 64DW
+	buffer_load_dword   v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset lds:1 offset:256	       // second 64DW
+    end
+    s_add_u32	    m0, m0, 256*2						// 128 DW
+    s_add_u32	    s_restore_mem_offset, s_restore_mem_offset, 256*2		//mem offset increased by 128DW
+    s_cmp_lt_u32    m0, s_restore_alloc_size					//scc=(m0 < s_restore_alloc_size) ? 1 : 0
+    s_cbranch_scc1  L_RESTORE_LDS_LOOP							    //LDS restore is complete?
+
+
+    /*		restore VGPRs	    */
+    //////////////////////////////
+  L_RESTORE_VGPR:
+	// VGPR SR memory offset : 0
+    s_mov_b32	    s_restore_mem_offset, 0x0
+    s_mov_b32	    exec_lo, 0xFFFFFFFF							    //need every thread from now on   //be consistent with SAVE although can be moved ahead
+    s_mov_b32	    exec_hi, 0xFFFFFFFF
+
+    s_getreg_b32    s_restore_alloc_size, hwreg(HW_REG_GPR_ALLOC,SQ_WAVE_GPR_ALLOC_VGPR_SIZE_SHIFT,SQ_WAVE_GPR_ALLOC_VGPR_SIZE_SIZE)	//vpgr_size
+    s_add_u32	    s_restore_alloc_size, s_restore_alloc_size, 1
+    s_lshl_b32	    s_restore_alloc_size, s_restore_alloc_size, 2			    //Number of VGPRs = (vgpr_size + 1) * 4    (non-zero value)
+    s_lshl_b32	    s_restore_buf_rsrc2,  s_restore_alloc_size, 8			    //NUM_RECORDS in bytes (64 threads*4)
+    if (SWIZZLE_EN)
+	s_add_u32	s_restore_buf_rsrc2, s_restore_buf_rsrc2, 0x0			    //FIXME need to use swizzle to enable bounds checking?
+    else
+	s_mov_b32	s_restore_buf_rsrc2,  0x1000000					    //NUM_RECORDS in bytes
+    end
+
+if G8SR_VGPR_SR_IN_DWX4
+     get_vgpr_size_bytes(s_restore_mem_offset)
+     s_sub_u32	       s_restore_mem_offset, s_restore_mem_offset, 256*4
+
+     // the const stride for DWx4 is 4*4 bytes
+     s_and_b32 s_restore_buf_rsrc1, s_restore_buf_rsrc1, 0x0000FFFF   // reset const stride to 0
+     s_or_b32  s_restore_buf_rsrc1, s_restore_buf_rsrc1, G8SR_RESTORE_BUF_RSRC_WORD1_STRIDE_DWx4  // const stride to 4*4 bytes
+
+     s_mov_b32	       m0, s_restore_alloc_size
+     s_set_gpr_idx_on  m0, 0x8	  // Note.. This will change m0
+
+L_RESTORE_VGPR_LOOP:
+     buffer_load_dwordx4 v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset slc:1 glc:1
+     s_waitcnt vmcnt(0)
+     s_sub_u32	       m0, m0, 4
+     v_mov_b32	       v0, v0	// v[0+m0] = v0
+     v_mov_b32	       v1, v1
+     v_mov_b32	       v2, v2
+     v_mov_b32	       v3, v3
+     s_sub_u32	       s_restore_mem_offset, s_restore_mem_offset, 256*4
+     s_cmp_eq_u32      m0, 0x8000
+     s_cbranch_scc0    L_RESTORE_VGPR_LOOP
+     s_set_gpr_idx_off
+
+     s_and_b32 s_restore_buf_rsrc1, s_restore_buf_rsrc1, 0x0000FFFF   // reset const stride to 0
+     s_or_b32  s_restore_buf_rsrc1, s_restore_buf_rsrc1, S_RESTORE_BUF_RSRC_WORD1_STRIDE  // const stride to 4*4 bytes
+
+else
+    // VGPR load using dw burst
+    s_mov_b32	    s_restore_mem_offset_save, s_restore_mem_offset	// restore start with v1, v0 will be the last
+    s_add_u32	    s_restore_mem_offset, s_restore_mem_offset, 256*4
+    s_mov_b32	    m0, 4				//VGPR initial index value = 1
+    s_set_gpr_idx_on  m0, 0x8			    //M0[7:0] = M0[7:0] and M0[15:12] = 0x8
+    s_add_u32	    s_restore_alloc_size, s_restore_alloc_size, 0x8000			    //add 0x8000 since we compare m0 against it later
+
+  L_RESTORE_VGPR_LOOP:
+    if(USE_MTBUF_INSTEAD_OF_MUBUF)
+	tbuffer_load_format_x v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset format:BUF_NUM_FORMAT_FLOAT format: BUF_DATA_FORMAT_32 slc:1 glc:1
+    else
+	buffer_load_dword v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset slc:1 glc:1
+	buffer_load_dword v1, v0, s_restore_buf_rsrc0, s_restore_mem_offset slc:1 glc:1 offset:256
+	buffer_load_dword v2, v0, s_restore_buf_rsrc0, s_restore_mem_offset slc:1 glc:1 offset:256*2
+	buffer_load_dword v3, v0, s_restore_buf_rsrc0, s_restore_mem_offset slc:1 glc:1 offset:256*3
+    end
+    s_waitcnt	    vmcnt(0)								    //ensure data ready
+    v_mov_b32	    v0, v0								    //v[0+m0] = v0
+    v_mov_b32	    v1, v1
+    v_mov_b32	    v2, v2
+    v_mov_b32	    v3, v3
+    s_add_u32	    m0, m0, 4								    //next vgpr index
+    s_add_u32	    s_restore_mem_offset, s_restore_mem_offset, 256*4				//every buffer_load_dword does 256 bytes
+    s_cmp_lt_u32    m0, s_restore_alloc_size						    //scc = (m0 < s_restore_alloc_size) ? 1 : 0
+    s_cbranch_scc1  L_RESTORE_VGPR_LOOP							    //VGPR restore (except v0) is complete?
+    s_set_gpr_idx_off
+											    /* VGPR restore on v0 */
+    if(USE_MTBUF_INSTEAD_OF_MUBUF)
+	tbuffer_load_format_x v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset_save format:BUF_NUM_FORMAT_FLOAT format: BUF_DATA_FORMAT_32 slc:1 glc:1
+    else
+	buffer_load_dword v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset_save    slc:1 glc:1
+	buffer_load_dword v1, v0, s_restore_buf_rsrc0, s_restore_mem_offset_save    slc:1 glc:1 offset:256
+	buffer_load_dword v2, v0, s_restore_buf_rsrc0, s_restore_mem_offset_save    slc:1 glc:1 offset:256*2
+	buffer_load_dword v3, v0, s_restore_buf_rsrc0, s_restore_mem_offset_save    slc:1 glc:1 offset:256*3
+    end
+
+end
+
+    /*		restore SGPRs	    */
+    //////////////////////////////
+
+    // SGPR SR memory offset : size(VGPR)
+    get_vgpr_size_bytes(s_restore_mem_offset)
+    get_sgpr_size_bytes(s_restore_tmp)
+    s_add_u32 s_restore_mem_offset, s_restore_mem_offset, s_restore_tmp
+    s_sub_u32 s_restore_mem_offset, s_restore_mem_offset, 16*4	   // restore SGPR from S[n] to S[0], by 16 sgprs group
+    // TODO, change RSRC word to rearrange memory layout for SGPRS
+
+    s_getreg_b32    s_restore_alloc_size, hwreg(HW_REG_GPR_ALLOC,SQ_WAVE_GPR_ALLOC_SGPR_SIZE_SHIFT,SQ_WAVE_GPR_ALLOC_SGPR_SIZE_SIZE)		    //spgr_size
+    s_add_u32	    s_restore_alloc_size, s_restore_alloc_size, 1
+    s_lshl_b32	    s_restore_alloc_size, s_restore_alloc_size, 4			    //Number of SGPRs = (sgpr_size + 1) * 16   (non-zero value)
+
+    if (SGPR_SAVE_USE_SQC)
+	s_lshl_b32	s_restore_buf_rsrc2,	s_restore_alloc_size, 2			    //NUM_RECORDS in bytes
+    else
+	s_lshl_b32	s_restore_buf_rsrc2,	s_restore_alloc_size, 8			    //NUM_RECORDS in bytes (64 threads)
+    end
+    if (SWIZZLE_EN)
+	s_add_u32	s_restore_buf_rsrc2, s_restore_buf_rsrc2, 0x0			    //FIXME need to use swizzle to enable bounds checking?
+    else
+	s_mov_b32	s_restore_buf_rsrc2,  0x1000000					    //NUM_RECORDS in bytes
+    end
+
+    s_mov_b32 m0, s_restore_alloc_size
+
+ L_RESTORE_SGPR_LOOP:
+    read_16sgpr_from_mem(s0, s_restore_buf_rsrc0, s_restore_mem_offset)	 //PV: further performance improvement can be made
+    s_waitcnt	    lgkmcnt(0)								    //ensure data ready
+
+    s_sub_u32 m0, m0, 16    // Restore from S[n] to S[0]
+    s_nop 0 // hazard SALU M0=> S_MOVREL
+
+    s_movreld_b64   s0, s0	//s[0+m0] = s0
+    s_movreld_b64   s2, s2
+    s_movreld_b64   s4, s4
+    s_movreld_b64   s6, s6
+    s_movreld_b64   s8, s8
+    s_movreld_b64   s10, s10
+    s_movreld_b64   s12, s12
+    s_movreld_b64   s14, s14
+
+    s_cmp_eq_u32    m0, 0		//scc = (m0 < s_restore_alloc_size) ? 1 : 0
+    s_cbranch_scc0  L_RESTORE_SGPR_LOOP		    //SGPR restore (except s0) is complete?
+
+    /*	    restore HW registers    */
+    //////////////////////////////
+  L_RESTORE_HWREG:
+
+
+if G8SR_DEBUG_TIMESTAMP
+      s_mov_b32 s_g8sr_ts_restore_s[0], s_restore_pc_lo
+      s_mov_b32 s_g8sr_ts_restore_s[1], s_restore_pc_hi
+end
+
+    // HWREG SR memory offset : size(VGPR)+size(SGPR)
+    get_vgpr_size_bytes(s_restore_mem_offset)
+    get_sgpr_size_bytes(s_restore_tmp)
+    s_add_u32 s_restore_mem_offset, s_restore_mem_offset, s_restore_tmp
+
+
+    s_mov_b32	    s_restore_buf_rsrc2, 0x4						    //NUM_RECORDS   in bytes
+    if (SWIZZLE_EN)
+	s_add_u32	s_restore_buf_rsrc2, s_restore_buf_rsrc2, 0x0			    //FIXME need to use swizzle to enable bounds checking?
+    else
+	s_mov_b32	s_restore_buf_rsrc2,  0x1000000					    //NUM_RECORDS in bytes
+    end
+
+    read_hwreg_from_mem(s_restore_m0, s_restore_buf_rsrc0, s_restore_mem_offset)		    //M0
+    read_hwreg_from_mem(s_restore_pc_lo, s_restore_buf_rsrc0, s_restore_mem_offset)		//PC
+    read_hwreg_from_mem(s_restore_pc_hi, s_restore_buf_rsrc0, s_restore_mem_offset)
+    read_hwreg_from_mem(s_restore_exec_lo, s_restore_buf_rsrc0, s_restore_mem_offset)		    //EXEC
+    read_hwreg_from_mem(s_restore_exec_hi, s_restore_buf_rsrc0, s_restore_mem_offset)
+    read_hwreg_from_mem(s_restore_status, s_restore_buf_rsrc0, s_restore_mem_offset)		    //STATUS
+    read_hwreg_from_mem(s_restore_trapsts, s_restore_buf_rsrc0, s_restore_mem_offset)		    //TRAPSTS
+    read_hwreg_from_mem(xnack_mask_lo, s_restore_buf_rsrc0, s_restore_mem_offset)		    //XNACK_MASK_LO
+    read_hwreg_from_mem(xnack_mask_hi, s_restore_buf_rsrc0, s_restore_mem_offset)		    //XNACK_MASK_HI
+    read_hwreg_from_mem(s_restore_mode, s_restore_buf_rsrc0, s_restore_mem_offset)		//MODE
+
+    s_waitcnt	    lgkmcnt(0)											    //from now on, it is safe to restore STATUS and IB_STS
+
+    s_and_b32 s_restore_pc_hi, s_restore_pc_hi, 0x0000ffff	//pc[47:32]	   //Do it here in order not to affect STATUS
+
+    //for normal save & restore, the saved PC points to the next inst to execute, no adjustment needs to be made, otherwise:
+    if ((EMU_RUN_HACK) && (!EMU_RUN_HACK_RESTORE_NORMAL))
+	s_add_u32 s_restore_pc_lo, s_restore_pc_lo, 8		 //pc[31:0]+8	  //two back-to-back s_trap are used (first for save and second for restore)
+	s_addc_u32  s_restore_pc_hi, s_restore_pc_hi, 0x0	 //carry bit over
+    end
+    if ((EMU_RUN_HACK) && (EMU_RUN_HACK_RESTORE_NORMAL))
+	s_add_u32 s_restore_pc_lo, s_restore_pc_lo, 4		 //pc[31:0]+4	  // save is hack through s_trap but restore is normal
+	s_addc_u32  s_restore_pc_hi, s_restore_pc_hi, 0x0	 //carry bit over
+    end
+
+    s_mov_b32	    m0,		s_restore_m0
+    s_mov_b32	    exec_lo,	s_restore_exec_lo
+    s_mov_b32	    exec_hi,	s_restore_exec_hi
+
+    s_and_b32	    s_restore_m0, SQ_WAVE_TRAPSTS_PRE_SAVECTX_MASK, s_restore_trapsts
+    s_setreg_b32    hwreg(HW_REG_TRAPSTS, SQ_WAVE_TRAPSTS_PRE_SAVECTX_SHIFT, SQ_WAVE_TRAPSTS_PRE_SAVECTX_SIZE), s_restore_m0
+    s_and_b32	    s_restore_m0, SQ_WAVE_TRAPSTS_POST_SAVECTX_MASK, s_restore_trapsts
+    s_lshr_b32	    s_restore_m0, s_restore_m0, SQ_WAVE_TRAPSTS_POST_SAVECTX_SHIFT
+    s_setreg_b32    hwreg(HW_REG_TRAPSTS, SQ_WAVE_TRAPSTS_POST_SAVECTX_SHIFT, SQ_WAVE_TRAPSTS_POST_SAVECTX_SIZE), s_restore_m0
+    //s_setreg_b32  hwreg(HW_REG_TRAPSTS),  s_restore_trapsts	   //don't overwrite SAVECTX bit as it may be set through external SAVECTX during restore
+    s_setreg_b32    hwreg(HW_REG_MODE),	    s_restore_mode
+
+    // Restore trap temporaries 6-11, 13-15 initialized by SPI debug dispatch logic
+    // ttmp SR memory offset : size(VGPR)+size(SGPR)+0x40
+    get_vgpr_size_bytes(s_restore_ttmps_lo)
+    get_sgpr_size_bytes(s_restore_ttmps_hi)
+    s_add_u32	    s_restore_ttmps_lo, s_restore_ttmps_lo, s_restore_ttmps_hi
+    s_add_u32	    s_restore_ttmps_lo, s_restore_ttmps_lo, s_restore_buf_rsrc0
+    s_addc_u32	    s_restore_ttmps_hi, s_restore_buf_rsrc1, 0x0
+    s_and_b32	    s_restore_ttmps_hi, s_restore_ttmps_hi, 0xFFFF
+    s_load_dwordx2  [ttmp6, ttmp7], [s_restore_ttmps_lo, s_restore_ttmps_hi], 0x40 glc:1
+    s_load_dwordx4  [ttmp8, ttmp9, ttmp10, ttmp11], [s_restore_ttmps_lo, s_restore_ttmps_hi], 0x48 glc:1
+    s_load_dword    ttmp13, [s_restore_ttmps_lo, s_restore_ttmps_hi], 0x58 glc:1
+    s_load_dwordx2  [ttmp14, ttmp15], [s_restore_ttmps_lo, s_restore_ttmps_hi], 0x5C glc:1
+    s_waitcnt	    lgkmcnt(0)
+
+    //reuse s_restore_m0 as a temp register
+    s_and_b32	    s_restore_m0, s_restore_pc_hi, S_SAVE_PC_HI_RCNT_MASK
+    s_lshr_b32	    s_restore_m0, s_restore_m0, S_SAVE_PC_HI_RCNT_SHIFT
+    s_lshl_b32	    s_restore_m0, s_restore_m0, SQ_WAVE_IB_STS_RCNT_SHIFT
+    s_mov_b32	    s_restore_tmp, 0x0										    //IB_STS is zero
+    s_or_b32	    s_restore_tmp, s_restore_tmp, s_restore_m0
+    s_and_b32	    s_restore_m0, s_restore_pc_hi, S_SAVE_PC_HI_FIRST_REPLAY_MASK
+    s_lshr_b32	    s_restore_m0, s_restore_m0, S_SAVE_PC_HI_FIRST_REPLAY_SHIFT
+    s_lshl_b32	    s_restore_m0, s_restore_m0, SQ_WAVE_IB_STS_FIRST_REPLAY_SHIFT
+    s_or_b32	    s_restore_tmp, s_restore_tmp, s_restore_m0
+    s_and_b32	    s_restore_m0, s_restore_status, SQ_WAVE_STATUS_INST_ATC_MASK
+    s_lshr_b32	    s_restore_m0, s_restore_m0, SQ_WAVE_STATUS_INST_ATC_SHIFT
+    s_setreg_b32    hwreg(HW_REG_IB_STS),   s_restore_tmp
+
+    s_and_b64	 exec, exec, exec  // Restore STATUS.EXECZ, not writable by s_setreg_b32
+    s_and_b64	 vcc, vcc, vcc	// Restore STATUS.VCCZ, not writable by s_setreg_b32
+    s_setreg_b32    hwreg(HW_REG_STATUS),   s_restore_status	 // SCC is included, which is changed by previous salu
+
+    s_barrier							//barrier to ensure the readiness of LDS before access attempts from any other wave in the same TG //FIXME not performance-optimal at this time
+
+if G8SR_DEBUG_TIMESTAMP
+    s_memrealtime s_g8sr_ts_restore_d
+    s_waitcnt lgkmcnt(0)
+end
+
+//  s_rfe_b64 s_restore_pc_lo					//Return to the main shader program and resume execution
+    s_rfe_restore_b64  s_restore_pc_lo, s_restore_m0		// s_restore_m0[0] is used to set STATUS.inst_atc
+
+
+/**************************************************************************/
+/*			the END						  */
+/**************************************************************************/
+L_END_PGM:
+    s_endpgm
+
+end
+
+
+/**************************************************************************/
+/*			the helper functions				  */
+/**************************************************************************/
+
+//Only for save hwreg to mem
+function write_hwreg_to_mem(s, s_rsrc, s_mem_offset)
+	s_mov_b32 exec_lo, m0			//assuming exec_lo is not needed anymore from this point on
+	s_mov_b32 m0, s_mem_offset
+	s_buffer_store_dword s, s_rsrc, m0	glc:1
+	ack_sqc_store_workaround()
+	s_add_u32	s_mem_offset, s_mem_offset, 4
+	s_mov_b32   m0, exec_lo
+end
+
+
+// HWREG are saved before SGPRs, so all HWREG could be use.
+function write_16sgpr_to_mem(s, s_rsrc, s_mem_offset)
+
+	s_buffer_store_dwordx4 s[0], s_rsrc, 0	glc:1
+	ack_sqc_store_workaround()
+	s_buffer_store_dwordx4 s[4], s_rsrc, 16	 glc:1
+	ack_sqc_store_workaround()
+	s_buffer_store_dwordx4 s[8], s_rsrc, 32	 glc:1
+	ack_sqc_store_workaround()
+	s_buffer_store_dwordx4 s[12], s_rsrc, 48 glc:1
+	ack_sqc_store_workaround()
+	s_add_u32	s_rsrc[0], s_rsrc[0], 4*16
+	s_addc_u32	s_rsrc[1], s_rsrc[1], 0x0	      // +scc
+end
+
+
+function read_hwreg_from_mem(s, s_rsrc, s_mem_offset)
+    s_buffer_load_dword s, s_rsrc, s_mem_offset	    glc:1
+    s_add_u32	    s_mem_offset, s_mem_offset, 4
+end
+
+function read_16sgpr_from_mem(s, s_rsrc, s_mem_offset)
+    s_buffer_load_dwordx16 s, s_rsrc, s_mem_offset	glc:1
+    s_sub_u32	    s_mem_offset, s_mem_offset, 4*16
+end
+
+
+
+function get_lds_size_bytes(s_lds_size_byte)
+    // SQ LDS granularity is 64DW, while PGM_RSRC2.lds_size is in granularity 128DW
+    s_getreg_b32   s_lds_size_byte, hwreg(HW_REG_LDS_ALLOC, SQ_WAVE_LDS_ALLOC_LDS_SIZE_SHIFT, SQ_WAVE_LDS_ALLOC_LDS_SIZE_SIZE)		// lds_size
+    s_lshl_b32	   s_lds_size_byte, s_lds_size_byte, 8			    //LDS size in dwords = lds_size * 64 *4Bytes    // granularity 64DW
+end
+
+function get_vgpr_size_bytes(s_vgpr_size_byte)
+    s_getreg_b32   s_vgpr_size_byte, hwreg(HW_REG_GPR_ALLOC,SQ_WAVE_GPR_ALLOC_VGPR_SIZE_SHIFT,SQ_WAVE_GPR_ALLOC_VGPR_SIZE_SIZE)	 //vpgr_size
+    s_add_u32	   s_vgpr_size_byte, s_vgpr_size_byte, 1
+    s_lshl_b32	   s_vgpr_size_byte, s_vgpr_size_byte, (2+8) //Number of VGPRs = (vgpr_size + 1) * 4 * 64 * 4	(non-zero value)   //FIXME for GFX, zero is possible
+end
+
+function get_sgpr_size_bytes(s_sgpr_size_byte)
+    s_getreg_b32   s_sgpr_size_byte, hwreg(HW_REG_GPR_ALLOC,SQ_WAVE_GPR_ALLOC_SGPR_SIZE_SHIFT,SQ_WAVE_GPR_ALLOC_SGPR_SIZE_SIZE)	 //spgr_size
+    s_add_u32	   s_sgpr_size_byte, s_sgpr_size_byte, 1
+    s_lshl_b32	   s_sgpr_size_byte, s_sgpr_size_byte, 6 //Number of SGPRs = (sgpr_size + 1) * 16 *4   (non-zero value)
+end
+
+function get_hwreg_size_bytes
+    return 128 //HWREG size 128 bytes
+end
+
+function ack_sqc_store_workaround
+    if ACK_SQC_STORE
+        s_waitcnt lgkmcnt(0)
+    end
+end
+
+
+#endif
+
+static const uint32_t cwsr_trap_gfx9_hex[] = {
+	0xbf820001, 0xbf820158,
+	0xb8f8f802, 0x89788678,
+	0xb8f1f803, 0x866eff71,
+	0x00000400, 0xbf850034,
+	0x866eff71, 0x00000800,
+	0xbf850003, 0x866eff71,
+	0x00000100, 0xbf840008,
+	0x866eff78, 0x00002000,
+	0xbf840001, 0xbf810000,
+	0x8778ff78, 0x00002000,
+	0x80ec886c, 0x82ed806d,
+	0xb8eef807, 0x866fff6e,
+	0x001f8000, 0x8e6f8b6f,
+	0x8977ff77, 0xfc000000,
+	0x87776f77, 0x896eff6e,
+	0x001f8000, 0xb96ef807,
+	0xb8f0f812, 0xb8f1f813,
+	0x8ef08870, 0xc0071bb8,
+	0x00000000, 0xbf8cc07f,
+	0xc0071c38, 0x00000008,
+	0xbf8cc07f, 0x86ee6e6e,
+	0xbf840001, 0xbe801d6e,
+	0xb8f1f803, 0x8671ff71,
+	0x000001ff, 0xbf850002,
+	0x806c846c, 0x826d806d,
+	0x866dff6d, 0x0000ffff,
+	0x8f6e8b77, 0x866eff6e,
+	0x001f8000, 0xb96ef807,
+	0x86fe7e7e, 0x86ea6a6a,
+	0xb978f802, 0xbe801f6c,
+	0x866dff6d, 0x0000ffff,
+	0xbef00080, 0xb9700283,
+	0xb8f02407, 0x8e709c70,
+	0x876d706d, 0xb8f003c7,
+	0x8e709b70, 0x876d706d,
+	0xb8f0f807, 0x8670ff70,
+	0x00007fff, 0xb970f807,
+	0xbeee007e, 0xbeef007f,
+	0xbefe0180, 0xbf900004,
+	0xbf8e0002, 0xbf88fffe,
+	0xb8f02a05, 0x80708170,
+	0x8e708a70, 0xb8f11605,
+	0x80718171, 0x8e718671,
+	0x80707170, 0x80707e70,
+	0x8271807f, 0x8671ff71,
+	0x0000ffff, 0xc0471cb8,
+	0x00000040, 0xbf8cc07f,
+	0xc04b1d38, 0x00000048,
+	0xbf8cc07f, 0xc0431e78,
+	0x00000058, 0xbf8cc07f,
+	0xc0471eb8, 0x0000005c,
+	0xbf8cc07f, 0xbef4007e,
+	0x8675ff7f, 0x0000ffff,
+	0x8775ff75, 0x00040000,
+	0xbef60080, 0xbef700ff,
+	0x00807fac, 0x8670ff7f,
+	0x08000000, 0x8f708370,
+	0x87777077, 0x8670ff7f,
+	0x70000000, 0x8f708170,
+	0x87777077, 0xbefb007c,
+	0xbefa0080, 0xb8fa2a05,
+	0x807a817a, 0x8e7a8a7a,
+	0xb8f01605, 0x80708170,
+	0x8e708670, 0x807a707a,
+	0xbef60084, 0xbef600ff,
+	0x01000000, 0xbefe007c,
+	0xbefc007a, 0xc0611efa,
+	0x0000007c, 0xbf8cc07f,
+	0x807a847a, 0xbefc007e,
+	0xbefe007c, 0xbefc007a,
+	0xc0611b3a, 0x0000007c,
+	0xbf8cc07f, 0x807a847a,
+	0xbefc007e, 0xbefe007c,
+	0xbefc007a, 0xc0611b7a,
+	0x0000007c, 0xbf8cc07f,
+	0x807a847a, 0xbefc007e,
+	0xbefe007c, 0xbefc007a,
+	0xc0611bba, 0x0000007c,
+	0xbf8cc07f, 0x807a847a,
+	0xbefc007e, 0xbefe007c,
+	0xbefc007a, 0xc0611bfa,
+	0x0000007c, 0xbf8cc07f,
+	0x807a847a, 0xbefc007e,
+	0xbefe007c, 0xbefc007a,
+	0xc0611e3a, 0x0000007c,
+	0xbf8cc07f, 0x807a847a,
+	0xbefc007e, 0xb8f1f803,
+	0xbefe007c, 0xbefc007a,
+	0xc0611c7a, 0x0000007c,
+	0xbf8cc07f, 0x807a847a,
+	0xbefc007e, 0xbefe007c,
+	0xbefc007a, 0xc0611a3a,
+	0x0000007c, 0xbf8cc07f,
+	0x807a847a, 0xbefc007e,
+	0xbefe007c, 0xbefc007a,
+	0xc0611a7a, 0x0000007c,
+	0xbf8cc07f, 0x807a847a,
+	0xbefc007e, 0xb8fbf801,
+	0xbefe007c, 0xbefc007a,
+	0xc0611efa, 0x0000007c,
+	0xbf8cc07f, 0x807a847a,
+	0xbefc007e, 0x8670ff7f,
+	0x04000000, 0xbeef0080,
+	0x876f6f70, 0xb8fa2a05,
+	0x807a817a, 0x8e7a8a7a,
+	0xb8f11605, 0x80718171,
+	0x8e718471, 0x8e768271,
+	0xbef600ff, 0x01000000,
+	0xbef20174, 0x80747a74,
+	0x82758075, 0xbefc0080,
+	0xbf800000, 0xbe802b00,
+	0xbe822b02, 0xbe842b04,
+	0xbe862b06, 0xbe882b08,
+	0xbe8a2b0a, 0xbe8c2b0c,
+	0xbe8e2b0e, 0xc06b003a,
+	0x00000000, 0xbf8cc07f,
+	0xc06b013a, 0x00000010,
+	0xbf8cc07f, 0xc06b023a,
+	0x00000020, 0xbf8cc07f,
+	0xc06b033a, 0x00000030,
+	0xbf8cc07f, 0x8074c074,
+	0x82758075, 0x807c907c,
+	0xbf0a717c, 0xbf85ffe7,
+	0xbef40172, 0xbefa0080,
+	0xbefe00c1, 0xbeff00c1,
+	0xbee80080, 0xbee90080,
+	0xbef600ff, 0x01000000,
+	0xe0724000, 0x7a1d0000,
+	0xe0724100, 0x7a1d0100,
+	0xe0724200, 0x7a1d0200,
+	0xe0724300, 0x7a1d0300,
+	0xbefe00c1, 0xbeff00c1,
+	0xb8f14306, 0x8671c171,
+	0xbf84002c, 0xbf8a0000,
+	0x8670ff6f, 0x04000000,
+	0xbf840028, 0x8e718671,
+	0x8e718271, 0xbef60071,
+	0xb8fa2a05, 0x807a817a,
+	0x8e7a8a7a, 0xb8f01605,
+	0x80708170, 0x8e708670,
+	0x807a707a, 0x807aff7a,
+	0x00000080, 0xbef600ff,
+	0x01000000, 0xbefc0080,
+	0xd28c0002, 0x000100c1,
+	0xd28d0003, 0x000204c1,
+	0xd1060002, 0x00011103,
+	0x7e0602ff, 0x00000200,
+	0xbefc00ff, 0x00010000,
+	0xbe800077, 0x8677ff77,
+	0xff7fffff, 0x8777ff77,
+	0x00058000, 0xd8ec0000,
+	0x00000002, 0xbf8cc07f,
+	0xe0765000, 0x7a1d0002,
+	0x68040702, 0xd0c9006a,
+	0x0000e302, 0xbf87fff7,
+	0xbef70000, 0xbefa00ff,
+	0x00000400, 0xbefe00c1,
+	0xbeff00c1, 0xb8f12a05,
+	0x80718171, 0x8e718271,
+	0x8e768871, 0xbef600ff,
+	0x01000000, 0xbefc0084,
+	0xbf0a717c, 0xbf840015,
+	0xbf11017c, 0x8071ff71,
+	0x00001000, 0x7e000300,
+	0x7e020301, 0x7e040302,
+	0x7e060303, 0xe0724000,
+	0x7a1d0000, 0xe0724100,
+	0x7a1d0100, 0xe0724200,
+	0x7a1d0200, 0xe0724300,
+	0x7a1d0300, 0x807c847c,
+	0x807aff7a, 0x00000400,
+	0xbf0a717c, 0xbf85ffef,
+	0xbf9c0000, 0xbf8200d9,
+	0xbef4007e, 0x8675ff7f,
+	0x0000ffff, 0x8775ff75,
+	0x00040000, 0xbef60080,
+	0xbef700ff, 0x00807fac,
+	0x866eff7f, 0x08000000,
+	0x8f6e836e, 0x87776e77,
+	0x866eff7f, 0x70000000,
+	0x8f6e816e, 0x87776e77,
+	0x866eff7f, 0x04000000,
+	0xbf84001e, 0xbefe00c1,
+	0xbeff00c1, 0xb8ef4306,
+	0x866fc16f, 0xbf840019,
+	0x8e6f866f, 0x8e6f826f,
+	0xbef6006f, 0xb8f82a05,
+	0x80788178, 0x8e788a78,
+	0xb8ee1605, 0x806e816e,
+	0x8e6e866e, 0x80786e78,
+	0x8078ff78, 0x00000080,
+	0xbef600ff, 0x01000000,
+	0xbefc0080, 0xe0510000,
+	0x781d0000, 0xe0510100,
+	0x781d0000, 0x807cff7c,
+	0x00000200, 0x8078ff78,
+	0x00000200, 0xbf0a6f7c,
+	0xbf85fff6, 0xbef80080,
+	0xbefe00c1, 0xbeff00c1,
+	0xb8ef2a05, 0x806f816f,
+	0x8e6f826f, 0x8e76886f,
+	0xbef600ff, 0x01000000,
+	0xbeee0078, 0x8078ff78,
+	0x00000400, 0xbefc0084,
+	0xbf11087c, 0x806fff6f,
+	0x00008000, 0xe0524000,
+	0x781d0000, 0xe0524100,
+	0x781d0100, 0xe0524200,
+	0x781d0200, 0xe0524300,
+	0x781d0300, 0xbf8c0f70,
+	0x7e000300, 0x7e020301,
+	0x7e040302, 0x7e060303,
+	0x807c847c, 0x8078ff78,
+	0x00000400, 0xbf0a6f7c,
+	0xbf85ffee, 0xbf9c0000,
+	0xe0524000, 0x6e1d0000,
+	0xe0524100, 0x6e1d0100,
+	0xe0524200, 0x6e1d0200,
+	0xe0524300, 0x6e1d0300,
+	0xb8f82a05, 0x80788178,
+	0x8e788a78, 0xb8ee1605,
+	0x806e816e, 0x8e6e866e,
+	0x80786e78, 0x80f8c078,
+	0xb8ef1605, 0x806f816f,
+	0x8e6f846f, 0x8e76826f,
+	0xbef600ff, 0x01000000,
+	0xbefc006f, 0xc031003a,
+	0x00000078, 0x80f8c078,
+	0xbf8cc07f, 0x80fc907c,
+	0xbf800000, 0xbe802d00,
+	0xbe822d02, 0xbe842d04,
+	0xbe862d06, 0xbe882d08,
+	0xbe8a2d0a, 0xbe8c2d0c,
+	0xbe8e2d0e, 0xbf06807c,
+	0xbf84fff0, 0xb8f82a05,
+	0x80788178, 0x8e788a78,
+	0xb8ee1605, 0x806e816e,
+	0x8e6e866e, 0x80786e78,
+	0xbef60084, 0xbef600ff,
+	0x01000000, 0xc0211bfa,
+	0x00000078, 0x80788478,
+	0xc0211b3a, 0x00000078,
+	0x80788478, 0xc0211b7a,
+	0x00000078, 0x80788478,
+	0xc0211eba, 0x00000078,
+	0x80788478, 0xc0211efa,
+	0x00000078, 0x80788478,
+	0xc0211c3a, 0x00000078,
+	0x80788478, 0xc0211c7a,
+	0x00000078, 0x80788478,
+	0xc0211a3a, 0x00000078,
+	0x80788478, 0xc0211a7a,
+	0x00000078, 0x80788478,
+	0xc0211cfa, 0x00000078,
+	0x80788478, 0xbf8cc07f,
+	0x866dff6d, 0x0000ffff,
+	0xbefc006f, 0xbefe007a,
+	0xbeff007b, 0x866f71ff,
+	0x000003ff, 0xb96f4803,
+	0x866f71ff, 0xfffff800,
+	0x8f6f8b6f, 0xb96fa2c3,
+	0xb973f801, 0xb8ee2a05,
+	0x806e816e, 0x8e6e8a6e,
+	0xb8ef1605, 0x806f816f,
+	0x8e6f866f, 0x806e6f6e,
+	0x806e746e, 0x826f8075,
+	0x866fff6f, 0x0000ffff,
+	0xc0071cb7, 0x00000040,
+	0xc00b1d37, 0x00000048,
+	0xc0031e77, 0x00000058,
+	0xc0071eb7, 0x0000005c,
+	0xbf8cc07f, 0x866fff6d,
+	0xf0000000, 0x8f6f9c6f,
+	0x8e6f906f, 0xbeee0080,
+	0x876e6f6e, 0x866fff6d,
+	0x08000000, 0x8f6f9b6f,
+	0x8e6f8f6f, 0x876e6f6e,
+	0x866fff70, 0x00800000,
+	0x8f6f976f, 0xb96ef807,
+	0x86fe7e7e, 0x86ea6a6a,
+	0xb970f802, 0xbf8a0000,
+	0x95806f6c, 0xbf810000,
+};
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
index c368ce3..053f1d0 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
@@ -30,6 +30,7 @@
 #include "kfd_device_queue_manager.h"
 #include "kfd_pm4_headers_vi.h"
 #include "cwsr_trap_handler_gfx8.asm"
+#include "cwsr_trap_handler_gfx9.asm"
 #include "kfd_iommu.h"
 
 #define MQD_SIZE_ALIGNED 768
@@ -333,10 +334,16 @@ struct kfd_dev *kgd2kfd_probe(struct kgd_dev *kgd,
 static void kfd_cwsr_init(struct kfd_dev *kfd)
 {
 	if (cwsr_enable && kfd->device_info->supports_cwsr) {
-		BUILD_BUG_ON(sizeof(cwsr_trap_gfx8_hex) > PAGE_SIZE);
+		if (kfd->device_info->asic_family < CHIP_VEGA10) {
+			BUILD_BUG_ON(sizeof(cwsr_trap_gfx8_hex) > PAGE_SIZE);
+			kfd->cwsr_isa = cwsr_trap_gfx8_hex;
+			kfd->cwsr_isa_size = sizeof(cwsr_trap_gfx8_hex);
+		} else {
+			BUILD_BUG_ON(sizeof(cwsr_trap_gfx9_hex) > PAGE_SIZE);
+			kfd->cwsr_isa = cwsr_trap_gfx9_hex;
+			kfd->cwsr_isa_size = sizeof(cwsr_trap_gfx9_hex);
+		}
 
-		kfd->cwsr_isa = cwsr_trap_gfx8_hex;
-		kfd->cwsr_isa_size = sizeof(cwsr_trap_gfx8_hex);
 		kfd->cwsr_enabled = true;
 	}
 }
-- 
cgit v1.1


From 6106dce9559ec5a4b1a97302f2fcc508e40d2747 Mon Sep 17 00:00:00 2001
From: welu <Wei.Lu2@amd.com>
Date: Tue, 10 Apr 2018 17:33:17 -0400
Subject: drm/amdkfd: Try to enable atomics for all GPUs

Report failure to enable atomics only on GPUs that require them.
This allows GPUs that don't require atomics to function, but can
benefit if they are available. This is the case for Vega10, which
doesn't use atomics for basic functioning of the MEC, AQL and HWS
microcode. So it can work without atomics. But shader programs can
still use atomic instructions on systems that support PCIe atomics.

Signed-off-by: welu <Wei.Lu2@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/kfd_device.c | 27 +++++++++++++--------------
 1 file changed, 13 insertions(+), 14 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
index 053f1d0..0e64fb2 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
@@ -290,7 +290,7 @@ struct kfd_dev *kgd2kfd_probe(struct kgd_dev *kgd,
 	struct pci_dev *pdev, const struct kfd2kgd_calls *f2g)
 {
 	struct kfd_dev *kfd;
-
+	int ret;
 	const struct kfd_device_info *device_info =
 					lookup_device_info(pdev->device);
 
@@ -299,19 +299,18 @@ struct kfd_dev *kgd2kfd_probe(struct kgd_dev *kgd,
 		return NULL;
 	}
 
-	if (device_info->needs_pci_atomics) {
-		/* Allow BIF to recode atomics to PCIe 3.0
-		 * AtomicOps. 32 and 64-bit requests are possible and
-		 * must be supported.
-		 */
-		if (pci_enable_atomic_ops_to_root(pdev,
-				PCI_EXP_DEVCAP2_ATOMIC_COMP32 |
-				PCI_EXP_DEVCAP2_ATOMIC_COMP64) < 0) {
-			dev_info(kfd_device,
-				"skipped device %x:%x, PCI rejects atomics",
-				 pdev->vendor, pdev->device);
-			return NULL;
-		}
+	/* Allow BIF to recode atomics to PCIe 3.0 AtomicOps.
+	 * 32 and 64-bit requests are possible and must be
+	 * supported.
+	 */
+	ret = pci_enable_atomic_ops_to_root(pdev,
+			PCI_EXP_DEVCAP2_ATOMIC_COMP32 |
+			PCI_EXP_DEVCAP2_ATOMIC_COMP64);
+	if (device_info->needs_pci_atomics && ret < 0) {
+		dev_info(kfd_device,
+			 "skipped device %x:%x, PCI rejects atomics\n",
+			 pdev->vendor, pdev->device);
+		return NULL;
 	}
 
 	kfd = kzalloc(sizeof(*kfd), GFP_KERNEL);
-- 
cgit v1.1


From 389056e5fef477c838dc20a08d6f1de960cf027b Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Tue, 10 Apr 2018 17:33:18 -0400
Subject: drm/amdkfd: Add Vega10 topology and device info

* Report 64-bit doorbells as HSA_CAP_DOORBELL_TYPE_2_0 in topology
* Report cache information in topology (duplicates GFXv8 info for now)
* Add device info for Vega10 support in KFD

Raven is not enabled at this time as it needs additional changes in
DQM to work with a single SDMA engine.

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/kfd_crat.c     | 11 +++++++++
 drivers/gpu/drm/amd/amdkfd/kfd_device.c   | 37 +++++++++++++++++++++++++++++++
 drivers/gpu/drm/amd/amdkfd/kfd_topology.c |  6 +++++
 drivers/gpu/drm/amd/amdkfd/kfd_topology.h |  1 +
 4 files changed, 55 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
index 4f126ef..296b3f2 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
@@ -132,6 +132,9 @@ static struct kfd_gpu_cache_info carrizo_cache_info[] = {
 #define fiji_cache_info  carrizo_cache_info
 #define polaris10_cache_info carrizo_cache_info
 #define polaris11_cache_info carrizo_cache_info
+/* TODO - check & update Vega10 cache details */
+#define vega10_cache_info carrizo_cache_info
+#define raven_cache_info carrizo_cache_info
 
 static void kfd_populated_cu_info_cpu(struct kfd_topology_device *dev,
 		struct crat_subtype_computeunit *cu)
@@ -603,6 +606,14 @@ static int kfd_fill_gpu_cache_info(struct kfd_dev *kdev,
 		pcache_info = polaris11_cache_info;
 		num_of_cache_types = ARRAY_SIZE(polaris11_cache_info);
 		break;
+	case CHIP_VEGA10:
+		pcache_info = vega10_cache_info;
+		num_of_cache_types = ARRAY_SIZE(vega10_cache_info);
+		break;
+	case CHIP_RAVEN:
+		pcache_info = raven_cache_info;
+		num_of_cache_types = ARRAY_SIZE(raven_cache_info);
+		break;
 	default:
 		return -EINVAL;
 	}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
index 0e64fb2..dd6c753 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
@@ -182,6 +182,34 @@ static const struct kfd_device_info polaris11_device_info = {
 	.needs_pci_atomics = true,
 };
 
+static const struct kfd_device_info vega10_device_info = {
+	.asic_family = CHIP_VEGA10,
+	.max_pasid_bits = 16,
+	.max_no_of_hqd  = 24,
+	.doorbell_size  = 8,
+	.ih_ring_entry_size = 8 * sizeof(uint32_t),
+	.event_interrupt_class = &event_interrupt_class_v9,
+	.num_of_watch_points = 4,
+	.mqd_size_aligned = MQD_SIZE_ALIGNED,
+	.supports_cwsr = true,
+	.needs_iommu_device = false,
+	.needs_pci_atomics = false,
+};
+
+static const struct kfd_device_info vega10_vf_device_info = {
+	.asic_family = CHIP_VEGA10,
+	.max_pasid_bits = 16,
+	.max_no_of_hqd  = 24,
+	.doorbell_size  = 8,
+	.ih_ring_entry_size = 8 * sizeof(uint32_t),
+	.event_interrupt_class = &event_interrupt_class_v9,
+	.num_of_watch_points = 4,
+	.mqd_size_aligned = MQD_SIZE_ALIGNED,
+	.supports_cwsr = true,
+	.needs_iommu_device = false,
+	.needs_pci_atomics = false,
+};
+
 
 struct kfd_deviceid {
 	unsigned short did;
@@ -261,6 +289,15 @@ static const struct kfd_deviceid supported_devices[] = {
 	{ 0x67EB, &polaris11_device_info },	/* Polaris11 */
 	{ 0x67EF, &polaris11_device_info },	/* Polaris11 */
 	{ 0x67FF, &polaris11_device_info },	/* Polaris11 */
+	{ 0x6860, &vega10_device_info },	/* Vega10 */
+	{ 0x6861, &vega10_device_info },	/* Vega10 */
+	{ 0x6862, &vega10_device_info },	/* Vega10 */
+	{ 0x6863, &vega10_device_info },	/* Vega10 */
+	{ 0x6864, &vega10_device_info },	/* Vega10 */
+	{ 0x6867, &vega10_device_info },	/* Vega10 */
+	{ 0x6868, &vega10_device_info },	/* Vega10 */
+	{ 0x686C, &vega10_vf_device_info },	/* Vega10  vf*/
+	{ 0x687F, &vega10_device_info },	/* Vega10 */
 };
 
 static int kfd_gtt_sa_init(struct kfd_dev *kfd, unsigned int buf_size,
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
index ac28abc..bc95d4df 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
@@ -1239,6 +1239,12 @@ int kfd_topology_add_device(struct kfd_dev *gpu)
 			HSA_CAP_DOORBELL_TYPE_TOTALBITS_SHIFT) &
 			HSA_CAP_DOORBELL_TYPE_TOTALBITS_MASK);
 		break;
+	case CHIP_VEGA10:
+	case CHIP_RAVEN:
+		dev->node_props.capability |= ((HSA_CAP_DOORBELL_TYPE_2_0 <<
+			HSA_CAP_DOORBELL_TYPE_TOTALBITS_SHIFT) &
+			HSA_CAP_DOORBELL_TYPE_TOTALBITS_MASK);
+		break;
 	default:
 		WARN(1, "Unexpected ASIC family %u",
 		     dev->gpu->device_info->asic_family);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.h b/drivers/gpu/drm/amd/amdkfd/kfd_topology.h
index eb54cfc..7d9c3f9 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.h
@@ -45,6 +45,7 @@
 
 #define HSA_CAP_DOORBELL_TYPE_PRE_1_0		0x0
 #define HSA_CAP_DOORBELL_TYPE_1_0		0x1
+#define HSA_CAP_DOORBELL_TYPE_2_0		0x2
 #define HSA_CAP_AQL_QUEUE_DOUBLE_MAP		0x00004000
 
 struct kfd_node_properties {
-- 
cgit v1.1


From ffdf16edfbbe77f5f5c3c87fe8d7387ecd16241b Mon Sep 17 00:00:00 2001
From: Colin Ian King <colin.king@canonical.com>
Date: Tue, 10 Apr 2018 13:33:12 +0100
Subject: drm/i915/gvt: fix memory leak of a cmd_entry struct on error exit
 path

The error exit path when a duplicate is found does not kfree and cmd_entry
struct and hence there is a small memory leak.  Fix this by kfree'ing it.

Detected by CoverityScan, CID#1370198 ("Resource Leak")

Fixes: be1da7070aea ("drm/i915/gvt: vGPU command scanner")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
---
 drivers/gpu/drm/i915/gvt/cmd_parser.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/i915/gvt/cmd_parser.c b/drivers/gpu/drm/i915/gvt/cmd_parser.c
index db6b94d..6bd56ec 100644
--- a/drivers/gpu/drm/i915/gvt/cmd_parser.c
+++ b/drivers/gpu/drm/i915/gvt/cmd_parser.c
@@ -2863,6 +2863,7 @@ static int init_cmd_table(struct intel_gvt *gvt)
 		if (info) {
 			gvt_err("%s %s duplicated\n", e->info->name,
 					info->name);
+			kfree(e);
 			return -EEXIST;
 		}
 
-- 
cgit v1.1


From 86993018d7d23b934d1c884be0fbf0bcfa15b8c5 Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Thu, 15 Mar 2018 16:40:02 -0400
Subject: drm/amdgpu: Add CM_TEST_DEBUG regs for DCN

We'd like to use them for reading DCN debug status.

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../gpu/drm/amd/include/asic_reg/dcn/dcn_1_0_offset.h | 19 ++++++++++++++++---
 .../drm/amd/include/asic_reg/dcn/dcn_1_0_sh_mask.h    |  8 ++++++++
 2 files changed, 24 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_1_0_offset.h b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_1_0_offset.h
index 4ccf968..721c611 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_1_0_offset.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_1_0_offset.h
@@ -3895,6 +3895,10 @@
 #define mmCM0_CM_MEM_PWR_CTRL_BASE_IDX                                                                 2
 #define mmCM0_CM_MEM_PWR_STATUS                                                                        0x0d33
 #define mmCM0_CM_MEM_PWR_STATUS_BASE_IDX                                                               2
+#define mmCM0_CM_TEST_DEBUG_INDEX                                                                      0x0d35
+#define mmCM0_CM_TEST_DEBUG_INDEX_BASE_IDX                                                             2
+#define mmCM0_CM_TEST_DEBUG_DATA                                                                       0x0d36
+#define mmCM0_CM_TEST_DEBUG_DATA_BASE_IDX                                                              2
 
 
 // addressBlock: dce_dc_dpp0_dispdec_dpp_dcperfmon_dc_perfmon_dispdec
@@ -4367,7 +4371,10 @@
 #define mmCM1_CM_MEM_PWR_CTRL_BASE_IDX                                                                 2
 #define mmCM1_CM_MEM_PWR_STATUS                                                                        0x0e4e
 #define mmCM1_CM_MEM_PWR_STATUS_BASE_IDX                                                               2
-
+#define mmCM1_CM_TEST_DEBUG_INDEX                                                                      0x0e50
+#define mmCM1_CM_TEST_DEBUG_INDEX_BASE_IDX                                                             2
+#define mmCM1_CM_TEST_DEBUG_DATA                                                                       0x0e51
+#define mmCM1_CM_TEST_DEBUG_DATA_BASE_IDX                                                              2
 
 // addressBlock: dce_dc_dpp1_dispdec_dpp_dcperfmon_dc_perfmon_dispdec
 // base address: 0x399c
@@ -4839,7 +4846,10 @@
 #define mmCM2_CM_MEM_PWR_CTRL_BASE_IDX                                                                 2
 #define mmCM2_CM_MEM_PWR_STATUS                                                                        0x0f69
 #define mmCM2_CM_MEM_PWR_STATUS_BASE_IDX                                                               2
-
+#define mmCM2_CM_TEST_DEBUG_INDEX                                                                      0x0f6b
+#define mmCM2_CM_TEST_DEBUG_INDEX_BASE_IDX                                                             2
+#define mmCM2_CM_TEST_DEBUG_DATA                                                                       0x0f6c
+#define mmCM2_CM_TEST_DEBUG_DATA_BASE_IDX                                                              2
 
 // addressBlock: dce_dc_dpp2_dispdec_dpp_dcperfmon_dc_perfmon_dispdec
 // base address: 0x3e08
@@ -5311,7 +5321,10 @@
 #define mmCM3_CM_MEM_PWR_CTRL_BASE_IDX                                                                 2
 #define mmCM3_CM_MEM_PWR_STATUS                                                                        0x1084
 #define mmCM3_CM_MEM_PWR_STATUS_BASE_IDX                                                               2
-
+#define mmCM3_CM_TEST_DEBUG_INDEX                                                                      0x1086
+#define mmCM3_CM_TEST_DEBUG_INDEX_BASE_IDX                                                             2
+#define mmCM3_CM_TEST_DEBUG_DATA                                                                       0x1087
+#define mmCM3_CM_TEST_DEBUG_DATA_BASE_IDX                                                              2
 
 // addressBlock: dce_dc_dpp3_dispdec_dpp_dcperfmon_dc_perfmon_dispdec
 // base address: 0x4274
diff --git a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_1_0_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_1_0_sh_mask.h
index e2a2f11..e7c0cad 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_1_0_sh_mask.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_1_0_sh_mask.h
@@ -14049,6 +14049,14 @@
 #define CM0_CM_MEM_PWR_STATUS__RGAM_MEM_PWR_STATE__SHIFT                                                      0x2
 #define CM0_CM_MEM_PWR_STATUS__SHARED_MEM_PWR_STATE_MASK                                                      0x00000003L
 #define CM0_CM_MEM_PWR_STATUS__RGAM_MEM_PWR_STATE_MASK                                                        0x0000000CL
+//CM0_CM_TEST_DEBUG_INDEX
+#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_INDEX__SHIFT                                                   0x0
+#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_WRITE_EN__SHIFT                                                0x8
+#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_INDEX_MASK                                                     0x000000FFL
+#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_WRITE_EN_MASK                                                  0x00000100L
+//CM0_CM_TEST_DEBUG_DATA
+#define CM0_CM_TEST_DEBUG_DATA__CM_TEST_DEBUG_DATA__SHIFT                                                     0x0
+#define CM0_CM_TEST_DEBUG_DATA__CM_TEST_DEBUG_DATA_MASK                                                       0xFFFFFFFFL
 
 
 // addressBlock: dce_dc_dpp0_dispdec_dpp_dcperfmon_dc_perfmon_dispdec
-- 
cgit v1.1


From 35d13315957f906774013ec374ce2263b665706c Mon Sep 17 00:00:00 2001
From: Martin Tsai <martin.tsai@amd.com>
Date: Wed, 7 Mar 2018 04:22:03 +0800
Subject: drm/amd/display: correct the condition in setting cursor not visible
 beyond left edge

Signed-off-by: Martin Tsai <martin.tsai@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c  | 2 +-
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
index e305c28..3356125 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
@@ -416,7 +416,7 @@ void dpp1_set_cursor_position(
 	if (src_x_offset >= (int)param->viewport_width)
 		cur_en = 0;  /* not visible beyond right edge*/
 
-	if (src_x_offset + (int)width < 0)
+	if (src_x_offset + (int)width <= 0)
 		cur_en = 0;  /* not visible beyond left edge*/
 
 	REG_UPDATE(CURSOR0_CONTROL,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
index 39b72f6..81b81e6 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
@@ -897,7 +897,7 @@ void hubp1_cursor_set_position(
 	if (src_x_offset >= (int)param->viewport_width)
 		cur_en = 0;  /* not visible beyond right edge*/
 
-	if (src_x_offset + (int)hubp->curs_attr.width < 0)
+	if (src_x_offset + (int)hubp->curs_attr.width <= 0)
 		cur_en = 0;  /* not visible beyond left edge*/
 
 	if (cur_en && REG_READ(CURSOR_SURFACE_ADDRESS) == 0)
-- 
cgit v1.1


From ba0a5aaa867d77cedb2cd6ad9e647243d9ba9650 Mon Sep 17 00:00:00 2001
From: Tony Cheng <tony.cheng@amd.com>
Date: Wed, 21 Feb 2018 16:41:42 -0500
Subject: drm/amd/display: dal 3.1.39

Signed-off-by: Tony Cheng <tony.cheng@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dc.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index fa4b3c8..4d9da9d 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -38,7 +38,7 @@
 #include "inc/compressor.h"
 #include "dml/display_mode_lib.h"
 
-#define DC_VER "3.1.38"
+#define DC_VER "3.1.39"
 
 #define MAX_SURFACES 3
 #define MAX_STREAMS 6
-- 
cgit v1.1


From 3c1a312aa4e4201efa8719e70a6dccd3acd6eba4 Mon Sep 17 00:00:00 2001
From: Yongqiang Sun <yongqiang.sun@amd.com>
Date: Wed, 7 Mar 2018 09:12:53 -0500
Subject: drm/amd/display: Retry when read dpcd caps failed.

Some DP panel not detected intermittently due to read dpcd
caps failed when doing hot plug.
[root cause] DC_HPD_CONNECT_INT_DELAY is set to 0, not delay
after HPD toggle and read dpcd data, while some panel need 4ms defer
to read.
[solution] Add a retry when read failed.

Signed-off-by: Yongqiang Sun <yongqiang.sun@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c | 16 +++++++++++-----
 1 file changed, 11 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
index 3b50535..b86325b 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
@@ -2278,6 +2278,8 @@ static bool retrieve_link_cap(struct dc_link *link)
 	union edp_configuration_cap edp_config_cap;
 	union dp_downstream_port_present ds_port = { 0 };
 	enum dc_status status = DC_ERROR_UNEXPECTED;
+	uint32_t read_dpcd_retry_cnt = 3;
+	int i;
 
 	memset(dpcd_data, '\0', sizeof(dpcd_data));
 	memset(&down_strm_port_count,
@@ -2285,11 +2287,15 @@ static bool retrieve_link_cap(struct dc_link *link)
 	memset(&edp_config_cap, '\0',
 		sizeof(union edp_configuration_cap));
 
-	status = core_link_read_dpcd(
-			link,
-			DP_DPCD_REV,
-			dpcd_data,
-			sizeof(dpcd_data));
+	for (i = 0; i < read_dpcd_retry_cnt; i++) {
+		status = core_link_read_dpcd(
+				link,
+				DP_DPCD_REV,
+				dpcd_data,
+				sizeof(dpcd_data));
+		if (status == DC_OK)
+			break;
+	}
 
 	if (status != DC_OK) {
 		dm_error("%s: Read dpcd data failed.\n", __func__);
-- 
cgit v1.1


From b552204b10ef30940d374510a1572b2eb4e24af6 Mon Sep 17 00:00:00 2001
From: Nikola Cornij <nikola.cornij@amd.com>
Date: Tue, 6 Mar 2018 13:41:38 -0500
Subject: drm/amd/display: Update ASIC header files

Also separate register address initialization between ASICs for the
registers that were removed in scaled-down variation of the ASIC.

Signed-off-by: Nikola Cornij <nikola.cornij@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c | 20 ++++++++------
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h | 32 +++++++++++++++--------
 2 files changed, 33 insertions(+), 19 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
index 81b81e6..4ca9b6e 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
@@ -535,11 +535,13 @@ void hubp1_program_deadline(
 	REG_SET(VBLANK_PARAMETERS_3, 0,
 		REFCYC_PER_META_CHUNK_VBLANK_L, dlg_attr->refcyc_per_meta_chunk_vblank_l);
 
-	REG_SET(NOM_PARAMETERS_0, 0,
-		DST_Y_PER_PTE_ROW_NOM_L, dlg_attr->dst_y_per_pte_row_nom_l);
+	if (REG(NOM_PARAMETERS_0))
+		REG_SET(NOM_PARAMETERS_0, 0,
+			DST_Y_PER_PTE_ROW_NOM_L, dlg_attr->dst_y_per_pte_row_nom_l);
 
-	REG_SET(NOM_PARAMETERS_1, 0,
-		REFCYC_PER_PTE_GROUP_NOM_L, dlg_attr->refcyc_per_pte_group_nom_l);
+	if (REG(NOM_PARAMETERS_1))
+		REG_SET(NOM_PARAMETERS_1, 0,
+			REFCYC_PER_PTE_GROUP_NOM_L, dlg_attr->refcyc_per_pte_group_nom_l);
 
 	REG_SET(NOM_PARAMETERS_4, 0,
 		DST_Y_PER_META_ROW_NOM_L, dlg_attr->dst_y_per_meta_row_nom_l);
@@ -568,11 +570,13 @@ void hubp1_program_deadline(
 	REG_SET(VBLANK_PARAMETERS_4, 0,
 		REFCYC_PER_META_CHUNK_VBLANK_C, dlg_attr->refcyc_per_meta_chunk_vblank_c);
 
-	REG_SET(NOM_PARAMETERS_2, 0,
-		DST_Y_PER_PTE_ROW_NOM_C, dlg_attr->dst_y_per_pte_row_nom_c);
+	if (REG(NOM_PARAMETERS_2))
+		REG_SET(NOM_PARAMETERS_2, 0,
+			DST_Y_PER_PTE_ROW_NOM_C, dlg_attr->dst_y_per_pte_row_nom_c);
 
-	REG_SET(NOM_PARAMETERS_3, 0,
-		REFCYC_PER_PTE_GROUP_NOM_C, dlg_attr->refcyc_per_pte_group_nom_c);
+	if (REG(NOM_PARAMETERS_3))
+		REG_SET(NOM_PARAMETERS_3, 0,
+			REFCYC_PER_PTE_GROUP_NOM_C, dlg_attr->refcyc_per_pte_group_nom_c);
 
 	REG_SET(NOM_PARAMETERS_6, 0,
 		DST_Y_PER_META_ROW_NOM_C, dlg_attr->dst_y_per_meta_row_nom_c);
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
index 4a3703e..c794ce4 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
@@ -30,6 +30,7 @@
 #define TO_DCN10_HUBP(hubp)\
 	container_of(hubp, struct dcn10_hubp, base)
 
+/* Register address initialization macro for all ASICs (including those with reduced functionality) */
 #define HUBP_REG_LIST_DCN(id)\
 	SRI(DCHUBP_CNTL, HUBP, id),\
 	SRI(HUBPREQ_DEBUG_DB, HUBP, id),\
@@ -78,16 +79,12 @@
 	SRI(REF_FREQ_TO_PIX_FREQ, HUBPREQ, id),\
 	SRI(VBLANK_PARAMETERS_1, HUBPREQ, id),\
 	SRI(VBLANK_PARAMETERS_3, HUBPREQ, id),\
-	SRI(NOM_PARAMETERS_0, HUBPREQ, id),\
-	SRI(NOM_PARAMETERS_1, HUBPREQ, id),\
 	SRI(NOM_PARAMETERS_4, HUBPREQ, id),\
 	SRI(NOM_PARAMETERS_5, HUBPREQ, id),\
 	SRI(PER_LINE_DELIVERY_PRE, HUBPREQ, id),\
 	SRI(PER_LINE_DELIVERY, HUBPREQ, id),\
 	SRI(VBLANK_PARAMETERS_2, HUBPREQ, id),\
 	SRI(VBLANK_PARAMETERS_4, HUBPREQ, id),\
-	SRI(NOM_PARAMETERS_2, HUBPREQ, id),\
-	SRI(NOM_PARAMETERS_3, HUBPREQ, id),\
 	SRI(NOM_PARAMETERS_6, HUBPREQ, id),\
 	SRI(NOM_PARAMETERS_7, HUBPREQ, id),\
 	SRI(DCN_TTU_QOS_WM, HUBPREQ, id),\
@@ -96,11 +93,19 @@
 	SRI(DCN_SURF0_TTU_CNTL1, HUBPREQ, id),\
 	SRI(DCN_SURF1_TTU_CNTL0, HUBPREQ, id),\
 	SRI(DCN_SURF1_TTU_CNTL1, HUBPREQ, id),\
-	SRI(DCN_VM_MX_L1_TLB_CNTL, HUBPREQ, id),\
 	SRI(HUBP_CLK_CNTL, HUBP, id)
 
+/* Register address initialization macro for "generic" ASICs with full functionality */
+#define HUBP_REG_LIST_DCN_GEN(id)\
+	SRI(NOM_PARAMETERS_0, HUBPREQ, id),\
+	SRI(NOM_PARAMETERS_1, HUBPREQ, id),\
+	SRI(NOM_PARAMETERS_2, HUBPREQ, id),\
+	SRI(NOM_PARAMETERS_3, HUBPREQ, id),\
+	SRI(DCN_VM_MX_L1_TLB_CNTL, HUBPREQ, id)
+
 #define HUBP_REG_LIST_DCN10(id)\
 	HUBP_REG_LIST_DCN(id),\
+	HUBP_REG_LIST_DCN_GEN(id),\
 	SRI(PREFETCH_SETTINS, HUBPREQ, id),\
 	SRI(PREFETCH_SETTINS_C, HUBPREQ, id),\
 	SRI(DCN_VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_MSB, HUBPREQ, id),\
@@ -237,6 +242,7 @@
 #define HUBP_SF(reg_name, field_name, post_fix)\
 	.field_name = reg_name ## __ ## field_name ## post_fix
 
+/* Mask/shift struct generation macro for all ASICs (including those with reduced functionality) */
 #define HUBP_MASK_SH_LIST_DCN(mask_sh)\
 	HUBP_SF(HUBP0_DCHUBP_CNTL, HUBP_BLANK_EN, mask_sh),\
 	HUBP_SF(HUBP0_DCHUBP_CNTL, HUBP_TTU_DISABLE, mask_sh),\
@@ -335,8 +341,6 @@
 	HUBP_SF(HUBPREQ0_REF_FREQ_TO_PIX_FREQ, REF_FREQ_TO_PIX_FREQ, mask_sh),\
 	HUBP_SF(HUBPREQ0_VBLANK_PARAMETERS_1, REFCYC_PER_PTE_GROUP_VBLANK_L, mask_sh),\
 	HUBP_SF(HUBPREQ0_VBLANK_PARAMETERS_3, REFCYC_PER_META_CHUNK_VBLANK_L, mask_sh),\
-	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_0, DST_Y_PER_PTE_ROW_NOM_L, mask_sh),\
-	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_1, REFCYC_PER_PTE_GROUP_NOM_L, mask_sh),\
 	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_4, DST_Y_PER_META_ROW_NOM_L, mask_sh),\
 	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_5, REFCYC_PER_META_CHUNK_NOM_L, mask_sh),\
 	HUBP_SF(HUBPREQ0_PER_LINE_DELIVERY_PRE, REFCYC_PER_LINE_DELIVERY_PRE_L, mask_sh),\
@@ -345,8 +349,6 @@
 	HUBP_SF(HUBPREQ0_PER_LINE_DELIVERY, REFCYC_PER_LINE_DELIVERY_C, mask_sh),\
 	HUBP_SF(HUBPREQ0_VBLANK_PARAMETERS_2, REFCYC_PER_PTE_GROUP_VBLANK_C, mask_sh),\
 	HUBP_SF(HUBPREQ0_VBLANK_PARAMETERS_4, REFCYC_PER_META_CHUNK_VBLANK_C, mask_sh),\
-	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_2, DST_Y_PER_PTE_ROW_NOM_C, mask_sh),\
-	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_3, REFCYC_PER_PTE_GROUP_NOM_C, mask_sh),\
 	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_6, DST_Y_PER_META_ROW_NOM_C, mask_sh),\
 	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_7, REFCYC_PER_META_CHUNK_NOM_C, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCN_TTU_QOS_WM, QoS_LEVEL_LOW_WM, mask_sh),\
@@ -357,12 +359,20 @@
 	HUBP_SF(HUBPREQ0_DCN_SURF0_TTU_CNTL0, QoS_LEVEL_FIXED, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCN_SURF0_TTU_CNTL0, QoS_RAMP_DISABLE, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCN_SURF0_TTU_CNTL1, REFCYC_PER_REQ_DELIVERY_PRE, mask_sh),\
-	HUBP_SF(HUBPREQ0_DCN_VM_MX_L1_TLB_CNTL, ENABLE_L1_TLB, mask_sh),\
-	HUBP_SF(HUBPREQ0_DCN_VM_MX_L1_TLB_CNTL, SYSTEM_ACCESS_MODE, mask_sh),\
 	HUBP_SF(HUBP0_HUBP_CLK_CNTL, HUBP_CLOCK_ENABLE, mask_sh)
 
+/* Mask/shift struct generation macro for "generic" ASICs with full functionality */
+#define HUBP_MASK_SH_LIST_DCN_GEN(mask_sh)\
+	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_0, DST_Y_PER_PTE_ROW_NOM_L, mask_sh),\
+	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_1, REFCYC_PER_PTE_GROUP_NOM_L, mask_sh),\
+	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_2, DST_Y_PER_PTE_ROW_NOM_C, mask_sh),\
+	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_3, REFCYC_PER_PTE_GROUP_NOM_C, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCN_VM_MX_L1_TLB_CNTL, ENABLE_L1_TLB, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCN_VM_MX_L1_TLB_CNTL, SYSTEM_ACCESS_MODE, mask_sh)
+
 #define HUBP_MASK_SH_LIST_DCN10(mask_sh)\
 	HUBP_MASK_SH_LIST_DCN(mask_sh),\
+	HUBP_MASK_SH_LIST_DCN_GEN(mask_sh),\
 	HUBP_SF(HUBPREQ0_PREFETCH_SETTINS, DST_Y_PREFETCH, mask_sh),\
 	HUBP_SF(HUBPREQ0_PREFETCH_SETTINS, VRATIO_PREFETCH, mask_sh),\
 	HUBP_SF(HUBPREQ0_PREFETCH_SETTINS_C, VRATIO_PREFETCH_C, mask_sh),\
-- 
cgit v1.1


From e4b3f6f299436be812aca4845bd20f592eaf074e Mon Sep 17 00:00:00 2001
From: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Date: Thu, 8 Mar 2018 12:08:01 -0500
Subject: drm/amd/display: fix Polaris 12 bw bounding box

Signed-off-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c | 121 ++++++++++++++++++++++-
 drivers/gpu/drm/amd/display/dc/inc/dce_calcs.h   |   1 +
 2 files changed, 120 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c b/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c
index 0cbab81..821502b 100644
--- a/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c
+++ b/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c
@@ -52,10 +52,11 @@ static enum bw_calcs_version bw_calcs_version_from_asic_id(struct hw_asic_id asi
 		return BW_CALCS_VERSION_CARRIZO;
 
 	case FAMILY_VI:
+		if (ASIC_REV_IS_POLARIS12_V(asic_id.hw_internal_rev))
+			return BW_CALCS_VERSION_POLARIS12;
 		if (ASIC_REV_IS_POLARIS10_P(asic_id.hw_internal_rev))
 			return BW_CALCS_VERSION_POLARIS10;
-		if (ASIC_REV_IS_POLARIS11_M(asic_id.hw_internal_rev) ||
-				ASIC_REV_IS_POLARIS12_V(asic_id.hw_internal_rev))
+		if (ASIC_REV_IS_POLARIS11_M(asic_id.hw_internal_rev))
 			return BW_CALCS_VERSION_POLARIS11;
 		return BW_CALCS_VERSION_INVALID;
 
@@ -2373,6 +2374,122 @@ void bw_calcs_init(struct bw_calcs_dceip *bw_dceip,
 		dceip.scatter_gather_pte_request_rows_in_tiling_mode = 2;
 		dceip.mcifwr_all_surfaces_burst_time = bw_int_to_fixed(0);
 		break;
+	case BW_CALCS_VERSION_POLARIS12:
+		vbios.memory_type = bw_def_gddr5;
+		vbios.dram_channel_width_in_bits = 32;
+		vbios.number_of_dram_channels = asic_id.vram_width / vbios.dram_channel_width_in_bits;
+		vbios.number_of_dram_banks = 8;
+		vbios.high_yclk = bw_int_to_fixed(6000);
+		vbios.mid_yclk = bw_int_to_fixed(3200);
+		vbios.low_yclk = bw_int_to_fixed(1000);
+		vbios.low_sclk = bw_int_to_fixed(678);
+		vbios.mid1_sclk = bw_int_to_fixed(864);
+		vbios.mid2_sclk = bw_int_to_fixed(900);
+		vbios.mid3_sclk = bw_int_to_fixed(920);
+		vbios.mid4_sclk = bw_int_to_fixed(940);
+		vbios.mid5_sclk = bw_int_to_fixed(960);
+		vbios.mid6_sclk = bw_int_to_fixed(980);
+		vbios.high_sclk = bw_int_to_fixed(1049);
+		vbios.low_voltage_max_dispclk = bw_int_to_fixed(459);
+		vbios.mid_voltage_max_dispclk = bw_int_to_fixed(654);
+		vbios.high_voltage_max_dispclk = bw_int_to_fixed(1108);
+		vbios.low_voltage_max_phyclk = bw_int_to_fixed(540);
+		vbios.mid_voltage_max_phyclk = bw_int_to_fixed(810);
+		vbios.high_voltage_max_phyclk = bw_int_to_fixed(810);
+		vbios.data_return_bus_width = bw_int_to_fixed(32);
+		vbios.trc = bw_int_to_fixed(48);
+		if (vbios.number_of_dram_channels == 2) // 64-bit
+			vbios.dmifmc_urgent_latency = bw_int_to_fixed(4);
+		else
+			vbios.dmifmc_urgent_latency = bw_int_to_fixed(3);
+		vbios.stutter_self_refresh_exit_latency = bw_int_to_fixed(5);
+		vbios.stutter_self_refresh_entry_latency = bw_int_to_fixed(0);
+		vbios.nbp_state_change_latency = bw_int_to_fixed(250);
+		vbios.mcifwrmc_urgent_latency = bw_int_to_fixed(10);
+		vbios.scatter_gather_enable = false;
+		vbios.down_spread_percentage = bw_frc_to_fixed(5, 10);
+		vbios.cursor_width = 32;
+		vbios.average_compression_rate = 4;
+		vbios.number_of_request_slots_gmc_reserves_for_dmif_per_channel = 256;
+		vbios.blackout_duration = bw_int_to_fixed(0); /* us */
+		vbios.maximum_blackout_recovery_time = bw_int_to_fixed(0);
+
+		dceip.max_average_percent_of_ideal_port_bw_display_can_use_in_normal_system_operation = 100;
+		dceip.max_average_percent_of_ideal_drambw_display_can_use_in_normal_system_operation = 100;
+		dceip.percent_of_ideal_port_bw_received_after_urgent_latency = 100;
+		dceip.large_cursor = false;
+		dceip.dmif_request_buffer_size = bw_int_to_fixed(768);
+		dceip.dmif_pipe_en_fbc_chunk_tracker = false;
+		dceip.cursor_max_outstanding_group_num = 1;
+		dceip.lines_interleaved_into_lb = 2;
+		dceip.chunk_width = 256;
+		dceip.number_of_graphics_pipes = 5;
+		dceip.number_of_underlay_pipes = 0;
+		dceip.low_power_tiling_mode = 0;
+		dceip.display_write_back_supported = true;
+		dceip.argb_compression_support = true;
+		dceip.underlay_vscaler_efficiency6_bit_per_component =
+			bw_frc_to_fixed(35556, 10000);
+		dceip.underlay_vscaler_efficiency8_bit_per_component =
+			bw_frc_to_fixed(34286, 10000);
+		dceip.underlay_vscaler_efficiency10_bit_per_component =
+			bw_frc_to_fixed(32, 10);
+		dceip.underlay_vscaler_efficiency12_bit_per_component =
+			bw_int_to_fixed(3);
+		dceip.graphics_vscaler_efficiency6_bit_per_component =
+			bw_frc_to_fixed(35, 10);
+		dceip.graphics_vscaler_efficiency8_bit_per_component =
+			bw_frc_to_fixed(34286, 10000);
+		dceip.graphics_vscaler_efficiency10_bit_per_component =
+			bw_frc_to_fixed(32, 10);
+		dceip.graphics_vscaler_efficiency12_bit_per_component =
+			bw_int_to_fixed(3);
+		dceip.alpha_vscaler_efficiency = bw_int_to_fixed(3);
+		dceip.max_dmif_buffer_allocated = 4;
+		dceip.graphics_dmif_size = 12288;
+		dceip.underlay_luma_dmif_size = 19456;
+		dceip.underlay_chroma_dmif_size = 23552;
+		dceip.pre_downscaler_enabled = true;
+		dceip.underlay_downscale_prefetch_enabled = true;
+		dceip.lb_write_pixels_per_dispclk = bw_int_to_fixed(1);
+		dceip.lb_size_per_component444 = bw_int_to_fixed(245952);
+		dceip.graphics_lb_nodownscaling_multi_line_prefetching = true;
+		dceip.stutter_and_dram_clock_state_change_gated_before_cursor =
+			bw_int_to_fixed(1);
+		dceip.underlay420_luma_lb_size_per_component = bw_int_to_fixed(
+			82176);
+		dceip.underlay420_chroma_lb_size_per_component =
+			bw_int_to_fixed(164352);
+		dceip.underlay422_lb_size_per_component = bw_int_to_fixed(
+			82176);
+		dceip.cursor_chunk_width = bw_int_to_fixed(64);
+		dceip.cursor_dcp_buffer_lines = bw_int_to_fixed(4);
+		dceip.underlay_maximum_width_efficient_for_tiling =
+			bw_int_to_fixed(1920);
+		dceip.underlay_maximum_height_efficient_for_tiling =
+			bw_int_to_fixed(1080);
+		dceip.peak_pte_request_to_eviction_ratio_limiting_multiple_displays_or_single_rotated_display =
+			bw_frc_to_fixed(3, 10);
+		dceip.peak_pte_request_to_eviction_ratio_limiting_single_display_no_rotation =
+			bw_int_to_fixed(25);
+		dceip.minimum_outstanding_pte_request_limit = bw_int_to_fixed(
+			2);
+		dceip.maximum_total_outstanding_pte_requests_allowed_by_saw =
+			bw_int_to_fixed(128);
+		dceip.limit_excessive_outstanding_dmif_requests = true;
+		dceip.linear_mode_line_request_alternation_slice =
+			bw_int_to_fixed(64);
+		dceip.scatter_gather_lines_of_pte_prefetching_in_linear_mode =
+			32;
+		dceip.display_write_back420_luma_mcifwr_buffer_size = 12288;
+		dceip.display_write_back420_chroma_mcifwr_buffer_size = 8192;
+		dceip.request_efficiency = bw_frc_to_fixed(8, 10);
+		dceip.dispclk_per_request = bw_int_to_fixed(2);
+		dceip.dispclk_ramping_factor = bw_frc_to_fixed(105, 100);
+		dceip.display_pipe_throughput_factor = bw_frc_to_fixed(105, 100);
+		dceip.scatter_gather_pte_request_rows_in_tiling_mode = 2;
+		dceip.mcifwr_all_surfaces_burst_time = bw_int_to_fixed(0);
+		break;
 	case BW_CALCS_VERSION_STONEY:
 		vbios.memory_type = bw_def_gddr5;
 		vbios.dram_channel_width_in_bits = 64;
diff --git a/drivers/gpu/drm/amd/display/dc/inc/dce_calcs.h b/drivers/gpu/drm/amd/display/dc/inc/dce_calcs.h
index a9bfe9f..0bd87f2 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/dce_calcs.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/dce_calcs.h
@@ -42,6 +42,7 @@ enum bw_calcs_version {
 	BW_CALCS_VERSION_CARRIZO,
 	BW_CALCS_VERSION_POLARIS10,
 	BW_CALCS_VERSION_POLARIS11,
+	BW_CALCS_VERSION_POLARIS12,
 	BW_CALCS_VERSION_STONEY,
 	BW_CALCS_VERSION_VEGA10
 };
-- 
cgit v1.1


From deb0aac6af79265408c2b3c62b3d1150e7c46a1b Mon Sep 17 00:00:00 2001
From: Nikola Cornij <nikola.cornij@amd.com>
Date: Fri, 9 Mar 2018 14:45:07 -0500
Subject: drm/amd/display: Rename feature-specific register address init macro

Signed-off-by: Nikola Cornij <nikola.cornij@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
index c794ce4..e0d6d32 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
@@ -95,8 +95,8 @@
 	SRI(DCN_SURF1_TTU_CNTL1, HUBPREQ, id),\
 	SRI(HUBP_CLK_CNTL, HUBP, id)
 
-/* Register address initialization macro for "generic" ASICs with full functionality */
-#define HUBP_REG_LIST_DCN_GEN(id)\
+/* Register address initialization macro for ASICs with VM */
+#define HUBP_REG_LIST_DCN_VM(id)\
 	SRI(NOM_PARAMETERS_0, HUBPREQ, id),\
 	SRI(NOM_PARAMETERS_1, HUBPREQ, id),\
 	SRI(NOM_PARAMETERS_2, HUBPREQ, id),\
@@ -105,7 +105,7 @@
 
 #define HUBP_REG_LIST_DCN10(id)\
 	HUBP_REG_LIST_DCN(id),\
-	HUBP_REG_LIST_DCN_GEN(id),\
+	HUBP_REG_LIST_DCN_VM(id),\
 	SRI(PREFETCH_SETTINS, HUBPREQ, id),\
 	SRI(PREFETCH_SETTINS_C, HUBPREQ, id),\
 	SRI(DCN_VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_MSB, HUBPREQ, id),\
@@ -361,8 +361,8 @@
 	HUBP_SF(HUBPREQ0_DCN_SURF0_TTU_CNTL1, REFCYC_PER_REQ_DELIVERY_PRE, mask_sh),\
 	HUBP_SF(HUBP0_HUBP_CLK_CNTL, HUBP_CLOCK_ENABLE, mask_sh)
 
-/* Mask/shift struct generation macro for "generic" ASICs with full functionality */
-#define HUBP_MASK_SH_LIST_DCN_GEN(mask_sh)\
+/* Mask/shift struct generation macro for ASICs with VM */
+#define HUBP_MASK_SH_LIST_DCN_VM(mask_sh)\
 	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_0, DST_Y_PER_PTE_ROW_NOM_L, mask_sh),\
 	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_1, REFCYC_PER_PTE_GROUP_NOM_L, mask_sh),\
 	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_2, DST_Y_PER_PTE_ROW_NOM_C, mask_sh),\
@@ -372,7 +372,7 @@
 
 #define HUBP_MASK_SH_LIST_DCN10(mask_sh)\
 	HUBP_MASK_SH_LIST_DCN(mask_sh),\
-	HUBP_MASK_SH_LIST_DCN_GEN(mask_sh),\
+	HUBP_MASK_SH_LIST_DCN_VM(mask_sh),\
 	HUBP_SF(HUBPREQ0_PREFETCH_SETTINS, DST_Y_PREFETCH, mask_sh),\
 	HUBP_SF(HUBPREQ0_PREFETCH_SETTINS, VRATIO_PREFETCH, mask_sh),\
 	HUBP_SF(HUBPREQ0_PREFETCH_SETTINS_C, VRATIO_PREFETCH_C, mask_sh),\
-- 
cgit v1.1


From a12c3b7d4e2ac7837c23620ebc3e42b397c1c321 Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Thu, 8 Mar 2018 22:05:35 -0500
Subject: drm/amd/display: Don't read EDID in atomic_check

We shouldn't attempt to read EDID in atomic_check. We really shouldn't
even be modifying the connector object, or any other non-state object,
but this is a start at least.

Moving EDID cleanup to dm_dp_mst_connector_destroy from
dm_dp_destroy_mst_connector to ensure the EDID is still available for
headless mode.

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../amd/display/amdgpu_dm/amdgpu_dm_mst_types.c    | 32 +++++++---------------
 1 file changed, 10 insertions(+), 22 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
index 8291d74..305292a 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
@@ -161,6 +161,11 @@ dm_dp_mst_connector_destroy(struct drm_connector *connector)
 	struct amdgpu_dm_connector *amdgpu_dm_connector = to_amdgpu_dm_connector(connector);
 	struct amdgpu_encoder *amdgpu_encoder = amdgpu_dm_connector->mst_encoder;
 
+	if (amdgpu_dm_connector->edid) {
+		kfree(amdgpu_dm_connector->edid);
+		amdgpu_dm_connector->edid = NULL;
+	}
+
 	drm_encoder_cleanup(&amdgpu_encoder->base);
 	kfree(amdgpu_encoder);
 	drm_connector_cleanup(connector);
@@ -181,28 +186,22 @@ static const struct drm_connector_funcs dm_dp_mst_connector_funcs = {
 void dm_dp_mst_dc_sink_create(struct drm_connector *connector)
 {
 	struct amdgpu_dm_connector *aconnector = to_amdgpu_dm_connector(connector);
-	struct edid *edid;
 	struct dc_sink *dc_sink;
 	struct dc_sink_init_data init_params = {
 			.link = aconnector->dc_link,
 			.sink_signal = SIGNAL_TYPE_DISPLAY_PORT_MST };
 
+	/* FIXME none of this is safe. we shouldn't touch aconnector here in
+	 * atomic_check
+	 */
+
 	/*
 	 * TODO: Need to further figure out why ddc.algo is NULL while MST port exists
 	 */
 	if (!aconnector->port || !aconnector->port->aux.ddc.algo)
 		return;
 
-	edid = drm_dp_mst_get_edid(connector, &aconnector->mst_port->mst_mgr, aconnector->port);
-
-	if (!edid) {
-		drm_mode_connector_update_edid_property(
-			&aconnector->base,
-			NULL);
-		return;
-	}
-
-	aconnector->edid = edid;
+	ASSERT(aconnector->edid);
 
 	dc_sink = dc_link_add_remote_sink(
 		aconnector->dc_link,
@@ -215,9 +214,6 @@ void dm_dp_mst_dc_sink_create(struct drm_connector *connector)
 
 	amdgpu_dm_add_sink_to_freesync_module(
 			connector, aconnector->edid);
-
-	drm_mode_connector_update_edid_property(
-					&aconnector->base, aconnector->edid);
 }
 
 static int dm_dp_mst_get_modes(struct drm_connector *connector)
@@ -424,14 +420,6 @@ static void dm_dp_destroy_mst_connector(struct drm_dp_mst_topology_mgr *mgr,
 		dc_sink_release(aconnector->dc_sink);
 		aconnector->dc_sink = NULL;
 	}
-	if (aconnector->edid) {
-		kfree(aconnector->edid);
-		aconnector->edid = NULL;
-	}
-
-	drm_mode_connector_update_edid_property(
-			&aconnector->base,
-			NULL);
 
 	aconnector->mst_connected = false;
 }
-- 
cgit v1.1


From dfd01f299987e7ede74e27d422c43846d1326010 Mon Sep 17 00:00:00 2001
From: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Date: Wed, 24 Jan 2018 14:28:30 -0500
Subject: drm/amd/display: add mpc to dtn log

Signed-off-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Reviewed-by: Wesley Chalmers <Wesley.Chalmers@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c  | 31 ++++++++++++++++------
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.c   | 17 ++++++++++++
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.h   |  5 ++++
 drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h        | 15 +++++++++++
 4 files changed, 60 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
index 8b0f6b8..999190a 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -121,20 +121,19 @@ void dcn10_log_hw_state(struct dc *dc)
 
 	dcn10_log_hubbub_state(dc);
 
-	DTN_INFO("HUBP:\t format \t addr_hi \t width \t height \t "
-			"rotation \t mirror \t  sw_mode \t "
-			"dcc_en \t blank_en \t ttu_dis \t underflow \t "
-			"min_ttu_vblank \t qos_low_wm \t qos_high_wm \n");
-
+	DTN_INFO("HUBP:  format  addr_hi  width  height  "
+			"rotation  mirror  sw_mode  "
+			"dcc_en  blank_en  ttu_dis  underflow  "
+			"min_ttu_vblank  qos_low_wm  qos_high_wm\n");
 	for (i = 0; i < pool->pipe_count; i++) {
 		struct hubp *hubp = pool->hubps[i];
 		struct dcn_hubp_state s;
 
 		hubp1_read_state(TO_DCN10_HUBP(hubp), &s);
 
-		DTN_INFO("[%d]:\t %xh \t %xh \t %d \t %d \t "
-				"%xh \t %xh \t %xh \t "
-				"%d \t %d \t %d \t %xh \t",
+		DTN_INFO("[%-2d]:  %5xh  %6xh  %5d  %6d  "
+				"%7xh  %5xh  %6xh  "
+				"%6d  %8d  %7d  %8xh \t",
 				hubp->inst,
 				s.pixel_format,
 				s.inuse_addr_hi,
@@ -153,6 +152,22 @@ void dcn10_log_hw_state(struct dc *dc)
 		DTN_INFO("\n");
 	}
 	DTN_INFO("\n");
+	for (i = 0; i < pool->pipe_count; i++) {
+		struct output_pixel_processor *opp = pool->opps[i];
+		struct mpcc *mpcc = opp->mpc_tree_params.opp_list;
+		struct mpcc_state s = {0};
+
+		while (mpcc) {
+			ASSERT(opp->mpc_tree_params.opp_id == opp->inst);
+			pool->mpc->funcs->read_mpcc_state(pool->mpc, mpcc->mpcc_id, &s);
+			DTN_INFO("[OPP%d - MPCC%d]: DPP%d MPCCBOT%x MODE:%d ALPHA_MODE:%d PREMULT:%d OVERLAP_ONLY:%d\n",
+				s.opp_id, mpcc->mpcc_id, s.dpp_id, s.bot_mpcc_id,
+				s.mode, s.alpha_mode, s.pre_multiplied_alpha, s.overlap_only);
+			mpcc = mpcc->mpcc_bot;
+			ASSERT(!mpcc || mpcc->mpcc_id == s.bot_mpcc_id);
+		}
+	}
+	DTN_INFO("\n");
 
 	DTN_INFO("OTG:\t v_bs \t v_be \t v_ss \t v_se \t vpol \t vmax \t vmin \t "
 			"h_bs \t h_be \t h_ss \t h_se \t hpol \t htot \t vtot \t underflow\n");
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.c
index 179890b..29e15a9 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.c
@@ -409,7 +409,24 @@ void mpc1_init_mpcc_list_from_hw(
 	}
 }
 
+void mpc1_read_mpcc_state(
+		struct mpc *mpc,
+		int mpcc_inst,
+		struct mpcc_state *s)
+{
+	struct dcn10_mpc *mpc10 = TO_DCN10_MPC(mpc);
+
+	REG_GET(MPCC_OPP_ID[mpcc_inst], MPCC_OPP_ID, &s->opp_id);
+	REG_GET(MPCC_TOP_SEL[mpcc_inst], MPCC_TOP_SEL, &s->dpp_id);
+	REG_GET(MPCC_BOT_SEL[mpcc_inst], MPCC_BOT_SEL, &s->bot_mpcc_id);
+	REG_GET_4(MPCC_CONTROL[mpcc_inst], MPCC_MODE, &s->mode,
+			MPCC_ALPHA_BLND_MODE, &s->alpha_mode,
+			MPCC_ALPHA_MULTIPLIED_MODE, &s->pre_multiplied_alpha,
+			MPCC_BLND_ACTIVE_OVERLAP_ONLY, &s->pre_multiplied_alpha);
+}
+
 const struct mpc_funcs dcn10_mpc_funcs = {
+	.read_mpcc_state = mpc1_read_mpcc_state,
 	.insert_plane = mpc1_insert_plane,
 	.remove_mpcc = mpc1_remove_mpcc,
 	.mpc_init = mpc1_mpc_init,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.h
index 267a299..d3d16c4 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.h
@@ -183,4 +183,9 @@ struct mpcc *mpc1_get_mpcc_for_dpp(
 	struct mpc_tree *tree,
 	int dpp_id);
 
+void mpc1_read_mpcc_state(
+		struct mpc *mpc,
+		int mpcc_inst,
+		struct mpcc_state *s);
+
 #endif
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h b/drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h
index 23a8d5e..5caacab 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h
@@ -105,7 +105,22 @@ struct mpc {
 	struct mpcc mpcc_array[MAX_MPCC];
 };
 
+struct mpcc_state {
+	uint32_t opp_id;
+	uint32_t dpp_id;
+	uint32_t bot_mpcc_id;
+	uint32_t mode;
+	uint32_t alpha_mode;
+	uint32_t pre_multiplied_alpha;
+	uint32_t overlap_only;
+};
+
 struct mpc_funcs {
+	void (*read_mpcc_state)(
+			struct mpc *mpc,
+			int mpcc_inst,
+			struct mpcc_state *s);
+
 	/*
 	 * Insert DPP into MPC tree based on specified blending position.
 	 * Only used for planes that are part of blending chain for OPP output
-- 
cgit v1.1


From 1249acefefd43006127e58acf9c67de8038d770b Mon Sep 17 00:00:00 2001
From: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Date: Thu, 8 Mar 2018 14:58:11 -0500
Subject: drm/amd/display: Add debug prints for bandwidth calculations

Using the three functions we can print the dceip, vbios and data struct
for bandwidth calculations. This is useful for debugging bandwidth
calculation issues without a debugger

Signed-off-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../gpu/drm/amd/display/dc/calcs/calcs_logger.h    | 579 +++++++++++++++++++++
 drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c   |   6 +
 drivers/gpu/drm/amd/display/dc/dc.h                |   1 +
 3 files changed, 586 insertions(+)
 create mode 100644 drivers/gpu/drm/amd/display/dc/calcs/calcs_logger.h

diff --git a/drivers/gpu/drm/amd/display/dc/calcs/calcs_logger.h b/drivers/gpu/drm/amd/display/dc/calcs/calcs_logger.h
new file mode 100644
index 0000000..fc3f98f
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/calcs/calcs_logger.h
@@ -0,0 +1,579 @@
+/*
+ * Copyright 2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef _CALCS_CALCS_LOGGER_H_
+#define _CALCS_CALCS_LOGGER_H_
+#define DC_LOGGER \
+	logger
+
+static void print_bw_calcs_dceip(struct dal_logger *logger, const struct bw_calcs_dceip *dceip)
+{
+
+	DC_LOG_BANDWIDTH_CALCS("#####################################################################");
+	DC_LOG_BANDWIDTH_CALCS("struct bw_calcs_dceip");
+	DC_LOG_BANDWIDTH_CALCS("#####################################################################");
+	DC_LOG_BANDWIDTH_CALCS("	[enum]   bw_calcs_version version %d", dceip->version);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] large_cursor: %d", dceip->large_cursor);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] dmif_pipe_en_fbc_chunk_tracker: %d", dceip->dmif_pipe_en_fbc_chunk_tracker);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] display_write_back_supported: %d", dceip->display_write_back_supported);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] argb_compression_support: %d", dceip->argb_compression_support);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] pre_downscaler_enabled: %d", dceip->pre_downscaler_enabled);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] underlay_downscale_prefetch_enabled: %d",
+				dceip->underlay_downscale_prefetch_enabled);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] graphics_lb_nodownscaling_multi_line_prefetching: %d",
+				dceip->graphics_lb_nodownscaling_multi_line_prefetching);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] limit_excessive_outstanding_dmif_requests: %d",
+				dceip->limit_excessive_outstanding_dmif_requests);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] cursor_max_outstanding_group_num: %d",
+				dceip->cursor_max_outstanding_group_num);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] lines_interleaved_into_lb: %d", dceip->lines_interleaved_into_lb);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] low_power_tiling_mode: %d", dceip->low_power_tiling_mode);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] chunk_width: %d", dceip->chunk_width);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] number_of_graphics_pipes: %d", dceip->number_of_graphics_pipes);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] number_of_underlay_pipes: %d", dceip->number_of_underlay_pipes);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] max_dmif_buffer_allocated: %d", dceip->max_dmif_buffer_allocated);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] graphics_dmif_size: %d", dceip->graphics_dmif_size);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] underlay_luma_dmif_size: %d", dceip->underlay_luma_dmif_size);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] underlay_chroma_dmif_size: %d", dceip->underlay_chroma_dmif_size);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] scatter_gather_lines_of_pte_prefetching_in_linear_mode: %d",
+				dceip->scatter_gather_lines_of_pte_prefetching_in_linear_mode);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] display_write_back420_luma_mcifwr_buffer_size: %d",
+				dceip->display_write_back420_luma_mcifwr_buffer_size);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] display_write_back420_chroma_mcifwr_buffer_size: %d",
+				dceip->display_write_back420_chroma_mcifwr_buffer_size);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] scatter_gather_pte_request_rows_in_tiling_mode: %d",
+				dceip->scatter_gather_pte_request_rows_in_tiling_mode);
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] underlay_vscaler_efficiency10_bit_per_component: %d",
+				bw_fixed_to_int(dceip->underlay_vscaler_efficiency10_bit_per_component));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] underlay_vscaler_efficiency12_bit_per_component: %d",
+				bw_fixed_to_int(dceip->underlay_vscaler_efficiency12_bit_per_component));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] graphics_vscaler_efficiency6_bit_per_component: %d",
+				bw_fixed_to_int(dceip->graphics_vscaler_efficiency6_bit_per_component));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] graphics_vscaler_efficiency8_bit_per_component: %d",
+				bw_fixed_to_int(dceip->graphics_vscaler_efficiency8_bit_per_component));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] graphics_vscaler_efficiency10_bit_per_component: %d",
+				bw_fixed_to_int(dceip->graphics_vscaler_efficiency10_bit_per_component));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] graphics_vscaler_efficiency12_bit_per_component: %d",
+				bw_fixed_to_int(dceip->graphics_vscaler_efficiency12_bit_per_component));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] alpha_vscaler_efficiency: %d",
+				bw_fixed_to_int(dceip->alpha_vscaler_efficiency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] lb_write_pixels_per_dispclk: %d",
+				bw_fixed_to_int(dceip->lb_write_pixels_per_dispclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] lb_size_per_component444: %d",
+				bw_fixed_to_int(dceip->lb_size_per_component444));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] stutter_and_dram_clock_state_change_gated_before_cursor: %d",
+				bw_fixed_to_int(dceip->stutter_and_dram_clock_state_change_gated_before_cursor));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] underlay420_luma_lb_size_per_component: %d",
+				bw_fixed_to_int(dceip->underlay420_luma_lb_size_per_component));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] underlay420_chroma_lb_size_per_component: %d",
+				bw_fixed_to_int(dceip->underlay420_chroma_lb_size_per_component));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] underlay422_lb_size_per_component: %d",
+				bw_fixed_to_int(dceip->underlay422_lb_size_per_component));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] cursor_chunk_width: %d", bw_fixed_to_int(dceip->cursor_chunk_width));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] cursor_dcp_buffer_lines: %d",
+				bw_fixed_to_int(dceip->cursor_dcp_buffer_lines));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] underlay_maximum_width_efficient_for_tiling: %d",
+				bw_fixed_to_int(dceip->underlay_maximum_width_efficient_for_tiling));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] underlay_maximum_height_efficient_for_tiling: %d",
+				bw_fixed_to_int(dceip->underlay_maximum_height_efficient_for_tiling));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] peak_pte_request_to_eviction_ratio_limiting_multiple_displays_or_single_rotated_display: %d",
+				bw_fixed_to_int(dceip->peak_pte_request_to_eviction_ratio_limiting_multiple_displays_or_single_rotated_display));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] peak_pte_request_to_eviction_ratio_limiting_single_display_no_rotation: %d",
+				bw_fixed_to_int(dceip->peak_pte_request_to_eviction_ratio_limiting_single_display_no_rotation));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] minimum_outstanding_pte_request_limit: %d",
+				bw_fixed_to_int(dceip->minimum_outstanding_pte_request_limit));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] maximum_total_outstanding_pte_requests_allowed_by_saw: %d",
+				bw_fixed_to_int(dceip->maximum_total_outstanding_pte_requests_allowed_by_saw));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] linear_mode_line_request_alternation_slice: %d",
+				bw_fixed_to_int(dceip->linear_mode_line_request_alternation_slice));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] request_efficiency: %d", bw_fixed_to_int(dceip->request_efficiency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dispclk_per_request: %d", bw_fixed_to_int(dceip->dispclk_per_request));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dispclk_ramping_factor: %d",
+				bw_fixed_to_int(dceip->dispclk_ramping_factor));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] display_pipe_throughput_factor: %d",
+				bw_fixed_to_int(dceip->display_pipe_throughput_factor));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mcifwr_all_surfaces_burst_time: %d",
+				bw_fixed_to_int(dceip->mcifwr_all_surfaces_burst_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dmif_request_buffer_size: %d",
+				bw_fixed_to_int(dceip->dmif_request_buffer_size));
+
+
+}
+
+static void print_bw_calcs_vbios(struct dal_logger *logger, const struct bw_calcs_vbios *vbios)
+{
+
+	DC_LOG_BANDWIDTH_CALCS("#####################################################################");
+	DC_LOG_BANDWIDTH_CALCS("struct bw_calcs_vbios vbios");
+	DC_LOG_BANDWIDTH_CALCS("#####################################################################");
+	DC_LOG_BANDWIDTH_CALCS("	[enum] bw_defines memory_type: %d", vbios->memory_type);
+	DC_LOG_BANDWIDTH_CALCS("	[enum] bw_defines memory_type: %d", vbios->memory_type);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] dram_channel_width_in_bits: %d", vbios->dram_channel_width_in_bits);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] number_of_dram_channels: %d", vbios->number_of_dram_channels);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] number_of_dram_banks: %d", vbios->number_of_dram_banks);
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] low_yclk: %d", bw_fixed_to_int(vbios->low_yclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mid_yclk: %d", bw_fixed_to_int(vbios->mid_yclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] high_yclk: %d", bw_fixed_to_int(vbios->high_yclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] low_sclk: %d", bw_fixed_to_int(vbios->low_sclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mid1_sclk: %d", bw_fixed_to_int(vbios->mid1_sclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mid2_sclk: %d", bw_fixed_to_int(vbios->mid2_sclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mid3_sclk: %d", bw_fixed_to_int(vbios->mid3_sclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mid4_sclk: %d", bw_fixed_to_int(vbios->mid4_sclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mid5_sclk: %d", bw_fixed_to_int(vbios->mid5_sclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mid6_sclk: %d", bw_fixed_to_int(vbios->mid6_sclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] high_sclk: %d", bw_fixed_to_int(vbios->high_sclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] low_voltage_max_dispclk: %d",
+				bw_fixed_to_int(vbios->low_voltage_max_dispclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mid_voltage_max_dispclk;: %d",
+				bw_fixed_to_int(vbios->mid_voltage_max_dispclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] high_voltage_max_dispclk;: %d",
+				bw_fixed_to_int(vbios->high_voltage_max_dispclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] low_voltage_max_phyclk: %d",
+				bw_fixed_to_int(vbios->low_voltage_max_phyclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mid_voltage_max_phyclk: %d",
+				bw_fixed_to_int(vbios->mid_voltage_max_phyclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] high_voltage_max_phyclk: %d",
+				bw_fixed_to_int(vbios->high_voltage_max_phyclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] data_return_bus_width: %d", bw_fixed_to_int(vbios->data_return_bus_width));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] trc: %d", bw_fixed_to_int(vbios->trc));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dmifmc_urgent_latency: %d", bw_fixed_to_int(vbios->dmifmc_urgent_latency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] stutter_self_refresh_exit_latency: %d",
+				bw_fixed_to_int(vbios->stutter_self_refresh_exit_latency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] stutter_self_refresh_entry_latency: %d",
+				bw_fixed_to_int(vbios->stutter_self_refresh_entry_latency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] nbp_state_change_latency: %d",
+				bw_fixed_to_int(vbios->nbp_state_change_latency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mcifwrmc_urgent_latency: %d",
+				bw_fixed_to_int(vbios->mcifwrmc_urgent_latency));
+	DC_LOG_BANDWIDTH_CALCS("	[bool] scatter_gather_enable: %d", vbios->scatter_gather_enable);
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] down_spread_percentage: %d",
+				bw_fixed_to_int(vbios->down_spread_percentage));
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] cursor_width: %d", vbios->cursor_width);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] average_compression_rate: %d", vbios->average_compression_rate);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] number_of_request_slots_gmc_reserves_for_dmif_per_channel: %d",
+				vbios->number_of_request_slots_gmc_reserves_for_dmif_per_channel);
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] blackout_duration: %d", bw_fixed_to_int(vbios->blackout_duration));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] maximum_blackout_recovery_time: %d",
+				bw_fixed_to_int(vbios->maximum_blackout_recovery_time));
+
+
+}
+
+static void print_bw_calcs_data(struct dal_logger *logger, struct bw_calcs_data *data)
+{
+
+	int i, j, k;
+
+	DC_LOG_BANDWIDTH_CALCS("#####################################################################");
+	DC_LOG_BANDWIDTH_CALCS("struct bw_calcs_data data");
+	DC_LOG_BANDWIDTH_CALCS("#####################################################################");
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] number_of_displays: %d", data->number_of_displays);
+	DC_LOG_BANDWIDTH_CALCS("	[enum] bw_defines underlay_surface_type: %d", data->underlay_surface_type);
+	DC_LOG_BANDWIDTH_CALCS("	[enum] bw_defines panning_and_bezel_adjustment: %d",
+				data->panning_and_bezel_adjustment);
+	DC_LOG_BANDWIDTH_CALCS("	[enum] bw_defines graphics_tiling_mode: %d", data->graphics_tiling_mode);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] graphics_lb_bpc: %d", data->graphics_lb_bpc);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] underlay_lb_bpc: %d", data->underlay_lb_bpc);
+	DC_LOG_BANDWIDTH_CALCS("	[enum] bw_defines underlay_tiling_mode: %d", data->underlay_tiling_mode);
+	DC_LOG_BANDWIDTH_CALCS("	[enum] bw_defines d0_underlay_mode: %d", data->d0_underlay_mode);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] d1_display_write_back_dwb_enable: %d", data->d1_display_write_back_dwb_enable);
+	DC_LOG_BANDWIDTH_CALCS("	[enum] bw_defines d1_underlay_mode: %d", data->d1_underlay_mode);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] cpup_state_change_enable: %d", data->cpup_state_change_enable);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] cpuc_state_change_enable: %d", data->cpuc_state_change_enable);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] nbp_state_change_enable: %d", data->nbp_state_change_enable);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] stutter_mode_enable: %d", data->stutter_mode_enable);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] y_clk_level: %d", data->y_clk_level);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] sclk_level: %d", data->sclk_level);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] number_of_underlay_surfaces: %d", data->number_of_underlay_surfaces);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] number_of_dram_wrchannels: %d", data->number_of_dram_wrchannels);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] chunk_request_delay: %d", data->chunk_request_delay);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] number_of_dram_channels: %d", data->number_of_dram_channels);
+	DC_LOG_BANDWIDTH_CALCS("	[enum] bw_defines underlay_micro_tile_mode: %d", data->underlay_micro_tile_mode);
+	DC_LOG_BANDWIDTH_CALCS("	[enum] bw_defines graphics_micro_tile_mode: %d", data->graphics_micro_tile_mode);
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] max_phyclk: %d", bw_fixed_to_int(data->max_phyclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dram_efficiency: %d", bw_fixed_to_int(data->dram_efficiency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] src_width_after_surface_type: %d",
+				bw_fixed_to_int(data->src_width_after_surface_type));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] src_height_after_surface_type: %d",
+				bw_fixed_to_int(data->src_height_after_surface_type));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] hsr_after_surface_type: %d",
+				bw_fixed_to_int(data->hsr_after_surface_type));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] vsr_after_surface_type: %d", bw_fixed_to_int(data->vsr_after_surface_type));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] src_width_after_rotation: %d",
+				bw_fixed_to_int(data->src_width_after_rotation));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] src_height_after_rotation: %d",
+				bw_fixed_to_int(data->src_height_after_rotation));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] hsr_after_rotation: %d", bw_fixed_to_int(data->hsr_after_rotation));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] vsr_after_rotation: %d", bw_fixed_to_int(data->vsr_after_rotation));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] source_height_pixels: %d", bw_fixed_to_int(data->source_height_pixels));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] hsr_after_stereo: %d", bw_fixed_to_int(data->hsr_after_stereo));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] vsr_after_stereo: %d", bw_fixed_to_int(data->vsr_after_stereo));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] source_width_in_lb: %d", bw_fixed_to_int(data->source_width_in_lb));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] lb_line_pitch: %d", bw_fixed_to_int(data->lb_line_pitch));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] underlay_maximum_source_efficient_for_tiling: %d",
+				bw_fixed_to_int(data->underlay_maximum_source_efficient_for_tiling));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] num_lines_at_frame_start: %d",
+				bw_fixed_to_int(data->num_lines_at_frame_start));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] min_dmif_size_in_time: %d", bw_fixed_to_int(data->min_dmif_size_in_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] min_mcifwr_size_in_time: %d",
+				bw_fixed_to_int(data->min_mcifwr_size_in_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_requests_for_dmif_size: %d",
+				bw_fixed_to_int(data->total_requests_for_dmif_size));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] peak_pte_request_to_eviction_ratio_limiting: %d",
+				bw_fixed_to_int(data->peak_pte_request_to_eviction_ratio_limiting));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] useful_pte_per_pte_request: %d",
+				bw_fixed_to_int(data->useful_pte_per_pte_request));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] scatter_gather_pte_request_rows: %d",
+				bw_fixed_to_int(data->scatter_gather_pte_request_rows));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] scatter_gather_row_height: %d",
+				bw_fixed_to_int(data->scatter_gather_row_height));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] scatter_gather_pte_requests_in_vblank: %d",
+				bw_fixed_to_int(data->scatter_gather_pte_requests_in_vblank));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] inefficient_linear_pitch_in_bytes: %d",
+				bw_fixed_to_int(data->inefficient_linear_pitch_in_bytes));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] cursor_total_data: %d", bw_fixed_to_int(data->cursor_total_data));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] cursor_total_request_groups: %d",
+				bw_fixed_to_int(data->cursor_total_request_groups));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] scatter_gather_total_pte_requests: %d",
+				bw_fixed_to_int(data->scatter_gather_total_pte_requests));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] scatter_gather_total_pte_request_groups: %d",
+				bw_fixed_to_int(data->scatter_gather_total_pte_request_groups));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] tile_width_in_pixels: %d", bw_fixed_to_int(data->tile_width_in_pixels));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dmif_total_number_of_data_request_page_close_open: %d",
+				bw_fixed_to_int(data->dmif_total_number_of_data_request_page_close_open));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mcifwr_total_number_of_data_request_page_close_open: %d",
+				bw_fixed_to_int(data->mcifwr_total_number_of_data_request_page_close_open));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] bytes_per_page_close_open: %d",
+				bw_fixed_to_int(data->bytes_per_page_close_open));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mcifwr_total_page_close_open_time: %d",
+				bw_fixed_to_int(data->mcifwr_total_page_close_open_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_requests_for_adjusted_dmif_size: %d",
+				bw_fixed_to_int(data->total_requests_for_adjusted_dmif_size));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_dmifmc_urgent_trips: %d",
+				bw_fixed_to_int(data->total_dmifmc_urgent_trips));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_dmifmc_urgent_latency: %d",
+				bw_fixed_to_int(data->total_dmifmc_urgent_latency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_display_reads_required_data: %d",
+				bw_fixed_to_int(data->total_display_reads_required_data));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_display_reads_required_dram_access_data: %d",
+				bw_fixed_to_int(data->total_display_reads_required_dram_access_data));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_display_writes_required_data: %d",
+				bw_fixed_to_int(data->total_display_writes_required_data));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_display_writes_required_dram_access_data: %d",
+				bw_fixed_to_int(data->total_display_writes_required_dram_access_data));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] display_reads_required_data: %d",
+				bw_fixed_to_int(data->display_reads_required_data));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] display_reads_required_dram_access_data: %d",
+				bw_fixed_to_int(data->display_reads_required_dram_access_data));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dmif_total_page_close_open_time: %d",
+				bw_fixed_to_int(data->dmif_total_page_close_open_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] min_cursor_memory_interface_buffer_size_in_time: %d",
+				bw_fixed_to_int(data->min_cursor_memory_interface_buffer_size_in_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] min_read_buffer_size_in_time: %d",
+				bw_fixed_to_int(data->min_read_buffer_size_in_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] display_reads_time_for_data_transfer: %d",
+				bw_fixed_to_int(data->display_reads_time_for_data_transfer));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] display_writes_time_for_data_transfer: %d",
+				bw_fixed_to_int(data->display_writes_time_for_data_transfer));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dmif_required_dram_bandwidth: %d",
+				bw_fixed_to_int(data->dmif_required_dram_bandwidth));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mcifwr_required_dram_bandwidth: %d",
+				bw_fixed_to_int(data->mcifwr_required_dram_bandwidth));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] required_dmifmc_urgent_latency_for_page_close_open: %d",
+				bw_fixed_to_int(data->required_dmifmc_urgent_latency_for_page_close_open));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] required_mcifmcwr_urgent_latency: %d",
+				bw_fixed_to_int(data->required_mcifmcwr_urgent_latency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] required_dram_bandwidth_gbyte_per_second: %d",
+				bw_fixed_to_int(data->required_dram_bandwidth_gbyte_per_second));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dram_bandwidth: %d", bw_fixed_to_int(data->dram_bandwidth));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dmif_required_sclk: %d", bw_fixed_to_int(data->dmif_required_sclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mcifwr_required_sclk: %d", bw_fixed_to_int(data->mcifwr_required_sclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] required_sclk: %d", bw_fixed_to_int(data->required_sclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] downspread_factor: %d", bw_fixed_to_int(data->downspread_factor));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] v_scaler_efficiency: %d", bw_fixed_to_int(data->v_scaler_efficiency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] scaler_limits_factor: %d", bw_fixed_to_int(data->scaler_limits_factor));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] display_pipe_pixel_throughput: %d",
+				bw_fixed_to_int(data->display_pipe_pixel_throughput));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_dispclk_required_with_ramping: %d",
+				bw_fixed_to_int(data->total_dispclk_required_with_ramping));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_dispclk_required_without_ramping: %d",
+				bw_fixed_to_int(data->total_dispclk_required_without_ramping));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_read_request_bandwidth: %d",
+				bw_fixed_to_int(data->total_read_request_bandwidth));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_write_request_bandwidth: %d",
+				bw_fixed_to_int(data->total_write_request_bandwidth));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dispclk_required_for_total_read_request_bandwidth: %d",
+				bw_fixed_to_int(data->dispclk_required_for_total_read_request_bandwidth));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_dispclk_required_with_ramping_with_request_bandwidth: %d",
+				bw_fixed_to_int(data->total_dispclk_required_with_ramping_with_request_bandwidth));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_dispclk_required_without_ramping_with_request_bandwidth: %d",
+				bw_fixed_to_int(data->total_dispclk_required_without_ramping_with_request_bandwidth));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dispclk: %d", bw_fixed_to_int(data->dispclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] blackout_recovery_time: %d", bw_fixed_to_int(data->blackout_recovery_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] min_pixels_per_data_fifo_entry: %d",
+				bw_fixed_to_int(data->min_pixels_per_data_fifo_entry));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] sclk_deep_sleep: %d", bw_fixed_to_int(data->sclk_deep_sleep));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] chunk_request_time: %d", bw_fixed_to_int(data->chunk_request_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] cursor_request_time: %d", bw_fixed_to_int(data->cursor_request_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] line_source_pixels_transfer_time: %d",
+				bw_fixed_to_int(data->line_source_pixels_transfer_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dmifdram_access_efficiency: %d",
+				bw_fixed_to_int(data->dmifdram_access_efficiency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mcifwrdram_access_efficiency: %d",
+				bw_fixed_to_int(data->mcifwrdram_access_efficiency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_average_bandwidth_no_compression: %d",
+				bw_fixed_to_int(data->total_average_bandwidth_no_compression));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_average_bandwidth: %d",
+				bw_fixed_to_int(data->total_average_bandwidth));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_stutter_cycle_duration: %d",
+				bw_fixed_to_int(data->total_stutter_cycle_duration));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] stutter_burst_time: %d", bw_fixed_to_int(data->stutter_burst_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] time_in_self_refresh: %d", bw_fixed_to_int(data->time_in_self_refresh));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] stutter_efficiency: %d", bw_fixed_to_int(data->stutter_efficiency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] worst_number_of_trips_to_memory: %d",
+				bw_fixed_to_int(data->worst_number_of_trips_to_memory));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] immediate_flip_time: %d", bw_fixed_to_int(data->immediate_flip_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] latency_for_non_dmif_clients: %d",
+				bw_fixed_to_int(data->latency_for_non_dmif_clients));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] latency_for_non_mcifwr_clients: %d",
+				bw_fixed_to_int(data->latency_for_non_mcifwr_clients));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dmifmc_urgent_latency_supported_in_high_sclk_and_yclk: %d",
+				bw_fixed_to_int(data->dmifmc_urgent_latency_supported_in_high_sclk_and_yclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] nbp_state_dram_speed_change_margin: %d",
+				bw_fixed_to_int(data->nbp_state_dram_speed_change_margin));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] display_reads_time_for_data_transfer_and_urgent_latency: %d",
+				bw_fixed_to_int(data->display_reads_time_for_data_transfer_and_urgent_latency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dram_speed_change_margin: %d",
+				bw_fixed_to_int(data->dram_speed_change_margin));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] min_vblank_dram_speed_change_margin: %d",
+				bw_fixed_to_int(data->min_vblank_dram_speed_change_margin));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] min_stutter_refresh_duration: %d",
+				bw_fixed_to_int(data->min_stutter_refresh_duration));
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] total_stutter_dmif_buffer_size: %d", data->total_stutter_dmif_buffer_size);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] total_bytes_requested: %d", data->total_bytes_requested);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] min_stutter_dmif_buffer_size: %d", data->min_stutter_dmif_buffer_size);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] num_stutter_bursts: %d", data->num_stutter_bursts);
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] v_blank_nbp_state_dram_speed_change_latency_supported: %d",
+				bw_fixed_to_int(data->v_blank_nbp_state_dram_speed_change_latency_supported));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] nbp_state_dram_speed_change_latency_supported: %d",
+				bw_fixed_to_int(data->nbp_state_dram_speed_change_latency_supported));
+
+	for (i = 0; i < maximum_number_of_surfaces; i++) {
+		DC_LOG_BANDWIDTH_CALCS("	[bool] fbc_en[%d]:%d\n", i, data->fbc_en[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[bool] lpt_en[%d]:%d", i, data->lpt_en[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[bool] displays_match_flag[%d]:%d", i, data->displays_match_flag[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[bool] use_alpha[%d]:%d", i, data->use_alpha[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[bool] orthogonal_rotation[%d]:%d", i, data->orthogonal_rotation[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[bool] enable[%d]:%d", i, data->enable[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[bool] access_one_channel_only[%d]:%d", i, data->access_one_channel_only[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[bool] scatter_gather_enable_for_pipe[%d]:%d",
+					i, data->scatter_gather_enable_for_pipe[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[bool] interlace_mode[%d]:%d",
+					i, data->interlace_mode[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[bool] display_pstate_change_enable[%d]:%d",
+					i, data->display_pstate_change_enable[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[bool] line_buffer_prefetch[%d]:%d", i, data->line_buffer_prefetch[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[uint32_t] bytes_per_pixel[%d]:%d", i, data->bytes_per_pixel[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[uint32_t] max_chunks_non_fbc_mode[%d]:%d",
+					i, data->max_chunks_non_fbc_mode[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[uint32_t] lb_bpc[%d]:%d", i, data->lb_bpc[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[uint32_t] output_bpphdmi[%d]:%d", i, data->output_bpphdmi[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[uint32_t] output_bppdp4_lane_hbr[%d]:%d", i, data->output_bppdp4_lane_hbr[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[uint32_t] output_bppdp4_lane_hbr2[%d]:%d",
+					i, data->output_bppdp4_lane_hbr2[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[uint32_t] output_bppdp4_lane_hbr3[%d]:%d",
+					i, data->output_bppdp4_lane_hbr3[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[enum] bw_defines stereo_mode[%d]:%d", i, data->stereo_mode[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dmif_buffer_transfer_time[%d]:%d",
+					i, bw_fixed_to_int(data->dmif_buffer_transfer_time[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] displays_with_same_mode[%d]:%d",
+					i, bw_fixed_to_int(data->displays_with_same_mode[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] stutter_dmif_buffer_size[%d]:%d",
+					i, bw_fixed_to_int(data->stutter_dmif_buffer_size[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] stutter_refresh_duration[%d]:%d",
+					i, bw_fixed_to_int(data->stutter_refresh_duration[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] stutter_exit_watermark[%d]:%d",
+					i, bw_fixed_to_int(data->stutter_exit_watermark[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] stutter_entry_watermark[%d]:%d",
+					i, bw_fixed_to_int(data->stutter_entry_watermark[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] h_total[%d]:%d", i, bw_fixed_to_int(data->h_total[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] v_total[%d]:%d", i, bw_fixed_to_int(data->v_total[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] pixel_rate[%d]:%d", i, bw_fixed_to_int(data->pixel_rate[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] src_width[%d]:%d", i, bw_fixed_to_int(data->src_width[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] pitch_in_pixels[%d]:%d",
+					i, bw_fixed_to_int(data->pitch_in_pixels[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] pitch_in_pixels_after_surface_type[%d]:%d",
+					i, bw_fixed_to_int(data->pitch_in_pixels_after_surface_type[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] src_height[%d]:%d", i, bw_fixed_to_int(data->src_height[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] scale_ratio[%d]:%d", i, bw_fixed_to_int(data->scale_ratio[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] h_taps[%d]:%d", i, bw_fixed_to_int(data->h_taps[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] v_taps[%d]:%d", i, bw_fixed_to_int(data->v_taps[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] h_scale_ratio[%d]:%d", i, bw_fixed_to_int(data->h_scale_ratio[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] v_scale_ratio[%d]:%d", i, bw_fixed_to_int(data->v_scale_ratio[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] rotation_angle[%d]:%d",
+					i, bw_fixed_to_int(data->rotation_angle[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] compression_rate[%d]:%d",
+					i, bw_fixed_to_int(data->compression_rate[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] hsr[%d]:%d", i, bw_fixed_to_int(data->hsr[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] vsr[%d]:%d", i, bw_fixed_to_int(data->vsr[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] source_width_rounded_up_to_chunks[%d]:%d",
+					i, bw_fixed_to_int(data->source_width_rounded_up_to_chunks[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] source_width_pixels[%d]:%d",
+					i, bw_fixed_to_int(data->source_width_pixels[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] source_height_rounded_up_to_chunks[%d]:%d",
+					i, bw_fixed_to_int(data->source_height_rounded_up_to_chunks[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] display_bandwidth[%d]:%d",
+					i, bw_fixed_to_int(data->display_bandwidth[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] request_bandwidth[%d]:%d",
+					i, bw_fixed_to_int(data->request_bandwidth[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] bytes_per_request[%d]:%d",
+					i, bw_fixed_to_int(data->bytes_per_request[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] useful_bytes_per_request[%d]:%d",
+					i, bw_fixed_to_int(data->useful_bytes_per_request[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] lines_interleaved_in_mem_access[%d]:%d",
+					i, bw_fixed_to_int(data->lines_interleaved_in_mem_access[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] latency_hiding_lines[%d]:%d",
+					i, bw_fixed_to_int(data->latency_hiding_lines[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] lb_partitions[%d]:%d",
+					i, bw_fixed_to_int(data->lb_partitions[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] lb_partitions_max[%d]:%d",
+					i, bw_fixed_to_int(data->lb_partitions_max[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dispclk_required_with_ramping[%d]:%d",
+					i, bw_fixed_to_int(data->dispclk_required_with_ramping[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dispclk_required_without_ramping[%d]:%d",
+					i, bw_fixed_to_int(data->dispclk_required_without_ramping[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] data_buffer_size[%d]:%d",
+					i, bw_fixed_to_int(data->data_buffer_size[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] outstanding_chunk_request_limit[%d]:%d",
+					i, bw_fixed_to_int(data->outstanding_chunk_request_limit[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] urgent_watermark[%d]:%d",
+					i, bw_fixed_to_int(data->urgent_watermark[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] nbp_state_change_watermark[%d]:%d",
+					i, bw_fixed_to_int(data->nbp_state_change_watermark[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] v_filter_init[%d]:%d", i, bw_fixed_to_int(data->v_filter_init[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] stutter_cycle_duration[%d]:%d",
+					i, bw_fixed_to_int(data->stutter_cycle_duration[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] average_bandwidth[%d]:%d",
+					i, bw_fixed_to_int(data->average_bandwidth[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] average_bandwidth_no_compression[%d]:%d",
+					i, bw_fixed_to_int(data->average_bandwidth_no_compression[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] scatter_gather_pte_request_limit[%d]:%d",
+					i, bw_fixed_to_int(data->scatter_gather_pte_request_limit[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] lb_size_per_component[%d]:%d",
+					i, bw_fixed_to_int(data->lb_size_per_component[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] memory_chunk_size_in_bytes[%d]:%d",
+					i, bw_fixed_to_int(data->memory_chunk_size_in_bytes[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] pipe_chunk_size_in_bytes[%d]:%d",
+					i, bw_fixed_to_int(data->pipe_chunk_size_in_bytes[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] number_of_trips_to_memory_for_getting_apte_row[%d]:%d",
+					i, bw_fixed_to_int(data->number_of_trips_to_memory_for_getting_apte_row[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] adjusted_data_buffer_size[%d]:%d",
+					i, bw_fixed_to_int(data->adjusted_data_buffer_size[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] adjusted_data_buffer_size_in_memory[%d]:%d",
+					i, bw_fixed_to_int(data->adjusted_data_buffer_size_in_memory[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] pixels_per_data_fifo_entry[%d]:%d",
+					i, bw_fixed_to_int(data->pixels_per_data_fifo_entry[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] scatter_gather_pte_requests_in_row[%d]:%d",
+					i, bw_fixed_to_int(data->scatter_gather_pte_requests_in_row[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] pte_request_per_chunk[%d]:%d",
+					i, bw_fixed_to_int(data->pte_request_per_chunk[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] scatter_gather_page_width[%d]:%d",
+					i, bw_fixed_to_int(data->scatter_gather_page_width[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] scatter_gather_page_height[%d]:%d",
+					i, bw_fixed_to_int(data->scatter_gather_page_height[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] lb_lines_in_per_line_out_in_beginning_of_frame[%d]:%d",
+					i, bw_fixed_to_int(data->lb_lines_in_per_line_out_in_beginning_of_frame[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] lb_lines_in_per_line_out_in_middle_of_frame[%d]:%d",
+					i, bw_fixed_to_int(data->lb_lines_in_per_line_out_in_middle_of_frame[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] cursor_width_pixels[%d]:%d",
+					i, bw_fixed_to_int(data->cursor_width_pixels[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] minimum_latency_hiding[%d]:%d",
+					i, bw_fixed_to_int(data->minimum_latency_hiding[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] maximum_latency_hiding[%d]:%d",
+					i, bw_fixed_to_int(data->maximum_latency_hiding[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] minimum_latency_hiding_with_cursor[%d]:%d",
+					i, bw_fixed_to_int(data->minimum_latency_hiding_with_cursor[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] maximum_latency_hiding_with_cursor[%d]:%d",
+					i, bw_fixed_to_int(data->maximum_latency_hiding_with_cursor[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] src_pixels_for_first_output_pixel[%d]:%d",
+					i, bw_fixed_to_int(data->src_pixels_for_first_output_pixel[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] src_pixels_for_last_output_pixel[%d]:%d",
+					i, bw_fixed_to_int(data->src_pixels_for_last_output_pixel[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] src_data_for_first_output_pixel[%d]:%d",
+					i, bw_fixed_to_int(data->src_data_for_first_output_pixel[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] src_data_for_last_output_pixel[%d]:%d",
+					i, bw_fixed_to_int(data->src_data_for_last_output_pixel[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] active_time[%d]:%d", i, bw_fixed_to_int(data->active_time[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] horizontal_blank_and_chunk_granularity_factor[%d]:%d",
+					i, bw_fixed_to_int(data->horizontal_blank_and_chunk_granularity_factor[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] cursor_latency_hiding[%d]:%d",
+					i, bw_fixed_to_int(data->cursor_latency_hiding[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] v_blank_dram_speed_change_margin[%d]:%d",
+					i, bw_fixed_to_int(data->v_blank_dram_speed_change_margin[i]));
+		}
+
+	for (i = 0; i < maximum_number_of_surfaces; i++) {
+		for (j = 0; j < 3; j++) {
+			for (k = 0; k < 8; k++) {
+
+				DC_LOG_BANDWIDTH_CALCS("\n	[bw_fixed] line_source_transfer_time[%d][%d][%d]:%d",
+					i, j, k, bw_fixed_to_int(data->line_source_transfer_time[i][j][k]));
+				DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dram_speed_change_line_source_transfer_time[%d][%d][%d]:%d",
+					i, j, k,
+					bw_fixed_to_int(data->dram_speed_change_line_source_transfer_time[i][j][k]));
+			}
+		}
+	}
+
+	for (i = 0; i < 3; i++) {
+		for (j = 0; j < 8; j++) {
+
+			DC_LOG_BANDWIDTH_CALCS("\n	[uint32_t] num_displays_with_margin[%d][%d]:%d",
+					i, j, data->num_displays_with_margin[i][j]);
+			DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dmif_burst_time[%d][%d]:%d",
+					i, j, bw_fixed_to_int(data->dmif_burst_time[i][j]));
+			DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mcifwr_burst_time[%d][%d]:%d",
+					i, j, bw_fixed_to_int(data->mcifwr_burst_time[i][j]));
+			DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] min_dram_speed_change_margin[%d][%d]:%d",
+					i, j, bw_fixed_to_int(data->min_dram_speed_change_margin[i][j]));
+			DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dispclk_required_for_dram_speed_change[%d][%d]:%d",
+					i, j, bw_fixed_to_int(data->dispclk_required_for_dram_speed_change[i][j]));
+			DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] blackout_duration_margin[%d][%d]:%d",
+					i, j, bw_fixed_to_int(data->blackout_duration_margin[i][j]));
+			DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dispclk_required_for_blackout_duration[%d][%d]:%d",
+					i, j, bw_fixed_to_int(data->dispclk_required_for_blackout_duration[i][j]));
+			DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dispclk_required_for_blackout_recovery[%d][%d]:%d",
+					i, j, bw_fixed_to_int(data->dispclk_required_for_blackout_recovery[i][j]));
+		}
+	}
+
+	for (i = 0; i < 6; i++) {
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dmif_required_sclk_for_urgent_latency[%d]:%d",
+					i, bw_fixed_to_int(data->dmif_required_sclk_for_urgent_latency[i]));
+	}
+}
+;
+
+#endif /* _CALCS_CALCS_LOGGER_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c b/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c
index 821502b..59acb08 100644
--- a/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c
+++ b/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c
@@ -28,6 +28,7 @@
 #include "dc.h"
 #include "core_types.h"
 #include "dal_asic_id.h"
+#include "calcs_logger.h"
 
 /*
  * NOTE:
@@ -2990,6 +2991,11 @@ bool bw_calcs(struct dc_context *ctx,
 		struct bw_fixed mid_yclk = vbios->mid_yclk;
 		struct bw_fixed low_yclk = vbios->low_yclk;
 
+		if (ctx->dc->debug.bandwidth_calcs_trace) {
+			print_bw_calcs_dceip(ctx->logger, dceip);
+			print_bw_calcs_vbios(ctx->logger, vbios);
+			print_bw_calcs_data(ctx->logger, data);
+		}
 		calculate_bandwidth(dceip, vbios, data);
 
 		yclk_lvl = data->y_clk_level;
diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index 4d9da9d..bdc3cef 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -202,6 +202,7 @@ struct dc_debug {
 	bool timing_trace;
 	bool clock_trace;
 	bool validation_trace;
+	bool bandwidth_calcs_trace;
 
 	/* stutter efficiency related */
 	bool disable_stutter;
-- 
cgit v1.1


From c1f8d3fa4627ec7ec31cc538b471488980738631 Mon Sep 17 00:00:00 2001
From: Mikita Lipski <mikita.lipski@amd.com>
Date: Wed, 7 Mar 2018 16:59:43 -0500
Subject: drm/amd/display: Don't call amdgpu_dm_display_resume as it doesn't
 exist

amdgpu_dm_display_resume was merged into dm_resume.
No need to call these functions separately.

Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index e42a28e..bad9f09 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -1629,7 +1629,6 @@ static ssize_t s3_debug_store(struct device *device,
 	if (ret == 0) {
 		if (s3_state) {
 			dm_resume(adev);
-			amdgpu_dm_display_resume(adev);
 			drm_kms_helper_hotplug_event(adev->ddev);
 		} else
 			dm_suspend(adev);
-- 
cgit v1.1


From b361521f59764139067ada4ea9d6c213d583678f Mon Sep 17 00:00:00 2001
From: Mikita Lipski <mikita.lipski@amd.com>
Date: Mon, 12 Mar 2018 15:53:47 -0400
Subject: drm/amd/display: Adding stutter entry wm to dce bw struct

Adding the stutter_entry_wm object to dce_bw_output struct
and populating it with bw calculations data

Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c | 111 ++++++++++++++++++++++-
 drivers/gpu/drm/amd/display/dc/inc/core_types.h  |   1 +
 2 files changed, 110 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c b/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c
index 59acb08..4b71932 100644
--- a/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c
+++ b/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c
@@ -3091,7 +3091,33 @@ bool bw_calcs(struct dc_context *ctx,
 			bw_fixed_to_int(bw_mul(data->
 				stutter_exit_watermark[9], bw_int_to_fixed(1000)));
 
-
+		calcs_output->stutter_entry_wm_ns[0].a_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[4], bw_int_to_fixed(1000)));
+		calcs_output->stutter_entry_wm_ns[1].a_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[5], bw_int_to_fixed(1000)));
+		calcs_output->stutter_entry_wm_ns[2].a_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[6], bw_int_to_fixed(1000)));
+		if (ctx->dc->caps.max_slave_planes) {
+			calcs_output->stutter_entry_wm_ns[3].a_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[0], bw_int_to_fixed(1000)));
+			calcs_output->stutter_entry_wm_ns[4].a_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[1], bw_int_to_fixed(1000)));
+		} else {
+			calcs_output->stutter_entry_wm_ns[3].a_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[7], bw_int_to_fixed(1000)));
+			calcs_output->stutter_entry_wm_ns[4].a_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[8], bw_int_to_fixed(1000)));
+		}
+		calcs_output->stutter_entry_wm_ns[5].a_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[9], bw_int_to_fixed(1000)));
 
 		calcs_output->urgent_wm_ns[0].a_mark =
 			bw_fixed_to_int(bw_mul(data->
@@ -3186,7 +3212,33 @@ bool bw_calcs(struct dc_context *ctx,
 				bw_fixed_to_int(bw_mul(data->
 					stutter_exit_watermark[9], bw_int_to_fixed(1000)));
 
-
+		calcs_output->stutter_entry_wm_ns[0].b_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[4], bw_int_to_fixed(1000)));
+		calcs_output->stutter_entry_wm_ns[1].b_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[5], bw_int_to_fixed(1000)));
+		calcs_output->stutter_entry_wm_ns[2].b_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[6], bw_int_to_fixed(1000)));
+		if (ctx->dc->caps.max_slave_planes) {
+			calcs_output->stutter_entry_wm_ns[3].b_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[0], bw_int_to_fixed(1000)));
+			calcs_output->stutter_entry_wm_ns[4].b_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[1], bw_int_to_fixed(1000)));
+		} else {
+			calcs_output->stutter_entry_wm_ns[3].b_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[7], bw_int_to_fixed(1000)));
+			calcs_output->stutter_entry_wm_ns[4].b_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[8], bw_int_to_fixed(1000)));
+		}
+		calcs_output->stutter_entry_wm_ns[5].b_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[9], bw_int_to_fixed(1000)));
 
 			calcs_output->urgent_wm_ns[0].b_mark =
 				bw_fixed_to_int(bw_mul(data->
@@ -3279,6 +3331,34 @@ bool bw_calcs(struct dc_context *ctx,
 				bw_fixed_to_int(bw_mul(data->
 					stutter_exit_watermark[9], bw_int_to_fixed(1000)));
 
+		calcs_output->stutter_entry_wm_ns[0].c_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[4], bw_int_to_fixed(1000)));
+		calcs_output->stutter_entry_wm_ns[1].c_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[5], bw_int_to_fixed(1000)));
+		calcs_output->stutter_entry_wm_ns[2].c_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[6], bw_int_to_fixed(1000)));
+		if (ctx->dc->caps.max_slave_planes) {
+			calcs_output->stutter_entry_wm_ns[3].c_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[0], bw_int_to_fixed(1000)));
+			calcs_output->stutter_entry_wm_ns[4].c_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[1], bw_int_to_fixed(1000)));
+		} else {
+			calcs_output->stutter_entry_wm_ns[3].c_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[7], bw_int_to_fixed(1000)));
+			calcs_output->stutter_entry_wm_ns[4].c_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[8], bw_int_to_fixed(1000)));
+		}
+		calcs_output->stutter_entry_wm_ns[5].c_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[9], bw_int_to_fixed(1000)));
+
 			calcs_output->urgent_wm_ns[0].c_mark =
 				bw_fixed_to_int(bw_mul(data->
 					urgent_watermark[4], bw_int_to_fixed(1000)));
@@ -3383,6 +3463,33 @@ bool bw_calcs(struct dc_context *ctx,
 			bw_fixed_to_int(bw_mul(data->
 				stutter_exit_watermark[9], bw_int_to_fixed(1000)));
 
+		calcs_output->stutter_entry_wm_ns[0].d_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[4], bw_int_to_fixed(1000)));
+		calcs_output->stutter_entry_wm_ns[1].d_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[5], bw_int_to_fixed(1000)));
+		calcs_output->stutter_entry_wm_ns[2].d_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[6], bw_int_to_fixed(1000)));
+		if (ctx->dc->caps.max_slave_planes) {
+			calcs_output->stutter_entry_wm_ns[3].d_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[0], bw_int_to_fixed(1000)));
+			calcs_output->stutter_entry_wm_ns[4].d_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[1], bw_int_to_fixed(1000)));
+		} else {
+			calcs_output->stutter_entry_wm_ns[3].d_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[7], bw_int_to_fixed(1000)));
+			calcs_output->stutter_entry_wm_ns[4].d_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[8], bw_int_to_fixed(1000)));
+		}
+		calcs_output->stutter_entry_wm_ns[5].d_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[9], bw_int_to_fixed(1000)));
 
 		calcs_output->urgent_wm_ns[0].d_mark =
 			bw_fixed_to_int(bw_mul(data->
diff --git a/drivers/gpu/drm/amd/display/dc/inc/core_types.h b/drivers/gpu/drm/amd/display/dc/inc/core_types.h
index 8c51ad7..55f56bf 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/core_types.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/core_types.h
@@ -250,6 +250,7 @@ struct dce_bw_output {
 	bool all_displays_in_sync;
 	struct dce_watermarks urgent_wm_ns[MAX_PIPES];
 	struct dce_watermarks stutter_exit_wm_ns[MAX_PIPES];
+	struct dce_watermarks stutter_entry_wm_ns[MAX_PIPES];
 	struct dce_watermarks nbp_state_change_wm_ns[MAX_PIPES];
 	int sclk_khz;
 	int sclk_deep_sleep_khz;
-- 
cgit v1.1


From f8931ea730ffa6c84e98c970c173935cfd38c0aa Mon Sep 17 00:00:00 2001
From: Eric Bernstein <eric.bernstein@amd.com>
Date: Mon, 12 Mar 2018 17:07:24 -0400
Subject: drm/amd/display: Change wb_h/vratio to double

Signed-off-by: Eric Bernstein <eric.bernstein@amd.com>
Reviewed-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h
index 09affa1..e296de6 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h
+++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h
@@ -215,8 +215,8 @@ struct writeback_st {
 	int wb_vtaps_luma;
 	int wb_htaps_chroma;
 	int wb_vtaps_chroma;
-	int wb_hratio;
-	int wb_vratio;
+	double wb_hratio;
+	double wb_vratio;
 };
 
 struct	_vcs_dpi_display_output_params_st	{
-- 
cgit v1.1


From 6133470c8e2ffdc6a5d67a1d79a9a0c1c0a94a10 Mon Sep 17 00:00:00 2001
From: Julian Parkin <jparkin@amd.com>
Date: Tue, 13 Mar 2018 15:53:13 -0400
Subject: drm/amd/display: drop dc_validate_guaranteed

Block FP16 scaling in validate_resources codepath.

Signed-off-by: Julian Parkin <jparkin@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_resource.c  | 32 --------------
 drivers/gpu/drm/amd/display/dc/dc_stream.h         |  8 ----
 .../drm/amd/display/dc/dce100/dce100_resource.c    | 33 ---------------
 .../drm/amd/display/dc/dce110/dce110_resource.c    | 33 ---------------
 .../drm/amd/display/dc/dce112/dce112_resource.c    | 33 ---------------
 .../drm/amd/display/dc/dce112/dce112_resource.h    |  5 ---
 .../drm/amd/display/dc/dce120/dce120_resource.c    |  1 -
 .../gpu/drm/amd/display/dc/dce80/dce80_resource.c  | 49 ----------------------
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c   |  7 ++++
 .../gpu/drm/amd/display/dc/dcn10/dcn10_resource.c  | 31 --------------
 drivers/gpu/drm/amd/display/dc/inc/core_types.h    |  5 ---
 drivers/gpu/drm/amd/display/dc/inc/resource.h      |  4 --
 12 files changed, 7 insertions(+), 234 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
index ba3487e..cae78ee 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
@@ -1599,18 +1599,6 @@ enum dc_status dc_remove_stream_from_ctx(
 	return DC_OK;
 }
 
-static void copy_pipe_ctx(
-	const struct pipe_ctx *from_pipe_ctx, struct pipe_ctx *to_pipe_ctx)
-{
-	struct dc_plane_state *plane_state = to_pipe_ctx->plane_state;
-	struct dc_stream_state *stream = to_pipe_ctx->stream;
-
-	*to_pipe_ctx = *from_pipe_ctx;
-	to_pipe_ctx->stream = stream;
-	if (plane_state != NULL)
-		to_pipe_ctx->plane_state = plane_state;
-}
-
 static struct dc_stream_state *find_pll_sharable_stream(
 		struct dc_stream_state *stream_needs_pll,
 		struct dc_state *context)
@@ -1752,26 +1740,6 @@ enum dc_status resource_map_pool_resources(
 	return DC_ERROR_UNEXPECTED;
 }
 
-/* first stream in the context is used to populate the rest */
-void validate_guaranteed_copy_streams(
-		struct dc_state *context,
-		int max_streams)
-{
-	int i;
-
-	for (i = 1; i < max_streams; i++) {
-		context->streams[i] = context->streams[0];
-
-		copy_pipe_ctx(&context->res_ctx.pipe_ctx[0],
-			      &context->res_ctx.pipe_ctx[i]);
-		context->res_ctx.pipe_ctx[i].stream =
-				context->res_ctx.pipe_ctx[0].stream;
-
-		dc_stream_retain(context->streams[i]);
-		context->stream_count++;
-	}
-}
-
 void dc_resource_state_copy_construct_current(
 		const struct dc *dc,
 		struct dc_state *dst_ctx)
diff --git a/drivers/gpu/drm/amd/display/dc/dc_stream.h b/drivers/gpu/drm/amd/display/dc/dc_stream.h
index d017df5..3a7093e 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_stream.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_stream.h
@@ -209,14 +209,6 @@ bool dc_add_all_planes_for_stream(
 enum dc_status dc_validate_stream(struct dc *dc, struct dc_stream_state *stream);
 
 /*
- * This function takes a stream and checks if it is guaranteed to be supported.
- * Guaranteed means that MAX_COFUNC similar streams are supported.
- *
- * After this call:
- *   No hardware is programmed for call.  Only validation is done.
- */
-
-/*
  * Set up streams and links associated to drive sinks
  * The streams parameter is an absolute set of all active streams.
  *
diff --git a/drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c b/drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c
index 3092f76..38ec0d6 100644
--- a/drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c
@@ -733,38 +733,6 @@ enum dc_status dce100_add_stream_to_ctx(
 	return result;
 }
 
-enum dc_status dce100_validate_guaranteed(
-		struct dc  *dc,
-		struct dc_stream_state *dc_stream,
-		struct dc_state *context)
-{
-	enum dc_status result = DC_ERROR_UNEXPECTED;
-
-	context->streams[0] = dc_stream;
-	dc_stream_retain(context->streams[0]);
-	context->stream_count++;
-
-	result = resource_map_pool_resources(dc, context, dc_stream);
-
-	if (result == DC_OK)
-		result = resource_map_clock_resources(dc, context, dc_stream);
-
-	if (result == DC_OK)
-		result = build_mapped_resource(dc, context, dc_stream);
-
-	if (result == DC_OK) {
-		validate_guaranteed_copy_streams(
-				context, dc->caps.max_streams);
-		result = resource_build_scaling_params_for_context(dc, context);
-	}
-
-	if (result == DC_OK)
-		if (!dce100_validate_bandwidth(dc, context))
-			result = DC_FAIL_BANDWIDTH_VALIDATE;
-
-	return result;
-}
-
 static void dce100_destroy_resource_pool(struct resource_pool **pool)
 {
 	struct dce110_resource_pool *dce110_pool = TO_DCE110_RES_POOL(*pool);
@@ -786,7 +754,6 @@ enum dc_status dce100_validate_plane(const struct dc_plane_state *plane_state, s
 static const struct resource_funcs dce100_res_pool_funcs = {
 	.destroy = dce100_destroy_resource_pool,
 	.link_enc_create = dce100_link_encoder_create,
-	.validate_guaranteed = dce100_validate_guaranteed,
 	.validate_bandwidth = dce100_validate_bandwidth,
 	.validate_plane = dce100_validate_plane,
 	.add_stream_to_ctx = dce100_add_stream_to_ctx,
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c
index b1f14be..ee33786 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c
@@ -930,38 +930,6 @@ static enum dc_status dce110_add_stream_to_ctx(
 	return result;
 }
 
-static enum dc_status dce110_validate_guaranteed(
-		struct dc *dc,
-		struct dc_stream_state *dc_stream,
-		struct dc_state *context)
-{
-	enum dc_status result = DC_ERROR_UNEXPECTED;
-
-	context->streams[0] = dc_stream;
-	dc_stream_retain(context->streams[0]);
-	context->stream_count++;
-
-	result = resource_map_pool_resources(dc, context, dc_stream);
-
-	if (result == DC_OK)
-		result = resource_map_clock_resources(dc, context, dc_stream);
-
-	if (result == DC_OK)
-		result = build_mapped_resource(dc, context, dc_stream);
-
-	if (result == DC_OK) {
-		validate_guaranteed_copy_streams(
-				context, dc->caps.max_streams);
-		result = resource_build_scaling_params_for_context(dc, context);
-	}
-
-	if (result == DC_OK)
-		if (!dce110_validate_bandwidth(dc, context))
-			result = DC_FAIL_BANDWIDTH_VALIDATE;
-
-	return result;
-}
-
 static struct pipe_ctx *dce110_acquire_underlay(
 		struct dc_state *context,
 		const struct resource_pool *pool,
@@ -1036,7 +1004,6 @@ static void dce110_destroy_resource_pool(struct resource_pool **pool)
 static const struct resource_funcs dce110_res_pool_funcs = {
 	.destroy = dce110_destroy_resource_pool,
 	.link_enc_create = dce110_link_encoder_create,
-	.validate_guaranteed = dce110_validate_guaranteed,
 	.validate_bandwidth = dce110_validate_bandwidth,
 	.validate_plane = dce110_validate_plane,
 	.acquire_idle_pipe_for_layer = dce110_acquire_underlay,
diff --git a/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c b/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c
index cd1e3f7..0a47663 100644
--- a/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c
@@ -867,38 +867,6 @@ enum dc_status dce112_add_stream_to_ctx(
 	return result;
 }
 
-enum dc_status dce112_validate_guaranteed(
-		struct dc *dc,
-		struct dc_stream_state *stream,
-		struct dc_state *context)
-{
-	enum dc_status result = DC_ERROR_UNEXPECTED;
-
-	context->streams[0] = stream;
-	dc_stream_retain(context->streams[0]);
-	context->stream_count++;
-
-	result = resource_map_pool_resources(dc, context, stream);
-
-	if (result == DC_OK)
-		result = resource_map_phy_clock_resources(dc, context, stream);
-
-	if (result == DC_OK)
-		result = build_mapped_resource(dc, context, stream);
-
-	if (result == DC_OK) {
-		validate_guaranteed_copy_streams(
-				context, dc->caps.max_streams);
-		result = resource_build_scaling_params_for_context(dc, context);
-	}
-
-	if (result == DC_OK)
-		if (!dce112_validate_bandwidth(dc, context))
-			result = DC_FAIL_BANDWIDTH_VALIDATE;
-
-	return result;
-}
-
 enum dc_status dce112_validate_global(
 		struct dc *dc,
 		struct dc_state *context)
@@ -921,7 +889,6 @@ static void dce112_destroy_resource_pool(struct resource_pool **pool)
 static const struct resource_funcs dce112_res_pool_funcs = {
 	.destroy = dce112_destroy_resource_pool,
 	.link_enc_create = dce112_link_encoder_create,
-	.validate_guaranteed = dce112_validate_guaranteed,
 	.validate_bandwidth = dce112_validate_bandwidth,
 	.validate_plane = dce100_validate_plane,
 	.add_stream_to_ctx = dce112_add_stream_to_ctx,
diff --git a/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.h b/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.h
index d5c19d3..95a4033 100644
--- a/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.h
+++ b/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.h
@@ -42,11 +42,6 @@ enum dc_status dce112_validate_with_context(
 		struct dc_state *context,
 		struct dc_state *old_context);
 
-enum dc_status dce112_validate_guaranteed(
-		struct dc *dc,
-		struct dc_stream_state *dc_stream,
-		struct dc_state *context);
-
 bool dce112_validate_bandwidth(
 	struct dc *dc,
 	struct dc_state *context);
diff --git a/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c b/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c
index 4659a4b..567e6b4 100644
--- a/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c
@@ -684,7 +684,6 @@ static void dce120_destroy_resource_pool(struct resource_pool **pool)
 static const struct resource_funcs dce120_res_pool_funcs = {
 	.destroy = dce120_destroy_resource_pool,
 	.link_enc_create = dce120_link_encoder_create,
-	.validate_guaranteed = dce112_validate_guaranteed,
 	.validate_bandwidth = dce112_validate_bandwidth,
 	.validate_plane = dce100_validate_plane,
 	.add_stream_to_ctx = dce112_add_stream_to_ctx
diff --git a/drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c b/drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c
index 5d854a3..48a0689 100644
--- a/drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c
@@ -691,23 +691,6 @@ static void destruct(struct dce110_resource_pool *pool)
 	}
 }
 
-static enum dc_status build_mapped_resource(
-		const struct dc *dc,
-		struct dc_state *context,
-		struct dc_stream_state *stream)
-{
-	struct pipe_ctx *pipe_ctx = resource_get_head_pipe_for_stream(&context->res_ctx, stream);
-
-	if (!pipe_ctx)
-		return DC_ERROR_UNEXPECTED;
-
-	dce110_resource_build_pipe_hw_param(pipe_ctx);
-
-	resource_build_info_frame(pipe_ctx);
-
-	return DC_OK;
-}
-
 bool dce80_validate_bandwidth(
 	struct dc *dc,
 	struct dc_state *context)
@@ -749,37 +732,6 @@ enum dc_status dce80_validate_global(
 	return DC_OK;
 }
 
-enum dc_status dce80_validate_guaranteed(
-		struct dc *dc,
-		struct dc_stream_state *dc_stream,
-		struct dc_state *context)
-{
-	enum dc_status result = DC_ERROR_UNEXPECTED;
-
-	context->streams[0] = dc_stream;
-	dc_stream_retain(context->streams[0]);
-	context->stream_count++;
-
-	result = resource_map_pool_resources(dc, context, dc_stream);
-
-	if (result == DC_OK)
-		result = resource_map_clock_resources(dc, context, dc_stream);
-
-	if (result == DC_OK)
-		result = build_mapped_resource(dc, context, dc_stream);
-
-	if (result == DC_OK) {
-		validate_guaranteed_copy_streams(
-				context, dc->caps.max_streams);
-		result = resource_build_scaling_params_for_context(dc, context);
-	}
-
-	if (result == DC_OK)
-		result = dce80_validate_bandwidth(dc, context);
-
-	return result;
-}
-
 static void dce80_destroy_resource_pool(struct resource_pool **pool)
 {
 	struct dce110_resource_pool *dce110_pool = TO_DCE110_RES_POOL(*pool);
@@ -792,7 +744,6 @@ static void dce80_destroy_resource_pool(struct resource_pool **pool)
 static const struct resource_funcs dce80_res_pool_funcs = {
 	.destroy = dce80_destroy_resource_pool,
 	.link_enc_create = dce80_link_encoder_create,
-	.validate_guaranteed = dce80_validate_guaranteed,
 	.validate_bandwidth = dce80_validate_bandwidth,
 	.validate_plane = dce100_validate_plane,
 	.add_stream_to_ctx = dce100_add_stream_to_ctx,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
index 3356125..5f40a73 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
@@ -121,6 +121,13 @@ bool dpp_get_optimal_number_of_taps(
 	else
 		pixel_width = scl_data->viewport.width;
 
+	/* Some ASICs does not support  FP16 scaling, so we reject modes require this*/
+	if (scl_data->viewport.width  != scl_data->h_active &&
+		scl_data->viewport.height != scl_data->v_active &&
+		dpp->caps->dscl_data_proc_format == DSCL_DATA_PRCESSING_FIXED_FORMAT &&
+		scl_data->format == PIXEL_FORMAT_FP16)
+		return false;
+
 	/* TODO: add lb check */
 
 	/* No support for programming ratio of 4, drop to 3.99999.. */
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
index 02bd664..a3fe343 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
@@ -918,36 +918,6 @@ enum dc_status dcn10_add_stream_to_ctx(
 	return result;
 }
 
-enum dc_status dcn10_validate_guaranteed(
-		struct dc *dc,
-		struct dc_stream_state *dc_stream,
-		struct dc_state *context)
-{
-	enum dc_status result = DC_ERROR_UNEXPECTED;
-
-	context->streams[0] = dc_stream;
-	dc_stream_retain(context->streams[0]);
-	context->stream_count++;
-
-	result = resource_map_pool_resources(dc, context, dc_stream);
-
-	if (result == DC_OK)
-		result = resource_map_phy_clock_resources(dc, context, dc_stream);
-
-	if (result == DC_OK)
-		result = build_mapped_resource(dc, context, dc_stream);
-
-	if (result == DC_OK) {
-		validate_guaranteed_copy_streams(
-				context, dc->caps.max_streams);
-		result = resource_build_scaling_params_for_context(dc, context);
-	}
-	if (result == DC_OK && !dcn_validate_bandwidth(dc, context))
-		return DC_FAIL_BANDWIDTH_VALIDATE;
-
-	return result;
-}
-
 static struct pipe_ctx *dcn10_acquire_idle_pipe_for_layer(
 		struct dc_state *context,
 		const struct resource_pool *pool,
@@ -1233,7 +1203,6 @@ static struct dc_cap_funcs cap_funcs = {
 static struct resource_funcs dcn10_res_pool_funcs = {
 	.destroy = dcn10_destroy_resource_pool,
 	.link_enc_create = dcn10_link_encoder_create,
-	.validate_guaranteed = dcn10_validate_guaranteed,
 	.validate_bandwidth = dcn_validate_bandwidth,
 	.acquire_idle_pipe_for_layer = dcn10_acquire_idle_pipe_for_layer,
 	.validate_plane = dcn10_validate_plane,
diff --git a/drivers/gpu/drm/amd/display/dc/inc/core_types.h b/drivers/gpu/drm/amd/display/dc/inc/core_types.h
index 55f56bf..a94942d 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/core_types.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/core_types.h
@@ -95,11 +95,6 @@ struct resource_funcs {
 	struct link_encoder *(*link_enc_create)(
 			const struct encoder_init_data *init);
 
-	enum dc_status (*validate_guaranteed)(
-					struct dc *dc,
-					struct dc_stream_state *stream,
-					struct dc_state *context);
-
 	bool (*validate_bandwidth)(
 					struct dc *dc,
 					struct dc_state *context);
diff --git a/drivers/gpu/drm/amd/display/dc/inc/resource.h b/drivers/gpu/drm/amd/display/dc/inc/resource.h
index 5467332f..640a647 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/resource.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/resource.h
@@ -139,10 +139,6 @@ bool resource_validate_attach_surfaces(
 		struct dc_state *context,
 		const struct resource_pool *pool);
 
-void validate_guaranteed_copy_streams(
-		struct dc_state *context,
-		int max_streams);
-
 void resource_validate_ctx_update_pointer_after_copy(
 		const struct dc_state *src_ctx,
 		struct dc_state *dst_ctx);
-- 
cgit v1.1


From 3722c794641f91e0b960dd901d6c5d2f3cc24080 Mon Sep 17 00:00:00 2001
From: Mikita Lipski <mikita.lipski@amd.com>
Date: Wed, 21 Feb 2018 16:57:10 -0500
Subject: drm/amd/display: Implementing new bandwidth registers for DCE120

Registers are added and defined.
Programmed to default values.
Stutter level watermark register is being set to calculated value.
Urgent level registers are programmed to the same as urgency.
The programming of the registers is not expected to have any
functional difference in performance.

Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c | 58 +++++++++++++++-------
 drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.h |  9 ++++
 .../amd/display/dc/dce110/dce110_hw_sequencer.c    |  3 ++
 .../drm/amd/display/dc/dce110/dce110_mem_input_v.c |  1 +
 drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h  |  1 +
 5 files changed, 55 insertions(+), 17 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c b/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c
index 0790f25..04fc86b 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c
@@ -174,6 +174,25 @@ static void program_urgency_watermark(
 		URGENCY_HIGH_WATERMARK, urgency_high_wm);
 }
 
+static void dce120_program_urgency_watermark(
+	struct dce_mem_input *dce_mi,
+	uint32_t wm_select,
+	uint32_t urgency_low_wm,
+	uint32_t urgency_high_wm)
+{
+	REG_UPDATE(DPG_WATERMARK_MASK_CONTROL,
+		URGENCY_WATERMARK_MASK, wm_select);
+
+	REG_SET_2(DPG_PIPE_URGENCY_CONTROL, 0,
+		URGENCY_LOW_WATERMARK, urgency_low_wm,
+		URGENCY_HIGH_WATERMARK, urgency_high_wm);
+
+	REG_SET_2(DPG_PIPE_URGENT_LEVEL_CONTROL, 0,
+		URGENT_LEVEL_LOW_WATERMARK, urgency_low_wm,
+		URGENT_LEVEL_HIGH_WATERMARK, urgency_high_wm);
+
+}
+
 static void program_nbp_watermark(
 	struct dce_mem_input *dce_mi,
 	uint32_t wm_select,
@@ -209,23 +228,27 @@ static void program_nbp_watermark(
 static void program_stutter_watermark(
 	struct dce_mem_input *dce_mi,
 	uint32_t wm_select,
-	uint32_t stutter_mark)
+	uint32_t stutter_mark,
+	uint32_t stutter_entry)
 {
 	REG_UPDATE(DPG_WATERMARK_MASK_CONTROL,
 		STUTTER_EXIT_SELF_REFRESH_WATERMARK_MASK, wm_select);
 
 	if (REG(DPG_PIPE_STUTTER_CONTROL2))
-		REG_UPDATE(DPG_PIPE_STUTTER_CONTROL2,
-				STUTTER_EXIT_SELF_REFRESH_WATERMARK, stutter_mark);
+		REG_UPDATE_2(DPG_PIPE_STUTTER_CONTROL2,
+				STUTTER_EXIT_SELF_REFRESH_WATERMARK, stutter_mark,
+				STUTTER_ENTER_SELF_REFRESH_WATERMARK, stutter_entry);
 	else
-		REG_UPDATE(DPG_PIPE_STUTTER_CONTROL,
-				STUTTER_EXIT_SELF_REFRESH_WATERMARK, stutter_mark);
+		REG_UPDATE_2(DPG_PIPE_STUTTER_CONTROL,
+				STUTTER_EXIT_SELF_REFRESH_WATERMARK, stutter_mark,
+				STUTTER_ENTER_SELF_REFRESH_WATERMARK, stutter_entry);
 }
 
 static void dce_mi_program_display_marks(
 	struct mem_input *mi,
 	struct dce_watermarks nbp,
-	struct dce_watermarks stutter,
+	struct dce_watermarks stutter_exit,
+	struct dce_watermarks stutter_enter,
 	struct dce_watermarks urgent,
 	uint32_t total_dest_line_time_ns)
 {
@@ -243,26 +266,27 @@ static void dce_mi_program_display_marks(
 	program_nbp_watermark(dce_mi, 2, nbp.a_mark); /* set a */
 	program_nbp_watermark(dce_mi, 1, nbp.d_mark); /* set d */
 
-	program_stutter_watermark(dce_mi, 2, stutter.a_mark); /* set a */
-	program_stutter_watermark(dce_mi, 1, stutter.d_mark); /* set d */
+	program_stutter_watermark(dce_mi, 2, stutter_exit.a_mark, stutter_enter.a_mark); /* set a */
+	program_stutter_watermark(dce_mi, 1, stutter_exit.d_mark, stutter_enter.d_mark); /* set d */
 }
 
 static void dce120_mi_program_display_marks(struct mem_input *mi,
 	struct dce_watermarks nbp,
-	struct dce_watermarks stutter,
+	struct dce_watermarks stutter_exit,
+	struct dce_watermarks stutter_entry,
 	struct dce_watermarks urgent,
 	uint32_t total_dest_line_time_ns)
 {
 	struct dce_mem_input *dce_mi = TO_DCE_MEM_INPUT(mi);
 	uint32_t stutter_en = mi->ctx->dc->debug.disable_stutter ? 0 : 1;
 
-	program_urgency_watermark(dce_mi, 0, /* set a */
+	dce120_program_urgency_watermark(dce_mi, 0, /* set a */
 			urgent.a_mark, total_dest_line_time_ns);
-	program_urgency_watermark(dce_mi, 1, /* set b */
+	dce120_program_urgency_watermark(dce_mi, 1, /* set b */
 			urgent.b_mark, total_dest_line_time_ns);
-	program_urgency_watermark(dce_mi, 2, /* set c */
+	dce120_program_urgency_watermark(dce_mi, 2, /* set c */
 			urgent.c_mark, total_dest_line_time_ns);
-	program_urgency_watermark(dce_mi, 3, /* set d */
+	dce120_program_urgency_watermark(dce_mi, 3, /* set d */
 			urgent.d_mark, total_dest_line_time_ns);
 
 	REG_UPDATE_2(DPG_PIPE_STUTTER_CONTROL,
@@ -273,10 +297,10 @@ static void dce120_mi_program_display_marks(struct mem_input *mi,
 	program_nbp_watermark(dce_mi, 2, nbp.c_mark); /* set c */
 	program_nbp_watermark(dce_mi, 3, nbp.d_mark); /* set d */
 
-	program_stutter_watermark(dce_mi, 0, stutter.a_mark); /* set a */
-	program_stutter_watermark(dce_mi, 1, stutter.b_mark); /* set b */
-	program_stutter_watermark(dce_mi, 2, stutter.c_mark); /* set c */
-	program_stutter_watermark(dce_mi, 3, stutter.d_mark); /* set d */
+	program_stutter_watermark(dce_mi, 0, stutter_exit.a_mark, stutter_entry.a_mark); /* set a */
+	program_stutter_watermark(dce_mi, 1, stutter_exit.b_mark, stutter_entry.b_mark); /* set b */
+	program_stutter_watermark(dce_mi, 2, stutter_exit.c_mark, stutter_entry.c_mark); /* set c */
+	program_stutter_watermark(dce_mi, 3, stutter_exit.d_mark, stutter_entry.d_mark); /* set d */
 }
 
 static void program_tiling(
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.h b/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.h
index 05d39c0..e877e73 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.h
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.h
@@ -106,6 +106,7 @@ struct dce_mem_input_registers {
 	uint32_t DPG_PIPE_ARBITRATION_CONTROL1;
 	uint32_t DPG_WATERMARK_MASK_CONTROL;
 	uint32_t DPG_PIPE_URGENCY_CONTROL;
+	uint32_t DPG_PIPE_URGENT_LEVEL_CONTROL;
 	uint32_t DPG_PIPE_NB_PSTATE_CHANGE_CONTROL;
 	uint32_t DPG_PIPE_LOW_POWER_CONTROL;
 	uint32_t DPG_PIPE_STUTTER_CONTROL;
@@ -213,6 +214,11 @@ struct dce_mem_input_registers {
 
 #define MI_DCE12_DMIF_PG_MASK_SH_LIST(mask_sh, blk)\
 	SFB(blk, DPG_PIPE_STUTTER_CONTROL2, STUTTER_EXIT_SELF_REFRESH_WATERMARK, mask_sh),\
+	SFB(blk, DPG_PIPE_STUTTER_CONTROL2, STUTTER_ENTER_SELF_REFRESH_WATERMARK, mask_sh),\
+	SFB(blk, DPG_PIPE_URGENT_LEVEL_CONTROL, URGENT_LEVEL_LOW_WATERMARK, mask_sh),\
+	SFB(blk, DPG_PIPE_URGENT_LEVEL_CONTROL, URGENT_LEVEL_HIGH_WATERMARK, mask_sh),\
+	SFB(blk, DPG_PIPE_URGENCY_CONTROL, URGENCY_LOW_WATERMARK, mask_sh),\
+	SFB(blk, DPG_PIPE_URGENCY_CONTROL, URGENCY_HIGH_WATERMARK, mask_sh),\
 	SFB(blk, DPG_WATERMARK_MASK_CONTROL, PSTATE_CHANGE_WATERMARK_MASK, mask_sh),\
 	SFB(blk, DPG_PIPE_LOW_POWER_CONTROL, PSTATE_CHANGE_ENABLE, mask_sh),\
 	SFB(blk, DPG_PIPE_LOW_POWER_CONTROL, PSTATE_CHANGE_URGENT_DURING_REQUEST, mask_sh),\
@@ -286,6 +292,8 @@ struct dce_mem_input_registers {
 	type STUTTER_EXIT_SELF_REFRESH_WATERMARK_MASK; \
 	type URGENCY_LOW_WATERMARK; \
 	type URGENCY_HIGH_WATERMARK; \
+	type URGENT_LEVEL_LOW_WATERMARK;\
+	type URGENT_LEVEL_HIGH_WATERMARK;\
 	type NB_PSTATE_CHANGE_ENABLE; \
 	type NB_PSTATE_CHANGE_URGENT_DURING_REQUEST; \
 	type NB_PSTATE_CHANGE_NOT_SELF_REFRESH_DURING_REQUEST; \
@@ -297,6 +305,7 @@ struct dce_mem_input_registers {
 	type STUTTER_ENABLE; \
 	type STUTTER_IGNORE_FBC; \
 	type STUTTER_EXIT_SELF_REFRESH_WATERMARK; \
+	type STUTTER_ENTER_SELF_REFRESH_WATERMARK; \
 	type DMIF_BUFFERS_ALLOCATED; \
 	type DMIF_BUFFERS_ALLOCATION_COMPLETED; \
 	type ENABLE; /* MC_HUB_RDREQ_DMIF_LIMIT */\
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
index 30dd62f..daa4673 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
@@ -1539,6 +1539,7 @@ static void dce110_set_displaymarks(
 			pipe_ctx->plane_res.mi,
 			context->bw.dce.nbp_state_change_wm_ns[num_pipes],
 			context->bw.dce.stutter_exit_wm_ns[num_pipes],
+			context->bw.dce.stutter_entry_wm_ns[num_pipes],
 			context->bw.dce.urgent_wm_ns[num_pipes],
 			total_dest_line_time_ns);
 		if (i == underlay_idx) {
@@ -1564,6 +1565,7 @@ static void set_safe_displaymarks(
 		MAX_WATERMARK, MAX_WATERMARK, MAX_WATERMARK, MAX_WATERMARK };
 	struct dce_watermarks nbp_marks = {
 		SAFE_NBP_MARK, SAFE_NBP_MARK, SAFE_NBP_MARK, SAFE_NBP_MARK };
+	struct dce_watermarks min_marks = { 0, 0, 0, 0};
 
 	for (i = 0; i < MAX_PIPES; i++) {
 		if (res_ctx->pipe_ctx[i].stream == NULL || res_ctx->pipe_ctx[i].plane_res.mi == NULL)
@@ -1573,6 +1575,7 @@ static void set_safe_displaymarks(
 				res_ctx->pipe_ctx[i].plane_res.mi,
 				nbp_marks,
 				max_marks,
+				min_marks,
 				max_marks,
 				MAX_WATERMARK);
 
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c
index 7bab8c6..0564c8e 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c
@@ -923,6 +923,7 @@ void dce_mem_input_v_program_display_marks(
 	struct mem_input *mem_input,
 	struct dce_watermarks nbp,
 	struct dce_watermarks stutter,
+	struct dce_watermarks stutter_enter,
 	struct dce_watermarks urgent,
 	uint32_t total_dest_line_time_ns)
 {
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h b/drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h
index 3e1e7e6..47f1dc5 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h
@@ -104,6 +104,7 @@ struct mem_input_funcs {
 		struct mem_input *mem_input,
 		struct dce_watermarks nbp,
 		struct dce_watermarks stutter,
+		struct dce_watermarks stutter_enter,
 		struct dce_watermarks urgent,
 		uint32_t total_dest_line_time_ns);
 
-- 
cgit v1.1


From 8a6095e08e4ae279de570b175e5aee525ae1251e Mon Sep 17 00:00:00 2001
From: Tony Cheng <tony.cheng@amd.com>
Date: Wed, 21 Feb 2018 16:42:05 -0500
Subject: drm/amd/display: dal 3.1.40

Signed-off-by: Tony Cheng <tony.cheng@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dc.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index bdc3cef..63817ed 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -38,7 +38,7 @@
 #include "inc/compressor.h"
 #include "dml/display_mode_lib.h"
 
-#define DC_VER "3.1.39"
+#define DC_VER "3.1.40"
 
 #define MAX_SURFACES 3
 #define MAX_STREAMS 6
-- 
cgit v1.1


From a052a516de4c3e46f2e442ec118c391dbf9932e3 Mon Sep 17 00:00:00 2001
From: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Date: Tue, 13 Mar 2018 15:00:20 -0400
Subject: drm/amd/display: align dtn logs and add mpc idle bit print

Signed-off-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c  | 53 ++++++++++------------
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.c   |  4 +-
 drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h        |  2 +
 3 files changed, 28 insertions(+), 31 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
index 999190a..675a81a 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -56,16 +56,17 @@
 #define FN(reg_name, field_name) \
 	hws->shifts->field_name, hws->masks->field_name
 
+/*print is 17 wide, first two characters are spaces*/
 #define DTN_INFO_MICRO_SEC(ref_cycle) \
 	print_microsec(dc_ctx, ref_cycle)
 
 void print_microsec(struct dc_context *dc_ctx, uint32_t ref_cycle)
 {
-	static const uint32_t ref_clk_mhz = 48;
-	static const unsigned int frac = 10;
+	const uint32_t ref_clk_mhz = dc_ctx->dc->res_pool->ref_clock_inKhz / 1000;
+	static const unsigned int frac = 1000;
 	uint32_t us_x10 = (ref_cycle * frac) / ref_clk_mhz;
 
-	DTN_INFO("%d.%d \t ",
+	DTN_INFO("  %11d.%03d",
 			us_x10 / frac,
 			us_x10 % frac);
 }
@@ -92,14 +93,14 @@ void dcn10_log_hubbub_state(struct dc *dc)
 
 	hubbub1_wm_read_state(dc->res_pool->hubbub, &wm);
 
-	DTN_INFO("HUBBUB WM: \t data_urgent \t pte_meta_urgent \t "
-			"sr_enter \t sr_exit \t dram_clk_change \n");
+	DTN_INFO("HUBBUB WM:      data_urgent  pte_meta_urgent"
+			"         sr_enter          sr_exit  dram_clk_change\n");
 
 	for (i = 0; i < 4; i++) {
 		struct dcn_hubbub_wm_set *s;
 
 		s = &wm.sets[i];
-		DTN_INFO("WM_Set[%d]:\t ", s->wm_set);
+		DTN_INFO("WM_Set[%d]:", s->wm_set);
 		DTN_INFO_MICRO_SEC(s->data_urgent);
 		DTN_INFO_MICRO_SEC(s->pte_meta_urgent);
 		DTN_INFO_MICRO_SEC(s->sr_enter);
@@ -121,19 +122,17 @@ void dcn10_log_hw_state(struct dc *dc)
 
 	dcn10_log_hubbub_state(dc);
 
-	DTN_INFO("HUBP:  format  addr_hi  width  height  "
-			"rotation  mirror  sw_mode  "
-			"dcc_en  blank_en  ttu_dis  underflow  "
-			"min_ttu_vblank  qos_low_wm  qos_high_wm\n");
+	DTN_INFO("HUBP:  format  addr_hi  width  height"
+			"  rot  mir  sw_mode  dcc_en  blank_en  ttu_dis  underflow"
+			"   min_ttu_vblank       qos_low_wm      qos_high_wm\n");
 	for (i = 0; i < pool->pipe_count; i++) {
 		struct hubp *hubp = pool->hubps[i];
 		struct dcn_hubp_state s;
 
 		hubp1_read_state(TO_DCN10_HUBP(hubp), &s);
 
-		DTN_INFO("[%-2d]:  %5xh  %6xh  %5d  %6d  "
-				"%7xh  %5xh  %6xh  "
-				"%6d  %8d  %7d  %8xh \t",
+		DTN_INFO("[%2d]:  %5xh  %6xh  %5d  %6d  %2xh  %2xh  %6xh"
+				"  %6d  %8d  %7d  %8xh",
 				hubp->inst,
 				s.pixel_format,
 				s.inuse_addr_hi,
@@ -152,25 +151,21 @@ void dcn10_log_hw_state(struct dc *dc)
 		DTN_INFO("\n");
 	}
 	DTN_INFO("\n");
+
+	DTN_INFO("MPCC:  OPP  DPP  MPCCBOT  MODE  ALPHA_MODE  PREMULT  OVERLAP_ONLY  IDLE\n");
 	for (i = 0; i < pool->pipe_count; i++) {
-		struct output_pixel_processor *opp = pool->opps[i];
-		struct mpcc *mpcc = opp->mpc_tree_params.opp_list;
 		struct mpcc_state s = {0};
 
-		while (mpcc) {
-			ASSERT(opp->mpc_tree_params.opp_id == opp->inst);
-			pool->mpc->funcs->read_mpcc_state(pool->mpc, mpcc->mpcc_id, &s);
-			DTN_INFO("[OPP%d - MPCC%d]: DPP%d MPCCBOT%x MODE:%d ALPHA_MODE:%d PREMULT:%d OVERLAP_ONLY:%d\n",
-				s.opp_id, mpcc->mpcc_id, s.dpp_id, s.bot_mpcc_id,
-				s.mode, s.alpha_mode, s.pre_multiplied_alpha, s.overlap_only);
-			mpcc = mpcc->mpcc_bot;
-			ASSERT(!mpcc || mpcc->mpcc_id == s.bot_mpcc_id);
-		}
+		pool->mpc->funcs->read_mpcc_state(pool->mpc, i, &s);
+		DTN_INFO("[%2d]:  %2xh  %2xh  %6xh  %4d  %10d  %7d  %12d  %4d\n",
+			i, s.opp_id, s.dpp_id, s.bot_mpcc_id,
+			s.mode, s.alpha_mode, s.pre_multiplied_alpha, s.overlap_only,
+			s.idle);
 	}
 	DTN_INFO("\n");
 
-	DTN_INFO("OTG:\t v_bs \t v_be \t v_ss \t v_se \t vpol \t vmax \t vmin \t "
-			"h_bs \t h_be \t h_ss \t h_se \t hpol \t htot \t vtot \t underflow\n");
+	DTN_INFO("OTG:  v_bs  v_be  v_ss  v_se  vpol  vmax  vmin"
+			"  h_bs  h_be  h_ss  h_se  hpol  htot  vtot  underflow\n");
 
 	for (i = 0; i < pool->timing_generator_count; i++) {
 		struct timing_generator *tg = pool->timing_generators[i];
@@ -182,9 +177,8 @@ void dcn10_log_hw_state(struct dc *dc)
 		if ((s.otg_enabled & 1) == 0)
 			continue;
 
-		DTN_INFO("[%d]:\t %d \t %d \t %d \t %d \t "
-				"%d \t %d \t %d \t %d \t %d \t %d \t "
-				"%d \t %d \t %d \t %d \t %d \t ",
+		DTN_INFO("[%d]: %5d %5d %5d %5d %5d %5d %5d %5d %5d %5d"
+				" %5d %5d %5d %5d  %9d\n",
 				tg->inst,
 				s.v_blank_start,
 				s.v_blank_end,
@@ -201,7 +195,6 @@ void dcn10_log_hw_state(struct dc *dc)
 				s.h_total,
 				s.v_total,
 				s.underflow_occurred_status);
-		DTN_INFO("\n");
 	}
 	DTN_INFO("\n");
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.c
index 29e15a9..6f7016a 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.c
@@ -422,7 +422,9 @@ void mpc1_read_mpcc_state(
 	REG_GET_4(MPCC_CONTROL[mpcc_inst], MPCC_MODE, &s->mode,
 			MPCC_ALPHA_BLND_MODE, &s->alpha_mode,
 			MPCC_ALPHA_MULTIPLIED_MODE, &s->pre_multiplied_alpha,
-			MPCC_BLND_ACTIVE_OVERLAP_ONLY, &s->pre_multiplied_alpha);
+			MPCC_BLND_ACTIVE_OVERLAP_ONLY, &s->overlap_only);
+	REG_GET_2(MPCC_STATUS[mpcc_inst], MPCC_IDLE, &s->idle,
+			MPCC_BUSY, &s->busy);
 }
 
 const struct mpc_funcs dcn10_mpc_funcs = {
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h b/drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h
index 5caacab..caf74e3 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h
@@ -113,6 +113,8 @@ struct mpcc_state {
 	uint32_t alpha_mode;
 	uint32_t pre_multiplied_alpha;
 	uint32_t overlap_only;
+	uint32_t idle;
+	uint32_t busy;
 };
 
 struct mpc_funcs {
-- 
cgit v1.1


From 7a84077304e84d0254d505a76cc40971cb74c2de Mon Sep 17 00:00:00 2001
From: Roman Li <Roman.Li@amd.com>
Date: Wed, 14 Mar 2018 18:02:07 -0400
Subject: drm/amd/display: add assert in enable FBC

Adding assert to prevent possible null deref warning.
Only can happen under abnormal circumstances.

Signed-off-by: Roman Li <Roman.Li@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
index daa4673..075ab29 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
@@ -1801,6 +1801,9 @@ static bool should_enable_fbc(struct dc *dc,
 		}
 	}
 
+	/* Pipe context should be found */
+	ASSERT(pipe_ctx);
+
 	/* Only supports eDP */
 	if (pipe_ctx->stream->sink->link->connector_signal != SIGNAL_TYPE_EDP)
 		return false;
-- 
cgit v1.1


From d72eb20379022a948c219e1fc451b6b0200cc7c5 Mon Sep 17 00:00:00 2001
From: Mikita Lipski <mikita.lipski@amd.com>
Date: Wed, 14 Mar 2018 14:42:25 -0400
Subject: drm/amd/display: Separate mem input constuctors for dce 112 and 120

Override the memory input functions for dce120 not to program
new registers on dce112.
This will fix warnings thrown on Polaris asics.

Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c | 75 ++++++++++++++++++++--
 drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.h |  8 +++
 .../drm/amd/display/dc/dce120/dce120_resource.c    |  2 +-
 3 files changed, 77 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c b/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c
index 04fc86b..b235a75 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c
@@ -225,7 +225,7 @@ static void program_nbp_watermark(
 	}
 }
 
-static void program_stutter_watermark(
+static void dce120_program_stutter_watermark(
 	struct dce_mem_input *dce_mi,
 	uint32_t wm_select,
 	uint32_t stutter_mark,
@@ -244,6 +244,22 @@ static void program_stutter_watermark(
 				STUTTER_ENTER_SELF_REFRESH_WATERMARK, stutter_entry);
 }
 
+static void program_stutter_watermark(
+	struct dce_mem_input *dce_mi,
+	uint32_t wm_select,
+	uint32_t stutter_mark)
+{
+	REG_UPDATE(DPG_WATERMARK_MASK_CONTROL,
+		STUTTER_EXIT_SELF_REFRESH_WATERMARK_MASK, wm_select);
+
+	if (REG(DPG_PIPE_STUTTER_CONTROL2))
+		REG_UPDATE(DPG_PIPE_STUTTER_CONTROL2,
+				STUTTER_EXIT_SELF_REFRESH_WATERMARK, stutter_mark);
+	else
+		REG_UPDATE(DPG_PIPE_STUTTER_CONTROL,
+				STUTTER_EXIT_SELF_REFRESH_WATERMARK, stutter_mark);
+}
+
 static void dce_mi_program_display_marks(
 	struct mem_input *mi,
 	struct dce_watermarks nbp,
@@ -266,8 +282,41 @@ static void dce_mi_program_display_marks(
 	program_nbp_watermark(dce_mi, 2, nbp.a_mark); /* set a */
 	program_nbp_watermark(dce_mi, 1, nbp.d_mark); /* set d */
 
-	program_stutter_watermark(dce_mi, 2, stutter_exit.a_mark, stutter_enter.a_mark); /* set a */
-	program_stutter_watermark(dce_mi, 1, stutter_exit.d_mark, stutter_enter.d_mark); /* set d */
+	program_stutter_watermark(dce_mi, 2, stutter_exit.a_mark); /* set a */
+	program_stutter_watermark(dce_mi, 1, stutter_exit.d_mark); /* set d */
+}
+
+static void dce112_mi_program_display_marks(struct mem_input *mi,
+	struct dce_watermarks nbp,
+	struct dce_watermarks stutter_exit,
+	struct dce_watermarks stutter_entry,
+	struct dce_watermarks urgent,
+	uint32_t total_dest_line_time_ns)
+{
+	struct dce_mem_input *dce_mi = TO_DCE_MEM_INPUT(mi);
+	uint32_t stutter_en = mi->ctx->dc->debug.disable_stutter ? 0 : 1;
+
+	program_urgency_watermark(dce_mi, 0, /* set a */
+			urgent.a_mark, total_dest_line_time_ns);
+	program_urgency_watermark(dce_mi, 1, /* set b */
+			urgent.b_mark, total_dest_line_time_ns);
+	program_urgency_watermark(dce_mi, 2, /* set c */
+			urgent.c_mark, total_dest_line_time_ns);
+	program_urgency_watermark(dce_mi, 3, /* set d */
+			urgent.d_mark, total_dest_line_time_ns);
+
+	REG_UPDATE_2(DPG_PIPE_STUTTER_CONTROL,
+		STUTTER_ENABLE, stutter_en,
+		STUTTER_IGNORE_FBC, 1);
+	program_nbp_watermark(dce_mi, 0, nbp.a_mark); /* set a */
+	program_nbp_watermark(dce_mi, 1, nbp.b_mark); /* set b */
+	program_nbp_watermark(dce_mi, 2, nbp.c_mark); /* set c */
+	program_nbp_watermark(dce_mi, 3, nbp.d_mark); /* set d */
+
+	program_stutter_watermark(dce_mi, 0, stutter_exit.a_mark); /* set a */
+	program_stutter_watermark(dce_mi, 1, stutter_exit.b_mark); /* set b */
+	program_stutter_watermark(dce_mi, 2, stutter_exit.c_mark); /* set c */
+	program_stutter_watermark(dce_mi, 3, stutter_exit.d_mark); /* set d */
 }
 
 static void dce120_mi_program_display_marks(struct mem_input *mi,
@@ -297,10 +346,10 @@ static void dce120_mi_program_display_marks(struct mem_input *mi,
 	program_nbp_watermark(dce_mi, 2, nbp.c_mark); /* set c */
 	program_nbp_watermark(dce_mi, 3, nbp.d_mark); /* set d */
 
-	program_stutter_watermark(dce_mi, 0, stutter_exit.a_mark, stutter_entry.a_mark); /* set a */
-	program_stutter_watermark(dce_mi, 1, stutter_exit.b_mark, stutter_entry.b_mark); /* set b */
-	program_stutter_watermark(dce_mi, 2, stutter_exit.c_mark, stutter_entry.c_mark); /* set c */
-	program_stutter_watermark(dce_mi, 3, stutter_exit.d_mark, stutter_entry.d_mark); /* set d */
+	dce120_program_stutter_watermark(dce_mi, 0, stutter_exit.a_mark, stutter_entry.a_mark); /* set a */
+	dce120_program_stutter_watermark(dce_mi, 1, stutter_exit.b_mark, stutter_entry.b_mark); /* set b */
+	dce120_program_stutter_watermark(dce_mi, 2, stutter_exit.c_mark, stutter_entry.c_mark); /* set c */
+	dce120_program_stutter_watermark(dce_mi, 3, stutter_exit.d_mark, stutter_entry.d_mark); /* set d */
 }
 
 static void program_tiling(
@@ -720,5 +769,17 @@ void dce112_mem_input_construct(
 	const struct dce_mem_input_mask *mi_mask)
 {
 	dce_mem_input_construct(dce_mi, ctx, inst, regs, mi_shift, mi_mask);
+	dce_mi->base.funcs->mem_input_program_display_marks = dce112_mi_program_display_marks;
+}
+
+void dce120_mem_input_construct(
+	struct dce_mem_input *dce_mi,
+	struct dc_context *ctx,
+	int inst,
+	const struct dce_mem_input_registers *regs,
+	const struct dce_mem_input_shift *mi_shift,
+	const struct dce_mem_input_mask *mi_mask)
+{
+	dce_mem_input_construct(dce_mi, ctx, inst, regs, mi_shift, mi_mask);
 	dce_mi->base.funcs->mem_input_program_display_marks = dce120_mi_program_display_marks;
 }
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.h b/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.h
index e877e73..d15b0d7 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.h
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.h
@@ -353,4 +353,12 @@ void dce112_mem_input_construct(
 	const struct dce_mem_input_shift *mi_shift,
 	const struct dce_mem_input_mask *mi_mask);
 
+void dce120_mem_input_construct(
+	struct dce_mem_input *dce_mi,
+	struct dc_context *ctx,
+	int inst,
+	const struct dce_mem_input_registers *regs,
+	const struct dce_mem_input_shift *mi_shift,
+	const struct dce_mem_input_mask *mi_mask);
+
 #endif /*__DCE_MEM_INPUT_H__*/
diff --git a/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c b/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c
index 567e6b4..fda0157 100644
--- a/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c
@@ -652,7 +652,7 @@ static struct mem_input *dce120_mem_input_create(
 		return NULL;
 	}
 
-	dce112_mem_input_construct(dce_mi, ctx, inst, &mi_regs[inst], &mi_shifts, &mi_masks);
+	dce120_mem_input_construct(dce_mi, ctx, inst, &mi_regs[inst], &mi_shifts, &mi_masks);
 	return &dce_mi->base;
 }
 
-- 
cgit v1.1


From 4173c0bdd7b79ef46161037f8845654416dbaca9 Mon Sep 17 00:00:00 2001
From: Eric Yang <Eric.Yang2@amd.com>
Date: Wed, 14 Mar 2018 17:41:57 -0400
Subject: drm/amd/display: Only update mpc blend config if not full update

The current mpcc insert/remove logic does not support updating
only a single mpcc. So when pixel alpha changed but no full update
we can mistakenly shuffle the mpcc layering order. With this change
we will only insert/remove mpcc if there is full update.

Signed-off-by: Eric Yang <Eric.Yang2@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c | 9 +++++++++
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.c          | 5 +++--
 2 files changed, 12 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
index 675a81a..27ae88e3 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -1631,6 +1631,8 @@ static void update_mpcc(struct dc *dc, struct pipe_ctx *pipe_ctx)
 	struct mpc *mpc = dc->res_pool->mpc;
 	struct mpc_tree *mpc_tree_params = &(pipe_ctx->stream_res.opp->mpc_tree_params);
 
+
+
 	/* TODO: proper fix once fpga works */
 
 	if (dc->debug.surface_visual_confirm)
@@ -1657,6 +1659,7 @@ static void update_mpcc(struct dc *dc, struct pipe_ctx *pipe_ctx)
 			pipe_ctx->stream->output_color_space)
 					&& per_pixel_alpha;
 
+
 	/*
 	 * TODO: remove hack
 	 * Note: currently there is a bug in init_hw such that
@@ -1667,6 +1670,12 @@ static void update_mpcc(struct dc *dc, struct pipe_ctx *pipe_ctx)
 	 */
 	mpcc_id = hubp->inst;
 
+	/* If there is no full update, don't need to touch MPC tree*/
+	if (!pipe_ctx->plane_state->update_flags.bits.full_update) {
+		mpc->funcs->update_blending(mpc, &blnd_cfg, mpcc_id);
+		return;
+	}
+
 	/* check if this MPCC is already being used */
 	new_mpcc = mpc->funcs->get_mpcc_for_dpp(mpc_tree_params, mpcc_id);
 	/* remove MPCC if being used */
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.c
index 6f7016a..9ca51ae 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.c
@@ -65,6 +65,7 @@ static void mpc1_update_blending(
 	int mpcc_id)
 {
 	struct dcn10_mpc *mpc10 = TO_DCN10_MPC(mpc);
+	struct mpcc *mpcc = mpc1_get_mpcc(mpc, mpcc_id);
 
 	REG_UPDATE_5(MPCC_CONTROL[mpcc_id],
 			MPCC_ALPHA_BLND_MODE,		blnd_cfg->alpha_mode,
@@ -74,6 +75,7 @@ static void mpc1_update_blending(
 			MPCC_GLOBAL_GAIN,		blnd_cfg->global_gain);
 
 	mpc1_set_bg_color(mpc, &blnd_cfg->black_color, mpcc_id);
+	mpcc->blnd_cfg = *blnd_cfg;
 }
 
 void mpc1_update_stereo_mix(
@@ -235,8 +237,7 @@ struct mpcc *mpc1_insert_plane(
 	}
 
 	/* update the blending configuration */
-	new_mpcc->blnd_cfg = *blnd_cfg;
-	mpc->funcs->update_blending(mpc, &new_mpcc->blnd_cfg, mpcc_id);
+	mpc->funcs->update_blending(mpc, blnd_cfg, mpcc_id);
 
 	/* update the stereo mix settings, if provided */
 	if (sm_cfg != NULL) {
-- 
cgit v1.1


From 1ba2faf207b47e23b1d756e7be25e980724214a7 Mon Sep 17 00:00:00 2001
From: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Date: Thu, 15 Mar 2018 10:25:43 -0400
Subject: drm/amd/display: hide inconsistent mpcc programming from dtn log

Signed-off-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
index 27ae88e3..e214581 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -157,10 +157,11 @@ void dcn10_log_hw_state(struct dc *dc)
 		struct mpcc_state s = {0};
 
 		pool->mpc->funcs->read_mpcc_state(pool->mpc, i, &s);
-		DTN_INFO("[%2d]:  %2xh  %2xh  %6xh  %4d  %10d  %7d  %12d  %4d\n",
-			i, s.opp_id, s.dpp_id, s.bot_mpcc_id,
-			s.mode, s.alpha_mode, s.pre_multiplied_alpha, s.overlap_only,
-			s.idle);
+		if (s.opp_id != 0xf)
+			DTN_INFO("[%2d]:  %2xh  %2xh  %6xh  %4d  %10d  %7d  %12d  %4d\n",
+				i, s.opp_id, s.dpp_id, s.bot_mpcc_id,
+				s.mode, s.alpha_mode, s.pre_multiplied_alpha, s.overlap_only,
+				s.idle);
 	}
 	DTN_INFO("\n");
 
-- 
cgit v1.1


From 24238ee6549bcbebca2f6fc49c225c4f377df8c0 Mon Sep 17 00:00:00 2001
From: Vitaly Prosyak <vitaly.prosyak@amd.com>
Date: Tue, 13 Mar 2018 15:18:34 -0500
Subject: drm/amd/display: Add dc_lut_mode enum

Signed-off-by: Vitaly Prosyak <vitaly.prosyak@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h b/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h
index b221581..015e209 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h
@@ -191,4 +191,9 @@ enum controller_dp_test_pattern {
 	CONTROLLER_DP_TEST_PATTERN_COLORSQUARES_CEA
 };
 
+enum dc_lut_mode {
+	LUT_BYPASS,
+	LUT_RAM_A,
+	LUT_RAM_B
+};
 #endif /* __DAL_HW_SHARED_H__ */
-- 
cgit v1.1


From 3032deb52a6bf706657c39d6335c81ce3265974d Mon Sep 17 00:00:00 2001
From: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Date: Wed, 14 Mar 2018 11:19:15 -0400
Subject: drm/amd/display: Correct print types in DC_LOGS

Correct the types used for printing in logs. This is needed for adding
dynamic printing (LINUX), otherwise we get warnings.

Signed-off-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c   | 74 +++++++++++-----------
 drivers/gpu/drm/amd/display/dc/core/dc_debug.c     | 14 ++--
 drivers/gpu/drm/amd/display/dc/core/dc_link.c      |  8 +--
 .../amd/display/dc/dce110/dce110_hw_sequencer.c    |  4 +-
 .../dc/i2caux/dce110/i2c_hw_engine_dce110.c        |  2 +-
 5 files changed, 52 insertions(+), 50 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c b/drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c
index 4bb43a3..a102c19 100644
--- a/drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c
+++ b/drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c
@@ -1459,39 +1459,39 @@ void dcn_bw_notify_pplib_of_wm_ranges(struct dc *dc)
 void dcn_bw_sync_calcs_and_dml(struct dc *dc)
 {
 	kernel_fpu_begin();
-	DC_LOG_BANDWIDTH_CALCS("sr_exit_time: %d ns\n"
-			"sr_enter_plus_exit_time: %d ns\n"
-			"urgent_latency: %d ns\n"
-			"write_back_latency: %d ns\n"
-			"percent_of_ideal_drambw_received_after_urg_latency: %d %\n"
+	DC_LOG_BANDWIDTH_CALCS("sr_exit_time: %f ns\n"
+			"sr_enter_plus_exit_time: %f ns\n"
+			"urgent_latency: %f ns\n"
+			"write_back_latency: %f ns\n"
+			"percent_of_ideal_drambw_received_after_urg_latency: %f %%\n"
 			"max_request_size: %d bytes\n"
-			"dcfclkv_max0p9: %d kHz\n"
-			"dcfclkv_nom0p8: %d kHz\n"
-			"dcfclkv_mid0p72: %d kHz\n"
-			"dcfclkv_min0p65: %d kHz\n"
-			"max_dispclk_vmax0p9: %d kHz\n"
-			"max_dispclk_vnom0p8: %d kHz\n"
-			"max_dispclk_vmid0p72: %d kHz\n"
-			"max_dispclk_vmin0p65: %d kHz\n"
-			"max_dppclk_vmax0p9: %d kHz\n"
-			"max_dppclk_vnom0p8: %d kHz\n"
-			"max_dppclk_vmid0p72: %d kHz\n"
-			"max_dppclk_vmin0p65: %d kHz\n"
-			"socclk: %d kHz\n"
-			"fabric_and_dram_bandwidth_vmax0p9: %d MB/s\n"
-			"fabric_and_dram_bandwidth_vnom0p8: %d MB/s\n"
-			"fabric_and_dram_bandwidth_vmid0p72: %d MB/s\n"
-			"fabric_and_dram_bandwidth_vmin0p65: %d MB/s\n"
-			"phyclkv_max0p9: %d kHz\n"
-			"phyclkv_nom0p8: %d kHz\n"
-			"phyclkv_mid0p72: %d kHz\n"
-			"phyclkv_min0p65: %d kHz\n"
-			"downspreading: %d %\n"
+			"dcfclkv_max0p9: %f kHz\n"
+			"dcfclkv_nom0p8: %f kHz\n"
+			"dcfclkv_mid0p72: %f kHz\n"
+			"dcfclkv_min0p65: %f kHz\n"
+			"max_dispclk_vmax0p9: %f kHz\n"
+			"max_dispclk_vnom0p8: %f kHz\n"
+			"max_dispclk_vmid0p72: %f kHz\n"
+			"max_dispclk_vmin0p65: %f kHz\n"
+			"max_dppclk_vmax0p9: %f kHz\n"
+			"max_dppclk_vnom0p8: %f kHz\n"
+			"max_dppclk_vmid0p72: %f kHz\n"
+			"max_dppclk_vmin0p65: %f kHz\n"
+			"socclk: %f kHz\n"
+			"fabric_and_dram_bandwidth_vmax0p9: %f MB/s\n"
+			"fabric_and_dram_bandwidth_vnom0p8: %f MB/s\n"
+			"fabric_and_dram_bandwidth_vmid0p72: %f MB/s\n"
+			"fabric_and_dram_bandwidth_vmin0p65: %f MB/s\n"
+			"phyclkv_max0p9: %f kHz\n"
+			"phyclkv_nom0p8: %f kHz\n"
+			"phyclkv_mid0p72: %f kHz\n"
+			"phyclkv_min0p65: %f kHz\n"
+			"downspreading: %f %%\n"
 			"round_trip_ping_latency_cycles: %d DCFCLK Cycles\n"
 			"urgent_out_of_order_return_per_channel: %d Bytes\n"
 			"number_of_channels: %d\n"
 			"vmm_page_size: %d Bytes\n"
-			"dram_clock_change_latency: %d ns\n"
+			"dram_clock_change_latency: %f ns\n"
 			"return_bus_width: %d Bytes\n",
 			dc->dcn_soc->sr_exit_time * 1000,
 			dc->dcn_soc->sr_enter_plus_exit_time * 1000,
@@ -1527,11 +1527,11 @@ void dcn_bw_sync_calcs_and_dml(struct dc *dc)
 			dc->dcn_soc->vmm_page_size,
 			dc->dcn_soc->dram_clock_change_latency * 1000,
 			dc->dcn_soc->return_bus_width);
-	DC_LOG_BANDWIDTH_CALCS("rob_buffer_size_in_kbyte: %d\n"
-			"det_buffer_size_in_kbyte: %d\n"
-			"dpp_output_buffer_pixels: %d\n"
-			"opp_output_buffer_lines: %d\n"
-			"pixel_chunk_size_in_kbyte: %d\n"
+	DC_LOG_BANDWIDTH_CALCS("rob_buffer_size_in_kbyte: %f\n"
+			"det_buffer_size_in_kbyte: %f\n"
+			"dpp_output_buffer_pixels: %f\n"
+			"opp_output_buffer_lines: %f\n"
+			"pixel_chunk_size_in_kbyte: %f\n"
 			"pte_enable: %d\n"
 			"pte_chunk_size: %d kbytes\n"
 			"meta_chunk_size: %d kbytes\n"
@@ -1550,13 +1550,13 @@ void dcn_bw_sync_calcs_and_dml(struct dc *dc)
 			"max_pscl_tolb_throughput: %d pixels/dppclk\n"
 			"max_lb_tovscl_throughput: %d pixels/dppclk\n"
 			"max_vscl_tohscl_throughput: %d pixels/dppclk\n"
-			"max_hscl_ratio: %d\n"
-			"max_vscl_ratio: %d\n"
+			"max_hscl_ratio: %f\n"
+			"max_vscl_ratio: %f\n"
 			"max_hscl_taps: %d\n"
 			"max_vscl_taps: %d\n"
 			"pte_buffer_size_in_requests: %d\n"
-			"dispclk_ramping_margin: %d %\n"
-			"under_scan_factor: %d %\n"
+			"dispclk_ramping_margin: %f %%\n"
+			"under_scan_factor: %f %%\n"
 			"max_inter_dcn_tile_repeaters: %d\n"
 			"can_vstartup_lines_exceed_vsync_plus_back_porch_lines_minus_one: %d\n"
 			"bug_forcing_luma_and_chroma_request_to_same_size_fixed: %d\n"
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_debug.c b/drivers/gpu/drm/amd/display/dc/core/dc_debug.c
index 5a552cb3..71cc60f 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_debug.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_debug.c
@@ -72,8 +72,8 @@ void pre_surface_trace(
 				"plane_state->visible = %d;\n"
 				"plane_state->flip_immediate = %d;\n"
 				"plane_state->address.type = %d;\n"
-				"plane_state->address.grph.addr.quad_part = 0x%X;\n"
-				"plane_state->address.grph.meta_addr.quad_part = 0x%X;\n"
+				"plane_state->address.grph.addr.quad_part = 0x%llX;\n"
+				"plane_state->address.grph.meta_addr.quad_part = 0x%llX;\n"
 				"plane_state->scaling_quality.h_taps = %d;\n"
 				"plane_state->scaling_quality.v_taps = %d;\n"
 				"plane_state->scaling_quality.h_taps_c = %d;\n"
@@ -192,8 +192,8 @@ void update_surface_trace(
 		SURFACE_TRACE("Update %d\n", i);
 		if (update->flip_addr) {
 			SURFACE_TRACE("flip_addr->address.type = %d;\n"
-					"flip_addr->address.grph.addr.quad_part = 0x%X;\n"
-					"flip_addr->address.grph.meta_addr.quad_part = 0x%X;\n"
+					"flip_addr->address.grph.addr.quad_part = 0x%llX;\n"
+					"flip_addr->address.grph.meta_addr.quad_part = 0x%llX;\n"
 					"flip_addr->flip_immediate = %d;\n",
 					update->flip_addr->address.type,
 					update->flip_addr->address.grph.addr.quad_part,
@@ -211,7 +211,8 @@ void update_surface_trace(
 					"plane_info->plane_size.grph.surface_size.width = %d;\n"
 					"plane_info->plane_size.grph.surface_size.x = %d;\n"
 					"plane_info->plane_size.grph.surface_size.y = %d;\n"
-					"plane_info->rotation = %d;\n",
+					"plane_info->rotation = %d;\n"
+					"plane_info->stereo_format = %d;\n",
 					update->plane_info->color_space,
 					update->plane_info->input_tf,
 					update->plane_info->format,
@@ -371,6 +372,7 @@ void context_clock_trace(
 			context->bw.dcn.calc_clk.dppclk_khz,
 			context->bw.dcn.calc_clk.dcfclk_khz,
 			context->bw.dcn.calc_clk.dcfclk_deep_sleep_khz,
-			context->bw.dcn.calc_clk.fclk_khz);
+			context->bw.dcn.calc_clk.fclk_khz,
+			context->bw.dcn.calc_clk.socclk_khz);
 #endif
 }
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
index eeb0447..82ee9de 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
@@ -2165,11 +2165,11 @@ static enum dc_status allocate_mst_payload(struct pipe_ctx *pipe_ctx)
 			link->mst_stream_alloc_table.stream_count);
 
 	for (i = 0; i < MAX_CONTROLLER_NUM; i++) {
-		DC_LOG_MST("stream_enc[%d]: 0x%x      "
+		DC_LOG_MST("stream_enc[%d]: %p      "
 		"stream[%d].vcp_id: %d      "
 		"stream[%d].slot_count: %d\n",
 		i,
-		link->mst_stream_alloc_table.stream_allocations[i].stream_enc,
+		(void *) link->mst_stream_alloc_table.stream_allocations[i].stream_enc,
 		i,
 		link->mst_stream_alloc_table.stream_allocations[i].vcp_id,
 		i,
@@ -2255,11 +2255,11 @@ static enum dc_status deallocate_mst_payload(struct pipe_ctx *pipe_ctx)
 			link->mst_stream_alloc_table.stream_count);
 
 	for (i = 0; i < MAX_CONTROLLER_NUM; i++) {
-		DC_LOG_MST("stream_enc[%d]: 0x%x      "
+		DC_LOG_MST("stream_enc[%d]: %p      "
 		"stream[%d].vcp_id: %d      "
 		"stream[%d].slot_count: %d\n",
 		i,
-		link->mst_stream_alloc_table.stream_allocations[i].stream_enc,
+		(void *) link->mst_stream_alloc_table.stream_allocations[i].stream_enc,
 		i,
 		link->mst_stream_alloc_table.stream_allocations[i].vcp_id,
 		i,
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
index 075ab29..c621230 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
@@ -2777,13 +2777,13 @@ static void dce110_program_front_end_for_pipe(
 		dc->hwss.set_output_transfer_func(pipe_ctx, pipe_ctx->stream);
 
 	DC_LOG_SURFACE(
-			"Pipe:%d 0x%x: addr hi:0x%x, "
+			"Pipe:%d %p: addr hi:0x%x, "
 			"addr low:0x%x, "
 			"src: %d, %d, %d,"
 			" %d; dst: %d, %d, %d, %d;"
 			"clip: %d, %d, %d, %d\n",
 			pipe_ctx->pipe_idx,
-			pipe_ctx->plane_state,
+			(void *) pipe_ctx->plane_state,
 			pipe_ctx->plane_state->address.grph.addr.high_part,
 			pipe_ctx->plane_state->address.grph.addr.low_part,
 			pipe_ctx->plane_state->src_rect.x,
diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2c_hw_engine_dce110.c b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2c_hw_engine_dce110.c
index abd0095..b7256f5 100644
--- a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2c_hw_engine_dce110.c
+++ b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2c_hw_engine_dce110.c
@@ -527,7 +527,7 @@ static void construct(
 	REG_GET(MICROSECOND_TIME_BASE_DIV, XTAL_REF_DIV, &xtal_ref_div);
 
 	if (xtal_ref_div == 0) {
-		DC_LOG_WARNING("Invalid base timer divider\n",
+		DC_LOG_WARNING("Invalid base timer divider [%s]\n",
 				__func__);
 		xtal_ref_div = 2;
 	}
-- 
cgit v1.1


From 8d815b4635382dddd58bf03bbcfac9f4e5201151 Mon Sep 17 00:00:00 2001
From: Eric Bernstein <eric.bernstein@amd.com>
Date: Tue, 13 Mar 2018 16:40:51 -0400
Subject: drm/amd/display: Add num_active_wb to DML

Signed-off-by: Eric Bernstein <eric.bernstein@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h
index e296de6..ce750ed 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h
+++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h
@@ -224,6 +224,7 @@ struct	_vcs_dpi_display_output_params_st	{
 	int	output_bpp;
 	int	dsc_enable;
 	int	wb_enable;
+	int	num_active_wb;
 	int	opp_input_bpc;
 	int	output_type;
 	int	output_format;
-- 
cgit v1.1


From 7608f8569d8fee1372d4a3409aea5cca0b13b194 Mon Sep 17 00:00:00 2001
From: Xingyue Tao <xingyue.tao@amd.com>
Date: Wed, 14 Mar 2018 17:57:42 -0400
Subject: drm/amd/display: Add double buffer machanism to ICSC

- Video playback shows tearing when adjusting
brightness through radeon custom settings.
- Now added double buffer mechanism to switch
input CSC from register buffer ICSC and COMA
- Improved tab alignment

Signed-off-by: Xingyue Tao <xingyue.tao@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h   | 11 +++++-
 .../gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c    | 44 ++++++++++++++--------
 .../gpu/drm/amd/display/dc/dcn10/dcn10_resource.c  | 31 ++++++++++++++-
 3 files changed, 67 insertions(+), 19 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h
index 17b062a..b81b2aa 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h
@@ -108,6 +108,8 @@
 	SRI(CM_DGAM_LUT_DATA, CM, id), \
 	SRI(CM_CONTROL, CM, id), \
 	SRI(CM_DGAM_CONTROL, CM, id), \
+	SRI(CM_TEST_DEBUG_INDEX, CM, id), \
+	SRI(CM_TEST_DEBUG_DATA, CM, id), \
 	SRI(FORMAT_CONTROL, CNVC_CFG, id), \
 	SRI(CNVC_SURFACE_PIXEL_FORMAT, CNVC_CFG, id), \
 	SRI(CURSOR0_CONTROL, CNVC_CUR, id), \
@@ -300,6 +302,7 @@
 	TF_SF(CM0_CM_DGAM_LUT_INDEX, CM_DGAM_LUT_INDEX, mask_sh), \
 	TF_SF(CM0_CM_DGAM_LUT_DATA, CM_DGAM_LUT_DATA, mask_sh), \
 	TF_SF(CM0_CM_DGAM_CONTROL, CM_DGAM_LUT_MODE, mask_sh), \
+	TF_SF(CM0_CM_TEST_DEBUG_INDEX, CM_TEST_DEBUG_INDEX, mask_sh), \
 	TF_SF(CNVC_CFG0_FORMAT_CONTROL, CNVC_BYPASS, mask_sh), \
 	TF2_SF(CNVC_CFG0, FORMAT_CONTROL__ALPHA_EN, mask_sh), \
 	TF_SF(CNVC_CFG0_FORMAT_CONTROL, FORMAT_EXPANSION_MODE, mask_sh), \
@@ -1010,6 +1013,8 @@
 	type CUR0_EXPANSION_MODE; \
 	type CUR0_ENABLE; \
 	type CM_BYPASS; \
+	type CM_TEST_DEBUG_INDEX; \
+	type CM_TEST_DEBUG_DATA_ID9_ICSC_MODE; \
 	type FORMAT_CONTROL__ALPHA_EN; \
 	type CUR0_COLOR0; \
 	type CUR0_COLOR1; \
@@ -1255,6 +1260,8 @@ struct dcn_dpp_mask {
 	uint32_t CM_IGAM_LUT_RW_CONTROL; \
 	uint32_t CM_IGAM_LUT_RW_INDEX; \
 	uint32_t CM_IGAM_LUT_SEQ_COLOR; \
+	uint32_t CM_TEST_DEBUG_INDEX; \
+	uint32_t CM_TEST_DEBUG_DATA; \
 	uint32_t FORMAT_CONTROL; \
 	uint32_t CNVC_SURFACE_PIXEL_FORMAT; \
 	uint32_t CURSOR_CONTROL; \
@@ -1289,8 +1296,8 @@ struct dcn10_dpp {
 
 enum dcn10_input_csc_select {
 	INPUT_CSC_SELECT_BYPASS = 0,
-	INPUT_CSC_SELECT_ICSC,
-	INPUT_CSC_SELECT_COMA
+	INPUT_CSC_SELECT_ICSC = 1,
+	INPUT_CSC_SELECT_COMA = 2
 };
 
 void dpp1_set_cursor_attributes(
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c
index fb32975e..cc51141 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c
@@ -267,6 +267,7 @@ void dpp1_cm_set_output_csc_default(
 		BREAK_TO_DEBUGGER();
 		return;
 	}
+
 	dpp1_cm_program_color_matrix(dpp, regval);
 	REG_SET(CM_OCSC_CONTROL, 0, CM_OCSC_MODE, ocsc_mode);
 }
@@ -330,6 +331,7 @@ void dpp1_cm_set_output_csc_adjustment(
 {
 	struct dcn10_dpp *dpp = TO_DCN10_DPP(dpp_base);
 	uint32_t ocsc_mode = 4;
+
 	dpp1_cm_program_color_matrix(dpp, regval);
 	REG_SET(CM_OCSC_CONTROL, 0, CM_OCSC_MODE, ocsc_mode);
 }
@@ -437,17 +439,18 @@ void dpp1_cm_program_regamma_lutb_settings(
 void dpp1_program_input_csc(
 		struct dpp *dpp_base,
 		enum dc_color_space color_space,
-		enum dcn10_input_csc_select select,
+		enum dcn10_input_csc_select input_select,
 		const struct out_csc_color_matrix *tbl_entry)
 {
 	struct dcn10_dpp *dpp = TO_DCN10_DPP(dpp_base);
 	int i;
 	int arr_size = sizeof(dcn10_input_csc_matrix)/sizeof(struct dcn10_input_csc_matrix);
 	const uint16_t *regval = NULL;
-	uint32_t selection = 1;
+	uint32_t cur_select = 0;
+	enum dcn10_input_csc_select select;
 	struct color_matrices_reg gam_regs;
 
-	if (select == INPUT_CSC_SELECT_BYPASS) {
+	if (input_select == INPUT_CSC_SELECT_BYPASS) {
 		REG_SET(CM_ICSC_CONTROL, 0, CM_ICSC_MODE, 0);
 		return;
 	}
@@ -467,36 +470,45 @@ void dpp1_program_input_csc(
 		regval = tbl_entry->regval;
 	}
 
-	if (select == INPUT_CSC_SELECT_COMA)
-		selection = 2;
-	REG_SET(CM_ICSC_CONTROL, 0,
-			CM_ICSC_MODE, selection);
+	/* determine which CSC matrix (icsc or coma) we are using
+	 * currently.  select the alternate set to double buffer
+	 * the CSC update so CSC is updated on frame boundary
+	 */
+	REG_SET(CM_TEST_DEBUG_INDEX, 0,
+			CM_TEST_DEBUG_INDEX, 9);
+
+	REG_GET(CM_TEST_DEBUG_DATA,
+			CM_TEST_DEBUG_DATA_ID9_ICSC_MODE, &cur_select);
+
+	if (cur_select != INPUT_CSC_SELECT_ICSC)
+		select = INPUT_CSC_SELECT_ICSC;
+	else
+		select = INPUT_CSC_SELECT_COMA;
 
 	gam_regs.shifts.csc_c11 = dpp->tf_shift->CM_ICSC_C11;
 	gam_regs.masks.csc_c11  = dpp->tf_mask->CM_ICSC_C11;
 	gam_regs.shifts.csc_c12 = dpp->tf_shift->CM_ICSC_C12;
 	gam_regs.masks.csc_c12 = dpp->tf_mask->CM_ICSC_C12;
 
-
 	if (select == INPUT_CSC_SELECT_ICSC) {
 
 		gam_regs.csc_c11_c12 = REG(CM_ICSC_C11_C12);
 		gam_regs.csc_c33_c34 = REG(CM_ICSC_C33_C34);
 
-		cm_helper_program_color_matrices(
-				dpp->base.ctx,
-				regval,
-				&gam_regs);
 	} else {
 
 		gam_regs.csc_c11_c12 = REG(CM_COMA_C11_C12);
 		gam_regs.csc_c33_c34 = REG(CM_COMA_C33_C34);
 
-		cm_helper_program_color_matrices(
-				dpp->base.ctx,
-				regval,
-				&gam_regs);
 	}
+
+	cm_helper_program_color_matrices(
+			dpp->base.ctx,
+			regval,
+			&gam_regs);
+
+	REG_SET(CM_ICSC_CONTROL, 0,
+				CM_ICSC_MODE, select);
 }
 
 //keep here for now, decide multi dce support later
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
index a3fe343..d321da9 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
@@ -319,12 +319,41 @@ static const struct dcn_dpp_registers tf_regs[] = {
 	tf_regs(3),
 };
 
+/*
+ *
+	DCN1 CM debug status register definition
+
+	register :ID9_CM_STATUS do
+	implement_ref :cm
+	map to:  :cmdebugind, at: j
+	width 32
+	disclosure   NEVER
+
+		field :ID9_VUPDATE_CFG, [0], R
+		field :ID9_IGAM_LUT_MODE, [2..1], R
+		field :ID9_BNS_BYPASS, [3], R
+		field :ID9_ICSC_MODE, [5..4], R
+		field :ID9_DGAM_LUT_MODE, [8..6], R
+		field :ID9_HDR_BYPASS, [9], R
+		field :ID9_GAMUT_REMAP_MODE, [11..10], R
+		field :ID9_RGAM_LUT_MODE, [14..12], R
+		#1 free bit
+		field :ID9_OCSC_MODE, [18..16], R
+		field :ID9_DENORM_MODE, [21..19], R
+		field :ID9_ROUND_TRUNC_MODE, [25..22], R
+		field :ID9_DITHER_EN, [26], R
+		field :ID9_DITHER_MODE, [28..27], R
+	end
+*/
+
 static const struct dcn_dpp_shift tf_shift = {
-	TF_REG_LIST_SH_MASK_DCN10(__SHIFT)
+	TF_REG_LIST_SH_MASK_DCN10(__SHIFT),
+	.CM_TEST_DEBUG_DATA_ID9_ICSC_MODE = 0x4
 };
 
 static const struct dcn_dpp_mask tf_mask = {
 	TF_REG_LIST_SH_MASK_DCN10(_MASK),
+	.CM_TEST_DEBUG_DATA_ID9_ICSC_MODE = 0x30
 };
 
 static const struct dcn_mpc_registers mpc_regs = {
-- 
cgit v1.1


From f412e8307d0ac6cbffd1240fb655557c126a0f2c Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Thu, 15 Mar 2018 13:31:14 -0400
Subject: drm/amd/display: Couple bug fixes in stats module

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/modules/stats/stats.c | 39 ++++++++++++++++-------
 1 file changed, 28 insertions(+), 11 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/modules/stats/stats.c b/drivers/gpu/drm/amd/display/modules/stats/stats.c
index 041f87b..ed5f680 100644
--- a/drivers/gpu/drm/amd/display/modules/stats/stats.c
+++ b/drivers/gpu/drm/amd/display/modules/stats/stats.c
@@ -187,7 +187,7 @@ void mod_stats_dump(struct mod_stats *mod_stats)
 
 		for (int i = 0; i < core_stats->index && i < core_stats->entries; i++) {
 			dm_logger_write(logger, LOG_PROFILING,
-					"%u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u\n",
+					"%u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u",
 					time[i].render_time_in_us,
 					time[i].avg_render_time_in_us_last_ten,
 					time[i].min_window,
@@ -227,7 +227,7 @@ void mod_stats_reset_data(struct mod_stats *mod_stats)
 	memset(core_stats->time, 0,
 		sizeof(struct stats_time_cache) * core_stats->entries);
 
-	core_stats->index = 0;
+	core_stats->index = 1;
 }
 
 void mod_stats_update_flip(struct mod_stats *mod_stats,
@@ -250,7 +250,7 @@ void mod_stats_update_flip(struct mod_stats *mod_stats,
 
 	time[index].flip_timestamp_in_ns = timestamp_in_ns;
 	time[index].render_time_in_us =
-		timestamp_in_ns - time[index - 1].flip_timestamp_in_ns;
+		(timestamp_in_ns - time[index - 1].flip_timestamp_in_ns) / 1000;
 
 	if (index >= 10) {
 		for (unsigned int i = 0; i < 10; i++)
@@ -261,10 +261,12 @@ void mod_stats_update_flip(struct mod_stats *mod_stats,
 
 	if (time[index].num_vsync_between_flips > 0)
 		time[index].vsync_to_flip_time_in_us =
-			timestamp_in_ns - time[index].vupdate_timestamp_in_ns;
+			(timestamp_in_ns -
+				time[index].vupdate_timestamp_in_ns) / 1000;
 	else
 		time[index].vsync_to_flip_time_in_us =
-			timestamp_in_ns - time[index - 1].vupdate_timestamp_in_ns;
+			(timestamp_in_ns -
+				time[index - 1].vupdate_timestamp_in_ns) / 1000;
 
 	core_stats->index++;
 }
@@ -275,6 +277,8 @@ void mod_stats_update_vupdate(struct mod_stats *mod_stats,
 	struct core_stats *core_stats = NULL;
 	struct stats_time_cache *time = NULL;
 	unsigned int index = 0;
+	unsigned int num_vsyncs = 0;
+	unsigned int prev_vsync_in_ns = 0;
 
 	if (mod_stats == NULL)
 		return;
@@ -286,14 +290,27 @@ void mod_stats_update_vupdate(struct mod_stats *mod_stats,
 
 	time = core_stats->time;
 	index = core_stats->index;
+	num_vsyncs = time[index].num_vsync_between_flips;
+
+	if (num_vsyncs < MOD_STATS_NUM_VSYNCS) {
+		if (num_vsyncs == 0) {
+			prev_vsync_in_ns =
+				time[index - 1].vupdate_timestamp_in_ns;
+
+			time[index].flip_to_vsync_time_in_us =
+				(timestamp_in_ns -
+					time[index - 1].flip_timestamp_in_ns) /
+					1000;
+		} else {
+			prev_vsync_in_ns =
+				time[index].vupdate_timestamp_in_ns;
+		}
 
-	time[index].vupdate_timestamp_in_ns = timestamp_in_ns;
-	if (time[index].num_vsync_between_flips < MOD_STATS_NUM_VSYNCS)
-		time[index].v_sync_time_in_us[time[index].num_vsync_between_flips] =
-			timestamp_in_ns - time[index - 1].vupdate_timestamp_in_ns;
-	time[index].flip_to_vsync_time_in_us =
-		timestamp_in_ns - time[index - 1].flip_timestamp_in_ns;
+		time[index].v_sync_time_in_us[num_vsyncs] =
+			(timestamp_in_ns - prev_vsync_in_ns) / 1000;
+	}
 
+	time[index].vupdate_timestamp_in_ns = timestamp_in_ns;
 	time[index].num_vsync_between_flips++;
 }
 
-- 
cgit v1.1


From e09b6473c605119a5f7c451a93a9e812e216a824 Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Thu, 15 Mar 2018 14:18:18 -0400
Subject: drm/amd/display: Rename encoder_info_packet to dc_info_packet

Move this out of the HW includes to dc_types.h

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_resource.c  | 99 +++++++++++-----------
 drivers/gpu/drm/amd/display/dc/dc_types.h          |  9 ++
 .../drm/amd/display/dc/dce/dce_stream_encoder.c    |  4 +-
 .../gpu/drm/amd/display/dc/inc/hw/stream_encoder.h | 21 ++---
 4 files changed, 65 insertions(+), 68 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
index cae78ee..379b055 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
@@ -1811,7 +1811,7 @@ enum dc_status dc_validate_global_state(
 }
 
 static void patch_gamut_packet_checksum(
-		struct encoder_info_packet *gamut_packet)
+		struct dc_info_packet *gamut_packet)
 {
 	/* For gamut we recalc checksum */
 	if (gamut_packet->valid) {
@@ -1830,12 +1830,11 @@ static void patch_gamut_packet_checksum(
 }
 
 static void set_avi_info_frame(
-		struct encoder_info_packet *info_packet,
+		struct dc_info_packet *info_packet,
 		struct pipe_ctx *pipe_ctx)
 {
 	struct dc_stream_state *stream = pipe_ctx->stream;
 	enum dc_color_space color_space = COLOR_SPACE_UNKNOWN;
-	struct info_frame info_frame = { {0} };
 	uint32_t pixel_encoding = 0;
 	enum scanning_type scan_type = SCANNING_TYPE_NODATA;
 	enum dc_aspect_ratio aspect = ASPECT_RATIO_NO_DATA;
@@ -1845,7 +1844,7 @@ static void set_avi_info_frame(
 	unsigned int cn0_cn1_value = 0;
 	uint8_t *check_sum = NULL;
 	uint8_t byte_index = 0;
-	union hdmi_info_packet *hdmi_info = &info_frame.avi_info_packet.info_packet_hdmi;
+	union hdmi_info_packet hdmi_info = {0};
 	union display_content_support support = {0};
 	unsigned int vic = pipe_ctx->stream->timing.vic;
 	enum dc_timing_3d_format format;
@@ -1856,11 +1855,11 @@ static void set_avi_info_frame(
 			COLOR_SPACE_SRGB:COLOR_SPACE_YCBCR709;
 
 	/* Initialize header */
-	hdmi_info->bits.header.info_frame_type = HDMI_INFOFRAME_TYPE_AVI;
+	hdmi_info.bits.header.info_frame_type = HDMI_INFOFRAME_TYPE_AVI;
 	/* InfoFrameVersion_3 is defined by CEA861F (Section 6.4), but shall
 	* not be used in HDMI 2.0 (Section 10.1) */
-	hdmi_info->bits.header.version = 2;
-	hdmi_info->bits.header.length = HDMI_AVI_INFOFRAME_SIZE;
+	hdmi_info.bits.header.version = 2;
+	hdmi_info.bits.header.length = HDMI_AVI_INFOFRAME_SIZE;
 
 	/*
 	 * IDO-defined (Y2,Y1,Y0 = 1,1,1) shall not be used by devices built
@@ -1886,39 +1885,39 @@ static void set_avi_info_frame(
 
 	/* Y0_Y1_Y2 : The pixel encoding */
 	/* H14b AVI InfoFrame has extension on Y-field from 2 bits to 3 bits */
-	hdmi_info->bits.Y0_Y1_Y2 = pixel_encoding;
+	hdmi_info.bits.Y0_Y1_Y2 = pixel_encoding;
 
 	/* A0 = 1 Active Format Information valid */
-	hdmi_info->bits.A0 = ACTIVE_FORMAT_VALID;
+	hdmi_info.bits.A0 = ACTIVE_FORMAT_VALID;
 
 	/* B0, B1 = 3; Bar info data is valid */
-	hdmi_info->bits.B0_B1 = BAR_INFO_BOTH_VALID;
+	hdmi_info.bits.B0_B1 = BAR_INFO_BOTH_VALID;
 
-	hdmi_info->bits.SC0_SC1 = PICTURE_SCALING_UNIFORM;
+	hdmi_info.bits.SC0_SC1 = PICTURE_SCALING_UNIFORM;
 
 	/* S0, S1 : Underscan / Overscan */
 	/* TODO: un-hardcode scan type */
 	scan_type = SCANNING_TYPE_UNDERSCAN;
-	hdmi_info->bits.S0_S1 = scan_type;
+	hdmi_info.bits.S0_S1 = scan_type;
 
 	/* C0, C1 : Colorimetry */
 	if (color_space == COLOR_SPACE_YCBCR709 ||
 			color_space == COLOR_SPACE_YCBCR709_LIMITED)
-		hdmi_info->bits.C0_C1 = COLORIMETRY_ITU709;
+		hdmi_info.bits.C0_C1 = COLORIMETRY_ITU709;
 	else if (color_space == COLOR_SPACE_YCBCR601 ||
 			color_space == COLOR_SPACE_YCBCR601_LIMITED)
-		hdmi_info->bits.C0_C1 = COLORIMETRY_ITU601;
+		hdmi_info.bits.C0_C1 = COLORIMETRY_ITU601;
 	else {
-		hdmi_info->bits.C0_C1 = COLORIMETRY_NO_DATA;
+		hdmi_info.bits.C0_C1 = COLORIMETRY_NO_DATA;
 	}
 	if (color_space == COLOR_SPACE_2020_RGB_FULLRANGE ||
 			color_space == COLOR_SPACE_2020_RGB_LIMITEDRANGE ||
 			color_space == COLOR_SPACE_2020_YCBCR) {
-		hdmi_info->bits.EC0_EC2 = COLORIMETRYEX_BT2020RGBYCBCR;
-		hdmi_info->bits.C0_C1   = COLORIMETRY_EXTENDED;
+		hdmi_info.bits.EC0_EC2 = COLORIMETRYEX_BT2020RGBYCBCR;
+		hdmi_info.bits.C0_C1   = COLORIMETRY_EXTENDED;
 	} else if (color_space == COLOR_SPACE_ADOBERGB) {
-		hdmi_info->bits.EC0_EC2 = COLORIMETRYEX_ADOBERGB;
-		hdmi_info->bits.C0_C1   = COLORIMETRY_EXTENDED;
+		hdmi_info.bits.EC0_EC2 = COLORIMETRYEX_ADOBERGB;
+		hdmi_info.bits.C0_C1   = COLORIMETRY_EXTENDED;
 	}
 
 	/* TODO: un-hardcode aspect ratio */
@@ -1927,18 +1926,18 @@ static void set_avi_info_frame(
 	switch (aspect) {
 	case ASPECT_RATIO_4_3:
 	case ASPECT_RATIO_16_9:
-		hdmi_info->bits.M0_M1 = aspect;
+		hdmi_info.bits.M0_M1 = aspect;
 		break;
 
 	case ASPECT_RATIO_NO_DATA:
 	case ASPECT_RATIO_64_27:
 	case ASPECT_RATIO_256_135:
 	default:
-		hdmi_info->bits.M0_M1 = 0;
+		hdmi_info.bits.M0_M1 = 0;
 	}
 
 	/* Active Format Aspect ratio - same as Picture Aspect Ratio. */
-	hdmi_info->bits.R0_R3 = ACTIVE_FORMAT_ASPECT_RATIO_SAME_AS_PICTURE;
+	hdmi_info.bits.R0_R3 = ACTIVE_FORMAT_ASPECT_RATIO_SAME_AS_PICTURE;
 
 	/* TODO: un-hardcode cn0_cn1 and itc */
 
@@ -1981,8 +1980,8 @@ static void set_avi_info_frame(
 				}
 			}
 		}
-		hdmi_info->bits.CN0_CN1 = cn0_cn1_value;
-		hdmi_info->bits.ITC = itc_value;
+		hdmi_info.bits.CN0_CN1 = cn0_cn1_value;
+		hdmi_info.bits.ITC = itc_value;
 	}
 
 	/* TODO : We should handle YCC quantization */
@@ -1991,19 +1990,19 @@ static void set_avi_info_frame(
 			stream->sink->edid_caps.qy_bit == 1) {
 		if (color_space == COLOR_SPACE_SRGB ||
 			color_space == COLOR_SPACE_2020_RGB_FULLRANGE) {
-			hdmi_info->bits.Q0_Q1   = RGB_QUANTIZATION_FULL_RANGE;
-			hdmi_info->bits.YQ0_YQ1 = YYC_QUANTIZATION_FULL_RANGE;
+			hdmi_info.bits.Q0_Q1   = RGB_QUANTIZATION_FULL_RANGE;
+			hdmi_info.bits.YQ0_YQ1 = YYC_QUANTIZATION_FULL_RANGE;
 		} else if (color_space == COLOR_SPACE_SRGB_LIMITED ||
 					color_space == COLOR_SPACE_2020_RGB_LIMITEDRANGE) {
-			hdmi_info->bits.Q0_Q1   = RGB_QUANTIZATION_LIMITED_RANGE;
-			hdmi_info->bits.YQ0_YQ1 = YYC_QUANTIZATION_LIMITED_RANGE;
+			hdmi_info.bits.Q0_Q1   = RGB_QUANTIZATION_LIMITED_RANGE;
+			hdmi_info.bits.YQ0_YQ1 = YYC_QUANTIZATION_LIMITED_RANGE;
 		} else {
-			hdmi_info->bits.Q0_Q1   = RGB_QUANTIZATION_DEFAULT_RANGE;
-			hdmi_info->bits.YQ0_YQ1 = YYC_QUANTIZATION_LIMITED_RANGE;
+			hdmi_info.bits.Q0_Q1   = RGB_QUANTIZATION_DEFAULT_RANGE;
+			hdmi_info.bits.YQ0_YQ1 = YYC_QUANTIZATION_LIMITED_RANGE;
 		}
 	} else {
-		hdmi_info->bits.Q0_Q1   = RGB_QUANTIZATION_DEFAULT_RANGE;
-		hdmi_info->bits.YQ0_YQ1   = YYC_QUANTIZATION_LIMITED_RANGE;
+		hdmi_info.bits.Q0_Q1   = RGB_QUANTIZATION_DEFAULT_RANGE;
+		hdmi_info.bits.YQ0_YQ1   = YYC_QUANTIZATION_LIMITED_RANGE;
 	}
 
 	///VIC
@@ -2028,51 +2027,49 @@ static void set_avi_info_frame(
 			break;
 		}
 	}
-	hdmi_info->bits.VIC0_VIC7 = vic;
+	hdmi_info.bits.VIC0_VIC7 = vic;
 
 	/* pixel repetition
 	 * PR0 - PR3 start from 0 whereas pHwPathMode->mode.timing.flags.pixel
 	 * repetition start from 1 */
-	hdmi_info->bits.PR0_PR3 = 0;
+	hdmi_info.bits.PR0_PR3 = 0;
 
 	/* Bar Info
 	 * barTop:    Line Number of End of Top Bar.
 	 * barBottom: Line Number of Start of Bottom Bar.
 	 * barLeft:   Pixel Number of End of Left Bar.
 	 * barRight:  Pixel Number of Start of Right Bar. */
-	hdmi_info->bits.bar_top = stream->timing.v_border_top;
-	hdmi_info->bits.bar_bottom = (stream->timing.v_total
+	hdmi_info.bits.bar_top = stream->timing.v_border_top;
+	hdmi_info.bits.bar_bottom = (stream->timing.v_total
 			- stream->timing.v_border_bottom + 1);
-	hdmi_info->bits.bar_left  = stream->timing.h_border_left;
-	hdmi_info->bits.bar_right = (stream->timing.h_total
+	hdmi_info.bits.bar_left  = stream->timing.h_border_left;
+	hdmi_info.bits.bar_right = (stream->timing.h_total
 			- stream->timing.h_border_right + 1);
 
 	/* check_sum - Calculate AFMT_AVI_INFO0 ~ AFMT_AVI_INFO3 */
-	check_sum = &info_frame.avi_info_packet.info_packet_hdmi.packet_raw_data.sb[0];
+	check_sum = &hdmi_info.packet_raw_data.sb[0];
 
 	*check_sum = HDMI_INFOFRAME_TYPE_AVI + HDMI_AVI_INFOFRAME_SIZE + 2;
 
 	for (byte_index = 1; byte_index <= HDMI_AVI_INFOFRAME_SIZE; byte_index++)
-		*check_sum += hdmi_info->packet_raw_data.sb[byte_index];
+		*check_sum += hdmi_info.packet_raw_data.sb[byte_index];
 
 	/* one byte complement */
 	*check_sum = (uint8_t) (0x100 - *check_sum);
 
 	/* Store in hw_path_mode */
-	info_packet->hb0 = hdmi_info->packet_raw_data.hb0;
-	info_packet->hb1 = hdmi_info->packet_raw_data.hb1;
-	info_packet->hb2 = hdmi_info->packet_raw_data.hb2;
+	info_packet->hb0 = hdmi_info.packet_raw_data.hb0;
+	info_packet->hb1 = hdmi_info.packet_raw_data.hb1;
+	info_packet->hb2 = hdmi_info.packet_raw_data.hb2;
 
-	for (byte_index = 0; byte_index < sizeof(info_frame.avi_info_packet.
-				info_packet_hdmi.packet_raw_data.sb); byte_index++)
-		info_packet->sb[byte_index] = info_frame.avi_info_packet.
-				info_packet_hdmi.packet_raw_data.sb[byte_index];
+	for (byte_index = 0; byte_index < sizeof(hdmi_info.packet_raw_data.sb); byte_index++)
+		info_packet->sb[byte_index] = hdmi_info.packet_raw_data.sb[byte_index];
 
 	info_packet->valid = true;
 }
 
 static void set_vendor_info_packet(
-		struct encoder_info_packet *info_packet,
+		struct dc_info_packet *info_packet,
 		struct dc_stream_state *stream)
 {
 	uint32_t length = 0;
@@ -2185,7 +2182,7 @@ static void set_vendor_info_packet(
 }
 
 static void set_spd_info_packet(
-		struct encoder_info_packet *info_packet,
+		struct dc_info_packet *info_packet,
 		struct dc_stream_state *stream)
 {
 	/* SPD info packet for FreeSync */
@@ -2306,7 +2303,7 @@ static void set_spd_info_packet(
 }
 
 static void set_hdr_static_info_packet(
-		struct encoder_info_packet *info_packet,
+		struct dc_info_packet *info_packet,
 		struct dc_stream_state *stream)
 {
 	uint16_t i = 0;
@@ -2403,7 +2400,7 @@ static void set_hdr_static_info_packet(
 }
 
 static void set_vsc_info_packet(
-		struct encoder_info_packet *info_packet,
+		struct dc_info_packet *info_packet,
 		struct dc_stream_state *stream)
 {
 	unsigned int vscPacketRevision = 0;
diff --git a/drivers/gpu/drm/amd/display/dc/dc_types.h b/drivers/gpu/drm/amd/display/dc/dc_types.h
index 9441305..cd324bc 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_types.h
@@ -530,6 +530,15 @@ struct vrr_params {
 	uint32_t frame_counter;
 };
 
+struct dc_info_packet {
+	bool valid;
+	uint8_t hb0;
+	uint8_t hb1;
+	uint8_t hb2;
+	uint8_t hb3;
+	uint8_t sb[32];
+};
+
 #define DC_PLANE_UPDATE_TIMES_MAX 10
 
 struct dc_plane_flip_time {
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
index 444558c..b85fda5 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
@@ -80,7 +80,7 @@ enum {
 static void dce110_update_generic_info_packet(
 	struct dce110_stream_encoder *enc110,
 	uint32_t packet_index,
-	const struct encoder_info_packet *info_packet)
+	const struct dc_info_packet *info_packet)
 {
 	uint32_t regval;
 	/* TODOFPGA Figure out a proper number for max_retries polling for lock
@@ -196,7 +196,7 @@ static void dce110_update_generic_info_packet(
 static void dce110_update_hdmi_info_packet(
 	struct dce110_stream_encoder *enc110,
 	uint32_t packet_index,
-	const struct encoder_info_packet *info_packet)
+	const struct dc_info_packet *info_packet)
 {
 	uint32_t cont, send, line;
 
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h b/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h
index b5db169..5c21336 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h
@@ -34,26 +34,17 @@ struct dc_bios;
 struct dc_context;
 struct dc_crtc_timing;
 
-struct encoder_info_packet {
-	bool valid;
-	uint8_t hb0;
-	uint8_t hb1;
-	uint8_t hb2;
-	uint8_t hb3;
-	uint8_t sb[32];
-};
-
 struct encoder_info_frame {
 	/* auxiliary video information */
-	struct encoder_info_packet avi;
-	struct encoder_info_packet gamut;
-	struct encoder_info_packet vendor;
+	struct dc_info_packet avi;
+	struct dc_info_packet gamut;
+	struct dc_info_packet vendor;
 	/* source product description */
-	struct encoder_info_packet spd;
+	struct dc_info_packet spd;
 	/* video stream configuration */
-	struct encoder_info_packet vsc;
+	struct dc_info_packet vsc;
 	/* HDR Static MetaData */
-	struct encoder_info_packet hdrsmd;
+	struct dc_info_packet hdrsmd;
 };
 
 struct encoder_unblank_param {
-- 
cgit v1.1


From 6e5b3587dbf6aaf7f9eef4956a3fce12bf7e9ffa Mon Sep 17 00:00:00 2001
From: SivapiriyanKumarasamy <sivapiriyan.kumarasamy@amd.com>
Date: Wed, 14 Mar 2018 09:15:24 -0400
Subject: drm/amd/display: Add vline IRQ programming for DCN

Signed-off-by: SivapiriyanKumarasamy <sivapiriyan.kumarasamy@amd.com>
Reviewed-by: Anthony Koo <Anthony.Koo@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc.c           |  6 ++
 drivers/gpu/drm/amd/display/dc/dc_stream.h         |  4 ++
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c  | 76 ++++++++++++++++++++++
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h  | 14 ++++
 .../drm/amd/display/dc/inc/hw/timing_generator.h   |  3 +
 drivers/gpu/drm/amd/display/dc/irq_types.h         |  7 ++
 6 files changed, 110 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c
index 63a3d46..554cf97 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc.c
@@ -1375,6 +1375,12 @@ static void commit_planes_for_stream(struct dc *dc,
 					pipe_ctx->stream_res.abm->funcs->set_abm_level(
 							pipe_ctx->stream_res.abm, stream->abm_level);
 			}
+
+			if (stream_update && stream_update->periodic_fn_vsync_delta &&
+					pipe_ctx->stream_res.tg->funcs->program_vline_interrupt)
+				pipe_ctx->stream_res.tg->funcs->program_vline_interrupt(
+						pipe_ctx->stream_res.tg, &pipe_ctx->stream->timing,
+						pipe_ctx->stream->periodic_fn_vsync_delta);
 		}
 	}
 
diff --git a/drivers/gpu/drm/amd/display/dc/dc_stream.h b/drivers/gpu/drm/amd/display/dc/dc_stream.h
index 3a7093e..8d51610 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_stream.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_stream.h
@@ -70,6 +70,9 @@ struct dc_stream_state {
 	enum color_transfer_func output_tf;
 
 	bool ignore_msa_timing_param;
+
+	unsigned long long periodic_fn_vsync_delta;
+
 	/* TODO: custom INFO packets */
 	/* TODO: ABM info (DMCU) */
 	/* PSR info */
@@ -113,6 +116,7 @@ struct dc_stream_update {
 	struct dc_hdr_static_metadata *hdr_static_metadata;
 	enum color_transfer_func color_output_tf;
 	unsigned int *abm_level;
+	unsigned long long *periodic_fn_vsync_delta;
 };
 
 bool dc_is_stream_unchanged(
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c
index 4bf64d1..f56eac0 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c
@@ -93,6 +93,81 @@ static void optc1_disable_stereo(struct timing_generator *optc)
 		OTG_3D_STRUCTURE_STEREO_SEL_OVR, 0);
 }
 
+static uint32_t get_start_vline(struct timing_generator *optc, const struct dc_crtc_timing *dc_crtc_timing)
+{
+	struct dc_crtc_timing patched_crtc_timing;
+	uint32_t vesa_sync_start;
+	uint32_t asic_blank_end;
+	uint32_t interlace_factor;
+	uint32_t vertical_line_start;
+
+	patched_crtc_timing = *dc_crtc_timing;
+	optc1_apply_front_porch_workaround(optc, &patched_crtc_timing);
+
+	vesa_sync_start = patched_crtc_timing.h_addressable +
+			patched_crtc_timing.h_border_right +
+			patched_crtc_timing.h_front_porch;
+
+	asic_blank_end = patched_crtc_timing.h_total -
+			vesa_sync_start -
+			patched_crtc_timing.h_border_left;
+
+	interlace_factor = patched_crtc_timing.flags.INTERLACE ? 2 : 1;
+
+	vesa_sync_start = patched_crtc_timing.v_addressable +
+			patched_crtc_timing.v_border_bottom +
+			patched_crtc_timing.v_front_porch;
+
+	asic_blank_end = (patched_crtc_timing.v_total -
+			vesa_sync_start -
+			patched_crtc_timing.v_border_top)
+			* interlace_factor;
+
+	vertical_line_start = asic_blank_end - optc->dlg_otg_param.vstartup_start + 1;
+	if (vertical_line_start < 0) {
+		ASSERT(0);
+		vertical_line_start = 0;
+	}
+
+	return vertical_line_start;
+}
+
+void optc1_program_vline_interrupt(
+		struct timing_generator *optc,
+		const struct dc_crtc_timing *dc_crtc_timing,
+		unsigned long long vsync_delta)
+{
+
+	struct optc *optc1 = DCN10TG_FROM_TG(optc);
+
+	unsigned long long req_delta_tens_of_usec = div64_u64((vsync_delta + 9999), 10000);
+	unsigned long long pix_clk_hundreds_khz = div64_u64((dc_crtc_timing->pix_clk_khz + 99), 100);
+	uint32_t req_delta_lines = (uint32_t) div64_u64(
+			(req_delta_tens_of_usec * pix_clk_hundreds_khz + dc_crtc_timing->h_total - 1),
+								dc_crtc_timing->h_total);
+
+	uint32_t vsync_line = get_start_vline(optc, dc_crtc_timing);
+	uint32_t start_line = 0;
+	uint32_t endLine = 0;
+
+	if (req_delta_lines != 0)
+		req_delta_lines--;
+
+	if (req_delta_lines > vsync_line)
+		start_line = dc_crtc_timing->v_total - (req_delta_lines - vsync_line) - 1;
+	else
+		start_line = vsync_line - req_delta_lines;
+
+	endLine = start_line + 2;
+
+	if (endLine >= dc_crtc_timing->v_total)
+		endLine = 2;
+
+	REG_SET_2(OTG_VERTICAL_INTERRUPT0_POSITION, 0,
+			OTG_VERTICAL_INTERRUPT0_LINE_START, start_line,
+			OTG_VERTICAL_INTERRUPT0_LINE_END, endLine);
+}
+
 /**
  * program_timing_generator   used by mode timing set
  * Program CRTC Timing Registers - OTG_H_*, OTG_V_*, Pixel repetition.
@@ -1215,6 +1290,7 @@ static bool optc1_is_optc_underflow_occurred(struct timing_generator *optc)
 static const struct timing_generator_funcs dcn10_tg_funcs = {
 		.validate_timing = optc1_validate_timing,
 		.program_timing = optc1_program_timing,
+		.program_vline_interrupt = optc1_program_vline_interrupt,
 		.program_global_sync = optc1_program_global_sync,
 		.enable_crtc = optc1_enable_crtc,
 		.disable_crtc = optc1_disable_crtc,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h
index d25e7bf..5a9a73d 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h
@@ -65,6 +65,8 @@
 	SRI(OTG_NOM_VERT_POSITION, OTG, inst),\
 	SRI(OTG_BLACK_COLOR, OTG, inst),\
 	SRI(OTG_CLOCK_CONTROL, OTG, inst),\
+	SRI(OTG_VERTICAL_INTERRUPT0_CONTROL, OTG, inst),\
+	SRI(OTG_VERTICAL_INTERRUPT0_POSITION, OTG, inst),\
 	SRI(OTG_VERTICAL_INTERRUPT2_CONTROL, OTG, inst),\
 	SRI(OTG_VERTICAL_INTERRUPT2_POSITION, OTG, inst),\
 	SRI(OPTC_INPUT_CLOCK_CONTROL, ODM, inst),\
@@ -124,6 +126,8 @@ struct dcn_optc_registers {
 	uint32_t OTG_TEST_PATTERN_CONTROL;
 	uint32_t OTG_TEST_PATTERN_COLOR;
 	uint32_t OTG_CLOCK_CONTROL;
+	uint32_t OTG_VERTICAL_INTERRUPT0_CONTROL;
+	uint32_t OTG_VERTICAL_INTERRUPT0_POSITION;
 	uint32_t OTG_VERTICAL_INTERRUPT2_CONTROL;
 	uint32_t OTG_VERTICAL_INTERRUPT2_POSITION;
 	uint32_t OPTC_INPUT_CLOCK_CONTROL;
@@ -206,6 +210,9 @@ struct dcn_optc_registers {
 	SF(OTG0_OTG_CLOCK_CONTROL, OTG_CLOCK_EN, mask_sh),\
 	SF(OTG0_OTG_CLOCK_CONTROL, OTG_CLOCK_ON, mask_sh),\
 	SF(OTG0_OTG_CLOCK_CONTROL, OTG_CLOCK_GATE_DIS, mask_sh),\
+	SF(OTG0_OTG_VERTICAL_INTERRUPT0_CONTROL, OTG_VERTICAL_INTERRUPT0_INT_ENABLE, mask_sh),\
+	SF(OTG0_OTG_VERTICAL_INTERRUPT0_POSITION, OTG_VERTICAL_INTERRUPT0_LINE_START, mask_sh),\
+	SF(OTG0_OTG_VERTICAL_INTERRUPT0_POSITION, OTG_VERTICAL_INTERRUPT0_LINE_END, mask_sh),\
 	SF(OTG0_OTG_VERTICAL_INTERRUPT2_CONTROL, OTG_VERTICAL_INTERRUPT2_INT_ENABLE, mask_sh),\
 	SF(OTG0_OTG_VERTICAL_INTERRUPT2_POSITION, OTG_VERTICAL_INTERRUPT2_LINE_START, mask_sh),\
 	SF(ODM0_OPTC_INPUT_CLOCK_CONTROL, OPTC_INPUT_CLK_EN, mask_sh),\
@@ -323,6 +330,9 @@ struct dcn_optc_registers {
 	type OTG_CLOCK_EN;\
 	type OTG_CLOCK_ON;\
 	type OTG_CLOCK_GATE_DIS;\
+	type OTG_VERTICAL_INTERRUPT0_INT_ENABLE;\
+	type OTG_VERTICAL_INTERRUPT0_LINE_START;\
+	type OTG_VERTICAL_INTERRUPT0_LINE_END;\
 	type OTG_VERTICAL_INTERRUPT2_INT_ENABLE;\
 	type OTG_VERTICAL_INTERRUPT2_LINE_START;\
 	type OPTC_INPUT_CLK_EN;\
@@ -420,6 +430,10 @@ void optc1_program_timing(
 	const struct dc_crtc_timing *dc_crtc_timing,
 	bool use_vbios);
 
+void optc1_program_vline_interrupt(struct timing_generator *optc,
+		const struct dc_crtc_timing *dc_crtc_timing,
+		unsigned long long vsync_delta);
+
 void optc1_program_global_sync(
 		struct timing_generator *optc);
 
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h b/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h
index 3217b5b..69cb0a1 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h
@@ -140,6 +140,9 @@ struct timing_generator_funcs {
 	void (*program_timing)(struct timing_generator *tg,
 							const struct dc_crtc_timing *timing,
 							bool use_vbios);
+	void (*program_vline_interrupt)(struct timing_generator *optc,
+			const struct dc_crtc_timing *dc_crtc_timing,
+			unsigned long long vsync_delta);
 	bool (*enable_crtc)(struct timing_generator *tg);
 	bool (*disable_crtc)(struct timing_generator *tg);
 	bool (*is_counter_moving)(struct timing_generator *tg);
diff --git a/drivers/gpu/drm/amd/display/dc/irq_types.h b/drivers/gpu/drm/amd/display/dc/irq_types.h
index a506c2e..cc3b1bc 100644
--- a/drivers/gpu/drm/amd/display/dc/irq_types.h
+++ b/drivers/gpu/drm/amd/display/dc/irq_types.h
@@ -135,6 +135,13 @@ enum dc_irq_source {
 	DC_IRQ_SOURCE_VBLANK5,
 	DC_IRQ_SOURCE_VBLANK6,
 
+	DC_IRQ_SOURCE_DC1_VLINE0,
+	DC_IRQ_SOURCE_DC2_VLINE0,
+	DC_IRQ_SOURCE_DC3_VLINE0,
+	DC_IRQ_SOURCE_DC4_VLINE0,
+	DC_IRQ_SOURCE_DC5_VLINE0,
+	DC_IRQ_SOURCE_DC6_VLINE0,
+
 	DAL_IRQ_SOURCES_NUMBER
 };
 
-- 
cgit v1.1


From 5813dd1c0c4e06b3321142cd2da99909a1f41707 Mon Sep 17 00:00:00 2001
From: Xingyue Tao <xingyue.tao@amd.com>
Date: Fri, 16 Mar 2018 15:20:48 -0400
Subject: drm/amd/display: Add double buffer machanism to OCSC

- Added double buffer mechanism to output CSC
so that there's no tearing when adjusting brightness
from Radeon settings

Signed-off-by: Xingyue Tao <xingyue.tao@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h   | 36 +++++++++++++++++
 .../gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c    | 46 +++++++++++++---------
 .../gpu/drm/amd/display/dc/dcn10/dcn10_resource.c  | 32 ++-------------
 3 files changed, 67 insertions(+), 47 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h
index b81b2aa..9b5ff76 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h
@@ -420,6 +420,41 @@
 	TF_SF(CURSOR0_CURSOR_CONTROL, CURSOR_ENABLE, mask_sh), \
 	TF_SF(DPP_TOP0_DPP_CONTROL, DPPCLK_RATE_CONTROL, mask_sh)
 
+/*
+ *
+	DCN1 CM debug status register definition
+
+	register :ID9_CM_STATUS do
+	implement_ref :cm
+	map to:  :cmdebugind, at: j
+	width 32
+	disclosure   NEVER
+
+		field :ID9_VUPDATE_CFG, [0], R
+		field :ID9_IGAM_LUT_MODE, [2..1], R
+		field :ID9_BNS_BYPASS, [3], R
+		field :ID9_ICSC_MODE, [5..4], R
+		field :ID9_DGAM_LUT_MODE, [8..6], R
+		field :ID9_HDR_BYPASS, [9], R
+		field :ID9_GAMUT_REMAP_MODE, [11..10], R
+		field :ID9_RGAM_LUT_MODE, [14..12], R
+		#1 free bit
+		field :ID9_OCSC_MODE, [18..16], R
+		field :ID9_DENORM_MODE, [21..19], R
+		field :ID9_ROUND_TRUNC_MODE, [25..22], R
+		field :ID9_DITHER_EN, [26], R
+		field :ID9_DITHER_MODE, [28..27], R
+	end
+*/
+
+#define TF_DEBUG_REG_LIST_SH_DCN10 \
+	.CM_TEST_DEBUG_DATA_ID9_ICSC_MODE = 4, \
+	.CM_TEST_DEBUG_DATA_ID9_OCSC_MODE = 16
+
+#define TF_DEBUG_REG_LIST_MASK_DCN10 \
+	.CM_TEST_DEBUG_DATA_ID9_ICSC_MODE = 0x30, \
+	.CM_TEST_DEBUG_DATA_ID9_OCSC_MODE = 0x70000
+
 #define TF_REG_FIELD_LIST(type) \
 	type EXT_OVERSCAN_LEFT; \
 	type EXT_OVERSCAN_RIGHT; \
@@ -1015,6 +1050,7 @@
 	type CM_BYPASS; \
 	type CM_TEST_DEBUG_INDEX; \
 	type CM_TEST_DEBUG_DATA_ID9_ICSC_MODE; \
+	type CM_TEST_DEBUG_DATA_ID9_OCSC_MODE;\
 	type FORMAT_CONTROL__ALPHA_EN; \
 	type CUR0_COLOR0; \
 	type CUR0_COLOR1; \
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c
index cc51141..4f373c9 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c
@@ -216,41 +216,55 @@ static void dpp1_cm_program_color_matrix(
 		struct dcn10_dpp *dpp,
 		const uint16_t *regval)
 {
-	uint32_t mode;
+	uint32_t ocsc_mode;
+	uint32_t cur_mode;
 	struct color_matrices_reg gam_regs;
 
-	REG_GET(CM_OCSC_CONTROL, CM_OCSC_MODE, &mode);
-
 	if (regval == NULL) {
 		BREAK_TO_DEBUGGER();
 		return;
 	}
-	mode = 4;
+
+	/* determine which CSC matrix (ocsc or comb) we are using
+	 * currently.  select the alternate set to double buffer
+	 * the CSC update so CSC is updated on frame boundary
+	 */
+	REG_SET(CM_TEST_DEBUG_INDEX, 0,
+			CM_TEST_DEBUG_INDEX, 9);
+
+	REG_GET(CM_TEST_DEBUG_DATA,
+			CM_TEST_DEBUG_DATA_ID9_OCSC_MODE, &cur_mode);
+
+	if (cur_mode != 4)
+		ocsc_mode = 4;
+	else
+		ocsc_mode = 5;
+
+
 	gam_regs.shifts.csc_c11 = dpp->tf_shift->CM_OCSC_C11;
 	gam_regs.masks.csc_c11  = dpp->tf_mask->CM_OCSC_C11;
 	gam_regs.shifts.csc_c12 = dpp->tf_shift->CM_OCSC_C12;
 	gam_regs.masks.csc_c12 = dpp->tf_mask->CM_OCSC_C12;
 
-	if (mode == 4) {
+	if (ocsc_mode == 4) {
 
 		gam_regs.csc_c11_c12 = REG(CM_OCSC_C11_C12);
 		gam_regs.csc_c33_c34 = REG(CM_OCSC_C33_C34);
 
-		cm_helper_program_color_matrices(
-				dpp->base.ctx,
-				regval,
-				&gam_regs);
-
 	} else {
 
 		gam_regs.csc_c11_c12 = REG(CM_COMB_C11_C12);
 		gam_regs.csc_c33_c34 = REG(CM_COMB_C33_C34);
 
-		cm_helper_program_color_matrices(
-				dpp->base.ctx,
-				regval,
-				&gam_regs);
 	}
+
+	cm_helper_program_color_matrices(
+			dpp->base.ctx,
+			regval,
+			&gam_regs);
+
+	REG_SET(CM_OCSC_CONTROL, 0, CM_OCSC_MODE, ocsc_mode);
+
 }
 
 void dpp1_cm_set_output_csc_default(
@@ -260,7 +274,6 @@ void dpp1_cm_set_output_csc_default(
 	struct dcn10_dpp *dpp = TO_DCN10_DPP(dpp_base);
 	const uint16_t *regval = NULL;
 	int arr_size;
-	uint32_t ocsc_mode = 4;
 
 	regval = find_color_matrix(colorspace, &arr_size);
 	if (regval == NULL) {
@@ -269,7 +282,6 @@ void dpp1_cm_set_output_csc_default(
 	}
 
 	dpp1_cm_program_color_matrix(dpp, regval);
-	REG_SET(CM_OCSC_CONTROL, 0, CM_OCSC_MODE, ocsc_mode);
 }
 
 static void dpp1_cm_get_reg_field(
@@ -330,10 +342,8 @@ void dpp1_cm_set_output_csc_adjustment(
 		const uint16_t *regval)
 {
 	struct dcn10_dpp *dpp = TO_DCN10_DPP(dpp_base);
-	uint32_t ocsc_mode = 4;
 
 	dpp1_cm_program_color_matrix(dpp, regval);
-	REG_SET(CM_OCSC_CONTROL, 0, CM_OCSC_MODE, ocsc_mode);
 }
 
 void dpp1_cm_power_on_regamma_lut(struct dpp *dpp_base,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
index d321da9..7ad290c 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
@@ -319,41 +319,15 @@ static const struct dcn_dpp_registers tf_regs[] = {
 	tf_regs(3),
 };
 
-/*
- *
-	DCN1 CM debug status register definition
-
-	register :ID9_CM_STATUS do
-	implement_ref :cm
-	map to:  :cmdebugind, at: j
-	width 32
-	disclosure   NEVER
-
-		field :ID9_VUPDATE_CFG, [0], R
-		field :ID9_IGAM_LUT_MODE, [2..1], R
-		field :ID9_BNS_BYPASS, [3], R
-		field :ID9_ICSC_MODE, [5..4], R
-		field :ID9_DGAM_LUT_MODE, [8..6], R
-		field :ID9_HDR_BYPASS, [9], R
-		field :ID9_GAMUT_REMAP_MODE, [11..10], R
-		field :ID9_RGAM_LUT_MODE, [14..12], R
-		#1 free bit
-		field :ID9_OCSC_MODE, [18..16], R
-		field :ID9_DENORM_MODE, [21..19], R
-		field :ID9_ROUND_TRUNC_MODE, [25..22], R
-		field :ID9_DITHER_EN, [26], R
-		field :ID9_DITHER_MODE, [28..27], R
-	end
-*/
-
 static const struct dcn_dpp_shift tf_shift = {
 	TF_REG_LIST_SH_MASK_DCN10(__SHIFT),
-	.CM_TEST_DEBUG_DATA_ID9_ICSC_MODE = 0x4
+	TF_DEBUG_REG_LIST_SH_DCN10
+
 };
 
 static const struct dcn_dpp_mask tf_mask = {
 	TF_REG_LIST_SH_MASK_DCN10(_MASK),
-	.CM_TEST_DEBUG_DATA_ID9_ICSC_MODE = 0x30
+	TF_DEBUG_REG_LIST_MASK_DCN10
 };
 
 static const struct dcn_mpc_registers mpc_regs = {
-- 
cgit v1.1


From 87943159f4093d2dae22abccbe046dac0fbdad4f Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Thu, 15 Mar 2018 14:54:30 -0400
Subject: drm/amd/display: Only program MSA_TIMING_PARAM if it changed

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_link.c | 15 ++++++++++-----
 1 file changed, 10 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
index 82ee9de..c18f24a 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
@@ -1183,16 +1183,21 @@ static void enable_stream_features(struct pipe_ctx *pipe_ctx)
 {
 	struct dc_stream_state *stream = pipe_ctx->stream;
 	struct dc_link *link = stream->sink->link;
-	union down_spread_ctrl downspread;
+	union down_spread_ctrl old_downspread;
+	union down_spread_ctrl new_downspread;
 
 	core_link_read_dpcd(link, DP_DOWNSPREAD_CTRL,
-			&downspread.raw, sizeof(downspread));
+			&old_downspread.raw, sizeof(old_downspread));
 
-	downspread.bits.IGNORE_MSA_TIMING_PARAM =
+	new_downspread.raw = old_downspread.raw;
+
+	new_downspread.bits.IGNORE_MSA_TIMING_PARAM =
 			(stream->ignore_msa_timing_param) ? 1 : 0;
 
-	core_link_write_dpcd(link, DP_DOWNSPREAD_CTRL,
-			&downspread.raw, sizeof(downspread));
+	if (new_downspread.raw != old_downspread.raw) {
+		core_link_write_dpcd(link, DP_DOWNSPREAD_CTRL,
+			&new_downspread.raw, sizeof(new_downspread));
+	}
 }
 
 static enum dc_status enable_link_dp(
-- 
cgit v1.1


From 44d09c6a577c8ed4e0ef50257487c071ae5e0fa2 Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Thu, 15 Mar 2018 14:29:24 -0400
Subject: drm/amd/display: Move commit_planes_to_stream to amdgpu_dm

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 108 +++++++++++++++++++++-
 drivers/gpu/drm/amd/display/dc/core/dc.c          |  89 ------------------
 drivers/gpu/drm/amd/display/dc/dc_stream.h        |   7 --
 3 files changed, 103 insertions(+), 101 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index bad9f09..3ff3905 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -3977,6 +3977,97 @@ static void amdgpu_dm_do_flip(struct drm_crtc *crtc,
 	spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
 }
 
+/*
+ * TODO this whole function needs to go
+ *
+ * dc_surface_update is needlessly complex. See if we can just replace this
+ * with a dc_plane_state and follow the atomic model a bit more closely here.
+ */
+static bool commit_planes_to_stream(
+		struct dc *dc,
+		struct dc_plane_state **plane_states,
+		uint8_t new_plane_count,
+		struct dm_crtc_state *dm_new_crtc_state,
+		struct dm_crtc_state *dm_old_crtc_state,
+		struct dc_state *state)
+{
+	/* no need to dynamically allocate this. it's pretty small */
+	struct dc_surface_update updates[MAX_SURFACES];
+	struct dc_flip_addrs *flip_addr;
+	struct dc_plane_info *plane_info;
+	struct dc_scaling_info *scaling_info;
+	int i;
+	struct dc_stream_state *dc_stream = dm_new_crtc_state->stream;
+	struct dc_stream_update *stream_update =
+			kzalloc(sizeof(struct dc_stream_update), GFP_KERNEL);
+
+	if (!stream_update) {
+		BREAK_TO_DEBUGGER();
+		return false;
+	}
+
+	flip_addr = kcalloc(MAX_SURFACES, sizeof(struct dc_flip_addrs),
+			    GFP_KERNEL);
+	plane_info = kcalloc(MAX_SURFACES, sizeof(struct dc_plane_info),
+			     GFP_KERNEL);
+	scaling_info = kcalloc(MAX_SURFACES, sizeof(struct dc_scaling_info),
+			       GFP_KERNEL);
+
+	if (!flip_addr || !plane_info || !scaling_info) {
+		kfree(flip_addr);
+		kfree(plane_info);
+		kfree(scaling_info);
+		kfree(stream_update);
+		return false;
+	}
+
+	memset(updates, 0, sizeof(updates));
+
+	stream_update->src = dc_stream->src;
+	stream_update->dst = dc_stream->dst;
+	stream_update->out_transfer_func = dc_stream->out_transfer_func;
+
+	for (i = 0; i < new_plane_count; i++) {
+		updates[i].surface = plane_states[i];
+		updates[i].gamma =
+			(struct dc_gamma *)plane_states[i]->gamma_correction;
+		updates[i].in_transfer_func = plane_states[i]->in_transfer_func;
+		flip_addr[i].address = plane_states[i]->address;
+		flip_addr[i].flip_immediate = plane_states[i]->flip_immediate;
+		plane_info[i].color_space = plane_states[i]->color_space;
+		plane_info[i].input_tf = plane_states[i]->input_tf;
+		plane_info[i].format = plane_states[i]->format;
+		plane_info[i].plane_size = plane_states[i]->plane_size;
+		plane_info[i].rotation = plane_states[i]->rotation;
+		plane_info[i].horizontal_mirror = plane_states[i]->horizontal_mirror;
+		plane_info[i].stereo_format = plane_states[i]->stereo_format;
+		plane_info[i].tiling_info = plane_states[i]->tiling_info;
+		plane_info[i].visible = plane_states[i]->visible;
+		plane_info[i].per_pixel_alpha = plane_states[i]->per_pixel_alpha;
+		plane_info[i].dcc = plane_states[i]->dcc;
+		scaling_info[i].scaling_quality = plane_states[i]->scaling_quality;
+		scaling_info[i].src_rect = plane_states[i]->src_rect;
+		scaling_info[i].dst_rect = plane_states[i]->dst_rect;
+		scaling_info[i].clip_rect = plane_states[i]->clip_rect;
+
+		updates[i].flip_addr = &flip_addr[i];
+		updates[i].plane_info = &plane_info[i];
+		updates[i].scaling_info = &scaling_info[i];
+	}
+
+	dc_commit_updates_for_stream(
+			dc,
+			updates,
+			new_plane_count,
+			dc_stream, stream_update, plane_states, state);
+
+	kfree(flip_addr);
+	kfree(plane_info);
+	kfree(scaling_info);
+	kfree(stream_update);
+	return true;
+}
+
 static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
 				    struct drm_device *dev,
 				    struct amdgpu_display_manager *dm,
@@ -3992,6 +4083,8 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
 	struct drm_crtc_state *new_pcrtc_state =
 			drm_atomic_get_new_crtc_state(state, pcrtc);
 	struct dm_crtc_state *acrtc_state = to_dm_crtc_state(new_pcrtc_state);
+	struct dm_crtc_state *dm_old_crtc_state =
+			to_dm_crtc_state(drm_atomic_get_old_crtc_state(state, pcrtc));
 	struct dm_atomic_state *dm_state = to_dm_atomic_state(state);
 	int planes_count = 0;
 	unsigned long flags;
@@ -4070,10 +4163,12 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
 			spin_unlock_irqrestore(&pcrtc->dev->event_lock, flags);
 		}
 
-		if (false == dc_commit_planes_to_stream(dm->dc,
+
+		if (false == commit_planes_to_stream(dm->dc,
 							plane_states_constructed,
 							planes_count,
-							dc_stream_attach,
+							acrtc_state,
+							dm_old_crtc_state,
 							dm_state->context))
 			dm_error("%s: Failed to attach plane!\n", __func__);
 	} else {
@@ -4298,8 +4393,10 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state)
 		struct amdgpu_crtc *acrtc = to_amdgpu_crtc(dm_new_con_state->base.crtc);
 		struct dc_stream_status *status = NULL;
 
-		if (acrtc)
+		if (acrtc) {
 			new_crtc_state = drm_atomic_get_new_crtc_state(state, &acrtc->base);
+			old_crtc_state = drm_atomic_get_old_crtc_state(state, &acrtc->base);
+		}
 
 		/* Skip any modesets/resets */
 		if (!acrtc || drm_atomic_crtc_needs_modeset(new_crtc_state))
@@ -4322,11 +4419,12 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state)
 		WARN_ON(!status->plane_count);
 
 		/*TODO How it works with MPO ?*/
-		if (!dc_commit_planes_to_stream(
+		if (!commit_planes_to_stream(
 				dm->dc,
 				status->plane_states,
 				status->plane_count,
-				dm_new_crtc_state->stream,
+				dm_new_crtc_state,
+				to_dm_crtc_state(old_crtc_state),
 				dm_state->context))
 			dm_error("%s: Failed to update stream scaling!\n", __func__);
 	}
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c
index 554cf97..6f4ad67 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc.c
@@ -936,95 +936,6 @@ bool dc_post_update_surfaces_to_stream(struct dc *dc)
 	return true;
 }
 
-/*
- * TODO this whole function needs to go
- *
- * dc_surface_update is needlessly complex. See if we can just replace this
- * with a dc_plane_state and follow the atomic model a bit more closely here.
- */
-bool dc_commit_planes_to_stream(
-		struct dc *dc,
-		struct dc_plane_state **plane_states,
-		uint8_t new_plane_count,
-		struct dc_stream_state *dc_stream,
-		struct dc_state *state)
-{
-	/* no need to dynamically allocate this. it's pretty small */
-	struct dc_surface_update updates[MAX_SURFACES];
-	struct dc_flip_addrs *flip_addr;
-	struct dc_plane_info *plane_info;
-	struct dc_scaling_info *scaling_info;
-	int i;
-	struct dc_stream_update *stream_update =
-			kzalloc(sizeof(struct dc_stream_update), GFP_KERNEL);
-
-	if (!stream_update) {
-		BREAK_TO_DEBUGGER();
-		return false;
-	}
-
-	flip_addr = kcalloc(MAX_SURFACES, sizeof(struct dc_flip_addrs),
-			    GFP_KERNEL);
-	plane_info = kcalloc(MAX_SURFACES, sizeof(struct dc_plane_info),
-			     GFP_KERNEL);
-	scaling_info = kcalloc(MAX_SURFACES, sizeof(struct dc_scaling_info),
-			       GFP_KERNEL);
-
-	if (!flip_addr || !plane_info || !scaling_info) {
-		kfree(flip_addr);
-		kfree(plane_info);
-		kfree(scaling_info);
-		kfree(stream_update);
-		return false;
-	}
-
-	memset(updates, 0, sizeof(updates));
-
-	stream_update->src = dc_stream->src;
-	stream_update->dst = dc_stream->dst;
-	stream_update->out_transfer_func = dc_stream->out_transfer_func;
-
-	for (i = 0; i < new_plane_count; i++) {
-		updates[i].surface = plane_states[i];
-		updates[i].gamma =
-			(struct dc_gamma *)plane_states[i]->gamma_correction;
-		updates[i].in_transfer_func = plane_states[i]->in_transfer_func;
-		flip_addr[i].address = plane_states[i]->address;
-		flip_addr[i].flip_immediate = plane_states[i]->flip_immediate;
-		plane_info[i].color_space = plane_states[i]->color_space;
-		plane_info[i].input_tf = plane_states[i]->input_tf;
-		plane_info[i].format = plane_states[i]->format;
-		plane_info[i].plane_size = plane_states[i]->plane_size;
-		plane_info[i].rotation = plane_states[i]->rotation;
-		plane_info[i].horizontal_mirror = plane_states[i]->horizontal_mirror;
-		plane_info[i].stereo_format = plane_states[i]->stereo_format;
-		plane_info[i].tiling_info = plane_states[i]->tiling_info;
-		plane_info[i].visible = plane_states[i]->visible;
-		plane_info[i].per_pixel_alpha = plane_states[i]->per_pixel_alpha;
-		plane_info[i].dcc = plane_states[i]->dcc;
-		scaling_info[i].scaling_quality = plane_states[i]->scaling_quality;
-		scaling_info[i].src_rect = plane_states[i]->src_rect;
-		scaling_info[i].dst_rect = plane_states[i]->dst_rect;
-		scaling_info[i].clip_rect = plane_states[i]->clip_rect;
-
-		updates[i].flip_addr = &flip_addr[i];
-		updates[i].plane_info = &plane_info[i];
-		updates[i].scaling_info = &scaling_info[i];
-	}
-
-	dc_commit_updates_for_stream(
-			dc,
-			updates,
-			new_plane_count,
-			dc_stream, stream_update, plane_states, state);
-
-	kfree(flip_addr);
-	kfree(plane_info);
-	kfree(scaling_info);
-	kfree(stream_update);
-	return true;
-}
-
 struct dc_state *dc_create_state(void)
 {
 	struct dc_state *context = kzalloc(sizeof(struct dc_state),
diff --git a/drivers/gpu/drm/amd/display/dc/dc_stream.h b/drivers/gpu/drm/amd/display/dc/dc_stream.h
index 8d51610..2971cd0 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_stream.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_stream.h
@@ -135,13 +135,6 @@ bool dc_is_stream_scaling_unchanged(
  *   This does not trigger a flip.  No surface address is programmed.
  */
 
-bool dc_commit_planes_to_stream(
-		struct dc *dc,
-		struct dc_plane_state **plane_states,
-		uint8_t new_plane_count,
-		struct dc_stream_state *dc_stream,
-		struct dc_state *state);
-
 void dc_commit_updates_for_stream(struct dc *dc,
 		struct dc_surface_update *srf_updates,
 		int surface_count,
-- 
cgit v1.1


From 844de65e9108a03f2018a6bb827cc53bfa71693b Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Fri, 23 Mar 2018 10:45:00 +0800
Subject: drm/amd/pp: Remove useless fw load error handler on Polaris

Acked-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c | 12 ++----------
 1 file changed, 2 insertions(+), 10 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c
index 997a777..fe6854e 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c
@@ -301,19 +301,11 @@ static int polaris10_start_smu(struct pp_hwmgr *hwmgr)
 		smu_data->smu7_data.security_hard_key = (uint8_t) (PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, SMU_FIRMWARE, SMU_SEL));
 
 		/* Check if SMU is running in protected mode */
-		if (smu_data->protected_mode == 0) {
+		if (smu_data->protected_mode == 0)
 			result = polaris10_start_smu_in_non_protection_mode(hwmgr);
-		} else {
+		else
 			result = polaris10_start_smu_in_protection_mode(hwmgr);
 
-			/* If failed, try with different security Key. */
-			if (result != 0) {
-				smu_data->smu7_data.security_hard_key ^= 1;
-				cgs_rel_firmware(hwmgr->device, CGS_UCODE_ID_SMU);
-				result = polaris10_start_smu_in_protection_mode(hwmgr);
-			}
-		}
-
 		if (result != 0)
 			PP_ASSERT_WITH_CODE(0, "Failed to load SMU ucode.", return result);
 
-- 
cgit v1.1


From ba8ab90e6ac9322f39ab8368941b38b5bb12477c Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Thu, 22 Mar 2018 14:52:35 +0800
Subject: drm/amd/pp: Add hwmgr_sw_init/fini functioins

Clean up pp ip functions

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/amd_powerplay.c      | 368 +++++----------------
 .../gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c  |  14 +-
 drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c        |  74 ++++-
 drivers/gpu/drm/amd/powerplay/inc/hwmgr.h          |   7 +-
 4 files changed, 147 insertions(+), 316 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
index 7e8ad30..6503bbf 100644
--- a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
+++ b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
@@ -31,24 +31,11 @@
 #include "amdgpu.h"
 #include "hwmgr.h"
 
-#define PP_DPM_DISABLED 0xCCCC
-
 static int pp_dpm_dispatch_tasks(void *handle, enum amd_pp_task task_id,
 		enum amd_pm_state_type *user_state);
 
 static const struct amd_pm_funcs pp_dpm_funcs;
 
-static inline int pp_check(struct pp_hwmgr *hwmgr)
-{
-	if (hwmgr == NULL || hwmgr->smumgr_funcs == NULL)
-		return -EINVAL;
-
-	if (hwmgr->pm_en == 0 || hwmgr->hwmgr_func == NULL)
-		return PP_DPM_DISABLED;
-
-	return 0;
-}
-
 static int amd_powerplay_create(struct amdgpu_device *adev)
 {
 	struct pp_hwmgr *hwmgr;
@@ -73,7 +60,7 @@ static int amd_powerplay_create(struct amdgpu_device *adev)
 }
 
 
-static int amd_powerplay_destroy(struct amdgpu_device *adev)
+static void amd_powerplay_destroy(struct amdgpu_device *adev)
 {
 	struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle;
 
@@ -82,8 +69,6 @@ static int amd_powerplay_destroy(struct amdgpu_device *adev)
 
 	kfree(hwmgr);
 	hwmgr = NULL;
-
-	return 0;
 }
 
 static int pp_early_init(void *handle)
@@ -109,18 +94,9 @@ static int pp_sw_init(void *handle)
 	struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret >= 0) {
-		if (hwmgr->smumgr_funcs->smu_init == NULL)
-			return -EINVAL;
+	ret = hwmgr_sw_init(hwmgr);
 
-		ret = hwmgr->smumgr_funcs->smu_init(hwmgr);
-
-		phm_register_irq_handlers(hwmgr);
-
-		pr_debug("amdgpu: powerplay sw initialized\n");
-	}
+	pr_debug("powerplay sw init %s\n", ret ? "failed" : "successfully");
 
 	return ret;
 }
@@ -129,13 +105,8 @@ static int pp_sw_fini(void *handle)
 {
 	struct amdgpu_device *adev = handle;
 	struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle;
-	int ret = 0;
 
-	ret = pp_check(hwmgr);
-	if (ret >= 0) {
-		if (hwmgr->smumgr_funcs->smu_fini != NULL)
-			hwmgr->smumgr_funcs->smu_fini(hwmgr);
-	}
+	hwmgr_sw_fini(hwmgr);
 
 	if (adev->firmware.load_type == AMDGPU_FW_LOAD_SMU)
 		amdgpu_ucode_fini_bo(adev);
@@ -152,40 +123,20 @@ static int pp_hw_init(void *handle)
 	if (adev->firmware.load_type == AMDGPU_FW_LOAD_SMU)
 		amdgpu_ucode_init_bo(adev);
 
-	ret = pp_check(hwmgr);
+	ret = hwmgr_hw_init(hwmgr);
 
-	if (ret >= 0) {
-		if (hwmgr->smumgr_funcs->start_smu == NULL)
-			return -EINVAL;
+	if (ret)
+		pr_err("powerplay hw init failed\n");
 
-		if (hwmgr->smumgr_funcs->start_smu(hwmgr)) {
-			pr_err("smc start failed\n");
-			hwmgr->smumgr_funcs->smu_fini(hwmgr);
-			return -EINVAL;
-		}
-		if (ret == PP_DPM_DISABLED)
-			goto exit;
-		ret = hwmgr_hw_init(hwmgr);
-		if (ret)
-			goto exit;
-	}
 	return ret;
-exit:
-	hwmgr->pm_en = 0;
-	cgs_notify_dpm_enabled(hwmgr->device, false);
-	return 0;
-
 }
 
 static int pp_hw_fini(void *handle)
 {
 	struct amdgpu_device *adev = handle;
 	struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle;
-	int ret = 0;
 
-	ret = pp_check(hwmgr);
-	if (ret == 0)
-		hwmgr_hw_fini(hwmgr);
+	hwmgr_hw_fini(hwmgr);
 
 	return 0;
 }
@@ -194,11 +145,8 @@ static int pp_late_init(void *handle)
 {
 	struct amdgpu_device *adev = handle;
 	struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle;
-	int ret = 0;
-
-	ret = pp_check(hwmgr);
 
-	if (ret == 0)
+	if (hwmgr && hwmgr->pm_en)
 		pp_dpm_dispatch_tasks(hwmgr,
 					AMD_PP_TASK_COMPLETE_INIT, NULL);
 
@@ -233,12 +181,9 @@ static int pp_set_powergating_state(void *handle,
 {
 	struct amdgpu_device *adev = handle;
 	struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle;
-	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return 0;
 
 	if (hwmgr->hwmgr_func->enable_per_cu_power_gating == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -254,38 +199,16 @@ static int pp_suspend(void *handle)
 {
 	struct amdgpu_device *adev = handle;
 	struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle;
-	int ret = 0;
 
-	ret = pp_check(hwmgr);
-	if (ret == 0)
-		hwmgr_hw_suspend(hwmgr);
-	return 0;
+	return hwmgr_suspend(hwmgr);
 }
 
 static int pp_resume(void *handle)
 {
 	struct amdgpu_device *adev = handle;
 	struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle;
-	int ret;
-
-	ret = pp_check(hwmgr);
-
-	if (ret < 0)
-		return ret;
-
-	if (hwmgr->smumgr_funcs->start_smu == NULL)
-		return -EINVAL;
-
-	if (hwmgr->smumgr_funcs->start_smu(hwmgr)) {
-		pr_err("smc start failed\n");
-		hwmgr->smumgr_funcs->smu_fini(hwmgr);
-		return -EINVAL;
-	}
-
-	if (ret == PP_DPM_DISABLED)
-		return 0;
 
-	return hwmgr_hw_resume(hwmgr);
+	return hwmgr_resume(hwmgr);
 }
 
 static int pp_set_clockgating_state(void *handle,
@@ -334,12 +257,9 @@ static int pp_dpm_fw_loading_complete(void *handle)
 static int pp_set_clockgating_by_smu(void *handle, uint32_t msg_id)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
-
-	ret = pp_check(hwmgr);
 
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->update_clock_gatings == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -389,12 +309,9 @@ static int pp_dpm_force_performance_level(void *handle,
 					enum amd_dpm_forced_level level)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (level == hwmgr->dpm_level)
 		return 0;
@@ -412,13 +329,10 @@ static enum amd_dpm_forced_level pp_dpm_get_performance_level(
 								void *handle)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
 	enum amd_dpm_forced_level level;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
 	level = hwmgr->dpm_level;
@@ -429,13 +343,10 @@ static enum amd_dpm_forced_level pp_dpm_get_performance_level(
 static uint32_t pp_dpm_get_sclk(void *handle, bool low)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
 	uint32_t clk = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return 0;
 
 	if (hwmgr->hwmgr_func->get_sclk == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -450,13 +361,10 @@ static uint32_t pp_dpm_get_sclk(void *handle, bool low)
 static uint32_t pp_dpm_get_mclk(void *handle, bool low)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
 	uint32_t clk = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return 0;
 
 	if (hwmgr->hwmgr_func->get_mclk == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -471,11 +379,8 @@ static uint32_t pp_dpm_get_mclk(void *handle, bool low)
 static void pp_dpm_powergate_vce(void *handle, bool gate)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
-
-	ret = pp_check(hwmgr);
 
-	if (ret)
+	if (!hwmgr || !hwmgr->pm_en)
 		return;
 
 	if (hwmgr->hwmgr_func->powergate_vce == NULL) {
@@ -490,11 +395,8 @@ static void pp_dpm_powergate_vce(void *handle, bool gate)
 static void pp_dpm_powergate_uvd(void *handle, bool gate)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
-
-	ret = pp_check(hwmgr);
 
-	if (ret)
+	if (!hwmgr || !hwmgr->pm_en)
 		return;
 
 	if (hwmgr->hwmgr_func->powergate_uvd == NULL) {
@@ -512,10 +414,8 @@ static int pp_dpm_dispatch_tasks(void *handle, enum amd_pp_task task_id,
 	int ret = 0;
 	struct pp_hwmgr *hwmgr = handle;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
 	ret = hwmgr_handle_task(hwmgr, task_id, user_state);
@@ -528,15 +428,9 @@ static enum amd_pm_state_type pp_dpm_get_current_power_state(void *handle)
 {
 	struct pp_hwmgr *hwmgr = handle;
 	struct pp_power_state *state;
-	int ret = 0;
 	enum amd_pm_state_type pm_type;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
-
-	if (hwmgr->current_ps == NULL)
+	if (!hwmgr || !hwmgr->pm_en || !hwmgr->current_ps)
 		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
@@ -568,11 +462,8 @@ static enum amd_pm_state_type pp_dpm_get_current_power_state(void *handle)
 static void pp_dpm_set_fan_control_mode(void *handle, uint32_t mode)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
-
-	ret = pp_check(hwmgr);
 
-	if (ret)
+	if (!hwmgr || !hwmgr->pm_en)
 		return;
 
 	if (hwmgr->hwmgr_func->set_fan_control_mode == NULL) {
@@ -587,13 +478,10 @@ static void pp_dpm_set_fan_control_mode(void *handle, uint32_t mode)
 static uint32_t pp_dpm_get_fan_control_mode(void *handle)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
 	uint32_t mode = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return 0;
 
 	if (hwmgr->hwmgr_func->get_fan_control_mode == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -610,10 +498,8 @@ static int pp_dpm_set_fan_speed_percent(void *handle, uint32_t percent)
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->set_fan_speed_percent == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -630,10 +516,8 @@ static int pp_dpm_get_fan_speed_percent(void *handle, uint32_t *speed)
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->get_fan_speed_percent == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -651,10 +535,8 @@ static int pp_dpm_get_fan_speed_rpm(void *handle, uint32_t *rpm)
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->get_fan_speed_rpm == NULL)
 		return -EINVAL;
@@ -670,16 +552,10 @@ static int pp_dpm_get_pp_num_states(void *handle,
 {
 	struct pp_hwmgr *hwmgr = handle;
 	int i;
-	int ret = 0;
 
 	memset(data, 0, sizeof(*data));
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
-
-	if (hwmgr->ps == NULL)
+	if (!hwmgr || !hwmgr->pm_en ||!hwmgr->ps)
 		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
@@ -713,15 +589,9 @@ static int pp_dpm_get_pp_num_states(void *handle,
 static int pp_dpm_get_pp_table(void *handle, char **table)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
 	int size = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
-
-	if (!hwmgr->soft_pp_table)
+	if (!hwmgr || !hwmgr->pm_en ||!hwmgr->soft_pp_table)
 		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
@@ -736,10 +606,6 @@ static int amd_powerplay_reset(void *handle)
 	struct pp_hwmgr *hwmgr = handle;
 	int ret;
 
-	ret = pp_check(hwmgr);
-	if (ret)
-		return ret;
-
 	ret = hwmgr_hw_fini(hwmgr);
 	if (ret)
 		return ret;
@@ -756,10 +622,8 @@ static int pp_dpm_set_pp_table(void *handle, const char *buf, size_t size)
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
 	if (!hwmgr->hardcode_pp_table) {
@@ -796,10 +660,8 @@ static int pp_dpm_force_clock_level(void *handle,
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->force_clock_level == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -820,10 +682,8 @@ static int pp_dpm_print_clock_levels(void *handle,
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->print_clock_levels == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -840,10 +700,8 @@ static int pp_dpm_get_sclk_od(void *handle)
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->get_sclk_od == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -860,10 +718,8 @@ static int pp_dpm_set_sclk_od(void *handle, uint32_t value)
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->set_sclk_od == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -881,10 +737,8 @@ static int pp_dpm_get_mclk_od(void *handle)
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->get_mclk_od == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -901,10 +755,8 @@ static int pp_dpm_set_mclk_od(void *handle, uint32_t value)
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->set_mclk_od == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -922,11 +774,7 @@ static int pp_dpm_read_sensor(void *handle, int idx,
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-	if (ret)
-		return ret;
-
-	if (value == NULL)
+	if (!hwmgr || !hwmgr->pm_en || !value)
 		return -EINVAL;
 
 	switch (idx) {
@@ -948,14 +796,11 @@ static struct amd_vce_state*
 pp_dpm_get_vce_clock_state(void *handle, unsigned idx)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
-
-	ret = pp_check(hwmgr);
 
-	if (ret)
+	if (!hwmgr || !hwmgr->pm_en)
 		return NULL;
 
-	if (hwmgr && idx < hwmgr->num_vce_state_tables)
+	if (idx < hwmgr->num_vce_state_tables)
 		return &hwmgr->vce_states[idx];
 	return NULL;
 }
@@ -964,7 +809,7 @@ static int pp_get_power_profile_mode(void *handle, char *buf)
 {
 	struct pp_hwmgr *hwmgr = handle;
 
-	if (!buf || pp_check(hwmgr))
+	if (!hwmgr || !hwmgr->pm_en || !buf)
 		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->get_power_profile_mode == NULL) {
@@ -980,12 +825,12 @@ static int pp_set_power_profile_mode(void *handle, long *input, uint32_t size)
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = -EINVAL;
 
-	if (pp_check(hwmgr))
-		return -EINVAL;
+	if (!hwmgr || !hwmgr->pm_en)
+		return ret;
 
 	if (hwmgr->hwmgr_func->set_power_profile_mode == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
-		return -EINVAL;
+		return ret;
 	}
 	mutex_lock(&hwmgr->smu_lock);
 	if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_MANUAL)
@@ -998,7 +843,7 @@ static int pp_odn_edit_dpm_table(void *handle, uint32_t type, long *input, uint3
 {
 	struct pp_hwmgr *hwmgr = handle;
 
-	if (pp_check(hwmgr))
+	if (!hwmgr || !hwmgr->pm_en)
 		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->odn_edit_dpm_table == NULL) {
@@ -1016,7 +861,7 @@ static int pp_dpm_switch_power_profile(void *handle,
 	long workload;
 	uint32_t index;
 
-	if (pp_check(hwmgr))
+	if (!hwmgr || !hwmgr->pm_en)
 		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->set_power_profile_mode == NULL) {
@@ -1058,10 +903,8 @@ static int pp_dpm_notify_smu_memory_info(void *handle,
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->notify_cac_buffer_info == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -1082,12 +925,9 @@ static int pp_dpm_notify_smu_memory_info(void *handle,
 static int pp_set_power_limit(void *handle, uint32_t limit)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
-
-	ret = pp_check(hwmgr);
 
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->set_power_limit == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -1104,20 +944,14 @@ static int pp_set_power_limit(void *handle, uint32_t limit)
 	hwmgr->hwmgr_func->set_power_limit(hwmgr, limit);
 	hwmgr->power_limit = limit;
 	mutex_unlock(&hwmgr->smu_lock);
-	return ret;
+	return 0;
 }
 
 static int pp_get_power_limit(void *handle, uint32_t *limit, bool default_limit)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
-
-	ret = pp_check(hwmgr);
 
-	if (ret)
-		return ret;
-
-	if (limit == NULL)
+	if (!hwmgr || !hwmgr->pm_en ||!limit)
 		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
@@ -1129,19 +963,16 @@ static int pp_get_power_limit(void *handle, uint32_t *limit, bool default_limit)
 
 	mutex_unlock(&hwmgr->smu_lock);
 
-	return ret;
+	return 0;
 }
 
 static int pp_display_configuration_change(void *handle,
 	const struct amd_pp_display_configuration *display_config)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
-
-	ret = pp_check(hwmgr);
 
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
 	phm_store_dal_configuration_data(hwmgr, display_config);
@@ -1155,12 +986,7 @@ static int pp_get_display_power_level(void *handle,
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
-
-	if (output == NULL)
+	if (!hwmgr || !hwmgr->pm_en ||!output)
 		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
@@ -1177,10 +1003,8 @@ static int pp_get_current_clocks(void *handle,
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
 
@@ -1225,10 +1049,8 @@ static int pp_get_clock_by_type(void *handle, enum amd_pp_clock_type type, struc
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (clocks == NULL)
 		return -EINVAL;
@@ -1246,11 +1068,7 @@ static int pp_get_clock_by_type_with_latency(void *handle,
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-	if (ret)
-		return ret;
-
-	if (!clocks)
+	if (!hwmgr || !hwmgr->pm_en ||!clocks)
 		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
@@ -1266,11 +1084,7 @@ static int pp_get_clock_by_type_with_voltage(void *handle,
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-	if (ret)
-		return ret;
-
-	if (!clocks)
+	if (!hwmgr || !hwmgr->pm_en ||!clocks)
 		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
@@ -1287,11 +1101,7 @@ static int pp_set_watermarks_for_clocks_ranges(void *handle,
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-	if (ret)
-		return ret;
-
-	if (!wm_with_clock_ranges)
+	if (!hwmgr || !hwmgr->pm_en ||!wm_with_clock_ranges)
 		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
@@ -1308,11 +1118,7 @@ static int pp_display_clock_voltage_request(void *handle,
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-	if (ret)
-		return ret;
-
-	if (!clock)
+	if (!hwmgr || !hwmgr->pm_en ||!clock)
 		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
@@ -1328,12 +1134,7 @@ static int pp_get_display_mode_validation_clocks(void *handle,
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
-
-	if (clocks == NULL)
+	if (!hwmgr || !hwmgr->pm_en ||!clocks)
 		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
@@ -1348,12 +1149,9 @@ static int pp_get_display_mode_validation_clocks(void *handle,
 static int pp_set_mmhub_powergating_by_smu(void *handle)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
-
-	ret = pp_check(hwmgr);
 
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->set_mmhub_powergating_by_smu == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c
index ae2e933..dcceadb 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c
@@ -75,8 +75,7 @@ int phm_set_power_state(struct pp_hwmgr *hwmgr,
 
 int phm_enable_dynamic_state_management(struct pp_hwmgr *hwmgr)
 {
-	int ret = 1;
-	bool enabled;
+	int ret = -EINVAL;;
 	PHM_FUNC_CHECK(hwmgr);
 
 	if (smum_is_dpm_running(hwmgr)) {
@@ -87,17 +86,12 @@ int phm_enable_dynamic_state_management(struct pp_hwmgr *hwmgr)
 	if (NULL != hwmgr->hwmgr_func->dynamic_state_management_enable)
 		ret = hwmgr->hwmgr_func->dynamic_state_management_enable(hwmgr);
 
-	enabled = ret == 0;
-
-	cgs_notify_dpm_enabled(hwmgr->device, enabled);
-
 	return ret;
 }
 
 int phm_disable_dynamic_state_management(struct pp_hwmgr *hwmgr)
 {
-	int ret = -1;
-	bool enabled;
+	int ret = -EINVAL;
 
 	PHM_FUNC_CHECK(hwmgr);
 
@@ -109,10 +103,6 @@ int phm_disable_dynamic_state_management(struct pp_hwmgr *hwmgr)
 	if (hwmgr->hwmgr_func->dynamic_state_management_disable)
 		ret = hwmgr->hwmgr_func->dynamic_state_management_disable(hwmgr);
 
-	enabled = ret == 0 ? false : true;
-
-	cgs_notify_dpm_enabled(hwmgr->device, enabled);
-
 	return ret;
 }
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
index 4298205..30ff8a9 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
@@ -76,7 +76,7 @@ static void hwmgr_init_workload_prority(struct pp_hwmgr *hwmgr)
 
 int hwmgr_early_init(struct pp_hwmgr *hwmgr)
 {
-	if (hwmgr == NULL)
+	if (!hwmgr)
 		return -EINVAL;
 
 	hwmgr->usec_timeout = AMD_MAX_USEC_TIMEOUT;
@@ -170,17 +170,51 @@ int hwmgr_early_init(struct pp_hwmgr *hwmgr)
 	return 0;
 }
 
+int hwmgr_sw_init(struct pp_hwmgr *hwmgr)
+{
+	if (!hwmgr|| !hwmgr->smumgr_funcs || !hwmgr->smumgr_funcs->smu_init)
+		return -EINVAL;
+
+	phm_register_irq_handlers(hwmgr);
+
+	return hwmgr->smumgr_funcs->smu_init(hwmgr);
+}
+
+
+int hwmgr_sw_fini(struct pp_hwmgr *hwmgr)
+{
+	if (hwmgr && hwmgr->smumgr_funcs && hwmgr->smumgr_funcs->smu_fini)
+		hwmgr->smumgr_funcs->smu_fini(hwmgr);
+
+	return 0;
+}
+
 int hwmgr_hw_init(struct pp_hwmgr *hwmgr)
 {
 	int ret = 0;
 
-	if (hwmgr == NULL)
+	if (!hwmgr || !hwmgr->smumgr_funcs)
 		return -EINVAL;
 
-	if (hwmgr->pptable_func == NULL ||
-	    hwmgr->pptable_func->pptable_init == NULL ||
-	    hwmgr->hwmgr_func->backend_init == NULL)
-		return -EINVAL;
+	if (hwmgr->smumgr_funcs->start_smu) {
+		ret = hwmgr->smumgr_funcs->start_smu(hwmgr);
+		if (ret) {
+			pr_err("smc start failed\n");
+			return -EINVAL;
+		}
+	}
+
+	if (!hwmgr->pm_en)
+		return 0;
+
+	if (!hwmgr->pptable_func ||
+	    !hwmgr->pptable_func->pptable_init ||
+	    !hwmgr->hwmgr_func->backend_init) {
+		hwmgr->pm_en = false;
+		((struct amdgpu_device *)hwmgr->adev)->pm.dpm_enabled = false;
+		pr_info("dpm not supported \n");
+		return 0;
+	}
 
 	ret = hwmgr->pptable_func->pptable_init(hwmgr);
 	if (ret)
@@ -214,14 +248,13 @@ err1:
 	if (hwmgr->pptable_func->pptable_fini)
 		hwmgr->pptable_func->pptable_fini(hwmgr);
 err:
-	pr_err("amdgpu: powerplay initialization failed\n");
 	return ret;
 }
 
 int hwmgr_hw_fini(struct pp_hwmgr *hwmgr)
 {
-	if (hwmgr == NULL)
-		return -EINVAL;
+	if (!hwmgr || !hwmgr->pm_en)
+		return 0;
 
 	phm_stop_thermal_controller(hwmgr);
 	psm_set_boot_states(hwmgr);
@@ -236,12 +269,12 @@ int hwmgr_hw_fini(struct pp_hwmgr *hwmgr)
 	return psm_fini_power_state_table(hwmgr);
 }
 
-int hwmgr_hw_suspend(struct pp_hwmgr *hwmgr)
+int hwmgr_suspend(struct pp_hwmgr *hwmgr)
 {
 	int ret = 0;
 
-	if (hwmgr == NULL)
-		return -EINVAL;
+	if (!hwmgr || !hwmgr->pm_en)
+		return 0;
 
 	phm_disable_smc_firmware_ctf(hwmgr);
 	ret = psm_set_boot_states(hwmgr);
@@ -255,13 +288,23 @@ int hwmgr_hw_suspend(struct pp_hwmgr *hwmgr)
 	return ret;
 }
 
-int hwmgr_hw_resume(struct pp_hwmgr *hwmgr)
+int hwmgr_resume(struct pp_hwmgr *hwmgr)
 {
 	int ret = 0;
 
-	if (hwmgr == NULL)
+	if (!hwmgr)
 		return -EINVAL;
 
+	if (hwmgr->smumgr_funcs && hwmgr->smumgr_funcs->start_smu) {
+		if (hwmgr->smumgr_funcs->start_smu(hwmgr)) {
+			pr_err("smc start failed\n");
+			return -EINVAL;
+		}
+	}
+
+	if (!hwmgr->pm_en)
+		return 0;
+
 	ret = phm_setup_asic(hwmgr);
 	if (ret)
 		return ret;
@@ -270,9 +313,6 @@ int hwmgr_hw_resume(struct pp_hwmgr *hwmgr)
 	if (ret)
 		return ret;
 	ret = phm_start_thermal_controller(hwmgr);
-	if (ret)
-		return ret;
-
 	ret |= psm_set_performance_states(hwmgr);
 	if (ret)
 		return ret;
diff --git a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
index 17f811d..d6c9a3b 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
@@ -782,10 +782,13 @@ struct pp_hwmgr {
 };
 
 int hwmgr_early_init(struct pp_hwmgr *hwmgr);
+int hwmgr_sw_init(struct pp_hwmgr *hwmgr);
+int hwmgr_sw_fini(struct pp_hwmgr *hwmgr);
 int hwmgr_hw_init(struct pp_hwmgr *hwmgr);
 int hwmgr_hw_fini(struct pp_hwmgr *hwmgr);
-int hwmgr_hw_suspend(struct pp_hwmgr *hwmgr);
-int hwmgr_hw_resume(struct pp_hwmgr *hwmgr);
+int hwmgr_suspend(struct pp_hwmgr *hwmgr);
+int hwmgr_resume(struct pp_hwmgr *hwmgr);
+
 int hwmgr_handle_task(struct pp_hwmgr *hwmgr,
 				enum amd_pp_task task_id,
 				enum amd_pm_state_type *user_state);
-- 
cgit v1.1


From b61e54cb1881c7cb74787da6a5d39d8d48dcc075 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Thu, 22 Mar 2018 15:12:59 +0800
Subject: drm/amd/pp: Lock pm_funcs when set pp table

unlock mutex until set pp table completely to avoid
conflict if other pp functions were called simultaneously.

use hwmgr_handle_task instand of pp_dpm_dispatch_tasks.
It is not make sense that call pp_functions in ip_functions.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/amd_powerplay.c | 28 +++++++++++++--------------
 1 file changed, 14 insertions(+), 14 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
index 6503bbf..9ada102 100644
--- a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
+++ b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
@@ -31,8 +31,6 @@
 #include "amdgpu.h"
 #include "hwmgr.h"
 
-static int pp_dpm_dispatch_tasks(void *handle, enum amd_pp_task task_id,
-		enum amd_pm_state_type *user_state);
 
 static const struct amd_pm_funcs pp_dpm_funcs;
 
@@ -146,10 +144,12 @@ static int pp_late_init(void *handle)
 	struct amdgpu_device *adev = handle;
 	struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle;
 
-	if (hwmgr && hwmgr->pm_en)
-		pp_dpm_dispatch_tasks(hwmgr,
+	if (hwmgr && hwmgr->pm_en) {
+		mutex_lock(&hwmgr->smu_lock);
+		hwmgr_handle_task(hwmgr,
 					AMD_PP_TASK_COMPLETE_INIT, NULL);
-
+		mutex_unlock(&hwmgr->smu_lock);
+	}
 	return 0;
 }
 
@@ -620,7 +620,7 @@ static int amd_powerplay_reset(void *handle)
 static int pp_dpm_set_pp_table(void *handle, const char *buf, size_t size)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
+	int ret = -ENOMEM;
 
 	if (!hwmgr || !hwmgr->pm_en)
 		return -EINVAL;
@@ -630,28 +630,28 @@ static int pp_dpm_set_pp_table(void *handle, const char *buf, size_t size)
 		hwmgr->hardcode_pp_table = kmemdup(hwmgr->soft_pp_table,
 						   hwmgr->soft_pp_table_size,
 						   GFP_KERNEL);
-		if (!hwmgr->hardcode_pp_table) {
-			mutex_unlock(&hwmgr->smu_lock);
-			return -ENOMEM;
-		}
+		if (!hwmgr->hardcode_pp_table)
+			goto err;
 	}
 
 	memcpy(hwmgr->hardcode_pp_table, buf, size);
 
 	hwmgr->soft_pp_table = hwmgr->hardcode_pp_table;
-	mutex_unlock(&hwmgr->smu_lock);
 
 	ret = amd_powerplay_reset(handle);
 	if (ret)
-		return ret;
+		goto err;
 
 	if (hwmgr->hwmgr_func->avfs_control) {
 		ret = hwmgr->hwmgr_func->avfs_control(hwmgr, false);
 		if (ret)
-			return ret;
+			goto err;
 	}
-
+	mutex_unlock(&hwmgr->smu_lock);
 	return 0;
+err:
+	mutex_unlock(&hwmgr->smu_lock);
+	return ret;
 }
 
 static int pp_dpm_force_clock_level(void *handle,
-- 
cgit v1.1


From 8bb575a2d83af097980641d864401b303286755c Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Thu, 22 Mar 2018 15:46:47 +0800
Subject: drm/amd/pp: Save vf state in pp context

Store vf state in pp_context so we can
deprecate the cgs interface.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/amd_powerplay.c           |  3 ++-
 drivers/gpu/drm/amd/powerplay/inc/hwmgr.h               |  1 +
 drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c      |  7 +++----
 drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c |  3 +--
 drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c      | 10 +++++-----
 drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c     |  3 +--
 6 files changed, 13 insertions(+), 14 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
index 9ada102..337af78 100644
--- a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
+++ b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
@@ -46,7 +46,8 @@ static int amd_powerplay_create(struct amdgpu_device *adev)
 		return -ENOMEM;
 
 	hwmgr->adev = adev;
-	hwmgr->pm_en = (amdgpu_dpm != 0 && !amdgpu_sriov_vf(adev)) ? true : false;
+	hwmgr->not_vf = !amdgpu_sriov_vf(adev);
+	hwmgr->pm_en = (amdgpu_dpm && hwmgr->not_vf) ? true : false;
 	hwmgr->device = amdgpu_cgs_create_device(adev);
 	mutex_init(&hwmgr->smu_lock);
 	hwmgr->chip_family = adev->family;
diff --git a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
index d6c9a3b..d5cadc6 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
@@ -718,6 +718,7 @@ struct pp_hwmgr {
 	uint32_t chip_family;
 	uint32_t chip_id;
 	uint32_t smu_version;
+	bool not_vf;
 	bool pm_en;
 	struct mutex smu_lock;
 
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c
index faef783..35b947e 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c
@@ -288,8 +288,7 @@ static int fiji_start_smu(struct pp_hwmgr *hwmgr)
 	struct fiji_smumgr *priv = (struct fiji_smumgr *)(hwmgr->smu_backend);
 
 	/* Only start SMC if SMC RAM is not running */
-	if (!(smu7_is_smc_ram_running(hwmgr)
-		|| cgs_is_virtualization_enabled(hwmgr->device))) {
+	if (!smu7_is_smc_ram_running(hwmgr) && hwmgr->not_vf) {
 		/* Check if SMU is running in protected mode */
 		if (0 == PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device,
 				CGS_IND_REG__SMC,
@@ -335,8 +334,8 @@ static bool fiji_is_hw_avfs_present(struct pp_hwmgr *hwmgr)
 	uint32_t efuse = 0;
 	uint32_t mask = (1 << ((AVFS_EN_MSB - AVFS_EN_LSB) + 1)) - 1;
 
-	if (cgs_is_virtualization_enabled(hwmgr->device))
-		return 0;
+	if (!hwmgr->not_vf)
+		return false;
 
 	if (!atomctrl_read_efuse(hwmgr->device, AVFS_EN_LSB, AVFS_EN_MSB,
 			mask, &efuse)) {
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c
index fe6854e..05e60e8 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c
@@ -295,8 +295,7 @@ static int polaris10_start_smu(struct pp_hwmgr *hwmgr)
 	struct polaris10_smumgr *smu_data = (struct polaris10_smumgr *)(hwmgr->smu_backend);
 
 	/* Only start SMC if SMC RAM is not running */
-	if (!(smu7_is_smc_ram_running(hwmgr)
-		|| cgs_is_virtualization_enabled(hwmgr->device))) {
+	if (!smu7_is_smc_ram_running(hwmgr) && hwmgr->not_vf) {
 		smu_data->protected_mode = (uint8_t) (PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, SMU_FIRMWARE, SMU_MODE));
 		smu_data->smu7_data.security_hard_key = (uint8_t) (PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, SMU_FIRMWARE, SMU_SEL));
 
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
index 0399c10..3684822 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
@@ -375,7 +375,7 @@ static int smu7_populate_single_firmware_entry(struct pp_hwmgr *hwmgr,
 		entry->meta_data_addr_low = 0;
 
 		/* digest need be excluded out */
-		if (cgs_is_virtualization_enabled(hwmgr->device))
+		if (!hwmgr->not_vf)
 			info.image_size -= 20;
 		entry->data_size_byte = info.image_size;
 		entry->num_register_entries = 0;
@@ -409,7 +409,7 @@ int smu7_request_smu_load_fw(struct pp_hwmgr *hwmgr)
 					0x0);
 
 	if (hwmgr->chip_id > CHIP_TOPAZ) { /* add support for Topaz */
-		if (!cgs_is_virtualization_enabled(hwmgr->device)) {
+		if (hwmgr->not_vf) {
 			smu7_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SMU_DRAM_ADDR_HI,
 						upper_32_bits(smu_data->smu_buffer.mc_addr));
@@ -467,7 +467,7 @@ int smu7_request_smu_load_fw(struct pp_hwmgr *hwmgr)
 	PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr,
 				UCODE_ID_SDMA1, &toc->entry[toc->num_entries++]),
 				"Failed to Get Firmware Entry.", return -EINVAL);
-	if (cgs_is_virtualization_enabled(hwmgr->device))
+	if (!hwmgr->not_vf)
 		PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr,
 				UCODE_ID_MEC_STORAGE, &toc->entry[toc->num_entries++]),
 				"Failed to Get Firmware Entry.", return -EINVAL);
@@ -608,7 +608,7 @@ int smu7_init(struct pp_hwmgr *hwmgr)
 	smu_data->header = smu_data->header_buffer.kaddr;
 	smu_data->header_buffer.mc_addr = mc_addr;
 
-	if (cgs_is_virtualization_enabled(hwmgr->device))
+	if (!hwmgr->not_vf)
 		return 0;
 
 	smu_data->smu_buffer.data_size = 200*4096;
@@ -643,7 +643,7 @@ int smu7_smu_fini(struct pp_hwmgr *hwmgr)
 					&smu_data->header_buffer.mc_addr,
 					&smu_data->header_buffer.kaddr);
 
-	if (!cgs_is_virtualization_enabled(hwmgr->device))
+	if (hwmgr->not_vf)
 		amdgpu_bo_free_kernel(&smu_data->smu_buffer.handle,
 					&smu_data->smu_buffer.mc_addr,
 					&smu_data->smu_buffer.kaddr);
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c
index b51d746..2ba05d2 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c
@@ -199,8 +199,7 @@ static int tonga_start_smu(struct pp_hwmgr *hwmgr)
 	int result;
 
 	/* Only start SMC if SMC RAM is not running */
-	if (!(smu7_is_smc_ram_running(hwmgr) ||
-		cgs_is_virtualization_enabled(hwmgr->device))) {
+	if (!smu7_is_smc_ram_running(hwmgr) && hwmgr->not_vf) {
 		/*Check if SMU is running in protected mode*/
 		if (0 == PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
 					SMU_FIRMWARE, SMU_MODE)) {
-- 
cgit v1.1


From 64f6db77fcb81493988061587fa478e6612dc45b Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Thu, 22 Mar 2018 19:32:45 +0800
Subject: drm/amd/pp: Use release_firmware directly in powerplay

Use kernel api directly so we can deprecate the cgs interface.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/amd_powerplay.c      | 6 +++++-
 drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c   | 1 -
 drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c | 1 -
 3 files changed, 5 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
index 337af78..b91ef11 100644
--- a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
+++ b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
@@ -25,6 +25,7 @@
 #include <linux/kernel.h>
 #include <linux/gfp.h>
 #include <linux/slab.h>
+#include <linux/firmware.h>
 #include "amd_shared.h"
 #include "amd_powerplay.h"
 #include "power_state.h"
@@ -107,8 +108,11 @@ static int pp_sw_fini(void *handle)
 
 	hwmgr_sw_fini(hwmgr);
 
-	if (adev->firmware.load_type == AMDGPU_FW_LOAD_SMU)
+	if (adev->firmware.load_type == AMDGPU_FW_LOAD_SMU) {
+		release_firmware(adev->pm.fw);
+		adev->pm.fw = NULL;
 		amdgpu_ucode_fini_bo(adev);
+	}
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c
index 08d0001..e30a2ee 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c
@@ -2784,7 +2784,6 @@ static int ci_smu_fini(struct pp_hwmgr *hwmgr)
 {
 	kfree(hwmgr->smu_backend);
 	hwmgr->smu_backend = NULL;
-	cgs_rel_firmware(hwmgr->device, CGS_UCODE_ID_SMU);
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
index 3684822..41fab2d 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
@@ -650,6 +650,5 @@ int smu7_smu_fini(struct pp_hwmgr *hwmgr)
 
 	kfree(hwmgr->smu_backend);
 	hwmgr->smu_backend = NULL;
-	cgs_rel_firmware(hwmgr->device, CGS_UCODE_ID_SMU);
 	return 0;
 }
-- 
cgit v1.1


From b13aa1091fb2002a6854e0401df5fc6231fbca58 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Mon, 26 Mar 2018 16:18:34 +0800
Subject: drm/amdgpu: Use dpm_enabled as dpm state flag

driver will set dpm_enabled to true only when
module parameter amdgpu_dpm not equal to 0 and
smu hw initialize successfully.

Reviewed-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c | 2 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c     | 2 +-
 drivers/gpu/drm/amd/amdgpu/ci_dpm.c         | 2 +-
 drivers/gpu/drm/amd/amdgpu/kv_dpm.c         | 2 +-
 drivers/gpu/drm/amd/amdgpu/si_dpm.c         | 2 +-
 drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c       | 4 ++--
 drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c | 3 ++-
 7 files changed, 9 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
index 448d69f..c98e597 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
@@ -428,7 +428,7 @@ static ssize_t amdgpu_debugfs_sensor_read(struct file *f, char __user *buf,
 	if (size & 3 || *pos & 0x3)
 		return -EINVAL;
 
-	if (amdgpu_dpm == 0)
+	if (!adev->pm.dpm_enabled)
 		return -EINVAL;
 
 	/* convert offset to sensor number */
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
index 4b7824d..bd9e723 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
@@ -704,7 +704,7 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 		struct pp_gpu_power query = {0};
 		int query_size = sizeof(query);
 
-		if (amdgpu_dpm == 0)
+		if (!adev->pm.dpm_enabled)
 			return -ENOENT;
 
 		switch (info->sensor_info.type) {
diff --git a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c
index 47ef3e6..be6b199 100644
--- a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c
@@ -6255,7 +6255,7 @@ static int ci_dpm_late_init(void *handle)
 	int ret;
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	if (!amdgpu_dpm)
+	if (!adev->pm.dpm_enabled)
 		return 0;
 
 	/* init the sysfs and debugfs files late */
diff --git a/drivers/gpu/drm/amd/amdgpu/kv_dpm.c b/drivers/gpu/drm/amd/amdgpu/kv_dpm.c
index 26ba984..bc1720e 100644
--- a/drivers/gpu/drm/amd/amdgpu/kv_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/kv_dpm.c
@@ -2974,7 +2974,7 @@ static int kv_dpm_late_init(void *handle)
 	/* powerdown unused blocks for now */
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	if (!amdgpu_dpm)
+	if (!adev->pm.dpm_enabled)
 		return 0;
 
 	kv_dpm_powergate_acp(adev, true);
diff --git a/drivers/gpu/drm/amd/amdgpu/si_dpm.c b/drivers/gpu/drm/amd/amdgpu/si_dpm.c
index 797d505..b12d7c9 100644
--- a/drivers/gpu/drm/amd/amdgpu/si_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/si_dpm.c
@@ -7580,7 +7580,7 @@ static int si_dpm_late_init(void *handle)
 	int ret;
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	if (!amdgpu_dpm)
+	if (!adev->pm.dpm_enabled)
 		return 0;
 
 	ret = si_set_temperature_range(adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
index 948bb943..87cbb14 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
@@ -688,7 +688,7 @@ static int uvd_v4_2_set_powergating_state(void *handle,
 
 	if (state == AMD_PG_STATE_GATE) {
 		uvd_v4_2_stop(adev);
-		if (adev->pg_flags & AMD_PG_SUPPORT_UVD && amdgpu_dpm == 0) {
+		if (adev->pg_flags & AMD_PG_SUPPORT_UVD && !adev->pm.dpm_enabled) {
 			if (!(RREG32_SMC(ixCURRENT_PG_STATUS) &
 				CURRENT_PG_STATUS__UVD_PG_STATUS_MASK)) {
 				WREG32(mmUVD_PGFSM_CONFIG, (UVD_PGFSM_CONFIG__UVD_PGFSM_FSM_ADDR_MASK   |
@@ -699,7 +699,7 @@ static int uvd_v4_2_set_powergating_state(void *handle,
 		}
 		return 0;
 	} else {
-		if (adev->pg_flags & AMD_PG_SUPPORT_UVD && amdgpu_dpm == 0) {
+		if (adev->pg_flags & AMD_PG_SUPPORT_UVD && !adev->pm.dpm_enabled) {
 			if (RREG32_SMC(ixCURRENT_PG_STATUS) &
 				CURRENT_PG_STATUS__UVD_PG_STATUS_MASK) {
 				WREG32(mmUVD_PGFSM_CONFIG, (UVD_PGFSM_CONFIG__UVD_PGFSM_FSM_ADDR_MASK   |
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
index 30ff8a9..bca67df 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
@@ -211,7 +211,6 @@ int hwmgr_hw_init(struct pp_hwmgr *hwmgr)
 	    !hwmgr->pptable_func->pptable_init ||
 	    !hwmgr->hwmgr_func->backend_init) {
 		hwmgr->pm_en = false;
-		((struct amdgpu_device *)hwmgr->adev)->pm.dpm_enabled = false;
 		pr_info("dpm not supported \n");
 		return 0;
 	}
@@ -240,6 +239,8 @@ int hwmgr_hw_init(struct pp_hwmgr *hwmgr)
 	if (ret)
 		goto err2;
 
+	((struct amdgpu_device *)hwmgr->adev)->pm.dpm_enabled = true;
+
 	return 0;
 err2:
 	if (hwmgr->hwmgr_func->backend_fini)
-- 
cgit v1.1


From 986567e4ed81a21a66e841b9e87e708c435328d8 Mon Sep 17 00:00:00 2001
From: Colin Ian King <colin.king@canonical.com>
Date: Tue, 27 Mar 2018 09:32:57 +0100
Subject: drm/amd/pp: Fix spelling mistake: "suppported" -> "supported"

Trivial fix to spelling mistake in pr_warn warning message text

Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/pp_psm.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/pp_psm.c b/drivers/gpu/drm/amd/powerplay/hwmgr/pp_psm.c
index 0f2851b..308bff2 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/pp_psm.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/pp_psm.c
@@ -46,7 +46,7 @@ int psm_init_power_state_table(struct pp_hwmgr *hwmgr)
 					  sizeof(struct pp_power_state);
 
 	if (table_entries == 0 || size == 0) {
-		pr_warn("Please check whether power state management is suppported on this asic\n");
+		pr_warn("Please check whether power state management is supported on this asic\n");
 		return 0;
 	}
 
-- 
cgit v1.1


From 62fd51275e4d43e300f95f2148a41e5bf738ac29 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Fri, 23 Mar 2018 18:18:23 +0800
Subject: drm/amd/pp: Use gfx rlc funcs directly in powerplay
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

In order to remove cgs interfaces:
cgs_enter_safe_mode
cgs_lock_grbm_idx

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c   | 13 ++--
 .../gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c | 71 ++++++++++++----------
 2 files changed, 46 insertions(+), 38 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c
index 03bc745..a55ee16 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c
@@ -740,8 +740,8 @@ int smu7_enable_didt_config(struct pp_hwmgr *hwmgr)
 	    PP_CAP(PHM_PlatformCaps_TDRamping) ||
 	    PP_CAP(PHM_PlatformCaps_TCPRamping)) {
 
-		cgs_enter_safe_mode(hwmgr->device, true);
-		cgs_lock_grbm_idx(hwmgr->device, true);
+		adev->gfx.rlc.funcs->enter_safe_mode(adev);
+		mutex_lock(&adev->grbm_idx_mutex);
 		value = 0;
 		value2 = cgs_read_register(hwmgr->device, mmGRBM_GFX_INDEX);
 		for (count = 0; count < num_se; count++) {
@@ -781,8 +781,8 @@ int smu7_enable_didt_config(struct pp_hwmgr *hwmgr)
 			PP_ASSERT_WITH_CODE((0 == result),
 					"Failed to enable DPM DIDT.", return result);
 		}
-		cgs_lock_grbm_idx(hwmgr->device, false);
-		cgs_enter_safe_mode(hwmgr->device, false);
+		mutex_unlock(&adev->grbm_idx_mutex);
+		adev->gfx.rlc.funcs->exit_safe_mode(adev);
 	}
 
 	return 0;
@@ -791,13 +791,14 @@ int smu7_enable_didt_config(struct pp_hwmgr *hwmgr)
 int smu7_disable_didt_config(struct pp_hwmgr *hwmgr)
 {
 	int result;
+	struct amdgpu_device *adev = hwmgr->adev;
 
 	if (PP_CAP(PHM_PlatformCaps_SQRamping) ||
 	    PP_CAP(PHM_PlatformCaps_DBRamping) ||
 	    PP_CAP(PHM_PlatformCaps_TDRamping) ||
 	    PP_CAP(PHM_PlatformCaps_TCPRamping)) {
 
-		cgs_enter_safe_mode(hwmgr->device, true);
+		adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
 		result = smu7_enable_didt(hwmgr, false);
 		PP_ASSERT_WITH_CODE((result == 0),
@@ -809,7 +810,7 @@ int smu7_disable_didt_config(struct pp_hwmgr *hwmgr)
 			PP_ASSERT_WITH_CODE((0 == result),
 					"Failed to disable DPM DIDT.", return result);
 		}
-		cgs_enter_safe_mode(hwmgr->device, false);
+		adev->gfx.rlc.funcs->exit_safe_mode(adev);
 	}
 
 	return 0;
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c
index ba63fae..203a691 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c
@@ -930,16 +930,16 @@ static void vega10_didt_set_mask(struct pp_hwmgr *hwmgr, const bool enable)
 
 static int vega10_enable_cac_driving_se_didt_config(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	int result;
 	uint32_t num_se = 0, count, data;
-	struct amdgpu_device *adev = hwmgr->adev;
 	uint32_t reg;
 
 	num_se = adev->gfx.config.max_shader_engines;
 
-	cgs_enter_safe_mode(hwmgr->device, true);
+	adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
-	cgs_lock_grbm_idx(hwmgr->device, true);
+	mutex_lock(&adev->grbm_idx_mutex);
 	reg = soc15_get_register_offset(GC_HWID, 0, mmGRBM_GFX_INDEX_BASE_IDX, mmGRBM_GFX_INDEX);
 	for (count = 0; count < num_se; count++) {
 		data = GRBM_GFX_INDEX__INSTANCE_BROADCAST_WRITES_MASK | GRBM_GFX_INDEX__SH_BROADCAST_WRITES_MASK | ( count << GRBM_GFX_INDEX__SE_INDEX__SHIFT);
@@ -959,38 +959,40 @@ static int vega10_enable_cac_driving_se_didt_config(struct pp_hwmgr *hwmgr)
 			break;
 	}
 	cgs_write_register(hwmgr->device, reg, 0xE0000000);
-	cgs_lock_grbm_idx(hwmgr->device, false);
+	mutex_unlock(&adev->grbm_idx_mutex);
 
 	vega10_didt_set_mask(hwmgr, true);
 
-	cgs_enter_safe_mode(hwmgr->device, false);
+	adev->gfx.rlc.funcs->exit_safe_mode(adev);
 
 	return 0;
 }
 
 static int vega10_disable_cac_driving_se_didt_config(struct pp_hwmgr *hwmgr)
 {
-	cgs_enter_safe_mode(hwmgr->device, true);
+	struct amdgpu_device *adev = hwmgr->adev;
+
+	adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
 	vega10_didt_set_mask(hwmgr, false);
 
-	cgs_enter_safe_mode(hwmgr->device, false);
+	adev->gfx.rlc.funcs->exit_safe_mode(adev);
 
 	return 0;
 }
 
 static int vega10_enable_psm_gc_didt_config(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	int result;
 	uint32_t num_se = 0, count, data;
-	struct amdgpu_device *adev = hwmgr->adev;
 	uint32_t reg;
 
 	num_se = adev->gfx.config.max_shader_engines;
 
-	cgs_enter_safe_mode(hwmgr->device, true);
+	adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
-	cgs_lock_grbm_idx(hwmgr->device, true);
+	mutex_lock(&adev->grbm_idx_mutex);
 	reg = soc15_get_register_offset(GC_HWID, 0, mmGRBM_GFX_INDEX_BASE_IDX, mmGRBM_GFX_INDEX);
 	for (count = 0; count < num_se; count++) {
 		data = GRBM_GFX_INDEX__INSTANCE_BROADCAST_WRITES_MASK | GRBM_GFX_INDEX__SH_BROADCAST_WRITES_MASK | ( count << GRBM_GFX_INDEX__SE_INDEX__SHIFT);
@@ -1004,11 +1006,11 @@ static int vega10_enable_psm_gc_didt_config(struct pp_hwmgr *hwmgr)
 			break;
 	}
 	cgs_write_register(hwmgr->device, reg, 0xE0000000);
-	cgs_lock_grbm_idx(hwmgr->device, false);
+	mutex_unlock(&adev->grbm_idx_mutex);
 
 	vega10_didt_set_mask(hwmgr, true);
 
-	cgs_enter_safe_mode(hwmgr->device, false);
+	adev->gfx.rlc.funcs->exit_safe_mode(adev);
 
 	vega10_program_gc_didt_config_registers(hwmgr, GCDiDtDroopCtrlConfig_vega10);
 	if (PP_CAP(PHM_PlatformCaps_GCEDC))
@@ -1022,13 +1024,14 @@ static int vega10_enable_psm_gc_didt_config(struct pp_hwmgr *hwmgr)
 
 static int vega10_disable_psm_gc_didt_config(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	uint32_t data;
 
-	cgs_enter_safe_mode(hwmgr->device, true);
+	adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
 	vega10_didt_set_mask(hwmgr, false);
 
-	cgs_enter_safe_mode(hwmgr->device, false);
+	adev->gfx.rlc.funcs->exit_safe_mode(adev);
 
 	if (PP_CAP(PHM_PlatformCaps_GCEDC)) {
 		data = 0x00000000;
@@ -1043,16 +1046,16 @@ static int vega10_disable_psm_gc_didt_config(struct pp_hwmgr *hwmgr)
 
 static int vega10_enable_se_edc_config(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	int result;
 	uint32_t num_se = 0, count, data;
-	struct amdgpu_device *adev = hwmgr->adev;
 	uint32_t reg;
 
 	num_se = adev->gfx.config.max_shader_engines;
 
-	cgs_enter_safe_mode(hwmgr->device, true);
+	adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
-	cgs_lock_grbm_idx(hwmgr->device, true);
+	mutex_lock(&adev->grbm_idx_mutex);
 	reg = soc15_get_register_offset(GC_HWID, 0, mmGRBM_GFX_INDEX_BASE_IDX, mmGRBM_GFX_INDEX);
 	for (count = 0; count < num_se; count++) {
 		data = GRBM_GFX_INDEX__INSTANCE_BROADCAST_WRITES_MASK | GRBM_GFX_INDEX__SH_BROADCAST_WRITES_MASK | ( count << GRBM_GFX_INDEX__SE_INDEX__SHIFT);
@@ -1068,41 +1071,43 @@ static int vega10_enable_se_edc_config(struct pp_hwmgr *hwmgr)
 			break;
 	}
 	cgs_write_register(hwmgr->device, reg, 0xE0000000);
-	cgs_lock_grbm_idx(hwmgr->device, false);
+	mutex_unlock(&adev->grbm_idx_mutex);
 
 	vega10_didt_set_mask(hwmgr, true);
 
-	cgs_enter_safe_mode(hwmgr->device, false);
+	adev->gfx.rlc.funcs->exit_safe_mode(adev);
 
 	return 0;
 }
 
 static int vega10_disable_se_edc_config(struct pp_hwmgr *hwmgr)
 {
-	cgs_enter_safe_mode(hwmgr->device, true);
+	struct amdgpu_device *adev = hwmgr->adev;
+
+	adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
 	vega10_didt_set_mask(hwmgr, false);
 
-	cgs_enter_safe_mode(hwmgr->device, false);
+	adev->gfx.rlc.funcs->exit_safe_mode(adev);
 
 	return 0;
 }
 
 static int vega10_enable_psm_gc_edc_config(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	int result;
 	uint32_t num_se = 0;
 	uint32_t count, data;
-	struct amdgpu_device *adev = hwmgr->adev;
 	uint32_t reg;
 
 	num_se = adev->gfx.config.max_shader_engines;
 
-	cgs_enter_safe_mode(hwmgr->device, true);
+	adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
 	vega10_program_gc_didt_config_registers(hwmgr, AvfsPSMResetConfig_vega10);
 
-	cgs_lock_grbm_idx(hwmgr->device, true);
+	mutex_lock(&adev->grbm_idx_mutex);
 	reg = soc15_get_register_offset(GC_HWID, 0, mmGRBM_GFX_INDEX_BASE_IDX, mmGRBM_GFX_INDEX);
 	for (count = 0; count < num_se; count++) {
 		data = GRBM_GFX_INDEX__INSTANCE_BROADCAST_WRITES_MASK | GRBM_GFX_INDEX__SH_BROADCAST_WRITES_MASK | ( count << GRBM_GFX_INDEX__SE_INDEX__SHIFT);
@@ -1116,11 +1121,11 @@ static int vega10_enable_psm_gc_edc_config(struct pp_hwmgr *hwmgr)
 			break;
 	}
 	cgs_write_register(hwmgr->device, reg, 0xE0000000);
-	cgs_lock_grbm_idx(hwmgr->device, false);
+	mutex_unlock(&adev->grbm_idx_mutex);
 
 	vega10_didt_set_mask(hwmgr, true);
 
-	cgs_enter_safe_mode(hwmgr->device, false);
+	adev->gfx.rlc.funcs->exit_safe_mode(adev);
 
 	vega10_program_gc_didt_config_registers(hwmgr, PSMGCEDCDroopCtrlConfig_vega10);
 
@@ -1137,13 +1142,14 @@ static int vega10_enable_psm_gc_edc_config(struct pp_hwmgr *hwmgr)
 
 static int vega10_disable_psm_gc_edc_config(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	uint32_t data;
 
-	cgs_enter_safe_mode(hwmgr->device, true);
+	adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
 	vega10_didt_set_mask(hwmgr, false);
 
-	cgs_enter_safe_mode(hwmgr->device, false);
+	adev->gfx.rlc.funcs->exit_safe_mode(adev);
 
 	if (PP_CAP(PHM_PlatformCaps_GCEDC)) {
 		data = 0x00000000;
@@ -1158,15 +1164,16 @@ static int vega10_disable_psm_gc_edc_config(struct pp_hwmgr *hwmgr)
 
 static int vega10_enable_se_edc_force_stall_config(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	uint32_t reg;
 	int result;
 
-	cgs_enter_safe_mode(hwmgr->device, true);
+	adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
-	cgs_lock_grbm_idx(hwmgr->device, true);
+	mutex_lock(&adev->grbm_idx_mutex);
 	reg = soc15_get_register_offset(GC_HWID, 0, mmGRBM_GFX_INDEX_BASE_IDX, mmGRBM_GFX_INDEX);
 	cgs_write_register(hwmgr->device, reg, 0xE0000000);
-	cgs_lock_grbm_idx(hwmgr->device, false);
+	mutex_unlock(&adev->grbm_idx_mutex);
 
 	result = vega10_program_didt_config_registers(hwmgr, SEEDCForceStallPatternConfig_Vega10, VEGA10_CONFIGREG_DIDT);
 	result |= vega10_program_didt_config_registers(hwmgr, SEEDCCtrlForceStallConfig_Vega10, VEGA10_CONFIGREG_DIDT);
@@ -1175,7 +1182,7 @@ static int vega10_enable_se_edc_force_stall_config(struct pp_hwmgr *hwmgr)
 
 	vega10_didt_set_mask(hwmgr, false);
 
-	cgs_enter_safe_mode(hwmgr->device, false);
+	adev->gfx.rlc.funcs->exit_safe_mode(adev);
 
 	return 0;
 }
-- 
cgit v1.1


From d32d661770a455802afc8d2f9efed617cc8073ed Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Fri, 23 Mar 2018 18:36:51 +0800
Subject: drm/amdgpu: Get pci resource directly through adev
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

In order to remove the cgs wrapper function
cgs_get_pci_resource

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c | 11 +++++------
 1 file changed, 5 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c
index a29362f..03ee367 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c
@@ -290,12 +290,11 @@ static int acp_hw_init(void *handle)
 	else if (r)
 		return r;
 
-	r = cgs_get_pci_resource(adev->acp.cgs_device, CGS_RESOURCE_TYPE_MMIO,
-			0x5289, 0, &acp_base);
-	if (r == -ENODEV)
-		return 0;
-	else if (r)
-		return r;
+	if (adev->rmmio_size == 0 || adev->rmmio_size < 0x5289)
+		return -EINVAL;
+
+	acp_base = adev->rmmio_base;
+
 	if (adev->asic_type != CHIP_STONEY) {
 		adev->acp.acp_genpd = kzalloc(sizeof(struct acp_pm_domain), GFP_KERNEL);
 		if (adev->acp.acp_genpd == NULL)
-- 
cgit v1.1


From e8ee21d2a46d8d5cc85766fb49251c2c21871b30 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Mon, 26 Mar 2018 18:13:28 +0800
Subject: drm/amd/dc: Use atombios api directly in DC

In order to remove the cgs wrapper functions for atombios api.

Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../gpu/drm/amd/display/dc/bios/command_table.c    | 22 +++++++++---------
 .../gpu/drm/amd/display/dc/bios/command_table2.c   | 26 +++++++++++++---------
 2 files changed, 28 insertions(+), 20 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table.c b/drivers/gpu/drm/amd/display/dc/bios/command_table.c
index 4b5fdd5..651e1fd 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/command_table.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/command_table.c
@@ -24,7 +24,7 @@
  */
 
 #include "dm_services.h"
-
+#include "amdgpu.h"
 #include "atom.h"
 
 #include "include/bios_parser_interface.h"
@@ -35,16 +35,16 @@
 #include "bios_parser_types_internal.h"
 
 #define EXEC_BIOS_CMD_TABLE(command, params)\
-	(cgs_atom_exec_cmd_table(bp->base.ctx->cgs_device, \
+	(amdgpu_atom_execute_table(((struct amdgpu_device *)bp->base.ctx->driver_context)->mode_info.atom_context, \
 		GetIndexIntoMasterTable(COMMAND, command), \
-		&params) == 0)
+		(uint32_t *)&params) == 0)
 
 #define BIOS_CMD_TABLE_REVISION(command, frev, crev)\
-	cgs_atom_get_cmd_table_revs(bp->base.ctx->cgs_device, \
+	amdgpu_atom_parse_cmd_header(((struct amdgpu_device *)bp->base.ctx->driver_context)->mode_info.atom_context, \
 		GetIndexIntoMasterTable(COMMAND, command), &frev, &crev)
 
 #define BIOS_CMD_TABLE_PARA_REVISION(command)\
-	bios_cmd_table_para_revision(bp->base.ctx->cgs_device, \
+	bios_cmd_table_para_revision(bp->base.ctx->driver_context, \
 		GetIndexIntoMasterTable(COMMAND, command))
 
 static void init_dig_encoder_control(struct bios_parser *bp);
@@ -82,16 +82,18 @@ void dal_bios_parser_init_cmd_tbl(struct bios_parser *bp)
 	init_set_dce_clock(bp);
 }
 
-static uint32_t bios_cmd_table_para_revision(void *cgs_device,
+static uint32_t bios_cmd_table_para_revision(void *dev,
 					     uint32_t index)
 {
+	struct amdgpu_device *adev = dev;
 	uint8_t frev, crev;
 
-	if (cgs_atom_get_cmd_table_revs(cgs_device,
+	if (amdgpu_atom_parse_cmd_header(adev->mode_info.atom_context,
 					index,
-					&frev, &crev) != 0)
+					&frev, &crev))
+		return crev;
+	else
 		return 0;
-	return crev;
 }
 
 /*******************************************************************************
@@ -368,7 +370,7 @@ static void init_transmitter_control(struct bios_parser *bp)
 	uint8_t crev;
 
 	if (BIOS_CMD_TABLE_REVISION(UNIPHYTransmitterControl,
-			frev, crev) != 0)
+			frev, crev) == false)
 		BREAK_TO_DEBUGGER();
 	switch (crev) {
 	case 2:
diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table2.c b/drivers/gpu/drm/amd/display/dc/bios/command_table2.c
index 3f63f71..752b08a 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/command_table2.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/command_table2.c
@@ -26,14 +26,18 @@
 #include "dm_services.h"
 
 #include "ObjectID.h"
-#include "atomfirmware.h"
 
+#include "atomfirmware.h"
+#include "atom.h"
 #include "include/bios_parser_interface.h"
 
 #include "command_table2.h"
 #include "command_table_helper2.h"
 #include "bios_parser_helper.h"
 #include "bios_parser_types_internal2.h"
+#include "amdgpu.h"
+
+
 #define DC_LOGGER \
 	bp->base.ctx->logger
 
@@ -43,16 +47,16 @@
 		->FieldName)-(char *)0)/sizeof(uint16_t))
 
 #define EXEC_BIOS_CMD_TABLE(fname, params)\
-	(cgs_atom_exec_cmd_table(bp->base.ctx->cgs_device, \
+	(amdgpu_atom_execute_table(((struct amdgpu_device *)bp->base.ctx->driver_context)->mode_info.atom_context, \
 		GET_INDEX_INTO_MASTER_TABLE(command, fname), \
-		&params) == 0)
+		(uint32_t *)&params) == 0)
 
 #define BIOS_CMD_TABLE_REVISION(fname, frev, crev)\
-	cgs_atom_get_cmd_table_revs(bp->base.ctx->cgs_device, \
+	amdgpu_atom_parse_cmd_header(((struct amdgpu_device *)bp->base.ctx->driver_context)->mode_info.atom_context, \
 		GET_INDEX_INTO_MASTER_TABLE(command, fname), &frev, &crev)
 
 #define BIOS_CMD_TABLE_PARA_REVISION(fname)\
-	bios_cmd_table_para_revision(bp->base.ctx->cgs_device, \
+	bios_cmd_table_para_revision(bp->base.ctx->driver_context, \
 			GET_INDEX_INTO_MASTER_TABLE(command, fname))
 
 static void init_dig_encoder_control(struct bios_parser *bp);
@@ -86,16 +90,18 @@ void dal_firmware_parser_init_cmd_tbl(struct bios_parser *bp)
 	init_get_smu_clock_info(bp);
 }
 
-static uint32_t bios_cmd_table_para_revision(void *cgs_device,
+static uint32_t bios_cmd_table_para_revision(void *dev,
 					     uint32_t index)
 {
+	struct amdgpu_device *adev = dev;
 	uint8_t frev, crev;
 
-	if (cgs_atom_get_cmd_table_revs(cgs_device,
+	if (amdgpu_atom_parse_cmd_header(adev->mode_info.atom_context,
 					index,
-					&frev, &crev) != 0)
+					&frev, &crev))
+		return crev;
+	else
 		return 0;
-	return crev;
 }
 
 /******************************************************************************
@@ -201,7 +207,7 @@ static void init_transmitter_control(struct bios_parser *bp)
 	uint8_t frev;
 	uint8_t crev;
 
-	if (BIOS_CMD_TABLE_REVISION(dig1transmittercontrol, frev, crev) != 0)
+	if (BIOS_CMD_TABLE_REVISION(dig1transmittercontrol, frev, crev) == false)
 		BREAK_TO_DEBUGGER();
 	switch (crev) {
 	case 6:
-- 
cgit v1.1


From b3892e2bb519fe18225d0628f0dd255761f16502 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Mon, 26 Mar 2018 18:49:35 +0800
Subject: drm/amd/pp: Use atombios api directly in powerplay (v2)

In order to remove the cgs wrapper functions for atombios api.

v2: squash in whitespace cleanup (Alex)

Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c   | 145 +++++++++++----------
 drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h   |   2 +-
 drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c |  90 ++++++-------
 .../amd/powerplay/hwmgr/process_pptables_v1_0.c    |   2 +-
 .../gpu/drm/amd/powerplay/hwmgr/processpptables.c  |   4 +-
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c   |   3 +-
 drivers/gpu/drm/amd/powerplay/hwmgr/smu8_hwmgr.c   |   3 +-
 drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.c   |  16 +++
 drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.h   |   3 +
 .../amd/powerplay/hwmgr/vega10_processpptables.c   |   2 +-
 .../amd/powerplay/hwmgr/vega12_processpptables.c   |   2 +-
 drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c |   2 +-
 12 files changed, 150 insertions(+), 124 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c
index c6febbf..971fb5d 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c
@@ -23,7 +23,7 @@
 #include "pp_debug.h"
 #include <linux/module.h>
 #include <linux/slab.h>
-
+#include "atom.h"
 #include "ppatomctrl.h"
 #include "atombios.h"
 #include "cgs_common.h"
@@ -128,7 +128,6 @@ static int atomctrl_set_mc_reg_address_table(
 	return 0;
 }
 
-
 int atomctrl_initialize_mc_reg_table(
 		struct pp_hwmgr *hwmgr,
 		uint8_t module_index,
@@ -141,7 +140,7 @@ int atomctrl_initialize_mc_reg_table(
 	u16 size;
 
 	vram_info = (ATOM_VRAM_INFO_HEADER_V2_1 *)
-		cgs_atom_get_data_table(hwmgr->device,
+		smu_atom_get_data_table(hwmgr->adev,
 				GetIndexIntoMasterTable(DATA, VRAM_Info), &size, &frev, &crev);
 
 	if (module_index >= vram_info->ucNumOfVRAMModule) {
@@ -174,6 +173,8 @@ int atomctrl_set_engine_dram_timings_rv770(
 		uint32_t engine_clock,
 		uint32_t memory_clock)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
+
 	SET_ENGINE_CLOCK_PS_ALLOCATION engine_clock_parameters;
 
 	/* They are both in 10KHz Units. */
@@ -184,9 +185,10 @@ int atomctrl_set_engine_dram_timings_rv770(
 	/* in 10 khz units.*/
 	engine_clock_parameters.sReserved.ulClock =
 		cpu_to_le32(memory_clock & SET_CLOCK_FREQ_MASK);
-	return cgs_atom_exec_cmd_table(hwmgr->device,
+
+	return amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, DynamicMemorySettings),
-			&engine_clock_parameters);
+			(uint32_t *)&engine_clock_parameters);
 }
 
 /**
@@ -203,7 +205,7 @@ static ATOM_VOLTAGE_OBJECT_INFO *get_voltage_info_table(void *device)
 	union voltage_object_info *voltage_info;
 
 	voltage_info = (union voltage_object_info *)
-		cgs_atom_get_data_table(device, index,
+		smu_atom_get_data_table(device, index,
 			&size, &frev, &crev);
 
 	if (voltage_info != NULL)
@@ -247,16 +249,16 @@ int atomctrl_get_memory_pll_dividers_si(
 		pp_atomctrl_memory_clock_param *mpll_param,
 		bool strobe_mode)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	COMPUTE_MEMORY_CLOCK_PARAM_PARAMETERS_V2_1 mpll_parameters;
 	int result;
 
 	mpll_parameters.ulClock = cpu_to_le32(clock_value);
 	mpll_parameters.ucInputFlag = (uint8_t)((strobe_mode) ? 1 : 0);
 
-	result = cgs_atom_exec_cmd_table
-		(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 		 GetIndexIntoMasterTable(COMMAND, ComputeMemoryClockParam),
-		 &mpll_parameters);
+		(uint32_t *)&mpll_parameters);
 
 	if (0 == result) {
 		mpll_param->mpll_fb_divider.clk_frac =
@@ -295,14 +297,15 @@ int atomctrl_get_memory_pll_dividers_si(
 int atomctrl_get_memory_pll_dividers_vi(struct pp_hwmgr *hwmgr,
 		uint32_t clock_value, pp_atomctrl_memory_clock_param *mpll_param)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	COMPUTE_MEMORY_CLOCK_PARAM_PARAMETERS_V2_2 mpll_parameters;
 	int result;
 
 	mpll_parameters.ulClock.ulClock = cpu_to_le32(clock_value);
 
-	result = cgs_atom_exec_cmd_table(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, ComputeMemoryClockParam),
-			&mpll_parameters);
+			(uint32_t *)&mpll_parameters);
 
 	if (!result)
 		mpll_param->mpll_post_divider =
@@ -315,15 +318,15 @@ int atomctrl_get_engine_pll_dividers_kong(struct pp_hwmgr *hwmgr,
 					  uint32_t clock_value,
 					  pp_atomctrl_clock_dividers_kong *dividers)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	COMPUTE_MEMORY_ENGINE_PLL_PARAMETERS_V4 pll_parameters;
 	int result;
 
 	pll_parameters.ulClock = cpu_to_le32(clock_value);
 
-	result = cgs_atom_exec_cmd_table
-		(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 		 GetIndexIntoMasterTable(COMMAND, ComputeMemoryEnginePLL),
-		 &pll_parameters);
+		(uint32_t *)&pll_parameters);
 
 	if (0 == result) {
 		dividers->pll_post_divider = pll_parameters.ucPostDiv;
@@ -338,16 +341,16 @@ int atomctrl_get_engine_pll_dividers_vi(
 		uint32_t clock_value,
 		pp_atomctrl_clock_dividers_vi *dividers)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	COMPUTE_GPU_CLOCK_OUTPUT_PARAMETERS_V1_6 pll_patameters;
 	int result;
 
 	pll_patameters.ulClock.ulClock = cpu_to_le32(clock_value);
 	pll_patameters.ulClock.ucPostDiv = COMPUTE_GPUCLK_INPUT_FLAG_SCLK;
 
-	result = cgs_atom_exec_cmd_table
-		(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 		 GetIndexIntoMasterTable(COMMAND, ComputeMemoryEnginePLL),
-		 &pll_patameters);
+		(uint32_t *)&pll_patameters);
 
 	if (0 == result) {
 		dividers->pll_post_divider =
@@ -375,16 +378,16 @@ int atomctrl_get_engine_pll_dividers_ai(struct pp_hwmgr *hwmgr,
 		uint32_t clock_value,
 		pp_atomctrl_clock_dividers_ai *dividers)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	COMPUTE_GPU_CLOCK_OUTPUT_PARAMETERS_V1_7 pll_patameters;
 	int result;
 
 	pll_patameters.ulClock.ulClock = cpu_to_le32(clock_value);
 	pll_patameters.ulClock.ucPostDiv = COMPUTE_GPUCLK_INPUT_FLAG_SCLK;
 
-	result = cgs_atom_exec_cmd_table
-		(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 		 GetIndexIntoMasterTable(COMMAND, ComputeMemoryEnginePLL),
-		 &pll_patameters);
+		(uint32_t *)&pll_patameters);
 
 	if (0 == result) {
 		dividers->usSclk_fcw_frac     = le16_to_cpu(pll_patameters.usSclk_fcw_frac);
@@ -407,6 +410,7 @@ int atomctrl_get_dfs_pll_dividers_vi(
 		uint32_t clock_value,
 		pp_atomctrl_clock_dividers_vi *dividers)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	COMPUTE_GPU_CLOCK_OUTPUT_PARAMETERS_V1_6 pll_patameters;
 	int result;
 
@@ -414,10 +418,9 @@ int atomctrl_get_dfs_pll_dividers_vi(
 	pll_patameters.ulClock.ucPostDiv =
 		COMPUTE_GPUCLK_INPUT_FLAG_DEFAULT_GPUCLK;
 
-	result = cgs_atom_exec_cmd_table
-		(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 		 GetIndexIntoMasterTable(COMMAND, ComputeMemoryEnginePLL),
-		 &pll_patameters);
+		(uint32_t *)&pll_patameters);
 
 	if (0 == result) {
 		dividers->pll_post_divider =
@@ -452,7 +455,7 @@ uint32_t atomctrl_get_reference_clock(struct pp_hwmgr *hwmgr)
 	uint32_t clock;
 
 	fw_info = (ATOM_FIRMWARE_INFO *)
-		cgs_atom_get_data_table(hwmgr->device,
+		smu_atom_get_data_table(hwmgr->adev,
 			GetIndexIntoMasterTable(DATA, FirmwareInfo),
 			&size, &frev, &crev);
 
@@ -476,7 +479,7 @@ bool atomctrl_is_voltage_controlled_by_gpio_v3(
 		uint8_t voltage_mode)
 {
 	ATOM_VOLTAGE_OBJECT_INFO_V3_1 *voltage_info =
-		(ATOM_VOLTAGE_OBJECT_INFO_V3_1 *)get_voltage_info_table(hwmgr->device);
+		(ATOM_VOLTAGE_OBJECT_INFO_V3_1 *)get_voltage_info_table(hwmgr->adev);
 	bool ret;
 
 	PP_ASSERT_WITH_CODE((NULL != voltage_info),
@@ -495,7 +498,7 @@ int atomctrl_get_voltage_table_v3(
 		pp_atomctrl_voltage_table *voltage_table)
 {
 	ATOM_VOLTAGE_OBJECT_INFO_V3_1 *voltage_info =
-		(ATOM_VOLTAGE_OBJECT_INFO_V3_1 *)get_voltage_info_table(hwmgr->device);
+		(ATOM_VOLTAGE_OBJECT_INFO_V3_1 *)get_voltage_info_table(hwmgr->adev);
 	const ATOM_VOLTAGE_OBJECT_V3 *voltage_object;
 	unsigned int i;
 
@@ -572,7 +575,7 @@ static ATOM_GPIO_PIN_LUT *get_gpio_lookup_table(void *device)
 	void *table_address;
 
 	table_address = (ATOM_GPIO_PIN_LUT *)
-		cgs_atom_get_data_table(device,
+		smu_atom_get_data_table(device,
 				GetIndexIntoMasterTable(DATA, GPIO_Pin_LUT),
 				&size, &frev, &crev);
 
@@ -592,7 +595,7 @@ bool atomctrl_get_pp_assign_pin(
 {
 	bool bRet = false;
 	ATOM_GPIO_PIN_LUT *gpio_lookup_table =
-		get_gpio_lookup_table(hwmgr->device);
+		get_gpio_lookup_table(hwmgr->adev);
 
 	PP_ASSERT_WITH_CODE((NULL != gpio_lookup_table),
 			"Could not find GPIO lookup Table in BIOS.", return false);
@@ -613,7 +616,7 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 		bool debug)
 {
 	ATOM_ASIC_PROFILING_INFO_V3_4 *getASICProfilingInfo;
-
+	struct amdgpu_device *adev = hwmgr->adev;
 	EFUSE_LINEAR_FUNC_PARAM sRO_fuse;
 	EFUSE_LINEAR_FUNC_PARAM sCACm_fuse;
 	EFUSE_LINEAR_FUNC_PARAM sCACb_fuse;
@@ -640,7 +643,7 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 	int result;
 
 	getASICProfilingInfo = (ATOM_ASIC_PROFILING_INFO_V3_4 *)
-			cgs_atom_get_data_table(hwmgr->device,
+			smu_atom_get_data_table(hwmgr->adev,
 					GetIndexIntoMasterTable(DATA, ASIC_ProfilingInfo),
 					NULL, NULL, NULL);
 
@@ -706,9 +709,9 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 
 	sOutput_FuseValues.sEfuse = sInput_FuseValues;
 
-	result = cgs_atom_exec_cmd_table(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, ReadEfuseValue),
-			&sOutput_FuseValues);
+			(uint32_t *)&sOutput_FuseValues);
 
 	if (result)
 		return result;
@@ -727,9 +730,9 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 
 	sOutput_FuseValues.sEfuse = sInput_FuseValues;
 
-	result = cgs_atom_exec_cmd_table(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, ReadEfuseValue),
-			&sOutput_FuseValues);
+			(uint32_t *)&sOutput_FuseValues);
 
 	if (result)
 		return result;
@@ -747,9 +750,9 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 	sInput_FuseValues.ucBitLength = sCACb_fuse.ucEfuseLength;
 	sOutput_FuseValues.sEfuse = sInput_FuseValues;
 
-	result = cgs_atom_exec_cmd_table(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, ReadEfuseValue),
-			&sOutput_FuseValues);
+			(uint32_t *)&sOutput_FuseValues);
 
 	if (result)
 		return result;
@@ -768,9 +771,9 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 
 	sOutput_FuseValues.sEfuse = sInput_FuseValues;
 
-	result = cgs_atom_exec_cmd_table(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, ReadEfuseValue),
-			&sOutput_FuseValues);
+			(uint32_t *)&sOutput_FuseValues);
 
 	if (result)
 		return result;
@@ -790,9 +793,9 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 
 	sOutput_FuseValues.sEfuse = sInput_FuseValues;
 
-	result = cgs_atom_exec_cmd_table(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, ReadEfuseValue),
-			&sOutput_FuseValues);
+			(uint32_t *)&sOutput_FuseValues);
 	if (result)
 		return result;
 
@@ -811,9 +814,9 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 	sInput_FuseValues.ucBitLength = sKv_b_fuse.ucEfuseLength;
 	sOutput_FuseValues.sEfuse = sInput_FuseValues;
 
-	result = cgs_atom_exec_cmd_table(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, ReadEfuseValue),
-			&sOutput_FuseValues);
+			(uint32_t *)&sOutput_FuseValues);
 
 	if (result)
 		return result;
@@ -842,9 +845,9 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 
 	sOutput_FuseValues.sEfuse = sInput_FuseValues;
 
-	result = cgs_atom_exec_cmd_table(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, ReadEfuseValue),
-			&sOutput_FuseValues);
+			(uint32_t *)&sOutput_FuseValues);
 
 	if (result)
 		return result;
@@ -1053,8 +1056,9 @@ int atomctrl_get_voltage_evv_on_sclk(
 		uint32_t sclk, uint16_t virtual_voltage_Id,
 		uint16_t *voltage)
 {
-	int result;
+	struct amdgpu_device *adev = hwmgr->adev;
 	GET_VOLTAGE_INFO_INPUT_PARAMETER_V1_2 get_voltage_info_param_space;
+	int result;
 
 	get_voltage_info_param_space.ucVoltageType   =
 		voltage_type;
@@ -1065,9 +1069,9 @@ int atomctrl_get_voltage_evv_on_sclk(
 	get_voltage_info_param_space.ulSCLKFreq      =
 		cpu_to_le32(sclk);
 
-	result = cgs_atom_exec_cmd_table(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, GetVoltageInfo),
-			&get_voltage_info_param_space);
+			(uint32_t *)&get_voltage_info_param_space);
 
 	if (0 != result)
 		return result;
@@ -1088,9 +1092,10 @@ int atomctrl_get_voltage_evv(struct pp_hwmgr *hwmgr,
 			     uint16_t virtual_voltage_id,
 			     uint16_t *voltage)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
+	GET_VOLTAGE_INFO_INPUT_PARAMETER_V1_2 get_voltage_info_param_space;
 	int result;
 	int entry_id;
-	GET_VOLTAGE_INFO_INPUT_PARAMETER_V1_2 get_voltage_info_param_space;
 
 	/* search for leakage voltage ID 0xff01 ~ 0xff08 and sckl */
 	for (entry_id = 0; entry_id < hwmgr->dyn_state.vddc_dependency_on_sclk->count; entry_id++) {
@@ -1111,9 +1116,9 @@ int atomctrl_get_voltage_evv(struct pp_hwmgr *hwmgr,
 	get_voltage_info_param_space.ulSCLKFreq =
 		cpu_to_le32(hwmgr->dyn_state.vddc_dependency_on_sclk->entries[entry_id].clk);
 
-	result = cgs_atom_exec_cmd_table(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, GetVoltageInfo),
-			&get_voltage_info_param_space);
+			(uint32_t *)&get_voltage_info_param_space);
 
 	if (0 != result)
 		return result;
@@ -1135,7 +1140,7 @@ uint32_t atomctrl_get_mpll_reference_clock(struct pp_hwmgr *hwmgr)
 	u16 size;
 
 	fw_info = (ATOM_COMMON_TABLE_HEADER *)
-		cgs_atom_get_data_table(hwmgr->device,
+		smu_atom_get_data_table(hwmgr->adev,
 				GetIndexIntoMasterTable(DATA, FirmwareInfo),
 				&size, &frev, &crev);
 
@@ -1167,7 +1172,7 @@ static ATOM_ASIC_INTERNAL_SS_INFO *asic_internal_ss_get_ss_table(void *device)
 	u16 size;
 
 	table = (ATOM_ASIC_INTERNAL_SS_INFO *)
-		cgs_atom_get_data_table(device,
+		smu_atom_get_data_table(device,
 			GetIndexIntoMasterTable(DATA, ASIC_InternalSS_Info),
 			&size, &frev, &crev);
 
@@ -1188,7 +1193,7 @@ static int asic_internal_ss_get_ss_asignment(struct pp_hwmgr *hwmgr,
 
 	memset(ssEntry, 0x00, sizeof(pp_atomctrl_internal_ss_info));
 
-	table = asic_internal_ss_get_ss_table(hwmgr->device);
+	table = asic_internal_ss_get_ss_table(hwmgr->adev);
 
 	if (NULL == table)
 		return -1;
@@ -1260,9 +1265,10 @@ int atomctrl_get_engine_clock_spread_spectrum(
 			ASIC_INTERNAL_ENGINE_SS, engine_clock, ssInfo);
 }
 
-int atomctrl_read_efuse(void *device, uint16_t start_index,
+int atomctrl_read_efuse(struct pp_hwmgr *hwmgr, uint16_t start_index,
 		uint16_t end_index, uint32_t mask, uint32_t *efuse)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	int result;
 	READ_EFUSE_VALUE_PARAMETER efuse_param;
 
@@ -1272,9 +1278,9 @@ int atomctrl_read_efuse(void *device, uint16_t start_index,
 	efuse_param.sEfuse.ucBitLength  = (uint8_t)
 			((end_index - start_index) + 1);
 
-	result = cgs_atom_exec_cmd_table(device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, ReadEfuseValue),
-			&efuse_param);
+			(uint32_t *)&efuse_param);
 	if (!result)
 		*efuse = le32_to_cpu(efuse_param.ulEfuseValue) & mask;
 
@@ -1284,6 +1290,7 @@ int atomctrl_read_efuse(void *device, uint16_t start_index,
 int atomctrl_set_ac_timing_ai(struct pp_hwmgr *hwmgr, uint32_t memory_clock,
 			      uint8_t level)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	DYNAMICE_MEMORY_SETTINGS_PARAMETER_V2_1 memory_clock_parameters;
 	int result;
 
@@ -1293,10 +1300,9 @@ int atomctrl_set_ac_timing_ai(struct pp_hwmgr *hwmgr, uint32_t memory_clock,
 		ADJUST_MC_SETTING_PARAM;
 	memory_clock_parameters.asDPMMCReg.ucMclkDPMState = level;
 
-	result = cgs_atom_exec_cmd_table
-		(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 		 GetIndexIntoMasterTable(COMMAND, DynamicMemorySettings),
-		 &memory_clock_parameters);
+		(uint32_t *)&memory_clock_parameters);
 
 	return result;
 }
@@ -1304,7 +1310,7 @@ int atomctrl_set_ac_timing_ai(struct pp_hwmgr *hwmgr, uint32_t memory_clock,
 int atomctrl_get_voltage_evv_on_sclk_ai(struct pp_hwmgr *hwmgr, uint8_t voltage_type,
 				uint32_t sclk, uint16_t virtual_voltage_Id, uint32_t *voltage)
 {
-
+	struct amdgpu_device *adev = hwmgr->adev;
 	int result;
 	GET_VOLTAGE_INFO_INPUT_PARAMETER_V1_3 get_voltage_info_param_space;
 
@@ -1313,9 +1319,9 @@ int atomctrl_get_voltage_evv_on_sclk_ai(struct pp_hwmgr *hwmgr, uint8_t voltage_
 	get_voltage_info_param_space.usVoltageLevel = cpu_to_le16(virtual_voltage_Id);
 	get_voltage_info_param_space.ulSCLKFreq = cpu_to_le32(sclk);
 
-	result = cgs_atom_exec_cmd_table(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, GetVoltageInfo),
-			&get_voltage_info_param_space);
+			(uint32_t *)&get_voltage_info_param_space);
 
 	if (0 != result)
 		return result;
@@ -1334,7 +1340,7 @@ int atomctrl_get_smc_sclk_range_table(struct pp_hwmgr *hwmgr, struct pp_atom_ctr
 	u16 size;
 
 	ATOM_SMU_INFO_V2_1 *psmu_info =
-		(ATOM_SMU_INFO_V2_1 *)cgs_atom_get_data_table(hwmgr->device,
+		(ATOM_SMU_INFO_V2_1 *)smu_atom_get_data_table(hwmgr->adev,
 			GetIndexIntoMasterTable(DATA, SMU_Info),
 			&size, &frev, &crev);
 
@@ -1362,7 +1368,7 @@ int atomctrl_get_avfs_information(struct pp_hwmgr *hwmgr,
 		return -EINVAL;
 
 	profile = (ATOM_ASIC_PROFILING_INFO_V3_6 *)
-			cgs_atom_get_data_table(hwmgr->device,
+			smu_atom_get_data_table(hwmgr->adev,
 					GetIndexIntoMasterTable(DATA, ASIC_ProfilingInfo),
 					NULL, NULL, NULL);
 	if (!profile)
@@ -1402,7 +1408,7 @@ int  atomctrl_get_svi2_info(struct pp_hwmgr *hwmgr, uint8_t voltage_type,
 				uint16_t *load_line)
 {
 	ATOM_VOLTAGE_OBJECT_INFO_V3_1 *voltage_info =
-		(ATOM_VOLTAGE_OBJECT_INFO_V3_1 *)get_voltage_info_table(hwmgr->device);
+		(ATOM_VOLTAGE_OBJECT_INFO_V3_1 *)get_voltage_info_table(hwmgr->adev);
 
 	const ATOM_VOLTAGE_OBJECT_V3 *voltage_object;
 
@@ -1421,16 +1427,17 @@ int  atomctrl_get_svi2_info(struct pp_hwmgr *hwmgr, uint8_t voltage_type,
 
 int atomctrl_get_leakage_id_from_efuse(struct pp_hwmgr *hwmgr, uint16_t *virtual_voltage_id)
 {
-	int result;
+	struct amdgpu_device *adev = hwmgr->adev;
 	SET_VOLTAGE_PS_ALLOCATION allocation;
 	SET_VOLTAGE_PARAMETERS_V1_3 *voltage_parameters =
 			(SET_VOLTAGE_PARAMETERS_V1_3 *)&allocation.sASICSetVoltage;
+	int result;
 
 	voltage_parameters->ucVoltageMode = ATOM_GET_LEAKAGE_ID;
 
-	result = cgs_atom_exec_cmd_table(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, SetVoltage),
-			voltage_parameters);
+			(uint32_t *)voltage_parameters);
 
 	*virtual_voltage_id = voltage_parameters->usVoltageLevel;
 
@@ -1453,7 +1460,7 @@ int atomctrl_get_leakage_vddc_base_on_leakage(struct pp_hwmgr *hwmgr,
 	ix = GetIndexIntoMasterTable(DATA, ASIC_ProfilingInfo);
 
 	profile = (ATOM_ASIC_PROFILING_INFO_V2_1 *)
-			cgs_atom_get_data_table(hwmgr->device,
+			smu_atom_get_data_table(hwmgr->adev,
 					ix,
 					NULL, NULL, NULL);
 	if (!profile)
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h
index c44a920..c672a50 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h
@@ -298,7 +298,7 @@ extern int atomctrl_get_memory_pll_dividers_vi(struct pp_hwmgr *hwmgr,
 extern int atomctrl_get_engine_pll_dividers_kong(struct pp_hwmgr *hwmgr,
 						 uint32_t clock_value,
 						 pp_atomctrl_clock_dividers_kong *dividers);
-extern int atomctrl_read_efuse(void *device, uint16_t start_index,
+extern int atomctrl_read_efuse(struct pp_hwmgr *hwmgr, uint16_t start_index,
 		uint16_t end_index, uint32_t mask, uint32_t *efuse);
 extern int atomctrl_calculate_voltage_evv_on_sclk(struct pp_hwmgr *hwmgr, uint8_t voltage_type,
 		uint32_t sclk, uint16_t virtual_voltage_Id, uint16_t *voltage, uint16_t dpm_level, bool debug);
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c
index ad42caa..0adaf36 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c
@@ -23,9 +23,9 @@
 
 #include "ppatomfwctrl.h"
 #include "atomfirmware.h"
+#include "atom.h"
 #include "pp_debug.h"
 
-
 static const union atom_voltage_object_v4 *pp_atomfwctrl_lookup_voltage_type_v4(
 		const struct atom_voltage_objects_info_v4_1 *voltage_object_info_table,
 		uint8_t voltage_type, uint8_t voltage_mode)
@@ -38,35 +38,34 @@ static const union atom_voltage_object_v4 *pp_atomfwctrl_lookup_voltage_type_v4(
 
 	while (offset < size) {
 		const union atom_voltage_object_v4 *voltage_object =
-				(const union atom_voltage_object_v4 *)(start + offset);
+			(const union atom_voltage_object_v4 *)(start + offset);
 
-        if (voltage_type == voltage_object->gpio_voltage_obj.header.voltage_type &&
-            voltage_mode == voltage_object->gpio_voltage_obj.header.voltage_mode)
-            return voltage_object;
+		if (voltage_type == voltage_object->gpio_voltage_obj.header.voltage_type &&
+		    voltage_mode == voltage_object->gpio_voltage_obj.header.voltage_mode)
+			return voltage_object;
 
-        offset += le16_to_cpu(voltage_object->gpio_voltage_obj.header.object_size);
+		offset += le16_to_cpu(voltage_object->gpio_voltage_obj.header.object_size);
 
-    }
+	}
 
-    return NULL;
+	return NULL;
 }
 
 static struct atom_voltage_objects_info_v4_1 *pp_atomfwctrl_get_voltage_info_table(
 		struct pp_hwmgr *hwmgr)
 {
-    const void *table_address;
-    uint16_t idx;
+	const void *table_address;
+	uint16_t idx;
 
-    idx = GetIndexIntoMasterDataTable(voltageobject_info);
-    table_address =	cgs_atom_get_data_table(hwmgr->device,
-    		idx, NULL, NULL, NULL);
+	idx = GetIndexIntoMasterDataTable(voltageobject_info);
+	table_address = smu_atom_get_data_table(hwmgr->adev,
+						idx, NULL, NULL, NULL);
 
-    PP_ASSERT_WITH_CODE( 
-        table_address,
-        "Error retrieving BIOS Table Address!",
-        return NULL);
+	PP_ASSERT_WITH_CODE(table_address,
+			"Error retrieving BIOS Table Address!",
+			return NULL);
 
-    return (struct atom_voltage_objects_info_v4_1 *)table_address;
+	return (struct atom_voltage_objects_info_v4_1 *)table_address;
 }
 
 /**
@@ -167,7 +166,7 @@ static struct atom_gpio_pin_lut_v2_1 *pp_atomfwctrl_get_gpio_lookup_table(
 	uint16_t idx;
 
 	idx = GetIndexIntoMasterDataTable(gpio_pin_lut);
-	table_address =	cgs_atom_get_data_table(hwmgr->device,
+	table_address =	smu_atom_get_data_table(hwmgr->adev,
 			idx, NULL, NULL, NULL);
 	PP_ASSERT_WITH_CODE(table_address,
 			"Error retrieving BIOS Table Address!",
@@ -248,28 +247,30 @@ int pp_atomfwctrl_get_gpu_pll_dividers_vega10(struct pp_hwmgr *hwmgr,
 		uint32_t clock_type, uint32_t clock_value,
 		struct pp_atomfwctrl_clock_dividers_soc15 *dividers)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	struct compute_gpu_clock_input_parameter_v1_8 pll_parameters;
 	struct compute_gpu_clock_output_parameter_v1_8 *pll_output;
-	int result;
 	uint32_t idx;
 
 	pll_parameters.gpuclock_10khz = (uint32_t)clock_value;
 	pll_parameters.gpu_clock_type = clock_type;
 
 	idx = GetIndexIntoMasterCmdTable(computegpuclockparam);
-	result = cgs_atom_exec_cmd_table(hwmgr->device, idx, &pll_parameters);
-
-	if (!result) {
-		pll_output = (struct compute_gpu_clock_output_parameter_v1_8 *)
-				&pll_parameters;
-		dividers->ulClock = le32_to_cpu(pll_output->gpuclock_10khz);
-		dividers->ulDid = le32_to_cpu(pll_output->dfs_did);
-		dividers->ulPll_fb_mult = le32_to_cpu(pll_output->pll_fb_mult);
-		dividers->ulPll_ss_fbsmult = le32_to_cpu(pll_output->pll_ss_fbsmult);
-		dividers->usPll_ss_slew_frac = le16_to_cpu(pll_output->pll_ss_slew_frac);
-		dividers->ucPll_ss_enable = pll_output->pll_ss_enable;
-	}
-	return result;
+
+	if (amdgpu_atom_execute_table(
+		adev->mode_info.atom_context, idx, (uint32_t *)&pll_parameters))
+		return -EINVAL;
+
+	pll_output = (struct compute_gpu_clock_output_parameter_v1_8 *)
+			&pll_parameters;
+	dividers->ulClock = le32_to_cpu(pll_output->gpuclock_10khz);
+	dividers->ulDid = le32_to_cpu(pll_output->dfs_did);
+	dividers->ulPll_fb_mult = le32_to_cpu(pll_output->pll_fb_mult);
+	dividers->ulPll_ss_fbsmult = le32_to_cpu(pll_output->pll_ss_fbsmult);
+	dividers->usPll_ss_slew_frac = le16_to_cpu(pll_output->pll_ss_slew_frac);
+	dividers->ucPll_ss_enable = pll_output->pll_ss_enable;
+
+	return 0;
 }
 
 int pp_atomfwctrl_get_avfs_information(struct pp_hwmgr *hwmgr,
@@ -283,7 +284,7 @@ int pp_atomfwctrl_get_avfs_information(struct pp_hwmgr *hwmgr,
 
 	idx = GetIndexIntoMasterDataTable(asic_profiling_info);
 	profile = (struct atom_asic_profiling_info_v4_1 *)
-			cgs_atom_get_data_table(hwmgr->device,
+			smu_atom_get_data_table(hwmgr->adev,
 					idx, NULL, NULL, NULL);
 
 	if (!profile)
@@ -467,7 +468,7 @@ int pp_atomfwctrl_get_gpio_information(struct pp_hwmgr *hwmgr,
 
 	idx = GetIndexIntoMasterDataTable(smu_info);
 	info = (struct atom_smu_info_v3_1 *)
-		cgs_atom_get_data_table(hwmgr->device,
+		smu_atom_get_data_table(hwmgr->adev,
 				idx, NULL, NULL, NULL);
 
 	if (!info) {
@@ -489,6 +490,7 @@ int pp_atomfwctrl_get_gpio_information(struct pp_hwmgr *hwmgr,
 
 int pp_atomfwctrl__get_clk_information_by_clkid(struct pp_hwmgr *hwmgr, BIOS_CLKID id, uint32_t *frequency)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	struct atom_get_smu_clock_info_parameters_v3_1   parameters;
 	struct atom_get_smu_clock_info_output_parameters_v3_1 *output;
 	uint32_t ix;
@@ -497,13 +499,13 @@ int pp_atomfwctrl__get_clk_information_by_clkid(struct pp_hwmgr *hwmgr, BIOS_CLK
 	parameters.command = GET_SMU_CLOCK_INFO_V3_1_GET_CLOCK_FREQ;
 
 	ix = GetIndexIntoMasterCmdTable(getsmuclockinfo);
-	if (!cgs_atom_exec_cmd_table(hwmgr->device, ix, &parameters)) {
-		output = (struct atom_get_smu_clock_info_output_parameters_v3_1 *)&parameters;
-		*frequency = output->atom_smu_outputclkfreq.smu_clock_freq_hz / 10000;
-	} else {
-		pr_info("Error execute_table getsmuclockinfo!");
-		return -1;
-	}
+
+	if (amdgpu_atom_execute_table(
+		adev->mode_info.atom_context, ix, (uint32_t *)&parameters))
+		return -EINVAL;
+
+	output = (struct atom_get_smu_clock_info_output_parameters_v3_1 *)&parameters;
+	*frequency = output->atom_smu_outputclkfreq.smu_clock_freq_hz / 10000;
 
 	return 0;
 }
@@ -517,7 +519,7 @@ int pp_atomfwctrl_get_vbios_bootup_values(struct pp_hwmgr *hwmgr,
 
 	ix = GetIndexIntoMasterDataTable(firmwareinfo);
 	info = (struct atom_firmware_info_v3_1 *)
-		cgs_atom_get_data_table(hwmgr->device,
+		smu_atom_get_data_table(hwmgr->adev,
 				ix, NULL, NULL, NULL);
 
 	if (!info) {
@@ -553,7 +555,7 @@ int pp_atomfwctrl_get_smc_dpm_information(struct pp_hwmgr *hwmgr,
 
 	ix = GetIndexIntoMasterDataTable(smc_dpm_info);
 	info = (struct atom_smc_dpm_info_v4_1 *)
-		cgs_atom_get_data_table(hwmgr->device,
+		smu_atom_get_data_table(hwmgr->adev,
 				ix, NULL, NULL, NULL);
 	if (!info) {
 		pr_info("Error retrieving BIOS Table Address!");
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.c b/drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.c
index c9eecce..8516516 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.c
@@ -141,7 +141,7 @@ static const void *get_powerplay_table(struct pp_hwmgr *hwmgr)
 
 	if (!table_address) {
 		table_address = (ATOM_Tonga_POWERPLAYTABLE *)
-				cgs_atom_get_data_table(hwmgr->device,
+				smu_atom_get_data_table(hwmgr->adev,
 						index, &size, &frev, &crev);
 		hwmgr->soft_pp_table = table_address;	/*Cache the result in RAM.*/
 		hwmgr->soft_pp_table_size = size;
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c b/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c
index 36ca7c4..ce64dfa 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c
@@ -837,7 +837,7 @@ static const ATOM_PPLIB_POWERPLAYTABLE *get_powerplay_table(
 			hwmgr->soft_pp_table = &soft_dummy_pp_table[0];
 			hwmgr->soft_pp_table_size = sizeof(soft_dummy_pp_table);
 		} else {
-			table_addr = cgs_atom_get_data_table(hwmgr->device,
+			table_addr = smu_atom_get_data_table(hwmgr->adev,
 					GetIndexIntoMasterTable(DATA, PowerPlayInfo),
 					&size, &frev, &crev);
 			hwmgr->soft_pp_table = table_addr;
@@ -1058,7 +1058,7 @@ static int init_overdrive_limits(struct pp_hwmgr *hwmgr,
 		return 0;
 
 	/* We assume here that fw_info is unchanged if this call fails.*/
-	fw_info = cgs_atom_get_data_table(hwmgr->device,
+	fw_info = smu_atom_get_data_table(hwmgr->adev,
 			 GetIndexIntoMasterTable(DATA, FirmwareInfo),
 			 &size, &frev, &crev);
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index add9067..9087ef9 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -2957,8 +2957,7 @@ static int smu7_dpm_patch_boot_state(struct pp_hwmgr *hwmgr,
 	/* First retrieve the Boot clocks and VDDC from the firmware info table.
 	 * We assume here that fw_info is unchanged if this call fails.
 	 */
-	fw_info = (ATOM_FIRMWARE_INFO_V2_2 *)cgs_atom_get_data_table(
-			hwmgr->device, index,
+	fw_info = (ATOM_FIRMWARE_INFO_V2_2 *)smu_atom_get_data_table(hwmgr->adev, index,
 			&size, &frev, &crev);
 	if (!fw_info)
 		/* During a test, there is no firmware info table. */
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu8_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu8_hwmgr.c
index 7b26607..3ac07fa 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu8_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu8_hwmgr.c
@@ -314,8 +314,7 @@ static int smu8_get_system_info_data(struct pp_hwmgr *hwmgr)
 	uint8_t frev, crev;
 	uint16_t size;
 
-	info = (ATOM_INTEGRATED_SYSTEM_INFO_V1_9 *) cgs_atom_get_data_table(
-			hwmgr->device,
+	info = (ATOM_INTEGRATED_SYSTEM_INFO_V1_9 *)smu_atom_get_data_table(hwmgr->adev,
 			GetIndexIntoMasterTable(DATA, IntegratedSystemInfo),
 			&size, &frev, &crev);
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.c
index 5981228..529be3c 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.c
@@ -24,6 +24,7 @@
 #include "pp_debug.h"
 #include "ppatomctrl.h"
 #include "ppsmc.h"
+#include "atom.h"
 
 uint8_t convert_to_vid(uint16_t vddc)
 {
@@ -608,3 +609,18 @@ int smu9_register_irq_handlers(struct pp_hwmgr *hwmgr)
 
 	return 0;
 }
+
+void *smu_atom_get_data_table(void *dev, uint32_t table, uint16_t *size,
+						uint8_t *frev, uint8_t *crev)
+{
+	struct amdgpu_device *adev = dev;
+	uint16_t data_start;
+
+	if (amdgpu_atom_parse_data_header(
+		    adev->mode_info.atom_context, table, size,
+		    frev, crev, &data_start))
+		return (uint8_t *)adev->mode_info.atom_context->bios +
+			data_start;
+
+	return NULL;
+}
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.h b/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.h
index d37d16e..14ee162 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.h
@@ -82,6 +82,9 @@ int phm_irq_process(struct amdgpu_device *adev,
 
 int smu9_register_irq_handlers(struct pp_hwmgr *hwmgr);
 
+void *smu_atom_get_data_table(void *dev, uint32_t table, uint16_t *size,
+						uint8_t *frev, uint8_t *crev);
+
 #define PHM_FIELD_SHIFT(reg, field) reg##__##field##__SHIFT
 #define PHM_FIELD_MASK(reg, field) reg##__##field##_MASK
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c
index c61d074..0768d25 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c
@@ -52,7 +52,7 @@ static const void *get_powerplay_table(struct pp_hwmgr *hwmgr)
 
 	if (!table_address) {
 		table_address = (ATOM_Vega10_POWERPLAYTABLE *)
-				cgs_atom_get_data_table(hwmgr->device, index,
+				smu_atom_get_data_table(hwmgr->adev, index,
 						&size, &frev, &crev);
 
 		hwmgr->soft_pp_table = table_address;	/*Cache the result in RAM.*/
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_processpptables.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_processpptables.c
index b34113f..7fa1ba8 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_processpptables.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_processpptables.c
@@ -51,7 +51,7 @@ static const void *get_powerplay_table(struct pp_hwmgr *hwmgr)
 
 	if (!table_address) {
 		table_address = (ATOM_Vega12_POWERPLAYTABLE *)
-				cgs_atom_get_data_table(hwmgr->device, index,
+				smu_atom_get_data_table(hwmgr->adev, index,
 						&size, &frev, &crev);
 
 		hwmgr->soft_pp_table = table_address;	/*Cache the result in RAM.*/
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c
index 35b947e..1eec527 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c
@@ -337,7 +337,7 @@ static bool fiji_is_hw_avfs_present(struct pp_hwmgr *hwmgr)
 	if (!hwmgr->not_vf)
 		return false;
 
-	if (!atomctrl_read_efuse(hwmgr->device, AVFS_EN_LSB, AVFS_EN_MSB,
+	if (!atomctrl_read_efuse(hwmgr, AVFS_EN_LSB, AVFS_EN_MSB,
 			mask, &efuse)) {
 		if (efuse)
 			return true;
-- 
cgit v1.1


From d91ea4969bc5edbbe3bd723a1b3ae7d947f62a5a Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Mon, 26 Mar 2018 22:08:29 +0800
Subject: drm/amdgpu: Set pm_display_cfg in non-dc mode

those display informations are needed by powerplay.

Reviewed-by: Huang Rui <ray.huang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c | 20 ++++++++++++++++++++
 drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h |  1 +
 drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c  | 31 ++++++++++++++++---------------
 3 files changed, 37 insertions(+), 15 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c
index e997ebbe43..def1010 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c
@@ -115,6 +115,26 @@ void amdgpu_dpm_print_ps_status(struct amdgpu_device *adev,
 	pr_cont("\n");
 }
 
+void amdgpu_dpm_get_active_displays(struct amdgpu_device *adev)
+{
+	struct drm_device *ddev = adev->ddev;
+	struct drm_crtc *crtc;
+	struct amdgpu_crtc *amdgpu_crtc;
+
+	adev->pm.dpm.new_active_crtcs = 0;
+	adev->pm.dpm.new_active_crtc_count = 0;
+	if (adev->mode_info.num_crtc && adev->mode_info.mode_config_initialized) {
+		list_for_each_entry(crtc,
+				    &ddev->mode_config.crtc_list, head) {
+			amdgpu_crtc = to_amdgpu_crtc(crtc);
+			if (amdgpu_crtc->enabled) {
+				adev->pm.dpm.new_active_crtcs |= (1 << amdgpu_crtc->crtc_id);
+				adev->pm.dpm.new_active_crtc_count++;
+			}
+		}
+	}
+}
+
 
 u32 amdgpu_dpm_get_vblank_time(struct amdgpu_device *adev)
 {
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
index 643d008..b8c5177 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
@@ -482,6 +482,7 @@ void amdgpu_dpm_print_ps_status(struct amdgpu_device *adev,
 				struct amdgpu_ps *rps);
 u32 amdgpu_dpm_get_vblank_time(struct amdgpu_device *adev);
 u32 amdgpu_dpm_get_vrefresh(struct amdgpu_device *adev);
+void amdgpu_dpm_get_active_displays(struct amdgpu_device *adev);
 bool amdgpu_is_uvd_state(u32 class, u32 class2);
 void amdgpu_calculate_u_and_p(u32 i, u32 r_c, u32 p_b,
 			      u32 *p, u32 *u);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
index 361975c..e6e3658 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
@@ -1658,9 +1658,6 @@ void amdgpu_pm_sysfs_fini(struct amdgpu_device *adev)
 
 void amdgpu_pm_compute_clocks(struct amdgpu_device *adev)
 {
-	struct drm_device *ddev = adev->ddev;
-	struct drm_crtc *crtc;
-	struct amdgpu_crtc *amdgpu_crtc;
 	int i = 0;
 
 	if (!adev->pm.dpm_enabled)
@@ -1675,22 +1672,26 @@ void amdgpu_pm_compute_clocks(struct amdgpu_device *adev)
 			amdgpu_fence_wait_empty(ring);
 	}
 
+	if (!amdgpu_device_has_dc_support(adev)) {
+		mutex_lock(&adev->pm.mutex);
+		amdgpu_dpm_get_active_displays(adev);
+		adev->pm.pm_display_cfg.num_display = adev->pm.dpm.new_active_crtcs;
+		adev->pm.pm_display_cfg.vrefresh = amdgpu_dpm_get_vrefresh(adev);
+		adev->pm.pm_display_cfg.min_vblank_time = amdgpu_dpm_get_vblank_time(adev);
+		/* we have issues with mclk switching with refresh rates over 120 hz on the non-DC code. */
+		if (adev->pm.pm_display_cfg.vrefresh > 120)
+			adev->pm.pm_display_cfg.min_vblank_time = 0;
+		if (adev->powerplay.pp_funcs->display_configuration_change)
+			adev->powerplay.pp_funcs->display_configuration_change(
+							adev->powerplay.pp_handle,
+							&adev->pm.pm_display_cfg);
+		mutex_unlock(&adev->pm.mutex);
+	}
+
 	if (adev->powerplay.pp_funcs->dispatch_tasks) {
 		amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_DISPLAY_CONFIG_CHANGE, NULL);
 	} else {
 		mutex_lock(&adev->pm.mutex);
-		adev->pm.dpm.new_active_crtcs = 0;
-		adev->pm.dpm.new_active_crtc_count = 0;
-		if (adev->mode_info.num_crtc && adev->mode_info.mode_config_initialized) {
-			list_for_each_entry(crtc,
-					    &ddev->mode_config.crtc_list, head) {
-				amdgpu_crtc = to_amdgpu_crtc(crtc);
-				if (amdgpu_crtc->enabled) {
-					adev->pm.dpm.new_active_crtcs |= (1 << amdgpu_crtc->crtc_id);
-					adev->pm.dpm.new_active_crtc_count++;
-				}
-			}
-		}
 		/* update battery/ac status */
 		if (power_supply_is_system_supplied() > 0)
 			adev->pm.dpm.ac_power = true;
-- 
cgit v1.1


From 555fd70c59bc7f7acd8bc429d92bd59a66a7b83b Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Tue, 27 Mar 2018 13:32:02 +0800
Subject: drm/amd/pp: Not call cgs interface to get display info

DC/Non DC all will update display configuration
when the display state changed
No need to get display info through cgs interface

Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/amd_powerplay.c      |  1 +
 .../gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c  |  8 ++--
 drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c  |  2 +-
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c   | 56 ++++++----------------
 drivers/gpu/drm/amd/powerplay/hwmgr/smu8_hwmgr.c   | 14 ++----
 drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c | 51 ++++++--------------
 drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c | 29 ++++-------
 drivers/gpu/drm/amd/powerplay/inc/hwmgr.h          |  2 +-
 drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c   |  4 +-
 drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c |  4 +-
 .../gpu/drm/amd/powerplay/smumgr/iceland_smumgr.c  |  6 +--
 .../drm/amd/powerplay/smumgr/polaris10_smumgr.c    |  8 ++--
 .../gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c    | 12 ++---
 13 files changed, 61 insertions(+), 136 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
index b91ef11..1ca6a13 100644
--- a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
+++ b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
@@ -54,6 +54,7 @@ static int amd_powerplay_create(struct amdgpu_device *adev)
 	hwmgr->chip_family = adev->family;
 	hwmgr->chip_id = adev->asic_type;
 	hwmgr->feature_mask = amdgpu_pp_feature_mask;
+	hwmgr->display_config = &adev->pm.pm_display_cfg;
 	adev->powerplay.pp_handle = hwmgr;
 	adev->powerplay.pp_funcs = &pp_dpm_funcs;
 	return 0;
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c
index dcceadb..e411012 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c
@@ -265,13 +265,11 @@ int phm_store_dal_configuration_data(struct pp_hwmgr *hwmgr,
 	if (display_config == NULL)
 		return -EINVAL;
 
-	hwmgr->display_config = *display_config;
-
 	if (NULL != hwmgr->hwmgr_func->set_deep_sleep_dcefclk)
-		hwmgr->hwmgr_func->set_deep_sleep_dcefclk(hwmgr, hwmgr->display_config.min_dcef_deep_sleep_set_clk);
+		hwmgr->hwmgr_func->set_deep_sleep_dcefclk(hwmgr, display_config->min_dcef_deep_sleep_set_clk);
 
-	for (index = 0; index < hwmgr->display_config.num_path_including_non_display; index++) {
-		if (hwmgr->display_config.displays[index].controller_id != 0)
+	for (index = 0; index < display_config->num_path_including_non_display; index++) {
+		if (display_config->displays[index].controller_id != 0)
 			number_of_active_display++;
 	}
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
index 10253b8..055358b 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
@@ -161,7 +161,7 @@ static int smu10_set_clock_limit(struct pp_hwmgr *hwmgr, const void *input)
 	struct PP_Clocks clocks = {0};
 	struct pp_display_clock_request clock_req;
 
-	clocks.dcefClock = hwmgr->display_config.min_dcef_set_clk;
+	clocks.dcefClock = hwmgr->display_config->min_dcef_set_clk;
 	clock_req.clock_type = amd_pp_dcf_clock;
 	clock_req.clock_freq_in_khz = clocks.dcefClock * 10;
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index 9087ef9..1433215 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -2777,8 +2777,6 @@ static int smu7_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 	struct PP_Clocks minimum_clocks = {0};
 	bool disable_mclk_switching;
 	bool disable_mclk_switching_for_frame_lock;
-	struct cgs_display_info info = {0};
-	struct cgs_mode_info mode_info = {0};
 	const struct phm_clock_and_voltage_limits *max_limits;
 	uint32_t i;
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
@@ -2787,7 +2785,6 @@ static int smu7_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 	int32_t count;
 	int32_t stable_pstate_sclk = 0, stable_pstate_mclk = 0;
 
-	info.mode_info = &mode_info;
 	data->battery_state = (PP_StateUILabel_Battery ==
 			request_ps->classification.ui_label);
 
@@ -2809,10 +2806,8 @@ static int smu7_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 		}
 	}
 
-	cgs_get_active_displays_info(hwmgr->device, &info);
-
-	minimum_clocks.engineClock = hwmgr->display_config.min_core_set_clock;
-	minimum_clocks.memoryClock = hwmgr->display_config.min_mem_set_clock;
+	minimum_clocks.engineClock = hwmgr->display_config->min_core_set_clock;
+	minimum_clocks.memoryClock = hwmgr->display_config->min_mem_set_clock;
 
 	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
 			PHM_PlatformCaps_StablePState)) {
@@ -2843,12 +2838,12 @@ static int smu7_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 				    PHM_PlatformCaps_DisableMclkSwitchingForFrameLock);
 
 
-	if (info.display_count == 0)
+	if (hwmgr->display_config->num_display == 0)
 		disable_mclk_switching = false;
 	else
-		disable_mclk_switching = ((1 < info.display_count) ||
+		disable_mclk_switching = ((1 < hwmgr->display_config->num_display) ||
 					  disable_mclk_switching_for_frame_lock ||
-					  smu7_vblank_too_short(hwmgr, mode_info.vblank_time_us));
+					  smu7_vblank_too_short(hwmgr, hwmgr->display_config->min_vblank_time));
 
 	sclk = smu7_ps->performance_levels[0].engine_clock;
 	mclk = smu7_ps->performance_levels[0].memory_clock;
@@ -3479,7 +3474,6 @@ static int smu7_find_dpm_states_clocks_in_dpm_table(struct pp_hwmgr *hwmgr, cons
 			[smu7_ps->performance_level_count - 1].memory_clock;
 	struct PP_Clocks min_clocks = {0};
 	uint32_t i;
-	struct cgs_display_info info = {0};
 
 	for (i = 0; i < sclk_table->count; i++) {
 		if (sclk == sclk_table->dpm_levels[i].value)
@@ -3506,9 +3500,8 @@ static int smu7_find_dpm_states_clocks_in_dpm_table(struct pp_hwmgr *hwmgr, cons
 	if (i >= mclk_table->count)
 		data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_MCLK;
 
-	cgs_get_active_displays_info(hwmgr->device, &info);
 
-	if (data->display_timing.num_existing_displays != info.display_count)
+	if (data->display_timing.num_existing_displays != hwmgr->display_config->num_display)
 		data->need_update_smu7_dpm_table |= DPMTABLE_UPDATE_MCLK;
 
 	return 0;
@@ -3907,15 +3900,8 @@ smu7_notify_smc_display_change(struct pp_hwmgr *hwmgr, bool has_display)
 static int
 smu7_notify_smc_display_config_after_ps_adjustment(struct pp_hwmgr *hwmgr)
 {
-	uint32_t num_active_displays = 0;
-	struct cgs_display_info info = {0};
-
-	info.mode_info = NULL;
-	cgs_get_active_displays_info(hwmgr->device, &info);
-
-	num_active_displays = info.display_count;
-
-	if (num_active_displays > 1 && hwmgr->display_config.multi_monitor_in_sync != true)
+	if (hwmgr->display_config->num_display > 1 &&
+			!hwmgr->display_config->multi_monitor_in_sync)
 		smu7_notify_smc_display_change(hwmgr, false);
 
 	return 0;
@@ -3930,33 +3916,24 @@ smu7_notify_smc_display_config_after_ps_adjustment(struct pp_hwmgr *hwmgr)
 static int smu7_program_display_gap(struct pp_hwmgr *hwmgr)
 {
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
-	uint32_t num_active_displays = 0;
 	uint32_t display_gap = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixCG_DISPLAY_GAP_CNTL);
 	uint32_t display_gap2;
 	uint32_t pre_vbi_time_in_us;
 	uint32_t frame_time_in_us;
-	uint32_t ref_clock;
-	uint32_t refresh_rate = 0;
-	struct cgs_display_info info = {0};
-	struct cgs_mode_info mode_info = {0};
+	uint32_t ref_clock, refresh_rate;
 
-	info.mode_info = &mode_info;
-	cgs_get_active_displays_info(hwmgr->device, &info);
-	num_active_displays = info.display_count;
-
-	display_gap = PHM_SET_FIELD(display_gap, CG_DISPLAY_GAP_CNTL, DISP_GAP, (num_active_displays > 0) ? DISPLAY_GAP_VBLANK_OR_WM : DISPLAY_GAP_IGNORE);
+	display_gap = PHM_SET_FIELD(display_gap, CG_DISPLAY_GAP_CNTL, DISP_GAP, (hwmgr->display_config->num_display > 0) ? DISPLAY_GAP_VBLANK_OR_WM : DISPLAY_GAP_IGNORE);
 	cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixCG_DISPLAY_GAP_CNTL, display_gap);
 
 	ref_clock =  amdgpu_asic_get_xclk((struct amdgpu_device *)hwmgr->adev);
-
-	refresh_rate = mode_info.refresh_rate;
+	refresh_rate = hwmgr->display_config->vrefresh;
 
 	if (0 == refresh_rate)
 		refresh_rate = 60;
 
 	frame_time_in_us = 1000000 / refresh_rate;
 
-	pre_vbi_time_in_us = frame_time_in_us - 200 - mode_info.vblank_time_us;
+	pre_vbi_time_in_us = frame_time_in_us - 200 - hwmgr->display_config->min_vblank_time;
 
 	data->frame_time_x2 = frame_time_in_us * 2 / 100;
 
@@ -4036,17 +4013,14 @@ smu7_check_smc_update_required_for_display_configuration(struct pp_hwmgr *hwmgr)
 {
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
 	bool is_update_required = false;
-	struct cgs_display_info info = {0, 0, NULL};
-
-	cgs_get_active_displays_info(hwmgr->device, &info);
 
-	if (data->display_timing.num_existing_displays != info.display_count)
+	if (data->display_timing.num_existing_displays != hwmgr->display_config->num_display)
 		is_update_required = true;
 
 	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_SclkDeepSleep)) {
-		if (data->display_timing.min_clock_in_sr != hwmgr->display_config.min_core_set_clock_in_sr &&
+		if (data->display_timing.min_clock_in_sr != hwmgr->display_config->min_core_set_clock_in_sr &&
 			(data->display_timing.min_clock_in_sr >= SMU7_MINIMUM_ENGINE_CLOCK ||
-			hwmgr->display_config.min_core_set_clock_in_sr >= SMU7_MINIMUM_ENGINE_CLOCK))
+			hwmgr->display_config->min_core_set_clock_in_sr >= SMU7_MINIMUM_ENGINE_CLOCK))
 			is_update_required = true;
 	}
 	return is_update_required;
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu8_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu8_hwmgr.c
index 3ac07fa..c2f93aa 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu8_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu8_hwmgr.c
@@ -693,7 +693,7 @@ static int smu8_update_sclk_limit(struct pp_hwmgr *hwmgr)
 	else
 		data->sclk_dpm.soft_max_clk  = table->entries[table->count - 1].clk;
 
-	clock = hwmgr->display_config.min_core_set_clock;
+	clock = hwmgr->display_config->min_core_set_clock;
 	if (clock == 0)
 		pr_debug("min_core_set_clock not set\n");
 
@@ -748,7 +748,7 @@ static int smu8_set_deep_sleep_sclk_threshold(struct pp_hwmgr *hwmgr)
 {
 	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
 				PHM_PlatformCaps_SclkDeepSleep)) {
-		uint32_t clks = hwmgr->display_config.min_core_set_clock_in_sr;
+		uint32_t clks = hwmgr->display_config->min_core_set_clock_in_sr;
 		if (clks == 0)
 			clks = SMU8_MIN_DEEP_SLEEP_SCLK;
 
@@ -1040,25 +1040,21 @@ static int smu8_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 	struct smu8_hwmgr *data = hwmgr->backend;
 	struct PP_Clocks clocks = {0, 0, 0, 0};
 	bool force_high;
-	uint32_t  num_of_active_displays = 0;
-	struct cgs_display_info info = {0};
 
 	smu8_ps->need_dfs_bypass = true;
 
 	data->battery_state = (PP_StateUILabel_Battery == prequest_ps->classification.ui_label);
 
-	clocks.memoryClock = hwmgr->display_config.min_mem_set_clock != 0 ?
-				hwmgr->display_config.min_mem_set_clock :
+	clocks.memoryClock = hwmgr->display_config->min_mem_set_clock != 0 ?
+				hwmgr->display_config->min_mem_set_clock :
 				data->sys_info.nbp_memory_clock[1];
 
-	cgs_get_active_displays_info(hwmgr->device, &info);
-	num_of_active_displays = info.display_count;
 
 	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_StablePState))
 		clocks.memoryClock = hwmgr->dyn_state.max_clock_voltage_on_ac.mclk;
 
 	force_high = (clocks.memoryClock > data->sys_info.nbp_memory_clock[SMU8_NUM_NBPMEMORYCLOCK - 1])
-			|| (num_of_active_displays >= 3);
+			|| (hwmgr->display_config->num_display >= 3);
 
 	smu8_ps->action = smu8_current_ps->action;
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
index 7cbb56b..c9fb4b2 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
@@ -3028,7 +3028,6 @@ static int vega10_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 	bool disable_mclk_switching_for_frame_lock;
 	bool disable_mclk_switching_for_vr;
 	bool force_mclk_high;
-	struct cgs_display_info info = {0};
 	const struct phm_clock_and_voltage_limits *max_limits;
 	uint32_t i;
 	struct vega10_hwmgr *data = hwmgr->backend;
@@ -3063,11 +3062,9 @@ static int vega10_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 		}
 	}
 
-	cgs_get_active_displays_info(hwmgr->device, &info);
-
 	/* result = PHM_CheckVBlankTime(hwmgr, &vblankTooShort);*/
-	minimum_clocks.engineClock = hwmgr->display_config.min_core_set_clock;
-	minimum_clocks.memoryClock = hwmgr->display_config.min_mem_set_clock;
+	minimum_clocks.engineClock = hwmgr->display_config->min_core_set_clock;
+	minimum_clocks.memoryClock = hwmgr->display_config->min_mem_set_clock;
 
 	if (PP_CAP(PHM_PlatformCaps_StablePState)) {
 		stable_pstate_sclk_dpm_percentage =
@@ -3107,10 +3104,10 @@ static int vega10_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 		PP_CAP(PHM_PlatformCaps_DisableMclkSwitchForVR);
 	force_mclk_high = PP_CAP(PHM_PlatformCaps_ForceMclkHigh);
 
-	if (info.display_count == 0)
+	if (hwmgr->display_config->num_display == 0)
 		disable_mclk_switching = false;
 	else
-		disable_mclk_switching = (info.display_count > 1) ||
+		disable_mclk_switching = (hwmgr->display_config->num_display > 1) ||
 			disable_mclk_switching_for_frame_lock ||
 			disable_mclk_switching_for_vr ||
 			force_mclk_high;
@@ -3186,7 +3183,6 @@ static int vega10_find_dpm_states_clocks_in_dpm_table(struct pp_hwmgr *hwmgr, co
 			[vega10_ps->performance_level_count - 1].mem_clock;
 	struct PP_Clocks min_clocks = {0};
 	uint32_t i;
-	struct cgs_display_info info = {0};
 
 	data->need_update_dpm_table = 0;
 
@@ -3211,10 +3207,8 @@ static int vega10_find_dpm_states_clocks_in_dpm_table(struct pp_hwmgr *hwmgr, co
 				data->need_update_dpm_table |= DPMTABLE_UPDATE_SCLK;
 		}
 
-		cgs_get_active_displays_info(hwmgr->device, &info);
-
 		if (data->display_timing.num_existing_displays !=
-				info.display_count)
+				hwmgr->display_config->num_display)
 			data->need_update_dpm_table |= DPMTABLE_UPDATE_MCLK;
 	} else {
 		for (i = 0; i < sclk_table->count; i++) {
@@ -3242,13 +3236,11 @@ static int vega10_find_dpm_states_clocks_in_dpm_table(struct pp_hwmgr *hwmgr, co
 				break;
 		}
 
-		cgs_get_active_displays_info(hwmgr->device, &info);
-
 		if (i >= mclk_table->count)
 			data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_MCLK;
 
 		if (data->display_timing.num_existing_displays !=
-				info.display_count ||
+				hwmgr->display_config->num_display ||
 				i >= mclk_table->count)
 			data->need_update_dpm_table |= DPMTABLE_UPDATE_MCLK;
 	}
@@ -3956,26 +3948,18 @@ static int vega10_notify_smc_display_config_after_ps_adjustment(
 			(struct phm_ppt_v2_information *)hwmgr->pptable;
 	struct phm_ppt_v1_clock_voltage_dependency_table *mclk_table = table_info->vdd_dep_on_mclk;
 	uint32_t idx;
-	uint32_t num_active_disps = 0;
-	struct cgs_display_info info = {0};
 	struct PP_Clocks min_clocks = {0};
 	uint32_t i;
 	struct pp_display_clock_request clock_req;
 
-	info.mode_info = NULL;
-
-	cgs_get_active_displays_info(hwmgr->device, &info);
-
-	num_active_disps = info.display_count;
-
-	if (num_active_disps > 1)
+	if (hwmgr->display_config->num_display > 1)
 		vega10_notify_smc_display_change(hwmgr, false);
 	else
 		vega10_notify_smc_display_change(hwmgr, true);
 
-	min_clocks.dcefClock = hwmgr->display_config.min_dcef_set_clk;
-	min_clocks.dcefClockInSR = hwmgr->display_config.min_dcef_deep_sleep_set_clk;
-	min_clocks.memoryClock = hwmgr->display_config.min_mem_set_clock;
+	min_clocks.dcefClock = hwmgr->display_config->min_dcef_set_clk;
+	min_clocks.dcefClockInSR = hwmgr->display_config->min_dcef_deep_sleep_set_clk;
+	min_clocks.memoryClock = hwmgr->display_config->min_mem_set_clock;
 
 	for (i = 0; i < dpm_table->count; i++) {
 		if (dpm_table->dpm_levels[i].value == min_clocks.dcefClock)
@@ -4501,10 +4485,8 @@ static int vega10_print_clock_levels(struct pp_hwmgr *hwmgr,
 static int vega10_display_configuration_changed_task(struct pp_hwmgr *hwmgr)
 {
 	struct vega10_hwmgr *data = hwmgr->backend;
-	int result = 0;
-	uint32_t num_turned_on_displays = 1;
 	Watermarks_t *wm_table = &(data->smc_state_table.water_marks_table);
-	struct cgs_display_info info = {0};
+	int result = 0;
 
 	if ((data->water_marks_bitmap & WaterMarksExist) &&
 			!(data->water_marks_bitmap & WaterMarksLoaded)) {
@@ -4514,10 +4496,8 @@ static int vega10_display_configuration_changed_task(struct pp_hwmgr *hwmgr)
 	}
 
 	if (data->water_marks_bitmap & WaterMarksLoaded) {
-		cgs_get_active_displays_info(hwmgr->device, &info);
-		num_turned_on_displays = info.display_count;
 		smum_send_msg_to_smc_with_parameter(hwmgr,
-			PPSMC_MSG_NumOfDisplays, num_turned_on_displays);
+			PPSMC_MSG_NumOfDisplays, hwmgr->display_config->num_display);
 	}
 
 	return result;
@@ -4603,15 +4583,12 @@ vega10_check_smc_update_required_for_display_configuration(struct pp_hwmgr *hwmg
 {
 	struct vega10_hwmgr *data = hwmgr->backend;
 	bool is_update_required = false;
-	struct cgs_display_info info = {0, 0, NULL};
-
-	cgs_get_active_displays_info(hwmgr->device, &info);
 
-	if (data->display_timing.num_existing_displays != info.display_count)
+	if (data->display_timing.num_existing_displays != hwmgr->display_config->num_display)
 		is_update_required = true;
 
 	if (PP_CAP(PHM_PlatformCaps_SclkDeepSleep)) {
-		if (data->display_timing.min_clock_in_sr != hwmgr->display_config.min_core_set_clock_in_sr)
+		if (data->display_timing.min_clock_in_sr != hwmgr->display_config->min_core_set_clock_in_sr)
 			is_update_required = true;
 	}
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
index 200de46..6a85238 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
@@ -1260,23 +1260,18 @@ static int vega12_notify_smc_display_config_after_ps_adjustment(
 {
 	struct vega12_hwmgr *data =
 			(struct vega12_hwmgr *)(hwmgr->backend);
-	uint32_t num_active_disps = 0;
-	struct cgs_display_info info = {0};
 	struct PP_Clocks min_clocks = {0};
 	struct pp_display_clock_request clock_req;
 	uint32_t clk_request;
 
-	info.mode_info = NULL;
-	cgs_get_active_displays_info(hwmgr->device, &info);
-	num_active_disps = info.display_count;
-	if (num_active_disps > 1)
+	if (hwmgr->display_config->num_display > 1)
 		vega12_notify_smc_display_change(hwmgr, false);
 	else
 		vega12_notify_smc_display_change(hwmgr, true);
 
-	min_clocks.dcefClock = hwmgr->display_config.min_dcef_set_clk;
-	min_clocks.dcefClockInSR = hwmgr->display_config.min_dcef_deep_sleep_set_clk;
-	min_clocks.memoryClock = hwmgr->display_config.min_mem_set_clock;
+	min_clocks.dcefClock = hwmgr->display_config->min_dcef_set_clk;
+	min_clocks.dcefClockInSR = hwmgr->display_config->min_dcef_deep_sleep_set_clk;
+	min_clocks.memoryClock = hwmgr->display_config->min_mem_set_clock;
 
 	if (data->smu_features[GNLD_DPM_DCEFCLK].supported) {
 		clock_req.clock_type = amd_pp_dcef_clock;
@@ -1832,9 +1827,7 @@ static int vega12_display_configuration_changed_task(struct pp_hwmgr *hwmgr)
 {
 	struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend);
 	int result = 0;
-	uint32_t num_turned_on_displays = 1;
 	Watermarks_t *wm_table = &(data->smc_state_table.water_marks_table);
-	struct cgs_display_info info = {0};
 
 	if ((data->water_marks_bitmap & WaterMarksExist) &&
 			!(data->water_marks_bitmap & WaterMarksLoaded)) {
@@ -1846,12 +1839,9 @@ static int vega12_display_configuration_changed_task(struct pp_hwmgr *hwmgr)
 
 	if ((data->water_marks_bitmap & WaterMarksExist) &&
 		data->smu_features[GNLD_DPM_DCEFCLK].supported &&
-		data->smu_features[GNLD_DPM_SOCCLK].supported) {
-		cgs_get_active_displays_info(hwmgr->device, &info);
-		num_turned_on_displays = info.display_count;
+		data->smu_features[GNLD_DPM_SOCCLK].supported)
 		smum_send_msg_to_smc_with_parameter(hwmgr,
-			PPSMC_MSG_NumOfDisplays, num_turned_on_displays);
-	}
+			PPSMC_MSG_NumOfDisplays, hwmgr->display_config->num_display);
 
 	return result;
 }
@@ -1894,15 +1884,12 @@ vega12_check_smc_update_required_for_display_configuration(struct pp_hwmgr *hwmg
 {
 	struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend);
 	bool is_update_required = false;
-	struct cgs_display_info info = {0, 0, NULL};
-
-	cgs_get_active_displays_info(hwmgr->device, &info);
 
-	if (data->display_timing.num_existing_displays != info.display_count)
+	if (data->display_timing.num_existing_displays != hwmgr->display_config->num_display)
 		is_update_required = true;
 
 	if (data->registry_data.gfx_clk_deep_sleep_support) {
-		if (data->display_timing.min_clock_in_sr != hwmgr->display_config.min_core_set_clock_in_sr)
+		if (data->display_timing.min_clock_in_sr != hwmgr->display_config->min_core_set_clock_in_sr)
 			is_update_required = true;
 	}
 
diff --git a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
index d5cadc6..e450ec7 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
@@ -765,7 +765,7 @@ struct pp_hwmgr {
 	struct pp_power_state    *request_ps;
 	struct pp_power_state    *boot_ps;
 	struct pp_power_state    *uvd_ps;
-	struct amd_pp_display_configuration display_config;
+	const struct amd_pp_display_configuration *display_config;
 	uint32_t feature_mask;
 	bool avfs_supported;
 	/* UMD Pstate */
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c
index e30a2ee..c28b95f 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c
@@ -1182,7 +1182,6 @@ static int ci_populate_single_memory_level(
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
 	int result = 0;
 	bool dll_state_on;
-	struct cgs_display_info info = {0};
 	uint32_t mclk_edc_wr_enable_threshold = 40000;
 	uint32_t mclk_edc_enable_threshold = 40000;
 	uint32_t mclk_strobe_mode_threshold = 40000;
@@ -1236,8 +1235,7 @@ static int ci_populate_single_memory_level(
 	/* default set to low watermark. Highest level will be set to high later.*/
 	memory_level->DisplayWatermark = PPSMC_DISPLAY_WATERMARK_LOW;
 
-	cgs_get_active_displays_info(hwmgr->device, &info);
-	data->display_timing.num_existing_displays = info.display_count;
+	data->display_timing.num_existing_displays = hwmgr->display_config->num_display;
 
 	/* stutter mode not support on ci */
 
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c
index 1eec527..d023494 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c
@@ -988,11 +988,11 @@ static int fiji_populate_single_graphic_level(struct pp_hwmgr *hwmgr,
 
 	threshold = clock * data->fast_watermark_threshold / 100;
 
-	data->display_timing.min_clock_in_sr = hwmgr->display_config.min_core_set_clock_in_sr;
+	data->display_timing.min_clock_in_sr = hwmgr->display_config->min_core_set_clock_in_sr;
 
 	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_SclkDeepSleep))
 		level->DeepSleepDivId = smu7_get_sleep_divider_id_from_clock(clock,
-								hwmgr->display_config.min_core_set_clock_in_sr);
+								hwmgr->display_config->min_core_set_clock_in_sr);
 
 
 	/* Default to slow, highest DPM level will be
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/iceland_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/iceland_smumgr.c
index d4bb934..bc05e35 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/iceland_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/iceland_smumgr.c
@@ -932,7 +932,7 @@ static int iceland_populate_single_graphic_level(struct pp_hwmgr *hwmgr,
 	graphic_level->PowerThrottle = 0;
 
 	data->display_timing.min_clock_in_sr =
-			hwmgr->display_config.min_core_set_clock_in_sr;
+			hwmgr->display_config->min_core_set_clock_in_sr;
 
 	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
 			PHM_PlatformCaps_SclkDeepSleep))
@@ -1236,7 +1236,6 @@ static int iceland_populate_single_memory_level(
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
 	int result = 0;
 	bool dll_state_on;
-	struct cgs_display_info info = {0};
 	uint32_t mclk_edc_wr_enable_threshold = 40000;
 	uint32_t mclk_edc_enable_threshold = 40000;
 	uint32_t mclk_strobe_mode_threshold = 40000;
@@ -1283,8 +1282,7 @@ static int iceland_populate_single_memory_level(
 	/* default set to low watermark. Highest level will be set to high later.*/
 	memory_level->DisplayWatermark = PPSMC_DISPLAY_WATERMARK_LOW;
 
-	cgs_get_active_displays_info(hwmgr->device, &info);
-	data->display_timing.num_existing_displays = info.display_count;
+	data->display_timing.num_existing_displays = hwmgr->display_config->num_display;
 
 	/* stutter mode not support on iceland */
 
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c
index 05e60e8..d919228 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c
@@ -942,11 +942,11 @@ static int polaris10_populate_single_graphic_level(struct pp_hwmgr *hwmgr,
 	level->DownHyst = data->current_profile_setting.sclk_down_hyst;
 	level->VoltageDownHyst = 0;
 	level->PowerThrottle = 0;
-	data->display_timing.min_clock_in_sr = hwmgr->display_config.min_core_set_clock_in_sr;
+	data->display_timing.min_clock_in_sr = hwmgr->display_config->min_core_set_clock_in_sr;
 
 	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_SclkDeepSleep))
 		level->DeepSleepDivId = smu7_get_sleep_divider_id_from_clock(clock,
-								hwmgr->display_config.min_core_set_clock_in_sr);
+								hwmgr->display_config->min_core_set_clock_in_sr);
 
 	/* Default to slow, highest DPM level will be
 	 * set to PPSMC_DISPLAY_WATERMARK_LOW later.
@@ -1076,11 +1076,9 @@ static int polaris10_populate_single_memory_level(struct pp_hwmgr *hwmgr,
 	struct phm_ppt_v1_information *table_info =
 			(struct phm_ppt_v1_information *)(hwmgr->pptable);
 	int result = 0;
-	struct cgs_display_info info = {0, 0, NULL};
 	uint32_t mclk_stutter_mode_threshold = 40000;
 	phm_ppt_v1_clock_voltage_dependency_table *vdd_dep_table = NULL;
 
-	cgs_get_active_displays_info(hwmgr->device, &info);
 
 	if (hwmgr->od_enabled)
 		vdd_dep_table = (phm_ppt_v1_clock_voltage_dependency_table *)&data->odn_dpm_table.vdd_dependency_on_mclk;
@@ -1106,7 +1104,7 @@ static int polaris10_populate_single_memory_level(struct pp_hwmgr *hwmgr,
 	mem_level->StutterEnable = false;
 	mem_level->DisplayWatermark = PPSMC_DISPLAY_WATERMARK_LOW;
 
-	data->display_timing.num_existing_displays = info.display_count;
+	data->display_timing.num_existing_displays = hwmgr->display_config->num_display;
 
 	if (mclk_stutter_mode_threshold &&
 		(clock <= mclk_stutter_mode_threshold) &&
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c
index 2ba05d2..94ba304 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c
@@ -650,7 +650,7 @@ static int tonga_populate_single_graphic_level(struct pp_hwmgr *hwmgr,
 	graphic_level->PowerThrottle = 0;
 
 	data->display_timing.min_clock_in_sr =
-			hwmgr->display_config.min_core_set_clock_in_sr;
+			hwmgr->display_config->min_core_set_clock_in_sr;
 
 	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
 			PHM_PlatformCaps_SclkDeepSleep))
@@ -956,18 +956,17 @@ static int tonga_populate_single_memory_level(
 		SMU72_Discrete_MemoryLevel *memory_level
 		)
 {
-	uint32_t mvdd = 0;
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
 	struct phm_ppt_v1_information *pptable_info =
 			  (struct phm_ppt_v1_information *)(hwmgr->pptable);
-	int result = 0;
-	bool dll_state_on;
-	struct cgs_display_info info = {0};
 	uint32_t mclk_edc_wr_enable_threshold = 40000;
 	uint32_t mclk_stutter_mode_threshold = 30000;
 	uint32_t mclk_edc_enable_threshold = 40000;
 	uint32_t mclk_strobe_mode_threshold = 40000;
 	phm_ppt_v1_clock_voltage_dependency_table *vdd_dep_table = NULL;
+	int result = 0;
+	bool dll_state_on;
+	uint32_t mvdd = 0;
 
 	if (hwmgr->od_enabled)
 		vdd_dep_table = (phm_ppt_v1_clock_voltage_dependency_table *)&data->odn_dpm_table.vdd_dependency_on_mclk;
@@ -1008,8 +1007,7 @@ static int tonga_populate_single_memory_level(
 	/* default set to low watermark. Highest level will be set to high later.*/
 	memory_level->DisplayWatermark = PPSMC_DISPLAY_WATERMARK_LOW;
 
-	cgs_get_active_displays_info(hwmgr->device, &info);
-	data->display_timing.num_existing_displays = info.display_count;
+	data->display_timing.num_existing_displays = hwmgr->display_config->num_display;
 
 	if ((mclk_stutter_mode_threshold != 0) &&
 	    (memory_clock <= mclk_stutter_mode_threshold) &&
-- 
cgit v1.1


From 10b3f45c4a189ffa0e0b9566fce7a0b65b289322 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Thu, 22 Mar 2018 19:32:45 +0800
Subject: drm/amdgpu: Delete some cgs functions

Drop cgs wrappers that are no longer used.
1. cgs_rel_firmwar
2. cgs_is_virtualization_enabled
3. cgs_notify_dpm_enabled
4. cgs_atom_get_data_table
5. cgs_atom_get_cmd_table_revs
6. cgs_atom_exec_cmd_table
7. cgs_get_active_displays_info

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c  | 197 -------------------------------
 drivers/gpu/drm/amd/include/cgs_common.h | 139 ----------------------
 2 files changed, 336 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
index 71a57b2..dc28fa6 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
@@ -23,7 +23,6 @@
  */
 #include <linux/list.h>
 #include <linux/slab.h>
-#include <linux/pci.h>
 #include <drm/drmP.h>
 #include <linux/firmware.h>
 #include <drm/amdgpu_drm.h>
@@ -109,78 +108,6 @@ static void amdgpu_cgs_write_ind_register(struct cgs_device *cgs_device,
 	WARN(1, "Invalid indirect register space");
 }
 
-static int amdgpu_cgs_get_pci_resource(struct cgs_device *cgs_device,
-				       enum cgs_resource_type resource_type,
-				       uint64_t size,
-				       uint64_t offset,
-				       uint64_t *resource_base)
-{
-	CGS_FUNC_ADEV;
-
-	if (resource_base == NULL)
-		return -EINVAL;
-
-	switch (resource_type) {
-	case CGS_RESOURCE_TYPE_MMIO:
-		if (adev->rmmio_size == 0)
-			return -ENOENT;
-		if ((offset + size) > adev->rmmio_size)
-			return -EINVAL;
-		*resource_base = adev->rmmio_base;
-		return 0;
-	case CGS_RESOURCE_TYPE_DOORBELL:
-		if (adev->doorbell.size == 0)
-			return -ENOENT;
-		if ((offset + size) > adev->doorbell.size)
-			return -EINVAL;
-		*resource_base = adev->doorbell.base;
-		return 0;
-	case CGS_RESOURCE_TYPE_FB:
-	case CGS_RESOURCE_TYPE_IO:
-	case CGS_RESOURCE_TYPE_ROM:
-	default:
-		return -EINVAL;
-	}
-}
-
-static const void *amdgpu_cgs_atom_get_data_table(struct cgs_device *cgs_device,
-						  unsigned table, uint16_t *size,
-						  uint8_t *frev, uint8_t *crev)
-{
-	CGS_FUNC_ADEV;
-	uint16_t data_start;
-
-	if (amdgpu_atom_parse_data_header(
-		    adev->mode_info.atom_context, table, size,
-		    frev, crev, &data_start))
-		return (uint8_t*)adev->mode_info.atom_context->bios +
-			data_start;
-
-	return NULL;
-}
-
-static int amdgpu_cgs_atom_get_cmd_table_revs(struct cgs_device *cgs_device, unsigned table,
-					      uint8_t *frev, uint8_t *crev)
-{
-	CGS_FUNC_ADEV;
-
-	if (amdgpu_atom_parse_cmd_header(
-		    adev->mode_info.atom_context, table,
-		    frev, crev))
-		return 0;
-
-	return -EINVAL;
-}
-
-static int amdgpu_cgs_atom_exec_cmd_table(struct cgs_device *cgs_device, unsigned table,
-					  void *args)
-{
-	CGS_FUNC_ADEV;
-
-	return amdgpu_atom_execute_table(
-		adev->mode_info.atom_context, table, args);
-}
-
 static int amdgpu_cgs_set_clockgating_state(struct cgs_device *cgs_device,
 				  enum amd_ip_block_type block_type,
 				  enum amd_clockgating_state state)
@@ -223,7 +150,6 @@ static int amdgpu_cgs_set_powergating_state(struct cgs_device *cgs_device,
 	return r;
 }
 
-
 static uint32_t fw_type_convert(struct cgs_device *cgs_device, uint32_t fw_type)
 {
 	CGS_FUNC_ADEV;
@@ -271,18 +197,6 @@ static uint32_t fw_type_convert(struct cgs_device *cgs_device, uint32_t fw_type)
 	return result;
 }
 
-static int amdgpu_cgs_rel_firmware(struct cgs_device *cgs_device, enum cgs_ucode_id type)
-{
-	CGS_FUNC_ADEV;
-	if ((CGS_UCODE_ID_SMU == type) || (CGS_UCODE_ID_SMU_SK == type)) {
-		release_firmware(adev->pm.fw);
-		adev->pm.fw = NULL;
-		return 0;
-	}
-	/* cannot release other firmware because they are not created by cgs */
-	return -EINVAL;
-}
-
 static uint16_t amdgpu_get_firmware_version(struct cgs_device *cgs_device,
 					enum cgs_ucode_id type)
 {
@@ -326,34 +240,6 @@ static uint16_t amdgpu_get_firmware_version(struct cgs_device *cgs_device,
 	return fw_version;
 }
 
-static int amdgpu_cgs_enter_safe_mode(struct cgs_device *cgs_device,
-					bool en)
-{
-	CGS_FUNC_ADEV;
-
-	if (adev->gfx.rlc.funcs->enter_safe_mode == NULL ||
-		adev->gfx.rlc.funcs->exit_safe_mode == NULL)
-		return 0;
-
-	if (en)
-		adev->gfx.rlc.funcs->enter_safe_mode(adev);
-	else
-		adev->gfx.rlc.funcs->exit_safe_mode(adev);
-
-	return 0;
-}
-
-static void amdgpu_cgs_lock_grbm_idx(struct cgs_device *cgs_device,
-					bool lock)
-{
-	CGS_FUNC_ADEV;
-
-	if (lock)
-		mutex_lock(&adev->grbm_idx_mutex);
-	else
-		mutex_unlock(&adev->grbm_idx_mutex);
-}
-
 static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device,
 					enum cgs_ucode_id type,
 					struct cgs_firmware_info *info)
@@ -598,97 +484,14 @@ static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device,
 	return 0;
 }
 
-static int amdgpu_cgs_is_virtualization_enabled(void *cgs_device)
-{
-	CGS_FUNC_ADEV;
-	return amdgpu_sriov_vf(adev);
-}
-
-static int amdgpu_cgs_get_active_displays_info(struct cgs_device *cgs_device,
-					  struct cgs_display_info *info)
-{
-	CGS_FUNC_ADEV;
-	struct cgs_mode_info *mode_info;
-
-	if (info == NULL)
-		return -EINVAL;
-
-	mode_info = info->mode_info;
-	if (mode_info)
-		/* if the displays are off, vblank time is max */
-		mode_info->vblank_time_us = 0xffffffff;
-
-	if (!amdgpu_device_has_dc_support(adev)) {
-		struct amdgpu_crtc *amdgpu_crtc;
-		struct drm_device *ddev = adev->ddev;
-		struct drm_crtc *crtc;
-		uint32_t line_time_us, vblank_lines;
-
-		if (adev->mode_info.num_crtc && adev->mode_info.mode_config_initialized) {
-			list_for_each_entry(crtc,
-					&ddev->mode_config.crtc_list, head) {
-				amdgpu_crtc = to_amdgpu_crtc(crtc);
-				if (crtc->enabled) {
-					info->active_display_mask |= (1 << amdgpu_crtc->crtc_id);
-					info->display_count++;
-				}
-				if (mode_info != NULL &&
-					crtc->enabled && amdgpu_crtc->enabled &&
-					amdgpu_crtc->hw_mode.clock) {
-					line_time_us = (amdgpu_crtc->hw_mode.crtc_htotal * 1000) /
-								amdgpu_crtc->hw_mode.clock;
-					vblank_lines = amdgpu_crtc->hw_mode.crtc_vblank_end -
-								amdgpu_crtc->hw_mode.crtc_vdisplay +
-								(amdgpu_crtc->v_border * 2);
-					mode_info->vblank_time_us = vblank_lines * line_time_us;
-					mode_info->refresh_rate = drm_mode_vrefresh(&amdgpu_crtc->hw_mode);
-					/* we have issues with mclk switching with refresh rates
-					 * over 120 hz on the non-DC code.
-					 */
-					if (mode_info->refresh_rate > 120)
-						mode_info->vblank_time_us = 0;
-					mode_info = NULL;
-				}
-			}
-		}
-	} else {
-		info->display_count = adev->pm.pm_display_cfg.num_display;
-		if (mode_info != NULL) {
-			mode_info->vblank_time_us = adev->pm.pm_display_cfg.min_vblank_time;
-			mode_info->refresh_rate = adev->pm.pm_display_cfg.vrefresh;
-		}
-	}
-	return 0;
-}
-
-
-static int amdgpu_cgs_notify_dpm_enabled(struct cgs_device *cgs_device, bool enabled)
-{
-	CGS_FUNC_ADEV;
-
-	adev->pm.dpm_enabled = enabled;
-
-	return 0;
-}
-
 static const struct cgs_ops amdgpu_cgs_ops = {
 	.read_register = amdgpu_cgs_read_register,
 	.write_register = amdgpu_cgs_write_register,
 	.read_ind_register = amdgpu_cgs_read_ind_register,
 	.write_ind_register = amdgpu_cgs_write_ind_register,
-	.get_pci_resource = amdgpu_cgs_get_pci_resource,
-	.atom_get_data_table = amdgpu_cgs_atom_get_data_table,
-	.atom_get_cmd_table_revs = amdgpu_cgs_atom_get_cmd_table_revs,
-	.atom_exec_cmd_table = amdgpu_cgs_atom_exec_cmd_table,
 	.get_firmware_info = amdgpu_cgs_get_firmware_info,
-	.rel_firmware = amdgpu_cgs_rel_firmware,
 	.set_powergating_state = amdgpu_cgs_set_powergating_state,
 	.set_clockgating_state = amdgpu_cgs_set_clockgating_state,
-	.get_active_displays_info = amdgpu_cgs_get_active_displays_info,
-	.notify_dpm_enabled = amdgpu_cgs_notify_dpm_enabled,
-	.is_virtualization_enabled = amdgpu_cgs_is_virtualization_enabled,
-	.enter_safe_mode = amdgpu_cgs_enter_safe_mode,
-	.lock_grbm_idx = amdgpu_cgs_lock_grbm_idx,
 };
 
 struct cgs_device *amdgpu_cgs_create_device(struct amdgpu_device *adev)
diff --git a/drivers/gpu/drm/amd/include/cgs_common.h b/drivers/gpu/drm/amd/include/cgs_common.h
index f2814ae..cab34a4 100644
--- a/drivers/gpu/drm/amd/include/cgs_common.h
+++ b/drivers/gpu/drm/amd/include/cgs_common.h
@@ -76,17 +76,6 @@ enum cgs_ucode_id {
 	CGS_UCODE_ID_MAXIMUM,
 };
 
-/*
- * enum cgs_resource_type - GPU resource type
- */
-enum cgs_resource_type {
-	CGS_RESOURCE_TYPE_MMIO = 0,
-	CGS_RESOURCE_TYPE_FB,
-	CGS_RESOURCE_TYPE_IO,
-	CGS_RESOURCE_TYPE_DOORBELL,
-	CGS_RESOURCE_TYPE_ROM,
-};
-
 /**
  * struct cgs_firmware_info - Firmware information
  */
@@ -104,17 +93,6 @@ struct cgs_firmware_info {
 	bool			is_kicker;
 };
 
-struct cgs_mode_info {
-	uint32_t		refresh_rate;
-	uint32_t		vblank_time_us;
-};
-
-struct cgs_display_info {
-	uint32_t		display_count;
-	uint32_t		active_display_mask;
-	struct cgs_mode_info *mode_info;
-};
-
 typedef unsigned long cgs_handle_t;
 
 /**
@@ -170,73 +148,10 @@ typedef void (*cgs_write_ind_register_t)(struct cgs_device *cgs_device, enum cgs
 #define CGS_WREG32_FIELD_IND(device, space, reg, field, val)	\
 	cgs_write_ind_register(device, space, ix##reg, (cgs_read_ind_register(device, space, ix##reg) & ~CGS_REG_FIELD_MASK(reg, field)) | (val) << CGS_REG_FIELD_SHIFT(reg, field))
 
-/**
- * cgs_get_pci_resource() - provide access to a device resource (PCI BAR)
- * @cgs_device:	opaque device handle
- * @resource_type:	Type of Resource (MMIO, IO, ROM, FB, DOORBELL)
- * @size:	size of the region
- * @offset:	offset from the start of the region
- * @resource_base:	base address (not including offset) returned
- *
- * Return: 0 on success, -errno otherwise
- */
-typedef int (*cgs_get_pci_resource_t)(struct cgs_device *cgs_device,
-				      enum cgs_resource_type resource_type,
-				      uint64_t size,
-				      uint64_t offset,
-				      uint64_t *resource_base);
-
-/**
- * cgs_atom_get_data_table() - Get a pointer to an ATOM BIOS data table
- * @cgs_device:	opaque device handle
- * @table:	data table index
- * @size:	size of the table (output, may be NULL)
- * @frev:	table format revision (output, may be NULL)
- * @crev:	table content revision (output, may be NULL)
- *
- * Return: Pointer to start of the table, or NULL on failure
- */
-typedef const void *(*cgs_atom_get_data_table_t)(
-	struct cgs_device *cgs_device, unsigned table,
-	uint16_t *size, uint8_t *frev, uint8_t *crev);
-
-/**
- * cgs_atom_get_cmd_table_revs() - Get ATOM BIOS command table revisions
- * @cgs_device:	opaque device handle
- * @table:	data table index
- * @frev:	table format revision (output, may be NULL)
- * @crev:	table content revision (output, may be NULL)
- *
- * Return: 0 on success, -errno otherwise
- */
-typedef int (*cgs_atom_get_cmd_table_revs_t)(struct cgs_device *cgs_device, unsigned table,
-					     uint8_t *frev, uint8_t *crev);
-
-/**
- * cgs_atom_exec_cmd_table() - Execute an ATOM BIOS command table
- * @cgs_device: opaque device handle
- * @table:	command table index
- * @args:	arguments
- *
- * Return: 0 on success, -errno otherwise
- */
-typedef int (*cgs_atom_exec_cmd_table_t)(struct cgs_device *cgs_device,
-					 unsigned table, void *args);
-
-/**
- * cgs_get_firmware_info - Get the firmware information from core driver
- * @cgs_device: opaque device handle
- * @type: the firmware type
- * @info: returend firmware information
- *
- * Return: 0 on success, -errno otherwise
- */
 typedef int (*cgs_get_firmware_info)(struct cgs_device *cgs_device,
 				     enum cgs_ucode_id type,
 				     struct cgs_firmware_info *info);
 
-typedef int (*cgs_rel_firmware)(struct cgs_device *cgs_device,
-					 enum cgs_ucode_id type);
 
 typedef int(*cgs_set_powergating_state)(struct cgs_device *cgs_device,
 				  enum amd_ip_block_type block_type,
@@ -246,43 +161,17 @@ typedef int(*cgs_set_clockgating_state)(struct cgs_device *cgs_device,
 				  enum amd_ip_block_type block_type,
 				  enum amd_clockgating_state state);
 
-typedef int(*cgs_get_active_displays_info)(
-					struct cgs_device *cgs_device,
-					struct cgs_display_info *info);
-
-typedef int (*cgs_notify_dpm_enabled)(struct cgs_device *cgs_device, bool enabled);
-
-typedef int (*cgs_is_virtualization_enabled_t)(void *cgs_device);
-
-typedef int (*cgs_enter_safe_mode)(struct cgs_device *cgs_device, bool en);
-
-typedef void (*cgs_lock_grbm_idx)(struct cgs_device *cgs_device, bool lock);
-
 struct cgs_ops {
 	/* MMIO access */
 	cgs_read_register_t read_register;
 	cgs_write_register_t write_register;
 	cgs_read_ind_register_t read_ind_register;
 	cgs_write_ind_register_t write_ind_register;
-	/* PCI resources */
-	cgs_get_pci_resource_t get_pci_resource;
-	/* ATOM BIOS */
-	cgs_atom_get_data_table_t atom_get_data_table;
-	cgs_atom_get_cmd_table_revs_t atom_get_cmd_table_revs;
-	cgs_atom_exec_cmd_table_t atom_exec_cmd_table;
 	/* Firmware Info */
 	cgs_get_firmware_info get_firmware_info;
-	cgs_rel_firmware rel_firmware;
 	/* cg pg interface*/
 	cgs_set_powergating_state set_powergating_state;
 	cgs_set_clockgating_state set_clockgating_state;
-	/* display manager */
-	cgs_get_active_displays_info get_active_displays_info;
-	/* notify dpm enabled */
-	cgs_notify_dpm_enabled notify_dpm_enabled;
-	cgs_is_virtualization_enabled_t is_virtualization_enabled;
-	cgs_enter_safe_mode enter_safe_mode;
-	cgs_lock_grbm_idx lock_grbm_idx;
 };
 
 struct cgs_os_ops; /* To be define in OS-specific CGS header */
@@ -309,40 +198,12 @@ struct cgs_device
 #define cgs_write_ind_register(dev,space,index,value)		\
 	CGS_CALL(write_ind_register,dev,space,index,value)
 
-#define cgs_atom_get_data_table(dev,table,size,frev,crev)	\
-	CGS_CALL(atom_get_data_table,dev,table,size,frev,crev)
-#define cgs_atom_get_cmd_table_revs(dev,table,frev,crev)	\
-	CGS_CALL(atom_get_cmd_table_revs,dev,table,frev,crev)
-#define cgs_atom_exec_cmd_table(dev,table,args)		\
-	CGS_CALL(atom_exec_cmd_table,dev,table,args)
-
 #define cgs_get_firmware_info(dev, type, info)	\
 	CGS_CALL(get_firmware_info, dev, type, info)
-#define cgs_rel_firmware(dev, type)	\
-	CGS_CALL(rel_firmware, dev, type)
 #define cgs_set_powergating_state(dev, block_type, state)	\
 	CGS_CALL(set_powergating_state, dev, block_type, state)
 #define cgs_set_clockgating_state(dev, block_type, state)	\
 	CGS_CALL(set_clockgating_state, dev, block_type, state)
-#define cgs_notify_dpm_enabled(dev, enabled)	\
-	CGS_CALL(notify_dpm_enabled, dev, enabled)
-
-#define cgs_get_active_displays_info(dev, info)	\
-	CGS_CALL(get_active_displays_info, dev, info)
-
-#define cgs_get_pci_resource(cgs_device, resource_type, size, offset, \
-	resource_base) \
-	CGS_CALL(get_pci_resource, cgs_device, resource_type, size, offset, \
-	resource_base)
-
-#define cgs_is_virtualization_enabled(cgs_device) \
-		CGS_CALL(is_virtualization_enabled, cgs_device)
-
-#define cgs_enter_safe_mode(cgs_device, en) \
-		CGS_CALL(enter_safe_mode, cgs_device, en)
-
-#define cgs_lock_grbm_idx(cgs_device, lock) \
-		CGS_CALL(lock_grbm_idx, cgs_device, lock)
 
 
 #endif /* _CGS_COMMON_H */
-- 
cgit v1.1


From 43fa561fd07fe707815d1b72472f6f5829223a52 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Wed, 28 Mar 2018 13:42:45 -0500
Subject: drm/amdgpu: remove duplicate cg/pg wrapper functions
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König<christian.koenig@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu.h                |  4 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c            | 44 ----------------------
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c         |  6 ++-
 drivers/gpu/drm/amd/include/cgs_common.h           | 31 ---------------
 drivers/gpu/drm/amd/powerplay/amd_powerplay.c      |  8 ++--
 .../amd/powerplay/hwmgr/smu7_clockpowergating.c    | 16 ++++----
 drivers/gpu/drm/amd/powerplay/hwmgr/smu8_hwmgr.c   | 20 ++++------
 drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c |  8 ++--
 8 files changed, 30 insertions(+), 107 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index 0193f6c..3000c4a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -222,10 +222,10 @@ enum amdgpu_kiq_irq {
 	AMDGPU_CP_KIQ_IRQ_LAST
 };
 
-int amdgpu_device_ip_set_clockgating_state(struct amdgpu_device *adev,
+int amdgpu_device_ip_set_clockgating_state(void *dev,
 					   enum amd_ip_block_type block_type,
 					   enum amd_clockgating_state state);
-int amdgpu_device_ip_set_powergating_state(struct amdgpu_device *adev,
+int amdgpu_device_ip_set_powergating_state(void *dev,
 					   enum amd_ip_block_type block_type,
 					   enum amd_powergating_state state);
 void amdgpu_device_ip_get_clockgating_state(struct amdgpu_device *adev,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
index dc28fa6..a8a942c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
@@ -108,48 +108,6 @@ static void amdgpu_cgs_write_ind_register(struct cgs_device *cgs_device,
 	WARN(1, "Invalid indirect register space");
 }
 
-static int amdgpu_cgs_set_clockgating_state(struct cgs_device *cgs_device,
-				  enum amd_ip_block_type block_type,
-				  enum amd_clockgating_state state)
-{
-	CGS_FUNC_ADEV;
-	int i, r = -1;
-
-	for (i = 0; i < adev->num_ip_blocks; i++) {
-		if (!adev->ip_blocks[i].status.valid)
-			continue;
-
-		if (adev->ip_blocks[i].version->type == block_type) {
-			r = adev->ip_blocks[i].version->funcs->set_clockgating_state(
-								(void *)adev,
-									state);
-			break;
-		}
-	}
-	return r;
-}
-
-static int amdgpu_cgs_set_powergating_state(struct cgs_device *cgs_device,
-				  enum amd_ip_block_type block_type,
-				  enum amd_powergating_state state)
-{
-	CGS_FUNC_ADEV;
-	int i, r = -1;
-
-	for (i = 0; i < adev->num_ip_blocks; i++) {
-		if (!adev->ip_blocks[i].status.valid)
-			continue;
-
-		if (adev->ip_blocks[i].version->type == block_type) {
-			r = adev->ip_blocks[i].version->funcs->set_powergating_state(
-								(void *)adev,
-									state);
-			break;
-		}
-	}
-	return r;
-}
-
 static uint32_t fw_type_convert(struct cgs_device *cgs_device, uint32_t fw_type)
 {
 	CGS_FUNC_ADEV;
@@ -490,8 +448,6 @@ static const struct cgs_ops amdgpu_cgs_ops = {
 	.read_ind_register = amdgpu_cgs_read_ind_register,
 	.write_ind_register = amdgpu_cgs_write_ind_register,
 	.get_firmware_info = amdgpu_cgs_get_firmware_info,
-	.set_powergating_state = amdgpu_cgs_set_powergating_state,
-	.set_clockgating_state = amdgpu_cgs_set_clockgating_state,
 };
 
 struct cgs_device *amdgpu_cgs_create_device(struct amdgpu_device *adev)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 34af664..a539265 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -1039,10 +1039,11 @@ static const struct vga_switcheroo_client_ops amdgpu_switcheroo_ops = {
  * the hardware IP specified.
  * Returns the error code from the last instance.
  */
-int amdgpu_device_ip_set_clockgating_state(struct amdgpu_device *adev,
+int amdgpu_device_ip_set_clockgating_state(void *dev,
 					   enum amd_ip_block_type block_type,
 					   enum amd_clockgating_state state)
 {
+	struct amdgpu_device *adev = dev;
 	int i, r = 0;
 
 	for (i = 0; i < adev->num_ip_blocks; i++) {
@@ -1072,10 +1073,11 @@ int amdgpu_device_ip_set_clockgating_state(struct amdgpu_device *adev,
  * the hardware IP specified.
  * Returns the error code from the last instance.
  */
-int amdgpu_device_ip_set_powergating_state(struct amdgpu_device *adev,
+int amdgpu_device_ip_set_powergating_state(void *dev,
 					   enum amd_ip_block_type block_type,
 					   enum amd_powergating_state state)
 {
+	struct amdgpu_device *adev = dev;
 	int i, r = 0;
 
 	for (i = 0; i < adev->num_ip_blocks; i++) {
diff --git a/drivers/gpu/drm/amd/include/cgs_common.h b/drivers/gpu/drm/amd/include/cgs_common.h
index cab34a4..a69deb3 100644
--- a/drivers/gpu/drm/amd/include/cgs_common.h
+++ b/drivers/gpu/drm/amd/include/cgs_common.h
@@ -42,20 +42,6 @@ enum cgs_ind_reg {
 	CGS_IND_REG__AUDIO_ENDPT
 };
 
-/**
- * enum cgs_engine - Engines that can be statically power-gated
- */
-enum cgs_engine {
-	CGS_ENGINE__UVD,
-	CGS_ENGINE__VCE,
-	CGS_ENGINE__VP8,
-	CGS_ENGINE__ACP_DMA,
-	CGS_ENGINE__ACP_DSP0,
-	CGS_ENGINE__ACP_DSP1,
-	CGS_ENGINE__ISP,
-	/* ... */
-};
-
 /*
  * enum cgs_ucode_id - Firmware types for different IPs
  */
@@ -152,15 +138,6 @@ typedef int (*cgs_get_firmware_info)(struct cgs_device *cgs_device,
 				     enum cgs_ucode_id type,
 				     struct cgs_firmware_info *info);
 
-
-typedef int(*cgs_set_powergating_state)(struct cgs_device *cgs_device,
-				  enum amd_ip_block_type block_type,
-				  enum amd_powergating_state state);
-
-typedef int(*cgs_set_clockgating_state)(struct cgs_device *cgs_device,
-				  enum amd_ip_block_type block_type,
-				  enum amd_clockgating_state state);
-
 struct cgs_ops {
 	/* MMIO access */
 	cgs_read_register_t read_register;
@@ -169,9 +146,6 @@ struct cgs_ops {
 	cgs_write_ind_register_t write_ind_register;
 	/* Firmware Info */
 	cgs_get_firmware_info get_firmware_info;
-	/* cg pg interface*/
-	cgs_set_powergating_state set_powergating_state;
-	cgs_set_clockgating_state set_clockgating_state;
 };
 
 struct cgs_os_ops; /* To be define in OS-specific CGS header */
@@ -200,10 +174,5 @@ struct cgs_device
 
 #define cgs_get_firmware_info(dev, type, info)	\
 	CGS_CALL(get_firmware_info, dev, type, info)
-#define cgs_set_powergating_state(dev, block_type, state)	\
-	CGS_CALL(set_powergating_state, dev, block_type, state)
-#define cgs_set_clockgating_state(dev, block_type, state)	\
-	CGS_CALL(set_clockgating_state, dev, block_type, state)
-
 
 #endif /* _CGS_COMMON_H */
diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
index 1ca6a13..66c49b8 100644
--- a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
+++ b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
@@ -288,10 +288,10 @@ static void pp_dpm_en_umd_pstate(struct pp_hwmgr  *hwmgr,
 		if (*level & profile_mode_mask) {
 			hwmgr->saved_dpm_level = hwmgr->dpm_level;
 			hwmgr->en_umd_pstate = true;
-			cgs_set_clockgating_state(hwmgr->device,
+			amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 						AMD_IP_BLOCK_TYPE_GFX,
 						AMD_CG_STATE_UNGATE);
-			cgs_set_powergating_state(hwmgr->device,
+			amdgpu_device_ip_set_powergating_state(hwmgr->adev,
 					AMD_IP_BLOCK_TYPE_GFX,
 					AMD_PG_STATE_UNGATE);
 		}
@@ -301,10 +301,10 @@ static void pp_dpm_en_umd_pstate(struct pp_hwmgr  *hwmgr,
 			if (*level == AMD_DPM_FORCED_LEVEL_PROFILE_EXIT)
 				*level = hwmgr->saved_dpm_level;
 			hwmgr->en_umd_pstate = false;
-			cgs_set_clockgating_state(hwmgr->device,
+			amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 					AMD_IP_BLOCK_TYPE_GFX,
 					AMD_CG_STATE_GATE);
-			cgs_set_powergating_state(hwmgr->device,
+			amdgpu_device_ip_set_powergating_state(hwmgr->adev,
 					AMD_IP_BLOCK_TYPE_GFX,
 					AMD_PG_STATE_GATE);
 		}
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_clockpowergating.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_clockpowergating.c
index f4cbaee..6d72a56 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_clockpowergating.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_clockpowergating.c
@@ -147,20 +147,20 @@ void smu7_powergate_uvd(struct pp_hwmgr *hwmgr, bool bgate)
 	data->uvd_power_gated = bgate;
 
 	if (bgate) {
-		cgs_set_powergating_state(hwmgr->device,
+		amdgpu_device_ip_set_powergating_state(hwmgr->adev,
 						AMD_IP_BLOCK_TYPE_UVD,
 						AMD_PG_STATE_GATE);
-		cgs_set_clockgating_state(hwmgr->device,
+		amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 				AMD_IP_BLOCK_TYPE_UVD,
 				AMD_CG_STATE_GATE);
 		smu7_update_uvd_dpm(hwmgr, true);
 		smu7_powerdown_uvd(hwmgr);
 	} else {
 		smu7_powerup_uvd(hwmgr);
-		cgs_set_clockgating_state(hwmgr->device,
+		amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 				AMD_IP_BLOCK_TYPE_UVD,
 				AMD_CG_STATE_UNGATE);
-		cgs_set_powergating_state(hwmgr->device,
+		amdgpu_device_ip_set_powergating_state(hwmgr->adev,
 						AMD_IP_BLOCK_TYPE_UVD,
 						AMD_PG_STATE_UNGATE);
 		smu7_update_uvd_dpm(hwmgr, false);
@@ -175,20 +175,20 @@ void smu7_powergate_vce(struct pp_hwmgr *hwmgr, bool bgate)
 	data->vce_power_gated = bgate;
 
 	if (bgate) {
-		cgs_set_powergating_state(hwmgr->device,
+		amdgpu_device_ip_set_powergating_state(hwmgr->adev,
 						AMD_IP_BLOCK_TYPE_VCE,
 						AMD_PG_STATE_GATE);
-		cgs_set_clockgating_state(hwmgr->device,
+		amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 				AMD_IP_BLOCK_TYPE_VCE,
 				AMD_CG_STATE_GATE);
 		smu7_update_vce_dpm(hwmgr, true);
 		smu7_powerdown_vce(hwmgr);
 	} else {
 		smu7_powerup_vce(hwmgr);
-		cgs_set_clockgating_state(hwmgr->device,
+		amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 				AMD_IP_BLOCK_TYPE_VCE,
 				AMD_CG_STATE_UNGATE);
-		cgs_set_powergating_state(hwmgr->device,
+		amdgpu_device_ip_set_powergating_state(hwmgr->adev,
 						AMD_IP_BLOCK_TYPE_VCE,
 						AMD_PG_STATE_UNGATE);
 		smu7_update_vce_dpm(hwmgr, false);
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu8_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu8_hwmgr.c
index c2f93aa..50690c7 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu8_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu8_hwmgr.c
@@ -1892,20 +1892,20 @@ static void smu8_dpm_powergate_uvd(struct pp_hwmgr *hwmgr, bool bgate)
 	data->uvd_power_gated = bgate;
 
 	if (bgate) {
-		cgs_set_powergating_state(hwmgr->device,
+		amdgpu_device_ip_set_powergating_state(hwmgr->adev,
 						AMD_IP_BLOCK_TYPE_UVD,
 						AMD_PG_STATE_GATE);
-		cgs_set_clockgating_state(hwmgr->device,
+		amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 						AMD_IP_BLOCK_TYPE_UVD,
 						AMD_CG_STATE_GATE);
 		smu8_dpm_update_uvd_dpm(hwmgr, true);
 		smu8_dpm_powerdown_uvd(hwmgr);
 	} else {
 		smu8_dpm_powerup_uvd(hwmgr);
-		cgs_set_clockgating_state(hwmgr->device,
+		amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 						AMD_IP_BLOCK_TYPE_UVD,
 						AMD_CG_STATE_UNGATE);
-		cgs_set_powergating_state(hwmgr->device,
+		amdgpu_device_ip_set_powergating_state(hwmgr->adev,
 						AMD_IP_BLOCK_TYPE_UVD,
 						AMD_PG_STATE_UNGATE);
 		smu8_dpm_update_uvd_dpm(hwmgr, false);
@@ -1918,12 +1918,10 @@ static void smu8_dpm_powergate_vce(struct pp_hwmgr *hwmgr, bool bgate)
 	struct smu8_hwmgr *data = hwmgr->backend;
 
 	if (bgate) {
-		cgs_set_powergating_state(
-					hwmgr->device,
+		amdgpu_device_ip_set_powergating_state(hwmgr->adev,
 					AMD_IP_BLOCK_TYPE_VCE,
 					AMD_PG_STATE_GATE);
-		cgs_set_clockgating_state(
-					hwmgr->device,
+		amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 					AMD_IP_BLOCK_TYPE_VCE,
 					AMD_CG_STATE_GATE);
 		smu8_enable_disable_vce_dpm(hwmgr, false);
@@ -1932,12 +1930,10 @@ static void smu8_dpm_powergate_vce(struct pp_hwmgr *hwmgr, bool bgate)
 	} else {
 		smu8_dpm_powerup_vce(hwmgr);
 		data->vce_power_gated = false;
-		cgs_set_clockgating_state(
-					hwmgr->device,
+		amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 					AMD_IP_BLOCK_TYPE_VCE,
 					AMD_CG_STATE_UNGATE);
-		cgs_set_powergating_state(
-					hwmgr->device,
+		amdgpu_device_ip_set_powergating_state(hwmgr->adev,
 					AMD_IP_BLOCK_TYPE_VCE,
 					AMD_PG_STATE_UNGATE);
 		smu8_dpm_update_vce_dpm(hwmgr);
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c
index d023494..dae3422 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c
@@ -306,13 +306,13 @@ static int fiji_start_smu(struct pp_hwmgr *hwmgr)
 	}
 
 	/* To initialize all clock gating before RLC loaded and running.*/
-	cgs_set_clockgating_state(hwmgr->device,
+	amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 			AMD_IP_BLOCK_TYPE_GFX, AMD_CG_STATE_GATE);
-	cgs_set_clockgating_state(hwmgr->device,
+	amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 			AMD_IP_BLOCK_TYPE_GMC, AMD_CG_STATE_GATE);
-	cgs_set_clockgating_state(hwmgr->device,
+	amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 			AMD_IP_BLOCK_TYPE_SDMA, AMD_CG_STATE_GATE);
-	cgs_set_clockgating_state(hwmgr->device,
+	amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 			AMD_IP_BLOCK_TYPE_COMMON, AMD_CG_STATE_GATE);
 
 	/* Setup SoftRegsStart here for register lookup in case
-- 
cgit v1.1


From 2b816a1d773e755332733a89bdd276e08f935933 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Tue, 27 Mar 2018 16:37:30 -0500
Subject: drm/amdgpu/sdma4: use a helper for SDMA_OP_POLL_REGMEM
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Rather than opencoding it in a bunch of functions.

Reviewed-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c | 62 +++++++++++++++++++---------------
 1 file changed, 34 insertions(+), 28 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
index 399f876..2c618a1 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
@@ -360,6 +360,31 @@ static void sdma_v4_0_ring_emit_ib(struct amdgpu_ring *ring,
 
 }
 
+static void sdma_v4_0_wait_reg_mem(struct amdgpu_ring *ring,
+				   int mem_space, int hdp,
+				   uint32_t addr0, uint32_t addr1,
+				   uint32_t ref, uint32_t mask,
+				   uint32_t inv)
+{
+	amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_POLL_REGMEM) |
+			  SDMA_PKT_POLL_REGMEM_HEADER_HDP_FLUSH(hdp) |
+			  SDMA_PKT_POLL_REGMEM_HEADER_MEM_POLL(mem_space) |
+			  SDMA_PKT_POLL_REGMEM_HEADER_FUNC(3)); /* == */
+	if (mem_space) {
+		/* memory */
+		amdgpu_ring_write(ring, addr0);
+		amdgpu_ring_write(ring, addr1);
+	} else {
+		/* registers */
+		amdgpu_ring_write(ring, addr0 << 2);
+		amdgpu_ring_write(ring, addr1 << 2);
+	}
+	amdgpu_ring_write(ring, ref); /* reference */
+	amdgpu_ring_write(ring, mask); /* mask */
+	amdgpu_ring_write(ring, SDMA_PKT_POLL_REGMEM_DW5_RETRY_COUNT(0xfff) |
+			  SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(inv)); /* retry count, poll interval */
+}
+
 /**
  * sdma_v4_0_ring_emit_hdp_flush - emit an hdp flush on the DMA ring
  *
@@ -378,15 +403,10 @@ static void sdma_v4_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
 	else
 		ref_and_mask = nbio_hf_reg->ref_and_mask_sdma1;
 
-	amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_POLL_REGMEM) |
-			  SDMA_PKT_POLL_REGMEM_HEADER_HDP_FLUSH(1) |
-			  SDMA_PKT_POLL_REGMEM_HEADER_FUNC(3)); /* == */
-	amdgpu_ring_write(ring, (adev->nbio_funcs->get_hdp_flush_done_offset(adev)) << 2);
-	amdgpu_ring_write(ring, (adev->nbio_funcs->get_hdp_flush_req_offset(adev)) << 2);
-	amdgpu_ring_write(ring, ref_and_mask); /* reference */
-	amdgpu_ring_write(ring, ref_and_mask); /* mask */
-	amdgpu_ring_write(ring, SDMA_PKT_POLL_REGMEM_DW5_RETRY_COUNT(0xfff) |
-			  SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(10)); /* retry count, poll interval */
+	sdma_v4_0_wait_reg_mem(ring, 0, 1,
+			       adev->nbio_funcs->get_hdp_flush_done_offset(adev),
+			       adev->nbio_funcs->get_hdp_flush_req_offset(adev),
+			       ref_and_mask, ref_and_mask, 10);
 }
 
 /**
@@ -1114,16 +1134,10 @@ static void sdma_v4_0_ring_emit_pipeline_sync(struct amdgpu_ring *ring)
 	uint64_t addr = ring->fence_drv.gpu_addr;
 
 	/* wait for idle */
-	amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_POLL_REGMEM) |
-			  SDMA_PKT_POLL_REGMEM_HEADER_HDP_FLUSH(0) |
-			  SDMA_PKT_POLL_REGMEM_HEADER_FUNC(3) | /* equal */
-			  SDMA_PKT_POLL_REGMEM_HEADER_MEM_POLL(1));
-	amdgpu_ring_write(ring, addr & 0xfffffffc);
-	amdgpu_ring_write(ring, upper_32_bits(addr) & 0xffffffff);
-	amdgpu_ring_write(ring, seq); /* reference */
-	amdgpu_ring_write(ring, 0xffffffff); /* mask */
-	amdgpu_ring_write(ring, SDMA_PKT_POLL_REGMEM_DW5_RETRY_COUNT(0xfff) |
-			  SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(4)); /* retry count, poll interval */
+	sdma_v4_0_wait_reg_mem(ring, 1, 0,
+			       addr & 0xfffffffc,
+			       upper_32_bits(addr) & 0xffffffff,
+			       seq, 0xffffffff, 4);
 }
 
 
@@ -1154,15 +1168,7 @@ static void sdma_v4_0_ring_emit_wreg(struct amdgpu_ring *ring,
 static void sdma_v4_0_ring_emit_reg_wait(struct amdgpu_ring *ring, uint32_t reg,
 					 uint32_t val, uint32_t mask)
 {
-	amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_POLL_REGMEM) |
-			  SDMA_PKT_POLL_REGMEM_HEADER_HDP_FLUSH(0) |
-			  SDMA_PKT_POLL_REGMEM_HEADER_FUNC(3)); /* equal */
-	amdgpu_ring_write(ring, reg << 2);
-	amdgpu_ring_write(ring, 0);
-	amdgpu_ring_write(ring, val); /* reference */
-	amdgpu_ring_write(ring, mask); /* mask */
-	amdgpu_ring_write(ring, SDMA_PKT_POLL_REGMEM_DW5_RETRY_COUNT(0xfff) |
-			  SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(10));
+	sdma_v4_0_wait_reg_mem(ring, 0, 0, reg, 0, val, mask, 10);
 }
 
 static int sdma_v4_0_early_init(void *handle)
-- 
cgit v1.1


From 3ef1381d4e7ddd3e063cf6fd33df96badfb66839 Mon Sep 17 00:00:00 2001
From: Hawking Zhang <Hawking.Zhang@amd.com>
Date: Wed, 28 Mar 2018 16:23:28 +0800
Subject: drm/amdgpu: add df v1_7 header files

Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../drm/amd/include/asic_reg/df/df_1_7_default.h   | 26 ++++++++++++
 .../drm/amd/include/asic_reg/df/df_1_7_offset.h    | 33 +++++++++++++++
 .../drm/amd/include/asic_reg/df/df_1_7_sh_mask.h   | 48 ++++++++++++++++++++++
 3 files changed, 107 insertions(+)
 create mode 100644 drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_default.h
 create mode 100644 drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_offset.h
 create mode 100644 drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_sh_mask.h

diff --git a/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_default.h b/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_default.h
new file mode 100644
index 0000000..9e19e72
--- /dev/null
+++ b/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_default.h
@@ -0,0 +1,26 @@
+/*
+ * Copyright (C) 2018  Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included
+ * in all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
+ * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN
+ * AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ */
+#ifndef _df_1_7_DEFAULT_HEADER
+#define _df_1_7_DEFAULT_HEADER
+
+#define mmFabricConfigAccessControl_DEFAULT						0x00000000
+
+#endif
diff --git a/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_offset.h b/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_offset.h
new file mode 100644
index 0000000..2b305dd
--- /dev/null
+++ b/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_offset.h
@@ -0,0 +1,33 @@
+/*
+ * Copyright (C) 2018  Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included
+ * in all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
+ * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN
+ * AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ */
+#ifndef _df_1_7_OFFSET_HEADER
+#define _df_1_7_OFFSET_HEADER
+
+#define mmFabricConfigAccessControl									0x0410
+#define mmFabricConfigAccessControl_BASE_IDX								0
+
+#define mmDF_PIE_AON0_DfGlobalClkGater									0x00fc
+#define mmDF_PIE_AON0_DfGlobalClkGater_BASE_IDX								0
+
+#define mmDF_CS_AON0_DramBaseAddress0									0x0044
+#define mmDF_CS_AON0_DramBaseAddress0_BASE_IDX								0
+
+#endif
diff --git a/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_sh_mask.h
new file mode 100644
index 0000000..2ba8497
--- /dev/null
+++ b/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_sh_mask.h
@@ -0,0 +1,48 @@
+/*
+ * Copyright (C) 2018  Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included
+ * in all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
+ * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN
+ * AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ */
+#ifndef _df_1_7_SH_MASK_HEADER
+#define _df_1_7_SH_MASK_HEADER
+
+/* FabricConfigAccessControl */
+#define FabricConfigAccessControl__CfgRegInstAccEn__SHIFT						0x0
+#define FabricConfigAccessControl__CfgRegInstAccRegLock__SHIFT						0x1
+#define FabricConfigAccessControl__CfgRegInstID__SHIFT							0x10
+#define FabricConfigAccessControl__CfgRegInstAccEn_MASK							0x00000001L
+#define FabricConfigAccessControl__CfgRegInstAccRegLock_MASK						0x00000002L
+#define FabricConfigAccessControl__CfgRegInstID_MASK							0x00FF0000L
+
+/* DF_PIE_AON0_DfGlobalClkGater */
+#define DF_PIE_AON0_DfGlobalClkGater__MGCGMode__SHIFT							0x0
+#define DF_PIE_AON0_DfGlobalClkGater__MGCGMode_MASK							0x0000000FL
+
+/* DF_CS_AON0_DramBaseAddress0 */
+#define DF_CS_AON0_DramBaseAddress0__AddrRngVal__SHIFT							0x0
+#define DF_CS_AON0_DramBaseAddress0__LgcyMmioHoleEn__SHIFT						0x1
+#define DF_CS_AON0_DramBaseAddress0__IntLvNumChan__SHIFT						0x4
+#define DF_CS_AON0_DramBaseAddress0__IntLvAddrSel__SHIFT						0x8
+#define DF_CS_AON0_DramBaseAddress0__DramBaseAddr__SHIFT						0xc
+#define DF_CS_AON0_DramBaseAddress0__AddrRngVal_MASK							0x00000001L
+#define DF_CS_AON0_DramBaseAddress0__LgcyMmioHoleEn_MASK						0x00000002L
+#define DF_CS_AON0_DramBaseAddress0__IntLvNumChan_MASK							0x000000F0L
+#define DF_CS_AON0_DramBaseAddress0__IntLvAddrSel_MASK							0x00000700L
+#define DF_CS_AON0_DramBaseAddress0__DramBaseAddr_MASK							0xFFFFF000L
+
+#endif
-- 
cgit v1.1


From 634c96e3f3c7982d4b3ad14f8e004d11af184e91 Mon Sep 17 00:00:00 2001
From: Hawking Zhang <Hawking.Zhang@amd.com>
Date: Fri, 23 Mar 2018 11:37:25 +0800
Subject: drm/amdgpu: add df callback functions structure

Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu.h | 13 ++++++++++++-
 1 file changed, 12 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index 3000c4a..df409dd 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -1368,7 +1368,17 @@ struct amdgpu_nbio_funcs {
 	void (*detect_hw_virt)(struct amdgpu_device *adev);
 };
 
-
+struct amdgpu_df_funcs {
+	void (*init)(struct amdgpu_device *adev);
+	void (*enable_broadcast_mode)(struct amdgpu_device *adev,
+				      bool enable);
+	u32 (*get_fb_channel_number)(struct amdgpu_device *adev);
+	u32 (*get_hbm_channel_number)(struct amdgpu_device *adev);
+	void (*update_medium_grain_clock_gating)(struct amdgpu_device *adev,
+						 bool enable);
+	void (*get_clockgating_state)(struct amdgpu_device *adev,
+				      u32 *flags);
+};
 /* Define the HW IP blocks will be used in driver , add more if necessary */
 enum amd_hw_ip_block_type {
 	GC_HWIP = 1,
@@ -1588,6 +1598,7 @@ struct amdgpu_device {
 	uint32_t 		*reg_offset[MAX_HWIP][HWIP_MAX_INSTANCE];
 
 	const struct amdgpu_nbio_funcs	*nbio_funcs;
+	const struct amdgpu_df_funcs	*df_funcs;
 
 	/* delayed work_func for deferring clockgating during resume */
 	struct delayed_work     late_init_work;
-- 
cgit v1.1


From d99605ead70efa0dc259c28f9b258184e2b3e77c Mon Sep 17 00:00:00 2001
From: Hawking Zhang <Hawking.Zhang@amd.com>
Date: Wed, 28 Mar 2018 16:27:56 +0800
Subject: drm/amdgpu/df: implement df v1_7 callback functions

Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/Makefile  |   4 ++
 drivers/gpu/drm/amd/amdgpu/df_v1_7.c | 112 +++++++++++++++++++++++++++++++++++
 drivers/gpu/drm/amd/amdgpu/df_v1_7.h |  40 +++++++++++++
 3 files changed, 156 insertions(+)
 create mode 100644 drivers/gpu/drm/amd/amdgpu/df_v1_7.c
 create mode 100644 drivers/gpu/drm/amd/amdgpu/df_v1_7.h

diff --git a/drivers/gpu/drm/amd/amdgpu/Makefile b/drivers/gpu/drm/amd/amdgpu/Makefile
index 2ca2b51..2fe4a0b 100644
--- a/drivers/gpu/drm/amd/amdgpu/Makefile
+++ b/drivers/gpu/drm/amd/amdgpu/Makefile
@@ -64,6 +64,10 @@ amdgpu-$(CONFIG_DRM_AMDGPU_SI)+= si.o gmc_v6_0.o gfx_v6_0.o si_ih.o si_dma.o dce
 amdgpu-y += \
 	vi.o mxgpu_vi.o nbio_v6_1.o soc15.o emu_soc.o mxgpu_ai.o nbio_v7_0.o vega10_reg_init.o
 
+# add DF block
+amdgpu-y += \
+	df_v1_7.o
+
 # add GMC block
 amdgpu-y += \
 	gmc_v7_0.o \
diff --git a/drivers/gpu/drm/amd/amdgpu/df_v1_7.c b/drivers/gpu/drm/amd/amdgpu/df_v1_7.c
new file mode 100644
index 0000000..4ffda99
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/df_v1_7.c
@@ -0,0 +1,112 @@
+/*
+ * Copyright 2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#include "amdgpu.h"
+#include "df_v1_7.h"
+
+#include "df/df_1_7_default.h"
+#include "df/df_1_7_offset.h"
+#include "df/df_1_7_sh_mask.h"
+
+static u32 df_v1_7_channel_number[] = {1, 2, 0, 4, 0, 8, 0, 16, 2};
+
+static void df_v1_7_init (struct amdgpu_device *adev)
+{
+}
+
+static void df_v1_7_enable_broadcast_mode(struct amdgpu_device *adev,
+                                          bool enable)
+{
+	u32 tmp;
+
+	if (enable) {
+		tmp = RREG32_SOC15(DF, 0, mmFabricConfigAccessControl);
+		tmp &= ~FabricConfigAccessControl__CfgRegInstAccEn_MASK;
+		WREG32_SOC15(DF, 0, mmFabricConfigAccessControl, tmp);
+	} else
+		WREG32_SOC15(DF, 0, mmFabricConfigAccessControl,
+			     mmFabricConfigAccessControl_DEFAULT);
+}
+
+static u32 df_v1_7_get_fb_channel_number(struct amdgpu_device *adev)
+{
+	u32 tmp;
+
+	tmp = RREG32_SOC15(DF, 0, mmDF_CS_AON0_DramBaseAddress0);
+	tmp &= DF_CS_AON0_DramBaseAddress0__IntLvNumChan_MASK;
+	tmp >>= DF_CS_AON0_DramBaseAddress0__IntLvNumChan__SHIFT;
+
+	return tmp;
+}
+
+static u32 df_v1_7_get_hbm_channel_number(struct amdgpu_device *adev)
+{
+	int fb_channel_number;
+
+	fb_channel_number = adev->df_funcs->get_fb_channel_number(adev);
+
+	return df_v1_7_channel_number[fb_channel_number];
+}
+
+static void df_v1_7_update_medium_grain_clock_gating(struct amdgpu_device *adev,
+						     bool enable)
+{
+	u32 tmp;
+
+	/* Put DF on broadcast mode */
+	adev->df_funcs->enable_broadcast_mode(adev, true);
+
+	if (enable && (adev->cg_flags & AMD_CG_SUPPORT_DF_MGCG)) {
+		tmp = RREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater);
+		tmp &= ~DF_PIE_AON0_DfGlobalClkGater__MGCGMode_MASK;
+		tmp |= DF_V1_7_MGCG_ENABLE_15_CYCLE_DELAY;
+		WREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater, tmp);
+	} else {
+		tmp = RREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater);
+		tmp &= ~DF_PIE_AON0_DfGlobalClkGater__MGCGMode_MASK;
+		tmp |= DF_V1_7_MGCG_DISABLE;
+		WREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater, tmp);
+	}
+
+	/* Exit boradcast mode */
+	adev->df_funcs->enable_broadcast_mode(adev, false);
+}
+
+static void df_v1_7_get_clockgating_state(struct amdgpu_device *adev,
+					  u32 *flags)
+{
+	u32 tmp;
+
+	/* AMD_CG_SUPPORT_DF_MGCG */
+	tmp = RREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater);
+	if (tmp & DF_V1_7_MGCG_ENABLE_15_CYCLE_DELAY)
+		*flags |= AMD_CG_SUPPORT_DF_MGCG;
+}
+
+const struct amdgpu_df_funcs df_v1_7_funcs = {
+	.init = df_v1_7_init,
+	.enable_broadcast_mode = df_v1_7_enable_broadcast_mode,
+	.get_fb_channel_number = df_v1_7_get_fb_channel_number,
+	.get_hbm_channel_number = df_v1_7_get_hbm_channel_number,
+	.update_medium_grain_clock_gating = df_v1_7_update_medium_grain_clock_gating,
+	.get_clockgating_state = df_v1_7_get_clockgating_state,
+};
diff --git a/drivers/gpu/drm/amd/amdgpu/df_v1_7.h b/drivers/gpu/drm/amd/amdgpu/df_v1_7.h
new file mode 100644
index 0000000..7462110
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/df_v1_7.h
@@ -0,0 +1,40 @@
+/*
+ * Copyright 2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef __DF_V1_7_H__
+#define __DF_V1_7_H__
+
+#include "soc15_common.h"
+enum DF_V1_7_MGCG
+{
+	DF_V1_7_MGCG_DISABLE = 0,
+	DF_V1_7_MGCG_ENABLE_00_CYCLE_DELAY =1,
+	DF_V1_7_MGCG_ENABLE_01_CYCLE_DELAY =2,
+	DF_V1_7_MGCG_ENABLE_15_CYCLE_DELAY =13,
+	DF_V1_7_MGCG_ENABLE_31_CYCLE_DELAY =14,
+	DF_V1_7_MGCG_ENABLE_63_CYCLE_DELAY =15
+};
+
+extern const struct amdgpu_df_funcs df_v1_7_funcs;
+
+#endif
-- 
cgit v1.1


From 070706c03b3e67207cc41bd97b67ff0930d79cb3 Mon Sep 17 00:00:00 2001
From: Hawking Zhang <Hawking.Zhang@amd.com>
Date: Wed, 28 Mar 2018 17:08:04 +0800
Subject: drm/amdgpu: switch to use df callback functions

Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c | 35 +-------------------
 drivers/gpu/drm/amd/amdgpu/soc15.c    | 62 +++--------------------------------
 2 files changed, 5 insertions(+), 92 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
index e687363..070946e 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
@@ -714,7 +714,6 @@ static void gmc_v9_0_vram_gtt_location(struct amdgpu_device *adev,
  */
 static int gmc_v9_0_mc_init(struct amdgpu_device *adev)
 {
-	u32 tmp;
 	int chansize, numchan;
 	int r;
 
@@ -727,39 +726,7 @@ static int gmc_v9_0_mc_init(struct amdgpu_device *adev)
 		else
 			chansize = 128;
 
-		tmp = RREG32_SOC15(DF, 0, mmDF_CS_AON0_DramBaseAddress0);
-		tmp &= DF_CS_AON0_DramBaseAddress0__IntLvNumChan_MASK;
-		tmp >>= DF_CS_AON0_DramBaseAddress0__IntLvNumChan__SHIFT;
-		switch (tmp) {
-		case 0:
-		default:
-			numchan = 1;
-			break;
-		case 1:
-			numchan = 2;
-			break;
-		case 2:
-			numchan = 0;
-			break;
-		case 3:
-			numchan = 4;
-			break;
-		case 4:
-			numchan = 0;
-			break;
-		case 5:
-			numchan = 8;
-			break;
-		case 6:
-			numchan = 0;
-			break;
-		case 7:
-			numchan = 16;
-			break;
-		case 8:
-			numchan = 2;
-			break;
-		}
+		numchan = adev->df_funcs->get_hbm_channel_number(adev);
 		adev->gmc.vram_width = numchan * chansize;
 	}
 
diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index 51cf8a3..654b015 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -52,6 +52,7 @@
 #include "gmc_v9_0.h"
 #include "gfxhub_v1_0.h"
 #include "mmhub_v1_0.h"
+#include "df_v1_7.h"
 #include "vega10_ih.h"
 #include "sdma_v4_0.h"
 #include "uvd_v7_0.h"
@@ -60,33 +61,6 @@
 #include "dce_virtual.h"
 #include "mxgpu_ai.h"
 
-#define mmFabricConfigAccessControl                                                                    0x0410
-#define mmFabricConfigAccessControl_BASE_IDX                                                           0
-#define mmFabricConfigAccessControl_DEFAULT                                      0x00000000
-//FabricConfigAccessControl
-#define FabricConfigAccessControl__CfgRegInstAccEn__SHIFT                                                     0x0
-#define FabricConfigAccessControl__CfgRegInstAccRegLock__SHIFT                                                0x1
-#define FabricConfigAccessControl__CfgRegInstID__SHIFT                                                        0x10
-#define FabricConfigAccessControl__CfgRegInstAccEn_MASK                                                       0x00000001L
-#define FabricConfigAccessControl__CfgRegInstAccRegLock_MASK                                                  0x00000002L
-#define FabricConfigAccessControl__CfgRegInstID_MASK                                                          0x00FF0000L
-
-
-#define mmDF_PIE_AON0_DfGlobalClkGater                                                                 0x00fc
-#define mmDF_PIE_AON0_DfGlobalClkGater_BASE_IDX                                                        0
-//DF_PIE_AON0_DfGlobalClkGater
-#define DF_PIE_AON0_DfGlobalClkGater__MGCGMode__SHIFT                                                         0x0
-#define DF_PIE_AON0_DfGlobalClkGater__MGCGMode_MASK                                                           0x0000000FL
-
-enum {
-	DF_MGCG_DISABLE = 0,
-	DF_MGCG_ENABLE_00_CYCLE_DELAY =1,
-	DF_MGCG_ENABLE_01_CYCLE_DELAY =2,
-	DF_MGCG_ENABLE_15_CYCLE_DELAY =13,
-	DF_MGCG_ENABLE_31_CYCLE_DELAY =14,
-	DF_MGCG_ENABLE_63_CYCLE_DELAY =15
-};
-
 #define mmMP0_MISC_CGTT_CTRL0                                                                   0x01b9
 #define mmMP0_MISC_CGTT_CTRL0_BASE_IDX                                                          0
 #define mmMP0_MISC_LIGHT_SLEEP_CTRL                                                             0x01ba
@@ -521,6 +495,7 @@ int soc15_set_ip_blocks(struct amdgpu_device *adev)
 	else
 		adev->nbio_funcs = &nbio_v6_1_funcs;
 
+	adev->df_funcs = &df_v1_7_funcs;
 	adev->nbio_funcs->detect_hw_virt(adev);
 
 	if (amdgpu_sriov_vf(adev))
@@ -871,32 +846,6 @@ static void soc15_update_rom_medium_grain_clock_gating(struct amdgpu_device *ade
 		WREG32(SOC15_REG_OFFSET(SMUIO, 0, mmCGTT_ROM_CLK_CTRL0), data);
 }
 
-static void soc15_update_df_medium_grain_clock_gating(struct amdgpu_device *adev,
-						       bool enable)
-{
-	uint32_t data;
-
-	/* Put DF on broadcast mode */
-	data = RREG32(SOC15_REG_OFFSET(DF, 0, mmFabricConfigAccessControl));
-	data &= ~FabricConfigAccessControl__CfgRegInstAccEn_MASK;
-	WREG32(SOC15_REG_OFFSET(DF, 0, mmFabricConfigAccessControl), data);
-
-	if (enable && (adev->cg_flags & AMD_CG_SUPPORT_DF_MGCG)) {
-		data = RREG32(SOC15_REG_OFFSET(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater));
-		data &= ~DF_PIE_AON0_DfGlobalClkGater__MGCGMode_MASK;
-		data |= DF_MGCG_ENABLE_15_CYCLE_DELAY;
-		WREG32(SOC15_REG_OFFSET(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater), data);
-	} else {
-		data = RREG32(SOC15_REG_OFFSET(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater));
-		data &= ~DF_PIE_AON0_DfGlobalClkGater__MGCGMode_MASK;
-		data |= DF_MGCG_DISABLE;
-		WREG32(SOC15_REG_OFFSET(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater), data);
-	}
-
-	WREG32(SOC15_REG_OFFSET(DF, 0, mmFabricConfigAccessControl),
-	       mmFabricConfigAccessControl_DEFAULT);
-}
-
 static int soc15_common_set_clockgating_state(void *handle,
 					    enum amd_clockgating_state state)
 {
@@ -920,7 +869,7 @@ static int soc15_common_set_clockgating_state(void *handle,
 				state == AMD_CG_STATE_GATE ? true : false);
 		soc15_update_rom_medium_grain_clock_gating(adev,
 				state == AMD_CG_STATE_GATE ? true : false);
-		soc15_update_df_medium_grain_clock_gating(adev,
+		adev->df_funcs->update_medium_grain_clock_gating(adev,
 				state == AMD_CG_STATE_GATE ? true : false);
 		break;
 	case CHIP_RAVEN:
@@ -973,10 +922,7 @@ static void soc15_common_get_clockgating_state(void *handle, u32 *flags)
 	if (!(data & CGTT_ROM_CLK_CTRL0__SOFT_OVERRIDE0_MASK))
 		*flags |= AMD_CG_SUPPORT_ROM_MGCG;
 
-	/* AMD_CG_SUPPORT_DF_MGCG */
-	data = RREG32(SOC15_REG_OFFSET(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater));
-	if (data & DF_MGCG_ENABLE_15_CYCLE_DELAY)
-		*flags |= AMD_CG_SUPPORT_DF_MGCG;
+	adev->df_funcs->get_clockgating_state(adev, flags);
 }
 
 static int soc15_common_set_powergating_state(void *handle,
-- 
cgit v1.1


From c99c7d6ef9f610145308577ae5845cd6e14051e2 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Fri, 30 Mar 2018 13:05:44 +0800
Subject: drm/amd/display: Disentangle dc.h include from amdgpu.h

Use forward declaration in amdgpu_dm.h for struct dc instand
of include dc.h to make header files more standalone

Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h       | 2 +-
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c | 1 +
 2 files changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
index b68400c..3af699b 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
@@ -28,7 +28,6 @@
 
 #include <drm/drmP.h>
 #include <drm/drm_atomic.h>
-#include "dc.h"
 
 /*
  * This file contains the definition for amdgpu_display_manager
@@ -53,6 +52,7 @@
 struct amdgpu_device;
 struct drm_device;
 struct amdgpu_dm_irq_handler_data;
+struct dc;
 
 struct amdgpu_dm_prev_state {
 	struct drm_framebuffer *fb;
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
index f6cb502..ef5fad8 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
@@ -25,6 +25,7 @@
 
 #include "amdgpu_mode.h"
 #include "amdgpu_dm.h"
+#include "dc.h"
 #include "modules/color/color_gamma.h"
 
 #define MAX_DRM_LUT_VALUE 0xFFFF
-- 
cgit v1.1


From a110dfe3ab28fd4c52f785a7b80131ef5e8b61fb Mon Sep 17 00:00:00 2001
From: Daniel Stone <daniels@collabora.com>
Date: Fri, 30 Mar 2018 15:11:36 +0100
Subject: drm/radeon: Move GEM BO to drm_framebuffer
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Since drm_framebuffer can now store GEM objects directly, place them
there rather than in our own subclass. As this makes the framebuffer
create_handle and destroy functions the same as the GEM framebuffer
helper, we can reuse those.

Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Daniel Stone <daniels@collabora.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Cc: David (ChunMing) Zhou <David1.Zhou@amd.com>
Cc: amd-gfx@lists.freedesktop.org
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/radeon/atombios_crtc.c      | 10 +++++-----
 drivers/gpu/drm/radeon/radeon_device.c      |  4 ++--
 drivers/gpu/drm/radeon/radeon_display.c     | 31 +++++++----------------------
 drivers/gpu/drm/radeon/radeon_fb.c          |  8 ++++----
 drivers/gpu/drm/radeon/radeon_legacy_crtc.c | 11 ++++------
 drivers/gpu/drm/radeon/radeon_mode.h        |  1 -
 6 files changed, 22 insertions(+), 43 deletions(-)

diff --git a/drivers/gpu/drm/radeon/atombios_crtc.c b/drivers/gpu/drm/radeon/atombios_crtc.c
index 02baaaf..028a811 100644
--- a/drivers/gpu/drm/radeon/atombios_crtc.c
+++ b/drivers/gpu/drm/radeon/atombios_crtc.c
@@ -1176,7 +1176,7 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 	/* If atomic, assume fb object is pinned & idle & fenced and
 	 * just update base pointers
 	 */
-	obj = radeon_fb->obj;
+	obj = radeon_fb->base.obj[0];
 	rbo = gem_to_radeon_bo(obj);
 	r = radeon_bo_reserve(rbo, false);
 	if (unlikely(r != 0))
@@ -1442,7 +1442,7 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 
 	if (!atomic && fb && fb != crtc->primary->fb) {
 		radeon_fb = to_radeon_framebuffer(fb);
-		rbo = gem_to_radeon_bo(radeon_fb->obj);
+		rbo = gem_to_radeon_bo(radeon_fb->base.obj[0]);
 		r = radeon_bo_reserve(rbo, false);
 		if (unlikely(r != 0))
 			return r;
@@ -1490,7 +1490,7 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc,
 		target_fb = crtc->primary->fb;
 	}
 
-	obj = radeon_fb->obj;
+	obj = radeon_fb->base.obj[0];
 	rbo = gem_to_radeon_bo(obj);
 	r = radeon_bo_reserve(rbo, false);
 	if (unlikely(r != 0))
@@ -1642,7 +1642,7 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc,
 
 	if (!atomic && fb && fb != crtc->primary->fb) {
 		radeon_fb = to_radeon_framebuffer(fb);
-		rbo = gem_to_radeon_bo(radeon_fb->obj);
+		rbo = gem_to_radeon_bo(radeon_fb->base.obj[0]);
 		r = radeon_bo_reserve(rbo, false);
 		if (unlikely(r != 0))
 			return r;
@@ -2153,7 +2153,7 @@ static void atombios_crtc_disable(struct drm_crtc *crtc)
 		struct radeon_bo *rbo;
 
 		radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
-		rbo = gem_to_radeon_bo(radeon_fb->obj);
+		rbo = gem_to_radeon_bo(radeon_fb->base.obj[0]);
 		r = radeon_bo_reserve(rbo, false);
 		if (unlikely(r))
 			DRM_ERROR("failed to reserve rbo before unpin\n");
diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c
index e415d2c..30c5bc2 100644
--- a/drivers/gpu/drm/radeon/radeon_device.c
+++ b/drivers/gpu/drm/radeon/radeon_device.c
@@ -1599,10 +1599,10 @@ int radeon_suspend_kms(struct drm_device *dev, bool suspend,
 			}
 		}
 
-		if (rfb == NULL || rfb->obj == NULL) {
+		if (rfb == NULL || rfb->base.obj[0] == NULL) {
 			continue;
 		}
-		robj = gem_to_radeon_bo(rfb->obj);
+		robj = gem_to_radeon_bo(rfb->base.obj[0]);
 		/* don't unpin kernel fb objects */
 		if (!radeon_fbdev_robj_is_fb(rdev, robj)) {
 			r = radeon_bo_reserve(robj, false);
diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c
index 26129b2..dc30012 100644
--- a/drivers/gpu/drm/radeon/radeon_display.c
+++ b/drivers/gpu/drm/radeon/radeon_display.c
@@ -32,6 +32,7 @@
 
 #include <linux/pm_runtime.h>
 #include <drm/drm_crtc_helper.h>
+#include <drm/drm_gem_framebuffer_helper.h>
 #include <drm/drm_fb_helper.h>
 #include <drm/drm_plane_helper.h>
 #include <drm/drm_edid.h>
@@ -502,14 +503,14 @@ static int radeon_crtc_page_flip_target(struct drm_crtc *crtc,
 
 	/* schedule unpin of the old buffer */
 	old_radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
-	obj = old_radeon_fb->obj;
+	obj = old_radeon_fb->base.obj[0];
 
 	/* take a reference to the old object */
 	drm_gem_object_get(obj);
 	work->old_rbo = gem_to_radeon_bo(obj);
 
 	new_radeon_fb = to_radeon_framebuffer(fb);
-	obj = new_radeon_fb->obj;
+	obj = new_radeon_fb->base.obj[0];
 	new_rbo = gem_to_radeon_bo(obj);
 
 	/* pin the new buffer */
@@ -1285,27 +1286,9 @@ void radeon_compute_pll_legacy(struct radeon_pll *pll,
 
 }
 
-static void radeon_user_framebuffer_destroy(struct drm_framebuffer *fb)
-{
-	struct radeon_framebuffer *radeon_fb = to_radeon_framebuffer(fb);
-
-	drm_gem_object_put_unlocked(radeon_fb->obj);
-	drm_framebuffer_cleanup(fb);
-	kfree(radeon_fb);
-}
-
-static int radeon_user_framebuffer_create_handle(struct drm_framebuffer *fb,
-						  struct drm_file *file_priv,
-						  unsigned int *handle)
-{
-	struct radeon_framebuffer *radeon_fb = to_radeon_framebuffer(fb);
-
-	return drm_gem_handle_create(file_priv, radeon_fb->obj, handle);
-}
-
 static const struct drm_framebuffer_funcs radeon_fb_funcs = {
-	.destroy = radeon_user_framebuffer_destroy,
-	.create_handle = radeon_user_framebuffer_create_handle,
+	.destroy = drm_gem_fb_destroy,
+	.create_handle = drm_gem_fb_create_handle,
 };
 
 int
@@ -1315,11 +1298,11 @@ radeon_framebuffer_init(struct drm_device *dev,
 			struct drm_gem_object *obj)
 {
 	int ret;
-	rfb->obj = obj;
+	rfb->base.obj[0] = obj;
 	drm_helper_mode_fill_fb_struct(dev, &rfb->base, mode_cmd);
 	ret = drm_framebuffer_init(dev, &rfb->base, &radeon_fb_funcs);
 	if (ret) {
-		rfb->obj = NULL;
+		rfb->base.obj[0] = NULL;
 		return ret;
 	}
 	return 0;
diff --git a/drivers/gpu/drm/radeon/radeon_fb.c b/drivers/gpu/drm/radeon/radeon_fb.c
index 57c5404..6cd99f6 100644
--- a/drivers/gpu/drm/radeon/radeon_fb.c
+++ b/drivers/gpu/drm/radeon/radeon_fb.c
@@ -312,9 +312,9 @@ static int radeon_fbdev_destroy(struct drm_device *dev, struct radeon_fbdev *rfb
 
 	drm_fb_helper_unregister_fbi(&rfbdev->helper);
 
-	if (rfb->obj) {
-		radeonfb_destroy_pinned_object(rfb->obj);
-		rfb->obj = NULL;
+	if (rfb->base.obj[0]) {
+		radeonfb_destroy_pinned_object(rfb->base.obj[0]);
+		rfb->base.obj[0] = NULL;
 		drm_framebuffer_unregister_private(&rfb->base);
 		drm_framebuffer_cleanup(&rfb->base);
 	}
@@ -400,7 +400,7 @@ bool radeon_fbdev_robj_is_fb(struct radeon_device *rdev, struct radeon_bo *robj)
 	if (!rdev->mode_info.rfbdev)
 		return false;
 
-	if (robj == gem_to_radeon_bo(rdev->mode_info.rfbdev->rfb.obj))
+	if (robj == gem_to_radeon_bo(rdev->mode_info.rfbdev->rfb.base.obj[0]))
 		return true;
 	return false;
 }
diff --git a/drivers/gpu/drm/radeon/radeon_legacy_crtc.c b/drivers/gpu/drm/radeon/radeon_legacy_crtc.c
index 1f1856e..50b3f55 100644
--- a/drivers/gpu/drm/radeon/radeon_legacy_crtc.c
+++ b/drivers/gpu/drm/radeon/radeon_legacy_crtc.c
@@ -423,7 +423,7 @@ int radeon_crtc_do_set_base(struct drm_crtc *crtc,
 	}
 
 	/* Pin framebuffer & get tilling informations */
-	obj = radeon_fb->obj;
+	obj = radeon_fb->base.obj[0];
 	rbo = gem_to_radeon_bo(obj);
 retry:
 	r = radeon_bo_reserve(rbo, false);
@@ -451,7 +451,7 @@ retry:
 			struct radeon_bo *old_rbo;
 			unsigned long nsize, osize;
 
-			old_rbo = gem_to_radeon_bo(to_radeon_framebuffer(fb)->obj);
+			old_rbo = gem_to_radeon_bo(fb->obj[0]);
 			osize = radeon_bo_size(old_rbo);
 			nsize = radeon_bo_size(rbo);
 			if (nsize <= osize && !radeon_bo_reserve(old_rbo, false)) {
@@ -558,8 +558,7 @@ retry:
 	WREG32(RADEON_CRTC_PITCH + radeon_crtc->crtc_offset, crtc_pitch);
 
 	if (!atomic && fb && fb != crtc->primary->fb) {
-		radeon_fb = to_radeon_framebuffer(fb);
-		rbo = gem_to_radeon_bo(radeon_fb->obj);
+		rbo = gem_to_radeon_bo(fb->obj[0]);
 		r = radeon_bo_reserve(rbo, false);
 		if (unlikely(r != 0))
 			return r;
@@ -1093,11 +1092,9 @@ static void radeon_crtc_disable(struct drm_crtc *crtc)
 	radeon_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
 	if (crtc->primary->fb) {
 		int r;
-		struct radeon_framebuffer *radeon_fb;
 		struct radeon_bo *rbo;
 
-		radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
-		rbo = gem_to_radeon_bo(radeon_fb->obj);
+		rbo = gem_to_radeon_bo(crtc->primary->fb->obj[0]);
 		r = radeon_bo_reserve(rbo, false);
 		if (unlikely(r))
 			DRM_ERROR("failed to reserve rbo before unpin\n");
diff --git a/drivers/gpu/drm/radeon/radeon_mode.h b/drivers/gpu/drm/radeon/radeon_mode.h
index 3243e5e..cd93c80 100644
--- a/drivers/gpu/drm/radeon/radeon_mode.h
+++ b/drivers/gpu/drm/radeon/radeon_mode.h
@@ -576,7 +576,6 @@ struct radeon_connector {
 
 struct radeon_framebuffer {
 	struct drm_framebuffer base;
-	struct drm_gem_object *obj;
 };
 
 #define ENCODER_MODE_IS_DP(em) (((em) == ATOM_ENCODER_MODE_DP) || \
-- 
cgit v1.1


From 9a0f0c9d0c9de76324807571694f8c89b433e3f1 Mon Sep 17 00:00:00 2001
From: Daniel Stone <daniels@collabora.com>
Date: Fri, 30 Mar 2018 15:11:37 +0100
Subject: drm/radeon: radeon_framebuffer -> drm_framebuffer
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Since drm_framebuffer can now store GEM objects directly, place them
there rather than in our own subclass. As this makes the framebuffer
create_handle and destroy functions the same as the GEM framebuffer
helper, we can reuse those.

Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Daniel Stone <daniels@collabora.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Cc: David (ChunMing) Zhou <David1.Zhou@amd.com>
Cc: amd-gfx@lists.freedesktop.org
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/radeon/atombios_crtc.c      | 32 ++++++++---------------------
 drivers/gpu/drm/radeon/radeon_device.c      |  6 +++---
 drivers/gpu/drm/radeon/radeon_display.c     | 30 ++++++++++++---------------
 drivers/gpu/drm/radeon/radeon_fb.c          | 20 +++++++++---------
 drivers/gpu/drm/radeon/radeon_legacy_crtc.c | 11 +++-------
 drivers/gpu/drm/radeon/radeon_mode.h        |  7 +------
 6 files changed, 39 insertions(+), 67 deletions(-)

diff --git a/drivers/gpu/drm/radeon/atombios_crtc.c b/drivers/gpu/drm/radeon/atombios_crtc.c
index 028a811..efbd581 100644
--- a/drivers/gpu/drm/radeon/atombios_crtc.c
+++ b/drivers/gpu/drm/radeon/atombios_crtc.c
@@ -1145,7 +1145,6 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct radeon_device *rdev = dev->dev_private;
-	struct radeon_framebuffer *radeon_fb;
 	struct drm_framebuffer *target_fb;
 	struct drm_gem_object *obj;
 	struct radeon_bo *rbo;
@@ -1164,19 +1163,15 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 		return 0;
 	}
 
-	if (atomic) {
-		radeon_fb = to_radeon_framebuffer(fb);
+	if (atomic)
 		target_fb = fb;
-	}
-	else {
-		radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
+	else
 		target_fb = crtc->primary->fb;
-	}
 
 	/* If atomic, assume fb object is pinned & idle & fenced and
 	 * just update base pointers
 	 */
-	obj = radeon_fb->base.obj[0];
+	obj = target_fb->obj[0];
 	rbo = gem_to_radeon_bo(obj);
 	r = radeon_bo_reserve(rbo, false);
 	if (unlikely(r != 0))
@@ -1441,8 +1436,7 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 	WREG32(EVERGREEN_MASTER_UPDATE_MODE + radeon_crtc->crtc_offset, 0);
 
 	if (!atomic && fb && fb != crtc->primary->fb) {
-		radeon_fb = to_radeon_framebuffer(fb);
-		rbo = gem_to_radeon_bo(radeon_fb->base.obj[0]);
+		rbo = gem_to_radeon_bo(fb->obj[0]);
 		r = radeon_bo_reserve(rbo, false);
 		if (unlikely(r != 0))
 			return r;
@@ -1463,7 +1457,6 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc,
 	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct radeon_device *rdev = dev->dev_private;
-	struct radeon_framebuffer *radeon_fb;
 	struct drm_gem_object *obj;
 	struct radeon_bo *rbo;
 	struct drm_framebuffer *target_fb;
@@ -1481,16 +1474,12 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc,
 		return 0;
 	}
 
-	if (atomic) {
-		radeon_fb = to_radeon_framebuffer(fb);
+	if (atomic)
 		target_fb = fb;
-	}
-	else {
-		radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
+	else
 		target_fb = crtc->primary->fb;
-	}
 
-	obj = radeon_fb->base.obj[0];
+	obj = target_fb->obj[0];
 	rbo = gem_to_radeon_bo(obj);
 	r = radeon_bo_reserve(rbo, false);
 	if (unlikely(r != 0))
@@ -1641,8 +1630,7 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc,
 	WREG32(AVIVO_D1MODE_MASTER_UPDATE_MODE + radeon_crtc->crtc_offset, 3);
 
 	if (!atomic && fb && fb != crtc->primary->fb) {
-		radeon_fb = to_radeon_framebuffer(fb);
-		rbo = gem_to_radeon_bo(radeon_fb->base.obj[0]);
+		rbo = gem_to_radeon_bo(fb->obj[0]);
 		r = radeon_bo_reserve(rbo, false);
 		if (unlikely(r != 0))
 			return r;
@@ -2149,11 +2137,9 @@ static void atombios_crtc_disable(struct drm_crtc *crtc)
 	atombios_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
 	if (crtc->primary->fb) {
 		int r;
-		struct radeon_framebuffer *radeon_fb;
 		struct radeon_bo *rbo;
 
-		radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
-		rbo = gem_to_radeon_bo(radeon_fb->base.obj[0]);
+		rbo = gem_to_radeon_bo(crtc->primary->fb->obj[0]);
 		r = radeon_bo_reserve(rbo, false);
 		if (unlikely(r))
 			DRM_ERROR("failed to reserve rbo before unpin\n");
diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c
index 30c5bc2..90e17e2 100644
--- a/drivers/gpu/drm/radeon/radeon_device.c
+++ b/drivers/gpu/drm/radeon/radeon_device.c
@@ -1587,7 +1587,7 @@ int radeon_suspend_kms(struct drm_device *dev, bool suspend,
 	/* unpin the front buffers and cursors */
 	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
 		struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
-		struct radeon_framebuffer *rfb = to_radeon_framebuffer(crtc->primary->fb);
+		struct drm_framebuffer *fb = crtc->primary->fb;
 		struct radeon_bo *robj;
 
 		if (radeon_crtc->cursor_bo) {
@@ -1599,10 +1599,10 @@ int radeon_suspend_kms(struct drm_device *dev, bool suspend,
 			}
 		}
 
-		if (rfb == NULL || rfb->base.obj[0] == NULL) {
+		if (fb == NULL || fb->obj[0] == NULL) {
 			continue;
 		}
-		robj = gem_to_radeon_bo(rfb->base.obj[0]);
+		robj = gem_to_radeon_bo(fb->obj[0]);
 		/* don't unpin kernel fb objects */
 		if (!radeon_fbdev_robj_is_fb(rdev, robj)) {
 			r = radeon_bo_reserve(robj, false);
diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c
index dc30012..9d3ac8b 100644
--- a/drivers/gpu/drm/radeon/radeon_display.c
+++ b/drivers/gpu/drm/radeon/radeon_display.c
@@ -479,8 +479,6 @@ static int radeon_crtc_page_flip_target(struct drm_crtc *crtc,
 	struct drm_device *dev = crtc->dev;
 	struct radeon_device *rdev = dev->dev_private;
 	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
-	struct radeon_framebuffer *old_radeon_fb;
-	struct radeon_framebuffer *new_radeon_fb;
 	struct drm_gem_object *obj;
 	struct radeon_flip_work *work;
 	struct radeon_bo *new_rbo;
@@ -502,15 +500,13 @@ static int radeon_crtc_page_flip_target(struct drm_crtc *crtc,
 	work->async = (page_flip_flags & DRM_MODE_PAGE_FLIP_ASYNC) != 0;
 
 	/* schedule unpin of the old buffer */
-	old_radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
-	obj = old_radeon_fb->base.obj[0];
+	obj = crtc->primary->fb->obj[0];
 
 	/* take a reference to the old object */
 	drm_gem_object_get(obj);
 	work->old_rbo = gem_to_radeon_bo(obj);
 
-	new_radeon_fb = to_radeon_framebuffer(fb);
-	obj = new_radeon_fb->base.obj[0];
+	obj = fb->obj[0];
 	new_rbo = gem_to_radeon_bo(obj);
 
 	/* pin the new buffer */
@@ -1293,16 +1289,16 @@ static const struct drm_framebuffer_funcs radeon_fb_funcs = {
 
 int
 radeon_framebuffer_init(struct drm_device *dev,
-			struct radeon_framebuffer *rfb,
+			struct drm_framebuffer *fb,
 			const struct drm_mode_fb_cmd2 *mode_cmd,
 			struct drm_gem_object *obj)
 {
 	int ret;
-	rfb->base.obj[0] = obj;
-	drm_helper_mode_fill_fb_struct(dev, &rfb->base, mode_cmd);
-	ret = drm_framebuffer_init(dev, &rfb->base, &radeon_fb_funcs);
+	fb->obj[0] = obj;
+	drm_helper_mode_fill_fb_struct(dev, fb, mode_cmd);
+	ret = drm_framebuffer_init(dev, fb, &radeon_fb_funcs);
 	if (ret) {
-		rfb->base.obj[0] = NULL;
+		fb->obj[0] = NULL;
 		return ret;
 	}
 	return 0;
@@ -1314,7 +1310,7 @@ radeon_user_framebuffer_create(struct drm_device *dev,
 			       const struct drm_mode_fb_cmd2 *mode_cmd)
 {
 	struct drm_gem_object *obj;
-	struct radeon_framebuffer *radeon_fb;
+	struct drm_framebuffer *fb;
 	int ret;
 
 	obj = drm_gem_object_lookup(file_priv, mode_cmd->handles[0]);
@@ -1330,20 +1326,20 @@ radeon_user_framebuffer_create(struct drm_device *dev,
 		return ERR_PTR(-EINVAL);
 	}
 
-	radeon_fb = kzalloc(sizeof(*radeon_fb), GFP_KERNEL);
-	if (radeon_fb == NULL) {
+	fb = kzalloc(sizeof(*fb), GFP_KERNEL);
+	if (fb == NULL) {
 		drm_gem_object_put_unlocked(obj);
 		return ERR_PTR(-ENOMEM);
 	}
 
-	ret = radeon_framebuffer_init(dev, radeon_fb, mode_cmd, obj);
+	ret = radeon_framebuffer_init(dev, fb, mode_cmd, obj);
 	if (ret) {
-		kfree(radeon_fb);
+		kfree(fb);
 		drm_gem_object_put_unlocked(obj);
 		return ERR_PTR(ret);
 	}
 
-	return &radeon_fb->base;
+	return fb;
 }
 
 static const struct drm_mode_config_funcs radeon_mode_funcs = {
diff --git a/drivers/gpu/drm/radeon/radeon_fb.c b/drivers/gpu/drm/radeon/radeon_fb.c
index 6cd99f6..11790340 100644
--- a/drivers/gpu/drm/radeon/radeon_fb.c
+++ b/drivers/gpu/drm/radeon/radeon_fb.c
@@ -43,7 +43,7 @@
  */
 struct radeon_fbdev {
 	struct drm_fb_helper helper;
-	struct radeon_framebuffer rfb;
+	struct drm_framebuffer fb;
 	struct radeon_device *rdev;
 };
 
@@ -246,13 +246,13 @@ static int radeonfb_create(struct drm_fb_helper *helper,
 
 	info->par = rfbdev;
 
-	ret = radeon_framebuffer_init(rdev->ddev, &rfbdev->rfb, &mode_cmd, gobj);
+	ret = radeon_framebuffer_init(rdev->ddev, &rfbdev->fb, &mode_cmd, gobj);
 	if (ret) {
 		DRM_ERROR("failed to initialize framebuffer %d\n", ret);
 		goto out;
 	}
 
-	fb = &rfbdev->rfb.base;
+	fb = &rfbdev->fb;
 
 	/* setup helper */
 	rfbdev->helper.fb = fb;
@@ -308,15 +308,15 @@ out:
 
 static int radeon_fbdev_destroy(struct drm_device *dev, struct radeon_fbdev *rfbdev)
 {
-	struct radeon_framebuffer *rfb = &rfbdev->rfb;
+	struct drm_framebuffer *fb = &rfbdev->fb;
 
 	drm_fb_helper_unregister_fbi(&rfbdev->helper);
 
-	if (rfb->base.obj[0]) {
-		radeonfb_destroy_pinned_object(rfb->base.obj[0]);
-		rfb->base.obj[0] = NULL;
-		drm_framebuffer_unregister_private(&rfb->base);
-		drm_framebuffer_cleanup(&rfb->base);
+	if (fb->obj[0]) {
+		radeonfb_destroy_pinned_object(fb->obj[0]);
+		fb->obj[0] = NULL;
+		drm_framebuffer_unregister_private(fb);
+		drm_framebuffer_cleanup(fb);
 	}
 	drm_fb_helper_fini(&rfbdev->helper);
 
@@ -400,7 +400,7 @@ bool radeon_fbdev_robj_is_fb(struct radeon_device *rdev, struct radeon_bo *robj)
 	if (!rdev->mode_info.rfbdev)
 		return false;
 
-	if (robj == gem_to_radeon_bo(rdev->mode_info.rfbdev->rfb.base.obj[0]))
+	if (robj == gem_to_radeon_bo(rdev->mode_info.rfbdev->fb.obj[0]))
 		return true;
 	return false;
 }
diff --git a/drivers/gpu/drm/radeon/radeon_legacy_crtc.c b/drivers/gpu/drm/radeon/radeon_legacy_crtc.c
index 50b3f55..35a205a 100644
--- a/drivers/gpu/drm/radeon/radeon_legacy_crtc.c
+++ b/drivers/gpu/drm/radeon/radeon_legacy_crtc.c
@@ -374,7 +374,6 @@ int radeon_crtc_do_set_base(struct drm_crtc *crtc,
 	struct drm_device *dev = crtc->dev;
 	struct radeon_device *rdev = dev->dev_private;
 	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
-	struct radeon_framebuffer *radeon_fb;
 	struct drm_framebuffer *target_fb;
 	struct drm_gem_object *obj;
 	struct radeon_bo *rbo;
@@ -393,14 +392,10 @@ int radeon_crtc_do_set_base(struct drm_crtc *crtc,
 		return 0;
 	}
 
-	if (atomic) {
-		radeon_fb = to_radeon_framebuffer(fb);
+	if (atomic)
 		target_fb = fb;
-	}
-	else {
-		radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
+	else
 		target_fb = crtc->primary->fb;
-	}
 
 	switch (target_fb->format->cpp[0] * 8) {
 	case 8:
@@ -423,7 +418,7 @@ int radeon_crtc_do_set_base(struct drm_crtc *crtc,
 	}
 
 	/* Pin framebuffer & get tilling informations */
-	obj = radeon_fb->base.obj[0];
+	obj = target_fb->obj[0];
 	rbo = gem_to_radeon_bo(obj);
 retry:
 	r = radeon_bo_reserve(rbo, false);
diff --git a/drivers/gpu/drm/radeon/radeon_mode.h b/drivers/gpu/drm/radeon/radeon_mode.h
index cd93c80..fd470d6 100644
--- a/drivers/gpu/drm/radeon/radeon_mode.h
+++ b/drivers/gpu/drm/radeon/radeon_mode.h
@@ -46,7 +46,6 @@ struct radeon_device;
 #define to_radeon_crtc(x) container_of(x, struct radeon_crtc, base)
 #define to_radeon_connector(x) container_of(x, struct radeon_connector, base)
 #define to_radeon_encoder(x) container_of(x, struct radeon_encoder, base)
-#define to_radeon_framebuffer(x) container_of(x, struct radeon_framebuffer, base)
 
 #define RADEON_MAX_HPD_PINS 7
 #define RADEON_MAX_CRTCS 6
@@ -574,10 +573,6 @@ struct radeon_connector {
 	int enabled_attribs;
 };
 
-struct radeon_framebuffer {
-	struct drm_framebuffer base;
-};
-
 #define ENCODER_MODE_IS_DP(em) (((em) == ATOM_ENCODER_MODE_DP) || \
 				((em) == ATOM_ENCODER_MODE_DP_MST))
 
@@ -931,7 +926,7 @@ radeon_combios_encoder_crtc_scratch_regs(struct drm_encoder *encoder, int crtc);
 extern void
 radeon_combios_encoder_dpms_scratch_regs(struct drm_encoder *encoder, bool on);
 int radeon_framebuffer_init(struct drm_device *dev,
-			     struct radeon_framebuffer *rfb,
+			     struct drm_framebuffer *rfb,
 			     const struct drm_mode_fb_cmd2 *mode_cmd,
 			     struct drm_gem_object *obj);
 
-- 
cgit v1.1


From e68d14dd4ebaf596bf0c237ba82f815c2f561dec Mon Sep 17 00:00:00 2001
From: Daniel Stone <daniels@collabora.com>
Date: Fri, 30 Mar 2018 15:11:38 +0100
Subject: drm/amdgpu: Move GEM BO to drm_framebuffer
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Since drm_framebuffer can now store GEM objects directly, place them
there rather than in our own subclass. As this makes the framebuffer
create_handle and destroy functions the same as the GEM framebuffer
helper, we can reuse those.

Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Daniel Stone <daniels@collabora.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Cc: David (ChunMing) Zhou <David1.Zhou@amd.com>
Cc: amd-gfx@lists.freedesktop.org
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c        |  6 ++--
 drivers/gpu/drm/amd/amdgpu/amdgpu_display.c       | 36 +++++------------------
 drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c            | 10 +++----
 drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h          |  1 -
 drivers/gpu/drm/amd/amdgpu/dce_v10_0.c            | 17 ++++-------
 drivers/gpu/drm/amd/amdgpu/dce_v11_0.c            | 17 ++++-------
 drivers/gpu/drm/amd/amdgpu/dce_v6_0.c             | 17 ++++-------
 drivers/gpu/drm/amd/amdgpu/dce_v8_0.c             | 17 ++++-------
 drivers/gpu/drm/amd/amdgpu/dce_virtual.c          |  4 +--
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 11 +++----
 10 files changed, 40 insertions(+), 96 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index a539265..e0d6b1d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -2541,7 +2541,7 @@ int amdgpu_device_suspend(struct drm_device *dev, bool suspend, bool fbcon)
 	/* unpin the front buffers and cursors */
 	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
 		struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
-		struct amdgpu_framebuffer *rfb = to_amdgpu_framebuffer(crtc->primary->fb);
+		struct drm_framebuffer *fb = crtc->primary->fb;
 		struct amdgpu_bo *robj;
 
 		if (amdgpu_crtc->cursor_bo) {
@@ -2553,10 +2553,10 @@ int amdgpu_device_suspend(struct drm_device *dev, bool suspend, bool fbcon)
 			}
 		}
 
-		if (rfb == NULL || rfb->obj == NULL) {
+		if (fb == NULL || fb->obj[0] == NULL) {
 			continue;
 		}
-		robj = gem_to_amdgpu_bo(rfb->obj);
+		robj = gem_to_amdgpu_bo(fb->obj[0]);
 		/* don't unpin kernel fb objects */
 		if (!amdgpu_fbdev_robj_is_fb(adev, robj)) {
 			r = amdgpu_bo_reserve(robj, true);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c
index 93f700a..b83ae99 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c
@@ -35,6 +35,7 @@
 #include <linux/pm_runtime.h>
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_edid.h>
+#include <drm/drm_gem_framebuffer_helper.h>
 #include <drm/drm_fb_helper.h>
 
 static void amdgpu_display_flip_callback(struct dma_fence *f,
@@ -151,8 +152,6 @@ int amdgpu_display_crtc_page_flip_target(struct drm_crtc *crtc,
 	struct drm_device *dev = crtc->dev;
 	struct amdgpu_device *adev = dev->dev_private;
 	struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
-	struct amdgpu_framebuffer *old_amdgpu_fb;
-	struct amdgpu_framebuffer *new_amdgpu_fb;
 	struct drm_gem_object *obj;
 	struct amdgpu_flip_work *work;
 	struct amdgpu_bo *new_abo;
@@ -174,15 +173,13 @@ int amdgpu_display_crtc_page_flip_target(struct drm_crtc *crtc,
 	work->async = (page_flip_flags & DRM_MODE_PAGE_FLIP_ASYNC) != 0;
 
 	/* schedule unpin of the old buffer */
-	old_amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb);
-	obj = old_amdgpu_fb->obj;
+	obj = crtc->primary->fb->obj[0];
 
 	/* take a reference to the old object */
 	work->old_abo = gem_to_amdgpu_bo(obj);
 	amdgpu_bo_ref(work->old_abo);
 
-	new_amdgpu_fb = to_amdgpu_framebuffer(fb);
-	obj = new_amdgpu_fb->obj;
+	obj = fb->obj[0];
 	new_abo = gem_to_amdgpu_bo(obj);
 
 	/* pin the new buffer */
@@ -482,28 +479,9 @@ bool amdgpu_display_ddc_probe(struct amdgpu_connector *amdgpu_connector,
 	return true;
 }
 
-static void amdgpu_display_user_framebuffer_destroy(struct drm_framebuffer *fb)
-{
-	struct amdgpu_framebuffer *amdgpu_fb = to_amdgpu_framebuffer(fb);
-
-	drm_gem_object_put_unlocked(amdgpu_fb->obj);
-	drm_framebuffer_cleanup(fb);
-	kfree(amdgpu_fb);
-}
-
-static int amdgpu_display_user_framebuffer_create_handle(
-			struct drm_framebuffer *fb,
-			struct drm_file *file_priv,
-			unsigned int *handle)
-{
-	struct amdgpu_framebuffer *amdgpu_fb = to_amdgpu_framebuffer(fb);
-
-	return drm_gem_handle_create(file_priv, amdgpu_fb->obj, handle);
-}
-
 static const struct drm_framebuffer_funcs amdgpu_fb_funcs = {
-	.destroy = amdgpu_display_user_framebuffer_destroy,
-	.create_handle = amdgpu_display_user_framebuffer_create_handle,
+	.destroy = drm_gem_fb_destroy,
+	.create_handle = drm_gem_fb_create_handle,
 };
 
 uint32_t amdgpu_display_framebuffer_domains(struct amdgpu_device *adev)
@@ -526,11 +504,11 @@ int amdgpu_display_framebuffer_init(struct drm_device *dev,
 				    struct drm_gem_object *obj)
 {
 	int ret;
-	rfb->obj = obj;
+	rfb->base.obj[0] = obj;
 	drm_helper_mode_fill_fb_struct(dev, &rfb->base, mode_cmd);
 	ret = drm_framebuffer_init(dev, &rfb->base, &amdgpu_fb_funcs);
 	if (ret) {
-		rfb->obj = NULL;
+		rfb->base.obj[0] = NULL;
 		return ret;
 	}
 	return 0;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c
index 1206301..ff89e84 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c
@@ -292,9 +292,9 @@ static int amdgpu_fbdev_destroy(struct drm_device *dev, struct amdgpu_fbdev *rfb
 
 	drm_fb_helper_unregister_fbi(&rfbdev->helper);
 
-	if (rfb->obj) {
-		amdgpufb_destroy_pinned_object(rfb->obj);
-		rfb->obj = NULL;
+	if (rfb->base.obj[0]) {
+		amdgpufb_destroy_pinned_object(rfb->base.obj[0]);
+		rfb->base.obj[0] = NULL;
 		drm_framebuffer_unregister_private(&rfb->base);
 		drm_framebuffer_cleanup(&rfb->base);
 	}
@@ -377,7 +377,7 @@ int amdgpu_fbdev_total_size(struct amdgpu_device *adev)
 	if (!adev->mode_info.rfbdev)
 		return 0;
 
-	robj = gem_to_amdgpu_bo(adev->mode_info.rfbdev->rfb.obj);
+	robj = gem_to_amdgpu_bo(adev->mode_info.rfbdev->rfb.base.obj[0]);
 	size += amdgpu_bo_size(robj);
 	return size;
 }
@@ -386,7 +386,7 @@ bool amdgpu_fbdev_robj_is_fb(struct amdgpu_device *adev, struct amdgpu_bo *robj)
 {
 	if (!adev->mode_info.rfbdev)
 		return false;
-	if (robj == gem_to_amdgpu_bo(adev->mode_info.rfbdev->rfb.obj))
+	if (robj == gem_to_amdgpu_bo(adev->mode_info.rfbdev->rfb.base.obj[0]))
 		return true;
 	return false;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h
index d6416ee..b9e9e8b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h
@@ -308,7 +308,6 @@ struct amdgpu_display_funcs {
 
 struct amdgpu_framebuffer {
 	struct drm_framebuffer base;
-	struct drm_gem_object *obj;
 
 	/* caching for later use */
 	uint64_t address;
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
index 452f88e..ada241b 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
@@ -1823,7 +1823,6 @@ static int dce_v10_0_crtc_do_set_base(struct drm_crtc *crtc,
 	struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct amdgpu_device *adev = dev->dev_private;
-	struct amdgpu_framebuffer *amdgpu_fb;
 	struct drm_framebuffer *target_fb;
 	struct drm_gem_object *obj;
 	struct amdgpu_bo *abo;
@@ -1842,18 +1841,15 @@ static int dce_v10_0_crtc_do_set_base(struct drm_crtc *crtc,
 		return 0;
 	}
 
-	if (atomic) {
-		amdgpu_fb = to_amdgpu_framebuffer(fb);
+	if (atomic)
 		target_fb = fb;
-	} else {
-		amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb);
+	else
 		target_fb = crtc->primary->fb;
-	}
 
 	/* If atomic, assume fb object is pinned & idle & fenced and
 	 * just update base pointers
 	 */
-	obj = amdgpu_fb->obj;
+	obj = target_fb->obj[0];
 	abo = gem_to_amdgpu_bo(obj);
 	r = amdgpu_bo_reserve(abo, false);
 	if (unlikely(r != 0))
@@ -2043,8 +2039,7 @@ static int dce_v10_0_crtc_do_set_base(struct drm_crtc *crtc,
 	WREG32(mmMASTER_UPDATE_MODE + amdgpu_crtc->crtc_offset, 0);
 
 	if (!atomic && fb && fb != crtc->primary->fb) {
-		amdgpu_fb = to_amdgpu_framebuffer(fb);
-		abo = gem_to_amdgpu_bo(amdgpu_fb->obj);
+		abo = gem_to_amdgpu_bo(fb->obj[0]);
 		r = amdgpu_bo_reserve(abo, true);
 		if (unlikely(r != 0))
 			return r;
@@ -2526,11 +2521,9 @@ static void dce_v10_0_crtc_disable(struct drm_crtc *crtc)
 	dce_v10_0_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
 	if (crtc->primary->fb) {
 		int r;
-		struct amdgpu_framebuffer *amdgpu_fb;
 		struct amdgpu_bo *abo;
 
-		amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb);
-		abo = gem_to_amdgpu_bo(amdgpu_fb->obj);
+		abo = gem_to_amdgpu_bo(crtc->primary->fb->obj[0]);
 		r = amdgpu_bo_reserve(abo, true);
 		if (unlikely(r))
 			DRM_ERROR("failed to reserve abo before unpin\n");
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
index a7c1c58..d3ae508 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
@@ -1862,7 +1862,6 @@ static int dce_v11_0_crtc_do_set_base(struct drm_crtc *crtc,
 	struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct amdgpu_device *adev = dev->dev_private;
-	struct amdgpu_framebuffer *amdgpu_fb;
 	struct drm_framebuffer *target_fb;
 	struct drm_gem_object *obj;
 	struct amdgpu_bo *abo;
@@ -1881,18 +1880,15 @@ static int dce_v11_0_crtc_do_set_base(struct drm_crtc *crtc,
 		return 0;
 	}
 
-	if (atomic) {
-		amdgpu_fb = to_amdgpu_framebuffer(fb);
+	if (atomic)
 		target_fb = fb;
-	} else {
-		amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb);
+	else
 		target_fb = crtc->primary->fb;
-	}
 
 	/* If atomic, assume fb object is pinned & idle & fenced and
 	 * just update base pointers
 	 */
-	obj = amdgpu_fb->obj;
+	obj = target_fb->obj[0];
 	abo = gem_to_amdgpu_bo(obj);
 	r = amdgpu_bo_reserve(abo, false);
 	if (unlikely(r != 0))
@@ -2082,8 +2078,7 @@ static int dce_v11_0_crtc_do_set_base(struct drm_crtc *crtc,
 	WREG32(mmCRTC_MASTER_UPDATE_MODE + amdgpu_crtc->crtc_offset, 0);
 
 	if (!atomic && fb && fb != crtc->primary->fb) {
-		amdgpu_fb = to_amdgpu_framebuffer(fb);
-		abo = gem_to_amdgpu_bo(amdgpu_fb->obj);
+		abo = gem_to_amdgpu_bo(fb->obj[0]);
 		r = amdgpu_bo_reserve(abo, true);
 		if (unlikely(r != 0))
 			return r;
@@ -2601,11 +2596,9 @@ static void dce_v11_0_crtc_disable(struct drm_crtc *crtc)
 	dce_v11_0_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
 	if (crtc->primary->fb) {
 		int r;
-		struct amdgpu_framebuffer *amdgpu_fb;
 		struct amdgpu_bo *abo;
 
-		amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb);
-		abo = gem_to_amdgpu_bo(amdgpu_fb->obj);
+		abo = gem_to_amdgpu_bo(crtc->primary->fb->obj[0]);
 		r = amdgpu_bo_reserve(abo, true);
 		if (unlikely(r))
 			DRM_ERROR("failed to reserve abo before unpin\n");
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c
index 9f67b7f..394cc1e 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c
@@ -1780,7 +1780,6 @@ static int dce_v6_0_crtc_do_set_base(struct drm_crtc *crtc,
 	struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct amdgpu_device *adev = dev->dev_private;
-	struct amdgpu_framebuffer *amdgpu_fb;
 	struct drm_framebuffer *target_fb;
 	struct drm_gem_object *obj;
 	struct amdgpu_bo *abo;
@@ -1798,18 +1797,15 @@ static int dce_v6_0_crtc_do_set_base(struct drm_crtc *crtc,
 		return 0;
 	}
 
-	if (atomic) {
-		amdgpu_fb = to_amdgpu_framebuffer(fb);
+	if (atomic)
 		target_fb = fb;
-	} else {
-		amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb);
+	else
 		target_fb = crtc->primary->fb;
-	}
 
 	/* If atomic, assume fb object is pinned & idle & fenced and
 	 * just update base pointers
 	 */
-	obj = amdgpu_fb->obj;
+	obj = target_fb->obj[0];
 	abo = gem_to_amdgpu_bo(obj);
 	r = amdgpu_bo_reserve(abo, false);
 	if (unlikely(r != 0))
@@ -1978,8 +1974,7 @@ static int dce_v6_0_crtc_do_set_base(struct drm_crtc *crtc,
 	WREG32(mmMASTER_UPDATE_MODE + amdgpu_crtc->crtc_offset, 0);
 
 	if (!atomic && fb && fb != crtc->primary->fb) {
-		amdgpu_fb = to_amdgpu_framebuffer(fb);
-		abo = gem_to_amdgpu_bo(amdgpu_fb->obj);
+		abo = gem_to_amdgpu_bo(fb->obj[0]);
 		r = amdgpu_bo_reserve(abo, true);
 		if (unlikely(r != 0))
 			return r;
@@ -2414,11 +2409,9 @@ static void dce_v6_0_crtc_disable(struct drm_crtc *crtc)
 	dce_v6_0_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
 	if (crtc->primary->fb) {
 		int r;
-		struct amdgpu_framebuffer *amdgpu_fb;
 		struct amdgpu_bo *abo;
 
-		amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb);
-		abo = gem_to_amdgpu_bo(amdgpu_fb->obj);
+		abo = gem_to_amdgpu_bo(crtc->primary->fb->obj[0]);
 		r = amdgpu_bo_reserve(abo, true);
 		if (unlikely(r))
 			DRM_ERROR("failed to reserve abo before unpin\n");
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
index f55422c..c9b9ab8 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
@@ -1754,7 +1754,6 @@ static int dce_v8_0_crtc_do_set_base(struct drm_crtc *crtc,
 	struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct amdgpu_device *adev = dev->dev_private;
-	struct amdgpu_framebuffer *amdgpu_fb;
 	struct drm_framebuffer *target_fb;
 	struct drm_gem_object *obj;
 	struct amdgpu_bo *abo;
@@ -1773,18 +1772,15 @@ static int dce_v8_0_crtc_do_set_base(struct drm_crtc *crtc,
 		return 0;
 	}
 
-	if (atomic) {
-		amdgpu_fb = to_amdgpu_framebuffer(fb);
+	if (atomic)
 		target_fb = fb;
-	} else {
-		amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb);
+	else
 		target_fb = crtc->primary->fb;
-	}
 
 	/* If atomic, assume fb object is pinned & idle & fenced and
 	 * just update base pointers
 	 */
-	obj = amdgpu_fb->obj;
+	obj = target_fb->obj[0];
 	abo = gem_to_amdgpu_bo(obj);
 	r = amdgpu_bo_reserve(abo, false);
 	if (unlikely(r != 0))
@@ -1955,8 +1951,7 @@ static int dce_v8_0_crtc_do_set_base(struct drm_crtc *crtc,
 	WREG32(mmMASTER_UPDATE_MODE + amdgpu_crtc->crtc_offset, 0);
 
 	if (!atomic && fb && fb != crtc->primary->fb) {
-		amdgpu_fb = to_amdgpu_framebuffer(fb);
-		abo = gem_to_amdgpu_bo(amdgpu_fb->obj);
+		abo = gem_to_amdgpu_bo(fb->obj[0]);
 		r = amdgpu_bo_reserve(abo, true);
 		if (unlikely(r != 0))
 			return r;
@@ -2430,11 +2425,9 @@ static void dce_v8_0_crtc_disable(struct drm_crtc *crtc)
 	dce_v8_0_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
 	if (crtc->primary->fb) {
 		int r;
-		struct amdgpu_framebuffer *amdgpu_fb;
 		struct amdgpu_bo *abo;
 
-		amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb);
-		abo = gem_to_amdgpu_bo(amdgpu_fb->obj);
+		abo = gem_to_amdgpu_bo(crtc->primary->fb->obj[0]);
 		r = amdgpu_bo_reserve(abo, true);
 		if (unlikely(r))
 			DRM_ERROR("failed to reserve abo before unpin\n");
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
index b51f05d..89b2286 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
@@ -168,11 +168,9 @@ static void dce_virtual_crtc_disable(struct drm_crtc *crtc)
 	dce_virtual_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
 	if (crtc->primary->fb) {
 		int r;
-		struct amdgpu_framebuffer *amdgpu_fb;
 		struct amdgpu_bo *abo;
 
-		amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb);
-		abo = gem_to_amdgpu_bo(amdgpu_fb->obj);
+		abo = gem_to_amdgpu_bo(crtc->primary->fb->obj[0]);
 		r = amdgpu_bo_reserve(abo, true);
 		if (unlikely(r))
 			DRM_ERROR("failed to reserve abo before unpin\n");
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 3ff3905..077ee67 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -1819,7 +1819,7 @@ static bool fill_rects_from_plane_state(const struct drm_plane_state *state,
 static int get_fb_info(const struct amdgpu_framebuffer *amdgpu_fb,
 		       uint64_t *tiling_flags)
 {
-	struct amdgpu_bo *rbo = gem_to_amdgpu_bo(amdgpu_fb->obj);
+	struct amdgpu_bo *rbo = gem_to_amdgpu_bo(amdgpu_fb->base.obj[0]);
 	int r = amdgpu_bo_reserve(rbo, false);
 
 	if (unlikely(r)) {
@@ -3028,8 +3028,7 @@ static int dm_plane_helper_prepare_fb(struct drm_plane *plane,
 	}
 
 	afb = to_amdgpu_framebuffer(new_state->fb);
-
-	obj = afb->obj;
+	obj = new_state->fb->obj[0];
 	rbo = gem_to_amdgpu_bo(obj);
 	adev = amdgpu_ttm_adev(rbo->tbo.bdev);
 	r = amdgpu_bo_reserve(rbo, false);
@@ -3093,14 +3092,12 @@ static void dm_plane_helper_cleanup_fb(struct drm_plane *plane,
 				       struct drm_plane_state *old_state)
 {
 	struct amdgpu_bo *rbo;
-	struct amdgpu_framebuffer *afb;
 	int r;
 
 	if (!old_state->fb)
 		return;
 
-	afb = to_amdgpu_framebuffer(old_state->fb);
-	rbo = gem_to_amdgpu_bo(afb->obj);
+	rbo = gem_to_amdgpu_bo(old_state->fb->obj[0]);
 	r = amdgpu_bo_reserve(rbo, false);
 	if (unlikely(r)) {
 		DRM_ERROR("failed to reserve rbo before unpin\n");
@@ -3896,7 +3893,7 @@ static void amdgpu_dm_do_flip(struct drm_crtc *crtc,
 	int r, vpos, hpos;
 	struct amdgpu_crtc *acrtc = to_amdgpu_crtc(crtc);
 	struct amdgpu_framebuffer *afb = to_amdgpu_framebuffer(fb);
-	struct amdgpu_bo *abo = gem_to_amdgpu_bo(afb->obj);
+	struct amdgpu_bo *abo = gem_to_amdgpu_bo(fb->obj[0]);
 	struct amdgpu_device *adev = crtc->dev->dev_private;
 	bool async_flip = (crtc->state->pageflip_flags & DRM_MODE_PAGE_FLIP_ASYNC) != 0;
 	struct dc_flip_addrs addr = { {0} };
-- 
cgit v1.1


From 844c541951a00ddffa0248c72e1d7d3e4afaad30 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Mon, 26 Mar 2018 12:56:56 -0500
Subject: drm/amdgpu: add documentation on hwmon interfaces exposed (v3)

Provide detail on the currently exposed hwmon interfaces
for temperature, power, voltage, and fan.

v2: add power cap documentation
v3: add a comment about sensors tool

Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c | 40 ++++++++++++++++++++++++++++++++++
 1 file changed, 40 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
index e6e3658..e5f60fc 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
@@ -1109,6 +1109,46 @@ static ssize_t amdgpu_hwmon_set_power_cap(struct device *dev,
 	return count;
 }
 
+
+/**
+ * DOC: hwmon
+ *
+ * The amdgpu driver exposes the following sensor interfaces:
+ * - GPU temperature (via the on-die sensor)
+ * - GPU voltage
+ * - Northbridge voltage (APUs only)
+ * - GPU power
+ * - GPU fan
+ *
+ * hwmon interfaces for GPU temperature:
+ * - temp1_input: the on die GPU temperature in millidegrees Celsius
+ * - temp1_crit: temperature critical max value in millidegrees Celsius
+ * - temp1_crit_hyst: temperature hysteresis for critical limit in millidegrees Celsius
+ *
+ * hwmon interfaces for GPU voltage:
+ * - in0_input: the voltage on the GPU in millivolts
+ * - in1_input: the voltage on the Northbridge in millivolts
+ *
+ * hwmon interfaces for GPU power:
+ * - power1_average: average power used by the GPU in microWatts
+ * - power1_cap_min: minimum cap supported in microWatts
+ * - power1_cap_max: maximum cap supported in microWatts
+ * - power1_cap: selected power cap in microWatts
+ *
+ * hwmon interfaces for GPU fan:
+ * - pwm1: pulse width modulation fan level (0-255)
+ * - pwm1_enable: pulse width modulation fan control method
+ *                0: no fan speed control
+ *                1: manual fan speed control using pwm interface
+ *                2: automatic fan speed control
+ * - pwm1_min: pulse width modulation fan control minimum level (0)
+ * - pwm1_max: pulse width modulation fan control maximum level (255)
+ * - fan1_input: fan speed in RPM
+ *
+ * You can use hwmon tools like sensors to view this information on your system.
+ *
+ */
+
 static SENSOR_DEVICE_ATTR(temp1_input, S_IRUGO, amdgpu_hwmon_show_temp, NULL, 0);
 static SENSOR_DEVICE_ATTR(temp1_crit, S_IRUGO, amdgpu_hwmon_show_temp_thresh, NULL, 0);
 static SENSOR_DEVICE_ATTR(temp1_crit_hyst, S_IRUGO, amdgpu_hwmon_show_temp_thresh, NULL, 1);
-- 
cgit v1.1


From 6907069004216e630d30847bf2893ab18156ed0f Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Thu, 29 Mar 2018 13:51:28 -0500
Subject: drm/amdgpu: add asic need_full_reset callback
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Allow us to determine at the soc level whether the
asic requires full reset or if soft reset will work.

Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu.h | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index df409dd..21272ce 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -1204,6 +1204,8 @@ struct amdgpu_asic_funcs {
 	/* invalidate hdp read cache */
 	void (*invalidate_hdp)(struct amdgpu_device *adev,
 			       struct amdgpu_ring *ring);
+	/* check if the asic needs a full reset of if soft reset will work */
+	bool (*need_full_reset)(struct amdgpu_device *adev);
 };
 
 /*
@@ -1773,6 +1775,7 @@ amdgpu_get_sdma_instance(struct amdgpu_ring *ring)
 #define amdgpu_asic_get_config_memsize(adev) (adev)->asic_funcs->get_config_memsize((adev))
 #define amdgpu_asic_flush_hdp(adev, r) (adev)->asic_funcs->flush_hdp((adev), (r))
 #define amdgpu_asic_invalidate_hdp(adev, r) (adev)->asic_funcs->invalidate_hdp((adev), (r))
+#define amdgpu_asic_need_full_reset(adev) (adev)->asic_funcs->need_full_reset((adev))
 #define amdgpu_gmc_flush_gpu_tlb(adev, vmid) (adev)->gmc.gmc_funcs->flush_gpu_tlb((adev), (vmid))
 #define amdgpu_gmc_emit_flush_gpu_tlb(r, vmid, addr) (r)->adev->gmc.gmc_funcs->emit_flush_gpu_tlb((r), (vmid), (addr))
 #define amdgpu_gmc_emit_pasid_mapping(r, vmid, pasid) (r)->adev->gmc.gmc_funcs->emit_pasid_mapping((r), (vmid), (pasid))
-- 
cgit v1.1


From 0a881af83cf8d0a9d270f63dd378f4eefda60c48 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Tue, 3 Apr 2018 13:27:14 -0500
Subject: drm/amdgpu/si: implement asic need_full_reset callback
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Used to check on a per SoC basis whether the SoC needs
a full reset of a per IP soft reset.

Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/si.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/si.c b/drivers/gpu/drm/amd/amdgpu/si.c
index a675ec6..c364ef9 100644
--- a/drivers/gpu/drm/amd/amdgpu/si.c
+++ b/drivers/gpu/drm/amd/amdgpu/si.c
@@ -1252,6 +1252,12 @@ static void si_invalidate_hdp(struct amdgpu_device *adev,
 	}
 }
 
+static bool si_need_full_reset(struct amdgpu_device *adev)
+{
+	/* change this when we support soft reset */
+	return true;
+}
+
 static int si_get_pcie_lanes(struct amdgpu_device *adev)
 {
 	u32 link_width_cntl;
@@ -1332,6 +1338,7 @@ static const struct amdgpu_asic_funcs si_asic_funcs =
 	.get_config_memsize = &si_get_config_memsize,
 	.flush_hdp = &si_flush_hdp,
 	.invalidate_hdp = &si_invalidate_hdp,
+	.need_full_reset = &si_need_full_reset,
 };
 
 static uint32_t si_get_rev_id(struct amdgpu_device *adev)
-- 
cgit v1.1


From b7acb46f210b92005b8db16380eac6b3a4c61431 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Thu, 29 Mar 2018 14:39:10 -0500
Subject: drm/amdgpu/cik: implement asic need_full_reset callback
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Used to check on a per SoC basis whether the SoC needs
a full reset of a per IP soft reset.

Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/cik.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/cik.c b/drivers/gpu/drm/amd/amdgpu/cik.c
index 0df2203..8ff4c60 100644
--- a/drivers/gpu/drm/amd/amdgpu/cik.c
+++ b/drivers/gpu/drm/amd/amdgpu/cik.c
@@ -1735,6 +1735,12 @@ static void cik_invalidate_hdp(struct amdgpu_device *adev,
 	}
 }
 
+static bool cik_need_full_reset(struct amdgpu_device *adev)
+{
+	/* change this when we support soft reset */
+	return true;
+}
+
 static const struct amdgpu_asic_funcs cik_asic_funcs =
 {
 	.read_disabled_bios = &cik_read_disabled_bios,
@@ -1748,6 +1754,7 @@ static const struct amdgpu_asic_funcs cik_asic_funcs =
 	.get_config_memsize = &cik_get_config_memsize,
 	.flush_hdp = &cik_flush_hdp,
 	.invalidate_hdp = &cik_invalidate_hdp,
+	.need_full_reset = &cik_need_full_reset,
 };
 
 static int cik_common_early_init(void *handle)
-- 
cgit v1.1


From 06082d9b711fd5889c5f182c6fa629891e5b48c3 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Thu, 29 Mar 2018 14:39:28 -0500
Subject: drm/amdgpu/vi: implement asic need_full_reset callback
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Used to check on a per SoC basis whether the SoC needs
a full reset of a per IP soft reset.

Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/vi.c | 22 ++++++++++++++++++++++
 1 file changed, 22 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c
index 126f127..1b4ee24 100644
--- a/drivers/gpu/drm/amd/amdgpu/vi.c
+++ b/drivers/gpu/drm/amd/amdgpu/vi.c
@@ -876,6 +876,27 @@ static void vi_invalidate_hdp(struct amdgpu_device *adev,
 	}
 }
 
+static bool vi_need_full_reset(struct amdgpu_device *adev)
+{
+	switch (adev->asic_type) {
+	case CHIP_CARRIZO:
+	case CHIP_STONEY:
+		/* CZ has hang issues with full reset at the moment */
+		return false;
+	case CHIP_FIJI:
+	case CHIP_TONGA:
+		/* XXX: soft reset should work on fiji and tonga */
+		return true;
+	case CHIP_POLARIS10:
+	case CHIP_POLARIS11:
+	case CHIP_POLARIS12:
+	case CHIP_TOPAZ:
+	default:
+		/* change this when we support soft reset */
+		return true;
+	}
+}
+
 static const struct amdgpu_asic_funcs vi_asic_funcs =
 {
 	.read_disabled_bios = &vi_read_disabled_bios,
@@ -889,6 +910,7 @@ static const struct amdgpu_asic_funcs vi_asic_funcs =
 	.get_config_memsize = &vi_get_config_memsize,
 	.flush_hdp = &vi_flush_hdp,
 	.invalidate_hdp = &vi_invalidate_hdp,
+	.need_full_reset = &vi_need_full_reset,
 };
 
 #define CZ_REV_BRISTOL(rev)	 \
-- 
cgit v1.1


From adbd4f894f3615f04a4c0cfb931ed647c0280a5f Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Thu, 29 Mar 2018 14:39:46 -0500
Subject: drm/amdgpu/soc15: implement asic need_full_reset callback
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Used to check on a per SoC basis whether the SoC needs
a full reset of a per IP soft reset.

Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/soc15.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index 654b015..2e9ebe8 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -568,6 +568,12 @@ static void soc15_invalidate_hdp(struct amdgpu_device *adev,
 			HDP, 0, mmHDP_READ_CACHE_INVALIDATE), 1);
 }
 
+static bool soc15_need_full_reset(struct amdgpu_device *adev)
+{
+	/* change this when we implement soft reset */
+	return true;
+}
+
 static const struct amdgpu_asic_funcs soc15_asic_funcs =
 {
 	.read_disabled_bios = &soc15_read_disabled_bios,
@@ -581,6 +587,7 @@ static const struct amdgpu_asic_funcs soc15_asic_funcs =
 	.get_config_memsize = &soc15_get_config_memsize,
 	.flush_hdp = &soc15_flush_hdp,
 	.invalidate_hdp = &soc15_invalidate_hdp,
+	.need_full_reset = &soc15_need_full_reset,
 };
 
 static int soc15_common_early_init(void *handle)
-- 
cgit v1.1


From 8bc04c2965879c79bb84e3fc8410e6c90cecc96d Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Thu, 29 Mar 2018 14:48:37 -0500
Subject: drm/amdgpu: use new asic need_full_reset callback
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Use the new callback to determine whether to use full
asic reset or per IP soft reset.  Enables reset to
actually proceed on asics which don't support soft
reset yet.

Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index e0d6b1d..abc3346 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -2738,6 +2738,9 @@ static bool amdgpu_device_ip_check_soft_reset(struct amdgpu_device *adev)
 	if (amdgpu_sriov_vf(adev))
 		return true;
 
+	if (amdgpu_asic_need_full_reset(adev))
+		return true;
+
 	for (i = 0; i < adev->num_ip_blocks; i++) {
 		if (!adev->ip_blocks[i].status.valid)
 			continue;
@@ -2794,6 +2797,9 @@ static bool amdgpu_device_ip_need_full_reset(struct amdgpu_device *adev)
 {
 	int i;
 
+	if (amdgpu_asic_need_full_reset(adev))
+		return true;
+
 	for (i = 0; i < adev->num_ip_blocks; i++) {
 		if (!adev->ip_blocks[i].status.valid)
 			continue;
-- 
cgit v1.1


From e63f86735d9220c8ca6929dc07a4c78f111a6201 Mon Sep 17 00:00:00 2001
From: Colin Ian King <colin.king@canonical.com>
Date: Fri, 30 Mar 2018 17:00:47 +0100
Subject: drm/amd/display: fix spelling mistake: "Usupported" -> "Unsupported"

Trivial fix to spelling mistake in DRM_ERROR error message text

Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 077ee67..fbde450 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -1521,7 +1521,7 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
 		break;
 #endif
 	default:
-		DRM_ERROR("Usupported ASIC type: 0x%X\n", adev->asic_type);
+		DRM_ERROR("Unsupported ASIC type: 0x%X\n", adev->asic_type);
 		goto fail;
 	}
 
@@ -1714,7 +1714,7 @@ static int dm_early_init(void *handle)
 		break;
 #endif
 	default:
-		DRM_ERROR("Usupported ASIC type: 0x%X\n", adev->asic_type);
+		DRM_ERROR("Unsupported ASIC type: 0x%X\n", adev->asic_type);
 		return -EINVAL;
 	}
 
-- 
cgit v1.1


From ced5443502b682decd886ccda10f10862e418ae9 Mon Sep 17 00:00:00 2001
From: Nayan Deshmukh <nayan26deshmukh@gmail.com>
Date: Thu, 29 Mar 2018 22:36:31 +0530
Subject: drm/scheduler: fix param documentation
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

There is no @kernel parameter anymore and document the
@guilty parameter

Signed-off-by: Nayan Deshmukh <nayan26deshmukh@gmail.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/scheduler/gpu_scheduler.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/scheduler/gpu_scheduler.c b/drivers/gpu/drm/scheduler/gpu_scheduler.c
index 0d95888..1d368bc 100644
--- a/drivers/gpu/drm/scheduler/gpu_scheduler.c
+++ b/drivers/gpu/drm/scheduler/gpu_scheduler.c
@@ -117,8 +117,9 @@ drm_sched_rq_select_entity(struct drm_sched_rq *rq)
  * @sched	The pointer to the scheduler
  * @entity	The pointer to a valid drm_sched_entity
  * @rq		The run queue this entity belongs
- * @kernel	If this is an entity for the kernel
  * @jobs	The max number of jobs in the job queue
+ * @guilty      atomic_t set to 1 when a job on this queue
+ *              is found to be guilty causing a timeout
  *
  * return 0 if succeed. negative error code on failure
 */
-- 
cgit v1.1


From a70cdb9eddcfd4ba20d69b84149b4a38648455ac Mon Sep 17 00:00:00 2001
From: Nayan Deshmukh <nayan26deshmukh@gmail.com>
Date: Thu, 29 Mar 2018 22:36:33 +0530
Subject: drm/scheduler: move the tracepoints file from the include directory
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Move it with the scheduler code. This is mostly a straight forward
rename with no code change except for updating the TRACE_INCLUDE_PATH

Signed-off-by: Nayan Deshmukh <nayan26deshmukh@gmail.com>
Suggested-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Acked-by: Lucas Stach <l.stach@pengutronix.de>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/scheduler/gpu_scheduler.c       |  2 +-
 drivers/gpu/drm/scheduler/gpu_scheduler_trace.h | 82 +++++++++++++++++++++++++
 include/drm/gpu_scheduler_trace.h               | 82 -------------------------
 3 files changed, 83 insertions(+), 83 deletions(-)
 create mode 100644 drivers/gpu/drm/scheduler/gpu_scheduler_trace.h
 delete mode 100644 include/drm/gpu_scheduler_trace.h

diff --git a/drivers/gpu/drm/scheduler/gpu_scheduler.c b/drivers/gpu/drm/scheduler/gpu_scheduler.c
index 1d368bc..310275e 100644
--- a/drivers/gpu/drm/scheduler/gpu_scheduler.c
+++ b/drivers/gpu/drm/scheduler/gpu_scheduler.c
@@ -30,7 +30,7 @@
 #include <drm/spsc_queue.h>
 
 #define CREATE_TRACE_POINTS
-#include <drm/gpu_scheduler_trace.h>
+#include "gpu_scheduler_trace.h"
 
 #define to_drm_sched_job(sched_job)		\
 		container_of((sched_job), struct drm_sched_job, queue_node)
diff --git a/drivers/gpu/drm/scheduler/gpu_scheduler_trace.h b/drivers/gpu/drm/scheduler/gpu_scheduler_trace.h
new file mode 100644
index 0000000..4998ad9
--- /dev/null
+++ b/drivers/gpu/drm/scheduler/gpu_scheduler_trace.h
@@ -0,0 +1,82 @@
+/*
+ * Copyright 2017 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#if !defined(_GPU_SCHED_TRACE_H) || defined(TRACE_HEADER_MULTI_READ)
+#define _GPU_SCHED_TRACE_H_
+
+#include <linux/stringify.h>
+#include <linux/types.h>
+#include <linux/tracepoint.h>
+
+#include <drm/drmP.h>
+
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM gpu_scheduler
+#define TRACE_INCLUDE_FILE gpu_scheduler_trace
+
+TRACE_EVENT(drm_sched_job,
+	    TP_PROTO(struct drm_sched_job *sched_job, struct drm_sched_entity *entity),
+	    TP_ARGS(sched_job, entity),
+	    TP_STRUCT__entry(
+			     __field(struct drm_sched_entity *, entity)
+			     __field(struct dma_fence *, fence)
+			     __field(const char *, name)
+			     __field(uint64_t, id)
+			     __field(u32, job_count)
+			     __field(int, hw_job_count)
+			     ),
+
+	    TP_fast_assign(
+			   __entry->entity = entity;
+			   __entry->id = sched_job->id;
+			   __entry->fence = &sched_job->s_fence->finished;
+			   __entry->name = sched_job->sched->name;
+			   __entry->job_count = spsc_queue_count(&entity->job_queue);
+			   __entry->hw_job_count = atomic_read(
+				   &sched_job->sched->hw_rq_count);
+			   ),
+	    TP_printk("entity=%p, id=%llu, fence=%p, ring=%s, job count:%u, hw job count:%d",
+		      __entry->entity, __entry->id,
+		      __entry->fence, __entry->name,
+		      __entry->job_count, __entry->hw_job_count)
+);
+
+TRACE_EVENT(drm_sched_process_job,
+	    TP_PROTO(struct drm_sched_fence *fence),
+	    TP_ARGS(fence),
+	    TP_STRUCT__entry(
+		    __field(struct dma_fence *, fence)
+		    ),
+
+	    TP_fast_assign(
+		    __entry->fence = &fence->finished;
+		    ),
+	    TP_printk("fence=%p signaled", __entry->fence)
+);
+
+#endif
+
+/* This part must be outside protection */
+#undef TRACE_INCLUDE_PATH
+#define TRACE_INCLUDE_PATH ../../drivers/gpu/drm/scheduler
+#include <trace/define_trace.h>
diff --git a/include/drm/gpu_scheduler_trace.h b/include/drm/gpu_scheduler_trace.h
deleted file mode 100644
index 0789e8d..0000000
--- a/include/drm/gpu_scheduler_trace.h
+++ /dev/null
@@ -1,82 +0,0 @@
-/*
- * Copyright 2017 Advanced Micro Devices, Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- */
-
-#if !defined(_GPU_SCHED_TRACE_H) || defined(TRACE_HEADER_MULTI_READ)
-#define _GPU_SCHED_TRACE_H_
-
-#include <linux/stringify.h>
-#include <linux/types.h>
-#include <linux/tracepoint.h>
-
-#include <drm/drmP.h>
-
-#undef TRACE_SYSTEM
-#define TRACE_SYSTEM gpu_scheduler
-#define TRACE_INCLUDE_FILE gpu_scheduler_trace
-
-TRACE_EVENT(drm_sched_job,
-	    TP_PROTO(struct drm_sched_job *sched_job, struct drm_sched_entity *entity),
-	    TP_ARGS(sched_job, entity),
-	    TP_STRUCT__entry(
-			     __field(struct drm_sched_entity *, entity)
-			     __field(struct dma_fence *, fence)
-			     __field(const char *, name)
-			     __field(uint64_t, id)
-			     __field(u32, job_count)
-			     __field(int, hw_job_count)
-			     ),
-
-	    TP_fast_assign(
-			   __entry->entity = entity;
-			   __entry->id = sched_job->id;
-			   __entry->fence = &sched_job->s_fence->finished;
-			   __entry->name = sched_job->sched->name;
-			   __entry->job_count = spsc_queue_count(&entity->job_queue);
-			   __entry->hw_job_count = atomic_read(
-				   &sched_job->sched->hw_rq_count);
-			   ),
-	    TP_printk("entity=%p, id=%llu, fence=%p, ring=%s, job count:%u, hw job count:%d",
-		      __entry->entity, __entry->id,
-		      __entry->fence, __entry->name,
-		      __entry->job_count, __entry->hw_job_count)
-);
-
-TRACE_EVENT(drm_sched_process_job,
-	    TP_PROTO(struct drm_sched_fence *fence),
-	    TP_ARGS(fence),
-	    TP_STRUCT__entry(
-		    __field(struct dma_fence *, fence)
-		    ),
-
-	    TP_fast_assign(
-		    __entry->fence = &fence->finished;
-		    ),
-	    TP_printk("fence=%p signaled", __entry->fence)
-);
-
-#endif
-
-/* This part must be outside protection */
-#undef TRACE_INCLUDE_PATH
-#define TRACE_INCLUDE_PATH .
-#include <trace/define_trace.h>
-- 
cgit v1.1


From 5a8c102ac471c53da38b2c3c35417e9355d21215 Mon Sep 17 00:00:00 2001
From: Chunming Zhou <david1.zhou@amd.com>
Date: Fri, 16 Mar 2018 12:29:38 +0800
Subject: drm/amdgpu: Don't change preferred domian when fallback GTT v6
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

v2: add sanity checking
v3: make code open
v4: also handle visible to invisible fallback
v5: Since two fallback cases, re-use goto retry
v6: avoid bo is unref when retry, and only user BO can fallback

Signed-off-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com> (v5)
Acked-by: Felix Kuehling <Felix.Kuehling@amd.com>
Cc: felix.kuehling@amd.com
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c    | 16 ++-----------
 drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 37 +++++++++++++++++++-----------
 2 files changed, 26 insertions(+), 27 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
index 46b9ea4..28c2706 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
@@ -56,23 +56,11 @@ int amdgpu_gem_object_create(struct amdgpu_device *adev, unsigned long size,
 		alignment = PAGE_SIZE;
 	}
 
-retry:
 	r = amdgpu_bo_create(adev, size, alignment, initial_domain,
 			     flags, type, resv, &bo);
 	if (r) {
-		if (r != -ERESTARTSYS) {
-			if (flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED) {
-				flags &= ~AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED;
-				goto retry;
-			}
-
-			if (initial_domain == AMDGPU_GEM_DOMAIN_VRAM) {
-				initial_domain |= AMDGPU_GEM_DOMAIN_GTT;
-				goto retry;
-			}
-			DRM_DEBUG("Failed to allocate GEM object (%ld, %d, %u, %d)\n",
-				  size, initial_domain, alignment, r);
-		}
+		DRM_DEBUG("Failed to allocate GEM object (%ld, %d, %u, %d)\n",
+			  size, initial_domain, alignment, r);
 		return r;
 	}
 	*obj = &bo->gem_base;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
index 6d08cde..04d6830 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
@@ -356,6 +356,7 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev, unsigned long size,
 	struct amdgpu_bo *bo;
 	unsigned long page_align;
 	size_t acc_size;
+	u32 domains, preferred_domains, allowed_domains;
 	int r;
 
 	page_align = roundup(byte_align, PAGE_SIZE) >> PAGE_SHIFT;
@@ -369,22 +370,24 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev, unsigned long size,
 	acc_size = ttm_bo_dma_acc_size(&adev->mman.bdev, size,
 				       sizeof(struct amdgpu_bo));
 
+	preferred_domains = domain & (AMDGPU_GEM_DOMAIN_VRAM |
+				      AMDGPU_GEM_DOMAIN_GTT |
+				      AMDGPU_GEM_DOMAIN_CPU |
+				      AMDGPU_GEM_DOMAIN_GDS |
+				      AMDGPU_GEM_DOMAIN_GWS |
+				      AMDGPU_GEM_DOMAIN_OA);
+	allowed_domains = preferred_domains;
+	if (type != ttm_bo_type_kernel &&
+	    allowed_domains == AMDGPU_GEM_DOMAIN_VRAM)
+		allowed_domains |= AMDGPU_GEM_DOMAIN_GTT;
+	domains = preferred_domains;
+retry:
 	bo = kzalloc(sizeof(struct amdgpu_bo), GFP_KERNEL);
 	if (bo == NULL)
 		return -ENOMEM;
 	drm_gem_private_object_init(adev->ddev, &bo->gem_base, size);
 	INIT_LIST_HEAD(&bo->shadow_list);
 	INIT_LIST_HEAD(&bo->va);
-	bo->preferred_domains = domain & (AMDGPU_GEM_DOMAIN_VRAM |
-					 AMDGPU_GEM_DOMAIN_GTT |
-					 AMDGPU_GEM_DOMAIN_CPU |
-					 AMDGPU_GEM_DOMAIN_GDS |
-					 AMDGPU_GEM_DOMAIN_GWS |
-					 AMDGPU_GEM_DOMAIN_OA);
-	bo->allowed_domains = bo->preferred_domains;
-	if (type != ttm_bo_type_kernel &&
-	    bo->allowed_domains == AMDGPU_GEM_DOMAIN_VRAM)
-		bo->allowed_domains |= AMDGPU_GEM_DOMAIN_GTT;
 
 	bo->flags = flags;
 
@@ -417,12 +420,20 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev, unsigned long size,
 #endif
 
 	bo->tbo.bdev = &adev->mman.bdev;
-	amdgpu_ttm_placement_from_domain(bo, domain);
-
+	amdgpu_ttm_placement_from_domain(bo, domains);
 	r = ttm_bo_init_reserved(&adev->mman.bdev, &bo->tbo, size, type,
 				 &bo->placement, page_align, &ctx, acc_size,
 				 NULL, resv, &amdgpu_ttm_bo_destroy);
-	if (unlikely(r != 0))
+	if (unlikely(r && r != -ERESTARTSYS) && type == ttm_bo_type_device) {
+		if (flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED) {
+			flags &= ~AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED;
+			goto retry;
+		} else if (domains != allowed_domains) {
+			domains = allowed_domains;
+			goto retry;
+		}
+	}
+	if (unlikely(r))
 		return r;
 
 	if (adev->gmc.visible_vram_size < adev->gmc.real_vram_size &&
-- 
cgit v1.1


From 552825b28ddac200b6080d9e79f4121b68e1517d Mon Sep 17 00:00:00 2001
From: Chunming Zhou <david1.zhou@amd.com>
Date: Mon, 2 Apr 2018 11:20:44 +0800
Subject: drm/amdgpu: add new bo flag that indicates BOs don't need fallback
 (v2)

user cases:
1. KFD wraps amdgpu_bo_create, they have no fallback case which is different
with amdgpu_gem_object_create.
since upstream branch has no amdgpu_amdkfd_gpuvm.c, which need KFD
guys add this flag to __alloc_memory_of_gpu:
+       flags |= AMDGPU_GEM_CREATE_NO_FALLBACK;
2. UMD can specify this flag for their allocation as well if they like.

v2: squash in merge conflict fix (Chunming)

Signed-off-by: Chunming Zhou <david1.zhou@amd.com>
Acked-by: Felix Kuehling <Felix.Kuehling@amd.com>
Cc: felix.kuehling@amd.com
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c     | 3 ++-
 drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 5 ++++-
 include/uapi/drm/amdgpu_drm.h              | 2 ++
 3 files changed, 8 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
index dc34b50..d7d7ce1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
@@ -386,7 +386,8 @@ retry:
 	    bo->tbo.mem.start < adev->gmc.visible_vram_size >> PAGE_SHIFT)
 		p->bytes_moved_vis += ctx.bytes_moved;
 
-	if (unlikely(r == -ENOMEM) && domain != bo->allowed_domains) {
+	if (unlikely(r == -ENOMEM) && domain != bo->allowed_domains &&
+	    !(bo->flags & AMDGPU_GEM_CREATE_NO_FALLBACK)) {
 		domain = bo->allowed_domains;
 		goto retry;
 	}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
index 04d6830..9e23d6f 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
@@ -388,6 +388,8 @@ retry:
 	drm_gem_private_object_init(adev->ddev, &bo->gem_base, size);
 	INIT_LIST_HEAD(&bo->shadow_list);
 	INIT_LIST_HEAD(&bo->va);
+	bo->preferred_domains = preferred_domains;
+	bo->allowed_domains = allowed_domains;
 
 	bo->flags = flags;
 
@@ -424,7 +426,8 @@ retry:
 	r = ttm_bo_init_reserved(&adev->mman.bdev, &bo->tbo, size, type,
 				 &bo->placement, page_align, &ctx, acc_size,
 				 NULL, resv, &amdgpu_ttm_bo_destroy);
-	if (unlikely(r && r != -ERESTARTSYS) && type == ttm_bo_type_device) {
+	if (unlikely(r && r != -ERESTARTSYS) && type == ttm_bo_type_device &&
+	    !(flags & AMDGPU_GEM_CREATE_NO_FALLBACK)) {
 		if (flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED) {
 			flags &= ~AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED;
 			goto retry;
diff --git a/include/uapi/drm/amdgpu_drm.h b/include/uapi/drm/amdgpu_drm.h
index c363b67..4f5a27d 100644
--- a/include/uapi/drm/amdgpu_drm.h
+++ b/include/uapi/drm/amdgpu_drm.h
@@ -95,6 +95,8 @@ extern "C" {
 #define AMDGPU_GEM_CREATE_VM_ALWAYS_VALID	(1 << 6)
 /* Flag that BO sharing will be explicitly synchronized */
 #define AMDGPU_GEM_CREATE_EXPLICIT_SYNC		(1 << 7)
+/* Flag that BO doesn't need fallback */
+#define AMDGPU_GEM_CREATE_NO_FALLBACK		(1 << 8)
 
 struct drm_amdgpu_gem_create_in  {
 	/** the requested memory size */
-- 
cgit v1.1


From 1a61ee07211c543bf43e635fa703c162a78af0e1 Mon Sep 17 00:00:00 2001
From: Eric Anholt <eric@anholt.net>
Date: Wed, 4 Apr 2018 15:32:51 -0700
Subject: drm/sched: Extend the documentation.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

These comments answer all the questions I had for myself when
implementing a driver using the GPU scheduler.

Signed-off-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 include/drm/gpu_scheduler.h | 46 +++++++++++++++++++++++++++++++++++++++++----
 1 file changed, 42 insertions(+), 4 deletions(-)

diff --git a/include/drm/gpu_scheduler.h b/include/drm/gpu_scheduler.h
index dfd54fb..c053a32 100644
--- a/include/drm/gpu_scheduler.h
+++ b/include/drm/gpu_scheduler.h
@@ -43,10 +43,12 @@ enum drm_sched_priority {
 };
 
 /**
- * A scheduler entity is a wrapper around a job queue or a group
- * of other entities. Entities take turns emitting jobs from their
- * job queues to corresponding hardware ring based on scheduling
- * policy.
+ * drm_sched_entity - A wrapper around a job queue (typically attached
+ * to the DRM file_priv).
+ *
+ * Entities will emit jobs in order to their corresponding hardware
+ * ring, and the scheduler will alternate between entities based on
+ * scheduling policy.
 */
 struct drm_sched_entity {
 	struct list_head		list;
@@ -78,7 +80,18 @@ struct drm_sched_rq {
 
 struct drm_sched_fence {
 	struct dma_fence		scheduled;
+
+	/* This fence is what will be signaled by the scheduler when
+	 * the job is completed.
+	 *
+	 * When setting up an out fence for the job, you should use
+	 * this, since it's available immediately upon
+	 * drm_sched_job_init(), and the fence returned by the driver
+	 * from run_job() won't be created until the dependencies have
+	 * resolved.
+	 */
 	struct dma_fence		finished;
+
 	struct dma_fence_cb		cb;
 	struct dma_fence		*parent;
 	struct drm_gpu_scheduler	*sched;
@@ -88,6 +101,13 @@ struct drm_sched_fence {
 
 struct drm_sched_fence *to_drm_sched_fence(struct dma_fence *f);
 
+/**
+ * drm_sched_job - A job to be run by an entity.
+ *
+ * A job is created by the driver using drm_sched_job_init(), and
+ * should call drm_sched_entity_push_job() once it wants the scheduler
+ * to schedule the job.
+ */
 struct drm_sched_job {
 	struct spsc_node		queue_node;
 	struct drm_gpu_scheduler	*sched;
@@ -112,10 +132,28 @@ static inline bool drm_sched_invalidate_job(struct drm_sched_job *s_job,
  * these functions should be implemented in driver side
 */
 struct drm_sched_backend_ops {
+	/* Called when the scheduler is considering scheduling this
+	 * job next, to get another struct dma_fence for this job to
+	 * block on.  Once it returns NULL, run_job() may be called.
+	 */
 	struct dma_fence *(*dependency)(struct drm_sched_job *sched_job,
 					struct drm_sched_entity *s_entity);
+
+	/* Called to execute the job once all of the dependencies have
+	 * been resolved.  This may be called multiple times, if
+	 * timedout_job() has happened and drm_sched_job_recovery()
+	 * decides to try it again.
+	 */
 	struct dma_fence *(*run_job)(struct drm_sched_job *sched_job);
+
+	/* Called when a job has taken too long to execute, to trigger
+	 * GPU recovery.
+	 */
 	void (*timedout_job)(struct drm_sched_job *sched_job);
+
+	/* Called once the job's finished fence has been signaled and
+	 * it's time to clean it up.
+	 */
 	void (*free_job)(struct drm_sched_job *sched_job);
 };
 
-- 
cgit v1.1


From 1eb1547fd0267fbb5fabe4973210dfc295a92725 Mon Sep 17 00:00:00 2001
From: James Zhu <James.Zhu@amd.com>
Date: Tue, 3 Apr 2018 10:41:32 -0400
Subject: drm/amdgpu: Added support for MV packet
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Motion vector packet needs support in physical mode.

Signed-off-by: James Zhu <James.Zhu@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c | 24 ++++++++++++++++++++++++
 1 file changed, 24 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
index a33804b..d7261e0 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
@@ -755,6 +755,18 @@ int amdgpu_vce_ring_parse_cs(struct amdgpu_cs_parser *p, uint32_t ib_idx)
 			if (r)
 				goto out;
 			break;
+
+		case 0x0500000d: /* MV buffer */
+			r = amdgpu_vce_validate_bo(p, ib_idx, idx + 3,
+							idx + 2, 0, 0);
+			if (r)
+				goto out;
+
+			r = amdgpu_vce_validate_bo(p, ib_idx, idx + 8,
+							idx + 7, 0, 0);
+			if (r)
+				goto out;
+			break;
 		}
 
 		idx += len / 4;
@@ -860,6 +872,18 @@ int amdgpu_vce_ring_parse_cs(struct amdgpu_cs_parser *p, uint32_t ib_idx)
 				goto out;
 			break;
 
+		case 0x0500000d: /* MV buffer */
+			r = amdgpu_vce_cs_reloc(p, ib_idx, idx + 3,
+							idx + 2, *size, 0);
+			if (r)
+				goto out;
+
+			r = amdgpu_vce_cs_reloc(p, ib_idx, idx + 8,
+							idx + 7, *size / 12, 0);
+			if (r)
+				goto out;
+			break;
+
 		default:
 			DRM_ERROR("invalid VCE command (0x%x)!\n", cmd);
 			r = -EINVAL;
-- 
cgit v1.1


From 8218d7f1f70179a532639f01dfd32dc5dbb09ed3 Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Tue, 17 Oct 2017 12:02:01 -0400
Subject: drm/amd/display: Don't access legacy properties

We're an atomic driver and shouldn't access legacy properties. Doing so
will only scare users with stack traces.

Instead save the prop in the state and access it directly. Much simpler.

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 13 ++++++++++---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h |  1 +
 2 files changed, 11 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index fbde450..7483947 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -5118,17 +5118,24 @@ void amdgpu_dm_add_sink_to_freesync_module(struct drm_connector *connector,
 					   struct edid *edid)
 {
 	int i;
-	uint64_t val_capable;
 	bool edid_check_required;
 	struct detailed_timing *timing;
 	struct detailed_non_pixel *data;
 	struct detailed_data_monitor_range *range;
 	struct amdgpu_dm_connector *amdgpu_dm_connector =
 			to_amdgpu_dm_connector(connector);
+	struct dm_connector_state *dm_con_state;
 
 	struct drm_device *dev = connector->dev;
 	struct amdgpu_device *adev = dev->dev_private;
 
+	if (!connector->state) {
+		DRM_ERROR("%s - Connector has no state", __func__);
+		return;
+	}
+
+	dm_con_state = to_dm_connector_state(connector->state);
+
 	edid_check_required = false;
 	if (!amdgpu_dm_connector->dc_sink) {
 		DRM_ERROR("dc_sink NULL, could not add free_sync module.\n");
@@ -5147,7 +5154,7 @@ void amdgpu_dm_add_sink_to_freesync_module(struct drm_connector *connector,
 						amdgpu_dm_connector);
 		}
 	}
-	val_capable = 0;
+	dm_con_state->freesync_capable = false;
 	if (edid_check_required == true && (edid->version > 1 ||
 	   (edid->version == 1 && edid->revision > 1))) {
 		for (i = 0; i < 4; i++) {
@@ -5183,7 +5190,7 @@ void amdgpu_dm_add_sink_to_freesync_module(struct drm_connector *connector,
 					amdgpu_dm_connector->min_vfreq * 1000000;
 			amdgpu_dm_connector->caps.max_refresh_in_micro_hz =
 					amdgpu_dm_connector->max_vfreq * 1000000;
-				val_capable = 1;
+			dm_con_state->freesync_capable = true;
 		}
 	}
 
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
index 3af699b..005cf0d 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
@@ -220,6 +220,7 @@ struct dm_connector_state {
 	uint8_t underscan_hborder;
 	bool underscan_enable;
 	struct mod_freesync_user_enable user_enable;
+	bool freesync_capable;
 };
 
 #define to_dm_connector_state(x)\
-- 
cgit v1.1


From 742811b7121ec4e426edb4a21657ca5523955489 Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Mon, 12 Mar 2018 11:16:47 -0400
Subject: drm/amd/display: Only register backlight device if embedded panel
 connected

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 58 +++++++++++++----------
 1 file changed, 33 insertions(+), 25 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 7483947..6636f4e 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -1403,6 +1403,28 @@ static int initialize_plane(struct amdgpu_display_manager *dm,
 	return ret;
 }
 
+
+static void register_backlight_device(struct amdgpu_display_manager *dm,
+				      struct dc_link *link)
+{
+#if defined(CONFIG_BACKLIGHT_CLASS_DEVICE) ||\
+	defined(CONFIG_BACKLIGHT_CLASS_DEVICE_MODULE)
+
+	if ((link->connector_signal & (SIGNAL_TYPE_EDP | SIGNAL_TYPE_LVDS)) &&
+	    link->type != dc_connection_none) {
+		/* Event if registration failed, we should continue with
+		 * DM initialization because not having a backlight control
+		 * is better then a black screen.
+		 */
+		amdgpu_dm_register_backlight_device(dm);
+
+		if (dm->backlight_dev)
+			dm->backlight_link = link;
+	}
+#endif
+}
+
+
 /* In this architecture, the association
  * connector -> encoder -> crtc
  * id not really requried. The crtc and connector will hold the
@@ -1456,6 +1478,7 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
 
 	/* loops over all connectors on the board */
 	for (i = 0; i < link_cnt; i++) {
+		struct dc_link *link = NULL;
 
 		if (i > AMDGPU_DM_MAX_DISPLAY_INDEX) {
 			DRM_ERROR(
@@ -1482,9 +1505,14 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
 			goto fail;
 		}
 
-		if (dc_link_detect(dc_get_link_at_index(dm->dc, i),
-				DETECT_REASON_BOOT))
+		link = dc_get_link_at_index(dm->dc, i);
+
+		if (dc_link_detect(link, DETECT_REASON_BOOT)) {
 			amdgpu_dm_update_connector_after_detect(aconnector);
+			register_backlight_device(dm, link);
+		}
+
+
 	}
 
 	/* Software is initialized. Now we can register interrupt handlers. */
@@ -2684,7 +2712,8 @@ static void amdgpu_dm_connector_destroy(struct drm_connector *connector)
 #if defined(CONFIG_BACKLIGHT_CLASS_DEVICE) ||\
 	defined(CONFIG_BACKLIGHT_CLASS_DEVICE_MODULE)
 
-	if (link->connector_signal & (SIGNAL_TYPE_EDP | SIGNAL_TYPE_LVDS)) {
+	if ((link->connector_signal & (SIGNAL_TYPE_EDP | SIGNAL_TYPE_LVDS)) &&
+	    link->type != dc_connection_none) {
 		amdgpu_dm_register_backlight_device(dm);
 
 		if (dm->backlight_dev) {
@@ -3557,6 +3586,7 @@ create_i2c(struct ddc_service *ddc_service,
 	return i2c;
 }
 
+
 /* Note: this function assumes that dc_link_detect() was called for the
  * dc_link which will be represented by this aconnector.
  */
@@ -3626,28 +3656,6 @@ static int amdgpu_dm_connector_init(struct amdgpu_display_manager *dm,
 		|| connector_type == DRM_MODE_CONNECTOR_eDP)
 		amdgpu_dm_initialize_dp_connector(dm, aconnector);
 
-#if defined(CONFIG_BACKLIGHT_CLASS_DEVICE) ||\
-	defined(CONFIG_BACKLIGHT_CLASS_DEVICE_MODULE)
-
-	/* NOTE: this currently will create backlight device even if a panel
-	 * is not connected to the eDP/LVDS connector.
-	 *
-	 * This is less than ideal but we don't have sink information at this
-	 * stage since detection happens after. We can't do detection earlier
-	 * since MST detection needs connectors to be created first.
-	 */
-	if (link->connector_signal & (SIGNAL_TYPE_EDP | SIGNAL_TYPE_LVDS)) {
-		/* Event if registration failed, we should continue with
-		 * DM initialization because not having a backlight control
-		 * is better then a black screen.
-		 */
-		amdgpu_dm_register_backlight_device(dm);
-
-		if (dm->backlight_dev)
-			dm->backlight_link = link;
-	}
-#endif
-
 out_free:
 	if (res) {
 		kfree(i2c);
-- 
cgit v1.1


From 5cd29ed0cb9a1985d3334d52d3b54f0defb23de4 Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Mon, 12 Mar 2018 11:48:26 -0400
Subject: drm/amd/display: Don't register backlight on connector_destroy

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 12 ++++--------
 1 file changed, 4 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 6636f4e..7ecc224 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -2713,14 +2713,10 @@ static void amdgpu_dm_connector_destroy(struct drm_connector *connector)
 	defined(CONFIG_BACKLIGHT_CLASS_DEVICE_MODULE)
 
 	if ((link->connector_signal & (SIGNAL_TYPE_EDP | SIGNAL_TYPE_LVDS)) &&
-	    link->type != dc_connection_none) {
-		amdgpu_dm_register_backlight_device(dm);
-
-		if (dm->backlight_dev) {
-			backlight_device_unregister(dm->backlight_dev);
-			dm->backlight_dev = NULL;
-		}
-
+	    link->type != dc_connection_none &&
+	    dm->backlight_dev) {
+		backlight_device_unregister(dm->backlight_dev);
+		dm->backlight_dev = NULL;
 	}
 #endif
 	drm_connector_unregister(connector);
-- 
cgit v1.1


From 0c8df4bbc4de4789dde7fa622585803fd10dd8e4 Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Thu, 15 Mar 2018 13:46:50 -0400
Subject: drm/amd/display: Program v_total_min/max after v_total_cntl

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../drm/amd/display/dc/dce110/dce110_timing_generator.c  | 16 ++++++++--------
 .../drm/amd/display/dc/dce120/dce120_timing_generator.c  | 12 ++++++------
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c        | 12 ++++++------
 3 files changed, 20 insertions(+), 20 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.c
index be71539..1b2fe0d 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.c
@@ -431,14 +431,6 @@ void dce110_timing_generator_set_drr(
 			0,
 			CRTC_V_TOTAL_CONTROL,
 			CRTC_SET_V_TOTAL_MIN_MASK);
-		set_reg_field_value(v_total_min,
-				0,
-				CRTC_V_TOTAL_MIN,
-				CRTC_V_TOTAL_MIN);
-		set_reg_field_value(v_total_max,
-				0,
-				CRTC_V_TOTAL_MAX,
-				CRTC_V_TOTAL_MAX);
 		set_reg_field_value(v_total_cntl,
 				0,
 				CRTC_V_TOTAL_CONTROL,
@@ -447,6 +439,14 @@ void dce110_timing_generator_set_drr(
 				0,
 				CRTC_V_TOTAL_CONTROL,
 				CRTC_V_TOTAL_MAX_SEL);
+		set_reg_field_value(v_total_min,
+				0,
+				CRTC_V_TOTAL_MIN,
+				CRTC_V_TOTAL_MIN);
+		set_reg_field_value(v_total_max,
+				0,
+				CRTC_V_TOTAL_MAX,
+				CRTC_V_TOTAL_MAX);
 		set_reg_field_value(v_total_cntl,
 				0,
 				CRTC_V_TOTAL_CONTROL,
diff --git a/drivers/gpu/drm/amd/display/dc/dce120/dce120_timing_generator.c b/drivers/gpu/drm/amd/display/dc/dce120/dce120_timing_generator.c
index 7bee781..2ea490f 100644
--- a/drivers/gpu/drm/amd/display/dc/dce120/dce120_timing_generator.c
+++ b/drivers/gpu/drm/amd/display/dc/dce120/dce120_timing_generator.c
@@ -570,12 +570,6 @@ void dce120_timing_generator_set_drr(
 				0x180);
 
 	} else {
-		CRTC_REG_UPDATE(
-				CRTC0_CRTC_V_TOTAL_MIN,
-				CRTC_V_TOTAL_MIN, 0);
-		CRTC_REG_UPDATE(
-				CRTC0_CRTC_V_TOTAL_MAX,
-				CRTC_V_TOTAL_MAX, 0);
 		CRTC_REG_SET_N(CRTC0_CRTC_V_TOTAL_CONTROL, 5,
 				FD(CRTC0_CRTC_V_TOTAL_CONTROL__CRTC_V_TOTAL_MIN_SEL), 0,
 				FD(CRTC0_CRTC_V_TOTAL_CONTROL__CRTC_V_TOTAL_MAX_SEL), 0,
@@ -583,6 +577,12 @@ void dce120_timing_generator_set_drr(
 				FD(CRTC0_CRTC_V_TOTAL_CONTROL__CRTC_FORCE_LOCK_TO_MASTER_VSYNC), 0,
 				FD(CRTC0_CRTC_V_TOTAL_CONTROL__CRTC_SET_V_TOTAL_MIN_MASK), 0);
 		CRTC_REG_UPDATE(
+				CRTC0_CRTC_V_TOTAL_MIN,
+				CRTC_V_TOTAL_MIN, 0);
+		CRTC_REG_UPDATE(
+				CRTC0_CRTC_V_TOTAL_MAX,
+				CRTC_V_TOTAL_MAX, 0);
+		CRTC_REG_UPDATE(
 				CRTC0_CRTC_STATIC_SCREEN_CONTROL,
 				CRTC_STATIC_SCREEN_EVENT_MASK,
 				0);
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c
index f56eac0..dc92130 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c
@@ -855,17 +855,17 @@ void optc1_set_drr(
 				OTG_SET_V_TOTAL_MIN_MASK_EN, 0,
 				OTG_SET_V_TOTAL_MIN_MASK, 0);
 	} else {
-		REG_SET(OTG_V_TOTAL_MIN, 0,
-			OTG_V_TOTAL_MIN, 0);
-
-		REG_SET(OTG_V_TOTAL_MAX, 0,
-			OTG_V_TOTAL_MAX, 0);
-
 		REG_UPDATE_4(OTG_V_TOTAL_CONTROL,
 				OTG_SET_V_TOTAL_MIN_MASK, 0,
 				OTG_V_TOTAL_MIN_SEL, 0,
 				OTG_V_TOTAL_MAX_SEL, 0,
 				OTG_FORCE_LOCK_ON_EVENT, 0);
+
+		REG_SET(OTG_V_TOTAL_MIN, 0,
+			OTG_V_TOTAL_MIN, 0);
+
+		REG_SET(OTG_V_TOTAL_MAX, 0,
+			OTG_V_TOTAL_MAX, 0);
 	}
 }
 
-- 
cgit v1.1


From 9e3efe3eed47952e2e0209b3808989ae1cc6a31b Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Thu, 15 Mar 2018 15:08:04 -0400
Subject: drm/amd/display: Set ignore_msa_timing_param

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 7ecc224..a6039e5 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -2487,6 +2487,9 @@ create_stream_for_sink(struct amdgpu_dm_connector *aconnector,
 
 	update_stream_signal(stream);
 
+	if (dm_state && dm_state->freesync_capable)
+		stream->ignore_msa_timing_param = true;
+
 	return stream;
 }
 
-- 
cgit v1.1


From f110892ead622bdc9a7732a23aef3a08b0565608 Mon Sep 17 00:00:00 2001
From: Hersen Wu <hersenxs.wu@amd.com>
Date: Mon, 19 Mar 2018 15:22:51 -0400
Subject: drm/amd/display: Non-HDMI DP active dongle should not support YUV
 pixel format

Signed-off-by: Hersen Wu <hersenxs.wu@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Reviewed-by: Wesley Chalmers <Wesley.Chalmers@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_link.c | 19 ++++++++++++++++---
 1 file changed, 16 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
index c18f24a..e612841 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
@@ -1848,9 +1848,22 @@ static void disable_link(struct dc_link *link, enum signal_type signal)
 
 static bool dp_active_dongle_validate_timing(
 		const struct dc_crtc_timing *timing,
-		const struct dc_dongle_caps *dongle_caps)
+		const struct dpcd_caps *dpcd_caps)
 {
 	unsigned int required_pix_clk = timing->pix_clk_khz;
+	const struct dc_dongle_caps *dongle_caps = &dpcd_caps->dongle_caps;
+
+	switch (dpcd_caps->dongle_type) {
+	case DISPLAY_DONGLE_DP_VGA_CONVERTER:
+	case DISPLAY_DONGLE_DP_DVI_CONVERTER:
+	case DISPLAY_DONGLE_DP_DVI_DONGLE:
+		if (timing->pixel_encoding == PIXEL_ENCODING_RGB)
+			return true;
+		else
+			return false;
+	default:
+		break;
+	}
 
 	if (dongle_caps->dongle_type != DISPLAY_DONGLE_DP_HDMI_CONVERTER ||
 		dongle_caps->extendedCapValid == false)
@@ -1916,7 +1929,7 @@ enum dc_status dc_link_validate_mode_timing(
 		const struct dc_crtc_timing *timing)
 {
 	uint32_t max_pix_clk = stream->sink->dongle_max_pix_clk;
-	struct dc_dongle_caps *dongle_caps = &link->dpcd_caps.dongle_caps;
+	struct dpcd_caps *dpcd_caps = &link->dpcd_caps;
 
 	/* A hack to avoid failing any modes for EDID override feature on
 	 * topology change such as lower quality cable for DP or different dongle
@@ -1929,7 +1942,7 @@ enum dc_status dc_link_validate_mode_timing(
 		return DC_EXCEED_DONGLE_CAP;
 
 	/* Active Dongle*/
-	if (!dp_active_dongle_validate_timing(timing, dongle_caps))
+	if (!dp_active_dongle_validate_timing(timing, dpcd_caps))
 		return DC_EXCEED_DONGLE_CAP;
 
 	switch (stream->signal) {
-- 
cgit v1.1


From 63bd5444f6937bf6bd27a2ab79162fec784dd83c Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Wed, 14 Mar 2018 15:54:27 -0400
Subject: drm/amd/display: Fix potential access beyond end of array in CM

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
CC: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c
index 881a1bf..96d5878 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c
@@ -367,15 +367,15 @@ bool cm_helper_translate_curve_to_hw_format(
 
 	lut_params->hw_points_num = hw_points;
 
-	i = 1;
-	for (k = 0; k < MAX_REGIONS_NUMBER && i < MAX_REGIONS_NUMBER; k++) {
+	k = 0;
+	for (i = 1; i < MAX_REGIONS_NUMBER; i++) {
 		if (seg_distr[k] != -1) {
 			lut_params->arr_curve_points[k].segments_num =
 					seg_distr[k];
 			lut_params->arr_curve_points[i].offset =
 					lut_params->arr_curve_points[k].offset + (1 << seg_distr[k]);
 		}
-		i++;
+		k++;
 	}
 
 	if (seg_distr[k] != -1)
@@ -529,15 +529,15 @@ bool cm_helper_translate_curve_to_degamma_hw_format(
 
 	lut_params->hw_points_num = hw_points;
 
-	i = 1;
-	for (k = 0; k < MAX_REGIONS_NUMBER && i < MAX_REGIONS_NUMBER; k++) {
+	k = 0;
+	for (i = 1; i < MAX_REGIONS_NUMBER; i++) {
 		if (seg_distr[k] != -1) {
 			lut_params->arr_curve_points[k].segments_num =
 					seg_distr[k];
 			lut_params->arr_curve_points[i].offset =
 					lut_params->arr_curve_points[k].offset + (1 << seg_distr[k]);
 		}
-		i++;
+		k++;
 	}
 
 	if (seg_distr[k] != -1)
-- 
cgit v1.1


From 5d4b05ddd826d877327ecabf987b7c61ec3cb0c5 Mon Sep 17 00:00:00 2001
From: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Date: Thu, 15 Mar 2018 13:01:46 -0400
Subject: drm/amd/display: Add Dynamic debug prints

Created Macros for DC_LOG_XXX to pr_debug() & DRM_DEBUG_KMS.

Signed-off-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c  |  5 --
 drivers/gpu/drm/amd/display/dc/core/dc_debug.c     | 20 +++----
 drivers/gpu/drm/amd/display/dc/core/dc_link.c      | 19 ++++---
 drivers/gpu/drm/amd/display/dc/core/dc_resource.c  |  7 ++-
 drivers/gpu/drm/amd/display/dc/dce/dce_audio.c     |  9 +--
 .../gpu/drm/amd/display/dc/dce/dce_clock_source.c  | 11 ++--
 .../amd/display/dc/dce110/dce110_hw_sequencer.c    |  7 ++-
 .../drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c  | 14 ++---
 drivers/gpu/drm/amd/display/include/logger_types.h | 64 +++++++++++-----------
 9 files changed, 77 insertions(+), 79 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index a6039e5..2514d7b 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -433,11 +433,6 @@ static int amdgpu_dm_init(struct amdgpu_device *adev)
 
 	init_data.dce_environment = DCE_ENV_PRODUCTION_DRV;
 
-	if (amdgpu_dc_log)
-		init_data.log_mask = DC_DEFAULT_LOG_MASK;
-	else
-		init_data.log_mask = DC_MIN_LOG_MASK;
-
 	/*
 	 * TODO debug why this doesn't work on Raven
 	 */
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_debug.c b/drivers/gpu/drm/amd/display/dc/core/dc_debug.c
index 71cc60f..a3c8761 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_debug.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_debug.c
@@ -36,8 +36,9 @@
 #include "hw_sequencer.h"
 
 #include "resource.h"
-#define DC_LOGGER \
-	logger
+
+#define DC_LOGGER_INIT(logger)
+
 
 #define SURFACE_TRACE(...) do {\
 		if (dc->debug.surface_trace) \
@@ -60,8 +61,7 @@ void pre_surface_trace(
 		int surface_count)
 {
 	int i;
-	struct dc  *core_dc = dc;
-	struct dal_logger *logger =  core_dc->ctx->logger;
+	DC_LOGGER_INIT(dc->ctx->logger);
 
 	for (i = 0; i < surface_count; i++) {
 		const struct dc_plane_state *plane_state = plane_states[i];
@@ -183,8 +183,7 @@ void update_surface_trace(
 		int surface_count)
 {
 	int i;
-	struct dc  *core_dc = dc;
-	struct dal_logger *logger =  core_dc->ctx->logger;
+	DC_LOGGER_INIT(dc->ctx->logger);
 
 	for (i = 0; i < surface_count; i++) {
 		const struct dc_surface_update *update = &updates[i];
@@ -304,8 +303,7 @@ void update_surface_trace(
 
 void post_surface_trace(struct dc *dc)
 {
-	struct dc  *core_dc = dc;
-	struct dal_logger *logger =  core_dc->ctx->logger;
+	DC_LOGGER_INIT(dc->ctx->logger);
 
 	SURFACE_TRACE("post surface process.\n");
 
@@ -317,10 +315,10 @@ void context_timing_trace(
 {
 	int i;
 	struct dc  *core_dc = dc;
-	struct dal_logger *logger =  core_dc->ctx->logger;
 	int h_pos[MAX_PIPES], v_pos[MAX_PIPES];
 	struct crtc_position position;
 	unsigned int underlay_idx = core_dc->res_pool->underlay_pipe_index;
+	DC_LOGGER_INIT(dc->ctx->logger);
 
 
 	for (i = 0; i < core_dc->res_pool->pipe_count; i++) {
@@ -355,9 +353,7 @@ void context_clock_trace(
 		struct dc_state *context)
 {
 #if defined(CONFIG_DRM_AMD_DC_DCN1_0)
-	struct dc  *core_dc = dc;
-	struct dal_logger *logger =  core_dc->ctx->logger;
-
+	DC_LOGGER_INIT(dc->ctx->logger);
 	CLOCK_TRACE("Current: dispclk_khz:%d  max_dppclk_khz:%d  dcfclk_khz:%d\n"
 			"dcfclk_deep_sleep_khz:%d  fclk_khz:%d  socclk_khz:%d\n",
 			context->bw.dcn.calc_clk.dispclk_khz,
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
index e612841..d9efdd9 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
@@ -45,8 +45,9 @@
 #include "dce/dce_11_0_d.h"
 #include "dce/dce_11_0_enum.h"
 #include "dce/dce_11_0_sh_mask.h"
-#define DC_LOGGER \
-	dc_ctx->logger
+
+#define DC_LOGGER_INIT(logger)
+
 
 #define LINK_INFO(...) \
 	DC_LOG_HW_HOTPLUG(  \
@@ -561,7 +562,7 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 	struct dc_context *dc_ctx = link->ctx;
 	struct dc_sink *sink = NULL;
 	enum dc_connection_type new_connection_type = dc_connection_none;
-
+	DC_LOGGER_INIT(link->ctx->logger);
 	if (link->connector_signal == SIGNAL_TYPE_VIRTUAL)
 		return false;
 
@@ -927,6 +928,7 @@ static bool construct(
 	struct integrated_info info = {{{ 0 }}};
 	struct dc_bios *bios = init_params->dc->ctx->dc_bios;
 	const struct dc_vbios_funcs *bp_funcs = bios->funcs;
+	DC_LOGGER_INIT(dc_ctx->logger);
 
 	link->irq_source_hpd = DC_IRQ_SOURCE_INVALID;
 	link->irq_source_hpd_rx = DC_IRQ_SOURCE_INVALID;
@@ -1135,7 +1137,8 @@ static void dpcd_configure_panel_mode(
 {
 	union dpcd_edp_config edp_config_set;
 	bool panel_mode_edp = false;
-	struct dc_context *dc_ctx = link->ctx;
+	DC_LOGGER_INIT(link->ctx->logger);
+
 	memset(&edp_config_set, '\0', sizeof(union dpcd_edp_config));
 
 	if (DP_PANEL_MODE_DEFAULT != panel_mode) {
@@ -1968,10 +1971,10 @@ bool dc_link_set_backlight_level(const struct dc_link *link, uint32_t level,
 	struct dc  *core_dc = link->ctx->dc;
 	struct abm *abm = core_dc->res_pool->abm;
 	struct dmcu *dmcu = core_dc->res_pool->dmcu;
-	struct dc_context *dc_ctx = link->ctx;
 	unsigned int controller_id = 0;
 	bool use_smooth_brightness = true;
 	int i;
+	DC_LOGGER_INIT(link->ctx->logger);
 
 	if ((dmcu == NULL) ||
 		(abm == NULL) ||
@@ -2154,8 +2157,8 @@ static enum dc_status allocate_mst_payload(struct pipe_ctx *pipe_ctx)
 	struct fixed31_32 avg_time_slots_per_mtp;
 	struct fixed31_32 pbn;
 	struct fixed31_32 pbn_per_slot;
-	struct dc_context *dc_ctx = link->ctx;
 	uint8_t i;
+	DC_LOGGER_INIT(link->ctx->logger);
 
 	/* enable_link_dp_mst already check link->enabled_stream_count
 	 * and stream is in link->stream[]. This is called during set mode,
@@ -2234,7 +2237,7 @@ static enum dc_status deallocate_mst_payload(struct pipe_ctx *pipe_ctx)
 	struct fixed31_32 avg_time_slots_per_mtp = dal_fixed31_32_from_int(0);
 	uint8_t i;
 	bool mst_mode = (link->type == dc_connection_mst_branch);
-	struct dc_context *dc_ctx = link->ctx;
+	DC_LOGGER_INIT(link->ctx->logger);
 
 	/* deallocate_mst_payload is called before disable link. When mode or
 	 * disable/enable monitor, new stream is created which is not in link
@@ -2307,8 +2310,8 @@ void core_link_enable_stream(
 		struct pipe_ctx *pipe_ctx)
 {
 	struct dc  *core_dc = pipe_ctx->stream->ctx->dc;
-	struct dc_context *dc_ctx = pipe_ctx->stream->ctx;
 	enum dc_status status;
+	DC_LOGGER_INIT(pipe_ctx->stream->ctx->logger);
 
 	/* eDP lit up by bios already, no need to enable again. */
 	if (pipe_ctx->stream->signal == SIGNAL_TYPE_EDP &&
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
index 379b055..50b84f6 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
@@ -45,8 +45,9 @@
 #include "dcn10/dcn10_resource.h"
 #endif
 #include "dce120/dce120_resource.h"
-#define DC_LOGGER \
-	ctx->logger
+
+#define DC_LOGGER_INIT(logger)
+
 enum dce_version resource_parse_asic_id(struct hw_asic_id asic_id)
 {
 	enum dce_version dc_version = DCE_VERSION_UNKNOWN;
@@ -835,7 +836,7 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx)
 	struct dc_crtc_timing *timing = &pipe_ctx->stream->timing;
 	struct view recout_skip = { 0 };
 	bool res = false;
-	struct dc_context *ctx = pipe_ctx->stream->ctx;
+	DC_LOGGER_INIT(pipe_ctx->stream->ctx->logger);
 	/* Important: scaling ratio calculation requires pixel format,
 	 * lb depth calculation requires recout and taps require scaling ratios.
 	 * Inits require viewport, taps, ratios and recout of split pipe
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_audio.c b/drivers/gpu/drm/amd/display/dc/dce/dce_audio.c
index 6d5cdcd..7f6d724 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_audio.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_audio.c
@@ -33,8 +33,9 @@
 
 #define CTX \
 	aud->base.ctx
-#define DC_LOGGER \
-	aud->base.ctx->logger
+
+#define DC_LOGGER_INIT()
+
 #define REG(reg)\
 	(aud->regs->reg)
 
@@ -348,8 +349,8 @@ static void set_audio_latency(
 
 void dce_aud_az_enable(struct audio *audio)
 {
-	struct dce_audio *aud = DCE_AUD(audio);
 	uint32_t value = AZ_REG_READ(AZALIA_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL);
+	DC_LOGGER_INIT();
 
 	set_reg_field_value(value, 1,
 			    AZALIA_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL,
@@ -371,7 +372,7 @@ void dce_aud_az_enable(struct audio *audio)
 void dce_aud_az_disable(struct audio *audio)
 {
 	uint32_t value;
-	struct dce_audio *aud = DCE_AUD(audio);
+	DC_LOGGER_INIT();
 
 	value = AZ_REG_READ(AZALIA_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL);
 	set_reg_field_value(value, 1,
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c
index 0aa2cda..67dad7f 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c
@@ -41,8 +41,9 @@
 
 #define CTX \
 	clk_src->base.ctx
-#define DC_LOGGER \
-	calc_pll_cs->ctx->logger
+
+#define DC_LOGGER_INIT()
+
 #undef FN
 #define FN(reg_name, field_name) \
 	clk_src->cs_shift->field_name, clk_src->cs_mask->field_name
@@ -467,7 +468,7 @@ static uint32_t dce110_get_pix_clk_dividers_helper (
 {
 	uint32_t field = 0;
 	uint32_t pll_calc_error = MAX_PLL_CALC_ERROR;
-	struct calc_pll_clock_source *calc_pll_cs = &clk_src->calc_pll;
+	DC_LOGGER_INIT();
 	/* Check if reference clock is external (not pcie/xtalin)
 	* HW Dce80 spec:
 	* 00 - PCIE_REFCLK, 01 - XTALIN,    02 - GENERICA,    03 - GENERICB
@@ -557,8 +558,8 @@ static uint32_t dce110_get_pix_clk_dividers(
 		struct pll_settings *pll_settings)
 {
 	struct dce110_clk_src *clk_src = TO_DCE110_CLK_SRC(cs);
-	struct calc_pll_clock_source *calc_pll_cs = &clk_src->calc_pll;
 	uint32_t pll_calc_error = MAX_PLL_CALC_ERROR;
+	DC_LOGGER_INIT();
 
 	if (pix_clk_params == NULL || pll_settings == NULL
 			|| pix_clk_params->requested_pix_clk == 0) {
@@ -1054,7 +1055,7 @@ static void get_ss_info_from_atombios(
 	struct spread_spectrum_info *ss_info_cur;
 	struct spread_spectrum_data *ss_data_cur;
 	uint32_t i;
-	struct calc_pll_clock_source *calc_pll_cs = &clk_src->calc_pll;
+	DC_LOGGER_INIT();
 	if (ss_entries_num == NULL) {
 		DC_LOG_SYNC(
 			"Invalid entry !!!\n");
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
index c621230..e8df50f 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
@@ -70,8 +70,9 @@
 
 #define CTX \
 	hws->ctx
-#define DC_LOGGER \
-	ctx->logger
+
+#define DC_LOGGER_INIT()
+
 #define REG(reg)\
 	hws->regs->reg
 
@@ -2701,7 +2702,7 @@ static void dce110_program_front_end_for_pipe(
 	struct xfm_grph_csc_adjustment adjust;
 	struct out_csc_color_matrix tbl_entry;
 	unsigned int i;
-	struct dc_context *ctx = dc->ctx;
+	DC_LOGGER_INIT();
 	memset(&tbl_entry, 0, sizeof(tbl_entry));
 
 	if (dc->current_state)
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
index e214581..de5293d 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -45,8 +45,8 @@
 #include "dcn10_hubbub.h"
 #include "dcn10_cm_common.h"
 
-#define DC_LOGGER \
-	ctx->logger
+#define DC_LOGGER_INIT(logger)
+
 #define CTX \
 	hws->ctx
 #define REG(reg)\
@@ -363,7 +363,7 @@ static void power_on_plane(
 	struct dce_hwseq *hws,
 	int plane_id)
 {
-	struct dc_context *ctx = hws->ctx;
+	DC_LOGGER_INIT(hws->ctx->logger);
 	if (REG(DC_IP_REQUEST_CNTL)) {
 		REG_SET(DC_IP_REQUEST_CNTL, 0,
 				IP_REQUEST_EN, 1);
@@ -562,7 +562,7 @@ static void reset_back_end_for_pipe(
 		struct dc_state *context)
 {
 	int i;
-	struct dc_context *ctx = dc->ctx;
+	DC_LOGGER_INIT(dc->ctx->logger);
 	if (pipe_ctx->stream_res.stream_enc == NULL) {
 		pipe_ctx->stream = NULL;
 		return;
@@ -658,7 +658,7 @@ static void plane_atomic_power_down(struct dc *dc, struct pipe_ctx *pipe_ctx)
 {
 	struct dce_hwseq *hws = dc->hwseq;
 	struct dpp *dpp = pipe_ctx->plane_res.dpp;
-	struct dc_context *ctx = dc->ctx;
+	DC_LOGGER_INIT(dc->ctx->logger);
 
 	if (REG(DC_IP_REQUEST_CNTL)) {
 		REG_SET(DC_IP_REQUEST_CNTL, 0,
@@ -708,7 +708,7 @@ static void plane_atomic_disable(struct dc *dc, struct pipe_ctx *pipe_ctx)
 
 static void dcn10_disable_plane(struct dc *dc, struct pipe_ctx *pipe_ctx)
 {
-	struct dc_context *ctx = dc->ctx;
+	DC_LOGGER_INIT(dc->ctx->logger);
 
 	if (!pipe_ctx->plane_res.hubp || pipe_ctx->plane_res.hubp->power_gated)
 		return;
@@ -2001,9 +2001,9 @@ static void dcn10_apply_ctx_for_surface(
 	bool removed_pipe[4] = { false };
 	unsigned int ref_clk_mhz = dc->res_pool->ref_clock_inKhz/1000;
 	bool program_water_mark = false;
-	struct dc_context *ctx = dc->ctx;
 	struct pipe_ctx *top_pipe_to_program =
 			find_top_pipe_for_stream(dc, context, stream);
+	DC_LOGGER_INIT(dc->ctx->logger);
 
 	if (!top_pipe_to_program)
 		return;
diff --git a/drivers/gpu/drm/amd/display/include/logger_types.h b/drivers/gpu/drm/amd/display/include/logger_types.h
index 427796b..4f332e8 100644
--- a/drivers/gpu/drm/amd/display/include/logger_types.h
+++ b/drivers/gpu/drm/amd/display/include/logger_types.h
@@ -29,39 +29,39 @@
 #include "os_types.h"
 
 #define MAX_NAME_LEN 32
-#define DC_LOG_ERROR(a, ...) dm_logger_write(DC_LOGGER, LOG_ERROR, a, ## __VA_ARGS__)
-#define DC_LOG_WARNING(a, ...) dm_logger_write(DC_LOGGER, LOG_WARNING, a, ## __VA_ARGS__)
-#define DC_LOG_DEBUG(a, ...) dm_logger_write(DC_LOGGER, LOG_DEBUG, a, ## __VA_ARGS__)
-#define DC_LOG_DC(a, ...) dm_logger_write(DC_LOGGER, LOG_DC, a, ## __VA_ARGS__)
-#define DC_LOG_DTN(a, ...) dm_logger_write(DC_LOGGER, LOG_DTN, a, ## __VA_ARGS__)
-#define DC_LOG_SURFACE(a, ...) dm_logger_write(DC_LOGGER, LOG_SURFACE, a, ## __VA_ARGS__)
-#define DC_LOG_HW_HOTPLUG(a, ...) dm_logger_write(DC_LOGGER, LOG_HW_HOTPLUG, a, ## __VA_ARGS__)
-#define DC_LOG_HW_LINK_TRAINING(a, ...) dm_logger_write(DC_LOGGER, LOG_HW_LINK_TRAINING, a, ## __VA_ARGS__)
-#define DC_LOG_HW_SET_MODE(a, ...) dm_logger_write(DC_LOGGER, LOG_HW_SET_MODE, a, ## __VA_ARGS__)
-#define DC_LOG_HW_RESUME_S3(a, ...) dm_logger_write(DC_LOGGER, LOG_HW_RESUME_S3, a, ## __VA_ARGS__)
-#define DC_LOG_HW_AUDIO(a, ...) dm_logger_write(DC_LOGGER, LOG_HW_AUDIO, a, ## __VA_ARGS__)
-#define DC_LOG_HW_HPD_IRQ(a, ...) dm_logger_write(DC_LOGGER, LOG_HW_HPD_IRQ, a, ## __VA_ARGS__)
-#define DC_LOG_MST(a, ...) dm_logger_write(DC_LOGGER, LOG_MST, a, ## __VA_ARGS__)
-#define DC_LOG_SCALER(a, ...) dm_logger_write(DC_LOGGER, LOG_SCALER, a, ## __VA_ARGS__)
-#define DC_LOG_BIOS(a, ...) dm_logger_write(DC_LOGGER, LOG_BIOS, a, ## __VA_ARGS__)
-#define DC_LOG_BANDWIDTH_CALCS(a, ...) dm_logger_write(DC_LOGGER, LOG_BANDWIDTH_CALCS, a, ## __VA_ARGS__)
-#define DC_LOG_BANDWIDTH_VALIDATION(a, ...) dm_logger_write(DC_LOGGER, LOG_BANDWIDTH_VALIDATION, a, ## __VA_ARGS__)
-#define DC_LOG_I2C_AUX(a, ...) dm_logger_write(DC_LOGGER, LOG_I2C_AUX, a, ## __VA_ARGS__)
-#define DC_LOG_SYNC(a, ...) dm_logger_write(DC_LOGGER, LOG_SYNC, a, ## __VA_ARGS__)
-#define DC_LOG_BACKLIGHT(a, ...) dm_logger_write(DC_LOGGER, LOG_BACKLIGHT, a, ## __VA_ARGS__)
-#define DC_LOG_FEATURE_OVERRIDE(a, ...) dm_logger_write(DC_LOGGER, LOG_FEATURE_OVERRIDE, a, ## __VA_ARGS__)
-#define DC_LOG_DETECTION_EDID_PARSER(a, ...) dm_logger_write(DC_LOGGER, LOG_DETECTION_EDID_PARSER, a, ## __VA_ARGS__)
-#define DC_LOG_DETECTION_DP_CAPS(a, ...) dm_logger_write(DC_LOGGER, LOG_DETECTION_DP_CAPS, a, ## __VA_ARGS__)
-#define DC_LOG_RESOURCE(a, ...) dm_logger_write(DC_LOGGER, LOG_RESOURCE, a, ## __VA_ARGS__)
-#define DC_LOG_DML(a, ...) dm_logger_write(DC_LOGGER, LOG_DML, a, ## __VA_ARGS__)
-#define DC_LOG_EVENT_MODE_SET(a, ...) dm_logger_write(DC_LOGGER, LOG_EVENT_MODE_SET, a, ## __VA_ARGS__)
-#define DC_LOG_EVENT_DETECTION(a, ...) dm_logger_write(DC_LOGGER, LOG_EVENT_DETECTION, a, ## __VA_ARGS__)
-#define DC_LOG_EVENT_LINK_TRAINING(a, ...) dm_logger_write(DC_LOGGER, LOG_EVENT_LINK_TRAINING, a, ## __VA_ARGS__)
-#define DC_LOG_EVENT_LINK_LOSS(a, ...) dm_logger_write(DC_LOGGER, LOG_EVENT_LINK_LOSS, a, ## __VA_ARGS__)
-#define DC_LOG_EVENT_UNDERFLOW(a, ...) dm_logger_write(DC_LOGGER, LOG_EVENT_UNDERFLOW, a, ## __VA_ARGS__)
-#define DC_LOG_IF_TRACE(a, ...) dm_logger_write(DC_LOGGER, LOG_IF_TRACE, a, ## __VA_ARGS__)
-#define DC_LOG_PERF_TRACE(a, ...) dm_logger_write(DC_LOGGER, LOG_PERF_TRACE, a, ## __VA_ARGS__)
 
+#define DC_LOG_ERROR(...) DRM_ERROR(__VA_ARGS__)
+#define DC_LOG_WARNING(...) DRM_WARN(__VA_ARGS__)
+#define DC_LOG_DEBUG(...) DRM_INFO(__VA_ARGS__)
+#define DC_LOG_DC(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_DTN(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_SURFACE(...) pr_debug("[SURFACE]:"__VA_ARGS__)
+#define DC_LOG_HW_HOTPLUG(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_HW_LINK_TRAINING(...) pr_debug("[HW_LINK_TRAINING]:"__VA_ARGS__)
+#define DC_LOG_HW_SET_MODE(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_HW_RESUME_S3(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_HW_AUDIO(...) pr_debug("[HW_AUDIO]:"__VA_ARGS__)
+#define DC_LOG_HW_HPD_IRQ(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_MST(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_SCALER(...) pr_debug("[SCALER]:"__VA_ARGS__)
+#define DC_LOG_BIOS(...) pr_debug("[BIOS]:"__VA_ARGS__)
+#define DC_LOG_BANDWIDTH_CALCS(...) pr_debug("[BANDWIDTH_CALCS]:"__VA_ARGS__)
+#define DC_LOG_BANDWIDTH_VALIDATION(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_I2C_AUX(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_SYNC(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_BACKLIGHT(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_FEATURE_OVERRIDE(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_DETECTION_EDID_PARSER(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_DETECTION_DP_CAPS(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_RESOURCE(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_DML(...) pr_debug("[DML]:"__VA_ARGS__)
+#define DC_LOG_EVENT_MODE_SET(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_EVENT_DETECTION(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_EVENT_LINK_TRAINING(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_EVENT_LINK_LOSS(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_EVENT_UNDERFLOW(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_IF_TRACE(...) pr_debug("[IF_TRACE]:"__VA_ARGS__)
+#define DC_LOG_PERF_TRACE(...) DRM_DEBUG_KMS(__VA_ARGS__)
 
 struct dal_logger;
 
-- 
cgit v1.1


From 01fe3e4876d3799b37e6c712dcfed7cc2cafa3f0 Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Thu, 15 Mar 2018 13:34:16 -0400
Subject: drm/amd/display: Add vmax/min_sel prints to dcn10_log_hw_state

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c | 6 ++++--
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c         | 6 ++++++
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h         | 2 ++
 3 files changed, 12 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
index de5293d..f3341a2 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -165,7 +165,7 @@ void dcn10_log_hw_state(struct dc *dc)
 	}
 	DTN_INFO("\n");
 
-	DTN_INFO("OTG:  v_bs  v_be  v_ss  v_se  vpol  vmax  vmin"
+	DTN_INFO("OTG:  v_bs  v_be  v_ss  v_se  vpol  vmax  vmin  vmax_sel  vmin_sel"
 			"  h_bs  h_be  h_ss  h_se  hpol  htot  vtot  underflow\n");
 
 	for (i = 0; i < pool->timing_generator_count; i++) {
@@ -178,7 +178,7 @@ void dcn10_log_hw_state(struct dc *dc)
 		if ((s.otg_enabled & 1) == 0)
 			continue;
 
-		DTN_INFO("[%d]: %5d %5d %5d %5d %5d %5d %5d %5d %5d %5d"
+		DTN_INFO("[%d]: %5d %5d %5d %5d %5d %5d %5d %9d %9d %5d %5d %5d"
 				" %5d %5d %5d %5d  %9d\n",
 				tg->inst,
 				s.v_blank_start,
@@ -188,6 +188,8 @@ void dcn10_log_hw_state(struct dc *dc)
 				s.v_sync_a_pol,
 				s.v_total_max,
 				s.v_total_min,
+				s.v_total_max_sel,
+				s.v_total_min_sel,
 				s.h_blank_start,
 				s.h_blank_end,
 				s.h_sync_a_start,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c
index dc92130..2c5dbec 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c
@@ -1229,6 +1229,12 @@ void optc1_read_otg_state(struct optc *optc1,
 	REG_GET(OTG_V_TOTAL_MIN,
 			OTG_V_TOTAL_MIN, &s->v_total_min);
 
+	REG_GET(OTG_V_TOTAL_CONTROL,
+			OTG_V_TOTAL_MAX_SEL, &s->v_total_max_sel);
+
+	REG_GET(OTG_V_TOTAL_CONTROL,
+			OTG_V_TOTAL_MIN_SEL, &s->v_total_min_sel);
+
 	REG_GET_2(OTG_V_SYNC_A,
 			OTG_V_SYNC_A_START, &s->v_sync_a_start,
 			OTG_V_SYNC_A_END, &s->v_sync_a_end);
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h
index 5a9a73d..89e09e5 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h
@@ -406,6 +406,8 @@ struct dcn_otg_state {
 	uint32_t v_total;
 	uint32_t v_total_max;
 	uint32_t v_total_min;
+	uint32_t v_total_min_sel;
+	uint32_t v_total_max_sel;
 	uint32_t v_sync_a_start;
 	uint32_t v_sync_a_end;
 	uint32_t h_blank_start;
-- 
cgit v1.1


From 4c61af8afe855fcf65a09d47c8e330bb1fd1fb4a Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Fri, 23 Mar 2018 13:39:27 -0400
Subject: drm/amd/display: Implement dm_get_timestamp

We use this to ensure we wait at least 500ms in between eDP
disable/enable.

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c
index 89342b4..fe29125 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c
@@ -37,8 +37,10 @@
 
 unsigned long long dm_get_timestamp(struct dc_context *ctx)
 {
-	/* TODO: return actual timestamp */
-	return 0;
+	struct timespec64 time;
+
+	getrawmonotonic64(&time);
+	return timespec64_to_ns(&time);
 }
 
 void dm_perf_trace_timestamp(const char *func_name, unsigned int line)
-- 
cgit v1.1


From 78d5d04d118d55b6c51ca787d5debb9ad1b8a391 Mon Sep 17 00:00:00 2001
From: Charlene Liu <charlene.liu@amd.com>
Date: Tue, 20 Mar 2018 14:53:04 -0400
Subject: drm/amd/display: add delay between panel pwr off to on.

As per eDP 1.4 spec, there must be at least 500ms delay
between eDP power off and on.
This change added time stamp when edp power off, which can
be used to calculate duration time when edp power on.
If duration less than 500ms, add a wait.

Signed-off-by: Charlene Liu <charlene.liu@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../drm/amd/display/amdgpu_dm/amdgpu_dm_services.c |  7 ++++++
 drivers/gpu/drm/amd/display/dc/dc_link.h           |  9 +++++++
 .../amd/display/dc/dce110/dce110_hw_sequencer.c    | 29 +++++++++++++++++++++-
 drivers/gpu/drm/amd/display/dc/dm_services.h       |  4 +++
 4 files changed, 48 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c
index fe29125..0229c7ed 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c
@@ -43,6 +43,13 @@ unsigned long long dm_get_timestamp(struct dc_context *ctx)
 	return timespec64_to_ns(&time);
 }
 
+unsigned long long dm_get_elapse_time_in_ns(struct dc_context *ctx,
+		unsigned long long current_time_stamp,
+		unsigned long long last_time_stamp)
+{
+	return current_time_stamp - last_time_stamp;
+}
+
 void dm_perf_trace_timestamp(const char *func_name, unsigned int line)
 {
 }
diff --git a/drivers/gpu/drm/amd/display/dc/dc_link.h b/drivers/gpu/drm/amd/display/dc/dc_link.h
index fb4d9ea..eeff987 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_link.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_link.h
@@ -51,6 +51,14 @@ struct link_mst_stream_allocation_table {
 	struct link_mst_stream_allocation stream_allocations[MAX_CONTROLLER_NUM];
 };
 
+struct time_stamp {
+	uint64_t edp_poweroff;
+	uint64_t edp_poweron;
+};
+
+struct link_trace {
+	struct time_stamp time_stamp;
+};
 /*
  * A link contains one or more sinks and their connected status.
  * The currently active signal type (HDMI, DP-SST, DP-MST) is also reported.
@@ -114,6 +122,7 @@ struct dc_link {
 
 	struct dc_link_status link_status;
 
+	struct link_trace link_trace;
 };
 
 const struct dc_link_status *dc_link_get_status(const struct dc_link *dc_link);
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
index e8df50f..db2d15d 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
@@ -849,6 +849,28 @@ void hwss_edp_power_control(
 
 	if (power_up != is_panel_powered_on(hwseq)) {
 		/* Send VBIOS command to prompt eDP panel power */
+		if (power_up) {
+			unsigned long long current_ts = dm_get_timestamp(ctx);
+			unsigned long long duration_in_ms =
+					dm_get_elapse_time_in_ns(
+							ctx,
+							current_ts,
+							link->link_trace.time_stamp.edp_poweroff) / 1000000;
+			unsigned long long wait_time_ms = 0;
+
+			/* max 500ms from LCDVDD off to on */
+			if (link->link_trace.time_stamp.edp_poweroff == 0)
+				wait_time_ms = 500;
+			else if (duration_in_ms < 500)
+				wait_time_ms = 500 - duration_in_ms;
+
+			if (wait_time_ms) {
+				msleep(wait_time_ms);
+				dm_output_to_console("%s: wait %lld ms to power on eDP.\n",
+						__func__, wait_time_ms);
+			}
+
+		}
 
 		DC_LOG_HW_RESUME_S3(
 				"%s: Panel Power action: %s\n",
@@ -862,9 +884,14 @@ void hwss_edp_power_control(
 		cntl.coherent = false;
 		cntl.lanes_number = LANE_COUNT_FOUR;
 		cntl.hpd_sel = link->link_enc->hpd_source;
-
 		bp_result = link_transmitter_control(ctx->dc_bios, &cntl);
 
+		if (!power_up)
+			/*save driver power off time stamp*/
+			link->link_trace.time_stamp.edp_poweroff = dm_get_timestamp(ctx);
+		else
+			link->link_trace.time_stamp.edp_poweron = dm_get_timestamp(ctx);
+
 		if (bp_result != BP_RESULT_OK)
 			DC_LOG_ERROR(
 					"%s: Panel Power bp_result: %d\n",
diff --git a/drivers/gpu/drm/amd/display/dc/dm_services.h b/drivers/gpu/drm/amd/display/dc/dm_services.h
index 22e7ee7..8eafe1a 100644
--- a/drivers/gpu/drm/amd/display/dc/dm_services.h
+++ b/drivers/gpu/drm/amd/display/dc/dm_services.h
@@ -341,6 +341,10 @@ bool dm_dmcu_set_pipe(struct dc_context *ctx, unsigned int controller_id);
 
 unsigned long long dm_get_timestamp(struct dc_context *ctx);
 
+unsigned long long dm_get_elapse_time_in_ns(struct dc_context *ctx,
+		unsigned long long current_time_stamp,
+		unsigned long long last_time_stamp);
+
 /*
  * performance tracing
  */
-- 
cgit v1.1


From 1402c605173bbbb1c2f7e615a1708a6ee61c69c3 Mon Sep 17 00:00:00 2001
From: Eric Yang <Eric.Yang2@amd.com>
Date: Wed, 14 Mar 2018 17:56:58 -0400
Subject: drm/amd/display: Set all update flags when we have full update

To prevent future optimization related bugs, just set all update
flags when we have a full update, since we know we want to reprogram
everything in that case.

Signed-off-by: Eric Yang <Eric.Yang2@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c
index 6f4ad67..b331d9e 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc.c
@@ -1208,7 +1208,7 @@ enum surface_update_type dc_check_update_surfaces_for_stream(
 	type = check_update_surfaces_for_stream(dc, updates, surface_count, stream_update, stream_status);
 	if (type == UPDATE_TYPE_FULL)
 		for (i = 0; i < surface_count; i++)
-			updates[i].surface->update_flags.bits.full_update = 1;
+			updates[i].surface->update_flags.raw = 0xFFFFFFFF;
 
 	return type;
 }
-- 
cgit v1.1


From 0c41891c81c017b5040b211f0b294ff5eb440d44 Mon Sep 17 00:00:00 2001
From: Eric Bernstein <eric.bernstein@amd.com>
Date: Mon, 19 Mar 2018 14:41:59 -0400
Subject: drm/amd/display: Refactor stream encoder for HW review

Move DCN1 implementation of stream encoder to new file (instead
of common dce_stream_encoder.c).
Cleanup code related to different implementation due to register
definition differences.

Signed-off-by: Eric Bernstein <eric.bernstein@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn10/Makefile      |    2 +-
 .../gpu/drm/amd/display/dc/dcn10/dcn10_resource.c  |   36 +-
 .../amd/display/dc/dcn10/dcn10_stream_encoder.c    | 1505 ++++++++++++++++++++
 .../amd/display/dc/dcn10/dcn10_stream_encoder.h    |  584 ++++++++
 4 files changed, 2101 insertions(+), 26 deletions(-)
 create mode 100644 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c
 create mode 100644 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.h

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/Makefile b/drivers/gpu/drm/amd/display/dc/dcn10/Makefile
index 5469bdf..5c69743 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/Makefile
@@ -26,7 +26,7 @@ DCN10 = dcn10_resource.o dcn10_ipp.o dcn10_hw_sequencer.o \
 		dcn10_dpp.o dcn10_opp.o dcn10_optc.o \
 		dcn10_hubp.o dcn10_mpc.o \
 		dcn10_dpp_dscl.o dcn10_dpp_cm.o dcn10_cm_common.o \
-		dcn10_hubbub.o
+		dcn10_hubbub.o dcn10_stream_encoder.o
 
 AMD_DAL_DCN10 = $(addprefix $(AMDDALPATH)/dc/dcn10/,$(DCN10))
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
index 7ad290c..f305f65 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
@@ -39,7 +39,7 @@
 #include "dce110/dce110_hw_sequencer.h"
 #include "dcn10/dcn10_opp.h"
 #include "dce/dce_link_encoder.h"
-#include "dce/dce_stream_encoder.h"
+#include "dcn10/dcn10_stream_encoder.h"
 #include "dce/dce_clocks.h"
 #include "dce/dce_clock_source.h"
 #include "dce/dce_audio.h"
@@ -166,36 +166,22 @@ static const struct dce_abm_mask abm_mask = {
 
 #define stream_enc_regs(id)\
 [id] = {\
-	SE_DCN_REG_LIST(id),\
-	.TMDS_CNTL = 0,\
-	.AFMT_AVI_INFO0 = 0,\
-	.AFMT_AVI_INFO1 = 0,\
-	.AFMT_AVI_INFO2 = 0,\
-	.AFMT_AVI_INFO3 = 0,\
+	SE_DCN_REG_LIST(id)\
 }
 
-static const struct dce110_stream_enc_registers stream_enc_regs[] = {
+static const struct dcn10_stream_enc_registers stream_enc_regs[] = {
 	stream_enc_regs(0),
 	stream_enc_regs(1),
 	stream_enc_regs(2),
 	stream_enc_regs(3),
 };
 
-static const struct dce_stream_encoder_shift se_shift = {
+static const struct dcn10_stream_encoder_shift se_shift = {
 		SE_COMMON_MASK_SH_LIST_DCN10(__SHIFT)
 };
 
-static const struct dce_stream_encoder_mask se_mask = {
-		SE_COMMON_MASK_SH_LIST_DCN10(_MASK),
-		.AFMT_GENERIC0_UPDATE = 0,
-		.AFMT_GENERIC2_UPDATE = 0,
-		.DP_DYN_RANGE = 0,
-		.DP_YCBCR_RANGE = 0,
-		.HDMI_AVI_INFO_SEND = 0,
-		.HDMI_AVI_INFO_CONT = 0,
-		.HDMI_AVI_INFO_LINE = 0,
-		.DP_SEC_AVI_ENABLE = 0,
-		.AFMT_AVI_INFO_VERSION = 0
+static const struct dcn10_stream_encoder_mask se_mask = {
+		SE_COMMON_MASK_SH_LIST_DCN10(_MASK)
 };
 
 #define audio_regs(id)\
@@ -653,16 +639,16 @@ static struct stream_encoder *dcn10_stream_encoder_create(
 	enum engine_id eng_id,
 	struct dc_context *ctx)
 {
-	struct dce110_stream_encoder *enc110 =
-		kzalloc(sizeof(struct dce110_stream_encoder), GFP_KERNEL);
+	struct dcn10_stream_encoder *enc1 =
+		kzalloc(sizeof(struct dcn10_stream_encoder), GFP_KERNEL);
 
-	if (!enc110)
+	if (!enc1)
 		return NULL;
 
-	dce110_stream_encoder_construct(enc110, ctx, ctx->dc_bios, eng_id,
+	dcn10_stream_encoder_construct(enc1, ctx, ctx->dc_bios, eng_id,
 					&stream_enc_regs[eng_id],
 					&se_shift, &se_mask);
-	return &enc110->base;
+	return &enc1->base;
 }
 
 static const struct dce_hwseq_registers hwseq_reg = {
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c
new file mode 100644
index 0000000..0413c70
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c
@@ -0,0 +1,1505 @@
+/*
+ * Copyright 2012-15 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ *  and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+
+#include "dc_bios_types.h"
+#include "dcn10_stream_encoder.h"
+
+#include "reg_helper.h"
+#define DC_LOGGER \
+		enc1->base.ctx->logger
+enum DP_PIXEL_ENCODING {
+DP_PIXEL_ENCODING_RGB444                 = 0x00000000,
+DP_PIXEL_ENCODING_YCBCR422               = 0x00000001,
+DP_PIXEL_ENCODING_YCBCR444               = 0x00000002,
+DP_PIXEL_ENCODING_RGB_WIDE_GAMUT         = 0x00000003,
+DP_PIXEL_ENCODING_Y_ONLY                 = 0x00000004,
+DP_PIXEL_ENCODING_YCBCR420               = 0x00000005,
+DP_PIXEL_ENCODING_RESERVED               = 0x00000006,
+};
+
+
+enum DP_COMPONENT_DEPTH {
+DP_COMPONENT_DEPTH_6BPC                  = 0x00000000,
+DP_COMPONENT_DEPTH_8BPC                  = 0x00000001,
+DP_COMPONENT_DEPTH_10BPC                 = 0x00000002,
+DP_COMPONENT_DEPTH_12BPC                 = 0x00000003,
+DP_COMPONENT_DEPTH_16BPC                 = 0x00000004,
+DP_COMPONENT_DEPTH_RESERVED              = 0x00000005,
+};
+
+
+#define REG(reg)\
+	(enc1->regs->reg)
+
+#undef FN
+#define FN(reg_name, field_name) \
+	enc1->se_shift->field_name, enc1->se_mask->field_name
+
+#define VBI_LINE_0 0
+#define DP_BLANK_MAX_RETRY 20
+#define HDMI_CLOCK_CHANNEL_RATE_MORE_340M 340000
+
+
+enum {
+	DP_MST_UPDATE_MAX_RETRY = 50
+};
+
+#define CTX \
+	enc1->base.ctx
+
+static void enc1_update_generic_info_packet(
+	struct dcn10_stream_encoder *enc1,
+	uint32_t packet_index,
+	const struct dc_info_packet *info_packet)
+{
+	uint32_t regval;
+	/* TODOFPGA Figure out a proper number for max_retries polling for lock
+	 * use 50 for now.
+	 */
+	uint32_t max_retries = 50;
+
+	/*we need turn on clock before programming AFMT block*/
+	REG_UPDATE(AFMT_CNTL, AFMT_AUDIO_CLOCK_EN, 1);
+
+	if (packet_index >= 8)
+		ASSERT(0);
+
+	/* poll dig_update_lock is not locked -> asic internal signal
+	 * assume otg master lock will unlock it
+	 */
+/*		REG_WAIT(AFMT_VBI_PACKET_CONTROL, AFMT_GENERIC_LOCK_STATUS,
+			0, 10, max_retries);*/
+
+	/* check if HW reading GSP memory */
+	REG_WAIT(AFMT_VBI_PACKET_CONTROL, AFMT_GENERIC_CONFLICT,
+			0, 10, max_retries);
+
+	/* HW does is not reading GSP memory not reading too long ->
+	 * something wrong. clear GPS memory access and notify?
+	 * hw SW is writing to GSP memory
+	 */
+	REG_UPDATE(AFMT_VBI_PACKET_CONTROL, AFMT_GENERIC_CONFLICT_CLR, 1);
+
+	/* choose which generic packet to use */
+	regval = REG_READ(AFMT_VBI_PACKET_CONTROL);
+	REG_UPDATE(AFMT_VBI_PACKET_CONTROL,
+			AFMT_GENERIC_INDEX, packet_index);
+
+	/* write generic packet header
+	 * (4th byte is for GENERIC0 only)
+	 */
+	REG_SET_4(AFMT_GENERIC_HDR, 0,
+			AFMT_GENERIC_HB0, info_packet->hb0,
+			AFMT_GENERIC_HB1, info_packet->hb1,
+			AFMT_GENERIC_HB2, info_packet->hb2,
+			AFMT_GENERIC_HB3, info_packet->hb3);
+
+	/* write generic packet contents
+	 * (we never use last 4 bytes)
+	 * there are 8 (0-7) mmDIG0_AFMT_GENERIC0_x registers
+	 */
+	{
+		const uint32_t *content =
+			(const uint32_t *) &info_packet->sb[0];
+
+		REG_WRITE(AFMT_GENERIC_0, *content++);
+		REG_WRITE(AFMT_GENERIC_1, *content++);
+		REG_WRITE(AFMT_GENERIC_2, *content++);
+		REG_WRITE(AFMT_GENERIC_3, *content++);
+		REG_WRITE(AFMT_GENERIC_4, *content++);
+		REG_WRITE(AFMT_GENERIC_5, *content++);
+		REG_WRITE(AFMT_GENERIC_6, *content++);
+		REG_WRITE(AFMT_GENERIC_7, *content);
+	}
+
+	switch (packet_index) {
+	case 0:
+		REG_UPDATE(AFMT_VBI_PACKET_CONTROL1,
+				AFMT_GENERIC0_FRAME_UPDATE, 1);
+		break;
+	case 1:
+		REG_UPDATE(AFMT_VBI_PACKET_CONTROL1,
+				AFMT_GENERIC1_FRAME_UPDATE, 1);
+		break;
+	case 2:
+		REG_UPDATE(AFMT_VBI_PACKET_CONTROL1,
+				AFMT_GENERIC2_FRAME_UPDATE, 1);
+		break;
+	case 3:
+		REG_UPDATE(AFMT_VBI_PACKET_CONTROL1,
+				AFMT_GENERIC3_FRAME_UPDATE, 1);
+		break;
+	case 4:
+		REG_UPDATE(AFMT_VBI_PACKET_CONTROL1,
+				AFMT_GENERIC4_FRAME_UPDATE, 1);
+		break;
+	case 5:
+		REG_UPDATE(AFMT_VBI_PACKET_CONTROL1,
+				AFMT_GENERIC5_FRAME_UPDATE, 1);
+		break;
+	case 6:
+		REG_UPDATE(AFMT_VBI_PACKET_CONTROL1,
+				AFMT_GENERIC6_FRAME_UPDATE, 1);
+		break;
+	case 7:
+		REG_UPDATE(AFMT_VBI_PACKET_CONTROL1,
+				AFMT_GENERIC7_FRAME_UPDATE, 1);
+		break;
+	default:
+		break;
+	}
+}
+
+static void enc1_update_hdmi_info_packet(
+	struct dcn10_stream_encoder *enc1,
+	uint32_t packet_index,
+	const struct dc_info_packet *info_packet)
+{
+	uint32_t cont, send, line;
+
+	if (info_packet->valid) {
+		enc1_update_generic_info_packet(
+			enc1,
+			packet_index,
+			info_packet);
+
+		/* enable transmission of packet(s) -
+		 * packet transmission begins on the next frame
+		 */
+		cont = 1;
+		/* send packet(s) every frame */
+		send = 1;
+		/* select line number to send packets on */
+		line = 2;
+	} else {
+		cont = 0;
+		send = 0;
+		line = 0;
+	}
+
+	/* choose which generic packet control to use */
+	switch (packet_index) {
+	case 0:
+		REG_UPDATE_3(HDMI_GENERIC_PACKET_CONTROL0,
+				HDMI_GENERIC0_CONT, cont,
+				HDMI_GENERIC0_SEND, send,
+				HDMI_GENERIC0_LINE, line);
+		break;
+	case 1:
+		REG_UPDATE_3(HDMI_GENERIC_PACKET_CONTROL0,
+				HDMI_GENERIC1_CONT, cont,
+				HDMI_GENERIC1_SEND, send,
+				HDMI_GENERIC1_LINE, line);
+		break;
+	case 2:
+		REG_UPDATE_3(HDMI_GENERIC_PACKET_CONTROL1,
+				HDMI_GENERIC0_CONT, cont,
+				HDMI_GENERIC0_SEND, send,
+				HDMI_GENERIC0_LINE, line);
+		break;
+	case 3:
+		REG_UPDATE_3(HDMI_GENERIC_PACKET_CONTROL1,
+				HDMI_GENERIC1_CONT, cont,
+				HDMI_GENERIC1_SEND, send,
+				HDMI_GENERIC1_LINE, line);
+		break;
+	case 4:
+		REG_UPDATE_3(HDMI_GENERIC_PACKET_CONTROL2,
+				HDMI_GENERIC0_CONT, cont,
+				HDMI_GENERIC0_SEND, send,
+				HDMI_GENERIC0_LINE, line);
+		break;
+	case 5:
+		REG_UPDATE_3(HDMI_GENERIC_PACKET_CONTROL2,
+				HDMI_GENERIC1_CONT, cont,
+				HDMI_GENERIC1_SEND, send,
+				HDMI_GENERIC1_LINE, line);
+		break;
+	case 6:
+		REG_UPDATE_3(HDMI_GENERIC_PACKET_CONTROL3,
+				HDMI_GENERIC0_CONT, cont,
+				HDMI_GENERIC0_SEND, send,
+				HDMI_GENERIC0_LINE, line);
+		break;
+	case 7:
+		REG_UPDATE_3(HDMI_GENERIC_PACKET_CONTROL3,
+				HDMI_GENERIC1_CONT, cont,
+				HDMI_GENERIC1_SEND, send,
+				HDMI_GENERIC1_LINE, line);
+		break;
+	default:
+		/* invalid HW packet index */
+		DC_LOG_WARNING(
+			"Invalid HW packet index: %s()\n",
+			__func__);
+		return;
+	}
+}
+
+/* setup stream encoder in dp mode */
+static void enc1_stream_encoder_dp_set_stream_attribute(
+	struct stream_encoder *enc,
+	struct dc_crtc_timing *crtc_timing,
+	enum dc_color_space output_color_space)
+{
+	uint32_t h_active_start;
+	uint32_t v_active_start;
+	uint32_t misc0 = 0;
+	uint32_t misc1 = 0;
+	uint32_t h_blank;
+	uint32_t h_back_porch;
+	uint8_t synchronous_clock = 0; /* asynchronous mode */
+	uint8_t colorimetry_bpc;
+	uint8_t dynamic_range_rgb = 0; /*full range*/
+	uint8_t dynamic_range_ycbcr = 1; /*bt709*/
+
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+
+	REG_UPDATE(DP_DB_CNTL, DP_DB_DISABLE, 1);
+
+	/* set pixel encoding */
+	switch (crtc_timing->pixel_encoding) {
+	case PIXEL_ENCODING_YCBCR422:
+		REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING,
+				DP_PIXEL_ENCODING_YCBCR422);
+		break;
+	case PIXEL_ENCODING_YCBCR444:
+		REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING,
+				DP_PIXEL_ENCODING_YCBCR444);
+
+		if (crtc_timing->flags.Y_ONLY)
+			if (crtc_timing->display_color_depth != COLOR_DEPTH_666)
+				/* HW testing only, no use case yet.
+				 * Color depth of Y-only could be
+				 * 8, 10, 12, 16 bits
+				 */
+				REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING,
+						DP_PIXEL_ENCODING_Y_ONLY);
+		/* Note: DP_MSA_MISC1 bit 7 is the indicator
+		 * of Y-only mode.
+		 * This bit is set in HW if register
+		 * DP_PIXEL_ENCODING is programmed to 0x4
+		 */
+		break;
+	case PIXEL_ENCODING_YCBCR420:
+		REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING,
+				DP_PIXEL_ENCODING_YCBCR420);
+		REG_UPDATE(DP_VID_TIMING, DP_VID_N_MUL, 1);
+		break;
+	default:
+		REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING,
+				DP_PIXEL_ENCODING_RGB444);
+		break;
+	}
+
+	misc1 = REG_READ(DP_MSA_MISC);
+
+	/* set color depth */
+
+	switch (crtc_timing->display_color_depth) {
+	case COLOR_DEPTH_666:
+		REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH,
+				0);
+		break;
+	case COLOR_DEPTH_888:
+		REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH,
+				DP_COMPONENT_DEPTH_8BPC);
+		break;
+	case COLOR_DEPTH_101010:
+		REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH,
+				DP_COMPONENT_DEPTH_10BPC);
+
+		break;
+	case COLOR_DEPTH_121212:
+		REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH,
+				DP_COMPONENT_DEPTH_12BPC);
+		break;
+	default:
+		REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH,
+				DP_COMPONENT_DEPTH_6BPC);
+		break;
+	}
+
+	/* set dynamic range and YCbCr range */
+
+	switch (crtc_timing->display_color_depth) {
+	case COLOR_DEPTH_666:
+		colorimetry_bpc = 0;
+		break;
+	case COLOR_DEPTH_888:
+		colorimetry_bpc = 1;
+		break;
+	case COLOR_DEPTH_101010:
+		colorimetry_bpc = 2;
+		break;
+	case COLOR_DEPTH_121212:
+		colorimetry_bpc = 3;
+		break;
+	default:
+		colorimetry_bpc = 0;
+		break;
+	}
+
+	misc0 = misc0 | synchronous_clock;
+	misc0 = colorimetry_bpc << 5;
+
+	switch (output_color_space) {
+	case COLOR_SPACE_SRGB:
+		misc0 = misc0 | 0x0;
+		misc1 = misc1 & ~0x80; /* bit7 = 0*/
+		dynamic_range_rgb = 0; /*full range*/
+		break;
+	case COLOR_SPACE_SRGB_LIMITED:
+		misc0 = misc0 | 0x8; /* bit3=1 */
+		misc1 = misc1 & ~0x80; /* bit7 = 0*/
+		dynamic_range_rgb = 1; /*limited range*/
+		break;
+	case COLOR_SPACE_YCBCR601:
+	case COLOR_SPACE_YCBCR601_LIMITED:
+		misc0 = misc0 | 0x8; /* bit3=1, bit4=0 */
+		misc1 = misc1 & ~0x80; /* bit7 = 0*/
+		dynamic_range_ycbcr = 0; /*bt601*/
+		if (crtc_timing->pixel_encoding == PIXEL_ENCODING_YCBCR422)
+			misc0 = misc0 | 0x2; /* bit2=0, bit1=1 */
+		else if (crtc_timing->pixel_encoding == PIXEL_ENCODING_YCBCR444)
+			misc0 = misc0 | 0x4; /* bit2=1, bit1=0 */
+		break;
+	case COLOR_SPACE_YCBCR709:
+	case COLOR_SPACE_YCBCR709_LIMITED:
+		misc0 = misc0 | 0x18; /* bit3=1, bit4=1 */
+		misc1 = misc1 & ~0x80; /* bit7 = 0*/
+		dynamic_range_ycbcr = 1; /*bt709*/
+		if (crtc_timing->pixel_encoding == PIXEL_ENCODING_YCBCR422)
+			misc0 = misc0 | 0x2; /* bit2=0, bit1=1 */
+		else if (crtc_timing->pixel_encoding == PIXEL_ENCODING_YCBCR444)
+			misc0 = misc0 | 0x4; /* bit2=1, bit1=0 */
+		break;
+	case COLOR_SPACE_2020_RGB_LIMITEDRANGE:
+		dynamic_range_rgb = 1; /*limited range*/
+		break;
+	case COLOR_SPACE_2020_RGB_FULLRANGE:
+	case COLOR_SPACE_2020_YCBCR:
+	case COLOR_SPACE_XR_RGB:
+	case COLOR_SPACE_MSREF_SCRGB:
+	case COLOR_SPACE_ADOBERGB:
+	case COLOR_SPACE_DCIP3:
+	case COLOR_SPACE_XV_YCC_709:
+	case COLOR_SPACE_XV_YCC_601:
+	case COLOR_SPACE_DISPLAYNATIVE:
+	case COLOR_SPACE_DOLBYVISION:
+	case COLOR_SPACE_APPCTRL:
+	case COLOR_SPACE_CUSTOMPOINTS:
+	case COLOR_SPACE_UNKNOWN:
+		/* do nothing */
+		break;
+	}
+
+	REG_SET(DP_MSA_COLORIMETRY, 0, DP_MSA_MISC0, misc0);
+	REG_WRITE(DP_MSA_MISC, misc1);   /* MSA_MISC1 */
+
+	/* dcn new register
+	 * dc_crtc_timing is vesa dmt struct. data from edid
+	 */
+	REG_SET_2(DP_MSA_TIMING_PARAM1, 0,
+			DP_MSA_HTOTAL, crtc_timing->h_total,
+			DP_MSA_VTOTAL, crtc_timing->v_total);
+
+	/* calculate from vesa timing parameters
+	 * h_active_start related to leading edge of sync
+	 */
+
+	h_blank = crtc_timing->h_total - crtc_timing->h_border_left -
+			crtc_timing->h_addressable - crtc_timing->h_border_right;
+
+	h_back_porch = h_blank - crtc_timing->h_front_porch -
+			crtc_timing->h_sync_width;
+
+	/* start at beginning of left border */
+	h_active_start = crtc_timing->h_sync_width + h_back_porch;
+
+
+	v_active_start = crtc_timing->v_total - crtc_timing->v_border_top -
+			crtc_timing->v_addressable - crtc_timing->v_border_bottom -
+			crtc_timing->v_front_porch;
+
+
+	/* start at beginning of left border */
+	REG_SET_2(DP_MSA_TIMING_PARAM2, 0,
+		DP_MSA_HSTART, h_active_start,
+		DP_MSA_VSTART, v_active_start);
+
+	REG_SET_4(DP_MSA_TIMING_PARAM3, 0,
+			DP_MSA_HSYNCWIDTH,
+			crtc_timing->h_sync_width,
+			DP_MSA_HSYNCPOLARITY,
+			!crtc_timing->flags.HSYNC_POSITIVE_POLARITY,
+			DP_MSA_VSYNCWIDTH,
+			crtc_timing->v_sync_width,
+			DP_MSA_VSYNCPOLARITY,
+			!crtc_timing->flags.VSYNC_POSITIVE_POLARITY);
+
+	/* HWDITH include border or overscan */
+	REG_SET_2(DP_MSA_TIMING_PARAM4, 0,
+		DP_MSA_HWIDTH, crtc_timing->h_border_left +
+		crtc_timing->h_addressable + crtc_timing->h_border_right,
+		DP_MSA_VHEIGHT, crtc_timing->v_border_top +
+		crtc_timing->v_addressable + crtc_timing->v_border_bottom);
+}
+
+static void enc1_stream_encoder_set_stream_attribute_helper(
+		struct dcn10_stream_encoder *enc1,
+		struct dc_crtc_timing *crtc_timing)
+{
+	switch (crtc_timing->pixel_encoding) {
+	case PIXEL_ENCODING_YCBCR422:
+		REG_UPDATE(DIG_FE_CNTL, TMDS_PIXEL_ENCODING, 1);
+		break;
+	default:
+		REG_UPDATE(DIG_FE_CNTL, TMDS_PIXEL_ENCODING, 0);
+		break;
+	}
+	REG_UPDATE(DIG_FE_CNTL, TMDS_COLOR_FORMAT, 0);
+}
+
+/* setup stream encoder in hdmi mode */
+static void enc1_stream_encoder_hdmi_set_stream_attribute(
+	struct stream_encoder *enc,
+	struct dc_crtc_timing *crtc_timing,
+	int actual_pix_clk_khz,
+	bool enable_audio)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+	struct bp_encoder_control cntl = {0};
+
+	cntl.action = ENCODER_CONTROL_SETUP;
+	cntl.engine_id = enc1->base.id;
+	cntl.signal = SIGNAL_TYPE_HDMI_TYPE_A;
+	cntl.enable_dp_audio = enable_audio;
+	cntl.pixel_clock = actual_pix_clk_khz;
+	cntl.lanes_number = LANE_COUNT_FOUR;
+
+	if (enc1->base.bp->funcs->encoder_control(
+			enc1->base.bp, &cntl) != BP_RESULT_OK)
+		return;
+
+	enc1_stream_encoder_set_stream_attribute_helper(enc1, crtc_timing);
+
+	/* setup HDMI engine */
+	REG_UPDATE_5(HDMI_CONTROL,
+		HDMI_PACKET_GEN_VERSION, 1,
+		HDMI_KEEPOUT_MODE, 1,
+		HDMI_DEEP_COLOR_ENABLE, 0,
+		HDMI_DATA_SCRAMBLE_EN, 0,
+		HDMI_CLOCK_CHANNEL_RATE, 0);
+
+
+	switch (crtc_timing->display_color_depth) {
+	case COLOR_DEPTH_888:
+		REG_UPDATE(HDMI_CONTROL, HDMI_DEEP_COLOR_DEPTH, 0);
+		break;
+	case COLOR_DEPTH_101010:
+		if (crtc_timing->pixel_encoding == PIXEL_ENCODING_YCBCR422) {
+			REG_UPDATE_2(HDMI_CONTROL,
+					HDMI_DEEP_COLOR_DEPTH, 1,
+					HDMI_DEEP_COLOR_ENABLE, 0);
+		} else {
+			REG_UPDATE_2(HDMI_CONTROL,
+					HDMI_DEEP_COLOR_DEPTH, 1,
+					HDMI_DEEP_COLOR_ENABLE, 1);
+			}
+		break;
+	case COLOR_DEPTH_121212:
+		if (crtc_timing->pixel_encoding == PIXEL_ENCODING_YCBCR422) {
+			REG_UPDATE_2(HDMI_CONTROL,
+					HDMI_DEEP_COLOR_DEPTH, 2,
+					HDMI_DEEP_COLOR_ENABLE, 0);
+		} else {
+			REG_UPDATE_2(HDMI_CONTROL,
+					HDMI_DEEP_COLOR_DEPTH, 2,
+					HDMI_DEEP_COLOR_ENABLE, 1);
+			}
+		break;
+	case COLOR_DEPTH_161616:
+		REG_UPDATE_2(HDMI_CONTROL,
+				HDMI_DEEP_COLOR_DEPTH, 3,
+				HDMI_DEEP_COLOR_ENABLE, 1);
+		break;
+	default:
+		break;
+	}
+
+	if (actual_pix_clk_khz >= HDMI_CLOCK_CHANNEL_RATE_MORE_340M) {
+		/* enable HDMI data scrambler
+		 * HDMI_CLOCK_CHANNEL_RATE_MORE_340M
+		 * Clock channel frequency is 1/4 of character rate.
+		 */
+		REG_UPDATE_2(HDMI_CONTROL,
+			HDMI_DATA_SCRAMBLE_EN, 1,
+			HDMI_CLOCK_CHANNEL_RATE, 1);
+	} else if (crtc_timing->flags.LTE_340MCSC_SCRAMBLE) {
+
+		/* TODO: New feature for DCE11, still need to implement */
+
+		/* enable HDMI data scrambler
+		 * HDMI_CLOCK_CHANNEL_FREQ_EQUAL_TO_CHAR_RATE
+		 * Clock channel frequency is the same
+		 * as character rate
+		 */
+		REG_UPDATE_2(HDMI_CONTROL,
+			HDMI_DATA_SCRAMBLE_EN, 1,
+			HDMI_CLOCK_CHANNEL_RATE, 0);
+	}
+
+
+	REG_UPDATE_3(HDMI_VBI_PACKET_CONTROL,
+		HDMI_GC_CONT, 1,
+		HDMI_GC_SEND, 1,
+		HDMI_NULL_SEND, 1);
+
+	/* following belongs to audio */
+	REG_UPDATE(HDMI_INFOFRAME_CONTROL0, HDMI_AUDIO_INFO_SEND, 1);
+
+	REG_UPDATE(AFMT_INFOFRAME_CONTROL0, AFMT_AUDIO_INFO_UPDATE, 1);
+
+	REG_UPDATE(HDMI_INFOFRAME_CONTROL1, HDMI_AUDIO_INFO_LINE,
+				VBI_LINE_0 + 2);
+
+	REG_UPDATE(HDMI_GC, HDMI_GC_AVMUTE, 0);
+}
+
+/* setup stream encoder in dvi mode */
+static void enc1_stream_encoder_dvi_set_stream_attribute(
+	struct stream_encoder *enc,
+	struct dc_crtc_timing *crtc_timing,
+	bool is_dual_link)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+	struct bp_encoder_control cntl = {0};
+
+	cntl.action = ENCODER_CONTROL_SETUP;
+	cntl.engine_id = enc1->base.id;
+	cntl.signal = is_dual_link ?
+			SIGNAL_TYPE_DVI_DUAL_LINK : SIGNAL_TYPE_DVI_SINGLE_LINK;
+	cntl.enable_dp_audio = false;
+	cntl.pixel_clock = crtc_timing->pix_clk_khz;
+	cntl.lanes_number = (is_dual_link) ? LANE_COUNT_EIGHT : LANE_COUNT_FOUR;
+
+	if (enc1->base.bp->funcs->encoder_control(
+			enc1->base.bp, &cntl) != BP_RESULT_OK)
+		return;
+
+	ASSERT(crtc_timing->pixel_encoding == PIXEL_ENCODING_RGB);
+	ASSERT(crtc_timing->display_color_depth == COLOR_DEPTH_888);
+	enc1_stream_encoder_set_stream_attribute_helper(enc1, crtc_timing);
+}
+
+static void enc1_stream_encoder_set_mst_bandwidth(
+	struct stream_encoder *enc,
+	struct fixed31_32 avg_time_slots_per_mtp)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+	uint32_t x = dal_fixed31_32_floor(
+		avg_time_slots_per_mtp);
+	uint32_t y = dal_fixed31_32_ceil(
+		dal_fixed31_32_shl(
+			dal_fixed31_32_sub_int(
+				avg_time_slots_per_mtp,
+				x),
+			26));
+
+	REG_SET_2(DP_MSE_RATE_CNTL, 0,
+		DP_MSE_RATE_X, x,
+		DP_MSE_RATE_Y, y);
+
+	/* wait for update to be completed on the link */
+	/* i.e. DP_MSE_RATE_UPDATE_PENDING field (read only) */
+	/* is reset to 0 (not pending) */
+	REG_WAIT(DP_MSE_RATE_UPDATE, DP_MSE_RATE_UPDATE_PENDING,
+			0,
+			10, DP_MST_UPDATE_MAX_RETRY);
+}
+
+static void enc1_stream_encoder_update_hdmi_info_packets(
+	struct stream_encoder *enc,
+	const struct encoder_info_frame *info_frame)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+
+	/* for bring up, disable dp double  TODO */
+	REG_UPDATE(HDMI_DB_CONTROL, HDMI_DB_DISABLE, 1);
+
+	enc1_update_hdmi_info_packet(enc1, 0, &info_frame->avi);
+	enc1_update_hdmi_info_packet(enc1, 1, &info_frame->vendor);
+	enc1_update_hdmi_info_packet(enc1, 2, &info_frame->gamut);
+	enc1_update_hdmi_info_packet(enc1, 3, &info_frame->spd);
+	enc1_update_hdmi_info_packet(enc1, 4, &info_frame->hdrsmd);
+}
+
+static void enc1_stream_encoder_stop_hdmi_info_packets(
+	struct stream_encoder *enc)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+
+	/* stop generic packets 0 & 1 on HDMI */
+	REG_SET_6(HDMI_GENERIC_PACKET_CONTROL0, 0,
+		HDMI_GENERIC1_CONT, 0,
+		HDMI_GENERIC1_LINE, 0,
+		HDMI_GENERIC1_SEND, 0,
+		HDMI_GENERIC0_CONT, 0,
+		HDMI_GENERIC0_LINE, 0,
+		HDMI_GENERIC0_SEND, 0);
+
+	/* stop generic packets 2 & 3 on HDMI */
+	REG_SET_6(HDMI_GENERIC_PACKET_CONTROL1, 0,
+		HDMI_GENERIC0_CONT, 0,
+		HDMI_GENERIC0_LINE, 0,
+		HDMI_GENERIC0_SEND, 0,
+		HDMI_GENERIC1_CONT, 0,
+		HDMI_GENERIC1_LINE, 0,
+		HDMI_GENERIC1_SEND, 0);
+
+	/* stop generic packets 2 & 3 on HDMI */
+	REG_SET_6(HDMI_GENERIC_PACKET_CONTROL2, 0,
+		HDMI_GENERIC0_CONT, 0,
+		HDMI_GENERIC0_LINE, 0,
+		HDMI_GENERIC0_SEND, 0,
+		HDMI_GENERIC1_CONT, 0,
+		HDMI_GENERIC1_LINE, 0,
+		HDMI_GENERIC1_SEND, 0);
+
+	REG_SET_6(HDMI_GENERIC_PACKET_CONTROL3, 0,
+		HDMI_GENERIC0_CONT, 0,
+		HDMI_GENERIC0_LINE, 0,
+		HDMI_GENERIC0_SEND, 0,
+		HDMI_GENERIC1_CONT, 0,
+		HDMI_GENERIC1_LINE, 0,
+		HDMI_GENERIC1_SEND, 0);
+}
+
+static void enc1_stream_encoder_update_dp_info_packets(
+	struct stream_encoder *enc,
+	const struct encoder_info_frame *info_frame)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+	uint32_t value = REG_READ(DP_SEC_CNTL);
+
+	if (info_frame->vsc.valid)
+		enc1_update_generic_info_packet(
+					enc1,
+					0,  /* packetIndex */
+					&info_frame->vsc);
+
+	if (info_frame->spd.valid)
+		enc1_update_generic_info_packet(
+				enc1,
+				2,  /* packetIndex */
+				&info_frame->spd);
+
+	if (info_frame->hdrsmd.valid)
+		enc1_update_generic_info_packet(
+				enc1,
+				3,  /* packetIndex */
+				&info_frame->hdrsmd);
+
+	/* enable/disable transmission of packet(s).
+	 * If enabled, packet transmission begins on the next frame
+	 */
+	REG_UPDATE(DP_SEC_CNTL, DP_SEC_GSP0_ENABLE, info_frame->vsc.valid);
+	REG_UPDATE(DP_SEC_CNTL, DP_SEC_GSP2_ENABLE, info_frame->spd.valid);
+	REG_UPDATE(DP_SEC_CNTL, DP_SEC_GSP3_ENABLE, info_frame->hdrsmd.valid);
+
+	/* This bit is the master enable bit.
+	 * When enabling secondary stream engine,
+	 * this master bit must also be set.
+	 * This register shared with audio info frame.
+	 * Therefore we need to enable master bit
+	 * if at least on of the fields is not 0
+	 */
+	if (value)
+		REG_UPDATE(DP_SEC_CNTL, DP_SEC_STREAM_ENABLE, 1);
+}
+
+static void enc1_stream_encoder_stop_dp_info_packets(
+	struct stream_encoder *enc)
+{
+	/* stop generic packets on DP */
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+	uint32_t value = REG_READ(DP_SEC_CNTL);
+
+	REG_SET_10(DP_SEC_CNTL, 0,
+		DP_SEC_GSP0_ENABLE, 0,
+		DP_SEC_GSP1_ENABLE, 0,
+		DP_SEC_GSP2_ENABLE, 0,
+		DP_SEC_GSP3_ENABLE, 0,
+		DP_SEC_GSP4_ENABLE, 0,
+		DP_SEC_GSP5_ENABLE, 0,
+		DP_SEC_GSP6_ENABLE, 0,
+		DP_SEC_GSP7_ENABLE, 0,
+		DP_SEC_MPG_ENABLE, 0,
+		DP_SEC_STREAM_ENABLE, 0);
+
+	/* this register shared with audio info frame.
+	 * therefore we need to keep master enabled
+	 * if at least one of the fields is not 0 */
+
+	if (value)
+		REG_UPDATE(DP_SEC_CNTL, DP_SEC_STREAM_ENABLE, 1);
+
+}
+
+static void enc1_stream_encoder_dp_blank(
+	struct stream_encoder *enc)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+	uint32_t retries = 0;
+	uint32_t  reg1 = 0;
+	uint32_t max_retries = DP_BLANK_MAX_RETRY * 10;
+
+	/* Note: For CZ, we are changing driver default to disable
+	 * stream deferred to next VBLANK. If results are positive, we
+	 * will make the same change to all DCE versions. There are a
+	 * handful of panels that cannot handle disable stream at
+	 * HBLANK and will result in a white line flash across the
+	 * screen on stream disable.
+	 */
+	REG_GET(DP_VID_STREAM_CNTL, DP_VID_STREAM_ENABLE, &reg1);
+	if ((reg1 & 0x1) == 0)
+		/*stream not enabled*/
+		return;
+	/* Specify the video stream disable point
+	 * (2 = start of the next vertical blank)
+	 */
+	REG_UPDATE(DP_VID_STREAM_CNTL, DP_VID_STREAM_DIS_DEFER, 2);
+	/* Larger delay to wait until VBLANK - use max retry of
+	 * 10us*3000=30ms. This covers 16.6ms of typical 60 Hz mode +
+	 * a little more because we may not trust delay accuracy.
+	 */
+	max_retries = DP_BLANK_MAX_RETRY * 150;
+
+	/* disable DP stream */
+	REG_UPDATE(DP_VID_STREAM_CNTL, DP_VID_STREAM_ENABLE, 0);
+
+	/* the encoder stops sending the video stream
+	 * at the start of the vertical blanking.
+	 * Poll for DP_VID_STREAM_STATUS == 0
+	 */
+
+	REG_WAIT(DP_VID_STREAM_CNTL, DP_VID_STREAM_STATUS,
+			0,
+			10, max_retries);
+
+	ASSERT(retries <= max_retries);
+
+	/* Tell the DP encoder to ignore timing from CRTC, must be done after
+	 * the polling. If we set DP_STEER_FIFO_RESET before DP stream blank is
+	 * complete, stream status will be stuck in video stream enabled state,
+	 * i.e. DP_VID_STREAM_STATUS stuck at 1.
+	 */
+
+	REG_UPDATE(DP_STEER_FIFO, DP_STEER_FIFO_RESET, true);
+}
+
+/* output video stream to link encoder */
+static void enc1_stream_encoder_dp_unblank(
+	struct stream_encoder *enc,
+	const struct encoder_unblank_param *param)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+
+	if (param->link_settings.link_rate != LINK_RATE_UNKNOWN) {
+		uint32_t n_vid = 0x8000;
+		uint32_t m_vid;
+
+		/* M / N = Fstream / Flink
+		 * m_vid / n_vid = pixel rate / link rate
+		 */
+
+		uint64_t m_vid_l = n_vid;
+
+		m_vid_l *= param->pixel_clk_khz;
+		m_vid_l = div_u64(m_vid_l,
+			param->link_settings.link_rate
+				* LINK_RATE_REF_FREQ_IN_KHZ);
+
+		m_vid = (uint32_t) m_vid_l;
+
+		/* enable auto measurement */
+
+		REG_UPDATE(DP_VID_TIMING, DP_VID_M_N_GEN_EN, 0);
+
+		/* auto measurement need 1 full 0x8000 symbol cycle to kick in,
+		 * therefore program initial value for Mvid and Nvid
+		 */
+
+		REG_UPDATE(DP_VID_N, DP_VID_N, n_vid);
+
+		REG_UPDATE(DP_VID_M, DP_VID_M, m_vid);
+
+		REG_UPDATE(DP_VID_TIMING, DP_VID_M_N_GEN_EN, 1);
+	}
+
+	/* set DIG_START to 0x1 to resync FIFO */
+
+	REG_UPDATE(DIG_FE_CNTL, DIG_START, 1);
+
+	/* switch DP encoder to CRTC data */
+
+	REG_UPDATE(DP_STEER_FIFO, DP_STEER_FIFO_RESET, 0);
+
+	/* wait 100us for DIG/DP logic to prime
+	 * (i.e. a few video lines)
+	 */
+	udelay(100);
+
+	/* the hardware would start sending video at the start of the next DP
+	 * frame (i.e. rising edge of the vblank).
+	 * NOTE: We used to program DP_VID_STREAM_DIS_DEFER = 2 here, but this
+	 * register has no effect on enable transition! HW always guarantees
+	 * VID_STREAM enable at start of next frame, and this is not
+	 * programmable
+	 */
+
+	REG_UPDATE(DP_VID_STREAM_CNTL, DP_VID_STREAM_ENABLE, true);
+}
+
+static void enc1_stream_encoder_set_avmute(
+	struct stream_encoder *enc,
+	bool enable)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+	unsigned int value = enable ? 1 : 0;
+
+	REG_UPDATE(HDMI_GC, HDMI_GC_AVMUTE, value);
+}
+
+
+#define DP_SEC_AUD_N__DP_SEC_AUD_N__DEFAULT 0x8000
+#define DP_SEC_TIMESTAMP__DP_SEC_TIMESTAMP_MODE__AUTO_CALC 1
+
+#include "include/audio_types.h"
+
+/**
+* speakersToChannels
+*
+* @brief
+*  translate speakers to channels
+*
+*  FL  - Front Left
+*  FR  - Front Right
+*  RL  - Rear Left
+*  RR  - Rear Right
+*  RC  - Rear Center
+*  FC  - Front Center
+*  FLC - Front Left Center
+*  FRC - Front Right Center
+*  RLC - Rear Left Center
+*  RRC - Rear Right Center
+*  LFE - Low Freq Effect
+*
+*               FC
+*          FLC      FRC
+*    FL                    FR
+*
+*                    LFE
+*              ()
+*
+*
+*    RL                    RR
+*          RLC      RRC
+*               RC
+*
+*             ch  8   7   6   5   4   3   2   1
+* 0b00000011      -   -   -   -   -   -   FR  FL
+* 0b00000111      -   -   -   -   -   LFE FR  FL
+* 0b00001011      -   -   -   -   FC  -   FR  FL
+* 0b00001111      -   -   -   -   FC  LFE FR  FL
+* 0b00010011      -   -   -   RC  -   -   FR  FL
+* 0b00010111      -   -   -   RC  -   LFE FR  FL
+* 0b00011011      -   -   -   RC  FC  -   FR  FL
+* 0b00011111      -   -   -   RC  FC  LFE FR  FL
+* 0b00110011      -   -   RR  RL  -   -   FR  FL
+* 0b00110111      -   -   RR  RL  -   LFE FR  FL
+* 0b00111011      -   -   RR  RL  FC  -   FR  FL
+* 0b00111111      -   -   RR  RL  FC  LFE FR  FL
+* 0b01110011      -   RC  RR  RL  -   -   FR  FL
+* 0b01110111      -   RC  RR  RL  -   LFE FR  FL
+* 0b01111011      -   RC  RR  RL  FC  -   FR  FL
+* 0b01111111      -   RC  RR  RL  FC  LFE FR  FL
+* 0b11110011      RRC RLC RR  RL  -   -   FR  FL
+* 0b11110111      RRC RLC RR  RL  -   LFE FR  FL
+* 0b11111011      RRC RLC RR  RL  FC  -   FR  FL
+* 0b11111111      RRC RLC RR  RL  FC  LFE FR  FL
+* 0b11000011      FRC FLC -   -   -   -   FR  FL
+* 0b11000111      FRC FLC -   -   -   LFE FR  FL
+* 0b11001011      FRC FLC -   -   FC  -   FR  FL
+* 0b11001111      FRC FLC -   -   FC  LFE FR  FL
+* 0b11010011      FRC FLC -   RC  -   -   FR  FL
+* 0b11010111      FRC FLC -   RC  -   LFE FR  FL
+* 0b11011011      FRC FLC -   RC  FC  -   FR  FL
+* 0b11011111      FRC FLC -   RC  FC  LFE FR  FL
+* 0b11110011      FRC FLC RR  RL  -   -   FR  FL
+* 0b11110111      FRC FLC RR  RL  -   LFE FR  FL
+* 0b11111011      FRC FLC RR  RL  FC  -   FR  FL
+* 0b11111111      FRC FLC RR  RL  FC  LFE FR  FL
+*
+* @param
+*  speakers - speaker information as it comes from CEA audio block
+*/
+/* translate speakers to channels */
+
+union audio_cea_channels {
+	uint8_t all;
+	struct audio_cea_channels_bits {
+		uint32_t FL:1;
+		uint32_t FR:1;
+		uint32_t LFE:1;
+		uint32_t FC:1;
+		uint32_t RL_RC:1;
+		uint32_t RR:1;
+		uint32_t RC_RLC_FLC:1;
+		uint32_t RRC_FRC:1;
+	} channels;
+};
+
+struct audio_clock_info {
+	/* pixel clock frequency*/
+	uint32_t pixel_clock_in_10khz;
+	/* N - 32KHz audio */
+	uint32_t n_32khz;
+	/* CTS - 32KHz audio*/
+	uint32_t cts_32khz;
+	uint32_t n_44khz;
+	uint32_t cts_44khz;
+	uint32_t n_48khz;
+	uint32_t cts_48khz;
+};
+
+/* 25.2MHz/1.001*/
+/* 25.2MHz/1.001*/
+/* 25.2MHz*/
+/* 27MHz */
+/* 27MHz*1.001*/
+/* 27MHz*1.001*/
+/* 54MHz*/
+/* 54MHz*1.001*/
+/* 74.25MHz/1.001*/
+/* 74.25MHz*/
+/* 148.5MHz/1.001*/
+/* 148.5MHz*/
+
+static const struct audio_clock_info audio_clock_info_table[16] = {
+	{2517, 4576, 28125, 7007, 31250, 6864, 28125},
+	{2518, 4576, 28125, 7007, 31250, 6864, 28125},
+	{2520, 4096, 25200, 6272, 28000, 6144, 25200},
+	{2700, 4096, 27000, 6272, 30000, 6144, 27000},
+	{2702, 4096, 27027, 6272, 30030, 6144, 27027},
+	{2703, 4096, 27027, 6272, 30030, 6144, 27027},
+	{5400, 4096, 54000, 6272, 60000, 6144, 54000},
+	{5405, 4096, 54054, 6272, 60060, 6144, 54054},
+	{7417, 11648, 210937, 17836, 234375, 11648, 140625},
+	{7425, 4096, 74250, 6272, 82500, 6144, 74250},
+	{14835, 11648, 421875, 8918, 234375, 5824, 140625},
+	{14850, 4096, 148500, 6272, 165000, 6144, 148500},
+	{29670, 5824, 421875, 4459, 234375, 5824, 281250},
+	{29700, 3072, 222750, 4704, 247500, 5120, 247500},
+	{59340, 5824, 843750, 8918, 937500, 5824, 562500},
+	{59400, 3072, 445500, 9408, 990000, 6144, 594000}
+};
+
+static const struct audio_clock_info audio_clock_info_table_36bpc[14] = {
+	{2517,  9152,  84375,  7007,  48875,  9152,  56250},
+	{2518,  9152,  84375,  7007,  48875,  9152,  56250},
+	{2520,  4096,  37800,  6272,  42000,  6144,  37800},
+	{2700,  4096,  40500,  6272,  45000,  6144,  40500},
+	{2702,  8192,  81081,  6272,  45045,  8192,  54054},
+	{2703,  8192,  81081,  6272,  45045,  8192,  54054},
+	{5400,  4096,  81000,  6272,  90000,  6144,  81000},
+	{5405,  4096,  81081,  6272,  90090,  6144,  81081},
+	{7417, 11648, 316406, 17836, 351562, 11648, 210937},
+	{7425, 4096, 111375,  6272, 123750,  6144, 111375},
+	{14835, 11648, 632812, 17836, 703125, 11648, 421875},
+	{14850, 4096, 222750,  6272, 247500,  6144, 222750},
+	{29670, 5824, 632812,  8918, 703125,  5824, 421875},
+	{29700, 4096, 445500,  4704, 371250,  5120, 371250}
+};
+
+static const struct audio_clock_info audio_clock_info_table_48bpc[14] = {
+	{2517,  4576,  56250,  7007,  62500,  6864,  56250},
+	{2518,  4576,  56250,  7007,  62500,  6864,  56250},
+	{2520,  4096,  50400,  6272,  56000,  6144,  50400},
+	{2700,  4096,  54000,  6272,  60000,  6144,  54000},
+	{2702,  4096,  54054,  6267,  60060,  8192,  54054},
+	{2703,  4096,  54054,  6272,  60060,  8192,  54054},
+	{5400,  4096, 108000,  6272, 120000,  6144, 108000},
+	{5405,  4096, 108108,  6272, 120120,  6144, 108108},
+	{7417, 11648, 421875, 17836, 468750, 11648, 281250},
+	{7425,  4096, 148500,  6272, 165000,  6144, 148500},
+	{14835, 11648, 843750,  8918, 468750, 11648, 281250},
+	{14850, 4096, 297000,  6272, 330000,  6144, 297000},
+	{29670, 5824, 843750,  4459, 468750,  5824, 562500},
+	{29700, 3072, 445500,  4704, 495000,  5120, 495000}
+
+
+};
+
+static union audio_cea_channels speakers_to_channels(
+	struct audio_speaker_flags speaker_flags)
+{
+	union audio_cea_channels cea_channels = {0};
+
+	/* these are one to one */
+	cea_channels.channels.FL = speaker_flags.FL_FR;
+	cea_channels.channels.FR = speaker_flags.FL_FR;
+	cea_channels.channels.LFE = speaker_flags.LFE;
+	cea_channels.channels.FC = speaker_flags.FC;
+
+	/* if Rear Left and Right exist move RC speaker to channel 7
+	 * otherwise to channel 5
+	 */
+	if (speaker_flags.RL_RR) {
+		cea_channels.channels.RL_RC = speaker_flags.RL_RR;
+		cea_channels.channels.RR = speaker_flags.RL_RR;
+		cea_channels.channels.RC_RLC_FLC = speaker_flags.RC;
+	} else {
+		cea_channels.channels.RL_RC = speaker_flags.RC;
+	}
+
+	/* FRONT Left Right Center and REAR Left Right Center are exclusive */
+	if (speaker_flags.FLC_FRC) {
+		cea_channels.channels.RC_RLC_FLC = speaker_flags.FLC_FRC;
+		cea_channels.channels.RRC_FRC = speaker_flags.FLC_FRC;
+	} else {
+		cea_channels.channels.RC_RLC_FLC = speaker_flags.RLC_RRC;
+		cea_channels.channels.RRC_FRC = speaker_flags.RLC_RRC;
+	}
+
+	return cea_channels;
+}
+
+static uint32_t calc_max_audio_packets_per_line(
+	const struct audio_crtc_info *crtc_info)
+{
+	uint32_t max_packets_per_line;
+
+	max_packets_per_line =
+		crtc_info->h_total - crtc_info->h_active;
+
+	if (crtc_info->pixel_repetition)
+		max_packets_per_line *= crtc_info->pixel_repetition;
+
+	/* for other hdmi features */
+	max_packets_per_line -= 58;
+	/* for Control Period */
+	max_packets_per_line -= 16;
+	/* Number of Audio Packets per Line */
+	max_packets_per_line /= 32;
+
+	return max_packets_per_line;
+}
+
+static void get_audio_clock_info(
+	enum dc_color_depth color_depth,
+	uint32_t crtc_pixel_clock_in_khz,
+	uint32_t actual_pixel_clock_in_khz,
+	struct audio_clock_info *audio_clock_info)
+{
+	const struct audio_clock_info *clock_info;
+	uint32_t index;
+	uint32_t crtc_pixel_clock_in_10khz = crtc_pixel_clock_in_khz / 10;
+	uint32_t audio_array_size;
+
+	switch (color_depth) {
+	case COLOR_DEPTH_161616:
+		clock_info = audio_clock_info_table_48bpc;
+		audio_array_size = ARRAY_SIZE(
+				audio_clock_info_table_48bpc);
+		break;
+	case COLOR_DEPTH_121212:
+		clock_info = audio_clock_info_table_36bpc;
+		audio_array_size = ARRAY_SIZE(
+				audio_clock_info_table_36bpc);
+		break;
+	default:
+		clock_info = audio_clock_info_table;
+		audio_array_size = ARRAY_SIZE(
+				audio_clock_info_table);
+		break;
+	}
+
+	if (clock_info != NULL) {
+		/* search for exact pixel clock in table */
+		for (index = 0; index < audio_array_size; index++) {
+			if (clock_info[index].pixel_clock_in_10khz >
+				crtc_pixel_clock_in_10khz)
+				break;  /* not match */
+			else if (clock_info[index].pixel_clock_in_10khz ==
+					crtc_pixel_clock_in_10khz) {
+				/* match found */
+				*audio_clock_info = clock_info[index];
+				return;
+			}
+		}
+	}
+
+	/* not found */
+	if (actual_pixel_clock_in_khz == 0)
+		actual_pixel_clock_in_khz = crtc_pixel_clock_in_khz;
+
+	/* See HDMI spec  the table entry under
+	 *  pixel clock of "Other". */
+	audio_clock_info->pixel_clock_in_10khz =
+			actual_pixel_clock_in_khz / 10;
+	audio_clock_info->cts_32khz = actual_pixel_clock_in_khz;
+	audio_clock_info->cts_44khz = actual_pixel_clock_in_khz;
+	audio_clock_info->cts_48khz = actual_pixel_clock_in_khz;
+
+	audio_clock_info->n_32khz = 4096;
+	audio_clock_info->n_44khz = 6272;
+	audio_clock_info->n_48khz = 6144;
+}
+
+static void enc1_se_audio_setup(
+	struct stream_encoder *enc,
+	unsigned int az_inst,
+	struct audio_info *audio_info)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+
+	uint32_t speakers = 0;
+	uint32_t channels = 0;
+
+	ASSERT(audio_info);
+	if (audio_info == NULL)
+		/* This should not happen.it does so we don't get BSOD*/
+		return;
+
+	speakers = audio_info->flags.info.ALLSPEAKERS;
+	channels = speakers_to_channels(audio_info->flags.speaker_flags).all;
+
+	/* setup the audio stream source select (audio -> dig mapping) */
+	REG_SET(AFMT_AUDIO_SRC_CONTROL, 0, AFMT_AUDIO_SRC_SELECT, az_inst);
+
+	/* Channel allocation */
+	REG_UPDATE(AFMT_AUDIO_PACKET_CONTROL2, AFMT_AUDIO_CHANNEL_ENABLE, channels);
+}
+
+static void enc1_se_setup_hdmi_audio(
+	struct stream_encoder *enc,
+	const struct audio_crtc_info *crtc_info)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+
+	struct audio_clock_info audio_clock_info = {0};
+	uint32_t max_packets_per_line;
+
+	/* For now still do calculation, although this field is ignored when
+	 * above HDMI_PACKET_GEN_VERSION set to 1
+	 */
+	max_packets_per_line = calc_max_audio_packets_per_line(crtc_info);
+
+	/* HDMI_AUDIO_PACKET_CONTROL */
+	REG_UPDATE_2(HDMI_AUDIO_PACKET_CONTROL,
+			HDMI_AUDIO_PACKETS_PER_LINE, max_packets_per_line,
+			HDMI_AUDIO_DELAY_EN, 1);
+
+	/* AFMT_AUDIO_PACKET_CONTROL */
+	REG_UPDATE(AFMT_AUDIO_PACKET_CONTROL, AFMT_60958_CS_UPDATE, 1);
+
+	/* AFMT_AUDIO_PACKET_CONTROL2 */
+	REG_UPDATE_2(AFMT_AUDIO_PACKET_CONTROL2,
+			AFMT_AUDIO_LAYOUT_OVRD, 0,
+			AFMT_60958_OSF_OVRD, 0);
+
+	/* HDMI_ACR_PACKET_CONTROL */
+	REG_UPDATE_3(HDMI_ACR_PACKET_CONTROL,
+			HDMI_ACR_AUTO_SEND, 1,
+			HDMI_ACR_SOURCE, 0,
+			HDMI_ACR_AUDIO_PRIORITY, 0);
+
+	/* Program audio clock sample/regeneration parameters */
+	get_audio_clock_info(crtc_info->color_depth,
+			     crtc_info->requested_pixel_clock,
+			     crtc_info->calculated_pixel_clock,
+			     &audio_clock_info);
+	DC_LOG_HW_AUDIO(
+			"\n%s:Input::requested_pixel_clock = %d"	\
+			"calculated_pixel_clock = %d \n", __func__,	\
+			crtc_info->requested_pixel_clock,		\
+			crtc_info->calculated_pixel_clock);
+
+	/* HDMI_ACR_32_0__HDMI_ACR_CTS_32_MASK */
+	REG_UPDATE(HDMI_ACR_32_0, HDMI_ACR_CTS_32, audio_clock_info.cts_32khz);
+
+	/* HDMI_ACR_32_1__HDMI_ACR_N_32_MASK */
+	REG_UPDATE(HDMI_ACR_32_1, HDMI_ACR_N_32, audio_clock_info.n_32khz);
+
+	/* HDMI_ACR_44_0__HDMI_ACR_CTS_44_MASK */
+	REG_UPDATE(HDMI_ACR_44_0, HDMI_ACR_CTS_44, audio_clock_info.cts_44khz);
+
+	/* HDMI_ACR_44_1__HDMI_ACR_N_44_MASK */
+	REG_UPDATE(HDMI_ACR_44_1, HDMI_ACR_N_44, audio_clock_info.n_44khz);
+
+	/* HDMI_ACR_48_0__HDMI_ACR_CTS_48_MASK */
+	REG_UPDATE(HDMI_ACR_48_0, HDMI_ACR_CTS_48, audio_clock_info.cts_48khz);
+
+	/* HDMI_ACR_48_1__HDMI_ACR_N_48_MASK */
+	REG_UPDATE(HDMI_ACR_48_1, HDMI_ACR_N_48, audio_clock_info.n_48khz);
+
+	/* Video driver cannot know in advance which sample rate will
+	 * be used by HD Audio driver
+	 * HDMI_ACR_PACKET_CONTROL__HDMI_ACR_N_MULTIPLE field is
+	 * programmed below in interruppt callback
+	 */
+
+	/* AFMT_60958_0__AFMT_60958_CS_CHANNEL_NUMBER_L_MASK &
+	 * AFMT_60958_0__AFMT_60958_CS_CLOCK_ACCURACY_MASK
+	 */
+	REG_UPDATE_2(AFMT_60958_0,
+			AFMT_60958_CS_CHANNEL_NUMBER_L, 1,
+			AFMT_60958_CS_CLOCK_ACCURACY, 0);
+
+	/* AFMT_60958_1 AFMT_60958_CS_CHALNNEL_NUMBER_R */
+	REG_UPDATE(AFMT_60958_1, AFMT_60958_CS_CHANNEL_NUMBER_R, 2);
+
+	/* AFMT_60958_2 now keep this settings until
+	 * Programming guide comes out
+	 */
+	REG_UPDATE_6(AFMT_60958_2,
+			AFMT_60958_CS_CHANNEL_NUMBER_2, 3,
+			AFMT_60958_CS_CHANNEL_NUMBER_3, 4,
+			AFMT_60958_CS_CHANNEL_NUMBER_4, 5,
+			AFMT_60958_CS_CHANNEL_NUMBER_5, 6,
+			AFMT_60958_CS_CHANNEL_NUMBER_6, 7,
+			AFMT_60958_CS_CHANNEL_NUMBER_7, 8);
+}
+
+static void enc1_se_setup_dp_audio(
+	struct stream_encoder *enc)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+
+	/* --- DP Audio packet configurations --- */
+
+	/* ATP Configuration */
+	REG_SET(DP_SEC_AUD_N, 0,
+			DP_SEC_AUD_N, DP_SEC_AUD_N__DP_SEC_AUD_N__DEFAULT);
+
+	/* Async/auto-calc timestamp mode */
+	REG_SET(DP_SEC_TIMESTAMP, 0, DP_SEC_TIMESTAMP_MODE,
+			DP_SEC_TIMESTAMP__DP_SEC_TIMESTAMP_MODE__AUTO_CALC);
+
+	/* --- The following are the registers
+	 *  copied from the SetupHDMI ---
+	 */
+
+	/* AFMT_AUDIO_PACKET_CONTROL */
+	REG_UPDATE(AFMT_AUDIO_PACKET_CONTROL, AFMT_60958_CS_UPDATE, 1);
+
+	/* AFMT_AUDIO_PACKET_CONTROL2 */
+	/* Program the ATP and AIP next */
+	REG_UPDATE_2(AFMT_AUDIO_PACKET_CONTROL2,
+			AFMT_AUDIO_LAYOUT_OVRD, 0,
+			AFMT_60958_OSF_OVRD, 0);
+
+	/* AFMT_INFOFRAME_CONTROL0 */
+	REG_UPDATE(AFMT_INFOFRAME_CONTROL0, AFMT_AUDIO_INFO_UPDATE, 1);
+
+	/* AFMT_60958_0__AFMT_60958_CS_CLOCK_ACCURACY_MASK */
+	REG_UPDATE(AFMT_60958_0, AFMT_60958_CS_CLOCK_ACCURACY, 0);
+}
+
+static void enc1_se_enable_audio_clock(
+	struct stream_encoder *enc,
+	bool enable)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+
+	if (REG(AFMT_CNTL) == 0)
+		return;   /* DCE8/10 does not have this register */
+
+	REG_UPDATE(AFMT_CNTL, AFMT_AUDIO_CLOCK_EN, !!enable);
+
+	/* wait for AFMT clock to turn on,
+	 * expectation: this should complete in 1-2 reads
+	 *
+	 * REG_WAIT(AFMT_CNTL, AFMT_AUDIO_CLOCK_ON, !!enable, 1, 10);
+	 *
+	 * TODO: wait for clock_on does not work well. May need HW
+	 * program sequence. But audio seems work normally even without wait
+	 * for clock_on status change
+	 */
+}
+
+static void enc1_se_enable_dp_audio(
+	struct stream_encoder *enc)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+
+	/* Enable Audio packets */
+	REG_UPDATE(DP_SEC_CNTL, DP_SEC_ASP_ENABLE, 1);
+
+	/* Program the ATP and AIP next */
+	REG_UPDATE_2(DP_SEC_CNTL,
+			DP_SEC_ATP_ENABLE, 1,
+			DP_SEC_AIP_ENABLE, 1);
+
+	/* Program STREAM_ENABLE after all the other enables. */
+	REG_UPDATE(DP_SEC_CNTL, DP_SEC_STREAM_ENABLE, 1);
+}
+
+static void enc1_se_disable_dp_audio(
+	struct stream_encoder *enc)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+	uint32_t value = REG_READ(DP_SEC_CNTL);
+
+	/* Disable Audio packets */
+	REG_UPDATE_5(DP_SEC_CNTL,
+			DP_SEC_ASP_ENABLE, 0,
+			DP_SEC_ATP_ENABLE, 0,
+			DP_SEC_AIP_ENABLE, 0,
+			DP_SEC_ACM_ENABLE, 0,
+			DP_SEC_STREAM_ENABLE, 0);
+
+	/* This register shared with encoder info frame. Therefore we need to
+	 * keep master enabled if at least on of the fields is not 0
+	 */
+	if (value != 0)
+		REG_UPDATE(DP_SEC_CNTL, DP_SEC_STREAM_ENABLE, 1);
+
+}
+
+void enc1_se_audio_mute_control(
+	struct stream_encoder *enc,
+	bool mute)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+
+	REG_UPDATE(AFMT_AUDIO_PACKET_CONTROL, AFMT_AUDIO_SAMPLE_SEND, !mute);
+}
+
+void enc1_se_dp_audio_setup(
+	struct stream_encoder *enc,
+	unsigned int az_inst,
+	struct audio_info *info)
+{
+	enc1_se_audio_setup(enc, az_inst, info);
+}
+
+void enc1_se_dp_audio_enable(
+	struct stream_encoder *enc)
+{
+	enc1_se_enable_audio_clock(enc, true);
+	enc1_se_setup_dp_audio(enc);
+	enc1_se_enable_dp_audio(enc);
+}
+
+void enc1_se_dp_audio_disable(
+	struct stream_encoder *enc)
+{
+	enc1_se_disable_dp_audio(enc);
+	enc1_se_enable_audio_clock(enc, false);
+}
+
+void enc1_se_hdmi_audio_setup(
+	struct stream_encoder *enc,
+	unsigned int az_inst,
+	struct audio_info *info,
+	struct audio_crtc_info *audio_crtc_info)
+{
+	enc1_se_enable_audio_clock(enc, true);
+	enc1_se_setup_hdmi_audio(enc, audio_crtc_info);
+	enc1_se_audio_setup(enc, az_inst, info);
+}
+
+void enc1_se_hdmi_audio_disable(
+	struct stream_encoder *enc)
+{
+	enc1_se_enable_audio_clock(enc, false);
+}
+
+
+static void enc1_setup_stereo_sync(
+	struct stream_encoder *enc,
+	int tg_inst, bool enable)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+	REG_UPDATE(DIG_FE_CNTL, DIG_STEREOSYNC_SELECT, tg_inst);
+	REG_UPDATE(DIG_FE_CNTL, DIG_STEREOSYNC_GATE_EN, !enable);
+}
+
+
+static const struct stream_encoder_funcs dcn10_str_enc_funcs = {
+	.dp_set_stream_attribute =
+		enc1_stream_encoder_dp_set_stream_attribute,
+	.hdmi_set_stream_attribute =
+		enc1_stream_encoder_hdmi_set_stream_attribute,
+	.dvi_set_stream_attribute =
+		enc1_stream_encoder_dvi_set_stream_attribute,
+	.set_mst_bandwidth =
+		enc1_stream_encoder_set_mst_bandwidth,
+	.update_hdmi_info_packets =
+		enc1_stream_encoder_update_hdmi_info_packets,
+	.stop_hdmi_info_packets =
+		enc1_stream_encoder_stop_hdmi_info_packets,
+	.update_dp_info_packets =
+		enc1_stream_encoder_update_dp_info_packets,
+	.stop_dp_info_packets =
+		enc1_stream_encoder_stop_dp_info_packets,
+	.dp_blank =
+		enc1_stream_encoder_dp_blank,
+	.dp_unblank =
+		enc1_stream_encoder_dp_unblank,
+	.audio_mute_control = enc1_se_audio_mute_control,
+
+	.dp_audio_setup = enc1_se_dp_audio_setup,
+	.dp_audio_enable = enc1_se_dp_audio_enable,
+	.dp_audio_disable = enc1_se_dp_audio_disable,
+
+	.hdmi_audio_setup = enc1_se_hdmi_audio_setup,
+	.hdmi_audio_disable = enc1_se_hdmi_audio_disable,
+	.setup_stereo_sync  = enc1_setup_stereo_sync,
+	.set_avmute = enc1_stream_encoder_set_avmute,
+};
+
+void dcn10_stream_encoder_construct(
+	struct dcn10_stream_encoder *enc1,
+	struct dc_context *ctx,
+	struct dc_bios *bp,
+	enum engine_id eng_id,
+	const struct dcn10_stream_enc_registers *regs,
+	const struct dcn10_stream_encoder_shift *se_shift,
+	const struct dcn10_stream_encoder_mask *se_mask)
+{
+	enc1->base.funcs = &dcn10_str_enc_funcs;
+	enc1->base.ctx = ctx;
+	enc1->base.id = eng_id;
+	enc1->base.bp = bp;
+	enc1->regs = regs;
+	enc1->se_shift = se_shift;
+	enc1->se_mask = se_mask;
+}
+
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.h
new file mode 100644
index 0000000..86f8ee5
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.h
@@ -0,0 +1,584 @@
+/*
+ * Copyright 2012-15 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ *  and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef __DC_STREAM_ENCODER_DCN10_H__
+#define __DC_STREAM_ENCODER_DCN10_H__
+
+#include "stream_encoder.h"
+
+#define DCN10STRENC_FROM_STRENC(stream_encoder)\
+	container_of(stream_encoder, struct dcn10_stream_encoder, base)
+
+#define SE_COMMON_REG_LIST_BASE(id) \
+	SRI(AFMT_GENERIC_0, DIG, id), \
+	SRI(AFMT_GENERIC_1, DIG, id), \
+	SRI(AFMT_GENERIC_2, DIG, id), \
+	SRI(AFMT_GENERIC_3, DIG, id), \
+	SRI(AFMT_GENERIC_4, DIG, id), \
+	SRI(AFMT_GENERIC_5, DIG, id), \
+	SRI(AFMT_GENERIC_6, DIG, id), \
+	SRI(AFMT_GENERIC_7, DIG, id), \
+	SRI(AFMT_GENERIC_HDR, DIG, id), \
+	SRI(AFMT_INFOFRAME_CONTROL0, DIG, id), \
+	SRI(AFMT_VBI_PACKET_CONTROL, DIG, id), \
+	SRI(AFMT_AUDIO_PACKET_CONTROL, DIG, id), \
+	SRI(AFMT_AUDIO_PACKET_CONTROL2, DIG, id), \
+	SRI(AFMT_AUDIO_SRC_CONTROL, DIG, id), \
+	SRI(AFMT_60958_0, DIG, id), \
+	SRI(AFMT_60958_1, DIG, id), \
+	SRI(AFMT_60958_2, DIG, id), \
+	SRI(DIG_FE_CNTL, DIG, id), \
+	SRI(HDMI_CONTROL, DIG, id), \
+	SRI(HDMI_GC, DIG, id), \
+	SRI(HDMI_GENERIC_PACKET_CONTROL0, DIG, id), \
+	SRI(HDMI_GENERIC_PACKET_CONTROL1, DIG, id), \
+	SRI(HDMI_INFOFRAME_CONTROL0, DIG, id), \
+	SRI(HDMI_INFOFRAME_CONTROL1, DIG, id), \
+	SRI(HDMI_VBI_PACKET_CONTROL, DIG, id), \
+	SRI(HDMI_AUDIO_PACKET_CONTROL, DIG, id),\
+	SRI(HDMI_ACR_PACKET_CONTROL, DIG, id),\
+	SRI(HDMI_ACR_32_0, DIG, id),\
+	SRI(HDMI_ACR_32_1, DIG, id),\
+	SRI(HDMI_ACR_44_0, DIG, id),\
+	SRI(HDMI_ACR_44_1, DIG, id),\
+	SRI(HDMI_ACR_48_0, DIG, id),\
+	SRI(HDMI_ACR_48_1, DIG, id),\
+	SRI(TMDS_CNTL, DIG, id), \
+	SRI(DP_MSE_RATE_CNTL, DP, id), \
+	SRI(DP_MSE_RATE_UPDATE, DP, id), \
+	SRI(DP_PIXEL_FORMAT, DP, id), \
+	SRI(DP_SEC_CNTL, DP, id), \
+	SRI(DP_STEER_FIFO, DP, id), \
+	SRI(DP_VID_M, DP, id), \
+	SRI(DP_VID_N, DP, id), \
+	SRI(DP_VID_STREAM_CNTL, DP, id), \
+	SRI(DP_VID_TIMING, DP, id), \
+	SRI(DP_SEC_AUD_N, DP, id), \
+	SRI(DP_SEC_TIMESTAMP, DP, id)
+
+#define SE_DCN_REG_LIST(id)\
+	SE_COMMON_REG_LIST_BASE(id),\
+	SRI(AFMT_CNTL, DIG, id),\
+	SRI(AFMT_VBI_PACKET_CONTROL1, DIG, id),\
+	SRI(HDMI_GENERIC_PACKET_CONTROL2, DIG, id), \
+	SRI(HDMI_GENERIC_PACKET_CONTROL3, DIG, id), \
+	SRI(DP_DB_CNTL, DP, id), \
+	SRI(DP_MSA_MISC, DP, id), \
+	SRI(DP_MSA_COLORIMETRY, DP, id), \
+	SRI(DP_MSA_TIMING_PARAM1, DP, id), \
+	SRI(DP_MSA_TIMING_PARAM2, DP, id), \
+	SRI(DP_MSA_TIMING_PARAM3, DP, id), \
+	SRI(DP_MSA_TIMING_PARAM4, DP, id), \
+	SRI(HDMI_DB_CONTROL, DIG, id)
+
+#define SE_SF(reg_name, field_name, post_fix)\
+	.field_name = reg_name ## __ ## field_name ## post_fix
+
+#define SE_COMMON_MASK_SH_LIST_SOC_BASE(mask_sh)\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL, AFMT_GENERIC_INDEX, mask_sh),\
+	SE_SF(DIG0_AFMT_GENERIC_HDR, AFMT_GENERIC_HB0, mask_sh),\
+	SE_SF(DIG0_AFMT_GENERIC_HDR, AFMT_GENERIC_HB1, mask_sh),\
+	SE_SF(DIG0_AFMT_GENERIC_HDR, AFMT_GENERIC_HB2, mask_sh),\
+	SE_SF(DIG0_AFMT_GENERIC_HDR, AFMT_GENERIC_HB3, mask_sh),\
+	SE_SF(DP0_DP_PIXEL_FORMAT, DP_PIXEL_ENCODING, mask_sh),\
+	SE_SF(DP0_DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH, mask_sh),\
+	SE_SF(DIG0_HDMI_CONTROL, HDMI_PACKET_GEN_VERSION, mask_sh),\
+	SE_SF(DIG0_HDMI_CONTROL, HDMI_KEEPOUT_MODE, mask_sh),\
+	SE_SF(DIG0_HDMI_CONTROL, HDMI_DEEP_COLOR_ENABLE, mask_sh),\
+	SE_SF(DIG0_HDMI_CONTROL, HDMI_DEEP_COLOR_DEPTH, mask_sh),\
+	SE_SF(DIG0_HDMI_CONTROL, HDMI_DATA_SCRAMBLE_EN, mask_sh),\
+	SE_SF(DIG0_HDMI_VBI_PACKET_CONTROL, HDMI_GC_CONT, mask_sh),\
+	SE_SF(DIG0_HDMI_VBI_PACKET_CONTROL, HDMI_GC_SEND, mask_sh),\
+	SE_SF(DIG0_HDMI_VBI_PACKET_CONTROL, HDMI_NULL_SEND, mask_sh),\
+	SE_SF(DIG0_HDMI_INFOFRAME_CONTROL0, HDMI_AUDIO_INFO_SEND, mask_sh),\
+	SE_SF(DIG0_AFMT_INFOFRAME_CONTROL0, AFMT_AUDIO_INFO_UPDATE, mask_sh),\
+	SE_SF(DIG0_HDMI_INFOFRAME_CONTROL1, HDMI_AUDIO_INFO_LINE, mask_sh),\
+	SE_SF(DIG0_HDMI_GC, HDMI_GC_AVMUTE, mask_sh),\
+	SE_SF(DP0_DP_MSE_RATE_CNTL, DP_MSE_RATE_X, mask_sh),\
+	SE_SF(DP0_DP_MSE_RATE_CNTL, DP_MSE_RATE_Y, mask_sh),\
+	SE_SF(DP0_DP_MSE_RATE_UPDATE, DP_MSE_RATE_UPDATE_PENDING, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_GSP0_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_STREAM_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_GSP1_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_GSP2_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_GSP3_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_MPG_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_VID_STREAM_CNTL, DP_VID_STREAM_DIS_DEFER, mask_sh),\
+	SE_SF(DP0_DP_VID_STREAM_CNTL, DP_VID_STREAM_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_VID_STREAM_CNTL, DP_VID_STREAM_STATUS, mask_sh),\
+	SE_SF(DP0_DP_STEER_FIFO, DP_STEER_FIFO_RESET, mask_sh),\
+	SE_SF(DP0_DP_VID_TIMING, DP_VID_M_N_GEN_EN, mask_sh),\
+	SE_SF(DP0_DP_VID_N, DP_VID_N, mask_sh),\
+	SE_SF(DP0_DP_VID_M, DP_VID_M, mask_sh),\
+	SE_SF(DIG0_DIG_FE_CNTL, DIG_START, mask_sh),\
+	SE_SF(DIG0_AFMT_AUDIO_SRC_CONTROL, AFMT_AUDIO_SRC_SELECT, mask_sh),\
+	SE_SF(DIG0_AFMT_AUDIO_PACKET_CONTROL2, AFMT_AUDIO_CHANNEL_ENABLE, mask_sh),\
+	SE_SF(DIG0_HDMI_AUDIO_PACKET_CONTROL, HDMI_AUDIO_PACKETS_PER_LINE, mask_sh),\
+	SE_SF(DIG0_HDMI_AUDIO_PACKET_CONTROL, HDMI_AUDIO_DELAY_EN, mask_sh),\
+	SE_SF(DIG0_AFMT_AUDIO_PACKET_CONTROL, AFMT_60958_CS_UPDATE, mask_sh),\
+	SE_SF(DIG0_AFMT_AUDIO_PACKET_CONTROL2, AFMT_AUDIO_LAYOUT_OVRD, mask_sh),\
+	SE_SF(DIG0_AFMT_AUDIO_PACKET_CONTROL2, AFMT_60958_OSF_OVRD, mask_sh),\
+	SE_SF(DIG0_HDMI_ACR_PACKET_CONTROL, HDMI_ACR_AUTO_SEND, mask_sh),\
+	SE_SF(DIG0_HDMI_ACR_PACKET_CONTROL, HDMI_ACR_SOURCE, mask_sh),\
+	SE_SF(DIG0_HDMI_ACR_PACKET_CONTROL, HDMI_ACR_AUDIO_PRIORITY, mask_sh),\
+	SE_SF(DIG0_HDMI_ACR_32_0, HDMI_ACR_CTS_32, mask_sh),\
+	SE_SF(DIG0_HDMI_ACR_32_1, HDMI_ACR_N_32, mask_sh),\
+	SE_SF(DIG0_HDMI_ACR_44_0, HDMI_ACR_CTS_44, mask_sh),\
+	SE_SF(DIG0_HDMI_ACR_44_1, HDMI_ACR_N_44, mask_sh),\
+	SE_SF(DIG0_HDMI_ACR_48_0, HDMI_ACR_CTS_48, mask_sh),\
+	SE_SF(DIG0_HDMI_ACR_48_1, HDMI_ACR_N_48, mask_sh),\
+	SE_SF(DIG0_AFMT_60958_0, AFMT_60958_CS_CHANNEL_NUMBER_L, mask_sh),\
+	SE_SF(DIG0_AFMT_60958_0, AFMT_60958_CS_CLOCK_ACCURACY, mask_sh),\
+	SE_SF(DIG0_AFMT_60958_1, AFMT_60958_CS_CHANNEL_NUMBER_R, mask_sh),\
+	SE_SF(DIG0_AFMT_60958_2, AFMT_60958_CS_CHANNEL_NUMBER_2, mask_sh),\
+	SE_SF(DIG0_AFMT_60958_2, AFMT_60958_CS_CHANNEL_NUMBER_3, mask_sh),\
+	SE_SF(DIG0_AFMT_60958_2, AFMT_60958_CS_CHANNEL_NUMBER_4, mask_sh),\
+	SE_SF(DIG0_AFMT_60958_2, AFMT_60958_CS_CHANNEL_NUMBER_5, mask_sh),\
+	SE_SF(DIG0_AFMT_60958_2, AFMT_60958_CS_CHANNEL_NUMBER_6, mask_sh),\
+	SE_SF(DIG0_AFMT_60958_2, AFMT_60958_CS_CHANNEL_NUMBER_7, mask_sh),\
+	SE_SF(DP0_DP_SEC_AUD_N, DP_SEC_AUD_N, mask_sh),\
+	SE_SF(DP0_DP_SEC_TIMESTAMP, DP_SEC_TIMESTAMP_MODE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_ASP_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_ATP_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_AIP_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_ACM_ENABLE, mask_sh),\
+	SE_SF(DIG0_AFMT_AUDIO_PACKET_CONTROL, AFMT_AUDIO_SAMPLE_SEND, mask_sh),\
+	SE_SF(DIG0_AFMT_CNTL, AFMT_AUDIO_CLOCK_EN, mask_sh),\
+	SE_SF(DIG0_HDMI_CONTROL, HDMI_CLOCK_CHANNEL_RATE, mask_sh),\
+	SE_SF(DIG0_DIG_FE_CNTL, TMDS_PIXEL_ENCODING, mask_sh),\
+	SE_SF(DIG0_DIG_FE_CNTL, TMDS_COLOR_FORMAT, mask_sh),\
+	SE_SF(DIG0_DIG_FE_CNTL, DIG_STEREOSYNC_SELECT, mask_sh),\
+	SE_SF(DIG0_DIG_FE_CNTL, DIG_STEREOSYNC_GATE_EN, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL, AFMT_GENERIC_LOCK_STATUS, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL, AFMT_GENERIC_CONFLICT, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL, AFMT_GENERIC_CONFLICT_CLR, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC0_FRAME_UPDATE_PENDING, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC1_FRAME_UPDATE_PENDING, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC2_FRAME_UPDATE_PENDING, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC3_FRAME_UPDATE_PENDING, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC4_FRAME_UPDATE_PENDING, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC5_FRAME_UPDATE_PENDING, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC6_FRAME_UPDATE_PENDING, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC7_FRAME_UPDATE_PENDING, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC0_FRAME_UPDATE, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC1_FRAME_UPDATE, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC2_FRAME_UPDATE, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC3_FRAME_UPDATE, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC4_FRAME_UPDATE, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC5_FRAME_UPDATE, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC6_FRAME_UPDATE, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC7_FRAME_UPDATE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_GSP4_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_GSP5_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_GSP6_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_GSP7_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_DB_CNTL, DP_DB_DISABLE, mask_sh),\
+	SE_SF(DP0_DP_MSA_COLORIMETRY, DP_MSA_MISC0, mask_sh),\
+	SE_SF(DP0_DP_MSA_TIMING_PARAM1, DP_MSA_HTOTAL, mask_sh),\
+	SE_SF(DP0_DP_MSA_TIMING_PARAM1, DP_MSA_VTOTAL, mask_sh),\
+	SE_SF(DP0_DP_MSA_TIMING_PARAM2, DP_MSA_HSTART, mask_sh),\
+	SE_SF(DP0_DP_MSA_TIMING_PARAM2, DP_MSA_VSTART, mask_sh),\
+	SE_SF(DP0_DP_MSA_TIMING_PARAM3, DP_MSA_HSYNCWIDTH, mask_sh),\
+	SE_SF(DP0_DP_MSA_TIMING_PARAM3, DP_MSA_HSYNCPOLARITY, mask_sh),\
+	SE_SF(DP0_DP_MSA_TIMING_PARAM3, DP_MSA_VSYNCWIDTH, mask_sh),\
+	SE_SF(DP0_DP_MSA_TIMING_PARAM3, DP_MSA_VSYNCPOLARITY, mask_sh),\
+	SE_SF(DP0_DP_MSA_TIMING_PARAM4, DP_MSA_HWIDTH, mask_sh),\
+	SE_SF(DP0_DP_MSA_TIMING_PARAM4, DP_MSA_VHEIGHT, mask_sh),\
+	SE_SF(DIG0_HDMI_DB_CONTROL, HDMI_DB_DISABLE, mask_sh),\
+	SE_SF(DP0_DP_VID_TIMING, DP_VID_N_MUL, mask_sh)
+
+#define SE_COMMON_MASK_SH_LIST_SOC(mask_sh)\
+	SE_COMMON_MASK_SH_LIST_SOC_BASE(mask_sh)
+
+#define SE_COMMON_MASK_SH_LIST_DCN10(mask_sh)\
+	SE_COMMON_MASK_SH_LIST_SOC(mask_sh),\
+	SE_SF(DIG0_HDMI_GENERIC_PACKET_CONTROL0, HDMI_GENERIC0_CONT, mask_sh),\
+	SE_SF(DIG0_HDMI_GENERIC_PACKET_CONTROL0, HDMI_GENERIC0_SEND, mask_sh),\
+	SE_SF(DIG0_HDMI_GENERIC_PACKET_CONTROL0, HDMI_GENERIC0_LINE, mask_sh),\
+	SE_SF(DIG0_HDMI_GENERIC_PACKET_CONTROL0, HDMI_GENERIC1_CONT, mask_sh),\
+	SE_SF(DIG0_HDMI_GENERIC_PACKET_CONTROL0, HDMI_GENERIC1_SEND, mask_sh),\
+	SE_SF(DIG0_HDMI_GENERIC_PACKET_CONTROL0, HDMI_GENERIC1_LINE, mask_sh)
+
+struct dcn10_stream_encoder_shift {
+	uint8_t AFMT_GENERIC_INDEX;
+	uint8_t AFMT_GENERIC_HB0;
+	uint8_t AFMT_GENERIC_HB1;
+	uint8_t AFMT_GENERIC_HB2;
+	uint8_t AFMT_GENERIC_HB3;
+	uint8_t AFMT_GENERIC_LOCK_STATUS;
+	uint8_t AFMT_GENERIC_CONFLICT;
+	uint8_t AFMT_GENERIC_CONFLICT_CLR;
+	uint8_t AFMT_GENERIC0_FRAME_UPDATE_PENDING;
+	uint8_t AFMT_GENERIC1_FRAME_UPDATE_PENDING;
+	uint8_t AFMT_GENERIC2_FRAME_UPDATE_PENDING;
+	uint8_t AFMT_GENERIC3_FRAME_UPDATE_PENDING;
+	uint8_t AFMT_GENERIC4_FRAME_UPDATE_PENDING;
+	uint8_t AFMT_GENERIC5_FRAME_UPDATE_PENDING;
+	uint8_t AFMT_GENERIC6_FRAME_UPDATE_PENDING;
+	uint8_t AFMT_GENERIC7_FRAME_UPDATE_PENDING;
+	uint8_t AFMT_GENERIC0_FRAME_UPDATE;
+	uint8_t AFMT_GENERIC1_FRAME_UPDATE;
+	uint8_t AFMT_GENERIC2_FRAME_UPDATE;
+	uint8_t AFMT_GENERIC3_FRAME_UPDATE;
+	uint8_t AFMT_GENERIC4_FRAME_UPDATE;
+	uint8_t AFMT_GENERIC5_FRAME_UPDATE;
+	uint8_t AFMT_GENERIC6_FRAME_UPDATE;
+	uint8_t AFMT_GENERIC7_FRAME_UPDATE;
+	uint8_t HDMI_GENERIC0_CONT;
+	uint8_t HDMI_GENERIC0_SEND;
+	uint8_t HDMI_GENERIC0_LINE;
+	uint8_t HDMI_GENERIC1_CONT;
+	uint8_t HDMI_GENERIC1_SEND;
+	uint8_t HDMI_GENERIC1_LINE;
+	uint8_t HDMI_GENERIC2_CONT;
+	uint8_t HDMI_GENERIC2_SEND;
+	uint8_t HDMI_GENERIC2_LINE;
+	uint8_t HDMI_GENERIC3_CONT;
+	uint8_t HDMI_GENERIC3_SEND;
+	uint8_t HDMI_GENERIC3_LINE;
+	uint8_t HDMI_GENERIC4_CONT;
+	uint8_t HDMI_GENERIC4_SEND;
+	uint8_t HDMI_GENERIC4_LINE;
+	uint8_t HDMI_GENERIC5_CONT;
+	uint8_t HDMI_GENERIC5_SEND;
+	uint8_t HDMI_GENERIC5_LINE;
+	uint8_t HDMI_GENERIC6_CONT;
+	uint8_t HDMI_GENERIC6_SEND;
+	uint8_t HDMI_GENERIC6_LINE;
+	uint8_t HDMI_GENERIC7_CONT;
+	uint8_t HDMI_GENERIC7_SEND;
+	uint8_t HDMI_GENERIC7_LINE;
+	uint8_t DP_PIXEL_ENCODING;
+	uint8_t DP_COMPONENT_DEPTH;
+	uint8_t HDMI_PACKET_GEN_VERSION;
+	uint8_t HDMI_KEEPOUT_MODE;
+	uint8_t HDMI_DEEP_COLOR_ENABLE;
+	uint8_t HDMI_CLOCK_CHANNEL_RATE;
+	uint8_t HDMI_DEEP_COLOR_DEPTH;
+	uint8_t HDMI_GC_CONT;
+	uint8_t HDMI_GC_SEND;
+	uint8_t HDMI_NULL_SEND;
+	uint8_t HDMI_DATA_SCRAMBLE_EN;
+	uint8_t HDMI_AUDIO_INFO_SEND;
+	uint8_t AFMT_AUDIO_INFO_UPDATE;
+	uint8_t HDMI_AUDIO_INFO_LINE;
+	uint8_t HDMI_GC_AVMUTE;
+	uint8_t DP_MSE_RATE_X;
+	uint8_t DP_MSE_RATE_Y;
+	uint8_t DP_MSE_RATE_UPDATE_PENDING;
+	uint8_t DP_SEC_GSP0_ENABLE;
+	uint8_t DP_SEC_STREAM_ENABLE;
+	uint8_t DP_SEC_GSP1_ENABLE;
+	uint8_t DP_SEC_GSP2_ENABLE;
+	uint8_t DP_SEC_GSP3_ENABLE;
+	uint8_t DP_SEC_GSP4_ENABLE;
+	uint8_t DP_SEC_GSP5_ENABLE;
+	uint8_t DP_SEC_GSP6_ENABLE;
+	uint8_t DP_SEC_GSP7_ENABLE;
+	uint8_t DP_SEC_MPG_ENABLE;
+	uint8_t DP_VID_STREAM_DIS_DEFER;
+	uint8_t DP_VID_STREAM_ENABLE;
+	uint8_t DP_VID_STREAM_STATUS;
+	uint8_t DP_STEER_FIFO_RESET;
+	uint8_t DP_VID_M_N_GEN_EN;
+	uint8_t DP_VID_N;
+	uint8_t DP_VID_M;
+	uint8_t DIG_START;
+	uint8_t AFMT_AUDIO_SRC_SELECT;
+	uint8_t AFMT_AUDIO_CHANNEL_ENABLE;
+	uint8_t HDMI_AUDIO_PACKETS_PER_LINE;
+	uint8_t HDMI_AUDIO_DELAY_EN;
+	uint8_t AFMT_60958_CS_UPDATE;
+	uint8_t AFMT_AUDIO_LAYOUT_OVRD;
+	uint8_t AFMT_60958_OSF_OVRD;
+	uint8_t HDMI_ACR_AUTO_SEND;
+	uint8_t HDMI_ACR_SOURCE;
+	uint8_t HDMI_ACR_AUDIO_PRIORITY;
+	uint8_t HDMI_ACR_CTS_32;
+	uint8_t HDMI_ACR_N_32;
+	uint8_t HDMI_ACR_CTS_44;
+	uint8_t HDMI_ACR_N_44;
+	uint8_t HDMI_ACR_CTS_48;
+	uint8_t HDMI_ACR_N_48;
+	uint8_t AFMT_60958_CS_CHANNEL_NUMBER_L;
+	uint8_t AFMT_60958_CS_CLOCK_ACCURACY;
+	uint8_t AFMT_60958_CS_CHANNEL_NUMBER_R;
+	uint8_t AFMT_60958_CS_CHANNEL_NUMBER_2;
+	uint8_t AFMT_60958_CS_CHANNEL_NUMBER_3;
+	uint8_t AFMT_60958_CS_CHANNEL_NUMBER_4;
+	uint8_t AFMT_60958_CS_CHANNEL_NUMBER_5;
+	uint8_t AFMT_60958_CS_CHANNEL_NUMBER_6;
+	uint8_t AFMT_60958_CS_CHANNEL_NUMBER_7;
+	uint8_t DP_SEC_AUD_N;
+	uint8_t DP_SEC_TIMESTAMP_MODE;
+	uint8_t DP_SEC_ASP_ENABLE;
+	uint8_t DP_SEC_ATP_ENABLE;
+	uint8_t DP_SEC_AIP_ENABLE;
+	uint8_t DP_SEC_ACM_ENABLE;
+	uint8_t AFMT_AUDIO_SAMPLE_SEND;
+	uint8_t AFMT_AUDIO_CLOCK_EN;
+	uint8_t TMDS_PIXEL_ENCODING;
+	uint8_t TMDS_COLOR_FORMAT;
+	uint8_t DIG_STEREOSYNC_SELECT;
+	uint8_t DIG_STEREOSYNC_GATE_EN;
+	uint8_t DP_DB_DISABLE;
+	uint8_t DP_MSA_MISC0;
+	uint8_t DP_MSA_HTOTAL;
+	uint8_t DP_MSA_VTOTAL;
+	uint8_t DP_MSA_HSTART;
+	uint8_t DP_MSA_VSTART;
+	uint8_t DP_MSA_HSYNCWIDTH;
+	uint8_t DP_MSA_HSYNCPOLARITY;
+	uint8_t DP_MSA_VSYNCWIDTH;
+	uint8_t DP_MSA_VSYNCPOLARITY;
+	uint8_t DP_MSA_HWIDTH;
+	uint8_t DP_MSA_VHEIGHT;
+	uint8_t HDMI_DB_DISABLE;
+	uint8_t DP_VID_N_MUL;
+	uint8_t DP_VID_M_DOUBLE_VALUE_EN;
+};
+
+struct dcn10_stream_encoder_mask {
+	uint32_t AFMT_GENERIC_INDEX;
+	uint32_t AFMT_GENERIC_HB0;
+	uint32_t AFMT_GENERIC_HB1;
+	uint32_t AFMT_GENERIC_HB2;
+	uint32_t AFMT_GENERIC_HB3;
+	uint32_t AFMT_GENERIC_LOCK_STATUS;
+	uint32_t AFMT_GENERIC_CONFLICT;
+	uint32_t AFMT_GENERIC_CONFLICT_CLR;
+	uint32_t AFMT_GENERIC0_FRAME_UPDATE_PENDING;
+	uint32_t AFMT_GENERIC1_FRAME_UPDATE_PENDING;
+	uint32_t AFMT_GENERIC2_FRAME_UPDATE_PENDING;
+	uint32_t AFMT_GENERIC3_FRAME_UPDATE_PENDING;
+	uint32_t AFMT_GENERIC4_FRAME_UPDATE_PENDING;
+	uint32_t AFMT_GENERIC5_FRAME_UPDATE_PENDING;
+	uint32_t AFMT_GENERIC6_FRAME_UPDATE_PENDING;
+	uint32_t AFMT_GENERIC7_FRAME_UPDATE_PENDING;
+	uint32_t AFMT_GENERIC0_FRAME_UPDATE;
+	uint32_t AFMT_GENERIC1_FRAME_UPDATE;
+	uint32_t AFMT_GENERIC2_FRAME_UPDATE;
+	uint32_t AFMT_GENERIC3_FRAME_UPDATE;
+	uint32_t AFMT_GENERIC4_FRAME_UPDATE;
+	uint32_t AFMT_GENERIC5_FRAME_UPDATE;
+	uint32_t AFMT_GENERIC6_FRAME_UPDATE;
+	uint32_t AFMT_GENERIC7_FRAME_UPDATE;
+	uint32_t HDMI_GENERIC0_CONT;
+	uint32_t HDMI_GENERIC0_SEND;
+	uint32_t HDMI_GENERIC0_LINE;
+	uint32_t HDMI_GENERIC1_CONT;
+	uint32_t HDMI_GENERIC1_SEND;
+	uint32_t HDMI_GENERIC1_LINE;
+	uint32_t HDMI_GENERIC2_CONT;
+	uint32_t HDMI_GENERIC2_SEND;
+	uint32_t HDMI_GENERIC2_LINE;
+	uint32_t HDMI_GENERIC3_CONT;
+	uint32_t HDMI_GENERIC3_SEND;
+	uint32_t HDMI_GENERIC3_LINE;
+	uint32_t HDMI_GENERIC4_CONT;
+	uint32_t HDMI_GENERIC4_SEND;
+	uint32_t HDMI_GENERIC4_LINE;
+	uint32_t HDMI_GENERIC5_CONT;
+	uint32_t HDMI_GENERIC5_SEND;
+	uint32_t HDMI_GENERIC5_LINE;
+	uint32_t HDMI_GENERIC6_CONT;
+	uint32_t HDMI_GENERIC6_SEND;
+	uint32_t HDMI_GENERIC6_LINE;
+	uint32_t HDMI_GENERIC7_CONT;
+	uint32_t HDMI_GENERIC7_SEND;
+	uint32_t HDMI_GENERIC7_LINE;
+	uint32_t DP_PIXEL_ENCODING;
+	uint32_t DP_COMPONENT_DEPTH;
+	uint32_t HDMI_PACKET_GEN_VERSION;
+	uint32_t HDMI_KEEPOUT_MODE;
+	uint32_t HDMI_DEEP_COLOR_ENABLE;
+	uint32_t HDMI_CLOCK_CHANNEL_RATE;
+	uint32_t HDMI_DEEP_COLOR_DEPTH;
+	uint32_t HDMI_GC_CONT;
+	uint32_t HDMI_GC_SEND;
+	uint32_t HDMI_NULL_SEND;
+	uint32_t HDMI_DATA_SCRAMBLE_EN;
+	uint32_t HDMI_AUDIO_INFO_SEND;
+	uint32_t AFMT_AUDIO_INFO_UPDATE;
+	uint32_t HDMI_AUDIO_INFO_LINE;
+	uint32_t HDMI_GC_AVMUTE;
+	uint32_t DP_MSE_RATE_X;
+	uint32_t DP_MSE_RATE_Y;
+	uint32_t DP_MSE_RATE_UPDATE_PENDING;
+	uint32_t DP_SEC_GSP0_ENABLE;
+	uint32_t DP_SEC_STREAM_ENABLE;
+	uint32_t DP_SEC_GSP1_ENABLE;
+	uint32_t DP_SEC_GSP2_ENABLE;
+	uint32_t DP_SEC_GSP3_ENABLE;
+	uint32_t DP_SEC_GSP4_ENABLE;
+	uint32_t DP_SEC_GSP5_ENABLE;
+	uint32_t DP_SEC_GSP6_ENABLE;
+	uint32_t DP_SEC_GSP7_ENABLE;
+	uint32_t DP_SEC_MPG_ENABLE;
+	uint32_t DP_VID_STREAM_DIS_DEFER;
+	uint32_t DP_VID_STREAM_ENABLE;
+	uint32_t DP_VID_STREAM_STATUS;
+	uint32_t DP_STEER_FIFO_RESET;
+	uint32_t DP_VID_M_N_GEN_EN;
+	uint32_t DP_VID_N;
+	uint32_t DP_VID_M;
+	uint32_t DIG_START;
+	uint32_t AFMT_AUDIO_SRC_SELECT;
+	uint32_t AFMT_AUDIO_CHANNEL_ENABLE;
+	uint32_t HDMI_AUDIO_PACKETS_PER_LINE;
+	uint32_t HDMI_AUDIO_DELAY_EN;
+	uint32_t AFMT_60958_CS_UPDATE;
+	uint32_t AFMT_AUDIO_LAYOUT_OVRD;
+	uint32_t AFMT_60958_OSF_OVRD;
+	uint32_t HDMI_ACR_AUTO_SEND;
+	uint32_t HDMI_ACR_SOURCE;
+	uint32_t HDMI_ACR_AUDIO_PRIORITY;
+	uint32_t HDMI_ACR_CTS_32;
+	uint32_t HDMI_ACR_N_32;
+	uint32_t HDMI_ACR_CTS_44;
+	uint32_t HDMI_ACR_N_44;
+	uint32_t HDMI_ACR_CTS_48;
+	uint32_t HDMI_ACR_N_48;
+	uint32_t AFMT_60958_CS_CHANNEL_NUMBER_L;
+	uint32_t AFMT_60958_CS_CLOCK_ACCURACY;
+	uint32_t AFMT_60958_CS_CHANNEL_NUMBER_R;
+	uint32_t AFMT_60958_CS_CHANNEL_NUMBER_2;
+	uint32_t AFMT_60958_CS_CHANNEL_NUMBER_3;
+	uint32_t AFMT_60958_CS_CHANNEL_NUMBER_4;
+	uint32_t AFMT_60958_CS_CHANNEL_NUMBER_5;
+	uint32_t AFMT_60958_CS_CHANNEL_NUMBER_6;
+	uint32_t AFMT_60958_CS_CHANNEL_NUMBER_7;
+	uint32_t DP_SEC_AUD_N;
+	uint32_t DP_SEC_TIMESTAMP_MODE;
+	uint32_t DP_SEC_ASP_ENABLE;
+	uint32_t DP_SEC_ATP_ENABLE;
+	uint32_t DP_SEC_AIP_ENABLE;
+	uint32_t DP_SEC_ACM_ENABLE;
+	uint32_t AFMT_AUDIO_SAMPLE_SEND;
+	uint32_t AFMT_AUDIO_CLOCK_EN;
+	uint32_t TMDS_PIXEL_ENCODING;
+	uint32_t DIG_STEREOSYNC_SELECT;
+	uint32_t DIG_STEREOSYNC_GATE_EN;
+	uint32_t TMDS_COLOR_FORMAT;
+	uint32_t DP_DB_DISABLE;
+	uint32_t DP_MSA_MISC0;
+	uint32_t DP_MSA_HTOTAL;
+	uint32_t DP_MSA_VTOTAL;
+	uint32_t DP_MSA_HSTART;
+	uint32_t DP_MSA_VSTART;
+	uint32_t DP_MSA_HSYNCWIDTH;
+	uint32_t DP_MSA_HSYNCPOLARITY;
+	uint32_t DP_MSA_VSYNCWIDTH;
+	uint32_t DP_MSA_VSYNCPOLARITY;
+	uint32_t DP_MSA_HWIDTH;
+	uint32_t DP_MSA_VHEIGHT;
+	uint32_t HDMI_DB_DISABLE;
+	uint32_t DP_VID_N_MUL;
+	uint32_t DP_VID_M_DOUBLE_VALUE_EN;
+};
+
+struct dcn10_stream_enc_registers {
+	uint32_t AFMT_CNTL;
+	uint32_t AFMT_AVI_INFO0;
+	uint32_t AFMT_AVI_INFO1;
+	uint32_t AFMT_AVI_INFO2;
+	uint32_t AFMT_AVI_INFO3;
+	uint32_t AFMT_GENERIC_0;
+	uint32_t AFMT_GENERIC_1;
+	uint32_t AFMT_GENERIC_2;
+	uint32_t AFMT_GENERIC_3;
+	uint32_t AFMT_GENERIC_4;
+	uint32_t AFMT_GENERIC_5;
+	uint32_t AFMT_GENERIC_6;
+	uint32_t AFMT_GENERIC_7;
+	uint32_t AFMT_GENERIC_HDR;
+	uint32_t AFMT_INFOFRAME_CONTROL0;
+	uint32_t AFMT_VBI_PACKET_CONTROL;
+	uint32_t AFMT_VBI_PACKET_CONTROL1;
+	uint32_t AFMT_AUDIO_PACKET_CONTROL;
+	uint32_t AFMT_AUDIO_PACKET_CONTROL2;
+	uint32_t AFMT_AUDIO_SRC_CONTROL;
+	uint32_t AFMT_60958_0;
+	uint32_t AFMT_60958_1;
+	uint32_t AFMT_60958_2;
+	uint32_t DIG_FE_CNTL;
+	uint32_t DP_MSE_RATE_CNTL;
+	uint32_t DP_MSE_RATE_UPDATE;
+	uint32_t DP_PIXEL_FORMAT;
+	uint32_t DP_SEC_CNTL;
+	uint32_t DP_STEER_FIFO;
+	uint32_t DP_VID_M;
+	uint32_t DP_VID_N;
+	uint32_t DP_VID_STREAM_CNTL;
+	uint32_t DP_VID_TIMING;
+	uint32_t DP_SEC_AUD_N;
+	uint32_t DP_SEC_TIMESTAMP;
+	uint32_t HDMI_CONTROL;
+	uint32_t HDMI_GC;
+	uint32_t HDMI_GENERIC_PACKET_CONTROL0;
+	uint32_t HDMI_GENERIC_PACKET_CONTROL1;
+	uint32_t HDMI_GENERIC_PACKET_CONTROL2;
+	uint32_t HDMI_GENERIC_PACKET_CONTROL3;
+	uint32_t HDMI_GENERIC_PACKET_CONTROL4;
+	uint32_t HDMI_GENERIC_PACKET_CONTROL5;
+	uint32_t HDMI_INFOFRAME_CONTROL0;
+	uint32_t HDMI_INFOFRAME_CONTROL1;
+	uint32_t HDMI_VBI_PACKET_CONTROL;
+	uint32_t HDMI_AUDIO_PACKET_CONTROL;
+	uint32_t HDMI_ACR_PACKET_CONTROL;
+	uint32_t HDMI_ACR_32_0;
+	uint32_t HDMI_ACR_32_1;
+	uint32_t HDMI_ACR_44_0;
+	uint32_t HDMI_ACR_44_1;
+	uint32_t HDMI_ACR_48_0;
+	uint32_t HDMI_ACR_48_1;
+	uint32_t TMDS_CNTL;
+	uint32_t DP_DB_CNTL;
+	uint32_t DP_MSA_MISC;
+	uint32_t DP_MSA_COLORIMETRY;
+	uint32_t DP_MSA_TIMING_PARAM1;
+	uint32_t DP_MSA_TIMING_PARAM2;
+	uint32_t DP_MSA_TIMING_PARAM3;
+	uint32_t DP_MSA_TIMING_PARAM4;
+	uint32_t HDMI_DB_CONTROL;
+};
+
+struct dcn10_stream_encoder {
+	struct stream_encoder base;
+	const struct dcn10_stream_enc_registers *regs;
+	const struct dcn10_stream_encoder_shift *se_shift;
+	const struct dcn10_stream_encoder_mask *se_mask;
+};
+
+void dcn10_stream_encoder_construct(
+	struct dcn10_stream_encoder *enc1,
+	struct dc_context *ctx,
+	struct dc_bios *bp,
+	enum engine_id eng_id,
+	const struct dcn10_stream_enc_registers *regs,
+	const struct dcn10_stream_encoder_shift *se_shift,
+	const struct dcn10_stream_encoder_mask *se_mask);
+
+#endif /* __DC_STREAM_ENCODER_DCN10_H__ */
-- 
cgit v1.1


From 3dc8acad23519123bdecaf3184f2ae774c5775fc Mon Sep 17 00:00:00 2001
From: Jun Lei <Jun.Lei@amd.com>
Date: Thu, 1 Mar 2018 08:58:02 -0500
Subject: drm/amd/display: remove unused enum

Signed-off-by: Jun Lei <Jun.Lei@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dc_types.h | 6 ------
 1 file changed, 6 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dc_types.h b/drivers/gpu/drm/amd/display/dc/dc_types.h
index cd324bc..9defe3b 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_types.h
@@ -370,12 +370,6 @@ struct dc_csc_adjustments {
 	struct fixed31_32 hue;
 };
 
-enum {
-	MAX_LANES = 2,
-	MAX_COFUNC_PATH = 6,
-	LAYER_INDEX_PRIMARY = -1,
-};
-
 enum dpcd_downstream_port_max_bpc {
 	DOWN_STREAM_MAX_8BPC = 0,
 	DOWN_STREAM_MAX_10BPC,
-- 
cgit v1.1


From cf65ebeb687678812eb3ddd5ef253bacf7ef330a Mon Sep 17 00:00:00 2001
From: Eric Yang <Eric.Yang2@amd.com>
Date: Fri, 23 Mar 2018 13:56:16 -0400
Subject: drm/amd/display: fix link bw calculation for 422 and 420 encoding

Link bw required is reduced when we have chroma subsampling.

Signed-off-by: Eric Yang <Eric.Yang2@amd.com>
Reviewed-by: Charlene Liu <Charlene.Liu@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c | 53 ++++++++++++++----------
 1 file changed, 32 insertions(+), 21 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
index b86325b..07cc438 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
@@ -1378,37 +1378,48 @@ static uint32_t bandwidth_in_kbps_from_timing(
 {
 	uint32_t bits_per_channel = 0;
 	uint32_t kbps;
-	switch (timing->display_color_depth) {
 
-	case COLOR_DEPTH_666:
-		bits_per_channel = 6;
-		break;
-	case COLOR_DEPTH_888:
-		bits_per_channel = 8;
-		break;
-	case COLOR_DEPTH_101010:
-		bits_per_channel = 10;
-		break;
-	case COLOR_DEPTH_121212:
+	if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR422)
 		bits_per_channel = 12;
-		break;
-	case COLOR_DEPTH_141414:
-		bits_per_channel = 14;
-		break;
-	case COLOR_DEPTH_161616:
-		bits_per_channel = 16;
-		break;
-	default:
-		break;
+	else{
+
+		switch (timing->display_color_depth) {
+
+		case COLOR_DEPTH_666:
+			bits_per_channel = 6;
+			break;
+		case COLOR_DEPTH_888:
+			bits_per_channel = 8;
+			break;
+		case COLOR_DEPTH_101010:
+			bits_per_channel = 10;
+			break;
+		case COLOR_DEPTH_121212:
+			bits_per_channel = 12;
+			break;
+		case COLOR_DEPTH_141414:
+			bits_per_channel = 14;
+			break;
+		case COLOR_DEPTH_161616:
+			bits_per_channel = 16;
+			break;
+		default:
+			break;
+		}
 	}
 	ASSERT(bits_per_channel != 0);
 
 	kbps = timing->pix_clk_khz;
 	kbps *= bits_per_channel;
 
-	if (timing->flags.Y_ONLY != 1)
+	if (timing->flags.Y_ONLY != 1) {
 		/*Only YOnly make reduce bandwidth by 1/3 compares to RGB*/
 		kbps *= 3;
+		if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR420)
+			kbps /= 2;
+		else if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR422)
+			kbps = kbps * 2 / 3;
+	}
 
 	return kbps;
 
-- 
cgit v1.1


From 8f121fe281692ce4b7849ee7be9f3c0dcb079742 Mon Sep 17 00:00:00 2001
From: Jun Lei <Jun.Lei@amd.com>
Date: Mon, 26 Mar 2018 14:01:41 -0400
Subject: drm/amd/display: Fill calcs date from stream src/dst if available

We would otherwise fallback to the timing, which would always give us
identity.

Signed-off-by: Jun Lei <Jun.Lei@amd.com>
Reviewed-by: Wesley Chalmers <Wesley.Chalmers@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c b/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c
index 4b71932..56f46a0 100644
--- a/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c
+++ b/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c
@@ -2933,6 +2933,19 @@ static void populate_initial_data(
 				data->bytes_per_pixel[num_displays + 4] = 4;
 				break;
 			}
+		} else if (pipe[i].stream->dst.width != 0 &&
+					pipe[i].stream->dst.height != 0 &&
+					pipe[i].stream->src.width != 0 &&
+					pipe[i].stream->src.height != 0) {
+			data->src_width[num_displays + 4] = bw_int_to_fixed(pipe[i].stream->src.width);
+			data->pitch_in_pixels[num_displays + 4] = data->src_width[num_displays + 4];
+			data->src_height[num_displays + 4] = bw_int_to_fixed(pipe[i].stream->src.height);
+			data->h_taps[num_displays + 4] = pipe[i].stream->src.width == pipe[i].stream->dst.width ? bw_int_to_fixed(1) : bw_int_to_fixed(2);
+			data->v_taps[num_displays + 4] = pipe[i].stream->src.height == pipe[i].stream->dst.height ? bw_int_to_fixed(1) : bw_int_to_fixed(2);
+			data->h_scale_ratio[num_displays + 4] = bw_frc_to_fixed(pipe[i].stream->src.width, pipe[i].stream->dst.width);
+			data->v_scale_ratio[num_displays + 4] = bw_frc_to_fixed(pipe[i].stream->src.height, pipe[i].stream->dst.height);
+			data->rotation_angle[num_displays + 4] = bw_int_to_fixed(0);
+			data->bytes_per_pixel[num_displays + 4] = 4;
 		} else {
 			data->src_width[num_displays + 4] = bw_int_to_fixed(pipe[i].stream->timing.h_addressable);
 			data->pitch_in_pixels[num_displays + 4] = data->src_width[num_displays + 4];
-- 
cgit v1.1


From 5282cbe32ea5b8008f86a743922f018287113e2c Mon Sep 17 00:00:00 2001
From: Yongqiang Sun <yongqiang.sun@amd.com>
Date: Tue, 27 Mar 2018 10:05:10 -0400
Subject: drm/amd/display: Change disable backlight ramp change threshold from
 0 to maximum value.

Instead of user set brightness with range of percentage,
HLK test set brightness level with range of normal, this will result in
HLK test case set brightness from 0 to 255, DC set brightness with ramp is 0,
and disabled ramp change which will fail the HLK test.
Fix:
In case of unblank stream and turn on edp, change brightness level in
stream to 0xFFFFFFFF(actural maximum level is 0xFF), use that value as
a flag to recogonize this the case of resume from S3.

Signed-off-by: Yongqiang Sun <yongqiang.sun@amd.com>
Reviewed-by: Eric Yang <eric.yang2@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_link.c               | 2 +-
 drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c | 2 +-
 drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h           | 2 ++
 3 files changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
index d9efdd9..0cd286f 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
@@ -1982,7 +1982,7 @@ bool dc_link_set_backlight_level(const struct dc_link *link, uint32_t level,
 		return false;
 
 	if (stream) {
-		if (stream->bl_pwm_level == 0)
+		if (stream->bl_pwm_level == EDP_BACKLIGHT_RAMP_DISABLE_LEVEL)
 			frame_ramp = 0;
 
 		((struct dc_stream_state *)stream)->bl_pwm_level = level;
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
index db2d15d..78bf4fa 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
@@ -1036,7 +1036,7 @@ void dce110_unblank_stream(struct pipe_ctx *pipe_ctx,
 
 	if (link->local_sink && link->local_sink->sink_signal == SIGNAL_TYPE_EDP) {
 		link->dc->hwss.edp_backlight_control(link, true);
-		stream->bl_pwm_level = 0;
+		stream->bl_pwm_level = EDP_BACKLIGHT_RAMP_DISABLE_LEVEL;
 	}
 }
 void dce110_blank_stream(struct pipe_ctx *pipe_ctx)
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
index e764cba..f54d478 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
@@ -32,6 +32,8 @@
 #include "inc/hw/link_encoder.h"
 #include "core_status.h"
 
+#define EDP_BACKLIGHT_RAMP_DISABLE_LEVEL 0xFFFFFFFF
+
 enum pipe_gating_control {
 	PIPE_GATING_CONTROL_DISABLE = 0,
 	PIPE_GATING_CONTROL_ENABLE,
-- 
cgit v1.1


From dc002a2e4f6e164fd9e5c1353df795dc65784887 Mon Sep 17 00:00:00 2001
From: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Date: Mon, 26 Mar 2018 12:33:22 -0400
Subject: drm/amd/display: Update scaler v_active data if interlaced

Signed-off-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Reviewed-by: Nikola Cornij <Nikola.Cornij@amd.com>
Reviewed-by: Wesley Chalmers <Wesley.Chalmers@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_resource.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
index 50b84f6..eb8f479 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
@@ -844,6 +844,9 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx)
 	pipe_ctx->plane_res.scl_data.format = convert_pixel_format_to_dalsurface(
 			pipe_ctx->plane_state->format);
 
+	if (pipe_ctx->stream->timing.flags.INTERLACE)
+		pipe_ctx->stream->dst.height *= 2;
+
 	calculate_scaling_ratios(pipe_ctx);
 
 	calculate_viewport(pipe_ctx);
@@ -864,6 +867,8 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx)
 
 	pipe_ctx->plane_res.scl_data.h_active = timing->h_addressable + timing->h_border_left + timing->h_border_right;
 	pipe_ctx->plane_res.scl_data.v_active = timing->v_addressable + timing->v_border_top + timing->v_border_bottom;
+	if (pipe_ctx->stream->timing.flags.INTERLACE)
+		pipe_ctx->plane_res.scl_data.v_active *= 2;
 
 
 	/* Taps calculations */
@@ -909,6 +914,9 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx)
 				plane_state->dst_rect.x,
 				plane_state->dst_rect.y);
 
+	if (pipe_ctx->stream->timing.flags.INTERLACE)
+		pipe_ctx->stream->dst.height /= 2;
+
 	return res;
 }
 
-- 
cgit v1.1


From c5011872f6ad7fb8700117ae2fbdcd3ebbbe8402 Mon Sep 17 00:00:00 2001
From: Eric Bernstein <eric.bernstein@amd.com>
Date: Mon, 26 Mar 2018 16:28:03 -0400
Subject: drm/amd/display: Make DCN stream encoder shareable

Signed-off-by: Eric Bernstein <eric.bernstein@amd.com>
Reviewed-by: Charlene Liu <Charlene.Liu@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../drm/amd/display/dc/dce/dce_stream_encoder.c    |  39 +-
 .../amd/display/dc/dcn10/dcn10_stream_encoder.c    |  62 +-
 .../amd/display/dc/dcn10/dcn10_stream_encoder.h    | 646 ++++++++++-----------
 drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h  |  17 +
 4 files changed, 343 insertions(+), 421 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
index b85fda5..07c3242 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
@@ -26,27 +26,10 @@
 #include "dc_bios_types.h"
 #include "dce_stream_encoder.h"
 #include "reg_helper.h"
+#include "hw_shared.h"
+
 #define DC_LOGGER \
 		enc110->base.ctx->logger
-enum DP_PIXEL_ENCODING {
-DP_PIXEL_ENCODING_RGB444                 = 0x00000000,
-DP_PIXEL_ENCODING_YCBCR422               = 0x00000001,
-DP_PIXEL_ENCODING_YCBCR444               = 0x00000002,
-DP_PIXEL_ENCODING_RGB_WIDE_GAMUT         = 0x00000003,
-DP_PIXEL_ENCODING_Y_ONLY                 = 0x00000004,
-DP_PIXEL_ENCODING_YCBCR420               = 0x00000005,
-DP_PIXEL_ENCODING_RESERVED               = 0x00000006,
-};
-
-
-enum DP_COMPONENT_DEPTH {
-DP_COMPONENT_DEPTH_6BPC                  = 0x00000000,
-DP_COMPONENT_DEPTH_8BPC                  = 0x00000001,
-DP_COMPONENT_DEPTH_10BPC                 = 0x00000002,
-DP_COMPONENT_DEPTH_12BPC                 = 0x00000003,
-DP_COMPONENT_DEPTH_16BPC                 = 0x00000004,
-DP_COMPONENT_DEPTH_RESERVED              = 0x00000005,
-};
 
 
 #define REG(reg)\
@@ -314,11 +297,11 @@ static void dce110_stream_encoder_dp_set_stream_attribute(
 	switch (crtc_timing->pixel_encoding) {
 	case PIXEL_ENCODING_YCBCR422:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING,
-				DP_PIXEL_ENCODING_YCBCR422);
+				DP_PIXEL_ENCODING_TYPE_YCBCR422);
 		break;
 	case PIXEL_ENCODING_YCBCR444:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING,
-				DP_PIXEL_ENCODING_YCBCR444);
+				DP_PIXEL_ENCODING_TYPE_YCBCR444);
 
 		if (crtc_timing->flags.Y_ONLY)
 			if (crtc_timing->display_color_depth != COLOR_DEPTH_666)
@@ -326,7 +309,7 @@ static void dce110_stream_encoder_dp_set_stream_attribute(
 				 * Color depth of Y-only could be
 				 * 8, 10, 12, 16 bits */
 				REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING,
-						DP_PIXEL_ENCODING_Y_ONLY);
+						DP_PIXEL_ENCODING_TYPE_Y_ONLY);
 		/* Note: DP_MSA_MISC1 bit 7 is the indicator
 		 * of Y-only mode.
 		 * This bit is set in HW if register
@@ -334,7 +317,7 @@ static void dce110_stream_encoder_dp_set_stream_attribute(
 		break;
 	case PIXEL_ENCODING_YCBCR420:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING,
-				DP_PIXEL_ENCODING_YCBCR420);
+				DP_PIXEL_ENCODING_TYPE_YCBCR420);
 		if (enc110->se_mask->DP_VID_M_DOUBLE_VALUE_EN)
 			REG_UPDATE(DP_VID_TIMING, DP_VID_M_DOUBLE_VALUE_EN, 1);
 
@@ -345,7 +328,7 @@ static void dce110_stream_encoder_dp_set_stream_attribute(
 		break;
 	default:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING,
-				DP_PIXEL_ENCODING_RGB444);
+				DP_PIXEL_ENCODING_TYPE_RGB444);
 		break;
 	}
 
@@ -363,20 +346,20 @@ static void dce110_stream_encoder_dp_set_stream_attribute(
 		break;
 	case COLOR_DEPTH_888:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH,
-				DP_COMPONENT_DEPTH_8BPC);
+				DP_COMPONENT_PIXEL_DEPTH_8BPC);
 		break;
 	case COLOR_DEPTH_101010:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH,
-				DP_COMPONENT_DEPTH_10BPC);
+				DP_COMPONENT_PIXEL_DEPTH_10BPC);
 
 		break;
 	case COLOR_DEPTH_121212:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH,
-				DP_COMPONENT_DEPTH_12BPC);
+				DP_COMPONENT_PIXEL_DEPTH_12BPC);
 		break;
 	default:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH,
-				DP_COMPONENT_DEPTH_6BPC);
+				DP_COMPONENT_PIXEL_DEPTH_6BPC);
 		break;
 	}
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c
index 0413c70..9ec46f8 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c
@@ -26,29 +26,11 @@
 
 #include "dc_bios_types.h"
 #include "dcn10_stream_encoder.h"
-
 #include "reg_helper.h"
+#include "hw_shared.h"
+
 #define DC_LOGGER \
 		enc1->base.ctx->logger
-enum DP_PIXEL_ENCODING {
-DP_PIXEL_ENCODING_RGB444                 = 0x00000000,
-DP_PIXEL_ENCODING_YCBCR422               = 0x00000001,
-DP_PIXEL_ENCODING_YCBCR444               = 0x00000002,
-DP_PIXEL_ENCODING_RGB_WIDE_GAMUT         = 0x00000003,
-DP_PIXEL_ENCODING_Y_ONLY                 = 0x00000004,
-DP_PIXEL_ENCODING_YCBCR420               = 0x00000005,
-DP_PIXEL_ENCODING_RESERVED               = 0x00000006,
-};
-
-
-enum DP_COMPONENT_DEPTH {
-DP_COMPONENT_DEPTH_6BPC                  = 0x00000000,
-DP_COMPONENT_DEPTH_8BPC                  = 0x00000001,
-DP_COMPONENT_DEPTH_10BPC                 = 0x00000002,
-DP_COMPONENT_DEPTH_12BPC                 = 0x00000003,
-DP_COMPONENT_DEPTH_16BPC                 = 0x00000004,
-DP_COMPONENT_DEPTH_RESERVED              = 0x00000005,
-};
 
 
 #define REG(reg)\
@@ -70,7 +52,7 @@ enum {
 #define CTX \
 	enc1->base.ctx
 
-static void enc1_update_generic_info_packet(
+void enc1_update_generic_info_packet(
 	struct dcn10_stream_encoder *enc1,
 	uint32_t packet_index,
 	const struct dc_info_packet *info_packet)
@@ -260,7 +242,7 @@ static void enc1_update_hdmi_info_packet(
 }
 
 /* setup stream encoder in dp mode */
-static void enc1_stream_encoder_dp_set_stream_attribute(
+void enc1_stream_encoder_dp_set_stream_attribute(
 	struct stream_encoder *enc,
 	struct dc_crtc_timing *crtc_timing,
 	enum dc_color_space output_color_space)
@@ -284,11 +266,11 @@ static void enc1_stream_encoder_dp_set_stream_attribute(
 	switch (crtc_timing->pixel_encoding) {
 	case PIXEL_ENCODING_YCBCR422:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING,
-				DP_PIXEL_ENCODING_YCBCR422);
+				DP_PIXEL_ENCODING_TYPE_YCBCR422);
 		break;
 	case PIXEL_ENCODING_YCBCR444:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING,
-				DP_PIXEL_ENCODING_YCBCR444);
+				DP_PIXEL_ENCODING_TYPE_YCBCR444);
 
 		if (crtc_timing->flags.Y_ONLY)
 			if (crtc_timing->display_color_depth != COLOR_DEPTH_666)
@@ -297,7 +279,7 @@ static void enc1_stream_encoder_dp_set_stream_attribute(
 				 * 8, 10, 12, 16 bits
 				 */
 				REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING,
-						DP_PIXEL_ENCODING_Y_ONLY);
+						DP_PIXEL_ENCODING_TYPE_Y_ONLY);
 		/* Note: DP_MSA_MISC1 bit 7 is the indicator
 		 * of Y-only mode.
 		 * This bit is set in HW if register
@@ -306,12 +288,12 @@ static void enc1_stream_encoder_dp_set_stream_attribute(
 		break;
 	case PIXEL_ENCODING_YCBCR420:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING,
-				DP_PIXEL_ENCODING_YCBCR420);
+				DP_PIXEL_ENCODING_TYPE_YCBCR420);
 		REG_UPDATE(DP_VID_TIMING, DP_VID_N_MUL, 1);
 		break;
 	default:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING,
-				DP_PIXEL_ENCODING_RGB444);
+				DP_PIXEL_ENCODING_TYPE_RGB444);
 		break;
 	}
 
@@ -326,20 +308,20 @@ static void enc1_stream_encoder_dp_set_stream_attribute(
 		break;
 	case COLOR_DEPTH_888:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH,
-				DP_COMPONENT_DEPTH_8BPC);
+				DP_COMPONENT_PIXEL_DEPTH_8BPC);
 		break;
 	case COLOR_DEPTH_101010:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH,
-				DP_COMPONENT_DEPTH_10BPC);
+				DP_COMPONENT_PIXEL_DEPTH_10BPC);
 
 		break;
 	case COLOR_DEPTH_121212:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH,
-				DP_COMPONENT_DEPTH_12BPC);
+				DP_COMPONENT_PIXEL_DEPTH_12BPC);
 		break;
 	default:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH,
-				DP_COMPONENT_DEPTH_6BPC);
+				DP_COMPONENT_PIXEL_DEPTH_6BPC);
 		break;
 	}
 
@@ -485,7 +467,7 @@ static void enc1_stream_encoder_set_stream_attribute_helper(
 }
 
 /* setup stream encoder in hdmi mode */
-static void enc1_stream_encoder_hdmi_set_stream_attribute(
+void enc1_stream_encoder_hdmi_set_stream_attribute(
 	struct stream_encoder *enc,
 	struct dc_crtc_timing *crtc_timing,
 	int actual_pix_clk_khz,
@@ -591,7 +573,7 @@ static void enc1_stream_encoder_hdmi_set_stream_attribute(
 }
 
 /* setup stream encoder in dvi mode */
-static void enc1_stream_encoder_dvi_set_stream_attribute(
+void enc1_stream_encoder_dvi_set_stream_attribute(
 	struct stream_encoder *enc,
 	struct dc_crtc_timing *crtc_timing,
 	bool is_dual_link)
@@ -616,7 +598,7 @@ static void enc1_stream_encoder_dvi_set_stream_attribute(
 	enc1_stream_encoder_set_stream_attribute_helper(enc1, crtc_timing);
 }
 
-static void enc1_stream_encoder_set_mst_bandwidth(
+void enc1_stream_encoder_set_mst_bandwidth(
 	struct stream_encoder *enc,
 	struct fixed31_32 avg_time_slots_per_mtp)
 {
@@ -699,7 +681,7 @@ static void enc1_stream_encoder_stop_hdmi_info_packets(
 		HDMI_GENERIC1_SEND, 0);
 }
 
-static void enc1_stream_encoder_update_dp_info_packets(
+void enc1_stream_encoder_update_dp_info_packets(
 	struct stream_encoder *enc,
 	const struct encoder_info_frame *info_frame)
 {
@@ -742,7 +724,7 @@ static void enc1_stream_encoder_update_dp_info_packets(
 		REG_UPDATE(DP_SEC_CNTL, DP_SEC_STREAM_ENABLE, 1);
 }
 
-static void enc1_stream_encoder_stop_dp_info_packets(
+void enc1_stream_encoder_stop_dp_info_packets(
 	struct stream_encoder *enc)
 {
 	/* stop generic packets on DP */
@@ -770,7 +752,7 @@ static void enc1_stream_encoder_stop_dp_info_packets(
 
 }
 
-static void enc1_stream_encoder_dp_blank(
+void enc1_stream_encoder_dp_blank(
 	struct stream_encoder *enc)
 {
 	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
@@ -823,7 +805,7 @@ static void enc1_stream_encoder_dp_blank(
 }
 
 /* output video stream to link encoder */
-static void enc1_stream_encoder_dp_unblank(
+void enc1_stream_encoder_dp_unblank(
 	struct stream_encoder *enc,
 	const struct encoder_unblank_param *param)
 {
@@ -885,7 +867,7 @@ static void enc1_stream_encoder_dp_unblank(
 	REG_UPDATE(DP_VID_STREAM_CNTL, DP_VID_STREAM_ENABLE, true);
 }
 
-static void enc1_stream_encoder_set_avmute(
+void enc1_stream_encoder_set_avmute(
 	struct stream_encoder *enc,
 	bool enable)
 {
@@ -1442,7 +1424,7 @@ void enc1_se_hdmi_audio_disable(
 }
 
 
-static void enc1_setup_stereo_sync(
+void enc1_setup_stereo_sync(
 	struct stream_encoder *enc,
 	int tg_inst, bool enable)
 {
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.h
index 86f8ee5..6b3e4de 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.h
@@ -31,7 +31,8 @@
 #define DCN10STRENC_FROM_STRENC(stream_encoder)\
 	container_of(stream_encoder, struct dcn10_stream_encoder, base)
 
-#define SE_COMMON_REG_LIST_BASE(id) \
+#define SE_COMMON_DCN_REG_LIST(id) \
+	SRI(AFMT_CNTL, DIG, id), \
 	SRI(AFMT_GENERIC_0, DIG, id), \
 	SRI(AFMT_GENERIC_1, DIG, id), \
 	SRI(AFMT_GENERIC_2, DIG, id), \
@@ -43,6 +44,7 @@
 	SRI(AFMT_GENERIC_HDR, DIG, id), \
 	SRI(AFMT_INFOFRAME_CONTROL0, DIG, id), \
 	SRI(AFMT_VBI_PACKET_CONTROL, DIG, id), \
+	SRI(AFMT_VBI_PACKET_CONTROL1, DIG, id), \
 	SRI(AFMT_AUDIO_PACKET_CONTROL, DIG, id), \
 	SRI(AFMT_AUDIO_PACKET_CONTROL2, DIG, id), \
 	SRI(AFMT_AUDIO_SRC_CONTROL, DIG, id), \
@@ -51,9 +53,12 @@
 	SRI(AFMT_60958_2, DIG, id), \
 	SRI(DIG_FE_CNTL, DIG, id), \
 	SRI(HDMI_CONTROL, DIG, id), \
+	SRI(HDMI_DB_CONTROL, DIG, id), \
 	SRI(HDMI_GC, DIG, id), \
 	SRI(HDMI_GENERIC_PACKET_CONTROL0, DIG, id), \
 	SRI(HDMI_GENERIC_PACKET_CONTROL1, DIG, id), \
+	SRI(HDMI_GENERIC_PACKET_CONTROL2, DIG, id), \
+	SRI(HDMI_GENERIC_PACKET_CONTROL3, DIG, id), \
 	SRI(HDMI_INFOFRAME_CONTROL0, DIG, id), \
 	SRI(HDMI_INFOFRAME_CONTROL1, DIG, id), \
 	SRI(HDMI_VBI_PACKET_CONTROL, DIG, id), \
@@ -65,7 +70,13 @@
 	SRI(HDMI_ACR_44_1, DIG, id),\
 	SRI(HDMI_ACR_48_0, DIG, id),\
 	SRI(HDMI_ACR_48_1, DIG, id),\
-	SRI(TMDS_CNTL, DIG, id), \
+	SRI(DP_DB_CNTL, DP, id), \
+	SRI(DP_MSA_MISC, DP, id), \
+	SRI(DP_MSA_COLORIMETRY, DP, id), \
+	SRI(DP_MSA_TIMING_PARAM1, DP, id), \
+	SRI(DP_MSA_TIMING_PARAM2, DP, id), \
+	SRI(DP_MSA_TIMING_PARAM3, DP, id), \
+	SRI(DP_MSA_TIMING_PARAM4, DP, id), \
 	SRI(DP_MSE_RATE_CNTL, DP, id), \
 	SRI(DP_MSE_RATE_UPDATE, DP, id), \
 	SRI(DP_PIXEL_FORMAT, DP, id), \
@@ -79,19 +90,74 @@
 	SRI(DP_SEC_TIMESTAMP, DP, id)
 
 #define SE_DCN_REG_LIST(id)\
-	SE_COMMON_REG_LIST_BASE(id),\
-	SRI(AFMT_CNTL, DIG, id),\
-	SRI(AFMT_VBI_PACKET_CONTROL1, DIG, id),\
-	SRI(HDMI_GENERIC_PACKET_CONTROL2, DIG, id), \
-	SRI(HDMI_GENERIC_PACKET_CONTROL3, DIG, id), \
-	SRI(DP_DB_CNTL, DP, id), \
-	SRI(DP_MSA_MISC, DP, id), \
-	SRI(DP_MSA_COLORIMETRY, DP, id), \
-	SRI(DP_MSA_TIMING_PARAM1, DP, id), \
-	SRI(DP_MSA_TIMING_PARAM2, DP, id), \
-	SRI(DP_MSA_TIMING_PARAM3, DP, id), \
-	SRI(DP_MSA_TIMING_PARAM4, DP, id), \
-	SRI(HDMI_DB_CONTROL, DIG, id)
+	SE_COMMON_DCN_REG_LIST(id)
+
+
+struct dcn10_stream_enc_registers {
+	uint32_t AFMT_CNTL;
+	uint32_t AFMT_AVI_INFO0;
+	uint32_t AFMT_AVI_INFO1;
+	uint32_t AFMT_AVI_INFO2;
+	uint32_t AFMT_AVI_INFO3;
+	uint32_t AFMT_GENERIC_0;
+	uint32_t AFMT_GENERIC_1;
+	uint32_t AFMT_GENERIC_2;
+	uint32_t AFMT_GENERIC_3;
+	uint32_t AFMT_GENERIC_4;
+	uint32_t AFMT_GENERIC_5;
+	uint32_t AFMT_GENERIC_6;
+	uint32_t AFMT_GENERIC_7;
+	uint32_t AFMT_GENERIC_HDR;
+	uint32_t AFMT_INFOFRAME_CONTROL0;
+	uint32_t AFMT_VBI_PACKET_CONTROL;
+	uint32_t AFMT_VBI_PACKET_CONTROL1;
+	uint32_t AFMT_AUDIO_PACKET_CONTROL;
+	uint32_t AFMT_AUDIO_PACKET_CONTROL2;
+	uint32_t AFMT_AUDIO_SRC_CONTROL;
+	uint32_t AFMT_60958_0;
+	uint32_t AFMT_60958_1;
+	uint32_t AFMT_60958_2;
+	uint32_t DIG_FE_CNTL;
+	uint32_t DP_MSE_RATE_CNTL;
+	uint32_t DP_MSE_RATE_UPDATE;
+	uint32_t DP_PIXEL_FORMAT;
+	uint32_t DP_SEC_CNTL;
+	uint32_t DP_STEER_FIFO;
+	uint32_t DP_VID_M;
+	uint32_t DP_VID_N;
+	uint32_t DP_VID_STREAM_CNTL;
+	uint32_t DP_VID_TIMING;
+	uint32_t DP_SEC_AUD_N;
+	uint32_t DP_SEC_TIMESTAMP;
+	uint32_t HDMI_CONTROL;
+	uint32_t HDMI_GC;
+	uint32_t HDMI_GENERIC_PACKET_CONTROL0;
+	uint32_t HDMI_GENERIC_PACKET_CONTROL1;
+	uint32_t HDMI_GENERIC_PACKET_CONTROL2;
+	uint32_t HDMI_GENERIC_PACKET_CONTROL3;
+	uint32_t HDMI_GENERIC_PACKET_CONTROL4;
+	uint32_t HDMI_GENERIC_PACKET_CONTROL5;
+	uint32_t HDMI_INFOFRAME_CONTROL0;
+	uint32_t HDMI_INFOFRAME_CONTROL1;
+	uint32_t HDMI_VBI_PACKET_CONTROL;
+	uint32_t HDMI_AUDIO_PACKET_CONTROL;
+	uint32_t HDMI_ACR_PACKET_CONTROL;
+	uint32_t HDMI_ACR_32_0;
+	uint32_t HDMI_ACR_32_1;
+	uint32_t HDMI_ACR_44_0;
+	uint32_t HDMI_ACR_44_1;
+	uint32_t HDMI_ACR_48_0;
+	uint32_t HDMI_ACR_48_1;
+	uint32_t DP_DB_CNTL;
+	uint32_t DP_MSA_MISC;
+	uint32_t DP_MSA_COLORIMETRY;
+	uint32_t DP_MSA_TIMING_PARAM1;
+	uint32_t DP_MSA_TIMING_PARAM2;
+	uint32_t DP_MSA_TIMING_PARAM3;
+	uint32_t DP_MSA_TIMING_PARAM4;
+	uint32_t HDMI_DB_CONTROL;
+};
+
 
 #define SE_SF(reg_name, field_name, post_fix)\
 	.field_name = reg_name ## __ ## field_name ## post_fix
@@ -221,348 +287,151 @@
 	SE_SF(DIG0_HDMI_GENERIC_PACKET_CONTROL0, HDMI_GENERIC1_SEND, mask_sh),\
 	SE_SF(DIG0_HDMI_GENERIC_PACKET_CONTROL0, HDMI_GENERIC1_LINE, mask_sh)
 
+
+#define SE_REG_FIELD_LIST_DCN1_0(type) \
+	type AFMT_GENERIC_INDEX;\
+	type AFMT_GENERIC_HB0;\
+	type AFMT_GENERIC_HB1;\
+	type AFMT_GENERIC_HB2;\
+	type AFMT_GENERIC_HB3;\
+	type AFMT_GENERIC_LOCK_STATUS;\
+	type AFMT_GENERIC_CONFLICT;\
+	type AFMT_GENERIC_CONFLICT_CLR;\
+	type AFMT_GENERIC0_FRAME_UPDATE_PENDING;\
+	type AFMT_GENERIC1_FRAME_UPDATE_PENDING;\
+	type AFMT_GENERIC2_FRAME_UPDATE_PENDING;\
+	type AFMT_GENERIC3_FRAME_UPDATE_PENDING;\
+	type AFMT_GENERIC4_FRAME_UPDATE_PENDING;\
+	type AFMT_GENERIC5_FRAME_UPDATE_PENDING;\
+	type AFMT_GENERIC6_FRAME_UPDATE_PENDING;\
+	type AFMT_GENERIC7_FRAME_UPDATE_PENDING;\
+	type AFMT_GENERIC0_FRAME_UPDATE;\
+	type AFMT_GENERIC1_FRAME_UPDATE;\
+	type AFMT_GENERIC2_FRAME_UPDATE;\
+	type AFMT_GENERIC3_FRAME_UPDATE;\
+	type AFMT_GENERIC4_FRAME_UPDATE;\
+	type AFMT_GENERIC5_FRAME_UPDATE;\
+	type AFMT_GENERIC6_FRAME_UPDATE;\
+	type AFMT_GENERIC7_FRAME_UPDATE;\
+	type HDMI_GENERIC0_CONT;\
+	type HDMI_GENERIC0_SEND;\
+	type HDMI_GENERIC0_LINE;\
+	type HDMI_GENERIC1_CONT;\
+	type HDMI_GENERIC1_SEND;\
+	type HDMI_GENERIC1_LINE;\
+	type HDMI_GENERIC2_CONT;\
+	type HDMI_GENERIC2_SEND;\
+	type HDMI_GENERIC2_LINE;\
+	type HDMI_GENERIC3_CONT;\
+	type HDMI_GENERIC3_SEND;\
+	type HDMI_GENERIC3_LINE;\
+	type HDMI_GENERIC4_CONT;\
+	type HDMI_GENERIC4_SEND;\
+	type HDMI_GENERIC4_LINE;\
+	type HDMI_GENERIC5_CONT;\
+	type HDMI_GENERIC5_SEND;\
+	type HDMI_GENERIC5_LINE;\
+	type HDMI_GENERIC6_CONT;\
+	type HDMI_GENERIC6_SEND;\
+	type HDMI_GENERIC6_LINE;\
+	type HDMI_GENERIC7_CONT;\
+	type HDMI_GENERIC7_SEND;\
+	type HDMI_GENERIC7_LINE;\
+	type DP_PIXEL_ENCODING;\
+	type DP_COMPONENT_DEPTH;\
+	type HDMI_PACKET_GEN_VERSION;\
+	type HDMI_KEEPOUT_MODE;\
+	type HDMI_DEEP_COLOR_ENABLE;\
+	type HDMI_CLOCK_CHANNEL_RATE;\
+	type HDMI_DEEP_COLOR_DEPTH;\
+	type HDMI_GC_CONT;\
+	type HDMI_GC_SEND;\
+	type HDMI_NULL_SEND;\
+	type HDMI_DATA_SCRAMBLE_EN;\
+	type HDMI_AUDIO_INFO_SEND;\
+	type AFMT_AUDIO_INFO_UPDATE;\
+	type HDMI_AUDIO_INFO_LINE;\
+	type HDMI_GC_AVMUTE;\
+	type DP_MSE_RATE_X;\
+	type DP_MSE_RATE_Y;\
+	type DP_MSE_RATE_UPDATE_PENDING;\
+	type DP_SEC_GSP0_ENABLE;\
+	type DP_SEC_STREAM_ENABLE;\
+	type DP_SEC_GSP1_ENABLE;\
+	type DP_SEC_GSP2_ENABLE;\
+	type DP_SEC_GSP3_ENABLE;\
+	type DP_SEC_GSP4_ENABLE;\
+	type DP_SEC_GSP5_ENABLE;\
+	type DP_SEC_GSP6_ENABLE;\
+	type DP_SEC_GSP7_ENABLE;\
+	type DP_SEC_MPG_ENABLE;\
+	type DP_VID_STREAM_DIS_DEFER;\
+	type DP_VID_STREAM_ENABLE;\
+	type DP_VID_STREAM_STATUS;\
+	type DP_STEER_FIFO_RESET;\
+	type DP_VID_M_N_GEN_EN;\
+	type DP_VID_N;\
+	type DP_VID_M;\
+	type DIG_START;\
+	type AFMT_AUDIO_SRC_SELECT;\
+	type AFMT_AUDIO_CHANNEL_ENABLE;\
+	type HDMI_AUDIO_PACKETS_PER_LINE;\
+	type HDMI_AUDIO_DELAY_EN;\
+	type AFMT_60958_CS_UPDATE;\
+	type AFMT_AUDIO_LAYOUT_OVRD;\
+	type AFMT_60958_OSF_OVRD;\
+	type HDMI_ACR_AUTO_SEND;\
+	type HDMI_ACR_SOURCE;\
+	type HDMI_ACR_AUDIO_PRIORITY;\
+	type HDMI_ACR_CTS_32;\
+	type HDMI_ACR_N_32;\
+	type HDMI_ACR_CTS_44;\
+	type HDMI_ACR_N_44;\
+	type HDMI_ACR_CTS_48;\
+	type HDMI_ACR_N_48;\
+	type AFMT_60958_CS_CHANNEL_NUMBER_L;\
+	type AFMT_60958_CS_CLOCK_ACCURACY;\
+	type AFMT_60958_CS_CHANNEL_NUMBER_R;\
+	type AFMT_60958_CS_CHANNEL_NUMBER_2;\
+	type AFMT_60958_CS_CHANNEL_NUMBER_3;\
+	type AFMT_60958_CS_CHANNEL_NUMBER_4;\
+	type AFMT_60958_CS_CHANNEL_NUMBER_5;\
+	type AFMT_60958_CS_CHANNEL_NUMBER_6;\
+	type AFMT_60958_CS_CHANNEL_NUMBER_7;\
+	type DP_SEC_AUD_N;\
+	type DP_SEC_TIMESTAMP_MODE;\
+	type DP_SEC_ASP_ENABLE;\
+	type DP_SEC_ATP_ENABLE;\
+	type DP_SEC_AIP_ENABLE;\
+	type DP_SEC_ACM_ENABLE;\
+	type AFMT_AUDIO_SAMPLE_SEND;\
+	type AFMT_AUDIO_CLOCK_EN;\
+	type TMDS_PIXEL_ENCODING;\
+	type TMDS_COLOR_FORMAT;\
+	type DIG_STEREOSYNC_SELECT;\
+	type DIG_STEREOSYNC_GATE_EN;\
+	type DP_DB_DISABLE;\
+	type DP_MSA_MISC0;\
+	type DP_MSA_HTOTAL;\
+	type DP_MSA_VTOTAL;\
+	type DP_MSA_HSTART;\
+	type DP_MSA_VSTART;\
+	type DP_MSA_HSYNCWIDTH;\
+	type DP_MSA_HSYNCPOLARITY;\
+	type DP_MSA_VSYNCWIDTH;\
+	type DP_MSA_VSYNCPOLARITY;\
+	type DP_MSA_HWIDTH;\
+	type DP_MSA_VHEIGHT;\
+	type HDMI_DB_DISABLE;\
+	type DP_VID_N_MUL;\
+	type DP_VID_M_DOUBLE_VALUE_EN
+
 struct dcn10_stream_encoder_shift {
-	uint8_t AFMT_GENERIC_INDEX;
-	uint8_t AFMT_GENERIC_HB0;
-	uint8_t AFMT_GENERIC_HB1;
-	uint8_t AFMT_GENERIC_HB2;
-	uint8_t AFMT_GENERIC_HB3;
-	uint8_t AFMT_GENERIC_LOCK_STATUS;
-	uint8_t AFMT_GENERIC_CONFLICT;
-	uint8_t AFMT_GENERIC_CONFLICT_CLR;
-	uint8_t AFMT_GENERIC0_FRAME_UPDATE_PENDING;
-	uint8_t AFMT_GENERIC1_FRAME_UPDATE_PENDING;
-	uint8_t AFMT_GENERIC2_FRAME_UPDATE_PENDING;
-	uint8_t AFMT_GENERIC3_FRAME_UPDATE_PENDING;
-	uint8_t AFMT_GENERIC4_FRAME_UPDATE_PENDING;
-	uint8_t AFMT_GENERIC5_FRAME_UPDATE_PENDING;
-	uint8_t AFMT_GENERIC6_FRAME_UPDATE_PENDING;
-	uint8_t AFMT_GENERIC7_FRAME_UPDATE_PENDING;
-	uint8_t AFMT_GENERIC0_FRAME_UPDATE;
-	uint8_t AFMT_GENERIC1_FRAME_UPDATE;
-	uint8_t AFMT_GENERIC2_FRAME_UPDATE;
-	uint8_t AFMT_GENERIC3_FRAME_UPDATE;
-	uint8_t AFMT_GENERIC4_FRAME_UPDATE;
-	uint8_t AFMT_GENERIC5_FRAME_UPDATE;
-	uint8_t AFMT_GENERIC6_FRAME_UPDATE;
-	uint8_t AFMT_GENERIC7_FRAME_UPDATE;
-	uint8_t HDMI_GENERIC0_CONT;
-	uint8_t HDMI_GENERIC0_SEND;
-	uint8_t HDMI_GENERIC0_LINE;
-	uint8_t HDMI_GENERIC1_CONT;
-	uint8_t HDMI_GENERIC1_SEND;
-	uint8_t HDMI_GENERIC1_LINE;
-	uint8_t HDMI_GENERIC2_CONT;
-	uint8_t HDMI_GENERIC2_SEND;
-	uint8_t HDMI_GENERIC2_LINE;
-	uint8_t HDMI_GENERIC3_CONT;
-	uint8_t HDMI_GENERIC3_SEND;
-	uint8_t HDMI_GENERIC3_LINE;
-	uint8_t HDMI_GENERIC4_CONT;
-	uint8_t HDMI_GENERIC4_SEND;
-	uint8_t HDMI_GENERIC4_LINE;
-	uint8_t HDMI_GENERIC5_CONT;
-	uint8_t HDMI_GENERIC5_SEND;
-	uint8_t HDMI_GENERIC5_LINE;
-	uint8_t HDMI_GENERIC6_CONT;
-	uint8_t HDMI_GENERIC6_SEND;
-	uint8_t HDMI_GENERIC6_LINE;
-	uint8_t HDMI_GENERIC7_CONT;
-	uint8_t HDMI_GENERIC7_SEND;
-	uint8_t HDMI_GENERIC7_LINE;
-	uint8_t DP_PIXEL_ENCODING;
-	uint8_t DP_COMPONENT_DEPTH;
-	uint8_t HDMI_PACKET_GEN_VERSION;
-	uint8_t HDMI_KEEPOUT_MODE;
-	uint8_t HDMI_DEEP_COLOR_ENABLE;
-	uint8_t HDMI_CLOCK_CHANNEL_RATE;
-	uint8_t HDMI_DEEP_COLOR_DEPTH;
-	uint8_t HDMI_GC_CONT;
-	uint8_t HDMI_GC_SEND;
-	uint8_t HDMI_NULL_SEND;
-	uint8_t HDMI_DATA_SCRAMBLE_EN;
-	uint8_t HDMI_AUDIO_INFO_SEND;
-	uint8_t AFMT_AUDIO_INFO_UPDATE;
-	uint8_t HDMI_AUDIO_INFO_LINE;
-	uint8_t HDMI_GC_AVMUTE;
-	uint8_t DP_MSE_RATE_X;
-	uint8_t DP_MSE_RATE_Y;
-	uint8_t DP_MSE_RATE_UPDATE_PENDING;
-	uint8_t DP_SEC_GSP0_ENABLE;
-	uint8_t DP_SEC_STREAM_ENABLE;
-	uint8_t DP_SEC_GSP1_ENABLE;
-	uint8_t DP_SEC_GSP2_ENABLE;
-	uint8_t DP_SEC_GSP3_ENABLE;
-	uint8_t DP_SEC_GSP4_ENABLE;
-	uint8_t DP_SEC_GSP5_ENABLE;
-	uint8_t DP_SEC_GSP6_ENABLE;
-	uint8_t DP_SEC_GSP7_ENABLE;
-	uint8_t DP_SEC_MPG_ENABLE;
-	uint8_t DP_VID_STREAM_DIS_DEFER;
-	uint8_t DP_VID_STREAM_ENABLE;
-	uint8_t DP_VID_STREAM_STATUS;
-	uint8_t DP_STEER_FIFO_RESET;
-	uint8_t DP_VID_M_N_GEN_EN;
-	uint8_t DP_VID_N;
-	uint8_t DP_VID_M;
-	uint8_t DIG_START;
-	uint8_t AFMT_AUDIO_SRC_SELECT;
-	uint8_t AFMT_AUDIO_CHANNEL_ENABLE;
-	uint8_t HDMI_AUDIO_PACKETS_PER_LINE;
-	uint8_t HDMI_AUDIO_DELAY_EN;
-	uint8_t AFMT_60958_CS_UPDATE;
-	uint8_t AFMT_AUDIO_LAYOUT_OVRD;
-	uint8_t AFMT_60958_OSF_OVRD;
-	uint8_t HDMI_ACR_AUTO_SEND;
-	uint8_t HDMI_ACR_SOURCE;
-	uint8_t HDMI_ACR_AUDIO_PRIORITY;
-	uint8_t HDMI_ACR_CTS_32;
-	uint8_t HDMI_ACR_N_32;
-	uint8_t HDMI_ACR_CTS_44;
-	uint8_t HDMI_ACR_N_44;
-	uint8_t HDMI_ACR_CTS_48;
-	uint8_t HDMI_ACR_N_48;
-	uint8_t AFMT_60958_CS_CHANNEL_NUMBER_L;
-	uint8_t AFMT_60958_CS_CLOCK_ACCURACY;
-	uint8_t AFMT_60958_CS_CHANNEL_NUMBER_R;
-	uint8_t AFMT_60958_CS_CHANNEL_NUMBER_2;
-	uint8_t AFMT_60958_CS_CHANNEL_NUMBER_3;
-	uint8_t AFMT_60958_CS_CHANNEL_NUMBER_4;
-	uint8_t AFMT_60958_CS_CHANNEL_NUMBER_5;
-	uint8_t AFMT_60958_CS_CHANNEL_NUMBER_6;
-	uint8_t AFMT_60958_CS_CHANNEL_NUMBER_7;
-	uint8_t DP_SEC_AUD_N;
-	uint8_t DP_SEC_TIMESTAMP_MODE;
-	uint8_t DP_SEC_ASP_ENABLE;
-	uint8_t DP_SEC_ATP_ENABLE;
-	uint8_t DP_SEC_AIP_ENABLE;
-	uint8_t DP_SEC_ACM_ENABLE;
-	uint8_t AFMT_AUDIO_SAMPLE_SEND;
-	uint8_t AFMT_AUDIO_CLOCK_EN;
-	uint8_t TMDS_PIXEL_ENCODING;
-	uint8_t TMDS_COLOR_FORMAT;
-	uint8_t DIG_STEREOSYNC_SELECT;
-	uint8_t DIG_STEREOSYNC_GATE_EN;
-	uint8_t DP_DB_DISABLE;
-	uint8_t DP_MSA_MISC0;
-	uint8_t DP_MSA_HTOTAL;
-	uint8_t DP_MSA_VTOTAL;
-	uint8_t DP_MSA_HSTART;
-	uint8_t DP_MSA_VSTART;
-	uint8_t DP_MSA_HSYNCWIDTH;
-	uint8_t DP_MSA_HSYNCPOLARITY;
-	uint8_t DP_MSA_VSYNCWIDTH;
-	uint8_t DP_MSA_VSYNCPOLARITY;
-	uint8_t DP_MSA_HWIDTH;
-	uint8_t DP_MSA_VHEIGHT;
-	uint8_t HDMI_DB_DISABLE;
-	uint8_t DP_VID_N_MUL;
-	uint8_t DP_VID_M_DOUBLE_VALUE_EN;
+	SE_REG_FIELD_LIST_DCN1_0(uint8_t);
 };
 
 struct dcn10_stream_encoder_mask {
-	uint32_t AFMT_GENERIC_INDEX;
-	uint32_t AFMT_GENERIC_HB0;
-	uint32_t AFMT_GENERIC_HB1;
-	uint32_t AFMT_GENERIC_HB2;
-	uint32_t AFMT_GENERIC_HB3;
-	uint32_t AFMT_GENERIC_LOCK_STATUS;
-	uint32_t AFMT_GENERIC_CONFLICT;
-	uint32_t AFMT_GENERIC_CONFLICT_CLR;
-	uint32_t AFMT_GENERIC0_FRAME_UPDATE_PENDING;
-	uint32_t AFMT_GENERIC1_FRAME_UPDATE_PENDING;
-	uint32_t AFMT_GENERIC2_FRAME_UPDATE_PENDING;
-	uint32_t AFMT_GENERIC3_FRAME_UPDATE_PENDING;
-	uint32_t AFMT_GENERIC4_FRAME_UPDATE_PENDING;
-	uint32_t AFMT_GENERIC5_FRAME_UPDATE_PENDING;
-	uint32_t AFMT_GENERIC6_FRAME_UPDATE_PENDING;
-	uint32_t AFMT_GENERIC7_FRAME_UPDATE_PENDING;
-	uint32_t AFMT_GENERIC0_FRAME_UPDATE;
-	uint32_t AFMT_GENERIC1_FRAME_UPDATE;
-	uint32_t AFMT_GENERIC2_FRAME_UPDATE;
-	uint32_t AFMT_GENERIC3_FRAME_UPDATE;
-	uint32_t AFMT_GENERIC4_FRAME_UPDATE;
-	uint32_t AFMT_GENERIC5_FRAME_UPDATE;
-	uint32_t AFMT_GENERIC6_FRAME_UPDATE;
-	uint32_t AFMT_GENERIC7_FRAME_UPDATE;
-	uint32_t HDMI_GENERIC0_CONT;
-	uint32_t HDMI_GENERIC0_SEND;
-	uint32_t HDMI_GENERIC0_LINE;
-	uint32_t HDMI_GENERIC1_CONT;
-	uint32_t HDMI_GENERIC1_SEND;
-	uint32_t HDMI_GENERIC1_LINE;
-	uint32_t HDMI_GENERIC2_CONT;
-	uint32_t HDMI_GENERIC2_SEND;
-	uint32_t HDMI_GENERIC2_LINE;
-	uint32_t HDMI_GENERIC3_CONT;
-	uint32_t HDMI_GENERIC3_SEND;
-	uint32_t HDMI_GENERIC3_LINE;
-	uint32_t HDMI_GENERIC4_CONT;
-	uint32_t HDMI_GENERIC4_SEND;
-	uint32_t HDMI_GENERIC4_LINE;
-	uint32_t HDMI_GENERIC5_CONT;
-	uint32_t HDMI_GENERIC5_SEND;
-	uint32_t HDMI_GENERIC5_LINE;
-	uint32_t HDMI_GENERIC6_CONT;
-	uint32_t HDMI_GENERIC6_SEND;
-	uint32_t HDMI_GENERIC6_LINE;
-	uint32_t HDMI_GENERIC7_CONT;
-	uint32_t HDMI_GENERIC7_SEND;
-	uint32_t HDMI_GENERIC7_LINE;
-	uint32_t DP_PIXEL_ENCODING;
-	uint32_t DP_COMPONENT_DEPTH;
-	uint32_t HDMI_PACKET_GEN_VERSION;
-	uint32_t HDMI_KEEPOUT_MODE;
-	uint32_t HDMI_DEEP_COLOR_ENABLE;
-	uint32_t HDMI_CLOCK_CHANNEL_RATE;
-	uint32_t HDMI_DEEP_COLOR_DEPTH;
-	uint32_t HDMI_GC_CONT;
-	uint32_t HDMI_GC_SEND;
-	uint32_t HDMI_NULL_SEND;
-	uint32_t HDMI_DATA_SCRAMBLE_EN;
-	uint32_t HDMI_AUDIO_INFO_SEND;
-	uint32_t AFMT_AUDIO_INFO_UPDATE;
-	uint32_t HDMI_AUDIO_INFO_LINE;
-	uint32_t HDMI_GC_AVMUTE;
-	uint32_t DP_MSE_RATE_X;
-	uint32_t DP_MSE_RATE_Y;
-	uint32_t DP_MSE_RATE_UPDATE_PENDING;
-	uint32_t DP_SEC_GSP0_ENABLE;
-	uint32_t DP_SEC_STREAM_ENABLE;
-	uint32_t DP_SEC_GSP1_ENABLE;
-	uint32_t DP_SEC_GSP2_ENABLE;
-	uint32_t DP_SEC_GSP3_ENABLE;
-	uint32_t DP_SEC_GSP4_ENABLE;
-	uint32_t DP_SEC_GSP5_ENABLE;
-	uint32_t DP_SEC_GSP6_ENABLE;
-	uint32_t DP_SEC_GSP7_ENABLE;
-	uint32_t DP_SEC_MPG_ENABLE;
-	uint32_t DP_VID_STREAM_DIS_DEFER;
-	uint32_t DP_VID_STREAM_ENABLE;
-	uint32_t DP_VID_STREAM_STATUS;
-	uint32_t DP_STEER_FIFO_RESET;
-	uint32_t DP_VID_M_N_GEN_EN;
-	uint32_t DP_VID_N;
-	uint32_t DP_VID_M;
-	uint32_t DIG_START;
-	uint32_t AFMT_AUDIO_SRC_SELECT;
-	uint32_t AFMT_AUDIO_CHANNEL_ENABLE;
-	uint32_t HDMI_AUDIO_PACKETS_PER_LINE;
-	uint32_t HDMI_AUDIO_DELAY_EN;
-	uint32_t AFMT_60958_CS_UPDATE;
-	uint32_t AFMT_AUDIO_LAYOUT_OVRD;
-	uint32_t AFMT_60958_OSF_OVRD;
-	uint32_t HDMI_ACR_AUTO_SEND;
-	uint32_t HDMI_ACR_SOURCE;
-	uint32_t HDMI_ACR_AUDIO_PRIORITY;
-	uint32_t HDMI_ACR_CTS_32;
-	uint32_t HDMI_ACR_N_32;
-	uint32_t HDMI_ACR_CTS_44;
-	uint32_t HDMI_ACR_N_44;
-	uint32_t HDMI_ACR_CTS_48;
-	uint32_t HDMI_ACR_N_48;
-	uint32_t AFMT_60958_CS_CHANNEL_NUMBER_L;
-	uint32_t AFMT_60958_CS_CLOCK_ACCURACY;
-	uint32_t AFMT_60958_CS_CHANNEL_NUMBER_R;
-	uint32_t AFMT_60958_CS_CHANNEL_NUMBER_2;
-	uint32_t AFMT_60958_CS_CHANNEL_NUMBER_3;
-	uint32_t AFMT_60958_CS_CHANNEL_NUMBER_4;
-	uint32_t AFMT_60958_CS_CHANNEL_NUMBER_5;
-	uint32_t AFMT_60958_CS_CHANNEL_NUMBER_6;
-	uint32_t AFMT_60958_CS_CHANNEL_NUMBER_7;
-	uint32_t DP_SEC_AUD_N;
-	uint32_t DP_SEC_TIMESTAMP_MODE;
-	uint32_t DP_SEC_ASP_ENABLE;
-	uint32_t DP_SEC_ATP_ENABLE;
-	uint32_t DP_SEC_AIP_ENABLE;
-	uint32_t DP_SEC_ACM_ENABLE;
-	uint32_t AFMT_AUDIO_SAMPLE_SEND;
-	uint32_t AFMT_AUDIO_CLOCK_EN;
-	uint32_t TMDS_PIXEL_ENCODING;
-	uint32_t DIG_STEREOSYNC_SELECT;
-	uint32_t DIG_STEREOSYNC_GATE_EN;
-	uint32_t TMDS_COLOR_FORMAT;
-	uint32_t DP_DB_DISABLE;
-	uint32_t DP_MSA_MISC0;
-	uint32_t DP_MSA_HTOTAL;
-	uint32_t DP_MSA_VTOTAL;
-	uint32_t DP_MSA_HSTART;
-	uint32_t DP_MSA_VSTART;
-	uint32_t DP_MSA_HSYNCWIDTH;
-	uint32_t DP_MSA_HSYNCPOLARITY;
-	uint32_t DP_MSA_VSYNCWIDTH;
-	uint32_t DP_MSA_VSYNCPOLARITY;
-	uint32_t DP_MSA_HWIDTH;
-	uint32_t DP_MSA_VHEIGHT;
-	uint32_t HDMI_DB_DISABLE;
-	uint32_t DP_VID_N_MUL;
-	uint32_t DP_VID_M_DOUBLE_VALUE_EN;
-};
-
-struct dcn10_stream_enc_registers {
-	uint32_t AFMT_CNTL;
-	uint32_t AFMT_AVI_INFO0;
-	uint32_t AFMT_AVI_INFO1;
-	uint32_t AFMT_AVI_INFO2;
-	uint32_t AFMT_AVI_INFO3;
-	uint32_t AFMT_GENERIC_0;
-	uint32_t AFMT_GENERIC_1;
-	uint32_t AFMT_GENERIC_2;
-	uint32_t AFMT_GENERIC_3;
-	uint32_t AFMT_GENERIC_4;
-	uint32_t AFMT_GENERIC_5;
-	uint32_t AFMT_GENERIC_6;
-	uint32_t AFMT_GENERIC_7;
-	uint32_t AFMT_GENERIC_HDR;
-	uint32_t AFMT_INFOFRAME_CONTROL0;
-	uint32_t AFMT_VBI_PACKET_CONTROL;
-	uint32_t AFMT_VBI_PACKET_CONTROL1;
-	uint32_t AFMT_AUDIO_PACKET_CONTROL;
-	uint32_t AFMT_AUDIO_PACKET_CONTROL2;
-	uint32_t AFMT_AUDIO_SRC_CONTROL;
-	uint32_t AFMT_60958_0;
-	uint32_t AFMT_60958_1;
-	uint32_t AFMT_60958_2;
-	uint32_t DIG_FE_CNTL;
-	uint32_t DP_MSE_RATE_CNTL;
-	uint32_t DP_MSE_RATE_UPDATE;
-	uint32_t DP_PIXEL_FORMAT;
-	uint32_t DP_SEC_CNTL;
-	uint32_t DP_STEER_FIFO;
-	uint32_t DP_VID_M;
-	uint32_t DP_VID_N;
-	uint32_t DP_VID_STREAM_CNTL;
-	uint32_t DP_VID_TIMING;
-	uint32_t DP_SEC_AUD_N;
-	uint32_t DP_SEC_TIMESTAMP;
-	uint32_t HDMI_CONTROL;
-	uint32_t HDMI_GC;
-	uint32_t HDMI_GENERIC_PACKET_CONTROL0;
-	uint32_t HDMI_GENERIC_PACKET_CONTROL1;
-	uint32_t HDMI_GENERIC_PACKET_CONTROL2;
-	uint32_t HDMI_GENERIC_PACKET_CONTROL3;
-	uint32_t HDMI_GENERIC_PACKET_CONTROL4;
-	uint32_t HDMI_GENERIC_PACKET_CONTROL5;
-	uint32_t HDMI_INFOFRAME_CONTROL0;
-	uint32_t HDMI_INFOFRAME_CONTROL1;
-	uint32_t HDMI_VBI_PACKET_CONTROL;
-	uint32_t HDMI_AUDIO_PACKET_CONTROL;
-	uint32_t HDMI_ACR_PACKET_CONTROL;
-	uint32_t HDMI_ACR_32_0;
-	uint32_t HDMI_ACR_32_1;
-	uint32_t HDMI_ACR_44_0;
-	uint32_t HDMI_ACR_44_1;
-	uint32_t HDMI_ACR_48_0;
-	uint32_t HDMI_ACR_48_1;
-	uint32_t TMDS_CNTL;
-	uint32_t DP_DB_CNTL;
-	uint32_t DP_MSA_MISC;
-	uint32_t DP_MSA_COLORIMETRY;
-	uint32_t DP_MSA_TIMING_PARAM1;
-	uint32_t DP_MSA_TIMING_PARAM2;
-	uint32_t DP_MSA_TIMING_PARAM3;
-	uint32_t DP_MSA_TIMING_PARAM4;
-	uint32_t HDMI_DB_CONTROL;
+	SE_REG_FIELD_LIST_DCN1_0(uint32_t);
 };
 
 struct dcn10_stream_encoder {
@@ -581,4 +450,75 @@ void dcn10_stream_encoder_construct(
 	const struct dcn10_stream_encoder_shift *se_shift,
 	const struct dcn10_stream_encoder_mask *se_mask);
 
+void enc1_update_generic_info_packet(
+	struct dcn10_stream_encoder *enc1,
+	uint32_t packet_index,
+	const struct dc_info_packet *info_packet);
+
+void enc1_stream_encoder_dp_set_stream_attribute(
+	struct stream_encoder *enc,
+	struct dc_crtc_timing *crtc_timing,
+	enum dc_color_space output_color_space);
+
+void enc1_stream_encoder_hdmi_set_stream_attribute(
+	struct stream_encoder *enc,
+	struct dc_crtc_timing *crtc_timing,
+	int actual_pix_clk_khz,
+	bool enable_audio);
+
+void enc1_stream_encoder_dvi_set_stream_attribute(
+	struct stream_encoder *enc,
+	struct dc_crtc_timing *crtc_timing,
+	bool is_dual_link);
+
+void enc1_stream_encoder_set_mst_bandwidth(
+	struct stream_encoder *enc,
+	struct fixed31_32 avg_time_slots_per_mtp);
+
+void enc1_stream_encoder_update_dp_info_packets(
+	struct stream_encoder *enc,
+	const struct encoder_info_frame *info_frame);
+
+void enc1_stream_encoder_stop_dp_info_packets(
+	struct stream_encoder *enc);
+
+void enc1_stream_encoder_dp_blank(
+	struct stream_encoder *enc);
+
+void enc1_stream_encoder_dp_unblank(
+	struct stream_encoder *enc,
+	const struct encoder_unblank_param *param);
+
+void enc1_setup_stereo_sync(
+	struct stream_encoder *enc,
+	int tg_inst, bool enable);
+
+void enc1_stream_encoder_set_avmute(
+	struct stream_encoder *enc,
+	bool enable);
+
+void enc1_se_audio_mute_control(
+	struct stream_encoder *enc,
+	bool mute);
+
+void enc1_se_dp_audio_setup(
+	struct stream_encoder *enc,
+	unsigned int az_inst,
+	struct audio_info *info);
+
+void enc1_se_dp_audio_enable(
+	struct stream_encoder *enc);
+
+void enc1_se_dp_audio_disable(
+	struct stream_encoder *enc);
+
+void enc1_se_hdmi_audio_setup(
+	struct stream_encoder *enc,
+	unsigned int az_inst,
+	struct audio_info *info,
+	struct audio_crtc_info *audio_crtc_info);
+
+void enc1_se_hdmi_audio_disable(
+	struct stream_encoder *enc);
+
 #endif /* __DC_STREAM_ENCODER_DCN10_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h b/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h
index 015e209..93da445 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h
@@ -191,6 +191,23 @@ enum controller_dp_test_pattern {
 	CONTROLLER_DP_TEST_PATTERN_COLORSQUARES_CEA
 };
 
+enum dp_pixel_encoding_type {
+	DP_PIXEL_ENCODING_TYPE_RGB444		= 0x00000000,
+	DP_PIXEL_ENCODING_TYPE_YCBCR422		= 0x00000001,
+	DP_PIXEL_ENCODING_TYPE_YCBCR444		= 0x00000002,
+	DP_PIXEL_ENCODING_TYPE_RGB_WIDE_GAMUT	= 0x00000003,
+	DP_PIXEL_ENCODING_TYPE_Y_ONLY		= 0x00000004,
+	DP_PIXEL_ENCODING_TYPE_YCBCR420		= 0x00000005
+};
+
+enum dp_component_depth {
+	DP_COMPONENT_PIXEL_DEPTH_6BPC		= 0x00000000,
+	DP_COMPONENT_PIXEL_DEPTH_8BPC		= 0x00000001,
+	DP_COMPONENT_PIXEL_DEPTH_10BPC		= 0x00000002,
+	DP_COMPONENT_PIXEL_DEPTH_12BPC		= 0x00000003,
+	DP_COMPONENT_PIXEL_DEPTH_16BPC		= 0x00000004
+};
+
 enum dc_lut_mode {
 	LUT_BYPASS,
 	LUT_RAM_A,
-- 
cgit v1.1


From 35ad2254cb7d0a46f135eb57990ca6618f79510b Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Thu, 29 Mar 2018 11:23:37 -0400
Subject: drm/amd/display: csc updates require FULL update

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc.c | 12 ++++++++++--
 drivers/gpu/drm/amd/display/dc/dc.h      |  1 +
 2 files changed, 11 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c
index b331d9e..8f09f3a 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc.c
@@ -1154,12 +1154,20 @@ static enum surface_update_type det_surface_update(const struct dc *dc,
 	if (u->input_csc_color_matrix)
 		update_flags->bits.input_csc_change = 1;
 
-	if (update_flags->bits.in_transfer_func_change
-			|| update_flags->bits.input_csc_change) {
+	if (u->coeff_reduction_factor)
+		update_flags->bits.coeff_reduction_change = 1;
+
+	if (update_flags->bits.in_transfer_func_change) {
 		type = UPDATE_TYPE_MED;
 		elevate_update_type(&overall_type, type);
 	}
 
+	if (update_flags->bits.input_csc_change
+			|| update_flags->bits.coeff_reduction_change) {
+		type = UPDATE_TYPE_FULL;
+		elevate_update_type(&overall_type, type);
+	}
+
 	return overall_type;
 }
 
diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index 63817ed..7d1a3c5 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -429,6 +429,7 @@ union surface_update_flags {
 		uint32_t position_change:1;
 		uint32_t in_transfer_func_change:1;
 		uint32_t input_csc_change:1;
+		uint32_t coeff_reduction_change:1;
 		uint32_t output_tf_change:1;
 		uint32_t pixel_format_change:1;
 
-- 
cgit v1.1


From bb33b1842c3f2592a9be4e80c9d4afe6251a5da6 Mon Sep 17 00:00:00 2001
From: Roman Li <roman.li@amd.com>
Date: Thu, 29 Mar 2018 11:14:25 -0400
Subject: drm/amd/display: Fix FBC text console corruption

Signed-off-by: Roman Li <roman.li@amd.com>
Reviewed-by: Charlene Liu <Charlene.Liu@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../drm/amd/display/dc/dce110/dce110_compressor.c  | 67 +++++++++++++++++-----
 1 file changed, 54 insertions(+), 13 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c
index 775d3bf..9150d26 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c
@@ -102,6 +102,43 @@ static uint32_t align_to_chunks_number_per_line(uint32_t pixels)
 	return 256 * ((pixels + 255) / 256);
 }
 
+static void reset_lb_on_vblank(struct dc_context *ctx)
+{
+	uint32_t value, frame_count;
+	uint32_t retry = 0;
+	uint32_t status_pos =
+			dm_read_reg(ctx, mmCRTC_STATUS_POSITION);
+
+
+	/* Only if CRTC is enabled and counter is moving we wait for one frame. */
+	if (status_pos != dm_read_reg(ctx, mmCRTC_STATUS_POSITION)) {
+		/* Resetting LB on VBlank */
+		value = dm_read_reg(ctx, mmLB_SYNC_RESET_SEL);
+		set_reg_field_value(value, 3, LB_SYNC_RESET_SEL, LB_SYNC_RESET_SEL);
+		set_reg_field_value(value, 1, LB_SYNC_RESET_SEL, LB_SYNC_RESET_SEL2);
+		dm_write_reg(ctx, mmLB_SYNC_RESET_SEL, value);
+
+		frame_count = dm_read_reg(ctx, mmCRTC_STATUS_FRAME_COUNT);
+
+
+		for (retry = 100; retry > 0; retry--) {
+			if (frame_count != dm_read_reg(ctx, mmCRTC_STATUS_FRAME_COUNT))
+				break;
+			msleep(1);
+		}
+		if (!retry)
+			dm_error("Frame count did not increase for 100ms.\n");
+
+		/* Resetting LB on VBlank */
+		value = dm_read_reg(ctx, mmLB_SYNC_RESET_SEL);
+		set_reg_field_value(value, 2, LB_SYNC_RESET_SEL, LB_SYNC_RESET_SEL);
+		set_reg_field_value(value, 0, LB_SYNC_RESET_SEL, LB_SYNC_RESET_SEL2);
+		dm_write_reg(ctx, mmLB_SYNC_RESET_SEL, value);
+
+	}
+
+}
+
 static void wait_for_fbc_state_changed(
 	struct dce110_compressor *cp110,
 	bool enabled)
@@ -232,19 +269,23 @@ void dce110_compressor_disable_fbc(struct compressor *compressor)
 {
 	struct dce110_compressor *cp110 = TO_DCE110_COMPRESSOR(compressor);
 
-	if (compressor->options.bits.FBC_SUPPORT &&
-		dce110_compressor_is_fbc_enabled_in_hw(compressor, NULL)) {
-		uint32_t reg_data;
-		/* Turn off compression */
-		reg_data = dm_read_reg(compressor->ctx, mmFBC_CNTL);
-		set_reg_field_value(reg_data, 0, FBC_CNTL, FBC_GRPH_COMP_EN);
-		dm_write_reg(compressor->ctx, mmFBC_CNTL, reg_data);
-
-		/* Reset enum controller_id to undefined */
-		compressor->attached_inst = 0;
-		compressor->is_enabled = false;
-
-		wait_for_fbc_state_changed(cp110, false);
+	if (compressor->options.bits.FBC_SUPPORT) {
+		if (dce110_compressor_is_fbc_enabled_in_hw(compressor, NULL)) {
+			uint32_t reg_data;
+			/* Turn off compression */
+			reg_data = dm_read_reg(compressor->ctx, mmFBC_CNTL);
+			set_reg_field_value(reg_data, 0, FBC_CNTL, FBC_GRPH_COMP_EN);
+			dm_write_reg(compressor->ctx, mmFBC_CNTL, reg_data);
+
+			/* Reset enum controller_id to undefined */
+			compressor->attached_inst = 0;
+			compressor->is_enabled = false;
+
+			wait_for_fbc_state_changed(cp110, false);
+		}
+
+		/* Sync line buffer  - dce100/110 only*/
+		reset_lb_on_vblank(compressor->ctx);
 	}
 }
 
-- 
cgit v1.1


From 144de8944805aef45964a904fdfd537486b1ce82 Mon Sep 17 00:00:00 2001
From: Yongqiang Sun <yongqiang.sun@amd.com>
Date: Thu, 29 Mar 2018 13:11:10 -0400
Subject: drm/amd/display: dal 3.1.41

Signed-off-by: Yongqiang Sun <yongqiang.sun@amd.com>
Reviewed-by: Yongqiang Sun <yongqiang.sun@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dc.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index 7d1a3c5..2334914 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -38,7 +38,7 @@
 #include "inc/compressor.h"
 #include "dml/display_mode_lib.h"
 
-#define DC_VER "3.1.40"
+#define DC_VER "3.1.41"
 
 #define MAX_SURFACES 3
 #define MAX_STREAMS 6
-- 
cgit v1.1


From 6bd6ae639683c0b41f46990d5c64ff9fbfa019dc Mon Sep 17 00:00:00 2001
From: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Date: Fri, 6 Apr 2018 10:23:05 -0700
Subject: Input: leds - fix out of bound access

UI_SET_LEDBIT ioctl() causes the following KASAN splat when used with
led > LED_CHARGING:

[ 1274.663418] BUG: KASAN: slab-out-of-bounds in input_leds_connect+0x611/0x730 [input_leds]
[ 1274.663426] Write of size 8 at addr ffff88003377b2c0 by task ckb-next-daemon/5128

This happens because we were writing to the led structure before making
sure that it exists.

Reported-by: Tasos Sahanidis <tasos@tasossah.com>
Tested-by: Tasos Sahanidis <tasos@tasossah.com>
Cc: stable@vger.kernel.org
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
---
 drivers/input/input-leds.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/input/input-leds.c b/drivers/input/input-leds.c
index 766bf26..5f04b2d 100644
--- a/drivers/input/input-leds.c
+++ b/drivers/input/input-leds.c
@@ -88,6 +88,7 @@ static int input_leds_connect(struct input_handler *handler,
 			      const struct input_device_id *id)
 {
 	struct input_leds *leds;
+	struct input_led *led;
 	unsigned int num_leds;
 	unsigned int led_code;
 	int led_no;
@@ -119,14 +120,13 @@ static int input_leds_connect(struct input_handler *handler,
 
 	led_no = 0;
 	for_each_set_bit(led_code, dev->ledbit, LED_CNT) {
-		struct input_led *led = &leds->leds[led_no];
+		if (!input_led_info[led_code].name)
+			continue;
 
+		led = &leds->leds[led_no];
 		led->handle = &leds->handle;
 		led->code = led_code;
 
-		if (!input_led_info[led_code].name)
-			continue;
-
 		led->cdev.name = kasprintf(GFP_KERNEL, "%s::%s",
 					   dev_name(&dev->dev),
 					   input_led_info[led_code].name);
-- 
cgit v1.1


From 2f24636b4b12e4f1eca23e6e5dd86ba335fc8066 Mon Sep 17 00:00:00 2001
From: Changbin Du <changbin.du@intel.com>
Date: Wed, 11 Apr 2018 16:39:22 +0800
Subject: drm/i915/gvt: Fix the validation on size field of dp aux header

The assertion for len is wrong, so fix it. And for where to validate
user input, we should not warn by call trace.

[ 290.584739] WARNING: CPU: 0 PID: 1471 at drivers/gpu/drm/i915/gvt/handlers.c:969 dp_aux_ch_ctl_mmio_write+0x394/0x430 [i915]
[ 290.586113] task: ffff880111fe8000 task.stack: ffffc90044a9c000
[ 290.586192] RIP: e030:dp_aux_ch_ctl_mmio_write+0x394/0x430 [i915]
[ 290.586258] RSP: e02b:ffffc90044a9fd88 EFLAGS: 00010282
[ 290.586315] RAX: 0000000000000017 RBX: 0000000000000003 RCX: ffffffff82461148
[ 290.586391] RDX: 0000000000000000 RSI: 0000000000000001 RDI: 0000000000000201
[ 290.586468] RBP: ffffc90043ed1000 R08: 0000000000000248 R09: 00000000000003d8
[ 290.586544] R10: ffffc90044bdd314 R11: 0000000000000011 R12: 0000000000064310
[ 290.586621] R13: 00000000fe4003ff R14: ffffc900432d1008 R15: ffff88010fa7cb40
[ 290.586701] FS: 0000000000000000(0000) GS:ffff880123200000(0000) knlGS:0000000000000000
[ 290.586787] CS: e033 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 290.586849] CR2: 00007f67ea44e000 CR3: 0000000116078000 CR4: 0000000000042660
[ 290.586926] Call Trace:
[ 290.586958] ? __switch_to_asm+0x40/0x70
[ 290.587017] intel_vgpu_mmio_reg_rw+0x1ec/0x3c0 [i915]
[ 290.587087] intel_vgpu_emulate_mmio_write+0xa8/0x2c0 [i915]
[ 290.587151] xengt_emulation_thread+0x501/0x7a0 [xengt]
[ 290.587208] ? __schedule+0x3c6/0x890
[ 290.587250] ? wait_woken+0x80/0x80
[ 290.587290] kthread+0xfc/0x130
[ 290.587326] ? xengt_gpa_to_va+0x1f0/0x1f0 [xengt]
[ 290.587378] ? kthread_create_on_node+0x70/0x70
[ 290.587429] ? do_group_exit+0x3a/0xa0
[ 290.587471] ret_from_fork+0x35/0x40

Fixes: 04d348a ("drm/i915/gvt: vGPU display virtualization")
Signed-off-by: Changbin Du <changbin.du@intel.com>
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
---
 drivers/gpu/drm/i915/gvt/display.h  |  2 +-
 drivers/gpu/drm/i915/gvt/handlers.c | 13 +++++++++----
 2 files changed, 10 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/i915/gvt/display.h b/drivers/gpu/drm/i915/gvt/display.h
index b46b868..ea7c1c5 100644
--- a/drivers/gpu/drm/i915/gvt/display.h
+++ b/drivers/gpu/drm/i915/gvt/display.h
@@ -67,7 +67,7 @@
 #define AUX_NATIVE_REPLY_NAK    (0x1 << 4)
 #define AUX_NATIVE_REPLY_DEFER  (0x2 << 4)
 
-#define AUX_BURST_SIZE          16
+#define AUX_BURST_SIZE          20
 
 /* DPCD addresses */
 #define DPCD_REV			0x000
diff --git a/drivers/gpu/drm/i915/gvt/handlers.c b/drivers/gpu/drm/i915/gvt/handlers.c
index 9be639a..dea751e 100644
--- a/drivers/gpu/drm/i915/gvt/handlers.c
+++ b/drivers/gpu/drm/i915/gvt/handlers.c
@@ -898,11 +898,14 @@ static int dp_aux_ch_ctl_mmio_write(struct intel_vgpu *vgpu,
 		}
 
 		/*
-		 * Write request format: (command + address) occupies
-		 * 3 bytes, followed by (len + 1) bytes of data.
+		 * Write request format: Headr (command + address + size) occupies
+		 * 4 bytes, followed by (len + 1) bytes of data. See details at
+		 * intel_dp_aux_transfer().
 		 */
-		if (WARN_ON((len + 4) > AUX_BURST_SIZE))
+		if ((len + 1 + 4) > AUX_BURST_SIZE) {
+			gvt_vgpu_err("dp_aux_header: len %d is too large\n", len);
 			return -EINVAL;
+		}
 
 		/* unpack data from vreg to buf */
 		for (t = 0; t < 4; t++) {
@@ -966,8 +969,10 @@ static int dp_aux_ch_ctl_mmio_write(struct intel_vgpu *vgpu,
 		/*
 		 * Read reply format: ACK (1 byte) plus (len + 1) bytes of data.
 		 */
-		if (WARN_ON((len + 2) > AUX_BURST_SIZE))
+		if ((len + 2) > AUX_BURST_SIZE) {
+			gvt_vgpu_err("dp_aux_header: len %d is too large\n", len);
 			return -EINVAL;
+		}
 
 		/* read from virtual DPCD to vreg */
 		/* first 4 bytes: [ACK][addr][addr+1][addr+2] */
-- 
cgit v1.1


From 5bbed0b38dd02dfb58d195487ef5b23e74b8b6c1 Mon Sep 17 00:00:00 2001
From: "Piorkowski, Piotr" <piotr.piorkowski@intel.com>
Date: Fri, 13 Apr 2018 10:52:45 +0200
Subject: drm/i915/guc: Remove GUC_CTL_DEVICE_INFO parameter
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

It looks that GuC does not actively use GUC_CTL_DEVICE_INFO parameter
where we are passing GT type and Core family values.
Let's stop/remove setup of this parameter and remove related
definitions.

v2: (this time without squashed HAX)
  - New title and description
  - Remove also GUC_CORE_FAMILY_* definitions (Michel)
v3:
  - The removed define GUC_CTL_DEVICE_INFO has been restored (Michel)
  - Updated description (Sagar)
v4: rebase

Signed-off-by: Piotr Piórkowski <piotr.piorkowski@intel.com>
Cc: Sagar Arun Kamble <sagar.a.kamble@intel.com>
Cc: Michał Winiarski <michal.winiarski@intel.com>
Cc: John A Spotswood <john.a.spotswood@intel.com>
Cc: Michal Wajdeczko <michal.wajdeczko@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Michel Thierry <michel.thierry@intel.com>
Acked-by: Michel Thierry <michel.thierry@intel.com>
Reviewed-by: Sagar Arun Kamble <sagar.a.kamble@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20180413085245.57206-1-piotr.piorkowski@intel.com
---
 drivers/gpu/drm/i915/intel_guc.c      | 24 ------------------------
 drivers/gpu/drm/i915/intel_guc_fwif.h |  5 -----
 2 files changed, 29 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_guc.c b/drivers/gpu/drm/i915/intel_guc.c
index a00a59a..116f4cc 100644
--- a/drivers/gpu/drm/i915/intel_guc.c
+++ b/drivers/gpu/drm/i915/intel_guc.c
@@ -203,26 +203,6 @@ void intel_guc_fini(struct intel_guc *guc)
 	guc_shared_data_destroy(guc);
 }
 
-static u32 get_gt_type(struct drm_i915_private *dev_priv)
-{
-	/* XXX: GT type based on PCI device ID? field seems unused by fw */
-	return 0;
-}
-
-static u32 get_core_family(struct drm_i915_private *dev_priv)
-{
-	u32 gen = INTEL_GEN(dev_priv);
-
-	switch (gen) {
-	case 9:
-		return GUC_CORE_FAMILY_GEN9;
-
-	default:
-		MISSING_CASE(gen);
-		return GUC_CORE_FAMILY_UNKNOWN;
-	}
-}
-
 static u32 get_log_control_flags(void)
 {
 	u32 level = i915_modparams.guc_log_level;
@@ -255,10 +235,6 @@ void intel_guc_init_params(struct intel_guc *guc)
 
 	memset(params, 0, sizeof(params));
 
-	params[GUC_CTL_DEVICE_INFO] |=
-		(get_gt_type(dev_priv) << GUC_CTL_GT_TYPE_SHIFT) |
-		(get_core_family(dev_priv) << GUC_CTL_CORE_FAMILY_SHIFT);
-
 	/*
 	 * GuC ARAT increment is 10 ns. GuC default scheduler quantum is one
 	 * second. This ARAR is calculated by:
diff --git a/drivers/gpu/drm/i915/intel_guc_fwif.h b/drivers/gpu/drm/i915/intel_guc_fwif.h
index d73673f..0867ba7 100644
--- a/drivers/gpu/drm/i915/intel_guc_fwif.h
+++ b/drivers/gpu/drm/i915/intel_guc_fwif.h
@@ -23,9 +23,6 @@
 #ifndef _INTEL_GUC_FWIF_H
 #define _INTEL_GUC_FWIF_H
 
-#define GUC_CORE_FAMILY_GEN9		12
-#define GUC_CORE_FAMILY_UNKNOWN		0x7fffffff
-
 #define GUC_CLIENT_PRIORITY_KMD_HIGH	0
 #define GUC_CLIENT_PRIORITY_HIGH	1
 #define GUC_CLIENT_PRIORITY_KMD_NORMAL	2
@@ -82,8 +79,6 @@
 #define GUC_CTL_ARAT_LOW		2
 
 #define GUC_CTL_DEVICE_INFO		3
-#define   GUC_CTL_GT_TYPE_SHIFT		0
-#define   GUC_CTL_CORE_FAMILY_SHIFT	7
 
 #define GUC_CTL_LOG_PARAMS		4
 #define   GUC_LOG_VALID			(1 << 0)
-- 
cgit v1.1


From 8feaccf71dd61f2201493068055e0d1d699014df Mon Sep 17 00:00:00 2001
From: Dan Carpenter <dan.carpenter@oracle.com>
Date: Tue, 24 Apr 2018 16:35:49 +0300
Subject: drm/amdkfd: Integer overflows in ioctl

args->n_devices is a u32 that comes from the user.  The multiplication
could overflow on 32 bit systems possibly leading to privilege
escalation.

Fixes: 5ec7e02854b3 ("drm/amdkfd: Add ioctls for GPUVM memory management")
Signed-off-by: Dan Carpenter dan.carpenter@oracle.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/kfd_chardev.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
index 5694fbe..ce15baf 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
@@ -1303,8 +1303,8 @@ static int kfd_ioctl_map_memory_to_gpu(struct file *filep,
 		return -EINVAL;
 	}
 
-	devices_arr = kmalloc(args->n_devices * sizeof(*devices_arr),
-			      GFP_KERNEL);
+	devices_arr = kmalloc_array(args->n_devices, sizeof(*devices_arr),
+				    GFP_KERNEL);
 	if (!devices_arr)
 		return -ENOMEM;
 
@@ -1412,8 +1412,8 @@ static int kfd_ioctl_unmap_memory_from_gpu(struct file *filep,
 		return -EINVAL;
 	}
 
-	devices_arr = kmalloc(args->n_devices * sizeof(*devices_arr),
-			      GFP_KERNEL);
+	devices_arr = kmalloc_array(args->n_devices, sizeof(*devices_arr),
+				    GFP_KERNEL);
 	if (!devices_arr)
 		return -ENOMEM;
 
-- 
cgit v1.1


From 24f48a42038f5baaae49b181b64782ecfb703a9c Mon Sep 17 00:00:00 2001
From: Oak Zeng <Oak.Zeng@amd.com>
Date: Tue, 1 May 2018 17:56:01 -0400
Subject: drm/amdkfd: Dump HQD of HIQ

Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
index 9af94b1..668ad07 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
@@ -1713,6 +1713,18 @@ int dqm_debugfs_hqds(struct seq_file *m, void *data)
 	int pipe, queue;
 	int r = 0;
 
+	r = dqm->dev->kfd2kgd->hqd_dump(dqm->dev->kgd,
+		KFD_CIK_HIQ_PIPE, KFD_CIK_HIQ_QUEUE, &dump, &n_regs);
+	if (!r) {
+		seq_printf(m, "  HIQ on MEC %d Pipe %d Queue %d\n",
+				KFD_CIK_HIQ_PIPE/get_pipes_per_mec(dqm)+1,
+				KFD_CIK_HIQ_PIPE%get_pipes_per_mec(dqm),
+				KFD_CIK_HIQ_QUEUE);
+		seq_reg_dump(m, dump, n_regs);
+
+		kfree(dump);
+	}
+
 	for (pipe = 0; pipe < get_pipes_per_mec(dqm); pipe++) {
 		int pipe_offset = pipe * get_queues_per_pipe(dqm);
 
-- 
cgit v1.1


From 87e6d4e0777daf774ed9aa59ed25b6ebaaad7052 Mon Sep 17 00:00:00 2001
From: Jay Cornwall <Jay.Cornwall@amd.com>
Date: Tue, 1 May 2018 17:56:02 -0400
Subject: drm/amdkfd: Reduce priority of context-saving waves before spin-wait

Synchronization between context-saving wavefronts is achieved by
sending a SAVEWAVE message to the SPI and then spin-waiting for a
response. These spin-waiting wavefronts may inhibit the progress
of other wavefronts in the context save handler, leading to the
synchronization condition never being achieved.

Before spin-waiting reduce the priority of each wavefront to
guarantee foward progress in the others.

Signed-off-by: Jay Cornwall <Jay.Cornwall@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm | 10 ++++++++--
 drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm |  8 +++++++-
 2 files changed, 15 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm
index 997a383d..34eabcd 100644
--- a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm
+++ b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm
@@ -98,6 +98,7 @@ var SWIZZLE_EN                      =   0                   //whether we use swi
 /**************************************************************************/
 var SQ_WAVE_STATUS_INST_ATC_SHIFT  = 23
 var SQ_WAVE_STATUS_INST_ATC_MASK   = 0x00800000
+var SQ_WAVE_STATUS_SPI_PRIO_SHIFT  = 1
 var SQ_WAVE_STATUS_SPI_PRIO_MASK   = 0x00000006
 
 var SQ_WAVE_LDS_ALLOC_LDS_SIZE_SHIFT    = 12
@@ -319,6 +320,10 @@ end
         s_sendmsg   sendmsg(MSG_SAVEWAVE)  //send SPI a message and wait for SPI's write to EXEC
     end
 
+    // Set SPI_PRIO=2 to avoid starving instruction fetch in the waves we're waiting for.
+    s_or_b32 s_save_tmp, s_save_status, (2 << SQ_WAVE_STATUS_SPI_PRIO_SHIFT)
+    s_setreg_b32 hwreg(HW_REG_STATUS), s_save_tmp
+
   L_SLEEP:
     s_sleep 0x2                // sleep 1 (64clk) is not enough for 8 waves per SIMD, which will cause SQ hang, since the 7,8th wave could not get arbit to exec inst, while other waves are stuck into the sleep-loop and waiting for wrexec!=0
 
@@ -1132,7 +1137,7 @@ end
 #endif
 
 static const uint32_t cwsr_trap_gfx8_hex[] = {
-	0xbf820001, 0xbf820123,
+	0xbf820001, 0xbf820125,
 	0xb8f4f802, 0x89748674,
 	0xb8f5f803, 0x8675ff75,
 	0x00000400, 0xbf850011,
@@ -1158,7 +1163,8 @@ static const uint32_t cwsr_trap_gfx8_hex[] = {
 	0x867aff7a, 0x00007fff,
 	0xb97af807, 0xbef2007e,
 	0xbef3007f, 0xbefe0180,
-	0xbf900004, 0xbf8e0002,
+	0xbf900004, 0x877a8474,
+	0xb97af802, 0xbf8e0002,
 	0xbf88fffe, 0xbef8007e,
 	0x8679ff7f, 0x0000ffff,
 	0x8779ff79, 0x00040000,
diff --git a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm
index 033580c..cac8d49 100644
--- a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm
+++ b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm
@@ -97,6 +97,7 @@ var ACK_SQC_STORE		    =	1		    //workaround for suspected SQC store bug causing
 /**************************************************************************/
 var SQ_WAVE_STATUS_INST_ATC_SHIFT  = 23
 var SQ_WAVE_STATUS_INST_ATC_MASK   = 0x00800000
+var SQ_WAVE_STATUS_SPI_PRIO_SHIFT  = 1
 var SQ_WAVE_STATUS_SPI_PRIO_MASK   = 0x00000006
 var SQ_WAVE_STATUS_HALT_MASK       = 0x2000
 
@@ -362,6 +363,10 @@ end
 	s_sendmsg   sendmsg(MSG_SAVEWAVE)  //send SPI a message and wait for SPI's write to EXEC
     end
 
+    // Set SPI_PRIO=2 to avoid starving instruction fetch in the waves we're waiting for.
+    s_or_b32 s_save_tmp, s_save_status, (2 << SQ_WAVE_STATUS_SPI_PRIO_SHIFT)
+    s_setreg_b32 hwreg(HW_REG_STATUS), s_save_tmp
+
   L_SLEEP:
     s_sleep 0x2		       // sleep 1 (64clk) is not enough for 8 waves per SIMD, which will cause SQ hang, since the 7,8th wave could not get arbit to exec inst, while other waves are stuck into the sleep-loop and waiting for wrexec!=0
 
@@ -1210,7 +1215,7 @@ end
 #endif
 
 static const uint32_t cwsr_trap_gfx9_hex[] = {
-	0xbf820001, 0xbf820158,
+	0xbf820001, 0xbf82015a,
 	0xb8f8f802, 0x89788678,
 	0xb8f1f803, 0x866eff71,
 	0x00000400, 0xbf850034,
@@ -1249,6 +1254,7 @@ static const uint32_t cwsr_trap_gfx9_hex[] = {
 	0x00007fff, 0xb970f807,
 	0xbeee007e, 0xbeef007f,
 	0xbefe0180, 0xbf900004,
+	0x87708478, 0xb970f802,
 	0xbf8e0002, 0xbf88fffe,
 	0xb8f02a05, 0x80708170,
 	0x8e708a70, 0xb8f11605,
-- 
cgit v1.1


From 2774c63ef3dbb6052dd1d224b38a9decf89be61c Mon Sep 17 00:00:00 2001
From: Jay Cornwall <Jay.Cornwall@amd.com>
Date: Tue, 1 May 2018 17:56:03 -0400
Subject: drm/amdkfd: Use volatile MTYPE in default/alternate apertures

MTYPE_NC_NV (0) marks scalar/vector L1 cache lines as non-volatile.
Cache lines loaded through these apertures are intended to be
invalidated before (and sometimes during) a dispatch. The non-volatile
qualifier prevents these cache lines from being distinguished from
those loaded through the private aperture.

Use MTYPE_NC (1) instead on both Gfx7 and Gfx8. This allows the
compiler to use the BUFFER_WBINVL1_VOL instruction and is a precursor
to automatic per-dispatch scalar/vector L1 volatile invalidation.

Signed-off-by: Jay Cornwall <Jay.Cornwall@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/cik_regs.h | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/cik_regs.h b/drivers/gpu/drm/amd/amdkfd/cik_regs.h
index 48769d1..37ce6dd 100644
--- a/drivers/gpu/drm/amd/amdkfd/cik_regs.h
+++ b/drivers/gpu/drm/amd/amdkfd/cik_regs.h
@@ -33,7 +33,8 @@
 #define	APE1_MTYPE(x)					((x) << 7)
 
 /* valid for both DEFAULT_MTYPE and APE1_MTYPE */
-#define	MTYPE_CACHED					0
+#define	MTYPE_CACHED_NV					0
+#define	MTYPE_CACHED					1
 #define	MTYPE_NONCACHED					3
 
 #define	DEFAULT_CP_HQD_PERSISTENT_STATE			(0x33U << 8)
-- 
cgit v1.1


From fa7e65147e5dcafdf8d6c3787e5b22ec5f6bcbdc Mon Sep 17 00:00:00 2001
From: Philip Yang <Philip.Yang@amd.com>
Date: Tue, 1 May 2018 17:56:04 -0400
Subject: drm/amdkfd: use %px to print user space address instead of %p

Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/kfd_chardev.c | 2 +-
 drivers/gpu/drm/amd/amdkfd/kfd_queue.c   | 8 ++++----
 2 files changed, 5 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
index ce15baf..beaa613 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
@@ -233,7 +233,7 @@ static int set_queue_properties_from_user(struct queue_properties *q_properties,
 	pr_debug("Queue Size: 0x%llX, %u\n",
 			q_properties->queue_size, args->ring_size);
 
-	pr_debug("Queue r/w Pointers: %p, %p\n",
+	pr_debug("Queue r/w Pointers: %px, %px\n",
 			q_properties->read_ptr,
 			q_properties->write_ptr);
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_queue.c b/drivers/gpu/drm/amd/amdkfd/kfd_queue.c
index a5315d4..6dcd621 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_queue.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_queue.c
@@ -36,8 +36,8 @@ void print_queue_properties(struct queue_properties *q)
 	pr_debug("Queue Address: 0x%llX\n", q->queue_address);
 	pr_debug("Queue Id: %u\n", q->queue_id);
 	pr_debug("Queue Process Vmid: %u\n", q->vmid);
-	pr_debug("Queue Read Pointer: 0x%p\n", q->read_ptr);
-	pr_debug("Queue Write Pointer: 0x%p\n", q->write_ptr);
+	pr_debug("Queue Read Pointer: 0x%px\n", q->read_ptr);
+	pr_debug("Queue Write Pointer: 0x%px\n", q->write_ptr);
 	pr_debug("Queue Doorbell Pointer: 0x%p\n", q->doorbell_ptr);
 	pr_debug("Queue Doorbell Offset: %u\n", q->doorbell_off);
 }
@@ -53,8 +53,8 @@ void print_queue(struct queue *q)
 	pr_debug("Queue Address: 0x%llX\n", q->properties.queue_address);
 	pr_debug("Queue Id: %u\n", q->properties.queue_id);
 	pr_debug("Queue Process Vmid: %u\n", q->properties.vmid);
-	pr_debug("Queue Read Pointer: 0x%p\n", q->properties.read_ptr);
-	pr_debug("Queue Write Pointer: 0x%p\n", q->properties.write_ptr);
+	pr_debug("Queue Read Pointer: 0x%px\n", q->properties.read_ptr);
+	pr_debug("Queue Write Pointer: 0x%px\n", q->properties.write_ptr);
 	pr_debug("Queue Doorbell Pointer: 0x%p\n", q->properties.doorbell_ptr);
 	pr_debug("Queue Doorbell Offset: %u\n", q->properties.doorbell_off);
 	pr_debug("Queue MQD Address: 0x%p\n", q->mqd);
-- 
cgit v1.1


From a2e94158b83185c9dac430cb53bff26737a786ef Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Tue, 1 May 2018 17:56:05 -0400
Subject: drm/amdkfd: Remove redundant include of amd-iommu.h

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/kfd_device.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
index dd6c753..c1d9e27 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
@@ -20,9 +20,6 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  */
 
-#if defined(CONFIG_AMD_IOMMU_V2_MODULE) || defined(CONFIG_AMD_IOMMU_V2)
-#include <linux/amd-iommu.h>
-#endif
 #include <linux/bsearch.h>
 #include <linux/pci.h>
 #include <linux/slab.h>
-- 
cgit v1.1


From 0db54b24ad676c3f2d0cf5291c9d170e3e15f213 Mon Sep 17 00:00:00 2001
From: Yong Zhao <yong.zhao@amd.com>
Date: Tue, 1 May 2018 17:56:06 -0400
Subject: drm/amdkfd: Separate trap handler assembly code and its hex values

Since the assembly code is inside "#if 0", it is ineffective. Despite that,
during debugging, we need to change the assembly code, extract it into
a separate file and compile the new file into hex values using sp3.
That process also requires us to remove "#if 0" and modify lines starting
with "#", so that sp3 can successfully compile the new file.

With this change, all the above chore is no longer needed, and
cwsr_trap_handler_gfx*.asm can be directly used by sp3 to generate its
hex values.

Signed-off-by: Yong Zhao <yong.zhao@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h     | 560 +++++++++++++++++++++
 .../gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm  | 267 +---------
 .../gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm  | 300 +----------
 drivers/gpu/drm/amd/amdkfd/kfd_device.c            |   3 +-
 4 files changed, 575 insertions(+), 555 deletions(-)
 create mode 100644 drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h

diff --git a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h
new file mode 100644
index 0000000..a546a21
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h
@@ -0,0 +1,560 @@
+/*
+ * Copyright 2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+static const uint32_t cwsr_trap_gfx8_hex[] = {
+	0xbf820001, 0xbf820125,
+	0xb8f4f802, 0x89748674,
+	0xb8f5f803, 0x8675ff75,
+	0x00000400, 0xbf850011,
+	0xc00a1e37, 0x00000000,
+	0xbf8c007f, 0x87777978,
+	0xbf840002, 0xb974f802,
+	0xbe801d78, 0xb8f5f803,
+	0x8675ff75, 0x000001ff,
+	0xbf850002, 0x80708470,
+	0x82718071, 0x8671ff71,
+	0x0000ffff, 0xb974f802,
+	0xbe801f70, 0xb8f5f803,
+	0x8675ff75, 0x00000100,
+	0xbf840006, 0xbefa0080,
+	0xb97a0203, 0x8671ff71,
+	0x0000ffff, 0x80f08870,
+	0x82f18071, 0xbefa0080,
+	0xb97a0283, 0xbef60068,
+	0xbef70069, 0xb8fa1c07,
+	0x8e7a9c7a, 0x87717a71,
+	0xb8fa03c7, 0x8e7a9b7a,
+	0x87717a71, 0xb8faf807,
+	0x867aff7a, 0x00007fff,
+	0xb97af807, 0xbef2007e,
+	0xbef3007f, 0xbefe0180,
+	0xbf900004, 0x877a8474,
+	0xb97af802, 0xbf8e0002,
+	0xbf88fffe, 0xbef8007e,
+	0x8679ff7f, 0x0000ffff,
+	0x8779ff79, 0x00040000,
+	0xbefa0080, 0xbefb00ff,
+	0x00807fac, 0x867aff7f,
+	0x08000000, 0x8f7a837a,
+	0x877b7a7b, 0x867aff7f,
+	0x70000000, 0x8f7a817a,
+	0x877b7a7b, 0xbeef007c,
+	0xbeee0080, 0xb8ee2a05,
+	0x806e816e, 0x8e6e8a6e,
+	0xb8fa1605, 0x807a817a,
+	0x8e7a867a, 0x806e7a6e,
+	0xbefa0084, 0xbefa00ff,
+	0x01000000, 0xbefe007c,
+	0xbefc006e, 0xc0611bfc,
+	0x0000007c, 0x806e846e,
+	0xbefc007e, 0xbefe007c,
+	0xbefc006e, 0xc0611c3c,
+	0x0000007c, 0x806e846e,
+	0xbefc007e, 0xbefe007c,
+	0xbefc006e, 0xc0611c7c,
+	0x0000007c, 0x806e846e,
+	0xbefc007e, 0xbefe007c,
+	0xbefc006e, 0xc0611cbc,
+	0x0000007c, 0x806e846e,
+	0xbefc007e, 0xbefe007c,
+	0xbefc006e, 0xc0611cfc,
+	0x0000007c, 0x806e846e,
+	0xbefc007e, 0xbefe007c,
+	0xbefc006e, 0xc0611d3c,
+	0x0000007c, 0x806e846e,
+	0xbefc007e, 0xb8f5f803,
+	0xbefe007c, 0xbefc006e,
+	0xc0611d7c, 0x0000007c,
+	0x806e846e, 0xbefc007e,
+	0xbefe007c, 0xbefc006e,
+	0xc0611dbc, 0x0000007c,
+	0x806e846e, 0xbefc007e,
+	0xbefe007c, 0xbefc006e,
+	0xc0611dfc, 0x0000007c,
+	0x806e846e, 0xbefc007e,
+	0xb8eff801, 0xbefe007c,
+	0xbefc006e, 0xc0611bfc,
+	0x0000007c, 0x806e846e,
+	0xbefc007e, 0xbefe007c,
+	0xbefc006e, 0xc0611b3c,
+	0x0000007c, 0x806e846e,
+	0xbefc007e, 0xbefe007c,
+	0xbefc006e, 0xc0611b7c,
+	0x0000007c, 0x806e846e,
+	0xbefc007e, 0x867aff7f,
+	0x04000000, 0xbef30080,
+	0x8773737a, 0xb8ee2a05,
+	0x806e816e, 0x8e6e8a6e,
+	0xb8f51605, 0x80758175,
+	0x8e758475, 0x8e7a8275,
+	0xbefa00ff, 0x01000000,
+	0xbef60178, 0x80786e78,
+	0x82798079, 0xbefc0080,
+	0xbe802b00, 0xbe822b02,
+	0xbe842b04, 0xbe862b06,
+	0xbe882b08, 0xbe8a2b0a,
+	0xbe8c2b0c, 0xbe8e2b0e,
+	0xc06b003c, 0x00000000,
+	0xc06b013c, 0x00000010,
+	0xc06b023c, 0x00000020,
+	0xc06b033c, 0x00000030,
+	0x8078c078, 0x82798079,
+	0x807c907c, 0xbf0a757c,
+	0xbf85ffeb, 0xbef80176,
+	0xbeee0080, 0xbefe00c1,
+	0xbeff00c1, 0xbefa00ff,
+	0x01000000, 0xe0724000,
+	0x6e1e0000, 0xe0724100,
+	0x6e1e0100, 0xe0724200,
+	0x6e1e0200, 0xe0724300,
+	0x6e1e0300, 0xbefe00c1,
+	0xbeff00c1, 0xb8f54306,
+	0x8675c175, 0xbf84002c,
+	0xbf8a0000, 0x867aff73,
+	0x04000000, 0xbf840028,
+	0x8e758675, 0x8e758275,
+	0xbefa0075, 0xb8ee2a05,
+	0x806e816e, 0x8e6e8a6e,
+	0xb8fa1605, 0x807a817a,
+	0x8e7a867a, 0x806e7a6e,
+	0x806eff6e, 0x00000080,
+	0xbefa00ff, 0x01000000,
+	0xbefc0080, 0xd28c0002,
+	0x000100c1, 0xd28d0003,
+	0x000204c1, 0xd1060002,
+	0x00011103, 0x7e0602ff,
+	0x00000200, 0xbefc00ff,
+	0x00010000, 0xbe80007b,
+	0x867bff7b, 0xff7fffff,
+	0x877bff7b, 0x00058000,
+	0xd8ec0000, 0x00000002,
+	0xbf8c007f, 0xe0765000,
+	0x6e1e0002, 0x32040702,
+	0xd0c9006a, 0x0000eb02,
+	0xbf87fff7, 0xbefb0000,
+	0xbeee00ff, 0x00000400,
+	0xbefe00c1, 0xbeff00c1,
+	0xb8f52a05, 0x80758175,
+	0x8e758275, 0x8e7a8875,
+	0xbefa00ff, 0x01000000,
+	0xbefc0084, 0xbf0a757c,
+	0xbf840015, 0xbf11017c,
+	0x8075ff75, 0x00001000,
+	0x7e000300, 0x7e020301,
+	0x7e040302, 0x7e060303,
+	0xe0724000, 0x6e1e0000,
+	0xe0724100, 0x6e1e0100,
+	0xe0724200, 0x6e1e0200,
+	0xe0724300, 0x6e1e0300,
+	0x807c847c, 0x806eff6e,
+	0x00000400, 0xbf0a757c,
+	0xbf85ffef, 0xbf9c0000,
+	0xbf8200ca, 0xbef8007e,
+	0x8679ff7f, 0x0000ffff,
+	0x8779ff79, 0x00040000,
+	0xbefa0080, 0xbefb00ff,
+	0x00807fac, 0x8676ff7f,
+	0x08000000, 0x8f768376,
+	0x877b767b, 0x8676ff7f,
+	0x70000000, 0x8f768176,
+	0x877b767b, 0x8676ff7f,
+	0x04000000, 0xbf84001e,
+	0xbefe00c1, 0xbeff00c1,
+	0xb8f34306, 0x8673c173,
+	0xbf840019, 0x8e738673,
+	0x8e738273, 0xbefa0073,
+	0xb8f22a05, 0x80728172,
+	0x8e728a72, 0xb8f61605,
+	0x80768176, 0x8e768676,
+	0x80727672, 0x8072ff72,
+	0x00000080, 0xbefa00ff,
+	0x01000000, 0xbefc0080,
+	0xe0510000, 0x721e0000,
+	0xe0510100, 0x721e0000,
+	0x807cff7c, 0x00000200,
+	0x8072ff72, 0x00000200,
+	0xbf0a737c, 0xbf85fff6,
+	0xbef20080, 0xbefe00c1,
+	0xbeff00c1, 0xb8f32a05,
+	0x80738173, 0x8e738273,
+	0x8e7a8873, 0xbefa00ff,
+	0x01000000, 0xbef60072,
+	0x8072ff72, 0x00000400,
+	0xbefc0084, 0xbf11087c,
+	0x8073ff73, 0x00008000,
+	0xe0524000, 0x721e0000,
+	0xe0524100, 0x721e0100,
+	0xe0524200, 0x721e0200,
+	0xe0524300, 0x721e0300,
+	0xbf8c0f70, 0x7e000300,
+	0x7e020301, 0x7e040302,
+	0x7e060303, 0x807c847c,
+	0x8072ff72, 0x00000400,
+	0xbf0a737c, 0xbf85ffee,
+	0xbf9c0000, 0xe0524000,
+	0x761e0000, 0xe0524100,
+	0x761e0100, 0xe0524200,
+	0x761e0200, 0xe0524300,
+	0x761e0300, 0xb8f22a05,
+	0x80728172, 0x8e728a72,
+	0xb8f61605, 0x80768176,
+	0x8e768676, 0x80727672,
+	0x80f2c072, 0xb8f31605,
+	0x80738173, 0x8e738473,
+	0x8e7a8273, 0xbefa00ff,
+	0x01000000, 0xbefc0073,
+	0xc031003c, 0x00000072,
+	0x80f2c072, 0xbf8c007f,
+	0x80fc907c, 0xbe802d00,
+	0xbe822d02, 0xbe842d04,
+	0xbe862d06, 0xbe882d08,
+	0xbe8a2d0a, 0xbe8c2d0c,
+	0xbe8e2d0e, 0xbf06807c,
+	0xbf84fff1, 0xb8f22a05,
+	0x80728172, 0x8e728a72,
+	0xb8f61605, 0x80768176,
+	0x8e768676, 0x80727672,
+	0xbefa0084, 0xbefa00ff,
+	0x01000000, 0xc0211cfc,
+	0x00000072, 0x80728472,
+	0xc0211c3c, 0x00000072,
+	0x80728472, 0xc0211c7c,
+	0x00000072, 0x80728472,
+	0xc0211bbc, 0x00000072,
+	0x80728472, 0xc0211bfc,
+	0x00000072, 0x80728472,
+	0xc0211d3c, 0x00000072,
+	0x80728472, 0xc0211d7c,
+	0x00000072, 0x80728472,
+	0xc0211a3c, 0x00000072,
+	0x80728472, 0xc0211a7c,
+	0x00000072, 0x80728472,
+	0xc0211dfc, 0x00000072,
+	0x80728472, 0xc0211b3c,
+	0x00000072, 0x80728472,
+	0xc0211b7c, 0x00000072,
+	0x80728472, 0xbf8c007f,
+	0x8671ff71, 0x0000ffff,
+	0xbefc0073, 0xbefe006e,
+	0xbeff006f, 0x867375ff,
+	0x000003ff, 0xb9734803,
+	0x867375ff, 0xfffff800,
+	0x8f738b73, 0xb973a2c3,
+	0xb977f801, 0x8673ff71,
+	0xf0000000, 0x8f739c73,
+	0x8e739073, 0xbef60080,
+	0x87767376, 0x8673ff71,
+	0x08000000, 0x8f739b73,
+	0x8e738f73, 0x87767376,
+	0x8673ff74, 0x00800000,
+	0x8f739773, 0xb976f807,
+	0x86fe7e7e, 0x86ea6a6a,
+	0xb974f802, 0xbf8a0000,
+	0x95807370, 0xbf810000,
+};
+
+
+static const uint32_t cwsr_trap_gfx9_hex[] = {
+	0xbf820001, 0xbf82015a,
+	0xb8f8f802, 0x89788678,
+	0xb8f1f803, 0x866eff71,
+	0x00000400, 0xbf850034,
+	0x866eff71, 0x00000800,
+	0xbf850003, 0x866eff71,
+	0x00000100, 0xbf840008,
+	0x866eff78, 0x00002000,
+	0xbf840001, 0xbf810000,
+	0x8778ff78, 0x00002000,
+	0x80ec886c, 0x82ed806d,
+	0xb8eef807, 0x866fff6e,
+	0x001f8000, 0x8e6f8b6f,
+	0x8977ff77, 0xfc000000,
+	0x87776f77, 0x896eff6e,
+	0x001f8000, 0xb96ef807,
+	0xb8f0f812, 0xb8f1f813,
+	0x8ef08870, 0xc0071bb8,
+	0x00000000, 0xbf8cc07f,
+	0xc0071c38, 0x00000008,
+	0xbf8cc07f, 0x86ee6e6e,
+	0xbf840001, 0xbe801d6e,
+	0xb8f1f803, 0x8671ff71,
+	0x000001ff, 0xbf850002,
+	0x806c846c, 0x826d806d,
+	0x866dff6d, 0x0000ffff,
+	0x8f6e8b77, 0x866eff6e,
+	0x001f8000, 0xb96ef807,
+	0x86fe7e7e, 0x86ea6a6a,
+	0xb978f802, 0xbe801f6c,
+	0x866dff6d, 0x0000ffff,
+	0xbef00080, 0xb9700283,
+	0xb8f02407, 0x8e709c70,
+	0x876d706d, 0xb8f003c7,
+	0x8e709b70, 0x876d706d,
+	0xb8f0f807, 0x8670ff70,
+	0x00007fff, 0xb970f807,
+	0xbeee007e, 0xbeef007f,
+	0xbefe0180, 0xbf900004,
+	0x87708478, 0xb970f802,
+	0xbf8e0002, 0xbf88fffe,
+	0xb8f02a05, 0x80708170,
+	0x8e708a70, 0xb8f11605,
+	0x80718171, 0x8e718671,
+	0x80707170, 0x80707e70,
+	0x8271807f, 0x8671ff71,
+	0x0000ffff, 0xc0471cb8,
+	0x00000040, 0xbf8cc07f,
+	0xc04b1d38, 0x00000048,
+	0xbf8cc07f, 0xc0431e78,
+	0x00000058, 0xbf8cc07f,
+	0xc0471eb8, 0x0000005c,
+	0xbf8cc07f, 0xbef4007e,
+	0x8675ff7f, 0x0000ffff,
+	0x8775ff75, 0x00040000,
+	0xbef60080, 0xbef700ff,
+	0x00807fac, 0x8670ff7f,
+	0x08000000, 0x8f708370,
+	0x87777077, 0x8670ff7f,
+	0x70000000, 0x8f708170,
+	0x87777077, 0xbefb007c,
+	0xbefa0080, 0xb8fa2a05,
+	0x807a817a, 0x8e7a8a7a,
+	0xb8f01605, 0x80708170,
+	0x8e708670, 0x807a707a,
+	0xbef60084, 0xbef600ff,
+	0x01000000, 0xbefe007c,
+	0xbefc007a, 0xc0611efa,
+	0x0000007c, 0xbf8cc07f,
+	0x807a847a, 0xbefc007e,
+	0xbefe007c, 0xbefc007a,
+	0xc0611b3a, 0x0000007c,
+	0xbf8cc07f, 0x807a847a,
+	0xbefc007e, 0xbefe007c,
+	0xbefc007a, 0xc0611b7a,
+	0x0000007c, 0xbf8cc07f,
+	0x807a847a, 0xbefc007e,
+	0xbefe007c, 0xbefc007a,
+	0xc0611bba, 0x0000007c,
+	0xbf8cc07f, 0x807a847a,
+	0xbefc007e, 0xbefe007c,
+	0xbefc007a, 0xc0611bfa,
+	0x0000007c, 0xbf8cc07f,
+	0x807a847a, 0xbefc007e,
+	0xbefe007c, 0xbefc007a,
+	0xc0611e3a, 0x0000007c,
+	0xbf8cc07f, 0x807a847a,
+	0xbefc007e, 0xb8f1f803,
+	0xbefe007c, 0xbefc007a,
+	0xc0611c7a, 0x0000007c,
+	0xbf8cc07f, 0x807a847a,
+	0xbefc007e, 0xbefe007c,
+	0xbefc007a, 0xc0611a3a,
+	0x0000007c, 0xbf8cc07f,
+	0x807a847a, 0xbefc007e,
+	0xbefe007c, 0xbefc007a,
+	0xc0611a7a, 0x0000007c,
+	0xbf8cc07f, 0x807a847a,
+	0xbefc007e, 0xb8fbf801,
+	0xbefe007c, 0xbefc007a,
+	0xc0611efa, 0x0000007c,
+	0xbf8cc07f, 0x807a847a,
+	0xbefc007e, 0x8670ff7f,
+	0x04000000, 0xbeef0080,
+	0x876f6f70, 0xb8fa2a05,
+	0x807a817a, 0x8e7a8a7a,
+	0xb8f11605, 0x80718171,
+	0x8e718471, 0x8e768271,
+	0xbef600ff, 0x01000000,
+	0xbef20174, 0x80747a74,
+	0x82758075, 0xbefc0080,
+	0xbf800000, 0xbe802b00,
+	0xbe822b02, 0xbe842b04,
+	0xbe862b06, 0xbe882b08,
+	0xbe8a2b0a, 0xbe8c2b0c,
+	0xbe8e2b0e, 0xc06b003a,
+	0x00000000, 0xbf8cc07f,
+	0xc06b013a, 0x00000010,
+	0xbf8cc07f, 0xc06b023a,
+	0x00000020, 0xbf8cc07f,
+	0xc06b033a, 0x00000030,
+	0xbf8cc07f, 0x8074c074,
+	0x82758075, 0x807c907c,
+	0xbf0a717c, 0xbf85ffe7,
+	0xbef40172, 0xbefa0080,
+	0xbefe00c1, 0xbeff00c1,
+	0xbee80080, 0xbee90080,
+	0xbef600ff, 0x01000000,
+	0xe0724000, 0x7a1d0000,
+	0xe0724100, 0x7a1d0100,
+	0xe0724200, 0x7a1d0200,
+	0xe0724300, 0x7a1d0300,
+	0xbefe00c1, 0xbeff00c1,
+	0xb8f14306, 0x8671c171,
+	0xbf84002c, 0xbf8a0000,
+	0x8670ff6f, 0x04000000,
+	0xbf840028, 0x8e718671,
+	0x8e718271, 0xbef60071,
+	0xb8fa2a05, 0x807a817a,
+	0x8e7a8a7a, 0xb8f01605,
+	0x80708170, 0x8e708670,
+	0x807a707a, 0x807aff7a,
+	0x00000080, 0xbef600ff,
+	0x01000000, 0xbefc0080,
+	0xd28c0002, 0x000100c1,
+	0xd28d0003, 0x000204c1,
+	0xd1060002, 0x00011103,
+	0x7e0602ff, 0x00000200,
+	0xbefc00ff, 0x00010000,
+	0xbe800077, 0x8677ff77,
+	0xff7fffff, 0x8777ff77,
+	0x00058000, 0xd8ec0000,
+	0x00000002, 0xbf8cc07f,
+	0xe0765000, 0x7a1d0002,
+	0x68040702, 0xd0c9006a,
+	0x0000e302, 0xbf87fff7,
+	0xbef70000, 0xbefa00ff,
+	0x00000400, 0xbefe00c1,
+	0xbeff00c1, 0xb8f12a05,
+	0x80718171, 0x8e718271,
+	0x8e768871, 0xbef600ff,
+	0x01000000, 0xbefc0084,
+	0xbf0a717c, 0xbf840015,
+	0xbf11017c, 0x8071ff71,
+	0x00001000, 0x7e000300,
+	0x7e020301, 0x7e040302,
+	0x7e060303, 0xe0724000,
+	0x7a1d0000, 0xe0724100,
+	0x7a1d0100, 0xe0724200,
+	0x7a1d0200, 0xe0724300,
+	0x7a1d0300, 0x807c847c,
+	0x807aff7a, 0x00000400,
+	0xbf0a717c, 0xbf85ffef,
+	0xbf9c0000, 0xbf8200d9,
+	0xbef4007e, 0x8675ff7f,
+	0x0000ffff, 0x8775ff75,
+	0x00040000, 0xbef60080,
+	0xbef700ff, 0x00807fac,
+	0x866eff7f, 0x08000000,
+	0x8f6e836e, 0x87776e77,
+	0x866eff7f, 0x70000000,
+	0x8f6e816e, 0x87776e77,
+	0x866eff7f, 0x04000000,
+	0xbf84001e, 0xbefe00c1,
+	0xbeff00c1, 0xb8ef4306,
+	0x866fc16f, 0xbf840019,
+	0x8e6f866f, 0x8e6f826f,
+	0xbef6006f, 0xb8f82a05,
+	0x80788178, 0x8e788a78,
+	0xb8ee1605, 0x806e816e,
+	0x8e6e866e, 0x80786e78,
+	0x8078ff78, 0x00000080,
+	0xbef600ff, 0x01000000,
+	0xbefc0080, 0xe0510000,
+	0x781d0000, 0xe0510100,
+	0x781d0000, 0x807cff7c,
+	0x00000200, 0x8078ff78,
+	0x00000200, 0xbf0a6f7c,
+	0xbf85fff6, 0xbef80080,
+	0xbefe00c1, 0xbeff00c1,
+	0xb8ef2a05, 0x806f816f,
+	0x8e6f826f, 0x8e76886f,
+	0xbef600ff, 0x01000000,
+	0xbeee0078, 0x8078ff78,
+	0x00000400, 0xbefc0084,
+	0xbf11087c, 0x806fff6f,
+	0x00008000, 0xe0524000,
+	0x781d0000, 0xe0524100,
+	0x781d0100, 0xe0524200,
+	0x781d0200, 0xe0524300,
+	0x781d0300, 0xbf8c0f70,
+	0x7e000300, 0x7e020301,
+	0x7e040302, 0x7e060303,
+	0x807c847c, 0x8078ff78,
+	0x00000400, 0xbf0a6f7c,
+	0xbf85ffee, 0xbf9c0000,
+	0xe0524000, 0x6e1d0000,
+	0xe0524100, 0x6e1d0100,
+	0xe0524200, 0x6e1d0200,
+	0xe0524300, 0x6e1d0300,
+	0xb8f82a05, 0x80788178,
+	0x8e788a78, 0xb8ee1605,
+	0x806e816e, 0x8e6e866e,
+	0x80786e78, 0x80f8c078,
+	0xb8ef1605, 0x806f816f,
+	0x8e6f846f, 0x8e76826f,
+	0xbef600ff, 0x01000000,
+	0xbefc006f, 0xc031003a,
+	0x00000078, 0x80f8c078,
+	0xbf8cc07f, 0x80fc907c,
+	0xbf800000, 0xbe802d00,
+	0xbe822d02, 0xbe842d04,
+	0xbe862d06, 0xbe882d08,
+	0xbe8a2d0a, 0xbe8c2d0c,
+	0xbe8e2d0e, 0xbf06807c,
+	0xbf84fff0, 0xb8f82a05,
+	0x80788178, 0x8e788a78,
+	0xb8ee1605, 0x806e816e,
+	0x8e6e866e, 0x80786e78,
+	0xbef60084, 0xbef600ff,
+	0x01000000, 0xc0211bfa,
+	0x00000078, 0x80788478,
+	0xc0211b3a, 0x00000078,
+	0x80788478, 0xc0211b7a,
+	0x00000078, 0x80788478,
+	0xc0211eba, 0x00000078,
+	0x80788478, 0xc0211efa,
+	0x00000078, 0x80788478,
+	0xc0211c3a, 0x00000078,
+	0x80788478, 0xc0211c7a,
+	0x00000078, 0x80788478,
+	0xc0211a3a, 0x00000078,
+	0x80788478, 0xc0211a7a,
+	0x00000078, 0x80788478,
+	0xc0211cfa, 0x00000078,
+	0x80788478, 0xbf8cc07f,
+	0x866dff6d, 0x0000ffff,
+	0xbefc006f, 0xbefe007a,
+	0xbeff007b, 0x866f71ff,
+	0x000003ff, 0xb96f4803,
+	0x866f71ff, 0xfffff800,
+	0x8f6f8b6f, 0xb96fa2c3,
+	0xb973f801, 0xb8ee2a05,
+	0x806e816e, 0x8e6e8a6e,
+	0xb8ef1605, 0x806f816f,
+	0x8e6f866f, 0x806e6f6e,
+	0x806e746e, 0x826f8075,
+	0x866fff6f, 0x0000ffff,
+	0xc0071cb7, 0x00000040,
+	0xc00b1d37, 0x00000048,
+	0xc0031e77, 0x00000058,
+	0xc0071eb7, 0x0000005c,
+	0xbf8cc07f, 0x866fff6d,
+	0xf0000000, 0x8f6f9c6f,
+	0x8e6f906f, 0xbeee0080,
+	0x876e6f6e, 0x866fff6d,
+	0x08000000, 0x8f6f9b6f,
+	0x8e6f8f6f, 0x876e6f6e,
+	0x866fff70, 0x00800000,
+	0x8f6f976f, 0xb96ef807,
+	0x86fe7e7e, 0x86ea6a6a,
+	0xb970f802, 0xbf8a0000,
+	0x95806f6c, 0xbf810000,
+};
diff --git a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm
index 34eabcd..658a4c6 100644
--- a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm
+++ b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm
@@ -20,9 +20,12 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  */
 
-#if 0
-HW (VI) source code for CWSR trap handler
-#Version 18 + multiple trap handler
+/* To compile this assembly code:
+ * PROJECT=vi ./sp3 cwsr_trap_handler_gfx8.asm -hex tmp.hex
+ */
+
+/* HW (VI) source code for CWSR trap handler */
+/* Version 18 + multiple trap handler */
 
 // this performance-optimal version was originally from Seven Xu at SRDC
 
@@ -150,7 +153,7 @@ var s_save_spi_init_lo              =   exec_lo
 var s_save_spi_init_hi              =   exec_hi
 
                                                 //tba_lo and tba_hi need to be saved/restored
-var s_save_pc_lo            =   ttmp0           //{TTMP1, TTMP0} = {3??h0,pc_rewind[3:0], HT[0],trapID[7:0], PC[47:0]}
+var s_save_pc_lo            =   ttmp0           //{TTMP1, TTMP0} = {3'h0,pc_rewind[3:0], HT[0],trapID[7:0], PC[47:0]}
 var s_save_pc_hi            =   ttmp1
 var s_save_exec_lo          =   ttmp2
 var s_save_exec_hi          =   ttmp3
@@ -1132,259 +1135,3 @@ end
 function get_hwreg_size_bytes
     return 128 //HWREG size 128 bytes
 end
-
-
-#endif
-
-static const uint32_t cwsr_trap_gfx8_hex[] = {
-	0xbf820001, 0xbf820125,
-	0xb8f4f802, 0x89748674,
-	0xb8f5f803, 0x8675ff75,
-	0x00000400, 0xbf850011,
-	0xc00a1e37, 0x00000000,
-	0xbf8c007f, 0x87777978,
-	0xbf840002, 0xb974f802,
-	0xbe801d78, 0xb8f5f803,
-	0x8675ff75, 0x000001ff,
-	0xbf850002, 0x80708470,
-	0x82718071, 0x8671ff71,
-	0x0000ffff, 0xb974f802,
-	0xbe801f70, 0xb8f5f803,
-	0x8675ff75, 0x00000100,
-	0xbf840006, 0xbefa0080,
-	0xb97a0203, 0x8671ff71,
-	0x0000ffff, 0x80f08870,
-	0x82f18071, 0xbefa0080,
-	0xb97a0283, 0xbef60068,
-	0xbef70069, 0xb8fa1c07,
-	0x8e7a9c7a, 0x87717a71,
-	0xb8fa03c7, 0x8e7a9b7a,
-	0x87717a71, 0xb8faf807,
-	0x867aff7a, 0x00007fff,
-	0xb97af807, 0xbef2007e,
-	0xbef3007f, 0xbefe0180,
-	0xbf900004, 0x877a8474,
-	0xb97af802, 0xbf8e0002,
-	0xbf88fffe, 0xbef8007e,
-	0x8679ff7f, 0x0000ffff,
-	0x8779ff79, 0x00040000,
-	0xbefa0080, 0xbefb00ff,
-	0x00807fac, 0x867aff7f,
-	0x08000000, 0x8f7a837a,
-	0x877b7a7b, 0x867aff7f,
-	0x70000000, 0x8f7a817a,
-	0x877b7a7b, 0xbeef007c,
-	0xbeee0080, 0xb8ee2a05,
-	0x806e816e, 0x8e6e8a6e,
-	0xb8fa1605, 0x807a817a,
-	0x8e7a867a, 0x806e7a6e,
-	0xbefa0084, 0xbefa00ff,
-	0x01000000, 0xbefe007c,
-	0xbefc006e, 0xc0611bfc,
-	0x0000007c, 0x806e846e,
-	0xbefc007e, 0xbefe007c,
-	0xbefc006e, 0xc0611c3c,
-	0x0000007c, 0x806e846e,
-	0xbefc007e, 0xbefe007c,
-	0xbefc006e, 0xc0611c7c,
-	0x0000007c, 0x806e846e,
-	0xbefc007e, 0xbefe007c,
-	0xbefc006e, 0xc0611cbc,
-	0x0000007c, 0x806e846e,
-	0xbefc007e, 0xbefe007c,
-	0xbefc006e, 0xc0611cfc,
-	0x0000007c, 0x806e846e,
-	0xbefc007e, 0xbefe007c,
-	0xbefc006e, 0xc0611d3c,
-	0x0000007c, 0x806e846e,
-	0xbefc007e, 0xb8f5f803,
-	0xbefe007c, 0xbefc006e,
-	0xc0611d7c, 0x0000007c,
-	0x806e846e, 0xbefc007e,
-	0xbefe007c, 0xbefc006e,
-	0xc0611dbc, 0x0000007c,
-	0x806e846e, 0xbefc007e,
-	0xbefe007c, 0xbefc006e,
-	0xc0611dfc, 0x0000007c,
-	0x806e846e, 0xbefc007e,
-	0xb8eff801, 0xbefe007c,
-	0xbefc006e, 0xc0611bfc,
-	0x0000007c, 0x806e846e,
-	0xbefc007e, 0xbefe007c,
-	0xbefc006e, 0xc0611b3c,
-	0x0000007c, 0x806e846e,
-	0xbefc007e, 0xbefe007c,
-	0xbefc006e, 0xc0611b7c,
-	0x0000007c, 0x806e846e,
-	0xbefc007e, 0x867aff7f,
-	0x04000000, 0xbef30080,
-	0x8773737a, 0xb8ee2a05,
-	0x806e816e, 0x8e6e8a6e,
-	0xb8f51605, 0x80758175,
-	0x8e758475, 0x8e7a8275,
-	0xbefa00ff, 0x01000000,
-	0xbef60178, 0x80786e78,
-	0x82798079, 0xbefc0080,
-	0xbe802b00, 0xbe822b02,
-	0xbe842b04, 0xbe862b06,
-	0xbe882b08, 0xbe8a2b0a,
-	0xbe8c2b0c, 0xbe8e2b0e,
-	0xc06b003c, 0x00000000,
-	0xc06b013c, 0x00000010,
-	0xc06b023c, 0x00000020,
-	0xc06b033c, 0x00000030,
-	0x8078c078, 0x82798079,
-	0x807c907c, 0xbf0a757c,
-	0xbf85ffeb, 0xbef80176,
-	0xbeee0080, 0xbefe00c1,
-	0xbeff00c1, 0xbefa00ff,
-	0x01000000, 0xe0724000,
-	0x6e1e0000, 0xe0724100,
-	0x6e1e0100, 0xe0724200,
-	0x6e1e0200, 0xe0724300,
-	0x6e1e0300, 0xbefe00c1,
-	0xbeff00c1, 0xb8f54306,
-	0x8675c175, 0xbf84002c,
-	0xbf8a0000, 0x867aff73,
-	0x04000000, 0xbf840028,
-	0x8e758675, 0x8e758275,
-	0xbefa0075, 0xb8ee2a05,
-	0x806e816e, 0x8e6e8a6e,
-	0xb8fa1605, 0x807a817a,
-	0x8e7a867a, 0x806e7a6e,
-	0x806eff6e, 0x00000080,
-	0xbefa00ff, 0x01000000,
-	0xbefc0080, 0xd28c0002,
-	0x000100c1, 0xd28d0003,
-	0x000204c1, 0xd1060002,
-	0x00011103, 0x7e0602ff,
-	0x00000200, 0xbefc00ff,
-	0x00010000, 0xbe80007b,
-	0x867bff7b, 0xff7fffff,
-	0x877bff7b, 0x00058000,
-	0xd8ec0000, 0x00000002,
-	0xbf8c007f, 0xe0765000,
-	0x6e1e0002, 0x32040702,
-	0xd0c9006a, 0x0000eb02,
-	0xbf87fff7, 0xbefb0000,
-	0xbeee00ff, 0x00000400,
-	0xbefe00c1, 0xbeff00c1,
-	0xb8f52a05, 0x80758175,
-	0x8e758275, 0x8e7a8875,
-	0xbefa00ff, 0x01000000,
-	0xbefc0084, 0xbf0a757c,
-	0xbf840015, 0xbf11017c,
-	0x8075ff75, 0x00001000,
-	0x7e000300, 0x7e020301,
-	0x7e040302, 0x7e060303,
-	0xe0724000, 0x6e1e0000,
-	0xe0724100, 0x6e1e0100,
-	0xe0724200, 0x6e1e0200,
-	0xe0724300, 0x6e1e0300,
-	0x807c847c, 0x806eff6e,
-	0x00000400, 0xbf0a757c,
-	0xbf85ffef, 0xbf9c0000,
-	0xbf8200ca, 0xbef8007e,
-	0x8679ff7f, 0x0000ffff,
-	0x8779ff79, 0x00040000,
-	0xbefa0080, 0xbefb00ff,
-	0x00807fac, 0x8676ff7f,
-	0x08000000, 0x8f768376,
-	0x877b767b, 0x8676ff7f,
-	0x70000000, 0x8f768176,
-	0x877b767b, 0x8676ff7f,
-	0x04000000, 0xbf84001e,
-	0xbefe00c1, 0xbeff00c1,
-	0xb8f34306, 0x8673c173,
-	0xbf840019, 0x8e738673,
-	0x8e738273, 0xbefa0073,
-	0xb8f22a05, 0x80728172,
-	0x8e728a72, 0xb8f61605,
-	0x80768176, 0x8e768676,
-	0x80727672, 0x8072ff72,
-	0x00000080, 0xbefa00ff,
-	0x01000000, 0xbefc0080,
-	0xe0510000, 0x721e0000,
-	0xe0510100, 0x721e0000,
-	0x807cff7c, 0x00000200,
-	0x8072ff72, 0x00000200,
-	0xbf0a737c, 0xbf85fff6,
-	0xbef20080, 0xbefe00c1,
-	0xbeff00c1, 0xb8f32a05,
-	0x80738173, 0x8e738273,
-	0x8e7a8873, 0xbefa00ff,
-	0x01000000, 0xbef60072,
-	0x8072ff72, 0x00000400,
-	0xbefc0084, 0xbf11087c,
-	0x8073ff73, 0x00008000,
-	0xe0524000, 0x721e0000,
-	0xe0524100, 0x721e0100,
-	0xe0524200, 0x721e0200,
-	0xe0524300, 0x721e0300,
-	0xbf8c0f70, 0x7e000300,
-	0x7e020301, 0x7e040302,
-	0x7e060303, 0x807c847c,
-	0x8072ff72, 0x00000400,
-	0xbf0a737c, 0xbf85ffee,
-	0xbf9c0000, 0xe0524000,
-	0x761e0000, 0xe0524100,
-	0x761e0100, 0xe0524200,
-	0x761e0200, 0xe0524300,
-	0x761e0300, 0xb8f22a05,
-	0x80728172, 0x8e728a72,
-	0xb8f61605, 0x80768176,
-	0x8e768676, 0x80727672,
-	0x80f2c072, 0xb8f31605,
-	0x80738173, 0x8e738473,
-	0x8e7a8273, 0xbefa00ff,
-	0x01000000, 0xbefc0073,
-	0xc031003c, 0x00000072,
-	0x80f2c072, 0xbf8c007f,
-	0x80fc907c, 0xbe802d00,
-	0xbe822d02, 0xbe842d04,
-	0xbe862d06, 0xbe882d08,
-	0xbe8a2d0a, 0xbe8c2d0c,
-	0xbe8e2d0e, 0xbf06807c,
-	0xbf84fff1, 0xb8f22a05,
-	0x80728172, 0x8e728a72,
-	0xb8f61605, 0x80768176,
-	0x8e768676, 0x80727672,
-	0xbefa0084, 0xbefa00ff,
-	0x01000000, 0xc0211cfc,
-	0x00000072, 0x80728472,
-	0xc0211c3c, 0x00000072,
-	0x80728472, 0xc0211c7c,
-	0x00000072, 0x80728472,
-	0xc0211bbc, 0x00000072,
-	0x80728472, 0xc0211bfc,
-	0x00000072, 0x80728472,
-	0xc0211d3c, 0x00000072,
-	0x80728472, 0xc0211d7c,
-	0x00000072, 0x80728472,
-	0xc0211a3c, 0x00000072,
-	0x80728472, 0xc0211a7c,
-	0x00000072, 0x80728472,
-	0xc0211dfc, 0x00000072,
-	0x80728472, 0xc0211b3c,
-	0x00000072, 0x80728472,
-	0xc0211b7c, 0x00000072,
-	0x80728472, 0xbf8c007f,
-	0x8671ff71, 0x0000ffff,
-	0xbefc0073, 0xbefe006e,
-	0xbeff006f, 0x867375ff,
-	0x000003ff, 0xb9734803,
-	0x867375ff, 0xfffff800,
-	0x8f738b73, 0xb973a2c3,
-	0xb977f801, 0x8673ff71,
-	0xf0000000, 0x8f739c73,
-	0x8e739073, 0xbef60080,
-	0x87767376, 0x8673ff71,
-	0x08000000, 0x8f739b73,
-	0x8e738f73, 0x87767376,
-	0x8673ff74, 0x00800000,
-	0x8f739773, 0xb976f807,
-	0x86fe7e7e, 0x86ea6a6a,
-	0xb974f802, 0xbf8a0000,
-	0x95807370, 0xbf810000,
-};
-
diff --git a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm
index cac8d49..065f55a 100644
--- a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm
+++ b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm
@@ -20,9 +20,12 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  */
 
-#if 0
-HW (GFX9) source code for CWSR trap handler
-#Version 18 + multiple trap handler
+/* To compile this assembly code:
+ * PROJECT=greenland ./sp3 cwsr_trap_handler_gfx9.asm -hex tmp.hex
+ */
+
+/* HW (GFX9) source code for CWSR trap handler */
+/* Version 18 + multiple trap handler */
 
 // this performance-optimal version was originally from Seven Xu at SRDC
 
@@ -151,7 +154,7 @@ var S_SAVE_PC_HI_FIRST_REPLAY_MASK	=   0x08000000		//FIXME
 var s_save_spi_init_lo		    =	exec_lo
 var s_save_spi_init_hi		    =	exec_hi
 
-var s_save_pc_lo	    =	ttmp0		//{TTMP1, TTMP0} = {3??h0,pc_rewind[3:0], HT[0],trapID[7:0], PC[47:0]}
+var s_save_pc_lo	    =	ttmp0		//{TTMP1, TTMP0} = {3'h0,pc_rewind[3:0], HT[0],trapID[7:0], PC[47:0]}
 var s_save_pc_hi	    =	ttmp1
 var s_save_exec_lo	    =	ttmp2
 var s_save_exec_hi	    =	ttmp3
@@ -1210,292 +1213,3 @@ function ack_sqc_store_workaround
         s_waitcnt lgkmcnt(0)
     end
 end
-
-
-#endif
-
-static const uint32_t cwsr_trap_gfx9_hex[] = {
-	0xbf820001, 0xbf82015a,
-	0xb8f8f802, 0x89788678,
-	0xb8f1f803, 0x866eff71,
-	0x00000400, 0xbf850034,
-	0x866eff71, 0x00000800,
-	0xbf850003, 0x866eff71,
-	0x00000100, 0xbf840008,
-	0x866eff78, 0x00002000,
-	0xbf840001, 0xbf810000,
-	0x8778ff78, 0x00002000,
-	0x80ec886c, 0x82ed806d,
-	0xb8eef807, 0x866fff6e,
-	0x001f8000, 0x8e6f8b6f,
-	0x8977ff77, 0xfc000000,
-	0x87776f77, 0x896eff6e,
-	0x001f8000, 0xb96ef807,
-	0xb8f0f812, 0xb8f1f813,
-	0x8ef08870, 0xc0071bb8,
-	0x00000000, 0xbf8cc07f,
-	0xc0071c38, 0x00000008,
-	0xbf8cc07f, 0x86ee6e6e,
-	0xbf840001, 0xbe801d6e,
-	0xb8f1f803, 0x8671ff71,
-	0x000001ff, 0xbf850002,
-	0x806c846c, 0x826d806d,
-	0x866dff6d, 0x0000ffff,
-	0x8f6e8b77, 0x866eff6e,
-	0x001f8000, 0xb96ef807,
-	0x86fe7e7e, 0x86ea6a6a,
-	0xb978f802, 0xbe801f6c,
-	0x866dff6d, 0x0000ffff,
-	0xbef00080, 0xb9700283,
-	0xb8f02407, 0x8e709c70,
-	0x876d706d, 0xb8f003c7,
-	0x8e709b70, 0x876d706d,
-	0xb8f0f807, 0x8670ff70,
-	0x00007fff, 0xb970f807,
-	0xbeee007e, 0xbeef007f,
-	0xbefe0180, 0xbf900004,
-	0x87708478, 0xb970f802,
-	0xbf8e0002, 0xbf88fffe,
-	0xb8f02a05, 0x80708170,
-	0x8e708a70, 0xb8f11605,
-	0x80718171, 0x8e718671,
-	0x80707170, 0x80707e70,
-	0x8271807f, 0x8671ff71,
-	0x0000ffff, 0xc0471cb8,
-	0x00000040, 0xbf8cc07f,
-	0xc04b1d38, 0x00000048,
-	0xbf8cc07f, 0xc0431e78,
-	0x00000058, 0xbf8cc07f,
-	0xc0471eb8, 0x0000005c,
-	0xbf8cc07f, 0xbef4007e,
-	0x8675ff7f, 0x0000ffff,
-	0x8775ff75, 0x00040000,
-	0xbef60080, 0xbef700ff,
-	0x00807fac, 0x8670ff7f,
-	0x08000000, 0x8f708370,
-	0x87777077, 0x8670ff7f,
-	0x70000000, 0x8f708170,
-	0x87777077, 0xbefb007c,
-	0xbefa0080, 0xb8fa2a05,
-	0x807a817a, 0x8e7a8a7a,
-	0xb8f01605, 0x80708170,
-	0x8e708670, 0x807a707a,
-	0xbef60084, 0xbef600ff,
-	0x01000000, 0xbefe007c,
-	0xbefc007a, 0xc0611efa,
-	0x0000007c, 0xbf8cc07f,
-	0x807a847a, 0xbefc007e,
-	0xbefe007c, 0xbefc007a,
-	0xc0611b3a, 0x0000007c,
-	0xbf8cc07f, 0x807a847a,
-	0xbefc007e, 0xbefe007c,
-	0xbefc007a, 0xc0611b7a,
-	0x0000007c, 0xbf8cc07f,
-	0x807a847a, 0xbefc007e,
-	0xbefe007c, 0xbefc007a,
-	0xc0611bba, 0x0000007c,
-	0xbf8cc07f, 0x807a847a,
-	0xbefc007e, 0xbefe007c,
-	0xbefc007a, 0xc0611bfa,
-	0x0000007c, 0xbf8cc07f,
-	0x807a847a, 0xbefc007e,
-	0xbefe007c, 0xbefc007a,
-	0xc0611e3a, 0x0000007c,
-	0xbf8cc07f, 0x807a847a,
-	0xbefc007e, 0xb8f1f803,
-	0xbefe007c, 0xbefc007a,
-	0xc0611c7a, 0x0000007c,
-	0xbf8cc07f, 0x807a847a,
-	0xbefc007e, 0xbefe007c,
-	0xbefc007a, 0xc0611a3a,
-	0x0000007c, 0xbf8cc07f,
-	0x807a847a, 0xbefc007e,
-	0xbefe007c, 0xbefc007a,
-	0xc0611a7a, 0x0000007c,
-	0xbf8cc07f, 0x807a847a,
-	0xbefc007e, 0xb8fbf801,
-	0xbefe007c, 0xbefc007a,
-	0xc0611efa, 0x0000007c,
-	0xbf8cc07f, 0x807a847a,
-	0xbefc007e, 0x8670ff7f,
-	0x04000000, 0xbeef0080,
-	0x876f6f70, 0xb8fa2a05,
-	0x807a817a, 0x8e7a8a7a,
-	0xb8f11605, 0x80718171,
-	0x8e718471, 0x8e768271,
-	0xbef600ff, 0x01000000,
-	0xbef20174, 0x80747a74,
-	0x82758075, 0xbefc0080,
-	0xbf800000, 0xbe802b00,
-	0xbe822b02, 0xbe842b04,
-	0xbe862b06, 0xbe882b08,
-	0xbe8a2b0a, 0xbe8c2b0c,
-	0xbe8e2b0e, 0xc06b003a,
-	0x00000000, 0xbf8cc07f,
-	0xc06b013a, 0x00000010,
-	0xbf8cc07f, 0xc06b023a,
-	0x00000020, 0xbf8cc07f,
-	0xc06b033a, 0x00000030,
-	0xbf8cc07f, 0x8074c074,
-	0x82758075, 0x807c907c,
-	0xbf0a717c, 0xbf85ffe7,
-	0xbef40172, 0xbefa0080,
-	0xbefe00c1, 0xbeff00c1,
-	0xbee80080, 0xbee90080,
-	0xbef600ff, 0x01000000,
-	0xe0724000, 0x7a1d0000,
-	0xe0724100, 0x7a1d0100,
-	0xe0724200, 0x7a1d0200,
-	0xe0724300, 0x7a1d0300,
-	0xbefe00c1, 0xbeff00c1,
-	0xb8f14306, 0x8671c171,
-	0xbf84002c, 0xbf8a0000,
-	0x8670ff6f, 0x04000000,
-	0xbf840028, 0x8e718671,
-	0x8e718271, 0xbef60071,
-	0xb8fa2a05, 0x807a817a,
-	0x8e7a8a7a, 0xb8f01605,
-	0x80708170, 0x8e708670,
-	0x807a707a, 0x807aff7a,
-	0x00000080, 0xbef600ff,
-	0x01000000, 0xbefc0080,
-	0xd28c0002, 0x000100c1,
-	0xd28d0003, 0x000204c1,
-	0xd1060002, 0x00011103,
-	0x7e0602ff, 0x00000200,
-	0xbefc00ff, 0x00010000,
-	0xbe800077, 0x8677ff77,
-	0xff7fffff, 0x8777ff77,
-	0x00058000, 0xd8ec0000,
-	0x00000002, 0xbf8cc07f,
-	0xe0765000, 0x7a1d0002,
-	0x68040702, 0xd0c9006a,
-	0x0000e302, 0xbf87fff7,
-	0xbef70000, 0xbefa00ff,
-	0x00000400, 0xbefe00c1,
-	0xbeff00c1, 0xb8f12a05,
-	0x80718171, 0x8e718271,
-	0x8e768871, 0xbef600ff,
-	0x01000000, 0xbefc0084,
-	0xbf0a717c, 0xbf840015,
-	0xbf11017c, 0x8071ff71,
-	0x00001000, 0x7e000300,
-	0x7e020301, 0x7e040302,
-	0x7e060303, 0xe0724000,
-	0x7a1d0000, 0xe0724100,
-	0x7a1d0100, 0xe0724200,
-	0x7a1d0200, 0xe0724300,
-	0x7a1d0300, 0x807c847c,
-	0x807aff7a, 0x00000400,
-	0xbf0a717c, 0xbf85ffef,
-	0xbf9c0000, 0xbf8200d9,
-	0xbef4007e, 0x8675ff7f,
-	0x0000ffff, 0x8775ff75,
-	0x00040000, 0xbef60080,
-	0xbef700ff, 0x00807fac,
-	0x866eff7f, 0x08000000,
-	0x8f6e836e, 0x87776e77,
-	0x866eff7f, 0x70000000,
-	0x8f6e816e, 0x87776e77,
-	0x866eff7f, 0x04000000,
-	0xbf84001e, 0xbefe00c1,
-	0xbeff00c1, 0xb8ef4306,
-	0x866fc16f, 0xbf840019,
-	0x8e6f866f, 0x8e6f826f,
-	0xbef6006f, 0xb8f82a05,
-	0x80788178, 0x8e788a78,
-	0xb8ee1605, 0x806e816e,
-	0x8e6e866e, 0x80786e78,
-	0x8078ff78, 0x00000080,
-	0xbef600ff, 0x01000000,
-	0xbefc0080, 0xe0510000,
-	0x781d0000, 0xe0510100,
-	0x781d0000, 0x807cff7c,
-	0x00000200, 0x8078ff78,
-	0x00000200, 0xbf0a6f7c,
-	0xbf85fff6, 0xbef80080,
-	0xbefe00c1, 0xbeff00c1,
-	0xb8ef2a05, 0x806f816f,
-	0x8e6f826f, 0x8e76886f,
-	0xbef600ff, 0x01000000,
-	0xbeee0078, 0x8078ff78,
-	0x00000400, 0xbefc0084,
-	0xbf11087c, 0x806fff6f,
-	0x00008000, 0xe0524000,
-	0x781d0000, 0xe0524100,
-	0x781d0100, 0xe0524200,
-	0x781d0200, 0xe0524300,
-	0x781d0300, 0xbf8c0f70,
-	0x7e000300, 0x7e020301,
-	0x7e040302, 0x7e060303,
-	0x807c847c, 0x8078ff78,
-	0x00000400, 0xbf0a6f7c,
-	0xbf85ffee, 0xbf9c0000,
-	0xe0524000, 0x6e1d0000,
-	0xe0524100, 0x6e1d0100,
-	0xe0524200, 0x6e1d0200,
-	0xe0524300, 0x6e1d0300,
-	0xb8f82a05, 0x80788178,
-	0x8e788a78, 0xb8ee1605,
-	0x806e816e, 0x8e6e866e,
-	0x80786e78, 0x80f8c078,
-	0xb8ef1605, 0x806f816f,
-	0x8e6f846f, 0x8e76826f,
-	0xbef600ff, 0x01000000,
-	0xbefc006f, 0xc031003a,
-	0x00000078, 0x80f8c078,
-	0xbf8cc07f, 0x80fc907c,
-	0xbf800000, 0xbe802d00,
-	0xbe822d02, 0xbe842d04,
-	0xbe862d06, 0xbe882d08,
-	0xbe8a2d0a, 0xbe8c2d0c,
-	0xbe8e2d0e, 0xbf06807c,
-	0xbf84fff0, 0xb8f82a05,
-	0x80788178, 0x8e788a78,
-	0xb8ee1605, 0x806e816e,
-	0x8e6e866e, 0x80786e78,
-	0xbef60084, 0xbef600ff,
-	0x01000000, 0xc0211bfa,
-	0x00000078, 0x80788478,
-	0xc0211b3a, 0x00000078,
-	0x80788478, 0xc0211b7a,
-	0x00000078, 0x80788478,
-	0xc0211eba, 0x00000078,
-	0x80788478, 0xc0211efa,
-	0x00000078, 0x80788478,
-	0xc0211c3a, 0x00000078,
-	0x80788478, 0xc0211c7a,
-	0x00000078, 0x80788478,
-	0xc0211a3a, 0x00000078,
-	0x80788478, 0xc0211a7a,
-	0x00000078, 0x80788478,
-	0xc0211cfa, 0x00000078,
-	0x80788478, 0xbf8cc07f,
-	0x866dff6d, 0x0000ffff,
-	0xbefc006f, 0xbefe007a,
-	0xbeff007b, 0x866f71ff,
-	0x000003ff, 0xb96f4803,
-	0x866f71ff, 0xfffff800,
-	0x8f6f8b6f, 0xb96fa2c3,
-	0xb973f801, 0xb8ee2a05,
-	0x806e816e, 0x8e6e8a6e,
-	0xb8ef1605, 0x806f816f,
-	0x8e6f866f, 0x806e6f6e,
-	0x806e746e, 0x826f8075,
-	0x866fff6f, 0x0000ffff,
-	0xc0071cb7, 0x00000040,
-	0xc00b1d37, 0x00000048,
-	0xc0031e77, 0x00000058,
-	0xc0071eb7, 0x0000005c,
-	0xbf8cc07f, 0x866fff6d,
-	0xf0000000, 0x8f6f9c6f,
-	0x8e6f906f, 0xbeee0080,
-	0x876e6f6e, 0x866fff6d,
-	0x08000000, 0x8f6f9b6f,
-	0x8e6f8f6f, 0x876e6f6e,
-	0x866fff70, 0x00800000,
-	0x8f6f976f, 0xb96ef807,
-	0x86fe7e7e, 0x86ea6a6a,
-	0xb970f802, 0xbf8a0000,
-	0x95806f6c, 0xbf810000,
-};
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
index c1d9e27..7ee6cec 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
@@ -26,8 +26,7 @@
 #include "kfd_priv.h"
 #include "kfd_device_queue_manager.h"
 #include "kfd_pm4_headers_vi.h"
-#include "cwsr_trap_handler_gfx8.asm"
-#include "cwsr_trap_handler_gfx9.asm"
+#include "cwsr_trap_handler.h"
 #include "kfd_iommu.h"
 
 #define MQD_SIZE_ALIGNED 768
-- 
cgit v1.1


From f8ea72d097965617bba0d6773fd29d44070c5e1a Mon Sep 17 00:00:00 2001
From: Yong Zhao <yong.zhao@amd.com>
Date: Tue, 1 May 2018 17:56:07 -0400
Subject: drm/amdkfd: Fix CP soft hang on APUs

The problem happens on Raven and Carrizo. The context save handler
should not clear the high bits of PC_HI before extracting the bits
of IB_STS.

The bug is not relevant to VEGA10 until we enable demand paging.

Signed-off-by: Jay Cornwall <Jay.Cornwall@amd.com>
Signed-off-by: Yong Zhao <yong.zhao@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h        | 4 ++--
 drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm | 3 +--
 drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm | 3 +--
 3 files changed, 4 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h
index a546a21..f68aef0 100644
--- a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h
+++ b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h
@@ -253,7 +253,6 @@ static const uint32_t cwsr_trap_gfx8_hex[] = {
 	0x00000072, 0x80728472,
 	0xc0211b7c, 0x00000072,
 	0x80728472, 0xbf8c007f,
-	0x8671ff71, 0x0000ffff,
 	0xbefc0073, 0xbefe006e,
 	0xbeff006f, 0x867375ff,
 	0x000003ff, 0xb9734803,
@@ -267,6 +266,7 @@ static const uint32_t cwsr_trap_gfx8_hex[] = {
 	0x8e738f73, 0x87767376,
 	0x8673ff74, 0x00800000,
 	0x8f739773, 0xb976f807,
+	0x8671ff71, 0x0000ffff,
 	0x86fe7e7e, 0x86ea6a6a,
 	0xb974f802, 0xbf8a0000,
 	0x95807370, 0xbf810000,
@@ -530,7 +530,6 @@ static const uint32_t cwsr_trap_gfx9_hex[] = {
 	0x00000078, 0x80788478,
 	0xc0211cfa, 0x00000078,
 	0x80788478, 0xbf8cc07f,
-	0x866dff6d, 0x0000ffff,
 	0xbefc006f, 0xbefe007a,
 	0xbeff007b, 0x866f71ff,
 	0x000003ff, 0xb96f4803,
@@ -554,6 +553,7 @@ static const uint32_t cwsr_trap_gfx9_hex[] = {
 	0x8e6f8f6f, 0x876e6f6e,
 	0x866fff70, 0x00800000,
 	0x8f6f976f, 0xb96ef807,
+	0x866dff6d, 0x0000ffff,
 	0x86fe7e7e, 0x86ea6a6a,
 	0xb970f802, 0xbf8a0000,
 	0x95806f6c, 0xbf810000,
diff --git a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm
index 658a4c6..a2a04bb 100644
--- a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm
+++ b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm
@@ -1015,8 +1015,6 @@ end
 
     s_waitcnt       lgkmcnt(0)                                                                                      //from now on, it is safe to restore STATUS and IB_STS
 
-    s_and_b32 s_restore_pc_hi, s_restore_pc_hi, 0x0000ffff      //pc[47:32]        //Do it here in order not to affect STATUS
-
     //for normal save & restore, the saved PC points to the next inst to execute, no adjustment needs to be made, otherwise:
     if ((EMU_RUN_HACK) && (!EMU_RUN_HACK_RESTORE_NORMAL))
         s_add_u32 s_restore_pc_lo, s_restore_pc_lo, 8            //pc[31:0]+8     //two back-to-back s_trap are used (first for save and second for restore)
@@ -1052,6 +1050,7 @@ end
     s_lshr_b32      s_restore_m0, s_restore_m0, SQ_WAVE_STATUS_INST_ATC_SHIFT
     s_setreg_b32    hwreg(HW_REG_IB_STS),   s_restore_tmp
 
+    s_and_b32 s_restore_pc_hi, s_restore_pc_hi, 0x0000ffff      //pc[47:32]        //Do it here in order not to affect STATUS
     s_and_b64    exec, exec, exec  // Restore STATUS.EXECZ, not writable by s_setreg_b32
     s_and_b64    vcc, vcc, vcc  // Restore STATUS.VCCZ, not writable by s_setreg_b32
     s_setreg_b32    hwreg(HW_REG_STATUS),   s_restore_status     // SCC is included, which is changed by previous salu
diff --git a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm
index 065f55a..998be96 100644
--- a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm
+++ b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm
@@ -1067,8 +1067,6 @@ end
 
     s_waitcnt	    lgkmcnt(0)											    //from now on, it is safe to restore STATUS and IB_STS
 
-    s_and_b32 s_restore_pc_hi, s_restore_pc_hi, 0x0000ffff	//pc[47:32]	   //Do it here in order not to affect STATUS
-
     //for normal save & restore, the saved PC points to the next inst to execute, no adjustment needs to be made, otherwise:
     if ((EMU_RUN_HACK) && (!EMU_RUN_HACK_RESTORE_NORMAL))
 	s_add_u32 s_restore_pc_lo, s_restore_pc_lo, 8		 //pc[31:0]+8	  //two back-to-back s_trap are used (first for save and second for restore)
@@ -1119,6 +1117,7 @@ end
     s_lshr_b32	    s_restore_m0, s_restore_m0, SQ_WAVE_STATUS_INST_ATC_SHIFT
     s_setreg_b32    hwreg(HW_REG_IB_STS),   s_restore_tmp
 
+    s_and_b32 s_restore_pc_hi, s_restore_pc_hi, 0x0000ffff	//pc[47:32]	   //Do it here in order not to affect STATUS
     s_and_b64	 exec, exec, exec  // Restore STATUS.EXECZ, not writable by s_setreg_b32
     s_and_b64	 vcc, vcc, vcc	// Restore STATUS.VCCZ, not writable by s_setreg_b32
     s_setreg_b32    hwreg(HW_REG_STATUS),   s_restore_status	 // SCC is included, which is changed by previous salu
-- 
cgit v1.1


From eeb27b7eb3826c23cc5688c47845e7309f20fc32 Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Tue, 1 May 2018 17:56:08 -0400
Subject: drm/amdkfd: Fix signal handling performance again

It turns out that idr_for_each_entry is really slow compared to just
iterating over the slots. Based on measurements the difference is
estimated to be about a factor 64. That means using idr_for_each_entry
is only worth it with very few allocated events.

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/kfd_events.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_events.c b/drivers/gpu/drm/amd/amdkfd/kfd_events.c
index bccf2f7..5562e94 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_events.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_events.c
@@ -496,7 +496,7 @@ void kfd_signal_event_interrupt(unsigned int pasid, uint32_t partial_id,
 			pr_debug_ratelimited("Partial ID invalid: %u (%u valid bits)\n",
 					     partial_id, valid_id_bits);
 
-		if (p->signal_event_count < KFD_SIGNAL_EVENT_LIMIT/2) {
+		if (p->signal_event_count < KFD_SIGNAL_EVENT_LIMIT / 64) {
 			/* With relatively few events, it's faster to
 			 * iterate over the event IDR
 			 */
-- 
cgit v1.1


From ccb76b149e1c849c0aee6b5043aed74d41064ad6 Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Tue, 1 May 2018 17:56:09 -0400
Subject: drm/amdkfd: Remove initialization of cp_hqd_ib_control on CIK

The initialization is not necessary. amd-kfd-staging and ROCm
releases have worked without it for two years.

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c | 4 ----
 1 file changed, 4 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c
index 2bc49c6..06eaa21 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c
@@ -79,10 +79,6 @@ static int init_mqd(struct mqd_manager *mm, void **mqd,
 	m->cp_mqd_base_addr_lo        = lower_32_bits(addr);
 	m->cp_mqd_base_addr_hi        = upper_32_bits(addr);
 
-	m->cp_hqd_ib_control = DEFAULT_MIN_IB_AVAIL_SIZE | IB_ATC_EN;
-	/* Although WinKFD writes this, I suspect it should not be necessary */
-	m->cp_hqd_ib_control = IB_ATC_EN | DEFAULT_MIN_IB_AVAIL_SIZE;
-
 	m->cp_hqd_quantum = QUANTUM_EN | QUANTUM_SCALE_1MS |
 				QUANTUM_DURATION(10);
 
-- 
cgit v1.1


From bfdcbfd25516eba6cd7b9862779a325ec26006ad Mon Sep 17 00:00:00 2001
From: Ben Goz <ben.goz@amd.com>
Date: Tue, 1 May 2018 17:56:10 -0400
Subject: drm/amdkfd: Locking PM mutex while allocating IB buffer

Signed-off-by: Ben Goz <ben.goz@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c
index 91f0350..c317feb4 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c
@@ -94,12 +94,14 @@ static int pm_allocate_runlist_ib(struct packet_manager *pm,
 
 	pm_calc_rlib_size(pm, rl_buffer_size, is_over_subscription);
 
+	mutex_lock(&pm->lock);
+
 	retval = kfd_gtt_sa_allocate(pm->dqm->dev, *rl_buffer_size,
 					&pm->ib_buffer_obj);
 
 	if (retval) {
 		pr_err("Failed to allocate runlist IB\n");
-		return retval;
+		goto out;
 	}
 
 	*(void **)rl_buffer = pm->ib_buffer_obj->cpu_ptr;
@@ -107,6 +109,9 @@ static int pm_allocate_runlist_ib(struct packet_manager *pm,
 
 	memset(*rl_buffer, 0, *rl_buffer_size);
 	pm->allocated = true;
+
+out:
+	mutex_unlock(&pm->lock);
 	return retval;
 }
 
-- 
cgit v1.1


From 2533f0741e5f7259393d7edecb4bca3106c583c2 Mon Sep 17 00:00:00 2001
From: Shaoyun Liu <Shaoyun.Liu@amd.com>
Date: Tue, 1 May 2018 17:56:11 -0400
Subject: drm/amdkfd: Remove queue node when destroy queue failed

HWS may hang in the middle of destroy queue, remove the queue from the
process queue list so it won't be freed again in the future

Signed-off-by: Shaoyun Liu <Shaoyun.Liu@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c | 10 +++++++---
 1 file changed, 7 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
index 3045aeb..d65ce04 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
@@ -241,7 +241,8 @@ int pqm_create_queue(struct process_queue_manager *pqm,
 	}
 
 	if (retval != 0) {
-		pr_err("DQM create queue failed\n");
+		pr_err("Pasid %d DQM create queue %d failed. ret %d\n",
+			pqm->process->pasid, type, retval);
 		goto err_create_queue;
 	}
 
@@ -319,8 +320,11 @@ int pqm_destroy_queue(struct process_queue_manager *pqm, unsigned int qid)
 		dqm = pqn->q->device->dqm;
 		retval = dqm->ops.destroy_queue(dqm, &pdd->qpd, pqn->q);
 		if (retval) {
-			pr_debug("Destroy queue failed, returned %d\n", retval);
-			goto err_destroy_queue;
+			pr_err("Pasid %d destroy queue %d failed, ret %d\n",
+				pqm->process->pasid,
+				pqn->q->properties.queue_id, retval);
+			if (retval != -ETIME)
+				goto err_destroy_queue;
 		}
 		uninit_queue(pqn->q);
 	}
-- 
cgit v1.1


From c129db1206bd11ab0531a4d91a455a0809acae0e Mon Sep 17 00:00:00 2001
From: Felix Kuehling <Felix.Kuehling@amd.com>
Date: Tue, 1 May 2018 17:56:12 -0400
Subject: drm/amdkfd: Add sanity checks in IRQ handlers

Only accept interrupts from KFD VMIDs. Just checking for a PASID may
not be enough because amdgpu started using PASIDs to map VM faults
to processes.

Warn if an IRQ doesn't have a valid PASID (indicating a firmware bug).

Suggested-by: Shaoyun Liu <Shaoyun.Liu@amd.com>
Suggested-by: Oak Zeng <Oak.Zeng@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/cik_event_interrupt.c | 20 +++++++++---
 drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c  | 40 ++++++++++++++----------
 2 files changed, 39 insertions(+), 21 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/cik_event_interrupt.c b/drivers/gpu/drm/amd/amdkfd/cik_event_interrupt.c
index 3d5ccb3..49df6c7 100644
--- a/drivers/gpu/drm/amd/amdkfd/cik_event_interrupt.c
+++ b/drivers/gpu/drm/amd/amdkfd/cik_event_interrupt.c
@@ -27,18 +27,28 @@
 static bool cik_event_interrupt_isr(struct kfd_dev *dev,
 					const uint32_t *ih_ring_entry)
 {
-	unsigned int pasid;
 	const struct cik_ih_ring_entry *ihre =
 			(const struct cik_ih_ring_entry *)ih_ring_entry;
+	unsigned int vmid, pasid;
+
+	/* Only handle interrupts from KFD VMIDs */
+	vmid  = (ihre->ring_id & 0x0000ff00) >> 8;
+	if (vmid < dev->vm_info.first_vmid_kfd ||
+	    vmid > dev->vm_info.last_vmid_kfd)
+		return 0;
 
+	/* If there is no valid PASID, it's likely a firmware bug */
 	pasid = (ihre->ring_id & 0xffff0000) >> 16;
+	if (WARN_ONCE(pasid == 0, "FW bug: No PASID in KFD interrupt"))
+		return 0;
 
-	/* Do not process in ISR, just request it to be forwarded to WQ. */
-	return (pasid != 0) &&
-		(ihre->source_id == CIK_INTSRC_CP_END_OF_PIPE ||
+	/* Interrupt types we care about: various signals and faults.
+	 * They will be forwarded to a work queue (see below).
+	 */
+	return ihre->source_id == CIK_INTSRC_CP_END_OF_PIPE ||
 		ihre->source_id == CIK_INTSRC_SDMA_TRAP ||
 		ihre->source_id == CIK_INTSRC_SQ_INTERRUPT_MSG ||
-		ihre->source_id == CIK_INTSRC_CP_BAD_OPCODE);
+		ihre->source_id == CIK_INTSRC_CP_BAD_OPCODE;
 }
 
 static void cik_event_interrupt_wq(struct kfd_dev *dev,
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c
index 39d4115..37029ba 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c
@@ -29,27 +29,35 @@ static bool event_interrupt_isr_v9(struct kfd_dev *dev,
 					const uint32_t *ih_ring_entry)
 {
 	uint16_t source_id, client_id, pasid, vmid;
+	const uint32_t *data = ih_ring_entry;
 
-	source_id = SOC15_SOURCE_ID_FROM_IH_ENTRY(ih_ring_entry);
-	client_id = SOC15_CLIENT_ID_FROM_IH_ENTRY(ih_ring_entry);
-	pasid = SOC15_PASID_FROM_IH_ENTRY(ih_ring_entry);
+	/* Only handle interrupts from KFD VMIDs */
 	vmid = SOC15_VMID_FROM_IH_ENTRY(ih_ring_entry);
+	if (vmid < dev->vm_info.first_vmid_kfd ||
+	    vmid > dev->vm_info.last_vmid_kfd)
+		return 0;
+
+	/* If there is no valid PASID, it's likely a firmware bug */
+	pasid = SOC15_PASID_FROM_IH_ENTRY(ih_ring_entry);
+	if (WARN_ONCE(pasid == 0, "FW bug: No PASID in KFD interrupt"))
+		return 0;
 
-	if (pasid) {
-		const uint32_t *data = ih_ring_entry;
+	source_id = SOC15_SOURCE_ID_FROM_IH_ENTRY(ih_ring_entry);
+	client_id = SOC15_CLIENT_ID_FROM_IH_ENTRY(ih_ring_entry);
 
-		pr_debug("client id 0x%x, source id %d, pasid 0x%x. raw data:\n",
-			 client_id, source_id, pasid);
-		pr_debug("%8X, %8X, %8X, %8X, %8X, %8X, %8X, %8X.\n",
-			 data[0], data[1], data[2], data[3],
-			 data[4], data[5], data[6], data[7]);
-	}
+	pr_debug("client id 0x%x, source id %d, pasid 0x%x. raw data:\n",
+		 client_id, source_id, pasid);
+	pr_debug("%8X, %8X, %8X, %8X, %8X, %8X, %8X, %8X.\n",
+		 data[0], data[1], data[2], data[3],
+		 data[4], data[5], data[6], data[7]);
 
-	return (pasid != 0) &&
-		(source_id == SOC15_INTSRC_CP_END_OF_PIPE ||
-		 source_id == SOC15_INTSRC_SDMA_TRAP ||
-		 source_id == SOC15_INTSRC_SQ_INTERRUPT_MSG ||
-		 source_id == SOC15_INTSRC_CP_BAD_OPCODE);
+	/* Interrupt types we care about: various signals and faults.
+	 * They will be forwarded to a work queue (see below).
+	 */
+	return source_id == SOC15_INTSRC_CP_END_OF_PIPE ||
+		source_id == SOC15_INTSRC_SDMA_TRAP ||
+		source_id == SOC15_INTSRC_SQ_INTERRUPT_MSG ||
+		source_id == SOC15_INTSRC_CP_BAD_OPCODE;
 }
 
 static void event_interrupt_wq_v9(struct kfd_dev *dev,
-- 
cgit v1.1


From af47b390273f1068bdb1d01263a81948c4e2f97a Mon Sep 17 00:00:00 2001
From: Laura Abbott <labbott@redhat.com>
Date: Fri, 13 Apr 2018 14:24:12 -0700
Subject: drm/amdkfd: Remove vla
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

There's an ongoing effort to remove VLAs[1] from the kernel to eventually
turn on -Wvla. Switch to a constant value that covers all hardware.

[1] https://lkml.org/lkml/2018/3/7/621

Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Laura Abbott <labbott@redhat.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c | 8 +++++---
 drivers/gpu/drm/amd/amdkfd/kfd_priv.h      | 2 ++
 2 files changed, 7 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c b/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c
index 035c351..db6d933 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c
@@ -139,10 +139,12 @@ static void interrupt_wq(struct work_struct *work)
 {
 	struct kfd_dev *dev = container_of(work, struct kfd_dev,
 						interrupt_work);
+	uint32_t ih_ring_entry[KFD_MAX_RING_ENTRY_SIZE];
 
-	uint32_t ih_ring_entry[DIV_ROUND_UP(
-				dev->device_info->ih_ring_entry_size,
-				sizeof(uint32_t))];
+	if (dev->device_info->ih_ring_entry_size > sizeof(ih_ring_entry)) {
+		dev_err_once(kfd_chardev(), "Ring entry too small\n");
+		return;
+	}
 
 	while (dequeue_ih_ring_entry(dev, ih_ring_entry))
 		dev->device_info->event_interrupt_class->interrupt_wq(dev,
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
index 10d5b54..5e3990b 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
@@ -39,6 +39,8 @@
 
 #include "amd_shared.h"
 
+#define KFD_MAX_RING_ENTRY_SIZE	8
+
 #define KFD_SYSFS_FILE_MODE 0444
 
 /* GPU ID hash width in bits */
-- 
cgit v1.1


From f4ecfbfc32ed0cb502374164638d14c4fb03e916 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Sat, 14 Apr 2018 13:27:54 +0100
Subject: drm/i915: Check whitelist registers across resets

Add a selftest to ensure that we restore the whitelisted registers after
rewrite the registers everytime they might be scrubbed, e.g. module
load, reset and resume. For the other volatile workaround registers, we
export their presence via debugfs and check in igt/gem_workarounds.
However, we don't export the whitelist and rather than do so, let's test
them directly in the kernel.

The test we use is to read the registers back from the CS (this helps us
be sure that the registers will be valid for MI_LRI etc). In order to
generate the expected list, we split intel_whitelist_workarounds_emit
into two phases, the first to build the list and the second to apply.
Inside the test, we only build the list and then check that list against
the hw.

v2: Filter out pre-gen8 as they do not have RING_NONPRIV.
v3: Drop unused engine parameter, no plans to use it now or future.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Oscar Mateo <oscar.mateo@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Oscar Mateo <oscar.mateo@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180414122754.569-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/i915_debugfs.c                |  14 +-
 drivers/gpu/drm/i915/i915_drv.h                    |   1 -
 drivers/gpu/drm/i915/intel_lrc.c                   |   8 +-
 drivers/gpu/drm/i915/intel_ringbuffer.c            |   4 +-
 drivers/gpu/drm/i915/intel_workarounds.c           | 206 +++++++--------
 drivers/gpu/drm/i915/intel_workarounds.h           |   2 +-
 .../gpu/drm/i915/selftests/i915_live_selftests.h   |   1 +
 drivers/gpu/drm/i915/selftests/intel_workarounds.c | 284 +++++++++++++++++++++
 8 files changed, 380 insertions(+), 140 deletions(-)
 create mode 100644 drivers/gpu/drm/i915/selftests/intel_workarounds.c

diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index 2e6652a..e0274f4 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -3304,24 +3304,13 @@ static int i915_shared_dplls_info(struct seq_file *m, void *unused)
 
 static int i915_wa_registers(struct seq_file *m, void *unused)
 {
-	int i;
-	int ret;
-	struct intel_engine_cs *engine;
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
-	struct drm_device *dev = &dev_priv->drm;
 	struct i915_workarounds *workarounds = &dev_priv->workarounds;
-	enum intel_engine_id id;
-
-	ret = mutex_lock_interruptible(&dev->struct_mutex);
-	if (ret)
-		return ret;
+	int i;
 
 	intel_runtime_pm_get(dev_priv);
 
 	seq_printf(m, "Workarounds applied: %d\n", workarounds->count);
-	for_each_engine(engine, dev_priv, id)
-		seq_printf(m, "HW whitelist count for %s: %d\n",
-			   engine->name, workarounds->hw_whitelist_count[id]);
 	for (i = 0; i < workarounds->count; ++i) {
 		i915_reg_t addr;
 		u32 mask, value, read;
@@ -3337,7 +3326,6 @@ static int i915_wa_registers(struct seq_file *m, void *unused)
 	}
 
 	intel_runtime_pm_put(dev_priv);
-	mutex_unlock(&dev->struct_mutex);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index e50d958..8e8667d 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -1297,7 +1297,6 @@ struct i915_wa_reg {
 struct i915_workarounds {
 	struct i915_wa_reg reg[I915_MAX_WA_REGS];
 	u32 count;
-	u32 hw_whitelist_count[I915_NUM_ENGINES];
 };
 
 struct i915_virtual_gpu {
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index c7c8513..4f72858 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -1744,9 +1744,7 @@ static int gen8_init_render_ring(struct intel_engine_cs *engine)
 	if (ret)
 		return ret;
 
-	ret = intel_whitelist_workarounds_apply(engine);
-	if (ret)
-		return ret;
+	intel_whitelist_workarounds_apply(engine);
 
 	/* We need to disable the AsyncFlip performance optimisations in order
 	 * to use MI_WAIT_FOR_EVENT within the CS. It should already be
@@ -1769,9 +1767,7 @@ static int gen9_init_render_ring(struct intel_engine_cs *engine)
 	if (ret)
 		return ret;
 
-	ret = intel_whitelist_workarounds_apply(engine);
-	if (ret)
-		return ret;
+	intel_whitelist_workarounds_apply(engine);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 757bb09..c68ac60 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -618,9 +618,7 @@ static int init_render_ring(struct intel_engine_cs *engine)
 	if (ret)
 		return ret;
 
-	ret = intel_whitelist_workarounds_apply(engine);
-	if (ret)
-		return ret;
+	intel_whitelist_workarounds_apply(engine);
 
 	/* WaTimedSingleVertexDispatch:cl,bw,ctg,elk,ilk,snb */
 	if (IS_GEN(dev_priv, 4, 6))
diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c
index bbbf4ed..ec9d340 100644
--- a/drivers/gpu/drm/i915/intel_workarounds.c
+++ b/drivers/gpu/drm/i915/intel_workarounds.c
@@ -687,170 +687,144 @@ void intel_gt_workarounds_apply(struct drm_i915_private *dev_priv)
 		MISSING_CASE(INTEL_GEN(dev_priv));
 }
 
-static int wa_ring_whitelist_reg(struct intel_engine_cs *engine,
-				 i915_reg_t reg)
-{
-	struct drm_i915_private *dev_priv = engine->i915;
-	struct i915_workarounds *wa = &dev_priv->workarounds;
-	const unsigned int index = wa->hw_whitelist_count[engine->id];
-
-	if (WARN_ON(index >= RING_MAX_NONPRIV_SLOTS))
-		return -EINVAL;
+struct whitelist {
+	i915_reg_t reg[RING_MAX_NONPRIV_SLOTS];
+	unsigned int count;
+	u32 nopid;
+};
 
-	I915_WRITE(RING_FORCE_TO_NONPRIV(engine->mmio_base, index),
-		   i915_mmio_reg_offset(reg));
-	wa->hw_whitelist_count[engine->id]++;
+static void whitelist_reg(struct whitelist *w, i915_reg_t reg)
+{
+	if (GEM_WARN_ON(w->count >= RING_MAX_NONPRIV_SLOTS))
+		return;
 
-	return 0;
+	w->reg[w->count++] = reg;
 }
 
-static int bdw_whitelist_workarounds_apply(struct intel_engine_cs *engine)
+static void bdw_whitelist_build(struct whitelist *w)
 {
-	return 0;
 }
 
-static int chv_whitelist_workarounds_apply(struct intel_engine_cs *engine)
+static void chv_whitelist_build(struct whitelist *w)
 {
-	return 0;
 }
 
-static int gen9_whitelist_workarounds_apply(struct intel_engine_cs *engine)
+static void gen9_whitelist_build(struct whitelist *w)
 {
-	int ret;
-
 	/* WaVFEStateAfterPipeControlwithMediaStateClear:skl,bxt,glk,cfl */
-	ret = wa_ring_whitelist_reg(engine, GEN9_CTX_PREEMPT_REG);
-	if (ret)
-		return ret;
+	whitelist_reg(w, GEN9_CTX_PREEMPT_REG);
 
 	/* WaEnablePreemptionGranularityControlByUMD:skl,bxt,kbl,cfl,[cnl] */
-	ret = wa_ring_whitelist_reg(engine, GEN8_CS_CHICKEN1);
-	if (ret)
-		return ret;
+	whitelist_reg(w, GEN8_CS_CHICKEN1);
 
 	/* WaAllowUMDToModifyHDCChicken1:skl,bxt,kbl,glk,cfl */
-	ret = wa_ring_whitelist_reg(engine, GEN8_HDC_CHICKEN1);
-	if (ret)
-		return ret;
-
-	return 0;
+	whitelist_reg(w, GEN8_HDC_CHICKEN1);
 }
 
-static int skl_whitelist_workarounds_apply(struct intel_engine_cs *engine)
+static void skl_whitelist_build(struct whitelist *w)
 {
-	int ret;
-
-	ret = gen9_whitelist_workarounds_apply(engine);
-	if (ret)
-		return ret;
+	gen9_whitelist_build(w);
 
 	/* WaDisableLSQCROPERFforOCL:skl */
-	ret = wa_ring_whitelist_reg(engine, GEN8_L3SQCREG4);
-	if (ret)
-		return ret;
-
-	return 0;
+	whitelist_reg(w, GEN8_L3SQCREG4);
 }
 
-static int bxt_whitelist_workarounds_apply(struct intel_engine_cs *engine)
+static void bxt_whitelist_build(struct whitelist *w)
 {
-	int ret;
-
-	ret = gen9_whitelist_workarounds_apply(engine);
-	if (ret)
-		return ret;
-
-	return 0;
+	gen9_whitelist_build(w);
 }
 
-static int kbl_whitelist_workarounds_apply(struct intel_engine_cs *engine)
+static void kbl_whitelist_build(struct whitelist *w)
 {
-	int ret;
-
-	ret = gen9_whitelist_workarounds_apply(engine);
-	if (ret)
-		return ret;
+	gen9_whitelist_build(w);
 
 	/* WaDisableLSQCROPERFforOCL:kbl */
-	ret = wa_ring_whitelist_reg(engine, GEN8_L3SQCREG4);
-	if (ret)
-		return ret;
-
-	return 0;
+	whitelist_reg(w, GEN8_L3SQCREG4);
 }
 
-static int glk_whitelist_workarounds_apply(struct intel_engine_cs *engine)
+static void glk_whitelist_build(struct whitelist *w)
 {
-	int ret;
-
-	ret = gen9_whitelist_workarounds_apply(engine);
-	if (ret)
-		return ret;
+	gen9_whitelist_build(w);
 
 	/* WA #0862: Userspace has to set "Barrier Mode" to avoid hangs. */
-	ret = wa_ring_whitelist_reg(engine, GEN9_SLICE_COMMON_ECO_CHICKEN1);
-	if (ret)
-		return ret;
-
-	return 0;
+	whitelist_reg(w, GEN9_SLICE_COMMON_ECO_CHICKEN1);
 }
 
-static int cfl_whitelist_workarounds_apply(struct intel_engine_cs *engine)
+static void cfl_whitelist_build(struct whitelist *w)
 {
-	int ret;
-
-	ret = gen9_whitelist_workarounds_apply(engine);
-	if (ret)
-		return ret;
-
-	return 0;
+	gen9_whitelist_build(w);
 }
 
-static int cnl_whitelist_workarounds_apply(struct intel_engine_cs *engine)
+static void cnl_whitelist_build(struct whitelist *w)
 {
-	int ret;
-
 	/* WaEnablePreemptionGranularityControlByUMD:cnl */
-	ret = wa_ring_whitelist_reg(engine, GEN8_CS_CHICKEN1);
-	if (ret)
-		return ret;
+	whitelist_reg(w, GEN8_CS_CHICKEN1);
+}
+
+static struct whitelist *whitelist_build(struct intel_engine_cs *engine,
+					 struct whitelist *w)
+{
+	struct drm_i915_private *i915 = engine->i915;
+
+	GEM_BUG_ON(engine->id != RCS);
+
+	w->count = 0;
+	w->nopid = i915_mmio_reg_offset(RING_NOPID(engine->mmio_base));
+
+	if (INTEL_GEN(i915) < 8)
+		return NULL;
+	else if (IS_BROADWELL(i915))
+		bdw_whitelist_build(w);
+	else if (IS_CHERRYVIEW(i915))
+		chv_whitelist_build(w);
+	else if (IS_SKYLAKE(i915))
+		skl_whitelist_build(w);
+	else if (IS_BROXTON(i915))
+		bxt_whitelist_build(w);
+	else if (IS_KABYLAKE(i915))
+		kbl_whitelist_build(w);
+	else if (IS_GEMINILAKE(i915))
+		glk_whitelist_build(w);
+	else if (IS_COFFEELAKE(i915))
+		cfl_whitelist_build(w);
+	else if (IS_CANNONLAKE(i915))
+		cnl_whitelist_build(w);
+	else
+		MISSING_CASE(INTEL_GEN(i915));
 
-	return 0;
+	return w;
 }
 
-int intel_whitelist_workarounds_apply(struct intel_engine_cs *engine)
+static void whitelist_apply(struct intel_engine_cs *engine,
+			    const struct whitelist *w)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
-	int err = 0;
+	const u32 base = engine->mmio_base;
+	unsigned int i;
+
+	if (!w)
+		return;
 
-	WARN_ON(engine->id != RCS);
+	intel_uncore_forcewake_get(engine->i915, FORCEWAKE_ALL);
 
-	dev_priv->workarounds.hw_whitelist_count[engine->id] = 0;
+	for (i = 0; i < w->count; i++)
+		I915_WRITE_FW(RING_FORCE_TO_NONPRIV(base, i),
+			      i915_mmio_reg_offset(w->reg[i]));
 
-	if (INTEL_GEN(dev_priv) < 8)
-		err = 0;
-	else if (IS_BROADWELL(dev_priv))
-		err = bdw_whitelist_workarounds_apply(engine);
-	else if (IS_CHERRYVIEW(dev_priv))
-		err = chv_whitelist_workarounds_apply(engine);
-	else if (IS_SKYLAKE(dev_priv))
-		err = skl_whitelist_workarounds_apply(engine);
-	else if (IS_BROXTON(dev_priv))
-		err = bxt_whitelist_workarounds_apply(engine);
-	else if (IS_KABYLAKE(dev_priv))
-		err = kbl_whitelist_workarounds_apply(engine);
-	else if (IS_GEMINILAKE(dev_priv))
-		err = glk_whitelist_workarounds_apply(engine);
-	else if (IS_COFFEELAKE(dev_priv))
-		err = cfl_whitelist_workarounds_apply(engine);
-	else if (IS_CANNONLAKE(dev_priv))
-		err = cnl_whitelist_workarounds_apply(engine);
-	else
-		MISSING_CASE(INTEL_GEN(dev_priv));
-	if (err)
-		return err;
+	/* And clear the rest just in case of garbage */
+	for (; i < RING_MAX_NONPRIV_SLOTS; i++)
+		I915_WRITE_FW(RING_FORCE_TO_NONPRIV(base, i), w->nopid);
 
-	DRM_DEBUG_DRIVER("%s: Number of whitelist w/a: %d\n", engine->name,
-			 dev_priv->workarounds.hw_whitelist_count[engine->id]);
-	return 0;
+	intel_uncore_forcewake_put(engine->i915, FORCEWAKE_ALL);
 }
+
+void intel_whitelist_workarounds_apply(struct intel_engine_cs *engine)
+{
+	struct whitelist w;
+
+	whitelist_apply(engine, whitelist_build(engine, &w));
+}
+
+#if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
+#include "selftests/intel_workarounds.c"
+#endif
diff --git a/drivers/gpu/drm/i915/intel_workarounds.h b/drivers/gpu/drm/i915/intel_workarounds.h
index d9b0cc5..b11d0623 100644
--- a/drivers/gpu/drm/i915/intel_workarounds.h
+++ b/drivers/gpu/drm/i915/intel_workarounds.h
@@ -12,6 +12,6 @@ int intel_ctx_workarounds_emit(struct i915_request *rq);
 
 void intel_gt_workarounds_apply(struct drm_i915_private *dev_priv);
 
-int intel_whitelist_workarounds_apply(struct intel_engine_cs *engine);
+void intel_whitelist_workarounds_apply(struct intel_engine_cs *engine);
 
 #endif
diff --git a/drivers/gpu/drm/i915/selftests/i915_live_selftests.h b/drivers/gpu/drm/i915/selftests/i915_live_selftests.h
index 8bf6aa5..a00e2bd 100644
--- a/drivers/gpu/drm/i915/selftests/i915_live_selftests.h
+++ b/drivers/gpu/drm/i915/selftests/i915_live_selftests.h
@@ -11,6 +11,7 @@
  */
 selftest(sanitycheck, i915_live_sanitycheck) /* keep first (igt selfcheck) */
 selftest(uncore, intel_uncore_live_selftests)
+selftest(workarounds, intel_workarounds_live_selftests)
 selftest(requests, i915_request_live_selftests)
 selftest(objects, i915_gem_object_live_selftests)
 selftest(dmabuf, i915_gem_dmabuf_live_selftests)
diff --git a/drivers/gpu/drm/i915/selftests/intel_workarounds.c b/drivers/gpu/drm/i915/selftests/intel_workarounds.c
new file mode 100644
index 0000000..fe7deca
--- /dev/null
+++ b/drivers/gpu/drm/i915/selftests/intel_workarounds.c
@@ -0,0 +1,284 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2018 Intel Corporation
+ */
+
+#include "../i915_selftest.h"
+
+#include "mock_context.h"
+
+static struct drm_i915_gem_object *
+read_nonprivs(struct i915_gem_context *ctx, struct intel_engine_cs *engine)
+{
+	struct drm_i915_gem_object *result;
+	struct i915_request *rq;
+	struct i915_vma *vma;
+	const u32 base = engine->mmio_base;
+	u32 srm, *cs;
+	int err;
+	int i;
+
+	result = i915_gem_object_create_internal(engine->i915, PAGE_SIZE);
+	if (IS_ERR(result))
+		return result;
+
+	i915_gem_object_set_cache_level(result, I915_CACHE_LLC);
+
+	cs = i915_gem_object_pin_map(result, I915_MAP_WB);
+	if (IS_ERR(cs)) {
+		err = PTR_ERR(cs);
+		goto err_obj;
+	}
+	memset(cs, 0xc5, PAGE_SIZE);
+	i915_gem_object_unpin_map(result);
+
+	vma = i915_vma_instance(result, &engine->i915->ggtt.base, NULL);
+	if (IS_ERR(vma)) {
+		err = PTR_ERR(vma);
+		goto err_obj;
+	}
+
+	err = i915_vma_pin(vma, 0, 0, PIN_GLOBAL);
+	if (err)
+		goto err_obj;
+
+	rq = i915_request_alloc(engine, ctx);
+	if (IS_ERR(rq)) {
+		err = PTR_ERR(rq);
+		goto err_pin;
+	}
+
+	srm = MI_STORE_REGISTER_MEM | MI_SRM_LRM_GLOBAL_GTT;
+	if (INTEL_GEN(ctx->i915) >= 8)
+		srm++;
+
+	cs = intel_ring_begin(rq, 4 * RING_MAX_NONPRIV_SLOTS);
+	for (i = 0; i < RING_MAX_NONPRIV_SLOTS; i++) {
+		*cs++ = srm;
+		*cs++ = i915_mmio_reg_offset(RING_FORCE_TO_NONPRIV(base, i));
+		*cs++ = i915_ggtt_offset(vma) + sizeof(u32) * i;
+		*cs++ = 0;
+	}
+	intel_ring_advance(rq, cs);
+
+	i915_vma_move_to_active(vma, rq, EXEC_OBJECT_WRITE);
+	reservation_object_lock(vma->resv, NULL);
+	reservation_object_add_excl_fence(vma->resv, &rq->fence);
+	reservation_object_unlock(vma->resv);
+
+	i915_gem_object_get(result);
+	i915_gem_object_set_active_reference(result);
+
+	__i915_request_add(rq, true);
+	i915_vma_unpin(vma);
+
+	return result;
+
+err_pin:
+	i915_vma_unpin(vma);
+err_obj:
+	i915_gem_object_put(result);
+	return ERR_PTR(err);
+}
+
+static u32 get_whitelist_reg(const struct whitelist *w, unsigned int i)
+{
+	return i < w->count ? i915_mmio_reg_offset(w->reg[i]) : w->nopid;
+}
+
+static void print_results(const struct whitelist *w, const u32 *results)
+{
+	unsigned int i;
+
+	for (i = 0; i < RING_MAX_NONPRIV_SLOTS; i++) {
+		u32 expected = get_whitelist_reg(w, i);
+		u32 actual = results[i];
+
+		pr_info("RING_NONPRIV[%d]: expected 0x%08x, found 0x%08x\n",
+			i, expected, actual);
+	}
+}
+
+static int check_whitelist(const struct whitelist *w,
+			   struct i915_gem_context *ctx,
+			   struct intel_engine_cs *engine)
+{
+	struct drm_i915_gem_object *results;
+	u32 *vaddr;
+	int err;
+	int i;
+
+	results = read_nonprivs(ctx, engine);
+	if (IS_ERR(results))
+		return PTR_ERR(results);
+
+	err = i915_gem_object_set_to_cpu_domain(results, false);
+	if (err)
+		goto out_put;
+
+	vaddr = i915_gem_object_pin_map(results, I915_MAP_WB);
+	if (IS_ERR(vaddr)) {
+		err = PTR_ERR(vaddr);
+		goto out_put;
+	}
+
+	for (i = 0; i < RING_MAX_NONPRIV_SLOTS; i++) {
+		u32 expected = get_whitelist_reg(w, i);
+		u32 actual = vaddr[i];
+
+		if (expected != actual) {
+			print_results(w, vaddr);
+			pr_err("Invalid RING_NONPRIV[%d], expected 0x%08x, found 0x%08x\n",
+			       i, expected, actual);
+
+			err = -EINVAL;
+			break;
+		}
+	}
+
+	i915_gem_object_unpin_map(results);
+out_put:
+	i915_gem_object_put(results);
+	return err;
+}
+
+static int do_device_reset(struct intel_engine_cs *engine)
+{
+	i915_reset(engine->i915, ENGINE_MASK(engine->id), NULL);
+	return 0;
+}
+
+static int do_engine_reset(struct intel_engine_cs *engine)
+{
+	return i915_reset_engine(engine, NULL);
+}
+
+static int switch_to_scratch_context(struct intel_engine_cs *engine)
+{
+	struct i915_gem_context *ctx;
+	struct i915_request *rq;
+
+	ctx = kernel_context(engine->i915);
+	if (IS_ERR(ctx))
+		return PTR_ERR(ctx);
+
+	rq = i915_request_alloc(engine, ctx);
+	kernel_context_close(ctx);
+	if (IS_ERR(rq))
+		return PTR_ERR(rq);
+
+	i915_request_add(rq);
+
+	return 0;
+}
+
+static int check_whitelist_across_reset(struct intel_engine_cs *engine,
+					int (*reset)(struct intel_engine_cs *),
+					const struct whitelist *w,
+					const char *name)
+{
+	struct i915_gem_context *ctx;
+	int err;
+
+	ctx = kernel_context(engine->i915);
+	if (IS_ERR(ctx))
+		return PTR_ERR(ctx);
+
+	err = check_whitelist(w, ctx, engine);
+	if (err) {
+		pr_err("Invalid whitelist *before* %s reset!\n", name);
+		goto out;
+	}
+
+	err = switch_to_scratch_context(engine);
+	if (err)
+		goto out;
+
+	err = reset(engine);
+	if (err) {
+		pr_err("%s reset failed\n", name);
+		goto out;
+	}
+
+	err = check_whitelist(w, ctx, engine);
+	if (err) {
+		pr_err("Whitelist not preserved in context across %s reset!\n",
+		       name);
+		goto out;
+	}
+
+	kernel_context_close(ctx);
+
+	ctx = kernel_context(engine->i915);
+	if (IS_ERR(ctx))
+		return PTR_ERR(ctx);
+
+	err = check_whitelist(w, ctx, engine);
+	if (err) {
+		pr_err("Invalid whitelist *after* %s reset in fresh context!\n",
+		       name);
+		goto out;
+	}
+
+out:
+	kernel_context_close(ctx);
+	return err;
+}
+
+static int live_reset_whitelist(void *arg)
+{
+	struct drm_i915_private *i915 = arg;
+	struct intel_engine_cs *engine = i915->engine[RCS];
+	struct i915_gpu_error *error = &i915->gpu_error;
+	struct whitelist w;
+	int err;
+
+	/* If we reset the gpu, we should not lose the RING_NONPRIV */
+
+	if (!engine)
+		return 0;
+
+	if (!whitelist_build(engine, &w))
+		return 0;
+
+	pr_info("Checking %d whitelisted registers (RING_NONPRIV)\n", w.count);
+
+	set_bit(I915_RESET_BACKOFF, &error->flags);
+	set_bit(I915_RESET_ENGINE + engine->id, &error->flags);
+
+	if (intel_has_reset_engine(i915)) {
+		err = check_whitelist_across_reset(engine,
+						   do_engine_reset, &w,
+						   "engine");
+		if (err)
+			goto out;
+	}
+
+	if (intel_has_gpu_reset(i915)) {
+		err = check_whitelist_across_reset(engine,
+						   do_device_reset, &w,
+						   "device");
+		if (err)
+			goto out;
+	}
+
+out:
+	clear_bit(I915_RESET_ENGINE + engine->id, &error->flags);
+	clear_bit(I915_RESET_BACKOFF, &error->flags);
+	return err;
+}
+
+int intel_workarounds_live_selftests(struct drm_i915_private *i915)
+{
+	static const struct i915_subtest tests[] = {
+		SUBTEST(live_reset_whitelist),
+	};
+	int err;
+
+	mutex_lock(&i915->drm.struct_mutex);
+	err = i915_subtests(tests, i915);
+	mutex_unlock(&i915->drm.struct_mutex);
+
+	return err;
+}
-- 
cgit v1.1


From 9f172f6fbd243759c808d97bd83c95e49325b2c9 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Sat, 14 Apr 2018 10:12:33 +0100
Subject: drm/i915: Call i915_perf_fini() on init_hw error unwind

We have to cleanup after i915_perf_init(), even on the error path, as it
passes a pointer into the module to the sysfs core. If we fail to
unregister the sysctl table, we leave a dangling pointer which then may
explode anytime later.

Fixes: 9f9b2792b6d3 ("drm/i915/perf: reuse timestamp frequency from device info")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180414091233.32224-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/i915_drv.c | 27 +++++++++++++++------------
 1 file changed, 15 insertions(+), 12 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index f770be1..8400206 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -1101,30 +1101,32 @@ static int i915_driver_init_hw(struct drm_i915_private *dev_priv)
 
 	ret = i915_ggtt_probe_hw(dev_priv);
 	if (ret)
-		return ret;
+		goto err_perf;
 
-	/* WARNING: Apparently we must kick fbdev drivers before vgacon,
-	 * otherwise the vga fbdev driver falls over. */
+	/*
+	 * WARNING: Apparently we must kick fbdev drivers before vgacon,
+	 * otherwise the vga fbdev driver falls over.
+	 */
 	ret = i915_kick_out_firmware_fb(dev_priv);
 	if (ret) {
 		DRM_ERROR("failed to remove conflicting framebuffer drivers\n");
-		goto out_ggtt;
+		goto err_ggtt;
 	}
 
 	ret = i915_kick_out_vgacon(dev_priv);
 	if (ret) {
 		DRM_ERROR("failed to remove conflicting VGA console\n");
-		goto out_ggtt;
+		goto err_ggtt;
 	}
 
 	ret = i915_ggtt_init_hw(dev_priv);
 	if (ret)
-		return ret;
+		goto err_ggtt;
 
 	ret = i915_ggtt_enable_hw(dev_priv);
 	if (ret) {
 		DRM_ERROR("failed to enable GGTT\n");
-		goto out_ggtt;
+		goto err_ggtt;
 	}
 
 	pci_set_master(pdev);
@@ -1135,7 +1137,7 @@ static int i915_driver_init_hw(struct drm_i915_private *dev_priv)
 		if (ret) {
 			DRM_ERROR("failed to set DMA mask\n");
 
-			goto out_ggtt;
+			goto err_ggtt;
 		}
 	}
 
@@ -1153,7 +1155,7 @@ static int i915_driver_init_hw(struct drm_i915_private *dev_priv)
 		if (ret) {
 			DRM_ERROR("failed to set DMA mask\n");
 
-			goto out_ggtt;
+			goto err_ggtt;
 		}
 	}
 
@@ -1186,13 +1188,14 @@ static int i915_driver_init_hw(struct drm_i915_private *dev_priv)
 
 	ret = intel_gvt_init(dev_priv);
 	if (ret)
-		goto out_ggtt;
+		goto err_ggtt;
 
 	return 0;
 
-out_ggtt:
+err_ggtt:
 	i915_ggtt_cleanup_hw(dev_priv);
-
+err_perf:
+	i915_perf_fini(dev_priv);
 	return ret;
 }
 
-- 
cgit v1.1


From b48c05ab5d32af2af4bc63851c153782d1c6ba42 Mon Sep 17 00:00:00 2001
From: Steffen Klassert <steffen.klassert@secunet.com>
Date: Mon, 16 Apr 2018 07:50:09 +0200
Subject: xfrm: Fix warning in xfrm6_tunnel_net_exit.

We need to make sure that all states are really deleted
before we check that the state lists are empty. Otherwise
we trigger a warning.

Fixes: baeb0dbbb5659 ("xfrm6_tunnel: exit_net cleanup check added")
Reported-and-tested-by:syzbot+777bf170a89e7b326405@syzkaller.appspotmail.com
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
---
 include/net/xfrm.h      | 1 +
 net/ipv6/xfrm6_tunnel.c | 3 +++
 net/xfrm/xfrm_state.c   | 6 ++++++
 3 files changed, 10 insertions(+)

diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index a872379..45e75c3 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -375,6 +375,7 @@ struct xfrm_input_afinfo {
 int xfrm_input_register_afinfo(const struct xfrm_input_afinfo *afinfo);
 int xfrm_input_unregister_afinfo(const struct xfrm_input_afinfo *afinfo);
 
+void xfrm_flush_gc(void);
 void xfrm_state_delete_tunnel(struct xfrm_state *x);
 
 struct xfrm_type {
diff --git a/net/ipv6/xfrm6_tunnel.c b/net/ipv6/xfrm6_tunnel.c
index f85f0d7..4a46df8 100644
--- a/net/ipv6/xfrm6_tunnel.c
+++ b/net/ipv6/xfrm6_tunnel.c
@@ -341,6 +341,9 @@ static void __net_exit xfrm6_tunnel_net_exit(struct net *net)
 	struct xfrm6_tunnel_net *xfrm6_tn = xfrm6_tunnel_pernet(net);
 	unsigned int i;
 
+	xfrm_state_flush(net, IPSEC_PROTO_ANY, false);
+	xfrm_flush_gc();
+
 	for (i = 0; i < XFRM6_TUNNEL_SPI_BYADDR_HSIZE; i++)
 		WARN_ON_ONCE(!hlist_empty(&xfrm6_tn->spi_byaddr[i]));
 
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index f9d2f22..6c177ae 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -2175,6 +2175,12 @@ struct xfrm_state_afinfo *xfrm_state_get_afinfo(unsigned int family)
 	return afinfo;
 }
 
+void xfrm_flush_gc(void)
+{
+	flush_work(&xfrm_state_gc_work);
+}
+EXPORT_SYMBOL(xfrm_flush_gc);
+
 /* Temporarily located here until net/xfrm/xfrm_tunnel.c is created */
 void xfrm_state_delete_tunnel(struct xfrm_state *x)
 {
-- 
cgit v1.1


From d54e79340ff8d65b6c63ac278158add2fe211fd0 Mon Sep 17 00:00:00 2001
From: Xiong Zhang <xiong.y.zhang@intel.com>
Date: Fri, 13 Apr 2018 10:26:16 +0800
Subject: drm/i915/gvt: Dereference msi eventfd_ctx when it isn't used anymore

kvmgt get msi eventfd_ctx at qemu vfio set irq eventfd, then
msi eventfd_ctx should be put at some point.
The first point is kvmgt handle qemu vfio_disable_irqindex()
call which has DATA_NONE and ACTION_TRIGGER in flags.
If qemu doesn't call vfio_disable_irqindex(), the second point
is vgpu release function.

v2: Don't inject msi interrupt into guest if eventfd_ctx is dereferenced

Signed-off-by: Xiong Zhang <xiong.y.zhang@intel.com>
Reviewed-by: Zhenyu Wang <zhenyuw@linux.intel.com>
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
---
 drivers/gpu/drm/i915/gvt/kvmgt.c | 28 +++++++++++++++++++++++++++-
 1 file changed, 27 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/gvt/kvmgt.c b/drivers/gpu/drm/i915/gvt/kvmgt.c
index 021f722..a7487f4 100644
--- a/drivers/gpu/drm/i915/gvt/kvmgt.c
+++ b/drivers/gpu/drm/i915/gvt/kvmgt.c
@@ -566,6 +566,17 @@ out:
 	return ret;
 }
 
+static void intel_vgpu_release_msi_eventfd_ctx(struct intel_vgpu *vgpu)
+{
+	struct eventfd_ctx *trigger;
+
+	trigger = vgpu->vdev.msi_trigger;
+	if (trigger) {
+		eventfd_ctx_put(trigger);
+		vgpu->vdev.msi_trigger = NULL;
+	}
+}
+
 static void __intel_vgpu_release(struct intel_vgpu *vgpu)
 {
 	struct kvmgt_guest_info *info;
@@ -590,6 +601,8 @@ static void __intel_vgpu_release(struct intel_vgpu *vgpu)
 	info = (struct kvmgt_guest_info *)vgpu->handle;
 	kvmgt_guest_exit(info);
 
+	intel_vgpu_release_msi_eventfd_ctx(vgpu);
+
 	vgpu->vdev.kvm = NULL;
 	vgpu->handle = 0;
 }
@@ -970,7 +983,8 @@ static int intel_vgpu_set_msi_trigger(struct intel_vgpu *vgpu,
 			return PTR_ERR(trigger);
 		}
 		vgpu->vdev.msi_trigger = trigger;
-	}
+	} else if ((flags & VFIO_IRQ_SET_DATA_NONE) && !count)
+		intel_vgpu_release_msi_eventfd_ctx(vgpu);
 
 	return 0;
 }
@@ -1566,6 +1580,18 @@ static int kvmgt_inject_msi(unsigned long handle, u32 addr, u16 data)
 	info = (struct kvmgt_guest_info *)handle;
 	vgpu = info->vgpu;
 
+	/*
+	 * When guest is poweroff, msi_trigger is set to NULL, but vgpu's
+	 * config and mmio register isn't restored to default during guest
+	 * poweroff. If this vgpu is still used in next vm, this vgpu's pipe
+	 * may be enabled, then once this vgpu is active, it will get inject
+	 * vblank interrupt request. But msi_trigger is null until msi is
+	 * enabled by guest. so if msi_trigger is null, success is still
+	 * returned and don't inject interrupt into guest.
+	 */
+	if (vgpu->vdev.msi_trigger == NULL)
+		return 0;
+
 	if (eventfd_signal(vgpu->vdev.msi_trigger, 1) == 1)
 		return 0;
 
-- 
cgit v1.1


From 99bf8f27f3f94d2a37291354b8dc83f13728f75f Mon Sep 17 00:00:00 2001
From: Lukasz Majewski <lukma@denx.de>
Date: Wed, 4 Apr 2018 09:52:04 +0200
Subject: doc: Add vendor prefix for Kieback & Peter GmbH

The 'kiebackpeter' entry has been added to vendor-prefixes.txt to indicate
products from Kieback & Peter GmbH.

Signed-off-by: Lukasz Majewski <lukma@denx.de>
Signed-off-by: Rob Herring <robh@kernel.org>
---
 Documentation/devicetree/bindings/vendor-prefixes.txt | 1 +
 1 file changed, 1 insertion(+)

diff --git a/Documentation/devicetree/bindings/vendor-prefixes.txt b/Documentation/devicetree/bindings/vendor-prefixes.txt
index b5f978a..a38d8bf 100644
--- a/Documentation/devicetree/bindings/vendor-prefixes.txt
+++ b/Documentation/devicetree/bindings/vendor-prefixes.txt
@@ -182,6 +182,7 @@ karo	Ka-Ro electronics GmbH
 keithkoep	Keith & Koep GmbH
 keymile	Keymile GmbH
 khadas	Khadas
+kiebackpeter    Kieback & Peter GmbH
 kinetic Kinetic Technologies
 kingnovel	Kingnovel Technology Co., Ltd.
 kosagi	Sutajio Ko-Usagi PTE Ltd.
-- 
cgit v1.1


From b614e905a0bc8fc5d4fa72665ac26ae00c874a4e Mon Sep 17 00:00:00 2001
From: Matheus Castello <matheus@castello.eng.br>
Date: Wed, 11 Apr 2018 01:17:03 -0400
Subject: dt-bindings: pinctrl: sunxi: Fix reference to driver

Bindings describe hardware, not drivers.
Use reference to hardware Allwinner A1X Pin Controller instead driver.

Signed-off-by: Matheus Castello <matheus@castello.eng.br>
Signed-off-by: Rob Herring <robh@kernel.org>
---
 .../devicetree/bindings/pinctrl/allwinner,sunxi-pinctrl.txt         | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/Documentation/devicetree/bindings/pinctrl/allwinner,sunxi-pinctrl.txt b/Documentation/devicetree/bindings/pinctrl/allwinner,sunxi-pinctrl.txt
index ed5eb54..64bc5c2 100644
--- a/Documentation/devicetree/bindings/pinctrl/allwinner,sunxi-pinctrl.txt
+++ b/Documentation/devicetree/bindings/pinctrl/allwinner,sunxi-pinctrl.txt
@@ -56,9 +56,9 @@ pins it needs, and how they should be configured, with regard to muxer
 configuration, drive strength and pullups. If one of these options is
 not set, its actual value will be unspecified.
 
-This driver supports the generic pin multiplexing and configuration
-bindings. For details on each properties, you can refer to
-./pinctrl-bindings.txt.
+Allwinner A1X Pin Controller supports the generic pin multiplexing and
+configuration bindings. For details on each properties, you can refer to
+ ./pinctrl-bindings.txt.
 
 Required sub-node properties:
   - pins
-- 
cgit v1.1


From 1a862488729a6ea9cfd285d2c90f8738949ae7d2 Mon Sep 17 00:00:00 2001
From: Jacopo Mondi <jacopo+renesas@jmondi.org>
Date: Mon, 16 Apr 2018 15:55:17 +0200
Subject: dt-bindings: net: ravb: Add support for r8a77965 SoC

Add documentation for r8a77965 compatible string to renesas ravb device
tree bindings documentation.

Signed-off-by: Jacopo Mondi <jacopo+renesas@jmondi.org>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
Acked-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: Rob Herring <robh@kernel.org>
---
 Documentation/devicetree/bindings/net/renesas,ravb.txt | 1 +
 1 file changed, 1 insertion(+)

diff --git a/Documentation/devicetree/bindings/net/renesas,ravb.txt b/Documentation/devicetree/bindings/net/renesas,ravb.txt
index c306f55..890526d 100644
--- a/Documentation/devicetree/bindings/net/renesas,ravb.txt
+++ b/Documentation/devicetree/bindings/net/renesas,ravb.txt
@@ -18,6 +18,7 @@ Required properties:
 
       - "renesas,etheravb-r8a7795" for the R8A7795 SoC.
       - "renesas,etheravb-r8a7796" for the R8A7796 SoC.
+      - "renesas,etheravb-r8a77965" for the R8A77965 SoC.
       - "renesas,etheravb-r8a77970" for the R8A77970 SoC.
       - "renesas,etheravb-r8a77980" for the R8A77980 SoC.
       - "renesas,etheravb-r8a77995" for the R8A77995 SoC.
-- 
cgit v1.1


From 7de5b7e5f6a67c285b86d1478e8e150929c93482 Mon Sep 17 00:00:00 2001
From: Jacopo Mondi <jacopo+renesas@jmondi.org>
Date: Mon, 16 Apr 2018 15:55:28 +0200
Subject: dt-bindings: serial: sh-sci: Add support for r8a77965 (H)SCIF

Add documentation for r8a77965 compatible string to Renesas sci-serial
device tree bindings documentation.

Signed-off-by: Jacopo Mondi <jacopo+renesas@jmondi.org>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Rob Herring <robh@kernel.org>
---
 Documentation/devicetree/bindings/serial/renesas,sci-serial.txt | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/Documentation/devicetree/bindings/serial/renesas,sci-serial.txt b/Documentation/devicetree/bindings/serial/renesas,sci-serial.txt
index ad962f4..0cc5417 100644
--- a/Documentation/devicetree/bindings/serial/renesas,sci-serial.txt
+++ b/Documentation/devicetree/bindings/serial/renesas,sci-serial.txt
@@ -41,6 +41,8 @@ Required properties:
     - "renesas,hscif-r8a7795" for R8A7795 (R-Car H3) HSCIF compatible UART.
     - "renesas,scif-r8a7796" for R8A7796 (R-Car M3-W) SCIF compatible UART.
     - "renesas,hscif-r8a7796" for R8A7796 (R-Car M3-W) HSCIF compatible UART.
+    - "renesas,scif-r8a77965" for R8A77965 (R-Car M3-N) SCIF compatible UART.
+    - "renesas,hscif-r8a77965" for R8A77965 (R-Car M3-N) HSCIF compatible UART.
     - "renesas,scif-r8a77970" for R8A77970 (R-Car V3M) SCIF compatible UART.
     - "renesas,hscif-r8a77970" for R8A77970 (R-Car V3M) HSCIF compatible UART.
     - "renesas,scif-r8a77980" for R8A77980 (R-Car V3H) SCIF compatible UART.
-- 
cgit v1.1


From b89bc283286b105e50aab9ab35992c0237ac77d8 Mon Sep 17 00:00:00 2001
From: Jacopo Mondi <jacopo+renesas@jmondi.org>
Date: Mon, 16 Apr 2018 15:56:08 +0200
Subject: dt-bindings: dmaengine: rcar-dmac: document R8A77965 support

Add documentation for r8a77965 compatible string to rcar-dmac device
tree bindings documentation.

Signed-off-by: Jacopo Mondi <jacopo+renesas@jmondi.org>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
Signed-off-by: Rob Herring <robh@kernel.org>
---
 Documentation/devicetree/bindings/dma/renesas,rcar-dmac.txt | 1 +
 1 file changed, 1 insertion(+)

diff --git a/Documentation/devicetree/bindings/dma/renesas,rcar-dmac.txt b/Documentation/devicetree/bindings/dma/renesas,rcar-dmac.txt
index aadfb23..61315ea 100644
--- a/Documentation/devicetree/bindings/dma/renesas,rcar-dmac.txt
+++ b/Documentation/devicetree/bindings/dma/renesas,rcar-dmac.txt
@@ -26,6 +26,7 @@ Required Properties:
 		- "renesas,dmac-r8a7794" (R-Car E2)
 		- "renesas,dmac-r8a7795" (R-Car H3)
 		- "renesas,dmac-r8a7796" (R-Car M3-W)
+		- "renesas,dmac-r8a77965" (R-Car M3-N)
 		- "renesas,dmac-r8a77970" (R-Car V3M)
 		- "renesas,dmac-r8a77980" (R-Car V3H)
 
-- 
cgit v1.1


From ef95a90ae6f4f21990e1f7ced6719784a409e811 Mon Sep 17 00:00:00 2001
From: Shamir Rabinovitch <shamir.rabinovitch@oracle.com>
Date: Tue, 10 Apr 2018 10:26:23 -0400
Subject: RDMA/ucma: ucma_context reference leak in error path

Validating input parameters should be done before getting the cm_id
otherwise it can leak a cm_id reference.

Fixes: 6a21dfc0d0db ("RDMA/ucma: Limit possible option size")
Signed-off-by: Shamir Rabinovitch <shamir.rabinovitch@oracle.com>
Reviewed-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
---
 drivers/infiniband/core/ucma.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/infiniband/core/ucma.c b/drivers/infiniband/core/ucma.c
index 7432948..680b353 100644
--- a/drivers/infiniband/core/ucma.c
+++ b/drivers/infiniband/core/ucma.c
@@ -1316,13 +1316,13 @@ static ssize_t ucma_set_option(struct ucma_file *file, const char __user *inbuf,
 	if (copy_from_user(&cmd, inbuf, sizeof(cmd)))
 		return -EFAULT;
 
+	if (unlikely(cmd.optlen > KMALLOC_MAX_SIZE))
+		return -EINVAL;
+
 	ctx = ucma_get_ctx(file, cmd.id);
 	if (IS_ERR(ctx))
 		return PTR_ERR(ctx);
 
-	if (unlikely(cmd.optlen > KMALLOC_MAX_SIZE))
-		return -EINVAL;
-
 	optval = memdup_user(u64_to_user_ptr(cmd.optval),
 			     cmd.optlen);
 	if (IS_ERR(optval)) {
-- 
cgit v1.1


From a21b85944c9653f2a3388d0da962c7c70690c30f Mon Sep 17 00:00:00 2001
From: Arnd Bergmann <arnd@arndb.de>
Date: Mon, 9 Apr 2018 12:54:34 +0200
Subject: clk: cs2000: mark resume function as __maybe_unused

When power management is disabled, we get a harmless warning:

drivers/clk/clk-cs2000-cp.c:544:12: error: 'cs2000_resume' defined but not used [-Werror=unused-function]

Marking the function as __maybe_unused lets the compiler silently
drop it instead.

Fixes: eade4ccdb087 ("clk: cs2000: set pm_ops in hibernate-compatible way")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
---
 drivers/clk/clk-cs2000-cp.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/clk/clk-cs2000-cp.c b/drivers/clk/clk-cs2000-cp.c
index c580197..a2f8c42 100644
--- a/drivers/clk/clk-cs2000-cp.c
+++ b/drivers/clk/clk-cs2000-cp.c
@@ -541,7 +541,7 @@ probe_err:
 	return ret;
 }
 
-static int cs2000_resume(struct device *dev)
+static int __maybe_unused cs2000_resume(struct device *dev)
 {
 	struct cs2000_priv *priv = dev_get_drvdata(dev);
 
-- 
cgit v1.1


From 4ad69b80e886a845f56ce0a3d10211208693d92b Mon Sep 17 00:00:00 2001
From: Jerome Brunet <jbrunet@baylibre.com>
Date: Mon, 9 Apr 2018 15:59:20 +0200
Subject: clk: honor CLK_MUX_ROUND_CLOSEST in generic clk mux

CLK_MUX_ROUND_CLOSEST is part of the clk_mux documentation but clk_mux
directly calls __clk_mux_determine_rate(), which overrides the flag.
As result, if clk_mux is instantiated with CLK_MUX_ROUND_CLOSEST, the
flag will be ignored and the clock rounded down.

To solve this, this patch expose clk_mux_determine_rate_flags() in the
clk-provider API and uses it in the determine_rate() callback of clk_mux.

Fixes: 15a02c1f6dd7 ("clk: Add __clk_mux_determine_rate_closest")
Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
---
 drivers/clk/clk-mux.c        | 10 +++++++++-
 drivers/clk/clk.c            |  7 ++++---
 include/linux/clk-provider.h |  3 +++
 3 files changed, 16 insertions(+), 4 deletions(-)

diff --git a/drivers/clk/clk-mux.c b/drivers/clk/clk-mux.c
index ac4a042..1628b93 100644
--- a/drivers/clk/clk-mux.c
+++ b/drivers/clk/clk-mux.c
@@ -112,10 +112,18 @@ static int clk_mux_set_parent(struct clk_hw *hw, u8 index)
 	return 0;
 }
 
+static int clk_mux_determine_rate(struct clk_hw *hw,
+				  struct clk_rate_request *req)
+{
+	struct clk_mux *mux = to_clk_mux(hw);
+
+	return clk_mux_determine_rate_flags(hw, req, mux->flags);
+}
+
 const struct clk_ops clk_mux_ops = {
 	.get_parent = clk_mux_get_parent,
 	.set_parent = clk_mux_set_parent,
-	.determine_rate = __clk_mux_determine_rate,
+	.determine_rate = clk_mux_determine_rate,
 };
 EXPORT_SYMBOL_GPL(clk_mux_ops);
 
diff --git a/drivers/clk/clk.c b/drivers/clk/clk.c
index ea67ac8..7af555f 100644
--- a/drivers/clk/clk.c
+++ b/drivers/clk/clk.c
@@ -426,9 +426,9 @@ static bool mux_is_better_rate(unsigned long rate, unsigned long now,
 	return now <= rate && now > best;
 }
 
-static int
-clk_mux_determine_rate_flags(struct clk_hw *hw, struct clk_rate_request *req,
-			     unsigned long flags)
+int clk_mux_determine_rate_flags(struct clk_hw *hw,
+				 struct clk_rate_request *req,
+				 unsigned long flags)
 {
 	struct clk_core *core = hw->core, *parent, *best_parent = NULL;
 	int i, num_parents, ret;
@@ -488,6 +488,7 @@ out:
 
 	return 0;
 }
+EXPORT_SYMBOL_GPL(clk_mux_determine_rate_flags);
 
 struct clk *__clk_lookup(const char *name)
 {
diff --git a/include/linux/clk-provider.h b/include/linux/clk-provider.h
index 210a890..1d25e14 100644
--- a/include/linux/clk-provider.h
+++ b/include/linux/clk-provider.h
@@ -765,6 +765,9 @@ int __clk_mux_determine_rate(struct clk_hw *hw,
 int __clk_determine_rate(struct clk_hw *core, struct clk_rate_request *req);
 int __clk_mux_determine_rate_closest(struct clk_hw *hw,
 				     struct clk_rate_request *req);
+int clk_mux_determine_rate_flags(struct clk_hw *hw,
+				 struct clk_rate_request *req,
+				 unsigned long flags);
 void clk_hw_reparent(struct clk_hw *hw, struct clk_hw *new_parent);
 void clk_hw_set_rate_range(struct clk_hw *hw, unsigned long min_rate,
 			   unsigned long max_rate);
-- 
cgit v1.1


From 6cc1eb507855612a983f9a66a9a13cdd48e16852 Mon Sep 17 00:00:00 2001
From: Jerome Brunet <jbrunet@baylibre.com>
Date: Mon, 9 Apr 2018 15:59:21 +0200
Subject: clk: meson: honor CLK_MUX_ROUND_CLOSEST in clk_regmap

Using __clk_mux_determine_rate effectively ignores CLK_MUX_ROUND_CLOSEST
if set the related clk_regmap mux instance.

Use clk_mux_determine_rate_flags() to make sure the flag is honored.

Fixes: ea11dda9e091 ("clk: meson: add regmap clocks")
Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
---
 drivers/clk/meson/clk-regmap.c | 11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)

diff --git a/drivers/clk/meson/clk-regmap.c b/drivers/clk/meson/clk-regmap.c
index 3645fdb6..ab7a355 100644
--- a/drivers/clk/meson/clk-regmap.c
+++ b/drivers/clk/meson/clk-regmap.c
@@ -153,10 +153,19 @@ static int clk_regmap_mux_set_parent(struct clk_hw *hw, u8 index)
 				  val << mux->shift);
 }
 
+static int clk_regmap_mux_determine_rate(struct clk_hw *hw,
+					 struct clk_rate_request *req)
+{
+	struct clk_regmap *clk = to_clk_regmap(hw);
+	struct clk_regmap_mux_data *mux = clk_get_regmap_mux_data(clk);
+
+	return clk_mux_determine_rate_flags(hw, req, mux->flags);
+}
+
 const struct clk_ops clk_regmap_mux_ops = {
 	.get_parent = clk_regmap_mux_get_parent,
 	.set_parent = clk_regmap_mux_set_parent,
-	.determine_rate = __clk_mux_determine_rate,
+	.determine_rate = clk_regmap_mux_determine_rate,
 };
 EXPORT_SYMBOL_GPL(clk_regmap_mux_ops);
 
-- 
cgit v1.1


From 25d0e2db3d8f4f79fddd436dcc848d912e98b485 Mon Sep 17 00:00:00 2001
From: Zhu Yanjun <yanjun.zhu@oracle.com>
Date: Sat, 14 Apr 2018 21:16:54 -0400
Subject: IB/mlx5: remove duplicate header file

The header file fs_helpers.h is included twice. So it should be removed.

Fixes: 802c2125689d ("IB/mlx5: Add IPsec support for egress and ingress")
CC: Srinivas Eeda <srinivas.eeda@oracle.com>
CC: Junxiao Bi <junxiao.bi@oracle.com>
Signed-off-by: Zhu Yanjun <yanjun.zhu@oracle.com>
Acked-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
---
 drivers/infiniband/hw/mlx5/main.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index daa919e..6a749c0 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -52,7 +52,6 @@
 #include <linux/mlx5/port.h>
 #include <linux/mlx5/vport.h>
 #include <linux/mlx5/fs.h>
-#include <linux/mlx5/fs_helpers.h>
 #include <linux/list.h>
 #include <rdma/ib_smi.h>
 #include <rdma/ib_umem.h>
-- 
cgit v1.1


From 39b4cbadb9a95bf3f13ea102d6ec841940916ee2 Mon Sep 17 00:00:00 2001
From: Changbin Du <changbin.du@intel.com>
Date: Fri, 30 Mar 2018 15:35:19 +0800
Subject: drm/i915/kvmgt: Check the pfn got from vfio_pin_pages

This can fix below oops. The target pfn must be mem backed.

[ 3639.109674] BUG: unable to handle kernel paging request at ffff8c44832a3000
[ 3639.109681] IP: memcpy_erms+0x6/0x10
[ 3639.109682] PGD 0 P4D 0
[ 3639.109685] Oops: 0000 1 SMP PTI
[ 3639.109726] CPU: 2 PID: 1724 Comm: qemu-system-x86 Not tainted 4.16.0-rc5+ #1
[ 3639.109727] Hardware name: /NUC7i7BNB, BIOS BNKBL357.86A.0050.2017.0816.2002 08/16/2017
[ 3639.109729] RIP: 0010:memcpy_erms+0x6/0x10
[ 3639.109730] RSP: 0018:ffffb1b7c3fbbbf0 EFLAGS: 00010246
[ 3639.109731] RAX: ffff8a44b6460000 RBX: 0000000036460000 RCX: 0000000000001000
[ 3639.109732] RDX: 0000000000001000 RSI: ffff8c44832a3000 RDI: ffff8a44b6460000
[ 3639.109733] RBP: 000000000006c8c0 R08: ffff8a44b6460000 R09: 0000000000000000
[ 3639.109734] R10: ffffb1b7c3fbbcd0 R11: ffff8a4d102018c0 R12: 0000000000000000
[ 3639.109734] R13: 0000000000000002 R14: 0000000000200000 R15: 0000000000000000
[ 3639.109736] FS: 00007f37f6d09700(0000) GS:ffff8a4d36d00000(0000) knlGS:0000000000000000
[ 3639.109737] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 3639.109738] CR2: ffff8c44832a3000 CR3: 000000088b7b8004 CR4: 00000000003626e0
[ 3639.109739] Call Trace:
[ 3639.109743] swiotlb_tbl_map_single+0x2bb/0x300
[ 3639.109746] map_single+0x30/0x80
[ 3639.109748] swiotlb_map_page+0x87/0x150
[ 3639.109751] kvmgt_dma_map_guest_page+0x329/0x3a0 [kvmgt]
[ 3639.109764] ? kvm_write_guest_offset_cached+0x84/0xe0 [kvm]
[ 3639.109789] intel_vgpu_emulate_ggtt_mmio_write+0x1f4/0x250 [i915]
[ 3639.109808] intel_vgpu_emulate_mmio_write+0x162/0x230 [i915]
[ 3639.109811] intel_vgpu_rw+0x1fc/0x240 [kvmgt]
[ 3639.109813] intel_vgpu_write+0x164/0x1f0 [kvmgt]
[ 3639.109816] __vfs_write+0x33/0x170
[ 3639.109818] ? do_vfs_ioctl+0x9f/0x5f0
[ 3639.109820] vfs_write+0xb3/0x1a0
[ 3639.109822] SyS_pwrite64+0x90/0xb0
[ 3639.109825] do_syscall_64+0x68/0x120
[ 3639.109827] entry_SYSCALL_64_after_hwframe+0x3d/0xa2
[ 3639.109829] RIP: 0033:0x7f3802b2d873
[ 3639.109830] RSP: 002b:00007f37f6d08670 EFLAGS: 00000293 ORIG_RAX: 0000000000000012
[ 3639.109831] RAX: ffffffffffffffda RBX: 0000000000000001 RCX: 00007f3802b2d873
[ 3639.109832] RDX: 0000000000000008 RSI: 00007f37f6d086a0 RDI: 000000000000001a
[ 3639.109833] RBP: 00007f37f6d086c0 R08: 0000000000000008 R09: ffffffffffffffff
[ 3639.109834] R10: 00000000008041c8 R11: 0000000000000293 R12: 00007ffd8bbf92ae
[ 3639.109835] R13: 00007ffd8bbf92af R14: 00007f37f6d09700 R15: 00007f37f6d099c0

v2: add Fixes tag.

Signed-off-by: Changbin Du <changbin.du@intel.com>
Fixes: cf4ee73 ("drm/i915/gvt: Fix guest vGPU hang caused by very high dma setup overhead")
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
---
 drivers/gpu/drm/i915/gvt/kvmgt.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/gpu/drm/i915/gvt/kvmgt.c b/drivers/gpu/drm/i915/gvt/kvmgt.c
index a2a59ff..29ec05b 100644
--- a/drivers/gpu/drm/i915/gvt/kvmgt.c
+++ b/drivers/gpu/drm/i915/gvt/kvmgt.c
@@ -123,6 +123,12 @@ static int gvt_dma_map_page(struct intel_vgpu *vgpu, unsigned long gfn,
 		return -EINVAL;
 	}
 
+	if (!pfn_valid(pfn)) {
+		gvt_vgpu_err("pfn 0x%lx is not mem backed\n", pfn);
+		vfio_unpin_pages(mdev_dev(vgpu->vdev.mdev), &gfn, 1);
+		return -EINVAL;
+	}
+
 	/* Setup DMA mapping. */
 	page = pfn_to_page(pfn);
 	*dma_addr = dma_map_page(dev, page, 0, PAGE_SIZE,
-- 
cgit v1.1


From 94f8dfc6cdfc3c48c3aea59ce528fa93cb54a69f Mon Sep 17 00:00:00 2001
From: Oscar Mateo <oscar.mateo@intel.com>
Date: Mon, 16 Apr 2018 14:57:01 -0700
Subject: drm/i915/selftests: Handle a potential failure of intel_ring_begin

Silence smatch over:

drivers/gpu/drm/i915/selftests/intel_workarounds.c:58 read_nonprivs() error: 'cs' dereferencing possible ERR_PTR()

by handling a potential (but unlikely) failure of intel_ring_begin.

Fixes: f4ecfbfc32ed ("drm/i915: Check whitelist registers across resets")
Signed-off-by: Oscar Mateo <oscar.mateo@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/1523915821-30624-1-git-send-email-oscar.mateo@intel.com
---
 drivers/gpu/drm/i915/selftests/intel_workarounds.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/drivers/gpu/drm/i915/selftests/intel_workarounds.c b/drivers/gpu/drm/i915/selftests/intel_workarounds.c
index fe7deca..5455b26 100644
--- a/drivers/gpu/drm/i915/selftests/intel_workarounds.c
+++ b/drivers/gpu/drm/i915/selftests/intel_workarounds.c
@@ -54,6 +54,11 @@ read_nonprivs(struct i915_gem_context *ctx, struct intel_engine_cs *engine)
 		srm++;
 
 	cs = intel_ring_begin(rq, 4 * RING_MAX_NONPRIV_SLOTS);
+	if (IS_ERR(cs)) {
+		err = PTR_ERR(cs);
+		goto err_req;
+	}
+
 	for (i = 0; i < RING_MAX_NONPRIV_SLOTS; i++) {
 		*cs++ = srm;
 		*cs++ = i915_mmio_reg_offset(RING_FORCE_TO_NONPRIV(base, i));
@@ -75,6 +80,8 @@ read_nonprivs(struct i915_gem_context *ctx, struct intel_engine_cs *engine)
 
 	return result;
 
+err_req:
+	i915_request_add(rq);
 err_pin:
 	i915_vma_unpin(vma);
 err_obj:
-- 
cgit v1.1


From b42db0860e13067fcc7cbfba3966c9e652668bbc Mon Sep 17 00:00:00 2001
From: Eric Sandeen <sandeen@sandeen.net>
Date: Mon, 16 Apr 2018 23:06:53 -0700
Subject: xfs: enhance dinode verifier

Add several more validations to xfs_dinode_verify:

- For LOCAL data fork formats, di_nextents must be 0.
- For LOCAL attr fork formats, di_anextents must be 0.
- For inodes with no attr fork offset,
  - format must be XFS_DINODE_FMT_EXTENTS if set at all
  - di_anextents must be 0.

Thanks to dchinner for pointing out a couple related checks I had
forgotten to add.

Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=199377
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/libxfs/xfs_inode_buf.c | 21 +++++++++++++++++++++
 1 file changed, 21 insertions(+)

diff --git a/fs/xfs/libxfs/xfs_inode_buf.c b/fs/xfs/libxfs/xfs_inode_buf.c
index ef68b1d..1201107 100644
--- a/fs/xfs/libxfs/xfs_inode_buf.c
+++ b/fs/xfs/libxfs/xfs_inode_buf.c
@@ -466,6 +466,8 @@ xfs_dinode_verify(
 				return __this_address;
 			if (di_size > XFS_DFORK_DSIZE(dip, mp))
 				return __this_address;
+			if (dip->di_nextents)
+				return __this_address;
 			/* fall through */
 		case XFS_DINODE_FMT_EXTENTS:
 		case XFS_DINODE_FMT_BTREE:
@@ -484,12 +486,31 @@ xfs_dinode_verify(
 	if (XFS_DFORK_Q(dip)) {
 		switch (dip->di_aformat) {
 		case XFS_DINODE_FMT_LOCAL:
+			if (dip->di_anextents)
+				return __this_address;
+		/* fall through */
 		case XFS_DINODE_FMT_EXTENTS:
 		case XFS_DINODE_FMT_BTREE:
 			break;
 		default:
 			return __this_address;
 		}
+	} else {
+		/*
+		 * If there is no fork offset, this may be a freshly-made inode
+		 * in a new disk cluster, in which case di_aformat is zeroed.
+		 * Otherwise, such an inode must be in EXTENTS format; this goes
+		 * for freed inodes as well.
+		 */
+		switch (dip->di_aformat) {
+		case 0:
+		case XFS_DINODE_FMT_EXTENTS:
+			break;
+		default:
+			return __this_address;
+		}
+		if (dip->di_anextents)
+			return __this_address;
 	}
 
 	/* only version 3 or greater inodes are extensively verified here */
-- 
cgit v1.1


From 2c4306f719b083d17df2963bc761777576b8ad1b Mon Sep 17 00:00:00 2001
From: Eric Sandeen <sandeen@redhat.com>
Date: Mon, 16 Apr 2018 23:07:27 -0700
Subject: xfs: set format back to extents if xfs_bmap_extents_to_btree

If xfs_bmap_extents_to_btree fails in a mode where we call
xfs_iroot_realloc(-1) to de-allocate the root, set the
format back to extents.

Otherwise we can assume we can dereference ifp->if_broot
based on the XFS_DINODE_FMT_BTREE format, and crash.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=199423
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/libxfs/xfs_bmap.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/fs/xfs/libxfs/xfs_bmap.c b/fs/xfs/libxfs/xfs_bmap.c
index 6a7c2f0..040eeda 100644
--- a/fs/xfs/libxfs/xfs_bmap.c
+++ b/fs/xfs/libxfs/xfs_bmap.c
@@ -725,12 +725,16 @@ xfs_bmap_extents_to_btree(
 	*logflagsp = 0;
 	if ((error = xfs_alloc_vextent(&args))) {
 		xfs_iroot_realloc(ip, -1, whichfork);
+		ASSERT(ifp->if_broot == NULL);
+		XFS_IFORK_FMT_SET(ip, whichfork, XFS_DINODE_FMT_EXTENTS);
 		xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
 		return error;
 	}
 
 	if (WARN_ON_ONCE(args.fsbno == NULLFSBLOCK)) {
 		xfs_iroot_realloc(ip, -1, whichfork);
+		ASSERT(ifp->if_broot == NULL);
+		XFS_IFORK_FMT_SET(ip, whichfork, XFS_DINODE_FMT_EXTENTS);
 		xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
 		return -ENOSPC;
 	}
-- 
cgit v1.1


From 7d83fb14258b9961920cd86f0b921caaeb3ebe85 Mon Sep 17 00:00:00 2001
From: "Darrick J. Wong" <darrick.wong@oracle.com>
Date: Mon, 16 Apr 2018 23:07:45 -0700
Subject: xfs: prevent creating negative-sized file via INSERT_RANGE

During the "insert range" fallocate operation, i_size grows by the
specified 'len' bytes.  XFS verifies that i_size + len < s_maxbytes, as
it should.  But this comparison is done using the signed 'loff_t', and
'i_size + len' can wrap around to a negative value, causing the check to
incorrectly pass, resulting in an inode with "negative" i_size.  This is
possible on 64-bit platforms, where XFS sets s_maxbytes = LLONG_MAX.
ext4 and f2fs don't run into this because they set a smaller s_maxbytes.

Fix it by using subtraction instead.

Reproducer:
    xfs_io -f file -c "truncate $(((1<<63)-1))" -c "finsert 0 4096"

Fixes: a904b1ca5751 ("xfs: Add support FALLOC_FL_INSERT_RANGE for fallocate")
Cc: <stable@vger.kernel.org> # v4.1+
Originally-From: Eric Biggers <ebiggers@google.com>
Signed-off-by: Eric Biggers <ebiggers@google.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
[darrick: fix signed integer addition overflow too]
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/xfs_file.c | 14 +++++++++-----
 1 file changed, 9 insertions(+), 5 deletions(-)

diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c
index 299aee4..eed073c 100644
--- a/fs/xfs/xfs_file.c
+++ b/fs/xfs/xfs_file.c
@@ -778,22 +778,26 @@ xfs_file_fallocate(
 		if (error)
 			goto out_unlock;
 	} else if (mode & FALLOC_FL_INSERT_RANGE) {
-		unsigned int blksize_mask = i_blocksize(inode) - 1;
+		unsigned int	blksize_mask = i_blocksize(inode) - 1;
+		loff_t		isize = i_size_read(inode);
 
-		new_size = i_size_read(inode) + len;
 		if (offset & blksize_mask || len & blksize_mask) {
 			error = -EINVAL;
 			goto out_unlock;
 		}
 
-		/* check the new inode size does not wrap through zero */
-		if (new_size > inode->i_sb->s_maxbytes) {
+		/*
+		 * New inode size must not exceed ->s_maxbytes, accounting for
+		 * possible signed overflow.
+		 */
+		if (inode->i_sb->s_maxbytes - isize < len) {
 			error = -EFBIG;
 			goto out_unlock;
 		}
+		new_size = isize + len;
 
 		/* Offset should be less than i_size */
-		if (offset >= i_size_read(inode)) {
+		if (offset >= isize) {
 			error = -EINVAL;
 			goto out_unlock;
 		}
-- 
cgit v1.1


From b3fe6c62bc66868c45b5bb16050e6bcb333af337 Mon Sep 17 00:00:00 2001
From: Randy Dunlap <rdunlap@infradead.org>
Date: Mon, 16 Apr 2018 18:51:50 -0700
Subject: infiniband: mlx5: fix build errors when INFINIBAND_USER_ACCESS=m

Fix build errors when INFINIBAND_USER_ACCESS=m and MLX5_INFINIBAND=y.
The build error occurs when the mlx5 driver code attempts to use
USER_ACCESS interfaces, which are built as a loadable module.

Fixes these build errors:

drivers/infiniband/hw/mlx5/main.o: In function `populate_specs_root':
../drivers/infiniband/hw/mlx5/main.c:4982: undefined reference to `uverbs_default_get_objects'
../drivers/infiniband/hw/mlx5/main.c:4994: undefined reference to `uverbs_alloc_spec_tree'
drivers/infiniband/hw/mlx5/main.o: In function `depopulate_specs_root':
../drivers/infiniband/hw/mlx5/main.c:5001: undefined reference to `uverbs_free_spec_tree'

Build-tested with multiple config combinations.

Fixes: 8c84660bb437 ("IB/mlx5: Initialize the parsing tree root without the help of uverbs")
Cc: stable@vger.kernel.org # reported against 4.16
Reported-by: kbuild test robot <lkp@intel.com>
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
---
 drivers/infiniband/hw/mlx5/Kconfig | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/infiniband/hw/mlx5/Kconfig b/drivers/infiniband/hw/mlx5/Kconfig
index bce263b..fb4d77b 100644
--- a/drivers/infiniband/hw/mlx5/Kconfig
+++ b/drivers/infiniband/hw/mlx5/Kconfig
@@ -1,6 +1,7 @@
 config MLX5_INFINIBAND
 	tristate "Mellanox Connect-IB HCA support"
 	depends on NETDEVICES && ETHERNET && PCI && MLX5_CORE
+	depends on INFINIBAND_USER_ACCESS || INFINIBAND_USER_ACCESS=n
 	---help---
 	  This driver provides low-level InfiniBand support for
 	  Mellanox Connect-IB PCI Express host channel adapters (HCAs).
-- 
cgit v1.1


From 7b38460dc8e4eafba06c78f8e37099d3b34d473c Mon Sep 17 00:00:00 2001
From: "Darrick J. Wong" <darrick.wong@oracle.com>
Date: Tue, 17 Apr 2018 19:10:15 -0700
Subject: xfs: don't fail when converting shortform attr to long form during
 ATTR_REPLACE

Kanda Motohiro reported that expanding a tiny xattr into a large xattr
fails on XFS because we remove the tiny xattr from a shortform fork and
then try to re-add it after converting the fork to extents format having
not removed the ATTR_REPLACE flag.  This fails because the attr is no
longer present, causing a fs shutdown.

This is derived from the patch in his bug report, but we really
shouldn't ignore a nonzero retval from the remove call.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=199119
Reported-by: kanda.motohiro@gmail.com
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/xfs/libxfs/xfs_attr.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/fs/xfs/libxfs/xfs_attr.c b/fs/xfs/libxfs/xfs_attr.c
index ce4a34a..35a1244 100644
--- a/fs/xfs/libxfs/xfs_attr.c
+++ b/fs/xfs/libxfs/xfs_attr.c
@@ -511,7 +511,14 @@ xfs_attr_shortform_addname(xfs_da_args_t *args)
 		if (args->flags & ATTR_CREATE)
 			return retval;
 		retval = xfs_attr_shortform_remove(args);
-		ASSERT(retval == 0);
+		if (retval)
+			return retval;
+		/*
+		 * Since we have removed the old attr, clear ATTR_REPLACE so
+		 * that the leaf format add routine won't trip over the attr
+		 * not being around.
+		 */
+		args->flags &= ~ATTR_REPLACE;
 	}
 
 	if (args->namelen >= XFS_ATTR_SF_ENTSIZE_MAX ||
-- 
cgit v1.1


From e33514f2e930ad800fa52db19e889bb0fba25419 Mon Sep 17 00:00:00 2001
From: Geert Uytterhoeven <geert@linux-m68k.org>
Date: Mon, 9 Apr 2018 16:52:47 +0200
Subject: IB/uverbs: Add missing braces in anonymous union initializers
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

With gcc-4.1.2:

    drivers/infiniband/core/uverbs_std_types_flow_action.c:366: error: unknown field ‘ptr’ specified in initializer
    drivers/infiniband/core/uverbs_std_types_flow_action.c:367: error: unknown field ‘type’ specified in initializer
    drivers/infiniband/core/uverbs_std_types_flow_action.c:367: warning: missing braces around initializer
    drivers/infiniband/core/uverbs_std_types_flow_action.c:367: warning: (near initialization for ‘uverbs_flow_action_esp_keymat[0].<anonymous>.<anonymous>’)
    drivers/infiniband/core/uverbs_std_types_flow_action.c:368: error: unknown field ‘min_len’ specified in initializer
    drivers/infiniband/core/uverbs_std_types_flow_action.c:368: warning: excess elements in union initializer
    drivers/infiniband/core/uverbs_std_types_flow_action.c:368: warning: (near initialization for ‘uverbs_flow_action_esp_keymat[0].<anonymous>’)
    drivers/infiniband/core/uverbs_std_types_flow_action.c:368: error: unknown field ‘len’ specified in initializer
    drivers/infiniband/core/uverbs_std_types_flow_action.c:368: warning: excess elements in union initializer
    drivers/infiniband/core/uverbs_std_types_flow_action.c:368: warning: (near initialization for ‘uverbs_flow_action_esp_keymat[0].<anonymous>’)
    drivers/infiniband/core/uverbs_std_types_flow_action.c:369: error: unknown field ‘flags’ specified in initializer
    drivers/infiniband/core/uverbs_std_types_flow_action.c:369: warning: excess elements in union initializer
    drivers/infiniband/core/uverbs_std_types_flow_action.c:369: warning: (near initialization for ‘uverbs_flow_action_esp_keymat[0].<anonymous>’)
    drivers/infiniband/core/uverbs_std_types_flow_action.c:376: error: unknown field ‘ptr’ specified in initializer
    drivers/infiniband/core/uverbs_std_types_flow_action.c:377: error: unknown field ‘type’ specified in initializer
    drivers/infiniband/core/uverbs_std_types_flow_action.c:377: warning: missing braces around initializer
    drivers/infiniband/core/uverbs_std_types_flow_action.c:377: warning: (near initialization for ‘uverbs_flow_action_esp_replay[0].<anonymous>.<anonymous>’)
    drivers/infiniband/core/uverbs_std_types_flow_action.c:379: error: unknown field ‘len’ specified in initializer
    drivers/infiniband/core/uverbs_std_types_flow_action.c:379: warning: excess elements in union initializer
    drivers/infiniband/core/uverbs_std_types_flow_action.c:379: warning: (near initialization for ‘uverbs_flow_action_esp_replay[0].<anonymous>’)
    drivers/infiniband/core/uverbs_std_types_flow_action.c:383: error: unknown field ‘ptr’ specified in initializer
    drivers/infiniband/core/uverbs_std_types_flow_action.c:384: error: unknown field ‘type’ specified in initializer
    drivers/infiniband/core/uverbs_std_types_flow_action.c:385: error: unknown field ‘min_len’ specified in initializer
    drivers/infiniband/core/uverbs_std_types_flow_action.c:385: warning: excess elements in union initializer
    drivers/infiniband/core/uverbs_std_types_flow_action.c:385: warning: (near initialization for ‘uverbs_flow_action_esp_replay[1].<anonymous>’)
    drivers/infiniband/core/uverbs_std_types_flow_action.c:385: error: unknown field ‘len’ specified in initializer
    drivers/infiniband/core/uverbs_std_types_flow_action.c:385: warning: excess elements in union initializer
    drivers/infiniband/core/uverbs_std_types_flow_action.c:385: warning: (near initialization for ‘uverbs_flow_action_esp_replay[1].<anonymous>’)
    drivers/infiniband/core/uverbs_std_types_flow_action.c:386: error: unknown field ‘flags’ specified in initializer
    drivers/infiniband/core/uverbs_std_types_flow_action.c:386: warning: excess elements in union initializer
    drivers/infiniband/core/uverbs_std_types_flow_action.c:386: warning: (near initialization for ‘uverbs_flow_action_esp_replay[1].<anonymous>’)

Add the missing braces to fix this.

Fixes: 2eb9beaee5d7 ("IB/uverbs: Add flow_action create and destroy verbs")
Fixes: 7d12f8d5a164 ("IB/uverbs: Add modify ESP flow_action")
Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
---
 drivers/infiniband/core/uverbs_std_types_flow_action.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/drivers/infiniband/core/uverbs_std_types_flow_action.c b/drivers/infiniband/core/uverbs_std_types_flow_action.c
index cbcec3da..b4f016d 100644
--- a/drivers/infiniband/core/uverbs_std_types_flow_action.c
+++ b/drivers/infiniband/core/uverbs_std_types_flow_action.c
@@ -363,28 +363,28 @@ static int UVERBS_HANDLER(UVERBS_METHOD_FLOW_ACTION_ESP_MODIFY)(struct ib_device
 
 static const struct uverbs_attr_spec uverbs_flow_action_esp_keymat[] = {
 	[IB_UVERBS_FLOW_ACTION_ESP_KEYMAT_AES_GCM] = {
-		.ptr = {
+		{ .ptr = {
 			.type = UVERBS_ATTR_TYPE_PTR_IN,
 			UVERBS_ATTR_TYPE(struct ib_uverbs_flow_action_esp_keymat_aes_gcm),
 			.flags = UVERBS_ATTR_SPEC_F_MIN_SZ_OR_ZERO,
-		},
+		} },
 	},
 };
 
 static const struct uverbs_attr_spec uverbs_flow_action_esp_replay[] = {
 	[IB_UVERBS_FLOW_ACTION_ESP_REPLAY_NONE] = {
-		.ptr = {
+		{ .ptr = {
 			.type = UVERBS_ATTR_TYPE_PTR_IN,
 			/* No need to specify any data */
 			.len = 0,
-		}
+		} }
 	},
 	[IB_UVERBS_FLOW_ACTION_ESP_REPLAY_BMP] = {
-		.ptr = {
+		{ .ptr = {
 			.type = UVERBS_ATTR_TYPE_PTR_IN,
 			UVERBS_ATTR_STRUCT(struct ib_uverbs_flow_action_esp_replay_bmp, size),
 			.flags = UVERBS_ATTR_SPEC_F_MIN_SZ_OR_ZERO,
-		}
+		} }
 	},
 };
 
-- 
cgit v1.1


From 8221229046e862977ae93ec9d34aa583fbd10397 Mon Sep 17 00:00:00 2001
From: Gaurav K Singh <gaurav.k.singh@intel.com>
Date: Tue, 17 Apr 2018 23:52:18 +0530
Subject: drm/i915/audio: Fix audio detection issue on GLK

On Geminilake, sometimes audio card is not getting
detected after reboot. This is a spurious issue happening on
Geminilake. HW codec and HD audio controller link was going
out of sync for which there was a fix in i915 driver but
was not getting invoked for GLK. Extending this fix to GLK as well.

Tested by Du,Wenkai on GLK board.

Bspec: 21829

v2: Instead of checking GEN9_BC, BXT and GLK macros, use IS_GEN9 macro (Jani N)

Cc: <stable@vger.kernel.org> # b651bd2a3ae3 ("drm/i915/audio: Fix audio enumeration issue on BXT")
Cc: <stable@vger.kernel.org>
Signed-off-by: Gaurav K Singh <gaurav.k.singh@intel.com>
Reviewed-by: Abhay Kumar <abhay.Kumar@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1523989338-29677-1-git-send-email-gaurav.k.singh@intel.com
---
 drivers/gpu/drm/i915/intel_audio.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/intel_audio.c b/drivers/gpu/drm/i915/intel_audio.c
index 656f6c9..3ea566f 100644
--- a/drivers/gpu/drm/i915/intel_audio.c
+++ b/drivers/gpu/drm/i915/intel_audio.c
@@ -729,7 +729,7 @@ static void i915_audio_component_codec_wake_override(struct device *kdev,
 	struct drm_i915_private *dev_priv = kdev_to_i915(kdev);
 	u32 tmp;
 
-	if (!IS_GEN9_BC(dev_priv) && !IS_BROXTON(dev_priv))
+	if (!IS_GEN9(dev_priv))
 		return;
 
 	i915_audio_component_get_power(kdev);
-- 
cgit v1.1


From 89e050c87dab3e933dcb90677947752d19aeafa4 Mon Sep 17 00:00:00 2001
From: Helge Deller <deller@gmx.de>
Date: Wed, 18 Apr 2018 07:49:19 +0200
Subject: parisc: Make bzImage default build target

Debian uses "make all" to build the Linux kernel, thus to be able to use
the self-decompressing kernel as default debian kernel we need to make
bzImage the default build target.

Signed-off-by: Helge Deller <deller@gmx.de>
---
 arch/parisc/Makefile | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/arch/parisc/Makefile b/arch/parisc/Makefile
index e2364ff..34ac503 100644
--- a/arch/parisc/Makefile
+++ b/arch/parisc/Makefile
@@ -123,6 +123,9 @@ INSTALL_TARGETS = zinstall install
 
 PHONY += bzImage $(BOOT_TARGETS) $(INSTALL_TARGETS)
 
+# Default kernel to build
+all: bzImage
+
 zImage: vmlinuz
 Image: vmlinux
 
-- 
cgit v1.1


From 41dbee81c875300072f6154d6722f6e8e1353cca Mon Sep 17 00:00:00 2001
From: Helge Deller <deller@gmx.de>
Date: Wed, 18 Apr 2018 15:46:41 +0200
Subject: parisc: Document rules regarding checksum of HPMC handler

Signed-off-by: Helge Deller <deller@gmx.de>
---
 arch/parisc/kernel/traps.c | 11 +++++++++++
 1 file changed, 11 insertions(+)

diff --git a/arch/parisc/kernel/traps.c b/arch/parisc/kernel/traps.c
index 68e671a..71d3127 100644
--- a/arch/parisc/kernel/traps.c
+++ b/arch/parisc/kernel/traps.c
@@ -837,6 +837,17 @@ void __init initialize_ivt(const void *iva)
 	if (pdc_instr(&instr) == PDC_OK)
 		ivap[0] = instr;
 
+	/*
+	 * Rules for the checksum of the HPMC handler:
+	 * 1. The IVA does not point to PDC/PDH space (ie: the OS has installed
+	 *    its own IVA).
+	 * 2. The word at IVA + 32 is nonzero.
+	 * 3. If Length (IVA + 60) is not zero, then Length (IVA + 60) and
+	 *    Address (IVA + 56) are word-aligned.
+	 * 4. The checksum of the 8 words starting at IVA + 32 plus the sum of
+	 *    the Length/4 words starting at Address is zero.
+	 */
+
 	/* Compute Checksum for HPMC handler */
 	length = os_hpmc_size;
 	ivap[7] = length;
-- 
cgit v1.1


From 98ff5c78307b4177b7e44783a04b208189e21418 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Wed, 18 Apr 2018 19:40:50 +0100
Subject: drm/i915: Move the priotree struct to its own headers

Over time the priotree has grown from a sorted list to a more
complicated structure for propagating constraints along the dependency
chain to try and resolve priority inversion. Start to segregate this
information from the rest of the request/fence tracking.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180418184052.7129-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/i915_request.h   | 39 +-----------------------
 drivers/gpu/drm/i915/i915_scheduler.h | 57 +++++++++++++++++++++++++++++++++++
 2 files changed, 58 insertions(+), 38 deletions(-)
 create mode 100644 drivers/gpu/drm/i915/i915_scheduler.h

diff --git a/drivers/gpu/drm/i915/i915_request.h b/drivers/gpu/drm/i915/i915_request.h
index 7d6eb82..e6f7c5f 100644
--- a/drivers/gpu/drm/i915/i915_request.h
+++ b/drivers/gpu/drm/i915/i915_request.h
@@ -28,6 +28,7 @@
 #include <linux/dma-fence.h>
 
 #include "i915_gem.h"
+#include "i915_scheduler.h"
 #include "i915_sw_fence.h"
 
 #include <uapi/drm/i915_drm.h>
@@ -48,44 +49,6 @@ struct intel_signal_node {
 	struct list_head link;
 };
 
-struct i915_dependency {
-	struct i915_priotree *signaler;
-	struct list_head signal_link;
-	struct list_head wait_link;
-	struct list_head dfs_link;
-	unsigned long flags;
-#define I915_DEPENDENCY_ALLOC BIT(0)
-};
-
-/*
- * "People assume that time is a strict progression of cause to effect, but
- * actually, from a nonlinear, non-subjective viewpoint, it's more like a big
- * ball of wibbly-wobbly, timey-wimey ... stuff." -The Doctor, 2015
- *
- * Requests exist in a complex web of interdependencies. Each request
- * has to wait for some other request to complete before it is ready to be run
- * (e.g. we have to wait until the pixels have been rendering into a texture
- * before we can copy from it). We track the readiness of a request in terms
- * of fences, but we also need to keep the dependency tree for the lifetime
- * of the request (beyond the life of an individual fence). We use the tree
- * at various points to reorder the requests whilst keeping the requests
- * in order with respect to their various dependencies.
- */
-struct i915_priotree {
-	struct list_head signalers_list; /* those before us, we depend upon */
-	struct list_head waiters_list; /* those after us, they depend upon us */
-	struct list_head link;
-	int priority;
-};
-
-enum {
-	I915_PRIORITY_MIN = I915_CONTEXT_MIN_USER_PRIORITY - 1,
-	I915_PRIORITY_NORMAL = I915_CONTEXT_DEFAULT_PRIORITY,
-	I915_PRIORITY_MAX = I915_CONTEXT_MAX_USER_PRIORITY + 1,
-
-	I915_PRIORITY_INVALID = INT_MIN
-};
-
 struct i915_capture_list {
 	struct i915_capture_list *next;
 	struct i915_vma *vma;
diff --git a/drivers/gpu/drm/i915/i915_scheduler.h b/drivers/gpu/drm/i915/i915_scheduler.h
new file mode 100644
index 0000000..9d6ea9f
--- /dev/null
+++ b/drivers/gpu/drm/i915/i915_scheduler.h
@@ -0,0 +1,57 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2018 Intel Corporation
+ */
+
+#ifndef _I915_SCHEDULER_H_
+#define _I915_SCHEDULER_H_
+
+#include <linux/bitops.h>
+
+#include <uapi/drm/i915_drm.h>
+
+enum {
+	I915_PRIORITY_MIN = I915_CONTEXT_MIN_USER_PRIORITY - 1,
+	I915_PRIORITY_NORMAL = I915_CONTEXT_DEFAULT_PRIORITY,
+	I915_PRIORITY_MAX = I915_CONTEXT_MAX_USER_PRIORITY + 1,
+
+	I915_PRIORITY_INVALID = INT_MIN
+};
+
+/*
+ * "People assume that time is a strict progression of cause to effect, but
+ * actually, from a nonlinear, non-subjective viewpoint, it's more like a big
+ * ball of wibbly-wobbly, timey-wimey ... stuff." -The Doctor, 2015
+ *
+ * Requests exist in a complex web of interdependencies. Each request
+ * has to wait for some other request to complete before it is ready to be run
+ * (e.g. we have to wait until the pixels have been rendering into a texture
+ * before we can copy from it). We track the readiness of a request in terms
+ * of fences, but we also need to keep the dependency tree for the lifetime
+ * of the request (beyond the life of an individual fence). We use the tree
+ * at various points to reorder the requests whilst keeping the requests
+ * in order with respect to their various dependencies.
+ *
+ * There is no active component to the "scheduler". As we know the dependency
+ * DAG of each request, we are able to insert it into a sorted queue when it
+ * is ready, and are able to reorder its portion of the graph to accommodate
+ * dynamic priority changes.
+ */
+struct i915_priotree {
+	struct list_head signalers_list; /* those before us, we depend upon */
+	struct list_head waiters_list; /* those after us, they depend upon us */
+	struct list_head link;
+	int priority;
+};
+
+struct i915_dependency {
+	struct i915_priotree *signaler;
+	struct list_head signal_link;
+	struct list_head wait_link;
+	struct list_head dfs_link;
+	unsigned long flags;
+#define I915_DEPENDENCY_ALLOC BIT(0)
+};
+
+#endif /* _I915_SCHEDULER_H_ */
-- 
cgit v1.1


From 0c7112a00272c633a79cad91ea9c1a0f40330f5d Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Wed, 18 Apr 2018 19:40:51 +0100
Subject: drm/i915: Rename priotree to sched

Having moved the priotree struct into i915_scheduler.h, identify it as
the scheduling element and rebrand into i915_sched. This becomes more
useful as we start attaching more information we require to propagate
through the scheduler.

v2: Use i915_sched_node for future distinctiveness

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180418184052.7129-2-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/i915_gpu_error.c       |  2 +-
 drivers/gpu/drm/i915/i915_request.c         | 66 +++++++++++++------------
 drivers/gpu/drm/i915/i915_request.h         |  6 +--
 drivers/gpu/drm/i915/i915_scheduler.h       |  4 +-
 drivers/gpu/drm/i915/intel_engine_cs.c      |  4 +-
 drivers/gpu/drm/i915/intel_guc_submission.c |  8 +--
 drivers/gpu/drm/i915/intel_lrc.c            | 77 +++++++++++++++--------------
 7 files changed, 85 insertions(+), 82 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c
index effaf98..6b5b9b3 100644
--- a/drivers/gpu/drm/i915/i915_gpu_error.c
+++ b/drivers/gpu/drm/i915/i915_gpu_error.c
@@ -1278,7 +1278,7 @@ static void record_request(struct i915_request *request,
 			   struct drm_i915_error_request *erq)
 {
 	erq->context = request->ctx->hw_id;
-	erq->priority = request->priotree.priority;
+	erq->priority = request->sched.priority;
 	erq->ban_score = atomic_read(&request->ctx->ban_score);
 	erq->seqno = request->global_seqno;
 	erq->jiffies = request->emitted_jiffies;
diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c
index 9ca9c24..dfcc6a0 100644
--- a/drivers/gpu/drm/i915/i915_request.c
+++ b/drivers/gpu/drm/i915/i915_request.c
@@ -125,22 +125,22 @@ i915_dependency_free(struct drm_i915_private *i915,
 }
 
 static void
-__i915_priotree_add_dependency(struct i915_priotree *pt,
-			       struct i915_priotree *signal,
-			       struct i915_dependency *dep,
-			       unsigned long flags)
+__i915_sched_node_add_dependency(struct i915_sched_node *node,
+				 struct i915_sched_node *signal,
+				 struct i915_dependency *dep,
+				 unsigned long flags)
 {
 	INIT_LIST_HEAD(&dep->dfs_link);
 	list_add(&dep->wait_link, &signal->waiters_list);
-	list_add(&dep->signal_link, &pt->signalers_list);
+	list_add(&dep->signal_link, &node->signalers_list);
 	dep->signaler = signal;
 	dep->flags = flags;
 }
 
 static int
-i915_priotree_add_dependency(struct drm_i915_private *i915,
-			     struct i915_priotree *pt,
-			     struct i915_priotree *signal)
+i915_sched_node_add_dependency(struct drm_i915_private *i915,
+			       struct i915_sched_node *node,
+			       struct i915_sched_node *signal)
 {
 	struct i915_dependency *dep;
 
@@ -148,16 +148,18 @@ i915_priotree_add_dependency(struct drm_i915_private *i915,
 	if (!dep)
 		return -ENOMEM;
 
-	__i915_priotree_add_dependency(pt, signal, dep, I915_DEPENDENCY_ALLOC);
+	__i915_sched_node_add_dependency(node, signal, dep,
+					 I915_DEPENDENCY_ALLOC);
 	return 0;
 }
 
 static void
-i915_priotree_fini(struct drm_i915_private *i915, struct i915_priotree *pt)
+i915_sched_node_fini(struct drm_i915_private *i915,
+		     struct i915_sched_node *node)
 {
-	struct i915_dependency *dep, *next;
+	struct i915_dependency *dep, *tmp;
 
-	GEM_BUG_ON(!list_empty(&pt->link));
+	GEM_BUG_ON(!list_empty(&node->link));
 
 	/*
 	 * Everyone we depended upon (the fences we wait to be signaled)
@@ -165,8 +167,8 @@ i915_priotree_fini(struct drm_i915_private *i915, struct i915_priotree *pt)
 	 * However, retirement is run independently on each timeline and
 	 * so we may be called out-of-order.
 	 */
-	list_for_each_entry_safe(dep, next, &pt->signalers_list, signal_link) {
-		GEM_BUG_ON(!i915_priotree_signaled(dep->signaler));
+	list_for_each_entry_safe(dep, tmp, &node->signalers_list, signal_link) {
+		GEM_BUG_ON(!i915_sched_node_signaled(dep->signaler));
 		GEM_BUG_ON(!list_empty(&dep->dfs_link));
 
 		list_del(&dep->wait_link);
@@ -175,8 +177,8 @@ i915_priotree_fini(struct drm_i915_private *i915, struct i915_priotree *pt)
 	}
 
 	/* Remove ourselves from everyone who depends upon us */
-	list_for_each_entry_safe(dep, next, &pt->waiters_list, wait_link) {
-		GEM_BUG_ON(dep->signaler != pt);
+	list_for_each_entry_safe(dep, tmp, &node->waiters_list, wait_link) {
+		GEM_BUG_ON(dep->signaler != node);
 		GEM_BUG_ON(!list_empty(&dep->dfs_link));
 
 		list_del(&dep->signal_link);
@@ -186,12 +188,12 @@ i915_priotree_fini(struct drm_i915_private *i915, struct i915_priotree *pt)
 }
 
 static void
-i915_priotree_init(struct i915_priotree *pt)
+i915_sched_node_init(struct i915_sched_node *node)
 {
-	INIT_LIST_HEAD(&pt->signalers_list);
-	INIT_LIST_HEAD(&pt->waiters_list);
-	INIT_LIST_HEAD(&pt->link);
-	pt->priority = I915_PRIORITY_INVALID;
+	INIT_LIST_HEAD(&node->signalers_list);
+	INIT_LIST_HEAD(&node->waiters_list);
+	INIT_LIST_HEAD(&node->link);
+	node->priority = I915_PRIORITY_INVALID;
 }
 
 static int reset_all_global_seqno(struct drm_i915_private *i915, u32 seqno)
@@ -422,7 +424,7 @@ static void i915_request_retire(struct i915_request *request)
 	}
 	spin_unlock_irq(&request->lock);
 
-	i915_priotree_fini(request->i915, &request->priotree);
+	i915_sched_node_fini(request->i915, &request->sched);
 	i915_request_put(request);
 }
 
@@ -725,7 +727,7 @@ i915_request_alloc(struct intel_engine_cs *engine, struct i915_gem_context *ctx)
 	i915_sw_fence_init(&i915_request_get(rq)->submit, submit_notify);
 	init_waitqueue_head(&rq->execute);
 
-	i915_priotree_init(&rq->priotree);
+	i915_sched_node_init(&rq->sched);
 
 	INIT_LIST_HEAD(&rq->active_list);
 	rq->i915 = i915;
@@ -777,8 +779,8 @@ err_unwind:
 
 	/* Make sure we didn't add ourselves to external state before freeing */
 	GEM_BUG_ON(!list_empty(&rq->active_list));
-	GEM_BUG_ON(!list_empty(&rq->priotree.signalers_list));
-	GEM_BUG_ON(!list_empty(&rq->priotree.waiters_list));
+	GEM_BUG_ON(!list_empty(&rq->sched.signalers_list));
+	GEM_BUG_ON(!list_empty(&rq->sched.waiters_list));
 
 	kmem_cache_free(i915->requests, rq);
 err_unreserve:
@@ -800,9 +802,9 @@ i915_request_await_request(struct i915_request *to, struct i915_request *from)
 		return 0;
 
 	if (to->engine->schedule) {
-		ret = i915_priotree_add_dependency(to->i915,
-						   &to->priotree,
-						   &from->priotree);
+		ret = i915_sched_node_add_dependency(to->i915,
+						     &to->sched,
+						     &from->sched);
 		if (ret < 0)
 			return ret;
 	}
@@ -1033,10 +1035,10 @@ void __i915_request_add(struct i915_request *request, bool flush_caches)
 		i915_sw_fence_await_sw_fence(&request->submit, &prev->submit,
 					     &request->submitq);
 		if (engine->schedule)
-			__i915_priotree_add_dependency(&request->priotree,
-						       &prev->priotree,
-						       &request->dep,
-						       0);
+			__i915_sched_node_add_dependency(&request->sched,
+							 &prev->sched,
+							 &request->dep,
+							 0);
 	}
 
 	spin_lock_irq(&timeline->lock);
diff --git a/drivers/gpu/drm/i915/i915_request.h b/drivers/gpu/drm/i915/i915_request.h
index e6f7c5f..35b8a98 100644
--- a/drivers/gpu/drm/i915/i915_request.h
+++ b/drivers/gpu/drm/i915/i915_request.h
@@ -117,7 +117,7 @@ struct i915_request {
 	 * to retirement), i.e. bidirectional dependency information for the
 	 * request not tied to individual fences.
 	 */
-	struct i915_priotree priotree;
+	struct i915_sched_node sched;
 	struct i915_dependency dep;
 
 	/**
@@ -306,10 +306,10 @@ static inline bool i915_request_started(const struct i915_request *rq)
 				 seqno - 1);
 }
 
-static inline bool i915_priotree_signaled(const struct i915_priotree *pt)
+static inline bool i915_sched_node_signaled(const struct i915_sched_node *node)
 {
 	const struct i915_request *rq =
-		container_of(pt, const struct i915_request, priotree);
+		container_of(node, const struct i915_request, sched);
 
 	return i915_request_completed(rq);
 }
diff --git a/drivers/gpu/drm/i915/i915_scheduler.h b/drivers/gpu/drm/i915/i915_scheduler.h
index 9d6ea9f..754243e 100644
--- a/drivers/gpu/drm/i915/i915_scheduler.h
+++ b/drivers/gpu/drm/i915/i915_scheduler.h
@@ -38,7 +38,7 @@ enum {
  * is ready, and are able to reorder its portion of the graph to accommodate
  * dynamic priority changes.
  */
-struct i915_priotree {
+struct i915_sched_node {
 	struct list_head signalers_list; /* those before us, we depend upon */
 	struct list_head waiters_list; /* those after us, they depend upon us */
 	struct list_head link;
@@ -46,7 +46,7 @@ struct i915_priotree {
 };
 
 struct i915_dependency {
-	struct i915_priotree *signaler;
+	struct i915_sched_node *signaler;
 	struct list_head signal_link;
 	struct list_head wait_link;
 	struct list_head dfs_link;
diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
index 1a83707..b542b1a 100644
--- a/drivers/gpu/drm/i915/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/intel_engine_cs.c
@@ -1123,7 +1123,7 @@ static void print_request(struct drm_printer *m,
 		   rq->global_seqno,
 		   i915_request_completed(rq) ? "!" : "",
 		   rq->fence.context, rq->fence.seqno,
-		   rq->priotree.priority,
+		   rq->sched.priority,
 		   jiffies_to_msecs(jiffies - rq->emitted_jiffies),
 		   name);
 }
@@ -1367,7 +1367,7 @@ void intel_engine_dump(struct intel_engine_cs *engine,
 		struct i915_priolist *p =
 			rb_entry(rb, typeof(*p), node);
 
-		list_for_each_entry(rq, &p->requests, priotree.link)
+		list_for_each_entry(rq, &p->requests, sched.link)
 			print_request(m, rq, "\t\tQ ");
 	}
 	spin_unlock_irq(&engine->timeline->lock);
diff --git a/drivers/gpu/drm/i915/intel_guc_submission.c b/drivers/gpu/drm/i915/intel_guc_submission.c
index 9712123..0755f5c 100644
--- a/drivers/gpu/drm/i915/intel_guc_submission.c
+++ b/drivers/gpu/drm/i915/intel_guc_submission.c
@@ -659,7 +659,7 @@ static void port_assign(struct execlist_port *port, struct i915_request *rq)
 
 static inline int rq_prio(const struct i915_request *rq)
 {
-	return rq->priotree.priority;
+	return rq->sched.priority;
 }
 
 static inline int port_prio(const struct execlist_port *port)
@@ -706,11 +706,11 @@ static void guc_dequeue(struct intel_engine_cs *engine)
 		struct i915_priolist *p = to_priolist(rb);
 		struct i915_request *rq, *rn;
 
-		list_for_each_entry_safe(rq, rn, &p->requests, priotree.link) {
+		list_for_each_entry_safe(rq, rn, &p->requests, sched.link) {
 			if (last && rq->ctx != last->ctx) {
 				if (port == last_port) {
 					__list_del_many(&p->requests,
-							&rq->priotree.link);
+							&rq->sched.link);
 					goto done;
 				}
 
@@ -719,7 +719,7 @@ static void guc_dequeue(struct intel_engine_cs *engine)
 				port++;
 			}
 
-			INIT_LIST_HEAD(&rq->priotree.link);
+			INIT_LIST_HEAD(&rq->sched.link);
 
 			__i915_request_submit(rq);
 			trace_i915_request_in(rq, port_index(port, execlists));
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 4f72858..062ed6e 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -177,7 +177,7 @@ static inline struct i915_priolist *to_priolist(struct rb_node *rb)
 
 static inline int rq_prio(const struct i915_request *rq)
 {
-	return rq->priotree.priority;
+	return rq->sched.priority;
 }
 
 static inline bool need_preempt(const struct intel_engine_cs *engine,
@@ -258,7 +258,7 @@ intel_lr_context_descriptor_update(struct i915_gem_context *ctx,
 
 static struct i915_priolist *
 lookup_priolist(struct intel_engine_cs *engine,
-		struct i915_priotree *pt,
+		struct i915_sched_node *node,
 		int prio)
 {
 	struct intel_engine_execlists * const execlists = &engine->execlists;
@@ -344,10 +344,10 @@ static void __unwind_incomplete_requests(struct intel_engine_cs *engine)
 		GEM_BUG_ON(rq_prio(rq) == I915_PRIORITY_INVALID);
 		if (rq_prio(rq) != last_prio) {
 			last_prio = rq_prio(rq);
-			p = lookup_priolist(engine, &rq->priotree, last_prio);
+			p = lookup_priolist(engine, &rq->sched, last_prio);
 		}
 
-		list_add(&rq->priotree.link, &p->requests);
+		list_add(&rq->sched.link, &p->requests);
 	}
 }
 
@@ -654,7 +654,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 		struct i915_priolist *p = to_priolist(rb);
 		struct i915_request *rq, *rn;
 
-		list_for_each_entry_safe(rq, rn, &p->requests, priotree.link) {
+		list_for_each_entry_safe(rq, rn, &p->requests, sched.link) {
 			/*
 			 * Can we combine this request with the current port?
 			 * It has to be the same context/ringbuffer and not
@@ -674,7 +674,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 				 */
 				if (port == last_port) {
 					__list_del_many(&p->requests,
-							&rq->priotree.link);
+							&rq->sched.link);
 					goto done;
 				}
 
@@ -688,7 +688,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 				if (ctx_single_port_submission(last->ctx) ||
 				    ctx_single_port_submission(rq->ctx)) {
 					__list_del_many(&p->requests,
-							&rq->priotree.link);
+							&rq->sched.link);
 					goto done;
 				}
 
@@ -701,7 +701,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 				GEM_BUG_ON(port_isset(port));
 			}
 
-			INIT_LIST_HEAD(&rq->priotree.link);
+			INIT_LIST_HEAD(&rq->sched.link);
 			__i915_request_submit(rq);
 			trace_i915_request_in(rq, port_index(port, execlists));
 			last = rq;
@@ -882,8 +882,8 @@ static void execlists_cancel_requests(struct intel_engine_cs *engine)
 	while (rb) {
 		struct i915_priolist *p = to_priolist(rb);
 
-		list_for_each_entry_safe(rq, rn, &p->requests, priotree.link) {
-			INIT_LIST_HEAD(&rq->priotree.link);
+		list_for_each_entry_safe(rq, rn, &p->requests, sched.link) {
+			INIT_LIST_HEAD(&rq->sched.link);
 
 			dma_fence_set_error(&rq->fence, -EIO);
 			__i915_request_submit(rq);
@@ -1116,10 +1116,11 @@ static void execlists_submission_tasklet(unsigned long data)
 }
 
 static void queue_request(struct intel_engine_cs *engine,
-			  struct i915_priotree *pt,
+			  struct i915_sched_node *node,
 			  int prio)
 {
-	list_add_tail(&pt->link, &lookup_priolist(engine, pt, prio)->requests);
+	list_add_tail(&node->link,
+		      &lookup_priolist(engine, node, prio)->requests);
 }
 
 static void __submit_queue(struct intel_engine_cs *engine, int prio)
@@ -1142,24 +1143,24 @@ static void execlists_submit_request(struct i915_request *request)
 	/* Will be called from irq-context when using foreign fences. */
 	spin_lock_irqsave(&engine->timeline->lock, flags);
 
-	queue_request(engine, &request->priotree, rq_prio(request));
+	queue_request(engine, &request->sched, rq_prio(request));
 	submit_queue(engine, rq_prio(request));
 
 	GEM_BUG_ON(!engine->execlists.first);
-	GEM_BUG_ON(list_empty(&request->priotree.link));
+	GEM_BUG_ON(list_empty(&request->sched.link));
 
 	spin_unlock_irqrestore(&engine->timeline->lock, flags);
 }
 
-static struct i915_request *pt_to_request(struct i915_priotree *pt)
+static struct i915_request *sched_to_request(struct i915_sched_node *node)
 {
-	return container_of(pt, struct i915_request, priotree);
+	return container_of(node, struct i915_request, sched);
 }
 
 static struct intel_engine_cs *
-pt_lock_engine(struct i915_priotree *pt, struct intel_engine_cs *locked)
+sched_lock_engine(struct i915_sched_node *node, struct intel_engine_cs *locked)
 {
-	struct intel_engine_cs *engine = pt_to_request(pt)->engine;
+	struct intel_engine_cs *engine = sched_to_request(node)->engine;
 
 	GEM_BUG_ON(!locked);
 
@@ -1183,23 +1184,23 @@ static void execlists_schedule(struct i915_request *request, int prio)
 	if (i915_request_completed(request))
 		return;
 
-	if (prio <= READ_ONCE(request->priotree.priority))
+	if (prio <= READ_ONCE(request->sched.priority))
 		return;
 
 	/* Need BKL in order to use the temporary link inside i915_dependency */
 	lockdep_assert_held(&request->i915->drm.struct_mutex);
 
-	stack.signaler = &request->priotree;
+	stack.signaler = &request->sched;
 	list_add(&stack.dfs_link, &dfs);
 
 	/*
 	 * Recursively bump all dependent priorities to match the new request.
 	 *
 	 * A naive approach would be to use recursion:
-	 * static void update_priorities(struct i915_priotree *pt, prio) {
-	 *	list_for_each_entry(dep, &pt->signalers_list, signal_link)
+	 * static void update_priorities(struct i915_sched_node *node, prio) {
+	 *	list_for_each_entry(dep, &node->signalers_list, signal_link)
 	 *		update_priorities(dep->signal, prio)
-	 *	queue_request(pt);
+	 *	queue_request(node);
 	 * }
 	 * but that may have unlimited recursion depth and so runs a very
 	 * real risk of overunning the kernel stack. Instead, we build
@@ -1211,7 +1212,7 @@ static void execlists_schedule(struct i915_request *request, int prio)
 	 * last element in the list is the request we must execute first.
 	 */
 	list_for_each_entry(dep, &dfs, dfs_link) {
-		struct i915_priotree *pt = dep->signaler;
+		struct i915_sched_node *node = dep->signaler;
 
 		/*
 		 * Within an engine, there can be no cycle, but we may
@@ -1219,13 +1220,13 @@ static void execlists_schedule(struct i915_request *request, int prio)
 		 * (redundant dependencies are not eliminated) and across
 		 * engines.
 		 */
-		list_for_each_entry(p, &pt->signalers_list, signal_link) {
+		list_for_each_entry(p, &node->signalers_list, signal_link) {
 			GEM_BUG_ON(p == dep); /* no cycles! */
 
-			if (i915_priotree_signaled(p->signaler))
+			if (i915_sched_node_signaled(p->signaler))
 				continue;
 
-			GEM_BUG_ON(p->signaler->priority < pt->priority);
+			GEM_BUG_ON(p->signaler->priority < node->priority);
 			if (prio > READ_ONCE(p->signaler->priority))
 				list_move_tail(&p->dfs_link, &dfs);
 		}
@@ -1237,9 +1238,9 @@ static void execlists_schedule(struct i915_request *request, int prio)
 	 * execlists_submit_request()), we can set our own priority and skip
 	 * acquiring the engine locks.
 	 */
-	if (request->priotree.priority == I915_PRIORITY_INVALID) {
-		GEM_BUG_ON(!list_empty(&request->priotree.link));
-		request->priotree.priority = prio;
+	if (request->sched.priority == I915_PRIORITY_INVALID) {
+		GEM_BUG_ON(!list_empty(&request->sched.link));
+		request->sched.priority = prio;
 		if (stack.dfs_link.next == stack.dfs_link.prev)
 			return;
 		__list_del_entry(&stack.dfs_link);
@@ -1250,23 +1251,23 @@ static void execlists_schedule(struct i915_request *request, int prio)
 
 	/* Fifo and depth-first replacement ensure our deps execute before us */
 	list_for_each_entry_safe_reverse(dep, p, &dfs, dfs_link) {
-		struct i915_priotree *pt = dep->signaler;
+		struct i915_sched_node *node = dep->signaler;
 
 		INIT_LIST_HEAD(&dep->dfs_link);
 
-		engine = pt_lock_engine(pt, engine);
+		engine = sched_lock_engine(node, engine);
 
-		if (prio <= pt->priority)
+		if (prio <= node->priority)
 			continue;
 
-		pt->priority = prio;
-		if (!list_empty(&pt->link)) {
-			__list_del_entry(&pt->link);
-			queue_request(engine, pt, prio);
+		node->priority = prio;
+		if (!list_empty(&node->link)) {
+			__list_del_entry(&node->link);
+			queue_request(engine, node, prio);
 		}
 
 		if (prio > engine->execlists.queue_priority &&
-		    i915_sw_fence_done(&pt_to_request(pt)->submit))
+		    i915_sw_fence_done(&sched_to_request(node)->submit))
 			__submit_queue(engine, prio);
 	}
 
-- 
cgit v1.1


From b7268c5eed0ab4f052d614b4b0e3fe8a51c9d5a1 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Wed, 18 Apr 2018 19:40:52 +0100
Subject: drm/i915: Pack params to engine->schedule() into a struct

Today we only want to pass along the priority to engine->schedule(), but
in the future we want to have much more control over the various aspects
of the GPU during a context's execution, for example controlling the
frequency allowed. As we need an ever growing number of parameters for
scheduling, move those into a struct for convenience.

v2: Move the anonymous struct into its own function for legibility and
ye olde gcc.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180418184052.7129-3-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/gvt/scheduler.c             |  2 +-
 drivers/gpu/drm/i915/i915_drv.h                  |  3 ++-
 drivers/gpu/drm/i915/i915_gem.c                  | 18 ++++++++++--------
 drivers/gpu/drm/i915/i915_gem_context.c          |  8 ++++----
 drivers/gpu/drm/i915/i915_gem_context.h          | 13 +------------
 drivers/gpu/drm/i915/i915_gpu_error.c            |  8 ++++----
 drivers/gpu/drm/i915/i915_gpu_error.h            |  5 +++--
 drivers/gpu/drm/i915/i915_request.c              |  4 ++--
 drivers/gpu/drm/i915/i915_request.h              |  1 +
 drivers/gpu/drm/i915/i915_scheduler.h            | 17 ++++++++++++++++-
 drivers/gpu/drm/i915/intel_display.c             | 11 ++++++++++-
 drivers/gpu/drm/i915/intel_engine_cs.c           | 18 +++++++++++++++---
 drivers/gpu/drm/i915/intel_guc_submission.c      |  2 +-
 drivers/gpu/drm/i915/intel_lrc.c                 | 20 +++++++++++---------
 drivers/gpu/drm/i915/intel_ringbuffer.h          |  4 +++-
 drivers/gpu/drm/i915/selftests/intel_hangcheck.c |  4 ++--
 drivers/gpu/drm/i915/selftests/intel_lrc.c       |  8 +++++---
 17 files changed, 91 insertions(+), 55 deletions(-)

diff --git a/drivers/gpu/drm/i915/gvt/scheduler.c b/drivers/gpu/drm/i915/gvt/scheduler.c
index 638abe8..f3d2184 100644
--- a/drivers/gpu/drm/i915/gvt/scheduler.c
+++ b/drivers/gpu/drm/i915/gvt/scheduler.c
@@ -1135,7 +1135,7 @@ int intel_vgpu_setup_submission(struct intel_vgpu *vgpu)
 		return PTR_ERR(s->shadow_ctx);
 
 	if (HAS_LOGICAL_RING_PREEMPTION(vgpu->gvt->dev_priv))
-		s->shadow_ctx->priority = INT_MAX;
+		s->shadow_ctx->sched.priority = INT_MAX;
 
 	bitmap_zero(s->shadow_ctx_desc_updated, I915_NUM_ENGINES);
 
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 8e8667d..0286911 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -75,6 +75,7 @@
 #include "i915_gem_timeline.h"
 #include "i915_gpu_error.h"
 #include "i915_request.h"
+#include "i915_scheduler.h"
 #include "i915_vma.h"
 
 #include "intel_gvt.h"
@@ -3158,7 +3159,7 @@ int i915_gem_object_wait(struct drm_i915_gem_object *obj,
 			 struct intel_rps_client *rps);
 int i915_gem_object_wait_priority(struct drm_i915_gem_object *obj,
 				  unsigned int flags,
-				  int priority);
+				  const struct i915_sched_attr *attr);
 #define I915_PRIORITY_DISPLAY I915_PRIORITY_MAX
 
 int __must_check
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 4c9d2a6..795ca83 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -564,7 +564,8 @@ i915_gem_object_wait_reservation(struct reservation_object *resv,
 	return timeout;
 }
 
-static void __fence_set_priority(struct dma_fence *fence, int prio)
+static void __fence_set_priority(struct dma_fence *fence,
+				 const struct i915_sched_attr *attr)
 {
 	struct i915_request *rq;
 	struct intel_engine_cs *engine;
@@ -577,11 +578,12 @@ static void __fence_set_priority(struct dma_fence *fence, int prio)
 
 	rcu_read_lock();
 	if (engine->schedule)
-		engine->schedule(rq, prio);
+		engine->schedule(rq, attr);
 	rcu_read_unlock();
 }
 
-static void fence_set_priority(struct dma_fence *fence, int prio)
+static void fence_set_priority(struct dma_fence *fence,
+			       const struct i915_sched_attr *attr)
 {
 	/* Recurse once into a fence-array */
 	if (dma_fence_is_array(fence)) {
@@ -589,16 +591,16 @@ static void fence_set_priority(struct dma_fence *fence, int prio)
 		int i;
 
 		for (i = 0; i < array->num_fences; i++)
-			__fence_set_priority(array->fences[i], prio);
+			__fence_set_priority(array->fences[i], attr);
 	} else {
-		__fence_set_priority(fence, prio);
+		__fence_set_priority(fence, attr);
 	}
 }
 
 int
 i915_gem_object_wait_priority(struct drm_i915_gem_object *obj,
 			      unsigned int flags,
-			      int prio)
+			      const struct i915_sched_attr *attr)
 {
 	struct dma_fence *excl;
 
@@ -613,7 +615,7 @@ i915_gem_object_wait_priority(struct drm_i915_gem_object *obj,
 			return ret;
 
 		for (i = 0; i < count; i++) {
-			fence_set_priority(shared[i], prio);
+			fence_set_priority(shared[i], attr);
 			dma_fence_put(shared[i]);
 		}
 
@@ -623,7 +625,7 @@ i915_gem_object_wait_priority(struct drm_i915_gem_object *obj,
 	}
 
 	if (excl) {
-		fence_set_priority(excl, prio);
+		fence_set_priority(excl, attr);
 		dma_fence_put(excl);
 	}
 	return 0;
diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c
index 9b3834a..74435af 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/i915_gem_context.c
@@ -281,7 +281,7 @@ __create_hw_context(struct drm_i915_private *dev_priv,
 	kref_init(&ctx->ref);
 	list_add_tail(&ctx->link, &dev_priv->contexts.list);
 	ctx->i915 = dev_priv;
-	ctx->priority = I915_PRIORITY_NORMAL;
+	ctx->sched.priority = I915_PRIORITY_NORMAL;
 
 	INIT_RADIX_TREE(&ctx->handles_vma, GFP_KERNEL);
 	INIT_LIST_HEAD(&ctx->handles_list);
@@ -431,7 +431,7 @@ i915_gem_context_create_kernel(struct drm_i915_private *i915, int prio)
 		return ctx;
 
 	i915_gem_context_clear_bannable(ctx);
-	ctx->priority = prio;
+	ctx->sched.priority = prio;
 	ctx->ring_size = PAGE_SIZE;
 
 	GEM_BUG_ON(!i915_gem_context_is_kernel(ctx));
@@ -753,7 +753,7 @@ int i915_gem_context_getparam_ioctl(struct drm_device *dev, void *data,
 		args->value = i915_gem_context_is_bannable(ctx);
 		break;
 	case I915_CONTEXT_PARAM_PRIORITY:
-		args->value = ctx->priority;
+		args->value = ctx->sched.priority;
 		break;
 	default:
 		ret = -EINVAL;
@@ -826,7 +826,7 @@ int i915_gem_context_setparam_ioctl(struct drm_device *dev, void *data,
 				 !capable(CAP_SYS_NICE))
 				ret = -EPERM;
 			else
-				ctx->priority = priority;
+				ctx->sched.priority = priority;
 		}
 		break;
 
diff --git a/drivers/gpu/drm/i915/i915_gem_context.h b/drivers/gpu/drm/i915/i915_gem_context.h
index 7854262..b12a8a8 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.h
+++ b/drivers/gpu/drm/i915/i915_gem_context.h
@@ -137,18 +137,7 @@ struct i915_gem_context {
 	 */
 	u32 user_handle;
 
-	/**
-	 * @priority: execution and service priority
-	 *
-	 * All clients are equal, but some are more equal than others!
-	 *
-	 * Requests from a context with a greater (more positive) value of
-	 * @priority will be executed before those with a lower @priority
-	 * value, forming a simple QoS.
-	 *
-	 * The &drm_i915_private.kernel_context is assigned the lowest priority.
-	 */
-	int priority;
+	struct i915_sched_attr sched;
 
 	/** ggtt_offset_bias: placement restriction for context objects */
 	u32 ggtt_offset_bias;
diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c
index 6b5b9b3..671ffa3 100644
--- a/drivers/gpu/drm/i915/i915_gpu_error.c
+++ b/drivers/gpu/drm/i915/i915_gpu_error.c
@@ -411,7 +411,7 @@ static void error_print_request(struct drm_i915_error_state_buf *m,
 
 	err_printf(m, "%s pid %d, ban score %d, seqno %8x:%08x, prio %d, emitted %dms ago, head %08x, tail %08x\n",
 		   prefix, erq->pid, erq->ban_score,
-		   erq->context, erq->seqno, erq->priority,
+		   erq->context, erq->seqno, erq->sched_attr.priority,
 		   jiffies_to_msecs(jiffies - erq->jiffies),
 		   erq->head, erq->tail);
 }
@@ -422,7 +422,7 @@ static void error_print_context(struct drm_i915_error_state_buf *m,
 {
 	err_printf(m, "%s%s[%d] user_handle %d hw_id %d, prio %d, ban score %d%s guilty %d active %d\n",
 		   header, ctx->comm, ctx->pid, ctx->handle, ctx->hw_id,
-		   ctx->priority, ctx->ban_score, bannable(ctx),
+		   ctx->sched_attr.priority, ctx->ban_score, bannable(ctx),
 		   ctx->guilty, ctx->active);
 }
 
@@ -1278,7 +1278,7 @@ static void record_request(struct i915_request *request,
 			   struct drm_i915_error_request *erq)
 {
 	erq->context = request->ctx->hw_id;
-	erq->priority = request->sched.priority;
+	erq->sched_attr = request->sched.attr;
 	erq->ban_score = atomic_read(&request->ctx->ban_score);
 	erq->seqno = request->global_seqno;
 	erq->jiffies = request->emitted_jiffies;
@@ -1372,7 +1372,7 @@ static void record_context(struct drm_i915_error_context *e,
 
 	e->handle = ctx->user_handle;
 	e->hw_id = ctx->hw_id;
-	e->priority = ctx->priority;
+	e->sched_attr = ctx->sched;
 	e->ban_score = atomic_read(&ctx->ban_score);
 	e->bannable = i915_gem_context_is_bannable(ctx);
 	e->guilty = atomic_read(&ctx->guilty_count);
diff --git a/drivers/gpu/drm/i915/i915_gpu_error.h b/drivers/gpu/drm/i915/i915_gpu_error.h
index c05b603..5d6fdcb 100644
--- a/drivers/gpu/drm/i915/i915_gpu_error.h
+++ b/drivers/gpu/drm/i915/i915_gpu_error.h
@@ -20,6 +20,7 @@
 #include "i915_gem.h"
 #include "i915_gem_gtt.h"
 #include "i915_params.h"
+#include "i915_scheduler.h"
 
 struct drm_i915_private;
 struct intel_overlay_error_state;
@@ -122,11 +123,11 @@ struct i915_gpu_state {
 			pid_t pid;
 			u32 handle;
 			u32 hw_id;
-			int priority;
 			int ban_score;
 			int active;
 			int guilty;
 			bool bannable;
+			struct i915_sched_attr sched_attr;
 		} context;
 
 		struct drm_i915_error_object {
@@ -147,11 +148,11 @@ struct i915_gpu_state {
 			long jiffies;
 			pid_t pid;
 			u32 context;
-			int priority;
 			int ban_score;
 			u32 seqno;
 			u32 head;
 			u32 tail;
+			struct i915_sched_attr sched_attr;
 		} *requests, execlist[EXECLIST_MAX_PORTS];
 		unsigned int num_ports;
 
diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c
index dfcc6a0..b692a9f 100644
--- a/drivers/gpu/drm/i915/i915_request.c
+++ b/drivers/gpu/drm/i915/i915_request.c
@@ -193,7 +193,7 @@ i915_sched_node_init(struct i915_sched_node *node)
 	INIT_LIST_HEAD(&node->signalers_list);
 	INIT_LIST_HEAD(&node->waiters_list);
 	INIT_LIST_HEAD(&node->link);
-	node->priority = I915_PRIORITY_INVALID;
+	node->attr.priority = I915_PRIORITY_INVALID;
 }
 
 static int reset_all_global_seqno(struct drm_i915_private *i915, u32 seqno)
@@ -1064,7 +1064,7 @@ void __i915_request_add(struct i915_request *request, bool flush_caches)
 	 */
 	rcu_read_lock();
 	if (engine->schedule)
-		engine->schedule(request, request->ctx->priority);
+		engine->schedule(request, &request->ctx->sched);
 	rcu_read_unlock();
 
 	local_bh_disable();
diff --git a/drivers/gpu/drm/i915/i915_request.h b/drivers/gpu/drm/i915/i915_request.h
index 35b8a98..8f31ca8 100644
--- a/drivers/gpu/drm/i915/i915_request.h
+++ b/drivers/gpu/drm/i915/i915_request.h
@@ -30,6 +30,7 @@
 #include "i915_gem.h"
 #include "i915_scheduler.h"
 #include "i915_sw_fence.h"
+#include "i915_scheduler.h"
 
 #include <uapi/drm/i915_drm.h>
 
diff --git a/drivers/gpu/drm/i915/i915_scheduler.h b/drivers/gpu/drm/i915/i915_scheduler.h
index 754243e..70a4222 100644
--- a/drivers/gpu/drm/i915/i915_scheduler.h
+++ b/drivers/gpu/drm/i915/i915_scheduler.h
@@ -19,6 +19,21 @@ enum {
 	I915_PRIORITY_INVALID = INT_MIN
 };
 
+struct i915_sched_attr {
+	/**
+	 * @priority: execution and service priority
+	 *
+	 * All clients are equal, but some are more equal than others!
+	 *
+	 * Requests from a context with a greater (more positive) value of
+	 * @priority will be executed before those with a lower @priority
+	 * value, forming a simple QoS.
+	 *
+	 * The &drm_i915_private.kernel_context is assigned the lowest priority.
+	 */
+	int priority;
+};
+
 /*
  * "People assume that time is a strict progression of cause to effect, but
  * actually, from a nonlinear, non-subjective viewpoint, it's more like a big
@@ -42,7 +57,7 @@ struct i915_sched_node {
 	struct list_head signalers_list; /* those before us, we depend upon */
 	struct list_head waiters_list; /* those after us, they depend upon us */
 	struct list_head link;
-	int priority;
+	struct i915_sched_attr attr;
 };
 
 struct i915_dependency {
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index e04050e..43d54c7 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -12761,6 +12761,15 @@ static void intel_plane_unpin_fb(struct intel_plane_state *old_plane_state)
 		intel_unpin_fb_vma(vma, old_plane_state->flags);
 }
 
+static void fb_obj_bump_render_priority(struct drm_i915_gem_object *obj)
+{
+	struct i915_sched_attr attr = {
+		.priority = I915_PRIORITY_DISPLAY,
+	};
+
+	i915_gem_object_wait_priority(obj, 0, &attr);
+}
+
 /**
  * intel_prepare_plane_fb - Prepare fb for usage on plane
  * @plane: drm plane to prepare for
@@ -12837,7 +12846,7 @@ intel_prepare_plane_fb(struct drm_plane *plane,
 
 	ret = intel_plane_pin_fb(to_intel_plane_state(new_state));
 
-	i915_gem_object_wait_priority(obj, 0, I915_PRIORITY_DISPLAY);
+	fb_obj_bump_render_priority(obj);
 
 	mutex_unlock(&dev_priv->drm.struct_mutex);
 	i915_gem_object_unpin_pages(obj);
diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
index b542b1a..be608f7 100644
--- a/drivers/gpu/drm/i915/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/intel_engine_cs.c
@@ -1113,17 +1113,29 @@ unsigned int intel_engines_has_context_isolation(struct drm_i915_private *i915)
 	return which;
 }
 
+static void print_sched_attr(struct drm_printer *m,
+			     const struct drm_i915_private *i915,
+			     const struct i915_sched_attr *attr)
+{
+	if (attr->priority == I915_PRIORITY_INVALID)
+		return;
+
+	drm_printf(m, "prio=%d", attr->priority);
+}
+
 static void print_request(struct drm_printer *m,
 			  struct i915_request *rq,
 			  const char *prefix)
 {
 	const char *name = rq->fence.ops->get_timeline_name(&rq->fence);
 
-	drm_printf(m, "%s%x%s [%llx:%x] prio=%d @ %dms: %s\n", prefix,
+	drm_printf(m, "%s%x%s [%llx:%x] ",
+		   prefix,
 		   rq->global_seqno,
 		   i915_request_completed(rq) ? "!" : "",
-		   rq->fence.context, rq->fence.seqno,
-		   rq->sched.priority,
+		   rq->fence.context, rq->fence.seqno);
+	print_sched_attr(m, rq->i915, &rq->sched.attr);
+	drm_printf(m, " @ %dms: %s\n",
 		   jiffies_to_msecs(jiffies - rq->emitted_jiffies),
 		   name);
 }
diff --git a/drivers/gpu/drm/i915/intel_guc_submission.c b/drivers/gpu/drm/i915/intel_guc_submission.c
index 0755f5c..02da058 100644
--- a/drivers/gpu/drm/i915/intel_guc_submission.c
+++ b/drivers/gpu/drm/i915/intel_guc_submission.c
@@ -659,7 +659,7 @@ static void port_assign(struct execlist_port *port, struct i915_request *rq)
 
 static inline int rq_prio(const struct i915_request *rq)
 {
-	return rq->sched.priority;
+	return rq->sched.attr.priority;
 }
 
 static inline int port_prio(const struct execlist_port *port)
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 062ed6e..029901a 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -177,7 +177,7 @@ static inline struct i915_priolist *to_priolist(struct rb_node *rb)
 
 static inline int rq_prio(const struct i915_request *rq)
 {
-	return rq->sched.priority;
+	return rq->sched.attr.priority;
 }
 
 static inline bool need_preempt(const struct intel_engine_cs *engine,
@@ -1172,11 +1172,13 @@ sched_lock_engine(struct i915_sched_node *node, struct intel_engine_cs *locked)
 	return engine;
 }
 
-static void execlists_schedule(struct i915_request *request, int prio)
+static void execlists_schedule(struct i915_request *request,
+			       const struct i915_sched_attr *attr)
 {
 	struct intel_engine_cs *engine;
 	struct i915_dependency *dep, *p;
 	struct i915_dependency stack;
+	const int prio = attr->priority;
 	LIST_HEAD(dfs);
 
 	GEM_BUG_ON(prio == I915_PRIORITY_INVALID);
@@ -1184,7 +1186,7 @@ static void execlists_schedule(struct i915_request *request, int prio)
 	if (i915_request_completed(request))
 		return;
 
-	if (prio <= READ_ONCE(request->sched.priority))
+	if (prio <= READ_ONCE(request->sched.attr.priority))
 		return;
 
 	/* Need BKL in order to use the temporary link inside i915_dependency */
@@ -1226,8 +1228,8 @@ static void execlists_schedule(struct i915_request *request, int prio)
 			if (i915_sched_node_signaled(p->signaler))
 				continue;
 
-			GEM_BUG_ON(p->signaler->priority < node->priority);
-			if (prio > READ_ONCE(p->signaler->priority))
+			GEM_BUG_ON(p->signaler->attr.priority < node->attr.priority);
+			if (prio > READ_ONCE(p->signaler->attr.priority))
 				list_move_tail(&p->dfs_link, &dfs);
 		}
 	}
@@ -1238,9 +1240,9 @@ static void execlists_schedule(struct i915_request *request, int prio)
 	 * execlists_submit_request()), we can set our own priority and skip
 	 * acquiring the engine locks.
 	 */
-	if (request->sched.priority == I915_PRIORITY_INVALID) {
+	if (request->sched.attr.priority == I915_PRIORITY_INVALID) {
 		GEM_BUG_ON(!list_empty(&request->sched.link));
-		request->sched.priority = prio;
+		request->sched.attr = *attr;
 		if (stack.dfs_link.next == stack.dfs_link.prev)
 			return;
 		__list_del_entry(&stack.dfs_link);
@@ -1257,10 +1259,10 @@ static void execlists_schedule(struct i915_request *request, int prio)
 
 		engine = sched_lock_engine(node, engine);
 
-		if (prio <= node->priority)
+		if (prio <= node->attr.priority)
 			continue;
 
-		node->priority = prio;
+		node->attr.priority = prio;
 		if (!list_empty(&node->link)) {
 			__list_del_entry(&node->link);
 			queue_request(engine, node, prio);
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index 7170416..c5e2790 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -14,6 +14,7 @@
 #include "intel_gpu_commands.h"
 
 struct drm_printer;
+struct i915_sched_attr;
 
 #define I915_CMD_HASH_ORDER 9
 
@@ -460,7 +461,8 @@ struct intel_engine_cs {
 	 *
 	 * Called under the struct_mutex.
 	 */
-	void		(*schedule)(struct i915_request *request, int priority);
+	void		(*schedule)(struct i915_request *request,
+				    const struct i915_sched_attr *attr);
 
 	/*
 	 * Cancel all requests on the hardware, or queued for execution.
diff --git a/drivers/gpu/drm/i915/selftests/intel_hangcheck.c b/drivers/gpu/drm/i915/selftests/intel_hangcheck.c
index 24f913f..f7ee54e 100644
--- a/drivers/gpu/drm/i915/selftests/intel_hangcheck.c
+++ b/drivers/gpu/drm/i915/selftests/intel_hangcheck.c
@@ -628,7 +628,7 @@ static int active_engine(void *data)
 		}
 
 		if (arg->flags & TEST_PRIORITY)
-			ctx[idx]->priority =
+			ctx[idx]->sched.priority =
 				i915_prandom_u32_max_state(512, &prng);
 
 		rq[idx] = i915_request_get(new);
@@ -683,7 +683,7 @@ static int __igt_reset_engines(struct drm_i915_private *i915,
 			return err;
 
 		if (flags & TEST_PRIORITY)
-			h.ctx->priority = 1024;
+			h.ctx->sched.priority = 1024;
 	}
 
 	for_each_engine(engine, i915, id) {
diff --git a/drivers/gpu/drm/i915/selftests/intel_lrc.c b/drivers/gpu/drm/i915/selftests/intel_lrc.c
index 0481e2e..ee7e22d1 100644
--- a/drivers/gpu/drm/i915/selftests/intel_lrc.c
+++ b/drivers/gpu/drm/i915/selftests/intel_lrc.c
@@ -335,12 +335,12 @@ static int live_preempt(void *arg)
 	ctx_hi = kernel_context(i915);
 	if (!ctx_hi)
 		goto err_spin_lo;
-	ctx_hi->priority = I915_CONTEXT_MAX_USER_PRIORITY;
+	ctx_hi->sched.priority = I915_CONTEXT_MAX_USER_PRIORITY;
 
 	ctx_lo = kernel_context(i915);
 	if (!ctx_lo)
 		goto err_ctx_hi;
-	ctx_lo->priority = I915_CONTEXT_MIN_USER_PRIORITY;
+	ctx_lo->sched.priority = I915_CONTEXT_MIN_USER_PRIORITY;
 
 	for_each_engine(engine, i915, id) {
 		struct i915_request *rq;
@@ -407,6 +407,7 @@ static int live_late_preempt(void *arg)
 	struct i915_gem_context *ctx_hi, *ctx_lo;
 	struct spinner spin_hi, spin_lo;
 	struct intel_engine_cs *engine;
+	struct i915_sched_attr attr = {};
 	enum intel_engine_id id;
 	int err = -ENOMEM;
 
@@ -458,7 +459,8 @@ static int live_late_preempt(void *arg)
 			goto err_wedged;
 		}
 
-		engine->schedule(rq, I915_PRIORITY_MAX);
+		attr.priority = I915_PRIORITY_MAX;
+		engine->schedule(rq, &attr);
 
 		if (!wait_for_spinner(&spin_hi, rq)) {
 			pr_err("High priority context failed to preempt the low priority context\n");
-- 
cgit v1.1


From 2a5b95b448485e143ec3e004eabe53b31db78eb3 Mon Sep 17 00:00:00 2001
From: Abhay Kumar <abhay.kumar@intel.com>
Date: Wed, 18 Apr 2018 13:37:07 +0300
Subject: drm/i915/audio: set minimum CD clock to twice the BCLK
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

In GLK when the device boots with only 1366x768 panel without audio, HDA
codec doesn't come up. In this case, the CDCLK is less than twice the
BCLK. Even though audio isn't being enabled, having a too low CDCLK
leads to audio probe failing altogether.

Require CDCLK to be at least twice the BLCK regardless of audio. This is
a minimal fix to improve things. Unfortunately, this a) leads to too
high CDCLK being used when audio is not used, and b) is still not enough
to fix audio probe when no outputs are connected at probe time.

The proper fix would be to increase CDCLK dynamically from the audio
component hooks.

v2:
    - Address comment (Jani)
    - New design approach
v3: - Typo fix on top of v1

v4 by Jani: rewrite commit message, add comment in code

Cc: stable@vger.kernel.org
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Dhinakaran Pandiyan <dhinakaran.pandiyan@gmail.com>
Cc: Wenkai Du <wenkai.du@intel.com>
Reviewed-by: Wenkai Du <wenkai.du@intel.com>
Tested-by: Wenkai Du <wenkai.du@intel.com>
Acked-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=102937
Signed-off-by: Abhay Kumar <abhay.kumar@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180418103707.14645-1-jani.nikula@intel.com
---
 drivers/gpu/drm/i915/intel_cdclk.c | 16 ++++++++++++++--
 1 file changed, 14 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_cdclk.c b/drivers/gpu/drm/i915/intel_cdclk.c
index dc7db8a..ebca83a 100644
--- a/drivers/gpu/drm/i915/intel_cdclk.c
+++ b/drivers/gpu/drm/i915/intel_cdclk.c
@@ -2140,10 +2140,22 @@ int intel_crtc_compute_min_cdclk(const struct intel_crtc_state *crtc_state)
 		}
 	}
 
-	/* According to BSpec, "The CD clock frequency must be at least twice
+	/*
+	 * According to BSpec, "The CD clock frequency must be at least twice
 	 * the frequency of the Azalia BCLK." and BCLK is 96 MHz by default.
+	 *
+	 * FIXME: Check the actual, not default, BCLK being used.
+	 *
+	 * FIXME: This does not depend on ->has_audio because the higher CDCLK
+	 * is required for audio probe, also when there are no audio capable
+	 * displays connected at probe time. This leads to unnecessarily high
+	 * CDCLK when audio is not required.
+	 *
+	 * FIXME: This limit is only applied when there are displays connected
+	 * at probe time. If we probe without displays, we'll still end up using
+	 * the platform minimum CDCLK, failing audio probe.
 	 */
-	if (crtc_state->has_audio && INTEL_GEN(dev_priv) >= 9)
+	if (INTEL_GEN(dev_priv) >= 9)
 		min_cdclk = max(2 * 96000, min_cdclk);
 
 	/*
-- 
cgit v1.1


From e307126a2c8e792a4b426ee3ab827d1285544e12 Mon Sep 17 00:00:00 2001
From: Jani Nikula <jani.nikula@intel.com>
Date: Thu, 19 Apr 2018 11:59:39 +0300
Subject: drm/i915/dsi: improve dphy param limits logging
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Move the limit checks near the calculations for each field, and actually
log the values that exceed limits.

Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180419085940.21505-1-jani.nikula@intel.com
---
 drivers/gpu/drm/i915/intel_dsi_vbt.c | 34 ++++++++++++++++++----------------
 1 file changed, 18 insertions(+), 16 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_dsi_vbt.c b/drivers/gpu/drm/i915/intel_dsi_vbt.c
index 91c07b0..4d6ffa7 100644
--- a/drivers/gpu/drm/i915/intel_dsi_vbt.c
+++ b/drivers/gpu/drm/i915/intel_dsi_vbt.c
@@ -647,6 +647,11 @@ bool intel_dsi_vbt_init(struct intel_dsi *intel_dsi, u16 panel_id)
 	/* prepare count */
 	prepare_cnt = DIV_ROUND_UP(ths_prepare_ns * ui_den, ui_num * mul);
 
+	if (prepare_cnt > PREPARE_CNT_MAX) {
+		DRM_DEBUG_KMS("prepare count too high %u\n", prepare_cnt);
+		prepare_cnt = PREPARE_CNT_MAX;
+	}
+
 	/* exit zero count */
 	exit_zero_cnt = DIV_ROUND_UP(
 				(ths_prepare_hszero - ths_prepare_ns) * ui_den,
@@ -662,32 +667,29 @@ bool intel_dsi_vbt_init(struct intel_dsi *intel_dsi, u16 panel_id)
 	if (exit_zero_cnt < (55 * ui_den / ui_num) && (55 * ui_den) % ui_num)
 		exit_zero_cnt += 1;
 
+	if (exit_zero_cnt > EXIT_ZERO_CNT_MAX) {
+		DRM_DEBUG_KMS("exit zero count too high %u\n", exit_zero_cnt);
+		exit_zero_cnt = EXIT_ZERO_CNT_MAX;
+	}
+
 	/* clk zero count */
 	clk_zero_cnt = DIV_ROUND_UP(
 				(tclk_prepare_clkzero -	ths_prepare_ns)
 				* ui_den, ui_num * mul);
 
+	if (clk_zero_cnt > CLK_ZERO_CNT_MAX) {
+		DRM_DEBUG_KMS("clock zero count too high %u\n", clk_zero_cnt);
+		clk_zero_cnt = CLK_ZERO_CNT_MAX;
+	}
+
 	/* trail count */
 	tclk_trail_ns = max(mipi_config->tclk_trail, mipi_config->ths_trail);
 	trail_cnt = DIV_ROUND_UP(tclk_trail_ns * ui_den, ui_num * mul);
 
-	if (prepare_cnt > PREPARE_CNT_MAX ||
-		exit_zero_cnt > EXIT_ZERO_CNT_MAX ||
-		clk_zero_cnt > CLK_ZERO_CNT_MAX ||
-		trail_cnt > TRAIL_CNT_MAX)
-		DRM_DEBUG_DRIVER("Values crossing maximum limits, restricting to max values\n");
-
-	if (prepare_cnt > PREPARE_CNT_MAX)
-		prepare_cnt = PREPARE_CNT_MAX;
-
-	if (exit_zero_cnt > EXIT_ZERO_CNT_MAX)
-		exit_zero_cnt = EXIT_ZERO_CNT_MAX;
-
-	if (clk_zero_cnt > CLK_ZERO_CNT_MAX)
-		clk_zero_cnt = CLK_ZERO_CNT_MAX;
-
-	if (trail_cnt > TRAIL_CNT_MAX)
+	if (trail_cnt > TRAIL_CNT_MAX) {
+		DRM_DEBUG_KMS("trail count too high %u\n", trail_cnt);
 		trail_cnt = TRAIL_CNT_MAX;
+	}
 
 	/* B080 */
 	intel_dsi->dphy_reg = exit_zero_cnt << 24 | trail_cnt << 16 |
-- 
cgit v1.1


From 911a26484c33e10de6237228ca1d7293548e9f49 Mon Sep 17 00:00:00 2001
From: Ilan Peer <ilan.peer@intel.com>
Date: Tue, 3 Apr 2018 11:35:22 +0300
Subject: mac80211: Fix condition validating WMM IE

Commit c470bdc1aaf3 ("mac80211: don't WARN on bad WMM parameters from
buggy APs") handled cases where an AP reports a zeroed WMM
IE. However, the condition that checks the validity accessed the wrong
index in the ieee80211_tx_queue_params array, thus wrongly deducing
that the parameters are invalid. Fix it.

Fixes: c470bdc1aaf3 ("mac80211: don't WARN on bad WMM parameters from buggy APs")
Signed-off-by: Ilan Peer <ilan.peer@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
---
 net/mac80211/mlme.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c
index 69449db..6fe72ef 100644
--- a/net/mac80211/mlme.c
+++ b/net/mac80211/mlme.c
@@ -1787,7 +1787,7 @@ static bool ieee80211_sta_wmm_params(struct ieee80211_local *local,
 		params[ac].acm = acm;
 		params[ac].uapsd = uapsd;
 
-		if (params->cw_min == 0 ||
+		if (params[ac].cw_min == 0 ||
 		    params[ac].cw_min > params[ac].cw_max) {
 			sdata_info(sdata,
 				   "AP has invalid WMM params (CWmin/max=%d/%d for ACI %d), using defaults\n",
-- 
cgit v1.1


From a7cfebcb7594a24609268f91299ab85ba064bf82 Mon Sep 17 00:00:00 2001
From: Johannes Berg <johannes.berg@intel.com>
Date: Tue, 3 Apr 2018 14:33:49 +0200
Subject: cfg80211: limit wiphy names to 128 bytes

There's currently no limit on wiphy names, other than netlink
message size and memory limitations, but that causes issues when,
for example, the wiphy name is used in a uevent, e.g. in rfkill
where we use the same name for the rfkill instance, and then the
buffer there is "only" 2k for the environment variables.

This was reported by syzkaller, which used a 4k name.

Limit the name to something reasonable, I randomly picked 128.

Reported-by: syzbot+230d9e642a85d3fec29c@syzkaller.appspotmail.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
---
 include/uapi/linux/nl80211.h | 2 ++
 net/wireless/core.c          | 3 +++
 2 files changed, 5 insertions(+)

diff --git a/include/uapi/linux/nl80211.h b/include/uapi/linux/nl80211.h
index 15daf5e..9c36301 100644
--- a/include/uapi/linux/nl80211.h
+++ b/include/uapi/linux/nl80211.h
@@ -2698,6 +2698,8 @@ enum nl80211_attrs {
 #define NL80211_ATTR_KEYS NL80211_ATTR_KEYS
 #define NL80211_ATTR_FEATURE_FLAGS NL80211_ATTR_FEATURE_FLAGS
 
+#define NL80211_WIPHY_NAME_MAXLEN		128
+
 #define NL80211_MAX_SUPP_RATES			32
 #define NL80211_MAX_SUPP_HT_RATES		77
 #define NL80211_MAX_SUPP_REG_RULES		64
diff --git a/net/wireless/core.c b/net/wireless/core.c
index a6f3cac..c0fd8a8 100644
--- a/net/wireless/core.c
+++ b/net/wireless/core.c
@@ -95,6 +95,9 @@ static int cfg80211_dev_check_name(struct cfg80211_registered_device *rdev,
 
 	ASSERT_RTNL();
 
+	if (strlen(newname) > NL80211_WIPHY_NAME_MAXLEN)
+		return -EINVAL;
+
 	/* prohibit calling the thing phy%d when %d is not its number */
 	sscanf(newname, PHY_NAME "%d%n", &wiphy_idx, &taken);
 	if (taken == strlen(newname) && wiphy_idx != rdev->wiphy_idx) {
-- 
cgit v1.1


From 83826469e36b9c8219f88c16713a755b2cea6ff5 Mon Sep 17 00:00:00 2001
From: "weiyongjun (A)" <weiyongjun1@huawei.com>
Date: Fri, 30 Mar 2018 02:07:05 +0000
Subject: cfg80211: fix possible memory leak in regdb_query_country()

'wmm_ptrs' is malloced in regdb_query_country() and should be freed
before leaving from the error handling cases, otherwise it will cause
memory leak.

Fixes: 230ebaa189af ("cfg80211: read wmm rules from regulatory database")
Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
[johannes: add Fixes tag]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
---
 net/wireless/reg.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/net/wireless/reg.c b/net/wireless/reg.c
index 16c7e4e..ac3e12c 100644
--- a/net/wireless/reg.c
+++ b/net/wireless/reg.c
@@ -1026,6 +1026,7 @@ static int regdb_query_country(const struct fwdb_header *db,
 
 			if (!tmp_rd) {
 				kfree(regdom);
+				kfree(wmm_ptrs);
 				return -ENOMEM;
 			}
 			regdom = tmp_rd;
-- 
cgit v1.1


From eadd2721d080e276f35abc1daab15cddd121e40f Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= <ville.syrjala@linux.intel.com>
Date: Fri, 16 Mar 2018 20:36:25 +0200
Subject: drm/i915: Protect PIPE_CONF_CHECK macros with do {} while(0)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Make the PIPE_CONF_CHECK macros a bit more robust by wrapping them
in do {} while(0). Avoids funky sirprises when you try put an 'else'
after a PIPE_CONF_CHECK invocation...

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180316183625.16316-1-ville.syrjala@linux.intel.com
Reviewed-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> #irc
---
 drivers/gpu/drm/i915/intel_display.c | 45 +++++++++++++++++++++---------------
 1 file changed, 27 insertions(+), 18 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 43d54c7..687e701 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -11141,39 +11141,42 @@ intel_pipe_config_compare(struct drm_i915_private *dev_priv,
 		(current_config->base.mode.private_flags & I915_MODE_FLAG_INHERITED) &&
 		!(pipe_config->base.mode.private_flags & I915_MODE_FLAG_INHERITED);
 
-#define PIPE_CONF_CHECK_X(name)	\
+#define PIPE_CONF_CHECK_X(name) do { \
 	if (current_config->name != pipe_config->name) { \
 		pipe_config_err(adjust, __stringify(name), \
 			  "(expected 0x%08x, found 0x%08x)\n", \
 			  current_config->name, \
 			  pipe_config->name); \
 		ret = false; \
-	}
+	} \
+} while (0)
 
-#define PIPE_CONF_CHECK_I(name)	\
+#define PIPE_CONF_CHECK_I(name) do { \
 	if (current_config->name != pipe_config->name) { \
 		pipe_config_err(adjust, __stringify(name), \
 			  "(expected %i, found %i)\n", \
 			  current_config->name, \
 			  pipe_config->name); \
 		ret = false; \
-	}
+	} \
+} while (0)
 
-#define PIPE_CONF_CHECK_BOOL(name)	\
+#define PIPE_CONF_CHECK_BOOL(name) do { \
 	if (current_config->name != pipe_config->name) { \
 		pipe_config_err(adjust, __stringify(name), \
 			  "(expected %s, found %s)\n", \
 			  yesno(current_config->name), \
 			  yesno(pipe_config->name)); \
 		ret = false; \
-	}
+	} \
+} while (0)
 
 /*
  * Checks state where we only read out the enabling, but not the entire
  * state itself (like full infoframes or ELD for audio). These states
  * require a full modeset on bootup to fix up.
  */
-#define PIPE_CONF_CHECK_BOOL_INCOMPLETE(name) \
+#define PIPE_CONF_CHECK_BOOL_INCOMPLETE(name) do { \
 	if (!fixup_inherited || (!current_config->name && !pipe_config->name)) { \
 		PIPE_CONF_CHECK_BOOL(name); \
 	} else { \
@@ -11182,18 +11185,20 @@ intel_pipe_config_compare(struct drm_i915_private *dev_priv,
 			  yesno(current_config->name), \
 			  yesno(pipe_config->name)); \
 		ret = false; \
-	}
+	} \
+} while (0)
 
-#define PIPE_CONF_CHECK_P(name)	\
+#define PIPE_CONF_CHECK_P(name) do { \
 	if (current_config->name != pipe_config->name) { \
 		pipe_config_err(adjust, __stringify(name), \
 			  "(expected %p, found %p)\n", \
 			  current_config->name, \
 			  pipe_config->name); \
 		ret = false; \
-	}
+	} \
+} while (0)
 
-#define PIPE_CONF_CHECK_M_N(name) \
+#define PIPE_CONF_CHECK_M_N(name) do { \
 	if (!intel_compare_link_m_n(&current_config->name, \
 				    &pipe_config->name,\
 				    adjust)) { \
@@ -11211,14 +11216,15 @@ intel_pipe_config_compare(struct drm_i915_private *dev_priv,
 			  pipe_config->name.link_m, \
 			  pipe_config->name.link_n); \
 		ret = false; \
-	}
+	} \
+} while (0)
 
 /* This is required for BDW+ where there is only one set of registers for
  * switching between high and low RR.
  * This macro can be used whenever a comparison has to be made between one
  * hw state and multiple sw state variables.
  */
-#define PIPE_CONF_CHECK_M_N_ALT(name, alt_name) \
+#define PIPE_CONF_CHECK_M_N_ALT(name, alt_name) do { \
 	if (!intel_compare_link_m_n(&current_config->name, \
 				    &pipe_config->name, adjust) && \
 	    !intel_compare_link_m_n(&current_config->alt_name, \
@@ -11243,9 +11249,10 @@ intel_pipe_config_compare(struct drm_i915_private *dev_priv,
 			  pipe_config->name.link_m, \
 			  pipe_config->name.link_n); \
 		ret = false; \
-	}
+	} \
+} while (0)
 
-#define PIPE_CONF_CHECK_FLAGS(name, mask)	\
+#define PIPE_CONF_CHECK_FLAGS(name, mask) do { \
 	if ((current_config->name ^ pipe_config->name) & (mask)) { \
 		pipe_config_err(adjust, __stringify(name), \
 			  "(%x) (expected %i, found %i)\n", \
@@ -11253,16 +11260,18 @@ intel_pipe_config_compare(struct drm_i915_private *dev_priv,
 			  current_config->name & (mask), \
 			  pipe_config->name & (mask)); \
 		ret = false; \
-	}
+	} \
+} while (0)
 
-#define PIPE_CONF_CHECK_CLOCK_FUZZY(name) \
+#define PIPE_CONF_CHECK_CLOCK_FUZZY(name) do { \
 	if (!intel_fuzzy_clock_check(current_config->name, pipe_config->name)) { \
 		pipe_config_err(adjust, __stringify(name), \
 			  "(expected %i, found %i)\n", \
 			  current_config->name, \
 			  pipe_config->name); \
 		ret = false; \
-	}
+	} \
+} while (0)
 
 #define PIPE_CONF_QUIRK(quirk)	\
 	((current_config->quirks | pipe_config->quirks) & (quirk))
-- 
cgit v1.1


From eb0b4aa89cf21b69e15168010189e9d9c7483e54 Mon Sep 17 00:00:00 2001
From: Laura Abbott <labbott@redhat.com>
Date: Wed, 18 Apr 2018 10:08:32 -0700
Subject: x86/xen: Remove use of VLAs

There's an ongoing effort to remove VLAs[1] from the kernel to eventually
turn on -Wvla. It turns out, the few VLAs in use in Xen produce only a
single entry array that is always bounded by GDT_SIZE. Clean up the code to
get rid of the VLA and the loop.

[1] https://lkml.org/lkml/2018/3/7/621

Signed-off-by: Laura Abbott <labbott@redhat.com>
Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>

[boris: Use BUG_ON(size>PAGE_SIZE) instead of GDT_SIZE]
Signed-off-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
---
 arch/x86/xen/enlighten_pv.c | 86 ++++++++++++++++-----------------------------
 1 file changed, 31 insertions(+), 55 deletions(-)

diff --git a/arch/x86/xen/enlighten_pv.c b/arch/x86/xen/enlighten_pv.c
index c36d23a..357969a 100644
--- a/arch/x86/xen/enlighten_pv.c
+++ b/arch/x86/xen/enlighten_pv.c
@@ -421,45 +421,33 @@ static void xen_load_gdt(const struct desc_ptr *dtr)
 {
 	unsigned long va = dtr->address;
 	unsigned int size = dtr->size + 1;
-	unsigned pages = DIV_ROUND_UP(size, PAGE_SIZE);
-	unsigned long frames[pages];
-	int f;
-
-	/*
-	 * A GDT can be up to 64k in size, which corresponds to 8192
-	 * 8-byte entries, or 16 4k pages..
-	 */
+	unsigned long pfn, mfn;
+	int level;
+	pte_t *ptep;
+	void *virt;
 
-	BUG_ON(size > 65536);
+	/* @size should be at most GDT_SIZE which is smaller than PAGE_SIZE. */
+	BUG_ON(size > PAGE_SIZE);
 	BUG_ON(va & ~PAGE_MASK);
 
-	for (f = 0; va < dtr->address + size; va += PAGE_SIZE, f++) {
-		int level;
-		pte_t *ptep;
-		unsigned long pfn, mfn;
-		void *virt;
-
-		/*
-		 * The GDT is per-cpu and is in the percpu data area.
-		 * That can be virtually mapped, so we need to do a
-		 * page-walk to get the underlying MFN for the
-		 * hypercall.  The page can also be in the kernel's
-		 * linear range, so we need to RO that mapping too.
-		 */
-		ptep = lookup_address(va, &level);
-		BUG_ON(ptep == NULL);
-
-		pfn = pte_pfn(*ptep);
-		mfn = pfn_to_mfn(pfn);
-		virt = __va(PFN_PHYS(pfn));
+	/*
+	 * The GDT is per-cpu and is in the percpu data area.
+	 * That can be virtually mapped, so we need to do a
+	 * page-walk to get the underlying MFN for the
+	 * hypercall.  The page can also be in the kernel's
+	 * linear range, so we need to RO that mapping too.
+	 */
+	ptep = lookup_address(va, &level);
+	BUG_ON(ptep == NULL);
 
-		frames[f] = mfn;
+	pfn = pte_pfn(*ptep);
+	mfn = pfn_to_mfn(pfn);
+	virt = __va(PFN_PHYS(pfn));
 
-		make_lowmem_page_readonly((void *)va);
-		make_lowmem_page_readonly(virt);
-	}
+	make_lowmem_page_readonly((void *)va);
+	make_lowmem_page_readonly(virt);
 
-	if (HYPERVISOR_set_gdt(frames, size / sizeof(struct desc_struct)))
+	if (HYPERVISOR_set_gdt(&mfn, size / sizeof(struct desc_struct)))
 		BUG();
 }
 
@@ -470,34 +458,22 @@ static void __init xen_load_gdt_boot(const struct desc_ptr *dtr)
 {
 	unsigned long va = dtr->address;
 	unsigned int size = dtr->size + 1;
-	unsigned pages = DIV_ROUND_UP(size, PAGE_SIZE);
-	unsigned long frames[pages];
-	int f;
-
-	/*
-	 * A GDT can be up to 64k in size, which corresponds to 8192
-	 * 8-byte entries, or 16 4k pages..
-	 */
+	unsigned long pfn, mfn;
+	pte_t pte;
 
-	BUG_ON(size > 65536);
+	/* @size should be at most GDT_SIZE which is smaller than PAGE_SIZE. */
+	BUG_ON(size > PAGE_SIZE);
 	BUG_ON(va & ~PAGE_MASK);
 
-	for (f = 0; va < dtr->address + size; va += PAGE_SIZE, f++) {
-		pte_t pte;
-		unsigned long pfn, mfn;
+	pfn = virt_to_pfn(va);
+	mfn = pfn_to_mfn(pfn);
 
-		pfn = virt_to_pfn(va);
-		mfn = pfn_to_mfn(pfn);
+	pte = pfn_pte(pfn, PAGE_KERNEL_RO);
 
-		pte = pfn_pte(pfn, PAGE_KERNEL_RO);
-
-		if (HYPERVISOR_update_va_mapping((unsigned long)va, pte, 0))
-			BUG();
-
-		frames[f] = mfn;
-	}
+	if (HYPERVISOR_update_va_mapping((unsigned long)va, pte, 0))
+		BUG();
 
-	if (HYPERVISOR_set_gdt(frames, size / sizeof(struct desc_struct)))
+	if (HYPERVISOR_set_gdt(&mfn, size / sizeof(struct desc_struct)))
 		BUG();
 }
 
-- 
cgit v1.1


From 8b77586bd8fe600d97f922c79f7222c46f37c118 Mon Sep 17 00:00:00 2001
From: Jason Gunthorpe <jgg@mellanox.com>
Date: Wed, 4 Apr 2018 21:00:01 -0600
Subject: RDMA/ucma: Check for a cm_id->device in all user calls that need it

This is done by auditing all callers of ucma_get_ctx and switching the
ones that unconditionally touch ->device to ucma_get_ctx_dev. This covers
a little less than  half of the call sites.

The 11 remaining call sites to ucma_get_ctx() were manually audited.

Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Reviewed-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/ucma.c | 36 ++++++++++++++++++++++++------------
 1 file changed, 24 insertions(+), 12 deletions(-)

diff --git a/drivers/infiniband/core/ucma.c b/drivers/infiniband/core/ucma.c
index 680b353..0efa0e2 100644
--- a/drivers/infiniband/core/ucma.c
+++ b/drivers/infiniband/core/ucma.c
@@ -159,6 +159,23 @@ static void ucma_put_ctx(struct ucma_context *ctx)
 		complete(&ctx->comp);
 }
 
+/*
+ * Same as ucm_get_ctx but requires that ->cm_id->device is valid, eg that the
+ * CM_ID is bound.
+ */
+static struct ucma_context *ucma_get_ctx_dev(struct ucma_file *file, int id)
+{
+	struct ucma_context *ctx = ucma_get_ctx(file, id);
+
+	if (IS_ERR(ctx))
+		return ctx;
+	if (!ctx->cm_id->device) {
+		ucma_put_ctx(ctx);
+		return ERR_PTR(-EINVAL);
+	}
+	return ctx;
+}
+
 static void ucma_close_event_id(struct work_struct *work)
 {
 	struct ucma_event *uevent_close =  container_of(work, struct ucma_event, close_work);
@@ -734,7 +751,7 @@ static ssize_t ucma_resolve_route(struct ucma_file *file,
 	if (copy_from_user(&cmd, inbuf, sizeof(cmd)))
 		return -EFAULT;
 
-	ctx = ucma_get_ctx(file, cmd.id);
+	ctx = ucma_get_ctx_dev(file, cmd.id);
 	if (IS_ERR(ctx))
 		return PTR_ERR(ctx);
 
@@ -1050,7 +1067,7 @@ static ssize_t ucma_connect(struct ucma_file *file, const char __user *inbuf,
 	if (!cmd.conn_param.valid)
 		return -EINVAL;
 
-	ctx = ucma_get_ctx(file, cmd.id);
+	ctx = ucma_get_ctx_dev(file, cmd.id);
 	if (IS_ERR(ctx))
 		return PTR_ERR(ctx);
 
@@ -1092,7 +1109,7 @@ static ssize_t ucma_accept(struct ucma_file *file, const char __user *inbuf,
 	if (copy_from_user(&cmd, inbuf, sizeof(cmd)))
 		return -EFAULT;
 
-	ctx = ucma_get_ctx(file, cmd.id);
+	ctx = ucma_get_ctx_dev(file, cmd.id);
 	if (IS_ERR(ctx))
 		return PTR_ERR(ctx);
 
@@ -1120,7 +1137,7 @@ static ssize_t ucma_reject(struct ucma_file *file, const char __user *inbuf,
 	if (copy_from_user(&cmd, inbuf, sizeof(cmd)))
 		return -EFAULT;
 
-	ctx = ucma_get_ctx(file, cmd.id);
+	ctx = ucma_get_ctx_dev(file, cmd.id);
 	if (IS_ERR(ctx))
 		return PTR_ERR(ctx);
 
@@ -1139,7 +1156,7 @@ static ssize_t ucma_disconnect(struct ucma_file *file, const char __user *inbuf,
 	if (copy_from_user(&cmd, inbuf, sizeof(cmd)))
 		return -EFAULT;
 
-	ctx = ucma_get_ctx(file, cmd.id);
+	ctx = ucma_get_ctx_dev(file, cmd.id);
 	if (IS_ERR(ctx))
 		return PTR_ERR(ctx);
 
@@ -1167,15 +1184,10 @@ static ssize_t ucma_init_qp_attr(struct ucma_file *file,
 	if (cmd.qp_state > IB_QPS_ERR)
 		return -EINVAL;
 
-	ctx = ucma_get_ctx(file, cmd.id);
+	ctx = ucma_get_ctx_dev(file, cmd.id);
 	if (IS_ERR(ctx))
 		return PTR_ERR(ctx);
 
-	if (!ctx->cm_id->device) {
-		ret = -EINVAL;
-		goto out;
-	}
-
 	resp.qp_attr_mask = 0;
 	memset(&qp_attr, 0, sizeof qp_attr);
 	qp_attr.qp_state = cmd.qp_state;
@@ -1384,7 +1396,7 @@ static ssize_t ucma_process_join(struct ucma_file *file,
 	else
 		return -EINVAL;
 
-	ctx = ucma_get_ctx(file, cmd->id);
+	ctx = ucma_get_ctx_dev(file, cmd->id);
 	if (IS_ERR(ctx))
 		return PTR_ERR(ctx);
 
-- 
cgit v1.1


From 2f0605a697f4b9f5b1c1571c7ec6a16df4dc2616 Mon Sep 17 00:00:00 2001
From: Srinivas Dasari <dasaris@codeaurora.org>
Date: Fri, 20 Apr 2018 11:41:14 +0530
Subject: nl80211: Free connkeys on external authentication failure

The failure scenario while processing
NL80211_ATTR_EXTERNAL_AUTH_SUPPORT does not free
the connkeys. This commit addresses the same.

Signed-off-by: Srinivas Dasari <dasaris@codeaurora.org>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
---
 net/wireless/nl80211.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c
index ff28f8f..a052693 100644
--- a/net/wireless/nl80211.c
+++ b/net/wireless/nl80211.c
@@ -9214,6 +9214,7 @@ static int nl80211_connect(struct sk_buff *skb, struct genl_info *info)
 
 	if (nla_get_flag(info->attrs[NL80211_ATTR_EXTERNAL_AUTH_SUPPORT])) {
 		if (!info->attrs[NL80211_ATTR_SOCKET_OWNER]) {
+			kzfree(connkeys);
 			GENL_SET_ERR_MSG(info,
 					 "external auth requires connection ownership");
 			return -EINVAL;
-- 
cgit v1.1


From b49be6622f08187129561cff0409f7b06b33de57 Mon Sep 17 00:00:00 2001
From: Imre Deak <imre.deak@intel.com>
Date: Thu, 19 Apr 2018 18:51:09 +0300
Subject: drm/i915: Enable display WA#1183 from its correct spot
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The DMC FW specific part of display WA#1183 is supposed to be enabled
whenever enabling DC5 or DC6, so move it to the DC6 enable function
from the DC6 disable function.

I noticed this after Daniel's patch to remove the unused
skl_disable_dc6() function.

Fixes: 53421c2fe99c ("drm/i915: Apply Display WA #1183 on skl, kbl, and cfl")
Cc: Lucas De Marchi <lucas.demarchi@intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: <stable@vger.kernel.org>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180419155109.29451-1-imre.deak@intel.com
---
 drivers/gpu/drm/i915/intel_runtime_pm.c | 11 +++++------
 1 file changed, 5 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_runtime_pm.c b/drivers/gpu/drm/i915/intel_runtime_pm.c
index 53ea564..66de4b2 100644
--- a/drivers/gpu/drm/i915/intel_runtime_pm.c
+++ b/drivers/gpu/drm/i915/intel_runtime_pm.c
@@ -641,19 +641,18 @@ void skl_enable_dc6(struct drm_i915_private *dev_priv)
 
 	DRM_DEBUG_KMS("Enabling DC6\n");
 
-	gen9_set_dc_state(dev_priv, DC_STATE_EN_UPTO_DC6);
+	/* Wa Display #1183: skl,kbl,cfl */
+	if (IS_GEN9_BC(dev_priv))
+		I915_WRITE(GEN8_CHICKEN_DCPR_1, I915_READ(GEN8_CHICKEN_DCPR_1) |
+			   SKL_SELECT_ALTERNATE_DC_EXIT);
 
+	gen9_set_dc_state(dev_priv, DC_STATE_EN_UPTO_DC6);
 }
 
 void skl_disable_dc6(struct drm_i915_private *dev_priv)
 {
 	DRM_DEBUG_KMS("Disabling DC6\n");
 
-	/* Wa Display #1183: skl,kbl,cfl */
-	if (IS_GEN9_BC(dev_priv))
-		I915_WRITE(GEN8_CHICKEN_DCPR_1, I915_READ(GEN8_CHICKEN_DCPR_1) |
-			   SKL_SELECT_ALTERNATE_DC_EXIT);
-
 	gen9_set_dc_state(dev_priv, DC_STATE_DISABLE);
 }
 
-- 
cgit v1.1


From c4c252590951704947d216a2565ee9dec21f704d Mon Sep 17 00:00:00 2001
From: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Tue, 17 Apr 2018 12:02:25 +0200
Subject: drm/i915: Remove skl dc6 enable/disable functions

One is outright unused, other can be made static.

Drive-by cleanup while accidentally reading dc code.

Cc: Imre Deak <imre.deak@intel.com>
Acked-by: Imre Deak <imre.deak@intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180417100225.12286-1-daniel.vetter@ffwll.ch
---
 drivers/gpu/drm/i915/intel_drv.h        | 2 --
 drivers/gpu/drm/i915/intel_runtime_pm.c | 9 +--------
 2 files changed, 1 insertion(+), 10 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
index 5bd2263..8b20824 100644
--- a/drivers/gpu/drm/i915/intel_drv.h
+++ b/drivers/gpu/drm/i915/intel_drv.h
@@ -1580,8 +1580,6 @@ void bxt_enable_dc9(struct drm_i915_private *dev_priv);
 void bxt_disable_dc9(struct drm_i915_private *dev_priv);
 void gen9_enable_dc5(struct drm_i915_private *dev_priv);
 unsigned int skl_cdclk_get_vco(unsigned int freq);
-void skl_enable_dc6(struct drm_i915_private *dev_priv);
-void skl_disable_dc6(struct drm_i915_private *dev_priv);
 void intel_dp_get_m_n(struct intel_crtc *crtc,
 		      struct intel_crtc_state *pipe_config);
 void intel_dp_set_m_n(struct intel_crtc *crtc, enum link_m_n_set m_n);
diff --git a/drivers/gpu/drm/i915/intel_runtime_pm.c b/drivers/gpu/drm/i915/intel_runtime_pm.c
index 66de4b2..ec59992 100644
--- a/drivers/gpu/drm/i915/intel_runtime_pm.c
+++ b/drivers/gpu/drm/i915/intel_runtime_pm.c
@@ -635,7 +635,7 @@ static void assert_can_enable_dc6(struct drm_i915_private *dev_priv)
 	assert_csr_loaded(dev_priv);
 }
 
-void skl_enable_dc6(struct drm_i915_private *dev_priv)
+static void skl_enable_dc6(struct drm_i915_private *dev_priv)
 {
 	assert_can_enable_dc6(dev_priv);
 
@@ -649,13 +649,6 @@ void skl_enable_dc6(struct drm_i915_private *dev_priv)
 	gen9_set_dc_state(dev_priv, DC_STATE_EN_UPTO_DC6);
 }
 
-void skl_disable_dc6(struct drm_i915_private *dev_priv)
-{
-	DRM_DEBUG_KMS("Disabling DC6\n");
-
-	gen9_set_dc_state(dev_priv, DC_STATE_DISABLE);
-}
-
 static void hsw_power_well_sync_hw(struct drm_i915_private *dev_priv,
 				   struct i915_power_well *power_well)
 {
-- 
cgit v1.1


From f76cdd00ef0e39d880139b074e3b247594dff95a Mon Sep 17 00:00:00 2001
From: Baolin Wang <baolin.wang@linaro.org>
Date: Thu, 19 Apr 2018 14:51:03 +0800
Subject: parisc: time: Convert read_persistent_clock() to
 read_persistent_clock64()

The read_persistent_clock() uses a timespec, which is not year 2038 safe
on 32bit systems. On parisc architecture, we have implemented generic
RTC drivers that can be used to compensate the system suspend time, but
the RTC time can not represent the nanosecond resolution, so this patch
just converts to read_persistent_clock64() with timespec64.

Signed-off-by: Baolin Wang <baolin.wang@linaro.org>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Helge Deller <deller@gmx.de>
---
 arch/parisc/kernel/time.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/parisc/kernel/time.c b/arch/parisc/kernel/time.c
index c383040..a1e772f 100644
--- a/arch/parisc/kernel/time.c
+++ b/arch/parisc/kernel/time.c
@@ -205,7 +205,7 @@ static int __init rtc_init(void)
 device_initcall(rtc_init);
 #endif
 
-void read_persistent_clock(struct timespec *ts)
+void read_persistent_clock64(struct timespec64 *ts)
 {
 	static struct pdc_tod tod_data;
 	if (pdc_tod_read(&tod_data) == 0) {
-- 
cgit v1.1


From f286299c1d0ba5e2ca0377610307b370fe178767 Mon Sep 17 00:00:00 2001
From: Long Li <longli@microsoft.com>
Date: Thu, 22 Mar 2018 14:47:18 -0700
Subject: scsi: storvsc: Set up correct queue depth values for IDE devices

Unlike SCSI and FC, we don't use multiple channels for IDE.  Also fix
the calculation for sub-channels.

Signed-off-by: Long Li <longli@microsoft.com>
Reviewed-by: Michael Kelley <mikelley@microsoft.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
---
 drivers/scsi/storvsc_drv.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/drivers/scsi/storvsc_drv.c b/drivers/scsi/storvsc_drv.c
index 8c51d62..a2ec0bc 100644
--- a/drivers/scsi/storvsc_drv.c
+++ b/drivers/scsi/storvsc_drv.c
@@ -1722,11 +1722,14 @@ static int storvsc_probe(struct hv_device *device,
 		max_targets = STORVSC_MAX_TARGETS;
 		max_channels = STORVSC_MAX_CHANNELS;
 		/*
-		 * On Windows8 and above, we support sub-channels for storage.
+		 * On Windows8 and above, we support sub-channels for storage
+		 * on SCSI and FC controllers.
 		 * The number of sub-channels offerred is based on the number of
 		 * VCPUs in the guest.
 		 */
-		max_sub_channels = (num_cpus / storvsc_vcpus_per_sub_channel);
+		if (!dev_is_ide)
+			max_sub_channels =
+				(num_cpus - 1) / storvsc_vcpus_per_sub_channel;
 	}
 
 	scsi_driver.can_queue = (max_outstanding_req_per_channel *
-- 
cgit v1.1


From fc34044248b611ea3f8b6e55b4ed404192a4f295 Mon Sep 17 00:00:00 2001
From: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Thu, 5 Apr 2018 15:00:23 -0700
Subject: drm/i915: Enable edp psr error interrupts on hsw
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The definitions for the error register should be valid on bdw/skl too,
but there we haven't even enabled DE_MISC handling yet.

Somewhat confusing the the moved register offset on bdw is only for
the _CTL/_AUX register, and that _IIR/IMR stayed where they have been
on bdw.

v2: Fixes from Ville.

v3: From DK
 * Rebased on drm-tip
 * Removed BDW IIR bit definition, looks like an unintentional change that
should be in the following patch.

v4: From DK
 * Don't mask REG_WRITE.

References: bspec/11974 [SRD Interrupt Bit Definition DevHSW]
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Daniel Vetter <daniel.vetter@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Signed-off-by: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Reviewed-by: Jose Roberto de Souza <jose.souza@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180405220023.9449-1-dhinakaran.pandiyan@intel.com
---
 drivers/gpu/drm/i915/i915_irq.c | 34 ++++++++++++++++++++++++++++++++++
 drivers/gpu/drm/i915/i915_reg.h |  8 ++++++++
 2 files changed, 42 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index b03d185..630fc6f 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -2452,6 +2452,26 @@ static void ilk_display_irq_handler(struct drm_i915_private *dev_priv,
 		ironlake_rps_change_irq_handler(dev_priv);
 }
 
+static void hsw_edp_psr_irq_handler(struct drm_i915_private *dev_priv)
+{
+	u32 edp_psr_iir = I915_READ(EDP_PSR_IIR);
+
+	if (edp_psr_iir & EDP_PSR_ERROR)
+		DRM_DEBUG_KMS("PSR error\n");
+
+	if (edp_psr_iir & EDP_PSR_PRE_ENTRY) {
+		DRM_DEBUG_KMS("PSR prepare entry in 2 vblanks\n");
+		I915_WRITE(EDP_PSR_IMR, EDP_PSR_PRE_ENTRY);
+	}
+
+	if (edp_psr_iir & EDP_PSR_POST_EXIT) {
+		DRM_DEBUG_KMS("PSR exit completed\n");
+		I915_WRITE(EDP_PSR_IMR, 0);
+	}
+
+	I915_WRITE(EDP_PSR_IIR, edp_psr_iir);
+}
+
 static void ivb_display_irq_handler(struct drm_i915_private *dev_priv,
 				    u32 de_iir)
 {
@@ -2464,6 +2484,9 @@ static void ivb_display_irq_handler(struct drm_i915_private *dev_priv,
 	if (de_iir & DE_ERR_INT_IVB)
 		ivb_err_int_handler(dev_priv);
 
+	if (de_iir & DE_EDP_PSR_INT_HSW)
+		hsw_edp_psr_irq_handler(dev_priv);
+
 	if (de_iir & DE_AUX_CHANNEL_A_IVB)
 		dp_aux_irq_handler(dev_priv);
 
@@ -3348,6 +3371,11 @@ static void ironlake_irq_reset(struct drm_device *dev)
 	if (IS_GEN7(dev_priv))
 		I915_WRITE(GEN7_ERR_INT, 0xffffffff);
 
+	if (IS_HASWELL(dev_priv)) {
+		I915_WRITE(EDP_PSR_IMR, 0xffffffff);
+		I915_WRITE(EDP_PSR_IIR, 0xffffffff);
+	}
+
 	gen5_gt_irq_reset(dev_priv);
 
 	ibx_irq_reset(dev_priv);
@@ -3762,6 +3790,12 @@ static int ironlake_irq_postinstall(struct drm_device *dev)
 			      DE_DP_A_HOTPLUG);
 	}
 
+	if (IS_HASWELL(dev_priv)) {
+		gen3_assert_iir_is_zero(dev_priv, EDP_PSR_IIR);
+		I915_WRITE(EDP_PSR_IMR, 0);
+		display_mask |= DE_EDP_PSR_INT_HSW;
+	}
+
 	dev_priv->irq_mask = ~display_mask;
 
 	ibx_irq_pre_postinstall(dev);
diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index fb10602..9f61d38 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -4026,6 +4026,13 @@ enum {
 #define   EDP_PSR_TP1_TIME_0us			(3<<4)
 #define   EDP_PSR_IDLE_FRAME_SHIFT		0
 
+/* Bspec claims those aren't shifted but stay at 0x64800 */
+#define EDP_PSR_IMR				_MMIO(0x64834)
+#define EDP_PSR_IIR				_MMIO(0x64838)
+#define   EDP_PSR_ERROR				(1<<2)
+#define   EDP_PSR_POST_EXIT			(1<<1)
+#define   EDP_PSR_PRE_ENTRY			(1<<0)
+
 #define EDP_PSR_AUX_CTL				_MMIO(dev_priv->psr_mmio_base + 0x10)
 #define   EDP_PSR_AUX_CTL_TIME_OUT_MASK		(3 << 26)
 #define   EDP_PSR_AUX_CTL_MESSAGE_SIZE_MASK	(0x1f << 20)
@@ -6837,6 +6844,7 @@ enum {
 #define DE_PCH_EVENT_IVB		(1<<28)
 #define DE_DP_A_HOTPLUG_IVB		(1<<27)
 #define DE_AUX_CHANNEL_A_IVB		(1<<26)
+#define DE_EDP_PSR_INT_HSW		(1<<19)
 #define DE_SPRITEC_FLIP_DONE_IVB	(1<<14)
 #define DE_PLANEC_FLIP_DONE_IVB		(1<<13)
 #define DE_PIPEC_VBLANK_IVB		(1<<10)
-- 
cgit v1.1


From e04f7ece1c4530b4f0db182b5596fadf48628f22 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= <ville.syrjala@linux.intel.com>
Date: Tue, 3 Apr 2018 14:24:18 -0700
Subject: drm/i915: Enable edp psr error interrupts on bdw+
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Plug in the bdw+ irq handling for PSR interrupts. bdw+ supports psr on
any transcoder in theory, though the we don't currenty enable PSR except
on the EDP transcoder.

v2: From DK
 * Rebased on drm-tip
v3: Switched author to Ville based on IRC discussion.

Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Reviewed-by: Jose Roberto de Souza <jose.souza@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180403212420.25007-2-dhinakaran.pandiyan@intel.com
---
 drivers/gpu/drm/i915/i915_irq.c      | 57 ++++++++++++++++++++++++++++--------
 drivers/gpu/drm/i915/i915_reg.h      |  7 +++--
 drivers/gpu/drm/i915/intel_display.h |  4 +++
 3 files changed, 52 insertions(+), 16 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index 630fc6f..ab9aac8 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -2455,20 +2455,34 @@ static void ilk_display_irq_handler(struct drm_i915_private *dev_priv,
 static void hsw_edp_psr_irq_handler(struct drm_i915_private *dev_priv)
 {
 	u32 edp_psr_iir = I915_READ(EDP_PSR_IIR);
+	u32 edp_psr_imr = I915_READ(EDP_PSR_IMR);
+	u32 mask = BIT(TRANSCODER_EDP);
+	enum transcoder cpu_transcoder;
 
-	if (edp_psr_iir & EDP_PSR_ERROR)
-		DRM_DEBUG_KMS("PSR error\n");
-
-	if (edp_psr_iir & EDP_PSR_PRE_ENTRY) {
-		DRM_DEBUG_KMS("PSR prepare entry in 2 vblanks\n");
-		I915_WRITE(EDP_PSR_IMR, EDP_PSR_PRE_ENTRY);
-	}
+	if (INTEL_GEN(dev_priv) >= 8)
+		mask |= BIT(TRANSCODER_A) |
+			BIT(TRANSCODER_B) |
+			BIT(TRANSCODER_C);
+
+	for_each_cpu_transcoder_masked(dev_priv, cpu_transcoder, mask) {
+		if (edp_psr_iir & EDP_PSR_ERROR(cpu_transcoder))
+			DRM_DEBUG_KMS("Transcoder %s PSR error\n",
+				      transcoder_name(cpu_transcoder));
+
+		if (edp_psr_iir & EDP_PSR_PRE_ENTRY(cpu_transcoder)) {
+			DRM_DEBUG_KMS("Transcoder %s PSR prepare entry in 2 vblanks\n",
+				      transcoder_name(cpu_transcoder));
+			edp_psr_imr |= EDP_PSR_PRE_ENTRY(cpu_transcoder);
+		}
 
-	if (edp_psr_iir & EDP_PSR_POST_EXIT) {
-		DRM_DEBUG_KMS("PSR exit completed\n");
-		I915_WRITE(EDP_PSR_IMR, 0);
+		if (edp_psr_iir & EDP_PSR_POST_EXIT(cpu_transcoder)) {
+			DRM_DEBUG_KMS("Transcoder %s PSR exit completed\n",
+				      transcoder_name(cpu_transcoder));
+			edp_psr_imr &= ~EDP_PSR_PRE_ENTRY(cpu_transcoder);
+		}
 	}
 
+	I915_WRITE(EDP_PSR_IMR, edp_psr_imr);
 	I915_WRITE(EDP_PSR_IIR, edp_psr_iir);
 }
 
@@ -2616,11 +2630,22 @@ gen8_de_irq_handler(struct drm_i915_private *dev_priv, u32 master_ctl)
 	if (master_ctl & GEN8_DE_MISC_IRQ) {
 		iir = I915_READ(GEN8_DE_MISC_IIR);
 		if (iir) {
+			bool found = false;
+
 			I915_WRITE(GEN8_DE_MISC_IIR, iir);
 			ret = IRQ_HANDLED;
-			if (iir & GEN8_DE_MISC_GSE)
+
+			if (iir & GEN8_DE_MISC_GSE) {
 				intel_opregion_asle_intr(dev_priv);
-			else
+				found = true;
+			}
+
+			if (iir & GEN8_DE_EDP_PSR) {
+				hsw_edp_psr_irq_handler(dev_priv);
+				found = true;
+			}
+
+			if (!found)
 				DRM_ERROR("Unexpected DE Misc interrupt\n");
 		}
 		else
@@ -3414,6 +3439,9 @@ static void gen8_irq_reset(struct drm_device *dev)
 
 	gen8_gt_irq_reset(dev_priv);
 
+	I915_WRITE(EDP_PSR_IMR, 0xffffffff);
+	I915_WRITE(EDP_PSR_IIR, 0xffffffff);
+
 	for_each_pipe(dev_priv, pipe)
 		if (intel_display_power_is_enabled(dev_priv,
 						   POWER_DOMAIN_PIPE(pipe)))
@@ -3906,7 +3934,7 @@ static void gen8_de_irq_postinstall(struct drm_i915_private *dev_priv)
 	uint32_t de_pipe_enables;
 	u32 de_port_masked = GEN8_AUX_CHANNEL_A;
 	u32 de_port_enables;
-	u32 de_misc_masked = GEN8_DE_MISC_GSE;
+	u32 de_misc_masked = GEN8_DE_MISC_GSE | GEN8_DE_EDP_PSR;
 	enum pipe pipe;
 
 	if (INTEL_GEN(dev_priv) >= 9) {
@@ -3931,6 +3959,9 @@ static void gen8_de_irq_postinstall(struct drm_i915_private *dev_priv)
 	else if (IS_BROADWELL(dev_priv))
 		de_port_enables |= GEN8_PORT_DP_A_HOTPLUG;
 
+	gen3_assert_iir_is_zero(dev_priv, EDP_PSR_IIR);
+	I915_WRITE(EDP_PSR_IMR, 0);
+
 	for_each_pipe(dev_priv, pipe) {
 		dev_priv->de_irq_mask[pipe] = ~de_pipe_masked;
 
diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 9f61d38..2dad655 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -4029,9 +4029,9 @@ enum {
 /* Bspec claims those aren't shifted but stay at 0x64800 */
 #define EDP_PSR_IMR				_MMIO(0x64834)
 #define EDP_PSR_IIR				_MMIO(0x64838)
-#define   EDP_PSR_ERROR				(1<<2)
-#define   EDP_PSR_POST_EXIT			(1<<1)
-#define   EDP_PSR_PRE_ENTRY			(1<<0)
+#define   EDP_PSR_ERROR(trans)			(1 << (((trans) * 8 + 10) & 31))
+#define   EDP_PSR_POST_EXIT(trans)		(1 << (((trans) * 8 + 9) & 31))
+#define   EDP_PSR_PRE_ENTRY(trans)		(1 << (((trans) * 8 + 8) & 31))
 
 #define EDP_PSR_AUX_CTL				_MMIO(dev_priv->psr_mmio_base + 0x10)
 #define   EDP_PSR_AUX_CTL_TIME_OUT_MASK		(3 << 26)
@@ -6969,6 +6969,7 @@ enum {
 #define GEN8_DE_MISC_IIR _MMIO(0x44468)
 #define GEN8_DE_MISC_IER _MMIO(0x4446c)
 #define  GEN8_DE_MISC_GSE		(1 << 27)
+#define  GEN8_DE_EDP_PSR		(1 << 19)
 
 #define GEN8_PCU_ISR _MMIO(0x444e0)
 #define GEN8_PCU_IMR _MMIO(0x444e4)
diff --git a/drivers/gpu/drm/i915/intel_display.h b/drivers/gpu/drm/i915/intel_display.h
index 4e7418b..2ef3161 100644
--- a/drivers/gpu/drm/i915/intel_display.h
+++ b/drivers/gpu/drm/i915/intel_display.h
@@ -218,6 +218,10 @@ struct intel_link_m_n {
 	for ((__p) = 0; (__p) < INTEL_INFO(__dev_priv)->num_pipes; (__p)++) \
 		for_each_if((__mask) & BIT(__p))
 
+#define for_each_cpu_transcoder_masked(__dev_priv, __t, __mask) \
+	for ((__t) = 0; (__t) < I915_MAX_TRANSCODERS; (__t)++)	\
+		for_each_if ((__mask) & (1 << (__t)))
+
 #define for_each_universal_plane(__dev_priv, __pipe, __p)		\
 	for ((__p) = 0;							\
 	     (__p) < INTEL_INFO(__dev_priv)->num_sprites[(__pipe)] + 1;	\
-- 
cgit v1.1


From 54fd3149598cc2f74cf0708d614470da2331a374 Mon Sep 17 00:00:00 2001
From: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Date: Wed, 4 Apr 2018 18:37:17 -0700
Subject: drm/i915/psr: Control PSR interrupts via debugfs
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Interrupts other than the one for AUX errors are required only for debug,
so unmask them via debugfs when the user requests debug.

User can make such a request with
echo 1 > <DEBUG_FS>/dri/0/i915_edp_psr_debug

There are no locks to serialize PSR debug enabling from
irq_postinstall() and debugfs for simplicity. As irq_postinstall() is
called only during module initialization/resume and IGT subtests
aren't expected to modify PSR debug at those times, we should be safe.

v2: Unroll loops (Ville)
    Avoid resetting error mask bits.

v3: Unmask interrupts in postinstall() if debug was still enabled.
    Avoid RMW (Ville)

v4: Avoid extra IMR write introduced in the previous version.(Jose)
    Style changes, renames (Jose).

Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Reviewed-by: Jose Roberto de Souza <jose.souza@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180405013717.24254-1-dhinakaran.pandiyan@intel.com
---
 drivers/gpu/drm/i915/i915_debugfs.c | 36 ++++++++++++++++++++++-
 drivers/gpu/drm/i915/i915_drv.h     |  1 +
 drivers/gpu/drm/i915/i915_irq.c     | 51 ++++++++------------------------
 drivers/gpu/drm/i915/intel_drv.h    |  2 ++
 drivers/gpu/drm/i915/intel_psr.c    | 58 +++++++++++++++++++++++++++++++++++++
 5 files changed, 108 insertions(+), 40 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index e0274f4..a6c70ff 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -2690,6 +2690,39 @@ static int i915_edp_psr_status(struct seq_file *m, void *data)
 	return 0;
 }
 
+static int
+i915_edp_psr_debug_set(void *data, u64 val)
+{
+	struct drm_i915_private *dev_priv = data;
+
+	if (!CAN_PSR(dev_priv))
+		return -ENODEV;
+
+	DRM_DEBUG_KMS("PSR debug %s\n", enableddisabled(val));
+
+	intel_runtime_pm_get(dev_priv);
+	intel_psr_irq_control(dev_priv, !!val);
+	intel_runtime_pm_put(dev_priv);
+
+	return 0;
+}
+
+static int
+i915_edp_psr_debug_get(void *data, u64 *val)
+{
+	struct drm_i915_private *dev_priv = data;
+
+	if (!CAN_PSR(dev_priv))
+		return -ENODEV;
+
+	*val = READ_ONCE(dev_priv->psr.debug);
+	return 0;
+}
+
+DEFINE_SIMPLE_ATTRIBUTE(i915_edp_psr_debug_fops,
+			i915_edp_psr_debug_get, i915_edp_psr_debug_set,
+			"%llu\n");
+
 static int i915_sink_crc(struct seq_file *m, void *data)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
@@ -4862,7 +4895,8 @@ static const struct i915_debugfs_files {
 	{"i915_guc_log_relay", &i915_guc_log_relay_fops},
 	{"i915_hpd_storm_ctl", &i915_hpd_storm_ctl_fops},
 	{"i915_ipc_status", &i915_ipc_status_fops},
-	{"i915_drrs_ctl", &i915_drrs_ctl_fops}
+	{"i915_drrs_ctl", &i915_drrs_ctl_fops},
+	{"i915_edp_psr_debug", &i915_edp_psr_debug_fops}
 };
 
 int i915_debugfs_register(struct drm_i915_private *dev_priv)
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 0286911..e67e217 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -610,6 +610,7 @@ struct i915_psr {
 	bool has_hw_tracking;
 	bool psr2_enabled;
 	u8 sink_sync_latency;
+	bool debug;
 
 	void (*enable_source)(struct intel_dp *,
 			      const struct intel_crtc_state *);
diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index ab9aac8..96547e0 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -2452,40 +2452,6 @@ static void ilk_display_irq_handler(struct drm_i915_private *dev_priv,
 		ironlake_rps_change_irq_handler(dev_priv);
 }
 
-static void hsw_edp_psr_irq_handler(struct drm_i915_private *dev_priv)
-{
-	u32 edp_psr_iir = I915_READ(EDP_PSR_IIR);
-	u32 edp_psr_imr = I915_READ(EDP_PSR_IMR);
-	u32 mask = BIT(TRANSCODER_EDP);
-	enum transcoder cpu_transcoder;
-
-	if (INTEL_GEN(dev_priv) >= 8)
-		mask |= BIT(TRANSCODER_A) |
-			BIT(TRANSCODER_B) |
-			BIT(TRANSCODER_C);
-
-	for_each_cpu_transcoder_masked(dev_priv, cpu_transcoder, mask) {
-		if (edp_psr_iir & EDP_PSR_ERROR(cpu_transcoder))
-			DRM_DEBUG_KMS("Transcoder %s PSR error\n",
-				      transcoder_name(cpu_transcoder));
-
-		if (edp_psr_iir & EDP_PSR_PRE_ENTRY(cpu_transcoder)) {
-			DRM_DEBUG_KMS("Transcoder %s PSR prepare entry in 2 vblanks\n",
-				      transcoder_name(cpu_transcoder));
-			edp_psr_imr |= EDP_PSR_PRE_ENTRY(cpu_transcoder);
-		}
-
-		if (edp_psr_iir & EDP_PSR_POST_EXIT(cpu_transcoder)) {
-			DRM_DEBUG_KMS("Transcoder %s PSR exit completed\n",
-				      transcoder_name(cpu_transcoder));
-			edp_psr_imr &= ~EDP_PSR_PRE_ENTRY(cpu_transcoder);
-		}
-	}
-
-	I915_WRITE(EDP_PSR_IMR, edp_psr_imr);
-	I915_WRITE(EDP_PSR_IIR, edp_psr_iir);
-}
-
 static void ivb_display_irq_handler(struct drm_i915_private *dev_priv,
 				    u32 de_iir)
 {
@@ -2498,8 +2464,12 @@ static void ivb_display_irq_handler(struct drm_i915_private *dev_priv,
 	if (de_iir & DE_ERR_INT_IVB)
 		ivb_err_int_handler(dev_priv);
 
-	if (de_iir & DE_EDP_PSR_INT_HSW)
-		hsw_edp_psr_irq_handler(dev_priv);
+	if (de_iir & DE_EDP_PSR_INT_HSW) {
+		u32 psr_iir = I915_READ(EDP_PSR_IIR);
+
+		intel_psr_irq_handler(dev_priv, psr_iir);
+		I915_WRITE(EDP_PSR_IIR, psr_iir);
+	}
 
 	if (de_iir & DE_AUX_CHANNEL_A_IVB)
 		dp_aux_irq_handler(dev_priv);
@@ -2641,7 +2611,10 @@ gen8_de_irq_handler(struct drm_i915_private *dev_priv, u32 master_ctl)
 			}
 
 			if (iir & GEN8_DE_EDP_PSR) {
-				hsw_edp_psr_irq_handler(dev_priv);
+				u32 psr_iir = I915_READ(EDP_PSR_IIR);
+
+				intel_psr_irq_handler(dev_priv, psr_iir);
+				I915_WRITE(EDP_PSR_IIR, psr_iir);
 				found = true;
 			}
 
@@ -3820,7 +3793,7 @@ static int ironlake_irq_postinstall(struct drm_device *dev)
 
 	if (IS_HASWELL(dev_priv)) {
 		gen3_assert_iir_is_zero(dev_priv, EDP_PSR_IIR);
-		I915_WRITE(EDP_PSR_IMR, 0);
+		intel_psr_irq_control(dev_priv, dev_priv->psr.debug);
 		display_mask |= DE_EDP_PSR_INT_HSW;
 	}
 
@@ -3960,7 +3933,7 @@ static void gen8_de_irq_postinstall(struct drm_i915_private *dev_priv)
 		de_port_enables |= GEN8_PORT_DP_A_HOTPLUG;
 
 	gen3_assert_iir_is_zero(dev_priv, EDP_PSR_IIR);
-	I915_WRITE(EDP_PSR_IMR, 0);
+	intel_psr_irq_control(dev_priv, dev_priv->psr.debug);
 
 	for_each_pipe(dev_priv, pipe) {
 		dev_priv->de_irq_mask[pipe] = ~de_pipe_masked;
diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
index 8b20824..44ed248 100644
--- a/drivers/gpu/drm/i915/intel_drv.h
+++ b/drivers/gpu/drm/i915/intel_drv.h
@@ -1899,6 +1899,8 @@ void intel_psr_single_frame_update(struct drm_i915_private *dev_priv,
 				   unsigned frontbuffer_bits);
 void intel_psr_compute_config(struct intel_dp *intel_dp,
 			      struct intel_crtc_state *crtc_state);
+void intel_psr_irq_control(struct drm_i915_private *dev_priv, bool debug);
+void intel_psr_irq_handler(struct drm_i915_private *dev_priv, u32 psr_iir);
 
 /* intel_runtime_pm.c */
 int intel_power_domains_init(struct drm_i915_private *);
diff --git a/drivers/gpu/drm/i915/intel_psr.c b/drivers/gpu/drm/i915/intel_psr.c
index 69a5b27..ae6a916 100644
--- a/drivers/gpu/drm/i915/intel_psr.c
+++ b/drivers/gpu/drm/i915/intel_psr.c
@@ -93,6 +93,64 @@ static void psr_aux_io_power_put(struct intel_dp *intel_dp)
 	intel_display_power_put(dev_priv, psr_aux_domain(intel_dp));
 }
 
+void intel_psr_irq_control(struct drm_i915_private *dev_priv, bool debug)
+{
+	u32 debug_mask, mask;
+
+	/* No PSR interrupts on VLV/CHV */
+	if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
+		return;
+
+	mask = EDP_PSR_ERROR(TRANSCODER_EDP);
+	debug_mask = EDP_PSR_POST_EXIT(TRANSCODER_EDP) |
+		     EDP_PSR_PRE_ENTRY(TRANSCODER_EDP);
+
+	if (INTEL_GEN(dev_priv) >= 8) {
+		mask |= EDP_PSR_ERROR(TRANSCODER_A) |
+			EDP_PSR_ERROR(TRANSCODER_B) |
+			EDP_PSR_ERROR(TRANSCODER_C);
+
+		debug_mask |= EDP_PSR_POST_EXIT(TRANSCODER_A) |
+			      EDP_PSR_PRE_ENTRY(TRANSCODER_A) |
+			      EDP_PSR_POST_EXIT(TRANSCODER_B) |
+			      EDP_PSR_PRE_ENTRY(TRANSCODER_B) |
+			      EDP_PSR_POST_EXIT(TRANSCODER_C) |
+			      EDP_PSR_PRE_ENTRY(TRANSCODER_C);
+	}
+
+	if (debug)
+		mask |= debug_mask;
+
+	WRITE_ONCE(dev_priv->psr.debug, debug);
+	I915_WRITE(EDP_PSR_IMR, ~mask);
+}
+
+void intel_psr_irq_handler(struct drm_i915_private *dev_priv, u32 psr_iir)
+{
+	u32 transcoders = BIT(TRANSCODER_EDP);
+	enum transcoder cpu_transcoder;
+
+	if (INTEL_GEN(dev_priv) >= 8)
+		transcoders |= BIT(TRANSCODER_A) |
+			       BIT(TRANSCODER_B) |
+			       BIT(TRANSCODER_C);
+
+	for_each_cpu_transcoder_masked(dev_priv, cpu_transcoder, transcoders) {
+		/* FIXME: Exit PSR and link train manually when this happens. */
+		if (psr_iir & EDP_PSR_ERROR(cpu_transcoder))
+			DRM_DEBUG_KMS("[transcoder %s] PSR aux error\n",
+				      transcoder_name(cpu_transcoder));
+
+		if (psr_iir & EDP_PSR_PRE_ENTRY(cpu_transcoder))
+			DRM_DEBUG_KMS("[transcoder %s] PSR entry attempt in 2 vblanks\n",
+				      transcoder_name(cpu_transcoder));
+
+		if (psr_iir & EDP_PSR_POST_EXIT(cpu_transcoder))
+			DRM_DEBUG_KMS("[transcoder %s] PSR exit completed\n",
+				      transcoder_name(cpu_transcoder));
+	}
+}
+
 static bool intel_dp_get_y_coord_required(struct intel_dp *intel_dp)
 {
 	uint8_t psr_caps = 0;
-- 
cgit v1.1


From 3f983e54fdad452582843b9fbc22df0eb67daacd Mon Sep 17 00:00:00 2001
From: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Date: Tue, 3 Apr 2018 14:24:20 -0700
Subject: drm/i915/psr: Timestamps for PSR entry and exit interrupts.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Timestamps are useful for IGT tests that trigger PSR exit and/or wait for
PSR entry.

v2: Removed seqlock (Ville)
    Removed erroneous warning in irq loop (Chris)

Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Reviewed-by: Jose Roberto de Souza <jose.souza@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180403212420.25007-4-dhinakaran.pandiyan@intel.com
---
 drivers/gpu/drm/i915/i915_debugfs.c | 7 +++++++
 drivers/gpu/drm/i915/i915_drv.h     | 2 ++
 drivers/gpu/drm/i915/intel_psr.c    | 9 +++++++--
 3 files changed, 16 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index a6c70ff..2f05f52 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -2686,6 +2686,13 @@ static int i915_edp_psr_status(struct seq_file *m, void *data)
 	}
 	mutex_unlock(&dev_priv->psr.lock);
 
+	if (READ_ONCE(dev_priv->psr.debug)) {
+		seq_printf(m, "Last attempted entry at: %lld\n",
+			   dev_priv->psr.last_entry_attempt);
+		seq_printf(m, "Last exit at: %lld\n",
+			   dev_priv->psr.last_exit);
+	}
+
 	intel_runtime_pm_put(dev_priv);
 	return 0;
 }
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index e67e217..8444ca8 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -611,6 +611,8 @@ struct i915_psr {
 	bool psr2_enabled;
 	u8 sink_sync_latency;
 	bool debug;
+	ktime_t last_entry_attempt;
+	ktime_t last_exit;
 
 	void (*enable_source)(struct intel_dp *,
 			      const struct intel_crtc_state *);
diff --git a/drivers/gpu/drm/i915/intel_psr.c b/drivers/gpu/drm/i915/intel_psr.c
index ae6a916..0d54829 100644
--- a/drivers/gpu/drm/i915/intel_psr.c
+++ b/drivers/gpu/drm/i915/intel_psr.c
@@ -129,6 +129,7 @@ void intel_psr_irq_handler(struct drm_i915_private *dev_priv, u32 psr_iir)
 {
 	u32 transcoders = BIT(TRANSCODER_EDP);
 	enum transcoder cpu_transcoder;
+	ktime_t time_ns =  ktime_get();
 
 	if (INTEL_GEN(dev_priv) >= 8)
 		transcoders |= BIT(TRANSCODER_A) |
@@ -141,13 +142,17 @@ void intel_psr_irq_handler(struct drm_i915_private *dev_priv, u32 psr_iir)
 			DRM_DEBUG_KMS("[transcoder %s] PSR aux error\n",
 				      transcoder_name(cpu_transcoder));
 
-		if (psr_iir & EDP_PSR_PRE_ENTRY(cpu_transcoder))
+		if (psr_iir & EDP_PSR_PRE_ENTRY(cpu_transcoder)) {
+			dev_priv->psr.last_entry_attempt = time_ns;
 			DRM_DEBUG_KMS("[transcoder %s] PSR entry attempt in 2 vblanks\n",
 				      transcoder_name(cpu_transcoder));
+		}
 
-		if (psr_iir & EDP_PSR_POST_EXIT(cpu_transcoder))
+		if (psr_iir & EDP_PSR_POST_EXIT(cpu_transcoder)) {
+			dev_priv->psr.last_exit = time_ns;
 			DRM_DEBUG_KMS("[transcoder %s] PSR exit completed\n",
 				      transcoder_name(cpu_transcoder));
+		}
 	}
 }
 
-- 
cgit v1.1


From 4bc83b3f272fe8f36450f9c003df49cf07ffe5fd Mon Sep 17 00:00:00 2001
From: Colin Ian King <colin.king@canonical.com>
Date: Fri, 20 Apr 2018 10:57:16 +0100
Subject: scsi: isci: Fix infinite loop in while loop

In the case when the phy_mask is bitwise anded with the phy_index bit is
zero the continue statement currently jumps to the next iteration of the
while loop and phy_index is never actually incremented, potentially
causing an infinite loop if phy_index is less than SCI_MAX_PHS. Fix this
by turning the while loop into a for loop.

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
---
 drivers/scsi/isci/port_config.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/scsi/isci/port_config.c b/drivers/scsi/isci/port_config.c
index edb7be7..9e8de14 100644
--- a/drivers/scsi/isci/port_config.c
+++ b/drivers/scsi/isci/port_config.c
@@ -291,7 +291,7 @@ sci_mpc_agent_validate_phy_configuration(struct isci_host *ihost,
 		 * Note: We have not moved the current phy_index so we will actually
 		 *       compare the startting phy with itself.
 		 *       This is expected and required to add the phy to the port. */
-		while (phy_index < SCI_MAX_PHYS) {
+		for (; phy_index < SCI_MAX_PHYS; phy_index++) {
 			if ((phy_mask & (1 << phy_index)) == 0)
 				continue;
 			sci_phy_get_sas_address(&ihost->phys[phy_index],
@@ -311,7 +311,6 @@ sci_mpc_agent_validate_phy_configuration(struct isci_host *ihost,
 					      &ihost->phys[phy_index]);
 
 			assigned_phy_mask |= (1 << phy_index);
-			phy_index++;
 		}
 
 	}
-- 
cgit v1.1


From f5957dade4f373b04fa1f5315a489f18cc2c4cb4 Mon Sep 17 00:00:00 2001
From: Bryant G Ly <bryantly@linux.vnet.ibm.com>
Date: Tue, 17 Apr 2018 10:33:21 -0500
Subject: scsi: target: Fix fortify_panic kernel exception

memcmp() requires the two buffers passed as arguments to be at least
'size' bytes long, otherwise a fortify_panic will trigger.

Use memchr_inv() instead of memcmp() to determine whether the received
payload is zeroed or not.

The bug was found by running a block backstore via LIO.

[  496.212958] Call Trace:
[  496.212960] [c0000007e58e3800] [c000000000cbbefc] fortify_panic+0x24/0x38 (unreliable)
[  496.212965] [c0000007e58e3860] [d00000000f150c28] iblock_execute_write_same+0x3b8/0x3c0 [target_core_iblock]
[  496.212976] [c0000007e58e3910] [d000000006c737d4] __target_execute_cmd+0x54/0x150 [target_core_mod]
[  496.212982] [c0000007e58e3940] [d000000006d32ce4] ibmvscsis_write_pending+0x74/0xe0 [ibmvscsis]
[  496.212991] [c0000007e58e39b0] [d000000006c74fc8] transport_generic_new_cmd+0x318/0x370 [target_core_mod]
[  496.213001] [c0000007e58e3a30] [d000000006c75084] transport_handle_cdb_direct+0x64/0xd0 [target_core_mod]
[  496.213011] [c0000007e58e3aa0] [d000000006c75298] target_submit_cmd_map_sgls+0x1a8/0x320 [target_core_mod]
[  496.213021] [c0000007e58e3b30] [d000000006c75458] target_submit_cmd+0x48/0x60 [target_core_mod]
[  496.213026] [c0000007e58e3bd0] [d000000006d34c20] ibmvscsis_scheduler+0x370/0x600 [ibmvscsis]
[  496.213031] [c0000007e58e3c90] [c00000000013135c] process_one_work+0x1ec/0x580
[  496.213035] [c0000007e58e3d20] [c000000000131798] worker_thread+0xa8/0x600
[  496.213039] [c0000007e58e3dc0] [c00000000013a468] kthread+0x168/0x1b0
[  496.213044] [c0000007e58e3e30] [c00000000000b528] ret_from_kernel_thread+0x5c/0xb4

[mkp: tweaked commit message]

Fixes: 2237498f0b5c ("target/iblock: Convert WRITE_SAME to blkdev_issue_zeroout")
Signed-off-by: Bryant G. Ly <bryantly@linux.vnet.ibm.com>
Reviewed-by: Steven Royer <seroyer@linux.vnet.ibm.com>
Tested-by: Taylor Jakobson <tjakobs@us.ibm.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Nicholas Bellinger <nab@linux-iscsi.org>
Cc: <stable@vger.kernel.org> # v4.13+
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
---
 drivers/target/target_core_iblock.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/target/target_core_iblock.c b/drivers/target/target_core_iblock.c
index 07c814c..6042901 100644
--- a/drivers/target/target_core_iblock.c
+++ b/drivers/target/target_core_iblock.c
@@ -427,8 +427,8 @@ iblock_execute_zero_out(struct block_device *bdev, struct se_cmd *cmd)
 {
 	struct se_device *dev = cmd->se_dev;
 	struct scatterlist *sg = &cmd->t_data_sg[0];
-	unsigned char *buf, zero = 0x00, *p = &zero;
-	int rc, ret;
+	unsigned char *buf, *not_zero;
+	int ret;
 
 	buf = kmap(sg_page(sg)) + sg->offset;
 	if (!buf)
@@ -437,10 +437,10 @@ iblock_execute_zero_out(struct block_device *bdev, struct se_cmd *cmd)
 	 * Fall back to block_execute_write_same() slow-path if
 	 * incoming WRITE_SAME payload does not contain zeros.
 	 */
-	rc = memcmp(buf, p, cmd->data_length);
+	not_zero = memchr_inv(buf, 0x00, cmd->data_length);
 	kunmap(sg_page(sg));
 
-	if (rc)
+	if (not_zero)
 		return TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE;
 
 	ret = blkdev_issue_zeroout(bdev,
-- 
cgit v1.1


From 011f22eb545a35f972036bb6a245c95c2e7e15a0 Mon Sep 17 00:00:00 2001
From: Hans de Goede <j.w.r.degoede@gmail.com>
Date: Fri, 20 Apr 2018 11:59:33 +0200
Subject: drm/i915: Do NOT skip the first 4k of stolen memory for pre-allocated
 buffers v2

Before this commit the WaSkipStolenMemoryFirstPage workaround code was
skipping the first 4k by passing 4096 as start of the address range passed
to drm_mm_init(). This means that calling drm_mm_reserve_node() to try and
reserve the firmware framebuffer so that we can inherit it would always
fail, as the firmware framebuffer starts at address 0.

Commit d43537610470 ("drm/i915: skip the first 4k of stolen memory on
everything >= gen8") says in its commit message: "This is confirmed to fix
Skylake screen flickering issues (probably caused by the fact that we
initialized a ring in the first page of stolen, but I didn't 100% confirm
this theory)."

Which suggests that it is safe to use the first page for a linear
framebuffer as the firmware is doing (see note below).

This commit always passes 0 as start to drm_mm_init() and works around
WaSkipStolenMemoryFirstPage in i915_gem_stolen_insert_node_in_range()
by insuring the start address passed by to drm_mm_insert_node_in_range()
is always 4k or more. All entry points to i915_gem_stolen.c go through
i915_gem_stolen_insert_node_in_range(), so that any newly allocated
objects such as ring-buffers will not be allocated in the first 4k.

The one exception is i915_gem_object_create_stolen_for_preallocated()
which directly calls drm_mm_reserve_node() which now will be able to
use the first 4k.

This fixes the i915 driver no longer being able to inherit the firmware
framebuffer on gen8+, which fixes the video output changing from the
vendor logo to a black screen as soon as the i915 driver is loaded
(on systems without fbcon).

Some notes about the mapping of the BIOS framebuffer:

v1 led to some discussion if the assumption of the intel_display.c code
that the firmware framebuffer is a linear mapping of the stolen memory
starting at offset 0 is still correct, because that would mean that the
GOP does not implement the WaSkipStolenMemoryFirstPage workaround.

To verify this the following code was added at the end of
i915_gem_object_create_stolen_for_preallocated() :

pr_err("first ggtt entry before bind: 0x%016llx\n",
       readq(dev_priv->ggtt.gsm));
ret = i915_vma_bind(vma,
            HAS_LLC(dev_priv) ? I915_CACHE_LLC : I915_CACHE_NONE,
            PIN_UPDATE);
pr_err("i915_vma_bind ret %d\n", ret);
pr_err("first ggtt entry after bind: 0x%016llx\n",
       readq(dev_priv->ggtt.gsm));

Which prints the mapping of the first page, then does a vma_bind() to
force update the mapping with our linear view of the framebuffer and
then prints the mapping of the first page again.

On an Asrock B150M Pro4S/D3 mainboard with i5-6500 CPU this prints:

[    1.651141] first ggtt entry before bind: 0x0000000078c00001
[    1.651151] i915_vma_bind ret 0
[    1.651152] first ggtt entry after bind: 0x0000000078c00083

And "sudo cat /proc/iomem | grep Stolen" gives:
  78c00000-88bfffff : Graphics Stolen Memory

There are no visual changes with this patch (BIOS vendor logo still
stays in place when we inherit the BIOS framebuffer), so the vma_bind()
does not impact which memory is being scanned out.

The address of the first ggtt entry matches with the start of stolen
and the i915_vma_bind call only changes the first gtt entry's flags,
or-ing in _PAGE_RW (BIT(1)) and PPAT_CACHED (BIT(7)), which perfectly
matches what we would expect based on gen8_pte_encode()'s behavior.

So it seems that the GOP indeed does NOT implement the wa and the i915's
code assuming a linear mapping at the start of stolen for the BIOS fb
still holds true for gen8+.

I've also tested this on a Cherry Trail based device (a GPD Win)
with identical results (the flags are 0x1b after the vma_bind
on CHT, which matches with I915_CACHE_NONE).

Changed in v2: No code changes, extended the commit message with the
verification that the intel_display.c BIOS framebuffer mapping is still
correct.

Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180420095933.16442-1-hdegoede@redhat.com
---
 drivers/gpu/drm/i915/i915_gem_stolen.c | 15 ++++++---------
 1 file changed, 6 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem_stolen.c b/drivers/gpu/drm/i915/i915_gem_stolen.c
index af915d0..ad949cc 100644
--- a/drivers/gpu/drm/i915/i915_gem_stolen.c
+++ b/drivers/gpu/drm/i915/i915_gem_stolen.c
@@ -51,6 +51,10 @@ int i915_gem_stolen_insert_node_in_range(struct drm_i915_private *dev_priv,
 	if (!drm_mm_initialized(&dev_priv->mm.stolen))
 		return -ENODEV;
 
+	/* WaSkipStolenMemoryFirstPage:bdw+ */
+	if (INTEL_GEN(dev_priv) >= 8 && start < 4096)
+		start = 4096;
+
 	mutex_lock(&dev_priv->mm.stolen_lock);
 	ret = drm_mm_insert_node_in_range(&dev_priv->mm.stolen, node,
 					  size, alignment, 0,
@@ -343,7 +347,6 @@ int i915_gem_init_stolen(struct drm_i915_private *dev_priv)
 {
 	resource_size_t reserved_base, stolen_top;
 	resource_size_t reserved_total, reserved_size;
-	resource_size_t stolen_usable_start;
 
 	mutex_init(&dev_priv->mm.stolen_lock);
 
@@ -435,17 +438,11 @@ int i915_gem_init_stolen(struct drm_i915_private *dev_priv)
 			 (u64)resource_size(&dev_priv->dsm) >> 10,
 			 ((u64)resource_size(&dev_priv->dsm) - reserved_total) >> 10);
 
-	stolen_usable_start = 0;
-	/* WaSkipStolenMemoryFirstPage:bdw+ */
-	if (INTEL_GEN(dev_priv) >= 8)
-		stolen_usable_start = 4096;
-
 	dev_priv->stolen_usable_size =
-		resource_size(&dev_priv->dsm) - reserved_total - stolen_usable_start;
+		resource_size(&dev_priv->dsm) - reserved_total;
 
 	/* Basic memrange allocator for stolen space. */
-	drm_mm_init(&dev_priv->mm.stolen, stolen_usable_start,
-		    dev_priv->stolen_usable_size);
+	drm_mm_init(&dev_priv->mm.stolen, 0, dev_priv->stolen_usable_size);
 
 	return 0;
 }
-- 
cgit v1.1


From 292bb0d38a5714440b59ef910404408d5e9a8017 Mon Sep 17 00:00:00 2001
From: Zhipeng Gong <zhipeng.gong@intel.com>
Date: Wed, 4 Apr 2018 08:43:52 +0800
Subject: drm/i915/gvt: Use real time to do timer check

intel_gvt_schedule check timer through a counter and is supposed
to wake up to increase the counter every ms.
In a system with heavy workload, gvt_service_thread can not get
a chance to run right after wake up and will be delayed several
milliseconds. As a result, one hundred counter interval means
several hundred milliseconds in real time.

This patch use real time instead of counter to do timer check.

v2: remove static variable. (Zhenyu)
v3: correct expire_time update. (Zhenyu)

Signed-off-by: Zhipeng Gong <zhipeng.gong@intel.com>
Cc: Zhenyu Wang <zhenyuw@linux.intel.com>
Cc: Min He <min.he@intel.com>
Signed-off-by: Zhi Wang <zhi.a.wang@intel.com>
---
 drivers/gpu/drm/i915/gvt/sched_policy.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/i915/gvt/sched_policy.c b/drivers/gpu/drm/i915/gvt/sched_policy.c
index 75b7bc7..8876a57 100644
--- a/drivers/gpu/drm/i915/gvt/sched_policy.c
+++ b/drivers/gpu/drm/i915/gvt/sched_policy.c
@@ -66,6 +66,7 @@ struct gvt_sched_data {
 	struct hrtimer timer;
 	unsigned long period;
 	struct list_head lru_runq_head;
+	ktime_t expire_time;
 };
 
 static void vgpu_update_timeslice(struct intel_vgpu *pre_vgpu)
@@ -226,14 +227,18 @@ out:
 void intel_gvt_schedule(struct intel_gvt *gvt)
 {
 	struct gvt_sched_data *sched_data = gvt->scheduler.sched_data;
-	static uint64_t timer_check;
 
 	mutex_lock(&gvt->lock);
 
 	if (test_and_clear_bit(INTEL_GVT_REQUEST_SCHED,
 				(void *)&gvt->service_request)) {
-		if (!(timer_check++ % GVT_TS_BALANCE_PERIOD_MS))
+		ktime_t cur_time = ktime_get();
+
+		if (cur_time >= sched_data->expire_time) {
 			gvt_balance_timeslice(sched_data);
+			sched_data->expire_time = ktime_add_ms(
+				cur_time, GVT_TS_BALANCE_PERIOD_MS);
+		}
 	}
 	clear_bit(INTEL_GVT_REQUEST_EVENT_SCHED, (void *)&gvt->service_request);
 
-- 
cgit v1.1


From 89babe7cf18e4f93c6ba1e6abfe2e5aa5e4fc66c Mon Sep 17 00:00:00 2001
From: Zhipeng Gong <zhipeng.gong@intel.com>
Date: Wed, 4 Apr 2018 08:43:53 +0800
Subject: drm/i915/gvt: Update time slice more frequently

When there is only one vGPU in GVT-g and it submits workloads
continuously, it will not be scheduled out, vgpu_update_timeslice
is not called and its sched_in_time is not updated in a long time,
which can be several seconds or longer.
Once GVT-g pauses to submit workload for this vGPU due to heavy
host CPU workload, this vGPU get scheduled out and
vgpu_update_timeslice is called, its left_ts will be subtract
by a big value from sched_out_time - sched_in_time.
When GVT-g is going to submit workload for this vGPU again,
it will not be scheduled in until gvt_balance_timeslice reaches
stage 0 and reset its left_ts, which introduces several
hunderand milliseconds latency.

This patch updates time slice in every ms to update sched_in_time
timely.

v2: revise commit message
v3: use more concise expr. (Zhenyu)

Signed-off-by: Zhipeng Gong <zhipeng.gong@intel.com>
Cc: Zhenyu Wang <zhenyuw@linux.intel.com>
Cc: Min He <min.he@intel.com>
Signed-off-by: Zhi Wang <zhi.a.wang@intel.com>
---
 drivers/gpu/drm/i915/gvt/sched_policy.c | 26 +++++++++++++-------------
 1 file changed, 13 insertions(+), 13 deletions(-)

diff --git a/drivers/gpu/drm/i915/gvt/sched_policy.c b/drivers/gpu/drm/i915/gvt/sched_policy.c
index 8876a57..d053cbe 100644
--- a/drivers/gpu/drm/i915/gvt/sched_policy.c
+++ b/drivers/gpu/drm/i915/gvt/sched_policy.c
@@ -53,7 +53,6 @@ struct vgpu_sched_data {
 	bool active;
 
 	ktime_t sched_in_time;
-	ktime_t sched_out_time;
 	ktime_t sched_time;
 	ktime_t left_ts;
 	ktime_t allocated_ts;
@@ -69,15 +68,19 @@ struct gvt_sched_data {
 	ktime_t expire_time;
 };
 
-static void vgpu_update_timeslice(struct intel_vgpu *pre_vgpu)
+static void vgpu_update_timeslice(struct intel_vgpu *vgpu, ktime_t cur_time)
 {
 	ktime_t delta_ts;
-	struct vgpu_sched_data *vgpu_data = pre_vgpu->sched_data;
+	struct vgpu_sched_data *vgpu_data;
 
-	delta_ts = vgpu_data->sched_out_time - vgpu_data->sched_in_time;
+	if (!vgpu || vgpu == vgpu->gvt->idle_vgpu)
+		return;
 
-	vgpu_data->sched_time += delta_ts;
-	vgpu_data->left_ts -= delta_ts;
+	vgpu_data = vgpu->sched_data;
+	delta_ts = ktime_sub(cur_time, vgpu_data->sched_in_time);
+	vgpu_data->sched_time = ktime_add(vgpu_data->sched_time, delta_ts);
+	vgpu_data->left_ts = ktime_sub(vgpu_data->left_ts, delta_ts);
+	vgpu_data->sched_in_time = cur_time;
 }
 
 #define GVT_TS_BALANCE_PERIOD_MS 100
@@ -151,11 +154,7 @@ static void try_to_schedule_next_vgpu(struct intel_gvt *gvt)
 	}
 
 	cur_time = ktime_get();
-	if (scheduler->current_vgpu) {
-		vgpu_data = scheduler->current_vgpu->sched_data;
-		vgpu_data->sched_out_time = cur_time;
-		vgpu_update_timeslice(scheduler->current_vgpu);
-	}
+	vgpu_update_timeslice(scheduler->current_vgpu, cur_time);
 	vgpu_data = scheduler->next_vgpu->sched_data;
 	vgpu_data->sched_in_time = cur_time;
 
@@ -227,13 +226,13 @@ out:
 void intel_gvt_schedule(struct intel_gvt *gvt)
 {
 	struct gvt_sched_data *sched_data = gvt->scheduler.sched_data;
+	ktime_t cur_time;
 
 	mutex_lock(&gvt->lock);
+	cur_time = ktime_get();
 
 	if (test_and_clear_bit(INTEL_GVT_REQUEST_SCHED,
 				(void *)&gvt->service_request)) {
-		ktime_t cur_time = ktime_get();
-
 		if (cur_time >= sched_data->expire_time) {
 			gvt_balance_timeslice(sched_data);
 			sched_data->expire_time = ktime_add_ms(
@@ -242,6 +241,7 @@ void intel_gvt_schedule(struct intel_gvt *gvt)
 	}
 	clear_bit(INTEL_GVT_REQUEST_EVENT_SCHED, (void *)&gvt->service_request);
 
+	vgpu_update_timeslice(gvt->scheduler.current_vgpu, cur_time);
 	tbs_sched_func(sched_data);
 
 	mutex_unlock(&gvt->lock);
-- 
cgit v1.1


From 96bebe39b2f4533af14c509061cd2b551ca81e8d Mon Sep 17 00:00:00 2001
From: Zhao Yan <yan.y.zhao@intel.com>
Date: Wed, 4 Apr 2018 13:57:09 +0800
Subject: drm/i915/gvt: scan non-privileged batch buffer for debug purpose

For perfomance purpose, scanning of non-privileged batch buffer is turned
off by default. But for debugging purpose, it can be turned on via debugfs.
After scanning, we submit the original non-privileged batch buffer into
hardware, so that the scanning is only a peeking window of guest submitted
commands and will not affect the execution results.

v4:
- refine debugfs print format&content (zhenyu wang)
- print engine id instread of engine name to prevent potential memory leak
  in debugfs warning message. (zhenyu wang)

v3:
- change vgpu->scan_nonprivbb from type bool to u32, so it is able to
  selectively turn on/off scanning of non-privileged batch buffer on engine
  level. e.g.
  if vgpu->scan_nonprivbb=3, then it will scan non-privileged batch buffer
  on engine 0 and 1.
- in debugfs interface to set vgpu->scan_nonprivbb, print warning message
  to warn user and explicitly tell state change in kernel log (zhenyu wang)
v2:
- rebase
- update comments for start_gma_offset (henry)

Signed-off-by: Zhao Yan <yan.y.zhao@intel.com>
Reviewed-by: Zhenyu Wang <zhenyuw@linux.intel.com>
Signed-off-by: Zhi Wang <zhi.a.wang@intel.com>
---
 drivers/gpu/drm/i915/gvt/cmd_parser.c | 55 +++++++++++++++++++++-------
 drivers/gpu/drm/i915/gvt/debugfs.c    | 67 +++++++++++++++++++++++++++++++++++
 drivers/gpu/drm/i915/gvt/gvt.h        |  1 +
 drivers/gpu/drm/i915/gvt/scheduler.c  | 64 ++++++++++++++++++++++-----------
 drivers/gpu/drm/i915/gvt/scheduler.h  |  1 +
 drivers/gpu/drm/i915/gvt/trace.h      | 24 ++++++++++---
 6 files changed, 173 insertions(+), 39 deletions(-)

diff --git a/drivers/gpu/drm/i915/gvt/cmd_parser.c b/drivers/gpu/drm/i915/gvt/cmd_parser.c
index db6b94d..9ec2cd9 100644
--- a/drivers/gpu/drm/i915/gvt/cmd_parser.c
+++ b/drivers/gpu/drm/i915/gvt/cmd_parser.c
@@ -1603,7 +1603,8 @@ static int batch_buffer_needs_scan(struct parser_exec_state *s)
 	if (IS_BROADWELL(gvt->dev_priv) || IS_SKYLAKE(gvt->dev_priv)
 		|| IS_KABYLAKE(gvt->dev_priv)) {
 		/* BDW decides privilege based on address space */
-		if (cmd_val(s, 0) & (1 << 8))
+		if (cmd_val(s, 0) & (1 << 8) &&
+			!(s->vgpu->scan_nonprivbb & (1 << s->ring_id)))
 			return 0;
 	}
 	return 1;
@@ -1617,6 +1618,8 @@ static int find_bb_size(struct parser_exec_state *s, unsigned long *bb_size)
 	bool bb_end = false;
 	struct intel_vgpu *vgpu = s->vgpu;
 	u32 cmd;
+	struct intel_vgpu_mm *mm = (s->buf_addr_type == GTT_BUFFER) ?
+		s->vgpu->gtt.ggtt_mm : s->workload->shadow_mm;
 
 	*bb_size = 0;
 
@@ -1628,18 +1631,22 @@ static int find_bb_size(struct parser_exec_state *s, unsigned long *bb_size)
 	cmd = cmd_val(s, 0);
 	info = get_cmd_info(s->vgpu->gvt, cmd, s->ring_id);
 	if (info == NULL) {
-		gvt_vgpu_err("unknown cmd 0x%x, opcode=0x%x\n",
-				cmd, get_opcode(cmd, s->ring_id));
+		gvt_vgpu_err("unknown cmd 0x%x, opcode=0x%x, addr_type=%s, ring %d, workload=%p\n",
+				cmd, get_opcode(cmd, s->ring_id),
+				(s->buf_addr_type == PPGTT_BUFFER) ?
+				"ppgtt" : "ggtt", s->ring_id, s->workload);
 		return -EBADRQC;
 	}
 	do {
-		if (copy_gma_to_hva(s->vgpu, s->vgpu->gtt.ggtt_mm,
+		if (copy_gma_to_hva(s->vgpu, mm,
 				gma, gma + 4, &cmd) < 0)
 			return -EFAULT;
 		info = get_cmd_info(s->vgpu->gvt, cmd, s->ring_id);
 		if (info == NULL) {
-			gvt_vgpu_err("unknown cmd 0x%x, opcode=0x%x\n",
-				cmd, get_opcode(cmd, s->ring_id));
+			gvt_vgpu_err("unknown cmd 0x%x, opcode=0x%x, addr_type=%s, ring %d, workload=%p\n",
+				cmd, get_opcode(cmd, s->ring_id),
+				(s->buf_addr_type == PPGTT_BUFFER) ?
+				"ppgtt" : "ggtt", s->ring_id, s->workload);
 			return -EBADRQC;
 		}
 
@@ -1665,6 +1672,9 @@ static int perform_bb_shadow(struct parser_exec_state *s)
 	unsigned long gma = 0;
 	unsigned long bb_size;
 	int ret = 0;
+	struct intel_vgpu_mm *mm = (s->buf_addr_type == GTT_BUFFER) ?
+		s->vgpu->gtt.ggtt_mm : s->workload->shadow_mm;
+	unsigned long gma_start_offset = 0;
 
 	/* get the start gm address of the batch buffer */
 	gma = get_gma_bb_from_cmd(s, 1);
@@ -1679,8 +1689,24 @@ static int perform_bb_shadow(struct parser_exec_state *s)
 	if (!bb)
 		return -ENOMEM;
 
+	bb->ppgtt = (s->buf_addr_type == GTT_BUFFER) ? false : true;
+
+	/* the gma_start_offset stores the batch buffer's start gma's
+	 * offset relative to page boundary. so for non-privileged batch
+	 * buffer, the shadowed gem object holds exactly the same page
+	 * layout as original gem object. This is for the convience of
+	 * replacing the whole non-privilged batch buffer page to this
+	 * shadowed one in PPGTT at the same gma address. (this replacing
+	 * action is not implemented yet now, but may be necessary in
+	 * future).
+	 * for prileged batch buffer, we just change start gma address to
+	 * that of shadowed page.
+	 */
+	if (bb->ppgtt)
+		gma_start_offset = gma & ~I915_GTT_PAGE_MASK;
+
 	bb->obj = i915_gem_object_create(s->vgpu->gvt->dev_priv,
-					 roundup(bb_size, PAGE_SIZE));
+			 roundup(bb_size + gma_start_offset, PAGE_SIZE));
 	if (IS_ERR(bb->obj)) {
 		ret = PTR_ERR(bb->obj);
 		goto err_free_bb;
@@ -1701,9 +1727,9 @@ static int perform_bb_shadow(struct parser_exec_state *s)
 		bb->clflush &= ~CLFLUSH_BEFORE;
 	}
 
-	ret = copy_gma_to_hva(s->vgpu, s->vgpu->gtt.ggtt_mm,
+	ret = copy_gma_to_hva(s->vgpu, mm,
 			      gma, gma + bb_size,
-			      bb->va);
+			      bb->va + gma_start_offset);
 	if (ret < 0) {
 		gvt_vgpu_err("fail to copy guest ring buffer\n");
 		ret = -EFAULT;
@@ -1729,7 +1755,7 @@ static int perform_bb_shadow(struct parser_exec_state *s)
 	 * buffer's gma in pair. After all, we don't want to pin the shadow
 	 * buffer here (too early).
 	 */
-	s->ip_va = bb->va;
+	s->ip_va = bb->va + gma_start_offset;
 	s->ip_gma = gma;
 	return 0;
 err_unmap:
@@ -2468,15 +2494,18 @@ static int cmd_parser_exec(struct parser_exec_state *s)
 
 	info = get_cmd_info(s->vgpu->gvt, cmd, s->ring_id);
 	if (info == NULL) {
-		gvt_vgpu_err("unknown cmd 0x%x, opcode=0x%x\n",
-				cmd, get_opcode(cmd, s->ring_id));
+		gvt_vgpu_err("unknown cmd 0x%x, opcode=0x%x, addr_type=%s, ring %d, workload=%p\n",
+				cmd, get_opcode(cmd, s->ring_id),
+				(s->buf_addr_type == PPGTT_BUFFER) ?
+				"ppgtt" : "ggtt", s->ring_id, s->workload);
 		return -EBADRQC;
 	}
 
 	s->info = info;
 
 	trace_gvt_command(vgpu->id, s->ring_id, s->ip_gma, s->ip_va,
-			  cmd_length(s), s->buf_type);
+			  cmd_length(s), s->buf_type, s->buf_addr_type,
+			  s->workload, info->name);
 
 	if (info->handler) {
 		ret = info->handler(s);
diff --git a/drivers/gpu/drm/i915/gvt/debugfs.c b/drivers/gpu/drm/i915/gvt/debugfs.c
index f7d0078..2ec89bc 100644
--- a/drivers/gpu/drm/i915/gvt/debugfs.c
+++ b/drivers/gpu/drm/i915/gvt/debugfs.c
@@ -124,6 +124,68 @@ static int vgpu_mmio_diff_show(struct seq_file *s, void *unused)
 }
 DEFINE_SHOW_ATTRIBUTE(vgpu_mmio_diff);
 
+static int
+vgpu_scan_nonprivbb_get(void *data, u64 *val)
+{
+	struct intel_vgpu *vgpu = (struct intel_vgpu *)data;
+	*val = vgpu->scan_nonprivbb;
+	return 0;
+}
+
+/*
+ * set/unset bit engine_id of vgpu->scan_nonprivbb to turn on/off scanning
+ * of non-privileged batch buffer. e.g.
+ * if vgpu->scan_nonprivbb=3, then it will scan non-privileged batch buffer
+ * on engine 0 and 1.
+ */
+static int
+vgpu_scan_nonprivbb_set(void *data, u64 val)
+{
+	struct intel_vgpu *vgpu = (struct intel_vgpu *)data;
+	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
+	enum intel_engine_id id;
+	char buf[128], *s;
+	int len;
+
+	val &= (1 << I915_NUM_ENGINES) - 1;
+
+	if (vgpu->scan_nonprivbb == val)
+		return 0;
+
+	if (!val)
+		goto done;
+
+	len = sprintf(buf,
+		"gvt: vgpu %d turns on non-privileged batch buffers scanning on Engines:",
+		vgpu->id);
+
+	s = buf + len;
+
+	for (id = 0; id < I915_NUM_ENGINES; id++) {
+		struct intel_engine_cs *engine;
+
+		engine = dev_priv->engine[id];
+		if (engine && (val & (1 << id))) {
+			len = snprintf(s, 4, "%d, ", engine->id);
+			s += len;
+		} else
+			val &=  ~(1 << id);
+	}
+
+	if (val)
+		sprintf(s, "low performance expected.");
+
+	pr_warn("%s\n", buf);
+
+done:
+	vgpu->scan_nonprivbb = val;
+	return 0;
+}
+
+DEFINE_SIMPLE_ATTRIBUTE(vgpu_scan_nonprivbb_fops,
+			vgpu_scan_nonprivbb_get, vgpu_scan_nonprivbb_set,
+			"0x%llx\n");
+
 /**
  * intel_gvt_debugfs_add_vgpu - register debugfs entries for a vGPU
  * @vgpu: a vGPU
@@ -151,6 +213,11 @@ int intel_gvt_debugfs_add_vgpu(struct intel_vgpu *vgpu)
 	if (!ent)
 		return -ENOMEM;
 
+	ent = debugfs_create_file("scan_nonprivbb", 0644, vgpu->debugfs,
+				 vgpu, &vgpu_scan_nonprivbb_fops);
+	if (!ent)
+		return -ENOMEM;
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/i915/gvt/gvt.h b/drivers/gpu/drm/i915/gvt/gvt.h
index efacd8a..6ec8888 100644
--- a/drivers/gpu/drm/i915/gvt/gvt.h
+++ b/drivers/gpu/drm/i915/gvt/gvt.h
@@ -226,6 +226,7 @@ struct intel_vgpu {
 
 	struct completion vblank_done;
 
+	u32 scan_nonprivbb;
 };
 
 /* validating GM healthy status*/
diff --git a/drivers/gpu/drm/i915/gvt/scheduler.c b/drivers/gpu/drm/i915/gvt/scheduler.c
index 638abe8..1bd7aa0 100644
--- a/drivers/gpu/drm/i915/gvt/scheduler.c
+++ b/drivers/gpu/drm/i915/gvt/scheduler.c
@@ -452,12 +452,6 @@ static int prepare_shadow_batch_buffer(struct intel_vgpu_workload *workload)
 	int ret;
 
 	list_for_each_entry(bb, &workload->shadow_bb, list) {
-		bb->vma = i915_gem_object_ggtt_pin(bb->obj, NULL, 0, 0, 0);
-		if (IS_ERR(bb->vma)) {
-			ret = PTR_ERR(bb->vma);
-			goto err;
-		}
-
 		/* For privilge batch buffer and not wa_ctx, the bb_start_cmd_va
 		 * is only updated into ring_scan_buffer, not real ring address
 		 * allocated in later copy_workload_to_ring_buffer. pls be noted
@@ -469,25 +463,53 @@ static int prepare_shadow_batch_buffer(struct intel_vgpu_workload *workload)
 			bb->bb_start_cmd_va = workload->shadow_ring_buffer_va
 				+ bb->bb_offset;
 
-		/* relocate shadow batch buffer */
-		bb->bb_start_cmd_va[1] = i915_ggtt_offset(bb->vma);
-		if (gmadr_bytes == 8)
-			bb->bb_start_cmd_va[2] = 0;
+		if (bb->ppgtt) {
+			/* for non-priv bb, scan&shadow is only for
+			 * debugging purpose, so the content of shadow bb
+			 * is the same as original bb. Therefore,
+			 * here, rather than switch to shadow bb's gma
+			 * address, we directly use original batch buffer's
+			 * gma address, and send original bb to hardware
+			 * directly
+			 */
+			if (bb->clflush & CLFLUSH_AFTER) {
+				drm_clflush_virt_range(bb->va,
+						bb->obj->base.size);
+				bb->clflush &= ~CLFLUSH_AFTER;
+			}
+			i915_gem_obj_finish_shmem_access(bb->obj);
+			bb->accessing = false;
+
+		} else {
+			bb->vma = i915_gem_object_ggtt_pin(bb->obj,
+					NULL, 0, 0, 0);
+			if (IS_ERR(bb->vma)) {
+				ret = PTR_ERR(bb->vma);
+				goto err;
+			}
 
-		/* No one is going to touch shadow bb from now on. */
-		if (bb->clflush & CLFLUSH_AFTER) {
-			drm_clflush_virt_range(bb->va, bb->obj->base.size);
-			bb->clflush &= ~CLFLUSH_AFTER;
-		}
+			/* relocate shadow batch buffer */
+			bb->bb_start_cmd_va[1] = i915_ggtt_offset(bb->vma);
+			if (gmadr_bytes == 8)
+				bb->bb_start_cmd_va[2] = 0;
 
-		ret = i915_gem_object_set_to_gtt_domain(bb->obj, false);
-		if (ret)
-			goto err;
+			/* No one is going to touch shadow bb from now on. */
+			if (bb->clflush & CLFLUSH_AFTER) {
+				drm_clflush_virt_range(bb->va,
+						bb->obj->base.size);
+				bb->clflush &= ~CLFLUSH_AFTER;
+			}
 
-		i915_gem_obj_finish_shmem_access(bb->obj);
-		bb->accessing = false;
+			ret = i915_gem_object_set_to_gtt_domain(bb->obj,
+					false);
+			if (ret)
+				goto err;
 
-		i915_vma_move_to_active(bb->vma, workload->req, 0);
+			i915_gem_obj_finish_shmem_access(bb->obj);
+			bb->accessing = false;
+
+			i915_vma_move_to_active(bb->vma, workload->req, 0);
+		}
 	}
 	return 0;
 err:
diff --git a/drivers/gpu/drm/i915/gvt/scheduler.h b/drivers/gpu/drm/i915/gvt/scheduler.h
index 486ed57..6c64478 100644
--- a/drivers/gpu/drm/i915/gvt/scheduler.h
+++ b/drivers/gpu/drm/i915/gvt/scheduler.h
@@ -125,6 +125,7 @@ struct intel_vgpu_shadow_bb {
 	unsigned int clflush;
 	bool accessing;
 	unsigned long bb_offset;
+	bool ppgtt;
 };
 
 #define workload_q_head(vgpu, ring_id) \
diff --git a/drivers/gpu/drm/i915/gvt/trace.h b/drivers/gpu/drm/i915/gvt/trace.h
index 82093f1..1fd6420 100644
--- a/drivers/gpu/drm/i915/gvt/trace.h
+++ b/drivers/gpu/drm/i915/gvt/trace.h
@@ -224,19 +224,25 @@ TRACE_EVENT(oos_sync,
 	TP_printk("%s", __entry->buf)
 );
 
+#define GVT_CMD_STR_LEN 40
 TRACE_EVENT(gvt_command,
-	TP_PROTO(u8 vgpu_id, u8 ring_id, u32 ip_gma, u32 *cmd_va, u32 cmd_len,
-		 u32 buf_type),
+	TP_PROTO(u8 vgpu_id, u8 ring_id, u32 ip_gma, u32 *cmd_va,
+		u32 cmd_len,  u32 buf_type, u32 buf_addr_type,
+		void *workload, char *cmd_name),
 
-	TP_ARGS(vgpu_id, ring_id, ip_gma, cmd_va, cmd_len, buf_type),
+	TP_ARGS(vgpu_id, ring_id, ip_gma, cmd_va, cmd_len, buf_type,
+		buf_addr_type, workload, cmd_name),
 
 	TP_STRUCT__entry(
 		__field(u8, vgpu_id)
 		__field(u8, ring_id)
 		__field(u32, ip_gma)
 		__field(u32, buf_type)
+		__field(u32, buf_addr_type)
 		__field(u32, cmd_len)
+		__field(void*, workload)
 		__dynamic_array(u32, raw_cmd, cmd_len)
+		__array(char, cmd_name, GVT_CMD_STR_LEN)
 	),
 
 	TP_fast_assign(
@@ -244,17 +250,25 @@ TRACE_EVENT(gvt_command,
 		__entry->ring_id = ring_id;
 		__entry->ip_gma = ip_gma;
 		__entry->buf_type = buf_type;
+		__entry->buf_addr_type = buf_addr_type;
 		__entry->cmd_len = cmd_len;
+		__entry->workload = workload;
+		snprintf(__entry->cmd_name, GVT_CMD_STR_LEN, "%s", cmd_name);
 		memcpy(__get_dynamic_array(raw_cmd), cmd_va, cmd_len * sizeof(*cmd_va));
 	),
 
 
-	TP_printk("vgpu%d ring %d: buf_type %u, ip_gma %08x, raw cmd %s",
+	TP_printk("vgpu%d ring %d: address_type %u, buf_type %u, ip_gma %08x,cmd (name=%s,len=%u,raw cmd=%s), workload=%p\n",
 		__entry->vgpu_id,
 		__entry->ring_id,
+		__entry->buf_addr_type,
 		__entry->buf_type,
 		__entry->ip_gma,
-		__print_array(__get_dynamic_array(raw_cmd), __entry->cmd_len, 4))
+		__entry->cmd_name,
+		__entry->cmd_len,
+		__print_array(__get_dynamic_array(raw_cmd),
+			__entry->cmd_len, 4),
+		__entry->workload)
 );
 
 #define GVT_TEMP_STR_LEN 10
-- 
cgit v1.1


From 41e7ccc19c2a52b13b77f2a489a466b140d2d235 Mon Sep 17 00:00:00 2001
From: "Gustavo A. R. Silva" <gustavo@embeddedor.com>
Date: Thu, 22 Mar 2018 13:21:54 -0500
Subject: drm/i915/gvt/scheduler: Remove unnecessary NULL checks in sr_oa_regs

The checks are misleading and not required [1].

[1] https://lkml.org/lkml/2018/3/19/1792

Addresses-Coverity-ID: 1466017
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Zhi Wang <zhi.a.wang@intel.com>
---
 drivers/gpu/drm/i915/gvt/scheduler.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/gvt/scheduler.c b/drivers/gpu/drm/i915/gvt/scheduler.c
index 1bd7aa0..d1a8fd8 100644
--- a/drivers/gpu/drm/i915/gvt/scheduler.c
+++ b/drivers/gpu/drm/i915/gvt/scheduler.c
@@ -97,7 +97,7 @@ static void sr_oa_regs(struct intel_vgpu_workload *workload,
 		i915_mmio_reg_offset(EU_PERF_CNTL6),
 	};
 
-	if (!workload || !reg_state || workload->ring_id != RCS)
+	if (workload->ring_id != RCS)
 		return;
 
 	if (save) {
-- 
cgit v1.1


From 3eda0d22ead04f81ea59c9584bcbf5b496745e92 Mon Sep 17 00:00:00 2001
From: "Gustavo A. R. Silva" <gustavo@embeddedor.com>
Date: Thu, 22 Mar 2018 12:27:54 -0500
Subject: drm/i915/gvt: Mark expected switch fall-through in
 handle_g2v_notification

In preparation to enabling -Wimplicit-fallthrough, mark switch cases
where we are expecting to fall through.

Addresses-Coverity-ID: 1466154 ("Missing break in switch")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: Zhi Wang <zhi.a.wang@intel.com>
---
 drivers/gpu/drm/i915/gvt/handlers.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/i915/gvt/handlers.c b/drivers/gpu/drm/i915/gvt/handlers.c
index 8c5d5d0..a33c1c3e 100644
--- a/drivers/gpu/drm/i915/gvt/handlers.c
+++ b/drivers/gpu/drm/i915/gvt/handlers.c
@@ -1150,6 +1150,7 @@ static int handle_g2v_notification(struct intel_vgpu *vgpu, int notification)
 	switch (notification) {
 	case VGT_G2V_PPGTT_L3_PAGE_TABLE_CREATE:
 		root_entry_type = GTT_TYPE_PPGTT_ROOT_L3_ENTRY;
+		/* fall through */
 	case VGT_G2V_PPGTT_L4_PAGE_TABLE_CREATE:
 		mm = intel_vgpu_get_ppgtt_mm(vgpu, root_entry_type, pdps);
 		return PTR_ERR_OR_ZERO(mm);
-- 
cgit v1.1


From 09abfe7b5b2f442a85f4c4d59ecf582ad76088d7 Mon Sep 17 00:00:00 2001
From: Roland Dreier <roland@purestorage.com>
Date: Thu, 19 Apr 2018 08:28:11 -0700
Subject: RDMA/ucma: Allow resolving address w/o specifying source address

The RDMA CM will select a source device and address by consulting
the routing table if no source address is passed into
rdma_resolve_address().  Userspace will ask for this by passing an
all-zero source address in the RESOLVE_IP command.  Unfortunately
the new check for non-zero address size rejects this with EINVAL,
which breaks valid userspace applications.

Fix this by explicitly allowing a zero address family for the source.

Fixes: 2975d5de6428 ("RDMA/ucma: Check AF family prior resolving address")
Cc: <stable@vger.kernel.org>
Signed-off-by: Roland Dreier <roland@purestorage.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/ucma.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/infiniband/core/ucma.c b/drivers/infiniband/core/ucma.c
index 0efa0e2..eab43b1 100644
--- a/drivers/infiniband/core/ucma.c
+++ b/drivers/infiniband/core/ucma.c
@@ -700,7 +700,7 @@ static ssize_t ucma_resolve_ip(struct ucma_file *file,
 	if (copy_from_user(&cmd, inbuf, sizeof(cmd)))
 		return -EFAULT;
 
-	if (!rdma_addr_size_in6(&cmd.src_addr) ||
+	if ((cmd.src_addr.sin6_family && !rdma_addr_size_in6(&cmd.src_addr)) ||
 	    !rdma_addr_size_in6(&cmd.dst_addr))
 		return -EINVAL;
 
-- 
cgit v1.1


From d50e14abe2d0024aa527b89c7990147df5d531a5 Mon Sep 17 00:00:00 2001
From: Jason Gunthorpe <jgg@mellanox.com>
Date: Fri, 20 Apr 2018 09:49:10 -0600
Subject: uapi: Fix SPDX tags for files referring to the 'OpenIB.org' license

Based on discussion with Kate Stewart this license is not a
BSD-2-Clause, but is now formally identified as Linux-OpenIB
by SPDX.

The key difference between the licenses is in the 'warranty'
paragraph.

if_infiniband.h refers to the 'OpenIB.org' license, but
does not include the text, instead it links to an obsolete
web site that contains a license that matches the BSD-2-Clause
SPX. There is no 'three clause' version of the OpenIB.org
license.

Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Acked-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 include/uapi/linux/if_infiniband.h      | 2 +-
 include/uapi/linux/rds.h                | 2 +-
 include/uapi/linux/tls.h                | 2 +-
 include/uapi/rdma/cxgb3-abi.h           | 2 +-
 include/uapi/rdma/cxgb4-abi.h           | 2 +-
 include/uapi/rdma/hns-abi.h             | 2 +-
 include/uapi/rdma/ib_user_cm.h          | 2 +-
 include/uapi/rdma/ib_user_ioctl_verbs.h | 2 +-
 include/uapi/rdma/ib_user_mad.h         | 2 +-
 include/uapi/rdma/ib_user_sa.h          | 2 +-
 include/uapi/rdma/ib_user_verbs.h       | 2 +-
 include/uapi/rdma/mlx4-abi.h            | 2 +-
 include/uapi/rdma/mlx5-abi.h            | 2 +-
 include/uapi/rdma/mthca-abi.h           | 2 +-
 include/uapi/rdma/nes-abi.h             | 2 +-
 include/uapi/rdma/qedr-abi.h            | 2 +-
 include/uapi/rdma/rdma_user_cm.h        | 2 +-
 include/uapi/rdma/rdma_user_ioctl.h     | 2 +-
 include/uapi/rdma/rdma_user_rxe.h       | 2 +-
 19 files changed, 19 insertions(+), 19 deletions(-)

diff --git a/include/uapi/linux/if_infiniband.h b/include/uapi/linux/if_infiniband.h
index 050b92d..0fc33bf 100644
--- a/include/uapi/linux/if_infiniband.h
+++ b/include/uapi/linux/if_infiniband.h
@@ -1,4 +1,4 @@
-/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-3-Clause) */
+/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */
 /*
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU
diff --git a/include/uapi/linux/rds.h b/include/uapi/linux/rds.h
index a66b213..20c6bd0 100644
--- a/include/uapi/linux/rds.h
+++ b/include/uapi/linux/rds.h
@@ -1,4 +1,4 @@
-/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */
+/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */
 /*
  * Copyright (c) 2008 Oracle.  All rights reserved.
  *
diff --git a/include/uapi/linux/tls.h b/include/uapi/linux/tls.h
index c6633e9..ff02287 100644
--- a/include/uapi/linux/tls.h
+++ b/include/uapi/linux/tls.h
@@ -1,4 +1,4 @@
-/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */
+/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */
 /*
  * Copyright (c) 2016-2017, Mellanox Technologies. All rights reserved.
  *
diff --git a/include/uapi/rdma/cxgb3-abi.h b/include/uapi/rdma/cxgb3-abi.h
index 9acb4b7..85aed67 100644
--- a/include/uapi/rdma/cxgb3-abi.h
+++ b/include/uapi/rdma/cxgb3-abi.h
@@ -1,4 +1,4 @@
-/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */
+/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */
 /*
  * Copyright (c) 2006 Chelsio, Inc. All rights reserved.
  *
diff --git a/include/uapi/rdma/cxgb4-abi.h b/include/uapi/rdma/cxgb4-abi.h
index 1fefd01..a159ba8 100644
--- a/include/uapi/rdma/cxgb4-abi.h
+++ b/include/uapi/rdma/cxgb4-abi.h
@@ -1,4 +1,4 @@
-/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */
+/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */
 /*
  * Copyright (c) 2009-2010 Chelsio, Inc. All rights reserved.
  *
diff --git a/include/uapi/rdma/hns-abi.h b/include/uapi/rdma/hns-abi.h
index 7092c8d..78613b6 100644
--- a/include/uapi/rdma/hns-abi.h
+++ b/include/uapi/rdma/hns-abi.h
@@ -1,4 +1,4 @@
-/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */
+/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */
 /*
  * Copyright (c) 2016 Hisilicon Limited.
  *
diff --git a/include/uapi/rdma/ib_user_cm.h b/include/uapi/rdma/ib_user_cm.h
index 4a8f956..e2709bb 100644
--- a/include/uapi/rdma/ib_user_cm.h
+++ b/include/uapi/rdma/ib_user_cm.h
@@ -1,4 +1,4 @@
-/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */
+/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */
 /*
  * Copyright (c) 2005 Topspin Communications.  All rights reserved.
  * Copyright (c) 2005 Intel Corporation.  All rights reserved.
diff --git a/include/uapi/rdma/ib_user_ioctl_verbs.h b/include/uapi/rdma/ib_user_ioctl_verbs.h
index 04e46ea..625545d 100644
--- a/include/uapi/rdma/ib_user_ioctl_verbs.h
+++ b/include/uapi/rdma/ib_user_ioctl_verbs.h
@@ -1,4 +1,4 @@
-/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */
+/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */
 /*
  * Copyright (c) 2017-2018, Mellanox Technologies inc.  All rights reserved.
  *
diff --git a/include/uapi/rdma/ib_user_mad.h b/include/uapi/rdma/ib_user_mad.h
index ef92118..90c0cf2 100644
--- a/include/uapi/rdma/ib_user_mad.h
+++ b/include/uapi/rdma/ib_user_mad.h
@@ -1,4 +1,4 @@
-/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */
+/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */
 /*
  * Copyright (c) 2004 Topspin Communications.  All rights reserved.
  * Copyright (c) 2005 Voltaire, Inc. All rights reserved.
diff --git a/include/uapi/rdma/ib_user_sa.h b/include/uapi/rdma/ib_user_sa.h
index 0d2607f..435155d 100644
--- a/include/uapi/rdma/ib_user_sa.h
+++ b/include/uapi/rdma/ib_user_sa.h
@@ -1,4 +1,4 @@
-/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */
+/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */
 /*
  * Copyright (c) 2005 Intel Corporation.  All rights reserved.
  *
diff --git a/include/uapi/rdma/ib_user_verbs.h b/include/uapi/rdma/ib_user_verbs.h
index 9be0739..6aeb033 100644
--- a/include/uapi/rdma/ib_user_verbs.h
+++ b/include/uapi/rdma/ib_user_verbs.h
@@ -1,4 +1,4 @@
-/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */
+/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */
 /*
  * Copyright (c) 2005 Topspin Communications.  All rights reserved.
  * Copyright (c) 2005, 2006 Cisco Systems.  All rights reserved.
diff --git a/include/uapi/rdma/mlx4-abi.h b/include/uapi/rdma/mlx4-abi.h
index 04f64bc..f745575 100644
--- a/include/uapi/rdma/mlx4-abi.h
+++ b/include/uapi/rdma/mlx4-abi.h
@@ -1,4 +1,4 @@
-/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */
+/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */
 /*
  * Copyright (c) 2007 Cisco Systems, Inc. All rights reserved.
  * Copyright (c) 2007, 2008 Mellanox Technologies. All rights reserved.
diff --git a/include/uapi/rdma/mlx5-abi.h b/include/uapi/rdma/mlx5-abi.h
index cb4a02c..fdaf00e 100644
--- a/include/uapi/rdma/mlx5-abi.h
+++ b/include/uapi/rdma/mlx5-abi.h
@@ -1,4 +1,4 @@
-/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */
+/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */
 /*
  * Copyright (c) 2013-2015, Mellanox Technologies. All rights reserved.
  *
diff --git a/include/uapi/rdma/mthca-abi.h b/include/uapi/rdma/mthca-abi.h
index ac756cd..91b12e1 100644
--- a/include/uapi/rdma/mthca-abi.h
+++ b/include/uapi/rdma/mthca-abi.h
@@ -1,4 +1,4 @@
-/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */
+/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */
 /*
  * Copyright (c) 2005 Topspin Communications.  All rights reserved.
  * Copyright (c) 2005, 2006 Cisco Systems.  All rights reserved.
diff --git a/include/uapi/rdma/nes-abi.h b/include/uapi/rdma/nes-abi.h
index 35bfd40..f80495b 100644
--- a/include/uapi/rdma/nes-abi.h
+++ b/include/uapi/rdma/nes-abi.h
@@ -1,4 +1,4 @@
-/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */
+/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */
 /*
  * Copyright (c) 2006 - 2011 Intel Corporation.  All rights reserved.
  * Copyright (c) 2005 Topspin Communications.  All rights reserved.
diff --git a/include/uapi/rdma/qedr-abi.h b/include/uapi/rdma/qedr-abi.h
index 8ba0989..24c658b 100644
--- a/include/uapi/rdma/qedr-abi.h
+++ b/include/uapi/rdma/qedr-abi.h
@@ -1,4 +1,4 @@
-/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */
+/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */
 /* QLogic qedr NIC Driver
  * Copyright (c) 2015-2016  QLogic Corporation
  *
diff --git a/include/uapi/rdma/rdma_user_cm.h b/include/uapi/rdma/rdma_user_cm.h
index e126902..0d1e78e 100644
--- a/include/uapi/rdma/rdma_user_cm.h
+++ b/include/uapi/rdma/rdma_user_cm.h
@@ -1,4 +1,4 @@
-/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */
+/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */
 /*
  * Copyright (c) 2005-2006 Intel Corporation.  All rights reserved.
  *
diff --git a/include/uapi/rdma/rdma_user_ioctl.h b/include/uapi/rdma/rdma_user_ioctl.h
index d223f41..d92d272 100644
--- a/include/uapi/rdma/rdma_user_ioctl.h
+++ b/include/uapi/rdma/rdma_user_ioctl.h
@@ -1,4 +1,4 @@
-/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */
+/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */
 /*
  * Copyright (c) 2016 Mellanox Technologies, LTD. All rights reserved.
  *
diff --git a/include/uapi/rdma/rdma_user_rxe.h b/include/uapi/rdma/rdma_user_rxe.h
index 1f8a9e7..44ef6a3 100644
--- a/include/uapi/rdma/rdma_user_rxe.h
+++ b/include/uapi/rdma/rdma_user_rxe.h
@@ -1,4 +1,4 @@
-/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) */
+/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR Linux-OpenIB) */
 /*
  * Copyright (c) 2016 Mellanox Technologies Ltd. All rights reserved.
  *
-- 
cgit v1.1


From 94912e8df4e5005cb5fa3f3603741eb5e8bf3334 Mon Sep 17 00:00:00 2001
From: "Gustavo A. R. Silva" <gustavo@embeddedor.com>
Date: Thu, 5 Apr 2018 11:20:06 -0500
Subject: ieee802154: mcr20a: Fix memory leak in mcr20a_probe

Free allocated memory for pdata before return.

Addresses-Coverity-ID: 1466096 ("Resource leak")
Fixes: 8c6ad9cc5157 ("ieee802154: Add NXP MCR20A IEEE 802.15.4 transceiver driver")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Acked-by: Xue Liu <liuxuenetmail@gmail.com>
Signed-off-by: Stefan Schmidt <stefan@osg.samsung.com>
---
 drivers/net/ieee802154/mcr20a.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ieee802154/mcr20a.c b/drivers/net/ieee802154/mcr20a.c
index 55a22c7..944470d 100644
--- a/drivers/net/ieee802154/mcr20a.c
+++ b/drivers/net/ieee802154/mcr20a.c
@@ -1267,7 +1267,7 @@ mcr20a_probe(struct spi_device *spi)
 	ret = mcr20a_get_platform_data(spi, pdata);
 	if (ret < 0) {
 		dev_crit(&spi->dev, "mcr20a_get_platform_data failed.\n");
-		return ret;
+		goto free_pdata;
 	}
 
 	/* init reset gpio */
@@ -1275,7 +1275,7 @@ mcr20a_probe(struct spi_device *spi)
 		ret = devm_gpio_request_one(&spi->dev, pdata->rst_gpio,
 					    GPIOF_OUT_INIT_HIGH, "reset");
 		if (ret)
-			return ret;
+			goto free_pdata;
 	}
 
 	/* reset mcr20a */
@@ -1291,7 +1291,8 @@ mcr20a_probe(struct spi_device *spi)
 	hw = ieee802154_alloc_hw(sizeof(*lp), &mcr20a_hw_ops);
 	if (!hw) {
 		dev_crit(&spi->dev, "ieee802154_alloc_hw failed\n");
-		return -ENOMEM;
+		ret = -ENOMEM;
+		goto free_pdata;
 	}
 
 	/* init mcr20a local data */
@@ -1366,6 +1367,8 @@ mcr20a_probe(struct spi_device *spi)
 
 free_dev:
 	ieee802154_free_hw(lp->hw);
+free_pdata:
+	kfree(pdata);
 
 	return ret;
 }
-- 
cgit v1.1


From f9e628a6645ed4f42a15dfa3c5af45a916a228e4 Mon Sep 17 00:00:00 2001
From: Jia-Ju Bai <baijiaju1990@gmail.com>
Date: Wed, 11 Apr 2018 10:14:10 +0800
Subject: net: ieee802154: atusb: Replace GFP_ATOMIC with GFP_KERNEL in
 atusb_probe

atusb_probe() is never called in atomic context.
This function is only set as ".probe" in struct usb_driver.

Despite never getting called from atomic context,
atusb_probe() calls usb_alloc_urb() with GFP_ATOMIC,
which does not sleep for allocation.
GFP_ATOMIC is not necessary and can be replaced with GFP_KERNEL,
which can sleep and improve the possibility of sucessful allocation.

This is found by a static analysis tool named DCNS written by myself.
And I also manually check it.

Signed-off-by: Jia-Ju Bai <baijiaju1990@gmail.com>
Signed-off-by: Stefan Schmidt <stefan@osg.samsung.com>
---
 drivers/net/ieee802154/atusb.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ieee802154/atusb.c b/drivers/net/ieee802154/atusb.c
index 9fb9b56..4f684cb 100644
--- a/drivers/net/ieee802154/atusb.c
+++ b/drivers/net/ieee802154/atusb.c
@@ -1045,7 +1045,7 @@ static int atusb_probe(struct usb_interface *interface,
 	atusb->tx_dr.bRequest = ATUSB_TX;
 	atusb->tx_dr.wValue = cpu_to_le16(0);
 
-	atusb->tx_urb = usb_alloc_urb(0, GFP_ATOMIC);
+	atusb->tx_urb = usb_alloc_urb(0, GFP_KERNEL);
 	if (!atusb->tx_urb)
 		goto fail;
 
-- 
cgit v1.1


From 7e0ffee1b2e7d26cc2147be0d9d5186823e923f3 Mon Sep 17 00:00:00 2001
From: Stefan Schmidt <stefan@osg.samsung.com>
Date: Thu, 12 Apr 2018 14:28:49 +0200
Subject: net: ieee802154: mcr20a: do not leak resources on error path

We already allocated the device and platform  data at this
point. Instead of simply return from the probe function we
need to cleanup the resources first.

Signed-off-by: Stefan Schmidt <stefan@osg.samsung.com>
Acked-by: Xue Liu <liuxuenetmail@gmail.com>
---
 drivers/net/ieee802154/mcr20a.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ieee802154/mcr20a.c b/drivers/net/ieee802154/mcr20a.c
index 944470d..de0d7f2 100644
--- a/drivers/net/ieee802154/mcr20a.c
+++ b/drivers/net/ieee802154/mcr20a.c
@@ -1309,8 +1309,10 @@ mcr20a_probe(struct spi_device *spi)
 	/* init buf */
 	lp->buf = devm_kzalloc(&spi->dev, SPI_COMMAND_BUFFER, GFP_KERNEL);
 
-	if (!lp->buf)
-		return -ENOMEM;
+	if (!lp->buf) {
+		ret = -ENOMEM;
+		goto free_dev;
+	}
 
 	mcr20a_setup_tx_spi_messages(lp);
 	mcr20a_setup_rx_spi_messages(lp);
-- 
cgit v1.1


From f18fa5de5ba7f1d6650951502bb96a6e4715a948 Mon Sep 17 00:00:00 2001
From: Alexander Aring <aring@mojatatu.com>
Date: Fri, 20 Apr 2018 14:54:13 -0400
Subject: net: ieee802154: 6lowpan: fix frag reassembly

This patch initialize stack variables which are used in
frag_lowpan_compare_key to zero. In my case there are padding bytes in the
structures ieee802154_addr as well in frag_lowpan_compare_key. Otherwise
the key variable contains random bytes. The result is that a compare of
two keys by memcmp works incorrect.

Fixes: 648700f76b03 ("inet: frags: use rhashtables for reassembly units")
Signed-off-by: Alexander Aring <aring@mojatatu.com>
Reported-by: Stefan Schmidt <stefan@osg.samsung.com>
Signed-off-by: Stefan Schmidt <stefan@osg.samsung.com>
---
 net/ieee802154/6lowpan/6lowpan_i.h  |  4 ++--
 net/ieee802154/6lowpan/reassembly.c | 14 +++++++-------
 2 files changed, 9 insertions(+), 9 deletions(-)

diff --git a/net/ieee802154/6lowpan/6lowpan_i.h b/net/ieee802154/6lowpan/6lowpan_i.h
index b8d95cb..44a7e16 100644
--- a/net/ieee802154/6lowpan/6lowpan_i.h
+++ b/net/ieee802154/6lowpan/6lowpan_i.h
@@ -20,8 +20,8 @@ typedef unsigned __bitwise lowpan_rx_result;
 struct frag_lowpan_compare_key {
 	u16 tag;
 	u16 d_size;
-	const struct ieee802154_addr src;
-	const struct ieee802154_addr dst;
+	struct ieee802154_addr src;
+	struct ieee802154_addr dst;
 };
 
 /* Equivalent of ipv4 struct ipq
diff --git a/net/ieee802154/6lowpan/reassembly.c b/net/ieee802154/6lowpan/reassembly.c
index 1790b65..2cc2241 100644
--- a/net/ieee802154/6lowpan/reassembly.c
+++ b/net/ieee802154/6lowpan/reassembly.c
@@ -75,14 +75,14 @@ fq_find(struct net *net, const struct lowpan_802154_cb *cb,
 {
 	struct netns_ieee802154_lowpan *ieee802154_lowpan =
 		net_ieee802154_lowpan(net);
-	struct frag_lowpan_compare_key key = {
-		.tag = cb->d_tag,
-		.d_size = cb->d_size,
-		.src = *src,
-		.dst = *dst,
-	};
+	struct frag_lowpan_compare_key key = {};
 	struct inet_frag_queue *q;
 
+	key.tag = cb->d_tag;
+	key.d_size = cb->d_size;
+	key.src = *src;
+	key.dst = *dst;
+
 	q = inet_frag_find(&ieee802154_lowpan->frags, &key);
 	if (!q)
 		return NULL;
@@ -372,7 +372,7 @@ int lowpan_frag_rcv(struct sk_buff *skb, u8 frag_type)
 	struct lowpan_frag_queue *fq;
 	struct net *net = dev_net(skb->dev);
 	struct lowpan_802154_cb *cb = lowpan_802154_cb(skb);
-	struct ieee802154_hdr hdr;
+	struct ieee802154_hdr hdr = {};
 	int err;
 
 	if (ieee802154_hdr_peek_addrs(skb, &hdr) < 0)
-- 
cgit v1.1


From 6899b32b5b2dee358936b82b8363b716607a138f Mon Sep 17 00:00:00 2001
From: Roman Gushchin <guro@fb.com>
Date: Mon, 23 Apr 2018 18:09:21 +0100
Subject: bpf: disable and restore preemption in __BPF_PROG_RUN_ARRAY

Running bpf programs requires disabled preemption,
however at least some* of the BPF_PROG_RUN_ARRAY users
do not follow this rule.

To fix this bug, and also to make it not happen in the future,
let's add explicit preemption disabling/re-enabling
to the __BPF_PROG_RUN_ARRAY code.

* for example:
 [   17.624472] RIP: 0010:__cgroup_bpf_run_filter_sk+0x1c4/0x1d0
 ...
 [   17.640890]  inet6_create+0x3eb/0x520
 [   17.641405]  __sock_create+0x242/0x340
 [   17.641939]  __sys_socket+0x57/0xe0
 [   17.642370]  ? trace_hardirqs_off_thunk+0x1a/0x1c
 [   17.642944]  SyS_socket+0xa/0x10
 [   17.643357]  do_syscall_64+0x79/0x220
 [   17.643879]  entry_SYSCALL_64_after_hwframe+0x42/0xb7

Signed-off-by: Roman Gushchin <guro@fb.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
---
 include/linux/bpf.h | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/include/linux/bpf.h b/include/linux/bpf.h
index 486e65e..dc586cc 100644
--- a/include/linux/bpf.h
+++ b/include/linux/bpf.h
@@ -351,6 +351,7 @@ int bpf_prog_array_copy(struct bpf_prog_array __rcu *old_array,
 		struct bpf_prog **_prog, *__prog;	\
 		struct bpf_prog_array *_array;		\
 		u32 _ret = 1;				\
+		preempt_disable();			\
 		rcu_read_lock();			\
 		_array = rcu_dereference(array);	\
 		if (unlikely(check_non_null && !_array))\
@@ -362,6 +363,7 @@ int bpf_prog_array_copy(struct bpf_prog_array __rcu *old_array,
 		}					\
 _out:							\
 		rcu_read_unlock();			\
+		preempt_enable_no_resched();		\
 		_ret;					\
 	 })
 
-- 
cgit v1.1


From a66ed149b0da5b97fa7486ee2bcd00f3f9df5442 Mon Sep 17 00:00:00 2001
From: Parav Pandit <parav@mellanox.com>
Date: Mon, 23 Apr 2018 16:58:17 +0300
Subject: IB/core: Don't allow default GID addition at non reseved slots

Default GIDs are marked reserved at the start of the GID table at index
0 and 1 by gid_table_reserve_default().  Currently when default GID is
requested, it can still allocates an empty slot which was not marked as
RESERVED for default GID, which is incorrect.

At least in current code flow of roce_gid_mgmt.c, in theory we can
still request to allocate more than one/two default GIDs depending
on how upper devices are setup.

Therefore, it is better for cache layer to only allow our reserved slots
to be used by default GID allocation requests.

Fixes: 598ff6bae689 ("IB/core: Refactor GID modify code for RoCE")
Signed-off-by: Parav Pandit <parav@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/cache.c | 20 ++++++++++++--------
 1 file changed, 12 insertions(+), 8 deletions(-)

diff --git a/drivers/infiniband/core/cache.c b/drivers/infiniband/core/cache.c
index e337b08..92ec845 100644
--- a/drivers/infiniband/core/cache.c
+++ b/drivers/infiniband/core/cache.c
@@ -291,14 +291,18 @@ static int find_gid(struct ib_gid_table *table, const union ib_gid *gid,
 		 * so lookup free slot only if requested.
 		 */
 		if (pempty && empty < 0) {
-			if (data->props & GID_TABLE_ENTRY_INVALID) {
-				/* Found an invalid (free) entry; allocate it */
-				if (data->props & GID_TABLE_ENTRY_DEFAULT) {
-					if (default_gid)
-						empty = curr_index;
-				} else {
-					empty = curr_index;
-				}
+			if (data->props & GID_TABLE_ENTRY_INVALID &&
+			    (default_gid ==
+			     !!(data->props & GID_TABLE_ENTRY_DEFAULT))) {
+				/*
+				 * Found an invalid (free) entry; allocate it.
+				 * If default GID is requested, then our
+				 * found slot must be one of the DEFAULT
+				 * reserved slots or we fail.
+				 * This ensures that only DEFAULT reserved
+				 * slots are used for default property GIDs.
+				 */
+				empty = curr_index;
 			}
 		}
 
-- 
cgit v1.1


From 22c01ee4b8a8c000c490dfc479e175404e64167b Mon Sep 17 00:00:00 2001
From: Parav Pandit <parav@mellanox.com>
Date: Mon, 23 Apr 2018 16:58:18 +0300
Subject: IB/core: Fix to avoid deleting IPv6 look alike default GIDs

When IPv6 link local address is removed, if it matches with the default
GID, default GID(s)s gets removed which may not be a desired behavior.
This behavior is introduced by refactor work in Fixes tag.

When IPv6 link address is removed, removing its equivalent RoCEv2 GID
which exactly matches with default RoCEv2 GID, is right thing to do.
However achieving it correctly requires lot more changes, likely in
roce_gid_mgmt.c and core/cache.c. This should be done as independent
patch.

Therefore, this patch preserves behavior of not deleteing default GIDs.
This is done by providing explicit hint to consider default GID property
using mask and default_gid; similar to add_gid().

Fixes: 598ff6bae68 ("IB/core: Refactor GID modify code for RoCE")
Signed-off-by: Parav Pandit <parav@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/cache.c | 17 +++++++++++++----
 1 file changed, 13 insertions(+), 4 deletions(-)

diff --git a/drivers/infiniband/core/cache.c b/drivers/infiniband/core/cache.c
index 92ec845..8e6a4f0 100644
--- a/drivers/infiniband/core/cache.c
+++ b/drivers/infiniband/core/cache.c
@@ -424,8 +424,10 @@ int ib_cache_gid_add(struct ib_device *ib_dev, u8 port,
 	return ret;
 }
 
-int ib_cache_gid_del(struct ib_device *ib_dev, u8 port,
-		     union ib_gid *gid, struct ib_gid_attr *attr)
+static int
+_ib_cache_gid_del(struct ib_device *ib_dev, u8 port,
+		  union ib_gid *gid, struct ib_gid_attr *attr,
+		  bool default_gid)
 {
 	struct ib_gid_table *table;
 	int ret = 0;
@@ -435,9 +437,10 @@ int ib_cache_gid_del(struct ib_device *ib_dev, u8 port,
 
 	mutex_lock(&table->lock);
 
-	ix = find_gid(table, gid, attr, false,
+	ix = find_gid(table, gid, attr, default_gid,
 		      GID_ATTR_FIND_MASK_GID	  |
 		      GID_ATTR_FIND_MASK_GID_TYPE |
+		      GID_ATTR_FIND_MASK_DEFAULT  |
 		      GID_ATTR_FIND_MASK_NETDEV,
 		      NULL);
 	if (ix < 0) {
@@ -456,6 +459,12 @@ out_unlock:
 	return ret;
 }
 
+int ib_cache_gid_del(struct ib_device *ib_dev, u8 port,
+		     union ib_gid *gid, struct ib_gid_attr *attr)
+{
+	return _ib_cache_gid_del(ib_dev, port, gid, attr, false);
+}
+
 int ib_cache_gid_del_all_netdev_gids(struct ib_device *ib_dev, u8 port,
 				     struct net_device *ndev)
 {
@@ -756,7 +765,7 @@ void ib_cache_gid_set_default_gid(struct ib_device *ib_dev, u8 port,
 			__ib_cache_gid_add(ib_dev, port, &gid,
 					   &gid_attr, mask, true);
 		} else if (mode == IB_CACHE_GID_DEFAULT_MODE_DELETE) {
-			ib_cache_gid_del(ib_dev, port, &gid, &gid_attr);
+			_ib_cache_gid_del(ib_dev, port, &gid, &gid_attr, true);
 		}
 	}
 }
-- 
cgit v1.1


From dc5640f294e4ff6b89047cb4a0dfa931d5f0cd1f Mon Sep 17 00:00:00 2001
From: Parav Pandit <parav@mellanox.com>
Date: Mon, 23 Apr 2018 16:58:19 +0300
Subject: IB/core: Fix deleting default GIDs when changing mac adddress

Before [1], When MAC address of the netdevice is changed, default GID is
supposed to get deleted and added back which affects the node and/or port
GUID in below sequence.

netdevice_event()
-> NETDEV_CHANGEADDR
   default_del_cmd()
      del_netdev_default_ips()
          bond_delete_netdev_default_gids()
              ib_cache_gid_set_default_gid()
                  ib_cache_gid_del()
   add_cmd()
   [..]

However, ib_cache_gid_del() was not getting invoked in non bonding
scenarios because event_ndev and rdma_ndev are same.
Therefore, fix such condition to ignore checking upper device when event
ndev and rdma_dev are same; similar to bond_set_netdev_default_gids().

Which this fix ib_cache_gid_del() is invoked correctly; however
ib_cache_gid_del() doesn't find the default GID for deletion because
find_gid() was given default_gid = false with
GID_ATTR_FIND_MASK_DEFAULT set.
But it was getting overwritten by ib_cache_gid_set_default_gid() later
on as part of add_cmd().
Therefore, mac address change used to work for default GID.

With refactor series [1], this incorrect behavior is detected.

Therefore,
when deleting default GID, set default_gid and set MASK flag.
when deleting IP based GID, clear default_gid and set MASK flag.

[1] https://patchwork.kernel.org/patch/10319151/

Fixes: 238fdf48f2b5 ("IB/core: Add RoCE table bonding support")
Fixes: 598ff6bae689 ("IB/core: Refactor GID modify code for RoCE")
Signed-off-by: Parav Pandit <parav@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/cache.c         | 28 +++++++++++++++-------------
 drivers/infiniband/core/roce_gid_mgmt.c | 26 ++++++++++++++------------
 2 files changed, 29 insertions(+), 25 deletions(-)

diff --git a/drivers/infiniband/core/cache.c b/drivers/infiniband/core/cache.c
index 8e6a4f0..fb2d347 100644
--- a/drivers/infiniband/core/cache.c
+++ b/drivers/infiniband/core/cache.c
@@ -427,7 +427,7 @@ int ib_cache_gid_add(struct ib_device *ib_dev, u8 port,
 static int
 _ib_cache_gid_del(struct ib_device *ib_dev, u8 port,
 		  union ib_gid *gid, struct ib_gid_attr *attr,
-		  bool default_gid)
+		  unsigned long mask, bool default_gid)
 {
 	struct ib_gid_table *table;
 	int ret = 0;
@@ -437,12 +437,7 @@ _ib_cache_gid_del(struct ib_device *ib_dev, u8 port,
 
 	mutex_lock(&table->lock);
 
-	ix = find_gid(table, gid, attr, default_gid,
-		      GID_ATTR_FIND_MASK_GID	  |
-		      GID_ATTR_FIND_MASK_GID_TYPE |
-		      GID_ATTR_FIND_MASK_DEFAULT  |
-		      GID_ATTR_FIND_MASK_NETDEV,
-		      NULL);
+	ix = find_gid(table, gid, attr, default_gid, mask, NULL);
 	if (ix < 0) {
 		ret = -EINVAL;
 		goto out_unlock;
@@ -462,7 +457,12 @@ out_unlock:
 int ib_cache_gid_del(struct ib_device *ib_dev, u8 port,
 		     union ib_gid *gid, struct ib_gid_attr *attr)
 {
-	return _ib_cache_gid_del(ib_dev, port, gid, attr, false);
+	unsigned long mask = GID_ATTR_FIND_MASK_GID	  |
+			     GID_ATTR_FIND_MASK_GID_TYPE |
+			     GID_ATTR_FIND_MASK_DEFAULT  |
+			     GID_ATTR_FIND_MASK_NETDEV;
+
+	return _ib_cache_gid_del(ib_dev, port, gid, attr, mask, false);
 }
 
 int ib_cache_gid_del_all_netdev_gids(struct ib_device *ib_dev, u8 port,
@@ -741,7 +741,7 @@ void ib_cache_gid_set_default_gid(struct ib_device *ib_dev, u8 port,
 				  unsigned long gid_type_mask,
 				  enum ib_cache_gid_default_mode mode)
 {
-	union ib_gid gid;
+	union ib_gid gid = { };
 	struct ib_gid_attr gid_attr;
 	struct ib_gid_table *table;
 	unsigned int gid_type;
@@ -749,7 +749,9 @@ void ib_cache_gid_set_default_gid(struct ib_device *ib_dev, u8 port,
 
 	table = ib_dev->cache.ports[port - rdma_start_port(ib_dev)].gid;
 
-	make_default_gid(ndev, &gid);
+	mask = GID_ATTR_FIND_MASK_GID_TYPE |
+	       GID_ATTR_FIND_MASK_DEFAULT |
+	       GID_ATTR_FIND_MASK_NETDEV;
 	memset(&gid_attr, 0, sizeof(gid_attr));
 	gid_attr.ndev = ndev;
 
@@ -760,12 +762,12 @@ void ib_cache_gid_set_default_gid(struct ib_device *ib_dev, u8 port,
 		gid_attr.gid_type = gid_type;
 
 		if (mode == IB_CACHE_GID_DEFAULT_MODE_SET) {
-			mask = GID_ATTR_FIND_MASK_GID_TYPE |
-			       GID_ATTR_FIND_MASK_DEFAULT;
+			make_default_gid(ndev, &gid);
 			__ib_cache_gid_add(ib_dev, port, &gid,
 					   &gid_attr, mask, true);
 		} else if (mode == IB_CACHE_GID_DEFAULT_MODE_DELETE) {
-			_ib_cache_gid_del(ib_dev, port, &gid, &gid_attr, true);
+			_ib_cache_gid_del(ib_dev, port, &gid,
+					  &gid_attr, mask, true);
 		}
 	}
 }
diff --git a/drivers/infiniband/core/roce_gid_mgmt.c b/drivers/infiniband/core/roce_gid_mgmt.c
index cc29663..c0e4fd5 100644
--- a/drivers/infiniband/core/roce_gid_mgmt.c
+++ b/drivers/infiniband/core/roce_gid_mgmt.c
@@ -255,6 +255,7 @@ static void bond_delete_netdev_default_gids(struct ib_device *ib_dev,
 					    struct net_device *rdma_ndev)
 {
 	struct net_device *real_dev = rdma_vlan_dev_real_dev(event_ndev);
+	unsigned long gid_type_mask;
 
 	if (!rdma_ndev)
 		return;
@@ -264,21 +265,22 @@ static void bond_delete_netdev_default_gids(struct ib_device *ib_dev,
 
 	rcu_read_lock();
 
-	if (rdma_is_upper_dev_rcu(rdma_ndev, event_ndev) &&
-	    is_eth_active_slave_of_bonding_rcu(rdma_ndev, real_dev) ==
-	    BONDING_SLAVE_STATE_INACTIVE) {
-		unsigned long gid_type_mask;
-
+	if (((rdma_ndev != event_ndev &&
+	      !rdma_is_upper_dev_rcu(rdma_ndev, event_ndev)) ||
+	     is_eth_active_slave_of_bonding_rcu(rdma_ndev, real_dev)
+						 ==
+	     BONDING_SLAVE_STATE_INACTIVE)) {
 		rcu_read_unlock();
+		return;
+	}
 
-		gid_type_mask = roce_gid_type_mask_support(ib_dev, port);
+	rcu_read_unlock();
 
-		ib_cache_gid_set_default_gid(ib_dev, port, rdma_ndev,
-					     gid_type_mask,
-					     IB_CACHE_GID_DEFAULT_MODE_DELETE);
-	} else {
-		rcu_read_unlock();
-	}
+	gid_type_mask = roce_gid_type_mask_support(ib_dev, port);
+
+	ib_cache_gid_set_default_gid(ib_dev, port, rdma_ndev,
+				     gid_type_mask,
+				     IB_CACHE_GID_DEFAULT_MODE_DELETE);
 }
 
 static void enum_netdev_ipv4_ips(struct ib_device *ib_dev,
-- 
cgit v1.1


From 514d6c1959f9b396f1b51850925900adedffb951 Mon Sep 17 00:00:00 2001
From: John Fastabend <john.fastabend@gmail.com>
Date: Mon, 23 Apr 2018 12:11:02 -0700
Subject: bpf: Document sockmap '-target bpf' requirement for PROG_TYPE_SK_MSG

BPF_PROG_TYPE_SK_MSG programs use a 'void *' for both data and the
data_end pointers. Additionally, the verifier ensures that every
accesses into the values is a __u64 read. This correctly maps on
to the BPF 64-bit architecture.

However, to ensure that when building on 32bit architectures that
clang uses correct types the '-target bpf' option _must_ be
specified. To make this clear add a note to the Documentation.

Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
---
 Documentation/bpf/bpf_devel_QA.txt | 10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)

diff --git a/Documentation/bpf/bpf_devel_QA.txt b/Documentation/bpf/bpf_devel_QA.txt
index 1a0b704..da57601 100644
--- a/Documentation/bpf/bpf_devel_QA.txt
+++ b/Documentation/bpf/bpf_devel_QA.txt
@@ -557,6 +557,14 @@ A: Although LLVM IR generation and optimization try to stay architecture
        pulls in some header files containing file scope host assembly codes.
      - You can add "-fno-jump-tables" to work around the switch table issue.
 
-   Otherwise, you can use bpf target.
+   Otherwise, you can use bpf target. Additionally, you _must_ use bpf target
+   when:
+
+     - Your program uses data structures with pointer or long / unsigned long
+       types that interface with BPF helpers or context data structures. Access
+       into these structures is verified by the BPF verifier and may result
+       in verification failures if the native architecture is not aligned with
+       the BPF architecture, e.g. 64-bit. An example of this is
+       BPF_PROG_TYPE_SK_MSG require '-target bpf'
 
 Happy BPF hacking!
-- 
cgit v1.1


From 4dfe1bb95235c553e216222cf0c377faf191dacd Mon Sep 17 00:00:00 2001
From: John Fastabend <john.fastabend@gmail.com>
Date: Mon, 23 Apr 2018 12:11:08 -0700
Subject: bpf: sockmap sample use clang flag, -target bpf

Per Documentation/bpf/bpf_devel_QA.txt add the -target flag to the
sockmap Makefile. Relevant text quoted here,

   Otherwise, you can use bpf target. Additionally, you _must_ use
   bpf target when:

 - Your program uses data structures with pointer or long / unsigned
   long types that interface with BPF helpers or context data
   structures. Access into these structures is verified by the BPF
   verifier and may result in verification failures if the native
   architecture is not aligned with the BPF architecture, e.g. 64-bit.
   An example of this is BPF_PROG_TYPE_SK_MSG require '-target bpf'

Fixes: 69e8cc134bcb ("bpf: sockmap sample program")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
---
 samples/sockmap/Makefile | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/samples/sockmap/Makefile b/samples/sockmap/Makefile
index 9bf2881..fa53f4d 100644
--- a/samples/sockmap/Makefile
+++ b/samples/sockmap/Makefile
@@ -65,11 +65,14 @@ $(src)/*.c: verify_target_bpf
 # asm/sysreg.h - inline assembly used by it is incompatible with llvm.
 # But, there is no easy way to fix it, so just exclude it since it is
 # useless for BPF samples.
+#
+# -target bpf option required with SK_MSG programs, this is to ensure
+#  reading 'void *' data types for data and data_end are __u64 reads.
 $(obj)/%.o: $(src)/%.c
 	$(CLANG) $(NOSTDINC_FLAGS) $(LINUXINCLUDE) $(EXTRA_CFLAGS) -I$(obj) \
 		-D__KERNEL__ -D__ASM_SYSREG_H -Wno-unused-value -Wno-pointer-sign \
 		-Wno-compare-distinct-pointer-types \
 		-Wno-gnu-variable-sized-type-not-at-end \
 		-Wno-address-of-packed-member -Wno-tautological-compare \
-		-Wno-unknown-warning-option \
-		-O2 -emit-llvm -c $< -o -| $(LLC) -march=bpf -filetype=obj -o $@
+		-Wno-unknown-warning-option -O2 -target bpf \
+		-emit-llvm -c $< -o -| $(LLC) -march=bpf -filetype=obj -o $@
-- 
cgit v1.1


From ba6b8de423f8d0dee48d6030288ed81c03ddf9f0 Mon Sep 17 00:00:00 2001
From: John Fastabend <john.fastabend@gmail.com>
Date: Mon, 23 Apr 2018 15:39:23 -0700
Subject: bpf: sockmap, map_release does not hold refcnt for pinned maps

Relying on map_release hook to decrement the reference counts when a
map is removed only works if the map is not being pinned. In the
pinned case the ref is decremented immediately and the BPF programs
released. After this BPF programs may not be in-use which is not
what the user would expect.

This patch moves the release logic into bpf_map_put_uref() and brings
sockmap in-line with how a similar case is handled in prog array maps.

Fixes: 3d9e952697de ("bpf: sockmap, fix leaking maps with attached but not detached progs")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
---
 include/linux/bpf.h   | 2 +-
 kernel/bpf/arraymap.c | 3 ++-
 kernel/bpf/sockmap.c  | 4 ++--
 kernel/bpf/syscall.c  | 4 ++--
 4 files changed, 7 insertions(+), 6 deletions(-)

diff --git a/include/linux/bpf.h b/include/linux/bpf.h
index dc586cc..469b20e 100644
--- a/include/linux/bpf.h
+++ b/include/linux/bpf.h
@@ -31,6 +31,7 @@ struct bpf_map_ops {
 	void (*map_release)(struct bpf_map *map, struct file *map_file);
 	void (*map_free)(struct bpf_map *map);
 	int (*map_get_next_key)(struct bpf_map *map, void *key, void *next_key);
+	void (*map_release_uref)(struct bpf_map *map);
 
 	/* funcs callable from userspace and from eBPF programs */
 	void *(*map_lookup_elem)(struct bpf_map *map, void *key);
@@ -436,7 +437,6 @@ int bpf_stackmap_copy(struct bpf_map *map, void *key, void *value);
 int bpf_fd_array_map_update_elem(struct bpf_map *map, struct file *map_file,
 				 void *key, void *value, u64 map_flags);
 int bpf_fd_array_map_lookup_elem(struct bpf_map *map, void *key, u32 *value);
-void bpf_fd_array_map_clear(struct bpf_map *map);
 int bpf_fd_htab_map_update_elem(struct bpf_map *map, struct file *map_file,
 				void *key, void *value, u64 map_flags);
 int bpf_fd_htab_map_lookup_elem(struct bpf_map *map, void *key, u32 *value);
diff --git a/kernel/bpf/arraymap.c b/kernel/bpf/arraymap.c
index 14750e7..027107f 100644
--- a/kernel/bpf/arraymap.c
+++ b/kernel/bpf/arraymap.c
@@ -476,7 +476,7 @@ static u32 prog_fd_array_sys_lookup_elem(void *ptr)
 }
 
 /* decrement refcnt of all bpf_progs that are stored in this map */
-void bpf_fd_array_map_clear(struct bpf_map *map)
+static void bpf_fd_array_map_clear(struct bpf_map *map)
 {
 	struct bpf_array *array = container_of(map, struct bpf_array, map);
 	int i;
@@ -495,6 +495,7 @@ const struct bpf_map_ops prog_array_map_ops = {
 	.map_fd_get_ptr = prog_fd_array_get_ptr,
 	.map_fd_put_ptr = prog_fd_array_put_ptr,
 	.map_fd_sys_lookup_elem = prog_fd_array_sys_lookup_elem,
+	.map_release_uref = bpf_fd_array_map_clear,
 };
 
 static struct bpf_event_entry *bpf_event_entry_gen(struct file *perf_file,
diff --git a/kernel/bpf/sockmap.c b/kernel/bpf/sockmap.c
index a3b2138..a73d484 100644
--- a/kernel/bpf/sockmap.c
+++ b/kernel/bpf/sockmap.c
@@ -1831,7 +1831,7 @@ static int sock_map_update_elem(struct bpf_map *map,
 	return err;
 }
 
-static void sock_map_release(struct bpf_map *map, struct file *map_file)
+static void sock_map_release(struct bpf_map *map)
 {
 	struct bpf_stab *stab = container_of(map, struct bpf_stab, map);
 	struct bpf_prog *orig;
@@ -1855,7 +1855,7 @@ const struct bpf_map_ops sock_map_ops = {
 	.map_get_next_key = sock_map_get_next_key,
 	.map_update_elem = sock_map_update_elem,
 	.map_delete_elem = sock_map_delete_elem,
-	.map_release = sock_map_release,
+	.map_release_uref = sock_map_release,
 };
 
 BPF_CALL_4(bpf_sock_map_update, struct bpf_sock_ops_kern *, bpf_sock,
diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c
index 4ca46df..ebfe9f2 100644
--- a/kernel/bpf/syscall.c
+++ b/kernel/bpf/syscall.c
@@ -257,8 +257,8 @@ static void bpf_map_free_deferred(struct work_struct *work)
 static void bpf_map_put_uref(struct bpf_map *map)
 {
 	if (atomic_dec_and_test(&map->usercnt)) {
-		if (map->map_type == BPF_MAP_TYPE_PROG_ARRAY)
-			bpf_fd_array_map_clear(map);
+		if (map->ops->map_release_uref)
+			map->ops->map_release_uref(map);
 	}
 }
 
-- 
cgit v1.1


From e20f7334837ae47341d8ec4e3170d0b4336a3676 Mon Sep 17 00:00:00 2001
From: John Fastabend <john.fastabend@gmail.com>
Date: Mon, 23 Apr 2018 15:39:28 -0700
Subject: bpf: sockmap, sk_wait_event needed to handle blocking cases

In the recvmsg handler we need to add a wait event to support the
blocking use cases. Without this we return zero and may confuse
user applications. In the wait event any data received on the
sk either via sk_receive_queue or the psock ingress list will
wake up the sock.

Fixes: fa246693a111 ("bpf: sockmap, BPF_F_INGRESS flag for BPF_SK_SKB_STREAM_VERDICT")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
---
 kernel/bpf/sockmap.c | 44 ++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 44 insertions(+)

diff --git a/kernel/bpf/sockmap.c b/kernel/bpf/sockmap.c
index a73d484..aaf50ec 100644
--- a/kernel/bpf/sockmap.c
+++ b/kernel/bpf/sockmap.c
@@ -43,6 +43,7 @@
 #include <net/tcp.h>
 #include <linux/ptr_ring.h>
 #include <net/inet_common.h>
+#include <linux/sched/signal.h>
 
 #define SOCK_CREATE_FLAG_MASK \
 	(BPF_F_NUMA_NODE | BPF_F_RDONLY | BPF_F_WRONLY)
@@ -732,6 +733,26 @@ out_err:
 	return err;
 }
 
+static int bpf_wait_data(struct sock *sk,
+			 struct smap_psock *psk, int flags,
+			 long timeo, int *err)
+{
+	int rc;
+
+	DEFINE_WAIT_FUNC(wait, woken_wake_function);
+
+	add_wait_queue(sk_sleep(sk), &wait);
+	sk_set_bit(SOCKWQ_ASYNC_WAITDATA, sk);
+	rc = sk_wait_event(sk, &timeo,
+			   !list_empty(&psk->ingress) ||
+			   !skb_queue_empty(&sk->sk_receive_queue),
+			   &wait);
+	sk_clear_bit(SOCKWQ_ASYNC_WAITDATA, sk);
+	remove_wait_queue(sk_sleep(sk), &wait);
+
+	return rc;
+}
+
 static int bpf_tcp_recvmsg(struct sock *sk, struct msghdr *msg, size_t len,
 			   int nonblock, int flags, int *addr_len)
 {
@@ -755,6 +776,7 @@ static int bpf_tcp_recvmsg(struct sock *sk, struct msghdr *msg, size_t len,
 		return tcp_recvmsg(sk, msg, len, nonblock, flags, addr_len);
 
 	lock_sock(sk);
+bytes_ready:
 	while (copied != len) {
 		struct scatterlist *sg;
 		struct sk_msg_buff *md;
@@ -809,6 +831,28 @@ static int bpf_tcp_recvmsg(struct sock *sk, struct msghdr *msg, size_t len,
 		}
 	}
 
+	if (!copied) {
+		long timeo;
+		int data;
+		int err = 0;
+
+		timeo = sock_rcvtimeo(sk, nonblock);
+		data = bpf_wait_data(sk, psock, flags, timeo, &err);
+
+		if (data) {
+			if (!skb_queue_empty(&sk->sk_receive_queue)) {
+				release_sock(sk);
+				smap_release_sock(psock, sk);
+				copied = tcp_recvmsg(sk, msg, len, nonblock, flags, addr_len);
+				return copied;
+			}
+			goto bytes_ready;
+		}
+
+		if (err)
+			copied = err;
+	}
+
 	release_sock(sk);
 	smap_release_sock(psock, sk);
 	return copied;
-- 
cgit v1.1


From 4fcfdfb83391c74e62683469289db42a143440ac Mon Sep 17 00:00:00 2001
From: John Fastabend <john.fastabend@gmail.com>
Date: Mon, 23 Apr 2018 15:39:33 -0700
Subject: bpf: sockmap, fix double page_put on ENOMEM error in redirect path

In the case where the socket memory boundary is hit the redirect
path returns an ENOMEM error. However, before checking for this
condition the redirect scatterlist buffer is setup with a valid
page and length. This is never unwound so when the buffers are
released latter in the error path we do a put_page() and clear
the scatterlist fields. But, because the initial error happens
before completing the scatterlist buffer we end up with both the
original buffer and the redirect buffer pointing to the same page
resulting in duplicate put_page() calls.

To fix this simply move the initial configuration of the redirect
scatterlist buffer below the sock memory check.

Found this while running TCP_STREAM test with netperf using Cilium.

Fixes: fa246693a111 ("bpf: sockmap, BPF_F_INGRESS flag for BPF_SK_SKB_STREAM_VERDICT")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
---
 kernel/bpf/sockmap.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/kernel/bpf/sockmap.c b/kernel/bpf/sockmap.c
index aaf50ec..634415c 100644
--- a/kernel/bpf/sockmap.c
+++ b/kernel/bpf/sockmap.c
@@ -524,8 +524,6 @@ static int bpf_tcp_ingress(struct sock *sk, int apply_bytes,
 	i = md->sg_start;
 
 	do {
-		r->sg_data[i] = md->sg_data[i];
-
 		size = (apply && apply_bytes < md->sg_data[i].length) ?
 			apply_bytes : md->sg_data[i].length;
 
@@ -536,6 +534,7 @@ static int bpf_tcp_ingress(struct sock *sk, int apply_bytes,
 		}
 
 		sk_mem_charge(sk, size);
+		r->sg_data[i] = md->sg_data[i];
 		r->sg_data[i].length = size;
 		md->sg_data[i].length -= size;
 		md->sg_data[i].offset += size;
-- 
cgit v1.1


From 672e314b21dc614894e69bb56a2b55cc7d256810 Mon Sep 17 00:00:00 2001
From: Matt Atwood <matthew.s.atwood@intel.com>
Date: Mon, 23 Apr 2018 15:28:03 -0700
Subject: drm/i915/kbl: Add KBL GT2 sku

Adding a missing GT2 sku discovered off hardware.

Signed-off-by: Matt Atwood <matthew.s.atwood@intel.com>
Reviewed-by: Clint Taylor <clinton.a.taylor@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1524522483-19987-1-git-send-email-matthew.s.atwood@intel.com
---
 include/drm/i915_pciids.h | 1 +
 1 file changed, 1 insertion(+)

diff --git a/include/drm/i915_pciids.h b/include/drm/i915_pciids.h
index 70f0c25..bab70ff 100644
--- a/include/drm/i915_pciids.h
+++ b/include/drm/i915_pciids.h
@@ -349,6 +349,7 @@
 #define INTEL_KBL_GT2_IDS(info)	\
 	INTEL_VGA_DEVICE(0x5916, info), /* ULT GT2 */ \
 	INTEL_VGA_DEVICE(0x5917, info), /* Mobile GT2 */ \
+	INTEL_VGA_DEVICE(0x591C, info), /* ULX GT2 */ \
 	INTEL_VGA_DEVICE(0x5921, info), /* ULT GT2F */ \
 	INTEL_VGA_DEVICE(0x591E, info), /* ULX GT2 */ \
 	INTEL_VGA_DEVICE(0x5912, info), /* DT  GT2 */ \
-- 
cgit v1.1


From 9c71b2c53bf56efa2f8aeab4247a86c0178fdcd7 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Pali=20Roh=C3=A1r?= <pali.rohar@gmail.com>
Date: Mon, 23 Apr 2018 16:25:58 -0700
Subject: Input: alps - fix reporting pressure of v3 trackstick
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

According to documentation, all 7 lower bits represents trackpoint pressure.

Fixes: 4621c9660459 ("Input: alps - report pressure of v3 and v7 trackstick")
Signed-off-by: Pali Rohár <pali.rohar@gmail.com>
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
---
 drivers/input/mouse/alps.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/input/mouse/alps.c b/drivers/input/mouse/alps.c
index 0a67f23..38f9501 100644
--- a/drivers/input/mouse/alps.c
+++ b/drivers/input/mouse/alps.c
@@ -583,7 +583,7 @@ static void alps_process_trackstick_packet_v3(struct psmouse *psmouse)
 
 	x = (s8)(((packet[0] & 0x20) << 2) | (packet[1] & 0x7f));
 	y = (s8)(((packet[0] & 0x10) << 3) | (packet[2] & 0x7f));
-	z = packet[4] & 0x7c;
+	z = packet[4] & 0x7f;
 
 	/*
 	 * The x and y values tend to be quite large, and when used
-- 
cgit v1.1


From 6f226cff7a1e80de2a67175d26156c6354734a73 Mon Sep 17 00:00:00 2001
From: Masanari Iida <standby24x7@gmail.com>
Date: Mon, 23 Apr 2018 12:01:21 -0700
Subject: Input: hideep_ts - fix a typo in Kconfig

This patch fixes a spelling error found in Kconfig.

Signed-off-by: Masanari Iida <standby24x7@gmail.com>
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
---
 drivers/input/touchscreen/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/input/touchscreen/Kconfig b/drivers/input/touchscreen/Kconfig
index 4f15496..3e613af 100644
--- a/drivers/input/touchscreen/Kconfig
+++ b/drivers/input/touchscreen/Kconfig
@@ -362,7 +362,7 @@ config TOUCHSCREEN_HIDEEP
 
 	  If unsure, say N.
 
-	  To compile this driver as a moudle, choose M here : the
+	  To compile this driver as a module, choose M here : the
 	  module will be called hideep_ts.
 
 config TOUCHSCREEN_ILI210X
-- 
cgit v1.1


From 596ea7aad431cb88d7a6f3a2b6019cf3221b8d02 Mon Sep 17 00:00:00 2001
From: Wolfram Sang <wsa@the-dreams.de>
Date: Mon, 23 Apr 2018 12:02:49 -0700
Subject: MAINTAINERS: Rakesh Iyer can't be reached anymore

The current mail address is rejected, last activity (with a different
address) in git-history is from 2012. Remove this.

Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
Acked-by: Jon Hunter <jonathanh@nvidia.com>
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
---
 MAINTAINERS | 1 -
 1 file changed, 1 deletion(-)

diff --git a/MAINTAINERS b/MAINTAINERS
index 4623caf..17323bf 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -13640,7 +13640,6 @@ S:	Supported
 F:	drivers/iommu/tegra*
 
 TEGRA KBC DRIVER
-M:	Rakesh Iyer <riyer@nvidia.com>
 M:	Laxman Dewangan <ldewangan@nvidia.com>
 S:	Supported
 F:	drivers/input/keyboard/tegra-kbc.c
-- 
cgit v1.1


From f0316f93897c4c4e67278b175bfbfd3a95ba650a Mon Sep 17 00:00:00 2001
From: Russell King <rmk+kernel@armlinux.org.uk>
Date: Sat, 5 Dec 2015 18:41:28 +0000
Subject: drm/i2c: tda9950: add CEC driver

Add a CEC driver for the TDA9950, which is a stand-alone I2C CEC device,
but is also integrated into HDMI transceivers such as the TDA9989 and
TDA19989.

The TDA9950 contains a command processor which handles retransmissions
and the low level bus protocol.  The driver just has to read and write
the messages, and handle error conditions.

Reviewed-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
---
 drivers/gpu/drm/i2c/Kconfig           |   5 +
 drivers/gpu/drm/i2c/Makefile          |   1 +
 drivers/gpu/drm/i2c/tda9950.c         | 509 ++++++++++++++++++++++++++++++++++
 include/linux/platform_data/tda9950.h |  16 ++
 4 files changed, 531 insertions(+)
 create mode 100644 drivers/gpu/drm/i2c/tda9950.c
 create mode 100644 include/linux/platform_data/tda9950.h

diff --git a/drivers/gpu/drm/i2c/Kconfig b/drivers/gpu/drm/i2c/Kconfig
index a6c92be..3a232f5 100644
--- a/drivers/gpu/drm/i2c/Kconfig
+++ b/drivers/gpu/drm/i2c/Kconfig
@@ -26,4 +26,9 @@ config DRM_I2C_NXP_TDA998X
 	help
 	  Support for NXP Semiconductors TDA998X HDMI encoders.
 
+config DRM_I2C_NXP_TDA9950
+	tristate "NXP Semiconductors TDA9950/TDA998X HDMI CEC"
+	select CEC_NOTIFIER
+	select CEC_CORE
+
 endmenu
diff --git a/drivers/gpu/drm/i2c/Makefile b/drivers/gpu/drm/i2c/Makefile
index b20100c..a962f6f 100644
--- a/drivers/gpu/drm/i2c/Makefile
+++ b/drivers/gpu/drm/i2c/Makefile
@@ -7,3 +7,4 @@ obj-$(CONFIG_DRM_I2C_SIL164) += sil164.o
 
 tda998x-y := tda998x_drv.o
 obj-$(CONFIG_DRM_I2C_NXP_TDA998X) += tda998x.o
+obj-$(CONFIG_DRM_I2C_NXP_TDA9950) += tda9950.o
diff --git a/drivers/gpu/drm/i2c/tda9950.c b/drivers/gpu/drm/i2c/tda9950.c
new file mode 100644
index 0000000..3f7396ca
--- /dev/null
+++ b/drivers/gpu/drm/i2c/tda9950.c
@@ -0,0 +1,509 @@
+/*
+ *  TDA9950 Consumer Electronics Control driver
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * The NXP TDA9950 implements the HDMI Consumer Electronics Control
+ * interface.  The host interface is similar to a mailbox: the data
+ * registers starting at REG_CDR0 are written to send a command to the
+ * internal CPU, and replies are read from these registers.
+ *
+ * As the data registers represent a mailbox, they must be accessed
+ * as a single I2C transaction.  See the TDA9950 data sheet for details.
+ */
+#include <linux/delay.h>
+#include <linux/i2c.h>
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/platform_data/tda9950.h>
+#include <linux/slab.h>
+#include <drm/drm_edid.h>
+#include <media/cec.h>
+#include <media/cec-notifier.h>
+
+enum {
+	REG_CSR = 0x00,
+	CSR_BUSY = BIT(7),
+	CSR_INT  = BIT(6),
+	CSR_ERR  = BIT(5),
+
+	REG_CER = 0x01,
+
+	REG_CVR = 0x02,
+
+	REG_CCR = 0x03,
+	CCR_RESET = BIT(7),
+	CCR_ON    = BIT(6),
+
+	REG_ACKH = 0x04,
+	REG_ACKL = 0x05,
+
+	REG_CCONR = 0x06,
+	CCONR_ENABLE_ERROR = BIT(4),
+	CCONR_RETRY_MASK = 7,
+
+	REG_CDR0 = 0x07,
+
+	CDR1_REQ = 0x00,
+	CDR1_CNF = 0x01,
+	CDR1_IND = 0x81,
+	CDR1_ERR = 0x82,
+	CDR1_IER = 0x83,
+
+	CDR2_CNF_SUCCESS    = 0x00,
+	CDR2_CNF_OFF_STATE  = 0x80,
+	CDR2_CNF_BAD_REQ    = 0x81,
+	CDR2_CNF_CEC_ACCESS = 0x82,
+	CDR2_CNF_ARB_ERROR  = 0x83,
+	CDR2_CNF_BAD_TIMING = 0x84,
+	CDR2_CNF_NACK_ADDR  = 0x85,
+	CDR2_CNF_NACK_DATA  = 0x86,
+};
+
+struct tda9950_priv {
+	struct i2c_client *client;
+	struct device *hdmi;
+	struct cec_adapter *adap;
+	struct tda9950_glue *glue;
+	u16 addresses;
+	struct cec_msg rx_msg;
+	struct cec_notifier *notify;
+	bool open;
+};
+
+static int tda9950_write_range(struct i2c_client *client, u8 addr, u8 *p, int cnt)
+{
+	struct i2c_msg msg;
+	u8 buf[cnt + 1];
+	int ret;
+
+	buf[0] = addr;
+	memcpy(buf + 1, p, cnt);
+
+	msg.addr = client->addr;
+	msg.flags = 0;
+	msg.len = cnt + 1;
+	msg.buf = buf;
+
+	dev_dbg(&client->dev, "wr 0x%02x: %*ph\n", addr, cnt, p);
+
+	ret = i2c_transfer(client->adapter, &msg, 1);
+	if (ret < 0)
+		dev_err(&client->dev, "Error %d writing to cec:0x%x\n", ret, addr);
+	return ret < 0 ? ret : 0;
+}
+
+static void tda9950_write(struct i2c_client *client, u8 addr, u8 val)
+{
+	tda9950_write_range(client, addr, &val, 1);
+}
+
+static int tda9950_read_range(struct i2c_client *client, u8 addr, u8 *p, int cnt)
+{
+	struct i2c_msg msg[2];
+	int ret;
+
+	msg[0].addr = client->addr;
+	msg[0].flags = 0;
+	msg[0].len = 1;
+	msg[0].buf = &addr;
+	msg[1].addr = client->addr;
+	msg[1].flags = I2C_M_RD;
+	msg[1].len = cnt;
+	msg[1].buf = p;
+
+	ret = i2c_transfer(client->adapter, msg, 2);
+	if (ret < 0)
+		dev_err(&client->dev, "Error %d reading from cec:0x%x\n", ret, addr);
+
+	dev_dbg(&client->dev, "rd 0x%02x: %*ph\n", addr, cnt, p);
+
+	return ret;
+}
+
+static u8 tda9950_read(struct i2c_client *client, u8 addr)
+{
+	int ret;
+	u8 val;
+
+	ret = tda9950_read_range(client, addr, &val, 1);
+	if (ret < 0)
+		val = 0;
+
+	return val;
+}
+
+static irqreturn_t tda9950_irq(int irq, void *data)
+{
+	struct tda9950_priv *priv = data;
+	unsigned int tx_status;
+	u8 csr, cconr, buf[19];
+	u8 arb_lost_cnt, nack_cnt, err_cnt;
+
+	if (!priv->open)
+		return IRQ_NONE;
+
+	csr = tda9950_read(priv->client, REG_CSR);
+	if (!(csr & CSR_INT))
+		return IRQ_NONE;
+
+	cconr = tda9950_read(priv->client, REG_CCONR) & CCONR_RETRY_MASK;
+
+	tda9950_read_range(priv->client, REG_CDR0, buf, sizeof(buf));
+
+	/*
+	 * This should never happen: the data sheet says that there will
+	 * always be a valid message if the interrupt line is asserted.
+	 */
+	if (buf[0] == 0) {
+		dev_warn(&priv->client->dev, "interrupt pending, but no message?\n");
+		return IRQ_NONE;
+	}
+
+	switch (buf[1]) {
+	case CDR1_CNF: /* transmit result */
+		arb_lost_cnt = nack_cnt = err_cnt = 0;
+		switch (buf[2]) {
+		case CDR2_CNF_SUCCESS:
+			tx_status = CEC_TX_STATUS_OK;
+			break;
+
+		case CDR2_CNF_ARB_ERROR:
+			tx_status = CEC_TX_STATUS_ARB_LOST;
+			arb_lost_cnt = cconr;
+			break;
+
+		case CDR2_CNF_NACK_ADDR:
+			tx_status = CEC_TX_STATUS_NACK;
+			nack_cnt = cconr;
+			break;
+
+		default: /* some other error, refer to TDA9950 docs */
+			dev_err(&priv->client->dev, "CNF reply error 0x%02x\n",
+				buf[2]);
+			tx_status = CEC_TX_STATUS_ERROR;
+			err_cnt = cconr;
+			break;
+		}
+		/* TDA9950 executes all retries for us */
+		tx_status |= CEC_TX_STATUS_MAX_RETRIES;
+		cec_transmit_done(priv->adap, tx_status, arb_lost_cnt,
+				  nack_cnt, 0, err_cnt);
+		break;
+
+	case CDR1_IND:
+		priv->rx_msg.len = buf[0] - 2;
+		if (priv->rx_msg.len > CEC_MAX_MSG_SIZE)
+			priv->rx_msg.len = CEC_MAX_MSG_SIZE;
+
+		memcpy(priv->rx_msg.msg, buf + 2, priv->rx_msg.len);
+		cec_received_msg(priv->adap, &priv->rx_msg);
+		break;
+
+	default: /* unknown */
+		dev_err(&priv->client->dev, "unknown service id 0x%02x\n",
+			buf[1]);
+		break;
+	}
+
+	return IRQ_HANDLED;
+}
+
+static int tda9950_cec_transmit(struct cec_adapter *adap, u8 attempts,
+				u32 signal_free_time, struct cec_msg *msg)
+{
+	struct tda9950_priv *priv = adap->priv;
+	u8 buf[CEC_MAX_MSG_SIZE + 2];
+
+	buf[0] = 2 + msg->len;
+	buf[1] = CDR1_REQ;
+	memcpy(buf + 2, msg->msg, msg->len);
+
+	if (attempts > 5)
+		attempts = 5;
+
+	tda9950_write(priv->client, REG_CCONR, attempts);
+
+	return tda9950_write_range(priv->client, REG_CDR0, buf, 2 + msg->len);
+}
+
+static int tda9950_cec_adap_log_addr(struct cec_adapter *adap, u8 addr)
+{
+	struct tda9950_priv *priv = adap->priv;
+	u16 addresses;
+	u8 buf[2];
+
+	if (addr == CEC_LOG_ADDR_INVALID)
+		addresses = priv->addresses = 0;
+	else
+		addresses = priv->addresses |= BIT(addr);
+
+	/* TDA9950 doesn't want address 15 set */
+	addresses &= 0x7fff;
+	buf[0] = addresses >> 8;
+	buf[1] = addresses;
+
+	return tda9950_write_range(priv->client, REG_ACKH, buf, 2);
+}
+
+/*
+ * When operating as part of the TDA998x, we need additional handling
+ * to initialise and shut down the TDA9950 part of the device.  These
+ * two hooks are provided to allow the TDA998x code to perform those
+ * activities.
+ */
+static int tda9950_glue_open(struct tda9950_priv *priv)
+{
+	int ret = 0;
+
+	if (priv->glue && priv->glue->open)
+		ret = priv->glue->open(priv->glue->data);
+
+	priv->open = true;
+
+	return ret;
+}
+
+static void tda9950_glue_release(struct tda9950_priv *priv)
+{
+	priv->open = false;
+
+	if (priv->glue && priv->glue->release)
+		priv->glue->release(priv->glue->data);
+}
+
+static int tda9950_open(struct tda9950_priv *priv)
+{
+	struct i2c_client *client = priv->client;
+	int ret;
+
+	ret = tda9950_glue_open(priv);
+	if (ret)
+		return ret;
+
+	/* Reset the TDA9950, and wait 250ms for it to recover */
+	tda9950_write(client, REG_CCR, CCR_RESET);
+	msleep(250);
+
+	tda9950_cec_adap_log_addr(priv->adap, CEC_LOG_ADDR_INVALID);
+
+	/* Start the command processor */
+	tda9950_write(client, REG_CCR, CCR_ON);
+
+	return 0;
+}
+
+static void tda9950_release(struct tda9950_priv *priv)
+{
+	struct i2c_client *client = priv->client;
+	int timeout = 50;
+	u8 csr;
+
+	/* Stop the command processor */
+	tda9950_write(client, REG_CCR, 0);
+
+	/* Wait up to .5s for it to signal non-busy */
+	do {
+		csr = tda9950_read(client, REG_CSR);
+		if (!(csr & CSR_BUSY) || --timeout)
+			break;
+		msleep(10);
+	} while (1);
+
+	/* Warn the user that their IRQ may die if it's shared. */
+	if (csr & CSR_BUSY)
+		dev_warn(&client->dev, "command processor failed to stop, irq%d may die (csr=0x%02x)\n",
+			 client->irq, csr);
+
+	tda9950_glue_release(priv);
+}
+
+static int tda9950_cec_adap_enable(struct cec_adapter *adap, bool enable)
+{
+	struct tda9950_priv *priv = adap->priv;
+
+	if (!enable) {
+		tda9950_release(priv);
+		return 0;
+	} else {
+		return tda9950_open(priv);
+	}
+}
+
+static const struct cec_adap_ops tda9950_cec_ops = {
+	.adap_enable = tda9950_cec_adap_enable,
+	.adap_log_addr = tda9950_cec_adap_log_addr,
+	.adap_transmit = tda9950_cec_transmit,
+};
+
+/*
+ * When operating as part of the TDA998x, we need to claim additional
+ * resources.  These two hooks permit the management of those resources.
+ */
+static void tda9950_devm_glue_exit(void *data)
+{
+	struct tda9950_glue *glue = data;
+
+	if (glue && glue->exit)
+		glue->exit(glue->data);
+}
+
+static int tda9950_devm_glue_init(struct device *dev, struct tda9950_glue *glue)
+{
+	int ret;
+
+	if (glue && glue->init) {
+		ret = glue->init(glue->data);
+		if (ret)
+			return ret;
+	}
+
+	ret = devm_add_action(dev, tda9950_devm_glue_exit, glue);
+	if (ret)
+		tda9950_devm_glue_exit(glue);
+
+	return ret;
+}
+
+static void tda9950_cec_del(void *data)
+{
+	struct tda9950_priv *priv = data;
+
+	cec_delete_adapter(priv->adap);
+}
+
+static int tda9950_probe(struct i2c_client *client,
+			 const struct i2c_device_id *id)
+{
+	struct tda9950_glue *glue = client->dev.platform_data;
+	struct device *dev = &client->dev;
+	struct tda9950_priv *priv;
+	unsigned long irqflags;
+	int ret;
+	u8 cvr;
+
+	/*
+	 * We must have I2C functionality: our multi-byte accesses
+	 * must be performed as a single contiguous transaction.
+	 */
+	if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C)) {
+		dev_err(&client->dev,
+			"adapter does not support I2C functionality\n");
+		return -ENXIO;
+	}
+
+	/* We must have an interrupt to be functional. */
+	if (client->irq <= 0) {
+		dev_err(&client->dev, "driver requires an interrupt\n");
+		return -ENXIO;
+	}
+
+	priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
+	if (!priv)
+		return -ENOMEM;
+
+	priv->client = client;
+	priv->glue = glue;
+
+	i2c_set_clientdata(client, priv);
+
+	/*
+	 * If we're part of a TDA998x, we want the class devices to be
+	 * associated with the HDMI Tx so we have a tight relationship
+	 * between the HDMI interface and the CEC interface.
+	 */
+	priv->hdmi = dev;
+	if (glue && glue->parent)
+		priv->hdmi = glue->parent;
+
+	priv->adap = cec_allocate_adapter(&tda9950_cec_ops, priv, "tda9950",
+					  CEC_CAP_DEFAULTS,
+					  CEC_MAX_LOG_ADDRS);
+	if (IS_ERR(priv->adap))
+		return PTR_ERR(priv->adap);
+
+	ret = devm_add_action(dev, tda9950_cec_del, priv);
+	if (ret) {
+		cec_delete_adapter(priv->adap);
+		return ret;
+	}
+
+	ret = tda9950_devm_glue_init(dev, glue);
+	if (ret)
+		return ret;
+
+	ret = tda9950_glue_open(priv);
+	if (ret)
+		return ret;
+
+	cvr = tda9950_read(client, REG_CVR);
+
+	dev_info(&client->dev,
+		 "TDA9950 CEC interface, hardware version %u.%u\n",
+		 cvr >> 4, cvr & 15);
+
+	tda9950_glue_release(priv);
+
+	irqflags = IRQF_TRIGGER_FALLING;
+	if (glue)
+		irqflags = glue->irq_flags;
+
+	ret = devm_request_threaded_irq(dev, client->irq, NULL, tda9950_irq,
+					irqflags | IRQF_SHARED | IRQF_ONESHOT,
+					dev_name(&client->dev), priv);
+	if (ret < 0)
+		return ret;
+
+	priv->notify = cec_notifier_get(priv->hdmi);
+	if (!priv->notify)
+		return -ENOMEM;
+
+	ret = cec_register_adapter(priv->adap, priv->hdmi);
+	if (ret < 0) {
+		cec_notifier_put(priv->notify);
+		return ret;
+	}
+
+	/*
+	 * CEC documentation says we must not call cec_delete_adapter
+	 * after a successful call to cec_register_adapter().
+	 */
+	devm_remove_action(dev, tda9950_cec_del, priv);
+
+	cec_register_cec_notifier(priv->adap, priv->notify);
+
+	return 0;
+}
+
+static int tda9950_remove(struct i2c_client *client)
+{
+	struct tda9950_priv *priv = i2c_get_clientdata(client);
+
+	cec_unregister_adapter(priv->adap);
+	cec_notifier_put(priv->notify);
+
+	return 0;
+}
+
+static struct i2c_device_id tda9950_ids[] = {
+	{ "tda9950", 0 },
+	{ },
+};
+MODULE_DEVICE_TABLE(i2c, tda9950_ids);
+
+static struct i2c_driver tda9950_driver = {
+	.probe = tda9950_probe,
+	.remove = tda9950_remove,
+	.driver = {
+		.name = "tda9950",
+	},
+	.id_table = tda9950_ids,
+};
+
+module_i2c_driver(tda9950_driver);
+
+MODULE_AUTHOR("Russell King <rmk+kernel@armlinux.org.uk>");
+MODULE_DESCRIPTION("TDA9950/TDA998x Consumer Electronics Control Driver");
+MODULE_LICENSE("GPL v2");
diff --git a/include/linux/platform_data/tda9950.h b/include/linux/platform_data/tda9950.h
new file mode 100644
index 0000000..c65efd4
--- /dev/null
+++ b/include/linux/platform_data/tda9950.h
@@ -0,0 +1,16 @@
+#ifndef LINUX_PLATFORM_DATA_TDA9950_H
+#define LINUX_PLATFORM_DATA_TDA9950_H
+
+struct device;
+
+struct tda9950_glue {
+	struct device *parent;
+	unsigned long irq_flags;
+	void *data;
+	int (*init)(void *);
+	void (*exit)(void *);
+	int (*open)(void *);
+	void (*release)(void *);
+};
+
+#endif
-- 
cgit v1.1


From 7e8675f000bc7e20f4efb72cf624f4109301002b Mon Sep 17 00:00:00 2001
From: Russell King <rmk+kernel@armlinux.org.uk>
Date: Wed, 5 Oct 2016 12:47:50 +0100
Subject: drm/i2c: tda998x: add CEC support

The TDA998x is a HDMI transmitter with a TDA9950 CEC engine integrated
onto the same die.  Add support for the TDA9950 CEC engine to the
TDA998x driver.

Reviewed-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
---
 drivers/gpu/drm/i2c/Kconfig       |   1 +
 drivers/gpu/drm/i2c/tda998x_drv.c | 195 ++++++++++++++++++++++++++++++++++++--
 2 files changed, 187 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/i2c/Kconfig b/drivers/gpu/drm/i2c/Kconfig
index 3a232f5..65d3acb 100644
--- a/drivers/gpu/drm/i2c/Kconfig
+++ b/drivers/gpu/drm/i2c/Kconfig
@@ -22,6 +22,7 @@ config DRM_I2C_SIL164
 config DRM_I2C_NXP_TDA998X
 	tristate "NXP Semiconductors TDA998X HDMI encoder"
 	default m if DRM_TILCDC
+	select CEC_CORE if CEC_NOTIFIER
 	select SND_SOC_HDMI_CODEC if SND_SOC
 	help
 	  Support for NXP Semiconductors TDA998X HDMI encoders.
diff --git a/drivers/gpu/drm/i2c/tda998x_drv.c b/drivers/gpu/drm/i2c/tda998x_drv.c
index 16e0439..eb9916b 100644
--- a/drivers/gpu/drm/i2c/tda998x_drv.c
+++ b/drivers/gpu/drm/i2c/tda998x_drv.c
@@ -16,8 +16,10 @@
  */
 
 #include <linux/component.h>
+#include <linux/gpio/consumer.h>
 #include <linux/hdmi.h>
 #include <linux/module.h>
+#include <linux/platform_data/tda9950.h>
 #include <linux/irq.h>
 #include <sound/asoundef.h>
 #include <sound/hdmi-codec.h>
@@ -29,6 +31,8 @@
 #include <drm/drm_of.h>
 #include <drm/i2c/tda998x.h>
 
+#include <media/cec-notifier.h>
+
 #define DBG(fmt, ...) DRM_DEBUG(fmt"\n", ##__VA_ARGS__)
 
 struct tda998x_audio_port {
@@ -55,6 +59,7 @@ struct tda998x_priv {
 	struct platform_device *audio_pdev;
 	struct mutex audio_mutex;
 
+	struct mutex edid_mutex;
 	wait_queue_head_t wq_edid;
 	volatile int wq_edid_wait;
 
@@ -67,6 +72,9 @@ struct tda998x_priv {
 	struct drm_connector connector;
 
 	struct tda998x_audio_port audio_port[2];
+	struct tda9950_glue cec_glue;
+	struct gpio_desc *calib;
+	struct cec_notifier *cec_notify;
 };
 
 #define conn_to_tda998x_priv(x) \
@@ -345,6 +353,12 @@ struct tda998x_priv {
 #define REG_CEC_INTSTATUS	  0xee		      /* read */
 # define CEC_INTSTATUS_CEC	  (1 << 0)
 # define CEC_INTSTATUS_HDMI	  (1 << 1)
+#define REG_CEC_CAL_XOSC_CTRL1    0xf2
+# define CEC_CAL_XOSC_CTRL1_ENA_CAL	BIT(0)
+#define REG_CEC_DES_FREQ2         0xf5
+# define CEC_DES_FREQ2_DIS_AUTOCAL BIT(7)
+#define REG_CEC_CLK               0xf6
+# define CEC_CLK_FRO              0x11
 #define REG_CEC_FRO_IM_CLK_CTRL   0xfb                /* read/write */
 # define CEC_FRO_IM_CLK_CTRL_GHOST_DIS (1 << 7)
 # define CEC_FRO_IM_CLK_CTRL_ENA_OTP   (1 << 6)
@@ -359,6 +373,7 @@ struct tda998x_priv {
 # define CEC_RXSHPDLEV_HPD        (1 << 1)
 
 #define REG_CEC_ENAMODS           0xff                /* read/write */
+# define CEC_ENAMODS_EN_CEC_CLK   (1 << 7)
 # define CEC_ENAMODS_DIS_FRO      (1 << 6)
 # define CEC_ENAMODS_DIS_CCLK     (1 << 5)
 # define CEC_ENAMODS_EN_RXSENS    (1 << 2)
@@ -417,6 +432,114 @@ cec_read(struct tda998x_priv *priv, u8 addr)
 	return val;
 }
 
+static void cec_enamods(struct tda998x_priv *priv, u8 mods, bool enable)
+{
+	int val = cec_read(priv, REG_CEC_ENAMODS);
+
+	if (val < 0)
+		return;
+
+	if (enable)
+		val |= mods;
+	else
+		val &= ~mods;
+
+	cec_write(priv, REG_CEC_ENAMODS, val);
+}
+
+static void tda998x_cec_set_calibration(struct tda998x_priv *priv, bool enable)
+{
+	if (enable) {
+		u8 val;
+
+		cec_write(priv, 0xf3, 0xc0);
+		cec_write(priv, 0xf4, 0xd4);
+
+		/* Enable automatic calibration mode */
+		val = cec_read(priv, REG_CEC_DES_FREQ2);
+		val &= ~CEC_DES_FREQ2_DIS_AUTOCAL;
+		cec_write(priv, REG_CEC_DES_FREQ2, val);
+
+		/* Enable free running oscillator */
+		cec_write(priv, REG_CEC_CLK, CEC_CLK_FRO);
+		cec_enamods(priv, CEC_ENAMODS_DIS_FRO, false);
+
+		cec_write(priv, REG_CEC_CAL_XOSC_CTRL1,
+			  CEC_CAL_XOSC_CTRL1_ENA_CAL);
+	} else {
+		cec_write(priv, REG_CEC_CAL_XOSC_CTRL1, 0);
+	}
+}
+
+/*
+ * Calibration for the internal oscillator: we need to set calibration mode,
+ * and then pulse the IRQ line low for a 10ms ± 1% period.
+ */
+static void tda998x_cec_calibration(struct tda998x_priv *priv)
+{
+	struct gpio_desc *calib = priv->calib;
+
+	mutex_lock(&priv->edid_mutex);
+	if (priv->hdmi->irq > 0)
+		disable_irq(priv->hdmi->irq);
+	gpiod_direction_output(calib, 1);
+	tda998x_cec_set_calibration(priv, true);
+
+	local_irq_disable();
+	gpiod_set_value(calib, 0);
+	mdelay(10);
+	gpiod_set_value(calib, 1);
+	local_irq_enable();
+
+	tda998x_cec_set_calibration(priv, false);
+	gpiod_direction_input(calib);
+	if (priv->hdmi->irq > 0)
+		enable_irq(priv->hdmi->irq);
+	mutex_unlock(&priv->edid_mutex);
+}
+
+static int tda998x_cec_hook_init(void *data)
+{
+	struct tda998x_priv *priv = data;
+	struct gpio_desc *calib;
+
+	calib = gpiod_get(&priv->hdmi->dev, "nxp,calib", GPIOD_ASIS);
+	if (IS_ERR(calib)) {
+		dev_warn(&priv->hdmi->dev, "failed to get calibration gpio: %ld\n",
+			 PTR_ERR(calib));
+		return PTR_ERR(calib);
+	}
+
+	priv->calib = calib;
+
+	return 0;
+}
+
+static void tda998x_cec_hook_exit(void *data)
+{
+	struct tda998x_priv *priv = data;
+
+	gpiod_put(priv->calib);
+	priv->calib = NULL;
+}
+
+static int tda998x_cec_hook_open(void *data)
+{
+	struct tda998x_priv *priv = data;
+
+	cec_enamods(priv, CEC_ENAMODS_EN_CEC_CLK | CEC_ENAMODS_EN_CEC, true);
+	tda998x_cec_calibration(priv);
+
+	return 0;
+}
+
+static void tda998x_cec_hook_release(void *data)
+{
+	struct tda998x_priv *priv = data;
+
+	cec_enamods(priv, CEC_ENAMODS_EN_CEC_CLK | CEC_ENAMODS_EN_CEC, false);
+}
+
 static int
 set_page(struct tda998x_priv *priv, u16 reg)
 {
@@ -657,10 +780,13 @@ static irqreturn_t tda998x_irq_thread(int irq, void *data)
 			sta, cec, lvl, flag0, flag1, flag2);
 
 		if (cec & CEC_RXSHPDINT_HPD) {
-			if (lvl & CEC_RXSHPDLEV_HPD)
+			if (lvl & CEC_RXSHPDLEV_HPD) {
 				tda998x_edid_delay_start(priv);
-			else
+			} else {
 				schedule_work(&priv->detect_work);
+				cec_notifier_set_phys_addr(priv->cec_notify,
+						   CEC_PHYS_ADDR_INVALID);
+			}
 
 			handled = true;
 		}
@@ -981,6 +1107,8 @@ static int tda998x_connector_fill_modes(struct drm_connector *connector,
 	if (connector->edid_blob_ptr) {
 		struct edid *edid = (void *)connector->edid_blob_ptr->data;
 
+		cec_notifier_set_phys_addr_from_edid(priv->cec_notify, edid);
+
 		priv->sink_has_audio = drm_detect_monitor_audio(edid);
 	} else {
 		priv->sink_has_audio = false;
@@ -1024,6 +1152,8 @@ static int read_edid_block(void *data, u8 *buf, unsigned int blk, size_t length)
 	offset = (blk & 1) ? 128 : 0;
 	segptr = blk / 2;
 
+	mutex_lock(&priv->edid_mutex);
+
 	reg_write(priv, REG_DDC_ADDR, 0xa0);
 	reg_write(priv, REG_DDC_OFFS, offset);
 	reg_write(priv, REG_DDC_SEGM_ADDR, 0x60);
@@ -1043,14 +1173,15 @@ static int read_edid_block(void *data, u8 *buf, unsigned int blk, size_t length)
 					msecs_to_jiffies(100));
 		if (i < 0) {
 			dev_err(&priv->hdmi->dev, "read edid wait err %d\n", i);
-			return i;
+			ret = i;
+			goto failed;
 		}
 	} else {
 		for (i = 100; i > 0; i--) {
 			msleep(1);
 			ret = reg_read(priv, REG_INT_FLAGS_2);
 			if (ret < 0)
-				return ret;
+				goto failed;
 			if (ret & INT_FLAGS_2_EDID_BLK_RD)
 				break;
 		}
@@ -1058,17 +1189,22 @@ static int read_edid_block(void *data, u8 *buf, unsigned int blk, size_t length)
 
 	if (i == 0) {
 		dev_err(&priv->hdmi->dev, "read edid timeout\n");
-		return -ETIMEDOUT;
+		ret = -ETIMEDOUT;
+		goto failed;
 	}
 
 	ret = reg_read_range(priv, REG_EDID_DATA_0, buf, length);
 	if (ret != length) {
 		dev_err(&priv->hdmi->dev, "failed to read edid block %d: %d\n",
 			blk, ret);
-		return ret;
+		goto failed;
 	}
 
-	return 0;
+	ret = 0;
+
+ failed:
+	mutex_unlock(&priv->edid_mutex);
+	return ret;
 }
 
 static int tda998x_connector_get_modes(struct drm_connector *connector)
@@ -1423,6 +1559,9 @@ static void tda998x_destroy(struct tda998x_priv *priv)
 	cancel_work_sync(&priv->detect_work);
 
 	i2c_unregister_device(priv->cec);
+
+	if (priv->cec_notify)
+		cec_notifier_put(priv->cec_notify);
 }
 
 /* I2C driver functions */
@@ -1472,11 +1611,13 @@ static int tda998x_get_audio_ports(struct tda998x_priv *priv,
 static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 {
 	struct device_node *np = client->dev.of_node;
+	struct i2c_board_info cec_info;
 	u32 video;
 	int rev_lo, rev_hi, ret;
 
 	mutex_init(&priv->mutex);	/* protect the page access */
 	mutex_init(&priv->audio_mutex); /* protect access from audio thread */
+	mutex_init(&priv->edid_mutex);
 	init_waitqueue_head(&priv->edid_delay_waitq);
 	timer_setup(&priv->edid_delay_timer, tda998x_edid_delay_done, 0);
 	INIT_WORK(&priv->detect_work, tda998x_detect_work);
@@ -1564,6 +1705,9 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 
 		irq_flags =
 			irqd_get_trigger_type(irq_get_irq_data(client->irq));
+
+		priv->cec_glue.irq_flags = irq_flags;
+
 		irq_flags |= IRQF_SHARED | IRQF_ONESHOT;
 		ret = request_threaded_irq(client->irq, NULL,
 					   tda998x_irq_thread, irq_flags,
@@ -1579,7 +1723,34 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 		cec_write(priv, REG_CEC_RXSHPDINTENA, CEC_RXSHPDLEV_HPD);
 	}
 
-	priv->cec = i2c_new_dummy(client->adapter, priv->cec_addr);
+	priv->cec_notify = cec_notifier_get(&client->dev);
+	if (!priv->cec_notify) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	priv->cec_glue.parent = &client->dev;
+	priv->cec_glue.data = priv;
+	priv->cec_glue.init = tda998x_cec_hook_init;
+	priv->cec_glue.exit = tda998x_cec_hook_exit;
+	priv->cec_glue.open = tda998x_cec_hook_open;
+	priv->cec_glue.release = tda998x_cec_hook_release;
+
+	/*
+	 * Some TDA998x are actually two I2C devices merged onto one piece
+	 * of silicon: TDA9989 and TDA19989 combine the HDMI transmitter
+	 * with a slightly modified TDA9950 CEC device.  The CEC device
+	 * is at the TDA9950 address, with the address pins strapped across
+	 * to the TDA998x address pins.  Hence, it always has the same
+	 * offset.
+	 */
+	memset(&cec_info, 0, sizeof(cec_info));
+	strlcpy(cec_info.type, "tda9950", sizeof(cec_info.type));
+	cec_info.addr = priv->cec_addr;
+	cec_info.platform_data = &priv->cec_glue;
+	cec_info.irq = client->irq;
+
+	priv->cec = i2c_new_device(client->adapter, &cec_info);
 	if (!priv->cec) {
 		ret = -ENODEV;
 		goto fail;
@@ -1609,10 +1780,16 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 	return 0;
 
 fail:
+	/* if encoder_init fails, the encoder slave is never registered,
+	 * so cleanup here:
+	 */
+	if (priv->cec)
+		i2c_unregister_device(priv->cec);
+	if (priv->cec_notify)
+		cec_notifier_put(priv->cec_notify);
 	if (client->irq)
 		free_irq(client->irq, priv);
 err_irq:
-	i2c_unregister_device(priv->cec);
 	return ret;
 }
 
-- 
cgit v1.1


From ba52762fb1430b2a2ea8127c1a292c15f13b8dac Mon Sep 17 00:00:00 2001
From: Russell King <rmk+kernel@armlinux.org.uk>
Date: Wed, 29 Nov 2017 10:27:12 +0000
Subject: dt-bindings: tda998x: add the calibration gpio

Add the optional calibration gpio for integrated TDA9950 CEC support.
This GPIO corresponds with the interrupt from the TDA998x, as the
calibration requires driving the interrupt pin low.

Reviewed-by: Rob Herring <robh@kernel.org>
Acked-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
---
 Documentation/devicetree/bindings/display/bridge/tda998x.txt | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/Documentation/devicetree/bindings/display/bridge/tda998x.txt b/Documentation/devicetree/bindings/display/bridge/tda998x.txt
index 24cc246..1a4eaca 100644
--- a/Documentation/devicetree/bindings/display/bridge/tda998x.txt
+++ b/Documentation/devicetree/bindings/display/bridge/tda998x.txt
@@ -27,6 +27,9 @@ Optional properties:
 	in question is used. The implementation allows one or two DAIs. If two
 	DAIs are defined, they must be of different type.
 
+  - nxp,calib-gpios: calibration GPIO, which must correspond with the
+	gpio used for the TDA998x interrupt pin.
+
 [1] Documentation/sound/alsa/soc/DAI.txt
 [2] include/dt-bindings/display/tda998x.h
 
-- 
cgit v1.1


From a083429e133df63bf2e618f51e4061649fb3c65e Mon Sep 17 00:00:00 2001
From: Luca Coelho <luciano.coelho@intel.com>
Date: Mon, 23 Apr 2018 16:01:31 +0300
Subject: iwlwifi: mvm: fix old scan version sizes

When version 8 of the scan command API was introduced, only the size
of version 7 was updated, causing older versions of the firmware to
throw BAD_COMMAND errors.

Calculating the old version based on the size of the latest version
got too complicated and the size of the older versions will never
change anyway, so it's better to just hardcoded the sizes.

Fixes: 66fa2424df16 ("iwlwifi: fw api: support the new scan request FW API version")
Reported-by: Scott Register <sreg@sreg.io>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
---
 drivers/net/wireless/intel/iwlwifi/fw/api/scan.h | 13 +++++--------
 1 file changed, 5 insertions(+), 8 deletions(-)

diff --git a/drivers/net/wireless/intel/iwlwifi/fw/api/scan.h b/drivers/net/wireless/intel/iwlwifi/fw/api/scan.h
index 7af3a0f5..a17c4a7 100644
--- a/drivers/net/wireless/intel/iwlwifi/fw/api/scan.h
+++ b/drivers/net/wireless/intel/iwlwifi/fw/api/scan.h
@@ -8,6 +8,7 @@
  * Copyright(c) 2012 - 2014 Intel Corporation. All rights reserved.
  * Copyright(c) 2013 - 2015 Intel Mobile Communications GmbH
  * Copyright(c) 2016 - 2017 Intel Deutschland GmbH
+ * Copyright(c) 2018        Intel Corporation
  *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of version 2 of the GNU General Public License as
@@ -30,7 +31,7 @@
  * Copyright(c) 2012 - 2014 Intel Corporation. All rights reserved.
  * Copyright(c) 2013 - 2015 Intel Mobile Communications GmbH
  * Copyright(c) 2016 - 2017 Intel Deutschland GmbH
- * Copyright(c) 2018 Intel Corporation
+ * Copyright(c) 2018        Intel Corporation
  * All rights reserved.
  *
  * Redistribution and use in source and binary forms, with or without
@@ -749,13 +750,9 @@ struct iwl_scan_req_umac {
 } __packed;
 
 #define IWL_SCAN_REQ_UMAC_SIZE_V8 sizeof(struct iwl_scan_req_umac)
-#define IWL_SCAN_REQ_UMAC_SIZE_V7 (sizeof(struct iwl_scan_req_umac) - \
-					 4 * sizeof(u8))
-#define IWL_SCAN_REQ_UMAC_SIZE_V6 (sizeof(struct iwl_scan_req_umac) - \
-				   2 * sizeof(u8) - sizeof(__le16))
-#define IWL_SCAN_REQ_UMAC_SIZE_V1 (sizeof(struct iwl_scan_req_umac) - \
-				   2 * sizeof(__le32) - 2 * sizeof(u8) - \
-				   sizeof(__le16))
+#define IWL_SCAN_REQ_UMAC_SIZE_V7 48
+#define IWL_SCAN_REQ_UMAC_SIZE_V6 44
+#define IWL_SCAN_REQ_UMAC_SIZE_V1 36
 
 /**
  * struct iwl_umac_scan_abort
-- 
cgit v1.1


From af8a41cccf8f469165c6debc8fe07c5fd2ca501a Mon Sep 17 00:00:00 2001
From: Ping-Ke Shih <pkshih@realtek.com>
Date: Fri, 20 Apr 2018 10:30:09 +0800
Subject: rtlwifi: cleanup 8723be ant_sel definition

Some HP laptops have only a single wifi antenna. This would not be a
problem except that they were shipped with an incorrectly encoded
EFUSE. It should have been possible to open the computer and transfer
the antenna connection to the other terminal except that such action
might void the warranty, and moving the antenna broke the Windows
driver. The fix was to add a module option that would override the
EFUSE encoding. That was done with commit c18d8f509571 ("rtlwifi:
rtl8723be: Add antenna select module parameter"). There was still a
problem with Bluetooth coexistence, which was addressed with commit
baa170229095 ("rtlwifi: btcoexist: Implement antenna selection").
There were still problems, thus there were commit 0ff78adeef11
("rtlwifi: rtl8723be: fix ant_sel code") and commit 6d6226928369
("rtlwifi: btcoexist: Fix antenna selection code"). Despite all these
attempts at fixing the problem, the code is not yet right. A proper
fix is important as there are now instances of laptops having
RTL8723DE chips with the same problem.

The module parameter ant_sel is used to control antenna number and path.
At present enum ANT_{X2,X1} is used to define the antenna number, but
this choice is not intuitive, thus change to a new enum ANT_{MAIN,AUX}
to make it more readable. This change showed examples where incorrect
values were used. It was also possible to remove a workaround in
halbtcoutsrc.c.

The experimental results with single antenna connected to specific path
are now as follows:
  ant_sel  ANT_MAIN(#1)  ANT_AUX(#2)
     0        -8            -62
     1        -62           -10
     2        -6            -60

Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Fixes: c18d8f509571 ("rtlwifi: rtl8723be: Add antenna select module parameter")
Fixes: baa170229095 ("rtlwifi: btcoexist: Implement antenna selection")
Fixes: 0ff78adeef11 ("rtlwifi: rtl8723be: fix ant_sel code")
Fixes: 6d6226928369 ("rtlwifi: btcoexist: Fix antenna selection code")
Cc: Stable <stable@vger.kernel.org> # 4.7+
Reviewed-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
---
 .../net/wireless/realtek/rtlwifi/btcoexist/halbtcoutsrc.c | 15 ---------------
 drivers/net/wireless/realtek/rtlwifi/rtl8723be/hw.c       | 11 +++++++----
 drivers/net/wireless/realtek/rtlwifi/wifi.h               |  5 +++++
 3 files changed, 12 insertions(+), 19 deletions(-)

diff --git a/drivers/net/wireless/realtek/rtlwifi/btcoexist/halbtcoutsrc.c b/drivers/net/wireless/realtek/rtlwifi/btcoexist/halbtcoutsrc.c
index 8b6b07a..b026e80 100644
--- a/drivers/net/wireless/realtek/rtlwifi/btcoexist/halbtcoutsrc.c
+++ b/drivers/net/wireless/realtek/rtlwifi/btcoexist/halbtcoutsrc.c
@@ -158,16 +158,6 @@ static u8 halbtc_get_wifi_central_chnl(struct btc_coexist *btcoexist)
 
 static u8 rtl_get_hwpg_single_ant_path(struct rtl_priv *rtlpriv)
 {
-	struct rtl_mod_params *mod_params = rtlpriv->cfg->mod_params;
-
-	/* override ant_num / ant_path */
-	if (mod_params->ant_sel) {
-		rtlpriv->btcoexist.btc_info.ant_num =
-			(mod_params->ant_sel == 1 ? ANT_X2 : ANT_X1);
-
-		rtlpriv->btcoexist.btc_info.single_ant_path =
-			(mod_params->ant_sel == 1 ? 0 : 1);
-	}
 	return rtlpriv->btcoexist.btc_info.single_ant_path;
 }
 
@@ -178,7 +168,6 @@ static u8 rtl_get_hwpg_bt_type(struct rtl_priv *rtlpriv)
 
 static u8 rtl_get_hwpg_ant_num(struct rtl_priv *rtlpriv)
 {
-	struct rtl_mod_params *mod_params = rtlpriv->cfg->mod_params;
 	u8 num;
 
 	if (rtlpriv->btcoexist.btc_info.ant_num == ANT_X2)
@@ -186,10 +175,6 @@ static u8 rtl_get_hwpg_ant_num(struct rtl_priv *rtlpriv)
 	else
 		num = 1;
 
-	/* override ant_num / ant_path */
-	if (mod_params->ant_sel)
-		num = (mod_params->ant_sel == 1 ? ANT_X2 : ANT_X1) + 1;
-
 	return num;
 }
 
diff --git a/drivers/net/wireless/realtek/rtlwifi/rtl8723be/hw.c b/drivers/net/wireless/realtek/rtlwifi/rtl8723be/hw.c
index e7bbbc9..b4f3f91 100644
--- a/drivers/net/wireless/realtek/rtlwifi/rtl8723be/hw.c
+++ b/drivers/net/wireless/realtek/rtlwifi/rtl8723be/hw.c
@@ -848,6 +848,9 @@ static bool _rtl8723be_init_mac(struct ieee80211_hw *hw)
 		return false;
 	}
 
+	if (rtlpriv->cfg->ops->get_btc_status())
+		rtlpriv->btcoexist.btc_ops->btc_power_on_setting(rtlpriv);
+
 	bytetmp = rtl_read_byte(rtlpriv, REG_MULTI_FUNC_CTRL);
 	rtl_write_byte(rtlpriv, REG_MULTI_FUNC_CTRL, bytetmp | BIT(3));
 
@@ -2696,21 +2699,21 @@ void rtl8723be_read_bt_coexist_info_from_hwpg(struct ieee80211_hw *hw,
 		rtlpriv->btcoexist.btc_info.bt_type = BT_RTL8723B;
 		rtlpriv->btcoexist.btc_info.ant_num = (value & 0x1);
 		rtlpriv->btcoexist.btc_info.single_ant_path =
-			 (value & 0x40);	/*0xc3[6]*/
+			 (value & 0x40 ? ANT_AUX : ANT_MAIN);	/*0xc3[6]*/
 	} else {
 		rtlpriv->btcoexist.btc_info.btcoexist = 0;
 		rtlpriv->btcoexist.btc_info.bt_type = BT_RTL8723B;
 		rtlpriv->btcoexist.btc_info.ant_num = ANT_X2;
-		rtlpriv->btcoexist.btc_info.single_ant_path = 0;
+		rtlpriv->btcoexist.btc_info.single_ant_path = ANT_MAIN;
 	}
 
 	/* override ant_num / ant_path */
 	if (mod_params->ant_sel) {
 		rtlpriv->btcoexist.btc_info.ant_num =
-			(mod_params->ant_sel == 1 ? ANT_X2 : ANT_X1);
+			(mod_params->ant_sel == 1 ? ANT_X1 : ANT_X2);
 
 		rtlpriv->btcoexist.btc_info.single_ant_path =
-			(mod_params->ant_sel == 1 ? 0 : 1);
+			(mod_params->ant_sel == 1 ? ANT_AUX : ANT_MAIN);
 	}
 }
 
diff --git a/drivers/net/wireless/realtek/rtlwifi/wifi.h b/drivers/net/wireless/realtek/rtlwifi/wifi.h
index d27e339..ce17540 100644
--- a/drivers/net/wireless/realtek/rtlwifi/wifi.h
+++ b/drivers/net/wireless/realtek/rtlwifi/wifi.h
@@ -2823,6 +2823,11 @@ enum bt_ant_num {
 	ANT_X1 = 1,
 };
 
+enum bt_ant_path {
+	ANT_MAIN = 0,
+	ANT_AUX = 1,
+};
+
 enum bt_co_type {
 	BT_2WIRE = 0,
 	BT_ISSC_3WIRE = 1,
-- 
cgit v1.1


From 3085982c6b45d7d22f76e3aa018affbc143a7370 Mon Sep 17 00:00:00 2001
From: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Date: Mon, 23 Apr 2018 14:37:53 +0300
Subject: drm/i915: Use ktime on wait_for
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

We use jiffies to determine when wait expires. However
Imre did find out that jiffies can and will do a >1
increments on certain situations [1]. When this happens
in a wait_for loop, we return timeout errorneously
much earlier than what the real wallclock would say.

We can't afford our waits to timeout prematurely.
Discard jiffies and change to ktime to detect timeouts.

v2: added bugzilla entry (Imre), added stable (Chris)

Reported-by: Imre Deak <imre.deak@intel.com>
References: https://lkml.org/lkml/2018/4/18/798 [1]
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105771
Cc: Imre Deak <imre.deak@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20180423113754.28424-1-mika.kuoppala@linux.intel.com
---
 drivers/gpu/drm/i915/intel_drv.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
index 44ed248..33ff263 100644
--- a/drivers/gpu/drm/i915/intel_drv.h
+++ b/drivers/gpu/drm/i915/intel_drv.h
@@ -49,12 +49,12 @@
  * check the condition before the timeout.
  */
 #define __wait_for(OP, COND, US, Wmin, Wmax) ({ \
-	unsigned long timeout__ = jiffies + usecs_to_jiffies(US) + 1;	\
+	const ktime_t end__ = ktime_add_ns(ktime_get_raw(), 1000ll * (US)); \
 	long wait__ = (Wmin); /* recommended min for usleep is 10 us */	\
 	int ret__;							\
 	might_sleep();							\
 	for (;;) {							\
-		bool expired__ = time_after(jiffies, timeout__);	\
+		const bool expired__ = ktime_after(ktime_get_raw(), end__); \
 		OP;							\
 		if (COND) {						\
 			ret__ = 0;					\
-- 
cgit v1.1


From 1c3c1dc66a9664a9b7048c3869fa7863be9123a0 Mon Sep 17 00:00:00 2001
From: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Date: Mon, 23 Apr 2018 14:37:54 +0300
Subject: drm/i915: Add compiler barrier to wait_for

We need to be careful to not let compiler evaluate
the expiration and the operation on it's terms.

Document and enforce that COND will be evaluated
before checking timeout expiration.

Suggested-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20180423113754.28424-2-mika.kuoppala@linux.intel.com
---
 drivers/gpu/drm/i915/intel_drv.h | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
index 33ff263..58868b9 100644
--- a/drivers/gpu/drm/i915/intel_drv.h
+++ b/drivers/gpu/drm/i915/intel_drv.h
@@ -56,6 +56,8 @@
 	for (;;) {							\
 		const bool expired__ = ktime_after(ktime_get_raw(), end__); \
 		OP;							\
+		/* Guarantee COND check prior to timeout */		\
+		barrier();						\
 		if (COND) {						\
 			ret__ = 0;					\
 			break;						\
@@ -96,6 +98,8 @@
 		u64 now = local_clock(); \
 		if (!(ATOMIC)) \
 			preempt_enable(); \
+		/* Guarantee COND check prior to timeout */ \
+		barrier(); \
 		if (COND) { \
 			ret = 0; \
 			break; \
-- 
cgit v1.1


From 970f04c840f6004a1a956c1a836792a341d63eae Mon Sep 17 00:00:00 2001
From: Rob Herring <robh@kernel.org>
Date: Fri, 20 Apr 2018 08:08:23 -0500
Subject: dtc: checks: drop warning for missing PCI bridge bus-range

Cherry-picked from dtc upstream commit e1f139ea4900fd0324c646822b4061fec6e08321.

Having a 'bus-range' property for PCI bridges should not be required,
so remove the warning when missing. There was some confusion with the
Linux kernel printing a message that no property is present and the OS
assigned the bus number. This message was intended to be informational
rather than a warning.

When the firmware doesn't enumerate the PCI bus and leaves it up to the
OS to do, then it is perfectly fine for the OS to assign bus numbers
and bus-range is not necessary.

There are a few cases where bus-range is needed or useful as Arnd
Bergmann summarized:

- Traditionally Linux avoided using multiple PCI domains, but instead
  configured separate PCI host bridges to have non-overlapping
  bus ranges so we can present them to user space as a single
  domain, and run the kernel without CONFIG_PCI_DOMAINS.
  Specifying the bus ranges this way would and give stable bus
  numbers across boots when the probe order is not fixed.

- On certain ARM64 systems, we must only use the first
  128 bus numbers based on the way the IOMMU identifies
  the device with truncated bus/dev/fn number. There are probably
  others like this, with various limitations.

- To leave some room for hotplugged devices, each slot on
  a host bridge can in theory get a range of bus numbers
  that are available when assigning bus numbers at boot time

Cc: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Rob Herring <robh@kernel.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
---
 scripts/dtc/checks.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/scripts/dtc/checks.c b/scripts/dtc/checks.c
index c07ba4d..815eaf1 100644
--- a/scripts/dtc/checks.c
+++ b/scripts/dtc/checks.c
@@ -787,10 +787,9 @@ static void check_pci_bridge(struct check *c, struct dt_info *dti, struct node *
 		FAIL(c, dti, node, "incorrect #size-cells for PCI bridge");
 
 	prop = get_property(node, "bus-range");
-	if (!prop) {
-		FAIL(c, dti, node, "missing bus-range for PCI bridge");
+	if (!prop)
 		return;
-	}
+
 	if (prop->val.len != (sizeof(cell_t) * 2)) {
 		FAIL_PROP(c, dti, node, prop, "value must be 2 cells");
 		return;
-- 
cgit v1.1


From 13b86f50eaaddaea4bdd2fe476fd12e6a0951add Mon Sep 17 00:00:00 2001
From: Hans de Goede <hdegoede@redhat.com>
Date: Sun, 22 Apr 2018 19:56:17 +0200
Subject: thermal: int3403_thermal: Fix NULL pointer deref on module load /
 probe

Starting with kernel 4.17 thermal_cooling_device_register() will call the
get_max_state() op during register.

Since we deref priv->priv in int3403_get_max_state() this means we must
set priv->priv before calling thermal_cooling_device_register().

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
---
 drivers/thermal/int340x_thermal/int3403_thermal.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/thermal/int340x_thermal/int3403_thermal.c b/drivers/thermal/int340x_thermal/int3403_thermal.c
index 8a7f24d..0c19fcd5 100644
--- a/drivers/thermal/int340x_thermal/int3403_thermal.c
+++ b/drivers/thermal/int340x_thermal/int3403_thermal.c
@@ -194,6 +194,7 @@ static int int3403_cdev_add(struct int3403_priv *priv)
 		return -EFAULT;
 	}
 
+	priv->priv = obj;
 	obj->max_state = p->package.count - 1;
 	obj->cdev =
 		thermal_cooling_device_register(acpi_device_bid(priv->adev),
@@ -201,8 +202,6 @@ static int int3403_cdev_add(struct int3403_priv *priv)
 	if (IS_ERR(obj->cdev))
 		result = PTR_ERR(obj->cdev);
 
-	priv->priv = obj;
-
 	kfree(buf.pointer);
 	/* TODO: add ACPI notification support */
 
-- 
cgit v1.1


From 247870ac8ea72916cd26f89e9bc211b97141ecd5 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Tue, 24 Apr 2018 02:08:39 +0100
Subject: drm/i915: Build request info on stack before printk

printk unhelpfully inserts a '\n' between consecutive calls, and since
our drm_printf wrapper may be emitting info a seq_file instead,
KERN_CONT is not an option. To work with any drm_printf destination, we
need to build up the output into a temporary buf on the stack and then
feed the complete line in a single call to printk.

Fixes: b7268c5eed0a ("drm/i915: Pack params to engine->schedule() into a struct")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180424010839.22860-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/intel_engine_cs.c | 24 +++++++++++++++---------
 1 file changed, 15 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
index be608f7..66cddd0 100644
--- a/drivers/gpu/drm/i915/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/intel_engine_cs.c
@@ -1113,14 +1113,17 @@ unsigned int intel_engines_has_context_isolation(struct drm_i915_private *i915)
 	return which;
 }
 
-static void print_sched_attr(struct drm_printer *m,
-			     const struct drm_i915_private *i915,
-			     const struct i915_sched_attr *attr)
+static int print_sched_attr(struct drm_i915_private *i915,
+			    const struct i915_sched_attr *attr,
+			    char *buf, int x, int len)
 {
 	if (attr->priority == I915_PRIORITY_INVALID)
-		return;
+		return x;
+
+	x += snprintf(buf + x, len - x,
+		      " prio=%d", attr->priority);
 
-	drm_printf(m, "prio=%d", attr->priority);
+	return x;
 }
 
 static void print_request(struct drm_printer *m,
@@ -1128,14 +1131,17 @@ static void print_request(struct drm_printer *m,
 			  const char *prefix)
 {
 	const char *name = rq->fence.ops->get_timeline_name(&rq->fence);
+	char buf[80];
+	int x = 0;
+
+	x = print_sched_attr(rq->i915, &rq->sched.attr, buf, x, sizeof(buf));
 
-	drm_printf(m, "%s%x%s [%llx:%x] ",
+	drm_printf(m, "%s%x%s [%llx:%x]%s @ %dms: %s\n",
 		   prefix,
 		   rq->global_seqno,
 		   i915_request_completed(rq) ? "!" : "",
-		   rq->fence.context, rq->fence.seqno);
-	print_sched_attr(m, rq->i915, &rq->sched.attr);
-	drm_printf(m, " @ %dms: %s\n",
+		   rq->fence.context, rq->fence.seqno,
+		   buf,
 		   jiffies_to_msecs(jiffies - rq->emitted_jiffies),
 		   name);
 }
-- 
cgit v1.1


From 56021f48dbea69a00b96a53d6450b0950f9c811f Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Tue, 24 Apr 2018 09:16:00 +0100
Subject: drm/i915: Don't dump umpteen thousand requests

If we have more than a few, possibly several thousand request in the
queue, don't show the central portion, just the first few and the last
being executed and/or queued. The first few should be enough to help
identify a problem in execution, and most often comparing the first/last
in the queue is enough to identify problems in the scheduling.

We may need some fine tuning to set MAX_REQUESTS_TO_SHOW for common
debug scenarios, but for the moment if we can avoiding spending more
than a few seconds dumping the GPU state that will avoid a nasty
livelock (where hangcheck spends so long dumping the state, it fires
again and starts to dump the state again in parallel, ad infinitum).

v2: Remember to print last not the stale rq iter after the loop.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180424081600.27544-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/intel_engine_cs.c | 43 ++++++++++++++++++++++++++++++----
 1 file changed, 38 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
index 66cddd0..2398ea7 100644
--- a/drivers/gpu/drm/i915/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/intel_engine_cs.c
@@ -1307,11 +1307,13 @@ void intel_engine_dump(struct intel_engine_cs *engine,
 		       struct drm_printer *m,
 		       const char *header, ...)
 {
+	const int MAX_REQUESTS_TO_SHOW = 8;
 	struct intel_breadcrumbs * const b = &engine->breadcrumbs;
 	const struct intel_engine_execlists * const execlists = &engine->execlists;
 	struct i915_gpu_error * const error = &engine->i915->gpu_error;
-	struct i915_request *rq;
+	struct i915_request *rq, *last;
 	struct rb_node *rb;
+	int count;
 
 	if (header) {
 		va_list ap;
@@ -1378,16 +1380,47 @@ void intel_engine_dump(struct intel_engine_cs *engine,
 	}
 
 	spin_lock_irq(&engine->timeline->lock);
-	list_for_each_entry(rq, &engine->timeline->requests, link)
-		print_request(m, rq, "\t\tE ");
+
+	last = NULL;
+	count = 0;
+	list_for_each_entry(rq, &engine->timeline->requests, link) {
+		if (count++ < MAX_REQUESTS_TO_SHOW - 1)
+			print_request(m, rq, "\t\tE ");
+		else
+			last = rq;
+	}
+	if (last) {
+		if (count > MAX_REQUESTS_TO_SHOW) {
+			drm_printf(m,
+				   "\t\t...skipping %d executing requests...\n",
+				   count - MAX_REQUESTS_TO_SHOW);
+		}
+		print_request(m, last, "\t\tE ");
+	}
+
+	last = NULL;
+	count = 0;
 	drm_printf(m, "\t\tQueue priority: %d\n", execlists->queue_priority);
 	for (rb = execlists->first; rb; rb = rb_next(rb)) {
 		struct i915_priolist *p =
 			rb_entry(rb, typeof(*p), node);
 
-		list_for_each_entry(rq, &p->requests, sched.link)
-			print_request(m, rq, "\t\tQ ");
+		list_for_each_entry(rq, &p->requests, sched.link) {
+			if (count++ < MAX_REQUESTS_TO_SHOW - 1)
+				print_request(m, rq, "\t\tQ ");
+			else
+				last = rq;
+		}
 	}
+	if (last) {
+		if (count > MAX_REQUESTS_TO_SHOW) {
+			drm_printf(m,
+				   "\t\t...skipping %d queued requests...\n",
+				   count - MAX_REQUESTS_TO_SHOW);
+		}
+		print_request(m, last, "\t\tQ ");
+	}
+
 	spin_unlock_irq(&engine->timeline->lock);
 
 	spin_lock_irq(&b->rb_lock);
-- 
cgit v1.1


From aaab22bcd157a0d081d3a747475fb26f4c549157 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Tue, 24 Apr 2018 12:52:36 +0100
Subject: drm/i915: Skip printing global offsets for per-engine scratch pages

Knowing the offset of the per-engine scratch/HWS page during boot is not
very informative, so remove the DRM_DEBUG.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180424115236.2022-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/intel_engine_cs.c | 5 -----
 1 file changed, 5 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
index 2398ea7..58be7fa 100644
--- a/drivers/gpu/drm/i915/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/intel_engine_cs.c
@@ -541,8 +541,6 @@ int intel_engine_create_scratch(struct intel_engine_cs *engine, int size)
 		goto err_unref;
 
 	engine->scratch = vma;
-	DRM_DEBUG_DRIVER("%s pipe control offset: 0x%08x\n",
-			 engine->name, i915_ggtt_offset(vma));
 	return 0;
 
 err_unref:
@@ -636,9 +634,6 @@ static int init_status_page(struct intel_engine_cs *engine)
 	engine->status_page.vma = vma;
 	engine->status_page.ggtt_offset = i915_ggtt_offset(vma);
 	engine->status_page.page_addr = memset(vaddr, 0, PAGE_SIZE);
-
-	DRM_DEBUG_DRIVER("%s hws offset: 0x%08x\n",
-			 engine->name, i915_ggtt_offset(vma));
 	return 0;
 
 err_unpin:
-- 
cgit v1.1


From df9e6521749ab33cde306e8a4350b0ac7889220a Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jos=C3=A9=20Roberto=20de=20Souza?= <jose.souza@intel.com>
Date: Wed, 18 Apr 2018 16:41:58 -0700
Subject: drm/i915/fbdev: Enable late fbdev initial configuration
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

If the initial fbdev configuration (intel_fbdev_initial_config()) runs
and there still no sink connected it will cause
drm_fb_helper_initial_config() to return 0 as no error happened (but
internally the return is -EAGAIN).  Because no framebuffer was
allocated, when a sink is connected intel_fbdev_output_poll_changed()
will not execute drm_fb_helper_hotplug_event() that would trigger
another try to do the initial fbdev configuration.

So here allowing drm_fb_helper_hotplug_event() to be executed when there
is no framebuffer allocated and fbdev was not set up yet.

This issue also happens when a MST DP sink is connected since boot, as
the MST topology is discovered in parallel if
intel_fbdev_initial_config() is executed before the first sink MST is
discovered it will cause this same issue.

This is a follow-up patch of
https://patchwork.freedesktop.org/patch/196089/

Changes from v1:
- not creating a dump framebuffer anymore, instead just allowing
  drm_fb_helper_hotplug_event() to execute when fbdev is not setup yet.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=104158
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=104425
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: stable@vger.kernel.org # v4.15+
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Tested-by: Paul Menzel <pmenzel@molgen.mpg.de>
Tested-by: frederik <frederik.schwan@linux.com> # 4.15.17
Tested-by: Ian Pilcher <arequipeno@gmail.com>
Acked-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180418234158.9388-1-jose.souza@intel.com
---
 drivers/gpu/drm/i915/intel_fbdev.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/intel_fbdev.c b/drivers/gpu/drm/i915/intel_fbdev.c
index 65a3313..c1c31b4 100644
--- a/drivers/gpu/drm/i915/intel_fbdev.c
+++ b/drivers/gpu/drm/i915/intel_fbdev.c
@@ -807,7 +807,7 @@ void intel_fbdev_output_poll_changed(struct drm_device *dev)
 		return;
 
 	intel_fbdev_sync(ifbdev);
-	if (ifbdev->vma)
+	if (ifbdev->vma || ifbdev->helper.deferred_setup)
 		drm_fb_helper_hotplug_event(&ifbdev->helper);
 }
 
-- 
cgit v1.1


From a3997159133d56e444f0c0f56ab1ae59863912a8 Mon Sep 17 00:00:00 2001
From: "Gustavo A. R. Silva" <gustavo@embeddedor.com>
Date: Tue, 24 Apr 2018 08:15:45 -0500
Subject: drm/i915/selftests: Fix uninitialized variable

There is a potential execution path in which variable err is
returned without being properly initialized previously.

Fix this by initializing variable err to 0.

Addresses-Coverity-ID: 1468362 ("Uninitialized scalar variable")
Fixes: f4ecfbfc32ed ("drm/i915: Check whitelist registers across resets")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20180424131545.GA4053@embeddedor.com
---
 drivers/gpu/drm/i915/selftests/intel_workarounds.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/selftests/intel_workarounds.c b/drivers/gpu/drm/i915/selftests/intel_workarounds.c
index 5455b26..17444a3 100644
--- a/drivers/gpu/drm/i915/selftests/intel_workarounds.c
+++ b/drivers/gpu/drm/i915/selftests/intel_workarounds.c
@@ -239,7 +239,7 @@ static int live_reset_whitelist(void *arg)
 	struct intel_engine_cs *engine = i915->engine[RCS];
 	struct i915_gpu_error *error = &i915->gpu_error;
 	struct whitelist w;
-	int err;
+	int err = 0;
 
 	/* If we reset the gpu, we should not lose the RING_NONPRIV */
 
-- 
cgit v1.1


From 36a501a199336fbf87299ebbe598d0af00922949 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Tue, 24 Apr 2018 15:29:45 +0100
Subject: drm/i915/breadcrumbs: Keep the fake irq armed across reset

Instead of synchronously cancelling the timer and re-enabling it inside
the reset callbacks, keep the timer enabled and let it die on its next
wakeup if no longer required. This allows
intel_engine_reset_breadcrumbs() to be used from an atomic
(timer/softirq) context such as required for resetting an engine.

It also allows us to react better to the user poking around debugfs for
testing missed irqs.

v2: Tighten the order of del_timer_sync as the fake_irq timer
may trigger the hangcheck timer, and so we should cancel it first and
then cancel the hangcheck (Mika)

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180424142945.6787-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/intel_breadcrumbs.c | 29 ++++++++++++++++++++---------
 1 file changed, 20 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_breadcrumbs.c b/drivers/gpu/drm/i915/intel_breadcrumbs.c
index 671a6d6..5ce4f51 100644
--- a/drivers/gpu/drm/i915/intel_breadcrumbs.c
+++ b/drivers/gpu/drm/i915/intel_breadcrumbs.c
@@ -130,11 +130,12 @@ static void intel_breadcrumbs_hangcheck(struct timer_list *t)
 
 static void intel_breadcrumbs_fake_irq(struct timer_list *t)
 {
-	struct intel_engine_cs *engine = from_timer(engine, t,
-						    breadcrumbs.fake_irq);
+	struct intel_engine_cs *engine =
+		from_timer(engine, t, breadcrumbs.fake_irq);
 	struct intel_breadcrumbs *b = &engine->breadcrumbs;
 
-	/* The timer persists in case we cannot enable interrupts,
+	/*
+	 * The timer persists in case we cannot enable interrupts,
 	 * or if we have previously seen seqno/interrupt incoherency
 	 * ("missed interrupt" syndrome, better known as a "missed breadcrumb").
 	 * Here the worker will wake up every jiffie in order to kick the
@@ -148,6 +149,12 @@ static void intel_breadcrumbs_fake_irq(struct timer_list *t)
 	if (!b->irq_armed)
 		return;
 
+	/* If the user has disabled the fake-irq, restore the hangchecking */
+	if (!test_bit(engine->id, &engine->i915->gpu_error.missed_irq_rings)) {
+		mod_timer(&b->hangcheck, wait_timeout());
+		return;
+	}
+
 	mod_timer(&b->fake_irq, jiffies + 1);
 }
 
@@ -831,8 +838,8 @@ static void cancel_fake_irq(struct intel_engine_cs *engine)
 {
 	struct intel_breadcrumbs *b = &engine->breadcrumbs;
 
+	del_timer_sync(&b->fake_irq); /* may queue b->hangcheck */
 	del_timer_sync(&b->hangcheck);
-	del_timer_sync(&b->fake_irq);
 	clear_bit(engine->id, &engine->i915->gpu_error.missed_irq_rings);
 }
 
@@ -840,15 +847,22 @@ void intel_engine_reset_breadcrumbs(struct intel_engine_cs *engine)
 {
 	struct intel_breadcrumbs *b = &engine->breadcrumbs;
 
-	cancel_fake_irq(engine);
 	spin_lock_irq(&b->irq_lock);
 
+	/*
+	 * Leave the fake_irq timer enabled (if it is running), but clear the
+	 * bit so that it turns itself off on its next wake up and goes back
+	 * to the long hangcheck interval if still required.
+	 */
+	clear_bit(engine->id, &engine->i915->gpu_error.missed_irq_rings);
+
 	if (b->irq_enabled)
 		irq_enable(engine);
 	else
 		irq_disable(engine);
 
-	/* We set the IRQ_BREADCRUMB bit when we enable the irq presuming the
+	/*
+	 * We set the IRQ_BREADCRUMB bit when we enable the irq presuming the
 	 * GPU is active and may have already executed the MI_USER_INTERRUPT
 	 * before the CPU is ready to receive. However, the engine is currently
 	 * idle (we haven't started it yet), there is no possibility for a
@@ -857,9 +871,6 @@ void intel_engine_reset_breadcrumbs(struct intel_engine_cs *engine)
 	 */
 	clear_bit(ENGINE_IRQ_BREADCRUMB, &engine->irq_posted);
 
-	if (b->irq_armed)
-		enable_fake_irq(b);
-
 	spin_unlock_irq(&b->irq_lock);
 }
 
-- 
cgit v1.1


From 14d4e522f0cbef2a10c3e5c243786a84d91cfb7b Mon Sep 17 00:00:00 2001
From: Lyude Paul <lyude@redhat.com>
Date: Wed, 11 Apr 2018 19:42:40 -0400
Subject: drm/atomic: Print debug message on atomic check failure
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Does what it says on the label, it's a little confusing debugging atomic
check failures otherwise.

Cc: Manasi Navare <manasi.d.navare@intel.com>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Lyude Paul <lyude@redhat.com>
Reviewed-by: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180411234302.2896-2-lyude@redhat.com
---
 drivers/gpu/drm/drm_atomic.c | 10 +++++++---
 1 file changed, 7 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index 7d25c42..0da8c5e 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -1702,11 +1702,15 @@ int drm_atomic_check_only(struct drm_atomic_state *state)
 		}
 	}
 
-	if (config->funcs->atomic_check)
+	if (config->funcs->atomic_check) {
 		ret = config->funcs->atomic_check(state->dev, state);
 
-	if (ret)
-		return ret;
+		if (ret) {
+			DRM_DEBUG_ATOMIC("atomic driver check for %p failed: %d\n",
+					 state, ret);
+			return ret;
+		}
+	}
 
 	if (!state->allow_modeset) {
 		for_each_new_crtc_in_state(state, crtc, crtc_state, i) {
-- 
cgit v1.1


From 1e10911539a61a048524bbb54ccfb6329b30b6d4 Mon Sep 17 00:00:00 2001
From: "Gomonovych, Vasyl" <gomonovych@gmail.com>
Date: Tue, 21 Nov 2017 23:31:33 +0100
Subject: drm/mediatek: Use ERR_CAST instead of ERR_PTR(PTR_ERR())

Use ERR_CAST inlined function instead of ERR_PTR(PTR_ERR(...)).

drivers/gpu/drm/mediatek/mtk_drm_gem.c:223:9-16: WARNING: ERR_CAST can be used with mtk_gem
Generated by: scripts/coccinelle/api/err_cast.cocci

Signed-off-by: Vasyl Gomonovych <gomonovych@gmail.com>
Acked-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: CK Hu <ck.hu@mediatek.com>
---
 drivers/gpu/drm/mediatek/mtk_drm_gem.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/mediatek/mtk_drm_gem.c b/drivers/gpu/drm/mediatek/mtk_drm_gem.c
index f595ac8..259b7b0 100644
--- a/drivers/gpu/drm/mediatek/mtk_drm_gem.c
+++ b/drivers/gpu/drm/mediatek/mtk_drm_gem.c
@@ -220,7 +220,7 @@ struct drm_gem_object *mtk_gem_prime_import_sg_table(struct drm_device *dev,
 	mtk_gem = mtk_drm_gem_init(dev, attach->dmabuf->size);
 
 	if (IS_ERR(mtk_gem))
-		return ERR_PTR(PTR_ERR(mtk_gem));
+		return ERR_CAST(mtk_gem);
 
 	expected = sg_dma_address(sg->sgl);
 	for_each_sg(sg->sgl, s, sg->nents, i) {
-- 
cgit v1.1


From 5540fbf43845868defcb599ec91c678275a20671 Mon Sep 17 00:00:00 2001
From: William Tu <u9012063@gmail.com>
Date: Tue, 24 Apr 2018 23:46:59 -0700
Subject: bpf: clear the ip_tunnel_info.

The percpu metadata_dst might carry the stale ip_tunnel_info
and cause incorrect behavior.  When mixing tests using ipv4/ipv6
bpf vxlan and geneve tunnel, the ipv6 tunnel info incorrectly uses
ipv4's src ip addr as its ipv6 src address, because the previous
tunnel info does not clean up.  The patch zeros the fields in
ip_tunnel_info.

Signed-off-by: William Tu <u9012063@gmail.com>
Reported-by: Yifeng Sun <pkusunyifeng@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
---
 net/core/filter.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/net/core/filter.c b/net/core/filter.c
index d31aff9..e77c30c 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -3240,6 +3240,7 @@ BPF_CALL_4(bpf_skb_set_tunnel_key, struct sk_buff *, skb,
 	skb_dst_set(skb, (struct dst_entry *) md);
 
 	info = &md->u.tun_info;
+	memset(info, 0, sizeof(*info));
 	info->mode = IP_TUNNEL_INFO_TX;
 
 	info->key.tun_flags = TUNNEL_KEY | TUNNEL_CSUM | TUNNEL_NOCACHE;
-- 
cgit v1.1


From 197143feede3038350056cd1d6e7c0524fc532dd Mon Sep 17 00:00:00 2001
From: Yixun Lan <yixun.lan@amlogic.com>
Date: Fri, 23 Mar 2018 22:38:10 +0800
Subject: clk: meson: drop meson_aoclk_gate_regmap_ops

let's remove the unused meson_aoclk_gate_regmap_ops

Fixes: 1f932d99710d ("clk: meson: remove superseded aoclk_gate_regmap")
Signed-off-by: Yixun Lan <yixun.lan@amlogic.com>
Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
---
 drivers/clk/meson/gxbb-aoclk.h | 2 --
 1 file changed, 2 deletions(-)

diff --git a/drivers/clk/meson/gxbb-aoclk.h b/drivers/clk/meson/gxbb-aoclk.h
index 0be7838..badc4c2 100644
--- a/drivers/clk/meson/gxbb-aoclk.h
+++ b/drivers/clk/meson/gxbb-aoclk.h
@@ -17,8 +17,6 @@
 #define AO_RTC_ALT_CLK_CNTL0	0x94
 #define AO_RTC_ALT_CLK_CNTL1	0x98
 
-extern const struct clk_ops meson_aoclk_gate_regmap_ops;
-
 struct aoclk_cec_32k {
 	struct clk_hw hw;
 	struct regmap *regmap;
-- 
cgit v1.1


From b251e4c88fb443b3a44c3d04268f70e2260f1f8a Mon Sep 17 00:00:00 2001
From: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Date: Sun, 22 Apr 2018 12:28:43 +0200
Subject: clk: meson: meson8b: fix meson8b_fclk_div3_div clock name

The names of all fclk divider gate clocks follow the naming schema
"fclk_divN" and the name of all fclk fixed dividers follow the naming
schema "fclk_divN_div".
There's one exception to this rule: meson8b_fclk_div3_div's name is
"fclk_div_div3". It's child clock meson8b_fclk_div3 however references
it as "fclk_div3_div" (following the naming schema explained above).

Fix the naming of the meson8b_fclk_div3_div clock to follow the naming
schema. This also fixes serial console on my Meson8m2 board because
"clk81" uses fclk_div3 as parent. However, since the hierarchy stops at
meson8b_fclk_div3 there's no known parent clock and the rate of "clk81"
and all of it's children (UART clock, SDIO MMC controller clock, ...)
are all 0.

Fixes: 05f814402d6174 ("clk: meson: add fdiv clock gates")
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
---
 drivers/clk/meson/meson8b.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/clk/meson/meson8b.c b/drivers/clk/meson/meson8b.c
index cc29924..2c4ff61 100644
--- a/drivers/clk/meson/meson8b.c
+++ b/drivers/clk/meson/meson8b.c
@@ -253,7 +253,7 @@ static struct clk_fixed_factor meson8b_fclk_div3_div = {
 	.mult = 1,
 	.div = 3,
 	.hw.init = &(struct clk_init_data){
-		.name = "fclk_div_div3",
+		.name = "fclk_div3_div",
 		.ops = &clk_fixed_factor_ops,
 		.parent_names = (const char *[]){ "fixed_pll" },
 		.num_parents = 1,
-- 
cgit v1.1


From 5b33139b1a08eabcba7b39d8a4babd7fc2d3b534 Mon Sep 17 00:00:00 2001
From: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Date: Mon, 23 Apr 2018 21:30:29 +0200
Subject: clk: meson: meson8b: fix meson8b_cpu_clk parent clock name

meson8b_cpu_clk has two parent clocks:
- meson8b_xtal
- meson8b_cpu_scale_out_sel

The name of the "xtal" clock parent is specified correctly. However,
there is a typo in the name of the second parent clock. The
meson8b_cpu_scale_out_sel definition uses the name "cpu_scale_out_sel"
(which matches the name from the datasheet). However, the mux parent
definition uses the name "cpu_out_sel" which does not match any existing
clock.

Fixes: 251b6fd38bcb9c ("clk: meson: rework meson8b cpu clock")
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
---
 drivers/clk/meson/meson8b.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/clk/meson/meson8b.c b/drivers/clk/meson/meson8b.c
index 2c4ff61..d0524ec 100644
--- a/drivers/clk/meson/meson8b.c
+++ b/drivers/clk/meson/meson8b.c
@@ -632,7 +632,8 @@ static struct clk_regmap meson8b_cpu_clk = {
 	.hw.init = &(struct clk_init_data){
 		.name = "cpu_clk",
 		.ops = &clk_regmap_mux_ro_ops,
-		.parent_names = (const char *[]){ "xtal", "cpu_out_sel" },
+		.parent_names = (const char *[]){ "xtal",
+						  "cpu_scale_out_sel" },
 		.num_parents = 2,
 		.flags = (CLK_SET_RATE_PARENT |
 			  CLK_SET_RATE_NO_REPARENT),
-- 
cgit v1.1


From a230cd52b8a2be39cd6e9a13b3e62af57f21372a Mon Sep 17 00:00:00 2001
From: pgzh <peter.ganzhorn@gmail.com>
Date: Thu, 12 Apr 2018 19:36:47 +0200
Subject: HID: lenovo: Add support for IBM/Lenovo Scrollpoint mice

The IBM/Lenovo Scrollpoint mice feature a trackpoint-like stick instead of a
scrolling wheel capable of 2-D (vertical+horizontal) scrolling.  hid-generic
does only expose 1-D (vertical) scrolling functionality for these mice.  This
patch adds support for horizontal scrolling for the IBM/Lenovo Scrollpoint mice
to hid-lenovo.

[jkosina@suse.cz: remove change versioning from git changelog]
Signed-off-by: Peter Ganzhorn <peter.ganzhorn@gmail.com>
Reviewed-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: Peter De Wachter <pdewacht@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
---
 drivers/hid/Kconfig      |  7 ++++---
 drivers/hid/hid-ids.h    |  8 ++++++++
 drivers/hid/hid-lenovo.c | 36 ++++++++++++++++++++++++++++++++++++
 3 files changed, 48 insertions(+), 3 deletions(-)

diff --git a/drivers/hid/Kconfig b/drivers/hid/Kconfig
index 60252fd..0000434 100644
--- a/drivers/hid/Kconfig
+++ b/drivers/hid/Kconfig
@@ -462,10 +462,11 @@ config HID_LENOVO
 	select NEW_LEDS
 	select LEDS_CLASS
 	---help---
-	Support for Lenovo devices that are not fully compliant with HID standard.
+	Support for IBM/Lenovo devices that are not fully compliant with HID standard.
 
-	Say Y if you want support for the non-compliant features of the Lenovo
-	Thinkpad standalone keyboards, e.g:
+	Say Y if you want support for horizontal scrolling of the IBM/Lenovo
+	Scrollpoint mice or the non-compliant features of the Lenovo Thinkpad
+	standalone keyboards, e.g:
 	- ThinkPad USB Keyboard with TrackPoint (supports extra LEDs and trackpoint
 	  configuration)
 	- ThinkPad Compact Bluetooth Keyboard with TrackPoint (supports Fn keys)
diff --git a/drivers/hid/hid-ids.h b/drivers/hid/hid-ids.h
index 0b5cc91..ec73aa4 100644
--- a/drivers/hid/hid-ids.h
+++ b/drivers/hid/hid-ids.h
@@ -552,6 +552,13 @@
 #define USB_VENDOR_ID_HUION		0x256c
 #define USB_DEVICE_ID_HUION_TABLET	0x006e
 
+#define USB_VENDOR_ID_IBM					0x04b3
+#define USB_DEVICE_ID_IBM_SCROLLPOINT_III			0x3100
+#define USB_DEVICE_ID_IBM_SCROLLPOINT_PRO			0x3103
+#define USB_DEVICE_ID_IBM_SCROLLPOINT_OPTICAL			0x3105
+#define USB_DEVICE_ID_IBM_SCROLLPOINT_800DPI_OPTICAL		0x3108
+#define USB_DEVICE_ID_IBM_SCROLLPOINT_800DPI_OPTICAL_PRO	0x3109
+
 #define USB_VENDOR_ID_IDEACOM		0x1cb6
 #define USB_DEVICE_ID_IDEACOM_IDC6650	0x6650
 #define USB_DEVICE_ID_IDEACOM_IDC6651	0x6651
@@ -684,6 +691,7 @@
 #define USB_DEVICE_ID_LENOVO_TPKBD	0x6009
 #define USB_DEVICE_ID_LENOVO_CUSBKBD	0x6047
 #define USB_DEVICE_ID_LENOVO_CBTKBD	0x6048
+#define USB_DEVICE_ID_LENOVO_SCROLLPOINT_OPTICAL	0x6049
 #define USB_DEVICE_ID_LENOVO_TPPRODOCK	0x6067
 #define USB_DEVICE_ID_LENOVO_X1_COVER	0x6085
 #define USB_DEVICE_ID_LENOVO_X1_TAB	0x60a3
diff --git a/drivers/hid/hid-lenovo.c b/drivers/hid/hid-lenovo.c
index 1ac4ff4..643b6eb 100644
--- a/drivers/hid/hid-lenovo.c
+++ b/drivers/hid/hid-lenovo.c
@@ -6,6 +6,17 @@
  *
  *  Copyright (c) 2012 Bernhard Seibold
  *  Copyright (c) 2014 Jamie Lentin <jm@lentin.co.uk>
+ *
+ * Linux IBM/Lenovo Scrollpoint mouse driver:
+ * - IBM Scrollpoint III
+ * - IBM Scrollpoint Pro
+ * - IBM Scrollpoint Optical
+ * - IBM Scrollpoint Optical 800dpi
+ * - IBM Scrollpoint Optical 800dpi Pro
+ * - Lenovo Scrollpoint Optical
+ *
+ *  Copyright (c) 2012 Peter De Wachter <pdewacht@gmail.com>
+ *  Copyright (c) 2018 Peter Ganzhorn <peter.ganzhorn@gmail.com>
  */
 
 /*
@@ -160,6 +171,17 @@ static int lenovo_input_mapping_cptkbd(struct hid_device *hdev,
 	return 0;
 }
 
+static int lenovo_input_mapping_scrollpoint(struct hid_device *hdev,
+		struct hid_input *hi, struct hid_field *field,
+		struct hid_usage *usage, unsigned long **bit, int *max)
+{
+	if (usage->hid == HID_GD_Z) {
+		hid_map_usage(hi, usage, bit, max, EV_REL, REL_HWHEEL);
+		return 1;
+	}
+	return 0;
+}
+
 static int lenovo_input_mapping(struct hid_device *hdev,
 		struct hid_input *hi, struct hid_field *field,
 		struct hid_usage *usage, unsigned long **bit, int *max)
@@ -172,6 +194,14 @@ static int lenovo_input_mapping(struct hid_device *hdev,
 	case USB_DEVICE_ID_LENOVO_CBTKBD:
 		return lenovo_input_mapping_cptkbd(hdev, hi, field,
 							usage, bit, max);
+	case USB_DEVICE_ID_IBM_SCROLLPOINT_III:
+	case USB_DEVICE_ID_IBM_SCROLLPOINT_PRO:
+	case USB_DEVICE_ID_IBM_SCROLLPOINT_OPTICAL:
+	case USB_DEVICE_ID_IBM_SCROLLPOINT_800DPI_OPTICAL:
+	case USB_DEVICE_ID_IBM_SCROLLPOINT_800DPI_OPTICAL_PRO:
+	case USB_DEVICE_ID_LENOVO_SCROLLPOINT_OPTICAL:
+		return lenovo_input_mapping_scrollpoint(hdev, hi, field,
+							usage, bit, max);
 	default:
 		return 0;
 	}
@@ -883,6 +913,12 @@ static const struct hid_device_id lenovo_devices[] = {
 	{ HID_USB_DEVICE(USB_VENDOR_ID_LENOVO, USB_DEVICE_ID_LENOVO_CUSBKBD) },
 	{ HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_LENOVO, USB_DEVICE_ID_LENOVO_CBTKBD) },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_LENOVO, USB_DEVICE_ID_LENOVO_TPPRODOCK) },
+	{ HID_USB_DEVICE(USB_VENDOR_ID_IBM, USB_DEVICE_ID_IBM_SCROLLPOINT_III) },
+	{ HID_USB_DEVICE(USB_VENDOR_ID_IBM, USB_DEVICE_ID_IBM_SCROLLPOINT_PRO) },
+	{ HID_USB_DEVICE(USB_VENDOR_ID_IBM, USB_DEVICE_ID_IBM_SCROLLPOINT_OPTICAL) },
+	{ HID_USB_DEVICE(USB_VENDOR_ID_IBM, USB_DEVICE_ID_IBM_SCROLLPOINT_800DPI_OPTICAL) },
+	{ HID_USB_DEVICE(USB_VENDOR_ID_IBM, USB_DEVICE_ID_IBM_SCROLLPOINT_800DPI_OPTICAL_PRO) },
+	{ HID_USB_DEVICE(USB_VENDOR_ID_LENOVO, USB_DEVICE_ID_LENOVO_SCROLLPOINT_OPTICAL) },
 	{ }
 };
 
-- 
cgit v1.1


From 097b8f62dd793e08f1732fc74dbb64596c7fbff9 Mon Sep 17 00:00:00 2001
From: Arvind Yadav <arvind.yadav.cs@gmail.com>
Date: Tue, 24 Apr 2018 13:33:03 +0530
Subject: HID: wacom: Release device resource data obtained by devres_alloc()

Free device resource data, if __wacom_devm_sysfs_create_group
is not successful.

Signed-off-by: Arvind Yadav <arvind.yadav.cs@gmail.com>
Reviewed-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
---
 drivers/hid/wacom_sys.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/hid/wacom_sys.c b/drivers/hid/wacom_sys.c
index b54ef1f..ee7a37e 100644
--- a/drivers/hid/wacom_sys.c
+++ b/drivers/hid/wacom_sys.c
@@ -1213,8 +1213,10 @@ static int __wacom_devm_sysfs_create_group(struct wacom *wacom,
 	devres->root = root;
 
 	error = sysfs_create_group(devres->root, group);
-	if (error)
+	if (error) {
+		devres_free(devres);
 		return error;
+	}
 
 	devres_add(&wacom->hdev->dev, devres);
 
-- 
cgit v1.1


From 37ba3c350e64adcbdd483e81bf194c05d9573515 Mon Sep 17 00:00:00 2001
From: Hans de Goede <hdegoede@redhat.com>
Date: Sat, 14 Apr 2018 17:06:44 +0200
Subject: HID: intel_ish-hid: Move header size check to inside the loop

With the headersize check outside of the loop, the second time through
the loop the: "payload_len = recv_msg->hdr.size;" statement may deref
recv_msg while it is pointing outside of our input buffer.

Move the headersize check to inside the loop to fix this.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Acked-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
---
 drivers/hid/intel-ish-hid/ishtp-hid-client.c | 20 ++++++++++----------
 1 file changed, 10 insertions(+), 10 deletions(-)

diff --git a/drivers/hid/intel-ish-hid/ishtp-hid-client.c b/drivers/hid/intel-ish-hid/ishtp-hid-client.c
index 157b44a..6ce1856 100644
--- a/drivers/hid/intel-ish-hid/ishtp-hid-client.c
+++ b/drivers/hid/intel-ish-hid/ishtp-hid-client.c
@@ -77,21 +77,21 @@ static void process_recv(struct ishtp_cl *hid_ishtp_cl, void *recv_buf,
 	struct ishtp_cl_data *client_data = hid_ishtp_cl->client_data;
 	int curr_hid_dev = client_data->cur_hid_dev;
 
-	if (data_len < sizeof(struct hostif_msg_hdr)) {
-		dev_err(&client_data->cl_device->dev,
-			"[hid-ish]: error, received %u which is less than data header %u\n",
-			(unsigned int)data_len,
-			(unsigned int)sizeof(struct hostif_msg_hdr));
-		++client_data->bad_recv_cnt;
-		ish_hw_reset(hid_ishtp_cl->dev);
-		return;
-	}
-
 	payload = recv_buf + sizeof(struct hostif_msg_hdr);
 	total_len = data_len;
 	cur_pos = 0;
 
 	do {
+		if (cur_pos + sizeof(struct hostif_msg) > total_len) {
+			dev_err(&client_data->cl_device->dev,
+				"[hid-ish]: error, received %u which is less than data header %u\n",
+				(unsigned int)data_len,
+				(unsigned int)sizeof(struct hostif_msg_hdr));
+			++client_data->bad_recv_cnt;
+			ish_hw_reset(hid_ishtp_cl->dev);
+			break;
+		}
+
 		recv_msg = (struct hostif_msg *)(recv_buf + cur_pos);
 		payload_len = recv_msg->hdr.size;
 
-- 
cgit v1.1


From 749ab300d41bcdbb8edf776d20e9a9891e79eba4 Mon Sep 17 00:00:00 2001
From: Hans de Goede <hdegoede@redhat.com>
Date: Sat, 14 Apr 2018 17:06:45 +0200
Subject: HID: intel_ish-hid: Stop using a static local buffer in get_report()

hid_ishtp_get_report() may be called by multiple callers at the same
time, causing trouble with the static local buffer used.

Also there is no reason to use a non stack buffer, the buffer is tiny
and ishtp_cl_send() copies its contents so the lifetime is not an
issue either.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Acked-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
---
 drivers/hid/intel-ish-hid/ishtp-hid-client.c | 16 ++++++----------
 1 file changed, 6 insertions(+), 10 deletions(-)

diff --git a/drivers/hid/intel-ish-hid/ishtp-hid-client.c b/drivers/hid/intel-ish-hid/ishtp-hid-client.c
index 6ce1856..acc2536 100644
--- a/drivers/hid/intel-ish-hid/ishtp-hid-client.c
+++ b/drivers/hid/intel-ish-hid/ishtp-hid-client.c
@@ -412,9 +412,7 @@ void hid_ishtp_get_report(struct hid_device *hid, int report_id,
 {
 	struct ishtp_hid_data *hid_data =  hid->driver_data;
 	struct ishtp_cl_data *client_data = hid_data->client_data;
-	static unsigned char	buf[10];
-	unsigned int	len;
-	struct hostif_msg_to_sensor *msg = (struct hostif_msg_to_sensor *)buf;
+	struct hostif_msg_to_sensor msg = {};
 	int	rv;
 	int	i;
 
@@ -426,14 +424,11 @@ void hid_ishtp_get_report(struct hid_device *hid, int report_id,
 		return;
 	}
 
-	len = sizeof(struct hostif_msg_to_sensor);
-
-	memset(msg, 0, sizeof(struct hostif_msg_to_sensor));
-	msg->hdr.command = (report_type == HID_FEATURE_REPORT) ?
+	msg.hdr.command = (report_type == HID_FEATURE_REPORT) ?
 		HOSTIF_GET_FEATURE_REPORT : HOSTIF_GET_INPUT_REPORT;
 	for (i = 0; i < client_data->num_hid_devices; ++i) {
 		if (hid == client_data->hid_sensor_hubs[i]) {
-			msg->hdr.device_id =
+			msg.hdr.device_id =
 				client_data->hid_devices[i].dev_id;
 			break;
 		}
@@ -442,8 +437,9 @@ void hid_ishtp_get_report(struct hid_device *hid, int report_id,
 	if (i == client_data->num_hid_devices)
 		return;
 
-	msg->report_id = report_id;
-	rv = ishtp_cl_send(client_data->hid_ishtp_cl, buf, len);
+	msg.report_id = report_id;
+	rv = ishtp_cl_send(client_data->hid_ishtp_cl, (uint8_t *)&msg,
+			    sizeof(msg));
 	if (rv)
 		hid_ishtp_trace(client_data,  "%s hid %p send failed\n",
 				__func__, hid);
-- 
cgit v1.1


From 064b40754291f95121bbdb05645a3a110d0ae346 Mon Sep 17 00:00:00 2001
From: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Date: Tue, 24 Apr 2018 15:18:41 +0200
Subject: usb: gadget: f_phonet: fix pn_net_xmit()'s return type

The method ndo_start_xmit() is defined as returning an 'netdev_tx_t',
which is a typedef for an enum type, but the implementation in this
driver returns an 'int'.

Fix this by returning 'netdev_tx_t' in this driver too.

Signed-off-by: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>
---
 drivers/usb/gadget/function/f_phonet.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/usb/gadget/function/f_phonet.c b/drivers/usb/gadget/function/f_phonet.c
index 7889bcc..8b72b19 100644
--- a/drivers/usb/gadget/function/f_phonet.c
+++ b/drivers/usb/gadget/function/f_phonet.c
@@ -221,7 +221,7 @@ static void pn_tx_complete(struct usb_ep *ep, struct usb_request *req)
 	netif_wake_queue(dev);
 }
 
-static int pn_net_xmit(struct sk_buff *skb, struct net_device *dev)
+static netdev_tx_t pn_net_xmit(struct sk_buff *skb, struct net_device *dev)
 {
 	struct phonet_port *port = netdev_priv(dev);
 	struct f_phonet *fp;
-- 
cgit v1.1


From 438fea2a6325933868aebc20279e2669c9a21207 Mon Sep 17 00:00:00 2001
From: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Date: Mon, 26 Mar 2018 11:00:01 +0200
Subject: usb: dwc2: dwc2_vbus_supply_init: fix error check

devm_regulator_get_optional returns -ENODEV if the regulator isn't
there, so if that's the case we have to make sure not to leave -ENODEV
in the regulator pointer.

Also, make sure we return 0 in that case, but correctly propagate any
other errors. Also propagate the error from _dwc2_hcd_start.

Fixes: 531ef5ebea96 ("usb: dwc2: add support for host mode external vbus supply")
Cc: Amelie Delaunay <amelie.delaunay@st.com>
Reviewed-by: Amelie Delaunay <amelie.delaunay@st.com>
Reviewed-by: Heiko Stuebner <heiko@sntech.de>
Reviewed-by: Grigor Tovmasyan <tovmasya@synopsys.com>
Tested-by: Heiko Stuebner <heiko@sntech.de>
Acked-by: Minas Harutyunyan <hminas@synopsys.com>
Signed-off-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>
---
 drivers/usb/dwc2/hcd.c | 13 ++++++++-----
 1 file changed, 8 insertions(+), 5 deletions(-)

diff --git a/drivers/usb/dwc2/hcd.c b/drivers/usb/dwc2/hcd.c
index 190f959..c51b73b 100644
--- a/drivers/usb/dwc2/hcd.c
+++ b/drivers/usb/dwc2/hcd.c
@@ -358,9 +358,14 @@ static void dwc2_gusbcfg_init(struct dwc2_hsotg *hsotg)
 
 static int dwc2_vbus_supply_init(struct dwc2_hsotg *hsotg)
 {
+	int ret;
+
 	hsotg->vbus_supply = devm_regulator_get_optional(hsotg->dev, "vbus");
-	if (IS_ERR(hsotg->vbus_supply))
-		return 0;
+	if (IS_ERR(hsotg->vbus_supply)) {
+		ret = PTR_ERR(hsotg->vbus_supply);
+		hsotg->vbus_supply = NULL;
+		return ret == -ENODEV ? 0 : ret;
+	}
 
 	return regulator_enable(hsotg->vbus_supply);
 }
@@ -4342,9 +4347,7 @@ static int _dwc2_hcd_start(struct usb_hcd *hcd)
 
 	spin_unlock_irqrestore(&hsotg->lock, flags);
 
-	dwc2_vbus_supply_init(hsotg);
-
-	return 0;
+	return dwc2_vbus_supply_init(hsotg);
 }
 
 /*
-- 
cgit v1.1


From 0ebf2a2c02b0976e939e795b1b9a53d32aeac8bc Mon Sep 17 00:00:00 2001
From: Artur Petrosyan <Arthur.Petrosyan@synopsys.com>
Date: Mon, 16 Apr 2018 08:45:31 -0400
Subject: usb: dwc2: WA for Full speed ISOC IN in DDMA mode.

By clearing NAK status of EP, core will send ZLP
to IN token and assert NAK interrupt relying
on TxFIFO status only.

The WA applies only to core versions from 2.72a
to 4.00a (including both). Also for FS_IOT_1.00a
and HS_IOT_1.00a.

Signed-off-by: Artur Petrosyan <arturp@synopsys.com>
Signed-off-by: Minas Harutyunyan <hminas@synopsys.com>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>
---
 drivers/usb/dwc2/core.h   |  2 ++
 drivers/usb/dwc2/gadget.c | 21 +++++++++++++++++++++
 2 files changed, 23 insertions(+)

diff --git a/drivers/usb/dwc2/core.h b/drivers/usb/dwc2/core.h
index d83be56..a666e07 100644
--- a/drivers/usb/dwc2/core.h
+++ b/drivers/usb/dwc2/core.h
@@ -985,6 +985,7 @@ struct dwc2_hsotg {
 
 	/* DWC OTG HW Release versions */
 #define DWC2_CORE_REV_2_71a	0x4f54271a
+#define DWC2_CORE_REV_2_72a     0x4f54272a
 #define DWC2_CORE_REV_2_80a	0x4f54280a
 #define DWC2_CORE_REV_2_90a	0x4f54290a
 #define DWC2_CORE_REV_2_91a	0x4f54291a
@@ -992,6 +993,7 @@ struct dwc2_hsotg {
 #define DWC2_CORE_REV_2_94a	0x4f54294a
 #define DWC2_CORE_REV_3_00a	0x4f54300a
 #define DWC2_CORE_REV_3_10a	0x4f54310a
+#define DWC2_CORE_REV_4_00a	0x4f54400a
 #define DWC2_FS_IOT_REV_1_00a	0x5531100a
 #define DWC2_HS_IOT_REV_1_00a	0x5532100a
 
diff --git a/drivers/usb/dwc2/gadget.c b/drivers/usb/dwc2/gadget.c
index 6c32bf2..83cb557 100644
--- a/drivers/usb/dwc2/gadget.c
+++ b/drivers/usb/dwc2/gadget.c
@@ -3928,6 +3928,27 @@ static int dwc2_hsotg_ep_enable(struct usb_ep *ep,
 	if (index && !hs_ep->isochronous)
 		epctrl |= DXEPCTL_SETD0PID;
 
+	/* WA for Full speed ISOC IN in DDMA mode.
+	 * By Clear NAK status of EP, core will send ZLP
+	 * to IN token and assert NAK interrupt relying
+	 * on TxFIFO status only
+	 */
+
+	if (hsotg->gadget.speed == USB_SPEED_FULL &&
+	    hs_ep->isochronous && dir_in) {
+		/* The WA applies only to core versions from 2.72a
+		 * to 4.00a (including both). Also for FS_IOT_1.00a
+		 * and HS_IOT_1.00a.
+		 */
+		u32 gsnpsid = dwc2_readl(hsotg->regs + GSNPSID);
+
+		if ((gsnpsid >= DWC2_CORE_REV_2_72a &&
+		     gsnpsid <= DWC2_CORE_REV_4_00a) ||
+		     gsnpsid == DWC2_FS_IOT_REV_1_00a ||
+		     gsnpsid == DWC2_HS_IOT_REV_1_00a)
+			epctrl |= DXEPCTL_CNAK;
+	}
+
 	dev_dbg(hsotg->dev, "%s: write DxEPCTL=0x%08x\n",
 		__func__, epctrl);
 
-- 
cgit v1.1


From a53097c2319c9df16dff6109392c50aa422265bf Mon Sep 17 00:00:00 2001
From: Wei Yongjun <weiyongjun1@huawei.com>
Date: Wed, 28 Mar 2018 13:41:13 +0000
Subject: usb: dwc2: pci: Fix error return code in dwc2_pci_probe()

Fix to return error code -ENOMEM from the alloc fail error handling
case instead of 0, as done elsewhere in this function.

Fixes: ecd29dabb2ba ("usb: dwc2: pci: Handle error cleanup in probe")
Reviewed-by: Grigor Tovmasyan <tovmasya@synopsys.com>
Acked-by: Minas Harutyunyan <hminas@synopsys.com>
Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>
---
 drivers/usb/dwc2/pci.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/usb/dwc2/pci.c b/drivers/usb/dwc2/pci.c
index 7f21747..bea2e8e 100644
--- a/drivers/usb/dwc2/pci.c
+++ b/drivers/usb/dwc2/pci.c
@@ -141,8 +141,10 @@ static int dwc2_pci_probe(struct pci_dev *pci,
 		goto err;
 
 	glue = devm_kzalloc(dev, sizeof(*glue), GFP_KERNEL);
-	if (!glue)
+	if (!glue) {
+		ret = -ENOMEM;
 		goto err;
+	}
 
 	ret = platform_device_add(dwc2);
 	if (ret) {
-- 
cgit v1.1


From 9cffd15de0c322c574085759793dbbcfafa91746 Mon Sep 17 00:00:00 2001
From: Wei Yongjun <weiyongjun1@huawei.com>
Date: Thu, 29 Mar 2018 02:20:10 +0000
Subject: usb: dwc3: gadget: dwc3_gadget_del_and_unmap_request() can be static

Fixes the following sparse warning:

drivers/usb/dwc3/gadget.c:169:6: warning:
 symbol 'dwc3_gadget_del_and_unmap_request' was not declared. Should it be static?

Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>
---
 drivers/usb/dwc3/gadget.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/usb/dwc3/gadget.c b/drivers/usb/dwc3/gadget.c
index 8796a5ee..206ecc7 100644
--- a/drivers/usb/dwc3/gadget.c
+++ b/drivers/usb/dwc3/gadget.c
@@ -166,7 +166,7 @@ static void dwc3_ep_inc_deq(struct dwc3_ep *dep)
 	dwc3_ep_inc_trb(&dep->trb_dequeue);
 }
 
-void dwc3_gadget_del_and_unmap_request(struct dwc3_ep *dep,
+static void dwc3_gadget_del_and_unmap_request(struct dwc3_ep *dep,
 		struct dwc3_request *req, int status)
 {
 	struct dwc3			*dwc = dep->dwc;
-- 
cgit v1.1


From 96bd39df29c29d348d65311e5954c0b7d3a2a790 Mon Sep 17 00:00:00 2001
From: Mayank Rana <mrana@codeaurora.org>
Date: Fri, 23 Mar 2018 10:05:33 -0700
Subject: usb: dwc3: gadget: Fix list_del corruption in dwc3_ep_dequeue

dwc3_ep_dequeue() waits for completion of End Transfer command using
wait_event_lock_irq(), which will release the dwc3->lock while waiting
and reacquire after completion. This allows a potential race condition
with ep_disable() which also removes all requests from started_list
and pending_list.

The check for NULL r->trb should catch this but currently it exits to
the wrong 'out1' label which calls dwc3_gadget_giveback(). Since its
list entry was already removed, if CONFIG_DEBUG_LIST is enabled a
'list_del corruption' bug is thrown since its next/prev pointers are
already LIST_POISON1/2. If r->trb is NULL it should simply exit to
'out0'.

Fixes: cf3113d893d4 ("usb: dwc3: gadget: properly increment dequeue pointer on ep_dequeue")
Cc: stable@vger.kernel.org # v4.12+
Signed-off-by: Mayank Rana <mrana@codeaurora.org>
Signed-off-by: Jack Pham <jackp@codeaurora.org>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>
---
 drivers/usb/dwc3/gadget.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/usb/dwc3/gadget.c b/drivers/usb/dwc3/gadget.c
index 206ecc7..0dedf8a 100644
--- a/drivers/usb/dwc3/gadget.c
+++ b/drivers/usb/dwc3/gadget.c
@@ -1424,7 +1424,7 @@ static int dwc3_gadget_ep_dequeue(struct usb_ep *ep,
 					dwc->lock);
 
 			if (!r->trb)
-				goto out1;
+				goto out0;
 
 			if (r->num_pending_sgs) {
 				struct dwc3_trb *trb;
-- 
cgit v1.1


From 1f177a131b2c106dd15ab51dda5c50f80b70cc72 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Wed, 25 Apr 2018 13:37:18 +0100
Subject: drm/i915: Use memset64() to align the ring with MI_NOOP

When filling the ring to align the emit pointer to the next cacheline,
use memset64() rather than open-coding it. As we know that we always
have an even number of dwords, we can replace the dword loop with the
qword equivalent.

v2: s/0/MI_NOOP<<32 | MI_NOOP/

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180425123718.16366-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/intel_ringbuffer.c | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index c68ac60..c06c22c 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -1717,22 +1717,24 @@ u32 *intel_ring_begin(struct i915_request *rq, unsigned int num_dwords)
 /* Align the ring tail to a cacheline boundary */
 int intel_ring_cacheline_align(struct i915_request *rq)
 {
-	int num_dwords = (rq->ring->emit & (CACHELINE_BYTES - 1)) / sizeof(u32);
-	u32 *cs;
+	int num_dwords;
+	void *cs;
 
+	num_dwords = (rq->ring->emit & (CACHELINE_BYTES - 1)) / sizeof(u32);
 	if (num_dwords == 0)
 		return 0;
 
-	num_dwords = CACHELINE_BYTES / sizeof(u32) - num_dwords;
+	num_dwords = CACHELINE_DWORDS - num_dwords;
+	GEM_BUG_ON(num_dwords & 1);
+
 	cs = intel_ring_begin(rq, num_dwords);
 	if (IS_ERR(cs))
 		return PTR_ERR(cs);
 
-	while (num_dwords--)
-		*cs++ = MI_NOOP;
-
+	memset64(cs, (u64)MI_NOOP << 32 | MI_NOOP, num_dwords / 2);
 	intel_ring_advance(rq, cs);
 
+	GEM_BUG_ON(rq->ring->emit & (CACHELINE_BYTES - 1));
 	return 0;
 }
 
-- 
cgit v1.1


From 1612a981b76688c598dc944bbfbe29a2b33e3973 Mon Sep 17 00:00:00 2001
From: Gianluca Borello <g.borello@gmail.com>
Date: Wed, 25 Apr 2018 05:42:16 +0000
Subject: bpf, x64: fix JIT emission for dead code

Commit 2a5418a13fcf ("bpf: improve dead code sanitizing") replaced dead
code with a series of ja-1 instructions, for safety. That made JIT
compilation much more complex for some BPF programs. One instance of such
programs is, for example:

bool flag = false
...
/* A bunch of other code */
...
if (flag)
        do_something()

In some cases llvm is not able to remove at compile time the code for
do_something(), so the generated BPF program ends up with a large amount
of dead instructions. In one specific real life example, there are two
series of ~500 and ~1000 dead instructions in the program. When the
verifier replaces them with a series of ja-1 instructions, it causes an
interesting behavior at JIT time.

During the first pass, since all the instructions are estimated at 64
bytes, the ja-1 instructions end up being translated as 5 bytes JMP
instructions (0xE9), since the jump offsets become increasingly large (>
127) as each instruction gets discovered to be 5 bytes instead of the
estimated 64.

Starting from the second pass, the first N instructions of the ja-1
sequence get translated into 2 bytes JMPs (0xEB) because the jump offsets
become <= 127 this time. In particular, N is defined as roughly 127 / (5
- 2) ~= 42. So, each further pass will make the subsequent N JMP
instructions shrink from 5 to 2 bytes, making the image shrink every time.
This means that in order to have the entire program converge, there need
to be, in the real example above, at least ~1000 / 42 ~= 24 passes just
for translating the dead code. If we add this number to the passes needed
to translate the other non dead code, it brings such program to 40+
passes, and JIT doesn't complete. Ultimately the userspace loader fails
because such BPF program was supposed to be part of a prog array owner
being JITed.

While it is certainly possible to try to refactor such programs to help
the compiler remove dead code, the behavior is not really intuitive and it
puts further burden on the BPF developer who is not expecting such
behavior. To make things worse, such programs are working just fine in all
the kernel releases prior to the ja-1 fix.

A possible approach to mitigate this behavior consists into noticing that
for ja-1 instructions we don't really need to rely on the estimated size
of the previous and current instructions, we know that a -1 BPF jump
offset can be safely translated into a 0xEB instruction with a jump offset
of -2.

Such fix brings the BPF program in the previous example to complete again
in ~9 passes.

Fixes: 2a5418a13fcf ("bpf: improve dead code sanitizing")
Signed-off-by: Gianluca Borello <g.borello@gmail.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
---
 arch/x86/net/bpf_jit_comp.c | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)

diff --git a/arch/x86/net/bpf_jit_comp.c b/arch/x86/net/bpf_jit_comp.c
index b7251541..abce27c 100644
--- a/arch/x86/net/bpf_jit_comp.c
+++ b/arch/x86/net/bpf_jit_comp.c
@@ -1027,7 +1027,17 @@ emit_cond_jmp:		/* convert BPF opcode to x86 */
 			break;
 
 		case BPF_JMP | BPF_JA:
-			jmp_offset = addrs[i + insn->off] - addrs[i];
+			if (insn->off == -1)
+				/* -1 jmp instructions will always jump
+				 * backwards two bytes. Explicitly handling
+				 * this case avoids wasting too many passes
+				 * when there are long sequences of replaced
+				 * dead code.
+				 */
+				jmp_offset = -2;
+			else
+				jmp_offset = addrs[i + insn->off] - addrs[i];
+
 			if (!jmp_offset)
 				/* optimize out nop jumps */
 				break;
-- 
cgit v1.1


From f6f109155f0deca7fae7ab503d1f12a806c2c095 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Wed, 25 Apr 2018 15:23:34 +0100
Subject: drm/i915: Remove obsolete min/max freq setters from debugfs

A more complete, and more importantly stable, interface for controlling
the RPS frequency range is available in sysfs, obsoleting the unstable
debugfs.

It's presence seems to trick people into using it, forgetting it is not
ABI.

References: https://bugs.freedesktop.org/show_bug.cgi?id=106237
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Sagar Arun Kamble <sagar.a.kamble@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180425142334.27113-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/i915_debugfs.c | 115 ------------------------------------
 1 file changed, 115 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index 2f05f52..1c88805 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -4205,119 +4205,6 @@ DEFINE_SIMPLE_ATTRIBUTE(i915_drop_caches_fops,
 			"0x%08llx\n");
 
 static int
-i915_max_freq_get(void *data, u64 *val)
-{
-	struct drm_i915_private *dev_priv = data;
-
-	if (INTEL_GEN(dev_priv) < 6)
-		return -ENODEV;
-
-	*val = intel_gpu_freq(dev_priv, dev_priv->gt_pm.rps.max_freq_softlimit);
-	return 0;
-}
-
-static int
-i915_max_freq_set(void *data, u64 val)
-{
-	struct drm_i915_private *dev_priv = data;
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-	u32 hw_max, hw_min;
-	int ret;
-
-	if (INTEL_GEN(dev_priv) < 6)
-		return -ENODEV;
-
-	DRM_DEBUG_DRIVER("Manually setting max freq to %llu\n", val);
-
-	ret = mutex_lock_interruptible(&dev_priv->pcu_lock);
-	if (ret)
-		return ret;
-
-	/*
-	 * Turbo will still be enabled, but won't go above the set value.
-	 */
-	val = intel_freq_opcode(dev_priv, val);
-
-	hw_max = rps->max_freq;
-	hw_min = rps->min_freq;
-
-	if (val < hw_min || val > hw_max || val < rps->min_freq_softlimit) {
-		mutex_unlock(&dev_priv->pcu_lock);
-		return -EINVAL;
-	}
-
-	rps->max_freq_softlimit = val;
-
-	if (intel_set_rps(dev_priv, val))
-		DRM_DEBUG_DRIVER("failed to update RPS to new softlimit\n");
-
-	mutex_unlock(&dev_priv->pcu_lock);
-
-	return 0;
-}
-
-DEFINE_SIMPLE_ATTRIBUTE(i915_max_freq_fops,
-			i915_max_freq_get, i915_max_freq_set,
-			"%llu\n");
-
-static int
-i915_min_freq_get(void *data, u64 *val)
-{
-	struct drm_i915_private *dev_priv = data;
-
-	if (INTEL_GEN(dev_priv) < 6)
-		return -ENODEV;
-
-	*val = intel_gpu_freq(dev_priv, dev_priv->gt_pm.rps.min_freq_softlimit);
-	return 0;
-}
-
-static int
-i915_min_freq_set(void *data, u64 val)
-{
-	struct drm_i915_private *dev_priv = data;
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-	u32 hw_max, hw_min;
-	int ret;
-
-	if (INTEL_GEN(dev_priv) < 6)
-		return -ENODEV;
-
-	DRM_DEBUG_DRIVER("Manually setting min freq to %llu\n", val);
-
-	ret = mutex_lock_interruptible(&dev_priv->pcu_lock);
-	if (ret)
-		return ret;
-
-	/*
-	 * Turbo will still be enabled, but won't go below the set value.
-	 */
-	val = intel_freq_opcode(dev_priv, val);
-
-	hw_max = rps->max_freq;
-	hw_min = rps->min_freq;
-
-	if (val < hw_min ||
-	    val > hw_max || val > rps->max_freq_softlimit) {
-		mutex_unlock(&dev_priv->pcu_lock);
-		return -EINVAL;
-	}
-
-	rps->min_freq_softlimit = val;
-
-	if (intel_set_rps(dev_priv, val))
-		DRM_DEBUG_DRIVER("failed to update RPS to new softlimit\n");
-
-	mutex_unlock(&dev_priv->pcu_lock);
-
-	return 0;
-}
-
-DEFINE_SIMPLE_ATTRIBUTE(i915_min_freq_fops,
-			i915_min_freq_get, i915_min_freq_set,
-			"%llu\n");
-
-static int
 i915_cache_sharing_get(void *data, u64 *val)
 {
 	struct drm_i915_private *dev_priv = data;
@@ -4878,8 +4765,6 @@ static const struct i915_debugfs_files {
 	const struct file_operations *fops;
 } i915_debugfs_files[] = {
 	{"i915_wedged", &i915_wedged_fops},
-	{"i915_max_freq", &i915_max_freq_fops},
-	{"i915_min_freq", &i915_min_freq_fops},
 	{"i915_cache_sharing", &i915_cache_sharing_fops},
 	{"i915_ring_missed_irq", &i915_ring_missed_irq_fops},
 	{"i915_ring_test_irq", &i915_ring_test_irq_fops},
-- 
cgit v1.1


From ff047a87cfacf0a530960171a0779f5b19a3b1b8 Mon Sep 17 00:00:00 2001
From: Oscar Mateo <oscar.mateo@intel.com>
Date: Tue, 24 Apr 2018 14:39:55 -0700
Subject: drm/i915/icl: Correctly clear lost ctx-switch interrupts across reset
 for Gen11

Interrupt handling in Gen11 is quite different from previous platforms.

v2: Rebased (Michel)
v3: Rebased with wiggle
v4: Rebased, remove TODO warning correctly (Daniele)
v5: Rebased, made gen11_gtiir const while at it (Michel)
v6: Rebased
v7: Adapt to the style currently in upstream

Suggested-by: Michel Thierry <michel.thierry@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Michel Thierry <michel.thierry@intel.com>
Signed-off-by: Oscar Mateo <oscar.mateo@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Michel Thierry <michel.thierry@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/1524605995-22324-1-git-send-email-oscar.mateo@intel.com
---
 drivers/gpu/drm/i915/i915_irq.c  |  6 ++--
 drivers/gpu/drm/i915/intel_drv.h |  3 ++
 drivers/gpu/drm/i915/intel_lrc.c | 60 ++++++++++++++++++++++++++++------------
 3 files changed, 48 insertions(+), 21 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index 96547e0..f9bc3aa 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -247,9 +247,9 @@ static u32
 gen11_gt_engine_identity(struct drm_i915_private * const i915,
 			 const unsigned int bank, const unsigned int bit);
 
-static bool gen11_reset_one_iir(struct drm_i915_private * const i915,
-				const unsigned int bank,
-				const unsigned int bit)
+bool gen11_reset_one_iir(struct drm_i915_private * const i915,
+			 const unsigned int bank,
+			 const unsigned int bit)
 {
 	void __iomem * const regs = i915->regs;
 	u32 dw;
diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
index 58868b9..9bba035 100644
--- a/drivers/gpu/drm/i915/intel_drv.h
+++ b/drivers/gpu/drm/i915/intel_drv.h
@@ -1333,6 +1333,9 @@ void intel_check_cpu_fifo_underruns(struct drm_i915_private *dev_priv);
 void intel_check_pch_fifo_underruns(struct drm_i915_private *dev_priv);
 
 /* i915_irq.c */
+bool gen11_reset_one_iir(struct drm_i915_private * const i915,
+			 const unsigned int bank,
+			 const unsigned int bit);
 void gen5_enable_gt_irq(struct drm_i915_private *dev_priv, uint32_t mask);
 void gen5_disable_gt_irq(struct drm_i915_private *dev_priv, uint32_t mask);
 void gen6_mask_pm_irq(struct drm_i915_private *dev_priv, u32 mask);
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 029901a..87eb3a6 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -789,22 +789,9 @@ execlists_cancel_port_requests(struct intel_engine_execlists * const execlists)
 
 static void clear_gtiir(struct intel_engine_cs *engine)
 {
-	static const u8 gtiir[] = {
-		[RCS]  = 0,
-		[BCS]  = 0,
-		[VCS]  = 1,
-		[VCS2] = 1,
-		[VECS] = 3,
-	};
 	struct drm_i915_private *dev_priv = engine->i915;
 	int i;
 
-	/* TODO: correctly reset irqs for gen11 */
-	if (WARN_ON_ONCE(INTEL_GEN(engine->i915) >= 11))
-		return;
-
-	GEM_BUG_ON(engine->id >= ARRAY_SIZE(gtiir));
-
 	/*
 	 * Clear any pending interrupt state.
 	 *
@@ -812,13 +799,50 @@ static void clear_gtiir(struct intel_engine_cs *engine)
 	 * double buffered, and so if we only reset it once there may
 	 * still be an interrupt pending.
 	 */
-	for (i = 0; i < 2; i++) {
-		I915_WRITE(GEN8_GT_IIR(gtiir[engine->id]),
+	if (INTEL_GEN(dev_priv) >= 11) {
+		static const struct {
+			u8 bank;
+			u8 bit;
+		} gen11_gtiir[] = {
+			[RCS] = {0, GEN11_RCS0},
+			[BCS] = {0, GEN11_BCS},
+			[_VCS(0)] = {1, GEN11_VCS(0)},
+			[_VCS(1)] = {1, GEN11_VCS(1)},
+			[_VCS(2)] = {1, GEN11_VCS(2)},
+			[_VCS(3)] = {1, GEN11_VCS(3)},
+			[_VECS(0)] = {1, GEN11_VECS(0)},
+			[_VECS(1)] = {1, GEN11_VECS(1)},
+		};
+		unsigned long irqflags;
+
+		GEM_BUG_ON(engine->id >= ARRAY_SIZE(gen11_gtiir));
+
+		spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
+		for (i = 0; i < 2; i++) {
+			gen11_reset_one_iir(dev_priv,
+					    gen11_gtiir[engine->id].bank,
+					    gen11_gtiir[engine->id].bit);
+		}
+		spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+	} else {
+		static const u8 gtiir[] = {
+			[RCS]  = 0,
+			[BCS]  = 0,
+			[VCS]  = 1,
+			[VCS2] = 1,
+			[VECS] = 3,
+		};
+
+		GEM_BUG_ON(engine->id >= ARRAY_SIZE(gtiir));
+
+		for (i = 0; i < 2; i++) {
+			I915_WRITE(GEN8_GT_IIR(gtiir[engine->id]),
+				   engine->irq_keep_mask);
+			POSTING_READ(GEN8_GT_IIR(gtiir[engine->id]));
+		}
+		GEM_BUG_ON(I915_READ(GEN8_GT_IIR(gtiir[engine->id])) &
 			   engine->irq_keep_mask);
-		POSTING_READ(GEN8_GT_IIR(gtiir[engine->id]));
 	}
-	GEM_BUG_ON(I915_READ(GEN8_GT_IIR(gtiir[engine->id])) &
-		   engine->irq_keep_mask);
 }
 
 static void reset_irq(struct intel_engine_cs *engine)
-- 
cgit v1.1


From 02a6efcab675fe32815d824837784c3f42a7d892 Mon Sep 17 00:00:00 2001
From: Alexandre Belloni <alexandre.belloni@bootlin.com>
Date: Tue, 24 Apr 2018 18:09:04 +0200
Subject: net: phy: allow scanning busses with missing phys

Some MDIO busses will error out when trying to read a phy address with no
phy present at that address. In that case, probing the bus will fail
because __mdiobus_register() is scanning the bus for all possible phys
addresses.

In case MII_PHYSID1 returns -EIO or -ENODEV, consider there is no phy at
this address and set the phy ID to 0xffffffff which is then properly
handled in get_phy_device().

Suggested-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/phy/phy_device.c | 11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)

diff --git a/drivers/net/phy/phy_device.c b/drivers/net/phy/phy_device.c
index ac23322..9e4ba8e 100644
--- a/drivers/net/phy/phy_device.c
+++ b/drivers/net/phy/phy_device.c
@@ -535,8 +535,17 @@ static int get_phy_id(struct mii_bus *bus, int addr, u32 *phy_id,
 
 	/* Grab the bits from PHYIR1, and put them in the upper half */
 	phy_reg = mdiobus_read(bus, addr, MII_PHYSID1);
-	if (phy_reg < 0)
+	if (phy_reg < 0) {
+		/* if there is no device, return without an error so scanning
+		 * the bus works properly
+		 */
+		if (phy_reg == -EIO || phy_reg == -ENODEV) {
+			*phy_id = 0xffffffff;
+			return 0;
+		}
+
 		return -EIO;
+	}
 
 	*phy_id = (phy_reg & 0xffff) << 16;
 
-- 
cgit v1.1


From 070204a34884110ac5e19c1e2e036fcfd033f8e3 Mon Sep 17 00:00:00 2001
From: Ursula Braun <ubraun@linux.ibm.com>
Date: Wed, 25 Apr 2018 12:48:58 +0200
Subject: net/smc: keep clcsock reference in smc_tcp_listen_work()

The internal CLC socket should exist till the SMC-socket is released.
Function tcp_listen_worker() releases the internal CLC socket of a
listen socket, if an smc_close_active() is called. This function
is called for the final release(), but it is called for shutdown
SHUT_RDWR as well. This opens a door for protection faults, if
socket calls using the internal CLC socket are called for a
shutdown listen socket.

With the changes of
commit 3d502067599f ("net/smc: simplify wait when closing listen socket")
there is no need anymore to release the internal CLC socket in
function tcp_listen_worker((). It is sufficient to release it in
smc_release().

Fixes: 127f49705823 ("net/smc: release clcsock from tcp_listen_worker")
Signed-off-by: Ursula Braun <ubraun@linux.ibm.com>
Reported-by: syzbot+9045fc589fcd196ef522@syzkaller.appspotmail.com
Reported-by: syzbot+28a2c86cf19c81d871fa@syzkaller.appspotmail.com
Reported-by: syzbot+9605e6cace1b5efd4a0a@syzkaller.appspotmail.com
Reported-by: syzbot+cf9012c597c8379d535c@syzkaller.appspotmail.com
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/smc/af_smc.c | 4 ----
 1 file changed, 4 deletions(-)

diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c
index f5d4b69..4470501 100644
--- a/net/smc/af_smc.c
+++ b/net/smc/af_smc.c
@@ -978,10 +978,6 @@ static void smc_tcp_listen_work(struct work_struct *work)
 	}
 
 out:
-	if (lsmc->clcsock) {
-		sock_release(lsmc->clcsock);
-		lsmc->clcsock = NULL;
-	}
 	release_sock(lsk);
 	sock_put(&lsmc->sk); /* sock_hold in smc_listen */
 }
-- 
cgit v1.1


From 91a825290ca4eae88603bc811bf74a45f94a3f46 Mon Sep 17 00:00:00 2001
From: Dag Moxnes <dag.moxnes@oracle.com>
Date: Wed, 25 Apr 2018 13:22:01 +0200
Subject: rds: ib: Fix missing call to rds_ib_dev_put in rds_ib_setup_qp
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The function rds_ib_setup_qp is calling rds_ib_get_client_data and
should correspondingly call rds_ib_dev_put. This call was lost in
the non-error path with the introduction of error handling done in
commit 3b12f73a5c29 ("rds: ib: add error handle")

Signed-off-by: Dag Moxnes <dag.moxnes@oracle.com>
Reviewed-by: Håkon Bugge <haakon.bugge@oracle.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/rds/ib_cm.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/net/rds/ib_cm.c b/net/rds/ib_cm.c
index eea1d86..13b38ad 100644
--- a/net/rds/ib_cm.c
+++ b/net/rds/ib_cm.c
@@ -547,7 +547,7 @@ static int rds_ib_setup_qp(struct rds_connection *conn)
 	rdsdebug("conn %p pd %p cq %p %p\n", conn, ic->i_pd,
 		 ic->i_send_cq, ic->i_recv_cq);
 
-	return ret;
+	goto out;
 
 sends_out:
 	vfree(ic->i_sends);
@@ -572,6 +572,7 @@ send_cq_out:
 		ic->i_send_cq = NULL;
 rds_ibdev_out:
 	rds_ib_remove_conn(rds_ibdev, conn);
+out:
 	rds_ib_dev_put(rds_ibdev);
 
 	return ret;
-- 
cgit v1.1


From 9c299a32ede98dc9faafb267034ed830a15304db Mon Sep 17 00:00:00 2001
From: John Fastabend <john.fastabend@gmail.com>
Date: Wed, 25 Apr 2018 14:22:45 -0700
Subject: bpf: fix for lex/yacc build error with gcc-5
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Fix build error found with Ubuntu shipped gcc-5

~/git/bpf/tools/bpf$ make all

Auto-detecting system features:
...                        libbfd: [ OFF ]
...        disassembler-four-args: [ OFF ]

  CC       bpf_jit_disasm.o
  LINK     bpf_jit_disasm
  CC       bpf_dbg.o
/home/john/git/bpf/tools/bpf/bpf_dbg.c: In function ‘cmd_load’:
/home/john/git/bpf/tools/bpf/bpf_dbg.c:1077:13: warning: ‘cont’ may be used uninitialized in this function [-Wmaybe-uninitialized]
  } else if (matches(subcmd, "pcap") == 0) {
             ^
  LINK     bpf_dbg
  CC       bpf_asm.o
make: *** No rule to make target `bpf_exp.yacc.o', needed by `bpf_asm'.  Stop.

Fixes: 5a8997f20715 ("tools: bpf: respect output directory during build")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
---
 tools/bpf/Makefile | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/tools/bpf/Makefile b/tools/bpf/Makefile
index 1ea5459..53b60ad 100644
--- a/tools/bpf/Makefile
+++ b/tools/bpf/Makefile
@@ -76,6 +76,8 @@ $(OUTPUT)bpf_asm: $(OUTPUT)bpf_asm.o $(OUTPUT)bpf_exp.yacc.o $(OUTPUT)bpf_exp.le
 	$(QUIET_LINK)$(CC) $(CFLAGS) -o $@ $^
 
 $(OUTPUT)bpf_exp.lex.c: $(OUTPUT)bpf_exp.yacc.c
+$(OUTPUT)bpf_exp.yacc.o: $(OUTPUT)bpf_exp.yacc.c
+$(OUTPUT)bpf_exp.lex.o: $(OUTPUT)bpf_exp.lex.c
 
 clean: bpftool_clean
 	$(call QUIET_CLEAN, bpf-progs)
-- 
cgit v1.1


From fcd58037f28bf70eb17157a51fbf94d466634a7d Mon Sep 17 00:00:00 2001
From: Arnaud Pouliquen <arnaud.pouliquen@st.com>
Date: Tue, 10 Apr 2018 10:57:25 +0200
Subject: remoteproc: fix crashed parameter logic on stop call

Fix rproc_add_subdev parameter name and inverse the crashed logic.

Fixes: 880f5b388252 ("remoteproc: Pass type of shutdown to subdev remove")
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Arnaud Pouliquen <arnaud.pouliquen@st.com>
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
---
 drivers/remoteproc/remoteproc_core.c | 4 ++--
 include/linux/remoteproc.h           | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/remoteproc/remoteproc_core.c b/drivers/remoteproc/remoteproc_core.c
index 6d9c583..a9609d9 100644
--- a/drivers/remoteproc/remoteproc_core.c
+++ b/drivers/remoteproc/remoteproc_core.c
@@ -1163,7 +1163,7 @@ int rproc_trigger_recovery(struct rproc *rproc)
 	if (ret)
 		return ret;
 
-	ret = rproc_stop(rproc, false);
+	ret = rproc_stop(rproc, true);
 	if (ret)
 		goto unlock_mutex;
 
@@ -1316,7 +1316,7 @@ void rproc_shutdown(struct rproc *rproc)
 	if (!atomic_dec_and_test(&rproc->power))
 		goto out;
 
-	ret = rproc_stop(rproc, true);
+	ret = rproc_stop(rproc, false);
 	if (ret) {
 		atomic_inc(&rproc->power);
 		goto out;
diff --git a/include/linux/remoteproc.h b/include/linux/remoteproc.h
index d09a9c7..dfdaede9 100644
--- a/include/linux/remoteproc.h
+++ b/include/linux/remoteproc.h
@@ -569,7 +569,7 @@ static inline struct rproc *vdev_to_rproc(struct virtio_device *vdev)
 void rproc_add_subdev(struct rproc *rproc,
 		      struct rproc_subdev *subdev,
 		      int (*probe)(struct rproc_subdev *subdev),
-		      void (*remove)(struct rproc_subdev *subdev, bool graceful));
+		      void (*remove)(struct rproc_subdev *subdev, bool crashed));
 
 void rproc_remove_subdev(struct rproc *rproc, struct rproc_subdev *subdev);
 
-- 
cgit v1.1


From 278d744c46fd4f1925aec77752d18a0e4a9cbec3 Mon Sep 17 00:00:00 2001
From: Tobias Jordan <Tobias.Jordan@elektrobit.com>
Date: Thu, 15 Feb 2018 16:12:55 +0100
Subject: remoteproc: qcom: Fix potential device node leaks

Add missing of_node_put()s at two places for device nodes returned by
of_parse_phandle().

Fixes: 051fb70fd4ea ("remoteproc: qcom: Driver for the self-authenticating
 Hexagon v5")
Signed-off-by: Tobias Jordan <Tobias.Jordan@elektrobit.com>
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
---
 drivers/remoteproc/qcom_q6v5_pil.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/remoteproc/qcom_q6v5_pil.c b/drivers/remoteproc/qcom_q6v5_pil.c
index 8e70a62..cbbafdc 100644
--- a/drivers/remoteproc/qcom_q6v5_pil.c
+++ b/drivers/remoteproc/qcom_q6v5_pil.c
@@ -1083,6 +1083,7 @@ static int q6v5_alloc_memory_region(struct q6v5 *qproc)
 		dev_err(qproc->dev, "unable to resolve mba region\n");
 		return ret;
 	}
+	of_node_put(node);
 
 	qproc->mba_phys = r.start;
 	qproc->mba_size = resource_size(&r);
@@ -1100,6 +1101,7 @@ static int q6v5_alloc_memory_region(struct q6v5 *qproc)
 		dev_err(qproc->dev, "unable to resolve mpss region\n");
 		return ret;
 	}
+	of_node_put(node);
 
 	qproc->mpss_phys = qproc->mpss_reloc = r.start;
 	qproc->mpss_size = resource_size(&r);
-- 
cgit v1.1


From 93dd4e73c0d9cc32f835d76a54257020b0bfc75a Mon Sep 17 00:00:00 2001
From: Ramon Fried <rfried@codeaurora.org>
Date: Fri, 23 Mar 2018 00:09:12 -0400
Subject: rpmsg: added MODULE_ALIAS for rpmsg_char

Added "rpmsg:rpmsg_chrdev" MODULE_ALIAS to autoload
rpmg_chrdev module automatically.

Signed-off-by: Ramon Fried <rfried@codeaurora.org>
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
---
 drivers/rpmsg/rpmsg_char.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/rpmsg/rpmsg_char.c b/drivers/rpmsg/rpmsg_char.c
index 64b6de9..1efdf9f 100644
--- a/drivers/rpmsg/rpmsg_char.c
+++ b/drivers/rpmsg/rpmsg_char.c
@@ -581,4 +581,6 @@ static void rpmsg_chrdev_exit(void)
 	unregister_chrdev_region(rpmsg_major, RPMSG_DEV_MAX);
 }
 module_exit(rpmsg_chrdev_exit);
+
+MODULE_ALIAS("rpmsg:rpmsg_chrdev");
 MODULE_LICENSE("GPL v2");
-- 
cgit v1.1


From 21fbd085e62ff9a11318f5a9aba55e56e8503173 Mon Sep 17 00:00:00 2001
From: Thomas Hellstrom <thellstrom@vmware.com>
Date: Thu, 26 Apr 2018 09:48:55 +0200
Subject: drm/vmwgfx: Clean up fbdev modeset locking

At least since the atomic port, the vmwgfx fbdev code is taking
a number of unnecessary modeset locks. In particular the
kms_set_config() function will grab its own locks, leading to
locking retries. So avoid drm_modeset_lock_all() and instead
provide a local acquire context for kms_set_config(). Also have the
vmw_kms_fbdev_init data itself grab the lock that it needs.

This also fixed a long standing problem that vmw_fb_close() didn't
provide an acquire context for kms_set_config(), causing potential
warnings and hangs during driver unload. This problem was uncovered by the
recent commit "drm/vmwgfx: Improve on hibernation"

Testing done:
Repeated driver load and unload on Ubuntu 16.04.2

Fixes: c3b9b1657344 ("drm/vmwgfx: Improve on hibernation")
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Deepak Rawat <drawat@vmware.com>
Reviewed-by: Sinclair Yeh <syeh@vmware.com>
---
 drivers/gpu/drm/vmwgfx/vmwgfx_fb.c  | 31 ++++++++++---------------------
 drivers/gpu/drm/vmwgfx/vmwgfx_kms.c | 13 ++++++++++---
 2 files changed, 20 insertions(+), 24 deletions(-)

diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_fb.c b/drivers/gpu/drm/vmwgfx/vmwgfx_fb.c
index 2582ffd..ba0cdb7 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_fb.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_fb.c
@@ -441,11 +441,11 @@ static int vmwgfx_set_config_internal(struct drm_mode_set *set)
 	struct drm_crtc *crtc = set->crtc;
 	struct drm_framebuffer *fb;
 	struct drm_crtc *tmp;
-	struct drm_modeset_acquire_ctx *ctx;
 	struct drm_device *dev = set->crtc->dev;
+	struct drm_modeset_acquire_ctx ctx;
 	int ret;
 
-	ctx = dev->mode_config.acquire_ctx;
+	drm_modeset_acquire_init(&ctx, 0);
 
 restart:
 	/*
@@ -458,7 +458,7 @@ restart:
 
 	fb = set->fb;
 
-	ret = crtc->funcs->set_config(set, ctx);
+	ret = crtc->funcs->set_config(set, &ctx);
 	if (ret == 0) {
 		crtc->primary->crtc = crtc;
 		crtc->primary->fb = fb;
@@ -473,20 +473,13 @@ restart:
 	}
 
 	if (ret == -EDEADLK) {
-		dev->mode_config.acquire_ctx = NULL;
-
-retry_locking:
-		drm_modeset_backoff(ctx);
-
-		ret = drm_modeset_lock_all_ctx(dev, ctx);
-		if (ret)
-			goto retry_locking;
-
-		dev->mode_config.acquire_ctx = ctx;
-
+		drm_modeset_backoff(&ctx);
 		goto restart;
 	}
 
+	drm_modeset_drop_locks(&ctx);
+	drm_modeset_acquire_fini(&ctx);
+
 	return ret;
 }
 
@@ -624,7 +617,6 @@ static int vmw_fb_set_par(struct fb_info *info)
 	}
 
 	mutex_lock(&par->bo_mutex);
-	drm_modeset_lock_all(vmw_priv->dev);
 	ret = vmw_fb_kms_framebuffer(info);
 	if (ret)
 		goto out_unlock;
@@ -657,7 +649,6 @@ out_unlock:
 		drm_mode_destroy(vmw_priv->dev, old_mode);
 	par->set_mode = mode;
 
-	drm_modeset_unlock_all(vmw_priv->dev);
 	mutex_unlock(&par->bo_mutex);
 
 	return ret;
@@ -713,18 +704,14 @@ int vmw_fb_init(struct vmw_private *vmw_priv)
 	par->max_width = fb_width;
 	par->max_height = fb_height;
 
-	drm_modeset_lock_all(vmw_priv->dev);
 	ret = vmw_kms_fbdev_init_data(vmw_priv, 0, par->max_width,
 				      par->max_height, &par->con,
 				      &par->crtc, &init_mode);
-	if (ret) {
-		drm_modeset_unlock_all(vmw_priv->dev);
+	if (ret)
 		goto err_kms;
-	}
 
 	info->var.xres = init_mode->hdisplay;
 	info->var.yres = init_mode->vdisplay;
-	drm_modeset_unlock_all(vmw_priv->dev);
 
 	/*
 	 * Create buffers and alloc memory
@@ -832,7 +819,9 @@ int vmw_fb_close(struct vmw_private *vmw_priv)
 	cancel_delayed_work_sync(&par->local_work);
 	unregister_framebuffer(info);
 
+	mutex_lock(&par->bo_mutex);
 	(void) vmw_fb_kms_detach(par, true, true);
+	mutex_unlock(&par->bo_mutex);
 
 	vfree(par->vmalloc);
 	framebuffer_release(info);
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
index f11601b..aacc930 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
@@ -2680,7 +2680,9 @@ int vmw_kms_fbdev_init_data(struct vmw_private *dev_priv,
 	struct vmw_display_unit *du;
 	struct drm_display_mode *mode;
 	int i = 0;
+	int ret = 0;
 
+	mutex_lock(&dev_priv->dev->mode_config.mutex);
 	list_for_each_entry(con, &dev_priv->dev->mode_config.connector_list,
 			    head) {
 		if (i == unit)
@@ -2691,7 +2693,8 @@ int vmw_kms_fbdev_init_data(struct vmw_private *dev_priv,
 
 	if (i != unit) {
 		DRM_ERROR("Could not find initial display unit.\n");
-		return -EINVAL;
+		ret = -EINVAL;
+		goto out_unlock;
 	}
 
 	if (list_empty(&con->modes))
@@ -2699,7 +2702,8 @@ int vmw_kms_fbdev_init_data(struct vmw_private *dev_priv,
 
 	if (list_empty(&con->modes)) {
 		DRM_ERROR("Could not find initial display mode.\n");
-		return -EINVAL;
+		ret = -EINVAL;
+		goto out_unlock;
 	}
 
 	du = vmw_connector_to_du(con);
@@ -2720,7 +2724,10 @@ int vmw_kms_fbdev_init_data(struct vmw_private *dev_priv,
 					   head);
 	}
 
-	return 0;
+ out_unlock:
+	mutex_unlock(&dev_priv->dev->mode_config.mutex);
+
+	return ret;
 }
 
 /**
-- 
cgit v1.1


From 13f149d47392782baafd96d54d4e65f3b5ca342f Mon Sep 17 00:00:00 2001
From: Thomas Hellstrom <thellstrom@vmware.com>
Date: Thu, 26 Apr 2018 09:59:30 +0200
Subject: drm/vmwgfx: Fix a buffer object leak

A buffer object leak was introduced when fixing a premature buffer
object release. Fix this.

Cc: <stable@vger.kernel.org>
Fixes: 73a88250b709 ("Fix a destoy-while-held mutex problem.")
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Deepak Rawat <drawat@vmware.com>
Reviewed-by: Sinclair Yeh <syeh@vmware.com>
---
 drivers/gpu/drm/vmwgfx/vmwgfx_kms.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
index aacc930..96fd7a0 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
@@ -2595,6 +2595,7 @@ void vmw_kms_helper_resource_finish(struct vmw_validation_ctx *ctx,
 		vmw_kms_helper_buffer_finish(res->dev_priv, NULL, ctx->buf,
 					     out_fence, NULL);
 
+	vmw_dmabuf_unreference(&ctx->buf);
 	vmw_resource_unreserve(res, false, NULL, 0);
 	mutex_unlock(&res->dev_priv->cmdbuf_mutex);
 }
-- 
cgit v1.1


From 741258cdd297e94b5f7167408b958b55795abaf8 Mon Sep 17 00:00:00 2001
From: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Date: Thu, 26 Apr 2018 08:47:16 +0100
Subject: drm/i915: Use seqlock in engine stats

We can convert engine stats from a spinlock to seqlock to ensure interrupt
processing is never even a tiny bit delayed by parallel readers.

There is a smidgen bit more cost on the write lock side, and an extremely
unlikely chance that readers will have to retry a few times in face of
heavy interrupt load. But it should be extremely unlikely given how
lightweight read side section is compared to the interrupt processing
side, and also compared to the rest of the code paths which can lead into
it. Furthermore, writer is the ones doing the real, latency sensitive
work, while readers are only informative.

Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Suggested-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20180426074716.7352-1-tvrtko.ursulin@linux.intel.com
---
 drivers/gpu/drm/i915/intel_engine_cs.c  | 19 ++++++++++---------
 drivers/gpu/drm/i915/intel_ringbuffer.h | 11 ++++++-----
 2 files changed, 16 insertions(+), 14 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
index 58be7fa..ac009f1 100644
--- a/drivers/gpu/drm/i915/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/intel_engine_cs.c
@@ -306,7 +306,7 @@ intel_engine_setup(struct drm_i915_private *dev_priv,
 	/* Nothing to do here, execute in order of dependencies */
 	engine->schedule = NULL;
 
-	spin_lock_init(&engine->stats.lock);
+	seqlock_init(&engine->stats.lock);
 
 	ATOMIC_INIT_NOTIFIER_HEAD(&engine->context_status_notifier);
 
@@ -1481,7 +1481,7 @@ int intel_enable_engine_stats(struct intel_engine_cs *engine)
 		return -ENODEV;
 
 	tasklet_disable(&execlists->tasklet);
-	spin_lock_irqsave(&engine->stats.lock, flags);
+	write_seqlock_irqsave(&engine->stats.lock, flags);
 
 	if (unlikely(engine->stats.enabled == ~0)) {
 		err = -EBUSY;
@@ -1505,7 +1505,7 @@ int intel_enable_engine_stats(struct intel_engine_cs *engine)
 	}
 
 unlock:
-	spin_unlock_irqrestore(&engine->stats.lock, flags);
+	write_sequnlock_irqrestore(&engine->stats.lock, flags);
 	tasklet_enable(&execlists->tasklet);
 
 	return err;
@@ -1534,12 +1534,13 @@ static ktime_t __intel_engine_get_busy_time(struct intel_engine_cs *engine)
  */
 ktime_t intel_engine_get_busy_time(struct intel_engine_cs *engine)
 {
+	unsigned int seq;
 	ktime_t total;
-	unsigned long flags;
 
-	spin_lock_irqsave(&engine->stats.lock, flags);
-	total = __intel_engine_get_busy_time(engine);
-	spin_unlock_irqrestore(&engine->stats.lock, flags);
+	do {
+		seq = read_seqbegin(&engine->stats.lock);
+		total = __intel_engine_get_busy_time(engine);
+	} while (read_seqretry(&engine->stats.lock, seq));
 
 	return total;
 }
@@ -1557,13 +1558,13 @@ void intel_disable_engine_stats(struct intel_engine_cs *engine)
 	if (!intel_engine_supports_stats(engine))
 		return;
 
-	spin_lock_irqsave(&engine->stats.lock, flags);
+	write_seqlock_irqsave(&engine->stats.lock, flags);
 	WARN_ON_ONCE(engine->stats.enabled == 0);
 	if (--engine->stats.enabled == 0) {
 		engine->stats.total = __intel_engine_get_busy_time(engine);
 		engine->stats.active = 0;
 	}
-	spin_unlock_irqrestore(&engine->stats.lock, flags);
+	write_sequnlock_irqrestore(&engine->stats.lock, flags);
 }
 
 #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index c5e2790..24af3f1 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -3,6 +3,7 @@
 #define _INTEL_RINGBUFFER_H_
 
 #include <linux/hashtable.h>
+#include <linux/seqlock.h>
 
 #include "i915_gem_batch_pool.h"
 #include "i915_gem_timeline.h"
@@ -595,7 +596,7 @@ struct intel_engine_cs {
 		/**
 		 * @lock: Lock protecting the below fields.
 		 */
-		spinlock_t lock;
+		seqlock_t lock;
 		/**
 		 * @enabled: Reference count indicating number of listeners.
 		 */
@@ -1064,7 +1065,7 @@ static inline void intel_engine_context_in(struct intel_engine_cs *engine)
 	if (READ_ONCE(engine->stats.enabled) == 0)
 		return;
 
-	spin_lock_irqsave(&engine->stats.lock, flags);
+	write_seqlock_irqsave(&engine->stats.lock, flags);
 
 	if (engine->stats.enabled > 0) {
 		if (engine->stats.active++ == 0)
@@ -1072,7 +1073,7 @@ static inline void intel_engine_context_in(struct intel_engine_cs *engine)
 		GEM_BUG_ON(engine->stats.active == 0);
 	}
 
-	spin_unlock_irqrestore(&engine->stats.lock, flags);
+	write_sequnlock_irqrestore(&engine->stats.lock, flags);
 }
 
 static inline void intel_engine_context_out(struct intel_engine_cs *engine)
@@ -1082,7 +1083,7 @@ static inline void intel_engine_context_out(struct intel_engine_cs *engine)
 	if (READ_ONCE(engine->stats.enabled) == 0)
 		return;
 
-	spin_lock_irqsave(&engine->stats.lock, flags);
+	write_seqlock_irqsave(&engine->stats.lock, flags);
 
 	if (engine->stats.enabled > 0) {
 		ktime_t last;
@@ -1109,7 +1110,7 @@ static inline void intel_engine_context_out(struct intel_engine_cs *engine)
 		}
 	}
 
-	spin_unlock_irqrestore(&engine->stats.lock, flags);
+	write_sequnlock_irqrestore(&engine->stats.lock, flags);
 }
 
 int intel_enable_engine_stats(struct intel_engine_cs *engine);
-- 
cgit v1.1


From f241632fd087d3d9fbd5450f4d8c8604badd8348 Mon Sep 17 00:00:00 2001
From: Govert Overgaauw <govert.overgaauw@prodrive-technologies.com>
Date: Fri, 6 Apr 2018 14:41:35 +0200
Subject: gpio: fix aspeed_gpio unmask irq

The unmask function disables all interrupts in a bank when unmasking an
interrupt. Only disable the given interrupt.

Cc: stable@vger.kernel.org
Signed-off-by: Govert Overgaauw <govert.overgaauw@prodrive-technologies.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
---
 drivers/gpio/gpio-aspeed.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpio/gpio-aspeed.c b/drivers/gpio/gpio-aspeed.c
index 77e4855..6f693b7 100644
--- a/drivers/gpio/gpio-aspeed.c
+++ b/drivers/gpio/gpio-aspeed.c
@@ -384,7 +384,7 @@ static void aspeed_gpio_irq_set_mask(struct irq_data *d, bool set)
 	if (set)
 		reg |= bit;
 	else
-		reg &= bit;
+		reg &= ~bit;
 	iowrite32(reg, addr);
 
 	spin_unlock_irqrestore(&gpio->lock, flags);
-- 
cgit v1.1


From ab3dbcf78f60f46d6a0ad63b1f4b690b7a427140 Mon Sep 17 00:00:00 2001
From: Timur Tabi <timur@codeaurora.org>
Date: Thu, 29 Mar 2018 13:29:12 -0500
Subject: gpioib: do not free unrequested descriptors

If the main loop in linehandle_create() encounters an error, it
unwinds completely by freeing all previously requested GPIO
descriptors.  However, if the error occurs in the beginning of
the loop before that GPIO is requested, then the exit code
attempts to free a null descriptor.  If extrachecks is enabled,
gpiod_free() triggers a WARN_ON.

Instead, keep a separate count of legitimate GPIOs so that only
those are freed.

Cc: stable@vger.kernel.org
Fixes: d7c51b47ac11 ("gpio: userspace ABI for reading/writing GPIO lines")
Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Timur Tabi <timur@codeaurora.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
---
 drivers/gpio/gpiolib.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/gpio/gpiolib.c b/drivers/gpio/gpiolib.c
index 43aeb07..d077717 100644
--- a/drivers/gpio/gpiolib.c
+++ b/drivers/gpio/gpiolib.c
@@ -497,7 +497,7 @@ static int linehandle_create(struct gpio_device *gdev, void __user *ip)
 	struct gpiohandle_request handlereq;
 	struct linehandle_state *lh;
 	struct file *file;
-	int fd, i, ret;
+	int fd, i, count = 0, ret;
 	u32 lflags;
 
 	if (copy_from_user(&handlereq, ip, sizeof(handlereq)))
@@ -558,6 +558,7 @@ static int linehandle_create(struct gpio_device *gdev, void __user *ip)
 		if (ret)
 			goto out_free_descs;
 		lh->descs[i] = desc;
+		count = i;
 
 		if (lflags & GPIOHANDLE_REQUEST_ACTIVE_LOW)
 			set_bit(FLAG_ACTIVE_LOW, &desc->flags);
@@ -628,7 +629,7 @@ static int linehandle_create(struct gpio_device *gdev, void __user *ip)
 out_put_unused_fd:
 	put_unused_fd(fd);
 out_free_descs:
-	for (; i >= 0; i--)
+	for (i = 0; i < count; i++)
 		gpiod_free(lh->descs[i]);
 	kfree(lh->label);
 out_free_lh:
-- 
cgit v1.1


From 75a07f399cd43bc7fb41a13723fbe04e61c5c470 Mon Sep 17 00:00:00 2001
From: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Date: Wed, 17 Jan 2018 22:18:41 +0200
Subject: drm: rcar-du: Zero-out sg_tables when duplicating plane state

The state structure for VSP-backed planes, rcar_du_vsp_plane_state,
contains sg tables that track framebuffer mapping performed in the
.prepare_fb() operation to unmap them in .cleanup_fb(). The tables are
incorrectly copied when duplicating state, which can result :

Zero-out sg_tables in original plane, effectively introducing move
semantic. Seems, this fixes issue with double-free,
when rcar_du_vsp_plane_cleanup_fb() freed the same sg_table
both in original plane and in the copy.

Reported-by: Volodymyr Babchuk <vlad.babchuk@gmail.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
---
 drivers/gpu/drm/rcar-du/rcar_du_vsp.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/rcar-du/rcar_du_vsp.c b/drivers/gpu/drm/rcar-du/rcar_du_vsp.c
index 2c260c3..4a01a99 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_vsp.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_vsp.c
@@ -299,18 +299,17 @@ static const struct drm_plane_helper_funcs rcar_du_vsp_plane_helper_funcs = {
 static struct drm_plane_state *
 rcar_du_vsp_plane_atomic_duplicate_state(struct drm_plane *plane)
 {
-	struct rcar_du_vsp_plane_state *state;
 	struct rcar_du_vsp_plane_state *copy;
 
 	if (WARN_ON(!plane->state))
 		return NULL;
 
-	state = to_rcar_vsp_plane_state(plane->state);
-	copy = kmemdup(state, sizeof(*state), GFP_KERNEL);
+	copy = kzalloc(sizeof(*copy), GFP_KERNEL);
 	if (copy == NULL)
 		return NULL;
 
 	__drm_atomic_helper_plane_duplicate_state(plane, &copy->state);
+	copy->alpha = to_rcar_vsp_plane_state(plane->state)->alpha;
 
 	return &copy->state;
 }
-- 
cgit v1.1


From 17515f1b764df36271f3166c714f5a78301fbaa7 Mon Sep 17 00:00:00 2001
From: Qu Wenruo <wqu@suse.com>
Date: Mon, 23 Apr 2018 17:32:04 +0800
Subject: btrfs: Fix wrong first_key parameter in replace_path

Commit 581c1760415c ("btrfs: Validate child tree block's level and first
key") introduced new @first_key parameter for read_tree_block(), however
caller in replace_path() is parasing wrong key to read_tree_block().

It should use parameter @first_key other than @key.

Normally it won't expose problem as @key is normally initialzied to the
same value of @first_key we expect.
However in relocation recovery case, @key can be set to (0, 0, 0), and
since no valid key in relocation tree can be (0, 0, 0), it will cause
read_tree_block() to return -EUCLEAN and interrupt relocation recovery.

Fix it by setting @first_key correctly.

Fixes: 581c1760415c ("btrfs: Validate child tree block's level and first key")
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
---
 fs/btrfs/relocation.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
index 00b7d32..b041b94 100644
--- a/fs/btrfs/relocation.c
+++ b/fs/btrfs/relocation.c
@@ -1841,7 +1841,7 @@ again:
 		old_bytenr = btrfs_node_blockptr(parent, slot);
 		blocksize = fs_info->nodesize;
 		old_ptr_gen = btrfs_node_ptr_generation(parent, slot);
-		btrfs_node_key_to_cpu(parent, &key, slot);
+		btrfs_node_key_to_cpu(parent, &first_key, slot);
 
 		if (level <= max_level) {
 			eb = path->nodes[level];
-- 
cgit v1.1


From a4eb490a41a0da3b1275fc7427084cf9ae2c3c1c Mon Sep 17 00:00:00 2001
From: Arvind Yadav <arvind.yadav.cs@gmail.com>
Date: Fri, 30 Mar 2018 16:56:10 +0530
Subject: HID: intel-ish-hid: use put_device() instead of kfree()

Never directly free @dev after calling device_register(), even
if it returned an error. Always use put_device() to give up the
reference initialized.

Signed-off-by: Arvind Yadav <arvind.yadav.cs@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
---
 drivers/hid/intel-ish-hid/ishtp/bus.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/hid/intel-ish-hid/ishtp/bus.c b/drivers/hid/intel-ish-hid/ishtp/bus.c
index f272cdd..2623a56 100644
--- a/drivers/hid/intel-ish-hid/ishtp/bus.c
+++ b/drivers/hid/intel-ish-hid/ishtp/bus.c
@@ -418,7 +418,7 @@ static struct ishtp_cl_device *ishtp_bus_add_device(struct ishtp_device *dev,
 		list_del(&device->device_link);
 		spin_unlock_irqrestore(&dev->device_list_lock, flags);
 		dev_err(dev->devc, "Failed to register ISHTP client device\n");
-		kfree(device);
+		put_device(&device->dev);
 		return NULL;
 	}
 
-- 
cgit v1.1


From 7f961d799fe4e0e515225c10b19177280d72c25e Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Thu, 26 Apr 2018 11:32:19 +0100
Subject: drm/i915: Compile out engine debug for release

The majority of the engine state dumping is too voluminous to be useful
outside of a controlled setup, though a few do accompany severe errors.
Keep the debug dumps next to the errors, but hide the others behind a CI
compile flag. This becomes more useful when adding more dumps to latency
sensitive paths.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180426103219.22181-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/i915_gem.c          | 2 +-
 drivers/gpu/drm/i915/i915_gem.h          | 6 ++++++
 drivers/gpu/drm/i915/intel_breadcrumbs.c | 2 +-
 drivers/gpu/drm/i915/intel_hangcheck.c   | 2 +-
 4 files changed, 9 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 795ca83..6b0c67a 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -3312,7 +3312,7 @@ void i915_gem_set_wedged(struct drm_i915_private *i915)
 
 	GEM_TRACE("start\n");
 
-	if (drm_debug & DRM_UT_DRIVER) {
+	if (GEM_SHOW_DEBUG()) {
 		struct drm_printer p = drm_debug_printer(__func__);
 
 		for_each_engine(engine, i915, id)
diff --git a/drivers/gpu/drm/i915/i915_gem.h b/drivers/gpu/drm/i915/i915_gem.h
index deaf78d..5259204 100644
--- a/drivers/gpu/drm/i915/i915_gem.h
+++ b/drivers/gpu/drm/i915/i915_gem.h
@@ -30,6 +30,9 @@
 struct drm_i915_private;
 
 #ifdef CONFIG_DRM_I915_DEBUG_GEM
+
+#define GEM_SHOW_DEBUG() (drm_debug & DRM_UT_DRIVER)
+
 #define GEM_BUG_ON(condition) do { if (unlikely((condition))) {	\
 		pr_err("%s:%d GEM_BUG_ON(%s)\n", \
 		       __func__, __LINE__, __stringify(condition)); \
@@ -45,6 +48,9 @@ struct drm_i915_private;
 #define GEM_DEBUG_BUG_ON(expr) GEM_BUG_ON(expr)
 
 #else
+
+#define GEM_SHOW_DEBUG() (0)
+
 #define GEM_BUG_ON(expr) BUILD_BUG_ON_INVALID(expr)
 #define GEM_WARN_ON(expr) (BUILD_BUG_ON_INVALID(expr), 0)
 
diff --git a/drivers/gpu/drm/i915/intel_breadcrumbs.c b/drivers/gpu/drm/i915/intel_breadcrumbs.c
index 5ce4f51..18e643d 100644
--- a/drivers/gpu/drm/i915/intel_breadcrumbs.c
+++ b/drivers/gpu/drm/i915/intel_breadcrumbs.c
@@ -82,7 +82,7 @@ static unsigned long wait_timeout(void)
 
 static noinline void missed_breadcrumb(struct intel_engine_cs *engine)
 {
-	if (drm_debug & DRM_UT_DRIVER) {
+	if (GEM_SHOW_DEBUG()) {
 		struct drm_printer p = drm_debug_printer(__func__);
 
 		intel_engine_dump(engine, &p,
diff --git a/drivers/gpu/drm/i915/intel_hangcheck.c b/drivers/gpu/drm/i915/intel_hangcheck.c
index fd0ffb8..309e38b 100644
--- a/drivers/gpu/drm/i915/intel_hangcheck.c
+++ b/drivers/gpu/drm/i915/intel_hangcheck.c
@@ -356,7 +356,7 @@ static void hangcheck_accumulate_sample(struct intel_engine_cs *engine,
 		break;
 
 	case ENGINE_DEAD:
-		if (drm_debug & DRM_UT_DRIVER) {
+		if (GEM_SHOW_DEBUG()) {
 			struct drm_printer p = drm_debug_printer("hangcheck");
 			intel_engine_dump(engine, &p, "%s\n", engine->name);
 		}
-- 
cgit v1.1


From 6131837b1de66116459ef4413e26fdbc70d066dc Mon Sep 17 00:00:00 2001
From: Omar Sandoval <osandov@fb.com>
Date: Thu, 26 Apr 2018 00:21:58 -0700
Subject: blk-mq: count allocated but not started requests in iostats inflight

In the legacy block case, we increment the counter right after we
allocate the request, not when the driver handles it. In both the legacy
and blk-mq cases, part_inc_in_flight() is called from
blk_account_io_start() right after we've allocated the request. blk-mq
only considers requests started requests as inflight, but this is
inconsistent with the legacy definition and the intention in the code.
This removes the started condition and instead counts all allocated
requests.

Fixes: f299b7c7a9de ("blk-mq: provide internal in-flight variant")
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 block/blk-mq.c | 21 +++++++++------------
 1 file changed, 9 insertions(+), 12 deletions(-)

diff --git a/block/blk-mq.c b/block/blk-mq.c
index c362145..5450cbc 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -95,18 +95,15 @@ static void blk_mq_check_inflight(struct blk_mq_hw_ctx *hctx,
 {
 	struct mq_inflight *mi = priv;
 
-	if (blk_mq_rq_state(rq) == MQ_RQ_IN_FLIGHT) {
-		/*
-		 * index[0] counts the specific partition that was asked
-		 * for. index[1] counts the ones that are active on the
-		 * whole device, so increment that if mi->part is indeed
-		 * a partition, and not a whole device.
-		 */
-		if (rq->part == mi->part)
-			mi->inflight[0]++;
-		if (mi->part->partno)
-			mi->inflight[1]++;
-	}
+	/*
+	 * index[0] counts the specific partition that was asked for. index[1]
+	 * counts the ones that are active on the whole device, so increment
+	 * that if mi->part is indeed a partition, and not a whole device.
+	 */
+	if (rq->part == mi->part)
+		mi->inflight[0]++;
+	if (mi->part->partno)
+		mi->inflight[1]++;
 }
 
 void blk_mq_in_flight(struct request_queue *q, struct hd_struct *part,
-- 
cgit v1.1


From bf0ddaba65ddbb2715af97041da8e7a45b2d8628 Mon Sep 17 00:00:00 2001
From: Omar Sandoval <osandov@fb.com>
Date: Thu, 26 Apr 2018 00:21:59 -0700
Subject: blk-mq: fix sysfs inflight counter

When the blk-mq inflight implementation was added, /proc/diskstats was
converted to use it, but /sys/block/$dev/inflight was not. Fix it by
adding another helper to count in-flight requests by data direction.

Fixes: f299b7c7a9de ("blk-mq: provide internal in-flight variant")
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 block/blk-mq.c            | 19 +++++++++++++++++++
 block/blk-mq.h            |  4 +++-
 block/genhd.c             | 12 ++++++++++++
 block/partition-generic.c | 10 ++++++----
 include/linux/genhd.h     |  4 +++-
 5 files changed, 43 insertions(+), 6 deletions(-)

diff --git a/block/blk-mq.c b/block/blk-mq.c
index 5450cbc..9ce9cac 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -115,6 +115,25 @@ void blk_mq_in_flight(struct request_queue *q, struct hd_struct *part,
 	blk_mq_queue_tag_busy_iter(q, blk_mq_check_inflight, &mi);
 }
 
+static void blk_mq_check_inflight_rw(struct blk_mq_hw_ctx *hctx,
+				     struct request *rq, void *priv,
+				     bool reserved)
+{
+	struct mq_inflight *mi = priv;
+
+	if (rq->part == mi->part)
+		mi->inflight[rq_data_dir(rq)]++;
+}
+
+void blk_mq_in_flight_rw(struct request_queue *q, struct hd_struct *part,
+			 unsigned int inflight[2])
+{
+	struct mq_inflight mi = { .part = part, .inflight = inflight, };
+
+	inflight[0] = inflight[1] = 0;
+	blk_mq_queue_tag_busy_iter(q, blk_mq_check_inflight_rw, &mi);
+}
+
 void blk_freeze_queue_start(struct request_queue *q)
 {
 	int freeze_depth;
diff --git a/block/blk-mq.h b/block/blk-mq.h
index 89b5cd3..e1bb420 100644
--- a/block/blk-mq.h
+++ b/block/blk-mq.h
@@ -188,7 +188,9 @@ static inline bool blk_mq_hw_queue_mapped(struct blk_mq_hw_ctx *hctx)
 }
 
 void blk_mq_in_flight(struct request_queue *q, struct hd_struct *part,
-			unsigned int inflight[2]);
+		      unsigned int inflight[2]);
+void blk_mq_in_flight_rw(struct request_queue *q, struct hd_struct *part,
+			 unsigned int inflight[2]);
 
 static inline void blk_mq_put_dispatch_budget(struct blk_mq_hw_ctx *hctx)
 {
diff --git a/block/genhd.c b/block/genhd.c
index dc7e089..c4513fe 100644
--- a/block/genhd.c
+++ b/block/genhd.c
@@ -82,6 +82,18 @@ void part_in_flight(struct request_queue *q, struct hd_struct *part,
 	}
 }
 
+void part_in_flight_rw(struct request_queue *q, struct hd_struct *part,
+		       unsigned int inflight[2])
+{
+	if (q->mq_ops) {
+		blk_mq_in_flight_rw(q, part, inflight);
+		return;
+	}
+
+	inflight[0] = atomic_read(&part->in_flight[0]);
+	inflight[1] = atomic_read(&part->in_flight[1]);
+}
+
 struct hd_struct *__disk_get_part(struct gendisk *disk, int partno)
 {
 	struct disk_part_tbl *ptbl = rcu_dereference(disk->part_tbl);
diff --git a/block/partition-generic.c b/block/partition-generic.c
index 08dabcd..db57cce 100644
--- a/block/partition-generic.c
+++ b/block/partition-generic.c
@@ -145,13 +145,15 @@ ssize_t part_stat_show(struct device *dev,
 		jiffies_to_msecs(part_stat_read(p, time_in_queue)));
 }
 
-ssize_t part_inflight_show(struct device *dev,
-			struct device_attribute *attr, char *buf)
+ssize_t part_inflight_show(struct device *dev, struct device_attribute *attr,
+			   char *buf)
 {
 	struct hd_struct *p = dev_to_part(dev);
+	struct request_queue *q = part_to_disk(p)->queue;
+	unsigned int inflight[2];
 
-	return sprintf(buf, "%8u %8u\n", atomic_read(&p->in_flight[0]),
-		atomic_read(&p->in_flight[1]));
+	part_in_flight_rw(q, p, inflight);
+	return sprintf(buf, "%8u %8u\n", inflight[0], inflight[1]);
 }
 
 #ifdef CONFIG_FAIL_MAKE_REQUEST
diff --git a/include/linux/genhd.h b/include/linux/genhd.h
index c826b0b..6cb8a57 100644
--- a/include/linux/genhd.h
+++ b/include/linux/genhd.h
@@ -368,7 +368,9 @@ static inline void free_part_stats(struct hd_struct *part)
 	part_stat_add(cpu, gendiskp, field, -subnd)
 
 void part_in_flight(struct request_queue *q, struct hd_struct *part,
-			unsigned int inflight[2]);
+		    unsigned int inflight[2]);
+void part_in_flight_rw(struct request_queue *q, struct hd_struct *part,
+		       unsigned int inflight[2]);
 void part_dec_in_flight(struct request_queue *q, struct hd_struct *part,
 			int rw);
 void part_inc_in_flight(struct request_queue *q, struct hd_struct *part,
-- 
cgit v1.1


From f60fa4087a8abac258b0c8facddf2ee46c4b6b36 Mon Sep 17 00:00:00 2001
From: Jani Nikula <jani.nikula@intel.com>
Date: Thu, 26 Apr 2018 14:35:21 +0300
Subject: drm/i915: prefer INTEL_GEN() over INTEL_INFO()->gen

Prefer INTEL_GEN() over INTEL_INFO()->gen except in special
circumstances.

v2: don't change device info dump (Chris)

Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180426113521.28417-1-jani.nikula@intel.com
---
 drivers/gpu/drm/i915/intel_device_info.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/intel_device_info.c b/drivers/gpu/drm/i915/intel_device_info.c
index a32ba72..0fd13df 100644
--- a/drivers/gpu/drm/i915/intel_device_info.c
+++ b/drivers/gpu/drm/i915/intel_device_info.c
@@ -848,7 +848,7 @@ void intel_device_info_runtime_init(struct intel_device_info *info)
 		gen9_sseu_info_init(dev_priv);
 	else if (INTEL_GEN(dev_priv) == 10)
 		gen10_sseu_info_init(dev_priv);
-	else if (INTEL_INFO(dev_priv)->gen >= 11)
+	else if (INTEL_GEN(dev_priv) >= 11)
 		gen11_sseu_info_init(dev_priv);
 
 	/* Initialize command stream timestamp frequency */
-- 
cgit v1.1


From 87251120553c82eda4ea109a8a52efe6c8ee0cc4 Mon Sep 17 00:00:00 2001
From: Jani Nikula <jani.nikula@intel.com>
Date: Thu, 26 Apr 2018 11:25:24 +0300
Subject: drm/i915/dp: remove stale comment about bw constants

We haven't used the DP bw constants here for a while. No functional
changes.

Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Manasi Navare <manasi.d.navare@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1dc7763cdc70c7f64c0a01f76f218d9ac0717227.1524730974.git.jani.nikula@intel.com
---
 drivers/gpu/drm/i915/intel_dp.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index 62f82c4..5f4b30f 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -1701,7 +1701,6 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 	int lane_count, clock;
 	int min_lane_count = 1;
 	int max_lane_count = intel_dp_max_lane_count(intel_dp);
-	/* Conveniently, the link BW constants become indices with a shift...*/
 	int min_clock = 0;
 	int max_clock;
 	int bpp, mode_rate;
-- 
cgit v1.1


From dd519418f5130ce9ca08256b3383bf0e529dbf81 Mon Sep 17 00:00:00 2001
From: Jani Nikula <jani.nikula@intel.com>
Date: Thu, 26 Apr 2018 11:25:25 +0300
Subject: drm/i915/dp: move link_bw and rate_select debugging where used

We call intel_dp_compute_rate() in intel_dp_compute_config() only to be
able to debug log the link_bw and rate_select parameters; we don't use
the parameters here for anything else. We call intel_dp_compute_rate()
again during link training where we actually need and use the
parameters.

Move the debug logging of link_bw and rate_select to
intel_dp_link_training_clock_recovery(), and clean up the extra
intel_dp_compute_rate() call and extra clutter from the already
overcrowded intel_dp_compute_config().

v2: Rewrote commit message (Rodrigo, Manasi)

Reviewed-by: Manasi Navare <manasi.d.navare@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/c5cf6a179e2d244eceb6bb80a792765d9efbee4f.1524730974.git.jani.nikula@intel.com
---
 drivers/gpu/drm/i915/intel_dp.c               | 9 ++-------
 drivers/gpu/drm/i915/intel_dp_link_training.c | 5 +++++
 2 files changed, 7 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index 5f4b30f..81cf363 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -1706,7 +1706,6 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 	int bpp, mode_rate;
 	int link_avail, link_clock;
 	int common_len;
-	uint8_t link_bw, rate_select;
 	bool reduce_m_n = drm_dp_has_quirk(&intel_dp->desc,
 					   DP_DPCD_QUIRK_LIMITED_M_N);
 
@@ -1852,12 +1851,8 @@ found:
 	pipe_config->pipe_bpp = bpp;
 	pipe_config->port_clock = intel_dp->common_rates[clock];
 
-	intel_dp_compute_rate(intel_dp, pipe_config->port_clock,
-			      &link_bw, &rate_select);
-
-	DRM_DEBUG_KMS("DP link bw %02x rate select %02x lane count %d clock %d bpp %d\n",
-		      link_bw, rate_select, pipe_config->lane_count,
-		      pipe_config->port_clock, bpp);
+	DRM_DEBUG_KMS("DP lane count %d clock %d bpp %d\n",
+		      pipe_config->lane_count, pipe_config->port_clock, bpp);
 	DRM_DEBUG_KMS("DP link bw required %i available %i\n",
 		      mode_rate, link_avail);
 
diff --git a/drivers/gpu/drm/i915/intel_dp_link_training.c b/drivers/gpu/drm/i915/intel_dp_link_training.c
index f59b59b..3fcaa98 100644
--- a/drivers/gpu/drm/i915/intel_dp_link_training.c
+++ b/drivers/gpu/drm/i915/intel_dp_link_training.c
@@ -139,6 +139,11 @@ intel_dp_link_training_clock_recovery(struct intel_dp *intel_dp)
 	intel_dp_compute_rate(intel_dp, intel_dp->link_rate,
 			      &link_bw, &rate_select);
 
+	if (link_bw)
+		DRM_DEBUG_KMS("Using LINK_BW_SET value %02x\n", link_bw);
+	else
+		DRM_DEBUG_KMS("Using LINK_RATE_SET value %02x\n", rate_select);
+
 	/* Write the link configuration data */
 	link_config[0] = link_bw;
 	link_config[1] = intel_dp->lane_count;
-- 
cgit v1.1


From 981a63eb2725eca63eee7b317ad9ff586d9e74b2 Mon Sep 17 00:00:00 2001
From: Jani Nikula <jani.nikula@intel.com>
Date: Thu, 26 Apr 2018 11:25:26 +0300
Subject: drm/i915/dp: abstract dp link config computation from the rest

Abstract a new intel_dp_compute_link_config() from
intel_dp_compute_config(), with the parts related to link configuration,
i.e. bpp, link rate, and lane count selection. No functional changes.

v2: Fix a checkpatch warn about spacing.

Reviewed-by: Manasi Navare <manasi.d.navare@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/80f99a625633f87f44d38d487ba3b32ff9a26b07.1524730974.git.jani.nikula@intel.com
---
 drivers/gpu/drm/i915/intel_dp.c | 161 ++++++++++++++++++++++------------------
 1 file changed, 88 insertions(+), 73 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index 81cf363..81da96b 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -1685,19 +1685,14 @@ static bool intel_edp_compare_alt_mode(struct drm_display_mode *m1,
 	return bres;
 }
 
-bool
-intel_dp_compute_config(struct intel_encoder *encoder,
-			struct intel_crtc_state *pipe_config,
-			struct drm_connector_state *conn_state)
+static bool
+intel_dp_compute_link_config(struct intel_encoder *encoder,
+			     struct intel_crtc_state *pipe_config)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
-	enum port port = encoder->port;
-	struct intel_crtc *intel_crtc = to_intel_crtc(pipe_config->base.crtc);
 	struct intel_connector *intel_connector = intel_dp->attached_connector;
-	struct intel_digital_connector_state *intel_conn_state =
-		to_intel_digital_connector_state(conn_state);
 	int lane_count, clock;
 	int min_lane_count = 1;
 	int max_lane_count = intel_dp_max_lane_count(intel_dp);
@@ -1706,9 +1701,6 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 	int bpp, mode_rate;
 	int link_avail, link_clock;
 	int common_len;
-	bool reduce_m_n = drm_dp_has_quirk(&intel_dp->desc,
-					   DP_DPCD_QUIRK_LIMITED_M_N);
-
 	common_len = intel_dp_common_len_rate_limit(intel_dp,
 						    intel_dp->max_link_rate);
 
@@ -1717,51 +1709,6 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 
 	max_clock = common_len - 1;
 
-	if (HAS_PCH_SPLIT(dev_priv) && !HAS_DDI(dev_priv) && port != PORT_A)
-		pipe_config->has_pch_encoder = true;
-
-	pipe_config->has_drrs = false;
-	if (IS_G4X(dev_priv) || port == PORT_A)
-		pipe_config->has_audio = false;
-	else if (intel_conn_state->force_audio == HDMI_AUDIO_AUTO)
-		pipe_config->has_audio = intel_dp->has_audio;
-	else
-		pipe_config->has_audio = intel_conn_state->force_audio == HDMI_AUDIO_ON;
-
-	if (intel_dp_is_edp(intel_dp) && intel_connector->panel.fixed_mode) {
-		struct drm_display_mode *panel_mode =
-			intel_connector->panel.alt_fixed_mode;
-		struct drm_display_mode *req_mode = &pipe_config->base.mode;
-
-		if (!intel_edp_compare_alt_mode(req_mode, panel_mode))
-			panel_mode = intel_connector->panel.fixed_mode;
-
-		drm_mode_debug_printmodeline(panel_mode);
-
-		intel_fixed_panel_mode(panel_mode, adjusted_mode);
-
-		if (INTEL_GEN(dev_priv) >= 9) {
-			int ret;
-			ret = skl_update_scaler_crtc(pipe_config);
-			if (ret)
-				return ret;
-		}
-
-		if (HAS_GMCH_DISPLAY(dev_priv))
-			intel_gmch_panel_fitting(intel_crtc, pipe_config,
-						 conn_state->scaling_mode);
-		else
-			intel_pch_panel_fitting(intel_crtc, pipe_config,
-						conn_state->scaling_mode);
-	}
-
-	if ((IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) &&
-	    adjusted_mode->flags & DRM_MODE_FLAG_INTERLACE)
-		return false;
-
-	if (adjusted_mode->flags & DRM_MODE_FLAG_DBLCLK)
-		return false;
-
 	/* Use values requested by Compliance Test Request */
 	if (intel_dp->compliance.test_type == DP_TEST_LINK_TRAINING) {
 		int index;
@@ -1831,6 +1778,83 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 	return false;
 
 found:
+	pipe_config->lane_count = lane_count;
+	pipe_config->pipe_bpp = bpp;
+	pipe_config->port_clock = intel_dp->common_rates[clock];
+
+	DRM_DEBUG_KMS("DP lane count %d clock %d bpp %d\n",
+		      pipe_config->lane_count, pipe_config->port_clock, bpp);
+	DRM_DEBUG_KMS("DP link bw required %i available %i\n",
+		      mode_rate, link_avail);
+
+	return true;
+}
+
+bool
+intel_dp_compute_config(struct intel_encoder *encoder,
+			struct intel_crtc_state *pipe_config,
+			struct drm_connector_state *conn_state)
+{
+	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
+	struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
+	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
+	enum port port = encoder->port;
+	struct intel_crtc *intel_crtc = to_intel_crtc(pipe_config->base.crtc);
+	struct intel_connector *intel_connector = intel_dp->attached_connector;
+	struct intel_digital_connector_state *intel_conn_state =
+		to_intel_digital_connector_state(conn_state);
+	bool reduce_m_n = drm_dp_has_quirk(&intel_dp->desc,
+					   DP_DPCD_QUIRK_LIMITED_M_N);
+
+	if (HAS_PCH_SPLIT(dev_priv) && !HAS_DDI(dev_priv) && port != PORT_A)
+		pipe_config->has_pch_encoder = true;
+
+	pipe_config->has_drrs = false;
+	if (IS_G4X(dev_priv) || port == PORT_A)
+		pipe_config->has_audio = false;
+	else if (intel_conn_state->force_audio == HDMI_AUDIO_AUTO)
+		pipe_config->has_audio = intel_dp->has_audio;
+	else
+		pipe_config->has_audio = intel_conn_state->force_audio == HDMI_AUDIO_ON;
+
+	if (intel_dp_is_edp(intel_dp) && intel_connector->panel.fixed_mode) {
+		struct drm_display_mode *panel_mode =
+			intel_connector->panel.alt_fixed_mode;
+		struct drm_display_mode *req_mode = &pipe_config->base.mode;
+
+		if (!intel_edp_compare_alt_mode(req_mode, panel_mode))
+			panel_mode = intel_connector->panel.fixed_mode;
+
+		drm_mode_debug_printmodeline(panel_mode);
+
+		intel_fixed_panel_mode(panel_mode, adjusted_mode);
+
+		if (INTEL_GEN(dev_priv) >= 9) {
+			int ret;
+
+			ret = skl_update_scaler_crtc(pipe_config);
+			if (ret)
+				return ret;
+		}
+
+		if (HAS_GMCH_DISPLAY(dev_priv))
+			intel_gmch_panel_fitting(intel_crtc, pipe_config,
+						 conn_state->scaling_mode);
+		else
+			intel_pch_panel_fitting(intel_crtc, pipe_config,
+						conn_state->scaling_mode);
+	}
+
+	if ((IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) &&
+	    adjusted_mode->flags & DRM_MODE_FLAG_INTERLACE)
+		return false;
+
+	if (adjusted_mode->flags & DRM_MODE_FLAG_DBLCLK)
+		return false;
+
+	if (!intel_dp_compute_link_config(encoder, pipe_config))
+		return false;
+
 	if (intel_conn_state->broadcast_rgb == INTEL_BROADCAST_RGB_AUTO) {
 		/*
 		 * See:
@@ -1838,7 +1862,7 @@ found:
 		 * VESA DisplayPort Ver.1.2a - 5.1.1.1 Video Colorimetry
 		 */
 		pipe_config->limited_color_range =
-			bpp != 18 &&
+			pipe_config->pipe_bpp != 18 &&
 			drm_default_rgb_quant_range(adjusted_mode) ==
 			HDMI_QUANTIZATION_RANGE_LIMITED;
 	} else {
@@ -1846,17 +1870,7 @@ found:
 			intel_conn_state->broadcast_rgb == INTEL_BROADCAST_RGB_LIMITED;
 	}
 
-	pipe_config->lane_count = lane_count;
-
-	pipe_config->pipe_bpp = bpp;
-	pipe_config->port_clock = intel_dp->common_rates[clock];
-
-	DRM_DEBUG_KMS("DP lane count %d clock %d bpp %d\n",
-		      pipe_config->lane_count, pipe_config->port_clock, bpp);
-	DRM_DEBUG_KMS("DP link bw required %i available %i\n",
-		      mode_rate, link_avail);
-
-	intel_link_compute_m_n(bpp, lane_count,
+	intel_link_compute_m_n(pipe_config->pipe_bpp, pipe_config->lane_count,
 			       adjusted_mode->crtc_clock,
 			       pipe_config->port_clock,
 			       &pipe_config->dp_m_n,
@@ -1865,11 +1879,12 @@ found:
 	if (intel_connector->panel.downclock_mode != NULL &&
 		dev_priv->drrs.type == SEAMLESS_DRRS_SUPPORT) {
 			pipe_config->has_drrs = true;
-			intel_link_compute_m_n(bpp, lane_count,
-				intel_connector->panel.downclock_mode->clock,
-				pipe_config->port_clock,
-				&pipe_config->dp_m2_n2,
-				reduce_m_n);
+			intel_link_compute_m_n(pipe_config->pipe_bpp,
+					       pipe_config->lane_count,
+					       intel_connector->panel.downclock_mode->clock,
+					       pipe_config->port_clock,
+					       &pipe_config->dp_m2_n2,
+					       reduce_m_n);
 	}
 
 	/*
-- 
cgit v1.1


From ef32659a78df0b10cd25076d483dcb6240274c90 Mon Sep 17 00:00:00 2001
From: Jani Nikula <jani.nikula@intel.com>
Date: Thu, 26 Apr 2018 11:25:27 +0300
Subject: drm/i915/dp: move eDP VBT bpp clamping code to intel_dp_compute_bpp()

Keep related things together. No functional changes.

v2: Fix a typo in patch subject, fix a checkpatch alignment warning.

Reviewed-by: Manasi Navare <manasi.d.navare@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/f24d44547a586a0e342f24e69ab4d576a2474891.1524730974.git.jani.nikula@intel.com
---
 drivers/gpu/drm/i915/intel_dp.c | 24 +++++++++++++-----------
 1 file changed, 13 insertions(+), 11 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index 81da96b..430c206 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -1650,6 +1650,8 @@ void intel_dp_compute_rate(struct intel_dp *intel_dp, int port_clock,
 static int intel_dp_compute_bpp(struct intel_dp *intel_dp,
 				struct intel_crtc_state *pipe_config)
 {
+	struct drm_i915_private *dev_priv = to_i915(intel_dp_to_dev(intel_dp));
+	struct intel_connector *intel_connector = intel_dp->attached_connector;
 	int bpp, bpc;
 
 	bpp = pipe_config->pipe_bpp;
@@ -1665,6 +1667,17 @@ static int intel_dp_compute_bpp(struct intel_dp *intel_dp,
 		DRM_DEBUG_KMS("Setting pipe_bpp to %d\n",
 			      pipe_config->pipe_bpp);
 	}
+
+	if (intel_dp_is_edp(intel_dp)) {
+		/* Get bpp from vbt only for panels that dont have bpp in edid */
+		if (intel_connector->base.display_info.bpc == 0 &&
+		    dev_priv->vbt.edp.bpp && dev_priv->vbt.edp.bpp < bpp) {
+			DRM_DEBUG_KMS("clamping bpp for eDP panel to BIOS-provided %i\n",
+				      dev_priv->vbt.edp.bpp);
+			bpp = dev_priv->vbt.edp.bpp;
+		}
+	}
+
 	return bpp;
 }
 
@@ -1689,10 +1702,8 @@ static bool
 intel_dp_compute_link_config(struct intel_encoder *encoder,
 			     struct intel_crtc_state *pipe_config)
 {
-	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
-	struct intel_connector *intel_connector = intel_dp->attached_connector;
 	int lane_count, clock;
 	int min_lane_count = 1;
 	int max_lane_count = intel_dp_max_lane_count(intel_dp);
@@ -1735,15 +1746,6 @@ intel_dp_compute_link_config(struct intel_encoder *encoder,
 	 * bpc in between. */
 	bpp = intel_dp_compute_bpp(intel_dp, pipe_config);
 	if (intel_dp_is_edp(intel_dp)) {
-
-		/* Get bpp from vbt only for panels that dont have bpp in edid */
-		if (intel_connector->base.display_info.bpc == 0 &&
-			(dev_priv->vbt.edp.bpp && dev_priv->vbt.edp.bpp < bpp)) {
-			DRM_DEBUG_KMS("clamping bpp for eDP panel to BIOS-provided %i\n",
-				      dev_priv->vbt.edp.bpp);
-			bpp = dev_priv->vbt.edp.bpp;
-		}
-
 		/*
 		 * Use the maximum clock and number of lanes the eDP panel
 		 * advertizes being capable of. The panels are generally
-- 
cgit v1.1


From 7c2781e41ec8893ad9ef62f28821e2dbd4bd0420 Mon Sep 17 00:00:00 2001
From: Jani Nikula <jani.nikula@intel.com>
Date: Thu, 26 Apr 2018 11:25:28 +0300
Subject: drm/i915/dp: group link config limits in a struct

Also use same min/max model for bpp, and adjust debug logging while at
it.

Reviewed-by: Manasi Navare <manasi.d.navare@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/72f78c7ae0cd1810798bd94cbf5e574c78da83f8.1524730974.git.jani.nikula@intel.com
---
 drivers/gpu/drm/i915/intel_dp.c | 57 ++++++++++++++++++++++++-----------------
 1 file changed, 33 insertions(+), 24 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index 430c206..9ef29b6 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -1647,6 +1647,12 @@ void intel_dp_compute_rate(struct intel_dp *intel_dp, int port_clock,
 	}
 }
 
+struct link_config_limits {
+	int min_clock, max_clock;
+	int min_lane_count, max_lane_count;
+	int min_bpp, max_bpp;
+};
+
 static int intel_dp_compute_bpp(struct intel_dp *intel_dp,
 				struct intel_crtc_state *pipe_config)
 {
@@ -1704,21 +1710,25 @@ intel_dp_compute_link_config(struct intel_encoder *encoder,
 {
 	struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
-	int lane_count, clock;
-	int min_lane_count = 1;
-	int max_lane_count = intel_dp_max_lane_count(intel_dp);
-	int min_clock = 0;
-	int max_clock;
-	int bpp, mode_rate;
-	int link_avail, link_clock;
+	struct link_config_limits limits;
+	int bpp, clock, lane_count;
+	int mode_rate, link_avail, link_clock;
 	int common_len;
+
 	common_len = intel_dp_common_len_rate_limit(intel_dp,
 						    intel_dp->max_link_rate);
 
 	/* No common link rates between source and sink */
 	WARN_ON(common_len <= 0);
 
-	max_clock = common_len - 1;
+	limits.min_clock = 0;
+	limits.max_clock = common_len - 1;
+
+	limits.min_lane_count = 1;
+	limits.max_lane_count = intel_dp_max_lane_count(intel_dp);
+
+	limits.min_bpp = 6 * 3;
+	limits.max_bpp = intel_dp_compute_bpp(intel_dp, pipe_config);
 
 	/* Use values requested by Compliance Test Request */
 	if (intel_dp->compliance.test_type == DP_TEST_LINK_TRAINING) {
@@ -1733,18 +1743,11 @@ intel_dp_compute_link_config(struct intel_encoder *encoder,
 						    intel_dp->num_common_rates,
 						    intel_dp->compliance.test_link_rate);
 			if (index >= 0)
-				min_clock = max_clock = index;
-			min_lane_count = max_lane_count = intel_dp->compliance.test_lane_count;
+				limits.min_clock = limits.max_clock = index;
+			limits.min_lane_count = limits.max_lane_count = intel_dp->compliance.test_lane_count;
 		}
 	}
-	DRM_DEBUG_KMS("DP link computation with max lane count %i "
-		      "max bw %d pixel clock %iKHz\n",
-		      max_lane_count, intel_dp->common_rates[max_clock],
-		      adjusted_mode->crtc_clock);
 
-	/* Walk through all bpp values. Luckily they're all nicely spaced with 2
-	 * bpc in between. */
-	bpp = intel_dp_compute_bpp(intel_dp, pipe_config);
 	if (intel_dp_is_edp(intel_dp)) {
 		/*
 		 * Use the maximum clock and number of lanes the eDP panel
@@ -1753,18 +1756,24 @@ intel_dp_compute_link_config(struct intel_encoder *encoder,
 		 * configuration, and typically these values correspond to the
 		 * native resolution of the panel.
 		 */
-		min_lane_count = max_lane_count;
-		min_clock = max_clock;
+		limits.min_lane_count = limits.max_lane_count;
+		limits.min_clock = limits.max_clock;
 	}
 
-	for (; bpp >= 6*3; bpp -= 2*3) {
+	DRM_DEBUG_KMS("DP link computation with max lane count %i "
+		      "max rate %d max bpp %d pixel clock %iKHz\n",
+		      limits.max_lane_count,
+		      intel_dp->common_rates[limits.max_clock],
+		      limits.max_bpp, adjusted_mode->crtc_clock);
+
+	for (bpp = limits.max_bpp; bpp >= limits.min_bpp; bpp -= 2 * 3) {
 		mode_rate = intel_dp_link_required(adjusted_mode->crtc_clock,
 						   bpp);
 
-		for (clock = min_clock; clock <= max_clock; clock++) {
-			for (lane_count = min_lane_count;
-				lane_count <= max_lane_count;
-				lane_count <<= 1) {
+		for (clock = limits.min_clock; clock <= limits.max_clock; clock++) {
+			for (lane_count = limits.min_lane_count;
+			     lane_count <= limits.max_lane_count;
+			     lane_count <<= 1) {
 
 				link_clock = intel_dp->common_rates[clock];
 				link_avail = intel_dp_max_data_rate(link_clock,
-- 
cgit v1.1


From 3acd115d08f70615debd5a8d37a8b97dc17a9cbb Mon Sep 17 00:00:00 2001
From: Jani Nikula <jani.nikula@intel.com>
Date: Thu, 26 Apr 2018 11:25:29 +0300
Subject: drm/i915/dp: abstract link config selection

For now, there's just the one link config selection, optimizing for slow
and wide link. No functional changes.

Keep the debug logging in the caller, to avoid duplication later on if
alternative link confing selection gets added.

v2: Improved commit message

Reviewed-by: Manasi Navare <manasi.d.navare@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/64848b76bf90d6ceecd7ec6b5add28531e0b1a41.1524730974.git.jani.nikula@intel.com
---
 drivers/gpu/drm/i915/intel_dp.c | 81 +++++++++++++++++++++++++----------------
 1 file changed, 50 insertions(+), 31 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index 9ef29b6..d622db7 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -1704,6 +1704,42 @@ static bool intel_edp_compare_alt_mode(struct drm_display_mode *m1,
 	return bres;
 }
 
+/* Optimize link config in order: max bpp, min clock, min lanes */
+static bool
+intel_dp_compute_link_config_wide(struct intel_dp *intel_dp,
+				  struct intel_crtc_state *pipe_config,
+				  const struct link_config_limits *limits)
+{
+	struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
+	int bpp, clock, lane_count;
+	int mode_rate, link_clock, link_avail;
+
+	for (bpp = limits->max_bpp; bpp >= limits->min_bpp; bpp -= 2 * 3) {
+		mode_rate = intel_dp_link_required(adjusted_mode->crtc_clock,
+						   bpp);
+
+		for (clock = limits->min_clock; clock <= limits->max_clock; clock++) {
+			for (lane_count = limits->min_lane_count;
+			     lane_count <= limits->max_lane_count;
+			     lane_count <<= 1) {
+				link_clock = intel_dp->common_rates[clock];
+				link_avail = intel_dp_max_data_rate(link_clock,
+								    lane_count);
+
+				if (mode_rate <= link_avail) {
+					pipe_config->lane_count = lane_count;
+					pipe_config->pipe_bpp = bpp;
+					pipe_config->port_clock = link_clock;
+
+					return true;
+				}
+			}
+		}
+	}
+
+	return false;
+}
+
 static bool
 intel_dp_compute_link_config(struct intel_encoder *encoder,
 			     struct intel_crtc_state *pipe_config)
@@ -1711,8 +1747,6 @@ intel_dp_compute_link_config(struct intel_encoder *encoder,
 	struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
 	struct link_config_limits limits;
-	int bpp, clock, lane_count;
-	int mode_rate, link_avail, link_clock;
 	int common_len;
 
 	common_len = intel_dp_common_len_rate_limit(intel_dp,
@@ -1766,37 +1800,22 @@ intel_dp_compute_link_config(struct intel_encoder *encoder,
 		      intel_dp->common_rates[limits.max_clock],
 		      limits.max_bpp, adjusted_mode->crtc_clock);
 
-	for (bpp = limits.max_bpp; bpp >= limits.min_bpp; bpp -= 2 * 3) {
-		mode_rate = intel_dp_link_required(adjusted_mode->crtc_clock,
-						   bpp);
-
-		for (clock = limits.min_clock; clock <= limits.max_clock; clock++) {
-			for (lane_count = limits.min_lane_count;
-			     lane_count <= limits.max_lane_count;
-			     lane_count <<= 1) {
-
-				link_clock = intel_dp->common_rates[clock];
-				link_avail = intel_dp_max_data_rate(link_clock,
-								    lane_count);
-
-				if (mode_rate <= link_avail) {
-					goto found;
-				}
-			}
-		}
-	}
-
-	return false;
-
-found:
-	pipe_config->lane_count = lane_count;
-	pipe_config->pipe_bpp = bpp;
-	pipe_config->port_clock = intel_dp->common_rates[clock];
+	/*
+	 * Optimize for slow and wide. This is the place to add alternative
+	 * optimization policy.
+	 */
+	if (!intel_dp_compute_link_config_wide(intel_dp, pipe_config, &limits))
+		return false;
 
 	DRM_DEBUG_KMS("DP lane count %d clock %d bpp %d\n",
-		      pipe_config->lane_count, pipe_config->port_clock, bpp);
-	DRM_DEBUG_KMS("DP link bw required %i available %i\n",
-		      mode_rate, link_avail);
+		      pipe_config->lane_count, pipe_config->port_clock,
+		      pipe_config->pipe_bpp);
+
+	DRM_DEBUG_KMS("DP link rate required %i available %i\n",
+		      intel_dp_link_required(adjusted_mode->crtc_clock,
+					     pipe_config->pipe_bpp),
+		      intel_dp_max_data_rate(pipe_config->port_clock,
+					     pipe_config->lane_count));
 
 	return true;
 }
-- 
cgit v1.1


From a49714531be37111b1540d85a70bb55d0958272c Mon Sep 17 00:00:00 2001
From: Jani Nikula <jani.nikula@intel.com>
Date: Thu, 26 Apr 2018 11:25:30 +0300
Subject: drm/i915/dp: fix compliance test adjustments

Abstract compliance test adjustments to a single function. Also make the
bpc adjustments affect the limits, actually forcing the bpc. Seems like
directly changing the pipe_bpp in the past could not have been
effective.

Reviewed-by: Manasi Navare <manasi.d.navare@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/ef61e76003ab7719c82810b742f3fb5765c0e14c.1524730974.git.jani.nikula@intel.com
---
 drivers/gpu/drm/i915/intel_dp.c | 64 ++++++++++++++++++++++++-----------------
 1 file changed, 38 insertions(+), 26 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index d622db7..83da50b 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -1666,14 +1666,6 @@ static int intel_dp_compute_bpp(struct intel_dp *intel_dp,
 	if (bpc > 0)
 		bpp = min(bpp, 3*bpc);
 
-	/* For DP Compliance we override the computed bpp for the pipe */
-	if (intel_dp->compliance.test_data.bpc != 0) {
-		pipe_config->pipe_bpp =	3*intel_dp->compliance.test_data.bpc;
-		pipe_config->dither_force_disable = pipe_config->pipe_bpp == 6*3;
-		DRM_DEBUG_KMS("Setting pipe_bpp to %d\n",
-			      pipe_config->pipe_bpp);
-	}
-
 	if (intel_dp_is_edp(intel_dp)) {
 		/* Get bpp from vbt only for panels that dont have bpp in edid */
 		if (intel_connector->base.display_info.bpc == 0 &&
@@ -1704,6 +1696,42 @@ static bool intel_edp_compare_alt_mode(struct drm_display_mode *m1,
 	return bres;
 }
 
+/* Adjust link config limits based on compliance test requests. */
+static void
+intel_dp_adjust_compliance_config(struct intel_dp *intel_dp,
+				  struct intel_crtc_state *pipe_config,
+				  struct link_config_limits *limits)
+{
+	/* For DP Compliance we override the computed bpp for the pipe */
+	if (intel_dp->compliance.test_data.bpc != 0) {
+		int bpp = 3 * intel_dp->compliance.test_data.bpc;
+
+		limits->min_bpp = limits->max_bpp = bpp;
+		pipe_config->dither_force_disable = bpp == 6 * 3;
+
+		DRM_DEBUG_KMS("Setting pipe_bpp to %d\n", bpp);
+	}
+
+	/* Use values requested by Compliance Test Request */
+	if (intel_dp->compliance.test_type == DP_TEST_LINK_TRAINING) {
+		int index;
+
+		/* Validate the compliance test data since max values
+		 * might have changed due to link train fallback.
+		 */
+		if (intel_dp_link_params_valid(intel_dp, intel_dp->compliance.test_link_rate,
+					       intel_dp->compliance.test_lane_count)) {
+			index = intel_dp_rate_index(intel_dp->common_rates,
+						    intel_dp->num_common_rates,
+						    intel_dp->compliance.test_link_rate);
+			if (index >= 0)
+				limits->min_clock = limits->max_clock = index;
+			limits->min_lane_count = limits->max_lane_count =
+				intel_dp->compliance.test_lane_count;
+		}
+	}
+}
+
 /* Optimize link config in order: max bpp, min clock, min lanes */
 static bool
 intel_dp_compute_link_config_wide(struct intel_dp *intel_dp,
@@ -1764,24 +1792,6 @@ intel_dp_compute_link_config(struct intel_encoder *encoder,
 	limits.min_bpp = 6 * 3;
 	limits.max_bpp = intel_dp_compute_bpp(intel_dp, pipe_config);
 
-	/* Use values requested by Compliance Test Request */
-	if (intel_dp->compliance.test_type == DP_TEST_LINK_TRAINING) {
-		int index;
-
-		/* Validate the compliance test data since max values
-		 * might have changed due to link train fallback.
-		 */
-		if (intel_dp_link_params_valid(intel_dp, intel_dp->compliance.test_link_rate,
-					       intel_dp->compliance.test_lane_count)) {
-			index = intel_dp_rate_index(intel_dp->common_rates,
-						    intel_dp->num_common_rates,
-						    intel_dp->compliance.test_link_rate);
-			if (index >= 0)
-				limits.min_clock = limits.max_clock = index;
-			limits.min_lane_count = limits.max_lane_count = intel_dp->compliance.test_lane_count;
-		}
-	}
-
 	if (intel_dp_is_edp(intel_dp)) {
 		/*
 		 * Use the maximum clock and number of lanes the eDP panel
@@ -1794,6 +1804,8 @@ intel_dp_compute_link_config(struct intel_encoder *encoder,
 		limits.min_clock = limits.max_clock;
 	}
 
+	intel_dp_adjust_compliance_config(intel_dp, pipe_config, &limits);
+
 	DRM_DEBUG_KMS("DP link computation with max lane count %i "
 		      "max rate %d max bpp %d pixel clock %iKHz\n",
 		      limits.max_lane_count,
-- 
cgit v1.1


From 935dff1a218c2162aad8f0e681cbb5d601742412 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Wed, 11 Apr 2018 13:03:46 +0100
Subject: drm/i915/selftests: Wait for idle between idle resets as well
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Even though we weren't injecting guilty requests to be reset, we could
still fall over the issue of resetting the same request too fast -- where
the GPU refuses to start again. (Although it is interesting to note that
reloading the driver is sufficient, suggesting that we could recover if
we delayed the setup after reset?) Continue to paper over the problem by
adding a small delay by waiting for the engine to idle between tests,
and ensure that the engines are idle before starting the idle tests.

v2: Replace single instance of 50 with a magic macro.

References: 028666793a02 ("drm/i915/selftests: Avoid repeatedly harming the same innocent context")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Michał Winiarski <michal.winiarski@intel.com>
Cc: Michel Thierry <michel.thierry@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Michał Winiarski <michal.winiarski@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180411120346.27618-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/selftests/intel_hangcheck.c | 50 +++++++++++++++++++++++-
 1 file changed, 49 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/selftests/intel_hangcheck.c b/drivers/gpu/drm/i915/selftests/intel_hangcheck.c
index f7ee54e..c61bf65 100644
--- a/drivers/gpu/drm/i915/selftests/intel_hangcheck.c
+++ b/drivers/gpu/drm/i915/selftests/intel_hangcheck.c
@@ -30,6 +30,8 @@
 #include "mock_context.h"
 #include "mock_drm.h"
 
+#define IGT_IDLE_TIMEOUT 50 /* ms; time to wait after flushing between tests */
+
 struct hang {
 	struct drm_i915_private *i915;
 	struct drm_i915_gem_object *hws;
@@ -454,6 +456,11 @@ static int igt_global_reset(void *arg)
 	return err;
 }
 
+static bool wait_for_idle(struct intel_engine_cs *engine)
+{
+	return wait_for(intel_engine_is_idle(engine), IGT_IDLE_TIMEOUT) == 0;
+}
+
 static int __igt_reset_engine(struct drm_i915_private *i915, bool active)
 {
 	struct intel_engine_cs *engine;
@@ -481,6 +488,13 @@ static int __igt_reset_engine(struct drm_i915_private *i915, bool active)
 		if (active && !intel_engine_can_store_dword(engine))
 			continue;
 
+		if (!wait_for_idle(engine)) {
+			pr_err("%s failed to idle before reset\n",
+			       engine->name);
+			err = -EIO;
+			break;
+		}
+
 		reset_count = i915_reset_count(&i915->gpu_error);
 		reset_engine_count = i915_reset_engine_count(&i915->gpu_error,
 							     engine);
@@ -542,6 +556,19 @@ static int __igt_reset_engine(struct drm_i915_private *i915, bool active)
 				err = -EINVAL;
 				break;
 			}
+
+			if (!wait_for_idle(engine)) {
+				struct drm_printer p =
+					drm_info_printer(i915->drm.dev);
+
+				pr_err("%s failed to idle after reset\n",
+				       engine->name);
+				intel_engine_dump(engine, &p,
+						  "%s\n", engine->name);
+
+				err = -EIO;
+				break;
+			}
 		} while (time_before(jiffies, end_time));
 		clear_bit(I915_RESET_ENGINE + id, &i915->gpu_error.flags);
 
@@ -696,6 +723,13 @@ static int __igt_reset_engines(struct drm_i915_private *i915,
 		    !intel_engine_can_store_dword(engine))
 			continue;
 
+		if (!wait_for_idle(engine)) {
+			pr_err("i915_reset_engine(%s:%s): failed to idle before reset\n",
+			       engine->name, test_name);
+			err = -EIO;
+			break;
+		}
+
 		memset(threads, 0, sizeof(threads));
 		for_each_engine(other, i915, tmp) {
 			struct task_struct *tsk;
@@ -772,6 +806,20 @@ static int __igt_reset_engines(struct drm_i915_private *i915,
 				i915_request_wait(rq, 0, MAX_SCHEDULE_TIMEOUT);
 				i915_request_put(rq);
 			}
+
+			if (!(flags & TEST_SELF) && !wait_for_idle(engine)) {
+				struct drm_printer p =
+					drm_info_printer(i915->drm.dev);
+
+				pr_err("i915_reset_engine(%s:%s):"
+				       " failed to idle after reset\n",
+				       engine->name, test_name);
+				intel_engine_dump(engine, &p,
+						  "%s\n", engine->name);
+
+				err = -EIO;
+				break;
+			}
 		} while (time_before(jiffies, end_time));
 		clear_bit(I915_RESET_ENGINE + id, &i915->gpu_error.flags);
 		pr_info("i915_reset_engine(%s:%s): %lu resets\n",
@@ -981,7 +1029,7 @@ static int wait_for_others(struct drm_i915_private *i915,
 		if (engine == exclude)
 			continue;
 
-		if (wait_for(intel_engine_is_idle(engine), 10))
+		if (!wait_for_idle(engine))
 			return -EIO;
 	}
 
-- 
cgit v1.1


From fa89f53bd7288d6aa7a982841119e7123faf5a53 Mon Sep 17 00:00:00 2001
From: Evan Wang <xswang@marvell.com>
Date: Fri, 13 Apr 2018 12:32:30 +0800
Subject: libahci: Allow drivers to override stop_engine

Marvell armada37xx, armada7k and armada8k share the same
AHCI sata controller IP, and currently there is an issue
(Errata Ref#226)that the SATA can not be detected via SATA
Port-MultiPlayer(PMP). After debugging, the reason is
found that the value of Port-x FIS-based Switching Control
(PxFBS@0x40) became wrong.
According to design, the bits[11:8, 0] of register PxFBS
are cleared when Port Command and Status (0x18) bit[0]
changes its value from 1 to 0, i.e. falling edge of Port
Command and Status bit[0] sends PULSE that resets PxFBS
bits[11:8; 0].
So it needs save the port PxFBS register before PxCMD
ST write and restore the port PxFBS register afterwards
in ahci_stop_engine().

This commit allows drivers to override ahci_stop_engine
behavior for use by the Marvell AHCI driver(and potentially
other drivers in the future).

Signed-off-by: Evan Wang <xswang@marvell.com>
Cc: Ofer Heifetz <oferh@marvell.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
---
 drivers/ata/ahci.c          |  6 +++---
 drivers/ata/ahci.h          |  7 +++++++
 drivers/ata/ahci_qoriq.c    |  2 +-
 drivers/ata/ahci_xgene.c    |  4 ++--
 drivers/ata/libahci.c       | 20 ++++++++++++--------
 drivers/ata/sata_highbank.c |  2 +-
 6 files changed, 26 insertions(+), 15 deletions(-)

diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c
index 1ff1779..6389c88 100644
--- a/drivers/ata/ahci.c
+++ b/drivers/ata/ahci.c
@@ -698,7 +698,7 @@ static int ahci_vt8251_hardreset(struct ata_link *link, unsigned int *class,
 
 	DPRINTK("ENTER\n");
 
-	ahci_stop_engine(ap);
+	hpriv->stop_engine(ap);
 
 	rc = sata_link_hardreset(link, sata_ehc_deb_timing(&link->eh_context),
 				 deadline, &online, NULL);
@@ -724,7 +724,7 @@ static int ahci_p5wdh_hardreset(struct ata_link *link, unsigned int *class,
 	bool online;
 	int rc;
 
-	ahci_stop_engine(ap);
+	hpriv->stop_engine(ap);
 
 	/* clear D2H reception area to properly wait for D2H FIS */
 	ata_tf_init(link->device, &tf);
@@ -788,7 +788,7 @@ static int ahci_avn_hardreset(struct ata_link *link, unsigned int *class,
 
 	DPRINTK("ENTER\n");
 
-	ahci_stop_engine(ap);
+	hpriv->stop_engine(ap);
 
 	for (i = 0; i < 2; i++) {
 		u16 val;
diff --git a/drivers/ata/ahci.h b/drivers/ata/ahci.h
index a9d996e..824bd39 100644
--- a/drivers/ata/ahci.h
+++ b/drivers/ata/ahci.h
@@ -365,6 +365,13 @@ struct ahci_host_priv {
 	 * be overridden anytime before the host is activated.
 	 */
 	void			(*start_engine)(struct ata_port *ap);
+	/*
+	 * Optional ahci_stop_engine override, if not set this gets set to the
+	 * default ahci_stop_engine during ahci_save_initial_config, this can
+	 * be overridden anytime before the host is activated.
+	 */
+	int			(*stop_engine)(struct ata_port *ap);
+
 	irqreturn_t 		(*irq_handler)(int irq, void *dev_instance);
 
 	/* only required for per-port MSI(-X) support */
diff --git a/drivers/ata/ahci_qoriq.c b/drivers/ata/ahci_qoriq.c
index 2685f28..cfdef4d 100644
--- a/drivers/ata/ahci_qoriq.c
+++ b/drivers/ata/ahci_qoriq.c
@@ -96,7 +96,7 @@ static int ahci_qoriq_hardreset(struct ata_link *link, unsigned int *class,
 
 	DPRINTK("ENTER\n");
 
-	ahci_stop_engine(ap);
+	hpriv->stop_engine(ap);
 
 	/*
 	 * There is a errata on ls1021a Rev1.0 and Rev2.0 which is:
diff --git a/drivers/ata/ahci_xgene.c b/drivers/ata/ahci_xgene.c
index c2b5941..ad58da7 100644
--- a/drivers/ata/ahci_xgene.c
+++ b/drivers/ata/ahci_xgene.c
@@ -165,7 +165,7 @@ static int xgene_ahci_restart_engine(struct ata_port *ap)
 				    PORT_CMD_ISSUE, 0x0, 1, 100))
 		  return -EBUSY;
 
-	ahci_stop_engine(ap);
+	hpriv->stop_engine(ap);
 	ahci_start_fis_rx(ap);
 
 	/*
@@ -421,7 +421,7 @@ static int xgene_ahci_hardreset(struct ata_link *link, unsigned int *class,
 	portrxfis_saved = readl(port_mmio + PORT_FIS_ADDR);
 	portrxfishi_saved = readl(port_mmio + PORT_FIS_ADDR_HI);
 
-	ahci_stop_engine(ap);
+	hpriv->stop_engine(ap);
 
 	rc = xgene_ahci_do_hardreset(link, deadline, &online);
 
diff --git a/drivers/ata/libahci.c b/drivers/ata/libahci.c
index 7adcf3c..e5d9097 100644
--- a/drivers/ata/libahci.c
+++ b/drivers/ata/libahci.c
@@ -560,6 +560,9 @@ void ahci_save_initial_config(struct device *dev, struct ahci_host_priv *hpriv)
 	if (!hpriv->start_engine)
 		hpriv->start_engine = ahci_start_engine;
 
+	if (!hpriv->stop_engine)
+		hpriv->stop_engine = ahci_stop_engine;
+
 	if (!hpriv->irq_handler)
 		hpriv->irq_handler = ahci_single_level_irq_intr;
 }
@@ -897,9 +900,10 @@ static void ahci_start_port(struct ata_port *ap)
 static int ahci_deinit_port(struct ata_port *ap, const char **emsg)
 {
 	int rc;
+	struct ahci_host_priv *hpriv = ap->host->private_data;
 
 	/* disable DMA */
-	rc = ahci_stop_engine(ap);
+	rc = hpriv->stop_engine(ap);
 	if (rc) {
 		*emsg = "failed to stop engine";
 		return rc;
@@ -1310,7 +1314,7 @@ int ahci_kick_engine(struct ata_port *ap)
 	int busy, rc;
 
 	/* stop engine */
-	rc = ahci_stop_engine(ap);
+	rc = hpriv->stop_engine(ap);
 	if (rc)
 		goto out_restart;
 
@@ -1549,7 +1553,7 @@ int ahci_do_hardreset(struct ata_link *link, unsigned int *class,
 
 	DPRINTK("ENTER\n");
 
-	ahci_stop_engine(ap);
+	hpriv->stop_engine(ap);
 
 	/* clear D2H reception area to properly wait for D2H FIS */
 	ata_tf_init(link->device, &tf);
@@ -2075,14 +2079,14 @@ void ahci_error_handler(struct ata_port *ap)
 
 	if (!(ap->pflags & ATA_PFLAG_FROZEN)) {
 		/* restart engine */
-		ahci_stop_engine(ap);
+		hpriv->stop_engine(ap);
 		hpriv->start_engine(ap);
 	}
 
 	sata_pmp_error_handler(ap);
 
 	if (!ata_dev_enabled(ap->link.device))
-		ahci_stop_engine(ap);
+		hpriv->stop_engine(ap);
 }
 EXPORT_SYMBOL_GPL(ahci_error_handler);
 
@@ -2129,7 +2133,7 @@ static void ahci_set_aggressive_devslp(struct ata_port *ap, bool sleep)
 		return;
 
 	/* set DITO, MDAT, DETO and enable DevSlp, need to stop engine first */
-	rc = ahci_stop_engine(ap);
+	rc = hpriv->stop_engine(ap);
 	if (rc)
 		return;
 
@@ -2189,7 +2193,7 @@ static void ahci_enable_fbs(struct ata_port *ap)
 		return;
 	}
 
-	rc = ahci_stop_engine(ap);
+	rc = hpriv->stop_engine(ap);
 	if (rc)
 		return;
 
@@ -2222,7 +2226,7 @@ static void ahci_disable_fbs(struct ata_port *ap)
 		return;
 	}
 
-	rc = ahci_stop_engine(ap);
+	rc = hpriv->stop_engine(ap);
 	if (rc)
 		return;
 
diff --git a/drivers/ata/sata_highbank.c b/drivers/ata/sata_highbank.c
index aafb8cc..e67815b 100644
--- a/drivers/ata/sata_highbank.c
+++ b/drivers/ata/sata_highbank.c
@@ -410,7 +410,7 @@ static int ahci_highbank_hardreset(struct ata_link *link, unsigned int *class,
 	int rc;
 	int retry = 100;
 
-	ahci_stop_engine(ap);
+	hpriv->stop_engine(ap);
 
 	/* clear D2H reception area to properly wait for D2H FIS */
 	ata_tf_init(link->device, &tf);
-- 
cgit v1.1


From daa2e3bdbb0b3e691cf20a042350817310cb8cb5 Mon Sep 17 00:00:00 2001
From: Evan Wang <xswang@marvell.com>
Date: Fri, 13 Apr 2018 12:32:31 +0800
Subject: ata: ahci: mvebu: override ahci_stop_engine for mvebu AHCI

There is an issue(Errata Ref#226) that the SATA can not be
detected via SATA Port-MultiPlayer(PMP) with following
error log:
  ata1.15: PMP product ID mismatch
  ata1.15: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
  ata1.15: Port Multiplier vendor mismatch '0x1b4b'!='0x0'
  ata1.15: PMP revalidation failed (errno=-19)

After debugging, the reason is found that the value Port-x
FIS-based Switching Control(PxFBS@0x40) become wrong.
According to design, the bits[11:8, 0] of register PxFBS
are cleared when Port Command and Status (0x18) bit[0]
changes its value from 1 to 0, i.e. falling edge of Port
Command and Status bit[0] sends PULSE that resets PxFBS
bits[11:8; 0].
So it needs a mvebu SATA WA to save the port PxFBS register
before PxCMD ST write and restore it afterwards.

This patch implements the WA in a separate function of
ahci_mvebu_stop_engine to override ahci_stop_gngine.

Signed-off-by: Evan Wang <xswang@marvell.com>
Cc: Ofer Heifetz <oferh@marvell.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
---
 drivers/ata/ahci_mvebu.c | 56 ++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 56 insertions(+)

diff --git a/drivers/ata/ahci_mvebu.c b/drivers/ata/ahci_mvebu.c
index de7128d..0045dac 100644
--- a/drivers/ata/ahci_mvebu.c
+++ b/drivers/ata/ahci_mvebu.c
@@ -62,6 +62,60 @@ static void ahci_mvebu_regret_option(struct ahci_host_priv *hpriv)
 	writel(0x80, hpriv->mmio + AHCI_VENDOR_SPECIFIC_0_DATA);
 }
 
+/**
+ * ahci_mvebu_stop_engine
+ *
+ * @ap:	Target ata port
+ *
+ * Errata Ref#226 - SATA Disk HOT swap issue when connected through
+ * Port Multiplier in FIS-based Switching mode.
+ *
+ * To avoid the issue, according to design, the bits[11:8, 0] of
+ * register PxFBS are cleared when Port Command and Status (0x18) bit[0]
+ * changes its value from 1 to 0, i.e. falling edge of Port
+ * Command and Status bit[0] sends PULSE that resets PxFBS
+ * bits[11:8; 0].
+ *
+ * This function is used to override function of "ahci_stop_engine"
+ * from libahci.c by adding the mvebu work around(WA) to save PxFBS
+ * value before the PxCMD ST write of 0, then restore PxFBS value.
+ *
+ * Return: 0 on success; Error code otherwise.
+ */
+int ahci_mvebu_stop_engine(struct ata_port *ap)
+{
+	void __iomem *port_mmio = ahci_port_base(ap);
+	u32 tmp, port_fbs;
+
+	tmp = readl(port_mmio + PORT_CMD);
+
+	/* check if the HBA is idle */
+	if ((tmp & (PORT_CMD_START | PORT_CMD_LIST_ON)) == 0)
+		return 0;
+
+	/* save the port PxFBS register for later restore */
+	port_fbs = readl(port_mmio + PORT_FBS);
+
+	/* setting HBA to idle */
+	tmp &= ~PORT_CMD_START;
+	writel(tmp, port_mmio + PORT_CMD);
+
+	/*
+	 * bit #15 PxCMD signal doesn't clear PxFBS,
+	 * restore the PxFBS register right after clearing the PxCMD ST,
+	 * no need to wait for the PxCMD bit #15.
+	 */
+	writel(port_fbs, port_mmio + PORT_FBS);
+
+	/* wait for engine to stop. This could be as long as 500 msec */
+	tmp = ata_wait_register(ap, port_mmio + PORT_CMD,
+				PORT_CMD_LIST_ON, PORT_CMD_LIST_ON, 1, 500);
+	if (tmp & PORT_CMD_LIST_ON)
+		return -EIO;
+
+	return 0;
+}
+
 #ifdef CONFIG_PM_SLEEP
 static int ahci_mvebu_suspend(struct platform_device *pdev, pm_message_t state)
 {
@@ -112,6 +166,8 @@ static int ahci_mvebu_probe(struct platform_device *pdev)
 	if (rc)
 		return rc;
 
+	hpriv->stop_engine = ahci_mvebu_stop_engine;
+
 	if (of_device_is_compatible(pdev->dev.of_node,
 				    "marvell,armada-380-ahci")) {
 		dram = mv_mbus_dram_info();
-- 
cgit v1.1


From b5b4d3a52c8fd6e3fc6469c5a64ca0139c07229e Mon Sep 17 00:00:00 2001
From: Hans de Goede <hdegoede@redhat.com>
Date: Tue, 24 Apr 2018 11:19:07 +0200
Subject: libata: Apply NOLPM quirk for SAMSUNG MZMPC128HBFU-000MV SSD

Kevin Shanahan reports the following repeating errors when using LPM,
causing long delays accessing the disk:

  Apr 23 10:21:43 link kernel: ata1.00: exception Emask 0x0 SAct 0x0 SErr 0x50000 action 0x6 frozen
  Apr 23 10:21:43 link kernel: ata1: SError: { PHYRdyChg CommWake }
  Apr 23 10:21:43 link kernel: ata1.00: failed command: WRITE DMA
  Apr 23 10:21:43 link kernel: ata1.00: cmd ca/00:08:60:5d:cd/00:00:00:00:00/e1 tag 9 dma 4096 out
                                        res 50/01:01:01:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
  Apr 23 10:21:43 link kernel: ata1.00: status: { DRDY }
  Apr 23 10:21:43 link kernel: ata1.00: error: { AMNF }
  Apr 23 10:21:43 link kernel: ata1: hard resetting link
  Apr 23 10:21:43 link kernel: ata1: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
  Apr 23 10:21:43 link kernel: ata1.00: configured for UDMA/133
  Apr 23 10:21:43 link kernel: ata1: EH complete

These go away when switching from med_power_with_dipm to medium_power.

This is somewhat weird as the PM830 datasheet explicitly mentions DIPM
being supported and the idle power-consumption is specified with DIPM
enabled.

There are many OEM customized firmware versions for the PM830, so for now
lets assume this is firmware version specific and blacklist LPM based on
the firmware version.

Cc: Kevin Shanahan <kevin@shanahan.id.au>
Reported-by: Kevin Shanahan <kevin@shanahan.id.au>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
---
 drivers/ata/libata-core.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 8bc71ca..6e400ff 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -4549,6 +4549,9 @@ static const struct ata_blacklist_entry ata_device_blacklist [] = {
 						ATA_HORKAGE_ZERO_AFTER_TRIM |
 						ATA_HORKAGE_NOLPM, },
 
+	/* This specific Samsung model/firmware-rev does not handle LPM well */
+	{ "SAMSUNG MZMPC128HBFU-000MV", "CXM14M1Q", ATA_HORKAGE_NOLPM, },
+
 	/* devices that don't properly handle queued TRIM commands */
 	{ "Micron_M500_*",		NULL,	ATA_HORKAGE_NO_NCQ_TRIM |
 						ATA_HORKAGE_ZERO_AFTER_TRIM, },
-- 
cgit v1.1


From 0c92c7a3c5d416f47b32c5f20a611dfeca5d5f2e Mon Sep 17 00:00:00 2001
From: Song Liu <songliubraving@fb.com>
Date: Mon, 23 Apr 2018 10:21:34 -0700
Subject: tracing: Fix bad use of igrab in trace_uprobe.c

As Miklos reported and suggested:

  This pattern repeats two times in trace_uprobe.c and in
  kernel/events/core.c as well:

      ret = kern_path(filename, LOOKUP_FOLLOW, &path);
      if (ret)
          goto fail_address_parse;

      inode = igrab(d_inode(path.dentry));
      path_put(&path);

  And it's wrong.  You can only hold a reference to the inode if you
  have an active ref to the superblock as well (which is normally
  through path.mnt) or holding s_umount.

  This way unmounting the containing filesystem while the tracepoint is
  active will give you the "VFS: Busy inodes after unmount..." message
  and a crash when the inode is finally put.

  Solution: store path instead of inode.

This patch fixes two instances in trace_uprobe.c. struct path is added to
struct trace_uprobe to keep the inode and containing mount point
referenced.

Link: http://lkml.kernel.org/r/20180423172135.4050588-1-songliubraving@fb.com

Fixes: f3f096cfedf8 ("tracing: Provide trace events interface for uprobes")
Fixes: 33ea4b24277b ("perf/core: Implement the 'perf_uprobe' PMU")
Cc: stable@vger.kernel.org
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Howard McLauchlan <hmclauchlan@fb.com>
Cc: Josef Bacik <jbacik@fb.com>
Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Acked-by: Miklos Szeredi <mszeredi@redhat.com>
Reported-by: Miklos Szeredi <miklos@szeredi.hu>
Signed-off-by: Song Liu <songliubraving@fb.com>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
---
 kernel/trace/trace_uprobe.c | 35 ++++++++++++++---------------------
 1 file changed, 14 insertions(+), 21 deletions(-)

diff --git a/kernel/trace/trace_uprobe.c b/kernel/trace/trace_uprobe.c
index 34fd0e0..ac89287 100644
--- a/kernel/trace/trace_uprobe.c
+++ b/kernel/trace/trace_uprobe.c
@@ -55,6 +55,7 @@ struct trace_uprobe {
 	struct list_head		list;
 	struct trace_uprobe_filter	filter;
 	struct uprobe_consumer		consumer;
+	struct path			path;
 	struct inode			*inode;
 	char				*filename;
 	unsigned long			offset;
@@ -289,7 +290,7 @@ static void free_trace_uprobe(struct trace_uprobe *tu)
 	for (i = 0; i < tu->tp.nr_args; i++)
 		traceprobe_free_probe_arg(&tu->tp.args[i]);
 
-	iput(tu->inode);
+	path_put(&tu->path);
 	kfree(tu->tp.call.class->system);
 	kfree(tu->tp.call.name);
 	kfree(tu->filename);
@@ -363,7 +364,6 @@ end:
 static int create_trace_uprobe(int argc, char **argv)
 {
 	struct trace_uprobe *tu;
-	struct inode *inode;
 	char *arg, *event, *group, *filename;
 	char buf[MAX_EVENT_NAME_LEN];
 	struct path path;
@@ -371,7 +371,6 @@ static int create_trace_uprobe(int argc, char **argv)
 	bool is_delete, is_return;
 	int i, ret;
 
-	inode = NULL;
 	ret = 0;
 	is_delete = false;
 	is_return = false;
@@ -437,21 +436,16 @@ static int create_trace_uprobe(int argc, char **argv)
 	}
 	/* Find the last occurrence, in case the path contains ':' too. */
 	arg = strrchr(argv[1], ':');
-	if (!arg) {
-		ret = -EINVAL;
-		goto fail_address_parse;
-	}
+	if (!arg)
+		return -EINVAL;
 
 	*arg++ = '\0';
 	filename = argv[1];
 	ret = kern_path(filename, LOOKUP_FOLLOW, &path);
 	if (ret)
-		goto fail_address_parse;
-
-	inode = igrab(d_real_inode(path.dentry));
-	path_put(&path);
+		return ret;
 
-	if (!inode || !S_ISREG(inode->i_mode)) {
+	if (!d_is_reg(path.dentry)) {
 		ret = -EINVAL;
 		goto fail_address_parse;
 	}
@@ -490,7 +484,7 @@ static int create_trace_uprobe(int argc, char **argv)
 		goto fail_address_parse;
 	}
 	tu->offset = offset;
-	tu->inode = inode;
+	tu->path = path;
 	tu->filename = kstrdup(filename, GFP_KERNEL);
 
 	if (!tu->filename) {
@@ -558,7 +552,7 @@ error:
 	return ret;
 
 fail_address_parse:
-	iput(inode);
+	path_put(&path);
 
 	pr_info("Failed to parse address or file.\n");
 
@@ -922,6 +916,7 @@ probe_event_enable(struct trace_uprobe *tu, struct trace_event_file *file,
 		goto err_flags;
 
 	tu->consumer.filter = filter;
+	tu->inode = d_real_inode(tu->path.dentry);
 	ret = uprobe_register(tu->inode, tu->offset, &tu->consumer);
 	if (ret)
 		goto err_buffer;
@@ -967,6 +962,7 @@ probe_event_disable(struct trace_uprobe *tu, struct trace_event_file *file)
 	WARN_ON(!uprobe_filter_is_empty(&tu->filter));
 
 	uprobe_unregister(tu->inode, tu->offset, &tu->consumer);
+	tu->inode = NULL;
 	tu->tp.flags &= file ? ~TP_FLAG_TRACE : ~TP_FLAG_PROFILE;
 
 	uprobe_buffer_disable();
@@ -1337,7 +1333,6 @@ struct trace_event_call *
 create_local_trace_uprobe(char *name, unsigned long offs, bool is_return)
 {
 	struct trace_uprobe *tu;
-	struct inode *inode;
 	struct path path;
 	int ret;
 
@@ -1345,11 +1340,8 @@ create_local_trace_uprobe(char *name, unsigned long offs, bool is_return)
 	if (ret)
 		return ERR_PTR(ret);
 
-	inode = igrab(d_inode(path.dentry));
-	path_put(&path);
-
-	if (!inode || !S_ISREG(inode->i_mode)) {
-		iput(inode);
+	if (!d_is_reg(path.dentry)) {
+		path_put(&path);
 		return ERR_PTR(-EINVAL);
 	}
 
@@ -1364,11 +1356,12 @@ create_local_trace_uprobe(char *name, unsigned long offs, bool is_return)
 	if (IS_ERR(tu)) {
 		pr_info("Failed to allocate trace_uprobe.(%d)\n",
 			(int)PTR_ERR(tu));
+		path_put(&path);
 		return ERR_CAST(tu);
 	}
 
 	tu->offset = offs;
-	tu->inode = inode;
+	tu->path = path;
 	tu->filename = kstrdup(name, GFP_KERNEL);
 	init_trace_event_call(tu, &tu->tp.call);
 
-- 
cgit v1.1


From 61f94203c9efcaf44a7435298697caf406476c79 Mon Sep 17 00:00:00 2001
From: Song Liu <songliubraving@fb.com>
Date: Mon, 23 Apr 2018 10:21:35 -0700
Subject: tracing: Remove igrab() iput() call from uprobes.c

Caller of uprobe_register is required to keep the inode and containing
mount point referenced.

There was misuse of igrab() in uprobes.c and trace_uprobe.c. This is
because igrab() will not prevent umount of the containing mount point.
To fix this, we added path to struct trace_uprobe, which keeps the inode
and containing mount reference.

For uprobes.c, it is not necessary to call igrab() in uprobe_register(),
as the caller is required to keep the inode reference. The igrab() is
removed and comments on this requirement is added to uprobe_register().

Link: http://lkml.kernel.org/r/CAELBmZB2XX=qEOLAdvGG4cPx4GEntcSnWQquJLUK1ongRj35cA@mail.gmail.com
Link: http://lkml.kernel.org/r/20180423172135.4050588-2-songliubraving@fb.com

Cc: Ingo Molnar <mingo@redhat.com>
Cc: Howard McLauchlan <hmclauchlan@fb.com>
Cc: Josef Bacik <jbacik@fb.com>
Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Acked-by: Miklos Szeredi <mszeredi@redhat.com>
Signed-off-by: Song Liu <songliubraving@fb.com>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
---
 kernel/events/uprobes.c | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)

diff --git a/kernel/events/uprobes.c b/kernel/events/uprobes.c
index ce6848e..1725b90 100644
--- a/kernel/events/uprobes.c
+++ b/kernel/events/uprobes.c
@@ -491,7 +491,7 @@ static struct uprobe *alloc_uprobe(struct inode *inode, loff_t offset)
 	if (!uprobe)
 		return NULL;
 
-	uprobe->inode = igrab(inode);
+	uprobe->inode = inode;
 	uprobe->offset = offset;
 	init_rwsem(&uprobe->register_rwsem);
 	init_rwsem(&uprobe->consumer_rwsem);
@@ -502,7 +502,6 @@ static struct uprobe *alloc_uprobe(struct inode *inode, loff_t offset)
 	if (cur_uprobe) {
 		kfree(uprobe);
 		uprobe = cur_uprobe;
-		iput(inode);
 	}
 
 	return uprobe;
@@ -701,7 +700,6 @@ static void delete_uprobe(struct uprobe *uprobe)
 	rb_erase(&uprobe->rb_node, &uprobes_tree);
 	spin_unlock(&uprobes_treelock);
 	RB_CLEAR_NODE(&uprobe->rb_node); /* for uprobe_is_active() */
-	iput(uprobe->inode);
 	put_uprobe(uprobe);
 }
 
@@ -873,7 +871,8 @@ static void __uprobe_unregister(struct uprobe *uprobe, struct uprobe_consumer *u
  * tuple).  Creation refcount stops uprobe_unregister from freeing the
  * @uprobe even before the register operation is complete. Creation
  * refcount is released when the last @uc for the @uprobe
- * unregisters.
+ * unregisters. Caller of uprobe_register() is required to keep @inode
+ * (and the containing mount) referenced.
  *
  * Return errno if it cannot successully install probes
  * else return 0 (success)
-- 
cgit v1.1


From 0566e40ce7c493d39006cdd7edf17bfdc52eb2ac Mon Sep 17 00:00:00 2001
From: Rishabh Bhatnagar <rishabhb@codeaurora.org>
Date: Wed, 25 Apr 2018 16:42:35 -0700
Subject: tracing: initcall: Ordered comparison of function pointers

Using initcall_t in the __field macro generates the following warning
with clang version 6.0:

include/trace/events/initcall.h:34:3: warning: ordered comparison of
function pointers ('initcall_t' (aka 'int (*)(void)') and 'initcall_t')

__field macro expands to __field_ext macro which does is_signed_type
check on the type argument. Since initcall_t is defined as a function
pointer, using it as the type in the __field macro, leads to an ordered
comparison of function pointer warning, inside the check. Using
__field_struct macro avoids the issue.

Link: http://lkml.kernel.org/r/1524699755-29388-1-git-send-email-rishabhb@codeaurora.org

Signed-off-by: Rishabh Bhatnagar <rishabhb@codeaurora.org>
[ Added comment to why we are using field_struct() ]
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
---
 include/trace/events/initcall.h | 14 +++++++++++---
 1 file changed, 11 insertions(+), 3 deletions(-)

diff --git a/include/trace/events/initcall.h b/include/trace/events/initcall.h
index 8d6cf10..eb903c3 100644
--- a/include/trace/events/initcall.h
+++ b/include/trace/events/initcall.h
@@ -31,7 +31,11 @@ TRACE_EVENT(initcall_start,
 	TP_ARGS(func),
 
 	TP_STRUCT__entry(
-		__field(initcall_t, func)
+		/*
+		 * Use field_struct to avoid is_signed_type()
+		 * comparison of a function pointer
+		 */
+		__field_struct(initcall_t, func)
 	),
 
 	TP_fast_assign(
@@ -48,8 +52,12 @@ TRACE_EVENT(initcall_finish,
 	TP_ARGS(func, ret),
 
 	TP_STRUCT__entry(
-		__field(initcall_t,	func)
-		__field(int,		ret)
+		/*
+		 * Use field_struct to avoid is_signed_type()
+		 * comparison of a function pointer
+		 */
+		__field_struct(initcall_t,	func)
+		__field(int,			ret)
 	),
 
 	TP_fast_assign(
-- 
cgit v1.1


From 1ccef350db2f13715040a10df77ae672206004cf Mon Sep 17 00:00:00 2001
From: Jianbo Liu <jianbol@mellanox.com>
Date: Tue, 27 Mar 2018 09:22:16 +0000
Subject: net/mlx5e: Allow offloading ipv4 header re-write for icmp

For ICMPv4, the checksum is calculated from the ICMP headers and data.
Since the ICMPv4 checksum doesn't cover the IP header, we can allow to
do L3 header re-write for this protocol.

Fixes: bdd66ac0aeed ('net/mlx5e: Disallow TC offloading of unsupported match/action combinations')
Signed-off-by: Jianbo Liu <jianbol@mellanox.com>
Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 4197001..3c534fc 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -1864,7 +1864,8 @@ static bool modify_header_match_supported(struct mlx5_flow_spec *spec,
 	}
 
 	ip_proto = MLX5_GET(fte_match_set_lyr_2_4, headers_v, ip_protocol);
-	if (modify_ip_header && ip_proto != IPPROTO_TCP && ip_proto != IPPROTO_UDP) {
+	if (modify_ip_header && ip_proto != IPPROTO_TCP &&
+	    ip_proto != IPPROTO_UDP && ip_proto != IPPROTO_ICMP) {
 		pr_info("can't offload re-write of ip proto %d\n", ip_proto);
 		return false;
 	}
-- 
cgit v1.1


From 35f80acb24cd53dabd65e0660e46afdf5c45991d Mon Sep 17 00:00:00 2001
From: Huy Nguyen <huyn@mellanox.com>
Date: Thu, 21 Dec 2017 15:15:24 -0600
Subject: net/mlx5e: DCBNL fix min inline header size for dscp

When the trust state is set to dscp and the netdev is down, the inline
header size is not updated. When netdev is up, the inline header size
stays at L2 instead of IP.

Fix this issue by updating the private parameter when the netdev is in
down so that when netdev is up, it picks up the right header size.

Fixes: fbcb127e89ba ("net/mlx5e: Support DSCP trust state ...")
Signed-off-by: Huy Nguyen <huyn@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c b/drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c
index 3d46ef4..c641d56 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c
@@ -1007,12 +1007,14 @@ static void mlx5e_trust_update_sq_inline_mode(struct mlx5e_priv *priv)
 
 	mutex_lock(&priv->state_lock);
 
-	if (!test_bit(MLX5E_STATE_OPENED, &priv->state))
-		goto out;
-
 	new_channels.params = priv->channels.params;
 	mlx5e_trust_update_tx_min_inline_mode(priv, &new_channels.params);
 
+	if (!test_bit(MLX5E_STATE_OPENED, &priv->state)) {
+		priv->channels.params = new_channels.params;
+		goto out;
+	}
+
 	/* Skip if tx_min_inline is the same */
 	if (new_channels.params.tx_min_inline_mode ==
 	    priv->channels.params.tx_min_inline_mode)
-- 
cgit v1.1


From 6082d9c9c94a408d7409b5f2e4e42ac9e8b16d0d Mon Sep 17 00:00:00 2001
From: Israel Rukshin <israelr@mellanox.com>
Date: Thu, 12 Apr 2018 09:49:11 +0000
Subject: net/mlx5: Fix mlx5_get_vector_affinity function

Adding the vector offset when calling to mlx5_vector2eqn() is wrong.
This is because mlx5_vector2eqn() checks if EQ index is equal to vector number
and the fact that the internal completion vectors that mlx5 allocates
don't get an EQ index.

The second problem here is that using effective_affinity_mask gives the same
CPU for different vectors.
This leads to unmapped queues when calling it from blk_mq_rdma_map_queues().
This doesn't happen when using affinity_hint mask.

Fixes: 2572cf57d75a ("mlx5: fix mlx5_get_vector_affinity to start from completion vector 0")
Fixes: 05e0cc84e00c ("net/mlx5: Fix get vector affinity helper function")
Signed-off-by: Israel Rukshin <israelr@mellanox.com>
Reviewed-by: Max Gurtovoy <maxg@mellanox.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
---
 drivers/infiniband/hw/mlx5/main.c |  2 +-
 include/linux/mlx5/driver.h       | 12 +++---------
 2 files changed, 4 insertions(+), 10 deletions(-)

diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index daa919e..241cf4f 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -4757,7 +4757,7 @@ mlx5_ib_get_vector_affinity(struct ib_device *ibdev, int comp_vector)
 {
 	struct mlx5_ib_dev *dev = to_mdev(ibdev);
 
-	return mlx5_get_vector_affinity(dev->mdev, comp_vector);
+	return mlx5_get_vector_affinity_hint(dev->mdev, comp_vector);
 }
 
 /* The mlx5_ib_multiport_mutex should be held when calling this function */
diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h
index 767d193..2a156c5 100644
--- a/include/linux/mlx5/driver.h
+++ b/include/linux/mlx5/driver.h
@@ -1284,25 +1284,19 @@ enum {
 };
 
 static inline const struct cpumask *
-mlx5_get_vector_affinity(struct mlx5_core_dev *dev, int vector)
+mlx5_get_vector_affinity_hint(struct mlx5_core_dev *dev, int vector)
 {
-	const struct cpumask *mask;
 	struct irq_desc *desc;
 	unsigned int irq;
 	int eqn;
 	int err;
 
-	err = mlx5_vector2eqn(dev, MLX5_EQ_VEC_COMP_BASE + vector, &eqn, &irq);
+	err = mlx5_vector2eqn(dev, vector, &eqn, &irq);
 	if (err)
 		return NULL;
 
 	desc = irq_to_desc(irq);
-#ifdef CONFIG_GENERIC_IRQ_EFFECTIVE_AFF_MASK
-	mask = irq_data_get_effective_affinity_mask(&desc->irq_data);
-#else
-	mask = desc->irq_common_data.affinity;
-#endif
-	return mask;
+	return desc->affinity_hint;
 }
 
 #endif /* MLX5_DRIVER_H */
-- 
cgit v1.1


From d9a96ec362e3da878c378854e25321c85bac52c2 Mon Sep 17 00:00:00 2001
From: Tariq Toukan <tariqt@mellanox.com>
Date: Tue, 20 Mar 2018 18:17:25 +0200
Subject: net/mlx5e: TX, Use correct counter in dma_map error flow

In case of a dma_mapping_error, do not use wi->num_dma
as a parameter for dma unmap function because it's yet
to be set, and holds an out-of-date value.
Use actual value (local variable num_dma) instead.

Fixes: 34802a42b352 ("net/mlx5e: Do not modify the TX SKB")
Fixes: e586b3b0baee ("net/mlx5: Ethernet Datapath files")
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_tx.c | 20 ++++++++++----------
 1 file changed, 10 insertions(+), 10 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c
index 2029710..5532aa3 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c
@@ -255,7 +255,7 @@ mlx5e_txwqe_build_dsegs(struct mlx5e_txqsq *sq, struct sk_buff *skb,
 		dma_addr = dma_map_single(sq->pdev, skb_data, headlen,
 					  DMA_TO_DEVICE);
 		if (unlikely(dma_mapping_error(sq->pdev, dma_addr)))
-			return -ENOMEM;
+			goto dma_unmap_wqe_err;
 
 		dseg->addr       = cpu_to_be64(dma_addr);
 		dseg->lkey       = sq->mkey_be;
@@ -273,7 +273,7 @@ mlx5e_txwqe_build_dsegs(struct mlx5e_txqsq *sq, struct sk_buff *skb,
 		dma_addr = skb_frag_dma_map(sq->pdev, frag, 0, fsz,
 					    DMA_TO_DEVICE);
 		if (unlikely(dma_mapping_error(sq->pdev, dma_addr)))
-			return -ENOMEM;
+			goto dma_unmap_wqe_err;
 
 		dseg->addr       = cpu_to_be64(dma_addr);
 		dseg->lkey       = sq->mkey_be;
@@ -285,6 +285,10 @@ mlx5e_txwqe_build_dsegs(struct mlx5e_txqsq *sq, struct sk_buff *skb,
 	}
 
 	return num_dma;
+
+dma_unmap_wqe_err:
+	mlx5e_dma_unmap_wqe_err(sq, num_dma);
+	return -ENOMEM;
 }
 
 static inline void
@@ -380,17 +384,15 @@ static netdev_tx_t mlx5e_sq_xmit(struct mlx5e_txqsq *sq, struct sk_buff *skb,
 	num_dma = mlx5e_txwqe_build_dsegs(sq, skb, skb_data, headlen,
 					  (struct mlx5_wqe_data_seg *)cseg + ds_cnt);
 	if (unlikely(num_dma < 0))
-		goto dma_unmap_wqe_err;
+		goto err_drop;
 
 	mlx5e_txwqe_complete(sq, skb, opcode, ds_cnt + num_dma,
 			     num_bytes, num_dma, wi, cseg);
 
 	return NETDEV_TX_OK;
 
-dma_unmap_wqe_err:
+err_drop:
 	sq->stats.dropped++;
-	mlx5e_dma_unmap_wqe_err(sq, wi->num_dma);
-
 	dev_kfree_skb_any(skb);
 
 	return NETDEV_TX_OK;
@@ -645,17 +647,15 @@ netdev_tx_t mlx5i_sq_xmit(struct mlx5e_txqsq *sq, struct sk_buff *skb,
 	num_dma = mlx5e_txwqe_build_dsegs(sq, skb, skb_data, headlen,
 					  (struct mlx5_wqe_data_seg *)cseg + ds_cnt);
 	if (unlikely(num_dma < 0))
-		goto dma_unmap_wqe_err;
+		goto err_drop;
 
 	mlx5e_txwqe_complete(sq, skb, opcode, ds_cnt + num_dma,
 			     num_bytes, num_dma, wi, cseg);
 
 	return NETDEV_TX_OK;
 
-dma_unmap_wqe_err:
+err_drop:
 	sq->stats.dropped++;
-	mlx5e_dma_unmap_wqe_err(sq, wi->num_dma);
-
 	dev_kfree_skb_any(skb);
 
 	return NETDEV_TX_OK;
-- 
cgit v1.1


From 9c26f5f89d01ca21560c6b8a8e4054c271cc3a9c Mon Sep 17 00:00:00 2001
From: Talat Batheesh <talatb@mellanox.com>
Date: Sun, 15 Apr 2018 11:26:19 +0300
Subject: net/mlx5: Avoid cleaning flow steering table twice during error flow

When we fail to initialize the RX root namespace, we need
to clean only that and not the entire flow steering.

Currently the code may try to clean the flow steering twice
on error witch leads to null pointer deference.
Make sure we clean correctly.

Fixes: fba53f7b5719 ("net/mlx5: Introduce mlx5_flow_steering structure")
Signed-off-by: Talat Batheesh <talatb@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/fs_core.c | 23 ++++++++++++++---------
 1 file changed, 14 insertions(+), 9 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
index de51e7c..2595c67 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
@@ -187,6 +187,7 @@ static void del_sw_ns(struct fs_node *node);
 static void del_sw_hw_rule(struct fs_node *node);
 static bool mlx5_flow_dests_cmp(struct mlx5_flow_destination *d1,
 				struct mlx5_flow_destination *d2);
+static void cleanup_root_ns(struct mlx5_flow_root_namespace *root_ns);
 static struct mlx5_flow_rule *
 find_flow_rule(struct fs_fte *fte,
 	       struct mlx5_flow_destination *dest);
@@ -2351,23 +2352,27 @@ static int create_anchor_flow_table(struct mlx5_flow_steering *steering)
 
 static int init_root_ns(struct mlx5_flow_steering *steering)
 {
+	int err;
+
 	steering->root_ns = create_root_ns(steering, FS_FT_NIC_RX);
 	if (!steering->root_ns)
-		goto cleanup;
+		return -ENOMEM;
 
-	if (init_root_tree(steering, &root_fs, &steering->root_ns->ns.node))
-		goto cleanup;
+	err = init_root_tree(steering, &root_fs, &steering->root_ns->ns.node);
+	if (err)
+		goto out_err;
 
 	set_prio_attrs(steering->root_ns);
-
-	if (create_anchor_flow_table(steering))
-		goto cleanup;
+	err = create_anchor_flow_table(steering);
+	if (err)
+		goto out_err;
 
 	return 0;
 
-cleanup:
-	mlx5_cleanup_fs(steering->dev);
-	return -ENOMEM;
+out_err:
+	cleanup_root_ns(steering->root_ns);
+	steering->root_ns = NULL;
+	return err;
 }
 
 static void clean_tree(struct fs_node *node)
-- 
cgit v1.1


From 99beaa22f11152e128861f1c681744ca4749e9f8 Mon Sep 17 00:00:00 2001
From: Shahar Klein <shahark@mellanox.com>
Date: Sun, 8 Apr 2018 09:50:53 +0300
Subject: net/mlx5e: Fix traffic between VF and representor

After the cited commit, WQE RQ size is calculated based on sw_mtu but it
was not set for representors. This commit fixes that.

Fixes: 472a1e44b349 ("net/mlx5e: Save MTU in channels params")
Signed-off-by: Shahar Klein <shahark@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_rep.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
index d8f68e4..876c3e4 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
@@ -877,13 +877,14 @@ static const struct net_device_ops mlx5e_netdev_ops_rep = {
 };
 
 static void mlx5e_build_rep_params(struct mlx5_core_dev *mdev,
-				   struct mlx5e_params *params)
+				   struct mlx5e_params *params, u16 mtu)
 {
 	u8 cq_period_mode = MLX5_CAP_GEN(mdev, cq_period_start_from_cqe) ?
 					 MLX5_CQ_PERIOD_MODE_START_FROM_CQE :
 					 MLX5_CQ_PERIOD_MODE_START_FROM_EQE;
 
 	params->hard_mtu    = MLX5E_ETH_HARD_MTU;
+	params->sw_mtu      = mtu;
 	params->log_sq_size = MLX5E_REP_PARAMS_LOG_SQ_SIZE;
 	params->rq_wq_type  = MLX5_WQ_TYPE_LINKED_LIST;
 	params->log_rq_mtu_frames = MLX5E_REP_PARAMS_LOG_RQ_SIZE;
@@ -931,7 +932,7 @@ static void mlx5e_init_rep(struct mlx5_core_dev *mdev,
 
 	priv->channels.params.num_channels = profile->max_nch(mdev);
 
-	mlx5e_build_rep_params(mdev, &priv->channels.params);
+	mlx5e_build_rep_params(mdev, &priv->channels.params, netdev->mtu);
 	mlx5e_build_rep_netdev(netdev);
 
 	mlx5e_timestamp_init(priv);
-- 
cgit v1.1


From 202854e9f4df99df1f79962a9e8f94a7de602f7b Mon Sep 17 00:00:00 2001
From: Chris Mi <chrism@mellanox.com>
Date: Mon, 16 Apr 2018 10:36:51 +0900
Subject: net/mlx5: Properly deal with flow counters when deleting rules

When deleting a flow counter, the modify mask should be the action and
the flow counter. Otherwise the flow counter is not deleted and we'll
get a firmware warning when deleting the remaining destinations on the
same FTE.

It only happens in the presence of flow counter and multiple vport
destinations. If there is only one vport destination, there is no
need to update the FTE when deleting the only vport destination,
we just delete the FTE.

Fixes: ae05831424ed ("net/mlx5: Add option to add fwd rule with counter")
Signed-off-by: Chris Mi <chrism@mellanox.com>
Signed-off-by: Jianbo Liu <jianbol@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/fs_core.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
index 2595c67..c39c169 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
@@ -482,7 +482,8 @@ static void del_sw_hw_rule(struct fs_node *node)
 
 	if (rule->dest_attr.type == MLX5_FLOW_DESTINATION_TYPE_COUNTER  &&
 	    --fte->dests_size) {
-		modify_mask = BIT(MLX5_SET_FTE_MODIFY_ENABLE_MASK_ACTION);
+		modify_mask = BIT(MLX5_SET_FTE_MODIFY_ENABLE_MASK_ACTION) |
+			      BIT(MLX5_SET_FTE_MODIFY_ENABLE_MASK_FLOW_COUNTERS);
 		fte->action.action &= ~MLX5_FLOW_CONTEXT_ACTION_COUNT;
 		update_fte = true;
 		goto out;
-- 
cgit v1.1


From 815425567dea6c54494e85050631d6bdda907c5d Mon Sep 17 00:00:00 2001
From: John Fastabend <john.fastabend@gmail.com>
Date: Wed, 25 Apr 2018 15:08:53 -0700
Subject: bpf: fix uninitialized variable in bpf tools
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Here the variable cont is used as the saved_pointer for a call to
strtok_r(). It is safe to use the value uninitialized in this
context however and the later reference is only ever used if
the strtok_r is successful. But, 'gcc-5' at least doesn't have all
this knowledge so initialize cont to NULL. Additionally, do the
natural NULL check before accessing just for completness.

The warning is the following:

./bpf/tools/bpf/bpf_dbg.c: In function ‘cmd_load’:
./bpf/tools/bpf/bpf_dbg.c:1077:13: warning: ‘cont’ may be used uninitialized in this function [-Wmaybe-uninitialized]
  } else if (matches(subcmd, "pcap") == 0) {

Fixes: fd981e3c321a "filter: bpf_dbg: add minimal bpf debugger"
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
---
 tools/bpf/bpf_dbg.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/tools/bpf/bpf_dbg.c b/tools/bpf/bpf_dbg.c
index 4f254bc..61b9aa5 100644
--- a/tools/bpf/bpf_dbg.c
+++ b/tools/bpf/bpf_dbg.c
@@ -1063,7 +1063,7 @@ static int cmd_load_pcap(char *file)
 
 static int cmd_load(char *arg)
 {
-	char *subcmd, *cont, *tmp = strdup(arg);
+	char *subcmd, *cont = NULL, *tmp = strdup(arg);
 	int ret = CMD_OK;
 
 	subcmd = strtok_r(tmp, " ", &cont);
@@ -1073,7 +1073,10 @@ static int cmd_load(char *arg)
 		bpf_reset();
 		bpf_reset_breakpoints();
 
-		ret = cmd_load_bpf(cont);
+		if (!cont)
+			ret = CMD_ERR;
+		else
+			ret = cmd_load_bpf(cont);
 	} else if (matches(subcmd, "pcap") == 0) {
 		ret = cmd_load_pcap(cont);
 	} else {
-- 
cgit v1.1


From 184add2ca23ce5edcac0ab9c3b9be13f91e7b567 Mon Sep 17 00:00:00 2001
From: Hans de Goede <hdegoede@redhat.com>
Date: Thu, 26 Apr 2018 22:32:21 +0200
Subject: libata: Apply NOLPM quirk for SanDisk SD7UB3Q*G1001 SSDs

Richard Jones has reported that using med_power_with_dipm on a T450s
with a Sandisk SD7UB3Q256G1001 SSD (firmware version X2180501) is
causing the machine to hang.

Switching the LPM to max_performance fixes this, so it seems that
this Sandisk SSD does not handle LPM well.

Note in the past there have been bug-reports about the following
Sandisk models not working with min_power, so we may need to extend
the quirk list in the future: name - firmware
Sandisk SD6SB2M512G1022I   - X210400
Sandisk SD6PP4M-256G-1006  - A200906

Cc: stable@vger.kernel.org
Cc: Richard W.M. Jones <rjones@redhat.com>
Reported-and-tested-by: Richard W.M. Jones <rjones@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
---
 drivers/ata/libata-core.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 6e400ff..68596bd4 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -4552,6 +4552,9 @@ static const struct ata_blacklist_entry ata_device_blacklist [] = {
 	/* This specific Samsung model/firmware-rev does not handle LPM well */
 	{ "SAMSUNG MZMPC128HBFU-000MV", "CXM14M1Q", ATA_HORKAGE_NOLPM, },
 
+	/* Sandisk devices which are known to not handle LPM well */
+	{ "SanDisk SD7UB3Q*G1001",	NULL,	ATA_HORKAGE_NOLPM, },
+
 	/* devices that don't properly handle queued TRIM commands */
 	{ "Micron_M500_*",		NULL,	ATA_HORKAGE_NO_NCQ_TRIM |
 						ATA_HORKAGE_ZERO_AFTER_TRIM, },
-- 
cgit v1.1


From 75cbec033c08f6d41c4775784f66ab860d02a6b5 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?os=C3=A9=20Roberto=20de=20Souza?= <jose.souza@intel.com>
Date: Wed, 25 Apr 2018 14:23:31 -0700
Subject: drm/i915/psr: Prevent PSR exit when a non-pipe related register is
 written
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Any write in any display register was causing HW to exit PSR,
masking it to allow more power savings. Writes to pipe related
registers will still cause HW to exit PSR.
This is already masked for PSR2.

It also do not break the Display WA #0884, writes to CURSURFLIVE
are still causing hardware to exit PSR. This was tested in CNL machine
by triggering a write to CURSURFLIVE when a debugfs was read by user.

Bspec: 7721 and 8042

v4: Checked that it do not breaks WA #0884 and added this information
to the commit message.

Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180425212334.21109-1-jose.souza@intel.com
---
 drivers/gpu/drm/i915/intel_psr.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/intel_psr.c b/drivers/gpu/drm/i915/intel_psr.c
index 0d54829..e35a3b9 100644
--- a/drivers/gpu/drm/i915/intel_psr.c
+++ b/drivers/gpu/drm/i915/intel_psr.c
@@ -667,7 +667,8 @@ static void hsw_psr_enable_source(struct intel_dp *intel_dp,
 		I915_WRITE(EDP_PSR_DEBUG,
 			   EDP_PSR_DEBUG_MASK_MEMUP |
 			   EDP_PSR_DEBUG_MASK_HPD |
-			   EDP_PSR_DEBUG_MASK_LPSP);
+			   EDP_PSR_DEBUG_MASK_LPSP |
+			   EDP_PSR_DEBUG_MASK_DISP_REG_WRITE);
 	}
 }
 
-- 
cgit v1.1


From bc18b4df0fcb9fa4a50a19576723f14bfdd10c26 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jos=C3=A9=20Roberto=20de=20Souza?= <jose.souza@intel.com>
Date: Wed, 25 Apr 2018 14:23:32 -0700
Subject: drm/i915/psr/skl+: Print information about what caused a PSR exit
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This will be helpful to debug what hardware is actually tracking
and causing PSR to exit.

BSpec: 7721

v4:
- Using _MMIO_TRANS2() in PSR_EVENT
- Cleaning events before printing

Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Cc: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180425212334.21109-2-jose.souza@intel.com
---
 drivers/gpu/drm/i915/i915_reg.h  | 23 ++++++++++++++++++++
 drivers/gpu/drm/i915/intel_psr.c | 45 ++++++++++++++++++++++++++++++++++++++++
 2 files changed, 68 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 2dad655..391825a 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -4095,6 +4095,29 @@ enum {
 #define   EDP_PSR2_IDLE_FRAME_MASK	0xf
 #define   EDP_PSR2_IDLE_FRAME_SHIFT	0
 
+#define _PSR_EVENT_TRANS_A			0x60848
+#define _PSR_EVENT_TRANS_B			0x61848
+#define _PSR_EVENT_TRANS_C			0x62848
+#define _PSR_EVENT_TRANS_D			0x63848
+#define _PSR_EVENT_TRANS_EDP			0x6F848
+#define PSR_EVENT(trans)			_MMIO_TRANS2(trans, _PSR_EVENT_TRANS_A)
+#define  PSR_EVENT_PSR2_WD_TIMER_EXPIRE		(1 << 17)
+#define  PSR_EVENT_PSR2_DISABLED		(1 << 16)
+#define  PSR_EVENT_SU_DIRTY_FIFO_UNDERRUN	(1 << 15)
+#define  PSR_EVENT_SU_CRC_FIFO_UNDERRUN		(1 << 14)
+#define  PSR_EVENT_GRAPHICS_RESET		(1 << 12)
+#define  PSR_EVENT_PCH_INTERRUPT		(1 << 11)
+#define  PSR_EVENT_MEMORY_UP			(1 << 10)
+#define  PSR_EVENT_FRONT_BUFFER_MODIFY		(1 << 9)
+#define  PSR_EVENT_WD_TIMER_EXPIRE		(1 << 8)
+#define  PSR_EVENT_PIPE_REGISTERS_UPDATE	(1 << 6)
+#define  PSR_EVENT_REGISTER_UPDATE		(1 << 5)
+#define  PSR_EVENT_HDCP_ENABLE			(1 << 4)
+#define  PSR_EVENT_KVMR_SESSION_ENABLE		(1 << 3)
+#define  PSR_EVENT_VBI_ENABLE			(1 << 2)
+#define  PSR_EVENT_LPSP_MODE_EXIT		(1 << 1)
+#define  PSR_EVENT_PSR_DISABLE			(1 << 0)
+
 #define EDP_PSR2_STATUS			_MMIO(0x6f940)
 #define EDP_PSR2_STATUS_STATE_MASK     (0xf<<28)
 #define EDP_PSR2_STATUS_STATE_SHIFT    28
diff --git a/drivers/gpu/drm/i915/intel_psr.c b/drivers/gpu/drm/i915/intel_psr.c
index e35a3b9..c8d5cdc 100644
--- a/drivers/gpu/drm/i915/intel_psr.c
+++ b/drivers/gpu/drm/i915/intel_psr.c
@@ -125,6 +125,43 @@ void intel_psr_irq_control(struct drm_i915_private *dev_priv, bool debug)
 	I915_WRITE(EDP_PSR_IMR, ~mask);
 }
 
+static void psr_event_print(u32 val, bool psr2_enabled)
+{
+	DRM_DEBUG_KMS("PSR exit events: 0x%x\n", val);
+	if (val & PSR_EVENT_PSR2_WD_TIMER_EXPIRE)
+		DRM_DEBUG_KMS("\tPSR2 watchdog timer expired\n");
+	if ((val & PSR_EVENT_PSR2_DISABLED) && psr2_enabled)
+		DRM_DEBUG_KMS("\tPSR2 disabled\n");
+	if (val & PSR_EVENT_SU_DIRTY_FIFO_UNDERRUN)
+		DRM_DEBUG_KMS("\tSU dirty FIFO underrun\n");
+	if (val & PSR_EVENT_SU_CRC_FIFO_UNDERRUN)
+		DRM_DEBUG_KMS("\tSU CRC FIFO underrun\n");
+	if (val & PSR_EVENT_GRAPHICS_RESET)
+		DRM_DEBUG_KMS("\tGraphics reset\n");
+	if (val & PSR_EVENT_PCH_INTERRUPT)
+		DRM_DEBUG_KMS("\tPCH interrupt\n");
+	if (val & PSR_EVENT_MEMORY_UP)
+		DRM_DEBUG_KMS("\tMemory up\n");
+	if (val & PSR_EVENT_FRONT_BUFFER_MODIFY)
+		DRM_DEBUG_KMS("\tFront buffer modification\n");
+	if (val & PSR_EVENT_WD_TIMER_EXPIRE)
+		DRM_DEBUG_KMS("\tPSR watchdog timer expired\n");
+	if (val & PSR_EVENT_PIPE_REGISTERS_UPDATE)
+		DRM_DEBUG_KMS("\tPIPE registers updated\n");
+	if (val & PSR_EVENT_REGISTER_UPDATE)
+		DRM_DEBUG_KMS("\tRegister updated\n");
+	if (val & PSR_EVENT_HDCP_ENABLE)
+		DRM_DEBUG_KMS("\tHDCP enabled\n");
+	if (val & PSR_EVENT_KVMR_SESSION_ENABLE)
+		DRM_DEBUG_KMS("\tKVMR session enabled\n");
+	if (val & PSR_EVENT_VBI_ENABLE)
+		DRM_DEBUG_KMS("\tVBI enabled\n");
+	if (val & PSR_EVENT_LPSP_MODE_EXIT)
+		DRM_DEBUG_KMS("\tLPSP mode exited\n");
+	if ((val & PSR_EVENT_PSR_DISABLE) && !psr2_enabled)
+		DRM_DEBUG_KMS("\tPSR disabled\n");
+}
+
 void intel_psr_irq_handler(struct drm_i915_private *dev_priv, u32 psr_iir)
 {
 	u32 transcoders = BIT(TRANSCODER_EDP);
@@ -152,6 +189,14 @@ void intel_psr_irq_handler(struct drm_i915_private *dev_priv, u32 psr_iir)
 			dev_priv->psr.last_exit = time_ns;
 			DRM_DEBUG_KMS("[transcoder %s] PSR exit completed\n",
 				      transcoder_name(cpu_transcoder));
+
+			if (INTEL_GEN(dev_priv) >= 9) {
+				u32 val = I915_READ(PSR_EVENT(cpu_transcoder));
+				bool psr2_enabled = dev_priv->psr.psr2_enabled;
+
+				I915_WRITE(PSR_EVENT(cpu_transcoder), val);
+				psr_event_print(val, psr2_enabled);
+			}
 		}
 	}
 }
-- 
cgit v1.1


From d0bc86231a55d72860ac3a07a0f676fc18f2dd9d Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jos=C3=A9=20Roberto=20de=20Souza?= <jose.souza@intel.com>
Date: Wed, 25 Apr 2018 14:23:33 -0700
Subject: drm/i915/debugfs: Print sink PSR status
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

IGT tests could be improved with sink status, knowing for sure that
hardware have activate or exit PSR.

v3:
Reading i915_edp_psr_status was causing PSR to exit but now with
'drm/i915/psr: Prevent PSR exit when a non-pipe related register is
written' it is fixed.

Reviewed-by: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180425212334.21109-3-jose.souza@intel.com
---
 drivers/gpu/drm/i915/i915_debugfs.c | 29 +++++++++++++++++++++++++++++
 1 file changed, 29 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index 1c88805..cb1a804 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -2603,6 +2603,26 @@ static const char *psr2_live_status(u32 val)
 	return "unknown";
 }
 
+static const char *psr_sink_status(u8 val)
+{
+	static const char * const sink_status[] = {
+		"inactive",
+		"transition to active, capture and display",
+		"active, display from RFB",
+		"active, capture and display on sink device timings",
+		"transition to inactive, capture and display, timing re-sync",
+		"reserved",
+		"reserved",
+		"sink internal error"
+	};
+
+	val &= DP_PSR_SINK_STATE_MASK;
+	if (val < ARRAY_SIZE(sink_status))
+		return sink_status[val];
+
+	return "unknown";
+}
+
 static int i915_edp_psr_status(struct seq_file *m, void *data)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
@@ -2684,6 +2704,15 @@ static int i915_edp_psr_status(struct seq_file *m, void *data)
 		seq_printf(m, "EDP_PSR2_STATUS: %x [%s]\n",
 			   psr2, psr2_live_status(psr2));
 	}
+
+	if (dev_priv->psr.enabled) {
+		struct drm_dp_aux *aux = &dev_priv->psr.enabled->aux;
+		u8 val;
+
+		if (drm_dp_dpcd_readb(aux, DP_PSR_STATUS, &val) == 1)
+			seq_printf(m, "Sink PSR status: 0x%x [%s]\n", val,
+				   psr_sink_status(val));
+	}
 	mutex_unlock(&dev_priv->psr.lock);
 
 	if (READ_ONCE(dev_priv->psr.debug)) {
-- 
cgit v1.1


From 2a34b0054bf4a5224347e4135019b63d91cfb5d4 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jos=C3=A9=20Roberto=20de=20Souza?= <jose.souza@intel.com>
Date: Wed, 25 Apr 2018 14:23:34 -0700
Subject: drm/i915/psr/cnl: Set y-coordinate as valid in SDP
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This was my bad, spec says that the name of this bit is
'Y-coordinate valid' but the values for it is:
0: Include Y-coordinate valid eDP1.4a
1: Do not include Y-coordinate valid eDP 1.4
So not setting it.

BSpec: 7713

Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180425212334.21109-4-jose.souza@intel.com
---
 drivers/gpu/drm/i915/intel_psr.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_psr.c b/drivers/gpu/drm/i915/intel_psr.c
index c8d5cdc..6233a32 100644
--- a/drivers/gpu/drm/i915/intel_psr.c
+++ b/drivers/gpu/drm/i915/intel_psr.c
@@ -508,9 +508,8 @@ static void hsw_activate_psr2(struct intel_dp *intel_dp)
 	 * mesh at all with our frontbuffer tracking. And the hw alone isn't
 	 * good enough. */
 	val |= EDP_PSR2_ENABLE | EDP_SU_TRACK_ENABLE;
-	if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv)) {
-		val |= EDP_Y_COORDINATE_VALID | EDP_Y_COORDINATE_ENABLE;
-	}
+	if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv))
+		val |= EDP_Y_COORDINATE_ENABLE;
 
 	val |= EDP_PSR2_FRAME_BEFORE_SU(dev_priv->psr.sink_sync_latency + 1);
 
-- 
cgit v1.1


From f001cc351ad3309ec8736c374e90e5a4bc472d41 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig?= <u.kleine-koenig@pengutronix.de>
Date: Mon, 16 Apr 2018 13:17:53 +0200
Subject: gpio: fix error path in lineevent_create
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

If gpiod_request() fails the cleanup must not call gpiod_free().

Cc: stable@vger.kernel.org
Fixes: 61f922db7221 ("gpio: userspace ABI for reading GPIO line events")
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
---
 drivers/gpio/gpiolib.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpio/gpiolib.c b/drivers/gpio/gpiolib.c
index d077717..d8ccb50 100644
--- a/drivers/gpio/gpiolib.c
+++ b/drivers/gpio/gpiolib.c
@@ -903,7 +903,7 @@ static int lineevent_create(struct gpio_device *gdev, void __user *ip)
 	desc = &gdev->descs[offset];
 	ret = gpiod_request(desc, le->label);
 	if (ret)
-		goto out_free_desc;
+		goto out_free_label;
 	le->desc = desc;
 	le->eflags = eflags;
 
-- 
cgit v1.1


From aaf96e51de117cdfa2dc04735639895b46a3da3f Mon Sep 17 00:00:00 2001
From: William Breathitt Gray <vilhelm.gray@gmail.com>
Date: Wed, 18 Apr 2018 08:53:10 -0400
Subject: gpio: pci-idio-16: Fix port memory offset for get_multiple callback

The ioread8 function expects a memory offset argument. This patch fixes
the ports array to provide the memory addresses of the respective device
I/O registers.

Fixes: 810ebfc5efca ("gpio: pci-idio-16: Implement get_multiple callback")
Signed-off-by: William Breathitt Gray <vilhelm.gray@gmail.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
---
 drivers/gpio/gpio-pci-idio-16.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/gpio/gpio-pci-idio-16.c b/drivers/gpio/gpio-pci-idio-16.c
index 1948724..25d16b2 100644
--- a/drivers/gpio/gpio-pci-idio-16.c
+++ b/drivers/gpio/gpio-pci-idio-16.c
@@ -116,9 +116,9 @@ static int idio_16_gpio_get_multiple(struct gpio_chip *chip,
 	unsigned long word_mask;
 	const unsigned long port_mask = GENMASK(gpio_reg_size - 1, 0);
 	unsigned long port_state;
-	u8 __iomem ports[] = {
-		idio16gpio->reg->out0_7, idio16gpio->reg->out8_15,
-		idio16gpio->reg->in0_7, idio16gpio->reg->in8_15,
+	void __iomem *ports[] = {
+		&idio16gpio->reg->out0_7, &idio16gpio->reg->out8_15,
+		&idio16gpio->reg->in0_7, &idio16gpio->reg->in8_15,
 	};
 
 	/* clear bits array to a clean slate */
@@ -143,7 +143,7 @@ static int idio_16_gpio_get_multiple(struct gpio_chip *chip,
 		}
 
 		/* read bits from current gpio port */
-		port_state = ioread8(ports + i);
+		port_state = ioread8(ports[i]);
 
 		/* store acquired bits at respective bits array offset */
 		bits[word_index] |= port_state << word_offset;
-- 
cgit v1.1


From 304440aa96c6e5cc37eaa7a46ff4dd862e4c21be Mon Sep 17 00:00:00 2001
From: William Breathitt Gray <vilhelm.gray@gmail.com>
Date: Wed, 18 Apr 2018 08:53:21 -0400
Subject: gpio: pcie-idio-24: Fix port memory offset for
 get_multiple/set_multiple callbacks

The ioread8/iowrite8 functions expect a memory offset argument. This
patch fixes the ports array to provide the memory addresses of the
respective device I/O registers.

Fixes: ca37081595a2 ("gpio: pcie-idio-24: Implement get_multiple/set_multiple callbacks")
Signed-off-by: William Breathitt Gray <vilhelm.gray@gmail.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
---
 drivers/gpio/gpio-pcie-idio-24.c | 20 ++++++++++----------
 1 file changed, 10 insertions(+), 10 deletions(-)

diff --git a/drivers/gpio/gpio-pcie-idio-24.c b/drivers/gpio/gpio-pcie-idio-24.c
index 835607e..3e77c2a 100644
--- a/drivers/gpio/gpio-pcie-idio-24.c
+++ b/drivers/gpio/gpio-pcie-idio-24.c
@@ -206,10 +206,10 @@ static int idio_24_gpio_get_multiple(struct gpio_chip *chip,
 	unsigned long word_mask;
 	const unsigned long port_mask = GENMASK(gpio_reg_size - 1, 0);
 	unsigned long port_state;
-	u8 __iomem ports[] = {
-		idio24gpio->reg->out0_7, idio24gpio->reg->out8_15,
-		idio24gpio->reg->out16_23, idio24gpio->reg->in0_7,
-		idio24gpio->reg->in8_15, idio24gpio->reg->in16_23,
+	void __iomem *ports[] = {
+		&idio24gpio->reg->out0_7, &idio24gpio->reg->out8_15,
+		&idio24gpio->reg->out16_23, &idio24gpio->reg->in0_7,
+		&idio24gpio->reg->in8_15, &idio24gpio->reg->in16_23,
 	};
 	const unsigned long out_mode_mask = BIT(1);
 
@@ -236,7 +236,7 @@ static int idio_24_gpio_get_multiple(struct gpio_chip *chip,
 
 		/* read bits from current gpio port (port 6 is TTL GPIO) */
 		if (i < 6)
-			port_state = ioread8(ports + i);
+			port_state = ioread8(ports[i]);
 		else if (ioread8(&idio24gpio->reg->ctl) & out_mode_mask)
 			port_state = ioread8(&idio24gpio->reg->ttl_out0_7);
 		else
@@ -301,9 +301,9 @@ static void idio_24_gpio_set_multiple(struct gpio_chip *chip,
 	const unsigned long port_mask = GENMASK(gpio_reg_size, 0);
 	unsigned long flags;
 	unsigned int out_state;
-	u8 __iomem ports[] = {
-		idio24gpio->reg->out0_7, idio24gpio->reg->out8_15,
-		idio24gpio->reg->out16_23
+	void __iomem *ports[] = {
+		&idio24gpio->reg->out0_7, &idio24gpio->reg->out8_15,
+		&idio24gpio->reg->out16_23
 	};
 	const unsigned long out_mode_mask = BIT(1);
 	const unsigned int ttl_offset = 48;
@@ -327,9 +327,9 @@ static void idio_24_gpio_set_multiple(struct gpio_chip *chip,
 		raw_spin_lock_irqsave(&idio24gpio->lock, flags);
 
 		/* process output lines */
-		out_state = ioread8(ports + i) & ~gpio_mask;
+		out_state = ioread8(ports[i]) & ~gpio_mask;
 		out_state |= (*bits >> bits_offset) & gpio_mask;
-		iowrite8(out_state, ports + i);
+		iowrite8(out_state, ports[i]);
 
 		raw_spin_unlock_irqrestore(&idio24gpio->lock, flags);
 	}
-- 
cgit v1.1


From 608940dabe1bd2ce4c97524004ec86637cf80f2c Mon Sep 17 00:00:00 2001
From: Tom Zanussi <tom.zanussi@linux.intel.com>
Date: Thu, 26 Apr 2018 20:04:47 -0500
Subject: tracing: Restore proper field flag printing when displaying triggers

The flag-printing code used when displaying hist triggers somehow got
dropped during refactoring of the inter-event patchset.  This restores
it.

Below are a couple examples - in the first case, .usecs wasn't being
displayed properly for common_timestamps and the second illustrates
the same for other flags such as .execname.

Before:

  # echo 'hist:key=common_pid.execname:val=count:sort=count' > /sys/kernel/debug/tracing/events/syscalls/sys_enter_read/trigger
  # cat /sys/kernel/debug/tracing/events/syscalls/sys_enter_read/trigger
  hist:keys=common_pid:vals=hitcount,count:sort=count:size=2048 [active]

  # echo 'hist:keys=pid:ts0=common_timestamp.usecs if comm=="cyclictest"' >> /sys/kernel/debug/tracing/events/sched/sched_wakeup/trigger
  # cat /sys/kernel/debug/tracing/events/sched/sched_wakeup/trigger
  hist:keys=pid:vals=hitcount:ts0=common_timestamp:sort=hitcount:size=2048:clock=global if comm=="cyclictest" [active]

After:

  # echo 'hist:key=common_pid.execname:val=count:sort=count' > /sys/kernel/debug/tracing/events/syscalls/sys_enter_read/trigger
  # cat /sys/kernel/debug/tracing/events/syscalls/sys_enter_read/trigger
  hist:keys=common_pid.execname:vals=hitcount,count:sort=count:size=2048 [active]

  # echo 'hist:keys=pid:ts0=common_timestamp.usecs if comm=="cyclictest"' >> /sys/kernel/debug/tracing/events/sched/sched_wakeup/trigger
  # cat /sys/kernel/debug/tracing/events/sched/sched_wakeup/trigger
  hist:keys=pid:vals=hitcount:ts0=common_timestamp.usecs:sort=hitcount:size=2048:clock=global if comm=="cyclictest" [active]

Link: http://lkml.kernel.org/r/492bab42ff21806600af98a8ea901af10efbee0c.1524790601.git.tom.zanussi@linux.intel.com

Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
---
 kernel/trace/trace_events_hist.c | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/kernel/trace/trace_events_hist.c b/kernel/trace/trace_events_hist.c
index 0d7b3ff..66c87be 100644
--- a/kernel/trace/trace_events_hist.c
+++ b/kernel/trace/trace_events_hist.c
@@ -4913,6 +4913,16 @@ static void hist_field_print(struct seq_file *m, struct hist_field *hist_field)
 		seq_printf(m, "%s", field_name);
 	} else if (hist_field->flags & HIST_FIELD_FL_TIMESTAMP)
 		seq_puts(m, "common_timestamp");
+
+	if (hist_field->flags) {
+		if (!(hist_field->flags & HIST_FIELD_FL_VAR_REF) &&
+		    !(hist_field->flags & HIST_FIELD_FL_EXPR)) {
+			const char *flags = get_hist_field_flags(hist_field);
+
+			if (flags)
+				seq_printf(m, ".%s", flags);
+		}
+	}
 }
 
 static int event_hist_trigger_print(struct seq_file *m,
-- 
cgit v1.1


From 5ec432d7bf9dd3b4a2b84f8974e3adb71f45fb1d Mon Sep 17 00:00:00 2001
From: Tom Zanussi <tom.zanussi@linux.intel.com>
Date: Thu, 26 Apr 2018 20:04:48 -0500
Subject: tracing: Add field parsing hist error for hist triggers

If the user specifies a nonexistent field for a hist trigger, the
current code correctly flags that as an error, but doesn't tell the
user what happened.

Fix this by invoking hist_err() with an appropriate message when
nonexistent fields are specified.

Before:

  # echo 'hist:keys=pid:ts0=common_timestamp.usecs' >> /sys/kernel/debug/tracing/events/sched/sched_switch/trigger
  -su: echo: write error: Invalid argument
  # cat /sys/kernel/debug/tracing/events/sched/sched_switch/hist

After:

  # echo 'hist:keys=pid:ts0=common_timestamp.usecs' >> /sys/kernel/debug/tracing/events/sched/sched_switch/trigger
  -su: echo: write error: Invalid argument
  # cat /sys/kernel/debug/tracing/events/sched/sched_switch/hist
  ERROR: Couldn't find field: pid
    Last command: keys=pid:ts0=common_timestamp.usecs

Link: http://lkml.kernel.org/r/fdc8746969d16906120f162b99dd71c741e0b62c.1524790601.git.tom.zanussi@linux.intel.com

Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>
Reported-by: Masami Hiramatsu <mhiramat@kernel.org>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
---
 kernel/trace/trace_events_hist.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/kernel/trace/trace_events_hist.c b/kernel/trace/trace_events_hist.c
index 66c87be..f231fa2 100644
--- a/kernel/trace/trace_events_hist.c
+++ b/kernel/trace/trace_events_hist.c
@@ -2481,6 +2481,7 @@ parse_field(struct hist_trigger_data *hist_data, struct trace_event_file *file,
 	else {
 		field = trace_find_event_field(file->event_call, field_name);
 		if (!field || !field->size) {
+			hist_err("Couldn't find field: ", field_name);
 			field = ERR_PTR(-EINVAL);
 			goto out;
 		}
-- 
cgit v1.1


From dcf234577cd31fa16874e828b90659166ad6b80d Mon Sep 17 00:00:00 2001
From: Tom Zanussi <tom.zanussi@linux.intel.com>
Date: Thu, 26 Apr 2018 20:04:49 -0500
Subject: tracing: Add field modifier parsing hist error for hist triggers

If the user specifies an invalid field modifier for a hist trigger,
the current code correctly flags that as an error, but doesn't tell
the user what happened.

Fix this by invoking hist_err() with an appropriate message when
invalid modifiers are specified.

Before:

  # echo 'hist:keys=pid:ts0=common_timestamp.junkusecs' >> /sys/kernel/debug/tracing/events/sched/sched_wakeup/trigger
  -su: echo: write error: Invalid argument
  # cat /sys/kernel/debug/tracing/events/sched/sched_wakeup/hist

After:

  # echo 'hist:keys=pid:ts0=common_timestamp.junkusecs' >> /sys/kernel/debug/tracing/events/sched/sched_wakeup/trigger
  -su: echo: write error: Invalid argument
  # cat /sys/kernel/debug/tracing/events/sched/sched_wakeup/hist
  ERROR: Invalid field modifier: junkusecs
    Last command: keys=pid:ts0=common_timestamp.junkusecs

Link: http://lkml.kernel.org/r/b043c59fa79acd06a5f14a1d44dee9e5a3cd1248.1524790601.git.tom.zanussi@linux.intel.com

Signed-off-by: Tom Zanussi <tom.zanussi@linux.intel.com>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
---
 kernel/trace/trace_events_hist.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/kernel/trace/trace_events_hist.c b/kernel/trace/trace_events_hist.c
index f231fa2..b9061ed 100644
--- a/kernel/trace/trace_events_hist.c
+++ b/kernel/trace/trace_events_hist.c
@@ -2466,6 +2466,7 @@ parse_field(struct hist_trigger_data *hist_data, struct trace_event_file *file,
 		else if (strcmp(modifier, "usecs") == 0)
 			*flags |= HIST_FIELD_FL_TIMESTAMP_USECS;
 		else {
+			hist_err("Invalid field modifier: ", modifier);
 			field = ERR_PTR(-EINVAL);
 			goto out;
 		}
-- 
cgit v1.1


From 83ef4777f5ff3689e6e52d3913a13d79aa25f1b5 Mon Sep 17 00:00:00 2001
From: Jan Kiszka <jan.kiszka@siemens.com>
Date: Thu, 26 Apr 2018 13:00:30 +0200
Subject: of: overlay: Stop leaking resources on overlay removal

Only the overlay notifier callbacks have a chance to potentially get
hold of references to those two resources, but they are not supposed to
store them beyond OF_OVERLAY_POST_REMOVE.

Document the overlay notifier API, its constraint regarding pointer
lifetime, and then remove intentional leaks of ovcs->overlay_tree and
ovcs->fdt from free_overlay_changeset.

See also https://lkml.org/lkml/2018/4/23/1063 and following.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Reviewed-by: Frank Rowand <frowand.list@gmail.com>
Signed-off-by: Rob Herring <robh@kernel.org>
---
 Documentation/devicetree/overlay-notes.txt |  8 ++++++++
 drivers/of/overlay.c                       | 30 +++++++++++++++++++++---------
 2 files changed, 29 insertions(+), 9 deletions(-)

diff --git a/Documentation/devicetree/overlay-notes.txt b/Documentation/devicetree/overlay-notes.txt
index a4feb6d..725fb8d 100644
--- a/Documentation/devicetree/overlay-notes.txt
+++ b/Documentation/devicetree/overlay-notes.txt
@@ -98,6 +98,14 @@ Finally, if you need to remove all overlays in one-go, just call
 of_overlay_remove_all() which will remove every single one in the correct
 order.
 
+In addition, there is the option to register notifiers that get called on
+overlay operations. See of_overlay_notifier_register/unregister and
+enum of_overlay_notify_action for details.
+
+Note that a notifier callback is not supposed to store pointers to a device
+tree node or its content beyond OF_OVERLAY_POST_REMOVE corresponding to the
+respective node it received.
+
 Overlay DTS Format
 ------------------
 
diff --git a/drivers/of/overlay.c b/drivers/of/overlay.c
index b35fe88..7baa53e 100644
--- a/drivers/of/overlay.c
+++ b/drivers/of/overlay.c
@@ -102,12 +102,28 @@ static DEFINE_IDR(ovcs_idr);
 
 static BLOCKING_NOTIFIER_HEAD(overlay_notify_chain);
 
+/**
+ * of_overlay_notifier_register() - Register notifier for overlay operations
+ * @nb:		Notifier block to register
+ *
+ * Register for notification on overlay operations on device tree nodes. The
+ * reported actions definied by @of_reconfig_change. The notifier callback
+ * furthermore receives a pointer to the affected device tree node.
+ *
+ * Note that a notifier callback is not supposed to store pointers to a device
+ * tree node or its content beyond @OF_OVERLAY_POST_REMOVE corresponding to the
+ * respective node it received.
+ */
 int of_overlay_notifier_register(struct notifier_block *nb)
 {
 	return blocking_notifier_chain_register(&overlay_notify_chain, nb);
 }
 EXPORT_SYMBOL_GPL(of_overlay_notifier_register);
 
+/**
+ * of_overlay_notifier_register() - Unregister notifier for overlay operations
+ * @nb:		Notifier block to unregister
+ */
 int of_overlay_notifier_unregister(struct notifier_block *nb)
 {
 	return blocking_notifier_chain_unregister(&overlay_notify_chain, nb);
@@ -671,17 +687,13 @@ static void free_overlay_changeset(struct overlay_changeset *ovcs)
 		of_node_put(ovcs->fragments[i].overlay);
 	}
 	kfree(ovcs->fragments);
-
 	/*
-	 * TODO
-	 *
-	 * would like to: kfree(ovcs->overlay_tree);
-	 * but can not since drivers may have pointers into this data
-	 *
-	 * would like to: kfree(ovcs->fdt);
-	 * but can not since drivers may have pointers into this data
+	 * There should be no live pointers into ovcs->overlay_tree and
+	 * ovcs->fdt due to the policy that overlay notifiers are not allowed
+	 * to retain pointers into the overlay devicetree.
 	 */
-
+	kfree(ovcs->overlay_tree);
+	kfree(ovcs->fdt);
 	kfree(ovcs);
 }
 
-- 
cgit v1.1


From b4331a681822b420511b3258f1c3db35001fde48 Mon Sep 17 00:00:00 2001
From: Stefano Brivio <sbrivio@redhat.com>
Date: Thu, 26 Apr 2018 19:39:09 +0200
Subject: vti6: Change minimum MTU to IPV4_MIN_MTU, vti6 can carry IPv4 too

A vti6 interface can carry IPv4 as well, so it makes no sense to
enforce a minimum MTU of IPV6_MIN_MTU.

If the user sets an MTU below IPV6_MIN_MTU, IPv6 will be
disabled on the interface, courtesy of addrconf_notify().

Reported-by: Xin Long <lucien.xin@gmail.com>
Fixes: b96f9afee4eb ("ipv4/6: use core net MTU range checking")
Fixes: c6741fbed6dc ("vti6: Properly adjust vti6 MTU from MTU of lower device")
Fixes: 53c81e95df17 ("ip6_vti: adjust vti mtu according to mtu of lower device")
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
---
 net/ipv6/ip6_vti.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/net/ipv6/ip6_vti.c b/net/ipv6/ip6_vti.c
index c214ffe..ca957dd 100644
--- a/net/ipv6/ip6_vti.c
+++ b/net/ipv6/ip6_vti.c
@@ -669,7 +669,7 @@ static void vti6_link_config(struct ip6_tnl *t, bool keep_mtu)
 	else
 		mtu = ETH_DATA_LEN - LL_MAX_HEADER - sizeof(struct ipv6hdr);
 
-	dev->mtu = max_t(int, mtu, IPV6_MIN_MTU);
+	dev->mtu = max_t(int, mtu, IPV4_MIN_MTU);
 }
 
 /**
@@ -881,7 +881,7 @@ static void vti6_dev_setup(struct net_device *dev)
 	dev->priv_destructor = vti6_dev_free;
 
 	dev->type = ARPHRD_TUNNEL6;
-	dev->min_mtu = IPV6_MIN_MTU;
+	dev->min_mtu = IPV4_MIN_MTU;
 	dev->max_mtu = IP_MAX_MTU - sizeof(struct ipv6hdr);
 	dev->flags |= IFF_NOARP;
 	dev->addr_len = sizeof(struct in6_addr);
-- 
cgit v1.1


From ed769520727edbf526e0f510e6c501fc6ba3824a Mon Sep 17 00:00:00 2001
From: Joel Pepper <joel.pepper@rwth-aachen.de>
Date: Thu, 26 Apr 2018 20:26:08 +0200
Subject: usb: gadget: composite Allow for larger configuration descriptors

The composite framework allows us to create gadgets composed from many
different functions, which need to fit into a single configuration
descriptor.

Some functions (like uvc) can produce configuration descriptors upwards
of 2500 bytes on their own.

This patch increases the limit from 1024 bytes to 4096.

Signed-off-by: Joel Pepper <joel.pepper@rwth-aachen.de>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>
---
 include/linux/usb/composite.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/include/linux/usb/composite.h b/include/linux/usb/composite.h
index 4b6b928..8675e14 100644
--- a/include/linux/usb/composite.h
+++ b/include/linux/usb/composite.h
@@ -52,7 +52,7 @@
 #define USB_GADGET_DELAYED_STATUS       0x7fff	/* Impossibly large value */
 
 /* big enough to hold our biggest descriptor */
-#define USB_COMP_EP0_BUFSIZ	1024
+#define USB_COMP_EP0_BUFSIZ	4096
 
 /* OS feature descriptor length <= 4kB */
 #define USB_COMP_EP0_OS_DESC_BUFSIZ	4096
-- 
cgit v1.1


From f6d3e06f074721ad3a231df745d85b60428c1f03 Mon Sep 17 00:00:00 2001
From: Ian W MORRISON <ianwmorrison@gmail.com>
Date: Wed, 11 Apr 2018 14:42:13 +1000
Subject: drm/i915/glk: Add MODULE_FIRMWARE for Geminilake

As the Geminilake firmware is now merged to linux-firmware.git
use MODUE_FIRMWARE to load the firmware.

This removes the error message in the dmesg log:

    i915 0000:00:02.0: Direct firmware load for
        i915/glk_dmc_ver1_04.bin failed with error -2
    i915 0000:00:02.0: Failed to load DMC firmware
        i915/glk_dmc_ver1_04.bin. Disabling runtime power management.
    i915 0000:00:02.0: DMC firmware homepage:
        https://01.org/linuxgraphics/downloads/firmware

and now shows that the firmware has correctly loaded:

    [drm] Finished loading DMC firmware i915/glk_dmc_ver1_04.bin (v1.4)

Signed-off-by: Ian W MORRISON <ianwmorrison@gmail.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180411044213.383-1-ianwmorrison@gmail.com
---
 drivers/gpu/drm/i915/intel_csr.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/i915/intel_csr.c b/drivers/gpu/drm/i915/intel_csr.c
index 41e6c75..f9550ea 100644
--- a/drivers/gpu/drm/i915/intel_csr.c
+++ b/drivers/gpu/drm/i915/intel_csr.c
@@ -35,6 +35,7 @@
  */
 
 #define I915_CSR_GLK "i915/glk_dmc_ver1_04.bin"
+MODULE_FIRMWARE(I915_CSR_GLK);
 #define GLK_CSR_VERSION_REQUIRED	CSR_VERSION(1, 4)
 
 #define I915_CSR_CNL "i915/cnl_dmc_ver1_07.bin"
-- 
cgit v1.1


From 53692908b0f594285aba18ab848318262332ed25 Mon Sep 17 00:00:00 2001
From: Marc Zyngier <marc.zyngier@arm.com>
Date: Wed, 18 Apr 2018 10:39:04 +0100
Subject: KVM: arm/arm64: vgic: Fix source vcpu issues for GICv2 SGI

Now that we make sure we don't inject multiple instances of the
same GICv2 SGI at the same time, we've made another bug more
obvious:

If we exit with an active SGI, we completely lose track of which
vcpu it came from. On the next entry, we restore it with 0 as a
source, and if that wasn't the right one, too bad. While this
doesn't seem to trouble GIC-400, the architectural model gets
offended and doesn't deactivate the interrupt on EOI.

Another connected issue is that we will happilly make pending
an interrupt from another vcpu, overriding the above zero with
something that is just as inconsistent. Don't do that.

The final issue is that we signal a maintenance interrupt when
no pending interrupts are present in the LR. Assuming we've fixed
the two issues above, we end-up in a situation where we keep
exiting as soon as we've reached the active state, and not be
able to inject the following pending.

The fix comes in 3 parts:
- GICv2 SGIs have their source vcpu saved if they are active on
  exit, and restored on entry
- Multi-SGIs cannot go via the Pending+Active state, as this would
  corrupt the source field
- Multi-SGIs are converted to using MI on EOI instead of NPIE

Fixes: 16ca6a607d84bef0 ("KVM: arm/arm64: vgic: Don't populate multiple LRs with the same vintid")
Reported-by: Mark Rutland <mark.rutland@arm.com>
Tested-by: Mark Rutland <mark.rutland@arm.com>
Reviewed-by: Christoffer Dall <christoffer.dall@arm.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
 include/kvm/arm_vgic.h        |  1 +
 virt/kvm/arm/vgic/vgic-mmio.c | 10 +++++++--
 virt/kvm/arm/vgic/vgic-v2.c   | 38 +++++++++++++++++++--------------
 virt/kvm/arm/vgic/vgic-v3.c   | 49 +++++++++++++++++++++++++------------------
 virt/kvm/arm/vgic/vgic.c      | 30 +++++++-------------------
 virt/kvm/arm/vgic/vgic.h      | 14 +++++++++++++
 6 files changed, 81 insertions(+), 61 deletions(-)

diff --git a/include/kvm/arm_vgic.h b/include/kvm/arm_vgic.h
index 24f0394..e7efe12 100644
--- a/include/kvm/arm_vgic.h
+++ b/include/kvm/arm_vgic.h
@@ -131,6 +131,7 @@ struct vgic_irq {
 		u32 mpidr;			/* GICv3 target VCPU */
 	};
 	u8 source;			/* GICv2 SGIs only */
+	u8 active_source;		/* GICv2 SGIs only */
 	u8 priority;
 	enum vgic_irq_config config;	/* Level or edge */
 
diff --git a/virt/kvm/arm/vgic/vgic-mmio.c b/virt/kvm/arm/vgic/vgic-mmio.c
index dbe99d6..ff9655c 100644
--- a/virt/kvm/arm/vgic/vgic-mmio.c
+++ b/virt/kvm/arm/vgic/vgic-mmio.c
@@ -289,10 +289,16 @@ static void vgic_mmio_change_active(struct kvm_vcpu *vcpu, struct vgic_irq *irq,
 	       irq->vcpu->cpu != -1) /* VCPU thread is running */
 		cond_resched_lock(&irq->irq_lock);
 
-	if (irq->hw)
+	if (irq->hw) {
 		vgic_hw_irq_change_active(vcpu, irq, active, !requester_vcpu);
-	else
+	} else {
+		u32 model = vcpu->kvm->arch.vgic.vgic_model;
+
 		irq->active = active;
+		if (model == KVM_DEV_TYPE_ARM_VGIC_V2 &&
+		    active && vgic_irq_is_sgi(irq->intid))
+			irq->active_source = requester_vcpu->vcpu_id;
+	}
 
 	if (irq->active)
 		vgic_queue_irq_unlock(vcpu->kvm, irq, flags);
diff --git a/virt/kvm/arm/vgic/vgic-v2.c b/virt/kvm/arm/vgic/vgic-v2.c
index 45aa433..a5f2e44 100644
--- a/virt/kvm/arm/vgic/vgic-v2.c
+++ b/virt/kvm/arm/vgic/vgic-v2.c
@@ -37,13 +37,6 @@ void vgic_v2_init_lrs(void)
 		vgic_v2_write_lr(i, 0);
 }
 
-void vgic_v2_set_npie(struct kvm_vcpu *vcpu)
-{
-	struct vgic_v2_cpu_if *cpuif = &vcpu->arch.vgic_cpu.vgic_v2;
-
-	cpuif->vgic_hcr |= GICH_HCR_NPIE;
-}
-
 void vgic_v2_set_underflow(struct kvm_vcpu *vcpu)
 {
 	struct vgic_v2_cpu_if *cpuif = &vcpu->arch.vgic_cpu.vgic_v2;
@@ -71,13 +64,18 @@ void vgic_v2_fold_lr_state(struct kvm_vcpu *vcpu)
 	int lr;
 	unsigned long flags;
 
-	cpuif->vgic_hcr &= ~(GICH_HCR_UIE | GICH_HCR_NPIE);
+	cpuif->vgic_hcr &= ~GICH_HCR_UIE;
 
 	for (lr = 0; lr < vgic_cpu->used_lrs; lr++) {
 		u32 val = cpuif->vgic_lr[lr];
-		u32 intid = val & GICH_LR_VIRTUALID;
+		u32 cpuid, intid = val & GICH_LR_VIRTUALID;
 		struct vgic_irq *irq;
 
+		/* Extract the source vCPU id from the LR */
+		cpuid = val & GICH_LR_PHYSID_CPUID;
+		cpuid >>= GICH_LR_PHYSID_CPUID_SHIFT;
+		cpuid &= 7;
+
 		/* Notify fds when the guest EOI'ed a level-triggered SPI */
 		if (lr_signals_eoi_mi(val) && vgic_valid_spi(vcpu->kvm, intid))
 			kvm_notify_acked_irq(vcpu->kvm, 0,
@@ -90,17 +88,16 @@ void vgic_v2_fold_lr_state(struct kvm_vcpu *vcpu)
 		/* Always preserve the active bit */
 		irq->active = !!(val & GICH_LR_ACTIVE_BIT);
 
+		if (irq->active && vgic_irq_is_sgi(intid))
+			irq->active_source = cpuid;
+
 		/* Edge is the only case where we preserve the pending bit */
 		if (irq->config == VGIC_CONFIG_EDGE &&
 		    (val & GICH_LR_PENDING_BIT)) {
 			irq->pending_latch = true;
 
-			if (vgic_irq_is_sgi(intid)) {
-				u32 cpuid = val & GICH_LR_PHYSID_CPUID;
-
-				cpuid >>= GICH_LR_PHYSID_CPUID_SHIFT;
+			if (vgic_irq_is_sgi(intid))
 				irq->source |= (1 << cpuid);
-			}
 		}
 
 		/*
@@ -152,8 +149,15 @@ void vgic_v2_populate_lr(struct kvm_vcpu *vcpu, struct vgic_irq *irq, int lr)
 	u32 val = irq->intid;
 	bool allow_pending = true;
 
-	if (irq->active)
+	if (irq->active) {
 		val |= GICH_LR_ACTIVE_BIT;
+		if (vgic_irq_is_sgi(irq->intid))
+			val |= irq->active_source << GICH_LR_PHYSID_CPUID_SHIFT;
+		if (vgic_irq_is_multi_sgi(irq)) {
+			allow_pending = false;
+			val |= GICH_LR_EOI;
+		}
+	}
 
 	if (irq->hw) {
 		val |= GICH_LR_HW;
@@ -190,8 +194,10 @@ void vgic_v2_populate_lr(struct kvm_vcpu *vcpu, struct vgic_irq *irq, int lr)
 			BUG_ON(!src);
 			val |= (src - 1) << GICH_LR_PHYSID_CPUID_SHIFT;
 			irq->source &= ~(1 << (src - 1));
-			if (irq->source)
+			if (irq->source) {
 				irq->pending_latch = true;
+				val |= GICH_LR_EOI;
+			}
 		}
 	}
 
diff --git a/virt/kvm/arm/vgic/vgic-v3.c b/virt/kvm/arm/vgic/vgic-v3.c
index 8195f52..c7423f3 100644
--- a/virt/kvm/arm/vgic/vgic-v3.c
+++ b/virt/kvm/arm/vgic/vgic-v3.c
@@ -27,13 +27,6 @@ static bool group1_trap;
 static bool common_trap;
 static bool gicv4_enable;
 
-void vgic_v3_set_npie(struct kvm_vcpu *vcpu)
-{
-	struct vgic_v3_cpu_if *cpuif = &vcpu->arch.vgic_cpu.vgic_v3;
-
-	cpuif->vgic_hcr |= ICH_HCR_NPIE;
-}
-
 void vgic_v3_set_underflow(struct kvm_vcpu *vcpu)
 {
 	struct vgic_v3_cpu_if *cpuif = &vcpu->arch.vgic_cpu.vgic_v3;
@@ -55,17 +48,23 @@ void vgic_v3_fold_lr_state(struct kvm_vcpu *vcpu)
 	int lr;
 	unsigned long flags;
 
-	cpuif->vgic_hcr &= ~(ICH_HCR_UIE | ICH_HCR_NPIE);
+	cpuif->vgic_hcr &= ~ICH_HCR_UIE;
 
 	for (lr = 0; lr < vgic_cpu->used_lrs; lr++) {
 		u64 val = cpuif->vgic_lr[lr];
-		u32 intid;
+		u32 intid, cpuid;
 		struct vgic_irq *irq;
+		bool is_v2_sgi = false;
 
-		if (model == KVM_DEV_TYPE_ARM_VGIC_V3)
+		cpuid = val & GICH_LR_PHYSID_CPUID;
+		cpuid >>= GICH_LR_PHYSID_CPUID_SHIFT;
+
+		if (model == KVM_DEV_TYPE_ARM_VGIC_V3) {
 			intid = val & ICH_LR_VIRTUAL_ID_MASK;
-		else
+		} else {
 			intid = val & GICH_LR_VIRTUALID;
+			is_v2_sgi = vgic_irq_is_sgi(intid);
+		}
 
 		/* Notify fds when the guest EOI'ed a level-triggered IRQ */
 		if (lr_signals_eoi_mi(val) && vgic_valid_spi(vcpu->kvm, intid))
@@ -81,18 +80,16 @@ void vgic_v3_fold_lr_state(struct kvm_vcpu *vcpu)
 		/* Always preserve the active bit */
 		irq->active = !!(val & ICH_LR_ACTIVE_BIT);
 
+		if (irq->active && is_v2_sgi)
+			irq->active_source = cpuid;
+
 		/* Edge is the only case where we preserve the pending bit */
 		if (irq->config == VGIC_CONFIG_EDGE &&
 		    (val & ICH_LR_PENDING_BIT)) {
 			irq->pending_latch = true;
 
-			if (vgic_irq_is_sgi(intid) &&
-			    model == KVM_DEV_TYPE_ARM_VGIC_V2) {
-				u32 cpuid = val & GICH_LR_PHYSID_CPUID;
-
-				cpuid >>= GICH_LR_PHYSID_CPUID_SHIFT;
+			if (is_v2_sgi)
 				irq->source |= (1 << cpuid);
-			}
 		}
 
 		/*
@@ -133,10 +130,20 @@ void vgic_v3_populate_lr(struct kvm_vcpu *vcpu, struct vgic_irq *irq, int lr)
 {
 	u32 model = vcpu->kvm->arch.vgic.vgic_model;
 	u64 val = irq->intid;
-	bool allow_pending = true;
+	bool allow_pending = true, is_v2_sgi;
 
-	if (irq->active)
+	is_v2_sgi = (vgic_irq_is_sgi(irq->intid) &&
+		     model == KVM_DEV_TYPE_ARM_VGIC_V2);
+
+	if (irq->active) {
 		val |= ICH_LR_ACTIVE_BIT;
+		if (is_v2_sgi)
+			val |= irq->active_source << GICH_LR_PHYSID_CPUID_SHIFT;
+		if (vgic_irq_is_multi_sgi(irq)) {
+			allow_pending = false;
+			val |= ICH_LR_EOI;
+		}
+	}
 
 	if (irq->hw) {
 		val |= ICH_LR_HW;
@@ -174,8 +181,10 @@ void vgic_v3_populate_lr(struct kvm_vcpu *vcpu, struct vgic_irq *irq, int lr)
 			BUG_ON(!src);
 			val |= (src - 1) << GICH_LR_PHYSID_CPUID_SHIFT;
 			irq->source &= ~(1 << (src - 1));
-			if (irq->source)
+			if (irq->source) {
 				irq->pending_latch = true;
+				val |= ICH_LR_EOI;
+			}
 		}
 	}
 
diff --git a/virt/kvm/arm/vgic/vgic.c b/virt/kvm/arm/vgic/vgic.c
index 4b6d729..568c65f 100644
--- a/virt/kvm/arm/vgic/vgic.c
+++ b/virt/kvm/arm/vgic/vgic.c
@@ -719,14 +719,6 @@ static inline void vgic_set_underflow(struct kvm_vcpu *vcpu)
 		vgic_v3_set_underflow(vcpu);
 }
 
-static inline void vgic_set_npie(struct kvm_vcpu *vcpu)
-{
-	if (kvm_vgic_global_state.type == VGIC_V2)
-		vgic_v2_set_npie(vcpu);
-	else
-		vgic_v3_set_npie(vcpu);
-}
-
 /* Requires the ap_list_lock to be held. */
 static int compute_ap_list_depth(struct kvm_vcpu *vcpu,
 				 bool *multi_sgi)
@@ -740,17 +732,15 @@ static int compute_ap_list_depth(struct kvm_vcpu *vcpu,
 	DEBUG_SPINLOCK_BUG_ON(!spin_is_locked(&vgic_cpu->ap_list_lock));
 
 	list_for_each_entry(irq, &vgic_cpu->ap_list_head, ap_list) {
+		int w;
+
 		spin_lock(&irq->irq_lock);
 		/* GICv2 SGIs can count for more than one... */
-		if (vgic_irq_is_sgi(irq->intid) && irq->source) {
-			int w = hweight8(irq->source);
-
-			count += w;
-			*multi_sgi |= (w > 1);
-		} else {
-			count++;
-		}
+		w = vgic_irq_get_lr_count(irq);
 		spin_unlock(&irq->irq_lock);
+
+		count += w;
+		*multi_sgi |= (w > 1);
 	}
 	return count;
 }
@@ -761,7 +751,6 @@ static void vgic_flush_lr_state(struct kvm_vcpu *vcpu)
 	struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
 	struct vgic_irq *irq;
 	int count;
-	bool npie = false;
 	bool multi_sgi;
 	u8 prio = 0xff;
 
@@ -791,10 +780,8 @@ static void vgic_flush_lr_state(struct kvm_vcpu *vcpu)
 		if (likely(vgic_target_oracle(irq) == vcpu)) {
 			vgic_populate_lr(vcpu, irq, count++);
 
-			if (irq->source) {
-				npie = true;
+			if (irq->source)
 				prio = irq->priority;
-			}
 		}
 
 		spin_unlock(&irq->irq_lock);
@@ -807,9 +794,6 @@ static void vgic_flush_lr_state(struct kvm_vcpu *vcpu)
 		}
 	}
 
-	if (npie)
-		vgic_set_npie(vcpu);
-
 	vcpu->arch.vgic_cpu.used_lrs = count;
 
 	/* Nuke remaining LRs */
diff --git a/virt/kvm/arm/vgic/vgic.h b/virt/kvm/arm/vgic/vgic.h
index 830e815..32c25d4 100644
--- a/virt/kvm/arm/vgic/vgic.h
+++ b/virt/kvm/arm/vgic/vgic.h
@@ -110,6 +110,20 @@ static inline bool vgic_irq_is_mapped_level(struct vgic_irq *irq)
 	return irq->config == VGIC_CONFIG_LEVEL && irq->hw;
 }
 
+static inline int vgic_irq_get_lr_count(struct vgic_irq *irq)
+{
+	/* Account for the active state as an interrupt */
+	if (vgic_irq_is_sgi(irq->intid) && irq->source)
+		return hweight8(irq->source) + irq->active;
+
+	return irq_is_pending(irq) || irq->active;
+}
+
+static inline bool vgic_irq_is_multi_sgi(struct vgic_irq *irq)
+{
+	return vgic_irq_get_lr_count(irq) > 1;
+}
+
 /*
  * This struct provides an intermediate representation of the fields contained
  * in the GICH_VMCR and ICH_VMCR registers, such that code exporting the GIC
-- 
cgit v1.1


From b4678df184b314a2bd47d2329feca2c2534aa12b Mon Sep 17 00:00:00 2001
From: Matthew Wilcox <willy@infradead.org>
Date: Tue, 24 Apr 2018 14:02:57 -0700
Subject: errseq: Always report a writeback error once

The errseq_t infrastructure assumes that errors which occurred before
the file descriptor was opened are of no interest to the application.
This turns out to be a regression for some applications, notably Postgres.

Before errseq_t, a writeback error would be reported exactly once (as
long as the inode remained in memory), so Postgres could open a file,
call fsync() and find out whether there had been a writeback error on
that file from another process.

This patch changes the errseq infrastructure to report errors to all
file descriptors which are opened after the error occurred, but before
it was reported to any file descriptor.  This restores the user-visible
behaviour.

Cc: stable@vger.kernel.org
Fixes: 5660e13d2fd6 ("fs: new infrastructure for writeback error handling and reporting")
Signed-off-by: Matthew Wilcox <mawilcox@microsoft.com>
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
---
 lib/errseq.c | 23 +++++++++--------------
 1 file changed, 9 insertions(+), 14 deletions(-)

diff --git a/lib/errseq.c b/lib/errseq.c
index df78241..81f9e33 100644
--- a/lib/errseq.c
+++ b/lib/errseq.c
@@ -111,27 +111,22 @@ EXPORT_SYMBOL(errseq_set);
  * errseq_sample() - Grab current errseq_t value.
  * @eseq: Pointer to errseq_t to be sampled.
  *
- * This function allows callers to sample an errseq_t value, marking it as
- * "seen" if required.
+ * This function allows callers to initialise their errseq_t variable.
+ * If the error has been "seen", new callers will not see an old error.
+ * If there is an unseen error in @eseq, the caller of this function will
+ * see it the next time it checks for an error.
  *
+ * Context: Any context.
  * Return: The current errseq value.
  */
 errseq_t errseq_sample(errseq_t *eseq)
 {
 	errseq_t old = READ_ONCE(*eseq);
-	errseq_t new = old;
 
-	/*
-	 * For the common case of no errors ever having been set, we can skip
-	 * marking the SEEN bit. Once an error has been set, the value will
-	 * never go back to zero.
-	 */
-	if (old != 0) {
-		new |= ERRSEQ_SEEN;
-		if (old != new)
-			cmpxchg(eseq, old, new);
-	}
-	return new;
+	/* If nobody has seen this error yet, then we can be the first. */
+	if (!(old & ERRSEQ_SEEN))
+		old = 0;
+	return old;
 }
 EXPORT_SYMBOL(errseq_sample);
 
-- 
cgit v1.1


From 88fc6f73fddf64eb507b04f7b2bd01d7291db514 Mon Sep 17 00:00:00 2001
From: Marek Szyprowski <m.szyprowski@samsung.com>
Date: Mon, 16 Apr 2018 12:11:52 +0200
Subject: thermal: exynos: Reading temperature makes sense only when TMU is
 turned on

When thermal sensor is not yet enabled, reading temperature might return
random value. This might even result in stopping system booting when such
temperature is higher than the critical value. Fix this by checking if TMU
has been actually enabled before reading the temperature.

This change fixes booting of Exynos4210-based board with TMU enabled (for
example Samsung Trats board), which was broken since v4.4 kernel release.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Fixes: 9e4249b40340 ("thermal: exynos: Fix first temperature read after registering sensor")
CC: stable@vger.kernel.org # v4.6+
Signed-off-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
---
 drivers/thermal/samsung/exynos_tmu.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/thermal/samsung/exynos_tmu.c b/drivers/thermal/samsung/exynos_tmu.c
index ed805c7..986cbd0 100644
--- a/drivers/thermal/samsung/exynos_tmu.c
+++ b/drivers/thermal/samsung/exynos_tmu.c
@@ -185,6 +185,7 @@
  * @regulator: pointer to the TMU regulator structure.
  * @reg_conf: pointer to structure to register with core thermal.
  * @ntrip: number of supported trip points.
+ * @enabled: current status of TMU device
  * @tmu_initialize: SoC specific TMU initialization method
  * @tmu_control: SoC specific TMU control method
  * @tmu_read: SoC specific TMU temperature read method
@@ -205,6 +206,7 @@ struct exynos_tmu_data {
 	struct regulator *regulator;
 	struct thermal_zone_device *tzd;
 	unsigned int ntrip;
+	bool enabled;
 
 	int (*tmu_initialize)(struct platform_device *pdev);
 	void (*tmu_control)(struct platform_device *pdev, bool on);
@@ -398,6 +400,7 @@ static void exynos_tmu_control(struct platform_device *pdev, bool on)
 	mutex_lock(&data->lock);
 	clk_enable(data->clk);
 	data->tmu_control(pdev, on);
+	data->enabled = on;
 	clk_disable(data->clk);
 	mutex_unlock(&data->lock);
 }
@@ -890,7 +893,7 @@ static int exynos_get_temp(void *p, int *temp)
 {
 	struct exynos_tmu_data *data = p;
 
-	if (!data || !data->tmu_read)
+	if (!data || !data->tmu_read || !data->enabled)
 		return -EINVAL;
 
 	mutex_lock(&data->lock);
-- 
cgit v1.1


From c8da6cdef57b459ac0fd5d9d348f8460a575ae90 Mon Sep 17 00:00:00 2001
From: Marek Szyprowski <m.szyprowski@samsung.com>
Date: Mon, 16 Apr 2018 12:11:53 +0200
Subject: thermal: exynos: Propagate error value from tmu_read()

tmu_read() in case of Exynos4210 might return error for out of bound
values. Current code ignores such value, what leads to reporting critical
temperature value. Add proper error code propagation to exynos_get_temp()
function.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
CC: stable@vger.kernel.org # v4.6+
Signed-off-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Signed-off-by: Eduardo Valentin <edubezval@gmail.com>
---
 drivers/thermal/samsung/exynos_tmu.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/drivers/thermal/samsung/exynos_tmu.c b/drivers/thermal/samsung/exynos_tmu.c
index 986cbd0..ac83f72 100644
--- a/drivers/thermal/samsung/exynos_tmu.c
+++ b/drivers/thermal/samsung/exynos_tmu.c
@@ -892,6 +892,7 @@ static void exynos7_tmu_control(struct platform_device *pdev, bool on)
 static int exynos_get_temp(void *p, int *temp)
 {
 	struct exynos_tmu_data *data = p;
+	int value, ret = 0;
 
 	if (!data || !data->tmu_read || !data->enabled)
 		return -EINVAL;
@@ -899,12 +900,16 @@ static int exynos_get_temp(void *p, int *temp)
 	mutex_lock(&data->lock);
 	clk_enable(data->clk);
 
-	*temp = code_to_temp(data, data->tmu_read(data)) * MCELSIUS;
+	value = data->tmu_read(data);
+	if (value < 0)
+		ret = value;
+	else
+		*temp = code_to_temp(data, value) * MCELSIUS;
 
 	clk_disable(data->clk);
 	mutex_unlock(&data->lock);
 
-	return 0;
+	return ret;
 }
 
 #ifdef CONFIG_THERMAL_EMULATION
-- 
cgit v1.1


From b4bd701ac469075d94ed9699a28755f2862252b9 Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leonro@mellanox.com>
Date: Mon, 23 Apr 2018 17:01:52 +0300
Subject: RDMA/mlx5: Fix multiple NULL-ptr deref errors in rereg_mr flow

Failure in rereg MR releases UMEM but leaves the MR to be destroyed
by the user. As a result the following scenario may happen:
"create MR -> rereg MR with failure -> call to rereg MR again" and
hit "NULL-ptr deref or user memory access" errors.

Ensure that rereg MR is only performed on a non-dead MR.

Cc: syzkaller <syzkaller@googlegroups.com>
Cc: <stable@vger.kernel.org> # 4.5
Fixes: 395a8e4c32ea ("IB/mlx5: Refactoring register MR code")
Reported-by: Noa Osherovich <noaos@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/mr.c | 32 +++++++++++++++++++++++---------
 1 file changed, 23 insertions(+), 9 deletions(-)

diff --git a/drivers/infiniband/hw/mlx5/mr.c b/drivers/infiniband/hw/mlx5/mr.c
index 1520a2f..90a9c46 100644
--- a/drivers/infiniband/hw/mlx5/mr.c
+++ b/drivers/infiniband/hw/mlx5/mr.c
@@ -866,25 +866,28 @@ static int mr_umem_get(struct ib_pd *pd, u64 start, u64 length,
 		       int *order)
 {
 	struct mlx5_ib_dev *dev = to_mdev(pd->device);
+	struct ib_umem *u;
 	int err;
 
-	*umem = ib_umem_get(pd->uobject->context, start, length,
-			    access_flags, 0);
-	err = PTR_ERR_OR_ZERO(*umem);
+	*umem = NULL;
+
+	u = ib_umem_get(pd->uobject->context, start, length, access_flags, 0);
+	err = PTR_ERR_OR_ZERO(u);
 	if (err) {
-		*umem = NULL;
-		mlx5_ib_err(dev, "umem get failed (%d)\n", err);
+		mlx5_ib_dbg(dev, "umem get failed (%d)\n", err);
 		return err;
 	}
 
-	mlx5_ib_cont_pages(*umem, start, MLX5_MKEY_PAGE_SHIFT_MASK, npages,
+	mlx5_ib_cont_pages(u, start, MLX5_MKEY_PAGE_SHIFT_MASK, npages,
 			   page_shift, ncont, order);
 	if (!*npages) {
 		mlx5_ib_warn(dev, "avoid zero region\n");
-		ib_umem_release(*umem);
+		ib_umem_release(u);
 		return -EINVAL;
 	}
 
+	*umem = u;
+
 	mlx5_ib_dbg(dev, "npages %d, ncont %d, order %d, page_shift %d\n",
 		    *npages, *ncont, *order, *page_shift);
 
@@ -1458,13 +1461,12 @@ int mlx5_ib_rereg_user_mr(struct ib_mr *ib_mr, int flags, u64 start,
 	int access_flags = flags & IB_MR_REREG_ACCESS ?
 			    new_access_flags :
 			    mr->access_flags;
-	u64 addr = (flags & IB_MR_REREG_TRANS) ? virt_addr : mr->umem->address;
-	u64 len = (flags & IB_MR_REREG_TRANS) ? length : mr->umem->length;
 	int page_shift = 0;
 	int upd_flags = 0;
 	int npages = 0;
 	int ncont = 0;
 	int order = 0;
+	u64 addr, len;
 	int err;
 
 	mlx5_ib_dbg(dev, "start 0x%llx, virt_addr 0x%llx, length 0x%llx, access_flags 0x%x\n",
@@ -1472,6 +1474,17 @@ int mlx5_ib_rereg_user_mr(struct ib_mr *ib_mr, int flags, u64 start,
 
 	atomic_sub(mr->npages, &dev->mdev->priv.reg_pages);
 
+	if (!mr->umem)
+		return -EINVAL;
+
+	if (flags & IB_MR_REREG_TRANS) {
+		addr = virt_addr;
+		len = length;
+	} else {
+		addr = mr->umem->address;
+		len = mr->umem->length;
+	}
+
 	if (flags != IB_MR_REREG_PD) {
 		/*
 		 * Replace umem. This needs to be done whether or not UMR is
@@ -1479,6 +1492,7 @@ int mlx5_ib_rereg_user_mr(struct ib_mr *ib_mr, int flags, u64 start,
 		 */
 		flags |= IB_MR_REREG_TRANS;
 		ib_umem_release(mr->umem);
+		mr->umem = NULL;
 		err = mr_umem_get(pd, addr, len, access_flags, &mr->umem,
 				  &npages, &page_shift, &ncont, &order);
 		if (err)
-- 
cgit v1.1


From 002bf2282b2d7318e444dca9ffcb994afc5d5f15 Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leonro@mellanox.com>
Date: Mon, 23 Apr 2018 17:01:53 +0300
Subject: RDMA/mlx5: Protect from shift operand overflow

Ensure that user didn't supply values too large that can cause overflow.

UBSAN: Undefined behaviour in drivers/infiniband/hw/mlx5/qp.c:263:23
shift exponent -2147483648 is negative
CPU: 0 PID: 292 Comm: syzkaller612609 Not tainted 4.16.0-rc1+ #131
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.11.0-0-g63451fca13-prebuilt.qemu-project.org 04/01/2014 Call
Trace:
dump_stack+0xde/0x164
ubsan_epilogue+0xe/0x81
set_rq_size+0x7c2/0xa90
create_qp_common+0xc18/0x43c0
mlx5_ib_create_qp+0x379/0x1ca0
create_qp.isra.5+0xc94/0x2260
ib_uverbs_create_qp+0x21b/0x2a0
ib_uverbs_write+0xc2c/0x1010
vfs_write+0x1b0/0x550
SyS_write+0xc7/0x1a0
do_syscall_64+0x1aa/0x740
entry_SYSCALL_64_after_hwframe+0x26/0x9b
RIP: 0033:0x433569
RSP: 002b:00007ffc6e62f448 EFLAGS: 00000217 ORIG_RAX: 0000000000000001
RAX: ffffffffffffffda RBX: 00000000004002f8 RCX: 0000000000433569
RDX: 0000000000000070 RSI: 00000000200042c0 RDI: 0000000000000003
RBP: 00000000006d5018 R08: 00000000004002f8 R09: 00000000004002f8
R10: 00000000004002f8 R11: 0000000000000217 R12: 0000000000000000
R13: 000000000040c9f0 R14: 000000000040ca80 R15: 0000000000000006

Cc: <stable@vger.kernel.org> # 3.10
Fixes: e126ba97dba9 ("mlx5: Add driver for Mellanox Connect-IB adapters")
Cc: syzkaller <syzkaller@googlegroups.com>
Reported-by: Noa Osherovich <noaos@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/qp.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c
index 7ed4b70..e6219a5 100644
--- a/drivers/infiniband/hw/mlx5/qp.c
+++ b/drivers/infiniband/hw/mlx5/qp.c
@@ -259,7 +259,11 @@ static int set_rq_size(struct mlx5_ib_dev *dev, struct ib_qp_cap *cap,
 	} else {
 		if (ucmd) {
 			qp->rq.wqe_cnt = ucmd->rq_wqe_count;
+			if (ucmd->rq_wqe_shift > BITS_PER_BYTE * sizeof(ucmd->rq_wqe_shift))
+				return -EINVAL;
 			qp->rq.wqe_shift = ucmd->rq_wqe_shift;
+			if ((1 << qp->rq.wqe_shift) / sizeof(struct mlx5_wqe_data_seg) < qp->wq_sig)
+				return -EINVAL;
 			qp->rq.max_gs = (1 << qp->rq.wqe_shift) / sizeof(struct mlx5_wqe_data_seg) - qp->wq_sig;
 			qp->rq.max_post = qp->rq.wqe_cnt;
 		} else {
-- 
cgit v1.1


From 4f32ac2e452c2180cd2df581cbadac183e27ecd0 Mon Sep 17 00:00:00 2001
From: Danit Goldberg <danitg@mellanox.com>
Date: Mon, 23 Apr 2018 17:01:54 +0300
Subject: IB/mlx5: Use unlimited rate when static rate is not supported

Before the change, if the user passed a static rate value different
than zero and the FW doesn't support static rate,
it would end up configuring rate of 2.5 GBps.

Fix this by using rate 0; unlimited, in cases where FW
doesn't support static rate configuration.

Cc: <stable@vger.kernel.org> # 3.10
Fixes: e126ba97dba9 ("mlx5: Add driver for Mellanox Connect-IB adapters")
Reviewed-by: Majd Dibbiny <majd@mellanox.com>
Signed-off-by: Danit Goldberg <danitg@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/qp.c | 18 +++++++++---------
 1 file changed, 9 insertions(+), 9 deletions(-)

diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c
index e6219a5..87b7c1b 100644
--- a/drivers/infiniband/hw/mlx5/qp.c
+++ b/drivers/infiniband/hw/mlx5/qp.c
@@ -2455,18 +2455,18 @@ enum {
 
 static int ib_rate_to_mlx5(struct mlx5_ib_dev *dev, u8 rate)
 {
-	if (rate == IB_RATE_PORT_CURRENT) {
+	if (rate == IB_RATE_PORT_CURRENT)
 		return 0;
-	} else if (rate < IB_RATE_2_5_GBPS || rate > IB_RATE_300_GBPS) {
+
+	if (rate < IB_RATE_2_5_GBPS || rate > IB_RATE_300_GBPS)
 		return -EINVAL;
-	} else {
-		while (rate != IB_RATE_2_5_GBPS &&
-		       !(1 << (rate + MLX5_STAT_RATE_OFFSET) &
-			 MLX5_CAP_GEN(dev->mdev, stat_rate_support)))
-			--rate;
-	}
 
-	return rate + MLX5_STAT_RATE_OFFSET;
+	while (rate != IB_RATE_PORT_CURRENT &&
+	       !(1 << (rate + MLX5_STAT_RATE_OFFSET) &
+		 MLX5_CAP_GEN(dev->mdev, stat_rate_support)))
+		--rate;
+
+	return rate ? rate + MLX5_STAT_RATE_OFFSET : rate;
 }
 
 static int modify_raw_packet_eth_prio(struct mlx5_core_dev *dev,
-- 
cgit v1.1


From 84a6a7a99c0ac2f67366288c0625c9fba176b264 Mon Sep 17 00:00:00 2001
From: Parav Pandit <parav@mellanox.com>
Date: Mon, 23 Apr 2018 17:01:55 +0300
Subject: IB/mlx5: Fix represent correct netdevice in dual port RoCE

In commit bcf87f1dbbec ("IB/mlx5: Listen to netdev register/unresiter events in switchdev mode")
incorrectly mapped primary device's netdevice to 2nd port netdevice.
It always represented primary port's netdevice for 2nd port netdevice
when ib representors were not used.

This results into failing to process CM request arriving on 2nd port due
to incorrect mapping of netdevice.

This fix corrects it by considering the right mdev.

Cc: <stable@vger.kernel.org> # 4.16
Fixes: bcf87f1dbbec ("IB/mlx5: Listen to netdev register/unresiter events in switchdev mode")
Reviewed-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Parav Pandit <parav@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/main.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index 6a749c0..78a4b27 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -179,7 +179,7 @@ static int mlx5_netdev_event(struct notifier_block *this,
 			if (rep_ndev == ndev)
 				roce->netdev = (event == NETDEV_UNREGISTER) ?
 					NULL : ndev;
-		} else if (ndev->dev.parent == &ibdev->mdev->pdev->dev) {
+		} else if (ndev->dev.parent == &mdev->pdev->dev) {
 			roce->netdev = (event == NETDEV_UNREGISTER) ?
 				NULL : ndev;
 		}
-- 
cgit v1.1


From 444261ca6ff201fa03de97a5041237e67a9d8d31 Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leonro@mellanox.com>
Date: Mon, 23 Apr 2018 17:01:56 +0300
Subject: RDMA/mlx5: Properly check return value of mlx5_get_uars_page

Starting from commit 72f36be06138 ("net/mlx5: Fix mlx5_get_uars_page to
return error code") the mlx5_get_uars_page() call returns error in case
of failure, but it was mistakenly overlooked in the merge commit.

Fixes: e7996a9a77fc ("Merge tag v4.15 of git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git")
Reported-by: Alaa Hleihel <alaa@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/main.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index 78a4b27..a42c6b1 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -5426,9 +5426,7 @@ static void mlx5_ib_stage_cong_debugfs_cleanup(struct mlx5_ib_dev *dev)
 static int mlx5_ib_stage_uar_init(struct mlx5_ib_dev *dev)
 {
 	dev->mdev->priv.uar = mlx5_get_uars_page(dev->mdev);
-	if (!dev->mdev->priv.uar)
-		return -ENOMEM;
-	return 0;
+	return PTR_ERR_OR_ZERO(dev->mdev->priv.uar);
 }
 
 static void mlx5_ib_stage_uar_cleanup(struct mlx5_ib_dev *dev)
-- 
cgit v1.1


From 7dbc73e6124ce4d0cfbdd6166de388e9367c47ad Mon Sep 17 00:00:00 2001
From: Jon Maloy <jon.maloy@ericsson.com>
Date: Wed, 25 Apr 2018 18:29:25 +0200
Subject: tipc: fix bug in function tipc_nl_node_dump_monitor

Commit 36a50a989ee8 ("tipc: fix infinite loop when dumping link monitor
summary") intended to fix a problem with user tool looping when max
number of bearers are enabled.

Unfortunately, the wrong version of the commit was posted, so the
problem was not solved at all.

This commit adds the missing part.

Fixes: 36a50a989ee8 ("tipc: fix infinite loop when dumping link monitor summary")
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/tipc/node.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/tipc/node.c b/net/tipc/node.c
index 6f98b56..baaf93f 100644
--- a/net/tipc/node.c
+++ b/net/tipc/node.c
@@ -2244,7 +2244,7 @@ int tipc_nl_node_dump_monitor(struct sk_buff *skb, struct netlink_callback *cb)
 
 	rtnl_lock();
 	for (bearer_id = prev_bearer; bearer_id < MAX_BEARERS; bearer_id++) {
-		err = __tipc_nl_add_monitor(net, &msg, prev_bearer);
+		err = __tipc_nl_add_monitor(net, &msg, bearer_id);
 		if (err)
 			break;
 	}
-- 
cgit v1.1


From c55ca688ed99a9cb79367aee2ed2ff6cb80fc039 Mon Sep 17 00:00:00 2001
From: Jakub Kicinski <jakub.kicinski@netronome.com>
Date: Wed, 25 Apr 2018 11:21:08 -0700
Subject: nfp: don't depend on eth_tbl being available

For very very old generation of the management FW Ethernet port
information table may theoretically not be available.  This in
turn will cause the nfp_port structures to not be allocated.

Make sure we don't crash the kernel when there is no eth_tbl:

RIP: 0010:nfp_net_pci_probe+0xf2/0xb40 [nfp]
...
Call Trace:
  nfp_pci_probe+0x6de/0xab0 [nfp]
  local_pci_probe+0x47/0xa0
  work_for_cpu_fn+0x1a/0x30
  process_one_work+0x1de/0x3e0

Found while working with broken/development version of management FW.

Fixes: a5950182c00e ("nfp: map mac_stats and vf_cfg BARs")
Fixes: 93da7d9660ee ("nfp: provide nfp_port to of nfp_net_get_mac_addr()")
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Dirk van der Merwe <dirk.vandermerwe@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/netronome/nfp/flower/main.c  |  2 +-
 drivers/net/ethernet/netronome/nfp/nfp_app_nic.c  |  2 +-
 drivers/net/ethernet/netronome/nfp/nfp_main.h     |  4 ++-
 drivers/net/ethernet/netronome/nfp/nfp_net_main.c | 31 +++++++++++++----------
 4 files changed, 23 insertions(+), 16 deletions(-)

diff --git a/drivers/net/ethernet/netronome/nfp/flower/main.c b/drivers/net/ethernet/netronome/nfp/flower/main.c
index ad02592..a997e34 100644
--- a/drivers/net/ethernet/netronome/nfp/flower/main.c
+++ b/drivers/net/ethernet/netronome/nfp/flower/main.c
@@ -360,7 +360,7 @@ nfp_flower_spawn_phy_reprs(struct nfp_app *app, struct nfp_flower_priv *priv)
 		}
 
 		SET_NETDEV_DEV(repr, &priv->nn->pdev->dev);
-		nfp_net_get_mac_addr(app->pf, port);
+		nfp_net_get_mac_addr(app->pf, repr, port);
 
 		cmsg_port_id = nfp_flower_cmsg_phys_port(phys_port);
 		err = nfp_repr_init(app, repr,
diff --git a/drivers/net/ethernet/netronome/nfp/nfp_app_nic.c b/drivers/net/ethernet/netronome/nfp/nfp_app_nic.c
index 2a2f2fb..b9618c3 100644
--- a/drivers/net/ethernet/netronome/nfp/nfp_app_nic.c
+++ b/drivers/net/ethernet/netronome/nfp/nfp_app_nic.c
@@ -69,7 +69,7 @@ int nfp_app_nic_vnic_alloc(struct nfp_app *app, struct nfp_net *nn,
 	if (err)
 		return err < 0 ? err : 0;
 
-	nfp_net_get_mac_addr(app->pf, nn->port);
+	nfp_net_get_mac_addr(app->pf, nn->dp.netdev, nn->port);
 
 	return 0;
 }
diff --git a/drivers/net/ethernet/netronome/nfp/nfp_main.h b/drivers/net/ethernet/netronome/nfp/nfp_main.h
index add46e2..4221108 100644
--- a/drivers/net/ethernet/netronome/nfp/nfp_main.h
+++ b/drivers/net/ethernet/netronome/nfp/nfp_main.h
@@ -171,7 +171,9 @@ void nfp_net_pci_remove(struct nfp_pf *pf);
 int nfp_hwmon_register(struct nfp_pf *pf);
 void nfp_hwmon_unregister(struct nfp_pf *pf);
 
-void nfp_net_get_mac_addr(struct nfp_pf *pf, struct nfp_port *port);
+void
+nfp_net_get_mac_addr(struct nfp_pf *pf, struct net_device *netdev,
+		     struct nfp_port *port);
 
 bool nfp_ctrl_tx(struct nfp_net *nn, struct sk_buff *skb);
 
diff --git a/drivers/net/ethernet/netronome/nfp/nfp_net_main.c b/drivers/net/ethernet/netronome/nfp/nfp_net_main.c
index 15fa47f..45cd209 100644
--- a/drivers/net/ethernet/netronome/nfp/nfp_net_main.c
+++ b/drivers/net/ethernet/netronome/nfp/nfp_net_main.c
@@ -67,23 +67,26 @@
 /**
  * nfp_net_get_mac_addr() - Get the MAC address.
  * @pf:       NFP PF handle
+ * @netdev:   net_device to set MAC address on
  * @port:     NFP port structure
  *
  * First try to get the MAC address from NSP ETH table. If that
  * fails generate a random address.
  */
-void nfp_net_get_mac_addr(struct nfp_pf *pf, struct nfp_port *port)
+void
+nfp_net_get_mac_addr(struct nfp_pf *pf, struct net_device *netdev,
+		     struct nfp_port *port)
 {
 	struct nfp_eth_table_port *eth_port;
 
 	eth_port = __nfp_port_get_eth_port(port);
 	if (!eth_port) {
-		eth_hw_addr_random(port->netdev);
+		eth_hw_addr_random(netdev);
 		return;
 	}
 
-	ether_addr_copy(port->netdev->dev_addr, eth_port->mac_addr);
-	ether_addr_copy(port->netdev->perm_addr, eth_port->mac_addr);
+	ether_addr_copy(netdev->dev_addr, eth_port->mac_addr);
+	ether_addr_copy(netdev->perm_addr, eth_port->mac_addr);
 }
 
 static struct nfp_eth_table_port *
@@ -511,16 +514,18 @@ static int nfp_net_pci_map_mem(struct nfp_pf *pf)
 		return PTR_ERR(mem);
 	}
 
-	min_size =  NFP_MAC_STATS_SIZE * (pf->eth_tbl->max_index + 1);
-	pf->mac_stats_mem = nfp_rtsym_map(pf->rtbl, "_mac_stats",
-					  "net.macstats", min_size,
-					  &pf->mac_stats_bar);
-	if (IS_ERR(pf->mac_stats_mem)) {
-		if (PTR_ERR(pf->mac_stats_mem) != -ENOENT) {
-			err = PTR_ERR(pf->mac_stats_mem);
-			goto err_unmap_ctrl;
+	if (pf->eth_tbl) {
+		min_size =  NFP_MAC_STATS_SIZE * (pf->eth_tbl->max_index + 1);
+		pf->mac_stats_mem = nfp_rtsym_map(pf->rtbl, "_mac_stats",
+						  "net.macstats", min_size,
+						  &pf->mac_stats_bar);
+		if (IS_ERR(pf->mac_stats_mem)) {
+			if (PTR_ERR(pf->mac_stats_mem) != -ENOENT) {
+				err = PTR_ERR(pf->mac_stats_mem);
+				goto err_unmap_ctrl;
+			}
+			pf->mac_stats_mem = NULL;
 		}
-		pf->mac_stats_mem = NULL;
 	}
 
 	pf->vf_cfg_mem = nfp_net_pf_map_rtsym(pf, "net.vfcfg",
-- 
cgit v1.1


From 3af7a156bdc356946098e13180be66b6420619bf Mon Sep 17 00:00:00 2001
From: Greg Thelen <gthelen@google.com>
Date: Thu, 26 Apr 2018 11:19:30 -0700
Subject: nvme: depend on INFINIBAND_ADDR_TRANS

NVME_RDMA code depends on INFINIBAND_ADDR_TRANS provided symbols.  So
declare the kconfig dependency.  This is necessary to allow for enabling
INFINIBAND without INFINIBAND_ADDR_TRANS.

Signed-off-by: Greg Thelen <gthelen@google.com>
Cc: Tarick Bedeir <tarick@google.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/nvme/host/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/nvme/host/Kconfig b/drivers/nvme/host/Kconfig
index b979cf3..88a8b59 100644
--- a/drivers/nvme/host/Kconfig
+++ b/drivers/nvme/host/Kconfig
@@ -27,7 +27,7 @@ config NVME_FABRICS
 
 config NVME_RDMA
 	tristate "NVM Express over Fabrics RDMA host driver"
-	depends on INFINIBAND && BLOCK
+	depends on INFINIBAND && INFINIBAND_ADDR_TRANS && BLOCK
 	select NVME_CORE
 	select NVME_FABRICS
 	select SG_POOL
-- 
cgit v1.1


From d6fc6a22fc7d3df987666725496ed5dd2dd30f23 Mon Sep 17 00:00:00 2001
From: Greg Thelen <gthelen@google.com>
Date: Thu, 26 Apr 2018 11:19:31 -0700
Subject: nvmet-rdma: depend on INFINIBAND_ADDR_TRANS

NVME_TARGET_RDMA code depends on INFINIBAND_ADDR_TRANS provided symbols.
So declare the kconfig dependency.  This is necessary to allow for
enabling INFINIBAND without INFINIBAND_ADDR_TRANS.

Signed-off-by: Greg Thelen <gthelen@google.com>
Cc: Tarick Bedeir <tarick@google.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/nvme/target/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/nvme/target/Kconfig b/drivers/nvme/target/Kconfig
index 5f4f8b1..3c7b61d 100644
--- a/drivers/nvme/target/Kconfig
+++ b/drivers/nvme/target/Kconfig
@@ -27,7 +27,7 @@ config NVME_TARGET_LOOP
 
 config NVME_TARGET_RDMA
 	tristate "NVMe over Fabrics RDMA target support"
-	depends on INFINIBAND
+	depends on INFINIBAND && INFINIBAND_ADDR_TRANS
 	depends on NVME_TARGET
 	select SGL_ALLOC
 	help
-- 
cgit v1.1


From 346a47b65d10e450778ec0d21e4a9409f25daaa8 Mon Sep 17 00:00:00 2001
From: Greg Thelen <gthelen@google.com>
Date: Thu, 26 Apr 2018 11:19:32 -0700
Subject: ib_srpt: depend on INFINIBAND_ADDR_TRANS

INFINIBAND_SRPT code depends on INFINIBAND_ADDR_TRANS provided symbols.
So declare the kconfig dependency.  This is necessary to allow for
enabling INFINIBAND without INFINIBAND_ADDR_TRANS.

Signed-off-by: Greg Thelen <gthelen@google.com>
Cc: Tarick Bedeir <tarick@google.com>
Reviewed-by: Bart Van Assche <bart.vanassche@wdc.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/ulp/srpt/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/infiniband/ulp/srpt/Kconfig b/drivers/infiniband/ulp/srpt/Kconfig
index 31ee83d..fb8b718 100644
--- a/drivers/infiniband/ulp/srpt/Kconfig
+++ b/drivers/infiniband/ulp/srpt/Kconfig
@@ -1,6 +1,6 @@
 config INFINIBAND_SRPT
 	tristate "InfiniBand SCSI RDMA Protocol target support"
-	depends on INFINIBAND && TARGET_CORE
+	depends on INFINIBAND && INFINIBAND_ADDR_TRANS && TARGET_CORE
 	---help---
 
 	  Support for the SCSI RDMA Protocol (SRP) Target driver. The
-- 
cgit v1.1


From 3c6b03d18df657d677808d7090b4d03bc6026efd Mon Sep 17 00:00:00 2001
From: Greg Thelen <gthelen@google.com>
Date: Thu, 26 Apr 2018 11:19:33 -0700
Subject: cifs: smbd: depend on INFINIBAND_ADDR_TRANS

CIFS_SMB_DIRECT code depends on INFINIBAND_ADDR_TRANS provided symbols.
So declare the kconfig dependency.  This is necessary to allow for
enabling INFINIBAND without INFINIBAND_ADDR_TRANS.

Signed-off-by: Greg Thelen <gthelen@google.com>
Cc: Tarick Bedeir <tarick@google.com>
Reviewed-by: Long Li <longli@microsoft.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 fs/cifs/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/cifs/Kconfig b/fs/cifs/Kconfig
index 741749a..5f132d5 100644
--- a/fs/cifs/Kconfig
+++ b/fs/cifs/Kconfig
@@ -197,7 +197,7 @@ config CIFS_SMB311
 
 config CIFS_SMB_DIRECT
 	bool "SMB Direct support (Experimental)"
-	depends on CIFS=m && INFINIBAND || CIFS=y && INFINIBAND=y
+	depends on CIFS=m && INFINIBAND && INFINIBAND_ADDR_TRANS || CIFS=y && INFINIBAND=y && INFINIBAND_ADDR_TRANS=y
 	help
 	  Enables SMB Direct experimental support for SMB 3.0, 3.02 and 3.1.1.
 	  SMB Direct allows transferring SMB packets over RDMA. If unsure,
-- 
cgit v1.1


From 5a3bc8a4abbd2d553430218d3a320400dce811b7 Mon Sep 17 00:00:00 2001
From: Greg Thelen <gthelen@google.com>
Date: Thu, 26 Apr 2018 11:19:34 -0700
Subject: ib_srp: depend on INFINIBAND_ADDR_TRANS

INFINIBAND_SRP code depends on INFINIBAND_ADDR_TRANS provided symbols.
So declare the kconfig dependency.  This is necessary to allow for
enabling INFINIBAND without INFINIBAND_ADDR_TRANS.

Signed-off-by: Greg Thelen <gthelen@google.com>
Cc: Tarick Bedeir <tarick@google.com>
Reviewed-by: Bart Van Assche <bart.vanassche@wdc.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/ulp/srp/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/infiniband/ulp/srp/Kconfig b/drivers/infiniband/ulp/srp/Kconfig
index c74ee96..99db8fe 100644
--- a/drivers/infiniband/ulp/srp/Kconfig
+++ b/drivers/infiniband/ulp/srp/Kconfig
@@ -1,6 +1,6 @@
 config INFINIBAND_SRP
 	tristate "InfiniBand SCSI RDMA Protocol"
-	depends on SCSI
+	depends on SCSI && INFINIBAND_ADDR_TRANS
 	select SCSI_SRP_ATTRS
 	---help---
 	  Support for the SCSI RDMA Protocol over InfiniBand.  This
-- 
cgit v1.1


From f7cb7b85be55a4906b4b4b30596db1043dae6335 Mon Sep 17 00:00:00 2001
From: Greg Thelen <gthelen@google.com>
Date: Thu, 26 Apr 2018 11:19:35 -0700
Subject: IB: make INFINIBAND_ADDR_TRANS configurable

Allow INFINIBAND without INFINIBAND_ADDR_TRANS because fuzzing has been
finding fair number of CM bugs.  So provide option to disable it.

Signed-off-by: Greg Thelen <gthelen@google.com>
Cc: Tarick Bedeir <tarick@google.com>
Reviewed-by: Bart Van Assche <bart.vanassche@wdc.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/Kconfig | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/infiniband/Kconfig b/drivers/infiniband/Kconfig
index ee270e0..2a972ed6 100644
--- a/drivers/infiniband/Kconfig
+++ b/drivers/infiniband/Kconfig
@@ -61,9 +61,12 @@ config INFINIBAND_ON_DEMAND_PAGING
 	  pages on demand instead.
 
 config INFINIBAND_ADDR_TRANS
-	bool
+	bool "RDMA/CM"
 	depends on INFINIBAND
 	default y
+	---help---
+	  Support for RDMA communication manager (CM).
+	  This allows for a generic connection abstraction over RDMA.
 
 config INFINIBAND_ADDR_TRANS_CONFIGFS
 	bool
-- 
cgit v1.1


From 6a30abaa40b62aed46ef12ea4c16c48565bdb376 Mon Sep 17 00:00:00 2001
From: Takashi Iwai <tiwai@suse.de>
Date: Fri, 27 Apr 2018 17:17:35 +0200
Subject: ALSA: hda - Fix incorrect usage of IS_REACHABLE()

The commit c469652bb5e8 ("ALSA: hda - Use IS_REACHABLE() for
dependency on input") simplified the dependencies with IS_REACHABLE()
macro, but it broke due to its incorrect usage: it should have been
IS_REACHABLE(CONFIG_INPUT) instead of IS_REACHABLE(INPUT).

Fixes: c469652bb5e8 ("ALSA: hda - Use IS_REACHABLE() for dependency on input")
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
---
 sound/pci/hda/patch_realtek.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c
index 8c238e5..2dd34dd 100644
--- a/sound/pci/hda/patch_realtek.c
+++ b/sound/pci/hda/patch_realtek.c
@@ -3832,7 +3832,7 @@ static void alc280_fixup_hp_gpio4(struct hda_codec *codec,
 	}
 }
 
-#if IS_REACHABLE(INPUT)
+#if IS_REACHABLE(CONFIG_INPUT)
 static void gpio2_mic_hotkey_event(struct hda_codec *codec,
 				   struct hda_jack_callback *event)
 {
-- 
cgit v1.1


From 45f972adb7f4db2d7f02af728ccd104113336074 Mon Sep 17 00:00:00 2001
From: Maxime Chevallier <maxime.chevallier@bootlin.com>
Date: Wed, 25 Apr 2018 20:21:16 +0200
Subject: net: mvpp2: Fix clk error path in mvpp2_probe

When clk_prepare_enable fails for the axi_clk, the mg_clk isn't properly
cleaned up. Add another jump label to handle that case, and make sure we
jump to it in the later error cases.

Fixes: 4792ea04bcd0 ("net: mvpp2: Fix clock resource by adding an optional bus clock")
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Acked-by: Gregory CLEMENT <gregory.clement@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/marvell/mvpp2.c | 15 ++++++++-------
 1 file changed, 8 insertions(+), 7 deletions(-)

diff --git a/drivers/net/ethernet/marvell/mvpp2.c b/drivers/net/ethernet/marvell/mvpp2.c
index 4202f9b..0c2f048 100644
--- a/drivers/net/ethernet/marvell/mvpp2.c
+++ b/drivers/net/ethernet/marvell/mvpp2.c
@@ -8774,12 +8774,12 @@ static int mvpp2_probe(struct platform_device *pdev)
 		if (IS_ERR(priv->axi_clk)) {
 			err = PTR_ERR(priv->axi_clk);
 			if (err == -EPROBE_DEFER)
-				goto err_gop_clk;
+				goto err_mg_clk;
 			priv->axi_clk = NULL;
 		} else {
 			err = clk_prepare_enable(priv->axi_clk);
 			if (err < 0)
-				goto err_gop_clk;
+				goto err_mg_clk;
 		}
 
 		/* Get system's tclk rate */
@@ -8793,7 +8793,7 @@ static int mvpp2_probe(struct platform_device *pdev)
 	if (priv->hw_version == MVPP22) {
 		err = dma_set_mask(&pdev->dev, MVPP2_DESC_DMA_MASK);
 		if (err)
-			goto err_mg_clk;
+			goto err_axi_clk;
 		/* Sadly, the BM pools all share the same register to
 		 * store the high 32 bits of their address. So they
 		 * must all have the same high 32 bits, which forces
@@ -8801,14 +8801,14 @@ static int mvpp2_probe(struct platform_device *pdev)
 		 */
 		err = dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(32));
 		if (err)
-			goto err_mg_clk;
+			goto err_axi_clk;
 	}
 
 	/* Initialize network controller */
 	err = mvpp2_init(pdev, priv);
 	if (err < 0) {
 		dev_err(&pdev->dev, "failed to initialize controller\n");
-		goto err_mg_clk;
+		goto err_axi_clk;
 	}
 
 	/* Initialize ports */
@@ -8821,7 +8821,7 @@ static int mvpp2_probe(struct platform_device *pdev)
 	if (priv->port_count == 0) {
 		dev_err(&pdev->dev, "no ports enabled\n");
 		err = -ENODEV;
-		goto err_mg_clk;
+		goto err_axi_clk;
 	}
 
 	/* Statistics must be gathered regularly because some of them (like
@@ -8849,8 +8849,9 @@ err_port_probe:
 			mvpp2_port_remove(priv->port_list[i]);
 		i++;
 	}
-err_mg_clk:
+err_axi_clk:
 	clk_disable_unprepare(priv->axi_clk);
+err_mg_clk:
 	if (priv->hw_version == MVPP22)
 		clk_disable_unprepare(priv->mg_clk);
 err_gop_clk:
-- 
cgit v1.1


From 9af771ced473f92b5e57d086a0c2453fc0cb149c Mon Sep 17 00:00:00 2001
From: Maxime Chevallier <maxime.chevallier@bootlin.com>
Date: Wed, 25 Apr 2018 20:21:17 +0200
Subject: net: mvpp2: Fix clock resource by adding missing mg_core_clk

Marvell's PPv2.2 IP needs an additional clock named "MG Core clock".
This is required on Armada 7K and 8K.

This commit adds the required clock in mvpp2, making sure it's only
used on PPv2.2.

Fixes: c7e92def1ef4 ("clk: mvebu: cp110: Fix clock tree representation")
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/marvell/mvpp2.c | 19 +++++++++++++++++--
 1 file changed, 17 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/marvell/mvpp2.c b/drivers/net/ethernet/marvell/mvpp2.c
index 0c2f048..6f41023 100644
--- a/drivers/net/ethernet/marvell/mvpp2.c
+++ b/drivers/net/ethernet/marvell/mvpp2.c
@@ -942,6 +942,7 @@ struct mvpp2 {
 	struct clk *pp_clk;
 	struct clk *gop_clk;
 	struct clk *mg_clk;
+	struct clk *mg_core_clk;
 	struct clk *axi_clk;
 
 	/* List of pointers to port structures */
@@ -8768,18 +8769,27 @@ static int mvpp2_probe(struct platform_device *pdev)
 			err = clk_prepare_enable(priv->mg_clk);
 			if (err < 0)
 				goto err_gop_clk;
+
+			priv->mg_core_clk = devm_clk_get(&pdev->dev, "mg_core_clk");
+			if (IS_ERR(priv->mg_core_clk)) {
+				priv->mg_core_clk = NULL;
+			} else {
+				err = clk_prepare_enable(priv->mg_core_clk);
+				if (err < 0)
+					goto err_mg_clk;
+			}
 		}
 
 		priv->axi_clk = devm_clk_get(&pdev->dev, "axi_clk");
 		if (IS_ERR(priv->axi_clk)) {
 			err = PTR_ERR(priv->axi_clk);
 			if (err == -EPROBE_DEFER)
-				goto err_mg_clk;
+				goto err_mg_core_clk;
 			priv->axi_clk = NULL;
 		} else {
 			err = clk_prepare_enable(priv->axi_clk);
 			if (err < 0)
-				goto err_mg_clk;
+				goto err_mg_core_clk;
 		}
 
 		/* Get system's tclk rate */
@@ -8851,6 +8861,10 @@ err_port_probe:
 	}
 err_axi_clk:
 	clk_disable_unprepare(priv->axi_clk);
+
+err_mg_core_clk:
+	if (priv->hw_version == MVPP22)
+		clk_disable_unprepare(priv->mg_core_clk);
 err_mg_clk:
 	if (priv->hw_version == MVPP22)
 		clk_disable_unprepare(priv->mg_clk);
@@ -8898,6 +8912,7 @@ static int mvpp2_remove(struct platform_device *pdev)
 		return 0;
 
 	clk_disable_unprepare(priv->axi_clk);
+	clk_disable_unprepare(priv->mg_core_clk);
 	clk_disable_unprepare(priv->mg_clk);
 	clk_disable_unprepare(priv->pp_clk);
 	clk_disable_unprepare(priv->gop_clk);
-- 
cgit v1.1


From 16ae6aa1705299789f71fdea59bfb119c1fbd9c0 Mon Sep 17 00:00:00 2001
From: Yuchung Cheng <ycheng@google.com>
Date: Wed, 25 Apr 2018 11:33:08 -0700
Subject: tcp: ignore Fast Open on repair mode

The TCP repair sequence of operation is to first set the socket in
repair mode, then inject the TCP stats into the socket with repair
socket options, then call connect() to re-activate the socket. The
connect syscall simply returns and set state to ESTABLISHED
mode. As a result Fast Open is meaningless for TCP repair.

However allowing sendto() system call with MSG_FASTOPEN flag half-way
during the repair operation could unexpectedly cause data to be
sent, before the operation finishes changing the internal TCP stats
(e.g. MSS).  This in turn triggers TCP warnings on inconsistent
packet accounting.

The fix is to simply disallow Fast Open operation once the socket
is in the repair mode.

Reported-by: syzbot <syzkaller@googlegroups.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Reviewed-by: Neal Cardwell <ncardwell@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/ipv4/tcp.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 9ce1c72..4b18ad4 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -1204,7 +1204,8 @@ int tcp_sendmsg_locked(struct sock *sk, struct msghdr *msg, size_t size)
 			uarg->zerocopy = 0;
 	}
 
-	if (unlikely(flags & MSG_FASTOPEN || inet_sk(sk)->defer_connect)) {
+	if (unlikely(flags & MSG_FASTOPEN || inet_sk(sk)->defer_connect) &&
+	    !tp->repair) {
 		err = tcp_sendmsg_fastopen(sk, msg, &copied_syn, size);
 		if (err == -EINPROGRESS && copied_syn > 0)
 			goto out;
-- 
cgit v1.1


From 8f22e52528cc372b218b5f100457469615c733ce Mon Sep 17 00:00:00 2001
From: Takashi Iwai <tiwai@suse.de>
Date: Thu, 26 Apr 2018 09:17:45 +0200
Subject: ALSA: seq: Fix races at MIDI encoding in snd_virmidi_output_trigger()

The sequencer virmidi code has an open race at its output trigger
callback: namely, virmidi keeps only one event packet for processing
while it doesn't protect for concurrent output trigger calls.

snd_virmidi_output_trigger() tries to process the previously
unfinished event before starting encoding the given MIDI stream, but
this is done without any lock.  Meanwhile, if another rawmidi stream
starts the output trigger, this proceeds further, and overwrites the
event package that is being processed in another thread.  This
eventually corrupts and may lead to the invalid memory access if the
event type is like SYSEX.

The fix is just to move the spinlock to cover both the pending event
and the new stream.

The bug was spotted by a new fuzzer, RaceFuzzer.

BugLink: http://lkml.kernel.org/r/20180426045223.GA15307@dragonet.kaist.ac.kr
Reported-by: DaeRyong Jeong <threeearcat@gmail.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
---
 sound/core/seq/seq_virmidi.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/sound/core/seq/seq_virmidi.c b/sound/core/seq/seq_virmidi.c
index f48a4cd..289ae6b 100644
--- a/sound/core/seq/seq_virmidi.c
+++ b/sound/core/seq/seq_virmidi.c
@@ -174,12 +174,12 @@ static void snd_virmidi_output_trigger(struct snd_rawmidi_substream *substream,
 			}
 			return;
 		}
+		spin_lock_irqsave(&substream->runtime->lock, flags);
 		if (vmidi->event.type != SNDRV_SEQ_EVENT_NONE) {
 			if (snd_seq_kernel_client_dispatch(vmidi->client, &vmidi->event, in_atomic(), 0) < 0)
-				return;
+				goto out;
 			vmidi->event.type = SNDRV_SEQ_EVENT_NONE;
 		}
-		spin_lock_irqsave(&substream->runtime->lock, flags);
 		while (1) {
 			count = __snd_rawmidi_transmit_peek(substream, buf, sizeof(buf));
 			if (count <= 0)
-- 
cgit v1.1


From 9faedd643fd9f3a53f10ca270d3bbd436b908766 Mon Sep 17 00:00:00 2001
From: Anders Roxell <anders.roxell@linaro.org>
Date: Wed, 25 Apr 2018 23:32:06 +0200
Subject: selftests: net: add in_netns.sh TEST_GEN_PROGS_EXTENDED

Script in_netns.sh is a utility function and not its own test so it
shouldn't be part of the TEST_PROGS. The in_netns.sh get used by
run_afpackettests.
To install in_netns.sh without being added to the main run_kselftest.sh
script use the TEST_GEN_PROGS_EXTENDED variable.

Fixes: 5ff9c1a3dd92 ("selftests: net: add in_netns.sh to TEST_PROGS")
Signed-off-by: Anders Roxell <anders.roxell@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 tools/testing/selftests/net/Makefile | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/tools/testing/selftests/net/Makefile b/tools/testing/selftests/net/Makefile
index 8f1e13d..daf5eff 100644
--- a/tools/testing/selftests/net/Makefile
+++ b/tools/testing/selftests/net/Makefile
@@ -5,7 +5,8 @@ CFLAGS =  -Wall -Wl,--no-as-needed -O2 -g
 CFLAGS += -I../../../../usr/include/
 
 TEST_PROGS := run_netsocktests run_afpackettests test_bpf.sh netdevice.sh rtnetlink.sh
-TEST_PROGS += fib_tests.sh fib-onlink-tests.sh in_netns.sh pmtu.sh
+TEST_PROGS += fib_tests.sh fib-onlink-tests.sh pmtu.sh
+TEST_GEN_PROGS_EXTENDED := in_netns.sh
 TEST_GEN_FILES =  socket
 TEST_GEN_FILES += psock_fanout psock_tpacket msg_zerocopy
 TEST_GEN_PROGS = reuseport_bpf reuseport_bpf_cpu reuseport_bpf_numa
-- 
cgit v1.1


From 1f3ccc3c3fc26468be00392ef0b2c215f9c9d054 Mon Sep 17 00:00:00 2001
From: Florian Fainelli <f.fainelli@gmail.com>
Date: Wed, 25 Apr 2018 16:21:51 -0700
Subject: net: systemport: Correclty disambiguate driver instances

While adding the DSA notifier, we will be sending DSA notifications with
info->master that is going to point to a particular net_device instance.

Our logic in bcm_sysport_map_queues() correctly disambiguates net_device
instances that are not covered by our own driver, but it will not make
sure that info->master points to a particular driver instance that we
are interested in. In a system where e.g: two or more SYSTEMPORT
instances are registered, this would lead in programming two or more
times the queue mapping, completely messing with the logic which does
the queue/port allocation and tracking.

Fix this by looking at the notifier_block pointer which is unique per
instance and allows us to go back to our driver private structure, and
in turn to the backing net_device instance.

Fixes: d156576362c0 ("net: systemport: Establish lower/upper queue mapping")
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/broadcom/bcmsysport.c | 16 ++++++++++++----
 1 file changed, 12 insertions(+), 4 deletions(-)

diff --git a/drivers/net/ethernet/broadcom/bcmsysport.c b/drivers/net/ethernet/broadcom/bcmsysport.c
index f9a3c1a..0c2b0fa 100644
--- a/drivers/net/ethernet/broadcom/bcmsysport.c
+++ b/drivers/net/ethernet/broadcom/bcmsysport.c
@@ -2144,14 +2144,21 @@ static const struct net_device_ops bcm_sysport_netdev_ops = {
 	.ndo_select_queue	= bcm_sysport_select_queue,
 };
 
-static int bcm_sysport_map_queues(struct net_device *dev,
+static int bcm_sysport_map_queues(struct notifier_block *nb,
 				  struct dsa_notifier_register_info *info)
 {
-	struct bcm_sysport_priv *priv = netdev_priv(dev);
 	struct bcm_sysport_tx_ring *ring;
+	struct bcm_sysport_priv *priv;
 	struct net_device *slave_dev;
 	unsigned int num_tx_queues;
 	unsigned int q, start, port;
+	struct net_device *dev;
+
+	priv = container_of(nb, struct bcm_sysport_priv, dsa_notifier);
+	if (priv->netdev != info->master)
+		return 0;
+
+	dev = info->master;
 
 	/* We can't be setting up queue inspection for non directly attached
 	 * switches
@@ -2174,6 +2181,7 @@ static int bcm_sysport_map_queues(struct net_device *dev,
 	if (priv->is_lite)
 		netif_set_real_num_tx_queues(slave_dev,
 					     slave_dev->num_tx_queues / 2);
+
 	num_tx_queues = slave_dev->real_num_tx_queues;
 
 	if (priv->per_port_num_tx_queues &&
@@ -2201,7 +2209,7 @@ static int bcm_sysport_map_queues(struct net_device *dev,
 	return 0;
 }
 
-static int bcm_sysport_dsa_notifier(struct notifier_block *unused,
+static int bcm_sysport_dsa_notifier(struct notifier_block *nb,
 				    unsigned long event, void *ptr)
 {
 	struct dsa_notifier_register_info *info;
@@ -2211,7 +2219,7 @@ static int bcm_sysport_dsa_notifier(struct notifier_block *unused,
 
 	info = ptr;
 
-	return notifier_from_errno(bcm_sysport_map_queues(info->master, info));
+	return notifier_from_errno(bcm_sysport_map_queues(nb, info));
 }
 
 #define REV_FMT	"v%2x.%02x"
-- 
cgit v1.1


From d625329b06e46bd20baf9ee40847d11982569204 Mon Sep 17 00:00:00 2001
From: Xin Long <lucien.xin@gmail.com>
Date: Thu, 26 Apr 2018 14:13:57 +0800
Subject: sctp: handle two v4 addrs comparison in sctp_inet6_cmp_addr

Since sctp ipv6 socket also supports v4 addrs, it's possible to
compare two v4 addrs in pf v6 .cmp_addr, sctp_inet6_cmp_addr.

However after Commit 1071ec9d453a ("sctp: do not check port in
sctp_inet6_cmp_addr"), it no longer calls af1->cmp_addr, which
in this case is sctp_v4_cmp_addr, but calls __sctp_v6_cmp_addr
where it handles them as two v6 addrs. It would cause a out of
bounds crash.

syzbot found this crash when trying to bind two v4 addrs to a
v6 socket.

This patch fixes it by adding the process for two v4 addrs in
sctp_inet6_cmp_addr.

Fixes: 1071ec9d453a ("sctp: do not check port in sctp_inet6_cmp_addr")
Reported-by: syzbot+cd494c1dd681d4d93ebb@syzkaller.appspotmail.com
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/sctp/ipv6.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/net/sctp/ipv6.c b/net/sctp/ipv6.c
index 2e3f7b7..4224711 100644
--- a/net/sctp/ipv6.c
+++ b/net/sctp/ipv6.c
@@ -895,6 +895,9 @@ static int sctp_inet6_cmp_addr(const union sctp_addr *addr1,
 	if (sctp_is_any(sk, addr1) || sctp_is_any(sk, addr2))
 		return 1;
 
+	if (addr1->sa.sa_family == AF_INET && addr2->sa.sa_family == AF_INET)
+		return addr1->v4.sin_addr.s_addr == addr2->v4.sin_addr.s_addr;
+
 	return __sctp_v6_cmp_addr(addr1, addr2);
 }
 
-- 
cgit v1.1


From 9306b38e42cb266f98bff6f6f4c1c652aa79ba45 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?SZ=20Lin=20=28=E6=9E=97=E4=B8=8A=E6=99=BA=29?=
 <sz.lin@moxa.com>
Date: Thu, 26 Apr 2018 14:30:13 +0800
Subject: NET: usb: qmi_wwan: add support for ublox R410M PID 0x90b2
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This patch adds support for PID 0x90b2 of ublox R410M.

qmicli -d /dev/cdc-wdm0 --dms-get-manufacturer
[/dev/cdc-wdm0] Device manufacturer retrieved:
        Manufacturer: 'u-blox'

qmicli -d /dev/cdc-wdm0 --dms-get-model
[/dev/cdc-wdm0] Device model retrieved:
        Model: 'SARA-R410M-02B'

Signed-off-by: SZ Lin (林上智) <sz.lin@moxa.com>
Cc: stable <stable@vger.kernel.org>
Acked-by: Bjørn Mork <bjorn@mork.no>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/usb/qmi_wwan.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/net/usb/qmi_wwan.c b/drivers/net/usb/qmi_wwan.c
index c853e74..51c68fc 100644
--- a/drivers/net/usb/qmi_wwan.c
+++ b/drivers/net/usb/qmi_wwan.c
@@ -1098,6 +1098,7 @@ static const struct usb_device_id products[] = {
 	{QMI_FIXED_INTF(0x05c6, 0x9080, 8)},
 	{QMI_FIXED_INTF(0x05c6, 0x9083, 3)},
 	{QMI_FIXED_INTF(0x05c6, 0x9084, 4)},
+	{QMI_FIXED_INTF(0x05c6, 0x90b2, 3)},    /* ublox R410M */
 	{QMI_FIXED_INTF(0x05c6, 0x920d, 0)},
 	{QMI_FIXED_INTF(0x05c6, 0x920d, 5)},
 	{QMI_QUIRK_SET_DTR(0x05c6, 0x9625, 4)},	/* YUGA CLM920-NC5 */
-- 
cgit v1.1


From 6a9a27d5397fc6c52f90c09ddab91e65053584aa Mon Sep 17 00:00:00 2001
From: Xin Long <lucien.xin@gmail.com>
Date: Thu, 26 Apr 2018 15:21:44 +0800
Subject: sctp: clear the new asoc's stream outcnt in sctp_stream_update

When processing a duplicate cookie-echo chunk, sctp moves the new
temp asoc's stream out/in into the old asoc, and later frees this
new temp asoc.

But now after this move, the new temp asoc's stream->outcnt is not
cleared while stream->out is set to NULL, which would cause a same
crash as the one fixed in Commit 79d0895140e9 ("sctp: fix error
path in sctp_stream_init") when freeing this asoc later.

This fix is to clear this outcnt in sctp_stream_update.

Fixes: f952be79cebd ("sctp: introduce struct sctp_stream_out_ext")
Reported-by: Jianwen Ji <jiji@redhat.com>
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/sctp/stream.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/net/sctp/stream.c b/net/sctp/stream.c
index f799043..f1f1d1b 100644
--- a/net/sctp/stream.c
+++ b/net/sctp/stream.c
@@ -240,6 +240,8 @@ void sctp_stream_update(struct sctp_stream *stream, struct sctp_stream *new)
 
 	new->out = NULL;
 	new->in  = NULL;
+	new->outcnt = 0;
+	new->incnt  = 0;
 }
 
 static int sctp_send_reconf(struct sctp_association *asoc,
-- 
cgit v1.1


From c7f46cca8c73a44311e4164b9196b4d791f59ac7 Mon Sep 17 00:00:00 2001
From: Ido Schimmel <idosch@mellanox.com>
Date: Thu, 26 Apr 2018 11:46:29 +0300
Subject: mlxsw: spectrum_switchdev: Do not remove mrouter port from MDB's
 ports list

When IGMP snooping is enabled on a bridge, traffic forwarded by an MDB
entry should be sent to both ports member in the MDB's ports list and
mrouter ports.

In case a port needs to be removed from an MDB's ports list, but this
port is also configured as an mrouter port, then do not update the
device so that it will continue to forward traffic through that port.

Fix a copy-paste error that checked that IGMP snooping is enabled twice
instead of checking the port's mrouter state.

Fixes: ded711c87a04 ("mlxsw: spectrum_switchdev: Consider mrouter status for mdb changes")
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reported-by: Colin King <colin.king@canonical.com>
Reviewed-by: Nogah Frankel <nogahf@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c | 12 +++++-------
 1 file changed, 5 insertions(+), 7 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c
index c11c9a6..4ed0118 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c
@@ -1718,13 +1718,11 @@ __mlxsw_sp_port_mdb_del(struct mlxsw_sp_port *mlxsw_sp_port,
 	struct net_device *dev = mlxsw_sp_port->dev;
 	int err;
 
-	if (bridge_port->bridge_device->multicast_enabled) {
-		if (bridge_port->bridge_device->multicast_enabled) {
-			err = mlxsw_sp_port_smid_set(mlxsw_sp_port, mid->mid,
-						     false);
-			if (err)
-				netdev_err(dev, "Unable to remove port from SMID\n");
-		}
+	if (bridge_port->bridge_device->multicast_enabled &&
+	    !bridge_port->mrouter) {
+		err = mlxsw_sp_port_smid_set(mlxsw_sp_port, mid->mid, false);
+		if (err)
+			netdev_err(dev, "Unable to remove port from SMID\n");
 	}
 
 	err = mlxsw_sp_port_remove_from_mid(mlxsw_sp_port, mid);
-- 
cgit v1.1


From 26bff1bd74a4f7417509a83295614e9dab995b2a Mon Sep 17 00:00:00 2001
From: Raju Rangoju <rajur@chelsio.com>
Date: Mon, 23 Apr 2018 21:42:37 +0530
Subject: RDMA/cxgb4: release hw resources on device removal

The c4iw_rdev_close() logic was not releasing all the hw
resources (PBL and RQT memory) during the device removal
event (driver unload / system reboot). This can cause panic
in gen_pool_destroy().

The module remove function will wait for all the hw
resources to be released during the device removal event.

Fixes c12a67fe(iw_cxgb4: free EQ queue memory on last deref)
Signed-off-by: Raju Rangoju <rajur@chelsio.com>
Reviewed-by: Steve Wise <swise@opengridcomputing.com>
Cc: stable@vger.kernel.org
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/cxgb4/device.c   |  9 ++++++++-
 drivers/infiniband/hw/cxgb4/iw_cxgb4.h |  4 ++++
 drivers/infiniband/hw/cxgb4/resource.c | 26 ++++++++++++++++++++++++--
 3 files changed, 36 insertions(+), 3 deletions(-)

diff --git a/drivers/infiniband/hw/cxgb4/device.c b/drivers/infiniband/hw/cxgb4/device.c
index feeb8ee..44161ca 100644
--- a/drivers/infiniband/hw/cxgb4/device.c
+++ b/drivers/infiniband/hw/cxgb4/device.c
@@ -875,6 +875,11 @@ static int c4iw_rdev_open(struct c4iw_rdev *rdev)
 
 	rdev->status_page->db_off = 0;
 
+	init_completion(&rdev->rqt_compl);
+	init_completion(&rdev->pbl_compl);
+	kref_init(&rdev->rqt_kref);
+	kref_init(&rdev->pbl_kref);
+
 	return 0;
 err_free_status_page_and_wr_log:
 	if (c4iw_wr_log && rdev->wr_log)
@@ -893,13 +898,15 @@ destroy_resource:
 
 static void c4iw_rdev_close(struct c4iw_rdev *rdev)
 {
-	destroy_workqueue(rdev->free_workq);
 	kfree(rdev->wr_log);
 	c4iw_release_dev_ucontext(rdev, &rdev->uctx);
 	free_page((unsigned long)rdev->status_page);
 	c4iw_pblpool_destroy(rdev);
 	c4iw_rqtpool_destroy(rdev);
+	wait_for_completion(&rdev->pbl_compl);
+	wait_for_completion(&rdev->rqt_compl);
 	c4iw_ocqp_pool_destroy(rdev);
+	destroy_workqueue(rdev->free_workq);
 	c4iw_destroy_resource(&rdev->resource);
 }
 
diff --git a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h
index cc92900..a60def2 100644
--- a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h
+++ b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h
@@ -185,6 +185,10 @@ struct c4iw_rdev {
 	struct wr_log_entry *wr_log;
 	int wr_log_size;
 	struct workqueue_struct *free_workq;
+	struct completion rqt_compl;
+	struct completion pbl_compl;
+	struct kref rqt_kref;
+	struct kref pbl_kref;
 };
 
 static inline int c4iw_fatal_error(struct c4iw_rdev *rdev)
diff --git a/drivers/infiniband/hw/cxgb4/resource.c b/drivers/infiniband/hw/cxgb4/resource.c
index 3cf2599..0ef25ae 100644
--- a/drivers/infiniband/hw/cxgb4/resource.c
+++ b/drivers/infiniband/hw/cxgb4/resource.c
@@ -260,12 +260,22 @@ u32 c4iw_pblpool_alloc(struct c4iw_rdev *rdev, int size)
 		rdev->stats.pbl.cur += roundup(size, 1 << MIN_PBL_SHIFT);
 		if (rdev->stats.pbl.cur > rdev->stats.pbl.max)
 			rdev->stats.pbl.max = rdev->stats.pbl.cur;
+		kref_get(&rdev->pbl_kref);
 	} else
 		rdev->stats.pbl.fail++;
 	mutex_unlock(&rdev->stats.lock);
 	return (u32)addr;
 }
 
+static void destroy_pblpool(struct kref *kref)
+{
+	struct c4iw_rdev *rdev;
+
+	rdev = container_of(kref, struct c4iw_rdev, pbl_kref);
+	gen_pool_destroy(rdev->pbl_pool);
+	complete(&rdev->pbl_compl);
+}
+
 void c4iw_pblpool_free(struct c4iw_rdev *rdev, u32 addr, int size)
 {
 	pr_debug("addr 0x%x size %d\n", addr, size);
@@ -273,6 +283,7 @@ void c4iw_pblpool_free(struct c4iw_rdev *rdev, u32 addr, int size)
 	rdev->stats.pbl.cur -= roundup(size, 1 << MIN_PBL_SHIFT);
 	mutex_unlock(&rdev->stats.lock);
 	gen_pool_free(rdev->pbl_pool, (unsigned long)addr, size);
+	kref_put(&rdev->pbl_kref, destroy_pblpool);
 }
 
 int c4iw_pblpool_create(struct c4iw_rdev *rdev)
@@ -310,7 +321,7 @@ int c4iw_pblpool_create(struct c4iw_rdev *rdev)
 
 void c4iw_pblpool_destroy(struct c4iw_rdev *rdev)
 {
-	gen_pool_destroy(rdev->pbl_pool);
+	kref_put(&rdev->pbl_kref, destroy_pblpool);
 }
 
 /*
@@ -331,12 +342,22 @@ u32 c4iw_rqtpool_alloc(struct c4iw_rdev *rdev, int size)
 		rdev->stats.rqt.cur += roundup(size << 6, 1 << MIN_RQT_SHIFT);
 		if (rdev->stats.rqt.cur > rdev->stats.rqt.max)
 			rdev->stats.rqt.max = rdev->stats.rqt.cur;
+		kref_get(&rdev->rqt_kref);
 	} else
 		rdev->stats.rqt.fail++;
 	mutex_unlock(&rdev->stats.lock);
 	return (u32)addr;
 }
 
+static void destroy_rqtpool(struct kref *kref)
+{
+	struct c4iw_rdev *rdev;
+
+	rdev = container_of(kref, struct c4iw_rdev, rqt_kref);
+	gen_pool_destroy(rdev->rqt_pool);
+	complete(&rdev->rqt_compl);
+}
+
 void c4iw_rqtpool_free(struct c4iw_rdev *rdev, u32 addr, int size)
 {
 	pr_debug("addr 0x%x size %d\n", addr, size << 6);
@@ -344,6 +365,7 @@ void c4iw_rqtpool_free(struct c4iw_rdev *rdev, u32 addr, int size)
 	rdev->stats.rqt.cur -= roundup(size << 6, 1 << MIN_RQT_SHIFT);
 	mutex_unlock(&rdev->stats.lock);
 	gen_pool_free(rdev->rqt_pool, (unsigned long)addr, size << 6);
+	kref_put(&rdev->rqt_kref, destroy_rqtpool);
 }
 
 int c4iw_rqtpool_create(struct c4iw_rdev *rdev)
@@ -380,7 +402,7 @@ int c4iw_rqtpool_create(struct c4iw_rdev *rdev)
 
 void c4iw_rqtpool_destroy(struct c4iw_rdev *rdev)
 {
-	gen_pool_destroy(rdev->rqt_pool);
+	kref_put(&rdev->rqt_kref, destroy_rqtpool);
 }
 
 /*
-- 
cgit v1.1


From f604db645a66b7ba4f21c426fe73253928dada41 Mon Sep 17 00:00:00 2001
From: Matan Barak <matanb@mellanox.com>
Date: Tue, 24 Apr 2018 08:15:20 +0000
Subject: IB/uverbs: Fix validating mandatory attributes

Previously, if a method contained mandatory attributes in a namespace
that wasn't given by the user, these attributes weren't validated.
Fixing this by iterating over all specification namespaces.

Fixes: fac9658cabb9 ("IB/core: Add new ioctl interface")
Signed-off-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/uverbs_ioctl.c | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/drivers/infiniband/core/uverbs_ioctl.c b/drivers/infiniband/core/uverbs_ioctl.c
index 8c93970..8d32c4a 100644
--- a/drivers/infiniband/core/uverbs_ioctl.c
+++ b/drivers/infiniband/core/uverbs_ioctl.c
@@ -234,6 +234,15 @@ static int uverbs_validate_kernel_mandatory(const struct uverbs_method_spec *met
 			return -EINVAL;
 	}
 
+	for (; i < method_spec->num_buckets; i++) {
+		struct uverbs_attr_spec_hash *attr_spec_bucket =
+			method_spec->attr_buckets[i];
+
+		if (!bitmap_empty(attr_spec_bucket->mandatory_attrs_bitmask,
+				  attr_spec_bucket->num_attrs))
+			return -EINVAL;
+	}
+
 	return 0;
 }
 
-- 
cgit v1.1


From 2918c1a900252b4a0c730715ec205437c7daf79d Mon Sep 17 00:00:00 2001
From: Parav Pandit <parav@mellanox.com>
Date: Tue, 24 Apr 2018 20:13:45 +0300
Subject: RDMA/cma: Fix use after destroy access to net namespace for IPoIB

There are few issues with validation of netdevice and listen id lookup
for IB (IPoIB) while processing incoming CM request as below.

1. While performing lookup of bind_list in cma_ps_find(), net namespace
of the netdevice can get deleted in cma_exit_net(), resulting in use
after free access of idr and/or net namespace structures.
This lookup occurs from the workqueue context (and not userspace
context where net namespace is always valid).

           CPU0                              CPU1
           ====                              ====

 bind_list = cma_ps_find();
                                     move netdevice to new namespace
                                     delete net namespace
                                        cma_exit_net()
                                           idr_destroy(idr);

 [..]
 cma_find_listener(bind_list, ..);

2. While netdevice is validated for IP address in given net namespace,
netdevice's net namespace and/or ifindex can change in
cma_get_net_dev() and cma_match_net_dev().

Above issues are overcome by using rcu lock along with netdevice
UP/DOWN state as described below.
When a net namespace is getting deleted, netdevice is closed and
shutdown before moving it back to init_net namespace.
change_net_namespace() synchronizes with any existing use of netdevice
before changing the netdev properties such as net or ifindex.
Once netdevice IFF_UP flags is cleared, such fields are not guaranteed
to be valid.
Therefore, rcu lock along with netdevice state check ensures that,
while route lookup and cm_id lookup is in progress, netdevice of
interest won't migrate to any other net namespace.
This ensures that associated net namespace of netdevice won't get
deleted while rcu lock is held for netdevice which is in IFF_UP state.

Fixes: fa20105e09e9 ("IB/cma: Add support for network namespaces")
Fixes: 4be74b42a6d0 ("IB/cma: Separate port allocation to network namespaces")
Fixes: f887f2ac87c2 ("IB/cma: Validate routing of incoming requests")
Signed-off-by: Parav Pandit <parav@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/cma.c | 53 +++++++++++++++++++++++++++++++++++--------
 1 file changed, 43 insertions(+), 10 deletions(-)

diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c
index 51a6410..8364223 100644
--- a/drivers/infiniband/core/cma.c
+++ b/drivers/infiniband/core/cma.c
@@ -382,6 +382,8 @@ struct cma_hdr {
 #define CMA_VERSION 0x00
 
 struct cma_req_info {
+	struct sockaddr_storage listen_addr_storage;
+	struct sockaddr_storage src_addr_storage;
 	struct ib_device *device;
 	int port;
 	union ib_gid local_gid;
@@ -1340,11 +1342,11 @@ static bool validate_net_dev(struct net_device *net_dev,
 }
 
 static struct net_device *cma_get_net_dev(struct ib_cm_event *ib_event,
-					  const struct cma_req_info *req)
+					  struct cma_req_info *req)
 {
-	struct sockaddr_storage listen_addr_storage, src_addr_storage;
-	struct sockaddr *listen_addr = (struct sockaddr *)&listen_addr_storage,
-			*src_addr = (struct sockaddr *)&src_addr_storage;
+	struct sockaddr *listen_addr =
+			(struct sockaddr *)&req->listen_addr_storage;
+	struct sockaddr *src_addr = (struct sockaddr *)&req->src_addr_storage;
 	struct net_device *net_dev;
 	const union ib_gid *gid = req->has_gid ? &req->local_gid : NULL;
 	int err;
@@ -1359,11 +1361,6 @@ static struct net_device *cma_get_net_dev(struct ib_cm_event *ib_event,
 	if (!net_dev)
 		return ERR_PTR(-ENODEV);
 
-	if (!validate_net_dev(net_dev, listen_addr, src_addr)) {
-		dev_put(net_dev);
-		return ERR_PTR(-EHOSTUNREACH);
-	}
-
 	return net_dev;
 }
 
@@ -1490,15 +1487,51 @@ static struct rdma_id_private *cma_id_from_event(struct ib_cm_id *cm_id,
 		}
 	}
 
+	/*
+	 * Net namespace might be getting deleted while route lookup,
+	 * cm_id lookup is in progress. Therefore, perform netdevice
+	 * validation, cm_id lookup under rcu lock.
+	 * RCU lock along with netdevice state check, synchronizes with
+	 * netdevice migrating to different net namespace and also avoids
+	 * case where net namespace doesn't get deleted while lookup is in
+	 * progress.
+	 * If the device state is not IFF_UP, its properties such as ifindex
+	 * and nd_net cannot be trusted to remain valid without rcu lock.
+	 * net/core/dev.c change_net_namespace() ensures to synchronize with
+	 * ongoing operations on net device after device is closed using
+	 * synchronize_net().
+	 */
+	rcu_read_lock();
+	if (*net_dev) {
+		/*
+		 * If netdevice is down, it is likely that it is administratively
+		 * down or it might be migrating to different namespace.
+		 * In that case avoid further processing, as the net namespace
+		 * or ifindex may change.
+		 */
+		if (((*net_dev)->flags & IFF_UP) == 0) {
+			id_priv = ERR_PTR(-EHOSTUNREACH);
+			goto err;
+		}
+
+		if (!validate_net_dev(*net_dev,
+				 (struct sockaddr *)&req.listen_addr_storage,
+				 (struct sockaddr *)&req.src_addr_storage)) {
+			id_priv = ERR_PTR(-EHOSTUNREACH);
+			goto err;
+		}
+	}
+
 	bind_list = cma_ps_find(*net_dev ? dev_net(*net_dev) : &init_net,
 				rdma_ps_from_service_id(req.service_id),
 				cma_port_from_service_id(req.service_id));
 	id_priv = cma_find_listener(bind_list, cm_id, ib_event, &req, *net_dev);
+err:
+	rcu_read_unlock();
 	if (IS_ERR(id_priv) && *net_dev) {
 		dev_put(*net_dev);
 		*net_dev = NULL;
 	}
-
 	return id_priv;
 }
 
-- 
cgit v1.1


From c192a12ce82ca0a951bd5449a21d53c5f3f88697 Mon Sep 17 00:00:00 2001
From: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Date: Tue, 24 Apr 2018 15:15:45 +0200
Subject: IB/nes: fix nes_netdev_start_xmit()'s return type

The method ndo_start_xmit() is defined as returning an 'netdev_tx_t',
which is a typedef for an enum type, but the implementation in this
driver returns an 'int'.

Fix this by returning 'netdev_tx_t' in this driver too.

Signed-off-by: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Reviewed-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/nes/nes_nic.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/infiniband/hw/nes/nes_nic.c b/drivers/infiniband/hw/nes/nes_nic.c
index 0a75164..007d5e8 100644
--- a/drivers/infiniband/hw/nes/nes_nic.c
+++ b/drivers/infiniband/hw/nes/nes_nic.c
@@ -461,7 +461,7 @@ static bool nes_nic_send(struct sk_buff *skb, struct net_device *netdev)
 /**
  * nes_netdev_start_xmit
  */
-static int nes_netdev_start_xmit(struct sk_buff *skb, struct net_device *netdev)
+static netdev_tx_t nes_netdev_start_xmit(struct sk_buff *skb, struct net_device *netdev)
 {
 	struct nes_vnic *nesvnic = netdev_priv(netdev);
 	struct nes_device *nesdev = nesvnic->nesdev;
-- 
cgit v1.1


From 47a3968a985e1686f41a55b4099fd1b5e16a5969 Mon Sep 17 00:00:00 2001
From: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Date: Tue, 24 Apr 2018 15:15:47 +0200
Subject: IB/ipoib: fix ipoib_start_xmit()'s return type

The method ndo_start_xmit() is defined as returning an 'netdev_tx_t',
which is a typedef for an enum type, but the implementation in this
driver returns an 'int'.

Fix this by returning 'netdev_tx_t' in this driver too.

Signed-off-by: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Reviewed-by: Yuval Shaia <yuval.shaia@oracle.com>
Reviewed-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/ulp/ipoib/ipoib_main.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c
index 161ba8c..cf291f9 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_main.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c
@@ -1094,7 +1094,7 @@ drop_and_unlock:
 	spin_unlock_irqrestore(&priv->lock, flags);
 }
 
-static int ipoib_start_xmit(struct sk_buff *skb, struct net_device *dev)
+static netdev_tx_t ipoib_start_xmit(struct sk_buff *skb, struct net_device *dev)
 {
 	struct ipoib_dev_priv *priv = ipoib_priv(dev);
 	struct rdma_netdev *rn = netdev_priv(dev);
-- 
cgit v1.1


From f96416cea7bce9afe619c15e87fced70f93f9098 Mon Sep 17 00:00:00 2001
From: Colin Ian King <colin.king@canonical.com>
Date: Wed, 25 Apr 2018 17:24:04 +0100
Subject: RDMA/iwpm: fix memory leak on map_info

In the cases where iwpm_hash_bucket is NULL and where function
get_mapinfo_hash_bucket returns NULL then the map_info is never added
to hash_bucket_head and hence there is a leak of map_info. Fix this
by nullifying hash_bucket_head and if that is null we know that
that map_info was not added to hash_bucket_head and hence map_info
should be free'd.

Detected by CoverityScan, CID#1222481 ("Resource Leak")

Fixes: 30dc5e63d6a5 ("RDMA/core: Add support for iWARP Port Mapper user space service")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/iwpm_util.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/infiniband/core/iwpm_util.c b/drivers/infiniband/core/iwpm_util.c
index 9821ae9..da12da1 100644
--- a/drivers/infiniband/core/iwpm_util.c
+++ b/drivers/infiniband/core/iwpm_util.c
@@ -114,7 +114,7 @@ int iwpm_create_mapinfo(struct sockaddr_storage *local_sockaddr,
 			struct sockaddr_storage *mapped_sockaddr,
 			u8 nl_client)
 {
-	struct hlist_head *hash_bucket_head;
+	struct hlist_head *hash_bucket_head = NULL;
 	struct iwpm_mapping_info *map_info;
 	unsigned long flags;
 	int ret = -EINVAL;
@@ -142,6 +142,9 @@ int iwpm_create_mapinfo(struct sockaddr_storage *local_sockaddr,
 		}
 	}
 	spin_unlock_irqrestore(&iwpm_mapinfo_lock, flags);
+
+	if (!hash_bucket_head)
+		kfree(map_info);
 	return ret;
 }
 
-- 
cgit v1.1


From 2da36d44a9d54a2c6e1f8da1f7ccc26b0bc6cfec Mon Sep 17 00:00:00 2001
From: Jianchao Wang <jianchao.w.wang@oracle.com>
Date: Thu, 26 Apr 2018 11:52:39 +0800
Subject: IB/rxe: add RXE_START_MASK for rxe_opcode IB_OPCODE_RC_SEND_ONLY_INV

w/o RXE_START_MASK, the last_psn of IB_OPCODE_RC_SEND_ONLY_INV
will not be updated in update_wqe_psn, and the corresponding
wqe will not be acked in rxe_completer due to its last_psn is
zero. Finally, the other wqe will also not be able to be acked,
because the wqe of IB_OPCODE_RC_SEND_ONLY_INV with last_psn 0
is still there. This causes large amount of io timeout when
nvmeof is over rxe.

Add RXE_START_MASK for IB_OPCODE_RC_SEND_ONLY_INV to fix this.

Signed-off-by: Jianchao Wang <jianchao.w.wang@oracle.com>
Reviewed-by: Zhu Yanjun <yanjun.zhu@oracle.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rxe/rxe_opcode.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/infiniband/sw/rxe/rxe_opcode.c b/drivers/infiniband/sw/rxe/rxe_opcode.c
index 61927c1..4cf1106 100644
--- a/drivers/infiniband/sw/rxe/rxe_opcode.c
+++ b/drivers/infiniband/sw/rxe/rxe_opcode.c
@@ -390,7 +390,7 @@ struct rxe_opcode_info rxe_opcode[RXE_NUM_OPCODE] = {
 		.name	= "IB_OPCODE_RC_SEND_ONLY_INV",
 		.mask	= RXE_IETH_MASK | RXE_PAYLOAD_MASK | RXE_REQ_MASK
 				| RXE_COMP_MASK | RXE_RWR_MASK | RXE_SEND_MASK
-				| RXE_END_MASK,
+				| RXE_END_MASK  | RXE_START_MASK,
 		.length = RXE_BTH_BYTES + RXE_IETH_BYTES,
 		.offset = {
 			[RXE_BTH]	= 0,
-- 
cgit v1.1


From 9fd4350ba8953804f05215999e11a6cfb7b41f2b Mon Sep 17 00:00:00 2001
From: Zhu Yanjun <yanjun.zhu@oracle.com>
Date: Thu, 26 Apr 2018 00:41:10 -0400
Subject: IB/rxe: avoid double kfree_skb

When skb is sent, it will pass the following functions in soft roce.

rxe_send [rdma_rxe]
    ip_local_out
        __ip_local_out
        ip_output
            ip_finish_output
                ip_finish_output2
                    dev_queue_xmit
                        __dev_queue_xmit
                            dev_hard_start_xmit

In the above functions, if error occurs in the above functions or
iptables rules drop skb after ip_local_out, kfree_skb will be called.
So it is not necessary to call kfree_skb in soft roce module again.
Or else crash will occur.

The steps to reproduce:

     server                       client
    ---------                    ---------
    |1.1.1.1|<----rxe-channel--->|1.1.1.2|
    ---------                    ---------

On server: rping -s -a 1.1.1.1 -v -C 10000 -S 512
On client: rping -c -a 1.1.1.1 -v -C 10000 -S 512

The kernel configs CONFIG_DEBUG_KMEMLEAK and
CONFIG_DEBUG_OBJECTS are enabled on both server and client.

When rping runs, run the following command in server:

iptables -I OUTPUT -p udp  --dport 4791 -j DROP

Without this patch, crash will occur.

CC: Srinivas Eeda <srinivas.eeda@oracle.com>
CC: Junxiao Bi <junxiao.bi@oracle.com>
Signed-off-by: Zhu Yanjun <yanjun.zhu@oracle.com>
Reviewed-by: Yuval Shaia <yuval.shaia@oracle.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rxe/rxe_req.c  | 1 -
 drivers/infiniband/sw/rxe/rxe_resp.c | 6 +-----
 2 files changed, 1 insertion(+), 6 deletions(-)

diff --git a/drivers/infiniband/sw/rxe/rxe_req.c b/drivers/infiniband/sw/rxe/rxe_req.c
index 7bdaf71..7851999 100644
--- a/drivers/infiniband/sw/rxe/rxe_req.c
+++ b/drivers/infiniband/sw/rxe/rxe_req.c
@@ -728,7 +728,6 @@ next_wqe:
 		rollback_state(wqe, qp, &rollback_wqe, rollback_psn);
 
 		if (ret == -EAGAIN) {
-			kfree_skb(skb);
 			rxe_run_task(&qp->req.task, 1);
 			goto exit;
 		}
diff --git a/drivers/infiniband/sw/rxe/rxe_resp.c b/drivers/infiniband/sw/rxe/rxe_resp.c
index a65c996..955ff3b 100644
--- a/drivers/infiniband/sw/rxe/rxe_resp.c
+++ b/drivers/infiniband/sw/rxe/rxe_resp.c
@@ -742,7 +742,6 @@ static enum resp_states read_reply(struct rxe_qp *qp,
 	err = rxe_xmit_packet(rxe, qp, &ack_pkt, skb);
 	if (err) {
 		pr_err("Failed sending RDMA reply.\n");
-		kfree_skb(skb);
 		return RESPST_ERR_RNR;
 	}
 
@@ -954,10 +953,8 @@ static int send_ack(struct rxe_qp *qp, struct rxe_pkt_info *pkt,
 	}
 
 	err = rxe_xmit_packet(rxe, qp, &ack_pkt, skb);
-	if (err) {
+	if (err)
 		pr_err_ratelimited("Failed sending ack\n");
-		kfree_skb(skb);
-	}
 
 err1:
 	return err;
@@ -1141,7 +1138,6 @@ static enum resp_states duplicate_request(struct rxe_qp *qp,
 			if (rc) {
 				pr_err("Failed resending result. This flow is not handled - skb ignored\n");
 				rxe_drop_ref(qp);
-				kfree_skb(skb_copy);
 				rc = RESPST_CLEANUP;
 				goto out;
 			}
-- 
cgit v1.1


From 215a8c09e5e2aa6ae1fbcef87f8f27d65d5d1ca4 Mon Sep 17 00:00:00 2001
From: oulijun <oulijun@huawei.com>
Date: Thu, 26 Apr 2018 14:46:15 +0800
Subject: RDMA/hns: Bugfix for init hem table

During init hem table, type should be used instead of
table->type which is finally initializaed with type.

Signed-off-by: Lijun Ou <oulijun@huawei.com>
Signed-off-by: Yixian Liu <liuyixian@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_hem.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/infiniband/hw/hns/hns_roce_hem.c b/drivers/infiniband/hw/hns/hns_roce_hem.c
index 0eeabfb..0d8c113 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hem.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hem.c
@@ -912,7 +912,7 @@ int hns_roce_init_hem_table(struct hns_roce_dev *hr_dev,
 		obj_per_chunk = buf_chunk_size / obj_size;
 		num_hem = (nobj + obj_per_chunk - 1) / obj_per_chunk;
 		bt_chunk_num = bt_chunk_size / 8;
-		if (table->type >= HEM_TYPE_MTT)
+		if (type >= HEM_TYPE_MTT)
 			num_bt_l0 = bt_chunk_num;
 
 		table->hem = kcalloc(num_hem, sizeof(*table->hem),
@@ -920,7 +920,7 @@ int hns_roce_init_hem_table(struct hns_roce_dev *hr_dev,
 		if (!table->hem)
 			goto err_kcalloc_hem_buf;
 
-		if (check_whether_bt_num_3(table->type, hop_num)) {
+		if (check_whether_bt_num_3(type, hop_num)) {
 			unsigned long num_bt_l1;
 
 			num_bt_l1 = (num_hem + bt_chunk_num - 1) /
@@ -939,8 +939,8 @@ int hns_roce_init_hem_table(struct hns_roce_dev *hr_dev,
 				goto err_kcalloc_l1_dma;
 		}
 
-		if (check_whether_bt_num_2(table->type, hop_num) ||
-			check_whether_bt_num_3(table->type, hop_num)) {
+		if (check_whether_bt_num_2(type, hop_num) ||
+			check_whether_bt_num_3(type, hop_num)) {
 			table->bt_l0 = kcalloc(num_bt_l0, sizeof(*table->bt_l0),
 					       GFP_KERNEL);
 			if (!table->bt_l0)
-- 
cgit v1.1


From 328d405b3d4c8dd1f06bfd77f498e23281ae348c Mon Sep 17 00:00:00 2001
From: oulijun <oulijun@huawei.com>
Date: Thu, 26 Apr 2018 14:46:16 +0800
Subject: RDMA/hns: Intercept illegal RDMA operation when use inline data

RDMA read operation is not supported inline data. If user cofigures
issue a RDMA read and use inline data, it will happen a hardware
error.

Signed-off-by: Lijun Ou <oulijun@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
index 8b84ab7..aa5f9b3 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
@@ -71,6 +71,11 @@ static int set_rwqe_data_seg(struct ib_qp *ibqp, struct ib_send_wr *wr,
 			return -EINVAL;
 		}
 
+		if (wr->opcode == IB_WR_RDMA_READ) {
+			dev_err(hr_dev->dev, "Not support inline data!\n");
+			return -EINVAL;
+		}
+
 		for (i = 0; i < wr->num_sge; i++) {
 			memcpy(wqe, ((void *)wr->sg_list[i].addr),
 			       wr->sg_list[i].length);
-- 
cgit v1.1


From 6e1a70943cecdca9bb13b601b1a9772a7bdcc2c3 Mon Sep 17 00:00:00 2001
From: oulijun <oulijun@huawei.com>
Date: Thu, 26 Apr 2018 14:46:17 +0800
Subject: RDMA/hns: Fix the qp context state diagram

According to RoCE protocol, it is possible to
transition from error to error state for modifying
qp in hip08. This patch fix it.

Signed-off-by: Lijun Ou <oulijun@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
index aa5f9b3..a4eea70 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
@@ -3166,7 +3166,8 @@ static int hns_roce_v2_modify_qp(struct ib_qp *ibqp,
 		   (cur_state == IB_QPS_RTR && new_state == IB_QPS_ERR) ||
 		   (cur_state == IB_QPS_RTS && new_state == IB_QPS_ERR) ||
 		   (cur_state == IB_QPS_SQD && new_state == IB_QPS_ERR) ||
-		   (cur_state == IB_QPS_SQE && new_state == IB_QPS_ERR)) {
+		   (cur_state == IB_QPS_SQE && new_state == IB_QPS_ERR) ||
+		   (cur_state == IB_QPS_ERR && new_state == IB_QPS_ERR)) {
 		/* Nothing */
 		;
 	} else {
-- 
cgit v1.1


From 6852af86627c7bd8de11c9ad3eb5cca7d99e5884 Mon Sep 17 00:00:00 2001
From: oulijun <oulijun@huawei.com>
Date: Thu, 26 Apr 2018 14:46:18 +0800
Subject: RDMA/hns: Only assign mtu if IB_QP_PATH_MTU bit is set

Only when the IB_QP_PATH_MTU flag of attr_mask is set
it is valid to assign the mtu field of qp context when
qp type is not GSI and UD.

Signed-off-by: Lijun Ou <oulijun@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
index a4eea70..9a31481 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
@@ -2850,7 +2850,7 @@ static int modify_qp_init_to_rtr(struct ib_qp *ibqp,
 	if (ibqp->qp_type == IB_QPT_GSI || ibqp->qp_type == IB_QPT_UD)
 		roce_set_field(context->byte_24_mtu_tc, V2_QPC_BYTE_24_MTU_M,
 			       V2_QPC_BYTE_24_MTU_S, IB_MTU_4096);
-	else
+	else if (attr_mask & IB_QP_PATH_MTU)
 		roce_set_field(context->byte_24_mtu_tc, V2_QPC_BYTE_24_MTU_M,
 			       V2_QPC_BYTE_24_MTU_S, attr->path_mtu);
 
-- 
cgit v1.1


From 734f38638d7b91c173ddfe46ceadb0382fae9b1f Mon Sep 17 00:00:00 2001
From: oulijun <oulijun@huawei.com>
Date: Thu, 26 Apr 2018 14:46:19 +0800
Subject: RDMA/hns: Remove some unnecessary attr_mask judgement

This patch deletes some unnecessary attr_mask if condition
in hip08 according to the IB protocol.

Signed-off-by: Lijun Ou <oulijun@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 11 ++++-------
 1 file changed, 4 insertions(+), 7 deletions(-)

diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
index 9a31481..97850f4 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
@@ -2655,8 +2655,7 @@ static int modify_qp_init_to_rtr(struct ib_qp *ibqp,
 		return -EINVAL;
 	}
 
-	if ((attr_mask & IB_QP_ALT_PATH) || (attr_mask & IB_QP_ACCESS_FLAGS) ||
-	    (attr_mask & IB_QP_PKEY_INDEX) || (attr_mask & IB_QP_QKEY)) {
+	if (attr_mask & IB_QP_ALT_PATH) {
 		dev_err(dev, "INIT2RTR attr_mask (0x%x) error\n", attr_mask);
 		return -EINVAL;
 	}
@@ -2927,11 +2926,9 @@ static int modify_qp_rtr_to_rts(struct ib_qp *ibqp,
 		return -EINVAL;
 	}
 
-	/* If exist optional param, return error */
-	if ((attr_mask & IB_QP_ALT_PATH) || (attr_mask & IB_QP_ACCESS_FLAGS) ||
-	    (attr_mask & IB_QP_QKEY) || (attr_mask & IB_QP_PATH_MIG_STATE) ||
-	    (attr_mask & IB_QP_CUR_STATE) ||
-	    (attr_mask & IB_QP_MIN_RNR_TIMER)) {
+	/* Not support alternate path and path migration */
+	if ((attr_mask & IB_QP_ALT_PATH) ||
+	    (attr_mask & IB_QP_PATH_MIG_STATE)) {
 		dev_err(dev, "RTR2RTS attr_mask (0x%x)error\n", attr_mask);
 		return -EINVAL;
 	}
-- 
cgit v1.1


From b6dd9b34838506f5307850d8d320d7b1ac6761d3 Mon Sep 17 00:00:00 2001
From: oulijun <oulijun@huawei.com>
Date: Thu, 26 Apr 2018 14:46:20 +0800
Subject: RDMA/hns: Only assign dqpn if IB_QP_PATH_DEST_QPN bit is set

Only when the IB_QP_PATH_DEST_QPN flag of attr_mask is set
is it valid to assign the dqpn field of qp context

Signed-off-by: Lijun Ou <oulijun@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 20 ++++++++++++--------
 1 file changed, 12 insertions(+), 8 deletions(-)

diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
index 97850f4..51ec60a 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
@@ -2597,10 +2597,12 @@ static void modify_qp_init_to_init(struct ib_qp *ibqp,
 	roce_set_field(qpc_mask->byte_4_sqpn_tst, V2_QPC_BYTE_4_SQPN_M,
 		       V2_QPC_BYTE_4_SQPN_S, 0);
 
-	roce_set_field(context->byte_56_dqpn_err, V2_QPC_BYTE_56_DQPN_M,
-		       V2_QPC_BYTE_56_DQPN_S, hr_qp->qpn);
-	roce_set_field(qpc_mask->byte_56_dqpn_err, V2_QPC_BYTE_56_DQPN_M,
-		       V2_QPC_BYTE_56_DQPN_S, 0);
+	if (attr_mask & IB_QP_DEST_QPN) {
+		roce_set_field(context->byte_56_dqpn_err, V2_QPC_BYTE_56_DQPN_M,
+			       V2_QPC_BYTE_56_DQPN_S, hr_qp->qpn);
+		roce_set_field(qpc_mask->byte_56_dqpn_err,
+			       V2_QPC_BYTE_56_DQPN_M, V2_QPC_BYTE_56_DQPN_S, 0);
+	}
 	roce_set_field(context->byte_168_irrl_idx,
 		       V2_QPC_BYTE_168_SQ_SHIFT_BAK_M,
 		       V2_QPC_BYTE_168_SQ_SHIFT_BAK_S,
@@ -2804,10 +2806,12 @@ static int modify_qp_init_to_rtr(struct ib_qp *ibqp,
 			       V2_QPC_BYTE_140_RR_MAX_S, 0);
 	}
 
-	roce_set_field(context->byte_56_dqpn_err, V2_QPC_BYTE_56_DQPN_M,
-		       V2_QPC_BYTE_56_DQPN_S, attr->dest_qp_num);
-	roce_set_field(qpc_mask->byte_56_dqpn_err, V2_QPC_BYTE_56_DQPN_M,
-		       V2_QPC_BYTE_56_DQPN_S, 0);
+	if (attr_mask & IB_QP_DEST_QPN) {
+		roce_set_field(context->byte_56_dqpn_err, V2_QPC_BYTE_56_DQPN_M,
+			       V2_QPC_BYTE_56_DQPN_S, attr->dest_qp_num);
+		roce_set_field(qpc_mask->byte_56_dqpn_err,
+			       V2_QPC_BYTE_56_DQPN_M, V2_QPC_BYTE_56_DQPN_S, 0);
+	}
 
 	/* Configure GID index */
 	port_num = rdma_ah_get_port_num(&attr->ah_attr);
-- 
cgit v1.1


From ae25db00285bf664fbd78222491b29aac6242902 Mon Sep 17 00:00:00 2001
From: oulijun <oulijun@huawei.com>
Date: Thu, 26 Apr 2018 14:46:21 +0800
Subject: RDMA/hns: Adjust the order of cleanup hem table

This patch update the order of cleaning hem table for trrl_table and irrl_table
as well as mtt_cqe_table and mtt_table.

Signed-off-by: Lijun Ou <oulijun@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_hem.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/infiniband/hw/hns/hns_roce_hem.c b/drivers/infiniband/hw/hns/hns_roce_hem.c
index 0d8c113..63b5b3e 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hem.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hem.c
@@ -1039,14 +1039,14 @@ void hns_roce_cleanup_hem_table(struct hns_roce_dev *hr_dev,
 void hns_roce_cleanup_hem(struct hns_roce_dev *hr_dev)
 {
 	hns_roce_cleanup_hem_table(hr_dev, &hr_dev->cq_table.table);
-	hns_roce_cleanup_hem_table(hr_dev, &hr_dev->qp_table.irrl_table);
 	if (hr_dev->caps.trrl_entry_sz)
 		hns_roce_cleanup_hem_table(hr_dev,
 					   &hr_dev->qp_table.trrl_table);
+	hns_roce_cleanup_hem_table(hr_dev, &hr_dev->qp_table.irrl_table);
 	hns_roce_cleanup_hem_table(hr_dev, &hr_dev->qp_table.qp_table);
 	hns_roce_cleanup_hem_table(hr_dev, &hr_dev->mr_table.mtpt_table);
-	hns_roce_cleanup_hem_table(hr_dev, &hr_dev->mr_table.mtt_table);
 	if (hns_roce_check_whether_mhop(hr_dev, HEM_TYPE_CQE))
 		hns_roce_cleanup_hem_table(hr_dev,
 					   &hr_dev->mr_table.mtt_cqe_table);
+	hns_roce_cleanup_hem_table(hr_dev, &hr_dev->mr_table.mtt_table);
 }
-- 
cgit v1.1


From 634f63902277700d0a02eaa101cc685958c35aa4 Mon Sep 17 00:00:00 2001
From: oulijun <oulijun@huawei.com>
Date: Thu, 26 Apr 2018 14:46:22 +0800
Subject: RDMA/hns: Update assignment method for owner field of send wqe

When posting a work reqeust, it need to update the owner bit of send
wqe. This patch mainly fix the bug when posting multiply work
request.

Signed-off-by: Lijun Ou <oulijun@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
index 51ec60a..8ea5470 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
@@ -187,7 +187,8 @@ static int hns_roce_v2_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 		qp->sq.wrid[(qp->sq.head + nreq) & (qp->sq.wqe_cnt - 1)] =
 								      wr->wr_id;
 
-		owner_bit = ~(qp->sq.head >> ilog2(qp->sq.wqe_cnt)) & 0x1;
+		owner_bit =
+		       ~(((qp->sq.head + nreq) >> ilog2(qp->sq.wqe_cnt)) & 0x1);
 
 		/* Corresponding to the QP type, wqe process separately */
 		if (ibqp->qp_type == IB_QPT_GSI) {
-- 
cgit v1.1


From 137ae3208416278aabef3b71e0ea1052940ca362 Mon Sep 17 00:00:00 2001
From: oulijun <oulijun@huawei.com>
Date: Thu, 26 Apr 2018 14:46:23 +0800
Subject: RDMA/hns: Submit bad wr

When generated bad work reqeust, it needs to
report to user. This patch mainly fixes it.

Signed-off-by: Lijun Ou <oulijun@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
index 8ea5470..38a0b6b 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
@@ -153,7 +153,7 @@ static int hns_roce_v2_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 		     ibqp->qp_type != IB_QPT_GSI &&
 		     ibqp->qp_type != IB_QPT_UD)) {
 		dev_err(dev, "Not supported QP(0x%x)type!\n", ibqp->qp_type);
-		*bad_wr = NULL;
+		*bad_wr = wr;
 		return -EOPNOTSUPP;
 	}
 
@@ -462,6 +462,7 @@ static int hns_roce_v2_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 		} else {
 			dev_err(dev, "Illegal qp_type(0x%x)\n", ibqp->qp_type);
 			spin_unlock_irqrestore(&qp->sq.lock, flags);
+			*bad_wr = wr;
 			return -EOPNOTSUPP;
 		}
 	}
-- 
cgit v1.1


From ab17884903f97054c719a3c68017513b922efe43 Mon Sep 17 00:00:00 2001
From: oulijun <oulijun@huawei.com>
Date: Thu, 26 Apr 2018 14:46:24 +0800
Subject: RDMA/hns: Fix a couple misspellings

This patch fixes two spelling errors.

Signed-off-by: Lijun Ou <oulijun@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 2 +-
 drivers/infiniband/hw/hns/hns_roce_qp.c    | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
index 38a0b6b..25916e8 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
@@ -4487,7 +4487,7 @@ static int hns_roce_v2_create_eq(struct hns_roce_dev *hr_dev,
 	ret = hns_roce_cmd_mbox(hr_dev, mailbox->dma, 0, eq->eqn, 0,
 				eq_cmd, HNS_ROCE_CMD_TIMEOUT_MSECS);
 	if (ret) {
-		dev_err(dev, "[mailbox cmd] creat eqc failed.\n");
+		dev_err(dev, "[mailbox cmd] create eqc failed.\n");
 		goto err_cmd_mbox;
 	}
 
diff --git a/drivers/infiniband/hw/hns/hns_roce_qp.c b/drivers/infiniband/hw/hns/hns_roce_qp.c
index e289a92..d4aad34 100644
--- a/drivers/infiniband/hw/hns/hns_roce_qp.c
+++ b/drivers/infiniband/hw/hns/hns_roce_qp.c
@@ -620,7 +620,7 @@ static int hns_roce_create_qp_common(struct hns_roce_dev *hr_dev,
 					to_hr_ucontext(ib_pd->uobject->context),
 					ucmd.db_addr, &hr_qp->rdb);
 			if (ret) {
-				dev_err(dev, "rp record doorbell map failed!\n");
+				dev_err(dev, "rq record doorbell map failed!\n");
 				goto err_mtt;
 			}
 		}
-- 
cgit v1.1


From 4f9ca2d8686ecfdd40ca4f0294a3d94f83f05cea Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leonro@mellanox.com>
Date: Thu, 26 Apr 2018 15:37:48 +0300
Subject: RDMA/mlx4: Add missed RSS hash inner header flag

Despite being advertised to user space application, the RSS inner
header flag was filtered by checks at the beginning of QP creation
routine.

Cc: <stable@vger.kernel.org> # 4.15
Fixes: 4d02ebd9bbbd ("IB/mlx4: Fix RSS hash fields restrictions")
Fixes: 07d84f7b6adf ("IB/mlx4: Add support to RSS hash for inner headers")
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx4/qp.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c
index 50af891..199648a 100644
--- a/drivers/infiniband/hw/mlx4/qp.c
+++ b/drivers/infiniband/hw/mlx4/qp.c
@@ -673,7 +673,8 @@ static int set_qp_rss(struct mlx4_ib_dev *dev, struct mlx4_ib_rss *rss_ctx,
 					  MLX4_IB_RX_HASH_SRC_PORT_TCP	|
 					  MLX4_IB_RX_HASH_DST_PORT_TCP	|
 					  MLX4_IB_RX_HASH_SRC_PORT_UDP	|
-					  MLX4_IB_RX_HASH_DST_PORT_UDP)) {
+					  MLX4_IB_RX_HASH_DST_PORT_UDP  |
+					  MLX4_IB_RX_HASH_INNER)) {
 		pr_debug("RX Hash fields_mask has unsupported mask (0x%llx)\n",
 			 ucmd->rx_hash_fields_mask);
 		return (-EOPNOTSUPP);
-- 
cgit v1.1


From 5ccbf63f87a39c279729670d2e31166844a335c6 Mon Sep 17 00:00:00 2001
From: Ariel Levkovich <lariel@mellanox.com>
Date: Thu, 26 Apr 2018 15:42:54 +0300
Subject: IB/uverbs: Prevent reregistration of DM_MR to regular MR

This patch adds a check in the ib_uverbs_rereg_mr flow to make
sure there's no attempt to rereg a device memory MR to regular MR.
In such case the command will fail with -EINVAL status.

fixes: be934cca9e98 ("IB/uverbs: Add device memory registration ioctl support")
Signed-off-by: Ariel Levkovich <lariel@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/uverbs_cmd.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c
index 13cb5e4..1837924 100644
--- a/drivers/infiniband/core/uverbs_cmd.c
+++ b/drivers/infiniband/core/uverbs_cmd.c
@@ -765,6 +765,11 @@ ssize_t ib_uverbs_rereg_mr(struct ib_uverbs_file *file,
 
 	mr = uobj->object;
 
+	if (mr->dm) {
+		ret = -EINVAL;
+		goto put_uobjs;
+	}
+
 	if (cmd.flags & IB_MR_REREG_ACCESS) {
 		ret = ib_check_mr_access(cmd.access_flags);
 		if (ret)
-- 
cgit v1.1


From 54e7e48b13c85d9a730b989fe7dc5250199a4f81 Mon Sep 17 00:00:00 2001
From: Ariel Levkovich <lariel@mellanox.com>
Date: Thu, 26 Apr 2018 15:42:55 +0300
Subject: IB/uverbs: Fix kernel crash during MR deregistration flow

This patch fixes a crash that happens due to access to an
uninitialized DM pointer within the MR object.

The change makes sure the DM pointer in the MR object is set to
NULL during a non-DM MR creation to prevent a false indication
that this MR is related to a DM in the dereg flow.

Fixes: be934cca9e98 ("IB/uverbs: Add device memory registration ioctl support")
Reported-by: Lijun Ou <oulijun@huawei.com>
Signed-off-by: Ariel Levkovich <lariel@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/uverbs_cmd.c | 1 +
 drivers/infiniband/core/verbs.c      | 1 +
 2 files changed, 2 insertions(+)

diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c
index 1837924..21a887c9 100644
--- a/drivers/infiniband/core/uverbs_cmd.c
+++ b/drivers/infiniband/core/uverbs_cmd.c
@@ -691,6 +691,7 @@ ssize_t ib_uverbs_reg_mr(struct ib_uverbs_file *file,
 
 	mr->device  = pd->device;
 	mr->pd      = pd;
+	mr->dm	    = NULL;
 	mr->uobject = uobj;
 	atomic_inc(&pd->usecnt);
 	mr->res.type = RDMA_RESTRACK_MR;
diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c
index 7eff3ae..6ddfb1f 100644
--- a/drivers/infiniband/core/verbs.c
+++ b/drivers/infiniband/core/verbs.c
@@ -1656,6 +1656,7 @@ struct ib_mr *ib_alloc_mr(struct ib_pd *pd,
 	if (!IS_ERR(mr)) {
 		mr->device  = pd->device;
 		mr->pd      = pd;
+		mr->dm      = NULL;
 		mr->uobject = NULL;
 		atomic_inc(&pd->usecnt);
 		mr->need_inval = false;
-- 
cgit v1.1


From 2df19e19ae90d94fd8724083f161f368a2797537 Mon Sep 17 00:00:00 2001
From: Bharat Potnuri <bharat@chelsio.com>
Date: Fri, 27 Apr 2018 16:41:16 +0530
Subject: iw_cxgb4: Atomically flush per QP HW CQEs

When a CQ is shared by multiple QPs, c4iw_flush_hw_cq() needs to acquire
corresponding QP lock before moving the CQEs into its corresponding SW
queue and accessing the SQ contents for completing a WR.
Ignore CQEs if corresponding QP is already flushed.

Cc: stable@vger.kernel.org
Signed-off-by: Potnuri Bharat Teja <bharat@chelsio.com>
Reviewed-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/cxgb4/cq.c       | 11 ++++++++++-
 drivers/infiniband/hw/cxgb4/iw_cxgb4.h |  2 +-
 drivers/infiniband/hw/cxgb4/qp.c       |  4 ++--
 3 files changed, 13 insertions(+), 4 deletions(-)

diff --git a/drivers/infiniband/hw/cxgb4/cq.c b/drivers/infiniband/hw/cxgb4/cq.c
index 6f2b261..2be2e1a 100644
--- a/drivers/infiniband/hw/cxgb4/cq.c
+++ b/drivers/infiniband/hw/cxgb4/cq.c
@@ -315,7 +315,7 @@ static void advance_oldest_read(struct t4_wq *wq)
  * Deal with out-of-order and/or completions that complete
  * prior unsignalled WRs.
  */
-void c4iw_flush_hw_cq(struct c4iw_cq *chp)
+void c4iw_flush_hw_cq(struct c4iw_cq *chp, struct c4iw_qp *flush_qhp)
 {
 	struct t4_cqe *hw_cqe, *swcqe, read_cqe;
 	struct c4iw_qp *qhp;
@@ -339,6 +339,13 @@ void c4iw_flush_hw_cq(struct c4iw_cq *chp)
 		if (qhp == NULL)
 			goto next_cqe;
 
+		if (flush_qhp != qhp) {
+			spin_lock(&qhp->lock);
+
+			if (qhp->wq.flushed == 1)
+				goto next_cqe;
+		}
+
 		if (CQE_OPCODE(hw_cqe) == FW_RI_TERMINATE)
 			goto next_cqe;
 
@@ -390,6 +397,8 @@ void c4iw_flush_hw_cq(struct c4iw_cq *chp)
 next_cqe:
 		t4_hwcq_consume(&chp->cq);
 		ret = t4_next_hw_cqe(&chp->cq, &hw_cqe);
+		if (qhp && flush_qhp != qhp)
+			spin_unlock(&qhp->lock);
 	}
 }
 
diff --git a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h
index a60def2..8310277 100644
--- a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h
+++ b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h
@@ -1053,7 +1053,7 @@ u32 c4iw_pblpool_alloc(struct c4iw_rdev *rdev, int size);
 void c4iw_pblpool_free(struct c4iw_rdev *rdev, u32 addr, int size);
 u32 c4iw_ocqp_pool_alloc(struct c4iw_rdev *rdev, int size);
 void c4iw_ocqp_pool_free(struct c4iw_rdev *rdev, u32 addr, int size);
-void c4iw_flush_hw_cq(struct c4iw_cq *chp);
+void c4iw_flush_hw_cq(struct c4iw_cq *chp, struct c4iw_qp *flush_qhp);
 void c4iw_count_rcqes(struct t4_cq *cq, struct t4_wq *wq, int *count);
 int c4iw_ep_disconnect(struct c4iw_ep *ep, int abrupt, gfp_t gfp);
 int c4iw_flush_rq(struct t4_wq *wq, struct t4_cq *cq, int count);
diff --git a/drivers/infiniband/hw/cxgb4/qp.c b/drivers/infiniband/hw/cxgb4/qp.c
index de77b60..ae167b6 100644
--- a/drivers/infiniband/hw/cxgb4/qp.c
+++ b/drivers/infiniband/hw/cxgb4/qp.c
@@ -1343,12 +1343,12 @@ static void __flush_qp(struct c4iw_qp *qhp, struct c4iw_cq *rchp,
 	qhp->wq.flushed = 1;
 	t4_set_wq_in_error(&qhp->wq);
 
-	c4iw_flush_hw_cq(rchp);
+	c4iw_flush_hw_cq(rchp, qhp);
 	c4iw_count_rcqes(&rchp->cq, &qhp->wq, &count);
 	rq_flushed = c4iw_flush_rq(&qhp->wq, &rchp->cq, count);
 
 	if (schp != rchp)
-		c4iw_flush_hw_cq(schp);
+		c4iw_flush_hw_cq(schp, qhp);
 	sq_flushed = c4iw_flush_sq(qhp);
 
 	spin_unlock(&qhp->lock);
-- 
cgit v1.1


From 0b21bca04551906485e5d4140ccb2d875c45daa0 Mon Sep 17 00:00:00 2001
From: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Date: Thu, 26 Apr 2018 19:47:35 -0400
Subject: MAINTAINERS: add davem in NETWORKING DRIVERS

"./scripts/get_maintainer.pl -f" does not actually show us David as the
maintainer of drivers/net directories such as team, bonding, phy or dsa.
Adding him in an M: entry of NETWORKING DRIVERS fixes this.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 MAINTAINERS | 1 +
 1 file changed, 1 insertion(+)

diff --git a/MAINTAINERS b/MAINTAINERS
index 92be777..2310341b 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -9733,6 +9733,7 @@ W:	https://fedorahosted.org/dropwatch/
 F:	net/core/drop_monitor.c
 
 NETWORKING DRIVERS
+M:	"David S. Miller" <davem@davemloft.net>
 L:	netdev@vger.kernel.org
 W:	http://www.linuxfoundation.org/en/Net
 Q:	http://patchwork.ozlabs.org/project/netdev/list/
-- 
cgit v1.1


From 077ef1f09c2528b81366ae9a2a969ea35c475027 Mon Sep 17 00:00:00 2001
From: James Ausmus <james.ausmus@intel.com>
Date: Wed, 28 Mar 2018 14:57:56 -0700
Subject: drm/i915/icl: Don't set pipe CSC/Gamma in PLANE_COLOR_CTL
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

These fields have been deprecated and moved in ICL+. Stop setting the
bits.

They have moved to GAMMA_MODE and CSC_MODE, respectively. This patch
is just to stop incorrectly setting bits in PLANE_COLOR_CTL while
we're waiting for the new replacement functionality to be done.

v2: Drop useless comment, and change !(GEN >= 11) to (GEN < 11). (Ville)

v3: No changes

v4 (from Paulo): Rebase.

Cc: Paulo Zanoni <paulo.r.zanoni@intel.com>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: James Ausmus <james.ausmus@intel.com>
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180328215803.13835-2-paulo.r.zanoni@intel.com
---
 drivers/gpu/drm/i915/i915_reg.h      | 4 ++--
 drivers/gpu/drm/i915/intel_display.c | 8 ++++++--
 2 files changed, 8 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 391825a..8c322ff 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -6407,9 +6407,9 @@ enum {
 #define _PLANE_COLOR_CTL_1_A			0x701CC /* GLK+ */
 #define _PLANE_COLOR_CTL_2_A			0x702CC /* GLK+ */
 #define _PLANE_COLOR_CTL_3_A			0x703CC /* GLK+ */
-#define   PLANE_COLOR_PIPE_GAMMA_ENABLE		(1 << 30)
+#define   PLANE_COLOR_PIPE_GAMMA_ENABLE		(1 << 30) /* Pre-ICL */
 #define   PLANE_COLOR_YUV_RANGE_CORRECTION_DISABLE	(1 << 28)
-#define   PLANE_COLOR_PIPE_CSC_ENABLE		(1 << 23)
+#define   PLANE_COLOR_PIPE_CSC_ENABLE		(1 << 23) /* Pre-ICL */
 #define   PLANE_COLOR_CSC_MODE_BYPASS			(0 << 17)
 #define   PLANE_COLOR_CSC_MODE_YUV601_TO_RGB709		(1 << 17)
 #define   PLANE_COLOR_CSC_MODE_YUV709_TO_RGB709		(2 << 17)
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 687e701..efa8822 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -3623,11 +3623,15 @@ u32 skl_plane_ctl(const struct intel_crtc_state *crtc_state,
 u32 glk_plane_color_ctl(const struct intel_crtc_state *crtc_state,
 			const struct intel_plane_state *plane_state)
 {
+	struct drm_i915_private *dev_priv =
+		to_i915(plane_state->base.plane->dev);
 	const struct drm_framebuffer *fb = plane_state->base.fb;
 	u32 plane_color_ctl = 0;
 
-	plane_color_ctl |= PLANE_COLOR_PIPE_GAMMA_ENABLE;
-	plane_color_ctl |= PLANE_COLOR_PIPE_CSC_ENABLE;
+	if (INTEL_GEN(dev_priv) < 11) {
+		plane_color_ctl |= PLANE_COLOR_PIPE_GAMMA_ENABLE;
+		plane_color_ctl |= PLANE_COLOR_PIPE_CSC_ENABLE;
+	}
 	plane_color_ctl |= PLANE_COLOR_PLANE_GAMMA_DISABLE;
 	plane_color_ctl |= glk_plane_color_ctl_alpha(fb->format->format);
 
-- 
cgit v1.1


From 779857e1870178e9f4aee600f17e44d81ff630b2 Mon Sep 17 00:00:00 2001
From: Geert Uytterhoeven <geert+renesas@glider.be>
Date: Mon, 23 Apr 2018 09:32:40 +0200
Subject: dt-bindings: meson-uart: DT fix s/clocks-names/clock-names/

Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Acked-by: Neil Armstrong <narmstrong@baylibre.com>
Signed-off-by: Rob Herring <robh@kernel.org>
---
 Documentation/devicetree/bindings/serial/amlogic,meson-uart.txt | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/Documentation/devicetree/bindings/serial/amlogic,meson-uart.txt b/Documentation/devicetree/bindings/serial/amlogic,meson-uart.txt
index 8ff65fa..c06c045 100644
--- a/Documentation/devicetree/bindings/serial/amlogic,meson-uart.txt
+++ b/Documentation/devicetree/bindings/serial/amlogic,meson-uart.txt
@@ -21,7 +21,7 @@ Required properties:
 - interrupts : identifier to the device interrupt
 - clocks : a list of phandle + clock-specifier pairs, one for each
 	   entry in clock names.
-- clocks-names :
+- clock-names :
    * "xtal" for external xtal clock identifier
    * "pclk" for the bus core clock, either the clk81 clock or the gate clock
    * "baud" for the source of the baudrate generator, can be either the xtal
-- 
cgit v1.1


From a33554401e4746cc33307910a1baad63ce3fd650 Mon Sep 17 00:00:00 2001
From: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Date: Fri, 27 Apr 2018 17:55:41 -0400
Subject: selftests: Fix lib.mk run_tests target shell script

Within run_tests target, the whole script needs to be executed within
the same shell and not as separate subshells, so the initial test_num
variable set to 0 is still present when executing "test_num=`echo
$$test_num+1 | bc`;".

Demonstration of the issue (make run_tests):

TAP version 13
(standard_in) 1: syntax error
selftests: basic_test
========================================
ok 1.. selftests: basic_test [PASS]
(standard_in) 1: syntax error
selftests: basic_percpu_ops_test
========================================
ok 1.. selftests: basic_percpu_ops_test [PASS]
(standard_in) 1: syntax error
selftests: param_test
========================================
ok 1.. selftests: param_test [PASS]

With fix applied:

TAP version 13
selftests: basic_test
========================================
ok 1..1 selftests: basic_test [PASS]
selftests: basic_percpu_ops_test
========================================
ok 1..2 selftests: basic_percpu_ops_test [PASS]
selftests: param_test
========================================
ok 1..3 selftests: param_test [PASS]

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Fixes: 1f87c7c15d7 ("selftests: lib.mk: change RUN_TESTS to print messages in TAP13 format")
CC: Shuah Khan <shuahkh@osg.samsung.com>
CC: linux-kselftest@vger.kernel.org
Signed-off-by: Shuah Khan (Samsung OSG) <shuah@kernel.org>
---
 tools/testing/selftests/lib.mk | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/tools/testing/selftests/lib.mk b/tools/testing/selftests/lib.mk
index 195e9d4..c1b1a4d 100644
--- a/tools/testing/selftests/lib.mk
+++ b/tools/testing/selftests/lib.mk
@@ -20,10 +20,10 @@ all: $(TEST_GEN_PROGS) $(TEST_GEN_PROGS_EXTENDED) $(TEST_GEN_FILES)
 
 .ONESHELL:
 define RUN_TESTS
-	@export KSFT_TAP_LEVEL=`echo 1`;
-	@test_num=`echo 0`;
-	@echo "TAP version 13";
-	@for TEST in $(1); do				\
+	@export KSFT_TAP_LEVEL=`echo 1`;		\
+	test_num=`echo 0`;				\
+	echo "TAP version 13";				\
+	for TEST in $(1); do				\
 		BASENAME_TEST=`basename $$TEST`;	\
 		test_num=`echo $$test_num+1 | bc`;	\
 		echo "selftests: $$BASENAME_TEST";	\
-- 
cgit v1.1


From 988bf7243e03ef69238381594e0334a79cef74a6 Mon Sep 17 00:00:00 2001
From: Lance Richardson <lance.richardson.net@gmail.com>
Date: Wed, 25 Apr 2018 10:21:54 -0400
Subject: net: support compat 64-bit time in {s,g}etsockopt

For the x32 ABI, struct timeval has two 64-bit fields. However
the kernel currently interprets the user-space values used for
the SO_RCVTIMEO and SO_SNDTIMEO socket options as having a pair
of 32-bit fields.

When the seconds portion of the requested timeout is less than 2**32,
the seconds portion of the effective timeout is correct but the
microseconds portion is zero.  When the seconds portion of the
requested timeout is zero and the microseconds portion is non-zero,
the kernel interprets the timeout as zero (never timeout).

Fix by using 64-bit time for SO_RCVTIMEO/SO_SNDTIMEO as required
for the ABI.

The code included below demonstrates the problem.

Results before patch:
    $ gcc -m64 -Wall -O2 -o socktmo socktmo.c && ./socktmo
    recv time: 2.008181 seconds
    send time: 2.015985 seconds

    $ gcc -m32 -Wall -O2 -o socktmo socktmo.c && ./socktmo
    recv time: 2.016763 seconds
    send time: 2.016062 seconds

    $ gcc -mx32 -Wall -O2 -o socktmo socktmo.c && ./socktmo
    recv time: 1.007239 seconds
    send time: 1.023890 seconds

Results after patch:
    $ gcc -m64 -O2 -Wall -o socktmo socktmo.c && ./socktmo
    recv time: 2.010062 seconds
    send time: 2.015836 seconds

    $ gcc -m32 -O2 -Wall -o socktmo socktmo.c && ./socktmo
    recv time: 2.013974 seconds
    send time: 2.015981 seconds

    $ gcc -mx32 -O2 -Wall -o socktmo socktmo.c && ./socktmo
    recv time: 2.030257 seconds
    send time: 2.013383 seconds

 #include <stdio.h>
 #include <stdlib.h>
 #include <sys/socket.h>
 #include <sys/types.h>
 #include <sys/time.h>

 void checkrc(char *str, int rc)
 {
         if (rc >= 0)
                 return;

         perror(str);
         exit(1);
 }

 static char buf[1024];
 int main(int argc, char **argv)
 {
         int rc;
         int socks[2];
         struct timeval tv;
         struct timeval start, end, delta;

         rc = socketpair(AF_UNIX, SOCK_STREAM, 0, socks);
         checkrc("socketpair", rc);

         /* set timeout to 1.999999 seconds */
         tv.tv_sec = 1;
         tv.tv_usec = 999999;
         rc = setsockopt(socks[0], SOL_SOCKET, SO_RCVTIMEO, &tv, sizeof tv);
         rc = setsockopt(socks[0], SOL_SOCKET, SO_SNDTIMEO, &tv, sizeof tv);
         checkrc("setsockopt", rc);

         /* measure actual receive timeout */
         gettimeofday(&start, NULL);
         rc = recv(socks[0], buf, sizeof buf, 0);
         gettimeofday(&end, NULL);
         timersub(&end, &start, &delta);

         printf("recv time: %ld.%06ld seconds\n",
                (long)delta.tv_sec, (long)delta.tv_usec);

         /* fill send buffer */
         do {
                 rc = send(socks[0], buf, sizeof buf, 0);
         } while (rc > 0);

         /* measure actual send timeout */
         gettimeofday(&start, NULL);
         rc = send(socks[0], buf, sizeof buf, 0);
         gettimeofday(&end, NULL);
         timersub(&end, &start, &delta);

         printf("send time: %ld.%06ld seconds\n",
                (long)delta.tv_sec, (long)delta.tv_usec);
         exit(0);
 }

Fixes: 515c7af85ed9 ("x32: Use compat shims for {g,s}etsockopt")
Reported-by: Gopal RajagopalSai <gopalsr83@gmail.com>
Signed-off-by: Lance Richardson <lance.richardson.net@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/compat.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/net/compat.c b/net/compat.c
index 5ae7437d..7242cce 100644
--- a/net/compat.c
+++ b/net/compat.c
@@ -377,7 +377,8 @@ static int compat_sock_setsockopt(struct socket *sock, int level, int optname,
 	    optname == SO_ATTACH_REUSEPORT_CBPF)
 		return do_set_attach_filter(sock, level, optname,
 					    optval, optlen);
-	if (optname == SO_RCVTIMEO || optname == SO_SNDTIMEO)
+	if (!COMPAT_USE_64BIT_TIME &&
+	    (optname == SO_RCVTIMEO || optname == SO_SNDTIMEO))
 		return do_set_sock_timeout(sock, level, optname, optval, optlen);
 
 	return sock_setsockopt(sock, level, optname, optval, optlen);
@@ -448,7 +449,8 @@ static int do_get_sock_timeout(struct socket *sock, int level, int optname,
 static int compat_sock_getsockopt(struct socket *sock, int level, int optname,
 				char __user *optval, int __user *optlen)
 {
-	if (optname == SO_RCVTIMEO || optname == SO_SNDTIMEO)
+	if (!COMPAT_USE_64BIT_TIME &&
+	    (optname == SO_RCVTIMEO || optname == SO_SNDTIMEO))
 		return do_get_sock_timeout(sock, level, optname, optval, optlen);
 	return sock_getsockopt(sock, level, optname, optval, optlen);
 }
-- 
cgit v1.1


From 74bd8004e475d67eb41f6795cda5efac03d010b8 Mon Sep 17 00:00:00 2001
From: Mahesh Kumar <mahesh1.kumar@intel.com>
Date: Thu, 26 Apr 2018 19:55:15 +0530
Subject: drm/i915/icl: track dbuf slice-2 status

This patch adds support to start tracking status of DBUF slices.
This is foundation to introduce support for enabling/disabling second
DBUF slice dynamically for ICL.

Changes Since V1:
 - use kernel type u8 over uint8_t

Signed-off-by: Mahesh Kumar <mahesh1.kumar@intel.com>
Reviewed-by: James Ausmus <james.ausmus@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180426142517.16643-2-mahesh1.kumar@intel.com
---
 drivers/gpu/drm/i915/i915_drv.h         |  1 +
 drivers/gpu/drm/i915/intel_display.c    |  5 +++++
 drivers/gpu/drm/i915/intel_pm.c         | 20 ++++++++++++++++++++
 drivers/gpu/drm/i915/intel_runtime_pm.c |  4 ++++
 4 files changed, 30 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 8444ca8..193176b 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -1189,6 +1189,7 @@ struct skl_ddb_allocation {
 	/* packed/y */
 	struct skl_ddb_entry plane[I915_MAX_PIPES][I915_MAX_PLANES];
 	struct skl_ddb_entry uv_plane[I915_MAX_PIPES][I915_MAX_PLANES];
+	u8 enabled_slices; /* GEN11 has configurable 2 slices */
 };
 
 struct skl_ddb_values {
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index efa8822..338570e 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -11447,6 +11447,11 @@ static void verify_wm_state(struct drm_crtc *crtc,
 	skl_ddb_get_hw_state(dev_priv, &hw_ddb);
 	sw_ddb = &dev_priv->wm.skl_hw.ddb;
 
+	if (INTEL_GEN(dev_priv) >= 11)
+		if (hw_ddb.enabled_slices != sw_ddb->enabled_slices)
+			DRM_ERROR("mismatch in DBUF Slices (expected %u, got %u)\n",
+				  sw_ddb->enabled_slices,
+				  hw_ddb.enabled_slices);
 	/* planes */
 	for_each_universal_plane(dev_priv, pipe, plane) {
 		hw_plane_wm = &hw_wm.planes[plane];
diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index 4baab85..a29e6d5 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -3567,6 +3567,23 @@ bool ilk_disable_lp_wm(struct drm_device *dev)
 	return _ilk_disable_lp_wm(dev_priv, WM_DIRTY_LP_ALL);
 }
 
+static u8 intel_enabled_dbuf_slices_num(struct drm_i915_private *dev_priv)
+{
+	u8 enabled_slices;
+
+	/* Slice 1 will always be enabled */
+	enabled_slices = 1;
+
+	/* Gen prior to GEN11 have only one DBuf slice */
+	if (INTEL_GEN(dev_priv) < 11)
+		return enabled_slices;
+
+	if (I915_READ(DBUF_CTL_S2) & DBUF_POWER_STATE)
+		enabled_slices++;
+
+	return enabled_slices;
+}
+
 /*
  * FIXME: We still don't have the proper code detect if we need to apply the WA,
  * so assume we'll always need it in order to avoid underruns.
@@ -3870,6 +3887,8 @@ void skl_ddb_get_hw_state(struct drm_i915_private *dev_priv,
 
 	memset(ddb, 0, sizeof(*ddb));
 
+	ddb->enabled_slices = intel_enabled_dbuf_slices_num(dev_priv);
+
 	for_each_intel_crtc(&dev_priv->drm, crtc) {
 		enum intel_display_power_domain power_domain;
 		enum plane_id plane_id;
@@ -5088,6 +5107,7 @@ skl_copy_ddb_for_pipe(struct skl_ddb_values *dst,
 	       sizeof(dst->ddb.uv_plane[pipe]));
 	memcpy(dst->ddb.plane[pipe], src->ddb.plane[pipe],
 	       sizeof(dst->ddb.plane[pipe]));
+	dst->ddb.enabled_slices = src->ddb.enabled_slices;
 }
 
 static void
diff --git a/drivers/gpu/drm/i915/intel_runtime_pm.c b/drivers/gpu/drm/i915/intel_runtime_pm.c
index ec59992..afc6ef8 100644
--- a/drivers/gpu/drm/i915/intel_runtime_pm.c
+++ b/drivers/gpu/drm/i915/intel_runtime_pm.c
@@ -2656,6 +2656,8 @@ static void icl_dbuf_enable(struct drm_i915_private *dev_priv)
 	if (!(I915_READ(DBUF_CTL_S1) & DBUF_POWER_STATE) ||
 	    !(I915_READ(DBUF_CTL_S2) & DBUF_POWER_STATE))
 		DRM_ERROR("DBuf power enable timeout\n");
+	else
+		dev_priv->wm.skl_hw.ddb.enabled_slices = 2;
 }
 
 static void icl_dbuf_disable(struct drm_i915_private *dev_priv)
@@ -2669,6 +2671,8 @@ static void icl_dbuf_disable(struct drm_i915_private *dev_priv)
 	if ((I915_READ(DBUF_CTL_S1) & DBUF_POWER_STATE) ||
 	    (I915_READ(DBUF_CTL_S2) & DBUF_POWER_STATE))
 		DRM_ERROR("DBuf power disable timeout!\n");
+	else
+		dev_priv->wm.skl_hw.ddb.enabled_slices = 0;
 }
 
 static void icl_mbus_init(struct drm_i915_private *dev_priv)
-- 
cgit v1.1


From aa9664ffe863f470efdbe40ea20ce96f2887ebcd Mon Sep 17 00:00:00 2001
From: Mahesh Kumar <mahesh1.kumar@intel.com>
Date: Thu, 26 Apr 2018 19:55:16 +0530
Subject: drm/i915/icl: Enable 2nd DBuf slice only when needed

ICL has two slices of DBuf, each slice of size 1024 blocks.
We should not always enable slice-2. It should be enabled only if
display total required BW is > 12GBps OR more than 1 pipes are enabled.

Changes since V1:
 - typecast total_data_rate to u64 before multiplication to solve any
   possible overflow (Rodrigo)
 - fix where skl_wm_get_hw_state was memsetting ddb, resulting
   enabled_slices to become zero
 - Fix the logic of calculating ddb_size
Changes since V2:
 - If no-crtc is part of commit required_slices will have value "0",
   don't try to disable DBuf slice.
Changes since V3:
 - Create a generic helper to enable/disable slice
 - don't return early if total_data_rate is 0, it may be cursor only
   commit, or atomic modeset without any plane.
Changes since V4:
 - Solve checkpatch warnings
 - use kernel types u8/u64 instead of uint8_t/uint64_t
Changes since V5:
 - Rebase

Signed-off-by: Mahesh Kumar <mahesh1.kumar@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180426142517.16643-3-mahesh1.kumar@intel.com
---
 drivers/gpu/drm/i915/intel_display.c    | 10 +++++
 drivers/gpu/drm/i915/intel_drv.h        |  6 +++
 drivers/gpu/drm/i915/intel_pm.c         | 57 +++++++++++++++++++++++------
 drivers/gpu/drm/i915/intel_runtime_pm.c | 65 ++++++++++++++++++++++++++-------
 4 files changed, 113 insertions(+), 25 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 338570e..84ce66b 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -12258,6 +12258,8 @@ static void skl_update_crtcs(struct drm_atomic_state *state)
 	bool progress;
 	enum pipe pipe;
 	int i;
+	u8 hw_enabled_slices = dev_priv->wm.skl_hw.ddb.enabled_slices;
+	u8 required_slices = intel_state->wm_results.ddb.enabled_slices;
 
 	const struct skl_ddb_entry *entries[I915_MAX_PIPES] = {};
 
@@ -12266,6 +12268,10 @@ static void skl_update_crtcs(struct drm_atomic_state *state)
 		if (new_crtc_state->active)
 			entries[i] = &to_intel_crtc_state(old_crtc_state)->wm.skl.ddb;
 
+	/* If 2nd DBuf slice required, enable it here */
+	if (INTEL_GEN(dev_priv) >= 11 && required_slices > hw_enabled_slices)
+		icl_dbuf_slices_update(dev_priv, required_slices);
+
 	/*
 	 * Whenever the number of active pipes changes, we need to make sure we
 	 * update the pipes in the right order so that their ddb allocations
@@ -12316,6 +12322,10 @@ static void skl_update_crtcs(struct drm_atomic_state *state)
 			progress = true;
 		}
 	} while (progress);
+
+	/* If 2nd DBuf slice is no more required disable it */
+	if (INTEL_GEN(dev_priv) >= 11 && required_slices < hw_enabled_slices)
+		icl_dbuf_slices_update(dev_priv, required_slices);
 }
 
 static void intel_atomic_helper_free_state(struct drm_i915_private *dev_priv)
diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
index 9bba035..11a1932 100644
--- a/drivers/gpu/drm/i915/intel_drv.h
+++ b/drivers/gpu/drm/i915/intel_drv.h
@@ -144,6 +144,10 @@
 #define KHz(x) (1000 * (x))
 #define MHz(x) KHz(1000 * (x))
 
+#define KBps(x) (1000 * (x))
+#define MBps(x) KBps(1000 * (x))
+#define GBps(x) ((u64)1000 * MBps((x)))
+
 /*
  * Display related stuff
  */
@@ -1931,6 +1935,8 @@ bool intel_display_power_get_if_enabled(struct drm_i915_private *dev_priv,
 					enum intel_display_power_domain domain);
 void intel_display_power_put(struct drm_i915_private *dev_priv,
 			     enum intel_display_power_domain domain);
+void icl_dbuf_slices_update(struct drm_i915_private *dev_priv,
+			    u8 req_slices);
 
 static inline void
 assert_rpm_device_not_suspended(struct drm_i915_private *dev_priv)
diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index a29e6d5..3e72e9e 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -3771,9 +3771,42 @@ bool intel_can_enable_sagv(struct drm_atomic_state *state)
 	return true;
 }
 
+static unsigned int intel_get_ddb_size(struct drm_i915_private *dev_priv,
+				       const struct intel_crtc_state *cstate,
+				       const unsigned int total_data_rate,
+				       const int num_active,
+				       struct skl_ddb_allocation *ddb)
+{
+	const struct drm_display_mode *adjusted_mode;
+	u64 total_data_bw;
+	u16 ddb_size = INTEL_INFO(dev_priv)->ddb_size;
+
+	WARN_ON(ddb_size == 0);
+
+	if (INTEL_GEN(dev_priv) < 11)
+		return ddb_size - 4; /* 4 blocks for bypass path allocation */
+
+	adjusted_mode = &cstate->base.adjusted_mode;
+	total_data_bw = (u64)total_data_rate * drm_mode_vrefresh(adjusted_mode);
+
+	/*
+	 * 12GB/s is maximum BW supported by single DBuf slice.
+	 */
+	if (total_data_bw >= GBps(12) || num_active > 1) {
+		ddb->enabled_slices = 2;
+	} else {
+		ddb->enabled_slices = 1;
+		ddb_size /= 2;
+	}
+
+	return ddb_size;
+}
+
 static void
 skl_ddb_get_pipe_allocation_limits(struct drm_device *dev,
 				   const struct intel_crtc_state *cstate,
+				   const unsigned int total_data_rate,
+				   struct skl_ddb_allocation *ddb,
 				   struct skl_ddb_entry *alloc, /* out */
 				   int *num_active /* out */)
 {
@@ -3796,11 +3829,8 @@ skl_ddb_get_pipe_allocation_limits(struct drm_device *dev,
 	else
 		*num_active = hweight32(dev_priv->active_crtcs);
 
-	ddb_size = INTEL_INFO(dev_priv)->ddb_size;
-	WARN_ON(ddb_size == 0);
-
-	if (INTEL_GEN(dev_priv) < 11)
-		ddb_size -= 4; /* 4 blocks for bypass path allocation */
+	ddb_size = intel_get_ddb_size(dev_priv, cstate, total_data_rate,
+				      *num_active, ddb);
 
 	/*
 	 * If the state doesn't change the active CRTC's, then there's
@@ -4261,7 +4291,11 @@ skl_allocate_pipe_ddb(struct intel_crtc_state *cstate,
 		return 0;
 	}
 
-	skl_ddb_get_pipe_allocation_limits(dev, cstate, alloc, &num_active);
+	total_data_rate = skl_get_total_relative_data_rate(cstate,
+							   plane_data_rate,
+							   uv_plane_data_rate);
+	skl_ddb_get_pipe_allocation_limits(dev, cstate, total_data_rate, ddb,
+					   alloc, &num_active);
 	alloc_size = skl_ddb_entry_size(alloc);
 	if (alloc_size == 0)
 		return 0;
@@ -4296,9 +4330,6 @@ skl_allocate_pipe_ddb(struct intel_crtc_state *cstate,
 	 *
 	 * FIXME: we may not allocate every single block here.
 	 */
-	total_data_rate = skl_get_total_relative_data_rate(cstate,
-							   plane_data_rate,
-							   uv_plane_data_rate);
 	if (total_data_rate == 0)
 		return 0;
 
@@ -5492,8 +5523,12 @@ void skl_wm_get_hw_state(struct drm_device *dev)
 		/* Fully recompute DDB on first atomic commit */
 		dev_priv->wm.distrust_bios_wm = true;
 	} else {
-		/* Easy/common case; just sanitize DDB now if everything off */
-		memset(ddb, 0, sizeof(*ddb));
+		/*
+		 * Easy/common case; just sanitize DDB now if everything off
+		 * Keep dbuf slice info intact
+		 */
+		memset(ddb->plane, 0, sizeof(ddb->plane));
+		memset(ddb->uv_plane, 0, sizeof(ddb->uv_plane));
 	}
 }
 
diff --git a/drivers/gpu/drm/i915/intel_runtime_pm.c b/drivers/gpu/drm/i915/intel_runtime_pm.c
index afc6ef8..3fffbfe 100644
--- a/drivers/gpu/drm/i915/intel_runtime_pm.c
+++ b/drivers/gpu/drm/i915/intel_runtime_pm.c
@@ -2619,32 +2619,69 @@ static void intel_power_domains_sync_hw(struct drm_i915_private *dev_priv)
 	mutex_unlock(&power_domains->lock);
 }
 
-static void gen9_dbuf_enable(struct drm_i915_private *dev_priv)
+static inline
+bool intel_dbuf_slice_set(struct drm_i915_private *dev_priv,
+			  i915_reg_t reg, bool enable)
 {
-	I915_WRITE(DBUF_CTL, I915_READ(DBUF_CTL) | DBUF_POWER_REQUEST);
-	POSTING_READ(DBUF_CTL);
+	u32 val, status;
 
+	val = I915_READ(reg);
+	val = enable ? (val | DBUF_POWER_REQUEST) : (val & ~DBUF_POWER_REQUEST);
+	I915_WRITE(reg, val);
+	POSTING_READ(reg);
 	udelay(10);
 
-	if (!(I915_READ(DBUF_CTL) & DBUF_POWER_STATE))
-		DRM_ERROR("DBuf power enable timeout\n");
+	status = I915_READ(reg) & DBUF_POWER_STATE;
+	if ((enable && !status) || (!enable && status)) {
+		DRM_ERROR("DBus power %s timeout!\n",
+			  enable ? "enable" : "disable");
+		return false;
+	}
+	return true;
+}
+
+static void gen9_dbuf_enable(struct drm_i915_private *dev_priv)
+{
+	intel_dbuf_slice_set(dev_priv, DBUF_CTL, true);
 }
 
 static void gen9_dbuf_disable(struct drm_i915_private *dev_priv)
 {
-	I915_WRITE(DBUF_CTL, I915_READ(DBUF_CTL) & ~DBUF_POWER_REQUEST);
-	POSTING_READ(DBUF_CTL);
+	intel_dbuf_slice_set(dev_priv, DBUF_CTL, false);
+}
 
-	udelay(10);
+static u8 intel_dbuf_max_slices(struct drm_i915_private *dev_priv)
+{
+	if (INTEL_GEN(dev_priv) < 11)
+		return 1;
+	return 2;
+}
 
-	if (I915_READ(DBUF_CTL) & DBUF_POWER_STATE)
-		DRM_ERROR("DBuf power disable timeout!\n");
+void icl_dbuf_slices_update(struct drm_i915_private *dev_priv,
+			    u8 req_slices)
+{
+	u8 hw_enabled_slices = dev_priv->wm.skl_hw.ddb.enabled_slices;
+	u32 val;
+	bool ret;
+
+	if (req_slices > intel_dbuf_max_slices(dev_priv)) {
+		DRM_ERROR("Invalid number of dbuf slices requested\n");
+		return;
+	}
+
+	if (req_slices == hw_enabled_slices || req_slices == 0)
+		return;
+
+	val = I915_READ(DBUF_CTL_S2);
+	if (req_slices > hw_enabled_slices)
+		ret = intel_dbuf_slice_set(dev_priv, DBUF_CTL_S2, true);
+	else
+		ret = intel_dbuf_slice_set(dev_priv, DBUF_CTL_S2, false);
+
+	if (ret)
+		dev_priv->wm.skl_hw.ddb.enabled_slices = req_slices;
 }
 
-/*
- * TODO: we shouldn't always enable DBUF_CTL_S2, we should only enable it when
- * needed and keep it disabled as much as possible.
- */
 static void icl_dbuf_enable(struct drm_i915_private *dev_priv)
 {
 	I915_WRITE(DBUF_CTL_S1, I915_READ(DBUF_CTL_S1) | DBUF_POWER_REQUEST);
-- 
cgit v1.1


From 37cde11ba720cc485bbc784e9a34878d40a34e96 Mon Sep 17 00:00:00 2001
From: Mahesh Kumar <mahesh1.kumar@intel.com>
Date: Thu, 26 Apr 2018 19:55:17 +0530
Subject: drm/i915/icl: update ddb entry start/end mask during hw ddb readout

Gen11/ICL onward ddb entry start/end mask is increased from 10 bits to
11 bits. This patch make changes to use proper mask for ICL+ during
hardware ddb value readout.

Changes since V1:
 - Use _MASK & _SHIFT macro (James)
Changes since V2:
 - use kernel type u8 instead of uint8_t
Changes since V3:
 - Rebase

Signed-off-by: Mahesh Kumar <mahesh1.kumar@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180426142517.16643-4-mahesh1.kumar@intel.com
---
 drivers/gpu/drm/i915/i915_reg.h |  3 +++
 drivers/gpu/drm/i915/intel_pm.c | 26 +++++++++++++++++++-------
 2 files changed, 22 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 8c322ff..115d7be 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -6504,6 +6504,9 @@ enum {
 
 #define _PLANE_BUF_CFG_1_B			0x7127c
 #define _PLANE_BUF_CFG_2_B			0x7137c
+#define  SKL_DDB_ENTRY_MASK			0x3FF
+#define  ICL_DDB_ENTRY_MASK			0x7FF
+#define  DDB_ENTRY_END_SHIFT			16
 #define _PLANE_BUF_CFG_1(pipe)	\
 	_PIPE(pipe, _PLANE_BUF_CFG_1_A, _PLANE_BUF_CFG_1_B)
 #define _PLANE_BUF_CFG_2(pipe)	\
diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index 3e72e9e..4126132 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -3864,10 +3864,18 @@ static unsigned int skl_cursor_allocation(int num_active)
 	return 8;
 }
 
-static void skl_ddb_entry_init_from_hw(struct skl_ddb_entry *entry, u32 reg)
+static void skl_ddb_entry_init_from_hw(struct drm_i915_private *dev_priv,
+				       struct skl_ddb_entry *entry, u32 reg)
 {
-	entry->start = reg & 0x3ff;
-	entry->end = (reg >> 16) & 0x3ff;
+	u16 mask;
+
+	if (INTEL_GEN(dev_priv) >= 11)
+		mask = ICL_DDB_ENTRY_MASK;
+	else
+		mask = SKL_DDB_ENTRY_MASK;
+	entry->start = reg & mask;
+	entry->end = (reg >> DDB_ENTRY_END_SHIFT) & mask;
+
 	if (entry->end)
 		entry->end += 1;
 }
@@ -3884,7 +3892,8 @@ skl_ddb_get_hw_plane_state(struct drm_i915_private *dev_priv,
 	/* Cursor doesn't support NV12/planar, so no extra calculation needed */
 	if (plane_id == PLANE_CURSOR) {
 		val = I915_READ(CUR_BUF_CFG(pipe));
-		skl_ddb_entry_init_from_hw(&ddb->plane[pipe][plane_id], val);
+		skl_ddb_entry_init_from_hw(dev_priv,
+					   &ddb->plane[pipe][plane_id], val);
 		return;
 	}
 
@@ -3903,10 +3912,13 @@ skl_ddb_get_hw_plane_state(struct drm_i915_private *dev_priv,
 	val2 = I915_READ(PLANE_NV12_BUF_CFG(pipe, plane_id));
 
 	if (fourcc == DRM_FORMAT_NV12) {
-		skl_ddb_entry_init_from_hw(&ddb->plane[pipe][plane_id], val2);
-		skl_ddb_entry_init_from_hw(&ddb->uv_plane[pipe][plane_id], val);
+		skl_ddb_entry_init_from_hw(dev_priv,
+					   &ddb->plane[pipe][plane_id], val2);
+		skl_ddb_entry_init_from_hw(dev_priv,
+					   &ddb->uv_plane[pipe][plane_id], val);
 	} else {
-		skl_ddb_entry_init_from_hw(&ddb->plane[pipe][plane_id], val);
+		skl_ddb_entry_init_from_hw(dev_priv,
+					   &ddb->plane[pipe][plane_id], val);
 	}
 }
 
-- 
cgit v1.1


From ded8b9c761c50d147b20d7de18766fcdb8f5b621 Mon Sep 17 00:00:00 2001
From: Edward Cree <ecree@solarflare.com>
Date: Fri, 27 Apr 2018 15:08:41 +0100
Subject: sfc: Use filter index rather than ID for rps_flow_id table

efx->type->filter_insert() returns an ID rather than the index that
 efx->type->filter_async_insert() used to, which causes it to exceed
 efx->type->max_rx_ip_filters on some EF10 configurations, leading to out-
 of-bounds array writes.
So, in efx_filter_rfs_work(), convert this back into an index (which is
 what the remove call in the expiry path expects, anyway).

Fixes: 3af0f34290f6 ("sfc: replace asynchronous filter operations")
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/sfc/rx.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/net/ethernet/sfc/rx.c b/drivers/net/ethernet/sfc/rx.c
index 64a94f2..d2e254f 100644
--- a/drivers/net/ethernet/sfc/rx.c
+++ b/drivers/net/ethernet/sfc/rx.c
@@ -839,6 +839,8 @@ static void efx_filter_rfs_work(struct work_struct *data)
 	int rc;
 
 	rc = efx->type->filter_insert(efx, &req->spec, true);
+	if (rc >= 0)
+		rc %= efx->type->max_rx_ip_filters;
 	if (efx->rps_hash_table) {
 		spin_lock_bh(&efx->rps_hash_lock);
 		rule = efx_rps_hash_find(efx, &req->spec);
-- 
cgit v1.1


From 987c658a61f432804c4662b736dbd5fc5939af1f Mon Sep 17 00:00:00 2001
From: Edward Cree <ecree@solarflare.com>
Date: Fri, 27 Apr 2018 15:08:57 +0100
Subject: sfc: fix ARFS expiry check on EF10

Owing to a missing conditional, the result of rps_may_expire_flow() was
 being ignored and filters were being removed even if we'd decided not to
 expire them.

Fixes: f8d6203780b7 ("sfc: ARFS filter IDs")
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/sfc/ef10.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/sfc/ef10.c b/drivers/net/ethernet/sfc/ef10.c
index 63036d9..d90a7b1 100644
--- a/drivers/net/ethernet/sfc/ef10.c
+++ b/drivers/net/ethernet/sfc/ef10.c
@@ -4784,8 +4784,9 @@ expire:
 	 * will set rule->filter_id to EFX_ARFS_FILTER_ID_PENDING, meaning that
 	 * the rule is not removed by efx_rps_hash_del() below.
 	 */
-	ret = efx_ef10_filter_remove_internal(efx, 1U << spec->priority,
-					      filter_idx, true) == 0;
+	if (ret)
+		ret = efx_ef10_filter_remove_internal(efx, 1U << spec->priority,
+						      filter_idx, true) == 0;
 	/* While we can't safely dereference rule (we dropped the lock), we can
 	 * still test it for NULL.
 	 */
-- 
cgit v1.1


From 78b60ce7b96cf1869b51cee916a40041e400d6ce Mon Sep 17 00:00:00 2001
From: Paulo Zanoni <paulo.r.zanoni@intel.com>
Date: Wed, 28 Mar 2018 14:57:57 -0700
Subject: drm/i915/icl: add definitions for the ICL PLL registers

There's a lot of code for the PLL enabling, so let's first only
introduce the register definitions in order to make patch reviewing a
little easier.

v2: Coding style (Jani).
v3: Preparation for upstreaming.
v4: Fix MG_CLKTOP2_CORECLKCTL1 address and random typos (James).

Cc: James Ausmus <james.ausmus@intel.com>
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: James Ausmus <james.ausmus@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180328215803.13835-3-paulo.r.zanoni@intel.com
---
 drivers/gpu/drm/i915/i915_reg.h | 149 ++++++++++++++++++++++++++++++++++++++++
 1 file changed, 149 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 115d7be..197c966 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -8759,6 +8759,12 @@ enum skl_power_gate {
 #define  PORT_CLK_SEL_NONE		(7<<29)
 #define  PORT_CLK_SEL_MASK		(7<<29)
 
+/* On ICL+ this is the same as PORT_CLK_SEL, but all bits change. */
+#define DDI_CLK_SEL(port)		PORT_CLK_SEL(port)
+#define  DDI_CLK_SEL_NONE		(0x0 << 28)
+#define  DDI_CLK_SEL_MG			(0x8 << 28)
+#define  DDI_CLK_SEL_MASK		(0xF << 28)
+
 /* Transcoder clock selection */
 #define _TRANS_CLK_SEL_A		0x46140
 #define _TRANS_CLK_SEL_B		0x46144
@@ -8889,6 +8895,7 @@ enum skl_power_gate {
  * CNL Clocks
  */
 #define DPCLKA_CFGCR0				_MMIO(0x6C200)
+#define DPCLKA_CFGCR0_ICL			_MMIO(0x164280)
 #define  DPCLKA_CFGCR0_DDI_CLK_OFF(port)	(1 << ((port) ==  PORT_F ? 23 : \
 						      (port)+10))
 #define  DPCLKA_CFGCR0_DDI_CLK_SEL_SHIFT(port)	((port) == PORT_F ? 21 : \
@@ -8905,10 +8912,141 @@ enum skl_power_gate {
 #define  PLL_POWER_STATE	(1 << 26)
 #define CNL_DPLL_ENABLE(pll)	_MMIO_PLL(pll, DPLL0_ENABLE, DPLL1_ENABLE)
 
+#define _MG_PLL1_ENABLE		0x46030
+#define _MG_PLL2_ENABLE		0x46034
+#define _MG_PLL3_ENABLE		0x46038
+#define _MG_PLL4_ENABLE		0x4603C
+/* Bits are the same as DPLL0_ENABLE */
+#define MG_PLL_ENABLE(port)	_MMIO_PORT((port) - PORT_C, _MG_PLL1_ENABLE, \
+					   _MG_PLL2_ENABLE)
+
+#define _MG_REFCLKIN_CTL_PORT1				0x16892C
+#define _MG_REFCLKIN_CTL_PORT2				0x16992C
+#define _MG_REFCLKIN_CTL_PORT3				0x16A92C
+#define _MG_REFCLKIN_CTL_PORT4				0x16B92C
+#define   MG_REFCLKIN_CTL_OD_2_MUX(x)			((x) << 8)
+#define MG_REFCLKIN_CTL(port) _MMIO_PORT((port) - PORT_C, \
+					 _MG_REFCLKIN_CTL_PORT1, \
+					 _MG_REFCLKIN_CTL_PORT2)
+
+#define _MG_CLKTOP2_CORECLKCTL1_PORT1			0x1688D8
+#define _MG_CLKTOP2_CORECLKCTL1_PORT2			0x1698D8
+#define _MG_CLKTOP2_CORECLKCTL1_PORT3			0x16A8D8
+#define _MG_CLKTOP2_CORECLKCTL1_PORT4			0x16B8D8
+#define   MG_CLKTOP2_CORECLKCTL1_B_DIVRATIO(x)		((x) << 16)
+#define   MG_CLKTOP2_CORECLKCTL1_A_DIVRATIO(x)		((x) << 8)
+#define MG_CLKTOP2_CORECLKCTL1(port) _MMIO_PORT((port) - PORT_C, \
+						_MG_CLKTOP2_CORECLKCTL1_PORT1, \
+						_MG_CLKTOP2_CORECLKCTL1_PORT2)
+
+#define _MG_CLKTOP2_HSCLKCTL_PORT1			0x1688D4
+#define _MG_CLKTOP2_HSCLKCTL_PORT2			0x1698D4
+#define _MG_CLKTOP2_HSCLKCTL_PORT3			0x16A8D4
+#define _MG_CLKTOP2_HSCLKCTL_PORT4			0x16B8D4
+#define   MG_CLKTOP2_HSCLKCTL_CORE_INPUTSEL(x)		((x) << 16)
+#define   MG_CLKTOP2_HSCLKCTL_TLINEDRV_CLKSEL(x)	((x) << 14)
+#define   MG_CLKTOP2_HSCLKCTL_HSDIV_RATIO(x)		((x) << 12)
+#define   MG_CLKTOP2_HSCLKCTL_DSDIV_RATIO(x)		((x) << 8)
+#define MG_CLKTOP2_HSCLKCTL(port) _MMIO_PORT((port) - PORT_C, \
+					     _MG_CLKTOP2_HSCLKCTL_PORT1, \
+					     _MG_CLKTOP2_HSCLKCTL_PORT2)
+
+#define _MG_PLL_DIV0_PORT1				0x168A00
+#define _MG_PLL_DIV0_PORT2				0x169A00
+#define _MG_PLL_DIV0_PORT3				0x16AA00
+#define _MG_PLL_DIV0_PORT4				0x16BA00
+#define   MG_PLL_DIV0_FRACNEN_H				(1 << 30)
+#define   MG_PLL_DIV0_FBDIV_FRAC(x)			((x) << 8)
+#define   MG_PLL_DIV0_FBDIV_INT(x)			((x) << 0)
+#define MG_PLL_DIV0(port) _MMIO_PORT((port) - PORT_C, _MG_PLL_DIV0_PORT1, \
+				     _MG_PLL_DIV0_PORT2)
+
+#define _MG_PLL_DIV1_PORT1				0x168A04
+#define _MG_PLL_DIV1_PORT2				0x169A04
+#define _MG_PLL_DIV1_PORT3				0x16AA04
+#define _MG_PLL_DIV1_PORT4				0x16BA04
+#define   MG_PLL_DIV1_IREF_NDIVRATIO(x)			((x) << 16)
+#define   MG_PLL_DIV1_DITHER_DIV_1			(0 << 12)
+#define   MG_PLL_DIV1_DITHER_DIV_2			(1 << 12)
+#define   MG_PLL_DIV1_DITHER_DIV_4			(2 << 12)
+#define   MG_PLL_DIV1_DITHER_DIV_8			(3 << 12)
+#define   MG_PLL_DIV1_NDIVRATIO(x)			((x) << 4)
+#define   MG_PLL_DIV1_FBPREDIV(x)			((x) << 0)
+#define MG_PLL_DIV1(port) _MMIO_PORT((port) - PORT_C, _MG_PLL_DIV1_PORT1, \
+				     _MG_PLL_DIV1_PORT2)
+
+#define _MG_PLL_LF_PORT1				0x168A08
+#define _MG_PLL_LF_PORT2				0x169A08
+#define _MG_PLL_LF_PORT3				0x16AA08
+#define _MG_PLL_LF_PORT4				0x16BA08
+#define   MG_PLL_LF_TDCTARGETCNT(x)			((x) << 24)
+#define   MG_PLL_LF_AFCCNTSEL_256			(0 << 20)
+#define   MG_PLL_LF_AFCCNTSEL_512			(1 << 20)
+#define   MG_PLL_LF_GAINCTRL(x)				((x) << 16)
+#define   MG_PLL_LF_INT_COEFF(x)			((x) << 8)
+#define   MG_PLL_LF_PROP_COEFF(x)			((x) << 0)
+#define MG_PLL_LF(port) _MMIO_PORT((port) - PORT_C, _MG_PLL_LF_PORT1, \
+				   _MG_PLL_LF_PORT2)
+
+#define _MG_PLL_FRAC_LOCK_PORT1				0x168A0C
+#define _MG_PLL_FRAC_LOCK_PORT2				0x169A0C
+#define _MG_PLL_FRAC_LOCK_PORT3				0x16AA0C
+#define _MG_PLL_FRAC_LOCK_PORT4				0x16BA0C
+#define   MG_PLL_FRAC_LOCK_TRUELOCK_CRIT_32		(1 << 18)
+#define   MG_PLL_FRAC_LOCK_EARLYLOCK_CRIT_32		(1 << 16)
+#define   MG_PLL_FRAC_LOCK_LOCKTHRESH(x)		((x) << 11)
+#define   MG_PLL_FRAC_LOCK_DCODITHEREN			(1 << 10)
+#define   MG_PLL_FRAC_LOCK_FEEDFWRDCAL_EN		(1 << 8)
+#define   MG_PLL_FRAC_LOCK_FEEDFWRDGAIN(x)		((x) << 0)
+#define MG_PLL_FRAC_LOCK(port) _MMIO_PORT((port) - PORT_C, \
+					  _MG_PLL_FRAC_LOCK_PORT1, \
+					  _MG_PLL_FRAC_LOCK_PORT2)
+
+#define _MG_PLL_SSC_PORT1				0x168A10
+#define _MG_PLL_SSC_PORT2				0x169A10
+#define _MG_PLL_SSC_PORT3				0x16AA10
+#define _MG_PLL_SSC_PORT4				0x16BA10
+#define   MG_PLL_SSC_EN					(1 << 28)
+#define   MG_PLL_SSC_TYPE(x)				((x) << 26)
+#define   MG_PLL_SSC_STEPLENGTH(x)			((x) << 16)
+#define   MG_PLL_SSC_STEPNUM(x)				((x) << 10)
+#define   MG_PLL_SSC_FLLEN				(1 << 9)
+#define   MG_PLL_SSC_STEPSIZE(x)			((x) << 0)
+#define MG_PLL_SSC(port) _MMIO_PORT((port) - PORT_C, _MG_PLL_SSC_PORT1, \
+				    _MG_PLL_SSC_PORT2)
+
+#define _MG_PLL_BIAS_PORT1				0x168A14
+#define _MG_PLL_BIAS_PORT2				0x169A14
+#define _MG_PLL_BIAS_PORT3				0x16AA14
+#define _MG_PLL_BIAS_PORT4				0x16BA14
+#define   MG_PLL_BIAS_BIAS_GB_SEL(x)			((x) << 30)
+#define   MG_PLL_BIAS_INIT_DCOAMP(x)			((x) << 24)
+#define   MG_PLL_BIAS_BIAS_BONUS(x)			((x) << 16)
+#define   MG_PLL_BIAS_BIASCAL_EN			(1 << 15)
+#define   MG_PLL_BIAS_CTRIM(x)				((x) << 8)
+#define   MG_PLL_BIAS_VREF_RDAC(x)			((x) << 5)
+#define   MG_PLL_BIAS_IREFTRIM(x)			((x) << 0)
+#define MG_PLL_BIAS(port) _MMIO_PORT((port) - PORT_C, _MG_PLL_BIAS_PORT1, \
+				     _MG_PLL_BIAS_PORT2)
+
+#define _MG_PLL_TDC_COLDST_BIAS_PORT1			0x168A18
+#define _MG_PLL_TDC_COLDST_BIAS_PORT2			0x169A18
+#define _MG_PLL_TDC_COLDST_BIAS_PORT3			0x16AA18
+#define _MG_PLL_TDC_COLDST_BIAS_PORT4			0x16BA18
+#define   MG_PLL_TDC_COLDST_IREFINT_EN			(1 << 27)
+#define   MG_PLL_TDC_COLDST_REFBIAS_START_PULSE_W(x)	((x) << 17)
+#define   MG_PLL_TDC_COLDST_COLDSTART			(1 << 16)
+#define   MG_PLL_TDC_TDCOVCCORR_EN			(1 << 2)
+#define   MG_PLL_TDC_TDCSEL(x)				((x) << 0)
+#define MG_PLL_TDC_COLDST_BIAS(port) _MMIO_PORT((port) - PORT_C, \
+						_MG_PLL_TDC_COLDST_BIAS_PORT1, \
+						_MG_PLL_TDC_COLDST_BIAS_PORT2)
+
 #define _CNL_DPLL0_CFGCR0		0x6C000
 #define _CNL_DPLL1_CFGCR0		0x6C080
 #define  DPLL_CFGCR0_HDMI_MODE		(1 << 30)
 #define  DPLL_CFGCR0_SSC_ENABLE		(1 << 29)
+#define  DPLL_CFGCR0_SSC_ENABLE_ICL	(1 << 25)
 #define  DPLL_CFGCR0_LINK_RATE_MASK	(0xf << 25)
 #define  DPLL_CFGCR0_LINK_RATE_2700	(0 << 25)
 #define  DPLL_CFGCR0_LINK_RATE_1350	(1 << 25)
@@ -8942,8 +9080,19 @@ enum skl_power_gate {
 #define  DPLL_CFGCR1_PDIV_5		(4 << 2)
 #define  DPLL_CFGCR1_PDIV_7		(8 << 2)
 #define  DPLL_CFGCR1_CENTRAL_FREQ	(3 << 0)
+#define  DPLL_CFGCR1_CENTRAL_FREQ_8400	(3 << 0)
 #define CNL_DPLL_CFGCR1(pll)		_MMIO_PLL(pll, _CNL_DPLL0_CFGCR1, _CNL_DPLL1_CFGCR1)
 
+#define _ICL_DPLL0_CFGCR0		0x164000
+#define _ICL_DPLL1_CFGCR0		0x164080
+#define ICL_DPLL_CFGCR0(pll)		_MMIO_PLL(pll, _ICL_DPLL0_CFGCR0, \
+						  _ICL_DPLL1_CFGCR0)
+
+#define _ICL_DPLL0_CFGCR1		0x164004
+#define _ICL_DPLL1_CFGCR1		0x164084
+#define ICL_DPLL_CFGCR1(pll)		_MMIO_PLL(pll, _ICL_DPLL0_CFGCR1, \
+						  _ICL_DPLL1_CFGCR1)
+
 /* BXT display engine PLL */
 #define BXT_DE_PLL_CTL			_MMIO(0x6d000)
 #define   BXT_DE_PLL_RATIO(x)		(x)	/* {60,65,100} * 19.2MHz */
-- 
cgit v1.1


From 52759c0963510a2843774aac9b65ccaed3308dc0 Mon Sep 17 00:00:00 2001
From: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Date: Sun, 29 Apr 2018 15:01:46 +0900
Subject: ALSA: dice: fix kernel NULL pointer dereference due to invalid
 calculation for array index

At a commit f91c9d7610a ('ALSA: firewire-lib: cache maximum length of
payload to reduce function calls'), maximum size of payload for tx
isochronous packet is cached to reduce the number of function calls.

This cache was programmed to updated at a first callback of ohci1394 IR
context. However, the maximum size is required to queueing packets before
starting the isochronous context.

As a result, the cached value is reused to queue packets in next time to
starting the isochronous context. Then the cache is updated in a first
callback of the isochronous context. This can cause kernel NULL pointer
dereference in a below call graph:

(sound/firewire/amdtp-stream.c)
amdtp_stream_start()
->queue_in_packet()
  ->queue_packet()
    (drivers/firewire/core-iso.c)
    ->fw_iso_context_queue()
      ->struct fw_card_driver.queue_iso()
      (drivers/firewire/ohci.c)
      = ohci_queue_iso()
        ->queue_iso_packet_per_buffer()
          buffer->pages[page]

The issued dereference occurs in a case that:
 - target unit supports different stream formats for sampling transmission
   frequency.
 - maximum length of payload for tx stream in a first trial is bigger
   than the length in a second trial.

In this case, correct number of pages are allocated for DMA and the 'pages'
array has enough elements, while index of the element is wrongly calculated
according to the old value of length of payload in a call of
'queue_in_packet()'. Then it causes the issue.

This commit fixes the critical bug. This affects all of drivers in ALSA
firewire stack in Linux kernel v4.12 or later.

[12665.302360] BUG: unable to handle kernel NULL pointer dereference at 0000000000000030
[12665.302415] IP: ohci_queue_iso+0x47c/0x800 [firewire_ohci]
[12665.302439] PGD 0
[12665.302440] P4D 0
[12665.302450]
[12665.302470] Oops: 0000 [#1] SMP PTI
[12665.302487] Modules linked in: ...
[12665.303096] CPU: 1 PID: 12760 Comm: jackd Tainted: P           OE   4.13.0-38-generic #43-Ubuntu
[12665.303154] Hardware name:                  /DH77DF, BIOS KCH7710H.86A.0069.2012.0224.1825 02/24/2012
[12665.303215] task: ffff9ce87da2ae80 task.stack: ffffb5b8823d0000
[12665.303258] RIP: 0010:ohci_queue_iso+0x47c/0x800 [firewire_ohci]
[12665.303301] RSP: 0018:ffffb5b8823d3ab8 EFLAGS: 00010086
[12665.303337] RAX: ffff9ce4f4876930 RBX: 0000000000000008 RCX: ffff9ce88a3955e0
[12665.303384] RDX: 0000000000000000 RSI: 0000000034877f00 RDI: 0000000000000000
[12665.303427] RBP: ffffb5b8823d3b68 R08: ffff9ce8ccb390a0 R09: ffff9ce877639ab0
[12665.303475] R10: 0000000000000108 R11: 0000000000000000 R12: 0000000000000003
[12665.303513] R13: 0000000000000000 R14: ffff9ce4f4876950 R15: 0000000000000000
[12665.303554] FS:  00007f2ec467f8c0(0000) GS:ffff9ce8df280000(0000) knlGS:0000000000000000
[12665.303600] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[12665.303633] CR2: 0000000000000030 CR3: 00000002dcf90004 CR4: 00000000000606e0
[12665.303674] Call Trace:
[12665.303698]  fw_iso_context_queue+0x18/0x20 [firewire_core]
[12665.303735]  queue_packet+0x88/0xe0 [snd_firewire_lib]
[12665.303770]  amdtp_stream_start+0x19b/0x270 [snd_firewire_lib]
[12665.303811]  start_streams+0x276/0x3c0 [snd_dice]
[12665.303840]  snd_dice_stream_start_duplex+0x1bf/0x480 [snd_dice]
[12665.303882]  ? vma_gap_callbacks_rotate+0x1e/0x30
[12665.303914]  ? __rb_insert_augmented+0xab/0x240
[12665.303936]  capture_prepare+0x3c/0x70 [snd_dice]
[12665.303961]  snd_pcm_do_prepare+0x1d/0x30 [snd_pcm]
[12665.303985]  snd_pcm_action_single+0x3b/0x90 [snd_pcm]
[12665.304009]  snd_pcm_action_nonatomic+0x68/0x70 [snd_pcm]
[12665.304035]  snd_pcm_prepare+0x68/0x90 [snd_pcm]
[12665.304058]  snd_pcm_common_ioctl1+0x4c0/0x940 [snd_pcm]
[12665.304083]  snd_pcm_capture_ioctl1+0x19b/0x250 [snd_pcm]
[12665.304108]  snd_pcm_capture_ioctl+0x27/0x40 [snd_pcm]
[12665.304131]  do_vfs_ioctl+0xa8/0x630
[12665.304148]  ? entry_SYSCALL_64_after_hwframe+0xe9/0x139
[12665.304172]  ? entry_SYSCALL_64_after_hwframe+0xe2/0x139
[12665.304195]  ? entry_SYSCALL_64_after_hwframe+0xdb/0x139
[12665.304218]  ? entry_SYSCALL_64_after_hwframe+0xd4/0x139
[12665.304242]  ? entry_SYSCALL_64_after_hwframe+0xcd/0x139
[12665.304265]  ? entry_SYSCALL_64_after_hwframe+0xc6/0x139
[12665.304288]  ? entry_SYSCALL_64_after_hwframe+0xbf/0x139
[12665.304312]  ? entry_SYSCALL_64_after_hwframe+0xb8/0x139
[12665.304335]  ? entry_SYSCALL_64_after_hwframe+0xb1/0x139
[12665.304358]  SyS_ioctl+0x79/0x90
[12665.304374]  ? entry_SYSCALL_64_after_hwframe+0x72/0x139
[12665.304397]  entry_SYSCALL_64_fastpath+0x24/0xab
[12665.304417] RIP: 0033:0x7f2ec3750ef7
[12665.304433] RSP: 002b:00007fff99e31388 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
[12665.304465] RAX: ffffffffffffffda RBX: 00007fff99e312f0 RCX: 00007f2ec3750ef7
[12665.304494] RDX: 0000000000000000 RSI: 0000000000004140 RDI: 0000000000000007
[12665.304522] RBP: 0000556ebc63fd60 R08: 0000556ebc640560 R09: 0000000000000000
[12665.304553] R10: 0000000000000001 R11: 0000000000000246 R12: 0000556ebc63fcf0
[12665.304584] R13: 0000000000000000 R14: 0000000000000007 R15: 0000000000000000
[12665.304612] Code: 01 00 00 44 89 eb 45 31 ed 45 31 db 66 41 89 1e 66 41 89 5e 0c 66 45 89 5e 0e 49 8b 49 08 49 63 d4 4d 85 c0 49 63 ff 48 8b 14 d1 <48> 8b 72 30 41 8d 14 37 41 89 56 04 48 63 d3 0f 84 ce 00 00 00
[12665.304713] RIP: ohci_queue_iso+0x47c/0x800 [firewire_ohci] RSP: ffffb5b8823d3ab8
[12665.304743] CR2: 0000000000000030
[12665.317701] ---[ end trace 9d55b056dd52a19f ]---

Fixes: f91c9d7610a ('ALSA: firewire-lib: cache maximum length of payload to reduce function calls')
Cc: <stable@vger.kernel.org> # v4.12+
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
---
 sound/firewire/amdtp-stream.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/sound/firewire/amdtp-stream.c b/sound/firewire/amdtp-stream.c
index 4a1dc14..cb9acfe 100644
--- a/sound/firewire/amdtp-stream.c
+++ b/sound/firewire/amdtp-stream.c
@@ -773,8 +773,6 @@ static void amdtp_stream_first_callback(struct fw_iso_context *context,
 	u32 cycle;
 	unsigned int packets;
 
-	s->max_payload_length = amdtp_stream_get_max_payload(s);
-
 	/*
 	 * For in-stream, first packet has come.
 	 * For out-stream, prepared to transmit first packet
@@ -879,6 +877,9 @@ int amdtp_stream_start(struct amdtp_stream *s, int channel, int speed)
 
 	amdtp_stream_update(s);
 
+	if (s->direction == AMDTP_IN_STREAM)
+		s->max_payload_length = amdtp_stream_get_max_payload(s);
+
 	if (s->flags & CIP_NO_HEADER)
 		s->tag = TAG_NO_CIP_HEADER;
 	else
-- 
cgit v1.1


From e8238fc2bd7b4c3c7554fa2df067e796610212fc Mon Sep 17 00:00:00 2001
From: Hangbin Liu <liuhangbin@gmail.com>
Date: Fri, 27 Apr 2018 20:59:24 +0800
Subject: bridge: check iface upper dev when setting master via ioctl

When we set a bond slave's master to bridge via ioctl, we only check
the IFF_BRIDGE_PORT flag. Although we will find the slave's real master
at netdev_master_upper_dev_link() later, it already does some settings
and allocates some resources. It would be better to return as early
as possible.

v1 -> v2:
use netdev_master_upper_dev_get() instead of netdev_has_any_upper_dev()
to check if we have a master, because not all upper devs are masters,
e.g. vlan device.

Reported-by: syzbot+de73361ee4971b6e6f75@syzkaller.appspotmail.com
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Acked-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/bridge/br_if.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/net/bridge/br_if.c b/net/bridge/br_if.c
index 82c1a6f..5bb6681 100644
--- a/net/bridge/br_if.c
+++ b/net/bridge/br_if.c
@@ -518,8 +518,8 @@ int br_add_if(struct net_bridge *br, struct net_device *dev,
 		return -ELOOP;
 	}
 
-	/* Device is already being bridged */
-	if (br_port_exists(dev))
+	/* Device has master upper dev */
+	if (netdev_master_upper_dev_get(dev))
 		return -EBUSY;
 
 	/* No bridging devices that dislike that (e.g. wireless) */
-- 
cgit v1.1


From 14b7dc18ee1d9ae79eb615ea6a918d15bfddd220 Mon Sep 17 00:00:00 2001
From: Colin Ian King <colin.king@canonical.com>
Date: Fri, 27 Apr 2018 20:09:25 +0100
Subject: net: systemport: fix spelling mistake: "asymetric" -> "asymmetric"

Trivial fix to spelling mistake in netdev_warn warning message

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/broadcom/bcmsysport.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/broadcom/bcmsysport.c b/drivers/net/ethernet/broadcom/bcmsysport.c
index 0c2b0fa..f33b25f 100644
--- a/drivers/net/ethernet/broadcom/bcmsysport.c
+++ b/drivers/net/ethernet/broadcom/bcmsysport.c
@@ -2186,7 +2186,7 @@ static int bcm_sysport_map_queues(struct notifier_block *nb,
 
 	if (priv->per_port_num_tx_queues &&
 	    priv->per_port_num_tx_queues != num_tx_queues)
-		netdev_warn(slave_dev, "asymetric number of per-port queues\n");
+		netdev_warn(slave_dev, "asymmetric number of per-port queues\n");
 
 	priv->per_port_num_tx_queues = num_tx_queues;
 
-- 
cgit v1.1


From 2cb5fb1454ef4990f44f3070226ee29201bd5c87 Mon Sep 17 00:00:00 2001
From: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Date: Fri, 27 Apr 2018 16:46:11 -0300
Subject: MAINTAINERS: add myself as SCTP co-maintainer

Signed-off-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 MAINTAINERS | 1 +
 1 file changed, 1 insertion(+)

diff --git a/MAINTAINERS b/MAINTAINERS
index 2310341b..6d21f2f 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -12508,6 +12508,7 @@ F:	drivers/scsi/st_*.h
 SCTP PROTOCOL
 M:	Vlad Yasevich <vyasevich@gmail.com>
 M:	Neil Horman <nhorman@tuxdriver.com>
+M:	Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
 L:	linux-sctp@vger.kernel.org
 W:	http://lksctp.sourceforge.net
 S:	Maintained
-- 
cgit v1.1


From b607990c76ceda0a7a7ceacabab174cdc8b9beee Mon Sep 17 00:00:00 2001
From: Ian W MORRISON <ianwmorrison@gmail.com>
Date: Wed, 11 Apr 2018 14:42:13 +1000
Subject: drm/i915/glk: Add MODULE_FIRMWARE for Geminilake

As the Geminilake firmware is now merged to linux-firmware.git
use MODUE_FIRMWARE to load the firmware.

This removes the error message in the dmesg log:

    i915 0000:00:02.0: Direct firmware load for
        i915/glk_dmc_ver1_04.bin failed with error -2
    i915 0000:00:02.0: Failed to load DMC firmware
        i915/glk_dmc_ver1_04.bin. Disabling runtime power management.
    i915 0000:00:02.0: DMC firmware homepage:
        https://01.org/linuxgraphics/downloads/firmware

and now shows that the firmware has correctly loaded:

    [drm] Finished loading DMC firmware i915/glk_dmc_ver1_04.bin (v1.4)

Signed-off-by: Ian W MORRISON <ianwmorrison@gmail.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180411044213.383-1-ianwmorrison@gmail.com
(cherry picked from commit f6d3e06f074721ad3a231df745d85b60428c1f03)
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
---
 drivers/gpu/drm/i915/intel_csr.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/i915/intel_csr.c b/drivers/gpu/drm/i915/intel_csr.c
index 41e6c75..f9550ea 100644
--- a/drivers/gpu/drm/i915/intel_csr.c
+++ b/drivers/gpu/drm/i915/intel_csr.c
@@ -35,6 +35,7 @@
  */
 
 #define I915_CSR_GLK "i915/glk_dmc_ver1_04.bin"
+MODULE_FIRMWARE(I915_CSR_GLK);
 #define GLK_CSR_VERSION_REQUIRED	CSR_VERSION(1, 4)
 
 #define I915_CSR_CNL "i915/cnl_dmc_ver1_07.bin"
-- 
cgit v1.1


From 76b3421b39bd610546931fc923edcf90c18fa395 Mon Sep 17 00:00:00 2001
From: Takashi Iwai <tiwai@suse.de>
Date: Mon, 30 Apr 2018 10:06:48 +0200
Subject: ALSA: aloop: Add missing cable lock to ctl API callbacks

Some control API callbacks in aloop driver are too lazy to take the
loopback->cable_lock and it results in possible races of cable access
while it's being freed.  It eventually lead to a UAF, as reported by
fuzzer recently.

This patch covers such control API callbacks and add the proper mutex
locks.

Reported-by: DaeRyong Jeong <threeearcat@gmail.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
---
 sound/drivers/aloop.c | 17 +++++++++++++++--
 1 file changed, 15 insertions(+), 2 deletions(-)

diff --git a/sound/drivers/aloop.c b/sound/drivers/aloop.c
index 58e349f..eab7f59 100644
--- a/sound/drivers/aloop.c
+++ b/sound/drivers/aloop.c
@@ -831,9 +831,11 @@ static int loopback_rate_shift_get(struct snd_kcontrol *kcontrol,
 {
 	struct loopback *loopback = snd_kcontrol_chip(kcontrol);
 	
+	mutex_lock(&loopback->cable_lock);
 	ucontrol->value.integer.value[0] =
 		loopback->setup[kcontrol->id.subdevice]
 			       [kcontrol->id.device].rate_shift;
+	mutex_unlock(&loopback->cable_lock);
 	return 0;
 }
 
@@ -865,9 +867,11 @@ static int loopback_notify_get(struct snd_kcontrol *kcontrol,
 {
 	struct loopback *loopback = snd_kcontrol_chip(kcontrol);
 	
+	mutex_lock(&loopback->cable_lock);
 	ucontrol->value.integer.value[0] =
 		loopback->setup[kcontrol->id.subdevice]
 			       [kcontrol->id.device].notify;
+	mutex_unlock(&loopback->cable_lock);
 	return 0;
 }
 
@@ -879,12 +883,14 @@ static int loopback_notify_put(struct snd_kcontrol *kcontrol,
 	int change = 0;
 
 	val = ucontrol->value.integer.value[0] ? 1 : 0;
+	mutex_lock(&loopback->cable_lock);
 	if (val != loopback->setup[kcontrol->id.subdevice]
 				[kcontrol->id.device].notify) {
 		loopback->setup[kcontrol->id.subdevice]
 			[kcontrol->id.device].notify = val;
 		change = 1;
 	}
+	mutex_unlock(&loopback->cable_lock);
 	return change;
 }
 
@@ -892,15 +898,18 @@ static int loopback_active_get(struct snd_kcontrol *kcontrol,
 			       struct snd_ctl_elem_value *ucontrol)
 {
 	struct loopback *loopback = snd_kcontrol_chip(kcontrol);
-	struct loopback_cable *cable = loopback->cables
-			[kcontrol->id.subdevice][kcontrol->id.device ^ 1];
+	struct loopback_cable *cable;
+
 	unsigned int val = 0;
 
+	mutex_lock(&loopback->cable_lock);
+	cable = loopback->cables[kcontrol->id.subdevice][kcontrol->id.device ^ 1];
 	if (cable != NULL) {
 		unsigned int running = cable->running ^ cable->pause;
 
 		val = (running & (1 << SNDRV_PCM_STREAM_PLAYBACK)) ? 1 : 0;
 	}
+	mutex_unlock(&loopback->cable_lock);
 	ucontrol->value.integer.value[0] = val;
 	return 0;
 }
@@ -943,9 +952,11 @@ static int loopback_rate_get(struct snd_kcontrol *kcontrol,
 {
 	struct loopback *loopback = snd_kcontrol_chip(kcontrol);
 	
+	mutex_lock(&loopback->cable_lock);
 	ucontrol->value.integer.value[0] =
 		loopback->setup[kcontrol->id.subdevice]
 			       [kcontrol->id.device].rate;
+	mutex_unlock(&loopback->cable_lock);
 	return 0;
 }
 
@@ -965,9 +976,11 @@ static int loopback_channels_get(struct snd_kcontrol *kcontrol,
 {
 	struct loopback *loopback = snd_kcontrol_chip(kcontrol);
 	
+	mutex_lock(&loopback->cable_lock);
 	ucontrol->value.integer.value[0] =
 		loopback->setup[kcontrol->id.subdevice]
 			       [kcontrol->id.device].channels;
+	mutex_unlock(&loopback->cable_lock);
 	return 0;
 }
 
-- 
cgit v1.1


From d4f3388afd488ed15368fa7413b8bd6d1f98bb1d Mon Sep 17 00:00:00 2001
From: Prashanth Prakash <pprakash@codeaurora.org>
Date: Fri, 27 Apr 2018 11:35:27 -0600
Subject: cpufreq / CPPC: Set platform specific transition_delay_us

Add support to specify platform specific transition_delay_us instead
of using the transition delay derived from PCC.

With commit 3d41386d556d (cpufreq: CPPC: Use transition_delay_us
depending transition_latency) we are setting transition_delay_us
directly and not applying the LATENCY_MULTIPLIER. Because of that,
on Qualcomm Centriq we can end up with a very high rate of frequency
change requests when using the schedutil governor (default
rate_limit_us=10 compared to an earlier value of 10000).

The PCC subspace describes the rate at which the platform can accept
commands on the CPPC's PCC channel. This includes read and write
command on the PCC channel that can be used for reasons other than
frequency transitions. Moreover the same PCC subspace can be used by
multiple freq domains and deriving transition_delay_us from it as we
do now can be sub-optimal.

Moreover if a platform does not use PCC for desired_perf register then
there is no way to compute the transition latency or the delay_us.

CPPC does not have a standard defined mechanism to get the transition
rate or the latency at the moment.

Given the above limitations, it is simpler to have a platform specific
transition_delay_us and rely on PCC derived value only if a platform
specific value is not available.

Signed-off-by: Prashanth Prakash <pprakash@codeaurora.org>
Cc: 4.14+ <stable@vger.kernel.org> # 4.14+
Fixes: 3d41386d556d (cpufreq: CPPC: Use transition_delay_us depending transition_latency)
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
---
 drivers/cpufreq/cppc_cpufreq.c | 46 ++++++++++++++++++++++++++++++++++++++++--
 1 file changed, 44 insertions(+), 2 deletions(-)

diff --git a/drivers/cpufreq/cppc_cpufreq.c b/drivers/cpufreq/cppc_cpufreq.c
index bc5fc16..b15115a 100644
--- a/drivers/cpufreq/cppc_cpufreq.c
+++ b/drivers/cpufreq/cppc_cpufreq.c
@@ -126,6 +126,49 @@ static void cppc_cpufreq_stop_cpu(struct cpufreq_policy *policy)
 				cpu->perf_caps.lowest_perf, cpu_num, ret);
 }
 
+/*
+ * The PCC subspace describes the rate at which platform can accept commands
+ * on the shared PCC channel (including READs which do not count towards freq
+ * trasition requests), so ideally we need to use the PCC values as a fallback
+ * if we don't have a platform specific transition_delay_us
+ */
+#ifdef CONFIG_ARM64
+#include <asm/cputype.h>
+
+static unsigned int cppc_cpufreq_get_transition_delay_us(int cpu)
+{
+	unsigned long implementor = read_cpuid_implementor();
+	unsigned long part_num = read_cpuid_part_number();
+	unsigned int delay_us = 0;
+
+	switch (implementor) {
+	case ARM_CPU_IMP_QCOM:
+		switch (part_num) {
+		case QCOM_CPU_PART_FALKOR_V1:
+		case QCOM_CPU_PART_FALKOR:
+			delay_us = 10000;
+			break;
+		default:
+			delay_us = cppc_get_transition_latency(cpu) / NSEC_PER_USEC;
+			break;
+		}
+		break;
+	default:
+		delay_us = cppc_get_transition_latency(cpu) / NSEC_PER_USEC;
+		break;
+	}
+
+	return delay_us;
+}
+
+#else
+
+static unsigned int cppc_cpufreq_get_transition_delay_us(int cpu)
+{
+	return cppc_get_transition_latency(cpu) / NSEC_PER_USEC;
+}
+#endif
+
 static int cppc_cpufreq_cpu_init(struct cpufreq_policy *policy)
 {
 	struct cppc_cpudata *cpu;
@@ -162,8 +205,7 @@ static int cppc_cpufreq_cpu_init(struct cpufreq_policy *policy)
 		cpu->perf_caps.highest_perf;
 	policy->cpuinfo.max_freq = cppc_dmi_max_khz;
 
-	policy->transition_delay_us = cppc_get_transition_latency(cpu_num) /
-		NSEC_PER_USEC;
+	policy->transition_delay_us = cppc_cpufreq_get_transition_delay_us(cpu_num);
 	policy->shared_type = cpu->shared_type;
 
 	if (policy->shared_type == CPUFREQ_SHARED_TYPE_ANY) {
-- 
cgit v1.1


From e026646c178d8292de563fbecc247bada059c282 Mon Sep 17 00:00:00 2001
From: William Breathitt Gray <vilhelm.gray@gmail.com>
Date: Fri, 20 Apr 2018 09:50:20 -0400
Subject: gpio: pcie-idio-24: Fix off-by-one error in get_multiple loop

The PCIe-IDIO-24 features 8 bits of TTL GPIO which may be configured for
output or input. This patch fixes an off-by-one error in the loop
conditional for the get_multiple callback so that the TTL GPIO are
handled.

Fixes: ca37081595a2 ("gpio: pcie-idio-24: Implement get_multiple/set_multiple callbacks")
Signed-off-by: William Breathitt Gray <vilhelm.gray@gmail.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
---
 drivers/gpio/gpio-pcie-idio-24.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpio/gpio-pcie-idio-24.c b/drivers/gpio/gpio-pcie-idio-24.c
index 3e77c2a..f953541 100644
--- a/drivers/gpio/gpio-pcie-idio-24.c
+++ b/drivers/gpio/gpio-pcie-idio-24.c
@@ -217,7 +217,7 @@ static int idio_24_gpio_get_multiple(struct gpio_chip *chip,
 	bitmap_zero(bits, chip->ngpio);
 
 	/* get bits are evaluated a gpio port register at a time */
-	for (i = 0; i < ARRAY_SIZE(ports); i++) {
+	for (i = 0; i < ARRAY_SIZE(ports) + 1; i++) {
 		/* gpio offset in bits array */
 		bits_offset = i * gpio_reg_size;
 
-- 
cgit v1.1


From 544a591668813583021474fa5c7ff4942244d654 Mon Sep 17 00:00:00 2001
From: Hans de Goede <hdegoede@redhat.com>
Date: Thu, 26 Apr 2018 14:18:19 +0200
Subject: Revert "Bluetooth: btusb: Fix quirk for Atheros 1525/QCA6174"

Commit f44cb4b19ed4 ("Bluetooth: btusb: Fix quirk for Atheros
1525/QCA6174") is causing bluetooth to no longer work for several
people, see: https://bugzilla.redhat.com/show_bug.cgi?id=1568911

So lets revert it for now and try to find another solution for
devices which need the modified quirk.

Cc: stable@vger.kernel.org
Cc: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
---
 drivers/bluetooth/btusb.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/bluetooth/btusb.c b/drivers/bluetooth/btusb.c
index c8c8b0b..ee31c99 100644
--- a/drivers/bluetooth/btusb.c
+++ b/drivers/bluetooth/btusb.c
@@ -231,6 +231,7 @@ static const struct usb_device_id blacklist_table[] = {
 	{ USB_DEVICE(0x0930, 0x0227), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x0b05, 0x17d0), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x0cf3, 0x0036), .driver_info = BTUSB_ATH3012 },
+	{ USB_DEVICE(0x0cf3, 0x3004), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x0cf3, 0x3008), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x0cf3, 0x311d), .driver_info = BTUSB_ATH3012 },
 	{ USB_DEVICE(0x0cf3, 0x311e), .driver_info = BTUSB_ATH3012 },
@@ -263,7 +264,6 @@ static const struct usb_device_id blacklist_table[] = {
 	{ USB_DEVICE(0x0489, 0xe03c), .driver_info = BTUSB_ATH3012 },
 
 	/* QCA ROME chipset */
-	{ USB_DEVICE(0x0cf3, 0x3004), .driver_info = BTUSB_QCA_ROME },
 	{ USB_DEVICE(0x0cf3, 0xe007), .driver_info = BTUSB_QCA_ROME },
 	{ USB_DEVICE(0x0cf3, 0xe009), .driver_info = BTUSB_QCA_ROME },
 	{ USB_DEVICE(0x0cf3, 0xe010), .driver_info = BTUSB_QCA_ROME },
-- 
cgit v1.1


From fc54910280eb38bde923cdf0898e74687d8e6989 Mon Sep 17 00:00:00 2001
From: Hans de Goede <hdegoede@redhat.com>
Date: Fri, 27 Apr 2018 11:26:43 +0200
Subject: Bluetooth: btusb: Only check needs_reset_resume DMI table for QCA
 rome chipsets

Jeremy Cline correctly points out in rhbz#1514836 that a device where the
QCA rome chipset needs the USB_QUIRK_RESET_RESUME quirk, may also ship
with a different wifi/bt chipset in some configurations.

If that is the case then we are needlessly penalizing those other chipsets
with a reset-resume quirk, typically causing 0.4W extra power use because
this disables runtime-pm.

This commit moves the DMI table check to a btusb_check_needs_reset_resume()
helper (so that we can easily also call it for other chipsets) and calls
this new helper only for QCA_ROME chipsets for now.

BugLink: https://bugzilla.redhat.com/show_bug.cgi?id=1514836
Cc: stable@vger.kernel.org
Cc: Jeremy Cline <jcline@redhat.com>
Suggested-by: Jeremy Cline <jcline@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
---
 drivers/bluetooth/btusb.c | 10 +++++++---
 1 file changed, 7 insertions(+), 3 deletions(-)

diff --git a/drivers/bluetooth/btusb.c b/drivers/bluetooth/btusb.c
index ee31c99..3d10e83 100644
--- a/drivers/bluetooth/btusb.c
+++ b/drivers/bluetooth/btusb.c
@@ -2852,6 +2852,12 @@ static int btusb_config_oob_wake(struct hci_dev *hdev)
 }
 #endif
 
+static void btusb_check_needs_reset_resume(struct usb_interface *intf)
+{
+	if (dmi_check_system(btusb_needs_reset_resume_table))
+		interface_to_usbdev(intf)->quirks |= USB_QUIRK_RESET_RESUME;
+}
+
 static int btusb_probe(struct usb_interface *intf,
 		       const struct usb_device_id *id)
 {
@@ -2974,9 +2980,6 @@ static int btusb_probe(struct usb_interface *intf,
 	hdev->send   = btusb_send_frame;
 	hdev->notify = btusb_notify;
 
-	if (dmi_check_system(btusb_needs_reset_resume_table))
-		interface_to_usbdev(intf)->quirks |= USB_QUIRK_RESET_RESUME;
-
 #ifdef CONFIG_PM
 	err = btusb_config_oob_wake(hdev);
 	if (err)
@@ -3064,6 +3067,7 @@ static int btusb_probe(struct usb_interface *intf,
 		data->setup_on_usb = btusb_setup_qca;
 		hdev->set_bdaddr = btusb_set_bdaddr_ath3012;
 		set_bit(HCI_QUIRK_SIMULTANEOUS_DISCOVERY, &hdev->quirks);
+		btusb_check_needs_reset_resume(intf);
 	}
 
 #ifdef CONFIG_BT_HCIBTUSB_RTL
-- 
cgit v1.1


From 596b07a9a22656493726edf1739569102bd3e136 Mon Sep 17 00:00:00 2001
From: Hans de Goede <hdegoede@redhat.com>
Date: Thu, 26 Apr 2018 20:52:06 +0200
Subject: Bluetooth: btusb: Add Dell XPS 13 9360 to
 btusb_needs_reset_resume_table

The Dell XPS 13 9360 uses a QCA Rome chip which needs to be reset
(and have its firmware reloaded) for bluetooth to work after
suspend/resume.

BugLink: https://bugzilla.redhat.com/show_bug.cgi?id=1514836
Cc: stable@vger.kernel.org
Cc: Garrett LeSage <glesage@redhat.com>
Reported-and-tested-by: Garrett LeSage <glesage@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
---
 drivers/bluetooth/btusb.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/drivers/bluetooth/btusb.c b/drivers/bluetooth/btusb.c
index 3d10e83..b937cc1 100644
--- a/drivers/bluetooth/btusb.c
+++ b/drivers/bluetooth/btusb.c
@@ -399,6 +399,13 @@ static const struct dmi_system_id btusb_needs_reset_resume_table[] = {
 			DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex 3060"),
 		},
 	},
+	{
+		/* Dell XPS 9360 (QCA ROME device 0cf3:e300) */
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
+			DMI_MATCH(DMI_PRODUCT_NAME, "XPS 13 9360"),
+		},
+	},
 	{}
 };
 
-- 
cgit v1.1


From 5692251c254a3d561316c4e8e10c77e470b60658 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Sat, 28 Apr 2018 12:15:32 +0100
Subject: drm/i915/lrc: Scrub the GPU state of the guilty hanging request
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Previously, we just reset the ring register in the context image such
that we could skip over the broken batch and emit the closing
breadcrumb. However, on resume the context image and GPU state would be
reloaded, which may have been left in an inconsistent state by the
reset. The presumption was that at worst it would just cause another
reset and skip again until it recovered, however it seems just as likely
to cause an unrecoverable hang. Instead of risking loading an incomplete
context image, restore it back to the default state.

v2: Fix up off-by-one from including the ppHSWP in with the register
state.
v3: Use a ring local to compact a few lines.
v4: Beware setting the ring local before checking for a NULL request.

References: https://bugs.freedesktop.org/show_bug.cgi?id=105304
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Michał Winiarski <michal.winiarski@intel.com>
Cc: Michel Thierry <michel.thierry@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Michel Thierry <michel.thierry@intel.com> #v2
Link: https://patchwork.freedesktop.org/patch/msgid/20180428111532.15819-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/intel_lrc.c | 24 +++++++++++++++++-------
 1 file changed, 17 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 87eb3a6..58cad24 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -1803,8 +1803,8 @@ static void reset_common_ring(struct intel_engine_cs *engine,
 			      struct i915_request *request)
 {
 	struct intel_engine_execlists * const execlists = &engine->execlists;
-	struct intel_context *ce;
 	unsigned long flags;
+	u32 *regs;
 
 	GEM_TRACE("%s request global=%x, current=%d\n",
 		  engine->name, request ? request->global_seqno : 0,
@@ -1854,14 +1854,24 @@ static void reset_common_ring(struct intel_engine_cs *engine,
 	 * future request will be after userspace has had the opportunity
 	 * to recreate its own state.
 	 */
-	ce = &request->ctx->engine[engine->id];
-	execlists_init_reg_state(ce->lrc_reg_state,
-				 request->ctx, engine, ce->ring);
+	regs = request->ctx->engine[engine->id].lrc_reg_state;
+	if (engine->default_state) {
+		void *defaults;
+
+		defaults = i915_gem_object_pin_map(engine->default_state,
+						   I915_MAP_WB);
+		if (!IS_ERR(defaults)) {
+			memcpy(regs, /* skip restoring the vanilla PPHWSP */
+			       defaults + LRC_STATE_PN * PAGE_SIZE,
+			       engine->context_size - PAGE_SIZE);
+			i915_gem_object_unpin_map(engine->default_state);
+		}
+	}
+	execlists_init_reg_state(regs, request->ctx, engine, request->ring);
 
 	/* Move the RING_HEAD onto the breadcrumb, past the hanging batch */
-	ce->lrc_reg_state[CTX_RING_BUFFER_START+1] =
-		i915_ggtt_offset(ce->ring->vma);
-	ce->lrc_reg_state[CTX_RING_HEAD+1] = request->postfix;
+	regs[CTX_RING_BUFFER_START + 1] = i915_ggtt_offset(request->ring->vma);
+	regs[CTX_RING_HEAD + 1] = request->postfix;
 
 	request->ring->head = request->postfix;
 	intel_ring_update_space(request->ring);
-- 
cgit v1.1


From ff81de73e4284649e78df1df0f63e2670ad87e7f Mon Sep 17 00:00:00 2001
From: Colin Ian King <colin.king@canonical.com>
Date: Sat, 28 Apr 2018 10:43:20 +0100
Subject: qed: fix spelling mistake: "checksumed" -> "checksummed"

Trivial fix to spelling mistake in DP_INFO message text

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/qlogic/qed/qed_ll2.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/qlogic/qed/qed_ll2.c b/drivers/net/ethernet/qlogic/qed/qed_ll2.c
index 74fc626..3850281 100644
--- a/drivers/net/ethernet/qlogic/qed/qed_ll2.c
+++ b/drivers/net/ethernet/qlogic/qed/qed_ll2.c
@@ -2370,7 +2370,7 @@ static int qed_ll2_start_xmit(struct qed_dev *cdev, struct sk_buff *skb)
 	u8 flags = 0;
 
 	if (unlikely(skb->ip_summed != CHECKSUM_NONE)) {
-		DP_INFO(cdev, "Cannot transmit a checksumed packet\n");
+		DP_INFO(cdev, "Cannot transmit a checksummed packet\n");
 		return -EINVAL;
 	}
 
-- 
cgit v1.1


From f944ad1b2b66bbec8ffc1d6d0a45565b12846308 Mon Sep 17 00:00:00 2001
From: Colin Ian King <colin.king@canonical.com>
Date: Sat, 28 Apr 2018 10:57:07 +0100
Subject: net: ethernet: ucc: fix spelling mistake: "tx-late-collsion" ->
 "tx-late-collision"

Trivial fix to spelling mistake in tx_fw_stat_gstrings text

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/freescale/ucc_geth_ethtool.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/freescale/ucc_geth_ethtool.c b/drivers/net/ethernet/freescale/ucc_geth_ethtool.c
index 4df282e..0beee2c 100644
--- a/drivers/net/ethernet/freescale/ucc_geth_ethtool.c
+++ b/drivers/net/ethernet/freescale/ucc_geth_ethtool.c
@@ -61,7 +61,7 @@ static const char hw_stat_gstrings[][ETH_GSTRING_LEN] = {
 static const char tx_fw_stat_gstrings[][ETH_GSTRING_LEN] = {
 	"tx-single-collision",
 	"tx-multiple-collision",
-	"tx-late-collsion",
+	"tx-late-collision",
 	"tx-aborted-frames",
 	"tx-lost-frames",
 	"tx-carrier-sense-errors",
-- 
cgit v1.1


From 52d7f16e5543ca892ae2393a716083d209ce3b36 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Mon, 30 Apr 2018 14:15:00 +0100
Subject: drm/i915: Stop tracking timeline->inflight_seqnos

In commit 9b6586ae9f6b ("drm/i915: Keep a global seqno per-engine"), we
moved from a global inflight counter to per-engine counters in the
hope that will be easy to run concurrently in future. However, with the
advent of the desire to move requests between engines, we do need a
global counter to preserve the semantics that no engine wraps in the
middle of a submit. (Although this semantic is now only required for gen7
semaphore support, which only supports greater-then comparisons!)

v2: Keep a global counter of all requests ever submitted and force the
reset when it wraps.

References: 9b6586ae9f6b ("drm/i915: Keep a global seqno per-engine")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180430131503.5375-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/i915_debugfs.c      |  5 ++---
 drivers/gpu/drm/i915/i915_drv.h          |  1 +
 drivers/gpu/drm/i915/i915_gem_timeline.h |  6 ------
 drivers/gpu/drm/i915/i915_request.c      | 33 ++++++++++++++++----------------
 drivers/gpu/drm/i915/intel_engine_cs.c   |  5 ++---
 5 files changed, 22 insertions(+), 28 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index cb1a804..747dad2 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -1340,10 +1340,9 @@ static int i915_hangcheck_info(struct seq_file *m, void *unused)
 		struct rb_node *rb;
 
 		seq_printf(m, "%s:\n", engine->name);
-		seq_printf(m, "\tseqno = %x [current %x, last %x], inflight %d\n",
+		seq_printf(m, "\tseqno = %x [current %x, last %x]\n",
 			   engine->hangcheck.seqno, seqno[id],
-			   intel_engine_last_submit(engine),
-			   engine->timeline->inflight_seqnos);
+			   intel_engine_last_submit(engine));
 		seq_printf(m, "\twaiters? %s, fake irq active? %s, stalled? %s\n",
 			   yesno(intel_engine_has_waiter(engine)),
 			   yesno(test_bit(engine->id,
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 193176b..dd4d6b9 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -2062,6 +2062,7 @@ struct drm_i915_private {
 		struct list_head timelines;
 		struct i915_gem_timeline global_timeline;
 		u32 active_requests;
+		u32 request_serial;
 
 		/**
 		 * Is the GPU currently considered idle, or busy executing
diff --git a/drivers/gpu/drm/i915/i915_gem_timeline.h b/drivers/gpu/drm/i915/i915_gem_timeline.h
index 33e01bf..6e82119 100644
--- a/drivers/gpu/drm/i915/i915_gem_timeline.h
+++ b/drivers/gpu/drm/i915/i915_gem_timeline.h
@@ -37,12 +37,6 @@ struct intel_timeline {
 	u64 fence_context;
 	u32 seqno;
 
-	/**
-	 * Count of outstanding requests, from the time they are constructed
-	 * to the moment they are retired. Loosely coupled to hardware.
-	 */
-	u32 inflight_seqnos;
-
 	spinlock_t lock;
 
 	/**
diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c
index b692a9f..b1993d4 100644
--- a/drivers/gpu/drm/i915/i915_request.c
+++ b/drivers/gpu/drm/i915/i915_request.c
@@ -241,6 +241,7 @@ static int reset_all_global_seqno(struct drm_i915_private *i915, u32 seqno)
 			       sizeof(timeline->engine[id].global_sync));
 	}
 
+	i915->gt.request_serial = seqno;
 	return 0;
 }
 
@@ -257,18 +258,22 @@ int i915_gem_set_global_seqno(struct drm_device *dev, u32 seqno)
 	return reset_all_global_seqno(i915, seqno - 1);
 }
 
-static int reserve_engine(struct intel_engine_cs *engine)
+static int reserve_gt(struct drm_i915_private *i915)
 {
-	struct drm_i915_private *i915 = engine->i915;
-	u32 active = ++engine->timeline->inflight_seqnos;
-	u32 seqno = engine->timeline->seqno;
 	int ret;
 
-	/* Reservation is fine until we need to wrap around */
-	if (unlikely(add_overflows(seqno, active))) {
+	/*
+	 * Reservation is fine until we may need to wrap around
+	 *
+	 * By incrementing the serial for every request, we know that no
+	 * individual engine may exceed that serial (as each is reset to 0
+	 * on any wrap). This protects even the most pessimistic of migrations
+	 * of every request from all engines onto just one.
+	 */
+	while (unlikely(++i915->gt.request_serial == 0)) {
 		ret = reset_all_global_seqno(i915, 0);
 		if (ret) {
-			engine->timeline->inflight_seqnos--;
+			i915->gt.request_serial--;
 			return ret;
 		}
 	}
@@ -279,15 +284,10 @@ static int reserve_engine(struct intel_engine_cs *engine)
 	return 0;
 }
 
-static void unreserve_engine(struct intel_engine_cs *engine)
+static void unreserve_gt(struct drm_i915_private *i915)
 {
-	struct drm_i915_private *i915 = engine->i915;
-
 	if (!--i915->gt.active_requests)
 		i915_gem_park(i915);
-
-	GEM_BUG_ON(!engine->timeline->inflight_seqnos);
-	engine->timeline->inflight_seqnos--;
 }
 
 void i915_gem_retire_noop(struct i915_gem_active *active,
@@ -362,7 +362,6 @@ static void i915_request_retire(struct i915_request *request)
 	list_del_init(&request->link);
 	spin_unlock_irq(&engine->timeline->lock);
 
-	unreserve_engine(request->engine);
 	advance_ring(request);
 
 	free_capture_list(request);
@@ -424,6 +423,8 @@ static void i915_request_retire(struct i915_request *request)
 	}
 	spin_unlock_irq(&request->lock);
 
+	unreserve_gt(request->i915);
+
 	i915_sched_node_fini(request->i915, &request->sched);
 	i915_request_put(request);
 }
@@ -642,7 +643,7 @@ i915_request_alloc(struct intel_engine_cs *engine, struct i915_gem_context *ctx)
 		return ERR_CAST(ring);
 	GEM_BUG_ON(!ring);
 
-	ret = reserve_engine(engine);
+	ret = reserve_gt(i915);
 	if (ret)
 		goto err_unpin;
 
@@ -784,7 +785,7 @@ err_unwind:
 
 	kmem_cache_free(i915->requests, rq);
 err_unreserve:
-	unreserve_engine(engine);
+	unreserve_gt(i915);
 err_unpin:
 	engine->context_unpin(engine, ctx);
 	return ERR_PTR(ret);
diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
index ac009f1..eba81d5 100644
--- a/drivers/gpu/drm/i915/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/intel_engine_cs.c
@@ -1321,12 +1321,11 @@ void intel_engine_dump(struct intel_engine_cs *engine,
 	if (i915_terminally_wedged(&engine->i915->gpu_error))
 		drm_printf(m, "*** WEDGED ***\n");
 
-	drm_printf(m, "\tcurrent seqno %x, last %x, hangcheck %x [%d ms], inflight %d\n",
+	drm_printf(m, "\tcurrent seqno %x, last %x, hangcheck %x [%d ms]\n",
 		   intel_engine_get_seqno(engine),
 		   intel_engine_last_submit(engine),
 		   engine->hangcheck.seqno,
-		   jiffies_to_msecs(jiffies - engine->hangcheck.action_timestamp),
-		   engine->timeline->inflight_seqnos);
+		   jiffies_to_msecs(jiffies - engine->hangcheck.action_timestamp));
 	drm_printf(m, "\tReset count: %d (global %d)\n",
 		   i915_reset_engine_count(error, engine),
 		   i915_reset_count(error));
-- 
cgit v1.1


From ab82a0635cdf0b91a134aaae34abd4e864595c5b Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Mon, 30 Apr 2018 14:15:01 +0100
Subject: drm/i915: Wrap engine->context_pin() and engine->context_unpin()

Make life easier in upcoming patches by moving the context_pin and
context_unpin vfuncs into inline helpers.

v2: Fixup mock_engine to mark the context as pinned on use.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180430131503.5375-2-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/gvt/mmio_context.c      |  2 +-
 drivers/gpu/drm/i915/gvt/scheduler.c         | 20 +++++++++----------
 drivers/gpu/drm/i915/i915_debugfs.c          | 20 +++++++++++--------
 drivers/gpu/drm/i915/i915_gem.c              |  4 ++--
 drivers/gpu/drm/i915/i915_gem_context.c      |  8 ++++----
 drivers/gpu/drm/i915/i915_gem_context.h      | 30 +++++++++++++++++++++++++++-
 drivers/gpu/drm/i915/i915_gpu_error.c        |  3 ++-
 drivers/gpu/drm/i915/i915_perf.c             |  9 +++++----
 drivers/gpu/drm/i915/i915_request.c          |  6 +++---
 drivers/gpu/drm/i915/intel_engine_cs.c       | 13 ++++++------
 drivers/gpu/drm/i915/intel_guc_ads.c         |  3 ++-
 drivers/gpu/drm/i915/intel_guc_submission.c  |  5 +++--
 drivers/gpu/drm/i915/intel_lrc.c             | 29 ++++++++++++++++-----------
 drivers/gpu/drm/i915/intel_lrc.h             |  2 +-
 drivers/gpu/drm/i915/intel_ringbuffer.c      | 19 +++++++++---------
 drivers/gpu/drm/i915/selftests/mock_engine.c | 13 +++++++++---
 16 files changed, 117 insertions(+), 69 deletions(-)

diff --git a/drivers/gpu/drm/i915/gvt/mmio_context.c b/drivers/gpu/drm/i915/gvt/mmio_context.c
index a5bac83..0f94955 100644
--- a/drivers/gpu/drm/i915/gvt/mmio_context.c
+++ b/drivers/gpu/drm/i915/gvt/mmio_context.c
@@ -448,7 +448,7 @@ static void switch_mocs(struct intel_vgpu *pre, struct intel_vgpu *next,
 
 bool is_inhibit_context(struct i915_gem_context *ctx, int ring_id)
 {
-	u32 *reg_state = ctx->engine[ring_id].lrc_reg_state;
+	u32 *reg_state = ctx->__engine[ring_id].lrc_reg_state;
 	u32 inhibit_mask =
 		_MASKED_BIT_ENABLE(CTX_CTRL_ENGINE_CTX_RESTORE_INHIBIT);
 
diff --git a/drivers/gpu/drm/i915/gvt/scheduler.c b/drivers/gpu/drm/i915/gvt/scheduler.c
index 35f7cfd..ffb45a9 100644
--- a/drivers/gpu/drm/i915/gvt/scheduler.c
+++ b/drivers/gpu/drm/i915/gvt/scheduler.c
@@ -58,7 +58,7 @@ static void update_shadow_pdps(struct intel_vgpu_workload *workload)
 	int ring_id = workload->ring_id;
 	struct i915_gem_context *shadow_ctx = vgpu->submission.shadow_ctx;
 	struct drm_i915_gem_object *ctx_obj =
-		shadow_ctx->engine[ring_id].state->obj;
+		shadow_ctx->__engine[ring_id].state->obj;
 	struct execlist_ring_context *shadow_ring_context;
 	struct page *page;
 
@@ -130,7 +130,7 @@ static int populate_shadow_context(struct intel_vgpu_workload *workload)
 	int ring_id = workload->ring_id;
 	struct i915_gem_context *shadow_ctx = vgpu->submission.shadow_ctx;
 	struct drm_i915_gem_object *ctx_obj =
-		shadow_ctx->engine[ring_id].state->obj;
+		shadow_ctx->__engine[ring_id].state->obj;
 	struct execlist_ring_context *shadow_ring_context;
 	struct page *page;
 	void *dst;
@@ -283,7 +283,7 @@ static int shadow_context_status_change(struct notifier_block *nb,
 static void shadow_context_descriptor_update(struct i915_gem_context *ctx,
 		struct intel_engine_cs *engine)
 {
-	struct intel_context *ce = &ctx->engine[engine->id];
+	struct intel_context *ce = to_intel_context(ctx, engine);
 	u64 desc = 0;
 
 	desc = ce->lrc_desc;
@@ -389,7 +389,7 @@ int intel_gvt_scan_and_shadow_workload(struct intel_vgpu_workload *workload)
 	 * shadow_ctx pages invalid. So gvt need to pin itself. After update
 	 * the guest context, gvt can unpin the shadow_ctx safely.
 	 */
-	ring = engine->context_pin(engine, shadow_ctx);
+	ring = intel_context_pin(shadow_ctx, engine);
 	if (IS_ERR(ring)) {
 		ret = PTR_ERR(ring);
 		gvt_vgpu_err("fail to pin shadow context\n");
@@ -403,7 +403,7 @@ int intel_gvt_scan_and_shadow_workload(struct intel_vgpu_workload *workload)
 	return 0;
 
 err_unpin:
-	engine->context_unpin(engine, shadow_ctx);
+	intel_context_unpin(shadow_ctx, engine);
 err_shadow:
 	release_shadow_wa_ctx(&workload->wa_ctx);
 err_scan:
@@ -437,7 +437,7 @@ static int intel_gvt_generate_request(struct intel_vgpu_workload *workload)
 	return 0;
 
 err_unpin:
-	engine->context_unpin(engine, shadow_ctx);
+	intel_context_unpin(shadow_ctx, engine);
 	release_shadow_wa_ctx(&workload->wa_ctx);
 	return ret;
 }
@@ -526,7 +526,7 @@ static int update_wa_ctx_2_shadow_ctx(struct intel_shadow_wa_ctx *wa_ctx)
 	struct intel_vgpu_submission *s = &workload->vgpu->submission;
 	struct i915_gem_context *shadow_ctx = s->shadow_ctx;
 	struct drm_i915_gem_object *ctx_obj =
-		shadow_ctx->engine[ring_id].state->obj;
+		shadow_ctx->__engine[ring_id].state->obj;
 	struct execlist_ring_context *shadow_ring_context;
 	struct page *page;
 
@@ -688,7 +688,7 @@ static int dispatch_workload(struct intel_vgpu_workload *workload)
 
 	ret = prepare_workload(workload);
 	if (ret) {
-		engine->context_unpin(engine, shadow_ctx);
+		intel_context_unpin(shadow_ctx, engine);
 		goto out;
 	}
 
@@ -771,7 +771,7 @@ static void update_guest_context(struct intel_vgpu_workload *workload)
 	struct i915_gem_context *shadow_ctx = s->shadow_ctx;
 	int ring_id = workload->ring_id;
 	struct drm_i915_gem_object *ctx_obj =
-		shadow_ctx->engine[ring_id].state->obj;
+		shadow_ctx->__engine[ring_id].state->obj;
 	struct execlist_ring_context *shadow_ring_context;
 	struct page *page;
 	void *src;
@@ -898,7 +898,7 @@ static void complete_current_workload(struct intel_gvt *gvt, int ring_id)
 		}
 		mutex_lock(&dev_priv->drm.struct_mutex);
 		/* unpin shadow ctx as the shadow_ctx update is done */
-		engine->context_unpin(engine, s->shadow_ctx);
+		intel_context_unpin(s->shadow_ctx, engine);
 		mutex_unlock(&dev_priv->drm.struct_mutex);
 	}
 
diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index 747dad2..85911bc 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -377,16 +377,19 @@ static void print_batch_pool_stats(struct seq_file *m,
 	print_file_stats(m, "[k]batch pool", stats);
 }
 
-static int per_file_ctx_stats(int id, void *ptr, void *data)
+static int per_file_ctx_stats(int idx, void *ptr, void *data)
 {
 	struct i915_gem_context *ctx = ptr;
-	int n;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+
+	for_each_engine(engine, ctx->i915, id) {
+		struct intel_context *ce = to_intel_context(ctx, engine);
 
-	for (n = 0; n < ARRAY_SIZE(ctx->engine); n++) {
-		if (ctx->engine[n].state)
-			per_file_stats(0, ctx->engine[n].state->obj, data);
-		if (ctx->engine[n].ring)
-			per_file_stats(0, ctx->engine[n].ring->vma->obj, data);
+		if (ce->state)
+			per_file_stats(0, ce->state->obj, data);
+		if (ce->ring)
+			per_file_stats(0, ce->ring->vma->obj, data);
 	}
 
 	return 0;
@@ -1959,7 +1962,8 @@ static int i915_context_status(struct seq_file *m, void *unused)
 		seq_putc(m, '\n');
 
 		for_each_engine(engine, dev_priv, id) {
-			struct intel_context *ce = &ctx->engine[engine->id];
+			struct intel_context *ce =
+				to_intel_context(ctx, engine);
 
 			seq_printf(m, "%s: ", engine->name);
 			if (ce->state)
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 6b0c67a..4090bfd 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -3234,7 +3234,7 @@ void i915_gem_reset(struct drm_i915_private *dev_priv,
 				      stalled_mask & ENGINE_MASK(id));
 		ctx = fetch_and_zero(&engine->last_retired_context);
 		if (ctx)
-			engine->context_unpin(engine, ctx);
+			intel_context_unpin(ctx, engine);
 
 		/*
 		 * Ostensibily, we always want a context loaded for powersaving,
@@ -5291,7 +5291,7 @@ static int __intel_engines_record_defaults(struct drm_i915_private *i915)
 	for_each_engine(engine, i915, id) {
 		struct i915_vma *state;
 
-		state = ctx->engine[id].state;
+		state = to_intel_context(ctx, engine)->state;
 		if (!state)
 			continue;
 
diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c
index 74435af..59d4bd4 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/i915_gem_context.c
@@ -117,15 +117,15 @@ static void lut_close(struct i915_gem_context *ctx)
 
 static void i915_gem_context_free(struct i915_gem_context *ctx)
 {
-	int i;
+	unsigned int n;
 
 	lockdep_assert_held(&ctx->i915->drm.struct_mutex);
 	GEM_BUG_ON(!i915_gem_context_is_closed(ctx));
 
 	i915_ppgtt_put(ctx->ppgtt);
 
-	for (i = 0; i < I915_NUM_ENGINES; i++) {
-		struct intel_context *ce = &ctx->engine[i];
+	for (n = 0; n < ARRAY_SIZE(ctx->__engine); n++) {
+		struct intel_context *ce = &ctx->__engine[n];
 
 		if (!ce->state)
 			continue;
@@ -521,7 +521,7 @@ void i915_gem_contexts_lost(struct drm_i915_private *dev_priv)
 		if (!engine->last_retired_context)
 			continue;
 
-		engine->context_unpin(engine, engine->last_retired_context);
+		intel_context_unpin(engine->last_retired_context, engine);
 		engine->last_retired_context = NULL;
 	}
 }
diff --git a/drivers/gpu/drm/i915/i915_gem_context.h b/drivers/gpu/drm/i915/i915_gem_context.h
index b12a8a8..ace3b12 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.h
+++ b/drivers/gpu/drm/i915/i915_gem_context.h
@@ -149,7 +149,7 @@ struct i915_gem_context {
 		u32 *lrc_reg_state;
 		u64 lrc_desc;
 		int pin_count;
-	} engine[I915_NUM_ENGINES];
+	} __engine[I915_NUM_ENGINES];
 
 	/** ring_size: size for allocating the per-engine ring buffer */
 	u32 ring_size;
@@ -256,6 +256,34 @@ static inline bool i915_gem_context_is_kernel(struct i915_gem_context *ctx)
 	return !ctx->file_priv;
 }
 
+static inline struct intel_context *
+to_intel_context(struct i915_gem_context *ctx,
+		 const struct intel_engine_cs *engine)
+{
+	return &ctx->__engine[engine->id];
+}
+
+static inline struct intel_ring *
+intel_context_pin(struct i915_gem_context *ctx, struct intel_engine_cs *engine)
+{
+	return engine->context_pin(engine, ctx);
+}
+
+static inline void __intel_context_pin(struct i915_gem_context *ctx,
+				       const struct intel_engine_cs *engine)
+{
+	struct intel_context *ce = to_intel_context(ctx, engine);
+
+	GEM_BUG_ON(!ce->pin_count);
+	ce->pin_count++;
+}
+
+static inline void intel_context_unpin(struct i915_gem_context *ctx,
+				       struct intel_engine_cs *engine)
+{
+	engine->context_unpin(engine, ctx);
+}
+
 /* i915_gem_context.c */
 int __must_check i915_gem_contexts_init(struct drm_i915_private *dev_priv);
 void i915_gem_contexts_lost(struct drm_i915_private *dev_priv);
diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c
index 671ffa3..c012796 100644
--- a/drivers/gpu/drm/i915/i915_gpu_error.c
+++ b/drivers/gpu/drm/i915/i915_gpu_error.c
@@ -1472,7 +1472,8 @@ static void gem_record_rings(struct i915_gpu_state *error)
 
 			ee->ctx =
 				i915_error_object_create(i915,
-							 request->ctx->engine[i].state);
+							 to_intel_context(request->ctx,
+									  engine)->state);
 
 			error->simulated |=
 				i915_gem_context_no_error_capture(request->ctx);
diff --git a/drivers/gpu/drm/i915/i915_perf.c b/drivers/gpu/drm/i915/i915_perf.c
index bfc906c..4b1da01 100644
--- a/drivers/gpu/drm/i915/i915_perf.c
+++ b/drivers/gpu/drm/i915/i915_perf.c
@@ -1234,7 +1234,7 @@ static int oa_get_render_ctx_id(struct i915_perf_stream *stream)
 		 *
 		 * NB: implied RCS engine...
 		 */
-		ring = engine->context_pin(engine, stream->ctx);
+		ring = intel_context_pin(stream->ctx, engine);
 		mutex_unlock(&dev_priv->drm.struct_mutex);
 		if (IS_ERR(ring))
 			return PTR_ERR(ring);
@@ -1246,7 +1246,7 @@ static int oa_get_render_ctx_id(struct i915_perf_stream *stream)
 		 * with gen8+ and execlists
 		 */
 		dev_priv->perf.oa.specific_ctx_id =
-			i915_ggtt_offset(stream->ctx->engine[engine->id].state);
+			i915_ggtt_offset(to_intel_context(stream->ctx, engine)->state);
 	}
 
 	return 0;
@@ -1271,7 +1271,7 @@ static void oa_put_render_ctx_id(struct i915_perf_stream *stream)
 		mutex_lock(&dev_priv->drm.struct_mutex);
 
 		dev_priv->perf.oa.specific_ctx_id = INVALID_CTX_ID;
-		engine->context_unpin(engine, stream->ctx);
+		intel_context_unpin(stream->ctx, engine);
 
 		mutex_unlock(&dev_priv->drm.struct_mutex);
 	}
@@ -1759,6 +1759,7 @@ static int gen8_switch_to_updated_kernel_context(struct drm_i915_private *dev_pr
 static int gen8_configure_all_contexts(struct drm_i915_private *dev_priv,
 				       const struct i915_oa_config *oa_config)
 {
+	struct intel_engine_cs *engine = dev_priv->engine[RCS];
 	struct i915_gem_context *ctx;
 	int ret;
 	unsigned int wait_flags = I915_WAIT_LOCKED;
@@ -1789,7 +1790,7 @@ static int gen8_configure_all_contexts(struct drm_i915_private *dev_priv,
 
 	/* Update all contexts now that we've stalled the submission. */
 	list_for_each_entry(ctx, &dev_priv->contexts.list, link) {
-		struct intel_context *ce = &ctx->engine[RCS];
+		struct intel_context *ce = to_intel_context(ctx, engine);
 		u32 *regs;
 
 		/* OA settings will be set upon first use */
diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c
index b1993d4..9358f2c 100644
--- a/drivers/gpu/drm/i915/i915_request.c
+++ b/drivers/gpu/drm/i915/i915_request.c
@@ -409,7 +409,7 @@ static void i915_request_retire(struct i915_request *request)
 	 * the subsequent request.
 	 */
 	if (engine->last_retired_context)
-		engine->context_unpin(engine, engine->last_retired_context);
+		intel_context_unpin(engine->last_retired_context, engine);
 	engine->last_retired_context = request->ctx;
 
 	spin_lock_irq(&request->lock);
@@ -638,7 +638,7 @@ i915_request_alloc(struct intel_engine_cs *engine, struct i915_gem_context *ctx)
 	 * GGTT space, so do this first before we reserve a seqno for
 	 * ourselves.
 	 */
-	ring = engine->context_pin(engine, ctx);
+	ring = intel_context_pin(ctx, engine);
 	if (IS_ERR(ring))
 		return ERR_CAST(ring);
 	GEM_BUG_ON(!ring);
@@ -787,7 +787,7 @@ err_unwind:
 err_unreserve:
 	unreserve_gt(i915);
 err_unpin:
-	engine->context_unpin(engine, ctx);
+	intel_context_unpin(ctx, engine);
 	return ERR_PTR(ret);
 }
 
diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
index eba81d5..238c8d3 100644
--- a/drivers/gpu/drm/i915/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/intel_engine_cs.c
@@ -685,7 +685,7 @@ int intel_engine_init_common(struct intel_engine_cs *engine)
 	 * be available. To avoid this we always pin the default
 	 * context.
 	 */
-	ring = engine->context_pin(engine, engine->i915->kernel_context);
+	ring = intel_context_pin(engine->i915->kernel_context, engine);
 	if (IS_ERR(ring))
 		return PTR_ERR(ring);
 
@@ -694,8 +694,7 @@ int intel_engine_init_common(struct intel_engine_cs *engine)
 	 * we can interrupt the engine at any time.
 	 */
 	if (engine->i915->preempt_context) {
-		ring = engine->context_pin(engine,
-					   engine->i915->preempt_context);
+		ring = intel_context_pin(engine->i915->preempt_context, engine);
 		if (IS_ERR(ring)) {
 			ret = PTR_ERR(ring);
 			goto err_unpin_kernel;
@@ -719,9 +718,9 @@ err_breadcrumbs:
 	intel_engine_fini_breadcrumbs(engine);
 err_unpin_preempt:
 	if (engine->i915->preempt_context)
-		engine->context_unpin(engine, engine->i915->preempt_context);
+		intel_context_unpin(engine->i915->preempt_context, engine);
 err_unpin_kernel:
-	engine->context_unpin(engine, engine->i915->kernel_context);
+	intel_context_unpin(engine->i915->kernel_context, engine);
 	return ret;
 }
 
@@ -749,8 +748,8 @@ void intel_engine_cleanup_common(struct intel_engine_cs *engine)
 		i915_gem_object_put(engine->default_state);
 
 	if (engine->i915->preempt_context)
-		engine->context_unpin(engine, engine->i915->preempt_context);
-	engine->context_unpin(engine, engine->i915->kernel_context);
+		intel_context_unpin(engine->i915->preempt_context, engine);
+	intel_context_unpin(engine->i915->kernel_context, engine);
 }
 
 u64 intel_engine_get_active_head(const struct intel_engine_cs *engine)
diff --git a/drivers/gpu/drm/i915/intel_guc_ads.c b/drivers/gpu/drm/i915/intel_guc_ads.c
index 334cb52..dcaa3fb7 100644
--- a/drivers/gpu/drm/i915/intel_guc_ads.c
+++ b/drivers/gpu/drm/i915/intel_guc_ads.c
@@ -121,7 +121,8 @@ int intel_guc_ads_create(struct intel_guc *guc)
 	 * to find it. Note that we have to skip our header (1 page),
 	 * because our GuC shared data is there.
 	 */
-	kernel_ctx_vma = dev_priv->kernel_context->engine[RCS].state;
+	kernel_ctx_vma = to_intel_context(dev_priv->kernel_context,
+					  dev_priv->engine[RCS])->state;
 	blob->ads.golden_context_lrca =
 		intel_guc_ggtt_offset(guc, kernel_ctx_vma) + skipped_offset;
 
diff --git a/drivers/gpu/drm/i915/intel_guc_submission.c b/drivers/gpu/drm/i915/intel_guc_submission.c
index 02da058..6e6ed0f 100644
--- a/drivers/gpu/drm/i915/intel_guc_submission.c
+++ b/drivers/gpu/drm/i915/intel_guc_submission.c
@@ -362,7 +362,7 @@ static void guc_stage_desc_init(struct intel_guc *guc,
 	desc->db_id = client->doorbell_id;
 
 	for_each_engine_masked(engine, dev_priv, client->engines, tmp) {
-		struct intel_context *ce = &ctx->engine[engine->id];
+		struct intel_context *ce = to_intel_context(ctx, engine);
 		u32 guc_engine_id = engine->guc_id;
 		struct guc_execlist_context *lrc = &desc->lrc[guc_engine_id];
 
@@ -990,7 +990,8 @@ static void guc_fill_preempt_context(struct intel_guc *guc)
 	enum intel_engine_id id;
 
 	for_each_engine(engine, dev_priv, id) {
-		struct intel_context *ce = &client->owner->engine[id];
+		struct intel_context *ce =
+			to_intel_context(client->owner, engine);
 		u32 addr = intel_hws_preempt_done_address(engine);
 		u32 *cs;
 
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 58cad24..0999956 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -223,7 +223,7 @@ static void
 intel_lr_context_descriptor_update(struct i915_gem_context *ctx,
 				   struct intel_engine_cs *engine)
 {
-	struct intel_context *ce = &ctx->engine[engine->id];
+	struct intel_context *ce = to_intel_context(ctx, engine);
 	u64 desc;
 
 	BUILD_BUG_ON(MAX_CONTEXT_HW_ID > (BIT(GEN8_CTX_ID_WIDTH)));
@@ -414,7 +414,7 @@ execlists_update_context_pdps(struct i915_hw_ppgtt *ppgtt, u32 *reg_state)
 
 static u64 execlists_update_context(struct i915_request *rq)
 {
-	struct intel_context *ce = &rq->ctx->engine[rq->engine->id];
+	struct intel_context *ce = to_intel_context(rq->ctx, rq->engine);
 	struct i915_hw_ppgtt *ppgtt =
 		rq->ctx->ppgtt ?: rq->i915->mm.aliasing_ppgtt;
 	u32 *reg_state = ce->lrc_reg_state;
@@ -523,7 +523,7 @@ static void inject_preempt_context(struct intel_engine_cs *engine)
 {
 	struct intel_engine_execlists *execlists = &engine->execlists;
 	struct intel_context *ce =
-		&engine->i915->preempt_context->engine[engine->id];
+		to_intel_context(engine->i915->preempt_context, engine);
 	unsigned int n;
 
 	GEM_BUG_ON(execlists->preempt_complete_status !=
@@ -1327,7 +1327,7 @@ static struct intel_ring *
 execlists_context_pin(struct intel_engine_cs *engine,
 		      struct i915_gem_context *ctx)
 {
-	struct intel_context *ce = &ctx->engine[engine->id];
+	struct intel_context *ce = to_intel_context(ctx, engine);
 	void *vaddr;
 	int ret;
 
@@ -1380,7 +1380,7 @@ err:
 static void execlists_context_unpin(struct intel_engine_cs *engine,
 				    struct i915_gem_context *ctx)
 {
-	struct intel_context *ce = &ctx->engine[engine->id];
+	struct intel_context *ce = to_intel_context(ctx, engine);
 
 	lockdep_assert_held(&ctx->i915->drm.struct_mutex);
 	GEM_BUG_ON(ce->pin_count == 0);
@@ -1399,8 +1399,8 @@ static void execlists_context_unpin(struct intel_engine_cs *engine,
 
 static int execlists_request_alloc(struct i915_request *request)
 {
-	struct intel_engine_cs *engine = request->engine;
-	struct intel_context *ce = &request->ctx->engine[engine->id];
+	struct intel_context *ce =
+		to_intel_context(request->ctx, request->engine);
 	int ret;
 
 	GEM_BUG_ON(!ce->pin_count);
@@ -1854,7 +1854,7 @@ static void reset_common_ring(struct intel_engine_cs *engine,
 	 * future request will be after userspace has had the opportunity
 	 * to recreate its own state.
 	 */
-	regs = request->ctx->engine[engine->id].lrc_reg_state;
+	regs = to_intel_context(request->ctx, engine)->lrc_reg_state;
 	if (engine->default_state) {
 		void *defaults;
 
@@ -2305,9 +2305,13 @@ static int logical_ring_init(struct intel_engine_cs *engine)
 	}
 
 	engine->execlists.preempt_complete_status = ~0u;
-	if (engine->i915->preempt_context)
+	if (engine->i915->preempt_context) {
+		struct intel_context *ce =
+			to_intel_context(engine->i915->preempt_context, engine);
+
 		engine->execlists.preempt_complete_status =
-			upper_32_bits(engine->i915->preempt_context->engine[engine->id].lrc_desc);
+			upper_32_bits(ce->lrc_desc);
+	}
 
 	return 0;
 
@@ -2589,7 +2593,7 @@ static int execlists_context_deferred_alloc(struct i915_gem_context *ctx,
 					    struct intel_engine_cs *engine)
 {
 	struct drm_i915_gem_object *ctx_obj;
-	struct intel_context *ce = &ctx->engine[engine->id];
+	struct intel_context *ce = to_intel_context(ctx, engine);
 	struct i915_vma *vma;
 	uint32_t context_size;
 	struct intel_ring *ring;
@@ -2660,7 +2664,8 @@ void intel_lr_context_resume(struct drm_i915_private *dev_priv)
 	 */
 	list_for_each_entry(ctx, &dev_priv->contexts.list, link) {
 		for_each_engine(engine, dev_priv, id) {
-			struct intel_context *ce = &ctx->engine[engine->id];
+			struct intel_context *ce =
+				to_intel_context(ctx, engine);
 			u32 *reg;
 
 			if (!ce->state)
diff --git a/drivers/gpu/drm/i915/intel_lrc.h b/drivers/gpu/drm/i915/intel_lrc.h
index 59d7b86..4ec7d8d 100644
--- a/drivers/gpu/drm/i915/intel_lrc.h
+++ b/drivers/gpu/drm/i915/intel_lrc.h
@@ -108,7 +108,7 @@ static inline uint64_t
 intel_lr_context_descriptor(struct i915_gem_context *ctx,
 			    struct intel_engine_cs *engine)
 {
-	return ctx->engine[engine->id].lrc_desc;
+	return to_intel_context(ctx, engine)->lrc_desc;
 }
 
 #endif /* _INTEL_LRC_H_ */
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index c06c22c..69ffc0d 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -558,7 +558,8 @@ static void reset_ring_common(struct intel_engine_cs *engine,
 	 */
 	if (request) {
 		struct drm_i915_private *dev_priv = request->i915;
-		struct intel_context *ce = &request->ctx->engine[engine->id];
+		struct intel_context *ce = to_intel_context(request->ctx,
+							    engine);
 		struct i915_hw_ppgtt *ppgtt;
 
 		if (ce->state) {
@@ -1163,9 +1164,9 @@ intel_ring_free(struct intel_ring *ring)
 	kfree(ring);
 }
 
-static int context_pin(struct i915_gem_context *ctx)
+static int context_pin(struct intel_context *ce)
 {
-	struct i915_vma *vma = ctx->engine[RCS].state;
+	struct i915_vma *vma = ce->state;
 	int ret;
 
 	/*
@@ -1256,7 +1257,7 @@ static struct intel_ring *
 intel_ring_context_pin(struct intel_engine_cs *engine,
 		       struct i915_gem_context *ctx)
 {
-	struct intel_context *ce = &ctx->engine[engine->id];
+	struct intel_context *ce = to_intel_context(ctx, engine);
 	int ret;
 
 	lockdep_assert_held(&ctx->i915->drm.struct_mutex);
@@ -1278,7 +1279,7 @@ intel_ring_context_pin(struct intel_engine_cs *engine,
 	}
 
 	if (ce->state) {
-		ret = context_pin(ctx);
+		ret = context_pin(ce);
 		if (ret)
 			goto err;
 
@@ -1299,7 +1300,7 @@ err:
 static void intel_ring_context_unpin(struct intel_engine_cs *engine,
 				     struct i915_gem_context *ctx)
 {
-	struct intel_context *ce = &ctx->engine[engine->id];
+	struct intel_context *ce = to_intel_context(ctx, engine);
 
 	lockdep_assert_held(&ctx->i915->drm.struct_mutex);
 	GEM_BUG_ON(ce->pin_count == 0);
@@ -1427,7 +1428,7 @@ static inline int mi_set_context(struct i915_request *rq, u32 flags)
 
 	*cs++ = MI_NOOP;
 	*cs++ = MI_SET_CONTEXT;
-	*cs++ = i915_ggtt_offset(rq->ctx->engine[RCS].state) | flags;
+	*cs++ = i915_ggtt_offset(to_intel_context(rq->ctx, engine)->state) | flags;
 	/*
 	 * w/a: MI_SET_CONTEXT must always be followed by MI_NOOP
 	 * WaMiSetContext_Hang:snb,ivb,vlv
@@ -1518,7 +1519,7 @@ static int switch_context(struct i915_request *rq)
 		hw_flags = MI_FORCE_RESTORE;
 	}
 
-	if (to_ctx->engine[engine->id].state &&
+	if (to_intel_context(to_ctx, engine)->state &&
 	    (to_ctx != from_ctx || hw_flags & MI_FORCE_RESTORE)) {
 		GEM_BUG_ON(engine->id != RCS);
 
@@ -1566,7 +1567,7 @@ static int ring_request_alloc(struct i915_request *request)
 {
 	int ret;
 
-	GEM_BUG_ON(!request->ctx->engine[request->engine->id].pin_count);
+	GEM_BUG_ON(!to_intel_context(request->ctx, request->engine)->pin_count);
 
 	/* Flush enough space to reduce the likelihood of waiting after
 	 * we start building the request - in which case we will just
diff --git a/drivers/gpu/drm/i915/selftests/mock_engine.c b/drivers/gpu/drm/i915/selftests/mock_engine.c
index 78a89ef..b82420c 100644
--- a/drivers/gpu/drm/i915/selftests/mock_engine.c
+++ b/drivers/gpu/drm/i915/selftests/mock_engine.c
@@ -71,14 +71,21 @@ static struct intel_ring *
 mock_context_pin(struct intel_engine_cs *engine,
 		 struct i915_gem_context *ctx)
 {
-	i915_gem_context_get(ctx);
+	struct intel_context *ce = to_intel_context(ctx, engine);
+
+	if (!ce->pin_count++)
+		i915_gem_context_get(ctx);
+
 	return engine->buffer;
 }
 
 static void mock_context_unpin(struct intel_engine_cs *engine,
 			       struct i915_gem_context *ctx)
 {
-	i915_gem_context_put(ctx);
+	struct intel_context *ce = to_intel_context(ctx, engine);
+
+	if (!--ce->pin_count)
+		i915_gem_context_put(ctx);
 }
 
 static int mock_request_alloc(struct i915_request *request)
@@ -217,7 +224,7 @@ void mock_engine_free(struct intel_engine_cs *engine)
 	GEM_BUG_ON(timer_pending(&mock->hw_delay));
 
 	if (engine->last_retired_context)
-		engine->context_unpin(engine, engine->last_retired_context);
+		intel_context_unpin(engine->last_retired_context, engine);
 
 	intel_engine_fini_breadcrumbs(engine);
 
-- 
cgit v1.1


From b887d61546245389c0304d8b1371bab9af8106c2 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Mon, 30 Apr 2018 14:15:02 +0100
Subject: drm/i915: Retire requests along rings

In the next patch, rings are the central timeline as requests may jump
between engines. Therefore in the future as we retire in order along the
engine timeline, we may retire out-of-order within a ring (as the ring now
occurs along multiple engines), leading to much hilarity in miscomputing
the position of ring->head.

As an added bonus, retiring along the ring reduces the penalty of having
one execlists client do cleanup for another (old legacy submission
shares a ring between all clients). The downside is that slow and
irregular (off the critical path) process of cleaning up stale requests
after userspace becomes a modicum less efficient.

In the long run, it will become apparent that the ordered
ring->request_list matches the ring->timeline, a fun challenge for the
future will be unifying the two lists to avoid duplication!

v2: We need both engine-order and ring-order processing to maintain our
knowledge of where individual rings have completed upto as well as
knowing what was last executing on any engine. And finally by decoupling
retiring the contexts on the engine and the timelines along the rings,
we do have to keep a reference to the context on each request
(previously it was guaranteed by the context being pinned).

v3: Not just a reference to the context, but we need to keep it pinned
as we manipulate the rings; i.e. we need a pin for both the manipulation
of the engine state during its retirements, and a separate pin for the
manipulation of the ring state.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180430131503.5375-3-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/i915_drv.h                  |   3 +-
 drivers/gpu/drm/i915/i915_gem.c                  |   1 +
 drivers/gpu/drm/i915/i915_request.c              | 150 ++++++++++++++---------
 drivers/gpu/drm/i915/i915_utils.h                |   6 +
 drivers/gpu/drm/i915/intel_ringbuffer.c          |   6 +-
 drivers/gpu/drm/i915/intel_ringbuffer.h          |   1 +
 drivers/gpu/drm/i915/selftests/mock_engine.c     |  27 ++--
 drivers/gpu/drm/i915/selftests/mock_gem_device.c |   2 +
 8 files changed, 131 insertions(+), 65 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index dd4d6b9..edc33e0 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -2059,8 +2059,9 @@ struct drm_i915_private {
 		void (*resume)(struct drm_i915_private *);
 		void (*cleanup_engine)(struct intel_engine_cs *engine);
 
-		struct list_head timelines;
 		struct i915_gem_timeline global_timeline;
+		struct list_head timelines;
+		struct list_head rings;
 		u32 active_requests;
 		u32 request_serial;
 
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 4090bfd..f0644d1 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -5600,6 +5600,7 @@ int i915_gem_init_early(struct drm_i915_private *dev_priv)
 		goto err_dependencies;
 
 	mutex_lock(&dev_priv->drm.struct_mutex);
+	INIT_LIST_HEAD(&dev_priv->gt.rings);
 	INIT_LIST_HEAD(&dev_priv->gt.timelines);
 	err = i915_gem_timeline_init__global(dev_priv);
 	mutex_unlock(&dev_priv->drm.struct_mutex);
diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c
index 9358f2c..e653525 100644
--- a/drivers/gpu/drm/i915/i915_request.c
+++ b/drivers/gpu/drm/i915/i915_request.c
@@ -286,6 +286,7 @@ static int reserve_gt(struct drm_i915_private *i915)
 
 static void unreserve_gt(struct drm_i915_private *i915)
 {
+	GEM_BUG_ON(!i915->gt.active_requests);
 	if (!--i915->gt.active_requests)
 		i915_gem_park(i915);
 }
@@ -298,6 +299,7 @@ void i915_gem_retire_noop(struct i915_gem_active *active,
 
 static void advance_ring(struct i915_request *request)
 {
+	struct intel_ring *ring = request->ring;
 	unsigned int tail;
 
 	/*
@@ -309,7 +311,8 @@ static void advance_ring(struct i915_request *request)
 	 * Note this requires that we are always called in request
 	 * completion order.
 	 */
-	if (list_is_last(&request->ring_link, &request->ring->request_list)) {
+	GEM_BUG_ON(!list_is_first(&request->ring_link, &ring->request_list));
+	if (list_is_last(&request->ring_link, &ring->request_list)) {
 		/*
 		 * We may race here with execlists resubmitting this request
 		 * as we retire it. The resubmission will move the ring->tail
@@ -322,9 +325,9 @@ static void advance_ring(struct i915_request *request)
 	} else {
 		tail = request->postfix;
 	}
-	list_del(&request->ring_link);
+	list_del_init(&request->ring_link);
 
-	request->ring->head = tail;
+	ring->head = tail;
 }
 
 static void free_capture_list(struct i915_request *request)
@@ -340,30 +343,84 @@ static void free_capture_list(struct i915_request *request)
 	}
 }
 
+static void __retire_engine_request(struct intel_engine_cs *engine,
+				    struct i915_request *rq)
+{
+	GEM_TRACE("%s(%s) fence %llx:%d, global=%d, current %d\n",
+		  __func__, engine->name,
+		  rq->fence.context, rq->fence.seqno,
+		  rq->global_seqno,
+		  intel_engine_get_seqno(engine));
+
+	GEM_BUG_ON(!i915_request_completed(rq));
+
+	local_irq_disable();
+
+	spin_lock(&engine->timeline->lock);
+	GEM_BUG_ON(!list_is_first(&rq->link, &engine->timeline->requests));
+	list_del_init(&rq->link);
+	spin_unlock(&engine->timeline->lock);
+
+	spin_lock(&rq->lock);
+	if (!test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &rq->fence.flags))
+		dma_fence_signal_locked(&rq->fence);
+	if (test_bit(DMA_FENCE_FLAG_ENABLE_SIGNAL_BIT, &rq->fence.flags))
+		intel_engine_cancel_signaling(rq);
+	if (rq->waitboost) {
+		GEM_BUG_ON(!atomic_read(&rq->i915->gt_pm.rps.num_waiters));
+		atomic_dec(&rq->i915->gt_pm.rps.num_waiters);
+	}
+	spin_unlock(&rq->lock);
+
+	local_irq_enable();
+
+	/*
+	 * The backing object for the context is done after switching to the
+	 * *next* context. Therefore we cannot retire the previous context until
+	 * the next context has already started running. However, since we
+	 * cannot take the required locks at i915_request_submit() we
+	 * defer the unpinning of the active context to now, retirement of
+	 * the subsequent request.
+	 */
+	if (engine->last_retired_context)
+		intel_context_unpin(engine->last_retired_context, engine);
+	engine->last_retired_context = rq->ctx;
+}
+
+static void __retire_engine_upto(struct intel_engine_cs *engine,
+				 struct i915_request *rq)
+{
+	struct i915_request *tmp;
+
+	if (list_empty(&rq->link))
+		return;
+
+	do {
+		tmp = list_first_entry(&engine->timeline->requests,
+				       typeof(*tmp), link);
+
+		GEM_BUG_ON(tmp->engine != engine);
+		__retire_engine_request(engine, tmp);
+	} while (tmp != rq);
+}
+
 static void i915_request_retire(struct i915_request *request)
 {
-	struct intel_engine_cs *engine = request->engine;
 	struct i915_gem_active *active, *next;
 
 	GEM_TRACE("%s fence %llx:%d, global=%d, current %d\n",
-		  engine->name,
+		  request->engine->name,
 		  request->fence.context, request->fence.seqno,
 		  request->global_seqno,
-		  intel_engine_get_seqno(engine));
+		  intel_engine_get_seqno(request->engine));
 
 	lockdep_assert_held(&request->i915->drm.struct_mutex);
 	GEM_BUG_ON(!i915_sw_fence_signaled(&request->submit));
 	GEM_BUG_ON(!i915_request_completed(request));
-	GEM_BUG_ON(!request->i915->gt.active_requests);
 
 	trace_i915_request_retire(request);
 
-	spin_lock_irq(&engine->timeline->lock);
-	list_del_init(&request->link);
-	spin_unlock_irq(&engine->timeline->lock);
-
 	advance_ring(request);
-
 	free_capture_list(request);
 
 	/*
@@ -399,29 +456,9 @@ static void i915_request_retire(struct i915_request *request)
 
 	/* Retirement decays the ban score as it is a sign of ctx progress */
 	atomic_dec_if_positive(&request->ctx->ban_score);
+	intel_context_unpin(request->ctx, request->engine);
 
-	/*
-	 * The backing object for the context is done after switching to the
-	 * *next* context. Therefore we cannot retire the previous context until
-	 * the next context has already started running. However, since we
-	 * cannot take the required locks at i915_request_submit() we
-	 * defer the unpinning of the active context to now, retirement of
-	 * the subsequent request.
-	 */
-	if (engine->last_retired_context)
-		intel_context_unpin(engine->last_retired_context, engine);
-	engine->last_retired_context = request->ctx;
-
-	spin_lock_irq(&request->lock);
-	if (!test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &request->fence.flags))
-		dma_fence_signal_locked(&request->fence);
-	if (test_bit(DMA_FENCE_FLAG_ENABLE_SIGNAL_BIT, &request->fence.flags))
-		intel_engine_cancel_signaling(request);
-	if (request->waitboost) {
-		GEM_BUG_ON(!atomic_read(&request->i915->gt_pm.rps.num_waiters));
-		atomic_dec(&request->i915->gt_pm.rps.num_waiters);
-	}
-	spin_unlock_irq(&request->lock);
+	__retire_engine_upto(request->engine, request);
 
 	unreserve_gt(request->i915);
 
@@ -431,18 +468,24 @@ static void i915_request_retire(struct i915_request *request)
 
 void i915_request_retire_upto(struct i915_request *rq)
 {
-	struct intel_engine_cs *engine = rq->engine;
+	struct intel_ring *ring = rq->ring;
 	struct i915_request *tmp;
 
+	GEM_TRACE("%s fence %llx:%d, global=%d, current %d\n",
+		  rq->engine->name,
+		  rq->fence.context, rq->fence.seqno,
+		  rq->global_seqno,
+		  intel_engine_get_seqno(rq->engine));
+
 	lockdep_assert_held(&rq->i915->drm.struct_mutex);
 	GEM_BUG_ON(!i915_request_completed(rq));
 
-	if (list_empty(&rq->link))
+	if (list_empty(&rq->ring_link))
 		return;
 
 	do {
-		tmp = list_first_entry(&engine->timeline->requests,
-				       typeof(*tmp), link);
+		tmp = list_first_entry(&ring->request_list,
+				       typeof(*tmp), ring_link);
 
 		i915_request_retire(tmp);
 	} while (tmp != rq);
@@ -651,9 +694,9 @@ i915_request_alloc(struct intel_engine_cs *engine, struct i915_gem_context *ctx)
 	if (ret)
 		goto err_unreserve;
 
-	/* Move the oldest request to the slab-cache (if not in use!) */
-	rq = list_first_entry_or_null(&engine->timeline->requests,
-				      typeof(*rq), link);
+	/* Move our oldest request to the slab-cache (if not in use!) */
+	rq = list_first_entry_or_null(&ring->request_list,
+				      typeof(*rq), ring_link);
 	if (rq && i915_request_completed(rq))
 		i915_request_retire(rq);
 
@@ -771,6 +814,9 @@ i915_request_alloc(struct intel_engine_cs *engine, struct i915_gem_context *ctx)
 	if (ret)
 		goto err_unwind;
 
+	/* Keep a second pin for the dual retirement along engine and ring */
+	__intel_context_pin(rq->ctx, engine);
+
 	/* Check that we didn't interrupt ourselves with a new request */
 	GEM_BUG_ON(rq->timeline->seqno != rq->fence.seqno);
 	return rq;
@@ -1357,38 +1403,30 @@ complete:
 	return timeout;
 }
 
-static void engine_retire_requests(struct intel_engine_cs *engine)
+static void ring_retire_requests(struct intel_ring *ring)
 {
 	struct i915_request *request, *next;
-	u32 seqno = intel_engine_get_seqno(engine);
-	LIST_HEAD(retire);
 
-	spin_lock_irq(&engine->timeline->lock);
 	list_for_each_entry_safe(request, next,
-				 &engine->timeline->requests, link) {
-		if (!i915_seqno_passed(seqno, request->global_seqno))
+				 &ring->request_list, ring_link) {
+		if (!i915_request_completed(request))
 			break;
 
-		list_move_tail(&request->link, &retire);
-	}
-	spin_unlock_irq(&engine->timeline->lock);
-
-	list_for_each_entry_safe(request, next, &retire, link)
 		i915_request_retire(request);
+	}
 }
 
 void i915_retire_requests(struct drm_i915_private *i915)
 {
-	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
+	struct intel_ring *ring, *next;
 
 	lockdep_assert_held(&i915->drm.struct_mutex);
 
 	if (!i915->gt.active_requests)
 		return;
 
-	for_each_engine(engine, i915, id)
-		engine_retire_requests(engine);
+	list_for_each_entry_safe(ring, next, &i915->gt.rings, link)
+		ring_retire_requests(ring);
 }
 
 #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
diff --git a/drivers/gpu/drm/i915/i915_utils.h b/drivers/gpu/drm/i915/i915_utils.h
index 0695717..00165ad 100644
--- a/drivers/gpu/drm/i915/i915_utils.h
+++ b/drivers/gpu/drm/i915/i915_utils.h
@@ -120,6 +120,12 @@ static inline u64 ptr_to_u64(const void *ptr)
 
 #include <linux/list.h>
 
+static inline int list_is_first(const struct list_head *list,
+				const struct list_head *head)
+{
+	return head->next == list;
+}
+
 static inline void __list_del_many(struct list_head *head,
 				   struct list_head *first)
 {
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 69ffc0d..ae89580 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -1066,7 +1066,6 @@ err:
 
 void intel_ring_reset(struct intel_ring *ring, u32 tail)
 {
-	GEM_BUG_ON(!list_empty(&ring->request_list));
 	ring->tail = tail;
 	ring->head = tail;
 	ring->emit = tail;
@@ -1125,6 +1124,7 @@ intel_engine_create_ring(struct intel_engine_cs *engine, int size)
 
 	GEM_BUG_ON(!is_power_of_2(size));
 	GEM_BUG_ON(RING_CTL_SIZE(size) & ~RING_NR_PAGES);
+	lockdep_assert_held(&engine->i915->drm.struct_mutex);
 
 	ring = kzalloc(sizeof(*ring), GFP_KERNEL);
 	if (!ring)
@@ -1150,6 +1150,8 @@ intel_engine_create_ring(struct intel_engine_cs *engine, int size)
 	}
 	ring->vma = vma;
 
+	list_add(&ring->link, &engine->i915->gt.rings);
+
 	return ring;
 }
 
@@ -1161,6 +1163,8 @@ intel_ring_free(struct intel_ring *ring)
 	i915_vma_close(ring->vma);
 	__i915_gem_object_release_unless_active(obj);
 
+	list_del(&ring->link);
+
 	kfree(ring);
 }
 
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index 24af3f1..deb80d0 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -130,6 +130,7 @@ struct intel_ring {
 	void *vaddr;
 
 	struct list_head request_list;
+	struct list_head link;
 
 	u32 head;
 	u32 tail;
diff --git a/drivers/gpu/drm/i915/selftests/mock_engine.c b/drivers/gpu/drm/i915/selftests/mock_engine.c
index b82420c..d95fc48 100644
--- a/drivers/gpu/drm/i915/selftests/mock_engine.c
+++ b/drivers/gpu/drm/i915/selftests/mock_engine.c
@@ -147,9 +147,18 @@ static struct intel_ring *mock_ring(struct intel_engine_cs *engine)
 	INIT_LIST_HEAD(&ring->request_list);
 	intel_ring_update_space(ring);
 
+	list_add(&ring->link, &engine->i915->gt.rings);
+
 	return ring;
 }
 
+static void mock_ring_free(struct intel_ring *ring)
+{
+	list_del(&ring->link);
+
+	kfree(ring);
+}
+
 struct intel_engine_cs *mock_engine(struct drm_i915_private *i915,
 				    const char *name,
 				    int id)
@@ -162,12 +171,6 @@ struct intel_engine_cs *mock_engine(struct drm_i915_private *i915,
 	if (!engine)
 		return NULL;
 
-	engine->base.buffer = mock_ring(&engine->base);
-	if (!engine->base.buffer) {
-		kfree(engine);
-		return NULL;
-	}
-
 	/* minimal engine setup for requests */
 	engine->base.i915 = i915;
 	snprintf(engine->base.name, sizeof(engine->base.name), "%s", name);
@@ -192,7 +195,16 @@ struct intel_engine_cs *mock_engine(struct drm_i915_private *i915,
 	timer_setup(&engine->hw_delay, hw_delay_complete, 0);
 	INIT_LIST_HEAD(&engine->hw_queue);
 
+	engine->base.buffer = mock_ring(&engine->base);
+	if (!engine->base.buffer)
+		goto err_breadcrumbs;
+
 	return &engine->base;
+
+err_breadcrumbs:
+	intel_engine_fini_breadcrumbs(&engine->base);
+	kfree(engine);
+	return NULL;
 }
 
 void mock_engine_flush(struct intel_engine_cs *engine)
@@ -226,8 +238,9 @@ void mock_engine_free(struct intel_engine_cs *engine)
 	if (engine->last_retired_context)
 		intel_context_unpin(engine->last_retired_context, engine);
 
+	mock_ring_free(engine->buffer);
+
 	intel_engine_fini_breadcrumbs(engine);
 
-	kfree(engine->buffer);
 	kfree(engine);
 }
diff --git a/drivers/gpu/drm/i915/selftests/mock_gem_device.c b/drivers/gpu/drm/i915/selftests/mock_gem_device.c
index e6d4b88..ac4bacf 100644
--- a/drivers/gpu/drm/i915/selftests/mock_gem_device.c
+++ b/drivers/gpu/drm/i915/selftests/mock_gem_device.c
@@ -44,6 +44,7 @@ void mock_device_flush(struct drm_i915_private *i915)
 		mock_engine_flush(engine);
 
 	i915_retire_requests(i915);
+	GEM_BUG_ON(i915->gt.active_requests);
 }
 
 static void mock_device_release(struct drm_device *dev)
@@ -224,6 +225,7 @@ struct drm_i915_private *mock_gem_device(void)
 		goto err_dependencies;
 
 	mutex_lock(&i915->drm.struct_mutex);
+	INIT_LIST_HEAD(&i915->gt.rings);
 	INIT_LIST_HEAD(&i915->gt.timelines);
 	err = i915_gem_timeline_init__global(i915);
 	if (err) {
-- 
cgit v1.1


From 643b450a594e9cb57fbd2534d1571d244faddd01 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Mon, 30 Apr 2018 14:15:03 +0100
Subject: drm/i915: Only track live rings for retiring

We don't need to track every ring for its lifetime as they are managed
by the contexts/engines. What we do want to track are the live rings so
that we can sporadically clean up requests if userspace falls behind. We
can simply restrict the gt->rings list to being only gt->live_rings.

v2: s/live/active/ for consistency with gt.active_requests

Suggested-by: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180430131503.5375-4-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/i915_drv.h                  |  3 ++-
 drivers/gpu/drm/i915/i915_gem.c                  |  6 ++++--
 drivers/gpu/drm/i915/i915_request.c              | 10 ++++++++--
 drivers/gpu/drm/i915/intel_ringbuffer.c          |  4 ----
 drivers/gpu/drm/i915/intel_ringbuffer.h          |  2 +-
 drivers/gpu/drm/i915/selftests/mock_engine.c     |  4 ----
 drivers/gpu/drm/i915/selftests/mock_gem_device.c |  5 +++--
 7 files changed, 18 insertions(+), 16 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index edc33e0..6268a51 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -2061,7 +2061,8 @@ struct drm_i915_private {
 
 		struct i915_gem_timeline global_timeline;
 		struct list_head timelines;
-		struct list_head rings;
+
+		struct list_head active_rings;
 		u32 active_requests;
 		u32 request_serial;
 
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index f0644d1..fa1d94a 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -141,6 +141,7 @@ static u32 __i915_gem_park(struct drm_i915_private *i915)
 {
 	lockdep_assert_held(&i915->drm.struct_mutex);
 	GEM_BUG_ON(i915->gt.active_requests);
+	GEM_BUG_ON(!list_empty(&i915->gt.active_rings));
 
 	if (!i915->gt.awake)
 		return I915_EPOCH_INVALID;
@@ -5599,9 +5600,10 @@ int i915_gem_init_early(struct drm_i915_private *dev_priv)
 	if (!dev_priv->priorities)
 		goto err_dependencies;
 
-	mutex_lock(&dev_priv->drm.struct_mutex);
-	INIT_LIST_HEAD(&dev_priv->gt.rings);
 	INIT_LIST_HEAD(&dev_priv->gt.timelines);
+	INIT_LIST_HEAD(&dev_priv->gt.active_rings);
+
+	mutex_lock(&dev_priv->drm.struct_mutex);
 	err = i915_gem_timeline_init__global(dev_priv);
 	mutex_unlock(&dev_priv->drm.struct_mutex);
 	if (err)
diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c
index e653525..c8fc4b3 100644
--- a/drivers/gpu/drm/i915/i915_request.c
+++ b/drivers/gpu/drm/i915/i915_request.c
@@ -322,6 +322,7 @@ static void advance_ring(struct i915_request *request)
 		 * noops - they are safe to be replayed on a reset.
 		 */
 		tail = READ_ONCE(request->tail);
+		list_del(&ring->active_link);
 	} else {
 		tail = request->postfix;
 	}
@@ -1096,6 +1097,8 @@ void __i915_request_add(struct i915_request *request, bool flush_caches)
 	i915_gem_active_set(&timeline->last_request, request);
 
 	list_add_tail(&request->ring_link, &ring->request_list);
+	if (list_is_first(&request->ring_link, &ring->request_list))
+		list_add(&ring->active_link, &request->i915->gt.active_rings);
 	request->emitted_jiffies = jiffies;
 
 	/*
@@ -1418,14 +1421,17 @@ static void ring_retire_requests(struct intel_ring *ring)
 
 void i915_retire_requests(struct drm_i915_private *i915)
 {
-	struct intel_ring *ring, *next;
+	struct intel_ring *ring, *tmp;
 
 	lockdep_assert_held(&i915->drm.struct_mutex);
 
 	if (!i915->gt.active_requests)
 		return;
 
-	list_for_each_entry_safe(ring, next, &i915->gt.rings, link)
+	/* An outstanding request must be on a still active ring somewhere */
+	GEM_BUG_ON(list_empty(&i915->gt.active_rings));
+
+	list_for_each_entry_safe(ring, tmp, &i915->gt.active_rings, active_link)
 		ring_retire_requests(ring);
 }
 
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index ae89580..007449c 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -1150,8 +1150,6 @@ intel_engine_create_ring(struct intel_engine_cs *engine, int size)
 	}
 	ring->vma = vma;
 
-	list_add(&ring->link, &engine->i915->gt.rings);
-
 	return ring;
 }
 
@@ -1163,8 +1161,6 @@ intel_ring_free(struct intel_ring *ring)
 	i915_vma_close(ring->vma);
 	__i915_gem_object_release_unless_active(obj);
 
-	list_del(&ring->link);
-
 	kfree(ring);
 }
 
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index deb80d0..fd679ce 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -130,7 +130,7 @@ struct intel_ring {
 	void *vaddr;
 
 	struct list_head request_list;
-	struct list_head link;
+	struct list_head active_link;
 
 	u32 head;
 	u32 tail;
diff --git a/drivers/gpu/drm/i915/selftests/mock_engine.c b/drivers/gpu/drm/i915/selftests/mock_engine.c
index d95fc48..19175dd 100644
--- a/drivers/gpu/drm/i915/selftests/mock_engine.c
+++ b/drivers/gpu/drm/i915/selftests/mock_engine.c
@@ -147,15 +147,11 @@ static struct intel_ring *mock_ring(struct intel_engine_cs *engine)
 	INIT_LIST_HEAD(&ring->request_list);
 	intel_ring_update_space(ring);
 
-	list_add(&ring->link, &engine->i915->gt.rings);
-
 	return ring;
 }
 
 static void mock_ring_free(struct intel_ring *ring)
 {
-	list_del(&ring->link);
-
 	kfree(ring);
 }
 
diff --git a/drivers/gpu/drm/i915/selftests/mock_gem_device.c b/drivers/gpu/drm/i915/selftests/mock_gem_device.c
index ac4bacf..f22a2b3 100644
--- a/drivers/gpu/drm/i915/selftests/mock_gem_device.c
+++ b/drivers/gpu/drm/i915/selftests/mock_gem_device.c
@@ -224,9 +224,10 @@ struct drm_i915_private *mock_gem_device(void)
 	if (!i915->priorities)
 		goto err_dependencies;
 
-	mutex_lock(&i915->drm.struct_mutex);
-	INIT_LIST_HEAD(&i915->gt.rings);
 	INIT_LIST_HEAD(&i915->gt.timelines);
+	INIT_LIST_HEAD(&i915->gt.active_rings);
+
+	mutex_lock(&i915->drm.struct_mutex);
 	err = i915_gem_timeline_init__global(i915);
 	if (err) {
 		mutex_unlock(&i915->drm.struct_mutex);
-- 
cgit v1.1


From fc8cec113904a47396bf0a1afc62920d66319d36 Mon Sep 17 00:00:00 2001
From: Mikulas Patocka <mpatocka@redhat.com>
Date: Tue, 17 Apr 2018 18:32:26 -0400
Subject: dm integrity: use kvfree for kvmalloc'd memory

Use kvfree instead of kfree because the array is allocated with kvmalloc.

Fixes: 7eada909bfd7a ("dm: add integrity target")
Cc: stable@vger.kernel.org # v4.12+
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
---
 drivers/md/dm-integrity.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/md/dm-integrity.c b/drivers/md/dm-integrity.c
index 77d9fe5..514fb4a 100644
--- a/drivers/md/dm-integrity.c
+++ b/drivers/md/dm-integrity.c
@@ -2440,7 +2440,7 @@ static void dm_integrity_free_journal_scatterlist(struct dm_integrity_c *ic, str
 	unsigned i;
 	for (i = 0; i < ic->journal_sections; i++)
 		kvfree(sl[i]);
-	kfree(sl);
+	kvfree(sl);
 }
 
 static struct scatterlist **dm_integrity_alloc_journal_scatterlist(struct dm_integrity_c *ic, struct page_list *pl)
-- 
cgit v1.1


From f7879b4cea4b7867e7382efdbd805fbe35835337 Mon Sep 17 00:00:00 2001
From: Mikulas Patocka <mpatocka@redhat.com>
Date: Thu, 19 Apr 2018 08:33:00 -0400
Subject: dm bufio: fix buffer alignment

Commit 6b5e718cc138 ("dm bufio: relax alignment constraint on slab
cache") relaxed alignment on dm-bufio cache, however it may break
dm-crypt or dm-integrity.

dm-crypt and dm-integrity require that the size of bio vector entries
(bv_len) is aligned on its sector size. bv_offset doesn't have to be
aligned, but bv_len must be. XFS sends unaligned bios, but they do not
cross page boundary, so the requirement for aligned bv_len is met.

Commit 6b5e718cc138 made dm-bufio send unaligned bios that cross page
boundary, this could break dm-crypt and dm-integrity.

Reinstates the alignment. Note that misaligned entries only happen when
we use slab/slub debugging. Without debugging, the entries are always
aligned.

Fixes: 6b5e718cc138 ("dm bufio: relax alignment constraint on slab cache")
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
---
 drivers/md/dm-bufio.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/md/dm-bufio.c b/drivers/md/dm-bufio.c
index 12aa9ca..dc385b7 100644
--- a/drivers/md/dm-bufio.c
+++ b/drivers/md/dm-bufio.c
@@ -1681,8 +1681,9 @@ struct dm_bufio_client *dm_bufio_client_create(struct block_device *bdev, unsign
 
 	if (block_size <= KMALLOC_MAX_SIZE &&
 	    (block_size < PAGE_SIZE || !is_power_of_2(block_size))) {
-		snprintf(slab_name, sizeof slab_name, "dm_bufio_cache-%u", c->block_size);
-		c->slab_cache = kmem_cache_create(slab_name, c->block_size, ARCH_KMALLOC_MINALIGN,
+		unsigned align = min(1U << __ffs(block_size), (unsigned)PAGE_SIZE);
+		snprintf(slab_name, sizeof slab_name, "dm_bufio_cache-%u", block_size);
+		c->slab_cache = kmem_cache_create(slab_name, block_size, align,
 						  SLAB_RECLAIM_ACCOUNT, NULL);
 		if (!c->slab_cache) {
 			r = -ENOMEM;
-- 
cgit v1.1


From d66a270be3310d7aa132fec0cea77d3d32a0ff75 Mon Sep 17 00:00:00 2001
From: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Date: Thu, 15 Mar 2018 08:44:24 -0400
Subject: tracepoint: Do not warn on ENOMEM

Tracepoint should only warn when a kernel API user does not respect the
required preconditions (e.g. same tracepoint enabled twice, or called
to remove a tracepoint that does not exist).

Silence warning in out-of-memory conditions, given that the error is
returned to the caller.

This ensures that out-of-memory error-injection testing does not trigger
warnings in tracepoint.c, which were seen by syzbot.

Link: https://lkml.kernel.org/r/001a114465e241a8720567419a72@google.com
Link: https://lkml.kernel.org/r/001a1140e0de15fc910567464190@google.com
Link: http://lkml.kernel.org/r/20180315124424.32319-1-mathieu.desnoyers@efficios.com

CC: Peter Zijlstra <peterz@infradead.org>
CC: Jiri Olsa <jolsa@redhat.com>
CC: Arnaldo Carvalho de Melo <acme@kernel.org>
CC: Alexander Shishkin <alexander.shishkin@linux.intel.com>
CC: Namhyung Kim <namhyung@kernel.org>
CC: stable@vger.kernel.org
Fixes: de7b2973903c6 ("tracepoint: Use struct pointer instead of name hash for reg/unreg tracepoints")
Reported-by: syzbot+9c0d616860575a73166a@syzkaller.appspotmail.com
Reported-by: syzbot+4e9ae7fa46233396f64d@syzkaller.appspotmail.com
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
---
 kernel/tracepoint.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/kernel/tracepoint.c b/kernel/tracepoint.c
index 671b134..1e37da2 100644
--- a/kernel/tracepoint.c
+++ b/kernel/tracepoint.c
@@ -207,7 +207,7 @@ static int tracepoint_add_func(struct tracepoint *tp,
 			lockdep_is_held(&tracepoints_mutex));
 	old = func_add(&tp_funcs, func, prio);
 	if (IS_ERR(old)) {
-		WARN_ON_ONCE(1);
+		WARN_ON_ONCE(PTR_ERR(old) != -ENOMEM);
 		return PTR_ERR(old);
 	}
 
@@ -239,7 +239,7 @@ static int tracepoint_remove_func(struct tracepoint *tp,
 			lockdep_is_held(&tracepoints_mutex));
 	old = func_remove(&tp_funcs, func);
 	if (IS_ERR(old)) {
-		WARN_ON_ONCE(1);
+		WARN_ON_ONCE(PTR_ERR(old) != -ENOMEM);
 		return PTR_ERR(old);
 	}
 
-- 
cgit v1.1


From 2b63f1329df2cd814c1f8353fae4853ace6521d1 Mon Sep 17 00:00:00 2001
From: Bin Liu <b-liu@ti.com>
Date: Mon, 30 Apr 2018 11:20:53 -0500
Subject: usb: musb: host: fix potential NULL pointer dereference

musb_start_urb() doesn't check the pass-in parameter if it is NULL.  But
in musb_bulk_nak_timeout() the parameter passed to musb_start_urb() is
returned from first_qh(), which could be NULL.

So wrap the musb_start_urb() call here with a if condition check to
avoid the potential NULL pointer dereference.

Fixes: f283862f3b5c ("usb: musb: NAK timeout scheme on bulk TX endpoint")
Cc: stable@vger.kernel.org # v3.7+
Signed-off-by: Bin Liu <b-liu@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 drivers/usb/musb/musb_host.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/usb/musb/musb_host.c b/drivers/usb/musb/musb_host.c
index 4fa372c..e7f99d5 100644
--- a/drivers/usb/musb/musb_host.c
+++ b/drivers/usb/musb/musb_host.c
@@ -990,7 +990,9 @@ static void musb_bulk_nak_timeout(struct musb *musb, struct musb_hw_ep *ep,
 			/* set tx_reinit and schedule the next qh */
 			ep->tx_reinit = 1;
 		}
-		musb_start_urb(musb, is_in, next_qh);
+
+		if (next_qh)
+			musb_start_urb(musb, is_in, next_qh);
 	}
 }
 
-- 
cgit v1.1


From 9aea9b6cc78d2b99b23d84fb2e0bc6e464c6569e Mon Sep 17 00:00:00 2001
From: Bin Liu <b-liu@ti.com>
Date: Mon, 30 Apr 2018 11:20:54 -0500
Subject: usb: musb: trace: fix NULL pointer dereference in musb_g_tx()

The usb_request pointer could be NULL in musb_g_tx(), where the
tracepoint call would trigger the NULL pointer dereference failure when
parsing the members of the usb_request pointer.

Move the tracepoint call to where the usb_request pointer is already
checked to solve the issue.

Fixes: fc78003e5345 ("usb: musb: gadget: add usb-request tracepoints")
Cc: stable@vger.kernel.org # v4.8+
Signed-off-by: Bin Liu <b-liu@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 drivers/usb/musb/musb_gadget.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/usb/musb/musb_gadget.c b/drivers/usb/musb/musb_gadget.c
index e564695..71c5835 100644
--- a/drivers/usb/musb/musb_gadget.c
+++ b/drivers/usb/musb/musb_gadget.c
@@ -417,7 +417,6 @@ void musb_g_tx(struct musb *musb, u8 epnum)
 	req = next_request(musb_ep);
 	request = &req->request;
 
-	trace_musb_req_tx(req);
 	csr = musb_readw(epio, MUSB_TXCSR);
 	musb_dbg(musb, "<== %s, txcsr %04x", musb_ep->end_point.name, csr);
 
@@ -456,6 +455,8 @@ void musb_g_tx(struct musb *musb, u8 epnum)
 		u8	is_dma = 0;
 		bool	short_packet = false;
 
+		trace_musb_req_tx(req);
+
 		if (dma && (csr & MUSB_TXCSR_DMAENAB)) {
 			is_dma = 1;
 			csr |= MUSB_TXCSR_P_WZC_BITS;
-- 
cgit v1.1


From db82476f37413eaeff5f836a9d8b022d6544accf Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?H=C3=A5kon=20Bugge?= <haakon.bugge@oracle.com>
Date: Wed, 18 Apr 2018 16:24:50 +0200
Subject: IB/core: Make ib_mad_client_id atomic
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Currently, the kernel protects access to the agent ID allocator on a per
port basis using a spinlock, so it is impossible for two apps/threads on
the same port to get the same TID, but it is entirely possible for two
threads on different ports to end up with the same TID.

As this can be confusing (regardless of it being legal according to the
IB Spec 1.3, C13-18.1.1, in section 13.4.6.4 - TransactionID usage),
and as the rdma-core user space API for /dev/umad devices implies unique
TIDs even across ports, make the TID an atomic type so that no two
allocations, regardless of port number, will be the same.

Signed-off-by: Håkon Bugge <haakon.bugge@oracle.com>
Reviewed-by: Jack Morgenstein <jackm@dev.mellanox.co.il>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Zhu Yanjun <yanjun.zhu@oracle.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/mad.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/infiniband/core/mad.c b/drivers/infiniband/core/mad.c
index c50596f..b28452a 100644
--- a/drivers/infiniband/core/mad.c
+++ b/drivers/infiniband/core/mad.c
@@ -59,7 +59,7 @@ module_param_named(recv_queue_size, mad_recvq_size, int, 0444);
 MODULE_PARM_DESC(recv_queue_size, "Size of receive queue in number of work requests");
 
 static struct list_head ib_mad_port_list;
-static u32 ib_mad_client_id = 0;
+static atomic_t ib_mad_client_id = ATOMIC_INIT(0);
 
 /* Port list lock */
 static DEFINE_SPINLOCK(ib_mad_port_list_lock);
@@ -377,7 +377,7 @@ struct ib_mad_agent *ib_register_mad_agent(struct ib_device *device,
 	}
 
 	spin_lock_irqsave(&port_priv->reg_lock, flags);
-	mad_agent_priv->agent.hi_tid = ++ib_mad_client_id;
+	mad_agent_priv->agent.hi_tid = atomic_inc_return(&ib_mad_client_id);
 
 	/*
 	 * Make sure MAD registration (if supplied)
-- 
cgit v1.1


From f7aef1c207092770d06d0df21dceafdca2b49c39 Mon Sep 17 00:00:00 2001
From: Boris Brezillon <boris.brezillon@bootlin.com>
Date: Mon, 30 Apr 2018 15:32:32 +0200
Subject: drm/vc4: Make sure vc4_bo_{inc,dec}_usecnt() calls are balanced

Commit b9f19259b84d ("drm/vc4: Add the DRM_IOCTL_VC4_GEM_MADVISE ioctl")
introduced a mechanism to mark some BOs as purgeable to allow the driver
to drop them under memory pressure. In order to implement this feature
we had to add a mechanism to mark BOs as currently used by a piece of
hardware which materialized through the ->usecnt counter.

Plane code is supposed to increment usecnt when it attaches a BO to a
plane and decrement it when it's done with this BO, which was done in
the ->prepare_fb() and ->cleanup_fb() hooks. The problem is, async page
flip logic does not go through the regular atomic update path, and
->prepare_fb() and ->cleanup_fb() are not called in this case.

Fix that by manually calling vc4_bo_{inc,dec}_usecnt() in the
async-page-flip path.

Note that all this should go away as soon as we get generic async page
flip support in the core, in the meantime, this fix should do the
trick.

Fixes: b9f19259b84d ("drm/vc4: Add the DRM_IOCTL_VC4_GEM_MADVISE ioctl")
Reported-by: Peter Robinson <pbrobinson@gmail.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Boris Brezillon <boris.brezillon@bootlin.com>
Signed-off-by: Eric Anholt <eric@anholt.net>
Link: https://patchwork.freedesktop.org/patch/msgid/20180430133232.32457-1-boris.brezillon@bootlin.com
Link: https://patchwork.freedesktop.org/patch/msgid/20180430133232.32457-1-boris.brezillon@bootlin.com
---
 drivers/gpu/drm/vc4/vc4_crtc.c | 46 +++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 45 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/vc4/vc4_crtc.c b/drivers/gpu/drm/vc4/vc4_crtc.c
index bf46674..c61dff5 100644
--- a/drivers/gpu/drm/vc4/vc4_crtc.c
+++ b/drivers/gpu/drm/vc4/vc4_crtc.c
@@ -760,6 +760,7 @@ static irqreturn_t vc4_crtc_irq_handler(int irq, void *data)
 struct vc4_async_flip_state {
 	struct drm_crtc *crtc;
 	struct drm_framebuffer *fb;
+	struct drm_framebuffer *old_fb;
 	struct drm_pending_vblank_event *event;
 
 	struct vc4_seqno_cb cb;
@@ -789,6 +790,23 @@ vc4_async_page_flip_complete(struct vc4_seqno_cb *cb)
 
 	drm_crtc_vblank_put(crtc);
 	drm_framebuffer_put(flip_state->fb);
+
+	/* Decrement the BO usecnt in order to keep the inc/dec calls balanced
+	 * when the planes are updated through the async update path.
+	 * FIXME: we should move to generic async-page-flip when it's
+	 * available, so that we can get rid of this hand-made cleanup_fb()
+	 * logic.
+	 */
+	if (flip_state->old_fb) {
+		struct drm_gem_cma_object *cma_bo;
+		struct vc4_bo *bo;
+
+		cma_bo = drm_fb_cma_get_gem_obj(flip_state->old_fb, 0);
+		bo = to_vc4_bo(&cma_bo->base);
+		vc4_bo_dec_usecnt(bo);
+		drm_framebuffer_put(flip_state->old_fb);
+	}
+
 	kfree(flip_state);
 
 	up(&vc4->async_modeset);
@@ -813,9 +831,22 @@ static int vc4_async_page_flip(struct drm_crtc *crtc,
 	struct drm_gem_cma_object *cma_bo = drm_fb_cma_get_gem_obj(fb, 0);
 	struct vc4_bo *bo = to_vc4_bo(&cma_bo->base);
 
+	/* Increment the BO usecnt here, so that we never end up with an
+	 * unbalanced number of vc4_bo_{dec,inc}_usecnt() calls when the
+	 * plane is later updated through the non-async path.
+	 * FIXME: we should move to generic async-page-flip when it's
+	 * available, so that we can get rid of this hand-made prepare_fb()
+	 * logic.
+	 */
+	ret = vc4_bo_inc_usecnt(bo);
+	if (ret)
+		return ret;
+
 	flip_state = kzalloc(sizeof(*flip_state), GFP_KERNEL);
-	if (!flip_state)
+	if (!flip_state) {
+		vc4_bo_dec_usecnt(bo);
 		return -ENOMEM;
+	}
 
 	drm_framebuffer_get(fb);
 	flip_state->fb = fb;
@@ -826,10 +857,23 @@ static int vc4_async_page_flip(struct drm_crtc *crtc,
 	ret = down_interruptible(&vc4->async_modeset);
 	if (ret) {
 		drm_framebuffer_put(fb);
+		vc4_bo_dec_usecnt(bo);
 		kfree(flip_state);
 		return ret;
 	}
 
+	/* Save the current FB before it's replaced by the new one in
+	 * drm_atomic_set_fb_for_plane(). We'll need the old FB in
+	 * vc4_async_page_flip_complete() to decrement the BO usecnt and keep
+	 * it consistent.
+	 * FIXME: we should move to generic async-page-flip when it's
+	 * available, so that we can get rid of this hand-made cleanup_fb()
+	 * logic.
+	 */
+	flip_state->old_fb = plane->state->fb;
+	if (flip_state->old_fb)
+		drm_framebuffer_get(flip_state->old_fb);
+
 	WARN_ON(drm_crtc_vblank_get(crtc) != 0);
 
 	/* Immediately update the plane's legacy fb pointer, so that later
-- 
cgit v1.1


From 280884fadc5bd81628da3f158ae6821ea7386432 Mon Sep 17 00:00:00 2001
From: Mike Snitzer <snitzer@redhat.com>
Date: Mon, 30 Apr 2018 15:40:40 -0400
Subject: dm cache background tracker: fix sparse warning

Fix drivers/md/dm-cache-background-tracker.c:169:16: warning: symbol
'alloc_work' was not declared. Should it be static?

Signed-off-by: Mike Snitzer <snitzer@redhat.com>
---
 drivers/md/dm-cache-background-tracker.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/md/dm-cache-background-tracker.c b/drivers/md/dm-cache-background-tracker.c
index 1d0af0a..84814e8 100644
--- a/drivers/md/dm-cache-background-tracker.c
+++ b/drivers/md/dm-cache-background-tracker.c
@@ -166,7 +166,7 @@ static bool max_work_reached(struct background_tracker *b)
 		atomic_read(&b->pending_demotes) >= b->max_work;
 }
 
-struct bt_work *alloc_work(struct background_tracker *b)
+static struct bt_work *alloc_work(struct background_tracker *b)
 {
 	if (max_work_reached(b))
 		return NULL;
-- 
cgit v1.1


From d3c68d0b4183ddfa05577348a231dd117e6563df Mon Sep 17 00:00:00 2001
From: Rob Gardner <rob.gardner@oracle.com>
Date: Fri, 20 Apr 2018 12:48:25 -0600
Subject: sparc64: Fix mistake in oradax license text

The license text in both oradax files mistakenly specifies "version 3" of
the GNU General Public License.  This is corrected to specify "version 2".

Signed-off-by: Rob Gardner <rob.gardner@oracle.com>
Signed-off-by: Jonathan Helman <jonathan.helman@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 arch/sparc/include/uapi/asm/oradax.h | 2 +-
 drivers/sbus/char/oradax.c           | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/sparc/include/uapi/asm/oradax.h b/arch/sparc/include/uapi/asm/oradax.h
index 7229519..4f6676f 100644
--- a/arch/sparc/include/uapi/asm/oradax.h
+++ b/arch/sparc/include/uapi/asm/oradax.h
@@ -3,7 +3,7 @@
  *
  * This program is free software: you can redistribute it and/or modify
  * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation, either version 3 of the License, or
+ * the Free Software Foundation, either version 2 of the License, or
  * (at your option) any later version.
  *
  * This program is distributed in the hope that it will be useful,
diff --git a/drivers/sbus/char/oradax.c b/drivers/sbus/char/oradax.c
index c44d7c7..1754f55 100644
--- a/drivers/sbus/char/oradax.c
+++ b/drivers/sbus/char/oradax.c
@@ -3,7 +3,7 @@
  *
  * This program is free software: you can redistribute it and/or modify
  * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation, either version 3 of the License, or
+ * the Free Software Foundation, either version 2 of the License, or
  * (at your option) any later version.
  *
  * This program is distributed in the hope that it will be useful,
-- 
cgit v1.1


From 00ad691ab140b54ab9f5de5e74cb994f552e8124 Mon Sep 17 00:00:00 2001
From: Arvind Yadav <arvind.yadav.cs@gmail.com>
Date: Wed, 25 Apr 2018 20:26:14 +0530
Subject: sparc: vio: use put_device() instead of kfree()

Never directly free @dev after calling device_register(), even
if it returned an error. Always use put_device() to give up the
reference initialized.

Signed-off-by: Arvind Yadav <arvind.yadav.cs@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 arch/sparc/kernel/vio.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/sparc/kernel/vio.c b/arch/sparc/kernel/vio.c
index 1a0fa10..32bae68 100644
--- a/arch/sparc/kernel/vio.c
+++ b/arch/sparc/kernel/vio.c
@@ -403,7 +403,7 @@ static struct vio_dev *vio_create_one(struct mdesc_handle *hp, u64 mp,
 	if (err) {
 		printk(KERN_ERR "VIO: Could not register device %s, err=%d\n",
 		       dev_name(&vdev->dev), err);
-		kfree(vdev);
+		put_device(&vdev->dev);
 		return NULL;
 	}
 	if (vdev->dp)
-- 
cgit v1.1


From 3d97c829edd43262e7e9d720fa82c2241ba685a3 Mon Sep 17 00:00:00 2001
From: Mike Snitzer <snitzer@redhat.com>
Date: Mon, 30 Apr 2018 16:06:28 -0400
Subject: dm: fix some sparse warnings and whitespace in dax methods

Eliminate these sparse warnings:
drivers/md/dm.c:1062:9: warning: context imbalance in 'dm_dax_direct_access' - unexpected unlock
drivers/md/dm.c:1086:9: warning: context imbalance in 'dm_dax_copy_from_iter' - unexpected unlock

Signed-off-by: Mike Snitzer <snitzer@redhat.com>
---
 drivers/md/dm.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/drivers/md/dm.c b/drivers/md/dm.c
index 4ea404d..0a7b010 100644
--- a/drivers/md/dm.c
+++ b/drivers/md/dm.c
@@ -1020,7 +1020,8 @@ int dm_set_target_max_io_len(struct dm_target *ti, sector_t len)
 EXPORT_SYMBOL_GPL(dm_set_target_max_io_len);
 
 static struct dm_target *dm_dax_get_live_target(struct mapped_device *md,
-		sector_t sector, int *srcu_idx)
+						sector_t sector, int *srcu_idx)
+	__acquires(md->io_barrier)
 {
 	struct dm_table *map;
 	struct dm_target *ti;
@@ -1037,7 +1038,7 @@ static struct dm_target *dm_dax_get_live_target(struct mapped_device *md,
 }
 
 static long dm_dax_direct_access(struct dax_device *dax_dev, pgoff_t pgoff,
-		long nr_pages, void **kaddr, pfn_t *pfn)
+				 long nr_pages, void **kaddr, pfn_t *pfn)
 {
 	struct mapped_device *md = dax_get_private(dax_dev);
 	sector_t sector = pgoff * PAGE_SECTORS;
@@ -1065,7 +1066,7 @@ static long dm_dax_direct_access(struct dax_device *dax_dev, pgoff_t pgoff,
 }
 
 static size_t dm_dax_copy_from_iter(struct dax_device *dax_dev, pgoff_t pgoff,
-		void *addr, size_t bytes, struct iov_iter *i)
+				    void *addr, size_t bytes, struct iov_iter *i)
 {
 	struct mapped_device *md = dax_get_private(dax_dev);
 	sector_t sector = pgoff * PAGE_SECTORS;
-- 
cgit v1.1


From fb5c8e9d4350cb20eba1d692213d9efbb7298256 Mon Sep 17 00:00:00 2001
From: Manasi Navare <manasi.d.navare@intel.com>
Date: Wed, 28 Mar 2018 14:58:02 -0700
Subject: drm/i915/icl: Implement voltage swing programming sequence for Combo
 PHY DDI

This is an important part of the DDI initalization as well as
for changing the voltage during DisplayPort link training.

The Voltage swing seqeuence is similar to Cannonlake.
However it has different register definitions and hence
it makes sense to create a separate vswing sequence and
program functions for ICL to leave room for more changes
in case the Bspec changes later and deviates from CNL sequence.

v2:
Use ~TAP3_DISABLE for enbaling that bit (Jani Nikula)

v3:
* Use dw4_scaling column for PORT_TX_DW4 values (Rodrigo)

v4:
* Call it combo_vswing, use switch statement (Paulo)

v5 (from Paulo):
* Fix a typo.
* s/rate < 600000/rate <= 600000/.
* Don't remove blank lines that should be there.

v6:
* Rebased by Rodrigo on top of Cannonlake changes
  where non vswing sequences are not aligned with iboost
  anymore.

v7: Another rebase after an upstream rework.

v8 (from Paulo):
* Adjust the code to the upstream output type changes.
* Squash the patch that moved some functions up.
* Merge both get_combo_buf_trans functions in order to simplify the
  code.
* Change the changelog format.

v9 (from Paulo):
* Use RTERM_SELECT instead of SCALING_MODE_SEL.
* Adjust the output type handling according to how the other platforms
  do it now.

v10 (from Paulo):
* Fix comment left out from v9 changes (Rodrigo).

Cc: Jani Nikula <jani.nikula@linux.intel.com>
Cc: James Ausmus <james.ausmus@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Manasi Navare <manasi.d.navare@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180328215803.13835-8-paulo.r.zanoni@intel.com
---
 drivers/gpu/drm/i915/intel_ddi.c | 191 ++++++++++++++++++++++++++++++++++++++-
 1 file changed, 188 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c
index 92cb26b..0edbdb6 100644
--- a/drivers/gpu/drm/i915/intel_ddi.c
+++ b/drivers/gpu/drm/i915/intel_ddi.c
@@ -870,6 +870,45 @@ cnl_get_buf_trans_edp(struct drm_i915_private *dev_priv, int *n_entries)
 	}
 }
 
+static const struct icl_combo_phy_ddi_buf_trans *
+icl_get_combo_buf_trans(struct drm_i915_private *dev_priv, enum port port,
+			int type, int *n_entries)
+{
+	u32 voltage = I915_READ(ICL_PORT_COMP_DW3(port)) & VOLTAGE_INFO_MASK;
+
+	if (type == INTEL_OUTPUT_EDP && dev_priv->vbt.edp.low_vswing) {
+		switch (voltage) {
+		case VOLTAGE_INFO_0_85V:
+			*n_entries = ARRAY_SIZE(icl_combo_phy_ddi_translations_edp_0_85V);
+			return icl_combo_phy_ddi_translations_edp_0_85V;
+		case VOLTAGE_INFO_0_95V:
+			*n_entries = ARRAY_SIZE(icl_combo_phy_ddi_translations_edp_0_95V);
+			return icl_combo_phy_ddi_translations_edp_0_95V;
+		case VOLTAGE_INFO_1_05V:
+			*n_entries = ARRAY_SIZE(icl_combo_phy_ddi_translations_edp_1_05V);
+			return icl_combo_phy_ddi_translations_edp_1_05V;
+		default:
+			MISSING_CASE(voltage);
+			return NULL;
+		}
+	} else {
+		switch (voltage) {
+		case VOLTAGE_INFO_0_85V:
+			*n_entries = ARRAY_SIZE(icl_combo_phy_ddi_translations_dp_hdmi_0_85V);
+			return icl_combo_phy_ddi_translations_dp_hdmi_0_85V;
+		case VOLTAGE_INFO_0_95V:
+			*n_entries = ARRAY_SIZE(icl_combo_phy_ddi_translations_dp_hdmi_0_95V);
+			return icl_combo_phy_ddi_translations_dp_hdmi_0_95V;
+		case VOLTAGE_INFO_1_05V:
+			*n_entries = ARRAY_SIZE(icl_combo_phy_ddi_translations_dp_hdmi_1_05V);
+			return icl_combo_phy_ddi_translations_dp_hdmi_1_05V;
+		default:
+			MISSING_CASE(voltage);
+			return NULL;
+		}
+	}
+}
+
 static int intel_ddi_hdmi_level(struct drm_i915_private *dev_priv, enum port port)
 {
 	int n_entries, level, default_entry;
@@ -2182,6 +2221,146 @@ static void cnl_ddi_vswing_sequence(struct intel_encoder *encoder,
 	I915_WRITE(CNL_PORT_TX_DW5_GRP(port), val);
 }
 
+static void icl_ddi_combo_vswing_program(struct drm_i915_private *dev_priv,
+					 u32 level, enum port port, int type)
+{
+	const struct icl_combo_phy_ddi_buf_trans *ddi_translations = NULL;
+	u32 n_entries, val;
+	int ln;
+
+	ddi_translations = icl_get_combo_buf_trans(dev_priv, port, type,
+						   &n_entries);
+	if (!ddi_translations)
+		return;
+
+	if (level >= n_entries) {
+		DRM_DEBUG_KMS("DDI translation not found for level %d. Using %d instead.", level, n_entries - 1);
+		level = n_entries - 1;
+	}
+
+	/* Set PORT_TX_DW5 Rterm Sel to 110b. */
+	val = I915_READ(ICL_PORT_TX_DW5_LN0(port));
+	val &= ~RTERM_SELECT_MASK;
+	val |= RTERM_SELECT(0x6);
+	I915_WRITE(ICL_PORT_TX_DW5_GRP(port), val);
+
+	/* Program PORT_TX_DW5 */
+	val = I915_READ(ICL_PORT_TX_DW5_LN0(port));
+	/* Set DisableTap2 and DisableTap3 if MIPI DSI
+	 * Clear DisableTap2 and DisableTap3 for all other Ports
+	 */
+	if (type == INTEL_OUTPUT_DSI) {
+		val |= TAP2_DISABLE;
+		val |= TAP3_DISABLE;
+	} else {
+		val &= ~TAP2_DISABLE;
+		val &= ~TAP3_DISABLE;
+	}
+	I915_WRITE(ICL_PORT_TX_DW5_GRP(port), val);
+
+	/* Program PORT_TX_DW2 */
+	val = I915_READ(ICL_PORT_TX_DW2_LN0(port));
+	val &= ~(SWING_SEL_LOWER_MASK | SWING_SEL_UPPER_MASK |
+		 RCOMP_SCALAR_MASK);
+	val |= SWING_SEL_UPPER(ddi_translations[level].dw2_swing_select);
+	val |= SWING_SEL_LOWER(ddi_translations[level].dw2_swing_select);
+	/* Program Rcomp scalar for every table entry */
+	val |= RCOMP_SCALAR(ddi_translations[level].dw2_swing_scalar);
+	I915_WRITE(ICL_PORT_TX_DW2_GRP(port), val);
+
+	/* Program PORT_TX_DW4 */
+	/* We cannot write to GRP. It would overwrite individual loadgen. */
+	for (ln = 0; ln <= 3; ln++) {
+		val = I915_READ(ICL_PORT_TX_DW4_LN(port, ln));
+		val &= ~(POST_CURSOR_1_MASK | POST_CURSOR_2_MASK |
+			 CURSOR_COEFF_MASK);
+		val |= ddi_translations[level].dw4_scaling;
+		I915_WRITE(ICL_PORT_TX_DW4_LN(port, ln), val);
+	}
+}
+
+static void icl_combo_phy_ddi_vswing_sequence(struct intel_encoder *encoder,
+					      u32 level,
+					      enum intel_output_type type)
+{
+	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
+	enum port port = encoder->port;
+	int width = 0;
+	int rate = 0;
+	u32 val;
+	int ln = 0;
+
+	if (type == INTEL_OUTPUT_HDMI) {
+		width = 4;
+		/* Rate is always < than 6GHz for HDMI */
+	} else {
+		struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
+
+		width = intel_dp->lane_count;
+		rate = intel_dp->link_rate;
+	}
+
+	/*
+	 * 1. If port type is eDP or DP,
+	 * set PORT_PCS_DW1 cmnkeeper_enable to 1b,
+	 * else clear to 0b.
+	 */
+	val = I915_READ(ICL_PORT_PCS_DW1_LN0(port));
+	if (type == INTEL_OUTPUT_HDMI)
+		val &= ~COMMON_KEEPER_EN;
+	else
+		val |= COMMON_KEEPER_EN;
+	I915_WRITE(ICL_PORT_PCS_DW1_GRP(port), val);
+
+	/* 2. Program loadgen select */
+	/*
+	 * Program PORT_TX_DW4_LN depending on Bit rate and used lanes
+	 * <= 6 GHz and 4 lanes (LN0=0, LN1=1, LN2=1, LN3=1)
+	 * <= 6 GHz and 1,2 lanes (LN0=0, LN1=1, LN2=1, LN3=0)
+	 * > 6 GHz (LN0=0, LN1=0, LN2=0, LN3=0)
+	 */
+	for (ln = 0; ln <= 3; ln++) {
+		val = I915_READ(ICL_PORT_TX_DW4_LN(port, ln));
+		val &= ~LOADGEN_SELECT;
+
+		if ((rate <= 600000 && width == 4 && ln >= 1) ||
+		    (rate <= 600000 && width < 4 && (ln == 1 || ln == 2))) {
+			val |= LOADGEN_SELECT;
+		}
+		I915_WRITE(ICL_PORT_TX_DW4_LN(port, ln), val);
+	}
+
+	/* 3. Set PORT_CL_DW5 SUS Clock Config to 11b */
+	val = I915_READ(ICL_PORT_CL_DW5(port));
+	val |= SUS_CLOCK_CONFIG;
+	I915_WRITE(ICL_PORT_CL_DW5(port), val);
+
+	/* 4. Clear training enable to change swing values */
+	val = I915_READ(ICL_PORT_TX_DW5_LN0(port));
+	val &= ~TX_TRAINING_EN;
+	I915_WRITE(ICL_PORT_TX_DW5_GRP(port), val);
+
+	/* 5. Program swing and de-emphasis */
+	icl_ddi_combo_vswing_program(dev_priv, level, port, type);
+
+	/* 6. Set training enable to trigger update */
+	val = I915_READ(ICL_PORT_TX_DW5_LN0(port));
+	val |= TX_TRAINING_EN;
+	I915_WRITE(ICL_PORT_TX_DW5_GRP(port), val);
+}
+
+static void icl_ddi_vswing_sequence(struct intel_encoder *encoder, u32 level,
+				    enum intel_output_type type)
+{
+	enum port port = encoder->port;
+
+	if (port == PORT_A || port == PORT_B)
+		icl_combo_phy_ddi_vswing_sequence(encoder, level, type);
+	else
+		/* Not Implemented Yet */
+		WARN_ON(1);
+}
+
 static uint32_t translate_signal_level(int signal_levels)
 {
 	int i;
@@ -2213,7 +2392,9 @@ u32 bxt_signal_levels(struct intel_dp *intel_dp)
 	struct intel_encoder *encoder = &dport->base;
 	int level = intel_ddi_dp_level(intel_dp);
 
-	if (IS_CANNONLAKE(dev_priv))
+	if (IS_ICELAKE(dev_priv))
+		icl_ddi_vswing_sequence(encoder, level, encoder->type);
+	else if (IS_CANNONLAKE(dev_priv))
 		cnl_ddi_vswing_sequence(encoder, level, encoder->type);
 	else
 		bxt_ddi_vswing_sequence(encoder, level, encoder->type);
@@ -2316,7 +2497,9 @@ static void intel_ddi_pre_enable_dp(struct intel_encoder *encoder,
 
 	intel_display_power_get(dev_priv, dig_port->ddi_io_power_domain);
 
-	if (IS_CANNONLAKE(dev_priv))
+	if (IS_ICELAKE(dev_priv))
+		icl_ddi_vswing_sequence(encoder, level, encoder->type);
+	else if (IS_CANNONLAKE(dev_priv))
 		cnl_ddi_vswing_sequence(encoder, level, encoder->type);
 	else if (IS_GEN9_LP(dev_priv))
 		bxt_ddi_vswing_sequence(encoder, level, encoder->type);
@@ -2347,7 +2530,9 @@ static void intel_ddi_pre_enable_hdmi(struct intel_encoder *encoder,
 
 	intel_display_power_get(dev_priv, dig_port->ddi_io_power_domain);
 
-	if (IS_CANNONLAKE(dev_priv))
+	if (IS_ICELAKE(dev_priv))
+		icl_ddi_vswing_sequence(encoder, level, INTEL_OUTPUT_HDMI);
+	else if (IS_CANNONLAKE(dev_priv))
 		cnl_ddi_vswing_sequence(encoder, level, INTEL_OUTPUT_HDMI);
 	else if (IS_GEN9_LP(dev_priv))
 		bxt_ddi_vswing_sequence(encoder, level, INTEL_OUTPUT_HDMI);
-- 
cgit v1.1


From 36cf89f53b0ccdbd6bdaedfe1435a574609f0de5 Mon Sep 17 00:00:00 2001
From: Manasi Navare <manasi.d.navare@intel.com>
Date: Wed, 28 Mar 2018 14:58:03 -0700
Subject: drm/i915/icl: Fix the DP Max Voltage for ICL

On clock recovery this function is called to find out
the max voltage swing level that we could go.

However gen 9 functions use the old buffer translation tables
to figure that out. ICL uses different set of tables for eDP
and DP for both Combo and MG PHY ports. This patch adds the hook
for ICL for getting this information from appropriate buf trans tables.

v5 (from Paulo):
* New rebase after changes to earlier patches.
v4:
* Rebase.
v3:
* Follow the coding conventions here
(https://cgit.freedesktop.org/drm-intel/tree/Documentation/process/codin
g-style.rst#n191) (Paulo)
v2:
* Rebase after patch that adds voltage check inside buf trans
function (Rodrigo)

Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Paulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Manasi Navare <manasi.d.navare@intel.com>
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180328215803.13835-9-paulo.r.zanoni@intel.com
---
 drivers/gpu/drm/i915/intel_ddi.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c
index 0edbdb6..8225d22 100644
--- a/drivers/gpu/drm/i915/intel_ddi.c
+++ b/drivers/gpu/drm/i915/intel_ddi.c
@@ -2064,7 +2064,13 @@ u8 intel_ddi_dp_voltage_max(struct intel_encoder *encoder)
 	enum port port = encoder->port;
 	int n_entries;
 
-	if (IS_CANNONLAKE(dev_priv)) {
+	if (IS_ICELAKE(dev_priv)) {
+		if (port == PORT_A || port == PORT_B)
+			icl_get_combo_buf_trans(dev_priv, port, encoder->type,
+						&n_entries);
+		else
+			n_entries = ARRAY_SIZE(icl_mg_phy_ddi_translations);
+	} else if (IS_CANNONLAKE(dev_priv)) {
 		if (encoder->type == INTEL_OUTPUT_EDP)
 			cnl_get_buf_trans_edp(dev_priv, &n_entries);
 		else
-- 
cgit v1.1


From 05c58752f9dce11e396676eb731a620541590ed0 Mon Sep 17 00:00:00 2001
From: CHANDAN VN <chandan.vn@samsung.com>
Date: Mon, 30 Apr 2018 09:50:18 +0530
Subject: arm64: To remove initrd reserved area entry from memblock

INITRD reserved area entry is not removed from memblock
even though initrd reserved area is freed. After freeing
the memory it is released from memblock. The same can be
checked from /sys/kernel/debug/memblock/reserved.

The patch makes sure that the initrd entry is removed from
memblock when keepinitrd is not enabled.

The patch only affects accounting and debugging. This does not
fix any memory leak.

Acked-by: Laura Abbott <labbott@redhat.com>
Signed-off-by: CHANDAN VN <chandan.vn@samsung.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
---
 arch/arm64/mm/init.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c
index 9f3c47a..1b18b47 100644
--- a/arch/arm64/mm/init.c
+++ b/arch/arm64/mm/init.c
@@ -646,8 +646,10 @@ static int keep_initrd __initdata;
 
 void __init free_initrd_mem(unsigned long start, unsigned long end)
 {
-	if (!keep_initrd)
+	if (!keep_initrd) {
 		free_reserved_area((void *)start, (void *)end, 0, "initrd");
+		memblock_free(__virt_to_phys(start), end - start);
+	}
 }
 
 static int __init keepinitrd_setup(char *__unused)
-- 
cgit v1.1


From 7b240e44d08c4a94faf3ed02c04c16b52fc045de Mon Sep 17 00:00:00 2001
From: Geert Uytterhoeven <geert+renesas@glider.be>
Date: Mon, 23 Apr 2018 09:35:16 +0200
Subject: dt-bindings: mvebu-uart: DT fix s/interrupts-names/interrupt-names/

Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Rob Herring <robh@kernel.org>
---
 Documentation/devicetree/bindings/serial/mvebu-uart.txt | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/Documentation/devicetree/bindings/serial/mvebu-uart.txt b/Documentation/devicetree/bindings/serial/mvebu-uart.txt
index 2ae2fee..b7e0e32 100644
--- a/Documentation/devicetree/bindings/serial/mvebu-uart.txt
+++ b/Documentation/devicetree/bindings/serial/mvebu-uart.txt
@@ -24,7 +24,7 @@ Required properties:
     - Must contain two elements for the extended variant of the IP
       (marvell,armada-3700-uart-ext): "uart-tx" and "uart-rx",
       respectively the UART TX interrupt and the UART RX interrupt. A
-      corresponding interrupts-names property must be defined.
+      corresponding interrupt-names property must be defined.
     - For backward compatibility reasons, a single element interrupts
       property is also supported for the standard variant of the IP,
       containing only the UART sum interrupt. This form is deprecated
-- 
cgit v1.1


From f130307054a59ca21d2396f386be77ebd2e8ca96 Mon Sep 17 00:00:00 2001
From: Geert Uytterhoeven <geert+renesas@glider.be>
Date: Wed, 25 Apr 2018 09:49:38 +0200
Subject: dt-bindings: panel: lvds: Fix path to display timing bindings

Fixes: 14da3ed8dd08c581 ("devicetree/bindings: display: Document common
panel properties")
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Rob Herring <robh@kernel.org>
---
 Documentation/devicetree/bindings/display/panel/panel-common.txt | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/Documentation/devicetree/bindings/display/panel/panel-common.txt b/Documentation/devicetree/bindings/display/panel/panel-common.txt
index 557fa76..5d2519a 100644
--- a/Documentation/devicetree/bindings/display/panel/panel-common.txt
+++ b/Documentation/devicetree/bindings/display/panel/panel-common.txt
@@ -38,7 +38,7 @@ Display Timings
   require specific display timings. The panel-timing subnode expresses those
   timings as specified in the timing subnode section of the display timing
   bindings defined in
-  Documentation/devicetree/bindings/display/display-timing.txt.
+  Documentation/devicetree/bindings/display/panel/display-timing.txt.
 
 
 Connectivity
-- 
cgit v1.1


From cea67a2dd6b2419dcc13a39309b9a79a1f773193 Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Sun, 29 Apr 2018 09:54:59 -0700
Subject: ipv6: fix uninit-value in ip6_multipath_l3_keys()

syzbot/KMSAN reported an uninit-value in ip6_multipath_l3_keys(),
root caused to a bad assumption of ICMP header being already
pulled in skb->head

ip_multipath_l3_keys() does the correct thing, so it is an IPv6 only bug.

BUG: KMSAN: uninit-value in ip6_multipath_l3_keys net/ipv6/route.c:1830 [inline]
BUG: KMSAN: uninit-value in rt6_multipath_hash+0x5c4/0x640 net/ipv6/route.c:1858
CPU: 0 PID: 4507 Comm: syz-executor661 Not tainted 4.16.0+ #87
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Call Trace:
 __dump_stack lib/dump_stack.c:17 [inline]
 dump_stack+0x185/0x1d0 lib/dump_stack.c:53
 kmsan_report+0x142/0x240 mm/kmsan/kmsan.c:1067
 __msan_warning_32+0x6c/0xb0 mm/kmsan/kmsan_instr.c:683
 ip6_multipath_l3_keys net/ipv6/route.c:1830 [inline]
 rt6_multipath_hash+0x5c4/0x640 net/ipv6/route.c:1858
 ip6_route_input+0x65a/0x920 net/ipv6/route.c:1884
 ip6_rcv_finish+0x413/0x6e0 net/ipv6/ip6_input.c:69
 NF_HOOK include/linux/netfilter.h:288 [inline]
 ipv6_rcv+0x1e16/0x2340 net/ipv6/ip6_input.c:208
 __netif_receive_skb_core+0x47df/0x4a90 net/core/dev.c:4562
 __netif_receive_skb net/core/dev.c:4627 [inline]
 netif_receive_skb_internal+0x49d/0x630 net/core/dev.c:4701
 netif_receive_skb+0x230/0x240 net/core/dev.c:4725
 tun_rx_batched drivers/net/tun.c:1555 [inline]
 tun_get_user+0x740f/0x7c60 drivers/net/tun.c:1962
 tun_chr_write_iter+0x1d4/0x330 drivers/net/tun.c:1990
 call_write_iter include/linux/fs.h:1782 [inline]
 new_sync_write fs/read_write.c:469 [inline]
 __vfs_write+0x7fb/0x9f0 fs/read_write.c:482
 vfs_write+0x463/0x8d0 fs/read_write.c:544
 SYSC_write+0x172/0x360 fs/read_write.c:589
 SyS_write+0x55/0x80 fs/read_write.c:581
 do_syscall_64+0x309/0x430 arch/x86/entry/common.c:287
 entry_SYSCALL_64_after_hwframe+0x3d/0xa2

Fixes: 23aebdacb05d ("ipv6: Compute multipath hash for ICMP errors from offending packet")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: syzbot <syzkaller@googlegroups.com>
Cc: Jakub Sitnicki <jkbs@redhat.com>
Acked-by: Jakub Sitnicki <jkbs@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/ipv6/route.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index cde7d82..f4d6173 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -1835,11 +1835,16 @@ static void ip6_multipath_l3_keys(const struct sk_buff *skb,
 	const struct ipv6hdr *inner_iph;
 	const struct icmp6hdr *icmph;
 	struct ipv6hdr _inner_iph;
+	struct icmp6hdr _icmph;
 
 	if (likely(outer_iph->nexthdr != IPPROTO_ICMPV6))
 		goto out;
 
-	icmph = icmp6_hdr(skb);
+	icmph = skb_header_pointer(skb, skb_transport_offset(skb),
+				   sizeof(_icmph), &_icmph);
+	if (!icmph)
+		goto out;
+
 	if (icmph->icmp6_type != ICMPV6_DEST_UNREACH &&
 	    icmph->icmp6_type != ICMPV6_PKT_TOOBIG &&
 	    icmph->icmp6_type != ICMPV6_TIME_EXCEED &&
-- 
cgit v1.1


From bf2acc943a45d2b2e8a9f1a5ddff6b6e43cc69d9 Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Sun, 29 Apr 2018 18:55:20 -0700
Subject: tcp: fix TCP_REPAIR_QUEUE bound checking

syzbot is able to produce a nasty WARN_ON() in tcp_verify_left_out()
with following C-repro :

socket(PF_INET, SOCK_STREAM, IPPROTO_IP) = 3
setsockopt(3, SOL_TCP, TCP_REPAIR, [1], 4) = 0
setsockopt(3, SOL_TCP, TCP_REPAIR_QUEUE, [-1], 4) = 0
bind(3, {sa_family=AF_INET, sin_port=htons(20002), sin_addr=inet_addr("0.0.0.0")}, 16) = 0
sendto(3, "\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"...,
	1242, MSG_FASTOPEN, {sa_family=AF_INET, sin_port=htons(20002), sin_addr=inet_addr("127.0.0.1")}, 16) = 1242
setsockopt(3, SOL_TCP, TCP_REPAIR_WINDOW, "\4\0\0@+\205\0\0\377\377\0\0\377\377\377\177\0\0\0\0", 20) = 0
writev(3, [{"\270", 1}], 1)             = 1
setsockopt(3, SOL_TCP, TCP_REPAIR_OPTIONS, "\10\0\0\0\0\0\0\0\0\0\0\0|\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 386) = 0
writev(3, [{"\210v\r[\226\320t\231qwQ\204\264l\254\t\1\20\245\214p\350H\223\254;\\\37\345\307p$"..., 3144}], 1) = 3144

The 3rd system call looks odd :
setsockopt(3, SOL_TCP, TCP_REPAIR_QUEUE, [-1], 4) = 0

This patch makes sure bound checking is using an unsigned compare.

Fixes: ee9952831cfd ("tcp: Initial repair mode")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: syzbot <syzkaller@googlegroups.com>
Cc: Pavel Emelyanov <xemul@parallels.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/ipv4/tcp.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 4b18ad4..44be7f4 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -2674,7 +2674,7 @@ static int do_tcp_setsockopt(struct sock *sk, int level,
 	case TCP_REPAIR_QUEUE:
 		if (!tp->repair)
 			err = -EPERM;
-		else if (val < TCP_QUEUES_NR)
+		else if ((unsigned int)val < TCP_QUEUES_NR)
 			tp->repair_queue = val;
 		else
 			err = -EINVAL;
-- 
cgit v1.1


From 054f155721d7af1f343ed52bea246626d8450ca8 Mon Sep 17 00:00:00 2001
From: Chuck Lever <chuck.lever@oracle.com>
Date: Tue, 1 May 2018 11:37:14 -0400
Subject: xprtrdma: Fix list corruption / DMAR errors during MR recovery

The ro_release_mr methods check whether mr->mr_list is empty.
Therefore, be sure to always use list_del_init when removing an MR
linked into a list using that field. Otherwise, when recovering from
transport failures or device removal, list corruption can result, or
MRs can get mapped or unmapped an odd number of times, resulting in
IOMMU-related failures.

In general this fix is appropriate back to v4.8. However, code
changes since then make it impossible to apply this patch directly
to stable kernels. The fix would have to be applied by hand or
reworked for kernels earlier than v4.16.

Backport guidance -- there are several cases:
- When creating an MR, initialize mr_list so that using list_empty
  on an as-yet-unused MR is safe.
- When an MR is being handled by the remote invalidation path,
  ensure that mr_list is reinitialized when it is removed from
  rl_registered.
- When an MR is being handled by rpcrdma_destroy_mrs, it is removed
  from mr_all, but it may still be on an rl_registered list. In
  that case, the MR needs to be removed from that list before being
  released.
- Other cases are covered by using list_del_init in rpcrdma_mr_pop.

Fixes: 9d6b04097882 ('xprtrdma: Place registered MWs on a ... ')
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
---
 net/sunrpc/xprtrdma/fmr_ops.c   | 5 +----
 net/sunrpc/xprtrdma/frwr_ops.c  | 9 +++------
 net/sunrpc/xprtrdma/verbs.c     | 5 +++++
 net/sunrpc/xprtrdma/xprt_rdma.h | 2 +-
 4 files changed, 10 insertions(+), 11 deletions(-)

diff --git a/net/sunrpc/xprtrdma/fmr_ops.c b/net/sunrpc/xprtrdma/fmr_ops.c
index 5cc68a8..f2f6395 100644
--- a/net/sunrpc/xprtrdma/fmr_ops.c
+++ b/net/sunrpc/xprtrdma/fmr_ops.c
@@ -72,6 +72,7 @@ fmr_op_init_mr(struct rpcrdma_ia *ia, struct rpcrdma_mr *mr)
 	if (IS_ERR(mr->fmr.fm_mr))
 		goto out_fmr_err;
 
+	INIT_LIST_HEAD(&mr->mr_list);
 	return 0;
 
 out_fmr_err:
@@ -102,10 +103,6 @@ fmr_op_release_mr(struct rpcrdma_mr *mr)
 	LIST_HEAD(unmap_list);
 	int rc;
 
-	/* Ensure MW is not on any rl_registered list */
-	if (!list_empty(&mr->mr_list))
-		list_del(&mr->mr_list);
-
 	kfree(mr->fmr.fm_physaddrs);
 	kfree(mr->mr_sg);
 
diff --git a/net/sunrpc/xprtrdma/frwr_ops.c b/net/sunrpc/xprtrdma/frwr_ops.c
index c5743a0..c59c5c7 100644
--- a/net/sunrpc/xprtrdma/frwr_ops.c
+++ b/net/sunrpc/xprtrdma/frwr_ops.c
@@ -110,6 +110,7 @@ frwr_op_init_mr(struct rpcrdma_ia *ia, struct rpcrdma_mr *mr)
 	if (!mr->mr_sg)
 		goto out_list_err;
 
+	INIT_LIST_HEAD(&mr->mr_list);
 	sg_init_table(mr->mr_sg, depth);
 	init_completion(&frwr->fr_linv_done);
 	return 0;
@@ -133,10 +134,6 @@ frwr_op_release_mr(struct rpcrdma_mr *mr)
 {
 	int rc;
 
-	/* Ensure MR is not on any rl_registered list */
-	if (!list_empty(&mr->mr_list))
-		list_del(&mr->mr_list);
-
 	rc = ib_dereg_mr(mr->frwr.fr_mr);
 	if (rc)
 		pr_err("rpcrdma: final ib_dereg_mr for %p returned %i\n",
@@ -195,7 +192,7 @@ frwr_op_recover_mr(struct rpcrdma_mr *mr)
 	return;
 
 out_release:
-	pr_err("rpcrdma: FRWR reset failed %d, %p release\n", rc, mr);
+	pr_err("rpcrdma: FRWR reset failed %d, %p released\n", rc, mr);
 	r_xprt->rx_stats.mrs_orphaned++;
 
 	spin_lock(&r_xprt->rx_buf.rb_mrlock);
@@ -476,7 +473,7 @@ frwr_op_reminv(struct rpcrdma_rep *rep, struct list_head *mrs)
 
 	list_for_each_entry(mr, mrs, mr_list)
 		if (mr->mr_handle == rep->rr_inv_rkey) {
-			list_del(&mr->mr_list);
+			list_del_init(&mr->mr_list);
 			trace_xprtrdma_remoteinv(mr);
 			mr->frwr.fr_state = FRWR_IS_INVALID;
 			rpcrdma_mr_unmap_and_put(mr);
diff --git a/net/sunrpc/xprtrdma/verbs.c b/net/sunrpc/xprtrdma/verbs.c
index fe5eaca..c345d36 100644
--- a/net/sunrpc/xprtrdma/verbs.c
+++ b/net/sunrpc/xprtrdma/verbs.c
@@ -1254,6 +1254,11 @@ rpcrdma_mrs_destroy(struct rpcrdma_buffer *buf)
 		list_del(&mr->mr_all);
 
 		spin_unlock(&buf->rb_mrlock);
+
+		/* Ensure MW is not on any rl_registered list */
+		if (!list_empty(&mr->mr_list))
+			list_del(&mr->mr_list);
+
 		ia->ri_ops->ro_release_mr(mr);
 		count++;
 		spin_lock(&buf->rb_mrlock);
diff --git a/net/sunrpc/xprtrdma/xprt_rdma.h b/net/sunrpc/xprtrdma/xprt_rdma.h
index 3d3b423..cb41b12 100644
--- a/net/sunrpc/xprtrdma/xprt_rdma.h
+++ b/net/sunrpc/xprtrdma/xprt_rdma.h
@@ -380,7 +380,7 @@ rpcrdma_mr_pop(struct list_head *list)
 	struct rpcrdma_mr *mr;
 
 	mr = list_first_entry(list, struct rpcrdma_mr, mr_list);
-	list_del(&mr->mr_list);
+	list_del_init(&mr->mr_list);
 	return mr;
 }
 
-- 
cgit v1.1


From 7e437d61514bc3e05167da263977e3cdd698f08a Mon Sep 17 00:00:00 2001
From: Colin Ian King <colin.king@canonical.com>
Date: Sun, 29 Apr 2018 13:01:11 +0100
Subject: ata: fix spelling mistake: "directon" -> "direction"

Trivial fix to spelling mistake in sil24_cerr_info message text

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
---
 drivers/ata/sata_sil24.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/ata/sata_sil24.c b/drivers/ata/sata_sil24.c
index 4b1995e..010ca10 100644
--- a/drivers/ata/sata_sil24.c
+++ b/drivers/ata/sata_sil24.c
@@ -285,13 +285,13 @@ static const struct sil24_cerr_info {
 	[PORT_CERR_INCONSISTENT] = { AC_ERR_HSM, ATA_EH_RESET,
 				     "protocol mismatch" },
 	[PORT_CERR_DIRECTION]	= { AC_ERR_HSM, ATA_EH_RESET,
-				    "data directon mismatch" },
+				    "data direction mismatch" },
 	[PORT_CERR_UNDERRUN]	= { AC_ERR_HSM, ATA_EH_RESET,
 				    "ran out of SGEs while writing" },
 	[PORT_CERR_OVERRUN]	= { AC_ERR_HSM, ATA_EH_RESET,
 				    "ran out of SGEs while reading" },
 	[PORT_CERR_PKT_PROT]	= { AC_ERR_HSM, ATA_EH_RESET,
-				    "invalid data directon for ATAPI CDB" },
+				    "invalid data direction for ATAPI CDB" },
 	[PORT_CERR_SGT_BOUNDARY] = { AC_ERR_SYSTEM, ATA_EH_RESET,
 				     "SGT not on qword boundary" },
 	[PORT_CERR_SGT_TGTABRT]	= { AC_ERR_HOST_BUS, ATA_EH_RESET,
-- 
cgit v1.1


From 98eb6cf25f0317395d9a799d18f3d46ba26a00d3 Mon Sep 17 00:00:00 2001
From: Chuck Lever <chuck.lever@oracle.com>
Date: Tue, 1 May 2018 11:37:19 -0400
Subject: sunrpc: Fix latency trace point crashes

If the rpc_task survived longer than the transport, task->tk_xprt
points to freed memory by the time rpc_count_iostats_metrics runs.
Replace the references to task->tk_xprt with references to the
task's tk_client.

Reported-by: syzbot+27db1f90e2b972a5f2d3@syzkaller.appspotmail.com
Fixes: 40bf7eb304b5 ('sunrpc: Add static trace point to report ...')
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
---
 include/trace/events/sunrpc.h | 16 ++++++----------
 1 file changed, 6 insertions(+), 10 deletions(-)

diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h
index 76887d6..7f1204a 100644
--- a/include/trace/events/sunrpc.h
+++ b/include/trace/events/sunrpc.h
@@ -224,6 +224,8 @@ TRACE_EVENT(rpc_stats_latency,
 	TP_ARGS(task, backlog, rtt, execute),
 
 	TP_STRUCT__entry(
+		__field(unsigned int, task_id)
+		__field(unsigned int, client_id)
 		__field(u32, xid)
 		__field(int, version)
 		__string(progname, task->tk_client->cl_program->name)
@@ -231,13 +233,11 @@ TRACE_EVENT(rpc_stats_latency,
 		__field(unsigned long, backlog)
 		__field(unsigned long, rtt)
 		__field(unsigned long, execute)
-		__string(addr,
-			 task->tk_xprt->address_strings[RPC_DISPLAY_ADDR])
-		__string(port,
-			 task->tk_xprt->address_strings[RPC_DISPLAY_PORT])
 	),
 
 	TP_fast_assign(
+		__entry->client_id = task->tk_client->cl_clid;
+		__entry->task_id = task->tk_pid;
 		__entry->xid = be32_to_cpu(task->tk_rqstp->rq_xid);
 		__entry->version = task->tk_client->cl_vers;
 		__assign_str(progname, task->tk_client->cl_program->name)
@@ -245,14 +245,10 @@ TRACE_EVENT(rpc_stats_latency,
 		__entry->backlog = ktime_to_us(backlog);
 		__entry->rtt = ktime_to_us(rtt);
 		__entry->execute = ktime_to_us(execute);
-		__assign_str(addr,
-			     task->tk_xprt->address_strings[RPC_DISPLAY_ADDR]);
-		__assign_str(port,
-			     task->tk_xprt->address_strings[RPC_DISPLAY_PORT]);
 	),
 
-	TP_printk("peer=[%s]:%s xid=0x%08x %sv%d %s backlog=%lu rtt=%lu execute=%lu",
-		__get_str(addr), __get_str(port), __entry->xid,
+	TP_printk("task:%u@%d xid=0x%08x %sv%d %s backlog=%lu rtt=%lu execute=%lu",
+		__entry->task_id, __entry->client_id, __entry->xid,
 		__get_str(progname), __entry->version, __get_str(procname),
 		__entry->backlog, __entry->rtt, __entry->execute)
 );
-- 
cgit v1.1


From de08481a253ac658433a8304a303ce9f018d71e5 Mon Sep 17 00:00:00 2001
From: "Michael S. Tsirkin" <mst@redhat.com>
Date: Fri, 27 Apr 2018 19:02:05 +0300
Subject: vhost: make msg padding explicit

There's a 32 bit hole just after type. It's best to
give it a name, this way compiler is forced to initialize
it with rest of the structure.

Reported-by: Kevin Easton <kevin@guarana.org>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/uapi/linux/vhost.h | 1 +
 1 file changed, 1 insertion(+)

diff --git a/include/uapi/linux/vhost.h b/include/uapi/linux/vhost.h
index c51f8e5..5a8ad06 100644
--- a/include/uapi/linux/vhost.h
+++ b/include/uapi/linux/vhost.h
@@ -68,6 +68,7 @@ struct vhost_iotlb_msg {
 
 struct vhost_msg {
 	int type;
+	int padding0;
 	union {
 		struct vhost_iotlb_msg iotlb;
 		__u8 padding[64];
-- 
cgit v1.1


From 26ff75857e5953720409ea531d42f902defb130a Mon Sep 17 00:00:00 2001
From: Colin Ian King <colin.king@canonical.com>
Date: Mon, 30 Apr 2018 17:29:45 +0100
Subject: net/mlx4: fix spelling mistake: "failedi" -> "failed"

trivial fix to spelling mistake in mlx4_warn message.

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/mellanox/mlx4/main.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/mellanox/mlx4/main.c b/drivers/net/ethernet/mellanox/mlx4/main.c
index bfef692..211578f 100644
--- a/drivers/net/ethernet/mellanox/mlx4/main.c
+++ b/drivers/net/ethernet/mellanox/mlx4/main.c
@@ -1317,7 +1317,7 @@ static int mlx4_mf_unbond(struct mlx4_dev *dev)
 
 	ret = mlx4_unbond_fs_rules(dev);
 	if (ret)
-		mlx4_warn(dev, "multifunction unbond for flow rules failedi (%d)\n", ret);
+		mlx4_warn(dev, "multifunction unbond for flow rules failed (%d)\n", ret);
 	ret1 = mlx4_unbond_mac_table(dev);
 	if (ret1) {
 		mlx4_warn(dev, "multifunction unbond for MAC table failed (%d)\n", ret1);
-- 
cgit v1.1


From d656fe49e33df48ee6bc19e871f5862f49895c9e Mon Sep 17 00:00:00 2001
From: Wenwen Wang <wang6495@umn.edu>
Date: Mon, 30 Apr 2018 12:31:13 -0500
Subject: ethtool: fix a potential missing-check bug

In ethtool_get_rxnfc(), the object "info" is firstly copied from
user-space. If the FLOW_RSS flag is set in the member field flow_type of
"info" (and cmd is ETHTOOL_GRXFH), info needs to be copied again from
user-space because FLOW_RSS is newer and has new definition, as mentioned
in the comment. However, given that the user data resides in user-space, a
malicious user can race to change the data after the first copy. By doing
so, the user can inject inconsistent data. For example, in the second
copy, the FLOW_RSS flag could be cleared in the field flow_type of "info".
In the following execution, "info" will be used in the function
ops->get_rxnfc(). Such inconsistent data can potentially lead to unexpected
information leakage since ops->get_rxnfc() will prepare various types of
data according to flow_type, and the prepared data will be eventually
copied to user-space. This inconsistent data may also cause undefined
behaviors based on how ops->get_rxnfc() is implemented.

This patch simply re-verifies the flow_type field of "info" after the
second copy. If the value is not as expected, an error code will be
returned.

Signed-off-by: Wenwen Wang <wang6495@umn.edu>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/core/ethtool.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/net/core/ethtool.c b/net/core/ethtool.c
index 03416e6..ba02f0d 100644
--- a/net/core/ethtool.c
+++ b/net/core/ethtool.c
@@ -1032,6 +1032,11 @@ static noinline_for_stack int ethtool_get_rxnfc(struct net_device *dev,
 		info_size = sizeof(info);
 		if (copy_from_user(&info, useraddr, info_size))
 			return -EFAULT;
+		/* Since malicious users may modify the original data,
+		 * we need to check whether FLOW_RSS is still requested.
+		 */
+		if (!(info.flow_type & FLOW_RSS))
+			return -EINVAL;
 	}
 
 	if (info.cmd == ETHTOOL_GRXCLSRLALL) {
-- 
cgit v1.1


From edd7ceb78296fb1574958991b6655c3c2cedf124 Mon Sep 17 00:00:00 2001
From: Thomas Winter <Thomas.Winter@alliedtelesis.co.nz>
Date: Tue, 1 May 2018 09:15:29 +1200
Subject: ipv6: Allow non-gateway ECMP for IPv6

It is valid to have static routes where the nexthop
is an interface not an address such as tunnels.
For IPv4 it was possible to use ECMP on these routes
but not for IPv6.

Signed-off-by: Thomas Winter <Thomas.Winter@alliedtelesis.co.nz>
Cc: David Ahern <dsahern@gmail.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Alexey Kuznetsov <kuznet@ms2.inr.ac.ru>
Cc: Hideaki YOSHIFUJI <yoshfuji@linux-ipv6.org>
Acked-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/ip6_route.h | 3 +--
 net/ipv6/ip6_fib.c      | 3 ---
 2 files changed, 1 insertion(+), 5 deletions(-)

diff --git a/include/net/ip6_route.h b/include/net/ip6_route.h
index 08b1323..abceb58 100644
--- a/include/net/ip6_route.h
+++ b/include/net/ip6_route.h
@@ -68,8 +68,7 @@ static inline bool rt6_need_strict(const struct in6_addr *daddr)
 
 static inline bool rt6_qualify_for_ecmp(const struct rt6_info *rt)
 {
-	return (rt->rt6i_flags & (RTF_GATEWAY|RTF_ADDRCONF|RTF_DYNAMIC)) ==
-	       RTF_GATEWAY;
+	return (rt->rt6i_flags & (RTF_ADDRCONF | RTF_DYNAMIC)) == 0;
 }
 
 void ip6_route_input(struct sk_buff *skb);
diff --git a/net/ipv6/ip6_fib.c b/net/ipv6/ip6_fib.c
index deab2db..3c97c29 100644
--- a/net/ipv6/ip6_fib.c
+++ b/net/ipv6/ip6_fib.c
@@ -934,9 +934,6 @@ static int fib6_add_rt2node(struct fib6_node *fn, struct rt6_info *rt,
 			 * list.
 			 * Only static routes (which don't have flag
 			 * RTF_EXPIRES) are used for ECMPv6.
-			 *
-			 * To avoid long list, we only had siblings if the
-			 * route have a gateway.
 			 */
 			if (rt_can_ecmp &&
 			    rt6_qualify_for_ecmp(iter))
-- 
cgit v1.1


From f372b81101e6895252298e563d634d5e44ae81e7 Mon Sep 17 00:00:00 2001
From: "Vittorio Gambaletta (VittGam)" <linuxbugs@vittgam.net>
Date: Wed, 25 Apr 2018 15:22:13 -0700
Subject: Input: atmel_mxt_ts - add touchpad button mapping for Samsung
 Chromebook Pro

This patch adds the correct platform data information for the Caroline
Chromebook, so that the mouse button does not get stuck in pressed state
after the first click.

The Samus button keymap and platform data definition are the correct
ones for Caroline, so they have been reused here.

Signed-off-by: Vittorio Gambaletta <linuxbugs@vittgam.net>
Signed-off-by: Salvatore Bellizzi <lkml@seppia.net>
Tested-by: Guenter Roeck <groeck@chromium.org>
Cc: stable@vger.kernel.org
[dtor: adjusted vendor spelling to match shipping firmware]
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
---
 drivers/input/touchscreen/atmel_mxt_ts.c | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/drivers/input/touchscreen/atmel_mxt_ts.c b/drivers/input/touchscreen/atmel_mxt_ts.c
index 7659bc4..429b694 100644
--- a/drivers/input/touchscreen/atmel_mxt_ts.c
+++ b/drivers/input/touchscreen/atmel_mxt_ts.c
@@ -3031,6 +3031,15 @@ static const struct dmi_system_id mxt_dmi_table[] = {
 		.driver_data = samus_platform_data,
 	},
 	{
+		/* Samsung Chromebook Pro */
+		.ident = "Samsung Chromebook Pro",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "Google"),
+			DMI_MATCH(DMI_PRODUCT_NAME, "Caroline"),
+		},
+		.driver_data = samus_platform_data,
+	},
+	{
 		/* Other Google Chromebooks */
 		.ident = "Chromebook",
 		.matches = {
-- 
cgit v1.1


From 068bdb67ef74df0ad1627b7247a163e3e252ac11 Mon Sep 17 00:00:00 2001
From: Nick Dyer <nick@shmanahar.org>
Date: Tue, 1 May 2018 11:40:18 -0700
Subject: Input: atmel_mxt_ts - fix the firmware update

The automatic update mechanism will trigger an update if the
info block CRCs are different between maxtouch configuration
file (maxtouch.cfg) and chip.

The driver compared the CRCs without retrieving the chip CRC,
resulting always in a failure and firmware flashing action
triggered. Fix this issue by retrieving the chip info block
CRC before the check.

Note that this solution has the benefit that by reading the
information block and the object table into a contiguous region
of memory, we can verify the checksum at probe time. This means
we make sure that we are indeed talking to a chip that supports
object protocol correctly.

Using this patch on a kevin chromebook, the touchscreen and
touchpad drivers are able to match the CRC:

  atmel_mxt_ts 3-004b: Family: 164 Variant: 14 Firmware V2.3.AA Objects: 40
  atmel_mxt_ts 5-004a: Family: 164 Variant: 17 Firmware V2.0.AA Objects: 31
  atmel_mxt_ts 3-004b: Resetting device
  atmel_mxt_ts 5-004a: Resetting device
  atmel_mxt_ts 3-004b: Config CRC 0x573E89: OK
  atmel_mxt_ts 3-004b: Touchscreen size X4095Y2729
  input: Atmel maXTouch Touchscreen as /devices/platform/ff130000.i2c/i2c-3/3-004b/input/input5
  atmel_mxt_ts 5-004a: Config CRC 0x0AF6BA: OK
  atmel_mxt_ts 5-004a: Touchscreen size X1920Y1080
  input: Atmel maXTouch Touchpad as /devices/platform/ff140000.i2c/i2c-5/5-004a/input/input6

Signed-off-by: Nick Dyer <nick.dyer@shmanahar.org>
Acked-by: Benson Leung <bleung@chromium.org>
[Ezequiel: minor patch massage]
Signed-off-by: Ezequiel Garcia <ezequiel@collabora.com>
Tested-by: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
---
 drivers/input/touchscreen/atmel_mxt_ts.c | 186 ++++++++++++++++++-------------
 1 file changed, 110 insertions(+), 76 deletions(-)

diff --git a/drivers/input/touchscreen/atmel_mxt_ts.c b/drivers/input/touchscreen/atmel_mxt_ts.c
index 429b694..fc149ea 100644
--- a/drivers/input/touchscreen/atmel_mxt_ts.c
+++ b/drivers/input/touchscreen/atmel_mxt_ts.c
@@ -275,7 +275,8 @@ struct mxt_data {
 	char phys[64];		/* device physical location */
 	const struct mxt_platform_data *pdata;
 	struct mxt_object *object_table;
-	struct mxt_info info;
+	struct mxt_info *info;
+	void *raw_info_block;
 	unsigned int irq;
 	unsigned int max_x;
 	unsigned int max_y;
@@ -450,12 +451,13 @@ static int mxt_lookup_bootloader_address(struct mxt_data *data, bool retry)
 {
 	u8 appmode = data->client->addr;
 	u8 bootloader;
+	u8 family_id = data->info ? data->info->family_id : 0;
 
 	switch (appmode) {
 	case 0x4a:
 	case 0x4b:
 		/* Chips after 1664S use different scheme */
-		if (retry || data->info.family_id >= 0xa2) {
+		if (retry || family_id >= 0xa2) {
 			bootloader = appmode - 0x24;
 			break;
 		}
@@ -682,7 +684,7 @@ mxt_get_object(struct mxt_data *data, u8 type)
 	struct mxt_object *object;
 	int i;
 
-	for (i = 0; i < data->info.object_num; i++) {
+	for (i = 0; i < data->info->object_num; i++) {
 		object = data->object_table + i;
 		if (object->type == type)
 			return object;
@@ -1453,12 +1455,12 @@ static int mxt_update_cfg(struct mxt_data *data, const struct firmware *cfg)
 		data_pos += offset;
 	}
 
-	if (cfg_info.family_id != data->info.family_id) {
+	if (cfg_info.family_id != data->info->family_id) {
 		dev_err(dev, "Family ID mismatch!\n");
 		return -EINVAL;
 	}
 
-	if (cfg_info.variant_id != data->info.variant_id) {
+	if (cfg_info.variant_id != data->info->variant_id) {
 		dev_err(dev, "Variant ID mismatch!\n");
 		return -EINVAL;
 	}
@@ -1503,7 +1505,7 @@ static int mxt_update_cfg(struct mxt_data *data, const struct firmware *cfg)
 
 	/* Malloc memory to store configuration */
 	cfg_start_ofs = MXT_OBJECT_START +
-			data->info.object_num * sizeof(struct mxt_object) +
+			data->info->object_num * sizeof(struct mxt_object) +
 			MXT_INFO_CHECKSUM_SIZE;
 	config_mem_size = data->mem_size - cfg_start_ofs;
 	config_mem = kzalloc(config_mem_size, GFP_KERNEL);
@@ -1554,20 +1556,6 @@ release_mem:
 	return ret;
 }
 
-static int mxt_get_info(struct mxt_data *data)
-{
-	struct i2c_client *client = data->client;
-	struct mxt_info *info = &data->info;
-	int error;
-
-	/* Read 7-byte info block starting at address 0 */
-	error = __mxt_read_reg(client, 0, sizeof(*info), info);
-	if (error)
-		return error;
-
-	return 0;
-}
-
 static void mxt_free_input_device(struct mxt_data *data)
 {
 	if (data->input_dev) {
@@ -1582,9 +1570,10 @@ static void mxt_free_object_table(struct mxt_data *data)
 	video_unregister_device(&data->dbg.vdev);
 	v4l2_device_unregister(&data->dbg.v4l2);
 #endif
-
-	kfree(data->object_table);
 	data->object_table = NULL;
+	data->info = NULL;
+	kfree(data->raw_info_block);
+	data->raw_info_block = NULL;
 	kfree(data->msg_buf);
 	data->msg_buf = NULL;
 	data->T5_address = 0;
@@ -1600,34 +1589,18 @@ static void mxt_free_object_table(struct mxt_data *data)
 	data->max_reportid = 0;
 }
 
-static int mxt_get_object_table(struct mxt_data *data)
+static int mxt_parse_object_table(struct mxt_data *data,
+				  struct mxt_object *object_table)
 {
 	struct i2c_client *client = data->client;
-	size_t table_size;
-	struct mxt_object *object_table;
-	int error;
 	int i;
 	u8 reportid;
 	u16 end_address;
 
-	table_size = data->info.object_num * sizeof(struct mxt_object);
-	object_table = kzalloc(table_size, GFP_KERNEL);
-	if (!object_table) {
-		dev_err(&data->client->dev, "Failed to allocate memory\n");
-		return -ENOMEM;
-	}
-
-	error = __mxt_read_reg(client, MXT_OBJECT_START, table_size,
-			object_table);
-	if (error) {
-		kfree(object_table);
-		return error;
-	}
-
 	/* Valid Report IDs start counting from 1 */
 	reportid = 1;
 	data->mem_size = 0;
-	for (i = 0; i < data->info.object_num; i++) {
+	for (i = 0; i < data->info->object_num; i++) {
 		struct mxt_object *object = object_table + i;
 		u8 min_id, max_id;
 
@@ -1651,8 +1624,8 @@ static int mxt_get_object_table(struct mxt_data *data)
 
 		switch (object->type) {
 		case MXT_GEN_MESSAGE_T5:
-			if (data->info.family_id == 0x80 &&
-			    data->info.version < 0x20) {
+			if (data->info->family_id == 0x80 &&
+			    data->info->version < 0x20) {
 				/*
 				 * On mXT224 firmware versions prior to V2.0
 				 * read and discard unused CRC byte otherwise
@@ -1707,24 +1680,102 @@ static int mxt_get_object_table(struct mxt_data *data)
 	/* If T44 exists, T5 position has to be directly after */
 	if (data->T44_address && (data->T5_address != data->T44_address + 1)) {
 		dev_err(&client->dev, "Invalid T44 position\n");
-		error = -EINVAL;
-		goto free_object_table;
+		return -EINVAL;
 	}
 
 	data->msg_buf = kcalloc(data->max_reportid,
 				data->T5_msg_size, GFP_KERNEL);
-	if (!data->msg_buf) {
-		dev_err(&client->dev, "Failed to allocate message buffer\n");
+	if (!data->msg_buf)
+		return -ENOMEM;
+
+	return 0;
+}
+
+static int mxt_read_info_block(struct mxt_data *data)
+{
+	struct i2c_client *client = data->client;
+	int error;
+	size_t size;
+	void *id_buf, *buf;
+	uint8_t num_objects;
+	u32 calculated_crc;
+	u8 *crc_ptr;
+
+	/* If info block already allocated, free it */
+	if (data->raw_info_block)
+		mxt_free_object_table(data);
+
+	/* Read 7-byte ID information block starting at address 0 */
+	size = sizeof(struct mxt_info);
+	id_buf = kzalloc(size, GFP_KERNEL);
+	if (!id_buf)
+		return -ENOMEM;
+
+	error = __mxt_read_reg(client, 0, size, id_buf);
+	if (error)
+		goto err_free_mem;
+
+	/* Resize buffer to give space for rest of info block */
+	num_objects = ((struct mxt_info *)id_buf)->object_num;
+	size += (num_objects * sizeof(struct mxt_object))
+		+ MXT_INFO_CHECKSUM_SIZE;
+
+	buf = krealloc(id_buf, size, GFP_KERNEL);
+	if (!buf) {
 		error = -ENOMEM;
-		goto free_object_table;
+		goto err_free_mem;
+	}
+	id_buf = buf;
+
+	/* Read rest of info block */
+	error = __mxt_read_reg(client, MXT_OBJECT_START,
+			       size - MXT_OBJECT_START,
+			       id_buf + MXT_OBJECT_START);
+	if (error)
+		goto err_free_mem;
+
+	/* Extract & calculate checksum */
+	crc_ptr = id_buf + size - MXT_INFO_CHECKSUM_SIZE;
+	data->info_crc = crc_ptr[0] | (crc_ptr[1] << 8) | (crc_ptr[2] << 16);
+
+	calculated_crc = mxt_calculate_crc(id_buf, 0,
+					   size - MXT_INFO_CHECKSUM_SIZE);
+
+	/*
+	 * CRC mismatch can be caused by data corruption due to I2C comms
+	 * issue or else device is not using Object Based Protocol (eg i2c-hid)
+	 */
+	if ((data->info_crc == 0) || (data->info_crc != calculated_crc)) {
+		dev_err(&client->dev,
+			"Info Block CRC error calculated=0x%06X read=0x%06X\n",
+			calculated_crc, data->info_crc);
+		error = -EIO;
+		goto err_free_mem;
+	}
+
+	data->raw_info_block = id_buf;
+	data->info = (struct mxt_info *)id_buf;
+
+	dev_info(&client->dev,
+		 "Family: %u Variant: %u Firmware V%u.%u.%02X Objects: %u\n",
+		 data->info->family_id, data->info->variant_id,
+		 data->info->version >> 4, data->info->version & 0xf,
+		 data->info->build, data->info->object_num);
+
+	/* Parse object table information */
+	error = mxt_parse_object_table(data, id_buf + MXT_OBJECT_START);
+	if (error) {
+		dev_err(&client->dev, "Error %d parsing object table\n", error);
+		mxt_free_object_table(data);
+		goto err_free_mem;
 	}
 
-	data->object_table = object_table;
+	data->object_table = (struct mxt_object *)(id_buf + MXT_OBJECT_START);
 
 	return 0;
 
-free_object_table:
-	mxt_free_object_table(data);
+err_free_mem:
+	kfree(id_buf);
 	return error;
 }
 
@@ -2039,7 +2090,7 @@ static int mxt_initialize(struct mxt_data *data)
 	int error;
 
 	while (1) {
-		error = mxt_get_info(data);
+		error = mxt_read_info_block(data);
 		if (!error)
 			break;
 
@@ -2070,16 +2121,9 @@ static int mxt_initialize(struct mxt_data *data)
 		msleep(MXT_FW_RESET_TIME);
 	}
 
-	/* Get object table information */
-	error = mxt_get_object_table(data);
-	if (error) {
-		dev_err(&client->dev, "Error %d reading object table\n", error);
-		return error;
-	}
-
 	error = mxt_acquire_irq(data);
 	if (error)
-		goto err_free_object_table;
+		return error;
 
 	error = request_firmware_nowait(THIS_MODULE, true, MXT_CFG_NAME,
 					&client->dev, GFP_KERNEL, data,
@@ -2087,14 +2131,10 @@ static int mxt_initialize(struct mxt_data *data)
 	if (error) {
 		dev_err(&client->dev, "Failed to invoke firmware loader: %d\n",
 			error);
-		goto err_free_object_table;
+		return error;
 	}
 
 	return 0;
-
-err_free_object_table:
-	mxt_free_object_table(data);
-	return error;
 }
 
 static int mxt_set_t7_power_cfg(struct mxt_data *data, u8 sleep)
@@ -2155,7 +2195,7 @@ recheck:
 static u16 mxt_get_debug_value(struct mxt_data *data, unsigned int x,
 			       unsigned int y)
 {
-	struct mxt_info *info = &data->info;
+	struct mxt_info *info = data->info;
 	struct mxt_dbg *dbg = &data->dbg;
 	unsigned int ofs, page;
 	unsigned int col = 0;
@@ -2483,7 +2523,7 @@ static const struct video_device mxt_video_device = {
 
 static void mxt_debug_init(struct mxt_data *data)
 {
-	struct mxt_info *info = &data->info;
+	struct mxt_info *info = data->info;
 	struct mxt_dbg *dbg = &data->dbg;
 	struct mxt_object *object;
 	int error;
@@ -2569,7 +2609,6 @@ static int mxt_configure_objects(struct mxt_data *data,
 				 const struct firmware *cfg)
 {
 	struct device *dev = &data->client->dev;
-	struct mxt_info *info = &data->info;
 	int error;
 
 	error = mxt_init_t7_power_cfg(data);
@@ -2594,11 +2633,6 @@ static int mxt_configure_objects(struct mxt_data *data,
 
 	mxt_debug_init(data);
 
-	dev_info(dev,
-		 "Family: %u Variant: %u Firmware V%u.%u.%02X Objects: %u\n",
-		 info->family_id, info->variant_id, info->version >> 4,
-		 info->version & 0xf, info->build, info->object_num);
-
 	return 0;
 }
 
@@ -2607,7 +2641,7 @@ static ssize_t mxt_fw_version_show(struct device *dev,
 				   struct device_attribute *attr, char *buf)
 {
 	struct mxt_data *data = dev_get_drvdata(dev);
-	struct mxt_info *info = &data->info;
+	struct mxt_info *info = data->info;
 	return scnprintf(buf, PAGE_SIZE, "%u.%u.%02X\n",
 			 info->version >> 4, info->version & 0xf, info->build);
 }
@@ -2617,7 +2651,7 @@ static ssize_t mxt_hw_version_show(struct device *dev,
 				   struct device_attribute *attr, char *buf)
 {
 	struct mxt_data *data = dev_get_drvdata(dev);
-	struct mxt_info *info = &data->info;
+	struct mxt_info *info = data->info;
 	return scnprintf(buf, PAGE_SIZE, "%u.%u\n",
 			 info->family_id, info->variant_id);
 }
@@ -2656,7 +2690,7 @@ static ssize_t mxt_object_show(struct device *dev,
 		return -ENOMEM;
 
 	error = 0;
-	for (i = 0; i < data->info.object_num; i++) {
+	for (i = 0; i < data->info->object_num; i++) {
 		object = data->object_table + i;
 
 		if (!mxt_object_readable(object->type))
-- 
cgit v1.1


From c5ce3b8df6c758169b5b5df5ee9adc4c39505d9b Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Tue, 1 May 2018 13:21:31 +0100
Subject: drm/i915/execlists: Don't trigger preemption if complete

Due to the latency of the tasklet running from ksoftirqd, by the time we
process the execlist dequeue may be a long time behind the GPU. If the
request was completed when we ran reschedule, we will not have tweaked
its priority, but if it is still listed as being in-flight for dequeue
we will use it as a reference for the rest of the queue, including
requests from its own context which will now be at higher priority. This
can cause us to issue a preempt-to-idle request, even though the request
we want to preempt is already complete.

Reported-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180501122131.19435-1-chris@chris-wilson.co.uk
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
---
 drivers/gpu/drm/i915/intel_lrc.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 0999956..774b2ad 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -185,7 +185,8 @@ static inline bool need_preempt(const struct intel_engine_cs *engine,
 				int prio)
 {
 	return (intel_engine_has_preemption(engine) &&
-		__execlists_need_preempt(prio, rq_prio(last)));
+		__execlists_need_preempt(prio, rq_prio(last)) &&
+		!i915_request_completed(last));
 }
 
 /**
-- 
cgit v1.1


From a57ab96ef9dde231d4d46edba4d5f73720edc16a Mon Sep 17 00:00:00 2001
From: Arnd Bergmann <arnd@arndb.de>
Date: Fri, 6 Apr 2018 16:28:22 +0200
Subject: hexagon: add memset_io() helper

We already have memcpy_toio(), but not memset_io(), so let's
add the obvious version to allow building an allmodconfig kernel
without errors like

drivers/gpu/drm/ttm/ttm_bo_util.c: In function 'ttm_bo_move_memcpy':
drivers/gpu/drm/ttm/ttm_bo_util.c:390:3: error: implicit declaration of function 'memset_io' [-Werror=implicit-function-declaration]

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Richard Kuo <rkuo@codeaurora.org>
---
 arch/hexagon/include/asm/io.h | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/arch/hexagon/include/asm/io.h b/arch/hexagon/include/asm/io.h
index 9e8621d..e17262a 100644
--- a/arch/hexagon/include/asm/io.h
+++ b/arch/hexagon/include/asm/io.h
@@ -216,6 +216,12 @@ static inline void memcpy_toio(volatile void __iomem *dst, const void *src,
 	memcpy((void *) dst, src, count);
 }
 
+static inline void memset_io(volatile void __iomem *addr, int value,
+			     size_t size)
+{
+	memset((void __force *)addr, value, size);
+}
+
 #define PCI_IO_ADDR	(volatile void __iomem *)
 
 /*
-- 
cgit v1.1


From 330e261c35dfb969c48f996dbbc8b334b5ee8d9d Mon Sep 17 00:00:00 2001
From: Arnd Bergmann <arnd@arndb.de>
Date: Fri, 6 Apr 2018 16:28:23 +0200
Subject: hexagon: export csum_partial_copy_nocheck

This is needed to link ipv6 as a loadable module, which in turn happens
in allmodconfig.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Richard Kuo <rkuo@codeaurora.org>
---
 arch/hexagon/lib/checksum.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/hexagon/lib/checksum.c b/arch/hexagon/lib/checksum.c
index 617506d..7cd0a22 100644
--- a/arch/hexagon/lib/checksum.c
+++ b/arch/hexagon/lib/checksum.c
@@ -199,3 +199,4 @@ csum_partial_copy_nocheck(const void *src, void *dst, int len, __wsum sum)
 	memcpy(dst, src, len);
 	return csum_partial(dst, len, sum);
 }
+EXPORT_SYMBOL(csum_partial_copy_nocheck);
-- 
cgit v1.1


From a4e21ff8d9a311b16cfa9e41aed0c627a47149a4 Mon Sep 17 00:00:00 2001
From: Song Liu <songliubraving@fb.com>
Date: Tue, 1 May 2018 10:20:24 -0700
Subject: bpf: minor fix to selftest test_stacktrace_build_id()

1. remove useless parameter list to ./urandom_read
2. add missing "\n" to the end of an error message

Fixes: 81f77fd0deeb ("bpf: add selftest for stackmap with BPF_F_STACK_BUILD_ID")
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Song Liu <songliubraving@fb.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
---
 tools/testing/selftests/bpf/test_progs.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/tools/testing/selftests/bpf/test_progs.c b/tools/testing/selftests/bpf/test_progs.c
index faadbe2..4123d0a 100644
--- a/tools/testing/selftests/bpf/test_progs.c
+++ b/tools/testing/selftests/bpf/test_progs.c
@@ -1108,7 +1108,7 @@ static void test_stacktrace_build_id(void)
 
 	assert(system("dd if=/dev/urandom of=/dev/zero count=4 2> /dev/null")
 	       == 0);
-	assert(system("./urandom_read if=/dev/urandom of=/dev/zero count=4 2> /dev/null") == 0);
+	assert(system("./urandom_read") == 0);
 	/* disable stack trace collection */
 	key = 0;
 	val = 1;
@@ -1158,7 +1158,7 @@ static void test_stacktrace_build_id(void)
 	} while (bpf_map_get_next_key(stackmap_fd, &previous_key, &key) == 0);
 
 	CHECK(build_id_matches < 1, "build id match",
-	      "Didn't find expected build ID from the map");
+	      "Didn't find expected build ID from the map\n");
 
 disable_pmu:
 	ioctl(pmu_fd, PERF_EVENT_IOC_DISABLE);
-- 
cgit v1.1


From f6eeb9e548572c0e64632672dbd04363cbdd4eed Mon Sep 17 00:00:00 2001
From: Javier Martinez Canillas <javierm@redhat.com>
Date: Tue, 1 May 2018 11:44:53 -0700
Subject: Input: atmel_mxt_ts - add missing compatible strings to OF device
 table

Commit af503716ac14 ("i2c: core: report OF style module alias for devices
registered via OF") fixed how the I2C core reports the module alias when
devices are registered via OF.

But the atmel_mxt_ts driver only has an "atmel,maxtouch" compatible in its
OF device ID table, so if a Device Tree is using a different one, autoload
won't be working for the module (the matching works because the I2C device
ID table is used as a fallback).

So add compatible strings for each of the entries in the I2C device table.

Fixes: af503716ac14 ("i2c: core: report OF style module alias for devices registered via OF")
Reported-by: Enric Balletbo i Serra <enric.balletbo@collabora.com>
Signed-off-by: Javier Martinez Canillas <javierm@redhat.com>
Tested-by: Enric Balletbo i Serra <enric.balletbo@collabora.com>
Reviewed-by: Rob Herring <robh@kernel.org>
[dtor: document which compatibles are deprecated and should not be used]
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
---
 Documentation/devicetree/bindings/input/atmel,maxtouch.txt | 7 +++++++
 drivers/input/touchscreen/atmel_mxt_ts.c                   | 5 +++++
 2 files changed, 12 insertions(+)

diff --git a/Documentation/devicetree/bindings/input/atmel,maxtouch.txt b/Documentation/devicetree/bindings/input/atmel,maxtouch.txt
index 23e3abc..c889194 100644
--- a/Documentation/devicetree/bindings/input/atmel,maxtouch.txt
+++ b/Documentation/devicetree/bindings/input/atmel,maxtouch.txt
@@ -4,6 +4,13 @@ Required properties:
 - compatible:
     atmel,maxtouch
 
+    The following compatibles have been used in various products but are
+    deprecated:
+	atmel,qt602240_ts
+	atmel,atmel_mxt_ts
+	atmel,atmel_mxt_tp
+	atmel,mXT224
+
 - reg: The I2C address of the device
 
 - interrupts: The sink for the touchpad's IRQ output
diff --git a/drivers/input/touchscreen/atmel_mxt_ts.c b/drivers/input/touchscreen/atmel_mxt_ts.c
index fc149ea..f04e4fb 100644
--- a/drivers/input/touchscreen/atmel_mxt_ts.c
+++ b/drivers/input/touchscreen/atmel_mxt_ts.c
@@ -3282,6 +3282,11 @@ static SIMPLE_DEV_PM_OPS(mxt_pm_ops, mxt_suspend, mxt_resume);
 
 static const struct of_device_id mxt_of_match[] = {
 	{ .compatible = "atmel,maxtouch", },
+	/* Compatibles listed below are deprecated */
+	{ .compatible = "atmel,qt602240_ts", },
+	{ .compatible = "atmel,atmel_mxt_ts", },
+	{ .compatible = "atmel,atmel_mxt_tp", },
+	{ .compatible = "atmel,mXT224", },
 	{},
 };
 MODULE_DEVICE_TABLE(of, mxt_of_match);
-- 
cgit v1.1


From c212d2c7fc4736d49be102fb7a1a545cdc2f1fea Mon Sep 17 00:00:00 2001
From: Dave Watson <davejwatson@fb.com>
Date: Tue, 1 May 2018 13:05:39 -0700
Subject: net/tls: Don't recursively call push_record during tls_write_space
 callbacks

It is reported that in some cases, write_space may be called in
do_tcp_sendpages, such that we recursively invoke do_tcp_sendpages again:

[  660.468802]  ? do_tcp_sendpages+0x8d/0x580
[  660.468826]  ? tls_push_sg+0x74/0x130 [tls]
[  660.468852]  ? tls_push_record+0x24a/0x390 [tls]
[  660.468880]  ? tls_write_space+0x6a/0x80 [tls]
...

tls_push_sg already does a loop over all sending sg's, so ignore
any tls_write_space notifications until we are done sending.
We then have to call the previous write_space to wake up
poll() waiters after we are done with the send loop.

Reported-by: Andre Tomt <andre@tomt.net>
Signed-off-by: Dave Watson <davejwatson@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/tls.h  | 1 +
 net/tls/tls_main.c | 7 +++++++
 2 files changed, 8 insertions(+)

diff --git a/include/net/tls.h b/include/net/tls.h
index 3da8e13..b400d0b 100644
--- a/include/net/tls.h
+++ b/include/net/tls.h
@@ -148,6 +148,7 @@ struct tls_context {
 	struct scatterlist *partially_sent_record;
 	u16 partially_sent_offset;
 	unsigned long flags;
+	bool in_tcp_sendpages;
 
 	u16 pending_open_record_frags;
 	int (*push_pending_record)(struct sock *sk, int flags);
diff --git a/net/tls/tls_main.c b/net/tls/tls_main.c
index 0d37997..cc03e00 100644
--- a/net/tls/tls_main.c
+++ b/net/tls/tls_main.c
@@ -114,6 +114,7 @@ int tls_push_sg(struct sock *sk,
 	size = sg->length - offset;
 	offset += sg->offset;
 
+	ctx->in_tcp_sendpages = true;
 	while (1) {
 		if (sg_is_last(sg))
 			sendpage_flags = flags;
@@ -148,6 +149,8 @@ retry:
 	}
 
 	clear_bit(TLS_PENDING_CLOSED_RECORD, &ctx->flags);
+	ctx->in_tcp_sendpages = false;
+	ctx->sk_write_space(sk);
 
 	return 0;
 }
@@ -217,6 +220,10 @@ static void tls_write_space(struct sock *sk)
 {
 	struct tls_context *ctx = tls_get_ctx(sk);
 
+	/* We are already sending pages, ignore notification */
+	if (ctx->in_tcp_sendpages)
+		return;
+
 	if (!sk->sk_write_pending && tls_is_pending_closed_record(ctx)) {
 		gfp_t sk_allocation = sk->sk_allocation;
 		int rc;
-- 
cgit v1.1


From 50a5852a657f793a8482fe3af4a141b460d3499e Mon Sep 17 00:00:00 2001
From: John Hurley <john.hurley@netronome.com>
Date: Tue, 1 May 2018 15:49:49 -0700
Subject: nfp: flower: set tunnel ttl value to net default

Firmware requires that the ttl value for an encapsulating ipv4 tunnel
header be included as an action field. Prior to the support of Geneve
tunnel encap (when ttl set was removed completely), ttl value was
extracted from the tunnel key. However, tests have shown that this can
still produce a ttl of 0.

Fix the issue by setting the namespace default value for each new tunnel.
Follow up patch for net-next will do a full route lookup.

Fixes: 3ca3059dc3a9 ("nfp: flower: compile Geneve encap actions")
Fixes: b27d6a95a70d ("nfp: compile flower vxlan tunnel set actions")
Signed-off-by: John Hurley <john.hurley@netronome.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/netronome/nfp/flower/action.c | 10 ++++++++--
 drivers/net/ethernet/netronome/nfp/flower/cmsg.h   |  5 ++++-
 2 files changed, 12 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ethernet/netronome/nfp/flower/action.c b/drivers/net/ethernet/netronome/nfp/flower/action.c
index b3567a5..80df9a5 100644
--- a/drivers/net/ethernet/netronome/nfp/flower/action.c
+++ b/drivers/net/ethernet/netronome/nfp/flower/action.c
@@ -183,17 +183,21 @@ static int
 nfp_fl_set_ipv4_udp_tun(struct nfp_fl_set_ipv4_udp_tun *set_tun,
 			const struct tc_action *action,
 			struct nfp_fl_pre_tunnel *pre_tun,
-			enum nfp_flower_tun_type tun_type)
+			enum nfp_flower_tun_type tun_type,
+			struct net_device *netdev)
 {
 	size_t act_size = sizeof(struct nfp_fl_set_ipv4_udp_tun);
 	struct ip_tunnel_info *ip_tun = tcf_tunnel_info(action);
 	u32 tmp_set_ip_tun_type_index = 0;
 	/* Currently support one pre-tunnel so index is always 0. */
 	int pretun_idx = 0;
+	struct net *net;
 
 	if (ip_tun->options_len)
 		return -EOPNOTSUPP;
 
+	net = dev_net(netdev);
+
 	set_tun->head.jump_id = NFP_FL_ACTION_OPCODE_SET_IPV4_TUNNEL;
 	set_tun->head.len_lw = act_size >> NFP_FL_LW_SIZ;
 
@@ -204,6 +208,7 @@ nfp_fl_set_ipv4_udp_tun(struct nfp_fl_set_ipv4_udp_tun *set_tun,
 
 	set_tun->tun_type_index = cpu_to_be32(tmp_set_ip_tun_type_index);
 	set_tun->tun_id = ip_tun->key.tun_id;
+	set_tun->ttl = net->ipv4.sysctl_ip_default_ttl;
 
 	/* Complete pre_tunnel action. */
 	pre_tun->ipv4_dst = ip_tun->key.u.ipv4.dst;
@@ -511,7 +516,8 @@ nfp_flower_loop_action(const struct tc_action *a,
 		*a_len += sizeof(struct nfp_fl_pre_tunnel);
 
 		set_tun = (void *)&nfp_fl->action_data[*a_len];
-		err = nfp_fl_set_ipv4_udp_tun(set_tun, a, pre_tun, *tun_type);
+		err = nfp_fl_set_ipv4_udp_tun(set_tun, a, pre_tun, *tun_type,
+					      netdev);
 		if (err)
 			return err;
 		*a_len += sizeof(struct nfp_fl_set_ipv4_udp_tun);
diff --git a/drivers/net/ethernet/netronome/nfp/flower/cmsg.h b/drivers/net/ethernet/netronome/nfp/flower/cmsg.h
index b6c0fd0..bee4367 100644
--- a/drivers/net/ethernet/netronome/nfp/flower/cmsg.h
+++ b/drivers/net/ethernet/netronome/nfp/flower/cmsg.h
@@ -190,7 +190,10 @@ struct nfp_fl_set_ipv4_udp_tun {
 	__be16 reserved;
 	__be64 tun_id __packed;
 	__be32 tun_type_index;
-	__be32 extra[3];
+	__be16 reserved2;
+	u8 ttl;
+	u8 reserved3;
+	__be32 extra[2];
 };
 
 /* Metadata with L2 (1W/4B)
-- 
cgit v1.1


From 72ac6969033dc9f5e526566240a3a7934f0916ee Mon Sep 17 00:00:00 2001
From: Satendra Singh Thakur <thakursatendra2003@yahoo.co.in>
Date: Sat, 31 Mar 2018 20:17:58 +0530
Subject: drm/mediatek: Using the function drm_display_mode_to_videomode

This patch uses existing method drm_display_mode_to_videomode for
calculating front/back porches, sync lengths for mediatek dsi/dpi
drivers; instead of manually calculating them

Signed-off-by: Satendra Singh Thakur <thakursatendra2003@yahoo.co.in>
Signed-off-by: CK Hu <ck.hu@mediatek.com>
---
 drivers/gpu/drm/mediatek/Kconfig   |  1 +
 drivers/gpu/drm/mediatek/mtk_dpi.c | 60 +++++++++++++++++++-------------------
 drivers/gpu/drm/mediatek/mtk_dsi.c | 14 ++-------
 3 files changed, 33 insertions(+), 42 deletions(-)

diff --git a/drivers/gpu/drm/mediatek/Kconfig b/drivers/gpu/drm/mediatek/Kconfig
index 294de45..119ec0a 100644
--- a/drivers/gpu/drm/mediatek/Kconfig
+++ b/drivers/gpu/drm/mediatek/Kconfig
@@ -11,6 +11,7 @@ config DRM_MEDIATEK
 	select DRM_PANEL
 	select MEMORY
 	select MTK_SMI
+	select VIDEOMODE_HELPERS
 	help
 	  Choose this option if you have a Mediatek SoCs.
 	  The module will be called mediatek-drm
diff --git a/drivers/gpu/drm/mediatek/mtk_dpi.c b/drivers/gpu/drm/mediatek/mtk_dpi.c
index e80a603..6c0ea39 100644
--- a/drivers/gpu/drm/mediatek/mtk_dpi.c
+++ b/drivers/gpu/drm/mediatek/mtk_dpi.c
@@ -22,6 +22,7 @@
 #include <linux/interrupt.h>
 #include <linux/types.h>
 #include <linux/clk.h>
+#include <video/videomode.h>
 
 #include "mtk_dpi_regs.h"
 #include "mtk_drm_ddp_comp.h"
@@ -429,34 +430,35 @@ static int mtk_dpi_set_display_mode(struct mtk_dpi *dpi,
 	struct mtk_dpi_sync_param vsync_leven = { 0 };
 	struct mtk_dpi_sync_param vsync_rodd = { 0 };
 	struct mtk_dpi_sync_param vsync_reven = { 0 };
-	unsigned long pix_rate;
+	struct videomode vm = { 0 };
 	unsigned long pll_rate;
 	unsigned int factor;
 
 	/* let pll_rate can fix the valid range of tvdpll (1G~2GHz) */
-	pix_rate = 1000UL * mode->clock;
+
 	if (mode->clock <= 27000)
-		factor = 16 * 3;
+		factor = 3 << 4;
 	else if (mode->clock <= 84000)
-		factor = 8 * 3;
+		factor = 3 << 3;
 	else if (mode->clock <= 167000)
-		factor = 4 * 3;
+		factor = 3 << 2;
 	else
-		factor = 2 * 3;
-	pll_rate = pix_rate * factor;
+		factor = 3 << 1;
+	drm_display_mode_to_videomode(mode, &vm);
+	pll_rate = vm.pixelclock * factor;
 
 	dev_dbg(dpi->dev, "Want PLL %lu Hz, pixel clock %lu Hz\n",
-		pll_rate, pix_rate);
+		pll_rate, vm.pixelclock);
 
 	clk_set_rate(dpi->tvd_clk, pll_rate);
 	pll_rate = clk_get_rate(dpi->tvd_clk);
 
-	pix_rate = pll_rate / factor;
-	clk_set_rate(dpi->pixel_clk, pix_rate);
-	pix_rate = clk_get_rate(dpi->pixel_clk);
+	vm.pixelclock = pll_rate / factor;
+	clk_set_rate(dpi->pixel_clk, vm.pixelclock);
+	vm.pixelclock = clk_get_rate(dpi->pixel_clk);
 
 	dev_dbg(dpi->dev, "Got  PLL %lu Hz, pixel clock %lu Hz\n",
-		pll_rate, pix_rate);
+		pll_rate, vm.pixelclock);
 
 	limit.c_bottom = 0x0010;
 	limit.c_top = 0x0FE0;
@@ -465,33 +467,31 @@ static int mtk_dpi_set_display_mode(struct mtk_dpi *dpi,
 
 	dpi_pol.ck_pol = MTK_DPI_POLARITY_FALLING;
 	dpi_pol.de_pol = MTK_DPI_POLARITY_RISING;
-	dpi_pol.hsync_pol = mode->flags & DRM_MODE_FLAG_PHSYNC ?
+	dpi_pol.hsync_pol = vm.flags & DISPLAY_FLAGS_HSYNC_HIGH ?
 			    MTK_DPI_POLARITY_FALLING : MTK_DPI_POLARITY_RISING;
-	dpi_pol.vsync_pol = mode->flags & DRM_MODE_FLAG_PVSYNC ?
+	dpi_pol.vsync_pol = vm.flags & DISPLAY_FLAGS_VSYNC_HIGH ?
 			    MTK_DPI_POLARITY_FALLING : MTK_DPI_POLARITY_RISING;
-
-	hsync.sync_width = mode->hsync_end - mode->hsync_start;
-	hsync.back_porch = mode->htotal - mode->hsync_end;
-	hsync.front_porch = mode->hsync_start - mode->hdisplay;
+	hsync.sync_width = vm.hsync_len;
+	hsync.back_porch = vm.hback_porch;
+	hsync.front_porch = vm.hfront_porch;
 	hsync.shift_half_line = false;
-
-	vsync_lodd.sync_width = mode->vsync_end - mode->vsync_start;
-	vsync_lodd.back_porch = mode->vtotal - mode->vsync_end;
-	vsync_lodd.front_porch = mode->vsync_start - mode->vdisplay;
+	vsync_lodd.sync_width = vm.vsync_len;
+	vsync_lodd.back_porch = vm.vback_porch;
+	vsync_lodd.front_porch = vm.vfront_porch;
 	vsync_lodd.shift_half_line = false;
 
-	if (mode->flags & DRM_MODE_FLAG_INTERLACE &&
+	if (vm.flags & DISPLAY_FLAGS_INTERLACED &&
 	    mode->flags & DRM_MODE_FLAG_3D_MASK) {
 		vsync_leven = vsync_lodd;
 		vsync_rodd = vsync_lodd;
 		vsync_reven = vsync_lodd;
 		vsync_leven.shift_half_line = true;
 		vsync_reven.shift_half_line = true;
-	} else if (mode->flags & DRM_MODE_FLAG_INTERLACE &&
+	} else if (vm.flags & DISPLAY_FLAGS_INTERLACED &&
 		   !(mode->flags & DRM_MODE_FLAG_3D_MASK)) {
 		vsync_leven = vsync_lodd;
 		vsync_leven.shift_half_line = true;
-	} else if (!(mode->flags & DRM_MODE_FLAG_INTERLACE) &&
+	} else if (!(vm.flags & DISPLAY_FLAGS_INTERLACED) &&
 		   mode->flags & DRM_MODE_FLAG_3D_MASK) {
 		vsync_rodd = vsync_lodd;
 	}
@@ -505,12 +505,12 @@ static int mtk_dpi_set_display_mode(struct mtk_dpi *dpi,
 	mtk_dpi_config_vsync_reven(dpi, &vsync_reven);
 
 	mtk_dpi_config_3d(dpi, !!(mode->flags & DRM_MODE_FLAG_3D_MASK));
-	mtk_dpi_config_interface(dpi, !!(mode->flags &
-					 DRM_MODE_FLAG_INTERLACE));
-	if (mode->flags & DRM_MODE_FLAG_INTERLACE)
-		mtk_dpi_config_fb_size(dpi, mode->hdisplay, mode->vdisplay / 2);
+	mtk_dpi_config_interface(dpi, !!(vm.flags &
+					 DISPLAY_FLAGS_INTERLACED));
+	if (vm.flags & DISPLAY_FLAGS_INTERLACED)
+		mtk_dpi_config_fb_size(dpi, vm.hactive, vm.vactive >> 1);
 	else
-		mtk_dpi_config_fb_size(dpi, mode->hdisplay, mode->vdisplay);
+		mtk_dpi_config_fb_size(dpi, vm.hactive, vm.vactive);
 
 	mtk_dpi_config_channel_limit(dpi, &limit);
 	mtk_dpi_config_bit_num(dpi, dpi->bit_num);
diff --git a/drivers/gpu/drm/mediatek/mtk_dsi.c b/drivers/gpu/drm/mediatek/mtk_dsi.c
index 7e5e24c..aa0943e 100644
--- a/drivers/gpu/drm/mediatek/mtk_dsi.c
+++ b/drivers/gpu/drm/mediatek/mtk_dsi.c
@@ -551,13 +551,12 @@ static int mtk_dsi_poweron(struct mtk_dsi *dsi)
 	}
 
 	/**
-	 * vm.pixelclock is in kHz, pixel_clock unit is Hz, so multiply by 1000
 	 * htotal_time = htotal * byte_per_pixel / num_lanes
 	 * overhead_time = lpx + hs_prepare + hs_zero + hs_trail + hs_exit
 	 * mipi_ratio = (htotal_time + overhead_time) / htotal_time
 	 * data_rate = pixel_clock * bit_per_pixel * mipi_ratio / num_lanes;
 	 */
-	pixel_clock = dsi->vm.pixelclock * 1000;
+	pixel_clock = dsi->vm.pixelclock;
 	htotal = dsi->vm.hactive + dsi->vm.hback_porch + dsi->vm.hfront_porch +
 			dsi->vm.hsync_len;
 	htotal_bits = htotal * bit_per_pixel;
@@ -725,16 +724,7 @@ static void mtk_dsi_encoder_mode_set(struct drm_encoder *encoder,
 {
 	struct mtk_dsi *dsi = encoder_to_dsi(encoder);
 
-	dsi->vm.pixelclock = adjusted->clock;
-	dsi->vm.hactive = adjusted->hdisplay;
-	dsi->vm.hback_porch = adjusted->htotal - adjusted->hsync_end;
-	dsi->vm.hfront_porch = adjusted->hsync_start - adjusted->hdisplay;
-	dsi->vm.hsync_len = adjusted->hsync_end - adjusted->hsync_start;
-
-	dsi->vm.vactive = adjusted->vdisplay;
-	dsi->vm.vback_porch = adjusted->vtotal - adjusted->vsync_end;
-	dsi->vm.vfront_porch = adjusted->vsync_start - adjusted->vdisplay;
-	dsi->vm.vsync_len = adjusted->vsync_end - adjusted->vsync_start;
+	drm_display_mode_to_videomode(adjusted, &dsi->vm);
 }
 
 static void mtk_dsi_encoder_disable(struct drm_encoder *encoder)
-- 
cgit v1.1


From f13876e2c33a657a71bcbb10f767c0951b165020 Mon Sep 17 00:00:00 2001
From: Takashi Iwai <tiwai@suse.de>
Date: Wed, 2 May 2018 08:48:46 +0200
Subject: ALSA: pcm: Check PCM state at xfern compat ioctl

Since snd_pcm_ioctl_xfern_compat() has no PCM state check, it may go
further and hit the sanity check pcm_sanity_check() when the ioctl is
called right after open.  It may eventually spew a kernel warning, as
triggered by syzbot, depending on kconfig.

The lack of PCM state check there was just an oversight.  Although
it's no real crash, the spurious kernel warning is annoying, so let's
add the proper check.

Reported-by: syzbot+1dac3a4f6bc9c1c675d4@syzkaller.appspotmail.com
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
---
 sound/core/pcm_compat.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/sound/core/pcm_compat.c b/sound/core/pcm_compat.c
index 06d7c40..6491afb 100644
--- a/sound/core/pcm_compat.c
+++ b/sound/core/pcm_compat.c
@@ -423,6 +423,8 @@ static int snd_pcm_ioctl_xfern_compat(struct snd_pcm_substream *substream,
 		return -ENOTTY;
 	if (substream->stream != dir)
 		return -EINVAL;
+	if (substream->runtime->status->state == SNDRV_PCM_STATE_OPEN)
+		return -EBADFD;
 
 	if ((ch = substream->runtime->channels) > 128)
 		return -EINVAL;
-- 
cgit v1.1


From 4205cb01f6e9ef2ae6daa7be4e8ac1edeb4c9d64 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?SZ=20Lin=20=28=E6=9E=97=E4=B8=8A=E6=99=BA=29?=
 <sz.lin@moxa.com>
Date: Thu, 26 Apr 2018 14:28:31 +0800
Subject: USB: serial: option: adding support for ublox R410M
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This patch adds support for ublox R410M PID 0x90b2 USB modem to option
driver, this module supports LTE Cat M1 / NB1.

Interface layout:
0: QCDM/DIAG
1: ADB
2: AT
3: RMNET

Signed-off-by: SZ Lin (林上智) <sz.lin@moxa.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Johan Hovold <johan@kernel.org>
---
 drivers/usb/serial/option.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/usb/serial/option.c b/drivers/usb/serial/option.c
index c3f2522..2058852 100644
--- a/drivers/usb/serial/option.c
+++ b/drivers/usb/serial/option.c
@@ -233,6 +233,8 @@ static void option_instat_callback(struct urb *urb);
 /* These Quectel products use Qualcomm's vendor ID */
 #define QUECTEL_PRODUCT_UC20			0x9003
 #define QUECTEL_PRODUCT_UC15			0x9090
+/* These u-blox products use Qualcomm's vendor ID */
+#define UBLOX_PRODUCT_R410M			0x90b2
 /* These Yuga products use Qualcomm's vendor ID */
 #define YUGA_PRODUCT_CLM920_NC5			0x9625
 
@@ -1065,6 +1067,9 @@ static const struct usb_device_id option_ids[] = {
 	/* Yuga products use Qualcomm vendor ID */
 	{ USB_DEVICE(QUALCOMM_VENDOR_ID, YUGA_PRODUCT_CLM920_NC5),
 	  .driver_info = RSVD(1) | RSVD(4) },
+	/* u-blox products using Qualcomm vendor ID */
+	{ USB_DEVICE(QUALCOMM_VENDOR_ID, UBLOX_PRODUCT_R410M),
+	  .driver_info = RSVD(1) | RSVD(3) },
 	/* Quectel products using Quectel vendor ID */
 	{ USB_DEVICE(QUECTEL_VENDOR_ID, QUECTEL_PRODUCT_EC21),
 	  .driver_info = RSVD(4) },
-- 
cgit v1.1


From 4842ed5bfcb9daf6660537d70503c18d38dbdbb8 Mon Sep 17 00:00:00 2001
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Date: Sun, 29 Apr 2018 17:41:55 +0200
Subject: USB: serial: visor: handle potential invalid device configuration

If we get an invalid device configuration from a palm 3 type device, we
might incorrectly parse things, and we have the potential to crash in
"interesting" ways.

Fix this up by verifying the size of the configuration passed to us by
the device, and only if it is correct, will we handle it.

Note that this also fixes an information leak of slab data.

Reported-by: Andrey Konovalov <andreyknvl@google.com>
Reviewed-by: Andrey Konovalov <andreyknvl@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ johan: add comment about the info leak ]
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Johan Hovold <johan@kernel.org>
---
 drivers/usb/serial/visor.c | 69 +++++++++++++++++++++++-----------------------
 1 file changed, 35 insertions(+), 34 deletions(-)

diff --git a/drivers/usb/serial/visor.c b/drivers/usb/serial/visor.c
index f5373ed..8ddbecc 100644
--- a/drivers/usb/serial/visor.c
+++ b/drivers/usb/serial/visor.c
@@ -335,47 +335,48 @@ static int palm_os_3_probe(struct usb_serial *serial,
 		goto exit;
 	}
 
-	if (retval == sizeof(*connection_info)) {
-			connection_info = (struct visor_connection_info *)
-							transfer_buffer;
-
-		num_ports = le16_to_cpu(connection_info->num_ports);
-		for (i = 0; i < num_ports; ++i) {
-			switch (
-			   connection_info->connections[i].port_function_id) {
-			case VISOR_FUNCTION_GENERIC:
-				string = "Generic";
-				break;
-			case VISOR_FUNCTION_DEBUGGER:
-				string = "Debugger";
-				break;
-			case VISOR_FUNCTION_HOTSYNC:
-				string = "HotSync";
-				break;
-			case VISOR_FUNCTION_CONSOLE:
-				string = "Console";
-				break;
-			case VISOR_FUNCTION_REMOTE_FILE_SYS:
-				string = "Remote File System";
-				break;
-			default:
-				string = "unknown";
-				break;
-			}
-			dev_info(dev, "%s: port %d, is for %s use\n",
-				serial->type->description,
-				connection_info->connections[i].port, string);
-		}
+	if (retval != sizeof(*connection_info)) {
+		dev_err(dev, "Invalid connection information received from device\n");
+		retval = -ENODEV;
+		goto exit;
 	}
-	/*
-	* Handle devices that report invalid stuff here.
-	*/
+
+	connection_info = (struct visor_connection_info *)transfer_buffer;
+
+	num_ports = le16_to_cpu(connection_info->num_ports);
+
+	/* Handle devices that report invalid stuff here. */
 	if (num_ports == 0 || num_ports > 2) {
 		dev_warn(dev, "%s: No valid connect info available\n",
 			serial->type->description);
 		num_ports = 2;
 	}
 
+	for (i = 0; i < num_ports; ++i) {
+		switch (connection_info->connections[i].port_function_id) {
+		case VISOR_FUNCTION_GENERIC:
+			string = "Generic";
+			break;
+		case VISOR_FUNCTION_DEBUGGER:
+			string = "Debugger";
+			break;
+		case VISOR_FUNCTION_HOTSYNC:
+			string = "HotSync";
+			break;
+		case VISOR_FUNCTION_CONSOLE:
+			string = "Console";
+			break;
+		case VISOR_FUNCTION_REMOTE_FILE_SYS:
+			string = "Remote File System";
+			break;
+		default:
+			string = "unknown";
+			break;
+		}
+		dev_info(dev, "%s: port %d, is for %s use\n",
+			serial->type->description,
+			connection_info->connections[i].port, string);
+	}
 	dev_info(dev, "%s: Number of ports: %d\n", serial->type->description,
 		num_ports);
 
-- 
cgit v1.1


From 043477b088d2af61a0937754c6560002237e6741 Mon Sep 17 00:00:00 2001
From: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Date: Mon, 30 Apr 2018 10:52:59 +0300
Subject: drm/i915: Print error state times relative to capture

Using plain jiffies in error state output makes the output
time differences relative to the current system time. This
is wrong as it makes output time differences dependent
of when the error state is printed rather than when it is
captured.

Store capture jiffies into error state and use it
when outputting the state to fix time differences output.

v2: use engine timestamp as epoch, output formatting (Chris)
v3: pass epoch to print_engine/request (Chris)

Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20180430075259.4476-1-mika.kuoppala@linux.intel.com
---
 drivers/gpu/drm/i915/i915_gpu_error.c | 47 ++++++++++++++++++++++++++++-------
 drivers/gpu/drm/i915/i915_gpu_error.h |  2 ++
 2 files changed, 40 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c
index c012796..1176d06 100644
--- a/drivers/gpu/drm/i915/i915_gpu_error.c
+++ b/drivers/gpu/drm/i915/i915_gpu_error.c
@@ -404,15 +404,16 @@ static const char *bannable(const struct drm_i915_error_context *ctx)
 
 static void error_print_request(struct drm_i915_error_state_buf *m,
 				const char *prefix,
-				const struct drm_i915_error_request *erq)
+				const struct drm_i915_error_request *erq,
+				const unsigned long epoch)
 {
 	if (!erq->seqno)
 		return;
 
-	err_printf(m, "%s pid %d, ban score %d, seqno %8x:%08x, prio %d, emitted %dms ago, head %08x, tail %08x\n",
+	err_printf(m, "%s pid %d, ban score %d, seqno %8x:%08x, prio %d, emitted %dms, head %08x, tail %08x\n",
 		   prefix, erq->pid, erq->ban_score,
 		   erq->context, erq->seqno, erq->sched_attr.priority,
-		   jiffies_to_msecs(jiffies - erq->jiffies),
+		   jiffies_to_msecs(erq->jiffies - epoch),
 		   erq->head, erq->tail);
 }
 
@@ -427,7 +428,8 @@ static void error_print_context(struct drm_i915_error_state_buf *m,
 }
 
 static void error_print_engine(struct drm_i915_error_state_buf *m,
-			       const struct drm_i915_error_engine *ee)
+			       const struct drm_i915_error_engine *ee,
+			       const unsigned long epoch)
 {
 	int n;
 
@@ -497,14 +499,15 @@ static void error_print_engine(struct drm_i915_error_state_buf *m,
 	err_printf(m, "  hangcheck stall: %s\n", yesno(ee->hangcheck_stalled));
 	err_printf(m, "  hangcheck action: %s\n",
 		   hangcheck_action_to_str(ee->hangcheck_action));
-	err_printf(m, "  hangcheck action timestamp: %lu, %u ms ago\n",
+	err_printf(m, "  hangcheck action timestamp: %dms (%lu%s)\n",
+		   jiffies_to_msecs(ee->hangcheck_timestamp - epoch),
 		   ee->hangcheck_timestamp,
-		   jiffies_to_msecs(jiffies - ee->hangcheck_timestamp));
+		   ee->hangcheck_timestamp == epoch ? "; epoch" : "");
 	err_printf(m, "  engine reset count: %u\n", ee->reset_count);
 
 	for (n = 0; n < ee->num_ports; n++) {
 		err_printf(m, "  ELSP[%d]:", n);
-		error_print_request(m, " ", &ee->execlist[n]);
+		error_print_request(m, " ", &ee->execlist[n], epoch);
 	}
 
 	error_print_context(m, "  Active context: ", &ee->context);
@@ -650,6 +653,11 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
 	ts = ktime_to_timespec64(error->uptime);
 	err_printf(m, "Uptime: %lld s %ld us\n",
 		   (s64)ts.tv_sec, ts.tv_nsec / NSEC_PER_USEC);
+	err_printf(m, "Epoch: %lu jiffies (%u HZ)\n", error->epoch, HZ);
+	err_printf(m, "Capture: %lu jiffies; %d ms ago, %d ms after epoch\n",
+		   error->capture,
+		   jiffies_to_msecs(jiffies - error->capture),
+		   jiffies_to_msecs(error->capture - error->epoch));
 
 	for (i = 0; i < ARRAY_SIZE(error->engine); i++) {
 		if (error->engine[i].hangcheck_stalled &&
@@ -710,7 +718,7 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
 
 	for (i = 0; i < ARRAY_SIZE(error->engine); i++) {
 		if (error->engine[i].engine_id != -1)
-			error_print_engine(m, &error->engine[i]);
+			error_print_engine(m, &error->engine[i], error->epoch);
 	}
 
 	for (i = 0; i < ARRAY_SIZE(error->active_vm); i++) {
@@ -769,7 +777,9 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
 				   dev_priv->engine[i]->name,
 				   ee->num_requests);
 			for (j = 0; j < ee->num_requests; j++)
-				error_print_request(m, " ", &ee->requests[j]);
+				error_print_request(m, " ",
+						    &ee->requests[j],
+						    error->epoch);
 		}
 
 		if (IS_ERR(ee->waiters)) {
@@ -1736,6 +1746,22 @@ static void capture_params(struct i915_gpu_state *error)
 #undef DUP
 }
 
+static unsigned long capture_find_epoch(const struct i915_gpu_state *error)
+{
+	unsigned long epoch = error->capture;
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(error->engine); i++) {
+		const struct drm_i915_error_engine *ee = &error->engine[i];
+
+		if (ee->hangcheck_stalled &&
+		    time_before(ee->hangcheck_timestamp, epoch))
+			epoch = ee->hangcheck_timestamp;
+	}
+
+	return epoch;
+}
+
 static int capture(void *data)
 {
 	struct i915_gpu_state *error = data;
@@ -1744,6 +1770,7 @@ static int capture(void *data)
 	error->boottime = ktime_get_boottime();
 	error->uptime = ktime_sub(ktime_get(),
 				  error->i915->gt.last_init_time);
+	error->capture = jiffies;
 
 	capture_params(error);
 	capture_gen_state(error);
@@ -1757,6 +1784,8 @@ static int capture(void *data)
 	error->overlay = intel_overlay_capture_error_state(error->i915);
 	error->display = intel_display_capture_error_state(error->i915);
 
+	error->epoch = capture_find_epoch(error);
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/i915/i915_gpu_error.h b/drivers/gpu/drm/i915/i915_gpu_error.h
index 5d6fdcb..0accd2e 100644
--- a/drivers/gpu/drm/i915/i915_gpu_error.h
+++ b/drivers/gpu/drm/i915/i915_gpu_error.h
@@ -31,6 +31,8 @@ struct i915_gpu_state {
 	ktime_t time;
 	ktime_t boottime;
 	ktime_t uptime;
+	unsigned long capture;
+	unsigned long epoch;
 
 	struct drm_i915_private *i915;
 
-- 
cgit v1.1


From 998ac6d21cfd6efd58f5edf420bae8839dda9f2a Mon Sep 17 00:00:00 2001
From: ethanwu <ethanwu@synology.com>
Date: Sun, 29 Apr 2018 15:59:42 +0800
Subject: btrfs: Take trans lock before access running trans in
 check_delayed_ref

In preivous patch:
Btrfs: kill trans in run_delalloc_nocow and btrfs_cross_ref_exist
We avoid starting btrfs transaction and get this information from
fs_info->running_transaction directly.

When accessing running_transaction in check_delayed_ref, there's a
chance that current transaction will be freed by commit transaction
after the NULL pointer check of running_transaction is passed.

After looking all the other places using fs_info->running_transaction,
they are either protected by trans_lock or holding the transactions.

Fix this by using trans_lock and increasing the use_count.

Fixes: e4c3b2dcd144 ("Btrfs: kill trans in run_delalloc_nocow and btrfs_cross_ref_exist")
CC: stable@vger.kernel.org # 4.14+
Signed-off-by: ethanwu <ethanwu@synology.com>
Signed-off-by: David Sterba <dsterba@suse.com>
---
 fs/btrfs/extent-tree.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index f991020..3871658 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -3142,7 +3142,11 @@ static noinline int check_delayed_ref(struct btrfs_root *root,
 	struct rb_node *node;
 	int ret = 0;
 
+	spin_lock(&root->fs_info->trans_lock);
 	cur_trans = root->fs_info->running_transaction;
+	if (cur_trans)
+		refcount_inc(&cur_trans->use_count);
+	spin_unlock(&root->fs_info->trans_lock);
 	if (!cur_trans)
 		return 0;
 
@@ -3151,6 +3155,7 @@ static noinline int check_delayed_ref(struct btrfs_root *root,
 	head = btrfs_find_delayed_ref_head(delayed_refs, bytenr);
 	if (!head) {
 		spin_unlock(&delayed_refs->lock);
+		btrfs_put_transaction(cur_trans);
 		return 0;
 	}
 
@@ -3167,6 +3172,7 @@ static noinline int check_delayed_ref(struct btrfs_root *root,
 		mutex_lock(&head->mutex);
 		mutex_unlock(&head->mutex);
 		btrfs_put_delayed_ref_head(head);
+		btrfs_put_transaction(cur_trans);
 		return -EAGAIN;
 	}
 	spin_unlock(&delayed_refs->lock);
@@ -3199,6 +3205,7 @@ static noinline int check_delayed_ref(struct btrfs_root *root,
 	}
 	spin_unlock(&head->lock);
 	mutex_unlock(&head->mutex);
+	btrfs_put_transaction(cur_trans);
 	return ret;
 }
 
-- 
cgit v1.1


From a6aa10c70bf72fb28504cb5de5deac75da78b0f5 Mon Sep 17 00:00:00 2001
From: Filipe Manana <fdmanana@suse.com>
Date: Mon, 30 Apr 2018 19:05:07 +0100
Subject: Btrfs: send, fix missing truncate for inode with prealloc extent past
 eof

An incremental send operation can miss a truncate operation when an inode
has an increased size in the send snapshot and a prealloc extent beyond
its size.

Consider the following scenario where a necessary truncate operation is
missing in the incremental send stream:

1) In the parent snapshot an inode has a size of 1282957 bytes and it has
   no prealloc extents beyond its size;

2) In the the send snapshot it has a size of 5738496 bytes and has a new
   extent at offsets 1884160 (length of 106496 bytes) and a prealloc
   extent beyond eof at offset 6729728 (and a length of 339968 bytes);

3) When processing the prealloc extent, at offset 6729728, we end up at
   send.c:send_write_or_clone() and set the @len variable to a value of
   18446744073708560384 because @offset plus the original @len value is
   larger then the inode's size (6729728 + 339968 > 5738496). We then
   call send_extent_data(), with that @offset and @len, which in turn
   calls send_write(), and then the later calls fill_read_buf(). Because
   the offset passed to fill_read_buf() is greater then inode's i_size,
   this function returns 0 immediately, which makes send_write() and
   send_extent_data() do nothing and return immediately as well. When
   we get back to send.c:send_write_or_clone() we adjust the value
   of sctx->cur_inode_next_write_offset to @offset plus @len, which
   corresponds to 6729728 + 18446744073708560384 = 5738496, which is
   precisely the the size of the inode in the send snapshot;

4) Later when at send.c:finish_inode_if_needed() we determine that
   we don't need to issue a truncate operation because the value of
   sctx->cur_inode_next_write_offset corresponds to the inode's new
   size, 5738496 bytes. This is wrong because the last write operation
   that was issued started at offset 1884160 with a length of 106496
   bytes, so the correct value for sctx->cur_inode_next_write_offset
   should be 1990656 (1884160 + 106496), so that a truncate operation
   with a value of 5738496 bytes would have been sent to insert a
   trailing hole at the destination.

So fix the issue by making send.c:send_write_or_clone() not attempt
to send write or clone operations for extents that start beyond the
inode's size, since such attempts do nothing but waste time by
calling helper functions and allocating path structures, and send
currently has no fallocate command in order to create prealloc extents
at the destination (either beyond a file's eof or not).

The issue was found running the test btrfs/007 from fstests using a seed
value of 1524346151 for fsstress.

Reported-by: Gu, Jinxiang <gujx@cn.fujitsu.com>
Fixes: ffa7c4296e93 ("Btrfs: send, do not issue unnecessary truncate operations")
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
---
 fs/btrfs/send.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/fs/btrfs/send.c b/fs/btrfs/send.c
index 221e5cd..c0074d2 100644
--- a/fs/btrfs/send.c
+++ b/fs/btrfs/send.c
@@ -5236,6 +5236,10 @@ static int send_write_or_clone(struct send_ctx *sctx,
 		len = btrfs_file_extent_num_bytes(path->nodes[0], ei);
 	}
 
+	if (offset >= sctx->cur_inode_size) {
+		ret = 0;
+		goto out;
+	}
 	if (offset + len > sctx->cur_inode_size)
 		len = sctx->cur_inode_size - offset;
 	if (len == 0) {
-- 
cgit v1.1


From f425d08bf17a68ade12424f420feadb70113f23a Mon Sep 17 00:00:00 2001
From: Jani Nikula <jani.nikula@intel.com>
Date: Tue, 24 Apr 2018 15:20:16 +0300
Subject: drm/i915: add support for specifying DMC firmware override by module
 param

Use i915.dmc_firmware_path to override default firmware for the platform
and bypassing version checks.

v2: add missing param struct member declaration (David)

Tested-by: David Weinehall <david.weinehall@linux.intel.com>
Reviewed-by: David Weinehall <david.weinehall@linux.intel.com>
Cc: Anusha Srivatsa <anusha.srivatsa@intel.com>
Cc: David Weinehall <david.weinehall@linux.intel.com>
Acked-by: Anusha Srivatsa <anusha.srivatsa@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180424122016.2416-1-jani.nikula@intel.com
---
 drivers/gpu/drm/i915/i915_params.c | 3 +++
 drivers/gpu/drm/i915/i915_params.h | 1 +
 drivers/gpu/drm/i915/intel_csr.c   | 9 +++++++--
 3 files changed, 11 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_params.c b/drivers/gpu/drm/i915/i915_params.c
index 08108ce..66ea355 100644
--- a/drivers/gpu/drm/i915/i915_params.c
+++ b/drivers/gpu/drm/i915/i915_params.c
@@ -164,6 +164,9 @@ i915_param_named_unsafe(guc_firmware_path, charp, 0400,
 i915_param_named_unsafe(huc_firmware_path, charp, 0400,
 	"HuC firmware path to use instead of the default one");
 
+i915_param_named_unsafe(dmc_firmware_path, charp, 0400,
+	"DMC firmware path to use instead of the default one");
+
 i915_param_named_unsafe(enable_dp_mst, bool, 0600,
 	"Enable multi-stream transport (MST) for new DisplayPort sinks. (default: true)");
 
diff --git a/drivers/gpu/drm/i915/i915_params.h b/drivers/gpu/drm/i915/i915_params.h
index c963603..6684025 100644
--- a/drivers/gpu/drm/i915/i915_params.h
+++ b/drivers/gpu/drm/i915/i915_params.h
@@ -51,6 +51,7 @@ struct drm_printer;
 	param(int, guc_log_level, -1) \
 	param(char *, guc_firmware_path, NULL) \
 	param(char *, huc_firmware_path, NULL) \
+	param(char *, dmc_firmware_path, NULL) \
 	param(int, mmio_debug, 0) \
 	param(int, edp_vswing, 0) \
 	param(int, reset, 2) \
diff --git a/drivers/gpu/drm/i915/intel_csr.c b/drivers/gpu/drm/i915/intel_csr.c
index f9550ea..cf9b600 100644
--- a/drivers/gpu/drm/i915/intel_csr.c
+++ b/drivers/gpu/drm/i915/intel_csr.c
@@ -298,7 +298,10 @@ static uint32_t *parse_csr_fw(struct drm_i915_private *dev_priv,
 
 	csr->version = css_header->version;
 
-	if (IS_CANNONLAKE(dev_priv)) {
+	if (csr->fw_path == i915_modparams.dmc_firmware_path) {
+		/* Bypass version check for firmware override. */
+		required_version = csr->version;
+	} else if (IS_CANNONLAKE(dev_priv)) {
 		required_version = CNL_CSR_VERSION_REQUIRED;
 	} else if (IS_GEMINILAKE(dev_priv)) {
 		required_version = GLK_CSR_VERSION_REQUIRED;
@@ -453,7 +456,9 @@ void intel_csr_ucode_init(struct drm_i915_private *dev_priv)
 	if (!HAS_CSR(dev_priv))
 		return;
 
-	if (IS_CANNONLAKE(dev_priv))
+	if (i915_modparams.dmc_firmware_path)
+		csr->fw_path = i915_modparams.dmc_firmware_path;
+	else if (IS_CANNONLAKE(dev_priv))
 		csr->fw_path = I915_CSR_CNL;
 	else if (IS_GEMINILAKE(dev_priv))
 		csr->fw_path = I915_CSR_GLK;
-- 
cgit v1.1


From 77cbe925bf77bd3159f49c4db0ea89a2045d9071 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Tue, 17 Apr 2018 18:06:38 +0100
Subject: drm/i915/selftests: Fix error checking for wait_var_timeout

The old wait_on_atomic_t used a custom callback to perform the
schedule(), which used my return semantics of reporting an error code on
timeout. wait_var_event_timeout() uses the schedule() return semantics
of reporting the remaining jiffies (1 if it timed out with 0 jiffies
remaining!) and 0 on failure. This semantic mismatch lead to us falsely
claiming a time out occurred.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=106085
Fixes: d224985a5e31 ("sched/wait, drivers/drm: Convert wait_on_atomic_t() usage to the new wait_var_event() API")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180417170638.20550-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/selftests/intel_breadcrumbs.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/i915/selftests/intel_breadcrumbs.c b/drivers/gpu/drm/i915/selftests/intel_breadcrumbs.c
index 4658002..d6926e7 100644
--- a/drivers/gpu/drm/i915/selftests/intel_breadcrumbs.c
+++ b/drivers/gpu/drm/i915/selftests/intel_breadcrumbs.c
@@ -412,10 +412,11 @@ static int igt_wakeup(void *arg)
 		 * that they are ready for the next test. We wait until all
 		 * threads are complete and waiting for us (i.e. not a seqno).
 		 */
-		err = wait_var_event_timeout(&done, !atomic_read(&done), 10 * HZ);
-		if (err) {
+		if (!wait_var_event_timeout(&done,
+					    !atomic_read(&done), 10 * HZ)) {
 			pr_err("Timed out waiting for %d remaining waiters\n",
 			       atomic_read(&done));
+			err = -ETIMEDOUT;
 			break;
 		}
 
-- 
cgit v1.1


From b84e54616a946f24eeeca8762cb70a9074b045e7 Mon Sep 17 00:00:00 2001
From: Yixun Lan <yixun.lan@amlogic.com>
Date: Tue, 17 Apr 2018 17:08:24 +0000
Subject: pinctrl: meson-axg: fix the range of aobus bank

The GPIOAO bank is range from GPIOAO_0 to GPIOAO_13.

Fixes: 83c566806a68 ("pinctrl: meson-axg: Add new pinctrl driver for Meson AXG SoC")
Reported-by: Xingyu Chen <xingyu.chen@amlogic.com>
Signed-off-by: Yixun Lan <yixun.lan@amlogic.com>
Acked-by: Kevin Hilman <khilman@baylibre.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
---
 drivers/pinctrl/meson/pinctrl-meson-axg.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/pinctrl/meson/pinctrl-meson-axg.c b/drivers/pinctrl/meson/pinctrl-meson-axg.c
index 4b91ff7..99a6cea 100644
--- a/drivers/pinctrl/meson/pinctrl-meson-axg.c
+++ b/drivers/pinctrl/meson/pinctrl-meson-axg.c
@@ -898,7 +898,7 @@ static struct meson_bank meson_axg_periphs_banks[] = {
 
 static struct meson_bank meson_axg_aobus_banks[] = {
 	/*   name    first      last      irq	pullen  pull    dir     out     in  */
-	BANK("AO",   GPIOAO_0,  GPIOAO_9, 0, 13, 0,  16,  0, 0,  0,  0,  0, 16,  1,  0),
+	BANK("AO",   GPIOAO_0,  GPIOAO_13, 0, 13, 0,  16,  0, 0,  0,  0,  0, 16,  1,  0),
 };
 
 static struct meson_pmx_bank meson_axg_periphs_pmx_banks[] = {
-- 
cgit v1.1


From 83b9dc11312f48a561594a895672abb6cb2a2250 Mon Sep 17 00:00:00 2001
From: Mika Westerberg <mika.westerberg@linux.intel.com>
Date: Wed, 25 Apr 2018 13:32:11 +0300
Subject: pinctrl: cherryview: Associate IRQ descriptors to irqdomain

When we dropped the custom Linux GPIO translation it resulted that the
IRQ numbers changed slightly as well. Normally this would be fine
because everyone is expected to use controller relative GPIO numbers and
ACPI GpioIo/GpioInt resources. However, there is a certain set of
Intel_Strago based Chromebooks where i8042 keyboard controller IRQ
number is hardcoded be 182 (this is corrected with newer coreboot but
the older ones still have the hardcoded Linux IRQ number). Because of
this hardcoded IRQ number keyboard on those systems accidentally broke
again.

Fix this by iteratively associating IRQ descriptors to the chip irqdomain
so that there are no gaps on those systems. Other systems are not
affected.

Fixes: 03c4749dd6c7 ("gpio / ACPI: Drop unnecessary ACPI GPIO to Linux GPIO translation")
Link: https://bugzilla.kernel.org/show_bug.cgi?id=199463
Reported-by: Sultan Alsawaf <sultanxda@gmail.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
---
 drivers/pinctrl/intel/pinctrl-cherryview.c | 16 ++++++++++++----
 1 file changed, 12 insertions(+), 4 deletions(-)

diff --git a/drivers/pinctrl/intel/pinctrl-cherryview.c b/drivers/pinctrl/intel/pinctrl-cherryview.c
index b1ae161..fee9225 100644
--- a/drivers/pinctrl/intel/pinctrl-cherryview.c
+++ b/drivers/pinctrl/intel/pinctrl-cherryview.c
@@ -1622,22 +1622,30 @@ static int chv_gpio_probe(struct chv_pinctrl *pctrl, int irq)
 
 	if (!need_valid_mask) {
 		irq_base = devm_irq_alloc_descs(pctrl->dev, -1, 0,
-						chip->ngpio, NUMA_NO_NODE);
+						community->npins, NUMA_NO_NODE);
 		if (irq_base < 0) {
 			dev_err(pctrl->dev, "Failed to allocate IRQ numbers\n");
 			return irq_base;
 		}
-	} else {
-		irq_base = 0;
 	}
 
-	ret = gpiochip_irqchip_add(chip, &chv_gpio_irqchip, irq_base,
+	ret = gpiochip_irqchip_add(chip, &chv_gpio_irqchip, 0,
 				   handle_bad_irq, IRQ_TYPE_NONE);
 	if (ret) {
 		dev_err(pctrl->dev, "failed to add IRQ chip\n");
 		return ret;
 	}
 
+	if (!need_valid_mask) {
+		for (i = 0; i < community->ngpio_ranges; i++) {
+			range = &community->gpio_ranges[i];
+
+			irq_domain_associate_many(chip->irq.domain, irq_base,
+						  range->base, range->npins);
+			irq_base += range->npins;
+		}
+	}
+
 	gpiochip_set_chained_irqchip(chip, &chv_gpio_irqchip, irq,
 				     chv_gpio_irq_handler);
 	return 0;
-- 
cgit v1.1


From c41eb2c7f93531b8ea689b1e2bfe7e6b884a7213 Mon Sep 17 00:00:00 2001
From: Mika Westerberg <mika.westerberg@linux.intel.com>
Date: Thu, 29 Mar 2018 17:37:25 +0300
Subject: pinctrl: sunrisepoint: Align GPIO number space with Windows

It turns out that the Windows GPIO driver for Sunrisepoint PCH-H uses
similar bank structure than it does for Cannon Lake with the exception
that here the bank size is always 24 pins. Starting from pad group E the
BIOS/Windows GPIO numbering does not match the hardware anymore but
instead there are gaps to make each pad group ("bank") consume exactly
24 pins. Because of this Linux does not use correct pins for
GpioIo/GpioIo resources exposed by the BIOS.

This patch aligns the GPIO number space with BIOS/Windows to make sure
the same numbering scheme is used in Linux as well following what we did
already for Intel Cannon Lake.

Link: https://bugzilla.redhat.com/show_bug.cgi?id=1543769
Reported-by: Vivien FRASCA <vivien.frasca@gmail.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
---
 drivers/pinctrl/intel/pinctrl-sunrisepoint.c | 45 ++++++++++++++++++++++++++--
 1 file changed, 42 insertions(+), 3 deletions(-)

diff --git a/drivers/pinctrl/intel/pinctrl-sunrisepoint.c b/drivers/pinctrl/intel/pinctrl-sunrisepoint.c
index 8870a41..fee3435 100644
--- a/drivers/pinctrl/intel/pinctrl-sunrisepoint.c
+++ b/drivers/pinctrl/intel/pinctrl-sunrisepoint.c
@@ -36,6 +36,27 @@
 		.npins = ((e) - (s) + 1),		\
 	}
 
+#define SPTH_GPP(r, s, e, g)				\
+	{						\
+		.reg_num = (r),				\
+		.base = (s),				\
+		.size = ((e) - (s) + 1),		\
+		.gpio_base = (g),			\
+	}
+
+#define SPTH_COMMUNITY(b, s, e, g)			\
+	{						\
+		.barno = (b),				\
+		.padown_offset = SPT_PAD_OWN,		\
+		.padcfglock_offset = SPT_PADCFGLOCK,	\
+		.hostown_offset = SPT_HOSTSW_OWN,	\
+		.ie_offset = SPT_GPI_IE,		\
+		.pin_base = (s),			\
+		.npins = ((e) - (s) + 1),		\
+		.gpps = (g),				\
+		.ngpps = ARRAY_SIZE(g),			\
+	}
+
 /* Sunrisepoint-LP */
 static const struct pinctrl_pin_desc sptlp_pins[] = {
 	/* GPP_A */
@@ -531,10 +552,28 @@ static const struct intel_function spth_functions[] = {
 	FUNCTION("i2c2", spth_i2c2_groups),
 };
 
+static const struct intel_padgroup spth_community0_gpps[] = {
+	SPTH_GPP(0, 0, 23, 0),		/* GPP_A */
+	SPTH_GPP(1, 24, 47, 24),	/* GPP_B */
+};
+
+static const struct intel_padgroup spth_community1_gpps[] = {
+	SPTH_GPP(0, 48, 71, 48),	/* GPP_C */
+	SPTH_GPP(1, 72, 95, 72),	/* GPP_D */
+	SPTH_GPP(2, 96, 108, 96),	/* GPP_E */
+	SPTH_GPP(3, 109, 132, 120),	/* GPP_F */
+	SPTH_GPP(4, 133, 156, 144),	/* GPP_G */
+	SPTH_GPP(5, 157, 180, 168),	/* GPP_H */
+};
+
+static const struct intel_padgroup spth_community3_gpps[] = {
+	SPTH_GPP(0, 181, 191, 192),	/* GPP_I */
+};
+
 static const struct intel_community spth_communities[] = {
-	SPT_COMMUNITY(0, 0, 47),
-	SPT_COMMUNITY(1, 48, 180),
-	SPT_COMMUNITY(2, 181, 191),
+	SPTH_COMMUNITY(0, 0, 47, spth_community0_gpps),
+	SPTH_COMMUNITY(1, 48, 180, spth_community1_gpps),
+	SPTH_COMMUNITY(2, 181, 191, spth_community3_gpps),
 };
 
 static const struct intel_pinctrl_soc_data spth_soc_data = {
-- 
cgit v1.1


From 892a0be43edd63e1cd228af3453a064e9e94f08e Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Michel=20D=C3=A4nzer?= <michel.daenzer@amd.com>
Date: Tue, 1 May 2018 15:24:11 +0200
Subject: swiotlb: fix inversed DMA_ATTR_NO_WARN test
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The result was printing the warning only when we were explicitly asked
not to.

Cc: stable@vger.kernel.org
Fixes: 0176adb004065d6815a8e67946752df4cd947c5b "swiotlb: refactor
 coherent buffer allocation"
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>.
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 lib/swiotlb.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/lib/swiotlb.c b/lib/swiotlb.c
index fece575..12fbaa4 100644
--- a/lib/swiotlb.c
+++ b/lib/swiotlb.c
@@ -737,7 +737,7 @@ out_unmap:
 	swiotlb_tbl_unmap_single(dev, phys_addr, size, DMA_TO_DEVICE,
 			DMA_ATTR_SKIP_CPU_SYNC);
 out_warn:
-	if ((attrs & DMA_ATTR_NO_WARN) && printk_ratelimit()) {
+	if (!(attrs & DMA_ATTR_NO_WARN) && printk_ratelimit()) {
 		dev_warn(dev,
 			"swiotlb: coherent allocation failed, size=%zu\n",
 			size);
-- 
cgit v1.1


From 6732cfd4cac514b556f36b518670af91c8bdf19a Mon Sep 17 00:00:00 2001
From: Ladislav Michl <ladis@linux-mips.org>
Date: Wed, 2 May 2018 12:41:32 +0200
Subject: mtd: onenand: omap2: Disable DMA for HIGHMEM buffers

dma_map_single does not work for vmalloc-ed buffers,
so disable DMA in this case.

Signed-off-by: Ladislav Michl <ladis@linux-mips.org>
Reported-by: "H. Nikolaus Schaller" <hns@goldelico.com>
Tested-by: "H. Nikolaus Schaller" <hns@goldelico.com>
Reviewed-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: Boris Brezillon <boris.brezillon@bootlin.com>
---
 drivers/mtd/nand/onenand/omap2.c | 105 ++++++++++++++-------------------------
 1 file changed, 38 insertions(+), 67 deletions(-)

diff --git a/drivers/mtd/nand/onenand/omap2.c b/drivers/mtd/nand/onenand/omap2.c
index 9c159f0..3211371 100644
--- a/drivers/mtd/nand/onenand/omap2.c
+++ b/drivers/mtd/nand/onenand/omap2.c
@@ -375,56 +375,42 @@ static int omap2_onenand_read_bufferram(struct mtd_info *mtd, int area,
 {
 	struct omap2_onenand *c = container_of(mtd, struct omap2_onenand, mtd);
 	struct onenand_chip *this = mtd->priv;
-	dma_addr_t dma_src, dma_dst;
-	int bram_offset;
+	struct device *dev = &c->pdev->dev;
 	void *buf = (void *)buffer;
+	dma_addr_t dma_src, dma_dst;
+	int bram_offset, err;
 	size_t xtra;
-	int ret;
 
 	bram_offset = omap2_onenand_bufferram_offset(mtd, area) + area + offset;
-	if (bram_offset & 3 || (size_t)buf & 3 || count < 384)
-		goto out_copy;
-
-	/* panic_write() may be in an interrupt context */
-	if (in_interrupt() || oops_in_progress)
+	/*
+	 * If the buffer address is not DMA-able, len is not long enough to make
+	 * DMA transfers profitable or panic_write() may be in an interrupt
+	 * context fallback to PIO mode.
+	 */
+	if (!virt_addr_valid(buf) || bram_offset & 3 || (size_t)buf & 3 ||
+	    count < 384 || in_interrupt() || oops_in_progress )
 		goto out_copy;
 
-	if (buf >= high_memory) {
-		struct page *p1;
-
-		if (((size_t)buf & PAGE_MASK) !=
-		    ((size_t)(buf + count - 1) & PAGE_MASK))
-			goto out_copy;
-		p1 = vmalloc_to_page(buf);
-		if (!p1)
-			goto out_copy;
-		buf = page_address(p1) + ((size_t)buf & ~PAGE_MASK);
-	}
-
 	xtra = count & 3;
 	if (xtra) {
 		count -= xtra;
 		memcpy(buf + count, this->base + bram_offset + count, xtra);
 	}
 
+	dma_dst = dma_map_single(dev, buf, count, DMA_FROM_DEVICE);
 	dma_src = c->phys_base + bram_offset;
-	dma_dst = dma_map_single(&c->pdev->dev, buf, count, DMA_FROM_DEVICE);
-	if (dma_mapping_error(&c->pdev->dev, dma_dst)) {
-		dev_err(&c->pdev->dev,
-			"Couldn't DMA map a %d byte buffer\n",
-			count);
-		goto out_copy;
-	}
 
-	ret = omap2_onenand_dma_transfer(c, dma_src, dma_dst, count);
-	dma_unmap_single(&c->pdev->dev, dma_dst, count, DMA_FROM_DEVICE);
-
-	if (ret) {
-		dev_err(&c->pdev->dev, "timeout waiting for DMA\n");
+	if (dma_mapping_error(dev, dma_dst)) {
+		dev_err(dev, "Couldn't DMA map a %d byte buffer\n", count);
 		goto out_copy;
 	}
 
-	return 0;
+	err = omap2_onenand_dma_transfer(c, dma_src, dma_dst, count);
+	dma_unmap_single(dev, dma_dst, count, DMA_FROM_DEVICE);
+	if (!err)
+		return 0;
+
+	dev_err(dev, "timeout waiting for DMA\n");
 
 out_copy:
 	memcpy(buf, this->base + bram_offset, count);
@@ -437,49 +423,34 @@ static int omap2_onenand_write_bufferram(struct mtd_info *mtd, int area,
 {
 	struct omap2_onenand *c = container_of(mtd, struct omap2_onenand, mtd);
 	struct onenand_chip *this = mtd->priv;
-	dma_addr_t dma_src, dma_dst;
-	int bram_offset;
+	struct device *dev = &c->pdev->dev;
 	void *buf = (void *)buffer;
-	int ret;
+	dma_addr_t dma_src, dma_dst;
+	int bram_offset, err;
 
 	bram_offset = omap2_onenand_bufferram_offset(mtd, area) + area + offset;
-	if (bram_offset & 3 || (size_t)buf & 3 || count < 384)
-		goto out_copy;
-
-	/* panic_write() may be in an interrupt context */
-	if (in_interrupt() || oops_in_progress)
+	/*
+	 * If the buffer address is not DMA-able, len is not long enough to make
+	 * DMA transfers profitable or panic_write() may be in an interrupt
+	 * context fallback to PIO mode.
+	 */
+	if (!virt_addr_valid(buf) || bram_offset & 3 || (size_t)buf & 3 ||
+	    count < 384 || in_interrupt() || oops_in_progress )
 		goto out_copy;
 
-	if (buf >= high_memory) {
-		struct page *p1;
-
-		if (((size_t)buf & PAGE_MASK) !=
-		    ((size_t)(buf + count - 1) & PAGE_MASK))
-			goto out_copy;
-		p1 = vmalloc_to_page(buf);
-		if (!p1)
-			goto out_copy;
-		buf = page_address(p1) + ((size_t)buf & ~PAGE_MASK);
-	}
-
-	dma_src = dma_map_single(&c->pdev->dev, buf, count, DMA_TO_DEVICE);
+	dma_src = dma_map_single(dev, buf, count, DMA_TO_DEVICE);
 	dma_dst = c->phys_base + bram_offset;
-	if (dma_mapping_error(&c->pdev->dev, dma_src)) {
-		dev_err(&c->pdev->dev,
-			"Couldn't DMA map a %d byte buffer\n",
-			count);
-		return -1;
-	}
-
-	ret = omap2_onenand_dma_transfer(c, dma_src, dma_dst, count);
-	dma_unmap_single(&c->pdev->dev, dma_src, count, DMA_TO_DEVICE);
-
-	if (ret) {
-		dev_err(&c->pdev->dev, "timeout waiting for DMA\n");
+	if (dma_mapping_error(dev, dma_src)) {
+		dev_err(dev, "Couldn't DMA map a %d byte buffer\n", count);
 		goto out_copy;
 	}
 
-	return 0;
+	err = omap2_onenand_dma_transfer(c, dma_src, dma_dst, count);
+	dma_unmap_page(dev, dma_src, count, DMA_TO_DEVICE);
+	if (!err)
+		return 0;
+
+	dev_err(dev, "timeout waiting for DMA\n");
 
 out_copy:
 	memcpy(this->base + bram_offset, buf, count);
-- 
cgit v1.1


From 1bc2463cee92ef0e2034c813d5e511adeb58b5fd Mon Sep 17 00:00:00 2001
From: Agustin Vega-Frias <agustinv@codeaurora.org>
Date: Tue, 1 May 2018 10:14:50 -0400
Subject: irqchip/qcom: Fix check for spurious interrupts

When the interrupts for a combiner span multiple registers it must be
checked if any interrupts have been asserted on each register before
checking for spurious interrupts.

Checking each register seperately leads to false positive warnings.

[ tglx: Massaged changelog ]

Fixes: f20cc9b00c7b ("irqchip/qcom: Add IRQ combiner driver")
Signed-off-by: Agustin Vega-Frias <agustinv@codeaurora.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jason Cooper <jason@lakedaemon.net>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: timur@codeaurora.org
Cc: linux-arm-kernel@lists.infradead.org
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/1525184090-26143-1-git-send-email-agustinv@codeaurora.org
---
 drivers/irqchip/qcom-irq-combiner.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/irqchip/qcom-irq-combiner.c b/drivers/irqchip/qcom-irq-combiner.c
index f312659..7f0c0be 100644
--- a/drivers/irqchip/qcom-irq-combiner.c
+++ b/drivers/irqchip/qcom-irq-combiner.c
@@ -1,4 +1,4 @@
-/* Copyright (c) 2015-2016, The Linux Foundation. All rights reserved.
+/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved.
  *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of the GNU General Public License version 2 and
@@ -68,7 +68,7 @@ static void combiner_handle_irq(struct irq_desc *desc)
 
 		bit = readl_relaxed(combiner->regs[reg].addr);
 		status = bit & combiner->regs[reg].enabled;
-		if (!status)
+		if (bit && !status)
 			pr_warn_ratelimited("Unexpected IRQ on CPU%d: (%08x %08lx %p)\n",
 					    smp_processor_id(), bit,
 					    combiner->regs[reg].enabled,
-- 
cgit v1.1


From e9088adda13cd23249d4b0abb97ff8a81bf5573a Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <peterz@infradead.org>
Date: Mon, 30 Apr 2018 12:00:09 +0200
Subject: x86/tsc: Always unregister clocksource_tsc_early

Don't leave the tsc-early clocksource registered if it errors out
early.

This was reported by Diego, who on his Core2 era machine got TSC
invalidated while it was running with tsc-early (due to C-states).
This results in keeping tsc-early with very bad effects.

Reported-and-Tested-by: Diego Viola <diego.viola@gmail.com>
Fixes: aa83c45762a2 ("x86/tsc: Introduce early tsc clocksource")
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Cc: len.brown@intel.com
Cc: rjw@rjwysocki.net
Cc: diego.viola@gmail.com
Cc: rui.zhang@intel.com
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20180430100344.350507853@infradead.org
---
 arch/x86/kernel/tsc.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c
index 91e6da4..23f36bc 100644
--- a/arch/x86/kernel/tsc.c
+++ b/arch/x86/kernel/tsc.c
@@ -1244,7 +1244,7 @@ static void tsc_refine_calibration_work(struct work_struct *work)
 
 	/* Don't bother refining TSC on unstable systems */
 	if (tsc_unstable)
-		return;
+		goto unreg;
 
 	/*
 	 * Since the work is started early in boot, we may be
@@ -1297,11 +1297,12 @@ static void tsc_refine_calibration_work(struct work_struct *work)
 
 out:
 	if (tsc_unstable)
-		return;
+		goto unreg;
 
 	if (boot_cpu_has(X86_FEATURE_ART))
 		art_related_clocksource = &clocksource_tsc;
 	clocksource_register_khz(&clocksource_tsc, tsc_khz);
+unreg:
 	clocksource_unregister(&clocksource_tsc_early);
 }
 
@@ -1311,8 +1312,8 @@ static int __init init_tsc_clocksource(void)
 	if (!boot_cpu_has(X86_FEATURE_TSC) || tsc_disabled > 0 || !tsc_khz)
 		return 0;
 
-	if (check_tsc_unstable())
-		return 0;
+	if (tsc_unstable)
+		goto unreg;
 
 	if (tsc_clocksource_reliable)
 		clocksource_tsc.flags &= ~CLOCK_SOURCE_MUST_VERIFY;
@@ -1328,6 +1329,7 @@ static int __init init_tsc_clocksource(void)
 		if (boot_cpu_has(X86_FEATURE_ART))
 			art_related_clocksource = &clocksource_tsc;
 		clocksource_register_khz(&clocksource_tsc, tsc_khz);
+unreg:
 		clocksource_unregister(&clocksource_tsc_early);
 		return 0;
 	}
-- 
cgit v1.1


From 2aae7bcfa4104b770e6f612356adb8d66c6144d6 Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <peterz@infradead.org>
Date: Mon, 23 Apr 2018 17:28:55 +0200
Subject: clocksource: Allow clocksource_mark_unstable() on unregistered
 clocksources

Because of how the code flips between tsc-early and tsc clocksources
it might need to mark one or both unstable. The current code in
mark_tsc_unstable() only worked because previously it registered the
tsc clocksource once and then never touched it.

Since it now unregisters the tsc-early clocksource, it needs to know
if a clocksource got unregistered and the current cs->mult test
doesn't work for that. Instead use list_empty(&cs->list) to test for
registration.

Furthermore, since clocksource_mark_unstable() needs to place the cs
on the wd_list, it links the cs->list and cs->wd_list serialization.
It must not see a clocsource registered (!empty cs->list) but already
past dequeue_watchdog(). So place {en,de}queue{,_watchdog}() under the
same lock.

Provided cs->list is initialized to empty, this then allows us to
unconditionally use clocksource_mark_unstable(), regardless of the
registration state.

Fixes: aa83c45762a2 ("x86/tsc: Introduce early tsc clocksource")
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Tested-by: Diego Viola <diego.viola@gmail.com>
Cc: len.brown@intel.com
Cc: rjw@rjwysocki.net
Cc: diego.viola@gmail.com
Cc: rui.zhang@intel.com
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20180502135312.GS12217@hirez.programming.kicks-ass.net
---
 kernel/time/clocksource.c | 50 ++++++++++++++++++++++++++++++++---------------
 1 file changed, 34 insertions(+), 16 deletions(-)

diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c
index 0e974cf..c3d2b94 100644
--- a/kernel/time/clocksource.c
+++ b/kernel/time/clocksource.c
@@ -119,6 +119,16 @@ static DEFINE_SPINLOCK(watchdog_lock);
 static int watchdog_running;
 static atomic_t watchdog_reset_pending;
 
+static void inline clocksource_watchdog_lock(unsigned long *flags)
+{
+	spin_lock_irqsave(&watchdog_lock, *flags);
+}
+
+static void inline clocksource_watchdog_unlock(unsigned long *flags)
+{
+	spin_unlock_irqrestore(&watchdog_lock, *flags);
+}
+
 static int clocksource_watchdog_kthread(void *data);
 static void __clocksource_change_rating(struct clocksource *cs, int rating);
 
@@ -142,6 +152,9 @@ static void __clocksource_unstable(struct clocksource *cs)
 	cs->flags &= ~(CLOCK_SOURCE_VALID_FOR_HRES | CLOCK_SOURCE_WATCHDOG);
 	cs->flags |= CLOCK_SOURCE_UNSTABLE;
 
+	if (list_empty(&cs->list))
+		return;
+
 	if (cs->mark_unstable)
 		cs->mark_unstable(cs);
 
@@ -164,7 +177,7 @@ void clocksource_mark_unstable(struct clocksource *cs)
 
 	spin_lock_irqsave(&watchdog_lock, flags);
 	if (!(cs->flags & CLOCK_SOURCE_UNSTABLE)) {
-		if (list_empty(&cs->wd_list))
+		if (!list_empty(&cs->list) && list_empty(&cs->wd_list))
 			list_add(&cs->wd_list, &watchdog_list);
 		__clocksource_unstable(cs);
 	}
@@ -319,9 +332,6 @@ static void clocksource_resume_watchdog(void)
 
 static void clocksource_enqueue_watchdog(struct clocksource *cs)
 {
-	unsigned long flags;
-
-	spin_lock_irqsave(&watchdog_lock, flags);
 	if (cs->flags & CLOCK_SOURCE_MUST_VERIFY) {
 		/* cs is a clocksource to be watched. */
 		list_add(&cs->wd_list, &watchdog_list);
@@ -331,7 +341,6 @@ static void clocksource_enqueue_watchdog(struct clocksource *cs)
 		if (cs->flags & CLOCK_SOURCE_IS_CONTINUOUS)
 			cs->flags |= CLOCK_SOURCE_VALID_FOR_HRES;
 	}
-	spin_unlock_irqrestore(&watchdog_lock, flags);
 }
 
 static void clocksource_select_watchdog(bool fallback)
@@ -373,9 +382,6 @@ static void clocksource_select_watchdog(bool fallback)
 
 static void clocksource_dequeue_watchdog(struct clocksource *cs)
 {
-	unsigned long flags;
-
-	spin_lock_irqsave(&watchdog_lock, flags);
 	if (cs != watchdog) {
 		if (cs->flags & CLOCK_SOURCE_MUST_VERIFY) {
 			/* cs is a watched clocksource. */
@@ -384,21 +390,19 @@ static void clocksource_dequeue_watchdog(struct clocksource *cs)
 			clocksource_stop_watchdog();
 		}
 	}
-	spin_unlock_irqrestore(&watchdog_lock, flags);
 }
 
 static int __clocksource_watchdog_kthread(void)
 {
 	struct clocksource *cs, *tmp;
 	unsigned long flags;
-	LIST_HEAD(unstable);
 	int select = 0;
 
 	spin_lock_irqsave(&watchdog_lock, flags);
 	list_for_each_entry_safe(cs, tmp, &watchdog_list, wd_list) {
 		if (cs->flags & CLOCK_SOURCE_UNSTABLE) {
 			list_del_init(&cs->wd_list);
-			list_add(&cs->wd_list, &unstable);
+			__clocksource_change_rating(cs, 0);
 			select = 1;
 		}
 		if (cs->flags & CLOCK_SOURCE_RESELECT) {
@@ -410,11 +414,6 @@ static int __clocksource_watchdog_kthread(void)
 	clocksource_stop_watchdog();
 	spin_unlock_irqrestore(&watchdog_lock, flags);
 
-	/* Needs to be done outside of watchdog lock */
-	list_for_each_entry_safe(cs, tmp, &unstable, wd_list) {
-		list_del_init(&cs->wd_list);
-		__clocksource_change_rating(cs, 0);
-	}
 	return select;
 }
 
@@ -447,6 +446,9 @@ static inline int __clocksource_watchdog_kthread(void) { return 0; }
 static bool clocksource_is_watchdog(struct clocksource *cs) { return false; }
 void clocksource_mark_unstable(struct clocksource *cs) { }
 
+static void inline clocksource_watchdog_lock(unsigned long *flags) { }
+static void inline clocksource_watchdog_unlock(unsigned long *flags) { }
+
 #endif /* CONFIG_CLOCKSOURCE_WATCHDOG */
 
 /**
@@ -779,14 +781,19 @@ EXPORT_SYMBOL_GPL(__clocksource_update_freq_scale);
  */
 int __clocksource_register_scale(struct clocksource *cs, u32 scale, u32 freq)
 {
+	unsigned long flags;
 
 	/* Initialize mult/shift and max_idle_ns */
 	__clocksource_update_freq_scale(cs, scale, freq);
 
 	/* Add clocksource to the clocksource list */
 	mutex_lock(&clocksource_mutex);
+
+	clocksource_watchdog_lock(&flags);
 	clocksource_enqueue(cs);
 	clocksource_enqueue_watchdog(cs);
+	clocksource_watchdog_unlock(&flags);
+
 	clocksource_select();
 	clocksource_select_watchdog(false);
 	mutex_unlock(&clocksource_mutex);
@@ -808,8 +815,13 @@ static void __clocksource_change_rating(struct clocksource *cs, int rating)
  */
 void clocksource_change_rating(struct clocksource *cs, int rating)
 {
+	unsigned long flags;
+
 	mutex_lock(&clocksource_mutex);
+	clocksource_watchdog_lock(&flags);
 	__clocksource_change_rating(cs, rating);
+	clocksource_watchdog_unlock(&flags);
+
 	clocksource_select();
 	clocksource_select_watchdog(false);
 	mutex_unlock(&clocksource_mutex);
@@ -821,6 +833,8 @@ EXPORT_SYMBOL(clocksource_change_rating);
  */
 static int clocksource_unbind(struct clocksource *cs)
 {
+	unsigned long flags;
+
 	if (clocksource_is_watchdog(cs)) {
 		/* Select and try to install a replacement watchdog. */
 		clocksource_select_watchdog(true);
@@ -834,8 +848,12 @@ static int clocksource_unbind(struct clocksource *cs)
 		if (curr_clocksource == cs)
 			return -EBUSY;
 	}
+
+	clocksource_watchdog_lock(&flags);
 	clocksource_dequeue_watchdog(cs);
 	list_del_init(&cs->list);
+	clocksource_watchdog_unlock(&flags);
+
 	return 0;
 }
 
-- 
cgit v1.1


From 5b9e886a4af97574ca3ce1147f35545da0e7afc7 Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <peterz@infradead.org>
Date: Mon, 30 Apr 2018 12:00:11 +0200
Subject: clocksource: Initialize cs->wd_list

A number of places relies on list_empty(&cs->wd_list), however the
list_head does not get initialized. Do so upon registration, such that
thereafter it is possible to rely on list_empty() correctly reflecting
the list membership status.

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Diego Viola <diego.viola@gmail.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Cc: stable@vger.kernel.org
Cc: len.brown@intel.com
Cc: rjw@rjwysocki.net
Cc: rui.zhang@intel.com
Link: https://lkml.kernel.org/r/20180430100344.472662715@infradead.org
---
 kernel/time/clocksource.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c
index c3d2b94..935f39e 100644
--- a/kernel/time/clocksource.c
+++ b/kernel/time/clocksource.c
@@ -332,6 +332,8 @@ static void clocksource_resume_watchdog(void)
 
 static void clocksource_enqueue_watchdog(struct clocksource *cs)
 {
+	INIT_LIST_HEAD(&cs->wd_list);
+
 	if (cs->flags & CLOCK_SOURCE_MUST_VERIFY) {
 		/* cs is a clocksource to be watched. */
 		list_add(&cs->wd_list, &watchdog_list);
-- 
cgit v1.1


From e3b4f79025e0a4eb7e2a2c7d24dadfa1e38893b0 Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <peterz@infradead.org>
Date: Mon, 30 Apr 2018 12:00:12 +0200
Subject: x86/tsc: Fix mark_tsc_unstable()

mark_tsc_unstable() also needs to affect tsc_early, Now that
clocksource_mark_unstable() can be used on a clocksource irrespective of
its registration state, use it on both tsc_early and tsc.

This does however require cs->list to be initialized empty, otherwise it
cannot tell the registation state before registation.

Fixes: aa83c45762a2 ("x86/tsc: Introduce early tsc clocksource")
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Diego Viola <diego.viola@gmail.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Cc: len.brown@intel.com
Cc: rjw@rjwysocki.net
Cc: rui.zhang@intel.com
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20180430100344.533326547@infradead.org
---
 arch/x86/kernel/tsc.c | 12 +++++-------
 1 file changed, 5 insertions(+), 7 deletions(-)

diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c
index 23f36bc..74392d9 100644
--- a/arch/x86/kernel/tsc.c
+++ b/arch/x86/kernel/tsc.c
@@ -1067,6 +1067,7 @@ static struct clocksource clocksource_tsc_early = {
 	.resume			= tsc_resume,
 	.mark_unstable		= tsc_cs_mark_unstable,
 	.tick_stable		= tsc_cs_tick_stable,
+	.list			= LIST_HEAD_INIT(clocksource_tsc_early.list),
 };
 
 /*
@@ -1086,6 +1087,7 @@ static struct clocksource clocksource_tsc = {
 	.resume			= tsc_resume,
 	.mark_unstable		= tsc_cs_mark_unstable,
 	.tick_stable		= tsc_cs_tick_stable,
+	.list			= LIST_HEAD_INIT(clocksource_tsc.list),
 };
 
 void mark_tsc_unstable(char *reason)
@@ -1098,13 +1100,9 @@ void mark_tsc_unstable(char *reason)
 		clear_sched_clock_stable();
 	disable_sched_clock_irqtime();
 	pr_info("Marking TSC unstable due to %s\n", reason);
-	/* Change only the rating, when not registered */
-	if (clocksource_tsc.mult) {
-		clocksource_mark_unstable(&clocksource_tsc);
-	} else {
-		clocksource_tsc.flags |= CLOCK_SOURCE_UNSTABLE;
-		clocksource_tsc.rating = 0;
-	}
+
+	clocksource_mark_unstable(&clocksource_tsc_early);
+	clocksource_mark_unstable(&clocksource_tsc);
 }
 
 EXPORT_SYMBOL_GPL(mark_tsc_unstable);
-- 
cgit v1.1


From cd2af07d823e5287cd6c91d54337348c2a873462 Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <peterz@infradead.org>
Date: Mon, 30 Apr 2018 12:00:13 +0200
Subject: clocksource: Consistent de-rate when marking unstable

When a registered clocksource gets marked unstable the watchdog_kthread
will de-rate and re-select the clocksource. Ensure it also de-rates
when getting called on an unregistered clocksource.

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Cc: len.brown@intel.com
Cc: rjw@rjwysocki.net
Cc: diego.viola@gmail.com
Cc: rui.zhang@intel.com
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20180430100344.594904898@infradead.org
---
 kernel/time/clocksource.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c
index 935f39e..605656d 100644
--- a/kernel/time/clocksource.c
+++ b/kernel/time/clocksource.c
@@ -152,12 +152,19 @@ static void __clocksource_unstable(struct clocksource *cs)
 	cs->flags &= ~(CLOCK_SOURCE_VALID_FOR_HRES | CLOCK_SOURCE_WATCHDOG);
 	cs->flags |= CLOCK_SOURCE_UNSTABLE;
 
-	if (list_empty(&cs->list))
+	/*
+	 * If the clocksource is registered clocksource_watchdog_kthread() will
+	 * re-rate and re-select.
+	 */
+	if (list_empty(&cs->list)) {
+		cs->rating = 0;
 		return;
+	}
 
 	if (cs->mark_unstable)
 		cs->mark_unstable(cs);
 
+	/* kick clocksource_watchdog_kthread() */
 	if (finished_booting)
 		schedule_work(&watchdog_work);
 }
-- 
cgit v1.1


From 7dba33c6346c337aac3f7cd188137d4a6d3d1f3a Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <peterz@infradead.org>
Date: Mon, 30 Apr 2018 12:00:14 +0200
Subject: clocksource: Rework stale comment

AFAICS the hotplug code no longer uses this function.

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Cc: len.brown@intel.com
Cc: rjw@rjwysocki.net
Cc: diego.viola@gmail.com
Cc: rui.zhang@intel.com
Link: https://lkml.kernel.org/r/20180430100344.656525644@infradead.org
---
 kernel/time/clocksource.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c
index 605656d..84f3742 100644
--- a/kernel/time/clocksource.c
+++ b/kernel/time/clocksource.c
@@ -173,10 +173,8 @@ static void __clocksource_unstable(struct clocksource *cs)
  * clocksource_mark_unstable - mark clocksource unstable via watchdog
  * @cs:		clocksource to be marked unstable
  *
- * This function is called instead of clocksource_change_rating from
- * cpu hotplug code to avoid a deadlock between the clocksource mutex
- * and the cpu hotplug mutex. It defers the update of the clocksource
- * to the watchdog thread.
+ * This function is called by the x86 TSC code to mark clocksources as unstable;
+ * it defers demotion and re-selection to a kthread.
  */
 void clocksource_mark_unstable(struct clocksource *cs)
 {
-- 
cgit v1.1


From c818aa88d2d0cfc4938bfa9e226c0792af2dc45f Mon Sep 17 00:00:00 2001
From: "Michael S. Tsirkin" <mst@redhat.com>
Date: Wed, 2 May 2018 17:19:05 +0300
Subject: Revert "vhost: make msg padding explicit"

This reverts commit 93c0d549c4c5a7382ad70de6b86610b7aae57406.

Unfortunately the padding will break 32 bit userspace.
Ouch. Need to add some compat code, revert for now.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/uapi/linux/vhost.h | 1 -
 1 file changed, 1 deletion(-)

diff --git a/include/uapi/linux/vhost.h b/include/uapi/linux/vhost.h
index 5a8ad06..c51f8e5 100644
--- a/include/uapi/linux/vhost.h
+++ b/include/uapi/linux/vhost.h
@@ -68,7 +68,6 @@ struct vhost_iotlb_msg {
 
 struct vhost_msg {
 	int type;
-	int padding0;
 	union {
 		struct vhost_iotlb_msg iotlb;
 		__u8 padding[64];
-- 
cgit v1.1


From 3a068721a97320b2ffdbf0fc0685cc300dce5388 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Wed, 2 May 2018 11:41:50 +0100
Subject: drm/i915: Show ring->start for the ELSP context/request queue

Since the advent of execlists, the HW no longer executes from a single
statically assigned ring, but instead switches to a different ring for
each context (logical ringbuffer contexts as it is called). So a good way
to tally the executing context against what we have queued is by
comparing the RING_START register against our requests. Make it so.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180502104150.29874-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/i915_gpu_error.c  | 5 +++--
 drivers/gpu/drm/i915/i915_gpu_error.h  | 1 +
 drivers/gpu/drm/i915/intel_engine_cs.c | 5 +++--
 3 files changed, 7 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c
index 1176d06..9449399 100644
--- a/drivers/gpu/drm/i915/i915_gpu_error.c
+++ b/drivers/gpu/drm/i915/i915_gpu_error.c
@@ -410,11 +410,11 @@ static void error_print_request(struct drm_i915_error_state_buf *m,
 	if (!erq->seqno)
 		return;
 
-	err_printf(m, "%s pid %d, ban score %d, seqno %8x:%08x, prio %d, emitted %dms, head %08x, tail %08x\n",
+	err_printf(m, "%s pid %d, ban score %d, seqno %8x:%08x, prio %d, emitted %dms, start %08x, head %08x, tail %08x\n",
 		   prefix, erq->pid, erq->ban_score,
 		   erq->context, erq->seqno, erq->sched_attr.priority,
 		   jiffies_to_msecs(erq->jiffies - epoch),
-		   erq->head, erq->tail);
+		   erq->start, erq->head, erq->tail);
 }
 
 static void error_print_context(struct drm_i915_error_state_buf *m,
@@ -1292,6 +1292,7 @@ static void record_request(struct i915_request *request,
 	erq->ban_score = atomic_read(&request->ctx->ban_score);
 	erq->seqno = request->global_seqno;
 	erq->jiffies = request->emitted_jiffies;
+	erq->start = i915_ggtt_offset(request->ring->vma);
 	erq->head = request->head;
 	erq->tail = request->tail;
 
diff --git a/drivers/gpu/drm/i915/i915_gpu_error.h b/drivers/gpu/drm/i915/i915_gpu_error.h
index 0accd2e..dac0f8c 100644
--- a/drivers/gpu/drm/i915/i915_gpu_error.h
+++ b/drivers/gpu/drm/i915/i915_gpu_error.h
@@ -152,6 +152,7 @@ struct i915_gpu_state {
 			u32 context;
 			int ban_score;
 			u32 seqno;
+			u32 start;
 			u32 head;
 			u32 tail;
 			struct i915_sched_attr sched_attr;
diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
index 238c8d3..9164e6d 100644
--- a/drivers/gpu/drm/i915/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/intel_engine_cs.c
@@ -1278,8 +1278,9 @@ static void intel_engine_print_registers(const struct intel_engine_cs *engine,
 				char hdr[80];
 
 				snprintf(hdr, sizeof(hdr),
-					 "\t\tELSP[%d] count=%d, rq: ",
-					 idx, count);
+					 "\t\tELSP[%d] count=%d, ring->start=%08x, rq: ",
+					 idx, count,
+					 i915_ggtt_offset(rq->ring->vma));
 				print_request(m, rq, hdr);
 			} else {
 				drm_printf(m, "\t\tELSP[%d] idle\n", idx);
-- 
cgit v1.1


From c65732e4f72124ca5a3a0dd3bee0d3cee39c7170 Mon Sep 17 00:00:00 2001
From: Thomas Gleixner <tglx@linutronix.de>
Date: Mon, 30 Apr 2018 21:47:46 +0200
Subject: x86/cpu: Restore CPUID_8000_0008_EBX reload
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The recent commt which addresses the x86_phys_bits corruption with
encrypted memory on CPUID reload after a microcode update lost the reload
of CPUID_8000_0008_EBX as well.

As a consequence IBRS and IBRS_FW are not longer detected

Restore the behaviour by bringing the reload of CPUID_8000_0008_EBX
back. This restore has a twist due to the convoluted way the cpuid analysis
works:

CPUID_8000_0008_EBX is used by AMD to enumerate IBRB, IBRS, STIBP. On Intel
EBX is not used. But the speculation control code sets the AMD bits when
running on Intel depending on the Intel specific speculation control
bits. This was done to use the same bits for alternatives.

The change which moved the 8000_0008 evaluation out of get_cpu_cap() broke
this nasty scheme due to ordering. So that on Intel the store to
CPUID_8000_0008_EBX clears the IBRB, IBRS, STIBP bits which had been set
before by software.

So the actual CPUID_8000_0008_EBX needs to go back to the place where it
was and the phys/virt address space calculation cannot touch it.

In hindsight this should have used completely synthetic bits for IBRB,
IBRS, STIBP instead of reusing the AMD bits, but that's for 4.18.

/me needs to find time to cleanup that steaming pile of ...

Fixes: d94a155c59c9 ("x86/cpu: Prevent cpuinfo_x86::x86_phys_bits adjustment corruption")
Reported-by: Jörg Otte <jrg.otte@gmail.com>
Reported-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Jörg Otte <jrg.otte@gmail.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: kirill.shutemov@linux.intel.com
Cc: Borislav Petkov <bp@alien8.de
Link: https://lkml.kernel.org/r/alpine.DEB.2.21.1805021043510.1668@nanos.tec.linutronix.de
---
 arch/x86/kernel/cpu/common.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c
index 8a5b185..ce243f7 100644
--- a/arch/x86/kernel/cpu/common.c
+++ b/arch/x86/kernel/cpu/common.c
@@ -848,6 +848,11 @@ void get_cpu_cap(struct cpuinfo_x86 *c)
 		c->x86_power = edx;
 	}
 
+	if (c->extended_cpuid_level >= 0x80000008) {
+		cpuid(0x80000008, &eax, &ebx, &ecx, &edx);
+		c->x86_capability[CPUID_8000_0008_EBX] = ebx;
+	}
+
 	if (c->extended_cpuid_level >= 0x8000000a)
 		c->x86_capability[CPUID_8000_000A_EDX] = cpuid_edx(0x8000000a);
 
@@ -871,7 +876,6 @@ static void get_cpu_address_sizes(struct cpuinfo_x86 *c)
 
 		c->x86_virt_bits = (eax >> 8) & 0xff;
 		c->x86_phys_bits = eax & 0xff;
-		c->x86_capability[CPUID_8000_0008_EBX] = ebx;
 	}
 #ifdef CONFIG_X86_32
 	else if (cpu_has(c, X86_FEATURE_PAE) || cpu_has(c, X86_FEATURE_PSE36))
-- 
cgit v1.1


From 5e5add172ea81152d518b161ec5706503ad3d799 Mon Sep 17 00:00:00 2001
From: Grygorii Strashko <grygorii.strashko@ti.com>
Date: Tue, 1 May 2018 12:41:22 -0500
Subject: net: ethernet: ti: cpsw: fix packet leaking in dual_mac mode

In dual_mac mode packets arrived on one port should not be forwarded by
switch hw to another port. Only Linux Host can forward packets between
ports. The below test case (reported in [1]) shows that packet arrived on
one port can be leaked to anoter (reproducible with dual port evms):
 - connect port 1 (eth0) to linux Host 0 and run tcpdump or Wireshark
 - connect port 2 (eth1) to linux Host 1 with vlan 1 configured
 - ping <IPx> from Host 1 through vlan 1 interface.
ARP packets will be seen on Host 0.

Issue happens because dual_mac mode is implemnted using two vlans: 1 (Port
1+Port 0) and 2 (Port 2+Port 0), so there are vlan records created for for
each vlan. By default, the ALE will find valid vlan record in its table
when vlan 1 tagged packet arrived on Port 2 and so forwards packet to all
ports which are vlan 1 members (like Port.

To avoid such behaviorr the ALE VLAN ID Ingress Check need to be enabled
for each external CPSW port (ALE_PORTCTLn.VID_INGRESS_CHECK) so ALE will
drop ingress packets if Rx port is not VLAN member.

Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/ti/cpsw.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/net/ethernet/ti/cpsw.c b/drivers/net/ethernet/ti/cpsw.c
index 74f8284..28d893b 100644
--- a/drivers/net/ethernet/ti/cpsw.c
+++ b/drivers/net/ethernet/ti/cpsw.c
@@ -1340,6 +1340,8 @@ static inline void cpsw_add_dual_emac_def_ale_entries(
 	cpsw_ale_add_ucast(cpsw->ale, priv->mac_addr,
 			   HOST_PORT_NUM, ALE_VLAN |
 			   ALE_SECURE, slave->port_vlan);
+	cpsw_ale_control_set(cpsw->ale, slave_port,
+			     ALE_PORT_DROP_UNKNOWN_VLAN, 1);
 }
 
 static void soft_reset_slave(struct cpsw_slave *slave)
-- 
cgit v1.1


From e6e6a278b1eaffa19d42186bfacd1ffc15a50b3f Mon Sep 17 00:00:00 2001
From: Neal Cardwell <ncardwell@google.com>
Date: Tue, 1 May 2018 21:45:41 -0400
Subject: tcp_bbr: fix to zero idle_restart only upon S/ACKed data

Previously the bbr->idle_restart tracking was zeroing out the
bbr->idle_restart bit upon ACKs that did not SACK or ACK anything,
e.g. receiving incoming data or receiver window updates. In such
situations BBR would forget that this was a restart-from-idle
situation, and if the min_rtt had expired it would unnecessarily enter
PROBE_RTT (even though we were actually restarting from idle but had
merely forgotten that fact).

The fix is simple: we need to remember we are restarting from idle
until we receive a S/ACK for some data (a S/ACK for the first flight
of data we send as we are restarting).

This commit is a stable candidate for kernels back as far as 4.9.

Fixes: 0f8782ea1497 ("tcp_bbr: add BBR congestion control")
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: Priyaranjan Jha <priyarjha@google.com>
Signed-off-by: Yousuk Seung <ysseung@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/ipv4/tcp_bbr.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/net/ipv4/tcp_bbr.c b/net/ipv4/tcp_bbr.c
index 158d105..58e2f47 100644
--- a/net/ipv4/tcp_bbr.c
+++ b/net/ipv4/tcp_bbr.c
@@ -806,7 +806,9 @@ static void bbr_update_min_rtt(struct sock *sk, const struct rate_sample *rs)
 			}
 		}
 	}
-	bbr->idle_restart = 0;
+	/* Restart after idle ends only once we process a new S/ACK for data */
+	if (rs->delivered > 0)
+		bbr->idle_restart = 0;
 }
 
 static void bbr_update_model(struct sock *sk, const struct rate_sample *rs)
-- 
cgit v1.1


From 4842a08fb80bc09b7b089af42c58353dfaa8f88f Mon Sep 17 00:00:00 2001
From: Xin Long <lucien.xin@gmail.com>
Date: Wed, 2 May 2018 13:37:44 +0800
Subject: sctp: init active key for the new asoc in dupcook_a and dupcook_b

When processing a duplicate cookie-echo chunk, for case 'A' and 'B',
after sctp_process_init for the new asoc, if auth is enabled for the
cookie-ack chunk, the active key should also be initialized.

Otherwise, the cookie-ack chunk made later can not be set with auth
shkey properly, and a crash can even be caused by this, as after
Commit 1b1e0bc99474 ("sctp: add refcnt support for sh_key"), sctp
needs to hold the shkey when making control chunks.

Fixes: 1b1e0bc99474 ("sctp: add refcnt support for sh_key")
Reported-by: Jianwen Ji <jiji@redhat.com>
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/sctp/sm_statefuns.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/net/sctp/sm_statefuns.c b/net/sctp/sm_statefuns.c
index dd0594a..98acfed 100644
--- a/net/sctp/sm_statefuns.c
+++ b/net/sctp/sm_statefuns.c
@@ -1794,6 +1794,9 @@ static enum sctp_disposition sctp_sf_do_dupcook_a(
 			       GFP_ATOMIC))
 		goto nomem;
 
+	if (sctp_auth_asoc_init_active_key(new_asoc, GFP_ATOMIC))
+		goto nomem;
+
 	/* Make sure no new addresses are being added during the
 	 * restart.  Though this is a pretty complicated attack
 	 * since you'd have to get inside the cookie.
@@ -1906,6 +1909,9 @@ static enum sctp_disposition sctp_sf_do_dupcook_b(
 			       GFP_ATOMIC))
 		goto nomem;
 
+	if (sctp_auth_asoc_init_active_key(new_asoc, GFP_ATOMIC))
+		goto nomem;
+
 	/* Update the content of current association.  */
 	sctp_add_cmd_sf(commands, SCTP_CMD_UPDATE_ASSOC, SCTP_ASOC(new_asoc));
 	sctp_add_cmd_sf(commands, SCTP_CMD_NEW_STATE,
-- 
cgit v1.1


From 46e16d4b956867013e0bbd7f2bad206f4aa55752 Mon Sep 17 00:00:00 2001
From: Xin Long <lucien.xin@gmail.com>
Date: Wed, 2 May 2018 13:39:46 +0800
Subject: sctp: use the old asoc when making the cookie-ack chunk in dupcook_d

When processing a duplicate cookie-echo chunk, for case 'D', sctp will
not process the param from this chunk. It means old asoc has nothing
to be updated, and the new temp asoc doesn't have the complete info.

So there's no reason to use the new asoc when creating the cookie-ack
chunk. Otherwise, like when auth is enabled for cookie-ack, the chunk
can not be set with auth, and it will definitely be dropped by peer.

This issue is there since very beginning, and we fix it by using the
old asoc instead.

Signed-off-by: Xin Long <lucien.xin@gmail.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/sctp/sm_statefuns.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/sctp/sm_statefuns.c b/net/sctp/sm_statefuns.c
index 98acfed..28c070e 100644
--- a/net/sctp/sm_statefuns.c
+++ b/net/sctp/sm_statefuns.c
@@ -2056,7 +2056,7 @@ static enum sctp_disposition sctp_sf_do_dupcook_d(
 		}
 	}
 
-	repl = sctp_make_cookie_ack(new_asoc, chunk);
+	repl = sctp_make_cookie_ack(asoc, chunk);
 	if (!repl)
 		goto nomem;
 
-- 
cgit v1.1


From ce402f044e4e432c296f90eaabb8dbe8f3624391 Mon Sep 17 00:00:00 2001
From: Xin Long <lucien.xin@gmail.com>
Date: Wed, 2 May 2018 13:45:12 +0800
Subject: sctp: fix the issue that the cookie-ack with auth can't get processed

When auth is enabled for cookie-ack chunk, in sctp_inq_pop, sctp
processes auth chunk first, then continues to the next chunk in
this packet if chunk_end + chunk_hdr size < skb_tail_pointer().
Otherwise, it will go to the next packet or discard this chunk.

However, it missed the fact that cookie-ack chunk's size is equal
to chunk_hdr size, which couldn't match that check, and thus this
chunk would not get processed.

This patch fixes it by changing the check to chunk_end + chunk_hdr
size <= skb_tail_pointer().

Fixes: 26b87c788100 ("net: sctp: fix remote memory pressure from excessive queueing")
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/sctp/inqueue.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/sctp/inqueue.c b/net/sctp/inqueue.c
index 23ebc53..eb93ffe 100644
--- a/net/sctp/inqueue.c
+++ b/net/sctp/inqueue.c
@@ -217,7 +217,7 @@ new_skb:
 	skb_pull(chunk->skb, sizeof(*ch));
 	chunk->subh.v = NULL; /* Subheader is no longer valid.  */
 
-	if (chunk->chunk_end + sizeof(*ch) < skb_tail_pointer(chunk->skb)) {
+	if (chunk->chunk_end + sizeof(*ch) <= skb_tail_pointer(chunk->skb)) {
 		/* This is not a singleton */
 		chunk->singleton = 0;
 	} else if (chunk->chunk_end > skb_tail_pointer(chunk->skb)) {
-- 
cgit v1.1


From 46e2068081e96472b1d7d2456fa0655aa148b32b Mon Sep 17 00:00:00 2001
From: Matthias Kaehlcke <mka@chromium.org>
Date: Tue, 1 May 2018 11:24:40 -0700
Subject: drm/i915: Disable some extra clang warnings

Commit 39bf4de89ff7 ("drm/i915: Add -Wall -Wextra to our build, set
warnings to full") enabled extra warnings for i915 to spot possible
bugs in new code, and then disabled a subset of these warnings to keep
the current code building without warnings (with gcc). Enabling the
extra warnings also enabled some additional clang-only warnings, as a
result building i915 with clang currently is extremely noisy. For now
also disable the clang warnings sign-compare, sometimes-uninitialized,
unneeded-internal-declaration and initializer-overrides. If desired
they can be re-enabled after the code has been fixed.

Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20180501182440.70121-1-mka@chromium.org
---
 drivers/gpu/drm/i915/Makefile | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile
index 9bee52a9..dfe0145 100644
--- a/drivers/gpu/drm/i915/Makefile
+++ b/drivers/gpu/drm/i915/Makefile
@@ -18,6 +18,10 @@ subdir-ccflags-y += $(call cc-disable-warning, type-limits)
 subdir-ccflags-y += $(call cc-disable-warning, missing-field-initializers)
 subdir-ccflags-y += $(call cc-disable-warning, implicit-fallthrough)
 subdir-ccflags-y += $(call cc-disable-warning, unused-but-set-variable)
+# clang warnings
+subdir-ccflags-y += $(call cc-disable-warning, sign-compare)
+subdir-ccflags-y += $(call cc-disable-warning, sometimes-uninitialized)
+subdir-ccflags-y += $(call cc-disable-warning, initializer-overrides)
 subdir-ccflags-$(CONFIG_DRM_I915_WERROR) += -Werror
 
 # Fine grained warnings disable
-- 
cgit v1.1


From 9fc347678dde73abd75bce81d5dd82c4f4505184 Mon Sep 17 00:00:00 2001
From: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Date: Thu, 22 Mar 2018 22:05:23 +0100
Subject: modpost: delete stale comment

Commit 7840fea200cd ("kbuild: Fix computing srcversion for modules")
fixed the comment above parse_source_files to refer to the new source_
line, but left this one behind that could still give the impression that
drivers/net/dummy.c appears in the deps_ variable.

Signed-off-by: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
---
 scripts/mod/sumversion.c | 9 +--------
 1 file changed, 1 insertion(+), 8 deletions(-)

diff --git a/scripts/mod/sumversion.c b/scripts/mod/sumversion.c
index 944418d..0f6dcb40 100644
--- a/scripts/mod/sumversion.c
+++ b/scripts/mod/sumversion.c
@@ -330,14 +330,7 @@ static int parse_source_files(const char *objfile, struct md4_ctx *md)
 		goto out;
 	}
 
-	/* There will be a line like so:
-		deps_drivers/net/dummy.o := \
-		  drivers/net/dummy.c \
-		    $(wildcard include/config/net/fastroute.h) \
-		  include/linux/module.h \
-
-	   Sum all files in the same dir or subdirs.
-	*/
+	/* Sum all files in the same dir or subdirs. */
 	while ((line = get_next_line(&pos, file, flen)) != NULL) {
 		char* p = line;
 
-- 
cgit v1.1


From 021ba8e98fe5c6691b3cc3669faafa02403aa211 Mon Sep 17 00:00:00 2001
From: "Darrick J. Wong" <darrick.wong@oracle.com>
Date: Mon, 16 Apr 2018 23:07:36 -0700
Subject: xfs: cap the length of deduplication requests

Since deduplication potentially has to read in all the pages in both
files in order to compare the contents, cap the deduplication request
length at MAX_RW_COUNT/2 (roughly 1GB) so that we have /some/ upper bound
on the request length and can't just lock up the kernel forever.  Found
by running generic/304 after commit 1ddae54555b62 ("common/rc: add
missing 'local' keywords").

Reported-by: matorola@gmail.com
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com>
---
 fs/xfs/xfs_file.c | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c
index eed073c..e70fb8c 100644
--- a/fs/xfs/xfs_file.c
+++ b/fs/xfs/xfs_file.c
@@ -880,8 +880,18 @@ xfs_file_dedupe_range(
 	struct file	*dst_file,
 	u64		dst_loff)
 {
+	struct inode	*srci = file_inode(src_file);
+	u64		max_dedupe;
 	int		error;
 
+	/*
+	 * Since we have to read all these pages in to compare them, cut
+	 * it off at MAX_RW_COUNT/2 rounded down to the nearest block.
+	 * That means we won't do more than MAX_RW_COUNT IO per request.
+	 */
+	max_dedupe = (MAX_RW_COUNT >> 1) & ~(i_blocksize(srci) - 1);
+	if (len > max_dedupe)
+		len = max_dedupe;
 	error = xfs_reflink_remap_range(src_file, loff, dst_file, dst_loff,
 				     len, true);
 	if (error)
-- 
cgit v1.1


From 0d4b78b3d2c0cb570abe086bce8c17ea3c474095 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Tue, 1 May 2018 08:52:03 +0100
Subject: drm/i915/guc: Assert we have the doorbell before setting it up
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

As our early doorbell is split between early allocation and a late setup
after we have a channel to the GuC, it may happen due to a lapse of
programmer judgement that we try to setup an invalid doorbell. Make use
of our has_doorbell() function to check the doorbell does exist for the
client before we try and tell the guc about it. In doing so, we prevent
the compiler from warning about the otherwise unused function in some
configurations.

Reported-by: Matthias Kaehlcke <mka@chromium.org>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Michał Winiarski <michal.winiarski@intel.com>
Cc: Michal Wajdeczko <michal.wajdeczko@intel.com>
Cc: Michel Thierry <michel.thierry@intel.com>
Reviewed-by: Michel Thierry <michel.thierry@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180501075203.12458-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/intel_guc_submission.c | 22 ++++++++++++----------
 1 file changed, 12 insertions(+), 10 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_guc_submission.c b/drivers/gpu/drm/i915/intel_guc_submission.c
index 6e6ed0f..c6bb5be 100644
--- a/drivers/gpu/drm/i915/intel_guc_submission.c
+++ b/drivers/gpu/drm/i915/intel_guc_submission.c
@@ -124,9 +124,17 @@ static int reserve_doorbell(struct intel_guc_client *client)
 	return 0;
 }
 
+static bool has_doorbell(struct intel_guc_client *client)
+{
+	if (client->doorbell_id == GUC_DOORBELL_INVALID)
+		return false;
+
+	return test_bit(client->doorbell_id, client->guc->doorbell_bitmap);
+}
+
 static void unreserve_doorbell(struct intel_guc_client *client)
 {
-	GEM_BUG_ON(client->doorbell_id == GUC_DOORBELL_INVALID);
+	GEM_BUG_ON(!has_doorbell(client));
 
 	__clear_bit(client->doorbell_id, client->guc->doorbell_bitmap);
 	client->doorbell_id = GUC_DOORBELL_INVALID;
@@ -184,14 +192,6 @@ static struct guc_doorbell_info *__get_doorbell(struct intel_guc_client *client)
 	return client->vaddr + client->doorbell_offset;
 }
 
-static bool has_doorbell(struct intel_guc_client *client)
-{
-	if (client->doorbell_id == GUC_DOORBELL_INVALID)
-		return false;
-
-	return test_bit(client->doorbell_id, client->guc->doorbell_bitmap);
-}
-
 static void __create_doorbell(struct intel_guc_client *client)
 {
 	struct guc_doorbell_info *doorbell;
@@ -207,7 +207,6 @@ static void __destroy_doorbell(struct intel_guc_client *client)
 	struct guc_doorbell_info *doorbell;
 	u16 db_id = client->doorbell_id;
 
-
 	doorbell = __get_doorbell(client);
 	doorbell->db_status = GUC_DOORBELL_DISABLED;
 	doorbell->cookie = 0;
@@ -224,6 +223,9 @@ static int create_doorbell(struct intel_guc_client *client)
 {
 	int ret;
 
+	if (WARN_ON(!has_doorbell(client)))
+		return -ENODEV; /* internal setup error, should never happen */
+
 	__update_doorbell_desc(client, client->doorbell_id);
 	__create_doorbell(client);
 
-- 
cgit v1.1


From af3e0fcf78879f718c5f73df0814951bd7057d34 Mon Sep 17 00:00:00 2001
From: Ingo Molnar <mingo@elte.hu>
Date: Wed, 2 May 2018 13:30:57 +0200
Subject: 8139too: Use disable_irq_nosync() in rtl8139_poll_controller()

Use disable_irq_nosync() instead of disable_irq() as this might be
called in atomic context with netpoll.

Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/realtek/8139too.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/realtek/8139too.c b/drivers/net/ethernet/realtek/8139too.c
index d24b47b..d118da5 100644
--- a/drivers/net/ethernet/realtek/8139too.c
+++ b/drivers/net/ethernet/realtek/8139too.c
@@ -2224,7 +2224,7 @@ static void rtl8139_poll_controller(struct net_device *dev)
 	struct rtl8139_private *tp = netdev_priv(dev);
 	const int irq = tp->pci_dev->irq;
 
-	disable_irq(irq);
+	disable_irq_nosync(irq);
 	rtl8139_interrupt(irq, dev);
 	enable_irq(irq);
 }
-- 
cgit v1.1


From 784813aed6ba24a1f24e7e11d9d0f208cee37a7d Mon Sep 17 00:00:00 2001
From: Ursula Braun <ubraun@linux.ibm.com>
Date: Wed, 2 May 2018 16:53:56 +0200
Subject: net/smc: restrict non-blocking connect finish

The smc_poll code tries to finish connect() if the socket is in
state SMC_INIT and polling of the internal CLC-socket returns with
EPOLLOUT. This makes sense for a select/poll call following a connect
call, but not without preceding connect().
With this patch smc_poll starts connect logic only, if the CLC-socket
is no longer in its initial state TCP_CLOSE.

In addition, a poll error on the internal CLC-socket is always
propagated to the SMC socket.

With this patch the code path mentioned by syzbot
https://syzkaller.appspot.com/bug?extid=03faa2dc16b8b64be396
is no longer possible.

Signed-off-by: Ursula Braun <ubraun@linux.ibm.com>
Reported-by: syzbot+03faa2dc16b8b64be396@syzkaller.appspotmail.com
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/smc/af_smc.c | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)

diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c
index 4470501..8b4c059 100644
--- a/net/smc/af_smc.c
+++ b/net/smc/af_smc.c
@@ -1166,13 +1166,15 @@ static __poll_t smc_poll(struct file *file, struct socket *sock,
 		/* delegate to CLC child sock */
 		release_sock(sk);
 		mask = smc->clcsock->ops->poll(file, smc->clcsock, wait);
-		/* if non-blocking connect finished ... */
 		lock_sock(sk);
-		if ((sk->sk_state == SMC_INIT) && (mask & EPOLLOUT)) {
-			sk->sk_err = smc->clcsock->sk->sk_err;
-			if (sk->sk_err) {
-				mask |= EPOLLERR;
-			} else {
+		sk->sk_err = smc->clcsock->sk->sk_err;
+		if (sk->sk_err) {
+			mask |= EPOLLERR;
+		} else {
+			/* if non-blocking connect finished ... */
+			if (sk->sk_state == SMC_INIT &&
+			    mask & EPOLLOUT &&
+			    smc->clcsock->sk->sk_state != TCP_CLOSE) {
 				rc = smc_connect_rdma(smc);
 				if (rc < 0)
 					mask |= EPOLLERR;
-- 
cgit v1.1


From 49ceda9de2da4d1827941d06701f3017c27c1855 Mon Sep 17 00:00:00 2001
From: Sean Paul <seanpaul@chromium.org>
Date: Fri, 20 Apr 2018 14:59:59 -0400
Subject: drm/bridge: vga-dac: Fix edid memory leak

edid should be freed once it's finished being used.

Fixes: 56fe8b6f4991 ("drm/bridge: Add RGB to VGA bridge support")
Cc: Rob Herring <robh@kernel.org>
Cc: Sean Paul <seanpaul@chromium.org>
Cc: Maxime Ripard <maxime.ripard@free-electrons.com>
Cc: Archit Taneja <architt@codeaurora.org>
Cc: Andrzej Hajda <a.hajda@samsung.com>
Cc: Laurent Pinchart <Laurent.pinchart@ideasonboard.com>
Cc: <stable@vger.kernel.org> # v4.9+
Reviewed-by: Maxime Ripard <maxime.ripard@bootlin.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20180420190007.1572-1-seanpaul@chromium.org
---
 drivers/gpu/drm/bridge/dumb-vga-dac.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/bridge/dumb-vga-dac.c b/drivers/gpu/drm/bridge/dumb-vga-dac.c
index 498d594..9837c8d 100644
--- a/drivers/gpu/drm/bridge/dumb-vga-dac.c
+++ b/drivers/gpu/drm/bridge/dumb-vga-dac.c
@@ -56,7 +56,9 @@ static int dumb_vga_get_modes(struct drm_connector *connector)
 	}
 
 	drm_mode_connector_update_edid_property(connector, edid);
-	return drm_add_edid_modes(connector, edid);
+	ret = drm_add_edid_modes(connector, edid);
+	kfree(edid);
+	return ret;
 
 fallback:
 	/*
-- 
cgit v1.1


From 3aab8884c9eb99189a3569ac4e6b205371c9ac0b Mon Sep 17 00:00:00 2001
From: Daniel Borkmann <daniel@iogearbox.net>
Date: Wed, 2 May 2018 20:12:22 +0200
Subject: bpf, x64: fix memleak when not converging after image

While reviewing x64 JIT code, I noticed that we leak the prior allocated
JIT image in the case where proglen != oldproglen during the JIT passes.
Prior to the commit e0ee9c12157d ("x86: bpf_jit: fix two bugs in eBPF JIT
compiler") we would just break out of the loop, and using the image as the
JITed prog since it could only shrink in size anyway. After e0ee9c12157d,
we would bail out to out_addrs label where we free addrs and jit_data but
not the image coming from bpf_jit_binary_alloc().

Fixes: e0ee9c12157d ("x86: bpf_jit: fix two bugs in eBPF JIT compiler")
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
---
 arch/x86/net/bpf_jit_comp.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/x86/net/bpf_jit_comp.c b/arch/x86/net/bpf_jit_comp.c
index abce27c..9ae7b93 100644
--- a/arch/x86/net/bpf_jit_comp.c
+++ b/arch/x86/net/bpf_jit_comp.c
@@ -1236,6 +1236,7 @@ skip_init_addrs:
 	for (pass = 0; pass < 20 || image; pass++) {
 		proglen = do_jit(prog, addrs, image, oldproglen, &ctx);
 		if (proglen <= 0) {
+out_image:
 			image = NULL;
 			if (header)
 				bpf_jit_binary_free(header);
@@ -1246,8 +1247,7 @@ skip_init_addrs:
 			if (proglen != oldproglen) {
 				pr_err("bpf_jit: proglen=%d != oldproglen=%d\n",
 				       proglen, oldproglen);
-				prog = orig_prog;
-				goto out_addrs;
+				goto out_image;
 			}
 			break;
 		}
-- 
cgit v1.1


From 39f56ca945af86112753646316c4c92dcd4acd82 Mon Sep 17 00:00:00 2001
From: Daniel Borkmann <daniel@iogearbox.net>
Date: Wed, 2 May 2018 20:12:23 +0200
Subject: bpf, x64: fix memleak when not converging on calls

The JIT logic in jit_subprogs() is as follows: for all subprogs we
allocate a bpf_prog_alloc(), populate it (prog->is_func = 1 here),
and pass it to bpf_int_jit_compile(). If a failure occurred during
JIT and prog->jited is not set, then we bail out from attempting to
JIT the whole program, and punt to the interpreter instead. In case
JITing went successful, we fixup BPF call offsets and do another
pass to bpf_int_jit_compile() (extra_pass is true at that point) to
complete JITing calls. Given that requires to pass JIT context around
addrs and jit_data from x86 JIT are freed in the extra_pass in
bpf_int_jit_compile() when calls are involved (if not, they can
be freed immediately). However, if in the original pass, the JIT
image didn't converge then we leak addrs and jit_data since image
itself is NULL, the prog->is_func is set and extra_pass is false
in that case, meaning both will become unreachable and are never
cleaned up, therefore we need to free as well on !image. Only x64
JIT is affected.

Fixes: 1c2a088a6626 ("bpf: x64: add JIT support for multi-function programs")
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
---
 arch/x86/net/bpf_jit_comp.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/x86/net/bpf_jit_comp.c b/arch/x86/net/bpf_jit_comp.c
index 9ae7b93..263c845 100644
--- a/arch/x86/net/bpf_jit_comp.c
+++ b/arch/x86/net/bpf_jit_comp.c
@@ -1283,7 +1283,7 @@ out_image:
 		prog = orig_prog;
 	}
 
-	if (!prog->is_func || extra_pass) {
+	if (!image || !prog->is_func || extra_pass) {
 out_addrs:
 		kfree(addrs);
 		kfree(jit_data);
-- 
cgit v1.1


From b819439fea305a0bfd6ca23a7994fd1a8847c0d8 Mon Sep 17 00:00:00 2001
From: Helge Deller <deller@gmx.de>
Date: Fri, 20 Apr 2018 23:19:17 +0200
Subject: parisc: drivers.c: Fix section mismatches

Fix two section mismatches in drivers.c:
1) Section mismatch in reference from the function alloc_tree_node() to
   the function .init.text:create_tree_node().
2) Section mismatch in reference from the function walk_native_bus() to
   the function .init.text:alloc_pa_dev().

Signed-off-by: Helge Deller <deller@gmx.de>
---
 arch/parisc/kernel/drivers.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/arch/parisc/kernel/drivers.c b/arch/parisc/kernel/drivers.c
index 3b8507f..ee5a78a 100644
--- a/arch/parisc/kernel/drivers.c
+++ b/arch/parisc/kernel/drivers.c
@@ -448,7 +448,8 @@ static int match_by_id(struct device * dev, void * data)
  * Checks all the children of @parent for a matching @id.  If none
  * found, it allocates a new device and returns it.
  */
-static struct parisc_device * alloc_tree_node(struct device *parent, char id)
+static struct parisc_device * __init alloc_tree_node(
+			struct device *parent, char id)
 {
 	struct match_id_data d = {
 		.id = id,
@@ -825,8 +826,8 @@ static void walk_lower_bus(struct parisc_device *dev)
  * devices which are not physically connected (such as extra serial &
  * keyboard ports).  This problem is not yet solved.
  */
-static void walk_native_bus(unsigned long io_io_low, unsigned long io_io_high,
-                            struct device *parent)
+static void __init walk_native_bus(unsigned long io_io_low,
+	unsigned long io_io_high, struct device *parent)
 {
 	int i, devices_found = 0;
 	unsigned long hpa = io_io_low;
-- 
cgit v1.1


From 8d73b18079613baf75019a920ce6a1ac0dac8b5b Mon Sep 17 00:00:00 2001
From: Helge Deller <deller@gmx.de>
Date: Fri, 20 Apr 2018 23:23:37 +0200
Subject: parisc: Fix section mismatches

Fix three section mismatches:
1) Section mismatch in reference from the function ioread8() to the
   function .init.text:pcibios_init_bridge()
2) Section mismatch in reference from the function free_initmem() to the
   function .init.text:map_pages()
3) Section mismatch in reference from the function ccio_ioc_init() to
   the function .init.text:count_parisc_driver()

Signed-off-by: Helge Deller <deller@gmx.de>
---
 arch/parisc/kernel/pci.c  | 2 +-
 arch/parisc/mm/init.c     | 2 +-
 drivers/parisc/ccio-dma.c | 2 +-
 3 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/arch/parisc/kernel/pci.c b/arch/parisc/kernel/pci.c
index 13ee356..ae684ac 100644
--- a/arch/parisc/kernel/pci.c
+++ b/arch/parisc/kernel/pci.c
@@ -174,7 +174,7 @@ void pcibios_set_master(struct pci_dev *dev)
  * pcibios_init_bridge() initializes cache line and default latency
  * for pci controllers and pci-pci bridges
  */
-void __init pcibios_init_bridge(struct pci_dev *dev)
+void __ref pcibios_init_bridge(struct pci_dev *dev)
 {
 	unsigned short bridge_ctl, bridge_ctl_new;
 
diff --git a/arch/parisc/mm/init.c b/arch/parisc/mm/init.c
index cab32ee..2607d2d 100644
--- a/arch/parisc/mm/init.c
+++ b/arch/parisc/mm/init.c
@@ -516,7 +516,7 @@ static void __init map_pages(unsigned long start_vaddr,
 	}
 }
 
-void free_initmem(void)
+void __ref free_initmem(void)
 {
 	unsigned long init_begin = (unsigned long)__init_begin;
 	unsigned long init_end = (unsigned long)__init_end;
diff --git a/drivers/parisc/ccio-dma.c b/drivers/parisc/ccio-dma.c
index acba1f5..126cf19 100644
--- a/drivers/parisc/ccio-dma.c
+++ b/drivers/parisc/ccio-dma.c
@@ -1263,7 +1263,7 @@ static struct parisc_driver ccio_driver __refdata = {
  * I/O Page Directory, the resource map, and initalizing the
  * U2/Uturn chip into virtual mode.
  */
-static void
+static void __init
 ccio_ioc_init(struct ioc *ioc)
 {
 	int i;
-- 
cgit v1.1


From 30ca22e4a5d0063dd9a9cdf35cd139c5807cbeb3 Mon Sep 17 00:00:00 2001
From: Ido Schimmel <idosch@mellanox.com>
Date: Wed, 2 May 2018 22:41:56 +0300
Subject: ipv6: Revert "ipv6: Allow non-gateway ECMP for IPv6"

This reverts commit edd7ceb78296 ("ipv6: Allow non-gateway ECMP for
IPv6").

Eric reported a division by zero in rt6_multipath_rebalance() which is
caused by above commit that considers identical local routes to be
siblings. The division by zero happens because a nexthop weight is not
set for local routes.

Revert the commit as it does not fix a bug and has side effects.

To reproduce:

# ip -6 address add 2001:db8::1/64 dev dummy0
# ip -6 address add 2001:db8::1/64 dev dummy1

Fixes: edd7ceb78296 ("ipv6: Allow non-gateway ECMP for IPv6")
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reported-by: Eric Dumazet <eric.dumazet@gmail.com>
Tested-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/ip6_route.h | 3 ++-
 net/ipv6/ip6_fib.c      | 3 +++
 2 files changed, 5 insertions(+), 1 deletion(-)

diff --git a/include/net/ip6_route.h b/include/net/ip6_route.h
index abceb58..08b1323 100644
--- a/include/net/ip6_route.h
+++ b/include/net/ip6_route.h
@@ -68,7 +68,8 @@ static inline bool rt6_need_strict(const struct in6_addr *daddr)
 
 static inline bool rt6_qualify_for_ecmp(const struct rt6_info *rt)
 {
-	return (rt->rt6i_flags & (RTF_ADDRCONF | RTF_DYNAMIC)) == 0;
+	return (rt->rt6i_flags & (RTF_GATEWAY|RTF_ADDRCONF|RTF_DYNAMIC)) ==
+	       RTF_GATEWAY;
 }
 
 void ip6_route_input(struct sk_buff *skb);
diff --git a/net/ipv6/ip6_fib.c b/net/ipv6/ip6_fib.c
index 3c97c29..deab2db 100644
--- a/net/ipv6/ip6_fib.c
+++ b/net/ipv6/ip6_fib.c
@@ -934,6 +934,9 @@ static int fib6_add_rt2node(struct fib6_node *fn, struct rt6_info *rt,
 			 * list.
 			 * Only static routes (which don't have flag
 			 * RTF_EXPIRES) are used for ECMPv6.
+			 *
+			 * To avoid long list, we only had siblings if the
+			 * route have a gateway.
 			 */
 			if (rt_can_ecmp &&
 			    rt6_qualify_for_ecmp(iter))
-- 
cgit v1.1


From 7df40c2673a1307c3260aab6f9d4b9bf97ca8fd7 Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Wed, 2 May 2018 10:03:30 -0700
Subject: net_sched: fq: take care of throttled flows before reuse

Normally, a socket can not be freed/reused unless all its TX packets
left qdisc and were TX-completed. However connect(AF_UNSPEC) allows
this to happen.

With commit fc59d5bdf1e3 ("pkt_sched: fq: clear time_next_packet for
reused flows") we cleared f->time_next_packet but took no special
action if the flow was still in the throttled rb-tree.

Since f->time_next_packet is the key used in the rb-tree searches,
blindly clearing it might break rb-tree integrity. We need to make
sure the flow is no longer in the rb-tree to avoid this problem.

Fixes: fc59d5bdf1e3 ("pkt_sched: fq: clear time_next_packet for reused flows")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/sched/sch_fq.c | 37 +++++++++++++++++++++++++------------
 1 file changed, 25 insertions(+), 12 deletions(-)

diff --git a/net/sched/sch_fq.c b/net/sched/sch_fq.c
index a366e4c..4808713 100644
--- a/net/sched/sch_fq.c
+++ b/net/sched/sch_fq.c
@@ -128,6 +128,28 @@ static bool fq_flow_is_detached(const struct fq_flow *f)
 	return f->next == &detached;
 }
 
+static bool fq_flow_is_throttled(const struct fq_flow *f)
+{
+	return f->next == &throttled;
+}
+
+static void fq_flow_add_tail(struct fq_flow_head *head, struct fq_flow *flow)
+{
+	if (head->first)
+		head->last->next = flow;
+	else
+		head->first = flow;
+	head->last = flow;
+	flow->next = NULL;
+}
+
+static void fq_flow_unset_throttled(struct fq_sched_data *q, struct fq_flow *f)
+{
+	rb_erase(&f->rate_node, &q->delayed);
+	q->throttled_flows--;
+	fq_flow_add_tail(&q->old_flows, f);
+}
+
 static void fq_flow_set_throttled(struct fq_sched_data *q, struct fq_flow *f)
 {
 	struct rb_node **p = &q->delayed.rb_node, *parent = NULL;
@@ -155,15 +177,6 @@ static void fq_flow_set_throttled(struct fq_sched_data *q, struct fq_flow *f)
 
 static struct kmem_cache *fq_flow_cachep __read_mostly;
 
-static void fq_flow_add_tail(struct fq_flow_head *head, struct fq_flow *flow)
-{
-	if (head->first)
-		head->last->next = flow;
-	else
-		head->first = flow;
-	head->last = flow;
-	flow->next = NULL;
-}
 
 /* limit number of collected flows per round */
 #define FQ_GC_MAX 8
@@ -267,6 +280,8 @@ static struct fq_flow *fq_classify(struct sk_buff *skb, struct fq_sched_data *q)
 				     f->socket_hash != sk->sk_hash)) {
 				f->credit = q->initial_quantum;
 				f->socket_hash = sk->sk_hash;
+				if (fq_flow_is_throttled(f))
+					fq_flow_unset_throttled(q, f);
 				f->time_next_packet = 0ULL;
 			}
 			return f;
@@ -438,9 +453,7 @@ static void fq_check_throttled(struct fq_sched_data *q, u64 now)
 			q->time_next_delayed_flow = f->time_next_packet;
 			break;
 		}
-		rb_erase(p, &q->delayed);
-		q->throttled_flows--;
-		fq_flow_add_tail(&q->old_flows, f);
+		fq_flow_unset_throttled(q, f);
 	}
 }
 
-- 
cgit v1.1


From 81cf8b74b0e0728589fdaa37cb3ae42a561bfd5c Mon Sep 17 00:00:00 2001
From: Anusha Srivatsa <anusha.srivatsa@intel.com>
Date: Mon, 30 Apr 2018 15:59:28 -0700
Subject: drm/i915/firmware: Correct URL for firmware

Replace 01.org URL with upstream linux-firmware repo URL.
We no longer release firmware to 01.org.
linux-firmware.git is the ultimate place to find
the i915 firmwares.

Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Anusha Srivatsa <anusha.srivatsa@intel.com>
Acked-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1525129168-529-1-git-send-email-anusha.srivatsa@intel.com
---
 drivers/gpu/drm/i915/intel_uc_fw.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/intel_uc_fw.h b/drivers/gpu/drm/i915/intel_uc_fw.h
index dc33b12..87910aa 100644
--- a/drivers/gpu/drm/i915/intel_uc_fw.h
+++ b/drivers/gpu/drm/i915/intel_uc_fw.h
@@ -30,7 +30,7 @@ struct drm_i915_private;
 struct i915_vma;
 
 /* Home of GuC, HuC and DMC firmwares */
-#define INTEL_UC_FIRMWARE_URL "https://01.org/linuxgraphics/downloads/firmware"
+#define INTEL_UC_FIRMWARE_URL "https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/tree/i915"
 
 enum intel_uc_fw_status {
 	INTEL_UC_FIRMWARE_FAIL = -1,
-- 
cgit v1.1


From 3cc9a472d625f31f981063882b07e96229b9e71a Mon Sep 17 00:00:00 2001
From: John Fastabend <john.fastabend@gmail.com>
Date: Wed, 2 May 2018 13:50:19 -0700
Subject: bpf: sockmap, fix scatterlist update on error path in send with apply

When the call to do_tcp_sendpage() fails to send the complete block
requested we either retry if only a partial send was completed or
abort if we receive a error less than or equal to zero. Before
returning though we must update the scatterlist length/offset to
account for any partial send completed.

Before this patch we did this at the end of the retry loop, but
this was buggy when used while applying a verdict to fewer bytes
than in the scatterlist. When the scatterlist length was being set
we forgot to account for the apply logic reducing the size variable.
So the result was we chopped off some bytes in the scatterlist without
doing proper cleanup on them. This results in a WARNING when the
sock is tore down because the bytes have previously been charged to
the socket but are never uncharged.

The simple fix is to simply do the accounting inside the retry loop
subtracting from the absolute scatterlist values rather than trying
to accumulate the totals and subtract at the end.

Reported-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
---
 kernel/bpf/sockmap.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/kernel/bpf/sockmap.c b/kernel/bpf/sockmap.c
index 634415c..943929a 100644
--- a/kernel/bpf/sockmap.c
+++ b/kernel/bpf/sockmap.c
@@ -326,6 +326,9 @@ retry:
 			if (ret > 0) {
 				if (apply)
 					apply_bytes -= ret;
+
+				sg->offset += ret;
+				sg->length -= ret;
 				size -= ret;
 				offset += ret;
 				if (uncharge)
@@ -333,8 +336,6 @@ retry:
 				goto retry;
 			}
 
-			sg->length = size;
-			sg->offset = offset;
 			return ret;
 		}
 
-- 
cgit v1.1


From fec51d40ea65dd8f51a3e27fc69b4e7dc4f17776 Mon Sep 17 00:00:00 2001
From: John Fastabend <john.fastabend@gmail.com>
Date: Wed, 2 May 2018 13:50:24 -0700
Subject: bpf: sockmap, zero sg_size on error when buffer is released

When an error occurs during a redirect we have two cases that need
to be handled (i) we have a cork'ed buffer (ii) we have a normal
sendmsg buffer.

In the cork'ed buffer case we don't currently support recovering from
errors in a redirect action. So the buffer is released and the error
should _not_ be pushed back to the caller of sendmsg/sendpage. The
rationale here is the user will get an error that relates to old
data that may have been sent by some arbitrary thread on that sock.
Instead we simple consume the data and tell the user that the data
has been consumed. We may add proper error recovery in the future.
However, this patch fixes a bug where the bytes outstanding counter
sg_size was not zeroed. This could result in a case where if the user
has both a cork'ed action and apply action in progress we may
incorrectly call into the BPF program when the user expected an
old verdict to be applied via the apply action. I don't have a use
case where using apply and cork at the same time is valid but we
never explicitly reject it because it should work fine. This patch
ensures the sg_size is zeroed so we don't have this case.

In the normal sendmsg buffer case (no cork data) we also do not
zero sg_size. Again this can confuse the apply logic when the logic
calls into the BPF program when the BPF programmer expected the old
verdict to remain. So ensure we set sg_size to zero here as well. And
additionally to keep the psock state in-sync with the sk_msg_buff
release all the memory as well. Previously we did this before
returning to the user but this left a gap where psock and sk_msg_buff
states were out of sync which seems fragile. No additional overhead
is taken here except for a call to check the length and realize its
already been freed. This is in the error path as well so in my
opinion lets have robust code over optimized error paths.

Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
---
 kernel/bpf/sockmap.c | 15 +++++++++++----
 1 file changed, 11 insertions(+), 4 deletions(-)

diff --git a/kernel/bpf/sockmap.c b/kernel/bpf/sockmap.c
index 943929a..052c313 100644
--- a/kernel/bpf/sockmap.c
+++ b/kernel/bpf/sockmap.c
@@ -701,15 +701,22 @@ more_data:
 		err = bpf_tcp_sendmsg_do_redirect(redir, send, m, flags);
 		lock_sock(sk);
 
+		if (unlikely(err < 0)) {
+			free_start_sg(sk, m);
+			psock->sg_size = 0;
+			if (!cork)
+				*copied -= send;
+		} else {
+			psock->sg_size -= send;
+		}
+
 		if (cork) {
 			free_start_sg(sk, m);
+			psock->sg_size = 0;
 			kfree(m);
 			m = NULL;
+			err = 0;
 		}
-		if (unlikely(err))
-			*copied -= err;
-		else
-			psock->sg_size -= send;
 		break;
 	case __SK_DROP:
 	default:
-- 
cgit v1.1


From abaeb096ca38cad02c8a68c49ddd7efc043c319a Mon Sep 17 00:00:00 2001
From: John Fastabend <john.fastabend@gmail.com>
Date: Wed, 2 May 2018 13:50:29 -0700
Subject: bpf: sockmap, fix error handling in redirect failures

When a redirect failure happens we release the buffers in-flight
without calling a sk_mem_uncharge(), the uncharge is called before
dropping the sock lock for the redirecte, however we missed updating
the ring start index. When no apply actions are in progress this
is OK because we uncharge the entire buffer before the redirect.
But, when we have apply logic running its possible that only a
portion of the buffer is being redirected. In this case we only
do memory accounting for the buffer slice being redirected and
expect to be able to loop over the BPF program again and/or if
a sock is closed uncharge the memory at sock destruct time.

With an invalid start index however the program logic looks at
the start pointer index, checks the length, and when seeing the
length is zero (from the initial release and failure to update
the pointer) aborts without uncharging/releasing the remaining
memory.

The fix for this is simply to update the start index. To avoid
fixing this error in two locations we do a small refactor and
remove one case where it is open-coded. Then fix it in the
single function.

Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
---
 kernel/bpf/sockmap.c | 28 ++++++++++++----------------
 1 file changed, 12 insertions(+), 16 deletions(-)

diff --git a/kernel/bpf/sockmap.c b/kernel/bpf/sockmap.c
index 052c313..098eca5 100644
--- a/kernel/bpf/sockmap.c
+++ b/kernel/bpf/sockmap.c
@@ -393,7 +393,8 @@ static void return_mem_sg(struct sock *sk, int bytes, struct sk_msg_buff *md)
 	} while (i != md->sg_end);
 }
 
-static void free_bytes_sg(struct sock *sk, int bytes, struct sk_msg_buff *md)
+static void free_bytes_sg(struct sock *sk, int bytes,
+			  struct sk_msg_buff *md, bool charge)
 {
 	struct scatterlist *sg = md->sg_data;
 	int i = md->sg_start, free;
@@ -403,11 +404,13 @@ static void free_bytes_sg(struct sock *sk, int bytes, struct sk_msg_buff *md)
 		if (bytes < free) {
 			sg[i].length -= bytes;
 			sg[i].offset += bytes;
-			sk_mem_uncharge(sk, bytes);
+			if (charge)
+				sk_mem_uncharge(sk, bytes);
 			break;
 		}
 
-		sk_mem_uncharge(sk, sg[i].length);
+		if (charge)
+			sk_mem_uncharge(sk, sg[i].length);
 		put_page(sg_page(&sg[i]));
 		bytes -= sg[i].length;
 		sg[i].length = 0;
@@ -418,6 +421,7 @@ static void free_bytes_sg(struct sock *sk, int bytes, struct sk_msg_buff *md)
 		if (i == MAX_SKB_FRAGS)
 			i = 0;
 	}
+	md->sg_start = i;
 }
 
 static int free_sg(struct sock *sk, int start, struct sk_msg_buff *md)
@@ -576,10 +580,10 @@ static int bpf_tcp_sendmsg_do_redirect(struct sock *sk, int send,
 				       struct sk_msg_buff *md,
 				       int flags)
 {
+	bool ingress = !!(md->flags & BPF_F_INGRESS);
 	struct smap_psock *psock;
 	struct scatterlist *sg;
-	int i, err, free = 0;
-	bool ingress = !!(md->flags & BPF_F_INGRESS);
+	int err = 0;
 
 	sg = md->sg_data;
 
@@ -607,16 +611,8 @@ static int bpf_tcp_sendmsg_do_redirect(struct sock *sk, int send,
 out_rcu:
 	rcu_read_unlock();
 out:
-	i = md->sg_start;
-	while (sg[i].length) {
-		free += sg[i].length;
-		put_page(sg_page(&sg[i]));
-		sg[i].length = 0;
-		i++;
-		if (i == MAX_SKB_FRAGS)
-			i = 0;
-	}
-	return free;
+	free_bytes_sg(NULL, send, md, false);
+	return err;
 }
 
 static inline void bpf_md_init(struct smap_psock *psock)
@@ -720,7 +716,7 @@ more_data:
 		break;
 	case __SK_DROP:
 	default:
-		free_bytes_sg(sk, send, m);
+		free_bytes_sg(sk, send, m, true);
 		apply_bytes_dec(psock, send);
 		*copied -= send;
 		psock->sg_size -= send;
-- 
cgit v1.1


From 65fcb8064dd0e54d4674e8e2c6bf6ed7264a29e9 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Wed, 2 May 2018 17:38:38 +0100
Subject: drm/i915: Move timeline from GTT to ring

In the future, we want to move a request between engines. To achieve
this, we first realise that we have two timelines in effect here. The
first runs through the GTT is required for ordering vma access, which is
tracked currently by engine. The second is implied by sequential
execution of commands inside the ringbuffer. This timeline is one that
maps to userspace's expectations when submitting requests (i.e. given the
same context, batch A is executed before batch B). As the rings's
timelines map to userspace and the GTT timeline an implementation
detail, move the timeline from the GTT into the ring itself (per-context
in logical-ring-contexts/execlists, or a global per-engine timeline for
the shared ringbuffers in legacy submission.

The two timelines are still assumed to be equivalent at the moment (no
migrating requests between engines yet) and so we can simply move from
one to the other without adding extra ordering.

v2: Reinforce that one isn't allowed to mix the engine execution
timeline with the client timeline from userspace (on the ring).

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180502163839.3248-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/i915_drv.h                   | 13 +-----
 drivers/gpu/drm/i915/i915_gem.c                   |  9 ++--
 drivers/gpu/drm/i915/i915_gem_context.c           | 15 ++++++-
 drivers/gpu/drm/i915/i915_gem_context.h           |  2 +
 drivers/gpu/drm/i915/i915_gem_gtt.c               |  3 --
 drivers/gpu/drm/i915/i915_gem_gtt.h               |  1 -
 drivers/gpu/drm/i915/i915_gem_timeline.c          | 54 ++++++++++++++++++++---
 drivers/gpu/drm/i915/i915_gem_timeline.h          |  4 ++
 drivers/gpu/drm/i915/i915_request.c               | 13 +++---
 drivers/gpu/drm/i915/intel_engine_cs.c            |  3 +-
 drivers/gpu/drm/i915/intel_lrc.c                  |  2 +-
 drivers/gpu/drm/i915/intel_ringbuffer.c           | 10 ++++-
 drivers/gpu/drm/i915/intel_ringbuffer.h           |  5 ++-
 drivers/gpu/drm/i915/selftests/i915_gem_context.c | 12 +++++
 drivers/gpu/drm/i915/selftests/mock_engine.c      |  5 ++-
 drivers/gpu/drm/i915/selftests/mock_gem_device.c  |  4 +-
 drivers/gpu/drm/i915/selftests/mock_gtt.c         |  1 -
 17 files changed, 115 insertions(+), 41 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 6268a51..ffa87ae 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -2059,7 +2059,8 @@ struct drm_i915_private {
 		void (*resume)(struct drm_i915_private *);
 		void (*cleanup_engine)(struct intel_engine_cs *engine);
 
-		struct i915_gem_timeline global_timeline;
+		struct i915_gem_timeline execution_timeline;
+		struct i915_gem_timeline legacy_timeline;
 		struct list_head timelines;
 
 		struct list_head active_rings;
@@ -3235,16 +3236,6 @@ i915_gem_context_lookup(struct drm_i915_file_private *file_priv, u32 id)
 	return ctx;
 }
 
-static inline struct intel_timeline *
-i915_gem_context_lookup_timeline(struct i915_gem_context *ctx,
-				 struct intel_engine_cs *engine)
-{
-	struct i915_address_space *vm;
-
-	vm = ctx->ppgtt ? &ctx->ppgtt->base : &ctx->i915->ggtt.base;
-	return &vm->timeline.engine[engine->id];
-}
-
 int i915_perf_open_ioctl(struct drm_device *dev, void *data,
 			 struct drm_file *file);
 int i915_perf_add_config_ioctl(struct drm_device *dev, void *data,
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index fa1d94a..438a2fc 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -3110,10 +3110,10 @@ static void engine_skip_context(struct i915_request *request)
 {
 	struct intel_engine_cs *engine = request->engine;
 	struct i915_gem_context *hung_ctx = request->ctx;
-	struct intel_timeline *timeline;
+	struct intel_timeline *timeline = request->timeline;
 	unsigned long flags;
 
-	timeline = i915_gem_context_lookup_timeline(hung_ctx, engine);
+	GEM_BUG_ON(timeline == engine->timeline);
 
 	spin_lock_irqsave(&engine->timeline->lock, flags);
 	spin_lock(&timeline->lock);
@@ -3782,7 +3782,7 @@ int i915_gem_wait_for_idle(struct drm_i915_private *i915, unsigned int flags)
 
 		ret = wait_for_engines(i915);
 	} else {
-		ret = wait_for_timeline(&i915->gt.global_timeline, flags);
+		ret = wait_for_timeline(&i915->gt.execution_timeline, flags);
 	}
 
 	return ret;
@@ -5652,7 +5652,8 @@ void i915_gem_cleanup_early(struct drm_i915_private *dev_priv)
 	WARN_ON(dev_priv->mm.object_count);
 
 	mutex_lock(&dev_priv->drm.struct_mutex);
-	i915_gem_timeline_fini(&dev_priv->gt.global_timeline);
+	i915_gem_timeline_fini(&dev_priv->gt.legacy_timeline);
+	i915_gem_timeline_fini(&dev_priv->gt.execution_timeline);
 	WARN_ON(!list_empty(&dev_priv->gt.timelines));
 	mutex_unlock(&dev_priv->drm.struct_mutex);
 
diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c
index 59d4bd4..1f4987d 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/i915_gem_context.c
@@ -122,6 +122,7 @@ static void i915_gem_context_free(struct i915_gem_context *ctx)
 	lockdep_assert_held(&ctx->i915->drm.struct_mutex);
 	GEM_BUG_ON(!i915_gem_context_is_closed(ctx));
 
+	i915_gem_timeline_free(ctx->timeline);
 	i915_ppgtt_put(ctx->ppgtt);
 
 	for (n = 0; n < ARRAY_SIZE(ctx->__engine); n++) {
@@ -376,6 +377,18 @@ i915_gem_create_context(struct drm_i915_private *dev_priv,
 		ctx->desc_template = default_desc_template(dev_priv, ppgtt);
 	}
 
+	if (HAS_EXECLISTS(dev_priv)) {
+		struct i915_gem_timeline *timeline;
+
+		timeline = i915_gem_timeline_create(dev_priv, ctx->name);
+		if (IS_ERR(timeline)) {
+			__destroy_hw_context(ctx, file_priv);
+			return ERR_CAST(timeline);
+		}
+
+		ctx->timeline = timeline;
+	}
+
 	trace_i915_context_create(ctx);
 
 	return ctx;
@@ -584,7 +597,7 @@ static bool engine_has_idle_kernel_context(struct intel_engine_cs *engine)
 	list_for_each_entry(timeline, &engine->i915->gt.timelines, link) {
 		struct intel_timeline *tl;
 
-		if (timeline == &engine->i915->gt.global_timeline)
+		if (timeline == &engine->i915->gt.execution_timeline)
 			continue;
 
 		tl = &timeline->engine[engine->id];
diff --git a/drivers/gpu/drm/i915/i915_gem_context.h b/drivers/gpu/drm/i915/i915_gem_context.h
index ace3b12..ec53ba0 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.h
+++ b/drivers/gpu/drm/i915/i915_gem_context.h
@@ -58,6 +58,8 @@ struct i915_gem_context {
 	/** file_priv: owning file descriptor */
 	struct drm_i915_file_private *file_priv;
 
+	struct i915_gem_timeline *timeline;
+
 	/**
 	 * @ppgtt: unique address space (GTT)
 	 *
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index 21d72f6..e9d8283 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -2111,8 +2111,6 @@ static void i915_address_space_init(struct i915_address_space *vm,
 				    struct drm_i915_private *dev_priv,
 				    const char *name)
 {
-	i915_gem_timeline_init(dev_priv, &vm->timeline, name);
-
 	drm_mm_init(&vm->mm, 0, vm->total);
 	vm->mm.head_node.color = I915_COLOR_UNEVICTABLE;
 
@@ -2129,7 +2127,6 @@ static void i915_address_space_fini(struct i915_address_space *vm)
 	if (pagevec_count(&vm->free_pages))
 		vm_free_pages_release(vm, true);
 
-	i915_gem_timeline_fini(&vm->timeline);
 	drm_mm_takedown(&vm->mm);
 	list_del(&vm->global_link);
 }
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.h b/drivers/gpu/drm/i915/i915_gem_gtt.h
index 6efc017..9810792 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.h
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.h
@@ -257,7 +257,6 @@ struct i915_pml4 {
 
 struct i915_address_space {
 	struct drm_mm mm;
-	struct i915_gem_timeline timeline;
 	struct drm_i915_private *i915;
 	struct device *dma;
 	/* Every address space belongs to a struct file - except for the global
diff --git a/drivers/gpu/drm/i915/i915_gem_timeline.c b/drivers/gpu/drm/i915/i915_gem_timeline.c
index e9fd876..24f4068 100644
--- a/drivers/gpu/drm/i915/i915_gem_timeline.c
+++ b/drivers/gpu/drm/i915/i915_gem_timeline.c
@@ -95,12 +95,28 @@ int i915_gem_timeline_init(struct drm_i915_private *i915,
 
 int i915_gem_timeline_init__global(struct drm_i915_private *i915)
 {
-	static struct lock_class_key class;
+	static struct lock_class_key class1, class2;
+	int err;
+
+	err = __i915_gem_timeline_init(i915,
+				       &i915->gt.execution_timeline,
+				       "[execution]", &class1,
+				       "i915_execution_timeline");
+	if (err)
+		return err;
+
+	err = __i915_gem_timeline_init(i915,
+				       &i915->gt.legacy_timeline,
+				       "[global]", &class2,
+				       "i915_global_timeline");
+	if (err)
+		goto err_exec_timeline;
+
+	return 0;
 
-	return __i915_gem_timeline_init(i915,
-					&i915->gt.global_timeline,
-					"[execution]",
-					&class, "&global_timeline->lock");
+err_exec_timeline:
+	i915_gem_timeline_fini(&i915->gt.execution_timeline);
+	return err;
 }
 
 /**
@@ -148,6 +164,34 @@ void i915_gem_timeline_fini(struct i915_gem_timeline *timeline)
 	kfree(timeline->name);
 }
 
+struct i915_gem_timeline *
+i915_gem_timeline_create(struct drm_i915_private *i915, const char *name)
+{
+	struct i915_gem_timeline *timeline;
+	int err;
+
+	timeline = kzalloc(sizeof(*timeline), GFP_KERNEL);
+	if (!timeline)
+		return ERR_PTR(-ENOMEM);
+
+	err = i915_gem_timeline_init(i915, timeline, name);
+	if (err) {
+		kfree(timeline);
+		return ERR_PTR(err);
+	}
+
+	return timeline;
+}
+
+void i915_gem_timeline_free(struct i915_gem_timeline *timeline)
+{
+	if (!timeline)
+		return;
+
+	i915_gem_timeline_fini(timeline);
+	kfree(timeline);
+}
+
 #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
 #include "selftests/mock_timeline.c"
 #include "selftests/i915_gem_timeline.c"
diff --git a/drivers/gpu/drm/i915/i915_gem_timeline.h b/drivers/gpu/drm/i915/i915_gem_timeline.h
index 6e82119..780ed46 100644
--- a/drivers/gpu/drm/i915/i915_gem_timeline.h
+++ b/drivers/gpu/drm/i915/i915_gem_timeline.h
@@ -90,6 +90,10 @@ int i915_gem_timeline_init__global(struct drm_i915_private *i915);
 void i915_gem_timelines_park(struct drm_i915_private *i915);
 void i915_gem_timeline_fini(struct i915_gem_timeline *tl);
 
+struct i915_gem_timeline *
+i915_gem_timeline_create(struct drm_i915_private *i915, const char *name);
+void i915_gem_timeline_free(struct i915_gem_timeline *timeline);
+
 static inline int __intel_timeline_sync_set(struct intel_timeline *tl,
 					    u64 context, u32 seqno)
 {
diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c
index c8fc4b3..7bb613c 100644
--- a/drivers/gpu/drm/i915/i915_request.c
+++ b/drivers/gpu/drm/i915/i915_request.c
@@ -758,7 +758,12 @@ i915_request_alloc(struct intel_engine_cs *engine, struct i915_gem_context *ctx)
 		}
 	}
 
-	rq->timeline = i915_gem_context_lookup_timeline(ctx, engine);
+	INIT_LIST_HEAD(&rq->active_list);
+	rq->i915 = i915;
+	rq->engine = engine;
+	rq->ctx = ctx;
+	rq->ring = ring;
+	rq->timeline = ring->timeline;
 	GEM_BUG_ON(rq->timeline == engine->timeline);
 
 	spin_lock_init(&rq->lock);
@@ -774,12 +779,6 @@ i915_request_alloc(struct intel_engine_cs *engine, struct i915_gem_context *ctx)
 
 	i915_sched_node_init(&rq->sched);
 
-	INIT_LIST_HEAD(&rq->active_list);
-	rq->i915 = i915;
-	rq->engine = engine;
-	rq->ctx = ctx;
-	rq->ring = ring;
-
 	/* No zalloc, must clear what we need by hand */
 	rq->global_seqno = 0;
 	rq->signaling.wait.seqno = 0;
diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
index 9164e6d..7af5fe8 100644
--- a/drivers/gpu/drm/i915/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/intel_engine_cs.c
@@ -453,7 +453,8 @@ void intel_engine_init_global_seqno(struct intel_engine_cs *engine, u32 seqno)
 
 static void intel_engine_init_timeline(struct intel_engine_cs *engine)
 {
-	engine->timeline = &engine->i915->gt.global_timeline.engine[engine->id];
+	engine->timeline =
+		&engine->i915->gt.execution_timeline.engine[engine->id];
 }
 
 static void intel_engine_init_batch_pool(struct intel_engine_cs *engine)
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 57396a2..9b24077 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -2624,7 +2624,7 @@ static int execlists_context_deferred_alloc(struct i915_gem_context *ctx,
 		goto error_deref_obj;
 	}
 
-	ring = intel_engine_create_ring(engine, ctx->ring_size);
+	ring = intel_engine_create_ring(engine, ctx->timeline, ctx->ring_size);
 	if (IS_ERR(ring)) {
 		ret = PTR_ERR(ring);
 		goto error_deref_obj;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 007449c..b73e700 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -1117,13 +1117,16 @@ err:
 }
 
 struct intel_ring *
-intel_engine_create_ring(struct intel_engine_cs *engine, int size)
+intel_engine_create_ring(struct intel_engine_cs *engine,
+			 struct i915_gem_timeline *timeline,
+			 int size)
 {
 	struct intel_ring *ring;
 	struct i915_vma *vma;
 
 	GEM_BUG_ON(!is_power_of_2(size));
 	GEM_BUG_ON(RING_CTL_SIZE(size) & ~RING_NR_PAGES);
+	GEM_BUG_ON(&timeline->engine[engine->id] == engine->timeline);
 	lockdep_assert_held(&engine->i915->drm.struct_mutex);
 
 	ring = kzalloc(sizeof(*ring), GFP_KERNEL);
@@ -1131,6 +1134,7 @@ intel_engine_create_ring(struct intel_engine_cs *engine, int size)
 		return ERR_PTR(-ENOMEM);
 
 	INIT_LIST_HEAD(&ring->request_list);
+	ring->timeline = &timeline->engine[engine->id];
 
 	ring->size = size;
 	/* Workaround an erratum on the i830 which causes a hang if
@@ -1327,7 +1331,9 @@ static int intel_init_ring_buffer(struct intel_engine_cs *engine)
 	if (err)
 		goto err;
 
-	ring = intel_engine_create_ring(engine, 32 * PAGE_SIZE);
+	ring = intel_engine_create_ring(engine,
+					&engine->i915->gt.legacy_timeline,
+					32 * PAGE_SIZE);
 	if (IS_ERR(ring)) {
 		err = PTR_ERR(ring);
 		goto err;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index fd679ce..da53aa2 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -129,6 +129,7 @@ struct intel_ring {
 	struct i915_vma *vma;
 	void *vaddr;
 
+	struct intel_timeline *timeline;
 	struct list_head request_list;
 	struct list_head active_link;
 
@@ -768,7 +769,9 @@ intel_write_status_page(struct intel_engine_cs *engine, int reg, u32 value)
 #define CNL_HWS_CSB_WRITE_INDEX		0x2f
 
 struct intel_ring *
-intel_engine_create_ring(struct intel_engine_cs *engine, int size);
+intel_engine_create_ring(struct intel_engine_cs *engine,
+			 struct i915_gem_timeline *timeline,
+			 int size);
 int intel_ring_pin(struct intel_ring *ring,
 		   struct drm_i915_private *i915,
 		   unsigned int offset_bias);
diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_context.c b/drivers/gpu/drm/i915/selftests/i915_gem_context.c
index 7ecaed5..24ac648 100644
--- a/drivers/gpu/drm/i915/selftests/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/selftests/i915_gem_context.c
@@ -355,6 +355,18 @@ static int igt_ctx_exec(void *arg)
 
 		if (first_shared_gtt) {
 			ctx = __create_hw_context(i915, file->driver_priv);
+			if (!IS_ERR(ctx) && HAS_EXECLISTS(i915)) {
+				struct i915_gem_timeline *timeline;
+
+				timeline = i915_gem_timeline_create(i915, ctx->name);
+				if (IS_ERR(timeline)) {
+					__destroy_hw_context(ctx, file->driver_priv);
+					ctx = ERR_CAST(timeline);
+				} else {
+					ctx->timeline = timeline;
+				}
+			}
+
 			first_shared_gtt = false;
 		} else {
 			ctx = i915_gem_create_context(i915, file->driver_priv);
diff --git a/drivers/gpu/drm/i915/selftests/mock_engine.c b/drivers/gpu/drm/i915/selftests/mock_engine.c
index 19175dd..6752498 100644
--- a/drivers/gpu/drm/i915/selftests/mock_engine.c
+++ b/drivers/gpu/drm/i915/selftests/mock_engine.c
@@ -140,6 +140,8 @@ static struct intel_ring *mock_ring(struct intel_engine_cs *engine)
 	if (!ring)
 		return NULL;
 
+	ring->timeline = &engine->i915->gt.legacy_timeline.engine[engine->id];
+
 	ring->size = sz;
 	ring->effective_size = sz;
 	ring->vaddr = (void *)(ring + 1);
@@ -180,8 +182,7 @@ struct intel_engine_cs *mock_engine(struct drm_i915_private *i915,
 	engine->base.emit_breadcrumb = mock_emit_breadcrumb;
 	engine->base.submit_request = mock_submit_request;
 
-	engine->base.timeline =
-		&i915->gt.global_timeline.engine[engine->base.id];
+	intel_engine_init_timeline(&engine->base);
 
 	intel_engine_init_breadcrumbs(&engine->base);
 	engine->base.breadcrumbs.mock = true; /* prevent touching HW for irqs */
diff --git a/drivers/gpu/drm/i915/selftests/mock_gem_device.c b/drivers/gpu/drm/i915/selftests/mock_gem_device.c
index f22a2b3..f11c83e 100644
--- a/drivers/gpu/drm/i915/selftests/mock_gem_device.c
+++ b/drivers/gpu/drm/i915/selftests/mock_gem_device.c
@@ -73,7 +73,9 @@ static void mock_device_release(struct drm_device *dev)
 
 	mutex_lock(&i915->drm.struct_mutex);
 	mock_fini_ggtt(i915);
-	i915_gem_timeline_fini(&i915->gt.global_timeline);
+	i915_gem_timeline_fini(&i915->gt.legacy_timeline);
+	i915_gem_timeline_fini(&i915->gt.execution_timeline);
+	WARN_ON(!list_empty(&i915->gt.timelines));
 	mutex_unlock(&i915->drm.struct_mutex);
 
 	destroy_workqueue(i915->wq);
diff --git a/drivers/gpu/drm/i915/selftests/mock_gtt.c b/drivers/gpu/drm/i915/selftests/mock_gtt.c
index e96873f..36c1120 100644
--- a/drivers/gpu/drm/i915/selftests/mock_gtt.c
+++ b/drivers/gpu/drm/i915/selftests/mock_gtt.c
@@ -76,7 +76,6 @@ mock_ppgtt(struct drm_i915_private *i915,
 
 	INIT_LIST_HEAD(&ppgtt->base.global_link);
 	drm_mm_init(&ppgtt->base.mm, 0, ppgtt->base.total);
-	i915_gem_timeline_init(i915, &ppgtt->base.timeline, name);
 
 	ppgtt->base.clear_range = nop_clear_range;
 	ppgtt->base.insert_page = mock_insert_page;
-- 
cgit v1.1


From a89d1f921c15932b4c9a70861d134290f1a14a10 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Wed, 2 May 2018 17:38:39 +0100
Subject: drm/i915: Split i915_gem_timeline into individual timelines

We need to move to a more flexible timeline that doesn't assume one
fence context per engine, and so allow for a single timeline to be used
across a combination of engines. This means that preallocating a fence
context per engine is now a hindrance, and so we want to introduce the
singular timeline. From the code perspective, this has the notable
advantage of clearing up a lot of mirky semantics and some clumsy
pointer chasing.

By splitting the timeline up into a single entity rather than an array
of per-engine timelines, we can realise the goal of the previous patch
of tracking the timeline alongside the ring.

v2: Tweak wait_for_idle to stop the compiling thinking that ret may be
uninitialised.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180502163839.3248-2-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/Makefile                      |   2 +-
 drivers/gpu/drm/i915/i915_drv.h                    |   4 +-
 drivers/gpu/drm/i915/i915_gem.c                    | 129 ++++-----
 drivers/gpu/drm/i915/i915_gem_context.c            |  48 ++--
 drivers/gpu/drm/i915/i915_gem_context.h            |   2 -
 drivers/gpu/drm/i915/i915_gem_gtt.h                |   3 +-
 drivers/gpu/drm/i915/i915_gem_timeline.c           | 198 --------------
 drivers/gpu/drm/i915/i915_gem_timeline.h           | 121 ---------
 drivers/gpu/drm/i915/i915_gpu_error.c              |   4 +-
 drivers/gpu/drm/i915/i915_perf.c                   |  10 +-
 drivers/gpu/drm/i915/i915_request.c                |  68 +++--
 drivers/gpu/drm/i915/i915_request.h                |   3 +-
 drivers/gpu/drm/i915/i915_timeline.c               | 105 ++++++++
 drivers/gpu/drm/i915/i915_timeline.h               | 126 +++++++++
 drivers/gpu/drm/i915/intel_engine_cs.c             |  27 +-
 drivers/gpu/drm/i915/intel_guc_submission.c        |   4 +-
 drivers/gpu/drm/i915/intel_lrc.c                   |  48 ++--
 drivers/gpu/drm/i915/intel_ringbuffer.c            |  25 +-
 drivers/gpu/drm/i915/intel_ringbuffer.h            |  11 +-
 drivers/gpu/drm/i915/selftests/i915_gem_context.c  |  12 -
 drivers/gpu/drm/i915/selftests/i915_gem_timeline.c | 299 ---------------------
 drivers/gpu/drm/i915/selftests/i915_timeline.c     | 267 ++++++++++++++++++
 drivers/gpu/drm/i915/selftests/mock_engine.c       |  32 ++-
 drivers/gpu/drm/i915/selftests/mock_gem_device.c   |  10 +-
 drivers/gpu/drm/i915/selftests/mock_timeline.c     |  45 +---
 drivers/gpu/drm/i915/selftests/mock_timeline.h     |  28 +-
 26 files changed, 723 insertions(+), 908 deletions(-)
 delete mode 100644 drivers/gpu/drm/i915/i915_gem_timeline.c
 delete mode 100644 drivers/gpu/drm/i915/i915_gem_timeline.h
 create mode 100644 drivers/gpu/drm/i915/i915_timeline.c
 create mode 100644 drivers/gpu/drm/i915/i915_timeline.h
 delete mode 100644 drivers/gpu/drm/i915/selftests/i915_gem_timeline.c
 create mode 100644 drivers/gpu/drm/i915/selftests/i915_timeline.c

diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile
index dfe0145..00c1338 100644
--- a/drivers/gpu/drm/i915/Makefile
+++ b/drivers/gpu/drm/i915/Makefile
@@ -71,11 +71,11 @@ i915-y += i915_cmd_parser.o \
 	  i915_gem_shrinker.o \
 	  i915_gem_stolen.o \
 	  i915_gem_tiling.o \
-	  i915_gem_timeline.o \
 	  i915_gem_userptr.o \
 	  i915_gemfs.o \
 	  i915_query.o \
 	  i915_request.o \
+	  i915_timeline.o \
 	  i915_trace_points.o \
 	  i915_vma.o \
 	  intel_breadcrumbs.o \
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index ffa87ae..11ff84e 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -72,10 +72,10 @@
 #include "i915_gem_fence_reg.h"
 #include "i915_gem_object.h"
 #include "i915_gem_gtt.h"
-#include "i915_gem_timeline.h"
 #include "i915_gpu_error.h"
 #include "i915_request.h"
 #include "i915_scheduler.h"
+#include "i915_timeline.h"
 #include "i915_vma.h"
 
 #include "intel_gvt.h"
@@ -2059,8 +2059,6 @@ struct drm_i915_private {
 		void (*resume)(struct drm_i915_private *);
 		void (*cleanup_engine)(struct intel_engine_cs *engine);
 
-		struct i915_gem_timeline execution_timeline;
-		struct i915_gem_timeline legacy_timeline;
 		struct list_head timelines;
 
 		struct list_head active_rings;
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 438a2fc..484354f 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -162,7 +162,7 @@ static u32 __i915_gem_park(struct drm_i915_private *i915)
 	synchronize_irq(i915->drm.irq);
 
 	intel_engines_park(i915);
-	i915_gem_timelines_park(i915);
+	i915_timelines_park(i915);
 
 	i915_pmu_gt_parked(i915);
 
@@ -2977,8 +2977,8 @@ i915_gem_find_active_request(struct intel_engine_cs *engine)
 	 * extra delay for a recent interrupt is pointless. Hence, we do
 	 * not need an engine->irq_seqno_barrier() before the seqno reads.
 	 */
-	spin_lock_irqsave(&engine->timeline->lock, flags);
-	list_for_each_entry(request, &engine->timeline->requests, link) {
+	spin_lock_irqsave(&engine->timeline.lock, flags);
+	list_for_each_entry(request, &engine->timeline.requests, link) {
 		if (__i915_request_completed(request, request->global_seqno))
 			continue;
 
@@ -2989,7 +2989,7 @@ i915_gem_find_active_request(struct intel_engine_cs *engine)
 		active = request;
 		break;
 	}
-	spin_unlock_irqrestore(&engine->timeline->lock, flags);
+	spin_unlock_irqrestore(&engine->timeline.lock, flags);
 
 	return active;
 }
@@ -3110,15 +3110,15 @@ static void engine_skip_context(struct i915_request *request)
 {
 	struct intel_engine_cs *engine = request->engine;
 	struct i915_gem_context *hung_ctx = request->ctx;
-	struct intel_timeline *timeline = request->timeline;
+	struct i915_timeline *timeline = request->timeline;
 	unsigned long flags;
 
-	GEM_BUG_ON(timeline == engine->timeline);
+	GEM_BUG_ON(timeline == &engine->timeline);
 
-	spin_lock_irqsave(&engine->timeline->lock, flags);
+	spin_lock_irqsave(&engine->timeline.lock, flags);
 	spin_lock(&timeline->lock);
 
-	list_for_each_entry_continue(request, &engine->timeline->requests, link)
+	list_for_each_entry_continue(request, &engine->timeline.requests, link)
 		if (request->ctx == hung_ctx)
 			skip_request(request);
 
@@ -3126,7 +3126,7 @@ static void engine_skip_context(struct i915_request *request)
 		skip_request(request);
 
 	spin_unlock(&timeline->lock);
-	spin_unlock_irqrestore(&engine->timeline->lock, flags);
+	spin_unlock_irqrestore(&engine->timeline.lock, flags);
 }
 
 /* Returns the request if it was guilty of the hang */
@@ -3183,11 +3183,11 @@ i915_gem_reset_request(struct intel_engine_cs *engine,
 			dma_fence_set_error(&request->fence, -EAGAIN);
 
 			/* Rewind the engine to replay the incomplete rq */
-			spin_lock_irq(&engine->timeline->lock);
+			spin_lock_irq(&engine->timeline.lock);
 			request = list_prev_entry(request, link);
-			if (&request->link == &engine->timeline->requests)
+			if (&request->link == &engine->timeline.requests)
 				request = NULL;
-			spin_unlock_irq(&engine->timeline->lock);
+			spin_unlock_irq(&engine->timeline.lock);
 		}
 	}
 
@@ -3300,10 +3300,10 @@ static void nop_complete_submit_request(struct i915_request *request)
 		  request->fence.context, request->fence.seqno);
 	dma_fence_set_error(&request->fence, -EIO);
 
-	spin_lock_irqsave(&request->engine->timeline->lock, flags);
+	spin_lock_irqsave(&request->engine->timeline.lock, flags);
 	__i915_request_submit(request);
 	intel_engine_init_global_seqno(request->engine, request->global_seqno);
-	spin_unlock_irqrestore(&request->engine->timeline->lock, flags);
+	spin_unlock_irqrestore(&request->engine->timeline.lock, flags);
 }
 
 void i915_gem_set_wedged(struct drm_i915_private *i915)
@@ -3372,10 +3372,10 @@ void i915_gem_set_wedged(struct drm_i915_private *i915)
 		 * (lockless) lookup doesn't try and wait upon the request as we
 		 * reset it.
 		 */
-		spin_lock_irqsave(&engine->timeline->lock, flags);
+		spin_lock_irqsave(&engine->timeline.lock, flags);
 		intel_engine_init_global_seqno(engine,
 					       intel_engine_last_submit(engine));
-		spin_unlock_irqrestore(&engine->timeline->lock, flags);
+		spin_unlock_irqrestore(&engine->timeline.lock, flags);
 
 		i915_gem_reset_finish_engine(engine);
 	}
@@ -3387,8 +3387,7 @@ void i915_gem_set_wedged(struct drm_i915_private *i915)
 
 bool i915_gem_unset_wedged(struct drm_i915_private *i915)
 {
-	struct i915_gem_timeline *tl;
-	int i;
+	struct i915_timeline *tl;
 
 	lockdep_assert_held(&i915->drm.struct_mutex);
 	if (!test_bit(I915_WEDGED, &i915->gpu_error.flags))
@@ -3407,29 +3406,27 @@ bool i915_gem_unset_wedged(struct drm_i915_private *i915)
 	 * No more can be submitted until we reset the wedged bit.
 	 */
 	list_for_each_entry(tl, &i915->gt.timelines, link) {
-		for (i = 0; i < ARRAY_SIZE(tl->engine); i++) {
-			struct i915_request *rq;
+		struct i915_request *rq;
 
-			rq = i915_gem_active_peek(&tl->engine[i].last_request,
-						  &i915->drm.struct_mutex);
-			if (!rq)
-				continue;
+		rq = i915_gem_active_peek(&tl->last_request,
+					  &i915->drm.struct_mutex);
+		if (!rq)
+			continue;
 
-			/*
-			 * We can't use our normal waiter as we want to
-			 * avoid recursively trying to handle the current
-			 * reset. The basic dma_fence_default_wait() installs
-			 * a callback for dma_fence_signal(), which is
-			 * triggered by our nop handler (indirectly, the
-			 * callback enables the signaler thread which is
-			 * woken by the nop_submit_request() advancing the seqno
-			 * and when the seqno passes the fence, the signaler
-			 * then signals the fence waking us up).
-			 */
-			if (dma_fence_default_wait(&rq->fence, true,
-						   MAX_SCHEDULE_TIMEOUT) < 0)
-				return false;
-		}
+		/*
+		 * We can't use our normal waiter as we want to
+		 * avoid recursively trying to handle the current
+		 * reset. The basic dma_fence_default_wait() installs
+		 * a callback for dma_fence_signal(), which is
+		 * triggered by our nop handler (indirectly, the
+		 * callback enables the signaler thread which is
+		 * woken by the nop_submit_request() advancing the seqno
+		 * and when the seqno passes the fence, the signaler
+		 * then signals the fence waking us up).
+		 */
+		if (dma_fence_default_wait(&rq->fence, true,
+					   MAX_SCHEDULE_TIMEOUT) < 0)
+			return false;
 	}
 	i915_retire_requests(i915);
 	GEM_BUG_ON(i915->gt.active_requests);
@@ -3734,17 +3731,9 @@ i915_gem_wait_ioctl(struct drm_device *dev, void *data, struct drm_file *file)
 	return ret;
 }
 
-static int wait_for_timeline(struct i915_gem_timeline *tl, unsigned int flags)
+static int wait_for_timeline(struct i915_timeline *tl, unsigned int flags)
 {
-	int ret, i;
-
-	for (i = 0; i < ARRAY_SIZE(tl->engine); i++) {
-		ret = i915_gem_active_wait(&tl->engine[i].last_request, flags);
-		if (ret)
-			return ret;
-	}
-
-	return 0;
+	return i915_gem_active_wait(&tl->last_request, flags);
 }
 
 static int wait_for_engines(struct drm_i915_private *i915)
@@ -3762,30 +3751,37 @@ static int wait_for_engines(struct drm_i915_private *i915)
 
 int i915_gem_wait_for_idle(struct drm_i915_private *i915, unsigned int flags)
 {
-	int ret;
-
 	/* If the device is asleep, we have no requests outstanding */
 	if (!READ_ONCE(i915->gt.awake))
 		return 0;
 
 	if (flags & I915_WAIT_LOCKED) {
-		struct i915_gem_timeline *tl;
+		struct i915_timeline *tl;
+		int err;
 
 		lockdep_assert_held(&i915->drm.struct_mutex);
 
 		list_for_each_entry(tl, &i915->gt.timelines, link) {
-			ret = wait_for_timeline(tl, flags);
-			if (ret)
-				return ret;
+			err = wait_for_timeline(tl, flags);
+			if (err)
+				return err;
 		}
 		i915_retire_requests(i915);
 
-		ret = wait_for_engines(i915);
+		return wait_for_engines(i915);
 	} else {
-		ret = wait_for_timeline(&i915->gt.execution_timeline, flags);
-	}
+		struct intel_engine_cs *engine;
+		enum intel_engine_id id;
+		int err;
 
-	return ret;
+		for_each_engine(engine, i915, id) {
+			err = wait_for_timeline(&engine->timeline, flags);
+			if (err)
+				return err;
+		}
+
+		return 0;
+	}
 }
 
 static void __i915_gem_object_flush_for_display(struct drm_i915_gem_object *obj)
@@ -4954,7 +4950,7 @@ static void assert_kernel_context_is_current(struct drm_i915_private *i915)
 	enum intel_engine_id id;
 
 	for_each_engine(engine, i915, id) {
-		GEM_BUG_ON(__i915_gem_active_peek(&engine->timeline->last_request));
+		GEM_BUG_ON(__i915_gem_active_peek(&engine->timeline.last_request));
 		GEM_BUG_ON(engine->last_retired_context != kernel_context);
 	}
 }
@@ -5603,12 +5599,6 @@ int i915_gem_init_early(struct drm_i915_private *dev_priv)
 	INIT_LIST_HEAD(&dev_priv->gt.timelines);
 	INIT_LIST_HEAD(&dev_priv->gt.active_rings);
 
-	mutex_lock(&dev_priv->drm.struct_mutex);
-	err = i915_gem_timeline_init__global(dev_priv);
-	mutex_unlock(&dev_priv->drm.struct_mutex);
-	if (err)
-		goto err_priorities;
-
 	i915_gem_init__mm(dev_priv);
 
 	INIT_DELAYED_WORK(&dev_priv->gt.retire_work,
@@ -5628,8 +5618,6 @@ int i915_gem_init_early(struct drm_i915_private *dev_priv)
 
 	return 0;
 
-err_priorities:
-	kmem_cache_destroy(dev_priv->priorities);
 err_dependencies:
 	kmem_cache_destroy(dev_priv->dependencies);
 err_requests:
@@ -5650,12 +5638,7 @@ void i915_gem_cleanup_early(struct drm_i915_private *dev_priv)
 	GEM_BUG_ON(!llist_empty(&dev_priv->mm.free_list));
 	GEM_BUG_ON(atomic_read(&dev_priv->mm.free_count));
 	WARN_ON(dev_priv->mm.object_count);
-
-	mutex_lock(&dev_priv->drm.struct_mutex);
-	i915_gem_timeline_fini(&dev_priv->gt.legacy_timeline);
-	i915_gem_timeline_fini(&dev_priv->gt.execution_timeline);
 	WARN_ON(!list_empty(&dev_priv->gt.timelines));
-	mutex_unlock(&dev_priv->drm.struct_mutex);
 
 	kmem_cache_destroy(dev_priv->priorities);
 	kmem_cache_destroy(dev_priv->dependencies);
diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c
index 1f4987d..33f8a4b 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/i915_gem_context.c
@@ -122,7 +122,6 @@ static void i915_gem_context_free(struct i915_gem_context *ctx)
 	lockdep_assert_held(&ctx->i915->drm.struct_mutex);
 	GEM_BUG_ON(!i915_gem_context_is_closed(ctx));
 
-	i915_gem_timeline_free(ctx->timeline);
 	i915_ppgtt_put(ctx->ppgtt);
 
 	for (n = 0; n < ARRAY_SIZE(ctx->__engine); n++) {
@@ -377,18 +376,6 @@ i915_gem_create_context(struct drm_i915_private *dev_priv,
 		ctx->desc_template = default_desc_template(dev_priv, ppgtt);
 	}
 
-	if (HAS_EXECLISTS(dev_priv)) {
-		struct i915_gem_timeline *timeline;
-
-		timeline = i915_gem_timeline_create(dev_priv, ctx->name);
-		if (IS_ERR(timeline)) {
-			__destroy_hw_context(ctx, file_priv);
-			return ERR_CAST(timeline);
-		}
-
-		ctx->timeline = timeline;
-	}
-
 	trace_i915_context_create(ctx);
 
 	return ctx;
@@ -590,19 +577,29 @@ void i915_gem_context_close(struct drm_file *file)
 	idr_destroy(&file_priv->context_idr);
 }
 
-static bool engine_has_idle_kernel_context(struct intel_engine_cs *engine)
+static struct i915_request *
+last_request_on_engine(struct i915_timeline *timeline,
+		       struct intel_engine_cs *engine)
 {
-	struct i915_gem_timeline *timeline;
+	struct i915_request *rq;
 
-	list_for_each_entry(timeline, &engine->i915->gt.timelines, link) {
-		struct intel_timeline *tl;
+	if (timeline == &engine->timeline)
+		return NULL;
 
-		if (timeline == &engine->i915->gt.execution_timeline)
-			continue;
+	rq = i915_gem_active_raw(&timeline->last_request,
+				 &engine->i915->drm.struct_mutex);
+	if (rq && rq->engine == engine)
+		return rq;
+
+	return NULL;
+}
 
-		tl = &timeline->engine[engine->id];
-		if (i915_gem_active_peek(&tl->last_request,
-					 &engine->i915->drm.struct_mutex))
+static bool engine_has_idle_kernel_context(struct intel_engine_cs *engine)
+{
+	struct i915_timeline *timeline;
+
+	list_for_each_entry(timeline, &engine->i915->gt.timelines, link) {
+		if (last_request_on_engine(timeline, engine))
 			return false;
 	}
 
@@ -612,7 +609,7 @@ static bool engine_has_idle_kernel_context(struct intel_engine_cs *engine)
 int i915_gem_switch_to_kernel_context(struct drm_i915_private *dev_priv)
 {
 	struct intel_engine_cs *engine;
-	struct i915_gem_timeline *timeline;
+	struct i915_timeline *timeline;
 	enum intel_engine_id id;
 
 	lockdep_assert_held(&dev_priv->drm.struct_mutex);
@@ -632,11 +629,8 @@ int i915_gem_switch_to_kernel_context(struct drm_i915_private *dev_priv)
 		/* Queue this switch after all other activity */
 		list_for_each_entry(timeline, &dev_priv->gt.timelines, link) {
 			struct i915_request *prev;
-			struct intel_timeline *tl;
 
-			tl = &timeline->engine[engine->id];
-			prev = i915_gem_active_raw(&tl->last_request,
-						   &dev_priv->drm.struct_mutex);
+			prev = last_request_on_engine(timeline, engine);
 			if (prev)
 				i915_sw_fence_await_sw_fence_gfp(&rq->submit,
 								 &prev->submit,
diff --git a/drivers/gpu/drm/i915/i915_gem_context.h b/drivers/gpu/drm/i915/i915_gem_context.h
index ec53ba0..ace3b12 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.h
+++ b/drivers/gpu/drm/i915/i915_gem_context.h
@@ -58,8 +58,6 @@ struct i915_gem_context {
 	/** file_priv: owning file descriptor */
 	struct drm_i915_file_private *file_priv;
 
-	struct i915_gem_timeline *timeline;
-
 	/**
 	 * @ppgtt: unique address space (GTT)
 	 *
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.h b/drivers/gpu/drm/i915/i915_gem_gtt.h
index 9810792..1db0ded 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.h
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.h
@@ -38,10 +38,9 @@
 #include <linux/mm.h>
 #include <linux/pagevec.h>
 
-#include "i915_gem_timeline.h"
-
 #include "i915_request.h"
 #include "i915_selftest.h"
+#include "i915_timeline.h"
 
 #define I915_GTT_PAGE_SIZE_4K BIT(12)
 #define I915_GTT_PAGE_SIZE_64K BIT(16)
diff --git a/drivers/gpu/drm/i915/i915_gem_timeline.c b/drivers/gpu/drm/i915/i915_gem_timeline.c
deleted file mode 100644
index 24f4068..0000000
--- a/drivers/gpu/drm/i915/i915_gem_timeline.c
+++ /dev/null
@@ -1,198 +0,0 @@
-/*
- * Copyright © 2016 Intel Corporation
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice (including the next
- * paragraph) shall be included in all copies or substantial portions of the
- * Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
- * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
- * IN THE SOFTWARE.
- *
- */
-
-#include "i915_drv.h"
-#include "i915_syncmap.h"
-
-static void __intel_timeline_init(struct intel_timeline *tl,
-				  struct i915_gem_timeline *parent,
-				  u64 context,
-				  struct lock_class_key *lockclass,
-				  const char *lockname)
-{
-	tl->fence_context = context;
-	tl->common = parent;
-	spin_lock_init(&tl->lock);
-	lockdep_set_class_and_name(&tl->lock, lockclass, lockname);
-	init_request_active(&tl->last_request, NULL);
-	INIT_LIST_HEAD(&tl->requests);
-	i915_syncmap_init(&tl->sync);
-}
-
-static void __intel_timeline_fini(struct intel_timeline *tl)
-{
-	GEM_BUG_ON(!list_empty(&tl->requests));
-
-	i915_syncmap_free(&tl->sync);
-}
-
-static int __i915_gem_timeline_init(struct drm_i915_private *i915,
-				    struct i915_gem_timeline *timeline,
-				    const char *name,
-				    struct lock_class_key *lockclass,
-				    const char *lockname)
-{
-	unsigned int i;
-	u64 fences;
-
-	lockdep_assert_held(&i915->drm.struct_mutex);
-
-	/*
-	 * Ideally we want a set of engines on a single leaf as we expect
-	 * to mostly be tracking synchronisation between engines. It is not
-	 * a huge issue if this is not the case, but we may want to mitigate
-	 * any page crossing penalties if they become an issue.
-	 */
-	BUILD_BUG_ON(KSYNCMAP < I915_NUM_ENGINES);
-
-	timeline->i915 = i915;
-	timeline->name = kstrdup(name ?: "[kernel]", GFP_KERNEL);
-	if (!timeline->name)
-		return -ENOMEM;
-
-	list_add(&timeline->link, &i915->gt.timelines);
-
-	/* Called during early_init before we know how many engines there are */
-	fences = dma_fence_context_alloc(ARRAY_SIZE(timeline->engine));
-	for (i = 0; i < ARRAY_SIZE(timeline->engine); i++)
-		__intel_timeline_init(&timeline->engine[i],
-				      timeline, fences++,
-				      lockclass, lockname);
-
-	return 0;
-}
-
-int i915_gem_timeline_init(struct drm_i915_private *i915,
-			   struct i915_gem_timeline *timeline,
-			   const char *name)
-{
-	static struct lock_class_key class;
-
-	return __i915_gem_timeline_init(i915, timeline, name,
-					&class, "&timeline->lock");
-}
-
-int i915_gem_timeline_init__global(struct drm_i915_private *i915)
-{
-	static struct lock_class_key class1, class2;
-	int err;
-
-	err = __i915_gem_timeline_init(i915,
-				       &i915->gt.execution_timeline,
-				       "[execution]", &class1,
-				       "i915_execution_timeline");
-	if (err)
-		return err;
-
-	err = __i915_gem_timeline_init(i915,
-				       &i915->gt.legacy_timeline,
-				       "[global]", &class2,
-				       "i915_global_timeline");
-	if (err)
-		goto err_exec_timeline;
-
-	return 0;
-
-err_exec_timeline:
-	i915_gem_timeline_fini(&i915->gt.execution_timeline);
-	return err;
-}
-
-/**
- * i915_gem_timelines_park - called when the driver idles
- * @i915: the drm_i915_private device
- *
- * When the driver is completely idle, we know that all of our sync points
- * have been signaled and our tracking is then entirely redundant. Any request
- * to wait upon an older sync point will be completed instantly as we know
- * the fence is signaled and therefore we will not even look them up in the
- * sync point map.
- */
-void i915_gem_timelines_park(struct drm_i915_private *i915)
-{
-	struct i915_gem_timeline *timeline;
-	int i;
-
-	lockdep_assert_held(&i915->drm.struct_mutex);
-
-	list_for_each_entry(timeline, &i915->gt.timelines, link) {
-		for (i = 0; i < ARRAY_SIZE(timeline->engine); i++) {
-			struct intel_timeline *tl = &timeline->engine[i];
-
-			/*
-			 * All known fences are completed so we can scrap
-			 * the current sync point tracking and start afresh,
-			 * any attempt to wait upon a previous sync point
-			 * will be skipped as the fence was signaled.
-			 */
-			i915_syncmap_free(&tl->sync);
-		}
-	}
-}
-
-void i915_gem_timeline_fini(struct i915_gem_timeline *timeline)
-{
-	int i;
-
-	lockdep_assert_held(&timeline->i915->drm.struct_mutex);
-
-	for (i = 0; i < ARRAY_SIZE(timeline->engine); i++)
-		__intel_timeline_fini(&timeline->engine[i]);
-
-	list_del(&timeline->link);
-	kfree(timeline->name);
-}
-
-struct i915_gem_timeline *
-i915_gem_timeline_create(struct drm_i915_private *i915, const char *name)
-{
-	struct i915_gem_timeline *timeline;
-	int err;
-
-	timeline = kzalloc(sizeof(*timeline), GFP_KERNEL);
-	if (!timeline)
-		return ERR_PTR(-ENOMEM);
-
-	err = i915_gem_timeline_init(i915, timeline, name);
-	if (err) {
-		kfree(timeline);
-		return ERR_PTR(err);
-	}
-
-	return timeline;
-}
-
-void i915_gem_timeline_free(struct i915_gem_timeline *timeline)
-{
-	if (!timeline)
-		return;
-
-	i915_gem_timeline_fini(timeline);
-	kfree(timeline);
-}
-
-#if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
-#include "selftests/mock_timeline.c"
-#include "selftests/i915_gem_timeline.c"
-#endif
diff --git a/drivers/gpu/drm/i915/i915_gem_timeline.h b/drivers/gpu/drm/i915/i915_gem_timeline.h
deleted file mode 100644
index 780ed46..0000000
--- a/drivers/gpu/drm/i915/i915_gem_timeline.h
+++ /dev/null
@@ -1,121 +0,0 @@
-/*
- * Copyright © 2016 Intel Corporation
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice (including the next
- * paragraph) shall be included in all copies or substantial portions of the
- * Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
- * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
- * IN THE SOFTWARE.
- *
- */
-
-#ifndef I915_GEM_TIMELINE_H
-#define I915_GEM_TIMELINE_H
-
-#include <linux/list.h>
-
-#include "i915_request.h"
-#include "i915_syncmap.h"
-#include "i915_utils.h"
-
-struct i915_gem_timeline;
-
-struct intel_timeline {
-	u64 fence_context;
-	u32 seqno;
-
-	spinlock_t lock;
-
-	/**
-	 * List of breadcrumbs associated with GPU requests currently
-	 * outstanding.
-	 */
-	struct list_head requests;
-
-	/* Contains an RCU guarded pointer to the last request. No reference is
-	 * held to the request, users must carefully acquire a reference to
-	 * the request using i915_gem_active_get_request_rcu(), or hold the
-	 * struct_mutex.
-	 */
-	struct i915_gem_active last_request;
-
-	/**
-	 * We track the most recent seqno that we wait on in every context so
-	 * that we only have to emit a new await and dependency on a more
-	 * recent sync point. As the contexts may be executed out-of-order, we
-	 * have to track each individually and can not rely on an absolute
-	 * global_seqno. When we know that all tracked fences are completed
-	 * (i.e. when the driver is idle), we know that the syncmap is
-	 * redundant and we can discard it without loss of generality.
-	 */
-	struct i915_syncmap *sync;
-	/**
-	 * Separately to the inter-context seqno map above, we track the last
-	 * barrier (e.g. semaphore wait) to the global engine timelines. Note
-	 * that this tracks global_seqno rather than the context.seqno, and
-	 * so it is subject to the limitations of hw wraparound and that we
-	 * may need to revoke global_seqno (on pre-emption).
-	 */
-	u32 global_sync[I915_NUM_ENGINES];
-
-	struct i915_gem_timeline *common;
-};
-
-struct i915_gem_timeline {
-	struct list_head link;
-
-	struct drm_i915_private *i915;
-	const char *name;
-
-	struct intel_timeline engine[I915_NUM_ENGINES];
-};
-
-int i915_gem_timeline_init(struct drm_i915_private *i915,
-			   struct i915_gem_timeline *tl,
-			   const char *name);
-int i915_gem_timeline_init__global(struct drm_i915_private *i915);
-void i915_gem_timelines_park(struct drm_i915_private *i915);
-void i915_gem_timeline_fini(struct i915_gem_timeline *tl);
-
-struct i915_gem_timeline *
-i915_gem_timeline_create(struct drm_i915_private *i915, const char *name);
-void i915_gem_timeline_free(struct i915_gem_timeline *timeline);
-
-static inline int __intel_timeline_sync_set(struct intel_timeline *tl,
-					    u64 context, u32 seqno)
-{
-	return i915_syncmap_set(&tl->sync, context, seqno);
-}
-
-static inline int intel_timeline_sync_set(struct intel_timeline *tl,
-					  const struct dma_fence *fence)
-{
-	return __intel_timeline_sync_set(tl, fence->context, fence->seqno);
-}
-
-static inline bool __intel_timeline_sync_is_later(struct intel_timeline *tl,
-						  u64 context, u32 seqno)
-{
-	return i915_syncmap_is_later(&tl->sync, context, seqno);
-}
-
-static inline bool intel_timeline_sync_is_later(struct intel_timeline *tl,
-						const struct dma_fence *fence)
-{
-	return __intel_timeline_sync_is_later(tl, fence->context, fence->seqno);
-}
-
-#endif
diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c
index 9449399..df234dc 100644
--- a/drivers/gpu/drm/i915/i915_gpu_error.c
+++ b/drivers/gpu/drm/i915/i915_gpu_error.c
@@ -1310,7 +1310,7 @@ static void engine_record_requests(struct intel_engine_cs *engine,
 
 	count = 0;
 	request = first;
-	list_for_each_entry_from(request, &engine->timeline->requests, link)
+	list_for_each_entry_from(request, &engine->timeline.requests, link)
 		count++;
 	if (!count)
 		return;
@@ -1323,7 +1323,7 @@ static void engine_record_requests(struct intel_engine_cs *engine,
 
 	count = 0;
 	request = first;
-	list_for_each_entry_from(request, &engine->timeline->requests, link) {
+	list_for_each_entry_from(request, &engine->timeline.requests, link) {
 		if (count >= ee->num_requests) {
 			/*
 			 * If the ring request list was changed in
diff --git a/drivers/gpu/drm/i915/i915_perf.c b/drivers/gpu/drm/i915/i915_perf.c
index 4b1da01..d934141 100644
--- a/drivers/gpu/drm/i915/i915_perf.c
+++ b/drivers/gpu/drm/i915/i915_perf.c
@@ -1695,7 +1695,7 @@ static int gen8_switch_to_updated_kernel_context(struct drm_i915_private *dev_pr
 						 const struct i915_oa_config *oa_config)
 {
 	struct intel_engine_cs *engine = dev_priv->engine[RCS];
-	struct i915_gem_timeline *timeline;
+	struct i915_timeline *timeline;
 	struct i915_request *rq;
 	int ret;
 
@@ -1716,15 +1716,11 @@ static int gen8_switch_to_updated_kernel_context(struct drm_i915_private *dev_pr
 	/* Queue this switch after all other activity */
 	list_for_each_entry(timeline, &dev_priv->gt.timelines, link) {
 		struct i915_request *prev;
-		struct intel_timeline *tl;
 
-		tl = &timeline->engine[engine->id];
-		prev = i915_gem_active_raw(&tl->last_request,
+		prev = i915_gem_active_raw(&timeline->last_request,
 					   &dev_priv->drm.struct_mutex);
 		if (prev)
-			i915_sw_fence_await_sw_fence_gfp(&rq->submit,
-							 &prev->submit,
-							 GFP_KERNEL);
+			i915_request_await_dma_fence(rq, &prev->fence);
 	}
 
 	i915_request_add(rq);
diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c
index 7bb613c..5acf869f 100644
--- a/drivers/gpu/drm/i915/i915_request.c
+++ b/drivers/gpu/drm/i915/i915_request.c
@@ -49,7 +49,7 @@ static const char *i915_fence_get_timeline_name(struct dma_fence *fence)
 	if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags))
 		return "signaled";
 
-	return to_request(fence)->timeline->common->name;
+	return to_request(fence)->timeline->name;
 }
 
 static bool i915_fence_signaled(struct dma_fence *fence)
@@ -199,6 +199,7 @@ i915_sched_node_init(struct i915_sched_node *node)
 static int reset_all_global_seqno(struct drm_i915_private *i915, u32 seqno)
 {
 	struct intel_engine_cs *engine;
+	struct i915_timeline *timeline;
 	enum intel_engine_id id;
 	int ret;
 
@@ -213,16 +214,13 @@ static int reset_all_global_seqno(struct drm_i915_private *i915, u32 seqno)
 
 	/* If the seqno wraps around, we need to clear the breadcrumb rbtree */
 	for_each_engine(engine, i915, id) {
-		struct i915_gem_timeline *timeline;
-		struct intel_timeline *tl = engine->timeline;
-
 		GEM_TRACE("%s seqno %d (current %d) -> %d\n",
 			  engine->name,
-			  tl->seqno,
+			  engine->timeline.seqno,
 			  intel_engine_get_seqno(engine),
 			  seqno);
 
-		if (!i915_seqno_passed(seqno, tl->seqno)) {
+		if (!i915_seqno_passed(seqno, engine->timeline.seqno)) {
 			/* Flush any waiters before we reuse the seqno */
 			intel_engine_disarm_breadcrumbs(engine);
 			GEM_BUG_ON(!list_empty(&engine->breadcrumbs.signals));
@@ -230,18 +228,18 @@ static int reset_all_global_seqno(struct drm_i915_private *i915, u32 seqno)
 
 		/* Check we are idle before we fiddle with hw state! */
 		GEM_BUG_ON(!intel_engine_is_idle(engine));
-		GEM_BUG_ON(i915_gem_active_isset(&engine->timeline->last_request));
+		GEM_BUG_ON(i915_gem_active_isset(&engine->timeline.last_request));
 
 		/* Finally reset hw state */
 		intel_engine_init_global_seqno(engine, seqno);
-		tl->seqno = seqno;
-
-		list_for_each_entry(timeline, &i915->gt.timelines, link)
-			memset(timeline->engine[id].global_sync, 0,
-			       sizeof(timeline->engine[id].global_sync));
+		engine->timeline.seqno = seqno;
 	}
 
+	list_for_each_entry(timeline, &i915->gt.timelines, link)
+		memset(timeline->global_sync, 0, sizeof(timeline->global_sync));
+
 	i915->gt.request_serial = seqno;
+
 	return 0;
 }
 
@@ -357,10 +355,10 @@ static void __retire_engine_request(struct intel_engine_cs *engine,
 
 	local_irq_disable();
 
-	spin_lock(&engine->timeline->lock);
-	GEM_BUG_ON(!list_is_first(&rq->link, &engine->timeline->requests));
+	spin_lock(&engine->timeline.lock);
+	GEM_BUG_ON(!list_is_first(&rq->link, &engine->timeline.requests));
 	list_del_init(&rq->link);
-	spin_unlock(&engine->timeline->lock);
+	spin_unlock(&engine->timeline.lock);
 
 	spin_lock(&rq->lock);
 	if (!test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &rq->fence.flags))
@@ -397,7 +395,7 @@ static void __retire_engine_upto(struct intel_engine_cs *engine,
 		return;
 
 	do {
-		tmp = list_first_entry(&engine->timeline->requests,
+		tmp = list_first_entry(&engine->timeline.requests,
 				       typeof(*tmp), link);
 
 		GEM_BUG_ON(tmp->engine != engine);
@@ -492,16 +490,16 @@ void i915_request_retire_upto(struct i915_request *rq)
 	} while (tmp != rq);
 }
 
-static u32 timeline_get_seqno(struct intel_timeline *tl)
+static u32 timeline_get_seqno(struct i915_timeline *tl)
 {
 	return ++tl->seqno;
 }
 
 static void move_to_timeline(struct i915_request *request,
-			     struct intel_timeline *timeline)
+			     struct i915_timeline *timeline)
 {
-	GEM_BUG_ON(request->timeline == request->engine->timeline);
-	lockdep_assert_held(&request->engine->timeline->lock);
+	GEM_BUG_ON(request->timeline == &request->engine->timeline);
+	lockdep_assert_held(&request->engine->timeline.lock);
 
 	spin_lock(&request->timeline->lock);
 	list_move_tail(&request->link, &timeline->requests);
@@ -516,15 +514,15 @@ void __i915_request_submit(struct i915_request *request)
 	GEM_TRACE("%s fence %llx:%d -> global=%d, current %d\n",
 		  engine->name,
 		  request->fence.context, request->fence.seqno,
-		  engine->timeline->seqno + 1,
+		  engine->timeline.seqno + 1,
 		  intel_engine_get_seqno(engine));
 
 	GEM_BUG_ON(!irqs_disabled());
-	lockdep_assert_held(&engine->timeline->lock);
+	lockdep_assert_held(&engine->timeline.lock);
 
 	GEM_BUG_ON(request->global_seqno);
 
-	seqno = timeline_get_seqno(engine->timeline);
+	seqno = timeline_get_seqno(&engine->timeline);
 	GEM_BUG_ON(!seqno);
 	GEM_BUG_ON(i915_seqno_passed(intel_engine_get_seqno(engine), seqno));
 
@@ -539,7 +537,7 @@ void __i915_request_submit(struct i915_request *request)
 				request->ring->vaddr + request->postfix);
 
 	/* Transfer from per-context onto the global per-engine timeline */
-	move_to_timeline(request, engine->timeline);
+	move_to_timeline(request, &engine->timeline);
 
 	trace_i915_request_execute(request);
 
@@ -552,11 +550,11 @@ void i915_request_submit(struct i915_request *request)
 	unsigned long flags;
 
 	/* Will be called from irq-context when using foreign fences. */
-	spin_lock_irqsave(&engine->timeline->lock, flags);
+	spin_lock_irqsave(&engine->timeline.lock, flags);
 
 	__i915_request_submit(request);
 
-	spin_unlock_irqrestore(&engine->timeline->lock, flags);
+	spin_unlock_irqrestore(&engine->timeline.lock, flags);
 }
 
 void __i915_request_unsubmit(struct i915_request *request)
@@ -570,17 +568,17 @@ void __i915_request_unsubmit(struct i915_request *request)
 		  intel_engine_get_seqno(engine));
 
 	GEM_BUG_ON(!irqs_disabled());
-	lockdep_assert_held(&engine->timeline->lock);
+	lockdep_assert_held(&engine->timeline.lock);
 
 	/*
 	 * Only unwind in reverse order, required so that the per-context list
 	 * is kept in seqno/ring order.
 	 */
 	GEM_BUG_ON(!request->global_seqno);
-	GEM_BUG_ON(request->global_seqno != engine->timeline->seqno);
+	GEM_BUG_ON(request->global_seqno != engine->timeline.seqno);
 	GEM_BUG_ON(i915_seqno_passed(intel_engine_get_seqno(engine),
 				     request->global_seqno));
-	engine->timeline->seqno--;
+	engine->timeline.seqno--;
 
 	/* We may be recursing from the signal callback of another i915 fence */
 	spin_lock_nested(&request->lock, SINGLE_DEPTH_NESTING);
@@ -607,11 +605,11 @@ void i915_request_unsubmit(struct i915_request *request)
 	unsigned long flags;
 
 	/* Will be called from irq-context when using foreign fences. */
-	spin_lock_irqsave(&engine->timeline->lock, flags);
+	spin_lock_irqsave(&engine->timeline.lock, flags);
 
 	__i915_request_unsubmit(request);
 
-	spin_unlock_irqrestore(&engine->timeline->lock, flags);
+	spin_unlock_irqrestore(&engine->timeline.lock, flags);
 }
 
 static int __i915_sw_fence_call
@@ -764,7 +762,7 @@ i915_request_alloc(struct intel_engine_cs *engine, struct i915_gem_context *ctx)
 	rq->ctx = ctx;
 	rq->ring = ring;
 	rq->timeline = ring->timeline;
-	GEM_BUG_ON(rq->timeline == engine->timeline);
+	GEM_BUG_ON(rq->timeline == &engine->timeline);
 
 	spin_lock_init(&rq->lock);
 	dma_fence_init(&rq->fence,
@@ -929,7 +927,7 @@ i915_request_await_dma_fence(struct i915_request *rq, struct dma_fence *fence)
 
 		/* Squash repeated waits to the same timelines */
 		if (fence->context != rq->i915->mm.unordered_timeline &&
-		    intel_timeline_sync_is_later(rq->timeline, fence))
+		    i915_timeline_sync_is_later(rq->timeline, fence))
 			continue;
 
 		if (dma_fence_is_i915(fence))
@@ -943,7 +941,7 @@ i915_request_await_dma_fence(struct i915_request *rq, struct dma_fence *fence)
 
 		/* Record the latest fence used against each timeline */
 		if (fence->context != rq->i915->mm.unordered_timeline)
-			intel_timeline_sync_set(rq->timeline, fence);
+			i915_timeline_sync_set(rq->timeline, fence);
 	} while (--nchild);
 
 	return 0;
@@ -1020,7 +1018,7 @@ void __i915_request_add(struct i915_request *request, bool flush_caches)
 {
 	struct intel_engine_cs *engine = request->engine;
 	struct intel_ring *ring = request->ring;
-	struct intel_timeline *timeline = request->timeline;
+	struct i915_timeline *timeline = request->timeline;
 	struct i915_request *prev;
 	u32 *cs;
 	int err;
diff --git a/drivers/gpu/drm/i915/i915_request.h b/drivers/gpu/drm/i915/i915_request.h
index 8f31ca8..eddbd424 100644
--- a/drivers/gpu/drm/i915/i915_request.h
+++ b/drivers/gpu/drm/i915/i915_request.h
@@ -37,6 +37,7 @@
 struct drm_file;
 struct drm_i915_gem_object;
 struct i915_request;
+struct i915_timeline;
 
 struct intel_wait {
 	struct rb_node node;
@@ -95,7 +96,7 @@ struct i915_request {
 	struct i915_gem_context *ctx;
 	struct intel_engine_cs *engine;
 	struct intel_ring *ring;
-	struct intel_timeline *timeline;
+	struct i915_timeline *timeline;
 	struct intel_signal_node signaling;
 
 	/*
diff --git a/drivers/gpu/drm/i915/i915_timeline.c b/drivers/gpu/drm/i915/i915_timeline.c
new file mode 100644
index 0000000..4667cc0
--- /dev/null
+++ b/drivers/gpu/drm/i915/i915_timeline.c
@@ -0,0 +1,105 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2016-2018 Intel Corporation
+ */
+
+#include "i915_drv.h"
+
+#include "i915_timeline.h"
+#include "i915_syncmap.h"
+
+void i915_timeline_init(struct drm_i915_private *i915,
+			struct i915_timeline *timeline,
+			const char *name)
+{
+	lockdep_assert_held(&i915->drm.struct_mutex);
+
+	/*
+	 * Ideally we want a set of engines on a single leaf as we expect
+	 * to mostly be tracking synchronisation between engines. It is not
+	 * a huge issue if this is not the case, but we may want to mitigate
+	 * any page crossing penalties if they become an issue.
+	 */
+	BUILD_BUG_ON(KSYNCMAP < I915_NUM_ENGINES);
+
+	timeline->name = name;
+
+	list_add(&timeline->link, &i915->gt.timelines);
+
+	/* Called during early_init before we know how many engines there are */
+
+	timeline->fence_context = dma_fence_context_alloc(1);
+
+	spin_lock_init(&timeline->lock);
+
+	init_request_active(&timeline->last_request, NULL);
+	INIT_LIST_HEAD(&timeline->requests);
+
+	i915_syncmap_init(&timeline->sync);
+}
+
+/**
+ * i915_timelines_park - called when the driver idles
+ * @i915: the drm_i915_private device
+ *
+ * When the driver is completely idle, we know that all of our sync points
+ * have been signaled and our tracking is then entirely redundant. Any request
+ * to wait upon an older sync point will be completed instantly as we know
+ * the fence is signaled and therefore we will not even look them up in the
+ * sync point map.
+ */
+void i915_timelines_park(struct drm_i915_private *i915)
+{
+	struct i915_timeline *timeline;
+
+	lockdep_assert_held(&i915->drm.struct_mutex);
+
+	list_for_each_entry(timeline, &i915->gt.timelines, link) {
+		/*
+		 * All known fences are completed so we can scrap
+		 * the current sync point tracking and start afresh,
+		 * any attempt to wait upon a previous sync point
+		 * will be skipped as the fence was signaled.
+		 */
+		i915_syncmap_free(&timeline->sync);
+	}
+}
+
+void i915_timeline_fini(struct i915_timeline *timeline)
+{
+	GEM_BUG_ON(!list_empty(&timeline->requests));
+
+	i915_syncmap_free(&timeline->sync);
+
+	list_del(&timeline->link);
+}
+
+struct i915_timeline *
+i915_timeline_create(struct drm_i915_private *i915, const char *name)
+{
+	struct i915_timeline *timeline;
+
+	timeline = kzalloc(sizeof(*timeline), GFP_KERNEL);
+	if (!timeline)
+		return ERR_PTR(-ENOMEM);
+
+	i915_timeline_init(i915, timeline, name);
+	kref_init(&timeline->kref);
+
+	return timeline;
+}
+
+void __i915_timeline_free(struct kref *kref)
+{
+	struct i915_timeline *timeline =
+		container_of(kref, typeof(*timeline), kref);
+
+	i915_timeline_fini(timeline);
+	kfree(timeline);
+}
+
+#if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
+#include "selftests/mock_timeline.c"
+#include "selftests/i915_timeline.c"
+#endif
diff --git a/drivers/gpu/drm/i915/i915_timeline.h b/drivers/gpu/drm/i915/i915_timeline.h
new file mode 100644
index 0000000..dc2a463
--- /dev/null
+++ b/drivers/gpu/drm/i915/i915_timeline.h
@@ -0,0 +1,126 @@
+/*
+ * Copyright © 2016 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ */
+
+#ifndef I915_TIMELINE_H
+#define I915_TIMELINE_H
+
+#include <linux/list.h>
+#include <linux/kref.h>
+
+#include "i915_request.h"
+#include "i915_syncmap.h"
+#include "i915_utils.h"
+
+struct i915_timeline {
+	u64 fence_context;
+	u32 seqno;
+
+	spinlock_t lock;
+
+	/**
+	 * List of breadcrumbs associated with GPU requests currently
+	 * outstanding.
+	 */
+	struct list_head requests;
+
+	/* Contains an RCU guarded pointer to the last request. No reference is
+	 * held to the request, users must carefully acquire a reference to
+	 * the request using i915_gem_active_get_request_rcu(), or hold the
+	 * struct_mutex.
+	 */
+	struct i915_gem_active last_request;
+
+	/**
+	 * We track the most recent seqno that we wait on in every context so
+	 * that we only have to emit a new await and dependency on a more
+	 * recent sync point. As the contexts may be executed out-of-order, we
+	 * have to track each individually and can not rely on an absolute
+	 * global_seqno. When we know that all tracked fences are completed
+	 * (i.e. when the driver is idle), we know that the syncmap is
+	 * redundant and we can discard it without loss of generality.
+	 */
+	struct i915_syncmap *sync;
+	/**
+	 * Separately to the inter-context seqno map above, we track the last
+	 * barrier (e.g. semaphore wait) to the global engine timelines. Note
+	 * that this tracks global_seqno rather than the context.seqno, and
+	 * so it is subject to the limitations of hw wraparound and that we
+	 * may need to revoke global_seqno (on pre-emption).
+	 */
+	u32 global_sync[I915_NUM_ENGINES];
+
+	struct list_head link;
+	const char *name;
+
+	struct kref kref;
+};
+
+void i915_timeline_init(struct drm_i915_private *i915,
+			struct i915_timeline *tl,
+			const char *name);
+void i915_timeline_fini(struct i915_timeline *tl);
+
+struct i915_timeline *
+i915_timeline_create(struct drm_i915_private *i915, const char *name);
+
+static inline struct i915_timeline *
+i915_timeline_get(struct i915_timeline *timeline)
+{
+	kref_get(&timeline->kref);
+	return timeline;
+}
+
+void __i915_timeline_free(struct kref *kref);
+static inline void i915_timeline_put(struct i915_timeline *timeline)
+{
+	kref_put(&timeline->kref, __i915_timeline_free);
+}
+
+static inline int __i915_timeline_sync_set(struct i915_timeline *tl,
+					   u64 context, u32 seqno)
+{
+	return i915_syncmap_set(&tl->sync, context, seqno);
+}
+
+static inline int i915_timeline_sync_set(struct i915_timeline *tl,
+					 const struct dma_fence *fence)
+{
+	return __i915_timeline_sync_set(tl, fence->context, fence->seqno);
+}
+
+static inline bool __i915_timeline_sync_is_later(struct i915_timeline *tl,
+						 u64 context, u32 seqno)
+{
+	return i915_syncmap_is_later(&tl->sync, context, seqno);
+}
+
+static inline bool i915_timeline_sync_is_later(struct i915_timeline *tl,
+					       const struct dma_fence *fence)
+{
+	return __i915_timeline_sync_is_later(tl, fence->context, fence->seqno);
+}
+
+void i915_timelines_park(struct drm_i915_private *i915);
+
+#endif
diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
index 7af5fe8..a90769b 100644
--- a/drivers/gpu/drm/i915/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/intel_engine_cs.c
@@ -451,12 +451,6 @@ void intel_engine_init_global_seqno(struct intel_engine_cs *engine, u32 seqno)
 	GEM_BUG_ON(intel_engine_get_seqno(engine) != seqno);
 }
 
-static void intel_engine_init_timeline(struct intel_engine_cs *engine)
-{
-	engine->timeline =
-		&engine->i915->gt.execution_timeline.engine[engine->id];
-}
-
 static void intel_engine_init_batch_pool(struct intel_engine_cs *engine)
 {
 	i915_gem_batch_pool_init(&engine->batch_pool, engine);
@@ -508,8 +502,9 @@ static void intel_engine_init_execlist(struct intel_engine_cs *engine)
  */
 void intel_engine_setup_common(struct intel_engine_cs *engine)
 {
+	i915_timeline_init(engine->i915, &engine->timeline, engine->name);
+
 	intel_engine_init_execlist(engine);
-	intel_engine_init_timeline(engine);
 	intel_engine_init_hangcheck(engine);
 	intel_engine_init_batch_pool(engine);
 	intel_engine_init_cmd_parser(engine);
@@ -751,6 +746,8 @@ void intel_engine_cleanup_common(struct intel_engine_cs *engine)
 	if (engine->i915->preempt_context)
 		intel_context_unpin(engine->i915->preempt_context, engine);
 	intel_context_unpin(engine->i915->kernel_context, engine);
+
+	i915_timeline_fini(&engine->timeline);
 }
 
 u64 intel_engine_get_active_head(const struct intel_engine_cs *engine)
@@ -1003,7 +1000,7 @@ bool intel_engine_has_kernel_context(const struct intel_engine_cs *engine)
 	 * the last request that remains in the timeline. When idle, it is
 	 * the last executed context as tracked by retirement.
 	 */
-	rq = __i915_gem_active_peek(&engine->timeline->last_request);
+	rq = __i915_gem_active_peek(&engine->timeline.last_request);
 	if (rq)
 		return rq->ctx == kernel_context;
 	else
@@ -1335,14 +1332,14 @@ void intel_engine_dump(struct intel_engine_cs *engine,
 
 	drm_printf(m, "\tRequests:\n");
 
-	rq = list_first_entry(&engine->timeline->requests,
+	rq = list_first_entry(&engine->timeline.requests,
 			      struct i915_request, link);
-	if (&rq->link != &engine->timeline->requests)
+	if (&rq->link != &engine->timeline.requests)
 		print_request(m, rq, "\t\tfirst  ");
 
-	rq = list_last_entry(&engine->timeline->requests,
+	rq = list_last_entry(&engine->timeline.requests,
 			     struct i915_request, link);
-	if (&rq->link != &engine->timeline->requests)
+	if (&rq->link != &engine->timeline.requests)
 		print_request(m, rq, "\t\tlast   ");
 
 	rq = i915_gem_find_active_request(engine);
@@ -1374,11 +1371,11 @@ void intel_engine_dump(struct intel_engine_cs *engine,
 		drm_printf(m, "\tDevice is asleep; skipping register dump\n");
 	}
 
-	spin_lock_irq(&engine->timeline->lock);
+	spin_lock_irq(&engine->timeline.lock);
 
 	last = NULL;
 	count = 0;
-	list_for_each_entry(rq, &engine->timeline->requests, link) {
+	list_for_each_entry(rq, &engine->timeline.requests, link) {
 		if (count++ < MAX_REQUESTS_TO_SHOW - 1)
 			print_request(m, rq, "\t\tE ");
 		else
@@ -1416,7 +1413,7 @@ void intel_engine_dump(struct intel_engine_cs *engine,
 		print_request(m, last, "\t\tQ ");
 	}
 
-	spin_unlock_irq(&engine->timeline->lock);
+	spin_unlock_irq(&engine->timeline.lock);
 
 	spin_lock_irq(&b->rb_lock);
 	for (rb = rb_first(&b->waiters); rb; rb = rb_next(rb)) {
diff --git a/drivers/gpu/drm/i915/intel_guc_submission.c b/drivers/gpu/drm/i915/intel_guc_submission.c
index c6bb5be..62828e3 100644
--- a/drivers/gpu/drm/i915/intel_guc_submission.c
+++ b/drivers/gpu/drm/i915/intel_guc_submission.c
@@ -679,7 +679,7 @@ static void guc_dequeue(struct intel_engine_cs *engine)
 	bool submit = false;
 	struct rb_node *rb;
 
-	spin_lock_irq(&engine->timeline->lock);
+	spin_lock_irq(&engine->timeline.lock);
 	rb = execlists->first;
 	GEM_BUG_ON(rb_first(&execlists->queue) != rb);
 
@@ -750,7 +750,7 @@ done:
 	GEM_BUG_ON(execlists->first && !port_isset(execlists->port));
 
 unlock:
-	spin_unlock_irq(&engine->timeline->lock);
+	spin_unlock_irq(&engine->timeline.lock);
 }
 
 static void guc_submission_tasklet(unsigned long data)
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 9b24077..e04798e 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -331,10 +331,10 @@ static void __unwind_incomplete_requests(struct intel_engine_cs *engine)
 	struct i915_priolist *uninitialized_var(p);
 	int last_prio = I915_PRIORITY_INVALID;
 
-	lockdep_assert_held(&engine->timeline->lock);
+	lockdep_assert_held(&engine->timeline.lock);
 
 	list_for_each_entry_safe_reverse(rq, rn,
-					 &engine->timeline->requests,
+					 &engine->timeline.requests,
 					 link) {
 		if (i915_request_completed(rq))
 			return;
@@ -358,9 +358,9 @@ execlists_unwind_incomplete_requests(struct intel_engine_execlists *execlists)
 	struct intel_engine_cs *engine =
 		container_of(execlists, typeof(*engine), execlists);
 
-	spin_lock_irq(&engine->timeline->lock);
+	spin_lock_irq(&engine->timeline.lock);
 	__unwind_incomplete_requests(engine);
-	spin_unlock_irq(&engine->timeline->lock);
+	spin_unlock_irq(&engine->timeline.lock);
 }
 
 static inline void
@@ -584,7 +584,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 	 * and context switches) submission.
 	 */
 
-	spin_lock_irq(&engine->timeline->lock);
+	spin_lock_irq(&engine->timeline.lock);
 	rb = execlists->first;
 	GEM_BUG_ON(rb_first(&execlists->queue) != rb);
 
@@ -744,7 +744,7 @@ done:
 	GEM_BUG_ON(execlists->first && !port_isset(execlists->port));
 
 unlock:
-	spin_unlock_irq(&engine->timeline->lock);
+	spin_unlock_irq(&engine->timeline.lock);
 
 	if (submit) {
 		execlists_user_begin(execlists, execlists->port);
@@ -894,10 +894,10 @@ static void execlists_cancel_requests(struct intel_engine_cs *engine)
 	execlists_cancel_port_requests(execlists);
 	reset_irq(engine);
 
-	spin_lock(&engine->timeline->lock);
+	spin_lock(&engine->timeline.lock);
 
 	/* Mark all executing requests as skipped. */
-	list_for_each_entry(rq, &engine->timeline->requests, link) {
+	list_for_each_entry(rq, &engine->timeline.requests, link) {
 		GEM_BUG_ON(!rq->global_seqno);
 		if (!i915_request_completed(rq))
 			dma_fence_set_error(&rq->fence, -EIO);
@@ -929,7 +929,7 @@ static void execlists_cancel_requests(struct intel_engine_cs *engine)
 	execlists->first = NULL;
 	GEM_BUG_ON(port_isset(execlists->port));
 
-	spin_unlock(&engine->timeline->lock);
+	spin_unlock(&engine->timeline.lock);
 
 	local_irq_restore(flags);
 }
@@ -1167,7 +1167,7 @@ static void execlists_submit_request(struct i915_request *request)
 	unsigned long flags;
 
 	/* Will be called from irq-context when using foreign fences. */
-	spin_lock_irqsave(&engine->timeline->lock, flags);
+	spin_lock_irqsave(&engine->timeline.lock, flags);
 
 	queue_request(engine, &request->sched, rq_prio(request));
 	submit_queue(engine, rq_prio(request));
@@ -1175,7 +1175,7 @@ static void execlists_submit_request(struct i915_request *request)
 	GEM_BUG_ON(!engine->execlists.first);
 	GEM_BUG_ON(list_empty(&request->sched.link));
 
-	spin_unlock_irqrestore(&engine->timeline->lock, flags);
+	spin_unlock_irqrestore(&engine->timeline.lock, flags);
 }
 
 static struct i915_request *sched_to_request(struct i915_sched_node *node)
@@ -1191,8 +1191,8 @@ sched_lock_engine(struct i915_sched_node *node, struct intel_engine_cs *locked)
 	GEM_BUG_ON(!locked);
 
 	if (engine != locked) {
-		spin_unlock(&locked->timeline->lock);
-		spin_lock(&engine->timeline->lock);
+		spin_unlock(&locked->timeline.lock);
+		spin_lock(&engine->timeline.lock);
 	}
 
 	return engine;
@@ -1275,7 +1275,7 @@ static void execlists_schedule(struct i915_request *request,
 	}
 
 	engine = request->engine;
-	spin_lock_irq(&engine->timeline->lock);
+	spin_lock_irq(&engine->timeline.lock);
 
 	/* Fifo and depth-first replacement ensure our deps execute before us */
 	list_for_each_entry_safe_reverse(dep, p, &dfs, dfs_link) {
@@ -1299,7 +1299,7 @@ static void execlists_schedule(struct i915_request *request,
 			__submit_queue(engine, prio);
 	}
 
-	spin_unlock_irq(&engine->timeline->lock);
+	spin_unlock_irq(&engine->timeline.lock);
 }
 
 static int __context_pin(struct i915_gem_context *ctx, struct i915_vma *vma)
@@ -1828,9 +1828,9 @@ static void reset_common_ring(struct intel_engine_cs *engine,
 	reset_irq(engine);
 
 	/* Push back any incomplete requests for replay after the reset. */
-	spin_lock(&engine->timeline->lock);
+	spin_lock(&engine->timeline.lock);
 	__unwind_incomplete_requests(engine);
-	spin_unlock(&engine->timeline->lock);
+	spin_unlock(&engine->timeline.lock);
 
 	local_irq_restore(flags);
 
@@ -2599,6 +2599,7 @@ static int execlists_context_deferred_alloc(struct i915_gem_context *ctx,
 	struct i915_vma *vma;
 	uint32_t context_size;
 	struct intel_ring *ring;
+	struct i915_timeline *timeline;
 	int ret;
 
 	if (ce->state)
@@ -2614,8 +2615,8 @@ static int execlists_context_deferred_alloc(struct i915_gem_context *ctx,
 
 	ctx_obj = i915_gem_object_create(ctx->i915, context_size);
 	if (IS_ERR(ctx_obj)) {
-		DRM_DEBUG_DRIVER("Alloc LRC backing obj failed.\n");
-		return PTR_ERR(ctx_obj);
+		ret = PTR_ERR(ctx_obj);
+		goto error_deref_obj;
 	}
 
 	vma = i915_vma_instance(ctx_obj, &ctx->i915->ggtt.base, NULL);
@@ -2624,7 +2625,14 @@ static int execlists_context_deferred_alloc(struct i915_gem_context *ctx,
 		goto error_deref_obj;
 	}
 
-	ring = intel_engine_create_ring(engine, ctx->timeline, ctx->ring_size);
+	timeline = i915_timeline_create(ctx->i915, ctx->name);
+	if (IS_ERR(timeline)) {
+		ret = PTR_ERR(timeline);
+		goto error_deref_obj;
+	}
+
+	ring = intel_engine_create_ring(engine, timeline, ctx->ring_size);
+	i915_timeline_put(timeline);
 	if (IS_ERR(ring)) {
 		ret = PTR_ERR(ring);
 		goto error_deref_obj;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index b73e700..8f19349 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -697,17 +697,17 @@ static void cancel_requests(struct intel_engine_cs *engine)
 	struct i915_request *request;
 	unsigned long flags;
 
-	spin_lock_irqsave(&engine->timeline->lock, flags);
+	spin_lock_irqsave(&engine->timeline.lock, flags);
 
 	/* Mark all submitted requests as skipped. */
-	list_for_each_entry(request, &engine->timeline->requests, link) {
+	list_for_each_entry(request, &engine->timeline.requests, link) {
 		GEM_BUG_ON(!request->global_seqno);
 		if (!i915_request_completed(request))
 			dma_fence_set_error(&request->fence, -EIO);
 	}
 	/* Remaining _unready_ requests will be nop'ed when submitted */
 
-	spin_unlock_irqrestore(&engine->timeline->lock, flags);
+	spin_unlock_irqrestore(&engine->timeline.lock, flags);
 }
 
 static void i9xx_submit_request(struct i915_request *request)
@@ -1118,7 +1118,7 @@ err:
 
 struct intel_ring *
 intel_engine_create_ring(struct intel_engine_cs *engine,
-			 struct i915_gem_timeline *timeline,
+			 struct i915_timeline *timeline,
 			 int size)
 {
 	struct intel_ring *ring;
@@ -1126,7 +1126,7 @@ intel_engine_create_ring(struct intel_engine_cs *engine,
 
 	GEM_BUG_ON(!is_power_of_2(size));
 	GEM_BUG_ON(RING_CTL_SIZE(size) & ~RING_NR_PAGES);
-	GEM_BUG_ON(&timeline->engine[engine->id] == engine->timeline);
+	GEM_BUG_ON(timeline == &engine->timeline);
 	lockdep_assert_held(&engine->i915->drm.struct_mutex);
 
 	ring = kzalloc(sizeof(*ring), GFP_KERNEL);
@@ -1134,7 +1134,7 @@ intel_engine_create_ring(struct intel_engine_cs *engine,
 		return ERR_PTR(-ENOMEM);
 
 	INIT_LIST_HEAD(&ring->request_list);
-	ring->timeline = &timeline->engine[engine->id];
+	ring->timeline = i915_timeline_get(timeline);
 
 	ring->size = size;
 	/* Workaround an erratum on the i830 which causes a hang if
@@ -1165,6 +1165,7 @@ intel_ring_free(struct intel_ring *ring)
 	i915_vma_close(ring->vma);
 	__i915_gem_object_release_unless_active(obj);
 
+	i915_timeline_put(ring->timeline);
 	kfree(ring);
 }
 
@@ -1323,6 +1324,7 @@ static void intel_ring_context_unpin(struct intel_engine_cs *engine,
 static int intel_init_ring_buffer(struct intel_engine_cs *engine)
 {
 	struct intel_ring *ring;
+	struct i915_timeline *timeline;
 	int err;
 
 	intel_engine_setup_common(engine);
@@ -1331,9 +1333,14 @@ static int intel_init_ring_buffer(struct intel_engine_cs *engine)
 	if (err)
 		goto err;
 
-	ring = intel_engine_create_ring(engine,
-					&engine->i915->gt.legacy_timeline,
-					32 * PAGE_SIZE);
+	timeline = i915_timeline_create(engine->i915, engine->name);
+	if (IS_ERR(timeline)) {
+		err = PTR_ERR(timeline);
+		goto err;
+	}
+
+	ring = intel_engine_create_ring(engine, timeline, 32 * PAGE_SIZE);
+	i915_timeline_put(timeline);
 	if (IS_ERR(ring)) {
 		err = PTR_ERR(ring);
 		goto err;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index da53aa2..010750e 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -6,12 +6,12 @@
 #include <linux/seqlock.h>
 
 #include "i915_gem_batch_pool.h"
-#include "i915_gem_timeline.h"
 
 #include "i915_reg.h"
 #include "i915_pmu.h"
 #include "i915_request.h"
 #include "i915_selftest.h"
+#include "i915_timeline.h"
 #include "intel_gpu_commands.h"
 
 struct drm_printer;
@@ -129,7 +129,7 @@ struct intel_ring {
 	struct i915_vma *vma;
 	void *vaddr;
 
-	struct intel_timeline *timeline;
+	struct i915_timeline *timeline;
 	struct list_head request_list;
 	struct list_head active_link;
 
@@ -338,7 +338,8 @@ struct intel_engine_cs {
 	u32 mmio_base;
 
 	struct intel_ring *buffer;
-	struct intel_timeline *timeline;
+
+	struct i915_timeline timeline;
 
 	struct drm_i915_gem_object *default_state;
 
@@ -770,7 +771,7 @@ intel_write_status_page(struct intel_engine_cs *engine, int reg, u32 value)
 
 struct intel_ring *
 intel_engine_create_ring(struct intel_engine_cs *engine,
-			 struct i915_gem_timeline *timeline,
+			 struct i915_timeline *timeline,
 			 int size);
 int intel_ring_pin(struct intel_ring *ring,
 		   struct drm_i915_private *i915,
@@ -889,7 +890,7 @@ static inline u32 intel_engine_last_submit(struct intel_engine_cs *engine)
 	 * wtih serialising this hint with anything, so document it as
 	 * a hint and nothing more.
 	 */
-	return READ_ONCE(engine->timeline->seqno);
+	return READ_ONCE(engine->timeline.seqno);
 }
 
 void intel_engine_get_instdone(struct intel_engine_cs *engine,
diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_context.c b/drivers/gpu/drm/i915/selftests/i915_gem_context.c
index 24ac648..7ecaed5 100644
--- a/drivers/gpu/drm/i915/selftests/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/selftests/i915_gem_context.c
@@ -355,18 +355,6 @@ static int igt_ctx_exec(void *arg)
 
 		if (first_shared_gtt) {
 			ctx = __create_hw_context(i915, file->driver_priv);
-			if (!IS_ERR(ctx) && HAS_EXECLISTS(i915)) {
-				struct i915_gem_timeline *timeline;
-
-				timeline = i915_gem_timeline_create(i915, ctx->name);
-				if (IS_ERR(timeline)) {
-					__destroy_hw_context(ctx, file->driver_priv);
-					ctx = ERR_CAST(timeline);
-				} else {
-					ctx->timeline = timeline;
-				}
-			}
-
 			first_shared_gtt = false;
 		} else {
 			ctx = i915_gem_create_context(i915, file->driver_priv);
diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_timeline.c b/drivers/gpu/drm/i915/selftests/i915_gem_timeline.c
deleted file mode 100644
index 3000e6a..0000000
--- a/drivers/gpu/drm/i915/selftests/i915_gem_timeline.c
+++ /dev/null
@@ -1,299 +0,0 @@
-/*
- * Copyright © 2017 Intel Corporation
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice (including the next
- * paragraph) shall be included in all copies or substantial portions of the
- * Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
- * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
- * IN THE SOFTWARE.
- *
- */
-
-#include "../i915_selftest.h"
-#include "i915_random.h"
-
-#include "mock_gem_device.h"
-#include "mock_timeline.h"
-
-struct __igt_sync {
-	const char *name;
-	u32 seqno;
-	bool expected;
-	bool set;
-};
-
-static int __igt_sync(struct intel_timeline *tl,
-		      u64 ctx,
-		      const struct __igt_sync *p,
-		      const char *name)
-{
-	int ret;
-
-	if (__intel_timeline_sync_is_later(tl, ctx, p->seqno) != p->expected) {
-		pr_err("%s: %s(ctx=%llu, seqno=%u) expected passed %s but failed\n",
-		       name, p->name, ctx, p->seqno, yesno(p->expected));
-		return -EINVAL;
-	}
-
-	if (p->set) {
-		ret = __intel_timeline_sync_set(tl, ctx, p->seqno);
-		if (ret)
-			return ret;
-	}
-
-	return 0;
-}
-
-static int igt_sync(void *arg)
-{
-	const struct __igt_sync pass[] = {
-		{ "unset", 0, false, false },
-		{ "new", 0, false, true },
-		{ "0a", 0, true, true },
-		{ "1a", 1, false, true },
-		{ "1b", 1, true, true },
-		{ "0b", 0, true, false },
-		{ "2a", 2, false, true },
-		{ "4", 4, false, true },
-		{ "INT_MAX", INT_MAX, false, true },
-		{ "INT_MAX-1", INT_MAX-1, true, false },
-		{ "INT_MAX+1", (u32)INT_MAX+1, false, true },
-		{ "INT_MAX", INT_MAX, true, false },
-		{ "UINT_MAX", UINT_MAX, false, true },
-		{ "wrap", 0, false, true },
-		{ "unwrap", UINT_MAX, true, false },
-		{},
-	}, *p;
-	struct intel_timeline *tl;
-	int order, offset;
-	int ret = -ENODEV;
-
-	tl = mock_timeline(0);
-	if (!tl)
-		return -ENOMEM;
-
-	for (p = pass; p->name; p++) {
-		for (order = 1; order < 64; order++) {
-			for (offset = -1; offset <= (order > 1); offset++) {
-				u64 ctx = BIT_ULL(order) + offset;
-
-				ret = __igt_sync(tl, ctx, p, "1");
-				if (ret)
-					goto out;
-			}
-		}
-	}
-	mock_timeline_destroy(tl);
-
-	tl = mock_timeline(0);
-	if (!tl)
-		return -ENOMEM;
-
-	for (order = 1; order < 64; order++) {
-		for (offset = -1; offset <= (order > 1); offset++) {
-			u64 ctx = BIT_ULL(order) + offset;
-
-			for (p = pass; p->name; p++) {
-				ret = __igt_sync(tl, ctx, p, "2");
-				if (ret)
-					goto out;
-			}
-		}
-	}
-
-out:
-	mock_timeline_destroy(tl);
-	return ret;
-}
-
-static unsigned int random_engine(struct rnd_state *rnd)
-{
-	return i915_prandom_u32_max_state(I915_NUM_ENGINES, rnd);
-}
-
-static int bench_sync(void *arg)
-{
-	struct rnd_state prng;
-	struct intel_timeline *tl;
-	unsigned long end_time, count;
-	u64 prng32_1M;
-	ktime_t kt;
-	int order, last_order;
-
-	tl = mock_timeline(0);
-	if (!tl)
-		return -ENOMEM;
-
-	/* Lookups from cache are very fast and so the random number generation
-	 * and the loop itself becomes a significant factor in the per-iteration
-	 * timings. We try to compensate the results by measuring the overhead
-	 * of the prng and subtract it from the reported results.
-	 */
-	prandom_seed_state(&prng, i915_selftest.random_seed);
-	count = 0;
-	kt = ktime_get();
-	end_time = jiffies + HZ/10;
-	do {
-		u32 x;
-
-		/* Make sure the compiler doesn't optimise away the prng call */
-		WRITE_ONCE(x, prandom_u32_state(&prng));
-
-		count++;
-	} while (!time_after(jiffies, end_time));
-	kt = ktime_sub(ktime_get(), kt);
-	pr_debug("%s: %lu random evaluations, %lluns/prng\n",
-		 __func__, count, (long long)div64_ul(ktime_to_ns(kt), count));
-	prng32_1M = div64_ul(ktime_to_ns(kt) << 20, count);
-
-	/* Benchmark (only) setting random context ids */
-	prandom_seed_state(&prng, i915_selftest.random_seed);
-	count = 0;
-	kt = ktime_get();
-	end_time = jiffies + HZ/10;
-	do {
-		u64 id = i915_prandom_u64_state(&prng);
-
-		__intel_timeline_sync_set(tl, id, 0);
-		count++;
-	} while (!time_after(jiffies, end_time));
-	kt = ktime_sub(ktime_get(), kt);
-	kt = ktime_sub_ns(kt, (count * prng32_1M * 2) >> 20);
-	pr_info("%s: %lu random insertions, %lluns/insert\n",
-		__func__, count, (long long)div64_ul(ktime_to_ns(kt), count));
-
-	/* Benchmark looking up the exact same context ids as we just set */
-	prandom_seed_state(&prng, i915_selftest.random_seed);
-	end_time = count;
-	kt = ktime_get();
-	while (end_time--) {
-		u64 id = i915_prandom_u64_state(&prng);
-
-		if (!__intel_timeline_sync_is_later(tl, id, 0)) {
-			mock_timeline_destroy(tl);
-			pr_err("Lookup of %llu failed\n", id);
-			return -EINVAL;
-		}
-	}
-	kt = ktime_sub(ktime_get(), kt);
-	kt = ktime_sub_ns(kt, (count * prng32_1M * 2) >> 20);
-	pr_info("%s: %lu random lookups, %lluns/lookup\n",
-		__func__, count, (long long)div64_ul(ktime_to_ns(kt), count));
-
-	mock_timeline_destroy(tl);
-	cond_resched();
-
-	tl = mock_timeline(0);
-	if (!tl)
-		return -ENOMEM;
-
-	/* Benchmark setting the first N (in order) contexts */
-	count = 0;
-	kt = ktime_get();
-	end_time = jiffies + HZ/10;
-	do {
-		__intel_timeline_sync_set(tl, count++, 0);
-	} while (!time_after(jiffies, end_time));
-	kt = ktime_sub(ktime_get(), kt);
-	pr_info("%s: %lu in-order insertions, %lluns/insert\n",
-		__func__, count, (long long)div64_ul(ktime_to_ns(kt), count));
-
-	/* Benchmark looking up the exact same context ids as we just set */
-	end_time = count;
-	kt = ktime_get();
-	while (end_time--) {
-		if (!__intel_timeline_sync_is_later(tl, end_time, 0)) {
-			pr_err("Lookup of %lu failed\n", end_time);
-			mock_timeline_destroy(tl);
-			return -EINVAL;
-		}
-	}
-	kt = ktime_sub(ktime_get(), kt);
-	pr_info("%s: %lu in-order lookups, %lluns/lookup\n",
-		__func__, count, (long long)div64_ul(ktime_to_ns(kt), count));
-
-	mock_timeline_destroy(tl);
-	cond_resched();
-
-	tl = mock_timeline(0);
-	if (!tl)
-		return -ENOMEM;
-
-	/* Benchmark searching for a random context id and maybe changing it */
-	prandom_seed_state(&prng, i915_selftest.random_seed);
-	count = 0;
-	kt = ktime_get();
-	end_time = jiffies + HZ/10;
-	do {
-		u32 id = random_engine(&prng);
-		u32 seqno = prandom_u32_state(&prng);
-
-		if (!__intel_timeline_sync_is_later(tl, id, seqno))
-			__intel_timeline_sync_set(tl, id, seqno);
-
-		count++;
-	} while (!time_after(jiffies, end_time));
-	kt = ktime_sub(ktime_get(), kt);
-	kt = ktime_sub_ns(kt, (count * prng32_1M * 2) >> 20);
-	pr_info("%s: %lu repeated insert/lookups, %lluns/op\n",
-		__func__, count, (long long)div64_ul(ktime_to_ns(kt), count));
-	mock_timeline_destroy(tl);
-	cond_resched();
-
-	/* Benchmark searching for a known context id and changing the seqno */
-	for (last_order = 1, order = 1; order < 32;
-	     ({ int tmp = last_order; last_order = order; order += tmp; })) {
-		unsigned int mask = BIT(order) - 1;
-
-		tl = mock_timeline(0);
-		if (!tl)
-			return -ENOMEM;
-
-		count = 0;
-		kt = ktime_get();
-		end_time = jiffies + HZ/10;
-		do {
-			/* Without assuming too many details of the underlying
-			 * implementation, try to identify its phase-changes
-			 * (if any)!
-			 */
-			u64 id = (u64)(count & mask) << order;
-
-			__intel_timeline_sync_is_later(tl, id, 0);
-			__intel_timeline_sync_set(tl, id, 0);
-
-			count++;
-		} while (!time_after(jiffies, end_time));
-		kt = ktime_sub(ktime_get(), kt);
-		pr_info("%s: %lu cyclic/%d insert/lookups, %lluns/op\n",
-			__func__, count, order,
-			(long long)div64_ul(ktime_to_ns(kt), count));
-		mock_timeline_destroy(tl);
-		cond_resched();
-	}
-
-	return 0;
-}
-
-int i915_gem_timeline_mock_selftests(void)
-{
-	static const struct i915_subtest tests[] = {
-		SUBTEST(igt_sync),
-		SUBTEST(bench_sync),
-	};
-
-	return i915_subtests(tests, NULL);
-}
diff --git a/drivers/gpu/drm/i915/selftests/i915_timeline.c b/drivers/gpu/drm/i915/selftests/i915_timeline.c
new file mode 100644
index 0000000..19f1c6a
--- /dev/null
+++ b/drivers/gpu/drm/i915/selftests/i915_timeline.c
@@ -0,0 +1,267 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2017-2018 Intel Corporation
+ */
+
+#include "../i915_selftest.h"
+#include "i915_random.h"
+
+#include "mock_gem_device.h"
+#include "mock_timeline.h"
+
+struct __igt_sync {
+	const char *name;
+	u32 seqno;
+	bool expected;
+	bool set;
+};
+
+static int __igt_sync(struct i915_timeline *tl,
+		      u64 ctx,
+		      const struct __igt_sync *p,
+		      const char *name)
+{
+	int ret;
+
+	if (__i915_timeline_sync_is_later(tl, ctx, p->seqno) != p->expected) {
+		pr_err("%s: %s(ctx=%llu, seqno=%u) expected passed %s but failed\n",
+		       name, p->name, ctx, p->seqno, yesno(p->expected));
+		return -EINVAL;
+	}
+
+	if (p->set) {
+		ret = __i915_timeline_sync_set(tl, ctx, p->seqno);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
+static int igt_sync(void *arg)
+{
+	const struct __igt_sync pass[] = {
+		{ "unset", 0, false, false },
+		{ "new", 0, false, true },
+		{ "0a", 0, true, true },
+		{ "1a", 1, false, true },
+		{ "1b", 1, true, true },
+		{ "0b", 0, true, false },
+		{ "2a", 2, false, true },
+		{ "4", 4, false, true },
+		{ "INT_MAX", INT_MAX, false, true },
+		{ "INT_MAX-1", INT_MAX-1, true, false },
+		{ "INT_MAX+1", (u32)INT_MAX+1, false, true },
+		{ "INT_MAX", INT_MAX, true, false },
+		{ "UINT_MAX", UINT_MAX, false, true },
+		{ "wrap", 0, false, true },
+		{ "unwrap", UINT_MAX, true, false },
+		{},
+	}, *p;
+	struct i915_timeline tl;
+	int order, offset;
+	int ret = -ENODEV;
+
+	mock_timeline_init(&tl, 0);
+	for (p = pass; p->name; p++) {
+		for (order = 1; order < 64; order++) {
+			for (offset = -1; offset <= (order > 1); offset++) {
+				u64 ctx = BIT_ULL(order) + offset;
+
+				ret = __igt_sync(&tl, ctx, p, "1");
+				if (ret)
+					goto out;
+			}
+		}
+	}
+	mock_timeline_fini(&tl);
+
+	mock_timeline_init(&tl, 0);
+	for (order = 1; order < 64; order++) {
+		for (offset = -1; offset <= (order > 1); offset++) {
+			u64 ctx = BIT_ULL(order) + offset;
+
+			for (p = pass; p->name; p++) {
+				ret = __igt_sync(&tl, ctx, p, "2");
+				if (ret)
+					goto out;
+			}
+		}
+	}
+
+out:
+	mock_timeline_fini(&tl);
+	return ret;
+}
+
+static unsigned int random_engine(struct rnd_state *rnd)
+{
+	return i915_prandom_u32_max_state(I915_NUM_ENGINES, rnd);
+}
+
+static int bench_sync(void *arg)
+{
+	struct rnd_state prng;
+	struct i915_timeline tl;
+	unsigned long end_time, count;
+	u64 prng32_1M;
+	ktime_t kt;
+	int order, last_order;
+
+	mock_timeline_init(&tl, 0);
+
+	/* Lookups from cache are very fast and so the random number generation
+	 * and the loop itself becomes a significant factor in the per-iteration
+	 * timings. We try to compensate the results by measuring the overhead
+	 * of the prng and subtract it from the reported results.
+	 */
+	prandom_seed_state(&prng, i915_selftest.random_seed);
+	count = 0;
+	kt = ktime_get();
+	end_time = jiffies + HZ/10;
+	do {
+		u32 x;
+
+		/* Make sure the compiler doesn't optimise away the prng call */
+		WRITE_ONCE(x, prandom_u32_state(&prng));
+
+		count++;
+	} while (!time_after(jiffies, end_time));
+	kt = ktime_sub(ktime_get(), kt);
+	pr_debug("%s: %lu random evaluations, %lluns/prng\n",
+		 __func__, count, (long long)div64_ul(ktime_to_ns(kt), count));
+	prng32_1M = div64_ul(ktime_to_ns(kt) << 20, count);
+
+	/* Benchmark (only) setting random context ids */
+	prandom_seed_state(&prng, i915_selftest.random_seed);
+	count = 0;
+	kt = ktime_get();
+	end_time = jiffies + HZ/10;
+	do {
+		u64 id = i915_prandom_u64_state(&prng);
+
+		__i915_timeline_sync_set(&tl, id, 0);
+		count++;
+	} while (!time_after(jiffies, end_time));
+	kt = ktime_sub(ktime_get(), kt);
+	kt = ktime_sub_ns(kt, (count * prng32_1M * 2) >> 20);
+	pr_info("%s: %lu random insertions, %lluns/insert\n",
+		__func__, count, (long long)div64_ul(ktime_to_ns(kt), count));
+
+	/* Benchmark looking up the exact same context ids as we just set */
+	prandom_seed_state(&prng, i915_selftest.random_seed);
+	end_time = count;
+	kt = ktime_get();
+	while (end_time--) {
+		u64 id = i915_prandom_u64_state(&prng);
+
+		if (!__i915_timeline_sync_is_later(&tl, id, 0)) {
+			mock_timeline_fini(&tl);
+			pr_err("Lookup of %llu failed\n", id);
+			return -EINVAL;
+		}
+	}
+	kt = ktime_sub(ktime_get(), kt);
+	kt = ktime_sub_ns(kt, (count * prng32_1M * 2) >> 20);
+	pr_info("%s: %lu random lookups, %lluns/lookup\n",
+		__func__, count, (long long)div64_ul(ktime_to_ns(kt), count));
+
+	mock_timeline_fini(&tl);
+	cond_resched();
+
+	mock_timeline_init(&tl, 0);
+
+	/* Benchmark setting the first N (in order) contexts */
+	count = 0;
+	kt = ktime_get();
+	end_time = jiffies + HZ/10;
+	do {
+		__i915_timeline_sync_set(&tl, count++, 0);
+	} while (!time_after(jiffies, end_time));
+	kt = ktime_sub(ktime_get(), kt);
+	pr_info("%s: %lu in-order insertions, %lluns/insert\n",
+		__func__, count, (long long)div64_ul(ktime_to_ns(kt), count));
+
+	/* Benchmark looking up the exact same context ids as we just set */
+	end_time = count;
+	kt = ktime_get();
+	while (end_time--) {
+		if (!__i915_timeline_sync_is_later(&tl, end_time, 0)) {
+			pr_err("Lookup of %lu failed\n", end_time);
+			mock_timeline_fini(&tl);
+			return -EINVAL;
+		}
+	}
+	kt = ktime_sub(ktime_get(), kt);
+	pr_info("%s: %lu in-order lookups, %lluns/lookup\n",
+		__func__, count, (long long)div64_ul(ktime_to_ns(kt), count));
+
+	mock_timeline_fini(&tl);
+	cond_resched();
+
+	mock_timeline_init(&tl, 0);
+
+	/* Benchmark searching for a random context id and maybe changing it */
+	prandom_seed_state(&prng, i915_selftest.random_seed);
+	count = 0;
+	kt = ktime_get();
+	end_time = jiffies + HZ/10;
+	do {
+		u32 id = random_engine(&prng);
+		u32 seqno = prandom_u32_state(&prng);
+
+		if (!__i915_timeline_sync_is_later(&tl, id, seqno))
+			__i915_timeline_sync_set(&tl, id, seqno);
+
+		count++;
+	} while (!time_after(jiffies, end_time));
+	kt = ktime_sub(ktime_get(), kt);
+	kt = ktime_sub_ns(kt, (count * prng32_1M * 2) >> 20);
+	pr_info("%s: %lu repeated insert/lookups, %lluns/op\n",
+		__func__, count, (long long)div64_ul(ktime_to_ns(kt), count));
+	mock_timeline_fini(&tl);
+	cond_resched();
+
+	/* Benchmark searching for a known context id and changing the seqno */
+	for (last_order = 1, order = 1; order < 32;
+	     ({ int tmp = last_order; last_order = order; order += tmp; })) {
+		unsigned int mask = BIT(order) - 1;
+
+		mock_timeline_init(&tl, 0);
+
+		count = 0;
+		kt = ktime_get();
+		end_time = jiffies + HZ/10;
+		do {
+			/* Without assuming too many details of the underlying
+			 * implementation, try to identify its phase-changes
+			 * (if any)!
+			 */
+			u64 id = (u64)(count & mask) << order;
+
+			__i915_timeline_sync_is_later(&tl, id, 0);
+			__i915_timeline_sync_set(&tl, id, 0);
+
+			count++;
+		} while (!time_after(jiffies, end_time));
+		kt = ktime_sub(ktime_get(), kt);
+		pr_info("%s: %lu cyclic/%d insert/lookups, %lluns/op\n",
+			__func__, count, order,
+			(long long)div64_ul(ktime_to_ns(kt), count));
+		mock_timeline_fini(&tl);
+		cond_resched();
+	}
+
+	return 0;
+}
+
+int i915_gem_timeline_mock_selftests(void)
+{
+	static const struct i915_subtest tests[] = {
+		SUBTEST(igt_sync),
+		SUBTEST(bench_sync),
+	};
+
+	return i915_subtests(tests, NULL);
+}
diff --git a/drivers/gpu/drm/i915/selftests/mock_engine.c b/drivers/gpu/drm/i915/selftests/mock_engine.c
index 6752498..26bf29d 100644
--- a/drivers/gpu/drm/i915/selftests/mock_engine.c
+++ b/drivers/gpu/drm/i915/selftests/mock_engine.c
@@ -25,6 +25,11 @@
 #include "mock_engine.h"
 #include "mock_request.h"
 
+struct mock_ring {
+	struct intel_ring base;
+	struct i915_timeline timeline;
+};
+
 static struct mock_request *first_request(struct mock_engine *engine)
 {
 	return list_first_entry_or_null(&engine->hw_queue,
@@ -132,7 +137,7 @@ static void mock_submit_request(struct i915_request *request)
 static struct intel_ring *mock_ring(struct intel_engine_cs *engine)
 {
 	const unsigned long sz = PAGE_SIZE / 2;
-	struct intel_ring *ring;
+	struct mock_ring *ring;
 
 	BUILD_BUG_ON(MIN_SPACE_FOR_ADD_REQUEST > sz);
 
@@ -140,20 +145,24 @@ static struct intel_ring *mock_ring(struct intel_engine_cs *engine)
 	if (!ring)
 		return NULL;
 
-	ring->timeline = &engine->i915->gt.legacy_timeline.engine[engine->id];
+	i915_timeline_init(engine->i915, &ring->timeline, engine->name);
 
-	ring->size = sz;
-	ring->effective_size = sz;
-	ring->vaddr = (void *)(ring + 1);
+	ring->base.size = sz;
+	ring->base.effective_size = sz;
+	ring->base.vaddr = (void *)(ring + 1);
+	ring->base.timeline = &ring->timeline;
 
-	INIT_LIST_HEAD(&ring->request_list);
-	intel_ring_update_space(ring);
+	INIT_LIST_HEAD(&ring->base.request_list);
+	intel_ring_update_space(&ring->base);
 
-	return ring;
+	return &ring->base;
 }
 
-static void mock_ring_free(struct intel_ring *ring)
+static void mock_ring_free(struct intel_ring *base)
 {
+	struct mock_ring *ring = container_of(base, typeof(*ring), base);
+
+	i915_timeline_fini(&ring->timeline);
 	kfree(ring);
 }
 
@@ -182,8 +191,7 @@ struct intel_engine_cs *mock_engine(struct drm_i915_private *i915,
 	engine->base.emit_breadcrumb = mock_emit_breadcrumb;
 	engine->base.submit_request = mock_submit_request;
 
-	intel_engine_init_timeline(&engine->base);
-
+	i915_timeline_init(i915, &engine->base.timeline, engine->base.name);
 	intel_engine_init_breadcrumbs(&engine->base);
 	engine->base.breadcrumbs.mock = true; /* prevent touching HW for irqs */
 
@@ -200,6 +208,7 @@ struct intel_engine_cs *mock_engine(struct drm_i915_private *i915,
 
 err_breadcrumbs:
 	intel_engine_fini_breadcrumbs(&engine->base);
+	i915_timeline_fini(&engine->base.timeline);
 	kfree(engine);
 	return NULL;
 }
@@ -238,6 +247,7 @@ void mock_engine_free(struct intel_engine_cs *engine)
 	mock_ring_free(engine->buffer);
 
 	intel_engine_fini_breadcrumbs(engine);
+	i915_timeline_fini(&engine->timeline);
 
 	kfree(engine);
 }
diff --git a/drivers/gpu/drm/i915/selftests/mock_gem_device.c b/drivers/gpu/drm/i915/selftests/mock_gem_device.c
index f11c83e..a662c04 100644
--- a/drivers/gpu/drm/i915/selftests/mock_gem_device.c
+++ b/drivers/gpu/drm/i915/selftests/mock_gem_device.c
@@ -73,10 +73,8 @@ static void mock_device_release(struct drm_device *dev)
 
 	mutex_lock(&i915->drm.struct_mutex);
 	mock_fini_ggtt(i915);
-	i915_gem_timeline_fini(&i915->gt.legacy_timeline);
-	i915_gem_timeline_fini(&i915->gt.execution_timeline);
-	WARN_ON(!list_empty(&i915->gt.timelines));
 	mutex_unlock(&i915->drm.struct_mutex);
+	WARN_ON(!list_empty(&i915->gt.timelines));
 
 	destroy_workqueue(i915->wq);
 
@@ -230,12 +228,6 @@ struct drm_i915_private *mock_gem_device(void)
 	INIT_LIST_HEAD(&i915->gt.active_rings);
 
 	mutex_lock(&i915->drm.struct_mutex);
-	err = i915_gem_timeline_init__global(i915);
-	if (err) {
-		mutex_unlock(&i915->drm.struct_mutex);
-		goto err_priorities;
-	}
-
 	mock_init_ggtt(i915);
 	mutex_unlock(&i915->drm.struct_mutex);
 
diff --git a/drivers/gpu/drm/i915/selftests/mock_timeline.c b/drivers/gpu/drm/i915/selftests/mock_timeline.c
index 47b1f47..dcf3b16 100644
--- a/drivers/gpu/drm/i915/selftests/mock_timeline.c
+++ b/drivers/gpu/drm/i915/selftests/mock_timeline.c
@@ -1,45 +1,28 @@
 /*
- * Copyright © 2017 Intel Corporation
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice (including the next
- * paragraph) shall be included in all copies or substantial portions of the
- * Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
- * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
- * IN THE SOFTWARE.
+ * SPDX-License-Identifier: MIT
  *
+ * Copyright © 2017-2018 Intel Corporation
  */
 
+#include "../i915_timeline.h"
+
 #include "mock_timeline.h"
 
-struct intel_timeline *mock_timeline(u64 context)
+void mock_timeline_init(struct i915_timeline *timeline, u64 context)
 {
-	static struct lock_class_key class;
-	struct intel_timeline *tl;
+	timeline->fence_context = context;
+
+	spin_lock_init(&timeline->lock);
 
-	tl = kzalloc(sizeof(*tl), GFP_KERNEL);
-	if (!tl)
-		return NULL;
+	init_request_active(&timeline->last_request, NULL);
+	INIT_LIST_HEAD(&timeline->requests);
 
-	__intel_timeline_init(tl, NULL, context, &class, "mock");
+	i915_syncmap_init(&timeline->sync);
 
-	return tl;
+	INIT_LIST_HEAD(&timeline->link);
 }
 
-void mock_timeline_destroy(struct intel_timeline *tl)
+void mock_timeline_fini(struct i915_timeline *timeline)
 {
-	__intel_timeline_fini(tl);
-	kfree(tl);
+	i915_timeline_fini(timeline);
 }
diff --git a/drivers/gpu/drm/i915/selftests/mock_timeline.h b/drivers/gpu/drm/i915/selftests/mock_timeline.h
index c27ff46..b6deaa6 100644
--- a/drivers/gpu/drm/i915/selftests/mock_timeline.h
+++ b/drivers/gpu/drm/i915/selftests/mock_timeline.h
@@ -1,33 +1,15 @@
 /*
- * Copyright © 2017 Intel Corporation
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice (including the next
- * paragraph) shall be included in all copies or substantial portions of the
- * Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
- * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
- * IN THE SOFTWARE.
+ * SPDX-License-Identifier: MIT
  *
+ * Copyright © 2017-2018 Intel Corporation
  */
 
 #ifndef __MOCK_TIMELINE__
 #define __MOCK_TIMELINE__
 
-#include "../i915_gem_timeline.h"
+struct i915_timeline;
 
-struct intel_timeline *mock_timeline(u64 context);
-void mock_timeline_destroy(struct intel_timeline *tl);
+void mock_timeline_init(struct i915_timeline *timeline, u64 context);
+void mock_timeline_fini(struct i915_timeline *timeline);
 
 #endif /* !__MOCK_TIMELINE__ */
-- 
cgit v1.1


From 94720e3aee6884d8c8beb678001629da60ec6366 Mon Sep 17 00:00:00 2001
From: Julian Anastasov <ja@ssi.bg>
Date: Wed, 2 May 2018 09:41:19 +0300
Subject: ipv4: fix fnhe usage by non-cached routes

Allow some non-cached routes to use non-expired fnhe:

1. ip_del_fnhe: moved above and now called by find_exception.
The 4.5+ commit deed49df7390 expires fnhe only when caching
routes. Change that to:

1.1. use fnhe for non-cached local output routes, with the help
from (2)

1.2. allow __mkroute_input to detect expired fnhe (outdated
fnhe_gw, for example) when do_cache is false, eg. when itag!=0
for unicast destinations.

2. __mkroute_output: keep fi to allow local routes with orig_oif != 0
to use fnhe info even when the new route will not be cached into fnhe.
After commit 839da4d98960 ("net: ipv4: set orig_oif based on fib
result for local traffic") it means all local routes will be affected
because they are not cached. This change is used to solve a PMTU
problem with IPVS (and probably Netfilter DNAT) setups that redirect
local clients from target local IP (local route to Virtual IP)
to new remote IP target, eg. IPVS TUN real server. Loopback has
64K MTU and we need to create fnhe on the local route that will
keep the reduced PMTU for the Virtual IP. Without this change
fnhe_pmtu is updated from ICMP but never exposed to non-cached
local routes. This includes routes with flowi4_oif!=0 for 4.6+ and
with flowi4_oif=any for 4.14+).

3. update_or_create_fnhe: make sure fnhe_expires is not 0 for
new entries

Fixes: 839da4d98960 ("net: ipv4: set orig_oif based on fib result for local traffic")
Fixes: d6d5e999e5df ("route: do not cache fib route info on local routes with oif")
Fixes: deed49df7390 ("route: check and remove route cache when we get route")
Cc: David Ahern <dsahern@gmail.com>
Cc: Xin Long <lucien.xin@gmail.com>
Signed-off-by: Julian Anastasov <ja@ssi.bg>
Acked-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/ipv4/route.c | 118 +++++++++++++++++++++++++------------------------------
 1 file changed, 53 insertions(+), 65 deletions(-)

diff --git a/net/ipv4/route.c b/net/ipv4/route.c
index ccb25d8..1412a7b 100644
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -709,7 +709,7 @@ static void update_or_create_fnhe(struct fib_nh *nh, __be32 daddr, __be32 gw,
 		fnhe->fnhe_gw = gw;
 		fnhe->fnhe_pmtu = pmtu;
 		fnhe->fnhe_mtu_locked = lock;
-		fnhe->fnhe_expires = expires;
+		fnhe->fnhe_expires = max(1UL, expires);
 
 		/* Exception created; mark the cached routes for the nexthop
 		 * stale, so anyone caching it rechecks if this exception
@@ -1297,6 +1297,36 @@ static unsigned int ipv4_mtu(const struct dst_entry *dst)
 	return mtu - lwtunnel_headroom(dst->lwtstate, mtu);
 }
 
+static void ip_del_fnhe(struct fib_nh *nh, __be32 daddr)
+{
+	struct fnhe_hash_bucket *hash;
+	struct fib_nh_exception *fnhe, __rcu **fnhe_p;
+	u32 hval = fnhe_hashfun(daddr);
+
+	spin_lock_bh(&fnhe_lock);
+
+	hash = rcu_dereference_protected(nh->nh_exceptions,
+					 lockdep_is_held(&fnhe_lock));
+	hash += hval;
+
+	fnhe_p = &hash->chain;
+	fnhe = rcu_dereference_protected(*fnhe_p, lockdep_is_held(&fnhe_lock));
+	while (fnhe) {
+		if (fnhe->fnhe_daddr == daddr) {
+			rcu_assign_pointer(*fnhe_p, rcu_dereference_protected(
+				fnhe->fnhe_next, lockdep_is_held(&fnhe_lock)));
+			fnhe_flush_routes(fnhe);
+			kfree_rcu(fnhe, rcu);
+			break;
+		}
+		fnhe_p = &fnhe->fnhe_next;
+		fnhe = rcu_dereference_protected(fnhe->fnhe_next,
+						 lockdep_is_held(&fnhe_lock));
+	}
+
+	spin_unlock_bh(&fnhe_lock);
+}
+
 static struct fib_nh_exception *find_exception(struct fib_nh *nh, __be32 daddr)
 {
 	struct fnhe_hash_bucket *hash = rcu_dereference(nh->nh_exceptions);
@@ -1310,8 +1340,14 @@ static struct fib_nh_exception *find_exception(struct fib_nh *nh, __be32 daddr)
 
 	for (fnhe = rcu_dereference(hash[hval].chain); fnhe;
 	     fnhe = rcu_dereference(fnhe->fnhe_next)) {
-		if (fnhe->fnhe_daddr == daddr)
+		if (fnhe->fnhe_daddr == daddr) {
+			if (fnhe->fnhe_expires &&
+			    time_after(jiffies, fnhe->fnhe_expires)) {
+				ip_del_fnhe(nh, daddr);
+				break;
+			}
 			return fnhe;
+		}
 	}
 	return NULL;
 }
@@ -1636,36 +1672,6 @@ static void ip_handle_martian_source(struct net_device *dev,
 #endif
 }
 
-static void ip_del_fnhe(struct fib_nh *nh, __be32 daddr)
-{
-	struct fnhe_hash_bucket *hash;
-	struct fib_nh_exception *fnhe, __rcu **fnhe_p;
-	u32 hval = fnhe_hashfun(daddr);
-
-	spin_lock_bh(&fnhe_lock);
-
-	hash = rcu_dereference_protected(nh->nh_exceptions,
-					 lockdep_is_held(&fnhe_lock));
-	hash += hval;
-
-	fnhe_p = &hash->chain;
-	fnhe = rcu_dereference_protected(*fnhe_p, lockdep_is_held(&fnhe_lock));
-	while (fnhe) {
-		if (fnhe->fnhe_daddr == daddr) {
-			rcu_assign_pointer(*fnhe_p, rcu_dereference_protected(
-				fnhe->fnhe_next, lockdep_is_held(&fnhe_lock)));
-			fnhe_flush_routes(fnhe);
-			kfree_rcu(fnhe, rcu);
-			break;
-		}
-		fnhe_p = &fnhe->fnhe_next;
-		fnhe = rcu_dereference_protected(fnhe->fnhe_next,
-						 lockdep_is_held(&fnhe_lock));
-	}
-
-	spin_unlock_bh(&fnhe_lock);
-}
-
 /* called in rcu_read_lock() section */
 static int __mkroute_input(struct sk_buff *skb,
 			   const struct fib_result *res,
@@ -1719,20 +1725,10 @@ static int __mkroute_input(struct sk_buff *skb,
 
 	fnhe = find_exception(&FIB_RES_NH(*res), daddr);
 	if (do_cache) {
-		if (fnhe) {
+		if (fnhe)
 			rth = rcu_dereference(fnhe->fnhe_rth_input);
-			if (rth && rth->dst.expires &&
-			    time_after(jiffies, rth->dst.expires)) {
-				ip_del_fnhe(&FIB_RES_NH(*res), daddr);
-				fnhe = NULL;
-			} else {
-				goto rt_cache;
-			}
-		}
-
-		rth = rcu_dereference(FIB_RES_NH(*res).nh_rth_input);
-
-rt_cache:
+		else
+			rth = rcu_dereference(FIB_RES_NH(*res).nh_rth_input);
 		if (rt_cache_valid(rth)) {
 			skb_dst_set_noref(skb, &rth->dst);
 			goto out;
@@ -2216,39 +2212,31 @@ static struct rtable *__mkroute_output(const struct fib_result *res,
 		 * the loopback interface and the IP_PKTINFO ipi_ifindex will
 		 * be set to the loopback interface as well.
 		 */
-		fi = NULL;
+		do_cache = false;
 	}
 
 	fnhe = NULL;
 	do_cache &= fi != NULL;
-	if (do_cache) {
+	if (fi) {
 		struct rtable __rcu **prth;
 		struct fib_nh *nh = &FIB_RES_NH(*res);
 
 		fnhe = find_exception(nh, fl4->daddr);
+		if (!do_cache)
+			goto add;
 		if (fnhe) {
 			prth = &fnhe->fnhe_rth_output;
-			rth = rcu_dereference(*prth);
-			if (rth && rth->dst.expires &&
-			    time_after(jiffies, rth->dst.expires)) {
-				ip_del_fnhe(nh, fl4->daddr);
-				fnhe = NULL;
-			} else {
-				goto rt_cache;
+		} else {
+			if (unlikely(fl4->flowi4_flags &
+				     FLOWI_FLAG_KNOWN_NH &&
+				     !(nh->nh_gw &&
+				       nh->nh_scope == RT_SCOPE_LINK))) {
+				do_cache = false;
+				goto add;
 			}
+			prth = raw_cpu_ptr(nh->nh_pcpu_rth_output);
 		}
-
-		if (unlikely(fl4->flowi4_flags &
-			     FLOWI_FLAG_KNOWN_NH &&
-			     !(nh->nh_gw &&
-			       nh->nh_scope == RT_SCOPE_LINK))) {
-			do_cache = false;
-			goto add;
-		}
-		prth = raw_cpu_ptr(nh->nh_pcpu_rth_output);
 		rth = rcu_dereference(*prth);
-
-rt_cache:
 		if (rt_cache_valid(rth) && dst_hold_safe(&rth->dst))
 			return rth;
 	}
-- 
cgit v1.1


From 0b26351b910fb8fe6a056f8a1bbccabe50c0e19f Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <peterz@infradead.org>
Date: Fri, 20 Apr 2018 11:50:05 +0200
Subject: stop_machine, sched: Fix migrate_swap() vs. active_balance() deadlock

Matt reported the following deadlock:

CPU0					CPU1

schedule(.prev=migrate/0)		<fault>
  pick_next_task()			  ...
    idle_balance()			    migrate_swap()
      active_balance()			      stop_two_cpus()
						spin_lock(stopper0->lock)
						spin_lock(stopper1->lock)
						ttwu(migrate/0)
						  smp_cond_load_acquire() -- waits for schedule()
        stop_one_cpu(1)
	  spin_lock(stopper1->lock) -- waits for stopper lock

Fix this deadlock by taking the wakeups out from under stopper->lock.
This allows the active_balance() to queue the stop work and finish the
context switch, which in turn allows the wakeup from migrate_swap() to
observe the context and complete the wakeup.

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reported-by: Matt Fleming <matt@codeblueprint.co.uk>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Matt Fleming <matt@codeblueprint.co.uk>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Galbraith <umgwanakikbuti@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20180420095005.GH4064@hirez.programming.kicks-ass.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 kernel/stop_machine.c | 19 ++++++++++++++-----
 1 file changed, 14 insertions(+), 5 deletions(-)

diff --git a/kernel/stop_machine.c b/kernel/stop_machine.c
index b759126..64c0291 100644
--- a/kernel/stop_machine.c
+++ b/kernel/stop_machine.c
@@ -21,6 +21,7 @@
 #include <linux/smpboot.h>
 #include <linux/atomic.h>
 #include <linux/nmi.h>
+#include <linux/sched/wake_q.h>
 
 /*
  * Structure to determine completion condition and record errors.  May
@@ -65,27 +66,31 @@ static void cpu_stop_signal_done(struct cpu_stop_done *done)
 }
 
 static void __cpu_stop_queue_work(struct cpu_stopper *stopper,
-					struct cpu_stop_work *work)
+					struct cpu_stop_work *work,
+					struct wake_q_head *wakeq)
 {
 	list_add_tail(&work->list, &stopper->works);
-	wake_up_process(stopper->thread);
+	wake_q_add(wakeq, stopper->thread);
 }
 
 /* queue @work to @stopper.  if offline, @work is completed immediately */
 static bool cpu_stop_queue_work(unsigned int cpu, struct cpu_stop_work *work)
 {
 	struct cpu_stopper *stopper = &per_cpu(cpu_stopper, cpu);
+	DEFINE_WAKE_Q(wakeq);
 	unsigned long flags;
 	bool enabled;
 
 	spin_lock_irqsave(&stopper->lock, flags);
 	enabled = stopper->enabled;
 	if (enabled)
-		__cpu_stop_queue_work(stopper, work);
+		__cpu_stop_queue_work(stopper, work, &wakeq);
 	else if (work->done)
 		cpu_stop_signal_done(work->done);
 	spin_unlock_irqrestore(&stopper->lock, flags);
 
+	wake_up_q(&wakeq);
+
 	return enabled;
 }
 
@@ -229,6 +234,7 @@ static int cpu_stop_queue_two_works(int cpu1, struct cpu_stop_work *work1,
 {
 	struct cpu_stopper *stopper1 = per_cpu_ptr(&cpu_stopper, cpu1);
 	struct cpu_stopper *stopper2 = per_cpu_ptr(&cpu_stopper, cpu2);
+	DEFINE_WAKE_Q(wakeq);
 	int err;
 retry:
 	spin_lock_irq(&stopper1->lock);
@@ -252,8 +258,8 @@ retry:
 			goto unlock;
 
 	err = 0;
-	__cpu_stop_queue_work(stopper1, work1);
-	__cpu_stop_queue_work(stopper2, work2);
+	__cpu_stop_queue_work(stopper1, work1, &wakeq);
+	__cpu_stop_queue_work(stopper2, work2, &wakeq);
 unlock:
 	spin_unlock(&stopper2->lock);
 	spin_unlock_irq(&stopper1->lock);
@@ -263,6 +269,9 @@ unlock:
 			cpu_relax();
 		goto retry;
 	}
+
+	wake_up_q(&wakeq);
+
 	return err;
 }
 /**
-- 
cgit v1.1


From 457be908c83637ee10bda085a23dc05afa3b14a0 Mon Sep 17 00:00:00 2001
From: Vincent Guittot <vincent.guittot@linaro.org>
Date: Thu, 26 Apr 2018 12:19:32 +0200
Subject: sched/fair: Fix the update of blocked load when newly idle

With commit:

  31e77c93e432 ("sched/fair: Update blocked load when newly idle")

... we release the rq->lock when updating blocked load of idle CPUs.

This opens a time window during which another CPU can add a task to this
CPU's cfs_rq.

The check for newly added task of idle_balance() is not in the common path.
Move the out label to include this check.

Reported-by: Heiner Kallweit <hkallweit1@gmail.com>
Tested-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Fixes: 31e77c93e432 ("sched/fair: Update blocked load when newly idle")
Link: http://lkml.kernel.org/r/20180426103133.GA6953@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 kernel/sched/fair.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 54dc31e..e3002e5a 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -9847,6 +9847,7 @@ static int idle_balance(struct rq *this_rq, struct rq_flags *rf)
 	if (curr_cost > this_rq->max_idle_balance_cost)
 		this_rq->max_idle_balance_cost = curr_cost;
 
+out:
 	/*
 	 * While browsing the domains, we released the rq lock, a task could
 	 * have been enqueued in the meantime. Since we're not going idle,
@@ -9855,7 +9856,6 @@ static int idle_balance(struct rq *this_rq, struct rq_flags *rf)
 	if (this_rq->cfs.h_nr_running && !pulled_task)
 		pulled_task = 1;
 
-out:
 	/* Move the next balance forward */
 	if (time_after(this_rq->next_balance, next_balance))
 		this_rq->next_balance = next_balance;
-- 
cgit v1.1


From 741a76b350897604c48fb12beff1c9b77724dc96 Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <peterz@infradead.org>
Date: Mon, 30 Apr 2018 14:50:22 +0200
Subject: kthread, sched/wait: Fix kthread_parkme() wait-loop

Gaurav reported a problem with __kthread_parkme() where a concurrent
try_to_wake_up() could result in competing stores to ->state which,
when the TASK_PARKED store got lost bad things would happen.

The comment near set_current_state() actually mentions this competing
store, but only mentions the case against TASK_RUNNING. This same
store, with different timing, can happen against a subsequent !RUNNING
store.

This normally is not a problem, because as per that same comment, the
!RUNNING state store is inside a condition based wait-loop:

  for (;;) {
    set_current_state(TASK_UNINTERRUPTIBLE);
    if (!need_sleep)
      break;
    schedule();
  }
  __set_current_state(TASK_RUNNING);

If we loose the (first) TASK_UNINTERRUPTIBLE store to a previous
(concurrent) wakeup, the schedule() will NO-OP and we'll go around the
loop once more.

The problem here is that the TASK_PARKED store is not inside the
KTHREAD_SHOULD_PARK condition wait-loop.

There is a genuine issue with sleeps that do not have a condition;
this is addressed in a subsequent patch.

Reported-by: Gaurav Kohli <gkohli@codeaurora.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Oleg Nesterov <oleg@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 kernel/kthread.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/kernel/kthread.c b/kernel/kthread.c
index cd50e99..cbee858 100644
--- a/kernel/kthread.c
+++ b/kernel/kthread.c
@@ -177,12 +177,13 @@ void *kthread_probe_data(struct task_struct *task)
 
 static void __kthread_parkme(struct kthread *self)
 {
-	__set_current_state(TASK_PARKED);
-	while (test_bit(KTHREAD_SHOULD_PARK, &self->flags)) {
+	for (;;) {
+		set_current_state(TASK_PARKED);
+		if (!test_bit(KTHREAD_SHOULD_PARK, &self->flags))
+			break;
 		if (!test_and_set_bit(KTHREAD_IS_PARKED, &self->flags))
 			complete(&self->parked);
 		schedule();
-		__set_current_state(TASK_PARKED);
 	}
 	clear_bit(KTHREAD_IS_PARKED, &self->flags);
 	__set_current_state(TASK_RUNNING);
-- 
cgit v1.1


From 85f1abe0019fcb3ea10df7029056cf42702283a8 Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <peterz@infradead.org>
Date: Tue, 1 May 2018 18:14:45 +0200
Subject: kthread, sched/wait: Fix kthread_parkme() completion issue

Even with the wait-loop fixed, there is a further issue with
kthread_parkme(). Upon hotplug, when we do takedown_cpu(),
smpboot_park_threads() can return before all those threads are in fact
blocked, due to the placement of the complete() in __kthread_parkme().

When that happens, sched_cpu_dying() -> migrate_tasks() can end up
migrating such a still runnable task onto another CPU.

Normally the task will have hit schedule() and gone to sleep by the
time we do kthread_unpark(), which will then do __kthread_bind() to
re-bind the task to the correct CPU.

However, when we loose the initial TASK_PARKED store to the concurrent
wakeup issue described previously, do the complete(), get migrated, it
is possible to either:

 - observe kthread_unpark()'s clearing of SHOULD_PARK and terminate
   the park and set TASK_RUNNING, or

 - __kthread_bind()'s wait_task_inactive() to observe the competing
   TASK_RUNNING store.

Either way the WARN() in __kthread_bind() will trigger and fail to
correctly set the CPU affinity.

Fix this by only issuing the complete() when the kthread has scheduled
out. This does away with all the icky 'still running' nonsense.

The alternative is to promote TASK_PARKED to a special state, this
guarantees wait_task_inactive() cannot observe a 'stale' TASK_RUNNING
and we'll end up doing the right thing, but this preserves the whole
icky business of potentially migating the still runnable thing.

Reported-by: Gaurav Kohli <gkohli@codeaurora.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 include/linux/kthread.h |  1 +
 kernel/kthread.c        | 43 +++++++++++++++++++------------------------
 kernel/sched/core.c     | 32 +++++++++++++++++++++-----------
 3 files changed, 41 insertions(+), 35 deletions(-)

diff --git a/include/linux/kthread.h b/include/linux/kthread.h
index c196176..2803264 100644
--- a/include/linux/kthread.h
+++ b/include/linux/kthread.h
@@ -62,6 +62,7 @@ void *kthread_probe_data(struct task_struct *k);
 int kthread_park(struct task_struct *k);
 void kthread_unpark(struct task_struct *k);
 void kthread_parkme(void);
+void kthread_park_complete(struct task_struct *k);
 
 int kthreadd(void *unused);
 extern struct task_struct *kthreadd_task;
diff --git a/kernel/kthread.c b/kernel/kthread.c
index cbee858..2017a39 100644
--- a/kernel/kthread.c
+++ b/kernel/kthread.c
@@ -55,7 +55,6 @@ enum KTHREAD_BITS {
 	KTHREAD_IS_PER_CPU = 0,
 	KTHREAD_SHOULD_STOP,
 	KTHREAD_SHOULD_PARK,
-	KTHREAD_IS_PARKED,
 };
 
 static inline void set_kthread_struct(void *kthread)
@@ -181,11 +180,8 @@ static void __kthread_parkme(struct kthread *self)
 		set_current_state(TASK_PARKED);
 		if (!test_bit(KTHREAD_SHOULD_PARK, &self->flags))
 			break;
-		if (!test_and_set_bit(KTHREAD_IS_PARKED, &self->flags))
-			complete(&self->parked);
 		schedule();
 	}
-	clear_bit(KTHREAD_IS_PARKED, &self->flags);
 	__set_current_state(TASK_RUNNING);
 }
 
@@ -195,6 +191,11 @@ void kthread_parkme(void)
 }
 EXPORT_SYMBOL_GPL(kthread_parkme);
 
+void kthread_park_complete(struct task_struct *k)
+{
+	complete(&to_kthread(k)->parked);
+}
+
 static int kthread(void *_create)
 {
 	/* Copy data: it's on kthread's stack */
@@ -451,22 +452,15 @@ void kthread_unpark(struct task_struct *k)
 {
 	struct kthread *kthread = to_kthread(k);
 
-	clear_bit(KTHREAD_SHOULD_PARK, &kthread->flags);
 	/*
-	 * We clear the IS_PARKED bit here as we don't wait
-	 * until the task has left the park code. So if we'd
-	 * park before that happens we'd see the IS_PARKED bit
-	 * which might be about to be cleared.
+	 * Newly created kthread was parked when the CPU was offline.
+	 * The binding was lost and we need to set it again.
 	 */
-	if (test_and_clear_bit(KTHREAD_IS_PARKED, &kthread->flags)) {
-		/*
-		 * Newly created kthread was parked when the CPU was offline.
-		 * The binding was lost and we need to set it again.
-		 */
-		if (test_bit(KTHREAD_IS_PER_CPU, &kthread->flags))
-			__kthread_bind(k, kthread->cpu, TASK_PARKED);
-		wake_up_state(k, TASK_PARKED);
-	}
+	if (test_bit(KTHREAD_IS_PER_CPU, &kthread->flags))
+		__kthread_bind(k, kthread->cpu, TASK_PARKED);
+
+	clear_bit(KTHREAD_SHOULD_PARK, &kthread->flags);
+	wake_up_state(k, TASK_PARKED);
 }
 EXPORT_SYMBOL_GPL(kthread_unpark);
 
@@ -489,12 +483,13 @@ int kthread_park(struct task_struct *k)
 	if (WARN_ON(k->flags & PF_EXITING))
 		return -ENOSYS;
 
-	if (!test_bit(KTHREAD_IS_PARKED, &kthread->flags)) {
-		set_bit(KTHREAD_SHOULD_PARK, &kthread->flags);
-		if (k != current) {
-			wake_up_process(k);
-			wait_for_completion(&kthread->parked);
-		}
+	if (WARN_ON_ONCE(test_bit(KTHREAD_SHOULD_PARK, &kthread->flags)))
+		return -EBUSY;
+
+	set_bit(KTHREAD_SHOULD_PARK, &kthread->flags);
+	if (k != current) {
+		wake_up_process(k);
+		wait_for_completion(&kthread->parked);
 	}
 
 	return 0;
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 5e10aae..7ad60e0 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -7,6 +7,8 @@
  */
 #include "sched.h"
 
+#include <linux/kthread.h>
+
 #include <asm/switch_to.h>
 #include <asm/tlb.h>
 
@@ -2718,20 +2720,28 @@ static struct rq *finish_task_switch(struct task_struct *prev)
 		membarrier_mm_sync_core_before_usermode(mm);
 		mmdrop(mm);
 	}
-	if (unlikely(prev_state == TASK_DEAD)) {
-		if (prev->sched_class->task_dead)
-			prev->sched_class->task_dead(prev);
+	if (unlikely(prev_state & (TASK_DEAD|TASK_PARKED))) {
+		switch (prev_state) {
+		case TASK_DEAD:
+			if (prev->sched_class->task_dead)
+				prev->sched_class->task_dead(prev);
 
-		/*
-		 * Remove function-return probe instances associated with this
-		 * task and put them back on the free list.
-		 */
-		kprobe_flush_task(prev);
+			/*
+			 * Remove function-return probe instances associated with this
+			 * task and put them back on the free list.
+			 */
+			kprobe_flush_task(prev);
 
-		/* Task is done with its stack. */
-		put_task_stack(prev);
+			/* Task is done with its stack. */
+			put_task_stack(prev);
 
-		put_task_struct(prev);
+			put_task_struct(prev);
+			break;
+
+		case TASK_PARKED:
+			kthread_park_complete(prev);
+			break;
+		}
 	}
 
 	tick_nohz_task_switch();
-- 
cgit v1.1


From b9b7742687085fc0fc0135d9104e0a8f30bea2e1 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Thu, 3 May 2018 00:02:02 +0100
Subject: drm/i915/execlists: Emit i915_trace_request_out for preemption

Move the tracepoint into the common execlists_context_schedule_out() and
call it from preemption completion as well. A small bit of refactoring
code should help with when tracing, or else we end up with requests
mysteriously disappearing and some being emitted to HW multiple times.

Reported-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180502230202.6848-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/intel_lrc.c | 19 +++++++++----------
 1 file changed, 9 insertions(+), 10 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index e04798e..3d747d1 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -398,10 +398,11 @@ execlists_context_schedule_in(struct i915_request *rq)
 }
 
 static inline void
-execlists_context_schedule_out(struct i915_request *rq)
+execlists_context_schedule_out(struct i915_request *rq, unsigned long status)
 {
 	intel_engine_context_out(rq->engine);
-	execlists_context_status_change(rq, INTEL_CONTEXT_SCHEDULE_OUT);
+	execlists_context_status_change(rq, status);
+	trace_i915_request_out(rq);
 }
 
 static void
@@ -772,12 +773,10 @@ execlists_cancel_port_requests(struct intel_engine_execlists * const execlists)
 			  intel_engine_get_seqno(rq->engine));
 
 		GEM_BUG_ON(!execlists->active);
-		intel_engine_context_out(rq->engine);
-
-		execlists_context_status_change(rq,
-						i915_request_completed(rq) ?
-						INTEL_CONTEXT_SCHEDULE_OUT :
-						INTEL_CONTEXT_SCHEDULE_PREEMPTED);
+		execlists_context_schedule_out(rq,
+					       i915_request_completed(rq) ?
+					       INTEL_CONTEXT_SCHEDULE_OUT :
+					       INTEL_CONTEXT_SCHEDULE_PREEMPTED);
 
 		i915_request_put(rq);
 
@@ -1105,8 +1104,8 @@ static void execlists_submission_tasklet(unsigned long data)
 				 */
 				GEM_BUG_ON(!i915_request_completed(rq));
 
-				execlists_context_schedule_out(rq);
-				trace_i915_request_out(rq);
+				execlists_context_schedule_out(rq,
+							       INTEL_CONTEXT_SCHEDULE_OUT);
 				i915_request_put(rq);
 
 				GEM_TRACE("%s completed ctx=%d\n",
-- 
cgit v1.1


From e01569ab962145b2fb46f6240bf1f07ec3a6261a Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Mon, 9 Apr 2018 10:49:05 +0100
Subject: drm/i915: Silence debugging DRM_ERROR for failing to suspend vlv
 powerwells
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

If we try to suspend a wedged device following a GPU reset failure, we
will also fail to turn off the rc6 powerwells (on vlv), leading to a
*ERROR*. This is quite expected in this case, so the best we can do is
shake our heads and reduce the *ERROR* to a debug so CI stops
complaining.

Testcase: igt/gem_eio/in-flight-suspend #vlv
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105583
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Imre Deak <imre.deak@intel.com>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Acked-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180409094905.4516-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/i915_drv.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index b7dbeba..8c29868 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -2468,10 +2468,13 @@ static void vlv_wait_for_gt_wells(struct drm_i915_private *dev_priv,
 	/*
 	 * RC6 transitioning can be delayed up to 2 msec (see
 	 * valleyview_enable_rps), use 3 msec for safety.
+	 *
+	 * This can fail to turn off the rc6 if the GPU is stuck after a failed
+	 * reset and we are trying to force the machine to sleep.
 	 */
 	if (vlv_wait_for_pw_status(dev_priv, mask, val))
-		DRM_ERROR("timeout waiting for GT wells to go %s\n",
-			  onoff(wait_for_on));
+		DRM_DEBUG_DRIVER("timeout waiting for GT wells to go %s\n",
+				 onoff(wait_for_on));
 }
 
 static void vlv_check_no_gt_access(struct drm_i915_private *dev_priv)
-- 
cgit v1.1


From ea491b23b2ffba069537a8216060d4d3400931a7 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Wed, 2 May 2018 23:03:12 +0100
Subject: drm/i915: Reset the hangcheck timestamp before repeating a seqno

In the unusual circumstance where we reuse a seqno (for example, in
igt), make sure that we reset the hangcheck timestamp before it sees the
same seqno again.

References: https://bugs.freedesktop.org/show_bug.cgi?id=106215
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180502220313.6459-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/i915_request.c    | 1 +
 drivers/gpu/drm/i915/intel_hangcheck.c | 1 +
 2 files changed, 2 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c
index 5acf869f..63bb610 100644
--- a/drivers/gpu/drm/i915/i915_request.c
+++ b/drivers/gpu/drm/i915/i915_request.c
@@ -223,6 +223,7 @@ static int reset_all_global_seqno(struct drm_i915_private *i915, u32 seqno)
 		if (!i915_seqno_passed(seqno, engine->timeline.seqno)) {
 			/* Flush any waiters before we reuse the seqno */
 			intel_engine_disarm_breadcrumbs(engine);
+			intel_engine_init_hangcheck(engine);
 			GEM_BUG_ON(!list_empty(&engine->breadcrumbs.signals));
 		}
 
diff --git a/drivers/gpu/drm/i915/intel_hangcheck.c b/drivers/gpu/drm/i915/intel_hangcheck.c
index 309e38b..d47e346 100644
--- a/drivers/gpu/drm/i915/intel_hangcheck.c
+++ b/drivers/gpu/drm/i915/intel_hangcheck.c
@@ -452,6 +452,7 @@ static void i915_hangcheck_elapsed(struct work_struct *work)
 void intel_engine_init_hangcheck(struct intel_engine_cs *engine)
 {
 	memset(&engine->hangcheck, 0, sizeof(engine->hangcheck));
+	engine->hangcheck.action_timestamp = jiffies;
 }
 
 void intel_hangcheck_init(struct drm_i915_private *i915)
-- 
cgit v1.1


From e21b141376f9b654e782757e156886c34bdf12ae Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Wed, 2 May 2018 23:03:13 +0100
Subject: drm/i915: Mark the hangcheck as idle when unparking the engines

As we unpark the engines and are about to begin a new cycle of activity,
mark the current status of the hangceck as idle so that we avoid
carrying over a stale timestamp/action into the next cycle.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180502220313.6459-2-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/intel_engine_cs.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
index a90769b..70325e0 100644
--- a/drivers/gpu/drm/i915/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/intel_engine_cs.c
@@ -1073,6 +1073,8 @@ void intel_engines_unpark(struct drm_i915_private *i915)
 	for_each_engine(engine, i915, id) {
 		if (engine->unpark)
 			engine->unpark(engine);
+
+		intel_engine_init_hangcheck(engine);
 	}
 }
 
-- 
cgit v1.1


From 3297234a05ab1e90091b0574db4c397ef0e90d5f Mon Sep 17 00:00:00 2001
From: Rodrigo Vivi <rodrigo.vivi@intel.com>
Date: Wed, 2 May 2018 10:52:55 -0700
Subject: drm/i915: Adjust eDP's logical vco in a reliable place.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

On intel_dp_compute_config() we were calculating the needed vco
for eDP on gen9 and we stashing it in
intel_atomic_state.cdclk.logical.vco

However few moments later on intel_modeset_checks() we fully
replace entire intel_atomic_state.cdclk.logical with
dev_priv->cdclk.logical fully overwriting the logical desired
vco for eDP on gen9.

So, with wrong VCO value we end up with wrong desired cdclk, but
also it will raise a lot of WARNs: On gen9, when we read
CDCLK_CTL to verify if we configured properly the desired
frequency the CD Frequency Select bits [27:26] == 10b can mean
337.5 or 308.57 MHz depending on the VCO. So if we have wrong
VCO value stashed we will believe the frequency selection didn't
stick and start to raise WARNs of cdclk mismatch.

[   42.857519] [drm:intel_dump_cdclk_state [i915]] Changing CDCLK to 308571 kHz, VCO 8640000 kHz, ref 24000 kHz, bypass 24000 kHz, voltage level 0
[   42.897269] cdclk state doesn't match!
[   42.901052] WARNING: CPU: 5 PID: 1116 at drivers/gpu/drm/i915/intel_cdclk.c:2084 intel_set_cdclk+0x5d/0x110 [i915]
[   42.938004] RIP: 0010:intel_set_cdclk+0x5d/0x110 [i915]
[   43.155253] WARNING: CPU: 5 PID: 1116 at drivers/gpu/drm/i915/intel_cdclk.c:2084 intel_set_cdclk+0x5d/0x110 [i915]
[   43.170277] [drm:intel_dump_cdclk_state [i915]] [hw state] 337500 kHz, VCO 8100000 kHz, ref 24000 kHz, bypass 24000 kHz, voltage level 0
[   43.182566] [drm:intel_dump_cdclk_state [i915]] [sw state] 308571 kHz, VCO 8640000 kHz, ref 24000 kHz, bypass 24000 kHz, voltage level 0

v2: Move the entire eDP's vco logical adjustment to inside
    the skl_modeset_calc_cdclk as suggested by Ville.

Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Fixes: bb0f4aab0e76 ("drm/i915: Track full cdclk state for the logical and actual cdclk frequencies")
Cc: <stable@vger.kernel.org> # v4.12+
Link: https://patchwork.freedesktop.org/patch/msgid/20180502175255.5344-1-rodrigo.vivi@intel.com
---
 drivers/gpu/drm/i915/intel_cdclk.c | 41 ++++++++++++++++++++++++++++++++++----
 drivers/gpu/drm/i915/intel_dp.c    | 20 -------------------
 2 files changed, 37 insertions(+), 24 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_cdclk.c b/drivers/gpu/drm/i915/intel_cdclk.c
index 32d24c6..704ddb4 100644
--- a/drivers/gpu/drm/i915/intel_cdclk.c
+++ b/drivers/gpu/drm/i915/intel_cdclk.c
@@ -2302,9 +2302,44 @@ static int bdw_modeset_calc_cdclk(struct drm_atomic_state *state)
 	return 0;
 }
 
+static int skl_dpll0_vco(struct intel_atomic_state *intel_state)
+{
+	struct drm_i915_private *dev_priv = to_i915(intel_state->base.dev);
+	struct intel_crtc *crtc;
+	struct intel_crtc_state *crtc_state;
+	int vco, i;
+
+	vco = intel_state->cdclk.logical.vco;
+	if (!vco)
+		vco = dev_priv->skl_preferred_vco_freq;
+
+	for_each_new_intel_crtc_in_state(intel_state, crtc, crtc_state, i) {
+		if (!crtc_state->base.enable)
+			continue;
+
+		if (!intel_crtc_has_type(crtc_state, INTEL_OUTPUT_EDP))
+			continue;
+
+		/*
+		 * DPLL0 VCO may need to be adjusted to get the correct
+		 * clock for eDP. This will affect cdclk as well.
+		 */
+		switch (crtc_state->port_clock / 2) {
+		case 108000:
+		case 216000:
+			vco = 8640000;
+			break;
+		default:
+			vco = 8100000;
+			break;
+		}
+	}
+
+	return vco;
+}
+
 static int skl_modeset_calc_cdclk(struct drm_atomic_state *state)
 {
-	struct drm_i915_private *dev_priv = to_i915(state->dev);
 	struct intel_atomic_state *intel_state = to_intel_atomic_state(state);
 	int min_cdclk, cdclk, vco;
 
@@ -2312,9 +2347,7 @@ static int skl_modeset_calc_cdclk(struct drm_atomic_state *state)
 	if (min_cdclk < 0)
 		return min_cdclk;
 
-	vco = intel_state->cdclk.logical.vco;
-	if (!vco)
-		vco = dev_priv->skl_preferred_vco_freq;
+	vco = skl_dpll0_vco(intel_state);
 
 	/*
 	 * FIXME should also account for plane ratio
diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index 83da50b..dde92e4 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -1929,26 +1929,6 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 					       reduce_m_n);
 	}
 
-	/*
-	 * DPLL0 VCO may need to be adjusted to get the correct
-	 * clock for eDP. This will affect cdclk as well.
-	 */
-	if (intel_dp_is_edp(intel_dp) && IS_GEN9_BC(dev_priv)) {
-		int vco;
-
-		switch (pipe_config->port_clock / 2) {
-		case 108000:
-		case 216000:
-			vco = 8640000;
-			break;
-		default:
-			vco = 8100000;
-			break;
-		}
-
-		to_intel_atomic_state(pipe_config->base.state)->cdclk.logical.vco = vco;
-	}
-
 	if (!HAS_DDI(dev_priv))
 		intel_dp_set_clock(encoder, pipe_config);
 
-- 
cgit v1.1


From aa7528fe3576d11f4a10237178a723a1f080a547 Mon Sep 17 00:00:00 2001
From: Jagannathan Raman <jag.raman@oracle.com>
Date: Tue, 6 Mar 2018 17:39:41 -0500
Subject: iommu/vt-d: Fix usage of force parameter in
 intel_ir_reconfigure_irte()

It was noticed that the IRTE configured for guest OS kernel
was over-written while the guest was running. As a result,
vt-d Posted Interrupts configured for the guest are not being
delivered directly, and instead bounces off the host. Every
interrupt delivery takes a VM Exit.

It was noticed that the following stack is doing the over-write:
[  147.463177]  modify_irte+0x171/0x1f0
[  147.463405]  intel_ir_set_affinity+0x5c/0x80
[  147.463641]  msi_domain_set_affinity+0x32/0x90
[  147.463881]  irq_do_set_affinity+0x37/0xd0
[  147.464125]  irq_set_affinity_locked+0x9d/0xb0
[  147.464374]  __irq_set_affinity+0x42/0x70
[  147.464627]  write_irq_affinity.isra.5+0xe1/0x110
[  147.464895]  proc_reg_write+0x38/0x70
[  147.465150]  __vfs_write+0x36/0x180
[  147.465408]  ? handle_mm_fault+0xdf/0x200
[  147.465671]  ? _cond_resched+0x15/0x30
[  147.465936]  vfs_write+0xad/0x1a0
[  147.466204]  SyS_write+0x52/0xc0
[  147.466472]  do_syscall_64+0x74/0x1a0
[  147.466744]  entry_SYSCALL_64_after_hwframe+0x3d/0xa2

reversing the sense of force check in intel_ir_reconfigure_irte()
restores proper posted interrupt functionality

Signed-off-by: Jagannathan Raman <jag.raman@oracle.com>
Fixes: d491bdff888e ('iommu/vt-d: Reevaluate vector configuration on activate()')
Signed-off-by: Joerg Roedel <jroedel@suse.de>
---
 drivers/iommu/intel_irq_remapping.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/iommu/intel_irq_remapping.c b/drivers/iommu/intel_irq_remapping.c
index 66f69af..3062a15 100644
--- a/drivers/iommu/intel_irq_remapping.c
+++ b/drivers/iommu/intel_irq_remapping.c
@@ -1136,7 +1136,7 @@ static void intel_ir_reconfigure_irte(struct irq_data *irqd, bool force)
 	irte->dest_id = IRTE_DEST(cfg->dest_apicid);
 
 	/* Update the hardware only if the interrupt is in remapped mode. */
-	if (!force || ir_data->irq_2_iommu.mode == IRQ_REMAPPING)
+	if (force || ir_data->irq_2_iommu.mode == IRQ_REMAPPING)
 		modify_irte(&ir_data->irq_2_iommu, irte);
 }
 
-- 
cgit v1.1


From 94c793accacdb0d33c1df66f3b324eec96d26e58 Mon Sep 17 00:00:00 2001
From: Arnd Bergmann <arnd@arndb.de>
Date: Wed, 4 Apr 2018 12:56:59 +0200
Subject: iommu/amd: Hide unused iommu_table_lock

The newly introduced lock is only used when CONFIG_IRQ_REMAP is enabled:

drivers/iommu/amd_iommu.c:86:24: error: 'iommu_table_lock' defined but not used [-Werror=unused-variable]
 static DEFINE_SPINLOCK(iommu_table_lock);

This moves the definition next to the user, within the #ifdef protected
section of the file.

Fixes: ea6166f4b83e ("iommu/amd: Split irq_lookup_table out of the amd_iommu_devtable_lock")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Joerg Roedel <jroedel@suse.de>
---
 drivers/iommu/amd_iommu.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/iommu/amd_iommu.c b/drivers/iommu/amd_iommu.c
index 2a99f0f..8fb8c73 100644
--- a/drivers/iommu/amd_iommu.c
+++ b/drivers/iommu/amd_iommu.c
@@ -83,7 +83,6 @@
 
 static DEFINE_SPINLOCK(amd_iommu_devtable_lock);
 static DEFINE_SPINLOCK(pd_bitmap_lock);
-static DEFINE_SPINLOCK(iommu_table_lock);
 
 /* List of all available dev_data structures */
 static LLIST_HEAD(dev_data_list);
@@ -3562,6 +3561,7 @@ EXPORT_SYMBOL(amd_iommu_device_info);
  *****************************************************************************/
 
 static struct irq_chip amd_ir_chip;
+static DEFINE_SPINLOCK(iommu_table_lock);
 
 static void set_dte_irq_entry(u16 devid, struct irq_remap_table *table)
 {
-- 
cgit v1.1


From 2f8c7f2e76bf040fcdaa2b4f4c2b9ef3b6e4ed25 Mon Sep 17 00:00:00 2001
From: Heiko Stuebner <heiko@sntech.de>
Date: Tue, 17 Apr 2018 14:09:15 +0200
Subject: iommu/rockchip: Make clock handling optional

iommu clocks are optional, so the driver should not fail if they are not
present. Instead just set the number of clocks to 0, which the clk-blk APIs
can handle just fine.

Fixes: f2e3a5f557ad ("iommu/rockchip: Control clocks needed to access the IOMMU")
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Reviewed-by: Robin Murphy <robin.murphy@arm.com>
Tested-by: Enric Balletbo i Serra <enric.balletbo@collabora.com>
Signed-off-by: Joerg Roedel <jroedel@suse.de>
---
 drivers/iommu/rockchip-iommu.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/drivers/iommu/rockchip-iommu.c b/drivers/iommu/rockchip-iommu.c
index 5fc8656..c0f03af 100644
--- a/drivers/iommu/rockchip-iommu.c
+++ b/drivers/iommu/rockchip-iommu.c
@@ -1175,8 +1175,15 @@ static int rk_iommu_probe(struct platform_device *pdev)
 	for (i = 0; i < iommu->num_clocks; ++i)
 		iommu->clocks[i].id = rk_iommu_clocks[i];
 
+	/*
+	 * iommu clocks should be present for all new devices and devicetrees
+	 * but there are older devicetrees without clocks out in the wild.
+	 * So clocks as optional for the time being.
+	 */
 	err = devm_clk_bulk_get(iommu->dev, iommu->num_clocks, iommu->clocks);
-	if (err)
+	if (err == -ENOENT)
+		iommu->num_clocks = 0;
+	else if (err)
 		return err;
 
 	err = clk_bulk_prepare(iommu->num_clocks, iommu->clocks);
-- 
cgit v1.1


From cd2c9fcf5c66fd21f7b83455806fdd14ab9236ee Mon Sep 17 00:00:00 2001
From: Shameer Kolothum <shameerali.kolothum.thodi@huawei.com>
Date: Wed, 18 Apr 2018 12:40:42 +0100
Subject: iommu/dma: Move PCI window region reservation back into dma specific
 path.

This pretty much reverts commit 273df9635385 ("iommu/dma: Make PCI
window reservation generic")  by moving the PCI window region
reservation back into the dma specific path so that these regions
doesn't get exposed via the IOMMU API interface. With this change,
the vfio interface will report only iommu specific reserved regions
to the user space.

Cc: Joerg Roedel <joro@8bytes.org>
Signed-off-by: Shameer Kolothum <shameerali.kolothum.thodi@huawei.com>
Reviewed-by: Robin Murphy <robin.murphy@arm.com>
Fixes: 273df9635385 ('iommu/dma: Make PCI window reservation generic')
Signed-off-by: Joerg Roedel <jroedel@suse.de>
---
 drivers/iommu/dma-iommu.c | 54 ++++++++++++++++++++++-------------------------
 1 file changed, 25 insertions(+), 29 deletions(-)

diff --git a/drivers/iommu/dma-iommu.c b/drivers/iommu/dma-iommu.c
index f05f3cf..ddcbbdb 100644
--- a/drivers/iommu/dma-iommu.c
+++ b/drivers/iommu/dma-iommu.c
@@ -167,40 +167,16 @@ EXPORT_SYMBOL(iommu_put_dma_cookie);
  * @list: Reserved region list from iommu_get_resv_regions()
  *
  * IOMMU drivers can use this to implement their .get_resv_regions callback
- * for general non-IOMMU-specific reservations. Currently, this covers host
- * bridge windows for PCI devices and GICv3 ITS region reservation on ACPI
- * based ARM platforms that may require HW MSI reservation.
+ * for general non-IOMMU-specific reservations. Currently, this covers GICv3
+ * ITS region reservation on ACPI based ARM platforms that may require HW MSI
+ * reservation.
  */
 void iommu_dma_get_resv_regions(struct device *dev, struct list_head *list)
 {
-	struct pci_host_bridge *bridge;
-	struct resource_entry *window;
-
-	if (!is_of_node(dev->iommu_fwspec->iommu_fwnode) &&
-		iort_iommu_msi_get_resv_regions(dev, list) < 0)
-		return;
-
-	if (!dev_is_pci(dev))
-		return;
-
-	bridge = pci_find_host_bridge(to_pci_dev(dev)->bus);
-	resource_list_for_each_entry(window, &bridge->windows) {
-		struct iommu_resv_region *region;
-		phys_addr_t start;
-		size_t length;
-
-		if (resource_type(window->res) != IORESOURCE_MEM)
-			continue;
 
-		start = window->res->start - window->offset;
-		length = window->res->end - window->res->start + 1;
-		region = iommu_alloc_resv_region(start, length, 0,
-				IOMMU_RESV_RESERVED);
-		if (!region)
-			return;
+	if (!is_of_node(dev->iommu_fwspec->iommu_fwnode))
+		iort_iommu_msi_get_resv_regions(dev, list);
 
-		list_add_tail(&region->list, list);
-	}
 }
 EXPORT_SYMBOL(iommu_dma_get_resv_regions);
 
@@ -229,6 +205,23 @@ static int cookie_init_hw_msi_region(struct iommu_dma_cookie *cookie,
 	return 0;
 }
 
+static void iova_reserve_pci_windows(struct pci_dev *dev,
+		struct iova_domain *iovad)
+{
+	struct pci_host_bridge *bridge = pci_find_host_bridge(dev->bus);
+	struct resource_entry *window;
+	unsigned long lo, hi;
+
+	resource_list_for_each_entry(window, &bridge->windows) {
+		if (resource_type(window->res) != IORESOURCE_MEM)
+			continue;
+
+		lo = iova_pfn(iovad, window->res->start - window->offset);
+		hi = iova_pfn(iovad, window->res->end - window->offset);
+		reserve_iova(iovad, lo, hi);
+	}
+}
+
 static int iova_reserve_iommu_regions(struct device *dev,
 		struct iommu_domain *domain)
 {
@@ -238,6 +231,9 @@ static int iova_reserve_iommu_regions(struct device *dev,
 	LIST_HEAD(resv_regions);
 	int ret = 0;
 
+	if (dev_is_pci(dev))
+		iova_reserve_pci_windows(to_pci_dev(dev), iovad);
+
 	iommu_get_resv_regions(dev, &resv_regions);
 	list_for_each_entry(region, &resv_regions, list) {
 		unsigned long lo, hi;
-- 
cgit v1.1


From 0dfc0c792d691f8056f38b5c30789f504be0e467 Mon Sep 17 00:00:00 2001
From: Changbin Du <changbin.du@intel.com>
Date: Fri, 20 Apr 2018 13:29:55 +0800
Subject: iommu/vt-d: fix shift-out-of-bounds in bug checking

It allows to flush more than 4GB of device TLBs. So the mask should be
64bit wide. UBSAN captured this fault as below.

[    3.760024] ================================================================================
[    3.768440] UBSAN: Undefined behaviour in drivers/iommu/dmar.c:1348:3
[    3.774864] shift exponent 64 is too large for 32-bit type 'int'
[    3.780853] CPU: 2 PID: 0 Comm: swapper/2 Tainted: G     U            4.17.0-rc1+ #89
[    3.788661] Hardware name: Dell Inc. OptiPlex 7040/0Y7WYT, BIOS 1.2.8 01/26/2016
[    3.796034] Call Trace:
[    3.798472]  <IRQ>
[    3.800479]  dump_stack+0x90/0xfb
[    3.803787]  ubsan_epilogue+0x9/0x40
[    3.807353]  __ubsan_handle_shift_out_of_bounds+0x10e/0x170
[    3.812916]  ? qi_flush_dev_iotlb+0x124/0x180
[    3.817261]  qi_flush_dev_iotlb+0x124/0x180
[    3.821437]  iommu_flush_dev_iotlb+0x94/0xf0
[    3.825698]  iommu_flush_iova+0x10b/0x1c0
[    3.829699]  ? fq_ring_free+0x1d0/0x1d0
[    3.833527]  iova_domain_flush+0x25/0x40
[    3.837448]  fq_flush_timeout+0x55/0x160
[    3.841368]  ? fq_ring_free+0x1d0/0x1d0
[    3.845200]  ? fq_ring_free+0x1d0/0x1d0
[    3.849034]  call_timer_fn+0xbe/0x310
[    3.852696]  ? fq_ring_free+0x1d0/0x1d0
[    3.856530]  run_timer_softirq+0x223/0x6e0
[    3.860625]  ? sched_clock+0x5/0x10
[    3.864108]  ? sched_clock+0x5/0x10
[    3.867594]  __do_softirq+0x1b5/0x6f5
[    3.871250]  irq_exit+0xd4/0x130
[    3.874470]  smp_apic_timer_interrupt+0xb8/0x2f0
[    3.879075]  apic_timer_interrupt+0xf/0x20
[    3.883159]  </IRQ>
[    3.885255] RIP: 0010:poll_idle+0x60/0xe7
[    3.889252] RSP: 0018:ffffb1b201943e30 EFLAGS: 00000246 ORIG_RAX: ffffffffffffff13
[    3.896802] RAX: 0000000080200000 RBX: 000000000000008e RCX: 000000000000001f
[    3.903918] RDX: 0000000000000000 RSI: 000000002819aa06 RDI: 0000000000000000
[    3.911031] RBP: ffff9e93c6b33280 R08: 00000010f717d567 R09: 000000000010d205
[    3.918146] R10: ffffb1b201943df8 R11: 0000000000000001 R12: 00000000e01b169d
[    3.925260] R13: 0000000000000000 R14: ffffffffb12aa400 R15: 0000000000000000
[    3.932382]  cpuidle_enter_state+0xb4/0x470
[    3.936558]  do_idle+0x222/0x310
[    3.939779]  cpu_startup_entry+0x78/0x90
[    3.943693]  start_secondary+0x205/0x2e0
[    3.947607]  secondary_startup_64+0xa5/0xb0
[    3.951783] ================================================================================

Signed-off-by: Changbin Du <changbin.du@intel.com>
Signed-off-by: Joerg Roedel <jroedel@suse.de>
---
 drivers/iommu/dmar.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/iommu/dmar.c b/drivers/iommu/dmar.c
index accf5838..e4ae600 100644
--- a/drivers/iommu/dmar.c
+++ b/drivers/iommu/dmar.c
@@ -1345,7 +1345,7 @@ void qi_flush_dev_iotlb(struct intel_iommu *iommu, u16 sid, u16 qdep,
 	struct qi_desc desc;
 
 	if (mask) {
-		BUG_ON(addr & ((1 << (VTD_PAGE_SHIFT + mask)) - 1));
+		BUG_ON(addr & ((1ULL << (VTD_PAGE_SHIFT + mask)) - 1));
 		addr |= (1ULL << (VTD_PAGE_SHIFT + mask - 1)) - 1;
 		desc.high = QI_DEV_IOTLB_ADDR(addr) | QI_DEV_IOTLB_SIZE;
 	} else
-- 
cgit v1.1


From a85894cd779455fab0086cfcb5c9f65c3706e1c6 Mon Sep 17 00:00:00 2001
From: Joerg Roedel <jroedel@suse.de>
Date: Thu, 3 May 2018 15:25:17 +0200
Subject: iommu/vt-d: Use WARN_ON_ONCE instead of BUG_ON in
 qi_flush_dev_iotlb()

A misaligned address is only worth a warning, and not
stopping the while execution path with a BUG_ON().

Signed-off-by: Joerg Roedel <jroedel@suse.de>
---
 drivers/iommu/dmar.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/iommu/dmar.c b/drivers/iommu/dmar.c
index e4ae600..460bed4 100644
--- a/drivers/iommu/dmar.c
+++ b/drivers/iommu/dmar.c
@@ -1345,7 +1345,7 @@ void qi_flush_dev_iotlb(struct intel_iommu *iommu, u16 sid, u16 qdep,
 	struct qi_desc desc;
 
 	if (mask) {
-		BUG_ON(addr & ((1ULL << (VTD_PAGE_SHIFT + mask)) - 1));
+		WARN_ON_ONCE(addr & ((1ULL << (VTD_PAGE_SHIFT + mask)) - 1));
 		addr |= (1ULL << (VTD_PAGE_SHIFT + mask - 1)) - 1;
 		desc.high = QI_DEV_IOTLB_ADDR(addr) | QI_DEV_IOTLB_SIZE;
 	} else
-- 
cgit v1.1


From 6e916a7eb1bc045f4e27355632ee7692014e6e60 Mon Sep 17 00:00:00 2001
From: Coly Li <colyli@suse.de>
Date: Thu, 3 May 2018 18:51:32 +0800
Subject: bcache: store disk name in struct cache and struct cached_dev

Current code uses bdevname() or bio_devname() to reference gendisk
disk name when bcache needs to display the disk names in kernel message.
It was safe before bcache device failure handling patch set merged in,
because when devices are failed, there was deadlock to prevent bcache
printing error messages with gendisk disk name. But after the failure
handling patch set merged, the deadlock is fixed, so it is possible
that the gendisk structure bdev->hd_disk is released when bdevname() is
called to reference bdev->bd_disk->disk_name[]. This is why I receive
bug report of NULL pointers deference panic.

This patch stores gendisk disk name in a buffer inside struct cache and
struct cached_dev, then print out the offline device name won't reference
bdev->hd_disk anymore. And this patch also avoids extra function calls
of bdevname() and bio_devnmae().

Changelog:
v3, add Reviewed-by from Hannes.
v2, call bdevname() earlier in register_bdev()
v1, first version with segguestion from Junhui Tang.

Fixes: c7b7bd07404c5 ("bcache: add io_disable to struct cached_dev")
Fixes: 5138ac6748e38 ("bcache: fix misleading error message in bch_count_io_errors()")
Signed-off-by: Coly Li <colyli@suse.de>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 drivers/md/bcache/bcache.h  |  4 ++++
 drivers/md/bcache/debug.c   |  3 +--
 drivers/md/bcache/io.c      |  8 +++-----
 drivers/md/bcache/request.c |  5 +----
 drivers/md/bcache/super.c   | 44 +++++++++++++++++++++-----------------------
 5 files changed, 30 insertions(+), 34 deletions(-)

diff --git a/drivers/md/bcache/bcache.h b/drivers/md/bcache/bcache.h
index d338b70..3a0cfb2 100644
--- a/drivers/md/bcache/bcache.h
+++ b/drivers/md/bcache/bcache.h
@@ -392,6 +392,8 @@ struct cached_dev {
 #define DEFAULT_CACHED_DEV_ERROR_LIMIT	64
 	atomic_t		io_errors;
 	unsigned		error_limit;
+
+	char			backing_dev_name[BDEVNAME_SIZE];
 };
 
 enum alloc_reserve {
@@ -464,6 +466,8 @@ struct cache {
 	atomic_long_t		meta_sectors_written;
 	atomic_long_t		btree_sectors_written;
 	atomic_long_t		sectors_written;
+
+	char			cache_dev_name[BDEVNAME_SIZE];
 };
 
 struct gc_stat {
diff --git a/drivers/md/bcache/debug.c b/drivers/md/bcache/debug.c
index 028f7b3..4e63c6f 100644
--- a/drivers/md/bcache/debug.c
+++ b/drivers/md/bcache/debug.c
@@ -106,7 +106,6 @@ void bch_btree_verify(struct btree *b)
 
 void bch_data_verify(struct cached_dev *dc, struct bio *bio)
 {
-	char name[BDEVNAME_SIZE];
 	struct bio *check;
 	struct bio_vec bv, cbv;
 	struct bvec_iter iter, citer = { 0 };
@@ -134,7 +133,7 @@ void bch_data_verify(struct cached_dev *dc, struct bio *bio)
 					bv.bv_len),
 				 dc->disk.c,
 				 "verify failed at dev %s sector %llu",
-				 bdevname(dc->bdev, name),
+				 dc->backing_dev_name,
 				 (uint64_t) bio->bi_iter.bi_sector);
 
 		kunmap_atomic(p1);
diff --git a/drivers/md/bcache/io.c b/drivers/md/bcache/io.c
index 7fac97a..2ddf851 100644
--- a/drivers/md/bcache/io.c
+++ b/drivers/md/bcache/io.c
@@ -52,7 +52,6 @@ void bch_submit_bbio(struct bio *bio, struct cache_set *c,
 /* IO errors */
 void bch_count_backing_io_errors(struct cached_dev *dc, struct bio *bio)
 {
-	char buf[BDEVNAME_SIZE];
 	unsigned errors;
 
 	WARN_ONCE(!dc, "NULL pointer of struct cached_dev");
@@ -60,7 +59,7 @@ void bch_count_backing_io_errors(struct cached_dev *dc, struct bio *bio)
 	errors = atomic_add_return(1, &dc->io_errors);
 	if (errors < dc->error_limit)
 		pr_err("%s: IO error on backing device, unrecoverable",
-			bio_devname(bio, buf));
+			dc->backing_dev_name);
 	else
 		bch_cached_dev_error(dc);
 }
@@ -105,19 +104,18 @@ void bch_count_io_errors(struct cache *ca,
 	}
 
 	if (error) {
-		char buf[BDEVNAME_SIZE];
 		unsigned errors = atomic_add_return(1 << IO_ERROR_SHIFT,
 						    &ca->io_errors);
 		errors >>= IO_ERROR_SHIFT;
 
 		if (errors < ca->set->error_limit)
 			pr_err("%s: IO error on %s%s",
-			       bdevname(ca->bdev, buf), m,
+			       ca->cache_dev_name, m,
 			       is_read ? ", recovering." : ".");
 		else
 			bch_cache_set_error(ca->set,
 					    "%s: too many IO errors %s",
-					    bdevname(ca->bdev, buf), m);
+					    ca->cache_dev_name, m);
 	}
 }
 
diff --git a/drivers/md/bcache/request.c b/drivers/md/bcache/request.c
index a65e336..8e3e865 100644
--- a/drivers/md/bcache/request.c
+++ b/drivers/md/bcache/request.c
@@ -649,11 +649,8 @@ static void backing_request_endio(struct bio *bio)
 		 */
 		if (unlikely(s->iop.writeback &&
 			     bio->bi_opf & REQ_PREFLUSH)) {
-			char buf[BDEVNAME_SIZE];
-
-			bio_devname(bio, buf);
 			pr_err("Can't flush %s: returned bi_status %i",
-				buf, bio->bi_status);
+				dc->backing_dev_name, bio->bi_status);
 		} else {
 			/* set to orig_bio->bi_status in bio_complete() */
 			s->iop.status = bio->bi_status;
diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c
index d90d9e5..8196b19 100644
--- a/drivers/md/bcache/super.c
+++ b/drivers/md/bcache/super.c
@@ -936,7 +936,6 @@ static void cancel_writeback_rate_update_dwork(struct cached_dev *dc)
 static void cached_dev_detach_finish(struct work_struct *w)
 {
 	struct cached_dev *dc = container_of(w, struct cached_dev, detach);
-	char buf[BDEVNAME_SIZE];
 	struct closure cl;
 	closure_init_stack(&cl);
 
@@ -967,7 +966,7 @@ static void cached_dev_detach_finish(struct work_struct *w)
 
 	mutex_unlock(&bch_register_lock);
 
-	pr_info("Caching disabled for %s", bdevname(dc->bdev, buf));
+	pr_info("Caching disabled for %s", dc->backing_dev_name);
 
 	/* Drop ref we took in cached_dev_detach() */
 	closure_put(&dc->disk.cl);
@@ -999,29 +998,28 @@ int bch_cached_dev_attach(struct cached_dev *dc, struct cache_set *c,
 {
 	uint32_t rtime = cpu_to_le32(get_seconds());
 	struct uuid_entry *u;
-	char buf[BDEVNAME_SIZE];
 	struct cached_dev *exist_dc, *t;
 
-	bdevname(dc->bdev, buf);
-
 	if ((set_uuid && memcmp(set_uuid, c->sb.set_uuid, 16)) ||
 	    (!set_uuid && memcmp(dc->sb.set_uuid, c->sb.set_uuid, 16)))
 		return -ENOENT;
 
 	if (dc->disk.c) {
-		pr_err("Can't attach %s: already attached", buf);
+		pr_err("Can't attach %s: already attached",
+		       dc->backing_dev_name);
 		return -EINVAL;
 	}
 
 	if (test_bit(CACHE_SET_STOPPING, &c->flags)) {
-		pr_err("Can't attach %s: shutting down", buf);
+		pr_err("Can't attach %s: shutting down",
+		       dc->backing_dev_name);
 		return -EINVAL;
 	}
 
 	if (dc->sb.block_size < c->sb.block_size) {
 		/* Will die */
 		pr_err("Couldn't attach %s: block size less than set's block size",
-		       buf);
+		       dc->backing_dev_name);
 		return -EINVAL;
 	}
 
@@ -1029,7 +1027,7 @@ int bch_cached_dev_attach(struct cached_dev *dc, struct cache_set *c,
 	list_for_each_entry_safe(exist_dc, t, &c->cached_devs, list) {
 		if (!memcmp(dc->sb.uuid, exist_dc->sb.uuid, 16)) {
 			pr_err("Tried to attach %s but duplicate UUID already attached",
-				buf);
+				dc->backing_dev_name);
 
 			return -EINVAL;
 		}
@@ -1047,13 +1045,15 @@ int bch_cached_dev_attach(struct cached_dev *dc, struct cache_set *c,
 
 	if (!u) {
 		if (BDEV_STATE(&dc->sb) == BDEV_STATE_DIRTY) {
-			pr_err("Couldn't find uuid for %s in set", buf);
+			pr_err("Couldn't find uuid for %s in set",
+			       dc->backing_dev_name);
 			return -ENOENT;
 		}
 
 		u = uuid_find_empty(c);
 		if (!u) {
-			pr_err("Not caching %s, no room for UUID", buf);
+			pr_err("Not caching %s, no room for UUID",
+			       dc->backing_dev_name);
 			return -EINVAL;
 		}
 	}
@@ -1112,7 +1112,8 @@ int bch_cached_dev_attach(struct cached_dev *dc, struct cache_set *c,
 	up_write(&dc->writeback_lock);
 
 	pr_info("Caching %s as %s on set %pU",
-		bdevname(dc->bdev, buf), dc->disk.disk->disk_name,
+		dc->backing_dev_name,
+		dc->disk.disk->disk_name,
 		dc->disk.c->sb.set_uuid);
 	return 0;
 }
@@ -1225,10 +1226,10 @@ static void register_bdev(struct cache_sb *sb, struct page *sb_page,
 				 struct block_device *bdev,
 				 struct cached_dev *dc)
 {
-	char name[BDEVNAME_SIZE];
 	const char *err = "cannot allocate memory";
 	struct cache_set *c;
 
+	bdevname(bdev, dc->backing_dev_name);
 	memcpy(&dc->sb, sb, sizeof(struct cache_sb));
 	dc->bdev = bdev;
 	dc->bdev->bd_holder = dc;
@@ -1237,6 +1238,7 @@ static void register_bdev(struct cache_sb *sb, struct page *sb_page,
 	bio_first_bvec_all(&dc->sb_bio)->bv_page = sb_page;
 	get_page(sb_page);
 
+
 	if (cached_dev_init(dc, sb->block_size << 9))
 		goto err;
 
@@ -1247,7 +1249,7 @@ static void register_bdev(struct cache_sb *sb, struct page *sb_page,
 	if (bch_cache_accounting_add_kobjs(&dc->accounting, &dc->disk.kobj))
 		goto err;
 
-	pr_info("registered backing device %s", bdevname(bdev, name));
+	pr_info("registered backing device %s", dc->backing_dev_name);
 
 	list_add(&dc->list, &uncached_devices);
 	list_for_each_entry(c, &bch_cache_sets, list)
@@ -1259,7 +1261,7 @@ static void register_bdev(struct cache_sb *sb, struct page *sb_page,
 
 	return;
 err:
-	pr_notice("error %s: %s", bdevname(bdev, name), err);
+	pr_notice("error %s: %s", dc->backing_dev_name, err);
 	bcache_device_stop(&dc->disk);
 }
 
@@ -1367,8 +1369,6 @@ int bch_flash_dev_create(struct cache_set *c, uint64_t size)
 
 bool bch_cached_dev_error(struct cached_dev *dc)
 {
-	char name[BDEVNAME_SIZE];
-
 	if (!dc || test_bit(BCACHE_DEV_CLOSING, &dc->disk.flags))
 		return false;
 
@@ -1377,7 +1377,7 @@ bool bch_cached_dev_error(struct cached_dev *dc)
 	smp_mb();
 
 	pr_err("stop %s: too many IO errors on backing device %s\n",
-		dc->disk.disk->disk_name, bdevname(dc->bdev, name));
+		dc->disk.disk->disk_name, dc->backing_dev_name);
 
 	bcache_device_stop(&dc->disk);
 	return true;
@@ -2003,12 +2003,10 @@ static int cache_alloc(struct cache *ca)
 static int register_cache(struct cache_sb *sb, struct page *sb_page,
 				struct block_device *bdev, struct cache *ca)
 {
-	char name[BDEVNAME_SIZE];
 	const char *err = NULL; /* must be set for any error case */
 	int ret = 0;
 
-	bdevname(bdev, name);
-
+	bdevname(bdev, ca->cache_dev_name);
 	memcpy(&ca->sb, sb, sizeof(struct cache_sb));
 	ca->bdev = bdev;
 	ca->bdev->bd_holder = ca;
@@ -2045,14 +2043,14 @@ static int register_cache(struct cache_sb *sb, struct page *sb_page,
 		goto out;
 	}
 
-	pr_info("registered cache device %s", name);
+	pr_info("registered cache device %s", ca->cache_dev_name);
 
 out:
 	kobject_put(&ca->kobj);
 
 err:
 	if (err)
-		pr_notice("error %s: %s", name, err);
+		pr_notice("error %s: %s", ca->cache_dev_name, err);
 
 	return ret;
 }
-- 
cgit v1.1


From 6147305c73e4511ca1a975b766b97a779d442567 Mon Sep 17 00:00:00 2001
From: Coly Li <colyli@suse.de>
Date: Thu, 3 May 2018 18:51:33 +0800
Subject: bcache: set CACHE_SET_IO_DISABLE in bch_cached_dev_error()

Commit c7b7bd07404c5 ("bcache: add io_disable to struct cached_dev") tries
to stop bcache device by calling bcache_device_stop() when too many I/O
errors happened on backing device. But if there is internal I/O happening
on cache device (writeback scan, garbage collection, etc), a regular I/O
request triggers the internal I/Os may still holds a refcount of dc->count,
and the refcount may only be dropped after the internal I/O stopped.

By this patch, bch_cached_dev_error() will check if the backing device is
attached to a cache set, if yes that CACHE_SET_IO_DISABLE will be set to
flags of this cache set. Then internal I/Os on cache device will be
rejected and stopped immediately, and the bcache device can be stopped.

For people who are not familiar with the interesting refcount dependance,
let me explain a bit more how the fix works. Example the writeback thread
will scan cache device for dirty data writeback purpose. Before it stopps,
it holds a refcount of dc->count. When CACHE_SET_IO_DISABLE bit is set,
the internal I/O will stopped and the while-loop in bch_writeback_thread()
quits and calls cached_dev_put() to drop dc->count. If this is the last
refcount to drop, then cached_dev_detach_finish() will be called. In this
call back function, in turn closure_put(dc->disk.cl) is called to drop a
refcount of closure dc->disk.cl. If this is the last refcount of this
closure to drop, then cached_dev_flush() will be called. Then the cached
device is freed. So if CACHE_SET_IO_DISABLE is not set, the bache device
can not be stopped until all inernal cache device I/O stopped. For large
size cache device, and writeback thread competes locks with gc thread,
there might be a quite long time to wait.

Fixes: c7b7bd07404c5 ("bcache: add io_disable to struct cached_dev")
Signed-off-by: Coly Li <colyli@suse.de>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 drivers/md/bcache/super.c | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)

diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c
index 8196b19..c017cd4 100644
--- a/drivers/md/bcache/super.c
+++ b/drivers/md/bcache/super.c
@@ -1369,6 +1369,8 @@ int bch_flash_dev_create(struct cache_set *c, uint64_t size)
 
 bool bch_cached_dev_error(struct cached_dev *dc)
 {
+	struct cache_set *c;
+
 	if (!dc || test_bit(BCACHE_DEV_CLOSING, &dc->disk.flags))
 		return false;
 
@@ -1379,6 +1381,21 @@ bool bch_cached_dev_error(struct cached_dev *dc)
 	pr_err("stop %s: too many IO errors on backing device %s\n",
 		dc->disk.disk->disk_name, dc->backing_dev_name);
 
+	/*
+	 * If the cached device is still attached to a cache set,
+	 * even dc->io_disable is true and no more I/O requests
+	 * accepted, cache device internal I/O (writeback scan or
+	 * garbage collection) may still prevent bcache device from
+	 * being stopped. So here CACHE_SET_IO_DISABLE should be
+	 * set to c->flags too, to make the internal I/O to cache
+	 * device rejected and stopped immediately.
+	 * If c is NULL, that means the bcache device is not attached
+	 * to any cache set, then no CACHE_SET_IO_DISABLE bit to set.
+	 */
+	c = dc->disk.c;
+	if (c && test_and_set_bit(CACHE_SET_IO_DISABLE, &c->flags))
+		pr_info("CACHE_SET_IO_DISABLE already set");
+
 	bcache_device_stop(&dc->disk);
 	return true;
 }
-- 
cgit v1.1


From bf78980fcc58bad2d61858ce342153a3dd097aa0 Mon Sep 17 00:00:00 2001
From: Coly Li <colyli@suse.de>
Date: Thu, 3 May 2018 18:51:34 +0800
Subject: bcache: count backing device I/O error for writeback I/O

Commit c7b7bd07404c5 ("bcache: add io_disable to struct cached_dev")
counts backing device I/O requets and set dc->io_disable to true if error
counters exceeds dc->io_error_limit. But it only counts I/O errors for
regular I/O request, neglects errors of write back I/Os when backing device
is offline.

This patch counts the errors of writeback I/Os, in dirty_endio() if
bio->bi_status is  not 0, it means error happens when writing dirty keys
to backing device, then bch_count_backing_io_errors() is called.

By this fix, even there is no reqular I/O request coming, if writeback I/O
errors exceed dc->io_error_limit, the bcache device may still be stopped
for the broken backing device.

Fixes: c7b7bd07404c5 ("bcache: add io_disable to struct cached_dev")
Signed-off-by: Coly Li <colyli@suse.de>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 drivers/md/bcache/writeback.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/md/bcache/writeback.c b/drivers/md/bcache/writeback.c
index 4a9547c..ad45ebe 100644
--- a/drivers/md/bcache/writeback.c
+++ b/drivers/md/bcache/writeback.c
@@ -244,8 +244,10 @@ static void dirty_endio(struct bio *bio)
 	struct keybuf_key *w = bio->bi_private;
 	struct dirty_io *io = w->private;
 
-	if (bio->bi_status)
+	if (bio->bi_status) {
 		SET_KEY_DIRTY(&w->key, false);
+		bch_count_backing_io_errors(io->dc, bio);
+	}
 
 	closure_put(&io->cl);
 }
-- 
cgit v1.1


From ecb2ba8cb83549f1bb06bc7e693ae8fed43c0e4f Mon Sep 17 00:00:00 2001
From: Coly Li <colyli@suse.de>
Date: Thu, 3 May 2018 18:51:35 +0800
Subject: bcache: add wait_for_kthread_stop() in bch_allocator_thread()

When CACHE_SET_IO_DISABLE is set on cache set flags, bcache allocator
thread routine bch_allocator_thread() may stop the while-loops and
exit. Then it is possible to observe the following kernel oops message,

[  631.068366] bcache: bch_btree_insert() error -5
[  631.069115] bcache: cached_dev_detach_finish() Caching disabled for sdf
[  631.070220] BUG: unable to handle kernel NULL pointer dereference at 0000000000000000
[  631.070250] PGD 0 P4D 0
[  631.070261] Oops: 0002 [#1] SMP PTI
[snipped]
[  631.070578] Workqueue: events cache_set_flush [bcache]
[  631.070597] RIP: 0010:exit_creds+0x1b/0x50
[  631.070610] RSP: 0018:ffffc9000705fe08 EFLAGS: 00010246
[  631.070626] RAX: 0000000000000001 RBX: ffff880a622ad300 RCX: 000000000000000b
[  631.070645] RDX: 0000000000000601 RSI: 000000000000000c RDI: 0000000000000000
[  631.070663] RBP: ffff880a622ad300 R08: ffffea00190c66e0 R09: 0000000000000200
[  631.070682] R10: ffff880a48123000 R11: ffff880000000000 R12: 0000000000000000
[  631.070700] R13: ffff880a4b160e40 R14: ffff880a4b160000 R15: 0ffff880667e2530
[  631.070719] FS:  0000000000000000(0000) GS:ffff880667e00000(0000) knlGS:0000000000000000
[  631.070740] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[  631.070755] CR2: 0000000000000000 CR3: 000000000200a001 CR4: 00000000003606e0
[  631.070774] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[  631.070793] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[  631.070811] Call Trace:
[  631.070828]  __put_task_struct+0x55/0x160
[  631.070845]  kthread_stop+0xee/0x100
[  631.070863]  cache_set_flush+0x11d/0x1a0 [bcache]
[  631.070879]  process_one_work+0x146/0x340
[  631.070892]  worker_thread+0x47/0x3e0
[  631.070906]  kthread+0xf5/0x130
[  631.070917]  ? max_active_store+0x60/0x60
[  631.070930]  ? kthread_bind+0x10/0x10
[  631.070945]  ret_from_fork+0x35/0x40
[snipped]
[  631.071017] RIP: exit_creds+0x1b/0x50 RSP: ffffc9000705fe08
[  631.071033] CR2: 0000000000000000
[  631.071045] ---[ end trace 011c63a24b22c927 ]---
[  631.071085] bcache: bcache_device_free() bcache0 stopped

The reason is when cache_set_flush() tries to call kthread_stop() to stop
allocator thread, but it exits already due to cache device I/O errors.

This patch adds wait_for_kthread_stop() at tail of bch_allocator_thread(),
to prevent the thread routine exiting directly. Then the allocator thread
can be blocked at wait_for_kthread_stop() and wait for cache_set_flush()
to stop it by calling kthread_stop().

changelog:
v3: add Reviewed-by from Hannnes.
v2: not directly return from allocator_wait(), move 'return 0' to tail of
    bch_allocator_thread().
v1: initial version.

Fixes: 771f393e8ffc ("bcache: add CACHE_SET_IO_DISABLE to struct cache_set flags")
Signed-off-by: Coly Li <colyli@suse.de>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 drivers/md/bcache/alloc.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/md/bcache/alloc.c b/drivers/md/bcache/alloc.c
index 004cc3c..7fa2631 100644
--- a/drivers/md/bcache/alloc.c
+++ b/drivers/md/bcache/alloc.c
@@ -290,7 +290,7 @@ do {									\
 		if (kthread_should_stop() ||				\
 		    test_bit(CACHE_SET_IO_DISABLE, &ca->set->flags)) {	\
 			set_current_state(TASK_RUNNING);		\
-			return 0;					\
+			goto out;					\
 		}							\
 									\
 		schedule();						\
@@ -378,6 +378,9 @@ retry_invalidate:
 			bch_prio_write(ca);
 		}
 	}
+out:
+	wait_for_kthread_stop();
+	return 0;
 }
 
 /* Allocation */
-- 
cgit v1.1


From 4fd8e13843978cbba48b8c21119da60c7fd5910d Mon Sep 17 00:00:00 2001
From: Coly Li <colyli@suse.de>
Date: Thu, 3 May 2018 18:51:36 +0800
Subject: bcache: set dc->io_disable to true in
 conditional_stop_bcache_device()

Commit 7e027ca4b534b ("bcache: add stop_when_cache_set_failed option to
backing device") adds stop_when_cache_set_failed option and stops bcache
device if stop_when_cache_set_failed is auto and there is dirty data on
broken cache device. There might exists a small time gap that the cache
set is released and set to NULL but bcache device is not released yet
(because they are released in parallel). During this time gap, dc->c is
NULL so CACHE_SET_IO_DISABLE won't be checked, and dc->io_disable is still
false, so new coming I/O requests will be accepted and directly go into
backing device as no cache set attached to. If there is dirty data on
cache device, this behavior may introduce potential inconsistent data.

This patch sets dc->io_disable to true before calling bcache_device_stop()
to make sure the backing device will reject new coming I/O request as
well, so even in the small time gap no I/O will directly go into backing
device to corrupt data consistency.

Fixes: 7e027ca4b534b ("bcache: add stop_when_cache_set_failed option to backing device")
Signed-off-by: Coly Li <colyli@suse.de>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 drivers/md/bcache/super.c | 14 ++++++++++++++
 1 file changed, 14 insertions(+)

diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c
index c017cd4..cedbb41 100644
--- a/drivers/md/bcache/super.c
+++ b/drivers/md/bcache/super.c
@@ -1556,6 +1556,20 @@ static void conditional_stop_bcache_device(struct cache_set *c,
 		 */
 		pr_warn("stop_when_cache_set_failed of %s is \"auto\" and cache is dirty, stop it to avoid potential data corruption.",
 			d->disk->disk_name);
+			/*
+			 * There might be a small time gap that cache set is
+			 * released but bcache device is not. Inside this time
+			 * gap, regular I/O requests will directly go into
+			 * backing device as no cache set attached to. This
+			 * behavior may also introduce potential inconsistence
+			 * data in writeback mode while cache is dirty.
+			 * Therefore before calling bcache_device_stop() due
+			 * to a broken cache device, dc->io_disable should be
+			 * explicitly set to true.
+			 */
+			dc->io_disable = true;
+			/* make others know io_disable is true earlier */
+			smp_mb();
 			bcache_device_stop(d);
 	} else {
 		/*
-- 
cgit v1.1


From 09a44ca2114737e0932257619c16a2b50c7807f1 Mon Sep 17 00:00:00 2001
From: Coly Li <colyli@suse.de>
Date: Thu, 3 May 2018 18:51:37 +0800
Subject: bcache: use pr_info() to inform duplicated CACHE_SET_IO_DISABLE set

It is possible that multiple I/O requests hits on failed cache device or
backing device, therefore it is quite common that CACHE_SET_IO_DISABLE is
set already when a task tries to set the bit from bch_cache_set_error().
Currently the message "CACHE_SET_IO_DISABLE already set" is printed by
pr_warn(), which might mislead users to think a serious fault happens in
source code.

This patch uses pr_info() to print the information in such situation,
avoid extra worries. This information is helpful to understand bcache
behavior in cache device failures, so I still keep them in source code.

Fixes: 771f393e8ffc9 ("bcache: add CACHE_SET_IO_DISABLE to struct cache_set flags")
Signed-off-by: Coly Li <colyli@suse.de>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 drivers/md/bcache/super.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c
index cedbb41..3dea06b 100644
--- a/drivers/md/bcache/super.c
+++ b/drivers/md/bcache/super.c
@@ -1412,7 +1412,7 @@ bool bch_cache_set_error(struct cache_set *c, const char *fmt, ...)
 		return false;
 
 	if (test_and_set_bit(CACHE_SET_IO_DISABLE, &c->flags))
-		pr_warn("CACHE_SET_IO_DISABLE already set");
+		pr_info("CACHE_SET_IO_DISABLE already set");
 
 	/* XXX: we can be called from atomic context
 	acquire_console_sem();
-- 
cgit v1.1


From 40fa84e10134ef5c892b628e02382349b5db3e0c Mon Sep 17 00:00:00 2001
From: Arnd Bergmann <arnd@arndb.de>
Date: Wed, 4 Apr 2018 12:23:53 +0200
Subject: iommu: rockchip: fix building without CONFIG_OF

We get a build error when compiling the iommu driver without CONFIG_OF:

drivers/iommu/rockchip-iommu.c: In function 'rk_iommu_of_xlate':
drivers/iommu/rockchip-iommu.c:1101:2: error: implicit declaration of function 'of_dev_put'; did you mean 'of_node_put'? [-Werror=implicit-function-declaration]

This replaces the of_dev_put() with the equivalent
platform_device_put().

Fixes: 5fd577c3eac3 ("iommu/rockchip: Use OF_IOMMU to attach devices automatically")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Joerg Roedel <jroedel@suse.de>
---
 drivers/iommu/rockchip-iommu.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/iommu/rockchip-iommu.c b/drivers/iommu/rockchip-iommu.c
index c0f03af..0468acf 100644
--- a/drivers/iommu/rockchip-iommu.c
+++ b/drivers/iommu/rockchip-iommu.c
@@ -1098,7 +1098,7 @@ static int rk_iommu_of_xlate(struct device *dev,
 	data->iommu = platform_get_drvdata(iommu_dev);
 	dev->archdata.iommu = data;
 
-	of_dev_put(iommu_dev);
+	platform_device_put(iommu_dev);
 
 	return 0;
 }
-- 
cgit v1.1


From 9ba59b79dc699af38624e1dc337f07d99b376c27 Mon Sep 17 00:00:00 2001
From: Tarun <tarun.vyas@intel.com>
Date: Wed, 2 May 2018 16:33:00 -0700
Subject: drm/i915: Remove redundant check for negative timeout while doing an
 atomic pipe update
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

No functional changes, just a minor knit. Stumbled across the kernel doc for
schedule_timeout() which quotes "In all cases the return value is guaranteed
to be non-negative". Also, the return code of schedule_timeout() already checks
for negative values "return timeout < 0 ? 0 : timeout;" and returns 0
in such cases. Furthermore, the msec_to_jiffies returns an ungined long
value. So, let's do away with the redundant check for an atomic
pipe update.

v2: Commit message changes (Manasi).

Reviewed-by: Manasi Navare <manasi.d.navare@intel.com>
Signed-off-by: Tarun Vyas <tarun.vyas@intel.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180502233300.81220-1-tarun.vyas@intel.com
---
 drivers/gpu/drm/i915/intel_sprite.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c
index aa1dfaa..9cd4be0 100644
--- a/drivers/gpu/drm/i915/intel_sprite.c
+++ b/drivers/gpu/drm/i915/intel_sprite.c
@@ -131,7 +131,7 @@ void intel_pipe_update_start(const struct intel_crtc_state *new_crtc_state)
 		if (scanline < min || scanline > max)
 			break;
 
-		if (timeout <= 0) {
+		if (!timeout) {
 			DRM_ERROR("Potential atomic update failure on pipe %c\n",
 				  pipe_name(crtc->pipe));
 			break;
-- 
cgit v1.1


From bd4cd03c81010dcd4e6f0e02e4c15f44aefe12d1 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= <ville.syrjala@linux.intel.com>
Date: Thu, 26 Apr 2018 19:30:15 +0300
Subject: drm/i915: Correctly populate user mode h/vdisplay with pipe src size
 during readout
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

During state readout we first read out the pipe src size, store
that information in the user mode h/vdisplay, but later on we overwrite
that with the actual crtc timings. That makes our read out crtc state
inconsistent with itself when the BIOS has enabled the panel fitter to
scale the pipe contents. Let's preserve the pipe src size based
information in the user mode to make things consistent again.

This fixes a problem introduced by commit a2936e3d9a9c ("drm/i915:
Use drm_mode_get_hv_timing() to populate plane clip rectangle")
where the inconsistent state is now leading the plane clipping code
to report a failure on account the plane dst coordinates not matching
the user mode size. Previously we did the plane clipping based on
the pipe src size instead and thus never noticed the inconsistency.

The failure manifests as a WARN:
[    0.762117] [drm:intel_dump_pipe_config [i915]] requested mode:
[    0.762142] [drm:drm_mode_debug_printmodeline [drm]] Modeline 0:"1366x768" 60 72143 1366 1414 1446 1526 768 771 777 784 0x40 0xa
...
[    0.762327] [drm:intel_dump_pipe_config [i915]] port clock: 72143, pipe src size: 1024x768, pixel rate 72143
...
[    0.764666] [drm:drm_atomic_helper_check_plane_state [drm_kms_helper]] Plane must cover entire CRTC
[    0.764690] [drm:drm_rect_debug_print [drm]] dst: 1024x768+0+0
[    0.764711] [drm:drm_rect_debug_print [drm]] clip: 1366x768+0+0
[    0.764713] ------------[ cut here ]------------
[    0.764714] Could not determine valid watermarks for inherited state
[    0.764792] WARNING: CPU: 4 PID: 159 at drivers/gpu/drm/i915/intel_display.c:14584 intel_modeset_init+0x3ce/0x19d0 [i915]
...

Cc: FadeMind <fademind@gmail.com>
Cc: Dave Jones <davej@codemonkey.org.uk>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Reported-by: FadeMind <fademind@gmail.com>
Reported-by: Dave Jones <davej@codemonkey.org.uk>
Tested-by: Dave Jones <davej@codemonkey.org.uk>
References: https://lists.freedesktop.org/archives/intel-gfx/2018-April/163186.html
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105992
Fixes: a2936e3d9a9c ("drm/i915: Use drm_mode_get_hv_timing() to populate plane clip rectangle")
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180426163015.14232-1-ville.syrjala@linux.intel.com
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Tested-by: Larry Finger <Larry.Finger@lwfinger.net>
Tested-by: FadeMind <fademind@gmail.com>
---
 drivers/gpu/drm/i915/intel_display.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 1087358..3fd249c 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -15307,6 +15307,8 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 		memset(&crtc->base.mode, 0, sizeof(crtc->base.mode));
 		if (crtc_state->base.active) {
 			intel_mode_from_pipe_config(&crtc->base.mode, crtc_state);
+			crtc->base.mode.hdisplay = crtc_state->pipe_src_w;
+			crtc->base.mode.vdisplay = crtc_state->pipe_src_h;
 			intel_mode_from_pipe_config(&crtc_state->base.adjusted_mode, crtc_state);
 			WARN_ON(drm_atomic_set_mode_for_crtc(crtc->base.state, &crtc->base.mode));
 
-- 
cgit v1.1


From 5697db4a696c41601a1d15c1922150b4dbf5726c Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Bj=C3=B8rn=20Mork?= <bjorn@mork.no>
Date: Wed, 2 May 2018 22:22:54 +0200
Subject: qmi_wwan: do not steal interfaces from class drivers
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The USB_DEVICE_INTERFACE_NUMBER matching macro assumes that
the { vendorid, productid, interfacenumber } set uniquely
identifies one specific function.  This has proven to fail
for some configurable devices. One example is the Quectel
EM06/EP06 where the same interface number can be either
QMI or MBIM, without the device ID changing either.

Fix by requiring the vendor-specific class for interface number
based matching.  Functions of other classes can and should use
class based matching instead.

Fixes: 03304bcb5ec4 ("net: qmi_wwan: use fixed interface number matching")
Signed-off-by: Bjørn Mork <bjorn@mork.no>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/usb/qmi_wwan.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)

diff --git a/drivers/net/usb/qmi_wwan.c b/drivers/net/usb/qmi_wwan.c
index 51c68fc..42565dd 100644
--- a/drivers/net/usb/qmi_wwan.c
+++ b/drivers/net/usb/qmi_wwan.c
@@ -1344,6 +1344,18 @@ static int qmi_wwan_probe(struct usb_interface *intf,
 		id->driver_info = (unsigned long)&qmi_wwan_info;
 	}
 
+	/* There are devices where the same interface number can be
+	 * configured as different functions. We should only bind to
+	 * vendor specific functions when matching on interface number
+	 */
+	if (id->match_flags & USB_DEVICE_ID_MATCH_INT_NUMBER &&
+	    desc->bInterfaceClass != USB_CLASS_VENDOR_SPEC) {
+		dev_dbg(&intf->dev,
+			"Rejecting interface number match for class %02x\n",
+			desc->bInterfaceClass);
+		return -ENODEV;
+	}
+
 	/* Quectel EC20 quirk where we've QMI on interface 4 instead of 0 */
 	if (quectel_ec20_detected(intf) && desc->bInterfaceNumber == 0) {
 		dev_dbg(&intf->dev, "Quectel EC20 quirk, skipping interface 0\n");
-- 
cgit v1.1


From 8236b0ae31c837d2b3a2565c5f8d77f637e824cc Mon Sep 17 00:00:00 2001
From: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Date: Wed, 2 May 2018 07:07:55 +0900
Subject: bdi: wake up concurrent wb_shutdown() callers.

syzbot is reporting hung tasks at wait_on_bit(WB_shutting_down) in
wb_shutdown() [1]. This seems to be because commit 5318ce7d46866e1d ("bdi:
Shutdown writeback on all cgwbs in cgwb_bdi_destroy()") forgot to call
wake_up_bit(WB_shutting_down) after clear_bit(WB_shutting_down).

Introduce a helper function clear_and_wake_up_bit() and use it, in order
to avoid similar errors in future.

[1] https://syzkaller.appspot.com/bug?id=b297474817af98d5796bc544e1bb806fc3da0e5e

Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Reported-by: syzbot <syzbot+c0cf869505e03bdf1a24@syzkaller.appspotmail.com>
Fixes: 5318ce7d46866e1d ("bdi: Shutdown writeback on all cgwbs in cgwb_bdi_destroy()")
Cc: Tejun Heo <tj@kernel.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 include/linux/wait_bit.h | 17 +++++++++++++++++
 mm/backing-dev.c         |  2 +-
 2 files changed, 18 insertions(+), 1 deletion(-)

diff --git a/include/linux/wait_bit.h b/include/linux/wait_bit.h
index 9318b21..2b0072f 100644
--- a/include/linux/wait_bit.h
+++ b/include/linux/wait_bit.h
@@ -305,4 +305,21 @@ do {									\
 	__ret;								\
 })
 
+/**
+ * clear_and_wake_up_bit - clear a bit and wake up anyone waiting on that bit
+ *
+ * @bit: the bit of the word being waited on
+ * @word: the word being waited on, a kernel virtual address
+ *
+ * You can use this helper if bitflags are manipulated atomically rather than
+ * non-atomically under a lock.
+ */
+static inline void clear_and_wake_up_bit(int bit, void *word)
+{
+	clear_bit_unlock(bit, word);
+	/* See wake_up_bit() for which memory barrier you need to use. */
+	smp_mb__after_atomic();
+	wake_up_bit(word, bit);
+}
+
 #endif /* _LINUX_WAIT_BIT_H */
diff --git a/mm/backing-dev.c b/mm/backing-dev.c
index 023190c..fa5e6d7 100644
--- a/mm/backing-dev.c
+++ b/mm/backing-dev.c
@@ -383,7 +383,7 @@ static void wb_shutdown(struct bdi_writeback *wb)
 	 * the barrier provided by test_and_clear_bit() above.
 	 */
 	smp_wmb();
-	clear_bit(WB_shutting_down, &wb->state);
+	clear_and_wake_up_bit(WB_shutting_down, &wb->state);
 }
 
 static void wb_exit(struct bdi_writeback *wb)
-- 
cgit v1.1


From eb80ca476ec11f67a62691a93604b405ffc7d80c Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Wed, 2 May 2018 14:53:39 -0700
Subject: rds: do not leak kernel memory to user land

syzbot/KMSAN reported an uninit-value in put_cmsg(), originating
from rds_cmsg_recv().

Simply clear the structure, since we have holes there, or since
rx_traces might be smaller than RDS_MSG_RX_DGRAM_TRACE_MAX.

BUG: KMSAN: uninit-value in copy_to_user include/linux/uaccess.h:184 [inline]
BUG: KMSAN: uninit-value in put_cmsg+0x600/0x870 net/core/scm.c:242
CPU: 0 PID: 4459 Comm: syz-executor582 Not tainted 4.16.0+ #87
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Call Trace:
 __dump_stack lib/dump_stack.c:17 [inline]
 dump_stack+0x185/0x1d0 lib/dump_stack.c:53
 kmsan_report+0x142/0x240 mm/kmsan/kmsan.c:1067
 kmsan_internal_check_memory+0x135/0x1e0 mm/kmsan/kmsan.c:1157
 kmsan_copy_to_user+0x69/0x160 mm/kmsan/kmsan.c:1199
 copy_to_user include/linux/uaccess.h:184 [inline]
 put_cmsg+0x600/0x870 net/core/scm.c:242
 rds_cmsg_recv net/rds/recv.c:570 [inline]
 rds_recvmsg+0x2db5/0x3170 net/rds/recv.c:657
 sock_recvmsg_nosec net/socket.c:803 [inline]
 sock_recvmsg+0x1d0/0x230 net/socket.c:810
 ___sys_recvmsg+0x3fb/0x810 net/socket.c:2205
 __sys_recvmsg net/socket.c:2250 [inline]
 SYSC_recvmsg+0x298/0x3c0 net/socket.c:2262
 SyS_recvmsg+0x54/0x80 net/socket.c:2257
 do_syscall_64+0x309/0x430 arch/x86/entry/common.c:287
 entry_SYSCALL_64_after_hwframe+0x3d/0xa2

Fixes: 3289025aedc0 ("RDS: add receive message trace used by application")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: syzbot <syzkaller@googlegroups.com>
Cc: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Cc: linux-rdma <linux-rdma@vger.kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/rds/recv.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/net/rds/recv.c b/net/rds/recv.c
index de50e21..dc67458 100644
--- a/net/rds/recv.c
+++ b/net/rds/recv.c
@@ -558,6 +558,7 @@ static int rds_cmsg_recv(struct rds_incoming *inc, struct msghdr *msg,
 		struct rds_cmsg_rx_trace t;
 		int i, j;
 
+		memset(&t, 0, sizeof(t));
 		inc->i_rx_lat_trace[RDS_MSG_RX_CMSG] = local_clock();
 		t.rx_traces =  rs->rs_rx_traces;
 		for (i = 0; i < rs->rs_rx_traces; i++) {
-- 
cgit v1.1


From 114f39feab360e6c7b0c4238697f223444d662a1 Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Wed, 2 May 2018 20:25:13 -0700
Subject: tcp: restore autocorking

When adding rb-tree for TCP retransmit queue, we inadvertently broke
TCP autocorking.

tcp_should_autocork() should really check if the rtx queue is not empty.

Tested:

Before the fix :
$ nstat -n;./netperf -H 10.246.7.152 -Cc -- -m 500;nstat | grep AutoCork
MIGRATED TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.246.7.152 () port 0 AF_INET
Recv   Send    Send                          Utilization       Service Demand
Socket Socket  Message  Elapsed              Send     Recv     Send    Recv
Size   Size    Size     Time     Throughput  local    remote   local   remote
bytes  bytes   bytes    secs.    10^6bits/s  % S      % S      us/KB   us/KB

540000 262144    500    10.00      2682.85   2.47     1.59     3.618   2.329
TcpExtTCPAutoCorking            33                 0.0

// Same test, but forcing TCP_NODELAY
$ nstat -n;./netperf -H 10.246.7.152 -Cc -- -D -m 500;nstat | grep AutoCork
MIGRATED TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.246.7.152 () port 0 AF_INET : nodelay
Recv   Send    Send                          Utilization       Service Demand
Socket Socket  Message  Elapsed              Send     Recv     Send    Recv
Size   Size    Size     Time     Throughput  local    remote   local   remote
bytes  bytes   bytes    secs.    10^6bits/s  % S      % S      us/KB   us/KB

540000 262144    500    10.00      1408.75   2.44     2.96     6.802   8.259
TcpExtTCPAutoCorking            1                  0.0

After the fix :
$ nstat -n;./netperf -H 10.246.7.152 -Cc -- -m 500;nstat | grep AutoCork
MIGRATED TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.246.7.152 () port 0 AF_INET
Recv   Send    Send                          Utilization       Service Demand
Socket Socket  Message  Elapsed              Send     Recv     Send    Recv
Size   Size    Size     Time     Throughput  local    remote   local   remote
bytes  bytes   bytes    secs.    10^6bits/s  % S      % S      us/KB   us/KB

540000 262144    500    10.00      5472.46   2.45     1.43     1.761   1.027
TcpExtTCPAutoCorking            361293             0.0

// With TCP_NODELAY option
$ nstat -n;./netperf -H 10.246.7.152 -Cc -- -D -m 500;nstat | grep AutoCork
MIGRATED TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.246.7.152 () port 0 AF_INET : nodelay
Recv   Send    Send                          Utilization       Service Demand
Socket Socket  Message  Elapsed              Send     Recv     Send    Recv
Size   Size    Size     Time     Throughput  local    remote   local   remote
bytes  bytes   bytes    secs.    10^6bits/s  % S      % S      us/KB   us/KB

540000 262144    500    10.00      5454.96   2.46     1.63     1.775   1.174
TcpExtTCPAutoCorking            315448             0.0

Fixes: 75c119afe14f ("tcp: implement rb-tree based retransmit queue")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Michael Wenig <mwenig@vmware.com>
Tested-by: Michael Wenig <mwenig@vmware.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Michael Wenig <mwenig@vmware.com>
Tested-by: Michael Wenig <mwenig@vmware.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/ipv4/tcp.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 44be7f4..c9d00ef 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -697,7 +697,7 @@ static bool tcp_should_autocork(struct sock *sk, struct sk_buff *skb,
 {
 	return skb->len < size_goal &&
 	       sock_net(sk)->ipv4.sysctl_tcp_autocorking &&
-	       skb != tcp_write_queue_head(sk) &&
+	       !tcp_rtx_queue_empty(sk) &&
 	       refcount_read(&sk->sk_wmem_alloc) > skb->truesize;
 }
 
-- 
cgit v1.1


From f53823c18131e755905b4f654196fd2cc3953f6e Mon Sep 17 00:00:00 2001
From: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Date: Mon, 23 Apr 2018 11:21:03 +0900
Subject: bdi: Fix use after free bug in debugfs_remove()

syzbot is reporting use after free bug in debugfs_remove() [1].

This is because fault injection made memory allocation for
debugfs_create_file() from bdi_debug_register() from bdi_register_va()
fail and continued with setting WB_registered. But when debugfs_remove()
is called from debugfs_remove(bdi->debug_dir) from bdi_debug_unregister()
 from bdi_unregister() from release_bdi() because WB_registered was set
by bdi_register_va(), IS_ERR_OR_NULL(bdi->debug_dir) == false despite
debugfs_remove(bdi->debug_dir) was already called from bdi_register_va().

Fix this by making IS_ERR_OR_NULL(bdi->debug_dir) == true.

[1] https://syzkaller.appspot.com/bug?id=5ab4efd91a96dcea9b68104f159adf4af2a6dfc1

Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Reported-by: syzbot <syzbot+049cb4ae097049dac137@syzkaller.appspotmail.com>
Fixes: 97f07697932e6faf ("bdi: convert bdi_debug_register to int")
Cc: weiping zhang <zhangweiping@didichuxing.com>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 mm/backing-dev.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/mm/backing-dev.c b/mm/backing-dev.c
index fa5e6d7..7441bd9 100644
--- a/mm/backing-dev.c
+++ b/mm/backing-dev.c
@@ -115,6 +115,7 @@ static int bdi_debug_register(struct backing_dev_info *bdi, const char *name)
 					       bdi, &bdi_debug_stats_fops);
 	if (!bdi->debug_stats) {
 		debugfs_remove(bdi->debug_dir);
+		bdi->debug_dir = NULL;
 		return -ENOMEM;
 	}
 
-- 
cgit v1.1


From 59a2f3f00fd744dbad22593f47552037d3154ca6 Mon Sep 17 00:00:00 2001
From: Chengguang Xu <cgxu519@gmx.com>
Date: Sat, 14 Apr 2018 20:06:19 +0800
Subject: nvme: fix potential memory leak in option parsing

When specifying same string type option several times,
current option parsing may cause memory leak. Hence,
call kfree for previous one in this case.

Signed-off-by: Chengguang Xu <cgxu519@gmx.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 drivers/nvme/host/fabrics.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/nvme/host/fabrics.c b/drivers/nvme/host/fabrics.c
index 124c458..7ae732a 100644
--- a/drivers/nvme/host/fabrics.c
+++ b/drivers/nvme/host/fabrics.c
@@ -668,6 +668,7 @@ static int nvmf_parse_options(struct nvmf_ctrl_options *opts,
 				ret = -ENOMEM;
 				goto out;
 			}
+			kfree(opts->transport);
 			opts->transport = p;
 			break;
 		case NVMF_OPT_NQN:
@@ -676,6 +677,7 @@ static int nvmf_parse_options(struct nvmf_ctrl_options *opts,
 				ret = -ENOMEM;
 				goto out;
 			}
+			kfree(opts->subsysnqn);
 			opts->subsysnqn = p;
 			nqnlen = strlen(opts->subsysnqn);
 			if (nqnlen >= NVMF_NQN_SIZE) {
@@ -698,6 +700,7 @@ static int nvmf_parse_options(struct nvmf_ctrl_options *opts,
 				ret = -ENOMEM;
 				goto out;
 			}
+			kfree(opts->traddr);
 			opts->traddr = p;
 			break;
 		case NVMF_OPT_TRSVCID:
@@ -706,6 +709,7 @@ static int nvmf_parse_options(struct nvmf_ctrl_options *opts,
 				ret = -ENOMEM;
 				goto out;
 			}
+			kfree(opts->trsvcid);
 			opts->trsvcid = p;
 			break;
 		case NVMF_OPT_QUEUE_SIZE:
@@ -792,6 +796,7 @@ static int nvmf_parse_options(struct nvmf_ctrl_options *opts,
 				ret = -EINVAL;
 				goto out;
 			}
+			nvmf_host_put(opts->host);
 			opts->host = nvmf_host_add(p);
 			kfree(p);
 			if (!opts->host) {
@@ -817,6 +822,7 @@ static int nvmf_parse_options(struct nvmf_ctrl_options *opts,
 				ret = -ENOMEM;
 				goto out;
 			}
+			kfree(opts->host_traddr);
 			opts->host_traddr = p;
 			break;
 		case NVMF_OPT_HOST_ID:
-- 
cgit v1.1


From f31a21103c03bb62846409fdc60cc9faf2398cfb Mon Sep 17 00:00:00 2001
From: Keith Busch <keith.busch@intel.com>
Date: Tue, 17 Apr 2018 14:42:44 -0600
Subject: nvme: Set integrity flag for user passthrough commands

If the command a separate metadata buffer attached, the request needs
to have the integrity flag set so the driver knows to map it.

Signed-off-by: Keith Busch <keith.busch@intel.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 drivers/nvme/host/core.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c
index 9df4f71..127a9cbf 100644
--- a/drivers/nvme/host/core.c
+++ b/drivers/nvme/host/core.c
@@ -764,6 +764,7 @@ static int nvme_submit_user_cmd(struct request_queue *q,
 				ret = PTR_ERR(meta);
 				goto out_unmap;
 			}
+			req->cmd_flags |= REQ_INTEGRITY;
 		}
 	}
 
-- 
cgit v1.1


From 5cadde8019a6a80550fdde92d5a3327565974eab Mon Sep 17 00:00:00 2001
From: Keith Busch <keith.busch@intel.com>
Date: Thu, 26 Apr 2018 14:24:29 -0600
Subject: nvme/multipath: Disable runtime writable enabling parameter

We can't allow the user to change multipath settings at runtime, as this
will create naming conflicts due to the different naming schemes used
for each mode.

Signed-off-by: Keith Busch <keith.busch@intel.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 drivers/nvme/host/multipath.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/nvme/host/multipath.c b/drivers/nvme/host/multipath.c
index 956e0b8..3ded009 100644
--- a/drivers/nvme/host/multipath.c
+++ b/drivers/nvme/host/multipath.c
@@ -15,7 +15,7 @@
 #include "nvme.h"
 
 static bool multipath = true;
-module_param(multipath, bool, 0644);
+module_param(multipath, bool, 0444);
 MODULE_PARM_DESC(multipath,
 	"turn on native support for multiple controllers per subsystem");
 
-- 
cgit v1.1


From a785dbccd95c37606c720580714f5a7a8b3255f1 Mon Sep 17 00:00:00 2001
From: Keith Busch <keith.busch@intel.com>
Date: Thu, 26 Apr 2018 14:22:41 -0600
Subject: nvme/multipath: Fix multipath disabled naming collisions

When CONFIG_NVME_MULTIPATH is set, but we're not using nvme to multipath,
namespaces with multiple paths were not creating unique names due to
reusing the same instance number from the namespace's head.

This patch fixes this by falling back to the non-multipath naming method
when the parameter disabled using multipath.

Reported-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Keith Busch <keith.busch@intel.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 drivers/nvme/host/core.c      | 26 +-------------------------
 drivers/nvme/host/multipath.c | 22 ++++++++++++++++++++++
 drivers/nvme/host/nvme.h      | 12 ++++++++++++
 3 files changed, 35 insertions(+), 25 deletions(-)

diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c
index 127a9cbf..a3771c5 100644
--- a/drivers/nvme/host/core.c
+++ b/drivers/nvme/host/core.c
@@ -2998,31 +2998,7 @@ static void nvme_alloc_ns(struct nvme_ctrl *ctrl, unsigned nsid)
 	if (nvme_init_ns_head(ns, nsid, id))
 		goto out_free_id;
 	nvme_setup_streams_ns(ctrl, ns);
-	
-#ifdef CONFIG_NVME_MULTIPATH
-	/*
-	 * If multipathing is enabled we need to always use the subsystem
-	 * instance number for numbering our devices to avoid conflicts
-	 * between subsystems that have multiple controllers and thus use
-	 * the multipath-aware subsystem node and those that have a single
-	 * controller and use the controller node directly.
-	 */
-	if (ns->head->disk) {
-		sprintf(disk_name, "nvme%dc%dn%d", ctrl->subsys->instance,
-				ctrl->cntlid, ns->head->instance);
-		flags = GENHD_FL_HIDDEN;
-	} else {
-		sprintf(disk_name, "nvme%dn%d", ctrl->subsys->instance,
-				ns->head->instance);
-	}
-#else
-	/*
-	 * But without the multipath code enabled, multiple controller per
-	 * subsystems are visible as devices and thus we cannot use the
-	 * subsystem instance.
-	 */
-	sprintf(disk_name, "nvme%dn%d", ctrl->instance, ns->head->instance);
-#endif
+	nvme_set_disk_name(disk_name, ns, ctrl, &flags);
 
 	if ((ctrl->quirks & NVME_QUIRK_LIGHTNVM) && id->vs[0] == 0x1) {
 		if (nvme_nvm_register(ns, disk_name, node)) {
diff --git a/drivers/nvme/host/multipath.c b/drivers/nvme/host/multipath.c
index 3ded009..d7b664a 100644
--- a/drivers/nvme/host/multipath.c
+++ b/drivers/nvme/host/multipath.c
@@ -19,6 +19,28 @@ module_param(multipath, bool, 0444);
 MODULE_PARM_DESC(multipath,
 	"turn on native support for multiple controllers per subsystem");
 
+/*
+ * If multipathing is enabled we need to always use the subsystem instance
+ * number for numbering our devices to avoid conflicts between subsystems that
+ * have multiple controllers and thus use the multipath-aware subsystem node
+ * and those that have a single controller and use the controller node
+ * directly.
+ */
+void nvme_set_disk_name(char *disk_name, struct nvme_ns *ns,
+			struct nvme_ctrl *ctrl, int *flags)
+{
+	if (!multipath) {
+		sprintf(disk_name, "nvme%dn%d", ctrl->instance, ns->head->instance);
+	} else if (ns->head->disk) {
+		sprintf(disk_name, "nvme%dc%dn%d", ctrl->subsys->instance,
+				ctrl->cntlid, ns->head->instance);
+		*flags = GENHD_FL_HIDDEN;
+	} else {
+		sprintf(disk_name, "nvme%dn%d", ctrl->subsys->instance,
+				ns->head->instance);
+	}
+}
+
 void nvme_failover_req(struct request *req)
 {
 	struct nvme_ns *ns = req->q->queuedata;
diff --git a/drivers/nvme/host/nvme.h b/drivers/nvme/host/nvme.h
index 061fecf..7ded7a5 100644
--- a/drivers/nvme/host/nvme.h
+++ b/drivers/nvme/host/nvme.h
@@ -436,6 +436,8 @@ extern const struct attribute_group nvme_ns_id_attr_group;
 extern const struct block_device_operations nvme_ns_head_ops;
 
 #ifdef CONFIG_NVME_MULTIPATH
+void nvme_set_disk_name(char *disk_name, struct nvme_ns *ns,
+			struct nvme_ctrl *ctrl, int *flags);
 void nvme_failover_req(struct request *req);
 bool nvme_req_needs_failover(struct request *req, blk_status_t error);
 void nvme_kick_requeue_lists(struct nvme_ctrl *ctrl);
@@ -461,6 +463,16 @@ static inline void nvme_mpath_check_last_path(struct nvme_ns *ns)
 }
 
 #else
+/*
+ * Without the multipath code enabled, multiple controller per subsystems are
+ * visible as devices and thus we cannot use the subsystem instance.
+ */
+static inline void nvme_set_disk_name(char *disk_name, struct nvme_ns *ns,
+				      struct nvme_ctrl *ctrl, int *flags)
+{
+	sprintf(disk_name, "nvme%dn%d", ctrl->instance, ns->head->instance);
+}
+
 static inline void nvme_failover_req(struct request *req)
 {
 }
-- 
cgit v1.1


From 8bfc3b4c6f9de815de4ab73784b9419348266a65 Mon Sep 17 00:00:00 2001
From: Johannes Thumshirn <jthumshirn@suse.de>
Date: Thu, 3 May 2018 17:00:35 +0200
Subject: nvmet: switch loopback target state to connecting when resetting

After commit bb06ec31452f ("nvme: expand nvmf_check_if_ready checks")
resetting of the loopback nvme target failed as we forgot to switch
it's state to NVME_CTRL_CONNECTING before we reconnect the admin
queues. Therefore the checks in nvmf_check_if_ready() choose to go to
the reject_io case and thus we couldn't sent out an identify
controller command to reconnect.

Change the controller state to NVME_CTRL_CONNECTING after tearing down
the old connection and before re-establishing the connection.

Fixes: bb06ec31452f ("nvme: expand nvmf_check_if_ready checks")
Signed-off-by: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 drivers/nvme/target/loop.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/nvme/target/loop.c b/drivers/nvme/target/loop.c
index 31fdfba..27a8561 100644
--- a/drivers/nvme/target/loop.c
+++ b/drivers/nvme/target/loop.c
@@ -469,6 +469,12 @@ static void nvme_loop_reset_ctrl_work(struct work_struct *work)
 	nvme_stop_ctrl(&ctrl->ctrl);
 	nvme_loop_shutdown_ctrl(ctrl);
 
+	if (!nvme_change_ctrl_state(&ctrl->ctrl, NVME_CTRL_CONNECTING)) {
+		/* state change failure should never happen */
+		WARN_ON_ONCE(1);
+		return;
+	}
+
 	ret = nvme_loop_configure_admin_queue(ctrl);
 	if (ret)
 		goto out_disable;
-- 
cgit v1.1


From 1ce0500d234f8ef880c399d55a886af646beec9a Mon Sep 17 00:00:00 2001
From: Chen LinX <linx.z.chen@intel.com>
Date: Wed, 3 Sep 2014 14:31:09 +0800
Subject: ftrace: Have set_graph_* files have normal file modes

The set_graph_function and set_graph_notrace file mode should be 0644
instead of 0444 as they are writeable. Note, the mode appears to be ignored
regardless, but they should at least look sane.

Link: http://lkml.kernel.org/r/1409725869-4501-1-git-send-email-linx.z.chen@intel.com

Acked-by: Namhyung Kim <namhyung@kernel.org>
Signed-off-by: Chen LinX <linx.z.chen@intel.com>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
---
 kernel/trace/ftrace.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c
index 16bbf06..8d83bcf 100644
--- a/kernel/trace/ftrace.c
+++ b/kernel/trace/ftrace.c
@@ -5514,10 +5514,10 @@ static __init int ftrace_init_dyn_tracefs(struct dentry *d_tracer)
 	ftrace_create_filter_files(&global_ops, d_tracer);
 
 #ifdef CONFIG_FUNCTION_GRAPH_TRACER
-	trace_create_file("set_graph_function", 0444, d_tracer,
+	trace_create_file("set_graph_function", 0644, d_tracer,
 				    NULL,
 				    &ftrace_graph_fops);
-	trace_create_file("set_graph_notrace", 0444, d_tracer,
+	trace_create_file("set_graph_notrace", 0644, d_tracer,
 				    NULL,
 				    &ftrace_graph_notrace_fops);
 #endif /* CONFIG_FUNCTION_GRAPH_TRACER */
-- 
cgit v1.1


From 44a182b9d17765514fa2b1cc911e4e65134eef93 Mon Sep 17 00:00:00 2001
From: Mathias Nyman <mathias.nyman@linux.intel.com>
Date: Thu, 3 May 2018 17:30:07 +0300
Subject: xhci: Fix use-after-free in xhci_free_virt_device

KASAN found a use-after-free in xhci_free_virt_device+0x33b/0x38e
where xhci_free_virt_device() sets slot id to 0 if udev exists:
if (dev->udev && dev->udev->slot_id)
	dev->udev->slot_id = 0;

dev->udev will be true even if udev is freed because dev->udev is
not set to NULL.

set dev->udev pointer to NULL in xhci_free_dev()

The original patch went to stable so this fix needs to be applied
there as well.

Fixes: a400efe455f7 ("xhci: zero usb device slot_id member when disabling and freeing a xhci slot")
Cc: <stable@vger.kernel.org>
Reported-by: Guenter Roeck <linux@roeck-us.net>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Tested-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 drivers/usb/host/xhci.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/usb/host/xhci.c b/drivers/usb/host/xhci.c
index 9b27798..711da33 100644
--- a/drivers/usb/host/xhci.c
+++ b/drivers/usb/host/xhci.c
@@ -3621,6 +3621,7 @@ static void xhci_free_dev(struct usb_hcd *hcd, struct usb_device *udev)
 		del_timer_sync(&virt_dev->eps[i].stop_cmd_timer);
 	}
 	xhci_debugfs_remove_slot(xhci, udev->slot_id);
+	virt_dev->udev = NULL;
 	ret = xhci_disable_slot(xhci, udev->slot_id);
 	if (ret)
 		xhci_free_virt_device(xhci, udev->slot_id);
-- 
cgit v1.1


From 0c5a9acc8b4e878e761f735e144d4a7e4477d4e6 Mon Sep 17 00:00:00 2001
From: Zhengyuan Liu <liuzhengyuan@kylinos.cn>
Date: Thu, 8 Feb 2018 09:41:53 +0800
Subject: tracing: Fix the file mode of stack tracer

It looks weird that the stack_trace_filter file can be written by root
but shows that it does not have write permission by ll command.

Link: http://lkml.kernel.org/r/1518054113-28096-1-git-send-email-liuzhengyuan@kylinos.cn

Signed-off-by: Zhengyuan Liu <liuzhengyuan@kylinos.cn>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
---
 kernel/trace/trace_stack.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/kernel/trace/trace_stack.c b/kernel/trace/trace_stack.c
index 3c7bfc4..4237eba 100644
--- a/kernel/trace/trace_stack.c
+++ b/kernel/trace/trace_stack.c
@@ -472,7 +472,7 @@ static __init int stack_trace_init(void)
 			NULL, &stack_trace_fops);
 
 #ifdef CONFIG_DYNAMIC_FTRACE
-	trace_create_file("stack_trace_filter", 0444, d_tracer,
+	trace_create_file("stack_trace_filter", 0644, d_tracer,
 			  &trace_ops, &stack_trace_filter_fops);
 #endif
 
-- 
cgit v1.1


From 4e11581c27a28503282e777ce75502f560c6f97b Mon Sep 17 00:00:00 2001
From: Colin Ian King <colin.king@canonical.com>
Date: Thu, 3 May 2018 10:12:53 +0100
Subject: net/mlx5e: fix spelling mistake: "loobpack" -> "loopback"

Trivial fix to spelling mistake in netdev_err error message

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_selftest.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_selftest.c b/drivers/net/ethernet/mellanox/mlx5/core/en_selftest.c
index 7079764..027f54a 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_selftest.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_selftest.c
@@ -290,7 +290,7 @@ static int mlx5e_test_loopback(struct mlx5e_priv *priv)
 
 	if (!test_bit(MLX5E_STATE_OPENED, &priv->state)) {
 		netdev_err(priv->netdev,
-			   "\tCan't perform loobpack test while device is down\n");
+			   "\tCan't perform loopback test while device is down\n");
 		return -ENODEV;
 	}
 
-- 
cgit v1.1


From fb5ee84ea72c5f1b6cabdd1c9d6e8648995ca7c6 Mon Sep 17 00:00:00 2001
From: Alan Stern <stern@rowland.harvard.edu>
Date: Thu, 3 May 2018 11:04:48 -0400
Subject: USB: Accept bulk endpoints with 1024-byte maxpacket

Some non-compliant high-speed USB devices have bulk endpoints with a
1024-byte maxpacket size.  Although such endpoints don't work with
xHCI host controllers, they do work with EHCI controllers.  We used to
accept these invalid sizes (with a warning), but we no longer do
because of an unintentional change introduced by commit aed9d65ac327
("USB: validate wMaxPacketValue entries in endpoint descriptors").

This patch restores the old behavior, so that people with these
peculiar devices can use them without patching their kernels by hand.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Suggested-by: Elvinas <elvinas@veikia.lt>
Fixes: aed9d65ac327 ("USB: validate wMaxPacketValue entries in endpoint descriptors")
CC: <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 drivers/usb/core/config.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/usb/core/config.c b/drivers/usb/core/config.c
index c821b4b..7b5cb28 100644
--- a/drivers/usb/core/config.c
+++ b/drivers/usb/core/config.c
@@ -191,7 +191,9 @@ static const unsigned short full_speed_maxpacket_maxes[4] = {
 static const unsigned short high_speed_maxpacket_maxes[4] = {
 	[USB_ENDPOINT_XFER_CONTROL] = 64,
 	[USB_ENDPOINT_XFER_ISOC] = 1024,
-	[USB_ENDPOINT_XFER_BULK] = 512,
+
+	/* Bulk should be 512, but some devices use 1024: we will warn below */
+	[USB_ENDPOINT_XFER_BULK] = 1024,
 	[USB_ENDPOINT_XFER_INT] = 1024,
 };
 static const unsigned short super_speed_maxpacket_maxes[4] = {
-- 
cgit v1.1


From 5671a4608c326b8ec16dec2f0f32e64a33cdd317 Mon Sep 17 00:00:00 2001
From: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Date: Mon, 30 Apr 2018 15:41:56 +0300
Subject: usb: typec: tcpm: Release the role mux when exiting

The ref count for the USB role switch device must be
released after we are done using the switch.

Fixes: c6962c29729c ("usb: typec: tcpm: Set USB role switch to device mode when configured as such")
Signed-off-by: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 drivers/usb/typec/tcpm.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/usb/typec/tcpm.c b/drivers/usb/typec/tcpm.c
index 677d121..ded49e3 100644
--- a/drivers/usb/typec/tcpm.c
+++ b/drivers/usb/typec/tcpm.c
@@ -3725,6 +3725,7 @@ void tcpm_unregister_port(struct tcpm_port *port)
 	for (i = 0; i < ARRAY_SIZE(port->port_altmode); i++)
 		typec_unregister_altmode(port->port_altmode[i]);
 	typec_unregister_port(port->typec_port);
+	usb_role_switch_put(port->role_sw);
 	tcpm_debugfs_exit(port);
 	destroy_workqueue(port->wq);
 }
-- 
cgit v1.1


From 1a2f474d328f292ee706414824ec4ca690cdf5ba Mon Sep 17 00:00:00 2001
From: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Date: Wed, 25 Apr 2018 17:22:09 +0300
Subject: usb: typec: tps6598x: handle block reads separately with plain-I2C
 adapters

If the I2C adapter that the PD controller is attached to
does not support SMBus protocol, the driver needs to handle
block reads separately. The first byte returned in block
read protocol will show the total number of bytes. It needs
to be stripped away.

This is handled separately in the driver only because right
now we have no way of requesting the used protocol with
regmap-i2c. This is in practice a workaround for what is
really a problem in regmap-i2c. The other option would have
been to register custom regmap, or not use regmap at all,
however, since the solution is very simple, I choose to use
it in this case for convenience. It is easy to remove once
we figure out how to handle this kind of cases in
regmap-i2c.

Fixes: 0a4c005bd171 ("usb: typec: driver for TI TPS6598x USB Power Delivery controllers")
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 drivers/usb/typec/tps6598x.c | 47 ++++++++++++++++++++++++++++++++++++--------
 1 file changed, 39 insertions(+), 8 deletions(-)

diff --git a/drivers/usb/typec/tps6598x.c b/drivers/usb/typec/tps6598x.c
index 8b84068..4b4c8d2 100644
--- a/drivers/usb/typec/tps6598x.c
+++ b/drivers/usb/typec/tps6598x.c
@@ -73,6 +73,7 @@ struct tps6598x {
 	struct device *dev;
 	struct regmap *regmap;
 	struct mutex lock; /* device lock */
+	u8 i2c_protocol:1;
 
 	struct typec_port *port;
 	struct typec_partner *partner;
@@ -80,19 +81,39 @@ struct tps6598x {
 	struct typec_capability typec_cap;
 };
 
+static int
+tps6598x_block_read(struct tps6598x *tps, u8 reg, void *val, size_t len)
+{
+	u8 data[len + 1];
+	int ret;
+
+	if (!tps->i2c_protocol)
+		return regmap_raw_read(tps->regmap, reg, val, len);
+
+	ret = regmap_raw_read(tps->regmap, reg, data, sizeof(data));
+	if (ret)
+		return ret;
+
+	if (data[0] < len)
+		return -EIO;
+
+	memcpy(val, &data[1], len);
+	return 0;
+}
+
 static inline int tps6598x_read16(struct tps6598x *tps, u8 reg, u16 *val)
 {
-	return regmap_raw_read(tps->regmap, reg, val, sizeof(u16));
+	return tps6598x_block_read(tps, reg, val, sizeof(u16));
 }
 
 static inline int tps6598x_read32(struct tps6598x *tps, u8 reg, u32 *val)
 {
-	return regmap_raw_read(tps->regmap, reg, val, sizeof(u32));
+	return tps6598x_block_read(tps, reg, val, sizeof(u32));
 }
 
 static inline int tps6598x_read64(struct tps6598x *tps, u8 reg, u64 *val)
 {
-	return regmap_raw_read(tps->regmap, reg, val, sizeof(u64));
+	return tps6598x_block_read(tps, reg, val, sizeof(u64));
 }
 
 static inline int tps6598x_write16(struct tps6598x *tps, u8 reg, u16 val)
@@ -121,8 +142,8 @@ static int tps6598x_read_partner_identity(struct tps6598x *tps)
 	struct tps6598x_rx_identity_reg id;
 	int ret;
 
-	ret = regmap_raw_read(tps->regmap, TPS_REG_RX_IDENTITY_SOP,
-			      &id, sizeof(id));
+	ret = tps6598x_block_read(tps, TPS_REG_RX_IDENTITY_SOP,
+				  &id, sizeof(id));
 	if (ret)
 		return ret;
 
@@ -224,13 +245,13 @@ static int tps6598x_exec_cmd(struct tps6598x *tps, const char *cmd,
 	} while (val);
 
 	if (out_len) {
-		ret = regmap_raw_read(tps->regmap, TPS_REG_DATA1,
-				      out_data, out_len);
+		ret = tps6598x_block_read(tps, TPS_REG_DATA1,
+					  out_data, out_len);
 		if (ret)
 			return ret;
 		val = out_data[0];
 	} else {
-		ret = regmap_read(tps->regmap, TPS_REG_DATA1, &val);
+		ret = tps6598x_block_read(tps, TPS_REG_DATA1, &val, sizeof(u8));
 		if (ret)
 			return ret;
 	}
@@ -385,6 +406,16 @@ static int tps6598x_probe(struct i2c_client *client)
 	if (!vid)
 		return -ENODEV;
 
+	/*
+	 * Checking can the adapter handle SMBus protocol. If it can not, the
+	 * driver needs to take care of block reads separately.
+	 *
+	 * FIXME: Testing with I2C_FUNC_I2C. regmap-i2c uses I2C protocol
+	 * unconditionally if the adapter has I2C_FUNC_I2C set.
+	 */
+	if (i2c_check_functionality(client->adapter, I2C_FUNC_I2C))
+		tps->i2c_protocol = true;
+
 	ret = tps6598x_read32(tps, TPS_REG_STATUS, &status);
 	if (ret < 0)
 		return ret;
-- 
cgit v1.1


From df80b8fb3c0ef510649d2d6e350cf11be240d15c Mon Sep 17 00:00:00 2001
From: Colin Ian King <colin.king@canonical.com>
Date: Thu, 3 May 2018 16:19:32 +0100
Subject: qed: fix spelling mistake: "offloded" -> "offloaded"

Trivial fix to spelling mistake in DP_NOTICE message

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/qlogic/qed/qed_roce.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/qlogic/qed/qed_roce.c b/drivers/net/ethernet/qlogic/qed/qed_roce.c
index fb7c2d1..6acfd43 100644
--- a/drivers/net/ethernet/qlogic/qed/qed_roce.c
+++ b/drivers/net/ethernet/qlogic/qed/qed_roce.c
@@ -848,7 +848,7 @@ int qed_roce_query_qp(struct qed_hwfn *p_hwfn,
 
 	if (!(qp->resp_offloaded)) {
 		DP_NOTICE(p_hwfn,
-			  "The responder's qp should be offloded before requester's\n");
+			  "The responder's qp should be offloaded before requester's\n");
 		return -EINVAL;
 	}
 
-- 
cgit v1.1


From e63a5f8c19d7807823d68830ebe8cfbd4419ab13 Mon Sep 17 00:00:00 2001
From: Karsten Graul <kgraul@linux.ibm.com>
Date: Thu, 3 May 2018 17:57:37 +0200
Subject: net/smc: call consolidation

Consolidate the call to smc_wr_reg_send() in a new function.
No functional changes.

Signed-off-by: Karsten Graul <kgraul@linux.ibm.com>
Signed-off-by: Ursula Braun <ubraun@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/smc/af_smc.c | 35 +++++++++++++++--------------------
 1 file changed, 15 insertions(+), 20 deletions(-)

diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c
index 8b4c059..fdb2976 100644
--- a/net/smc/af_smc.c
+++ b/net/smc/af_smc.c
@@ -292,6 +292,15 @@ static void smc_copy_sock_settings_to_smc(struct smc_sock *smc)
 	smc_copy_sock_settings(&smc->sk, smc->clcsock->sk, SK_FLAGS_CLC_TO_SMC);
 }
 
+/* register a new rmb */
+static int smc_reg_rmb(struct smc_link *link, struct smc_buf_desc *rmb_desc)
+{
+	/* register memory region for new rmb */
+	if (smc_wr_reg_send(link, rmb_desc->mr_rx[SMC_SINGLE_LINK]))
+		return -EFAULT;
+	return 0;
+}
+
 static int smc_clnt_conf_first_link(struct smc_sock *smc)
 {
 	struct smc_link_group *lgr = smc->conn.lgr;
@@ -321,9 +330,7 @@ static int smc_clnt_conf_first_link(struct smc_sock *smc)
 
 	smc_wr_remember_qp_attr(link);
 
-	rc = smc_wr_reg_send(link,
-			     smc->conn.rmb_desc->mr_rx[SMC_SINGLE_LINK]);
-	if (rc)
+	if (smc_reg_rmb(link, smc->conn.rmb_desc))
 		return SMC_CLC_DECL_INTERR;
 
 	/* send CONFIRM LINK response over RoCE fabric */
@@ -473,13 +480,8 @@ static int smc_connect_rdma(struct smc_sock *smc)
 			goto decline_rdma_unlock;
 		}
 	} else {
-		struct smc_buf_desc *buf_desc = smc->conn.rmb_desc;
-
-		if (!buf_desc->reused) {
-			/* register memory region for new rmb */
-			rc = smc_wr_reg_send(link,
-					     buf_desc->mr_rx[SMC_SINGLE_LINK]);
-			if (rc) {
+		if (!smc->conn.rmb_desc->reused) {
+			if (smc_reg_rmb(link, smc->conn.rmb_desc)) {
 				reason_code = SMC_CLC_DECL_INTERR;
 				goto decline_rdma_unlock;
 			}
@@ -719,9 +721,7 @@ static int smc_serv_conf_first_link(struct smc_sock *smc)
 
 	link = &lgr->lnk[SMC_SINGLE_LINK];
 
-	rc = smc_wr_reg_send(link,
-			     smc->conn.rmb_desc->mr_rx[SMC_SINGLE_LINK]);
-	if (rc)
+	if (smc_reg_rmb(link, smc->conn.rmb_desc))
 		return SMC_CLC_DECL_INTERR;
 
 	/* send CONFIRM LINK request to client over the RoCE fabric */
@@ -854,13 +854,8 @@ static void smc_listen_work(struct work_struct *work)
 	smc_rx_init(new_smc);
 
 	if (local_contact != SMC_FIRST_CONTACT) {
-		struct smc_buf_desc *buf_desc = new_smc->conn.rmb_desc;
-
-		if (!buf_desc->reused) {
-			/* register memory region for new rmb */
-			rc = smc_wr_reg_send(link,
-					     buf_desc->mr_rx[SMC_SINGLE_LINK]);
-			if (rc) {
+		if (!new_smc->conn.rmb_desc->reused) {
+			if (smc_reg_rmb(link, new_smc->conn.rmb_desc)) {
 				reason_code = SMC_CLC_DECL_INTERR;
 				goto decline_rdma_unlock;
 			}
-- 
cgit v1.1


From a6920d1d130c3de039be982eba42542d329dc64c Mon Sep 17 00:00:00 2001
From: Karsten Graul <kgraul@linux.ibm.com>
Date: Thu, 3 May 2018 17:57:38 +0200
Subject: net/smc: handle unregistered buffers

When smc_wr_reg_send() fails then tag (regerr) the affected buffer and
free it in smc_buf_unuse().

Signed-off-by: Karsten Graul <kgraul@linux.ibm.com>
Signed-off-by: Ursula Braun <ubraun@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/smc/af_smc.c   |  4 +++-
 net/smc/smc_core.c | 22 +++++++++++++++++++---
 net/smc/smc_core.h |  3 ++-
 3 files changed, 24 insertions(+), 5 deletions(-)

diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c
index fdb2976..d03b8d2 100644
--- a/net/smc/af_smc.c
+++ b/net/smc/af_smc.c
@@ -296,8 +296,10 @@ static void smc_copy_sock_settings_to_smc(struct smc_sock *smc)
 static int smc_reg_rmb(struct smc_link *link, struct smc_buf_desc *rmb_desc)
 {
 	/* register memory region for new rmb */
-	if (smc_wr_reg_send(link, rmb_desc->mr_rx[SMC_SINGLE_LINK]))
+	if (smc_wr_reg_send(link, rmb_desc->mr_rx[SMC_SINGLE_LINK])) {
+		rmb_desc->regerr = 1;
 		return -EFAULT;
+	}
 	return 0;
 }
 
diff --git a/net/smc/smc_core.c b/net/smc/smc_core.c
index f44f680..d4bd01b 100644
--- a/net/smc/smc_core.c
+++ b/net/smc/smc_core.c
@@ -32,6 +32,9 @@
 
 static u32 smc_lgr_num;			/* unique link group number */
 
+static void smc_buf_free(struct smc_buf_desc *buf_desc, struct smc_link *lnk,
+			 bool is_rmb);
+
 static void smc_lgr_schedule_free_work(struct smc_link_group *lgr)
 {
 	/* client link group creation always follows the server link group
@@ -234,9 +237,22 @@ static void smc_buf_unuse(struct smc_connection *conn)
 		conn->sndbuf_size = 0;
 	}
 	if (conn->rmb_desc) {
-		conn->rmb_desc->reused = true;
-		conn->rmb_desc->used = 0;
-		conn->rmbe_size = 0;
+		if (!conn->rmb_desc->regerr) {
+			conn->rmb_desc->reused = 1;
+			conn->rmb_desc->used = 0;
+			conn->rmbe_size = 0;
+		} else {
+			/* buf registration failed, reuse not possible */
+			struct smc_link_group *lgr = conn->lgr;
+			struct smc_link *lnk;
+
+			write_lock_bh(&lgr->rmbs_lock);
+			list_del(&conn->rmb_desc->list);
+			write_unlock_bh(&lgr->rmbs_lock);
+
+			lnk = &lgr->lnk[SMC_SINGLE_LINK];
+			smc_buf_free(conn->rmb_desc, lnk, true);
+		}
 	}
 }
 
diff --git a/net/smc/smc_core.h b/net/smc/smc_core.h
index 07e2a39..5dfcb15 100644
--- a/net/smc/smc_core.h
+++ b/net/smc/smc_core.h
@@ -123,7 +123,8 @@ struct smc_buf_desc {
 						 */
 	u32			order;		/* allocation order */
 	u32			used;		/* currently used / unused */
-	bool			reused;		/* new created / reused */
+	u8			reused	: 1;	/* new created / reused */
+	u8			regerr	: 1;	/* err during registration */
 };
 
 struct smc_rtoken {				/* address/key of remote RMB */
-- 
cgit v1.1


From bda27ff5c4526f80a7620a94ecfe8dca153e3696 Mon Sep 17 00:00:00 2001
From: Stefan Raspl <stefan.raspl@linux.ibm.com>
Date: Thu, 3 May 2018 17:57:39 +0200
Subject: smc: fix sendpage() call

The sendpage() call grabs the sock lock before calling the default
implementation - which tries to grab it once again.

Signed-off-by: Stefan Raspl <raspl@linux.ibm.com>
Signed-off-by: Ursula Braun <ubraun@linux.ibm.com><
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/smc/af_smc.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c
index d03b8d2..544bab4 100644
--- a/net/smc/af_smc.c
+++ b/net/smc/af_smc.c
@@ -1315,8 +1315,11 @@ static ssize_t smc_sendpage(struct socket *sock, struct page *page,
 
 	smc = smc_sk(sk);
 	lock_sock(sk);
-	if (sk->sk_state != SMC_ACTIVE)
+	if (sk->sk_state != SMC_ACTIVE) {
+		release_sock(sk);
 		goto out;
+	}
+	release_sock(sk);
 	if (smc->use_fallback)
 		rc = kernel_sendpage(smc->clcsock, page, offset,
 				     size, flags);
@@ -1324,7 +1327,6 @@ static ssize_t smc_sendpage(struct socket *sock, struct page *page,
 		rc = sock_no_sendpage(sock, page, offset, size, flags);
 
 out:
-	release_sock(sk);
 	return rc;
 }
 
-- 
cgit v1.1


From a8d7aa17bbc970971ccdf71988ea19230ab368b1 Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Thu, 3 May 2018 09:39:20 -0700
Subject: dccp: fix tasklet usage

syzbot reported a crash in tasklet_action_common() caused by dccp.

dccp needs to make sure socket wont disappear before tasklet handler
has completed.

This patch takes a reference on the socket when arming the tasklet,
and moves the sock_put() from dccp_write_xmit_timer() to dccp_write_xmitlet()

kernel BUG at kernel/softirq.c:514!
invalid opcode: 0000 [#1] SMP KASAN
Dumping ftrace buffer:
   (ftrace buffer empty)
Modules linked in:
CPU: 1 PID: 17 Comm: ksoftirqd/1 Not tainted 4.17.0-rc3+ #30
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
RIP: 0010:tasklet_action_common.isra.19+0x6db/0x700 kernel/softirq.c:515
RSP: 0018:ffff8801d9b3faf8 EFLAGS: 00010246
dccp_close: ABORT with 65423 bytes unread
RAX: 1ffff1003b367f6b RBX: ffff8801daf1f3f0 RCX: 0000000000000000
RDX: ffff8801cf895498 RSI: 0000000000000004 RDI: 0000000000000000
RBP: ffff8801d9b3fc40 R08: ffffed0039f12a95 R09: ffffed0039f12a94
dccp_close: ABORT with 65423 bytes unread
R10: ffffed0039f12a94 R11: ffff8801cf8954a3 R12: 0000000000000000
R13: ffff8801d9b3fc18 R14: dffffc0000000000 R15: ffff8801cf895490
FS:  0000000000000000(0000) GS:ffff8801daf00000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000001b2bc28000 CR3: 00000001a08a9000 CR4: 00000000001406e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
 tasklet_action+0x1d/0x20 kernel/softirq.c:533
 __do_softirq+0x2e0/0xaf5 kernel/softirq.c:285
dccp_close: ABORT with 65423 bytes unread
 run_ksoftirqd+0x86/0x100 kernel/softirq.c:646
 smpboot_thread_fn+0x417/0x870 kernel/smpboot.c:164
 kthread+0x345/0x410 kernel/kthread.c:238
 ret_from_fork+0x3a/0x50 arch/x86/entry/entry_64.S:412
Code: 48 8b 85 e8 fe ff ff 48 8b 95 f0 fe ff ff e9 94 fb ff ff 48 89 95 f0 fe ff ff e8 81 53 6e 00 48 8b 95 f0 fe ff ff e9 62 fb ff ff <0f> 0b 48 89 cf 48 89 8d e8 fe ff ff e8 64 53 6e 00 48 8b 8d e8
RIP: tasklet_action_common.isra.19+0x6db/0x700 kernel/softirq.c:515 RSP: ffff8801d9b3faf8

Fixes: dc841e30eaea ("dccp: Extend CCID packet dequeueing interface")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: syzbot <syzkaller@googlegroups.com>
Cc: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Cc: dccp@vger.kernel.org
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/dccp/ccids/ccid2.c | 14 ++++++++++++--
 net/dccp/timer.c       |  2 +-
 2 files changed, 13 insertions(+), 3 deletions(-)

diff --git a/net/dccp/ccids/ccid2.c b/net/dccp/ccids/ccid2.c
index 92d016e..385f153 100644
--- a/net/dccp/ccids/ccid2.c
+++ b/net/dccp/ccids/ccid2.c
@@ -126,6 +126,16 @@ static void ccid2_change_l_seq_window(struct sock *sk, u64 val)
 						  DCCPF_SEQ_WMAX));
 }
 
+static void dccp_tasklet_schedule(struct sock *sk)
+{
+	struct tasklet_struct *t = &dccp_sk(sk)->dccps_xmitlet;
+
+	if (!test_and_set_bit(TASKLET_STATE_SCHED, &t->state)) {
+		sock_hold(sk);
+		__tasklet_schedule(t);
+	}
+}
+
 static void ccid2_hc_tx_rto_expire(struct timer_list *t)
 {
 	struct ccid2_hc_tx_sock *hc = from_timer(hc, t, tx_rtotimer);
@@ -166,7 +176,7 @@ static void ccid2_hc_tx_rto_expire(struct timer_list *t)
 
 	/* if we were blocked before, we may now send cwnd=1 packet */
 	if (sender_was_blocked)
-		tasklet_schedule(&dccp_sk(sk)->dccps_xmitlet);
+		dccp_tasklet_schedule(sk);
 	/* restart backed-off timer */
 	sk_reset_timer(sk, &hc->tx_rtotimer, jiffies + hc->tx_rto);
 out:
@@ -706,7 +716,7 @@ static void ccid2_hc_tx_packet_recv(struct sock *sk, struct sk_buff *skb)
 done:
 	/* check if incoming Acks allow pending packets to be sent */
 	if (sender_was_blocked && !ccid2_cwnd_network_limited(hc))
-		tasklet_schedule(&dccp_sk(sk)->dccps_xmitlet);
+		dccp_tasklet_schedule(sk);
 	dccp_ackvec_parsed_cleanup(&hc->tx_av_chunks);
 }
 
diff --git a/net/dccp/timer.c b/net/dccp/timer.c
index b50a873..1501a20 100644
--- a/net/dccp/timer.c
+++ b/net/dccp/timer.c
@@ -232,6 +232,7 @@ static void dccp_write_xmitlet(unsigned long data)
 	else
 		dccp_write_xmit(sk);
 	bh_unlock_sock(sk);
+	sock_put(sk);
 }
 
 static void dccp_write_xmit_timer(struct timer_list *t)
@@ -240,7 +241,6 @@ static void dccp_write_xmit_timer(struct timer_list *t)
 	struct sock *sk = &dp->dccps_inet_connection.icsk_inet.sk;
 
 	dccp_write_xmitlet((unsigned long)sk);
-	sock_put(sk);
 }
 
 void dccp_init_xmit_timers(struct sock *sk)
-- 
cgit v1.1


From f59fb9e05109b836230813e45f71c9ecc2d5dbe6 Mon Sep 17 00:00:00 2001
From: Mike Marciniszyn <mike.marciniszyn@intel.com>
Date: Tue, 1 May 2018 05:35:36 -0700
Subject: IB/hfi1: Fix handling of FECN marked multicast packet

The code for handling a marked UD packet unconditionally returns the
dlid in the header of the FECN marked packet.  This is not correct
for multicast packets where the DLID is in the multicast range.

The subsequent attempt to send the CNP with the multicast lid will
cause the chip to halt the ack send context because the source
lid doesn't match the chip programming.   The send context will
be halted and flush any other pending packets in the pio ring causing
the CNP to not be sent.

A part of investigating the fix, it was determined that the 16B work
broke the FECN routine badly with inconsistent use of 16 bit and 32 bits
types for lids and pkeys.  Since the port's source lid was correctly 32
bits the type mixmatches need to be dealt with at the same time as
fixing the CNP header issue.

Fix these issues by:
- Using the ports lid for as the SLID for responding to FECN marked UD
  packets
- Insure pkey is always 16 bit in this and subordinate routines
- Insure lids are 32 bits in this and subordinate routines

Cc: <stable@vger.kernel.org> # 4.14.x
Fixes: 88733e3b8450 ("IB/hfi1: Add 16B UD support")
Reviewed-by: Don Hiatt <don.hiatt@intel.com>
Reviewed-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/driver.c | 19 +++++++++++++++----
 drivers/infiniband/hw/hfi1/hfi.h    |  8 ++++----
 drivers/infiniband/hw/hfi1/ud.c     |  4 ++--
 3 files changed, 21 insertions(+), 10 deletions(-)

diff --git a/drivers/infiniband/hw/hfi1/driver.c b/drivers/infiniband/hw/hfi1/driver.c
index 46d1475..bd837a0 100644
--- a/drivers/infiniband/hw/hfi1/driver.c
+++ b/drivers/infiniband/hw/hfi1/driver.c
@@ -433,31 +433,43 @@ void hfi1_process_ecn_slowpath(struct rvt_qp *qp, struct hfi1_packet *pkt,
 			       bool do_cnp)
 {
 	struct hfi1_ibport *ibp = to_iport(qp->ibqp.device, qp->port_num);
+	struct hfi1_pportdata *ppd = ppd_from_ibp(ibp);
 	struct ib_other_headers *ohdr = pkt->ohdr;
 	struct ib_grh *grh = pkt->grh;
 	u32 rqpn = 0, bth1;
-	u16 pkey, rlid, dlid = ib_get_dlid(pkt->hdr);
+	u16 pkey;
+	u32 rlid, slid, dlid = 0;
 	u8 hdr_type, sc, svc_type;
 	bool is_mcast = false;
 
+	/* can be called from prescan */
 	if (pkt->etype == RHF_RCV_TYPE_BYPASS) {
 		is_mcast = hfi1_is_16B_mcast(dlid);
 		pkey = hfi1_16B_get_pkey(pkt->hdr);
 		sc = hfi1_16B_get_sc(pkt->hdr);
+		dlid = hfi1_16B_get_dlid(pkt->hdr);
+		slid = hfi1_16B_get_slid(pkt->hdr);
 		hdr_type = HFI1_PKT_TYPE_16B;
 	} else {
 		is_mcast = (dlid > be16_to_cpu(IB_MULTICAST_LID_BASE)) &&
 			   (dlid != be16_to_cpu(IB_LID_PERMISSIVE));
 		pkey = ib_bth_get_pkey(ohdr);
 		sc = hfi1_9B_get_sc5(pkt->hdr, pkt->rhf);
+		dlid = ib_get_dlid(pkt->hdr);
+		slid = ib_get_slid(pkt->hdr);
 		hdr_type = HFI1_PKT_TYPE_9B;
 	}
 
 	switch (qp->ibqp.qp_type) {
+	case IB_QPT_UD:
+		dlid = ppd->lid;
+		rlid = slid;
+		rqpn = ib_get_sqpn(pkt->ohdr);
+		svc_type = IB_CC_SVCTYPE_UD;
+		break;
 	case IB_QPT_SMI:
 	case IB_QPT_GSI:
-	case IB_QPT_UD:
-		rlid = ib_get_slid(pkt->hdr);
+		rlid = slid;
 		rqpn = ib_get_sqpn(pkt->ohdr);
 		svc_type = IB_CC_SVCTYPE_UD;
 		break;
@@ -482,7 +494,6 @@ void hfi1_process_ecn_slowpath(struct rvt_qp *qp, struct hfi1_packet *pkt,
 					      dlid, rlid, sc, grh);
 
 	if (!is_mcast && (bth1 & IB_BECN_SMASK)) {
-		struct hfi1_pportdata *ppd = ppd_from_ibp(ibp);
 		u32 lqpn = bth1 & RVT_QPN_MASK;
 		u8 sl = ibp->sc_to_sl[sc];
 
diff --git a/drivers/infiniband/hw/hfi1/hfi.h b/drivers/infiniband/hw/hfi1/hfi.h
index 32c4826..cac2c62 100644
--- a/drivers/infiniband/hw/hfi1/hfi.h
+++ b/drivers/infiniband/hw/hfi1/hfi.h
@@ -1537,13 +1537,13 @@ void set_link_ipg(struct hfi1_pportdata *ppd);
 void process_becn(struct hfi1_pportdata *ppd, u8 sl, u32 rlid, u32 lqpn,
 		  u32 rqpn, u8 svc_type);
 void return_cnp(struct hfi1_ibport *ibp, struct rvt_qp *qp, u32 remote_qpn,
-		u32 pkey, u32 slid, u32 dlid, u8 sc5,
+		u16 pkey, u32 slid, u32 dlid, u8 sc5,
 		const struct ib_grh *old_grh);
 void return_cnp_16B(struct hfi1_ibport *ibp, struct rvt_qp *qp,
-		    u32 remote_qpn, u32 pkey, u32 slid, u32 dlid,
+		    u32 remote_qpn, u16 pkey, u32 slid, u32 dlid,
 		    u8 sc5, const struct ib_grh *old_grh);
 typedef void (*hfi1_handle_cnp)(struct hfi1_ibport *ibp, struct rvt_qp *qp,
-				u32 remote_qpn, u32 pkey, u32 slid, u32 dlid,
+				u32 remote_qpn, u16 pkey, u32 slid, u32 dlid,
 				u8 sc5, const struct ib_grh *old_grh);
 
 #define PKEY_CHECK_INVALID -1
@@ -2437,7 +2437,7 @@ static inline void hfi1_make_16b_hdr(struct hfi1_16b_header *hdr,
 		((slid >> OPA_16B_SLID_SHIFT) << OPA_16B_SLID_HIGH_SHIFT);
 	lrh2 = (lrh2 & ~OPA_16B_DLID_MASK) |
 		((dlid >> OPA_16B_DLID_SHIFT) << OPA_16B_DLID_HIGH_SHIFT);
-	lrh2 = (lrh2 & ~OPA_16B_PKEY_MASK) | (pkey << OPA_16B_PKEY_SHIFT);
+	lrh2 = (lrh2 & ~OPA_16B_PKEY_MASK) | ((u32)pkey << OPA_16B_PKEY_SHIFT);
 	lrh2 = (lrh2 & ~OPA_16B_L4_MASK) | l4;
 
 	hdr->lrh[0] = lrh0;
diff --git a/drivers/infiniband/hw/hfi1/ud.c b/drivers/infiniband/hw/hfi1/ud.c
index bcf3b0b..69c17a5 100644
--- a/drivers/infiniband/hw/hfi1/ud.c
+++ b/drivers/infiniband/hw/hfi1/ud.c
@@ -628,7 +628,7 @@ int hfi1_lookup_pkey_idx(struct hfi1_ibport *ibp, u16 pkey)
 }
 
 void return_cnp_16B(struct hfi1_ibport *ibp, struct rvt_qp *qp,
-		    u32 remote_qpn, u32 pkey, u32 slid, u32 dlid,
+		    u32 remote_qpn, u16 pkey, u32 slid, u32 dlid,
 		    u8 sc5, const struct ib_grh *old_grh)
 {
 	u64 pbc, pbc_flags = 0;
@@ -687,7 +687,7 @@ void return_cnp_16B(struct hfi1_ibport *ibp, struct rvt_qp *qp,
 }
 
 void return_cnp(struct hfi1_ibport *ibp, struct rvt_qp *qp, u32 remote_qpn,
-		u32 pkey, u32 slid, u32 dlid, u8 sc5,
+		u16 pkey, u32 slid, u32 dlid, u8 sc5,
 		const struct ib_grh *old_grh)
 {
 	u64 pbc, pbc_flags = 0;
-- 
cgit v1.1


From 5da9e742be44d9b7c68b1bf6e1aaf46a1aa7a52b Mon Sep 17 00:00:00 2001
From: "Michael J. Ruhl" <michael.j.ruhl@intel.com>
Date: Tue, 1 May 2018 05:35:43 -0700
Subject: IB/hfi1 Use correct type for num_user_context

The module parameter num_user_context is defined as 'int' and
defaults to -1.  The module_param_named() says that it is uint.

Correct module_param_named() type information and update the modinfo
text to reflect the default value.

Reviewed-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/init.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/infiniband/hw/hfi1/init.c b/drivers/infiniband/hw/hfi1/init.c
index 33eba23..c45cca5 100644
--- a/drivers/infiniband/hw/hfi1/init.c
+++ b/drivers/infiniband/hw/hfi1/init.c
@@ -88,9 +88,9 @@
  * pio buffers per ctxt, etc.)  Zero means use one user context per CPU.
  */
 int num_user_contexts = -1;
-module_param_named(num_user_contexts, num_user_contexts, uint, S_IRUGO);
+module_param_named(num_user_contexts, num_user_contexts, int, 0444);
 MODULE_PARM_DESC(
-	num_user_contexts, "Set max number of user contexts to use");
+	num_user_contexts, "Set max number of user contexts to use (default: -1 will use the real (non-HT) CPU count)");
 
 uint krcvqs[RXE_NUM_DATA_VL];
 int krcvqsset;
-- 
cgit v1.1


From 0a0bcb046b2f0c15b89f8c1b08ad3de601a83c66 Mon Sep 17 00:00:00 2001
From: Mike Marciniszyn <mike.marciniszyn@intel.com>
Date: Tue, 1 May 2018 05:35:51 -0700
Subject: IB/hfi1: Fix loss of BECN with AHG

AHG may be armed to use the stored header, which by design is limited
to edits in the PSN/A 32 bit word (bth2).

When the code is trying to send a BECN, the use of the stored header
will lose the BECN bit.

Fix by avoiding AHG when getting ready to send a BECN. This is
accomplished by always claiming the packet is not a middle packet which
is an AHG precursor.  BECNs are not a normal case and this should not
hurt AHG optimizations.

Cc: <stable@vger.kernel.org> # 4.14.x
Reviewed-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/ruc.c | 50 ++++++++++++++++++++++++++++++++--------
 1 file changed, 40 insertions(+), 10 deletions(-)

diff --git a/drivers/infiniband/hw/hfi1/ruc.c b/drivers/infiniband/hw/hfi1/ruc.c
index 3daa94b..c0071ca 100644
--- a/drivers/infiniband/hw/hfi1/ruc.c
+++ b/drivers/infiniband/hw/hfi1/ruc.c
@@ -733,6 +733,20 @@ static inline void hfi1_make_ruc_bth(struct rvt_qp *qp,
 	ohdr->bth[2] = cpu_to_be32(bth2);
 }
 
+/**
+ * hfi1_make_ruc_header_16B - build a 16B header
+ * @qp: the queue pair
+ * @ohdr: a pointer to the destination header memory
+ * @bth0: bth0 passed in from the RC/UC builder
+ * @bth2: bth2 passed in from the RC/UC builder
+ * @middle: non zero implies indicates ahg "could" be used
+ * @ps: the current packet state
+ *
+ * This routine may disarm ahg under these situations:
+ * - packet needs a GRH
+ * - BECN needed
+ * - migration state not IB_MIG_MIGRATED
+ */
 static inline void hfi1_make_ruc_header_16B(struct rvt_qp *qp,
 					    struct ib_other_headers *ohdr,
 					    u32 bth0, u32 bth2, int middle,
@@ -777,6 +791,12 @@ static inline void hfi1_make_ruc_header_16B(struct rvt_qp *qp,
 	else
 		middle = 0;
 
+	if (qp->s_flags & RVT_S_ECN) {
+		qp->s_flags &= ~RVT_S_ECN;
+		/* we recently received a FECN, so return a BECN */
+		becn = true;
+		middle = 0;
+	}
 	if (middle)
 		build_ahg(qp, bth2);
 	else
@@ -784,11 +804,6 @@ static inline void hfi1_make_ruc_header_16B(struct rvt_qp *qp,
 
 	bth0 |= pkey;
 	bth0 |= extra_bytes << 20;
-	if (qp->s_flags & RVT_S_ECN) {
-		qp->s_flags &= ~RVT_S_ECN;
-		/* we recently received a FECN, so return a BECN */
-		becn = true;
-	}
 	hfi1_make_ruc_bth(qp, ohdr, bth0, bth1, bth2);
 
 	if (!ppd->lid)
@@ -806,6 +821,20 @@ static inline void hfi1_make_ruc_header_16B(struct rvt_qp *qp,
 			  pkey, becn, 0, l4, priv->s_sc);
 }
 
+/**
+ * hfi1_make_ruc_header_9B - build a 9B header
+ * @qp: the queue pair
+ * @ohdr: a pointer to the destination header memory
+ * @bth0: bth0 passed in from the RC/UC builder
+ * @bth2: bth2 passed in from the RC/UC builder
+ * @middle: non zero implies indicates ahg "could" be used
+ * @ps: the current packet state
+ *
+ * This routine may disarm ahg under these situations:
+ * - packet needs a GRH
+ * - BECN needed
+ * - migration state not IB_MIG_MIGRATED
+ */
 static inline void hfi1_make_ruc_header_9B(struct rvt_qp *qp,
 					   struct ib_other_headers *ohdr,
 					   u32 bth0, u32 bth2, int middle,
@@ -839,6 +868,12 @@ static inline void hfi1_make_ruc_header_9B(struct rvt_qp *qp,
 	else
 		middle = 0;
 
+	if (qp->s_flags & RVT_S_ECN) {
+		qp->s_flags &= ~RVT_S_ECN;
+		/* we recently received a FECN, so return a BECN */
+		bth1 |= (IB_BECN_MASK << IB_BECN_SHIFT);
+		middle = 0;
+	}
 	if (middle)
 		build_ahg(qp, bth2);
 	else
@@ -846,11 +881,6 @@ static inline void hfi1_make_ruc_header_9B(struct rvt_qp *qp,
 
 	bth0 |= pkey;
 	bth0 |= extra_bytes << 20;
-	if (qp->s_flags & RVT_S_ECN) {
-		qp->s_flags &= ~RVT_S_ECN;
-		/* we recently received a FECN, so return a BECN */
-		bth1 |= (IB_BECN_MASK << IB_BECN_SHIFT);
-	}
 	hfi1_make_ruc_bth(qp, ohdr, bth0, bth1, bth2);
 	hfi1_make_ib_hdr(&ps->s_txreq->phdr.hdr.ibh,
 			 lrh0,
-- 
cgit v1.1


From 45d924571a5e1329580811f2419da61b07ac3613 Mon Sep 17 00:00:00 2001
From: Sebastian Sanchez <sebastian.sanchez@intel.com>
Date: Tue, 1 May 2018 05:35:58 -0700
Subject: IB/hfi1: Fix NULL pointer dereference when invalid num_vls is used

When an invalid num_vls is used as a module parameter, the code
execution follows an exception path where the macro dd_dev_err()
expects dd->pcidev->dev not to be NULL in hfi1_init_dd(). This
causes a NULL pointer dereference.

Fix hfi1_init_dd() by initializing dd->pcidev and dd->pcidev->dev
earlier in the code. If a dd exists, then dd->pcidev and
dd->pcidev->dev always exists.

BUG: unable to handle kernel NULL pointer dereference
at 00000000000000f0
IP: __dev_printk+0x15/0x90
Workqueue: events work_for_cpu_fn
RIP: 0010:__dev_printk+0x15/0x90
Call Trace:
 dev_err+0x6c/0x90
 ? hfi1_init_pportdata+0x38d/0x3f0 [hfi1]
 hfi1_init_dd+0xdd/0x2530 [hfi1]
 ? pci_conf1_read+0xb2/0xf0
 ? pci_read_config_word.part.9+0x64/0x80
 ? pci_conf1_write+0xb0/0xf0
 ? pcie_capability_clear_and_set_word+0x57/0x80
 init_one+0x141/0x490 [hfi1]
 local_pci_probe+0x3f/0xa0
 work_for_cpu_fn+0x10/0x20
 process_one_work+0x152/0x350
 worker_thread+0x1cf/0x3e0
 kthread+0xf5/0x130
 ? max_active_store+0x80/0x80
 ? kthread_bind+0x10/0x10
 ? do_syscall_64+0x6e/0x1a0
 ? SyS_exit_group+0x10/0x10
 ret_from_fork+0x35/0x40

Cc: <stable@vger.kernel.org> # 4.9.x
Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Reviewed-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Signed-off-by: Sebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/init.c | 2 ++
 drivers/infiniband/hw/hfi1/pcie.c | 3 ---
 2 files changed, 2 insertions(+), 3 deletions(-)

diff --git a/drivers/infiniband/hw/hfi1/init.c b/drivers/infiniband/hw/hfi1/init.c
index c45cca5..b417e3b 100644
--- a/drivers/infiniband/hw/hfi1/init.c
+++ b/drivers/infiniband/hw/hfi1/init.c
@@ -1265,6 +1265,8 @@ struct hfi1_devdata *hfi1_alloc_devdata(struct pci_dev *pdev, size_t extra)
 		return ERR_PTR(-ENOMEM);
 	dd->num_pports = nports;
 	dd->pport = (struct hfi1_pportdata *)(dd + 1);
+	dd->pcidev = pdev;
+	pci_set_drvdata(pdev, dd);
 
 	INIT_LIST_HEAD(&dd->list);
 	idr_preload(GFP_KERNEL);
diff --git a/drivers/infiniband/hw/hfi1/pcie.c b/drivers/infiniband/hw/hfi1/pcie.c
index 83d66e8..c1c9829 100644
--- a/drivers/infiniband/hw/hfi1/pcie.c
+++ b/drivers/infiniband/hw/hfi1/pcie.c
@@ -163,9 +163,6 @@ int hfi1_pcie_ddinit(struct hfi1_devdata *dd, struct pci_dev *pdev)
 	resource_size_t addr;
 	int ret = 0;
 
-	dd->pcidev = pdev;
-	pci_set_drvdata(pdev, dd);
-
 	addr = pci_resource_start(pdev, 0);
 	len = pci_resource_len(pdev, 0);
 
-- 
cgit v1.1


From e9777ad4399c26c70318c4945f94efac2ed95391 Mon Sep 17 00:00:00 2001
From: Sebastian Sanchez <sebastian.sanchez@intel.com>
Date: Tue, 1 May 2018 05:36:06 -0700
Subject: IB/{hfi1, rdmavt}: Fix memory leak in hfi1_alloc_devdata() upon
 failure

When allocating device data, if there's an allocation failure, the
already allocated memory won't be freed such as per-cpu counters.

Fix memory leaks in exception path by creating a common reentrant
clean up function hfi1_clean_devdata() to be used at driver unload
time and device data allocation failure.

To accomplish this, free_platform_config() and clean_up_i2c() are
changed to be reentrant to remove dependencies when they are called
in different order. This helps avoid NULL pointer dereferences
introduced by this patch if those two functions weren't reentrant.

In addition, set dd->int_counter, dd->rcv_limit,
dd->send_schedule and dd->tx_opstats to NULL after they're freed in
hfi1_clean_devdata(), so that hfi1_clean_devdata() is fully reentrant.

Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Reviewed-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Signed-off-by: Sebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/init.c     | 37 +++++++++++++++++++++++++----------
 drivers/infiniband/hw/hfi1/platform.c |  1 +
 drivers/infiniband/hw/hfi1/qsfp.c     |  2 ++
 3 files changed, 30 insertions(+), 10 deletions(-)

diff --git a/drivers/infiniband/hw/hfi1/init.c b/drivers/infiniband/hw/hfi1/init.c
index b417e3b..6309edf 100644
--- a/drivers/infiniband/hw/hfi1/init.c
+++ b/drivers/infiniband/hw/hfi1/init.c
@@ -1209,30 +1209,49 @@ static void finalize_asic_data(struct hfi1_devdata *dd,
 	kfree(ad);
 }
 
-static void __hfi1_free_devdata(struct kobject *kobj)
+/**
+ * hfi1_clean_devdata - cleans up per-unit data structure
+ * @dd: pointer to a valid devdata structure
+ *
+ * It cleans up all data structures set up by
+ * by hfi1_alloc_devdata().
+ */
+static void hfi1_clean_devdata(struct hfi1_devdata *dd)
 {
-	struct hfi1_devdata *dd =
-		container_of(kobj, struct hfi1_devdata, kobj);
 	struct hfi1_asic_data *ad;
 	unsigned long flags;
 
 	spin_lock_irqsave(&hfi1_devs_lock, flags);
-	idr_remove(&hfi1_unit_table, dd->unit);
-	list_del(&dd->list);
+	if (!list_empty(&dd->list)) {
+		idr_remove(&hfi1_unit_table, dd->unit);
+		list_del_init(&dd->list);
+	}
 	ad = release_asic_data(dd);
 	spin_unlock_irqrestore(&hfi1_devs_lock, flags);
-	if (ad)
-		finalize_asic_data(dd, ad);
+
+	finalize_asic_data(dd, ad);
 	free_platform_config(dd);
 	rcu_barrier(); /* wait for rcu callbacks to complete */
 	free_percpu(dd->int_counter);
 	free_percpu(dd->rcv_limit);
 	free_percpu(dd->send_schedule);
 	free_percpu(dd->tx_opstats);
+	dd->int_counter   = NULL;
+	dd->rcv_limit     = NULL;
+	dd->send_schedule = NULL;
+	dd->tx_opstats    = NULL;
 	sdma_clean(dd, dd->num_sdma);
 	rvt_dealloc_device(&dd->verbs_dev.rdi);
 }
 
+static void __hfi1_free_devdata(struct kobject *kobj)
+{
+	struct hfi1_devdata *dd =
+		container_of(kobj, struct hfi1_devdata, kobj);
+
+	hfi1_clean_devdata(dd);
+}
+
 static struct kobj_type hfi1_devdata_type = {
 	.release = __hfi1_free_devdata,
 };
@@ -1333,9 +1352,7 @@ struct hfi1_devdata *hfi1_alloc_devdata(struct pci_dev *pdev, size_t extra)
 	return dd;
 
 bail:
-	if (!list_empty(&dd->list))
-		list_del_init(&dd->list);
-	rvt_dealloc_device(&dd->verbs_dev.rdi);
+	hfi1_clean_devdata(dd);
 	return ERR_PTR(ret);
 }
 
diff --git a/drivers/infiniband/hw/hfi1/platform.c b/drivers/infiniband/hw/hfi1/platform.c
index d486355..cbf7faa 100644
--- a/drivers/infiniband/hw/hfi1/platform.c
+++ b/drivers/infiniband/hw/hfi1/platform.c
@@ -199,6 +199,7 @@ void free_platform_config(struct hfi1_devdata *dd)
 {
 	/* Release memory allocated for eprom or fallback file read. */
 	kfree(dd->platform_config.data);
+	dd->platform_config.data = NULL;
 }
 
 void get_port_type(struct hfi1_pportdata *ppd)
diff --git a/drivers/infiniband/hw/hfi1/qsfp.c b/drivers/infiniband/hw/hfi1/qsfp.c
index 1869f63..b596699 100644
--- a/drivers/infiniband/hw/hfi1/qsfp.c
+++ b/drivers/infiniband/hw/hfi1/qsfp.c
@@ -204,6 +204,8 @@ static void clean_i2c_bus(struct hfi1_i2c_bus *bus)
 
 void clean_up_i2c(struct hfi1_devdata *dd, struct hfi1_asic_data *ad)
 {
+	if (!ad)
+		return;
 	clean_i2c_bus(ad->i2c_bus0);
 	ad->i2c_bus0 = NULL;
 	clean_i2c_bus(ad->i2c_bus1);
-- 
cgit v1.1


From 59482a14918b282ca2a98f38c69da5ebeb1107d2 Mon Sep 17 00:00:00 2001
From: Sebastian Sanchez <sebastian.sanchez@intel.com>
Date: Tue, 1 May 2018 05:36:13 -0700
Subject: IB/hfi1: Fix memory leak in exception path in get_irq_affinity()

When IRQ affinity is set and the interrupt type is unknown, a cpu
mask allocated within the function is never freed. Fix this memory
leak by allocating memory within the scope where it is used.

Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Reviewed-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Signed-off-by: Sebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/affinity.c | 11 +++++------
 1 file changed, 5 insertions(+), 6 deletions(-)

diff --git a/drivers/infiniband/hw/hfi1/affinity.c b/drivers/infiniband/hw/hfi1/affinity.c
index a97055d..b5fab55 100644
--- a/drivers/infiniband/hw/hfi1/affinity.c
+++ b/drivers/infiniband/hw/hfi1/affinity.c
@@ -412,7 +412,6 @@ static void hfi1_cleanup_sdma_notifier(struct hfi1_msix_entry *msix)
 static int get_irq_affinity(struct hfi1_devdata *dd,
 			    struct hfi1_msix_entry *msix)
 {
-	int ret;
 	cpumask_var_t diff;
 	struct hfi1_affinity_node *entry;
 	struct cpu_mask_set *set = NULL;
@@ -424,10 +423,6 @@ static int get_irq_affinity(struct hfi1_devdata *dd,
 	extra[0] = '\0';
 	cpumask_clear(&msix->mask);
 
-	ret = zalloc_cpumask_var(&diff, GFP_KERNEL);
-	if (!ret)
-		return -ENOMEM;
-
 	entry = node_affinity_lookup(dd->node);
 
 	switch (msix->type) {
@@ -458,6 +453,9 @@ static int get_irq_affinity(struct hfi1_devdata *dd,
 	 * finds its CPU here.
 	 */
 	if (cpu == -1 && set) {
+		if (!zalloc_cpumask_var(&diff, GFP_KERNEL))
+			return -ENOMEM;
+
 		if (cpumask_equal(&set->mask, &set->used)) {
 			/*
 			 * We've used up all the CPUs, bump up the generation
@@ -469,6 +467,8 @@ static int get_irq_affinity(struct hfi1_devdata *dd,
 		cpumask_andnot(diff, &set->mask, &set->used);
 		cpu = cpumask_first(diff);
 		cpumask_set_cpu(cpu, &set->used);
+
+		free_cpumask_var(diff);
 	}
 
 	cpumask_set_cpu(cpu, &msix->mask);
@@ -482,7 +482,6 @@ static int get_irq_affinity(struct hfi1_devdata *dd,
 		hfi1_setup_sdma_notifier(msix);
 	}
 
-	free_cpumask_var(diff);
 	return 0;
 }
 
-- 
cgit v1.1


From b03bcde962606d2ee59a4e9dd470db9ad53c5418 Mon Sep 17 00:00:00 2001
From: Jack Morgenstein <jackm@dev.mellanox.co.il>
Date: Wed, 2 May 2018 13:04:25 +0300
Subject: IB/mlx4: Fix integer overflow when calculating optimal MTT size

When the kernel was compiled using the UBSAN option,
we saw the following stack trace:

[ 1184.827917] UBSAN: Undefined behaviour in drivers/infiniband/hw/mlx4/mr.c:349:27
[ 1184.828114] signed integer overflow:
[ 1184.828247] -2147483648 - 1 cannot be represented in type 'int'

The problem was caused by calling round_up in procedure
mlx4_ib_umem_calc_optimal_mtt_size (on line 349, as noted in the stack
trace) with the second parameter (1 << block_shift) (which is an int).
The second parameter should have been (1ULL << block_shift) (which
is an unsigned long long).

(1 << block_shift) is treated by the compiler as an int (because 1 is
an integer).

Now, local variable block_shift is initialized to 31.
If block_shift is 31, 1 << block_shift is 1 << 31 = 0x80000000=-214748368.
This is the most negative int value.

Inside the round_up macro, there is a cast applied to ((1 << 31) - 1).
However, this cast is applied AFTER ((1 << 31) - 1) is calculated.
Since (1 << 31) is treated as an int, we get the negative overflow
identified by UBSAN in the process of calculating ((1 << 31) - 1).

The fix is to change (1 << block_shift) to (1ULL << block_shift) on
line 349.

Fixes: 9901abf58368 ("IB/mlx4: Use optimal numbers of MTT entries")
Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx4/mr.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/infiniband/hw/mlx4/mr.c b/drivers/infiniband/hw/mlx4/mr.c
index 17f4f15..61d8b06 100644
--- a/drivers/infiniband/hw/mlx4/mr.c
+++ b/drivers/infiniband/hw/mlx4/mr.c
@@ -346,7 +346,7 @@ int mlx4_ib_umem_calc_optimal_mtt_size(struct ib_umem *umem, u64 start_va,
 	/* Add to the first block the misalignment that it suffers from. */
 	total_len += (first_block_start & ((1ULL << block_shift) - 1ULL));
 	last_block_end = current_block_start + current_block_len;
-	last_block_aligned_end = round_up(last_block_end, 1 << block_shift);
+	last_block_aligned_end = round_up(last_block_end, 1ULL << block_shift);
 	total_len += (last_block_aligned_end - last_block_end);
 
 	if (total_len & ((1ULL << block_shift) - 1ULL))
-- 
cgit v1.1


From 9aa169213d1166d30ae357a44abbeae93459339d Mon Sep 17 00:00:00 2001
From: Parav Pandit <parav@mellanox.com>
Date: Wed, 2 May 2018 13:18:59 +0300
Subject: RDMA/cma: Do not query GID during QP state transition to RTR

When commit [1] was added, SGID was queried to derive the SMAC address.
Then, later on during a refactor [2], SMAC was no longer needed. However,
the now useless GID query remained.  Then during additional code changes
later on, the GID query was being done in such a way that it caused iWARP
queries to start breaking.  Remove the useless GID query and resolve the
iWARP breakage at the same time.

This is discussed in [3].

[1] commit dd5f03beb4f7 ("IB/core: Ethernet L2 attributes in verbs/cm structures")
[2] commit 5c266b2304fb ("IB/cm: Remove the usage of smac and vid of qp_attr and cm_av")
[3] https://www.spinics.net/lists/linux-rdma/msg63951.html

Suggested-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Parav Pandit <parav@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/cma.c | 7 -------
 1 file changed, 7 deletions(-)

diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c
index 8364223..a693fcd 100644
--- a/drivers/infiniband/core/cma.c
+++ b/drivers/infiniband/core/cma.c
@@ -868,7 +868,6 @@ static int cma_modify_qp_rtr(struct rdma_id_private *id_priv,
 {
 	struct ib_qp_attr qp_attr;
 	int qp_attr_mask, ret;
-	union ib_gid sgid;
 
 	mutex_lock(&id_priv->qp_mutex);
 	if (!id_priv->id.qp) {
@@ -891,12 +890,6 @@ static int cma_modify_qp_rtr(struct rdma_id_private *id_priv,
 	if (ret)
 		goto out;
 
-	ret = ib_query_gid(id_priv->id.device, id_priv->id.port_num,
-			   rdma_ah_read_grh(&qp_attr.ah_attr)->sgid_index,
-			   &sgid, NULL);
-	if (ret)
-		goto out;
-
 	BUG_ON(id_priv->cma_dev->device != id_priv->id.device);
 
 	if (conn_param)
-- 
cgit v1.1


From b8b784958eccbf8f51ebeee65282ca3fd59ea391 Mon Sep 17 00:00:00 2001
From: Jan Kara <jack@suse.cz>
Date: Thu, 3 May 2018 18:26:26 +0200
Subject: bdi: Fix oops in wb_workfn()

Syzbot has reported that it can hit a NULL pointer dereference in
wb_workfn() due to wb->bdi->dev being NULL. This indicates that
wb_workfn() was called for an already unregistered bdi which should not
happen as wb_shutdown() called from bdi_unregister() should make sure
all pending writeback works are completed before bdi is unregistered.
Except that wb_workfn() itself can requeue the work with:

	mod_delayed_work(bdi_wq, &wb->dwork, 0);

and if this happens while wb_shutdown() is waiting in:

	flush_delayed_work(&wb->dwork);

the dwork can get executed after wb_shutdown() has finished and
bdi_unregister() has cleared wb->bdi->dev.

Make wb_workfn() use wakeup_wb() for requeueing the work which takes all
the necessary precautions against racing with bdi unregistration.

CC: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
CC: Tejun Heo <tj@kernel.org>
Fixes: 839a8e8660b6777e7fe4e80af1a048aebe2b5977
Reported-by: syzbot <syzbot+9873874c735f2892e7e9@syzkaller.appspotmail.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 fs/fs-writeback.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
index 47d7c15..471d863 100644
--- a/fs/fs-writeback.c
+++ b/fs/fs-writeback.c
@@ -1961,7 +1961,7 @@ void wb_workfn(struct work_struct *work)
 	}
 
 	if (!list_empty(&wb->work_list))
-		mod_delayed_work(bdi_wq, &wb->dwork, 0);
+		wb_wakeup(wb);
 	else if (wb_has_dirty_io(wb) && dirty_writeback_interval)
 		wb_wakeup_delayed(wb);
 
-- 
cgit v1.1


From 9ef09e35e521bf0df5325cc9cffa726a8f5f3c1b Mon Sep 17 00:00:00 2001
From: Mark Rutland <mark.rutland@arm.com>
Date: Thu, 3 May 2018 17:04:59 +0100
Subject: bpf: fix possible spectre-v1 in find_and_alloc_map()

It's possible for userspace to control attr->map_type. Sanitize it when
using it as an array index to prevent an out-of-bounds value being used
under speculation.

Found by smatch.

Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Dan Carpenter <dan.carpenter@oracle.com>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: netdev@vger.kernel.org
Acked-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
---
 kernel/bpf/syscall.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c
index ebfe9f2..8f43448 100644
--- a/kernel/bpf/syscall.c
+++ b/kernel/bpf/syscall.c
@@ -26,6 +26,7 @@
 #include <linux/cred.h>
 #include <linux/timekeeping.h>
 #include <linux/ctype.h>
+#include <linux/nospec.h>
 
 #define IS_FD_ARRAY(map) ((map)->map_type == BPF_MAP_TYPE_PROG_ARRAY || \
 			   (map)->map_type == BPF_MAP_TYPE_PERF_EVENT_ARRAY || \
@@ -102,12 +103,14 @@ const struct bpf_map_ops bpf_map_offload_ops = {
 static struct bpf_map *find_and_alloc_map(union bpf_attr *attr)
 {
 	const struct bpf_map_ops *ops;
+	u32 type = attr->map_type;
 	struct bpf_map *map;
 	int err;
 
-	if (attr->map_type >= ARRAY_SIZE(bpf_map_types))
+	if (type >= ARRAY_SIZE(bpf_map_types))
 		return ERR_PTR(-EINVAL);
-	ops = bpf_map_types[attr->map_type];
+	type = array_index_nospec(type, ARRAY_SIZE(bpf_map_types));
+	ops = bpf_map_types[type];
 	if (!ops)
 		return ERR_PTR(-EINVAL);
 
@@ -122,7 +125,7 @@ static struct bpf_map *find_and_alloc_map(union bpf_attr *attr)
 	if (IS_ERR(map))
 		return map;
 	map->ops = ops;
-	map->map_type = attr->map_type;
+	map->map_type = type;
 	return map;
 }
 
-- 
cgit v1.1


From 74ba22ead5971e25c2e01bd08f5d814fd8ee41b3 Mon Sep 17 00:00:00 2001
From: Tomasz Lis <tomasz.lis@intel.com>
Date: Wed, 2 May 2018 15:31:42 -0700
Subject: drm/i915/icl: Add configuring MOCS in new Icelake engines

In Icelake, there are more engines on which Memory Object Control
States need to be configured. Besides adding Icelake under Skylake
config, the patch makes sure MOCS register addresses for the new
engines are properly defined.

Additional patch might be need later, in case the specification will
propose different MOCS config values for Icelake than in previous
gens.

v2: Restricted comments to gen11, updated description, renamed
defines.

v3: Used proper engine indexes for gen11.

v4: Ensure patch is Icelake only.

v5: Style fixes (proposed by mwajdeczko)

v6 (from Paulo): fix checkpatch's COMMIT_LOG_LONG_LINE (Checkpatch).

BSpec: 19405
BSpec: 21140
Cc: Oscar Mateo Lozano <oscar.mateo@intel.com>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Reviewed-by: Michel Thierry <michel.thierry@intel.com>
Signed-off-by: Tomasz Lis <tomasz.lis@intel.com>
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180502223142.3891-1-paulo.r.zanoni@intel.com
---
 drivers/gpu/drm/i915/i915_reg.h   | 2 ++
 drivers/gpu/drm/i915/intel_mocs.c | 5 ++++-
 2 files changed, 6 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 197c966..085928c 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -9864,6 +9864,8 @@ enum skl_power_gate {
 #define GEN9_MFX1_MOCS(i)	_MMIO(0xca00 + (i) * 4)	/* Media 1 MOCS registers */
 #define GEN9_VEBOX_MOCS(i)	_MMIO(0xcb00 + (i) * 4)	/* Video MOCS registers */
 #define GEN9_BLT_MOCS(i)	_MMIO(0xcc00 + (i) * 4)	/* Blitter MOCS registers */
+/* Media decoder 2 MOCS registers */
+#define GEN11_MFX2_MOCS(i)	_MMIO(0x10000 + (i) * 4)
 
 /* gamt regs */
 #define GEN8_L3_LRA_1_GPGPU _MMIO(0x4dd4)
diff --git a/drivers/gpu/drm/i915/intel_mocs.c b/drivers/gpu/drm/i915/intel_mocs.c
index c0b34b7..9f0bd6a4 100644
--- a/drivers/gpu/drm/i915/intel_mocs.c
+++ b/drivers/gpu/drm/i915/intel_mocs.c
@@ -178,7 +178,8 @@ static bool get_mocs_settings(struct drm_i915_private *dev_priv,
 {
 	bool result = false;
 
-	if (IS_GEN9_BC(dev_priv) || IS_CANNONLAKE(dev_priv)) {
+	if (IS_GEN9_BC(dev_priv) || IS_CANNONLAKE(dev_priv) ||
+	    IS_ICELAKE(dev_priv)) {
 		table->size  = ARRAY_SIZE(skylake_mocs_table);
 		table->table = skylake_mocs_table;
 		result = true;
@@ -217,6 +218,8 @@ static i915_reg_t mocs_register(enum intel_engine_id engine_id, int index)
 		return GEN9_VEBOX_MOCS(index);
 	case VCS2:
 		return GEN9_MFX1_MOCS(index);
+	case VCS3:
+		return GEN11_MFX2_MOCS(index);
 	default:
 		MISSING_CASE(engine_id);
 		return INVALID_MMIO_REG;
-- 
cgit v1.1


From 2eced8e917b060587fc8ed46df41c364957a5050 Mon Sep 17 00:00:00 2001
From: Andrzej Hajda <a.hajda@samsung.com>
Date: Fri, 2 Feb 2018 16:11:22 +0100
Subject: drm/exynos/mixer: fix synchronization check in interlaced mode

In case of interlace mode video processor registers and mixer config
register must be check to ensure internal state is in sync with shadow
registers.
This patch fixes page-faults in interlaced mode.

Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
---
 drivers/gpu/drm/exynos/exynos_mixer.c | 10 ++++++++++
 drivers/gpu/drm/exynos/regs-mixer.h   |  1 +
 2 files changed, 11 insertions(+)

diff --git a/drivers/gpu/drm/exynos/exynos_mixer.c b/drivers/gpu/drm/exynos/exynos_mixer.c
index 257299e..a8d978d 100644
--- a/drivers/gpu/drm/exynos/exynos_mixer.c
+++ b/drivers/gpu/drm/exynos/exynos_mixer.c
@@ -482,6 +482,7 @@ static void vp_video_buffer(struct mixer_context *ctx,
 
 	spin_lock_irqsave(&ctx->reg_slock, flags);
 
+	vp_reg_write(ctx, VP_SHADOW_UPDATE, 1);
 	/* interlace or progressive scan mode */
 	val = (test_bit(MXR_BIT_INTERLACE, &ctx->flags) ? ~0 : 0);
 	vp_reg_writemask(ctx, VP_MODE, val, VP_MODE_LINE_SKIP);
@@ -699,6 +700,15 @@ static irqreturn_t mixer_irq_handler(int irq, void *arg)
 
 		/* interlace scan need to check shadow register */
 		if (test_bit(MXR_BIT_INTERLACE, &ctx->flags)) {
+			if (test_bit(MXR_BIT_VP_ENABLED, &ctx->flags) &&
+			    vp_reg_read(ctx, VP_SHADOW_UPDATE))
+				goto out;
+
+			base = mixer_reg_read(ctx, MXR_CFG);
+			shadow = mixer_reg_read(ctx, MXR_CFG_S);
+			if (base != shadow)
+				goto out;
+
 			base = mixer_reg_read(ctx, MXR_GRAPHIC_BASE(0));
 			shadow = mixer_reg_read(ctx, MXR_GRAPHIC_BASE_S(0));
 			if (base != shadow)
diff --git a/drivers/gpu/drm/exynos/regs-mixer.h b/drivers/gpu/drm/exynos/regs-mixer.h
index c311f57..189cfa2 100644
--- a/drivers/gpu/drm/exynos/regs-mixer.h
+++ b/drivers/gpu/drm/exynos/regs-mixer.h
@@ -47,6 +47,7 @@
 #define MXR_MO				0x0304
 #define MXR_RESOLUTION			0x0310
 
+#define MXR_CFG_S			0x2004
 #define MXR_GRAPHIC0_BASE_S		0x2024
 #define MXR_GRAPHIC1_BASE_S		0x2044
 
-- 
cgit v1.1


From 0ccc1c8f0282e237a0bd6dca7cdac4ed5e318ee7 Mon Sep 17 00:00:00 2001
From: Tobias Jakobi <tjakobi@math.uni-bielefeld.de>
Date: Fri, 2 Feb 2018 16:11:23 +0100
Subject: drm/exynos: mixer: avoid Oops in vp_video_buffer()

If an interlaced video mode is selected, a IOMMU pagefault is
triggered by vp_video_buffer().

Fix the most apparent bugs:
- pitch value for chroma plane
- divide by two of height and vpos of source and destination

Signed-off-by: Tobias Jakobi <tjakobi@math.uni-bielefeld.de>
[ a.hajda: Halved also destination height and vpos, updated commit message ]
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
---
 drivers/gpu/drm/exynos/exynos_mixer.c | 12 +++++++-----
 1 file changed, 7 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/exynos/exynos_mixer.c b/drivers/gpu/drm/exynos/exynos_mixer.c
index a8d978d..272c79f 100644
--- a/drivers/gpu/drm/exynos/exynos_mixer.c
+++ b/drivers/gpu/drm/exynos/exynos_mixer.c
@@ -473,7 +473,7 @@ static void vp_video_buffer(struct mixer_context *ctx,
 			chroma_addr[1] = chroma_addr[0] + 0x40;
 		} else {
 			luma_addr[1] = luma_addr[0] + fb->pitches[0];
-			chroma_addr[1] = chroma_addr[0] + fb->pitches[0];
+			chroma_addr[1] = chroma_addr[0] + fb->pitches[1];
 		}
 	} else {
 		luma_addr[1] = 0;
@@ -496,21 +496,23 @@ static void vp_video_buffer(struct mixer_context *ctx,
 	vp_reg_write(ctx, VP_IMG_SIZE_Y, VP_IMG_HSIZE(fb->pitches[0]) |
 		VP_IMG_VSIZE(fb->height));
 	/* chroma plane for NV12/NV21 is half the height of the luma plane */
-	vp_reg_write(ctx, VP_IMG_SIZE_C, VP_IMG_HSIZE(fb->pitches[0]) |
+	vp_reg_write(ctx, VP_IMG_SIZE_C, VP_IMG_HSIZE(fb->pitches[1]) |
 		VP_IMG_VSIZE(fb->height / 2));
 
 	vp_reg_write(ctx, VP_SRC_WIDTH, state->src.w);
-	vp_reg_write(ctx, VP_SRC_HEIGHT, state->src.h);
 	vp_reg_write(ctx, VP_SRC_H_POSITION,
 			VP_SRC_H_POSITION_VAL(state->src.x));
-	vp_reg_write(ctx, VP_SRC_V_POSITION, state->src.y);
-
 	vp_reg_write(ctx, VP_DST_WIDTH, state->crtc.w);
 	vp_reg_write(ctx, VP_DST_H_POSITION, state->crtc.x);
+
 	if (test_bit(MXR_BIT_INTERLACE, &ctx->flags)) {
+		vp_reg_write(ctx, VP_SRC_HEIGHT, state->src.h / 2);
+		vp_reg_write(ctx, VP_SRC_V_POSITION, state->src.y / 2);
 		vp_reg_write(ctx, VP_DST_HEIGHT, state->crtc.h / 2);
 		vp_reg_write(ctx, VP_DST_V_POSITION, state->crtc.y / 2);
 	} else {
+		vp_reg_write(ctx, VP_SRC_HEIGHT, state->src.h);
+		vp_reg_write(ctx, VP_SRC_V_POSITION, state->src.y);
 		vp_reg_write(ctx, VP_DST_HEIGHT, state->crtc.h);
 		vp_reg_write(ctx, VP_DST_V_POSITION, state->crtc.y);
 	}
-- 
cgit v1.1


From d0f1a451e33d9ca834422622da30aa68daade56b Mon Sep 17 00:00:00 2001
From: Daniel Borkmann <daniel@iogearbox.net>
Date: Fri, 4 May 2018 02:13:57 +0200
Subject: bpf: use array_index_nospec in find_prog_type

Commit 9ef09e35e521 ("bpf: fix possible spectre-v1 in find_and_alloc_map()")
converted find_and_alloc_map() over to use array_index_nospec() to sanitize
map type that user space passes on map creation, and this patch does an
analogous conversion for progs in find_prog_type() as it's also passed from
user space when loading progs as attr->prog_type.

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Cc: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
---
 kernel/bpf/syscall.c | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)

diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c
index 8f43448..016ef90 100644
--- a/kernel/bpf/syscall.c
+++ b/kernel/bpf/syscall.c
@@ -874,11 +874,17 @@ static const struct bpf_prog_ops * const bpf_prog_types[] = {
 
 static int find_prog_type(enum bpf_prog_type type, struct bpf_prog *prog)
 {
-	if (type >= ARRAY_SIZE(bpf_prog_types) || !bpf_prog_types[type])
+	const struct bpf_prog_ops *ops;
+
+	if (type >= ARRAY_SIZE(bpf_prog_types))
+		return -EINVAL;
+	type = array_index_nospec(type, ARRAY_SIZE(bpf_prog_types));
+	ops = bpf_prog_types[type];
+	if (!ops)
 		return -EINVAL;
 
 	if (!bpf_prog_is_dev_bound(prog->aux))
-		prog->aux->ops = bpf_prog_types[type];
+		prog->aux->ops = ops;
 	else
 		prog->aux->ops = &bpf_offload_prog_ops;
 	prog->type = type;
-- 
cgit v1.1


From b5bf9a90bbebffba888c9144c5a8a10317b04064 Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <peterz@infradead.org>
Date: Mon, 30 Apr 2018 14:51:01 +0200
Subject: sched/core: Introduce set_special_state()

Gaurav reported a perceived problem with TASK_PARKED, which turned out
to be a broken wait-loop pattern in __kthread_parkme(), but the
reported issue can (and does) in fact happen for states that do not do
condition based sleeps.

When the 'current->state = TASK_RUNNING' store of a previous
(concurrent) try_to_wake_up() collides with the setting of a 'special'
sleep state, we can loose the sleep state.

Normal condition based wait-loops are immune to this problem, but for
sleep states that are not condition based are subject to this problem.

There already is a fix for TASK_DEAD. Abstract that and also apply it
to TASK_STOPPED and TASK_TRACED, both of which are also without
condition based wait-loop.

Reported-by: Gaurav Kohli <gkohli@codeaurora.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Oleg Nesterov <oleg@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 include/linux/sched.h        | 50 +++++++++++++++++++++++++++++++++++++++-----
 include/linux/sched/signal.h |  2 +-
 kernel/sched/core.c          | 17 +--------------
 kernel/signal.c              | 17 +++++++++++++--
 4 files changed, 62 insertions(+), 24 deletions(-)

diff --git a/include/linux/sched.h b/include/linux/sched.h
index b3d697f..c241370 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -112,17 +112,36 @@ struct task_group;
 
 #ifdef CONFIG_DEBUG_ATOMIC_SLEEP
 
+/*
+ * Special states are those that do not use the normal wait-loop pattern. See
+ * the comment with set_special_state().
+ */
+#define is_special_task_state(state)				\
+	((state) & (__TASK_STOPPED | __TASK_TRACED | TASK_DEAD))
+
 #define __set_current_state(state_value)			\
 	do {							\
+		WARN_ON_ONCE(is_special_task_state(state_value));\
 		current->task_state_change = _THIS_IP_;		\
 		current->state = (state_value);			\
 	} while (0)
+
 #define set_current_state(state_value)				\
 	do {							\
+		WARN_ON_ONCE(is_special_task_state(state_value));\
 		current->task_state_change = _THIS_IP_;		\
 		smp_store_mb(current->state, (state_value));	\
 	} while (0)
 
+#define set_special_state(state_value)					\
+	do {								\
+		unsigned long flags; /* may shadow */			\
+		WARN_ON_ONCE(!is_special_task_state(state_value));	\
+		raw_spin_lock_irqsave(&current->pi_lock, flags);	\
+		current->task_state_change = _THIS_IP_;			\
+		current->state = (state_value);				\
+		raw_spin_unlock_irqrestore(&current->pi_lock, flags);	\
+	} while (0)
 #else
 /*
  * set_current_state() includes a barrier so that the write of current->state
@@ -144,8 +163,8 @@ struct task_group;
  *
  * The above is typically ordered against the wakeup, which does:
  *
- *	need_sleep = false;
- *	wake_up_state(p, TASK_UNINTERRUPTIBLE);
+ *   need_sleep = false;
+ *   wake_up_state(p, TASK_UNINTERRUPTIBLE);
  *
  * Where wake_up_state() (and all other wakeup primitives) imply enough
  * barriers to order the store of the variable against wakeup.
@@ -154,12 +173,33 @@ struct task_group;
  * once it observes the TASK_UNINTERRUPTIBLE store the waking CPU can issue a
  * TASK_RUNNING store which can collide with __set_current_state(TASK_RUNNING).
  *
- * This is obviously fine, since they both store the exact same value.
+ * However, with slightly different timing the wakeup TASK_RUNNING store can
+ * also collide with the TASK_UNINTERRUPTIBLE store. Loosing that store is not
+ * a problem either because that will result in one extra go around the loop
+ * and our @cond test will save the day.
  *
  * Also see the comments of try_to_wake_up().
  */
-#define __set_current_state(state_value) do { current->state = (state_value); } while (0)
-#define set_current_state(state_value)	 smp_store_mb(current->state, (state_value))
+#define __set_current_state(state_value)				\
+	current->state = (state_value)
+
+#define set_current_state(state_value)					\
+	smp_store_mb(current->state, (state_value))
+
+/*
+ * set_special_state() should be used for those states when the blocking task
+ * can not use the regular condition based wait-loop. In that case we must
+ * serialize against wakeups such that any possible in-flight TASK_RUNNING stores
+ * will not collide with our state change.
+ */
+#define set_special_state(state_value)					\
+	do {								\
+		unsigned long flags; /* may shadow */			\
+		raw_spin_lock_irqsave(&current->pi_lock, flags);	\
+		current->state = (state_value);				\
+		raw_spin_unlock_irqrestore(&current->pi_lock, flags);	\
+	} while (0)
+
 #endif
 
 /* Task command name length: */
diff --git a/include/linux/sched/signal.h b/include/linux/sched/signal.h
index a7ce74c..113d1ad 100644
--- a/include/linux/sched/signal.h
+++ b/include/linux/sched/signal.h
@@ -280,7 +280,7 @@ static inline void kernel_signal_stop(void)
 {
 	spin_lock_irq(&current->sighand->siglock);
 	if (current->jobctl & JOBCTL_STOP_DEQUEUED)
-		__set_current_state(TASK_STOPPED);
+		set_special_state(TASK_STOPPED);
 	spin_unlock_irq(&current->sighand->siglock);
 
 	schedule();
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 7ad60e0..ffde9ee 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -3508,23 +3508,8 @@ static void __sched notrace __schedule(bool preempt)
 
 void __noreturn do_task_dead(void)
 {
-	/*
-	 * The setting of TASK_RUNNING by try_to_wake_up() may be delayed
-	 * when the following two conditions become true.
-	 *   - There is race condition of mmap_sem (It is acquired by
-	 *     exit_mm()), and
-	 *   - SMI occurs before setting TASK_RUNINNG.
-	 *     (or hypervisor of virtual machine switches to other guest)
-	 *  As a result, we may become TASK_RUNNING after becoming TASK_DEAD
-	 *
-	 * To avoid it, we have to wait for releasing tsk->pi_lock which
-	 * is held by try_to_wake_up()
-	 */
-	raw_spin_lock_irq(&current->pi_lock);
-	raw_spin_unlock_irq(&current->pi_lock);
-
 	/* Causes final put_task_struct in finish_task_switch(): */
-	__set_current_state(TASK_DEAD);
+	set_special_state(TASK_DEAD);
 
 	/* Tell freezer to ignore us: */
 	current->flags |= PF_NOFREEZE;
diff --git a/kernel/signal.c b/kernel/signal.c
index d4ccea5..9c33163 100644
--- a/kernel/signal.c
+++ b/kernel/signal.c
@@ -1961,14 +1961,27 @@ static void ptrace_stop(int exit_code, int why, int clear_code, siginfo_t *info)
 			return;
 	}
 
+	set_special_state(TASK_TRACED);
+
 	/*
 	 * We're committing to trapping.  TRACED should be visible before
 	 * TRAPPING is cleared; otherwise, the tracer might fail do_wait().
 	 * Also, transition to TRACED and updates to ->jobctl should be
 	 * atomic with respect to siglock and should be done after the arch
 	 * hook as siglock is released and regrabbed across it.
+	 *
+	 *     TRACER				    TRACEE
+	 *
+	 *     ptrace_attach()
+	 * [L]   wait_on_bit(JOBCTL_TRAPPING)	[S] set_special_state(TRACED)
+	 *     do_wait()
+	 *       set_current_state()                smp_wmb();
+	 *       ptrace_do_wait()
+	 *         wait_task_stopped()
+	 *           task_stopped_code()
+	 * [L]         task_is_traced()		[S] task_clear_jobctl_trapping();
 	 */
-	set_current_state(TASK_TRACED);
+	smp_wmb();
 
 	current->last_siginfo = info;
 	current->exit_code = exit_code;
@@ -2176,7 +2189,7 @@ static bool do_signal_stop(int signr)
 		if (task_participate_group_stop(current))
 			notify = CLD_STOPPED;
 
-		__set_current_state(TASK_STOPPED);
+		set_special_state(TASK_STOPPED);
 		spin_unlock_irq(&current->sighand->siglock);
 
 		/*
-- 
cgit v1.1


From dc74f6fec68daa7cb34ad9155da3782c0f9bf86a Mon Sep 17 00:00:00 2001
From: Colin Ian King <colin.king@canonical.com>
Date: Thu, 3 May 2018 16:45:10 +0100
Subject: drm/i915/selftests: fix spelling mistake: "parmaters" -> "parameters"

Trivial fix to spelling mistake in pr_err error message

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20180503154510.708-1-colin.king@canonical.com
---
 drivers/gpu/drm/i915/selftests/i915_vma.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/selftests/i915_vma.c b/drivers/gpu/drm/i915/selftests/i915_vma.c
index eb89e30..e90f972 100644
--- a/drivers/gpu/drm/i915/selftests/i915_vma.c
+++ b/drivers/gpu/drm/i915/selftests/i915_vma.c
@@ -81,7 +81,7 @@ checked_vma_instance(struct drm_i915_gem_object *obj,
 	}
 
 	if (i915_vma_compare(vma, vm, view)) {
-		pr_err("i915_vma_compare failed with create parmaters!\n");
+		pr_err("i915_vma_compare failed with create parameters!\n");
 		return ERR_PTR(-EINVAL);
 	}
 
-- 
cgit v1.1


From 3365e2268b6bc3d9fa6550f2deaf1b6a537f8732 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Thu, 3 May 2018 20:51:14 +0100
Subject: drm/i915: Lazily unbind vma on close

When userspace is passing around swapbuffers using DRI, we frequently
have to open and close the same object in the foreign address space.
This shows itself as the same object being rebound at roughly 30fps
(with a second object also being rebound at 30fps), which involves us
having to rewrite the page tables and maintain the drm_mm range manager
every time.

However, since the object still exists and it is only the local handle
that disappears, if we are lazy and do not unbind the VMA immediately
when the local user closes the object but defer it until the GPU is
idle, then we can reuse the same VMA binding. We still have to be
careful to mark the handle and lookup tables as closed to maintain the
uABI, just allowing the underlying VMA to be resurrected if the user is
able to access the same object from the same context again.

If the object itself is destroyed (neither userspace keeping a handle to
it), the VMA will be reaped immediately as usual.

In the future, this will be even more useful as instantiating a new VMA
for use on the GPU will become heavier. A nuisance indeed, so nip it in
the bud.

v2: s/__i915_vma_final_close/i915_vma_destroy/ etc.
v3: Leave a hint as to why we deferred the unbind on close.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180503195115.22309-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/i915_drv.h                  |  1 +
 drivers/gpu/drm/i915/i915_gem.c                  |  4 +-
 drivers/gpu/drm/i915/i915_gem_execbuffer.c       |  3 +-
 drivers/gpu/drm/i915/i915_gem_gtt.c              | 14 +++--
 drivers/gpu/drm/i915/i915_vma.c                  | 73 ++++++++++++++++++------
 drivers/gpu/drm/i915/i915_vma.h                  |  6 ++
 drivers/gpu/drm/i915/selftests/huge_pages.c      |  2 +-
 drivers/gpu/drm/i915/selftests/mock_gem_device.c |  1 +
 8 files changed, 79 insertions(+), 25 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 11ff84e..04e2780 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -2062,6 +2062,7 @@ struct drm_i915_private {
 		struct list_head timelines;
 
 		struct list_head active_rings;
+		struct list_head closed_vma;
 		u32 active_requests;
 		u32 request_serial;
 
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 484354f..5ece6ae 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -165,6 +165,7 @@ static u32 __i915_gem_park(struct drm_i915_private *i915)
 	i915_timelines_park(i915);
 
 	i915_pmu_gt_parked(i915);
+	i915_vma_parked(i915);
 
 	i915->gt.awake = false;
 
@@ -4795,7 +4796,7 @@ static void __i915_gem_free_objects(struct drm_i915_private *i915,
 					 &obj->vma_list, obj_link) {
 			GEM_BUG_ON(i915_vma_is_active(vma));
 			vma->flags &= ~I915_VMA_PIN_MASK;
-			i915_vma_close(vma);
+			i915_vma_destroy(vma);
 		}
 		GEM_BUG_ON(!list_empty(&obj->vma_list));
 		GEM_BUG_ON(!RB_EMPTY_ROOT(&obj->vma_tree));
@@ -5598,6 +5599,7 @@ int i915_gem_init_early(struct drm_i915_private *dev_priv)
 
 	INIT_LIST_HEAD(&dev_priv->gt.timelines);
 	INIT_LIST_HEAD(&dev_priv->gt.active_rings);
+	INIT_LIST_HEAD(&dev_priv->gt.closed_vma);
 
 	i915_gem_init__mm(dev_priv);
 
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index c74f5df..f627a8c 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -762,7 +762,8 @@ static int eb_lookup_vmas(struct i915_execbuffer *eb)
 		}
 
 		/* transfer ref to ctx */
-		vma->open_count++;
+		if (!vma->open_count++)
+			i915_vma_reopen(vma);
 		list_add(&lut->obj_link, &obj->lut_list);
 		list_add(&lut->ctx_link, &eb->ctx->handles_list);
 		lut->ctx = eb->ctx;
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index e9d8283..272d6bb 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -2219,6 +2219,12 @@ i915_ppgtt_create(struct drm_i915_private *dev_priv,
 
 void i915_ppgtt_close(struct i915_address_space *vm)
 {
+	GEM_BUG_ON(vm->closed);
+	vm->closed = true;
+}
+
+static void ppgtt_destroy_vma(struct i915_address_space *vm)
+{
 	struct list_head *phases[] = {
 		&vm->active_list,
 		&vm->inactive_list,
@@ -2226,15 +2232,12 @@ void i915_ppgtt_close(struct i915_address_space *vm)
 		NULL,
 	}, **phase;
 
-	GEM_BUG_ON(vm->closed);
 	vm->closed = true;
-
 	for (phase = phases; *phase; phase++) {
 		struct i915_vma *vma, *vn;
 
 		list_for_each_entry_safe(vma, vn, *phase, vm_link)
-			if (!i915_vma_is_closed(vma))
-				i915_vma_close(vma);
+			i915_vma_destroy(vma);
 	}
 }
 
@@ -2245,7 +2248,8 @@ void i915_ppgtt_release(struct kref *kref)
 
 	trace_i915_ppgtt_release(&ppgtt->base);
 
-	/* vmas should already be unbound and destroyed */
+	ppgtt_destroy_vma(&ppgtt->base);
+
 	GEM_BUG_ON(!list_empty(&ppgtt->base.active_list));
 	GEM_BUG_ON(!list_empty(&ppgtt->base.inactive_list));
 	GEM_BUG_ON(!list_empty(&ppgtt->base.unbound_list));
diff --git a/drivers/gpu/drm/i915/i915_vma.c b/drivers/gpu/drm/i915/i915_vma.c
index 4bda3bd..9324d47 100644
--- a/drivers/gpu/drm/i915/i915_vma.c
+++ b/drivers/gpu/drm/i915/i915_vma.c
@@ -46,8 +46,6 @@ i915_vma_retire(struct i915_gem_active *active, struct i915_request *rq)
 
 	GEM_BUG_ON(!drm_mm_node_allocated(&vma->node));
 	list_move_tail(&vma->vm_link, &vma->vm->inactive_list);
-	if (unlikely(i915_vma_is_closed(vma) && !i915_vma_is_pinned(vma)))
-		WARN_ON(i915_vma_unbind(vma));
 
 	GEM_BUG_ON(!i915_gem_object_is_active(obj));
 	if (--obj->active_count)
@@ -232,7 +230,6 @@ i915_vma_instance(struct drm_i915_gem_object *obj,
 	if (!vma)
 		vma = vma_create(obj, vm, view);
 
-	GEM_BUG_ON(!IS_ERR(vma) && i915_vma_is_closed(vma));
 	GEM_BUG_ON(!IS_ERR(vma) && i915_vma_compare(vma, vm, view));
 	GEM_BUG_ON(!IS_ERR(vma) && vma_lookup(obj, vm, view) != vma);
 	return vma;
@@ -684,13 +681,43 @@ err_unpin:
 	return ret;
 }
 
-static void i915_vma_destroy(struct i915_vma *vma)
+void i915_vma_close(struct i915_vma *vma)
+{
+	lockdep_assert_held(&vma->vm->i915->drm.struct_mutex);
+
+	GEM_BUG_ON(i915_vma_is_closed(vma));
+	vma->flags |= I915_VMA_CLOSED;
+
+	/*
+	 * We defer actually closing, unbinding and destroying the VMA until
+	 * the next idle point, or if the object is freed in the meantime. By
+	 * postponing the unbind, we allow for it to be resurrected by the
+	 * client, avoiding the work required to rebind the VMA. This is
+	 * advantageous for DRI, where the client/server pass objects
+	 * between themselves, temporarily opening a local VMA to the
+	 * object, and then closing it again. The same object is then reused
+	 * on the next frame (or two, depending on the depth of the swap queue)
+	 * causing us to rebind the VMA once more. This ends up being a lot
+	 * of wasted work for the steady state.
+	 */
+	list_add_tail(&vma->closed_link, &vma->vm->i915->gt.closed_vma);
+}
+
+void i915_vma_reopen(struct i915_vma *vma)
+{
+	lockdep_assert_held(&vma->vm->i915->drm.struct_mutex);
+
+	if (vma->flags & I915_VMA_CLOSED) {
+		vma->flags &= ~I915_VMA_CLOSED;
+		list_del(&vma->closed_link);
+	}
+}
+
+static void __i915_vma_destroy(struct i915_vma *vma)
 {
 	int i;
 
 	GEM_BUG_ON(vma->node.allocated);
-	GEM_BUG_ON(i915_vma_is_active(vma));
-	GEM_BUG_ON(!i915_vma_is_closed(vma));
 	GEM_BUG_ON(vma->fence);
 
 	for (i = 0; i < ARRAY_SIZE(vma->last_read); i++)
@@ -699,6 +726,7 @@ static void i915_vma_destroy(struct i915_vma *vma)
 
 	list_del(&vma->obj_link);
 	list_del(&vma->vm_link);
+	rb_erase(&vma->obj_node, &vma->obj->vma_tree);
 
 	if (!i915_vma_is_ggtt(vma))
 		i915_ppgtt_put(i915_vm_to_ppgtt(vma->vm));
@@ -706,15 +734,30 @@ static void i915_vma_destroy(struct i915_vma *vma)
 	kmem_cache_free(to_i915(vma->obj->base.dev)->vmas, vma);
 }
 
-void i915_vma_close(struct i915_vma *vma)
+void i915_vma_destroy(struct i915_vma *vma)
 {
-	GEM_BUG_ON(i915_vma_is_closed(vma));
-	vma->flags |= I915_VMA_CLOSED;
+	lockdep_assert_held(&vma->vm->i915->drm.struct_mutex);
 
-	rb_erase(&vma->obj_node, &vma->obj->vma_tree);
+	GEM_BUG_ON(i915_vma_is_active(vma));
+	GEM_BUG_ON(i915_vma_is_pinned(vma));
+
+	if (i915_vma_is_closed(vma))
+		list_del(&vma->closed_link);
+
+	WARN_ON(i915_vma_unbind(vma));
+	__i915_vma_destroy(vma);
+}
+
+void i915_vma_parked(struct drm_i915_private *i915)
+{
+	struct i915_vma *vma, *next;
 
-	if (!i915_vma_is_active(vma) && !i915_vma_is_pinned(vma))
-		WARN_ON(i915_vma_unbind(vma));
+	list_for_each_entry_safe(vma, next, &i915->gt.closed_vma, closed_link) {
+		GEM_BUG_ON(!i915_vma_is_closed(vma));
+		i915_vma_destroy(vma);
+	}
+
+	GEM_BUG_ON(!list_empty(&i915->gt.closed_vma));
 }
 
 static void __i915_vma_iounmap(struct i915_vma *vma)
@@ -804,7 +847,7 @@ int i915_vma_unbind(struct i915_vma *vma)
 		return -EBUSY;
 
 	if (!drm_mm_node_allocated(&vma->node))
-		goto destroy;
+		return 0;
 
 	GEM_BUG_ON(obj->bind_count == 0);
 	GEM_BUG_ON(!i915_gem_object_has_pinned_pages(obj));
@@ -841,10 +884,6 @@ int i915_vma_unbind(struct i915_vma *vma)
 
 	i915_vma_remove(vma);
 
-destroy:
-	if (unlikely(i915_vma_is_closed(vma)))
-		i915_vma_destroy(vma);
-
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/i915/i915_vma.h b/drivers/gpu/drm/i915/i915_vma.h
index 8c50220..fc4294c 100644
--- a/drivers/gpu/drm/i915/i915_vma.h
+++ b/drivers/gpu/drm/i915/i915_vma.h
@@ -119,6 +119,8 @@ struct i915_vma {
 	/** This vma's place in the eviction list */
 	struct list_head evict_link;
 
+	struct list_head closed_link;
+
 	/**
 	 * Used for performing relocations during execbuffer insertion.
 	 */
@@ -285,6 +287,8 @@ void i915_vma_revoke_mmap(struct i915_vma *vma);
 int __must_check i915_vma_unbind(struct i915_vma *vma);
 void i915_vma_unlink_ctx(struct i915_vma *vma);
 void i915_vma_close(struct i915_vma *vma);
+void i915_vma_reopen(struct i915_vma *vma);
+void i915_vma_destroy(struct i915_vma *vma);
 
 int __i915_vma_do_pin(struct i915_vma *vma,
 		      u64 size, u64 alignment, u64 flags);
@@ -408,6 +412,8 @@ i915_vma_unpin_fence(struct i915_vma *vma)
 		__i915_vma_unpin_fence(vma);
 }
 
+void i915_vma_parked(struct drm_i915_private *i915);
+
 #define for_each_until(cond) if (cond) break; else
 
 /**
diff --git a/drivers/gpu/drm/i915/selftests/huge_pages.c b/drivers/gpu/drm/i915/selftests/huge_pages.c
index 05bbef3..d7c8ef8 100644
--- a/drivers/gpu/drm/i915/selftests/huge_pages.c
+++ b/drivers/gpu/drm/i915/selftests/huge_pages.c
@@ -1091,7 +1091,7 @@ static int __igt_write_huge(struct i915_gem_context *ctx,
 out_vma_unpin:
 	i915_vma_unpin(vma);
 out_vma_close:
-	i915_vma_close(vma);
+	i915_vma_destroy(vma);
 
 	return err;
 }
diff --git a/drivers/gpu/drm/i915/selftests/mock_gem_device.c b/drivers/gpu/drm/i915/selftests/mock_gem_device.c
index a662c04..4b6622c 100644
--- a/drivers/gpu/drm/i915/selftests/mock_gem_device.c
+++ b/drivers/gpu/drm/i915/selftests/mock_gem_device.c
@@ -226,6 +226,7 @@ struct drm_i915_private *mock_gem_device(void)
 
 	INIT_LIST_HEAD(&i915->gt.timelines);
 	INIT_LIST_HEAD(&i915->gt.active_rings);
+	INIT_LIST_HEAD(&i915->gt.closed_vma);
 
 	mutex_lock(&i915->drm.struct_mutex);
 	mock_init_ggtt(i915);
-- 
cgit v1.1


From 7c572e1bdf8dea0c84ce8da01a84cdaa26d8e138 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Thu, 3 May 2018 20:51:15 +0100
Subject: drm/i915: Keep one request in our ring_list

Don't pre-emptively retire the oldest request in our ring's list if it
is the only request. We keep various bits of state alive using the
active reference from the request and would rather transfer that state
over to a new request rather than the more involved process of retiring
and reacquiring it.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180503195115.22309-2-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/i915_request.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c
index 63bb610..d68739b 100644
--- a/drivers/gpu/drm/i915/i915_request.c
+++ b/drivers/gpu/drm/i915/i915_request.c
@@ -695,9 +695,9 @@ i915_request_alloc(struct intel_engine_cs *engine, struct i915_gem_context *ctx)
 		goto err_unreserve;
 
 	/* Move our oldest request to the slab-cache (if not in use!) */
-	rq = list_first_entry_or_null(&ring->request_list,
-				      typeof(*rq), ring_link);
-	if (rq && i915_request_completed(rq))
+	rq = list_first_entry(&ring->request_list, typeof(*rq), ring_link);
+	if (!list_is_last(&rq->ring_link, &ring->request_list) &&
+	    i915_request_completed(rq))
 		i915_request_retire(rq);
 
 	/*
-- 
cgit v1.1


From 74f9474124ea53f98866e434a8080f1538b8c2b7 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Thu, 3 May 2018 20:54:16 +0100
Subject: drm/i915/execlists: Drop preemption arbitrations points along the
 ring
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Limit the arbitration (where preemption may occur) to inside the batch,
and prevent it from happening on the pipecontrols/flushes we use to
write the breadcrumb seqno. Once the user batch is complete, we have
nothing left to do but serialise and emit the breadcrumb; switching
contexts at this point is futile so don't.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Michał Winiarski <michal.winiarski@intel.com>
Cc: Michel Thierry <michel.thierry@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Michel Thierry <michel.thierry@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180503195416.22498-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/intel_lrc.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 3d747d1..9f3cce0 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -1933,7 +1933,7 @@ static int gen8_emit_bb_start(struct i915_request *rq,
 		rq->ctx->ppgtt->pd_dirty_rings &= ~intel_engine_flag(rq->engine);
 	}
 
-	cs = intel_ring_begin(rq, 4);
+	cs = intel_ring_begin(rq, 6);
 	if (IS_ERR(cs))
 		return PTR_ERR(cs);
 
@@ -1962,6 +1962,9 @@ static int gen8_emit_bb_start(struct i915_request *rq,
 		(flags & I915_DISPATCH_RS ? MI_BATCH_RESOURCE_STREAMER : 0);
 	*cs++ = lower_32_bits(offset);
 	*cs++ = upper_32_bits(offset);
+
+	*cs++ = MI_ARB_ON_OFF | MI_ARB_DISABLE;
+	*cs++ = MI_NOOP;
 	intel_ring_advance(rq, cs);
 
 	return 0;
@@ -2104,7 +2107,7 @@ static void gen8_emit_breadcrumb(struct i915_request *request, u32 *cs)
 	cs = gen8_emit_ggtt_write(cs, request->global_seqno,
 				  intel_hws_seqno_address(request->engine));
 	*cs++ = MI_USER_INTERRUPT;
-	*cs++ = MI_NOOP;
+	*cs++ = MI_ARB_ON_OFF | MI_ARB_ENABLE;
 	request->tail = intel_ring_offset(request, cs);
 	assert_ring_tail_valid(request->ring, request->tail);
 
@@ -2120,7 +2123,7 @@ static void gen8_emit_breadcrumb_rcs(struct i915_request *request, u32 *cs)
 	cs = gen8_emit_ggtt_write_rcs(cs, request->global_seqno,
 				      intel_hws_seqno_address(request->engine));
 	*cs++ = MI_USER_INTERRUPT;
-	*cs++ = MI_NOOP;
+	*cs++ = MI_ARB_ON_OFF | MI_ARB_ENABLE;
 	request->tail = intel_ring_offset(request, cs);
 	assert_ring_tail_valid(request->ring, request->tail);
 
-- 
cgit v1.1


From 47d4cb8ae8e7f70363ece0bcc4ffc76a9e4638dc Mon Sep 17 00:00:00 2001
From: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Date: Thu, 3 May 2018 21:17:06 +0300
Subject: i915: Convert to use match_string() helper

The new helper returns index of the matching string in an array.
We are going to use it here.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180503181706.22120-1-andriy.shevchenko@linux.intel.com
---
 drivers/gpu/drm/i915/intel_pipe_crc.c | 22 ++++++++++------------
 1 file changed, 10 insertions(+), 12 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_pipe_crc.c b/drivers/gpu/drm/i915/intel_pipe_crc.c
index 4f367c1..39a4e4e 100644
--- a/drivers/gpu/drm/i915/intel_pipe_crc.c
+++ b/drivers/gpu/drm/i915/intel_pipe_crc.c
@@ -766,13 +766,12 @@ display_crc_ctl_parse_object(const char *buf, enum intel_pipe_crc_object *o)
 {
 	int i;
 
-	for (i = 0; i < ARRAY_SIZE(pipe_crc_objects); i++)
-		if (!strcmp(buf, pipe_crc_objects[i])) {
-			*o = i;
-			return 0;
-		}
+	i = match_string(pipe_crc_objects, ARRAY_SIZE(pipe_crc_objects), buf);
+	if (i < 0)
+		return i;
 
-	return -EINVAL;
+	*o = i;
+	return 0;
 }
 
 static int display_crc_ctl_parse_pipe(struct drm_i915_private *dev_priv,
@@ -798,13 +797,12 @@ display_crc_ctl_parse_source(const char *buf, enum intel_pipe_crc_source *s)
 		return 0;
 	}
 
-	for (i = 0; i < ARRAY_SIZE(pipe_crc_sources); i++)
-		if (!strcmp(buf, pipe_crc_sources[i])) {
-			*s = i;
-			return 0;
-		}
+	i = match_string(pipe_crc_sources, ARRAY_SIZE(pipe_crc_sources), buf);
+	if (i < 0)
+		return i;
 
-	return -EINVAL;
+	*s = i;
+	return 0;
 }
 
 static int display_crc_ctl_parse(struct drm_i915_private *dev_priv,
-- 
cgit v1.1


From b76401fc4ba720f0f38f7b1f9d54d5c2308bc18d Mon Sep 17 00:00:00 2001
From: Chris Packham <chris.packham@alliedtelesis.co.nz>
Date: Thu, 3 May 2018 14:21:28 +1200
Subject: mtd: rawnand: marvell: pass ms delay to wait_op

marvell_nfc_wait_op() expects the delay to be expressed in milliseconds
but nand_sdr_timings uses picoseconds. Use PSEC_TO_MSEC when passing
tPROG_max to marvell_nfc_wait_op().

Fixes: 02f26ecf8c772 ("mtd: nand: add reworked Marvell NAND controller driver")
Cc: stable@vger.kernel.org
Signed-off-by: Chris Packham <chris.packham@alliedtelesis.co.nz>
Reviewed-by: Miquel Raynal <miquel.raynal@bootlin.com>
Signed-off-by: Boris Brezillon <boris.brezillon@bootlin.com>
---
 drivers/mtd/nand/raw/marvell_nand.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/mtd/nand/raw/marvell_nand.c b/drivers/mtd/nand/raw/marvell_nand.c
index 1d779a3..e4b964f 100644
--- a/drivers/mtd/nand/raw/marvell_nand.c
+++ b/drivers/mtd/nand/raw/marvell_nand.c
@@ -1074,7 +1074,7 @@ static int marvell_nfc_hw_ecc_hmg_do_write_page(struct nand_chip *chip,
 		return ret;
 
 	ret = marvell_nfc_wait_op(chip,
-				  chip->data_interface.timings.sdr.tPROG_max);
+				  PSEC_TO_MSEC(chip->data_interface.timings.sdr.tPROG_max));
 	return ret;
 }
 
@@ -1494,7 +1494,7 @@ static int marvell_nfc_hw_ecc_bch_write_page(struct mtd_info *mtd,
 	}
 
 	ret = marvell_nfc_wait_op(chip,
-				  chip->data_interface.timings.sdr.tPROG_max);
+				  PSEC_TO_MSEC(chip->data_interface.timings.sdr.tPROG_max));
 
 	marvell_nfc_disable_hw_ecc(chip);
 
-- 
cgit v1.1


From a2ee41fd953e7c3ff6c55a3038c80354d191a318 Mon Sep 17 00:00:00 2001
From: Miquel Raynal <miquel.raynal@bootlin.com>
Date: Thu, 3 May 2018 12:00:27 +0200
Subject: mtd: rawnand: marvell: fix command xtype in BCH write hook

One layout supported by the Marvell NAND controller supports NAND pages
of 2048 bytes, all handled in one single chunk when using BCH with a
strength of 4-bit per 512 bytes. In this case, instead of the generic
XTYPE_WRITE_DISPATCH/XTYPE_LAST_NAKED_RW couple, the controller expects
to receive XTYPE_MONOLITHIC_RW.

This fixes problems at boot like:

[    1.315475] Scanning device for bad blocks
[    3.203108] marvell-nfc f10d0000.flash: Timeout waiting for RB signal
[    3.209564] nand_bbt: error while writing BBT block -110
[    4.243106] marvell-nfc f10d0000.flash: Timeout waiting for RB signal
[    5.283106] marvell-nfc f10d0000.flash: Timeout waiting for RB signal
[    5.289562] nand_bbt: error -110 while marking block 2047 bad
[    6.323106] marvell-nfc f10d0000.flash: Timeout waiting for RB signal
[    6.329559] nand_bbt: error while writing BBT block -110
[    7.363106] marvell-nfc f10d0000.flash: Timeout waiting for RB signal
[    8.403105] marvell-nfc f10d0000.flash: Timeout waiting for RB signal
[    8.409559] nand_bbt: error -110 while marking block 2046 bad
...

Fixes: 02f26ecf8c772 ("mtd: nand: add reworked Marvell NAND controller driver")
Cc: stable@vger.kernel.org
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Tested-by: Chris Packham <chris.packham@alliedtelesis.co.nz>
Signed-off-by: Boris Brezillon <boris.brezillon@bootlin.com>
---
 drivers/mtd/nand/raw/marvell_nand.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/drivers/mtd/nand/raw/marvell_nand.c b/drivers/mtd/nand/raw/marvell_nand.c
index e4b964f..db5ec4e 100644
--- a/drivers/mtd/nand/raw/marvell_nand.c
+++ b/drivers/mtd/nand/raw/marvell_nand.c
@@ -1408,6 +1408,7 @@ marvell_nfc_hw_ecc_bch_write_chunk(struct nand_chip *chip, int chunk,
 	struct marvell_nand_chip *marvell_nand = to_marvell_nand(chip);
 	struct marvell_nfc *nfc = to_marvell_nfc(chip->controller);
 	const struct marvell_hw_ecc_layout *lt = to_marvell_nand(chip)->layout;
+	u32 xtype;
 	int ret;
 	struct marvell_nfc_op nfc_op = {
 		.ndcb[0] = NDCB0_CMD_TYPE(TYPE_WRITE) | NDCB0_LEN_OVRD,
@@ -1423,7 +1424,12 @@ marvell_nfc_hw_ecc_bch_write_chunk(struct nand_chip *chip, int chunk,
 	 * last naked write.
 	 */
 	if (chunk == 0) {
-		nfc_op.ndcb[0] |= NDCB0_CMD_XTYPE(XTYPE_WRITE_DISPATCH) |
+		if (lt->nchunks == 1)
+			xtype = XTYPE_MONOLITHIC_RW;
+		else
+			xtype = XTYPE_WRITE_DISPATCH;
+
+		nfc_op.ndcb[0] |= NDCB0_CMD_XTYPE(xtype) |
 				  NDCB0_ADDR_CYC(marvell_nand->addr_cyc) |
 				  NDCB0_CMD1(NAND_CMD_SEQIN);
 		nfc_op.ndcb[1] |= NDCB1_ADDRS_PAGE(page);
-- 
cgit v1.1


From 755396163148b50fe1afb4bdd3365e47f3ff7a42 Mon Sep 17 00:00:00 2001
From: Jiri Slaby <jslaby@suse.cz>
Date: Tue, 24 Apr 2018 09:43:44 +0200
Subject: tools: power/acpi, revert to LD = gcc

Commit 7ed1c1901fe5 (tools: fix cross-compile var clobbering) removed
setting of LD to $(CROSS_COMPILE)gcc. This broke build of acpica
(acpidump) in power/acpi:
 ld: unrecognized option '-D_LINUX'

The tools pass CFLAGS to the linker (incl. -D_LINUX), so revert this
particular change and let LD be $(CC) again. Note that the old behaviour
was a bit different, it used $(CROSS_COMPILE)gcc which was eliminated by
the commit 7ed1c1901fe5. We use $(CC) for that reason.

Fixes: 7ed1c1901fe5 (tools: fix cross-compile var clobbering)
Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Cc: 4.16+ <stable@vger.kernel.org> # 4.16+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
---
 tools/power/acpi/Makefile.config | 1 +
 1 file changed, 1 insertion(+)

diff --git a/tools/power/acpi/Makefile.config b/tools/power/acpi/Makefile.config
index 2cccbba..f304be7 100644
--- a/tools/power/acpi/Makefile.config
+++ b/tools/power/acpi/Makefile.config
@@ -56,6 +56,7 @@ INSTALL_SCRIPT = ${INSTALL_PROGRAM}
 # to compile vs uClibc, that can be done here as well.
 CROSS = #/usr/i386-linux-uclibc/usr/bin/i386-uclibc-
 CROSS_COMPILE ?= $(CROSS)
+LD = $(CC)
 HOSTCC = gcc
 
 # check if compiler option is supported
-- 
cgit v1.1


From b1a3dc0b85bde4d8d549ea3aa31106b599694f37 Mon Sep 17 00:00:00 2001
From: Stefan Schake <stschake@gmail.com>
Date: Wed, 18 Apr 2018 03:40:19 -0700
Subject: drm/tegra: hub: Use state directly

Using drm_atomic_get_private_obj_state() after state has been swapped
will return old state.

Fixes: 0281c4149021 ("drm/tegra: hub: Use private object for global state")
Signed-off-by: Stefan Schake <stschake@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/hub.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/tegra/hub.c b/drivers/gpu/drm/tegra/hub.c
index 9a3f23d..8f4fcbb 100644
--- a/drivers/gpu/drm/tegra/hub.c
+++ b/drivers/gpu/drm/tegra/hub.c
@@ -687,7 +687,7 @@ void tegra_display_hub_atomic_commit(struct drm_device *drm,
 	struct device *dev = hub->client.dev;
 	int err;
 
-	hub_state = tegra_display_hub_get_state(hub, state);
+	hub_state = to_tegra_display_hub_state(hub->base.state);
 
 	if (hub_state->clk) {
 		err = clk_set_rate(hub_state->clk, hub_state->rate);
-- 
cgit v1.1


From 2b7db29b79190f7ad5c32f63594ba08b9b9171ea Mon Sep 17 00:00:00 2001
From: "From: Gustavo A. R. Silva" <gustavo@embeddedor.com>
Date: Mon, 16 Apr 2018 13:28:56 -0400
Subject: media: imx-media-csi: Fix inconsistent IS_ERR and PTR_ERR

Fix inconsistent IS_ERR and PTR_ERR in imx_csi_probe.
The proper pointer to be passed as argument is pinctrl
instead of priv->vdev.

This issue was detected with the help of Coccinelle.

Fixes: 52e17089d185 ("media: imx: Don't initialize vars that won't be used")

Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
Reviewed-by: Philipp Zabel <p.zabel@pengutronix.de>
Tested-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
---
 drivers/staging/media/imx/imx-media-csi.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/staging/media/imx/imx-media-csi.c b/drivers/staging/media/imx/imx-media-csi.c
index 16cab40..aeab05f 100644
--- a/drivers/staging/media/imx/imx-media-csi.c
+++ b/drivers/staging/media/imx/imx-media-csi.c
@@ -1799,7 +1799,7 @@ static int imx_csi_probe(struct platform_device *pdev)
 	priv->dev->of_node = pdata->of_node;
 	pinctrl = devm_pinctrl_get_select_default(priv->dev);
 	if (IS_ERR(pinctrl)) {
-		ret = PTR_ERR(priv->vdev);
+		ret = PTR_ERR(pinctrl);
 		dev_dbg(priv->dev,
 			"devm_pinctrl_get_select_default() failed: %d\n", ret);
 		if (ret != -ENODEV)
-- 
cgit v1.1


From 3259081991a9398434f6f49468b960f136ac0158 Mon Sep 17 00:00:00 2001
From: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
Date: Wed, 25 Apr 2018 05:34:48 -0400
Subject: MAINTAINERS & files: Canonize the e-mails I use at files

From now on, I'll start using my @kernel.org as my development e-mail.

As such, let's remove the entries that point to the old
mchehab@s-opensource.com at MAINTAINERS file.

For the files written with a copyright with mchehab@s-opensource,
let's keep Samsung on their names, using mchehab+samsung@kernel.org,
in order to keep pointing to my employer, with sponsors the work.

For the files written before I join Samsung (on July, 4 2013),
let's just use mchehab@kernel.org.

For bug reports, we can simply point to just kernel.org, as
this will reach my mchehab+samsung inbox anyway.

Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Signed-off-by: Brian Warner <brian.warner@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
---
 Documentation/doc-guide/parse-headers.rst               |  4 ++--
 Documentation/media/uapi/rc/keytable.c.rst              |  2 +-
 Documentation/media/uapi/v4l/v4l2grab.c.rst             |  2 +-
 Documentation/sphinx/parse-headers.pl                   |  4 ++--
 .../translations/zh_CN/video4linux/v4l2-framework.txt   |  4 ++--
 MAINTAINERS                                             | 17 -----------------
 drivers/media/i2c/saa7115.c                             |  2 +-
 drivers/media/i2c/saa711x_regs.h                        |  2 +-
 drivers/media/i2c/tda7432.c                             |  2 +-
 drivers/media/i2c/tvp5150.c                             |  2 +-
 drivers/media/i2c/tvp5150_reg.h                         |  2 +-
 drivers/media/i2c/tvp7002.c                             |  2 +-
 drivers/media/i2c/tvp7002_reg.h                         |  2 +-
 drivers/media/media-devnode.c                           |  2 +-
 drivers/media/pci/bt8xx/bttv-audio-hook.c               |  2 +-
 drivers/media/pci/bt8xx/bttv-audio-hook.h               |  2 +-
 drivers/media/pci/bt8xx/bttv-cards.c                    |  4 ++--
 drivers/media/pci/bt8xx/bttv-driver.c                   |  2 +-
 drivers/media/pci/bt8xx/bttv-i2c.c                      |  2 +-
 drivers/media/pci/cx23885/cx23885-input.c               |  2 +-
 drivers/media/pci/cx88/cx88-alsa.c                      |  4 ++--
 drivers/media/pci/cx88/cx88-blackbird.c                 |  2 +-
 drivers/media/pci/cx88/cx88-core.c                      |  2 +-
 drivers/media/pci/cx88/cx88-i2c.c                       |  2 +-
 drivers/media/pci/cx88/cx88-video.c                     |  2 +-
 drivers/media/radio/radio-aimslab.c                     |  2 +-
 drivers/media/radio/radio-aztech.c                      |  2 +-
 drivers/media/radio/radio-gemtek.c                      |  2 +-
 drivers/media/radio/radio-maxiradio.c                   |  2 +-
 drivers/media/radio/radio-rtrack2.c                     |  2 +-
 drivers/media/radio/radio-sf16fmi.c                     |  2 +-
 drivers/media/radio/radio-terratec.c                    |  2 +-
 drivers/media/radio/radio-trust.c                       |  2 +-
 drivers/media/radio/radio-typhoon.c                     |  2 +-
 drivers/media/radio/radio-zoltrix.c                     |  2 +-
 drivers/media/rc/keymaps/rc-avermedia-m135a.c           |  2 +-
 drivers/media/rc/keymaps/rc-encore-enltv-fm53.c         |  2 +-
 drivers/media/rc/keymaps/rc-encore-enltv2.c             |  2 +-
 drivers/media/rc/keymaps/rc-kaiomy.c                    |  2 +-
 drivers/media/rc/keymaps/rc-kworld-plus-tv-analog.c     |  2 +-
 drivers/media/rc/keymaps/rc-pixelview-new.c             |  2 +-
 drivers/media/tuners/tea5761.c                          |  4 ++--
 drivers/media/tuners/tea5767.c                          |  4 ++--
 drivers/media/tuners/tuner-xc2028-types.h               |  2 +-
 drivers/media/tuners/tuner-xc2028.c                     |  4 ++--
 drivers/media/tuners/tuner-xc2028.h                     |  2 +-
 drivers/media/usb/em28xx/em28xx-camera.c                |  2 +-
 drivers/media/usb/em28xx/em28xx-cards.c                 |  2 +-
 drivers/media/usb/em28xx/em28xx-core.c                  |  4 ++--
 drivers/media/usb/em28xx/em28xx-dvb.c                   |  4 ++--
 drivers/media/usb/em28xx/em28xx-i2c.c                   |  2 +-
 drivers/media/usb/em28xx/em28xx-input.c                 |  2 +-
 drivers/media/usb/em28xx/em28xx-video.c                 |  4 ++--
 drivers/media/usb/em28xx/em28xx.h                       |  2 +-
 drivers/media/usb/gspca/zc3xx-reg.h                     |  2 +-
 drivers/media/usb/tm6000/tm6000-cards.c                 |  2 +-
 drivers/media/usb/tm6000/tm6000-core.c                  |  2 +-
 drivers/media/usb/tm6000/tm6000-i2c.c                   |  2 +-
 drivers/media/usb/tm6000/tm6000-regs.h                  |  2 +-
 drivers/media/usb/tm6000/tm6000-usb-isoc.h              |  2 +-
 drivers/media/usb/tm6000/tm6000-video.c                 |  2 +-
 drivers/media/usb/tm6000/tm6000.h                       |  2 +-
 drivers/media/v4l2-core/v4l2-dev.c                      |  4 ++--
 drivers/media/v4l2-core/v4l2-ioctl.c                    |  2 +-
 drivers/media/v4l2-core/videobuf-core.c                 |  6 +++---
 drivers/media/v4l2-core/videobuf-dma-contig.c           |  2 +-
 drivers/media/v4l2-core/videobuf-dma-sg.c               |  6 +++---
 drivers/media/v4l2-core/videobuf-vmalloc.c              |  4 ++--
 include/media/i2c/tvp7002.h                             |  2 +-
 include/media/videobuf-core.h                           |  4 ++--
 include/media/videobuf-dma-sg.h                         |  4 ++--
 include/media/videobuf-vmalloc.h                        |  2 +-
 scripts/extract_xc3028.pl                               |  2 +-
 scripts/split-man.pl                                    |  2 +-
 74 files changed, 92 insertions(+), 109 deletions(-)

diff --git a/Documentation/doc-guide/parse-headers.rst b/Documentation/doc-guide/parse-headers.rst
index 96a0423..82a3e43 100644
--- a/Documentation/doc-guide/parse-headers.rst
+++ b/Documentation/doc-guide/parse-headers.rst
@@ -177,14 +177,14 @@ BUGS
 ****
 
 
-Report bugs to Mauro Carvalho Chehab <mchehab@s-opensource.com>
+Report bugs to Mauro Carvalho Chehab <mchehab@kernel.org>
 
 
 COPYRIGHT
 *********
 
 
-Copyright (c) 2016 by Mauro Carvalho Chehab <mchehab@s-opensource.com>.
+Copyright (c) 2016 by Mauro Carvalho Chehab <mchehab+samsung@kernel.org>.
 
 License GPLv2: GNU GPL version 2 <http://gnu.org/licenses/gpl.html>.
 
diff --git a/Documentation/media/uapi/rc/keytable.c.rst b/Documentation/media/uapi/rc/keytable.c.rst
index e6ce1e3f..217237f 100644
--- a/Documentation/media/uapi/rc/keytable.c.rst
+++ b/Documentation/media/uapi/rc/keytable.c.rst
@@ -7,7 +7,7 @@ file: uapi/v4l/keytable.c
 
     /* keytable.c - This program allows checking/replacing keys at IR
 
-       Copyright (C) 2006-2009 Mauro Carvalho Chehab <mchehab@infradead.org>
+       Copyright (C) 2006-2009 Mauro Carvalho Chehab <mchehab@kernel.org>
 
        This program is free software; you can redistribute it and/or modify
        it under the terms of the GNU General Public License as published by
diff --git a/Documentation/media/uapi/v4l/v4l2grab.c.rst b/Documentation/media/uapi/v4l/v4l2grab.c.rst
index 5aabd0b..f0d0ab6 100644
--- a/Documentation/media/uapi/v4l/v4l2grab.c.rst
+++ b/Documentation/media/uapi/v4l/v4l2grab.c.rst
@@ -6,7 +6,7 @@ file: media/v4l/v4l2grab.c
 .. code-block:: c
 
     /* V4L2 video picture grabber
-       Copyright (C) 2009 Mauro Carvalho Chehab <mchehab@infradead.org>
+       Copyright (C) 2009 Mauro Carvalho Chehab <mchehab@kernel.org>
 
        This program is free software; you can redistribute it and/or modify
        it under the terms of the GNU General Public License as published by
diff --git a/Documentation/sphinx/parse-headers.pl b/Documentation/sphinx/parse-headers.pl
index a958d8b..d410f47 100755
--- a/Documentation/sphinx/parse-headers.pl
+++ b/Documentation/sphinx/parse-headers.pl
@@ -387,11 +387,11 @@ tree for more details.
 
 =head1 BUGS
 
-Report bugs to Mauro Carvalho Chehab <mchehab@s-opensource.com>
+Report bugs to Mauro Carvalho Chehab <mchehab@kernel.org>
 
 =head1 COPYRIGHT
 
-Copyright (c) 2016 by Mauro Carvalho Chehab <mchehab@s-opensource.com>.
+Copyright (c) 2016 by Mauro Carvalho Chehab <mchehab+samsung@kernel.org>.
 
 License GPLv2: GNU GPL version 2 <http://gnu.org/licenses/gpl.html>.
 
diff --git a/Documentation/translations/zh_CN/video4linux/v4l2-framework.txt b/Documentation/translations/zh_CN/video4linux/v4l2-framework.txt
index 698660b..c77c0f0 100644
--- a/Documentation/translations/zh_CN/video4linux/v4l2-framework.txt
+++ b/Documentation/translations/zh_CN/video4linux/v4l2-framework.txt
@@ -6,7 +6,7 @@ communicating in English you can also ask the Chinese maintainer for
 help.  Contact the Chinese maintainer if this translation is outdated
 or if there is a problem with the translation.
 
-Maintainer: Mauro Carvalho Chehab <mchehab@infradead.org>
+Maintainer: Mauro Carvalho Chehab <mchehab@kernel.org>
 Chinese maintainer: Fu Wei <tekkamanninja@gmail.com>
 ---------------------------------------------------------------------
 Documentation/video4linux/v4l2-framework.txt 的中文翻译
@@ -14,7 +14,7 @@ Documentation/video4linux/v4l2-framework.txt 的中文翻译
 如果想评论或更新本文的内容，请直接联系原文档的维护者。如果你使用英文
 交流有困难的话，也可以向中文版维护者求助。如果本翻译更新不及时或者翻
 译存在问题，请联系中文版维护者。
-英文版维护者： Mauro Carvalho Chehab <mchehab@infradead.org>
+英文版维护者： Mauro Carvalho Chehab <mchehab@kernel.org>
 中文版维护者： 傅炜 Fu Wei <tekkamanninja@gmail.com>
 中文版翻译者： 傅炜 Fu Wei <tekkamanninja@gmail.com>
 中文版校译者： 傅炜 Fu Wei <tekkamanninja@gmail.com>
diff --git a/MAINTAINERS b/MAINTAINERS
index 79bb02f..64a3159 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -2554,7 +2554,6 @@ F:	Documentation/devicetree/bindings/sound/axentia,*
 F:	sound/soc/atmel/tse850-pcm5142.c
 
 AZ6007 DVB DRIVER
-M:	Mauro Carvalho Chehab <mchehab@s-opensource.com>
 M:	Mauro Carvalho Chehab <mchehab@kernel.org>
 L:	linux-media@vger.kernel.org
 W:	https://linuxtv.org
@@ -3083,7 +3082,6 @@ F:	include/linux/btrfs*
 F:	include/uapi/linux/btrfs*
 
 BTTV VIDEO4LINUX DRIVER
-M:	Mauro Carvalho Chehab <mchehab@s-opensource.com>
 M:	Mauro Carvalho Chehab <mchehab@kernel.org>
 L:	linux-media@vger.kernel.org
 W:	https://linuxtv.org
@@ -3812,7 +3810,6 @@ S:	Maintained
 F:	drivers/media/dvb-frontends/cx24120*
 
 CX88 VIDEO4LINUX DRIVER
-M:	Mauro Carvalho Chehab <mchehab@s-opensource.com>
 M:	Mauro Carvalho Chehab <mchehab@kernel.org>
 L:	linux-media@vger.kernel.org
 W:	https://linuxtv.org
@@ -5053,7 +5050,6 @@ F:	drivers/edac/thunderx_edac*
 
 EDAC-CORE
 M:	Borislav Petkov <bp@alien8.de>
-M:	Mauro Carvalho Chehab <mchehab@s-opensource.com>
 M:	Mauro Carvalho Chehab <mchehab@kernel.org>
 L:	linux-edac@vger.kernel.org
 T:	git git://git.kernel.org/pub/scm/linux/kernel/git/bp/bp.git for-next
@@ -5082,7 +5078,6 @@ S:	Maintained
 F:	drivers/edac/fsl_ddr_edac.*
 
 EDAC-GHES
-M:	Mauro Carvalho Chehab <mchehab@s-opensource.com>
 M:	Mauro Carvalho Chehab <mchehab@kernel.org>
 L:	linux-edac@vger.kernel.org
 S:	Maintained
@@ -5099,21 +5094,18 @@ S:	Maintained
 F:	drivers/edac/i5000_edac.c
 
 EDAC-I5400
-M:	Mauro Carvalho Chehab <mchehab@s-opensource.com>
 M:	Mauro Carvalho Chehab <mchehab@kernel.org>
 L:	linux-edac@vger.kernel.org
 S:	Maintained
 F:	drivers/edac/i5400_edac.c
 
 EDAC-I7300
-M:	Mauro Carvalho Chehab <mchehab@s-opensource.com>
 M:	Mauro Carvalho Chehab <mchehab@kernel.org>
 L:	linux-edac@vger.kernel.org
 S:	Maintained
 F:	drivers/edac/i7300_edac.c
 
 EDAC-I7CORE
-M:	Mauro Carvalho Chehab <mchehab@s-opensource.com>
 M:	Mauro Carvalho Chehab <mchehab@kernel.org>
 L:	linux-edac@vger.kernel.org
 S:	Maintained
@@ -5163,7 +5155,6 @@ S:	Maintained
 F:	drivers/edac/r82600_edac.c
 
 EDAC-SBRIDGE
-M:	Mauro Carvalho Chehab <mchehab@s-opensource.com>
 M:	Mauro Carvalho Chehab <mchehab@kernel.org>
 L:	linux-edac@vger.kernel.org
 S:	Maintained
@@ -5222,7 +5213,6 @@ S:	Maintained
 F:	drivers/net/ethernet/ibm/ehea/
 
 EM28XX VIDEO4LINUX DRIVER
-M:	Mauro Carvalho Chehab <mchehab@s-opensource.com>
 M:	Mauro Carvalho Chehab <mchehab@kernel.org>
 L:	linux-media@vger.kernel.org
 W:	https://linuxtv.org
@@ -8871,7 +8861,6 @@ F:	Documentation/devicetree/bindings/media/nvidia,tegra-vde.txt
 F:	drivers/staging/media/tegra-vde/
 
 MEDIA INPUT INFRASTRUCTURE (V4L/DVB)
-M:	Mauro Carvalho Chehab <mchehab@s-opensource.com>
 M:	Mauro Carvalho Chehab <mchehab@kernel.org>
 P:	LinuxTV.org Project
 L:	linux-media@vger.kernel.org
@@ -12259,7 +12248,6 @@ S:	Odd Fixes
 F:	drivers/media/i2c/saa6588*
 
 SAA7134 VIDEO4LINUX DRIVER
-M:	Mauro Carvalho Chehab <mchehab@s-opensource.com>
 M:	Mauro Carvalho Chehab <mchehab@kernel.org>
 L:	linux-media@vger.kernel.org
 W:	https://linuxtv.org
@@ -12763,7 +12751,6 @@ S:	Maintained
 F:	drivers/media/radio/si4713/radio-usb-si4713.c
 
 SIANO DVB DRIVER
-M:	Mauro Carvalho Chehab <mchehab@s-opensource.com>
 M:	Mauro Carvalho Chehab <mchehab@kernel.org>
 L:	linux-media@vger.kernel.org
 W:	https://linuxtv.org
@@ -13754,7 +13741,6 @@ S:	Maintained
 F:	drivers/media/i2c/tda9840*
 
 TEA5761 TUNER DRIVER
-M:	Mauro Carvalho Chehab <mchehab@s-opensource.com>
 M:	Mauro Carvalho Chehab <mchehab@kernel.org>
 L:	linux-media@vger.kernel.org
 W:	https://linuxtv.org
@@ -13763,7 +13749,6 @@ S:	Odd fixes
 F:	drivers/media/tuners/tea5761.*
 
 TEA5767 TUNER DRIVER
-M:	Mauro Carvalho Chehab <mchehab@s-opensource.com>
 M:	Mauro Carvalho Chehab <mchehab@kernel.org>
 L:	linux-media@vger.kernel.org
 W:	https://linuxtv.org
@@ -14180,7 +14165,6 @@ F:	Documentation/networking/tlan.txt
 F:	drivers/net/ethernet/ti/tlan.*
 
 TM6000 VIDEO4LINUX DRIVER
-M:	Mauro Carvalho Chehab <mchehab@s-opensource.com>
 M:	Mauro Carvalho Chehab <mchehab@kernel.org>
 L:	linux-media@vger.kernel.org
 W:	https://linuxtv.org
@@ -15407,7 +15391,6 @@ S:	Maintained
 F:	arch/x86/entry/vdso/
 
 XC2028/3028 TUNER DRIVER
-M:	Mauro Carvalho Chehab <mchehab@s-opensource.com>
 M:	Mauro Carvalho Chehab <mchehab@kernel.org>
 L:	linux-media@vger.kernel.org
 W:	https://linuxtv.org
diff --git a/drivers/media/i2c/saa7115.c b/drivers/media/i2c/saa7115.c
index e216cd7..b07114b 100644
--- a/drivers/media/i2c/saa7115.c
+++ b/drivers/media/i2c/saa7115.c
@@ -20,7 +20,7 @@
 //
 // VBI support (2004) and cleanups (2005) by Hans Verkuil <hverkuil@xs4all.nl>
 //
-// Copyright (c) 2005-2006 Mauro Carvalho Chehab <mchehab@infradead.org>
+// Copyright (c) 2005-2006 Mauro Carvalho Chehab <mchehab@kernel.org>
 //	SAA7111, SAA7113 and SAA7118 support
 
 #include "saa711x_regs.h"
diff --git a/drivers/media/i2c/saa711x_regs.h b/drivers/media/i2c/saa711x_regs.h
index a50d480..44fabe0 100644
--- a/drivers/media/i2c/saa711x_regs.h
+++ b/drivers/media/i2c/saa711x_regs.h
@@ -2,7 +2,7 @@
  * SPDX-License-Identifier: GPL-2.0+
  * saa711x - Philips SAA711x video decoder register specifications
  *
- * Copyright (c) 2006 Mauro Carvalho Chehab <mchehab@infradead.org>
+ * Copyright (c) 2006 Mauro Carvalho Chehab <mchehab@kernel.org>
  */
 
 #define R_00_CHIP_VERSION                             0x00
diff --git a/drivers/media/i2c/tda7432.c b/drivers/media/i2c/tda7432.c
index 1c5c61d..9b4f212 100644
--- a/drivers/media/i2c/tda7432.c
+++ b/drivers/media/i2c/tda7432.c
@@ -8,7 +8,7 @@
  * Muting and tone control by Jonathan Isom <jisom@ematic.com>
  *
  * Copyright (c) 2000 Eric Sandeen <eric_sandeen@bigfoot.com>
- * Copyright (c) 2006 Mauro Carvalho Chehab <mchehab@infradead.org>
+ * Copyright (c) 2006 Mauro Carvalho Chehab <mchehab@kernel.org>
  * This code is placed under the terms of the GNU General Public License
  * Based on tda9855.c by Steve VanDeBogart (vandebo@uclink.berkeley.edu)
  * Which was based on tda8425.c by Greg Alexander (c) 1998
diff --git a/drivers/media/i2c/tvp5150.c b/drivers/media/i2c/tvp5150.c
index 2476d81..1734ed4 100644
--- a/drivers/media/i2c/tvp5150.c
+++ b/drivers/media/i2c/tvp5150.c
@@ -2,7 +2,7 @@
 //
 // tvp5150 - Texas Instruments TVP5150A/AM1 and TVP5151 video decoder driver
 //
-// Copyright (c) 2005,2006 Mauro Carvalho Chehab <mchehab@infradead.org>
+// Copyright (c) 2005,2006 Mauro Carvalho Chehab <mchehab@kernel.org>
 
 #include <dt-bindings/media/tvp5150.h>
 #include <linux/i2c.h>
diff --git a/drivers/media/i2c/tvp5150_reg.h b/drivers/media/i2c/tvp5150_reg.h
index c43b7b8..d3a764c 100644
--- a/drivers/media/i2c/tvp5150_reg.h
+++ b/drivers/media/i2c/tvp5150_reg.h
@@ -3,7 +3,7 @@
  *
  * tvp5150 - Texas Instruments TVP5150A/AM1 video decoder registers
  *
- * Copyright (c) 2005,2006 Mauro Carvalho Chehab <mchehab@infradead.org>
+ * Copyright (c) 2005,2006 Mauro Carvalho Chehab <mchehab@kernel.org>
  */
 
 #define TVP5150_VD_IN_SRC_SEL_1      0x00 /* Video input source selection #1 */
diff --git a/drivers/media/i2c/tvp7002.c b/drivers/media/i2c/tvp7002.c
index a26c1a3..4599b7e 100644
--- a/drivers/media/i2c/tvp7002.c
+++ b/drivers/media/i2c/tvp7002.c
@@ -5,7 +5,7 @@
  * Author: Santiago Nunez-Corrales <santiago.nunez@ridgerun.com>
  *
  * This code is partially based upon the TVP5150 driver
- * written by Mauro Carvalho Chehab (mchehab@infradead.org),
+ * written by Mauro Carvalho Chehab <mchehab@kernel.org>,
  * the TVP514x driver written by Vaibhav Hiremath <hvaibhav@ti.com>
  * and the TVP7002 driver in the TI LSP 2.10.00.14. Revisions by
  * Muralidharan Karicheri and Snehaprabha Narnakaje (TI).
diff --git a/drivers/media/i2c/tvp7002_reg.h b/drivers/media/i2c/tvp7002_reg.h
index 3c8c8b0..7f56ba6 100644
--- a/drivers/media/i2c/tvp7002_reg.h
+++ b/drivers/media/i2c/tvp7002_reg.h
@@ -5,7 +5,7 @@
  * Author: Santiago Nunez-Corrales <santiago.nunez@ridgerun.com>
  *
  * This code is partially based upon the TVP5150 driver
- * written by Mauro Carvalho Chehab (mchehab@infradead.org),
+ * written by Mauro Carvalho Chehab <mchehab@kernel.org>,
  * the TVP514x driver written by Vaibhav Hiremath <hvaibhav@ti.com>
  * and the TVP7002 driver in the TI LSP 2.10.00.14
  *
diff --git a/drivers/media/media-devnode.c b/drivers/media/media-devnode.c
index 67ac51e..6b87a72 100644
--- a/drivers/media/media-devnode.c
+++ b/drivers/media/media-devnode.c
@@ -4,7 +4,7 @@
  * Copyright (C) 2010 Nokia Corporation
  *
  * Based on drivers/media/video/v4l2_dev.c code authored by
- *	Mauro Carvalho Chehab <mchehab@infradead.org> (version 2)
+ *	Mauro Carvalho Chehab <mchehab@kernel.org> (version 2)
  *	Alan Cox, <alan@lxorguk.ukuu.org.uk> (version 1)
  *
  * Contacts: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
diff --git a/drivers/media/pci/bt8xx/bttv-audio-hook.c b/drivers/media/pci/bt8xx/bttv-audio-hook.c
index 9f1f9169..346fc7f 100644
--- a/drivers/media/pci/bt8xx/bttv-audio-hook.c
+++ b/drivers/media/pci/bt8xx/bttv-audio-hook.c
@@ -1,7 +1,7 @@
 /*
  * Handlers for board audio hooks, splitted from bttv-cards
  *
- * Copyright (c) 2006 Mauro Carvalho Chehab (mchehab@infradead.org)
+ * Copyright (c) 2006 Mauro Carvalho Chehab <mchehab@kernel.org>
  * This code is placed under the terms of the GNU General Public License
  */
 
diff --git a/drivers/media/pci/bt8xx/bttv-audio-hook.h b/drivers/media/pci/bt8xx/bttv-audio-hook.h
index 159d07a..be16a53 100644
--- a/drivers/media/pci/bt8xx/bttv-audio-hook.h
+++ b/drivers/media/pci/bt8xx/bttv-audio-hook.h
@@ -1,7 +1,7 @@
 /*
  * Handlers for board audio hooks, splitted from bttv-cards
  *
- * Copyright (c) 2006 Mauro Carvalho Chehab (mchehab@infradead.org)
+ * Copyright (c) 2006 Mauro Carvalho Chehab <mchehab@kernel.org>
  * This code is placed under the terms of the GNU General Public License
  */
 
diff --git a/drivers/media/pci/bt8xx/bttv-cards.c b/drivers/media/pci/bt8xx/bttv-cards.c
index 1902732..2616243 100644
--- a/drivers/media/pci/bt8xx/bttv-cards.c
+++ b/drivers/media/pci/bt8xx/bttv-cards.c
@@ -2447,7 +2447,7 @@ struct tvcard bttv_tvcards[] = {
 	},
 		/* ---- card 0x88---------------------------------- */
 	[BTTV_BOARD_ACORP_Y878F] = {
-		/* Mauro Carvalho Chehab <mchehab@infradead.org> */
+		/* Mauro Carvalho Chehab <mchehab@kernel.org> */
 		.name		= "Acorp Y878F",
 		.video_inputs	= 3,
 		/* .audio_inputs= 1, */
@@ -2688,7 +2688,7 @@ struct tvcard bttv_tvcards[] = {
 	},
 	[BTTV_BOARD_ENLTV_FM_2] = {
 		/* Encore TV Tuner Pro ENL TV-FM-2
-		   Mauro Carvalho Chehab <mchehab@infradead.org */
+		   Mauro Carvalho Chehab <mchehab@kernel.org> */
 		.name           = "Encore ENL TV-FM-2",
 		.video_inputs   = 3,
 		/* .audio_inputs= 1, */
diff --git a/drivers/media/pci/bt8xx/bttv-driver.c b/drivers/media/pci/bt8xx/bttv-driver.c
index 707f57a..de3f44b 100644
--- a/drivers/media/pci/bt8xx/bttv-driver.c
+++ b/drivers/media/pci/bt8xx/bttv-driver.c
@@ -13,7 +13,7 @@
     (c) 2005-2006 Nickolay V. Shmyrev <nshmyrev@yandex.ru>
 
     Fixes to be fully V4L2 compliant by
-    (c) 2006 Mauro Carvalho Chehab <mchehab@infradead.org>
+    (c) 2006 Mauro Carvalho Chehab <mchehab@kernel.org>
 
     Cropping and overscan support
     Copyright (C) 2005, 2006 Michael H. Schimek <mschimek@gmx.at>
diff --git a/drivers/media/pci/bt8xx/bttv-i2c.c b/drivers/media/pci/bt8xx/bttv-i2c.c
index eccd1e3..c76823e 100644
--- a/drivers/media/pci/bt8xx/bttv-i2c.c
+++ b/drivers/media/pci/bt8xx/bttv-i2c.c
@@ -8,7 +8,7 @@
 			   & Marcus Metzler (mocm@thp.uni-koeln.de)
     (c) 1999-2003 Gerd Knorr <kraxel@bytesex.org>
 
-    (c) 2005 Mauro Carvalho Chehab <mchehab@infradead.org>
+    (c) 2005 Mauro Carvalho Chehab <mchehab@kernel.org>
 	- Multituner support and i2c address binding
 
     This program is free software; you can redistribute it and/or modify
diff --git a/drivers/media/pci/cx23885/cx23885-input.c b/drivers/media/pci/cx23885/cx23885-input.c
index be49589..395ff9b 100644
--- a/drivers/media/pci/cx23885/cx23885-input.c
+++ b/drivers/media/pci/cx23885/cx23885-input.c
@@ -13,7 +13,7 @@
  *  Copyright (C) 2008 <srinivasa.deevi at conexant dot com>
  *  Copyright (C) 2005 Ludovico Cavedon <cavedon@sssup.it>
  *		       Markus Rechberger <mrechberger@gmail.com>
- *		       Mauro Carvalho Chehab <mchehab@infradead.org>
+ *		       Mauro Carvalho Chehab <mchehab@kernel.org>
  *		       Sascha Sommer <saschasommer@freenet.de>
  *  Copyright (C) 2004, 2005 Chris Pascoe
  *  Copyright (C) 2003, 2004 Gerd Knorr
diff --git a/drivers/media/pci/cx88/cx88-alsa.c b/drivers/media/pci/cx88/cx88-alsa.c
index ab09bb5..8a28fda 100644
--- a/drivers/media/pci/cx88/cx88-alsa.c
+++ b/drivers/media/pci/cx88/cx88-alsa.c
@@ -4,7 +4,7 @@
  *
  *    (c) 2007 Trent Piepho <xyzzy@speakeasy.org>
  *    (c) 2005,2006 Ricardo Cerqueira <v4l@cerqueira.org>
- *    (c) 2005 Mauro Carvalho Chehab <mchehab@infradead.org>
+ *    (c) 2005 Mauro Carvalho Chehab <mchehab@kernel.org>
  *    Based on a dummy cx88 module by Gerd Knorr <kraxel@bytesex.org>
  *    Based on dummy.c by Jaroslav Kysela <perex@perex.cz>
  *
@@ -103,7 +103,7 @@ MODULE_PARM_DESC(index, "Index value for cx88x capture interface(s).");
 
 MODULE_DESCRIPTION("ALSA driver module for cx2388x based TV cards");
 MODULE_AUTHOR("Ricardo Cerqueira");
-MODULE_AUTHOR("Mauro Carvalho Chehab <mchehab@infradead.org>");
+MODULE_AUTHOR("Mauro Carvalho Chehab <mchehab@kernel.org>");
 MODULE_LICENSE("GPL");
 MODULE_VERSION(CX88_VERSION);
 
diff --git a/drivers/media/pci/cx88/cx88-blackbird.c b/drivers/media/pci/cx88/cx88-blackbird.c
index 0e0952e..7a4876c 100644
--- a/drivers/media/pci/cx88/cx88-blackbird.c
+++ b/drivers/media/pci/cx88/cx88-blackbird.c
@@ -5,7 +5,7 @@
  *    (c) 2004 Jelle Foks <jelle@foks.us>
  *    (c) 2004 Gerd Knorr <kraxel@bytesex.org>
  *
- *    (c) 2005-2006 Mauro Carvalho Chehab <mchehab@infradead.org>
+ *    (c) 2005-2006 Mauro Carvalho Chehab <mchehab@kernel.org>
  *        - video_ioctl2 conversion
  *
  *  Includes parts from the ivtv driver <http://sourceforge.net/projects/ivtv/>
diff --git a/drivers/media/pci/cx88/cx88-core.c b/drivers/media/pci/cx88/cx88-core.c
index 8bfa5b7..60988e9 100644
--- a/drivers/media/pci/cx88/cx88-core.c
+++ b/drivers/media/pci/cx88/cx88-core.c
@@ -4,7 +4,7 @@
  *
  * (c) 2003 Gerd Knorr <kraxel@bytesex.org> [SuSE Labs]
  *
- * (c) 2005-2006 Mauro Carvalho Chehab <mchehab@infradead.org>
+ * (c) 2005-2006 Mauro Carvalho Chehab <mchehab@kernel.org>
  *     - Multituner support
  *     - video_ioctl2 conversion
  *     - PAL/M fixes
diff --git a/drivers/media/pci/cx88/cx88-i2c.c b/drivers/media/pci/cx88/cx88-i2c.c
index f769277..99f88a0 100644
--- a/drivers/media/pci/cx88/cx88-i2c.c
+++ b/drivers/media/pci/cx88/cx88-i2c.c
@@ -8,7 +8,7 @@
  * (c) 2002 Yurij Sysoev <yurij@naturesoft.net>
  * (c) 1999-2003 Gerd Knorr <kraxel@bytesex.org>
  *
- * (c) 2005 Mauro Carvalho Chehab <mchehab@infradead.org>
+ * (c) 2005 Mauro Carvalho Chehab <mchehab@kernel.org>
  *	- Multituner support and i2c address binding
  *
  * This program is free software; you can redistribute it and/or modify
diff --git a/drivers/media/pci/cx88/cx88-video.c b/drivers/media/pci/cx88/cx88-video.c
index 9be682c..7b113ba 100644
--- a/drivers/media/pci/cx88/cx88-video.c
+++ b/drivers/media/pci/cx88/cx88-video.c
@@ -5,7 +5,7 @@
  *
  * (c) 2003-04 Gerd Knorr <kraxel@bytesex.org> [SuSE Labs]
  *
- * (c) 2005-2006 Mauro Carvalho Chehab <mchehab@infradead.org>
+ * (c) 2005-2006 Mauro Carvalho Chehab <mchehab@kernel.org>
  *	- Multituner support
  *	- video_ioctl2 conversion
  *	- PAL/M fixes
diff --git a/drivers/media/radio/radio-aimslab.c b/drivers/media/radio/radio-aimslab.c
index 5ef635e..4c52ac6 100644
--- a/drivers/media/radio/radio-aimslab.c
+++ b/drivers/media/radio/radio-aimslab.c
@@ -4,7 +4,7 @@
  * Copyright 1997 M. Kirkwood
  *
  * Converted to the radio-isa framework by Hans Verkuil <hans.verkuil@cisco.com>
- * Converted to V4L2 API by Mauro Carvalho Chehab <mchehab@infradead.org>
+ * Converted to V4L2 API by Mauro Carvalho Chehab <mchehab@kernel.org>
  * Converted to new API by Alan Cox <alan@lxorguk.ukuu.org.uk>
  * Various bugfixes and enhancements by Russell Kroll <rkroll@exploits.org>
  *
diff --git a/drivers/media/radio/radio-aztech.c b/drivers/media/radio/radio-aztech.c
index 9e12c60..840b7d6 100644
--- a/drivers/media/radio/radio-aztech.c
+++ b/drivers/media/radio/radio-aztech.c
@@ -2,7 +2,7 @@
  * radio-aztech.c - Aztech radio card driver
  *
  * Converted to the radio-isa framework by Hans Verkuil <hans.verkuil@xs4all.nl>
- * Converted to V4L2 API by Mauro Carvalho Chehab <mchehab@infradead.org>
+ * Converted to V4L2 API by Mauro Carvalho Chehab <mchehab@kernel.org>
  * Adapted to support the Video for Linux API by
  * Russell Kroll <rkroll@exploits.org>.  Based on original tuner code by:
  *
diff --git a/drivers/media/radio/radio-gemtek.c b/drivers/media/radio/radio-gemtek.c
index 3ff4c4e..f051f86 100644
--- a/drivers/media/radio/radio-gemtek.c
+++ b/drivers/media/radio/radio-gemtek.c
@@ -15,7 +15,7 @@
  *    Various bugfixes and enhancements by Russell Kroll <rkroll@exploits.org>
  *
  * Converted to the radio-isa framework by Hans Verkuil <hans.verkuil@cisco.com>
- * Converted to V4L2 API by Mauro Carvalho Chehab <mchehab@infradead.org>
+ * Converted to V4L2 API by Mauro Carvalho Chehab <mchehab@kernel.org>
  *
  * Note: this card seems to swap the left and right audio channels!
  *
diff --git a/drivers/media/radio/radio-maxiradio.c b/drivers/media/radio/radio-maxiradio.c
index 95f06f3..e4e7587 100644
--- a/drivers/media/radio/radio-maxiradio.c
+++ b/drivers/media/radio/radio-maxiradio.c
@@ -27,7 +27,7 @@
  * BUGS:
  *   - card unmutes if you change frequency
  *
- * (c) 2006, 2007 by Mauro Carvalho Chehab <mchehab@infradead.org>:
+ * (c) 2006, 2007 by Mauro Carvalho Chehab <mchehab@kernel.org>:
  *	- Conversion to V4L2 API
  *      - Uses video_ioctl2 for parsing and to add debug support
  */
diff --git a/drivers/media/radio/radio-rtrack2.c b/drivers/media/radio/radio-rtrack2.c
index abeaedd..5a1470e 100644
--- a/drivers/media/radio/radio-rtrack2.c
+++ b/drivers/media/radio/radio-rtrack2.c
@@ -7,7 +7,7 @@
  * Various bugfixes and enhancements by Russell Kroll <rkroll@exploits.org>
  *
  * Converted to the radio-isa framework by Hans Verkuil <hans.verkuil@cisco.com>
- * Converted to V4L2 API by Mauro Carvalho Chehab <mchehab@infradead.org>
+ * Converted to V4L2 API by Mauro Carvalho Chehab <mchehab@kernel.org>
  *
  * Fully tested with actual hardware and the v4l2-compliance tool.
  */
diff --git a/drivers/media/radio/radio-sf16fmi.c b/drivers/media/radio/radio-sf16fmi.c
index fc4e63d..4f9b97e 100644
--- a/drivers/media/radio/radio-sf16fmi.c
+++ b/drivers/media/radio/radio-sf16fmi.c
@@ -13,7 +13,7 @@
  *  No volume control - only mute/unmute - you have to use line volume
  *  control on SB-part of SF16-FMI/SF16-FMP/SF16-FMD
  *
- * Converted to V4L2 API by Mauro Carvalho Chehab <mchehab@infradead.org>
+ * Converted to V4L2 API by Mauro Carvalho Chehab <mchehab@kernel.org>
  */
 
 #include <linux/kernel.h>	/* __setup			*/
diff --git a/drivers/media/radio/radio-terratec.c b/drivers/media/radio/radio-terratec.c
index 4f116ea..1af8f29 100644
--- a/drivers/media/radio/radio-terratec.c
+++ b/drivers/media/radio/radio-terratec.c
@@ -17,7 +17,7 @@
  *  Volume Control is done digitally
  *
  * Converted to the radio-isa framework by Hans Verkuil <hans.verkuil@cisco.com>
- * Converted to V4L2 API by Mauro Carvalho Chehab <mchehab@infradead.org>
+ * Converted to V4L2 API by Mauro Carvalho Chehab <mchehab@kernel.org>
  */
 
 #include <linux/module.h>	/* Modules			*/
diff --git a/drivers/media/radio/radio-trust.c b/drivers/media/radio/radio-trust.c
index 26a8c60..a4bad32 100644
--- a/drivers/media/radio/radio-trust.c
+++ b/drivers/media/radio/radio-trust.c
@@ -12,7 +12,7 @@
  * Scott McGrath    (smcgrath@twilight.vtc.vsc.edu)
  * William McGrath  (wmcgrath@twilight.vtc.vsc.edu)
  *
- * Converted to V4L2 API by Mauro Carvalho Chehab <mchehab@infradead.org>
+ * Converted to V4L2 API by Mauro Carvalho Chehab <mchehab@kernel.org>
  */
 
 #include <stdarg.h>
diff --git a/drivers/media/radio/radio-typhoon.c b/drivers/media/radio/radio-typhoon.c
index eb72a4d..d0d67ad 100644
--- a/drivers/media/radio/radio-typhoon.c
+++ b/drivers/media/radio/radio-typhoon.c
@@ -25,7 +25,7 @@
  * The frequency change is necessary since the card never seems to be
  * completely silent.
  *
- * Converted to V4L2 API by Mauro Carvalho Chehab <mchehab@infradead.org>
+ * Converted to V4L2 API by Mauro Carvalho Chehab <mchehab@kernel.org>
  */
 
 #include <linux/module.h>	/* Modules                        */
diff --git a/drivers/media/radio/radio-zoltrix.c b/drivers/media/radio/radio-zoltrix.c
index 026e88e..6007cd0 100644
--- a/drivers/media/radio/radio-zoltrix.c
+++ b/drivers/media/radio/radio-zoltrix.c
@@ -27,7 +27,7 @@
  * 2002-07-15 - Fix Stereo typo
  *
  * 2006-07-24 - Converted to V4L2 API
- *		by Mauro Carvalho Chehab <mchehab@infradead.org>
+ *		by Mauro Carvalho Chehab <mchehab@kernel.org>
  *
  * Converted to the radio-isa framework by Hans Verkuil <hans.verkuil@cisco.com>
  *
diff --git a/drivers/media/rc/keymaps/rc-avermedia-m135a.c b/drivers/media/rc/keymaps/rc-avermedia-m135a.c
index f6977df..d275d98 100644
--- a/drivers/media/rc/keymaps/rc-avermedia-m135a.c
+++ b/drivers/media/rc/keymaps/rc-avermedia-m135a.c
@@ -12,7 +12,7 @@
  *
  * On Avermedia M135A with IR model RM-JX, the same codes exist on both
  * Positivo (BR) and original IR, initial version and remote control codes
- * added by Mauro Carvalho Chehab <mchehab@infradead.org>
+ * added by Mauro Carvalho Chehab <mchehab@kernel.org>
  *
  * Positivo also ships Avermedia M135A with model RM-K6, extra control
  * codes added by Herton Ronaldo Krzesinski <herton@mandriva.com.br>
diff --git a/drivers/media/rc/keymaps/rc-encore-enltv-fm53.c b/drivers/media/rc/keymaps/rc-encore-enltv-fm53.c
index e4e78c1..057c13b 100644
--- a/drivers/media/rc/keymaps/rc-encore-enltv-fm53.c
+++ b/drivers/media/rc/keymaps/rc-encore-enltv-fm53.c
@@ -9,7 +9,7 @@
 #include <linux/module.h>
 
 /* Encore ENLTV-FM v5.3
-   Mauro Carvalho Chehab <mchehab@infradead.org>
+   Mauro Carvalho Chehab <mchehab@kernel.org>
  */
 
 static struct rc_map_table encore_enltv_fm53[] = {
diff --git a/drivers/media/rc/keymaps/rc-encore-enltv2.c b/drivers/media/rc/keymaps/rc-encore-enltv2.c
index c3d4437..cd05559 100644
--- a/drivers/media/rc/keymaps/rc-encore-enltv2.c
+++ b/drivers/media/rc/keymaps/rc-encore-enltv2.c
@@ -9,7 +9,7 @@
 #include <linux/module.h>
 
 /* Encore ENLTV2-FM  - silver plastic - "Wand Media" written at the botton
-    Mauro Carvalho Chehab <mchehab@infradead.org> */
+    Mauro Carvalho Chehab <mchehab@kernel.org> */
 
 static struct rc_map_table encore_enltv2[] = {
 	{ 0x4c, KEY_POWER2 },
diff --git a/drivers/media/rc/keymaps/rc-kaiomy.c b/drivers/media/rc/keymaps/rc-kaiomy.c
index f0f88df..a000513 100644
--- a/drivers/media/rc/keymaps/rc-kaiomy.c
+++ b/drivers/media/rc/keymaps/rc-kaiomy.c
@@ -9,7 +9,7 @@
 #include <linux/module.h>
 
 /* Kaiomy TVnPC U2
-   Mauro Carvalho Chehab <mchehab@infradead.org>
+   Mauro Carvalho Chehab <mchehab@kernel.org>
  */
 
 static struct rc_map_table kaiomy[] = {
diff --git a/drivers/media/rc/keymaps/rc-kworld-plus-tv-analog.c b/drivers/media/rc/keymaps/rc-kworld-plus-tv-analog.c
index 453e0437..db5edde 100644
--- a/drivers/media/rc/keymaps/rc-kworld-plus-tv-analog.c
+++ b/drivers/media/rc/keymaps/rc-kworld-plus-tv-analog.c
@@ -9,7 +9,7 @@
 #include <linux/module.h>
 
 /* Kworld Plus TV Analog Lite PCI IR
-   Mauro Carvalho Chehab <mchehab@infradead.org>
+   Mauro Carvalho Chehab <mchehab@kernel.org>
  */
 
 static struct rc_map_table kworld_plus_tv_analog[] = {
diff --git a/drivers/media/rc/keymaps/rc-pixelview-new.c b/drivers/media/rc/keymaps/rc-pixelview-new.c
index 791130f..e4e34f2 100644
--- a/drivers/media/rc/keymaps/rc-pixelview-new.c
+++ b/drivers/media/rc/keymaps/rc-pixelview-new.c
@@ -9,7 +9,7 @@
 #include <linux/module.h>
 
 /*
-   Mauro Carvalho Chehab <mchehab@infradead.org>
+   Mauro Carvalho Chehab <mchehab@kernel.org>
    present on PV MPEG 8000GT
  */
 
diff --git a/drivers/media/tuners/tea5761.c b/drivers/media/tuners/tea5761.c
index 88b3e80..d78a2bd 100644
--- a/drivers/media/tuners/tea5761.c
+++ b/drivers/media/tuners/tea5761.c
@@ -2,7 +2,7 @@
 // For Philips TEA5761 FM Chip
 // I2C address is always 0x20 (0x10 at 7-bit mode).
 //
-// Copyright (c) 2005-2007 Mauro Carvalho Chehab (mchehab@infradead.org)
+// Copyright (c) 2005-2007 Mauro Carvalho Chehab <mchehab@kernel.org>
 
 #include <linux/i2c.h>
 #include <linux/slab.h>
@@ -337,5 +337,5 @@ EXPORT_SYMBOL_GPL(tea5761_attach);
 EXPORT_SYMBOL_GPL(tea5761_autodetection);
 
 MODULE_DESCRIPTION("Philips TEA5761 FM tuner driver");
-MODULE_AUTHOR("Mauro Carvalho Chehab <mchehab@infradead.org>");
+MODULE_AUTHOR("Mauro Carvalho Chehab <mchehab@kernel.org>");
 MODULE_LICENSE("GPL v2");
diff --git a/drivers/media/tuners/tea5767.c b/drivers/media/tuners/tea5767.c
index 2b2c064..016d0d5e 100644
--- a/drivers/media/tuners/tea5767.c
+++ b/drivers/media/tuners/tea5767.c
@@ -2,7 +2,7 @@
 // For Philips TEA5767 FM Chip used on some TV Cards like Prolink Pixelview
 // I2C address is always 0xC0.
 //
-// Copyright (c) 2005 Mauro Carvalho Chehab (mchehab@infradead.org)
+// Copyright (c) 2005 Mauro Carvalho Chehab <mchehab@kernel.org>
 //
 // tea5767 autodetection thanks to Torsten Seeboth and Atsushi Nakagawa
 // from their contributions on DScaler.
@@ -469,5 +469,5 @@ EXPORT_SYMBOL_GPL(tea5767_attach);
 EXPORT_SYMBOL_GPL(tea5767_autodetection);
 
 MODULE_DESCRIPTION("Philips TEA5767 FM tuner driver");
-MODULE_AUTHOR("Mauro Carvalho Chehab <mchehab@infradead.org>");
+MODULE_AUTHOR("Mauro Carvalho Chehab <mchehab@kernel.org>");
 MODULE_LICENSE("GPL v2");
diff --git a/drivers/media/tuners/tuner-xc2028-types.h b/drivers/media/tuners/tuner-xc2028-types.h
index bb0437c..50d017a 100644
--- a/drivers/media/tuners/tuner-xc2028-types.h
+++ b/drivers/media/tuners/tuner-xc2028-types.h
@@ -5,7 +5,7 @@
  * This file includes internal tipes to be used inside tuner-xc2028.
  * Shouldn't be included outside tuner-xc2028
  *
- * Copyright (c) 2007-2008 Mauro Carvalho Chehab (mchehab@infradead.org)
+ * Copyright (c) 2007-2008 Mauro Carvalho Chehab <mchehab@kernel.org>
  */
 
 /* xc3028 firmware types */
diff --git a/drivers/media/tuners/tuner-xc2028.c b/drivers/media/tuners/tuner-xc2028.c
index fca85e0..84744e1 100644
--- a/drivers/media/tuners/tuner-xc2028.c
+++ b/drivers/media/tuners/tuner-xc2028.c
@@ -1,7 +1,7 @@
 // SPDX-License-Identifier: GPL-2.0
 // tuner-xc2028
 //
-// Copyright (c) 2007-2008 Mauro Carvalho Chehab (mchehab@infradead.org)
+// Copyright (c) 2007-2008 Mauro Carvalho Chehab <mchehab@kernel.org>
 //
 // Copyright (c) 2007 Michel Ludwig (michel.ludwig@gmail.com)
 //       - frontend interface
@@ -1518,7 +1518,7 @@ EXPORT_SYMBOL(xc2028_attach);
 
 MODULE_DESCRIPTION("Xceive xc2028/xc3028 tuner driver");
 MODULE_AUTHOR("Michel Ludwig <michel.ludwig@gmail.com>");
-MODULE_AUTHOR("Mauro Carvalho Chehab <mchehab@infradead.org>");
+MODULE_AUTHOR("Mauro Carvalho Chehab <mchehab@kernel.org>");
 MODULE_LICENSE("GPL v2");
 MODULE_FIRMWARE(XC2028_DEFAULT_FIRMWARE);
 MODULE_FIRMWARE(XC3028L_DEFAULT_FIRMWARE);
diff --git a/drivers/media/tuners/tuner-xc2028.h b/drivers/media/tuners/tuner-xc2028.h
index 03fd6d4..7b58bc0 100644
--- a/drivers/media/tuners/tuner-xc2028.h
+++ b/drivers/media/tuners/tuner-xc2028.h
@@ -2,7 +2,7 @@
  * SPDX-License-Identifier: GPL-2.0
  * tuner-xc2028
  *
- * Copyright (c) 2007-2008 Mauro Carvalho Chehab (mchehab@infradead.org)
+ * Copyright (c) 2007-2008 Mauro Carvalho Chehab <mchehab@kernel.org>
  */
 
 #ifndef __TUNER_XC2028_H__
diff --git a/drivers/media/usb/em28xx/em28xx-camera.c b/drivers/media/usb/em28xx/em28xx-camera.c
index 3c2694a..d1e66b5 100644
--- a/drivers/media/usb/em28xx/em28xx-camera.c
+++ b/drivers/media/usb/em28xx/em28xx-camera.c
@@ -2,7 +2,7 @@
 //
 // em28xx-camera.c - driver for Empia EM25xx/27xx/28xx USB video capture devices
 //
-// Copyright (C) 2009 Mauro Carvalho Chehab <mchehab@infradead.org>
+// Copyright (C) 2009 Mauro Carvalho Chehab <mchehab@kernel.org>
 // Copyright (C) 2013 Frank Schäfer <fschaefer.oss@googlemail.com>
 //
 // This program is free software; you can redistribute it and/or modify
diff --git a/drivers/media/usb/em28xx/em28xx-cards.c b/drivers/media/usb/em28xx/em28xx-cards.c
index 6e0e67d..7c3203d 100644
--- a/drivers/media/usb/em28xx/em28xx-cards.c
+++ b/drivers/media/usb/em28xx/em28xx-cards.c
@@ -5,7 +5,7 @@
 //
 // Copyright (C) 2005 Ludovico Cavedon <cavedon@sssup.it>
 //		      Markus Rechberger <mrechberger@gmail.com>
-//		      Mauro Carvalho Chehab <mchehab@infradead.org>
+//		      Mauro Carvalho Chehab <mchehab@kernel.org>
 //		      Sascha Sommer <saschasommer@freenet.de>
 // Copyright (C) 2012 Frank Schäfer <fschaefer.oss@googlemail.com>
 //
diff --git a/drivers/media/usb/em28xx/em28xx-core.c b/drivers/media/usb/em28xx/em28xx-core.c
index 36d341f..f289953 100644
--- a/drivers/media/usb/em28xx/em28xx-core.c
+++ b/drivers/media/usb/em28xx/em28xx-core.c
@@ -4,7 +4,7 @@
 //
 // Copyright (C) 2005 Ludovico Cavedon <cavedon@sssup.it>
 //		      Markus Rechberger <mrechberger@gmail.com>
-//		      Mauro Carvalho Chehab <mchehab@infradead.org>
+//		      Mauro Carvalho Chehab <mchehab@kernel.org>
 //		      Sascha Sommer <saschasommer@freenet.de>
 // Copyright (C) 2012 Frank Schäfer <fschaefer.oss@googlemail.com>
 //
@@ -32,7 +32,7 @@
 
 #define DRIVER_AUTHOR "Ludovico Cavedon <cavedon@sssup.it>, " \
 		      "Markus Rechberger <mrechberger@gmail.com>, " \
-		      "Mauro Carvalho Chehab <mchehab@infradead.org>, " \
+		      "Mauro Carvalho Chehab <mchehab@kernel.org>, " \
 		      "Sascha Sommer <saschasommer@freenet.de>"
 
 MODULE_AUTHOR(DRIVER_AUTHOR);
diff --git a/drivers/media/usb/em28xx/em28xx-dvb.c b/drivers/media/usb/em28xx/em28xx-dvb.c
index a54cb8d..3f493e0 100644
--- a/drivers/media/usb/em28xx/em28xx-dvb.c
+++ b/drivers/media/usb/em28xx/em28xx-dvb.c
@@ -2,7 +2,7 @@
 //
 // DVB device driver for em28xx
 //
-// (c) 2008-2011 Mauro Carvalho Chehab <mchehab@infradead.org>
+// (c) 2008-2011 Mauro Carvalho Chehab <mchehab@kernel.org>
 //
 // (c) 2008 Devin Heitmueller <devin.heitmueller@gmail.com>
 //	- Fixes for the driver to properly work with HVR-950
@@ -63,7 +63,7 @@
 #include "tc90522.h"
 #include "qm1d1c0042.h"
 
-MODULE_AUTHOR("Mauro Carvalho Chehab <mchehab@infradead.org>");
+MODULE_AUTHOR("Mauro Carvalho Chehab <mchehab@kernel.org>");
 MODULE_LICENSE("GPL v2");
 MODULE_DESCRIPTION(DRIVER_DESC " - digital TV interface");
 MODULE_VERSION(EM28XX_VERSION);
diff --git a/drivers/media/usb/em28xx/em28xx-i2c.c b/drivers/media/usb/em28xx/em28xx-i2c.c
index 9151bcc..6458682 100644
--- a/drivers/media/usb/em28xx/em28xx-i2c.c
+++ b/drivers/media/usb/em28xx/em28xx-i2c.c
@@ -4,7 +4,7 @@
 //
 // Copyright (C) 2005 Ludovico Cavedon <cavedon@sssup.it>
 //		      Markus Rechberger <mrechberger@gmail.com>
-//		      Mauro Carvalho Chehab <mchehab@infradead.org>
+//		      Mauro Carvalho Chehab <mchehab@kernel.org>
 //		      Sascha Sommer <saschasommer@freenet.de>
 // Copyright (C) 2013 Frank Schäfer <fschaefer.oss@googlemail.com>
 //
diff --git a/drivers/media/usb/em28xx/em28xx-input.c b/drivers/media/usb/em28xx/em28xx-input.c
index 2dc1be0..f84a120 100644
--- a/drivers/media/usb/em28xx/em28xx-input.c
+++ b/drivers/media/usb/em28xx/em28xx-input.c
@@ -4,7 +4,7 @@
 //
 // Copyright (C) 2005 Ludovico Cavedon <cavedon@sssup.it>
 //		      Markus Rechberger <mrechberger@gmail.com>
-//		      Mauro Carvalho Chehab <mchehab@infradead.org>
+//		      Mauro Carvalho Chehab <mchehab@kernel.org>
 //		      Sascha Sommer <saschasommer@freenet.de>
 //
 // This program is free software; you can redistribute it and/or modify
diff --git a/drivers/media/usb/em28xx/em28xx-video.c b/drivers/media/usb/em28xx/em28xx-video.c
index d70ee13..68571bf 100644
--- a/drivers/media/usb/em28xx/em28xx-video.c
+++ b/drivers/media/usb/em28xx/em28xx-video.c
@@ -5,7 +5,7 @@
 //
 // Copyright (C) 2005 Ludovico Cavedon <cavedon@sssup.it>
 //		      Markus Rechberger <mrechberger@gmail.com>
-//		      Mauro Carvalho Chehab <mchehab@infradead.org>
+//		      Mauro Carvalho Chehab <mchehab@kernel.org>
 //		      Sascha Sommer <saschasommer@freenet.de>
 // Copyright (C) 2012 Frank Schäfer <fschaefer.oss@googlemail.com>
 //
@@ -44,7 +44,7 @@
 
 #define DRIVER_AUTHOR "Ludovico Cavedon <cavedon@sssup.it>, " \
 		      "Markus Rechberger <mrechberger@gmail.com>, " \
-		      "Mauro Carvalho Chehab <mchehab@infradead.org>, " \
+		      "Mauro Carvalho Chehab <mchehab@kernel.org>, " \
 		      "Sascha Sommer <saschasommer@freenet.de>"
 
 static unsigned int isoc_debug;
diff --git a/drivers/media/usb/em28xx/em28xx.h b/drivers/media/usb/em28xx/em28xx.h
index 63c7c61..b0378e7 100644
--- a/drivers/media/usb/em28xx/em28xx.h
+++ b/drivers/media/usb/em28xx/em28xx.h
@@ -4,7 +4,7 @@
  *
  * Copyright (C) 2005 Markus Rechberger <mrechberger@gmail.com>
  *		      Ludovico Cavedon <cavedon@sssup.it>
- *		      Mauro Carvalho Chehab <mchehab@infradead.org>
+ *		      Mauro Carvalho Chehab <mchehab@kernel.org>
  * Copyright (C) 2012 Frank Schäfer <fschaefer.oss@googlemail.com>
  *
  * Based on the em2800 driver from Sascha Sommer <saschasommer@freenet.de>
diff --git a/drivers/media/usb/gspca/zc3xx-reg.h b/drivers/media/usb/gspca/zc3xx-reg.h
index a1bd94e..71fda38 100644
--- a/drivers/media/usb/gspca/zc3xx-reg.h
+++ b/drivers/media/usb/gspca/zc3xx-reg.h
@@ -1,7 +1,7 @@
 /*
  * zc030x registers
  *
- * Copyright (c) 2008 Mauro Carvalho Chehab <mchehab@infradead.org>
+ * Copyright (c) 2008 Mauro Carvalho Chehab <mchehab@kernel.org>
  *
  * The register aliases used here came from this driver:
  *	http://zc0302.sourceforge.net/zc0302.php
diff --git a/drivers/media/usb/tm6000/tm6000-cards.c b/drivers/media/usb/tm6000/tm6000-cards.c
index 70939e9..23df50a 100644
--- a/drivers/media/usb/tm6000/tm6000-cards.c
+++ b/drivers/media/usb/tm6000/tm6000-cards.c
@@ -1,7 +1,7 @@
 // SPDX-License-Identifier: GPL-2.0
 // tm6000-cards.c - driver for TM5600/TM6000/TM6010 USB video capture devices
 //
-// Copyright (c) 2006-2007 Mauro Carvalho Chehab <mchehab@infradead.org>
+// Copyright (c) 2006-2007 Mauro Carvalho Chehab <mchehab@kernel.org>
 
 #include <linux/init.h>
 #include <linux/module.h>
diff --git a/drivers/media/usb/tm6000/tm6000-core.c b/drivers/media/usb/tm6000/tm6000-core.c
index 23a1332..d3229aa 100644
--- a/drivers/media/usb/tm6000/tm6000-core.c
+++ b/drivers/media/usb/tm6000/tm6000-core.c
@@ -1,7 +1,7 @@
 // SPDX-License-Identifier: GPL-2.0
 // tm6000-core.c - driver for TM5600/TM6000/TM6010 USB video capture devices
 //
-// Copyright (c) 2006-2007 Mauro Carvalho Chehab <mchehab@infradead.org>
+// Copyright (c) 2006-2007 Mauro Carvalho Chehab <mchehab@kernel.org>
 //
 // Copyright (c) 2007 Michel Ludwig <michel.ludwig@gmail.com>
 //     - DVB-T support
diff --git a/drivers/media/usb/tm6000/tm6000-i2c.c b/drivers/media/usb/tm6000/tm6000-i2c.c
index c9a62bb..659b63f 100644
--- a/drivers/media/usb/tm6000/tm6000-i2c.c
+++ b/drivers/media/usb/tm6000/tm6000-i2c.c
@@ -1,7 +1,7 @@
 // SPDX-License-Identifier: GPL-2.0
 // tm6000-i2c.c - driver for TM5600/TM6000/TM6010 USB video capture devices
 //
-// Copyright (c) 2006-2007 Mauro Carvalho Chehab <mchehab@infradead.org>
+// Copyright (c) 2006-2007 Mauro Carvalho Chehab <mchehab@kernel.org>
 //
 // Copyright (c) 2007 Michel Ludwig <michel.ludwig@gmail.com>
 //	- Fix SMBus Read Byte command
diff --git a/drivers/media/usb/tm6000/tm6000-regs.h b/drivers/media/usb/tm6000/tm6000-regs.h
index 21587fc..d104246 100644
--- a/drivers/media/usb/tm6000/tm6000-regs.h
+++ b/drivers/media/usb/tm6000/tm6000-regs.h
@@ -2,7 +2,7 @@
  * SPDX-License-Identifier: GPL-2.0
  * tm6000-regs.h - driver for TM5600/TM6000/TM6010 USB video capture devices
  *
- * Copyright (c) 2006-2007 Mauro Carvalho Chehab <mchehab@infradead.org>
+ * Copyright (c) 2006-2007 Mauro Carvalho Chehab <mchehab@kernel.org>
  */
 
 /*
diff --git a/drivers/media/usb/tm6000/tm6000-usb-isoc.h b/drivers/media/usb/tm6000/tm6000-usb-isoc.h
index 5c615b0..b275dbc 100644
--- a/drivers/media/usb/tm6000/tm6000-usb-isoc.h
+++ b/drivers/media/usb/tm6000/tm6000-usb-isoc.h
@@ -2,7 +2,7 @@
  * SPDX-License-Identifier: GPL-2.0
  * tm6000-buf.c - driver for TM5600/TM6000/TM6010 USB video capture devices
  *
- * Copyright (c) 2006-2007 Mauro Carvalho Chehab <mchehab@infradead.org>
+ * Copyright (c) 2006-2007 Mauro Carvalho Chehab <mchehab@kernel.org>
  */
 
 #include <linux/videodev2.h>
diff --git a/drivers/media/usb/tm6000/tm6000-video.c b/drivers/media/usb/tm6000/tm6000-video.c
index b2399d4..aa85fe31 100644
--- a/drivers/media/usb/tm6000/tm6000-video.c
+++ b/drivers/media/usb/tm6000/tm6000-video.c
@@ -1,7 +1,7 @@
 // SPDX-License-Identifier: GPL-2.0
 // tm6000-video.c - driver for TM5600/TM6000/TM6010 USB video capture devices
 //
-// Copyright (c) 2006-2007 Mauro Carvalho Chehab <mchehab@infradead.org>
+// Copyright (c) 2006-2007 Mauro Carvalho Chehab <mchehab@kernel.org>
 //
 // Copyright (c) 2007 Michel Ludwig <michel.ludwig@gmail.com>
 //	- Fixed module load/unload
diff --git a/drivers/media/usb/tm6000/tm6000.h b/drivers/media/usb/tm6000/tm6000.h
index e1e4577..0864ed7 100644
--- a/drivers/media/usb/tm6000/tm6000.h
+++ b/drivers/media/usb/tm6000/tm6000.h
@@ -2,7 +2,7 @@
  * SPDX-License-Identifier: GPL-2.0
  * tm6000.h - driver for TM5600/TM6000/TM6010 USB video capture devices
  *
- * Copyright (c) 2006-2007 Mauro Carvalho Chehab <mchehab@infradead.org>
+ * Copyright (c) 2006-2007 Mauro Carvalho Chehab <mchehab@kernel.org>
  *
  * Copyright (c) 2007 Michel Ludwig <michel.ludwig@gmail.com>
  *	- DVB-T support
diff --git a/drivers/media/v4l2-core/v4l2-dev.c b/drivers/media/v4l2-core/v4l2-dev.c
index 1d0b220..c080dcc 100644
--- a/drivers/media/v4l2-core/v4l2-dev.c
+++ b/drivers/media/v4l2-core/v4l2-dev.c
@@ -10,7 +10,7 @@
  *	2 of the License, or (at your option) any later version.
  *
  * Authors:	Alan Cox, <alan@lxorguk.ukuu.org.uk> (version 1)
- *              Mauro Carvalho Chehab <mchehab@infradead.org> (version 2)
+ *              Mauro Carvalho Chehab <mchehab@kernel.org> (version 2)
  *
  * Fixes:	20000516  Claudio Matsuoka <claudio@conectiva.com>
  *		- Added procfs support
@@ -1072,7 +1072,7 @@ static void __exit videodev_exit(void)
 subsys_initcall(videodev_init);
 module_exit(videodev_exit)
 
-MODULE_AUTHOR("Alan Cox, Mauro Carvalho Chehab <mchehab@infradead.org>");
+MODULE_AUTHOR("Alan Cox, Mauro Carvalho Chehab <mchehab@kernel.org>");
 MODULE_DESCRIPTION("Device registrar for Video4Linux drivers v2");
 MODULE_LICENSE("GPL");
 MODULE_ALIAS_CHARDEV_MAJOR(VIDEO_MAJOR);
diff --git a/drivers/media/v4l2-core/v4l2-ioctl.c b/drivers/media/v4l2-core/v4l2-ioctl.c
index f48c505..de5d96d 100644
--- a/drivers/media/v4l2-core/v4l2-ioctl.c
+++ b/drivers/media/v4l2-core/v4l2-ioctl.c
@@ -9,7 +9,7 @@
  * 2 of the License, or (at your option) any later version.
  *
  * Authors:	Alan Cox, <alan@lxorguk.ukuu.org.uk> (version 1)
- *              Mauro Carvalho Chehab <mchehab@infradead.org> (version 2)
+ *              Mauro Carvalho Chehab <mchehab@kernel.org> (version 2)
  */
 
 #include <linux/mm.h>
diff --git a/drivers/media/v4l2-core/videobuf-core.c b/drivers/media/v4l2-core/videobuf-core.c
index 2b39818..7491b33 100644
--- a/drivers/media/v4l2-core/videobuf-core.c
+++ b/drivers/media/v4l2-core/videobuf-core.c
@@ -1,11 +1,11 @@
 /*
  * generic helper functions for handling video4linux capture buffers
  *
- * (c) 2007 Mauro Carvalho Chehab, <mchehab@infradead.org>
+ * (c) 2007 Mauro Carvalho Chehab, <mchehab@kernel.org>
  *
  * Highly based on video-buf written originally by:
  * (c) 2001,02 Gerd Knorr <kraxel@bytesex.org>
- * (c) 2006 Mauro Carvalho Chehab, <mchehab@infradead.org>
+ * (c) 2006 Mauro Carvalho Chehab, <mchehab@kernel.org>
  * (c) 2006 Ted Walther and John Sokol
  *
  * This program is free software; you can redistribute it and/or modify
@@ -38,7 +38,7 @@ static int debug;
 module_param(debug, int, 0644);
 
 MODULE_DESCRIPTION("helper module to manage video4linux buffers");
-MODULE_AUTHOR("Mauro Carvalho Chehab <mchehab@infradead.org>");
+MODULE_AUTHOR("Mauro Carvalho Chehab <mchehab@kernel.org>");
 MODULE_LICENSE("GPL");
 
 #define dprintk(level, fmt, arg...)					\
diff --git a/drivers/media/v4l2-core/videobuf-dma-contig.c b/drivers/media/v4l2-core/videobuf-dma-contig.c
index e02353e..f461325 100644
--- a/drivers/media/v4l2-core/videobuf-dma-contig.c
+++ b/drivers/media/v4l2-core/videobuf-dma-contig.c
@@ -7,7 +7,7 @@
  * Copyright (c) 2008 Magnus Damm
  *
  * Based on videobuf-vmalloc.c,
- * (c) 2007 Mauro Carvalho Chehab, <mchehab@infradead.org>
+ * (c) 2007 Mauro Carvalho Chehab, <mchehab@kernel.org>
  *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of the GNU General Public License as published by
diff --git a/drivers/media/v4l2-core/videobuf-dma-sg.c b/drivers/media/v4l2-core/videobuf-dma-sg.c
index add2edb..7770034 100644
--- a/drivers/media/v4l2-core/videobuf-dma-sg.c
+++ b/drivers/media/v4l2-core/videobuf-dma-sg.c
@@ -6,11 +6,11 @@
  * into PAGE_SIZE chunks).  They also assume the driver does not need
  * to touch the video data.
  *
- * (c) 2007 Mauro Carvalho Chehab, <mchehab@infradead.org>
+ * (c) 2007 Mauro Carvalho Chehab, <mchehab@kernel.org>
  *
  * Highly based on video-buf written originally by:
  * (c) 2001,02 Gerd Knorr <kraxel@bytesex.org>
- * (c) 2006 Mauro Carvalho Chehab, <mchehab@infradead.org>
+ * (c) 2006 Mauro Carvalho Chehab, <mchehab@kernel.org>
  * (c) 2006 Ted Walther and John Sokol
  *
  * This program is free software; you can redistribute it and/or modify
@@ -48,7 +48,7 @@ static int debug;
 module_param(debug, int, 0644);
 
 MODULE_DESCRIPTION("helper module to manage video4linux dma sg buffers");
-MODULE_AUTHOR("Mauro Carvalho Chehab <mchehab@infradead.org>");
+MODULE_AUTHOR("Mauro Carvalho Chehab <mchehab@kernel.org>");
 MODULE_LICENSE("GPL");
 
 #define dprintk(level, fmt, arg...)					\
diff --git a/drivers/media/v4l2-core/videobuf-vmalloc.c b/drivers/media/v4l2-core/videobuf-vmalloc.c
index 2ff7fcc..45fe781 100644
--- a/drivers/media/v4l2-core/videobuf-vmalloc.c
+++ b/drivers/media/v4l2-core/videobuf-vmalloc.c
@@ -6,7 +6,7 @@
  * into PAGE_SIZE chunks).  They also assume the driver does not need
  * to touch the video data.
  *
- * (c) 2007 Mauro Carvalho Chehab, <mchehab@infradead.org>
+ * (c) 2007 Mauro Carvalho Chehab, <mchehab@kernel.org>
  *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of the GNU General Public License as published by
@@ -41,7 +41,7 @@ static int debug;
 module_param(debug, int, 0644);
 
 MODULE_DESCRIPTION("helper module to manage video4linux vmalloc buffers");
-MODULE_AUTHOR("Mauro Carvalho Chehab <mchehab@infradead.org>");
+MODULE_AUTHOR("Mauro Carvalho Chehab <mchehab@kernel.org>");
 MODULE_LICENSE("GPL");
 
 #define dprintk(level, fmt, arg...)					\
diff --git a/include/media/i2c/tvp7002.h b/include/media/i2c/tvp7002.h
index 5ee007c..cb213c1 100644
--- a/include/media/i2c/tvp7002.h
+++ b/include/media/i2c/tvp7002.h
@@ -5,7 +5,7 @@
  * Author: Santiago Nunez-Corrales <santiago.nunez@ridgerun.com>
  *
  * This code is partially based upon the TVP5150 driver
- * written by Mauro Carvalho Chehab (mchehab@infradead.org),
+ * written by Mauro Carvalho Chehab <mchehab@kernel.org>,
  * the TVP514x driver written by Vaibhav Hiremath <hvaibhav@ti.com>
  * and the TVP7002 driver in the TI LSP 2.10.00.14
  *
diff --git a/include/media/videobuf-core.h b/include/media/videobuf-core.h
index 0bda0ad..60a664f 100644
--- a/include/media/videobuf-core.h
+++ b/include/media/videobuf-core.h
@@ -1,11 +1,11 @@
 /*
  * generic helper functions for handling video4linux capture buffers
  *
- * (c) 2007 Mauro Carvalho Chehab, <mchehab@infradead.org>
+ * (c) 2007 Mauro Carvalho Chehab, <mchehab@kernel.org>
  *
  * Highly based on video-buf written originally by:
  * (c) 2001,02 Gerd Knorr <kraxel@bytesex.org>
- * (c) 2006 Mauro Carvalho Chehab, <mchehab@infradead.org>
+ * (c) 2006 Mauro Carvalho Chehab, <mchehab@kernel.org>
  * (c) 2006 Ted Walther and John Sokol
  *
  * This program is free software; you can redistribute it and/or modify
diff --git a/include/media/videobuf-dma-sg.h b/include/media/videobuf-dma-sg.h
index d8b27854..01bd142 100644
--- a/include/media/videobuf-dma-sg.h
+++ b/include/media/videobuf-dma-sg.h
@@ -6,11 +6,11 @@
  * into PAGE_SIZE chunks).  They also assume the driver does not need
  * to touch the video data.
  *
- * (c) 2007 Mauro Carvalho Chehab, <mchehab@infradead.org>
+ * (c) 2007 Mauro Carvalho Chehab, <mchehab@kernel.org>
  *
  * Highly based on video-buf written originally by:
  * (c) 2001,02 Gerd Knorr <kraxel@bytesex.org>
- * (c) 2006 Mauro Carvalho Chehab, <mchehab@infradead.org>
+ * (c) 2006 Mauro Carvalho Chehab, <mchehab@kernel.org>
  * (c) 2006 Ted Walther and John Sokol
  *
  * This program is free software; you can redistribute it and/or modify
diff --git a/include/media/videobuf-vmalloc.h b/include/media/videobuf-vmalloc.h
index 486a97e..36c6a4a 100644
--- a/include/media/videobuf-vmalloc.h
+++ b/include/media/videobuf-vmalloc.h
@@ -6,7 +6,7 @@
  * into PAGE_SIZE chunks).  They also assume the driver does not need
  * to touch the video data.
  *
- * (c) 2007 Mauro Carvalho Chehab, <mchehab@infradead.org>
+ * (c) 2007 Mauro Carvalho Chehab, <mchehab@kernel.org>
  *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of the GNU General Public License as published by
diff --git a/scripts/extract_xc3028.pl b/scripts/extract_xc3028.pl
index 61d9b25..a1c51b7 100755
--- a/scripts/extract_xc3028.pl
+++ b/scripts/extract_xc3028.pl
@@ -1,6 +1,6 @@
 #!/usr/bin/env perl
 
-# Copyright (c) Mauro Carvalho Chehab <mchehab@infradead.org>
+# Copyright (c) Mauro Carvalho Chehab <mchehab@kernel.org>
 # Released under GPLv2
 #
 # In order to use, you need to:
diff --git a/scripts/split-man.pl b/scripts/split-man.pl
index bfe16cb..c3db607 100755
--- a/scripts/split-man.pl
+++ b/scripts/split-man.pl
@@ -1,7 +1,7 @@
 #!/usr/bin/perl
 # SPDX-License-Identifier: GPL-2.0
 #
-# Author: Mauro Carvalho Chehab <mchehab@s-opensource.com>
+# Author: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
 #
 # Produce manpages from kernel-doc.
 # See Documentation/doc-guide/kernel-doc.rst for instructions
-- 
cgit v1.1


From c258f91d8ae737cd46a1201aa63655ec8b878ed5 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Thu, 3 May 2018 22:29:56 +0100
Subject: drm/i915/gtt: Tidy up duplicate branches in gen8_gmch_probe()
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Following commit f773568b6ff8 ("drm/i915: nuke the duplicated stolen
discovery"), the if-else-chain for determining the GTT size is redundant
with the !chv branches all being the same.

Reported-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
References: f773568b6ff8 ("drm/i915: nuke the duplicated stolen discovery")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180503212956.3948-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/i915_gem_gtt.c | 8 ++------
 1 file changed, 2 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index 272d6bb..c879bfd 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -3326,14 +3326,10 @@ static int gen8_gmch_probe(struct i915_ggtt *ggtt)
 		DRM_ERROR("Can't set DMA mask/consistent mask (%d)\n", err);
 
 	pci_read_config_word(pdev, SNB_GMCH_CTRL, &snb_gmch_ctl);
-
-	if (INTEL_GEN(dev_priv) >= 9) {
-		size = gen8_get_total_gtt_size(snb_gmch_ctl);
-	} else if (IS_CHERRYVIEW(dev_priv)) {
+	if (IS_CHERRYVIEW(dev_priv))
 		size = chv_get_total_gtt_size(snb_gmch_ctl);
-	} else {
+	else
 		size = gen8_get_total_gtt_size(snb_gmch_ctl);
-	}
 
 	ggtt->base.total = (size / sizeof(gen8_pte_t)) << PAGE_SHIFT;
 	ggtt->base.cleanup = gen6_gmch_remove;
-- 
cgit v1.1


From 43c8c44105e30d912746a6dbd10c59ef42c230f0 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Fri, 4 May 2018 11:11:47 +0100
Subject: drm/i915: Remove assertion of active_rings must be non-empty if
 active_requests

"An outstanding request must still be on an active ring somewhere" is
only true if we haven't just been interrupted by the shrinker in the
middle of allocating the request itself. (At the start of
i915_request_alloc() we pin the context and prepare the GT for activity,
marking it as active, and then try to allocate the request. If this
allocation invokes the shrinker, we try to reclaim some space by calling
i915_retire_requests() which may then be confused by the pre-reservation
of active_requests.)

<3>[  125.472695] i915_retire_requests:1429 GEM_BUG_ON(list_empty(&i915->gt.active_rings))
<2>[  125.472792] kernel BUG at drivers/gpu/drm/i915/i915_request.c:1429!
<4>[  125.472822] invalid opcode: 0000 [#1] PREEMPT SMP KASAN PTI
<4>[  125.498764] Modules linked in: snd_hda_codec_hdmi x86_pkg_temp_thermal intel_powerclamp coretemp crct10dif_pclmul crc32_pclmul ghash_clmulni_intel btusb btrtl btbcm btintel cdc_ether snd_hda_codec_realtek bluetooth i915 snd_hda_codec_generic usbnet r8152 mii ecdh_generic lpc_ich mei_me snd_hda_intel snd_hda_codec mei snd_hwdep snd_hda_core snd_pcm prime_numbers
<4>[  125.498923] CPU: 0 PID: 1115 Comm: gem_exec_create Tainted: G     U            4.17.0-rc3-gc49cbe0d1eb8-kasan_32+ #1
<4>[  125.498955] Hardware name: GOOGLE Peppy/Peppy, BIOS MrChromebox 02/04/2018
<4>[  125.499074] RIP: 0010:i915_retire_requests+0x3f2/0x590 [i915]
<4>[  125.499095] RSP: 0018:ffff88004e5dec40 EFLAGS: 00010282
<4>[  125.499117] RAX: 0000000000000010 RBX: ffff8800458f0000 RCX: 0000000000000000
<4>[  125.499140] RDX: dffffc0000000000 RSI: 0000000000000008 RDI: ffff880060c2f6f0
<4>[  125.499164] RBP: ffff88004e5dee30 R08: ffffed000c185ee6 R09: ffffed000c185ee6
<4>[  125.499187] R10: 0000000000000001 R11: ffffed000c185ee5 R12: ffff8800553da160
<4>[  125.499210] R13: dffffc0000000000 R14: 0000000000000000 R15: ffff8800458faed0
<4>[  125.499235] FS:  00007fe18f052980(0000) GS:ffff880065400000(0000) knlGS:0000000000000000
<4>[  125.499262] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
<4>[  125.499282] CR2: 00007f01df11efb8 CR3: 00000000518d4001 CR4: 00000000000606f0
<4>[  125.499304] Call Trace:
<4>[  125.499417]  i915_gem_shrink+0x576/0xb50 [i915]
<4>[  125.499532]  ? i915_gem_shrinker_count+0x2f0/0x2f0 [i915]
<4>[  125.499561]  ? trace_hardirqs_on_thunk+0x1a/0x1c
<4>[  125.499671]  ? i915_gem_shrinker_count+0x1d6/0x2f0 [i915]
<4>[  125.499782]  ? i915_gem_shrinker_scan+0xc4/0x320 [i915]
<4>[  125.499889]  i915_gem_shrinker_scan+0xc4/0x320 [i915]
<4>[  125.499997]  ? i915_gem_shrinker_vmap+0x3a0/0x3a0 [i915]
<4>[  125.500021]  ? do_raw_spin_unlock+0x4f/0x240
<4>[  125.500042]  ? _raw_spin_unlock+0x29/0x40
<4>[  125.500149]  ? i915_gem_shrinker_count+0x1d6/0x2f0 [i915]
<4>[  125.500177]  shrink_slab.part.18+0x23e/0x8f0
<4>[  125.500202]  ? unregister_shrinker+0x1f0/0x1f0
<4>[  125.500226]  ? mem_cgroup_iter+0x379/0xcc0
<4>[  125.500249]  shrink_node+0xa7e/0x1180
<4>[  125.500276]  ? shrink_node_memcg+0x11f0/0x11f0
<4>[  125.500297]  ? __delayacct_freepages_start+0x38/0x80
<4>[  125.500319]  ? __is_insn_slot_addr+0xe3/0x1a0
<4>[  125.500342]  ? recalibrate_cpu_khz+0x10/0x10
<4>[  125.500361]  ? ktime_get+0xb2/0x140
<4>[  125.500382]  do_try_to_free_pages+0x2d3/0xe40
<4>[  125.500407]  ? allow_direct_reclaim.part.23+0x1e0/0x1e0
<4>[  125.500429]  ? shrink_node+0x1180/0x1180
<4>[  125.500450]  ? __read_once_size_nocheck.constprop.4+0x10/0x10
<4>[  125.500476]  try_to_free_pages+0x1af/0x560
<4>[  125.500497]  ? do_try_to_free_pages+0xe40/0xe40
<4>[  125.500525]  __alloc_pages_nodemask+0xadc/0x2130
<4>[  125.500553]  ? gfp_pfmemalloc_allowed+0x150/0x150
<4>[  125.500654]  ? i915_gem_do_execbuffer+0x219d/0x32e0 [i915]
<4>[  125.500678]  ? debug_check_no_locks_freed+0x2a0/0x2a0
<4>[  125.500701]  ? __debug_object_init+0x322/0xd90
<4>[  125.500722]  ? debug_check_no_locks_freed+0x2a0/0x2a0
<4>[  125.500827]  ? i915_gem_do_execbuffer+0xdc2/0x32e0 [i915]
<4>[  125.500942]  ? i915_request_alloc+0x5b5/0x13f0 [i915]
<4>[  125.500964]  ? page_frag_free+0x170/0x170
<4>[  125.500984]  ? debug_check_no_locks_freed+0x2a0/0x2a0
<4>[  125.501008]  new_slab+0x21d/0x5c0
<4>[  125.501029]  ___slab_alloc.constprop.35+0x322/0x3e0
<4>[  125.501052]  ? reservation_object_reserve_shared+0x10b/0x250
<4>[  125.501074]  ? __ww_mutex_lock.constprop.3+0x1104/0x2cf0
<4>[  125.501097]  ? _raw_spin_unlock_irqrestore+0x39/0x60
<4>[  125.501120]  ? fs_reclaim_acquire+0x10/0x10
<4>[  125.501138]  ? lock_acquire+0x138/0x3c0
<4>[  125.501156]  ? lock_acquire+0x3c0/0x3c0
<4>[  125.501176]  ? reservation_object_reserve_shared+0x10b/0x250
<4>[  125.501198]  ? __slab_alloc.isra.27.constprop.34+0x3d/0x70
<4>[  125.501219]  __slab_alloc.isra.27.constprop.34+0x3d/0x70
<4>[  125.501243]  ? reservation_object_reserve_shared+0x10b/0x250
<4>[  125.501265]  __kmalloc_track_caller+0x313/0x350
<4>[  125.501287]  krealloc+0x62/0xb0
<4>[  125.501305]  reservation_object_reserve_shared+0x10b/0x250
<4>[  125.501411]  i915_gem_do_execbuffer+0x2040/0x32e0 [i915]
<4>[  125.501522]  ? eb_relocate_slow+0xad0/0xad0 [i915]
<4>[  125.501544]  ? debug_check_no_locks_freed+0x2a0/0x2a0
<4>[  125.501646]  ? i915_gem_execbuffer2_ioctl+0x108/0x770 [i915]
<4>[  125.501755]  ? i915_gem_execbuffer2_ioctl+0x108/0x770 [i915]
<4>[  125.501779]  ? drm_dev_get+0x20/0x20
<4>[  125.501803]  ? __might_fault+0xea/0x1a0
<4>[  125.501902]  ? i915_gem_execbuffer2_ioctl+0x108/0x770 [i915]
<4>[  125.502012]  ? i915_gem_execbuffer_ioctl+0xb90/0xb90 [i915]
<4>[  125.502116]  ? i915_gem_execbuffer_ioctl+0xb90/0xb90 [i915]
<4>[  125.502218]  i915_gem_execbuffer2_ioctl+0x3c5/0x770 [i915]
<4>[  125.502243]  ? drm_dev_enter+0xe0/0xe0
<4>[  125.502260]  ? lock_acquire+0x138/0x3c0
<4>[  125.502362]  ? i915_gem_execbuffer_ioctl+0xb90/0xb90 [i915]
<4>[  125.502470]  ? i915_gem_object_create.part.28+0x570/0x570 [i915]
<4>[  125.502575]  ? i915_gem_execbuffer_ioctl+0xb90/0xb90 [i915]
<4>[  125.502680]  ? i915_gem_execbuffer_ioctl+0xb90/0xb90 [i915]
<4>[  125.502702]  drm_ioctl_kernel+0x151/0x200
<4>[  125.502721]  ? drm_ioctl_permit+0x2a0/0x2a0
<4>[  125.502746]  drm_ioctl+0x63a/0x920
<4>[  125.502844]  ? i915_gem_execbuffer_ioctl+0xb90/0xb90 [i915]
<4>[  125.502868]  ? drm_getstats+0x20/0x20
<4>[  125.502886]  ? trace_hardirqs_on_thunk+0x1a/0x1c
<4>[  125.502919]  do_vfs_ioctl+0x173/0xe90
<4>[  125.502936]  ? trace_hardirqs_on_thunk+0x1a/0x1c
<4>[  125.502957]  ? ioctl_preallocate+0x170/0x170
<4>[  125.502978]  ? trace_hardirqs_on_thunk+0x1a/0x1c
<4>[  125.503002]  ? retint_kernel+0x2d/0x2d
<4>[  125.503024]  ksys_ioctl+0x35/0x60
<4>[  125.503043]  __x64_sys_ioctl+0x6a/0xb0
<4>[  125.503061]  do_syscall_64+0x97/0x400
<4>[  125.503081]  entry_SYSCALL_64_after_hwframe+0x49/0xbe
<4>[  125.503101] RIP: 0033:0x7fe18e4f65d7
<4>[  125.503116] RSP: 002b:00007ffe2ffc06a8 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
<4>[  125.503145] RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007fe18e4f65d7
<4>[  125.503168] RDX: 00007ffe2ffc07f0 RSI: 0000000040406469 RDI: 0000000000000003
<4>[  125.503191] RBP: 00007ffe2ffc07f0 R08: 0000000000000004 R09: 00007ffe2ffcf080
<4>[  125.503215] R10: 000000000002c7de R11: 0000000000000246 R12: 0000000040406469
<4>[  125.503238] R13: 0000000000000003 R14: 0000000000000000 R15: 0000000000000000
<4>[  125.503268] Code: e8 18 a0 c9 da 48 8b 35 25 3a 47 00 49 c7 c0 a0 3b 88 c0 b9 95 05 00 00 48 c7 c2 e0 49 88 c0 48 c7 c7 8d 3b 5d c0 e8 ee 7e db da <0f> 0b 48 89 ef e8 a4 26 f5 da e9 51 fe ff ff e8 8a 26 f5 da e9
<1>[  125.503548] RIP: i915_retire_requests+0x3f2/0x590 [i915] RSP: ffff88004e5dec40

Fixes: 643b450a594e ("drm/i915: Only track live rings for retiring")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180504101147.26286-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/i915_request.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c
index d68739b..e4cf76e 100644
--- a/drivers/gpu/drm/i915/i915_request.c
+++ b/drivers/gpu/drm/i915/i915_request.c
@@ -1426,9 +1426,6 @@ void i915_retire_requests(struct drm_i915_private *i915)
 	if (!i915->gt.active_requests)
 		return;
 
-	/* An outstanding request must be on a still active ring somewhere */
-	GEM_BUG_ON(list_empty(&i915->gt.active_rings));
-
 	list_for_each_entry_safe(ring, tmp, &i915->gt.active_rings, active_link)
 		ring_retire_requests(ring);
 }
-- 
cgit v1.1


From 6f75b16b2683eb7c86ce2c8d150bf3fa759103b9 Mon Sep 17 00:00:00 2001
From: Dmitry Osipenko <digetx@gmail.com>
Date: Fri, 4 May 2018 02:47:19 +0300
Subject: drm/tegra: dc: Balance IOMMU group refcounting

Remove unneeded iommu_group_get() and add missing iommu_group_put(),
correcting IOMMU group refcount. This is a minor correction / cleanup that
doesn't really fix anything because Tegra's IOMMU driver are built-in and
hence groups refcounting can't hold IOMMU driver from unloading.

Signed-off-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/dc.c | 31 +++++++++++++++----------------
 drivers/gpu/drm/tegra/dc.h |  2 +-
 2 files changed, 16 insertions(+), 17 deletions(-)

diff --git a/drivers/gpu/drm/tegra/dc.c b/drivers/gpu/drm/tegra/dc.c
index 9f83a65..f20648f 100644
--- a/drivers/gpu/drm/tegra/dc.c
+++ b/drivers/gpu/drm/tegra/dc.c
@@ -1826,7 +1826,6 @@ static irqreturn_t tegra_dc_irq(int irq, void *data)
 static int tegra_dc_init(struct host1x_client *client)
 {
 	struct drm_device *drm = dev_get_drvdata(client->parent);
-	struct iommu_group *group = iommu_group_get(client->dev);
 	unsigned long flags = HOST1X_SYNCPT_CLIENT_MANAGED;
 	struct tegra_dc *dc = host1x_client_to_dc(client);
 	struct tegra_drm *tegra = drm->dev_private;
@@ -1838,20 +1837,21 @@ static int tegra_dc_init(struct host1x_client *client)
 	if (!dc->syncpt)
 		dev_warn(dc->dev, "failed to allocate syncpoint\n");
 
-	if (group && tegra->domain) {
-		if (group != tegra->group) {
-			err = iommu_attach_group(tegra->domain, group);
+	if (tegra->domain) {
+		dc->group = iommu_group_get(client->dev);
+
+		if (dc->group && dc->group != tegra->group) {
+			err = iommu_attach_group(tegra->domain, dc->group);
 			if (err < 0) {
 				dev_err(dc->dev,
 					"failed to attach to domain: %d\n",
 					err);
+				iommu_group_put(dc->group);
 				return err;
 			}
 
-			tegra->group = group;
+			tegra->group = dc->group;
 		}
-
-		dc->domain = tegra->domain;
 	}
 
 	if (dc->soc->wgrps)
@@ -1916,13 +1916,13 @@ cleanup:
 	if (!IS_ERR(primary))
 		drm_plane_cleanup(primary);
 
-	if (group && dc->domain) {
-		if (group == tegra->group) {
-			iommu_detach_group(dc->domain, group);
+	if (dc->group) {
+		if (dc->group == tegra->group) {
+			iommu_detach_group(tegra->domain, dc->group);
 			tegra->group = NULL;
 		}
 
-		dc->domain = NULL;
+		iommu_group_put(dc->group);
 	}
 
 	return err;
@@ -1931,7 +1931,6 @@ cleanup:
 static int tegra_dc_exit(struct host1x_client *client)
 {
 	struct drm_device *drm = dev_get_drvdata(client->parent);
-	struct iommu_group *group = iommu_group_get(client->dev);
 	struct tegra_dc *dc = host1x_client_to_dc(client);
 	struct tegra_drm *tegra = drm->dev_private;
 	int err;
@@ -1944,13 +1943,13 @@ static int tegra_dc_exit(struct host1x_client *client)
 		return err;
 	}
 
-	if (group && dc->domain) {
-		if (group == tegra->group) {
-			iommu_detach_group(dc->domain, group);
+	if (dc->group) {
+		if (dc->group == tegra->group) {
+			iommu_detach_group(tegra->domain, dc->group);
 			tegra->group = NULL;
 		}
 
-		dc->domain = NULL;
+		iommu_group_put(dc->group);
 	}
 
 	host1x_syncpt_free(dc->syncpt);
diff --git a/drivers/gpu/drm/tegra/dc.h b/drivers/gpu/drm/tegra/dc.h
index d2b50d3..7be786f 100644
--- a/drivers/gpu/drm/tegra/dc.h
+++ b/drivers/gpu/drm/tegra/dc.h
@@ -92,7 +92,7 @@ struct tegra_dc {
 
 	const struct tegra_dc_soc_info *soc;
 
-	struct iommu_domain *domain;
+	struct iommu_group *group;
 };
 
 static inline struct tegra_dc *
-- 
cgit v1.1


From 5fda01b50d769d600c34cb00ab15ce1b6a66c028 Mon Sep 17 00:00:00 2001
From: Dmitry Osipenko <digetx@gmail.com>
Date: Fri, 4 May 2018 02:47:20 +0300
Subject: drm/tegra: gr2d: Add IOMMU support

Attach GR2D to the display IOMMU group in order to provide GR2D access
to BO's IOVA.

Signed-off-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/gr2d.c | 31 +++++++++++++++++++++++++++++--
 1 file changed, 29 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/tegra/gr2d.c b/drivers/gpu/drm/tegra/gr2d.c
index 9a8ea93..8eb530a 100644
--- a/drivers/gpu/drm/tegra/gr2d.c
+++ b/drivers/gpu/drm/tegra/gr2d.c
@@ -7,12 +7,14 @@
  */
 
 #include <linux/clk.h>
+#include <linux/iommu.h>
 
 #include "drm.h"
 #include "gem.h"
 #include "gr2d.h"
 
 struct gr2d {
+	struct iommu_group *group;
 	struct tegra_drm_client client;
 	struct host1x_channel *channel;
 	struct clk *clk;
@@ -30,7 +32,9 @@ static int gr2d_init(struct host1x_client *client)
 	struct tegra_drm_client *drm = host1x_to_drm_client(client);
 	struct drm_device *dev = dev_get_drvdata(client->parent);
 	unsigned long flags = HOST1X_SYNCPT_HAS_BASE;
+	struct tegra_drm *tegra = dev->dev_private;
 	struct gr2d *gr2d = to_gr2d(drm);
+	int err;
 
 	gr2d->channel = host1x_channel_request(client->dev);
 	if (!gr2d->channel)
@@ -42,23 +46,46 @@ static int gr2d_init(struct host1x_client *client)
 		return -ENOMEM;
 	}
 
-	return tegra_drm_register_client(dev->dev_private, drm);
+	if (tegra->domain) {
+		gr2d->group = iommu_group_get(client->dev);
+
+		if (gr2d->group) {
+			err = iommu_attach_group(tegra->domain, gr2d->group);
+			if (err < 0) {
+				dev_err(client->dev,
+					"failed to attach to domain: %d\n",
+					err);
+				host1x_syncpt_free(client->syncpts[0]);
+				host1x_channel_put(gr2d->channel);
+				iommu_group_put(gr2d->group);
+				return err;
+			}
+		}
+	}
+
+	return tegra_drm_register_client(tegra, drm);
 }
 
 static int gr2d_exit(struct host1x_client *client)
 {
 	struct tegra_drm_client *drm = host1x_to_drm_client(client);
 	struct drm_device *dev = dev_get_drvdata(client->parent);
+	struct tegra_drm *tegra = dev->dev_private;
 	struct gr2d *gr2d = to_gr2d(drm);
 	int err;
 
-	err = tegra_drm_unregister_client(dev->dev_private, drm);
+	err = tegra_drm_unregister_client(tegra, drm);
 	if (err < 0)
 		return err;
 
 	host1x_syncpt_free(client->syncpts[0]);
 	host1x_channel_put(gr2d->channel);
 
+	if (gr2d->group) {
+		iommu_detach_group(tegra->domain, gr2d->group);
+		iommu_group_put(gr2d->group);
+	}
+
 	return 0;
 }
 
-- 
cgit v1.1


From c9ac52175b38e7f22fe37b9f943973d9095e53b7 Mon Sep 17 00:00:00 2001
From: Dmitry Osipenko <digetx@gmail.com>
Date: Fri, 4 May 2018 02:47:21 +0300
Subject: drm/tegra: gr3d: Add IOMMU support

Attach GR3D to the displays IOMMU group in order to provide GR3D access
to BO's IOVA.

Signed-off-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/gr3d.c | 27 +++++++++++++++++++++++++++
 1 file changed, 27 insertions(+)

diff --git a/drivers/gpu/drm/tegra/gr3d.c b/drivers/gpu/drm/tegra/gr3d.c
index 28c4ef6..ce51206 100644
--- a/drivers/gpu/drm/tegra/gr3d.c
+++ b/drivers/gpu/drm/tegra/gr3d.c
@@ -9,6 +9,7 @@
 
 #include <linux/clk.h>
 #include <linux/host1x.h>
+#include <linux/iommu.h>
 #include <linux/module.h>
 #include <linux/platform_device.h>
 #include <linux/reset.h>
@@ -20,6 +21,7 @@
 #include "gr3d.h"
 
 struct gr3d {
+	struct iommu_group *group;
 	struct tegra_drm_client client;
 	struct host1x_channel *channel;
 	struct clk *clk_secondary;
@@ -40,7 +42,9 @@ static int gr3d_init(struct host1x_client *client)
 	struct tegra_drm_client *drm = host1x_to_drm_client(client);
 	struct drm_device *dev = dev_get_drvdata(client->parent);
 	unsigned long flags = HOST1X_SYNCPT_HAS_BASE;
+	struct tegra_drm *tegra = dev->dev_private;
 	struct gr3d *gr3d = to_gr3d(drm);
+	int err;
 
 	gr3d->channel = host1x_channel_request(client->dev);
 	if (!gr3d->channel)
@@ -52,6 +56,23 @@ static int gr3d_init(struct host1x_client *client)
 		return -ENOMEM;
 	}
 
+	if (tegra->domain) {
+		gr3d->group = iommu_group_get(client->dev);
+
+		if (gr3d->group) {
+			err = iommu_attach_group(tegra->domain, gr3d->group);
+			if (err < 0) {
+				dev_err(client->dev,
+					"failed to attach to domain: %d\n",
+					err);
+				host1x_syncpt_free(client->syncpts[0]);
+				host1x_channel_put(gr3d->channel);
+				iommu_group_put(gr3d->group);
+				return err;
+			}
+		}
+	}
+
 	return tegra_drm_register_client(dev->dev_private, drm);
 }
 
@@ -59,6 +80,7 @@ static int gr3d_exit(struct host1x_client *client)
 {
 	struct tegra_drm_client *drm = host1x_to_drm_client(client);
 	struct drm_device *dev = dev_get_drvdata(client->parent);
+	struct tegra_drm *tegra = dev->dev_private;
 	struct gr3d *gr3d = to_gr3d(drm);
 	int err;
 
@@ -69,6 +91,11 @@ static int gr3d_exit(struct host1x_client *client)
 	host1x_syncpt_free(client->syncpts[0]);
 	host1x_channel_put(gr3d->channel);
 
+	if (gr3d->group) {
+		iommu_detach_group(tegra->domain, gr3d->group);
+		iommu_group_put(gr3d->group);
+	}
+
 	return 0;
 }
 
-- 
cgit v1.1


From 65972a6fa914b16cc15ffcffcb8bea8c64e78f49 Mon Sep 17 00:00:00 2001
From: Kees Cook <keescook@chromium.org>
Date: Tue, 10 Apr 2018 21:43:15 -0700
Subject: dm mirror: remove VLA usage

On the quest to remove all VLAs from the kernel[1], this avoids VLAs
in dm-raid1.c by just using the maximum size for the stack arrays.
The nr_mirrors value was already capped at 9, so this makes it a trivial
adjustment to the array sizes.

[1] https://lkml.org/lkml/2018/3/7/621

Signed-off-by: Kees Cook <keescook@chromium.org>
Acked-by: Heinz Mauelshagen <heinzm@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
---
 drivers/md/dm-raid1.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/drivers/md/dm-raid1.c b/drivers/md/dm-raid1.c
index 580c49c..5903e49 100644
--- a/drivers/md/dm-raid1.c
+++ b/drivers/md/dm-raid1.c
@@ -23,6 +23,8 @@
 
 #define MAX_RECOVERY 1	/* Maximum number of regions recovered in parallel. */
 
+#define MAX_NR_MIRRORS	(DM_KCOPYD_MAX_REGIONS + 1)
+
 #define DM_RAID1_HANDLE_ERRORS	0x01
 #define DM_RAID1_KEEP_LOG	0x02
 #define errors_handled(p)	((p)->features & DM_RAID1_HANDLE_ERRORS)
@@ -255,7 +257,7 @@ static int mirror_flush(struct dm_target *ti)
 	unsigned long error_bits;
 
 	unsigned int i;
-	struct dm_io_region io[ms->nr_mirrors];
+	struct dm_io_region io[MAX_NR_MIRRORS];
 	struct mirror *m;
 	struct dm_io_request io_req = {
 		.bi_op = REQ_OP_WRITE,
@@ -651,7 +653,7 @@ static void write_callback(unsigned long error, void *context)
 static void do_write(struct mirror_set *ms, struct bio *bio)
 {
 	unsigned int i;
-	struct dm_io_region io[ms->nr_mirrors], *dest = io;
+	struct dm_io_region io[MAX_NR_MIRRORS], *dest = io;
 	struct mirror *m;
 	struct dm_io_request io_req = {
 		.bi_op = REQ_OP_WRITE,
@@ -1083,7 +1085,7 @@ static int mirror_ctr(struct dm_target *ti, unsigned int argc, char **argv)
 	argc -= args_used;
 
 	if (!argc || sscanf(argv[0], "%u%c", &nr_mirrors, &dummy) != 1 ||
-	    nr_mirrors < 2 || nr_mirrors > DM_KCOPYD_MAX_REGIONS + 1) {
+	    nr_mirrors < 2 || nr_mirrors > MAX_NR_MIRRORS) {
 		ti->error = "Invalid number of mirrors";
 		dm_dirty_log_destroy(dl);
 		return -EINVAL;
@@ -1404,7 +1406,7 @@ static void mirror_status(struct dm_target *ti, status_type_t type,
 	int num_feature_args = 0;
 	struct mirror_set *ms = (struct mirror_set *) ti->private;
 	struct dm_dirty_log *log = dm_rh_dirty_log(ms->rh);
-	char buffer[ms->nr_mirrors + 1];
+	char buffer[MAX_NR_MIRRORS + 1];
 
 	switch (type) {
 	case STATUSTYPE_INFO:
-- 
cgit v1.1


From fd5ec0dc34dafa6c5bb46770ca283ae90a4db3c7 Mon Sep 17 00:00:00 2001
From: Thierry Reding <treding@nvidia.com>
Date: Fri, 4 May 2018 15:00:54 +0200
Subject: drm/tegra: dc: Free syncpoint on errors

If an error happens during display controller initialization, the host1x
syncpoint previously requested would be leaked. Properly clean up the
syncpoint along with the other resources.

Reviewed-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/dc.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/gpu/drm/tegra/dc.c b/drivers/gpu/drm/tegra/dc.c
index f20648f..c843f11 100644
--- a/drivers/gpu/drm/tegra/dc.c
+++ b/drivers/gpu/drm/tegra/dc.c
@@ -1925,6 +1925,8 @@ cleanup:
 		iommu_group_put(dc->group);
 	}
 
+	host1x_syncpt_free(dc->syncpt);
+
 	return err;
 }
 
-- 
cgit v1.1


From dd99b4b48833a55817d78a48034b606664b1fff8 Mon Sep 17 00:00:00 2001
From: Thierry Reding <treding@nvidia.com>
Date: Fri, 4 May 2018 14:58:26 +0200
Subject: drm/tegra: gr2d: Properly clean up resources

Failure to register the Tegra DRM client would leak the resources. Move
cleanup code to error unwinding gotos to fix that and share the cleanup
code with the other error paths.

Reviewed-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/gr2d.c | 28 ++++++++++++++++++++++------
 1 file changed, 22 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/tegra/gr2d.c b/drivers/gpu/drm/tegra/gr2d.c
index 8eb530a..0b42e99 100644
--- a/drivers/gpu/drm/tegra/gr2d.c
+++ b/drivers/gpu/drm/tegra/gr2d.c
@@ -42,8 +42,9 @@ static int gr2d_init(struct host1x_client *client)
 
 	client->syncpts[0] = host1x_syncpt_request(client, flags);
 	if (!client->syncpts[0]) {
-		host1x_channel_put(gr2d->channel);
-		return -ENOMEM;
+		err = -ENOMEM;
+		dev_err(client->dev, "failed to request syncpoint: %d\n", err);
+		goto put;
 	}
 
 	if (tegra->domain) {
@@ -55,15 +56,30 @@ static int gr2d_init(struct host1x_client *client)
 				dev_err(client->dev,
 					"failed to attach to domain: %d\n",
 					err);
-				host1x_syncpt_free(client->syncpts[0]);
-				host1x_channel_put(gr2d->channel);
 				iommu_group_put(gr2d->group);
-				return err;
+				goto free;
 			}
 		}
 	}
 
-	return tegra_drm_register_client(tegra, drm);
+	err = tegra_drm_register_client(tegra, drm);
+	if (err < 0) {
+		dev_err(client->dev, "failed to register client: %d\n", err);
+		goto detach;
+	}
+
+	return 0;
+
+detach:
+	if (gr2d->group) {
+		iommu_detach_group(tegra->domain, gr2d->group);
+		iommu_group_put(gr2d->group);
+	}
+free:
+	host1x_syncpt_free(client->syncpts[0]);
+put:
+	host1x_channel_put(gr2d->channel);
+	return err;
 }
 
 static int gr2d_exit(struct host1x_client *client)
-- 
cgit v1.1


From 1975fa56f1c85f5f47ab5cee903b9374a921b122 Mon Sep 17 00:00:00 2001
From: James Morse <james.morse@arm.com>
Date: Wed, 2 May 2018 12:17:02 +0100
Subject: KVM: arm64: Fix order of vcpu_write_sys_reg() arguments

A typo in kvm_vcpu_set_be()'s call:
| vcpu_write_sys_reg(vcpu, SCTLR_EL1, sctlr)
causes us to use the 32bit register value as an index into the sys_reg[]
array, and sail off the end of the linear map when we try to bring up
big-endian secondaries.

| Unable to handle kernel paging request at virtual address ffff80098b982c00
| Mem abort info:
|  ESR = 0x96000045
|  Exception class = DABT (current EL), IL = 32 bits
|   SET = 0, FnV = 0
|   EA = 0, S1PTW = 0
| Data abort info:
|   ISV = 0, ISS = 0x00000045
|   CM = 0, WnR = 1
| swapper pgtable: 4k pages, 48-bit VAs, pgdp = 000000002ea0571a
| [ffff80098b982c00] pgd=00000009ffff8803, pud=0000000000000000
| Internal error: Oops: 96000045 [#1] PREEMPT SMP
| Modules linked in:
| CPU: 2 PID: 1561 Comm: kvm-vcpu-0 Not tainted 4.17.0-rc3-00001-ga912e2261ca6-dirty #1323
| Hardware name: ARM Juno development board (r1) (DT)
| pstate: 60000005 (nZCv daif -PAN -UAO)
| pc : vcpu_write_sys_reg+0x50/0x134
| lr : vcpu_write_sys_reg+0x50/0x134

| Process kvm-vcpu-0 (pid: 1561, stack limit = 0x000000006df4728b)
| Call trace:
|  vcpu_write_sys_reg+0x50/0x134
|  kvm_psci_vcpu_on+0x14c/0x150
|  kvm_psci_0_2_call+0x244/0x2a4
|  kvm_hvc_call_handler+0x1cc/0x258
|  handle_hvc+0x20/0x3c
|  handle_exit+0x130/0x1ec
|  kvm_arch_vcpu_ioctl_run+0x340/0x614
|  kvm_vcpu_ioctl+0x4d0/0x840
|  do_vfs_ioctl+0xc8/0x8d0
|  ksys_ioctl+0x78/0xa8
|  sys_ioctl+0xc/0x18
|  el0_svc_naked+0x30/0x34
| Code: 73620291 604d00b0 00201891 1ab10194 (957a33f8)
|---[ end trace 4b4a4f9628596602 ]---

Fix the order of the arguments.

Fixes: 8d404c4c24613 ("KVM: arm64: Rewrite system register accessors to read/write functions")
CC: Christoffer Dall <cdall@cs.columbia.edu>
Signed-off-by: James Morse <james.morse@arm.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
 arch/arm64/include/asm/kvm_emulate.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/arm64/include/asm/kvm_emulate.h b/arch/arm64/include/asm/kvm_emulate.h
index 23b33e8..1dab3a9 100644
--- a/arch/arm64/include/asm/kvm_emulate.h
+++ b/arch/arm64/include/asm/kvm_emulate.h
@@ -333,7 +333,7 @@ static inline void kvm_vcpu_set_be(struct kvm_vcpu *vcpu)
 	} else {
 		u64 sctlr = vcpu_read_sys_reg(vcpu, SCTLR_EL1);
 		sctlr |= (1 << 25);
-		vcpu_write_sys_reg(vcpu, SCTLR_EL1, sctlr);
+		vcpu_write_sys_reg(vcpu, sctlr, SCTLR_EL1);
 	}
 }
 
-- 
cgit v1.1


From c3616a077190435cb540e134d5dfcd15207817ee Mon Sep 17 00:00:00 2001
From: Valentin Schneider <valentin.schneider@arm.com>
Date: Wed, 2 May 2018 11:53:03 +0100
Subject: KVM: arm/arm64: vgic_init: Cleanup reference to process_maintenance

One comment still mentioned process_maintenance operations after
commit af0614991ab6 ("KVM: arm/arm64: vgic: Get rid of unnecessary
process_maintenance operation")

Update the comment to point to vgic_fold_lr_state instead, which
is where maintenance interrupts are taken care of.

Acked-by: Christoffer Dall <christoffer.dall@arm.com>
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
 virt/kvm/arm/vgic/vgic-init.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/virt/kvm/arm/vgic/vgic-init.c b/virt/kvm/arm/vgic/vgic-init.c
index 68378fe..e07156c 100644
--- a/virt/kvm/arm/vgic/vgic-init.c
+++ b/virt/kvm/arm/vgic/vgic-init.c
@@ -423,7 +423,7 @@ static irqreturn_t vgic_maintenance_handler(int irq, void *data)
 	 * We cannot rely on the vgic maintenance interrupt to be
 	 * delivered synchronously. This means we can only use it to
 	 * exit the VM, and we perform the handling of EOIed
-	 * interrupts on the exit path (see vgic_process_maintenance).
+	 * interrupts on the exit path (see vgic_fold_lr_state).
 	 */
 	return IRQ_HANDLED;
 }
-- 
cgit v1.1


From b220244d41798c6592e7d17843256eb0bae456a0 Mon Sep 17 00:00:00 2001
From: James Morse <james.morse@arm.com>
Date: Fri, 4 May 2018 16:19:24 +0100
Subject: arm64: vgic-v2: Fix proxying of cpuif access

Proxying the cpuif accesses at EL2 makes use of vcpu_data_guest_to_host
and co, which check the endianness, which call into vcpu_read_sys_reg...
which isn't mapped at EL2 (it was inlined before, and got moved OoL
with the VHE optimizations).

The result is of course a nice panic. Let's add some specialized
cruft to keep the broken platforms that require this hack alive.

But, this code used vcpu_data_guest_to_host(), which expected us to
write the value to host memory, instead we have trapped the guest's
read or write to an mmio-device, and are about to replay it using the
host's readl()/writel() which also perform swabbing based on the host
endianness. This goes wrong when both host and guest are big-endian,
as readl()/writel() will undo the guest's swabbing, causing the
big-endian value to be written to device-memory.

What needs doing?
A big-endian guest will have pre-swabbed data before storing, undo this.
If its necessary for the host, writel() will re-swab it.

For a read a big-endian guest expects to swab the data after the load.
The hosts's readl() will correct for host endianness, giving us the
device-memory's value in the register. For a big-endian guest, swab it
as if we'd only done the load.

For a little-endian guest, nothing needs doing as readl()/writel() leave
the correct device-memory value in registers.

Tested on Juno with that rarest of things: a big-endian 64K host.
Based on a patch from Marc Zyngier.

Reported-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Fixes: bf8feb39642b ("arm64: KVM: vgic-v2: Add GICV access from HYP")
Signed-off-by: James Morse <james.morse@arm.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
 arch/arm64/kvm/hyp/vgic-v2-cpuif-proxy.c | 24 +++++++++++++++++++-----
 1 file changed, 19 insertions(+), 5 deletions(-)

diff --git a/arch/arm64/kvm/hyp/vgic-v2-cpuif-proxy.c b/arch/arm64/kvm/hyp/vgic-v2-cpuif-proxy.c
index 86801b6..39be799 100644
--- a/arch/arm64/kvm/hyp/vgic-v2-cpuif-proxy.c
+++ b/arch/arm64/kvm/hyp/vgic-v2-cpuif-proxy.c
@@ -18,11 +18,20 @@
 #include <linux/compiler.h>
 #include <linux/irqchip/arm-gic.h>
 #include <linux/kvm_host.h>
+#include <linux/swab.h>
 
 #include <asm/kvm_emulate.h>
 #include <asm/kvm_hyp.h>
 #include <asm/kvm_mmu.h>
 
+static bool __hyp_text __is_be(struct kvm_vcpu *vcpu)
+{
+	if (vcpu_mode_is_32bit(vcpu))
+		return !!(read_sysreg_el2(spsr) & COMPAT_PSR_E_BIT);
+
+	return !!(read_sysreg(SCTLR_EL1) & SCTLR_ELx_EE);
+}
+
 /*
  * __vgic_v2_perform_cpuif_access -- perform a GICV access on behalf of the
  *				     guest.
@@ -64,14 +73,19 @@ int __hyp_text __vgic_v2_perform_cpuif_access(struct kvm_vcpu *vcpu)
 	addr += fault_ipa - vgic->vgic_cpu_base;
 
 	if (kvm_vcpu_dabt_iswrite(vcpu)) {
-		u32 data = vcpu_data_guest_to_host(vcpu,
-						   vcpu_get_reg(vcpu, rd),
-						   sizeof(u32));
+		u32 data = vcpu_get_reg(vcpu, rd);
+		if (__is_be(vcpu)) {
+			/* guest pre-swabbed data, undo this for writel() */
+			data = swab32(data);
+		}
 		writel_relaxed(data, addr);
 	} else {
 		u32 data = readl_relaxed(addr);
-		vcpu_set_reg(vcpu, rd, vcpu_data_host_to_guest(vcpu, data,
-							       sizeof(u32)));
+		if (__is_be(vcpu)) {
+			/* guest expects swabbed data */
+			data = swab32(data);
+		}
+		vcpu_set_reg(vcpu, rd, data);
 	}
 
 	return 1;
-- 
cgit v1.1


From f2742e471281ee55180cfd49557a5cd26986c6b3 Mon Sep 17 00:00:00 2001
From: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Date: Fri, 4 May 2018 12:56:43 +0100
Subject: drm/i915: Include priority and completed status in request in/out
 tracepoints

It is useful to see the priority as requests are coming in and completed
status as requests are coming out of the GPU.

To achieve this in a more readable way we need to abandon the common
request_hw tracepoint class.

Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20180504115643.22437-1-tvrtko.ursulin@linux.intel.com
---
 drivers/gpu/drm/i915/i915_trace.h | 93 ++++++++++++++++++++++++---------------
 1 file changed, 58 insertions(+), 35 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_trace.h b/drivers/gpu/drm/i915/i915_trace.h
index 408827b..77ee5e5 100644
--- a/drivers/gpu/drm/i915/i915_trace.h
+++ b/drivers/gpu/drm/i915/i915_trace.h
@@ -679,45 +679,68 @@ DEFINE_EVENT(i915_request, i915_request_execute,
 	     TP_ARGS(rq)
 );
 
-DECLARE_EVENT_CLASS(i915_request_hw,
-		    TP_PROTO(struct i915_request *rq, unsigned int port),
-		    TP_ARGS(rq, port),
-
-		    TP_STRUCT__entry(
-				     __field(u32, dev)
-				     __field(u32, hw_id)
-				     __field(u32, ring)
-				     __field(u32, ctx)
-				     __field(u32, seqno)
-				     __field(u32, global_seqno)
-				     __field(u32, port)
-				    ),
-
-		    TP_fast_assign(
-				   __entry->dev = rq->i915->drm.primary->index;
-				   __entry->hw_id = rq->ctx->hw_id;
-				   __entry->ring = rq->engine->id;
-				   __entry->ctx = rq->fence.context;
-				   __entry->seqno = rq->fence.seqno;
-				   __entry->global_seqno = rq->global_seqno;
-				   __entry->port = port;
-				  ),
-
-		    TP_printk("dev=%u, hw_id=%u, ring=%u, ctx=%u, seqno=%u, global=%u, port=%u",
-			      __entry->dev, __entry->hw_id, __entry->ring,
-			      __entry->ctx, __entry->seqno,
-			      __entry->global_seqno, __entry->port)
-);
+TRACE_EVENT(i915_request_in,
+	    TP_PROTO(struct i915_request *rq, unsigned int port),
+	    TP_ARGS(rq, port),
+
+	    TP_STRUCT__entry(
+			     __field(u32, dev)
+			     __field(u32, hw_id)
+			     __field(u32, ring)
+			     __field(u32, ctx)
+			     __field(u32, seqno)
+			     __field(u32, global_seqno)
+			     __field(u32, port)
+			     __field(u32, prio)
+			    ),
 
-DEFINE_EVENT(i915_request_hw, i915_request_in,
-	     TP_PROTO(struct i915_request *rq, unsigned int port),
-	     TP_ARGS(rq, port)
+	    TP_fast_assign(
+			   __entry->dev = rq->i915->drm.primary->index;
+			   __entry->hw_id = rq->ctx->hw_id;
+			   __entry->ring = rq->engine->id;
+			   __entry->ctx = rq->fence.context;
+			   __entry->seqno = rq->fence.seqno;
+			   __entry->global_seqno = rq->global_seqno;
+			   __entry->prio = rq->sched.attr.priority;
+			   __entry->port = port;
+			   ),
+
+	    TP_printk("dev=%u, hw_id=%u, ring=%u, ctx=%u, seqno=%u, prio=%u, global=%u, port=%u",
+		      __entry->dev, __entry->hw_id, __entry->ring, __entry->ctx,
+		      __entry->seqno, __entry->prio, __entry->global_seqno,
+		      __entry->port)
 );
 
-DEFINE_EVENT(i915_request, i915_request_out,
-	     TP_PROTO(struct i915_request *rq),
-	     TP_ARGS(rq)
+TRACE_EVENT(i915_request_out,
+	    TP_PROTO(struct i915_request *rq),
+	    TP_ARGS(rq),
+
+	    TP_STRUCT__entry(
+			     __field(u32, dev)
+			     __field(u32, hw_id)
+			     __field(u32, ring)
+			     __field(u32, ctx)
+			     __field(u32, seqno)
+			     __field(u32, global_seqno)
+			     __field(u32, completed)
+			    ),
+
+	    TP_fast_assign(
+			   __entry->dev = rq->i915->drm.primary->index;
+			   __entry->hw_id = rq->ctx->hw_id;
+			   __entry->ring = rq->engine->id;
+			   __entry->ctx = rq->fence.context;
+			   __entry->seqno = rq->fence.seqno;
+			   __entry->global_seqno = rq->global_seqno;
+			   __entry->completed = i915_request_completed(rq);
+			   ),
+
+		    TP_printk("dev=%u, hw_id=%u, ring=%u, ctx=%u, seqno=%u, global=%u, completed?=%u",
+			      __entry->dev, __entry->hw_id, __entry->ring,
+			      __entry->ctx, __entry->seqno,
+			      __entry->global_seqno, __entry->completed)
 );
+
 #else
 #if !defined(TRACE_HEADER_MULTI_READ)
 static inline void
-- 
cgit v1.1


From 23b8392201e0681b76630c4cea68e1a2e1821ec6 Mon Sep 17 00:00:00 2001
From: Bhadram Varka <vbhadram@nvidia.com>
Date: Wed, 2 May 2018 20:43:58 +0530
Subject: net: phy: broadcom: add support for BCM89610 PHY

It adds support for BCM89610 (Single-Port 10/100/1000BASE-T)
transceiver which is used in P3310 Tegra186 platform.

Signed-off-by: Bhadram Varka <vbhadram@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/phy/broadcom.c | 10 ++++++++++
 include/linux/brcmphy.h    |  1 +
 2 files changed, 11 insertions(+)

diff --git a/drivers/net/phy/broadcom.c b/drivers/net/phy/broadcom.c
index 3bb6b66..f9c2591 100644
--- a/drivers/net/phy/broadcom.c
+++ b/drivers/net/phy/broadcom.c
@@ -720,6 +720,15 @@ static struct phy_driver broadcom_drivers[] = {
 	.get_strings	= bcm_phy_get_strings,
 	.get_stats	= bcm53xx_phy_get_stats,
 	.probe		= bcm53xx_phy_probe,
+}, {
+	.phy_id         = PHY_ID_BCM89610,
+	.phy_id_mask    = 0xfffffff0,
+	.name           = "Broadcom BCM89610",
+	.features       = PHY_GBIT_FEATURES,
+	.flags          = PHY_HAS_INTERRUPT,
+	.config_init    = bcm54xx_config_init,
+	.ack_interrupt  = bcm_phy_ack_intr,
+	.config_intr    = bcm_phy_config_intr,
 } };
 
 module_phy_driver(broadcom_drivers);
@@ -741,6 +750,7 @@ static struct mdio_device_id __maybe_unused broadcom_tbl[] = {
 	{ PHY_ID_BCMAC131, 0xfffffff0 },
 	{ PHY_ID_BCM5241, 0xfffffff0 },
 	{ PHY_ID_BCM5395, 0xfffffff0 },
+	{ PHY_ID_BCM89610, 0xfffffff0 },
 	{ }
 };
 
diff --git a/include/linux/brcmphy.h b/include/linux/brcmphy.h
index d3339dd..b324e01 100644
--- a/include/linux/brcmphy.h
+++ b/include/linux/brcmphy.h
@@ -25,6 +25,7 @@
 #define PHY_ID_BCM54612E		0x03625e60
 #define PHY_ID_BCM54616S		0x03625d10
 #define PHY_ID_BCM57780			0x03625d90
+#define PHY_ID_BCM89610			0x03625cd0
 
 #define PHY_ID_BCM7250			0xae025280
 #define PHY_ID_BCM7260			0xae025190
-- 
cgit v1.1


From 72f17baf2352ded6a1d3f4bb2d15da8c678cd2cb Mon Sep 17 00:00:00 2001
From: Stefano Brivio <sbrivio@redhat.com>
Date: Thu, 3 May 2018 18:13:25 +0200
Subject: openvswitch: Don't swap table in nlattr_set() after OVS_ATTR_NESTED
 is found

If an OVS_ATTR_NESTED attribute type is found while walking
through netlink attributes, we call nlattr_set() recursively
passing the length table for the following nested attributes, if
different from the current one.

However, once we're done with those sub-nested attributes, we
should continue walking through attributes using the current
table, instead of using the one related to the sub-nested
attributes.

For example, given this sequence:

1  OVS_KEY_ATTR_PRIORITY
2  OVS_KEY_ATTR_TUNNEL
3	OVS_TUNNEL_KEY_ATTR_ID
4	OVS_TUNNEL_KEY_ATTR_IPV4_SRC
5	OVS_TUNNEL_KEY_ATTR_IPV4_DST
6	OVS_TUNNEL_KEY_ATTR_TTL
7	OVS_TUNNEL_KEY_ATTR_TP_SRC
8	OVS_TUNNEL_KEY_ATTR_TP_DST
9  OVS_KEY_ATTR_IN_PORT
10 OVS_KEY_ATTR_SKB_MARK
11 OVS_KEY_ATTR_MPLS

we switch to the 'ovs_tunnel_key_lens' table on attribute #3,
and we don't switch back to 'ovs_key_lens' while setting
attributes #9 to #11 in the sequence. As OVS_KEY_ATTR_MPLS
evaluates to 21, and the array size of 'ovs_tunnel_key_lens' is
15, we also get this kind of KASan splat while accessing the
wrong table:

[ 7654.586496] ==================================================================
[ 7654.594573] BUG: KASAN: global-out-of-bounds in nlattr_set+0x164/0xde9 [openvswitch]
[ 7654.603214] Read of size 4 at addr ffffffffc169ecf0 by task handler29/87430
[ 7654.610983]
[ 7654.612644] CPU: 21 PID: 87430 Comm: handler29 Kdump: loaded Not tainted 3.10.0-866.el7.test.x86_64 #1
[ 7654.623030] Hardware name: Dell Inc. PowerEdge R730/072T6D, BIOS 2.1.7 06/16/2016
[ 7654.631379] Call Trace:
[ 7654.634108]  [<ffffffffb65a7c50>] dump_stack+0x19/0x1b
[ 7654.639843]  [<ffffffffb53ff373>] print_address_description+0x33/0x290
[ 7654.647129]  [<ffffffffc169b37b>] ? nlattr_set+0x164/0xde9 [openvswitch]
[ 7654.654607]  [<ffffffffb53ff812>] kasan_report.part.3+0x242/0x330
[ 7654.661406]  [<ffffffffb53ff9b4>] __asan_report_load4_noabort+0x34/0x40
[ 7654.668789]  [<ffffffffc169b37b>] nlattr_set+0x164/0xde9 [openvswitch]
[ 7654.676076]  [<ffffffffc167ef68>] ovs_nla_get_match+0x10c8/0x1900 [openvswitch]
[ 7654.684234]  [<ffffffffb61e9cc8>] ? genl_rcv+0x28/0x40
[ 7654.689968]  [<ffffffffb61e7733>] ? netlink_unicast+0x3f3/0x590
[ 7654.696574]  [<ffffffffc167dea0>] ? ovs_nla_put_tunnel_info+0xb0/0xb0 [openvswitch]
[ 7654.705122]  [<ffffffffb4f41b50>] ? unwind_get_return_address+0xb0/0xb0
[ 7654.712503]  [<ffffffffb65d9355>] ? system_call_fastpath+0x1c/0x21
[ 7654.719401]  [<ffffffffb4f41d79>] ? update_stack_state+0x229/0x370
[ 7654.726298]  [<ffffffffb4f41d79>] ? update_stack_state+0x229/0x370
[ 7654.733195]  [<ffffffffb53fe4b5>] ? kasan_unpoison_shadow+0x35/0x50
[ 7654.740187]  [<ffffffffb53fe62a>] ? kasan_kmalloc+0xaa/0xe0
[ 7654.746406]  [<ffffffffb53fec32>] ? kasan_slab_alloc+0x12/0x20
[ 7654.752914]  [<ffffffffb53fe711>] ? memset+0x31/0x40
[ 7654.758456]  [<ffffffffc165bf92>] ovs_flow_cmd_new+0x2b2/0xf00 [openvswitch]

[snip]

[ 7655.132484] The buggy address belongs to the variable:
[ 7655.138226]  ovs_tunnel_key_lens+0xf0/0xffffffffffffd400 [openvswitch]
[ 7655.145507]
[ 7655.147166] Memory state around the buggy address:
[ 7655.152514]  ffffffffc169eb80: 00 00 00 00 00 00 00 00 00 00 fa fa fa fa fa fa
[ 7655.160585]  ffffffffc169ec00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
[ 7655.168644] >ffffffffc169ec80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 fa fa
[ 7655.176701]                                                              ^
[ 7655.184372]  ffffffffc169ed00: fa fa fa fa 00 00 00 00 fa fa fa fa 00 00 00 05
[ 7655.192431]  ffffffffc169ed80: fa fa fa fa 00 00 00 00 00 00 00 00 00 00 00 00
[ 7655.200490] ==================================================================

Reported-by: Hangbin Liu <liuhangbin@gmail.com>
Fixes: 982b52700482 ("openvswitch: Fix mask generation for nested attributes.")
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Reviewed-by: Sabrina Dubroca <sd@queasysnail.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/openvswitch/flow_netlink.c | 9 +++------
 1 file changed, 3 insertions(+), 6 deletions(-)

diff --git a/net/openvswitch/flow_netlink.c b/net/openvswitch/flow_netlink.c
index 7322aa1..492ab0c 100644
--- a/net/openvswitch/flow_netlink.c
+++ b/net/openvswitch/flow_netlink.c
@@ -1712,13 +1712,10 @@ static void nlattr_set(struct nlattr *attr, u8 val,
 
 	/* The nlattr stream should already have been validated */
 	nla_for_each_nested(nla, attr, rem) {
-		if (tbl[nla_type(nla)].len == OVS_ATTR_NESTED) {
-			if (tbl[nla_type(nla)].next)
-				tbl = tbl[nla_type(nla)].next;
-			nlattr_set(nla, val, tbl);
-		} else {
+		if (tbl[nla_type(nla)].len == OVS_ATTR_NESTED)
+			nlattr_set(nla, val, tbl[nla_type(nla)].next ? : tbl);
+		else
 			memset(nla_data(nla), val, nla_len(nla));
-		}
 
 		if (nla_type(nla) == OVS_KEY_ATTR_CT_STATE)
 			*(u32 *)nla_data(nla) &= CT_SUPPORTED_MASK;
-- 
cgit v1.1


From 2be147f7459db5bbf292e0a6f135037b55e20b39 Mon Sep 17 00:00:00 2001
From: "Gustavo A. R. Silva" <gustavo@embeddedor.com>
Date: Thu, 3 May 2018 13:17:12 -0500
Subject: atm: zatm: Fix potential Spectre v1

pool can be indirectly controlled by user-space, hence leading to
a potential exploitation of the Spectre variant 1 vulnerability.

This issue was detected with the help of Smatch:

drivers/atm/zatm.c:1462 zatm_ioctl() warn: potential spectre issue
'zatm_dev->pool_info' (local cap)

Fix this by sanitizing pool before using it to index
zatm_dev->pool_info

Notice that given that speculation windows are large, the policy is
to kill the speculation on the first load and not worry if it can be
completed with a dependent load/store [1].

[1] https://marc.info/?l=linux-kernel&m=152449131114778&w=2

Cc: stable@vger.kernel.org
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/atm/zatm.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/atm/zatm.c b/drivers/atm/zatm.c
index 1ef67db..9c9a229 100644
--- a/drivers/atm/zatm.c
+++ b/drivers/atm/zatm.c
@@ -28,6 +28,7 @@
 #include <asm/io.h>
 #include <linux/atomic.h>
 #include <linux/uaccess.h>
+#include <linux/nospec.h>
 
 #include "uPD98401.h"
 #include "uPD98402.h"
@@ -1458,6 +1459,8 @@ static int zatm_ioctl(struct atm_dev *dev,unsigned int cmd,void __user *arg)
 					return -EFAULT;
 				if (pool < 0 || pool > ZATM_LAST_POOL)
 					return -EINVAL;
+				pool = array_index_nospec(pool,
+							  ZATM_LAST_POOL + 1);
 				spin_lock_irqsave(&zatm_dev->lock, flags);
 				info = zatm_dev->pool_info[pool];
 				if (cmd == ZATM_GETPOOLZ) {
-- 
cgit v1.1


From acf784bd0ce257fe43da7ca266f7a10b837479d2 Mon Sep 17 00:00:00 2001
From: "Gustavo A. R. Silva" <gustavo@embeddedor.com>
Date: Thu, 3 May 2018 13:45:58 -0500
Subject: net: atm: Fix potential Spectre v1

ioc_data.dev_num can be controlled by user-space, hence leading to
a potential exploitation of the Spectre variant 1 vulnerability.

This issue was detected with the help of Smatch:
net/atm/lec.c:702 lec_vcc_attach() warn: potential spectre issue
'dev_lec'

Fix this by sanitizing ioc_data.dev_num before using it to index
dev_lec. Also, notice that there is another instance in which array
dev_lec is being indexed using ioc_data.dev_num at line 705:
lec_vcc_added(netdev_priv(dev_lec[ioc_data.dev_num]),

Notice that given that speculation windows are large, the policy is
to kill the speculation on the first load and not worry if it can be
completed with a dependent load/store [1].

[1] https://marc.info/?l=linux-kernel&m=152449131114778&w=2

Cc: stable@vger.kernel.org
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/atm/lec.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/net/atm/lec.c b/net/atm/lec.c
index 01d5d20..3138a86 100644
--- a/net/atm/lec.c
+++ b/net/atm/lec.c
@@ -41,6 +41,9 @@ static unsigned char bridge_ula_lec[] = { 0x01, 0x80, 0xc2, 0x00, 0x00 };
 #include <linux/module.h>
 #include <linux/init.h>
 
+/* Hardening for Spectre-v1 */
+#include <linux/nospec.h>
+
 #include "lec.h"
 #include "lec_arpc.h"
 #include "resources.h"
@@ -687,8 +690,10 @@ static int lec_vcc_attach(struct atm_vcc *vcc, void __user *arg)
 	bytes_left = copy_from_user(&ioc_data, arg, sizeof(struct atmlec_ioc));
 	if (bytes_left != 0)
 		pr_info("copy from user failed for %d bytes\n", bytes_left);
-	if (ioc_data.dev_num < 0 || ioc_data.dev_num >= MAX_LEC_ITF ||
-	    !dev_lec[ioc_data.dev_num])
+	if (ioc_data.dev_num < 0 || ioc_data.dev_num >= MAX_LEC_ITF)
+		return -EINVAL;
+	ioc_data.dev_num = array_index_nospec(ioc_data.dev_num, MAX_LEC_ITF);
+	if (!dev_lec[ioc_data.dev_num])
 		return -EINVAL;
 	vpriv = kmalloc(sizeof(struct lec_vcc_priv), GFP_KERNEL);
 	if (!vpriv)
-- 
cgit v1.1


From af50e4ba34f4c45e92535364133d4deb5931c1c5 Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Thu, 3 May 2018 13:37:54 -0700
Subject: nsh: fix infinite loop

syzbot caught an infinite recursion in nsh_gso_segment().

Problem here is that we need to make sure the NSH header is of
reasonable length.

BUG: MAX_LOCK_DEPTH too low!
turning off the locking correctness validator.
depth: 48  max: 48!
48 locks held by syz-executor0/10189:
 #0:         (ptrval) (rcu_read_lock_bh){....}, at: __dev_queue_xmit+0x30f/0x34c0 net/core/dev.c:3517
 #1:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #1:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #2:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #2:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #3:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #3:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #4:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #4:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #5:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #5:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #6:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #6:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #7:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #7:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #8:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #8:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #9:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #9:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #10:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #10:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #11:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #11:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #12:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #12:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #13:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #13:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #14:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #14:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #15:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #15:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #16:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #16:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #17:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #17:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #18:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #18:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #19:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #19:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #20:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #20:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #21:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #21:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #22:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #22:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #23:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #23:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #24:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #24:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #25:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #25:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #26:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #26:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #27:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #27:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #28:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #28:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #29:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #29:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #30:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #30:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #31:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #31:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
dccp_close: ABORT with 65423 bytes unread
 #32:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #32:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #33:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #33:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #34:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #34:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #35:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #35:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #36:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #36:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #37:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #37:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #38:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #38:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #39:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #39:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #40:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #40:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #41:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #41:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #42:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #42:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #43:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #43:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #44:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #44:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #45:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #45:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #46:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #46:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
 #47:         (ptrval) (rcu_read_lock){....}, at: __skb_pull include/linux/skbuff.h:2080 [inline]
 #47:         (ptrval) (rcu_read_lock){....}, at: skb_mac_gso_segment+0x221/0x720 net/core/dev.c:2787
INFO: lockdep is turned off.
CPU: 1 PID: 10189 Comm: syz-executor0 Not tainted 4.17.0-rc2+ #26
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Call Trace:
 __dump_stack lib/dump_stack.c:77 [inline]
 dump_stack+0x1b9/0x294 lib/dump_stack.c:113
 __lock_acquire+0x1788/0x5140 kernel/locking/lockdep.c:3449
 lock_acquire+0x1dc/0x520 kernel/locking/lockdep.c:3920
 rcu_lock_acquire include/linux/rcupdate.h:246 [inline]
 rcu_read_lock include/linux/rcupdate.h:632 [inline]
 skb_mac_gso_segment+0x25b/0x720 net/core/dev.c:2789
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 nsh_gso_segment+0x405/0xb60 net/nsh/nsh.c:107
 skb_mac_gso_segment+0x3ad/0x720 net/core/dev.c:2792
 __skb_gso_segment+0x3bb/0x870 net/core/dev.c:2865
 skb_gso_segment include/linux/netdevice.h:4025 [inline]
 validate_xmit_skb+0x54d/0xd90 net/core/dev.c:3118
 validate_xmit_skb_list+0xbf/0x120 net/core/dev.c:3168
 sch_direct_xmit+0x354/0x11e0 net/sched/sch_generic.c:312
 qdisc_restart net/sched/sch_generic.c:399 [inline]
 __qdisc_run+0x741/0x1af0 net/sched/sch_generic.c:410
 __dev_xmit_skb net/core/dev.c:3243 [inline]
 __dev_queue_xmit+0x28ea/0x34c0 net/core/dev.c:3551
 dev_queue_xmit+0x17/0x20 net/core/dev.c:3616
 packet_snd net/packet/af_packet.c:2951 [inline]
 packet_sendmsg+0x40f8/0x6070 net/packet/af_packet.c:2976
 sock_sendmsg_nosec net/socket.c:629 [inline]
 sock_sendmsg+0xd5/0x120 net/socket.c:639
 __sys_sendto+0x3d7/0x670 net/socket.c:1789
 __do_sys_sendto net/socket.c:1801 [inline]
 __se_sys_sendto net/socket.c:1797 [inline]
 __x64_sys_sendto+0xe1/0x1a0 net/socket.c:1797
 do_syscall_64+0x1b1/0x800 arch/x86/entry/common.c:287
 entry_SYSCALL_64_after_hwframe+0x49/0xbe

Fixes: c411ed854584 ("nsh: add GSO support")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Jiri Benc <jbenc@redhat.com>
Reported-by: syzbot <syzkaller@googlegroups.com>
Acked-by: Jiri Benc <jbenc@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/nsh/nsh.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/net/nsh/nsh.c b/net/nsh/nsh.c
index d7da99a..9696ef9 100644
--- a/net/nsh/nsh.c
+++ b/net/nsh/nsh.c
@@ -57,6 +57,8 @@ int nsh_pop(struct sk_buff *skb)
 		return -ENOMEM;
 	nh = (struct nshhdr *)(skb->data);
 	length = nsh_hdr_len(nh);
+	if (length < NSH_BASE_HDR_LEN)
+		return -EINVAL;
 	inner_proto = tun_p_to_eth_p(nh->np);
 	if (!pskb_may_pull(skb, length))
 		return -ENOMEM;
@@ -90,6 +92,8 @@ static struct sk_buff *nsh_gso_segment(struct sk_buff *skb,
 	if (unlikely(!pskb_may_pull(skb, NSH_BASE_HDR_LEN)))
 		goto out;
 	nsh_len = nsh_hdr_len(nsh_hdr(skb));
+	if (nsh_len < NSH_BASE_HDR_LEN)
+		goto out;
 	if (unlikely(!pskb_may_pull(skb, nsh_len)))
 		goto out;
 
-- 
cgit v1.1


From d89a2adb8bfe6f8949ff389acdb9fa298b6e8e12 Mon Sep 17 00:00:00 2001
From: Michael Chan <michael.chan@broadcom.com>
Date: Thu, 3 May 2018 20:04:27 -0400
Subject: tg3: Fix vunmap() BUG_ON() triggered from tg3_free_consistent().

tg3_free_consistent() calls dma_free_coherent() to free tp->hw_stats
under spinlock and can trigger BUG_ON() in vunmap() because vunmap()
may sleep.  Fix it by removing the spinlock and relying on the
TG3_FLAG_INIT_COMPLETE flag to prevent race conditions between
tg3_get_stats64() and tg3_free_consistent().  TG3_FLAG_INIT_COMPLETE
is always cleared under tp->lock before tg3_free_consistent()
and therefore tg3_get_stats64() can safely access tp->hw_stats
under tp->lock if TG3_FLAG_INIT_COMPLETE is set.

Fixes: f5992b72ebe0 ("tg3: Fix race condition in tg3_get_stats64().")
Reported-by: Zumeng Chen <zumeng.chen@gmail.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/broadcom/tg3.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/drivers/net/ethernet/broadcom/tg3.c b/drivers/net/ethernet/broadcom/tg3.c
index 08bbb63..9f59b12 100644
--- a/drivers/net/ethernet/broadcom/tg3.c
+++ b/drivers/net/ethernet/broadcom/tg3.c
@@ -8733,14 +8733,15 @@ static void tg3_free_consistent(struct tg3 *tp)
 	tg3_mem_rx_release(tp);
 	tg3_mem_tx_release(tp);
 
-	/* Protect tg3_get_stats64() from reading freed tp->hw_stats. */
-	tg3_full_lock(tp, 0);
+	/* tp->hw_stats can be referenced safely:
+	 *     1. under rtnl_lock
+	 *     2. or under tp->lock if TG3_FLAG_INIT_COMPLETE is set.
+	 */
 	if (tp->hw_stats) {
 		dma_free_coherent(&tp->pdev->dev, sizeof(struct tg3_hw_stats),
 				  tp->hw_stats, tp->stats_mapping);
 		tp->hw_stats = NULL;
 	}
-	tg3_full_unlock(tp);
 }
 
 /*
@@ -14178,7 +14179,7 @@ static void tg3_get_stats64(struct net_device *dev,
 	struct tg3 *tp = netdev_priv(dev);
 
 	spin_lock_bh(&tp->lock);
-	if (!tp->hw_stats) {
+	if (!tp->hw_stats || !tg3_flag(tp, INIT_COMPLETE)) {
 		*stats = tp->net_stats_prev;
 		spin_unlock_bh(&tp->lock);
 		return;
-- 
cgit v1.1


From ae552ac2785d69189c865dcea7e71da02180c59c Mon Sep 17 00:00:00 2001
From: YU Bo <tsu.yubo@gmail.com>
Date: Thu, 3 May 2018 23:09:23 -0400
Subject: net/netlink: make sure the headers line up actual value output

Making sure the headers line up properly with the actual value output of the command
`cat /proc/net/netlink`

Before the patch:
<sk       Eth Pid    Groups   Rmem     Wmem     Dump     Locks     Drops     Inode
<ffff8cd2c2f7b000 0   909    00000550 0        0        0 2        0        18946

After the patch:
>sk               Eth Pid        Groups   Rmem     Wmem     Dump  Locks    Drops    Inode
>0000000033203952 0   897        00000113 0        0        0     2        0        14906

Signed-off-by: Bo YU <tsu.yubo@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/netlink/af_netlink.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
index 55342c4..2e2dd88 100644
--- a/net/netlink/af_netlink.c
+++ b/net/netlink/af_netlink.c
@@ -2606,13 +2606,13 @@ static int netlink_seq_show(struct seq_file *seq, void *v)
 {
 	if (v == SEQ_START_TOKEN) {
 		seq_puts(seq,
-			 "sk       Eth Pid    Groups   "
-			 "Rmem     Wmem     Dump     Locks     Drops     Inode\n");
+			 "sk               Eth Pid        Groups   "
+			 "Rmem     Wmem     Dump  Locks    Drops    Inode\n");
 	} else {
 		struct sock *s = v;
 		struct netlink_sock *nlk = nlk_sk(s);
 
-		seq_printf(seq, "%pK %-3d %-6u %08x %-8d %-8d %d %-8d %-8d %-8lu\n",
+		seq_printf(seq, "%pK %-3d %-10u %08x %-8d %-8d %-5d %-8d %-8d %-8lu\n",
 			   s,
 			   s->sk_protocol,
 			   nlk->portid,
-- 
cgit v1.1


From 14224923c3600bae2ac4dcae3bf0c3d4dc2812be Mon Sep 17 00:00:00 2001
From: Rob Taglang <rob@taglang.io>
Date: Thu, 3 May 2018 17:13:06 -0400
Subject: net: ethernet: sun: niu set correct packet size in skb

Currently, skb->len and skb->data_len are set to the page size, not
the packet size. This causes the frame check sequence to not be
located at the "end" of the packet resulting in ethernet frame check
errors. The driver does work currently, but stricter kernel facing
networking solutions like OpenVSwitch will drop these packets as
invalid.

These changes set the packet size correctly so that these errors no
longer occur. The length does not include the frame check sequence, so
that subtraction was removed.

Tested on Oracle/SUN Multithreaded 10-Gigabit Ethernet Network
Controller [108e:abcd] and validated in wireshark.

Signed-off-by: Rob Taglang <rob@taglang.io>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/sun/niu.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ethernet/sun/niu.c b/drivers/net/ethernet/sun/niu.c
index f081de4..88c1247 100644
--- a/drivers/net/ethernet/sun/niu.c
+++ b/drivers/net/ethernet/sun/niu.c
@@ -3443,7 +3443,7 @@ static int niu_process_rx_pkt(struct napi_struct *napi, struct niu *np,
 
 		len = (val & RCR_ENTRY_L2_LEN) >>
 			RCR_ENTRY_L2_LEN_SHIFT;
-		len -= ETH_FCS_LEN;
+		append_size = len + ETH_HLEN + ETH_FCS_LEN;
 
 		addr = (val & RCR_ENTRY_PKT_BUF_ADDR) <<
 			RCR_ENTRY_PKT_BUF_ADDR_SHIFT;
@@ -3453,7 +3453,6 @@ static int niu_process_rx_pkt(struct napi_struct *napi, struct niu *np,
 					 RCR_ENTRY_PKTBUFSZ_SHIFT];
 
 		off = addr & ~PAGE_MASK;
-		append_size = rcr_size;
 		if (num_rcr == 1) {
 			int ptype;
 
@@ -3466,7 +3465,7 @@ static int niu_process_rx_pkt(struct napi_struct *napi, struct niu *np,
 			else
 				skb_checksum_none_assert(skb);
 		} else if (!(val & RCR_ENTRY_MULTI))
-			append_size = len - skb->len;
+			append_size = append_size - skb->len;
 
 		niu_rx_skb_append(skb, page, off, append_size, rcr_size);
 		if ((page->index + rp->rbr_block_size) - rcr_size == addr) {
-- 
cgit v1.1


From 52c5cd1bf0cecf4b146ca07dc513cbe2f4583bb5 Mon Sep 17 00:00:00 2001
From: Antoine Tenart <antoine.tenart@bootlin.com>
Date: Fri, 4 May 2018 17:10:54 +0200
Subject: net: phy: sfp: fix the BR,min computation

In an SFP EEPROM values can be read to get information about a given SFP
module. One of those is the bitrate, which can be determined using a
nominal bitrate in addition with min and max values (in %). The SFP code
currently compute both BR,min and BR,max values thanks to this nominal
and min,max values.

This patch fixes the BR,min computation as the min value should be
subtracted to the nominal one, not added.

Fixes: 9962acf7fb8c ("sfp: add support for 1000Base-PX and 1000Base-BX10")
Signed-off-by: Antoine Tenart <antoine.tenart@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/phy/sfp-bus.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/phy/sfp-bus.c b/drivers/net/phy/sfp-bus.c
index 0381da7..fd6c23f 100644
--- a/drivers/net/phy/sfp-bus.c
+++ b/drivers/net/phy/sfp-bus.c
@@ -125,7 +125,7 @@ void sfp_parse_support(struct sfp_bus *bus, const struct sfp_eeprom_id *id,
 	if (id->base.br_nominal) {
 		if (id->base.br_nominal != 255) {
 			br_nom = id->base.br_nominal * 100;
-			br_min = br_nom + id->base.br_nominal * id->ext.br_min;
+			br_min = br_nom - id->base.br_nominal * id->ext.br_min;
 			br_max = br_nom + id->base.br_nominal * id->ext.br_max;
 		} else if (id->ext.br_max) {
 			br_nom = 250 * id->ext.br_max;
-- 
cgit v1.1


From 280b54ade5914d3b4abe4f0ebe083ddbd4603246 Mon Sep 17 00:00:00 2001
From: Florent Flament <contact@florentflament.com>
Date: Thu, 19 Apr 2018 19:07:00 +0300
Subject: drm/i915: Fix drm:intel_enable_lvds ERROR message in kernel log
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Fix `[drm:intel_enable_lvds] *ERROR* timed out waiting for panel to
power on` in kernel log at boot time.

Toshiba Satellite Z930 laptops needs between 1 and 2 seconds to power
on its screen during Intel i915 DRM initialization. This currently
results in a `[drm:intel_enable_lvds] *ERROR* timed out waiting for
panel to power on` message appearing in the kernel log during boot
time and when stopping the machine.

This change increases the timeout of the `intel_enable_lvds` function
from 1 to 5 seconds, letting enough time for the Satellite 930 LCD
screen to power on, and suppressing the error message from the kernel
log.

This patch has been successfully tested on Linux 4.14 running on a
Toshiba Satellite Z930.

[vsyrjala: bump the timeout from 2 to 5 seconds to match the DP
 code and properly cover the max hw timeout of ~4 seconds, and
 drop the comment about the specific machine since this is not
 a particulary surprising issue, nor specific to that one machine]

Signed-off-by: Florent Flament <contact@florentflament.com>
Cc: stable@vger.kernel.org
Cc: Pavel Petrovic <ppetrovic@acm.org>
Cc: Sérgio M. Basto <sergio@serjux.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=103414
References: https://bugzilla.kernel.org/show_bug.cgi?id=57591
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180419160700.19828-1-ville.syrjala@linux.intel.com
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
---
 drivers/gpu/drm/i915/intel_lvds.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/intel_lvds.c b/drivers/gpu/drm/i915/intel_lvds.c
index d35d2d5..8691c86 100644
--- a/drivers/gpu/drm/i915/intel_lvds.c
+++ b/drivers/gpu/drm/i915/intel_lvds.c
@@ -326,7 +326,8 @@ static void intel_enable_lvds(struct intel_encoder *encoder,
 
 	I915_WRITE(PP_CONTROL(0), I915_READ(PP_CONTROL(0)) | PANEL_POWER_ON);
 	POSTING_READ(lvds_encoder->reg);
-	if (intel_wait_for_register(dev_priv, PP_STATUS(0), PP_ON, PP_ON, 1000))
+
+	if (intel_wait_for_register(dev_priv, PP_STATUS(0), PP_ON, PP_ON, 5000))
 		DRM_ERROR("timed out waiting for panel to power on\n");
 
 	intel_panel_enable_backlight(pipe_config, conn_state);
-- 
cgit v1.1


From 52cc80146d935aa902a3e0fc54268a99fcf68ccf Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Fri, 4 May 2018 13:42:02 +0100
Subject: drm/i915/selftests: Skip the execlists tests on !execlists machines

Ignore the tests looking at the innards of execlists and its submission
tasklets on machines that don't support execlists!

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180504124202.24894-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/selftests/intel_lrc.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/gpu/drm/i915/selftests/intel_lrc.c b/drivers/gpu/drm/i915/selftests/intel_lrc.c
index ee7e22d1..b7460b5 100644
--- a/drivers/gpu/drm/i915/selftests/intel_lrc.c
+++ b/drivers/gpu/drm/i915/selftests/intel_lrc.c
@@ -505,5 +505,9 @@ int intel_execlists_live_selftests(struct drm_i915_private *i915)
 		SUBTEST(live_preempt),
 		SUBTEST(live_late_preempt),
 	};
+
+	if (!HAS_EXECLISTS(i915))
+		return 0;
+
 	return i915_subtests(tests, i915);
 }
-- 
cgit v1.1


From 9f0a93de9139c2b0a59299cd36b61564522458f8 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jo=C3=A3o=20Paulo=20Rechi=20Vita?= <jprvita@gmail.com>
Date: Thu, 19 Apr 2018 07:04:34 -0700
Subject: platform/x86: asus-wireless: Fix NULL pointer dereference
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

When the module is removed the led workqueue is destroyed in the remove
callback, before the led device is unregistered from the led subsystem.

This leads to a NULL pointer derefence when the led device is
unregistered automatically later as part of the module removal cleanup.
Bellow is the backtrace showing the problem.

  BUG: unable to handle kernel NULL pointer dereference at           (null)
  IP: __queue_work+0x8c/0x410
  PGD 0 P4D 0
  Oops: 0000 [#1] SMP NOPTI
  Modules linked in: ccm edac_mce_amd kvm_amd kvm irqbypass crct10dif_pclmul crc32_pclmul ghash_clmulni_intel pcbc aesni_intel aes_x86_64 joydev crypto_simd asus_nb_wmi glue_helper uvcvideo snd_hda_codec_conexant snd_hda_codec_generic snd_hda_codec_hdmi snd_hda_intel asus_wmi snd_hda_codec cryptd snd_hda_core sparse_keymap videobuf2_vmalloc arc4 videobuf2_memops snd_hwdep input_leds videobuf2_v4l2 ath9k psmouse videobuf2_core videodev ath9k_common snd_pcm ath9k_hw media fam15h_power ath k10temp snd_timer mac80211 i2c_piix4 r8169 mii mac_hid cfg80211 asus_wireless(-) snd soundcore wmi shpchp 8250_dw ip_tables x_tables amdkfd amd_iommu_v2 amdgpu radeon chash i2c_algo_bit drm_kms_helper syscopyarea serio_raw sysfillrect sysimgblt fb_sys_fops ahci ttm libahci drm video
  CPU: 3 PID: 2177 Comm: rmmod Not tainted 4.15.0-5-generic #6+dev94.b4287e5bem1-Endless
  Hardware name: ASUSTeK COMPUTER INC. X555DG/X555DG, BIOS 5.011 05/05/2015
  RIP: 0010:__queue_work+0x8c/0x410
  RSP: 0018:ffffbe8cc249fcd8 EFLAGS: 00010086
  RAX: ffff992ac6810800 RBX: 0000000000000000 RCX: 0000000000000008
  RDX: 0000000000000000 RSI: 0000000000000008 RDI: ffff992ac6400e18
  RBP: ffffbe8cc249fd18 R08: ffff992ac6400db0 R09: 0000000000000000
  R10: 0000000000000040 R11: ffff992ac6400dd8 R12: 0000000000002000
  R13: ffff992abd762e00 R14: ffff992abd763e38 R15: 000000000001ebe0
  FS:  00007f318203e700(0000) GS:ffff992aced80000(0000) knlGS:0000000000000000
  CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
  CR2: 0000000000000000 CR3: 00000001c720e000 CR4: 00000000001406e0
  Call Trace:
   queue_work_on+0x38/0x40
   led_state_set+0x2c/0x40 [asus_wireless]
   led_set_brightness_nopm+0x14/0x40
   led_set_brightness+0x37/0x60
   led_trigger_set+0xfc/0x1d0
   led_classdev_unregister+0x32/0xd0
   devm_led_classdev_release+0x11/0x20
   release_nodes+0x109/0x1f0
   devres_release_all+0x3c/0x50
   device_release_driver_internal+0x16d/0x220
   driver_detach+0x3f/0x80
   bus_remove_driver+0x55/0xd0
   driver_unregister+0x2c/0x40
   acpi_bus_unregister_driver+0x15/0x20
   asus_wireless_driver_exit+0x10/0xb7c [asus_wireless]
   SyS_delete_module+0x1da/0x2b0
   entry_SYSCALL_64_fastpath+0x24/0x87
  RIP: 0033:0x7f3181b65fd7
  RSP: 002b:00007ffe74bcbe18 EFLAGS: 00000206 ORIG_RAX: 00000000000000b0
  RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f3181b65fd7
  RDX: 000000000000000a RSI: 0000000000000800 RDI: 0000555ea2559258
  RBP: 0000555ea25591f0 R08: 00007ffe74bcad91 R09: 000000000000000a
  R10: 0000000000000000 R11: 0000000000000206 R12: 0000000000000003
  R13: 00007ffe74bcae00 R14: 0000000000000000 R15: 0000555ea25591f0
  Code: 01 00 00 02 0f 85 7d 01 00 00 48 63 45 d4 48 c7 c6 00 f4 fa 87 49 8b 9d 08 01 00 00 48 03 1c c6 4c 89 f7 e8 87 fb ff ff 48 85 c0 <48> 8b 3b 0f 84 c5 01 00 00 48 39 f8 0f 84 bc 01 00 00 48 89 c7
  RIP: __queue_work+0x8c/0x410 RSP: ffffbe8cc249fcd8
  CR2: 0000000000000000
  ---[ end trace 7aa4f4a232e9c39c ]---

Unregistering the led device on the remove callback before destroying the
workqueue avoids this problem.

https://bugzilla.kernel.org/show_bug.cgi?id=196097

Reported-by: Dun Hum <bitter.taste@gmx.com>
Cc: stable@vger.kernel.org
Signed-off-by: João Paulo Rechi Vita <jprvita@endlessm.com>
Signed-off-by: Darren Hart (VMware) <dvhart@infradead.org>
---
 drivers/platform/x86/asus-wireless.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/platform/x86/asus-wireless.c b/drivers/platform/x86/asus-wireless.c
index d4aeac3..f086469 100644
--- a/drivers/platform/x86/asus-wireless.c
+++ b/drivers/platform/x86/asus-wireless.c
@@ -178,8 +178,10 @@ static int asus_wireless_remove(struct acpi_device *adev)
 {
 	struct asus_wireless_data *data = acpi_driver_data(adev);
 
-	if (data->wq)
+	if (data->wq) {
+		devm_led_classdev_unregister(&adev->dev, &data->led);
 		destroy_workqueue(data->wq);
+	}
 	return 0;
 }
 
-- 
cgit v1.1


From 7fe3fa3b5ec8e75389cce4bf5d052a52e6198d59 Mon Sep 17 00:00:00 2001
From: Mario Limonciello <mario.limonciello@dell.com>
Date: Fri, 20 Apr 2018 12:42:11 -0500
Subject: platform/x86: Kconfig: Fix dell-laptop dependency chain.

As reported by Randy Dunlap:
>> WARNING: unmet direct dependencies detected for DELL_SMBIOS
>>   Depends on [m]: X86 [=y] && X86_PLATFORM_DEVICES [=y]
>>	&& (DCDBAS [=m] ||
>> DCDBAS [=m]=n) && (ACPI_WMI [=n] || ACPI_WMI [=n]=n)
>>   Selected by [y]:
>>   - DELL_LAPTOP [=y] && X86 [=y] && X86_PLATFORM_DEVICES [=y]
>> && DMI [=y]
>> && BACKLIGHT_CLASS_DEVICE [=y] && (ACPI_VIDEO [=n] ||
>>	ACPI_VIDEO [=n]=n)
>> && (RFKILL [=n] || RFKILL [=n]=n) && SERIO_I8042 [=y]
>>

Right now it's possible to set dell laptop to compile in but this
causes dell-smbios to compile in which breaks if dcdbas is a module.

Dell laptop shouldn't select dell-smbios anymore, but depend on it.

Fixes: 32d7b19bad96 (platform/x86: dell-smbios: Resolve dependency error on DCDBAS)
Reported-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Mario Limonciello <mario.limonciello@dell.com>
Cc: stable@vger.kernel.org
Signed-off-by: Darren Hart (VMware) <dvhart@infradead.org>
---
 drivers/platform/x86/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/platform/x86/Kconfig b/drivers/platform/x86/Kconfig
index 39d06dd..bc309c5 100644
--- a/drivers/platform/x86/Kconfig
+++ b/drivers/platform/x86/Kconfig
@@ -154,7 +154,7 @@ config DELL_LAPTOP
 	depends on ACPI_VIDEO || ACPI_VIDEO = n
 	depends on RFKILL || RFKILL = n
 	depends on SERIO_I8042
-	select DELL_SMBIOS
+	depends on DELL_SMBIOS
 	select POWER_SUPPLY
 	select LEDS_CLASS
 	select NEW_LEDS
-- 
cgit v1.1


From 43b78f1155c868208a413082179251f5fba78153 Mon Sep 17 00:00:00 2001
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Date: Fri, 4 May 2018 14:35:12 -0700
Subject: Revert "usb: host: ehci: Use dma_pool_zalloc()"

This reverts commit 22072e83ebd510fb6a090aef9d65ccfda9b1e7e4 as it is
broken.

Alan writes:
	What you can't see just from reading the patch is that in both
	cases (ehci->itd_pool and ehci->sitd_pool) there are two
	allocation paths -- the two branches of an "if" statement -- and
	only one of the paths calls dma_pool_[z]alloc.  However, the
	memset is needed for both paths, and so it can't be eliminated.
	Given that it must be present, there's no advantage to calling
	dma_pool_zalloc rather than dma_pool_alloc.

Reported-by: Erick Cafferata <erick@cafferata.me>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Souptick Joarder <jrdr.linux@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 drivers/usb/host/ehci-mem.c   | 3 ++-
 drivers/usb/host/ehci-sched.c | 6 ++++--
 2 files changed, 6 insertions(+), 3 deletions(-)

diff --git a/drivers/usb/host/ehci-mem.c b/drivers/usb/host/ehci-mem.c
index 4c6c08b..21307d8 100644
--- a/drivers/usb/host/ehci-mem.c
+++ b/drivers/usb/host/ehci-mem.c
@@ -73,9 +73,10 @@ static struct ehci_qh *ehci_qh_alloc (struct ehci_hcd *ehci, gfp_t flags)
 	if (!qh)
 		goto done;
 	qh->hw = (struct ehci_qh_hw *)
-		dma_pool_zalloc(ehci->qh_pool, flags, &dma);
+		dma_pool_alloc(ehci->qh_pool, flags, &dma);
 	if (!qh->hw)
 		goto fail;
+	memset(qh->hw, 0, sizeof *qh->hw);
 	qh->qh_dma = dma;
 	// INIT_LIST_HEAD (&qh->qh_list);
 	INIT_LIST_HEAD (&qh->qtd_list);
diff --git a/drivers/usb/host/ehci-sched.c b/drivers/usb/host/ehci-sched.c
index 28e2a33..e56db44 100644
--- a/drivers/usb/host/ehci-sched.c
+++ b/drivers/usb/host/ehci-sched.c
@@ -1287,7 +1287,7 @@ itd_urb_transaction(
 		} else {
  alloc_itd:
 			spin_unlock_irqrestore(&ehci->lock, flags);
-			itd = dma_pool_zalloc(ehci->itd_pool, mem_flags,
+			itd = dma_pool_alloc(ehci->itd_pool, mem_flags,
 					&itd_dma);
 			spin_lock_irqsave(&ehci->lock, flags);
 			if (!itd) {
@@ -1297,6 +1297,7 @@ itd_urb_transaction(
 			}
 		}
 
+		memset(itd, 0, sizeof(*itd));
 		itd->itd_dma = itd_dma;
 		itd->frame = NO_FRAME;
 		list_add(&itd->itd_list, &sched->td_list);
@@ -2080,7 +2081,7 @@ sitd_urb_transaction(
 		} else {
  alloc_sitd:
 			spin_unlock_irqrestore(&ehci->lock, flags);
-			sitd = dma_pool_zalloc(ehci->sitd_pool, mem_flags,
+			sitd = dma_pool_alloc(ehci->sitd_pool, mem_flags,
 					&sitd_dma);
 			spin_lock_irqsave(&ehci->lock, flags);
 			if (!sitd) {
@@ -2090,6 +2091,7 @@ sitd_urb_transaction(
 			}
 		}
 
+		memset(sitd, 0, sizeof(*sitd));
 		sitd->sitd_dma = sitd_dma;
 		sitd->frame = NO_FRAME;
 		list_add(&sitd->sitd_list, &iso_sched->td_list);
-- 
cgit v1.1


From 1cd4023bc3370578fae05abf48fb4a6ef79c9ea1 Mon Sep 17 00:00:00 2001
From: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Date: Thu, 22 Mar 2018 21:58:27 +0100
Subject: MAINTAINERS: Update Kbuild entry with a few paths

I managed to send some modpost patches to old addresses of both
Masahiro and Michal, and omitted linux-kbuild from cc, because my
tried and trusted scripts/get_maintainer wrapper failed me. Add the
modpost directory to the MAINTAINERS entry, and while at it make the
Makefile glob match scripts/Makefile itself, and add one matching the
Kbuild.include file as well.

Signed-off-by: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
---
 MAINTAINERS | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/MAINTAINERS b/MAINTAINERS
index 79bb02f..dd6a769 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -7677,9 +7677,11 @@ L:	linux-kbuild@vger.kernel.org
 S:	Maintained
 F:	Documentation/kbuild/
 F:	Makefile
-F:	scripts/Makefile.*
+F:	scripts/Kbuild*
+F:	scripts/Makefile*
 F:	scripts/basic/
 F:	scripts/mk*
+F:	scripts/mod/
 F:	scripts/package/
 
 KERNEL JANITORS
-- 
cgit v1.1


From 642ef99be932c4071274b28eaf3d3d85bbb6e78c Mon Sep 17 00:00:00 2001
From: Masahiro Yamada <yamada.masahiro@socionext.com>
Date: Fri, 13 Apr 2018 14:06:10 +0900
Subject: gcc-plugins: fix build condition of SANCOV plugin

Since commit d677a4d60193 ("Makefile: support flag
-fsanitizer-coverage=trace-cmp"), you miss to build the SANCOV
plugin under some circumstances.

  CONFIG_KCOV=y
  CONFIG_KCOV_ENABLE_COMPARISONS=y
  Your compiler does not support -fsanitize-coverage=trace-pc
  Your compiler does not support -fsanitize-coverage=trace-cmp

Under this condition, $(CFLAGS_KCOV) is not empty but contains a
space, so the following ifeq-conditional is false.

    ifeq ($(CFLAGS_KCOV),)

Then, scripts/Makefile.gcc-plugins misses to add sancov_plugin.so to
gcc-plugin-y while the SANCOV plugin is necessary as an alternative
means.

Fixes: d677a4d60193 ("Makefile: support flag -fsanitizer-coverage=trace-cmp")
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Acked-by: Kees Cook <keescook@chromium.org>
---
 scripts/Makefile.gcc-plugins | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/scripts/Makefile.gcc-plugins b/scripts/Makefile.gcc-plugins
index b2a95af..7f5c862 100644
--- a/scripts/Makefile.gcc-plugins
+++ b/scripts/Makefile.gcc-plugins
@@ -14,7 +14,7 @@ ifdef CONFIG_GCC_PLUGINS
   endif
 
   ifdef CONFIG_GCC_PLUGIN_SANCOV
-    ifeq ($(CFLAGS_KCOV),)
+    ifeq ($(strip $(CFLAGS_KCOV)),)
       # It is needed because of the gcc-plugin.sh and gcc version checks.
       gcc-plugin-$(CONFIG_GCC_PLUGIN_SANCOV)           += sancov_plugin.so
 
-- 
cgit v1.1


From d59fbbd09d5d6bdf4ea1f0b4090d175e081ff421 Mon Sep 17 00:00:00 2001
From: Masahiro Yamada <yamada.masahiro@socionext.com>
Date: Tue, 24 Apr 2018 20:07:13 +0900
Subject: kbuild: replace hardcoded bison in cmd_bison_h with $(YACC)

Commit 73a4f6dbe70a ("kbuild: add LEX and YACC variables") missed to
update cmd_bison_h somehow.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
---
 scripts/Makefile.lib | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/scripts/Makefile.lib b/scripts/Makefile.lib
index 07d0740..5af34a2b 100644
--- a/scripts/Makefile.lib
+++ b/scripts/Makefile.lib
@@ -196,7 +196,7 @@ $(obj)/%.tab.c: $(src)/%.y FORCE
 	$(call if_changed,bison)
 
 quiet_cmd_bison_h = YACC    $@
-      cmd_bison_h = bison -o/dev/null --defines=$@ -t -l $<
+      cmd_bison_h = $(YACC) -o/dev/null --defines=$@ -t -l $<
 
 $(obj)/%.tab.h: $(src)/%.y FORCE
 	$(call if_changed,bison_h)
-- 
cgit v1.1


From 0da7e43261142b93307b70da455376ad84414d0a Mon Sep 17 00:00:00 2001
From: Mauro Rossi <issor.oruam@gmail.com>
Date: Tue, 24 Apr 2018 20:08:18 +0900
Subject: genksyms: fix typo in parse.tab.{c,h} generation rules

'quet' is replaced by 'quiet' in scripts/genksyms/Makefile

Signed-off-by: Mauro Rossi <issor.oruam@gmail.com>
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
---
 scripts/genksyms/Makefile | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/scripts/genksyms/Makefile b/scripts/genksyms/Makefile
index ef0287e..03b7ce9 100644
--- a/scripts/genksyms/Makefile
+++ b/scripts/genksyms/Makefile
@@ -14,14 +14,14 @@ genksyms-objs	:= genksyms.o parse.tab.o lex.lex.o
 # so that 'bison: not found' will be displayed if it is missing.
 ifeq ($(findstring 1,$(KBUILD_ENABLE_EXTRA_GCC_CHECKS)),)
 
-quiet_cmd_bison_no_warn = $(quet_cmd_bison)
+quiet_cmd_bison_no_warn = $(quiet_cmd_bison)
       cmd_bison_no_warn = $(YACC) --version >/dev/null; \
 			  $(cmd_bison) 2>/dev/null
 
 $(obj)/parse.tab.c: $(src)/parse.y FORCE
 	$(call if_changed,bison_no_warn)
 
-quiet_cmd_bison_h_no_warn = $(quet_cmd_bison_h)
+quiet_cmd_bison_h_no_warn = $(quiet_cmd_bison_h)
       cmd_bison_h_no_warn = $(YACC) --version >/dev/null; \
 			    $(cmd_bison_h) 2>/dev/null
 
-- 
cgit v1.1


From 7281c8dec8a87685cb54d503d8cceef5a0fc2fdd Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <peterz@infradead.org>
Date: Fri, 20 Apr 2018 14:29:51 +0200
Subject: sched/core: Fix possible Spectre-v1 indexing for
 sched_prio_to_weight[]

> kernel/sched/core.c:6921 cpu_weight_nice_write_s64() warn: potential spectre issue 'sched_prio_to_weight'

Userspace controls @nice, so sanitize the value before using it to
index an array.

Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: <stable@kernel.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 kernel/sched/core.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index ffde9ee..092f7c4 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -8,6 +8,7 @@
 #include "sched.h"
 
 #include <linux/kthread.h>
+#include <linux/nospec.h>
 
 #include <asm/switch_to.h>
 #include <asm/tlb.h>
@@ -6923,11 +6924,15 @@ static int cpu_weight_nice_write_s64(struct cgroup_subsys_state *css,
 				     struct cftype *cft, s64 nice)
 {
 	unsigned long weight;
+	int idx;
 
 	if (nice < MIN_NICE || nice > MAX_NICE)
 		return -ERANGE;
 
-	weight = sched_prio_to_weight[NICE_TO_PRIO(nice) - MAX_RT_PRIO];
+	idx = NICE_TO_PRIO(nice) - MAX_RT_PRIO;
+	idx = array_index_nospec(idx, 40);
+	weight = sched_prio_to_weight[idx];
+
 	return sched_group_set_shares(css_tg(css), scale_load(weight));
 }
 #endif
-- 
cgit v1.1


From 354d7793070611b4df5a79fbb0f12752d0ed0cc5 Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <peterz@infradead.org>
Date: Fri, 20 Apr 2018 15:03:45 +0200
Subject: sched/autogroup: Fix possible Spectre-v1 indexing for
 sched_prio_to_weight[]

> kernel/sched/autogroup.c:230 proc_sched_autogroup_set_nice() warn: potential spectre issue 'sched_prio_to_weight'

Userspace controls @nice, sanitize the array index.

Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: <stable@kernel.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 kernel/sched/autogroup.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/kernel/sched/autogroup.c b/kernel/sched/autogroup.c
index 6be6c57..2d4ff53 100644
--- a/kernel/sched/autogroup.c
+++ b/kernel/sched/autogroup.c
@@ -2,6 +2,7 @@
 /*
  * Auto-group scheduling implementation:
  */
+#include <linux/nospec.h>
 #include "sched.h"
 
 unsigned int __read_mostly sysctl_sched_autogroup_enabled = 1;
@@ -209,7 +210,7 @@ int proc_sched_autogroup_set_nice(struct task_struct *p, int nice)
 	static unsigned long next = INITIAL_JIFFIES;
 	struct autogroup *ag;
 	unsigned long shares;
-	int err;
+	int err, idx;
 
 	if (nice < MIN_NICE || nice > MAX_NICE)
 		return -EINVAL;
@@ -227,7 +228,9 @@ int proc_sched_autogroup_set_nice(struct task_struct *p, int nice)
 
 	next = HZ / 10 + jiffies;
 	ag = autogroup_task_get(p);
-	shares = scale_load(sched_prio_to_weight[nice + 20]);
+
+	idx = array_index_nospec(nice + 20, 40);
+	shares = scale_load(sched_prio_to_weight[idx]);
 
 	down_write(&ag->lock);
 	err = sched_group_set_shares(ag->tg, shares);
-- 
cgit v1.1


From 4411ec1d1993e8dbff2898390e3fed280d88e446 Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <peterz@infradead.org>
Date: Fri, 20 Apr 2018 14:03:18 +0200
Subject: perf/core: Fix possible Spectre-v1 indexing for ->aux_pages[]

> kernel/events/ring_buffer.c:871 perf_mmap_to_page() warn: potential spectre issue 'rb->aux_pages'

Userspace controls @pgoff through the fault address. Sanitize the
array index before doing the array dereference.

Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: <stable@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 kernel/events/ring_buffer.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/kernel/events/ring_buffer.c b/kernel/events/ring_buffer.c
index 6c6b3c4..1d8ca9e 100644
--- a/kernel/events/ring_buffer.c
+++ b/kernel/events/ring_buffer.c
@@ -14,6 +14,7 @@
 #include <linux/slab.h>
 #include <linux/circ_buf.h>
 #include <linux/poll.h>
+#include <linux/nospec.h>
 
 #include "internal.h"
 
@@ -867,8 +868,10 @@ perf_mmap_to_page(struct ring_buffer *rb, unsigned long pgoff)
 			return NULL;
 
 		/* AUX space */
-		if (pgoff >= rb->aux_pgoff)
-			return virt_to_page(rb->aux_pages[pgoff - rb->aux_pgoff]);
+		if (pgoff >= rb->aux_pgoff) {
+			int aux_pgoff = array_index_nospec(pgoff - rb->aux_pgoff, rb->aux_nr_pages);
+			return virt_to_page(rb->aux_pages[aux_pgoff]);
+		}
 	}
 
 	return __perf_mmap_to_page(rb, pgoff);
-- 
cgit v1.1


From ef9ee4ad38445a30909c48998624861716f2a994 Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <peterz@infradead.org>
Date: Fri, 20 Apr 2018 14:06:29 +0200
Subject: perf/x86: Fix possible Spectre-v1 indexing for hw_perf_event cache_*

> arch/x86/events/core.c:319 set_ext_hw_attr() warn: potential spectre issue 'hw_cache_event_ids[cache_type]' (local cap)
> arch/x86/events/core.c:319 set_ext_hw_attr() warn: potential spectre issue 'hw_cache_event_ids' (local cap)
> arch/x86/events/core.c:328 set_ext_hw_attr() warn: potential spectre issue 'hw_cache_extra_regs[cache_type]' (local cap)
> arch/x86/events/core.c:328 set_ext_hw_attr() warn: potential spectre issue 'hw_cache_extra_regs' (local cap)

Userspace controls @config which contains 3 (byte) fields used for a 3
dimensional array deref.

Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: <stable@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 arch/x86/events/core.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/arch/x86/events/core.c b/arch/x86/events/core.c
index a6006e7..b1be0ac 100644
--- a/arch/x86/events/core.c
+++ b/arch/x86/events/core.c
@@ -304,17 +304,20 @@ set_ext_hw_attr(struct hw_perf_event *hwc, struct perf_event *event)
 
 	config = attr->config;
 
-	cache_type = (config >>  0) & 0xff;
+	cache_type = (config >> 0) & 0xff;
 	if (cache_type >= PERF_COUNT_HW_CACHE_MAX)
 		return -EINVAL;
+	cache_type = array_index_nospec(cache_type, PERF_COUNT_HW_CACHE_MAX);
 
 	cache_op = (config >>  8) & 0xff;
 	if (cache_op >= PERF_COUNT_HW_CACHE_OP_MAX)
 		return -EINVAL;
+	cache_op = array_index_nospec(cache_op, PERF_COUNT_HW_CACHE_OP_MAX);
 
 	cache_result = (config >> 16) & 0xff;
 	if (cache_result >= PERF_COUNT_HW_CACHE_RESULT_MAX)
 		return -EINVAL;
+	cache_result = array_index_nospec(cache_result, PERF_COUNT_HW_CACHE_RESULT_MAX);
 
 	val = hw_cache_event_ids[cache_type][cache_op][cache_result];
 
-- 
cgit v1.1


From 46b1b577229a091b137831becaa0fae8690ee15a Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <peterz@infradead.org>
Date: Fri, 20 Apr 2018 14:08:58 +0200
Subject: perf/x86: Fix possible Spectre-v1 indexing for x86_pmu::event_map()

> arch/x86/events/intel/cstate.c:307 cstate_pmu_event_init() warn: potential spectre issue 'pkg_msr' (local cap)
> arch/x86/events/intel/core.c:337 intel_pmu_event_map() warn: potential spectre issue 'intel_perfmon_event_map'
> arch/x86/events/intel/knc.c:122 knc_pmu_event_map() warn: potential spectre issue 'knc_perfmon_event_map'
> arch/x86/events/intel/p4.c:722 p4_pmu_event_map() warn: potential spectre issue 'p4_general_events'
> arch/x86/events/intel/p6.c:116 p6_pmu_event_map() warn: potential spectre issue 'p6_perfmon_event_map'
> arch/x86/events/amd/core.c:132 amd_pmu_event_map() warn: potential spectre issue 'amd_perfmon_event_map'

Userspace controls @attr, sanitize @attr->config before passing it on
to x86_pmu::event_map().

Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: <stable@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 arch/x86/events/core.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/arch/x86/events/core.c b/arch/x86/events/core.c
index b1be0ac..45b2b1c 100644
--- a/arch/x86/events/core.c
+++ b/arch/x86/events/core.c
@@ -27,6 +27,7 @@
 #include <linux/cpu.h>
 #include <linux/bitops.h>
 #include <linux/device.h>
+#include <linux/nospec.h>
 
 #include <asm/apic.h>
 #include <asm/stacktrace.h>
@@ -424,6 +425,8 @@ int x86_setup_perfctr(struct perf_event *event)
 	if (attr->config >= x86_pmu.max_events)
 		return -EINVAL;
 
+	attr->config = array_index_nospec((unsigned long)attr->config, x86_pmu.max_events);
+
 	/*
 	 * The generic map:
 	 */
-- 
cgit v1.1


From 06ce6e9b6d6c09d4129c6e24a1314a395d816c10 Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <peterz@infradead.org>
Date: Fri, 20 Apr 2018 14:23:36 +0200
Subject: perf/x86/msr: Fix possible Spectre-v1 indexing in the MSR driver

> arch/x86/events/msr.c:178 msr_event_init() warn: potential spectre issue 'msr' (local cap)

Userspace controls @attr, sanitize cfg (attr->config) before using it
to index an array.

Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: <stable@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 arch/x86/events/msr.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/arch/x86/events/msr.c b/arch/x86/events/msr.c
index e7edf19..b4771a6 100644
--- a/arch/x86/events/msr.c
+++ b/arch/x86/events/msr.c
@@ -1,5 +1,6 @@
 // SPDX-License-Identifier: GPL-2.0
 #include <linux/perf_event.h>
+#include <linux/nospec.h>
 #include <asm/intel-family.h>
 
 enum perf_msr_id {
@@ -158,9 +159,6 @@ static int msr_event_init(struct perf_event *event)
 	if (event->attr.type != event->pmu->type)
 		return -ENOENT;
 
-	if (cfg >= PERF_MSR_EVENT_MAX)
-		return -EINVAL;
-
 	/* unsupported modes and filters */
 	if (event->attr.exclude_user   ||
 	    event->attr.exclude_kernel ||
@@ -171,6 +169,11 @@ static int msr_event_init(struct perf_event *event)
 	    event->attr.sample_period) /* no sampling */
 		return -EINVAL;
 
+	if (cfg >= PERF_MSR_EVENT_MAX)
+		return -EINVAL;
+
+	cfg = array_index_nospec((unsigned long)cfg, PERF_MSR_EVENT_MAX);
+
 	if (!msr[cfg].attr)
 		return -EINVAL;
 
-- 
cgit v1.1


From a5f81290ce475489fa2551c01a07470c1a4c932e Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <peterz@infradead.org>
Date: Fri, 20 Apr 2018 14:25:48 +0200
Subject: perf/x86/cstate: Fix possible Spectre-v1 indexing for pkg_msr

> arch/x86/events/intel/cstate.c:307 cstate_pmu_event_init() warn: potential spectre issue 'pkg_msr' (local cap)

Userspace controls @attr, sanitize cfg (attr->config) before using it
to index an array.

Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: <stable@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 arch/x86/events/intel/cstate.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/arch/x86/events/intel/cstate.c b/arch/x86/events/intel/cstate.c
index 9aca448..9f8084f 100644
--- a/arch/x86/events/intel/cstate.c
+++ b/arch/x86/events/intel/cstate.c
@@ -92,6 +92,7 @@
 #include <linux/module.h>
 #include <linux/slab.h>
 #include <linux/perf_event.h>
+#include <linux/nospec.h>
 #include <asm/cpu_device_id.h>
 #include <asm/intel-family.h>
 #include "../perf_event.h"
@@ -302,6 +303,7 @@ static int cstate_pmu_event_init(struct perf_event *event)
 	} else if (event->pmu == &cstate_pkg_pmu) {
 		if (cfg >= PERF_CSTATE_PKG_EVENT_MAX)
 			return -EINVAL;
+		cfg = array_index_nospec((unsigned long)cfg, PERF_CSTATE_PKG_EVENT_MAX);
 		if (!pkg_msr[cfg].attr)
 			return -EINVAL;
 		event->hw.event_base = pkg_msr[cfg].msr;
-- 
cgit v1.1


From e0f6d1a526b6adfa9ca3b336b83ece0eed345033 Mon Sep 17 00:00:00 2001
From: Jann Horn <jannh@google.com>
Date: Fri, 4 May 2018 19:59:35 +0200
Subject: x86/vdso: Remove unused file

commit da861e18eccc ("x86, vdso: Get rid of the fake section mechanism")
left this file behind; nothing is using it anymore.

Signed-off-by: Jann Horn <jannh@google.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: luto@amacapital.net
Link: http://lkml.kernel.org/r/20180504175935.104085-1-jannh@google.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 arch/x86/entry/vdso/vdso32/vdso-fakesections.c | 1 -
 1 file changed, 1 deletion(-)
 delete mode 100644 arch/x86/entry/vdso/vdso32/vdso-fakesections.c

diff --git a/arch/x86/entry/vdso/vdso32/vdso-fakesections.c b/arch/x86/entry/vdso/vdso32/vdso-fakesections.c
deleted file mode 100644
index 541468e..0000000
--- a/arch/x86/entry/vdso/vdso32/vdso-fakesections.c
+++ /dev/null
@@ -1 +0,0 @@
-#include "../vdso-fakesections.c"
-- 
cgit v1.1


From c18e9a098605abe5a1dc1c5dd9cfeda322ed36d8 Mon Sep 17 00:00:00 2001
From: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Date: Tue, 24 Apr 2018 16:39:42 +0100
Subject: drm: rcar-du: of: Include header to define prototypes

The symbol 'rcar_du_of_init' is defined by the rcar_du_of module header,
but it is not included by the C implementation.

Include the header to correctly define the function prototypes.

Fixes the following warning:

linux/drivers/gpu/drm/rcar-du/rcar_du_of.c:319:13:
   warning: symbol 'rcar_du_of_init' was not declared. Should it be static?
    CC      drivers/gpu/drm/rcar-du/rcar_du_of.o

Fixes: 81c0e3dd8292 ("drm: rcar-du: Fix legacy DT to create LVDS encoder nodes")
Signed-off-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Reviewed-by: Vaishali Thakkar <vthakkar@vaishalithakkar.in>
Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
---
 drivers/gpu/drm/rcar-du/rcar_du_of.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/rcar-du/rcar_du_of.c b/drivers/gpu/drm/rcar-du/rcar_du_of.c
index 68a0b82..afef696 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_of.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_of.c
@@ -18,6 +18,7 @@
 
 #include "rcar_du_crtc.h"
 #include "rcar_du_drv.h"
+#include "rcar_du_of.h"
 
 /* -----------------------------------------------------------------------------
  * Generic Overlay Handling
-- 
cgit v1.1


From 3b0033eb39360dc655466c5f1e6852d5a83b384d Mon Sep 17 00:00:00 2001
From: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Date: Tue, 24 Apr 2018 16:40:03 +0100
Subject: drm: rcar-du: Use NULL for table initialisation

Replace the initialisation of the vsps table with a NULL specifier.

Fixes the following warning:
 linux/drivers/gpu/drm/rcar-du/rcar_du_kms.c:483:40:
    warning: Using plain integer as NULL pointer
      CC      drivers/gpu/drm/rcar-du/rcar_du_kms.o

Fixes: 3e81374e2014 ("drm: rcar-du: Support multiple sources from the same VSP")
Signed-off-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Reviewed-by: Vaishali Thakkar <vthakkar@vaishalithakkar.in>
Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
---
 drivers/gpu/drm/rcar-du/rcar_du_kms.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/rcar-du/rcar_du_kms.c b/drivers/gpu/drm/rcar-du/rcar_du_kms.c
index 0329b35..0c8b7e5 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_kms.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_kms.c
@@ -441,7 +441,7 @@ static int rcar_du_vsps_init(struct rcar_du_device *rcdu)
 	struct {
 		struct device_node *np;
 		unsigned int crtcs_mask;
-	} vsps[RCAR_DU_MAX_VSPS] = { { 0, }, };
+	} vsps[RCAR_DU_MAX_VSPS] = { { NULL, }, };
 	unsigned int vsps_count = 0;
 	unsigned int cells;
 	unsigned int i;
-- 
cgit v1.1


From a4af8423cfe50e5cafa1893fc40643245793e3cd Mon Sep 17 00:00:00 2001
From: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Date: Thu, 26 Apr 2018 17:53:30 +0100
Subject: dt-bindings: display: renesas: du: Increase indent in output table

The DU output table lists the port combinations for each supported DU
type.  Newer models of R-Car Gen3 platforms have an increased string
length.

Increase the table indentation in preparation for supporting new target
types.

Signed-off-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
---
 .../devicetree/bindings/display/renesas,du.txt     | 26 +++++++++++-----------
 1 file changed, 13 insertions(+), 13 deletions(-)

diff --git a/Documentation/devicetree/bindings/display/renesas,du.txt b/Documentation/devicetree/bindings/display/renesas,du.txt
index c9cd17f..a36a6e7 100644
--- a/Documentation/devicetree/bindings/display/renesas,du.txt
+++ b/Documentation/devicetree/bindings/display/renesas,du.txt
@@ -47,20 +47,20 @@ bindings specified in Documentation/devicetree/bindings/graph.txt.
 The following table lists for each supported model the port number
 corresponding to each DU output.
 
-                      Port0          Port1          Port2          Port3
+                        Port0          Port1          Port2          Port3
 -----------------------------------------------------------------------------
- R8A7743 (RZ/G1M)     DPAD 0         LVDS 0         -              -
- R8A7745 (RZ/G1E)     DPAD 0         DPAD 1         -              -
- R8A7779 (R-Car H1)   DPAD 0         DPAD 1         -              -
- R8A7790 (R-Car H2)   DPAD 0         LVDS 0         LVDS 1         -
- R8A7791 (R-Car M2-W) DPAD 0         LVDS 0         -              -
- R8A7792 (R-Car V2H)  DPAD 0         DPAD 1         -              -
- R8A7793 (R-Car M2-N) DPAD 0         LVDS 0         -              -
- R8A7794 (R-Car E2)   DPAD 0         DPAD 1         -              -
- R8A7795 (R-Car H3)   DPAD 0         HDMI 0         HDMI 1         LVDS 0
- R8A7796 (R-Car M3-W) DPAD 0         HDMI 0         LVDS 0         -
- R8A77970 (R-Car V3M) DPAD 0         LVDS 0         -              -
- R8A77995 (R-Car D3)  DPAD 0         LVDS 0         LVDS 1         -
+ R8A7743 (RZ/G1M)       DPAD 0         LVDS 0         -              -
+ R8A7745 (RZ/G1E)       DPAD 0         DPAD 1         -              -
+ R8A7779 (R-Car H1)     DPAD 0         DPAD 1         -              -
+ R8A7790 (R-Car H2)     DPAD 0         LVDS 0         LVDS 1         -
+ R8A7791 (R-Car M2-W)   DPAD 0         LVDS 0         -              -
+ R8A7792 (R-Car V2H)    DPAD 0         DPAD 1         -              -
+ R8A7793 (R-Car M2-N)   DPAD 0         LVDS 0         -              -
+ R8A7794 (R-Car E2)     DPAD 0         DPAD 1         -              -
+ R8A7795 (R-Car H3)     DPAD 0         HDMI 0         HDMI 1         LVDS 0
+ R8A7796 (R-Car M3-W)   DPAD 0         HDMI 0         LVDS 0         -
+ R8A77970 (R-Car V3M)   DPAD 0         LVDS 0         -              -
+ R8A77995 (R-Car D3)    DPAD 0         LVDS 0         LVDS 1         -
 
 
 Example: R8A7795 (R-Car H3) ES2.0 DU
-- 
cgit v1.1


From dc8142901befabea974393d49b803f131243feb4 Mon Sep 17 00:00:00 2001
From: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Date: Thu, 26 Apr 2018 17:53:31 +0100
Subject: dt-bindings: display: renesas: du: Document the r8a77965 bindings

Document the M3-N (r8a77965) SoC in the R-Car DU bindings.

Signed-off-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
---
 Documentation/devicetree/bindings/display/renesas,du.txt | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/Documentation/devicetree/bindings/display/renesas,du.txt b/Documentation/devicetree/bindings/display/renesas,du.txt
index a36a6e7..7c6854b 100644
--- a/Documentation/devicetree/bindings/display/renesas,du.txt
+++ b/Documentation/devicetree/bindings/display/renesas,du.txt
@@ -13,6 +13,7 @@ Required Properties:
     - "renesas,du-r8a7794" for R8A7794 (R-Car E2) compatible DU
     - "renesas,du-r8a7795" for R8A7795 (R-Car H3) compatible DU
     - "renesas,du-r8a7796" for R8A7796 (R-Car M3-W) compatible DU
+    - "renesas,du-r8a77965" for R8A77965 (R-Car M3-N) compatible DU
     - "renesas,du-r8a77970" for R8A77970 (R-Car V3M) compatible DU
     - "renesas,du-r8a77995" for R8A77995 (R-Car D3) compatible DU
 
@@ -59,6 +60,7 @@ corresponding to each DU output.
  R8A7794 (R-Car E2)     DPAD 0         DPAD 1         -              -
  R8A7795 (R-Car H3)     DPAD 0         HDMI 0         HDMI 1         LVDS 0
  R8A7796 (R-Car M3-W)   DPAD 0         HDMI 0         LVDS 0         -
+ R8A77965 (R-Car M3-N)  DPAD 0         HDMI 0         LVDS 0         -
  R8A77970 (R-Car V3M)   DPAD 0         LVDS 0         -              -
  R8A77995 (R-Car D3)    DPAD 0         LVDS 0         LVDS 1         -
 
-- 
cgit v1.1


From 4012532e040ba4c6bba0883c27b57adb1fd88db8 Mon Sep 17 00:00:00 2001
From: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Date: Thu, 26 Apr 2018 17:53:33 +0100
Subject: drm: rcar-du: Use the correct naming for ODPM fields in DEFR6

The naming of the fields for the ODPM signals in the DU extensional
function control register 6 (DEFR6) is incorrect against the data sheets
for both R-Car Gen2 and R-Car Gen3.

Rename the fields to match the datasheet.

Signed-off-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
---
 drivers/gpu/drm/rcar-du/rcar_du_group.c |  4 ++--
 drivers/gpu/drm/rcar-du/rcar_du_regs.h  | 16 ++++++++--------
 2 files changed, 10 insertions(+), 10 deletions(-)

diff --git a/drivers/gpu/drm/rcar-du/rcar_du_group.c b/drivers/gpu/drm/rcar-du/rcar_du_group.c
index 2f37ea9..eead202 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_group.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_group.c
@@ -46,10 +46,10 @@ void rcar_du_group_write(struct rcar_du_group *rgrp, u32 reg, u32 data)
 
 static void rcar_du_group_setup_pins(struct rcar_du_group *rgrp)
 {
-	u32 defr6 = DEFR6_CODE | DEFR6_ODPM12_DISP;
+	u32 defr6 = DEFR6_CODE | DEFR6_ODPM02_DISP;
 
 	if (rgrp->num_crtcs > 1)
-		defr6 |= DEFR6_ODPM22_DISP;
+		defr6 |= DEFR6_ODPM12_DISP;
 
 	rcar_du_group_write(rgrp, DEFR6, defr6);
 }
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_regs.h b/drivers/gpu/drm/rcar-du/rcar_du_regs.h
index d5bae99..9dfd220 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_regs.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_regs.h
@@ -187,14 +187,14 @@
 
 #define DEFR6			0x000e8
 #define DEFR6_CODE		(0x7778 << 16)
-#define DEFR6_ODPM22_DSMR	(0 << 10)
-#define DEFR6_ODPM22_DISP	(2 << 10)
-#define DEFR6_ODPM22_CDE	(3 << 10)
-#define DEFR6_ODPM22_MASK	(3 << 10)
-#define DEFR6_ODPM12_DSMR	(0 << 8)
-#define DEFR6_ODPM12_DISP	(2 << 8)
-#define DEFR6_ODPM12_CDE	(3 << 8)
-#define DEFR6_ODPM12_MASK	(3 << 8)
+#define DEFR6_ODPM12_DSMR	(0 << 10)
+#define DEFR6_ODPM12_DISP	(2 << 10)
+#define DEFR6_ODPM12_CDE	(3 << 10)
+#define DEFR6_ODPM12_MASK	(3 << 10)
+#define DEFR6_ODPM02_DSMR	(0 << 8)
+#define DEFR6_ODPM02_DISP	(2 << 8)
+#define DEFR6_ODPM02_CDE	(3 << 8)
+#define DEFR6_ODPM02_MASK	(3 << 8)
 #define DEFR6_TCNE1		(1 << 6)
 #define DEFR6_TCNE0		(1 << 4)
 #define DEFR6_MLOS1		(1 << 2)
-- 
cgit v1.1


From 425f33bdcd4f492546354cbe4daafe420c450a83 Mon Sep 17 00:00:00 2001
From: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Date: Fri, 27 Apr 2018 23:21:50 +0100
Subject: dt-bindings: display: renesas: Add R-Car M3-N HDMI TX DT bindings

The M3-N HDMI TX controller is compatible with the M3-W and H3. No
extension to the DT bindings are needed.

Add an SoC-specific compatible string in case differences between the IP
versions are found later and require model-specific handling.

Signed-off-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
---
 Documentation/devicetree/bindings/display/bridge/renesas,dw-hdmi.txt | 1 +
 1 file changed, 1 insertion(+)

diff --git a/Documentation/devicetree/bindings/display/bridge/renesas,dw-hdmi.txt b/Documentation/devicetree/bindings/display/bridge/renesas,dw-hdmi.txt
index 3a72a10..a41d280 100644
--- a/Documentation/devicetree/bindings/display/bridge/renesas,dw-hdmi.txt
+++ b/Documentation/devicetree/bindings/display/bridge/renesas,dw-hdmi.txt
@@ -14,6 +14,7 @@ Required properties:
 - compatible : Shall contain one or more of
   - "renesas,r8a7795-hdmi" for R8A7795 (R-Car H3) compatible HDMI TX
   - "renesas,r8a7796-hdmi" for R8A7796 (R-Car M3-W) compatible HDMI TX
+  - "renesas,r8a77965-hdmi" for R8A77965 (R-Car M3-N) compatible HDMI TX
   - "renesas,rcar-gen3-hdmi" for the generic R-Car Gen3 compatible HDMI TX
 
     When compatible with generic versions, nodes must list the SoC-specific
-- 
cgit v1.1


From 5361cc7f8e9146f393cfcb76890d8c80a4e73086 Mon Sep 17 00:00:00 2001
From: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Date: Fri, 27 Apr 2018 23:21:52 +0100
Subject: drm: rcar-du: Split CRTC handling to support hardware indexing

The DU CRTC driver does not support distinguishing between a hardware
index, and a software (CRTC) index in the event that a DU channel might
not be populated by the hardware.

Support this by adapting the rcar_du_device_info structure to store a
bitmask of available channels rather than a count of CRTCs. The count
can then be obtained by determining the hamming weight of the bitmask.

This allows the rcar_du_crtc_create() function to distinguish between
both index types, and non-populated DU channels will be skipped without
leaving a gap in the software CRTC indexes.

Signed-off-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
---
 drivers/gpu/drm/rcar-du/rcar_du_crtc.c | 26 ++++++++++++++------------
 drivers/gpu/drm/rcar-du/rcar_du_crtc.h |  3 ++-
 drivers/gpu/drm/rcar-du/rcar_du_drv.c  | 22 +++++++++++-----------
 drivers/gpu/drm/rcar-du/rcar_du_drv.h  |  4 ++--
 drivers/gpu/drm/rcar-du/rcar_du_kms.c  | 18 +++++++++++++-----
 5 files changed, 42 insertions(+), 31 deletions(-)

diff --git a/drivers/gpu/drm/rcar-du/rcar_du_crtc.c b/drivers/gpu/drm/rcar-du/rcar_du_crtc.c
index c442053..f2a0bd1 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_crtc.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_crtc.c
@@ -767,7 +767,8 @@ static irqreturn_t rcar_du_crtc_irq(int irq, void *arg)
  * Initialization
  */
 
-int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int index)
+int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int swindex,
+			unsigned int hwindex)
 {
 	static const unsigned int mmio_offsets[] = {
 		DU0_REG_OFFSET, DU1_REG_OFFSET, DU2_REG_OFFSET, DU3_REG_OFFSET
@@ -775,7 +776,7 @@ int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int index)
 
 	struct rcar_du_device *rcdu = rgrp->dev;
 	struct platform_device *pdev = to_platform_device(rcdu->dev);
-	struct rcar_du_crtc *rcrtc = &rcdu->crtcs[index];
+	struct rcar_du_crtc *rcrtc = &rcdu->crtcs[swindex];
 	struct drm_crtc *crtc = &rcrtc->crtc;
 	struct drm_plane *primary;
 	unsigned int irqflags;
@@ -787,7 +788,7 @@ int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int index)
 
 	/* Get the CRTC clock and the optional external clock. */
 	if (rcar_du_has(rcdu, RCAR_DU_FEATURE_CRTC_IRQ_CLOCK)) {
-		sprintf(clk_name, "du.%u", index);
+		sprintf(clk_name, "du.%u", hwindex);
 		name = clk_name;
 	} else {
 		name = NULL;
@@ -795,16 +796,16 @@ int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int index)
 
 	rcrtc->clock = devm_clk_get(rcdu->dev, name);
 	if (IS_ERR(rcrtc->clock)) {
-		dev_err(rcdu->dev, "no clock for CRTC %u\n", index);
+		dev_err(rcdu->dev, "no clock for DU channel %u\n", hwindex);
 		return PTR_ERR(rcrtc->clock);
 	}
 
-	sprintf(clk_name, "dclkin.%u", index);
+	sprintf(clk_name, "dclkin.%u", hwindex);
 	clk = devm_clk_get(rcdu->dev, clk_name);
 	if (!IS_ERR(clk)) {
 		rcrtc->extclock = clk;
 	} else if (PTR_ERR(rcrtc->clock) == -EPROBE_DEFER) {
-		dev_info(rcdu->dev, "can't get external clock %u\n", index);
+		dev_info(rcdu->dev, "can't get external clock %u\n", hwindex);
 		return -EPROBE_DEFER;
 	}
 
@@ -813,13 +814,13 @@ int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int index)
 	spin_lock_init(&rcrtc->vblank_lock);
 
 	rcrtc->group = rgrp;
-	rcrtc->mmio_offset = mmio_offsets[index];
-	rcrtc->index = index;
+	rcrtc->mmio_offset = mmio_offsets[hwindex];
+	rcrtc->index = hwindex;
 
 	if (rcar_du_has(rcdu, RCAR_DU_FEATURE_VSP1_SOURCE))
 		primary = &rcrtc->vsp->planes[rcrtc->vsp_pipe].plane;
 	else
-		primary = &rgrp->planes[index % 2].plane;
+		primary = &rgrp->planes[swindex % 2].plane;
 
 	ret = drm_crtc_init_with_planes(rcdu->ddev, crtc, primary,
 					NULL, &crtc_funcs, NULL);
@@ -833,7 +834,8 @@ int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int index)
 
 	/* Register the interrupt handler. */
 	if (rcar_du_has(rcdu, RCAR_DU_FEATURE_CRTC_IRQ_CLOCK)) {
-		irq = platform_get_irq(pdev, index);
+		/* The IRQ's are associated with the CRTC (sw)index. */
+		irq = platform_get_irq(pdev, swindex);
 		irqflags = 0;
 	} else {
 		irq = platform_get_irq(pdev, 0);
@@ -841,7 +843,7 @@ int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int index)
 	}
 
 	if (irq < 0) {
-		dev_err(rcdu->dev, "no IRQ for CRTC %u\n", index);
+		dev_err(rcdu->dev, "no IRQ for CRTC %u\n", swindex);
 		return irq;
 	}
 
@@ -849,7 +851,7 @@ int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int index)
 			       dev_name(rcdu->dev), rcrtc);
 	if (ret < 0) {
 		dev_err(rcdu->dev,
-			"failed to register IRQ for CRTC %u\n", index);
+			"failed to register IRQ for CRTC %u\n", swindex);
 		return ret;
 	}
 
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_crtc.h b/drivers/gpu/drm/rcar-du/rcar_du_crtc.h
index fdc2bf9..84b5e23 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_crtc.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_crtc.h
@@ -80,7 +80,8 @@ enum rcar_du_output {
 	RCAR_DU_OUTPUT_MAX,
 };
 
-int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int index);
+int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int swindex,
+			unsigned int hwindex);
 void rcar_du_crtc_suspend(struct rcar_du_crtc *rcrtc);
 void rcar_du_crtc_resume(struct rcar_du_crtc *rcrtc);
 
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_drv.c b/drivers/gpu/drm/rcar-du/rcar_du_drv.c
index 3917d83..2aa392b 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_drv.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_drv.c
@@ -40,7 +40,7 @@ static const struct rcar_du_device_info rzg1_du_r8a7743_info = {
 	.gen = 2,
 	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK
 		  | RCAR_DU_FEATURE_EXT_CTRL_REGS,
-	.num_crtcs = 2,
+	.channels_mask = BIT(1) | BIT(0),
 	.routes = {
 		/*
 		 * R8A7743 has one RGB output and one LVDS output
@@ -61,7 +61,7 @@ static const struct rcar_du_device_info rzg1_du_r8a7745_info = {
 	.gen = 2,
 	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK
 		  | RCAR_DU_FEATURE_EXT_CTRL_REGS,
-	.num_crtcs = 2,
+	.channels_mask = BIT(1) | BIT(0),
 	.routes = {
 		/*
 		 * R8A7745 has two RGB outputs
@@ -80,7 +80,7 @@ static const struct rcar_du_device_info rzg1_du_r8a7745_info = {
 static const struct rcar_du_device_info rcar_du_r8a7779_info = {
 	.gen = 2,
 	.features = 0,
-	.num_crtcs = 2,
+	.channels_mask = BIT(1) | BIT(0),
 	.routes = {
 		/*
 		 * R8A7779 has two RGB outputs and one (currently unsupported)
@@ -102,7 +102,7 @@ static const struct rcar_du_device_info rcar_du_r8a7790_info = {
 	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK
 		  | RCAR_DU_FEATURE_EXT_CTRL_REGS,
 	.quirks = RCAR_DU_QUIRK_ALIGN_128B,
-	.num_crtcs = 3,
+	.channels_mask = BIT(2) | BIT(1) | BIT(0),
 	.routes = {
 		/*
 		 * R8A7790 has one RGB output, two LVDS outputs and one
@@ -129,7 +129,7 @@ static const struct rcar_du_device_info rcar_du_r8a7791_info = {
 	.gen = 2,
 	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK
 		  | RCAR_DU_FEATURE_EXT_CTRL_REGS,
-	.num_crtcs = 2,
+	.channels_mask = BIT(1) | BIT(0),
 	.routes = {
 		/*
 		 * R8A779[13] has one RGB output, one LVDS output and one
@@ -151,7 +151,7 @@ static const struct rcar_du_device_info rcar_du_r8a7792_info = {
 	.gen = 2,
 	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK
 		  | RCAR_DU_FEATURE_EXT_CTRL_REGS,
-	.num_crtcs = 2,
+	.channels_mask = BIT(1) | BIT(0),
 	.routes = {
 		/* R8A7792 has two RGB outputs. */
 		[RCAR_DU_OUTPUT_DPAD0] = {
@@ -169,7 +169,7 @@ static const struct rcar_du_device_info rcar_du_r8a7794_info = {
 	.gen = 2,
 	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK
 		  | RCAR_DU_FEATURE_EXT_CTRL_REGS,
-	.num_crtcs = 2,
+	.channels_mask = BIT(1) | BIT(0),
 	.routes = {
 		/*
 		 * R8A7794 has two RGB outputs and one (currently unsupported)
@@ -191,7 +191,7 @@ static const struct rcar_du_device_info rcar_du_r8a7795_info = {
 	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK
 		  | RCAR_DU_FEATURE_EXT_CTRL_REGS
 		  | RCAR_DU_FEATURE_VSP1_SOURCE,
-	.num_crtcs = 4,
+	.channels_mask = BIT(3) | BIT(2) | BIT(1) | BIT(0),
 	.routes = {
 		/*
 		 * R8A7795 has one RGB output, two HDMI outputs and one
@@ -215,7 +215,7 @@ static const struct rcar_du_device_info rcar_du_r8a7795_info = {
 		},
 	},
 	.num_lvds = 1,
-	.dpll_ch =  BIT(1) | BIT(2),
+	.dpll_ch =  BIT(2) | BIT(1),
 };
 
 static const struct rcar_du_device_info rcar_du_r8a7796_info = {
@@ -223,7 +223,7 @@ static const struct rcar_du_device_info rcar_du_r8a7796_info = {
 	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK
 		  | RCAR_DU_FEATURE_EXT_CTRL_REGS
 		  | RCAR_DU_FEATURE_VSP1_SOURCE,
-	.num_crtcs = 3,
+	.channels_mask = BIT(2) | BIT(1) | BIT(0),
 	.routes = {
 		/*
 		 * R8A7796 has one RGB output, one LVDS output and one HDMI
@@ -251,7 +251,7 @@ static const struct rcar_du_device_info rcar_du_r8a77970_info = {
 	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK
 		  | RCAR_DU_FEATURE_EXT_CTRL_REGS
 		  | RCAR_DU_FEATURE_VSP1_SOURCE,
-	.num_crtcs = 1,
+	.channels_mask = BIT(0),
 	.routes = {
 		/* R8A77970 has one RGB output and one LVDS output. */
 		[RCAR_DU_OUTPUT_DPAD0] = {
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_drv.h b/drivers/gpu/drm/rcar-du/rcar_du_drv.h
index 5c7ec15..5385bb5 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_drv.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_drv.h
@@ -52,7 +52,7 @@ struct rcar_du_output_routing {
  * @gen: device generation (2 or 3)
  * @features: device features (RCAR_DU_FEATURE_*)
  * @quirks: device quirks (RCAR_DU_QUIRK_*)
- * @num_crtcs: total number of CRTCs
+ * @channels_mask: bit mask of available DU channels
  * @routes: array of CRTC to output routes, indexed by output (RCAR_DU_OUTPUT_*)
  * @num_lvds: number of internal LVDS encoders
  */
@@ -60,7 +60,7 @@ struct rcar_du_device_info {
 	unsigned int gen;
 	unsigned int features;
 	unsigned int quirks;
-	unsigned int num_crtcs;
+	unsigned int channels_mask;
 	struct rcar_du_output_routing routes[RCAR_DU_OUTPUT_MAX];
 	unsigned int num_lvds;
 	unsigned int dpll_ch;
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_kms.c b/drivers/gpu/drm/rcar-du/rcar_du_kms.c
index 0c8b7e5..b5e331c 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_kms.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_kms.c
@@ -520,6 +520,8 @@ int rcar_du_modeset_init(struct rcar_du_device *rcdu)
 	struct drm_fbdev_cma *fbdev;
 	unsigned int num_encoders;
 	unsigned int num_groups;
+	unsigned int swindex;
+	unsigned int hwindex;
 	unsigned int i;
 	int ret;
 
@@ -532,7 +534,7 @@ int rcar_du_modeset_init(struct rcar_du_device *rcdu)
 	dev->mode_config.funcs = &rcar_du_mode_config_funcs;
 	dev->mode_config.helper_private = &rcar_du_mode_config_helper;
 
-	rcdu->num_crtcs = rcdu->info->num_crtcs;
+	rcdu->num_crtcs = hweight8(rcdu->info->channels_mask);
 
 	ret = rcar_du_properties_init(rcdu);
 	if (ret < 0)
@@ -542,7 +544,7 @@ int rcar_du_modeset_init(struct rcar_du_device *rcdu)
 	 * Initialize vertical blanking interrupts handling. Start with vblank
 	 * disabled for all CRTCs.
 	 */
-	ret = drm_vblank_init(dev, (1 << rcdu->info->num_crtcs) - 1);
+	ret = drm_vblank_init(dev, (1 << rcdu->num_crtcs) - 1);
 	if (ret < 0)
 		return ret;
 
@@ -584,10 +586,16 @@ int rcar_du_modeset_init(struct rcar_du_device *rcdu)
 	}
 
 	/* Create the CRTCs. */
-	for (i = 0; i < rcdu->num_crtcs; ++i) {
-		struct rcar_du_group *rgrp = &rcdu->groups[i / 2];
+	for (swindex = 0, hwindex = 0; swindex < rcdu->num_crtcs; ++hwindex) {
+		struct rcar_du_group *rgrp;
+
+		/* Skip unpopulated DU channels. */
+		if (!(rcdu->info->channels_mask & BIT(hwindex)))
+			continue;
+
+		rgrp = &rcdu->groups[hwindex / 2];
 
-		ret = rcar_du_crtc_create(rgrp, i);
+		ret = rcar_du_crtc_create(rgrp, swindex++, hwindex);
 		if (ret < 0)
 			return ret;
 	}
-- 
cgit v1.1


From 7ae90455bc865ab1c30fb4db53ac56ec32741ab9 Mon Sep 17 00:00:00 2001
From: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Date: Fri, 27 Apr 2018 23:21:53 +0100
Subject: drm: rcar-du: Allow DU groups to work with hardware indexing

The group objects assume linear indexing, and more so always assume that
channel 0 of any active group is used.

Now that the CRTC objects support non-linear indexing, adapt the groups
to remove assumptions that channel 0 is utilised in each group by using
the channel mask provided in the device structures.

Finally ensure that the RGB routing is determined from the index of the
CRTC object (which represents the hardware DU channel index).

Signed-off-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
---
 drivers/gpu/drm/rcar-du/rcar_du_group.c | 14 +++++++++-----
 drivers/gpu/drm/rcar-du/rcar_du_group.h |  2 ++
 drivers/gpu/drm/rcar-du/rcar_du_kms.c   |  5 ++++-
 3 files changed, 15 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/rcar-du/rcar_du_group.c b/drivers/gpu/drm/rcar-du/rcar_du_group.c
index eead202..d539cb2 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_group.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_group.c
@@ -46,9 +46,12 @@ void rcar_du_group_write(struct rcar_du_group *rgrp, u32 reg, u32 data)
 
 static void rcar_du_group_setup_pins(struct rcar_du_group *rgrp)
 {
-	u32 defr6 = DEFR6_CODE | DEFR6_ODPM02_DISP;
+	u32 defr6 = DEFR6_CODE;
 
-	if (rgrp->num_crtcs > 1)
+	if (rgrp->channels_mask & BIT(0))
+		defr6 |= DEFR6_ODPM02_DISP;
+
+	if (rgrp->channels_mask & BIT(1))
 		defr6 |= DEFR6_ODPM12_DISP;
 
 	rcar_du_group_write(rgrp, DEFR6, defr6);
@@ -80,10 +83,11 @@ static void rcar_du_group_setup_defr8(struct rcar_du_group *rgrp)
 		 * On Gen3 VSPD routing can't be configured, but DPAD routing
 		 * needs to be set despite having a single option available.
 		 */
-		u32 crtc = ffs(possible_crtcs) - 1;
+		unsigned int rgb_crtc = ffs(possible_crtcs) - 1;
+		struct rcar_du_crtc *crtc = &rcdu->crtcs[rgb_crtc];
 
-		if (crtc / 2 == rgrp->index)
-			defr8 |= DEFR8_DRGBS_DU(crtc);
+		if (crtc->index / 2 == rgrp->index)
+			defr8 |= DEFR8_DRGBS_DU(crtc->index);
 	}
 
 	rcar_du_group_write(rgrp, DEFR8, defr8);
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_group.h b/drivers/gpu/drm/rcar-du/rcar_du_group.h
index 5e3adc6..42105ae 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_group.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_group.h
@@ -25,6 +25,7 @@ struct rcar_du_device;
  * @dev: the DU device
  * @mmio_offset: registers offset in the device memory map
  * @index: group index
+ * @channels_mask: bitmask of populated DU channels in this group
  * @num_crtcs: number of CRTCs in this group (1 or 2)
  * @use_count: number of users of the group (rcar_du_group_(get|put))
  * @used_crtcs: number of CRTCs currently in use
@@ -39,6 +40,7 @@ struct rcar_du_group {
 	unsigned int mmio_offset;
 	unsigned int index;
 
+	unsigned int channels_mask;
 	unsigned int num_crtcs;
 	unsigned int use_count;
 	unsigned int used_crtcs;
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_kms.c b/drivers/gpu/drm/rcar-du/rcar_du_kms.c
index b5e331c..34f1c27 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_kms.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_kms.c
@@ -559,7 +559,10 @@ int rcar_du_modeset_init(struct rcar_du_device *rcdu)
 		rgrp->dev = rcdu;
 		rgrp->mmio_offset = mmio_offsets[i];
 		rgrp->index = i;
-		rgrp->num_crtcs = min(rcdu->num_crtcs - 2 * i, 2U);
+		/* Extract the channel mask for this group only. */
+		rgrp->channels_mask = (rcdu->info->channels_mask >> (2 * i))
+				   & GENMASK(1, 0);
+		rgrp->num_crtcs = hweight8(rgrp->channels_mask);
 
 		/*
 		 * If we have more than one CRTCs in this group pre-associate
-- 
cgit v1.1


From f1e9a22ac3cff749077f40bf1a149aaaf587ae2d Mon Sep 17 00:00:00 2001
From: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Date: Fri, 27 Apr 2018 23:21:54 +0100
Subject: drm: rcar-du: Add R8A77965 support

The R8A77965 (M3-N) SoC provides RGB, HDMI and LVDS output.

This platform is unusual in that the RGB is connected to DU3 leaving DU2
unpopulated. This is reflected by the channels_mask accordingly.

Signed-off-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
---
 drivers/gpu/drm/rcar-du/rcar_du_drv.c | 29 +++++++++++++++++++++++++++++
 1 file changed, 29 insertions(+)

diff --git a/drivers/gpu/drm/rcar-du/rcar_du_drv.c b/drivers/gpu/drm/rcar-du/rcar_du_drv.c
index 2aa392b..02aee6c 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_drv.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_drv.c
@@ -246,6 +246,34 @@ static const struct rcar_du_device_info rcar_du_r8a7796_info = {
 	.dpll_ch =  BIT(1),
 };
 
+static const struct rcar_du_device_info rcar_du_r8a77965_info = {
+	.gen = 3,
+	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK
+		  | RCAR_DU_FEATURE_EXT_CTRL_REGS
+		  | RCAR_DU_FEATURE_VSP1_SOURCE,
+	.channels_mask = BIT(3) | BIT(1) | BIT(0),
+	.routes = {
+		/*
+		 * R8A77965 has one RGB output, one LVDS output and one HDMI
+		 * output.
+		 */
+		[RCAR_DU_OUTPUT_DPAD0] = {
+			.possible_crtcs = BIT(2),
+			.port = 0,
+		},
+		[RCAR_DU_OUTPUT_HDMI0] = {
+			.possible_crtcs = BIT(1),
+			.port = 1,
+		},
+		[RCAR_DU_OUTPUT_LVDS0] = {
+			.possible_crtcs = BIT(0),
+			.port = 2,
+		},
+	},
+	.num_lvds = 1,
+	.dpll_ch =  BIT(1),
+};
+
 static const struct rcar_du_device_info rcar_du_r8a77970_info = {
 	.gen = 3,
 	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK
@@ -277,6 +305,7 @@ static const struct of_device_id rcar_du_of_table[] = {
 	{ .compatible = "renesas,du-r8a7794", .data = &rcar_du_r8a7794_info },
 	{ .compatible = "renesas,du-r8a7795", .data = &rcar_du_r8a7795_info },
 	{ .compatible = "renesas,du-r8a7796", .data = &rcar_du_r8a7796_info },
+	{ .compatible = "renesas,du-r8a77965", .data = &rcar_du_r8a77965_info },
 	{ .compatible = "renesas,du-r8a77970", .data = &rcar_du_r8a77970_info },
 	{ }
 };
-- 
cgit v1.1


From b06078de418d6f77c81aa74516f787663f51a262 Mon Sep 17 00:00:00 2001
From: Emre Ucan <eucan@de.adit-jv.com>
Date: Mon, 30 Apr 2018 14:02:04 +0200
Subject: drm: rcar-du: Track dma-buf fences

We have to check dma-buf reservation objects of our framebuffers before
we use them. Otherwise, another driver might be writing on the same
buffer which we are using. This would cause visible tearing effects
on display.

We can use existing atomic helper functions to solve this problem.

Signed-off-by: Emre Ucan <eucan@de.adit-jv.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Reviewed-by: Daniel Vetter <daniel@ffwll.ch>
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
---
 drivers/gpu/drm/rcar-du/rcar_du_vsp.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/gpu/drm/rcar-du/rcar_du_vsp.c b/drivers/gpu/drm/rcar-du/rcar_du_vsp.c
index 4a01a99..3badf02 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_vsp.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_vsp.c
@@ -17,6 +17,7 @@
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_fb_cma_helper.h>
 #include <drm/drm_gem_cma_helper.h>
+#include <drm/drm_gem_framebuffer_helper.h>
 #include <drm/drm_plane_helper.h>
 
 #include <linux/bitops.h>
@@ -237,6 +238,10 @@ static int rcar_du_vsp_plane_prepare_fb(struct drm_plane *plane,
 		}
 	}
 
+	ret = drm_gem_fb_prepare_fb(plane, state);
+	if (ret)
+		goto fail;
+
 	return 0;
 
 fail:
-- 
cgit v1.1


From 6cb465972c4eb6741b3094a58a65e527fc63c100 Mon Sep 17 00:00:00 2001
From: Rich Felker <dalias@libc.org>
Date: Sat, 5 May 2018 16:40:23 -0400
Subject: sh: fix build failure for J2 cpu with SMP disabled

The sh asm/smp.h defines a fallback hard_smp_processor_id macro for
the !SMP case, but linux/smp.h never includes asm/smp.h in the !SMP
case.

Signed-off-by: Rich Felker <dalias@libc.org>
---
 arch/sh/kernel/cpu/sh2/probe.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/arch/sh/kernel/cpu/sh2/probe.c b/arch/sh/kernel/cpu/sh2/probe.c
index 4205f6d..a5bd036 100644
--- a/arch/sh/kernel/cpu/sh2/probe.c
+++ b/arch/sh/kernel/cpu/sh2/probe.c
@@ -43,7 +43,11 @@ void __ref cpu_probe(void)
 #endif
 
 #if defined(CONFIG_CPU_J2)
+#if defined(CONFIG_SMP)
 	unsigned cpu = hard_smp_processor_id();
+#else
+	unsigned cpu = 0;
+#endif
 	if (cpu == 0) of_scan_flat_dt(scan_cache, NULL);
 	if (j2_ccr_base) __raw_writel(0x80000303, j2_ccr_base + 4*cpu);
 	if (cpu != 0) return;
-- 
cgit v1.1


From b9826a4929bbd4dcb245429fd7434145c4fcdc9b Mon Sep 17 00:00:00 2001
From: Jacopo Mondi <jacopo+renesas@jmondi.org>
Date: Wed, 18 Apr 2018 11:42:35 +0200
Subject: sh: mm: Fix unprotected access to struct device

With commit ce88313069c36eef80f21fd7 ("arch/sh: make the DMA mapping
operations observe dev->dma_pfn_offset") the generic DMA allocation
function on which the SH 'dma_alloc_coherent()' function relies on,
accesses the 'dma_pfn_offset' field of struct device.

Unfortunately the 'dma_generic_alloc_coherent()' function is called from
several places with a NULL struct device argument, halting the CPU
during the boot process.

This patch fixes the issue by protecting access to dev->dma_pfn_offset,
with a trivial check for validity. It also passes a valid 'struct device'
in the 'platform_resource_setup_memory()' function which is the main user
of 'dma_alloc_coherent()', and inserts a WARN_ON() check to remind to future
(and existing) bogus users of this function to provide a valid 'struct device'
whenever possible.

Fixes: ce88313069c36eef80f21fd7 ("arch/sh: make the DMA mapping operations observe dev->dma_pfn_offset")
Signed-off-by: Jacopo Mondi <jacopo+renesas@jmondi.org>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Rich Felker <dalias@libc.org>
---
 arch/sh/mm/consistent.c | 11 ++++++++---
 1 file changed, 8 insertions(+), 3 deletions(-)

diff --git a/arch/sh/mm/consistent.c b/arch/sh/mm/consistent.c
index 8ce9869..f1b4469 100644
--- a/arch/sh/mm/consistent.c
+++ b/arch/sh/mm/consistent.c
@@ -59,7 +59,9 @@ void *dma_generic_alloc_coherent(struct device *dev, size_t size,
 
 	split_page(pfn_to_page(virt_to_phys(ret) >> PAGE_SHIFT), order);
 
-	*dma_handle = virt_to_phys(ret) - PFN_PHYS(dev->dma_pfn_offset);
+	*dma_handle = virt_to_phys(ret);
+	if (!WARN_ON(!dev))
+		*dma_handle -= PFN_PHYS(dev->dma_pfn_offset);
 
 	return ret_nocache;
 }
@@ -69,9 +71,12 @@ void dma_generic_free_coherent(struct device *dev, size_t size,
 			       unsigned long attrs)
 {
 	int order = get_order(size);
-	unsigned long pfn = (dma_handle >> PAGE_SHIFT) + dev->dma_pfn_offset;
+	unsigned long pfn = dma_handle >> PAGE_SHIFT;
 	int k;
 
+	if (!WARN_ON(!dev))
+		pfn += dev->dma_pfn_offset;
+
 	for (k = 0; k < (1 << order); k++)
 		__free_pages(pfn_to_page(pfn + k), 0);
 
@@ -143,7 +148,7 @@ int __init platform_resource_setup_memory(struct platform_device *pdev,
 	if (!memsize)
 		return 0;
 
-	buf = dma_alloc_coherent(NULL, memsize, &dma_handle, GFP_KERNEL);
+	buf = dma_alloc_coherent(&pdev->dev, memsize, &dma_handle, GFP_KERNEL);
 	if (!buf) {
 		pr_warning("%s: unable to allocate memory\n", name);
 		return -ENOMEM;
-- 
cgit v1.1


From ecf08dad723d3e000aecff6c396f54772d124733 Mon Sep 17 00:00:00 2001
From: Anthoine Bourgeois <anthoine.bourgeois@blade-group.com>
Date: Sun, 29 Apr 2018 22:05:58 +0000
Subject: KVM: x86: remove APIC Timer periodic/oneshot spikes
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Since the commit "8003c9ae204e: add APIC Timer periodic/oneshot mode VMX
preemption timer support", a Windows 10 guest has some erratic timer
spikes.

Here the results on a 150000 times 1ms timer without any load:
	  Before 8003c9ae204e | After 8003c9ae204e
Max           1834us          |  86000us
Mean          1100us          |   1021us
Deviation       59us          |    149us
Here the results on a 150000 times 1ms timer with a cpu-z stress test:
	  Before 8003c9ae204e | After 8003c9ae204e
Max          32000us          | 140000us
Mean          1006us          |   1997us
Deviation      140us          |  11095us

The root cause of the problem is starting hrtimer with an expiry time
already in the past can take more than 20 milliseconds to trigger the
timer function.  It can be solved by forward such past timers
immediately, rather than submitting them to hrtimer_start().
In case the timer is periodic, update the target expiration and call
hrtimer_start with it.

v2: Check if the tsc deadline is already expired. Thank you Mika.
v3: Execute the past timers immediately rather than submitting them to
hrtimer_start().
v4: Rearm the periodic timer with advance_periodic_target_expiration() a
simpler version of set_target_expiration(). Thank you Paolo.

Cc: Mika Penttilä <mika.penttila@nextfour.com>
Cc: Wanpeng Li <kernellwp@gmail.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: stable@vger.kernel.org
Signed-off-by: Anthoine Bourgeois <anthoine.bourgeois@blade-group.com>
8003c9ae204e ("KVM: LAPIC: add APIC Timer periodic/oneshot mode VMX preemption timer support")
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
---
 arch/x86/kvm/lapic.c | 37 ++++++++++++++++++++-----------------
 1 file changed, 20 insertions(+), 17 deletions(-)

diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
index 70dcb55..b74c9c1 100644
--- a/arch/x86/kvm/lapic.c
+++ b/arch/x86/kvm/lapic.c
@@ -1463,23 +1463,6 @@ static void start_sw_tscdeadline(struct kvm_lapic *apic)
 	local_irq_restore(flags);
 }
 
-static void start_sw_period(struct kvm_lapic *apic)
-{
-	if (!apic->lapic_timer.period)
-		return;
-
-	if (apic_lvtt_oneshot(apic) &&
-	    ktime_after(ktime_get(),
-			apic->lapic_timer.target_expiration)) {
-		apic_timer_expired(apic);
-		return;
-	}
-
-	hrtimer_start(&apic->lapic_timer.timer,
-		apic->lapic_timer.target_expiration,
-		HRTIMER_MODE_ABS_PINNED);
-}
-
 static void update_target_expiration(struct kvm_lapic *apic, uint32_t old_divisor)
 {
 	ktime_t now, remaining;
@@ -1546,6 +1529,26 @@ static void advance_periodic_target_expiration(struct kvm_lapic *apic)
 				apic->lapic_timer.period);
 }
 
+static void start_sw_period(struct kvm_lapic *apic)
+{
+	if (!apic->lapic_timer.period)
+		return;
+
+	if (ktime_after(ktime_get(),
+			apic->lapic_timer.target_expiration)) {
+		apic_timer_expired(apic);
+
+		if (apic_lvtt_oneshot(apic))
+			return;
+
+		advance_periodic_target_expiration(apic);
+	}
+
+	hrtimer_start(&apic->lapic_timer.timer,
+		apic->lapic_timer.target_expiration,
+		HRTIMER_MODE_ABS_PINNED);
+}
+
 bool kvm_lapic_hv_timer_in_use(struct kvm_vcpu *vcpu)
 {
 	if (!lapic_in_kernel(vcpu))
-- 
cgit v1.1


From 75bc37fefc4471e718ba8e651aa74673d4e0a9eb Mon Sep 17 00:00:00 2001
From: Linus Torvalds <torvalds@linux-foundation.org>
Date: Sun, 6 May 2018 16:57:38 -1000
Subject: Linux 4.17-rc4

---
 Makefile | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/Makefile b/Makefile
index 619a85a..d0d2652 100644
--- a/Makefile
+++ b/Makefile
@@ -2,8 +2,8 @@
 VERSION = 4
 PATCHLEVEL = 17
 SUBLEVEL = 0
-EXTRAVERSION = -rc3
-NAME = Fearless Coyote
+EXTRAVERSION = -rc4
+NAME = Merciless Moray
 
 # *DOCUMENTATION*
 # To see a list of typical targets execute "make help"
-- 
cgit v1.1


From c4ec1f0353b342473b93637fd0c3fb524bedbb2d Mon Sep 17 00:00:00 2001
From: Michael Ellerman <mpe@ellerman.id.au>
Date: Wed, 2 May 2018 16:57:54 +1000
Subject: powerpc/64: Remove unused paca->soft_enabled

In commit 4e26bc4a4ed6 ("powerpc/64: Rename soft_enabled to
irq_soft_mask") we renamed paca->soft_enabled. But then in commit
8e0b634b1327 ("powerpc/64s: Do not allocate lppaca if we are not
virtualized") we added it back. Oops. This happened because the two
patches were in flight at the same time and rebased vs each other
multiple times, and we missed it in review.

Fixes: 8e0b634b1327 ("powerpc/64s: Do not allocate lppaca if we are not virtualized")
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
---
 arch/powerpc/include/asm/paca.h | 1 -
 1 file changed, 1 deletion(-)

diff --git a/arch/powerpc/include/asm/paca.h b/arch/powerpc/include/asm/paca.h
index 4185f1c..3f109a3 100644
--- a/arch/powerpc/include/asm/paca.h
+++ b/arch/powerpc/include/asm/paca.h
@@ -165,7 +165,6 @@ struct paca_struct {
 	u64 saved_msr;			/* MSR saved here by enter_rtas */
 	u16 trap_save;			/* Used when bad stack is encountered */
 	u8 irq_soft_mask;		/* mask for irq soft masking */
-	u8 soft_enabled;		/* irq soft-enable flag */
 	u8 irq_happened;		/* irq happened while soft-disabled */
 	u8 io_sync;			/* writel() needs spin_unlock sync */
 	u8 irq_work_pending;		/* IRQ_WORK interrupt while soft-disable */
-- 
cgit v1.1


From 0b7758aaf6543b9a10c8671db559e9d374a3fd95 Mon Sep 17 00:00:00 2001
From: "Naveen N. Rao" <naveen.n.rao@linux.vnet.ibm.com>
Date: Fri, 4 May 2018 18:44:24 +0530
Subject: powerpc/trace/syscalls: Update syscall name matching logic

On powerpc64 ABIv1, we are enabling syscall tracing for only ~20
syscalls. This is due to commit e145242ea0df6 ("syscalls/core,
syscalls/x86: Clean up syscall stub naming convention") which has
changed the syscall entry wrapper prefix from "SyS" to "__se_sys".

Update the logic for ABIv1 to not just skip the initial dot, but also
the "__se_sys" prefix.

Fixes: commit e145242ea0df6 ("syscalls/core, syscalls/x86: Clean up syscall stub naming convention")
Reported-by: Michael Ellerman <mpe@ellerman.id.au>
Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
---
 arch/powerpc/include/asm/ftrace.h | 10 +++-------
 1 file changed, 3 insertions(+), 7 deletions(-)

diff --git a/arch/powerpc/include/asm/ftrace.h b/arch/powerpc/include/asm/ftrace.h
index 9abddde..24103fa 100644
--- a/arch/powerpc/include/asm/ftrace.h
+++ b/arch/powerpc/include/asm/ftrace.h
@@ -73,13 +73,9 @@ struct dyn_arch_ftrace {
 #define ARCH_HAS_SYSCALL_MATCH_SYM_NAME
 static inline bool arch_syscall_match_sym_name(const char *sym, const char *name)
 {
-	/*
-	 * Compare the symbol name with the system call name. Skip the .sys or .SyS
-	 * prefix from the symbol name and the sys prefix from the system call name and
-	 * just match the rest. This is only needed on ppc64 since symbol names on
-	 * 32bit do not start with a period so the generic function will work.
-	 */
-	return !strcmp(sym + 4, name + 3);
+	/* We need to skip past the initial dot, and the __se_sys alias */
+	return !strcmp(sym + 1, name) ||
+		(!strncmp(sym, ".__se_sys", 9) && !strcmp(sym + 6, name));
 }
 #endif
 #endif /* CONFIG_FTRACE_SYSCALLS && !__ASSEMBLY__ */
-- 
cgit v1.1


From edf6a2dfe3889daf97e7c164891a87832169e3e4 Mon Sep 17 00:00:00 2001
From: "Naveen N. Rao" <naveen.n.rao@linux.vnet.ibm.com>
Date: Fri, 4 May 2018 18:44:25 +0530
Subject: powerpc/trace/syscalls: Update syscall name matching logic to account
 for ppc_ prefix

Some syscall entry functions on powerpc are prefixed with
ppc_/ppc32_/ppc64_ rather than the usual sys_/__se_sys prefix. fork(),
clone(), swapcontext() are some examples of syscalls with such entry
points. We need to match against these names when initializing ftrace
syscall tracing.

Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
---
 arch/powerpc/include/asm/ftrace.h | 21 +++++++++++++++++++--
 1 file changed, 19 insertions(+), 2 deletions(-)

diff --git a/arch/powerpc/include/asm/ftrace.h b/arch/powerpc/include/asm/ftrace.h
index 24103fa..b2dabd0 100644
--- a/arch/powerpc/include/asm/ftrace.h
+++ b/arch/powerpc/include/asm/ftrace.h
@@ -69,13 +69,30 @@ struct dyn_arch_ftrace {
 #endif
 
 #if defined(CONFIG_FTRACE_SYSCALLS) && !defined(__ASSEMBLY__)
-#ifdef PPC64_ELF_ABI_v1
+/*
+ * Some syscall entry functions on powerpc start with "ppc_" (fork and clone,
+ * for instance) or ppc32_/ppc64_. We should also match the sys_ variant with
+ * those.
+ */
 #define ARCH_HAS_SYSCALL_MATCH_SYM_NAME
+#ifdef PPC64_ELF_ABI_v1
 static inline bool arch_syscall_match_sym_name(const char *sym, const char *name)
 {
 	/* We need to skip past the initial dot, and the __se_sys alias */
 	return !strcmp(sym + 1, name) ||
-		(!strncmp(sym, ".__se_sys", 9) && !strcmp(sym + 6, name));
+		(!strncmp(sym, ".__se_sys", 9) && !strcmp(sym + 6, name)) ||
+		(!strncmp(sym, ".ppc_", 5) && !strcmp(sym + 5, name + 4)) ||
+		(!strncmp(sym, ".ppc32_", 7) && !strcmp(sym + 7, name + 4)) ||
+		(!strncmp(sym, ".ppc64_", 7) && !strcmp(sym + 7, name + 4));
+}
+#else
+static inline bool arch_syscall_match_sym_name(const char *sym, const char *name)
+{
+	return !strcmp(sym, name) ||
+		(!strncmp(sym, "__se_sys", 8) && !strcmp(sym + 5, name)) ||
+		(!strncmp(sym, "ppc_", 4) && !strcmp(sym + 4, name + 4)) ||
+		(!strncmp(sym, "ppc32_", 6) && !strcmp(sym + 6, name + 4)) ||
+		(!strncmp(sym, "ppc64_", 6) && !strcmp(sym + 6, name + 4));
 }
 #endif
 #endif /* CONFIG_FTRACE_SYSCALLS && !__ASSEMBLY__ */
-- 
cgit v1.1


From e8f90c74e6ab64824f3a21521640de7b21050b9d Mon Sep 17 00:00:00 2001
From: YueHaibing <yuehaibing@huawei.com>
Date: Tue, 24 Apr 2018 11:08:35 +0800
Subject: mac80211_hwsim: fix a possible memory leak in hwsim_new_radio_nl()

'hwname' should be freed before leaving from the error handling cases,
otherwise it will cause mem leak

Fixes: cb1a5bae5684 ("mac80211_hwsim: add permanent mac address option for new radios")
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
---
 drivers/net/wireless/mac80211_hwsim.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/net/wireless/mac80211_hwsim.c b/drivers/net/wireless/mac80211_hwsim.c
index 96d26cf..4a017a0 100644
--- a/drivers/net/wireless/mac80211_hwsim.c
+++ b/drivers/net/wireless/mac80211_hwsim.c
@@ -3236,6 +3236,7 @@ static int hwsim_new_radio_nl(struct sk_buff *msg, struct genl_info *info)
 			GENL_SET_ERR_MSG(info,"MAC is no valid source addr");
 			NL_SET_BAD_ATTR(info->extack,
 					info->attrs[HWSIM_ATTR_PERM_ADDR]);
+			kfree(hwname);
 			return -EINVAL;
 		}
 
-- 
cgit v1.1


From 4bf01ca21e2e0e4561d1a03c48c3d740418702db Mon Sep 17 00:00:00 2001
From: Johan Hovold <johan@kernel.org>
Date: Thu, 26 Apr 2018 09:31:52 +0200
Subject: rfkill: gpio: fix memory leak in probe error path

Make sure to free the rfkill device in case registration fails during
probe.

Fixes: 5e7ca3937fbe ("net: rfkill: gpio: convert to resource managed allocation")
Cc: stable <stable@vger.kernel.org>	# 3.13
Cc: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Signed-off-by: Johan Hovold <johan@kernel.org>
Reviewed-by: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
---
 net/rfkill/rfkill-gpio.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/net/rfkill/rfkill-gpio.c b/net/rfkill/rfkill-gpio.c
index 41bd496..00192a9 100644
--- a/net/rfkill/rfkill-gpio.c
+++ b/net/rfkill/rfkill-gpio.c
@@ -137,13 +137,18 @@ static int rfkill_gpio_probe(struct platform_device *pdev)
 
 	ret = rfkill_register(rfkill->rfkill_dev);
 	if (ret < 0)
-		return ret;
+		goto err_destroy;
 
 	platform_set_drvdata(pdev, rfkill);
 
 	dev_info(&pdev->dev, "%s device registered.\n", rfkill->name);
 
 	return 0;
+
+err_destroy:
+	rfkill_destroy(rfkill->rfkill_dev);
+
+	return ret;
 }
 
 static int rfkill_gpio_remove(struct platform_device *pdev)
-- 
cgit v1.1


From d1361b32e6aec7440c01d5c8fcc54189930a342d Mon Sep 17 00:00:00 2001
From: Randy Dunlap <rdunlap@infradead.org>
Date: Thu, 26 Apr 2018 18:17:31 -0700
Subject: mac80211: fix kernel-doc "bad line" warning

Fix 88 instances of a kernel-doc warning:
  ../include/net/mac80211.h:2083: warning: bad line:  >

Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Cc: linux-wireless@vger.kernel.org
Cc: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
---
 include/net/mac80211.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/include/net/mac80211.h b/include/net/mac80211.h
index d2279b2..b2f3a0c 100644
--- a/include/net/mac80211.h
+++ b/include/net/mac80211.h
@@ -2080,7 +2080,7 @@ struct ieee80211_txq {
  *	virtual interface might not be given air time for the transmission of
  *	the frame, as it is not synced with the AP/P2P GO yet, and thus the
  *	deauthentication frame might not be transmitted.
- >
+ *
  * @IEEE80211_HW_DOESNT_SUPPORT_QOS_NDP: The driver (or firmware) doesn't
  *	support QoS NDP for AP probing - that's most likely a driver bug.
  *
-- 
cgit v1.1


From 407879b690ba3a6bf29be896d02dad63463bd1c0 Mon Sep 17 00:00:00 2001
From: Ilan Peer <ilan.peer@intel.com>
Date: Fri, 20 Apr 2018 13:49:20 +0300
Subject: mac80211: Adjust SAE authentication timeout

The IEEE P802.11-REVmd D1.0 specification updated the SAE authentication
timeout to be 2000 milliseconds (see dot11RSNASAERetransPeriod). Update
the SAE timeout setting accordingly.

While at it, reduce some code duplication in the timeout configuration.

Signed-off-by: Ilan Peer <ilan.peer@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
---
 net/mac80211/mlme.c | 25 ++++++++++++++++++-------
 1 file changed, 18 insertions(+), 7 deletions(-)

diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c
index 6fe72ef..2330687 100644
--- a/net/mac80211/mlme.c
+++ b/net/mac80211/mlme.c
@@ -36,6 +36,7 @@
 #define IEEE80211_AUTH_TIMEOUT		(HZ / 5)
 #define IEEE80211_AUTH_TIMEOUT_LONG	(HZ / 2)
 #define IEEE80211_AUTH_TIMEOUT_SHORT	(HZ / 10)
+#define IEEE80211_AUTH_TIMEOUT_SAE	(HZ * 2)
 #define IEEE80211_AUTH_MAX_TRIES	3
 #define IEEE80211_AUTH_WAIT_ASSOC	(HZ * 5)
 #define IEEE80211_ASSOC_TIMEOUT		(HZ / 5)
@@ -3814,16 +3815,19 @@ static int ieee80211_auth(struct ieee80211_sub_if_data *sdata)
 			    tx_flags);
 
 	if (tx_flags == 0) {
-		auth_data->timeout = jiffies + IEEE80211_AUTH_TIMEOUT;
-		auth_data->timeout_started = true;
-		run_again(sdata, auth_data->timeout);
+		if (auth_data->algorithm == WLAN_AUTH_SAE)
+			auth_data->timeout = jiffies +
+				IEEE80211_AUTH_TIMEOUT_SAE;
+		else
+			auth_data->timeout = jiffies + IEEE80211_AUTH_TIMEOUT;
 	} else {
 		auth_data->timeout =
 			round_jiffies_up(jiffies + IEEE80211_AUTH_TIMEOUT_LONG);
-		auth_data->timeout_started = true;
-		run_again(sdata, auth_data->timeout);
 	}
 
+	auth_data->timeout_started = true;
+	run_again(sdata, auth_data->timeout);
+
 	return 0;
 }
 
@@ -3894,8 +3898,15 @@ void ieee80211_sta_work(struct ieee80211_sub_if_data *sdata)
 		ifmgd->status_received = false;
 		if (ifmgd->auth_data && ieee80211_is_auth(fc)) {
 			if (status_acked) {
-				ifmgd->auth_data->timeout =
-					jiffies + IEEE80211_AUTH_TIMEOUT_SHORT;
+				if (ifmgd->auth_data->algorithm ==
+				    WLAN_AUTH_SAE)
+					ifmgd->auth_data->timeout =
+						jiffies +
+						IEEE80211_AUTH_TIMEOUT_SAE;
+				else
+					ifmgd->auth_data->timeout =
+						jiffies +
+						IEEE80211_AUTH_TIMEOUT_SHORT;
 				run_again(sdata, ifmgd->auth_data->timeout);
 			} else {
 				ifmgd->auth_data->timeout = jiffies - 1;
-- 
cgit v1.1


From f0b408eebc993310bea3f2daae286c40bd3f063b Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= <ville.syrjala@linux.intel.com>
Date: Wed, 2 May 2018 21:32:47 +0300
Subject: drm/atomic: Clean old_state/new_state in
 drm_atomic_state_default_clear()
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Clear the old_state and new_state pointers for every object in
drm_atomic_state_default_clear(). Otherwise
drm_atomic_get_{new,old}_*_state() will hand out stale pointers to
anyone who hasn't first confirmed that the object is in fact part of
the current atomic transcation, if they are called after we've done
the ww backoff dance while hanging on to the same drm_atomic_state.

For example, handle_conflicting_encoders() looks like it could hit
this since it iterates the full connector list and just calls
drm_atomic_get_new_connector_state() for each.

And I believe we have now witnessed this happening at least once in
i915 check_digital_port_conflicts(). Commit 8b69449d2663 ("drm/i915:
Remove last references to drm_atomic_get_existing* macros") changed
the safe drm_atomic_get_existing_connector_state() to the unsafe
drm_atomic_get_new_connector_state(), which opened the doors for
this particular bug there as well.

v2: Split private objs out to a separate patch (Daniel)

Cc: stable@vger.kernel.org
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Cc: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Cc: Abhay Kumar <abhay.kumar@intel.com>
Fixes: 581e49fe6b41 ("drm/atomic: Add new iterators over all state, v3.")
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180502183247.5746-1-ville.syrjala@linux.intel.com
Reviewed-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
---
 drivers/gpu/drm/drm_atomic.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index 7d25c42..4fa19ed 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -155,6 +155,8 @@ void drm_atomic_state_default_clear(struct drm_atomic_state *state)
 						       state->connectors[i].state);
 		state->connectors[i].ptr = NULL;
 		state->connectors[i].state = NULL;
+		state->connectors[i].old_state = NULL;
+		state->connectors[i].new_state = NULL;
 		drm_connector_put(connector);
 	}
 
@@ -169,6 +171,8 @@ void drm_atomic_state_default_clear(struct drm_atomic_state *state)
 
 		state->crtcs[i].ptr = NULL;
 		state->crtcs[i].state = NULL;
+		state->crtcs[i].old_state = NULL;
+		state->crtcs[i].new_state = NULL;
 	}
 
 	for (i = 0; i < config->num_total_plane; i++) {
@@ -181,6 +185,8 @@ void drm_atomic_state_default_clear(struct drm_atomic_state *state)
 						   state->planes[i].state);
 		state->planes[i].ptr = NULL;
 		state->planes[i].state = NULL;
+		state->planes[i].old_state = NULL;
+		state->planes[i].new_state = NULL;
 	}
 
 	for (i = 0; i < state->num_private_objs; i++) {
-- 
cgit v1.1


From b5cb2e5a1f64d882a155add7522247ab0523051e Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= <ville.syrjala@linux.intel.com>
Date: Wed, 2 May 2018 21:32:47 +0300
Subject: drm/atomic: Clean private obj old_state/new_state in
 drm_atomic_state_default_clear()
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Clear the old_state and new_state pointers for private objects
in drm_atomic_state_default_clear(). We don't actually have
functions to get the new/old state for private objects so
getting access to the potentially stale pointers requires a
bit more manual labour than for other object types. But let's
clear the pointers for private objects as well, if only to
avoid future surprises when someone decides to add the functions
to get at them.

v2: Split private objs to a separate patch (Daniel)

Cc: <stable@vger.kernel.org> # v4.14+
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Cc: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Cc: Abhay Kumar <abhay.kumar@intel.com>
Fixes: a4370c777406 (drm/atomic: Make private objs proper objects)
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180502183247.5746-1-ville.syrjala@linux.intel.com
Reviewed-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
---
 drivers/gpu/drm/drm_atomic.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index 4fa19ed..c825c76 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -196,6 +196,8 @@ void drm_atomic_state_default_clear(struct drm_atomic_state *state)
 						 state->private_objs[i].state);
 		state->private_objs[i].ptr = NULL;
 		state->private_objs[i].state = NULL;
+		state->private_objs[i].old_state = NULL;
+		state->private_objs[i].new_state = NULL;
 	}
 	state->num_private_objs = 0;
 
-- 
cgit v1.1


From 164c2416dd40770aba5814f93da835e8a9f7196d Mon Sep 17 00:00:00 2001
From: Eric Anholt <eric@anholt.net>
Date: Fri, 9 Mar 2018 15:32:56 -0800
Subject: drm/vc4: Fix oops dereferencing DPI's connector since panel_bridge.

In the cleanup, I didn't notice that we needed to dereference the
connector for the bus_format.  Fix the regression by looking up the
first (and only) connector attached to us, and assume that its
bus_format is what we want.  Some day it would be good to have that
part of display_info attached to the bridge, instead.

v2: Fix stray whitespace change

Signed-off-by: Eric Anholt <eric@anholt.net>
Fixes: 7b1298e05310 ("drm/vc4: Switch DPI to using the panel-bridge helper.")
Link: https://patchwork.freedesktop.org/patch/msgid/20180309233256.1667-1-eric@anholt.net
Reviewed-by: Sean Paul <seanpaul@chromium.org>
Reviewed-by: Boris Brezillon <boris.brezillon@bootlin.com>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
---
 drivers/gpu/drm/vc4/vc4_dpi.c | 25 ++++++++++++++++++++++---
 1 file changed, 22 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/vc4/vc4_dpi.c b/drivers/gpu/drm/vc4/vc4_dpi.c
index 72c9dbd..f185812 100644
--- a/drivers/gpu/drm/vc4/vc4_dpi.c
+++ b/drivers/gpu/drm/vc4/vc4_dpi.c
@@ -96,7 +96,6 @@ struct vc4_dpi {
 	struct platform_device *pdev;
 
 	struct drm_encoder *encoder;
-	struct drm_connector *connector;
 
 	void __iomem *regs;
 
@@ -164,14 +163,31 @@ static void vc4_dpi_encoder_disable(struct drm_encoder *encoder)
 
 static void vc4_dpi_encoder_enable(struct drm_encoder *encoder)
 {
+	struct drm_device *dev = encoder->dev;
 	struct drm_display_mode *mode = &encoder->crtc->mode;
 	struct vc4_dpi_encoder *vc4_encoder = to_vc4_dpi_encoder(encoder);
 	struct vc4_dpi *dpi = vc4_encoder->dpi;
+	struct drm_connector_list_iter conn_iter;
+	struct drm_connector *connector = NULL, *connector_scan;
 	u32 dpi_c = DPI_ENABLE | DPI_OUTPUT_ENABLE_MODE;
 	int ret;
 
-	if (dpi->connector->display_info.num_bus_formats) {
-		u32 bus_format = dpi->connector->display_info.bus_formats[0];
+	/* Look up the connector attached to DPI so we can get the
+	 * bus_format.  Ideally the bridge would tell us the
+	 * bus_format we want, but it doesn't yet, so assume that it's
+	 * uniform throughout the bridge chain.
+	 */
+	drm_connector_list_iter_begin(dev, &conn_iter);
+	drm_for_each_connector_iter(connector_scan, &conn_iter) {
+		if (connector_scan->encoder == encoder) {
+			connector = connector_scan;
+			break;
+		}
+	}
+	drm_connector_list_iter_end(&conn_iter);
+
+	if (connector && connector->display_info.num_bus_formats) {
+		u32 bus_format = connector->display_info.bus_formats[0];
 
 		switch (bus_format) {
 		case MEDIA_BUS_FMT_RGB888_1X24:
@@ -199,6 +215,9 @@ static void vc4_dpi_encoder_enable(struct drm_encoder *encoder)
 			DRM_ERROR("Unknown media bus format %d\n", bus_format);
 			break;
 		}
+	} else {
+		/* Default to 24bit if no connector found. */
+		dpi_c |= VC4_SET_FIELD(DPI_FORMAT_24BIT_888_RGB, DPI_FORMAT);
 	}
 
 	if (mode->flags & DRM_MODE_FLAG_NHSYNC)
-- 
cgit v1.1


From 4a9fbfcab19d3f71ad2bf0bcb653c4ee84e69c7f Mon Sep 17 00:00:00 2001
From: Dan Carpenter <dan.carpenter@oracle.com>
Date: Wed, 18 Apr 2018 17:29:37 +0300
Subject: drm/omap: silence unititialized variable warning

Smatch complains that "area_free" could be used without being
initialized.  This code is several years old and premusably works fine
so this can't be a very serious bug.  But it's easy enough to silence
the warning.  If "area_free" is false at the end of the function then
we return -ENOMEM.

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180418142937.GA13828@mwanda
Signed-off-by: Sean Paul <seanpaul@chromium.org>
---
 drivers/gpu/drm/omapdrm/tcm-sita.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/omapdrm/tcm-sita.c b/drivers/gpu/drm/omapdrm/tcm-sita.c
index d7f7bc9..817be3c 100644
--- a/drivers/gpu/drm/omapdrm/tcm-sita.c
+++ b/drivers/gpu/drm/omapdrm/tcm-sita.c
@@ -90,7 +90,7 @@ static int l2r_t2b(u16 w, u16 h, u16 a, s16 offset,
 {
 	int i;
 	unsigned long index;
-	bool area_free;
+	bool area_free = false;
 	unsigned long slots_per_band = PAGE_SIZE / slot_bytes;
 	unsigned long bit_offset = (offset > 0) ? offset / slot_bytes : 0;
 	unsigned long curr_bit = bit_offset;
-- 
cgit v1.1


From 77eeac24b10fc84d3ffd5b11a897dff88dde244d Mon Sep 17 00:00:00 2001
From: Tomi Valkeinen <tomi.valkeinen@ti.com>
Date: Thu, 29 Mar 2018 13:40:36 +0300
Subject: drm/omap: fix uninitialized ret variable

audio_config function for both HDMI4 and HDMI5 return uninitialized
value as the error code if the display is not currently enabled. For
some reason this has not caused any issues.

Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180329104038.29154-1-tomi.valkeinen@ti.com
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
---
 drivers/gpu/drm/omapdrm/dss/hdmi4.c | 2 +-
 drivers/gpu/drm/omapdrm/dss/hdmi5.c | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/omapdrm/dss/hdmi4.c b/drivers/gpu/drm/omapdrm/dss/hdmi4.c
index 97c8886..5879f45 100644
--- a/drivers/gpu/drm/omapdrm/dss/hdmi4.c
+++ b/drivers/gpu/drm/omapdrm/dss/hdmi4.c
@@ -679,7 +679,7 @@ static int hdmi_audio_config(struct device *dev,
 			     struct omap_dss_audio *dss_audio)
 {
 	struct omap_hdmi *hd = dev_get_drvdata(dev);
-	int ret;
+	int ret = 0;
 
 	mutex_lock(&hd->lock);
 
diff --git a/drivers/gpu/drm/omapdrm/dss/hdmi5.c b/drivers/gpu/drm/omapdrm/dss/hdmi5.c
index d28da9a..ae1a001 100644
--- a/drivers/gpu/drm/omapdrm/dss/hdmi5.c
+++ b/drivers/gpu/drm/omapdrm/dss/hdmi5.c
@@ -671,7 +671,7 @@ static int hdmi_audio_config(struct device *dev,
 			     struct omap_dss_audio *dss_audio)
 {
 	struct omap_hdmi *hd = dev_get_drvdata(dev);
-	int ret;
+	int ret = 0;
 
 	mutex_lock(&hd->lock);
 
-- 
cgit v1.1


From 6a0f0c55619f0b82a677cab72e77c3444a5eee58 Mon Sep 17 00:00:00 2001
From: Tomi Valkeinen <tomi.valkeinen@ti.com>
Date: Thu, 29 Mar 2018 13:40:37 +0300
Subject: drm/omap: fix possible NULL ref issue in tiler_reserve_2d

tiler_reserve_2d allocates memory but does not check if it got the
memory. Add the check and return ENOMEM on failure.

Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180329104038.29154-2-tomi.valkeinen@ti.com
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
---
 drivers/gpu/drm/omapdrm/omap_dmm_tiler.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/omapdrm/omap_dmm_tiler.c b/drivers/gpu/drm/omapdrm/omap_dmm_tiler.c
index f9fa1c9..401c02e 100644
--- a/drivers/gpu/drm/omapdrm/omap_dmm_tiler.c
+++ b/drivers/gpu/drm/omapdrm/omap_dmm_tiler.c
@@ -401,12 +401,16 @@ int tiler_unpin(struct tiler_block *block)
 struct tiler_block *tiler_reserve_2d(enum tiler_fmt fmt, u16 w,
 		u16 h, u16 align)
 {
-	struct tiler_block *block = kzalloc(sizeof(*block), GFP_KERNEL);
+	struct tiler_block *block;
 	u32 min_align = 128;
 	int ret;
 	unsigned long flags;
 	u32 slot_bytes;
 
+	block = kzalloc(sizeof(*block), GFP_KERNEL);
+	if (!block)
+		return ERR_PTR(-ENOMEM);
+
 	BUG_ON(!validfmt(fmt));
 
 	/* convert width/height to slots */
-- 
cgit v1.1


From 4d6cb5e2fee52af17001e92950f0894304706ee4 Mon Sep 17 00:00:00 2001
From: Tomi Valkeinen <tomi.valkeinen@ti.com>
Date: Wed, 2 May 2018 12:11:56 +0300
Subject: drm/omap: check return value from soc_device_match

soc_device_match() can return NULL, so add a check and fail if
soc_device_match() fails.

Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180502091159.7071-2-tomi.valkeinen@ti.com
Reviewed-by: Benoit Parrot <bparrot@ti.com>
Reviewed-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
---
 drivers/gpu/drm/omapdrm/dss/hdmi4_core.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/omapdrm/dss/hdmi4_core.c b/drivers/gpu/drm/omapdrm/dss/hdmi4_core.c
index 35ed2ad..813ba42 100644
--- a/drivers/gpu/drm/omapdrm/dss/hdmi4_core.c
+++ b/drivers/gpu/drm/omapdrm/dss/hdmi4_core.c
@@ -922,8 +922,13 @@ int hdmi4_core_init(struct platform_device *pdev, struct hdmi_core_data *core)
 {
 	const struct hdmi4_features *features;
 	struct resource *res;
+	const struct soc_device_attribute *soc;
 
-	features = soc_device_match(hdmi4_soc_devices)->data;
+	soc = soc_device_match(hdmi4_soc_devices);
+	if (!soc)
+		return -ENODEV;
+
+	features = soc->data;
 	core->cts_swmode = features->cts_swmode;
 	core->audio_use_mclk = features->audio_use_mclk;
 
-- 
cgit v1.1


From e1cdab6e5f9a52e0bdf34363ec0fdab0e2328ba9 Mon Sep 17 00:00:00 2001
From: Tomi Valkeinen <tomi.valkeinen@ti.com>
Date: Wed, 2 May 2018 12:11:57 +0300
Subject: drm/omap: handle error if scale coefs are not found

If get_scale_coef functions fail, they return NULL, but we never check
the return value and could do a NULL deref. This should not happen as we
ought to validate the amount of scaling already earlier, but to be safe,
add the necessary check.

Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180502091159.7071-3-tomi.valkeinen@ti.com
Reviewed-by: Benoit Parrot <bparrot@ti.com>
Reviewed-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
---
 drivers/gpu/drm/omapdrm/dss/dispc.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/gpu/drm/omapdrm/dss/dispc.c b/drivers/gpu/drm/omapdrm/dss/dispc.c
index 5e2e65e..b8fdb63 100644
--- a/drivers/gpu/drm/omapdrm/dss/dispc.c
+++ b/drivers/gpu/drm/omapdrm/dss/dispc.c
@@ -828,6 +828,12 @@ static void dispc_ovl_set_scale_coef(struct dispc_device *dispc,
 	h_coef = dispc_ovl_get_scale_coef(fir_hinc, true);
 	v_coef = dispc_ovl_get_scale_coef(fir_vinc, five_taps);
 
+	if (!h_coef || !v_coef) {
+		dev_err(&dispc->pdev->dev, "%s: failed to find scale coefs\n",
+			__func__);
+		return;
+	}
+
 	for (i = 0; i < 8; i++) {
 		u32 h, hv;
 
-- 
cgit v1.1


From 7f26eee572167926806b6fcb703bb5f8e452c398 Mon Sep 17 00:00:00 2001
From: Tomi Valkeinen <tomi.valkeinen@ti.com>
Date: Wed, 2 May 2018 12:11:58 +0300
Subject: drm/omap: add missing linefeeds to prints

A bunch of debug and error prints are missing linefeeds. Add those.

Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180502091159.7071-4-tomi.valkeinen@ti.com
Reviewed-by: Benoit Parrot <bparrot@ti.com>
Reviewed-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
---
 drivers/gpu/drm/omapdrm/dss/dispc.c | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/omapdrm/dss/dispc.c b/drivers/gpu/drm/omapdrm/dss/dispc.c
index b8fdb63..7f3ac6b 100644
--- a/drivers/gpu/drm/omapdrm/dss/dispc.c
+++ b/drivers/gpu/drm/omapdrm/dss/dispc.c
@@ -2348,7 +2348,7 @@ static int dispc_ovl_calc_scaling_24xx(struct dispc_device *dispc,
 	}
 
 	if (in_width > maxsinglelinewidth) {
-		DSSERR("Cannot scale max input width exceeded");
+		DSSERR("Cannot scale max input width exceeded\n");
 		return -EINVAL;
 	}
 	return 0;
@@ -2430,13 +2430,13 @@ again:
 	}
 
 	if (in_width > (maxsinglelinewidth * 2)) {
-		DSSERR("Cannot setup scaling");
-		DSSERR("width exceeds maximum width possible");
+		DSSERR("Cannot setup scaling\n");
+		DSSERR("width exceeds maximum width possible\n");
 		return -EINVAL;
 	}
 
 	if (in_width > maxsinglelinewidth && *five_taps) {
-		DSSERR("cannot setup scaling with five taps");
+		DSSERR("cannot setup scaling with five taps\n");
 		return -EINVAL;
 	}
 	return 0;
@@ -2478,7 +2478,7 @@ static int dispc_ovl_calc_scaling_44xx(struct dispc_device *dispc,
 			in_width > maxsinglelinewidth && ++*decim_x);
 
 	if (in_width > maxsinglelinewidth) {
-		DSSERR("Cannot scale width exceeds max line width");
+		DSSERR("Cannot scale width exceeds max line width\n");
 		return -EINVAL;
 	}
 
@@ -2496,7 +2496,7 @@ static int dispc_ovl_calc_scaling_44xx(struct dispc_device *dispc,
 		 * bandwidth. Despite what theory says this appears to
 		 * be true also for 16-bit color formats.
 		 */
-		DSSERR("Not enough bandwidth, too much downscaling (x-decimation factor %d > 4)", *decim_x);
+		DSSERR("Not enough bandwidth, too much downscaling (x-decimation factor %d > 4)\n", *decim_x);
 
 		return -EINVAL;
 	}
@@ -4639,7 +4639,7 @@ static int dispc_errata_i734_wa_init(struct dispc_device *dispc)
 						i734_buf.size, &i734_buf.paddr,
 						GFP_KERNEL);
 	if (!i734_buf.vaddr) {
-		dev_err(&dispc->pdev->dev, "%s: dma_alloc_writecombine failed",
+		dev_err(&dispc->pdev->dev, "%s: dma_alloc_writecombine failed\n",
 			__func__);
 		return -ENOMEM;
 	}
-- 
cgit v1.1


From 47aaaec818dfd1009d1358974a2931f05dd57203 Mon Sep 17 00:00:00 2001
From: Tomi Valkeinen <tomi.valkeinen@ti.com>
Date: Wed, 2 May 2018 12:11:59 +0300
Subject: drm/omap: handle alloc failures in omap_connector

Handle memory allocation failures in omap_connector to avoid NULL
derefs.

Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180502091159.7071-5-tomi.valkeinen@ti.com
Reviewed-by: Benoit Parrot <bparrot@ti.com>
Reviewed-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
---
 drivers/gpu/drm/omapdrm/omap_connector.c | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/drivers/gpu/drm/omapdrm/omap_connector.c b/drivers/gpu/drm/omapdrm/omap_connector.c
index a0d7b1d..5cde26a 100644
--- a/drivers/gpu/drm/omapdrm/omap_connector.c
+++ b/drivers/gpu/drm/omapdrm/omap_connector.c
@@ -121,6 +121,9 @@ static int omap_connector_get_modes(struct drm_connector *connector)
 	if (dssdrv->read_edid) {
 		void *edid = kzalloc(MAX_EDID, GFP_KERNEL);
 
+		if (!edid)
+			return 0;
+
 		if ((dssdrv->read_edid(dssdev, edid, MAX_EDID) > 0) &&
 				drm_edid_is_valid(edid)) {
 			drm_mode_connector_update_edid_property(
@@ -139,6 +142,9 @@ static int omap_connector_get_modes(struct drm_connector *connector)
 		struct drm_display_mode *mode = drm_mode_create(dev);
 		struct videomode vm = {0};
 
+		if (!mode)
+			return 0;
+
 		dssdrv->get_timings(dssdev, &vm);
 
 		drm_display_mode_from_videomode(&vm, mode);
@@ -200,6 +206,10 @@ static int omap_connector_mode_valid(struct drm_connector *connector,
 	if (!r) {
 		/* check if vrefresh is still valid */
 		new_mode = drm_mode_duplicate(dev, mode);
+
+		if (!new_mode)
+			return MODE_BAD;
+
 		new_mode->clock = vm.pixelclock / 1000;
 		new_mode->vrefresh = 0;
 		if (mode->vrefresh == drm_mode_vrefresh(new_mode))
-- 
cgit v1.1


From 7a109673899b6d5b147307999efe2392181acee3 Mon Sep 17 00:00:00 2001
From: Andrzej Hajda <a.hajda@samsung.com>
Date: Mon, 9 Apr 2018 08:27:08 +0200
Subject: drm/bridge/sii8620: add Kconfig dependency on extcon

The driver can work with or without extcon framework, but if extcon is
build as module, sii8620 should be build as module as well.

Fixes: 688838442147 ("drm/bridge/sii8620: use micro-USB cable detection logic to detect MHL")
Reported-by: kbuild test robot <lkp@intel.com>
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Reviewed-by: Chanwoo Choi <cw00.choi@samsung.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180409062708.4326-1-a.hajda@samsung.com
Signed-off-by: Sean Paul <seanpaul@chromium.org>
---
 drivers/gpu/drm/bridge/Kconfig | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/bridge/Kconfig b/drivers/gpu/drm/bridge/Kconfig
index 3aa65bd..684ac62 100644
--- a/drivers/gpu/drm/bridge/Kconfig
+++ b/drivers/gpu/drm/bridge/Kconfig
@@ -74,6 +74,7 @@ config DRM_SIL_SII8620
 	tristate "Silicon Image SII8620 HDMI/MHL bridge"
 	depends on OF && RC_CORE
 	select DRM_KMS_HELPER
+	imply EXTCON
 	help
 	  Silicon Image SII8620 HDMI/MHL bridge chip driver.
 
-- 
cgit v1.1


From 12d9f07022dcde261ad16e9a11f45096dc68b03c Mon Sep 17 00:00:00 2001
From: Jianchao Wang <jianchao.w.wang@oracle.com>
Date: Fri, 4 May 2018 16:01:57 +0800
Subject: nvme: fix use-after-free in nvme_free_ns_head

Currently only nvme_ctrl will take a reference counter of
nvme_subsystem, nvme_ns_head also needs it. Otherwise
nvme_free_ns_head will access the nvme_subsystem.ns_ida
which has been freed by __nvme_release_subsystem after all the
reference of nvme_subsystem have been released by nvme_free_ctrl.
This could cause memory corruption.

 BUG: KASAN: use-after-free in radix_tree_next_chunk+0x9f/0x4b0
 Read of size 8 at addr ffff88036494d2e8 by task fio/1815

 CPU: 1 PID: 1815 Comm: fio Kdump: loaded Tainted: G        W         4.17.0-rc1+ #18
 Hardware name: LENOVO 10MLS0E339/3106, BIOS M1AKT22A 06/27/2017
 Call Trace:
  dump_stack+0x91/0xeb
  print_address_description+0x6b/0x290
  kasan_report+0x261/0x360
  radix_tree_next_chunk+0x9f/0x4b0
  ida_remove+0x8b/0x180
  ida_simple_remove+0x26/0x40
  nvme_free_ns_head+0x58/0xc0
  __blkdev_put+0x30a/0x3a0
  blkdev_close+0x44/0x50
  __fput+0x184/0x380
  task_work_run+0xaf/0xe0
  do_exit+0x501/0x1440
  do_group_exit+0x89/0x140
  __x64_sys_exit_group+0x28/0x30
  do_syscall_64+0x72/0x230

Signed-off-by: Jianchao Wang <jianchao.w.wang@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Keith Busch <keith.busch@intel.com>
---
 drivers/nvme/host/core.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c
index a3771c5..2cbc378 100644
--- a/drivers/nvme/host/core.c
+++ b/drivers/nvme/host/core.c
@@ -99,6 +99,7 @@ static struct class *nvme_subsys_class;
 
 static void nvme_ns_remove(struct nvme_ns *ns);
 static int nvme_revalidate_disk(struct gendisk *disk);
+static void nvme_put_subsystem(struct nvme_subsystem *subsys);
 
 int nvme_reset_ctrl(struct nvme_ctrl *ctrl)
 {
@@ -350,6 +351,7 @@ static void nvme_free_ns_head(struct kref *ref)
 	ida_simple_remove(&head->subsys->ns_ida, head->instance);
 	list_del_init(&head->entry);
 	cleanup_srcu_struct(&head->srcu);
+	nvme_put_subsystem(head->subsys);
 	kfree(head);
 }
 
@@ -2861,6 +2863,9 @@ static struct nvme_ns_head *nvme_alloc_ns_head(struct nvme_ctrl *ctrl,
 		goto out_cleanup_srcu;
 
 	list_add_tail(&head->entry, &ctrl->subsys->nsheads);
+
+	kref_get(&ctrl->subsys->ref);
+
 	return head;
 out_cleanup_srcu:
 	cleanup_srcu_struct(&head->srcu);
-- 
cgit v1.1


From 13e1592f7379422c8d45ccd9c8c8a4698c97b39b Mon Sep 17 00:00:00 2001
From: Imre Deak <imre.deak@intel.com>
Date: Tue, 17 Apr 2018 14:31:47 +0300
Subject: drm/i915: Add documentation to gen9_set_dc_state()

Add documentation to gen9_set_dc_state() on what enabling a given DC
state means and at what point HW/DMC actually enters/exits these states.

Cc: Jani Nikula <jani.nikula@intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20180417113147.25120-1-imre.deak@intel.com
---
 drivers/gpu/drm/i915/intel_runtime_pm.c | 23 +++++++++++++++++++++++
 1 file changed, 23 insertions(+)

diff --git a/drivers/gpu/drm/i915/intel_runtime_pm.c b/drivers/gpu/drm/i915/intel_runtime_pm.c
index 3fffbfe..53a6eaa 100644
--- a/drivers/gpu/drm/i915/intel_runtime_pm.c
+++ b/drivers/gpu/drm/i915/intel_runtime_pm.c
@@ -542,6 +542,29 @@ void gen9_sanitize_dc_state(struct drm_i915_private *dev_priv)
 	dev_priv->csr.dc_state = val;
 }
 
+/**
+ * gen9_set_dc_state - set target display C power state
+ * @dev_priv: i915 device instance
+ * @state: target DC power state
+ * - DC_STATE_DISABLE
+ * - DC_STATE_EN_UPTO_DC5
+ * - DC_STATE_EN_UPTO_DC6
+ * - DC_STATE_EN_DC9
+ *
+ * Signal to DMC firmware/HW the target DC power state passed in @state.
+ * DMC/HW can turn off individual display clocks and power rails when entering
+ * a deeper DC power state (higher in number) and turns these back when exiting
+ * that state to a shallower power state (lower in number). The HW will decide
+ * when to actually enter a given state on an on-demand basis, for instance
+ * depending on the active state of display pipes. The state of display
+ * registers backed by affected power rails are saved/restored as needed.
+ *
+ * Based on the above enabling a deeper DC power state is asynchronous wrt.
+ * enabling it. Disabling a deeper power state is synchronous: for instance
+ * setting %DC_STATE_DISABLE won't complete until all HW resources are turned
+ * back on and register state is restored. This is guaranteed by the MMIO write
+ * to DC_STATE_EN blocking until the state is restored.
+ */
 static void gen9_set_dc_state(struct drm_i915_private *dev_priv, uint32_t state)
 {
 	uint32_t val;
-- 
cgit v1.1


From 2abb80dad3afa9170ae19ca03bb7b4cd1ec06d62 Mon Sep 17 00:00:00 2001
From: Yisheng Xie <xieyisheng1@huawei.com>
Date: Wed, 25 Apr 2018 16:25:07 +0800
Subject: perf bench numa: Fix typo in options

'R' means access the data via reads instead of writes, fix this typo.

Signed-off-by: Yisheng Xie <xieyisheng1@huawei.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1524644707-11030-1-git-send-email-xieyisheng1@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/bench/numa.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/perf/bench/numa.c b/tools/perf/bench/numa.c
index 944070e..63eb4908 100644
--- a/tools/perf/bench/numa.c
+++ b/tools/perf/bench/numa.c
@@ -175,7 +175,7 @@ static const struct option options[] = {
 	OPT_UINTEGER('s', "nr_secs"	, &p0.nr_secs,		"max number of seconds to run (default: 5 secs)"),
 	OPT_UINTEGER('u', "usleep"	, &p0.sleep_usecs,	"usecs to sleep per loop iteration"),
 
-	OPT_BOOLEAN('R', "data_reads"	, &p0.data_reads,	"access the data via writes (can be mixed with -W)"),
+	OPT_BOOLEAN('R', "data_reads"	, &p0.data_reads,	"access the data via reads (can be mixed with -W)"),
 	OPT_BOOLEAN('W', "data_writes"	, &p0.data_writes,	"access the data via writes (can be mixed with -R)"),
 	OPT_BOOLEAN('B', "data_backwards", &p0.data_backwards,	"access the data backwards as well"),
 	OPT_BOOLEAN('Z', "data_zero_memset", &p0.data_zero_memset,"access the data via glibc bzero only"),
-- 
cgit v1.1


From 0d74d872c3f8b9cb3d096fb932a063b43b37f188 Mon Sep 17 00:00:00 2001
From: Mathieu Malaterre <malat@debian.org>
Date: Sat, 5 May 2018 22:00:37 +0200
Subject: driver core: add __printf verification to __ata_ehi_pushv_desc
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

__printf is useful to verify format and arguments. Remove the following
warning (with W=1):

  drivers/ata/libata-eh.c:183:10: warning: function might be possible candidate for ‘gnu_printf’ format attribute [-Wsuggest-attribute=format]

Signed-off-by: Mathieu Malaterre <malat@debian.org>
Signed-off-by: Tejun Heo <tj@kernel.org>
---
 drivers/ata/libata-eh.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/ata/libata-eh.c b/drivers/ata/libata-eh.c
index c016829..513b260b 100644
--- a/drivers/ata/libata-eh.c
+++ b/drivers/ata/libata-eh.c
@@ -175,8 +175,8 @@ static void ata_eh_handle_port_resume(struct ata_port *ap)
 { }
 #endif /* CONFIG_PM */
 
-static void __ata_ehi_pushv_desc(struct ata_eh_info *ehi, const char *fmt,
-				 va_list args)
+static __printf(2, 0) void __ata_ehi_pushv_desc(struct ata_eh_info *ehi,
+				 const char *fmt, va_list args)
 {
 	ehi->desc_len += vscnprintf(ehi->desc + ehi->desc_len,
 				     ATA_EH_DESC_LEN - ehi->desc_len,
-- 
cgit v1.1


From f142f08bf7ecc41c3e71e05b765ea654047cf0c0 Mon Sep 17 00:00:00 2001
From: Florian La Roche <florian.laroche@googlemail.com>
Date: Sun, 6 May 2018 19:34:07 +0200
Subject: Fix typo in comment.

CONFIG_PRREMPT -> CONFIG_PREEMPT

Signed-off-by: Florian La Roche <Florian.LaRoche@googlemail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 init/main.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/init/main.c b/init/main.c
index b795aa3..a404936 100644
--- a/init/main.c
+++ b/init/main.c
@@ -423,7 +423,7 @@ static noinline void __ref rest_init(void)
 
 	/*
 	 * Enable might_sleep() and smp_processor_id() checks.
-	 * They cannot be enabled earlier because with CONFIG_PRREMPT=y
+	 * They cannot be enabled earlier because with CONFIG_PREEMPT=y
 	 * kernel_thread() would trigger might_sleep() splats. With
 	 * CONFIG_PREEMPT_VOLUNTARY=y the init task might have scheduled
 	 * already, but it's stuck on the kthreadd_done completion.
-- 
cgit v1.1


From ea9032fa2e4e91ae15facff5b7c4b2a84c1e40af Mon Sep 17 00:00:00 2001
From: William Cohen <wcohen@redhat.com>
Date: Thu, 3 May 2018 15:50:32 -0400
Subject: perf vendor events intel: Remove duplicated entry for westmereep-dp
 in mapfile.csv

Signed-off-by: William Cohen <wcohen@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20180503195032.28871-1-wcohen@redhat.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/pmu-events/arch/x86/mapfile.csv | 1 -
 1 file changed, 1 deletion(-)

diff --git a/tools/perf/pmu-events/arch/x86/mapfile.csv b/tools/perf/pmu-events/arch/x86/mapfile.csv
index 93656f2..7e3cce3 100644
--- a/tools/perf/pmu-events/arch/x86/mapfile.csv
+++ b/tools/perf/pmu-events/arch/x86/mapfile.csv
@@ -29,7 +29,6 @@ GenuineIntel-6-4D,v13,silvermont,core
 GenuineIntel-6-4C,v13,silvermont,core
 GenuineIntel-6-2A,v15,sandybridge,core
 GenuineIntel-6-2C,v2,westmereep-dp,core
-GenuineIntel-6-2C,v2,westmereep-dp,core
 GenuineIntel-6-25,v2,westmereep-sp,core
 GenuineIntel-6-2F,v2,westmereex,core
 GenuineIntel-6-55,v1,skylakex,core
-- 
cgit v1.1


From 605e71cd007b1897a4c5eda952acadd61b6a98f0 Mon Sep 17 00:00:00 2001
From: Arnaldo Carvalho de Melo <acme@redhat.com>
Date: Mon, 7 May 2018 12:22:37 -0300
Subject: tools headers: Sync x86 cpufeatures.h with the kernel sources

The 912413057395 ("x86/cpufeatures: Enumerate cldemote instruction")
doesn't requires changes in the tools, just copy it to silence this
warning:

  Warning: Kernel ABI header at 'tools/arch/x86/include/asm/cpufeatures.h' differs from latest version at 'arch/x86/include/asm/cpufeatures.h'

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: https://lkml.kernel.org/n/tip-1vo20y5z2drlujfpltjudwk8@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/arch/x86/include/asm/cpufeatures.h | 1 +
 1 file changed, 1 insertion(+)

diff --git a/tools/arch/x86/include/asm/cpufeatures.h b/tools/arch/x86/include/asm/cpufeatures.h
index d554c11..578793e 100644
--- a/tools/arch/x86/include/asm/cpufeatures.h
+++ b/tools/arch/x86/include/asm/cpufeatures.h
@@ -320,6 +320,7 @@
 #define X86_FEATURE_AVX512_VPOPCNTDQ	(16*32+14) /* POPCNT for vectors of DW/QW */
 #define X86_FEATURE_LA57		(16*32+16) /* 5-level page tables */
 #define X86_FEATURE_RDPID		(16*32+22) /* RDPID instruction */
+#define X86_FEATURE_CLDEMOTE		(16*32+25) /* CLDEMOTE instruction */
 
 /* AMD-defined CPU features, CPUID level 0x80000007 (EBX), word 17 */
 #define X86_FEATURE_OVERFLOW_RECOV	(17*32+ 0) /* MCA overflow recovery support */
-- 
cgit v1.1


From d0e9f4c1a4e777c4d1122758137fc47238f4f015 Mon Sep 17 00:00:00 2001
From: Arnaldo Carvalho de Melo <acme@redhat.com>
Date: Mon, 7 May 2018 12:26:23 -0300
Subject: tools headers kvm: Sync uapi/linux/kvm.h with the kernel sources
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The changes in 5e62493f1a70 ("x86/headers/UAPI: Move DISABLE_EXITS KVM
capability bits to the UAPI") do not requires changes in the tooling nor
will trigger the automatic update of used ioctl string tables, copy it
to silence this build warning:

  Warning: Kernel ABI header at 'tools/include/uapi/linux/kvm.h' differs from latest version at 'include/uapi/linux/kvm.h'

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: KarimAllah Ahmed <karahmed@amazon.de>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Radim Krčmář <rkrcmar@redhat.com>
Cc: Wang Nan <wangnan0@huawei.com>
Link: https://lkml.kernel.org/n/tip-8o5auh1lqglsgl1q97x00tlv@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/include/uapi/linux/kvm.h | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/tools/include/uapi/linux/kvm.h b/tools/include/uapi/linux/kvm.h
index 1065006..b02c41e 100644
--- a/tools/include/uapi/linux/kvm.h
+++ b/tools/include/uapi/linux/kvm.h
@@ -676,6 +676,13 @@ struct kvm_ioeventfd {
 	__u8  pad[36];
 };
 
+#define KVM_X86_DISABLE_EXITS_MWAIT          (1 << 0)
+#define KVM_X86_DISABLE_EXITS_HTL            (1 << 1)
+#define KVM_X86_DISABLE_EXITS_PAUSE          (1 << 2)
+#define KVM_X86_DISABLE_VALID_EXITS          (KVM_X86_DISABLE_EXITS_MWAIT | \
+                                              KVM_X86_DISABLE_EXITS_HTL | \
+                                              KVM_X86_DISABLE_EXITS_PAUSE)
+
 /* for KVM_ENABLE_CAP */
 struct kvm_enable_cap {
 	/* in */
-- 
cgit v1.1


From 5981ec36688c49b7262f399c1a10edecc6e55ed2 Mon Sep 17 00:00:00 2001
From: Arnaldo Carvalho de Melo <acme@redhat.com>
Date: Mon, 7 May 2018 13:23:19 -0300
Subject: tools headers kvm: Sync ARM UAPI headers with the kernel sources

To sync with the changes made in 85bd0ba1ff98 ("arm/arm64: KVM: Add PSCI
version selection API"), that do not cause any changes in the tools,
just to silence the build warning:

  Warning: Kernel ABI header at 'tools/arch/arm/include/uapi/asm/kvm.h' differs from latest version at 'arch/arm/include/uapi/asm/kvm.h'
  Warning: Kernel ABI header at 'tools/arch/arm64/include/uapi/asm/kvm.h' differs from latest version at 'arch/arm64/include/uapi/asm/kvm.h'

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: https://lkml.kernel.org/n/tip-7u37pv09xtvet1ll27840w73@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/arch/arm/include/uapi/asm/kvm.h   | 6 ++++++
 tools/arch/arm64/include/uapi/asm/kvm.h | 6 ++++++
 2 files changed, 12 insertions(+)

diff --git a/tools/arch/arm/include/uapi/asm/kvm.h b/tools/arch/arm/include/uapi/asm/kvm.h
index 2ba95d6..caae484 100644
--- a/tools/arch/arm/include/uapi/asm/kvm.h
+++ b/tools/arch/arm/include/uapi/asm/kvm.h
@@ -195,6 +195,12 @@ struct kvm_arch_memory_slot {
 #define KVM_REG_ARM_VFP_FPINST		0x1009
 #define KVM_REG_ARM_VFP_FPINST2		0x100A
 
+/* KVM-as-firmware specific pseudo-registers */
+#define KVM_REG_ARM_FW			(0x0014 << KVM_REG_ARM_COPROC_SHIFT)
+#define KVM_REG_ARM_FW_REG(r)		(KVM_REG_ARM | KVM_REG_SIZE_U64 | \
+					 KVM_REG_ARM_FW | ((r) & 0xffff))
+#define KVM_REG_ARM_PSCI_VERSION	KVM_REG_ARM_FW_REG(0)
+
 /* Device Control API: ARM VGIC */
 #define KVM_DEV_ARM_VGIC_GRP_ADDR	0
 #define KVM_DEV_ARM_VGIC_GRP_DIST_REGS	1
diff --git a/tools/arch/arm64/include/uapi/asm/kvm.h b/tools/arch/arm64/include/uapi/asm/kvm.h
index 9abbf30..04b3256 100644
--- a/tools/arch/arm64/include/uapi/asm/kvm.h
+++ b/tools/arch/arm64/include/uapi/asm/kvm.h
@@ -206,6 +206,12 @@ struct kvm_arch_memory_slot {
 #define KVM_REG_ARM_TIMER_CNT		ARM64_SYS_REG(3, 3, 14, 3, 2)
 #define KVM_REG_ARM_TIMER_CVAL		ARM64_SYS_REG(3, 3, 14, 0, 2)
 
+/* KVM-as-firmware specific pseudo-registers */
+#define KVM_REG_ARM_FW			(0x0014 << KVM_REG_ARM_COPROC_SHIFT)
+#define KVM_REG_ARM_FW_REG(r)		(KVM_REG_ARM64 | KVM_REG_SIZE_U64 | \
+					 KVM_REG_ARM_FW | ((r) & 0xffff))
+#define KVM_REG_ARM_PSCI_VERSION	KVM_REG_ARM_FW_REG(0)
+
 /* Device Control API: ARM VGIC */
 #define KVM_DEV_ARM_VGIC_GRP_ADDR	0
 #define KVM_DEV_ARM_VGIC_GRP_DIST_REGS	1
-- 
cgit v1.1


From 914eac248d876f9c00cd1792ffec3d182c863f13 Mon Sep 17 00:00:00 2001
From: Sara Sharon <sara.sharon@intel.com>
Date: Fri, 20 Apr 2018 13:49:19 +0300
Subject: mac80211: use timeout from the AddBA response instead of the request

2016 spec, section 10.24.2 specifies that the block ack
timeout in the ADD BA request is advisory.

That means we should check the value in the response and
act upon it (same as buffer size).

Signed-off-by: Sara Sharon <sara.sharon@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
---
 net/mac80211/agg-tx.c | 4 ++++
 net/mac80211/tx.c     | 3 ++-
 2 files changed, 6 insertions(+), 1 deletion(-)

diff --git a/net/mac80211/agg-tx.c b/net/mac80211/agg-tx.c
index 595c662..ac42952 100644
--- a/net/mac80211/agg-tx.c
+++ b/net/mac80211/agg-tx.c
@@ -8,6 +8,7 @@
  * Copyright 2007, Michael Wu <flamingice@sourmilk.net>
  * Copyright 2007-2010, Intel Corporation
  * Copyright(c) 2015-2017 Intel Deutschland GmbH
+ * Copyright (C) 2018 Intel Corporation
  *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of the GNU General Public License version 2 as
@@ -970,6 +971,9 @@ void ieee80211_process_addba_resp(struct ieee80211_local *local,
 
 		sta->ampdu_mlme.addba_req_num[tid] = 0;
 
+		tid_tx->timeout =
+			le16_to_cpu(mgmt->u.action.u.addba_resp.timeout);
+
 		if (tid_tx->timeout) {
 			mod_timer(&tid_tx->session_timer,
 				  TU_TO_EXP_TIME(tid_tx->timeout));
diff --git a/net/mac80211/tx.c b/net/mac80211/tx.c
index 535de31..05a265c 100644
--- a/net/mac80211/tx.c
+++ b/net/mac80211/tx.c
@@ -4,6 +4,7 @@
  * Copyright 2006-2007	Jiri Benc <jbenc@suse.cz>
  * Copyright 2007	Johannes Berg <johannes@sipsolutions.net>
  * Copyright 2013-2014  Intel Mobile Communications GmbH
+ * Copyright (C) 2018 Intel Corporation
  *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of the GNU General Public License version 2 as
@@ -1135,7 +1136,7 @@ static bool ieee80211_tx_prep_agg(struct ieee80211_tx_data *tx,
 	}
 
 	/* reset session timer */
-	if (reset_agg_timer && tid_tx->timeout)
+	if (reset_agg_timer)
 		tid_tx->last_tx = jiffies;
 
 	return queued;
-- 
cgit v1.1


From d1ecfa9d1f402366b1776fbf84e635678a51414f Mon Sep 17 00:00:00 2001
From: "van der Linden, Frank" <fllinden@amazon.com>
Date: Fri, 4 May 2018 16:11:00 -0400
Subject: x86/xen: Reset VCPU0 info pointer after shared_info remap

This patch fixes crashes during boot for HVM guests on older (pre HVM
vector callback) Xen versions. Without this, current kernels will always
fail to boot on those Xen versions.

Sample stack trace:

   BUG: unable to handle kernel paging request at ffffffffff200000
   IP: __xen_evtchn_do_upcall+0x1e/0x80
   PGD 1e0e067 P4D 1e0e067 PUD 1e10067 PMD 235c067 PTE 0
    Oops: 0002 [#1] SMP PTI
   Modules linked in:
   CPU: 0 PID: 512 Comm: kworker/u2:0 Not tainted 4.14.33-52.13.amzn1.x86_64 #1
   Hardware name: Xen HVM domU, BIOS 3.4.3.amazon 11/11/2016
   task: ffff88002531d700 task.stack: ffffc90000480000
   RIP: 0010:__xen_evtchn_do_upcall+0x1e/0x80
   RSP: 0000:ffff880025403ef0 EFLAGS: 00010046
   RAX: ffffffff813cc760 RBX: ffffffffff200000 RCX: ffffc90000483ef0
   RDX: ffff880020540a00 RSI: ffff880023c78000 RDI: 000000000000001c
   RBP: 0000000000000001 R08: 0000000000000000 R09: 0000000000000000
   R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000000
   R13: ffff880025403f5c R14: 0000000000000000 R15: 0000000000000000
   FS:  0000000000000000(0000) GS:ffff880025400000(0000) knlGS:0000000000000000
   CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
   CR2: ffffffffff200000 CR3: 0000000001e0a000 CR4: 00000000000006f0
    Call Trace:
   <IRQ>
   do_hvm_evtchn_intr+0xa/0x10
   __handle_irq_event_percpu+0x43/0x1a0
   handle_irq_event_percpu+0x20/0x50
   handle_irq_event+0x39/0x60
   handle_fasteoi_irq+0x80/0x140
   handle_irq+0xaf/0x120
   do_IRQ+0x41/0xd0
   common_interrupt+0x7d/0x7d
   </IRQ>

During boot, the HYPERVISOR_shared_info page gets remapped to make it work
with KASLR. This means that any pointer derived from it needs to be
adjusted.

The only value that this applies to is the vcpu_info pointer for VCPU 0.
For PV and HVM with the callback vector feature, this gets done via the
smp_ops prepare_boot_cpu callback. Older Xen versions do not support the
HVM callback vector, so there is no Xen-specific smp_ops set up in that
scenario. So, the vcpu_info pointer for VCPU 0 never gets set to the proper
value, and the first reference of it will be bad. Fix this by resetting it
immediately after the remap.

Signed-off-by: Frank van der Linden <fllinden@amazon.com>
Reviewed-by: Eduardo Valentin <eduval@amazon.com>
Reviewed-by: Alakesh Haloi <alakeshh@amazon.com>
Reviewed-by: Vallish Vaidyeshwara <vallish@amazon.com>
Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: Juergen Gross <jgross@suse.com>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: xen-devel@lists.xenproject.org
Signed-off-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
---
 arch/x86/xen/enlighten_hvm.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)

diff --git a/arch/x86/xen/enlighten_hvm.c b/arch/x86/xen/enlighten_hvm.c
index 8268987..19c1ff5 100644
--- a/arch/x86/xen/enlighten_hvm.c
+++ b/arch/x86/xen/enlighten_hvm.c
@@ -65,6 +65,19 @@ static void __init xen_hvm_init_mem_mapping(void)
 {
 	early_memunmap(HYPERVISOR_shared_info, PAGE_SIZE);
 	HYPERVISOR_shared_info = __va(PFN_PHYS(shared_info_pfn));
+
+	/*
+	 * The virtual address of the shared_info page has changed, so
+	 * the vcpu_info pointer for VCPU 0 is now stale.
+	 *
+	 * The prepare_boot_cpu callback will re-initialize it via
+	 * xen_vcpu_setup, but we can't rely on that to be called for
+	 * old Xen versions (xen_have_vector_callback == 0).
+	 *
+	 * It is, in any case, bad to have a stale vcpu_info pointer
+	 * so reset it now.
+	 */
+	xen_vcpu_info_reset(0);
 }
 
 static void __init init_hvm_pv_info(void)
-- 
cgit v1.1


From 4a35a9027f64d588d2fd9436dda4126e8d5647d7 Mon Sep 17 00:00:00 2001
From: Arnaldo Carvalho de Melo <acme@redhat.com>
Date: Mon, 7 May 2018 15:27:01 -0300
Subject: Revert "perf pmu: Fix pmu events parsing rule"

As reported by Adrian Hunter, this breaks intel_pt event parsing:

  # perf record -e intel_pt//u uname
  event syntax error: 'intel_pt//u'
                               \___ parser error
  Run 'perf list' for a list of valid events

   Usage: perf record [<options>] [<command>]
      or: perf record [<options>] -- <command> [<options>]

      -e, --event <event>   event selector. use 'perf list' to list available events
  #

This reverts commit 9a4a931ce847f4aaa12edf11b2e050e18bf45910.

Reported-by: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: https://lkml.kernel.org/n/tip-ye1o2mji7x68xotiot1tn1gp@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/parse-events.y | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/tools/perf/util/parse-events.y b/tools/perf/util/parse-events.y
index d14464c..7afeb80 100644
--- a/tools/perf/util/parse-events.y
+++ b/tools/perf/util/parse-events.y
@@ -224,15 +224,15 @@ event_def: event_pmu |
 	   event_bpf_file
 
 event_pmu:
-PE_NAME '/' event_config '/'
+PE_NAME opt_event_config
 {
 	struct list_head *list, *orig_terms, *terms;
 
-	if (parse_events_copy_term_list($3, &orig_terms))
+	if (parse_events_copy_term_list($2, &orig_terms))
 		YYABORT;
 
 	ALLOC_LIST(list);
-	if (parse_events_add_pmu(_parse_state, list, $1, $3, false)) {
+	if (parse_events_add_pmu(_parse_state, list, $1, $2, false)) {
 		struct perf_pmu *pmu = NULL;
 		int ok = 0;
 		char *pattern;
@@ -262,7 +262,7 @@ PE_NAME '/' event_config '/'
 		if (!ok)
 			YYABORT;
 	}
-	parse_events_terms__delete($3);
+	parse_events_terms__delete($2);
 	parse_events_terms__delete(orig_terms);
 	$$ = list;
 }
-- 
cgit v1.1


From c27e917e2bda748777b7927d7cb7c911bc2027c8 Mon Sep 17 00:00:00 2001
From: Paulo Zanoni <paulo.r.zanoni@intel.com>
Date: Fri, 27 Apr 2018 16:14:36 -0700
Subject: drm/i915/icl: add basic support for the ICL clocks

This commit introduces the definitions for the ICL clocks and adds the
basic functions to the shared DPLL framework. It adds code for the
Enable and Disable sequences for some PLLs, but it does not have the
code to compute the actual PLL values, which are marked as TODO
comments and should be introduced as separate commits.

Special thanks to James Ausmus for investigating and fixing a bug with
the placement of icl_unmap_plls_to_ports() function.

v2:
 - Rebase around dpll_lock changes.
v3:
 - The spec now says what the timeouts should be.
 - Touch DPCLKA_CFGCR0_ICL at the appropriate time so we don't freeze
   the machine.
 - Checkpatch found a white space problem.
 - Small adjustments before upstreaming.
v4:
 - Move the ICL checks out of the *map_plls_to_ports() functions
  (James)
 - Add extra encoder check (James)
 - Call icl_unmap_plls_to_ports() later (James)
v5:
 - Rebase after the pll struct changes.
v6:
 - Properly make the unmap function based on encoders_post_disable()
   with regarding to checks and iterators.
 - Address checkpatch comment on "min = max = x()".

Cc: James Ausmus <james.ausmus@intel.com>
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: James Ausmus <james.ausmus@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180427231436.9353-1-paulo.r.zanoni@intel.com
---
 drivers/gpu/drm/i915/i915_debugfs.c   |  22 +++
 drivers/gpu/drm/i915/intel_ddi.c      |  98 ++++++++++-
 drivers/gpu/drm/i915/intel_display.c  |  16 ++
 drivers/gpu/drm/i915/intel_dpll_mgr.c | 313 +++++++++++++++++++++++++++++++++-
 drivers/gpu/drm/i915/intel_dpll_mgr.h |  41 +++++
 drivers/gpu/drm/i915/intel_drv.h      |   6 +
 6 files changed, 491 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index 85911bc..13e7b9e 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -3368,6 +3368,28 @@ static int i915_shared_dplls_info(struct seq_file *m, void *unused)
 		seq_printf(m, " fp0:     0x%08x\n", pll->state.hw_state.fp0);
 		seq_printf(m, " fp1:     0x%08x\n", pll->state.hw_state.fp1);
 		seq_printf(m, " wrpll:   0x%08x\n", pll->state.hw_state.wrpll);
+		seq_printf(m, " cfgcr0:  0x%08x\n", pll->state.hw_state.cfgcr0);
+		seq_printf(m, " cfgcr1:  0x%08x\n", pll->state.hw_state.cfgcr1);
+		seq_printf(m, " mg_refclkin_ctl:        0x%08x\n",
+			   pll->state.hw_state.mg_refclkin_ctl);
+		seq_printf(m, " mg_clktop2_coreclkctl1: 0x%08x\n",
+			   pll->state.hw_state.mg_clktop2_coreclkctl1);
+		seq_printf(m, " mg_clktop2_hsclkctl:    0x%08x\n",
+			   pll->state.hw_state.mg_clktop2_hsclkctl);
+		seq_printf(m, " mg_pll_div0:  0x%08x\n",
+			   pll->state.hw_state.mg_pll_div0);
+		seq_printf(m, " mg_pll_div1:  0x%08x\n",
+			   pll->state.hw_state.mg_pll_div1);
+		seq_printf(m, " mg_pll_lf:    0x%08x\n",
+			   pll->state.hw_state.mg_pll_lf);
+		seq_printf(m, " mg_pll_frac_lock: 0x%08x\n",
+			   pll->state.hw_state.mg_pll_frac_lock);
+		seq_printf(m, " mg_pll_ssc:   0x%08x\n",
+			   pll->state.hw_state.mg_pll_ssc);
+		seq_printf(m, " mg_pll_bias:  0x%08x\n",
+			   pll->state.hw_state.mg_pll_bias);
+		seq_printf(m, " mg_pll_tdc_coldst_bias: 0x%08x\n",
+			   pll->state.hw_state.mg_pll_tdc_coldst_bias);
 	}
 	drm_modeset_unlock_all(dev);
 
diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c
index 8225d22..b98ac054 100644
--- a/drivers/gpu/drm/i915/intel_ddi.c
+++ b/drivers/gpu/drm/i915/intel_ddi.c
@@ -1052,6 +1052,25 @@ static uint32_t hsw_pll_to_ddi_pll_sel(const struct intel_shared_dpll *pll)
 	}
 }
 
+static uint32_t icl_pll_to_ddi_pll_sel(struct intel_encoder *encoder,
+				       const struct intel_shared_dpll *pll)
+{
+	const enum intel_dpll_id id = pll->info->id;
+
+	switch (id) {
+	default:
+		MISSING_CASE(id);
+	case DPLL_ID_ICL_DPLL0:
+	case DPLL_ID_ICL_DPLL1:
+		return DDI_CLK_SEL_NONE;
+	case DPLL_ID_ICL_MGPLL1:
+	case DPLL_ID_ICL_MGPLL2:
+	case DPLL_ID_ICL_MGPLL3:
+	case DPLL_ID_ICL_MGPLL4:
+		return DDI_CLK_SEL_MG;
+	}
+}
+
 /* Starting with Haswell, different DDI ports can work in FDI mode for
  * connection to the PCH-located connectors. For this, it is necessary to train
  * both the DDI port and PCH receiver for the desired DDI buffer settings.
@@ -2421,6 +2440,69 @@ uint32_t ddi_signal_levels(struct intel_dp *intel_dp)
 	return DDI_BUF_TRANS_SELECT(level);
 }
 
+void icl_map_plls_to_ports(struct drm_crtc *crtc,
+			   struct intel_crtc_state *crtc_state,
+			   struct drm_atomic_state *old_state)
+{
+	struct intel_shared_dpll *pll = crtc_state->shared_dpll;
+	struct drm_i915_private *dev_priv = to_i915(crtc->dev);
+	struct drm_connector_state *conn_state;
+	struct drm_connector *conn;
+	int i;
+
+	for_each_new_connector_in_state(old_state, conn, conn_state, i) {
+		struct intel_encoder *encoder =
+			to_intel_encoder(conn_state->best_encoder);
+		enum port port = encoder->port;
+		uint32_t val;
+
+		if (conn_state->crtc != crtc)
+			continue;
+
+		mutex_lock(&dev_priv->dpll_lock);
+
+		val = I915_READ(DPCLKA_CFGCR0_ICL);
+		WARN_ON((val & DPCLKA_CFGCR0_DDI_CLK_OFF(port)) == 0);
+
+		if (port == PORT_A || port == PORT_B) {
+			val &= ~DPCLKA_CFGCR0_DDI_CLK_SEL_MASK(port);
+			val |= DPCLKA_CFGCR0_DDI_CLK_SEL(pll->info->id, port);
+			I915_WRITE(DPCLKA_CFGCR0_ICL, val);
+			POSTING_READ(DPCLKA_CFGCR0_ICL);
+		}
+
+		val &= ~DPCLKA_CFGCR0_DDI_CLK_OFF(port);
+		I915_WRITE(DPCLKA_CFGCR0_ICL, val);
+
+		mutex_unlock(&dev_priv->dpll_lock);
+	}
+}
+
+void icl_unmap_plls_to_ports(struct drm_crtc *crtc,
+			     struct intel_crtc_state *crtc_state,
+			     struct drm_atomic_state *old_state)
+{
+	struct drm_i915_private *dev_priv = to_i915(crtc->dev);
+	struct drm_connector_state *old_conn_state;
+	struct drm_connector *conn;
+	int i;
+
+	for_each_old_connector_in_state(old_state, conn, old_conn_state, i) {
+		struct intel_encoder *encoder =
+			to_intel_encoder(old_conn_state->best_encoder);
+		enum port port = encoder->port;
+
+		if (old_conn_state->crtc != crtc)
+			continue;
+
+		mutex_lock(&dev_priv->dpll_lock);
+		I915_WRITE(DPCLKA_CFGCR0_ICL,
+			   I915_READ(DPCLKA_CFGCR0_ICL) |
+			   DPCLKA_CFGCR0_DDI_CLK_OFF(port));
+		mutex_unlock(&dev_priv->dpll_lock);
+	}
+}
+
 static void intel_ddi_clk_select(struct intel_encoder *encoder,
 				 const struct intel_shared_dpll *pll)
 {
@@ -2433,7 +2515,11 @@ static void intel_ddi_clk_select(struct intel_encoder *encoder,
 
 	mutex_lock(&dev_priv->dpll_lock);
 
-	if (IS_CANNONLAKE(dev_priv)) {
+	if (IS_ICELAKE(dev_priv)) {
+		if (port >= PORT_C)
+			I915_WRITE(DDI_CLK_SEL(port),
+				   icl_pll_to_ddi_pll_sel(encoder, pll));
+	} else if (IS_CANNONLAKE(dev_priv)) {
 		/* Configure DPCLKA_CFGCR0 to map the DPLL to the DDI. */
 		val = I915_READ(DPCLKA_CFGCR0);
 		val &= ~DPCLKA_CFGCR0_DDI_CLK_SEL_MASK(port);
@@ -2471,14 +2557,18 @@ static void intel_ddi_clk_disable(struct intel_encoder *encoder)
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	enum port port = encoder->port;
 
-	if (IS_CANNONLAKE(dev_priv))
+	if (IS_ICELAKE(dev_priv)) {
+		if (port >= PORT_C)
+			I915_WRITE(DDI_CLK_SEL(port), DDI_CLK_SEL_NONE);
+	} else if (IS_CANNONLAKE(dev_priv)) {
 		I915_WRITE(DPCLKA_CFGCR0, I915_READ(DPCLKA_CFGCR0) |
 			   DPCLKA_CFGCR0_DDI_CLK_OFF(port));
-	else if (IS_GEN9_BC(dev_priv))
+	} else if (IS_GEN9_BC(dev_priv)) {
 		I915_WRITE(DPLL_CTRL2, I915_READ(DPLL_CTRL2) |
 			   DPLL_CTRL2_DDI_CLK_OFF(port));
-	else if (INTEL_GEN(dev_priv) < 9)
+	} else if (INTEL_GEN(dev_priv) < 9) {
 		I915_WRITE(PORT_CLK_SEL(port), PORT_CLK_SEL_NONE);
+	}
 }
 
 static void intel_ddi_pre_enable_dp(struct intel_encoder *encoder,
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 3fd249c..cdfe095 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -5559,6 +5559,9 @@ static void haswell_crtc_enable(struct intel_crtc_state *pipe_config,
 	if (intel_crtc->config->shared_dpll)
 		intel_enable_shared_dpll(intel_crtc);
 
+	if (INTEL_GEN(dev_priv) >= 11)
+		icl_map_plls_to_ports(crtc, pipe_config, old_state);
+
 	if (intel_crtc_has_dp_encoder(intel_crtc->config))
 		intel_dp_set_m_n(intel_crtc, M1_N1);
 
@@ -5756,6 +5759,9 @@ static void haswell_crtc_disable(struct intel_crtc_state *old_crtc_state,
 		intel_ddi_disable_pipe_clock(intel_crtc->config);
 
 	intel_encoders_post_disable(crtc, old_crtc_state, old_state);
+
+	if (INTEL_GEN(dev_priv) >= 11)
+		icl_unmap_plls_to_ports(crtc, old_crtc_state, old_state);
 }
 
 static void i9xx_pfit_enable(struct intel_crtc *crtc)
@@ -11386,6 +11392,16 @@ intel_pipe_config_compare(struct drm_i915_private *dev_priv,
 	PIPE_CONF_CHECK_X(dpll_hw_state.pll9);
 	PIPE_CONF_CHECK_X(dpll_hw_state.pll10);
 	PIPE_CONF_CHECK_X(dpll_hw_state.pcsdw12);
+	PIPE_CONF_CHECK_X(dpll_hw_state.mg_refclkin_ctl);
+	PIPE_CONF_CHECK_X(dpll_hw_state.mg_clktop2_coreclkctl1);
+	PIPE_CONF_CHECK_X(dpll_hw_state.mg_clktop2_hsclkctl);
+	PIPE_CONF_CHECK_X(dpll_hw_state.mg_pll_div0);
+	PIPE_CONF_CHECK_X(dpll_hw_state.mg_pll_div1);
+	PIPE_CONF_CHECK_X(dpll_hw_state.mg_pll_lf);
+	PIPE_CONF_CHECK_X(dpll_hw_state.mg_pll_frac_lock);
+	PIPE_CONF_CHECK_X(dpll_hw_state.mg_pll_ssc);
+	PIPE_CONF_CHECK_X(dpll_hw_state.mg_pll_bias);
+	PIPE_CONF_CHECK_X(dpll_hw_state.mg_pll_tdc_coldst_bias);
 
 	PIPE_CONF_CHECK_X(dsi_pll.ctrl);
 	PIPE_CONF_CHECK_X(dsi_pll.div);
diff --git a/drivers/gpu/drm/i915/intel_dpll_mgr.c b/drivers/gpu/drm/i915/intel_dpll_mgr.c
index d5e114e..14f5414 100644
--- a/drivers/gpu/drm/i915/intel_dpll_mgr.c
+++ b/drivers/gpu/drm/i915/intel_dpll_mgr.c
@@ -2399,6 +2399,315 @@ static const struct intel_dpll_mgr cnl_pll_mgr = {
 	.dump_hw_state = cnl_dump_hw_state,
 };
 
+static bool icl_calc_dpll_state(struct intel_crtc_state *crtc_state,
+				struct intel_encoder *encoder, int clock,
+				struct intel_dpll_hw_state *pll_state)
+{
+	/* TODO */
+	return true;
+}
+
+static enum port icl_mg_pll_id_to_port(enum intel_dpll_id id)
+{
+	return id - DPLL_ID_ICL_MGPLL1 + PORT_C;
+}
+
+static enum intel_dpll_id icl_port_to_mg_pll_id(enum port port)
+{
+	return port - PORT_C + DPLL_ID_ICL_MGPLL1;
+}
+
+static bool icl_calc_mg_pll_state(struct intel_crtc_state *crtc_state,
+				  struct intel_encoder *encoder, int clock,
+				  struct intel_dpll_hw_state *pll_state)
+{
+	/* TODO */
+	return true;
+}
+
+static struct intel_shared_dpll *
+icl_get_dpll(struct intel_crtc *crtc, struct intel_crtc_state *crtc_state,
+	     struct intel_encoder *encoder)
+{
+	struct intel_shared_dpll *pll;
+	struct intel_dpll_hw_state pll_state = {};
+	enum port port = encoder->port;
+	enum intel_dpll_id min, max;
+	int clock = crtc_state->port_clock;
+	bool ret;
+
+	switch (port) {
+	case PORT_A:
+	case PORT_B:
+		min = DPLL_ID_ICL_DPLL0;
+		max = DPLL_ID_ICL_DPLL1;
+		ret = icl_calc_dpll_state(crtc_state, encoder, clock,
+					  &pll_state);
+		break;
+	case PORT_C:
+	case PORT_D:
+	case PORT_E:
+	case PORT_F:
+		min = icl_port_to_mg_pll_id(port);
+		max = min;
+		ret = icl_calc_mg_pll_state(crtc_state, encoder, clock,
+					    &pll_state);
+		break;
+	default:
+		MISSING_CASE(port);
+		return NULL;
+	}
+
+	if (!ret) {
+		DRM_DEBUG_KMS("Could not calculate PLL state.\n");
+		return NULL;
+	}
+
+	crtc_state->dpll_hw_state = pll_state;
+
+	pll = intel_find_shared_dpll(crtc, crtc_state, min, max);
+	if (!pll) {
+		DRM_DEBUG_KMS("No PLL selected\n");
+		return NULL;
+	}
+
+	intel_reference_shared_dpll(pll, crtc_state);
+
+	return pll;
+}
+
+static i915_reg_t icl_pll_id_to_enable_reg(enum intel_dpll_id id)
+{
+	switch (id) {
+	default:
+		MISSING_CASE(id);
+	case DPLL_ID_ICL_DPLL0:
+	case DPLL_ID_ICL_DPLL1:
+		return CNL_DPLL_ENABLE(id);
+	case DPLL_ID_ICL_MGPLL1:
+	case DPLL_ID_ICL_MGPLL2:
+	case DPLL_ID_ICL_MGPLL3:
+	case DPLL_ID_ICL_MGPLL4:
+		return MG_PLL_ENABLE(icl_mg_pll_id_to_port(id));
+	}
+}
+
+static bool icl_pll_get_hw_state(struct drm_i915_private *dev_priv,
+				 struct intel_shared_dpll *pll,
+				 struct intel_dpll_hw_state *hw_state)
+{
+	const enum intel_dpll_id id = pll->info->id;
+	uint32_t val;
+	enum port port;
+	bool ret = false;
+
+	if (!intel_display_power_get_if_enabled(dev_priv, POWER_DOMAIN_PLLS))
+		return false;
+
+	val = I915_READ(icl_pll_id_to_enable_reg(id));
+	if (!(val & PLL_ENABLE))
+		goto out;
+
+	switch (id) {
+	case DPLL_ID_ICL_DPLL0:
+	case DPLL_ID_ICL_DPLL1:
+		hw_state->cfgcr0 = I915_READ(ICL_DPLL_CFGCR0(id));
+		hw_state->cfgcr1 = I915_READ(ICL_DPLL_CFGCR1(id));
+		break;
+	case DPLL_ID_ICL_MGPLL1:
+	case DPLL_ID_ICL_MGPLL2:
+	case DPLL_ID_ICL_MGPLL3:
+	case DPLL_ID_ICL_MGPLL4:
+		port = icl_mg_pll_id_to_port(id);
+		hw_state->mg_refclkin_ctl = I915_READ(MG_REFCLKIN_CTL(port));
+		hw_state->mg_clktop2_coreclkctl1 =
+			I915_READ(MG_CLKTOP2_CORECLKCTL1(port));
+		hw_state->mg_clktop2_hsclkctl =
+			I915_READ(MG_CLKTOP2_HSCLKCTL(port));
+		hw_state->mg_pll_div0 = I915_READ(MG_PLL_DIV0(port));
+		hw_state->mg_pll_div1 = I915_READ(MG_PLL_DIV1(port));
+		hw_state->mg_pll_lf = I915_READ(MG_PLL_LF(port));
+		hw_state->mg_pll_frac_lock = I915_READ(MG_PLL_FRAC_LOCK(port));
+		hw_state->mg_pll_ssc = I915_READ(MG_PLL_SSC(port));
+		hw_state->mg_pll_bias = I915_READ(MG_PLL_BIAS(port));
+		hw_state->mg_pll_tdc_coldst_bias =
+			I915_READ(MG_PLL_TDC_COLDST_BIAS(port));
+		break;
+	default:
+		MISSING_CASE(id);
+	}
+
+	ret = true;
+out:
+	intel_display_power_put(dev_priv, POWER_DOMAIN_PLLS);
+	return ret;
+}
+
+static void icl_dpll_write(struct drm_i915_private *dev_priv,
+			   struct intel_shared_dpll *pll)
+{
+	struct intel_dpll_hw_state *hw_state = &pll->state.hw_state;
+	const enum intel_dpll_id id = pll->info->id;
+
+	I915_WRITE(ICL_DPLL_CFGCR0(id), hw_state->cfgcr0);
+	I915_WRITE(ICL_DPLL_CFGCR1(id), hw_state->cfgcr1);
+	POSTING_READ(ICL_DPLL_CFGCR1(id));
+}
+
+static void icl_mg_pll_write(struct drm_i915_private *dev_priv,
+			     struct intel_shared_dpll *pll)
+{
+	struct intel_dpll_hw_state *hw_state = &pll->state.hw_state;
+	enum port port = icl_mg_pll_id_to_port(pll->info->id);
+
+	I915_WRITE(MG_REFCLKIN_CTL(port), hw_state->mg_refclkin_ctl);
+	I915_WRITE(MG_CLKTOP2_CORECLKCTL1(port),
+		   hw_state->mg_clktop2_coreclkctl1);
+	I915_WRITE(MG_CLKTOP2_HSCLKCTL(port), hw_state->mg_clktop2_hsclkctl);
+	I915_WRITE(MG_PLL_DIV0(port), hw_state->mg_pll_div0);
+	I915_WRITE(MG_PLL_DIV1(port), hw_state->mg_pll_div1);
+	I915_WRITE(MG_PLL_LF(port), hw_state->mg_pll_lf);
+	I915_WRITE(MG_PLL_FRAC_LOCK(port), hw_state->mg_pll_frac_lock);
+	I915_WRITE(MG_PLL_SSC(port), hw_state->mg_pll_ssc);
+	I915_WRITE(MG_PLL_BIAS(port), hw_state->mg_pll_bias);
+	I915_WRITE(MG_PLL_TDC_COLDST_BIAS(port),
+		   hw_state->mg_pll_tdc_coldst_bias);
+	POSTING_READ(MG_PLL_TDC_COLDST_BIAS(port));
+}
+
+static void icl_pll_enable(struct drm_i915_private *dev_priv,
+			   struct intel_shared_dpll *pll)
+{
+	const enum intel_dpll_id id = pll->info->id;
+	i915_reg_t enable_reg = icl_pll_id_to_enable_reg(id);
+	uint32_t val;
+
+	val = I915_READ(enable_reg);
+	val |= PLL_POWER_ENABLE;
+	I915_WRITE(enable_reg, val);
+
+	/*
+	 * The spec says we need to "wait" but it also says it should be
+	 * immediate.
+	 */
+	if (intel_wait_for_register(dev_priv, enable_reg, PLL_POWER_STATE,
+				    PLL_POWER_STATE, 1))
+		DRM_ERROR("PLL %d Power not enabled\n", id);
+
+	switch (id) {
+	case DPLL_ID_ICL_DPLL0:
+	case DPLL_ID_ICL_DPLL1:
+		icl_dpll_write(dev_priv, pll);
+		break;
+	case DPLL_ID_ICL_MGPLL1:
+	case DPLL_ID_ICL_MGPLL2:
+	case DPLL_ID_ICL_MGPLL3:
+	case DPLL_ID_ICL_MGPLL4:
+		icl_mg_pll_write(dev_priv, pll);
+		break;
+	default:
+		MISSING_CASE(id);
+	}
+
+	/*
+	 * DVFS pre sequence would be here, but in our driver the cdclk code
+	 * paths should already be setting the appropriate voltage, hence we do
+	 * nothign here.
+	 */
+
+	val = I915_READ(enable_reg);
+	val |= PLL_ENABLE;
+	I915_WRITE(enable_reg, val);
+
+	if (intel_wait_for_register(dev_priv, enable_reg, PLL_LOCK, PLL_LOCK,
+				    1)) /* 600us actually. */
+		DRM_ERROR("PLL %d not locked\n", id);
+
+	/* DVFS post sequence would be here. See the comment above. */
+}
+
+static void icl_pll_disable(struct drm_i915_private *dev_priv,
+			    struct intel_shared_dpll *pll)
+{
+	const enum intel_dpll_id id = pll->info->id;
+	i915_reg_t enable_reg = icl_pll_id_to_enable_reg(id);
+	uint32_t val;
+
+	/* The first steps are done by intel_ddi_post_disable(). */
+
+	/*
+	 * DVFS pre sequence would be here, but in our driver the cdclk code
+	 * paths should already be setting the appropriate voltage, hence we do
+	 * nothign here.
+	 */
+
+	val = I915_READ(enable_reg);
+	val &= ~PLL_ENABLE;
+	I915_WRITE(enable_reg, val);
+
+	/* Timeout is actually 1us. */
+	if (intel_wait_for_register(dev_priv, enable_reg, PLL_LOCK, 0, 1))
+		DRM_ERROR("PLL %d locked\n", id);
+
+	/* DVFS post sequence would be here. See the comment above. */
+
+	val = I915_READ(enable_reg);
+	val &= ~PLL_POWER_ENABLE;
+	I915_WRITE(enable_reg, val);
+
+	/*
+	 * The spec says we need to "wait" but it also says it should be
+	 * immediate.
+	 */
+	if (intel_wait_for_register(dev_priv, enable_reg, PLL_POWER_STATE, 0,
+				    1))
+		DRM_ERROR("PLL %d Power not disabled\n", id);
+}
+
+static void icl_dump_hw_state(struct drm_i915_private *dev_priv,
+			      struct intel_dpll_hw_state *hw_state)
+{
+	DRM_DEBUG_KMS("dpll_hw_state: cfgcr0: 0x%x, cfgcr1: 0x%x, "
+		      "mg_refclkin_ctl: 0x%x, hg_clktop2_coreclkctl1: 0x%x, "
+		      "mg_clktop2_hsclkctl: 0x%x, mg_pll_div0: 0x%x, "
+		      "mg_pll_div2: 0x%x, mg_pll_lf: 0x%x, "
+		      "mg_pll_frac_lock: 0x%x, mg_pll_ssc: 0x%x, "
+		      "mg_pll_bias: 0x%x, mg_pll_tdc_coldst_bias: 0x%x\n",
+		      hw_state->cfgcr0, hw_state->cfgcr1,
+		      hw_state->mg_refclkin_ctl,
+		      hw_state->mg_clktop2_coreclkctl1,
+		      hw_state->mg_clktop2_hsclkctl,
+		      hw_state->mg_pll_div0,
+		      hw_state->mg_pll_div1,
+		      hw_state->mg_pll_lf,
+		      hw_state->mg_pll_frac_lock,
+		      hw_state->mg_pll_ssc,
+		      hw_state->mg_pll_bias,
+		      hw_state->mg_pll_tdc_coldst_bias);
+}
+
+static const struct intel_shared_dpll_funcs icl_pll_funcs = {
+	.enable = icl_pll_enable,
+	.disable = icl_pll_disable,
+	.get_hw_state = icl_pll_get_hw_state,
+};
+
+static const struct dpll_info icl_plls[] = {
+	{ "DPLL 0",   &icl_pll_funcs, DPLL_ID_ICL_DPLL0,  0 },
+	{ "DPLL 1",   &icl_pll_funcs, DPLL_ID_ICL_DPLL1,  0 },
+	{ "MG PLL 1", &icl_pll_funcs, DPLL_ID_ICL_MGPLL1, 0 },
+	{ "MG PLL 2", &icl_pll_funcs, DPLL_ID_ICL_MGPLL2, 0 },
+	{ "MG PLL 3", &icl_pll_funcs, DPLL_ID_ICL_MGPLL3, 0 },
+	{ "MG PLL 4", &icl_pll_funcs, DPLL_ID_ICL_MGPLL4, 0 },
+	{ },
+};
+
+static const struct intel_dpll_mgr icl_pll_mgr = {
+	.dpll_info = icl_plls,
+	.get_dpll = icl_get_dpll,
+	.dump_hw_state = icl_dump_hw_state,
+};
+
 /**
  * intel_shared_dpll_init - Initialize shared DPLLs
  * @dev: drm device
@@ -2412,7 +2721,9 @@ void intel_shared_dpll_init(struct drm_device *dev)
 	const struct dpll_info *dpll_info;
 	int i;
 
-	if (IS_CANNONLAKE(dev_priv))
+	if (IS_ICELAKE(dev_priv))
+		dpll_mgr = &icl_pll_mgr;
+	else if (IS_CANNONLAKE(dev_priv))
 		dpll_mgr = &cnl_pll_mgr;
 	else if (IS_GEN9_BC(dev_priv))
 		dpll_mgr = &skl_pll_mgr;
diff --git a/drivers/gpu/drm/i915/intel_dpll_mgr.h b/drivers/gpu/drm/i915/intel_dpll_mgr.h
index 4febfaa..7a0cd56 100644
--- a/drivers/gpu/drm/i915/intel_dpll_mgr.h
+++ b/drivers/gpu/drm/i915/intel_dpll_mgr.h
@@ -103,6 +103,32 @@ enum intel_dpll_id {
 	 * @DPLL_ID_SKL_DPLL3: SKL and later DPLL3
 	 */
 	DPLL_ID_SKL_DPLL3 = 3,
+
+
+	/**
+	 * @DPLL_ID_ICL_DPLL0: ICL combo PHY DPLL0
+	 */
+	DPLL_ID_ICL_DPLL0 = 0,
+	/**
+	 * @DPLL_ID_ICL_DPLL1: ICL combo PHY DPLL1
+	 */
+	DPLL_ID_ICL_DPLL1 = 1,
+	/**
+	 * @DPLL_ID_ICL_MGPLL1: ICL MG PLL 1 port 1 (C)
+	 */
+	DPLL_ID_ICL_MGPLL1 = 2,
+	/**
+	 * @DPLL_ID_ICL_MGPLL2: ICL MG PLL 1 port 2 (D)
+	 */
+	DPLL_ID_ICL_MGPLL2 = 3,
+	/**
+	 * @DPLL_ID_ICL_MGPLL3: ICL MG PLL 1 port 3 (E)
+	 */
+	DPLL_ID_ICL_MGPLL3 = 4,
+	/**
+	 * @DPLL_ID_ICL_MGPLL4: ICL MG PLL 1 port 4 (F)
+	 */
+	DPLL_ID_ICL_MGPLL4 = 5,
 };
 #define I915_NUM_PLLS 6
 
@@ -135,6 +161,21 @@ struct intel_dpll_hw_state {
 	/* bxt */
 	uint32_t ebb0, ebb4, pll0, pll1, pll2, pll3, pll6, pll8, pll9, pll10,
 		 pcsdw12;
+
+	/*
+	 * ICL uses the following, already defined:
+	 * uint32_t cfgcr0, cfgcr1;
+	 */
+	uint32_t mg_refclkin_ctl;
+	uint32_t mg_clktop2_coreclkctl1;
+	uint32_t mg_clktop2_hsclkctl;
+	uint32_t mg_pll_div0;
+	uint32_t mg_pll_div1;
+	uint32_t mg_pll_lf;
+	uint32_t mg_pll_frac_lock;
+	uint32_t mg_pll_ssc;
+	uint32_t mg_pll_bias;
+	uint32_t mg_pll_tdc_coldst_bias;
 };
 
 /**
diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
index 11a1932..52337f4 100644
--- a/drivers/gpu/drm/i915/intel_drv.h
+++ b/drivers/gpu/drm/i915/intel_drv.h
@@ -1409,6 +1409,12 @@ uint32_t ddi_signal_levels(struct intel_dp *intel_dp);
 u8 intel_ddi_dp_voltage_max(struct intel_encoder *encoder);
 int intel_ddi_toggle_hdcp_signalling(struct intel_encoder *intel_encoder,
 				     bool enable);
+void icl_map_plls_to_ports(struct drm_crtc *crtc,
+			   struct intel_crtc_state *crtc_state,
+			   struct drm_atomic_state *old_state);
+void icl_unmap_plls_to_ports(struct drm_crtc *crtc,
+			     struct intel_crtc_state *crtc_state,
+			     struct drm_atomic_state *old_state);
 
 unsigned int intel_fb_align_height(const struct drm_framebuffer *fb,
 				   int plane, unsigned int height);
-- 
cgit v1.1


From febafb93181e4fb4de19f4484df62ce2d04155aa Mon Sep 17 00:00:00 2001
From: Paulo Zanoni <paulo.r.zanoni@intel.com>
Date: Wed, 28 Mar 2018 14:57:59 -0700
Subject: drm/i915/icl: compute the combo PHY (DPLL) HDMI registers

HDMI mode DPLL programming on ICL is the same as CNL, so just reuse
the CNL code.

v2:
 - Properly detect HDMI crtcs.
 - Rebase after changes to the cnl function (clock * 1000).
v3:
 - Add a comment to clarify why we treat 38.4 as 19.2 (James).

Reviewed-by: James Ausmus <james.ausmus@intel.com>
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180328215803.13835-5-paulo.r.zanoni@intel.com
---
 drivers/gpu/drm/i915/intel_dpll_mgr.c | 39 ++++++++++++++++++++++++++++++++---
 1 file changed, 36 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_dpll_mgr.c b/drivers/gpu/drm/i915/intel_dpll_mgr.c
index 14f5414..a1c2bd1 100644
--- a/drivers/gpu/drm/i915/intel_dpll_mgr.c
+++ b/drivers/gpu/drm/i915/intel_dpll_mgr.c
@@ -2218,6 +2218,7 @@ cnl_ddi_calculate_wrpll(int clock,
 			struct skl_wrpll_params *wrpll_params)
 {
 	u32 afe_clock = clock * 5;
+	uint32_t ref_clock;
 	u32 dco_min = 7998000;
 	u32 dco_max = 10000000;
 	u32 dco_mid = (dco_min + dco_max) / 2;
@@ -2250,8 +2251,17 @@ cnl_ddi_calculate_wrpll(int clock,
 
 	cnl_wrpll_get_multipliers(best_div, &pdiv, &qdiv, &kdiv);
 
-	cnl_wrpll_params_populate(wrpll_params, best_dco,
-				  dev_priv->cdclk.hw.ref, pdiv, qdiv, kdiv);
+	ref_clock = dev_priv->cdclk.hw.ref;
+
+	/*
+	 * For ICL, the spec states: if reference frequency is 38.4, use 19.2
+	 * because the DPLL automatically divides that by 2.
+	 */
+	if (IS_ICELAKE(dev_priv) && ref_clock == 38400)
+		ref_clock = 19200;
+
+	cnl_wrpll_params_populate(wrpll_params, best_dco, ref_clock, pdiv, qdiv,
+				  kdiv);
 
 	return true;
 }
@@ -2403,7 +2413,30 @@ static bool icl_calc_dpll_state(struct intel_crtc_state *crtc_state,
 				struct intel_encoder *encoder, int clock,
 				struct intel_dpll_hw_state *pll_state)
 {
-	/* TODO */
+	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
+	uint32_t cfgcr0, cfgcr1;
+	struct skl_wrpll_params pll_params = { 0 };
+	bool ret;
+
+	if (intel_crtc_has_type(crtc_state, INTEL_OUTPUT_HDMI))
+		ret = cnl_ddi_calculate_wrpll(clock, dev_priv, &pll_params);
+	else
+		ret = false; /* TODO */
+
+	if (!ret)
+		return false;
+
+	cfgcr0 = DPLL_CFGCR0_DCO_FRACTION(pll_params.dco_fraction) |
+		 pll_params.dco_integer;
+
+	cfgcr1 = DPLL_CFGCR1_QDIV_RATIO(pll_params.qdiv_ratio) |
+		 DPLL_CFGCR1_QDIV_MODE(pll_params.qdiv_mode) |
+		 DPLL_CFGCR1_KDIV(pll_params.kdiv) |
+		 DPLL_CFGCR1_PDIV(pll_params.pdiv) |
+		 DPLL_CFGCR1_CENTRAL_FREQ_8400;
+
+	pll_state->cfgcr0 = cfgcr0;
+	pll_state->cfgcr1 = cfgcr1;
 	return true;
 }
 
-- 
cgit v1.1


From bb82139b4bbf8bdb825a7339d34d231632e67f27 Mon Sep 17 00:00:00 2001
From: Paulo Zanoni <paulo.r.zanoni@intel.com>
Date: Wed, 28 Mar 2018 14:58:00 -0700
Subject: drm/i915/icl: compute the combo PHY (DPLL) DP registers

Just use the hardcoded tables provided by our spec.

v2: Rebase.
v3: Clarify that 38.4 uses the 19.2 table (James).

Reviewed-by: James Ausmus <james.ausmus@intel.com>
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180328215803.13835-6-paulo.r.zanoni@intel.com
---
 drivers/gpu/drm/i915/intel_dpll_mgr.c | 87 ++++++++++++++++++++++++++++++++++-
 1 file changed, 86 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/intel_dpll_mgr.c b/drivers/gpu/drm/i915/intel_dpll_mgr.c
index a1c2bd1..30eca81 100644
--- a/drivers/gpu/drm/i915/intel_dpll_mgr.c
+++ b/drivers/gpu/drm/i915/intel_dpll_mgr.c
@@ -2409,6 +2409,91 @@ static const struct intel_dpll_mgr cnl_pll_mgr = {
 	.dump_hw_state = cnl_dump_hw_state,
 };
 
+/*
+ * These values alrea already adjusted: they're the bits we write to the
+ * registers, not the logical values.
+ */
+static const struct skl_wrpll_params icl_dp_combo_pll_24MHz_values[] = {
+	{ .dco_integer = 0x151, .dco_fraction = 0x4000,		/* [0]: 5.4 */
+	  .pdiv = 0x2 /* 3 */, .kdiv = 1, .qdiv_mode = 0, .qdiv_ratio = 0},
+	{ .dco_integer = 0x151, .dco_fraction = 0x4000,		/* [1]: 2.7 */
+	  .pdiv = 0x2 /* 3 */, .kdiv = 2, .qdiv_mode = 0, .qdiv_ratio = 0},
+	{ .dco_integer = 0x151, .dco_fraction = 0x4000,		/* [2]: 1.62 */
+	  .pdiv = 0x4 /* 5 */, .kdiv = 2, .qdiv_mode = 0, .qdiv_ratio = 0},
+	{ .dco_integer = 0x151, .dco_fraction = 0x4000,		/* [3]: 3.24 */
+	  .pdiv = 0x4 /* 5 */, .kdiv = 1, .qdiv_mode = 0, .qdiv_ratio = 0},
+	{ .dco_integer = 0x168, .dco_fraction = 0x0000,		/* [4]: 2.16 */
+	  .pdiv = 0x1 /* 2 */, .kdiv = 2, .qdiv_mode = 1, .qdiv_ratio = 2},
+	{ .dco_integer = 0x168, .dco_fraction = 0x0000,		/* [5]: 4.32 */
+	  .pdiv = 0x1 /* 2 */, .kdiv = 2, .qdiv_mode = 0, .qdiv_ratio = 0},
+	{ .dco_integer = 0x195, .dco_fraction = 0x0000,		/* [6]: 6.48 */
+	  .pdiv = 0x2 /* 3 */, .kdiv = 1, .qdiv_mode = 0, .qdiv_ratio = 0},
+	{ .dco_integer = 0x151, .dco_fraction = 0x4000,		/* [7]: 8.1 */
+	  .pdiv = 0x1 /* 2 */, .kdiv = 1, .qdiv_mode = 0, .qdiv_ratio = 0},
+};
+
+/* Also used for 38.4 MHz values. */
+static const struct skl_wrpll_params icl_dp_combo_pll_19_2MHz_values[] = {
+	{ .dco_integer = 0x1A5, .dco_fraction = 0x7000,		/* [0]: 5.4 */
+	  .pdiv = 0x2 /* 3 */, .kdiv = 1, .qdiv_mode = 0, .qdiv_ratio = 0},
+	{ .dco_integer = 0x1A5, .dco_fraction = 0x7000,		/* [1]: 2.7 */
+	  .pdiv = 0x2 /* 3 */, .kdiv = 2, .qdiv_mode = 0, .qdiv_ratio = 0},
+	{ .dco_integer = 0x1A5, .dco_fraction = 0x7000,		/* [2]: 1.62 */
+	  .pdiv = 0x4 /* 5 */, .kdiv = 2, .qdiv_mode = 0, .qdiv_ratio = 0},
+	{ .dco_integer = 0x1A5, .dco_fraction = 0x7000,		/* [3]: 3.24 */
+	  .pdiv = 0x4 /* 5 */, .kdiv = 1, .qdiv_mode = 0, .qdiv_ratio = 0},
+	{ .dco_integer = 0x1C2, .dco_fraction = 0x0000,		/* [4]: 2.16 */
+	  .pdiv = 0x1 /* 2 */, .kdiv = 2, .qdiv_mode = 1, .qdiv_ratio = 2},
+	{ .dco_integer = 0x1C2, .dco_fraction = 0x0000,		/* [5]: 4.32 */
+	  .pdiv = 0x1 /* 2 */, .kdiv = 2, .qdiv_mode = 0, .qdiv_ratio = 0},
+	{ .dco_integer = 0x1FA, .dco_fraction = 0x2000,		/* [6]: 6.48 */
+	  .pdiv = 0x2 /* 3 */, .kdiv = 1, .qdiv_mode = 0, .qdiv_ratio = 0},
+	{ .dco_integer = 0x1A5, .dco_fraction = 0x7000,		/* [7]: 8.1 */
+	  .pdiv = 0x1 /* 2 */, .kdiv = 1, .qdiv_mode = 0, .qdiv_ratio = 0},
+};
+
+static bool icl_calc_dp_combo_pll(struct drm_i915_private *dev_priv, int clock,
+				  struct skl_wrpll_params *pll_params)
+{
+	const struct skl_wrpll_params *params;
+
+	params = dev_priv->cdclk.hw.ref == 24000 ?
+			icl_dp_combo_pll_24MHz_values :
+			icl_dp_combo_pll_19_2MHz_values;
+
+	switch (clock) {
+	case 540000:
+		*pll_params = params[0];
+		break;
+	case 270000:
+		*pll_params = params[1];
+		break;
+	case 162000:
+		*pll_params = params[2];
+		break;
+	case 324000:
+		*pll_params = params[3];
+		break;
+	case 216000:
+		*pll_params = params[4];
+		break;
+	case 432000:
+		*pll_params = params[5];
+		break;
+	case 648000:
+		*pll_params = params[6];
+		break;
+	case 810000:
+		*pll_params = params[7];
+		break;
+	default:
+		MISSING_CASE(clock);
+		return false;
+	}
+
+	return true;
+}
+
 static bool icl_calc_dpll_state(struct intel_crtc_state *crtc_state,
 				struct intel_encoder *encoder, int clock,
 				struct intel_dpll_hw_state *pll_state)
@@ -2421,7 +2506,7 @@ static bool icl_calc_dpll_state(struct intel_crtc_state *crtc_state,
 	if (intel_crtc_has_type(crtc_state, INTEL_OUTPUT_HDMI))
 		ret = cnl_ddi_calculate_wrpll(clock, dev_priv, &pll_params);
 	else
-		ret = false; /* TODO */
+		ret = icl_calc_dp_combo_pll(dev_priv, clock, &pll_params);
 
 	if (!ret)
 		return false;
-- 
cgit v1.1


From 145ef0d17d57788293b5569cd03118fc0e220a61 Mon Sep 17 00:00:00 2001
From: Paulo Zanoni <paulo.r.zanoni@intel.com>
Date: Wed, 28 Mar 2018 14:58:01 -0700
Subject: drm/i915/icl: compute the MG PLL registers

This implements the "MG PLL Programming" sequence from our spec. The
biggest problem was that the spec assumes real numbers, so we had to
adjust some numbers and calculations due to the fact that the Kernel
prefers to deal with integers.

I recommend grabbing some coffee, a pen and paper before reviewing
this patch.

v2:
 - Correctly identify DP encoders after upstream change.
 - Small checkpatch issues.
 - Rebase.
v3:
 - Try to impove the comment on the tdc_targetcnt calculation based on
   Manasi's feedback (Manasi).
 - Rebase.

Reviewed-by: Manasi Navare <manasi.d.navare@intel.com>
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180328215803.13835-7-paulo.r.zanoni@intel.com
---
 drivers/gpu/drm/i915/intel_dpll_mgr.c | 223 +++++++++++++++++++++++++++++++++-
 1 file changed, 222 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/intel_dpll_mgr.c b/drivers/gpu/drm/i915/intel_dpll_mgr.c
index 30eca81..383fbc1 100644
--- a/drivers/gpu/drm/i915/intel_dpll_mgr.c
+++ b/drivers/gpu/drm/i915/intel_dpll_mgr.c
@@ -2535,11 +2535,232 @@ static enum intel_dpll_id icl_port_to_mg_pll_id(enum port port)
 	return port - PORT_C + DPLL_ID_ICL_MGPLL1;
 }
 
+static bool icl_mg_pll_find_divisors(int clock_khz, bool is_dp, bool use_ssc,
+				     uint32_t *target_dco_khz,
+				     struct intel_dpll_hw_state *state)
+{
+	uint32_t dco_min_freq, dco_max_freq;
+	int div1_vals[] = {7, 5, 3, 2};
+	unsigned int i;
+	int div2;
+
+	dco_min_freq = is_dp ? 8100000 : use_ssc ? 8000000 : 7992000;
+	dco_max_freq = is_dp ? 8100000 : 10000000;
+
+	for (i = 0; i < ARRAY_SIZE(div1_vals); i++) {
+		int div1 = div1_vals[i];
+
+		for (div2 = 10; div2 > 0; div2--) {
+			int dco = div1 * div2 * clock_khz * 5;
+			int a_divratio, tlinedrv, inputsel, hsdiv;
+
+			if (dco < dco_min_freq || dco > dco_max_freq)
+				continue;
+
+			if (div2 >= 2) {
+				a_divratio = is_dp ? 10 : 5;
+				tlinedrv = 2;
+			} else {
+				a_divratio = 5;
+				tlinedrv = 0;
+			}
+			inputsel = is_dp ? 0 : 1;
+
+			switch (div1) {
+			default:
+				MISSING_CASE(div1);
+			case 2:
+				hsdiv = 0;
+				break;
+			case 3:
+				hsdiv = 1;
+				break;
+			case 5:
+				hsdiv = 2;
+				break;
+			case 7:
+				hsdiv = 3;
+				break;
+			}
+
+			*target_dco_khz = dco;
+
+			state->mg_refclkin_ctl = MG_REFCLKIN_CTL_OD_2_MUX(1);
+
+			state->mg_clktop2_coreclkctl1 =
+				MG_CLKTOP2_CORECLKCTL1_A_DIVRATIO(a_divratio);
+
+			state->mg_clktop2_hsclkctl =
+				MG_CLKTOP2_HSCLKCTL_TLINEDRV_CLKSEL(tlinedrv) |
+				MG_CLKTOP2_HSCLKCTL_CORE_INPUTSEL(inputsel) |
+				MG_CLKTOP2_HSCLKCTL_HSDIV_RATIO(hsdiv) |
+				MG_CLKTOP2_HSCLKCTL_DSDIV_RATIO(div2);
+
+			return true;
+		}
+	}
+
+	return false;
+}
+
+/*
+ * The specification for this function uses real numbers, so the math had to be
+ * adapted to integer-only calculation, that's why it looks so different.
+ */
 static bool icl_calc_mg_pll_state(struct intel_crtc_state *crtc_state,
 				  struct intel_encoder *encoder, int clock,
 				  struct intel_dpll_hw_state *pll_state)
 {
-	/* TODO */
+	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
+	int refclk_khz = dev_priv->cdclk.hw.ref;
+	uint32_t dco_khz, m1div, m2div_int, m2div_rem, m2div_frac;
+	uint32_t iref_ndiv, iref_trim, iref_pulse_w;
+	uint32_t prop_coeff, int_coeff;
+	uint32_t tdc_targetcnt, feedfwgain;
+	uint64_t ssc_stepsize, ssc_steplen, ssc_steplog;
+	uint64_t tmp;
+	bool use_ssc = false;
+	bool is_dp = !intel_crtc_has_type(crtc_state, INTEL_OUTPUT_HDMI);
+
+	if (!icl_mg_pll_find_divisors(clock, is_dp, use_ssc, &dco_khz,
+				      pll_state)) {
+		DRM_DEBUG_KMS("Failed to find divisors for clock %d\n", clock);
+		return false;
+	}
+
+	m1div = 2;
+	m2div_int = dco_khz / (refclk_khz * m1div);
+	if (m2div_int > 255) {
+		m1div = 4;
+		m2div_int = dco_khz / (refclk_khz * m1div);
+		if (m2div_int > 255) {
+			DRM_DEBUG_KMS("Failed to find mdiv for clock %d\n",
+				      clock);
+			return false;
+		}
+	}
+	m2div_rem = dco_khz % (refclk_khz * m1div);
+
+	tmp = (uint64_t)m2div_rem * (1 << 22);
+	do_div(tmp, refclk_khz * m1div);
+	m2div_frac = tmp;
+
+	switch (refclk_khz) {
+	case 19200:
+		iref_ndiv = 1;
+		iref_trim = 28;
+		iref_pulse_w = 1;
+		break;
+	case 24000:
+		iref_ndiv = 1;
+		iref_trim = 25;
+		iref_pulse_w = 2;
+		break;
+	case 38400:
+		iref_ndiv = 2;
+		iref_trim = 28;
+		iref_pulse_w = 1;
+		break;
+	default:
+		MISSING_CASE(refclk_khz);
+		return false;
+	}
+
+	/*
+	 * tdc_res = 0.000003
+	 * tdc_targetcnt = int(2 / (tdc_res * 8 * 50 * 1.1) / refclk_mhz + 0.5)
+	 *
+	 * The multiplication by 1000 is due to refclk MHz to KHz conversion. It
+	 * was supposed to be a division, but we rearranged the operations of
+	 * the formula to avoid early divisions so we don't multiply the
+	 * rounding errors.
+	 *
+	 * 0.000003 * 8 * 50 * 1.1 = 0.00132, also known as 132 / 100000, which
+	 * we also rearrange to work with integers.
+	 *
+	 * The 0.5 transformed to 5 results in a multiplication by 10 and the
+	 * last division by 10.
+	 */
+	tdc_targetcnt = (2 * 1000 * 100000 * 10 / (132 * refclk_khz) + 5) / 10;
+
+	/*
+	 * Here we divide dco_khz by 10 in order to allow the dividend to fit in
+	 * 32 bits. That's not a problem since we round the division down
+	 * anyway.
+	 */
+	feedfwgain = (use_ssc || m2div_rem > 0) ?
+		m1div * 1000000 * 100 / (dco_khz * 3 / 10) : 0;
+
+	if (dco_khz >= 9000000) {
+		prop_coeff = 5;
+		int_coeff = 10;
+	} else {
+		prop_coeff = 4;
+		int_coeff = 8;
+	}
+
+	if (use_ssc) {
+		tmp = (uint64_t)dco_khz * 47 * 32;
+		do_div(tmp, refclk_khz * m1div * 10000);
+		ssc_stepsize = tmp;
+
+		tmp = (uint64_t)dco_khz * 1000;
+		ssc_steplen = DIV_ROUND_UP_ULL(tmp, 32 * 2 * 32);
+	} else {
+		ssc_stepsize = 0;
+		ssc_steplen = 0;
+	}
+	ssc_steplog = 4;
+
+	pll_state->mg_pll_div0 = (m2div_rem > 0 ? MG_PLL_DIV0_FRACNEN_H : 0) |
+				  MG_PLL_DIV0_FBDIV_FRAC(m2div_frac) |
+				  MG_PLL_DIV0_FBDIV_INT(m2div_int);
+
+	pll_state->mg_pll_div1 = MG_PLL_DIV1_IREF_NDIVRATIO(iref_ndiv) |
+				 MG_PLL_DIV1_DITHER_DIV_2 |
+				 MG_PLL_DIV1_NDIVRATIO(1) |
+				 MG_PLL_DIV1_FBPREDIV(m1div);
+
+	pll_state->mg_pll_lf = MG_PLL_LF_TDCTARGETCNT(tdc_targetcnt) |
+			       MG_PLL_LF_AFCCNTSEL_512 |
+			       MG_PLL_LF_GAINCTRL(1) |
+			       MG_PLL_LF_INT_COEFF(int_coeff) |
+			       MG_PLL_LF_PROP_COEFF(prop_coeff);
+
+	pll_state->mg_pll_frac_lock = MG_PLL_FRAC_LOCK_TRUELOCK_CRIT_32 |
+				      MG_PLL_FRAC_LOCK_EARLYLOCK_CRIT_32 |
+				      MG_PLL_FRAC_LOCK_LOCKTHRESH(10) |
+				      MG_PLL_FRAC_LOCK_DCODITHEREN |
+				      MG_PLL_FRAC_LOCK_FEEDFWRDGAIN(feedfwgain);
+	if (use_ssc || m2div_rem > 0)
+		pll_state->mg_pll_frac_lock |= MG_PLL_FRAC_LOCK_FEEDFWRDCAL_EN;
+
+	pll_state->mg_pll_ssc = (use_ssc ? MG_PLL_SSC_EN : 0) |
+				MG_PLL_SSC_TYPE(2) |
+				MG_PLL_SSC_STEPLENGTH(ssc_steplen) |
+				MG_PLL_SSC_STEPNUM(ssc_steplog) |
+				MG_PLL_SSC_FLLEN |
+				MG_PLL_SSC_STEPSIZE(ssc_stepsize);
+
+	pll_state->mg_pll_tdc_coldst_bias = MG_PLL_TDC_COLDST_COLDSTART;
+
+	if (refclk_khz != 38400) {
+		pll_state->mg_pll_tdc_coldst_bias |=
+			MG_PLL_TDC_COLDST_IREFINT_EN |
+			MG_PLL_TDC_COLDST_REFBIAS_START_PULSE_W(iref_pulse_w) |
+			MG_PLL_TDC_COLDST_COLDSTART |
+			MG_PLL_TDC_TDCOVCCORR_EN |
+			MG_PLL_TDC_TDCSEL(3);
+
+		pll_state->mg_pll_bias = MG_PLL_BIAS_BIAS_GB_SEL(3) |
+					 MG_PLL_BIAS_INIT_DCOAMP(0x3F) |
+					 MG_PLL_BIAS_BIAS_BONUS(10) |
+					 MG_PLL_BIAS_BIASCAL_EN |
+					 MG_PLL_BIAS_CTRIM(12) |
+					 MG_PLL_BIAS_VREF_RDAC(4) |
+					 MG_PLL_BIAS_IREFTRIM(iref_trim);
+	}
+
 	return true;
 }
 
-- 
cgit v1.1


From 1751eb42ddb56b1e0a28ade5c5f19b7961a6f0c2 Mon Sep 17 00:00:00 2001
From: Anders Roxell <anders.roxell@linaro.org>
Date: Fri, 4 May 2018 18:47:25 +0200
Subject: selftests: net: use TEST_PROGS_EXTENDED

When a script file that isn't generated uses the variable
TEST_GEN_PROGS_EXTENDED and a 'make -C tools/testing/selftests clean' is
performed the script file gets removed and git shows the file as
deleted. For script files that isn't generated TEST_PROGS_EXTENDED
should be used.

Fixes: 9faedd643fd9 ("selftests: net: add in_netns.sh TEST_GEN_PROGS_EXTENDED")
Signed-off-by: Anders Roxell <anders.roxell@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 tools/testing/selftests/net/Makefile | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/testing/selftests/net/Makefile b/tools/testing/selftests/net/Makefile
index daf5eff..3ff81a4 100644
--- a/tools/testing/selftests/net/Makefile
+++ b/tools/testing/selftests/net/Makefile
@@ -6,7 +6,7 @@ CFLAGS += -I../../../../usr/include/
 
 TEST_PROGS := run_netsocktests run_afpackettests test_bpf.sh netdevice.sh rtnetlink.sh
 TEST_PROGS += fib_tests.sh fib-onlink-tests.sh pmtu.sh
-TEST_GEN_PROGS_EXTENDED := in_netns.sh
+TEST_PROGS_EXTENDED := in_netns.sh
 TEST_GEN_FILES =  socket
 TEST_GEN_FILES += psock_fanout psock_tpacket msg_zerocopy
 TEST_GEN_PROGS = reuseport_bpf reuseport_bpf_cpu reuseport_bpf_numa
-- 
cgit v1.1


From abcd3d6fc640aff48b17900734eff134f27fdf2a Mon Sep 17 00:00:00 2001
From: Moritz Fischer <mdf@kernel.org>
Date: Fri, 4 May 2018 10:18:33 -0700
Subject: net: nixge: Fix error path for obtaining mac address

Fix issue where nixge_get_nvmem_address() returns a non-NULL
return value on a failed nvmem_cell_get() that causes an invalid
access when error value encoded in pointer is dereferenced.

Furthermore ensure that buffer allocated by nvmem_cell_read()
actually gets kfreed() if the function succeeds.

Fixes commit 492caffa8a1a ("net: ethernet: nixge: Add support for
National Instruments XGE netdev")
Reported-by: Alex Williams <alex.williams@ni.com>
Signed-off-by: Moritz Fischer <mdf@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/ni/nixge.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ethernet/ni/nixge.c b/drivers/net/ethernet/ni/nixge.c
index 27364b7..c41fea9 100644
--- a/drivers/net/ethernet/ni/nixge.c
+++ b/drivers/net/ethernet/ni/nixge.c
@@ -1170,7 +1170,7 @@ static void *nixge_get_nvmem_address(struct device *dev)
 
 	cell = nvmem_cell_get(dev, "address");
 	if (IS_ERR(cell))
-		return cell;
+		return NULL;
 
 	mac = nvmem_cell_read(cell, &cell_size);
 	nvmem_cell_put(cell);
@@ -1202,10 +1202,12 @@ static int nixge_probe(struct platform_device *pdev)
 	ndev->max_mtu = NIXGE_JUMBO_MTU;
 
 	mac_addr = nixge_get_nvmem_address(&pdev->dev);
-	if (mac_addr && is_valid_ether_addr(mac_addr))
+	if (mac_addr && is_valid_ether_addr(mac_addr)) {
 		ether_addr_copy(ndev->dev_addr, mac_addr);
-	else
+		kfree(mac_addr);
+	} else {
 		eth_hw_addr_random(ndev);
+	}
 
 	priv = netdev_priv(ndev);
 	priv->ndev = ndev;
-- 
cgit v1.1


From a86b74d363708d0b04305f356fd3429c7b560a64 Mon Sep 17 00:00:00 2001
From: Moritz Fischer <mdf@kernel.org>
Date: Fri, 4 May 2018 10:18:34 -0700
Subject: net: nixge: Address compiler warnings about signedness
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Fixes the following warnings:
warning: pointer targets in passing argument 1 of
‘is_valid_ether_addr’ differ in signedness [-Wpointer-sign]
  if (mac_addr && is_valid_ether_addr(mac_addr)) {
                                      ^~~~~~~~
expected ‘const u8 * {aka const unsigned char *}’ but argument
is of type ‘const char *’
 static inline bool is_valid_ether_addr(const u8 *addr)
                    ^~~~~~~~~~~~~~~~~~~
warning: pointer targets in passing argument 2 of
‘ether_addr_copy’ differ in signedness [-Wpointer-sign]
   ether_addr_copy(ndev->dev_addr, mac_addr);
                                   ^~~~~~~~
expected ‘const u8 * {aka const unsigned char *}’ but argument
is of type ‘const char *’
 static inline void ether_addr_copy(u8 *dst, const u8 *src)

Signed-off-by: Moritz Fischer <mdf@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/ni/nixge.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/ni/nixge.c b/drivers/net/ethernet/ni/nixge.c
index c41fea9..b092894 100644
--- a/drivers/net/ethernet/ni/nixge.c
+++ b/drivers/net/ethernet/ni/nixge.c
@@ -1183,7 +1183,7 @@ static int nixge_probe(struct platform_device *pdev)
 	struct nixge_priv *priv;
 	struct net_device *ndev;
 	struct resource *dmares;
-	const char *mac_addr;
+	const u8 *mac_addr;
 	int err;
 
 	ndev = alloc_etherdev(sizeof(*priv));
-- 
cgit v1.1


From 59d8d4434f429b4fa8a346fd889058bda427a837 Mon Sep 17 00:00:00 2001
From: Xin Long <lucien.xin@gmail.com>
Date: Sat, 5 May 2018 14:59:47 +0800
Subject: sctp: delay the authentication for the duplicated cookie-echo chunk

Now sctp only delays the authentication for the normal cookie-echo
chunk by setting chunk->auth_chunk in sctp_endpoint_bh_rcv(). But
for the duplicated one with auth, in sctp_assoc_bh_rcv(), it does
authentication first based on the old asoc, which will definitely
fail due to the different auth info in the old asoc.

The duplicated cookie-echo chunk will create a new asoc with the
auth info from this chunk, and the authentication should also be
done with the new asoc's auth info for all of the collision 'A',
'B' and 'D'. Otherwise, the duplicated cookie-echo chunk with auth
will never pass the authentication and create the new connection.

This issue exists since very beginning, and this fix is to make
sctp_assoc_bh_rcv() follow the way sctp_endpoint_bh_rcv() does
for the normal cookie-echo chunk to delay the authentication.

While at it, remove the unused params from sctp_sf_authenticate()
and define sctp_auth_chunk_verify() used for all the places that
do the delayed authentication.

v1->v2:
  fix the typo in changelog as Marcelo noticed.

Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/sctp/associola.c    | 30 ++++++++++++++++-
 net/sctp/sm_statefuns.c | 86 ++++++++++++++++++++++++++-----------------------
 2 files changed, 75 insertions(+), 41 deletions(-)

diff --git a/net/sctp/associola.c b/net/sctp/associola.c
index 837806d..a47179d 100644
--- a/net/sctp/associola.c
+++ b/net/sctp/associola.c
@@ -1024,8 +1024,9 @@ static void sctp_assoc_bh_rcv(struct work_struct *work)
 	struct sctp_endpoint *ep;
 	struct sctp_chunk *chunk;
 	struct sctp_inq *inqueue;
-	int state;
+	int first_time = 1;	/* is this the first time through the loop */
 	int error = 0;
+	int state;
 
 	/* The association should be held so we should be safe. */
 	ep = asoc->ep;
@@ -1036,6 +1037,30 @@ static void sctp_assoc_bh_rcv(struct work_struct *work)
 		state = asoc->state;
 		subtype = SCTP_ST_CHUNK(chunk->chunk_hdr->type);
 
+		/* If the first chunk in the packet is AUTH, do special
+		 * processing specified in Section 6.3 of SCTP-AUTH spec
+		 */
+		if (first_time && subtype.chunk == SCTP_CID_AUTH) {
+			struct sctp_chunkhdr *next_hdr;
+
+			next_hdr = sctp_inq_peek(inqueue);
+			if (!next_hdr)
+				goto normal;
+
+			/* If the next chunk is COOKIE-ECHO, skip the AUTH
+			 * chunk while saving a pointer to it so we can do
+			 * Authentication later (during cookie-echo
+			 * processing).
+			 */
+			if (next_hdr->type == SCTP_CID_COOKIE_ECHO) {
+				chunk->auth_chunk = skb_clone(chunk->skb,
+							      GFP_ATOMIC);
+				chunk->auth = 1;
+				continue;
+			}
+		}
+
+normal:
 		/* SCTP-AUTH, Section 6.3:
 		 *    The receiver has a list of chunk types which it expects
 		 *    to be received only after an AUTH-chunk.  This list has
@@ -1074,6 +1099,9 @@ static void sctp_assoc_bh_rcv(struct work_struct *work)
 		/* If there is an error on chunk, discard this packet. */
 		if (error && chunk)
 			chunk->pdiscard = 1;
+
+		if (first_time)
+			first_time = 0;
 	}
 	sctp_association_put(asoc);
 }
diff --git a/net/sctp/sm_statefuns.c b/net/sctp/sm_statefuns.c
index 28c070e..c9ae340 100644
--- a/net/sctp/sm_statefuns.c
+++ b/net/sctp/sm_statefuns.c
@@ -153,10 +153,7 @@ static enum sctp_disposition sctp_sf_violation_chunk(
 					struct sctp_cmd_seq *commands);
 
 static enum sctp_ierror sctp_sf_authenticate(
-					struct net *net,
-					const struct sctp_endpoint *ep,
 					const struct sctp_association *asoc,
-					const union sctp_subtype type,
 					struct sctp_chunk *chunk);
 
 static enum sctp_disposition __sctp_sf_do_9_1_abort(
@@ -626,6 +623,38 @@ enum sctp_disposition sctp_sf_do_5_1C_ack(struct net *net,
 	return SCTP_DISPOSITION_CONSUME;
 }
 
+static bool sctp_auth_chunk_verify(struct net *net, struct sctp_chunk *chunk,
+				   const struct sctp_association *asoc)
+{
+	struct sctp_chunk auth;
+
+	if (!chunk->auth_chunk)
+		return true;
+
+	/* SCTP-AUTH:  auth_chunk pointer is only set when the cookie-echo
+	 * is supposed to be authenticated and we have to do delayed
+	 * authentication.  We've just recreated the association using
+	 * the information in the cookie and now it's much easier to
+	 * do the authentication.
+	 */
+
+	/* Make sure that we and the peer are AUTH capable */
+	if (!net->sctp.auth_enable || !asoc->peer.auth_capable)
+		return false;
+
+	/* set-up our fake chunk so that we can process it */
+	auth.skb = chunk->auth_chunk;
+	auth.asoc = chunk->asoc;
+	auth.sctp_hdr = chunk->sctp_hdr;
+	auth.chunk_hdr = (struct sctp_chunkhdr *)
+				skb_push(chunk->auth_chunk,
+					 sizeof(struct sctp_chunkhdr));
+	skb_pull(chunk->auth_chunk, sizeof(struct sctp_chunkhdr));
+	auth.transport = chunk->transport;
+
+	return sctp_sf_authenticate(asoc, &auth) == SCTP_IERROR_NO_ERROR;
+}
+
 /*
  * Respond to a normal COOKIE ECHO chunk.
  * We are the side that is being asked for an association.
@@ -763,37 +792,9 @@ enum sctp_disposition sctp_sf_do_5_1D_ce(struct net *net,
 	if (error)
 		goto nomem_init;
 
-	/* SCTP-AUTH:  auth_chunk pointer is only set when the cookie-echo
-	 * is supposed to be authenticated and we have to do delayed
-	 * authentication.  We've just recreated the association using
-	 * the information in the cookie and now it's much easier to
-	 * do the authentication.
-	 */
-	if (chunk->auth_chunk) {
-		struct sctp_chunk auth;
-		enum sctp_ierror ret;
-
-		/* Make sure that we and the peer are AUTH capable */
-		if (!net->sctp.auth_enable || !new_asoc->peer.auth_capable) {
-			sctp_association_free(new_asoc);
-			return sctp_sf_pdiscard(net, ep, asoc, type, arg, commands);
-		}
-
-		/* set-up our fake chunk so that we can process it */
-		auth.skb = chunk->auth_chunk;
-		auth.asoc = chunk->asoc;
-		auth.sctp_hdr = chunk->sctp_hdr;
-		auth.chunk_hdr = (struct sctp_chunkhdr *)
-					skb_push(chunk->auth_chunk,
-						 sizeof(struct sctp_chunkhdr));
-		skb_pull(chunk->auth_chunk, sizeof(struct sctp_chunkhdr));
-		auth.transport = chunk->transport;
-
-		ret = sctp_sf_authenticate(net, ep, new_asoc, type, &auth);
-		if (ret != SCTP_IERROR_NO_ERROR) {
-			sctp_association_free(new_asoc);
-			return sctp_sf_pdiscard(net, ep, asoc, type, arg, commands);
-		}
+	if (!sctp_auth_chunk_verify(net, chunk, new_asoc)) {
+		sctp_association_free(new_asoc);
+		return sctp_sf_pdiscard(net, ep, asoc, type, arg, commands);
 	}
 
 	repl = sctp_make_cookie_ack(new_asoc, chunk);
@@ -1797,13 +1798,15 @@ static enum sctp_disposition sctp_sf_do_dupcook_a(
 	if (sctp_auth_asoc_init_active_key(new_asoc, GFP_ATOMIC))
 		goto nomem;
 
+	if (!sctp_auth_chunk_verify(net, chunk, new_asoc))
+		return SCTP_DISPOSITION_DISCARD;
+
 	/* Make sure no new addresses are being added during the
 	 * restart.  Though this is a pretty complicated attack
 	 * since you'd have to get inside the cookie.
 	 */
-	if (!sctp_sf_check_restart_addrs(new_asoc, asoc, chunk, commands)) {
+	if (!sctp_sf_check_restart_addrs(new_asoc, asoc, chunk, commands))
 		return SCTP_DISPOSITION_CONSUME;
-	}
 
 	/* If the endpoint is in the SHUTDOWN-ACK-SENT state and recognizes
 	 * the peer has restarted (Action A), it MUST NOT setup a new
@@ -1912,6 +1915,9 @@ static enum sctp_disposition sctp_sf_do_dupcook_b(
 	if (sctp_auth_asoc_init_active_key(new_asoc, GFP_ATOMIC))
 		goto nomem;
 
+	if (!sctp_auth_chunk_verify(net, chunk, new_asoc))
+		return SCTP_DISPOSITION_DISCARD;
+
 	/* Update the content of current association.  */
 	sctp_add_cmd_sf(commands, SCTP_CMD_UPDATE_ASSOC, SCTP_ASOC(new_asoc));
 	sctp_add_cmd_sf(commands, SCTP_CMD_NEW_STATE,
@@ -2009,6 +2015,9 @@ static enum sctp_disposition sctp_sf_do_dupcook_d(
 	 * a COOKIE ACK.
 	 */
 
+	if (!sctp_auth_chunk_verify(net, chunk, asoc))
+		return SCTP_DISPOSITION_DISCARD;
+
 	/* Don't accidentally move back into established state. */
 	if (asoc->state < SCTP_STATE_ESTABLISHED) {
 		sctp_add_cmd_sf(commands, SCTP_CMD_TIMER_STOP,
@@ -4171,10 +4180,7 @@ gen_shutdown:
  * The return value is the disposition of the chunk.
  */
 static enum sctp_ierror sctp_sf_authenticate(
-					struct net *net,
-					const struct sctp_endpoint *ep,
 					const struct sctp_association *asoc,
-					const union sctp_subtype type,
 					struct sctp_chunk *chunk)
 {
 	struct sctp_shared_key *sh_key = NULL;
@@ -4275,7 +4281,7 @@ enum sctp_disposition sctp_sf_eat_auth(struct net *net,
 						  commands);
 
 	auth_hdr = (struct sctp_authhdr *)chunk->skb->data;
-	error = sctp_sf_authenticate(net, ep, asoc, type, chunk);
+	error = sctp_sf_authenticate(asoc, chunk);
 	switch (error) {
 	case SCTP_IERROR_AUTH_BAD_HMAC:
 		/* Generate the ERROR chunk and discard the rest
-- 
cgit v1.1


From 98f0a39529e553bb45c0a7d775d2cc2afe9ed91d Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Sat, 5 May 2018 08:35:04 -0700
Subject: tls: fix use after free in tls_sk_proto_close

syzbot reported a use-after-free in tls_sk_proto_close

Add a boolean value to cleanup a bit this function.

BUG: KASAN: use-after-free in tls_sk_proto_close+0x8ab/0x9c0 net/tls/tls_main.c:297
Read of size 1 at addr ffff8801ae40a858 by task syz-executor363/4503

CPU: 0 PID: 4503 Comm: syz-executor363 Not tainted 4.17.0-rc3+ #34
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Call Trace:
 __dump_stack lib/dump_stack.c:77 [inline]
 dump_stack+0x1b9/0x294 lib/dump_stack.c:113
 print_address_description+0x6c/0x20b mm/kasan/report.c:256
 kasan_report_error mm/kasan/report.c:354 [inline]
 kasan_report.cold.7+0x242/0x2fe mm/kasan/report.c:412
 __asan_report_load1_noabort+0x14/0x20 mm/kasan/report.c:430
 tls_sk_proto_close+0x8ab/0x9c0 net/tls/tls_main.c:297
 inet_release+0x104/0x1f0 net/ipv4/af_inet.c:427
 inet6_release+0x50/0x70 net/ipv6/af_inet6.c:460
 sock_release+0x96/0x1b0 net/socket.c:594
 sock_close+0x16/0x20 net/socket.c:1149
 __fput+0x34d/0x890 fs/file_table.c:209
 ____fput+0x15/0x20 fs/file_table.c:243
 task_work_run+0x1e4/0x290 kernel/task_work.c:113
 exit_task_work include/linux/task_work.h:22 [inline]
 do_exit+0x1aee/0x2730 kernel/exit.c:865
 do_group_exit+0x16f/0x430 kernel/exit.c:968
 get_signal+0x886/0x1960 kernel/signal.c:2469
 do_signal+0x98/0x2040 arch/x86/kernel/signal.c:810
 exit_to_usermode_loop+0x28a/0x310 arch/x86/entry/common.c:162
 prepare_exit_to_usermode arch/x86/entry/common.c:196 [inline]
 syscall_return_slowpath arch/x86/entry/common.c:265 [inline]
 do_syscall_64+0x6ac/0x800 arch/x86/entry/common.c:290
 entry_SYSCALL_64_after_hwframe+0x49/0xbe
RIP: 0033:0x4457b9
RSP: 002b:00007fdf4d766da8 EFLAGS: 00000246 ORIG_RAX: 00000000000000ca
RAX: fffffffffffffe00 RBX: 00000000006dac3c RCX: 00000000004457b9
RDX: 0000000000000000 RSI: 0000000000000000 RDI: 00000000006dac3c
RBP: 0000000000000000 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 00000000006dac38
R13: 3692738801137283 R14: 6bf92c39443c4c1d R15: 0000000000000006

Allocated by task 4498:
 save_stack+0x43/0xd0 mm/kasan/kasan.c:448
 set_track mm/kasan/kasan.c:460 [inline]
 kasan_kmalloc+0xc4/0xe0 mm/kasan/kasan.c:553
 kmem_cache_alloc_trace+0x152/0x780 mm/slab.c:3620
 kmalloc include/linux/slab.h:512 [inline]
 kzalloc include/linux/slab.h:701 [inline]
 create_ctx net/tls/tls_main.c:521 [inline]
 tls_init+0x1f9/0xb00 net/tls/tls_main.c:633
 tcp_set_ulp+0x1bc/0x520 net/ipv4/tcp_ulp.c:153
 do_tcp_setsockopt.isra.39+0x44a/0x2600 net/ipv4/tcp.c:2588
 tcp_setsockopt+0xc1/0xe0 net/ipv4/tcp.c:2893
 sock_common_setsockopt+0x9a/0xe0 net/core/sock.c:3039
 __sys_setsockopt+0x1bd/0x390 net/socket.c:1903
 __do_sys_setsockopt net/socket.c:1914 [inline]
 __se_sys_setsockopt net/socket.c:1911 [inline]
 __x64_sys_setsockopt+0xbe/0x150 net/socket.c:1911
 do_syscall_64+0x1b1/0x800 arch/x86/entry/common.c:287
 entry_SYSCALL_64_after_hwframe+0x49/0xbe

Freed by task 4503:
 save_stack+0x43/0xd0 mm/kasan/kasan.c:448
 set_track mm/kasan/kasan.c:460 [inline]
 __kasan_slab_free+0x11a/0x170 mm/kasan/kasan.c:521
 kasan_slab_free+0xe/0x10 mm/kasan/kasan.c:528
 __cache_free mm/slab.c:3498 [inline]
 kfree+0xd9/0x260 mm/slab.c:3813
 tls_sw_free_resources+0x2a3/0x360 net/tls/tls_sw.c:1037
 tls_sk_proto_close+0x67c/0x9c0 net/tls/tls_main.c:288
 inet_release+0x104/0x1f0 net/ipv4/af_inet.c:427
 inet6_release+0x50/0x70 net/ipv6/af_inet6.c:460
 sock_release+0x96/0x1b0 net/socket.c:594
 sock_close+0x16/0x20 net/socket.c:1149
 __fput+0x34d/0x890 fs/file_table.c:209
 ____fput+0x15/0x20 fs/file_table.c:243
 task_work_run+0x1e4/0x290 kernel/task_work.c:113
 exit_task_work include/linux/task_work.h:22 [inline]
 do_exit+0x1aee/0x2730 kernel/exit.c:865
 do_group_exit+0x16f/0x430 kernel/exit.c:968
 get_signal+0x886/0x1960 kernel/signal.c:2469
 do_signal+0x98/0x2040 arch/x86/kernel/signal.c:810
 exit_to_usermode_loop+0x28a/0x310 arch/x86/entry/common.c:162
 prepare_exit_to_usermode arch/x86/entry/common.c:196 [inline]
 syscall_return_slowpath arch/x86/entry/common.c:265 [inline]
 do_syscall_64+0x6ac/0x800 arch/x86/entry/common.c:290
 entry_SYSCALL_64_after_hwframe+0x49/0xbe

The buggy address belongs to the object at ffff8801ae40a800
 which belongs to the cache kmalloc-256 of size 256
The buggy address is located 88 bytes inside of
 256-byte region [ffff8801ae40a800, ffff8801ae40a900)
The buggy address belongs to the page:
page:ffffea0006b90280 count:1 mapcount:0 mapping:ffff8801ae40a080 index:0x0
flags: 0x2fffc0000000100(slab)
raw: 02fffc0000000100 ffff8801ae40a080 0000000000000000 000000010000000c
raw: ffffea0006bea9e0 ffffea0006bc94a0 ffff8801da8007c0 0000000000000000
page dumped because: kasan: bad access detected

Fixes: dd0bed1665d6 ("tls: support for Inline tls record")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Atul Gupta <atul.gupta@chelsio.com>
Cc: Steve Wise <swise@opengridcomputing.com>
Cc: Ilya Lesokhin <ilyal@mellanox.com>
Cc: Aviad Yehezkel <aviadye@mellanox.com>
Cc: Dave Watson <davejwatson@fb.com>
Reported-by: syzbot <syzkaller@googlegroups.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/tls/tls_main.c | 11 ++++-------
 1 file changed, 4 insertions(+), 7 deletions(-)

diff --git a/net/tls/tls_main.c b/net/tls/tls_main.c
index cc03e00..74ed1e7 100644
--- a/net/tls/tls_main.c
+++ b/net/tls/tls_main.c
@@ -248,16 +248,13 @@ static void tls_sk_proto_close(struct sock *sk, long timeout)
 	struct tls_context *ctx = tls_get_ctx(sk);
 	long timeo = sock_sndtimeo(sk, 0);
 	void (*sk_proto_close)(struct sock *sk, long timeout);
+	bool free_ctx = false;
 
 	lock_sock(sk);
 	sk_proto_close = ctx->sk_proto_close;
 
-	if (ctx->conf == TLS_HW_RECORD)
-		goto skip_tx_cleanup;
-
-	if (ctx->conf == TLS_BASE) {
-		kfree(ctx);
-		ctx = NULL;
+	if (ctx->conf == TLS_BASE || ctx->conf == TLS_HW_RECORD) {
+		free_ctx = true;
 		goto skip_tx_cleanup;
 	}
 
@@ -294,7 +291,7 @@ skip_tx_cleanup:
 	/* free ctx for TLS_HW_RECORD, used by tcp_set_state
 	 * for sk->sk_prot->unhash [tls_hw_unhash]
 	 */
-	if (ctx && ctx->conf == TLS_HW_RECORD)
+	if (free_ctx)
 		kfree(ctx);
 }
 
-- 
cgit v1.1


From 9255bacd57a5a247be91ce2b81ff09e15a0b8822 Mon Sep 17 00:00:00 2001
From: Andrew Lunn <andrew@lunn.ch>
Date: Sat, 5 May 2018 20:58:22 +0200
Subject: net: dsa: mv88e6xxx: Fix PHY interrupts by parameterising PHY base
 address

Most of the mv88e6xxx switches have the PHYs at address 0, 1, 2, ...
The 6341 however has the PHYs at 0x10, 0x11, 0x12. Add a parameter to
the info structure for this base address.

Testing of 6f88284f3bd7 ("net: dsa: mv88e6xxx: Add MDIO interrupts for
internal PHYs") was performed on the 6341. So it works only on the
6341. Use this base information to correctly set the interrupt.

Fixes: 6f88284f3bd7 ("net: dsa: mv88e6xxx: Add MDIO interrupts for internal PHYs")
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/dsa/mv88e6xxx/chip.c    | 26 ++++++++++++++++++++++++++
 drivers/net/dsa/mv88e6xxx/chip.h    |  1 +
 drivers/net/dsa/mv88e6xxx/global2.c |  2 +-
 3 files changed, 28 insertions(+), 1 deletion(-)

diff --git a/drivers/net/dsa/mv88e6xxx/chip.c b/drivers/net/dsa/mv88e6xxx/chip.c
index 3d20910..5b4374f 100644
--- a/drivers/net/dsa/mv88e6xxx/chip.c
+++ b/drivers/net/dsa/mv88e6xxx/chip.c
@@ -3370,6 +3370,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_internal_phys = 5,
 		.max_vid = 4095,
 		.port_base_addr = 0x10,
+		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 15000,
@@ -3391,6 +3392,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_internal_phys = 0,
 		.max_vid = 4095,
 		.port_base_addr = 0x10,
+		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 15000,
@@ -3410,6 +3412,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_internal_phys = 8,
 		.max_vid = 4095,
 		.port_base_addr = 0x10,
+		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 15000,
@@ -3431,6 +3434,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_internal_phys = 5,
 		.max_vid = 4095,
 		.port_base_addr = 0x10,
+		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 15000,
@@ -3452,6 +3456,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_internal_phys = 0,
 		.max_vid = 4095,
 		.port_base_addr = 0x10,
+		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 15000,
@@ -3472,6 +3477,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_gpio = 11,
 		.max_vid = 4095,
 		.port_base_addr = 0x10,
+		.phy_base_addr = 0x10,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 3750,
@@ -3493,6 +3499,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_internal_phys = 5,
 		.max_vid = 4095,
 		.port_base_addr = 0x10,
+		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 15000,
@@ -3514,6 +3521,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_internal_phys = 0,
 		.max_vid = 4095,
 		.port_base_addr = 0x10,
+		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 15000,
@@ -3535,6 +3543,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_internal_phys = 5,
 		.max_vid = 4095,
 		.port_base_addr = 0x10,
+		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 15000,
@@ -3557,6 +3566,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_gpio = 15,
 		.max_vid = 4095,
 		.port_base_addr = 0x10,
+		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 15000,
@@ -3578,6 +3588,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_internal_phys = 5,
 		.max_vid = 4095,
 		.port_base_addr = 0x10,
+		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 15000,
@@ -3600,6 +3611,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_gpio = 15,
 		.max_vid = 4095,
 		.port_base_addr = 0x10,
+		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 15000,
@@ -3621,6 +3633,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_internal_phys = 0,
 		.max_vid = 4095,
 		.port_base_addr = 0x10,
+		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 15000,
@@ -3641,6 +3654,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_gpio = 16,
 		.max_vid = 8191,
 		.port_base_addr = 0x0,
+		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.tag_protocol = DSA_TAG_PROTO_DSA,
@@ -3663,6 +3677,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_gpio = 16,
 		.max_vid = 8191,
 		.port_base_addr = 0x0,
+		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 3750,
@@ -3684,6 +3699,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_internal_phys = 11,
 		.max_vid = 8191,
 		.port_base_addr = 0x0,
+		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 3750,
@@ -3707,6 +3723,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_gpio = 15,
 		.max_vid = 4095,
 		.port_base_addr = 0x10,
+		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 15000,
@@ -3730,6 +3747,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_gpio = 16,
 		.max_vid = 8191,
 		.port_base_addr = 0x0,
+		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 3750,
@@ -3753,6 +3771,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_gpio = 15,
 		.max_vid = 4095,
 		.port_base_addr = 0x10,
+		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 15000,
@@ -3776,6 +3795,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_gpio = 15,
 		.max_vid = 4095,
 		.port_base_addr = 0x10,
+		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 15000,
@@ -3798,6 +3818,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_gpio = 11,
 		.max_vid = 4095,
 		.port_base_addr = 0x10,
+		.phy_base_addr = 0x10,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 3750,
@@ -3820,6 +3841,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_internal_phys = 5,
 		.max_vid = 4095,
 		.port_base_addr = 0x10,
+		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 15000,
@@ -3841,6 +3863,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_internal_phys = 5,
 		.max_vid = 4095,
 		.port_base_addr = 0x10,
+		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 15000,
@@ -3863,6 +3886,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_gpio = 15,
 		.max_vid = 4095,
 		.port_base_addr = 0x10,
+		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 15000,
@@ -3885,6 +3909,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_gpio = 16,
 		.max_vid = 8191,
 		.port_base_addr = 0x0,
+		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 3750,
@@ -3907,6 +3932,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_gpio = 16,
 		.max_vid = 8191,
 		.port_base_addr = 0x0,
+		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 3750,
diff --git a/drivers/net/dsa/mv88e6xxx/chip.h b/drivers/net/dsa/mv88e6xxx/chip.h
index 80490f6..12b7f46 100644
--- a/drivers/net/dsa/mv88e6xxx/chip.h
+++ b/drivers/net/dsa/mv88e6xxx/chip.h
@@ -114,6 +114,7 @@ struct mv88e6xxx_info {
 	unsigned int num_gpio;
 	unsigned int max_vid;
 	unsigned int port_base_addr;
+	unsigned int phy_base_addr;
 	unsigned int global1_addr;
 	unsigned int global2_addr;
 	unsigned int age_time_coeff;
diff --git a/drivers/net/dsa/mv88e6xxx/global2.c b/drivers/net/dsa/mv88e6xxx/global2.c
index 0ce627f..8d22d66 100644
--- a/drivers/net/dsa/mv88e6xxx/global2.c
+++ b/drivers/net/dsa/mv88e6xxx/global2.c
@@ -1118,7 +1118,7 @@ int mv88e6xxx_g2_irq_mdio_setup(struct mv88e6xxx_chip *chip,
 			err = irq;
 			goto out;
 		}
-		bus->irq[chip->info->port_base_addr + phy] = irq;
+		bus->irq[chip->info->phy_base_addr + phy] = irq;
 	}
 	return 0;
 out:
-- 
cgit v1.1


From 53bc017f72b3b7f7f1bad3043c2d6dd87331ea44 Mon Sep 17 00:00:00 2001
From: Wolfram Sang <wsa+renesas@sang-engineering.com>
Date: Sun, 6 May 2018 13:23:52 +0200
Subject: net: flow_dissector: fix typo 'can by' to 'can be'

Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/flow_dissector.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/include/net/flow_dissector.h b/include/net/flow_dissector.h
index 9a07477..d1fcf24 100644
--- a/include/net/flow_dissector.h
+++ b/include/net/flow_dissector.h
@@ -251,7 +251,7 @@ extern struct flow_dissector flow_keys_buf_dissector;
  * This structure is used to hold a digest of the full flow keys. This is a
  * larger "hash" of a flow to allow definitively matching specific flows where
  * the 32 bit skb->hash is not large enough. The size is limited to 16 bytes so
- * that it can by used in CB of skb (see sch_choke for an example).
+ * that it can be used in CB of skb (see sch_choke for an example).
  */
 #define FLOW_KEYS_DIGEST_LEN	16
 struct flow_keys_digest {
-- 
cgit v1.1


From 080324c36ade319f57e505633ab54f6f53289b45 Mon Sep 17 00:00:00 2001
From: Andre Tomt <andre@tomt.net>
Date: Mon, 7 May 2018 04:24:39 +0200
Subject: net/tls: Fix connection stall on partial tls record

In the case of writing a partial tls record we forgot to clear the
ctx->in_tcp_sendpages flag, causing some connections to stall.

Fixes: c212d2c7fc47 ("net/tls: Don't recursively call push_record during tls_write_space callbacks")
Signed-off-by: Andre Tomt <andre@tomt.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/tls/tls_main.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/net/tls/tls_main.c b/net/tls/tls_main.c
index 74ed1e7..20cd93be62 100644
--- a/net/tls/tls_main.c
+++ b/net/tls/tls_main.c
@@ -135,6 +135,7 @@ retry:
 			offset -= sg->offset;
 			ctx->partially_sent_offset = offset;
 			ctx->partially_sent_record = (void *)sg;
+			ctx->in_tcp_sendpages = false;
 			return ret;
 		}
 
-- 
cgit v1.1


From 5ed3fde21331407a319e966a0582195988b4f4f0 Mon Sep 17 00:00:00 2001
From: Steffen Klassert <steffen.klassert@secunet.com>
Date: Mon, 7 May 2018 12:39:30 +0200
Subject: MAINTAINERS: Update the 3c59x network driver entry

Replace my old E-Mail address with a working one.
While at it, change the maintainance status to
'Odd Fixes'. I'm still around with some knowledge,
but don't actively maintain it anymore.

Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 MAINTAINERS | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/MAINTAINERS b/MAINTAINERS
index b1ccabd..b3cbf1c 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -137,9 +137,9 @@ Maintainers List (try to look for most precise areas first)
 		-----------------------------------
 
 3C59X NETWORK DRIVER
-M:	Steffen Klassert <klassert@mathematik.tu-chemnitz.de>
+M:	Steffen Klassert <klassert@kernel.org>
 L:	netdev@vger.kernel.org
-S:	Maintained
+S:	Odd Fixes
 F:	Documentation/networking/vortex.txt
 F:	drivers/net/ethernet/3com/3c59x.c
 
-- 
cgit v1.1


From a9f71d0de68db84de610b09cb72e8eb0f621d0e1 Mon Sep 17 00:00:00 2001
From: Georg Hofmann <georg@hofmannsweb.com>
Date: Mon, 7 May 2018 14:03:18 +0200
Subject: trivial: fix inconsistent help texts

This patch removes "experimental" from the help text where depends on
CONFIG_EXPERIMENTAL was already removed.

Signed-off-by: Georg Hofmann <georg@hofmannsweb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/ipv6/Kconfig | 9 ++++-----
 1 file changed, 4 insertions(+), 5 deletions(-)

diff --git a/net/ipv6/Kconfig b/net/ipv6/Kconfig
index 6794ddf..11e4e80 100644
--- a/net/ipv6/Kconfig
+++ b/net/ipv6/Kconfig
@@ -34,16 +34,15 @@ config IPV6_ROUTE_INFO
 	bool "IPv6: Route Information (RFC 4191) support"
 	depends on IPV6_ROUTER_PREF
 	---help---
-	  This is experimental support of Route Information.
+	  Support of Route Information.
 
 	  If unsure, say N.
 
 config IPV6_OPTIMISTIC_DAD
 	bool "IPv6: Enable RFC 4429 Optimistic DAD"
 	---help---
-	  This is experimental support for optimistic Duplicate
-	  Address Detection.  It allows for autoconfigured addresses
-	  to be used more quickly.
+	  Support for optimistic Duplicate Address Detection. It allows for
+	  autoconfigured addresses to be used more quickly.
 
 	  If unsure, say N.
 
@@ -280,7 +279,7 @@ config IPV6_MROUTE
 	depends on IPV6
 	select IP_MROUTE_COMMON
 	---help---
-	  Experimental support for IPv6 multicast forwarding.
+	  Support for IPv6 multicast forwarding.
 	  If unsure, say N.
 
 config IPV6_MROUTE_MULTIPLE_TABLES
-- 
cgit v1.1


From 8c61ab7f111a2b29d051348b9cb9a39804ebf1f8 Mon Sep 17 00:00:00 2001
From: Igor Russkikh <igor.russkikh@aquantia.com>
Date: Mon, 7 May 2018 16:10:38 +0300
Subject: net: aquantia: driver should correctly declare vlan_features bits

In particular, not reporting SG forced skbs to be linear for vlan
interfaces over atlantic NIC.

With this fix it is possible to enable SG feature on device and
therefore optimize performance.

Reported-by: Ma Yuying <yuma@redhat.com>
Signed-off-by: Igor Russkikh <igor.russkikh@aquantia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/aquantia/atlantic/aq_nic.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/net/ethernet/aquantia/atlantic/aq_nic.c b/drivers/net/ethernet/aquantia/atlantic/aq_nic.c
index 32f6d2e..720760d 100644
--- a/drivers/net/ethernet/aquantia/atlantic/aq_nic.c
+++ b/drivers/net/ethernet/aquantia/atlantic/aq_nic.c
@@ -246,6 +246,8 @@ void aq_nic_ndev_init(struct aq_nic_s *self)
 
 	self->ndev->hw_features |= aq_hw_caps->hw_features;
 	self->ndev->features = aq_hw_caps->hw_features;
+	self->ndev->vlan_features |= NETIF_F_HW_CSUM | NETIF_F_RXCSUM |
+				     NETIF_F_RXHASH | NETIF_F_SG | NETIF_F_LRO;
 	self->ndev->priv_flags = aq_hw_caps->hw_priv_flags;
 	self->ndev->priv_flags |= IFF_LIVE_ADDR_CHANGE;
 
-- 
cgit v1.1


From a09bd81b5413d1b4d705c6c5303b5d311069da22 Mon Sep 17 00:00:00 2001
From: Igor Russkikh <igor.russkikh@aquantia.com>
Date: Mon, 7 May 2018 16:10:39 +0300
Subject: net: aquantia: Limit number of vectors to actually allocated irqs

Driver should use pci_alloc_irq_vectors return value to correct number
of allocated vectors and napi instances. Otherwise it'll panic later
in pci_irq_vector.

Driver also should allow more than one MSI vectors to be allocated.

Error return path from pci_alloc_irq_vectors is also fixed to revert
resources in a correct sequence when error happens.

Reported-by: Long, Nicholas <nicholas.a.long@baesystems.com>
Fixes: 23ee07a ("net: aquantia: Cleanup pci functions module")
Signed-off-by: Igor Russkikh <igor.russkikh@aquantia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/aquantia/atlantic/aq_nic.c      |  1 +
 drivers/net/ethernet/aquantia/atlantic/aq_nic.h      |  1 +
 drivers/net/ethernet/aquantia/atlantic/aq_pci_func.c | 20 ++++++++++----------
 3 files changed, 12 insertions(+), 10 deletions(-)

diff --git a/drivers/net/ethernet/aquantia/atlantic/aq_nic.c b/drivers/net/ethernet/aquantia/atlantic/aq_nic.c
index 720760d..1a1a638 100644
--- a/drivers/net/ethernet/aquantia/atlantic/aq_nic.c
+++ b/drivers/net/ethernet/aquantia/atlantic/aq_nic.c
@@ -95,6 +95,7 @@ void aq_nic_cfg_start(struct aq_nic_s *self)
 	/*rss rings */
 	cfg->vecs = min(cfg->aq_hw_caps->vecs, AQ_CFG_VECS_DEF);
 	cfg->vecs = min(cfg->vecs, num_online_cpus());
+	cfg->vecs = min(cfg->vecs, self->irqvecs);
 	/* cfg->vecs should be power of 2 for RSS */
 	if (cfg->vecs >= 8U)
 		cfg->vecs = 8U;
diff --git a/drivers/net/ethernet/aquantia/atlantic/aq_nic.h b/drivers/net/ethernet/aquantia/atlantic/aq_nic.h
index 219b550..faa533a 100644
--- a/drivers/net/ethernet/aquantia/atlantic/aq_nic.h
+++ b/drivers/net/ethernet/aquantia/atlantic/aq_nic.h
@@ -80,6 +80,7 @@ struct aq_nic_s {
 
 	struct pci_dev *pdev;
 	unsigned int msix_entry_mask;
+	u32 irqvecs;
 };
 
 static inline struct device *aq_nic_get_dev(struct aq_nic_s *self)
diff --git a/drivers/net/ethernet/aquantia/atlantic/aq_pci_func.c b/drivers/net/ethernet/aquantia/atlantic/aq_pci_func.c
index ecc6306..a50e08b 100644
--- a/drivers/net/ethernet/aquantia/atlantic/aq_pci_func.c
+++ b/drivers/net/ethernet/aquantia/atlantic/aq_pci_func.c
@@ -267,16 +267,16 @@ static int aq_pci_probe(struct pci_dev *pdev,
 	numvecs = min(numvecs, num_online_cpus());
 	/*enable interrupts */
 #if !AQ_CFG_FORCE_LEGACY_INT
-	err = pci_alloc_irq_vectors(self->pdev, numvecs, numvecs,
-				    PCI_IRQ_MSIX);
-
-	if (err < 0) {
-		err = pci_alloc_irq_vectors(self->pdev, 1, 1,
-					    PCI_IRQ_MSI | PCI_IRQ_LEGACY);
-		if (err < 0)
-			goto err_hwinit;
+	numvecs = pci_alloc_irq_vectors(self->pdev, 1, numvecs,
+					PCI_IRQ_MSIX | PCI_IRQ_MSI |
+					PCI_IRQ_LEGACY);
+
+	if (numvecs < 0) {
+		err = numvecs;
+		goto err_hwinit;
 	}
 #endif
+	self->irqvecs = numvecs;
 
 	/* net device init */
 	aq_nic_cfg_start(self);
@@ -298,9 +298,9 @@ err_free_aq_hw:
 	kfree(self->aq_hw);
 err_ioremap:
 	free_netdev(ndev);
-err_pci_func:
-	pci_release_regions(pdev);
 err_ndev:
+	pci_release_regions(pdev);
+err_pci_func:
 	pci_disable_device(pdev);
 	return err;
 }
-- 
cgit v1.1


From 2c5d5b13c6eb79f5677e206b8aad59b3a2097f60 Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Mon, 7 May 2018 09:02:25 -0700
Subject: llc: better deal with too small mtu

syzbot loves to set very small mtu on devices, since it brings joy.
We must make llc_ui_sendmsg() fool proof.

usercopy: Kernel memory overwrite attempt detected to wrapped address (offset 0, size 18446612139802320068)!

kernel BUG at mm/usercopy.c:100!
invalid opcode: 0000 [#1] SMP KASAN
Dumping ftrace buffer:
   (ftrace buffer empty)
Modules linked in:
CPU: 0 PID: 17464 Comm: syz-executor1 Not tainted 4.17.0-rc3+ #36
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
RIP: 0010:usercopy_abort+0xbb/0xbd mm/usercopy.c:88
RSP: 0018:ffff8801868bf800 EFLAGS: 00010282
RAX: 000000000000006c RBX: ffffffff87d2fb00 RCX: 0000000000000000
RDX: 000000000000006c RSI: ffffffff81610731 RDI: ffffed0030d17ef6
RBP: ffff8801868bf858 R08: ffff88018daa4200 R09: ffffed003b5c4fb0
R10: ffffed003b5c4fb0 R11: ffff8801dae27d87 R12: ffffffff87d2f8e0
R13: ffffffff87d2f7a0 R14: ffffffff87d2f7a0 R15: ffffffff87d2f7a0
FS:  00007f56a14ac700(0000) GS:ffff8801dae00000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000001b2bc21000 CR3: 00000001abeb1000 CR4: 00000000001426f0
DR0: 0000000020000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000030602
Call Trace:
 check_bogus_address mm/usercopy.c:153 [inline]
 __check_object_size+0x5d9/0x5d9 mm/usercopy.c:256
 check_object_size include/linux/thread_info.h:108 [inline]
 check_copy_size include/linux/thread_info.h:139 [inline]
 copy_from_iter_full include/linux/uio.h:121 [inline]
 memcpy_from_msg include/linux/skbuff.h:3305 [inline]
 llc_ui_sendmsg+0x4b1/0x1530 net/llc/af_llc.c:941
 sock_sendmsg_nosec net/socket.c:629 [inline]
 sock_sendmsg+0xd5/0x120 net/socket.c:639
 __sys_sendto+0x3d7/0x670 net/socket.c:1789
 __do_sys_sendto net/socket.c:1801 [inline]
 __se_sys_sendto net/socket.c:1797 [inline]
 __x64_sys_sendto+0xe1/0x1a0 net/socket.c:1797
 do_syscall_64+0x1b1/0x800 arch/x86/entry/common.c:287
 entry_SYSCALL_64_after_hwframe+0x49/0xbe
RIP: 0033:0x455979
RSP: 002b:00007f56a14abc68 EFLAGS: 00000246 ORIG_RAX: 000000000000002c
RAX: ffffffffffffffda RBX: 00007f56a14ac6d4 RCX: 0000000000455979
RDX: 0000000000000000 RSI: 0000000020000000 RDI: 0000000000000018
RBP: 000000000072bea0 R08: 00000000200012c0 R09: 0000000000000010
R10: 0000000000000000 R11: 0000000000000246 R12: 00000000ffffffff
R13: 0000000000000548 R14: 00000000006fbf60 R15: 0000000000000000
Code: 55 c0 e8 c0 55 bb ff ff 75 c8 48 8b 55 c0 4d 89 f9 ff 75 d0 4d 89 e8 48 89 d9 4c 89 e6 41 56 48 c7 c7 80 fa d2 87 e8 a0 0b a3 ff <0f> 0b e8 95 55 bb ff e8 c0 a8 f7 ff 8b 95 14 ff ff ff 4d 89 e8
RIP: usercopy_abort+0xbb/0xbd mm/usercopy.c:88 RSP: ffff8801868bf800

Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: syzbot <syzkaller@googlegroups.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/llc/af_llc.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/net/llc/af_llc.c b/net/llc/af_llc.c
index cb80ebb..1beeea9 100644
--- a/net/llc/af_llc.c
+++ b/net/llc/af_llc.c
@@ -930,6 +930,9 @@ static int llc_ui_sendmsg(struct socket *sock, struct msghdr *msg, size_t len)
 	if (size > llc->dev->mtu)
 		size = llc->dev->mtu;
 	copied = size - hdrlen;
+	rc = -EINVAL;
+	if (copied < 0)
+		goto release;
 	release_sock(sk);
 	skb = sock_alloc_send_skb(sk, size, noblock, &rc);
 	lock_sock(sk);
-- 
cgit v1.1


From 6c0a8f6b5a45ac892a763b6299bd3c5324fc5e02 Mon Sep 17 00:00:00 2001
From: Michael Ellerman <mpe@ellerman.id.au>
Date: Tue, 8 May 2018 14:59:56 +1000
Subject: powerpc/pseries: Fix CONFIG_NUMA=n build

The build is failing with CONFIG_NUMA=n and some compiler versions:

  arch/powerpc/platforms/pseries/hotplug-cpu.o: In function `dlpar_online_cpu':
  hotplug-cpu.c:(.text+0x12c): undefined reference to `timed_topology_update'
  arch/powerpc/platforms/pseries/hotplug-cpu.o: In function `dlpar_cpu_remove':
  hotplug-cpu.c:(.text+0x400): undefined reference to `timed_topology_update'

Fix it by moving the empty version of timed_topology_update() into the
existing #ifdef block, which has the right guard of SPLPAR && NUMA.

Fixes: cee5405da402 ("powerpc/hotplug: Improve responsiveness of hotplug change")
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
---
 arch/powerpc/include/asm/topology.h | 13 +++++--------
 1 file changed, 5 insertions(+), 8 deletions(-)

diff --git a/arch/powerpc/include/asm/topology.h b/arch/powerpc/include/asm/topology.h
index 9f42164..16b0778 100644
--- a/arch/powerpc/include/asm/topology.h
+++ b/arch/powerpc/include/asm/topology.h
@@ -91,6 +91,7 @@ extern int start_topology_update(void);
 extern int stop_topology_update(void);
 extern int prrn_is_enabled(void);
 extern int find_and_online_cpu_nid(int cpu);
+extern int timed_topology_update(int nsecs);
 #else
 static inline int start_topology_update(void)
 {
@@ -108,16 +109,12 @@ static inline int find_and_online_cpu_nid(int cpu)
 {
 	return 0;
 }
+static inline int timed_topology_update(int nsecs)
+{
+	return 0;
+}
 #endif /* CONFIG_NUMA && CONFIG_PPC_SPLPAR */
 
-#if defined(CONFIG_HOTPLUG_CPU) || defined(CONFIG_NEED_MULTIPLE_NODES)
-#if defined(CONFIG_PPC_SPLPAR)
-extern int timed_topology_update(int nsecs);
-#else
-#define	timed_topology_update(nsecs)
-#endif /* CONFIG_PPC_SPLPAR */
-#endif /* CONFIG_HOTPLUG_CPU || CONFIG_NEED_MULTIPLE_NODES */
-
 #include <asm-generic/topology.h>
 
 #ifdef CONFIG_SMP
-- 
cgit v1.1


From 4e8507ba774f1fe5cd0c26a7cafc09afb8a6ba8e Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Sun, 6 May 2018 19:31:47 +0100
Subject: drm/i915: Don't request a bug report for unsafe module parameters

Unsafe module parameters are just that, unsafe. If the user is foolish
enough to try them and the kernel breaks, they get to keep both pieces.
Don't ask them to file a bug report if they broke it themselves.

References: https://bugs.freedesktop.org/show_bug.cgi?id=106423
Fixes: d15d7538c6d2 ("drm/i915: Tune down init error message due to failure injection")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Imre Deak <imre.deak@intel.com>
Cc: Jani Nikula <jani.nikula@linux.intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Acked-by: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180506183147.2690-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/i915_drv.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index 8c29868..2a96d08 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -101,7 +101,13 @@ __i915_printk(struct drm_i915_private *dev_priv, const char *level,
 		   __builtin_return_address(0), &vaf);
 
 	if (is_error && !shown_bug_once) {
-		dev_notice(kdev, "%s", FDO_BUG_MSG);
+		/*
+		 * Ask the user to file a bug report for the error, except
+		 * if they may have caused the bug by fiddling with unsafe
+		 * module parameters.
+		 */
+		if (!test_taint(TAINT_USER))
+			dev_notice(kdev, "%s", FDO_BUG_MSG);
 		shown_bug_once = true;
 	}
 
-- 
cgit v1.1


From 87c7acf867700a9f32db81b227174bf83fecfde3 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Tue, 8 May 2018 01:30:45 +0100
Subject: drm/i915/execlists: Drop unused parameter to lookup_priolist()

lookup_priolist() no longer attaches the request into the priolist, it
just returns the priolist for the given priority instead. Drop the
unused parameter.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180508003046.2633-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/intel_lrc.c | 8 +++-----
 1 file changed, 3 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 9f3cce0..bf9a44d 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -258,9 +258,7 @@ intel_lr_context_descriptor_update(struct i915_gem_context *ctx,
 }
 
 static struct i915_priolist *
-lookup_priolist(struct intel_engine_cs *engine,
-		struct i915_sched_node *node,
-		int prio)
+lookup_priolist(struct intel_engine_cs *engine, int prio)
 {
 	struct intel_engine_execlists * const execlists = &engine->execlists;
 	struct i915_priolist *p;
@@ -345,7 +343,7 @@ static void __unwind_incomplete_requests(struct intel_engine_cs *engine)
 		GEM_BUG_ON(rq_prio(rq) == I915_PRIORITY_INVALID);
 		if (rq_prio(rq) != last_prio) {
 			last_prio = rq_prio(rq);
-			p = lookup_priolist(engine, &rq->sched, last_prio);
+			p = lookup_priolist(engine, last_prio);
 		}
 
 		list_add(&rq->sched.link, &p->requests);
@@ -1145,7 +1143,7 @@ static void queue_request(struct intel_engine_cs *engine,
 			  int prio)
 {
 	list_add_tail(&node->link,
-		      &lookup_priolist(engine, node, prio)->requests);
+		      &lookup_priolist(engine, prio)->requests);
 }
 
 static void __submit_queue(struct intel_engine_cs *engine, int prio)
-- 
cgit v1.1


From a02eb975be78171f66a47c103e57e7940d0860a7 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Tue, 8 May 2018 01:30:46 +0100
Subject: drm/i915/execlists: Cache the priolist when rescheduling

When rescheduling a change of dependencies, they all need to be added to
the same priolist (at least the ones on the same engine!). Since we
likely want to move a batch of requests, keep the priolist around.

v2: Throw in an assert to catch trivial errors quickly.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180508003046.2633-2-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/intel_lrc.c | 13 ++++++++++---
 1 file changed, 10 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index bf9a44d..046adf3 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -346,6 +346,7 @@ static void __unwind_incomplete_requests(struct intel_engine_cs *engine)
 			p = lookup_priolist(engine, last_prio);
 		}
 
+		GEM_BUG_ON(p->priority != rq_prio(rq));
 		list_add(&rq->sched.link, &p->requests);
 	}
 }
@@ -1198,7 +1199,8 @@ sched_lock_engine(struct i915_sched_node *node, struct intel_engine_cs *locked)
 static void execlists_schedule(struct i915_request *request,
 			       const struct i915_sched_attr *attr)
 {
-	struct intel_engine_cs *engine;
+	struct i915_priolist *uninitialized_var(pl);
+	struct intel_engine_cs *engine, *last;
 	struct i915_dependency *dep, *p;
 	struct i915_dependency stack;
 	const int prio = attr->priority;
@@ -1271,6 +1273,7 @@ static void execlists_schedule(struct i915_request *request,
 		__list_del_entry(&stack.dfs_link);
 	}
 
+	last = NULL;
 	engine = request->engine;
 	spin_lock_irq(&engine->timeline.lock);
 
@@ -1287,8 +1290,12 @@ static void execlists_schedule(struct i915_request *request,
 
 		node->attr.priority = prio;
 		if (!list_empty(&node->link)) {
-			__list_del_entry(&node->link);
-			queue_request(engine, node, prio);
+			if (last != engine) {
+				pl = lookup_priolist(engine, prio);
+				last = engine;
+			}
+			GEM_BUG_ON(pl->priority != prio);
+			list_move_tail(&node->link, &pl->requests);
 		}
 
 		if (prio > engine->execlists.queue_priority &&
-- 
cgit v1.1


From 71c23a821c6bcacba71a094efe49ee689605906b Mon Sep 17 00:00:00 2001
From: Jakob Unterwurzacher <jakob.unterwurzacher@theobroma-systems.com>
Date: Wed, 18 Apr 2018 16:10:03 +0200
Subject: can: dev: increase bus-off message severity

bus-off is usually caused by hardware malfunction or configuration error
(baud rate mismatch) and causes a complete loss of communication.

Increase the "bus-off" message's severity from netdev_dbg() to
netdev_info() to make it visible to the user.

A can interface going into bus-off is similar in severity to ethernet's
"Link is Down" message, which is also printed at info level.

It is debatable whether the the "restarted" message should also be
changed to netdev_info() to make the interface state changes
comprehensible from the kernel log. I have chosen to keep the
"restarted" message at dbg for now as the "bus-off" message should be
enough for the user to notice and investigate the problem.

Signed-off-by: Jakob Unterwurzacher <jakob.unterwurzacher@theobroma-systems.com>
Cc: linux-can@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
---
 drivers/net/can/dev.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/can/dev.c b/drivers/net/can/dev.c
index b177956..3c71f1c 100644
--- a/drivers/net/can/dev.c
+++ b/drivers/net/can/dev.c
@@ -605,7 +605,7 @@ void can_bus_off(struct net_device *dev)
 {
 	struct can_priv *priv = netdev_priv(dev);
 
-	netdev_dbg(dev, "bus-off\n");
+	netdev_info(dev, "bus-off\n");
 
 	netif_carrier_off(dev);
 
-- 
cgit v1.1


From 0e030a373df3b8792b8991740fc31fe0629c6e58 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig?= <u.kleine-koenig@pengutronix.de>
Date: Wed, 25 Apr 2018 16:50:39 +0200
Subject: can: flexcan: fix endianess detection
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

In commit 88462d2a7830 ("can: flexcan: Remodel FlexCAN register r/w APIs
for big endian FlexCAN controllers.") the following logic was
implemented:

	if the dt property "big-endian" is given or
	   the device is compatible to "fsl,p1010-flexcan":
		use big-endian mode;
	else
		use little-endian mode;

This relies on commit d50f4630c2e1 ("arm: dts: Remove p1010-flexcan
compatible from imx series dts") which was applied a few commits later.
Without this commit (or an old device tree used for booting a new
kernel) the flexcan devices on i.MX25, i.MX28, i.MX35 and i.MX53 match
the 'the device is compatible to "fsl,p1010-flexcan"' test and so are
switched erroneously to big endian mode.

Instead of the check above put a quirk in devtype data and rely on
of_match_device yielding the most compatible match

Fixes: 88462d2a7830 ("can: flexcan: Remodel FlexCAN register r/w APIs for big endian FlexCAN controllers.")
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Tested-by: Gavin Schenk <g.schenk@eckelmann.de>
Cc: linux-stable <stable@vger.kernel.org> # >= v4.16
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
---
 drivers/net/can/flexcan.c | 26 ++++++++++++++------------
 1 file changed, 14 insertions(+), 12 deletions(-)

diff --git a/drivers/net/can/flexcan.c b/drivers/net/can/flexcan.c
index 634c51e..d53a45b 100644
--- a/drivers/net/can/flexcan.c
+++ b/drivers/net/can/flexcan.c
@@ -200,6 +200,7 @@
 #define FLEXCAN_QUIRK_DISABLE_MECR	BIT(4) /* Disable Memory error detection */
 #define FLEXCAN_QUIRK_USE_OFF_TIMESTAMP	BIT(5) /* Use timestamp based offloading */
 #define FLEXCAN_QUIRK_BROKEN_PERR_STATE	BIT(6) /* No interrupt for error passive */
+#define FLEXCAN_QUIRK_DEFAULT_BIG_ENDIAN	BIT(7) /* default to BE register access */
 
 /* Structure of the message buffer */
 struct flexcan_mb {
@@ -288,6 +289,12 @@ struct flexcan_priv {
 
 static const struct flexcan_devtype_data fsl_p1010_devtype_data = {
 	.quirks = FLEXCAN_QUIRK_BROKEN_WERR_STATE |
+		FLEXCAN_QUIRK_BROKEN_PERR_STATE |
+		FLEXCAN_QUIRK_DEFAULT_BIG_ENDIAN,
+};
+
+static const struct flexcan_devtype_data fsl_imx25_devtype_data = {
+	.quirks = FLEXCAN_QUIRK_BROKEN_WERR_STATE |
 		FLEXCAN_QUIRK_BROKEN_PERR_STATE,
 };
 
@@ -1251,9 +1258,9 @@ static void unregister_flexcandev(struct net_device *dev)
 static const struct of_device_id flexcan_of_match[] = {
 	{ .compatible = "fsl,imx6q-flexcan", .data = &fsl_imx6q_devtype_data, },
 	{ .compatible = "fsl,imx28-flexcan", .data = &fsl_imx28_devtype_data, },
-	{ .compatible = "fsl,imx53-flexcan", .data = &fsl_p1010_devtype_data, },
-	{ .compatible = "fsl,imx35-flexcan", .data = &fsl_p1010_devtype_data, },
-	{ .compatible = "fsl,imx25-flexcan", .data = &fsl_p1010_devtype_data, },
+	{ .compatible = "fsl,imx53-flexcan", .data = &fsl_imx25_devtype_data, },
+	{ .compatible = "fsl,imx35-flexcan", .data = &fsl_imx25_devtype_data, },
+	{ .compatible = "fsl,imx25-flexcan", .data = &fsl_imx25_devtype_data, },
 	{ .compatible = "fsl,p1010-flexcan", .data = &fsl_p1010_devtype_data, },
 	{ .compatible = "fsl,vf610-flexcan", .data = &fsl_vf610_devtype_data, },
 	{ .compatible = "fsl,ls1021ar2-flexcan", .data = &fsl_ls1021a_r2_devtype_data, },
@@ -1337,18 +1344,13 @@ static int flexcan_probe(struct platform_device *pdev)
 
 	priv = netdev_priv(dev);
 
-	if (of_property_read_bool(pdev->dev.of_node, "big-endian")) {
+	if (of_property_read_bool(pdev->dev.of_node, "big-endian") ||
+	    devtype_data->quirks & FLEXCAN_QUIRK_DEFAULT_BIG_ENDIAN) {
 		priv->read = flexcan_read_be;
 		priv->write = flexcan_write_be;
 	} else {
-		if (of_device_is_compatible(pdev->dev.of_node,
-					    "fsl,p1010-flexcan")) {
-			priv->read = flexcan_read_be;
-			priv->write = flexcan_write_be;
-		} else {
-			priv->read = flexcan_read_le;
-			priv->write = flexcan_write_le;
-		}
+		priv->read = flexcan_read_le;
+		priv->write = flexcan_write_le;
 	}
 
 	priv->can.clock.freq = clock_freq;
-- 
cgit v1.1


From 9a62dcf486c10daf5366f29df1c799f69b1510f9 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig?= <u.kleine-koenig@pengutronix.de>
Date: Wed, 25 Apr 2018 16:50:40 +0200
Subject: arm: dts: imx[35]*: declare flexcan devices to be compatible to
 imx25's flexcan
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Commit d50f4630c2e1 ("arm: dts: Remove p1010-flexcan compatible from imx
series dts") removed the fallback compatible "fsl,p1010-flexcan" from
the imx device trees. As the flexcan cores on i.MX25, i.MX35 and i.MX53
are identical, introduce the first as fallback for the two latter ones.

Fixes: d50f4630c2e1 ("arm: dts: Remove p1010-flexcan compatible from imx series dts")
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Cc: linux-stable <stable@vger.kernel.org> # >= v4.16
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
---
 arch/arm/boot/dts/imx35.dtsi | 4 ++--
 arch/arm/boot/dts/imx53.dtsi | 4 ++--
 2 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/arch/arm/boot/dts/imx35.dtsi b/arch/arm/boot/dts/imx35.dtsi
index bf343195..54111ed 100644
--- a/arch/arm/boot/dts/imx35.dtsi
+++ b/arch/arm/boot/dts/imx35.dtsi
@@ -303,7 +303,7 @@
 			};
 
 			can1: can@53fe4000 {
-				compatible = "fsl,imx35-flexcan";
+				compatible = "fsl,imx35-flexcan", "fsl,imx25-flexcan";
 				reg = <0x53fe4000 0x1000>;
 				clocks = <&clks 33>, <&clks 33>;
 				clock-names = "ipg", "per";
@@ -312,7 +312,7 @@
 			};
 
 			can2: can@53fe8000 {
-				compatible = "fsl,imx35-flexcan";
+				compatible = "fsl,imx35-flexcan", "fsl,imx25-flexcan";
 				reg = <0x53fe8000 0x1000>;
 				clocks = <&clks 34>, <&clks 34>;
 				clock-names = "ipg", "per";
diff --git a/arch/arm/boot/dts/imx53.dtsi b/arch/arm/boot/dts/imx53.dtsi
index 7d647d0..3d65c01 100644
--- a/arch/arm/boot/dts/imx53.dtsi
+++ b/arch/arm/boot/dts/imx53.dtsi
@@ -551,7 +551,7 @@
 			};
 
 			can1: can@53fc8000 {
-				compatible = "fsl,imx53-flexcan";
+				compatible = "fsl,imx53-flexcan", "fsl,imx25-flexcan";
 				reg = <0x53fc8000 0x4000>;
 				interrupts = <82>;
 				clocks = <&clks IMX5_CLK_CAN1_IPG_GATE>,
@@ -561,7 +561,7 @@
 			};
 
 			can2: can@53fcc000 {
-				compatible = "fsl,imx53-flexcan";
+				compatible = "fsl,imx53-flexcan", "fsl,imx25-flexcan";
 				reg = <0x53fcc000 0x4000>;
 				interrupts = <83>;
 				clocks = <&clks IMX5_CLK_CAN2_IPG_GATE>,
-- 
cgit v1.1


From 6ee00865ffe4e8c8ba4a68d26db53c7ec09bbb89 Mon Sep 17 00:00:00 2001
From: Jimmy Assarsson <extja@kvaser.com>
Date: Fri, 20 Apr 2018 14:38:46 +0200
Subject: can: kvaser_usb: Increase correct stats counter in
 kvaser_usb_rx_can_msg()

Increase rx_dropped, if alloc_can_skb() fails, not tx_dropped.

Signed-off-by: Jimmy Assarsson <extja@kvaser.com>
Cc: linux-stable <stable@vger.kernel.org>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
---
 drivers/net/can/usb/kvaser_usb.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/can/usb/kvaser_usb.c b/drivers/net/can/usb/kvaser_usb.c
index 63587b8..daed57d 100644
--- a/drivers/net/can/usb/kvaser_usb.c
+++ b/drivers/net/can/usb/kvaser_usb.c
@@ -1179,7 +1179,7 @@ static void kvaser_usb_rx_can_msg(const struct kvaser_usb *dev,
 
 	skb = alloc_can_skb(priv->netdev, &cf);
 	if (!skb) {
-		stats->tx_dropped++;
+		stats->rx_dropped++;
 		return;
 	}
 
-- 
cgit v1.1


From 1469c5f033a287dc25d113ea65c498c0603fbaa1 Mon Sep 17 00:00:00 2001
From: Geert Uytterhoeven <geert+renesas@glider.be>
Date: Thu, 3 May 2018 15:02:33 +0200
Subject: dt-bindings: can: rcar_can: Fix R8A7796 SoC name

R8A7796 is R-Car M3-W.

Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
---
 Documentation/devicetree/bindings/net/can/rcar_canfd.txt | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/Documentation/devicetree/bindings/net/can/rcar_canfd.txt b/Documentation/devicetree/bindings/net/can/rcar_canfd.txt
index 93c3a6a..1a4ee1d 100644
--- a/Documentation/devicetree/bindings/net/can/rcar_canfd.txt
+++ b/Documentation/devicetree/bindings/net/can/rcar_canfd.txt
@@ -5,7 +5,7 @@ Required properties:
 - compatible: Must contain one or more of the following:
   - "renesas,rcar-gen3-canfd" for R-Car Gen3 compatible controller.
   - "renesas,r8a7795-canfd" for R8A7795 (R-Car H3) compatible controller.
-  - "renesas,r8a7796-canfd" for R8A7796 (R-Car M3) compatible controller.
+  - "renesas,r8a7796-canfd" for R8A7796 (R-Car M3-W) compatible controller.
 
   When compatible with the generic version, nodes must list the
   SoC-specific version corresponding to the platform first, followed by the
-- 
cgit v1.1


From 0a4fe40efb04686529d998716d1680429d0b586b Mon Sep 17 00:00:00 2001
From: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Date: Thu, 26 Apr 2018 22:41:14 +0300
Subject: DT: net: can: rcar_canfd: document R8A77970 bindings

Document the R-Car V3M (R8A77970) SoC support in the R-Car CAN-FD bindings.

Signed-off-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Reviewed-by: Ramesh Shanmugasundaram <ramesh.shanmugasundaram@bp.renesas.com>
Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
---
 Documentation/devicetree/bindings/net/can/rcar_canfd.txt | 1 +
 1 file changed, 1 insertion(+)

diff --git a/Documentation/devicetree/bindings/net/can/rcar_canfd.txt b/Documentation/devicetree/bindings/net/can/rcar_canfd.txt
index 1a4ee1d..59dd13a 100644
--- a/Documentation/devicetree/bindings/net/can/rcar_canfd.txt
+++ b/Documentation/devicetree/bindings/net/can/rcar_canfd.txt
@@ -6,6 +6,7 @@ Required properties:
   - "renesas,rcar-gen3-canfd" for R-Car Gen3 compatible controller.
   - "renesas,r8a7795-canfd" for R8A7795 (R-Car H3) compatible controller.
   - "renesas,r8a7796-canfd" for R8A7796 (R-Car M3-W) compatible controller.
+  - "renesas,r8a77970-canfd" for R8A77970 (R-Car V3M) compatible controller.
 
   When compatible with the generic version, nodes must list the
   SoC-specific version corresponding to the platform first, followed by the
-- 
cgit v1.1


From 7a25ac2f71a409e77dd5c85cf3cbe1cbf2ae77f3 Mon Sep 17 00:00:00 2001
From: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Date: Fri, 27 Apr 2018 21:53:33 +0300
Subject: DT: net: can: rcar_canfd: document R8A77980 bindings

Document the R-Car V3H (R8A77980) SoC support in the R-Car CAN-FD bindings.

Signed-off-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
---
 Documentation/devicetree/bindings/net/can/rcar_canfd.txt | 1 +
 1 file changed, 1 insertion(+)

diff --git a/Documentation/devicetree/bindings/net/can/rcar_canfd.txt b/Documentation/devicetree/bindings/net/can/rcar_canfd.txt
index 59dd13a..ac71daa 100644
--- a/Documentation/devicetree/bindings/net/can/rcar_canfd.txt
+++ b/Documentation/devicetree/bindings/net/can/rcar_canfd.txt
@@ -7,6 +7,7 @@ Required properties:
   - "renesas,r8a7795-canfd" for R8A7795 (R-Car H3) compatible controller.
   - "renesas,r8a7796-canfd" for R8A7796 (R-Car M3-W) compatible controller.
   - "renesas,r8a77970-canfd" for R8A77970 (R-Car V3M) compatible controller.
+  - "renesas,r8a77980-canfd" for R8A77980 (R-Car V3H) compatible controller.
 
   When compatible with the generic version, nodes must list the
   SoC-specific version corresponding to the platform first, followed by the
-- 
cgit v1.1


From c11c7bfd213495784b22ef82a69b6489f8d0092f Mon Sep 17 00:00:00 2001
From: Matthew Auld <matthew.auld@intel.com>
Date: Wed, 2 May 2018 20:50:21 +0100
Subject: drm/i915/userptr: reject zero user_size

Operating on a zero sized GEM userptr object will lead to explosions.

Fixes: 5cc9ed4b9a7a ("drm/i915: Introduce mapping of user pages into video memory (userptr) ioctl")
Testcase: igt/gem_userptr_blits/input-checking
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20180502195021.30900-1-matthew.auld@intel.com
---
 drivers/gpu/drm/i915/i915_gem_userptr.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_gem_userptr.c b/drivers/gpu/drm/i915/i915_gem_userptr.c
index d596a83..854bd51 100644
--- a/drivers/gpu/drm/i915/i915_gem_userptr.c
+++ b/drivers/gpu/drm/i915/i915_gem_userptr.c
@@ -778,6 +778,9 @@ i915_gem_userptr_ioctl(struct drm_device *dev,
 			    I915_USERPTR_UNSYNCHRONIZED))
 		return -EINVAL;
 
+	if (!args->user_size)
+		return -EINVAL;
+
 	if (offset_in_page(args->user_ptr | args->user_size))
 		return -EINVAL;
 
-- 
cgit v1.1


From 98dc0454c023985cb31de2578c941391a900e941 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Sat, 5 May 2018 10:10:13 +0100
Subject: drm/i915/selftests: Refactor common flush_test()

Pull igt_flush_test() out into its own library before copying and
pasting the code for a third time.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180505091014.26126-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/Makefile                    |  3 +-
 drivers/gpu/drm/i915/selftests/igt_flush_test.c  | 64 ++++++++++++++++++++++
 drivers/gpu/drm/i915/selftests/igt_flush_test.h  | 14 +++++
 drivers/gpu/drm/i915/selftests/intel_hangcheck.c | 66 +++--------------------
 drivers/gpu/drm/i915/selftests/intel_lrc.c       | 68 +++---------------------
 5 files changed, 93 insertions(+), 122 deletions(-)
 create mode 100644 drivers/gpu/drm/i915/selftests/igt_flush_test.c
 create mode 100644 drivers/gpu/drm/i915/selftests/igt_flush_test.h

diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile
index 00c1338..4c6adae 100644
--- a/drivers/gpu/drm/i915/Makefile
+++ b/drivers/gpu/drm/i915/Makefile
@@ -158,7 +158,8 @@ i915-y += dvo_ch7017.o \
 i915-$(CONFIG_DRM_I915_CAPTURE_ERROR) += i915_gpu_error.o
 i915-$(CONFIG_DRM_I915_SELFTEST) += \
 	selftests/i915_random.o \
-	selftests/i915_selftest.o
+	selftests/i915_selftest.o \
+	selftests/igt_flush_test.o
 
 # virtual gpu code
 i915-y += i915_vgpu.o
diff --git a/drivers/gpu/drm/i915/selftests/igt_flush_test.c b/drivers/gpu/drm/i915/selftests/igt_flush_test.c
new file mode 100644
index 0000000..abff2f0
--- /dev/null
+++ b/drivers/gpu/drm/i915/selftests/igt_flush_test.c
@@ -0,0 +1,64 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2018 Intel Corporation
+ */
+
+#include "../i915_drv.h"
+
+#include "../i915_selftest.h"
+#include "igt_flush_test.h"
+
+struct wedge_me {
+	struct delayed_work work;
+	struct drm_i915_private *i915;
+	const void *symbol;
+};
+
+static void wedge_me(struct work_struct *work)
+{
+	struct wedge_me *w = container_of(work, typeof(*w), work.work);
+
+	pr_err("%pS timed out, cancelling all further testing.\n", w->symbol);
+
+	GEM_TRACE("%pS timed out.\n", w->symbol);
+	GEM_TRACE_DUMP();
+
+	i915_gem_set_wedged(w->i915);
+}
+
+static void __init_wedge(struct wedge_me *w,
+			 struct drm_i915_private *i915,
+			 long timeout,
+			 const void *symbol)
+{
+	w->i915 = i915;
+	w->symbol = symbol;
+
+	INIT_DELAYED_WORK_ONSTACK(&w->work, wedge_me);
+	schedule_delayed_work(&w->work, timeout);
+}
+
+static void __fini_wedge(struct wedge_me *w)
+{
+	cancel_delayed_work_sync(&w->work);
+	destroy_delayed_work_on_stack(&w->work);
+	w->i915 = NULL;
+}
+
+#define wedge_on_timeout(W, DEV, TIMEOUT)				\
+	for (__init_wedge((W), (DEV), (TIMEOUT), __builtin_return_address(0)); \
+	     (W)->i915;							\
+	     __fini_wedge((W)))
+
+int igt_flush_test(struct drm_i915_private *i915, unsigned int flags)
+{
+	struct wedge_me w;
+
+	cond_resched();
+
+	wedge_on_timeout(&w, i915, HZ)
+		i915_gem_wait_for_idle(i915, flags);
+
+	return i915_terminally_wedged(&i915->gpu_error) ? -EIO : 0;
+}
diff --git a/drivers/gpu/drm/i915/selftests/igt_flush_test.h b/drivers/gpu/drm/i915/selftests/igt_flush_test.h
new file mode 100644
index 0000000..63e0099
--- /dev/null
+++ b/drivers/gpu/drm/i915/selftests/igt_flush_test.h
@@ -0,0 +1,14 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2018 Intel Corporation
+ */
+
+#ifndef IGT_FLUSH_TEST_H
+#define IGT_FLUSH_TEST_H
+
+struct drm_i915_private;
+
+int igt_flush_test(struct drm_i915_private *i915, unsigned int flags);
+
+#endif /* IGT_FLUSH_TEST_H */
diff --git a/drivers/gpu/drm/i915/selftests/intel_hangcheck.c b/drivers/gpu/drm/i915/selftests/intel_hangcheck.c
index c61bf65..438e0b0 100644
--- a/drivers/gpu/drm/i915/selftests/intel_hangcheck.c
+++ b/drivers/gpu/drm/i915/selftests/intel_hangcheck.c
@@ -26,6 +26,7 @@
 
 #include "../i915_selftest.h"
 #include "i915_random.h"
+#include "igt_flush_test.h"
 
 #include "mock_context.h"
 #include "mock_drm.h"
@@ -253,61 +254,6 @@ static u32 hws_seqno(const struct hang *h, const struct i915_request *rq)
 	return READ_ONCE(h->seqno[rq->fence.context % (PAGE_SIZE/sizeof(u32))]);
 }
 
-struct wedge_me {
-	struct delayed_work work;
-	struct drm_i915_private *i915;
-	const void *symbol;
-};
-
-static void wedge_me(struct work_struct *work)
-{
-	struct wedge_me *w = container_of(work, typeof(*w), work.work);
-
-	pr_err("%pS timed out, cancelling all further testing.\n", w->symbol);
-
-	GEM_TRACE("%pS timed out.\n", w->symbol);
-	GEM_TRACE_DUMP();
-
-	i915_gem_set_wedged(w->i915);
-}
-
-static void __init_wedge(struct wedge_me *w,
-			 struct drm_i915_private *i915,
-			 long timeout,
-			 const void *symbol)
-{
-	w->i915 = i915;
-	w->symbol = symbol;
-
-	INIT_DELAYED_WORK_ONSTACK(&w->work, wedge_me);
-	schedule_delayed_work(&w->work, timeout);
-}
-
-static void __fini_wedge(struct wedge_me *w)
-{
-	cancel_delayed_work_sync(&w->work);
-	destroy_delayed_work_on_stack(&w->work);
-	w->i915 = NULL;
-}
-
-#define wedge_on_timeout(W, DEV, TIMEOUT)				\
-	for (__init_wedge((W), (DEV), (TIMEOUT), __builtin_return_address(0)); \
-	     (W)->i915;							\
-	     __fini_wedge((W)))
-
-static noinline int
-flush_test(struct drm_i915_private *i915, unsigned int flags)
-{
-	struct wedge_me w;
-
-	cond_resched();
-
-	wedge_on_timeout(&w, i915, HZ)
-		i915_gem_wait_for_idle(i915, flags);
-
-	return i915_terminally_wedged(&i915->gpu_error) ? -EIO : 0;
-}
-
 static void hang_fini(struct hang *h)
 {
 	*h->batch = MI_BATCH_BUFFER_END;
@@ -321,7 +267,7 @@ static void hang_fini(struct hang *h)
 
 	kernel_context_close(h->ctx);
 
-	flush_test(h->i915, I915_WAIT_LOCKED);
+	igt_flush_test(h->i915, I915_WAIT_LOCKED);
 }
 
 static bool wait_until_running(struct hang *h, struct i915_request *rq)
@@ -575,7 +521,7 @@ static int __igt_reset_engine(struct drm_i915_private *i915, bool active)
 		if (err)
 			break;
 
-		err = flush_test(i915, 0);
+		err = igt_flush_test(i915, 0);
 		if (err)
 			break;
 	}
@@ -874,7 +820,7 @@ unwind:
 		if (err)
 			break;
 
-		err = flush_test(i915, 0);
+		err = igt_flush_test(i915, 0);
 		if (err)
 			break;
 	}
@@ -1168,7 +1114,7 @@ static int igt_reset_queue(void *arg)
 
 		i915_request_put(prev);
 
-		err = flush_test(i915, I915_WAIT_LOCKED);
+		err = igt_flush_test(i915, I915_WAIT_LOCKED);
 		if (err)
 			break;
 	}
@@ -1280,7 +1226,7 @@ int intel_hangcheck_live_selftests(struct drm_i915_private *i915)
 	err = i915_subtests(tests, i915);
 
 	mutex_lock(&i915->drm.struct_mutex);
-	flush_test(i915, I915_WAIT_LOCKED);
+	igt_flush_test(i915, I915_WAIT_LOCKED);
 	mutex_unlock(&i915->drm.struct_mutex);
 
 	i915_modparams.enable_hangcheck = saved_hangcheck;
diff --git a/drivers/gpu/drm/i915/selftests/intel_lrc.c b/drivers/gpu/drm/i915/selftests/intel_lrc.c
index b7460b5..1b8a071 100644
--- a/drivers/gpu/drm/i915/selftests/intel_lrc.c
+++ b/drivers/gpu/drm/i915/selftests/intel_lrc.c
@@ -5,6 +5,7 @@
  */
 
 #include "../i915_selftest.h"
+#include "igt_flush_test.h"
 
 #include "mock_context.h"
 
@@ -168,61 +169,6 @@ static u32 hws_seqno(const struct spinner *spin, const struct i915_request *rq)
 	return READ_ONCE(*seqno);
 }
 
-struct wedge_me {
-	struct delayed_work work;
-	struct drm_i915_private *i915;
-	const void *symbol;
-};
-
-static void wedge_me(struct work_struct *work)
-{
-	struct wedge_me *w = container_of(work, typeof(*w), work.work);
-
-	pr_err("%pS timed out, cancelling all further testing.\n", w->symbol);
-
-	GEM_TRACE("%pS timed out.\n", w->symbol);
-	GEM_TRACE_DUMP();
-
-	i915_gem_set_wedged(w->i915);
-}
-
-static void __init_wedge(struct wedge_me *w,
-			 struct drm_i915_private *i915,
-			 long timeout,
-			 const void *symbol)
-{
-	w->i915 = i915;
-	w->symbol = symbol;
-
-	INIT_DELAYED_WORK_ONSTACK(&w->work, wedge_me);
-	schedule_delayed_work(&w->work, timeout);
-}
-
-static void __fini_wedge(struct wedge_me *w)
-{
-	cancel_delayed_work_sync(&w->work);
-	destroy_delayed_work_on_stack(&w->work);
-	w->i915 = NULL;
-}
-
-#define wedge_on_timeout(W, DEV, TIMEOUT)				\
-	for (__init_wedge((W), (DEV), (TIMEOUT), __builtin_return_address(0)); \
-	     (W)->i915;							\
-	     __fini_wedge((W)))
-
-static noinline int
-flush_test(struct drm_i915_private *i915, unsigned int flags)
-{
-	struct wedge_me w;
-
-	cond_resched();
-
-	wedge_on_timeout(&w, i915, HZ)
-		i915_gem_wait_for_idle(i915, flags);
-
-	return i915_terminally_wedged(&i915->gpu_error) ? -EIO : 0;
-}
-
 static void spinner_end(struct spinner *spin)
 {
 	*spin->batch = MI_BATCH_BUFFER_END;
@@ -295,7 +241,7 @@ static int live_sanitycheck(void *arg)
 		}
 
 		spinner_end(&spin);
-		if (flush_test(i915, I915_WAIT_LOCKED)) {
+		if (igt_flush_test(i915, I915_WAIT_LOCKED)) {
 			err = -EIO;
 			goto err_ctx;
 		}
@@ -307,7 +253,7 @@ err_ctx:
 err_spin:
 	spinner_fini(&spin);
 err_unlock:
-	flush_test(i915, I915_WAIT_LOCKED);
+	igt_flush_test(i915, I915_WAIT_LOCKED);
 	mutex_unlock(&i915->drm.struct_mutex);
 	return err;
 }
@@ -380,7 +326,7 @@ static int live_preempt(void *arg)
 
 		spinner_end(&spin_hi);
 		spinner_end(&spin_lo);
-		if (flush_test(i915, I915_WAIT_LOCKED)) {
+		if (igt_flush_test(i915, I915_WAIT_LOCKED)) {
 			err = -EIO;
 			goto err_ctx_lo;
 		}
@@ -396,7 +342,7 @@ err_spin_lo:
 err_spin_hi:
 	spinner_fini(&spin_hi);
 err_unlock:
-	flush_test(i915, I915_WAIT_LOCKED);
+	igt_flush_test(i915, I915_WAIT_LOCKED);
 	mutex_unlock(&i915->drm.struct_mutex);
 	return err;
 }
@@ -470,7 +416,7 @@ static int live_late_preempt(void *arg)
 
 		spinner_end(&spin_hi);
 		spinner_end(&spin_lo);
-		if (flush_test(i915, I915_WAIT_LOCKED)) {
+		if (igt_flush_test(i915, I915_WAIT_LOCKED)) {
 			err = -EIO;
 			goto err_ctx_lo;
 		}
@@ -486,7 +432,7 @@ err_spin_lo:
 err_spin_hi:
 	spinner_fini(&spin_hi);
 err_unlock:
-	flush_test(i915, I915_WAIT_LOCKED);
+	igt_flush_test(i915, I915_WAIT_LOCKED);
 	mutex_unlock(&i915->drm.struct_mutex);
 	return err;
 
-- 
cgit v1.1


From 7c2f5bc5f0f41a3e294f5fa3b010a10f47512706 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Sat, 5 May 2018 10:10:14 +0100
Subject: drm/i915/selftests: Flush GPU activity before completing
 live_contexts

igt_ctx_exec() expects that we retire all active requests/objects before
completing, so that when we clean up the files afterwards they are ready
to be freed. Before we do so, it is then prudent to ensure that we have
indeed retired the GPU activity, raising an error if it fails. If we do
not, we run the risk of triggering an assertion when freeing the object:

  __i915_gem_free_objects:4793 GEM_BUG_ON(i915_gem_object_is_active(obj))

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180505091014.26126-2-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/selftests/i915_gem_context.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_context.c b/drivers/gpu/drm/i915/selftests/i915_gem_context.c
index 7ecaed5..ddb03f0 100644
--- a/drivers/gpu/drm/i915/selftests/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/selftests/i915_gem_context.c
@@ -23,6 +23,7 @@
  */
 
 #include "../i915_selftest.h"
+#include "igt_flush_test.h"
 
 #include "mock_drm.h"
 #include "huge_gem_object.h"
@@ -411,6 +412,8 @@ static int igt_ctx_exec(void *arg)
 	}
 
 out_unlock:
+	if (igt_flush_test(i915, I915_WAIT_LOCKED))
+		err = -EIO;
 	mutex_unlock(&i915->drm.struct_mutex);
 
 	mock_file_free(i915, file);
-- 
cgit v1.1


From aaefa06a0ea845a7088585ca42259515769ea496 Mon Sep 17 00:00:00 2001
From: Matthew Auld <matthew.auld@intel.com>
Date: Thu, 1 Mar 2018 11:46:39 +0000
Subject: drm/i915: don't leak the pin_map on error

Add some onion to populate_lr_context.

v2: prefer err_unpin_ctx
    drop the fixes tag, worst case we just spew a warn before everything
    is cleaned up and balance is restored

Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20180301114639.510-1-matthew.auld@intel.com
---
 drivers/gpu/drm/i915/intel_lrc.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 046adf3..da08225 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -2574,8 +2574,10 @@ populate_lr_context(struct i915_gem_context *ctx,
 
 		defaults = i915_gem_object_pin_map(engine->default_state,
 						   I915_MAP_WB);
-		if (IS_ERR(defaults))
-			return PTR_ERR(defaults);
+		if (IS_ERR(defaults)) {
+			ret = PTR_ERR(defaults);
+			goto err_unpin_ctx;
+		}
 
 		memcpy(vaddr + start, defaults + start, engine->context_size);
 		i915_gem_object_unpin_map(engine->default_state);
@@ -2593,9 +2595,9 @@ populate_lr_context(struct i915_gem_context *ctx,
 			_MASKED_BIT_ENABLE(CTX_CTRL_ENGINE_CTX_RESTORE_INHIBIT |
 					   CTX_CTRL_ENGINE_CTX_SAVE_INHIBIT);
 
+err_unpin_ctx:
 	i915_gem_object_unpin_map(ctx_obj);
-
-	return 0;
+	return ret;
 }
 
 static int execlists_context_deferred_alloc(struct i915_gem_context *ctx,
-- 
cgit v1.1


From 9d219554d9bf59875b4e571a0392d620e8954879 Mon Sep 17 00:00:00 2001
From: Rodrigo Vivi <rodrigo.vivi@intel.com>
Date: Wed, 2 May 2018 10:52:55 -0700
Subject: drm/i915: Adjust eDP's logical vco in a reliable place.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

On intel_dp_compute_config() we were calculating the needed vco
for eDP on gen9 and we stashing it in
intel_atomic_state.cdclk.logical.vco

However few moments later on intel_modeset_checks() we fully
replace entire intel_atomic_state.cdclk.logical with
dev_priv->cdclk.logical fully overwriting the logical desired
vco for eDP on gen9.

So, with wrong VCO value we end up with wrong desired cdclk, but
also it will raise a lot of WARNs: On gen9, when we read
CDCLK_CTL to verify if we configured properly the desired
frequency the CD Frequency Select bits [27:26] == 10b can mean
337.5 or 308.57 MHz depending on the VCO. So if we have wrong
VCO value stashed we will believe the frequency selection didn't
stick and start to raise WARNs of cdclk mismatch.

[   42.857519] [drm:intel_dump_cdclk_state [i915]] Changing CDCLK to 308571 kHz, VCO 8640000 kHz, ref 24000 kHz, bypass 24000 kHz, voltage level 0
[   42.897269] cdclk state doesn't match!
[   42.901052] WARNING: CPU: 5 PID: 1116 at drivers/gpu/drm/i915/intel_cdclk.c:2084 intel_set_cdclk+0x5d/0x110 [i915]
[   42.938004] RIP: 0010:intel_set_cdclk+0x5d/0x110 [i915]
[   43.155253] WARNING: CPU: 5 PID: 1116 at drivers/gpu/drm/i915/intel_cdclk.c:2084 intel_set_cdclk+0x5d/0x110 [i915]
[   43.170277] [drm:intel_dump_cdclk_state [i915]] [hw state] 337500 kHz, VCO 8100000 kHz, ref 24000 kHz, bypass 24000 kHz, voltage level 0
[   43.182566] [drm:intel_dump_cdclk_state [i915]] [sw state] 308571 kHz, VCO 8640000 kHz, ref 24000 kHz, bypass 24000 kHz, voltage level 0

v2: Move the entire eDP's vco logical adjustment to inside
    the skl_modeset_calc_cdclk as suggested by Ville.

Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Fixes: bb0f4aab0e76 ("drm/i915: Track full cdclk state for the logical and actual cdclk frequencies")
Cc: <stable@vger.kernel.org> # v4.12+
Link: https://patchwork.freedesktop.org/patch/msgid/20180502175255.5344-1-rodrigo.vivi@intel.com
(cherry picked from commit 3297234a05ab1e90091b0574db4c397ef0e90d5f)
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
---
 drivers/gpu/drm/i915/intel_cdclk.c | 41 ++++++++++++++++++++++++++++++++++----
 drivers/gpu/drm/i915/intel_dp.c    | 20 -------------------
 2 files changed, 37 insertions(+), 24 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_cdclk.c b/drivers/gpu/drm/i915/intel_cdclk.c
index 32d24c6..704ddb4 100644
--- a/drivers/gpu/drm/i915/intel_cdclk.c
+++ b/drivers/gpu/drm/i915/intel_cdclk.c
@@ -2302,9 +2302,44 @@ static int bdw_modeset_calc_cdclk(struct drm_atomic_state *state)
 	return 0;
 }
 
+static int skl_dpll0_vco(struct intel_atomic_state *intel_state)
+{
+	struct drm_i915_private *dev_priv = to_i915(intel_state->base.dev);
+	struct intel_crtc *crtc;
+	struct intel_crtc_state *crtc_state;
+	int vco, i;
+
+	vco = intel_state->cdclk.logical.vco;
+	if (!vco)
+		vco = dev_priv->skl_preferred_vco_freq;
+
+	for_each_new_intel_crtc_in_state(intel_state, crtc, crtc_state, i) {
+		if (!crtc_state->base.enable)
+			continue;
+
+		if (!intel_crtc_has_type(crtc_state, INTEL_OUTPUT_EDP))
+			continue;
+
+		/*
+		 * DPLL0 VCO may need to be adjusted to get the correct
+		 * clock for eDP. This will affect cdclk as well.
+		 */
+		switch (crtc_state->port_clock / 2) {
+		case 108000:
+		case 216000:
+			vco = 8640000;
+			break;
+		default:
+			vco = 8100000;
+			break;
+		}
+	}
+
+	return vco;
+}
+
 static int skl_modeset_calc_cdclk(struct drm_atomic_state *state)
 {
-	struct drm_i915_private *dev_priv = to_i915(state->dev);
 	struct intel_atomic_state *intel_state = to_intel_atomic_state(state);
 	int min_cdclk, cdclk, vco;
 
@@ -2312,9 +2347,7 @@ static int skl_modeset_calc_cdclk(struct drm_atomic_state *state)
 	if (min_cdclk < 0)
 		return min_cdclk;
 
-	vco = intel_state->cdclk.logical.vco;
-	if (!vco)
-		vco = dev_priv->skl_preferred_vco_freq;
+	vco = skl_dpll0_vco(intel_state);
 
 	/*
 	 * FIXME should also account for plane ratio
diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index 9a4a51e..b7b4cfd 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -1881,26 +1881,6 @@ found:
 				reduce_m_n);
 	}
 
-	/*
-	 * DPLL0 VCO may need to be adjusted to get the correct
-	 * clock for eDP. This will affect cdclk as well.
-	 */
-	if (intel_dp_is_edp(intel_dp) && IS_GEN9_BC(dev_priv)) {
-		int vco;
-
-		switch (pipe_config->port_clock / 2) {
-		case 108000:
-		case 216000:
-			vco = 8640000;
-			break;
-		default:
-			vco = 8100000;
-			break;
-		}
-
-		to_intel_atomic_state(pipe_config->base.state)->cdclk.logical.vco = vco;
-	}
-
 	if (!HAS_DDI(dev_priv))
 		intel_dp_set_clock(encoder, pipe_config);
 
-- 
cgit v1.1


From 660d88e74cf6e16252e366616f158d84dc9dc6a3 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= <ville.syrjala@linux.intel.com>
Date: Thu, 26 Apr 2018 19:30:15 +0300
Subject: drm/i915: Correctly populate user mode h/vdisplay with pipe src size
 during readout
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

During state readout we first read out the pipe src size, store
that information in the user mode h/vdisplay, but later on we overwrite
that with the actual crtc timings. That makes our read out crtc state
inconsistent with itself when the BIOS has enabled the panel fitter to
scale the pipe contents. Let's preserve the pipe src size based
information in the user mode to make things consistent again.

This fixes a problem introduced by commit a2936e3d9a9c ("drm/i915:
Use drm_mode_get_hv_timing() to populate plane clip rectangle")
where the inconsistent state is now leading the plane clipping code
to report a failure on account the plane dst coordinates not matching
the user mode size. Previously we did the plane clipping based on
the pipe src size instead and thus never noticed the inconsistency.

The failure manifests as a WARN:
[    0.762117] [drm:intel_dump_pipe_config [i915]] requested mode:
[    0.762142] [drm:drm_mode_debug_printmodeline [drm]] Modeline 0:"1366x768" 60 72143 1366 1414 1446 1526 768 771 777 784 0x40 0xa
...
[    0.762327] [drm:intel_dump_pipe_config [i915]] port clock: 72143, pipe src size: 1024x768, pixel rate 72143
...
[    0.764666] [drm:drm_atomic_helper_check_plane_state [drm_kms_helper]] Plane must cover entire CRTC
[    0.764690] [drm:drm_rect_debug_print [drm]] dst: 1024x768+0+0
[    0.764711] [drm:drm_rect_debug_print [drm]] clip: 1366x768+0+0
[    0.764713] ------------[ cut here ]------------
[    0.764714] Could not determine valid watermarks for inherited state
[    0.764792] WARNING: CPU: 4 PID: 159 at drivers/gpu/drm/i915/intel_display.c:14584 intel_modeset_init+0x3ce/0x19d0 [i915]
...

Cc: FadeMind <fademind@gmail.com>
Cc: Dave Jones <davej@codemonkey.org.uk>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Reported-by: FadeMind <fademind@gmail.com>
Reported-by: Dave Jones <davej@codemonkey.org.uk>
Tested-by: Dave Jones <davej@codemonkey.org.uk>
References: https://lists.freedesktop.org/archives/intel-gfx/2018-April/163186.html
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105992
Fixes: a2936e3d9a9c ("drm/i915: Use drm_mode_get_hv_timing() to populate plane clip rectangle")
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180426163015.14232-1-ville.syrjala@linux.intel.com
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Tested-by: Larry Finger <Larry.Finger@lwfinger.net>
Tested-by: FadeMind <fademind@gmail.com>
(cherry picked from commit bd4cd03c81010dcd4e6f0e02e4c15f44aefe12d1)
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
---
 drivers/gpu/drm/i915/intel_display.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 3b48fd2..56004ff 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -15178,6 +15178,8 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 		memset(&crtc->base.mode, 0, sizeof(crtc->base.mode));
 		if (crtc_state->base.active) {
 			intel_mode_from_pipe_config(&crtc->base.mode, crtc_state);
+			crtc->base.mode.hdisplay = crtc_state->pipe_src_w;
+			crtc->base.mode.vdisplay = crtc_state->pipe_src_h;
 			intel_mode_from_pipe_config(&crtc_state->base.adjusted_mode, crtc_state);
 			WARN_ON(drm_atomic_set_mode_for_crtc(crtc->base.state, &crtc->base.mode));
 
-- 
cgit v1.1


From e8f48f96db7e482995743f461b3e8a5c1a102533 Mon Sep 17 00:00:00 2001
From: Florent Flament <contact@florentflament.com>
Date: Thu, 19 Apr 2018 19:07:00 +0300
Subject: drm/i915: Fix drm:intel_enable_lvds ERROR message in kernel log
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Fix `[drm:intel_enable_lvds] *ERROR* timed out waiting for panel to
power on` in kernel log at boot time.

Toshiba Satellite Z930 laptops needs between 1 and 2 seconds to power
on its screen during Intel i915 DRM initialization. This currently
results in a `[drm:intel_enable_lvds] *ERROR* timed out waiting for
panel to power on` message appearing in the kernel log during boot
time and when stopping the machine.

This change increases the timeout of the `intel_enable_lvds` function
from 1 to 5 seconds, letting enough time for the Satellite 930 LCD
screen to power on, and suppressing the error message from the kernel
log.

This patch has been successfully tested on Linux 4.14 running on a
Toshiba Satellite Z930.

[vsyrjala: bump the timeout from 2 to 5 seconds to match the DP
 code and properly cover the max hw timeout of ~4 seconds, and
 drop the comment about the specific machine since this is not
 a particulary surprising issue, nor specific to that one machine]

Signed-off-by: Florent Flament <contact@florentflament.com>
Cc: stable@vger.kernel.org
Cc: Pavel Petrovic <ppetrovic@acm.org>
Cc: Sérgio M. Basto <sergio@serjux.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=103414
References: https://bugzilla.kernel.org/show_bug.cgi?id=57591
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180419160700.19828-1-ville.syrjala@linux.intel.com
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
(cherry picked from commit 280b54ade5914d3b4abe4f0ebe083ddbd4603246)
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
---
 drivers/gpu/drm/i915/intel_lvds.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/intel_lvds.c b/drivers/gpu/drm/i915/intel_lvds.c
index d35d2d5..8691c86 100644
--- a/drivers/gpu/drm/i915/intel_lvds.c
+++ b/drivers/gpu/drm/i915/intel_lvds.c
@@ -326,7 +326,8 @@ static void intel_enable_lvds(struct intel_encoder *encoder,
 
 	I915_WRITE(PP_CONTROL(0), I915_READ(PP_CONTROL(0)) | PANEL_POWER_ON);
 	POSTING_READ(lvds_encoder->reg);
-	if (intel_wait_for_register(dev_priv, PP_STATUS(0), PP_ON, PP_ON, 1000))
+
+	if (intel_wait_for_register(dev_priv, PP_STATUS(0), PP_ON, PP_ON, 5000))
 		DRM_ERROR("timed out waiting for panel to power on\n");
 
 	intel_panel_enable_backlight(pipe_config, conn_state);
-- 
cgit v1.1


From 4cdf65ce8cc28e72089605250b887ab70e10f750 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Tue, 8 May 2018 12:53:12 +0100
Subject: drm/i915/selftests: Return to kernel context after each test

As we flush each test and wait for idle before the next, also switch
back to the kernel context. This helps limit the amount of collateral
damage a test may cause by resetting to the default state each time (and
also helps clean up temporaries used by the test).

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180508115312.12628-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/selftests/igt_flush_test.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/gpu/drm/i915/selftests/igt_flush_test.c b/drivers/gpu/drm/i915/selftests/igt_flush_test.c
index abff2f0..7f35bdd 100644
--- a/drivers/gpu/drm/i915/selftests/igt_flush_test.c
+++ b/drivers/gpu/drm/i915/selftests/igt_flush_test.c
@@ -57,6 +57,11 @@ int igt_flush_test(struct drm_i915_private *i915, unsigned int flags)
 
 	cond_resched();
 
+	if (i915_gem_switch_to_kernel_context(i915)) {
+		pr_err("Failed to switch back to kernel context; declaring wedged\n");
+		i915_gem_set_wedged(i915);
+	}
+
 	wedge_on_timeout(&w, i915, HZ)
 		i915_gem_wait_for_idle(i915, flags);
 
-- 
cgit v1.1


From 4f6d8fcf1a53d4d544d9a13f70cb2669572b7ecb Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Mon, 7 May 2018 14:57:25 +0100
Subject: drm/i915: Flush submission tasklet after bumping priority

When called from process context tasklet_schedule() defers itself to
ksoftirqd. From experience this may cause unacceptable latencies of over
200ms in executing the submission tasklet, our goal is to reprioritise
the HW execution queue and trigger HW preemption immediately, so disable
bh over the call to schedule and force the tasklet to run afterwards if
scheduled.

v2: Keep rcu_read_lock() around for PREEMPT_RCU

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180507135731.10587-1-chris@chris-wilson.co.uk
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
---
 drivers/gpu/drm/i915/i915_gem.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 5ece6ae..89bf5d6 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -578,10 +578,12 @@ static void __fence_set_priority(struct dma_fence *fence,
 	rq = to_request(fence);
 	engine = rq->engine;
 
-	rcu_read_lock();
+	local_bh_disable();
+	rcu_read_lock(); /* RCU serialisation for set-wedged protection */
 	if (engine->schedule)
 		engine->schedule(rq, attr);
 	rcu_read_unlock();
+	local_bh_enable(); /* kick the tasklets if queues were reprioritised */
 }
 
 static void fence_set_priority(struct dma_fence *fence,
-- 
cgit v1.1


From 71ace7ca2545d7cd7522988c16ad6c94e6169366 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Mon, 7 May 2018 14:57:26 +0100
Subject: drm/i915: Disable tasklet scheduling across initial scheduling

During request submission, we call the engine->schedule() function so
that we may reorder the active requests as required for inheriting the
new request's priority. This may schedule several tasklets to run on the
local CPU, but we will need to schedule the tasklets again for the new
request. Delay all the local tasklets until the end, so that we only
have to process the queue just once.

v2: Beware PREEMPT_RCU, as then local_bh_disable() is then not a
superset of rcu_read_lock().

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180507135731.10587-2-chris@chris-wilson.co.uk
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
---
 drivers/gpu/drm/i915/i915_request.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c
index e4cf76e..f336942 100644
--- a/drivers/gpu/drm/i915/i915_request.c
+++ b/drivers/gpu/drm/i915/i915_request.c
@@ -1110,12 +1110,11 @@ void __i915_request_add(struct i915_request *request, bool flush_caches)
 	 * decide whether to preempt the entire chain so that it is ready to
 	 * run at the earliest possible convenience.
 	 */
-	rcu_read_lock();
+	local_bh_disable();
+	rcu_read_lock(); /* RCU serialisation for set-wedged protection */
 	if (engine->schedule)
 		engine->schedule(request, &request->ctx->sched);
 	rcu_read_unlock();
-
-	local_bh_disable();
 	i915_sw_fence_commit(&request->submit);
 	local_bh_enable(); /* Kick the execlists tasklet if just scheduled */
 
-- 
cgit v1.1


From a33f084cc5eaa3a71d872baad83d6738cbf0f783 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Tue, 8 May 2018 16:15:52 +0100
Subject: drm/i915: Remove unused i915_flip tracepoints
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The i915_flip* tracepoints are no longer in use since the removal of CS
flip in commit 8b5d27b911d7 ("drm/i915: Remove intel_flip_work
infrastructure")

References: 8b5d27b911d7 ("drm/i915: Remove intel_flip_work infrastructure")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Daniel Vetter <daniel.vetter@intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180508151552.31024-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/i915_trace.h | 36 ------------------------------------
 1 file changed, 36 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_trace.h b/drivers/gpu/drm/i915/i915_trace.h
index 77ee5e5..8cc3a25 100644
--- a/drivers/gpu/drm/i915/i915_trace.h
+++ b/drivers/gpu/drm/i915/i915_trace.h
@@ -834,42 +834,6 @@ DEFINE_EVENT(i915_request, i915_request_wait_end,
 	    TP_ARGS(rq)
 );
 
-TRACE_EVENT(i915_flip_request,
-	    TP_PROTO(int plane, struct drm_i915_gem_object *obj),
-
-	    TP_ARGS(plane, obj),
-
-	    TP_STRUCT__entry(
-		    __field(int, plane)
-		    __field(struct drm_i915_gem_object *, obj)
-		    ),
-
-	    TP_fast_assign(
-		    __entry->plane = plane;
-		    __entry->obj = obj;
-		    ),
-
-	    TP_printk("plane=%d, obj=%p", __entry->plane, __entry->obj)
-);
-
-TRACE_EVENT(i915_flip_complete,
-	    TP_PROTO(int plane, struct drm_i915_gem_object *obj),
-
-	    TP_ARGS(plane, obj),
-
-	    TP_STRUCT__entry(
-		    __field(int, plane)
-		    __field(struct drm_i915_gem_object *, obj)
-		    ),
-
-	    TP_fast_assign(
-		    __entry->plane = plane;
-		    __entry->obj = obj;
-		    ),
-
-	    TP_printk("plane=%d, obj=%p", __entry->plane, __entry->obj)
-);
-
 TRACE_EVENT_CONDITION(i915_reg_rw,
 	TP_PROTO(bool write, i915_reg_t reg, u64 val, int len, bool trace),
 
-- 
cgit v1.1


From 0adb90d330bb5f0d7fba511af5af3fc1ba93fb7a Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Tue, 8 May 2018 16:35:14 +0100
Subject: drm/i915: Annotate timeline lock nesting

CI noticed

<4>[   23.430701] ============================================
<4>[   23.430706] WARNING: possible recursive locking detected
<4>[   23.430713] 4.17.0-rc4-CI-CI_DRM_4156+ #1 Not tainted
<4>[   23.430720] --------------------------------------------
<4>[   23.430725] systemd-udevd/169 is trying to acquire lock:
<4>[   23.430732]         (ptrval) (&(&timeline->lock)->rlock){....}, at: move_to_timeline+0x48/0x12c [i915]
<4>[   23.430888]
                  but task is already holding lock:
<4>[   23.430894]         (ptrval) (&(&timeline->lock)->rlock){....}, at: i915_request_submit+0x1a/0x40 [i915]
<4>[   23.430995]
                  other info that might help us debug this:
<4>[   23.431002]  Possible unsafe locking scenario:

<4>[   23.431007]        CPU0
<4>[   23.431010]        ----
<4>[   23.431013]   lock(&(&timeline->lock)->rlock);
<4>[   23.431021]   lock(&(&timeline->lock)->rlock);
<4>[   23.431028]
                   *** DEADLOCK ***

<4>[   23.431036]  May be due to missing lock nesting notation

<4>[   23.431044] 5 locks held by systemd-udevd/169:
<4>[   23.431049]  #0:         (ptrval) (&dev->mutex){....}, at: __driver_attach+0x42/0xe0
<4>[   23.431065]  #1:         (ptrval) (&dev->mutex){....}, at: __driver_attach+0x50/0xe0
<4>[   23.431078]  #2:         (ptrval) (&dev->struct_mutex){+.+.}, at: i915_gem_init+0xca/0x630 [i915]
<4>[   23.431174]  #3:         (ptrval) (rcu_read_lock){....}, at: submit_notify+0x35/0x124 [i915]
<4>[   23.431271]  #4:         (ptrval) (&(&timeline->lock)->rlock){....}, at: i915_request_submit+0x1a/0x40 [i915]
<4>[   23.431369]
                  stack backtrace:
<4>[   23.431377] CPU: 0 PID: 169 Comm: systemd-udevd Not tainted 4.17.0-rc4-CI-CI_DRM_4156+ #1
<4>[   23.431385] Hardware name: Dell Inc.                 OptiPlex GX280               /0G8310, BIOS A04 02/09/2005
<4>[   23.431394] Call Trace:
<4>[   23.431403]  dump_stack+0x67/0x9b
<4>[   23.431411]  __lock_acquire+0xc67/0x1b50
<4>[   23.431421]  ? ring_buffer_lock_reserve+0x154/0x3f0
<4>[   23.431429]  ? lock_acquire+0xa6/0x210
<4>[   23.431435]  lock_acquire+0xa6/0x210
<4>[   23.431530]  ? move_to_timeline+0x48/0x12c [i915]
<4>[   23.431540]  _raw_spin_lock+0x2a/0x40
<4>[   23.431634]  ? move_to_timeline+0x48/0x12c [i915]
<4>[   23.431730]  move_to_timeline+0x48/0x12c [i915]
<4>[   23.431826]  __i915_request_submit+0xfa/0x280 [i915]
<4>[   23.431923]  i915_request_submit+0x25/0x40 [i915]
<4>[   23.432024]  i9xx_submit_request+0x11/0x140 [i915]
<4>[   23.432120]  submit_notify+0x8d/0x124 [i915]
<4>[   23.432202]  __i915_sw_fence_complete+0x81/0x250 [i915]
<4>[   23.432300]  __i915_request_add+0x31c/0x7c0 [i915]
<4>[   23.432395]  i915_gem_init+0x621/0x630 [i915]
<4>[   23.432476]  i915_driver_load+0xbee/0x10b0 [i915]
<4>[   23.432485]  ? trace_hardirqs_on_caller+0xe0/0x1b0
<4>[   23.432566]  i915_pci_probe+0x29/0x90 [i915]
<4>[   23.432574]  pci_device_probe+0xa1/0x130
<4>[   23.432582]  driver_probe_device+0x306/0x480
<4>[   23.432589]  __driver_attach+0xb7/0xe0
<4>[   23.432596]  ? driver_probe_device+0x480/0x480
<4>[   23.432602]  ? driver_probe_device+0x480/0x480
<4>[   23.432609]  bus_for_each_dev+0x74/0xc0
<4>[   23.432616]  bus_add_driver+0x15f/0x250
<4>[   23.432623]  ? 0xffffffffa02d7000
<4>[   23.432629]  driver_register+0x52/0xc0
<4>[   23.432635]  ? 0xffffffffa02d7000
<4>[   23.432642]  do_one_initcall+0x58/0x370
<4>[   23.432653]  ? do_init_module+0x1d/0x1ea
<4>[   23.432660]  ? rcu_read_lock_sched_held+0x6f/0x80
<4>[   23.432667]  ? kmem_cache_alloc_trace+0x282/0x2e0
<4>[   23.432675]  do_init_module+0x56/0x1ea
<4>[   23.432682]  load_module+0x2435/0x2b20
<4>[   23.432694]  ? __se_sys_finit_module+0xd3/0xf0
<4>[   23.432701]  __se_sys_finit_module+0xd3/0xf0
<4>[   23.432710]  do_syscall_64+0x55/0x190
<4>[   23.432717]  entry_SYSCALL_64_after_hwframe+0x49/0xbe
<4>[   23.432724] RIP: 0033:0x7fa780782839
<4>[   23.432729] RSP: 002b:00007ffcea73e668 EFLAGS: 00000246 ORIG_RAX: 0000000000000139
<4>[   23.432738] RAX: ffffffffffffffda RBX: 0000561a472a4b30 RCX: 00007fa780782839
<4>[   23.432745] RDX: 0000000000000000 RSI: 00007fa7804610e5 RDI: 000000000000000e
<4>[   23.432752] RBP: 00007fa7804610e5 R08: 0000000000000000 R09: 00007ffcea73e780
<4>[   23.432758] R10: 000000000000000e R11: 0000000000000246 R12: 0000000000000000
<4>[   23.432765] R13: 0000561a47296450 R14: 0000000000020000 R15: 0000561a472a4b30

but did not report it as an issue as it only occurred during the first
module on boot. This is due to the removal of the distinct global
timeline, and its separate lock class. So instead mark up the expected
nesting. An alternative would be to define a separate lock class for the
engine, but since we only expect to have a single point of nesting, we
can avoid having multiple lock classes for the struct.

Fixes: a89d1f921c15 ("drm/i915: Split i915_gem_timeline into individual timelines")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Tested-by: Michel Thierry <michel.thierry@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180508153514.20251-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/i915_request.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c
index f336942..8928894 100644
--- a/drivers/gpu/drm/i915/i915_request.c
+++ b/drivers/gpu/drm/i915/i915_request.c
@@ -502,7 +502,7 @@ static void move_to_timeline(struct i915_request *request,
 	GEM_BUG_ON(request->timeline == &request->engine->timeline);
 	lockdep_assert_held(&request->engine->timeline.lock);
 
-	spin_lock(&request->timeline->lock);
+	spin_lock_nested(&request->timeline->lock, SINGLE_DEPTH_NESTING);
 	list_move_tail(&request->link, &timeline->requests);
 	spin_unlock(&request->timeline->lock);
 }
-- 
cgit v1.1


From 6f2db7dc901a1b89fbc50f7b38f0f7ee17205703 Mon Sep 17 00:00:00 2001
From: Peter Rosin <peda@axentia.se>
Date: Wed, 2 May 2018 09:40:25 +0200
Subject: drm/exynos: hdmi: avoid duplicating drm_bridge_attach

drm_bridge_attach takes care of these assignments, so there is no need
to open-code them a second time.

Signed-off-by: Peter Rosin <peda@axentia.se>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
---
 drivers/gpu/drm/exynos/exynos_hdmi.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/drivers/gpu/drm/exynos/exynos_hdmi.c b/drivers/gpu/drm/exynos/exynos_hdmi.c
index abd84cb..09c4bc0 100644
--- a/drivers/gpu/drm/exynos/exynos_hdmi.c
+++ b/drivers/gpu/drm/exynos/exynos_hdmi.c
@@ -954,8 +954,6 @@ static int hdmi_create_connector(struct drm_encoder *encoder)
 	drm_mode_connector_attach_encoder(connector, encoder);
 
 	if (hdata->bridge) {
-		encoder->bridge = hdata->bridge;
-		hdata->bridge->encoder = encoder;
 		ret = drm_bridge_attach(encoder, hdata->bridge, NULL);
 		if (ret)
 			DRM_ERROR("Failed to attach bridge\n");
-- 
cgit v1.1


From 8a8d9b2c38d9e050bec8d203ba2fb40c663c1b9c Mon Sep 17 00:00:00 2001
From: Souptick Joarder <jrdr.linux@gmail.com>
Date: Sat, 14 Apr 2018 21:34:29 +0530
Subject: gpu: drm: exynos: Change return type to vm_fault_t

Use new return type vm_fault_t for fault handler
in struct vm_operations_struct.

Signed-off-by: Souptick Joarder <jrdr.linux@gmail.com>
Reviewed-by: Matthew Wilcox <mawilcox@microsoft.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
---
 drivers/gpu/drm/exynos/exynos_drm_gem.c | 21 ++++-----------------
 drivers/gpu/drm/exynos/exynos_drm_gem.h |  3 ++-
 2 files changed, 6 insertions(+), 18 deletions(-)

diff --git a/drivers/gpu/drm/exynos/exynos_drm_gem.c b/drivers/gpu/drm/exynos/exynos_drm_gem.c
index 11cc01b..6e1494f 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_gem.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_gem.c
@@ -431,37 +431,24 @@ int exynos_drm_gem_dumb_create(struct drm_file *file_priv,
 	return 0;
 }
 
-int exynos_drm_gem_fault(struct vm_fault *vmf)
+vm_fault_t exynos_drm_gem_fault(struct vm_fault *vmf)
 {
 	struct vm_area_struct *vma = vmf->vma;
 	struct drm_gem_object *obj = vma->vm_private_data;
 	struct exynos_drm_gem *exynos_gem = to_exynos_gem(obj);
 	unsigned long pfn;
 	pgoff_t page_offset;
-	int ret;
 
 	page_offset = (vmf->address - vma->vm_start) >> PAGE_SHIFT;
 
 	if (page_offset >= (exynos_gem->size >> PAGE_SHIFT)) {
 		DRM_ERROR("invalid page offset\n");
-		ret = -EINVAL;
-		goto out;
+		return VM_FAULT_SIGBUS;
 	}
 
 	pfn = page_to_pfn(exynos_gem->pages[page_offset]);
-	ret = vm_insert_mixed(vma, vmf->address, __pfn_to_pfn_t(pfn, PFN_DEV));
-
-out:
-	switch (ret) {
-	case 0:
-	case -ERESTARTSYS:
-	case -EINTR:
-		return VM_FAULT_NOPAGE;
-	case -ENOMEM:
-		return VM_FAULT_OOM;
-	default:
-		return VM_FAULT_SIGBUS;
-	}
+	return vmf_insert_mixed(vma, vmf->address,
+			__pfn_to_pfn_t(pfn, PFN_DEV));
 }
 
 static int exynos_drm_gem_mmap_obj(struct drm_gem_object *obj,
diff --git a/drivers/gpu/drm/exynos/exynos_drm_gem.h b/drivers/gpu/drm/exynos/exynos_drm_gem.h
index 5a4c7de..9057d7f 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_gem.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_gem.h
@@ -13,6 +13,7 @@
 #define _EXYNOS_DRM_GEM_H_
 
 #include <drm/drm_gem.h>
+#include <linux/mm_types.h>
 
 #define to_exynos_gem(x)	container_of(x, struct exynos_drm_gem, base)
 
@@ -111,7 +112,7 @@ int exynos_drm_gem_dumb_create(struct drm_file *file_priv,
 			       struct drm_mode_create_dumb *args);
 
 /* page fault handler and mmap fault address(virtual) to physical memory. */
-int exynos_drm_gem_fault(struct vm_fault *vmf);
+vm_fault_t exynos_drm_gem_fault(struct vm_fault *vmf);
 
 /* set vm_flags and we can change the vm attribute to other one at here. */
 int exynos_drm_gem_mmap(struct file *filp, struct vm_area_struct *vma);
-- 
cgit v1.1


From fa50b7b4ba50f015acd0a6ca505582851e316d2a Mon Sep 17 00:00:00 2001
From: Tomasz Figa <tomasz.figa@gmail.com>
Date: Sat, 21 Apr 2018 19:26:10 +0200
Subject: drm/exynos: fimd: Add support for S5PV210 FIMD variant
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This patch adds support for FIMD variant found on S5PV210 SoC.
Except CLKSEL bit availability, it is identical to Exynos4210.

Tested-by: Paweł Chmiel <pawel.mikolaj.chmiel@gmail.com>
Signed-off-by: Tomasz Figa <tomasz.figa@gmail.com>
Signed-off-by: Paweł Chmiel <pawel.mikolaj.chmiel@gmail.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
---
 drivers/gpu/drm/exynos/exynos_drm_fimd.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/drivers/gpu/drm/exynos/exynos_drm_fimd.c b/drivers/gpu/drm/exynos/exynos_drm_fimd.c
index d42ae2b..01b1570 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_fimd.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_fimd.c
@@ -121,6 +121,12 @@ static struct fimd_driver_data s3c64xx_fimd_driver_data = {
 	.has_limited_fmt = 1,
 };
 
+static struct fimd_driver_data s5pv210_fimd_driver_data = {
+	.timing_base = 0x0,
+	.has_shadowcon = 1,
+	.has_clksel = 1,
+};
+
 static struct fimd_driver_data exynos3_fimd_driver_data = {
 	.timing_base = 0x20000,
 	.lcdblk_offset = 0x210,
@@ -193,6 +199,8 @@ struct fimd_context {
 static const struct of_device_id fimd_driver_dt_match[] = {
 	{ .compatible = "samsung,s3c6400-fimd",
 	  .data = &s3c64xx_fimd_driver_data },
+	{ .compatible = "samsung,s5pv210-fimd",
+	  .data = &s5pv210_fimd_driver_data },
 	{ .compatible = "samsung,exynos3250-fimd",
 	  .data = &exynos3_fimd_driver_data },
 	{ .compatible = "samsung,exynos4210-fimd",
-- 
cgit v1.1


From 5fae288d8ddd5b75d38d323cb4aa51ed2190ce17 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Pawe=C5=82=20Chmiel?= <pawel.mikolaj.chmiel@gmail.com>
Date: Sat, 21 Apr 2018 19:26:11 +0200
Subject: drm/exynos: Allow DRM_EXYNOS on s5pv210.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This patch brings back possibility to use drivers depending on
DRM_EXYNOS, on Samsung S5PV210/S5PC110 series based systems.

Fixes: dbbc925bb83a ("drm/exynos: depend on ARCH_EXYNOS for DRM_EXYNOS")
Signed-off-by: Paweł Chmiel <pawel.mikolaj.chmiel@gmail.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
---
 drivers/gpu/drm/exynos/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/exynos/Kconfig b/drivers/gpu/drm/exynos/Kconfig
index 735ce47..1548a78 100644
--- a/drivers/gpu/drm/exynos/Kconfig
+++ b/drivers/gpu/drm/exynos/Kconfig
@@ -1,6 +1,6 @@
 config DRM_EXYNOS
 	tristate "DRM Support for Samsung SoC EXYNOS Series"
-	depends on OF && DRM && (ARCH_S3C64XX || ARCH_EXYNOS || ARCH_MULTIPLATFORM)
+	depends on OF && DRM && (ARCH_S3C64XX || ARCH_S5PV210 || ARCH_EXYNOS || ARCH_MULTIPLATFORM)
 	select DRM_KMS_HELPER
 	select VIDEOMODE_HELPERS
 	select SND_SOC_HDMI_CODEC if SND_SOC
-- 
cgit v1.1


From 3148dedfe79e422f448a10250d3e2cdf8b7ee617 Mon Sep 17 00:00:00 2001
From: Heiner Kallweit <hkallweit1@gmail.com>
Date: Mon, 7 May 2018 21:11:21 +0200
Subject: r8169: fix powering up RTL8168h

Since commit a92a08499b1f "r8169: improve runtime pm in general and
suspend unused ports" interfaces w/o link are runtime-suspended after
10s. On systems where drivers take longer to load this can lead to the
situation that the interface is runtime-suspended already when it's
initially brought up.
This shouldn't be a problem because rtl_open() resumes MAC/PHY.
However with at least one chip version the interface doesn't properly
come up, as reported here:
https://bugzilla.kernel.org/show_bug.cgi?id=199549

The vendor driver uses a delay to give certain chip versions some
time to resume before starting the PHY configuration. So let's do
the same. I don't know which chip versions may be affected,
therefore apply this delay always.

This patch was reported to fix the issue for RTL8168h.
I was able to reproduce the issue on an Asus H310I-Plus which also
uses a RTL8168h. Also in my case the patch fixed the issue.

Reported-by: Slava Kardakov <ojab@ojab.ru>
Tested-by: Slava Kardakov <ojab@ojab.ru>
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/realtek/r8169.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/net/ethernet/realtek/r8169.c b/drivers/net/ethernet/realtek/r8169.c
index 604ae78..c7aac1f 100644
--- a/drivers/net/ethernet/realtek/r8169.c
+++ b/drivers/net/ethernet/realtek/r8169.c
@@ -4981,6 +4981,9 @@ static void rtl_pll_power_down(struct rtl8169_private *tp)
 static void rtl_pll_power_up(struct rtl8169_private *tp)
 {
 	rtl_generic_op(tp, tp->pll_power_ops.up);
+
+	/* give MAC/PHY some time to resume */
+	msleep(20);
 }
 
 static void rtl_init_pll_power_ops(struct rtl8169_private *tp)
-- 
cgit v1.1


From 1d7a99f5148fdcdb9d40367d6d0668a34df161d4 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Tue, 8 May 2018 22:10:56 +0100
Subject: drm/i915/selftests: Create mock_engine() under struct_mutex

Calling mock_engine() calls i915_timeline_init() and that requires
struct_mutex to be held as it adds itself to the global list of
timelines. This error was introduced by commit a89d1f921c15 ("drm/i915:
Split i915_gem_timeline into individual timelines") but the issue was
masked in CI by the earlier lockdep spam.

Fixes: a89d1f921c15 ("drm/i915: Split i915_gem_timeline into individual timelines")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Michel Thierry <michel.thierry@intel.com>
Reviewed-by: Michel Thierry <michel.thierry@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180508211056.17151-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/selftests/mock_gem_device.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/i915/selftests/mock_gem_device.c b/drivers/gpu/drm/i915/selftests/mock_gem_device.c
index 4b6622c..94baedf 100644
--- a/drivers/gpu/drm/i915/selftests/mock_gem_device.c
+++ b/drivers/gpu/drm/i915/selftests/mock_gem_device.c
@@ -229,18 +229,20 @@ struct drm_i915_private *mock_gem_device(void)
 	INIT_LIST_HEAD(&i915->gt.closed_vma);
 
 	mutex_lock(&i915->drm.struct_mutex);
+
 	mock_init_ggtt(i915);
-	mutex_unlock(&i915->drm.struct_mutex);
 
 	mkwrite_device_info(i915)->ring_mask = BIT(0);
 	i915->engine[RCS] = mock_engine(i915, "mock", RCS);
 	if (!i915->engine[RCS])
-		goto err_priorities;
+		goto err_unlock;
 
 	i915->kernel_context = mock_context(i915, NULL);
 	if (!i915->kernel_context)
 		goto err_engine;
 
+	mutex_unlock(&i915->drm.struct_mutex);
+
 	WARN_ON(i915_gemfs_init(i915));
 
 	return i915;
@@ -248,7 +250,8 @@ struct drm_i915_private *mock_gem_device(void)
 err_engine:
 	for_each_engine(engine, i915, id)
 		mock_engine_free(engine);
-err_priorities:
+err_unlock:
+	mutex_unlock(&i915->drm.struct_mutex);
 	kmem_cache_destroy(i915->priorities);
 err_dependencies:
 	kmem_cache_destroy(i915->dependencies);
-- 
cgit v1.1


From 9a0e9802217291e54c4dd1fc5462f189a4be14ec Mon Sep 17 00:00:00 2001
From: Boris Brezillon <boris.brezillon@bootlin.com>
Date: Mon, 7 May 2018 14:13:03 +0200
Subject: drm/vc4: Fix scaling of uni-planar formats

When using uni-planar formats (like RGB), the scaling parameters are
stored in plane 0, not plane 1.

Fixes: fc04023fafec ("drm/vc4: Add support for YUV planes.")
Cc: stable@vger.kernel.org
Signed-off-by: Boris Brezillon <boris.brezillon@bootlin.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Link: https://patchwork.freedesktop.org/patch/msgid/20180507121303.5610-1-boris.brezillon@bootlin.com
---
 drivers/gpu/drm/vc4/vc4_plane.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/vc4/vc4_plane.c b/drivers/gpu/drm/vc4/vc4_plane.c
index ce39390..13dcaad 100644
--- a/drivers/gpu/drm/vc4/vc4_plane.c
+++ b/drivers/gpu/drm/vc4/vc4_plane.c
@@ -503,7 +503,7 @@ static int vc4_plane_mode_set(struct drm_plane *plane,
 	 * the scl fields here.
 	 */
 	if (num_planes == 1) {
-		scl0 = vc4_get_scl_field(state, 1);
+		scl0 = vc4_get_scl_field(state, 0);
 		scl1 = scl0;
 	} else {
 		scl0 = vc4_get_scl_field(state, 1);
-- 
cgit v1.1


From b9777c6f86ac8c21f82211ab982ca48302042ede Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Wed, 9 May 2018 07:59:26 +0100
Subject: drm/i915/selftests: Only switch to kernel context when locked

In igt_flush_test() we try to switch back to the kernel context, but we
are only able to do so when we are called with struct_mutex held.

More of my CI fallout from lockdep being temporarily suppressed :(

Fixes: 4cdf65ce8cc2 ("drm/i915/selftests: Return to kernel context after each test")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180509065926.19207-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/selftests/igt_flush_test.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/selftests/igt_flush_test.c b/drivers/gpu/drm/i915/selftests/igt_flush_test.c
index 7f35bdd..0d06f55 100644
--- a/drivers/gpu/drm/i915/selftests/igt_flush_test.c
+++ b/drivers/gpu/drm/i915/selftests/igt_flush_test.c
@@ -57,7 +57,8 @@ int igt_flush_test(struct drm_i915_private *i915, unsigned int flags)
 
 	cond_resched();
 
-	if (i915_gem_switch_to_kernel_context(i915)) {
+	if (flags & I915_WAIT_LOCKED &&
+	    i915_gem_switch_to_kernel_context(i915)) {
 		pr_err("Failed to switch back to kernel context; declaring wedged\n");
 		i915_gem_set_wedged(i915);
 	}
-- 
cgit v1.1


From 4413c474b18f116f96157599b7cc418a6471ed31 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Tue, 8 May 2018 22:03:17 +0100
Subject: drm/i915/execlists: Make submission tasklet hardirq safe

Prepare to allow the execlists submission to be run from underneath a
hardirq timer context (and not just the current softirq context) as is
required for fast preemption resets and context switches.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180508210318.10274-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/intel_lrc.c | 42 +++++++++++++++++++++++++++-------------
 1 file changed, 29 insertions(+), 13 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index da08225..d3c00f6 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -356,10 +356,13 @@ execlists_unwind_incomplete_requests(struct intel_engine_execlists *execlists)
 {
 	struct intel_engine_cs *engine =
 		container_of(execlists, typeof(*engine), execlists);
+	unsigned long flags;
+
+	spin_lock_irqsave(&engine->timeline.lock, flags);
 
-	spin_lock_irq(&engine->timeline.lock);
 	__unwind_incomplete_requests(engine);
-	spin_unlock_irq(&engine->timeline.lock);
+
+	spin_unlock_irqrestore(&engine->timeline.lock, flags);
 }
 
 static inline void
@@ -553,7 +556,7 @@ static void inject_preempt_context(struct intel_engine_cs *engine)
 	execlists_set_active(&engine->execlists, EXECLISTS_ACTIVE_PREEMPT);
 }
 
-static void execlists_dequeue(struct intel_engine_cs *engine)
+static bool __execlists_dequeue(struct intel_engine_cs *engine)
 {
 	struct intel_engine_execlists * const execlists = &engine->execlists;
 	struct execlist_port *port = execlists->port;
@@ -563,6 +566,8 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 	struct rb_node *rb;
 	bool submit = false;
 
+	lockdep_assert_held(&engine->timeline.lock);
+
 	/* Hardware submission is through 2 ports. Conceptually each port
 	 * has a (RING_START, RING_HEAD, RING_TAIL) tuple. RING_START is
 	 * static for a context, and unique to each, so we only execute
@@ -584,7 +589,6 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 	 * and context switches) submission.
 	 */
 
-	spin_lock_irq(&engine->timeline.lock);
 	rb = execlists->first;
 	GEM_BUG_ON(rb_first(&execlists->queue) != rb);
 
@@ -599,7 +603,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 						EXECLISTS_ACTIVE_USER));
 		GEM_BUG_ON(!port_count(&port[0]));
 		if (port_count(&port[0]) > 1)
-			goto unlock;
+			return false;
 
 		/*
 		 * If we write to ELSP a second time before the HW has had
@@ -609,11 +613,11 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 		 * the HW to indicate that it has had a chance to respond.
 		 */
 		if (!execlists_is_active(execlists, EXECLISTS_ACTIVE_HWACK))
-			goto unlock;
+			return false;
 
 		if (need_preempt(engine, last, execlists->queue_priority)) {
 			inject_preempt_context(engine);
-			goto unlock;
+			return false;
 		}
 
 		/*
@@ -638,7 +642,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 		 * priorities of the ports haven't been switch.
 		 */
 		if (port_count(&port[1]))
-			goto unlock;
+			return false;
 
 		/*
 		 * WaIdleLiteRestore:bdw,skl
@@ -743,13 +747,25 @@ done:
 	/* We must always keep the beast fed if we have work piled up */
 	GEM_BUG_ON(execlists->first && !port_isset(execlists->port));
 
-unlock:
-	spin_unlock_irq(&engine->timeline.lock);
-
-	if (submit) {
+	/* Re-evaluate the executing context setup after each preemptive kick */
+	if (last)
 		execlists_user_begin(execlists, execlists->port);
+
+	return submit;
+}
+
+static void execlists_dequeue(struct intel_engine_cs *engine)
+{
+	struct intel_engine_execlists * const execlists = &engine->execlists;
+	unsigned long flags;
+	bool submit;
+
+	spin_lock_irqsave(&engine->timeline.lock, flags);
+	submit = __execlists_dequeue(engine);
+	spin_unlock_irqrestore(&engine->timeline.lock, flags);
+
+	if (submit)
 		execlists_submit_ports(engine);
-	}
 
 	GEM_BUG_ON(port_isset(execlists->port) &&
 		   !execlists_is_active(execlists, EXECLISTS_ACTIVE_USER));
-- 
cgit v1.1


From 6486d84b1cca1b4374286ba2685161824f0bfa7d Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Tue, 8 May 2018 22:03:18 +0100
Subject: drm/i915/guc: Make submission tasklet hardirq safe

Prepare to allow the GuC submission to be run from underneath a
hardirq timer context (and not just the current softirq context) as is
required for fast preemption resets and context switches.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180508210318.10274-2-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/intel_guc_submission.c | 34 +++++++++++++++++++++--------
 1 file changed, 25 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_guc_submission.c b/drivers/gpu/drm/i915/intel_guc_submission.c
index 62828e3..2feb650 100644
--- a/drivers/gpu/drm/i915/intel_guc_submission.c
+++ b/drivers/gpu/drm/i915/intel_guc_submission.c
@@ -669,7 +669,7 @@ static inline int port_prio(const struct execlist_port *port)
 	return rq_prio(port_request(port));
 }
 
-static void guc_dequeue(struct intel_engine_cs *engine)
+static bool __guc_dequeue(struct intel_engine_cs *engine)
 {
 	struct intel_engine_execlists * const execlists = &engine->execlists;
 	struct execlist_port *port = execlists->port;
@@ -679,7 +679,8 @@ static void guc_dequeue(struct intel_engine_cs *engine)
 	bool submit = false;
 	struct rb_node *rb;
 
-	spin_lock_irq(&engine->timeline.lock);
+	lockdep_assert_held(&engine->timeline.lock);
+
 	rb = execlists->first;
 	GEM_BUG_ON(rb_first(&execlists->queue) != rb);
 
@@ -694,13 +695,13 @@ static void guc_dequeue(struct intel_engine_cs *engine)
 						     EXECLISTS_ACTIVE_PREEMPT);
 				queue_work(engine->i915->guc.preempt_wq,
 					   &preempt_work->work);
-				goto unlock;
+				return false;
 			}
 		}
 
 		port++;
 		if (port_isset(port))
-			goto unlock;
+			return false;
 	}
 	GEM_BUG_ON(port_isset(port));
 
@@ -738,19 +739,34 @@ static void guc_dequeue(struct intel_engine_cs *engine)
 done:
 	execlists->queue_priority = rb ? to_priolist(rb)->priority : INT_MIN;
 	execlists->first = rb;
-	if (submit) {
+	if (submit)
 		port_assign(port, last);
+	if (last)
 		execlists_user_begin(execlists, execlists->port);
-		guc_submit(engine);
-	}
 
 	/* We must always keep the beast fed if we have work piled up */
 	GEM_BUG_ON(port_isset(execlists->port) &&
 		   !execlists_is_active(execlists, EXECLISTS_ACTIVE_USER));
 	GEM_BUG_ON(execlists->first && !port_isset(execlists->port));
 
-unlock:
-	spin_unlock_irq(&engine->timeline.lock);
+	return submit;
+}
+
+static void guc_dequeue(struct intel_engine_cs *engine)
+{
+	unsigned long flags;
+	bool submit;
+
+	local_irq_save(flags);
+
+	spin_lock(&engine->timeline.lock);
+	submit = __guc_dequeue(engine);
+	spin_unlock(&engine->timeline.lock);
+
+	if (submit)
+		guc_submit(engine);
+
+	local_irq_restore(flags);
 }
 
 static void guc_submission_tasklet(unsigned long data)
-- 
cgit v1.1


From c72a0ded8d4ac9d99c04200035bbfcc82e15842c Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jonathan=20Neusch=C3=A4fer?= <j.neuschaefer@gmx.net>
Date: Wed, 25 Apr 2018 12:07:03 +0200
Subject: PM: docs: sleep-states: Fix a typo ("includig")
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Fix a typo in admin-guide/pm/sleep-states.rst.

Signed-off-by: Jonathan Neuschäfer <j.neuschaefer@gmx.net>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
---
 Documentation/admin-guide/pm/sleep-states.rst | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/Documentation/admin-guide/pm/sleep-states.rst b/Documentation/admin-guide/pm/sleep-states.rst
index 1e5c0f0..dbf5acd 100644
--- a/Documentation/admin-guide/pm/sleep-states.rst
+++ b/Documentation/admin-guide/pm/sleep-states.rst
@@ -15,7 +15,7 @@ Sleep States That Can Be Supported
 ==================================
 
 Depending on its configuration and the capabilities of the platform it runs on,
-the Linux kernel can support up to four system sleep states, includig
+the Linux kernel can support up to four system sleep states, including
 hibernation and up to three variants of system suspend.  The sleep states that
 can be supported by the kernel are listed below.
 
-- 
cgit v1.1


From 13610c93488b3c290f393c76744b308445921094 Mon Sep 17 00:00:00 2001
From: Juri Lelli <juri.lelli@redhat.com>
Date: Tue, 8 May 2018 17:12:09 +0200
Subject: PM: docs: intel_pstate: fix Active Mode w/o HWP paragraph

P-state selection algorithm (powersave or performance) is selected by
echoing the desired choice to scaling_governor sysfs attribute and not
to scaling_cur_freq (as currently stated).

Fix it.

Signed-off-by: Juri Lelli <juri.lelli@redhat.com>
Reviewed-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
---
 Documentation/admin-guide/pm/intel_pstate.rst | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/Documentation/admin-guide/pm/intel_pstate.rst b/Documentation/admin-guide/pm/intel_pstate.rst
index d2b6fda..ab2fe0e 100644
--- a/Documentation/admin-guide/pm/intel_pstate.rst
+++ b/Documentation/admin-guide/pm/intel_pstate.rst
@@ -145,7 +145,7 @@ feature enabled.]
 
 In this mode ``intel_pstate`` registers utilization update callbacks with the
 CPU scheduler in order to run a P-state selection algorithm, either
-``powersave`` or ``performance``, depending on the ``scaling_cur_freq`` policy
+``powersave`` or ``performance``, depending on the ``scaling_governor`` policy
 setting in ``sysfs``.  The current CPU frequency information to be made
 available from the ``scaling_cur_freq`` policy attribute in ``sysfs`` is
 periodically updated by those utilization update callbacks too.
-- 
cgit v1.1


From a744490f12707d9f0b205272b29adf5bdb3ba193 Mon Sep 17 00:00:00 2001
From: Juri Lelli <juri.lelli@redhat.com>
Date: Wed, 9 May 2018 10:40:51 +0200
Subject: cpufreq: schedutil: remove stale comment

After commit 794a56ebd9a57 (sched/cpufreq: Change the worker kthread to
SCHED_DEADLINE) schedutil kthreads are "ignored" for a clock frequency
selection point of view, so the potential corner case for RT tasks is not
possible at all now.

Remove the stale comment mentioning it.

Signed-off-by: Juri Lelli <juri.lelli@redhat.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
---
 kernel/sched/cpufreq_schedutil.c | 13 -------------
 1 file changed, 13 deletions(-)

diff --git a/kernel/sched/cpufreq_schedutil.c b/kernel/sched/cpufreq_schedutil.c
index d2c6083..23ef190 100644
--- a/kernel/sched/cpufreq_schedutil.c
+++ b/kernel/sched/cpufreq_schedutil.c
@@ -396,19 +396,6 @@ static void sugov_irq_work(struct irq_work *irq_work)
 
 	sg_policy = container_of(irq_work, struct sugov_policy, irq_work);
 
-	/*
-	 * For RT tasks, the schedutil governor shoots the frequency to maximum.
-	 * Special care must be taken to ensure that this kthread doesn't result
-	 * in the same behavior.
-	 *
-	 * This is (mostly) guaranteed by the work_in_progress flag. The flag is
-	 * updated only at the end of the sugov_work() function and before that
-	 * the schedutil governor rejects all other frequency scaling requests.
-	 *
-	 * There is a very rare case though, where the RT thread yields right
-	 * after the work_in_progress flag is cleared. The effects of that are
-	 * neglected for now.
-	 */
 	kthread_queue_work(&sg_policy->worker, &sg_policy->work);
 }
 
-- 
cgit v1.1


From 97739501f207efe33145b918817f305b822987f8 Mon Sep 17 00:00:00 2001
From: "Rafael J. Wysocki" <rafael.j.wysocki@intel.com>
Date: Wed, 9 May 2018 11:44:56 +0200
Subject: cpufreq: schedutil: Avoid using invalid next_freq

If the next_freq field of struct sugov_policy is set to UINT_MAX,
it shouldn't be used for updating the CPU frequency (this is a
special "invalid" value), but after commit b7eaf1aab9f8 (cpufreq:
schedutil: Avoid reducing frequency of busy CPUs prematurely) it
may be passed as the new frequency to sugov_update_commit() in
sugov_update_single().

Fix that by adding an extra check for the special UINT_MAX value
of next_freq to sugov_update_single().

Fixes: b7eaf1aab9f8 (cpufreq: schedutil: Avoid reducing frequency of busy CPUs prematurely)
Reported-by: Viresh Kumar <viresh.kumar@linaro.org>
Cc: 4.12+ <stable@vger.kernel.org> # 4.12+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
---
 kernel/sched/cpufreq_schedutil.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/kernel/sched/cpufreq_schedutil.c b/kernel/sched/cpufreq_schedutil.c
index 23ef190..e13df95 100644
--- a/kernel/sched/cpufreq_schedutil.c
+++ b/kernel/sched/cpufreq_schedutil.c
@@ -305,7 +305,8 @@ static void sugov_update_single(struct update_util_data *hook, u64 time,
 	 * Do not reduce the frequency if the CPU has not been idle
 	 * recently, as the reduction is likely to be premature then.
 	 */
-	if (busy && next_f < sg_policy->next_freq) {
+	if (busy && next_f < sg_policy->next_freq &&
+	    sg_policy->next_freq != UINT_MAX) {
 		next_f = sg_policy->next_freq;
 
 		/* Reset cached freq as next_freq has changed */
-- 
cgit v1.1


From 8feaec33b9868582654cd3d5355225dcb79aeca6 Mon Sep 17 00:00:00 2001
From: Kai Heng Feng <kai.heng.feng@canonical.com>
Date: Mon, 7 May 2018 14:11:20 +0800
Subject: PCI / PM: Always check PME wakeup capability for runtime wakeup
 support

USB controller ASM1042 stops working after commit de3ef1eb1cd0 (PM /
core: Drop run_wake flag from struct dev_pm_info).

The device in question is not power managed by platform firmware,
furthermore, it only supports PME# from D3cold:
Capabilities: [78] Power Management version 3
       Flags: PMEClk- DSI- D1- D2- AuxCurrent=55mA PME(D0-,D1-,D2-,D3hot-,D3cold+)
       Status: D0 NoSoftRst+ PME-Enable- DSel=0 DScale=0 PME-

Before commit de3ef1eb1cd0, the device never gets runtime suspended.
After that commit, the device gets runtime suspended to D3hot, which can
not generate any PME#.

usb_hcd_pci_probe() unconditionally calls device_wakeup_enable(), hence
device_can_wakeup() in pci_dev_run_wake() always returns true.

So pci_dev_run_wake() needs to check PME wakeup capability as its first
condition.

In addition, change wakeup flag passed to pci_target_state() from false
to true, because we want to find the deepest state different from D3cold
that the device can still generate PME#. In this case, it's D0 for the
device in question.

Fixes: de3ef1eb1cd0 (PM / core: Drop run_wake flag from struct dev_pm_info)
Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
Cc: 4.13+ <stable@vger.kernel.org> # 4.13+
Acked-by: Bjorn Helgaas <bhelgaas@google.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
---
 drivers/pci/pci.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c
index a04197c..c2616ca 100644
--- a/drivers/pci/pci.c
+++ b/drivers/pci/pci.c
@@ -2138,16 +2138,16 @@ bool pci_dev_run_wake(struct pci_dev *dev)
 {
 	struct pci_bus *bus = dev->bus;
 
-	if (device_can_wakeup(&dev->dev))
-		return true;
-
 	if (!dev->pme_support)
 		return false;
 
 	/* PME-capable in principle, but not from the target power state */
-	if (!pci_pme_capable(dev, pci_target_state(dev, false)))
+	if (!pci_pme_capable(dev, pci_target_state(dev, true)))
 		return false;
 
+	if (device_can_wakeup(&dev->dev))
+		return true;
+
 	while (bus->parent) {
 		struct pci_dev *bridge = bus->self;
 
-- 
cgit v1.1


From 070b9637dd8fa85c3ba7ecc60fe57fa4da9c2d1d Mon Sep 17 00:00:00 2001
From: Hans de Goede <hdegoede@redhat.com>
Date: Thu, 3 May 2018 11:32:33 +0200
Subject: HID: i2c-hid: Add RESEND_REPORT_DESCR quirk for Toshiba Click Mini
 L9W-B

The 0457:10fb touchscreen found on the Toshiba Click Mini L9W-B needs
to have a report-decriptors command send to it on resume in order for
the touchscreen to start generating events again on resume.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
---
 drivers/hid/hid-ids.h         | 1 +
 drivers/hid/i2c-hid/i2c-hid.c | 2 ++
 2 files changed, 3 insertions(+)

diff --git a/drivers/hid/hid-ids.h b/drivers/hid/hid-ids.h
index ec73aa4..46f5ecd 100644
--- a/drivers/hid/hid-ids.h
+++ b/drivers/hid/hid-ids.h
@@ -972,6 +972,7 @@
 #define USB_DEVICE_ID_SIS817_TOUCH	0x0817
 #define USB_DEVICE_ID_SIS_TS		0x1013
 #define USB_DEVICE_ID_SIS1030_TOUCH	0x1030
+#define USB_DEVICE_ID_SIS10FB_TOUCH	0x10fb
 
 #define USB_VENDOR_ID_SKYCABLE			0x1223
 #define	USB_DEVICE_ID_SKYCABLE_WIRELESS_PRESENTER	0x3F07
diff --git a/drivers/hid/i2c-hid/i2c-hid.c b/drivers/hid/i2c-hid/i2c-hid.c
index 9633286..cc33622 100644
--- a/drivers/hid/i2c-hid/i2c-hid.c
+++ b/drivers/hid/i2c-hid/i2c-hid.c
@@ -174,6 +174,8 @@ static const struct i2c_hid_quirks {
 		I2C_HID_QUIRK_NO_IRQ_AFTER_RESET },
 	{ I2C_VENDOR_ID_RAYD, I2C_PRODUCT_ID_RAYD_3118,
 		I2C_HID_QUIRK_RESEND_REPORT_DESCR },
+	{ USB_VENDOR_ID_SIS_TOUCH, USB_DEVICE_ID_SIS10FB_TOUCH,
+		I2C_HID_QUIRK_RESEND_REPORT_DESCR },
 	{ 0, 0 }
 };
 
-- 
cgit v1.1


From 2bdd045e3a30f7ddda1604f835df9c8d14d6d048 Mon Sep 17 00:00:00 2001
From: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Date: Tue, 8 May 2018 17:35:24 -0700
Subject: drm/i915/psr: Check if VBT says PSR can be enabled.

Driver features data block has a boolean flag for PSR, use this to decide
whether PSR should be enabled on a platform. The module parameter can
still be used to override this.

Note: The feature currently remains disabled by default for all platforms
irrespective of what VBT says.

Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180509003524.3199-1-dhinakaran.pandiyan@intel.com
---
 drivers/gpu/drm/i915/i915_drv.h   | 1 +
 drivers/gpu/drm/i915/intel_bios.c | 1 +
 drivers/gpu/drm/i915/intel_psr.c  | 7 +++++--
 3 files changed, 7 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 04e2780..24c5e47 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -1073,6 +1073,7 @@ struct intel_vbt_data {
 	} edp;
 
 	struct {
+		bool enable;
 		bool full_link;
 		bool require_aux_wakeup;
 		int idle_frames;
diff --git a/drivers/gpu/drm/i915/intel_bios.c b/drivers/gpu/drm/i915/intel_bios.c
index 702d3fa..54270bd 100644
--- a/drivers/gpu/drm/i915/intel_bios.c
+++ b/drivers/gpu/drm/i915/intel_bios.c
@@ -530,6 +530,7 @@ parse_driver_features(struct drm_i915_private *dev_priv,
 	 */
 	if (!driver->drrs_enabled)
 		dev_priv->vbt.drrs_type = DRRS_NOT_SUPPORTED;
+	dev_priv->vbt.psr.enable = driver->psr_enabled;
 }
 
 static void
diff --git a/drivers/gpu/drm/i915/intel_psr.c b/drivers/gpu/drm/i915/intel_psr.c
index 6233a32..db27f2f 100644
--- a/drivers/gpu/drm/i915/intel_psr.c
+++ b/drivers/gpu/drm/i915/intel_psr.c
@@ -1173,9 +1173,12 @@ void intel_psr_init(struct drm_i915_private *dev_priv)
 	if (!dev_priv->psr.sink_support)
 		return;
 
-	/* Per platform default: all disabled. */
-	if (i915_modparams.enable_psr == -1)
+	if (i915_modparams.enable_psr == -1) {
+		i915_modparams.enable_psr = dev_priv->vbt.psr.enable;
+
+		/* Per platform default: all disabled. */
 		i915_modparams.enable_psr = 0;
+	}
 
 	/* Set link_standby x link_off defaults */
 	if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv))
-- 
cgit v1.1


From 1b06bd8dd95f7a19ab33fdf0f477c94950822ab3 Mon Sep 17 00:00:00 2001
From: David Gilhooley <dgilhooley@nvidia.com>
Date: Tue, 8 May 2018 15:49:42 -0700
Subject: arm64: Add MIDR encoding for NVIDIA CPUs

This patch adds the MIDR encodings for NVIDIA as well as
the Denver and Carmel CPUs used in Tegra SoCs.

Signed-off-by: David Gilhooley <dgilhooley@nvidia.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
---
 arch/arm64/include/asm/cputype.h | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/arch/arm64/include/asm/cputype.h b/arch/arm64/include/asm/cputype.h
index 30014a9..ea690b3 100644
--- a/arch/arm64/include/asm/cputype.h
+++ b/arch/arm64/include/asm/cputype.h
@@ -75,6 +75,7 @@
 #define ARM_CPU_IMP_CAVIUM		0x43
 #define ARM_CPU_IMP_BRCM		0x42
 #define ARM_CPU_IMP_QCOM		0x51
+#define ARM_CPU_IMP_NVIDIA		0x4E
 
 #define ARM_CPU_PART_AEM_V8		0xD0F
 #define ARM_CPU_PART_FOUNDATION		0xD00
@@ -99,6 +100,9 @@
 #define QCOM_CPU_PART_FALKOR		0xC00
 #define QCOM_CPU_PART_KRYO		0x200
 
+#define NVIDIA_CPU_PART_DENVER		0x003
+#define NVIDIA_CPU_PART_CARMEL		0x004
+
 #define MIDR_CORTEX_A53 MIDR_CPU_MODEL(ARM_CPU_IMP_ARM, ARM_CPU_PART_CORTEX_A53)
 #define MIDR_CORTEX_A57 MIDR_CPU_MODEL(ARM_CPU_IMP_ARM, ARM_CPU_PART_CORTEX_A57)
 #define MIDR_CORTEX_A72 MIDR_CPU_MODEL(ARM_CPU_IMP_ARM, ARM_CPU_PART_CORTEX_A72)
@@ -114,6 +118,8 @@
 #define MIDR_QCOM_FALKOR_V1 MIDR_CPU_MODEL(ARM_CPU_IMP_QCOM, QCOM_CPU_PART_FALKOR_V1)
 #define MIDR_QCOM_FALKOR MIDR_CPU_MODEL(ARM_CPU_IMP_QCOM, QCOM_CPU_PART_FALKOR)
 #define MIDR_QCOM_KRYO MIDR_CPU_MODEL(ARM_CPU_IMP_QCOM, QCOM_CPU_PART_KRYO)
+#define MIDR_NVIDIA_DENVER MIDR_CPU_MODEL(ARM_CPU_IMP_NVIDIA, NVIDIA_CPU_PART_DENVER)
+#define MIDR_NVIDIA_CARMEL MIDR_CPU_MODEL(ARM_CPU_IMP_NVIDIA, NVIDIA_CPU_PART_CARMEL)
 
 #ifndef __ASSEMBLY__
 
-- 
cgit v1.1


From 0583a4ef05987f7e0f3a7bdd3365e5dc36ca306d Mon Sep 17 00:00:00 2001
From: David Gilhooley <dgilhooley@nvidia.com>
Date: Tue, 8 May 2018 15:49:43 -0700
Subject: arm64: capabilities: Add NVIDIA Denver CPU to bp_harden list

The NVIDIA Denver CPU also needs a PSCI call to harden the branch
predictor.

Signed-off-by: David Gilhooley <dgilhooley@nvidia.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
---
 arch/arm64/kernel/cpu_errata.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/arm64/kernel/cpu_errata.c b/arch/arm64/kernel/cpu_errata.c
index a900bef..e4a1182 100644
--- a/arch/arm64/kernel/cpu_errata.c
+++ b/arch/arm64/kernel/cpu_errata.c
@@ -316,6 +316,7 @@ static const struct midr_range arm64_bp_harden_smccc_cpus[] = {
 	MIDR_ALL_VERSIONS(MIDR_CAVIUM_THUNDERX2),
 	MIDR_ALL_VERSIONS(MIDR_QCOM_FALKOR_V1),
 	MIDR_ALL_VERSIONS(MIDR_QCOM_FALKOR),
+	MIDR_ALL_VERSIONS(MIDR_NVIDIA_DENVER),
 	{},
 };
 
-- 
cgit v1.1


From 2796d303e3c5ec213c578ed3a66872205c126eb8 Mon Sep 17 00:00:00 2001
From: Long Li <longli@microsoft.com>
Date: Wed, 25 Apr 2018 11:30:04 -0700
Subject: cifs: Allocate validate negotiation request through kmalloc

The data buffer allocated on the stack can't be DMA'ed, ib_dma_map_page will
return an invalid DMA address for a buffer on stack. Even worse, this
incorrect address can't be detected by ib_dma_mapping_error. Sending data
from this address to hardware will not fail, but the remote peer will get
junk data.

Fix this by allocating the request on the heap in smb3_validate_negotiate.

Changes in v2:
Removed duplicated code on freeing buffers on function exit.
(Thanks to Parav Pandit <parav@mellanox.com>)
Fixed typo in the patch title.

Changes in v3:
Added "Fixes" to the patch.
Changed several sizeof() to use *pointer in place of struct.

Changes in v4:
Added detailed comments on the failure through RDMA.
Allocate request buffer using GPF_NOFS.
Fixed possible memory leak.

Changes in v5:
Removed variable ret for checking return value.
Changed to use pneg_inbuf->Dialects[0] to calculate unused space in pneg_inbuf.

Fixes: ff1c038addc4 ("Check SMB3 dialects against downgrade attacks")
Signed-off-by: Long Li <longli@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Reviewed-by: Ronnie Sahlberg <lsahlber@redhat.com>
Reviewed-by: Tom Talpey <ttalpey@microsoft.com>
---
 fs/cifs/smb2pdu.c | 68 +++++++++++++++++++++++++++++++------------------------
 1 file changed, 38 insertions(+), 30 deletions(-)

diff --git a/fs/cifs/smb2pdu.c b/fs/cifs/smb2pdu.c
index 60db51b..260e9c4 100644
--- a/fs/cifs/smb2pdu.c
+++ b/fs/cifs/smb2pdu.c
@@ -730,8 +730,8 @@ neg_exit:
 
 int smb3_validate_negotiate(const unsigned int xid, struct cifs_tcon *tcon)
 {
-	int rc = 0;
-	struct validate_negotiate_info_req vneg_inbuf;
+	int rc;
+	struct validate_negotiate_info_req *pneg_inbuf;
 	struct validate_negotiate_info_rsp *pneg_rsp = NULL;
 	u32 rsplen;
 	u32 inbuflen; /* max of 4 dialects */
@@ -765,63 +765,69 @@ int smb3_validate_negotiate(const unsigned int xid, struct cifs_tcon *tcon)
 	if (tcon->ses->session_flags & SMB2_SESSION_FLAG_IS_NULL)
 		cifs_dbg(VFS, "Unexpected null user (anonymous) auth flag sent by server\n");
 
-	vneg_inbuf.Capabilities =
+	pneg_inbuf = kmalloc(sizeof(*pneg_inbuf), GFP_NOFS);
+	if (!pneg_inbuf)
+		return -ENOMEM;
+
+	pneg_inbuf->Capabilities =
 			cpu_to_le32(tcon->ses->server->vals->req_capabilities);
-	memcpy(vneg_inbuf.Guid, tcon->ses->server->client_guid,
+	memcpy(pneg_inbuf->Guid, tcon->ses->server->client_guid,
 					SMB2_CLIENT_GUID_SIZE);
 
 	if (tcon->ses->sign)
-		vneg_inbuf.SecurityMode =
+		pneg_inbuf->SecurityMode =
 			cpu_to_le16(SMB2_NEGOTIATE_SIGNING_REQUIRED);
 	else if (global_secflags & CIFSSEC_MAY_SIGN)
-		vneg_inbuf.SecurityMode =
+		pneg_inbuf->SecurityMode =
 			cpu_to_le16(SMB2_NEGOTIATE_SIGNING_ENABLED);
 	else
-		vneg_inbuf.SecurityMode = 0;
+		pneg_inbuf->SecurityMode = 0;
 
 
 	if (strcmp(tcon->ses->server->vals->version_string,
 		SMB3ANY_VERSION_STRING) == 0) {
-		vneg_inbuf.Dialects[0] = cpu_to_le16(SMB30_PROT_ID);
-		vneg_inbuf.Dialects[1] = cpu_to_le16(SMB302_PROT_ID);
-		vneg_inbuf.DialectCount = cpu_to_le16(2);
+		pneg_inbuf->Dialects[0] = cpu_to_le16(SMB30_PROT_ID);
+		pneg_inbuf->Dialects[1] = cpu_to_le16(SMB302_PROT_ID);
+		pneg_inbuf->DialectCount = cpu_to_le16(2);
 		/* structure is big enough for 3 dialects, sending only 2 */
-		inbuflen = sizeof(struct validate_negotiate_info_req) - 2;
+		inbuflen = sizeof(*pneg_inbuf) -
+				sizeof(pneg_inbuf->Dialects[0]);
 	} else if (strcmp(tcon->ses->server->vals->version_string,
 		SMBDEFAULT_VERSION_STRING) == 0) {
-		vneg_inbuf.Dialects[0] = cpu_to_le16(SMB21_PROT_ID);
-		vneg_inbuf.Dialects[1] = cpu_to_le16(SMB30_PROT_ID);
-		vneg_inbuf.Dialects[2] = cpu_to_le16(SMB302_PROT_ID);
-		vneg_inbuf.DialectCount = cpu_to_le16(3);
+		pneg_inbuf->Dialects[0] = cpu_to_le16(SMB21_PROT_ID);
+		pneg_inbuf->Dialects[1] = cpu_to_le16(SMB30_PROT_ID);
+		pneg_inbuf->Dialects[2] = cpu_to_le16(SMB302_PROT_ID);
+		pneg_inbuf->DialectCount = cpu_to_le16(3);
 		/* structure is big enough for 3 dialects */
-		inbuflen = sizeof(struct validate_negotiate_info_req);
+		inbuflen = sizeof(*pneg_inbuf);
 	} else {
 		/* otherwise specific dialect was requested */
-		vneg_inbuf.Dialects[0] =
+		pneg_inbuf->Dialects[0] =
 			cpu_to_le16(tcon->ses->server->vals->protocol_id);
-		vneg_inbuf.DialectCount = cpu_to_le16(1);
+		pneg_inbuf->DialectCount = cpu_to_le16(1);
 		/* structure is big enough for 3 dialects, sending only 1 */
-		inbuflen = sizeof(struct validate_negotiate_info_req) - 4;
+		inbuflen = sizeof(*pneg_inbuf) -
+				sizeof(pneg_inbuf->Dialects[0]) * 2;
 	}
 
 	rc = SMB2_ioctl(xid, tcon, NO_FILE_ID, NO_FILE_ID,
 		FSCTL_VALIDATE_NEGOTIATE_INFO, true /* is_fsctl */,
-		(char *)&vneg_inbuf, sizeof(struct validate_negotiate_info_req),
-		(char **)&pneg_rsp, &rsplen);
+		(char *)pneg_inbuf, inbuflen, (char **)&pneg_rsp, &rsplen);
 
 	if (rc != 0) {
 		cifs_dbg(VFS, "validate protocol negotiate failed: %d\n", rc);
-		return -EIO;
+		rc = -EIO;
+		goto out_free_inbuf;
 	}
 
-	if (rsplen != sizeof(struct validate_negotiate_info_rsp)) {
+	rc = -EIO;
+	if (rsplen != sizeof(*pneg_rsp)) {
 		cifs_dbg(VFS, "invalid protocol negotiate response size: %d\n",
 			 rsplen);
 
 		/* relax check since Mac returns max bufsize allowed on ioctl */
-		if ((rsplen > CIFSMaxBufSize)
-		     || (rsplen < sizeof(struct validate_negotiate_info_rsp)))
-			goto err_rsp_free;
+		if (rsplen > CIFSMaxBufSize || rsplen < sizeof(*pneg_rsp))
+			goto out_free_rsp;
 	}
 
 	/* check validate negotiate info response matches what we got earlier */
@@ -838,15 +844,17 @@ int smb3_validate_negotiate(const unsigned int xid, struct cifs_tcon *tcon)
 		goto vneg_out;
 
 	/* validate negotiate successful */
+	rc = 0;
 	cifs_dbg(FYI, "validate negotiate info successful\n");
-	kfree(pneg_rsp);
-	return 0;
+	goto out_free_rsp;
 
 vneg_out:
 	cifs_dbg(VFS, "protocol revalidation - security settings mismatch\n");
-err_rsp_free:
+out_free_rsp:
 	kfree(pneg_rsp);
-	return -EIO;
+out_free_inbuf:
+	kfree(pneg_inbuf);
+	return rc;
 }
 
 enum securityEnum
-- 
cgit v1.1


From f7c439668a291ca94f358e44d3a3e9f2a2524b8a Mon Sep 17 00:00:00 2001
From: Long Li <longli@microsoft.com>
Date: Wed, 25 Apr 2018 11:30:05 -0700
Subject: cifs: smbd: Enable signing with smbdirect

Now signing is supported with RDMA transport.

Remove the code that disabled it.

Signed-off-by: Long Li <longli@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Reviewed-by: Ronnie Sahlberg <lsahlber@redhat.com>
---
 fs/cifs/connect.c | 8 --------
 fs/cifs/smb2pdu.c | 5 -----
 2 files changed, 13 deletions(-)

diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c
index a5aa158..7a10a5d 100644
--- a/fs/cifs/connect.c
+++ b/fs/cifs/connect.c
@@ -1977,14 +1977,6 @@ cifs_parse_mount_options(const char *mountdata, const char *devname,
 		goto cifs_parse_mount_err;
 	}
 
-#ifdef CONFIG_CIFS_SMB_DIRECT
-	if (vol->rdma && vol->sign) {
-		cifs_dbg(VFS, "Currently SMB direct doesn't support signing."
-			" This is being fixed\n");
-		goto cifs_parse_mount_err;
-	}
-#endif
-
 #ifndef CONFIG_KEYS
 	/* Muliuser mounts require CONFIG_KEYS support */
 	if (vol->multiuser) {
diff --git a/fs/cifs/smb2pdu.c b/fs/cifs/smb2pdu.c
index 260e9c4..0f48741 100644
--- a/fs/cifs/smb2pdu.c
+++ b/fs/cifs/smb2pdu.c
@@ -738,11 +738,6 @@ int smb3_validate_negotiate(const unsigned int xid, struct cifs_tcon *tcon)
 
 	cifs_dbg(FYI, "validate negotiate\n");
 
-#ifdef CONFIG_CIFS_SMB_DIRECT
-	if (tcon->ses->server->rdma)
-		return 0;
-#endif
-
 	/* In SMB3.11 preauth integrity supersedes validate negotiate */
 	if (tcon->ses->server->dialect == SMB311_PROT_ID)
 		return 0;
-- 
cgit v1.1


From ae2cd7fb478b8da707906ee1706ae1379968a8f9 Mon Sep 17 00:00:00 2001
From: Paulo Alcantara <palcantara@suse.de>
Date: Fri, 4 May 2018 11:25:26 -0300
Subject: cifs: smb2ops: Fix listxattr() when there are no EAs

As per listxattr(2):

       On success, a nonnegative number is returned indicating the size
       of the extended attribute name list.  On failure, -1 is returned
       and errno  is set appropriately.

In SMB1, when the server returns an empty EA list through a listxattr(),
it will correctly return 0 as there are no EAs for the given file.

However, in SMB2+, it returns -ENODATA in listxattr() which is wrong since
the request and response were sent successfully, although there's no actual
EA for the given file.

This patch fixes listxattr() for SMB2+ by returning 0 in cifs_listxattr()
when the server returns an empty list of EAs.

Signed-off-by: Paulo Alcantara <palcantara@suse.de>
Reviewed-by: Aurelien Aptel <aaptel@suse.com>
Signed-off-by: Steve French <smfrench@gmail.com>
---
 fs/cifs/smb2ops.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/fs/cifs/smb2ops.c b/fs/cifs/smb2ops.c
index b76b858..9c6d95f 100644
--- a/fs/cifs/smb2ops.c
+++ b/fs/cifs/smb2ops.c
@@ -589,9 +589,15 @@ smb2_query_eas(const unsigned int xid, struct cifs_tcon *tcon,
 
 	SMB2_close(xid, tcon, fid.persistent_fid, fid.volatile_fid);
 
+	/*
+	 * If ea_name is NULL (listxattr) and there are no EAs, return 0 as it's
+	 * not an error. Otherwise, the specified ea_name was not found.
+	 */
 	if (!rc)
 		rc = move_smb2_ea_to_cifs(ea_data, buf_size, smb2_data,
 					  SMB2_MAX_EA_BUF, ea_name);
+	else if (!ea_name && rc == -ENODATA)
+		rc = 0;
 
 	kfree(smb2_data);
 	return rc;
-- 
cgit v1.1


From a88ad3ded15daa0389106779c60b8a5e76d4b20a Mon Sep 17 00:00:00 2001
From: David Lechner <david@lechnology.com>
Date: Wed, 14 Mar 2018 17:58:45 -0500
Subject: drm/tilcdc: Fix setting clock divider for omap-l138

This fixes setting the clock divider on the TI OMAP-L138 LCDK board.

The clock drivers for OMAP-L138 are being covernted to the common clock
framework. When this happens, clk_set_rate() will no longer return an
error. However, on this SoC, the clock rate cannot actually be changed
because the clock has to maintain a fixed ratio to the ARM clock. So
after attempting to set the clock rate, we need to check to see if the
new rate is actually close enough. If not, then follow the previous
error path to adjust the divider in LCDC IP block to compensate for not
being able to change the parent clock rate.

Tested working on a TI OMAP-L138 LCDK board.

Signed-off-by: David Lechner <david@lechnology.com>
Signed-off-by: Jyri Sarha <jsarha@ti.com>
---
 drivers/gpu/drm/tilcdc/tilcdc_crtc.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/tilcdc/tilcdc_crtc.c b/drivers/gpu/drm/tilcdc/tilcdc_crtc.c
index 1b278a2..1067e70 100644
--- a/drivers/gpu/drm/tilcdc/tilcdc_crtc.c
+++ b/drivers/gpu/drm/tilcdc/tilcdc_crtc.c
@@ -224,7 +224,7 @@ static void tilcdc_crtc_set_clk(struct drm_crtc *crtc)
 
 	ret = clk_set_rate(priv->clk, req_rate * clkdiv);
 	clk_rate = clk_get_rate(priv->clk);
-	if (ret < 0) {
+	if (ret < 0 || tilcdc_pclk_diff(req_rate, clk_rate) > 5) {
 		/*
 		 * If we fail to set the clock rate (some architectures don't
 		 * use the common clock framework yet and may not implement
-- 
cgit v1.1


From c5191133405ac317d20d23c8510416e18842031d Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Fri, 20 Apr 2018 11:05:07 -0400
Subject: drm/amd/display: Add VG12 ASIC IDs

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/include/dal_asic_id.h | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/include/dal_asic_id.h b/drivers/gpu/drm/amd/display/include/dal_asic_id.h
index 9831cb5..9b0a04f 100644
--- a/drivers/gpu/drm/amd/display/include/dal_asic_id.h
+++ b/drivers/gpu/drm/amd/display/include/dal_asic_id.h
@@ -113,9 +113,14 @@
 
 #define AI_GREENLAND_P_A0 1
 #define AI_GREENLAND_P_A1 2
+#define AI_UNKNOWN 0xFF
 
-#define ASICREV_IS_GREENLAND_M(eChipRev)  (eChipRev < AI_UNKNOWN)
-#define ASICREV_IS_GREENLAND_P(eChipRev)  (eChipRev < AI_UNKNOWN)
+#define AI_VEGA12_P_A0 20
+#define ASICREV_IS_GREENLAND_M(eChipRev)  (eChipRev < AI_VEGA12_P_A0)
+#define ASICREV_IS_GREENLAND_P(eChipRev)  (eChipRev < AI_VEGA12_P_A0)
+
+#define ASICREV_IS_VEGA12_P(eChipRev) ((eChipRev >= AI_VEGA12_P_A0) && (eChipRev < AI_UNKNOWN))
+#define ASICREV_IS_VEGA12_p(eChipRev) ((eChipRev >= AI_VEGA12_P_A0) && (eChipRev < AI_UNKNOWN))
 
 /* DCN1_0 */
 #define INTERNAL_REV_RAVEN_A0             0x00    /* First spin of Raven */
-- 
cgit v1.1


From 60a5205fb5f3da3907b8b53561571a790e7b1e70 Mon Sep 17 00:00:00 2001
From: "Jerry (Fangzhi) Zuo" <Jerry.Zuo@amd.com>
Date: Mon, 5 Mar 2018 14:59:57 -0500
Subject: drm/amd: Add BIOS smu_info v3_3 required struct def.

Signed-off-by: Jerry (Fangzhi) Zuo <Jerry.Zuo@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/include/atomfirmware.h | 170 ++++++++++++++++++++++++++++-
 1 file changed, 168 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/include/atomfirmware.h b/drivers/gpu/drm/amd/include/atomfirmware.h
index 0f5ad54..de177ce 100644
--- a/drivers/gpu/drm/amd/include/atomfirmware.h
+++ b/drivers/gpu/drm/amd/include/atomfirmware.h
@@ -501,6 +501,32 @@ enum atom_cooling_solution_id{
   LIQUID_COOLING = 0x01
 };
 
+struct atom_firmware_info_v3_2 {
+  struct atom_common_table_header table_header;
+  uint32_t firmware_revision;
+  uint32_t bootup_sclk_in10khz;
+  uint32_t bootup_mclk_in10khz;
+  uint32_t firmware_capability;             // enum atombios_firmware_capability
+  uint32_t main_call_parser_entry;          /* direct address of main parser call in VBIOS binary. */
+  uint32_t bios_scratch_reg_startaddr;      // 1st bios scratch register dword address
+  uint16_t bootup_vddc_mv;
+  uint16_t bootup_vddci_mv;
+  uint16_t bootup_mvddc_mv;
+  uint16_t bootup_vddgfx_mv;
+  uint8_t  mem_module_id;
+  uint8_t  coolingsolution_id;              /*0: Air cooling; 1: Liquid cooling ... */
+  uint8_t  reserved1[2];
+  uint32_t mc_baseaddr_high;
+  uint32_t mc_baseaddr_low;
+  uint8_t  board_i2c_feature_id;            // enum of atom_board_i2c_feature_id_def
+  uint8_t  board_i2c_feature_gpio_id;       // i2c id find in gpio_lut data table gpio_id
+  uint8_t  board_i2c_feature_slave_addr;
+  uint8_t  reserved3;
+  uint16_t bootup_mvddq_mv;
+  uint16_t bootup_mvpp_mv;
+  uint32_t zfbstartaddrin16mb;
+  uint32_t reserved2[3];
+};
 
 /* 
   ***************************************************************************
@@ -1169,7 +1195,29 @@ struct  atom_gfx_info_v2_2
   uint32_t rlc_gpu_timer_refclk; 
 };
 
-
+struct  atom_gfx_info_v2_3 {
+  struct  atom_common_table_header  table_header;
+  uint8_t gfxip_min_ver;
+  uint8_t gfxip_max_ver;
+  uint8_t max_shader_engines;
+  uint8_t max_tile_pipes;
+  uint8_t max_cu_per_sh;
+  uint8_t max_sh_per_se;
+  uint8_t max_backends_per_se;
+  uint8_t max_texture_channel_caches;
+  uint32_t regaddr_cp_dma_src_addr;
+  uint32_t regaddr_cp_dma_src_addr_hi;
+  uint32_t regaddr_cp_dma_dst_addr;
+  uint32_t regaddr_cp_dma_dst_addr_hi;
+  uint32_t regaddr_cp_dma_command;
+  uint32_t regaddr_cp_status;
+  uint32_t regaddr_rlc_gpu_clock_32;
+  uint32_t rlc_gpu_timer_refclk;
+  uint8_t active_cu_per_sh;
+  uint8_t active_rb_per_se;
+  uint16_t gcgoldenoffset;
+  uint32_t rm21_sram_vmin_value;
+};
 
 /* 
   ***************************************************************************
@@ -1198,6 +1246,76 @@ struct atom_smu_info_v3_1
   uint8_t  fw_ctf_polarity;         // GPIO polarity for CTF
 };
 
+struct atom_smu_info_v3_2 {
+  struct   atom_common_table_header  table_header;
+  uint8_t  smuip_min_ver;
+  uint8_t  smuip_max_ver;
+  uint8_t  smu_rsd1;
+  uint8_t  gpuclk_ss_mode;
+  uint16_t sclk_ss_percentage;
+  uint16_t sclk_ss_rate_10hz;
+  uint16_t gpuclk_ss_percentage;    // in unit of 0.001%
+  uint16_t gpuclk_ss_rate_10hz;
+  uint32_t core_refclk_10khz;
+  uint8_t  ac_dc_gpio_bit;          // GPIO bit shift in SMU_GPIOPAD_A  configured for AC/DC switching, =0xff means invalid
+  uint8_t  ac_dc_polarity;          // GPIO polarity for AC/DC switching
+  uint8_t  vr0hot_gpio_bit;         // GPIO bit shift in SMU_GPIOPAD_A  configured for VR0 HOT event, =0xff means invalid
+  uint8_t  vr0hot_polarity;         // GPIO polarity for VR0 HOT event
+  uint8_t  vr1hot_gpio_bit;         // GPIO bit shift in SMU_GPIOPAD_A configured for VR1 HOT event , =0xff means invalid
+  uint8_t  vr1hot_polarity;         // GPIO polarity for VR1 HOT event
+  uint8_t  fw_ctf_gpio_bit;         // GPIO bit shift in SMU_GPIOPAD_A configured for CTF, =0xff means invalid
+  uint8_t  fw_ctf_polarity;         // GPIO polarity for CTF
+  uint8_t  pcc_gpio_bit;            // GPIO bit shift in SMU_GPIOPAD_A configured for PCC, =0xff means invalid
+  uint8_t  pcc_gpio_polarity;       // GPIO polarity for CTF
+  uint16_t smugoldenoffset;
+  uint32_t gpupll_vco_freq_10khz;
+  uint32_t bootup_smnclk_10khz;
+  uint32_t bootup_socclk_10khz;
+  uint32_t bootup_mp0clk_10khz;
+  uint32_t bootup_mp1clk_10khz;
+  uint32_t bootup_lclk_10khz;
+  uint32_t bootup_dcefclk_10khz;
+  uint32_t ctf_threshold_override_value;
+  uint32_t reserved[5];
+};
+
+struct atom_smu_info_v3_3 {
+  struct   atom_common_table_header  table_header;
+  uint8_t  smuip_min_ver;
+  uint8_t  smuip_max_ver;
+  uint8_t  smu_rsd1;
+  uint8_t  gpuclk_ss_mode;
+  uint16_t sclk_ss_percentage;
+  uint16_t sclk_ss_rate_10hz;
+  uint16_t gpuclk_ss_percentage;    // in unit of 0.001%
+  uint16_t gpuclk_ss_rate_10hz;
+  uint32_t core_refclk_10khz;
+  uint8_t  ac_dc_gpio_bit;          // GPIO bit shift in SMU_GPIOPAD_A  configured for AC/DC switching, =0xff means invalid
+  uint8_t  ac_dc_polarity;          // GPIO polarity for AC/DC switching
+  uint8_t  vr0hot_gpio_bit;         // GPIO bit shift in SMU_GPIOPAD_A  configured for VR0 HOT event, =0xff means invalid
+  uint8_t  vr0hot_polarity;         // GPIO polarity for VR0 HOT event
+  uint8_t  vr1hot_gpio_bit;         // GPIO bit shift in SMU_GPIOPAD_A configured for VR1 HOT event , =0xff means invalid
+  uint8_t  vr1hot_polarity;         // GPIO polarity for VR1 HOT event
+  uint8_t  fw_ctf_gpio_bit;         // GPIO bit shift in SMU_GPIOPAD_A configured for CTF, =0xff means invalid
+  uint8_t  fw_ctf_polarity;         // GPIO polarity for CTF
+  uint8_t  pcc_gpio_bit;            // GPIO bit shift in SMU_GPIOPAD_A configured for PCC, =0xff means invalid
+  uint8_t  pcc_gpio_polarity;       // GPIO polarity for CTF
+  uint16_t smugoldenoffset;
+  uint32_t gpupll_vco_freq_10khz;
+  uint32_t bootup_smnclk_10khz;
+  uint32_t bootup_socclk_10khz;
+  uint32_t bootup_mp0clk_10khz;
+  uint32_t bootup_mp1clk_10khz;
+  uint32_t bootup_lclk_10khz;
+  uint32_t bootup_dcefclk_10khz;
+  uint32_t ctf_threshold_override_value;
+  uint32_t syspll3_0_vco_freq_10khz;
+  uint32_t syspll3_1_vco_freq_10khz;
+  uint32_t bootup_fclk_10khz;
+  uint32_t bootup_waflclk_10khz;
+  uint32_t reserved[3];
+};
+
 /*
  ***************************************************************************
    Data Table smc_dpm_info  structure
@@ -1283,7 +1401,6 @@ struct atom_smc_dpm_info_v4_1
 	uint32_t boardreserved[10];
 };
 
-
 /* 
   ***************************************************************************
     Data Table asic_profiling_info  structure
@@ -1864,6 +1981,55 @@ enum atom_smu9_syspll0_clock_id
   SMU9_SYSPLL0_DISPCLK_ID  = 11,      //	DISPCLK
 };
 
+enum atom_smu11_syspll_id {
+  SMU11_SYSPLL0_ID            = 0,
+  SMU11_SYSPLL1_0_ID          = 1,
+  SMU11_SYSPLL1_1_ID          = 2,
+  SMU11_SYSPLL1_2_ID          = 3,
+  SMU11_SYSPLL2_ID            = 4,
+  SMU11_SYSPLL3_0_ID          = 5,
+  SMU11_SYSPLL3_1_ID          = 6,
+};
+
+
+enum atom_smu11_syspll0_clock_id {
+  SMU11_SYSPLL0_SOCCLK_ID   = 0,       //	SOCCLK
+  SMU11_SYSPLL0_MP0CLK_ID   = 1,       //	MP0CLK
+  SMU11_SYSPLL0_DCLK_ID     = 2,       //	DCLK
+  SMU11_SYSPLL0_VCLK_ID     = 3,       //	VCLK
+  SMU11_SYSPLL0_ECLK_ID     = 4,       //	ECLK
+  SMU11_SYSPLL0_DCEFCLK_ID  = 5,       //	DCEFCLK
+};
+
+
+enum atom_smu11_syspll1_0_clock_id {
+  SMU11_SYSPLL1_0_UCLKA_ID   = 0,       // UCLK_a
+};
+
+enum atom_smu11_syspll1_1_clock_id {
+  SMU11_SYSPLL1_0_UCLKB_ID   = 0,       // UCLK_b
+};
+
+enum atom_smu11_syspll1_2_clock_id {
+  SMU11_SYSPLL1_0_FCLK_ID   = 0,        // FCLK
+};
+
+enum atom_smu11_syspll2_clock_id {
+  SMU11_SYSPLL2_GFXCLK_ID   = 0,        // GFXCLK
+};
+
+enum atom_smu11_syspll3_0_clock_id {
+  SMU11_SYSPLL3_0_WAFCLK_ID = 0,       //	WAFCLK
+  SMU11_SYSPLL3_0_DISPCLK_ID = 1,      //	DISPCLK
+  SMU11_SYSPLL3_0_DPREFCLK_ID = 2,     //	DPREFCLK
+};
+
+enum atom_smu11_syspll3_1_clock_id {
+  SMU11_SYSPLL3_1_MP1CLK_ID = 0,       //	MP1CLK
+  SMU11_SYSPLL3_1_SMNCLK_ID = 1,       //	SMNCLK
+  SMU11_SYSPLL3_1_LCLK_ID = 2,         //	LCLK
+};
+
 struct  atom_get_smu_clock_info_output_parameters_v3_1
 {
   union {
-- 
cgit v1.1


From 6e65fb862064663ad3a08f964af1e8f3f2abf688 Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Fri, 20 Apr 2018 10:56:18 -0400
Subject: drm/amd/display: Add get_firmware_info_v3_2 for VG12

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c | 86 +++++++++++++++++++++-
 1 file changed, 85 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
index 985fe8c..10a5807 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
@@ -70,6 +70,10 @@ static enum bp_result get_firmware_info_v3_1(
 	struct bios_parser *bp,
 	struct dc_firmware_info *info);
 
+static enum bp_result get_firmware_info_v3_2(
+	struct bios_parser *bp,
+	struct dc_firmware_info *info);
+
 static struct atom_hpd_int_record *get_hpd_record(struct bios_parser *bp,
 		struct atom_display_object_path_v2 *object);
 
@@ -1321,9 +1325,11 @@ static enum bp_result bios_parser_get_firmware_info(
 		case 3:
 			switch (revision.minor) {
 			case 1:
-			case 2:
 				result = get_firmware_info_v3_1(bp, info);
 				break;
+			case 2:
+				result = get_firmware_info_v3_2(bp, info);
+				break;
 			default:
 				break;
 			}
@@ -1383,6 +1389,84 @@ static enum bp_result get_firmware_info_v3_1(
 	return BP_RESULT_OK;
 }
 
+static enum bp_result get_firmware_info_v3_2(
+	struct bios_parser *bp,
+	struct dc_firmware_info *info)
+{
+	struct atom_firmware_info_v3_2 *firmware_info;
+	struct atom_display_controller_info_v4_1 *dce_info = NULL;
+	struct atom_common_table_header *header;
+	struct atom_data_revision revision;
+	struct atom_smu_info_v3_2 *smu_info_v3_2 = NULL;
+	struct atom_smu_info_v3_3 *smu_info_v3_3 = NULL;
+
+	if (!info)
+		return BP_RESULT_BADINPUT;
+
+	firmware_info = GET_IMAGE(struct atom_firmware_info_v3_2,
+			DATA_TABLES(firmwareinfo));
+
+	dce_info = GET_IMAGE(struct atom_display_controller_info_v4_1,
+			DATA_TABLES(dce_info));
+
+	if (!firmware_info || !dce_info)
+		return BP_RESULT_BADBIOSTABLE;
+
+	memset(info, 0, sizeof(*info));
+
+	header = GET_IMAGE(struct atom_common_table_header,
+					DATA_TABLES(smu_info));
+	get_atom_data_table_revision(header, &revision);
+
+	if (revision.minor == 2) {
+		/* Vega12 */
+		smu_info_v3_2 = GET_IMAGE(struct atom_smu_info_v3_2,
+							DATA_TABLES(smu_info));
+
+		if (!smu_info_v3_2)
+			return BP_RESULT_BADBIOSTABLE;
+
+		info->default_engine_clk = smu_info_v3_2->bootup_dcefclk_10khz * 10;
+	} else if (revision.minor == 3) {
+		/* Vega20 */
+		smu_info_v3_3 = GET_IMAGE(struct atom_smu_info_v3_3,
+							DATA_TABLES(smu_info));
+
+		if (!smu_info_v3_3)
+			return BP_RESULT_BADBIOSTABLE;
+
+		info->default_engine_clk = smu_info_v3_3->bootup_dcefclk_10khz * 10;
+	}
+
+	 // We need to convert from 10KHz units into KHz units.
+	info->default_memory_clk = firmware_info->bootup_mclk_in10khz * 10;
+
+	 /* 27MHz for Vega10 & Vega12; 100MHz for Vega20 */
+	info->pll_info.crystal_frequency = dce_info->dce_refclk_10khz * 10;
+	/* Hardcode frequency if BIOS gives no DCE Ref Clk */
+	if (info->pll_info.crystal_frequency == 0) {
+		if (revision.minor == 2)
+			info->pll_info.crystal_frequency = 27000;
+		else if (revision.minor == 3)
+			info->pll_info.crystal_frequency = 100000;
+	}
+	/*dp_phy_ref_clk is not correct for atom_display_controller_info_v4_2, but we don't use it*/
+	info->dp_phy_ref_clk     = dce_info->dpphy_refclk_10khz * 10;
+	info->i2c_engine_ref_clk = dce_info->i2c_engine_refclk_10khz * 10;
+
+	/* Get GPU PLL VCO Clock */
+	if (bp->cmd_tbl.get_smu_clock_info != NULL) {
+		if (revision.minor == 2)
+			info->smu_gpu_pll_output_freq =
+					bp->cmd_tbl.get_smu_clock_info(bp, SMU9_SYSPLL0_ID) * 10;
+		else if (revision.minor == 3)
+			info->smu_gpu_pll_output_freq =
+					bp->cmd_tbl.get_smu_clock_info(bp, SMU11_SYSPLL3_0_ID) * 10;
+	}
+
+	return BP_RESULT_OK;
+}
+
 static enum bp_result bios_parser_get_encoder_cap_info(
 	struct dc_bios *dcb,
 	struct graphics_object_id object_id,
-- 
cgit v1.1


From 018d82e5f02ef3583411bcaa4e00c69786f46f19 Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Tue, 24 Apr 2018 10:49:20 -0400
Subject: drm/amd/display: Don't return ddc result and read_bytes in same
 return value

The two ranges overlap.

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c  | 20 ++++++++++++--------
 drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c    | 10 +++++++---
 drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h     |  5 +++--
 3 files changed, 22 insertions(+), 13 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
index ace9ad5..4304d9e 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
@@ -83,21 +83,22 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux,
 	enum i2c_mot_mode mot = (msg->request & DP_AUX_I2C_MOT) ?
 		I2C_MOT_TRUE : I2C_MOT_FALSE;
 	enum ddc_result res;
-	ssize_t read_bytes;
+	uint32_t read_bytes = msg->size;
 
 	if (WARN_ON(msg->size > 16))
 		return -E2BIG;
 
 	switch (msg->request & ~DP_AUX_I2C_MOT) {
 	case DP_AUX_NATIVE_READ:
-		read_bytes = dal_ddc_service_read_dpcd_data(
+		res = dal_ddc_service_read_dpcd_data(
 				TO_DM_AUX(aux)->ddc_service,
 				false,
 				I2C_MOT_UNDEF,
 				msg->address,
 				msg->buffer,
-				msg->size);
-		return read_bytes;
+				msg->size,
+				&read_bytes);
+		break;
 	case DP_AUX_NATIVE_WRITE:
 		res = dal_ddc_service_write_dpcd_data(
 				TO_DM_AUX(aux)->ddc_service,
@@ -108,14 +109,15 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux,
 				msg->size);
 		break;
 	case DP_AUX_I2C_READ:
-		read_bytes = dal_ddc_service_read_dpcd_data(
+		res = dal_ddc_service_read_dpcd_data(
 				TO_DM_AUX(aux)->ddc_service,
 				true,
 				mot,
 				msg->address,
 				msg->buffer,
-				msg->size);
-		return read_bytes;
+				msg->size,
+				&read_bytes);
+		break;
 	case DP_AUX_I2C_WRITE:
 		res = dal_ddc_service_write_dpcd_data(
 				TO_DM_AUX(aux)->ddc_service,
@@ -137,7 +139,9 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux,
 		 r == DDC_RESULT_SUCESSFULL);
 #endif
 
-	return msg->size;
+	if (res != DDC_RESULT_SUCESSFULL)
+		return -EIO;
+	return read_bytes;
 }
 
 static enum drm_connector_status
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c
index 49c2fac..ae48d60 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c
@@ -629,13 +629,14 @@ bool dal_ddc_service_query_ddc_data(
 	return ret;
 }
 
-ssize_t dal_ddc_service_read_dpcd_data(
+enum ddc_result dal_ddc_service_read_dpcd_data(
 	struct ddc_service *ddc,
 	bool i2c,
 	enum i2c_mot_mode mot,
 	uint32_t address,
 	uint8_t *data,
-	uint32_t len)
+	uint32_t len,
+	uint32_t *read)
 {
 	struct aux_payload read_payload = {
 		.i2c_over_aux = i2c,
@@ -652,6 +653,8 @@ ssize_t dal_ddc_service_read_dpcd_data(
 		.mot = mot
 	};
 
+	*read = 0;
+
 	if (len > DEFAULT_AUX_MAX_DATA_SIZE) {
 		BREAK_TO_DEBUGGER();
 		return DDC_RESULT_FAILED_INVALID_OPERATION;
@@ -661,7 +664,8 @@ ssize_t dal_ddc_service_read_dpcd_data(
 		ddc->ctx->i2caux,
 		ddc->ddc_pin,
 		&command)) {
-		return (ssize_t)command.payloads->length;
+		*read = command.payloads->length;
+		return DDC_RESULT_SUCESSFULL;
 	}
 
 	return DDC_RESULT_FAILED_OPERATION;
diff --git a/drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h b/drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h
index 090b7a8..30b3a08 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h
@@ -102,13 +102,14 @@ bool dal_ddc_service_query_ddc_data(
 		uint8_t *read_buf,
 		uint32_t read_size);
 
-ssize_t dal_ddc_service_read_dpcd_data(
+enum ddc_result dal_ddc_service_read_dpcd_data(
 		struct ddc_service *ddc,
 		bool i2c,
 		enum i2c_mot_mode mot,
 		uint32_t address,
 		uint8_t *data,
-		uint32_t len);
+		uint32_t len,
+		uint32_t *read);
 
 enum ddc_result dal_ddc_service_write_dpcd_data(
 		struct ddc_service *ddc,
-- 
cgit v1.1


From bd4caed47a19f25fe8674344ea06d469c27ac314 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Michel=20D=C3=A4nzer?= <michel.daenzer@amd.com>
Date: Tue, 17 Apr 2018 12:25:22 +0200
Subject: drm/amd/display: Use kvzalloc for potentially large allocations
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Allocating up to 32 physically contiguous pages can easily fail (and has
failed for me), and isn't necessary anyway.

Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_surface.c   | 14 ++---
 .../drm/amd/display/modules/color/color_gamma.c    | 72 ++++++++++++----------
 2 files changed, 45 insertions(+), 41 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_surface.c b/drivers/gpu/drm/amd/display/dc/core/dc_surface.c
index ade5b8e..132eef3 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_surface.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_surface.c
@@ -66,8 +66,8 @@ struct dc_plane_state *dc_create_plane_state(struct dc *dc)
 {
 	struct dc *core_dc = dc;
 
-	struct dc_plane_state *plane_state = kzalloc(sizeof(*plane_state),
-						     GFP_KERNEL);
+	struct dc_plane_state *plane_state = kvzalloc(sizeof(*plane_state),
+						      GFP_KERNEL);
 
 	if (NULL == plane_state)
 		return NULL;
@@ -120,7 +120,7 @@ static void dc_plane_state_free(struct kref *kref)
 {
 	struct dc_plane_state *plane_state = container_of(kref, struct dc_plane_state, refcount);
 	destruct(plane_state);
-	kfree(plane_state);
+	kvfree(plane_state);
 }
 
 void dc_plane_state_release(struct dc_plane_state *plane_state)
@@ -136,7 +136,7 @@ void dc_gamma_retain(struct dc_gamma *gamma)
 static void dc_gamma_free(struct kref *kref)
 {
 	struct dc_gamma *gamma = container_of(kref, struct dc_gamma, refcount);
-	kfree(gamma);
+	kvfree(gamma);
 }
 
 void dc_gamma_release(struct dc_gamma **gamma)
@@ -147,7 +147,7 @@ void dc_gamma_release(struct dc_gamma **gamma)
 
 struct dc_gamma *dc_create_gamma(void)
 {
-	struct dc_gamma *gamma = kzalloc(sizeof(*gamma), GFP_KERNEL);
+	struct dc_gamma *gamma = kvzalloc(sizeof(*gamma), GFP_KERNEL);
 
 	if (gamma == NULL)
 		goto alloc_fail;
@@ -167,7 +167,7 @@ void dc_transfer_func_retain(struct dc_transfer_func *tf)
 static void dc_transfer_func_free(struct kref *kref)
 {
 	struct dc_transfer_func *tf = container_of(kref, struct dc_transfer_func, refcount);
-	kfree(tf);
+	kvfree(tf);
 }
 
 void dc_transfer_func_release(struct dc_transfer_func *tf)
@@ -177,7 +177,7 @@ void dc_transfer_func_release(struct dc_transfer_func *tf)
 
 struct dc_transfer_func *dc_create_transfer_func(void)
 {
-	struct dc_transfer_func *tf = kzalloc(sizeof(*tf), GFP_KERNEL);
+	struct dc_transfer_func *tf = kvzalloc(sizeof(*tf), GFP_KERNEL);
 
 	if (tf == NULL)
 		goto alloc_fail;
diff --git a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c
index e7e374f..b3747a0 100644
--- a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c
+++ b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c
@@ -1093,19 +1093,19 @@ bool mod_color_calculate_regamma_params(struct dc_transfer_func *output_tf,
 
 	output_tf->type = TF_TYPE_DISTRIBUTED_POINTS;
 
-	rgb_user = kzalloc(sizeof(*rgb_user) * (ramp->num_entries + _EXTRA_POINTS),
-			   GFP_KERNEL);
+	rgb_user = kvzalloc(sizeof(*rgb_user) * (ramp->num_entries + _EXTRA_POINTS),
+			    GFP_KERNEL);
 	if (!rgb_user)
 		goto rgb_user_alloc_fail;
-	rgb_regamma = kzalloc(sizeof(*rgb_regamma) * (MAX_HW_POINTS + _EXTRA_POINTS),
-			GFP_KERNEL);
+	rgb_regamma = kvzalloc(sizeof(*rgb_regamma) * (MAX_HW_POINTS + _EXTRA_POINTS),
+			       GFP_KERNEL);
 	if (!rgb_regamma)
 		goto rgb_regamma_alloc_fail;
-	axix_x = kzalloc(sizeof(*axix_x) * (ramp->num_entries + 3),
-			 GFP_KERNEL);
+	axix_x = kvzalloc(sizeof(*axix_x) * (ramp->num_entries + 3),
+			  GFP_KERNEL);
 	if (!axix_x)
 		goto axix_x_alloc_fail;
-	coeff = kzalloc(sizeof(*coeff) * (MAX_HW_POINTS + _EXTRA_POINTS), GFP_KERNEL);
+	coeff = kvzalloc(sizeof(*coeff) * (MAX_HW_POINTS + _EXTRA_POINTS), GFP_KERNEL);
 	if (!coeff)
 		goto coeff_alloc_fail;
 
@@ -1157,13 +1157,13 @@ bool mod_color_calculate_regamma_params(struct dc_transfer_func *output_tf,
 
 	ret = true;
 
-	kfree(coeff);
+	kvfree(coeff);
 coeff_alloc_fail:
-	kfree(axix_x);
+	kvfree(axix_x);
 axix_x_alloc_fail:
-	kfree(rgb_regamma);
+	kvfree(rgb_regamma);
 rgb_regamma_alloc_fail:
-	kfree(rgb_user);
+	kvfree(rgb_user);
 rgb_user_alloc_fail:
 	return ret;
 }
@@ -1192,19 +1192,19 @@ bool mod_color_calculate_degamma_params(struct dc_transfer_func *input_tf,
 
 	input_tf->type = TF_TYPE_DISTRIBUTED_POINTS;
 
-	rgb_user = kzalloc(sizeof(*rgb_user) * (ramp->num_entries + _EXTRA_POINTS),
-			   GFP_KERNEL);
+	rgb_user = kvzalloc(sizeof(*rgb_user) * (ramp->num_entries + _EXTRA_POINTS),
+			    GFP_KERNEL);
 	if (!rgb_user)
 		goto rgb_user_alloc_fail;
-	curve = kzalloc(sizeof(*curve) * (MAX_HW_POINTS + _EXTRA_POINTS),
-			GFP_KERNEL);
+	curve = kvzalloc(sizeof(*curve) * (MAX_HW_POINTS + _EXTRA_POINTS),
+			 GFP_KERNEL);
 	if (!curve)
 		goto curve_alloc_fail;
-	axix_x = kzalloc(sizeof(*axix_x) * (ramp->num_entries + _EXTRA_POINTS),
-			 GFP_KERNEL);
+	axix_x = kvzalloc(sizeof(*axix_x) * (ramp->num_entries + _EXTRA_POINTS),
+			  GFP_KERNEL);
 	if (!axix_x)
 		goto axix_x_alloc_fail;
-	coeff = kzalloc(sizeof(*coeff) * (MAX_HW_POINTS + _EXTRA_POINTS), GFP_KERNEL);
+	coeff = kvzalloc(sizeof(*coeff) * (MAX_HW_POINTS + _EXTRA_POINTS), GFP_KERNEL);
 	if (!coeff)
 		goto coeff_alloc_fail;
 
@@ -1246,13 +1246,13 @@ bool mod_color_calculate_degamma_params(struct dc_transfer_func *input_tf,
 
 	ret = true;
 
-	kfree(coeff);
+	kvfree(coeff);
 coeff_alloc_fail:
-	kfree(axix_x);
+	kvfree(axix_x);
 axix_x_alloc_fail:
-	kfree(curve);
+	kvfree(curve);
 curve_alloc_fail:
-	kfree(rgb_user);
+	kvfree(rgb_user);
 rgb_user_alloc_fail:
 
 	return ret;
@@ -1281,8 +1281,9 @@ bool  mod_color_calculate_curve(enum dc_transfer_func_predefined trans,
 		}
 		ret = true;
 	} else if (trans == TRANSFER_FUNCTION_PQ) {
-		rgb_regamma = kzalloc(sizeof(*rgb_regamma) * (MAX_HW_POINTS +
-						_EXTRA_POINTS), GFP_KERNEL);
+		rgb_regamma = kvzalloc(sizeof(*rgb_regamma) *
+				       (MAX_HW_POINTS + _EXTRA_POINTS),
+				       GFP_KERNEL);
 		if (!rgb_regamma)
 			goto rgb_regamma_alloc_fail;
 		points->end_exponent = 7;
@@ -1302,11 +1303,12 @@ bool  mod_color_calculate_curve(enum dc_transfer_func_predefined trans,
 		}
 		ret = true;
 
-		kfree(rgb_regamma);
+		kvfree(rgb_regamma);
 	} else if (trans == TRANSFER_FUNCTION_SRGB ||
 			  trans == TRANSFER_FUNCTION_BT709) {
-		rgb_regamma = kzalloc(sizeof(*rgb_regamma) * (MAX_HW_POINTS +
-						_EXTRA_POINTS), GFP_KERNEL);
+		rgb_regamma = kvzalloc(sizeof(*rgb_regamma) *
+				       (MAX_HW_POINTS + _EXTRA_POINTS),
+				       GFP_KERNEL);
 		if (!rgb_regamma)
 			goto rgb_regamma_alloc_fail;
 		points->end_exponent = 0;
@@ -1324,7 +1326,7 @@ bool  mod_color_calculate_curve(enum dc_transfer_func_predefined trans,
 		}
 		ret = true;
 
-		kfree(rgb_regamma);
+		kvfree(rgb_regamma);
 	}
 rgb_regamma_alloc_fail:
 	return ret;
@@ -1348,8 +1350,9 @@ bool  mod_color_calculate_degamma_curve(enum dc_transfer_func_predefined trans,
 		}
 		ret = true;
 	} else if (trans == TRANSFER_FUNCTION_PQ) {
-		rgb_degamma = kzalloc(sizeof(*rgb_degamma) * (MAX_HW_POINTS +
-						_EXTRA_POINTS), GFP_KERNEL);
+		rgb_degamma = kvzalloc(sizeof(*rgb_degamma) *
+				       (MAX_HW_POINTS +	_EXTRA_POINTS),
+				       GFP_KERNEL);
 		if (!rgb_degamma)
 			goto rgb_degamma_alloc_fail;
 
@@ -1364,11 +1367,12 @@ bool  mod_color_calculate_degamma_curve(enum dc_transfer_func_predefined trans,
 		}
 		ret = true;
 
-		kfree(rgb_degamma);
+		kvfree(rgb_degamma);
 	} else if (trans == TRANSFER_FUNCTION_SRGB ||
 			  trans == TRANSFER_FUNCTION_BT709) {
-		rgb_degamma = kzalloc(sizeof(*rgb_degamma) * (MAX_HW_POINTS +
-						_EXTRA_POINTS), GFP_KERNEL);
+		rgb_degamma = kvzalloc(sizeof(*rgb_degamma) *
+				       (MAX_HW_POINTS + _EXTRA_POINTS),
+				       GFP_KERNEL);
 		if (!rgb_degamma)
 			goto rgb_degamma_alloc_fail;
 
@@ -1382,7 +1386,7 @@ bool  mod_color_calculate_degamma_curve(enum dc_transfer_func_predefined trans,
 		}
 		ret = true;
 
-		kfree(rgb_degamma);
+		kvfree(rgb_degamma);
 	}
 	points->end_exponent = 0;
 	points->x_point_at_y1_red = 1;
-- 
cgit v1.1


From da291320baec914f0bb4e65a9dccb86bd6c728f2 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Michel=20D=C3=A4nzer?= <michel.daenzer@amd.com>
Date: Wed, 25 Apr 2018 17:32:10 +0200
Subject: drm/ttm: Use GFP_TRANSHUGE_LIGHT for allocating huge pages
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

GFP_TRANSHUGE tries very hard to allocate huge pages, which can result
in long delays with high memory pressure. I have observed firefox
freezing for up to around a minute due to this while restic was taking
a full system backup.

Since we don't really need huge pages, use GFP_TRANSHUGE_LIGHT |
__GFP_NORETRY instead, in order to fail quickly when there are no huge
pages available.

Set __GFP_KSWAPD_RECLAIM as well, in order for huge pages to be freed
up in the background if necessary.

With these changes, I'm no longer seeing freezes during a restic backup.

Cc: stable@vger.kernel.org
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/ttm/ttm_page_alloc.c     | 11 ++++++++---
 drivers/gpu/drm/ttm/ttm_page_alloc_dma.c |  3 ++-
 2 files changed, 10 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/ttm/ttm_page_alloc.c b/drivers/gpu/drm/ttm/ttm_page_alloc.c
index f0481b7..06c94e3 100644
--- a/drivers/gpu/drm/ttm/ttm_page_alloc.c
+++ b/drivers/gpu/drm/ttm/ttm_page_alloc.c
@@ -910,7 +910,8 @@ static int ttm_get_pages(struct page **pages, unsigned npages, int flags,
 			while (npages >= HPAGE_PMD_NR) {
 				gfp_t huge_flags = gfp_flags;
 
-				huge_flags |= GFP_TRANSHUGE;
+				huge_flags |= GFP_TRANSHUGE_LIGHT | __GFP_NORETRY |
+					__GFP_KSWAPD_RECLAIM;
 				huge_flags &= ~__GFP_MOVABLE;
 				huge_flags &= ~__GFP_COMP;
 				p = alloc_pages(huge_flags, HPAGE_PMD_ORDER);
@@ -1027,11 +1028,15 @@ int ttm_page_alloc_init(struct ttm_mem_global *glob, unsigned max_pages)
 				  GFP_USER | GFP_DMA32, "uc dma", 0);
 
 	ttm_page_pool_init_locked(&_manager->wc_pool_huge,
-				  GFP_TRANSHUGE	& ~(__GFP_MOVABLE | __GFP_COMP),
+				  (GFP_TRANSHUGE_LIGHT | __GFP_NORETRY |
+				   __GFP_KSWAPD_RECLAIM) &
+				  ~(__GFP_MOVABLE | __GFP_COMP),
 				  "wc huge", order);
 
 	ttm_page_pool_init_locked(&_manager->uc_pool_huge,
-				  GFP_TRANSHUGE	& ~(__GFP_MOVABLE | __GFP_COMP)
+				  (GFP_TRANSHUGE_LIGHT | __GFP_NORETRY |
+				   __GFP_KSWAPD_RECLAIM) &
+				  ~(__GFP_MOVABLE | __GFP_COMP)
 				  , "uc huge", order);
 
 	_manager->options.max_size = max_pages;
diff --git a/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c b/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c
index 8a25d19..f63d99c 100644
--- a/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c
+++ b/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c
@@ -910,7 +910,8 @@ static gfp_t ttm_dma_pool_gfp_flags(struct ttm_dma_tt *ttm_dma, bool huge)
 		gfp_flags |= __GFP_ZERO;
 
 	if (huge) {
-		gfp_flags |= GFP_TRANSHUGE;
+		gfp_flags |= GFP_TRANSHUGE_LIGHT | __GFP_NORETRY |
+			__GFP_KSWAPD_RECLAIM;
 		gfp_flags &= ~__GFP_MOVABLE;
 		gfp_flags &= ~__GFP_COMP;
 	}
-- 
cgit v1.1


From e6a5b9f9aee145c2f2c24431d84edfbb0d49eea5 Mon Sep 17 00:00:00 2001
From: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Date: Mon, 30 Apr 2018 10:04:42 -0400
Subject: drm/amdgpu: Switch to interruptable wait to recover from ring hang.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

v2:
Use dma_fence_wait instead of dma_fence_wait_timeout(...,MAX_SCHEDULE_TIMEOUT)
Avoid printing error message for ERESTARTSYS

Originally-by: David Panariti <David.Panariti@amd.com>
Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
index 09d35051..3fabf9f9 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
@@ -419,9 +419,11 @@ int amdgpu_ctx_wait_prev_fence(struct amdgpu_ctx *ctx, unsigned ring_id)
 
 	if (other) {
 		signed long r;
-		r = dma_fence_wait_timeout(other, false, MAX_SCHEDULE_TIMEOUT);
+		r = dma_fence_wait(other, true);
 		if (r < 0) {
-			DRM_ERROR("Error (%ld) waiting for fence!\n", r);
+			if (r != -ERESTARTSYS)
+				DRM_ERROR("Error (%ld) waiting for fence!\n", r);
+
 			return r;
 		}
 	}
-- 
cgit v1.1


From 639f790223e62339b9cb7319ea3fae9e02c39bdb Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Mon, 7 May 2018 14:23:04 +0800
Subject: drm/amd/pp: Refine the output of pp_power_profile_mode on VI
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

In order to keep consist with Vega,
the output format of the pp_power_profile_mode would be
<integer><mode name string>< “*” for current profile>:"detail settings"
and remove the "CURRENT" mode line.

for example:
NUM        MODE_NAME     SCLK_UP_HYST   SCLK_DOWN_HYST SCLK_ACTIVE_LEVEL     MCLK_UP_HYST   MCLK_DOWN_HYST MCLK_ACTIVE_LEVEL
  0   3D_FULL_SCREEN:        0              100               30                0              100               10
  1     POWER_SAVING:       10                0               30                -                -                -
  2            VIDEO:        -                -                -               10               16               31
  3               VR:        0               11               50                0              100               10
  4          COMPUTE:        0                5               30                -                -                -
  5         CUSTOM *:        0                5               30                0              100               10
NUM        MODE_NAME     SCLK_UP_HYST   SCLK_DOWN_HYST SCLK_ACTIVE_LEVEL     MCLK_UP_HYST   MCLK_DOWN_HYST MCLK_ACTIVE_LEVEL
  0   3D_FULL_SCREEN:        0              100               30                0              100               10
  1   POWER_SAVING *:       10                0               30                0              100               10
  2            VIDEO:        -                -                -               10               16               31
  3               VR:        0               11               50                0              100               10
  4          COMPUTE:        0                5               30                -                -                -
  5           CUSTOM:        -                -                -                -                -                -

Reviewed-by: Evan Quan <evan.quan@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c | 52 +++++++++++-------------
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h |  1 -
 2 files changed, 23 insertions(+), 30 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index 26fbeaf..18b5b2f 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -79,12 +79,13 @@
 #define PCIE_BUS_CLK                10000
 #define TCLK                        (PCIE_BUS_CLK / 10)
 
-static const struct profile_mode_setting smu7_profiling[5] =
+static const struct profile_mode_setting smu7_profiling[6] =
 					{{1, 0, 100, 30, 1, 0, 100, 10},
 					 {1, 10, 0, 30, 0, 0, 0, 0},
 					 {0, 0, 0, 0, 1, 10, 16, 31},
 					 {1, 0, 11, 50, 1, 0, 100, 10},
 					 {1, 0, 5, 30, 0, 0, 0, 0},
+					 {0, 0, 0, 0, 0, 0, 0, 0},
 					};
 
 /** Values for the CG_THERMAL_CTRL::DPM_EVENT_SRC field. */
@@ -4864,6 +4865,17 @@ static int smu7_get_power_profile_mode(struct pp_hwmgr *hwmgr, char *buf)
 	len = sizeof(smu7_profiling) / sizeof(struct profile_mode_setting);
 
 	for (i = 0; i < len; i++) {
+		if (i == hwmgr->power_profile_mode) {
+			size += sprintf(buf + size, "%3d %14s %s: %8d %16d %16d %16d %16d %16d\n",
+			i, profile_name[i], "*",
+			data->current_profile_setting.sclk_up_hyst,
+			data->current_profile_setting.sclk_down_hyst,
+			data->current_profile_setting.sclk_activity,
+			data->current_profile_setting.mclk_up_hyst,
+			data->current_profile_setting.mclk_down_hyst,
+			data->current_profile_setting.mclk_activity);
+			continue;
+		}
 		if (smu7_profiling[i].bupdate_sclk)
 			size += sprintf(buf + size, "%3d %16s: %8d %16d %16d ",
 			i, profile_name[i], smu7_profiling[i].sclk_up_hyst,
@@ -4883,24 +4895,6 @@ static int smu7_get_power_profile_mode(struct pp_hwmgr *hwmgr, char *buf)
 			"-", "-", "-");
 	}
 
-	size += sprintf(buf + size, "%3d %16s: %8d %16d %16d %16d %16d %16d\n",
-			i, profile_name[i],
-			data->custom_profile_setting.sclk_up_hyst,
-			data->custom_profile_setting.sclk_down_hyst,
-			data->custom_profile_setting.sclk_activity,
-			data->custom_profile_setting.mclk_up_hyst,
-			data->custom_profile_setting.mclk_down_hyst,
-			data->custom_profile_setting.mclk_activity);
-
-	size += sprintf(buf + size, "%3s %16s: %8d %16d %16d %16d %16d %16d\n",
-			"*", "CURRENT",
-			data->current_profile_setting.sclk_up_hyst,
-			data->current_profile_setting.sclk_down_hyst,
-			data->current_profile_setting.sclk_activity,
-			data->current_profile_setting.mclk_up_hyst,
-			data->current_profile_setting.mclk_down_hyst,
-			data->current_profile_setting.mclk_activity);
-
 	return size;
 }
 
@@ -4939,16 +4933,16 @@ static int smu7_set_power_profile_mode(struct pp_hwmgr *hwmgr, long *input, uint
 		if (size < 8)
 			return -EINVAL;
 
-		data->custom_profile_setting.bupdate_sclk = input[0];
-		data->custom_profile_setting.sclk_up_hyst = input[1];
-		data->custom_profile_setting.sclk_down_hyst = input[2];
-		data->custom_profile_setting.sclk_activity = input[3];
-		data->custom_profile_setting.bupdate_mclk = input[4];
-		data->custom_profile_setting.mclk_up_hyst = input[5];
-		data->custom_profile_setting.mclk_down_hyst = input[6];
-		data->custom_profile_setting.mclk_activity = input[7];
-		if (!smum_update_dpm_settings(hwmgr, &data->custom_profile_setting)) {
-			memcpy(&data->current_profile_setting, &data->custom_profile_setting, sizeof(struct profile_mode_setting));
+		tmp.bupdate_sclk = input[0];
+		tmp.sclk_up_hyst = input[1];
+		tmp.sclk_down_hyst = input[2];
+		tmp.sclk_activity = input[3];
+		tmp.bupdate_mclk = input[4];
+		tmp.mclk_up_hyst = input[5];
+		tmp.mclk_down_hyst = input[6];
+		tmp.mclk_activity = input[7];
+		if (!smum_update_dpm_settings(hwmgr, &tmp)) {
+			memcpy(&data->current_profile_setting, &tmp, sizeof(struct profile_mode_setting));
 			hwmgr->power_profile_mode = mode;
 		}
 		break;
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h
index f40179c..b8d0bb3 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h
@@ -325,7 +325,6 @@ struct smu7_hwmgr {
 	uint16_t                              mem_latency_high;
 	uint16_t                              mem_latency_low;
 	uint32_t                              vr_config;
-	struct profile_mode_setting           custom_profile_setting;
 	struct profile_mode_setting           current_profile_setting;
 };
 
-- 
cgit v1.1


From 9913f74fe15705acd5163551ddf449568cf0048d Mon Sep 17 00:00:00 2001
From: Marek Szyprowski <m.szyprowski@samsung.com>
Date: Thu, 10 May 2018 08:46:36 +0900
Subject: drm/exynos: ipp: Add IPP v2 framework

This patch adds Exynos IPP v2 subsystem and userspace API.

New userspace API is focused ONLY on memory-to-memory image processing.
The two remainging operation modes of obsolete IPP v1 API (framebuffer
writeback and local-path output with image processing) can be implemented
using standard DRM features: writeback connectors and additional DRM planes
with scaling features.

V2 IPP userspace API is based on stateless approach, which much better fits
to memory-to-memory image processing model. It also provides support for
all image formats, which are both already defined in DRM API and supported
by the existing IPP hardware modules.

The API consists of the following ioctls:
- DRM_IOCTL_EXYNOS_IPP_GET_RESOURCES: to enumerate all available image
  processing modules,
- DRM_IOCTL_EXYNOS_IPP_GET_CAPS: to query capabilities and supported image
  formats of given IPP module,
- DRM_IOCTL_EXYNOS_IPP_GET_LIMITS: to query hardware limitiations for
  selected image format of given IPP module,
- DRM_IOCTL_EXYNOS_IPP_COMMIT: to perform operation described by the
  provided structures (source and destination buffers, operation rectangle,
  transformation, etc).

The proposed userspace API is extensible. In the future more advanced image
processing operations can be defined to support for example blending.

Userspace API is fully functional also on DRM render nodes, so it is not
limited to the root/privileged client.

Internal driver API also has been completely rewritten. New IPP core
performs all possible input validation, checks and object life-time
control. The drivers can focus only on writing configuration to hardware
registers. Stateless nature of DRM_IOCTL_EXYNOS_IPP_COMMIT ioctl simplifies
the driver API. Minimal driver needs to provide a single callback for
starting processing and an array with supported image formats.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Tested-by: Hoegeun Kwon <hoegeun.kwon@samsung.com>
Merge conflict so merged manually.
Signed-off-by: Inki Dae <inki.dae@samsung.com>
---
 drivers/gpu/drm/exynos/Kconfig          |   3 +
 drivers/gpu/drm/exynos/Makefile         |   1 +
 drivers/gpu/drm/exynos/exynos_drm_drv.c |  22 +-
 drivers/gpu/drm/exynos/exynos_drm_ipp.c | 916 ++++++++++++++++++++++++++++++++
 drivers/gpu/drm/exynos/exynos_drm_ipp.h | 175 ++++++
 include/uapi/drm/exynos_drm.h           | 240 +++++++++
 6 files changed, 1355 insertions(+), 2 deletions(-)
 create mode 100644 drivers/gpu/drm/exynos/exynos_drm_ipp.c
 create mode 100644 drivers/gpu/drm/exynos/exynos_drm_ipp.h

diff --git a/drivers/gpu/drm/exynos/Kconfig b/drivers/gpu/drm/exynos/Kconfig
index 1548a78..9e91465 100644
--- a/drivers/gpu/drm/exynos/Kconfig
+++ b/drivers/gpu/drm/exynos/Kconfig
@@ -95,6 +95,9 @@ config DRM_EXYNOS_G2D
 	help
 	  Choose this option if you want to use Exynos G2D for DRM.
 
+config DRM_EXYNOS_IPP
+	bool
+
 config DRM_EXYNOS_FIMC
 	bool "FIMC"
 	depends on BROKEN && MFD_SYSCON
diff --git a/drivers/gpu/drm/exynos/Makefile b/drivers/gpu/drm/exynos/Makefile
index a51c545..bdf4212 100644
--- a/drivers/gpu/drm/exynos/Makefile
+++ b/drivers/gpu/drm/exynos/Makefile
@@ -18,6 +18,7 @@ exynosdrm-$(CONFIG_DRM_EXYNOS_MIXER)	+= exynos_mixer.o
 exynosdrm-$(CONFIG_DRM_EXYNOS_HDMI)	+= exynos_hdmi.o
 exynosdrm-$(CONFIG_DRM_EXYNOS_VIDI)	+= exynos_drm_vidi.o
 exynosdrm-$(CONFIG_DRM_EXYNOS_G2D)	+= exynos_drm_g2d.o
+exynosdrm-$(CONFIG_DRM_EXYNOS_IPP)	+= exynos_drm_ipp.o
 exynosdrm-$(CONFIG_DRM_EXYNOS_FIMC)	+= exynos_drm_fimc.o
 exynosdrm-$(CONFIG_DRM_EXYNOS_ROTATOR)	+= exynos_drm_rotator.o
 exynosdrm-$(CONFIG_DRM_EXYNOS_GSC)	+= exynos_drm_gsc.o
diff --git a/drivers/gpu/drm/exynos/exynos_drm_drv.c b/drivers/gpu/drm/exynos/exynos_drm_drv.c
index a518e9c..37c0db7 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_drv.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_drv.c
@@ -27,15 +27,23 @@
 #include "exynos_drm_fb.h"
 #include "exynos_drm_gem.h"
 #include "exynos_drm_plane.h"
+#include "exynos_drm_ipp.h"
 #include "exynos_drm_vidi.h"
 #include "exynos_drm_g2d.h"
 #include "exynos_drm_iommu.h"
 
 #define DRIVER_NAME	"exynos"
 #define DRIVER_DESC	"Samsung SoC DRM"
-#define DRIVER_DATE	"20110530"
+#define DRIVER_DATE	"20180330"
+
+/*
+ * Interface history:
+ *
+ * 1.0 - Original version
+ * 1.1 - Upgrade IPP driver to version 2.0
+ */
 #define DRIVER_MAJOR	1
-#define DRIVER_MINOR	0
+#define DRIVER_MINOR	1
 
 int exynos_atomic_check(struct drm_device *dev,
 			struct drm_atomic_state *state)
@@ -108,6 +116,16 @@ static const struct drm_ioctl_desc exynos_ioctls[] = {
 			DRM_AUTH | DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(EXYNOS_G2D_EXEC, exynos_g2d_exec_ioctl,
 			DRM_AUTH | DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(EXYNOS_IPP_GET_RESOURCES,
+			exynos_drm_ipp_get_res_ioctl,
+			DRM_AUTH | DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(EXYNOS_IPP_GET_CAPS, exynos_drm_ipp_get_caps_ioctl,
+			DRM_AUTH | DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(EXYNOS_IPP_GET_LIMITS,
+			exynos_drm_ipp_get_limits_ioctl,
+			DRM_AUTH | DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(EXYNOS_IPP_COMMIT, exynos_drm_ipp_commit_ioctl,
+			DRM_AUTH | DRM_RENDER_ALLOW),
 };
 
 static const struct file_operations exynos_drm_driver_fops = {
diff --git a/drivers/gpu/drm/exynos/exynos_drm_ipp.c b/drivers/gpu/drm/exynos/exynos_drm_ipp.c
new file mode 100644
index 0000000..26374e5
--- /dev/null
+++ b/drivers/gpu/drm/exynos/exynos_drm_ipp.c
@@ -0,0 +1,916 @@
+/*
+ * Copyright (C) 2017 Samsung Electronics Co.Ltd
+ * Authors:
+ *	Marek Szyprowski <m.szyprowski@samsung.com>
+ *
+ * Exynos DRM Image Post Processing (IPP) related functions
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ */
+
+
+#include <drm/drmP.h>
+#include <drm/drm_mode.h>
+#include <uapi/drm/exynos_drm.h>
+
+#include "exynos_drm_drv.h"
+#include "exynos_drm_gem.h"
+#include "exynos_drm_ipp.h"
+
+static int num_ipp;
+static LIST_HEAD(ipp_list);
+
+/**
+ * exynos_drm_ipp_register - Register a new picture processor hardware module
+ * @dev: DRM device
+ * @ipp: ipp module to init
+ * @funcs: callbacks for the new ipp object
+ * @caps: bitmask of ipp capabilities (%DRM_EXYNOS_IPP_CAP_*)
+ * @formats: array of supported formats
+ * @num_formats: size of the supported formats array
+ * @name: name (for debugging purposes)
+ *
+ * Initializes a ipp module.
+ *
+ * Returns:
+ * Zero on success, error code on failure.
+ */
+int exynos_drm_ipp_register(struct drm_device *dev, struct exynos_drm_ipp *ipp,
+		const struct exynos_drm_ipp_funcs *funcs, unsigned int caps,
+		const struct exynos_drm_ipp_formats *formats,
+		unsigned int num_formats, const char *name)
+{
+	WARN_ON(!ipp);
+	WARN_ON(!funcs);
+	WARN_ON(!formats);
+	WARN_ON(!num_formats);
+
+	spin_lock_init(&ipp->lock);
+	INIT_LIST_HEAD(&ipp->todo_list);
+	init_waitqueue_head(&ipp->done_wq);
+	ipp->dev = dev;
+	ipp->funcs = funcs;
+	ipp->capabilities = caps;
+	ipp->name = name;
+	ipp->formats = formats;
+	ipp->num_formats = num_formats;
+
+	/* ipp_list modification is serialized by component framework */
+	list_add_tail(&ipp->head, &ipp_list);
+	ipp->id = num_ipp++;
+
+	DRM_DEBUG_DRIVER("Registered ipp %d\n", ipp->id);
+
+	return 0;
+}
+
+/**
+ * exynos_drm_ipp_unregister - Unregister the picture processor module
+ * @dev: DRM device
+ * @ipp: ipp module
+ */
+void exynos_drm_ipp_unregister(struct drm_device *dev,
+			       struct exynos_drm_ipp *ipp)
+{
+	WARN_ON(ipp->task);
+	WARN_ON(!list_empty(&ipp->todo_list));
+	list_del(&ipp->head);
+}
+
+/**
+ * exynos_drm_ipp_ioctl_get_res_ioctl - enumerate all ipp modules
+ * @dev: DRM device
+ * @data: ioctl data
+ * @file_priv: DRM file info
+ *
+ * Construct a list of ipp ids.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, negative errno on failure.
+ */
+int exynos_drm_ipp_get_res_ioctl(struct drm_device *dev, void *data,
+				 struct drm_file *file_priv)
+{
+	struct drm_exynos_ioctl_ipp_get_res *resp = data;
+	struct exynos_drm_ipp *ipp;
+	uint32_t __user *ipp_ptr = (uint32_t __user *)
+						(unsigned long)resp->ipp_id_ptr;
+	unsigned int count = num_ipp, copied = 0;
+
+	/*
+	 * This ioctl is called twice, once to determine how much space is
+	 * needed, and the 2nd time to fill it.
+	 */
+	if (count && resp->count_ipps >= count) {
+		list_for_each_entry(ipp, &ipp_list, head) {
+			if (put_user(ipp->id, ipp_ptr + copied))
+				return -EFAULT;
+			copied++;
+		}
+	}
+	resp->count_ipps = count;
+
+	return 0;
+}
+
+static inline struct exynos_drm_ipp *__ipp_get(uint32_t id)
+{
+	struct exynos_drm_ipp *ipp;
+
+	list_for_each_entry(ipp, &ipp_list, head)
+		if (ipp->id == id)
+			return ipp;
+	return NULL;
+}
+
+/**
+ * exynos_drm_ipp_ioctl_get_caps - get ipp module capabilities and formats
+ * @dev: DRM device
+ * @data: ioctl data
+ * @file_priv: DRM file info
+ *
+ * Construct a structure describing ipp module capabilities.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, negative errno on failure.
+ */
+int exynos_drm_ipp_get_caps_ioctl(struct drm_device *dev, void *data,
+				  struct drm_file *file_priv)
+{
+	struct drm_exynos_ioctl_ipp_get_caps *resp = data;
+	void __user *ptr = (void __user *)(unsigned long)resp->formats_ptr;
+	struct exynos_drm_ipp *ipp;
+	int i;
+
+	ipp = __ipp_get(resp->ipp_id);
+	if (!ipp)
+		return -ENOENT;
+
+	resp->ipp_id = ipp->id;
+	resp->capabilities = ipp->capabilities;
+
+	/*
+	 * This ioctl is called twice, once to determine how much space is
+	 * needed, and the 2nd time to fill it.
+	 */
+	if (resp->formats_count >= ipp->num_formats) {
+		for (i = 0; i < ipp->num_formats; i++) {
+			struct drm_exynos_ipp_format tmp = {
+				.fourcc = ipp->formats[i].fourcc,
+				.type = ipp->formats[i].type,
+				.modifier = ipp->formats[i].modifier,
+			};
+
+			if (copy_to_user(ptr, &tmp, sizeof(tmp)))
+				return -EFAULT;
+			ptr += sizeof(tmp);
+		}
+	}
+	resp->formats_count = ipp->num_formats;
+
+	return 0;
+}
+
+static inline const struct exynos_drm_ipp_formats *__ipp_format_get(
+				struct exynos_drm_ipp *ipp, uint32_t fourcc,
+				uint64_t mod, unsigned int type)
+{
+	int i;
+
+	for (i = 0; i < ipp->num_formats; i++) {
+		if ((ipp->formats[i].type & type) &&
+		    ipp->formats[i].fourcc == fourcc &&
+		    ipp->formats[i].modifier == mod)
+			return &ipp->formats[i];
+	}
+	return NULL;
+}
+
+/**
+ * exynos_drm_ipp_get_limits_ioctl - get ipp module limits
+ * @dev: DRM device
+ * @data: ioctl data
+ * @file_priv: DRM file info
+ *
+ * Construct a structure describing ipp module limitations for provided
+ * picture format.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, negative errno on failure.
+ */
+int exynos_drm_ipp_get_limits_ioctl(struct drm_device *dev, void *data,
+				    struct drm_file *file_priv)
+{
+	struct drm_exynos_ioctl_ipp_get_limits *resp = data;
+	void __user *ptr = (void __user *)(unsigned long)resp->limits_ptr;
+	const struct exynos_drm_ipp_formats *format;
+	struct exynos_drm_ipp *ipp;
+
+	if (resp->type != DRM_EXYNOS_IPP_FORMAT_SOURCE &&
+	    resp->type != DRM_EXYNOS_IPP_FORMAT_DESTINATION)
+		return -EINVAL;
+
+	ipp = __ipp_get(resp->ipp_id);
+	if (!ipp)
+		return -ENOENT;
+
+	format = __ipp_format_get(ipp, resp->fourcc, resp->modifier,
+				  resp->type);
+	if (!format)
+		return -EINVAL;
+
+	/*
+	 * This ioctl is called twice, once to determine how much space is
+	 * needed, and the 2nd time to fill it.
+	 */
+	if (format->num_limits && resp->limits_count >= format->num_limits)
+		if (copy_to_user((void __user *)ptr, format->limits,
+				 sizeof(*format->limits) * format->num_limits))
+			return -EFAULT;
+	resp->limits_count = format->num_limits;
+
+	return 0;
+}
+
+struct drm_pending_exynos_ipp_event {
+	struct drm_pending_event base;
+	struct drm_exynos_ipp_event event;
+};
+
+static inline struct exynos_drm_ipp_task *
+			exynos_drm_ipp_task_alloc(struct exynos_drm_ipp *ipp)
+{
+	struct exynos_drm_ipp_task *task;
+
+	task = kzalloc(sizeof(*task), GFP_KERNEL);
+	if (!task)
+		return NULL;
+
+	task->dev = ipp->dev;
+	task->ipp = ipp;
+
+	/* some defaults */
+	task->src.rect.w = task->dst.rect.w = UINT_MAX;
+	task->src.rect.h = task->dst.rect.h = UINT_MAX;
+	task->transform.rotation = DRM_MODE_ROTATE_0;
+
+	DRM_DEBUG_DRIVER("Allocated task %pK\n", task);
+
+	return task;
+}
+
+static const struct exynos_drm_param_map {
+	unsigned int id;
+	unsigned int size;
+	unsigned int offset;
+} exynos_drm_ipp_params_maps[] = {
+	{
+		DRM_EXYNOS_IPP_TASK_BUFFER | DRM_EXYNOS_IPP_TASK_TYPE_SOURCE,
+		sizeof(struct drm_exynos_ipp_task_buffer),
+		offsetof(struct exynos_drm_ipp_task, src.buf),
+	}, {
+		DRM_EXYNOS_IPP_TASK_BUFFER |
+			DRM_EXYNOS_IPP_TASK_TYPE_DESTINATION,
+		sizeof(struct drm_exynos_ipp_task_buffer),
+		offsetof(struct exynos_drm_ipp_task, dst.buf),
+	}, {
+		DRM_EXYNOS_IPP_TASK_RECTANGLE | DRM_EXYNOS_IPP_TASK_TYPE_SOURCE,
+		sizeof(struct drm_exynos_ipp_task_rect),
+		offsetof(struct exynos_drm_ipp_task, src.rect),
+	}, {
+		DRM_EXYNOS_IPP_TASK_RECTANGLE |
+			DRM_EXYNOS_IPP_TASK_TYPE_DESTINATION,
+		sizeof(struct drm_exynos_ipp_task_rect),
+		offsetof(struct exynos_drm_ipp_task, dst.rect),
+	}, {
+		DRM_EXYNOS_IPP_TASK_TRANSFORM,
+		sizeof(struct drm_exynos_ipp_task_transform),
+		offsetof(struct exynos_drm_ipp_task, transform),
+	}, {
+		DRM_EXYNOS_IPP_TASK_ALPHA,
+		sizeof(struct drm_exynos_ipp_task_alpha),
+		offsetof(struct exynos_drm_ipp_task, alpha),
+	},
+};
+
+static int exynos_drm_ipp_task_set(struct exynos_drm_ipp_task *task,
+				   struct drm_exynos_ioctl_ipp_commit *arg)
+{
+	const struct exynos_drm_param_map *map = exynos_drm_ipp_params_maps;
+	void __user *params = (void __user *)(unsigned long)arg->params_ptr;
+	unsigned int size = arg->params_size;
+	uint32_t id;
+	int i;
+
+	while (size) {
+		if (get_user(id, (uint32_t __user *)params))
+			return -EFAULT;
+
+		for (i = 0; i < ARRAY_SIZE(exynos_drm_ipp_params_maps); i++)
+			if (map[i].id == id)
+				break;
+		if (i == ARRAY_SIZE(exynos_drm_ipp_params_maps) ||
+		    map[i].size > size)
+			return -EINVAL;
+
+		if (copy_from_user((void *)task + map[i].offset, params,
+				   map[i].size))
+			return -EFAULT;
+
+		params += map[i].size;
+		size -= map[i].size;
+	}
+
+	DRM_DEBUG_DRIVER("Got task %pK configuration from userspace\n", task);
+	return 0;
+}
+
+static int exynos_drm_ipp_task_setup_buffer(struct exynos_drm_ipp_buffer *buf,
+					    struct drm_file *filp)
+{
+	int ret = 0;
+	int i;
+
+	/* basic checks */
+	if (buf->buf.width == 0 || buf->buf.height == 0)
+		return -EINVAL;
+	buf->format = drm_format_info(buf->buf.fourcc);
+	for (i = 0; i < buf->format->num_planes; i++) {
+		unsigned int width = (i == 0) ? buf->buf.width :
+			     DIV_ROUND_UP(buf->buf.width, buf->format->hsub);
+
+		if (buf->buf.pitch[i] == 0)
+			buf->buf.pitch[i] = width * buf->format->cpp[i];
+		if (buf->buf.pitch[i] < width * buf->format->cpp[i])
+			return -EINVAL;
+		if (!buf->buf.gem_id[i])
+			return -ENOENT;
+	}
+
+	/* pitch for additional planes must match */
+	if (buf->format->num_planes > 2 &&
+	    buf->buf.pitch[1] != buf->buf.pitch[2])
+		return -EINVAL;
+
+	/* get GEM buffers and check their size */
+	for (i = 0; i < buf->format->num_planes; i++) {
+		unsigned int height = (i == 0) ? buf->buf.height :
+			     DIV_ROUND_UP(buf->buf.height, buf->format->vsub);
+		unsigned long size = height * buf->buf.pitch[i];
+		struct drm_gem_object *obj = drm_gem_object_lookup(filp,
+							    buf->buf.gem_id[i]);
+		if (!obj) {
+			ret = -ENOENT;
+			goto gem_free;
+		}
+		buf->exynos_gem[i] = to_exynos_gem(obj);
+
+		if (size + buf->buf.offset[i] > buf->exynos_gem[i]->size) {
+			i++;
+			ret = -EINVAL;
+			goto gem_free;
+		}
+		buf->dma_addr[i] = buf->exynos_gem[i]->dma_addr +
+				   buf->buf.offset[i];
+	}
+
+	return 0;
+gem_free:
+	while (i--) {
+		drm_gem_object_put_unlocked(&buf->exynos_gem[i]->base);
+		buf->exynos_gem[i] = NULL;
+	}
+	return ret;
+}
+
+static void exynos_drm_ipp_task_release_buf(struct exynos_drm_ipp_buffer *buf)
+{
+	int i;
+
+	if (!buf->exynos_gem[0])
+		return;
+	for (i = 0; i < buf->format->num_planes; i++)
+		drm_gem_object_put_unlocked(&buf->exynos_gem[i]->base);
+}
+
+static void exynos_drm_ipp_task_free(struct exynos_drm_ipp *ipp,
+				 struct exynos_drm_ipp_task *task)
+{
+	DRM_DEBUG_DRIVER("Freeing task %pK\n", task);
+
+	exynos_drm_ipp_task_release_buf(&task->src);
+	exynos_drm_ipp_task_release_buf(&task->dst);
+	if (task->event)
+		drm_event_cancel_free(ipp->dev, &task->event->base);
+	kfree(task);
+}
+
+struct drm_ipp_limit {
+	struct drm_exynos_ipp_limit_val h;
+	struct drm_exynos_ipp_limit_val v;
+};
+
+enum drm_ipp_size_id {
+	IPP_LIMIT_BUFFER, IPP_LIMIT_AREA, IPP_LIMIT_ROTATED, IPP_LIMIT_MAX
+};
+
+static const enum drm_ipp_size_id limit_id_fallback[IPP_LIMIT_MAX][4] = {
+	[IPP_LIMIT_BUFFER]  = { DRM_EXYNOS_IPP_LIMIT_SIZE_BUFFER },
+	[IPP_LIMIT_AREA]    = { DRM_EXYNOS_IPP_LIMIT_SIZE_AREA,
+				DRM_EXYNOS_IPP_LIMIT_SIZE_BUFFER },
+	[IPP_LIMIT_ROTATED] = { DRM_EXYNOS_IPP_LIMIT_SIZE_ROTATED,
+				DRM_EXYNOS_IPP_LIMIT_SIZE_AREA,
+				DRM_EXYNOS_IPP_LIMIT_SIZE_BUFFER },
+};
+
+static inline void __limit_set_val(unsigned int *ptr, unsigned int val)
+{
+	if (!*ptr)
+		*ptr = val;
+}
+
+static void __get_size_limit(const struct drm_exynos_ipp_limit *limits,
+			     unsigned int num_limits, enum drm_ipp_size_id id,
+			     struct drm_ipp_limit *res)
+{
+	const struct drm_exynos_ipp_limit *l = limits;
+	int i = 0;
+
+	memset(res, 0, sizeof(*res));
+	for (i = 0; limit_id_fallback[id][i]; i++)
+		for (l = limits; l - limits < num_limits; l++) {
+			if (((l->type & DRM_EXYNOS_IPP_LIMIT_TYPE_MASK) !=
+			      DRM_EXYNOS_IPP_LIMIT_TYPE_SIZE) ||
+			    ((l->type & DRM_EXYNOS_IPP_LIMIT_SIZE_MASK) !=
+						     limit_id_fallback[id][i]))
+				continue;
+			__limit_set_val(&res->h.min, l->h.min);
+			__limit_set_val(&res->h.max, l->h.max);
+			__limit_set_val(&res->h.align, l->h.align);
+			__limit_set_val(&res->v.min, l->v.min);
+			__limit_set_val(&res->v.max, l->v.max);
+			__limit_set_val(&res->v.align, l->v.align);
+		}
+}
+
+static inline bool __align_check(unsigned int val, unsigned int align)
+{
+	if (align && (val & (align - 1))) {
+		DRM_DEBUG_DRIVER("Value %d exceeds HW limits (align %d)\n",
+				 val, align);
+		return false;
+	}
+	return true;
+}
+
+static inline bool __size_limit_check(unsigned int val,
+				 struct drm_exynos_ipp_limit_val *l)
+{
+	if ((l->min && val < l->min) || (l->max && val > l->max)) {
+		DRM_DEBUG_DRIVER("Value %d exceeds HW limits (min %d, max %d)\n",
+				 val, l->min, l->max);
+		return false;
+	}
+	return __align_check(val, l->align);
+}
+
+static int exynos_drm_ipp_check_size_limits(struct exynos_drm_ipp_buffer *buf,
+	const struct drm_exynos_ipp_limit *limits, unsigned int num_limits,
+	bool rotate, bool swap)
+{
+	enum drm_ipp_size_id id = rotate ? IPP_LIMIT_ROTATED : IPP_LIMIT_AREA;
+	struct drm_ipp_limit l;
+	struct drm_exynos_ipp_limit_val *lh = &l.h, *lv = &l.v;
+
+	if (!limits)
+		return 0;
+
+	__get_size_limit(limits, num_limits, IPP_LIMIT_BUFFER, &l);
+	if (!__size_limit_check(buf->buf.width, &l.h) ||
+	    !__size_limit_check(buf->buf.height, &l.v))
+		return -EINVAL;
+
+	if (swap) {
+		lv = &l.h;
+		lh = &l.v;
+	}
+	__get_size_limit(limits, num_limits, id, &l);
+	if (!__size_limit_check(buf->rect.w, lh) ||
+	    !__align_check(buf->rect.x, lh->align) ||
+	    !__size_limit_check(buf->rect.h, lv) ||
+	    !__align_check(buf->rect.y, lv->align))
+		return -EINVAL;
+
+	return 0;
+}
+
+static inline bool __scale_limit_check(unsigned int src, unsigned int dst,
+				       unsigned int min, unsigned int max)
+{
+	if ((max && (dst << 16) > src * max) ||
+	    (min && (dst << 16) < src * min)) {
+		DRM_DEBUG_DRIVER("Scale from %d to %d exceeds HW limits (ratio min %d.%05d, max %d.%05d)\n",
+			 src, dst,
+			 min >> 16, 100000 * (min & 0xffff) / (1 << 16),
+			 max >> 16, 100000 * (max & 0xffff) / (1 << 16));
+		return false;
+	}
+	return true;
+}
+
+static int exynos_drm_ipp_check_scale_limits(
+				struct drm_exynos_ipp_task_rect *src,
+				struct drm_exynos_ipp_task_rect *dst,
+				const struct drm_exynos_ipp_limit *limits,
+				unsigned int num_limits, bool swap)
+{
+	const struct drm_exynos_ipp_limit_val *lh, *lv;
+	int dw, dh;
+
+	for (; num_limits; limits++, num_limits--)
+		if ((limits->type & DRM_EXYNOS_IPP_LIMIT_TYPE_MASK) ==
+		    DRM_EXYNOS_IPP_LIMIT_TYPE_SCALE)
+			break;
+	if (!num_limits)
+		return 0;
+
+	lh = (!swap) ? &limits->h : &limits->v;
+	lv = (!swap) ? &limits->v : &limits->h;
+	dw = (!swap) ? dst->w : dst->h;
+	dh = (!swap) ? dst->h : dst->w;
+
+	if (!__scale_limit_check(src->w, dw, lh->min, lh->max) ||
+	    !__scale_limit_check(src->h, dh, lv->min, lv->max))
+		return -EINVAL;
+
+	return 0;
+}
+
+static int exynos_drm_ipp_task_check(struct exynos_drm_ipp_task *task)
+{
+	struct exynos_drm_ipp *ipp = task->ipp;
+	const struct exynos_drm_ipp_formats *src_fmt, *dst_fmt;
+	struct exynos_drm_ipp_buffer *src = &task->src, *dst = &task->dst;
+	unsigned int rotation = task->transform.rotation;
+	int ret = 0;
+	bool swap = drm_rotation_90_or_270(rotation);
+	bool rotate = (rotation != DRM_MODE_ROTATE_0);
+	bool scale = false;
+
+	DRM_DEBUG_DRIVER("Checking task %pK\n", task);
+
+	if (src->rect.w == UINT_MAX)
+		src->rect.w = src->buf.width;
+	if (src->rect.h == UINT_MAX)
+		src->rect.h = src->buf.height;
+	if (dst->rect.w == UINT_MAX)
+		dst->rect.w = dst->buf.width;
+	if (dst->rect.h == UINT_MAX)
+		dst->rect.h = dst->buf.height;
+
+	if (src->rect.x + src->rect.w > (src->buf.width) ||
+	    src->rect.y + src->rect.h > (src->buf.height) ||
+	    dst->rect.x + dst->rect.w > (dst->buf.width) ||
+	    dst->rect.y + dst->rect.h > (dst->buf.height)) {
+		DRM_DEBUG_DRIVER("Task %pK: defined area is outside provided buffers\n",
+				 task);
+		return -EINVAL;
+	}
+
+	if ((!swap && (src->rect.w != dst->rect.w ||
+		       src->rect.h != dst->rect.h)) ||
+	    (swap && (src->rect.w != dst->rect.h ||
+		      src->rect.h != dst->rect.w)))
+		scale = true;
+
+	if ((!(ipp->capabilities & DRM_EXYNOS_IPP_CAP_CROP) &&
+	     (src->rect.x || src->rect.y || dst->rect.x || dst->rect.y)) ||
+	    (!(ipp->capabilities & DRM_EXYNOS_IPP_CAP_ROTATE) && rotate) ||
+	    (!(ipp->capabilities & DRM_EXYNOS_IPP_CAP_SCALE) && scale) ||
+	    (!(ipp->capabilities & DRM_EXYNOS_IPP_CAP_CONVERT) &&
+	     src->buf.fourcc != dst->buf.fourcc)) {
+		DRM_DEBUG_DRIVER("Task %pK: hw capabilities exceeded\n", task);
+		return -EINVAL;
+	}
+
+	src_fmt = __ipp_format_get(ipp, src->buf.fourcc, src->buf.modifier,
+				   DRM_EXYNOS_IPP_FORMAT_SOURCE);
+	if (!src_fmt) {
+		DRM_DEBUG_DRIVER("Task %pK: src format not supported\n", task);
+		return -EINVAL;
+	}
+	ret = exynos_drm_ipp_check_size_limits(src, src_fmt->limits,
+					       src_fmt->num_limits,
+					       rotate, false);
+	if (ret)
+		return ret;
+	ret = exynos_drm_ipp_check_scale_limits(&src->rect, &dst->rect,
+						src_fmt->limits,
+						src_fmt->num_limits, swap);
+	if (ret)
+		return ret;
+
+	dst_fmt = __ipp_format_get(ipp, dst->buf.fourcc, dst->buf.modifier,
+				   DRM_EXYNOS_IPP_FORMAT_DESTINATION);
+	if (!dst_fmt) {
+		DRM_DEBUG_DRIVER("Task %pK: dst format not supported\n", task);
+		return -EINVAL;
+	}
+	ret = exynos_drm_ipp_check_size_limits(dst, dst_fmt->limits,
+					       dst_fmt->num_limits,
+					       false, swap);
+	if (ret)
+		return ret;
+	ret = exynos_drm_ipp_check_scale_limits(&src->rect, &dst->rect,
+						dst_fmt->limits,
+						dst_fmt->num_limits, swap);
+	if (ret)
+		return ret;
+
+	DRM_DEBUG_DRIVER("Task %pK: all checks done.\n", task);
+
+	return ret;
+}
+
+static int exynos_drm_ipp_task_setup_buffers(struct exynos_drm_ipp_task *task,
+				     struct drm_file *filp)
+{
+	struct exynos_drm_ipp_buffer *src = &task->src, *dst = &task->dst;
+	int ret = 0;
+
+	DRM_DEBUG_DRIVER("Setting buffer for task %pK\n", task);
+
+	ret = exynos_drm_ipp_task_setup_buffer(src, filp);
+	if (ret) {
+		DRM_DEBUG_DRIVER("Task %pK: src buffer setup failed\n", task);
+		return ret;
+	}
+	ret = exynos_drm_ipp_task_setup_buffer(dst, filp);
+	if (ret) {
+		DRM_DEBUG_DRIVER("Task %pK: dst buffer setup failed\n", task);
+		return ret;
+	}
+
+	DRM_DEBUG_DRIVER("Task %pK: buffers prepared.\n", task);
+
+	return ret;
+}
+
+
+static int exynos_drm_ipp_event_create(struct exynos_drm_ipp_task *task,
+				 struct drm_file *file_priv, uint64_t user_data)
+{
+	struct drm_pending_exynos_ipp_event *e = NULL;
+	int ret;
+
+	e = kzalloc(sizeof(*e), GFP_KERNEL);
+	if (!e)
+		return -ENOMEM;
+
+	e->event.base.type = DRM_EXYNOS_IPP_EVENT;
+	e->event.base.length = sizeof(e->event);
+	e->event.user_data = user_data;
+
+	ret = drm_event_reserve_init(task->dev, file_priv, &e->base,
+				     &e->event.base);
+	if (ret)
+		goto free;
+
+	task->event = e;
+	return 0;
+free:
+	kfree(e);
+	return ret;
+}
+
+static void exynos_drm_ipp_event_send(struct exynos_drm_ipp_task *task)
+{
+	struct timespec64 now;
+
+	ktime_get_ts64(&now);
+	task->event->event.tv_sec = now.tv_sec;
+	task->event->event.tv_usec = now.tv_nsec / NSEC_PER_USEC;
+	task->event->event.sequence = atomic_inc_return(&task->ipp->sequence);
+
+	drm_send_event(task->dev, &task->event->base);
+}
+
+static int exynos_drm_ipp_task_cleanup(struct exynos_drm_ipp_task *task)
+{
+	int ret = task->ret;
+
+	if (ret == 0 && task->event) {
+		exynos_drm_ipp_event_send(task);
+		/* ensure event won't be canceled on task free */
+		task->event = NULL;
+	}
+
+	exynos_drm_ipp_task_free(task->ipp, task);
+	return ret;
+}
+
+static void exynos_drm_ipp_cleanup_work(struct work_struct *work)
+{
+	struct exynos_drm_ipp_task *task = container_of(work,
+				      struct exynos_drm_ipp_task, cleanup_work);
+
+	exynos_drm_ipp_task_cleanup(task);
+}
+
+static void exynos_drm_ipp_next_task(struct exynos_drm_ipp *ipp);
+
+/**
+ * exynos_drm_ipp_task_done - finish given task and set return code
+ * @task: ipp task to finish
+ * @ret: error code or 0 if operation has been performed successfully
+ */
+void exynos_drm_ipp_task_done(struct exynos_drm_ipp_task *task, int ret)
+{
+	struct exynos_drm_ipp *ipp = task->ipp;
+	unsigned long flags;
+
+	DRM_DEBUG_DRIVER("ipp: %d, task %pK done: %d\n", ipp->id, task, ret);
+
+	spin_lock_irqsave(&ipp->lock, flags);
+	if (ipp->task == task)
+		ipp->task = NULL;
+	task->flags |= DRM_EXYNOS_IPP_TASK_DONE;
+	task->ret = ret;
+	spin_unlock_irqrestore(&ipp->lock, flags);
+
+	exynos_drm_ipp_next_task(ipp);
+	wake_up(&ipp->done_wq);
+
+	if (task->flags & DRM_EXYNOS_IPP_TASK_ASYNC) {
+		INIT_WORK(&task->cleanup_work, exynos_drm_ipp_cleanup_work);
+		schedule_work(&task->cleanup_work);
+	}
+}
+
+static void exynos_drm_ipp_next_task(struct exynos_drm_ipp *ipp)
+{
+	struct exynos_drm_ipp_task *task;
+	unsigned long flags;
+	int ret;
+
+	DRM_DEBUG_DRIVER("ipp: %d, try to run new task\n", ipp->id);
+
+	spin_lock_irqsave(&ipp->lock, flags);
+
+	if (ipp->task || list_empty(&ipp->todo_list)) {
+		spin_unlock_irqrestore(&ipp->lock, flags);
+		return;
+	}
+
+	task = list_first_entry(&ipp->todo_list, struct exynos_drm_ipp_task,
+				head);
+	list_del_init(&task->head);
+	ipp->task = task;
+
+	spin_unlock_irqrestore(&ipp->lock, flags);
+
+	DRM_DEBUG_DRIVER("ipp: %d, selected task %pK to run\n", ipp->id, task);
+
+	ret = ipp->funcs->commit(ipp, task);
+	if (ret)
+		exynos_drm_ipp_task_done(task, ret);
+}
+
+static void exynos_drm_ipp_schedule_task(struct exynos_drm_ipp *ipp,
+					 struct exynos_drm_ipp_task *task)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&ipp->lock, flags);
+	list_add(&task->head, &ipp->todo_list);
+	spin_unlock_irqrestore(&ipp->lock, flags);
+
+	exynos_drm_ipp_next_task(ipp);
+}
+
+static void exynos_drm_ipp_task_abort(struct exynos_drm_ipp *ipp,
+				      struct exynos_drm_ipp_task *task)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&ipp->lock, flags);
+	if (task->flags & DRM_EXYNOS_IPP_TASK_DONE) {
+		/* already completed task */
+		exynos_drm_ipp_task_cleanup(task);
+	} else if (ipp->task != task) {
+		/* task has not been scheduled for execution yet */
+		list_del_init(&task->head);
+		exynos_drm_ipp_task_cleanup(task);
+	} else {
+		/*
+		 * currently processed task, call abort() and perform
+		 * cleanup with async worker
+		 */
+		task->flags |= DRM_EXYNOS_IPP_TASK_ASYNC;
+		spin_unlock_irqrestore(&ipp->lock, flags);
+		if (ipp->funcs->abort)
+			ipp->funcs->abort(ipp, task);
+		return;
+	}
+	spin_unlock_irqrestore(&ipp->lock, flags);
+}
+
+/**
+ * exynos_drm_ipp_commit_ioctl - perform image processing operation
+ * @dev: DRM device
+ * @data: ioctl data
+ * @file_priv: DRM file info
+ *
+ * Construct a ipp task from the set of properties provided from the user
+ * and try to schedule it to framebuffer processor hardware.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, negative errno on failure.
+ */
+int exynos_drm_ipp_commit_ioctl(struct drm_device *dev, void *data,
+				struct drm_file *file_priv)
+{
+	struct drm_exynos_ioctl_ipp_commit *arg = data;
+	struct exynos_drm_ipp *ipp;
+	struct exynos_drm_ipp_task *task;
+	int ret = 0;
+
+	if ((arg->flags & ~DRM_EXYNOS_IPP_FLAGS) || arg->reserved)
+		return -EINVAL;
+
+	/* can't test and expect an event at the same time */
+	if ((arg->flags & DRM_EXYNOS_IPP_FLAG_TEST_ONLY) &&
+			(arg->flags & DRM_EXYNOS_IPP_FLAG_EVENT))
+		return -EINVAL;
+
+	ipp = __ipp_get(arg->ipp_id);
+	if (!ipp)
+		return -ENOENT;
+
+	task = exynos_drm_ipp_task_alloc(ipp);
+	if (!task)
+		return -ENOMEM;
+
+	ret = exynos_drm_ipp_task_set(task, arg);
+	if (ret)
+		goto free;
+
+	ret = exynos_drm_ipp_task_check(task);
+	if (ret)
+		goto free;
+
+	ret = exynos_drm_ipp_task_setup_buffers(task, file_priv);
+	if (ret || arg->flags & DRM_EXYNOS_IPP_FLAG_TEST_ONLY)
+		goto free;
+
+	if (arg->flags & DRM_EXYNOS_IPP_FLAG_EVENT) {
+		ret = exynos_drm_ipp_event_create(task, file_priv,
+						 arg->user_data);
+		if (ret)
+			goto free;
+	}
+
+	/*
+	 * Queue task for processing on the hardware. task object will be
+	 * then freed after exynos_drm_ipp_task_done()
+	 */
+	if (arg->flags & DRM_EXYNOS_IPP_FLAG_NONBLOCK) {
+		DRM_DEBUG_DRIVER("ipp: %d, nonblocking processing task %pK\n",
+				 ipp->id, task);
+
+		task->flags |= DRM_EXYNOS_IPP_TASK_ASYNC;
+		exynos_drm_ipp_schedule_task(task->ipp, task);
+		ret = 0;
+	} else {
+		DRM_DEBUG_DRIVER("ipp: %d, processing task %pK\n", ipp->id,
+				 task);
+		exynos_drm_ipp_schedule_task(ipp, task);
+		ret = wait_event_interruptible(ipp->done_wq,
+					task->flags & DRM_EXYNOS_IPP_TASK_DONE);
+		if (ret)
+			exynos_drm_ipp_task_abort(ipp, task);
+		else
+			ret = exynos_drm_ipp_task_cleanup(task);
+	}
+	return ret;
+free:
+	exynos_drm_ipp_task_free(ipp, task);
+
+	return ret;
+}
diff --git a/drivers/gpu/drm/exynos/exynos_drm_ipp.h b/drivers/gpu/drm/exynos/exynos_drm_ipp.h
new file mode 100644
index 0000000..0b27d4a
--- /dev/null
+++ b/drivers/gpu/drm/exynos/exynos_drm_ipp.h
@@ -0,0 +1,175 @@
+/*
+ * Copyright (c) 2017 Samsung Electronics Co., Ltd.
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ */
+
+#ifndef _EXYNOS_DRM_IPP_H_
+#define _EXYNOS_DRM_IPP_H_
+
+#include <drm/drmP.h>
+
+struct exynos_drm_ipp;
+struct exynos_drm_ipp_task;
+
+/**
+ * struct exynos_drm_ipp_funcs - exynos_drm_ipp control functions
+ */
+struct exynos_drm_ipp_funcs {
+	/**
+	 * @commit:
+	 *
+	 * This is the main entry point to start framebuffer processing
+	 * in the hardware. The exynos_drm_ipp_task has been already validated.
+	 * This function must not wait until the device finishes processing.
+	 * When the driver finishes processing, it has to call
+	 * exynos_exynos_drm_ipp_task_done() function.
+	 *
+	 * RETURNS:
+	 *
+	 * 0 on success or negative error codes in case of failure.
+	 */
+	int (*commit)(struct exynos_drm_ipp *ipp,
+		      struct exynos_drm_ipp_task *task);
+
+	/**
+	 * @abort:
+	 *
+	 * Informs the driver that it has to abort the currently running
+	 * task as soon as possible (i.e. as soon as it can stop the device
+	 * safely), even if the task would not have been finished by then.
+	 * After the driver performs the necessary steps, it has to call
+	 * exynos_drm_ipp_task_done() (as if the task ended normally).
+	 * This function does not have to (and will usually not) wait
+	 * until the device enters a state when it can be stopped.
+	 */
+	void (*abort)(struct exynos_drm_ipp *ipp,
+		      struct exynos_drm_ipp_task *task);
+};
+
+/**
+ * struct exynos_drm_ipp - central picture processor module structure
+ */
+struct exynos_drm_ipp {
+	struct drm_device *dev;
+	struct list_head head;
+	unsigned int id;
+
+	const char *name;
+	const struct exynos_drm_ipp_funcs *funcs;
+	unsigned int capabilities;
+	const struct exynos_drm_ipp_formats *formats;
+	unsigned int num_formats;
+	atomic_t sequence;
+
+	spinlock_t lock;
+	struct exynos_drm_ipp_task *task;
+	struct list_head todo_list;
+	wait_queue_head_t done_wq;
+};
+
+struct exynos_drm_ipp_buffer {
+	struct drm_exynos_ipp_task_buffer buf;
+	struct drm_exynos_ipp_task_rect rect;
+
+	struct exynos_drm_gem *exynos_gem[MAX_FB_BUFFER];
+	const struct drm_format_info *format;
+	dma_addr_t dma_addr[MAX_FB_BUFFER];
+};
+
+/**
+ * struct exynos_drm_ipp_task - a structure describing transformation that
+ * has to be performed by the picture processor hardware module
+ */
+struct exynos_drm_ipp_task {
+	struct drm_device *dev;
+	struct exynos_drm_ipp *ipp;
+	struct list_head head;
+
+	struct exynos_drm_ipp_buffer src;
+	struct exynos_drm_ipp_buffer dst;
+
+	struct drm_exynos_ipp_task_transform transform;
+	struct drm_exynos_ipp_task_alpha alpha;
+
+	struct work_struct cleanup_work;
+	unsigned int flags;
+	int ret;
+
+	struct drm_pending_exynos_ipp_event *event;
+};
+
+#define DRM_EXYNOS_IPP_TASK_DONE	(1 << 0)
+#define DRM_EXYNOS_IPP_TASK_ASYNC	(1 << 1)
+
+struct exynos_drm_ipp_formats {
+	uint32_t fourcc;
+	uint32_t type;
+	uint64_t modifier;
+	const struct drm_exynos_ipp_limit *limits;
+	unsigned int num_limits;
+};
+
+/* helper macros to set exynos_drm_ipp_formats structure and limits*/
+#define IPP_SRCDST_MFORMAT(f, m, l) \
+	.fourcc = DRM_FORMAT_##f, .modifier = m, .limits = l, \
+	.num_limits = ARRAY_SIZE(l), \
+	.type = (DRM_EXYNOS_IPP_FORMAT_SOURCE | \
+		 DRM_EXYNOS_IPP_FORMAT_DESTINATION)
+
+#define IPP_SRCDST_FORMAT(f, l) IPP_SRCDST_MFORMAT(f, 0, l)
+
+#define IPP_SIZE_LIMIT(l, val...)	\
+	.type = (DRM_EXYNOS_IPP_LIMIT_TYPE_SIZE | \
+		 DRM_EXYNOS_IPP_LIMIT_SIZE_##l), val
+
+#define IPP_SCALE_LIMIT(val...)		\
+	.type = (DRM_EXYNOS_IPP_LIMIT_TYPE_SCALE), val
+
+int exynos_drm_ipp_register(struct drm_device *dev, struct exynos_drm_ipp *ipp,
+		const struct exynos_drm_ipp_funcs *funcs, unsigned int caps,
+		const struct exynos_drm_ipp_formats *formats,
+		unsigned int num_formats, const char *name);
+void exynos_drm_ipp_unregister(struct drm_device *dev,
+			       struct exynos_drm_ipp *ipp);
+
+void exynos_drm_ipp_task_done(struct exynos_drm_ipp_task *task, int ret);
+
+#ifdef CONFIG_DRM_EXYNOS_IPP
+int exynos_drm_ipp_get_res_ioctl(struct drm_device *dev, void *data,
+				 struct drm_file *file_priv);
+int exynos_drm_ipp_get_caps_ioctl(struct drm_device *dev, void *data,
+				  struct drm_file *file_priv);
+int exynos_drm_ipp_get_limits_ioctl(struct drm_device *dev, void *data,
+				    struct drm_file *file_priv);
+int exynos_drm_ipp_commit_ioctl(struct drm_device *dev,
+				void *data, struct drm_file *file_priv);
+#else
+static inline int exynos_drm_ipp_get_res_ioctl(struct drm_device *dev,
+	 void *data, struct drm_file *file_priv)
+{
+	struct drm_exynos_ioctl_ipp_get_res *resp = data;
+
+	resp->count_ipps = 0;
+	return 0;
+}
+static inline int exynos_drm_ipp_get_caps_ioctl(struct drm_device *dev,
+	 void *data, struct drm_file *file_priv)
+{
+	return -ENODEV;
+}
+static inline int exynos_drm_ipp_get_limits_ioctl(struct drm_device *dev,
+	 void *data, struct drm_file *file_priv)
+{
+	return -ENODEV;
+}
+static inline int exynos_drm_ipp_commit_ioctl(struct drm_device *dev,
+	 void *data, struct drm_file *file_priv)
+{
+	return -ENODEV;
+}
+#endif
+#endif
diff --git a/include/uapi/drm/exynos_drm.h b/include/uapi/drm/exynos_drm.h
index 4a54305..3e59b83 100644
--- a/include/uapi/drm/exynos_drm.h
+++ b/include/uapi/drm/exynos_drm.h
@@ -135,6 +135,219 @@ struct drm_exynos_g2d_exec {
 	__u64					async;
 };
 
+/* Exynos DRM IPP v2 API */
+
+/**
+ * Enumerate available IPP hardware modules.
+ *
+ * @count_ipps: size of ipp_id array / number of ipp modules (set by driver)
+ * @reserved: padding
+ * @ipp_id_ptr: pointer to ipp_id array or NULL
+ */
+struct drm_exynos_ioctl_ipp_get_res {
+	__u32 count_ipps;
+	__u32 reserved;
+	__u64 ipp_id_ptr;
+};
+
+enum drm_exynos_ipp_format_type {
+	DRM_EXYNOS_IPP_FORMAT_SOURCE		= 0x01,
+	DRM_EXYNOS_IPP_FORMAT_DESTINATION	= 0x02,
+};
+
+struct drm_exynos_ipp_format {
+	__u32 fourcc;
+	__u32 type;
+	__u64 modifier;
+};
+
+enum drm_exynos_ipp_capability {
+	DRM_EXYNOS_IPP_CAP_CROP		= 0x01,
+	DRM_EXYNOS_IPP_CAP_ROTATE	= 0x02,
+	DRM_EXYNOS_IPP_CAP_SCALE	= 0x04,
+	DRM_EXYNOS_IPP_CAP_CONVERT	= 0x08,
+};
+
+/**
+ * Get IPP hardware capabilities and supported image formats.
+ *
+ * @ipp_id: id of IPP module to query
+ * @capabilities: bitmask of drm_exynos_ipp_capability (set by driver)
+ * @reserved: padding
+ * @formats_count: size of formats array (in entries) / number of filled
+ *		   formats (set by driver)
+ * @formats_ptr: pointer to formats array or NULL
+ */
+struct drm_exynos_ioctl_ipp_get_caps {
+	__u32 ipp_id;
+	__u32 capabilities;
+	__u32 reserved;
+	__u32 formats_count;
+	__u64 formats_ptr;
+};
+
+enum drm_exynos_ipp_limit_type {
+	/* size (horizontal/vertial) limits, in pixels (min, max, alignment) */
+	DRM_EXYNOS_IPP_LIMIT_TYPE_SIZE		= 0x0001,
+	/* scale ratio (horizonta/vertial), 16.16 fixed point (min, max) */
+	DRM_EXYNOS_IPP_LIMIT_TYPE_SCALE		= 0x0002,
+
+	/* image buffer area */
+	DRM_EXYNOS_IPP_LIMIT_SIZE_BUFFER	= 0x0001 << 16,
+	/* src/dst rectangle area */
+	DRM_EXYNOS_IPP_LIMIT_SIZE_AREA		= 0x0002 << 16,
+	/* src/dst rectangle area when rotation enabled */
+	DRM_EXYNOS_IPP_LIMIT_SIZE_ROTATED	= 0x0003 << 16,
+
+	DRM_EXYNOS_IPP_LIMIT_TYPE_MASK		= 0x000f,
+	DRM_EXYNOS_IPP_LIMIT_SIZE_MASK		= 0x000f << 16,
+};
+
+struct drm_exynos_ipp_limit_val {
+	__u32 min;
+	__u32 max;
+	__u32 align;
+	__u32 reserved;
+};
+
+/**
+ * IPP module limitation.
+ *
+ * @type: limit type (see drm_exynos_ipp_limit_type enum)
+ * @reserved: padding
+ * @h: horizontal limits
+ * @v: vertical limits
+ */
+struct drm_exynos_ipp_limit {
+	__u32 type;
+	__u32 reserved;
+	struct drm_exynos_ipp_limit_val h;
+	struct drm_exynos_ipp_limit_val v;
+};
+
+/**
+ * Get IPP limits for given image format.
+ *
+ * @ipp_id: id of IPP module to query
+ * @fourcc: image format code (see DRM_FORMAT_* in drm_fourcc.h)
+ * @modifier: image format modifier (see DRM_FORMAT_MOD_* in drm_fourcc.h)
+ * @type: source/destination identifier (drm_exynos_ipp_format_flag enum)
+ * @limits_count: size of limits array (in entries) / number of filled entries
+ *		 (set by driver)
+ * @limits_ptr: pointer to limits array or NULL
+ */
+struct drm_exynos_ioctl_ipp_get_limits {
+	__u32 ipp_id;
+	__u32 fourcc;
+	__u64 modifier;
+	__u32 type;
+	__u32 limits_count;
+	__u64 limits_ptr;
+};
+
+enum drm_exynos_ipp_task_id {
+	/* buffer described by struct drm_exynos_ipp_task_buffer */
+	DRM_EXYNOS_IPP_TASK_BUFFER		= 0x0001,
+	/* rectangle described by struct drm_exynos_ipp_task_rect */
+	DRM_EXYNOS_IPP_TASK_RECTANGLE		= 0x0002,
+	/* transformation described by struct drm_exynos_ipp_task_transform */
+	DRM_EXYNOS_IPP_TASK_TRANSFORM		= 0x0003,
+	/* alpha configuration described by struct drm_exynos_ipp_task_alpha */
+	DRM_EXYNOS_IPP_TASK_ALPHA		= 0x0004,
+
+	/* source image data (for buffer and rectangle chunks) */
+	DRM_EXYNOS_IPP_TASK_TYPE_SOURCE		= 0x0001 << 16,
+	/* destination image data (for buffer and rectangle chunks) */
+	DRM_EXYNOS_IPP_TASK_TYPE_DESTINATION	= 0x0002 << 16,
+};
+
+/**
+ * Memory buffer with image data.
+ *
+ * @id: must be DRM_EXYNOS_IPP_TASK_BUFFER
+ * other parameters are same as for AddFB2 generic DRM ioctl
+ */
+struct drm_exynos_ipp_task_buffer {
+	__u32	id;
+	__u32	fourcc;
+	__u32	width, height;
+	__u32	gem_id[4];
+	__u32	offset[4];
+	__u32	pitch[4];
+	__u64	modifier;
+};
+
+/**
+ * Rectangle for processing.
+ *
+ * @id: must be DRM_EXYNOS_IPP_TASK_RECTANGLE
+ * @reserved: padding
+ * @x,@y: left corner in pixels
+ * @w,@h: width/height in pixels
+ */
+struct drm_exynos_ipp_task_rect {
+	__u32	id;
+	__u32	reserved;
+	__u32	x;
+	__u32	y;
+	__u32	w;
+	__u32	h;
+};
+
+/**
+ * Image tranformation description.
+ *
+ * @id: must be DRM_EXYNOS_IPP_TASK_TRANSFORM
+ * @rotation: DRM_MODE_ROTATE_* and DRM_MODE_REFLECT_* values
+ */
+struct drm_exynos_ipp_task_transform {
+	__u32	id;
+	__u32	rotation;
+};
+
+/**
+ * Image global alpha configuration for formats without alpha values.
+ *
+ * @id: must be DRM_EXYNOS_IPP_TASK_ALPHA
+ * @value: global alpha value (0-255)
+ */
+struct drm_exynos_ipp_task_alpha {
+	__u32	id;
+	__u32	value;
+};
+
+enum drm_exynos_ipp_flag {
+	/* generate DRM event after processing */
+	DRM_EXYNOS_IPP_FLAG_EVENT	= 0x01,
+	/* dry run, only check task parameters */
+	DRM_EXYNOS_IPP_FLAG_TEST_ONLY	= 0x02,
+	/* non-blocking processing */
+	DRM_EXYNOS_IPP_FLAG_NONBLOCK	= 0x04,
+};
+
+#define DRM_EXYNOS_IPP_FLAGS (DRM_EXYNOS_IPP_FLAG_EVENT |\
+		DRM_EXYNOS_IPP_FLAG_TEST_ONLY | DRM_EXYNOS_IPP_FLAG_NONBLOCK)
+
+/**
+ * Perform image processing described by array of drm_exynos_ipp_task_*
+ * structures (parameters array).
+ *
+ * @ipp_id: id of IPP module to run the task
+ * @flags: bitmask of drm_exynos_ipp_flag values
+ * @reserved: padding
+ * @params_size: size of parameters array (in bytes)
+ * @params_ptr: pointer to parameters array or NULL
+ * @user_data: (optional) data for drm event
+ */
+struct drm_exynos_ioctl_ipp_commit {
+	__u32 ipp_id;
+	__u32 flags;
+	__u32 reserved;
+	__u32 params_size;
+	__u64 params_ptr;
+	__u64 user_data;
+};
+
 #define DRM_EXYNOS_GEM_CREATE		0x00
 #define DRM_EXYNOS_GEM_MAP		0x01
 /* Reserved 0x03 ~ 0x05 for exynos specific gem ioctl */
@@ -147,6 +360,11 @@ struct drm_exynos_g2d_exec {
 #define DRM_EXYNOS_G2D_EXEC		0x22
 
 /* Reserved 0x30 ~ 0x33 for obsolete Exynos IPP ioctls */
+/* IPP - Image Post Processing */
+#define DRM_EXYNOS_IPP_GET_RESOURCES	0x40
+#define DRM_EXYNOS_IPP_GET_CAPS		0x41
+#define DRM_EXYNOS_IPP_GET_LIMITS	0x42
+#define DRM_EXYNOS_IPP_COMMIT		0x43
 
 #define DRM_IOCTL_EXYNOS_GEM_CREATE		DRM_IOWR(DRM_COMMAND_BASE + \
 		DRM_EXYNOS_GEM_CREATE, struct drm_exynos_gem_create)
@@ -165,8 +383,20 @@ struct drm_exynos_g2d_exec {
 #define DRM_IOCTL_EXYNOS_G2D_EXEC		DRM_IOWR(DRM_COMMAND_BASE + \
 		DRM_EXYNOS_G2D_EXEC, struct drm_exynos_g2d_exec)
 
+#define DRM_IOCTL_EXYNOS_IPP_GET_RESOURCES	DRM_IOWR(DRM_COMMAND_BASE + \
+		DRM_EXYNOS_IPP_GET_RESOURCES, \
+		struct drm_exynos_ioctl_ipp_get_res)
+#define DRM_IOCTL_EXYNOS_IPP_GET_CAPS		DRM_IOWR(DRM_COMMAND_BASE + \
+		DRM_EXYNOS_IPP_GET_CAPS, struct drm_exynos_ioctl_ipp_get_caps)
+#define DRM_IOCTL_EXYNOS_IPP_GET_LIMITS		DRM_IOWR(DRM_COMMAND_BASE + \
+		DRM_EXYNOS_IPP_GET_LIMITS, \
+		struct drm_exynos_ioctl_ipp_get_limits)
+#define DRM_IOCTL_EXYNOS_IPP_COMMIT		DRM_IOWR(DRM_COMMAND_BASE + \
+		DRM_EXYNOS_IPP_COMMIT, struct drm_exynos_ioctl_ipp_commit)
+
 /* EXYNOS specific events */
 #define DRM_EXYNOS_G2D_EVENT		0x80000000
+#define DRM_EXYNOS_IPP_EVENT		0x80000002
 
 struct drm_exynos_g2d_event {
 	struct drm_event	base;
@@ -177,6 +407,16 @@ struct drm_exynos_g2d_event {
 	__u32			reserved;
 };
 
+struct drm_exynos_ipp_event {
+	struct drm_event	base;
+	__u64			user_data;
+	__u32			tv_sec;
+	__u32			tv_usec;
+	__u32			ipp_id;
+	__u32			sequence;
+	__u64			reserved;
+};
+
 #if defined(__cplusplus)
 }
 #endif
-- 
cgit v1.1


From d8cb9eeaa79fce028982589da8696df6bb10b903 Mon Sep 17 00:00:00 2001
From: Marek Szyprowski <m.szyprowski@samsung.com>
Date: Wed, 9 May 2018 10:59:23 +0200
Subject: drm/exynos: rotator: Convert driver to IPP v2 core API

This patch adapts Exynos DRM rotator driver to new IPP v2 core API.
The side effect of this conversion is a switch to driver component API
to register properly in the Exynos DRM core.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
---
 drivers/gpu/drm/exynos/Kconfig              |   2 +-
 drivers/gpu/drm/exynos/exynos_drm_drv.c     |   1 +
 drivers/gpu/drm/exynos/exynos_drm_rotator.c | 758 +++++++---------------------
 3 files changed, 190 insertions(+), 571 deletions(-)

diff --git a/drivers/gpu/drm/exynos/Kconfig b/drivers/gpu/drm/exynos/Kconfig
index 9e91465..63a27c1 100644
--- a/drivers/gpu/drm/exynos/Kconfig
+++ b/drivers/gpu/drm/exynos/Kconfig
@@ -106,7 +106,7 @@ config DRM_EXYNOS_FIMC
 
 config DRM_EXYNOS_ROTATOR
 	bool "Rotator"
-	depends on BROKEN
+	select DRM_EXYNOS_IPP
 	help
 	  Choose this option if you want to use Exynos Rotator for DRM.
 
diff --git a/drivers/gpu/drm/exynos/exynos_drm_drv.c b/drivers/gpu/drm/exynos/exynos_drm_drv.c
index 37c0db7..537a588 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_drv.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_drv.c
@@ -263,6 +263,7 @@ static struct exynos_drm_driver_info exynos_drm_drivers[] = {
 		DRV_PTR(fimc_driver, CONFIG_DRM_EXYNOS_FIMC),
 	}, {
 		DRV_PTR(rotator_driver, CONFIG_DRM_EXYNOS_ROTATOR),
+		DRM_COMPONENT_DRIVER
 	}, {
 		DRV_PTR(gsc_driver, CONFIG_DRM_EXYNOS_GSC),
 	}, {
diff --git a/drivers/gpu/drm/exynos/exynos_drm_rotator.c b/drivers/gpu/drm/exynos/exynos_drm_rotator.c
index 79282a8..1a76dd3 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_rotator.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_rotator.c
@@ -10,6 +10,7 @@
  */
 
 #include <linux/kernel.h>
+#include <linux/component.h>
 #include <linux/err.h>
 #include <linux/interrupt.h>
 #include <linux/io.h>
@@ -22,29 +23,18 @@
 #include <drm/exynos_drm.h>
 #include "regs-rotator.h"
 #include "exynos_drm_drv.h"
+#include "exynos_drm_iommu.h"
 #include "exynos_drm_ipp.h"
 
 /*
  * Rotator supports image crop/rotator and input/output DMA operations.
  * input DMA reads image data from the memory.
  * output DMA writes image data to memory.
- *
- * M2M operation : supports crop/scale/rotation/csc so on.
- * Memory ----> Rotator H/W ----> Memory.
  */
 
-/*
- * TODO
- * 1. check suspend/resume api if needed.
- * 2. need to check use case platform_device_id.
- * 3. check src/dst size with, height.
- * 4. need to add supported list in prop_list.
- */
+#define ROTATOR_AUTOSUSPEND_DELAY	2000
 
-#define get_rot_context(dev)	platform_get_drvdata(to_platform_device(dev))
-#define get_ctx_from_ippdrv(ippdrv)	container_of(ippdrv,\
-					struct rot_context, ippdrv);
-#define rot_read(offset)		readl(rot->regs + (offset))
+#define rot_read(offset)	readl(rot->regs + (offset))
 #define rot_write(cfg, offset)	writel(cfg, rot->regs + (offset))
 
 enum rot_irq_status {
@@ -52,54 +42,28 @@ enum rot_irq_status {
 	ROT_IRQ_STATUS_ILLEGAL	= 9,
 };
 
-/*
- * A structure of limitation.
- *
- * @min_w: minimum width.
- * @min_h: minimum height.
- * @max_w: maximum width.
- * @max_h: maximum height.
- * @align: align size.
- */
-struct rot_limit {
-	u32	min_w;
-	u32	min_h;
-	u32	max_w;
-	u32	max_h;
-	u32	align;
-};
-
-/*
- * A structure of limitation table.
- *
- * @ycbcr420_2p: case of YUV.
- * @rgb888: case of RGB.
- */
-struct rot_limit_table {
-	struct rot_limit	ycbcr420_2p;
-	struct rot_limit	rgb888;
+struct rot_variant {
+	const struct exynos_drm_ipp_formats *formats;
+	unsigned int	num_formats;
 };
 
 /*
  * A structure of rotator context.
  * @ippdrv: prepare initialization using ippdrv.
- * @regs_res: register resources.
  * @regs: memory mapped io registers.
  * @clock: rotator gate clock.
  * @limit_tbl: limitation of rotator.
  * @irq: irq number.
- * @cur_buf_id: current operation buffer id.
- * @suspended: suspended state.
  */
 struct rot_context {
-	struct exynos_drm_ippdrv	ippdrv;
-	struct resource	*regs_res;
+	struct exynos_drm_ipp ipp;
+	struct drm_device *drm_dev;
+	struct device	*dev;
 	void __iomem	*regs;
 	struct clk	*clock;
-	struct rot_limit_table	*limit_tbl;
-	int	irq;
-	int	cur_buf_id[EXYNOS_DRM_OPS_MAX];
-	bool	suspended;
+	const struct exynos_drm_ipp_formats *formats;
+	unsigned int	num_formats;
+	struct exynos_drm_ipp_task	*task;
 };
 
 static void rotator_reg_set_irq(struct rot_context *rot, bool enable)
@@ -114,15 +78,6 @@ static void rotator_reg_set_irq(struct rot_context *rot, bool enable)
 	rot_write(val, ROT_CONFIG);
 }
 
-static u32 rotator_reg_get_fmt(struct rot_context *rot)
-{
-	u32 val = rot_read(ROT_CONTROL);
-
-	val &= ROT_CONTROL_FMT_MASK;
-
-	return val;
-}
-
 static enum rot_irq_status rotator_reg_get_irq_status(struct rot_context *rot)
 {
 	u32 val = rot_read(ROT_STATUS);
@@ -138,9 +93,6 @@ static enum rot_irq_status rotator_reg_get_irq_status(struct rot_context *rot)
 static irqreturn_t rotator_irq_handler(int irq, void *arg)
 {
 	struct rot_context *rot = arg;
-	struct exynos_drm_ippdrv *ippdrv = &rot->ippdrv;
-	struct drm_exynos_ipp_cmd_node *c_node = ippdrv->c_node;
-	struct drm_exynos_ipp_event_work *event_work = c_node->event_work;
 	enum rot_irq_status irq_status;
 	u32 val;
 
@@ -152,56 +104,21 @@ static irqreturn_t rotator_irq_handler(int irq, void *arg)
 	val |= ROT_STATUS_IRQ_PENDING((u32)irq_status);
 	rot_write(val, ROT_STATUS);
 
-	if (irq_status == ROT_IRQ_STATUS_COMPLETE) {
-		event_work->ippdrv = ippdrv;
-		event_work->buf_id[EXYNOS_DRM_OPS_DST] =
-			rot->cur_buf_id[EXYNOS_DRM_OPS_DST];
-		queue_work(ippdrv->event_workq, &event_work->work);
-	} else {
-		DRM_ERROR("the SFR is set illegally\n");
+	if (rot->task) {
+		struct exynos_drm_ipp_task *task = rot->task;
+
+		rot->task = NULL;
+		pm_runtime_mark_last_busy(rot->dev);
+		pm_runtime_put_autosuspend(rot->dev);
+		exynos_drm_ipp_task_done(task,
+			irq_status == ROT_IRQ_STATUS_COMPLETE ? 0 : -EINVAL);
 	}
 
 	return IRQ_HANDLED;
 }
 
-static void rotator_align_size(struct rot_context *rot, u32 fmt, u32 *hsize,
-		u32 *vsize)
+static void rotator_src_set_fmt(struct rot_context *rot, u32 fmt)
 {
-	struct rot_limit_table *limit_tbl = rot->limit_tbl;
-	struct rot_limit *limit;
-	u32 mask, val;
-
-	/* Get size limit */
-	if (fmt == ROT_CONTROL_FMT_RGB888)
-		limit = &limit_tbl->rgb888;
-	else
-		limit = &limit_tbl->ycbcr420_2p;
-
-	/* Get mask for rounding to nearest aligned val */
-	mask = ~((1 << limit->align) - 1);
-
-	/* Set aligned width */
-	val = ROT_ALIGN(*hsize, limit->align, mask);
-	if (val < limit->min_w)
-		*hsize = ROT_MIN(limit->min_w, mask);
-	else if (val > limit->max_w)
-		*hsize = ROT_MAX(limit->max_w, mask);
-	else
-		*hsize = val;
-
-	/* Set aligned height */
-	val = ROT_ALIGN(*vsize, limit->align, mask);
-	if (val < limit->min_h)
-		*vsize = ROT_MIN(limit->min_h, mask);
-	else if (val > limit->max_h)
-		*vsize = ROT_MAX(limit->max_h, mask);
-	else
-		*vsize = val;
-}
-
-static int rotator_src_set_fmt(struct device *dev, u32 fmt)
-{
-	struct rot_context *rot = dev_get_drvdata(dev);
 	u32 val;
 
 	val = rot_read(ROT_CONTROL);
@@ -214,515 +131,176 @@ static int rotator_src_set_fmt(struct device *dev, u32 fmt)
 	case DRM_FORMAT_XRGB8888:
 		val |= ROT_CONTROL_FMT_RGB888;
 		break;
-	default:
-		DRM_ERROR("invalid image format\n");
-		return -EINVAL;
 	}
 
 	rot_write(val, ROT_CONTROL);
-
-	return 0;
 }
 
-static inline bool rotator_check_reg_fmt(u32 fmt)
+static void rotator_src_set_buf(struct rot_context *rot,
+				struct exynos_drm_ipp_buffer *buf)
 {
-	if ((fmt == ROT_CONTROL_FMT_YCBCR420_2P) ||
-	    (fmt == ROT_CONTROL_FMT_RGB888))
-		return true;
-
-	return false;
-}
-
-static int rotator_src_set_size(struct device *dev, int swap,
-		struct drm_exynos_pos *pos,
-		struct drm_exynos_sz *sz)
-{
-	struct rot_context *rot = dev_get_drvdata(dev);
-	u32 fmt, hsize, vsize;
 	u32 val;
 
-	/* Get format */
-	fmt = rotator_reg_get_fmt(rot);
-	if (!rotator_check_reg_fmt(fmt)) {
-		DRM_ERROR("invalid format.\n");
-		return -EINVAL;
-	}
-
-	/* Align buffer size */
-	hsize = sz->hsize;
-	vsize = sz->vsize;
-	rotator_align_size(rot, fmt, &hsize, &vsize);
-
 	/* Set buffer size configuration */
-	val = ROT_SET_BUF_SIZE_H(vsize) | ROT_SET_BUF_SIZE_W(hsize);
+	val = ROT_SET_BUF_SIZE_H(buf->buf.height) |
+	      ROT_SET_BUF_SIZE_W(buf->buf.pitch[0] / buf->format->cpp[0]);
 	rot_write(val, ROT_SRC_BUF_SIZE);
 
 	/* Set crop image position configuration */
-	val = ROT_CROP_POS_Y(pos->y) | ROT_CROP_POS_X(pos->x);
+	val = ROT_CROP_POS_Y(buf->rect.y) | ROT_CROP_POS_X(buf->rect.x);
 	rot_write(val, ROT_SRC_CROP_POS);
-	val = ROT_SRC_CROP_SIZE_H(pos->h) | ROT_SRC_CROP_SIZE_W(pos->w);
+	val = ROT_SRC_CROP_SIZE_H(buf->rect.h) |
+	      ROT_SRC_CROP_SIZE_W(buf->rect.w);
 	rot_write(val, ROT_SRC_CROP_SIZE);
 
-	return 0;
+	/* Set buffer DMA address */
+	rot_write(buf->dma_addr[0], ROT_SRC_BUF_ADDR(0));
+	rot_write(buf->dma_addr[1], ROT_SRC_BUF_ADDR(1));
 }
 
-static int rotator_src_set_addr(struct device *dev,
-		struct drm_exynos_ipp_buf_info *buf_info,
-		u32 buf_id, enum drm_exynos_ipp_buf_type buf_type)
+static void rotator_dst_set_transf(struct rot_context *rot,
+				   unsigned int rotation)
 {
-	struct rot_context *rot = dev_get_drvdata(dev);
-	dma_addr_t addr[EXYNOS_DRM_PLANAR_MAX];
-	u32 val, fmt, hsize, vsize;
-	int i;
-
-	/* Set current buf_id */
-	rot->cur_buf_id[EXYNOS_DRM_OPS_SRC] = buf_id;
-
-	switch (buf_type) {
-	case IPP_BUF_ENQUEUE:
-		/* Set address configuration */
-		for_each_ipp_planar(i)
-			addr[i] = buf_info->base[i];
-
-		/* Get format */
-		fmt = rotator_reg_get_fmt(rot);
-		if (!rotator_check_reg_fmt(fmt)) {
-			DRM_ERROR("invalid format.\n");
-			return -EINVAL;
-		}
-
-		/* Re-set cb planar for NV12 format */
-		if ((fmt == ROT_CONTROL_FMT_YCBCR420_2P) &&
-		    !addr[EXYNOS_DRM_PLANAR_CB]) {
-
-			val = rot_read(ROT_SRC_BUF_SIZE);
-			hsize = ROT_GET_BUF_SIZE_W(val);
-			vsize = ROT_GET_BUF_SIZE_H(val);
-
-			/* Set cb planar */
-			addr[EXYNOS_DRM_PLANAR_CB] =
-				addr[EXYNOS_DRM_PLANAR_Y] + hsize * vsize;
-		}
-
-		for_each_ipp_planar(i)
-			rot_write(addr[i], ROT_SRC_BUF_ADDR(i));
-		break;
-	case IPP_BUF_DEQUEUE:
-		for_each_ipp_planar(i)
-			rot_write(0x0, ROT_SRC_BUF_ADDR(i));
-		break;
-	default:
-		/* Nothing to do */
-		break;
-	}
-
-	return 0;
-}
-
-static int rotator_dst_set_transf(struct device *dev,
-		enum drm_exynos_degree degree,
-		enum drm_exynos_flip flip, bool *swap)
-{
-	struct rot_context *rot = dev_get_drvdata(dev);
 	u32 val;
 
 	/* Set transform configuration */
 	val = rot_read(ROT_CONTROL);
 	val &= ~ROT_CONTROL_FLIP_MASK;
 
-	switch (flip) {
-	case EXYNOS_DRM_FLIP_VERTICAL:
-		val |= ROT_CONTROL_FLIP_VERTICAL;
-		break;
-	case EXYNOS_DRM_FLIP_HORIZONTAL:
+	if (rotation & DRM_MODE_REFLECT_X)
 		val |= ROT_CONTROL_FLIP_HORIZONTAL;
-		break;
-	default:
-		/* Flip None */
-		break;
-	}
+	if (rotation & DRM_MODE_REFLECT_Y)
+		val |= ROT_CONTROL_FLIP_VERTICAL;
 
 	val &= ~ROT_CONTROL_ROT_MASK;
 
-	switch (degree) {
-	case EXYNOS_DRM_DEGREE_90:
+	if (rotation & DRM_MODE_ROTATE_90)
 		val |= ROT_CONTROL_ROT_90;
-		break;
-	case EXYNOS_DRM_DEGREE_180:
+	else if (rotation & DRM_MODE_ROTATE_180)
 		val |= ROT_CONTROL_ROT_180;
-		break;
-	case EXYNOS_DRM_DEGREE_270:
+	else if (rotation & DRM_MODE_ROTATE_270)
 		val |= ROT_CONTROL_ROT_270;
-		break;
-	default:
-		/* Rotation 0 Degree */
-		break;
-	}
 
 	rot_write(val, ROT_CONTROL);
-
-	/* Check degree for setting buffer size swap */
-	if ((degree == EXYNOS_DRM_DEGREE_90) ||
-	    (degree == EXYNOS_DRM_DEGREE_270))
-		*swap = true;
-	else
-		*swap = false;
-
-	return 0;
 }
 
-static int rotator_dst_set_size(struct device *dev, int swap,
-		struct drm_exynos_pos *pos,
-		struct drm_exynos_sz *sz)
+static void rotator_dst_set_buf(struct rot_context *rot,
+				struct exynos_drm_ipp_buffer *buf)
 {
-	struct rot_context *rot = dev_get_drvdata(dev);
-	u32 val, fmt, hsize, vsize;
-
-	/* Get format */
-	fmt = rotator_reg_get_fmt(rot);
-	if (!rotator_check_reg_fmt(fmt)) {
-		DRM_ERROR("invalid format.\n");
-		return -EINVAL;
-	}
-
-	/* Align buffer size */
-	hsize = sz->hsize;
-	vsize = sz->vsize;
-	rotator_align_size(rot, fmt, &hsize, &vsize);
+	u32 val;
 
 	/* Set buffer size configuration */
-	val = ROT_SET_BUF_SIZE_H(vsize) | ROT_SET_BUF_SIZE_W(hsize);
+	val = ROT_SET_BUF_SIZE_H(buf->buf.height) |
+	      ROT_SET_BUF_SIZE_W(buf->buf.pitch[0] / buf->format->cpp[0]);
 	rot_write(val, ROT_DST_BUF_SIZE);
 
 	/* Set crop image position configuration */
-	val = ROT_CROP_POS_Y(pos->y) | ROT_CROP_POS_X(pos->x);
+	val = ROT_CROP_POS_Y(buf->rect.y) | ROT_CROP_POS_X(buf->rect.x);
 	rot_write(val, ROT_DST_CROP_POS);
 
-	return 0;
+	/* Set buffer DMA address */
+	rot_write(buf->dma_addr[0], ROT_DST_BUF_ADDR(0));
+	rot_write(buf->dma_addr[1], ROT_DST_BUF_ADDR(1));
 }
 
-static int rotator_dst_set_addr(struct device *dev,
-		struct drm_exynos_ipp_buf_info *buf_info,
-		u32 buf_id, enum drm_exynos_ipp_buf_type buf_type)
+static void rotator_start(struct rot_context *rot)
 {
-	struct rot_context *rot = dev_get_drvdata(dev);
-	dma_addr_t addr[EXYNOS_DRM_PLANAR_MAX];
-	u32 val, fmt, hsize, vsize;
-	int i;
-
-	/* Set current buf_id */
-	rot->cur_buf_id[EXYNOS_DRM_OPS_DST] = buf_id;
-
-	switch (buf_type) {
-	case IPP_BUF_ENQUEUE:
-		/* Set address configuration */
-		for_each_ipp_planar(i)
-			addr[i] = buf_info->base[i];
-
-		/* Get format */
-		fmt = rotator_reg_get_fmt(rot);
-		if (!rotator_check_reg_fmt(fmt)) {
-			DRM_ERROR("invalid format.\n");
-			return -EINVAL;
-		}
-
-		/* Re-set cb planar for NV12 format */
-		if ((fmt == ROT_CONTROL_FMT_YCBCR420_2P) &&
-		    !addr[EXYNOS_DRM_PLANAR_CB]) {
-			/* Get buf size */
-			val = rot_read(ROT_DST_BUF_SIZE);
-
-			hsize = ROT_GET_BUF_SIZE_W(val);
-			vsize = ROT_GET_BUF_SIZE_H(val);
-
-			/* Set cb planar */
-			addr[EXYNOS_DRM_PLANAR_CB] =
-				addr[EXYNOS_DRM_PLANAR_Y] + hsize * vsize;
-		}
-
-		for_each_ipp_planar(i)
-			rot_write(addr[i], ROT_DST_BUF_ADDR(i));
-		break;
-	case IPP_BUF_DEQUEUE:
-		for_each_ipp_planar(i)
-			rot_write(0x0, ROT_DST_BUF_ADDR(i));
-		break;
-	default:
-		/* Nothing to do */
-		break;
-	}
+	u32 val;
 
-	return 0;
+	/* Set interrupt enable */
+	rotator_reg_set_irq(rot, true);
+
+	val = rot_read(ROT_CONTROL);
+	val |= ROT_CONTROL_START;
+	rot_write(val, ROT_CONTROL);
 }
 
-static struct exynos_drm_ipp_ops rot_src_ops = {
-	.set_fmt	=	rotator_src_set_fmt,
-	.set_size	=	rotator_src_set_size,
-	.set_addr	=	rotator_src_set_addr,
-};
+static int rotator_commit(struct exynos_drm_ipp *ipp,
+			  struct exynos_drm_ipp_task *task)
+{
+	struct rot_context *rot =
+			container_of(ipp, struct rot_context, ipp);
 
-static struct exynos_drm_ipp_ops rot_dst_ops = {
-	.set_transf	=	rotator_dst_set_transf,
-	.set_size	=	rotator_dst_set_size,
-	.set_addr	=	rotator_dst_set_addr,
-};
+	pm_runtime_get_sync(rot->dev);
+	rot->task = task;
 
-static int rotator_init_prop_list(struct exynos_drm_ippdrv *ippdrv)
-{
-	struct drm_exynos_ipp_prop_list *prop_list = &ippdrv->prop_list;
-
-	prop_list->version = 1;
-	prop_list->flip = (1 << EXYNOS_DRM_FLIP_VERTICAL) |
-				(1 << EXYNOS_DRM_FLIP_HORIZONTAL);
-	prop_list->degree = (1 << EXYNOS_DRM_DEGREE_0) |
-				(1 << EXYNOS_DRM_DEGREE_90) |
-				(1 << EXYNOS_DRM_DEGREE_180) |
-				(1 << EXYNOS_DRM_DEGREE_270);
-	prop_list->csc = 0;
-	prop_list->crop = 0;
-	prop_list->scale = 0;
+	rotator_src_set_fmt(rot, task->src.buf.fourcc);
+	rotator_src_set_buf(rot, &task->src);
+	rotator_dst_set_transf(rot, task->transform.rotation);
+	rotator_dst_set_buf(rot, &task->dst);
+	rotator_start(rot);
 
 	return 0;
 }
 
-static inline bool rotator_check_drm_fmt(u32 fmt)
-{
-	switch (fmt) {
-	case DRM_FORMAT_XRGB8888:
-	case DRM_FORMAT_NV12:
-		return true;
-	default:
-		DRM_DEBUG_KMS("not support format\n");
-		return false;
-	}
-}
-
-static inline bool rotator_check_drm_flip(enum drm_exynos_flip flip)
-{
-	switch (flip) {
-	case EXYNOS_DRM_FLIP_NONE:
-	case EXYNOS_DRM_FLIP_VERTICAL:
-	case EXYNOS_DRM_FLIP_HORIZONTAL:
-	case EXYNOS_DRM_FLIP_BOTH:
-		return true;
-	default:
-		DRM_DEBUG_KMS("invalid flip\n");
-		return false;
-	}
-}
+static const struct exynos_drm_ipp_funcs ipp_funcs = {
+	.commit = rotator_commit,
+};
 
-static int rotator_ippdrv_check_property(struct device *dev,
-		struct drm_exynos_ipp_property *property)
+static int rotator_bind(struct device *dev, struct device *master, void *data)
 {
-	struct drm_exynos_ipp_config *src_config =
-					&property->config[EXYNOS_DRM_OPS_SRC];
-	struct drm_exynos_ipp_config *dst_config =
-					&property->config[EXYNOS_DRM_OPS_DST];
-	struct drm_exynos_pos *src_pos = &src_config->pos;
-	struct drm_exynos_pos *dst_pos = &dst_config->pos;
-	struct drm_exynos_sz *src_sz = &src_config->sz;
-	struct drm_exynos_sz *dst_sz = &dst_config->sz;
-	bool swap = false;
-
-	/* Check format configuration */
-	if (src_config->fmt != dst_config->fmt) {
-		DRM_DEBUG_KMS("not support csc feature\n");
-		return -EINVAL;
-	}
-
-	if (!rotator_check_drm_fmt(dst_config->fmt)) {
-		DRM_DEBUG_KMS("invalid format\n");
-		return -EINVAL;
-	}
-
-	/* Check transform configuration */
-	if (src_config->degree != EXYNOS_DRM_DEGREE_0) {
-		DRM_DEBUG_KMS("not support source-side rotation\n");
-		return -EINVAL;
-	}
-
-	switch (dst_config->degree) {
-	case EXYNOS_DRM_DEGREE_90:
-	case EXYNOS_DRM_DEGREE_270:
-		swap = true;
-	case EXYNOS_DRM_DEGREE_0:
-	case EXYNOS_DRM_DEGREE_180:
-		/* No problem */
-		break;
-	default:
-		DRM_DEBUG_KMS("invalid degree\n");
-		return -EINVAL;
-	}
-
-	if (src_config->flip != EXYNOS_DRM_FLIP_NONE) {
-		DRM_DEBUG_KMS("not support source-side flip\n");
-		return -EINVAL;
-	}
+	struct rot_context *rot = dev_get_drvdata(dev);
+	struct drm_device *drm_dev = data;
+	struct exynos_drm_ipp *ipp = &rot->ipp;
 
-	if (!rotator_check_drm_flip(dst_config->flip)) {
-		DRM_DEBUG_KMS("invalid flip\n");
-		return -EINVAL;
-	}
+	rot->drm_dev = drm_dev;
+	drm_iommu_attach_device(drm_dev, dev);
 
-	/* Check size configuration */
-	if ((src_pos->x + src_pos->w > src_sz->hsize) ||
-		(src_pos->y + src_pos->h > src_sz->vsize)) {
-		DRM_DEBUG_KMS("out of source buffer bound\n");
-		return -EINVAL;
-	}
+	exynos_drm_ipp_register(drm_dev, ipp, &ipp_funcs,
+			   DRM_EXYNOS_IPP_CAP_CROP | DRM_EXYNOS_IPP_CAP_ROTATE,
+			   rot->formats, rot->num_formats, "rotator");
 
-	if (swap) {
-		if ((dst_pos->x + dst_pos->h > dst_sz->vsize) ||
-			(dst_pos->y + dst_pos->w > dst_sz->hsize)) {
-			DRM_DEBUG_KMS("out of destination buffer bound\n");
-			return -EINVAL;
-		}
-
-		if ((src_pos->w != dst_pos->h) || (src_pos->h != dst_pos->w)) {
-			DRM_DEBUG_KMS("not support scale feature\n");
-			return -EINVAL;
-		}
-	} else {
-		if ((dst_pos->x + dst_pos->w > dst_sz->hsize) ||
-			(dst_pos->y + dst_pos->h > dst_sz->vsize)) {
-			DRM_DEBUG_KMS("out of destination buffer bound\n");
-			return -EINVAL;
-		}
-
-		if ((src_pos->w != dst_pos->w) || (src_pos->h != dst_pos->h)) {
-			DRM_DEBUG_KMS("not support scale feature\n");
-			return -EINVAL;
-		}
-	}
+	dev_info(dev, "The exynos rotator has been probed successfully\n");
 
 	return 0;
 }
 
-static int rotator_ippdrv_start(struct device *dev, enum drm_exynos_ipp_cmd cmd)
+static void rotator_unbind(struct device *dev, struct device *master,
+			void *data)
 {
 	struct rot_context *rot = dev_get_drvdata(dev);
-	u32 val;
-
-	if (rot->suspended) {
-		DRM_ERROR("suspended state\n");
-		return -EPERM;
-	}
-
-	if (cmd != IPP_CMD_M2M) {
-		DRM_ERROR("not support cmd: %d\n", cmd);
-		return -EINVAL;
-	}
-
-	/* Set interrupt enable */
-	rotator_reg_set_irq(rot, true);
-
-	val = rot_read(ROT_CONTROL);
-	val |= ROT_CONTROL_START;
-
-	rot_write(val, ROT_CONTROL);
+	struct drm_device *drm_dev = data;
+	struct exynos_drm_ipp *ipp = &rot->ipp;
 
-	return 0;
+	exynos_drm_ipp_unregister(drm_dev, ipp);
+	drm_iommu_detach_device(rot->drm_dev, rot->dev);
 }
 
-static struct rot_limit_table rot_limit_tbl_4210 = {
-	.ycbcr420_2p = {
-		.min_w = 32,
-		.min_h = 32,
-		.max_w = SZ_64K,
-		.max_h = SZ_64K,
-		.align = 3,
-	},
-	.rgb888 = {
-		.min_w = 8,
-		.min_h = 8,
-		.max_w = SZ_16K,
-		.max_h = SZ_16K,
-		.align = 2,
-	},
-};
-
-static struct rot_limit_table rot_limit_tbl_4x12 = {
-	.ycbcr420_2p = {
-		.min_w = 32,
-		.min_h = 32,
-		.max_w = SZ_32K,
-		.max_h = SZ_32K,
-		.align = 3,
-	},
-	.rgb888 = {
-		.min_w = 8,
-		.min_h = 8,
-		.max_w = SZ_8K,
-		.max_h = SZ_8K,
-		.align = 2,
-	},
+static const struct component_ops rotator_component_ops = {
+	.bind	= rotator_bind,
+	.unbind = rotator_unbind,
 };
 
-static struct rot_limit_table rot_limit_tbl_5250 = {
-	.ycbcr420_2p = {
-		.min_w = 32,
-		.min_h = 32,
-		.max_w = SZ_32K,
-		.max_h = SZ_32K,
-		.align = 3,
-	},
-	.rgb888 = {
-		.min_w = 8,
-		.min_h = 8,
-		.max_w = SZ_8K,
-		.max_h = SZ_8K,
-		.align = 1,
-	},
-};
-
-static const struct of_device_id exynos_rotator_match[] = {
-	{
-		.compatible = "samsung,exynos4210-rotator",
-		.data = &rot_limit_tbl_4210,
-	},
-	{
-		.compatible = "samsung,exynos4212-rotator",
-		.data = &rot_limit_tbl_4x12,
-	},
-	{
-		.compatible = "samsung,exynos5250-rotator",
-		.data = &rot_limit_tbl_5250,
-	},
-	{},
-};
-MODULE_DEVICE_TABLE(of, exynos_rotator_match);
-
 static int rotator_probe(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;
+	struct resource	*regs_res;
 	struct rot_context *rot;
-	struct exynos_drm_ippdrv *ippdrv;
+	const struct rot_variant *variant;
+	int irq;
 	int ret;
 
-	if (!dev->of_node) {
-		dev_err(dev, "cannot find of_node.\n");
-		return -ENODEV;
-	}
-
 	rot = devm_kzalloc(dev, sizeof(*rot), GFP_KERNEL);
 	if (!rot)
 		return -ENOMEM;
 
-	rot->limit_tbl = (struct rot_limit_table *)
-				of_device_get_match_data(dev);
-	rot->regs_res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-	rot->regs = devm_ioremap_resource(dev, rot->regs_res);
+	variant = of_device_get_match_data(dev);
+	rot->formats = variant->formats;
+	rot->num_formats = variant->num_formats;
+	rot->dev = dev;
+	regs_res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	rot->regs = devm_ioremap_resource(dev, regs_res);
 	if (IS_ERR(rot->regs))
 		return PTR_ERR(rot->regs);
 
-	rot->irq = platform_get_irq(pdev, 0);
-	if (rot->irq < 0) {
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
 		dev_err(dev, "failed to get irq\n");
-		return rot->irq;
+		return irq;
 	}
 
-	ret = devm_request_threaded_irq(dev, rot->irq, NULL,
-			rotator_irq_handler, IRQF_ONESHOT, "drm_rotator", rot);
+	ret = devm_request_irq(dev, irq, rotator_irq_handler, 0, dev_name(dev),
+			       rot);
 	if (ret < 0) {
 		dev_err(dev, "failed to request irq\n");
 		return ret;
@@ -734,35 +312,19 @@ static int rotator_probe(struct platform_device *pdev)
 		return PTR_ERR(rot->clock);
 	}
 
+	pm_runtime_use_autosuspend(dev);
+	pm_runtime_set_autosuspend_delay(dev, ROTATOR_AUTOSUSPEND_DELAY);
 	pm_runtime_enable(dev);
-
-	ippdrv = &rot->ippdrv;
-	ippdrv->dev = dev;
-	ippdrv->ops[EXYNOS_DRM_OPS_SRC] = &rot_src_ops;
-	ippdrv->ops[EXYNOS_DRM_OPS_DST] = &rot_dst_ops;
-	ippdrv->check_property = rotator_ippdrv_check_property;
-	ippdrv->start = rotator_ippdrv_start;
-	ret = rotator_init_prop_list(ippdrv);
-	if (ret < 0) {
-		dev_err(dev, "failed to init property list.\n");
-		goto err_ippdrv_register;
-	}
-
-	DRM_DEBUG_KMS("ippdrv[%pK]\n", ippdrv);
-
 	platform_set_drvdata(pdev, rot);
 
-	ret = exynos_drm_ippdrv_register(ippdrv);
-	if (ret < 0) {
-		dev_err(dev, "failed to register drm rotator device\n");
-		goto err_ippdrv_register;
-	}
-
-	dev_info(dev, "The exynos rotator is probed successfully\n");
+	ret = component_add(dev, &rotator_component_ops);
+	if (ret)
+		goto err_component;
 
 	return 0;
 
-err_ippdrv_register:
+err_component:
+	pm_runtime_dont_use_autosuspend(dev);
 	pm_runtime_disable(dev);
 	return ret;
 }
@@ -770,45 +332,101 @@ err_ippdrv_register:
 static int rotator_remove(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;
-	struct rot_context *rot = dev_get_drvdata(dev);
-	struct exynos_drm_ippdrv *ippdrv = &rot->ippdrv;
-
-	exynos_drm_ippdrv_unregister(ippdrv);
 
+	component_del(dev, &rotator_component_ops);
+	pm_runtime_dont_use_autosuspend(dev);
 	pm_runtime_disable(dev);
 
 	return 0;
 }
 
 #ifdef CONFIG_PM
-static int rotator_clk_crtl(struct rot_context *rot, bool enable)
-{
-	if (enable) {
-		clk_prepare_enable(rot->clock);
-		rot->suspended = false;
-	} else {
-		clk_disable_unprepare(rot->clock);
-		rot->suspended = true;
-	}
-
-	return 0;
-}
-
 static int rotator_runtime_suspend(struct device *dev)
 {
 	struct rot_context *rot = dev_get_drvdata(dev);
 
-	return  rotator_clk_crtl(rot, false);
+	clk_disable_unprepare(rot->clock);
+	return 0;
 }
 
 static int rotator_runtime_resume(struct device *dev)
 {
 	struct rot_context *rot = dev_get_drvdata(dev);
 
-	return  rotator_clk_crtl(rot, true);
+	return clk_prepare_enable(rot->clock);
 }
 #endif
 
+static const struct drm_exynos_ipp_limit rotator_4210_rbg888_limits[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 8, SZ_16K }, .v = { 8, SZ_16K }) },
+	{ IPP_SIZE_LIMIT(AREA, .h.align = 4, .v.align = 4) },
+};
+
+static const struct drm_exynos_ipp_limit rotator_4412_rbg888_limits[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 8, SZ_8K }, .v = { 8, SZ_8K }) },
+	{ IPP_SIZE_LIMIT(AREA, .h.align = 4, .v.align = 4) },
+};
+
+static const struct drm_exynos_ipp_limit rotator_5250_rbg888_limits[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 8, SZ_8K }, .v = { 8, SZ_8K }) },
+	{ IPP_SIZE_LIMIT(AREA, .h.align = 2, .v.align = 2) },
+};
+
+static const struct drm_exynos_ipp_limit rotator_4210_yuv_limits[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 32, SZ_64K }, .v = { 32, SZ_64K }) },
+	{ IPP_SIZE_LIMIT(AREA, .h.align = 8, .v.align = 8) },
+};
+
+static const struct drm_exynos_ipp_limit rotator_4412_yuv_limits[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 32, SZ_32K }, .v = { 32, SZ_32K }) },
+	{ IPP_SIZE_LIMIT(AREA, .h.align = 8, .v.align = 8) },
+};
+
+static const struct exynos_drm_ipp_formats rotator_4210_formats[] = {
+	{ IPP_SRCDST_FORMAT(XRGB8888, rotator_4210_rbg888_limits) },
+	{ IPP_SRCDST_FORMAT(NV12, rotator_4210_yuv_limits) },
+};
+
+static const struct exynos_drm_ipp_formats rotator_4412_formats[] = {
+	{ IPP_SRCDST_FORMAT(XRGB8888, rotator_4412_rbg888_limits) },
+	{ IPP_SRCDST_FORMAT(NV12, rotator_4412_yuv_limits) },
+};
+
+static const struct exynos_drm_ipp_formats rotator_5250_formats[] = {
+	{ IPP_SRCDST_FORMAT(XRGB8888, rotator_5250_rbg888_limits) },
+	{ IPP_SRCDST_FORMAT(NV12, rotator_4412_yuv_limits) },
+};
+
+static const struct rot_variant rotator_4210_data = {
+	.formats = rotator_4210_formats,
+	.num_formats = ARRAY_SIZE(rotator_4210_formats),
+};
+
+static const struct rot_variant rotator_4412_data = {
+	.formats = rotator_4412_formats,
+	.num_formats = ARRAY_SIZE(rotator_4412_formats),
+};
+
+static const struct rot_variant rotator_5250_data = {
+	.formats = rotator_5250_formats,
+	.num_formats = ARRAY_SIZE(rotator_5250_formats),
+};
+
+static const struct of_device_id exynos_rotator_match[] = {
+	{
+		.compatible = "samsung,exynos4210-rotator",
+		.data = &rotator_4210_data,
+	}, {
+		.compatible = "samsung,exynos4212-rotator",
+		.data = &rotator_4412_data,
+	}, {
+		.compatible = "samsung,exynos5250-rotator",
+		.data = &rotator_5250_data,
+	}, {
+	},
+};
+MODULE_DEVICE_TABLE(of, exynos_rotator_match);
+
 static const struct dev_pm_ops rotator_pm_ops = {
 	SET_SYSTEM_SLEEP_PM_OPS(pm_runtime_force_suspend,
 				pm_runtime_force_resume)
@@ -820,7 +438,7 @@ struct platform_driver rotator_driver = {
 	.probe		= rotator_probe,
 	.remove		= rotator_remove,
 	.driver		= {
-		.name	= "exynos-rot",
+		.name	= "exynos-rotator",
 		.owner	= THIS_MODULE,
 		.pm	= &rotator_pm_ops,
 		.of_match_table = exynos_rotator_match,
-- 
cgit v1.1


From 8b7d3ec83aba6381bfc123c7aebcd78199635c3a Mon Sep 17 00:00:00 2001
From: Marek Szyprowski <m.szyprowski@samsung.com>
Date: Wed, 9 May 2018 10:59:24 +0200
Subject: drm/exynos: gsc: Convert driver to IPP v2 core API

This patch adapts Exynos DRM GScaler driver to new IPP v2 core API.
The side effect of this conversion is a switch to driver component API
to register properly in the Exynos DRM core. During the conversion
driver has been adapted to support more specific compatible strings
to distinguish between Exynos5250 and Exynos5420 (different hardware
limits). Support for Exynos5433 variant has been added too
(different limits table, removed dependency on ARCH_EXYNOS5).

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Tested-by: Hoegeun Kwon <hoegeun.kwon@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
---
 drivers/gpu/drm/exynos/Kconfig          |    3 +-
 drivers/gpu/drm/exynos/exynos_drm_drv.c |    1 +
 drivers/gpu/drm/exynos/exynos_drm_gsc.c | 1075 ++++++++++---------------------
 drivers/gpu/drm/exynos/exynos_drm_gsc.h |   24 -
 4 files changed, 338 insertions(+), 765 deletions(-)
 delete mode 100644 drivers/gpu/drm/exynos/exynos_drm_gsc.h

diff --git a/drivers/gpu/drm/exynos/Kconfig b/drivers/gpu/drm/exynos/Kconfig
index 63a27c1..5c21654 100644
--- a/drivers/gpu/drm/exynos/Kconfig
+++ b/drivers/gpu/drm/exynos/Kconfig
@@ -112,7 +112,8 @@ config DRM_EXYNOS_ROTATOR
 
 config DRM_EXYNOS_GSC
 	bool "GScaler"
-	depends on BROKEN && ARCH_EXYNOS5 && VIDEO_SAMSUNG_EXYNOS_GSC=n
+	depends on VIDEO_SAMSUNG_EXYNOS_GSC=n
+	select DRM_EXYNOS_IPP
 	help
 	  Choose this option if you want to use Exynos GSC for DRM.
 
diff --git a/drivers/gpu/drm/exynos/exynos_drm_drv.c b/drivers/gpu/drm/exynos/exynos_drm_drv.c
index 537a588..2dcb940 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_drv.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_drv.c
@@ -266,6 +266,7 @@ static struct exynos_drm_driver_info exynos_drm_drivers[] = {
 		DRM_COMPONENT_DRIVER
 	}, {
 		DRV_PTR(gsc_driver, CONFIG_DRM_EXYNOS_GSC),
+		DRM_COMPONENT_DRIVER
 	}, {
 		&exynos_drm_platform_driver,
 		DRM_VIRTUAL_DEVICE
diff --git a/drivers/gpu/drm/exynos/exynos_drm_gsc.c b/drivers/gpu/drm/exynos/exynos_drm_gsc.c
index 0506b2b..e99dd1e 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_gsc.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_gsc.c
@@ -12,18 +12,20 @@
  *
  */
 #include <linux/kernel.h>
+#include <linux/component.h>
 #include <linux/platform_device.h>
 #include <linux/clk.h>
 #include <linux/pm_runtime.h>
 #include <linux/mfd/syscon.h>
+#include <linux/of_device.h>
 #include <linux/regmap.h>
 
 #include <drm/drmP.h>
 #include <drm/exynos_drm.h>
 #include "regs-gsc.h"
 #include "exynos_drm_drv.h"
+#include "exynos_drm_iommu.h"
 #include "exynos_drm_ipp.h"
-#include "exynos_drm_gsc.h"
 
 /*
  * GSC stands for General SCaler and
@@ -31,26 +33,10 @@
  * input DMA reads image data from the memory.
  * output DMA writes image data to memory.
  * GSC supports image rotation and image effect functions.
- *
- * M2M operation : supports crop/scale/rotation/csc so on.
- * Memory ----> GSC H/W ----> Memory.
- * Writeback operation : supports cloned screen with FIMD.
- * FIMD ----> GSC H/W ----> Memory.
- * Output operation : supports direct display using local path.
- * Memory ----> GSC H/W ----> FIMD, Mixer.
  */
 
-/*
- * TODO
- * 1. check suspend/resume api if needed.
- * 2. need to check use case platform_device_id.
- * 3. check src/dst size with, height.
- * 4. added check_prepare api for right register.
- * 5. need to add supported list in prop_list.
- * 6. check prescaler/scaler optimization.
- */
 
-#define GSC_MAX_DEVS	4
+#define GSC_MAX_CLOCKS	8
 #define GSC_MAX_SRC		4
 #define GSC_MAX_DST		16
 #define GSC_RESET_TIMEOUT	50
@@ -65,8 +51,6 @@
 #define GSC_SC_DOWN_RATIO_4_8		131072
 #define GSC_SC_DOWN_RATIO_3_8		174762
 #define GSC_SC_DOWN_RATIO_2_8		262144
-#define GSC_REFRESH_MIN	12
-#define GSC_REFRESH_MAX	60
 #define GSC_CROP_MAX	8192
 #define GSC_CROP_MIN	32
 #define GSC_SCALE_MAX	4224
@@ -77,10 +61,9 @@
 #define GSC_COEF_H_8T	8
 #define GSC_COEF_V_4T	4
 #define GSC_COEF_DEPTH	3
+#define GSC_AUTOSUSPEND_DELAY		2000
 
 #define get_gsc_context(dev)	platform_get_drvdata(to_platform_device(dev))
-#define get_ctx_from_ippdrv(ippdrv)	container_of(ippdrv,\
-					struct gsc_context, ippdrv);
 #define gsc_read(offset)		readl(ctx->regs + (offset))
 #define gsc_write(cfg, offset)	writel(cfg, ctx->regs + (offset))
 
@@ -104,50 +87,47 @@ struct gsc_scaler {
 };
 
 /*
- * A structure of scaler capability.
- *
- * find user manual 49.2 features.
- * @tile_w: tile mode or rotation width.
- * @tile_h: tile mode or rotation height.
- * @w: other cases width.
- * @h: other cases height.
- */
-struct gsc_capability {
-	/* tile or rotation */
-	u32	tile_w;
-	u32	tile_h;
-	/* other cases */
-	u32	w;
-	u32	h;
-};
-
-/*
  * A structure of gsc context.
  *
- * @ippdrv: prepare initialization using ippdrv.
  * @regs_res: register resources.
  * @regs: memory mapped io registers.
- * @sysreg: handle to SYSREG block regmap.
- * @lock: locking of operations.
  * @gsc_clk: gsc gate clock.
  * @sc: scaler infomations.
  * @id: gsc id.
  * @irq: irq number.
  * @rotation: supports rotation of src.
- * @suspended: qos operations.
  */
 struct gsc_context {
-	struct exynos_drm_ippdrv	ippdrv;
+	struct exynos_drm_ipp ipp;
+	struct drm_device *drm_dev;
+	struct device	*dev;
+	struct exynos_drm_ipp_task	*task;
+	struct exynos_drm_ipp_formats	*formats;
+	unsigned int			num_formats;
+
 	struct resource	*regs_res;
 	void __iomem	*regs;
-	struct regmap	*sysreg;
-	struct mutex	lock;
-	struct clk	*gsc_clk;
+	const char	**clk_names;
+	struct clk	*clocks[GSC_MAX_CLOCKS];
+	int		num_clocks;
 	struct gsc_scaler	sc;
 	int	id;
 	int	irq;
 	bool	rotation;
-	bool	suspended;
+};
+
+/**
+ * struct gsc_driverdata - per device type driver data for init time.
+ *
+ * @limits: picture size limits array
+ * @clk_names: names of clocks needed by this variant
+ * @num_clocks: the number of clocks needed by this variant
+ */
+struct gsc_driverdata {
+	const struct drm_exynos_ipp_limit *limits;
+	int		num_limits;
+	const char	*clk_names[GSC_MAX_CLOCKS];
+	int		num_clocks;
 };
 
 /* 8-tap Filter Coefficient */
@@ -438,25 +418,6 @@ static int gsc_sw_reset(struct gsc_context *ctx)
 	return 0;
 }
 
-static void gsc_set_gscblk_fimd_wb(struct gsc_context *ctx, bool enable)
-{
-	unsigned int gscblk_cfg;
-
-	if (!ctx->sysreg)
-		return;
-
-	regmap_read(ctx->sysreg, SYSREG_GSCBLK_CFG1, &gscblk_cfg);
-
-	if (enable)
-		gscblk_cfg |= GSC_BLK_DISP1WB_DEST(ctx->id) |
-				GSC_BLK_GSCL_WB_IN_SRC_SEL(ctx->id) |
-				GSC_BLK_SW_RESET_WB_DEST(ctx->id);
-	else
-		gscblk_cfg |= GSC_BLK_PXLASYNC_LO_MASK_WB(ctx->id);
-
-	regmap_write(ctx->sysreg, SYSREG_GSCBLK_CFG1, gscblk_cfg);
-}
-
 static void gsc_handle_irq(struct gsc_context *ctx, bool enable,
 		bool overflow, bool done)
 {
@@ -487,10 +448,8 @@ static void gsc_handle_irq(struct gsc_context *ctx, bool enable,
 }
 
 
-static int gsc_src_set_fmt(struct device *dev, u32 fmt)
+static void gsc_src_set_fmt(struct gsc_context *ctx, u32 fmt)
 {
-	struct gsc_context *ctx = get_gsc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
 	u32 cfg;
 
 	DRM_DEBUG_KMS("fmt[0x%x]\n", fmt);
@@ -506,6 +465,7 @@ static int gsc_src_set_fmt(struct device *dev, u32 fmt)
 		cfg |= GSC_IN_RGB565;
 		break;
 	case DRM_FORMAT_XRGB8888:
+	case DRM_FORMAT_ARGB8888:
 		cfg |= GSC_IN_XRGB8888;
 		break;
 	case DRM_FORMAT_BGRX8888:
@@ -548,115 +508,84 @@ static int gsc_src_set_fmt(struct device *dev, u32 fmt)
 		cfg |= (GSC_IN_CHROMA_ORDER_CBCR |
 			GSC_IN_YUV420_2P);
 		break;
-	default:
-		dev_err(ippdrv->dev, "invalid target yuv order 0x%x.\n", fmt);
-		return -EINVAL;
 	}
 
 	gsc_write(cfg, GSC_IN_CON);
-
-	return 0;
 }
 
-static int gsc_src_set_transf(struct device *dev,
-		enum drm_exynos_degree degree,
-		enum drm_exynos_flip flip, bool *swap)
+static void gsc_src_set_transf(struct gsc_context *ctx, unsigned int rotation)
 {
-	struct gsc_context *ctx = get_gsc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
+	unsigned int degree = rotation & DRM_MODE_ROTATE_MASK;
 	u32 cfg;
 
-	DRM_DEBUG_KMS("degree[%d]flip[0x%x]\n", degree, flip);
-
 	cfg = gsc_read(GSC_IN_CON);
 	cfg &= ~GSC_IN_ROT_MASK;
 
 	switch (degree) {
-	case EXYNOS_DRM_DEGREE_0:
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
+	case DRM_MODE_ROTATE_0:
+		if (rotation & DRM_MODE_REFLECT_Y)
 			cfg |= GSC_IN_ROT_XFLIP;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
+		if (rotation & DRM_MODE_REFLECT_X)
 			cfg |= GSC_IN_ROT_YFLIP;
 		break;
-	case EXYNOS_DRM_DEGREE_90:
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
-			cfg |= GSC_IN_ROT_90_XFLIP;
-		else if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
-			cfg |= GSC_IN_ROT_90_YFLIP;
-		else
-			cfg |= GSC_IN_ROT_90;
+	case DRM_MODE_ROTATE_90:
+		cfg |= GSC_IN_ROT_90;
+		if (rotation & DRM_MODE_REFLECT_Y)
+			cfg |= GSC_IN_ROT_XFLIP;
+		if (rotation & DRM_MODE_REFLECT_X)
+			cfg |= GSC_IN_ROT_YFLIP;
 		break;
-	case EXYNOS_DRM_DEGREE_180:
+	case DRM_MODE_ROTATE_180:
 		cfg |= GSC_IN_ROT_180;
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
+		if (rotation & DRM_MODE_REFLECT_Y)
 			cfg &= ~GSC_IN_ROT_XFLIP;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
+		if (rotation & DRM_MODE_REFLECT_X)
 			cfg &= ~GSC_IN_ROT_YFLIP;
 		break;
-	case EXYNOS_DRM_DEGREE_270:
+	case DRM_MODE_ROTATE_270:
 		cfg |= GSC_IN_ROT_270;
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
+		if (rotation & DRM_MODE_REFLECT_Y)
 			cfg &= ~GSC_IN_ROT_XFLIP;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
+		if (rotation & DRM_MODE_REFLECT_X)
 			cfg &= ~GSC_IN_ROT_YFLIP;
 		break;
-	default:
-		dev_err(ippdrv->dev, "invalid degree value %d.\n", degree);
-		return -EINVAL;
 	}
 
 	gsc_write(cfg, GSC_IN_CON);
 
 	ctx->rotation = (cfg & GSC_IN_ROT_90) ? 1 : 0;
-	*swap = ctx->rotation;
-
-	return 0;
 }
 
-static int gsc_src_set_size(struct device *dev, int swap,
-		struct drm_exynos_pos *pos, struct drm_exynos_sz *sz)
+static void gsc_src_set_size(struct gsc_context *ctx,
+			     struct exynos_drm_ipp_buffer *buf)
 {
-	struct gsc_context *ctx = get_gsc_context(dev);
-	struct drm_exynos_pos img_pos = *pos;
 	struct gsc_scaler *sc = &ctx->sc;
 	u32 cfg;
 
-	DRM_DEBUG_KMS("swap[%d]x[%d]y[%d]w[%d]h[%d]\n",
-		swap, pos->x, pos->y, pos->w, pos->h);
-
-	if (swap) {
-		img_pos.w = pos->h;
-		img_pos.h = pos->w;
-	}
-
 	/* pixel offset */
-	cfg = (GSC_SRCIMG_OFFSET_X(img_pos.x) |
-		GSC_SRCIMG_OFFSET_Y(img_pos.y));
+	cfg = (GSC_SRCIMG_OFFSET_X(buf->rect.x) |
+		GSC_SRCIMG_OFFSET_Y(buf->rect.y));
 	gsc_write(cfg, GSC_SRCIMG_OFFSET);
 
 	/* cropped size */
-	cfg = (GSC_CROPPED_WIDTH(img_pos.w) |
-		GSC_CROPPED_HEIGHT(img_pos.h));
+	cfg = (GSC_CROPPED_WIDTH(buf->rect.w) |
+		GSC_CROPPED_HEIGHT(buf->rect.h));
 	gsc_write(cfg, GSC_CROPPED_SIZE);
 
-	DRM_DEBUG_KMS("hsize[%d]vsize[%d]\n", sz->hsize, sz->vsize);
-
 	/* original size */
 	cfg = gsc_read(GSC_SRCIMG_SIZE);
 	cfg &= ~(GSC_SRCIMG_HEIGHT_MASK |
 		GSC_SRCIMG_WIDTH_MASK);
 
-	cfg |= (GSC_SRCIMG_WIDTH(sz->hsize) |
-		GSC_SRCIMG_HEIGHT(sz->vsize));
+	cfg |= (GSC_SRCIMG_WIDTH(buf->buf.width) |
+		GSC_SRCIMG_HEIGHT(buf->buf.height));
 
 	gsc_write(cfg, GSC_SRCIMG_SIZE);
 
 	cfg = gsc_read(GSC_IN_CON);
 	cfg &= ~GSC_IN_RGB_TYPE_MASK;
 
-	DRM_DEBUG_KMS("width[%d]range[%d]\n", pos->w, sc->range);
-
-	if (pos->w >= GSC_WIDTH_ITU_709)
+	if (buf->rect.w >= GSC_WIDTH_ITU_709)
 		if (sc->range)
 			cfg |= GSC_IN_RGB_HD_WIDE;
 		else
@@ -668,103 +597,39 @@ static int gsc_src_set_size(struct device *dev, int swap,
 			cfg |= GSC_IN_RGB_SD_NARROW;
 
 	gsc_write(cfg, GSC_IN_CON);
-
-	return 0;
 }
 
-static int gsc_src_set_buf_seq(struct gsc_context *ctx, u32 buf_id,
-		enum drm_exynos_ipp_buf_type buf_type)
+static void gsc_src_set_buf_seq(struct gsc_context *ctx, u32 buf_id,
+			       bool enqueue)
 {
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	bool masked;
+	bool masked = !enqueue;
 	u32 cfg;
 	u32 mask = 0x00000001 << buf_id;
 
-	DRM_DEBUG_KMS("buf_id[%d]buf_type[%d]\n", buf_id, buf_type);
-
 	/* mask register set */
 	cfg = gsc_read(GSC_IN_BASE_ADDR_Y_MASK);
 
-	switch (buf_type) {
-	case IPP_BUF_ENQUEUE:
-		masked = false;
-		break;
-	case IPP_BUF_DEQUEUE:
-		masked = true;
-		break;
-	default:
-		dev_err(ippdrv->dev, "invalid buf ctrl parameter.\n");
-		return -EINVAL;
-	}
-
 	/* sequence id */
 	cfg &= ~mask;
 	cfg |= masked << buf_id;
 	gsc_write(cfg, GSC_IN_BASE_ADDR_Y_MASK);
 	gsc_write(cfg, GSC_IN_BASE_ADDR_CB_MASK);
 	gsc_write(cfg, GSC_IN_BASE_ADDR_CR_MASK);
-
-	return 0;
 }
 
-static int gsc_src_set_addr(struct device *dev,
-		struct drm_exynos_ipp_buf_info *buf_info, u32 buf_id,
-		enum drm_exynos_ipp_buf_type buf_type)
+static void gsc_src_set_addr(struct gsc_context *ctx, u32 buf_id,
+			    struct exynos_drm_ipp_buffer *buf)
 {
-	struct gsc_context *ctx = get_gsc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	struct drm_exynos_ipp_cmd_node *c_node = ippdrv->c_node;
-	struct drm_exynos_ipp_property *property;
-
-	if (!c_node) {
-		DRM_ERROR("failed to get c_node.\n");
-		return -EFAULT;
-	}
-
-	property = &c_node->property;
-
-	DRM_DEBUG_KMS("prop_id[%d]buf_id[%d]buf_type[%d]\n",
-		property->prop_id, buf_id, buf_type);
-
-	if (buf_id > GSC_MAX_SRC) {
-		dev_info(ippdrv->dev, "invalid buf_id %d.\n", buf_id);
-		return -EINVAL;
-	}
-
 	/* address register set */
-	switch (buf_type) {
-	case IPP_BUF_ENQUEUE:
-		gsc_write(buf_info->base[EXYNOS_DRM_PLANAR_Y],
-			GSC_IN_BASE_ADDR_Y(buf_id));
-		gsc_write(buf_info->base[EXYNOS_DRM_PLANAR_CB],
-			GSC_IN_BASE_ADDR_CB(buf_id));
-		gsc_write(buf_info->base[EXYNOS_DRM_PLANAR_CR],
-			GSC_IN_BASE_ADDR_CR(buf_id));
-		break;
-	case IPP_BUF_DEQUEUE:
-		gsc_write(0x0, GSC_IN_BASE_ADDR_Y(buf_id));
-		gsc_write(0x0, GSC_IN_BASE_ADDR_CB(buf_id));
-		gsc_write(0x0, GSC_IN_BASE_ADDR_CR(buf_id));
-		break;
-	default:
-		/* bypass */
-		break;
-	}
+	gsc_write(buf->dma_addr[0], GSC_IN_BASE_ADDR_Y(buf_id));
+	gsc_write(buf->dma_addr[1], GSC_IN_BASE_ADDR_CB(buf_id));
+	gsc_write(buf->dma_addr[2], GSC_IN_BASE_ADDR_CR(buf_id));
 
-	return gsc_src_set_buf_seq(ctx, buf_id, buf_type);
+	gsc_src_set_buf_seq(ctx, buf_id, true);
 }
 
-static struct exynos_drm_ipp_ops gsc_src_ops = {
-	.set_fmt = gsc_src_set_fmt,
-	.set_transf = gsc_src_set_transf,
-	.set_size = gsc_src_set_size,
-	.set_addr = gsc_src_set_addr,
-};
-
-static int gsc_dst_set_fmt(struct device *dev, u32 fmt)
+static void gsc_dst_set_fmt(struct gsc_context *ctx, u32 fmt)
 {
-	struct gsc_context *ctx = get_gsc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
 	u32 cfg;
 
 	DRM_DEBUG_KMS("fmt[0x%x]\n", fmt);
@@ -779,8 +644,9 @@ static int gsc_dst_set_fmt(struct device *dev, u32 fmt)
 	case DRM_FORMAT_RGB565:
 		cfg |= GSC_OUT_RGB565;
 		break;
+	case DRM_FORMAT_ARGB8888:
 	case DRM_FORMAT_XRGB8888:
-		cfg |= GSC_OUT_XRGB8888;
+		cfg |= (GSC_OUT_XRGB8888 | GSC_OUT_GLOBAL_ALPHA(0xff));
 		break;
 	case DRM_FORMAT_BGRX8888:
 		cfg |= (GSC_OUT_XRGB8888 | GSC_OUT_RB_SWAP);
@@ -819,69 +685,9 @@ static int gsc_dst_set_fmt(struct device *dev, u32 fmt)
 		cfg |= (GSC_OUT_CHROMA_ORDER_CBCR |
 			GSC_OUT_YUV420_2P);
 		break;
-	default:
-		dev_err(ippdrv->dev, "invalid target yuv order 0x%x.\n", fmt);
-		return -EINVAL;
 	}
 
 	gsc_write(cfg, GSC_OUT_CON);
-
-	return 0;
-}
-
-static int gsc_dst_set_transf(struct device *dev,
-		enum drm_exynos_degree degree,
-		enum drm_exynos_flip flip, bool *swap)
-{
-	struct gsc_context *ctx = get_gsc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	u32 cfg;
-
-	DRM_DEBUG_KMS("degree[%d]flip[0x%x]\n", degree, flip);
-
-	cfg = gsc_read(GSC_IN_CON);
-	cfg &= ~GSC_IN_ROT_MASK;
-
-	switch (degree) {
-	case EXYNOS_DRM_DEGREE_0:
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
-			cfg |= GSC_IN_ROT_XFLIP;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
-			cfg |= GSC_IN_ROT_YFLIP;
-		break;
-	case EXYNOS_DRM_DEGREE_90:
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
-			cfg |= GSC_IN_ROT_90_XFLIP;
-		else if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
-			cfg |= GSC_IN_ROT_90_YFLIP;
-		else
-			cfg |= GSC_IN_ROT_90;
-		break;
-	case EXYNOS_DRM_DEGREE_180:
-		cfg |= GSC_IN_ROT_180;
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
-			cfg &= ~GSC_IN_ROT_XFLIP;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
-			cfg &= ~GSC_IN_ROT_YFLIP;
-		break;
-	case EXYNOS_DRM_DEGREE_270:
-		cfg |= GSC_IN_ROT_270;
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
-			cfg &= ~GSC_IN_ROT_XFLIP;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
-			cfg &= ~GSC_IN_ROT_YFLIP;
-		break;
-	default:
-		dev_err(ippdrv->dev, "invalid degree value %d.\n", degree);
-		return -EINVAL;
-	}
-
-	gsc_write(cfg, GSC_IN_CON);
-
-	ctx->rotation = (cfg & GSC_IN_ROT_90) ? 1 : 0;
-	*swap = ctx->rotation;
-
-	return 0;
 }
 
 static int gsc_get_ratio_shift(u32 src, u32 dst, u32 *ratio)
@@ -919,9 +725,9 @@ static void gsc_get_prescaler_shfactor(u32 hratio, u32 vratio, u32 *shfactor)
 }
 
 static int gsc_set_prescaler(struct gsc_context *ctx, struct gsc_scaler *sc,
-		struct drm_exynos_pos *src, struct drm_exynos_pos *dst)
+			     struct drm_exynos_ipp_task_rect *src,
+			     struct drm_exynos_ipp_task_rect *dst)
 {
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
 	u32 cfg;
 	u32 src_w, src_h, dst_w, dst_h;
 	int ret = 0;
@@ -939,13 +745,13 @@ static int gsc_set_prescaler(struct gsc_context *ctx, struct gsc_scaler *sc,
 
 	ret = gsc_get_ratio_shift(src_w, dst_w, &sc->pre_hratio);
 	if (ret) {
-		dev_err(ippdrv->dev, "failed to get ratio horizontal.\n");
+		dev_err(ctx->dev, "failed to get ratio horizontal.\n");
 		return ret;
 	}
 
 	ret = gsc_get_ratio_shift(src_h, dst_h, &sc->pre_vratio);
 	if (ret) {
-		dev_err(ippdrv->dev, "failed to get ratio vertical.\n");
+		dev_err(ctx->dev, "failed to get ratio vertical.\n");
 		return ret;
 	}
 
@@ -1039,47 +845,37 @@ static void gsc_set_scaler(struct gsc_context *ctx, struct gsc_scaler *sc)
 	gsc_write(cfg, GSC_MAIN_V_RATIO);
 }
 
-static int gsc_dst_set_size(struct device *dev, int swap,
-		struct drm_exynos_pos *pos, struct drm_exynos_sz *sz)
+static void gsc_dst_set_size(struct gsc_context *ctx,
+			     struct exynos_drm_ipp_buffer *buf)
 {
-	struct gsc_context *ctx = get_gsc_context(dev);
-	struct drm_exynos_pos img_pos = *pos;
 	struct gsc_scaler *sc = &ctx->sc;
 	u32 cfg;
 
-	DRM_DEBUG_KMS("swap[%d]x[%d]y[%d]w[%d]h[%d]\n",
-		swap, pos->x, pos->y, pos->w, pos->h);
-
-	if (swap) {
-		img_pos.w = pos->h;
-		img_pos.h = pos->w;
-	}
-
 	/* pixel offset */
-	cfg = (GSC_DSTIMG_OFFSET_X(pos->x) |
-		GSC_DSTIMG_OFFSET_Y(pos->y));
+	cfg = (GSC_DSTIMG_OFFSET_X(buf->rect.x) |
+		GSC_DSTIMG_OFFSET_Y(buf->rect.y));
 	gsc_write(cfg, GSC_DSTIMG_OFFSET);
 
 	/* scaled size */
-	cfg = (GSC_SCALED_WIDTH(img_pos.w) | GSC_SCALED_HEIGHT(img_pos.h));
+	if (ctx->rotation)
+		cfg = (GSC_SCALED_WIDTH(buf->rect.h) |
+		       GSC_SCALED_HEIGHT(buf->rect.w));
+	else
+		cfg = (GSC_SCALED_WIDTH(buf->rect.w) |
+		       GSC_SCALED_HEIGHT(buf->rect.h));
 	gsc_write(cfg, GSC_SCALED_SIZE);
 
-	DRM_DEBUG_KMS("hsize[%d]vsize[%d]\n", sz->hsize, sz->vsize);
-
 	/* original size */
 	cfg = gsc_read(GSC_DSTIMG_SIZE);
-	cfg &= ~(GSC_DSTIMG_HEIGHT_MASK |
-		GSC_DSTIMG_WIDTH_MASK);
-	cfg |= (GSC_DSTIMG_WIDTH(sz->hsize) |
-		GSC_DSTIMG_HEIGHT(sz->vsize));
+	cfg &= ~(GSC_DSTIMG_HEIGHT_MASK | GSC_DSTIMG_WIDTH_MASK);
+	cfg |= GSC_DSTIMG_WIDTH(buf->buf.width) |
+	       GSC_DSTIMG_HEIGHT(buf->buf.height);
 	gsc_write(cfg, GSC_DSTIMG_SIZE);
 
 	cfg = gsc_read(GSC_OUT_CON);
 	cfg &= ~GSC_OUT_RGB_TYPE_MASK;
 
-	DRM_DEBUG_KMS("width[%d]range[%d]\n", pos->w, sc->range);
-
-	if (pos->w >= GSC_WIDTH_ITU_709)
+	if (buf->rect.w >= GSC_WIDTH_ITU_709)
 		if (sc->range)
 			cfg |= GSC_OUT_RGB_HD_WIDE;
 		else
@@ -1091,8 +887,6 @@ static int gsc_dst_set_size(struct device *dev, int swap,
 			cfg |= GSC_OUT_RGB_SD_NARROW;
 
 	gsc_write(cfg, GSC_OUT_CON);
-
-	return 0;
 }
 
 static int gsc_dst_get_buf_seq(struct gsc_context *ctx)
@@ -1111,35 +905,16 @@ static int gsc_dst_get_buf_seq(struct gsc_context *ctx)
 	return buf_num;
 }
 
-static int gsc_dst_set_buf_seq(struct gsc_context *ctx, u32 buf_id,
-		enum drm_exynos_ipp_buf_type buf_type)
+static void gsc_dst_set_buf_seq(struct gsc_context *ctx, u32 buf_id,
+				bool enqueue)
 {
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	bool masked;
+	bool masked = !enqueue;
 	u32 cfg;
 	u32 mask = 0x00000001 << buf_id;
-	int ret = 0;
-
-	DRM_DEBUG_KMS("buf_id[%d]buf_type[%d]\n", buf_id, buf_type);
-
-	mutex_lock(&ctx->lock);
 
 	/* mask register set */
 	cfg = gsc_read(GSC_OUT_BASE_ADDR_Y_MASK);
 
-	switch (buf_type) {
-	case IPP_BUF_ENQUEUE:
-		masked = false;
-		break;
-	case IPP_BUF_DEQUEUE:
-		masked = true;
-		break;
-	default:
-		dev_err(ippdrv->dev, "invalid buf ctrl parameter.\n");
-		ret =  -EINVAL;
-		goto err_unlock;
-	}
-
 	/* sequence id */
 	cfg &= ~mask;
 	cfg |= masked << buf_id;
@@ -1148,94 +923,29 @@ static int gsc_dst_set_buf_seq(struct gsc_context *ctx, u32 buf_id,
 	gsc_write(cfg, GSC_OUT_BASE_ADDR_CR_MASK);
 
 	/* interrupt enable */
-	if (buf_type == IPP_BUF_ENQUEUE &&
-	    gsc_dst_get_buf_seq(ctx) >= GSC_BUF_START)
+	if (enqueue && gsc_dst_get_buf_seq(ctx) >= GSC_BUF_START)
 		gsc_handle_irq(ctx, true, false, true);
 
 	/* interrupt disable */
-	if (buf_type == IPP_BUF_DEQUEUE &&
-	    gsc_dst_get_buf_seq(ctx) <= GSC_BUF_STOP)
+	if (!enqueue && gsc_dst_get_buf_seq(ctx) <= GSC_BUF_STOP)
 		gsc_handle_irq(ctx, false, false, true);
-
-err_unlock:
-	mutex_unlock(&ctx->lock);
-	return ret;
 }
 
-static int gsc_dst_set_addr(struct device *dev,
-		struct drm_exynos_ipp_buf_info *buf_info, u32 buf_id,
-		enum drm_exynos_ipp_buf_type buf_type)
+static void gsc_dst_set_addr(struct gsc_context *ctx,
+			     u32 buf_id, struct exynos_drm_ipp_buffer *buf)
 {
-	struct gsc_context *ctx = get_gsc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	struct drm_exynos_ipp_cmd_node *c_node = ippdrv->c_node;
-	struct drm_exynos_ipp_property *property;
-
-	if (!c_node) {
-		DRM_ERROR("failed to get c_node.\n");
-		return -EFAULT;
-	}
-
-	property = &c_node->property;
-
-	DRM_DEBUG_KMS("prop_id[%d]buf_id[%d]buf_type[%d]\n",
-		property->prop_id, buf_id, buf_type);
-
-	if (buf_id > GSC_MAX_DST) {
-		dev_info(ippdrv->dev, "invalid buf_id %d.\n", buf_id);
-		return -EINVAL;
-	}
-
 	/* address register set */
-	switch (buf_type) {
-	case IPP_BUF_ENQUEUE:
-		gsc_write(buf_info->base[EXYNOS_DRM_PLANAR_Y],
-			GSC_OUT_BASE_ADDR_Y(buf_id));
-		gsc_write(buf_info->base[EXYNOS_DRM_PLANAR_CB],
-			GSC_OUT_BASE_ADDR_CB(buf_id));
-		gsc_write(buf_info->base[EXYNOS_DRM_PLANAR_CR],
-			GSC_OUT_BASE_ADDR_CR(buf_id));
-		break;
-	case IPP_BUF_DEQUEUE:
-		gsc_write(0x0, GSC_OUT_BASE_ADDR_Y(buf_id));
-		gsc_write(0x0, GSC_OUT_BASE_ADDR_CB(buf_id));
-		gsc_write(0x0, GSC_OUT_BASE_ADDR_CR(buf_id));
-		break;
-	default:
-		/* bypass */
-		break;
-	}
+	gsc_write(buf->dma_addr[0], GSC_OUT_BASE_ADDR_Y(buf_id));
+	gsc_write(buf->dma_addr[1], GSC_OUT_BASE_ADDR_CB(buf_id));
+	gsc_write(buf->dma_addr[2], GSC_OUT_BASE_ADDR_CR(buf_id));
 
-	return gsc_dst_set_buf_seq(ctx, buf_id, buf_type);
-}
-
-static struct exynos_drm_ipp_ops gsc_dst_ops = {
-	.set_fmt = gsc_dst_set_fmt,
-	.set_transf = gsc_dst_set_transf,
-	.set_size = gsc_dst_set_size,
-	.set_addr = gsc_dst_set_addr,
-};
-
-static int gsc_clk_ctrl(struct gsc_context *ctx, bool enable)
-{
-	DRM_DEBUG_KMS("enable[%d]\n", enable);
-
-	if (enable) {
-		clk_prepare_enable(ctx->gsc_clk);
-		ctx->suspended = false;
-	} else {
-		clk_disable_unprepare(ctx->gsc_clk);
-		ctx->suspended = true;
-	}
-
-	return 0;
+	gsc_dst_set_buf_seq(ctx, buf_id, true);
 }
 
 static int gsc_get_src_buf_index(struct gsc_context *ctx)
 {
 	u32 cfg, curr_index, i;
 	u32 buf_id = GSC_MAX_SRC;
-	int ret;
 
 	DRM_DEBUG_KMS("gsc id[%d]\n", ctx->id);
 
@@ -1249,19 +959,15 @@ static int gsc_get_src_buf_index(struct gsc_context *ctx)
 		}
 	}
 
+	DRM_DEBUG_KMS("cfg[0x%x]curr_index[%d]buf_id[%d]\n", cfg,
+		curr_index, buf_id);
+
 	if (buf_id == GSC_MAX_SRC) {
 		DRM_ERROR("failed to get in buffer index.\n");
 		return -EINVAL;
 	}
 
-	ret = gsc_src_set_buf_seq(ctx, buf_id, IPP_BUF_DEQUEUE);
-	if (ret < 0) {
-		DRM_ERROR("failed to dequeue.\n");
-		return ret;
-	}
-
-	DRM_DEBUG_KMS("cfg[0x%x]curr_index[%d]buf_id[%d]\n", cfg,
-		curr_index, buf_id);
+	gsc_src_set_buf_seq(ctx, buf_id, false);
 
 	return buf_id;
 }
@@ -1270,7 +976,6 @@ static int gsc_get_dst_buf_index(struct gsc_context *ctx)
 {
 	u32 cfg, curr_index, i;
 	u32 buf_id = GSC_MAX_DST;
-	int ret;
 
 	DRM_DEBUG_KMS("gsc id[%d]\n", ctx->id);
 
@@ -1289,11 +994,7 @@ static int gsc_get_dst_buf_index(struct gsc_context *ctx)
 		return -EINVAL;
 	}
 
-	ret = gsc_dst_set_buf_seq(ctx, buf_id, IPP_BUF_DEQUEUE);
-	if (ret < 0) {
-		DRM_ERROR("failed to dequeue.\n");
-		return ret;
-	}
+	gsc_dst_set_buf_seq(ctx, buf_id, false);
 
 	DRM_DEBUG_KMS("cfg[0x%x]curr_index[%d]buf_id[%d]\n", cfg,
 		curr_index, buf_id);
@@ -1304,215 +1005,55 @@ static int gsc_get_dst_buf_index(struct gsc_context *ctx)
 static irqreturn_t gsc_irq_handler(int irq, void *dev_id)
 {
 	struct gsc_context *ctx = dev_id;
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	struct drm_exynos_ipp_cmd_node *c_node = ippdrv->c_node;
-	struct drm_exynos_ipp_event_work *event_work =
-		c_node->event_work;
 	u32 status;
-	int buf_id[EXYNOS_DRM_OPS_MAX];
+	int err = 0;
 
 	DRM_DEBUG_KMS("gsc id[%d]\n", ctx->id);
 
 	status = gsc_read(GSC_IRQ);
 	if (status & GSC_IRQ_STATUS_OR_IRQ) {
-		dev_err(ippdrv->dev, "occurred overflow at %d, status 0x%x.\n",
+		dev_err(ctx->dev, "occurred overflow at %d, status 0x%x.\n",
 			ctx->id, status);
-		return IRQ_NONE;
+		err = -EINVAL;
 	}
 
 	if (status & GSC_IRQ_STATUS_OR_FRM_DONE) {
-		dev_dbg(ippdrv->dev, "occurred frame done at %d, status 0x%x.\n",
-			ctx->id, status);
-
-		buf_id[EXYNOS_DRM_OPS_SRC] = gsc_get_src_buf_index(ctx);
-		if (buf_id[EXYNOS_DRM_OPS_SRC] < 0)
-			return IRQ_HANDLED;
-
-		buf_id[EXYNOS_DRM_OPS_DST] = gsc_get_dst_buf_index(ctx);
-		if (buf_id[EXYNOS_DRM_OPS_DST] < 0)
-			return IRQ_HANDLED;
-
-		DRM_DEBUG_KMS("buf_id_src[%d]buf_id_dst[%d]\n",
-			buf_id[EXYNOS_DRM_OPS_SRC], buf_id[EXYNOS_DRM_OPS_DST]);
-
-		event_work->ippdrv = ippdrv;
-		event_work->buf_id[EXYNOS_DRM_OPS_SRC] =
-			buf_id[EXYNOS_DRM_OPS_SRC];
-		event_work->buf_id[EXYNOS_DRM_OPS_DST] =
-			buf_id[EXYNOS_DRM_OPS_DST];
-		queue_work(ippdrv->event_workq, &event_work->work);
-	}
-
-	return IRQ_HANDLED;
-}
-
-static int gsc_init_prop_list(struct exynos_drm_ippdrv *ippdrv)
-{
-	struct drm_exynos_ipp_prop_list *prop_list = &ippdrv->prop_list;
-
-	prop_list->version = 1;
-	prop_list->writeback = 1;
-	prop_list->refresh_min = GSC_REFRESH_MIN;
-	prop_list->refresh_max = GSC_REFRESH_MAX;
-	prop_list->flip = (1 << EXYNOS_DRM_FLIP_VERTICAL) |
-				(1 << EXYNOS_DRM_FLIP_HORIZONTAL);
-	prop_list->degree = (1 << EXYNOS_DRM_DEGREE_0) |
-				(1 << EXYNOS_DRM_DEGREE_90) |
-				(1 << EXYNOS_DRM_DEGREE_180) |
-				(1 << EXYNOS_DRM_DEGREE_270);
-	prop_list->csc = 1;
-	prop_list->crop = 1;
-	prop_list->crop_max.hsize = GSC_CROP_MAX;
-	prop_list->crop_max.vsize = GSC_CROP_MAX;
-	prop_list->crop_min.hsize = GSC_CROP_MIN;
-	prop_list->crop_min.vsize = GSC_CROP_MIN;
-	prop_list->scale = 1;
-	prop_list->scale_max.hsize = GSC_SCALE_MAX;
-	prop_list->scale_max.vsize = GSC_SCALE_MAX;
-	prop_list->scale_min.hsize = GSC_SCALE_MIN;
-	prop_list->scale_min.vsize = GSC_SCALE_MIN;
-
-	return 0;
-}
-
-static inline bool gsc_check_drm_flip(enum drm_exynos_flip flip)
-{
-	switch (flip) {
-	case EXYNOS_DRM_FLIP_NONE:
-	case EXYNOS_DRM_FLIP_VERTICAL:
-	case EXYNOS_DRM_FLIP_HORIZONTAL:
-	case EXYNOS_DRM_FLIP_BOTH:
-		return true;
-	default:
-		DRM_DEBUG_KMS("invalid flip\n");
-		return false;
-	}
-}
-
-static int gsc_ippdrv_check_property(struct device *dev,
-		struct drm_exynos_ipp_property *property)
-{
-	struct gsc_context *ctx = get_gsc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	struct drm_exynos_ipp_prop_list *pp = &ippdrv->prop_list;
-	struct drm_exynos_ipp_config *config;
-	struct drm_exynos_pos *pos;
-	struct drm_exynos_sz *sz;
-	bool swap;
-	int i;
-
-	for_each_ipp_ops(i) {
-		if ((i == EXYNOS_DRM_OPS_SRC) &&
-			(property->cmd == IPP_CMD_WB))
-			continue;
+		int src_buf_id, dst_buf_id;
 
-		config = &property->config[i];
-		pos = &config->pos;
-		sz = &config->sz;
-
-		/* check for flip */
-		if (!gsc_check_drm_flip(config->flip)) {
-			DRM_ERROR("invalid flip.\n");
-			goto err_property;
-		}
-
-		/* check for degree */
-		switch (config->degree) {
-		case EXYNOS_DRM_DEGREE_90:
-		case EXYNOS_DRM_DEGREE_270:
-			swap = true;
-			break;
-		case EXYNOS_DRM_DEGREE_0:
-		case EXYNOS_DRM_DEGREE_180:
-			swap = false;
-			break;
-		default:
-			DRM_ERROR("invalid degree.\n");
-			goto err_property;
-		}
+		dev_dbg(ctx->dev, "occurred frame done at %d, status 0x%x.\n",
+			ctx->id, status);
 
-		/* check for buffer bound */
-		if ((pos->x + pos->w > sz->hsize) ||
-			(pos->y + pos->h > sz->vsize)) {
-			DRM_ERROR("out of buf bound.\n");
-			goto err_property;
-		}
+		src_buf_id = gsc_get_src_buf_index(ctx);
+		dst_buf_id = gsc_get_dst_buf_index(ctx);
 
-		/* check for crop */
-		if ((i == EXYNOS_DRM_OPS_SRC) && (pp->crop)) {
-			if (swap) {
-				if ((pos->h < pp->crop_min.hsize) ||
-					(sz->vsize > pp->crop_max.hsize) ||
-					(pos->w < pp->crop_min.vsize) ||
-					(sz->hsize > pp->crop_max.vsize)) {
-					DRM_ERROR("out of crop size.\n");
-					goto err_property;
-				}
-			} else {
-				if ((pos->w < pp->crop_min.hsize) ||
-					(sz->hsize > pp->crop_max.hsize) ||
-					(pos->h < pp->crop_min.vsize) ||
-					(sz->vsize > pp->crop_max.vsize)) {
-					DRM_ERROR("out of crop size.\n");
-					goto err_property;
-				}
-			}
-		}
+		DRM_DEBUG_KMS("buf_id_src[%d]buf_id_dst[%d]\n",	src_buf_id,
+			      dst_buf_id);
 
-		/* check for scale */
-		if ((i == EXYNOS_DRM_OPS_DST) && (pp->scale)) {
-			if (swap) {
-				if ((pos->h < pp->scale_min.hsize) ||
-					(sz->vsize > pp->scale_max.hsize) ||
-					(pos->w < pp->scale_min.vsize) ||
-					(sz->hsize > pp->scale_max.vsize)) {
-					DRM_ERROR("out of scale size.\n");
-					goto err_property;
-				}
-			} else {
-				if ((pos->w < pp->scale_min.hsize) ||
-					(sz->hsize > pp->scale_max.hsize) ||
-					(pos->h < pp->scale_min.vsize) ||
-					(sz->vsize > pp->scale_max.vsize)) {
-					DRM_ERROR("out of scale size.\n");
-					goto err_property;
-				}
-			}
-		}
+		if (src_buf_id < 0 || dst_buf_id < 0)
+			err = -EINVAL;
 	}
 
-	return 0;
-
-err_property:
-	for_each_ipp_ops(i) {
-		if ((i == EXYNOS_DRM_OPS_SRC) &&
-			(property->cmd == IPP_CMD_WB))
-			continue;
+	if (ctx->task) {
+		struct exynos_drm_ipp_task *task = ctx->task;
 
-		config = &property->config[i];
-		pos = &config->pos;
-		sz = &config->sz;
-
-		DRM_ERROR("[%s]f[%d]r[%d]pos[%d %d %d %d]sz[%d %d]\n",
-			i ? "dst" : "src", config->flip, config->degree,
-			pos->x, pos->y, pos->w, pos->h,
-			sz->hsize, sz->vsize);
+		ctx->task = NULL;
+		pm_runtime_mark_last_busy(ctx->dev);
+		pm_runtime_put_autosuspend(ctx->dev);
+		exynos_drm_ipp_task_done(task, err);
 	}
 
-	return -EINVAL;
+	return IRQ_HANDLED;
 }
 
-
-static int gsc_ippdrv_reset(struct device *dev)
+static int gsc_reset(struct gsc_context *ctx)
 {
-	struct gsc_context *ctx = get_gsc_context(dev);
 	struct gsc_scaler *sc = &ctx->sc;
 	int ret;
 
 	/* reset h/w block */
 	ret = gsc_sw_reset(ctx);
 	if (ret < 0) {
-		dev_err(dev, "failed to reset hardware.\n");
+		dev_err(ctx->dev, "failed to reset hardware.\n");
 		return ret;
 	}
 
@@ -1523,166 +1064,172 @@ static int gsc_ippdrv_reset(struct device *dev)
 	return 0;
 }
 
-static int gsc_ippdrv_start(struct device *dev, enum drm_exynos_ipp_cmd cmd)
+static void gsc_start(struct gsc_context *ctx)
 {
-	struct gsc_context *ctx = get_gsc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	struct drm_exynos_ipp_cmd_node *c_node = ippdrv->c_node;
-	struct drm_exynos_ipp_property *property;
-	struct drm_exynos_ipp_config *config;
-	struct drm_exynos_pos	img_pos[EXYNOS_DRM_OPS_MAX];
-	struct drm_exynos_ipp_set_wb set_wb;
 	u32 cfg;
-	int ret, i;
-
-	DRM_DEBUG_KMS("cmd[%d]\n", cmd);
-
-	if (!c_node) {
-		DRM_ERROR("failed to get c_node.\n");
-		return -EINVAL;
-	}
-
-	property = &c_node->property;
 
 	gsc_handle_irq(ctx, true, false, true);
 
-	for_each_ipp_ops(i) {
-		config = &property->config[i];
-		img_pos[i] = config->pos;
-	}
+	/* enable one shot */
+	cfg = gsc_read(GSC_ENABLE);
+	cfg &= ~(GSC_ENABLE_ON_CLEAR_MASK |
+		GSC_ENABLE_CLK_GATE_MODE_MASK);
+	cfg |= GSC_ENABLE_ON_CLEAR_ONESHOT;
+	gsc_write(cfg, GSC_ENABLE);
 
-	switch (cmd) {
-	case IPP_CMD_M2M:
-		/* enable one shot */
-		cfg = gsc_read(GSC_ENABLE);
-		cfg &= ~(GSC_ENABLE_ON_CLEAR_MASK |
-			GSC_ENABLE_CLK_GATE_MODE_MASK);
-		cfg |= GSC_ENABLE_ON_CLEAR_ONESHOT;
-		gsc_write(cfg, GSC_ENABLE);
-
-		/* src dma memory */
-		cfg = gsc_read(GSC_IN_CON);
-		cfg &= ~(GSC_IN_PATH_MASK | GSC_IN_LOCAL_SEL_MASK);
-		cfg |= GSC_IN_PATH_MEMORY;
-		gsc_write(cfg, GSC_IN_CON);
-
-		/* dst dma memory */
-		cfg = gsc_read(GSC_OUT_CON);
-		cfg |= GSC_OUT_PATH_MEMORY;
-		gsc_write(cfg, GSC_OUT_CON);
-		break;
-	case IPP_CMD_WB:
-		set_wb.enable = 1;
-		set_wb.refresh = property->refresh_rate;
-		gsc_set_gscblk_fimd_wb(ctx, set_wb.enable);
-		exynos_drm_ippnb_send_event(IPP_SET_WRITEBACK, (void *)&set_wb);
-
-		/* src local path */
-		cfg = gsc_read(GSC_IN_CON);
-		cfg &= ~(GSC_IN_PATH_MASK | GSC_IN_LOCAL_SEL_MASK);
-		cfg |= (GSC_IN_PATH_LOCAL | GSC_IN_LOCAL_FIMD_WB);
-		gsc_write(cfg, GSC_IN_CON);
-
-		/* dst dma memory */
-		cfg = gsc_read(GSC_OUT_CON);
-		cfg |= GSC_OUT_PATH_MEMORY;
-		gsc_write(cfg, GSC_OUT_CON);
-		break;
-	case IPP_CMD_OUTPUT:
-		/* src dma memory */
-		cfg = gsc_read(GSC_IN_CON);
-		cfg &= ~(GSC_IN_PATH_MASK | GSC_IN_LOCAL_SEL_MASK);
-		cfg |= GSC_IN_PATH_MEMORY;
-		gsc_write(cfg, GSC_IN_CON);
-
-		/* dst local path */
-		cfg = gsc_read(GSC_OUT_CON);
-		cfg |= GSC_OUT_PATH_MEMORY;
-		gsc_write(cfg, GSC_OUT_CON);
-		break;
-	default:
-		ret = -EINVAL;
-		dev_err(dev, "invalid operations.\n");
-		return ret;
-	}
+	/* src dma memory */
+	cfg = gsc_read(GSC_IN_CON);
+	cfg &= ~(GSC_IN_PATH_MASK | GSC_IN_LOCAL_SEL_MASK);
+	cfg |= GSC_IN_PATH_MEMORY;
+	gsc_write(cfg, GSC_IN_CON);
 
-	ret = gsc_set_prescaler(ctx, &ctx->sc,
-		&img_pos[EXYNOS_DRM_OPS_SRC],
-		&img_pos[EXYNOS_DRM_OPS_DST]);
-	if (ret) {
-		dev_err(dev, "failed to set prescaler.\n");
-		return ret;
-	}
+	/* dst dma memory */
+	cfg = gsc_read(GSC_OUT_CON);
+	cfg |= GSC_OUT_PATH_MEMORY;
+	gsc_write(cfg, GSC_OUT_CON);
 
 	gsc_set_scaler(ctx, &ctx->sc);
 
 	cfg = gsc_read(GSC_ENABLE);
 	cfg |= GSC_ENABLE_ON;
 	gsc_write(cfg, GSC_ENABLE);
+}
+
+static int gsc_commit(struct exynos_drm_ipp *ipp,
+			  struct exynos_drm_ipp_task *task)
+{
+	struct gsc_context *ctx = container_of(ipp, struct gsc_context, ipp);
+	int ret;
+
+	pm_runtime_get_sync(ctx->dev);
+	ctx->task = task;
+
+	ret = gsc_reset(ctx);
+	if (ret) {
+		pm_runtime_put_autosuspend(ctx->dev);
+		ctx->task = NULL;
+		return ret;
+	}
+
+	gsc_src_set_fmt(ctx, task->src.buf.fourcc);
+	gsc_src_set_transf(ctx, task->transform.rotation);
+	gsc_src_set_size(ctx, &task->src);
+	gsc_src_set_addr(ctx, 0, &task->src);
+	gsc_dst_set_fmt(ctx, task->dst.buf.fourcc);
+	gsc_dst_set_size(ctx, &task->dst);
+	gsc_dst_set_addr(ctx, 0, &task->dst);
+	gsc_set_prescaler(ctx, &ctx->sc, &task->src.rect, &task->dst.rect);
+	gsc_start(ctx);
 
 	return 0;
 }
 
-static void gsc_ippdrv_stop(struct device *dev, enum drm_exynos_ipp_cmd cmd)
+static void gsc_abort(struct exynos_drm_ipp *ipp,
+			  struct exynos_drm_ipp_task *task)
 {
-	struct gsc_context *ctx = get_gsc_context(dev);
-	struct drm_exynos_ipp_set_wb set_wb = {0, 0};
-	u32 cfg;
+	struct gsc_context *ctx =
+			container_of(ipp, struct gsc_context, ipp);
 
-	DRM_DEBUG_KMS("cmd[%d]\n", cmd);
+	gsc_reset(ctx);
+	if (ctx->task) {
+		struct exynos_drm_ipp_task *task = ctx->task;
 
-	switch (cmd) {
-	case IPP_CMD_M2M:
-		/* bypass */
-		break;
-	case IPP_CMD_WB:
-		gsc_set_gscblk_fimd_wb(ctx, set_wb.enable);
-		exynos_drm_ippnb_send_event(IPP_SET_WRITEBACK, (void *)&set_wb);
-		break;
-	case IPP_CMD_OUTPUT:
-	default:
-		dev_err(dev, "invalid operations.\n");
-		break;
+		ctx->task = NULL;
+		pm_runtime_mark_last_busy(ctx->dev);
+		pm_runtime_put_autosuspend(ctx->dev);
+		exynos_drm_ipp_task_done(task, -EIO);
 	}
+}
 
-	gsc_handle_irq(ctx, false, false, true);
+static struct exynos_drm_ipp_funcs ipp_funcs = {
+	.commit = gsc_commit,
+	.abort = gsc_abort,
+};
 
-	/* reset sequence */
-	gsc_write(0xff, GSC_OUT_BASE_ADDR_Y_MASK);
-	gsc_write(0xff, GSC_OUT_BASE_ADDR_CB_MASK);
-	gsc_write(0xff, GSC_OUT_BASE_ADDR_CR_MASK);
+static int gsc_bind(struct device *dev, struct device *master, void *data)
+{
+	struct gsc_context *ctx = dev_get_drvdata(dev);
+	struct drm_device *drm_dev = data;
+	struct exynos_drm_ipp *ipp = &ctx->ipp;
 
-	cfg = gsc_read(GSC_ENABLE);
-	cfg &= ~GSC_ENABLE_ON;
-	gsc_write(cfg, GSC_ENABLE);
+	ctx->drm_dev = drm_dev;
+	drm_iommu_attach_device(drm_dev, dev);
+
+	exynos_drm_ipp_register(drm_dev, ipp, &ipp_funcs,
+			DRM_EXYNOS_IPP_CAP_CROP | DRM_EXYNOS_IPP_CAP_ROTATE |
+			DRM_EXYNOS_IPP_CAP_SCALE | DRM_EXYNOS_IPP_CAP_CONVERT,
+			ctx->formats, ctx->num_formats, "gsc");
+
+	dev_info(dev, "The exynos gscaler has been probed successfully\n");
+
+	return 0;
+}
+
+static void gsc_unbind(struct device *dev, struct device *master,
+			void *data)
+{
+	struct gsc_context *ctx = dev_get_drvdata(dev);
+	struct drm_device *drm_dev = data;
+	struct exynos_drm_ipp *ipp = &ctx->ipp;
+
+	exynos_drm_ipp_unregister(drm_dev, ipp);
+	drm_iommu_detach_device(drm_dev, dev);
 }
 
+static const struct component_ops gsc_component_ops = {
+	.bind	= gsc_bind,
+	.unbind = gsc_unbind,
+};
+
+static const unsigned int gsc_formats[] = {
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_XRGB8888, DRM_FORMAT_RGB565, DRM_FORMAT_BGRX8888,
+	DRM_FORMAT_NV12, DRM_FORMAT_NV16, DRM_FORMAT_NV21, DRM_FORMAT_NV61,
+	DRM_FORMAT_UYVY, DRM_FORMAT_VYUY, DRM_FORMAT_YUYV, DRM_FORMAT_YVYU,
+	DRM_FORMAT_YUV420, DRM_FORMAT_YVU420, DRM_FORMAT_YUV422,
+};
+
 static int gsc_probe(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;
+	struct gsc_driverdata *driver_data;
+	struct exynos_drm_ipp_formats *formats;
 	struct gsc_context *ctx;
 	struct resource *res;
-	struct exynos_drm_ippdrv *ippdrv;
-	int ret;
+	int ret, i;
 
 	ctx = devm_kzalloc(dev, sizeof(*ctx), GFP_KERNEL);
 	if (!ctx)
 		return -ENOMEM;
 
-	if (dev->of_node) {
-		ctx->sysreg = syscon_regmap_lookup_by_phandle(dev->of_node,
-							"samsung,sysreg");
-		if (IS_ERR(ctx->sysreg)) {
-			dev_warn(dev, "failed to get system register.\n");
-			ctx->sysreg = NULL;
-		}
+	formats = devm_kzalloc(dev, sizeof(*formats) *
+			       (ARRAY_SIZE(gsc_formats)), GFP_KERNEL);
+	if (!formats)
+		return -ENOMEM;
+
+	driver_data = (struct gsc_driverdata *)of_device_get_match_data(dev);
+	ctx->dev = dev;
+	ctx->num_clocks = driver_data->num_clocks;
+	ctx->clk_names = driver_data->clk_names;
+
+	for (i = 0; i < ARRAY_SIZE(gsc_formats); i++) {
+		formats[i].fourcc = gsc_formats[i];
+		formats[i].type = DRM_EXYNOS_IPP_FORMAT_SOURCE |
+				  DRM_EXYNOS_IPP_FORMAT_DESTINATION;
+		formats[i].limits = driver_data->limits;
+		formats[i].num_limits = driver_data->num_limits;
 	}
+	ctx->formats = formats;
+	ctx->num_formats = ARRAY_SIZE(gsc_formats);
 
 	/* clock control */
-	ctx->gsc_clk = devm_clk_get(dev, "gscl");
-	if (IS_ERR(ctx->gsc_clk)) {
-		dev_err(dev, "failed to get gsc clock.\n");
-		return PTR_ERR(ctx->gsc_clk);
+	for (i = 0; i < ctx->num_clocks; i++) {
+		ctx->clocks[i] = devm_clk_get(dev, ctx->clk_names[i]);
+		if (IS_ERR(ctx->clocks[i])) {
+			dev_err(dev, "failed to get clock: %s\n",
+				ctx->clk_names[i]);
+			return PTR_ERR(ctx->clocks[i]);
+		}
 	}
 
 	/* resource memory */
@@ -1699,8 +1246,8 @@ static int gsc_probe(struct platform_device *pdev)
 	}
 
 	ctx->irq = res->start;
-	ret = devm_request_threaded_irq(dev, ctx->irq, NULL, gsc_irq_handler,
-		IRQF_ONESHOT, "drm_gsc", ctx);
+	ret = devm_request_irq(dev, ctx->irq, gsc_irq_handler, 0,
+			       dev_name(dev), ctx);
 	if (ret < 0) {
 		dev_err(dev, "failed to request irq.\n");
 		return ret;
@@ -1709,38 +1256,22 @@ static int gsc_probe(struct platform_device *pdev)
 	/* context initailization */
 	ctx->id = pdev->id;
 
-	ippdrv = &ctx->ippdrv;
-	ippdrv->dev = dev;
-	ippdrv->ops[EXYNOS_DRM_OPS_SRC] = &gsc_src_ops;
-	ippdrv->ops[EXYNOS_DRM_OPS_DST] = &gsc_dst_ops;
-	ippdrv->check_property = gsc_ippdrv_check_property;
-	ippdrv->reset = gsc_ippdrv_reset;
-	ippdrv->start = gsc_ippdrv_start;
-	ippdrv->stop = gsc_ippdrv_stop;
-	ret = gsc_init_prop_list(ippdrv);
-	if (ret < 0) {
-		dev_err(dev, "failed to init property list.\n");
-		return ret;
-	}
-
-	DRM_DEBUG_KMS("id[%d]ippdrv[%pK]\n", ctx->id, ippdrv);
-
-	mutex_init(&ctx->lock);
 	platform_set_drvdata(pdev, ctx);
 
+	pm_runtime_use_autosuspend(dev);
+	pm_runtime_set_autosuspend_delay(dev, GSC_AUTOSUSPEND_DELAY);
 	pm_runtime_enable(dev);
 
-	ret = exynos_drm_ippdrv_register(ippdrv);
-	if (ret < 0) {
-		dev_err(dev, "failed to register drm gsc device.\n");
-		goto err_ippdrv_register;
-	}
+	ret = component_add(dev, &gsc_component_ops);
+	if (ret)
+		goto err_pm_dis;
 
 	dev_info(dev, "drm gsc registered successfully.\n");
 
 	return 0;
 
-err_ippdrv_register:
+err_pm_dis:
+	pm_runtime_dont_use_autosuspend(dev);
 	pm_runtime_disable(dev);
 	return ret;
 }
@@ -1748,13 +1279,8 @@ err_ippdrv_register:
 static int gsc_remove(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;
-	struct gsc_context *ctx = get_gsc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
 
-	exynos_drm_ippdrv_unregister(ippdrv);
-	mutex_destroy(&ctx->lock);
-
-	pm_runtime_set_suspended(dev);
+	pm_runtime_dont_use_autosuspend(dev);
 	pm_runtime_disable(dev);
 
 	return 0;
@@ -1763,19 +1289,32 @@ static int gsc_remove(struct platform_device *pdev)
 static int __maybe_unused gsc_runtime_suspend(struct device *dev)
 {
 	struct gsc_context *ctx = get_gsc_context(dev);
+	int i;
 
 	DRM_DEBUG_KMS("id[%d]\n", ctx->id);
 
-	return  gsc_clk_ctrl(ctx, false);
+	for (i = ctx->num_clocks - 1; i >= 0; i--)
+		clk_disable_unprepare(ctx->clocks[i]);
+
+	return 0;
 }
 
 static int __maybe_unused gsc_runtime_resume(struct device *dev)
 {
 	struct gsc_context *ctx = get_gsc_context(dev);
+	int i, ret;
 
 	DRM_DEBUG_KMS("id[%d]\n", ctx->id);
 
-	return  gsc_clk_ctrl(ctx, true);
+	for (i = 0; i < ctx->num_clocks; i++) {
+		ret = clk_prepare_enable(ctx->clocks[i]);
+		if (ret) {
+			while (--i > 0)
+				clk_disable_unprepare(ctx->clocks[i]);
+			return ret;
+		}
+	}
+	return 0;
 }
 
 static const struct dev_pm_ops gsc_pm_ops = {
@@ -1784,9 +1323,66 @@ static const struct dev_pm_ops gsc_pm_ops = {
 	SET_RUNTIME_PM_OPS(gsc_runtime_suspend, gsc_runtime_resume, NULL)
 };
 
+static const struct drm_exynos_ipp_limit gsc_5250_limits[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 32, 4800, 8 }, .v = { 16, 3344, 8 }) },
+	{ IPP_SIZE_LIMIT(AREA, .h = { 16, 4800, 2 }, .v = { 8, 3344, 2 }) },
+	{ IPP_SIZE_LIMIT(ROTATED, .h = { 32, 2048 }, .v = { 16, 2048 }) },
+	{ IPP_SCALE_LIMIT(.h = { (1 << 16) / 16, (1 << 16) * 8 },
+			  .v = { (1 << 16) / 16, (1 << 16) * 8 }) },
+};
+
+static const struct drm_exynos_ipp_limit gsc_5420_limits[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 32, 4800, 8 }, .v = { 16, 3344, 8 }) },
+	{ IPP_SIZE_LIMIT(AREA, .h = { 16, 4800, 2 }, .v = { 8, 3344, 2 }) },
+	{ IPP_SIZE_LIMIT(ROTATED, .h = { 16, 2016 }, .v = { 8, 2016 }) },
+	{ IPP_SCALE_LIMIT(.h = { (1 << 16) / 16, (1 << 16) * 8 },
+			  .v = { (1 << 16) / 16, (1 << 16) * 8 }) },
+};
+
+static const struct drm_exynos_ipp_limit gsc_5433_limits[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 32, 8191, 2 }, .v = { 16, 8191, 2 }) },
+	{ IPP_SIZE_LIMIT(AREA, .h = { 16, 4800, 1 }, .v = { 8, 3344, 1 }) },
+	{ IPP_SIZE_LIMIT(ROTATED, .h = { 32, 2047 }, .v = { 8, 8191 }) },
+	{ IPP_SCALE_LIMIT(.h = { (1 << 16) / 16, (1 << 16) * 8 },
+			  .v = { (1 << 16) / 16, (1 << 16) * 8 }) },
+};
+
+static struct gsc_driverdata gsc_exynos5250_drvdata = {
+	.clk_names = {"gscl"},
+	.num_clocks = 1,
+	.limits = gsc_5250_limits,
+	.num_limits = ARRAY_SIZE(gsc_5250_limits),
+};
+
+static struct gsc_driverdata gsc_exynos5420_drvdata = {
+	.clk_names = {"gscl"},
+	.num_clocks = 1,
+	.limits = gsc_5420_limits,
+	.num_limits = ARRAY_SIZE(gsc_5420_limits),
+};
+
+static struct gsc_driverdata gsc_exynos5433_drvdata = {
+	.clk_names = {"pclk", "aclk", "aclk_xiu", "aclk_gsclbend"},
+	.num_clocks = 4,
+	.limits = gsc_5433_limits,
+	.num_limits = ARRAY_SIZE(gsc_5433_limits),
+};
+
 static const struct of_device_id exynos_drm_gsc_of_match[] = {
-	{ .compatible = "samsung,exynos5-gsc" },
-	{ },
+	{
+		.compatible = "samsung,exynos5-gsc",
+		.data = &gsc_exynos5250_drvdata,
+	}, {
+		.compatible = "samsung,exynos5250-gsc",
+		.data = &gsc_exynos5250_drvdata,
+	}, {
+		.compatible = "samsung,exynos5420-gsc",
+		.data = &gsc_exynos5420_drvdata,
+	}, {
+		.compatible = "samsung,exynos5433-gsc",
+		.data = &gsc_exynos5433_drvdata,
+	}, {
+	},
 };
 MODULE_DEVICE_TABLE(of, exynos_drm_gsc_of_match);
 
@@ -1800,4 +1396,3 @@ struct platform_driver gsc_driver = {
 		.of_match_table = of_match_ptr(exynos_drm_gsc_of_match),
 	},
 };
-
diff --git a/drivers/gpu/drm/exynos/exynos_drm_gsc.h b/drivers/gpu/drm/exynos/exynos_drm_gsc.h
deleted file mode 100644
index 29ec1c5..0000000
--- a/drivers/gpu/drm/exynos/exynos_drm_gsc.h
+++ /dev/null
@@ -1,24 +0,0 @@
-/*
- * Copyright (c) 2012 Samsung Electronics Co., Ltd.
- *
- * Authors:
- *	Eunchul Kim <chulspro.kim@samsung.com>
- *	Jinyoung Jeon <jy0.jeon@samsung.com>
- *	Sangmin Lee <lsmin.lee@samsung.com>
- *
- * This program is free software; you can redistribute  it and/or modify it
- * under  the terms of  the GNU General  Public License as published by the
- * Free Software Foundation;  either version 2 of the  License, or (at your
- * option) any later version.
- */
-
-#ifndef _EXYNOS_DRM_GSC_H_
-#define _EXYNOS_DRM_GSC_H_
-
-/*
- * TODO
- * FIMD output interface notifier callback.
- * Mixer output interface notifier callback.
- */
-
-#endif /* _EXYNOS_DRM_GSC_H_ */
-- 
cgit v1.1


From 7a2d5c77c55847f31945e5aa8337db2218a5a7c1 Mon Sep 17 00:00:00 2001
From: Marek Szyprowski <m.szyprowski@samsung.com>
Date: Thu, 10 May 2018 08:52:12 +0900
Subject: drm/exynos: fimc: Convert driver to IPP v2 core API

This patch adapts Exynos DRM FIMC driver to new IPP v2 core API.
The side effect of this conversion is a switch to driver component API
to register properly in the Exynos DRM core.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Merge conflict so merged manually.
Signed-off-by: Inki Dae <inki.dae@samsung.com>
---
 drivers/gpu/drm/exynos/Kconfig           |    2 +-
 drivers/gpu/drm/exynos/exynos_drm_drv.c  |    8 +-
 drivers/gpu/drm/exynos/exynos_drm_drv.h  |    9 +
 drivers/gpu/drm/exynos/exynos_drm_fimc.c | 1080 ++++++++++--------------------
 drivers/gpu/drm/exynos/exynos_drm_fimc.h |   23 -
 5 files changed, 366 insertions(+), 756 deletions(-)
 delete mode 100644 drivers/gpu/drm/exynos/exynos_drm_fimc.h

diff --git a/drivers/gpu/drm/exynos/Kconfig b/drivers/gpu/drm/exynos/Kconfig
index 5c21654..54f5703 100644
--- a/drivers/gpu/drm/exynos/Kconfig
+++ b/drivers/gpu/drm/exynos/Kconfig
@@ -100,7 +100,7 @@ config DRM_EXYNOS_IPP
 
 config DRM_EXYNOS_FIMC
 	bool "FIMC"
-	depends on BROKEN && MFD_SYSCON
+	select DRM_EXYNOS_IPP
 	help
 	  Choose this option if you want to use Exynos FIMC for DRM.
 
diff --git a/drivers/gpu/drm/exynos/exynos_drm_drv.c b/drivers/gpu/drm/exynos/exynos_drm_drv.c
index 2dcb940..7ba13c1 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_drv.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_drv.c
@@ -222,6 +222,7 @@ struct exynos_drm_driver_info {
 #define DRM_COMPONENT_DRIVER	BIT(0)	/* supports component framework */
 #define DRM_VIRTUAL_DEVICE	BIT(1)	/* create virtual platform device */
 #define DRM_DMA_DEVICE		BIT(2)	/* can be used for dma allocations */
+#define DRM_FIMC_DEVICE		BIT(3)	/* devices shared with V4L2 subsystem */
 
 #define DRV_PTR(drv, cond) (IS_ENABLED(cond) ? &drv : NULL)
 
@@ -261,6 +262,7 @@ static struct exynos_drm_driver_info exynos_drm_drivers[] = {
 		DRV_PTR(g2d_driver, CONFIG_DRM_EXYNOS_G2D),
 	}, {
 		DRV_PTR(fimc_driver, CONFIG_DRM_EXYNOS_FIMC),
+		DRM_COMPONENT_DRIVER | DRM_FIMC_DEVICE,
 	}, {
 		DRV_PTR(rotator_driver, CONFIG_DRM_EXYNOS_ROTATOR),
 		DRM_COMPONENT_DRIVER
@@ -294,7 +296,11 @@ static struct component_match *exynos_drm_match_add(struct device *dev)
 					    &info->driver->driver,
 					    (void *)platform_bus_type.match))) {
 			put_device(p);
-			component_match_add(dev, &match, compare_dev, d);
+
+			if (!(info->flags & DRM_FIMC_DEVICE) ||
+			    exynos_drm_check_fimc_device(d) == 0)
+				component_match_add(dev, &match,
+						    compare_dev, d);
 			p = d;
 		}
 		put_device(p);
diff --git a/drivers/gpu/drm/exynos/exynos_drm_drv.h b/drivers/gpu/drm/exynos/exynos_drm_drv.h
index df2262f..0834e7e 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_drv.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_drv.h
@@ -273,6 +273,15 @@ static inline int exynos_dpi_bind(struct drm_device *dev,
 }
 #endif
 
+#ifdef CONFIG_DRM_EXYNOS_FIMC
+int exynos_drm_check_fimc_device(struct device *dev);
+#else
+static inline int exynos_drm_check_fimc_device(struct device *dev)
+{
+	return 0;
+}
+#endif
+
 int exynos_atomic_commit(struct drm_device *dev, struct drm_atomic_state *state,
 			 bool nonblock);
 int exynos_atomic_check(struct drm_device *dev, struct drm_atomic_state *state);
diff --git a/drivers/gpu/drm/exynos/exynos_drm_fimc.c b/drivers/gpu/drm/exynos/exynos_drm_fimc.c
index 5b18b5c..4dfbfc7 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_fimc.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_fimc.c
@@ -12,6 +12,7 @@
  *
  */
 #include <linux/kernel.h>
+#include <linux/component.h>
 #include <linux/platform_device.h>
 #include <linux/mfd/syscon.h>
 #include <linux/regmap.h>
@@ -24,8 +25,8 @@
 #include <drm/exynos_drm.h>
 #include "regs-fimc.h"
 #include "exynos_drm_drv.h"
+#include "exynos_drm_iommu.h"
 #include "exynos_drm_ipp.h"
-#include "exynos_drm_fimc.h"
 
 /*
  * FIMC stands for Fully Interactive Mobile Camera and
@@ -33,23 +34,6 @@
  * input DMA reads image data from the memory.
  * output DMA writes image data to memory.
  * FIMC supports image rotation and image effect functions.
- *
- * M2M operation : supports crop/scale/rotation/csc so on.
- * Memory ----> FIMC H/W ----> Memory.
- * Writeback operation : supports cloned screen with FIMD.
- * FIMD ----> FIMC H/W ----> Memory.
- * Output operation : supports direct display using local path.
- * Memory ----> FIMC H/W ----> FIMD.
- */
-
-/*
- * TODO
- * 1. check suspend/resume api if needed.
- * 2. need to check use case platform_device_id.
- * 3. check src/dst size with, height.
- * 4. added check_prepare api for right register.
- * 5. need to add supported list in prop_list.
- * 6. check prescaler/scaler optimization.
  */
 
 #define FIMC_MAX_DEVS	4
@@ -59,29 +43,19 @@
 #define FIMC_BUF_STOP	1
 #define FIMC_BUF_START	2
 #define FIMC_WIDTH_ITU_709	1280
-#define FIMC_REFRESH_MAX	60
-#define FIMC_REFRESH_MIN	12
-#define FIMC_CROP_MAX	8192
-#define FIMC_CROP_MIN	32
-#define FIMC_SCALE_MAX	4224
-#define FIMC_SCALE_MIN	32
+#define FIMC_AUTOSUSPEND_DELAY	2000
+
+static unsigned int fimc_mask = 0xc;
+module_param_named(fimc_devs, fimc_mask, uint, 0644);
+MODULE_PARM_DESC(fimc_devs, "Alias mask for assigning FIMC devices to Exynos DRM");
 
 #define get_fimc_context(dev)	platform_get_drvdata(to_platform_device(dev))
-#define get_ctx_from_ippdrv(ippdrv)	container_of(ippdrv,\
-					struct fimc_context, ippdrv);
-enum fimc_wb {
-	FIMC_WB_NONE,
-	FIMC_WB_A,
-	FIMC_WB_B,
-};
 
 enum {
 	FIMC_CLK_LCLK,
 	FIMC_CLK_GATE,
 	FIMC_CLK_WB_A,
 	FIMC_CLK_WB_B,
-	FIMC_CLK_MUX,
-	FIMC_CLK_PARENT,
 	FIMC_CLKS_MAX
 };
 
@@ -90,12 +64,8 @@ static const char * const fimc_clock_names[] = {
 	[FIMC_CLK_GATE]   = "fimc",
 	[FIMC_CLK_WB_A]   = "pxl_async0",
 	[FIMC_CLK_WB_B]   = "pxl_async1",
-	[FIMC_CLK_MUX]    = "mux",
-	[FIMC_CLK_PARENT] = "parent",
 };
 
-#define FIMC_DEFAULT_LCLK_FREQUENCY 133000000UL
-
 /*
  * A structure of scaler.
  *
@@ -107,7 +77,7 @@ static const char * const fimc_clock_names[] = {
  * @vratio: vertical ratio.
  */
 struct fimc_scaler {
-	bool	range;
+	bool range;
 	bool bypass;
 	bool up_h;
 	bool up_v;
@@ -116,56 +86,32 @@ struct fimc_scaler {
 };
 
 /*
- * A structure of scaler capability.
- *
- * find user manual table 43-1.
- * @in_hori: scaler input horizontal size.
- * @bypass: scaler bypass mode.
- * @dst_h_wo_rot: target horizontal size without output rotation.
- * @dst_h_rot: target horizontal size with output rotation.
- * @rl_w_wo_rot: real width without input rotation.
- * @rl_h_rot: real height without output rotation.
- */
-struct fimc_capability {
-	/* scaler */
-	u32	in_hori;
-	u32	bypass;
-	/* output rotator */
-	u32	dst_h_wo_rot;
-	u32	dst_h_rot;
-	/* input rotator */
-	u32	rl_w_wo_rot;
-	u32	rl_h_rot;
-};
-
-/*
  * A structure of fimc context.
  *
- * @ippdrv: prepare initialization using ippdrv.
  * @regs_res: register resources.
  * @regs: memory mapped io registers.
  * @lock: locking of operations.
  * @clocks: fimc clocks.
- * @clk_frequency: LCLK clock frequency.
- * @sysreg: handle to SYSREG block regmap.
  * @sc: scaler infomations.
  * @pol: porarity of writeback.
  * @id: fimc id.
  * @irq: irq number.
- * @suspended: qos operations.
  */
 struct fimc_context {
-	struct exynos_drm_ippdrv	ippdrv;
+	struct exynos_drm_ipp ipp;
+	struct drm_device *drm_dev;
+	struct device	*dev;
+	struct exynos_drm_ipp_task	*task;
+	struct exynos_drm_ipp_formats	*formats;
+	unsigned int			num_formats;
+
 	struct resource	*regs_res;
 	void __iomem	*regs;
 	spinlock_t	lock;
 	struct clk	*clocks[FIMC_CLKS_MAX];
-	u32		clk_frequency;
-	struct regmap	*sysreg;
 	struct fimc_scaler	sc;
 	int	id;
 	int	irq;
-	bool	suspended;
 };
 
 static u32 fimc_read(struct fimc_context *ctx, u32 reg)
@@ -217,19 +163,10 @@ static void fimc_sw_reset(struct fimc_context *ctx)
 	fimc_write(ctx, 0x0, EXYNOS_CIFCNTSEQ);
 }
 
-static int fimc_set_camblk_fimd0_wb(struct fimc_context *ctx)
-{
-	return regmap_update_bits(ctx->sysreg, SYSREG_CAMERA_BLK,
-				  SYSREG_FIMD0WB_DEST_MASK,
-				  ctx->id << SYSREG_FIMD0WB_DEST_SHIFT);
-}
-
-static void fimc_set_type_ctrl(struct fimc_context *ctx, enum fimc_wb wb)
+static void fimc_set_type_ctrl(struct fimc_context *ctx)
 {
 	u32 cfg;
 
-	DRM_DEBUG_KMS("wb[%d]\n", wb);
-
 	cfg = fimc_read(ctx, EXYNOS_CIGCTRL);
 	cfg &= ~(EXYNOS_CIGCTRL_TESTPATTERN_MASK |
 		EXYNOS_CIGCTRL_SELCAM_ITU_MASK |
@@ -238,23 +175,10 @@ static void fimc_set_type_ctrl(struct fimc_context *ctx, enum fimc_wb wb)
 		EXYNOS_CIGCTRL_SELWB_CAMIF_MASK |
 		EXYNOS_CIGCTRL_SELWRITEBACK_MASK);
 
-	switch (wb) {
-	case FIMC_WB_A:
-		cfg |= (EXYNOS_CIGCTRL_SELWRITEBACK_A |
-			EXYNOS_CIGCTRL_SELWB_CAMIF_WRITEBACK);
-		break;
-	case FIMC_WB_B:
-		cfg |= (EXYNOS_CIGCTRL_SELWRITEBACK_B |
-			EXYNOS_CIGCTRL_SELWB_CAMIF_WRITEBACK);
-		break;
-	case FIMC_WB_NONE:
-	default:
-		cfg |= (EXYNOS_CIGCTRL_SELCAM_ITU_A |
-			EXYNOS_CIGCTRL_SELWRITEBACK_A |
-			EXYNOS_CIGCTRL_SELCAM_MIPI_A |
-			EXYNOS_CIGCTRL_SELCAM_FIMC_ITU);
-		break;
-	}
+	cfg |= (EXYNOS_CIGCTRL_SELCAM_ITU_A |
+		EXYNOS_CIGCTRL_SELWRITEBACK_A |
+		EXYNOS_CIGCTRL_SELCAM_MIPI_A |
+		EXYNOS_CIGCTRL_SELCAM_FIMC_ITU);
 
 	fimc_write(ctx, cfg, EXYNOS_CIGCTRL);
 }
@@ -296,7 +220,6 @@ static void fimc_clear_irq(struct fimc_context *ctx)
 
 static bool fimc_check_ovf(struct fimc_context *ctx)
 {
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
 	u32 status, flag;
 
 	status = fimc_read(ctx, EXYNOS_CISTATUS);
@@ -310,7 +233,7 @@ static bool fimc_check_ovf(struct fimc_context *ctx)
 			EXYNOS_CIWDOFST_CLROVFIY | EXYNOS_CIWDOFST_CLROVFICB |
 			EXYNOS_CIWDOFST_CLROVFICR);
 
-		dev_err(ippdrv->dev, "occurred overflow at %d, status 0x%x.\n",
+		dev_err(ctx->dev, "occurred overflow at %d, status 0x%x.\n",
 			ctx->id, status);
 		return true;
 	}
@@ -376,10 +299,8 @@ static void fimc_handle_lastend(struct fimc_context *ctx, bool enable)
 	fimc_write(ctx, cfg, EXYNOS_CIOCTRL);
 }
 
-
-static int fimc_src_set_fmt_order(struct fimc_context *ctx, u32 fmt)
+static void fimc_src_set_fmt_order(struct fimc_context *ctx, u32 fmt)
 {
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
 	u32 cfg;
 
 	DRM_DEBUG_KMS("fmt[0x%x]\n", fmt);
@@ -392,12 +313,12 @@ static int fimc_src_set_fmt_order(struct fimc_context *ctx, u32 fmt)
 	case DRM_FORMAT_RGB565:
 		cfg |= EXYNOS_CISCCTRL_INRGB_FMT_RGB565;
 		fimc_write(ctx, cfg, EXYNOS_CISCCTRL);
-		return 0;
+		return;
 	case DRM_FORMAT_RGB888:
 	case DRM_FORMAT_XRGB8888:
 		cfg |= EXYNOS_CISCCTRL_INRGB_FMT_RGB888;
 		fimc_write(ctx, cfg, EXYNOS_CISCCTRL);
-		return 0;
+		return;
 	default:
 		/* bypass */
 		break;
@@ -438,20 +359,13 @@ static int fimc_src_set_fmt_order(struct fimc_context *ctx, u32 fmt)
 		cfg |= (EXYNOS_MSCTRL_ORDER2P_LSB_CBCR |
 			EXYNOS_MSCTRL_C_INT_IN_2PLANE);
 		break;
-	default:
-		dev_err(ippdrv->dev, "invalid source yuv order 0x%x.\n", fmt);
-		return -EINVAL;
 	}
 
 	fimc_write(ctx, cfg, EXYNOS_MSCTRL);
-
-	return 0;
 }
 
-static int fimc_src_set_fmt(struct device *dev, u32 fmt)
+static void fimc_src_set_fmt(struct fimc_context *ctx, u32 fmt, bool tiled)
 {
-	struct fimc_context *ctx = get_fimc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
 	u32 cfg;
 
 	DRM_DEBUG_KMS("fmt[0x%x]\n", fmt);
@@ -485,9 +399,6 @@ static int fimc_src_set_fmt(struct device *dev, u32 fmt)
 	case DRM_FORMAT_NV21:
 		cfg |= EXYNOS_MSCTRL_INFORMAT_YCBCR420;
 		break;
-	default:
-		dev_err(ippdrv->dev, "invalid source format 0x%x.\n", fmt);
-		return -EINVAL;
 	}
 
 	fimc_write(ctx, cfg, EXYNOS_MSCTRL);
@@ -495,22 +406,22 @@ static int fimc_src_set_fmt(struct device *dev, u32 fmt)
 	cfg = fimc_read(ctx, EXYNOS_CIDMAPARAM);
 	cfg &= ~EXYNOS_CIDMAPARAM_R_MODE_MASK;
 
-	cfg |= EXYNOS_CIDMAPARAM_R_MODE_LINEAR;
+	if (tiled)
+		cfg |= EXYNOS_CIDMAPARAM_R_MODE_64X32;
+	else
+		cfg |= EXYNOS_CIDMAPARAM_R_MODE_LINEAR;
 
 	fimc_write(ctx, cfg, EXYNOS_CIDMAPARAM);
 
-	return fimc_src_set_fmt_order(ctx, fmt);
+	fimc_src_set_fmt_order(ctx, fmt);
 }
 
-static int fimc_src_set_transf(struct device *dev,
-		enum drm_exynos_degree degree,
-		enum drm_exynos_flip flip, bool *swap)
+static void fimc_src_set_transf(struct fimc_context *ctx, unsigned int rotation)
 {
-	struct fimc_context *ctx = get_fimc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
+	unsigned int degree = rotation & DRM_MODE_ROTATE_MASK;
 	u32 cfg1, cfg2;
 
-	DRM_DEBUG_KMS("degree[%d]flip[0x%x]\n", degree, flip);
+	DRM_DEBUG_KMS("rotation[%x]\n", rotation);
 
 	cfg1 = fimc_read(ctx, EXYNOS_MSCTRL);
 	cfg1 &= ~(EXYNOS_MSCTRL_FLIP_X_MIRROR |
@@ -520,61 +431,56 @@ static int fimc_src_set_transf(struct device *dev,
 	cfg2 &= ~EXYNOS_CITRGFMT_INROT90_CLOCKWISE;
 
 	switch (degree) {
-	case EXYNOS_DRM_DEGREE_0:
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
+	case DRM_MODE_ROTATE_0:
+		if (rotation & DRM_MODE_REFLECT_X)
 			cfg1 |= EXYNOS_MSCTRL_FLIP_X_MIRROR;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
+		if (rotation & DRM_MODE_REFLECT_Y)
 			cfg1 |= EXYNOS_MSCTRL_FLIP_Y_MIRROR;
 		break;
-	case EXYNOS_DRM_DEGREE_90:
+	case DRM_MODE_ROTATE_90:
 		cfg2 |= EXYNOS_CITRGFMT_INROT90_CLOCKWISE;
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
+		if (rotation & DRM_MODE_REFLECT_X)
 			cfg1 |= EXYNOS_MSCTRL_FLIP_X_MIRROR;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
+		if (rotation & DRM_MODE_REFLECT_Y)
 			cfg1 |= EXYNOS_MSCTRL_FLIP_Y_MIRROR;
 		break;
-	case EXYNOS_DRM_DEGREE_180:
+	case DRM_MODE_ROTATE_180:
 		cfg1 |= (EXYNOS_MSCTRL_FLIP_X_MIRROR |
 			EXYNOS_MSCTRL_FLIP_Y_MIRROR);
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
+		if (rotation & DRM_MODE_REFLECT_X)
 			cfg1 &= ~EXYNOS_MSCTRL_FLIP_X_MIRROR;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
+		if (rotation & DRM_MODE_REFLECT_Y)
 			cfg1 &= ~EXYNOS_MSCTRL_FLIP_Y_MIRROR;
 		break;
-	case EXYNOS_DRM_DEGREE_270:
+	case DRM_MODE_ROTATE_270:
 		cfg1 |= (EXYNOS_MSCTRL_FLIP_X_MIRROR |
 			EXYNOS_MSCTRL_FLIP_Y_MIRROR);
 		cfg2 |= EXYNOS_CITRGFMT_INROT90_CLOCKWISE;
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
+		if (rotation & DRM_MODE_REFLECT_X)
 			cfg1 &= ~EXYNOS_MSCTRL_FLIP_X_MIRROR;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
+		if (rotation & DRM_MODE_REFLECT_Y)
 			cfg1 &= ~EXYNOS_MSCTRL_FLIP_Y_MIRROR;
 		break;
-	default:
-		dev_err(ippdrv->dev, "invalid degree value %d.\n", degree);
-		return -EINVAL;
 	}
 
 	fimc_write(ctx, cfg1, EXYNOS_MSCTRL);
 	fimc_write(ctx, cfg2, EXYNOS_CITRGFMT);
-	*swap = (cfg2 & EXYNOS_CITRGFMT_INROT90_CLOCKWISE) ? 1 : 0;
-
-	return 0;
 }
 
-static int fimc_set_window(struct fimc_context *ctx,
-		struct drm_exynos_pos *pos, struct drm_exynos_sz *sz)
+static void fimc_set_window(struct fimc_context *ctx,
+			    struct exynos_drm_ipp_buffer *buf)
 {
 	u32 cfg, h1, h2, v1, v2;
 
 	/* cropped image */
-	h1 = pos->x;
-	h2 = sz->hsize - pos->w - pos->x;
-	v1 = pos->y;
-	v2 = sz->vsize - pos->h - pos->y;
+	h1 = buf->rect.x;
+	h2 = buf->buf.width - buf->rect.w - buf->rect.x;
+	v1 = buf->rect.y;
+	v2 = buf->buf.height - buf->rect.h - buf->rect.y;
 
 	DRM_DEBUG_KMS("x[%d]y[%d]w[%d]h[%d]hsize[%d]vsize[%d]\n",
-		pos->x, pos->y, pos->w, pos->h, sz->hsize, sz->vsize);
+		buf->rect.x, buf->rect.y, buf->rect.w, buf->rect.h,
+		buf->buf.width, buf->buf.height);
 	DRM_DEBUG_KMS("h1[%d]h2[%d]v1[%d]v2[%d]\n", h1, h2, v1, v2);
 
 	/*
@@ -592,42 +498,30 @@ static int fimc_set_window(struct fimc_context *ctx,
 	cfg = (EXYNOS_CIWDOFST2_WINHOROFST2(h2) |
 		EXYNOS_CIWDOFST2_WINVEROFST2(v2));
 	fimc_write(ctx, cfg, EXYNOS_CIWDOFST2);
-
-	return 0;
 }
 
-static int fimc_src_set_size(struct device *dev, int swap,
-		struct drm_exynos_pos *pos, struct drm_exynos_sz *sz)
+static void fimc_src_set_size(struct fimc_context *ctx,
+			      struct exynos_drm_ipp_buffer *buf)
 {
-	struct fimc_context *ctx = get_fimc_context(dev);
-	struct drm_exynos_pos img_pos = *pos;
-	struct drm_exynos_sz img_sz = *sz;
 	u32 cfg;
 
-	DRM_DEBUG_KMS("swap[%d]hsize[%d]vsize[%d]\n",
-		swap, sz->hsize, sz->vsize);
+	DRM_DEBUG_KMS("hsize[%d]vsize[%d]\n", buf->buf.width, buf->buf.height);
 
 	/* original size */
-	cfg = (EXYNOS_ORGISIZE_HORIZONTAL(img_sz.hsize) |
-		EXYNOS_ORGISIZE_VERTICAL(img_sz.vsize));
+	cfg = (EXYNOS_ORGISIZE_HORIZONTAL(buf->buf.width) |
+		EXYNOS_ORGISIZE_VERTICAL(buf->buf.height));
 
 	fimc_write(ctx, cfg, EXYNOS_ORGISIZE);
 
-	DRM_DEBUG_KMS("x[%d]y[%d]w[%d]h[%d]\n", pos->x, pos->y, pos->w, pos->h);
-
-	if (swap) {
-		img_pos.w = pos->h;
-		img_pos.h = pos->w;
-		img_sz.hsize = sz->vsize;
-		img_sz.vsize = sz->hsize;
-	}
+	DRM_DEBUG_KMS("x[%d]y[%d]w[%d]h[%d]\n", buf->rect.x, buf->rect.y,
+		buf->rect.w, buf->rect.h);
 
 	/* set input DMA image size */
 	cfg = fimc_read(ctx, EXYNOS_CIREAL_ISIZE);
 	cfg &= ~(EXYNOS_CIREAL_ISIZE_HEIGHT_MASK |
 		EXYNOS_CIREAL_ISIZE_WIDTH_MASK);
-	cfg |= (EXYNOS_CIREAL_ISIZE_WIDTH(img_pos.w) |
-		EXYNOS_CIREAL_ISIZE_HEIGHT(img_pos.h));
+	cfg |= (EXYNOS_CIREAL_ISIZE_WIDTH(buf->rect.w) |
+		EXYNOS_CIREAL_ISIZE_HEIGHT(buf->rect.h));
 	fimc_write(ctx, cfg, EXYNOS_CIREAL_ISIZE);
 
 	/*
@@ -635,91 +529,34 @@ static int fimc_src_set_size(struct device *dev, int swap,
 	 * for now, we support only ITU601 8 bit mode
 	 */
 	cfg = (EXYNOS_CISRCFMT_ITU601_8BIT |
-		EXYNOS_CISRCFMT_SOURCEHSIZE(img_sz.hsize) |
-		EXYNOS_CISRCFMT_SOURCEVSIZE(img_sz.vsize));
+		EXYNOS_CISRCFMT_SOURCEHSIZE(buf->buf.width) |
+		EXYNOS_CISRCFMT_SOURCEVSIZE(buf->buf.height));
 	fimc_write(ctx, cfg, EXYNOS_CISRCFMT);
 
 	/* offset Y(RGB), Cb, Cr */
-	cfg = (EXYNOS_CIIYOFF_HORIZONTAL(img_pos.x) |
-		EXYNOS_CIIYOFF_VERTICAL(img_pos.y));
+	cfg = (EXYNOS_CIIYOFF_HORIZONTAL(buf->rect.x) |
+		EXYNOS_CIIYOFF_VERTICAL(buf->rect.y));
 	fimc_write(ctx, cfg, EXYNOS_CIIYOFF);
-	cfg = (EXYNOS_CIICBOFF_HORIZONTAL(img_pos.x) |
-		EXYNOS_CIICBOFF_VERTICAL(img_pos.y));
+	cfg = (EXYNOS_CIICBOFF_HORIZONTAL(buf->rect.x) |
+		EXYNOS_CIICBOFF_VERTICAL(buf->rect.y));
 	fimc_write(ctx, cfg, EXYNOS_CIICBOFF);
-	cfg = (EXYNOS_CIICROFF_HORIZONTAL(img_pos.x) |
-		EXYNOS_CIICROFF_VERTICAL(img_pos.y));
+	cfg = (EXYNOS_CIICROFF_HORIZONTAL(buf->rect.x) |
+		EXYNOS_CIICROFF_VERTICAL(buf->rect.y));
 	fimc_write(ctx, cfg, EXYNOS_CIICROFF);
 
-	return fimc_set_window(ctx, &img_pos, &img_sz);
+	fimc_set_window(ctx, buf);
 }
 
-static int fimc_src_set_addr(struct device *dev,
-		struct drm_exynos_ipp_buf_info *buf_info, u32 buf_id,
-		enum drm_exynos_ipp_buf_type buf_type)
+static void fimc_src_set_addr(struct fimc_context *ctx,
+			      struct exynos_drm_ipp_buffer *buf)
 {
-	struct fimc_context *ctx = get_fimc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	struct drm_exynos_ipp_cmd_node *c_node = ippdrv->c_node;
-	struct drm_exynos_ipp_property *property;
-	struct drm_exynos_ipp_config *config;
-
-	if (!c_node) {
-		DRM_ERROR("failed to get c_node.\n");
-		return -EINVAL;
-	}
-
-	property = &c_node->property;
-
-	DRM_DEBUG_KMS("prop_id[%d]buf_id[%d]buf_type[%d]\n",
-		property->prop_id, buf_id, buf_type);
-
-	if (buf_id > FIMC_MAX_SRC) {
-		dev_info(ippdrv->dev, "invalid buf_id %d.\n", buf_id);
-		return -ENOMEM;
-	}
-
-	/* address register set */
-	switch (buf_type) {
-	case IPP_BUF_ENQUEUE:
-		config = &property->config[EXYNOS_DRM_OPS_SRC];
-		fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_Y],
-			EXYNOS_CIIYSA0);
-
-		if (config->fmt == DRM_FORMAT_YVU420) {
-			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CR],
-				EXYNOS_CIICBSA0);
-			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CB],
-				EXYNOS_CIICRSA0);
-		} else {
-			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CB],
-				EXYNOS_CIICBSA0);
-			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CR],
-				EXYNOS_CIICRSA0);
-		}
-		break;
-	case IPP_BUF_DEQUEUE:
-		fimc_write(ctx, 0x0, EXYNOS_CIIYSA0);
-		fimc_write(ctx, 0x0, EXYNOS_CIICBSA0);
-		fimc_write(ctx, 0x0, EXYNOS_CIICRSA0);
-		break;
-	default:
-		/* bypass */
-		break;
-	}
-
-	return 0;
+	fimc_write(ctx, buf->dma_addr[0], EXYNOS_CIIYSA(0));
+	fimc_write(ctx, buf->dma_addr[1], EXYNOS_CIICBSA(0));
+	fimc_write(ctx, buf->dma_addr[2], EXYNOS_CIICRSA(0));
 }
 
-static struct exynos_drm_ipp_ops fimc_src_ops = {
-	.set_fmt = fimc_src_set_fmt,
-	.set_transf = fimc_src_set_transf,
-	.set_size = fimc_src_set_size,
-	.set_addr = fimc_src_set_addr,
-};
-
-static int fimc_dst_set_fmt_order(struct fimc_context *ctx, u32 fmt)
+static void fimc_dst_set_fmt_order(struct fimc_context *ctx, u32 fmt)
 {
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
 	u32 cfg;
 
 	DRM_DEBUG_KMS("fmt[0x%x]\n", fmt);
@@ -732,11 +569,11 @@ static int fimc_dst_set_fmt_order(struct fimc_context *ctx, u32 fmt)
 	case DRM_FORMAT_RGB565:
 		cfg |= EXYNOS_CISCCTRL_OUTRGB_FMT_RGB565;
 		fimc_write(ctx, cfg, EXYNOS_CISCCTRL);
-		return 0;
+		return;
 	case DRM_FORMAT_RGB888:
 		cfg |= EXYNOS_CISCCTRL_OUTRGB_FMT_RGB888;
 		fimc_write(ctx, cfg, EXYNOS_CISCCTRL);
-		return 0;
+		return;
 	case DRM_FORMAT_XRGB8888:
 		cfg |= (EXYNOS_CISCCTRL_OUTRGB_FMT_RGB888 |
 			EXYNOS_CISCCTRL_EXTRGB_EXTENSION);
@@ -784,20 +621,13 @@ static int fimc_dst_set_fmt_order(struct fimc_context *ctx, u32 fmt)
 		cfg |= EXYNOS_CIOCTRL_ORDER2P_LSB_CBCR;
 		cfg |= EXYNOS_CIOCTRL_YCBCR_2PLANE;
 		break;
-	default:
-		dev_err(ippdrv->dev, "invalid target yuv order 0x%x.\n", fmt);
-		return -EINVAL;
 	}
 
 	fimc_write(ctx, cfg, EXYNOS_CIOCTRL);
-
-	return 0;
 }
 
-static int fimc_dst_set_fmt(struct device *dev, u32 fmt)
+static void fimc_dst_set_fmt(struct fimc_context *ctx, u32 fmt, bool tiled)
 {
-	struct fimc_context *ctx = get_fimc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
 	u32 cfg;
 
 	DRM_DEBUG_KMS("fmt[0x%x]\n", fmt);
@@ -837,10 +667,6 @@ static int fimc_dst_set_fmt(struct device *dev, u32 fmt)
 		case DRM_FORMAT_NV21:
 			cfg |= EXYNOS_CITRGFMT_OUTFORMAT_YCBCR420;
 			break;
-		default:
-			dev_err(ippdrv->dev, "invalid target format 0x%x.\n",
-				fmt);
-			return -EINVAL;
 		}
 
 		fimc_write(ctx, cfg, EXYNOS_CITRGFMT);
@@ -849,73 +675,67 @@ static int fimc_dst_set_fmt(struct device *dev, u32 fmt)
 	cfg = fimc_read(ctx, EXYNOS_CIDMAPARAM);
 	cfg &= ~EXYNOS_CIDMAPARAM_W_MODE_MASK;
 
-	cfg |= EXYNOS_CIDMAPARAM_W_MODE_LINEAR;
+	if (tiled)
+		cfg |= EXYNOS_CIDMAPARAM_W_MODE_64X32;
+	else
+		cfg |= EXYNOS_CIDMAPARAM_W_MODE_LINEAR;
 
 	fimc_write(ctx, cfg, EXYNOS_CIDMAPARAM);
 
-	return fimc_dst_set_fmt_order(ctx, fmt);
+	fimc_dst_set_fmt_order(ctx, fmt);
 }
 
-static int fimc_dst_set_transf(struct device *dev,
-		enum drm_exynos_degree degree,
-		enum drm_exynos_flip flip, bool *swap)
+static void fimc_dst_set_transf(struct fimc_context *ctx, unsigned int rotation)
 {
-	struct fimc_context *ctx = get_fimc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
+	unsigned int degree = rotation & DRM_MODE_ROTATE_MASK;
 	u32 cfg;
 
-	DRM_DEBUG_KMS("degree[%d]flip[0x%x]\n", degree, flip);
+	DRM_DEBUG_KMS("rotation[0x%x]\n", rotation);
 
 	cfg = fimc_read(ctx, EXYNOS_CITRGFMT);
 	cfg &= ~EXYNOS_CITRGFMT_FLIP_MASK;
 	cfg &= ~EXYNOS_CITRGFMT_OUTROT90_CLOCKWISE;
 
 	switch (degree) {
-	case EXYNOS_DRM_DEGREE_0:
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
+	case DRM_MODE_ROTATE_0:
+		if (rotation & DRM_MODE_REFLECT_X)
 			cfg |= EXYNOS_CITRGFMT_FLIP_X_MIRROR;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
+		if (rotation & DRM_MODE_REFLECT_Y)
 			cfg |= EXYNOS_CITRGFMT_FLIP_Y_MIRROR;
 		break;
-	case EXYNOS_DRM_DEGREE_90:
+	case DRM_MODE_ROTATE_90:
 		cfg |= EXYNOS_CITRGFMT_OUTROT90_CLOCKWISE;
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
+		if (rotation & DRM_MODE_REFLECT_X)
 			cfg |= EXYNOS_CITRGFMT_FLIP_X_MIRROR;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
+		if (rotation & DRM_MODE_REFLECT_Y)
 			cfg |= EXYNOS_CITRGFMT_FLIP_Y_MIRROR;
 		break;
-	case EXYNOS_DRM_DEGREE_180:
+	case DRM_MODE_ROTATE_180:
 		cfg |= (EXYNOS_CITRGFMT_FLIP_X_MIRROR |
 			EXYNOS_CITRGFMT_FLIP_Y_MIRROR);
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
+		if (rotation & DRM_MODE_REFLECT_X)
 			cfg &= ~EXYNOS_CITRGFMT_FLIP_X_MIRROR;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
+		if (rotation & DRM_MODE_REFLECT_Y)
 			cfg &= ~EXYNOS_CITRGFMT_FLIP_Y_MIRROR;
 		break;
-	case EXYNOS_DRM_DEGREE_270:
+	case DRM_MODE_ROTATE_270:
 		cfg |= (EXYNOS_CITRGFMT_OUTROT90_CLOCKWISE |
 			EXYNOS_CITRGFMT_FLIP_X_MIRROR |
 			EXYNOS_CITRGFMT_FLIP_Y_MIRROR);
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
+		if (rotation & DRM_MODE_REFLECT_X)
 			cfg &= ~EXYNOS_CITRGFMT_FLIP_X_MIRROR;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
+		if (rotation & DRM_MODE_REFLECT_Y)
 			cfg &= ~EXYNOS_CITRGFMT_FLIP_Y_MIRROR;
 		break;
-	default:
-		dev_err(ippdrv->dev, "invalid degree value %d.\n", degree);
-		return -EINVAL;
 	}
 
 	fimc_write(ctx, cfg, EXYNOS_CITRGFMT);
-	*swap = (cfg & EXYNOS_CITRGFMT_OUTROT90_CLOCKWISE) ? 1 : 0;
-
-	return 0;
 }
 
 static int fimc_set_prescaler(struct fimc_context *ctx, struct fimc_scaler *sc,
-		struct drm_exynos_pos *src, struct drm_exynos_pos *dst)
+			      struct drm_exynos_ipp_task_rect *src,
+			      struct drm_exynos_ipp_task_rect *dst)
 {
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
 	u32 cfg, cfg_ext, shfactor;
 	u32 pre_dst_width, pre_dst_height;
 	u32 hfactor, vfactor;
@@ -942,13 +762,13 @@ static int fimc_set_prescaler(struct fimc_context *ctx, struct fimc_scaler *sc,
 	/* fimc_ippdrv_check_property assures that dividers are not null */
 	hfactor = fls(src_w / dst_w / 2);
 	if (hfactor > FIMC_SHFACTOR / 2) {
-		dev_err(ippdrv->dev, "failed to get ratio horizontal.\n");
+		dev_err(ctx->dev, "failed to get ratio horizontal.\n");
 		return -EINVAL;
 	}
 
 	vfactor = fls(src_h / dst_h / 2);
 	if (vfactor > FIMC_SHFACTOR / 2) {
-		dev_err(ippdrv->dev, "failed to get ratio vertical.\n");
+		dev_err(ctx->dev, "failed to get ratio vertical.\n");
 		return -EINVAL;
 	}
 
@@ -1019,83 +839,77 @@ static void fimc_set_scaler(struct fimc_context *ctx, struct fimc_scaler *sc)
 	fimc_write(ctx, cfg_ext, EXYNOS_CIEXTEN);
 }
 
-static int fimc_dst_set_size(struct device *dev, int swap,
-		struct drm_exynos_pos *pos, struct drm_exynos_sz *sz)
+static void fimc_dst_set_size(struct fimc_context *ctx,
+			     struct exynos_drm_ipp_buffer *buf)
 {
-	struct fimc_context *ctx = get_fimc_context(dev);
-	struct drm_exynos_pos img_pos = *pos;
-	struct drm_exynos_sz img_sz = *sz;
-	u32 cfg;
+	u32 cfg, cfg_ext;
 
-	DRM_DEBUG_KMS("swap[%d]hsize[%d]vsize[%d]\n",
-		swap, sz->hsize, sz->vsize);
+	DRM_DEBUG_KMS("hsize[%d]vsize[%d]\n", buf->buf.width, buf->buf.height);
 
 	/* original size */
-	cfg = (EXYNOS_ORGOSIZE_HORIZONTAL(img_sz.hsize) |
-		EXYNOS_ORGOSIZE_VERTICAL(img_sz.vsize));
+	cfg = (EXYNOS_ORGOSIZE_HORIZONTAL(buf->buf.width) |
+		EXYNOS_ORGOSIZE_VERTICAL(buf->buf.height));
 
 	fimc_write(ctx, cfg, EXYNOS_ORGOSIZE);
 
-	DRM_DEBUG_KMS("x[%d]y[%d]w[%d]h[%d]\n", pos->x, pos->y, pos->w, pos->h);
+	DRM_DEBUG_KMS("x[%d]y[%d]w[%d]h[%d]\n", buf->rect.x, buf->rect.y,
+		buf->rect.w, buf->rect.h);
 
 	/* CSC ITU */
 	cfg = fimc_read(ctx, EXYNOS_CIGCTRL);
 	cfg &= ~EXYNOS_CIGCTRL_CSC_MASK;
 
-	if (sz->hsize >= FIMC_WIDTH_ITU_709)
+	if (buf->buf.width >= FIMC_WIDTH_ITU_709)
 		cfg |= EXYNOS_CIGCTRL_CSC_ITU709;
 	else
 		cfg |= EXYNOS_CIGCTRL_CSC_ITU601;
 
 	fimc_write(ctx, cfg, EXYNOS_CIGCTRL);
 
-	if (swap) {
-		img_pos.w = pos->h;
-		img_pos.h = pos->w;
-		img_sz.hsize = sz->vsize;
-		img_sz.vsize = sz->hsize;
-	}
+	cfg_ext = fimc_read(ctx, EXYNOS_CITRGFMT);
 
 	/* target image size */
 	cfg = fimc_read(ctx, EXYNOS_CITRGFMT);
 	cfg &= ~(EXYNOS_CITRGFMT_TARGETH_MASK |
 		EXYNOS_CITRGFMT_TARGETV_MASK);
-	cfg |= (EXYNOS_CITRGFMT_TARGETHSIZE(img_pos.w) |
-		EXYNOS_CITRGFMT_TARGETVSIZE(img_pos.h));
+	if (cfg_ext & EXYNOS_CITRGFMT_OUTROT90_CLOCKWISE)
+		cfg |= (EXYNOS_CITRGFMT_TARGETHSIZE(buf->rect.h) |
+			EXYNOS_CITRGFMT_TARGETVSIZE(buf->rect.w));
+	else
+		cfg |= (EXYNOS_CITRGFMT_TARGETHSIZE(buf->rect.w) |
+			EXYNOS_CITRGFMT_TARGETVSIZE(buf->rect.h));
 	fimc_write(ctx, cfg, EXYNOS_CITRGFMT);
 
 	/* target area */
-	cfg = EXYNOS_CITAREA_TARGET_AREA(img_pos.w * img_pos.h);
+	cfg = EXYNOS_CITAREA_TARGET_AREA(buf->rect.w * buf->rect.h);
 	fimc_write(ctx, cfg, EXYNOS_CITAREA);
 
 	/* offset Y(RGB), Cb, Cr */
-	cfg = (EXYNOS_CIOYOFF_HORIZONTAL(img_pos.x) |
-		EXYNOS_CIOYOFF_VERTICAL(img_pos.y));
+	cfg = (EXYNOS_CIOYOFF_HORIZONTAL(buf->rect.x) |
+		EXYNOS_CIOYOFF_VERTICAL(buf->rect.y));
 	fimc_write(ctx, cfg, EXYNOS_CIOYOFF);
-	cfg = (EXYNOS_CIOCBOFF_HORIZONTAL(img_pos.x) |
-		EXYNOS_CIOCBOFF_VERTICAL(img_pos.y));
+	cfg = (EXYNOS_CIOCBOFF_HORIZONTAL(buf->rect.x) |
+		EXYNOS_CIOCBOFF_VERTICAL(buf->rect.y));
 	fimc_write(ctx, cfg, EXYNOS_CIOCBOFF);
-	cfg = (EXYNOS_CIOCROFF_HORIZONTAL(img_pos.x) |
-		EXYNOS_CIOCROFF_VERTICAL(img_pos.y));
+	cfg = (EXYNOS_CIOCROFF_HORIZONTAL(buf->rect.x) |
+		EXYNOS_CIOCROFF_VERTICAL(buf->rect.y));
 	fimc_write(ctx, cfg, EXYNOS_CIOCROFF);
-
-	return 0;
 }
 
 static void fimc_dst_set_buf_seq(struct fimc_context *ctx, u32 buf_id,
-		enum drm_exynos_ipp_buf_type buf_type)
+		bool enqueue)
 {
 	unsigned long flags;
 	u32 buf_num;
 	u32 cfg;
 
-	DRM_DEBUG_KMS("buf_id[%d]buf_type[%d]\n", buf_id, buf_type);
+	DRM_DEBUG_KMS("buf_id[%d]enqueu[%d]\n", buf_id, enqueue);
 
 	spin_lock_irqsave(&ctx->lock, flags);
 
 	cfg = fimc_read(ctx, EXYNOS_CIFCNTSEQ);
 
-	if (buf_type == IPP_BUF_ENQUEUE)
+	if (enqueue)
 		cfg |= (1 << buf_id);
 	else
 		cfg &= ~(1 << buf_id);
@@ -1104,88 +918,29 @@ static void fimc_dst_set_buf_seq(struct fimc_context *ctx, u32 buf_id,
 
 	buf_num = hweight32(cfg);
 
-	if (buf_type == IPP_BUF_ENQUEUE && buf_num >= FIMC_BUF_START)
+	if (enqueue && buf_num >= FIMC_BUF_START)
 		fimc_mask_irq(ctx, true);
-	else if (buf_type == IPP_BUF_DEQUEUE && buf_num <= FIMC_BUF_STOP)
+	else if (!enqueue && buf_num <= FIMC_BUF_STOP)
 		fimc_mask_irq(ctx, false);
 
 	spin_unlock_irqrestore(&ctx->lock, flags);
 }
 
-static int fimc_dst_set_addr(struct device *dev,
-		struct drm_exynos_ipp_buf_info *buf_info, u32 buf_id,
-		enum drm_exynos_ipp_buf_type buf_type)
+static void fimc_dst_set_addr(struct fimc_context *ctx,
+			     struct exynos_drm_ipp_buffer *buf)
 {
-	struct fimc_context *ctx = get_fimc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	struct drm_exynos_ipp_cmd_node *c_node = ippdrv->c_node;
-	struct drm_exynos_ipp_property *property;
-	struct drm_exynos_ipp_config *config;
-
-	if (!c_node) {
-		DRM_ERROR("failed to get c_node.\n");
-		return -EINVAL;
-	}
-
-	property = &c_node->property;
-
-	DRM_DEBUG_KMS("prop_id[%d]buf_id[%d]buf_type[%d]\n",
-		property->prop_id, buf_id, buf_type);
+	fimc_write(ctx, buf->dma_addr[0], EXYNOS_CIOYSA(0));
+	fimc_write(ctx, buf->dma_addr[1], EXYNOS_CIOCBSA(0));
+	fimc_write(ctx, buf->dma_addr[2], EXYNOS_CIOCRSA(0));
 
-	if (buf_id > FIMC_MAX_DST) {
-		dev_info(ippdrv->dev, "invalid buf_id %d.\n", buf_id);
-		return -ENOMEM;
-	}
-
-	/* address register set */
-	switch (buf_type) {
-	case IPP_BUF_ENQUEUE:
-		config = &property->config[EXYNOS_DRM_OPS_DST];
-
-		fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_Y],
-			EXYNOS_CIOYSA(buf_id));
-
-		if (config->fmt == DRM_FORMAT_YVU420) {
-			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CR],
-				EXYNOS_CIOCBSA(buf_id));
-			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CB],
-				EXYNOS_CIOCRSA(buf_id));
-		} else {
-			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CB],
-				EXYNOS_CIOCBSA(buf_id));
-			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CR],
-				EXYNOS_CIOCRSA(buf_id));
-		}
-		break;
-	case IPP_BUF_DEQUEUE:
-		fimc_write(ctx, 0x0, EXYNOS_CIOYSA(buf_id));
-		fimc_write(ctx, 0x0, EXYNOS_CIOCBSA(buf_id));
-		fimc_write(ctx, 0x0, EXYNOS_CIOCRSA(buf_id));
-		break;
-	default:
-		/* bypass */
-		break;
-	}
-
-	fimc_dst_set_buf_seq(ctx, buf_id, buf_type);
-
-	return 0;
+	fimc_dst_set_buf_seq(ctx, 0, true);
 }
 
-static struct exynos_drm_ipp_ops fimc_dst_ops = {
-	.set_fmt = fimc_dst_set_fmt,
-	.set_transf = fimc_dst_set_transf,
-	.set_size = fimc_dst_set_size,
-	.set_addr = fimc_dst_set_addr,
-};
+static void fimc_stop(struct fimc_context *ctx);
 
 static irqreturn_t fimc_irq_handler(int irq, void *dev_id)
 {
 	struct fimc_context *ctx = dev_id;
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	struct drm_exynos_ipp_cmd_node *c_node = ippdrv->c_node;
-	struct drm_exynos_ipp_event_work *event_work =
-		c_node->event_work;
 	int buf_id;
 
 	DRM_DEBUG_KMS("fimc id[%d]\n", ctx->id);
@@ -1203,170 +958,19 @@ static irqreturn_t fimc_irq_handler(int irq, void *dev_id)
 
 	DRM_DEBUG_KMS("buf_id[%d]\n", buf_id);
 
-	fimc_dst_set_buf_seq(ctx, buf_id, IPP_BUF_DEQUEUE);
-
-	event_work->ippdrv = ippdrv;
-	event_work->buf_id[EXYNOS_DRM_OPS_DST] = buf_id;
-	queue_work(ippdrv->event_workq, &event_work->work);
-
-	return IRQ_HANDLED;
-}
-
-static int fimc_init_prop_list(struct exynos_drm_ippdrv *ippdrv)
-{
-	struct drm_exynos_ipp_prop_list *prop_list = &ippdrv->prop_list;
-
-	prop_list->version = 1;
-	prop_list->writeback = 1;
-	prop_list->refresh_min = FIMC_REFRESH_MIN;
-	prop_list->refresh_max = FIMC_REFRESH_MAX;
-	prop_list->flip = (1 << EXYNOS_DRM_FLIP_NONE) |
-				(1 << EXYNOS_DRM_FLIP_VERTICAL) |
-				(1 << EXYNOS_DRM_FLIP_HORIZONTAL);
-	prop_list->degree = (1 << EXYNOS_DRM_DEGREE_0) |
-				(1 << EXYNOS_DRM_DEGREE_90) |
-				(1 << EXYNOS_DRM_DEGREE_180) |
-				(1 << EXYNOS_DRM_DEGREE_270);
-	prop_list->csc = 1;
-	prop_list->crop = 1;
-	prop_list->crop_max.hsize = FIMC_CROP_MAX;
-	prop_list->crop_max.vsize = FIMC_CROP_MAX;
-	prop_list->crop_min.hsize = FIMC_CROP_MIN;
-	prop_list->crop_min.vsize = FIMC_CROP_MIN;
-	prop_list->scale = 1;
-	prop_list->scale_max.hsize = FIMC_SCALE_MAX;
-	prop_list->scale_max.vsize = FIMC_SCALE_MAX;
-	prop_list->scale_min.hsize = FIMC_SCALE_MIN;
-	prop_list->scale_min.vsize = FIMC_SCALE_MIN;
-
-	return 0;
-}
-
-static inline bool fimc_check_drm_flip(enum drm_exynos_flip flip)
-{
-	switch (flip) {
-	case EXYNOS_DRM_FLIP_NONE:
-	case EXYNOS_DRM_FLIP_VERTICAL:
-	case EXYNOS_DRM_FLIP_HORIZONTAL:
-	case EXYNOS_DRM_FLIP_BOTH:
-		return true;
-	default:
-		DRM_DEBUG_KMS("invalid flip\n");
-		return false;
-	}
-}
-
-static int fimc_ippdrv_check_property(struct device *dev,
-		struct drm_exynos_ipp_property *property)
-{
-	struct fimc_context *ctx = get_fimc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	struct drm_exynos_ipp_prop_list *pp = &ippdrv->prop_list;
-	struct drm_exynos_ipp_config *config;
-	struct drm_exynos_pos *pos;
-	struct drm_exynos_sz *sz;
-	bool swap;
-	int i;
-
-	for_each_ipp_ops(i) {
-		if ((i == EXYNOS_DRM_OPS_SRC) &&
-			(property->cmd == IPP_CMD_WB))
-			continue;
-
-		config = &property->config[i];
-		pos = &config->pos;
-		sz = &config->sz;
-
-		/* check for flip */
-		if (!fimc_check_drm_flip(config->flip)) {
-			DRM_ERROR("invalid flip.\n");
-			goto err_property;
-		}
-
-		/* check for degree */
-		switch (config->degree) {
-		case EXYNOS_DRM_DEGREE_90:
-		case EXYNOS_DRM_DEGREE_270:
-			swap = true;
-			break;
-		case EXYNOS_DRM_DEGREE_0:
-		case EXYNOS_DRM_DEGREE_180:
-			swap = false;
-			break;
-		default:
-			DRM_ERROR("invalid degree.\n");
-			goto err_property;
-		}
-
-		/* check for buffer bound */
-		if ((pos->x + pos->w > sz->hsize) ||
-			(pos->y + pos->h > sz->vsize)) {
-			DRM_ERROR("out of buf bound.\n");
-			goto err_property;
-		}
+	if (ctx->task) {
+		struct exynos_drm_ipp_task *task = ctx->task;
 
-		/* check for crop */
-		if ((i == EXYNOS_DRM_OPS_SRC) && (pp->crop)) {
-			if (swap) {
-				if ((pos->h < pp->crop_min.hsize) ||
-					(sz->vsize > pp->crop_max.hsize) ||
-					(pos->w < pp->crop_min.vsize) ||
-					(sz->hsize > pp->crop_max.vsize)) {
-					DRM_ERROR("out of crop size.\n");
-					goto err_property;
-				}
-			} else {
-				if ((pos->w < pp->crop_min.hsize) ||
-					(sz->hsize > pp->crop_max.hsize) ||
-					(pos->h < pp->crop_min.vsize) ||
-					(sz->vsize > pp->crop_max.vsize)) {
-					DRM_ERROR("out of crop size.\n");
-					goto err_property;
-				}
-			}
-		}
-
-		/* check for scale */
-		if ((i == EXYNOS_DRM_OPS_DST) && (pp->scale)) {
-			if (swap) {
-				if ((pos->h < pp->scale_min.hsize) ||
-					(sz->vsize > pp->scale_max.hsize) ||
-					(pos->w < pp->scale_min.vsize) ||
-					(sz->hsize > pp->scale_max.vsize)) {
-					DRM_ERROR("out of scale size.\n");
-					goto err_property;
-				}
-			} else {
-				if ((pos->w < pp->scale_min.hsize) ||
-					(sz->hsize > pp->scale_max.hsize) ||
-					(pos->h < pp->scale_min.vsize) ||
-					(sz->vsize > pp->scale_max.vsize)) {
-					DRM_ERROR("out of scale size.\n");
-					goto err_property;
-				}
-			}
-		}
+		ctx->task = NULL;
+		pm_runtime_mark_last_busy(ctx->dev);
+		pm_runtime_put_autosuspend(ctx->dev);
+		exynos_drm_ipp_task_done(task, 0);
 	}
 
-	return 0;
+	fimc_dst_set_buf_seq(ctx, buf_id, false);
+	fimc_stop(ctx);
 
-err_property:
-	for_each_ipp_ops(i) {
-		if ((i == EXYNOS_DRM_OPS_SRC) &&
-			(property->cmd == IPP_CMD_WB))
-			continue;
-
-		config = &property->config[i];
-		pos = &config->pos;
-		sz = &config->sz;
-
-		DRM_ERROR("[%s]f[%d]r[%d]pos[%d %d %d %d]sz[%d %d]\n",
-			i ? "dst" : "src", config->flip, config->degree,
-			pos->x, pos->y, pos->w, pos->h,
-			sz->hsize, sz->vsize);
-	}
-
-	return -EINVAL;
+	return IRQ_HANDLED;
 }
 
 static void fimc_clear_addr(struct fimc_context *ctx)
@@ -1386,10 +990,8 @@ static void fimc_clear_addr(struct fimc_context *ctx)
 	}
 }
 
-static int fimc_ippdrv_reset(struct device *dev)
+static void fimc_reset(struct fimc_context *ctx)
 {
-	struct fimc_context *ctx = get_fimc_context(dev);
-
 	/* reset h/w block */
 	fimc_sw_reset(ctx);
 
@@ -1397,82 +999,26 @@ static int fimc_ippdrv_reset(struct device *dev)
 	memset(&ctx->sc, 0x0, sizeof(ctx->sc));
 
 	fimc_clear_addr(ctx);
-
-	return 0;
 }
 
-static int fimc_ippdrv_start(struct device *dev, enum drm_exynos_ipp_cmd cmd)
+static void fimc_start(struct fimc_context *ctx)
 {
-	struct fimc_context *ctx = get_fimc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	struct drm_exynos_ipp_cmd_node *c_node = ippdrv->c_node;
-	struct drm_exynos_ipp_property *property;
-	struct drm_exynos_ipp_config *config;
-	struct drm_exynos_pos	img_pos[EXYNOS_DRM_OPS_MAX];
-	struct drm_exynos_ipp_set_wb set_wb;
-	int ret, i;
 	u32 cfg0, cfg1;
 
-	DRM_DEBUG_KMS("cmd[%d]\n", cmd);
-
-	if (!c_node) {
-		DRM_ERROR("failed to get c_node.\n");
-		return -EINVAL;
-	}
-
-	property = &c_node->property;
-
 	fimc_mask_irq(ctx, true);
 
-	for_each_ipp_ops(i) {
-		config = &property->config[i];
-		img_pos[i] = config->pos;
-	}
-
-	ret = fimc_set_prescaler(ctx, &ctx->sc,
-		&img_pos[EXYNOS_DRM_OPS_SRC],
-		&img_pos[EXYNOS_DRM_OPS_DST]);
-	if (ret) {
-		dev_err(dev, "failed to set prescaler.\n");
-		return ret;
-	}
-
-	/* If set ture, we can save jpeg about screen */
+	/* If set true, we can save jpeg about screen */
 	fimc_handle_jpeg(ctx, false);
 	fimc_set_scaler(ctx, &ctx->sc);
 
-	switch (cmd) {
-	case IPP_CMD_M2M:
-		fimc_set_type_ctrl(ctx, FIMC_WB_NONE);
-		fimc_handle_lastend(ctx, false);
-
-		/* setup dma */
-		cfg0 = fimc_read(ctx, EXYNOS_MSCTRL);
-		cfg0 &= ~EXYNOS_MSCTRL_INPUT_MASK;
-		cfg0 |= EXYNOS_MSCTRL_INPUT_MEMORY;
-		fimc_write(ctx, cfg0, EXYNOS_MSCTRL);
-		break;
-	case IPP_CMD_WB:
-		fimc_set_type_ctrl(ctx, FIMC_WB_A);
-		fimc_handle_lastend(ctx, true);
-
-		/* setup FIMD */
-		ret = fimc_set_camblk_fimd0_wb(ctx);
-		if (ret < 0) {
-			dev_err(dev, "camblk setup failed.\n");
-			return ret;
-		}
+	fimc_set_type_ctrl(ctx);
+	fimc_handle_lastend(ctx, false);
 
-		set_wb.enable = 1;
-		set_wb.refresh = property->refresh_rate;
-		exynos_drm_ippnb_send_event(IPP_SET_WRITEBACK, (void *)&set_wb);
-		break;
-	case IPP_CMD_OUTPUT:
-	default:
-		ret = -EINVAL;
-		dev_err(dev, "invalid operations.\n");
-		return ret;
-	}
+	/* setup dma */
+	cfg0 = fimc_read(ctx, EXYNOS_MSCTRL);
+	cfg0 &= ~EXYNOS_MSCTRL_INPUT_MASK;
+	cfg0 |= EXYNOS_MSCTRL_INPUT_MEMORY;
+	fimc_write(ctx, cfg0, EXYNOS_MSCTRL);
 
 	/* Reset status */
 	fimc_write(ctx, 0x0, EXYNOS_CISTATUS);
@@ -1498,36 +1044,18 @@ static int fimc_ippdrv_start(struct device *dev, enum drm_exynos_ipp_cmd cmd)
 
 	fimc_clear_bits(ctx, EXYNOS_CIOCTRL, EXYNOS_CIOCTRL_WEAVE_MASK);
 
-	if (cmd == IPP_CMD_M2M)
-		fimc_set_bits(ctx, EXYNOS_MSCTRL, EXYNOS_MSCTRL_ENVID);
-
-	return 0;
+	fimc_set_bits(ctx, EXYNOS_MSCTRL, EXYNOS_MSCTRL_ENVID);
 }
 
-static void fimc_ippdrv_stop(struct device *dev, enum drm_exynos_ipp_cmd cmd)
+static void fimc_stop(struct fimc_context *ctx)
 {
-	struct fimc_context *ctx = get_fimc_context(dev);
-	struct drm_exynos_ipp_set_wb set_wb = {0, 0};
 	u32 cfg;
 
-	DRM_DEBUG_KMS("cmd[%d]\n", cmd);
-
-	switch (cmd) {
-	case IPP_CMD_M2M:
-		/* Source clear */
-		cfg = fimc_read(ctx, EXYNOS_MSCTRL);
-		cfg &= ~EXYNOS_MSCTRL_INPUT_MASK;
-		cfg &= ~EXYNOS_MSCTRL_ENVID;
-		fimc_write(ctx, cfg, EXYNOS_MSCTRL);
-		break;
-	case IPP_CMD_WB:
-		exynos_drm_ippnb_send_event(IPP_SET_WRITEBACK, (void *)&set_wb);
-		break;
-	case IPP_CMD_OUTPUT:
-	default:
-		dev_err(dev, "invalid operations.\n");
-		break;
-	}
+	/* Source clear */
+	cfg = fimc_read(ctx, EXYNOS_MSCTRL);
+	cfg &= ~EXYNOS_MSCTRL_INPUT_MASK;
+	cfg &= ~EXYNOS_MSCTRL_ENVID;
+	fimc_write(ctx, cfg, EXYNOS_MSCTRL);
 
 	fimc_mask_irq(ctx, false);
 
@@ -1545,6 +1073,87 @@ static void fimc_ippdrv_stop(struct device *dev, enum drm_exynos_ipp_cmd cmd)
 	fimc_set_bits(ctx, EXYNOS_CIGCTRL, EXYNOS_CIGCTRL_IRQ_END_DISABLE);
 }
 
+static int fimc_commit(struct exynos_drm_ipp *ipp,
+			  struct exynos_drm_ipp_task *task)
+{
+	struct fimc_context *ctx =
+			container_of(ipp, struct fimc_context, ipp);
+
+	pm_runtime_get_sync(ctx->dev);
+	ctx->task = task;
+
+	fimc_src_set_fmt(ctx, task->src.buf.fourcc, task->src.buf.modifier);
+	fimc_src_set_size(ctx, &task->src);
+	fimc_src_set_transf(ctx, DRM_MODE_ROTATE_0);
+	fimc_src_set_addr(ctx, &task->src);
+	fimc_dst_set_fmt(ctx, task->dst.buf.fourcc, task->dst.buf.modifier);
+	fimc_dst_set_transf(ctx, task->transform.rotation);
+	fimc_dst_set_size(ctx, &task->dst);
+	fimc_dst_set_addr(ctx, &task->dst);
+	fimc_set_prescaler(ctx, &ctx->sc, &task->src.rect, &task->dst.rect);
+	fimc_start(ctx);
+
+	return 0;
+}
+
+static void fimc_abort(struct exynos_drm_ipp *ipp,
+			  struct exynos_drm_ipp_task *task)
+{
+	struct fimc_context *ctx =
+			container_of(ipp, struct fimc_context, ipp);
+
+	fimc_reset(ctx);
+
+	if (ctx->task) {
+		struct exynos_drm_ipp_task *task = ctx->task;
+
+		ctx->task = NULL;
+		pm_runtime_mark_last_busy(ctx->dev);
+		pm_runtime_put_autosuspend(ctx->dev);
+		exynos_drm_ipp_task_done(task, -EIO);
+	}
+}
+
+static struct exynos_drm_ipp_funcs ipp_funcs = {
+	.commit = fimc_commit,
+	.abort = fimc_abort,
+};
+
+static int fimc_bind(struct device *dev, struct device *master, void *data)
+{
+	struct fimc_context *ctx = dev_get_drvdata(dev);
+	struct drm_device *drm_dev = data;
+	struct exynos_drm_ipp *ipp = &ctx->ipp;
+
+	ctx->drm_dev = drm_dev;
+	drm_iommu_attach_device(drm_dev, dev);
+
+	exynos_drm_ipp_register(drm_dev, ipp, &ipp_funcs,
+			DRM_EXYNOS_IPP_CAP_CROP | DRM_EXYNOS_IPP_CAP_ROTATE |
+			DRM_EXYNOS_IPP_CAP_SCALE | DRM_EXYNOS_IPP_CAP_CONVERT,
+			ctx->formats, ctx->num_formats, "fimc");
+
+	dev_info(dev, "The exynos fimc has been probed successfully\n");
+
+	return 0;
+}
+
+static void fimc_unbind(struct device *dev, struct device *master,
+			void *data)
+{
+	struct fimc_context *ctx = dev_get_drvdata(dev);
+	struct drm_device *drm_dev = data;
+	struct exynos_drm_ipp *ipp = &ctx->ipp;
+
+	exynos_drm_ipp_unregister(drm_dev, ipp);
+	drm_iommu_detach_device(drm_dev, dev);
+}
+
+static const struct component_ops fimc_component_ops = {
+	.bind	= fimc_bind,
+	.unbind = fimc_unbind,
+};
+
 static void fimc_put_clocks(struct fimc_context *ctx)
 {
 	int i;
@@ -1559,7 +1168,7 @@ static void fimc_put_clocks(struct fimc_context *ctx)
 
 static int fimc_setup_clocks(struct fimc_context *ctx)
 {
-	struct device *fimc_dev = ctx->ippdrv.dev;
+	struct device *fimc_dev = ctx->dev;
 	struct device *dev;
 	int ret, i;
 
@@ -1574,8 +1183,6 @@ static int fimc_setup_clocks(struct fimc_context *ctx)
 
 		ctx->clocks[i] = clk_get(dev, fimc_clock_names[i]);
 		if (IS_ERR(ctx->clocks[i])) {
-			if (i >= FIMC_CLK_MUX)
-				break;
 			ret = PTR_ERR(ctx->clocks[i]);
 			dev_err(fimc_dev, "failed to get clock: %s\n",
 						fimc_clock_names[i]);
@@ -1583,20 +1190,6 @@ static int fimc_setup_clocks(struct fimc_context *ctx)
 		}
 	}
 
-	/* Optional FIMC LCLK parent clock setting */
-	if (!IS_ERR(ctx->clocks[FIMC_CLK_PARENT])) {
-		ret = clk_set_parent(ctx->clocks[FIMC_CLK_MUX],
-				     ctx->clocks[FIMC_CLK_PARENT]);
-		if (ret < 0) {
-			dev_err(fimc_dev, "failed to set parent.\n");
-			goto e_clk_free;
-		}
-	}
-
-	ret = clk_set_rate(ctx->clocks[FIMC_CLK_LCLK], ctx->clk_frequency);
-	if (ret < 0)
-		goto e_clk_free;
-
 	ret = clk_prepare_enable(ctx->clocks[FIMC_CLK_LCLK]);
 	if (!ret)
 		return ret;
@@ -1605,57 +1198,118 @@ e_clk_free:
 	return ret;
 }
 
-static int fimc_parse_dt(struct fimc_context *ctx)
+int exynos_drm_check_fimc_device(struct device *dev)
 {
-	struct device_node *node = ctx->ippdrv.dev->of_node;
+	unsigned int id = of_alias_get_id(dev->of_node, "fimc");
 
-	/* Handle only devices that support the LCD Writeback data path */
-	if (!of_property_read_bool(node, "samsung,lcd-wb"))
-		return -ENODEV;
+	if (id >= 0 && (BIT(id) & fimc_mask))
+		return 0;
+	return -ENODEV;
+}
 
-	if (of_property_read_u32(node, "clock-frequency",
-					&ctx->clk_frequency))
-		ctx->clk_frequency = FIMC_DEFAULT_LCLK_FREQUENCY;
+static const unsigned int fimc_formats[] = {
+	DRM_FORMAT_XRGB8888, DRM_FORMAT_RGB565,
+	DRM_FORMAT_NV12, DRM_FORMAT_NV16, DRM_FORMAT_NV21, DRM_FORMAT_NV61,
+	DRM_FORMAT_UYVY, DRM_FORMAT_VYUY, DRM_FORMAT_YUYV, DRM_FORMAT_YVYU,
+	DRM_FORMAT_YUV420, DRM_FORMAT_YVU420, DRM_FORMAT_YUV422,
+	DRM_FORMAT_YUV444,
+};
 
-	ctx->id = of_alias_get_id(node, "fimc");
+static const unsigned int fimc_tiled_formats[] = {
+	DRM_FORMAT_NV12, DRM_FORMAT_NV21,
+};
 
-	if (ctx->id < 0) {
-		dev_err(ctx->ippdrv.dev, "failed to get node alias id.\n");
-		return -EINVAL;
-	}
+static const struct drm_exynos_ipp_limit fimc_4210_limits_v1[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 16, 8192, 8 }, .v = { 16, 8192, 2 }) },
+	{ IPP_SIZE_LIMIT(AREA, .h = { 16, 4224, 2 }, .v = { 16, 0, 2 }) },
+	{ IPP_SIZE_LIMIT(ROTATED, .h = { 128, 1920 }, .v = { 128, 0 }) },
+	{ IPP_SCALE_LIMIT(.h = { (1 << 16) / 64, (1 << 16) * 64 },
+			  .v = { (1 << 16) / 64, (1 << 16) * 64 }) },
+};
 
-	return 0;
-}
+static const struct drm_exynos_ipp_limit fimc_4210_limits_v2[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 16, 8192, 8 }, .v = { 16, 8192, 2 }) },
+	{ IPP_SIZE_LIMIT(AREA, .h = { 16, 1920, 2 }, .v = { 16, 0, 2 }) },
+	{ IPP_SIZE_LIMIT(ROTATED, .h = { 128, 1366 }, .v = { 128, 0 }) },
+	{ IPP_SCALE_LIMIT(.h = { (1 << 16) / 64, (1 << 16) * 64 },
+			  .v = { (1 << 16) / 64, (1 << 16) * 64 }) },
+};
+
+static const struct drm_exynos_ipp_limit fimc_4210_limits_tiled_v1[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 128, 1920, 128 }, .v = { 32, 1920, 32 }) },
+	{ IPP_SIZE_LIMIT(AREA, .h = { 128, 1920, 2 }, .v = { 128, 0, 2 }) },
+	{ IPP_SCALE_LIMIT(.h = { (1 << 16) / 64, (1 << 16) * 64 },
+			  .v = { (1 << 16) / 64, (1 << 16) * 64 }) },
+};
+
+static const struct drm_exynos_ipp_limit fimc_4210_limits_tiled_v2[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 128, 1920, 128 }, .v = { 32, 1920, 32 }) },
+	{ IPP_SIZE_LIMIT(AREA, .h = { 128, 1366, 2 }, .v = { 128, 0, 2 }) },
+	{ IPP_SCALE_LIMIT(.h = { (1 << 16) / 64, (1 << 16) * 64 },
+			  .v = { (1 << 16) / 64, (1 << 16) * 64 }) },
+};
 
 static int fimc_probe(struct platform_device *pdev)
 {
+	const struct drm_exynos_ipp_limit *limits;
+	struct exynos_drm_ipp_formats *formats;
 	struct device *dev = &pdev->dev;
 	struct fimc_context *ctx;
 	struct resource *res;
-	struct exynos_drm_ippdrv *ippdrv;
 	int ret;
+	int i, j, num_limits, num_formats;
 
-	if (!dev->of_node) {
-		dev_err(dev, "device tree node not found.\n");
+	if (exynos_drm_check_fimc_device(dev) != 0)
 		return -ENODEV;
-	}
 
 	ctx = devm_kzalloc(dev, sizeof(*ctx), GFP_KERNEL);
 	if (!ctx)
 		return -ENOMEM;
 
-	ctx->ippdrv.dev = dev;
+	ctx->dev = dev;
+	ctx->id = of_alias_get_id(dev->of_node, "fimc");
 
-	ret = fimc_parse_dt(ctx);
-	if (ret < 0)
-		return ret;
+	/* construct formats/limits array */
+	num_formats = ARRAY_SIZE(fimc_formats) + ARRAY_SIZE(fimc_tiled_formats);
+	formats = devm_kzalloc(dev, sizeof(*formats) * num_formats, GFP_KERNEL);
+	if (!formats)
+		return -ENOMEM;
+
+	/* linear formats */
+	if (ctx->id < 3) {
+		limits = fimc_4210_limits_v1;
+		num_limits = ARRAY_SIZE(fimc_4210_limits_v1);
+	} else {
+		limits = fimc_4210_limits_v2;
+		num_limits = ARRAY_SIZE(fimc_4210_limits_v2);
+	}
+	for (i = 0; i < ARRAY_SIZE(fimc_formats); i++) {
+		formats[i].fourcc = fimc_formats[i];
+		formats[i].type = DRM_EXYNOS_IPP_FORMAT_SOURCE |
+				  DRM_EXYNOS_IPP_FORMAT_DESTINATION;
+		formats[i].limits = limits;
+		formats[i].num_limits = num_limits;
+	}
 
-	ctx->sysreg = syscon_regmap_lookup_by_phandle(dev->of_node,
-						"samsung,sysreg");
-	if (IS_ERR(ctx->sysreg)) {
-		dev_err(dev, "syscon regmap lookup failed.\n");
-		return PTR_ERR(ctx->sysreg);
+	/* tiled formats */
+	if (ctx->id < 3) {
+		limits = fimc_4210_limits_tiled_v1;
+		num_limits = ARRAY_SIZE(fimc_4210_limits_tiled_v1);
+	} else {
+		limits = fimc_4210_limits_tiled_v2;
+		num_limits = ARRAY_SIZE(fimc_4210_limits_tiled_v2);
 	}
+	for (j = i, i = 0; i < ARRAY_SIZE(fimc_tiled_formats); j++, i++) {
+		formats[j].fourcc = fimc_tiled_formats[i];
+		formats[j].modifier = DRM_FORMAT_MOD_SAMSUNG_64_32_TILE;
+		formats[j].type = DRM_EXYNOS_IPP_FORMAT_SOURCE |
+				  DRM_EXYNOS_IPP_FORMAT_DESTINATION;
+		formats[j].limits = limits;
+		formats[j].num_limits = num_limits;
+	}
+
+	ctx->formats = formats;
+	ctx->num_formats = num_formats;
 
 	/* resource memory */
 	ctx->regs_res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
@@ -1670,9 +1324,8 @@ static int fimc_probe(struct platform_device *pdev)
 		return -ENOENT;
 	}
 
-	ctx->irq = res->start;
-	ret = devm_request_threaded_irq(dev, ctx->irq, NULL, fimc_irq_handler,
-		IRQF_ONESHOT, "drm_fimc", ctx);
+	ret = devm_request_irq(dev, res->start, fimc_irq_handler,
+		0, dev_name(dev), ctx);
 	if (ret < 0) {
 		dev_err(dev, "failed to request irq.\n");
 		return ret;
@@ -1682,39 +1335,24 @@ static int fimc_probe(struct platform_device *pdev)
 	if (ret < 0)
 		return ret;
 
-	ippdrv = &ctx->ippdrv;
-	ippdrv->ops[EXYNOS_DRM_OPS_SRC] = &fimc_src_ops;
-	ippdrv->ops[EXYNOS_DRM_OPS_DST] = &fimc_dst_ops;
-	ippdrv->check_property = fimc_ippdrv_check_property;
-	ippdrv->reset = fimc_ippdrv_reset;
-	ippdrv->start = fimc_ippdrv_start;
-	ippdrv->stop = fimc_ippdrv_stop;
-	ret = fimc_init_prop_list(ippdrv);
-	if (ret < 0) {
-		dev_err(dev, "failed to init property list.\n");
-		goto err_put_clk;
-	}
-
-	DRM_DEBUG_KMS("id[%d]ippdrv[%pK]\n", ctx->id, ippdrv);
-
 	spin_lock_init(&ctx->lock);
 	platform_set_drvdata(pdev, ctx);
 
+	pm_runtime_use_autosuspend(dev);
+	pm_runtime_set_autosuspend_delay(dev, FIMC_AUTOSUSPEND_DELAY);
 	pm_runtime_enable(dev);
 
-	ret = exynos_drm_ippdrv_register(ippdrv);
-	if (ret < 0) {
-		dev_err(dev, "failed to register drm fimc device.\n");
+	ret = component_add(dev, &fimc_component_ops);
+	if (ret)
 		goto err_pm_dis;
-	}
 
 	dev_info(dev, "drm fimc registered successfully.\n");
 
 	return 0;
 
 err_pm_dis:
+	pm_runtime_dont_use_autosuspend(dev);
 	pm_runtime_disable(dev);
-err_put_clk:
 	fimc_put_clocks(ctx);
 
 	return ret;
@@ -1724,42 +1362,24 @@ static int fimc_remove(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;
 	struct fimc_context *ctx = get_fimc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
 
-	exynos_drm_ippdrv_unregister(ippdrv);
+	component_del(dev, &fimc_component_ops);
+	pm_runtime_dont_use_autosuspend(dev);
+	pm_runtime_disable(dev);
 
 	fimc_put_clocks(ctx);
-	pm_runtime_set_suspended(dev);
-	pm_runtime_disable(dev);
 
 	return 0;
 }
 
 #ifdef CONFIG_PM
-static int fimc_clk_ctrl(struct fimc_context *ctx, bool enable)
-{
-	DRM_DEBUG_KMS("enable[%d]\n", enable);
-
-	if (enable) {
-		clk_prepare_enable(ctx->clocks[FIMC_CLK_GATE]);
-		clk_prepare_enable(ctx->clocks[FIMC_CLK_WB_A]);
-		ctx->suspended = false;
-	} else {
-		clk_disable_unprepare(ctx->clocks[FIMC_CLK_GATE]);
-		clk_disable_unprepare(ctx->clocks[FIMC_CLK_WB_A]);
-		ctx->suspended = true;
-	}
-
-	return 0;
-}
-
 static int fimc_runtime_suspend(struct device *dev)
 {
 	struct fimc_context *ctx = get_fimc_context(dev);
 
 	DRM_DEBUG_KMS("id[%d]\n", ctx->id);
-
-	return  fimc_clk_ctrl(ctx, false);
+	clk_disable_unprepare(ctx->clocks[FIMC_CLK_GATE]);
+	return 0;
 }
 
 static int fimc_runtime_resume(struct device *dev)
@@ -1767,8 +1387,7 @@ static int fimc_runtime_resume(struct device *dev)
 	struct fimc_context *ctx = get_fimc_context(dev);
 
 	DRM_DEBUG_KMS("id[%d]\n", ctx->id);
-
-	return  fimc_clk_ctrl(ctx, true);
+	return clk_prepare_enable(ctx->clocks[FIMC_CLK_GATE]);
 }
 #endif
 
@@ -1795,4 +1414,3 @@ struct platform_driver fimc_driver = {
 		.pm	= &fimc_pm_ops,
 	},
 };
-
diff --git a/drivers/gpu/drm/exynos/exynos_drm_fimc.h b/drivers/gpu/drm/exynos/exynos_drm_fimc.h
deleted file mode 100644
index 127a424..0000000
--- a/drivers/gpu/drm/exynos/exynos_drm_fimc.h
+++ /dev/null
@@ -1,23 +0,0 @@
-/*
- * Copyright (c) 2012 Samsung Electronics Co., Ltd.
- *
- * Authors:
- *	Eunchul Kim <chulspro.kim@samsung.com>
- *	Jinyoung Jeon <jy0.jeon@samsung.com>
- *	Sangmin Lee <lsmin.lee@samsung.com>
- *
- * This program is free software; you can redistribute  it and/or modify it
- * under  the terms of  the GNU General  Public License as published by the
- * Free Software Foundation;  either version 2 of the  License, or (at your
- * option) any later version.
- */
-
-#ifndef _EXYNOS_DRM_FIMC_H_
-#define _EXYNOS_DRM_FIMC_H_
-
-/*
- * TODO
- * FIMD output interface notifier callback.
- */
-
-#endif /* _EXYNOS_DRM_FIMC_H_ */
-- 
cgit v1.1


From 01fb9185dc180940f90510215ef8764d6155d088 Mon Sep 17 00:00:00 2001
From: Andrzej Pietrasiewicz <andrzej.p@samsung.com>
Date: Wed, 9 May 2018 10:59:26 +0200
Subject: drm/exynos: Add driver for Exynos Scaler module

Exynos Scaler is a hardware module, which processes graphic data fetched
from memory and transfers the resultant dato another memory buffer.
Graphics data can be up/down-scaled, rotated, flipped and converted color
space. Scaler hardware modules are a part of Exynos5420 and newer Exynos
SoCs.

Signed-off-by: Andrzej Pietrasiewicz <andrzej.p@samsung.com>
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
---
 .../devicetree/bindings/gpu/samsung-scaler.txt     |  27 +
 drivers/gpu/drm/exynos/Kconfig                     |   6 +
 drivers/gpu/drm/exynos/Makefile                    |   1 +
 drivers/gpu/drm/exynos/exynos_drm_drv.c            |   3 +
 drivers/gpu/drm/exynos/exynos_drm_drv.h            |   1 +
 drivers/gpu/drm/exynos/exynos_drm_scaler.c         | 694 +++++++++++++++++++++
 drivers/gpu/drm/exynos/regs-scaler.h               | 426 +++++++++++++
 7 files changed, 1158 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/gpu/samsung-scaler.txt
 create mode 100644 drivers/gpu/drm/exynos/exynos_drm_scaler.c
 create mode 100644 drivers/gpu/drm/exynos/regs-scaler.h

diff --git a/Documentation/devicetree/bindings/gpu/samsung-scaler.txt b/Documentation/devicetree/bindings/gpu/samsung-scaler.txt
new file mode 100644
index 0000000..9c3d981
--- /dev/null
+++ b/Documentation/devicetree/bindings/gpu/samsung-scaler.txt
@@ -0,0 +1,27 @@
+* Samsung Exynos Image Scaler
+
+Required properties:
+  - compatible : value should be one of the following:
+	(a) "samsung,exynos5420-scaler" for Scaler IP in Exynos5420
+	(b) "samsung,exynos5433-scaler" for Scaler IP in Exynos5433
+
+  - reg : Physical base address of the IP registers and length of memory
+	  mapped region.
+
+  - interrupts : Interrupt specifier for scaler interrupt, according to format
+		 specific to interrupt parent.
+
+  - clocks : Clock specifier for scaler clock, according to generic clock
+	     bindings. (See Documentation/devicetree/bindings/clock/exynos*.txt)
+
+  - clock-names : Names of clocks. For exynos scaler, it should be "mscl"
+		  on 5420 and "pclk", "aclk" and "aclk_xiu" on 5433.
+
+Example:
+	scaler@12800000 {
+		compatible = "samsung,exynos5420-scaler";
+		reg = <0x12800000 0x1294>;
+		interrupts = <0 220 IRQ_TYPE_LEVEL_HIGH>;
+		clocks = <&clock CLK_MSCL0>;
+		clock-names = "mscl";
+	};
diff --git a/drivers/gpu/drm/exynos/Kconfig b/drivers/gpu/drm/exynos/Kconfig
index 54f5703..208bc27 100644
--- a/drivers/gpu/drm/exynos/Kconfig
+++ b/drivers/gpu/drm/exynos/Kconfig
@@ -110,6 +110,12 @@ config DRM_EXYNOS_ROTATOR
 	help
 	  Choose this option if you want to use Exynos Rotator for DRM.
 
+config DRM_EXYNOS_SCALER
+	bool "Scaler"
+	select DRM_EXYNOS_IPP
+	help
+	  Choose this option if you want to use Exynos Scaler for DRM.
+
 config DRM_EXYNOS_GSC
 	bool "GScaler"
 	depends on VIDEO_SAMSUNG_EXYNOS_GSC=n
diff --git a/drivers/gpu/drm/exynos/Makefile b/drivers/gpu/drm/exynos/Makefile
index bdf4212..3b323f1 100644
--- a/drivers/gpu/drm/exynos/Makefile
+++ b/drivers/gpu/drm/exynos/Makefile
@@ -21,6 +21,7 @@ exynosdrm-$(CONFIG_DRM_EXYNOS_G2D)	+= exynos_drm_g2d.o
 exynosdrm-$(CONFIG_DRM_EXYNOS_IPP)	+= exynos_drm_ipp.o
 exynosdrm-$(CONFIG_DRM_EXYNOS_FIMC)	+= exynos_drm_fimc.o
 exynosdrm-$(CONFIG_DRM_EXYNOS_ROTATOR)	+= exynos_drm_rotator.o
+exynosdrm-$(CONFIG_DRM_EXYNOS_SCALER)	+= exynos_drm_scaler.o
 exynosdrm-$(CONFIG_DRM_EXYNOS_GSC)	+= exynos_drm_gsc.o
 exynosdrm-$(CONFIG_DRM_EXYNOS_MIC)     += exynos_drm_mic.o
 
diff --git a/drivers/gpu/drm/exynos/exynos_drm_drv.c b/drivers/gpu/drm/exynos/exynos_drm_drv.c
index 7ba13c1..f55ce44 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_drv.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_drv.c
@@ -267,6 +267,9 @@ static struct exynos_drm_driver_info exynos_drm_drivers[] = {
 		DRV_PTR(rotator_driver, CONFIG_DRM_EXYNOS_ROTATOR),
 		DRM_COMPONENT_DRIVER
 	}, {
+		DRV_PTR(scaler_driver, CONFIG_DRM_EXYNOS_SCALER),
+		DRM_COMPONENT_DRIVER
+	}, {
 		DRV_PTR(gsc_driver, CONFIG_DRM_EXYNOS_GSC),
 		DRM_COMPONENT_DRIVER
 	}, {
diff --git a/drivers/gpu/drm/exynos/exynos_drm_drv.h b/drivers/gpu/drm/exynos/exynos_drm_drv.h
index 0834e7e..c07e6f3 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_drv.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_drv.h
@@ -298,6 +298,7 @@ extern struct platform_driver vidi_driver;
 extern struct platform_driver g2d_driver;
 extern struct platform_driver fimc_driver;
 extern struct platform_driver rotator_driver;
+extern struct platform_driver scaler_driver;
 extern struct platform_driver gsc_driver;
 extern struct platform_driver mic_driver;
 #endif
diff --git a/drivers/gpu/drm/exynos/exynos_drm_scaler.c b/drivers/gpu/drm/exynos/exynos_drm_scaler.c
new file mode 100644
index 0000000..63b05b7
--- /dev/null
+++ b/drivers/gpu/drm/exynos/exynos_drm_scaler.c
@@ -0,0 +1,694 @@
+/*
+ * Copyright (C) 2017 Samsung Electronics Co.Ltd
+ * Author:
+ *	Andrzej Pietrasiewicz <andrzej.p@samsung.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundationr
+ */
+
+#include <linux/kernel.h>
+#include <linux/component.h>
+#include <linux/err.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/platform_device.h>
+#include <linux/clk.h>
+#include <linux/of_device.h>
+#include <linux/pm_runtime.h>
+
+#include <drm/drmP.h>
+#include <drm/exynos_drm.h>
+#include "regs-scaler.h"
+#include "exynos_drm_fb.h"
+#include "exynos_drm_drv.h"
+#include "exynos_drm_iommu.h"
+#include "exynos_drm_ipp.h"
+
+#define scaler_read(offset)		readl(scaler->regs + (offset))
+#define scaler_write(cfg, offset)	writel(cfg, scaler->regs + (offset))
+#define SCALER_MAX_CLK			4
+#define SCALER_AUTOSUSPEND_DELAY	2000
+
+struct scaler_data {
+	const char	*clk_name[SCALER_MAX_CLK];
+	unsigned int	num_clk;
+	const struct exynos_drm_ipp_formats *formats;
+	unsigned int	num_formats;
+};
+
+struct scaler_context {
+	struct exynos_drm_ipp		ipp;
+	struct drm_device		*drm_dev;
+	struct device			*dev;
+	void __iomem			*regs;
+	struct clk			*clock[SCALER_MAX_CLK];
+	struct exynos_drm_ipp_task	*task;
+	const struct scaler_data	*scaler_data;
+};
+
+static u32 scaler_get_format(u32 drm_fmt)
+{
+	switch (drm_fmt) {
+	case DRM_FORMAT_NV21:
+		return SCALER_YUV420_2P_UV;
+	case DRM_FORMAT_NV12:
+		return SCALER_YUV420_2P_VU;
+	case DRM_FORMAT_YUV420:
+		return SCALER_YUV420_3P;
+	case DRM_FORMAT_YUYV:
+		return SCALER_YUV422_1P_YUYV;
+	case DRM_FORMAT_UYVY:
+		return SCALER_YUV422_1P_UYVY;
+	case DRM_FORMAT_YVYU:
+		return SCALER_YUV422_1P_YVYU;
+	case DRM_FORMAT_NV61:
+		return SCALER_YUV422_2P_UV;
+	case DRM_FORMAT_NV16:
+		return SCALER_YUV422_2P_VU;
+	case DRM_FORMAT_YUV422:
+		return SCALER_YUV422_3P;
+	case DRM_FORMAT_NV42:
+		return SCALER_YUV444_2P_UV;
+	case DRM_FORMAT_NV24:
+		return SCALER_YUV444_2P_VU;
+	case DRM_FORMAT_YUV444:
+		return SCALER_YUV444_3P;
+	case DRM_FORMAT_RGB565:
+		return SCALER_RGB_565;
+	case DRM_FORMAT_XRGB1555:
+		return SCALER_ARGB1555;
+	case DRM_FORMAT_ARGB1555:
+		return SCALER_ARGB1555;
+	case DRM_FORMAT_XRGB4444:
+		return SCALER_ARGB4444;
+	case DRM_FORMAT_ARGB4444:
+		return SCALER_ARGB4444;
+	case DRM_FORMAT_XRGB8888:
+		return SCALER_ARGB8888;
+	case DRM_FORMAT_ARGB8888:
+		return SCALER_ARGB8888;
+	case DRM_FORMAT_RGBX8888:
+		return SCALER_RGBA8888;
+	case DRM_FORMAT_RGBA8888:
+		return SCALER_RGBA8888;
+	default:
+		break;
+	}
+
+	return 0;
+}
+
+static inline void scaler_enable_int(struct scaler_context *scaler)
+{
+	u32 val;
+
+	val = SCALER_INT_EN_TIMEOUT |
+		SCALER_INT_EN_ILLEGAL_BLEND |
+		SCALER_INT_EN_ILLEGAL_RATIO |
+		SCALER_INT_EN_ILLEGAL_DST_HEIGHT |
+		SCALER_INT_EN_ILLEGAL_DST_WIDTH |
+		SCALER_INT_EN_ILLEGAL_DST_V_POS |
+		SCALER_INT_EN_ILLEGAL_DST_H_POS |
+		SCALER_INT_EN_ILLEGAL_DST_C_SPAN |
+		SCALER_INT_EN_ILLEGAL_DST_Y_SPAN |
+		SCALER_INT_EN_ILLEGAL_DST_CR_BASE |
+		SCALER_INT_EN_ILLEGAL_DST_CB_BASE |
+		SCALER_INT_EN_ILLEGAL_DST_Y_BASE |
+		SCALER_INT_EN_ILLEGAL_DST_COLOR |
+		SCALER_INT_EN_ILLEGAL_SRC_HEIGHT |
+		SCALER_INT_EN_ILLEGAL_SRC_WIDTH |
+		SCALER_INT_EN_ILLEGAL_SRC_CV_POS |
+		SCALER_INT_EN_ILLEGAL_SRC_CH_POS |
+		SCALER_INT_EN_ILLEGAL_SRC_YV_POS |
+		SCALER_INT_EN_ILLEGAL_SRC_YH_POS |
+		SCALER_INT_EN_ILLEGAL_DST_SPAN |
+		SCALER_INT_EN_ILLEGAL_SRC_Y_SPAN |
+		SCALER_INT_EN_ILLEGAL_SRC_CR_BASE |
+		SCALER_INT_EN_ILLEGAL_SRC_CB_BASE |
+		SCALER_INT_EN_ILLEGAL_SRC_Y_BASE |
+		SCALER_INT_EN_ILLEGAL_SRC_COLOR |
+		SCALER_INT_EN_FRAME_END;
+	scaler_write(val, SCALER_INT_EN);
+}
+
+static inline void scaler_set_src_fmt(struct scaler_context *scaler,
+	u32 src_fmt)
+{
+	u32 val;
+
+	val = SCALER_SRC_CFG_SET_COLOR_FORMAT(src_fmt);
+	scaler_write(val, SCALER_SRC_CFG);
+}
+
+static inline void scaler_set_src_base(struct scaler_context *scaler,
+	struct exynos_drm_ipp_buffer *src_buf)
+{
+	static unsigned int bases[] = {
+		SCALER_SRC_Y_BASE,
+		SCALER_SRC_CB_BASE,
+		SCALER_SRC_CR_BASE,
+	};
+	int i;
+
+	for (i = 0; i < src_buf->format->num_planes; ++i)
+		scaler_write(src_buf->dma_addr[i], bases[i]);
+}
+
+static inline void scaler_set_src_span(struct scaler_context *scaler,
+	struct exynos_drm_ipp_buffer *src_buf)
+{
+	u32 val;
+
+	val = SCALER_SRC_SPAN_SET_Y_SPAN(src_buf->buf.pitch[0] /
+		src_buf->format->cpp[0]);
+
+	if (src_buf->format->num_planes > 1)
+		val |= SCALER_SRC_SPAN_SET_C_SPAN(src_buf->buf.pitch[1]);
+
+	scaler_write(val, SCALER_SRC_SPAN);
+}
+
+static inline void scaler_set_src_luma_pos(struct scaler_context *scaler,
+	struct drm_exynos_ipp_task_rect *src_pos)
+{
+	u32 val;
+
+	val = SCALER_SRC_Y_POS_SET_YH_POS(src_pos->x << 2);
+	val |=  SCALER_SRC_Y_POS_SET_YV_POS(src_pos->y << 2);
+	scaler_write(val, SCALER_SRC_Y_POS);
+	scaler_write(val, SCALER_SRC_C_POS); /* ATTENTION! */
+}
+
+static inline void scaler_set_src_wh(struct scaler_context *scaler,
+	struct drm_exynos_ipp_task_rect *src_pos)
+{
+	u32 val;
+
+	val = SCALER_SRC_WH_SET_WIDTH(src_pos->w);
+	val |= SCALER_SRC_WH_SET_HEIGHT(src_pos->h);
+	scaler_write(val, SCALER_SRC_WH);
+}
+
+static inline void scaler_set_dst_fmt(struct scaler_context *scaler,
+	u32 dst_fmt)
+{
+	u32 val;
+
+	val = SCALER_DST_CFG_SET_COLOR_FORMAT(dst_fmt);
+	scaler_write(val, SCALER_DST_CFG);
+}
+
+static inline void scaler_set_dst_base(struct scaler_context *scaler,
+	struct exynos_drm_ipp_buffer *dst_buf)
+{
+	static unsigned int bases[] = {
+		SCALER_DST_Y_BASE,
+		SCALER_DST_CB_BASE,
+		SCALER_DST_CR_BASE,
+	};
+	int i;
+
+	for (i = 0; i < dst_buf->format->num_planes; ++i)
+		scaler_write(dst_buf->dma_addr[i], bases[i]);
+}
+
+static inline void scaler_set_dst_span(struct scaler_context *scaler,
+	struct exynos_drm_ipp_buffer *dst_buf)
+{
+	u32 val;
+
+	val = SCALER_DST_SPAN_SET_Y_SPAN(dst_buf->buf.pitch[0] /
+		dst_buf->format->cpp[0]);
+
+	if (dst_buf->format->num_planes > 1)
+		val |= SCALER_DST_SPAN_SET_C_SPAN(dst_buf->buf.pitch[1]);
+
+	scaler_write(val, SCALER_DST_SPAN);
+}
+
+static inline void scaler_set_dst_luma_pos(struct scaler_context *scaler,
+	struct drm_exynos_ipp_task_rect *dst_pos)
+{
+	u32 val;
+
+	val = SCALER_DST_WH_SET_WIDTH(dst_pos->w);
+	val |= SCALER_DST_WH_SET_HEIGHT(dst_pos->h);
+	scaler_write(val, SCALER_DST_WH);
+}
+
+static inline void scaler_set_dst_wh(struct scaler_context *scaler,
+	struct drm_exynos_ipp_task_rect *dst_pos)
+{
+	u32 val;
+
+	val = SCALER_DST_POS_SET_H_POS(dst_pos->x);
+	val |= SCALER_DST_POS_SET_V_POS(dst_pos->y);
+	scaler_write(val, SCALER_DST_POS);
+}
+
+static inline void scaler_set_hv_ratio(struct scaler_context *scaler,
+	unsigned int rotation,
+	struct drm_exynos_ipp_task_rect *src_pos,
+	struct drm_exynos_ipp_task_rect *dst_pos)
+{
+	u32 val, h_ratio, v_ratio;
+
+	if (drm_rotation_90_or_270(rotation)) {
+		h_ratio = (src_pos->h << 16) / dst_pos->w;
+		v_ratio = (src_pos->w << 16) / dst_pos->h;
+	} else {
+		h_ratio = (src_pos->w << 16) / dst_pos->w;
+		v_ratio = (src_pos->h << 16) / dst_pos->h;
+	}
+
+	val = SCALER_H_RATIO_SET(h_ratio);
+	scaler_write(val, SCALER_H_RATIO);
+
+	val = SCALER_V_RATIO_SET(v_ratio);
+	scaler_write(val, SCALER_V_RATIO);
+}
+
+static inline void scaler_set_rotation(struct scaler_context *scaler,
+	unsigned int rotation)
+{
+	u32 val = 0;
+
+	if (rotation & DRM_MODE_ROTATE_90)
+		val |= SCALER_ROT_CFG_SET_ROTMODE(SCALER_ROT_MODE_90);
+	else if (rotation & DRM_MODE_ROTATE_180)
+		val |= SCALER_ROT_CFG_SET_ROTMODE(SCALER_ROT_MODE_180);
+	else if (rotation & DRM_MODE_ROTATE_270)
+		val |= SCALER_ROT_CFG_SET_ROTMODE(SCALER_ROT_MODE_270);
+	if (rotation & DRM_MODE_REFLECT_X)
+		val |= SCALER_ROT_CFG_FLIP_X_EN;
+	if (rotation & DRM_MODE_REFLECT_Y)
+		val |= SCALER_ROT_CFG_FLIP_Y_EN;
+	scaler_write(val, SCALER_ROT_CFG);
+}
+
+static inline void scaler_set_csc(struct scaler_context *scaler,
+	const struct drm_format_info *fmt)
+{
+	static const u32 csc_mtx[2][3][3] = {
+		{ /* YCbCr to RGB */
+			{0x254, 0x000, 0x331},
+			{0x254, 0xf38, 0xe60},
+			{0x254, 0x409, 0x000},
+		},
+		{ /* RGB to YCbCr */
+			{0x084, 0x102, 0x032},
+			{0xfb4, 0xf6b, 0x0e1},
+			{0x0e1, 0xf44, 0xfdc},
+		},
+	};
+	int i, j, dir;
+
+	switch (fmt->format) {
+	case DRM_FORMAT_RGB565:
+	case DRM_FORMAT_XRGB1555:
+	case DRM_FORMAT_ARGB1555:
+	case DRM_FORMAT_XRGB4444:
+	case DRM_FORMAT_ARGB4444:
+	case DRM_FORMAT_XRGB8888:
+	case DRM_FORMAT_ARGB8888:
+	case DRM_FORMAT_RGBX8888:
+	case DRM_FORMAT_RGBA8888:
+		dir = 1;
+		break;
+	default:
+		dir = 0;
+	}
+
+	for (i = 0; i < 3; i++)
+		for (j = 0; j < 3; j++)
+			scaler_write(csc_mtx[dir][i][j], SCALER_CSC_COEF(j, i));
+}
+
+static inline void scaler_set_timer(struct scaler_context *scaler,
+	unsigned int timer, unsigned int divider)
+{
+	u32 val;
+
+	val = SCALER_TIMEOUT_CTRL_TIMER_ENABLE;
+	val |= SCALER_TIMEOUT_CTRL_SET_TIMER_VALUE(timer);
+	val |= SCALER_TIMEOUT_CTRL_SET_TIMER_DIV(divider);
+	scaler_write(val, SCALER_TIMEOUT_CTRL);
+}
+
+static inline void scaler_start_hw(struct scaler_context *scaler)
+{
+	scaler_write(SCALER_CFG_START_CMD, SCALER_CFG);
+}
+
+static int scaler_commit(struct exynos_drm_ipp *ipp,
+			  struct exynos_drm_ipp_task *task)
+{
+	struct scaler_context *scaler =
+			container_of(ipp, struct scaler_context, ipp);
+
+	u32 src_fmt = scaler_get_format(task->src.buf.fourcc);
+	struct drm_exynos_ipp_task_rect *src_pos = &task->src.rect;
+
+	u32 dst_fmt = scaler_get_format(task->dst.buf.fourcc);
+	struct drm_exynos_ipp_task_rect *dst_pos = &task->dst.rect;
+
+	scaler->task = task;
+
+	pm_runtime_get_sync(scaler->dev);
+
+	scaler_set_src_fmt(scaler, src_fmt);
+	scaler_set_src_base(scaler, &task->src);
+	scaler_set_src_span(scaler, &task->src);
+	scaler_set_src_luma_pos(scaler, src_pos);
+	scaler_set_src_wh(scaler, src_pos);
+
+	scaler_set_dst_fmt(scaler, dst_fmt);
+	scaler_set_dst_base(scaler, &task->dst);
+	scaler_set_dst_span(scaler, &task->dst);
+	scaler_set_dst_luma_pos(scaler, dst_pos);
+	scaler_set_dst_wh(scaler, dst_pos);
+
+	scaler_set_hv_ratio(scaler, task->transform.rotation, src_pos, dst_pos);
+	scaler_set_rotation(scaler, task->transform.rotation);
+
+	scaler_set_csc(scaler, task->src.format);
+
+	scaler_set_timer(scaler, 0xffff, 0xf);
+
+	scaler_enable_int(scaler);
+	scaler_start_hw(scaler);
+
+	return 0;
+}
+
+static struct exynos_drm_ipp_funcs ipp_funcs = {
+	.commit = scaler_commit,
+};
+
+static inline void scaler_disable_int(struct scaler_context *scaler)
+{
+	scaler_write(0, SCALER_INT_EN);
+}
+
+static inline u32 scaler_get_int_status(struct scaler_context *scaler)
+{
+	return scaler_read(SCALER_INT_STATUS);
+}
+
+static inline bool scaler_task_done(u32 val)
+{
+	return val & SCALER_INT_STATUS_FRAME_END ? 0 : -EINVAL;
+}
+
+static irqreturn_t scaler_irq_handler(int irq, void *arg)
+{
+	struct scaler_context *scaler = arg;
+
+	u32 val = scaler_get_int_status(scaler);
+
+	scaler_disable_int(scaler);
+
+	if (scaler->task) {
+		struct exynos_drm_ipp_task *task = scaler->task;
+
+		scaler->task = NULL;
+		pm_runtime_mark_last_busy(scaler->dev);
+		pm_runtime_put_autosuspend(scaler->dev);
+		exynos_drm_ipp_task_done(task, scaler_task_done(val));
+	}
+
+	return IRQ_HANDLED;
+}
+
+static int scaler_bind(struct device *dev, struct device *master, void *data)
+{
+	struct scaler_context *scaler = dev_get_drvdata(dev);
+	struct drm_device *drm_dev = data;
+	struct exynos_drm_ipp *ipp = &scaler->ipp;
+
+	scaler->drm_dev = drm_dev;
+	drm_iommu_attach_device(drm_dev, dev);
+
+	exynos_drm_ipp_register(drm_dev, ipp, &ipp_funcs,
+			DRM_EXYNOS_IPP_CAP_CROP | DRM_EXYNOS_IPP_CAP_ROTATE |
+			DRM_EXYNOS_IPP_CAP_SCALE | DRM_EXYNOS_IPP_CAP_CONVERT,
+			scaler->scaler_data->formats,
+			scaler->scaler_data->num_formats, "scaler");
+
+	dev_info(dev, "The exynos scaler has been probed successfully\n");
+
+	return 0;
+}
+
+static void scaler_unbind(struct device *dev, struct device *master,
+			void *data)
+{
+	struct scaler_context *scaler = dev_get_drvdata(dev);
+	struct drm_device *drm_dev = data;
+	struct exynos_drm_ipp *ipp = &scaler->ipp;
+
+	exynos_drm_ipp_unregister(drm_dev, ipp);
+	drm_iommu_detach_device(scaler->drm_dev, scaler->dev);
+}
+
+static const struct component_ops scaler_component_ops = {
+	.bind	= scaler_bind,
+	.unbind = scaler_unbind,
+};
+
+static int scaler_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct resource	*regs_res;
+	struct scaler_context *scaler;
+	int irq;
+	int ret, i;
+
+	scaler = devm_kzalloc(dev, sizeof(*scaler), GFP_KERNEL);
+	if (!scaler)
+		return -ENOMEM;
+
+	scaler->scaler_data =
+		(struct scaler_data *)of_device_get_match_data(dev);
+
+	scaler->dev = dev;
+	regs_res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	scaler->regs = devm_ioremap_resource(dev, regs_res);
+	if (IS_ERR(scaler->regs))
+		return PTR_ERR(scaler->regs);
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		dev_err(dev, "failed to get irq\n");
+		return irq;
+	}
+
+	ret = devm_request_threaded_irq(dev, irq, NULL,	scaler_irq_handler,
+					IRQF_ONESHOT, "drm_scaler", scaler);
+	if (ret < 0) {
+		dev_err(dev, "failed to request irq\n");
+		return ret;
+	}
+
+	for (i = 0; i < scaler->scaler_data->num_clk; ++i) {
+		scaler->clock[i] = devm_clk_get(dev,
+					      scaler->scaler_data->clk_name[i]);
+		if (IS_ERR(scaler->clock[i])) {
+			dev_err(dev, "failed to get clock\n");
+			return PTR_ERR(scaler->clock[i]);
+		}
+	}
+
+	pm_runtime_use_autosuspend(dev);
+	pm_runtime_set_autosuspend_delay(dev, SCALER_AUTOSUSPEND_DELAY);
+	pm_runtime_enable(dev);
+	platform_set_drvdata(pdev, scaler);
+
+	ret = component_add(dev, &scaler_component_ops);
+	if (ret)
+		goto err_ippdrv_register;
+
+	return 0;
+
+err_ippdrv_register:
+	pm_runtime_dont_use_autosuspend(dev);
+	pm_runtime_disable(dev);
+	return ret;
+}
+
+static int scaler_remove(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+
+	component_del(dev, &scaler_component_ops);
+	pm_runtime_dont_use_autosuspend(dev);
+	pm_runtime_disable(dev);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM
+
+static int clk_disable_unprepare_wrapper(struct clk *clk)
+{
+	clk_disable_unprepare(clk);
+
+	return 0;
+}
+
+static int scaler_clk_ctrl(struct scaler_context *scaler, bool enable)
+{
+	int (*clk_fun)(struct clk *clk), i;
+
+	clk_fun = enable ? clk_prepare_enable : clk_disable_unprepare_wrapper;
+
+	for (i = 0; i < scaler->scaler_data->num_clk; ++i)
+		clk_fun(scaler->clock[i]);
+
+	return 0;
+}
+
+static int scaler_runtime_suspend(struct device *dev)
+{
+	struct scaler_context *scaler = dev_get_drvdata(dev);
+
+	return  scaler_clk_ctrl(scaler, false);
+}
+
+static int scaler_runtime_resume(struct device *dev)
+{
+	struct scaler_context *scaler = dev_get_drvdata(dev);
+
+	return  scaler_clk_ctrl(scaler, true);
+}
+#endif
+
+static const struct dev_pm_ops scaler_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(pm_runtime_force_suspend,
+				pm_runtime_force_resume)
+	SET_RUNTIME_PM_OPS(scaler_runtime_suspend, scaler_runtime_resume, NULL)
+};
+
+static const struct drm_exynos_ipp_limit scaler_5420_two_pixel_hv_limits[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 16, SZ_8K }, .v = { 16, SZ_8K }) },
+	{ IPP_SIZE_LIMIT(AREA, .h.align = 2, .v.align = 2) },
+	{ IPP_SCALE_LIMIT(.h = { 65536 * 1 / 4, 65536 * 16 },
+			  .v = { 65536 * 1 / 4, 65536 * 16 }) },
+};
+
+static const struct drm_exynos_ipp_limit scaler_5420_two_pixel_h_limits[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 16, SZ_8K }, .v = { 16, SZ_8K }) },
+	{ IPP_SIZE_LIMIT(AREA, .h.align = 2, .v.align = 1) },
+	{ IPP_SCALE_LIMIT(.h = { 65536 * 1 / 4, 65536 * 16 },
+			  .v = { 65536 * 1 / 4, 65536 * 16 }) },
+};
+
+static const struct drm_exynos_ipp_limit scaler_5420_one_pixel_limits[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 16, SZ_8K }, .v = { 16, SZ_8K }) },
+	{ IPP_SCALE_LIMIT(.h = { 65536 * 1 / 4, 65536 * 16 },
+			  .v = { 65536 * 1 / 4, 65536 * 16 }) },
+};
+
+static const struct exynos_drm_ipp_formats exynos5420_formats[] = {
+	/* SCALER_YUV420_2P_UV */
+	{ IPP_SRCDST_FORMAT(NV21, scaler_5420_two_pixel_hv_limits) },
+
+	/* SCALER_YUV420_2P_VU */
+	{ IPP_SRCDST_FORMAT(NV12, scaler_5420_two_pixel_hv_limits) },
+
+	/* SCALER_YUV420_3P */
+	{ IPP_SRCDST_FORMAT(YUV420, scaler_5420_two_pixel_hv_limits) },
+
+	/* SCALER_YUV422_1P_YUYV */
+	{ IPP_SRCDST_FORMAT(YUYV, scaler_5420_two_pixel_h_limits) },
+
+	/* SCALER_YUV422_1P_UYVY */
+	{ IPP_SRCDST_FORMAT(UYVY, scaler_5420_two_pixel_h_limits) },
+
+	/* SCALER_YUV422_1P_YVYU */
+	{ IPP_SRCDST_FORMAT(YVYU, scaler_5420_two_pixel_h_limits) },
+
+	/* SCALER_YUV422_2P_UV */
+	{ IPP_SRCDST_FORMAT(NV61, scaler_5420_two_pixel_h_limits) },
+
+	/* SCALER_YUV422_2P_VU */
+	{ IPP_SRCDST_FORMAT(NV16, scaler_5420_two_pixel_h_limits) },
+
+	/* SCALER_YUV422_3P */
+	{ IPP_SRCDST_FORMAT(YUV422, scaler_5420_two_pixel_h_limits) },
+
+	/* SCALER_YUV444_2P_UV */
+	{ IPP_SRCDST_FORMAT(NV42, scaler_5420_one_pixel_limits) },
+
+	/* SCALER_YUV444_2P_VU */
+	{ IPP_SRCDST_FORMAT(NV24, scaler_5420_one_pixel_limits) },
+
+	/* SCALER_YUV444_3P */
+	{ IPP_SRCDST_FORMAT(YUV444, scaler_5420_one_pixel_limits) },
+
+	/* SCALER_RGB_565 */
+	{ IPP_SRCDST_FORMAT(RGB565, scaler_5420_one_pixel_limits) },
+
+	/* SCALER_ARGB1555 */
+	{ IPP_SRCDST_FORMAT(XRGB1555, scaler_5420_one_pixel_limits) },
+
+	/* SCALER_ARGB1555 */
+	{ IPP_SRCDST_FORMAT(ARGB1555, scaler_5420_one_pixel_limits) },
+
+	/* SCALER_ARGB4444 */
+	{ IPP_SRCDST_FORMAT(XRGB4444, scaler_5420_one_pixel_limits) },
+
+	/* SCALER_ARGB4444 */
+	{ IPP_SRCDST_FORMAT(ARGB4444, scaler_5420_one_pixel_limits) },
+
+	/* SCALER_ARGB8888 */
+	{ IPP_SRCDST_FORMAT(XRGB8888, scaler_5420_one_pixel_limits) },
+
+	/* SCALER_ARGB8888 */
+	{ IPP_SRCDST_FORMAT(ARGB8888, scaler_5420_one_pixel_limits) },
+
+	/* SCALER_RGBA8888 */
+	{ IPP_SRCDST_FORMAT(RGBX8888, scaler_5420_one_pixel_limits) },
+
+	/* SCALER_RGBA8888 */
+	{ IPP_SRCDST_FORMAT(RGBA8888, scaler_5420_one_pixel_limits) },
+};
+
+static const struct scaler_data exynos5420_data = {
+	.clk_name	= {"mscl"},
+	.num_clk	= 1,
+	.formats	= exynos5420_formats,
+	.num_formats	= ARRAY_SIZE(exynos5420_formats),
+};
+
+static const struct scaler_data exynos5433_data = {
+	.clk_name	= {"pclk", "aclk", "aclk_xiu"},
+	.num_clk	= 3,
+	.formats	= exynos5420_formats, /* intentional */
+	.num_formats	= ARRAY_SIZE(exynos5420_formats),
+};
+
+static const struct of_device_id exynos_scaler_match[] = {
+	{
+		.compatible = "samsung,exynos5420-scaler",
+		.data = &exynos5420_data,
+	}, {
+		.compatible = "samsung,exynos5433-scaler",
+		.data = &exynos5433_data,
+	}, {
+	},
+};
+MODULE_DEVICE_TABLE(of, exynos_scaler_match);
+
+struct platform_driver scaler_driver = {
+	.probe		= scaler_probe,
+	.remove		= scaler_remove,
+	.driver		= {
+		.name	= "exynos-scaler",
+		.owner	= THIS_MODULE,
+		.pm	= &scaler_pm_ops,
+		.of_match_table = exynos_scaler_match,
+	},
+};
diff --git a/drivers/gpu/drm/exynos/regs-scaler.h b/drivers/gpu/drm/exynos/regs-scaler.h
new file mode 100644
index 0000000..fc7ccad
--- /dev/null
+++ b/drivers/gpu/drm/exynos/regs-scaler.h
@@ -0,0 +1,426 @@
+/* drivers/gpu/drm/exynos/regs-scaler.h
+ *
+ * Copyright (c) 2017 Samsung Electronics Co., Ltd.
+ *		http://www.samsung.com/
+ * Author: Andrzej Pietrasiewicz <andrzej.p@samsung.com>
+ *
+ * Register definition file for Samsung scaler driver
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#ifndef EXYNOS_REGS_SCALER_H
+#define EXYNOS_REGS_SCALER_H
+
+/* Register part */
+
+/* Global setting */
+#define SCALER_STATUS			0x0	/* no shadow */
+#define SCALER_CFG			0x4
+
+/* Interrupt */
+#define SCALER_INT_EN			0x8	/* no shadow */
+#define SCALER_INT_STATUS		0xc	/* no shadow */
+
+/* SRC */
+#define SCALER_SRC_CFG			0x10
+#define SCALER_SRC_Y_BASE		0x14
+#define SCALER_SRC_CB_BASE		0x18
+#define SCALER_SRC_CR_BASE		0x294
+#define SCALER_SRC_SPAN			0x1c
+#define SCALER_SRC_Y_POS		0x20
+#define SCALER_SRC_WH			0x24
+#define SCALER_SRC_C_POS		0x28
+
+/* DST */
+#define SCALER_DST_CFG			0x30
+#define SCALER_DST_Y_BASE		0x34
+#define SCALER_DST_CB_BASE		0x38
+#define SCALER_DST_CR_BASE		0x298
+#define SCALER_DST_SPAN			0x3c
+#define SCALER_DST_WH			0x40
+#define SCALER_DST_POS			0x44
+
+/* Ratio */
+#define SCALER_H_RATIO			0x50
+#define SCALER_V_RATIO			0x54
+
+/* Rotation */
+#define SCALER_ROT_CFG			0x58
+
+/* Coefficient */
+/*
+ * YHCOEF_{x}{A|B|C|D}			CHCOEF_{x}{A|B|C|D}
+ *
+ *	A	B	C	D	A	B	C	D
+ * 0	60	64	68	6c	140	144	148	14c
+ * 1	70	74	78	7c	150	154	158	15c
+ * 2	80	84	88	8c	160	164	168	16c
+ * 3	90	94	98	9c	170	174	178	17c
+ * 4	a0	a4	a8	ac	180	184	188	18c
+ * 5	b0	b4	b8	bc	190	194	198	19c
+ * 6	c0	c4	c8	cc	1a0	1a4	1a8	1ac
+ * 7	d0	d4	d8	dc	1b0	1b4	1b8	1bc
+ * 8	e0	e4	e8	ec	1c0	1c4	1c8	1cc
+ *
+ *
+ * YVCOEF_{x}{A|B}			CVCOEF_{x}{A|B}
+ *
+ *	A	B			A	B
+ * 0	f0	f4			1d0	1d4
+ * 1	f8	fc			1d8	1dc
+ * 2	100	104			1e0	1e4
+ * 3	108	10c			1e8	1ec
+ * 4	110	114			1f0	1f4
+ * 5	118	11c			1f8	1fc
+ * 6	120	124			200	204
+ * 7	128	12c			208	20c
+ * 8	130	134			210	214
+ */
+#define _SCALER_HCOEF_DELTA(r, c)	((r) * 0x10 + (c) * 0x4)
+#define _SCALER_VCOEF_DELTA(r, c)	((r) * 0x8 + (c) * 0x4)
+
+#define SCALER_YHCOEF(r, c)		(0x60 + _SCALER_HCOEF_DELTA((r), (c)))
+#define SCALER_YVCOEF(r, c)		(0xf0 + _SCALER_VCOEF_DELTA((r), (c)))
+#define SCALER_CHCOEF(r, c)		(0x140 + _SCALER_HCOEF_DELTA((r), (c)))
+#define SCALER_CVCOEF(r, c)		(0x1d0 + _SCALER_VCOEF_DELTA((r), (c)))
+
+
+/* Color Space Conversion */
+#define SCALER_CSC_COEF(x, y)		(0x220 + (y) * 0xc + (x) * 0x4)
+
+/* Dithering */
+#define SCALER_DITH_CFG			0x250
+
+/* Version Number */
+#define SCALER_VER			0x260	/* no shadow */
+
+/* Cycle count and Timeout */
+#define SCALER_CYCLE_COUNT		0x278	/* no shadow */
+#define SCALER_TIMEOUT_CTRL		0x2c0	/* no shadow */
+#define SCALER_TIMEOUT_CNT		0x2c4	/* no shadow */
+
+/* Blending */
+#define SCALER_SRC_BLEND_COLOR		0x280
+#define SCALER_SRC_BLEND_ALPHA		0x284
+#define SCALER_DST_BLEND_COLOR		0x288
+#define SCALER_DST_BLEND_ALPHA		0x28c
+
+/* Color Fill */
+#define SCALER_FILL_COLOR		0x290
+
+/* Multiple Command Queue */
+#define SCALER_ADDR_Q_CONFIG		0x2a0	/* no shadow */
+#define SCALER_SRC_ADDR_Q_STATUS	0x2a4	/* no shadow */
+#define SCALER_SRC_ADDR_Q		0x2a8	/* no shadow */
+
+/* CRC */
+#define SCALER_CRC_COLOR00_10		0x2b0	/* no shadow */
+#define SCALER_CRC_COLOR20_30		0x2b4	/* no shadow */
+#define SCALER_CRC_COLOR01_11		0x2b8	/* no shadow */
+#define SCALER_CRC_COLOR21_31		0x2bc	/* no shadow */
+
+/* Shadow Registers */
+#define SCALER_SHADOW_OFFSET		0x1000
+
+
+/* Bit definition part */
+#define SCALER_MASK(hi_b, lo_b)		((1 << ((hi_b) - (lo_b) + 1)) - 1)
+#define SCALER_GET(reg, hi_b, lo_b)	\
+	(((reg) >> (lo_b)) & SCALER_MASK(hi_b, lo_b))
+#define SCALER_SET(val, hi_b, lo_b) \
+	(((val) & SCALER_MASK(hi_b, lo_b)) << lo_b)
+
+/* SCALER_STATUS */
+#define SCALER_STATUS_SCALER_RUNNING		(1 << 1)
+#define SCALER_STATUS_SCALER_READY_CLK_DOWN	(1 << 0)
+
+/* SCALER_CFG */
+#define SCALER_CFG_FILL_EN			(1 << 24)
+#define SCALER_CFG_BLEND_COLOR_DIVIDE_ALPHA_EN	(1 << 17)
+#define SCALER_CFG_BLEND_EN			(1 << 16)
+#define SCALER_CFG_CSC_Y_OFFSET_SRC_EN		(1 << 10)
+#define SCALER_CFG_CSC_Y_OFFSET_DST_EN		(1 << 9)
+#define SCALER_CFG_16_BURST_MODE		(1 << 8)
+#define SCALER_CFG_SOFT_RESET			(1 << 1)
+#define SCALER_CFG_START_CMD			(1 << 0)
+
+/* SCALER_INT_EN */
+#define SCALER_INT_EN_TIMEOUT			(1 << 31)
+#define SCALER_INT_EN_ILLEGAL_BLEND		(1 << 24)
+#define SCALER_INT_EN_ILLEGAL_RATIO		(1 << 23)
+#define SCALER_INT_EN_ILLEGAL_DST_HEIGHT	(1 << 22)
+#define SCALER_INT_EN_ILLEGAL_DST_WIDTH		(1 << 21)
+#define SCALER_INT_EN_ILLEGAL_DST_V_POS		(1 << 20)
+#define SCALER_INT_EN_ILLEGAL_DST_H_POS		(1 << 19)
+#define SCALER_INT_EN_ILLEGAL_DST_C_SPAN	(1 << 18)
+#define SCALER_INT_EN_ILLEGAL_DST_Y_SPAN	(1 << 17)
+#define SCALER_INT_EN_ILLEGAL_DST_CR_BASE	(1 << 16)
+#define SCALER_INT_EN_ILLEGAL_DST_CB_BASE	(1 << 15)
+#define SCALER_INT_EN_ILLEGAL_DST_Y_BASE	(1 << 14)
+#define SCALER_INT_EN_ILLEGAL_DST_COLOR		(1 << 13)
+#define SCALER_INT_EN_ILLEGAL_SRC_HEIGHT	(1 << 12)
+#define SCALER_INT_EN_ILLEGAL_SRC_WIDTH		(1 << 11)
+#define SCALER_INT_EN_ILLEGAL_SRC_CV_POS	(1 << 10)
+#define SCALER_INT_EN_ILLEGAL_SRC_CH_POS	(1 << 9)
+#define SCALER_INT_EN_ILLEGAL_SRC_YV_POS	(1 << 8)
+#define SCALER_INT_EN_ILLEGAL_SRC_YH_POS	(1 << 7)
+#define SCALER_INT_EN_ILLEGAL_DST_SPAN		(1 << 6)
+#define SCALER_INT_EN_ILLEGAL_SRC_Y_SPAN	(1 << 5)
+#define SCALER_INT_EN_ILLEGAL_SRC_CR_BASE	(1 << 4)
+#define SCALER_INT_EN_ILLEGAL_SRC_CB_BASE	(1 << 3)
+#define SCALER_INT_EN_ILLEGAL_SRC_Y_BASE	(1 << 2)
+#define SCALER_INT_EN_ILLEGAL_SRC_COLOR		(1 << 1)
+#define SCALER_INT_EN_FRAME_END			(1 << 0)
+
+/* SCALER_INT_STATUS */
+#define SCALER_INT_STATUS_TIMEOUT		(1 << 31)
+#define SCALER_INT_STATUS_ILLEGAL_BLEND		(1 << 24)
+#define SCALER_INT_STATUS_ILLEGAL_RATIO		(1 << 23)
+#define SCALER_INT_STATUS_ILLEGAL_DST_HEIGHT	(1 << 22)
+#define SCALER_INT_STATUS_ILLEGAL_DST_WIDTH	(1 << 21)
+#define SCALER_INT_STATUS_ILLEGAL_DST_V_POS	(1 << 20)
+#define SCALER_INT_STATUS_ILLEGAL_DST_H_POS	(1 << 19)
+#define SCALER_INT_STATUS_ILLEGAL_DST_C_SPAN	(1 << 18)
+#define SCALER_INT_STATUS_ILLEGAL_DST_Y_SPAN	(1 << 17)
+#define SCALER_INT_STATUS_ILLEGAL_DST_CR_BASE	(1 << 16)
+#define SCALER_INT_STATUS_ILLEGAL_DST_CB_BASE	(1 << 15)
+#define SCALER_INT_STATUS_ILLEGAL_DST_Y_BASE	(1 << 14)
+#define SCALER_INT_STATUS_ILLEGAL_DST_COLOR	(1 << 13)
+#define SCALER_INT_STATUS_ILLEGAL_SRC_HEIGHT	(1 << 12)
+#define SCALER_INT_STATUS_ILLEGAL_SRC_WIDTH	(1 << 11)
+#define SCALER_INT_STATUS_ILLEGAL_SRC_CV_POS	(1 << 10)
+#define SCALER_INT_STATUS_ILLEGAL_SRC_CH_POS	(1 << 9)
+#define SCALER_INT_STATUS_ILLEGAL_SRC_YV_POS	(1 << 8)
+#define SCALER_INT_STATUS_ILLEGAL_SRC_YH_POS	(1 << 7)
+#define SCALER_INT_STATUS_ILLEGAL_DST_SPAN	(1 << 6)
+#define SCALER_INT_STATUS_ILLEGAL_SRC_Y_SPAN	(1 << 5)
+#define SCALER_INT_STATUS_ILLEGAL_SRC_CR_BASE	(1 << 4)
+#define SCALER_INT_STATUS_ILLEGAL_SRC_CB_BASE	(1 << 3)
+#define SCALER_INT_STATUS_ILLEGAL_SRC_Y_BASE	(1 << 2)
+#define SCALER_INT_STATUS_ILLEGAL_SRC_COLOR	(1 << 1)
+#define SCALER_INT_STATUS_FRAME_END		(1 << 0)
+
+/* SCALER_SRC_CFG */
+#define SCALER_SRC_CFG_TILE_EN			(1 << 10)
+#define SCALER_SRC_CFG_GET_BYTE_SWAP(r)		SCALER_GET(r, 6, 5)
+#define SCALER_SRC_CFG_SET_BYTE_SWAP(v)		SCALER_SET(v, 6, 5)
+#define SCALER_SRC_CFG_GET_COLOR_FORMAT(r)	SCALER_GET(r, 4, 0)
+#define SCALER_SRC_CFG_SET_COLOR_FORMAT(v)	SCALER_SET(v, 4, 0)
+#define SCALER_YUV420_2P_UV			0
+#define SCALER_YUV422_2P_UV			2
+#define SCALER_YUV444_2P_UV			3
+#define SCALER_RGB_565				4
+#define SCALER_ARGB1555				5
+#define SCALER_ARGB8888				6
+#define SCALER_ARGB8888_PRE			7
+#define SCALER_YUV422_1P_YVYU			9
+#define SCALER_YUV422_1P_YUYV			10
+#define SCALER_YUV422_1P_UYVY			11
+#define SCALER_ARGB4444				12
+#define SCALER_L8A8				13
+#define SCALER_RGBA8888				14
+#define SCALER_L8				15
+#define SCALER_YUV420_2P_VU			16
+#define SCALER_YUV422_2P_VU			18
+#define SCALER_YUV444_2P_VU			19
+#define SCALER_YUV420_3P			20
+#define SCALER_YUV422_3P			22
+#define SCALER_YUV444_3P			23
+
+/* SCALER_SRC_SPAN */
+#define SCALER_SRC_SPAN_GET_C_SPAN(r)		SCALER_GET(r, 29, 16)
+#define SCALER_SRC_SPAN_SET_C_SPAN(v)		SCALER_SET(v, 29, 16)
+#define SCALER_SRC_SPAN_GET_Y_SPAN(r)		SCALER_GET(r, 13, 0)
+#define SCALER_SRC_SPAN_SET_Y_SPAN(v)		SCALER_SET(v, 13, 0)
+
+/* SCALER_SRC_Y_POS */
+#define SCALER_SRC_Y_POS_GET_YH_POS(r)		SCALER_GET(r, 31, 16)
+#define SCALER_SRC_Y_POS_SET_YH_POS(v)		SCALER_SET(v, 31, 16)
+#define SCALER_SRC_Y_POS_GET_YV_POS(r)		SCALER_GET(r, 15, 0)
+#define SCALER_SRC_Y_POS_SET_YV_POS(v)		SCALER_SET(v, 15, 0)
+
+/* SCALER_SRC_WH */
+#define SCALER_SRC_WH_GET_WIDTH(r)		SCALER_GET(r, 29, 16)
+#define SCALER_SRC_WH_SET_WIDTH(v)		SCALER_SET(v, 29, 16)
+#define SCALER_SRC_WH_GET_HEIGHT(r)		SCALER_GET(r, 13, 0)
+#define SCALER_SRC_WH_SET_HEIGHT(v)		SCALER_SET(v, 13, 0)
+
+/* SCALER_SRC_C_POS */
+#define SCALER_SRC_C_POS_GET_CH_POS(r)		SCALER_GET(r, 31, 16)
+#define SCALER_SRC_C_POS_SET_CH_POS(v)		SCALER_SET(v, 31, 16)
+#define SCALER_SRC_C_POS_GET_CV_POS(r)		SCALER_GET(r, 15, 0)
+#define SCALER_SRC_C_POS_SET_CV_POS(v)		SCALER_SET(v, 15, 0)
+
+/* SCALER_DST_CFG */
+#define SCALER_DST_CFG_GET_BYTE_SWAP(r)		SCALER_GET(r, 6, 5)
+#define SCALER_DST_CFG_SET_BYTE_SWAP(v)		SCALER_SET(v, 6, 5)
+#define SCALER_DST_CFG_GET_COLOR_FORMAT(r)	SCALER_GET(r, 4, 0)
+#define SCALER_DST_CFG_SET_COLOR_FORMAT(v)	SCALER_SET(v, 4, 0)
+
+/* SCALER_DST_SPAN */
+#define SCALER_DST_SPAN_GET_C_SPAN(r)		SCALER_GET(r, 29, 16)
+#define SCALER_DST_SPAN_SET_C_SPAN(v)		SCALER_SET(v, 29, 16)
+#define SCALER_DST_SPAN_GET_Y_SPAN(r)		SCALER_GET(r, 13, 0)
+#define SCALER_DST_SPAN_SET_Y_SPAN(v)		SCALER_SET(v, 13, 0)
+
+/* SCALER_DST_WH */
+#define SCALER_DST_WH_GET_WIDTH(r)		SCALER_GET(r, 29, 16)
+#define SCALER_DST_WH_SET_WIDTH(v)		SCALER_SET(v, 29, 16)
+#define SCALER_DST_WH_GET_HEIGHT(r)		SCALER_GET(r, 13, 0)
+#define SCALER_DST_WH_SET_HEIGHT(v)		SCALER_SET(v, 13, 0)
+
+/* SCALER_DST_POS */
+#define SCALER_DST_POS_GET_H_POS(r)		SCALER_GET(r, 29, 16)
+#define SCALER_DST_POS_SET_H_POS(v)		SCALER_SET(v, 29, 16)
+#define SCALER_DST_POS_GET_V_POS(r)		SCALER_GET(r, 13, 0)
+#define SCALER_DST_POS_SET_V_POS(v)		SCALER_SET(v, 13, 0)
+
+/* SCALER_H_RATIO */
+#define SCALER_H_RATIO_GET(r)			SCALER_GET(r, 18, 0)
+#define SCALER_H_RATIO_SET(v)			SCALER_SET(v, 18, 0)
+
+/* SCALER_V_RATIO */
+#define SCALER_V_RATIO_GET(r)			SCALER_GET(r, 18, 0)
+#define SCALER_V_RATIO_SET(v)			SCALER_SET(v, 18, 0)
+
+/* SCALER_ROT_CFG */
+#define SCALER_ROT_CFG_FLIP_X_EN		(1 << 3)
+#define SCALER_ROT_CFG_FLIP_Y_EN		(1 << 2)
+#define SCALER_ROT_CFG_GET_ROTMODE(r)		SCALER_GET(r, 1, 0)
+#define SCALER_ROT_CFG_SET_ROTMODE(v)		SCALER_SET(v, 1, 0)
+#define SCALER_ROT_MODE_90			1
+#define SCALER_ROT_MODE_180			2
+#define SCALER_ROT_MODE_270			3
+
+/* SCALER_HCOEF, SCALER_VCOEF */
+#define SCALER_COEF_SHIFT(i)			(16 * (1 - (i) % 2))
+#define SCALER_COEF_GET(r, i)			\
+				(((r) >> SCALER_COEF_SHIFT(i)) & 0x1ff)
+#define SCALER_COEF_SET(v, i)			\
+				(((v) & 0x1ff) << SCALER_COEF_SHIFT(i))
+
+/* SCALER_CSC_COEFxy */
+#define SCALER_CSC_COEF_GET(r)			SCALER_GET(r, 11, 0)
+#define SCALER_CSC_COEF_SET(v)			SCALER_SET(v, 11, 0)
+
+/* SCALER_DITH_CFG */
+#define SCALER_DITH_CFG_GET_R_TYPE(r)		SCALER_GET(r, 8, 6)
+#define SCALER_DITH_CFG_SET_R_TYPE(v)		SCALER_SET(v, 8, 6)
+#define SCALER_DITH_CFG_GET_G_TYPE(r)		SCALER_GET(r, 5, 3)
+#define SCALER_DITH_CFG_SET_G_TYPE(v)		SCALER_SET(v, 5, 3)
+#define SCALER_DITH_CFG_GET_B_TYPE(r)		SCALER_GET(r, 2, 0)
+#define SCALER_DITH_CFG_SET_B_TYPE(v)		SCALER_SET(v, 2, 0)
+
+/* SCALER_TIMEOUT_CTRL */
+#define SCALER_TIMEOUT_CTRL_GET_TIMER_VALUE(r)	SCALER_GET(r, 31, 16)
+#define SCALER_TIMEOUT_CTRL_SET_TIMER_VALUE(v)	SCALER_SET(v, 31, 16)
+#define SCALER_TIMEOUT_CTRL_GET_TIMER_DIV(r)	SCALER_GET(r, 7, 4)
+#define SCALER_TIMEOUT_CTRL_SET_TIMER_DIV(v)	SCALER_SET(v, 7, 4)
+#define SCALER_TIMEOUT_CTRL_TIMER_ENABLE	(1 << 0)
+
+/* SCALER_TIMEOUT_CNT */
+#define SCALER_TIMEOUT_CTRL_GET_TIMER_COUNT(r)	SCALER_GET(r, 31, 16)
+
+/* SCALER_SRC_BLEND_COLOR */
+#define SCALER_SRC_BLEND_COLOR_SEL_INV		(1 << 31)
+#define SCALER_SRC_BLEND_COLOR_GET_SEL(r)	SCALER_GET(r, 30, 29)
+#define SCALER_SRC_BLEND_COLOR_SET_SEL(v)	SCALER_SET(v, 30, 29)
+#define SCALER_SRC_BLEND_COLOR_OP_SEL_INV	(1 << 28)
+#define SCALER_SRC_BLEND_COLOR_GET_OP_SEL(r)	SCALER_GET(r, 27, 24)
+#define SCALER_SRC_BLEND_COLOR_SET_OP_SEL(v)	SCALER_SET(v, 27, 24)
+#define SCALER_SRC_BLEND_COLOR_GET_COLOR0(r)	SCALER_GET(r, 23, 16)
+#define SCALER_SRC_BLEND_COLOR_SET_COLOR0(v)	SCALER_SET(v, 23, 16)
+#define SCALER_SRC_BLEND_COLOR_GET_COLOR1(r)	SCALER_GET(r, 15, 8)
+#define SCALER_SRC_BLEND_COLOR_SET_COLOR1(v)	SCALER_SET(v, 15, 8)
+#define SCALER_SRC_BLEND_COLOR_GET_COLOR2(r)	SCALER_GET(r, 7, 0)
+#define SCALER_SRC_BLEND_COLOR_SET_COLOR2(v)	SCALER_SET(v, 7, 0)
+
+/* SCALER_SRC_BLEND_ALPHA */
+#define SCALER_SRC_BLEND_ALPHA_SEL_INV		(1 << 31)
+#define SCALER_SRC_BLEND_ALPHA_GET_SEL(r)	SCALER_GET(r, 30, 29)
+#define SCALER_SRC_BLEND_ALPHA_SET_SEL(v)	SCALER_SET(v, 30, 29)
+#define SCALER_SRC_BLEND_ALPHA_OP_SEL_INV	(1 << 28)
+#define SCALER_SRC_BLEND_ALPHA_GET_OP_SEL(r)	SCALER_GET(r, 27, 24)
+#define SCALER_SRC_BLEND_ALPHA_SET_OP_SEL(v)	SCALER_SET(v, 27, 24)
+#define SCALER_SRC_BLEND_ALPHA_GET_ALPHA(r)	SCALER_GET(r, 7, 0)
+#define SCALER_SRC_BLEND_ALPHA_SET_ALPHA(v)	SCALER_SET(v, 7, 0)
+
+/* SCALER_DST_BLEND_COLOR */
+#define SCALER_DST_BLEND_COLOR_SEL_INV		(1 << 31)
+#define SCALER_DST_BLEND_COLOR_GET_SEL(r)	SCALER_GET(r, 30, 29)
+#define SCALER_DST_BLEND_COLOR_SET_SEL(v)	SCALER_SET(v, 30, 29)
+#define SCALER_DST_BLEND_COLOR_OP_SEL_INV	(1 << 28)
+#define SCALER_DST_BLEND_COLOR_GET_OP_SEL(r)	SCALER_GET(r, 27, 24)
+#define SCALER_DST_BLEND_COLOR_SET_OP_SEL(v)	SCALER_SET(v, 27, 24)
+#define SCALER_DST_BLEND_COLOR_GET_COLOR0(r)	SCALER_GET(r, 23, 16)
+#define SCALER_DST_BLEND_COLOR_SET_COLOR0(v)	SCALER_SET(v, 23, 16)
+#define SCALER_DST_BLEND_COLOR_GET_COLOR1(r)	SCALER_GET(r, 15, 8)
+#define SCALER_DST_BLEND_COLOR_SET_COLOR1(v)	SCALER_SET(v, 15, 8)
+#define SCALER_DST_BLEND_COLOR_GET_COLOR2(r)	SCALER_GET(r, 7, 0)
+#define SCALER_DST_BLEND_COLOR_SET_COLOR2(v)	SCALER_SET(v, 7, 0)
+
+/* SCALER_DST_BLEND_ALPHA */
+#define SCALER_DST_BLEND_ALPHA_SEL_INV		(1 << 31)
+#define SCALER_DST_BLEND_ALPHA_GET_SEL(r)	SCALER_GET(r, 30, 29)
+#define SCALER_DST_BLEND_ALPHA_SET_SEL(v)	SCALER_SET(v, 30, 29)
+#define SCALER_DST_BLEND_ALPHA_OP_SEL_INV	(1 << 28)
+#define SCALER_DST_BLEND_ALPHA_GET_OP_SEL(r)	SCALER_GET(r, 27, 24)
+#define SCALER_DST_BLEND_ALPHA_SET_OP_SEL(v)	SCALER_SET(v, 27, 24)
+#define SCALER_DST_BLEND_ALPHA_GET_ALPHA(r)	SCALER_GET(r, 7, 0)
+#define SCALER_DST_BLEND_ALPHA_SET_ALPHA(v)	SCALER_SET(v, 7, 0)
+
+/* SCALER_FILL_COLOR */
+#define SCALER_FILL_COLOR_GET_ALPHA(r)		SCALER_GET(r, 31, 24)
+#define SCALER_FILL_COLOR_SET_ALPHA(v)		SCALER_SET(v, 31, 24)
+#define SCALER_FILL_COLOR_GET_FILL_COLOR0(r)	SCALER_GET(r, 23, 16)
+#define SCALER_FILL_COLOR_SET_FILL_COLOR0(v)	SCALER_SET(v, 23, 16)
+#define SCALER_FILL_COLOR_GET_FILL_COLOR1(r)	SCALER_GET(r, 15, 8)
+#define SCALER_FILL_COLOR_SET_FILL_COLOR1(v)	SCALER_SET(v, 15, 8)
+#define SCALER_FILL_COLOR_GET_FILL_COLOR2(r)	SCALER_GET(r, 7, 0)
+#define SCALER_FILL_COLOR_SET_FILL_COLOR2(v)	SCALER_SET(v, 7, 0)
+
+/* SCALER_ADDR_Q_CONFIG */
+#define SCALER_ADDR_Q_CONFIG_RST		(1 << 0)
+
+/* SCALER_SRC_ADDR_Q_STATUS */
+#define SCALER_SRC_ADDR_Q_STATUS_Y_FULL		(1 << 23)
+#define SCALER_SRC_ADDR_Q_STATUS_Y_EMPTY	(1 << 22)
+#define SCALER_SRC_ADDR_Q_STATUS_GET_Y_WR_IDX(r)	SCALER_GET(r, 21, 16)
+#define SCALER_SRC_ADDR_Q_STATUS_CB_FULL	(1 << 15)
+#define SCALER_SRC_ADDR_Q_STATUS_CB_EMPTY	(1 << 14)
+#define SCALER_SRC_ADDR_Q_STATUS_GET_CB_WR_IDX(r)	SCALER_GET(r, 13, 8)
+#define SCALER_SRC_ADDR_Q_STATUS_CR_FULL	(1 << 7)
+#define SCALER_SRC_ADDR_Q_STATUS_CR_EMPTY	(1 << 6)
+#define SCALER_SRC_ADDR_Q_STATUS_GET_CR_WR_IDX(r)	SCALER_GET(r, 5, 0)
+
+/* SCALER_DST_ADDR_Q_STATUS */
+#define SCALER_DST_ADDR_Q_STATUS_Y_FULL		(1 << 23)
+#define SCALER_DST_ADDR_Q_STATUS_Y_EMPTY	(1 << 22)
+#define SCALER_DST_ADDR_Q_STATUS_GET_Y_WR_IDX(r)	SCALER_GET(r, 21, 16)
+#define SCALER_DST_ADDR_Q_STATUS_CB_FULL	(1 << 15)
+#define SCALER_DST_ADDR_Q_STATUS_CB_EMPTY	(1 << 14)
+#define SCALER_DST_ADDR_Q_STATUS_GET_CB_WR_IDX(r)	SCALER_GET(r, 13, 8)
+#define SCALER_DST_ADDR_Q_STATUS_CR_FULL	(1 << 7)
+#define SCALER_DST_ADDR_Q_STATUS_CR_EMPTY	(1 << 6)
+#define SCALER_DST_ADDR_Q_STATUS_GET_CR_WR_IDX(r)	SCALER_GET(r, 5, 0)
+
+/* SCALER_CRC_COLOR00_10 */
+#define SCALER_CRC_COLOR00_10_GET_00(r)		SCALER_GET(r, 31, 16)
+#define SCALER_CRC_COLOR00_10_GET_10(r)		SCALER_GET(r, 15, 0)
+
+/* SCALER_CRC_COLOR20_30 */
+#define SCALER_CRC_COLOR20_30_GET_20(r)		SCALER_GET(r, 31, 16)
+#define SCALER_CRC_COLOR20_30_GET_30(r)		SCALER_GET(r, 15, 0)
+
+/* SCALER_CRC_COLOR01_11 */
+#define SCALER_CRC_COLOR01_11_GET_01(r)		SCALER_GET(r, 31, 16)
+#define SCALER_CRC_COLOR01_11_GET_11(r)		SCALER_GET(r, 15, 0)
+
+/* SCALER_CRC_COLOR21_31 */
+#define SCALER_CRC_COLOR21_31_GET_21(r)		SCALER_GET(r, 31, 16)
+#define SCALER_CRC_COLOR21_31_GET_31(r)		SCALER_GET(r, 15, 0)
+
+#endif /* EXYNOS_REGS_SCALER_H */
-- 
cgit v1.1


From dec60f3a9b7251f2657d743d96ba9a83dca02351 Mon Sep 17 00:00:00 2001
From: Mathieu Malaterre <malat@debian.org>
Date: Sat, 5 May 2018 21:54:05 +0200
Subject: agp: uninorth: make two functions static
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Both ‘uninorth_remove_memory’ and ‘null_cache_flush’ can be made
static. So make them.

Silence the following gcc warning (W=1):

  drivers/char/agp/uninorth-agp.c:198:5: warning: no previous prototype for ‘uninorth_remove_memory’ [-Wmissing-prototypes]

and

  drivers/char/agp/uninorth-agp.c:473:6: warning: no previous prototype for ‘null_cache_flush’ [-Wmissing-prototypes]

Signed-off-by: Mathieu Malaterre <malat@debian.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>
---
 drivers/char/agp/uninorth-agp.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/char/agp/uninorth-agp.c b/drivers/char/agp/uninorth-agp.c
index c381c8e..79d8c84 100644
--- a/drivers/char/agp/uninorth-agp.c
+++ b/drivers/char/agp/uninorth-agp.c
@@ -195,7 +195,7 @@ static int uninorth_insert_memory(struct agp_memory *mem, off_t pg_start, int ty
 	return 0;
 }
 
-int uninorth_remove_memory(struct agp_memory *mem, off_t pg_start, int type)
+static int uninorth_remove_memory(struct agp_memory *mem, off_t pg_start, int type)
 {
 	size_t i;
 	u32 *gp;
@@ -470,7 +470,7 @@ static int uninorth_free_gatt_table(struct agp_bridge_data *bridge)
 	return 0;
 }
 
-void null_cache_flush(void)
+static void null_cache_flush(void)
 {
 	mb();
 }
-- 
cgit v1.1


From 0d5a03c3d9254813ca76d7886ff9ed76a0aea545 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/ttm: don't dereference nvbo::cli, it can outlive client

Potentially responsible for some random OOPSes.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Cc: stable@vger.kernel.org [v4.15+]
---
 drivers/gpu/drm/nouveau/nouveau_bo.c  | 1 -
 drivers/gpu/drm/nouveau/nouveau_bo.h  | 2 --
 drivers/gpu/drm/nouveau/nouveau_ttm.c | 6 +++---
 3 files changed, 3 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.c b/drivers/gpu/drm/nouveau/nouveau_bo.c
index 6f402c4..ab61c03 100644
--- a/drivers/gpu/drm/nouveau/nouveau_bo.c
+++ b/drivers/gpu/drm/nouveau/nouveau_bo.c
@@ -214,7 +214,6 @@ nouveau_bo_new(struct nouveau_cli *cli, u64 size, int align,
 	INIT_LIST_HEAD(&nvbo->entry);
 	INIT_LIST_HEAD(&nvbo->vma_list);
 	nvbo->bo.bdev = &drm->ttm.bdev;
-	nvbo->cli = cli;
 
 	/* This is confusing, and doesn't actually mean we want an uncached
 	 * mapping, but is what NOUVEAU_GEM_DOMAIN_COHERENT gets translated
diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.h b/drivers/gpu/drm/nouveau/nouveau_bo.h
index be8e00b..73c4844 100644
--- a/drivers/gpu/drm/nouveau/nouveau_bo.h
+++ b/drivers/gpu/drm/nouveau/nouveau_bo.h
@@ -26,8 +26,6 @@ struct nouveau_bo {
 
 	struct list_head vma_list;
 
-	struct nouveau_cli *cli;
-
 	unsigned contig:1;
 	unsigned page:5;
 	unsigned kind:8;
diff --git a/drivers/gpu/drm/nouveau/nouveau_ttm.c b/drivers/gpu/drm/nouveau/nouveau_ttm.c
index dff51a0..8c093ca 100644
--- a/drivers/gpu/drm/nouveau/nouveau_ttm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_ttm.c
@@ -63,7 +63,7 @@ nouveau_vram_manager_new(struct ttm_mem_type_manager *man,
 			 struct ttm_mem_reg *reg)
 {
 	struct nouveau_bo *nvbo = nouveau_bo(bo);
-	struct nouveau_drm *drm = nvbo->cli->drm;
+	struct nouveau_drm *drm = nouveau_bdev(bo->bdev);
 	struct nouveau_mem *mem;
 	int ret;
 
@@ -103,7 +103,7 @@ nouveau_gart_manager_new(struct ttm_mem_type_manager *man,
 			 struct ttm_mem_reg *reg)
 {
 	struct nouveau_bo *nvbo = nouveau_bo(bo);
-	struct nouveau_drm *drm = nvbo->cli->drm;
+	struct nouveau_drm *drm = nouveau_bdev(bo->bdev);
 	struct nouveau_mem *mem;
 	int ret;
 
@@ -131,7 +131,7 @@ nv04_gart_manager_new(struct ttm_mem_type_manager *man,
 		      struct ttm_mem_reg *reg)
 {
 	struct nouveau_bo *nvbo = nouveau_bo(bo);
-	struct nouveau_drm *drm = nvbo->cli->drm;
+	struct nouveau_drm *drm = nouveau_bdev(bo->bdev);
 	struct nouveau_mem *mem;
 	int ret;
 
-- 
cgit v1.1


From 352672db857290ab5b0e2b6a99c414f92bee024c Mon Sep 17 00:00:00 2001
From: Lyude Paul <lyude@redhat.com>
Date: Wed, 2 May 2018 19:38:48 -0400
Subject: drm/nouveau: Fix deadlock in nv50_mstm_register_connector()

Currently; we're grabbing all of the modesetting locks before adding MST
connectors to fbdev. This isn't actually necessary, and causes a
deadlock as well:

======================================================
WARNING: possible circular locking dependency detected
4.17.0-rc3Lyude-Test+ #1 Tainted: G           O
------------------------------------------------------
kworker/1:0/18 is trying to acquire lock:
00000000c832f62d (&helper->lock){+.+.}, at: drm_fb_helper_add_one_connector+0x2a/0x60 [drm_kms_helper]

but task is already holding lock:
00000000942e28e2 (crtc_ww_class_mutex){+.+.}, at: drm_modeset_backoff+0x8e/0x1c0 [drm]

which lock already depends on the new lock.

the existing dependency chain (in reverse order) is:

-> #3 (crtc_ww_class_mutex){+.+.}:
       ww_mutex_lock+0x43/0x80
       drm_modeset_lock+0x71/0x130 [drm]
       drm_helper_probe_single_connector_modes+0x7d/0x6b0 [drm_kms_helper]
       drm_setup_crtcs+0x15e/0xc90 [drm_kms_helper]
       __drm_fb_helper_initial_config_and_unlock+0x29/0x480 [drm_kms_helper]
       nouveau_fbcon_init+0x138/0x1a0 [nouveau]
       nouveau_drm_load+0x173/0x7e0 [nouveau]
       drm_dev_register+0x134/0x1c0 [drm]
       drm_get_pci_dev+0x8e/0x160 [drm]
       nouveau_drm_probe+0x1a9/0x230 [nouveau]
       pci_device_probe+0xcd/0x150
       driver_probe_device+0x30b/0x480
       __driver_attach+0xbc/0xe0
       bus_for_each_dev+0x67/0x90
       bus_add_driver+0x164/0x260
       driver_register+0x57/0xc0
       do_one_initcall+0x4d/0x323
       do_init_module+0x5b/0x1f8
       load_module+0x20e5/0x2ac0
       __do_sys_finit_module+0xb7/0xd0
       do_syscall_64+0x60/0x1b0
       entry_SYSCALL_64_after_hwframe+0x49/0xbe

-> #2 (crtc_ww_class_acquire){+.+.}:
       drm_helper_probe_single_connector_modes+0x58/0x6b0 [drm_kms_helper]
       drm_setup_crtcs+0x15e/0xc90 [drm_kms_helper]
       __drm_fb_helper_initial_config_and_unlock+0x29/0x480 [drm_kms_helper]
       nouveau_fbcon_init+0x138/0x1a0 [nouveau]
       nouveau_drm_load+0x173/0x7e0 [nouveau]
       drm_dev_register+0x134/0x1c0 [drm]
       drm_get_pci_dev+0x8e/0x160 [drm]
       nouveau_drm_probe+0x1a9/0x230 [nouveau]
       pci_device_probe+0xcd/0x150
       driver_probe_device+0x30b/0x480
       __driver_attach+0xbc/0xe0
       bus_for_each_dev+0x67/0x90
       bus_add_driver+0x164/0x260
       driver_register+0x57/0xc0
       do_one_initcall+0x4d/0x323
       do_init_module+0x5b/0x1f8
       load_module+0x20e5/0x2ac0
       __do_sys_finit_module+0xb7/0xd0
       do_syscall_64+0x60/0x1b0
       entry_SYSCALL_64_after_hwframe+0x49/0xbe

-> #1 (&dev->mode_config.mutex){+.+.}:
       drm_setup_crtcs+0x10c/0xc90 [drm_kms_helper]
       __drm_fb_helper_initial_config_and_unlock+0x29/0x480 [drm_kms_helper]
       nouveau_fbcon_init+0x138/0x1a0 [nouveau]
       nouveau_drm_load+0x173/0x7e0 [nouveau]
       drm_dev_register+0x134/0x1c0 [drm]
       drm_get_pci_dev+0x8e/0x160 [drm]
       nouveau_drm_probe+0x1a9/0x230 [nouveau]
       pci_device_probe+0xcd/0x150
       driver_probe_device+0x30b/0x480
       __driver_attach+0xbc/0xe0
       bus_for_each_dev+0x67/0x90
       bus_add_driver+0x164/0x260
       driver_register+0x57/0xc0
       do_one_initcall+0x4d/0x323
       do_init_module+0x5b/0x1f8
       load_module+0x20e5/0x2ac0
       __do_sys_finit_module+0xb7/0xd0
       do_syscall_64+0x60/0x1b0
       entry_SYSCALL_64_after_hwframe+0x49/0xbe

-> #0 (&helper->lock){+.+.}:
       __mutex_lock+0x70/0x9d0
       drm_fb_helper_add_one_connector+0x2a/0x60 [drm_kms_helper]
       nv50_mstm_register_connector+0x2c/0x50 [nouveau]
       drm_dp_add_port+0x2f5/0x420 [drm_kms_helper]
       drm_dp_send_link_address+0x155/0x1e0 [drm_kms_helper]
       drm_dp_add_port+0x33f/0x420 [drm_kms_helper]
       drm_dp_send_link_address+0x155/0x1e0 [drm_kms_helper]
       drm_dp_check_and_send_link_address+0x87/0xd0 [drm_kms_helper]
       drm_dp_mst_link_probe_work+0x4d/0x80 [drm_kms_helper]
       process_one_work+0x20d/0x650
       worker_thread+0x3a/0x390
       kthread+0x11e/0x140
       ret_from_fork+0x3a/0x50

other info that might help us debug this:
Chain exists of:
  &helper->lock --> crtc_ww_class_acquire --> crtc_ww_class_mutex
 Possible unsafe locking scenario:
       CPU0                    CPU1
       ----                    ----
  lock(crtc_ww_class_mutex);
                               lock(crtc_ww_class_acquire);
                               lock(crtc_ww_class_mutex);
  lock(&helper->lock);

 *** DEADLOCK ***
5 locks held by kworker/1:0/18:
 #0: 000000004a05cd50 ((wq_completion)"events_long"){+.+.}, at: process_one_work+0x187/0x650
 #1: 00000000601c11d1 ((work_completion)(&mgr->work)){+.+.}, at: process_one_work+0x187/0x650
 #2: 00000000586ca0df (&dev->mode_config.mutex){+.+.}, at: drm_modeset_lock_all+0x3a/0x1b0 [drm]
 #3: 00000000d3ca0ffa (crtc_ww_class_acquire){+.+.}, at: drm_modeset_lock_all+0x44/0x1b0 [drm]
 #4: 00000000942e28e2 (crtc_ww_class_mutex){+.+.}, at: drm_modeset_backoff+0x8e/0x1c0 [drm]

stack backtrace:
CPU: 1 PID: 18 Comm: kworker/1:0 Tainted: G           O      4.17.0-rc3Lyude-Test+ #1
Hardware name: Gateway FX6840/FX6840, BIOS P01-A3         05/17/2010
Workqueue: events_long drm_dp_mst_link_probe_work [drm_kms_helper]
Call Trace:
 dump_stack+0x85/0xcb
 print_circular_bug.isra.38+0x1ce/0x1db
 __lock_acquire+0x128f/0x1350
 ? lock_acquire+0x9f/0x200
 ? lock_acquire+0x9f/0x200
 ? __ww_mutex_lock.constprop.13+0x8f/0x1000
 lock_acquire+0x9f/0x200
 ? drm_fb_helper_add_one_connector+0x2a/0x60 [drm_kms_helper]
 ? drm_fb_helper_add_one_connector+0x2a/0x60 [drm_kms_helper]
 __mutex_lock+0x70/0x9d0
 ? drm_fb_helper_add_one_connector+0x2a/0x60 [drm_kms_helper]
 ? ww_mutex_lock+0x43/0x80
 ? _cond_resched+0x15/0x30
 ? ww_mutex_lock+0x43/0x80
 ? drm_modeset_lock+0xb2/0x130 [drm]
 ? drm_fb_helper_add_one_connector+0x2a/0x60 [drm_kms_helper]
 drm_fb_helper_add_one_connector+0x2a/0x60 [drm_kms_helper]
 nv50_mstm_register_connector+0x2c/0x50 [nouveau]
 drm_dp_add_port+0x2f5/0x420 [drm_kms_helper]
 ? mark_held_locks+0x50/0x80
 ? kfree+0xcf/0x2a0
 ? drm_dp_check_mstb_guid+0xd6/0x120 [drm_kms_helper]
 ? trace_hardirqs_on_caller+0xed/0x180
 ? drm_dp_check_mstb_guid+0xd6/0x120 [drm_kms_helper]
 drm_dp_send_link_address+0x155/0x1e0 [drm_kms_helper]
 drm_dp_add_port+0x33f/0x420 [drm_kms_helper]
 ? nouveau_connector_aux_xfer+0x7c/0xb0 [nouveau]
 ? find_held_lock+0x2d/0x90
 ? drm_dp_dpcd_access+0xd9/0xf0 [drm_kms_helper]
 ? __mutex_unlock_slowpath+0x3b/0x280
 ? drm_dp_dpcd_access+0xd9/0xf0 [drm_kms_helper]
 drm_dp_send_link_address+0x155/0x1e0 [drm_kms_helper]
 drm_dp_check_and_send_link_address+0x87/0xd0 [drm_kms_helper]
 drm_dp_mst_link_probe_work+0x4d/0x80 [drm_kms_helper]
 process_one_work+0x20d/0x650
 worker_thread+0x3a/0x390
 ? process_one_work+0x650/0x650
 kthread+0x11e/0x140
 ? kthread_create_worker_on_cpu+0x50/0x50
 ret_from_fork+0x3a/0x50

Taking example from i915, the only time we need to hold any modesetting
locks is when changing the port on the mstc, and in that case we only
need to hold the connection mutex.

Signed-off-by: Lyude Paul <lyude@redhat.com>
Cc: Karol Herbst <kherbst@redhat.com>
Cc: stable@vger.kernel.org

Signed-off-by: Lyude Paul <lyude@redhat.com>

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nv50_display.c | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nv50_display.c b/drivers/gpu/drm/nouveau/nv50_display.c
index 8bd739c..2b3ccd8 100644
--- a/drivers/gpu/drm/nouveau/nv50_display.c
+++ b/drivers/gpu/drm/nouveau/nv50_display.c
@@ -3264,10 +3264,11 @@ nv50_mstm_destroy_connector(struct drm_dp_mst_topology_mgr *mgr,
 
 	drm_connector_unregister(&mstc->connector);
 
-	drm_modeset_lock_all(drm->dev);
 	drm_fb_helper_remove_one_connector(&drm->fbcon->helper, &mstc->connector);
+
+	drm_modeset_lock(&drm->dev->mode_config.connection_mutex, NULL);
 	mstc->port = NULL;
-	drm_modeset_unlock_all(drm->dev);
+	drm_modeset_unlock(&drm->dev->mode_config.connection_mutex);
 
 	drm_connector_unreference(&mstc->connector);
 }
@@ -3277,9 +3278,7 @@ nv50_mstm_register_connector(struct drm_connector *connector)
 {
 	struct nouveau_drm *drm = nouveau_drm(connector->dev);
 
-	drm_modeset_lock_all(drm->dev);
 	drm_fb_helper_add_one_connector(&drm->fbcon->helper, connector);
-	drm_modeset_unlock_all(drm->dev);
 
 	drm_connector_register(connector);
 }
-- 
cgit v1.1


From 3057fcef385348fe85173f1b0c824d89f1176f72 Mon Sep 17 00:00:00 2001
From: Boris Brezillon <boris.brezillon@bootlin.com>
Date: Fri, 4 May 2018 21:24:31 +0200
Subject: mtd: rawnand: Make sure we wait tWB before polling the STATUS reg

NAND chips require a bit of time to take the NAND operation into
account and set the BUSY bit in the STATUS reg. Make sure we don't poll
the STATUS reg too early in nand_soft_waitrdy().

Fixes: 8878b126df76 ("mtd: nand: add ->exec_op() implementation")
Cc: <stable@vger.kernel.org>
Signed-off-by: Boris Brezillon <boris.brezillon@bootlin.com>
Acked-by: Miquel Raynal <miquel.raynal@bootlin.com>
---
 drivers/mtd/nand/raw/nand_base.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/mtd/nand/raw/nand_base.c b/drivers/mtd/nand/raw/nand_base.c
index 72f3a89..f28c3a5 100644
--- a/drivers/mtd/nand/raw/nand_base.c
+++ b/drivers/mtd/nand/raw/nand_base.c
@@ -706,12 +706,17 @@ static void nand_wait_status_ready(struct mtd_info *mtd, unsigned long timeo)
  */
 int nand_soft_waitrdy(struct nand_chip *chip, unsigned long timeout_ms)
 {
+	const struct nand_sdr_timings *timings;
 	u8 status = 0;
 	int ret;
 
 	if (!chip->exec_op)
 		return -ENOTSUPP;
 
+	/* Wait tWB before polling the STATUS reg. */
+	timings = nand_get_sdr_timings(&chip->data_interface);
+	ndelay(PSEC_TO_NSEC(timings->tWB_max));
+
 	ret = nand_status_op(chip, NULL);
 	if (ret)
 		return ret;
-- 
cgit v1.1


From 3a15b38fd2efc1d648cb33186bf71e9138c93491 Mon Sep 17 00:00:00 2001
From: Ilya Dryomov <idryomov@gmail.com>
Date: Thu, 3 May 2018 16:10:09 +0200
Subject: ceph: fix rsize/wsize capping in ceph_direct_read_write()

rsize/wsize cap should be applied before ceph_osdc_new_request() is
called.  Otherwise, if the size is limited by the cap instead of the
stripe unit, ceph_osdc_new_request() would setup an extent op that is
bigger than what dio_get_pages_alloc() would pin and add to the page
vector, triggering asserts in the messenger.

Cc: stable@vger.kernel.org
Fixes: 95cca2b44e54 ("ceph: limit osd write size")
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Reviewed-by: "Yan, Zheng" <zyan@redhat.com>
---
 fs/ceph/file.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/fs/ceph/file.c b/fs/ceph/file.c
index f85040d..8ce7849 100644
--- a/fs/ceph/file.c
+++ b/fs/ceph/file.c
@@ -918,6 +918,11 @@ ceph_direct_read_write(struct kiocb *iocb, struct iov_iter *iter,
 		size_t start = 0;
 		ssize_t len;
 
+		if (write)
+			size = min_t(u64, size, fsc->mount_options->wsize);
+		else
+			size = min_t(u64, size, fsc->mount_options->rsize);
+
 		vino = ceph_vino(inode);
 		req = ceph_osdc_new_request(&fsc->client->osdc, &ci->i_layout,
 					    vino, pos, &size, 0,
@@ -933,11 +938,6 @@ ceph_direct_read_write(struct kiocb *iocb, struct iov_iter *iter,
 			break;
 		}
 
-		if (write)
-			size = min_t(u64, size, fsc->mount_options->wsize);
-		else
-			size = min_t(u64, size, fsc->mount_options->rsize);
-
 		len = size;
 		pages = dio_get_pages_alloc(iter, len, &start, &num_pages);
 		if (IS_ERR(pages)) {
-- 
cgit v1.1


From 0010f7052d6cb71c4b120238e28cd3fa413913d1 Mon Sep 17 00:00:00 2001
From: Ilya Dryomov <idryomov@gmail.com>
Date: Fri, 4 May 2018 16:57:30 +0200
Subject: libceph: add osd_req_op_extent_osd_data_bvecs()

... and store num_bvecs for client code's convenience.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Reviewed-by: Jeff Layton <jlayton@redhat.com>
Reviewed-by: "Yan, Zheng" <zyan@redhat.com>
---
 drivers/block/rbd.c             |  4 +++-
 include/linux/ceph/osd_client.h | 12 ++++++++++--
 net/ceph/osd_client.c           | 27 +++++++++++++++++++++++----
 3 files changed, 36 insertions(+), 7 deletions(-)

diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c
index 8e8b04c..33b36fe 100644
--- a/drivers/block/rbd.c
+++ b/drivers/block/rbd.c
@@ -2366,7 +2366,9 @@ static int rbd_obj_issue_copyup(struct rbd_obj_request *obj_req, u32 bytes)
 	osd_req_op_cls_init(obj_req->osd_req, 0, CEPH_OSD_OP_CALL, "rbd",
 			    "copyup");
 	osd_req_op_cls_request_data_bvecs(obj_req->osd_req, 0,
-					  obj_req->copyup_bvecs, bytes);
+					  obj_req->copyup_bvecs,
+					  obj_req->copyup_bvec_count,
+					  bytes);
 
 	switch (obj_req->img_request->op_type) {
 	case OBJ_OP_WRITE:
diff --git a/include/linux/ceph/osd_client.h b/include/linux/ceph/osd_client.h
index 528ccc9..96bb322 100644
--- a/include/linux/ceph/osd_client.h
+++ b/include/linux/ceph/osd_client.h
@@ -77,7 +77,10 @@ struct ceph_osd_data {
 			u32			bio_length;
 		};
 #endif /* CONFIG_BLOCK */
-		struct ceph_bvec_iter	bvec_pos;
+		struct {
+			struct ceph_bvec_iter	bvec_pos;
+			u32			num_bvecs;
+		};
 	};
 };
 
@@ -412,6 +415,10 @@ void osd_req_op_extent_osd_data_bio(struct ceph_osd_request *osd_req,
 				    struct ceph_bio_iter *bio_pos,
 				    u32 bio_length);
 #endif /* CONFIG_BLOCK */
+void osd_req_op_extent_osd_data_bvecs(struct ceph_osd_request *osd_req,
+				      unsigned int which,
+				      struct bio_vec *bvecs, u32 num_bvecs,
+				      u32 bytes);
 void osd_req_op_extent_osd_data_bvec_pos(struct ceph_osd_request *osd_req,
 					 unsigned int which,
 					 struct ceph_bvec_iter *bvec_pos);
@@ -426,7 +433,8 @@ extern void osd_req_op_cls_request_data_pages(struct ceph_osd_request *,
 					bool own_pages);
 void osd_req_op_cls_request_data_bvecs(struct ceph_osd_request *osd_req,
 				       unsigned int which,
-				       struct bio_vec *bvecs, u32 bytes);
+				       struct bio_vec *bvecs, u32 num_bvecs,
+				       u32 bytes);
 extern void osd_req_op_cls_response_data_pages(struct ceph_osd_request *,
 					unsigned int which,
 					struct page **pages, u64 length,
diff --git a/net/ceph/osd_client.c b/net/ceph/osd_client.c
index ea2a6c9f..d2667e5 100644
--- a/net/ceph/osd_client.c
+++ b/net/ceph/osd_client.c
@@ -157,10 +157,12 @@ static void ceph_osd_data_bio_init(struct ceph_osd_data *osd_data,
 #endif /* CONFIG_BLOCK */
 
 static void ceph_osd_data_bvecs_init(struct ceph_osd_data *osd_data,
-				     struct ceph_bvec_iter *bvec_pos)
+				     struct ceph_bvec_iter *bvec_pos,
+				     u32 num_bvecs)
 {
 	osd_data->type = CEPH_OSD_DATA_TYPE_BVECS;
 	osd_data->bvec_pos = *bvec_pos;
+	osd_data->num_bvecs = num_bvecs;
 }
 
 #define osd_req_op_data(oreq, whch, typ, fld)				\
@@ -237,6 +239,22 @@ void osd_req_op_extent_osd_data_bio(struct ceph_osd_request *osd_req,
 EXPORT_SYMBOL(osd_req_op_extent_osd_data_bio);
 #endif /* CONFIG_BLOCK */
 
+void osd_req_op_extent_osd_data_bvecs(struct ceph_osd_request *osd_req,
+				      unsigned int which,
+				      struct bio_vec *bvecs, u32 num_bvecs,
+				      u32 bytes)
+{
+	struct ceph_osd_data *osd_data;
+	struct ceph_bvec_iter it = {
+		.bvecs = bvecs,
+		.iter = { .bi_size = bytes },
+	};
+
+	osd_data = osd_req_op_data(osd_req, which, extent, osd_data);
+	ceph_osd_data_bvecs_init(osd_data, &it, num_bvecs);
+}
+EXPORT_SYMBOL(osd_req_op_extent_osd_data_bvecs);
+
 void osd_req_op_extent_osd_data_bvec_pos(struct ceph_osd_request *osd_req,
 					 unsigned int which,
 					 struct ceph_bvec_iter *bvec_pos)
@@ -244,7 +262,7 @@ void osd_req_op_extent_osd_data_bvec_pos(struct ceph_osd_request *osd_req,
 	struct ceph_osd_data *osd_data;
 
 	osd_data = osd_req_op_data(osd_req, which, extent, osd_data);
-	ceph_osd_data_bvecs_init(osd_data, bvec_pos);
+	ceph_osd_data_bvecs_init(osd_data, bvec_pos, 0);
 }
 EXPORT_SYMBOL(osd_req_op_extent_osd_data_bvec_pos);
 
@@ -287,7 +305,8 @@ EXPORT_SYMBOL(osd_req_op_cls_request_data_pages);
 
 void osd_req_op_cls_request_data_bvecs(struct ceph_osd_request *osd_req,
 				       unsigned int which,
-				       struct bio_vec *bvecs, u32 bytes)
+				       struct bio_vec *bvecs, u32 num_bvecs,
+				       u32 bytes)
 {
 	struct ceph_osd_data *osd_data;
 	struct ceph_bvec_iter it = {
@@ -296,7 +315,7 @@ void osd_req_op_cls_request_data_bvecs(struct ceph_osd_request *osd_req,
 	};
 
 	osd_data = osd_req_op_data(osd_req, which, cls, request_data);
-	ceph_osd_data_bvecs_init(osd_data, &it);
+	ceph_osd_data_bvecs_init(osd_data, &it, num_bvecs);
 	osd_req->r_ops[which].cls.indata_len += bytes;
 	osd_req->r_ops[which].indata_len += bytes;
 }
-- 
cgit v1.1


From fc218544fbc800d1c91348ec834cacfb257348f7 Mon Sep 17 00:00:00 2001
From: Ilya Dryomov <idryomov@gmail.com>
Date: Fri, 4 May 2018 16:57:31 +0200
Subject: ceph: fix iov_iter issues in ceph_direct_read_write()

dio_get_pagev_size() and dio_get_pages_alloc() introduced in commit
b5b98989dc7e ("ceph: combine as many iovec as possile into one OSD
request") assume that the passed iov_iter is ITER_IOVEC.  This isn't
the case with splice where it ends up poking into the guts of ITER_BVEC
or ITER_PIPE iterators, causing lockups and crashes easily reproduced
with generic/095.

Rather than trying to figure out gap alignment and stuff pages into
a page vector, add a helper for going from iov_iter to a bio_vec array
and make use of the new CEPH_OSD_DATA_TYPE_BVECS code.

Fixes: b5b98989dc7e ("ceph: combine as many iovec as possile into one OSD request")
Link: http://tracker.ceph.com/issues/18130
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Reviewed-by: Jeff Layton <jlayton@redhat.com>
Reviewed-by: "Yan, Zheng" <zyan@redhat.com>
Tested-by: Luis Henriques <lhenriques@suse.com>
---
 fs/ceph/file.c | 195 ++++++++++++++++++++++++++++++++++-----------------------
 1 file changed, 117 insertions(+), 78 deletions(-)

diff --git a/fs/ceph/file.c b/fs/ceph/file.c
index 8ce7849..cf0e45b 100644
--- a/fs/ceph/file.c
+++ b/fs/ceph/file.c
@@ -70,69 +70,104 @@ static __le32 ceph_flags_sys2wire(u32 flags)
  */
 
 /*
- * Calculate the length sum of direct io vectors that can
- * be combined into one page vector.
+ * How many pages to get in one call to iov_iter_get_pages().  This
+ * determines the size of the on-stack array used as a buffer.
  */
-static size_t dio_get_pagev_size(const struct iov_iter *it)
+#define ITER_GET_BVECS_PAGES	64
+
+static ssize_t __iter_get_bvecs(struct iov_iter *iter, size_t maxsize,
+				struct bio_vec *bvecs)
 {
-    const struct iovec *iov = it->iov;
-    const struct iovec *iovend = iov + it->nr_segs;
-    size_t size;
-
-    size = iov->iov_len - it->iov_offset;
-    /*
-     * An iov can be page vectored when both the current tail
-     * and the next base are page aligned.
-     */
-    while (PAGE_ALIGNED((iov->iov_base + iov->iov_len)) &&
-           (++iov < iovend && PAGE_ALIGNED((iov->iov_base)))) {
-        size += iov->iov_len;
-    }
-    dout("dio_get_pagevlen len = %zu\n", size);
-    return size;
+	size_t size = 0;
+	int bvec_idx = 0;
+
+	if (maxsize > iov_iter_count(iter))
+		maxsize = iov_iter_count(iter);
+
+	while (size < maxsize) {
+		struct page *pages[ITER_GET_BVECS_PAGES];
+		ssize_t bytes;
+		size_t start;
+		int idx = 0;
+
+		bytes = iov_iter_get_pages(iter, pages, maxsize - size,
+					   ITER_GET_BVECS_PAGES, &start);
+		if (bytes < 0)
+			return size ?: bytes;
+
+		iov_iter_advance(iter, bytes);
+		size += bytes;
+
+		for ( ; bytes; idx++, bvec_idx++) {
+			struct bio_vec bv = {
+				.bv_page = pages[idx],
+				.bv_len = min_t(int, bytes, PAGE_SIZE - start),
+				.bv_offset = start,
+			};
+
+			bvecs[bvec_idx] = bv;
+			bytes -= bv.bv_len;
+			start = 0;
+		}
+	}
+
+	return size;
 }
 
 /*
- * Allocate a page vector based on (@it, @nbytes).
- * The return value is the tuple describing a page vector,
- * that is (@pages, @page_align, @num_pages).
+ * iov_iter_get_pages() only considers one iov_iter segment, no matter
+ * what maxsize or maxpages are given.  For ITER_BVEC that is a single
+ * page.
+ *
+ * Attempt to get up to @maxsize bytes worth of pages from @iter.
+ * Return the number of bytes in the created bio_vec array, or an error.
  */
-static struct page **
-dio_get_pages_alloc(const struct iov_iter *it, size_t nbytes,
-		    size_t *page_align, int *num_pages)
+static ssize_t iter_get_bvecs_alloc(struct iov_iter *iter, size_t maxsize,
+				    struct bio_vec **bvecs, int *num_bvecs)
 {
-	struct iov_iter tmp_it = *it;
-	size_t align;
-	struct page **pages;
-	int ret = 0, idx, npages;
+	struct bio_vec *bv;
+	size_t orig_count = iov_iter_count(iter);
+	ssize_t bytes;
+	int npages;
 
-	align = (unsigned long)(it->iov->iov_base + it->iov_offset) &
-		(PAGE_SIZE - 1);
-	npages = calc_pages_for(align, nbytes);
-	pages = kvmalloc(sizeof(*pages) * npages, GFP_KERNEL);
-	if (!pages)
-		return ERR_PTR(-ENOMEM);
+	iov_iter_truncate(iter, maxsize);
+	npages = iov_iter_npages(iter, INT_MAX);
+	iov_iter_reexpand(iter, orig_count);
 
-	for (idx = 0; idx < npages; ) {
-		size_t start;
-		ret = iov_iter_get_pages(&tmp_it, pages + idx, nbytes,
-					 npages - idx, &start);
-		if (ret < 0)
-			goto fail;
+	/*
+	 * __iter_get_bvecs() may populate only part of the array -- zero it
+	 * out.
+	 */
+	bv = kvmalloc_array(npages, sizeof(*bv), GFP_KERNEL | __GFP_ZERO);
+	if (!bv)
+		return -ENOMEM;
 
-		iov_iter_advance(&tmp_it, ret);
-		nbytes -= ret;
-		idx += (ret + start + PAGE_SIZE - 1) / PAGE_SIZE;
+	bytes = __iter_get_bvecs(iter, maxsize, bv);
+	if (bytes < 0) {
+		/*
+		 * No pages were pinned -- just free the array.
+		 */
+		kvfree(bv);
+		return bytes;
 	}
 
-	BUG_ON(nbytes != 0);
-	*num_pages = npages;
-	*page_align = align;
-	dout("dio_get_pages_alloc: got %d pages align %zu\n", npages, align);
-	return pages;
-fail:
-	ceph_put_page_vector(pages, idx, false);
-	return ERR_PTR(ret);
+	*bvecs = bv;
+	*num_bvecs = npages;
+	return bytes;
+}
+
+static void put_bvecs(struct bio_vec *bvecs, int num_bvecs, bool should_dirty)
+{
+	int i;
+
+	for (i = 0; i < num_bvecs; i++) {
+		if (bvecs[i].bv_page) {
+			if (should_dirty)
+				set_page_dirty_lock(bvecs[i].bv_page);
+			put_page(bvecs[i].bv_page);
+		}
+	}
+	kvfree(bvecs);
 }
 
 /*
@@ -746,11 +781,12 @@ static void ceph_aio_complete_req(struct ceph_osd_request *req)
 	struct inode *inode = req->r_inode;
 	struct ceph_aio_request *aio_req = req->r_priv;
 	struct ceph_osd_data *osd_data = osd_req_op_extent_osd_data(req, 0);
-	int num_pages = calc_pages_for((u64)osd_data->alignment,
-				       osd_data->length);
 
-	dout("ceph_aio_complete_req %p rc %d bytes %llu\n",
-	     inode, rc, osd_data->length);
+	BUG_ON(osd_data->type != CEPH_OSD_DATA_TYPE_BVECS);
+	BUG_ON(!osd_data->num_bvecs);
+
+	dout("ceph_aio_complete_req %p rc %d bytes %u\n",
+	     inode, rc, osd_data->bvec_pos.iter.bi_size);
 
 	if (rc == -EOLDSNAPC) {
 		struct ceph_aio_work *aio_work;
@@ -768,9 +804,10 @@ static void ceph_aio_complete_req(struct ceph_osd_request *req)
 	} else if (!aio_req->write) {
 		if (rc == -ENOENT)
 			rc = 0;
-		if (rc >= 0 && osd_data->length > rc) {
-			int zoff = osd_data->alignment + rc;
-			int zlen = osd_data->length - rc;
+		if (rc >= 0 && osd_data->bvec_pos.iter.bi_size > rc) {
+			struct iov_iter i;
+			int zlen = osd_data->bvec_pos.iter.bi_size - rc;
+
 			/*
 			 * If read is satisfied by single OSD request,
 			 * it can pass EOF. Otherwise read is within
@@ -785,13 +822,16 @@ static void ceph_aio_complete_req(struct ceph_osd_request *req)
 				aio_req->total_len = rc + zlen;
 			}
 
-			if (zlen > 0)
-				ceph_zero_page_vector_range(zoff, zlen,
-							    osd_data->pages);
+			iov_iter_bvec(&i, ITER_BVEC, osd_data->bvec_pos.bvecs,
+				      osd_data->num_bvecs,
+				      osd_data->bvec_pos.iter.bi_size);
+			iov_iter_advance(&i, rc);
+			iov_iter_zero(zlen, &i);
 		}
 	}
 
-	ceph_put_page_vector(osd_data->pages, num_pages, aio_req->should_dirty);
+	put_bvecs(osd_data->bvec_pos.bvecs, osd_data->num_bvecs,
+		  aio_req->should_dirty);
 	ceph_osdc_put_request(req);
 
 	if (rc < 0)
@@ -879,7 +919,7 @@ ceph_direct_read_write(struct kiocb *iocb, struct iov_iter *iter,
 	struct ceph_fs_client *fsc = ceph_inode_to_client(inode);
 	struct ceph_vino vino;
 	struct ceph_osd_request *req;
-	struct page **pages;
+	struct bio_vec *bvecs;
 	struct ceph_aio_request *aio_req = NULL;
 	int num_pages = 0;
 	int flags;
@@ -914,8 +954,7 @@ ceph_direct_read_write(struct kiocb *iocb, struct iov_iter *iter,
 	}
 
 	while (iov_iter_count(iter) > 0) {
-		u64 size = dio_get_pagev_size(iter);
-		size_t start = 0;
+		u64 size = iov_iter_count(iter);
 		ssize_t len;
 
 		if (write)
@@ -938,13 +977,14 @@ ceph_direct_read_write(struct kiocb *iocb, struct iov_iter *iter,
 			break;
 		}
 
-		len = size;
-		pages = dio_get_pages_alloc(iter, len, &start, &num_pages);
-		if (IS_ERR(pages)) {
+		len = iter_get_bvecs_alloc(iter, size, &bvecs, &num_pages);
+		if (len < 0) {
 			ceph_osdc_put_request(req);
-			ret = PTR_ERR(pages);
+			ret = len;
 			break;
 		}
+		if (len != size)
+			osd_req_op_extent_update(req, 0, len);
 
 		/*
 		 * To simplify error handling, allow AIO when IO within i_size
@@ -977,8 +1017,7 @@ ceph_direct_read_write(struct kiocb *iocb, struct iov_iter *iter,
 			req->r_mtime = mtime;
 		}
 
-		osd_req_op_extent_osd_data_pages(req, 0, pages, len, start,
-						 false, false);
+		osd_req_op_extent_osd_data_bvecs(req, 0, bvecs, num_pages, len);
 
 		if (aio_req) {
 			aio_req->total_len += len;
@@ -991,7 +1030,6 @@ ceph_direct_read_write(struct kiocb *iocb, struct iov_iter *iter,
 			list_add_tail(&req->r_unsafe_item, &aio_req->osd_reqs);
 
 			pos += len;
-			iov_iter_advance(iter, len);
 			continue;
 		}
 
@@ -1004,25 +1042,26 @@ ceph_direct_read_write(struct kiocb *iocb, struct iov_iter *iter,
 			if (ret == -ENOENT)
 				ret = 0;
 			if (ret >= 0 && ret < len && pos + ret < size) {
+				struct iov_iter i;
 				int zlen = min_t(size_t, len - ret,
 						 size - pos - ret);
-				ceph_zero_page_vector_range(start + ret, zlen,
-							    pages);
+
+				iov_iter_bvec(&i, ITER_BVEC, bvecs, num_pages,
+					      len);
+				iov_iter_advance(&i, ret);
+				iov_iter_zero(zlen, &i);
 				ret += zlen;
 			}
 			if (ret >= 0)
 				len = ret;
 		}
 
-		ceph_put_page_vector(pages, num_pages, should_dirty);
-
+		put_bvecs(bvecs, num_pages, should_dirty);
 		ceph_osdc_put_request(req);
 		if (ret < 0)
 			break;
 
 		pos += len;
-		iov_iter_advance(iter, len);
-
 		if (!write && pos >= size)
 			break;
 
-- 
cgit v1.1


From 4a026da91caaa36004a53a844dd00959370ea8fc Mon Sep 17 00:00:00 2001
From: Sun Lianwen <sunlw.fnst@cn.fujitsu.com>
Date: Tue, 8 May 2018 09:49:38 +0800
Subject: net/9p: correct some comment errors in 9p file system code

There are follow comment errors:
1 The function name is wrong in p9_release_pages() comment.
2 The function name and variable name is wrong in p9_poll_workfn() comment.
3 There is no variable dm_mr and lkey in struct p9_trans_rdma.
4 The function name is wrong in rdma_create_trans() comment.
5 There is no variable initialized in struct virtio_chan.
6 The variable name is wrong in p9_virtio_zc_request() comment.

Signed-off-by: Sun Lianwen <sunlw.fnst@cn.fujitsu.com>
Reviewed-by: Randy Dunlap <rdunlap@infradead.org>
Reviewed-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/9p/trans_common.c | 2 +-
 net/9p/trans_fd.c     | 4 ++--
 net/9p/trans_rdma.c   | 4 +---
 net/9p/trans_virtio.c | 5 ++---
 4 files changed, 6 insertions(+), 9 deletions(-)

diff --git a/net/9p/trans_common.c b/net/9p/trans_common.c
index 38aa634..b718db2 100644
--- a/net/9p/trans_common.c
+++ b/net/9p/trans_common.c
@@ -16,7 +16,7 @@
 #include <linux/module.h>
 
 /**
- *  p9_release_req_pages - Release pages after the transaction.
+ *  p9_release_pages - Release pages after the transaction.
  */
 void p9_release_pages(struct page **pages, int nr_pages)
 {
diff --git a/net/9p/trans_fd.c b/net/9p/trans_fd.c
index 0cfba91..848969f 100644
--- a/net/9p/trans_fd.c
+++ b/net/9p/trans_fd.c
@@ -1092,8 +1092,8 @@ static struct p9_trans_module p9_fd_trans = {
 };
 
 /**
- * p9_poll_proc - poll worker thread
- * @a: thread state and arguments
+ * p9_poll_workfn - poll worker thread
+ * @work: work queue
  *
  * polls all v9fs transports for new events and queues the appropriate
  * work to the work queue
diff --git a/net/9p/trans_rdma.c b/net/9p/trans_rdma.c
index 6d8e303..3d414ac 100644
--- a/net/9p/trans_rdma.c
+++ b/net/9p/trans_rdma.c
@@ -68,8 +68,6 @@
  * @pd: Protection Domain pointer
  * @qp: Queue Pair pointer
  * @cq: Completion Queue pointer
- * @dm_mr: DMA Memory Region pointer
- * @lkey: The local access only memory region key
  * @timeout: Number of uSecs to wait for connection management events
  * @privport: Whether a privileged port may be used
  * @port: The port to use
@@ -632,7 +630,7 @@ static int p9_rdma_bind_privport(struct p9_trans_rdma *rdma)
 }
 
 /**
- * trans_create_rdma - Transport method for creating atransport instance
+ * rdma_create_trans - Transport method for creating a transport instance
  * @client: client instance
  * @addr: IP address string
  * @args: Mount options string
diff --git a/net/9p/trans_virtio.c b/net/9p/trans_virtio.c
index 3aa5a93..4d03722 100644
--- a/net/9p/trans_virtio.c
+++ b/net/9p/trans_virtio.c
@@ -60,7 +60,6 @@ static atomic_t vp_pinned = ATOMIC_INIT(0);
 
 /**
  * struct virtio_chan - per-instance transport information
- * @initialized: whether the channel is initialized
  * @inuse: whether the channel is in use
  * @lock: protects multiple elements within this structure
  * @client: client instance
@@ -385,8 +384,8 @@ static int p9_get_mapped_pages(struct virtio_chan *chan,
  * @uidata: user bffer that should be ued for zero copy read
  * @uodata: user buffer that shoud be user for zero copy write
  * @inlen: read buffer size
- * @olen: write buffer size
- * @hdrlen: reader header size, This is the size of response protocol data
+ * @outlen: write buffer size
+ * @in_hdr_len: reader header size, This is the size of response protocol data
  *
  */
 static int
-- 
cgit v1.1


From 94f6a80c0c11828cb7b3d79294459dd8d761ca89 Mon Sep 17 00:00:00 2001
From: Ying Xue <ying.xue@windriver.com>
Date: Tue, 8 May 2018 21:44:06 +0800
Subject: tipc: eliminate KMSAN uninit-value in strcmp complaint

When we get link properties through netlink interface with
tipc_nl_node_get_link(), we don't validate TIPC_NLA_LINK_NAME
attribute at all, instead we directly use it. As a consequence,
KMSAN detected the TIPC_NLA_LINK_NAME attribute was an uninitialized
value, and then posted the following complaint:

==================================================================
BUG: KMSAN: uninit-value in strcmp+0xf7/0x160 lib/string.c:329
CPU: 1 PID: 4527 Comm: syz-executor655 Not tainted 4.16.0+ #87
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS
Google 01/01/2011
Call Trace:
  __dump_stack lib/dump_stack.c:17 [inline]
  dump_stack+0x185/0x1d0 lib/dump_stack.c:53
  kmsan_report+0x142/0x240 mm/kmsan/kmsan.c:1067
  __msan_warning_32+0x6c/0xb0 mm/kmsan/kmsan_instr.c:683
  strcmp+0xf7/0x160 lib/string.c:329
  tipc_nl_node_get_link+0x220/0x6f0 net/tipc/node.c:1881
  genl_family_rcv_msg net/netlink/genetlink.c:599 [inline]
  genl_rcv_msg+0x1686/0x1810 net/netlink/genetlink.c:624
  netlink_rcv_skb+0x378/0x600 net/netlink/af_netlink.c:2447
  genl_rcv+0x63/0x80 net/netlink/genetlink.c:635
  netlink_unicast_kernel net/netlink/af_netlink.c:1311 [inline]
  netlink_unicast+0x166b/0x1740 net/netlink/af_netlink.c:1337
  netlink_sendmsg+0x1048/0x1310 net/netlink/af_netlink.c:1900
  sock_sendmsg_nosec net/socket.c:630 [inline]
  sock_sendmsg net/socket.c:640 [inline]
  ___sys_sendmsg+0xec0/0x1310 net/socket.c:2046
  __sys_sendmsg net/socket.c:2080 [inline]
  SYSC_sendmsg+0x2a3/0x3d0 net/socket.c:2091
  SyS_sendmsg+0x54/0x80 net/socket.c:2087
  do_syscall_64+0x309/0x430 arch/x86/entry/common.c:287
  entry_SYSCALL_64_after_hwframe+0x3d/0xa2
RIP: 0033:0x445589
RSP: 002b:00007fb7ee66cdb8 EFLAGS: 00000246 ORIG_RAX: 000000000000002e
RAX: ffffffffffffffda RBX: 00000000006dac24 RCX: 0000000000445589
RDX: 0000000000000000 RSI: 0000000020023000 RDI: 0000000000000003
RBP: 00000000006dac20 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
R13: 00007fffa2bf3f3f R14: 00007fb7ee66d9c0 R15: 0000000000000001

Uninit was created at:
  kmsan_save_stack_with_flags mm/kmsan/kmsan.c:278 [inline]
  kmsan_internal_poison_shadow+0xb8/0x1b0 mm/kmsan/kmsan.c:188
  kmsan_kmalloc+0x94/0x100 mm/kmsan/kmsan.c:314
  kmsan_slab_alloc+0x11/0x20 mm/kmsan/kmsan.c:321
  slab_post_alloc_hook mm/slab.h:445 [inline]
  slab_alloc_node mm/slub.c:2737 [inline]
  __kmalloc_node_track_caller+0xaed/0x11c0 mm/slub.c:4369
  __kmalloc_reserve net/core/skbuff.c:138 [inline]
  __alloc_skb+0x2cf/0x9f0 net/core/skbuff.c:206
  alloc_skb include/linux/skbuff.h:984 [inline]
  netlink_alloc_large_skb net/netlink/af_netlink.c:1183 [inline]
  netlink_sendmsg+0x9a6/0x1310 net/netlink/af_netlink.c:1875
  sock_sendmsg_nosec net/socket.c:630 [inline]
  sock_sendmsg net/socket.c:640 [inline]
  ___sys_sendmsg+0xec0/0x1310 net/socket.c:2046
  __sys_sendmsg net/socket.c:2080 [inline]
  SYSC_sendmsg+0x2a3/0x3d0 net/socket.c:2091
  SyS_sendmsg+0x54/0x80 net/socket.c:2087
  do_syscall_64+0x309/0x430 arch/x86/entry/common.c:287
  entry_SYSCALL_64_after_hwframe+0x3d/0xa2
==================================================================

To quiet the complaint, TIPC_NLA_LINK_NAME attribute has been
validated in tipc_nl_node_get_link() before it's used.

Reported-by: syzbot+df0257c92ffd4fcc58cd@syzkaller.appspotmail.com
Signed-off-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/tipc/node.c | 15 +++++++++++++--
 1 file changed, 13 insertions(+), 2 deletions(-)

diff --git a/net/tipc/node.c b/net/tipc/node.c
index baaf93f..f29549d 100644
--- a/net/tipc/node.c
+++ b/net/tipc/node.c
@@ -1950,6 +1950,7 @@ out:
 int tipc_nl_node_get_link(struct sk_buff *skb, struct genl_info *info)
 {
 	struct net *net = genl_info_net(info);
+	struct nlattr *attrs[TIPC_NLA_LINK_MAX + 1];
 	struct tipc_nl_msg msg;
 	char *name;
 	int err;
@@ -1957,9 +1958,19 @@ int tipc_nl_node_get_link(struct sk_buff *skb, struct genl_info *info)
 	msg.portid = info->snd_portid;
 	msg.seq = info->snd_seq;
 
-	if (!info->attrs[TIPC_NLA_LINK_NAME])
+	if (!info->attrs[TIPC_NLA_LINK])
 		return -EINVAL;
-	name = nla_data(info->attrs[TIPC_NLA_LINK_NAME]);
+
+	err = nla_parse_nested(attrs, TIPC_NLA_LINK_MAX,
+			       info->attrs[TIPC_NLA_LINK],
+			       tipc_nl_link_policy, info->extack);
+	if (err)
+		return err;
+
+	if (!attrs[TIPC_NLA_LINK_NAME])
+		return -EINVAL;
+
+	name = nla_data(attrs[TIPC_NLA_LINK_NAME]);
 
 	msg.skb = nlmsg_new(NLMSG_GOODSIZE, GFP_KERNEL);
 	if (!msg.skb)
-- 
cgit v1.1


From 7fc6311b174091e3283c28381e58bed3d12b6591 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Thu, 10 May 2018 19:51:09 +0800
Subject: drm/amd/pp: Fix performance drop on Fiji

The performance drop if the default TDP more than 256 Watt

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c
index 03bc745..d9e92e3 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c
@@ -852,12 +852,10 @@ int smu7_set_power_limit(struct pp_hwmgr *hwmgr, uint32_t n)
 {
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
 
-	n = (n & 0xff) << 8;
-
 	if (data->power_containment_features &
 			POWERCONTAINMENT_FEATURE_PkgPwrLimit)
 		return smum_send_msg_to_smc_with_parameter(hwmgr,
-				PPSMC_MSG_PkgPwrSetLimit, n);
+				PPSMC_MSG_PkgPwrSetLimit, n<<8);
 	return 0;
 }
 
-- 
cgit v1.1


From cfcadfaad7251d8b640713724b388164d75465b2 Mon Sep 17 00:00:00 2001
From: "Rafael J. Wysocki" <rafael.j.wysocki@intel.com>
Date: Wed, 9 May 2018 00:18:32 +0200
Subject: PCI / PM: Check device_may_wakeup() in pci_enable_wake()

Commit 0847684cfc5f0 (PCI / PM: Simplify device wakeup settings code)
went too far and dropped the device_may_wakeup() check from
pci_enable_wake() which causes wakeup to be enabled during system
suspend, hibernation or shutdown for some PCI devices that are not
allowed by user space to wake up the system from sleep (or power off).

As a result of this, excessive power is drawn by some of the affected
systems while in sleep states or off.

Restore the device_may_wakeup() check in pci_enable_wake(), but make
sure that the PCI bus type's runtime suspend callback will not call
device_may_wakeup() which is about system wakeup from sleep and not
about device wakeup from runtime suspend.

Fixes: 0847684cfc5f0 (PCI / PM: Simplify device wakeup settings code)
Reported-by: Joseph Salisbury <joseph.salisbury@canonical.com>
Cc: 4.13+ <stable@vger.kernel.org> # 4.13+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Bjorn Helgaas <bhelgaas@google.com>
---
 drivers/pci/pci.c | 29 +++++++++++++++++++++++------
 1 file changed, 23 insertions(+), 6 deletions(-)

diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c
index c2616ca..dbfe7c4 100644
--- a/drivers/pci/pci.c
+++ b/drivers/pci/pci.c
@@ -1910,7 +1910,7 @@ void pci_pme_active(struct pci_dev *dev, bool enable)
 EXPORT_SYMBOL(pci_pme_active);
 
 /**
- * pci_enable_wake - enable PCI device as wakeup event source
+ * __pci_enable_wake - enable PCI device as wakeup event source
  * @dev: PCI device affected
  * @state: PCI state from which device will issue wakeup events
  * @enable: True to enable event generation; false to disable
@@ -1928,7 +1928,7 @@ EXPORT_SYMBOL(pci_pme_active);
  * Error code depending on the platform is returned if both the platform and
  * the native mechanism fail to enable the generation of wake-up events
  */
-int pci_enable_wake(struct pci_dev *dev, pci_power_t state, bool enable)
+static int __pci_enable_wake(struct pci_dev *dev, pci_power_t state, bool enable)
 {
 	int ret = 0;
 
@@ -1969,6 +1969,23 @@ int pci_enable_wake(struct pci_dev *dev, pci_power_t state, bool enable)
 
 	return ret;
 }
+
+/**
+ * pci_enable_wake - change wakeup settings for a PCI device
+ * @pci_dev: Target device
+ * @state: PCI state from which device will issue wakeup events
+ * @enable: Whether or not to enable event generation
+ *
+ * If @enable is set, check device_may_wakeup() for the device before calling
+ * __pci_enable_wake() for it.
+ */
+int pci_enable_wake(struct pci_dev *pci_dev, pci_power_t state, bool enable)
+{
+	if (enable && !device_may_wakeup(&pci_dev->dev))
+		return -EINVAL;
+
+	return __pci_enable_wake(pci_dev, state, enable);
+}
 EXPORT_SYMBOL(pci_enable_wake);
 
 /**
@@ -1981,9 +1998,9 @@ EXPORT_SYMBOL(pci_enable_wake);
  * should not be called twice in a row to enable wake-up due to PCI PM vs ACPI
  * ordering constraints.
  *
- * This function only returns error code if the device is not capable of
- * generating PME# from both D3_hot and D3_cold, and the platform is unable to
- * enable wake-up power for it.
+ * This function only returns error code if the device is not allowed to wake
+ * up the system from sleep or it is not capable of generating PME# from both
+ * D3_hot and D3_cold and the platform is unable to enable wake-up power for it.
  */
 int pci_wake_from_d3(struct pci_dev *dev, bool enable)
 {
@@ -2114,7 +2131,7 @@ int pci_finish_runtime_suspend(struct pci_dev *dev)
 
 	dev->runtime_d3cold = target_state == PCI_D3cold;
 
-	pci_enable_wake(dev, target_state, pci_dev_run_wake(dev));
+	__pci_enable_wake(dev, target_state, pci_dev_run_wake(dev));
 
 	error = pci_set_power_state(dev, target_state);
 
-- 
cgit v1.1


From 28700a36232ba61fb6ac59466821546cb25aec69 Mon Sep 17 00:00:00 2001
From: Mike Snitzer <snitzer@redhat.com>
Date: Thu, 10 May 2018 11:18:49 -0400
Subject: dm thin: update Documentation to clarify when "read_only" is valid

Due to user confusion, clarify that it doesn't make sense to try to
create a thin-pool with "read_only" mode enabled.

Signed-off-by: Mike Snitzer <snitzer@redhat.com>
---
 Documentation/device-mapper/thin-provisioning.txt | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/Documentation/device-mapper/thin-provisioning.txt b/Documentation/device-mapper/thin-provisioning.txt
index 4bcd4b7..3d01948 100644
--- a/Documentation/device-mapper/thin-provisioning.txt
+++ b/Documentation/device-mapper/thin-provisioning.txt
@@ -264,7 +264,10 @@ i) Constructor
 			   data device, but just remove the mapping.
 
       read_only: Don't allow any changes to be made to the pool
-		 metadata.
+		 metadata.  This mode is only available after the
+		 thin-pool has been created and first used in full
+		 read/write mode.  It cannot be specified on initial
+		 thin-pool creation.
 
       error_if_no_space: Error IOs, instead of queueing, if no space.
 
-- 
cgit v1.1


From 5cec9425b41dcf834c3d48776900d6acb7e96f38 Mon Sep 17 00:00:00 2001
From: Lukas Wunner <lukas@wunner.de>
Date: Wed, 9 May 2018 14:38:43 +0200
Subject: can: hi311x: Acquire SPI lock on ->do_get_berr_counter

hi3110_get_berr_counter() may run concurrently to the rest of the driver
but neglects to acquire the lock protecting access to the SPI device.
As a result, it and the rest of the driver may clobber each other's tx
and rx buffers.

We became aware of this issue because transmission of packets with
"cangen -g 0 -i -x" frequently hung.  It turns out that agetty executes
->do_get_berr_counter every few seconds via the following call stack:

    CPU: 2 PID: 1605 Comm: agetty
    [<7f3f7500>] (hi3110_get_berr_counter [hi311x])
    [<7f130204>] (can_fill_info [can_dev])
    [<80693bc0>] (rtnl_fill_ifinfo)
    [<806949ec>] (rtnl_dump_ifinfo)
    [<806b4834>] (netlink_dump)
    [<806b4bc8>] (netlink_recvmsg)
    [<8065f180>] (sock_recvmsg)
    [<80660f90>] (___sys_recvmsg)
    [<80661e7c>] (__sys_recvmsg)
    [<80661ec0>] (SyS_recvmsg)
    [<80108b20>] (ret_fast_syscall+0x0/0x1c)

agetty listens to netlink messages in order to update the login prompt
when IP addresses change (if /etc/issue contains \4 or \6 escape codes):
https://git.kernel.org/pub/scm/utils/util-linux/util-linux.git/commit/?id=e36deb6424e8

It's a useful feature, though it seems questionable that it causes CAN
bit error statistics to be queried.

Be that as it may, if hi3110_get_berr_counter() is invoked while a frame
is sent by hi3110_hw_tx(), bogus SPI transfers like the following may
occur:

    => 12 00             (hi3110_get_berr_counter() wanted to transmit
                          EC 00 to query the transmit error counter,
                          but the first byte was overwritten by
                          hi3110_hw_tx_frame())

    => EA 00 3E 80 01 FB (hi3110_hw_tx_frame() wanted to transmit a
                          frame, but the first byte was overwritten by
                          hi3110_get_berr_counter() because it wanted
                          to query the receive error counter)

This sequence hangs the transmission because the driver believes it has
sent a frame and waits for the interrupt signaling completion, but in
reality the chip has never sent away the frame since the commands it
received were malformed.

Fix by acquiring the SPI lock in hi3110_get_berr_counter().

I've scrutinized the entire driver for further unlocked SPI accesses but
found no others.

Cc: Mathias Duckeck <m.duckeck@kunbus.de>
Cc: Akshay Bhat <akshay.bhat@timesys.com>
Cc: Casey Fitzpatrick <casey.fitzpatrick@timesys.com>
Cc: Stef Walter <stefw@redhat.com>
Cc: Karel Zak <kzak@redhat.com>
Cc: stable@vger.kernel.org # v4.12+
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Reviewed-by: Akshay Bhat <akshay.bhat@timesys.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
---
 drivers/net/can/spi/hi311x.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/net/can/spi/hi311x.c b/drivers/net/can/spi/hi311x.c
index 5590c55..c2cf254 100644
--- a/drivers/net/can/spi/hi311x.c
+++ b/drivers/net/can/spi/hi311x.c
@@ -427,8 +427,10 @@ static int hi3110_get_berr_counter(const struct net_device *net,
 	struct hi3110_priv *priv = netdev_priv(net);
 	struct spi_device *spi = priv->spi;
 
+	mutex_lock(&priv->hi3110_lock);
 	bec->txerr = hi3110_read(spi, HI3110_READ_TEC);
 	bec->rxerr = hi3110_read(spi, HI3110_READ_REC);
+	mutex_unlock(&priv->hi3110_lock);
 
 	return 0;
 }
-- 
cgit v1.1


From 32bee8f48fa048a3198109de50e51c092507ff52 Mon Sep 17 00:00:00 2001
From: Lukas Wunner <lukas@wunner.de>
Date: Wed, 9 May 2018 14:43:43 +0200
Subject: can: hi311x: Work around TX complete interrupt erratum

When sending packets as fast as possible using "cangen -g 0 -i -x", the
HI-3110 occasionally latches the interrupt pin high on completion of a
packet, but doesn't set the TXCPLT bit in the INTF register.  The INTF
register contains 0x00 as if no interrupt has occurred.  Even waiting
for a few milliseconds after the interrupt doesn't help.

Work around this apparent erratum by instead checking the TXMTY bit in
the STATF register ("TX FIFO empty").  We know that we've queued up a
packet for transmission if priv->tx_len is nonzero.  If the TX FIFO is
empty, transmission of that packet must have completed.

Note that this is congruent with our handling of received packets, which
likewise gleans from the STATF register whether a packet is waiting in
the RX FIFO, instead of looking at the INTF register.

Cc: Mathias Duckeck <m.duckeck@kunbus.de>
Cc: Akshay Bhat <akshay.bhat@timesys.com>
Cc: Casey Fitzpatrick <casey.fitzpatrick@timesys.com>
Cc: stable@vger.kernel.org # v4.12+
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Acked-by: Akshay Bhat <akshay.bhat@timesys.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
---
 drivers/net/can/spi/hi311x.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/drivers/net/can/spi/hi311x.c b/drivers/net/can/spi/hi311x.c
index c2cf254..53e320c 100644
--- a/drivers/net/can/spi/hi311x.c
+++ b/drivers/net/can/spi/hi311x.c
@@ -91,6 +91,7 @@
 #define HI3110_STAT_BUSOFF BIT(2)
 #define HI3110_STAT_ERRP BIT(3)
 #define HI3110_STAT_ERRW BIT(4)
+#define HI3110_STAT_TXMTY BIT(7)
 
 #define HI3110_BTR0_SJW_SHIFT 6
 #define HI3110_BTR0_BRP_SHIFT 0
@@ -737,10 +738,7 @@ static irqreturn_t hi3110_can_ist(int irq, void *dev_id)
 			}
 		}
 
-		if (intf == 0)
-			break;
-
-		if (intf & HI3110_INT_TXCPLT) {
+		if (priv->tx_len && statf & HI3110_STAT_TXMTY) {
 			net->stats.tx_packets++;
 			net->stats.tx_bytes += priv->tx_len - 1;
 			can_led_event(net, CAN_LED_EVENT_TX);
@@ -750,6 +748,9 @@ static irqreturn_t hi3110_can_ist(int irq, void *dev_id)
 			}
 			netif_wake_queue(net);
 		}
+
+		if (intf == 0)
+			break;
 	}
 	mutex_unlock(&priv->hi3110_lock);
 	return IRQ_HANDLED;
-- 
cgit v1.1


From af6858ee423a309d93054c361c61099b8eb12bbf Mon Sep 17 00:00:00 2001
From: Michal Kalderon <Michal.Kalderon@cavium.com>
Date: Tue, 8 May 2018 21:29:18 +0300
Subject: qed: Fix l2 initializations over iWARP personality

If qede driver was loaded on a device configured for iWARP
the l2 mutex wouldn't be allocated, and some l2 related
resources wouldn't be freed.

fixes: c851a9dc4359 ("qed: Introduce iWARP personality")
Signed-off-by: Michal Kalderon <Michal.Kalderon@cavium.com>
Signed-off-by: Sudarsana Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/qlogic/qed/qed_l2.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/drivers/net/ethernet/qlogic/qed/qed_l2.c b/drivers/net/ethernet/qlogic/qed/qed_l2.c
index e874504..8667799d 100644
--- a/drivers/net/ethernet/qlogic/qed/qed_l2.c
+++ b/drivers/net/ethernet/qlogic/qed/qed_l2.c
@@ -115,8 +115,7 @@ int qed_l2_alloc(struct qed_hwfn *p_hwfn)
 
 void qed_l2_setup(struct qed_hwfn *p_hwfn)
 {
-	if (p_hwfn->hw_info.personality != QED_PCI_ETH &&
-	    p_hwfn->hw_info.personality != QED_PCI_ETH_ROCE)
+	if (!QED_IS_L2_PERSONALITY(p_hwfn))
 		return;
 
 	mutex_init(&p_hwfn->p_l2_info->lock);
@@ -126,8 +125,7 @@ void qed_l2_free(struct qed_hwfn *p_hwfn)
 {
 	u32 i;
 
-	if (p_hwfn->hw_info.personality != QED_PCI_ETH &&
-	    p_hwfn->hw_info.personality != QED_PCI_ETH_ROCE)
+	if (!QED_IS_L2_PERSONALITY(p_hwfn))
 		return;
 
 	if (!p_hwfn->p_l2_info)
-- 
cgit v1.1


From 090477e4acb31c5dd674940c7c01d4f16bd1ac41 Mon Sep 17 00:00:00 2001
From: Michal Kalderon <Michal.Kalderon@cavium.com>
Date: Tue, 8 May 2018 21:29:19 +0300
Subject: qede: Fix gfp flags sent to rdma event node allocation

A previous commit 4609adc27175 ("qede: Fix qedr link update")
added a flow that could allocate rdma event objects from an
interrupt path (link notification). Therefore the kzalloc call
should be done with GFP_ATOMIC.

fixes: 4609adc27175 ("qede: Fix qedr link update")
Signed-off-by: Michal Kalderon <Michal.Kalderon@cavium.com>
Signed-off-by: Sudarsana Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/qlogic/qede/qede_rdma.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/qlogic/qede/qede_rdma.c b/drivers/net/ethernet/qlogic/qede/qede_rdma.c
index 50b142f..1900bf7 100644
--- a/drivers/net/ethernet/qlogic/qede/qede_rdma.c
+++ b/drivers/net/ethernet/qlogic/qede/qede_rdma.c
@@ -238,7 +238,7 @@ qede_rdma_get_free_event_node(struct qede_dev *edev)
 	}
 
 	if (!found) {
-		event_node = kzalloc(sizeof(*event_node), GFP_KERNEL);
+		event_node = kzalloc(sizeof(*event_node), GFP_ATOMIC);
 		if (!event_node) {
 			DP_NOTICE(edev,
 				  "qedr: Could not allocate memory for rdma work\n");
-- 
cgit v1.1


From 9af5573f3ae1b5ad7340b2f77bba81b066a426d7 Mon Sep 17 00:00:00 2001
From: Colin Ian King <colin.king@canonical.com>
Date: Tue, 8 May 2018 23:01:51 +0100
Subject: firestream: fix spelling mistake: "reseverd" -> "reserved"

Trivial fix to spelling mistake in res_strings string array

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/atm/firestream.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/atm/firestream.c b/drivers/atm/firestream.c
index d97c056..4e46dc9 100644
--- a/drivers/atm/firestream.c
+++ b/drivers/atm/firestream.c
@@ -191,7 +191,7 @@ static char *res_strings[] = {
 	"reserved 37",
 	"reserved 38",
 	"reserved 39",
-	"reseverd 40",
+	"reserved 40",
 	"reserved 41", 
 	"reserved 42", 
 	"reserved 43", 
-- 
cgit v1.1


From 39a2d5cbaa8ce03bdd8eb9d3891d7a1b3f5fff96 Mon Sep 17 00:00:00 2001
From: Colin Ian King <colin.king@canonical.com>
Date: Tue, 8 May 2018 23:24:28 +0100
Subject: sctp: fix spelling mistake: "max_retans" -> "max_retrans"

Trivial fix to spelling mistake in error string

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/sctp/sm_make_chunk.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/sctp/sm_make_chunk.c b/net/sctp/sm_make_chunk.c
index 5a4fb1d..e62addb 100644
--- a/net/sctp/sm_make_chunk.c
+++ b/net/sctp/sm_make_chunk.c
@@ -1152,7 +1152,7 @@ struct sctp_chunk *sctp_make_violation_max_retrans(
 					const struct sctp_association *asoc,
 					const struct sctp_chunk *chunk)
 {
-	static const char error[] = "Association exceeded its max_retans count";
+	static const char error[] = "Association exceeded its max_retrans count";
 	size_t payload_len = sizeof(error) + sizeof(struct sctp_errhdr);
 	struct sctp_chunk *retval;
 
-- 
cgit v1.1


From 680a284635808594b8ac34939a7513dd4320e907 Mon Sep 17 00:00:00 2001
From: Colin Ian King <colin.king@canonical.com>
Date: Wed, 9 May 2018 10:48:33 +0100
Subject: net/9p: fix spelling mistake: "suspsend" -> "suspend"

Trivial fix to spelling mistake in dev_warn message text

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/9p/trans_xen.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/9p/trans_xen.c b/net/9p/trans_xen.c
index 086a4ab..0f19960 100644
--- a/net/9p/trans_xen.c
+++ b/net/9p/trans_xen.c
@@ -485,7 +485,7 @@ static int xen_9pfs_front_probe(struct xenbus_device *dev,
 
 static int xen_9pfs_front_resume(struct xenbus_device *dev)
 {
-	dev_warn(&dev->dev, "suspsend/resume unsupported\n");
+	dev_warn(&dev->dev, "suspend/resume unsupported\n");
 	return 0;
 }
 
-- 
cgit v1.1


From df13c59b54a9f8d3e20a2fb565d54eee8dac8768 Mon Sep 17 00:00:00 2001
From: Pieter Jansen van Vuuren <pieter.jansenvanvuuren@netronome.com>
Date: Wed, 9 May 2018 00:18:58 -0700
Subject: nfp: flower: remove headroom from max MTU calculation

Since commit 29a5dcae2790 ("nfp: flower: offload phys port MTU change") we
take encapsulation headroom into account when calculating the max allowed
MTU.  This is unnecessary as the max MTU advertised by firmware should have
already accounted for encap headroom.

Subtracting headroom twice brings the max MTU below what's necessary for
some deployments.

Fixes: 29a5dcae2790 ("nfp: flower: offload phys port MTU change")
Signed-off-by: Pieter Jansen van Vuuren <pieter.jansenvanvuuren@netronome.com>
Reviewed-by: John Hurley <john.hurley@netronome.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/netronome/nfp/flower/main.c | 19 -------------------
 1 file changed, 19 deletions(-)

diff --git a/drivers/net/ethernet/netronome/nfp/flower/main.c b/drivers/net/ethernet/netronome/nfp/flower/main.c
index a997e34..84e3b9f 100644
--- a/drivers/net/ethernet/netronome/nfp/flower/main.c
+++ b/drivers/net/ethernet/netronome/nfp/flower/main.c
@@ -52,8 +52,6 @@
 
 #define NFP_FLOWER_ALLOWED_VER 0x0001000000010000UL
 
-#define NFP_FLOWER_FRAME_HEADROOM	158
-
 static const char *nfp_flower_extra_cap(struct nfp_app *app, struct nfp_net *nn)
 {
 	return "FLOWER";
@@ -559,22 +557,6 @@ static void nfp_flower_clean(struct nfp_app *app)
 	app->priv = NULL;
 }
 
-static int
-nfp_flower_check_mtu(struct nfp_app *app, struct net_device *netdev,
-		     int new_mtu)
-{
-	/* The flower fw reserves NFP_FLOWER_FRAME_HEADROOM bytes of the
-	 * supported max MTU to allow for appending tunnel headers. To prevent
-	 * unexpected behaviour this needs to be accounted for.
-	 */
-	if (new_mtu > netdev->max_mtu - NFP_FLOWER_FRAME_HEADROOM) {
-		nfp_err(app->cpp, "New MTU (%d) is not valid\n", new_mtu);
-		return -EINVAL;
-	}
-
-	return 0;
-}
-
 static bool nfp_flower_check_ack(struct nfp_flower_priv *app_priv)
 {
 	bool ret;
@@ -656,7 +638,6 @@ const struct nfp_app_type app_flower = {
 	.init		= nfp_flower_init,
 	.clean		= nfp_flower_clean,
 
-	.check_mtu	= nfp_flower_check_mtu,
 	.repr_change_mtu  = nfp_flower_repr_change_mtu,
 
 	.vnic_alloc	= nfp_flower_vnic_alloc,
-- 
cgit v1.1


From 55be9f25be1ca5bda75c39808fc77e42691bc07f Mon Sep 17 00:00:00 2001
From: Mohammed Gamal <mgamal@redhat.com>
Date: Wed, 9 May 2018 10:17:34 +0200
Subject: hv_netvsc: Fix net device attach on older Windows hosts

On older windows hosts the net_device instance is returned to
the caller of rndis_filter_device_add() without having the presence
bit set first. This would cause any subsequent calls to network device
operations (e.g. MTU change, channel change) to fail after the device
is detached once, returning -ENODEV.

Instead of returning the device instabce, we take the exit path where
we call netif_device_attach()

Fixes: 7b2ee50c0cd5 ("hv_netvsc: common detach logic")
Signed-off-by: Mohammed Gamal <mgamal@redhat.com>
Reviewed-by: Stephen Hemminger <sthemmin@microsoft.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/hyperv/rndis_filter.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/hyperv/rndis_filter.c b/drivers/net/hyperv/rndis_filter.c
index 6b127be..e7ca5b5 100644
--- a/drivers/net/hyperv/rndis_filter.c
+++ b/drivers/net/hyperv/rndis_filter.c
@@ -1288,7 +1288,7 @@ struct netvsc_device *rndis_filter_device_add(struct hv_device *dev,
 		   rndis_device->link_state ? "down" : "up");
 
 	if (net_device->nvsp_version < NVSP_PROTOCOL_VERSION_5)
-		return net_device;
+		goto out;
 
 	rndis_filter_query_link_speed(rndis_device, net_device);
 
-- 
cgit v1.1


From 0e8411e426e277f55bd21e287ec89fab6f8eacae Mon Sep 17 00:00:00 2001
From: Hangbin Liu <liuhangbin@gmail.com>
Date: Wed, 9 May 2018 18:06:44 +0800
Subject: ipv4: reset fnhe_mtu_locked after cache route flushed

After route cache is flushed via ipv4_sysctl_rtcache_flush(), we forget
to reset fnhe_mtu_locked in rt_bind_exception(). When pmtu is updated
in __ip_rt_update_pmtu(), it will return directly since the pmtu is
still locked. e.g.

+ ip netns exec client ping 10.10.1.1 -c 1 -s 1400 -M do
PING 10.10.1.1 (10.10.1.1) 1400(1428) bytes of data.
>From 10.10.0.254 icmp_seq=1 Frag needed and DF set (mtu = 0)

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/ipv4/route.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/net/ipv4/route.c b/net/ipv4/route.c
index 1412a7b..29268ef 100644
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -1375,6 +1375,7 @@ static bool rt_bind_exception(struct rtable *rt, struct fib_nh_exception *fnhe,
 			fnhe->fnhe_gw = 0;
 			fnhe->fnhe_pmtu = 0;
 			fnhe->fnhe_expires = 0;
+			fnhe->fnhe_mtu_locked = false;
 			fnhe_flush_routes(fnhe);
 			orig = NULL;
 		}
-- 
cgit v1.1


From 69678bcd4d2dedbc3e8fcd6d7d99f283d83c531a Mon Sep 17 00:00:00 2001
From: Paolo Abeni <pabeni@redhat.com>
Date: Wed, 9 May 2018 12:42:34 +0200
Subject: udp: fix SO_BINDTODEVICE

Damir reported a breakage of SO_BINDTODEVICE for UDP sockets.
In absence of VRF devices, after commit fb74c27735f0 ("net:
ipv4: add second dif to udp socket lookups") the dif mismatch
isn't fatal anymore for UDP socket lookup with non null
sk_bound_dev_if, breaking SO_BINDTODEVICE semantics.

This changeset addresses the issue making the dif match mandatory
again in the above scenario.

Reported-by: Damir Mansurov <dnman@oktetlabs.ru>
Fixes: fb74c27735f0 ("net: ipv4: add second dif to udp socket lookups")
Fixes: 1801b570dd2a ("net: ipv6: add second dif to udp socket lookups")
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Acked-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/ipv4/udp.c | 4 ++--
 net/ipv6/udp.c | 4 ++--
 2 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
index 24b5c59..c2a292d 100644
--- a/net/ipv4/udp.c
+++ b/net/ipv4/udp.c
@@ -401,9 +401,9 @@ static int compute_score(struct sock *sk, struct net *net,
 		bool dev_match = (sk->sk_bound_dev_if == dif ||
 				  sk->sk_bound_dev_if == sdif);
 
-		if (exact_dif && !dev_match)
+		if (!dev_match)
 			return -1;
-		if (sk->sk_bound_dev_if && dev_match)
+		if (sk->sk_bound_dev_if)
 			score += 4;
 	}
 
diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
index 4ec76a8..ea07300 100644
--- a/net/ipv6/udp.c
+++ b/net/ipv6/udp.c
@@ -148,9 +148,9 @@ static int compute_score(struct sock *sk, struct net *net,
 		bool dev_match = (sk->sk_bound_dev_if == dif ||
 				  sk->sk_bound_dev_if == sdif);
 
-		if (exact_dif && !dev_match)
+		if (!dev_match)
 			return -1;
-		if (sk->sk_bound_dev_if && dev_match)
+		if (sk->sk_bound_dev_if)
 			score++;
 	}
 
-- 
cgit v1.1


From 2b928749f9ba781085442d1f8832f057cbf458a1 Mon Sep 17 00:00:00 2001
From: Ganesh Goudar <ganeshgr@chelsio.com>
Date: Wed, 9 May 2018 18:40:09 +0530
Subject: cxgb4: zero the HMA memory

firmware expects HMA memory to be zeroed, use __GFP_ZERO
for HMA memory allocation.

Fixes: 8b4e6b3ca2ed ("cxgb4: Add HMA support")
Signed-off-by: Ganesh Goudar <ganeshgr@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
index 24d2865..c3ae575 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
@@ -3433,8 +3433,8 @@ static int adap_config_hma(struct adapter *adapter)
 	sgl = adapter->hma.sgt->sgl;
 	node = dev_to_node(adapter->pdev_dev);
 	for_each_sg(sgl, iter, sgt->orig_nents, i) {
-		newpage = alloc_pages_node(node, __GFP_NOWARN | GFP_KERNEL,
-					   page_order);
+		newpage = alloc_pages_node(node, __GFP_NOWARN | GFP_KERNEL |
+					   __GFP_ZERO, page_order);
 		if (!newpage) {
 			dev_err(adapter->pdev_dev,
 				"Not enough memory for HMA page allocation\n");
-- 
cgit v1.1


From aca06eafd09f48ca4d97f3c0b2a12c8d631116f0 Mon Sep 17 00:00:00 2001
From: Ganesh Goudar <ganeshgr@chelsio.com>
Date: Wed, 9 May 2018 19:00:35 +0530
Subject: cxgb4: copy mbox log size to PF0-3 adap instances

copy mbox size to adapter instances of PF0-3 to avoid
mbox log overflow. This fixes the possible protection
fault.

Fixes: baf5086840ab ("cxgb4: restructure VF mgmt code")
Signed-off-by: Casey Leedom <leedom@chelsio.com>
Signed-off-by: Ganesh Goudar <ganeshgr@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
index c3ae575..005283c 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
@@ -5474,6 +5474,7 @@ static int init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 	}
 	spin_lock_init(&adapter->mbox_lock);
 	INIT_LIST_HEAD(&adapter->mlist.list);
+	adapter->mbox_log->size = T4_OS_LOG_MBOX_CMDS;
 	pci_set_drvdata(pdev, adapter);
 
 	if (func != ent->driver_data) {
@@ -5508,8 +5509,6 @@ static int init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 		goto out_free_adapter;
 	}
 
-	adapter->mbox_log->size = T4_OS_LOG_MBOX_CMDS;
-
 	/* PCI device has been enabled */
 	adapter->flags |= DEV_ENABLED;
 	memset(adapter->chan_map, 0xff, sizeof(adapter->chan_map));
-- 
cgit v1.1


From 6ad4e91c6d796b38a7f0e724db1de28eeb122bad Mon Sep 17 00:00:00 2001
From: Moshe Shemesh <moshe@mellanox.com>
Date: Wed, 9 May 2018 18:35:13 +0300
Subject: net/mlx4_en: Verify coalescing parameters are in range

Add check of coalescing parameters received through ethtool are within
range of values supported by the HW.
Driver gets the coalescing rx/tx-usecs and rx/tx-frames as set by the
users through ethtool. The ethtool support up to 32 bit value for each.
However, mlx4 modify cq limits the coalescing time parameter and
coalescing frames parameters to 16 bits.
Return out of range error if user tries to set these parameters to
higher values.
Change type of sample-interval and adaptive_rx_coal parameters in mlx4
driver to u32 as the ethtool holds them as u32 and these parameters are
not limited due to mlx4 HW.

Fixes: c27a02cd94d6 ('mlx4_en: Add driver for Mellanox ConnectX 10GbE NIC')
Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/mellanox/mlx4/en_ethtool.c | 16 ++++++++++++++++
 drivers/net/ethernet/mellanox/mlx4/mlx4_en.h    |  7 +++++--
 2 files changed, 21 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c b/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c
index a30a2e9..f11b450 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c
@@ -1027,6 +1027,22 @@ static int mlx4_en_set_coalesce(struct net_device *dev,
 	if (!coal->tx_max_coalesced_frames_irq)
 		return -EINVAL;
 
+	if (coal->tx_coalesce_usecs > MLX4_EN_MAX_COAL_TIME ||
+	    coal->rx_coalesce_usecs > MLX4_EN_MAX_COAL_TIME ||
+	    coal->rx_coalesce_usecs_low > MLX4_EN_MAX_COAL_TIME ||
+	    coal->rx_coalesce_usecs_high > MLX4_EN_MAX_COAL_TIME) {
+		netdev_info(dev, "%s: maximum coalesce time supported is %d usecs\n",
+			    __func__, MLX4_EN_MAX_COAL_TIME);
+		return -ERANGE;
+	}
+
+	if (coal->tx_max_coalesced_frames > MLX4_EN_MAX_COAL_PKTS ||
+	    coal->rx_max_coalesced_frames > MLX4_EN_MAX_COAL_PKTS) {
+		netdev_info(dev, "%s: maximum coalesced frames supported is %d\n",
+			    __func__, MLX4_EN_MAX_COAL_PKTS);
+		return -ERANGE;
+	}
+
 	priv->rx_frames = (coal->rx_max_coalesced_frames ==
 			   MLX4_EN_AUTO_CONF) ?
 				MLX4_EN_RX_COAL_TARGET :
diff --git a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
index f7c8113..ace6545 100644
--- a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
+++ b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
@@ -132,6 +132,9 @@
 #define MLX4_EN_TX_COAL_PKTS	16
 #define MLX4_EN_TX_COAL_TIME	0x10
 
+#define MLX4_EN_MAX_COAL_PKTS	U16_MAX
+#define MLX4_EN_MAX_COAL_TIME	U16_MAX
+
 #define MLX4_EN_RX_RATE_LOW		400000
 #define MLX4_EN_RX_COAL_TIME_LOW	0
 #define MLX4_EN_RX_RATE_HIGH		450000
@@ -552,8 +555,8 @@ struct mlx4_en_priv {
 	u16 rx_usecs_low;
 	u32 pkt_rate_high;
 	u16 rx_usecs_high;
-	u16 sample_interval;
-	u16 adaptive_rx_coal;
+	u32 sample_interval;
+	u32 adaptive_rx_coal;
 	u32 msg_enable;
 	u32 loopback_ok;
 	u32 validate_loopback;
-- 
cgit v1.1


From f7017cafcdd7574680fc7faabcb73f91172a14ab Mon Sep 17 00:00:00 2001
From: Davide Caratti <dcaratti@redhat.com>
Date: Wed, 9 May 2018 18:45:42 +0200
Subject: tc-testing: fix tdc tests for 'bpf' action

- correct a typo in the value of 'matchPattern' of test 282d, potentially
 causing false negative
- allow errors when 'teardown' executes '$TC action flush action bpf' in
 test 282d, to fix false positive when it is run with act_bpf unloaded
- correct the value of 'matchPattern' in test e939, causing false positive
 in case the BPF JIT is enabled

Fixes: 440ea4ae1828 ("tc-testing: add selftests for 'bpf' action")
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Acked-by: Lucas Bates <lucasb@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 tools/testing/selftests/tc-testing/tc-tests/actions/bpf.json | 11 ++++++++---
 1 file changed, 8 insertions(+), 3 deletions(-)

diff --git a/tools/testing/selftests/tc-testing/tc-tests/actions/bpf.json b/tools/testing/selftests/tc-testing/tc-tests/actions/bpf.json
index 5b012f4..6f289a4 100644
--- a/tools/testing/selftests/tc-testing/tc-tests/actions/bpf.json
+++ b/tools/testing/selftests/tc-testing/tc-tests/actions/bpf.json
@@ -66,7 +66,7 @@
         "cmdUnderTest": "$TC action add action bpf object-file _b.o index 667",
         "expExitCode": "0",
         "verifyCmd": "$TC action get action bpf index 667",
-        "matchPattern": "action order [0-9]*: bpf _b.o:\\[action\\] id [0-9]* tag 3b185187f1855c4c default-action pipe.*index 667 ref",
+        "matchPattern": "action order [0-9]*: bpf _b.o:\\[action\\] id [0-9]* tag 3b185187f1855c4c( jited)? default-action pipe.*index 667 ref",
         "matchCount": "1",
         "teardown": [
             "$TC action flush action bpf",
@@ -92,10 +92,15 @@
         "cmdUnderTest": "$TC action add action bpf object-file _c.o index 667",
         "expExitCode": "255",
         "verifyCmd": "$TC action get action bpf index 667",
-        "matchPattern": "action order [0-9]*: bpf _b.o:\\[action\\] id [0-9].*index 667 ref",
+        "matchPattern": "action order [0-9]*: bpf _c.o:\\[action\\] id [0-9].*index 667 ref",
         "matchCount": "0",
         "teardown": [
-            "$TC action flush action bpf",
+            [
+                "$TC action flush action bpf",
+                0,
+                1,
+                255
+            ],
             "rm -f _c.o"
         ]
     },
-- 
cgit v1.1


From 09c8b9718a7af674036643fa2e0dbb2f09aba75e Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Wed, 9 May 2018 09:50:22 -0700
Subject: tipc: fix one byte leak in tipc_sk_set_orig_addr()

sysbot/KMSAN reported an uninit-value in recvmsg() that
I tracked down to tipc_sk_set_orig_addr(), missing
srcaddr->member.scope initialization.

This patches moves srcaddr->sock.scope init to follow
fields order and ease future verifications.

BUG: KMSAN: uninit-value in copy_to_user include/linux/uaccess.h:184 [inline]
BUG: KMSAN: uninit-value in move_addr_to_user+0x32e/0x530 net/socket.c:226
CPU: 0 PID: 4549 Comm: syz-executor287 Not tainted 4.17.0-rc3+ #88
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Call Trace:
 __dump_stack lib/dump_stack.c:77 [inline]
 dump_stack+0x185/0x1d0 lib/dump_stack.c:113
 kmsan_report+0x142/0x240 mm/kmsan/kmsan.c:1067
 kmsan_internal_check_memory+0x135/0x1e0 mm/kmsan/kmsan.c:1157
 kmsan_copy_to_user+0x69/0x160 mm/kmsan/kmsan.c:1199
 copy_to_user include/linux/uaccess.h:184 [inline]
 move_addr_to_user+0x32e/0x530 net/socket.c:226
 ___sys_recvmsg+0x4e2/0x810 net/socket.c:2285
 __sys_recvmsg net/socket.c:2328 [inline]
 __do_sys_recvmsg net/socket.c:2338 [inline]
 __se_sys_recvmsg net/socket.c:2335 [inline]
 __x64_sys_recvmsg+0x325/0x460 net/socket.c:2335
 do_syscall_64+0x154/0x220 arch/x86/entry/common.c:287
 entry_SYSCALL_64_after_hwframe+0x44/0xa9
RIP: 0033:0x4455e9
RSP: 002b:00007fe3bd36ddb8 EFLAGS: 00000246 ORIG_RAX: 000000000000002f
RAX: ffffffffffffffda RBX: 00000000006dac24 RCX: 00000000004455e9
RDX: 0000000000002002 RSI: 0000000020000400 RDI: 0000000000000003
RBP: 00000000006dac20 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
R13: 00007fff98ce4b6f R14: 00007fe3bd36e9c0 R15: 0000000000000003

Local variable description: ----addr@___sys_recvmsg
Variable was created at:
 ___sys_recvmsg+0xd5/0x810 net/socket.c:2246
 __sys_recvmsg net/socket.c:2328 [inline]
 __do_sys_recvmsg net/socket.c:2338 [inline]
 __se_sys_recvmsg net/socket.c:2335 [inline]
 __x64_sys_recvmsg+0x325/0x460 net/socket.c:2335

Byte 19 of 32 is uninitialized

Fixes: 31c82a2d9d51 ("tipc: add second source address to recvmsg()/recvfrom()")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: syzbot <syzkaller@googlegroups.com>
Cc: Jon Maloy <jon.maloy@ericsson.com>
Cc: Ying Xue <ying.xue@windriver.com>
Acked-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/tipc/socket.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/net/tipc/socket.c b/net/tipc/socket.c
index 252a52ae..6be2157 100644
--- a/net/tipc/socket.c
+++ b/net/tipc/socket.c
@@ -1516,10 +1516,10 @@ static void tipc_sk_set_orig_addr(struct msghdr *m, struct sk_buff *skb)
 
 	srcaddr->sock.family = AF_TIPC;
 	srcaddr->sock.addrtype = TIPC_ADDR_ID;
+	srcaddr->sock.scope = 0;
 	srcaddr->sock.addr.id.ref = msg_origport(hdr);
 	srcaddr->sock.addr.id.node = msg_orignode(hdr);
 	srcaddr->sock.addr.name.domain = 0;
-	srcaddr->sock.scope = 0;
 	m->msg_namelen = sizeof(struct sockaddr_tipc);
 
 	if (!msg_in_group(hdr))
@@ -1528,6 +1528,7 @@ static void tipc_sk_set_orig_addr(struct msghdr *m, struct sk_buff *skb)
 	/* Group message users may also want to know sending member's id */
 	srcaddr->member.family = AF_TIPC;
 	srcaddr->member.addrtype = TIPC_ADDR_NAME;
+	srcaddr->member.scope = 0;
 	srcaddr->member.addr.name.name.type = msg_nametype(hdr);
 	srcaddr->member.addr.name.name.instance = TIPC_SKB_CB(skb)->orig_member;
 	srcaddr->member.addr.name.domain = 0;
-- 
cgit v1.1


From 97f3efb64323beb0690576e9d74e94998ad6e82a Mon Sep 17 00:00:00 2001
From: Stephen Hemminger <stephen@networkplumber.org>
Date: Wed, 9 May 2018 14:09:04 -0700
Subject: hv_netvsc: set master device

The hyper-v transparent bonding should have used master_dev_link.
The netvsc device should look like a master bond device not
like the upper side of a tunnel.

This makes the semantics the same so that userspace applications
looking at network devices see the correct master relationshipship.

Fixes: 0c195567a8f6 ("netvsc: transparent VF management")
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/hyperv/netvsc_drv.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/net/hyperv/netvsc_drv.c b/drivers/net/hyperv/netvsc_drv.c
index ecc8495..da07ccd 100644
--- a/drivers/net/hyperv/netvsc_drv.c
+++ b/drivers/net/hyperv/netvsc_drv.c
@@ -1840,7 +1840,8 @@ static int netvsc_vf_join(struct net_device *vf_netdev,
 		goto rx_handler_failed;
 	}
 
-	ret = netdev_upper_dev_link(vf_netdev, ndev, NULL);
+	ret = netdev_master_upper_dev_link(vf_netdev, ndev,
+					   NULL, NULL, NULL);
 	if (ret != 0) {
 		netdev_err(vf_netdev,
 			   "can not set master device %s (err = %d)\n",
-- 
cgit v1.1


From a577d868b768a3baf16cdd4841ab8cfb165521d6 Mon Sep 17 00:00:00 2001
From: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Date: Thu, 10 May 2018 09:06:04 +0200
Subject: net/mlx4_en: Fix an error handling path in 'mlx4_en_init_netdev()'

If an error occurs, 'mlx4_en_destroy_netdev()' is called.
It then calls 'mlx4_en_free_resources()' which does the needed resources
cleanup.

So, doing some explicit kfree in the error handling path would lead to
some double kfree.

Simplify code to avoid such a case.

Fixes: 67f8b1dcb9ee ("net/mlx4_en: Refactor the XDP forwarding rings scheme")
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Reviewed-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/mellanox/mlx4/en_netdev.c | 8 +-------
 1 file changed, 1 insertion(+), 7 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
index e0adac4..9670b33 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
@@ -3324,12 +3324,11 @@ int mlx4_en_init_netdev(struct mlx4_en_dev *mdev, int port,
 					   MAX_TX_RINGS, GFP_KERNEL);
 		if (!priv->tx_ring[t]) {
 			err = -ENOMEM;
-			goto err_free_tx;
+			goto out;
 		}
 		priv->tx_cq[t] = kzalloc(sizeof(struct mlx4_en_cq *) *
 					 MAX_TX_RINGS, GFP_KERNEL);
 		if (!priv->tx_cq[t]) {
-			kfree(priv->tx_ring[t]);
 			err = -ENOMEM;
 			goto out;
 		}
@@ -3582,11 +3581,6 @@ int mlx4_en_init_netdev(struct mlx4_en_dev *mdev, int port,
 
 	return 0;
 
-err_free_tx:
-	while (t--) {
-		kfree(priv->tx_ring[t]);
-		kfree(priv->tx_cq[t]);
-	}
 out:
 	mlx4_en_destroy_netdev(dev);
 	return err;
-- 
cgit v1.1


From 6910e25de2257e2c82c7a2d126e3463cd8e50810 Mon Sep 17 00:00:00 2001
From: Xin Long <lucien.xin@gmail.com>
Date: Thu, 10 May 2018 17:34:13 +0800
Subject: sctp: remove sctp_chunk_put from fail_mark err path in
 sctp_ulpevent_make_rcvmsg

In Commit 1f45f78f8e51 ("sctp: allow GSO frags to access the chunk too"),
it held the chunk in sctp_ulpevent_make_rcvmsg to access it safely later
in recvmsg. However, it also added sctp_chunk_put in fail_mark err path,
which is only triggered before holding the chunk.

syzbot reported a use-after-free crash happened on this err path, where
it shouldn't call sctp_chunk_put.

This patch simply removes this call.

Fixes: 1f45f78f8e51 ("sctp: allow GSO frags to access the chunk too")
Reported-by: syzbot+141d898c5f24489db4aa@syzkaller.appspotmail.com
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/sctp/ulpevent.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/net/sctp/ulpevent.c b/net/sctp/ulpevent.c
index 84207ad..8cb7d98 100644
--- a/net/sctp/ulpevent.c
+++ b/net/sctp/ulpevent.c
@@ -715,7 +715,6 @@ struct sctp_ulpevent *sctp_ulpevent_make_rcvmsg(struct sctp_association *asoc,
 	return event;
 
 fail_mark:
-	sctp_chunk_put(chunk);
 	kfree_skb(skb);
 fail:
 	return NULL;
-- 
cgit v1.1


From 2fdae0349f587ffde4d1010c65fb543569bab787 Mon Sep 17 00:00:00 2001
From: Colin Ian King <colin.king@canonical.com>
Date: Thu, 10 May 2018 15:03:27 +0100
Subject: qed: fix spelling mistake: "taskelt" -> "tasklet"

Trivial fix to spelling mistake in DP_VERBOSE message text

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/qlogic/qed/qed_main.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/qlogic/qed/qed_main.c b/drivers/net/ethernet/qlogic/qed/qed_main.c
index 9854aa9..7870ae2 100644
--- a/drivers/net/ethernet/qlogic/qed/qed_main.c
+++ b/drivers/net/ethernet/qlogic/qed/qed_main.c
@@ -680,7 +680,7 @@ static int qed_nic_stop(struct qed_dev *cdev)
 			tasklet_disable(p_hwfn->sp_dpc);
 			p_hwfn->b_sp_dpc_enabled = false;
 			DP_VERBOSE(cdev, NETIF_MSG_IFDOWN,
-				   "Disabled sp taskelt [hwfn %d] at %p\n",
+				   "Disabled sp tasklet [hwfn %d] at %p\n",
 				   i, p_hwfn->sp_dpc);
 		}
 	}
-- 
cgit v1.1


From c54e43d752c7187595c8c62a231e0b0d53c7fded Mon Sep 17 00:00:00 2001
From: David Howells <dhowells@redhat.com>
Date: Thu, 10 May 2018 23:26:00 +0100
Subject: rxrpc: Fix missing start of call timeout

The expect_rx_by call timeout is supposed to be set when a call is started
to indicate that we need to receive a packet by that point.  This is
currently put back every time we receive a packet, but it isn't started
when we first send a packet.  Without this, the call may wait forever if
the server doesn't deign to reply.

Fix this by setting the timeout upon a successful UDP sendmsg call for the
first DATA packet.  The timeout is initiated only for initial transmission
and not for subsequent retries as we don't want the retry mechanism to
extend the timeout indefinitely.

Fixes: a158bdd3247b ("rxrpc: Fix call timeouts")
Reported-by: Marc Dionne <marc.dionne@auristor.com>
Signed-off-by: David Howells <dhowells@redhat.com>
---
 net/rxrpc/ar-internal.h |  1 +
 net/rxrpc/input.c       |  2 +-
 net/rxrpc/output.c      | 11 +++++++++++
 net/rxrpc/sendmsg.c     | 10 ++++++++++
 4 files changed, 23 insertions(+), 1 deletion(-)

diff --git a/net/rxrpc/ar-internal.h b/net/rxrpc/ar-internal.h
index 90d7079..19975d2 100644
--- a/net/rxrpc/ar-internal.h
+++ b/net/rxrpc/ar-internal.h
@@ -476,6 +476,7 @@ enum rxrpc_call_flag {
 	RXRPC_CALL_SEND_PING,		/* A ping will need to be sent */
 	RXRPC_CALL_PINGING,		/* Ping in process */
 	RXRPC_CALL_RETRANS_TIMEOUT,	/* Retransmission due to timeout occurred */
+	RXRPC_CALL_BEGAN_RX_TIMER,	/* We began the expect_rx_by timer */
 };
 
 /*
diff --git a/net/rxrpc/input.c b/net/rxrpc/input.c
index 0410d22..b5fd638 100644
--- a/net/rxrpc/input.c
+++ b/net/rxrpc/input.c
@@ -971,7 +971,7 @@ static void rxrpc_input_call_packet(struct rxrpc_call *call,
 	if (timo) {
 		unsigned long now = jiffies, expect_rx_by;
 
-		expect_rx_by = jiffies + timo;
+		expect_rx_by = now + timo;
 		WRITE_ONCE(call->expect_rx_by, expect_rx_by);
 		rxrpc_reduce_call_timer(call, expect_rx_by, now,
 					rxrpc_timer_set_for_normal);
diff --git a/net/rxrpc/output.c b/net/rxrpc/output.c
index 7f1fc04..6b9d27f 100644
--- a/net/rxrpc/output.c
+++ b/net/rxrpc/output.c
@@ -414,6 +414,17 @@ done:
 							rxrpc_timer_set_for_lost_ack);
 			}
 		}
+
+		if (sp->hdr.seq == 1 &&
+		    !test_and_set_bit(RXRPC_CALL_BEGAN_RX_TIMER,
+				      &call->flags)) {
+			unsigned long nowj = jiffies, expect_rx_by;
+
+			expect_rx_by = nowj + call->next_rx_timo;
+			WRITE_ONCE(call->expect_rx_by, expect_rx_by);
+			rxrpc_reduce_call_timer(call, expect_rx_by, nowj,
+						rxrpc_timer_set_for_normal);
+		}
 	}
 
 	rxrpc_set_keepalive(call);
diff --git a/net/rxrpc/sendmsg.c b/net/rxrpc/sendmsg.c
index 206e802..be01f9c 100644
--- a/net/rxrpc/sendmsg.c
+++ b/net/rxrpc/sendmsg.c
@@ -223,6 +223,15 @@ static void rxrpc_queue_packet(struct rxrpc_sock *rx, struct rxrpc_call *call,
 
 	ret = rxrpc_send_data_packet(call, skb, false);
 	if (ret < 0) {
+		switch (ret) {
+		case -ENETUNREACH:
+		case -EHOSTUNREACH:
+		case -ECONNREFUSED:
+			rxrpc_set_call_completion(call,
+						  RXRPC_CALL_LOCAL_ERROR,
+						  0, ret);
+			goto out;
+		}
 		_debug("need instant resend %d", ret);
 		rxrpc_instant_resend(call, ix);
 	} else {
@@ -241,6 +250,7 @@ static void rxrpc_queue_packet(struct rxrpc_sock *rx, struct rxrpc_call *call,
 					rxrpc_timer_set_for_send);
 	}
 
+out:
 	rxrpc_free_skb(skb, rxrpc_skb_tx_freed);
 	_leave("");
 }
-- 
cgit v1.1


From f2aeed3a591ff29a82495eeaa92ac4780bad7487 Mon Sep 17 00:00:00 2001
From: David Howells <dhowells@redhat.com>
Date: Thu, 10 May 2018 23:26:00 +0100
Subject: rxrpc: Fix error reception on AF_INET6 sockets

AF_RXRPC tries to turn on IP_RECVERR and IP_MTU_DISCOVER on the UDP socket
it just opened for communications with the outside world, regardless of the
type of socket.  Unfortunately, this doesn't work with an AF_INET6 socket.

Fix this by turning on IPV6_RECVERR and IPV6_MTU_DISCOVER instead if the
socket is of the AF_INET6 family.

Without this, kAFS server and address rotation doesn't work correctly
because the algorithm doesn't detect received network errors.

Fixes: 75b54cb57ca3 ("rxrpc: Add IPv6 support")
Signed-off-by: David Howells <dhowells@redhat.com>
---
 net/rxrpc/local_object.c | 57 +++++++++++++++++++++++++++++++++++-------------
 1 file changed, 42 insertions(+), 15 deletions(-)

diff --git a/net/rxrpc/local_object.c b/net/rxrpc/local_object.c
index 8b54e95..b493e6b 100644
--- a/net/rxrpc/local_object.c
+++ b/net/rxrpc/local_object.c
@@ -134,22 +134,49 @@ static int rxrpc_open_socket(struct rxrpc_local *local, struct net *net)
 		}
 	}
 
-	/* we want to receive ICMP errors */
-	opt = 1;
-	ret = kernel_setsockopt(local->socket, SOL_IP, IP_RECVERR,
-				(char *) &opt, sizeof(opt));
-	if (ret < 0) {
-		_debug("setsockopt failed");
-		goto error;
-	}
+	switch (local->srx.transport.family) {
+	case AF_INET:
+		/* we want to receive ICMP errors */
+		opt = 1;
+		ret = kernel_setsockopt(local->socket, SOL_IP, IP_RECVERR,
+					(char *) &opt, sizeof(opt));
+		if (ret < 0) {
+			_debug("setsockopt failed");
+			goto error;
+		}
 
-	/* we want to set the don't fragment bit */
-	opt = IP_PMTUDISC_DO;
-	ret = kernel_setsockopt(local->socket, SOL_IP, IP_MTU_DISCOVER,
-				(char *) &opt, sizeof(opt));
-	if (ret < 0) {
-		_debug("setsockopt failed");
-		goto error;
+		/* we want to set the don't fragment bit */
+		opt = IP_PMTUDISC_DO;
+		ret = kernel_setsockopt(local->socket, SOL_IP, IP_MTU_DISCOVER,
+					(char *) &opt, sizeof(opt));
+		if (ret < 0) {
+			_debug("setsockopt failed");
+			goto error;
+		}
+		break;
+
+	case AF_INET6:
+		/* we want to receive ICMP errors */
+		opt = 1;
+		ret = kernel_setsockopt(local->socket, SOL_IPV6, IPV6_RECVERR,
+					(char *) &opt, sizeof(opt));
+		if (ret < 0) {
+			_debug("setsockopt failed");
+			goto error;
+		}
+
+		/* we want to set the don't fragment bit */
+		opt = IPV6_PMTUDISC_DO;
+		ret = kernel_setsockopt(local->socket, SOL_IPV6, IPV6_MTU_DISCOVER,
+					(char *) &opt, sizeof(opt));
+		if (ret < 0) {
+			_debug("setsockopt failed");
+			goto error;
+		}
+		break;
+
+	default:
+		BUG();
 	}
 
 	/* set the socket up */
-- 
cgit v1.1


From 93864fc3ffcc4bf70e96cfb5cc6e941630419ad0 Mon Sep 17 00:00:00 2001
From: David Howells <dhowells@redhat.com>
Date: Thu, 10 May 2018 23:26:01 +0100
Subject: rxrpc: Fix the min security level for kernel calls

Fix the kernel call initiation to set the minimum security level for kernel
initiated calls (such as from kAFS) from the sockopt value.

Fixes: 19ffa01c9c45 ("rxrpc: Use structs to hold connection params and protocol info")
Signed-off-by: David Howells <dhowells@redhat.com>
---
 net/rxrpc/af_rxrpc.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/rxrpc/af_rxrpc.c b/net/rxrpc/af_rxrpc.c
index 9a2c8e7..2b46304 100644
--- a/net/rxrpc/af_rxrpc.c
+++ b/net/rxrpc/af_rxrpc.c
@@ -313,7 +313,7 @@ struct rxrpc_call *rxrpc_kernel_begin_call(struct socket *sock,
 	memset(&cp, 0, sizeof(cp));
 	cp.local		= rx->local;
 	cp.key			= key;
-	cp.security_level	= 0;
+	cp.security_level	= rx->min_sec_level;
 	cp.exclusive		= false;
 	cp.upgrade		= upgrade;
 	cp.service_id		= srx->srx_service;
-- 
cgit v1.1


From 494337c918403db5743af0348d963cafb9b20cee Mon Sep 17 00:00:00 2001
From: David Howells <dhowells@redhat.com>
Date: Thu, 10 May 2018 23:26:01 +0100
Subject: rxrpc: Add a tracepoint to log ICMP/ICMP6 and error messages

Add a tracepoint to log received ICMP/ICMP6 events and other error
messages.

Signed-off-by: David Howells <dhowells@redhat.com>
---
 include/trace/events/rxrpc.h | 30 +++++++++++++++++++++++++++++
 net/rxrpc/peer_event.c       | 46 ++++++++++++++++++++++----------------------
 2 files changed, 53 insertions(+), 23 deletions(-)

diff --git a/include/trace/events/rxrpc.h b/include/trace/events/rxrpc.h
index 9e96c2f..497d0b6 100644
--- a/include/trace/events/rxrpc.h
+++ b/include/trace/events/rxrpc.h
@@ -15,6 +15,7 @@
 #define _TRACE_RXRPC_H
 
 #include <linux/tracepoint.h>
+#include <linux/errqueue.h>
 
 /*
  * Define enums for tracing information.
@@ -1374,6 +1375,35 @@ TRACE_EVENT(rxrpc_resend,
 		      __entry->anno)
 	    );
 
+TRACE_EVENT(rxrpc_rx_icmp,
+	    TP_PROTO(struct rxrpc_peer *peer, struct sock_extended_err *ee,
+		     struct sockaddr_rxrpc *srx),
+
+	    TP_ARGS(peer, ee, srx),
+
+	    TP_STRUCT__entry(
+		    __field(unsigned int,			peer	)
+		    __field_struct(struct sock_extended_err,	ee	)
+		    __field_struct(struct sockaddr_rxrpc,	srx	)
+			     ),
+
+	    TP_fast_assign(
+		    __entry->peer = peer->debug_id;
+		    memcpy(&__entry->ee, ee, sizeof(__entry->ee));
+		    memcpy(&__entry->srx, srx, sizeof(__entry->srx));
+			   ),
+
+	    TP_printk("P=%08x o=%u t=%u c=%u i=%u d=%u e=%d %pISp",
+		      __entry->peer,
+		      __entry->ee.ee_origin,
+		      __entry->ee.ee_type,
+		      __entry->ee.ee_code,
+		      __entry->ee.ee_info,
+		      __entry->ee.ee_data,
+		      __entry->ee.ee_errno,
+		      &__entry->srx.transport)
+	    );
+
 #endif /* _TRACE_RXRPC_H */
 
 /* This part must be outside protection */
diff --git a/net/rxrpc/peer_event.c b/net/rxrpc/peer_event.c
index 78c2f95..0ed8b65 100644
--- a/net/rxrpc/peer_event.c
+++ b/net/rxrpc/peer_event.c
@@ -28,39 +28,39 @@ static void rxrpc_store_error(struct rxrpc_peer *, struct sock_exterr_skb *);
  * Find the peer associated with an ICMP packet.
  */
 static struct rxrpc_peer *rxrpc_lookup_peer_icmp_rcu(struct rxrpc_local *local,
-						     const struct sk_buff *skb)
+						     const struct sk_buff *skb,
+						     struct sockaddr_rxrpc *srx)
 {
 	struct sock_exterr_skb *serr = SKB_EXT_ERR(skb);
-	struct sockaddr_rxrpc srx;
 
 	_enter("");
 
-	memset(&srx, 0, sizeof(srx));
-	srx.transport_type = local->srx.transport_type;
-	srx.transport_len = local->srx.transport_len;
-	srx.transport.family = local->srx.transport.family;
+	memset(srx, 0, sizeof(*srx));
+	srx->transport_type = local->srx.transport_type;
+	srx->transport_len = local->srx.transport_len;
+	srx->transport.family = local->srx.transport.family;
 
 	/* Can we see an ICMP4 packet on an ICMP6 listening socket?  and vice
 	 * versa?
 	 */
-	switch (srx.transport.family) {
+	switch (srx->transport.family) {
 	case AF_INET:
-		srx.transport.sin.sin_port = serr->port;
+		srx->transport.sin.sin_port = serr->port;
 		switch (serr->ee.ee_origin) {
 		case SO_EE_ORIGIN_ICMP:
 			_net("Rx ICMP");
-			memcpy(&srx.transport.sin.sin_addr,
+			memcpy(&srx->transport.sin.sin_addr,
 			       skb_network_header(skb) + serr->addr_offset,
 			       sizeof(struct in_addr));
 			break;
 		case SO_EE_ORIGIN_ICMP6:
 			_net("Rx ICMP6 on v4 sock");
-			memcpy(&srx.transport.sin.sin_addr,
+			memcpy(&srx->transport.sin.sin_addr,
 			       skb_network_header(skb) + serr->addr_offset + 12,
 			       sizeof(struct in_addr));
 			break;
 		default:
-			memcpy(&srx.transport.sin.sin_addr, &ip_hdr(skb)->saddr,
+			memcpy(&srx->transport.sin.sin_addr, &ip_hdr(skb)->saddr,
 			       sizeof(struct in_addr));
 			break;
 		}
@@ -68,25 +68,25 @@ static struct rxrpc_peer *rxrpc_lookup_peer_icmp_rcu(struct rxrpc_local *local,
 
 #ifdef CONFIG_AF_RXRPC_IPV6
 	case AF_INET6:
-		srx.transport.sin6.sin6_port = serr->port;
+		srx->transport.sin6.sin6_port = serr->port;
 		switch (serr->ee.ee_origin) {
 		case SO_EE_ORIGIN_ICMP6:
 			_net("Rx ICMP6");
-			memcpy(&srx.transport.sin6.sin6_addr,
+			memcpy(&srx->transport.sin6.sin6_addr,
 			       skb_network_header(skb) + serr->addr_offset,
 			       sizeof(struct in6_addr));
 			break;
 		case SO_EE_ORIGIN_ICMP:
 			_net("Rx ICMP on v6 sock");
-			srx.transport.sin6.sin6_addr.s6_addr32[0] = 0;
-			srx.transport.sin6.sin6_addr.s6_addr32[1] = 0;
-			srx.transport.sin6.sin6_addr.s6_addr32[2] = htonl(0xffff);
-			memcpy(srx.transport.sin6.sin6_addr.s6_addr + 12,
+			srx->transport.sin6.sin6_addr.s6_addr32[0] = 0;
+			srx->transport.sin6.sin6_addr.s6_addr32[1] = 0;
+			srx->transport.sin6.sin6_addr.s6_addr32[2] = htonl(0xffff);
+			memcpy(srx->transport.sin6.sin6_addr.s6_addr + 12,
 			       skb_network_header(skb) + serr->addr_offset,
 			       sizeof(struct in_addr));
 			break;
 		default:
-			memcpy(&srx.transport.sin6.sin6_addr,
+			memcpy(&srx->transport.sin6.sin6_addr,
 			       &ipv6_hdr(skb)->saddr,
 			       sizeof(struct in6_addr));
 			break;
@@ -98,7 +98,7 @@ static struct rxrpc_peer *rxrpc_lookup_peer_icmp_rcu(struct rxrpc_local *local,
 		BUG();
 	}
 
-	return rxrpc_lookup_peer_rcu(local, &srx);
+	return rxrpc_lookup_peer_rcu(local, srx);
 }
 
 /*
@@ -146,6 +146,7 @@ static void rxrpc_adjust_mtu(struct rxrpc_peer *peer, struct sock_exterr_skb *se
 void rxrpc_error_report(struct sock *sk)
 {
 	struct sock_exterr_skb *serr;
+	struct sockaddr_rxrpc srx;
 	struct rxrpc_local *local = sk->sk_user_data;
 	struct rxrpc_peer *peer;
 	struct sk_buff *skb;
@@ -166,7 +167,7 @@ void rxrpc_error_report(struct sock *sk)
 	}
 
 	rcu_read_lock();
-	peer = rxrpc_lookup_peer_icmp_rcu(local, skb);
+	peer = rxrpc_lookup_peer_icmp_rcu(local, skb, &srx);
 	if (peer && !rxrpc_get_peer_maybe(peer))
 		peer = NULL;
 	if (!peer) {
@@ -176,6 +177,8 @@ void rxrpc_error_report(struct sock *sk)
 		return;
 	}
 
+	trace_rxrpc_rx_icmp(peer, &serr->ee, &srx);
+
 	if ((serr->ee.ee_origin == SO_EE_ORIGIN_ICMP &&
 	     serr->ee.ee_type == ICMP_DEST_UNREACH &&
 	     serr->ee.ee_code == ICMP_FRAG_NEEDED)) {
@@ -209,9 +212,6 @@ static void rxrpc_store_error(struct rxrpc_peer *peer,
 
 	ee = &serr->ee;
 
-	_net("Rx Error o=%d t=%d c=%d e=%d",
-	     ee->ee_origin, ee->ee_type, ee->ee_code, ee->ee_errno);
-
 	err = ee->ee_errno;
 
 	switch (ee->ee_origin) {
-- 
cgit v1.1


From 6b47fe1d1ca3aec3a1a8623439c22fbf51016cd8 Mon Sep 17 00:00:00 2001
From: David Howells <dhowells@redhat.com>
Date: Thu, 10 May 2018 23:26:01 +0100
Subject: rxrpc: Trace UDP transmission failure

Add a tracepoint to log transmission failure from the UDP transport socket
being used by AF_RXRPC.

Signed-off-by: David Howells <dhowells@redhat.com>
---
 include/trace/events/rxrpc.h | 55 ++++++++++++++++++++++++++++++++++++++++++++
 net/rxrpc/conn_event.c       | 11 ++++++---
 net/rxrpc/local_event.c      |  3 ++-
 net/rxrpc/output.c           | 23 ++++++++++++++++--
 net/rxrpc/rxkad.c            |  6 +++--
 5 files changed, 90 insertions(+), 8 deletions(-)

diff --git a/include/trace/events/rxrpc.h b/include/trace/events/rxrpc.h
index 497d0b6..077e664 100644
--- a/include/trace/events/rxrpc.h
+++ b/include/trace/events/rxrpc.h
@@ -211,6 +211,20 @@ enum rxrpc_congest_change {
 	rxrpc_cong_saw_nack,
 };
 
+enum rxrpc_tx_fail_trace {
+	rxrpc_tx_fail_call_abort,
+	rxrpc_tx_fail_call_ack,
+	rxrpc_tx_fail_call_data_frag,
+	rxrpc_tx_fail_call_data_nofrag,
+	rxrpc_tx_fail_call_final_resend,
+	rxrpc_tx_fail_conn_abort,
+	rxrpc_tx_fail_conn_challenge,
+	rxrpc_tx_fail_conn_response,
+	rxrpc_tx_fail_reject,
+	rxrpc_tx_fail_version_keepalive,
+	rxrpc_tx_fail_version_reply,
+};
+
 #endif /* end __RXRPC_DECLARE_TRACE_ENUMS_ONCE_ONLY */
 
 /*
@@ -438,6 +452,19 @@ enum rxrpc_congest_change {
 	EM(RXRPC_CALL_LOCAL_ERROR,		"LocalError") \
 	E_(RXRPC_CALL_NETWORK_ERROR,		"NetError")
 
+#define rxrpc_tx_fail_traces \
+	EM(rxrpc_tx_fail_call_abort,		"CallAbort") \
+	EM(rxrpc_tx_fail_call_ack,		"CallAck") \
+	EM(rxrpc_tx_fail_call_data_frag,	"CallDataFrag") \
+	EM(rxrpc_tx_fail_call_data_nofrag,	"CallDataNofrag") \
+	EM(rxrpc_tx_fail_call_final_resend,	"CallFinalResend") \
+	EM(rxrpc_tx_fail_conn_abort,		"ConnAbort") \
+	EM(rxrpc_tx_fail_conn_challenge,	"ConnChall") \
+	EM(rxrpc_tx_fail_conn_response,		"ConnResp") \
+	EM(rxrpc_tx_fail_reject,		"Reject") \
+	EM(rxrpc_tx_fail_version_keepalive,	"VerKeepalive") \
+	E_(rxrpc_tx_fail_version_reply,		"VerReply")
+
 /*
  * Export enum symbols via userspace.
  */
@@ -461,6 +488,7 @@ rxrpc_propose_ack_traces;
 rxrpc_propose_ack_outcomes;
 rxrpc_congest_modes;
 rxrpc_congest_changes;
+rxrpc_tx_fail_traces;
 
 /*
  * Now redefine the EM() and E_() macros to map the enums to the strings that
@@ -1404,6 +1432,33 @@ TRACE_EVENT(rxrpc_rx_icmp,
 		      &__entry->srx.transport)
 	    );
 
+TRACE_EVENT(rxrpc_tx_fail,
+	    TP_PROTO(unsigned int debug_id, rxrpc_serial_t serial, int ret,
+		     enum rxrpc_tx_fail_trace what),
+
+	    TP_ARGS(debug_id, serial, ret, what),
+
+	    TP_STRUCT__entry(
+		    __field(unsigned int,		debug_id	)
+		    __field(rxrpc_serial_t,		serial		)
+		    __field(int,			ret		)
+		    __field(enum rxrpc_tx_fail_trace,   what		)
+			     ),
+
+	    TP_fast_assign(
+		    __entry->debug_id = debug_id;
+		    __entry->serial = serial;
+		    __entry->ret = ret;
+		    __entry->what = what;
+			   ),
+
+	    TP_printk("c=%08x r=%x ret=%d %s",
+		      __entry->debug_id,
+		      __entry->serial,
+		      __entry->ret,
+		      __print_symbolic(__entry->what, rxrpc_tx_fail_traces))
+	    );
+
 #endif /* _TRACE_RXRPC_H */
 
 /* This part must be outside protection */
diff --git a/net/rxrpc/conn_event.c b/net/rxrpc/conn_event.c
index c717152..1350f1b 100644
--- a/net/rxrpc/conn_event.c
+++ b/net/rxrpc/conn_event.c
@@ -40,7 +40,7 @@ static void rxrpc_conn_retransmit_call(struct rxrpc_connection *conn,
 	} __attribute__((packed)) pkt;
 	struct rxrpc_ackinfo ack_info;
 	size_t len;
-	int ioc;
+	int ret, ioc;
 	u32 serial, mtu, call_id, padding;
 
 	_enter("%d", conn->debug_id);
@@ -135,10 +135,13 @@ static void rxrpc_conn_retransmit_call(struct rxrpc_connection *conn,
 		break;
 	}
 
-	kernel_sendmsg(conn->params.local->socket, &msg, iov, ioc, len);
+	ret = kernel_sendmsg(conn->params.local->socket, &msg, iov, ioc, len);
 	conn->params.peer->last_tx_at = ktime_get_real();
+	if (ret < 0)
+		trace_rxrpc_tx_fail(conn->debug_id, serial, ret,
+				    rxrpc_tx_fail_call_final_resend);
+
 	_leave("");
-	return;
 }
 
 /*
@@ -236,6 +239,8 @@ static int rxrpc_abort_connection(struct rxrpc_connection *conn,
 
 	ret = kernel_sendmsg(conn->params.local->socket, &msg, iov, 2, len);
 	if (ret < 0) {
+		trace_rxrpc_tx_fail(conn->debug_id, serial, ret,
+				    rxrpc_tx_fail_conn_abort);
 		_debug("sendmsg failed: %d", ret);
 		return -EAGAIN;
 	}
diff --git a/net/rxrpc/local_event.c b/net/rxrpc/local_event.c
index 93b5d91..8325f1b 100644
--- a/net/rxrpc/local_event.c
+++ b/net/rxrpc/local_event.c
@@ -71,7 +71,8 @@ static void rxrpc_send_version_request(struct rxrpc_local *local,
 
 	ret = kernel_sendmsg(local->socket, &msg, iov, 2, len);
 	if (ret < 0)
-		_debug("sendmsg failed: %d", ret);
+		trace_rxrpc_tx_fail(local->debug_id, 0, ret,
+				    rxrpc_tx_fail_version_reply);
 
 	_leave("");
 }
diff --git a/net/rxrpc/output.c b/net/rxrpc/output.c
index 6b9d27f..f03de1c 100644
--- a/net/rxrpc/output.c
+++ b/net/rxrpc/output.c
@@ -210,6 +210,9 @@ int rxrpc_send_ack_packet(struct rxrpc_call *call, bool ping,
 	if (ping)
 		call->ping_time = now;
 	conn->params.peer->last_tx_at = ktime_get_real();
+	if (ret < 0)
+		trace_rxrpc_tx_fail(call->debug_id, serial, ret,
+				    rxrpc_tx_fail_call_ack);
 
 	if (call->state < RXRPC_CALL_COMPLETE) {
 		if (ret < 0) {
@@ -294,6 +297,10 @@ int rxrpc_send_abort_packet(struct rxrpc_call *call)
 	ret = kernel_sendmsg(conn->params.local->socket,
 			     &msg, iov, 1, sizeof(pkt));
 	conn->params.peer->last_tx_at = ktime_get_real();
+	if (ret < 0)
+		trace_rxrpc_tx_fail(call->debug_id, serial, ret,
+				    rxrpc_tx_fail_call_abort);
+
 
 	rxrpc_put_connection(conn);
 	return ret;
@@ -387,6 +394,9 @@ int rxrpc_send_data_packet(struct rxrpc_call *call, struct sk_buff *skb,
 	conn->params.peer->last_tx_at = ktime_get_real();
 
 	up_read(&conn->params.local->defrag_sem);
+	if (ret < 0)
+		trace_rxrpc_tx_fail(call->debug_id, serial, ret,
+				    rxrpc_tx_fail_call_data_nofrag);
 	if (ret == -EMSGSIZE)
 		goto send_fragmentable;
 
@@ -476,6 +486,10 @@ send_fragmentable:
 #endif
 	}
 
+	if (ret < 0)
+		trace_rxrpc_tx_fail(call->debug_id, serial, ret,
+				    rxrpc_tx_fail_call_data_frag);
+
 	up_write(&conn->params.local->defrag_sem);
 	goto done;
 }
@@ -493,6 +507,7 @@ void rxrpc_reject_packets(struct rxrpc_local *local)
 	struct kvec iov[2];
 	size_t size;
 	__be32 code;
+	int ret;
 
 	_enter("%d", local->debug_id);
 
@@ -527,7 +542,10 @@ void rxrpc_reject_packets(struct rxrpc_local *local)
 			whdr.flags	^= RXRPC_CLIENT_INITIATED;
 			whdr.flags	&= RXRPC_CLIENT_INITIATED;
 
-			kernel_sendmsg(local->socket, &msg, iov, 2, size);
+			ret = kernel_sendmsg(local->socket, &msg, iov, 2, size);
+			if (ret < 0)
+				trace_rxrpc_tx_fail(local->debug_id, 0, ret,
+						    rxrpc_tx_fail_reject);
 		}
 
 		rxrpc_free_skb(skb, rxrpc_skb_rx_freed);
@@ -578,7 +596,8 @@ void rxrpc_send_keepalive(struct rxrpc_peer *peer)
 
 	ret = kernel_sendmsg(peer->local->socket, &msg, iov, 2, len);
 	if (ret < 0)
-		_debug("sendmsg failed: %d", ret);
+		trace_rxrpc_tx_fail(peer->debug_id, 0, ret,
+				    rxrpc_tx_fail_version_keepalive);
 
 	peer->last_tx_at = ktime_get_real();
 	_leave("");
diff --git a/net/rxrpc/rxkad.c b/net/rxrpc/rxkad.c
index 588fea0..6c0ae27 100644
--- a/net/rxrpc/rxkad.c
+++ b/net/rxrpc/rxkad.c
@@ -664,7 +664,8 @@ static int rxkad_issue_challenge(struct rxrpc_connection *conn)
 
 	ret = kernel_sendmsg(conn->params.local->socket, &msg, iov, 2, len);
 	if (ret < 0) {
-		_debug("sendmsg failed: %d", ret);
+		trace_rxrpc_tx_fail(conn->debug_id, serial, ret,
+				    rxrpc_tx_fail_conn_challenge);
 		return -EAGAIN;
 	}
 
@@ -719,7 +720,8 @@ static int rxkad_send_response(struct rxrpc_connection *conn,
 
 	ret = kernel_sendmsg(conn->params.local->socket, &msg, iov, 3, len);
 	if (ret < 0) {
-		_debug("sendmsg failed: %d", ret);
+		trace_rxrpc_tx_fail(conn->debug_id, serial, ret,
+				    rxrpc_tx_fail_conn_response);
 		return -EAGAIN;
 	}
 
-- 
cgit v1.1


From 1ef903bf795be01c91c10c93a0f9d9d6f2f7921b Mon Sep 17 00:00:00 2001
From: Daniel Jurgens <danielj@mellanox.com>
Date: Mon, 26 Mar 2018 13:35:29 -0500
Subject: net/mlx5: Free IRQs in shutdown path

Some platforms require IRQs to be free'd in the shutdown path. Otherwise
they will fail to be reallocated after a kexec.

Fixes: 8812c24d28f4 ("net/mlx5: Add fast unload support in shutdown flow")
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/eq.c       | 28 ++++++++++++++++++++++
 drivers/net/ethernet/mellanox/mlx5/core/main.c     |  8 +++++++
 .../net/ethernet/mellanox/mlx5/core/mlx5_core.h    |  2 ++
 3 files changed, 38 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eq.c b/drivers/net/ethernet/mellanox/mlx5/core/eq.c
index c1c9497..1814f80 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eq.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eq.c
@@ -34,6 +34,9 @@
 #include <linux/module.h>
 #include <linux/mlx5/driver.h>
 #include <linux/mlx5/cmd.h>
+#ifdef CONFIG_RFS_ACCEL
+#include <linux/cpu_rmap.h>
+#endif
 #include "mlx5_core.h"
 #include "fpga/core.h"
 #include "eswitch.h"
@@ -923,3 +926,28 @@ int mlx5_core_eq_query(struct mlx5_core_dev *dev, struct mlx5_eq *eq,
 	MLX5_SET(query_eq_in, in, eq_number, eq->eqn);
 	return mlx5_cmd_exec(dev, in, sizeof(in), out, outlen);
 }
+
+/* This function should only be called after mlx5_cmd_force_teardown_hca */
+void mlx5_core_eq_free_irqs(struct mlx5_core_dev *dev)
+{
+	struct mlx5_eq_table *table = &dev->priv.eq_table;
+	struct mlx5_eq *eq;
+
+#ifdef CONFIG_RFS_ACCEL
+	if (dev->rmap) {
+		free_irq_cpu_rmap(dev->rmap);
+		dev->rmap = NULL;
+	}
+#endif
+	list_for_each_entry(eq, &table->comp_eqs_list, list)
+		free_irq(eq->irqn, eq);
+
+	free_irq(table->pages_eq.irqn, &table->pages_eq);
+	free_irq(table->async_eq.irqn, &table->async_eq);
+	free_irq(table->cmd_eq.irqn, &table->cmd_eq);
+#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
+	if (MLX5_CAP_GEN(dev, pg))
+		free_irq(table->pfault_eq.irqn, &table->pfault_eq);
+#endif
+	pci_free_irq_vectors(dev->pdev);
+}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/main.c b/drivers/net/ethernet/mellanox/mlx5/core/main.c
index 63a8ea3..e2c465b 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/main.c
@@ -1587,6 +1587,14 @@ static int mlx5_try_fast_unload(struct mlx5_core_dev *dev)
 
 	mlx5_enter_error_state(dev, true);
 
+	/* Some platforms requiring freeing the IRQ's in the shutdown
+	 * flow. If they aren't freed they can't be allocated after
+	 * kexec. There is no need to cleanup the mlx5_core software
+	 * contexts.
+	 */
+	mlx5_irq_clear_affinity_hints(dev);
+	mlx5_core_eq_free_irqs(dev);
+
 	return 0;
 }
 
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h b/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h
index 7d001fe..023882d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h
@@ -128,6 +128,8 @@ int mlx5_core_eq_query(struct mlx5_core_dev *dev, struct mlx5_eq *eq,
 		       u32 *out, int outlen);
 int mlx5_start_eqs(struct mlx5_core_dev *dev);
 void mlx5_stop_eqs(struct mlx5_core_dev *dev);
+/* This function should only be called after mlx5_cmd_force_teardown_hca */
+void mlx5_core_eq_free_irqs(struct mlx5_core_dev *dev);
 struct mlx5_eq *mlx5_eqn2eq(struct mlx5_core_dev *dev, int eqn);
 u32 mlx5_eq_poll_irq_disabled(struct mlx5_eq *eq);
 void mlx5_cq_tasklet_cb(unsigned long data);
-- 
cgit v1.1


From 88d725bbb43cd63a40c8ef70dd373f1d38ead2e3 Mon Sep 17 00:00:00 2001
From: Adi Nissim <adin@mellanox.com>
Date: Wed, 25 Apr 2018 11:21:32 +0300
Subject: net/mlx5: E-Switch, Include VF RDMA stats in vport statistics

The host side reporting of VF vport statistics didn't include the VF
RDMA traffic.

Fixes: 3b751a2a418a ("net/mlx5: E-Switch, Introduce get vf statistics")
Signed-off-by: Adi Nissim <adin@mellanox.com>
Reported-by: Ariel Almog <ariela@mellanox.com>
Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/eswitch.c | 11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
index 332bc56..1352d13 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
@@ -2175,26 +2175,35 @@ int mlx5_eswitch_get_vport_stats(struct mlx5_eswitch *esw,
 	memset(vf_stats, 0, sizeof(*vf_stats));
 	vf_stats->rx_packets =
 		MLX5_GET_CTR(out, received_eth_unicast.packets) +
+		MLX5_GET_CTR(out, received_ib_unicast.packets) +
 		MLX5_GET_CTR(out, received_eth_multicast.packets) +
+		MLX5_GET_CTR(out, received_ib_multicast.packets) +
 		MLX5_GET_CTR(out, received_eth_broadcast.packets);
 
 	vf_stats->rx_bytes =
 		MLX5_GET_CTR(out, received_eth_unicast.octets) +
+		MLX5_GET_CTR(out, received_ib_unicast.octets) +
 		MLX5_GET_CTR(out, received_eth_multicast.octets) +
+		MLX5_GET_CTR(out, received_ib_multicast.octets) +
 		MLX5_GET_CTR(out, received_eth_broadcast.octets);
 
 	vf_stats->tx_packets =
 		MLX5_GET_CTR(out, transmitted_eth_unicast.packets) +
+		MLX5_GET_CTR(out, transmitted_ib_unicast.packets) +
 		MLX5_GET_CTR(out, transmitted_eth_multicast.packets) +
+		MLX5_GET_CTR(out, transmitted_ib_multicast.packets) +
 		MLX5_GET_CTR(out, transmitted_eth_broadcast.packets);
 
 	vf_stats->tx_bytes =
 		MLX5_GET_CTR(out, transmitted_eth_unicast.octets) +
+		MLX5_GET_CTR(out, transmitted_ib_unicast.octets) +
 		MLX5_GET_CTR(out, transmitted_eth_multicast.octets) +
+		MLX5_GET_CTR(out, transmitted_ib_multicast.octets) +
 		MLX5_GET_CTR(out, transmitted_eth_broadcast.octets);
 
 	vf_stats->multicast =
-		MLX5_GET_CTR(out, received_eth_multicast.packets);
+		MLX5_GET_CTR(out, received_eth_multicast.packets) +
+		MLX5_GET_CTR(out, received_ib_multicast.packets);
 
 	vf_stats->broadcast =
 		MLX5_GET_CTR(out, received_eth_broadcast.packets);
-- 
cgit v1.1


From f85900c3e13fdb61f040c9feecbcda601e0cdcfb Mon Sep 17 00:00:00 2001
From: Roi Dayan <roid@mellanox.com>
Date: Thu, 22 Mar 2018 18:51:37 +0200
Subject: net/mlx5e: Err if asked to offload TC match on frag being first

The HW doesn't support matching on frag first/later, return error if we are
asked to offload that.

Fixes: 3f7d0eb42d59 ("net/mlx5e: Offload TC matching on packets being IP fragments")
Signed-off-by: Roi Dayan <roid@mellanox.com>
Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 3c534fc..b94276d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -1261,6 +1261,10 @@ static int __parse_cls_flower(struct mlx5e_priv *priv,
 						  f->mask);
 		addr_type = key->addr_type;
 
+		/* the HW doesn't support frag first/later */
+		if (mask->flags & FLOW_DIS_FIRST_FRAG)
+			return -EOPNOTSUPP;
+
 		if (mask->flags & FLOW_DIS_IS_FRAGMENT) {
 			MLX5_SET(fte_match_set_lyr_2_4, headers_c, frag, 1);
 			MLX5_SET(fte_match_set_lyr_2_4, headers_v, frag,
-- 
cgit v1.1


From 6e70c267e68d77679534dcf4aaf84e66f2cf1425 Mon Sep 17 00:00:00 2001
From: Steve French <smfrench@gmail.com>
Date: Thu, 10 May 2018 10:59:37 -0500
Subject: smb3: directory sync should not return an error

As with NFS, which ignores sync on directory handles,
fsync on a directory handle is a noop for CIFS/SMB3.
Do not return an error on it.  It breaks some database
apps otherwise.

Signed-off-by: Steve French <smfrench@gmail.com>
CC: Stable <stable@vger.kernel.org>
Reviewed-by: Ronnie Sahlberg <lsahlber@redhat.com>
Reviewed-by: Pavel Shilovsky <pshilov@microsoft.com>
---
 fs/cifs/cifsfs.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)

diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c
index f715609..5a5a015 100644
--- a/fs/cifs/cifsfs.c
+++ b/fs/cifs/cifsfs.c
@@ -1047,6 +1047,18 @@ out:
 	return rc;
 }
 
+/*
+ * Directory operations under CIFS/SMB2/SMB3 are synchronous, so fsync()
+ * is a dummy operation.
+ */
+static int cifs_dir_fsync(struct file *file, loff_t start, loff_t end, int datasync)
+{
+	cifs_dbg(FYI, "Sync directory - name: %pD datasync: 0x%x\n",
+		 file, datasync);
+
+	return 0;
+}
+
 static ssize_t cifs_copy_file_range(struct file *src_file, loff_t off,
 				struct file *dst_file, loff_t destoff,
 				size_t len, unsigned int flags)
@@ -1181,6 +1193,7 @@ const struct file_operations cifs_dir_ops = {
 	.copy_file_range = cifs_copy_file_range,
 	.clone_file_range = cifs_clone_file_range,
 	.llseek = generic_file_llseek,
+	.fsync = cifs_dir_fsync,
 };
 
 static void
-- 
cgit v1.1


From 0a0b98734479aa5b3c671d5190e86273372cab95 Mon Sep 17 00:00:00 2001
From: Jann Horn <jannh@google.com>
Date: Fri, 11 May 2018 02:19:01 +0200
Subject: compat: fix 4-byte infoleak via uninitialized struct field

Commit 3a4d44b61625 ("ntp: Move adjtimex related compat syscalls to
native counterparts") removed the memset() in compat_get_timex().  Since
then, the compat adjtimex syscall can invoke do_adjtimex() with an
uninitialized ->tai.

If do_adjtimex() doesn't write to ->tai (e.g.  because the arguments are
invalid), compat_put_timex() then copies the uninitialized ->tai field
to userspace.

Fix it by adding the memset() back.

Fixes: 3a4d44b61625 ("ntp: Move adjtimex related compat syscalls to native counterparts")
Signed-off-by: Jann Horn <jannh@google.com>
Acked-by: Kees Cook <keescook@chromium.org>
Acked-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 kernel/compat.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/kernel/compat.c b/kernel/compat.c
index 6d21894..92d8c98 100644
--- a/kernel/compat.c
+++ b/kernel/compat.c
@@ -34,6 +34,7 @@ int compat_get_timex(struct timex *txc, const struct compat_timex __user *utp)
 {
 	struct compat_timex tx32;
 
+	memset(txc, 0, sizeof(struct timex));
 	if (copy_from_user(&tx32, utp, sizeof(struct compat_timex)))
 		return -EFAULT;
 
-- 
cgit v1.1


From 8e021151948c56eddf68affc6817965dabbcaddd Mon Sep 17 00:00:00 2001
From: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Date: Sat, 12 May 2018 03:03:12 +0530
Subject: drm/i915: Enable display workaround 827 for all planes, v2.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The workaround was applied only to the primary plane, but is required
on all planes. Iterate over all planes in the crtc atomic check to see
if the workaround is enabled, and only perform the actual toggling in
the pre/post plane update functions.

Changes since v1:
- Track active NV12 planes in a nv12_planes bitmask. (Ville)

v2: Removing BROXTON support for NV12 due to WA826

Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1526074397-10457-2-git-send-email-vidya.srinivas@intel.com
---
 drivers/gpu/drm/i915/intel_atomic_plane.c |  7 ++++-
 drivers/gpu/drm/i915/intel_display.c      | 43 +++++++++++++++++++------------
 drivers/gpu/drm/i915/intel_drv.h          |  1 +
 3 files changed, 33 insertions(+), 18 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_atomic_plane.c b/drivers/gpu/drm/i915/intel_atomic_plane.c
index 7481ce8..6d06878 100644
--- a/drivers/gpu/drm/i915/intel_atomic_plane.c
+++ b/drivers/gpu/drm/i915/intel_atomic_plane.c
@@ -183,11 +183,16 @@ int intel_plane_atomic_check_with_state(const struct intel_crtc_state *old_crtc_
 	}
 
 	/* FIXME pre-g4x don't work like this */
-	if (intel_state->base.visible)
+	if (state->visible)
 		crtc_state->active_planes |= BIT(intel_plane->id);
 	else
 		crtc_state->active_planes &= ~BIT(intel_plane->id);
 
+	if (state->visible && state->fb->format->format == DRM_FORMAT_NV12)
+		crtc_state->nv12_planes |= BIT(intel_plane->id);
+	else
+		crtc_state->nv12_planes &= ~BIT(intel_plane->id);
+
 	return intel_plane_atomic_calc_changes(old_crtc_state,
 					       &crtc_state->base,
 					       old_plane_state,
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index cdfe095..3bc1295 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -5142,6 +5142,22 @@ static bool hsw_post_update_enable_ips(const struct intel_crtc_state *old_crtc_s
 	return !old_crtc_state->ips_enabled;
 }
 
+static bool needs_nv12_wa(struct drm_i915_private *dev_priv,
+			  const struct intel_crtc_state *crtc_state)
+{
+	if (!crtc_state->nv12_planes)
+		return false;
+
+	if (IS_SKYLAKE(dev_priv) || IS_BROXTON(dev_priv))
+		return false;
+
+	if ((INTEL_GEN(dev_priv) == 9 && !IS_GEMINILAKE(dev_priv)) ||
+	    IS_CANNONLAKE(dev_priv))
+		return true;
+
+	return false;
+}
+
 static void intel_post_plane_update(struct intel_crtc_state *old_crtc_state)
 {
 	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->base.crtc);
@@ -5166,7 +5182,6 @@ static void intel_post_plane_update(struct intel_crtc_state *old_crtc_state)
 	if (old_primary_state) {
 		struct drm_plane_state *new_primary_state =
 			drm_atomic_get_new_plane_state(old_state, primary);
-		struct drm_framebuffer *fb = new_primary_state->fb;
 
 		intel_fbc_post_update(crtc);
 
@@ -5174,15 +5189,12 @@ static void intel_post_plane_update(struct intel_crtc_state *old_crtc_state)
 		    (needs_modeset(&pipe_config->base) ||
 		     !old_primary_state->visible))
 			intel_post_enable_primary(&crtc->base, pipe_config);
-
-		/* Display WA 827 */
-		if ((INTEL_GEN(dev_priv) == 9 && !IS_GEMINILAKE(dev_priv)) ||
-		    IS_CANNONLAKE(dev_priv)) {
-			if (fb && fb->format->format == DRM_FORMAT_NV12)
-				skl_wa_clkgate(dev_priv, crtc->pipe, false);
-		}
-
 	}
+
+	/* Display WA 827 */
+	if (needs_nv12_wa(dev_priv, old_crtc_state) &&
+	    !needs_nv12_wa(dev_priv, pipe_config))
+		skl_wa_clkgate(dev_priv, crtc->pipe, false);
 }
 
 static void intel_pre_plane_update(struct intel_crtc_state *old_crtc_state,
@@ -5206,14 +5218,6 @@ static void intel_pre_plane_update(struct intel_crtc_state *old_crtc_state,
 		struct intel_plane_state *new_primary_state =
 			intel_atomic_get_new_plane_state(old_intel_state,
 							 to_intel_plane(primary));
-		struct drm_framebuffer *fb = new_primary_state->base.fb;
-
-		/* Display WA 827 */
-		if ((INTEL_GEN(dev_priv) == 9 && !IS_GEMINILAKE(dev_priv)) ||
-		    IS_CANNONLAKE(dev_priv)) {
-			if (fb && fb->format->format == DRM_FORMAT_NV12)
-				skl_wa_clkgate(dev_priv, crtc->pipe, true);
-		}
 
 		intel_fbc_pre_update(crtc, pipe_config, new_primary_state);
 		/*
@@ -5225,6 +5229,11 @@ static void intel_pre_plane_update(struct intel_crtc_state *old_crtc_state,
 			intel_set_cpu_fifo_underrun_reporting(dev_priv, crtc->pipe, false);
 	}
 
+	/* Display WA 827 */
+	if (!needs_nv12_wa(dev_priv, old_crtc_state) &&
+	    needs_nv12_wa(dev_priv, pipe_config))
+		skl_wa_clkgate(dev_priv, crtc->pipe, true);
+
 	/*
 	 * Vblank time updates from the shadow to live plane control register
 	 * are blocked if the memory self-refresh mode is active at that
diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
index 52337f4..038870a 100644
--- a/drivers/gpu/drm/i915/intel_drv.h
+++ b/drivers/gpu/drm/i915/intel_drv.h
@@ -890,6 +890,7 @@ struct intel_crtc_state {
 
 	/* bitmask of visible planes (enum plane_id) */
 	u8 active_planes;
+	u8 nv12_planes;
 
 	/* HDMI scrambling status */
 	bool hdmi_scrambling;
-- 
cgit v1.1


From 6deef9b6057d9432e08dab3643be92517e0d15de Mon Sep 17 00:00:00 2001
From: Vidya Srinivas <vidya.srinivas@intel.com>
Date: Sat, 12 May 2018 03:03:13 +0530
Subject: drm/i915: Enable Display WA 0528

Possible hang with NV12 plane surface formats.
WA: When the plane source pixel format is NV12,
the CHICKEN_PIPESL_* register bit 22 must be set to 1
and the render decompression must not be enabled
on any of the planes in that pipe.

v2: removed unnecessary POSTING_READ

v3: Added RB from Maarten

v4: Removed support for NV12 for BROXTON

Credits-to: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Signed-off-by: Vidya Srinivas <vidya.srinivas@intel.com>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1526074397-10457-3-git-send-email-vidya.srinivas@intel.com
---
 drivers/gpu/drm/i915/intel_display.c | 22 +++++++++++++++++++---
 1 file changed, 19 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 3bc1295..9418953 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -489,9 +489,21 @@ static const struct intel_limit intel_limits_bxt = {
 };
 
 static void
+skl_wa_528(struct drm_i915_private *dev_priv, int pipe, bool enable)
+{
+	if (IS_SKYLAKE(dev_priv) || IS_BROXTON(dev_priv))
+		return;
+
+	if (enable)
+		I915_WRITE(CHICKEN_PIPESL_1(pipe), HSW_FBCQ_DIS);
+	else
+		I915_WRITE(CHICKEN_PIPESL_1(pipe), 0);
+}
+
+static void
 skl_wa_clkgate(struct drm_i915_private *dev_priv, int pipe, bool enable)
 {
-	if (IS_SKYLAKE(dev_priv))
+	if (IS_SKYLAKE(dev_priv) || IS_BROXTON(dev_priv))
 		return;
 
 	if (enable)
@@ -5193,8 +5205,10 @@ static void intel_post_plane_update(struct intel_crtc_state *old_crtc_state)
 
 	/* Display WA 827 */
 	if (needs_nv12_wa(dev_priv, old_crtc_state) &&
-	    !needs_nv12_wa(dev_priv, pipe_config))
+	    !needs_nv12_wa(dev_priv, pipe_config)) {
 		skl_wa_clkgate(dev_priv, crtc->pipe, false);
+		skl_wa_528(dev_priv, crtc->pipe, false);
+	}
 }
 
 static void intel_pre_plane_update(struct intel_crtc_state *old_crtc_state,
@@ -5231,8 +5245,10 @@ static void intel_pre_plane_update(struct intel_crtc_state *old_crtc_state,
 
 	/* Display WA 827 */
 	if (!needs_nv12_wa(dev_priv, old_crtc_state) &&
-	    needs_nv12_wa(dev_priv, pipe_config))
+	    needs_nv12_wa(dev_priv, pipe_config)) {
 		skl_wa_clkgate(dev_priv, crtc->pipe, true);
+		skl_wa_528(dev_priv, crtc->pipe, true);
+	}
 
 	/*
 	 * Vblank time updates from the shadow to live plane control register
-- 
cgit v1.1


From 5d794288fc562fd584c33a0834b45c134d6202c8 Mon Sep 17 00:00:00 2001
From: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Date: Sat, 12 May 2018 03:03:14 +0530
Subject: drm/i915: Add skl_check_nv12_surface for NV12

We skip src trunction/adjustments for
NV12 case and handle the sizes directly.
Without this, pipe fifo underruns are seen on APL/KBL.

v2: For NV12, making the src coordinates multiplier of 4

v3: Moving all the src coords handling code for NV12
to skl_check_nv12_surface

v4: Added RB from Mika

v5: Rebased the series. Removed checks of mult of 4 in
skl_update_scaler, Added NV12 condition in intel_check_sprite_plane
where src x/w is being checked for mult of 2 for yuv planes.

v6: Made changes to skl_check_nv12_surface as per WA#1106

Reviewed-by: Mika Kahola <mika.kahola@intel.com>
Signed-off-by: Vidya Srinivas <vidya.srinivas@intel.com>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1526074397-10457-4-git-send-email-vidya.srinivas@intel.com
---
 drivers/gpu/drm/i915/intel_display.c | 29 +++++++++++++++++++++++++++--
 drivers/gpu/drm/i915/intel_sprite.c  |  3 ++-
 2 files changed, 29 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 9418953..c8ff4b7 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -3102,6 +3102,29 @@ static int skl_check_main_surface(const struct intel_crtc_state *crtc_state,
 	return 0;
 }
 
+static int
+skl_check_nv12_surface(const struct intel_crtc_state *crtc_state,
+		       struct intel_plane_state *plane_state)
+{
+	/* Display WA #1106 */
+	if (plane_state->base.rotation !=
+	    (DRM_MODE_REFLECT_X | DRM_MODE_ROTATE_90) &&
+	    plane_state->base.rotation != DRM_MODE_ROTATE_270)
+		return 0;
+
+	/*
+	 * src coordinates are rotated here.
+	 * We check height but report it as width
+	 */
+	if (((drm_rect_height(&plane_state->base.src) >> 16) % 4) != 0) {
+		DRM_DEBUG_KMS("src width must be multiple "
+			      "of 4 for rotated NV12\n");
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
 static int skl_check_nv12_aux_surface(struct intel_plane_state *plane_state)
 {
 	const struct drm_framebuffer *fb = plane_state->base.fb;
@@ -3185,6 +3208,9 @@ int skl_check_plane_surface(const struct intel_crtc_state *crtc_state,
 	 * the main surface setup depends on it.
 	 */
 	if (fb->format->format == DRM_FORMAT_NV12) {
+		ret = skl_check_nv12_surface(crtc_state, plane_state);
+		if (ret)
+			return ret;
 		ret = skl_check_nv12_aux_surface(plane_state);
 		if (ret)
 			return ret;
@@ -4806,8 +4832,7 @@ skl_update_scaler(struct intel_crtc_state *crtc_state, bool force_detach,
 	}
 
 	if (plane_scaler_check && pixel_format == DRM_FORMAT_NV12 &&
-	    (src_h < SKL_MIN_YUV_420_SRC_H || (src_w % 4) != 0 ||
-	     (src_h % 4) != 0)) {
+	    (src_h < SKL_MIN_YUV_420_SRC_H || src_w < SKL_MIN_YUV_420_SRC_W)) {
 		DRM_DEBUG_KMS("NV12: src dimensions not met\n");
 		return -EINVAL;
 	}
diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c
index 9cd4be0..26de044 100644
--- a/drivers/gpu/drm/i915/intel_sprite.c
+++ b/drivers/gpu/drm/i915/intel_sprite.c
@@ -1060,7 +1060,8 @@ intel_check_sprite_plane(struct intel_plane *plane,
 		src_y = src->y1 >> 16;
 		src_h = drm_rect_height(src) >> 16;
 
-		if (intel_format_is_yuv(fb->format->format)) {
+		if (intel_format_is_yuv(fb->format->format) &&
+		    fb->format->format != DRM_FORMAT_NV12) {
 			src_x &= ~1;
 			src_w &= ~1;
 
-- 
cgit v1.1


From e44134f2673cf104c0ce42bfdebaf9f11bbef997 Mon Sep 17 00:00:00 2001
From: Chandra Konduru <chandra.konduru@intel.com>
Date: Sat, 12 May 2018 03:03:15 +0530
Subject: drm/i915: Add NV12 support to intel_framebuffer_init

This patch adds NV12 as supported format
to intel_framebuffer_init and performs various checks.

v2:
-Fix an issue in checks added (Chandra Konduru)

v3: rebased (me)

v4: Review comments by Ville addressed
Added platform check for NV12 in intel_framebuffer_init
Removed offset checks for NV12 case

v5: Addressed review comments by Clinton A Taylor
This NV12 support only correctly works on SKL.
Plane color space conversion is different on GLK and later platforms
causing the colors to display incorrectly.
Ville's plane color space property patch series
in review will fix this issue.
- Restricted the NV12 case in intel_framebuffer_init to
SKL and BXT only.

v6: Rebased (me)

v7: Addressed review comments by Ville
Restricting the NV12 to BXT for now.

v8: Rebased (me)
Restricting the NV12 changes to BXT and KBL for now.

v9: Rebased (me)

v10: NV12 supported by all GEN >= 9.
Making this change in intel_framebuffer_init. This is
part of addressing Maarten's review comments.
Comment under v8 no longer applicable

v11: Addressed review comments from Shashank Sharma

v12: Adding Reviewed By from Shashank Sharma

v13: Addressed review comments from Juha-Pekka Heikkila
"NV12 not to be supported by SKL"

v14: Addressed review comments from Maarten.
Add checks for fb width height for NV12 and fail the fb
creation if check fails. Added reviewed by from
Juha-Pekka Heikkila

v15: Rebased the series

v16: Setting the minimum value during fb creating to 16
as per Bspec for NV12. Earlier minimum was expected
to be > 16. Now changed it to >=16.

v17: Adding restriction to framebuffer_init - the fb
width and height should be a multiplier of 4

v18: Added RB from Maarten. Included Maarten's review comments
Dont allow CCS formats for fb creation of NV12

v19: Review comments from Maarten addressed -
Removing BROXTON support for NV12 due to WA826

Credits-to: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Tested-by: Clinton Taylor <clinton.a.taylor@intel.com>
Reviewed-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Shashank Sharma <shashank.sharma@intel.com>
Reviewed-by: Clinton Taylor <clinton.a.taylor@intel.com>
Reviewed-by: Juha-Pekka Heikkila <juhapekka.heikkila@gmail.com>
Signed-off-by: Chandra Konduru <chandra.konduru@intel.com>
Signed-off-by: Nabendu Maiti <nabendu.bikash.maiti@intel.com>
Signed-off-by: Vidya Srinivas <vidya.srinivas@intel.com>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1526074397-10457-5-git-send-email-vidya.srinivas@intel.com
---
 drivers/gpu/drm/i915/intel_display.c | 22 ++++++++++++++++++++++
 1 file changed, 22 insertions(+)

diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index c8ff4b7..1492cf9 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -14307,6 +14307,20 @@ static int intel_framebuffer_init(struct intel_framebuffer *intel_fb,
 			goto err;
 		}
 		break;
+	case DRM_FORMAT_NV12:
+		if (mode_cmd->modifier[0] == I915_FORMAT_MOD_Y_TILED_CCS ||
+		    mode_cmd->modifier[0] == I915_FORMAT_MOD_Yf_TILED_CCS) {
+			DRM_DEBUG_KMS("RC not to be enabled with NV12\n");
+			goto err;
+		}
+		if (INTEL_GEN(dev_priv) < 9 || IS_SKYLAKE(dev_priv) ||
+		    IS_BROXTON(dev_priv)) {
+			DRM_DEBUG_KMS("unsupported pixel format: %s\n",
+				      drm_get_format_name(mode_cmd->pixel_format,
+							  &format_name));
+			goto err;
+		}
+		break;
 	default:
 		DRM_DEBUG_KMS("unsupported pixel format: %s\n",
 			      drm_get_format_name(mode_cmd->pixel_format, &format_name));
@@ -14319,6 +14333,14 @@ static int intel_framebuffer_init(struct intel_framebuffer *intel_fb,
 
 	drm_helper_mode_fill_fb_struct(&dev_priv->drm, fb, mode_cmd);
 
+	if (fb->format->format == DRM_FORMAT_NV12 &&
+	    (fb->width < SKL_MIN_YUV_420_SRC_W ||
+	     fb->height < SKL_MIN_YUV_420_SRC_H ||
+	     (fb->width % 4) != 0 || (fb->height % 4) != 0)) {
+		DRM_DEBUG_KMS("src dimensions not correct for NV12\n");
+		return -EINVAL;
+	}
+
 	for (i = 0; i < fb->format->num_planes; i++) {
 		u32 stride_alignment;
 
-- 
cgit v1.1


From c0b56ab544c0aaefc0aa5526471f72e2324e002c Mon Sep 17 00:00:00 2001
From: Chandra Konduru <chandra.konduru@intel.com>
Date: Sat, 12 May 2018 03:03:16 +0530
Subject: drm/i915: Add NV12 as supported format for primary plane
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This patch adds NV12 to list of supported formats for
primary plane

v2: Rebased (Chandra Konduru)

v3: Rebased (me)

v4: Review comments by Ville addressed
Removed the skl_primary_formats_with_nv12 and
added NV12 case in existing skl_primary_formats

v5: Rebased (me)

v6: Missed the Tested-by/Reviewed-by in the previous series
Adding the same to commit message in this version.

v7: Review comments by Ville addressed
	Restricting the NV12 for BXT and on PIPE A and B
Rebased (me)

v8: Rebased (me)
Modified restricting the NV12 support for both BXT and KBL.

v9: Rebased (me)

v10: Addressed review comments from Maarten.
	Adding NV12 inside skl_primary_formats itself.

v11: Adding Reviewed By tag from Shashank Sharma

v12: Addressed review comments from Juha-Pekka Heikkila
"NV12 not to be supported by SKL"

v13: Addressed review comments from Ville
Added skl_pri_planar_formats to include NV12
and skl_plane_has_planar function to check for
NV12 support on plane. Added NV12 format to
skl_mod_supported. These were review comments
from Kristian Høgsberg <hoegsberg@gmail.com>

v14: Added reviewed by from Juha-Pekka Heikkila

v15: Rebased the series

v16: Added all tiling support under mod supported
for NV12. Credits to Megha Aggarwal

v17: Added RB by Maarten and Kristian

v18: Review comments from Maarten addressed -
Removing BROXTON support for NV12 due to WA826

v19: Addressed review comments from Maarten
Make changes to skl_mod_supported

Credits-to: Megha Aggarwal megha.aggarwal@intel.com
Credits-to: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Tested-by: Clinton Taylor <clinton.a.taylor@intel.com>
Reviewed-by: Kristian Høgsberg <hoegsberg@gmail.com>
Reviewed-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Juha-Pekka Heikkila <juhapekka.heikkila@gmail.com>
Reviewed-by: Clinton Taylor <clinton.a.taylor@intel.com>
Reviewed-by: Shashank Sharma <shashank.sharma@intel.com>
Signed-off-by: Chandra Konduru <chandra.konduru@intel.com>
Signed-off-by: Nabendu Maiti <nabendu.bikash.maiti@intel.com>
Signed-off-by: Vidya Srinivas <vidya.srinivas@intel.com>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1526074397-10457-6-git-send-email-vidya.srinivas@intel.com
---
 drivers/gpu/drm/i915/intel_display.c | 50 ++++++++++++++++++++++++++++++++++--
 drivers/gpu/drm/i915/intel_drv.h     |  2 ++
 2 files changed, 50 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 1492cf9..ad588d5 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -88,6 +88,22 @@ static const uint32_t skl_primary_formats[] = {
 	DRM_FORMAT_VYUY,
 };
 
+static const uint32_t skl_pri_planar_formats[] = {
+	DRM_FORMAT_C8,
+	DRM_FORMAT_RGB565,
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_XBGR8888,
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_ABGR8888,
+	DRM_FORMAT_XRGB2101010,
+	DRM_FORMAT_XBGR2101010,
+	DRM_FORMAT_YUYV,
+	DRM_FORMAT_YVYU,
+	DRM_FORMAT_UYVY,
+	DRM_FORMAT_VYUY,
+	DRM_FORMAT_NV12,
+};
+
 static const uint64_t skl_format_modifiers_noccs[] = {
 	I915_FORMAT_MOD_Yf_TILED,
 	I915_FORMAT_MOD_Y_TILED,
@@ -13218,6 +13234,7 @@ static bool skl_mod_supported(uint32_t format, uint64_t modifier)
 	case DRM_FORMAT_YVYU:
 	case DRM_FORMAT_UYVY:
 	case DRM_FORMAT_VYUY:
+	case DRM_FORMAT_NV12:
 		if (modifier == I915_FORMAT_MOD_Yf_TILED)
 			return true;
 		/* fall through */
@@ -13425,6 +13442,30 @@ static bool skl_plane_has_fbc(struct drm_i915_private *dev_priv,
 	return pipe == PIPE_A && plane_id == PLANE_PRIMARY;
 }
 
+bool skl_plane_has_planar(struct drm_i915_private *dev_priv,
+			  enum pipe pipe, enum plane_id plane_id)
+{
+	if (plane_id == PLANE_PRIMARY) {
+		if (IS_SKYLAKE(dev_priv) || IS_BROXTON(dev_priv))
+			return false;
+		else if ((INTEL_GEN(dev_priv) == 9 && pipe == PIPE_C) &&
+			 !IS_GEMINILAKE(dev_priv))
+			return false;
+	} else if (plane_id >= PLANE_SPRITE0) {
+		if (plane_id == PLANE_CURSOR)
+			return false;
+		if (IS_GEMINILAKE(dev_priv) || INTEL_GEN(dev_priv) == 10) {
+			if (plane_id != PLANE_SPRITE0)
+				return false;
+		} else {
+			if (plane_id != PLANE_SPRITE0 || pipe == PIPE_C ||
+			    IS_SKYLAKE(dev_priv) || IS_BROXTON(dev_priv))
+				return false;
+		}
+	}
+	return true;
+}
+
 static struct intel_plane *
 intel_primary_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe)
 {
@@ -13485,8 +13526,13 @@ intel_primary_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe)
 	primary->check_plane = intel_check_primary_plane;
 
 	if (INTEL_GEN(dev_priv) >= 9) {
-		intel_primary_formats = skl_primary_formats;
-		num_formats = ARRAY_SIZE(skl_primary_formats);
+		if (skl_plane_has_planar(dev_priv, pipe, PLANE_PRIMARY)) {
+			intel_primary_formats = skl_pri_planar_formats;
+			num_formats = ARRAY_SIZE(skl_pri_planar_formats);
+		} else {
+			intel_primary_formats = skl_primary_formats;
+			num_formats = ARRAY_SIZE(skl_primary_formats);
+		}
 
 		if (skl_plane_has_ccs(dev_priv, pipe, PLANE_PRIMARY))
 			modifiers = skl_format_modifiers_ccs;
diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
index 038870a..d7dbca1 100644
--- a/drivers/gpu/drm/i915/intel_drv.h
+++ b/drivers/gpu/drm/i915/intel_drv.h
@@ -2082,6 +2082,8 @@ bool skl_plane_get_hw_state(struct intel_plane *plane);
 bool skl_plane_has_ccs(struct drm_i915_private *dev_priv,
 		       enum pipe pipe, enum plane_id plane_id);
 bool intel_format_is_yuv(uint32_t format);
+bool skl_plane_has_planar(struct drm_i915_private *dev_priv,
+			  enum pipe pipe, enum plane_id plane_id);
 
 /* intel_tv.c */
 void intel_tv_init(struct drm_i915_private *dev_priv);
-- 
cgit v1.1


From 429204f1059909245d8f73b66aa729c6c2807cae Mon Sep 17 00:00:00 2001
From: Chandra Konduru <chandra.konduru@intel.com>
Date: Sat, 12 May 2018 03:03:17 +0530
Subject: drm/i915: Add NV12 as supported format for sprite plane
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This patch adds NV12 to list of supported formats for sprite plane.

v2: Rebased (me)

v3: Review comments by Ville addressed
- Removed skl_plane_formats_with_nv12 and added
NV12 case in existing skl_plane_formats
- Added the 10bpc RGB formats

v4: Addressed review comments from Clinton A Taylor
"Why are we adding 10 bit RGB formats with the NV12 series patches?
Trying to set XR30 or AB30 results in error returned even though
the modes are advertised for the planes"
- Removed 10bit RGB formats added previously with NV12 series

v5: Missed the Tested-by/Reviewed-by in the previous series
Adding the same to commit message in this version.
Addressed review comments from Clinton A Taylor
"Why are we adding 10 bit RGB formats with the NV12 series patches?
Trying to set XR30 or AB30 results in error returned even though
the modes are advertised for the planes"
- Previous version has 10bit RGB format removed from VLV formats
by mistake. Fixing that in this version.
Removed 10bit RGB formats added previously with NV12 series
for SKL.

v6: Addressed review comments by Ville
Restricting the NV12 to BXT and PIPE A and B

v7: Rebased (me)

v8: Rebased (me)
Restricting NV12 changes to BXT and KBL
Restricting NV12 changes for plane 0 (overlay)

v9: Rebased (me)

v10: Addressed review comments from Maarten.
Adding NV12 to skl_plane_formats itself.

v11: Addressed review comments from Shashank Sharma

v12: Addressed review comments from Shashank Sharma
Made the condition in intel_sprite_plane_create
simple and easy to read as suggested.

v13: Adding reviewed by tag from Shashank Sharma
Addressed review comments from Juha-Pekka Heikkila
"NV12 not to be supported by SKL"

v14: Addressed review comments from Ville
Added skl_planar_formats to include NV12
and a check skl_plane_has_planar in sprite create
Added NV12 format to skl_mod_supported. These were
review comments from Kristian Høgsberg <hoegsberg@gmail.com>

v15: Added reviewed by from Juha-Pekka Heikkila

v16: Rebased the series

v17: Added all tiling under mod supported for NV12
Credits to Megha Aggarwal

v18: Added RB by Maarten and Kristian

v19: Addressed review comments from Maarten
Made modification to skl_mod_supported

Credits-to: Megha Aggarwal <megha.aggarwal@intel.com>
Credits-to: Kristian Høgsberg <hoegsberg@gmail.com>
Reviewed-by: Kristian Høgsberg <hoegsberg@gmail.com>
Reviewed-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Tested-by: Clinton Taylor <clinton.a.taylor@intel.com>
Reviewed-by: Juha-Pekka Heikkila <juhapekka.heikkila@gmail.com>
Reviewed-by: Shashank Sharma <shashank.sharma@intel.com>
Reviewed-by: Clinton Taylor <clinton.a.taylor@intel.com>
Signed-off-by: Chandra Konduru <chandra.konduru@intel.com>
Signed-off-by: Nabendu Maiti <nabendu.bikash.maiti@intel.com>
Signed-off-by: Vidya Srinivas <vidya.srinivas@intel.com>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1526074397-10457-7-git-send-email-vidya.srinivas@intel.com
---
 drivers/gpu/drm/i915/intel_sprite.c | 24 ++++++++++++++++++++++--
 1 file changed, 22 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c
index 26de044..1597938 100644
--- a/drivers/gpu/drm/i915/intel_sprite.c
+++ b/drivers/gpu/drm/i915/intel_sprite.c
@@ -1254,6 +1254,19 @@ static uint32_t skl_plane_formats[] = {
 	DRM_FORMAT_VYUY,
 };
 
+static uint32_t skl_planar_formats[] = {
+	DRM_FORMAT_RGB565,
+	DRM_FORMAT_ABGR8888,
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_XBGR8888,
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_YUYV,
+	DRM_FORMAT_YVYU,
+	DRM_FORMAT_UYVY,
+	DRM_FORMAT_VYUY,
+	DRM_FORMAT_NV12,
+};
+
 static const uint64_t skl_plane_format_modifiers_noccs[] = {
 	I915_FORMAT_MOD_Yf_TILED,
 	I915_FORMAT_MOD_Y_TILED,
@@ -1348,6 +1361,7 @@ static bool skl_mod_supported(uint32_t format, uint64_t modifier)
 	case DRM_FORMAT_YVYU:
 	case DRM_FORMAT_UYVY:
 	case DRM_FORMAT_VYUY:
+	case DRM_FORMAT_NV12:
 		if (modifier == I915_FORMAT_MOD_Yf_TILED)
 			return true;
 		/* fall through */
@@ -1447,8 +1461,14 @@ intel_sprite_plane_create(struct drm_i915_private *dev_priv,
 		intel_plane->disable_plane = skl_disable_plane;
 		intel_plane->get_hw_state = skl_plane_get_hw_state;
 
-		plane_formats = skl_plane_formats;
-		num_plane_formats = ARRAY_SIZE(skl_plane_formats);
+		if (skl_plane_has_planar(dev_priv, pipe,
+					 PLANE_SPRITE0 + plane)) {
+			plane_formats = skl_planar_formats;
+			num_plane_formats = ARRAY_SIZE(skl_planar_formats);
+		} else {
+			plane_formats = skl_plane_formats;
+			num_plane_formats = ARRAY_SIZE(skl_plane_formats);
+		}
 
 		if (skl_plane_has_ccs(dev_priv, pipe, PLANE_SPRITE0 + plane))
 			modifiers = skl_plane_format_modifiers_ccs;
-- 
cgit v1.1


From ca6acc25250a1dc101c5a541b4f58bcc1dd65de5 Mon Sep 17 00:00:00 2001
From: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Date: Tue, 8 May 2018 15:41:54 +0300
Subject: drm/i915/gtt: Trust the uncached store to flush wcb

Not all architectures guarantee that uncached read will
flush the write combining buffer. So marking it explicitly
is recommended [1].

However we know the architecture we are operating on
and can avoid wmb as the UC store will flush the wcb [2].

Omit the wmb() before invalidate as redudant.

v2: squash combining and removal (Chris)
v3: remove obsolete comments about posting reads (Chris)

References: http://yarchive.net/comp/linux/write_combining.html [1]
References: http://download.intel.com/design/PentiumII/applnots/24442201.pdf [2]
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Matthew Auld <matthew.auld@intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20180508124154.14586-1-mika.kuoppala@linux.intel.com
---
 drivers/gpu/drm/i915/i915_gem_gtt.c | 18 ++++++++----------
 1 file changed, 8 insertions(+), 10 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index c879bfd..6eae9e1 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -110,7 +110,8 @@ i915_get_ggtt_vma_pages(struct i915_vma *vma);
 
 static void gen6_ggtt_invalidate(struct drm_i915_private *dev_priv)
 {
-	/* Note that as an uncached mmio write, this should flush the
+	/*
+	 * Note that as an uncached mmio write, this will flush the
 	 * WCB of the writes into the GGTT before it triggers the invalidate.
 	 */
 	I915_WRITE(GFX_FLSH_CNTL_GEN6, GFX_FLSH_CNTL_EN);
@@ -2418,11 +2419,9 @@ static void gen8_ggtt_insert_entries(struct i915_address_space *vm,
 	for_each_sgt_dma(addr, sgt_iter, vma->pages)
 		gen8_set_pte(gtt_entries++, pte_encode | addr);
 
-	wmb();
-
-	/* This next bit makes the above posting read even more important. We
-	 * want to flush the TLBs only after we're certain all the PTE updates
-	 * have finished.
+	/*
+	 * We want to flush the TLBs only after we're certain all the PTE
+	 * updates have finished.
 	 */
 	ggtt->invalidate(vm->i915);
 }
@@ -2460,11 +2459,10 @@ static void gen6_ggtt_insert_entries(struct i915_address_space *vm,
 	dma_addr_t addr;
 	for_each_sgt_dma(addr, iter, vma->pages)
 		iowrite32(vm->pte_encode(addr, level, flags), &entries[i++]);
-	wmb();
 
-	/* This next bit makes the above posting read even more important. We
-	 * want to flush the TLBs only after we're certain all the PTE updates
-	 * have finished.
+	/*
+	 * We want to flush the TLBs only after we're certain all the PTE
+	 * updates have finished.
 	 */
 	ggtt->invalidate(vm->i915);
 }
-- 
cgit v1.1


From cc38cae7c4e9350c93aa2da506086415fecd6e4a Mon Sep 17 00:00:00 2001
From: Oscar Mateo <oscar.mateo@intel.com>
Date: Tue, 8 May 2018 14:29:23 -0700
Subject: drm/i915/icl: Introduce initial Icelake Workarounds

Inherit workarounds from previous platforms that are still valid for
Icelake.

v2: GEN7_ROW_CHICKEN2 is masked
v3:
  - Since it has been fixed already in upstream, removed the TODO
    comment about WA_SET_BIT for WaInPlaceDecompressionHang.
  - Squashed with this patch:
      drm/i915/icl: add icelake_init_clock_gating()
    from Paulo Zanoni <paulo.r.zanoni@intel.com>
  - Squashed with this patch:
      drm/i915/icl: WaForceEnableNonCoherent
    from Oscar Mateo <oscar.mateo@intel.com>
  - WaPushConstantDereferenceHoldDisable is now Wa_1604370585 and
    applies to B0 as well.
  - WaPipeControlBefore3DStateSamplePattern WABB was being applied
    to ICL incorrectly.
v4:
  - Wrap the commit message
  - s/dev_priv/p to please checkpatch
v5: Rebased on top of the WA refactoring
v6: Rebased on top of further whitelist registers refactoring (Michel)
v7: Added WaRsForcewakeAddDelayForAck
v8: s/ICL_HDC_CHICKEN0/ICL_HDC_MODE (Mika)
v9:
  - C, not lisp (Chris)
  - WaIncreaseDefaultTLBEntries is the same for GEN > 9_LP (Tvrtko)

Signed-off-by: Oscar Mateo <oscar.mateo@intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Tomasz Lis <tomasz.lis@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1525814984-20039-2-git-send-email-oscar.mateo@intel.com
---
 drivers/gpu/drm/i915/i915_drv.h          |  9 +++++++
 drivers/gpu/drm/i915/i915_gem_gtt.c      |  6 ++---
 drivers/gpu/drm/i915/i915_reg.h          |  1 +
 drivers/gpu/drm/i915/intel_lrc.c         |  2 ++
 drivers/gpu/drm/i915/intel_pm.c          |  4 ++-
 drivers/gpu/drm/i915/intel_uncore.c      |  7 +++--
 drivers/gpu/drm/i915/intel_workarounds.c | 46 ++++++++++++++++++++++++++++++++
 7 files changed, 69 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 24c5e47..57fb3aa 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -2470,6 +2470,15 @@ intel_info(const struct drm_i915_private *dev_priv)
 #define IS_CNL_REVID(p, since, until) \
 	(IS_CANNONLAKE(p) && IS_REVID(p, since, until))
 
+#define ICL_REVID_A0		0x0
+#define ICL_REVID_A2		0x1
+#define ICL_REVID_B0		0x3
+#define ICL_REVID_B2		0x4
+#define ICL_REVID_C0		0x5
+
+#define IS_ICL_REVID(p, since, until) \
+	(IS_ICELAKE(p) && IS_REVID(p, since, until))
+
 /*
  * The genX designation typically refers to the render engine, so render
  * capability related checks should use IS_GEN, while display and other checks
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index 6eae9e1..c01d6db 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -2138,15 +2138,15 @@ static void gtt_write_workarounds(struct drm_i915_private *dev_priv)
 	 * called on driver load and after a GPU reset, so you can place
 	 * workarounds here even if they get overwritten by GPU reset.
 	 */
-	/* WaIncreaseDefaultTLBEntries:chv,bdw,skl,bxt,kbl,glk,cfl,cnl */
+	/* WaIncreaseDefaultTLBEntries:chv,bdw,skl,bxt,kbl,glk,cfl,cnl,icl */
 	if (IS_BROADWELL(dev_priv))
 		I915_WRITE(GEN8_L3_LRA_1_GPGPU, GEN8_L3_LRA_1_GPGPU_DEFAULT_VALUE_BDW);
 	else if (IS_CHERRYVIEW(dev_priv))
 		I915_WRITE(GEN8_L3_LRA_1_GPGPU, GEN8_L3_LRA_1_GPGPU_DEFAULT_VALUE_CHV);
-	else if (IS_GEN9_BC(dev_priv) || IS_GEN10(dev_priv))
-		I915_WRITE(GEN8_L3_LRA_1_GPGPU, GEN9_L3_LRA_1_GPGPU_DEFAULT_VALUE_SKL);
 	else if (IS_GEN9_LP(dev_priv))
 		I915_WRITE(GEN8_L3_LRA_1_GPGPU, GEN9_L3_LRA_1_GPGPU_DEFAULT_VALUE_BXT);
+	else if (INTEL_GEN(dev_priv) >= 9)
+		I915_WRITE(GEN8_L3_LRA_1_GPGPU, GEN9_L3_LRA_1_GPGPU_DEFAULT_VALUE_SKL);
 
 	/*
 	 * To support 64K PTEs we need to first enable the use of the
diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 085928c..2b22d4d 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -7238,6 +7238,7 @@ enum {
 /* GEN8 chicken */
 #define HDC_CHICKEN0				_MMIO(0x7300)
 #define CNL_HDC_CHICKEN0			_MMIO(0xE5F0)
+#define ICL_HDC_MODE				_MMIO(0xE5F4)
 #define  HDC_FORCE_CSR_NON_COHERENT_OVR_DISABLE	(1<<15)
 #define  HDC_FENCE_DEST_SLM_DISABLE		(1<<14)
 #define  HDC_DONOT_FETCH_MEM_WHEN_MASKED	(1<<11)
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index d3c00f6..243d403 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -1682,6 +1682,8 @@ static int intel_init_workaround_bb(struct intel_engine_cs *engine)
 		return -EINVAL;
 
 	switch (INTEL_GEN(engine->i915)) {
+	case 11:
+		return 0;
 	case 10:
 		wa_bb_fn[0] = gen10_init_indirectctx_bb;
 		wa_bb_fn[1] = NULL;
diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index 4126132..9c6e48c 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -9190,7 +9190,9 @@ static void nop_init_clock_gating(struct drm_i915_private *dev_priv)
  */
 void intel_init_clock_gating_hooks(struct drm_i915_private *dev_priv)
 {
-	if (IS_CANNONLAKE(dev_priv))
+	if (IS_ICELAKE(dev_priv))
+		dev_priv->display.init_clock_gating = nop_init_clock_gating;
+	else if (IS_CANNONLAKE(dev_priv))
 		dev_priv->display.init_clock_gating = cnl_init_clock_gating;
 	else if (IS_COFFEELAKE(dev_priv))
 		dev_priv->display.init_clock_gating = cfl_init_clock_gating;
diff --git a/drivers/gpu/drm/i915/intel_uncore.c b/drivers/gpu/drm/i915/intel_uncore.c
index d6e20f0..448293e 100644
--- a/drivers/gpu/drm/i915/intel_uncore.c
+++ b/drivers/gpu/drm/i915/intel_uncore.c
@@ -139,7 +139,9 @@ fw_domain_wait_ack_with_fallback(const struct drm_i915_private *i915,
 	 * in the hope that the original ack will be delivered along with
 	 * the fallback ack.
 	 *
-	 * This workaround is described in HSDES #1604254524
+	 * This workaround is described in HSDES #1604254524 and it's known as:
+	 * WaRsForcewakeAddDelayForAck:skl,bxt,kbl,glk,cfl,cnl,icl
+	 * although the name is a bit misleading.
 	 */
 
 	pass = 1;
@@ -1394,7 +1396,8 @@ static void intel_uncore_fw_domains_init(struct drm_i915_private *dev_priv)
 	if (INTEL_GEN(dev_priv) >= 11) {
 		int i;
 
-		dev_priv->uncore.funcs.force_wake_get = fw_domains_get;
+		dev_priv->uncore.funcs.force_wake_get =
+			fw_domains_get_with_fallback;
 		dev_priv->uncore.funcs.force_wake_put = fw_domains_put;
 		fw_domain_init(dev_priv, FW_DOMAIN_ID_RENDER,
 			       FORCEWAKE_RENDER_GEN9,
diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c
index ec9d340..73d02d3 100644
--- a/drivers/gpu/drm/i915/intel_workarounds.c
+++ b/drivers/gpu/drm/i915/intel_workarounds.c
@@ -441,6 +441,27 @@ static int cnl_ctx_workarounds_init(struct drm_i915_private *dev_priv)
 	return 0;
 }
 
+static int icl_ctx_workarounds_init(struct drm_i915_private *dev_priv)
+{
+	/* Wa_1604370585:icl (pre-prod)
+	 * Formerly known as WaPushConstantDereferenceHoldDisable
+	 */
+	if (IS_ICL_REVID(dev_priv, ICL_REVID_A0, ICL_REVID_B0))
+		WA_SET_BIT_MASKED(GEN7_ROW_CHICKEN2,
+				  PUSH_CONSTANT_DEREF_DISABLE);
+
+	/* WaForceEnableNonCoherent:icl
+	 * This is not the same workaround as in early Gen9 platforms, where
+	 * lacking this could cause system hangs, but coherency performance
+	 * overhead is high and only a few compute workloads really need it
+	 * (the register is whitelisted in hardware now, so UMDs can opt in
+	 * for coherency if they have a good reason).
+	 */
+	WA_SET_BIT_MASKED(ICL_HDC_MODE, HDC_FORCE_NON_COHERENT);
+
+	return 0;
+}
+
 int intel_ctx_workarounds_init(struct drm_i915_private *dev_priv)
 {
 	int err = 0;
@@ -465,6 +486,8 @@ int intel_ctx_workarounds_init(struct drm_i915_private *dev_priv)
 		err = cfl_ctx_workarounds_init(dev_priv);
 	else if (IS_CANNONLAKE(dev_priv))
 		err = cnl_ctx_workarounds_init(dev_priv);
+	else if (IS_ICELAKE(dev_priv))
+		err = icl_ctx_workarounds_init(dev_priv);
 	else
 		MISSING_CASE(INTEL_GEN(dev_priv));
 	if (err)
@@ -663,6 +686,21 @@ static void cnl_gt_workarounds_apply(struct drm_i915_private *dev_priv)
 		   _MASKED_BIT_ENABLE(GEN9_FFSC_PERCTX_PREEMPT_CTRL));
 }
 
+static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)
+{
+	/* This is not an Wa. Enable for better image quality */
+	I915_WRITE(_3D_CHICKEN3,
+		   _MASKED_BIT_ENABLE(_3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLE));
+
+	/* WaInPlaceDecompressionHang:icl */
+	I915_WRITE(GEN9_GAMT_ECO_REG_RW_IA, I915_READ(GEN9_GAMT_ECO_REG_RW_IA) |
+					    GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS);
+
+	/* WaPipelineFlushCoherentLines:icl */
+	I915_WRITE(GEN8_L3SQCREG4, I915_READ(GEN8_L3SQCREG4) |
+				   GEN8_LQSC_FLUSH_COHERENT_LINES);
+}
+
 void intel_gt_workarounds_apply(struct drm_i915_private *dev_priv)
 {
 	if (INTEL_GEN(dev_priv) < 8)
@@ -683,6 +721,8 @@ void intel_gt_workarounds_apply(struct drm_i915_private *dev_priv)
 		cfl_gt_workarounds_apply(dev_priv);
 	else if (IS_CANNONLAKE(dev_priv))
 		cnl_gt_workarounds_apply(dev_priv);
+	else if (IS_ICELAKE(dev_priv))
+		icl_gt_workarounds_apply(dev_priv);
 	else
 		MISSING_CASE(INTEL_GEN(dev_priv));
 }
@@ -761,6 +801,10 @@ static void cnl_whitelist_build(struct whitelist *w)
 	whitelist_reg(w, GEN8_CS_CHICKEN1);
 }
 
+static void icl_whitelist_build(struct whitelist *w)
+{
+}
+
 static struct whitelist *whitelist_build(struct intel_engine_cs *engine,
 					 struct whitelist *w)
 {
@@ -789,6 +833,8 @@ static struct whitelist *whitelist_build(struct intel_engine_cs *engine,
 		cfl_whitelist_build(w);
 	else if (IS_CANNONLAKE(i915))
 		cnl_whitelist_build(w);
+	else if (IS_ICELAKE(i915))
+		icl_whitelist_build(w);
 	else
 		MISSING_CASE(INTEL_GEN(i915));
 
-- 
cgit v1.1


From d65dc3e40b80ab63fb0d70c947558d0f49f912da Mon Sep 17 00:00:00 2001
From: Oscar Mateo <oscar.mateo@intel.com>
Date: Tue, 8 May 2018 14:29:24 -0700
Subject: drm/i915/icl: Enable Sampler DFR

Sampler Dynamic Frequency Rebalancing (DFR) aims to reduce Sampler
power by dynamically changing its clock frequency in low-throughput
conditions. This patches enables it by default on Gen11.

v2: Wrong operation to clear the bit (Praveen)
v3: Rebased on top of the WA refactoring
v4: Move to icl_init_clock_gating, since it's not a WA (Rodrigo)
v5: C, not lisp (Chris)

Signed-off-by: Oscar Mateo <oscar.mateo@intel.com>
Cc: Praveen Paneri <praveen.paneri@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Sagar Arun Kamble <sagar.a.kamble@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1525814984-20039-3-git-send-email-oscar.mateo@intel.com
---
 drivers/gpu/drm/i915/i915_reg.h | 3 +++
 drivers/gpu/drm/i915/intel_pm.c | 9 ++++++++-
 2 files changed, 11 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 2b22d4d..6aad16e 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -8253,6 +8253,9 @@ enum {
 #define GEN8_GARBCNTL                   _MMIO(0xB004)
 #define   GEN9_GAPS_TSV_CREDIT_DISABLE  (1<<7)
 
+#define GEN10_DFR_RATIO_EN_AND_CHICKEN	_MMIO(0x9550)
+#define   DFR_DISABLE			(1 << 9)
+
 /* IVYBRIDGE DPF */
 #define GEN7_L3CDERRST1(slice)		_MMIO(0xB008 + (slice) * 0x200) /* L3CD Error Status 1 */
 #define   GEN7_L3CDERRST1_ROW_MASK	(0x7ff<<14)
diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index 9c6e48c..b85229e 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -8664,6 +8664,13 @@ static void gen8_set_l3sqc_credits(struct drm_i915_private *dev_priv,
 	I915_WRITE(GEN7_MISCCPCTL, misccpctl);
 }
 
+static void icl_init_clock_gating(struct drm_i915_private *dev_priv)
+{
+	/* This is not an Wa. Enable to reduce Sampler power */
+	I915_WRITE(GEN10_DFR_RATIO_EN_AND_CHICKEN,
+		   I915_READ(GEN10_DFR_RATIO_EN_AND_CHICKEN) & ~DFR_DISABLE);
+}
+
 static void cnp_init_clock_gating(struct drm_i915_private *dev_priv)
 {
 	if (!HAS_PCH_CNP(dev_priv))
@@ -9191,7 +9198,7 @@ static void nop_init_clock_gating(struct drm_i915_private *dev_priv)
 void intel_init_clock_gating_hooks(struct drm_i915_private *dev_priv)
 {
 	if (IS_ICELAKE(dev_priv))
-		dev_priv->display.init_clock_gating = nop_init_clock_gating;
+		dev_priv->display.init_clock_gating = icl_init_clock_gating;
 	else if (IS_CANNONLAKE(dev_priv))
 		dev_priv->display.init_clock_gating = cnl_init_clock_gating;
 	else if (IS_COFFEELAKE(dev_priv))
-- 
cgit v1.1


From 5bcebe76704f43d598c8a8da8dd77ffd3afd754e Mon Sep 17 00:00:00 2001
From: Oscar Mateo <oscar.mateo@intel.com>
Date: Tue, 8 May 2018 14:29:25 -0700
Subject: drm/i915/icl: WaGAPZPriorityScheme

The default GAPZ arbitrer priority value at power-on has been found
to be incorrect.

v2: Now renamed to Wa_1405543622
v3: Rebased on top of the WA refactoring
v4: Added HSDES reference number (Mika)
v5:
  - Rebased
  - C, not lisp (Chris)

References: HSDES#1405543622
Signed-off-by: Oscar Mateo <oscar.mateo@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1525814984-20039-4-git-send-email-oscar.mateo@intel.com
---
 drivers/gpu/drm/i915/i915_reg.h          | 5 +++--
 drivers/gpu/drm/i915/intel_workarounds.c | 6 ++++++
 2 files changed, 9 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 6aad16e..c9c2ad5 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -8250,8 +8250,9 @@ enum {
 #define   GEN8_DOP_CLOCK_GATE_GUC_ENABLE	(1<<4)
 #define   GEN8_DOP_CLOCK_GATE_MEDIA_ENABLE     (1<<6)
 
-#define GEN8_GARBCNTL                   _MMIO(0xB004)
-#define   GEN9_GAPS_TSV_CREDIT_DISABLE  (1<<7)
+#define GEN8_GARBCNTL				_MMIO(0xB004)
+#define   GEN9_GAPS_TSV_CREDIT_DISABLE		(1 << 7)
+#define   GEN11_ARBITRATION_PRIO_ORDER_MASK	(0x3f << 22)
 
 #define GEN10_DFR_RATIO_EN_AND_CHICKEN	_MMIO(0x9550)
 #define   DFR_DISABLE			(1 << 9)
diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c
index 73d02d3..44ae0b4 100644
--- a/drivers/gpu/drm/i915/intel_workarounds.c
+++ b/drivers/gpu/drm/i915/intel_workarounds.c
@@ -699,6 +699,12 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)
 	/* WaPipelineFlushCoherentLines:icl */
 	I915_WRITE(GEN8_L3SQCREG4, I915_READ(GEN8_L3SQCREG4) |
 				   GEN8_LQSC_FLUSH_COHERENT_LINES);
+
+	/* Wa_1405543622:icl
+	 * Formerly known as WaGAPZPriorityScheme
+	 */
+	I915_WRITE(GEN8_GARBCNTL, I915_READ(GEN8_GARBCNTL) |
+				  GEN11_ARBITRATION_PRIO_ORDER_MASK);
 }
 
 void intel_gt_workarounds_apply(struct drm_i915_private *dev_priv)
-- 
cgit v1.1


From d41bab687999793d7331b7b8906dca7e1aeb64c7 Mon Sep 17 00:00:00 2001
From: Oscar Mateo <oscar.mateo@intel.com>
Date: Tue, 8 May 2018 14:29:26 -0700
Subject: drm/i915/icl: WaL3BankAddressHashing

Revert to an L3 non-hash model, for performance reasons.

v2:
  - Place the WA name above the actual change
  - Improve the register naming
v3:
  - Rebased
  - Renamed to Wa_1604223664
v4: Rebased on top of the WA refactoring
v5:
  - Added References (Mika)
  - Fixed wrong mask and value (Mika)
  - Do not apply together with another WA for the same
    register (not worth the hassle)
v6:
  - Rebased
  - C, not lisp (Chris)

References: HSDES#1604223664
Signed-off-by: Oscar Mateo <oscar.mateo@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1525814984-20039-5-git-send-email-oscar.mateo@intel.com
---
 drivers/gpu/drm/i915/i915_reg.h          |  6 ++++++
 drivers/gpu/drm/i915/intel_workarounds.c | 10 ++++++++++
 2 files changed, 16 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index c9c2ad5..81f1a8c 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -8253,6 +8253,12 @@ enum {
 #define GEN8_GARBCNTL				_MMIO(0xB004)
 #define   GEN9_GAPS_TSV_CREDIT_DISABLE		(1 << 7)
 #define   GEN11_ARBITRATION_PRIO_ORDER_MASK	(0x3f << 22)
+#define   GEN11_HASH_CTRL_EXCL_MASK		(0x7f << 0)
+#define   GEN11_HASH_CTRL_EXCL_BIT0		(1 << 0)
+
+#define GEN11_GLBLINVL				_MMIO(0xB404)
+#define   GEN11_BANK_HASH_ADDR_EXCL_MASK	(0x7f << 5)
+#define   GEN11_BANK_HASH_ADDR_EXCL_BIT0	(1 << 5)
 
 #define GEN10_DFR_RATIO_EN_AND_CHICKEN	_MMIO(0x9550)
 #define   DFR_DISABLE			(1 << 9)
diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c
index 44ae0b4..3b03729 100644
--- a/drivers/gpu/drm/i915/intel_workarounds.c
+++ b/drivers/gpu/drm/i915/intel_workarounds.c
@@ -705,6 +705,16 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)
 	 */
 	I915_WRITE(GEN8_GARBCNTL, I915_READ(GEN8_GARBCNTL) |
 				  GEN11_ARBITRATION_PRIO_ORDER_MASK);
+
+	/* Wa_1604223664:icl
+	 * Formerly known as WaL3BankAddressHashing
+	 */
+	I915_WRITE(GEN8_GARBCNTL,
+		   (I915_READ(GEN8_GARBCNTL) & ~GEN11_HASH_CTRL_EXCL_MASK) |
+		   GEN11_HASH_CTRL_EXCL_BIT0);
+	I915_WRITE(GEN11_GLBLINVL,
+		   (I915_READ(GEN11_GLBLINVL) & ~GEN11_BANK_HASH_ADDR_EXCL_MASK) |
+		   GEN11_BANK_HASH_ADDR_EXCL_BIT0);
 }
 
 void intel_gt_workarounds_apply(struct drm_i915_private *dev_priv)
-- 
cgit v1.1


From f4a357140a5693bfb0ab4e6608b7118a288a9ba9 Mon Sep 17 00:00:00 2001
From: Oscar Mateo <oscar.mateo@intel.com>
Date: Tue, 8 May 2018 14:29:27 -0700
Subject: drm/i915/icl: WaModifyGamTlbPartitioning

Adjust default GAM TLB partitioning for performance reasons.

v2: Only touch the bits that we really need
v3: Rebased on top of the WA refactoring
v4:
  - Added References (Mika)
  - Rebased
v5:
  - Rebased
  - C, not lisp (Chris)
  - Correct reference number (Mika)

References: HSDES#220160670
Signed-off-by: Oscar Mateo <oscar.mateo@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1525814984-20039-6-git-send-email-oscar.mateo@intel.com
---
 drivers/gpu/drm/i915/i915_reg.h          | 5 +++++
 drivers/gpu/drm/i915/intel_workarounds.c | 5 +++++
 2 files changed, 10 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 81f1a8c..7fe505c 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -8263,6 +8263,11 @@ enum {
 #define GEN10_DFR_RATIO_EN_AND_CHICKEN	_MMIO(0x9550)
 #define   DFR_DISABLE			(1 << 9)
 
+#define GEN11_GACB_PERF_CTRL			_MMIO(0x4B80)
+#define   GEN11_HASH_CTRL_MASK			(0x3 << 12 | 0xf << 0)
+#define   GEN11_HASH_CTRL_BIT0			(1 << 0)
+#define   GEN11_HASH_CTRL_BIT4			(1 << 12)
+
 /* IVYBRIDGE DPF */
 #define GEN7_L3CDERRST1(slice)		_MMIO(0xB008 + (slice) * 0x200) /* L3CD Error Status 1 */
 #define   GEN7_L3CDERRST1_ROW_MASK	(0x7ff<<14)
diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c
index 3b03729..2561c55 100644
--- a/drivers/gpu/drm/i915/intel_workarounds.c
+++ b/drivers/gpu/drm/i915/intel_workarounds.c
@@ -715,6 +715,11 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)
 	I915_WRITE(GEN11_GLBLINVL,
 		   (I915_READ(GEN11_GLBLINVL) & ~GEN11_BANK_HASH_ADDR_EXCL_MASK) |
 		   GEN11_BANK_HASH_ADDR_EXCL_BIT0);
+
+	/* WaModifyGamTlbPartitioning:icl */
+	I915_WRITE(GEN11_GACB_PERF_CTRL,
+		   (I915_READ(GEN11_GACB_PERF_CTRL) & ~GEN11_HASH_CTRL_MASK) |
+		   GEN11_HASH_CTRL_BIT0 | GEN11_HASH_CTRL_BIT4);
 }
 
 void intel_gt_workarounds_apply(struct drm_i915_private *dev_priv)
-- 
cgit v1.1


From 5246ae4bdb4ceae9778a7966db1d9522c6cb0ea7 Mon Sep 17 00:00:00 2001
From: Oscar Mateo <oscar.mateo@intel.com>
Date: Tue, 8 May 2018 14:29:28 -0700
Subject: drm/i915/icl: WaDisableCleanEvicts

Avoids an undefined LLC behavior.

BSpec: 9613

v2: Renamed to Wa_1405733216
v3: Spaces around '<<' and fix surrounding code
v4: Rebased on top of the WA refactoring
v5: Added References (Mika)
v6:
  - Rebased
  - C, not lisp (Chris)

References: HSDES#1405733216
Signed-off-by: Oscar Mateo <oscar.mateo@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1525814984-20039-7-git-send-email-oscar.mateo@intel.com
---
 drivers/gpu/drm/i915/i915_reg.h          | 5 +++--
 drivers/gpu/drm/i915/intel_workarounds.c | 6 ++++++
 2 files changed, 9 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 7fe505c..a7bd739 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -7232,8 +7232,9 @@ enum {
 #define  L3SQ_URB_READ_CAM_MATCH_DISABLE	(1<<27)
 
 #define GEN8_L3SQCREG4				_MMIO(0xb118)
-#define  GEN8_LQSC_RO_PERF_DIS			(1<<27)
-#define  GEN8_LQSC_FLUSH_COHERENT_LINES		(1<<21)
+#define  GEN11_LQSC_CLEAN_EVICT_DISABLE		(1 << 6)
+#define  GEN8_LQSC_RO_PERF_DIS			(1 << 27)
+#define  GEN8_LQSC_FLUSH_COHERENT_LINES		(1 << 21)
 
 /* GEN8 chicken */
 #define HDC_CHICKEN0				_MMIO(0x7300)
diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c
index 2561c55..7e8bcc2 100644
--- a/drivers/gpu/drm/i915/intel_workarounds.c
+++ b/drivers/gpu/drm/i915/intel_workarounds.c
@@ -720,6 +720,12 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)
 	I915_WRITE(GEN11_GACB_PERF_CTRL,
 		   (I915_READ(GEN11_GACB_PERF_CTRL) & ~GEN11_HASH_CTRL_MASK) |
 		   GEN11_HASH_CTRL_BIT0 | GEN11_HASH_CTRL_BIT4);
+
+	/* Wa_1405733216:icl
+	 * Formerly known as WaDisableCleanEvicts
+	 */
+	I915_WRITE(GEN8_L3SQCREG4, I915_READ(GEN8_L3SQCREG4) |
+				   GEN11_LQSC_CLEAN_EVICT_DISABLE);
 }
 
 void intel_gt_workarounds_apply(struct drm_i915_private *dev_priv)
-- 
cgit v1.1


From 6b967dc392090831954644549676409ca22fe8bf Mon Sep 17 00:00:00 2001
From: Oscar Mateo <oscar.mateo@intel.com>
Date: Tue, 8 May 2018 14:29:29 -0700
Subject: drm/i915/icl: WaCL2SFHalfMaxAlloc

This workarounds an issue with insufficient storage for the
CL2 and SF units.

v2: Renamed to Wa_1405766107
v3: Wrapped the commit message
v4: Rebased on top of the WA refactoring
v5: Added References (Mika)
v6:
  - Rebased
  - s/MACALLOC/MAXALLOC (Mika)
  - C, not lisp (Chris)

References: HSDES#1405766107
Signed-off-by: Oscar Mateo <oscar.mateo@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1525814984-20039-8-git-send-email-oscar.mateo@intel.com
---
 drivers/gpu/drm/i915/i915_reg.h          | 4 ++++
 drivers/gpu/drm/i915/intel_workarounds.c | 7 +++++++
 2 files changed, 11 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index a7bd739..d325fad 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -8269,6 +8269,10 @@ enum {
 #define   GEN11_HASH_CTRL_BIT0			(1 << 0)
 #define   GEN11_HASH_CTRL_BIT4			(1 << 12)
 
+#define GEN11_LSN_UNSLCVC				_MMIO(0xB43C)
+#define   GEN11_LSN_UNSLCVC_GAFS_HALF_CL2_MAXALLOC	(1 << 9)
+#define   GEN11_LSN_UNSLCVC_GAFS_HALF_SF_MAXALLOC	(1 << 7)
+
 /* IVYBRIDGE DPF */
 #define GEN7_L3CDERRST1(slice)		_MMIO(0xB008 + (slice) * 0x200) /* L3CD Error Status 1 */
 #define   GEN7_L3CDERRST1_ROW_MASK	(0x7ff<<14)
diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c
index 7e8bcc2..a6758bd 100644
--- a/drivers/gpu/drm/i915/intel_workarounds.c
+++ b/drivers/gpu/drm/i915/intel_workarounds.c
@@ -726,6 +726,13 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)
 	 */
 	I915_WRITE(GEN8_L3SQCREG4, I915_READ(GEN8_L3SQCREG4) |
 				   GEN11_LQSC_CLEAN_EVICT_DISABLE);
+
+	/* Wa_1405766107:icl
+	 * Formerly known as WaCL2SFHalfMaxAlloc
+	 */
+	I915_WRITE(GEN11_LSN_UNSLCVC, I915_READ(GEN11_LSN_UNSLCVC) |
+				      GEN11_LSN_UNSLCVC_GAFS_HALF_SF_MAXALLOC |
+				      GEN11_LSN_UNSLCVC_GAFS_HALF_CL2_MAXALLOC);
 }
 
 void intel_gt_workarounds_apply(struct drm_i915_private *dev_priv)
-- 
cgit v1.1


From 908ae05173637e9b39545636a12c244314d6fce1 Mon Sep 17 00:00:00 2001
From: Oscar Mateo <oscar.mateo@intel.com>
Date: Tue, 8 May 2018 14:29:30 -0700
Subject: drm/i915/icl: WaDisCtxReload

Revert to the legacy implementation to avoid a system hang.

v2: Correct the address for GAMW_ECO_DEV_RW_IA_REG
v3: Renamed to Wa_220166154
v4: Rebased on top of the WA refactoring
v5: Added References (Mika)
v6:
  - Rebased
  - C, not lisp (Chris)

References: HSDES#220166154
Signed-off-by: Oscar Mateo <oscar.mateo@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1525814984-20039-9-git-send-email-oscar.mateo@intel.com
---
 drivers/gpu/drm/i915/i915_reg.h          | 3 +++
 drivers/gpu/drm/i915/intel_workarounds.c | 6 ++++++
 2 files changed, 9 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index d325fad..dd23af3 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -8273,6 +8273,9 @@ enum {
 #define   GEN11_LSN_UNSLCVC_GAFS_HALF_CL2_MAXALLOC	(1 << 9)
 #define   GEN11_LSN_UNSLCVC_GAFS_HALF_SF_MAXALLOC	(1 << 7)
 
+#define GAMW_ECO_DEV_RW_IA_REG			_MMIO(0x4080)
+#define   GAMW_ECO_DEV_CTX_RELOAD_DISABLE	(1 << 7)
+
 /* IVYBRIDGE DPF */
 #define GEN7_L3CDERRST1(slice)		_MMIO(0xB008 + (slice) * 0x200) /* L3CD Error Status 1 */
 #define   GEN7_L3CDERRST1_ROW_MASK	(0x7ff<<14)
diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c
index a6758bd..3547403 100644
--- a/drivers/gpu/drm/i915/intel_workarounds.c
+++ b/drivers/gpu/drm/i915/intel_workarounds.c
@@ -733,6 +733,12 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)
 	I915_WRITE(GEN11_LSN_UNSLCVC, I915_READ(GEN11_LSN_UNSLCVC) |
 				      GEN11_LSN_UNSLCVC_GAFS_HALF_SF_MAXALLOC |
 				      GEN11_LSN_UNSLCVC_GAFS_HALF_CL2_MAXALLOC);
+
+	/* Wa_220166154:icl
+	 * Formerly known as WaDisCtxReload
+	 */
+	I915_WRITE(GAMW_ECO_DEV_RW_IA_REG, I915_READ(GAMW_ECO_DEV_RW_IA_REG) |
+					   GAMW_ECO_DEV_CTX_RELOAD_DISABLE);
 }
 
 void intel_gt_workarounds_apply(struct drm_i915_private *dev_priv)
-- 
cgit v1.1


From 0a437d4981650615ba3b4c7a9ce0c76894637217 Mon Sep 17 00:00:00 2001
From: Oscar Mateo <oscar.mateo@intel.com>
Date: Tue, 8 May 2018 14:29:31 -0700
Subject: drm/i915/icl: Wa_1405779004

Disable MSC clock gating to prevent data corruption.

BSpec: 19257

v2: Rebased on top of the WA refactoring
v3: Added References (Mika)
v4:
  - Rebased
  - C, not lisp (Chris)
  - A0 only (Mika)

References: HSDES#1405779004
Signed-off-by: Oscar Mateo <oscar.mateo@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1525814984-20039-10-git-send-email-oscar.mateo@intel.com
---
 drivers/gpu/drm/i915/i915_reg.h          | 1 +
 drivers/gpu/drm/i915/intel_workarounds.c | 6 ++++++
 2 files changed, 7 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index dd23af3..950ec8e 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -3840,6 +3840,7 @@ enum {
 #define SLICE_UNIT_LEVEL_CLKGATE	_MMIO(0x94d4)
 #define  SARBUNIT_CLKGATE_DIS		(1 << 5)
 #define  RCCUNIT_CLKGATE_DIS		(1 << 7)
+#define  MSCUNIT_CLKGATE_DIS		(1 << 10)
 
 #define SUBSLICE_UNIT_LEVEL_CLKGATE	_MMIO(0x9524)
 #define  GWUNIT_CLKGATE_DIS		(1 << 16)
diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c
index 3547403..469a83d 100644
--- a/drivers/gpu/drm/i915/intel_workarounds.c
+++ b/drivers/gpu/drm/i915/intel_workarounds.c
@@ -739,6 +739,12 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)
 	 */
 	I915_WRITE(GAMW_ECO_DEV_RW_IA_REG, I915_READ(GAMW_ECO_DEV_RW_IA_REG) |
 					   GAMW_ECO_DEV_CTX_RELOAD_DISABLE);
+
+	/* Wa_1405779004:icl (pre-prod) */
+	if (IS_ICL_REVID(dev_priv, ICL_REVID_A0, ICL_REVID_A0))
+		I915_WRITE(SLICE_UNIT_LEVEL_CLKGATE,
+			   I915_READ(SLICE_UNIT_LEVEL_CLKGATE) |
+			   MSCUNIT_CLKGATE_DIS);
 }
 
 void intel_gt_workarounds_apply(struct drm_i915_private *dev_priv)
-- 
cgit v1.1


From 36204d80bacb5382c7944fceb14975c727def102 Mon Sep 17 00:00:00 2001
From: Oscar Mateo <oscar.mateo@intel.com>
Date: Tue, 8 May 2018 14:29:32 -0700
Subject: drm/i915/icl: Wa_1406680159

Disable GWL clock gating to prevent an issue that might
cause hangs.

v2: Rebased on top of the WA refactoring
v3: Wa_2201832410 officially merged with Wa_1406680159
v4: Added References (Mika)
v5:
  - Rebased
  - C, not lisp (Chris)
  - Add reference where WA is better explained (Rodrigo)
  - Add reference to WA that got merged with this

References: HSDES#1406681710
References: HSDES#1406680159
References: HSDES#2201832410
Signed-off-by: Oscar Mateo <oscar.mateo@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1525814984-20039-11-git-send-email-oscar.mateo@intel.com
---
 drivers/gpu/drm/i915/intel_workarounds.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c
index 469a83d..a3fa01a 100644
--- a/drivers/gpu/drm/i915/intel_workarounds.c
+++ b/drivers/gpu/drm/i915/intel_workarounds.c
@@ -745,6 +745,11 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)
 		I915_WRITE(SLICE_UNIT_LEVEL_CLKGATE,
 			   I915_READ(SLICE_UNIT_LEVEL_CLKGATE) |
 			   MSCUNIT_CLKGATE_DIS);
+
+	/* Wa_1406680159:icl */
+	I915_WRITE(SUBSLICE_UNIT_LEVEL_CLKGATE,
+		   I915_READ(SUBSLICE_UNIT_LEVEL_CLKGATE) |
+		   GWUNIT_CLKGATE_DIS);
 }
 
 void intel_gt_workarounds_apply(struct drm_i915_private *dev_priv)
-- 
cgit v1.1


From 5215eef35fcbbc8f9bd68adff90eb813e8c3b7cf Mon Sep 17 00:00:00 2001
From: Oscar Mateo <oscar.mateo@intel.com>
Date: Tue, 8 May 2018 14:29:33 -0700
Subject: drm/i915/icl: Wa_1604302699

Disable I2M Write for performance reasons.

v2: Rebased on top of the WA refactoring
v3: Added References (Mika)
v4:
  - Rebased
  - C, not lisp (Chris)
  - GEN7 chicken bit in the wrong side of the fence (Mika)
  - Use two spaces to align bit macros

References: HSDES#1604302699
Signed-off-by: Oscar Mateo <oscar.mateo@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1525814984-20039-12-git-send-email-oscar.mateo@intel.com
---
 drivers/gpu/drm/i915/i915_reg.h          | 4 +++-
 drivers/gpu/drm/i915/intel_workarounds.c | 5 +++++
 2 files changed, 8 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 950ec8e..7cb2ddc 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -7227,7 +7227,9 @@ enum {
 #define GEN7_L3CNTLREG3				_MMIO(0xB024)
 
 #define GEN7_L3_CHICKEN_MODE_REGISTER		_MMIO(0xB030)
-#define  GEN7_WA_L3_CHICKEN_MODE				0x20000000
+#define   GEN7_WA_L3_CHICKEN_MODE		0x20000000
+#define GEN10_L3_CHICKEN_MODE_REGISTER		_MMIO(0xB114)
+#define   GEN11_I2M_WRITE_DISABLE		(1 << 28)
 
 #define GEN7_L3SQCREG4				_MMIO(0xb034)
 #define  L3SQ_URB_READ_CAM_MATCH_DISABLE	(1<<27)
diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c
index a3fa01a..2a4e3ee 100644
--- a/drivers/gpu/drm/i915/intel_workarounds.c
+++ b/drivers/gpu/drm/i915/intel_workarounds.c
@@ -750,6 +750,11 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)
 	I915_WRITE(SUBSLICE_UNIT_LEVEL_CLKGATE,
 		   I915_READ(SUBSLICE_UNIT_LEVEL_CLKGATE) |
 		   GWUNIT_CLKGATE_DIS);
+
+	/* Wa_1604302699:icl */
+	I915_WRITE(GEN10_L3_CHICKEN_MODE_REGISTER,
+		   I915_READ(GEN10_L3_CHICKEN_MODE_REGISTER) |
+		   GEN11_I2M_WRITE_DISABLE);
 }
 
 void intel_gt_workarounds_apply(struct drm_i915_private *dev_priv)
-- 
cgit v1.1


From 5ba700c73a89b0bace1a89a08e7a7eca5f011152 Mon Sep 17 00:00:00 2001
From: Oscar Mateo <oscar.mateo@intel.com>
Date: Tue, 8 May 2018 14:29:34 -0700
Subject: drm/i915/icl: Wa_1406838659

Disable CGPSF unit clock gating to prevent an issue.

v2: Rebased on top of the WA refactoring
v3: Added References (Mika)
v4:
  - Rebased
  - C, not lisp (Chris)
  - Remove unintentional whitespaces (Mika)
  - Fixed in C0 (Mika)

References: HSDES#1406838659
Signed-off-by: Oscar Mateo <oscar.mateo@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1525814984-20039-13-git-send-email-oscar.mateo@intel.com
---
 drivers/gpu/drm/i915/i915_reg.h          | 3 +++
 drivers/gpu/drm/i915/intel_workarounds.c | 6 ++++++
 2 files changed, 9 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 7cb2ddc..ce48427 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -3848,6 +3848,9 @@ enum {
 #define UNSLICE_UNIT_LEVEL_CLKGATE	_MMIO(0x9434)
 #define  VFUNIT_CLKGATE_DIS		(1 << 20)
 
+#define INF_UNIT_LEVEL_CLKGATE		_MMIO(0x9560)
+#define   CGPSF_CLKGATE_DIS		(1 << 3)
+
 /*
  * Display engine regs
  */
diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c
index 2a4e3ee..942d322 100644
--- a/drivers/gpu/drm/i915/intel_workarounds.c
+++ b/drivers/gpu/drm/i915/intel_workarounds.c
@@ -755,6 +755,12 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)
 	I915_WRITE(GEN10_L3_CHICKEN_MODE_REGISTER,
 		   I915_READ(GEN10_L3_CHICKEN_MODE_REGISTER) |
 		   GEN11_I2M_WRITE_DISABLE);
+
+	/* Wa_1406838659:icl (pre-prod) */
+	if (IS_ICL_REVID(dev_priv, ICL_REVID_A0, ICL_REVID_B0))
+		I915_WRITE(INF_UNIT_LEVEL_CLKGATE,
+			   I915_READ(INF_UNIT_LEVEL_CLKGATE) |
+			   CGPSF_CLKGATE_DIS);
 }
 
 void intel_gt_workarounds_apply(struct drm_i915_private *dev_priv)
-- 
cgit v1.1


From 73f4e8a338da114ec9e0b8c634a02fd85d4fa396 Mon Sep 17 00:00:00 2001
From: Oscar Mateo <oscar.mateo@intel.com>
Date: Tue, 8 May 2018 14:29:35 -0700
Subject: drm/i915/icl: WaForwardProgressSoftReset

Avoids a hang during soft reset.

v2: Rebased on top of the WA refactoring
v3: Added References (Mika)
v4:
  - Rebased
  - C, not lisp (Chris)
  - Which steppings affected by this are not clear.
    For the moment, apply unconditionally as per the
    BSpec (Mika)
  - Add reference to another HSD also related

References: HSDES#1405476379
References: HSDES#2006612137
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: Oscar Mateo <oscar.mateo@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1525814984-20039-14-git-send-email-oscar.mateo@intel.com
---
 drivers/gpu/drm/i915/i915_reg.h          | 5 +++++
 drivers/gpu/drm/i915/intel_workarounds.c | 7 +++++++
 2 files changed, 12 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index ce48427..1449178 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -9897,6 +9897,11 @@ enum skl_power_gate {
 /* Media decoder 2 MOCS registers */
 #define GEN11_MFX2_MOCS(i)	_MMIO(0x10000 + (i) * 4)
 
+#define GEN10_SCRATCH_LNCF2		_MMIO(0xb0a0)
+#define   PMFLUSHDONE_LNICRSDROP	(1 << 20)
+#define   PMFLUSH_GAPL3UNBLOCK		(1 << 21)
+#define   PMFLUSHDONE_LNEBLK		(1 << 22)
+
 /* gamt regs */
 #define GEN8_L3_LRA_1_GPGPU _MMIO(0x4dd4)
 #define   GEN8_L3_LRA_1_GPGPU_DEFAULT_VALUE_BDW  0x67F1427F /* max/min for LRA1/2 */
diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c
index 942d322..5eec4ce 100644
--- a/drivers/gpu/drm/i915/intel_workarounds.c
+++ b/drivers/gpu/drm/i915/intel_workarounds.c
@@ -761,6 +761,13 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)
 		I915_WRITE(INF_UNIT_LEVEL_CLKGATE,
 			   I915_READ(INF_UNIT_LEVEL_CLKGATE) |
 			   CGPSF_CLKGATE_DIS);
+
+	/* WaForwardProgressSoftReset:icl */
+	I915_WRITE(GEN10_SCRATCH_LNCF2,
+		   I915_READ(GEN10_SCRATCH_LNCF2) |
+		   PMFLUSHDONE_LNICRSDROP |
+		   PMFLUSH_GAPL3UNBLOCK |
+		   PMFLUSHDONE_LNEBLK);
 }
 
 void intel_gt_workarounds_apply(struct drm_i915_private *dev_priv)
-- 
cgit v1.1


From dc432c3d7f9bceb3de6f5b44fb9c657c9810ed6d Mon Sep 17 00:00:00 2001
From: "Steven Rostedt (VMware)" <rostedt@goodmis.org>
Date: Wed, 9 May 2018 11:59:32 -0400
Subject: tracing: Fix regex_match_front() to not over compare the test string

The regex match function regex_match_front() in the tracing filter logic,
was fixed to test just the pattern length from testing the entire test
string. That is, it went from strncmp(str, r->pattern, len) to
strcmp(str, r->pattern, r->len).

The issue is that str is not guaranteed to be nul terminated, and if r->len
is greater than the length of str, it can access more memory than is
allocated.

The solution is to add a simple test if (len < r->len) return 0.

Cc: stable@vger.kernel.org
Fixes: 285caad415f45 ("tracing/filters: Fix MATCH_FRONT_ONLY filter matching")
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
---
 kernel/trace/trace_events_filter.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/kernel/trace/trace_events_filter.c b/kernel/trace/trace_events_filter.c
index 1f951b3..7d306b7 100644
--- a/kernel/trace/trace_events_filter.c
+++ b/kernel/trace/trace_events_filter.c
@@ -762,6 +762,9 @@ static int regex_match_full(char *str, struct regex *r, int len)
 
 static int regex_match_front(char *str, struct regex *r, int len)
 {
+	if (len < r->len)
+		return 0;
+
 	if (strncmp(str, r->pattern, r->len) == 0)
 		return 1;
 	return 0;
-- 
cgit v1.1


From 77dfedb5be03779f9a5d83e323a1b36e32090105 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Fri, 11 May 2018 13:11:45 +0100
Subject: drm/i915/execlists: Use rmb() to order CSB reads
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

We assume that the CSB is written using the normal ringbuffer
coherency protocols, as outlined in kernel/events/ring_buffer.c:

    *   (HW)                              (DRIVER)
    *
    *   if (LOAD ->data_tail) {            LOAD ->data_head
    *                      (A)             smp_rmb()       (C)
    *      STORE $data                     LOAD $data
    *      smp_wmb()       (B)             smp_mb()        (D)
    *      STORE ->data_head               STORE ->data_tail
    *   }

So we assume that the HW fulfils its ordering requirements (B), and so
we should use a complimentary rmb (C) to ensure that our read of its
WRITE pointer is completed before we start accessing the data.

The final mb (D) is implied by the uncached mmio we perform to inform
the HW of our READ pointer.

References: https://bugs.freedesktop.org/show_bug.cgi?id=105064
References: https://bugs.freedesktop.org/show_bug.cgi?id=105888
References: https://bugs.freedesktop.org/show_bug.cgi?id=106185
Fixes: 767a983ab255 ("drm/i915/execlists: Read the context-status HEAD from the HWSP")
References: 61bf9719fa17 ("drm/i915/cnl: Use mmio access to context status buffer")
Suggested-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Michał Winiarski <michal.winiarski@intel.com>
Cc: Rafael Antognolli <rafael.antognolli@intel.com>
Cc: Michel Thierry <michel.thierry@intel.com>
Cc: Timo Aaltonen <tjaalton@ubuntu.com>
Tested-by: Timo Aaltonen <tjaalton@ubuntu.com>
Acked-by: Michel Thierry <michel.thierry@intel.com>
Acked-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180511121147.31915-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/intel_lrc.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 243d403..7c6164d 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -1009,6 +1009,7 @@ static void execlists_submission_tasklet(unsigned long data)
 
 			head = execlists->csb_head;
 			tail = READ_ONCE(buf[write_idx]);
+			rmb(); /* Hopefully paired with a wmb() in HW */
 		}
 		GEM_TRACE("%s cs-irq head=%d [%d%s], tail=%d [%d%s]\n",
 			  engine->name,
-- 
cgit v1.1


From e71a82d8c1fa28ab048227df929e4f07d98f1656 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Fri, 11 May 2018 13:11:46 +0100
Subject: Revert "drm/i915/cnl: Use mmio access to context status buffer"
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

In the previous patch (to include a rmb() after readig the CSB WRITE
pointer from the HWSP) we believe we have fixed the underlying bug, and
so can re-enable using the HWSP on Cannolake.

This reverts commit 61bf9719fa17 ("drm/i915/cnl: Use mmio access to
context status buffer").

References: https://bugs.freedesktop.org/show_bug.cgi?id=105888
References: https://bugs.freedesktop.org/show_bug.cgi?id=106185
References: 61bf9719fa17 ("drm/i915/cnl: Use mmio access to context status buffer")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Michał Winiarski <michal.winiarski@intel.com>
Cc: Rafael Antognolli <rafael.antognolli@intel.com>
Cc: Michel Thierry <michel.thierry@intel.com>
Cc: Timo Aaltonen <tjaalton@ubuntu.com>
Tested-by: Timo Aaltonen <tjaalton@ubuntu.com>
Acked-by: Michel Thierry <michel.thierry@intel.com>
Acked-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180511121147.31915-2-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/intel_engine_cs.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
index 70325e0..8303e05 100644
--- a/drivers/gpu/drm/i915/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/intel_engine_cs.c
@@ -470,9 +470,6 @@ static bool csb_force_mmio(struct drm_i915_private *i915)
 	if (intel_vgpu_active(i915) && !intel_vgpu_has_hwsp_emulation(i915))
 		return true;
 
-	if (IS_CANNONLAKE(i915))
-		return true;
-
 	return false;
 }
 
-- 
cgit v1.1


From 4fa8667ca3989ce14cf66301fa251544fbddbdd0 Mon Sep 17 00:00:00 2001
From: Debabrata Banerjee <dbanerje@akamai.com>
Date: Wed, 9 May 2018 19:32:10 -0400
Subject: bonding: do not allow rlb updates to invalid mac

Make sure multicast, broadcast, and zero mac's cannot be the output of rlb
updates, which should all be directed arps. Receive load balancing will be
collapsed if any of these happen, as the switch will broadcast.

Signed-off-by: Debabrata Banerjee <dbanerje@akamai.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/bonding/bond_alb.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/bonding/bond_alb.c b/drivers/net/bonding/bond_alb.c
index 1ed9529..3f6faa6 100644
--- a/drivers/net/bonding/bond_alb.c
+++ b/drivers/net/bonding/bond_alb.c
@@ -450,7 +450,7 @@ static void rlb_update_client(struct rlb_client_info *client_info)
 {
 	int i;
 
-	if (!client_info->slave)
+	if (!client_info->slave || !is_valid_ether_addr(client_info->mac_dst))
 		return;
 
 	for (i = 0; i < RLB_ARP_BURST_SIZE; i++) {
-- 
cgit v1.1


From 21706ee8a47d3ede7fdae0be6d7c0a0e31a83229 Mon Sep 17 00:00:00 2001
From: Debabrata Banerjee <dbanerje@akamai.com>
Date: Wed, 9 May 2018 19:32:11 -0400
Subject: bonding: send learning packets for vlans on slave

There was a regression at some point from the intended functionality of
commit f60c3704e87d ("bonding: Fix alb mode to only use first level
vlans.")

Given the return value vlan_get_encap_level() we need to store the nest
level of the bond device, and then compare the vlan's encap level to
this. Without this, this check always fails and learning packets are
never sent.

In addition, this same commit caused a regression in the behavior of
balance_alb, which requires learning packets be sent for all interfaces
using the slave's mac in order to load balance properly. For vlan's
that have not set a user mac, we can send after checking one bit.
Otherwise we need send the set mac, albeit defeating rx load balancing
for that vlan.

Signed-off-by: Debabrata Banerjee <dbanerje@akamai.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/bonding/bond_alb.c  | 13 ++++++++-----
 drivers/net/bonding/bond_main.c |  2 ++
 include/net/bonding.h           |  1 +
 3 files changed, 11 insertions(+), 5 deletions(-)

diff --git a/drivers/net/bonding/bond_alb.c b/drivers/net/bonding/bond_alb.c
index 3f6faa6..5eb0df2 100644
--- a/drivers/net/bonding/bond_alb.c
+++ b/drivers/net/bonding/bond_alb.c
@@ -943,6 +943,10 @@ static void alb_send_lp_vid(struct slave *slave, u8 mac_addr[],
 	skb->priority = TC_PRIO_CONTROL;
 	skb->dev = slave->dev;
 
+	netdev_dbg(slave->bond->dev,
+		   "Send learning packet: dev %s mac %pM vlan %d\n",
+		   slave->dev->name, mac_addr, vid);
+
 	if (vid)
 		__vlan_hwaccel_put_tag(skb, vlan_proto, vid);
 
@@ -965,14 +969,13 @@ static int alb_upper_dev_walk(struct net_device *upper, void *_data)
 	u8 *mac_addr = data->mac_addr;
 	struct bond_vlan_tag *tags;
 
-	if (is_vlan_dev(upper) && vlan_get_encap_level(upper) == 0) {
-		if (strict_match &&
-		    ether_addr_equal_64bits(mac_addr,
-					    upper->dev_addr)) {
+	if (is_vlan_dev(upper) &&
+	    bond->nest_level == vlan_get_encap_level(upper) - 1) {
+		if (upper->addr_assign_type == NET_ADDR_STOLEN) {
 			alb_send_lp_vid(slave, mac_addr,
 					vlan_dev_vlan_proto(upper),
 					vlan_dev_vlan_id(upper));
-		} else if (!strict_match) {
+		} else {
 			alb_send_lp_vid(slave, upper->dev_addr,
 					vlan_dev_vlan_proto(upper),
 					vlan_dev_vlan_id(upper));
diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
index 718e491..1f1e97b 100644
--- a/drivers/net/bonding/bond_main.c
+++ b/drivers/net/bonding/bond_main.c
@@ -1738,6 +1738,8 @@ int bond_enslave(struct net_device *bond_dev, struct net_device *slave_dev,
 	if (bond_mode_uses_xmit_hash(bond))
 		bond_update_slave_arr(bond, NULL);
 
+	bond->nest_level = dev_get_nest_level(bond_dev);
+
 	netdev_info(bond_dev, "Enslaving %s as %s interface with %s link\n",
 		    slave_dev->name,
 		    bond_is_active_slave(new_slave) ? "an active" : "a backup",
diff --git a/include/net/bonding.h b/include/net/bonding.h
index f801fc9..b522351 100644
--- a/include/net/bonding.h
+++ b/include/net/bonding.h
@@ -198,6 +198,7 @@ struct bonding {
 	struct   slave __rcu *primary_slave;
 	struct   bond_up_slave __rcu *slave_arr; /* Array of usable slaves */
 	bool     force_primary;
+	u32      nest_level;
 	s32      slave_cnt; /* never change this value outside the attach/detach wrappers */
 	int     (*recv_probe)(const struct sk_buff *, struct bonding *,
 			      struct slave *);
-- 
cgit v1.1


From 8ccc113172e4c1ebef45c2433f3c32ed6ae1b9c9 Mon Sep 17 00:00:00 2001
From: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Date: Thu, 10 May 2018 13:26:16 +0200
Subject: mlxsw: core: Fix an error handling path in
 'mlxsw_core_bus_device_register()'

Resources are not freed in the reverse order of the allocation.
Labels are also mixed-up.

Fix it and reorder code and labels in the error handling path of
'mlxsw_core_bus_device_register()'

Fixes: ef3116e5403e ("mlxsw: spectrum: Register KVD resources with devlink")
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/mellanox/mlxsw/core.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlxsw/core.c b/drivers/net/ethernet/mellanox/mlxsw/core.c
index 93ea566..e13ac3b 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/core.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/core.c
@@ -1100,11 +1100,11 @@ err_emad_init:
 err_alloc_lag_mapping:
 	mlxsw_ports_fini(mlxsw_core);
 err_ports_init:
-	mlxsw_bus->fini(bus_priv);
-err_bus_init:
 	if (!reload)
 		devlink_resources_unregister(devlink, NULL);
 err_register_resources:
+	mlxsw_bus->fini(bus_priv);
+err_bus_init:
 	if (!reload)
 		devlink_free(devlink);
 err_devlink_alloc:
-- 
cgit v1.1


From 1b97013bfb11d66f041de691de6f0fec748ce016 Mon Sep 17 00:00:00 2001
From: Andrey Ignatov <rdna@fb.com>
Date: Thu, 10 May 2018 10:59:34 -0700
Subject: ipv4: fix memory leaks in udp_sendmsg, ping_v4_sendmsg

Fix more memory leaks in ip_cmsg_send() callers. Part of them were fixed
earlier in 919483096bfe.

* udp_sendmsg one was there since the beginning when linux sources were
  first added to git;
* ping_v4_sendmsg one was copy/pasted in c319b4d76b9e.

Whenever return happens in udp_sendmsg() or ping_v4_sendmsg() IP options
have to be freed if they were allocated previously.

Add label so that future callers (if any) can use it instead of kfree()
before return that is easy to forget.

Fixes: c319b4d76b9e (net: ipv4: add IPPROTO_ICMP socket kind)
Signed-off-by: Andrey Ignatov <rdna@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/ipv4/ping.c | 7 +++++--
 net/ipv4/udp.c  | 7 +++++--
 2 files changed, 10 insertions(+), 4 deletions(-)

diff --git a/net/ipv4/ping.c b/net/ipv4/ping.c
index 05e47d7..56a0106 100644
--- a/net/ipv4/ping.c
+++ b/net/ipv4/ping.c
@@ -775,8 +775,10 @@ static int ping_v4_sendmsg(struct sock *sk, struct msghdr *msg, size_t len)
 	ipc.addr = faddr = daddr;
 
 	if (ipc.opt && ipc.opt->opt.srr) {
-		if (!daddr)
-			return -EINVAL;
+		if (!daddr) {
+			err = -EINVAL;
+			goto out_free;
+		}
 		faddr = ipc.opt->opt.faddr;
 	}
 	tos = get_rttos(&ipc, inet);
@@ -842,6 +844,7 @@ back_from_confirm:
 
 out:
 	ip_rt_put(rt);
+out_free:
 	if (free)
 		kfree(ipc.opt);
 	if (!err) {
diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
index c2a292d..b61a770 100644
--- a/net/ipv4/udp.c
+++ b/net/ipv4/udp.c
@@ -952,8 +952,10 @@ int udp_sendmsg(struct sock *sk, struct msghdr *msg, size_t len)
 	sock_tx_timestamp(sk, ipc.sockc.tsflags, &ipc.tx_flags);
 
 	if (ipc.opt && ipc.opt->opt.srr) {
-		if (!daddr)
-			return -EINVAL;
+		if (!daddr) {
+			err = -EINVAL;
+			goto out_free;
+		}
 		faddr = ipc.opt->opt.faddr;
 		connected = 0;
 	}
@@ -1074,6 +1076,7 @@ do_append_data:
 
 out:
 	ip_rt_put(rt);
+out_free:
 	if (free)
 		kfree(ipc.opt);
 	if (!err)
-- 
cgit v1.1


From e896d29a548d04371ce746f7d02a8488ff93d812 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Fri, 11 May 2018 14:52:07 +0100
Subject: drm/i915/oa: Check that OA is disabled before unpinning

Before we unpin the buffer used for OA reports and return it to the
system, we need to be sure that the HW has finished writing into it.
For lack of a better idea, poll OACONTROL to check it is switched off.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=106379
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Tested-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180511135207.12880-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/i915_perf.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_perf.c b/drivers/gpu/drm/i915/i915_perf.c
index d934141..019bd2d 100644
--- a/drivers/gpu/drm/i915/i915_perf.c
+++ b/drivers/gpu/drm/i915/i915_perf.c
@@ -1960,11 +1960,19 @@ static void i915_oa_stream_enable(struct i915_perf_stream *stream)
 static void gen7_oa_disable(struct drm_i915_private *dev_priv)
 {
 	I915_WRITE(GEN7_OACONTROL, 0);
+	if (intel_wait_for_register(dev_priv,
+				    GEN7_OACONTROL, GEN7_OACONTROL_ENABLE, 0,
+				    50))
+		DRM_ERROR("wait for OA to be disabled timed out\n");
 }
 
 static void gen8_oa_disable(struct drm_i915_private *dev_priv)
 {
 	I915_WRITE(GEN8_OACONTROL, 0);
+	if (intel_wait_for_register(dev_priv,
+				    GEN8_OACONTROL, GEN8_OA_COUNTER_ENABLE, 0,
+				    50))
+		DRM_ERROR("wait for OA to be disabled timed out\n");
 }
 
 /**
-- 
cgit v1.1


From 4e50d9ebaeaa3c6761d2b513ef7039510c8cf213 Mon Sep 17 00:00:00 2001
From: Charles Machalow <charles.machalow@intel.com>
Date: Thu, 10 May 2018 16:01:38 -0700
Subject: nvme: Fix sync controller reset return

If a controller reset is requested while the device has no namespaces,
we were incorrectly returning ENETRESET. This patch adds the check for
ADMIN_ONLY controller state to indicate a successful reset.

Fixes: 8000d1fdb0  ("nvme-rdma: fix sysfs invoked reset_ctrl error flow ")
Cc: <stable@vger.kernel.org>
Signed-off-by: Charles Machalow <charles.machalow@intel.com>
[changelog]
Signed-off-by: Keith Busch <keith.busch@intel.com>
---
 drivers/nvme/host/core.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c
index 2cbc378..99b857e 100644
--- a/drivers/nvme/host/core.c
+++ b/drivers/nvme/host/core.c
@@ -118,7 +118,8 @@ int nvme_reset_ctrl_sync(struct nvme_ctrl *ctrl)
 	ret = nvme_reset_ctrl(ctrl);
 	if (!ret) {
 		flush_work(&ctrl->reset_work);
-		if (ctrl->state != NVME_CTRL_LIVE)
+		if (ctrl->state != NVME_CTRL_LIVE &&
+		    ctrl->state != NVME_CTRL_ADMIN_ONLY)
 			ret = -ENETRESET;
 	}
 
-- 
cgit v1.1


From be83bbf806822b1b89e0a0f23cd87cddc409e429 Mon Sep 17 00:00:00 2001
From: Linus Torvalds <torvalds@linux-foundation.org>
Date: Fri, 11 May 2018 09:52:01 -0700
Subject: mmap: introduce sane default mmap limits

The internal VM "mmap()" interfaces are based on the mmap target doing
everything using page indexes rather than byte offsets, because
traditionally (ie 32-bit) we had the situation that the byte offset
didn't fit in a register.  So while the mmap virtual address was limited
by the word size of the architecture, the backing store was not.

So we're basically passing "pgoff" around as a page index, in order to
be able to describe backing store locations that are much bigger than
the word size (think files larger than 4GB etc).

But while this all makes a ton of sense conceptually, we've been dogged
by various drivers that don't really understand this, and internally
work with byte offsets, and then try to work with the page index by
turning it into a byte offset with "pgoff << PAGE_SHIFT".

Which obviously can overflow.

Adding the size of the mapping to it to get the byte offset of the end
of the backing store just exacerbates the problem, and if you then use
this overflow-prone value to check various limits of your device driver
mmap capability, you're just setting yourself up for problems.

The correct thing for drivers to do is to do their limit math in page
indices, the way the interface is designed.  Because the generic mmap
code _does_ test that the index doesn't overflow, since that's what the
mmap code really cares about.

HOWEVER.

Finding and fixing various random drivers is a sisyphean task, so let's
just see if we can just make the core mmap() code do the limiting for
us.  Realistically, the only "big" backing stores we need to care about
are regular files and block devices, both of which are known to do this
properly, and which have nice well-defined limits for how much data they
can access.

So let's special-case just those two known cases, and then limit other
random mmap users to a backing store that still fits in "unsigned long".
Realistically, that's not much of a limit at all on 64-bit, and on
32-bit architectures the only worry might be the GPU drivers, which can
have big physical address spaces.

To make it possible for drivers like that to say that they are 64-bit
clean, this patch does repurpose the "FMODE_UNSIGNED_OFFSET" bit in the
file flags to allow drivers to mark their file descriptors as safe in
the full 64-bit mmap address space.

[ The timing for doing this is less than optimal, and this should really
  go in a merge window. But realistically, this needs wide testing more
  than it needs anything else, and being main-line is the only way to do
  that.

  So the earlier the better, even if it's outside the proper development
  cycle        - Linus ]

Cc: Kees Cook <keescook@chromium.org>
Cc: Dan Carpenter <dan.carpenter@oracle.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Willy Tarreau <w@1wt.eu>
Cc: Dave Airlie <airlied@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 mm/mmap.c | 32 ++++++++++++++++++++++++++++++++
 1 file changed, 32 insertions(+)

diff --git a/mm/mmap.c b/mm/mmap.c
index 9d5968d..6fc4357 100644
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -1324,6 +1324,35 @@ static inline int mlock_future_check(struct mm_struct *mm,
 	return 0;
 }
 
+static inline u64 file_mmap_size_max(struct file *file, struct inode *inode)
+{
+	if (S_ISREG(inode->i_mode))
+		return inode->i_sb->s_maxbytes;
+
+	if (S_ISBLK(inode->i_mode))
+		return MAX_LFS_FILESIZE;
+
+	/* Special "we do even unsigned file positions" case */
+	if (file->f_mode & FMODE_UNSIGNED_OFFSET)
+		return 0;
+
+	/* Yes, random drivers might want more. But I'm tired of buggy drivers */
+	return ULONG_MAX;
+}
+
+static inline bool file_mmap_ok(struct file *file, struct inode *inode,
+				unsigned long pgoff, unsigned long len)
+{
+	u64 maxsize = file_mmap_size_max(file, inode);
+
+	if (maxsize && len > maxsize)
+		return false;
+	maxsize -= len;
+	if (pgoff > maxsize >> PAGE_SHIFT)
+		return false;
+	return true;
+}
+
 /*
  * The caller must hold down_write(&current->mm->mmap_sem).
  */
@@ -1409,6 +1438,9 @@ unsigned long do_mmap(struct file *file, unsigned long addr,
 		struct inode *inode = file_inode(file);
 		unsigned long flags_mask;
 
+		if (!file_mmap_ok(file, inode, pgoff, len))
+			return -EOVERFLOW;
+
 		flags_mask = LEGACY_MAP_MASK | file->f_op->mmap_supported_flags;
 
 		switch (flags & MAP_TYPE) {
-- 
cgit v1.1


From ac21fc2dcb405cf250ad3f1228f64f64930d9211 Mon Sep 17 00:00:00 2001
From: Rob Herring <robh@kernel.org>
Date: Fri, 11 May 2018 08:45:59 -0500
Subject: sh: switch to NO_BOOTMEM

Commit 0fa1c579349f ("of/fdt: use memblock_virt_alloc for early alloc")
inadvertently switched the DT unflattening allocations from memblock to
bootmem which doesn't work because the unflattening happens before
bootmem is initialized. Swapping the order of bootmem init and
unflattening could also fix this, but removing bootmem is desired. So
enable NO_BOOTMEM on SH like other architectures have done.

Fixes: 0fa1c579349f ("of/fdt: use memblock_virt_alloc for early alloc")
Reported-by: Rich Felker <dalias@libc.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Signed-off-by: Rob Herring <robh@kernel.org>
Signed-off-by: Rich Felker <dalias@libc.org>
---
 arch/sh/Kconfig        |  1 +
 arch/sh/kernel/setup.c |  1 -
 arch/sh/mm/init.c      | 68 +++++---------------------------------------------
 arch/sh/mm/numa.c      | 19 --------------
 4 files changed, 7 insertions(+), 82 deletions(-)

diff --git a/arch/sh/Kconfig b/arch/sh/Kconfig
index 97fe293..1851eae 100644
--- a/arch/sh/Kconfig
+++ b/arch/sh/Kconfig
@@ -9,6 +9,7 @@ config SUPERH
 	select HAVE_IDE if HAS_IOPORT_MAP
 	select HAVE_MEMBLOCK
 	select HAVE_MEMBLOCK_NODE_MAP
+	select NO_BOOTMEM
 	select ARCH_DISCARD_MEMBLOCK
 	select HAVE_OPROFILE
 	select HAVE_GENERIC_DMA_COHERENT
diff --git a/arch/sh/kernel/setup.c b/arch/sh/kernel/setup.c
index d34e998..c286cf5 100644
--- a/arch/sh/kernel/setup.c
+++ b/arch/sh/kernel/setup.c
@@ -11,7 +11,6 @@
 #include <linux/ioport.h>
 #include <linux/init.h>
 #include <linux/initrd.h>
-#include <linux/bootmem.h>
 #include <linux/console.h>
 #include <linux/root_dev.h>
 #include <linux/utsname.h>
diff --git a/arch/sh/mm/init.c b/arch/sh/mm/init.c
index ce0bbaa..4034035 100644
--- a/arch/sh/mm/init.c
+++ b/arch/sh/mm/init.c
@@ -211,59 +211,15 @@ void __init allocate_pgdat(unsigned int nid)
 
 	NODE_DATA(nid) = __va(phys);
 	memset(NODE_DATA(nid), 0, sizeof(struct pglist_data));
-
-	NODE_DATA(nid)->bdata = &bootmem_node_data[nid];
 #endif
 
 	NODE_DATA(nid)->node_start_pfn = start_pfn;
 	NODE_DATA(nid)->node_spanned_pages = end_pfn - start_pfn;
 }
 
-static void __init bootmem_init_one_node(unsigned int nid)
-{
-	unsigned long total_pages, paddr;
-	unsigned long end_pfn;
-	struct pglist_data *p;
-
-	p = NODE_DATA(nid);
-
-	/* Nothing to do.. */
-	if (!p->node_spanned_pages)
-		return;
-
-	end_pfn = pgdat_end_pfn(p);
-
-	total_pages = bootmem_bootmap_pages(p->node_spanned_pages);
-
-	paddr = memblock_alloc(total_pages << PAGE_SHIFT, PAGE_SIZE);
-	if (!paddr)
-		panic("Can't allocate bootmap for nid[%d]\n", nid);
-
-	init_bootmem_node(p, paddr >> PAGE_SHIFT, p->node_start_pfn, end_pfn);
-
-	free_bootmem_with_active_regions(nid, end_pfn);
-
-	/*
-	 * XXX Handle initial reservations for the system memory node
-	 * only for the moment, we'll refactor this later for handling
-	 * reservations in other nodes.
-	 */
-	if (nid == 0) {
-		struct memblock_region *reg;
-
-		/* Reserve the sections we're already using. */
-		for_each_memblock(reserved, reg) {
-			reserve_bootmem(reg->base, reg->size, BOOTMEM_DEFAULT);
-		}
-	}
-
-	sparse_memory_present_with_active_regions(nid);
-}
-
 static void __init do_init_bootmem(void)
 {
 	struct memblock_region *reg;
-	int i;
 
 	/* Add active regions with valid PFNs. */
 	for_each_memblock(memory, reg) {
@@ -279,9 +235,12 @@ static void __init do_init_bootmem(void)
 
 	plat_mem_setup();
 
-	for_each_online_node(i)
-		bootmem_init_one_node(i);
+	for_each_memblock(memory, reg) {
+		int nid = memblock_get_region_node(reg);
 
+		memory_present(nid, memblock_region_memory_base_pfn(reg),
+			memblock_region_memory_end_pfn(reg));
+	}
 	sparse_init();
 }
 
@@ -322,7 +281,6 @@ void __init paging_init(void)
 {
 	unsigned long max_zone_pfns[MAX_NR_ZONES];
 	unsigned long vaddr, end;
-	int nid;
 
 	sh_mv.mv_mem_init();
 
@@ -377,21 +335,7 @@ void __init paging_init(void)
 	kmap_coherent_init();
 
 	memset(max_zone_pfns, 0, sizeof(max_zone_pfns));
-
-	for_each_online_node(nid) {
-		pg_data_t *pgdat = NODE_DATA(nid);
-		unsigned long low, start_pfn;
-
-		start_pfn = pgdat->bdata->node_min_pfn;
-		low = pgdat->bdata->node_low_pfn;
-
-		if (max_zone_pfns[ZONE_NORMAL] < low)
-			max_zone_pfns[ZONE_NORMAL] = low;
-
-		printk("Node %u: start_pfn = 0x%lx, low = 0x%lx\n",
-		       nid, start_pfn, low);
-	}
-
+	max_zone_pfns[ZONE_NORMAL] = max_low_pfn;
 	free_area_init_nodes(max_zone_pfns);
 }
 
diff --git a/arch/sh/mm/numa.c b/arch/sh/mm/numa.c
index 05713d1..830e8b3 100644
--- a/arch/sh/mm/numa.c
+++ b/arch/sh/mm/numa.c
@@ -8,7 +8,6 @@
  * for more details.
  */
 #include <linux/module.h>
-#include <linux/bootmem.h>
 #include <linux/memblock.h>
 #include <linux/mm.h>
 #include <linux/numa.h>
@@ -26,9 +25,7 @@ EXPORT_SYMBOL_GPL(node_data);
  */
 void __init setup_bootmem_node(int nid, unsigned long start, unsigned long end)
 {
-	unsigned long bootmap_pages;
 	unsigned long start_pfn, end_pfn;
-	unsigned long bootmem_paddr;
 
 	/* Don't allow bogus node assignment */
 	BUG_ON(nid >= MAX_NUMNODES || nid <= 0);
@@ -48,25 +45,9 @@ void __init setup_bootmem_node(int nid, unsigned long start, unsigned long end)
 					     SMP_CACHE_BYTES, end));
 	memset(NODE_DATA(nid), 0, sizeof(struct pglist_data));
 
-	NODE_DATA(nid)->bdata = &bootmem_node_data[nid];
 	NODE_DATA(nid)->node_start_pfn = start_pfn;
 	NODE_DATA(nid)->node_spanned_pages = end_pfn - start_pfn;
 
-	/* Node-local bootmap */
-	bootmap_pages = bootmem_bootmap_pages(end_pfn - start_pfn);
-	bootmem_paddr = memblock_alloc_base(bootmap_pages << PAGE_SHIFT,
-				       PAGE_SIZE, end);
-	init_bootmem_node(NODE_DATA(nid), bootmem_paddr >> PAGE_SHIFT,
-			  start_pfn, end_pfn);
-
-	free_bootmem_with_active_regions(nid, end_pfn);
-
-	/* Reserve the pgdat and bootmap space with the bootmem allocator */
-	reserve_bootmem_node(NODE_DATA(nid), start_pfn << PAGE_SHIFT,
-			     sizeof(struct pglist_data), BOOTMEM_DEFAULT);
-	reserve_bootmem_node(NODE_DATA(nid), bootmem_paddr,
-			     bootmap_pages << PAGE_SHIFT, BOOTMEM_DEFAULT);
-
 	/* It's up */
 	node_set_online(nid);
 
-- 
cgit v1.1


From 04ac6fdba1afffad664377a324b017e63ac08bd8 Mon Sep 17 00:00:00 2001
From: Trond Myklebust <trondmy@gmail.com>
Date: Fri, 11 May 2018 14:13:57 -0400
Subject: Change Trond's email address in MAINTAINERS

Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
---
 MAINTAINERS | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/MAINTAINERS b/MAINTAINERS
index 4623caf..92e8db1 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -9744,7 +9744,7 @@ F:	include/linux/platform_data/nxp-nci.h
 F:	Documentation/devicetree/bindings/net/nfc/
 
 NFS, SUNRPC, AND LOCKD CLIENTS
-M:	Trond Myklebust <trond.myklebust@primarydata.com>
+M:	Trond Myklebust <trond.myklebust@hammerspace.com>
 M:	Anna Schumaker <anna.schumaker@netapp.com>
 L:	linux-nfs@vger.kernel.org
 W:	http://client.linux-nfs.org
-- 
cgit v1.1


From ea3beca422742ca7e8502496e2806421d5e9d6fd Mon Sep 17 00:00:00 2001
From: Jeff Shaw <jeffrey.b.shaw@intel.com>
Date: Wed, 18 Apr 2018 11:23:27 -0700
Subject: ice: Set rq_last_status when cleaning rq

Prior to this commit, the rq_last_status was only set when hardware
responded with an error. This leads to rq_last_status being invalid
in the future when hardware eventually responds without error. This
commit resolves the issue by unconditionally setting rq_last_status
with the value returned in the descriptor.

Fixes: 940b61af02f4 ("ice: Initialize PF and setup miscellaneous
interrupt")

Signed-off-by: Jeff Shaw <jeffrey.b.shaw@intel.com>
Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Tested-by: Tony Brelinski <tonyx.brelinski@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/ice/ice_controlq.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/intel/ice/ice_controlq.c b/drivers/net/ethernet/intel/ice/ice_controlq.c
index 5909a44..7c511f1 100644
--- a/drivers/net/ethernet/intel/ice/ice_controlq.c
+++ b/drivers/net/ethernet/intel/ice/ice_controlq.c
@@ -1014,10 +1014,10 @@ ice_clean_rq_elem(struct ice_hw *hw, struct ice_ctl_q_info *cq,
 	desc = ICE_CTL_Q_DESC(cq->rq, ntc);
 	desc_idx = ntc;
 
+	cq->rq_last_status = (enum ice_aq_err)le16_to_cpu(desc->retval);
 	flags = le16_to_cpu(desc->flags);
 	if (flags & ICE_AQ_FLAG_ERR) {
 		ret_code = ICE_ERR_AQ_ERROR;
-		cq->rq_last_status = (enum ice_aq_err)le16_to_cpu(desc->retval);
 		ice_debug(hw, ICE_DBG_AQ_MSG,
 			  "Control Receive Queue Event received with error 0x%x\n",
 			  cq->rq_last_status);
-- 
cgit v1.1


From bbb2707623f3ccc48695da2433f06d7c38193451 Mon Sep 17 00:00:00 2001
From: Emil Tantilov <emil.s.tantilov@intel.com>
Date: Thu, 19 Apr 2018 17:06:57 -0700
Subject: ixgbe: return error on unsupported SFP module when resetting

Add check for unsupported module and return the error code.
This fixes a Coverity hit due to unused return status from setup_sfp.

Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/ixgbe/ixgbe_x550.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_x550.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_x550.c
index 3123267..9592f3e 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_x550.c
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_x550.c
@@ -3427,6 +3427,9 @@ static s32 ixgbe_reset_hw_X550em(struct ixgbe_hw *hw)
 		hw->phy.sfp_setup_needed = false;
 	}
 
+	if (status == IXGBE_ERR_SFP_NOT_SUPPORTED)
+		return status;
+
 	/* Reset PHY */
 	if (!hw->phy.reset_disable && hw->phy.ops.reset)
 		hw->phy.ops.reset(hw);
-- 
cgit v1.1


From cf12aab67a2fb29342d833a6333a322aaccca2d0 Mon Sep 17 00:00:00 2001
From: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Date: Tue, 24 Apr 2018 15:16:48 +0200
Subject: ixgbevf: fix ixgbevf_xmit_frame()'s return type

The method ndo_start_xmit() is defined as returning an 'netdev_tx_t',
which is a typedef for an enum type, but the implementation in this
driver returns an 'int'.

Fix this by returning 'netdev_tx_t' in this driver too.

Signed-off-by: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c
index e3d04f2..850f8af 100644
--- a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c
+++ b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c
@@ -4137,7 +4137,7 @@ out_drop:
 	return NETDEV_TX_OK;
 }
 
-static int ixgbevf_xmit_frame(struct sk_buff *skb, struct net_device *netdev)
+static netdev_tx_t ixgbevf_xmit_frame(struct sk_buff *skb, struct net_device *netdev)
 {
 	struct ixgbevf_adapter *adapter = netdev_priv(netdev);
 	struct ixgbevf_ring *tx_ring;
-- 
cgit v1.1


From c89ebb968f04c71e16e86c91caeacb045dc8f908 Mon Sep 17 00:00:00 2001
From: Colin Ian King <colin.king@canonical.com>
Date: Wed, 9 May 2018 14:58:48 +0100
Subject: ixgbe: fix memory leak on ipsec allocation

The error clean up path kfree's adapter->ipsec and should be
instead kfree'ing ipsec. Fix this.  Also, the err1 error exit path
does not need to kfree ipsec because this failure path was for
the failed allocation of ipsec.

Detected by CoverityScan, CID#146424 ("Resource Leak")

Fixes: 63a67fe229ea ("ixgbe: add ipsec offload add and remove SA")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Acked-by: Shannon Nelson <shannon.nelson@oracle.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/ixgbe/ixgbe_ipsec.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_ipsec.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_ipsec.c
index 68af127..cead23e 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_ipsec.c
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_ipsec.c
@@ -943,8 +943,8 @@ err2:
 	kfree(ipsec->ip_tbl);
 	kfree(ipsec->rx_tbl);
 	kfree(ipsec->tx_tbl);
+	kfree(ipsec);
 err1:
-	kfree(adapter->ipsec);
 	netdev_err(adapter->netdev, "Unable to allocate memory for SA tables");
 }
 
-- 
cgit v1.1


From 9abd68ef454c824bfd18629033367b4382b5f390 Mon Sep 17 00:00:00 2001
From: Jens Axboe <axboe@kernel.dk>
Date: Tue, 8 May 2018 10:25:15 -0600
Subject: nvme: add quirk to force medium priority for SQ creation

Some P3100 drives have a bug where they think WRRU (weighted round robin)
is always enabled, even though the host doesn't set it. Since they think
it's enabled, they also look at the submission queue creation priority. We
used to set that to MEDIUM by default, but that was removed in commit
81c1cd98351b. This causes various issues on that drive. Add a quirk to
still set MEDIUM priority for that controller.

Fixes: 81c1cd98351b ("nvme/pci: Don't set reserved SQ create flags")
Cc: stable@vger.kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Keith Busch <keith.busch@intel.com>
---
 drivers/nvme/host/nvme.h |  5 +++++
 drivers/nvme/host/pci.c  | 12 +++++++++++-
 2 files changed, 16 insertions(+), 1 deletion(-)

diff --git a/drivers/nvme/host/nvme.h b/drivers/nvme/host/nvme.h
index 7ded7a5..17d2f7c 100644
--- a/drivers/nvme/host/nvme.h
+++ b/drivers/nvme/host/nvme.h
@@ -84,6 +84,11 @@ enum nvme_quirks {
 	 * Supports the LighNVM command set if indicated in vs[1].
 	 */
 	NVME_QUIRK_LIGHTNVM			= (1 << 6),
+
+	/*
+	 * Set MEDIUM priority on SQ creation
+	 */
+	NVME_QUIRK_MEDIUM_PRIO_SQ		= (1 << 7),
 };
 
 /*
diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index fbc71fa..17a0190 100644
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -1093,10 +1093,19 @@ static int adapter_alloc_cq(struct nvme_dev *dev, u16 qid,
 static int adapter_alloc_sq(struct nvme_dev *dev, u16 qid,
 						struct nvme_queue *nvmeq)
 {
+	struct nvme_ctrl *ctrl = &dev->ctrl;
 	struct nvme_command c;
 	int flags = NVME_QUEUE_PHYS_CONTIG;
 
 	/*
+	 * Some drives have a bug that auto-enables WRRU if MEDIUM isn't
+	 * set. Since URGENT priority is zeroes, it makes all queues
+	 * URGENT.
+	 */
+	if (ctrl->quirks & NVME_QUIRK_MEDIUM_PRIO_SQ)
+		flags |= NVME_SQ_PRIO_MEDIUM;
+
+	/*
 	 * Note: we (ab)use the fact that the prp fields survive if no data
 	 * is attached to the request.
 	 */
@@ -2701,7 +2710,8 @@ static const struct pci_device_id nvme_id_table[] = {
 		.driver_data = NVME_QUIRK_STRIPE_SIZE |
 				NVME_QUIRK_DEALLOCATE_ZEROES, },
 	{ PCI_VDEVICE(INTEL, 0xf1a5),	/* Intel 600P/P3100 */
-		.driver_data = NVME_QUIRK_NO_DEEPEST_PS },
+		.driver_data = NVME_QUIRK_NO_DEEPEST_PS |
+				NVME_QUIRK_MEDIUM_PRIO_SQ },
 	{ PCI_VDEVICE(INTEL, 0x5845),	/* Qemu emulated controller */
 		.driver_data = NVME_QUIRK_IDENTIFY_CNS, },
 	{ PCI_DEVICE(0x1c58, 0x0003),	/* HGST adapter */
-- 
cgit v1.1


From af5d01842fe1fbfb9f5e1c1d957ba02ab6f4569a Mon Sep 17 00:00:00 2001
From: Roman Mashak <mrv@mojatatu.com>
Date: Fri, 11 May 2018 10:55:09 -0400
Subject: net sched actions: fix invalid pointer dereferencing if skbedit flags
 missing

When application fails to pass flags in netlink TLV for a new skbedit action,
the kernel results in the following oops:

[    8.307732] BUG: unable to handle kernel paging request at 0000000000021130
[    8.309167] PGD 80000000193d1067 P4D 80000000193d1067 PUD 180e0067 PMD 0
[    8.310595] Oops: 0000 [#1] SMP PTI
[    8.311334] Modules linked in: kvm_intel kvm irqbypass crct10dif_pclmul crc32_pclmul ghash_clmulni_intel pcbc aesni_intel aes_x86_64 crypto_simd cryptd glue_helper serio_raw
[    8.314190] CPU: 1 PID: 397 Comm: tc Not tainted 4.17.0-rc3+ #357
[    8.315252] RIP: 0010:__tcf_idr_release+0x33/0x140
[    8.316203] RSP: 0018:ffffa0718038f840 EFLAGS: 00010246
[    8.317123] RAX: 0000000000000001 RBX: 0000000000021100 RCX: 0000000000000000
[    8.319831] RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000021100
[    8.321181] RBP: 0000000000000000 R08: 000000000004adf8 R09: 0000000000000122
[    8.322645] R10: 0000000000000000 R11: ffffffff9e5b01ed R12: 0000000000000000
[    8.324157] R13: ffffffff9e0d3cc0 R14: 0000000000000000 R15: 0000000000000000
[    8.325590] FS:  00007f591292e700(0000) GS:ffff8fcf5bc40000(0000) knlGS:0000000000000000
[    8.327001] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[    8.327987] CR2: 0000000000021130 CR3: 00000000180e6004 CR4: 00000000001606a0
[    8.329289] Call Trace:
[    8.329735]  tcf_skbedit_init+0xa7/0xb0
[    8.330423]  tcf_action_init_1+0x362/0x410
[    8.331139]  ? try_to_wake_up+0x44/0x430
[    8.331817]  tcf_action_init+0x103/0x190
[    8.332511]  tc_ctl_action+0x11a/0x220
[    8.333174]  rtnetlink_rcv_msg+0x23d/0x2e0
[    8.333902]  ? _cond_resched+0x16/0x40
[    8.334569]  ? __kmalloc_node_track_caller+0x5b/0x2c0
[    8.335440]  ? rtnl_calcit.isra.31+0xf0/0xf0
[    8.336178]  netlink_rcv_skb+0xdb/0x110
[    8.336855]  netlink_unicast+0x167/0x220
[    8.337550]  netlink_sendmsg+0x2a7/0x390
[    8.338258]  sock_sendmsg+0x30/0x40
[    8.338865]  ___sys_sendmsg+0x2c5/0x2e0
[    8.339531]  ? pagecache_get_page+0x27/0x210
[    8.340271]  ? filemap_fault+0xa2/0x630
[    8.340943]  ? page_add_file_rmap+0x108/0x200
[    8.341732]  ? alloc_set_pte+0x2aa/0x530
[    8.342573]  ? finish_fault+0x4e/0x70
[    8.343332]  ? __handle_mm_fault+0xbc1/0x10d0
[    8.344337]  ? __sys_sendmsg+0x53/0x80
[    8.345040]  __sys_sendmsg+0x53/0x80
[    8.345678]  do_syscall_64+0x4f/0x100
[    8.346339]  entry_SYSCALL_64_after_hwframe+0x44/0xa9
[    8.347206] RIP: 0033:0x7f591191da67
[    8.347831] RSP: 002b:00007fff745abd48 EFLAGS: 00000246 ORIG_RAX: 000000000000002e
[    8.349179] RAX: ffffffffffffffda RBX: 00007fff745abe70 RCX: 00007f591191da67
[    8.350431] RDX: 0000000000000000 RSI: 00007fff745abdc0 RDI: 0000000000000003
[    8.351659] RBP: 000000005af35251 R08: 0000000000000001 R09: 0000000000000000
[    8.352922] R10: 00000000000005f1 R11: 0000000000000246 R12: 0000000000000000
[    8.354183] R13: 00007fff745afed0 R14: 0000000000000001 R15: 00000000006767c0
[    8.355400] Code: 41 89 d4 53 89 f5 48 89 fb e8 aa 20 fd ff 85 c0 0f 84 ed 00
00 00 48 85 db 0f 84 cf 00 00 00 40 84 ed 0f 85 cd 00 00 00 45 84 e4 <8b> 53 30
74 0d 85 d2 b8 ff ff ff ff 0f 8f b3 00 00 00 8b 43 2c
[    8.358699] RIP: __tcf_idr_release+0x33/0x140 RSP: ffffa0718038f840
[    8.359770] CR2: 0000000000021130
[    8.360438] ---[ end trace 60c66be45dfc14f0 ]---

The caller calls action's ->init() and passes pointer to "struct tc_action *a",
which later may be initialized to point at the existing action, otherwise
"struct tc_action *a" is still invalid, and therefore dereferencing it is an
error as happens in tcf_idr_release, where refcnt is decremented.

So in case of missing flags tcf_idr_release must be called only for
existing actions.

v2:
    - prepare patch for net tree

Fixes: 5e1567aeb7fe ("net sched: skbedit action fix late binding")
Signed-off-by: Roman Mashak <mrv@mojatatu.com>
Acked-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/sched/act_skbedit.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/net/sched/act_skbedit.c b/net/sched/act_skbedit.c
index ddf69fc..6138d1d7 100644
--- a/net/sched/act_skbedit.c
+++ b/net/sched/act_skbedit.c
@@ -121,7 +121,8 @@ static int tcf_skbedit_init(struct net *net, struct nlattr *nla,
 		return 0;
 
 	if (!flags) {
-		tcf_idr_release(*a, bind);
+		if (exists)
+			tcf_idr_release(*a, bind);
 		return -EINVAL;
 	}
 
-- 
cgit v1.1


From d68d75fdc34b0253c2bded7ed18cd60eb5a9599b Mon Sep 17 00:00:00 2001
From: Jiri Pirko <jiri@mellanox.com>
Date: Fri, 11 May 2018 17:45:32 +0200
Subject: net: sched: fix error path in tcf_proto_create() when modules are not
 configured

In case modules are not configured, error out when tp->ops is null
and prevent later null pointer dereference.

Fixes: 33a48927c193 ("sched: push TC filter protocol creation into a separate function")
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Acked-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/sched/cls_api.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/sched/cls_api.c b/net/sched/cls_api.c
index b66754f..963e4bf 100644
--- a/net/sched/cls_api.c
+++ b/net/sched/cls_api.c
@@ -152,8 +152,8 @@ static struct tcf_proto *tcf_proto_create(const char *kind, u32 protocol,
 			NL_SET_ERR_MSG(extack, "TC classifier not found");
 			err = -ENOENT;
 		}
-		goto errout;
 #endif
+		goto errout;
 	}
 	tp->classify = tp->ops->classify;
 	tp->protocol = protocol;
-- 
cgit v1.1


From a52956dfc503f8cc5cfe6454959b7049fddb4413 Mon Sep 17 00:00:00 2001
From: Roman Mashak <mrv@mojatatu.com>
Date: Fri, 11 May 2018 14:35:33 -0400
Subject: net sched actions: fix refcnt leak in skbmod

When application fails to pass flags in netlink TLV when replacing
existing skbmod action, the kernel will leak refcnt:

$ tc actions get action skbmod index 1
total acts 0

        action order 0: skbmod pipe set smac 00:11:22:33:44:55
         index 1 ref 1 bind 0

For example, at this point a buggy application replaces the action with
index 1 with new smac 00:aa:22:33:44:55, it fails because of zero flags,
however refcnt gets bumped:

$ tc actions get actions skbmod index 1
total acts 0

        action order 0: skbmod pipe set smac 00:11:22:33:44:55
         index 1 ref 2 bind 0
$

Tha patch fixes this by calling tcf_idr_release() on existing actions.

Fixes: 86da71b57383d ("net_sched: Introduce skbmod action")
Signed-off-by: Roman Mashak <mrv@mojatatu.com>
Acked-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/sched/act_skbmod.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/net/sched/act_skbmod.c b/net/sched/act_skbmod.c
index bbcbdce..ad050d7 100644
--- a/net/sched/act_skbmod.c
+++ b/net/sched/act_skbmod.c
@@ -131,8 +131,11 @@ static int tcf_skbmod_init(struct net *net, struct nlattr *nla,
 	if (exists && bind)
 		return 0;
 
-	if (!lflags)
+	if (!lflags) {
+		if (exists)
+			tcf_idr_release(*a, bind);
 		return -EINVAL;
+	}
 
 	if (!exists) {
 		ret = tcf_idr_create(tn, parm->index, est, a,
-- 
cgit v1.1


From 1d1c8e5f0d95c2c97f29122741aed7d0dfb2b391 Mon Sep 17 00:00:00 2001
From: "Shuah Khan (Samsung OSG)" <shuah@kernel.org>
Date: Fri, 11 May 2018 16:01:32 -0700
Subject: MAINTAINERS: update Shuah's email address

Update email address in MAINTAINERS file due to IT infrastructure changes
at Samsung.

Link: http://lkml.kernel.org/r/20180501212815.25911-1-shuah@kernel.org
Signed-off-by: Shuah Khan (Samsung OSG) <shuah@kernel.org>
Cc: Mauro Carvalho Chehab <mchehab@kernel.org>
Cc: David S. Miller <davem@davemloft.net>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Linus Walleij <linus.walleij@linaro.org>
Cc: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 MAINTAINERS | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/MAINTAINERS b/MAINTAINERS
index df6e9bb..b495113 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -3691,7 +3691,6 @@ F:	drivers/cpufreq/arm_big_little_dt.c
 
 CPU POWER MONITORING SUBSYSTEM
 M:	Thomas Renninger <trenn@suse.com>
-M:	Shuah Khan <shuahkh@osg.samsung.com>
 M:	Shuah Khan <shuah@kernel.org>
 L:	linux-pm@vger.kernel.org
 S:	Maintained
@@ -7696,7 +7695,6 @@ F:	include/linux/sunrpc/
 F:	include/uapi/linux/sunrpc/
 
 KERNEL SELFTEST FRAMEWORK
-M:	Shuah Khan <shuahkh@osg.samsung.com>
 M:	Shuah Khan <shuah@kernel.org>
 L:	linux-kselftest@vger.kernel.org
 T:	git git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest.git
@@ -14650,7 +14648,6 @@ F:	drivers/usb/common/usb-otg-fsm.c
 
 USB OVER IP DRIVER
 M:	Valentina Manea <valentina.manea.m@gmail.com>
-M:	Shuah Khan <shuahkh@osg.samsung.com>
 M:	Shuah Khan <shuah@kernel.org>
 L:	linux-usb@vger.kernel.org
 S:	Maintained
-- 
cgit v1.1


From c9cf87ea6a4725bc586cc5776c3fb6042a264b7d Mon Sep 17 00:00:00 2001
From: Dmitry Vyukov <dvyukov@google.com>
Date: Fri, 11 May 2018 16:01:35 -0700
Subject: KASAN: prohibit KASAN+STRUCTLEAK combination

Currently STRUCTLEAK inserts initialization out of live scope of variables
from KASAN point of view.  This leads to KASAN false positive reports.
Prohibit this combination for now.

Link: http://lkml.kernel.org/r/20180419172451.104700-1-dvyukov@google.com
Signed-off-by: Dmitry Vyukov <dvyukov@google.com>
Acked-by: Kees Cook <keescook@chromium.org>
Cc: Fengguang Wu <fengguang.wu@intel.com>
Cc: Sergey Senozhatsky <sergey.senozhatsky.work@gmail.com>
Cc: Andrey Ryabinin <aryabinin@virtuozzo.com>
Cc: Dennis Zhou <dennisszhou@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 arch/Kconfig | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/arch/Kconfig b/arch/Kconfig
index 8e0d665..75dd23a 100644
--- a/arch/Kconfig
+++ b/arch/Kconfig
@@ -464,6 +464,10 @@ config GCC_PLUGIN_LATENT_ENTROPY
 config GCC_PLUGIN_STRUCTLEAK
 	bool "Force initialization of variables containing userspace addresses"
 	depends on GCC_PLUGINS
+	# Currently STRUCTLEAK inserts initialization out of live scope of
+	# variables from KASAN point of view. This leads to KASAN false
+	# positive reports. Prohibit this combination for now.
+	depends on !KASAN_EXTRA
 	help
 	  This plugin zero-initializes any structures containing a
 	  __user attribute. This can prevent some classes of information
-- 
cgit v1.1


From 4ba281d5bd9907355e6b79fb72049c9ed50cc670 Mon Sep 17 00:00:00 2001
From: Yury Norov <ynorov@caviumnetworks.com>
Date: Fri, 11 May 2018 16:01:39 -0700
Subject: lib/find_bit_benchmark.c: avoid soft lockup in test_find_first_bit()

test_find_first_bit() is intentionally sub-optimal, and may cause soft
lockup due to long time of run on some systems.  So decrease length of
bitmap to traverse to avoid lockup.

With the change below, time of test execution doesn't exceed 0.2 seconds
on my testing system.

Link: http://lkml.kernel.org/r/20180420171949.15710-1-ynorov@caviumnetworks.com
Fixes: 4441fca0a27f5 ("lib: test module for find_*_bit() functions")
Signed-off-by: Yury Norov <ynorov@caviumnetworks.com>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 lib/find_bit_benchmark.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/lib/find_bit_benchmark.c b/lib/find_bit_benchmark.c
index 5985a25..5367ffa 100644
--- a/lib/find_bit_benchmark.c
+++ b/lib/find_bit_benchmark.c
@@ -132,7 +132,12 @@ static int __init find_bit_test(void)
 	test_find_next_bit(bitmap, BITMAP_LEN);
 	test_find_next_zero_bit(bitmap, BITMAP_LEN);
 	test_find_last_bit(bitmap, BITMAP_LEN);
-	test_find_first_bit(bitmap, BITMAP_LEN);
+
+	/*
+	 * test_find_first_bit() may take some time, so
+	 * traverse only part of bitmap to avoid soft lockup.
+	 */
+	test_find_first_bit(bitmap, BITMAP_LEN / 10);
 	test_find_next_and_bit(bitmap, bitmap2, BITMAP_LEN);
 
 	pr_err("\nStart testing find_bit() with sparse bitmap\n");
-- 
cgit v1.1


From ae646f0b9ca135b87bc73ff606ef996c3029780a Mon Sep 17 00:00:00 2001
From: Jeffrey Hugo <jhugo@codeaurora.org>
Date: Fri, 11 May 2018 16:01:42 -0700
Subject: init: fix false positives in W+X checking

load_module() creates W+X mappings via __vmalloc_node_range() (from
layout_and_allocate()->move_module()->module_alloc()) by using
PAGE_KERNEL_EXEC.  These mappings are later cleaned up via
"call_rcu_sched(&freeinit->rcu, do_free_init)" from do_init_module().

This is a problem because call_rcu_sched() queues work, which can be run
after debug_checkwx() is run, resulting in a race condition.  If hit,
the race results in a nasty splat about insecure W+X mappings, which
results in a poor user experience as these are not the mappings that
debug_checkwx() is intended to catch.

This issue is observed on multiple arm64 platforms, and has been
artificially triggered on an x86 platform.

Address the race by flushing the queued work before running the
arch-defined mark_rodata_ro() which then calls debug_checkwx().

Link: http://lkml.kernel.org/r/1525103946-29526-1-git-send-email-jhugo@codeaurora.org
Fixes: e1a58320a38d ("x86/mm: Warn on W^X mappings")
Signed-off-by: Jeffrey Hugo <jhugo@codeaurora.org>
Reported-by: Timur Tabi <timur@codeaurora.org>
Reported-by: Jan Glauber <jan.glauber@caviumnetworks.com>
Acked-by: Kees Cook <keescook@chromium.org>
Acked-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Will Deacon <will.deacon@arm.com>
Acked-by: Laura Abbott <labbott@redhat.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Stephen Smalley <sds@tycho.nsa.gov>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 init/main.c     | 7 +++++++
 kernel/module.c | 5 +++++
 2 files changed, 12 insertions(+)

diff --git a/init/main.c b/init/main.c
index a404936..fd37315 100644
--- a/init/main.c
+++ b/init/main.c
@@ -1034,6 +1034,13 @@ __setup("rodata=", set_debug_rodata);
 static void mark_readonly(void)
 {
 	if (rodata_enabled) {
+		/*
+		 * load_module() results in W+X mappings, which are cleaned up
+		 * with call_rcu_sched().  Let's make sure that queued work is
+		 * flushed so that we don't hit false positives looking for
+		 * insecure pages which are W+X.
+		 */
+		rcu_barrier_sched();
 		mark_rodata_ro();
 		rodata_test();
 	} else
diff --git a/kernel/module.c b/kernel/module.c
index ce8066b..c9bea7f 100644
--- a/kernel/module.c
+++ b/kernel/module.c
@@ -3517,6 +3517,11 @@ static noinline int do_init_module(struct module *mod)
 	 * walking this with preempt disabled.  In all the failure paths, we
 	 * call synchronize_sched(), but we don't want to slow down the success
 	 * path, so use actual RCU here.
+	 * Note that module_alloc() on most architectures creates W+X page
+	 * mappings which won't be cleaned up until do_free_init() runs.  Any
+	 * code such as mark_rodata_ro() which depends on those mappings to
+	 * be cleaned up needs to sync with the queued work - ie
+	 * rcu_barrier_sched()
 	 */
 	call_rcu_sched(&freeinit->rcu, do_free_init);
 	mutex_unlock(&module_mutex);
-- 
cgit v1.1


From 6098d7e136692f9c6e23ae362c62ec822343e4d5 Mon Sep 17 00:00:00 2001
From: Vitaly Wool <vitalywool@gmail.com>
Date: Fri, 11 May 2018 16:01:46 -0700
Subject: z3fold: fix reclaim lock-ups

Do not try to optimize in-page object layout while the page is under
reclaim.  This fixes lock-ups on reclaim and improves reclaim
performance at the same time.

[akpm@linux-foundation.org: coding-style fixes]
Link: http://lkml.kernel.org/r/20180430125800.444cae9706489f412ad12621@gmail.com
Signed-off-by: Vitaly Wool <vitaly.vul@sony.com>
Reported-by: Guenter Roeck <linux@roeck-us.net>
Tested-by: Guenter Roeck <linux@roeck-us.net>
Cc: <Oleksiy.Avramchenko@sony.com>
Cc: Matthew Wilcox <mawilcox@microsoft.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 mm/z3fold.c | 42 ++++++++++++++++++++++++++++++------------
 1 file changed, 30 insertions(+), 12 deletions(-)

diff --git a/mm/z3fold.c b/mm/z3fold.c
index c0bca61..4b366d1 100644
--- a/mm/z3fold.c
+++ b/mm/z3fold.c
@@ -144,7 +144,8 @@ enum z3fold_page_flags {
 	PAGE_HEADLESS = 0,
 	MIDDLE_CHUNK_MAPPED,
 	NEEDS_COMPACTING,
-	PAGE_STALE
+	PAGE_STALE,
+	UNDER_RECLAIM
 };
 
 /*****************
@@ -173,6 +174,7 @@ static struct z3fold_header *init_z3fold_page(struct page *page,
 	clear_bit(MIDDLE_CHUNK_MAPPED, &page->private);
 	clear_bit(NEEDS_COMPACTING, &page->private);
 	clear_bit(PAGE_STALE, &page->private);
+	clear_bit(UNDER_RECLAIM, &page->private);
 
 	spin_lock_init(&zhdr->page_lock);
 	kref_init(&zhdr->refcount);
@@ -756,6 +758,10 @@ static void z3fold_free(struct z3fold_pool *pool, unsigned long handle)
 		atomic64_dec(&pool->pages_nr);
 		return;
 	}
+	if (test_bit(UNDER_RECLAIM, &page->private)) {
+		z3fold_page_unlock(zhdr);
+		return;
+	}
 	if (test_and_set_bit(NEEDS_COMPACTING, &page->private)) {
 		z3fold_page_unlock(zhdr);
 		return;
@@ -840,6 +846,8 @@ static int z3fold_reclaim_page(struct z3fold_pool *pool, unsigned int retries)
 			kref_get(&zhdr->refcount);
 			list_del_init(&zhdr->buddy);
 			zhdr->cpu = -1;
+			set_bit(UNDER_RECLAIM, &page->private);
+			break;
 		}
 
 		list_del_init(&page->lru);
@@ -887,25 +895,35 @@ static int z3fold_reclaim_page(struct z3fold_pool *pool, unsigned int retries)
 				goto next;
 		}
 next:
-		spin_lock(&pool->lock);
 		if (test_bit(PAGE_HEADLESS, &page->private)) {
 			if (ret == 0) {
-				spin_unlock(&pool->lock);
 				free_z3fold_page(page);
 				return 0;
 			}
-		} else if (kref_put(&zhdr->refcount, release_z3fold_page)) {
-			atomic64_dec(&pool->pages_nr);
+			spin_lock(&pool->lock);
+			list_add(&page->lru, &pool->lru);
+			spin_unlock(&pool->lock);
+		} else {
+			z3fold_page_lock(zhdr);
+			clear_bit(UNDER_RECLAIM, &page->private);
+			if (kref_put(&zhdr->refcount,
+					release_z3fold_page_locked)) {
+				atomic64_dec(&pool->pages_nr);
+				return 0;
+			}
+			/*
+			 * if we are here, the page is still not completely
+			 * free. Take the global pool lock then to be able
+			 * to add it back to the lru list
+			 */
+			spin_lock(&pool->lock);
+			list_add(&page->lru, &pool->lru);
 			spin_unlock(&pool->lock);
-			return 0;
+			z3fold_page_unlock(zhdr);
 		}
 
-		/*
-		 * Add to the beginning of LRU.
-		 * Pool lock has to be kept here to ensure the page has
-		 * not already been released
-		 */
-		list_add(&page->lru, &pool->lru);
+		/* We started off locked to we need to lock the pool back */
+		spin_lock(&pool->lock);
 	}
 	spin_unlock(&pool->lock);
 	return -EAGAIN;
-- 
cgit v1.1


From 27227c733852f71008e9bf165950bb2edaed3a90 Mon Sep 17 00:00:00 2001
From: Pavel Tatashin <pasha.tatashin@oracle.com>
Date: Fri, 11 May 2018 16:01:50 -0700
Subject: mm: sections are not offlined during memory hotremove

Memory hotplug and hotremove operate with per-block granularity.  If the
machine has a large amount of memory (more than 64G), the size of a
memory block can span multiple sections.  By mistake, during hotremove
we set only the first section to offline state.

The bug was discovered because kernel selftest started to fail:
  https://lkml.kernel.org/r/20180423011247.GK5563@yexl-desktop

After commit, "mm/memory_hotplug: optimize probe routine".  But, the bug
is older than this commit.  In this optimization we also added a check
for sections to be in a proper state during hotplug operation.

Link: http://lkml.kernel.org/r/20180427145257.15222-1-pasha.tatashin@oracle.com
Fixes: 2d070eab2e82 ("mm: consider zone which is not fully populated to have holes")
Signed-off-by: Pavel Tatashin <pasha.tatashin@oracle.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Steven Sistare <steven.sistare@oracle.com>
Cc: Daniel Jordan <daniel.m.jordan@oracle.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 mm/sparse.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/mm/sparse.c b/mm/sparse.c
index 62eef26..73dc2fc 100644
--- a/mm/sparse.c
+++ b/mm/sparse.c
@@ -629,7 +629,7 @@ void offline_mem_sections(unsigned long start_pfn, unsigned long end_pfn)
 	unsigned long pfn;
 
 	for (pfn = start_pfn; pfn < end_pfn; pfn += PAGES_PER_SECTION) {
-		unsigned long section_nr = pfn_to_section_nr(start_pfn);
+		unsigned long section_nr = pfn_to_section_nr(pfn);
 		struct mem_section *ms;
 
 		/*
-- 
cgit v1.1


From 7aaf7727235870f497eb928f728f7773d6df3b40 Mon Sep 17 00:00:00 2001
From: Roman Gushchin <guro@fb.com>
Date: Fri, 11 May 2018 16:01:53 -0700
Subject: mm: don't show nr_indirectly_reclaimable in /proc/vmstat

Don't show nr_indirectly_reclaimable in /proc/vmstat, because there is
no need to export this vm counter to userspace, and some changes are
expected in reclaimable object accounting, which can alter this counter.

Link: http://lkml.kernel.org/r/20180425191422.9159-1-guro@fb.com
Signed-off-by: Roman Gushchin <guro@fb.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: David Rientjes <rientjes@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 mm/vmstat.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/mm/vmstat.c b/mm/vmstat.c
index 536332e..a2b9518 100644
--- a/mm/vmstat.c
+++ b/mm/vmstat.c
@@ -1161,7 +1161,7 @@ const char * const vmstat_text[] = {
 	"nr_vmscan_immediate_reclaim",
 	"nr_dirtied",
 	"nr_written",
-	"nr_indirectly_reclaimable",
+	"", /* nr_indirectly_reclaimable */
 
 	/* enum writeback_stat_item counters */
 	"nr_dirty_threshold",
@@ -1740,6 +1740,10 @@ static int vmstat_show(struct seq_file *m, void *arg)
 	unsigned long *l = arg;
 	unsigned long off = l - (unsigned long *)m->private;
 
+	/* Skip hidden vmstat items. */
+	if (*vmstat_text[off] == '\0')
+		return 0;
+
 	seq_puts(m, vmstat_text[off]);
 	seq_put_decimal_ull(m, " ", *l);
 	seq_putc(m, '\n');
-- 
cgit v1.1


From 3955333df9a50e8783d115613a397ae55d905080 Mon Sep 17 00:00:00 2001
From: Laura Abbott <labbott@redhat.com>
Date: Fri, 11 May 2018 16:01:57 -0700
Subject: proc/kcore: don't bounds check against address 0

The existing kcore code checks for bad addresses against __va(0) with
the assumption that this is the lowest address on the system.  This may
not hold true on some systems (e.g.  arm64) and produce overflows and
crashes.  Switch to using other functions to validate the address range.

It's currently only seen on arm64 and it's not clear if anyone wants to
use that particular combination on a stable release.  So this is not
urgent for stable.

Link: http://lkml.kernel.org/r/20180501201143.15121-1-labbott@redhat.com
Signed-off-by: Laura Abbott <labbott@redhat.com>
Tested-by: Dave Anderson <anderson@redhat.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Alexey Dobriyan <adobriyan@gmail.com>a
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 fs/proc/kcore.c | 23 ++++++++++++++++-------
 1 file changed, 16 insertions(+), 7 deletions(-)

diff --git a/fs/proc/kcore.c b/fs/proc/kcore.c
index d1e8276..e64ecb9 100644
--- a/fs/proc/kcore.c
+++ b/fs/proc/kcore.c
@@ -209,25 +209,34 @@ kclist_add_private(unsigned long pfn, unsigned long nr_pages, void *arg)
 {
 	struct list_head *head = (struct list_head *)arg;
 	struct kcore_list *ent;
+	struct page *p;
+
+	if (!pfn_valid(pfn))
+		return 1;
+
+	p = pfn_to_page(pfn);
+	if (!memmap_valid_within(pfn, p, page_zone(p)))
+		return 1;
 
 	ent = kmalloc(sizeof(*ent), GFP_KERNEL);
 	if (!ent)
 		return -ENOMEM;
-	ent->addr = (unsigned long)__va((pfn << PAGE_SHIFT));
+	ent->addr = (unsigned long)page_to_virt(p);
 	ent->size = nr_pages << PAGE_SHIFT;
 
-	/* Sanity check: Can happen in 32bit arch...maybe */
-	if (ent->addr < (unsigned long) __va(0))
+	if (!virt_addr_valid(ent->addr))
 		goto free_out;
 
 	/* cut not-mapped area. ....from ppc-32 code. */
 	if (ULONG_MAX - ent->addr < ent->size)
 		ent->size = ULONG_MAX - ent->addr;
 
-	/* cut when vmalloc() area is higher than direct-map area */
-	if (VMALLOC_START > (unsigned long)__va(0)) {
-		if (ent->addr > VMALLOC_START)
-			goto free_out;
+	/*
+	 * We've already checked virt_addr_valid so we know this address
+	 * is a valid pointer, therefore we can check against it to determine
+	 * if we need to trim
+	 */
+	if (VMALLOC_START > ent->addr) {
 		if (VMALLOC_START - ent->addr < ent->size)
 			ent->size = VMALLOC_START - ent->addr;
 	}
-- 
cgit v1.1


From 013567be19761e2d14fc2a2676fe7686ac54c9ac Mon Sep 17 00:00:00 2001
From: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Date: Fri, 11 May 2018 16:02:00 -0700
Subject: mm: migrate: fix double call of radix_tree_replace_slot()

radix_tree_replace_slot() is called twice for head page, it's obviously
a bug.  Let's fix it.

Link: http://lkml.kernel.org/r/20180423072101.GA12157@hori1.linux.bs1.fc.nec.co.jp
Fixes: e71769ae5260 ("mm: enable thp migration for shmem thp")
Signed-off-by: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Reported-by: Matthew Wilcox <willy@infradead.org>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: "Kirill A. Shutemov" <kirill@shutemov.name>
Cc: Zi Yan <zi.yan@sent.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 mm/migrate.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/mm/migrate.c b/mm/migrate.c
index 5684330..8c0af0f 100644
--- a/mm/migrate.c
+++ b/mm/migrate.c
@@ -528,14 +528,12 @@ int migrate_page_move_mapping(struct address_space *mapping,
 		int i;
 		int index = page_index(page);
 
-		for (i = 0; i < HPAGE_PMD_NR; i++) {
+		for (i = 1; i < HPAGE_PMD_NR; i++) {
 			pslot = radix_tree_lookup_slot(&mapping->i_pages,
 						       index + i);
 			radix_tree_replace_slot(&mapping->i_pages, pslot,
 						newpage + i);
 		}
-	} else {
-		radix_tree_replace_slot(&mapping->i_pages, pslot, newpage);
 	}
 
 	/*
-- 
cgit v1.1


From 27ae357fa82be5ab73b2ef8d39dcb8ca2563483a Mon Sep 17 00:00:00 2001
From: David Rientjes <rientjes@google.com>
Date: Fri, 11 May 2018 16:02:04 -0700
Subject: mm, oom: fix concurrent munlock and oom reaper unmap, v3

Since exit_mmap() is done without the protection of mm->mmap_sem, it is
possible for the oom reaper to concurrently operate on an mm until
MMF_OOM_SKIP is set.

This allows munlock_vma_pages_all() to concurrently run while the oom
reaper is operating on a vma.  Since munlock_vma_pages_range() depends
on clearing VM_LOCKED from vm_flags before actually doing the munlock to
determine if any other vmas are locking the same memory, the check for
VM_LOCKED in the oom reaper is racy.

This is especially noticeable on architectures such as powerpc where
clearing a huge pmd requires serialize_against_pte_lookup().  If the pmd
is zapped by the oom reaper during follow_page_mask() after the check
for pmd_none() is bypassed, this ends up deferencing a NULL ptl or a
kernel oops.

Fix this by manually freeing all possible memory from the mm before
doing the munlock and then setting MMF_OOM_SKIP.  The oom reaper can not
run on the mm anymore so the munlock is safe to do in exit_mmap().  It
also matches the logic that the oom reaper currently uses for
determining when to set MMF_OOM_SKIP itself, so there's no new risk of
excessive oom killing.

This issue fixes CVE-2018-1000200.

Link: http://lkml.kernel.org/r/alpine.DEB.2.21.1804241526320.238665@chino.kir.corp.google.com
Fixes: 212925802454 ("mm: oom: let oom_reap_task and exit_mmap run concurrently")
Signed-off-by: David Rientjes <rientjes@google.com>
Suggested-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Acked-by: Michal Hocko <mhocko@suse.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: <stable@vger.kernel.org>	[4.14+]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/oom.h |  2 ++
 mm/mmap.c           | 44 +++++++++++++++++------------
 mm/oom_kill.c       | 81 ++++++++++++++++++++++++++++-------------------------
 3 files changed, 71 insertions(+), 56 deletions(-)

diff --git a/include/linux/oom.h b/include/linux/oom.h
index 5bad038..6adac11 100644
--- a/include/linux/oom.h
+++ b/include/linux/oom.h
@@ -95,6 +95,8 @@ static inline int check_stable_address_space(struct mm_struct *mm)
 	return 0;
 }
 
+void __oom_reap_task_mm(struct mm_struct *mm);
+
 extern unsigned long oom_badness(struct task_struct *p,
 		struct mem_cgroup *memcg, const nodemask_t *nodemask,
 		unsigned long totalpages);
diff --git a/mm/mmap.c b/mm/mmap.c
index 9d5968d..d683656 100644
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -3024,6 +3024,32 @@ void exit_mmap(struct mm_struct *mm)
 	/* mm's last user has gone, and its about to be pulled down */
 	mmu_notifier_release(mm);
 
+	if (unlikely(mm_is_oom_victim(mm))) {
+		/*
+		 * Manually reap the mm to free as much memory as possible.
+		 * Then, as the oom reaper does, set MMF_OOM_SKIP to disregard
+		 * this mm from further consideration.  Taking mm->mmap_sem for
+		 * write after setting MMF_OOM_SKIP will guarantee that the oom
+		 * reaper will not run on this mm again after mmap_sem is
+		 * dropped.
+		 *
+		 * Nothing can be holding mm->mmap_sem here and the above call
+		 * to mmu_notifier_release(mm) ensures mmu notifier callbacks in
+		 * __oom_reap_task_mm() will not block.
+		 *
+		 * This needs to be done before calling munlock_vma_pages_all(),
+		 * which clears VM_LOCKED, otherwise the oom reaper cannot
+		 * reliably test it.
+		 */
+		mutex_lock(&oom_lock);
+		__oom_reap_task_mm(mm);
+		mutex_unlock(&oom_lock);
+
+		set_bit(MMF_OOM_SKIP, &mm->flags);
+		down_write(&mm->mmap_sem);
+		up_write(&mm->mmap_sem);
+	}
+
 	if (mm->locked_vm) {
 		vma = mm->mmap;
 		while (vma) {
@@ -3045,24 +3071,6 @@ void exit_mmap(struct mm_struct *mm)
 	/* update_hiwater_rss(mm) here? but nobody should be looking */
 	/* Use -1 here to ensure all VMAs in the mm are unmapped */
 	unmap_vmas(&tlb, vma, 0, -1);
-
-	if (unlikely(mm_is_oom_victim(mm))) {
-		/*
-		 * Wait for oom_reap_task() to stop working on this
-		 * mm. Because MMF_OOM_SKIP is already set before
-		 * calling down_read(), oom_reap_task() will not run
-		 * on this "mm" post up_write().
-		 *
-		 * mm_is_oom_victim() cannot be set from under us
-		 * either because victim->mm is already set to NULL
-		 * under task_lock before calling mmput and oom_mm is
-		 * set not NULL by the OOM killer only if victim->mm
-		 * is found not NULL while holding the task_lock.
-		 */
-		set_bit(MMF_OOM_SKIP, &mm->flags);
-		down_write(&mm->mmap_sem);
-		up_write(&mm->mmap_sem);
-	}
 	free_pgtables(&tlb, vma, FIRST_USER_ADDRESS, USER_PGTABLES_CEILING);
 	tlb_finish_mmu(&tlb, 0, -1);
 
diff --git a/mm/oom_kill.c b/mm/oom_kill.c
index ff992fa..8ba6cb8 100644
--- a/mm/oom_kill.c
+++ b/mm/oom_kill.c
@@ -469,7 +469,6 @@ bool process_shares_mm(struct task_struct *p, struct mm_struct *mm)
 	return false;
 }
 
-
 #ifdef CONFIG_MMU
 /*
  * OOM Reaper kernel thread which tries to reap the memory used by the OOM
@@ -480,16 +479,54 @@ static DECLARE_WAIT_QUEUE_HEAD(oom_reaper_wait);
 static struct task_struct *oom_reaper_list;
 static DEFINE_SPINLOCK(oom_reaper_lock);
 
-static bool __oom_reap_task_mm(struct task_struct *tsk, struct mm_struct *mm)
+void __oom_reap_task_mm(struct mm_struct *mm)
 {
-	struct mmu_gather tlb;
 	struct vm_area_struct *vma;
+
+	/*
+	 * Tell all users of get_user/copy_from_user etc... that the content
+	 * is no longer stable. No barriers really needed because unmapping
+	 * should imply barriers already and the reader would hit a page fault
+	 * if it stumbled over a reaped memory.
+	 */
+	set_bit(MMF_UNSTABLE, &mm->flags);
+
+	for (vma = mm->mmap ; vma; vma = vma->vm_next) {
+		if (!can_madv_dontneed_vma(vma))
+			continue;
+
+		/*
+		 * Only anonymous pages have a good chance to be dropped
+		 * without additional steps which we cannot afford as we
+		 * are OOM already.
+		 *
+		 * We do not even care about fs backed pages because all
+		 * which are reclaimable have already been reclaimed and
+		 * we do not want to block exit_mmap by keeping mm ref
+		 * count elevated without a good reason.
+		 */
+		if (vma_is_anonymous(vma) || !(vma->vm_flags & VM_SHARED)) {
+			const unsigned long start = vma->vm_start;
+			const unsigned long end = vma->vm_end;
+			struct mmu_gather tlb;
+
+			tlb_gather_mmu(&tlb, mm, start, end);
+			mmu_notifier_invalidate_range_start(mm, start, end);
+			unmap_page_range(&tlb, vma, start, end, NULL);
+			mmu_notifier_invalidate_range_end(mm, start, end);
+			tlb_finish_mmu(&tlb, start, end);
+		}
+	}
+}
+
+static bool oom_reap_task_mm(struct task_struct *tsk, struct mm_struct *mm)
+{
 	bool ret = true;
 
 	/*
 	 * We have to make sure to not race with the victim exit path
 	 * and cause premature new oom victim selection:
-	 * __oom_reap_task_mm		exit_mm
+	 * oom_reap_task_mm		exit_mm
 	 *   mmget_not_zero
 	 *				  mmput
 	 *				    atomic_dec_and_test
@@ -534,39 +571,8 @@ static bool __oom_reap_task_mm(struct task_struct *tsk, struct mm_struct *mm)
 
 	trace_start_task_reaping(tsk->pid);
 
-	/*
-	 * Tell all users of get_user/copy_from_user etc... that the content
-	 * is no longer stable. No barriers really needed because unmapping
-	 * should imply barriers already and the reader would hit a page fault
-	 * if it stumbled over a reaped memory.
-	 */
-	set_bit(MMF_UNSTABLE, &mm->flags);
-
-	for (vma = mm->mmap ; vma; vma = vma->vm_next) {
-		if (!can_madv_dontneed_vma(vma))
-			continue;
+	__oom_reap_task_mm(mm);
 
-		/*
-		 * Only anonymous pages have a good chance to be dropped
-		 * without additional steps which we cannot afford as we
-		 * are OOM already.
-		 *
-		 * We do not even care about fs backed pages because all
-		 * which are reclaimable have already been reclaimed and
-		 * we do not want to block exit_mmap by keeping mm ref
-		 * count elevated without a good reason.
-		 */
-		if (vma_is_anonymous(vma) || !(vma->vm_flags & VM_SHARED)) {
-			const unsigned long start = vma->vm_start;
-			const unsigned long end = vma->vm_end;
-
-			tlb_gather_mmu(&tlb, mm, start, end);
-			mmu_notifier_invalidate_range_start(mm, start, end);
-			unmap_page_range(&tlb, vma, start, end, NULL);
-			mmu_notifier_invalidate_range_end(mm, start, end);
-			tlb_finish_mmu(&tlb, start, end);
-		}
-	}
 	pr_info("oom_reaper: reaped process %d (%s), now anon-rss:%lukB, file-rss:%lukB, shmem-rss:%lukB\n",
 			task_pid_nr(tsk), tsk->comm,
 			K(get_mm_counter(mm, MM_ANONPAGES)),
@@ -587,14 +593,13 @@ static void oom_reap_task(struct task_struct *tsk)
 	struct mm_struct *mm = tsk->signal->oom_mm;
 
 	/* Retry the down_read_trylock(mmap_sem) a few times */
-	while (attempts++ < MAX_OOM_REAP_RETRIES && !__oom_reap_task_mm(tsk, mm))
+	while (attempts++ < MAX_OOM_REAP_RETRIES && !oom_reap_task_mm(tsk, mm))
 		schedule_timeout_idle(HZ/10);
 
 	if (attempts <= MAX_OOM_REAP_RETRIES ||
 	    test_bit(MMF_OOM_SKIP, &mm->flags))
 		goto done;
 
-
 	pr_info("oom_reaper: unable to reap pid:%d (%s)\n",
 		task_pid_nr(tsk), tsk->comm);
 	debug_show_all_locks();
-- 
cgit v1.1


From e4383029201470523c3ffe339bd7d57e9b4a7d65 Mon Sep 17 00:00:00 2001
From: Ashish Samant <ashish.samant@oracle.com>
Date: Fri, 11 May 2018 16:02:07 -0700
Subject: ocfs2: take inode cluster lock before moving reflinked inode from
 orphan dir

While reflinking an inode, we create a new inode in orphan directory,
then take EX lock on it, reflink the original inode to orphan inode and
release EX lock.  Once the lock is released another node could request
it in EX mode from ocfs2_recover_orphans() which causes downconvert of
the lock, on this node, to NL mode.

Later we attempt to initialize security acl for the orphan inode and
move it to the reflink destination.  However, while doing this we dont
take EX lock on the inode.  This could potentially cause problems
because we could be starting transaction, accessing journal and
modifying metadata of the inode while holding NL lock and with another
node holding EX lock on the inode.

Fix this by taking orphan inode cluster lock in EX mode before
initializing security and moving orphan inode to reflink destination.
Use the __tracker variant while taking inode lock to avoid recursive
locking in the ocfs2_init_security_and_acl() call chain.

Link: http://lkml.kernel.org/r/1523475107-7639-1-git-send-email-ashish.samant@oracle.com
Signed-off-by: Ashish Samant <ashish.samant@oracle.com>
Reviewed-by: Joseph Qi <jiangqi903@gmail.com>
Reviewed-by: Junxiao Bi <junxiao.bi@oracle.com>
Acked-by: Jun Piao <piaojun@huawei.com>
Cc: Mark Fasheh <mark@fasheh.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Changwei Ge <ge.changwei@h3c.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 fs/ocfs2/refcounttree.c | 14 ++++++++++++--
 1 file changed, 12 insertions(+), 2 deletions(-)

diff --git a/fs/ocfs2/refcounttree.c b/fs/ocfs2/refcounttree.c
index 01c6b38..7869622 100644
--- a/fs/ocfs2/refcounttree.c
+++ b/fs/ocfs2/refcounttree.c
@@ -4250,10 +4250,11 @@ out:
 static int ocfs2_reflink(struct dentry *old_dentry, struct inode *dir,
 			 struct dentry *new_dentry, bool preserve)
 {
-	int error;
+	int error, had_lock;
 	struct inode *inode = d_inode(old_dentry);
 	struct buffer_head *old_bh = NULL;
 	struct inode *new_orphan_inode = NULL;
+	struct ocfs2_lock_holder oh;
 
 	if (!ocfs2_refcount_tree(OCFS2_SB(inode->i_sb)))
 		return -EOPNOTSUPP;
@@ -4295,6 +4296,14 @@ static int ocfs2_reflink(struct dentry *old_dentry, struct inode *dir,
 		goto out;
 	}
 
+	had_lock = ocfs2_inode_lock_tracker(new_orphan_inode, NULL, 1,
+					    &oh);
+	if (had_lock < 0) {
+		error = had_lock;
+		mlog_errno(error);
+		goto out;
+	}
+
 	/* If the security isn't preserved, we need to re-initialize them. */
 	if (!preserve) {
 		error = ocfs2_init_security_and_acl(dir, new_orphan_inode,
@@ -4302,14 +4311,15 @@ static int ocfs2_reflink(struct dentry *old_dentry, struct inode *dir,
 		if (error)
 			mlog_errno(error);
 	}
-out:
 	if (!error) {
 		error = ocfs2_mv_orphaned_inode_to_new(dir, new_orphan_inode,
 						       new_dentry);
 		if (error)
 			mlog_errno(error);
 	}
+	ocfs2_inode_unlock_tracker(new_orphan_inode, 1, &oh, had_lock);
 
+out:
 	if (new_orphan_inode) {
 		/*
 		 * We need to open_unlock the inode no matter whether we
-- 
cgit v1.1


From 78eb0c6356cda285c6ee6e29bea0c0188368103e Mon Sep 17 00:00:00 2001
From: Changbin Du <changbin.du@intel.com>
Date: Fri, 11 May 2018 16:02:11 -0700
Subject: scripts/faddr2line: fix error when addr2line output contains
 discriminator

When addr2line output contains discriminator, the current awk script
cannot parse it.  This patch fixes it by extracting key words using
regex which is more reliable.

  $ scripts/faddr2line vmlinux tlb_flush_mmu_free+0x26
  tlb_flush_mmu_free+0x26/0x50:
  tlb_flush_mmu_free at mm/memory.c:258 (discriminator 3)
  scripts/faddr2line: eval: line 173: unexpected EOF while looking for matching `)'

Link: http://lkml.kernel.org/r/1525323379-25193-1-git-send-email-changbin.du@intel.com
Fixes: 6870c0165feaa5 ("scripts/faddr2line: show the code context")
Signed-off-by: Changbin Du <changbin.du@intel.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Philippe Ombredanne <pombredanne@nexb.com>
Cc: NeilBrown <neilb@suse.com>
Cc: Richard Weinberger <richard@nod.at>
Cc: Kate Stewart <kstewart@linuxfoundation.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 scripts/faddr2line | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/scripts/faddr2line b/scripts/faddr2line
index 9e5735a..1876a74 100755
--- a/scripts/faddr2line
+++ b/scripts/faddr2line
@@ -170,7 +170,10 @@ __faddr2line() {
 		echo "$file_lines" | while read -r line
 		do
 			echo $line
-			eval $(echo $line | awk -F "[ :]" '{printf("n1=%d;n2=%d;f=%s",$NF-5, $NF+5, $(NF-1))}')
+			n=$(echo $line | sed 's/.*:\([0-9]\+\).*/\1/g')
+			n1=$[$n-5]
+			n2=$[$n+5]
+			f=$(echo $line | sed 's/.*at \(.\+\):.*/\1/g')
 			awk 'NR>=strtonum("'$n1'") && NR<=strtonum("'$n2'") {printf("%d\t%s\n", NR, $0)}' $f
 		done
 
-- 
cgit v1.1


From 2075b16e32c26e4031b9fd3cbe26c54676a8fcb5 Mon Sep 17 00:00:00 2001
From: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Date: Fri, 11 May 2018 16:02:14 -0700
Subject: rbtree: include rcu.h

Since commit c1adf20052d8 ("Introduce rb_replace_node_rcu()")
rbtree_augmented.h uses RCU related data structures but does not include
the header file.  It works as long as it gets somehow included before
that and fails otherwise.

Link: http://lkml.kernel.org/r/20180504103159.19938-1-bigeasy@linutronix.de
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: David Howells <dhowells@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/rbtree_augmented.h | 1 +
 include/linux/rbtree_latch.h     | 1 +
 2 files changed, 2 insertions(+)

diff --git a/include/linux/rbtree_augmented.h b/include/linux/rbtree_augmented.h
index 6bfd2b5..af8a61b 100644
--- a/include/linux/rbtree_augmented.h
+++ b/include/linux/rbtree_augmented.h
@@ -26,6 +26,7 @@
 
 #include <linux/compiler.h>
 #include <linux/rbtree.h>
+#include <linux/rcupdate.h>
 
 /*
  * Please note - only struct rb_augment_callbacks and the prototypes for
diff --git a/include/linux/rbtree_latch.h b/include/linux/rbtree_latch.h
index ece43e8..7d012fa 100644
--- a/include/linux/rbtree_latch.h
+++ b/include/linux/rbtree_latch.h
@@ -35,6 +35,7 @@
 
 #include <linux/rbtree.h>
 #include <linux/seqlock.h>
+#include <linux/rcupdate.h>
 
 struct latch_tree_node {
 	struct rb_node node[2];
-- 
cgit v1.1


From 789ba28013ce23dbf5e9f5f014f4233b35523bf3 Mon Sep 17 00:00:00 2001
From: Mel Gorman <mgorman@techsingularity.net>
Date: Wed, 9 May 2018 17:31:15 +0100
Subject: Revert "sched/numa: Delay retrying placement for automatic NUMA
 balance after wake_affine()"

This reverts commit 7347fc87dfe6b7315e74310ee1243dc222c68086.

Srikar Dronamra pointed out that while the commit in question did show
a performance improvement on ppc64, it did so at the cost of disabling
active CPU migration by automatic NUMA balancing which was not the intent.
The issue was that a serious flaw in the logic failed to ever active balance
if SD_WAKE_AFFINE was disabled on scheduler domains. Even when it's enabled,
the logic is still bizarre and against the original intent.

Investigation showed that fixing the patch in either the way he suggested,
using the correct comparison for jiffies values or introducing a new
numa_migrate_deferred variable in task_struct all perform similarly to a
revert with a mix of gains and losses depending on the workload, machine
and socket count.

The original intent of the commit was to handle a problem whereby
wake_affine, idle balancing and automatic NUMA balancing disagree on the
appropriate placement for a task. This was particularly true for cases where
a single task was a massive waker of tasks but where wake_wide logic did
not apply.  This was particularly noticeable when a futex (a barrier) woke
all worker threads and tried pulling the wakees to the waker nodes. In that
specific case, it could be handled by tuning MPI or openMP appropriately,
but the behavior is not illogical and was worth attempting to fix. However,
the approach was wrong. Given that we're at rc4 and a fix is not obvious,
it's better to play safe, revert this commit and retry later.

Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: efault@gmx.de
Cc: ggherdovich@suse.cz
Cc: hpa@zytor.com
Cc: matt@codeblueprint.co.uk
Cc: mpe@ellerman.id.au
Link: http://lkml.kernel.org/r/20180509163115.6fnnyeg4vdm2ct4v@techsingularity.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 kernel/sched/fair.c | 57 +----------------------------------------------------
 1 file changed, 1 insertion(+), 56 deletions(-)

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 54dc31e..f43627c 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -1854,7 +1854,6 @@ static int task_numa_migrate(struct task_struct *p)
 static void numa_migrate_preferred(struct task_struct *p)
 {
 	unsigned long interval = HZ;
-	unsigned long numa_migrate_retry;
 
 	/* This task has no NUMA fault statistics yet */
 	if (unlikely(p->numa_preferred_nid == -1 || !p->numa_faults))
@@ -1862,18 +1861,7 @@ static void numa_migrate_preferred(struct task_struct *p)
 
 	/* Periodically retry migrating the task to the preferred node */
 	interval = min(interval, msecs_to_jiffies(p->numa_scan_period) / 16);
-	numa_migrate_retry = jiffies + interval;
-
-	/*
-	 * Check that the new retry threshold is after the current one. If
-	 * the retry is in the future, it implies that wake_affine has
-	 * temporarily asked NUMA balancing to backoff from placement.
-	 */
-	if (numa_migrate_retry > p->numa_migrate_retry)
-		return;
-
-	/* Safe to try placing the task on the preferred node */
-	p->numa_migrate_retry = numa_migrate_retry;
+	p->numa_migrate_retry = jiffies + interval;
 
 	/* Success if task is already running on preferred CPU */
 	if (task_node(p) == p->numa_preferred_nid)
@@ -5922,48 +5910,6 @@ wake_affine_weight(struct sched_domain *sd, struct task_struct *p,
 	return this_eff_load < prev_eff_load ? this_cpu : nr_cpumask_bits;
 }
 
-#ifdef CONFIG_NUMA_BALANCING
-static void
-update_wa_numa_placement(struct task_struct *p, int prev_cpu, int target)
-{
-	unsigned long interval;
-
-	if (!static_branch_likely(&sched_numa_balancing))
-		return;
-
-	/* If balancing has no preference then continue gathering data */
-	if (p->numa_preferred_nid == -1)
-		return;
-
-	/*
-	 * If the wakeup is not affecting locality then it is neutral from
-	 * the perspective of NUMA balacing so continue gathering data.
-	 */
-	if (cpu_to_node(prev_cpu) == cpu_to_node(target))
-		return;
-
-	/*
-	 * Temporarily prevent NUMA balancing trying to place waker/wakee after
-	 * wakee has been moved by wake_affine. This will potentially allow
-	 * related tasks to converge and update their data placement. The
-	 * 4 * numa_scan_period is to allow the two-pass filter to migrate
-	 * hot data to the wakers node.
-	 */
-	interval = max(sysctl_numa_balancing_scan_delay,
-			 p->numa_scan_period << 2);
-	p->numa_migrate_retry = jiffies + msecs_to_jiffies(interval);
-
-	interval = max(sysctl_numa_balancing_scan_delay,
-			 current->numa_scan_period << 2);
-	current->numa_migrate_retry = jiffies + msecs_to_jiffies(interval);
-}
-#else
-static void
-update_wa_numa_placement(struct task_struct *p, int prev_cpu, int target)
-{
-}
-#endif
-
 static int wake_affine(struct sched_domain *sd, struct task_struct *p,
 		       int this_cpu, int prev_cpu, int sync)
 {
@@ -5979,7 +5925,6 @@ static int wake_affine(struct sched_domain *sd, struct task_struct *p,
 	if (target == nr_cpumask_bits)
 		return prev_cpu;
 
-	update_wa_numa_placement(p, prev_cpu, target);
 	schedstat_inc(sd->ttwu_move_affine);
 	schedstat_inc(p->se.statistics.nr_wakeups_affine);
 	return target;
-- 
cgit v1.1


From 05e13bb57e6f181d7605f8608181c7e6fb7f591d Mon Sep 17 00:00:00 2001
From: Jean Delvare <jdelvare@suse.de>
Date: Sat, 12 May 2018 11:57:37 +0200
Subject: swiotlb: silent unwanted warning "buffer is full"
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

If DMA_ATTR_NO_WARN is passed to swiotlb_alloc_buffer(), it should be
passed further down to swiotlb_tbl_map_single(). Otherwise we escape
half of the warnings but still log the other half.

This is one of the multiple causes of spurious warnings reported at:
https://bugs.freedesktop.org/show_bug.cgi?id=104082

Signed-off-by: Jean Delvare <jdelvare@suse.de>
Fixes: 0176adb00406 ("swiotlb: refactor coherent buffer allocation")
Cc: Christoph Hellwig <hch@lst.de>
Cc: Christian König <christian.koenig@amd.com>
Cc: Michel Dänzer <michel@daenzer.net>
Cc: Takashi Iwai <tiwai@suse.de>
Cc: stable@vger.kernel.org # v4.16
---
 lib/swiotlb.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/lib/swiotlb.c b/lib/swiotlb.c
index 12fbaa4..cc64058 100644
--- a/lib/swiotlb.c
+++ b/lib/swiotlb.c
@@ -714,7 +714,7 @@ swiotlb_alloc_buffer(struct device *dev, size_t size, dma_addr_t *dma_handle,
 
 	phys_addr = swiotlb_tbl_map_single(dev,
 			__phys_to_dma(dev, io_tlb_start),
-			0, size, DMA_FROM_DEVICE, 0);
+			0, size, DMA_FROM_DEVICE, attrs);
 	if (phys_addr == SWIOTLB_MAP_ERROR)
 		goto out_warn;
 
-- 
cgit v1.1


From f79401b477bc22914e4c37ea39c611117bd10b19 Mon Sep 17 00:00:00 2001
From: Matthew Auld <matthew.auld@intel.com>
Date: Fri, 11 May 2018 10:51:40 +0100
Subject: drm/i915/selftests: scrub 64K

We write all 4K page entries, even when using 64K pages. In order to
verify that the HW isn't cheating by using the 4K PTE instead of the 64K
PTE, we want to remove all the surplus entries. If the HW skipped the
64K PTE, it will read/write into the scratch page instead - which we
detect as missing results during selftests.

v2: much improved commentary (Chris)

Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Changbin Du <changbin.du@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20180511095140.25590-1-matthew.auld@intel.com
---
 drivers/gpu/drm/i915/i915_gem_gtt.c         | 21 +++++++++++++++++++++
 drivers/gpu/drm/i915/i915_gem_gtt.h         |  1 +
 drivers/gpu/drm/i915/selftests/huge_pages.c |  3 +++
 3 files changed, 25 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index c01d6db..996ab2a 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -1162,6 +1162,27 @@ static void gen8_ppgtt_insert_huge_entries(struct i915_vma *vma,
 			vaddr[idx.pde] |= GEN8_PDE_IPS_64K;
 			kunmap_atomic(vaddr);
 			page_size = I915_GTT_PAGE_SIZE_64K;
+
+			/*
+			 * We write all 4K page entries, even when using 64K
+			 * pages. In order to verify that the HW isn't cheating
+			 * by using the 4K PTE instead of the 64K PTE, we want
+			 * to remove all the surplus entries. If the HW skipped
+			 * the 64K PTE, it will read/write into the scratch page
+			 * instead - which we detect as missing results during
+			 * selftests.
+			 */
+			if (I915_SELFTEST_ONLY(vma->vm->scrub_64K)) {
+				u16 i;
+
+				encode = pte_encode | vma->vm->scratch_page.daddr;
+				vaddr = kmap_atomic_px(pd->page_table[idx.pde]);
+
+				for (i = 1; i < index; i += 16)
+					memset64(vaddr + i, encode, 15);
+
+				kunmap_atomic(vaddr);
+			}
 		}
 
 		vma->page_sizes.gtt |= page_size;
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.h b/drivers/gpu/drm/i915/i915_gem_gtt.h
index 1db0ded..aec4f73 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.h
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.h
@@ -342,6 +342,7 @@ struct i915_address_space {
 	void (*clear_pages)(struct i915_vma *vma);
 
 	I915_SELFTEST_DECLARE(struct fault_attr fault_attr);
+	I915_SELFTEST_DECLARE(bool scrub_64K);
 };
 
 #define i915_is_ggtt(V) (!(V)->file)
diff --git a/drivers/gpu/drm/i915/selftests/huge_pages.c b/drivers/gpu/drm/i915/selftests/huge_pages.c
index d7c8ef8..91c7291 100644
--- a/drivers/gpu/drm/i915/selftests/huge_pages.c
+++ b/drivers/gpu/drm/i915/selftests/huge_pages.c
@@ -1757,6 +1757,9 @@ int i915_gem_huge_page_live_selftests(struct drm_i915_private *dev_priv)
 		goto out_unlock;
 	}
 
+	if (ctx->ppgtt)
+		ctx->ppgtt->base.scrub_64K = true;
+
 	err = i915_subtests(tests, ctx);
 
 out_unlock:
-- 
cgit v1.1


From 0c79f9cb77eae28d48a4f9fc1b3341aacbbd260c Mon Sep 17 00:00:00 2001
From: Michel Thierry <michel.thierry@intel.com>
Date: Thu, 10 May 2018 13:07:08 -0700
Subject: drm/i915/gen9: Add WaClearHIZ_WM_CHICKEN3 for bxt and glk

Factor in clear values wherever required while updating destination
min/max.

References: HSDES#1604444184
Signed-off-by: Michel Thierry <michel.thierry@intel.com>
Cc: mesa-dev@lists.freedesktop.org
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Oscar Mateo <oscar.mateo@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20180510200708.18097-1-michel.thierry@intel.com
Cc: stable@vger.kernel.org
---
 drivers/gpu/drm/i915/i915_reg.h          | 3 +++
 drivers/gpu/drm/i915/intel_workarounds.c | 4 ++++
 2 files changed, 7 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 1449178..f11bb21 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -7259,6 +7259,9 @@ enum {
 #define SLICE_ECO_CHICKEN0			_MMIO(0x7308)
 #define   PIXEL_MASK_CAMMING_DISABLE		(1 << 14)
 
+#define GEN9_WM_CHICKEN3			_MMIO(0x5588)
+#define   GEN9_FACTOR_IN_CLR_VAL_HIZ		(1 << 9)
+
 /* WaCatErrorRejectionIssue */
 #define GEN7_SQ_CHICKEN_MBCUNIT_CONFIG		_MMIO(0x9030)
 #define  GEN7_SQ_CHICKEN_MBCUNIT_SQINTMOB	(1<<11)
diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c
index 5eec4ce..2df3538 100644
--- a/drivers/gpu/drm/i915/intel_workarounds.c
+++ b/drivers/gpu/drm/i915/intel_workarounds.c
@@ -270,6 +270,10 @@ static int gen9_ctx_workarounds_init(struct drm_i915_private *dev_priv)
 			    GEN9_PREEMPT_GPGPU_LEVEL_MASK,
 			    GEN9_PREEMPT_GPGPU_COMMAND_LEVEL);
 
+	/* WaClearHIZ_WM_CHICKEN3:bxt,glk */
+	if (IS_GEN9_LP(dev_priv))
+		WA_SET_BIT_MASKED(GEN9_WM_CHICKEN3, GEN9_FACTOR_IN_CLR_VAL_HIZ);
+
 	return 0;
 }
 
-- 
cgit v1.1


From b99f514f5dfa38e04ef0b628d82a97772945cae7 Mon Sep 17 00:00:00 2001
From: Changbin Du <changbin.du@intel.com>
Date: Thu, 19 Apr 2018 12:12:37 +0800
Subject: drm/i915/gvt: Remove disable_warn_untrack and print untracked mmio
 with debug level

The disable_warn_untrack never prevent gvt from printing untracked
mmio errors. We were disturbed by this error storm and the fix is
just adding them to the list with no essential new change.

This message is only useful for enabling new platform during
developing process. So lower the message level to debug and then
remove disable_warn_untrack.

Signed-off-by: Changbin Du <changbin.du@intel.com>
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
---
 drivers/gpu/drm/i915/gvt/gvt.h      |  1 -
 drivers/gpu/drm/i915/gvt/handlers.c | 11 +++--------
 drivers/gpu/drm/i915/gvt/mmio.c     |  2 --
 3 files changed, 3 insertions(+), 11 deletions(-)

diff --git a/drivers/gpu/drm/i915/gvt/gvt.h b/drivers/gpu/drm/i915/gvt/gvt.h
index 6ec8888..05d15a0 100644
--- a/drivers/gpu/drm/i915/gvt/gvt.h
+++ b/drivers/gpu/drm/i915/gvt/gvt.h
@@ -99,7 +99,6 @@ struct intel_vgpu_fence {
 struct intel_vgpu_mmio {
 	void *vreg;
 	void *sreg;
-	bool disable_warn_untrack;
 };
 
 #define INTEL_GVT_MAX_BAR_NUM 4
diff --git a/drivers/gpu/drm/i915/gvt/handlers.c b/drivers/gpu/drm/i915/gvt/handlers.c
index a33c1c3e..26c924b 100644
--- a/drivers/gpu/drm/i915/gvt/handlers.c
+++ b/drivers/gpu/drm/i915/gvt/handlers.c
@@ -191,6 +191,8 @@ static int sanitize_fence_mmio_access(struct intel_vgpu *vgpu,
 	unsigned int max_fence = vgpu_fence_sz(vgpu);
 
 	if (fence_num >= max_fence) {
+		gvt_vgpu_err("access oob fence reg %d/%d\n",
+			     fence_num, max_fence);
 
 		/* When guest access oob fence regs without access
 		 * pv_info first, we treat guest not supporting GVT,
@@ -200,11 +202,6 @@ static int sanitize_fence_mmio_access(struct intel_vgpu *vgpu,
 			enter_failsafe_mode(vgpu,
 					GVT_FAILSAFE_UNSUPPORTED_GUEST);
 
-		if (!vgpu->mmio.disable_warn_untrack) {
-			gvt_vgpu_err("found oob fence register access\n");
-			gvt_vgpu_err("total fence %d, access fence %d\n",
-				     max_fence, fence_num);
-		}
 		memset(p_data, 0, bytes);
 		return -EINVAL;
 	}
@@ -3092,9 +3089,7 @@ int intel_vgpu_mmio_reg_rw(struct intel_vgpu *vgpu, unsigned int offset,
 	 */
 	mmio_info = find_mmio_info(gvt, offset);
 	if (!mmio_info) {
-		if (!vgpu->mmio.disable_warn_untrack)
-			gvt_vgpu_err("untracked MMIO %08x len %d\n",
-				     offset, bytes);
+		gvt_dbg_mmio("untracked MMIO %08x len %d\n", offset, bytes);
 		goto default_rw;
 	}
 
diff --git a/drivers/gpu/drm/i915/gvt/mmio.c b/drivers/gpu/drm/i915/gvt/mmio.c
index 11b71b3..e4960af 100644
--- a/drivers/gpu/drm/i915/gvt/mmio.c
+++ b/drivers/gpu/drm/i915/gvt/mmio.c
@@ -244,8 +244,6 @@ void intel_vgpu_reset_mmio(struct intel_vgpu *vgpu, bool dmlr)
 
 		/* set the bit 0:2(Core C-State ) to C0 */
 		vgpu_vreg_t(vgpu, GEN6_GT_CORE_STATUS) = 0;
-
-		vgpu->mmio.disable_warn_untrack = false;
 	} else {
 #define GVT_GEN8_MMIO_RESET_OFFSET		(0x44200)
 		/* only reset the engine related, so starting with 0x44200
-- 
cgit v1.1


From 3d8b9e258b9dbbeb0cdeb1cf5885e40d63d564ab Mon Sep 17 00:00:00 2001
From: Zhao Yan <yan.y.zhao@intel.com>
Date: Tue, 8 May 2018 14:52:30 +0800
Subject: drm/i915/gvt: let NOPID be the default value of force_to_nonpriv
 registers

Each ring has a NOPID register and currently they are regarded as default
value of force_to_nonpriv registers in guest drivers

Signed-off-by: Zhao Yan <yan.y.zhao@intel.com>
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
---
 drivers/gpu/drm/i915/gvt/cmd_parser.c |  9 ++++++++-
 drivers/gpu/drm/i915/gvt/handlers.c   | 22 ++++++++++++++--------
 2 files changed, 22 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/i915/gvt/cmd_parser.c b/drivers/gpu/drm/i915/gvt/cmd_parser.c
index 9ec2cd9..737cc82 100644
--- a/drivers/gpu/drm/i915/gvt/cmd_parser.c
+++ b/drivers/gpu/drm/i915/gvt/cmd_parser.c
@@ -817,8 +817,15 @@ static int force_nonpriv_reg_handler(struct parser_exec_state *s,
 {
 	struct intel_gvt *gvt = s->vgpu->gvt;
 	unsigned int data = cmd_val(s, index + 1);
+	u32 ring_base;
+	u32 nopid;
+	struct drm_i915_private *dev_priv = s->vgpu->gvt->dev_priv;
+
+	ring_base = dev_priv->engine[s->ring_id]->mmio_base;
+	nopid = i915_mmio_reg_offset(RING_NOPID(ring_base));
 
-	if (!intel_gvt_in_force_nonpriv_whitelist(gvt, data)) {
+	if (!intel_gvt_in_force_nonpriv_whitelist(gvt, data) &&
+			data != nopid) {
 		gvt_err("Unexpected forcenonpriv 0x%x LRI write, value=0x%x\n",
 			offset, data);
 		return -EPERM;
diff --git a/drivers/gpu/drm/i915/gvt/handlers.c b/drivers/gpu/drm/i915/gvt/handlers.c
index 26c924b..bf2fa60 100644
--- a/drivers/gpu/drm/i915/gvt/handlers.c
+++ b/drivers/gpu/drm/i915/gvt/handlers.c
@@ -474,21 +474,27 @@ static int force_nonpriv_write(struct intel_vgpu *vgpu,
 	unsigned int offset, void *p_data, unsigned int bytes)
 {
 	u32 reg_nonpriv = *(u32 *)p_data;
+	int ring_id = intel_gvt_render_mmio_to_ring_id(vgpu->gvt, offset);
+	u32 ring_base;
+	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
 	int ret = -EINVAL;
 
-	if ((bytes != 4) || ((offset & (bytes - 1)) != 0)) {
-		gvt_err("vgpu(%d) Invalid FORCE_NONPRIV offset %x(%dB)\n",
-			vgpu->id, offset, bytes);
+	if ((bytes != 4) || ((offset & (bytes - 1)) != 0) || ring_id < 0) {
+		gvt_err("vgpu(%d) ring %d Invalid FORCE_NONPRIV offset %x(%dB)\n",
+			vgpu->id, ring_id, offset, bytes);
 		return ret;
 	}
 
-	if (in_whitelist(reg_nonpriv)) {
+	ring_base = dev_priv->engine[ring_id]->mmio_base;
+
+	if (in_whitelist(reg_nonpriv) ||
+		reg_nonpriv == i915_mmio_reg_offset(RING_NOPID(ring_base))) {
 		ret = intel_vgpu_default_mmio_write(vgpu, offset, p_data,
 			bytes);
-	} else {
-		gvt_err("vgpu(%d) Invalid FORCE_NONPRIV write %x\n",
-			vgpu->id, reg_nonpriv);
-	}
+	} else
+		gvt_err("vgpu(%d) Invalid FORCE_NONPRIV write %x at offset %x\n",
+			vgpu->id, reg_nonpriv, offset);
+
 	return ret;
 }
 
-- 
cgit v1.1


From 0438a1059877396319b90da289f1473c9c973cd8 Mon Sep 17 00:00:00 2001
From: Zhao Yan <yan.y.zhao@intel.com>
Date: Tue, 8 May 2018 14:52:42 +0800
Subject: drm/i915/gvt: do not return error on handling force_to_nonpriv
 registers

Return error will cause vm hang and enter failsafe mode.
However, we don't want that happen on detecting an wrong force_to_nonpriv
register write.
Therefore, we just omit the wrong write or patch it to default value.

v2: only return 0 on detecting lri write of registers outside whitelist,
but still return error on other error conditions.  (zhenyu wang)

Signed-off-by: Zhao Yan <yan.y.zhao@intel.com>
Reviewed-by: Zhang Yulei <yulei.zhang@intel.com>
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
---
 drivers/gpu/drm/i915/gvt/cmd_parser.c | 3 ++-
 drivers/gpu/drm/i915/gvt/handlers.c   | 2 +-
 2 files changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/i915/gvt/cmd_parser.c b/drivers/gpu/drm/i915/gvt/cmd_parser.c
index 737cc82..536cb69 100644
--- a/drivers/gpu/drm/i915/gvt/cmd_parser.c
+++ b/drivers/gpu/drm/i915/gvt/cmd_parser.c
@@ -828,7 +828,8 @@ static int force_nonpriv_reg_handler(struct parser_exec_state *s,
 			data != nopid) {
 		gvt_err("Unexpected forcenonpriv 0x%x LRI write, value=0x%x\n",
 			offset, data);
-		return -EPERM;
+		patch_value(s, cmd_ptr(s, index), nopid);
+		return 0;
 	}
 	return 0;
 }
diff --git a/drivers/gpu/drm/i915/gvt/handlers.c b/drivers/gpu/drm/i915/gvt/handlers.c
index bf2fa60..4b6532f 100644
--- a/drivers/gpu/drm/i915/gvt/handlers.c
+++ b/drivers/gpu/drm/i915/gvt/handlers.c
@@ -495,7 +495,7 @@ static int force_nonpriv_write(struct intel_vgpu *vgpu,
 		gvt_err("vgpu(%d) Invalid FORCE_NONPRIV write %x at offset %x\n",
 			vgpu->id, reg_nonpriv, offset);
 
-	return ret;
+	return 0;
 }
 
 static int ddi_buf_ctl_mmio_write(struct intel_vgpu *vgpu, unsigned int offset,
-- 
cgit v1.1


From cb8ba171ae6c1e4f5fa027162c06d50fc2b43055 Mon Sep 17 00:00:00 2001
From: Zhao Yan <yan.y.zhao@intel.com>
Date: Tue, 8 May 2018 14:52:50 +0800
Subject: drm/i915/gvt: let force_to_nonpriv cmd handler only valid for LRI cmd

the cmd_reg_handler() is called by cmds LRM, PIPE_CTRL, SRM...
for LRM, SRM, we cannot get write data in a simple way.
On other side,  the force_to_nonpriv reigsters will only be written in LRI
in current drivers. so we don't want to bother the handler to handle those
memory access cmds, just leave a print message here.

Signed-off-by: Zhao Yan <yan.y.zhao@intel.com>
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
---
 drivers/gpu/drm/i915/gvt/cmd_parser.c | 14 +++++++++++---
 1 file changed, 11 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/i915/gvt/cmd_parser.c b/drivers/gpu/drm/i915/gvt/cmd_parser.c
index 536cb69..36c6180 100644
--- a/drivers/gpu/drm/i915/gvt/cmd_parser.c
+++ b/drivers/gpu/drm/i915/gvt/cmd_parser.c
@@ -813,14 +813,22 @@ static inline bool is_force_nonpriv_mmio(unsigned int offset)
 }
 
 static int force_nonpriv_reg_handler(struct parser_exec_state *s,
-				     unsigned int offset, unsigned int index)
+		unsigned int offset, unsigned int index, char *cmd)
 {
 	struct intel_gvt *gvt = s->vgpu->gvt;
-	unsigned int data = cmd_val(s, index + 1);
+	unsigned int data;
 	u32 ring_base;
 	u32 nopid;
 	struct drm_i915_private *dev_priv = s->vgpu->gvt->dev_priv;
 
+	if (!strcmp(cmd, "lri"))
+		data = cmd_val(s, index + 1);
+	else {
+		gvt_err("Unexpected forcenonpriv 0x%x write from cmd %s\n",
+			offset, cmd);
+		return -EINVAL;
+	}
+
 	ring_base = dev_priv->engine[s->ring_id]->mmio_base;
 	nopid = i915_mmio_reg_offset(RING_NOPID(ring_base));
 
@@ -877,7 +885,7 @@ static int cmd_reg_handler(struct parser_exec_state *s,
 		return -EINVAL;
 
 	if (is_force_nonpriv_mmio(offset) &&
-		force_nonpriv_reg_handler(s, offset, index))
+		force_nonpriv_reg_handler(s, offset, index, cmd))
 		return -EPERM;
 
 	if (offset == i915_mmio_reg_offset(DERRMR) ||
-- 
cgit v1.1


From 41e403d04e7050c8d88682939febcdbe117d4c82 Mon Sep 17 00:00:00 2001
From: Weinan Li <weinan.z.li@intel.com>
Date: Wed, 21 Mar 2018 15:40:32 +0800
Subject: Revert "drm/i915/gvt: set max priority for gvt context"

This reverts commit 11474e9091cf2002e948647fd9f63a7f027e488a.

There are issues which will block the host preemption before, instead of
disabling it use one workaround "setting max priority for gvt context"
to avoid the gvt context be preempted by the host. Now the issues have been
cleared, so revert this patch to enable host preemption.

v2:
- refine description(Zhenyu)

Signed-off-by: Weinan Li <weinan.z.li@intel.com>
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
Signed-off-by: Zhi Wang <zhi.a.wang@intel.com>
---
 drivers/gpu/drm/i915/gvt/scheduler.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/drivers/gpu/drm/i915/gvt/scheduler.c b/drivers/gpu/drm/i915/gvt/scheduler.c
index ffb45a9..c2d183b 100644
--- a/drivers/gpu/drm/i915/gvt/scheduler.c
+++ b/drivers/gpu/drm/i915/gvt/scheduler.c
@@ -1156,9 +1156,6 @@ int intel_vgpu_setup_submission(struct intel_vgpu *vgpu)
 	if (IS_ERR(s->shadow_ctx))
 		return PTR_ERR(s->shadow_ctx);
 
-	if (HAS_LOGICAL_RING_PREEMPTION(vgpu->gvt->dev_priv))
-		s->shadow_ctx->sched.priority = INT_MAX;
-
 	bitmap_zero(s->shadow_ctx_desc_updated, I915_NUM_ENGINES);
 
 	s->workloads = kmem_cache_create_usercopy("gvt-g_vgpu_workload",
-- 
cgit v1.1


From e8929999fa718da5758ff877592f33fea368ca8a Mon Sep 17 00:00:00 2001
From: Andrzej Hajda <a.hajda@samsung.com>
Date: Mon, 7 May 2018 11:29:28 +0200
Subject: drm/exynos/dsi: remove mode_set callback

The callback was used only to copy provided mode to context for later
usage. Since the mode is always available from crtc atomic state this code
can be removed.

Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
---
 drivers/gpu/drm/exynos/exynos_drm_dsi.c | 40 +++++++++------------------------
 1 file changed, 10 insertions(+), 30 deletions(-)

diff --git a/drivers/gpu/drm/exynos/exynos_drm_dsi.c b/drivers/gpu/drm/exynos/exynos_drm_dsi.c
index 7904ffa..eae44fd 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_dsi.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_dsi.c
@@ -270,7 +270,6 @@ struct exynos_dsi {
 	u32 lanes;
 	u32 mode_flags;
 	u32 format;
-	struct videomode vm;
 
 	int state;
 	struct drm_property *brightness;
@@ -881,30 +880,30 @@ static int exynos_dsi_init_link(struct exynos_dsi *dsi)
 
 static void exynos_dsi_set_display_mode(struct exynos_dsi *dsi)
 {
-	struct videomode *vm = &dsi->vm;
+	struct drm_display_mode *m = &dsi->encoder.crtc->state->adjusted_mode;
 	unsigned int num_bits_resol = dsi->driver_data->num_bits_resol;
 	u32 reg;
 
 	if (dsi->mode_flags & MIPI_DSI_MODE_VIDEO) {
 		reg = DSIM_CMD_ALLOW(0xf)
-			| DSIM_STABLE_VFP(vm->vfront_porch)
-			| DSIM_MAIN_VBP(vm->vback_porch);
+			| DSIM_STABLE_VFP(m->vsync_start - m->vdisplay)
+			| DSIM_MAIN_VBP(m->vtotal - m->vsync_end);
 		exynos_dsi_write(dsi, DSIM_MVPORCH_REG, reg);
 
-		reg = DSIM_MAIN_HFP(vm->hfront_porch)
-			| DSIM_MAIN_HBP(vm->hback_porch);
+		reg = DSIM_MAIN_HFP(m->hsync_start - m->hdisplay)
+			| DSIM_MAIN_HBP(m->htotal - m->hsync_end);
 		exynos_dsi_write(dsi, DSIM_MHPORCH_REG, reg);
 
-		reg = DSIM_MAIN_VSA(vm->vsync_len)
-			| DSIM_MAIN_HSA(vm->hsync_len);
+		reg = DSIM_MAIN_VSA(m->vsync_end - m->vsync_start)
+			| DSIM_MAIN_HSA(m->hsync_end - m->hsync_start);
 		exynos_dsi_write(dsi, DSIM_MSYNC_REG, reg);
 	}
-	reg =  DSIM_MAIN_HRESOL(vm->hactive, num_bits_resol) |
-		DSIM_MAIN_VRESOL(vm->vactive, num_bits_resol);
+	reg =  DSIM_MAIN_HRESOL(m->hdisplay, num_bits_resol) |
+		DSIM_MAIN_VRESOL(m->vdisplay, num_bits_resol);
 
 	exynos_dsi_write(dsi, DSIM_MDRESOL_REG, reg);
 
-	dev_dbg(dsi->dev, "LCD size = %dx%d\n", vm->hactive, vm->vactive);
+	dev_dbg(dsi->dev, "LCD size = %dx%d\n", m->hdisplay, m->vdisplay);
 }
 
 static void exynos_dsi_set_display_enable(struct exynos_dsi *dsi, bool enable)
@@ -1485,26 +1484,7 @@ static int exynos_dsi_create_connector(struct drm_encoder *encoder)
 	return 0;
 }
 
-static void exynos_dsi_mode_set(struct drm_encoder *encoder,
-				struct drm_display_mode *mode,
-				struct drm_display_mode *adjusted_mode)
-{
-	struct exynos_dsi *dsi = encoder_to_dsi(encoder);
-	struct videomode *vm = &dsi->vm;
-	struct drm_display_mode *m = adjusted_mode;
-
-	vm->hactive = m->hdisplay;
-	vm->vactive = m->vdisplay;
-	vm->vfront_porch = m->vsync_start - m->vdisplay;
-	vm->vback_porch = m->vtotal - m->vsync_end;
-	vm->vsync_len = m->vsync_end - m->vsync_start;
-	vm->hfront_porch = m->hsync_start - m->hdisplay;
-	vm->hback_porch = m->htotal - m->hsync_end;
-	vm->hsync_len = m->hsync_end - m->hsync_start;
-}
-
 static const struct drm_encoder_helper_funcs exynos_dsi_encoder_helper_funcs = {
-	.mode_set = exynos_dsi_mode_set,
 	.enable = exynos_dsi_enable,
 	.disable = exynos_dsi_disable,
 };
-- 
cgit v1.1


From 67b8d5c7081221efa252e111cd52532ec6d4266f Mon Sep 17 00:00:00 2001
From: Linus Torvalds <torvalds@linux-foundation.org>
Date: Sun, 13 May 2018 16:15:17 -0700
Subject: Linux 4.17-rc5

---
 Makefile | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/Makefile b/Makefile
index d0d2652..ba3106b 100644
--- a/Makefile
+++ b/Makefile
@@ -2,7 +2,7 @@
 VERSION = 4
 PATCHLEVEL = 17
 SUBLEVEL = 0
-EXTRAVERSION = -rc4
+EXTRAVERSION = -rc5
 NAME = Merciless Moray
 
 # *DOCUMENTATION*
-- 
cgit v1.1


From 4db518e4e8286ca93bd5399f26549eafc87607ea Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Fri, 11 May 2018 13:11:47 +0100
Subject: drm/i915/execlists: Relax CSB force-mmio for VT-d
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The original switch to use CSB from the HWSP was plagued by the effect
of read ordering on VT-d; we would read the WRITE pointer from the HWSP
before it had completed writing the CSB contents. The mystery comes down
to the lack of rmb() for correct ordering with respect to the writes
from HW, and with that resolved we can remove the VT-d special casing.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Michał Winiarski <michal.winiarski@intel.com>
Cc: Michel Thierry <michel.thierry@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180511121147.31915-3-chris@chris-wilson.co.uk
Tested-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
---
 drivers/gpu/drm/i915/intel_engine_cs.c | 8 --------
 1 file changed, 8 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
index 8303e05..6bfd7e3 100644
--- a/drivers/gpu/drm/i915/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/intel_engine_cs.c
@@ -458,14 +458,6 @@ static void intel_engine_init_batch_pool(struct intel_engine_cs *engine)
 
 static bool csb_force_mmio(struct drm_i915_private *i915)
 {
-	/*
-	 * IOMMU adds unpredictable latency causing the CSB write (from the
-	 * GPU into the HWSP) to only be visible some time after the interrupt
-	 * (missed breadcrumb syndrome).
-	 */
-	if (intel_vtd_active())
-		return true;
-
 	/* Older GVT emulation depends upon intercepting CSB mmio */
 	if (intel_vgpu_active(i915) && !intel_vgpu_has_hwsp_emulation(i915))
 		return true;
-- 
cgit v1.1


From 0c591a40af1b369cc11dce4d558dd71bebbdc090 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Sat, 12 May 2018 09:49:57 +0100
Subject: drm/i915: Mark up nested spinlocks

When we process the outstanding requests upon banning a context, we need
to acquire both the engine and the client's timeline, nesting the locks.
This requires explicit markup as the two timelines are now of the same
class, since commit a89d1f921c15 ("drm/i915: Split i915_gem_timeline into
individual timelines").

Testcase: igt/gem_eio/banned
Fixes: a89d1f921c15 ("drm/i915: Split i915_gem_timeline into individual timelines")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Michel Thierry <michel.thierry@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180512084957.9829-1-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/i915_gem.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 89bf5d6..0a20701 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -3119,7 +3119,7 @@ static void engine_skip_context(struct i915_request *request)
 	GEM_BUG_ON(timeline == &engine->timeline);
 
 	spin_lock_irqsave(&engine->timeline.lock, flags);
-	spin_lock(&timeline->lock);
+	spin_lock_nested(&timeline->lock, SINGLE_DEPTH_NESTING);
 
 	list_for_each_entry_continue(request, &engine->timeline.requests, link)
 		if (request->ctx == hung_ctx)
-- 
cgit v1.1


From 01f83786f9ab9c8883ce634cb9a0de51086ad7ea Mon Sep 17 00:00:00 2001
From: Jani Nikula <jani.nikula@intel.com>
Date: Mon, 14 May 2018 15:28:05 +0300
Subject: drm/i915: Update DRIVER_DATE to 20180514

Signed-off-by: Jani Nikula <jani.nikula@intel.com>
---
 drivers/gpu/drm/i915/i915_drv.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 57fb3aa..34c125e 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -85,8 +85,8 @@
 
 #define DRIVER_NAME		"i915"
 #define DRIVER_DESC		"Intel Graphics"
-#define DRIVER_DATE		"20180413"
-#define DRIVER_TIMESTAMP	1523611258
+#define DRIVER_DATE		"20180514"
+#define DRIVER_TIMESTAMP	1526300884
 
 /* Use I915_STATE_WARN(x) and I915_STATE_WARN_ON() (rather than WARN() and
  * WARN_ON()) for hw state sanity checks to check for unexpected conditions
-- 
cgit v1.1


From 76ef6b28ea4f81c3d511866a9b31392caa833126 Mon Sep 17 00:00:00 2001
From: Dave Airlie <airlied@redhat.com>
Date: Tue, 15 May 2018 13:38:15 +1000
Subject: drm: set FMODE_UNSIGNED_OFFSET for drm files

Since we have the ttm and gem vma managers using a subset
of the file address space for objects, and these start at
0x100000000 they will overflow the new mmap checks.

I've checked all the mmap routines I could see for any
bad behaviour but overall most people use GEM/TTM VMA
managers even the legacy drivers have a hashtable.

Reported-and-Tested-by: Arthur Marsh (amarsh04 on #radeon)
Fixes: be83bbf8068 (mmap: introduce sane default mmap limits)
Signed-off-by: Dave Airlie <airlied@redhat.com>
---
 drivers/gpu/drm/drm_file.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/drm_file.c b/drivers/gpu/drm/drm_file.c
index e394799..6d9b945 100644
--- a/drivers/gpu/drm/drm_file.c
+++ b/drivers/gpu/drm/drm_file.c
@@ -212,6 +212,7 @@ static int drm_open_helper(struct file *filp, struct drm_minor *minor)
 		return -ENOMEM;
 
 	filp->private_data = priv;
+	filp->f_mode |= FMODE_UNSIGNED_OFFSET;
 	priv->filp = filp;
 	priv->pid = get_pid(task_pid(current));
 	priv->minor = minor;
-- 
cgit v1.1


From 0eeef69022b4ea503106f5f695fd5d8ae2c72706 Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Mon, 9 Apr 2018 14:55:17 -0500
Subject: drm/amd/display: Updated HDR Static Metadata to directly take info
 packet raw

Updated HDR Static Metadata to directly take info packet raw

Updating Infopacket does not require Passive

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Anthony Koo <Anthony.Koo@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_resource.c | 91 +----------------------
 drivers/gpu/drm/amd/display/dc/dc_stream.h        |  4 +-
 2 files changed, 5 insertions(+), 90 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
index eb8f479..e1036e4 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
@@ -2315,97 +2315,12 @@ static void set_hdr_static_info_packet(
 		struct dc_info_packet *info_packet,
 		struct dc_stream_state *stream)
 {
-	uint16_t i = 0;
-	enum signal_type signal = stream->signal;
-	uint32_t data;
+	/* HDR Static Metadata info packet for HDR10 */
 
-	if (!stream->hdr_static_metadata.hdr_supported)
+	if (!stream->hdr_static_metadata.valid)
 		return;
 
-	if (dc_is_hdmi_signal(signal)) {
-		info_packet->valid = true;
-
-		info_packet->hb0 = 0x87;
-		info_packet->hb1 = 0x01;
-		info_packet->hb2 = 0x1A;
-		i = 1;
-	} else if (dc_is_dp_signal(signal)) {
-		info_packet->valid = true;
-
-		info_packet->hb0 = 0x00;
-		info_packet->hb1 = 0x87;
-		info_packet->hb2 = 0x1D;
-		info_packet->hb3 = (0x13 << 2);
-		i = 2;
-	}
-
-	data = stream->hdr_static_metadata.is_hdr;
-	info_packet->sb[i++] = data ? 0x02 : 0x00;
-	info_packet->sb[i++] = 0x00;
-
-	data = stream->hdr_static_metadata.chromaticity_green_x / 2;
-	info_packet->sb[i++] = data & 0xFF;
-	info_packet->sb[i++] = (data & 0xFF00) >> 8;
-
-	data = stream->hdr_static_metadata.chromaticity_green_y / 2;
-	info_packet->sb[i++] = data & 0xFF;
-	info_packet->sb[i++] = (data & 0xFF00) >> 8;
-
-	data = stream->hdr_static_metadata.chromaticity_blue_x / 2;
-	info_packet->sb[i++] = data & 0xFF;
-	info_packet->sb[i++] = (data & 0xFF00) >> 8;
-
-	data = stream->hdr_static_metadata.chromaticity_blue_y / 2;
-	info_packet->sb[i++] = data & 0xFF;
-	info_packet->sb[i++] = (data & 0xFF00) >> 8;
-
-	data = stream->hdr_static_metadata.chromaticity_red_x / 2;
-	info_packet->sb[i++] = data & 0xFF;
-	info_packet->sb[i++] = (data & 0xFF00) >> 8;
-
-	data = stream->hdr_static_metadata.chromaticity_red_y / 2;
-	info_packet->sb[i++] = data & 0xFF;
-	info_packet->sb[i++] = (data & 0xFF00) >> 8;
-
-	data = stream->hdr_static_metadata.chromaticity_white_point_x / 2;
-	info_packet->sb[i++] = data & 0xFF;
-	info_packet->sb[i++] = (data & 0xFF00) >> 8;
-
-	data = stream->hdr_static_metadata.chromaticity_white_point_y / 2;
-	info_packet->sb[i++] = data & 0xFF;
-	info_packet->sb[i++] = (data & 0xFF00) >> 8;
-
-	data = stream->hdr_static_metadata.max_luminance;
-	info_packet->sb[i++] = data & 0xFF;
-	info_packet->sb[i++] = (data & 0xFF00) >> 8;
-
-	data = stream->hdr_static_metadata.min_luminance;
-	info_packet->sb[i++] = data & 0xFF;
-	info_packet->sb[i++] = (data & 0xFF00) >> 8;
-
-	data = stream->hdr_static_metadata.maximum_content_light_level;
-	info_packet->sb[i++] = data & 0xFF;
-	info_packet->sb[i++] = (data & 0xFF00) >> 8;
-
-	data = stream->hdr_static_metadata.maximum_frame_average_light_level;
-	info_packet->sb[i++] = data & 0xFF;
-	info_packet->sb[i++] = (data & 0xFF00) >> 8;
-
-	if (dc_is_hdmi_signal(signal)) {
-		uint32_t checksum = 0;
-
-		checksum += info_packet->hb0;
-		checksum += info_packet->hb1;
-		checksum += info_packet->hb2;
-
-		for (i = 1; i <= info_packet->hb2; i++)
-			checksum += info_packet->sb[i];
-
-		info_packet->sb[0] = 0x100 - checksum;
-	} else if (dc_is_dp_signal(signal)) {
-		info_packet->sb[0] = 0x01;
-		info_packet->sb[1] = 0x1A;
-	}
+	*info_packet = stream->hdr_static_metadata;
 }
 
 static void set_vsc_info_packet(
diff --git a/drivers/gpu/drm/amd/display/dc/dc_stream.h b/drivers/gpu/drm/amd/display/dc/dc_stream.h
index 2971cd0..08f1a45 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_stream.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_stream.h
@@ -58,7 +58,7 @@ struct dc_stream_state {
 
 	struct freesync_context freesync_ctx;
 
-	struct dc_hdr_static_metadata hdr_static_metadata;
+	struct dc_info_packet hdr_static_metadata;
 	struct dc_transfer_func *out_transfer_func;
 	struct colorspace_transform gamut_remap_matrix;
 	struct csc_transform csc_color_matrix;
@@ -113,8 +113,8 @@ struct dc_stream_update {
 	struct rect src;
 	struct rect dst;
 	struct dc_transfer_func *out_transfer_func;
-	struct dc_hdr_static_metadata *hdr_static_metadata;
 	enum color_transfer_func color_output_tf;
+	struct dc_info_packet *hdr_static_metadata;
 	unsigned int *abm_level;
 	unsigned long long *periodic_fn_vsync_delta;
 };
-- 
cgit v1.1


From 85b25034608e861ce60b771b988967ea039a06c6 Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Sun, 25 Mar 2018 16:41:06 -0400
Subject: drm/amd/display: Get rid of unused input_tf

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 1 -
 drivers/gpu/drm/amd/display/dc/core/dc.c          | 3 ---
 drivers/gpu/drm/amd/display/dc/core/dc_debug.c    | 4 ----
 drivers/gpu/drm/amd/display/dc/dc.h               | 5 -----
 4 files changed, 13 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 2514d7b..aa8e25a 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -4034,7 +4034,6 @@ static bool commit_planes_to_stream(
 		flip_addr[i].address = plane_states[i]->address;
 		flip_addr[i].flip_immediate = plane_states[i]->flip_immediate;
 		plane_info[i].color_space = plane_states[i]->color_space;
-		plane_info[i].input_tf = plane_states[i]->input_tf;
 		plane_info[i].format = plane_states[i]->format;
 		plane_info[i].plane_size = plane_states[i]->plane_size;
 		plane_info[i].rotation = plane_states[i]->rotation;
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c
index 8f09f3a..e593577 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc.c
@@ -1018,9 +1018,6 @@ static enum surface_update_type get_plane_info_update_type(const struct dc_surfa
 	if (u->plane_info->color_space != u->surface->color_space)
 		update_flags->bits.color_space_change = 1;
 
-	if (u->plane_info->input_tf != u->surface->input_tf)
-		update_flags->bits.input_tf_change = 1;
-
 	if (u->plane_info->horizontal_mirror != u->surface->horizontal_mirror)
 		update_flags->bits.horizontal_mirror_change = 1;
 
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_debug.c b/drivers/gpu/drm/amd/display/dc/core/dc_debug.c
index a3c8761..267c767 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_debug.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_debug.c
@@ -155,7 +155,6 @@ void pre_surface_trace(
 				"plane_state->tiling_info.gfx8.pipe_config = %d;\n"
 				"plane_state->tiling_info.gfx8.array_mode = %d;\n"
 				"plane_state->color_space = %d;\n"
-				"plane_state->input_tf = %d;\n"
 				"plane_state->dcc.enable = %d;\n"
 				"plane_state->format = %d;\n"
 				"plane_state->rotation = %d;\n"
@@ -163,7 +162,6 @@ void pre_surface_trace(
 				plane_state->tiling_info.gfx8.pipe_config,
 				plane_state->tiling_info.gfx8.array_mode,
 				plane_state->color_space,
-				plane_state->input_tf,
 				plane_state->dcc.enable,
 				plane_state->format,
 				plane_state->rotation,
@@ -203,7 +201,6 @@ void update_surface_trace(
 		if (update->plane_info) {
 			SURFACE_TRACE(
 					"plane_info->color_space = %d;\n"
-					"plane_info->input_tf = %d;\n"
 					"plane_info->format = %d;\n"
 					"plane_info->plane_size.grph.surface_pitch = %d;\n"
 					"plane_info->plane_size.grph.surface_size.height = %d;\n"
@@ -213,7 +210,6 @@ void update_surface_trace(
 					"plane_info->rotation = %d;\n"
 					"plane_info->stereo_format = %d;\n",
 					update->plane_info->color_space,
-					update->plane_info->input_tf,
 					update->plane_info->format,
 					update->plane_info->plane_size.grph.surface_pitch,
 					update->plane_info->plane_size.grph.surface_size.height,
diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index 2334914..6a47da3 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -333,7 +333,6 @@ enum {
 	TRANSFER_FUNC_POINTS = 1025
 };
 
-// Moved here from color module for linux
 enum color_transfer_func {
 	transfer_func_unknown,
 	transfer_func_srgb,
@@ -420,7 +419,6 @@ union surface_update_flags {
 		/* Medium updates */
 		uint32_t dcc_change:1;
 		uint32_t color_space_change:1;
-		uint32_t input_tf_change:1;
 		uint32_t horizontal_mirror_change:1;
 		uint32_t per_pixel_alpha_change:1;
 		uint32_t rotation_change:1;
@@ -470,7 +468,6 @@ struct dc_plane_state {
 	struct dc_hdr_static_metadata hdr_static_ctx;
 
 	enum dc_color_space color_space;
-	enum color_transfer_func input_tf;
 
 	enum surface_pixel_format format;
 	enum dc_rotation_angle rotation;
@@ -500,7 +497,6 @@ struct dc_plane_info {
 	enum dc_rotation_angle rotation;
 	enum plane_stereo_format stereo_format;
 	enum dc_color_space color_space;
-	enum color_transfer_func input_tf;
 	unsigned int sdr_white_level;
 	bool horizontal_mirror;
 	bool visible;
@@ -527,7 +523,6 @@ struct dc_surface_update {
 	 * null means no updates
 	 */
 	struct dc_gamma *gamma;
-	enum color_transfer_func color_input_tf;
 	struct dc_transfer_func *in_transfer_func;
 
 	struct csc_transform *input_csc_color_matrix;
-- 
cgit v1.1


From 477c000ece26a588752c9d1ed9904097e95de8c9 Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Sun, 25 Mar 2018 16:55:05 -0400
Subject: drm/amd/display: Remove unused fields

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dc.h | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index 6a47da3..5b81ae5 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -361,9 +361,6 @@ struct dc_hdr_static_metadata {
 	uint32_t max_luminance;
 	uint32_t maximum_content_light_level;
 	uint32_t maximum_frame_average_light_level;
-
-	bool hdr_supported;
-	bool is_hdr;
 };
 
 enum dc_transfer_func_type {
-- 
cgit v1.1


From 5c6161162a556e2260d3c9f61f2c02bd82ee1ae9 Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Mon, 26 Mar 2018 16:14:31 -0400
Subject: drm/amd/display: Do not use os types

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c | 158 ++++++++++-----------
 drivers/gpu/drm/amd/display/include/fixed31_32.h   |  40 +++---
 2 files changed, 98 insertions(+), 100 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c b/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c
index 8a9bba8..7191c32 100644
--- a/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c
+++ b/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c
@@ -26,13 +26,13 @@
 #include "dm_services.h"
 #include "include/fixed31_32.h"
 
-static inline uint64_t abs_i64(
-	int64_t arg)
+static inline unsigned long long abs_i64(
+	long long arg)
 {
 	if (arg > 0)
-		return (uint64_t)arg;
+		return (unsigned long long)arg;
 	else
-		return (uint64_t)(-arg);
+		return (unsigned long long)(-arg);
 }
 
 /*
@@ -40,12 +40,12 @@ static inline uint64_t abs_i64(
  * result = dividend / divisor
  * *remainder = dividend % divisor
  */
-static inline uint64_t complete_integer_division_u64(
-	uint64_t dividend,
-	uint64_t divisor,
-	uint64_t *remainder)
+static inline unsigned long long complete_integer_division_u64(
+	unsigned long long dividend,
+	unsigned long long divisor,
+	unsigned long long *remainder)
 {
-	uint64_t result;
+	unsigned long long result;
 
 	ASSERT(divisor);
 
@@ -65,29 +65,29 @@ static inline uint64_t complete_integer_division_u64(
 	(FRACTIONAL_PART_MASK & (x))
 
 struct fixed31_32 dal_fixed31_32_from_fraction(
-	int64_t numerator,
-	int64_t denominator)
+	long long numerator,
+	long long denominator)
 {
 	struct fixed31_32 res;
 
 	bool arg1_negative = numerator < 0;
 	bool arg2_negative = denominator < 0;
 
-	uint64_t arg1_value = arg1_negative ? -numerator : numerator;
-	uint64_t arg2_value = arg2_negative ? -denominator : denominator;
+	unsigned long long arg1_value = arg1_negative ? -numerator : numerator;
+	unsigned long long arg2_value = arg2_negative ? -denominator : denominator;
 
-	uint64_t remainder;
+	unsigned long long remainder;
 
 	/* determine integer part */
 
-	uint64_t res_value = complete_integer_division_u64(
+	unsigned long long res_value = complete_integer_division_u64(
 		arg1_value, arg2_value, &remainder);
 
 	ASSERT(res_value <= LONG_MAX);
 
 	/* determine fractional part */
 	{
-		uint32_t i = FIXED31_32_BITS_PER_FRACTIONAL_PART;
+		unsigned int i = FIXED31_32_BITS_PER_FRACTIONAL_PART;
 
 		do {
 			remainder <<= 1;
@@ -103,14 +103,14 @@ struct fixed31_32 dal_fixed31_32_from_fraction(
 
 	/* round up LSB */
 	{
-		uint64_t summand = (remainder << 1) >= arg2_value;
+		unsigned long long summand = (remainder << 1) >= arg2_value;
 
 		ASSERT(res_value <= LLONG_MAX - summand);
 
 		res_value += summand;
 	}
 
-	res.value = (int64_t)res_value;
+	res.value = (long long)res_value;
 
 	if (arg1_negative ^ arg2_negative)
 		res.value = -res.value;
@@ -119,7 +119,7 @@ struct fixed31_32 dal_fixed31_32_from_fraction(
 }
 
 struct fixed31_32 dal_fixed31_32_from_int_nonconst(
-	int64_t arg)
+	long long arg)
 {
 	struct fixed31_32 res;
 
@@ -132,7 +132,7 @@ struct fixed31_32 dal_fixed31_32_from_int_nonconst(
 
 struct fixed31_32 dal_fixed31_32_shl(
 	struct fixed31_32 arg,
-	uint8_t shift)
+	unsigned char shift)
 {
 	struct fixed31_32 res;
 
@@ -181,16 +181,16 @@ struct fixed31_32 dal_fixed31_32_mul(
 	bool arg1_negative = arg1.value < 0;
 	bool arg2_negative = arg2.value < 0;
 
-	uint64_t arg1_value = arg1_negative ? -arg1.value : arg1.value;
-	uint64_t arg2_value = arg2_negative ? -arg2.value : arg2.value;
+	unsigned long long arg1_value = arg1_negative ? -arg1.value : arg1.value;
+	unsigned long long arg2_value = arg2_negative ? -arg2.value : arg2.value;
 
-	uint64_t arg1_int = GET_INTEGER_PART(arg1_value);
-	uint64_t arg2_int = GET_INTEGER_PART(arg2_value);
+	unsigned long long arg1_int = GET_INTEGER_PART(arg1_value);
+	unsigned long long arg2_int = GET_INTEGER_PART(arg2_value);
 
-	uint64_t arg1_fra = GET_FRACTIONAL_PART(arg1_value);
-	uint64_t arg2_fra = GET_FRACTIONAL_PART(arg2_value);
+	unsigned long long arg1_fra = GET_FRACTIONAL_PART(arg1_value);
+	unsigned long long arg2_fra = GET_FRACTIONAL_PART(arg2_value);
 
-	uint64_t tmp;
+	unsigned long long tmp;
 
 	res.value = arg1_int * arg2_int;
 
@@ -200,22 +200,22 @@ struct fixed31_32 dal_fixed31_32_mul(
 
 	tmp = arg1_int * arg2_fra;
 
-	ASSERT(tmp <= (uint64_t)(LLONG_MAX - res.value));
+	ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value));
 
 	res.value += tmp;
 
 	tmp = arg2_int * arg1_fra;
 
-	ASSERT(tmp <= (uint64_t)(LLONG_MAX - res.value));
+	ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value));
 
 	res.value += tmp;
 
 	tmp = arg1_fra * arg2_fra;
 
 	tmp = (tmp >> FIXED31_32_BITS_PER_FRACTIONAL_PART) +
-		(tmp >= (uint64_t)dal_fixed31_32_half.value);
+		(tmp >= (unsigned long long)dal_fixed31_32_half.value);
 
-	ASSERT(tmp <= (uint64_t)(LLONG_MAX - res.value));
+	ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value));
 
 	res.value += tmp;
 
@@ -230,13 +230,13 @@ struct fixed31_32 dal_fixed31_32_sqr(
 {
 	struct fixed31_32 res;
 
-	uint64_t arg_value = abs_i64(arg.value);
+	unsigned long long arg_value = abs_i64(arg.value);
 
-	uint64_t arg_int = GET_INTEGER_PART(arg_value);
+	unsigned long long arg_int = GET_INTEGER_PART(arg_value);
 
-	uint64_t arg_fra = GET_FRACTIONAL_PART(arg_value);
+	unsigned long long arg_fra = GET_FRACTIONAL_PART(arg_value);
 
-	uint64_t tmp;
+	unsigned long long tmp;
 
 	res.value = arg_int * arg_int;
 
@@ -246,20 +246,20 @@ struct fixed31_32 dal_fixed31_32_sqr(
 
 	tmp = arg_int * arg_fra;
 
-	ASSERT(tmp <= (uint64_t)(LLONG_MAX - res.value));
+	ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value));
 
 	res.value += tmp;
 
-	ASSERT(tmp <= (uint64_t)(LLONG_MAX - res.value));
+	ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value));
 
 	res.value += tmp;
 
 	tmp = arg_fra * arg_fra;
 
 	tmp = (tmp >> FIXED31_32_BITS_PER_FRACTIONAL_PART) +
-		(tmp >= (uint64_t)dal_fixed31_32_half.value);
+		(tmp >= (unsigned long long)dal_fixed31_32_half.value);
 
-	ASSERT(tmp <= (uint64_t)(LLONG_MAX - res.value));
+	ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value));
 
 	res.value += tmp;
 
@@ -288,7 +288,7 @@ struct fixed31_32 dal_fixed31_32_sinc(
 
 	struct fixed31_32 res = dal_fixed31_32_one;
 
-	int32_t n = 27;
+	int n = 27;
 
 	struct fixed31_32 arg_norm = arg;
 
@@ -299,7 +299,7 @@ struct fixed31_32 dal_fixed31_32_sinc(
 			arg_norm,
 			dal_fixed31_32_mul_int(
 				dal_fixed31_32_two_pi,
-				(int32_t)div64_s64(
+				(int)div64_s64(
 					arg_norm.value,
 					dal_fixed31_32_two_pi.value)));
 	}
@@ -343,7 +343,7 @@ struct fixed31_32 dal_fixed31_32_cos(
 
 	struct fixed31_32 res = dal_fixed31_32_one;
 
-	int32_t n = 26;
+	int n = 26;
 
 	do {
 		res = dal_fixed31_32_sub(
@@ -370,7 +370,7 @@ struct fixed31_32 dal_fixed31_32_cos(
 static struct fixed31_32 fixed31_32_exp_from_taylor_series(
 	struct fixed31_32 arg)
 {
-	uint32_t n = 9;
+	unsigned int n = 9;
 
 	struct fixed31_32 res = dal_fixed31_32_from_fraction(
 		n + 2,
@@ -409,7 +409,7 @@ struct fixed31_32 dal_fixed31_32_exp(
 	if (dal_fixed31_32_le(
 		dal_fixed31_32_ln2_div_2,
 		dal_fixed31_32_abs(arg))) {
-		int32_t m = dal_fixed31_32_round(
+		int m = dal_fixed31_32_round(
 			dal_fixed31_32_div(
 				arg,
 				dal_fixed31_32_ln2));
@@ -429,7 +429,7 @@ struct fixed31_32 dal_fixed31_32_exp(
 		if (m > 0)
 			return dal_fixed31_32_shl(
 				fixed31_32_exp_from_taylor_series(r),
-				(uint8_t)m);
+				(unsigned char)m);
 		else
 			return dal_fixed31_32_div_int(
 				fixed31_32_exp_from_taylor_series(r),
@@ -482,50 +482,50 @@ struct fixed31_32 dal_fixed31_32_pow(
 			arg2));
 }
 
-int32_t dal_fixed31_32_floor(
+int dal_fixed31_32_floor(
 	struct fixed31_32 arg)
 {
-	uint64_t arg_value = abs_i64(arg.value);
+	unsigned long long arg_value = abs_i64(arg.value);
 
 	if (arg.value >= 0)
-		return (int32_t)GET_INTEGER_PART(arg_value);
+		return (int)GET_INTEGER_PART(arg_value);
 	else
-		return -(int32_t)GET_INTEGER_PART(arg_value);
+		return -(int)GET_INTEGER_PART(arg_value);
 }
 
-int32_t dal_fixed31_32_round(
+int dal_fixed31_32_round(
 	struct fixed31_32 arg)
 {
-	uint64_t arg_value = abs_i64(arg.value);
+	unsigned long long arg_value = abs_i64(arg.value);
 
-	const int64_t summand = dal_fixed31_32_half.value;
+	const long long summand = dal_fixed31_32_half.value;
 
-	ASSERT(LLONG_MAX - (int64_t)arg_value >= summand);
+	ASSERT(LLONG_MAX - (long long)arg_value >= summand);
 
 	arg_value += summand;
 
 	if (arg.value >= 0)
-		return (int32_t)GET_INTEGER_PART(arg_value);
+		return (int)GET_INTEGER_PART(arg_value);
 	else
-		return -(int32_t)GET_INTEGER_PART(arg_value);
+		return -(int)GET_INTEGER_PART(arg_value);
 }
 
-int32_t dal_fixed31_32_ceil(
+int dal_fixed31_32_ceil(
 	struct fixed31_32 arg)
 {
-	uint64_t arg_value = abs_i64(arg.value);
+	unsigned long long arg_value = abs_i64(arg.value);
 
-	const int64_t summand = dal_fixed31_32_one.value -
+	const long long summand = dal_fixed31_32_one.value -
 		dal_fixed31_32_epsilon.value;
 
-	ASSERT(LLONG_MAX - (int64_t)arg_value >= summand);
+	ASSERT(LLONG_MAX - (long long)arg_value >= summand);
 
 	arg_value += summand;
 
 	if (arg.value >= 0)
-		return (int32_t)GET_INTEGER_PART(arg_value);
+		return (int)GET_INTEGER_PART(arg_value);
 	else
-		return -(int32_t)GET_INTEGER_PART(arg_value);
+		return -(int)GET_INTEGER_PART(arg_value);
 }
 
 /* this function is a generic helper to translate fixed point value to
@@ -535,15 +535,15 @@ int32_t dal_fixed31_32_ceil(
  * part in 32 bits. It is used in hw programming (scaler)
  */
 
-static inline uint32_t ux_dy(
-	int64_t value,
-	uint32_t integer_bits,
-	uint32_t fractional_bits)
+static inline unsigned int ux_dy(
+	long long value,
+	unsigned int integer_bits,
+	unsigned int fractional_bits)
 {
 	/* 1. create mask of integer part */
-	uint32_t result = (1 << integer_bits) - 1;
+	unsigned int result = (1 << integer_bits) - 1;
 	/* 2. mask out fractional part */
-	uint32_t fractional_part = FRACTIONAL_PART_MASK & value;
+	unsigned int fractional_part = FRACTIONAL_PART_MASK & value;
 	/* 3. shrink fixed point integer part to be of integer_bits width*/
 	result &= GET_INTEGER_PART(value);
 	/* 4. make space for fractional part to be filled in after integer */
@@ -554,13 +554,13 @@ static inline uint32_t ux_dy(
 	return result | fractional_part;
 }
 
-static inline uint32_t clamp_ux_dy(
-	int64_t value,
-	uint32_t integer_bits,
-	uint32_t fractional_bits,
-	uint32_t min_clamp)
+static inline unsigned int clamp_ux_dy(
+	long long value,
+	unsigned int integer_bits,
+	unsigned int fractional_bits,
+	unsigned int min_clamp)
 {
-	uint32_t truncated_val = ux_dy(value, integer_bits, fractional_bits);
+	unsigned int truncated_val = ux_dy(value, integer_bits, fractional_bits);
 
 	if (value >= (1LL << (integer_bits + FIXED31_32_BITS_PER_FRACTIONAL_PART)))
 		return (1 << (integer_bits + fractional_bits)) - 1;
@@ -570,35 +570,35 @@ static inline uint32_t clamp_ux_dy(
 		return min_clamp;
 }
 
-uint32_t dal_fixed31_32_u2d19(
+unsigned int dal_fixed31_32_u2d19(
 	struct fixed31_32 arg)
 {
 	return ux_dy(arg.value, 2, 19);
 }
 
-uint32_t dal_fixed31_32_u0d19(
+unsigned int dal_fixed31_32_u0d19(
 	struct fixed31_32 arg)
 {
 	return ux_dy(arg.value, 0, 19);
 }
 
-uint32_t dal_fixed31_32_clamp_u0d14(
+unsigned int dal_fixed31_32_clamp_u0d14(
 	struct fixed31_32 arg)
 {
 	return clamp_ux_dy(arg.value, 0, 14, 1);
 }
 
-uint32_t dal_fixed31_32_clamp_u0d10(
+unsigned int dal_fixed31_32_clamp_u0d10(
 	struct fixed31_32 arg)
 {
 	return clamp_ux_dy(arg.value, 0, 10, 1);
 }
 
-int32_t dal_fixed31_32_s4d19(
+int dal_fixed31_32_s4d19(
 	struct fixed31_32 arg)
 {
 	if (arg.value < 0)
-		return -(int32_t)ux_dy(dal_fixed31_32_abs(arg).value, 4, 19);
+		return -(int)ux_dy(dal_fixed31_32_abs(arg).value, 4, 19);
 	else
 		return ux_dy(arg.value, 4, 19);
 }
diff --git a/drivers/gpu/drm/amd/display/include/fixed31_32.h b/drivers/gpu/drm/amd/display/include/fixed31_32.h
index 0de2586..16cbdb4 100644
--- a/drivers/gpu/drm/amd/display/include/fixed31_32.h
+++ b/drivers/gpu/drm/amd/display/include/fixed31_32.h
@@ -26,8 +26,6 @@
 #ifndef __DAL_FIXED31_32_H__
 #define __DAL_FIXED31_32_H__
 
-#include "os_types.h"
-
 #define FIXED31_32_BITS_PER_FRACTIONAL_PART 32
 
 /*
@@ -44,7 +42,7 @@
  */
 
 struct fixed31_32 {
-	int64_t value;
+	long long value;
 };
 
 /*
@@ -73,15 +71,15 @@ static const struct fixed31_32 dal_fixed31_32_ln2_div_2 = { 1488522236LL };
  * result = numerator / denominator
  */
 struct fixed31_32 dal_fixed31_32_from_fraction(
-	int64_t numerator,
-	int64_t denominator);
+	long long numerator,
+	long long denominator);
 
 /*
  * @brief
  * result = arg
  */
-struct fixed31_32 dal_fixed31_32_from_int_nonconst(int64_t arg);
-static inline struct fixed31_32 dal_fixed31_32_from_int(int64_t arg)
+struct fixed31_32 dal_fixed31_32_from_int_nonconst(long long arg);
+static inline struct fixed31_32 dal_fixed31_32_from_int(long long arg)
 {
 	if (__builtin_constant_p(arg)) {
 		struct fixed31_32 res;
@@ -213,7 +211,7 @@ static inline struct fixed31_32 dal_fixed31_32_clamp(
  */
 struct fixed31_32 dal_fixed31_32_shl(
 	struct fixed31_32 arg,
-	uint8_t shift);
+	unsigned char shift);
 
 /*
  * @brief
@@ -221,7 +219,7 @@ struct fixed31_32 dal_fixed31_32_shl(
  */
 static inline struct fixed31_32 dal_fixed31_32_shr(
 	struct fixed31_32 arg,
-	uint8_t shift)
+	unsigned char shift)
 {
 	struct fixed31_32 res;
 	res.value = arg.value >> shift;
@@ -246,7 +244,7 @@ struct fixed31_32 dal_fixed31_32_add(
  * result = arg1 + arg2
  */
 static inline struct fixed31_32 dal_fixed31_32_add_int(struct fixed31_32 arg1,
-						       int32_t arg2)
+						       int arg2)
 {
 	return dal_fixed31_32_add(arg1,
 				  dal_fixed31_32_from_int(arg2));
@@ -265,7 +263,7 @@ struct fixed31_32 dal_fixed31_32_sub(
  * result = arg1 - arg2
  */
 static inline struct fixed31_32 dal_fixed31_32_sub_int(struct fixed31_32 arg1,
-						       int32_t arg2)
+						       int arg2)
 {
 	return dal_fixed31_32_sub(arg1,
 				  dal_fixed31_32_from_int(arg2));
@@ -291,7 +289,7 @@ struct fixed31_32 dal_fixed31_32_mul(
  * result = arg1 * arg2
  */
 static inline struct fixed31_32 dal_fixed31_32_mul_int(struct fixed31_32 arg1,
-						       int32_t arg2)
+						       int arg2)
 {
 	return dal_fixed31_32_mul(arg1,
 				  dal_fixed31_32_from_int(arg2));
@@ -309,7 +307,7 @@ struct fixed31_32 dal_fixed31_32_sqr(
  * result = arg1 / arg2
  */
 static inline struct fixed31_32 dal_fixed31_32_div_int(struct fixed31_32 arg1,
-						       int64_t arg2)
+						       long long arg2)
 {
 	return dal_fixed31_32_from_fraction(arg1.value,
 					    dal_fixed31_32_from_int(arg2).value);
@@ -434,21 +432,21 @@ struct fixed31_32 dal_fixed31_32_pow(
  * @brief
  * result = floor(arg) := greatest integer lower than or equal to arg
  */
-int32_t dal_fixed31_32_floor(
+int dal_fixed31_32_floor(
 	struct fixed31_32 arg);
 
 /*
  * @brief
  * result = round(arg) := integer nearest to arg
  */
-int32_t dal_fixed31_32_round(
+int dal_fixed31_32_round(
 	struct fixed31_32 arg);
 
 /*
  * @brief
  * result = ceil(arg) := lowest integer greater than or equal to arg
  */
-int32_t dal_fixed31_32_ceil(
+int dal_fixed31_32_ceil(
 	struct fixed31_32 arg);
 
 /* the following two function are used in scaler hw programming to convert fixed
@@ -457,20 +455,20 @@ int32_t dal_fixed31_32_ceil(
  * fractional
  */
 
-uint32_t dal_fixed31_32_u2d19(
+unsigned int dal_fixed31_32_u2d19(
 	struct fixed31_32 arg);
 
-uint32_t dal_fixed31_32_u0d19(
+unsigned int dal_fixed31_32_u0d19(
 	struct fixed31_32 arg);
 
 
-uint32_t dal_fixed31_32_clamp_u0d14(
+unsigned int dal_fixed31_32_clamp_u0d14(
 	struct fixed31_32 arg);
 
-uint32_t dal_fixed31_32_clamp_u0d10(
+unsigned int dal_fixed31_32_clamp_u0d10(
 	struct fixed31_32 arg);
 
-int32_t dal_fixed31_32_s4d19(
+int dal_fixed31_32_s4d19(
 	struct fixed31_32 arg);
 
 #endif
-- 
cgit v1.1


From 586f27a3c2b04e041c2d51c6c2374fbfa3075407 Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Mon, 26 Mar 2018 16:19:18 -0400
Subject: drm/amd/display: csc_transform to dc_csc_transform

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dc.h               | 4 ++--
 drivers/gpu/drm/amd/display/dc/dc_hw_types.h      | 5 +++++
 drivers/gpu/drm/amd/display/dc/dc_stream.h        | 2 +-
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c  | 2 +-
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h  | 2 +-
 drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h       | 2 +-
 drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h | 5 -----
 drivers/gpu/drm/amd/display/dc/inc/hw/ipp.h       | 2 +-
 drivers/gpu/drm/amd/display/dc/inc/hw/transform.h | 2 +-
 9 files changed, 13 insertions(+), 13 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index 5b81ae5..588672c 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -457,7 +457,7 @@ struct dc_plane_state {
 	struct dc_gamma *gamma_correction;
 	struct dc_transfer_func *in_transfer_func;
 	struct dc_bias_and_scale *bias_and_scale;
-	struct csc_transform input_csc_color_matrix;
+	struct dc_csc_transform input_csc_color_matrix;
 	struct fixed31_32 coeff_reduction_factor;
 	uint32_t sdr_white_level;
 
@@ -522,7 +522,7 @@ struct dc_surface_update {
 	struct dc_gamma *gamma;
 	struct dc_transfer_func *in_transfer_func;
 
-	struct csc_transform *input_csc_color_matrix;
+	struct dc_csc_transform *input_csc_color_matrix;
 	struct fixed31_32 *coeff_reduction_factor;
 };
 
diff --git a/drivers/gpu/drm/amd/display/dc/dc_hw_types.h b/drivers/gpu/drm/amd/display/dc/dc_hw_types.h
index b83a7dc..b1f7057 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_hw_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_hw_types.h
@@ -423,6 +423,11 @@ enum dc_gamma_type {
 	GAMMA_CS_TFM_1D = 3,
 };
 
+struct dc_csc_transform {
+	uint16_t matrix[12];
+	bool enable_adjustment;
+};
+
 struct dc_gamma {
 	struct kref refcount;
 	enum dc_gamma_type type;
diff --git a/drivers/gpu/drm/amd/display/dc/dc_stream.h b/drivers/gpu/drm/amd/display/dc/dc_stream.h
index 08f1a45..ed3c39f 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_stream.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_stream.h
@@ -61,7 +61,7 @@ struct dc_stream_state {
 	struct dc_info_packet hdr_static_metadata;
 	struct dc_transfer_func *out_transfer_func;
 	struct colorspace_transform gamut_remap_matrix;
-	struct csc_transform csc_color_matrix;
+	struct dc_csc_transform csc_color_matrix;
 
 	enum dc_color_space output_color_space;
 	enum dc_dither_option dither_option;
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
index 5f40a73..c008a71 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
@@ -264,7 +264,7 @@ void dpp1_cnv_setup (
 		struct dpp *dpp_base,
 		enum surface_pixel_format format,
 		enum expansion_mode mode,
-		struct csc_transform input_csc_color_matrix,
+		struct dc_csc_transform input_csc_color_matrix,
 		enum dc_color_space input_color_space)
 {
 	uint32_t pixel_format;
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h
index 9b5ff76..3fccf99 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h
@@ -1451,7 +1451,7 @@ void dpp1_cnv_setup (
 		struct dpp *dpp_base,
 		enum surface_pixel_format format,
 		enum expansion_mode mode,
-		struct csc_transform input_csc_color_matrix,
+		struct dc_csc_transform input_csc_color_matrix,
 		enum dc_color_space input_color_space);
 
 void dpp1_full_bypass(struct dpp *dpp_base);
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h
index 9999560..bb7af1b 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h
@@ -117,7 +117,7 @@ struct dpp_funcs {
 			struct dpp *dpp_base,
 			enum surface_pixel_format format,
 			enum expansion_mode mode,
-			struct csc_transform input_csc_color_matrix,
+			struct dc_csc_transform input_csc_color_matrix,
 			enum dc_color_space input_color_space);
 
 	void (*dpp_full_bypass)(struct dpp *dpp_base);
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h b/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h
index 93da445..9fe7302 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h
@@ -140,11 +140,6 @@ enum opp_regamma {
 	OPP_REGAMMA_USER
 };
 
-struct csc_transform {
-	uint16_t matrix[12];
-	bool enable_adjustment;
-};
-
 struct dc_bias_and_scale {
 	uint16_t scale_red;
 	uint16_t bias_red;
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/ipp.h b/drivers/gpu/drm/amd/display/dc/inc/hw/ipp.h
index 2109eac..b2fa4c4 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/ipp.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/ipp.h
@@ -87,7 +87,7 @@ struct ipp_funcs {
 		struct input_pixel_processor *ipp,
 		enum surface_pixel_format format,
 		enum expansion_mode mode,
-		struct csc_transform input_csc_color_matrix,
+		struct dc_csc_transform input_csc_color_matrix,
 		enum dc_color_space input_color_space);
 
 	/* DCE function to setup IPP.  TODO: see if we can consolidate to setup */
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/transform.h b/drivers/gpu/drm/amd/display/dc/inc/hw/transform.h
index c5b3623..fecc80c 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/transform.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/transform.h
@@ -252,7 +252,7 @@ struct transform_funcs {
 			struct transform *xfm_base,
 			enum surface_pixel_format format,
 			enum expansion_mode mode,
-			struct csc_transform input_csc_color_matrix,
+			struct dc_csc_transform input_csc_color_matrix,
 			enum dc_color_space input_color_space);
 
 	void (*ipp_full_bypass)(struct transform *xfm_base);
-- 
cgit v1.1


From a2e8f540c4efa7a8f180c910d202469b3ed4f5ba Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Mon, 9 Apr 2018 14:57:47 -0500
Subject: drm/amd/display: Refactor color module

Remove some unnecessary TF definitions from update structures

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dc_stream.h | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dc_stream.h b/drivers/gpu/drm/amd/display/dc/dc_stream.h
index ed3c39f..d7e6d53 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_stream.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_stream.h
@@ -67,7 +67,6 @@ struct dc_stream_state {
 	enum dc_dither_option dither_option;
 
 	enum view_3d_format view_format;
-	enum color_transfer_func output_tf;
 
 	bool ignore_msa_timing_param;
 
@@ -113,9 +112,9 @@ struct dc_stream_update {
 	struct rect src;
 	struct rect dst;
 	struct dc_transfer_func *out_transfer_func;
-	enum color_transfer_func color_output_tf;
 	struct dc_info_packet *hdr_static_metadata;
 	unsigned int *abm_level;
+
 	unsigned long long *periodic_fn_vsync_delta;
 };
 
-- 
cgit v1.1


From 28177772cbf693a6960e92bba1f08a0e78acb048 Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Mon, 26 Mar 2018 16:29:51 -0400
Subject: drm/amd/display: move color_transfer_func to color mod

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dc.h | 14 +-------------
 1 file changed, 1 insertion(+), 13 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index 588672c..0f566a1 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -333,19 +333,6 @@ enum {
 	TRANSFER_FUNC_POINTS = 1025
 };
 
-enum color_transfer_func {
-	transfer_func_unknown,
-	transfer_func_srgb,
-	transfer_func_bt709,
-	transfer_func_pq2084,
-	transfer_func_pq2084_interim,
-	transfer_func_linear_0_1,
-	transfer_func_linear_0_125,
-	transfer_func_dolbyvision,
-	transfer_func_gamma_22,
-	transfer_func_gamma_26
-};
-
 struct dc_hdr_static_metadata {
 	/* display chromaticities and white point in units of 0.00001 */
 	unsigned int chromaticity_green_x;
@@ -693,6 +680,7 @@ struct dc_cursor {
 	struct dc_cursor_attributes attributes;
 };
 
+
 /*******************************************************************************
  * Interrupt interfaces
  ******************************************************************************/
-- 
cgit v1.1


From 754e3673201eb192be68b8a07f2e448d75f69dfe Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Tue, 27 Mar 2018 23:12:21 -0400
Subject: drm/amd/display: Fix structure initialization of hdmi_info_packet

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_resource.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
index e1036e4..8d7bc1f 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
@@ -1853,11 +1853,13 @@ static void set_avi_info_frame(
 	unsigned int cn0_cn1_value = 0;
 	uint8_t *check_sum = NULL;
 	uint8_t byte_index = 0;
-	union hdmi_info_packet hdmi_info = {0};
+	union hdmi_info_packet hdmi_info;
 	union display_content_support support = {0};
 	unsigned int vic = pipe_ctx->stream->timing.vic;
 	enum dc_timing_3d_format format;
 
+	memset(&hdmi_info, 0, sizeof(union hdmi_info_packet));
+
 	color_space = pipe_ctx->stream->output_color_space;
 	if (color_space == COLOR_SPACE_UNKNOWN)
 		color_space = (stream->timing.pixel_encoding == PIXEL_ENCODING_RGB) ?
-- 
cgit v1.1


From e43a432c018a9a2c2641e1f8c08a836cc83982cd Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Tue, 27 Mar 2018 16:43:56 -0400
Subject: drm/amd/display: Have DC manage its own allocation of gamma

Creating plane will also allocate gamma and input TF
Creating stream will also allocate outputTF

Fix issue with gamma not applied
OS may call SetGamma before surface committed, so need to store
in target and apply later.

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c         | 6 ++----
 drivers/gpu/drm/amd/display/dc/core/dc_stream.c           | 6 ++++--
 drivers/gpu/drm/amd/display/dc/core/dc_surface.c          | 8 +++++++-
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c | 5 ++---
 4 files changed, 15 insertions(+), 10 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index aa8e25a..18f221b 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -2200,7 +2200,6 @@ fill_stream_properties_from_drm_display_mode(struct dc_stream_state *stream,
 					     const struct drm_connector *connector)
 {
 	struct dc_crtc_timing *timing_out = &stream->timing;
-	struct dc_transfer_func *tf = dc_create_transfer_func();
 
 	memset(timing_out, 0, sizeof(struct dc_crtc_timing));
 
@@ -2244,9 +2243,8 @@ fill_stream_properties_from_drm_display_mode(struct dc_stream_state *stream,
 
 	stream->output_color_space = get_output_color_space(timing_out);
 
-	tf->type = TF_TYPE_PREDEFINED;
-	tf->tf = TRANSFER_FUNCTION_SRGB;
-	stream->out_transfer_func = tf;
+	stream->out_transfer_func->type = TF_TYPE_PREDEFINED;
+	stream->out_transfer_func->tf = TRANSFER_FUNCTION_SRGB;
 }
 
 static void fill_audio_info(struct audio_info *audio_info,
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c
index ce0747e..3b2ddbd 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c
@@ -101,14 +101,16 @@ static void construct(struct dc_stream_state *stream,
 	stream->status.link = stream->sink->link;
 
 	update_stream_signal(stream);
+
+	stream->out_transfer_func = dc_create_transfer_func();
+	stream->out_transfer_func->type = TF_TYPE_BYPASS;
 }
 
 static void destruct(struct dc_stream_state *stream)
 {
 	dc_sink_release(stream->sink);
 	if (stream->out_transfer_func != NULL) {
-		dc_transfer_func_release(
-				stream->out_transfer_func);
+		dc_transfer_func_release(stream->out_transfer_func);
 		stream->out_transfer_func = NULL;
 	}
 }
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_surface.c b/drivers/gpu/drm/amd/display/dc/core/dc_surface.c
index ade5b8e..9593877 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_surface.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_surface.c
@@ -38,6 +38,12 @@
 static void construct(struct dc_context *ctx, struct dc_plane_state *plane_state)
 {
 	plane_state->ctx = ctx;
+
+	plane_state->gamma_correction = dc_create_gamma();
+	plane_state->gamma_correction->is_identity = true;
+
+	plane_state->in_transfer_func = dc_create_transfer_func();
+	plane_state->in_transfer_func->type = TF_TYPE_BYPASS;
 }
 
 static void destruct(struct dc_plane_state *plane_state)
@@ -175,7 +181,7 @@ void dc_transfer_func_release(struct dc_transfer_func *tf)
 	kref_put(&tf->refcount, dc_transfer_func_free);
 }
 
-struct dc_transfer_func *dc_create_transfer_func(void)
+struct dc_transfer_func *dc_create_transfer_func()
 {
 	struct dc_transfer_func *tf = kzalloc(sizeof(*tf), GFP_KERNEL);
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
index f3341a2..a6cf9ad 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -956,9 +956,8 @@ static bool dcn10_set_input_transfer_func(struct pipe_ctx *pipe_ctx,
 		tf = plane_state->in_transfer_func;
 
 	if (plane_state->gamma_correction &&
-		plane_state->gamma_correction->is_identity)
-		dpp_base->funcs->dpp_set_degamma(dpp_base, IPP_DEGAMMA_MODE_BYPASS);
-	else if (plane_state->gamma_correction && dce_use_lut(plane_state->format))
+		!plane_state->gamma_correction->is_identity
+			&& dce_use_lut(plane_state->format))
 		dpp_base->funcs->dpp_program_input_lut(dpp_base, plane_state->gamma_correction);
 
 	if (tf == NULL)
-- 
cgit v1.1


From 84ffa80123f56f80145dc638f21dfcbedda5610d Mon Sep 17 00:00:00 2001
From: "Leo (Sunpeng) Li" <sunpeng.li@amd.com>
Date: Thu, 29 Mar 2018 17:04:12 -0400
Subject: drm/amd/display: Fix dim display on DCE11

Before programming the input gamma, check that we're not using the
identity correction.

Signed-off-by: Leo (Sunpeng) Li <sunpeng.li@amd.com>
Reviewed-by: Anthony Koo <Anthony.Koo@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
index 78bf4fa..52427ae 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
@@ -280,7 +280,9 @@ dce110_set_input_transfer_func(struct pipe_ctx *pipe_ctx,
 	build_prescale_params(&prescale_params, plane_state);
 	ipp->funcs->ipp_program_prescale(ipp, &prescale_params);
 
-	if (plane_state->gamma_correction && dce_use_lut(plane_state->format))
+	if (plane_state->gamma_correction &&
+			!plane_state->gamma_correction->is_identity &&
+			dce_use_lut(plane_state->format))
 		ipp->funcs->ipp_program_input_lut(ipp, plane_state->gamma_correction);
 
 	if (tf == NULL) {
-- 
cgit v1.1


From e405c2173e8c1c8e2e823592581ba3f774e7811c Mon Sep 17 00:00:00 2001
From: "Leo (Sunpeng) Li" <sunpeng.li@amd.com>
Date: Wed, 4 Apr 2018 16:01:30 -0400
Subject: drm/amd/display: Fix memleak on input transfer function

Input transfer function creation is now done when the plane is created.
This is done within the following change:

    Author: Anthony Koo <Anthony.Koo@amd.com>
        drm/amd/display: Have DC manage its own allocation of gamma

Therefore, we no longer need to create it when filling in the plane
attributes.

Signed-off-by: Leo (Sunpeng) Li <sunpeng.li@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 8 --------
 1 file changed, 8 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 18f221b..265f0166 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -2011,7 +2011,6 @@ static int fill_plane_attributes(struct amdgpu_device *adev,
 	const struct amdgpu_framebuffer *amdgpu_fb =
 		to_amdgpu_framebuffer(plane_state->fb);
 	const struct drm_crtc *crtc = plane_state->crtc;
-	struct dc_transfer_func *input_tf;
 	int ret = 0;
 
 	if (!fill_rects_from_plane_state(plane_state, dc_plane_state))
@@ -2025,13 +2024,6 @@ static int fill_plane_attributes(struct amdgpu_device *adev,
 	if (ret)
 		return ret;
 
-	input_tf = dc_create_transfer_func();
-
-	if (input_tf == NULL)
-		return -ENOMEM;
-
-	dc_plane_state->in_transfer_func = input_tf;
-
 	/*
 	 * Always set input transfer function, since plane state is refreshed
 	 * every time.
-- 
cgit v1.1


From eb40c86a83211827338e4a58b8e925a45a19d633 Mon Sep 17 00:00:00 2001
From: Nico Sneck <nicosneck@hotmail.com>
Date: Sat, 7 Apr 2018 15:13:04 +0000
Subject: drm/radeon: add PX quirk for Asus K73TK

With this the dGPU turns on correctly.

Signed-off-by: Nico Sneck <nicosneck@hotmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
---
 drivers/gpu/drm/radeon/radeon_device.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c
index 90e17e2..59c8a66 100644
--- a/drivers/gpu/drm/radeon/radeon_device.c
+++ b/drivers/gpu/drm/radeon/radeon_device.c
@@ -140,6 +140,10 @@ static struct radeon_px_quirk radeon_px_quirk_list[] = {
 	 * https://bugs.freedesktop.org/show_bug.cgi?id=101491
 	 */
 	{ PCI_VENDOR_ID_ATI, 0x6741, 0x1043, 0x2122, RADEON_PX_QUIRK_DISABLE_PX },
+	/* Asus K73TK laptop with AMD A6-3420M APU and Radeon 7670m GPU
+	 * https://bugzilla.kernel.org/show_bug.cgi?id=51381#c52
+	 */
+	{ PCI_VENDOR_ID_ATI, 0x6840, 0x1043, 0x2123, RADEON_PX_QUIRK_DISABLE_PX },
 	{ 0, 0, 0, 0, 0 },
 };
 
-- 
cgit v1.1


From 1bb5afd768b950e9ddcb62b3c31bce8bed3ef774 Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Sun, 8 Apr 2018 14:39:18 +0800
Subject: drm/amdgpu: fix null pointer panic with direct fw loading on gpu
 reset

When system uses fw direct loading, then psp context structure won't be
initiliazed. And it is also unable to execute mode reset.

[  434.601474] amdgpu 0000:0c:00.0: GPU reset begin!
[  434.694326] amdgpu 0000:0c:00.0: GPU reset
[  434.743152] BUG: unable to handle kernel NULL pointer dereference at
0000000000000058
[  434.838474] IP: psp_gpu_reset+0xc/0x30 [amdgpu]
[  434.893532] PGD 406ed9067
[  434.893533] P4D 406ed9067
[  434.926376] PUD 400b46067
[  434.959217] PMD 0
[  435.033379] Oops: 0000 [#1] SMP
[  435.072573] Modules linked in: amdgpu(OE) chash(OE) gpu_sched(OE) ttm(OE)
drm_kms_helper(OE) drm(OE) fb_sys_fops syscopyarea sysfillrect sysimgblt
rpcsec_gss_krb5 auth_rpcgss nfsv4 nfs lockd grace fscache snd_hda_codec_realtek
snd_hda_codec_generic snd_hda_codec_hdmi snd_hda_intel snd_hda_codec
snd_hda_core snd_hwdep snd_pcm edac_mce_amd snd_seq_midi snd_seq_midi_event
kvm_amd snd_rawmidi kvm irqbypass crct10dif_pclmul crc32_pclmul snd_seq
ghash_clmulni_intel snd_seq_device pcbc snd_timer eeepc_wmi aesni_intel snd
asus_wmi aes_x86_64 sparse_keymap crypto_simd glue_helper joydev soundcore
wmi_bmof cryptd video i2c_piix4 shpchp 8250_dw i2c_designware_platform mac_hid
i2c_designware_core sunrpc parport_pc ppdev lp parport autofs4 hid_generic igb
usbhid dca ptp mxm_wmi pps_core ahci hid i2c_algo_bit
[  435.931754]  libahci wmi

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
index 19e71f4..c7d43e0 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
@@ -505,6 +505,9 @@ failed:
 
 int psp_gpu_reset(struct amdgpu_device *adev)
 {
+	if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP)
+		return 0;
+
 	return psp_mode1_reset(&adev->psp);
 }
 
-- 
cgit v1.1


From f73f9e35a2a7c8cee3691a4b7313bbc3b95eec6b Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Wed, 4 Apr 2018 12:36:57 +0800
Subject: drm/amd/pp: Refine pp_atomfwctrl_get_vbios_bootup_values

In order to share pp_atomfwctrl_get_vbios_bootup_values
on asics with different BIOS_CLKID.
Not call function pp_atomfwctrl_get_clk_information_by_clkid in
pp_atomfwctrl_get_vbios_bootup_values.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c | 9 +--------
 drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.h | 2 ++
 drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c | 6 ++++++
 3 files changed, 9 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c
index 0adaf36..c97b0e5 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c
@@ -488,7 +488,7 @@ int pp_atomfwctrl_get_gpio_information(struct pp_hwmgr *hwmgr,
 	return 0;
 }
 
-int pp_atomfwctrl__get_clk_information_by_clkid(struct pp_hwmgr *hwmgr, BIOS_CLKID id, uint32_t *frequency)
+int pp_atomfwctrl_get_clk_information_by_clkid(struct pp_hwmgr *hwmgr, BIOS_CLKID id, uint32_t *frequency)
 {
 	struct amdgpu_device *adev = hwmgr->adev;
 	struct atom_get_smu_clock_info_parameters_v3_1   parameters;
@@ -515,7 +515,6 @@ int pp_atomfwctrl_get_vbios_bootup_values(struct pp_hwmgr *hwmgr,
 {
 	struct atom_firmware_info_v3_1 *info = NULL;
 	uint16_t ix;
-	uint32_t frequency = 0;
 
 	ix = GetIndexIntoMasterDataTable(firmwareinfo);
 	info = (struct atom_firmware_info_v3_1 *)
@@ -538,12 +537,6 @@ int pp_atomfwctrl_get_vbios_bootup_values(struct pp_hwmgr *hwmgr,
 	boot_values->ulSocClk   = 0;
 	boot_values->ulDCEFClk   = 0;
 
-	if (!pp_atomfwctrl__get_clk_information_by_clkid(hwmgr, SMU9_SYSPLL0_SOCCLK_ID, &frequency))
-		boot_values->ulSocClk   = frequency;
-
-	if (!pp_atomfwctrl__get_clk_information_by_clkid(hwmgr, SMU9_SYSPLL0_DCEFCLK_ID, &frequency))
-		boot_values->ulDCEFClk   = frequency;
-
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.h b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.h
index 8df1e84f..fe10aa4 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.h
@@ -230,6 +230,8 @@ int pp_atomfwctrl_get_vbios_bootup_values(struct pp_hwmgr *hwmgr,
 			struct pp_atomfwctrl_bios_boot_up_values *boot_values);
 int pp_atomfwctrl_get_smc_dpm_information(struct pp_hwmgr *hwmgr,
 			struct pp_atomfwctrl_smc_dpm_parameters *param);
+int pp_atomfwctrl_get_clk_information_by_clkid(struct pp_hwmgr *hwmgr,
+					BIOS_CLKID id, uint32_t *frequency);
 
 #endif
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
index c9fb4b2..ba29942 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
@@ -2481,6 +2481,12 @@ static int vega10_init_smc_table(struct pp_hwmgr *hwmgr)
 		data->vbios_boot_state.mvddc    = boot_up_values.usMvddc;
 		data->vbios_boot_state.gfx_clock = boot_up_values.ulGfxClk;
 		data->vbios_boot_state.mem_clock = boot_up_values.ulUClk;
+		pp_atomfwctrl_get_clk_information_by_clkid(hwmgr,
+				SMU9_SYSPLL0_SOCCLK_ID, &boot_up_values.ulSocClk);
+
+		pp_atomfwctrl_get_clk_information_by_clkid(hwmgr,
+				SMU9_SYSPLL0_DCEFCLK_ID, &boot_up_values.ulDCEFClk);
+
 		data->vbios_boot_state.soc_clock = boot_up_values.ulSocClk;
 		data->vbios_boot_state.dcef_clock = boot_up_values.ulDCEFClk;
 		if (0 != boot_up_values.usVddc) {
-- 
cgit v1.1


From e6636ae1b7aab30a1fb4ea7805b5b6b2494eca71 Mon Sep 17 00:00:00 2001
From: Evan Quan <evan.quan@amd.com>
Date: Tue, 10 Apr 2018 12:30:59 +0800
Subject: drm/amdgpu: add MP1 and THM hw ip base reg offset

Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu.h          | 2 ++
 drivers/gpu/drm/amd/amdgpu/vega10_reg_init.c | 3 ++-
 2 files changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index 21272ce..7e5defb 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -1391,6 +1391,7 @@ enum amd_hw_ip_block_type {
 	ATHUB_HWIP,
 	NBIO_HWIP,
 	MP0_HWIP,
+	MP1_HWIP,
 	UVD_HWIP,
 	VCN_HWIP = UVD_HWIP,
 	VCE_HWIP,
@@ -1400,6 +1401,7 @@ enum amd_hw_ip_block_type {
 	SMUIO_HWIP,
 	PWR_HWIP,
 	NBIF_HWIP,
+	THM_HWIP,
 	MAX_HWIP
 };
 
diff --git a/drivers/gpu/drm/amd/amdgpu/vega10_reg_init.c b/drivers/gpu/drm/amd/amdgpu/vega10_reg_init.c
index 4c45db7..45aafca 100644
--- a/drivers/gpu/drm/amd/amdgpu/vega10_reg_init.c
+++ b/drivers/gpu/drm/amd/amdgpu/vega10_reg_init.c
@@ -38,6 +38,7 @@ int vega10_reg_base_init(struct amdgpu_device *adev)
 		adev->reg_offset[ATHUB_HWIP][i] = (uint32_t *)(&(ATHUB_BASE.instance[i]));
 		adev->reg_offset[NBIO_HWIP][i] = (uint32_t *)(&(NBIO_BASE.instance[i]));
 		adev->reg_offset[MP0_HWIP][i] = (uint32_t *)(&(MP0_BASE.instance[i]));
+		adev->reg_offset[MP1_HWIP][i] = (uint32_t *)(&(MP1_BASE.instance[i]));
 		adev->reg_offset[UVD_HWIP][i] = (uint32_t *)(&(UVD_BASE.instance[i]));
 		adev->reg_offset[VCE_HWIP][i] = (uint32_t *)(&(VCE_BASE.instance[i]));
 		adev->reg_offset[VCN_HWIP][i] = (uint32_t *)(&(VCN_BASE.instance[i]));
@@ -49,7 +50,7 @@ int vega10_reg_base_init(struct amdgpu_device *adev)
 		adev->reg_offset[SMUIO_HWIP][i] = (uint32_t *)(&(SMUIO_BASE.instance[i]));
 		adev->reg_offset[PWR_HWIP][i] = (uint32_t *)(&(PWR_BASE.instance[i]));
 		adev->reg_offset[NBIF_HWIP][i] = (uint32_t *)(&(NBIF_BASE.instance[i]));
-
+		adev->reg_offset[THM_HWIP][i] = (uint32_t *)(&(THM_BASE.instance[i]));
 	}
 	return 0;
 }
-- 
cgit v1.1


From b8a5559112714bb328330dbf2a4a1912e8c7a462 Mon Sep 17 00:00:00 2001
From: Evan Quan <evan.quan@amd.com>
Date: Tue, 10 Apr 2018 12:32:16 +0800
Subject: drm/amd/pp: use soc15 common macros instead of vega10 specific

pp_soc15.h is vega10 specific. Update powerplay code to use soc15 common
macros defined in soc15_common.h.

Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c  |   7 +-
 drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c |  16 +--
 .../gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c |  50 ++++------
 .../gpu/drm/amd/powerplay/hwmgr/vega10_thermal.c   | 107 ++++++++-------------
 drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c |   1 -
 .../gpu/drm/amd/powerplay/hwmgr/vega12_thermal.c   |  37 +++----
 drivers/gpu/drm/amd/powerplay/inc/pp_soc15.h       |  52 ----------
 .../gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c    |  37 +++----
 .../gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c   |  50 ++++------
 .../gpu/drm/amd/powerplay/smumgr/vega12_smumgr.c   |  56 ++++-------
 10 files changed, 133 insertions(+), 280 deletions(-)
 delete mode 100644 drivers/gpu/drm/amd/powerplay/inc/pp_soc15.h

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
index 055358b..6ba3b1f 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
@@ -34,7 +34,7 @@
 #include "rv_ppsmc.h"
 #include "smu10_hwmgr.h"
 #include "power_state.h"
-#include "pp_soc15.h"
+#include "soc15_common.h"
 
 #define SMU10_MAX_DEEPSLEEP_DIVIDER_ID     5
 #define SMU10_MINIMUM_ENGINE_CLOCK         800   /* 8Mhz, the low boundary of engine clock allowed on this chip */
@@ -947,9 +947,8 @@ static int smu10_get_max_high_clocks(struct pp_hwmgr *hwmgr, struct amd_pp_simpl
 
 static int smu10_thermal_get_temperature(struct pp_hwmgr *hwmgr)
 {
-	uint32_t reg_offset = soc15_get_register_offset(THM_HWID, 0,
-			mmTHM_TCON_CUR_TMP_BASE_IDX, mmTHM_TCON_CUR_TMP);
-	uint32_t reg_value = cgs_read_register(hwmgr->device, reg_offset);
+	struct amdgpu_device *adev = hwmgr->adev;
+	uint32_t reg_value = RREG32_SOC15(THM, 0, mmTHM_TCON_CUR_TMP);
 	int cur_temp =
 		(reg_value & THM_TCON_CUR_TMP__CUR_TEMP_MASK) >> THM_TCON_CUR_TMP__CUR_TEMP__SHIFT;
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
index ba29942..f6427c8 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
@@ -36,7 +36,7 @@
 #include "smu9.h"
 #include "smu9_driver_if.h"
 #include "vega10_inc.h"
-#include "pp_soc15.h"
+#include "soc15_common.h"
 #include "pppcielanes.h"
 #include "vega10_hwmgr.h"
 #include "vega10_processpptables.h"
@@ -754,7 +754,6 @@ static int vega10_hwmgr_backend_init(struct pp_hwmgr *hwmgr)
 	uint32_t config_telemetry = 0;
 	struct pp_atomfwctrl_voltage_table vol_table;
 	struct amdgpu_device *adev = hwmgr->adev;
-	uint32_t reg;
 
 	data = kzalloc(sizeof(struct vega10_hwmgr), GFP_KERNEL);
 	if (data == NULL)
@@ -860,10 +859,7 @@ static int vega10_hwmgr_backend_init(struct pp_hwmgr *hwmgr)
 			advanceFanControlParameters.usFanPWMMinLimit *
 			hwmgr->thermal_controller.fanInfo.ulMaxRPM / 100;
 
-	reg = soc15_get_register_offset(DF_HWID, 0,
-			mmDF_CS_AON0_DramBaseAddress0_BASE_IDX,
-			mmDF_CS_AON0_DramBaseAddress0);
-	data->mem_channels = (cgs_read_register(hwmgr->device, reg) &
+	data->mem_channels = (RREG32_SOC15(DF, 0, mmDF_CS_AON0_DramBaseAddress0) &
 			DF_CS_AON0_DramBaseAddress0__IntLvNumChan_MASK) >>
 			DF_CS_AON0_DramBaseAddress0__IntLvNumChan__SHIFT;
 	PP_ASSERT_WITH_CODE(data->mem_channels < ARRAY_SIZE(channel_number),
@@ -3808,11 +3804,12 @@ static int vega10_get_gpu_power(struct pp_hwmgr *hwmgr,
 static int vega10_read_sensor(struct pp_hwmgr *hwmgr, int idx,
 			      void *value, int *size)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	uint32_t sclk_idx, mclk_idx, activity_percent = 0;
 	struct vega10_hwmgr *data = hwmgr->backend;
 	struct vega10_dpm_table *dpm_table = &data->dpm_table;
 	int ret = 0;
-	uint32_t reg, val_vid;
+	uint32_t val_vid;
 
 	switch (idx) {
 	case AMDGPU_PP_SENSOR_GFX_SCLK:
@@ -3862,10 +3859,7 @@ static int vega10_read_sensor(struct pp_hwmgr *hwmgr, int idx,
 		}
 		break;
 	case AMDGPU_PP_SENSOR_VDDGFX:
-		reg = soc15_get_register_offset(SMUIO_HWID, 0,
-			mmSMUSVI0_PLANE0_CURRENTVID_BASE_IDX,
-			mmSMUSVI0_PLANE0_CURRENTVID);
-		val_vid = (cgs_read_register(hwmgr->device, reg) &
+		val_vid = (RREG32_SOC15(SMUIO, 0, mmSMUSVI0_PLANE0_CURRENTVID) &
 			SMUSVI0_PLANE0_CURRENTVID__CURRENT_SVI0_PLANE0_VID_MASK) >>
 			SMUSVI0_PLANE0_CURRENTVID__CURRENT_SVI0_PLANE0_VID__SHIFT;
 		*((uint32_t *)value) = (uint32_t)convert_to_vddc((uint8_t)val_vid);
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c
index 203a691..a9efd855 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c
@@ -27,7 +27,7 @@
 #include "vega10_ppsmc.h"
 #include "vega10_inc.h"
 #include "pp_debug.h"
-#include "pp_soc15.h"
+#include "soc15_common.h"
 
 static const struct vega10_didt_config_reg SEDiDtTuningCtrlConfig_Vega10[] =
 {
@@ -888,36 +888,36 @@ static void vega10_didt_set_mask(struct pp_hwmgr *hwmgr, const bool enable)
 	if (PP_CAP(PHM_PlatformCaps_DiDtEDCEnable)) {
 		if (PP_CAP(PHM_PlatformCaps_SQRamping)) {
 			data = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_SQ_EDC_CTRL);
-			data = CGS_REG_SET_FIELD(data, DIDT_SQ_EDC_CTRL, EDC_EN, en);
-			data = CGS_REG_SET_FIELD(data, DIDT_SQ_EDC_CTRL, EDC_SW_RST, ~en);
+			data = REG_SET_FIELD(data, DIDT_SQ_EDC_CTRL, EDC_EN, en);
+			data = REG_SET_FIELD(data, DIDT_SQ_EDC_CTRL, EDC_SW_RST, ~en);
 			cgs_write_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_SQ_EDC_CTRL, data);
 		}
 
 		if (PP_CAP(PHM_PlatformCaps_DBRamping)) {
 			data = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_DB_EDC_CTRL);
-			data = CGS_REG_SET_FIELD(data, DIDT_DB_EDC_CTRL, EDC_EN, en);
-			data = CGS_REG_SET_FIELD(data, DIDT_DB_EDC_CTRL, EDC_SW_RST, ~en);
+			data = REG_SET_FIELD(data, DIDT_DB_EDC_CTRL, EDC_EN, en);
+			data = REG_SET_FIELD(data, DIDT_DB_EDC_CTRL, EDC_SW_RST, ~en);
 			cgs_write_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_DB_EDC_CTRL, data);
 		}
 
 		if (PP_CAP(PHM_PlatformCaps_TDRamping)) {
 			data = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_TD_EDC_CTRL);
-			data = CGS_REG_SET_FIELD(data, DIDT_TD_EDC_CTRL, EDC_EN, en);
-			data = CGS_REG_SET_FIELD(data, DIDT_TD_EDC_CTRL, EDC_SW_RST, ~en);
+			data = REG_SET_FIELD(data, DIDT_TD_EDC_CTRL, EDC_EN, en);
+			data = REG_SET_FIELD(data, DIDT_TD_EDC_CTRL, EDC_SW_RST, ~en);
 			cgs_write_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_TD_EDC_CTRL, data);
 		}
 
 		if (PP_CAP(PHM_PlatformCaps_TCPRamping)) {
 			data = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_TCP_EDC_CTRL);
-			data = CGS_REG_SET_FIELD(data, DIDT_TCP_EDC_CTRL, EDC_EN, en);
-			data = CGS_REG_SET_FIELD(data, DIDT_TCP_EDC_CTRL, EDC_SW_RST, ~en);
+			data = REG_SET_FIELD(data, DIDT_TCP_EDC_CTRL, EDC_EN, en);
+			data = REG_SET_FIELD(data, DIDT_TCP_EDC_CTRL, EDC_SW_RST, ~en);
 			cgs_write_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_TCP_EDC_CTRL, data);
 		}
 
 		if (PP_CAP(PHM_PlatformCaps_DBRRamping)) {
 			data = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_DBR_EDC_CTRL);
-			data = CGS_REG_SET_FIELD(data, DIDT_DBR_EDC_CTRL, EDC_EN, en);
-			data = CGS_REG_SET_FIELD(data, DIDT_DBR_EDC_CTRL, EDC_SW_RST, ~en);
+			data = REG_SET_FIELD(data, DIDT_DBR_EDC_CTRL, EDC_EN, en);
+			data = REG_SET_FIELD(data, DIDT_DBR_EDC_CTRL, EDC_SW_RST, ~en);
 			cgs_write_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_DBR_EDC_CTRL, data);
 		}
 	}
@@ -933,17 +933,15 @@ static int vega10_enable_cac_driving_se_didt_config(struct pp_hwmgr *hwmgr)
 	struct amdgpu_device *adev = hwmgr->adev;
 	int result;
 	uint32_t num_se = 0, count, data;
-	uint32_t reg;
 
 	num_se = adev->gfx.config.max_shader_engines;
 
 	adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
 	mutex_lock(&adev->grbm_idx_mutex);
-	reg = soc15_get_register_offset(GC_HWID, 0, mmGRBM_GFX_INDEX_BASE_IDX, mmGRBM_GFX_INDEX);
 	for (count = 0; count < num_se; count++) {
 		data = GRBM_GFX_INDEX__INSTANCE_BROADCAST_WRITES_MASK | GRBM_GFX_INDEX__SH_BROADCAST_WRITES_MASK | ( count << GRBM_GFX_INDEX__SE_INDEX__SHIFT);
-		cgs_write_register(hwmgr->device, reg, data);
+		WREG32_SOC15(GC, 0, mmGRBM_GFX_INDEX, data);
 
 		result =  vega10_program_didt_config_registers(hwmgr, SEDiDtStallCtrlConfig_vega10, VEGA10_CONFIGREG_DIDT);
 		result |= vega10_program_didt_config_registers(hwmgr, SEDiDtStallPatternConfig_vega10, VEGA10_CONFIGREG_DIDT);
@@ -958,7 +956,7 @@ static int vega10_enable_cac_driving_se_didt_config(struct pp_hwmgr *hwmgr)
 		if (0 != result)
 			break;
 	}
-	cgs_write_register(hwmgr->device, reg, 0xE0000000);
+	WREG32_SOC15(GC, 0, mmGRBM_GFX_INDEX, 0xE0000000);
 	mutex_unlock(&adev->grbm_idx_mutex);
 
 	vega10_didt_set_mask(hwmgr, true);
@@ -986,17 +984,15 @@ static int vega10_enable_psm_gc_didt_config(struct pp_hwmgr *hwmgr)
 	struct amdgpu_device *adev = hwmgr->adev;
 	int result;
 	uint32_t num_se = 0, count, data;
-	uint32_t reg;
 
 	num_se = adev->gfx.config.max_shader_engines;
 
 	adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
 	mutex_lock(&adev->grbm_idx_mutex);
-	reg = soc15_get_register_offset(GC_HWID, 0, mmGRBM_GFX_INDEX_BASE_IDX, mmGRBM_GFX_INDEX);
 	for (count = 0; count < num_se; count++) {
 		data = GRBM_GFX_INDEX__INSTANCE_BROADCAST_WRITES_MASK | GRBM_GFX_INDEX__SH_BROADCAST_WRITES_MASK | ( count << GRBM_GFX_INDEX__SE_INDEX__SHIFT);
-		cgs_write_register(hwmgr->device, reg, data);
+		WREG32_SOC15(GC, 0, mmGRBM_GFX_INDEX, data);
 
 		result = vega10_program_didt_config_registers(hwmgr, SEDiDtStallCtrlConfig_vega10, VEGA10_CONFIGREG_DIDT);
 		result |= vega10_program_didt_config_registers(hwmgr, SEDiDtStallPatternConfig_vega10, VEGA10_CONFIGREG_DIDT);
@@ -1005,7 +1001,7 @@ static int vega10_enable_psm_gc_didt_config(struct pp_hwmgr *hwmgr)
 		if (0 != result)
 			break;
 	}
-	cgs_write_register(hwmgr->device, reg, 0xE0000000);
+	WREG32_SOC15(GC, 0, mmGRBM_GFX_INDEX, 0xE0000000);
 	mutex_unlock(&adev->grbm_idx_mutex);
 
 	vega10_didt_set_mask(hwmgr, true);
@@ -1049,17 +1045,15 @@ static int vega10_enable_se_edc_config(struct pp_hwmgr *hwmgr)
 	struct amdgpu_device *adev = hwmgr->adev;
 	int result;
 	uint32_t num_se = 0, count, data;
-	uint32_t reg;
 
 	num_se = adev->gfx.config.max_shader_engines;
 
 	adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
 	mutex_lock(&adev->grbm_idx_mutex);
-	reg = soc15_get_register_offset(GC_HWID, 0, mmGRBM_GFX_INDEX_BASE_IDX, mmGRBM_GFX_INDEX);
 	for (count = 0; count < num_se; count++) {
 		data = GRBM_GFX_INDEX__INSTANCE_BROADCAST_WRITES_MASK | GRBM_GFX_INDEX__SH_BROADCAST_WRITES_MASK | ( count << GRBM_GFX_INDEX__SE_INDEX__SHIFT);
-		cgs_write_register(hwmgr->device, reg, data);
+		WREG32_SOC15(GC, 0, mmGRBM_GFX_INDEX, data);
 		result = vega10_program_didt_config_registers(hwmgr, SEDiDtWeightConfig_Vega10, VEGA10_CONFIGREG_DIDT);
 		result |= vega10_program_didt_config_registers(hwmgr, SEEDCStallPatternConfig_Vega10, VEGA10_CONFIGREG_DIDT);
 		result |= vega10_program_didt_config_registers(hwmgr, SEEDCStallDelayConfig_Vega10, VEGA10_CONFIGREG_DIDT);
@@ -1070,7 +1064,7 @@ static int vega10_enable_se_edc_config(struct pp_hwmgr *hwmgr)
 		if (0 != result)
 			break;
 	}
-	cgs_write_register(hwmgr->device, reg, 0xE0000000);
+	WREG32_SOC15(GC, 0, mmGRBM_GFX_INDEX, 0xE0000000);
 	mutex_unlock(&adev->grbm_idx_mutex);
 
 	vega10_didt_set_mask(hwmgr, true);
@@ -1099,7 +1093,6 @@ static int vega10_enable_psm_gc_edc_config(struct pp_hwmgr *hwmgr)
 	int result;
 	uint32_t num_se = 0;
 	uint32_t count, data;
-	uint32_t reg;
 
 	num_se = adev->gfx.config.max_shader_engines;
 
@@ -1108,10 +1101,9 @@ static int vega10_enable_psm_gc_edc_config(struct pp_hwmgr *hwmgr)
 	vega10_program_gc_didt_config_registers(hwmgr, AvfsPSMResetConfig_vega10);
 
 	mutex_lock(&adev->grbm_idx_mutex);
-	reg = soc15_get_register_offset(GC_HWID, 0, mmGRBM_GFX_INDEX_BASE_IDX, mmGRBM_GFX_INDEX);
 	for (count = 0; count < num_se; count++) {
 		data = GRBM_GFX_INDEX__INSTANCE_BROADCAST_WRITES_MASK | GRBM_GFX_INDEX__SH_BROADCAST_WRITES_MASK | ( count << GRBM_GFX_INDEX__SE_INDEX__SHIFT);
-		cgs_write_register(hwmgr->device, reg, data);
+		WREG32_SOC15(GC, 0, mmGRBM_GFX_INDEX, data);
 		result |= vega10_program_didt_config_registers(hwmgr, PSMSEEDCStallPatternConfig_Vega10, VEGA10_CONFIGREG_DIDT);
 		result |= vega10_program_didt_config_registers(hwmgr, PSMSEEDCStallDelayConfig_Vega10, VEGA10_CONFIGREG_DIDT);
 		result |= vega10_program_didt_config_registers(hwmgr, PSMSEEDCCtrlResetConfig_Vega10, VEGA10_CONFIGREG_DIDT);
@@ -1120,7 +1112,7 @@ static int vega10_enable_psm_gc_edc_config(struct pp_hwmgr *hwmgr)
 		if (0 != result)
 			break;
 	}
-	cgs_write_register(hwmgr->device, reg, 0xE0000000);
+	WREG32_SOC15(GC, 0, mmGRBM_GFX_INDEX, 0xE0000000);
 	mutex_unlock(&adev->grbm_idx_mutex);
 
 	vega10_didt_set_mask(hwmgr, true);
@@ -1165,14 +1157,12 @@ static int vega10_disable_psm_gc_edc_config(struct pp_hwmgr *hwmgr)
 static int vega10_enable_se_edc_force_stall_config(struct pp_hwmgr *hwmgr)
 {
 	struct amdgpu_device *adev = hwmgr->adev;
-	uint32_t reg;
 	int result;
 
 	adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
 	mutex_lock(&adev->grbm_idx_mutex);
-	reg = soc15_get_register_offset(GC_HWID, 0, mmGRBM_GFX_INDEX_BASE_IDX, mmGRBM_GFX_INDEX);
-	cgs_write_register(hwmgr->device, reg, 0xE0000000);
+	WREG32_SOC15(GC, 0, mmGRBM_GFX_INDEX, 0xE0000000);
 	mutex_unlock(&adev->grbm_idx_mutex);
 
 	result = vega10_program_didt_config_registers(hwmgr, SEEDCForceStallPatternConfig_Vega10, VEGA10_CONFIGREG_DIDT);
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_thermal.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_thermal.c
index 9f18226..aa044c1 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_thermal.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_thermal.c
@@ -25,7 +25,7 @@
 #include "vega10_hwmgr.h"
 #include "vega10_ppsmc.h"
 #include "vega10_inc.h"
-#include "pp_soc15.h"
+#include "soc15_common.h"
 #include "pp_debug.h"
 
 static int vega10_get_current_rpm(struct pp_hwmgr *hwmgr, uint32_t *current_rpm)
@@ -89,6 +89,7 @@ int vega10_fan_ctrl_get_fan_speed_percent(struct pp_hwmgr *hwmgr,
 
 int vega10_fan_ctrl_get_fan_speed_rpm(struct pp_hwmgr *hwmgr, uint32_t *speed)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	struct vega10_hwmgr *data = hwmgr->backend;
 	uint32_t tach_period;
 	uint32_t crystal_clock_freq;
@@ -100,10 +101,8 @@ int vega10_fan_ctrl_get_fan_speed_rpm(struct pp_hwmgr *hwmgr, uint32_t *speed)
 	if (data->smu_features[GNLD_FAN_CONTROL].supported) {
 		result = vega10_get_current_rpm(hwmgr, speed);
 	} else {
-		uint32_t reg = soc15_get_register_offset(THM_HWID, 0,
-				mmCG_TACH_STATUS_BASE_IDX, mmCG_TACH_STATUS);
 		tach_period =
-			CGS_REG_GET_FIELD(cgs_read_register(hwmgr->device, reg),
+			REG_GET_FIELD(RREG32_SOC15(THM, 0, mmCG_TACH_STATUS),
 					  CG_TACH_STATUS,
 					  TACH_PERIOD);
 
@@ -127,26 +126,23 @@ int vega10_fan_ctrl_get_fan_speed_rpm(struct pp_hwmgr *hwmgr, uint32_t *speed)
 */
 int vega10_fan_ctrl_set_static_mode(struct pp_hwmgr *hwmgr, uint32_t mode)
 {
-	uint32_t reg;
-
-	reg = soc15_get_register_offset(THM_HWID, 0,
-			mmCG_FDO_CTRL2_BASE_IDX, mmCG_FDO_CTRL2);
+	struct amdgpu_device *adev = hwmgr->adev;
 
 	if (hwmgr->fan_ctrl_is_in_default_mode) {
 		hwmgr->fan_ctrl_default_mode =
-			CGS_REG_GET_FIELD(cgs_read_register(hwmgr->device, reg),
+			REG_GET_FIELD(RREG32_SOC15(THM, 0, mmCG_FDO_CTRL2),
 				CG_FDO_CTRL2, FDO_PWM_MODE);
 		hwmgr->tmin =
-			CGS_REG_GET_FIELD(cgs_read_register(hwmgr->device, reg),
+			REG_GET_FIELD(RREG32_SOC15(THM, 0, mmCG_FDO_CTRL2),
 				CG_FDO_CTRL2, TMIN);
 		hwmgr->fan_ctrl_is_in_default_mode = false;
 	}
 
-	cgs_write_register(hwmgr->device, reg,
-			CGS_REG_SET_FIELD(cgs_read_register(hwmgr->device, reg),
+	WREG32_SOC15(THM, 0, mmCG_FDO_CTRL2,
+			REG_SET_FIELD(RREG32_SOC15(THM, 0, mmCG_FDO_CTRL2),
 				CG_FDO_CTRL2, TMIN, 0));
-	cgs_write_register(hwmgr->device, reg,
-			CGS_REG_SET_FIELD(cgs_read_register(hwmgr->device, reg),
+	WREG32_SOC15(THM, 0, mmCG_FDO_CTRL2,
+			REG_SET_FIELD(RREG32_SOC15(THM, 0, mmCG_FDO_CTRL2),
 				CG_FDO_CTRL2, FDO_PWM_MODE, mode));
 
 	return 0;
@@ -159,18 +155,15 @@ int vega10_fan_ctrl_set_static_mode(struct pp_hwmgr *hwmgr, uint32_t mode)
 */
 int vega10_fan_ctrl_set_default_mode(struct pp_hwmgr *hwmgr)
 {
-	uint32_t reg;
-
-	reg = soc15_get_register_offset(THM_HWID, 0,
-			mmCG_FDO_CTRL2_BASE_IDX, mmCG_FDO_CTRL2);
+	struct amdgpu_device *adev = hwmgr->adev;
 
 	if (!hwmgr->fan_ctrl_is_in_default_mode) {
-		cgs_write_register(hwmgr->device, reg,
-			CGS_REG_SET_FIELD(cgs_read_register(hwmgr->device, reg),
+		WREG32_SOC15(THM, 0, mmCG_FDO_CTRL2,
+			REG_SET_FIELD(RREG32_SOC15(THM, 0, mmCG_FDO_CTRL2),
 				CG_FDO_CTRL2, FDO_PWM_MODE,
 				hwmgr->fan_ctrl_default_mode));
-		cgs_write_register(hwmgr->device, reg,
-			CGS_REG_SET_FIELD(cgs_read_register(hwmgr->device, reg),
+		WREG32_SOC15(THM, 0, mmCG_FDO_CTRL2,
+			REG_SET_FIELD(RREG32_SOC15(THM, 0, mmCG_FDO_CTRL2),
 				CG_FDO_CTRL2, TMIN,
 				hwmgr->tmin << CG_FDO_CTRL2__TMIN__SHIFT));
 		hwmgr->fan_ctrl_is_in_default_mode = true;
@@ -257,10 +250,10 @@ int vega10_fan_ctrl_stop_smc_fan_control(struct pp_hwmgr *hwmgr)
 int vega10_fan_ctrl_set_fan_speed_percent(struct pp_hwmgr *hwmgr,
 		uint32_t speed)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	uint32_t duty100;
 	uint32_t duty;
 	uint64_t tmp64;
-	uint32_t reg;
 
 	if (hwmgr->thermal_controller.fanInfo.bNoFan)
 		return 0;
@@ -271,10 +264,7 @@ int vega10_fan_ctrl_set_fan_speed_percent(struct pp_hwmgr *hwmgr,
 	if (PP_CAP(PHM_PlatformCaps_MicrocodeFanControl))
 		vega10_fan_ctrl_stop_smc_fan_control(hwmgr);
 
-	reg = soc15_get_register_offset(THM_HWID, 0,
-			mmCG_FDO_CTRL1_BASE_IDX, mmCG_FDO_CTRL1);
-
-	duty100 = CGS_REG_GET_FIELD(cgs_read_register(hwmgr->device, reg),
+	duty100 = REG_GET_FIELD(RREG32_SOC15(THM, 0, mmCG_FDO_CTRL1),
 				    CG_FDO_CTRL1, FMAX_DUTY100);
 
 	if (duty100 == 0)
@@ -284,10 +274,8 @@ int vega10_fan_ctrl_set_fan_speed_percent(struct pp_hwmgr *hwmgr,
 	do_div(tmp64, 100);
 	duty = (uint32_t)tmp64;
 
-	reg = soc15_get_register_offset(THM_HWID, 0,
-			mmCG_FDO_CTRL0_BASE_IDX, mmCG_FDO_CTRL0);
-	cgs_write_register(hwmgr->device, reg,
-		CGS_REG_SET_FIELD(cgs_read_register(hwmgr->device, reg),
+	WREG32_SOC15(THM, 0, mmCG_FDO_CTRL0,
+		REG_SET_FIELD(RREG32_SOC15(THM, 0, mmCG_FDO_CTRL0),
 			CG_FDO_CTRL0, FDO_STATIC_DUTY, duty));
 
 	return vega10_fan_ctrl_set_static_mode(hwmgr, FDO_PWM_MODE_STATIC);
@@ -317,10 +305,10 @@ int vega10_fan_ctrl_reset_fan_speed_to_default(struct pp_hwmgr *hwmgr)
 */
 int vega10_fan_ctrl_set_fan_speed_rpm(struct pp_hwmgr *hwmgr, uint32_t speed)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	uint32_t tach_period;
 	uint32_t crystal_clock_freq;
 	int result = 0;
-	uint32_t reg;
 
 	if (hwmgr->thermal_controller.fanInfo.bNoFan ||
 	    (speed < hwmgr->thermal_controller.fanInfo.ulMinRPM) ||
@@ -333,10 +321,8 @@ int vega10_fan_ctrl_set_fan_speed_rpm(struct pp_hwmgr *hwmgr, uint32_t speed)
 	if (!result) {
 		crystal_clock_freq = amdgpu_asic_get_xclk((struct amdgpu_device *)hwmgr->adev);
 		tach_period = 60 * crystal_clock_freq * 10000 / (8 * speed);
-		reg = soc15_get_register_offset(THM_HWID, 0,
-				mmCG_TACH_STATUS_BASE_IDX, mmCG_TACH_STATUS);
-		cgs_write_register(hwmgr->device, reg,
-				CGS_REG_SET_FIELD(cgs_read_register(hwmgr->device, reg),
+		WREG32_SOC15(THM, 0, mmCG_TACH_STATUS,
+				REG_SET_FIELD(RREG32_SOC15(THM, 0, mmCG_TACH_STATUS),
 					CG_TACH_STATUS, TACH_PERIOD,
 					tach_period));
 	}
@@ -350,13 +336,10 @@ int vega10_fan_ctrl_set_fan_speed_rpm(struct pp_hwmgr *hwmgr, uint32_t speed)
 */
 int vega10_thermal_get_temperature(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	int temp;
-	uint32_t reg;
 
-	reg = soc15_get_register_offset(THM_HWID, 0,
-			mmCG_MULT_THERMAL_STATUS_BASE_IDX,  mmCG_MULT_THERMAL_STATUS);
-
-	temp = cgs_read_register(hwmgr->device, reg);
+	temp = RREG32_SOC15(THM, 0, mmCG_MULT_THERMAL_STATUS);
 
 	temp = (temp & CG_MULT_THERMAL_STATUS__CTF_TEMP_MASK) >>
 			CG_MULT_THERMAL_STATUS__CTF_TEMP__SHIFT;
@@ -379,11 +362,12 @@ int vega10_thermal_get_temperature(struct pp_hwmgr *hwmgr)
 static int vega10_thermal_set_temperature_range(struct pp_hwmgr *hwmgr,
 		struct PP_TemperatureRange *range)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	int low = VEGA10_THERMAL_MINIMUM_ALERT_TEMP *
 			PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
 	int high = VEGA10_THERMAL_MAXIMUM_ALERT_TEMP *
 			PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
-	uint32_t val, reg;
+	uint32_t val;
 
 	if (low < range->min)
 		low = range->min;
@@ -393,20 +377,17 @@ static int vega10_thermal_set_temperature_range(struct pp_hwmgr *hwmgr,
 	if (low > high)
 		return -EINVAL;
 
-	reg = soc15_get_register_offset(THM_HWID, 0,
-			mmTHM_THERMAL_INT_CTRL_BASE_IDX, mmTHM_THERMAL_INT_CTRL);
-
-	val = cgs_read_register(hwmgr->device, reg);
+	val = RREG32_SOC15(THM, 0, mmTHM_THERMAL_INT_CTRL);
 
-	val = CGS_REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, MAX_IH_CREDIT, 5);
-	val = CGS_REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, THERM_IH_HW_ENA, 1);
-	val = CGS_REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, DIG_THERM_INTH, (high / PP_TEMPERATURE_UNITS_PER_CENTIGRADES));
-	val = CGS_REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, DIG_THERM_INTL, (low / PP_TEMPERATURE_UNITS_PER_CENTIGRADES));
+	val = REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, MAX_IH_CREDIT, 5);
+	val = REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, THERM_IH_HW_ENA, 1);
+	val = REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, DIG_THERM_INTH, (high / PP_TEMPERATURE_UNITS_PER_CENTIGRADES));
+	val = REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, DIG_THERM_INTL, (low / PP_TEMPERATURE_UNITS_PER_CENTIGRADES));
 	val &= (~THM_THERMAL_INT_CTRL__THERM_TRIGGER_MASK_MASK) &
 			(~THM_THERMAL_INT_CTRL__THERM_INTH_MASK_MASK) &
 			(~THM_THERMAL_INT_CTRL__THERM_INTL_MASK_MASK);
 
-	cgs_write_register(hwmgr->device, reg, val);
+	WREG32_SOC15(THM, 0, mmTHM_THERMAL_INT_CTRL, val);
 
 	return 0;
 }
@@ -418,21 +399,17 @@ static int vega10_thermal_set_temperature_range(struct pp_hwmgr *hwmgr,
 */
 static int vega10_thermal_initialize(struct pp_hwmgr *hwmgr)
 {
-	uint32_t reg;
+	struct amdgpu_device *adev = hwmgr->adev;
 
 	if (hwmgr->thermal_controller.fanInfo.ucTachometerPulsesPerRevolution) {
-		reg = soc15_get_register_offset(THM_HWID, 0,
-				mmCG_TACH_CTRL_BASE_IDX, mmCG_TACH_CTRL);
-		cgs_write_register(hwmgr->device, reg,
-			CGS_REG_SET_FIELD(cgs_read_register(hwmgr->device, reg),
+		WREG32_SOC15(THM, 0, mmCG_TACH_CTRL,
+			REG_SET_FIELD(RREG32_SOC15(THM, 0, mmCG_TACH_CTRL),
 				CG_TACH_CTRL, EDGE_PER_REV,
 				hwmgr->thermal_controller.fanInfo.ucTachometerPulsesPerRevolution - 1));
 	}
 
-	reg = soc15_get_register_offset(THM_HWID, 0,
-			mmCG_FDO_CTRL2_BASE_IDX, mmCG_FDO_CTRL2);
-	cgs_write_register(hwmgr->device, reg,
-		CGS_REG_SET_FIELD(cgs_read_register(hwmgr->device, reg),
+	WREG32_SOC15(THM, 0, mmCG_FDO_CTRL2,
+		REG_SET_FIELD(RREG32_SOC15(THM, 0, mmCG_FDO_CTRL2),
 			CG_FDO_CTRL2, TACH_PWM_RESP_RATE, 0x28));
 
 	return 0;
@@ -445,9 +422,9 @@ static int vega10_thermal_initialize(struct pp_hwmgr *hwmgr)
 */
 static int vega10_thermal_enable_alert(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	struct vega10_hwmgr *data = hwmgr->backend;
 	uint32_t val = 0;
-	uint32_t reg;
 
 	if (data->smu_features[GNLD_FW_CTF].supported) {
 		if (data->smu_features[GNLD_FW_CTF].enabled)
@@ -465,8 +442,7 @@ static int vega10_thermal_enable_alert(struct pp_hwmgr *hwmgr)
 	val |= (1 << THM_THERMAL_INT_ENA__THERM_INTL_CLR__SHIFT);
 	val |= (1 << THM_THERMAL_INT_ENA__THERM_TRIGGER_CLR__SHIFT);
 
-	reg = soc15_get_register_offset(THM_HWID, 0, mmTHM_THERMAL_INT_ENA_BASE_IDX, mmTHM_THERMAL_INT_ENA);
-	cgs_write_register(hwmgr->device, reg, val);
+	WREG32_SOC15(THM, 0, mmTHM_THERMAL_INT_ENA, val);
 
 	return 0;
 }
@@ -477,8 +453,8 @@ static int vega10_thermal_enable_alert(struct pp_hwmgr *hwmgr)
 */
 int vega10_thermal_disable_alert(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	struct vega10_hwmgr *data = hwmgr->backend;
-	uint32_t reg;
 
 	if (data->smu_features[GNLD_FW_CTF].supported) {
 		if (!data->smu_features[GNLD_FW_CTF].enabled)
@@ -493,8 +469,7 @@ int vega10_thermal_disable_alert(struct pp_hwmgr *hwmgr)
 		data->smu_features[GNLD_FW_CTF].enabled = false;
 	}
 
-	reg = soc15_get_register_offset(THM_HWID, 0, mmTHM_THERMAL_INT_ENA_BASE_IDX, mmTHM_THERMAL_INT_ENA);
-	cgs_write_register(hwmgr->device, reg, 0);
+	WREG32_SOC15(THM, 0, mmTHM_THERMAL_INT_ENA, 0);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
index 6a85238..7dca75c 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
@@ -34,7 +34,6 @@
 #include "atomfirmware.h"
 #include "cgs_common.h"
 #include "vega12_inc.h"
-#include "pp_soc15.h"
 #include "pppcielanes.h"
 #include "vega12_hwmgr.h"
 #include "vega12_processpptables.h"
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_thermal.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_thermal.c
index df0fa81..cfd9e6c 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_thermal.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_thermal.c
@@ -26,7 +26,7 @@
 #include "vega12_smumgr.h"
 #include "vega12_ppsmc.h"
 #include "vega12_inc.h"
-#include "pp_soc15.h"
+#include "soc15_common.h"
 #include "pp_debug.h"
 
 static int vega12_get_current_rpm(struct pp_hwmgr *hwmgr, uint32_t *current_rpm)
@@ -147,13 +147,10 @@ int vega12_fan_ctrl_reset_fan_speed_to_default(struct pp_hwmgr *hwmgr)
 */
 int vega12_thermal_get_temperature(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	int temp = 0;
-	uint32_t reg;
 
-	reg = soc15_get_register_offset(THM_HWID, 0,
-			mmCG_MULT_THERMAL_STATUS_BASE_IDX,  mmCG_MULT_THERMAL_STATUS);
-
-	temp = cgs_read_register(hwmgr->device, reg);
+	temp = RREG32_SOC15(THM, 0, mmCG_MULT_THERMAL_STATUS);
 
 	temp = (temp & CG_MULT_THERMAL_STATUS__CTF_TEMP_MASK) >>
 			CG_MULT_THERMAL_STATUS__CTF_TEMP__SHIFT;
@@ -175,11 +172,12 @@ int vega12_thermal_get_temperature(struct pp_hwmgr *hwmgr)
 static int vega12_thermal_set_temperature_range(struct pp_hwmgr *hwmgr,
 		struct PP_TemperatureRange *range)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	int low = VEGA12_THERMAL_MINIMUM_ALERT_TEMP *
 			PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
 	int high = VEGA12_THERMAL_MAXIMUM_ALERT_TEMP *
 			PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
-	uint32_t val, reg;
+	uint32_t val;
 
 	if (low < range->min)
 		low = range->min;
@@ -189,18 +187,15 @@ static int vega12_thermal_set_temperature_range(struct pp_hwmgr *hwmgr,
 	if (low > high)
 		return -EINVAL;
 
-	reg = soc15_get_register_offset(THM_HWID, 0,
-			mmTHM_THERMAL_INT_CTRL_BASE_IDX, mmTHM_THERMAL_INT_CTRL);
-
-	val = cgs_read_register(hwmgr->device, reg);
+	val = RREG32_SOC15(THM, 0, mmTHM_THERMAL_INT_CTRL);
 
-	val = CGS_REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, MAX_IH_CREDIT, 5);
-	val = CGS_REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, THERM_IH_HW_ENA, 1);
-	val = CGS_REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, DIG_THERM_INTH, (high / PP_TEMPERATURE_UNITS_PER_CENTIGRADES));
-	val = CGS_REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, DIG_THERM_INTL, (low / PP_TEMPERATURE_UNITS_PER_CENTIGRADES));
+	val = REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, MAX_IH_CREDIT, 5);
+	val = REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, THERM_IH_HW_ENA, 1);
+	val = REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, DIG_THERM_INTH, (high / PP_TEMPERATURE_UNITS_PER_CENTIGRADES));
+	val = REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, DIG_THERM_INTL, (low / PP_TEMPERATURE_UNITS_PER_CENTIGRADES));
 	val = val & (~THM_THERMAL_INT_CTRL__THERM_TRIGGER_MASK_MASK);
 
-	cgs_write_register(hwmgr->device, reg, val);
+	WREG32_SOC15(THM, 0, mmTHM_THERMAL_INT_CTRL, val);
 
 	return 0;
 }
@@ -212,15 +207,14 @@ static int vega12_thermal_set_temperature_range(struct pp_hwmgr *hwmgr,
 */
 static int vega12_thermal_enable_alert(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	uint32_t val = 0;
-	uint32_t reg;
 
 	val |= (1 << THM_THERMAL_INT_ENA__THERM_INTH_CLR__SHIFT);
 	val |= (1 << THM_THERMAL_INT_ENA__THERM_INTL_CLR__SHIFT);
 	val |= (1 << THM_THERMAL_INT_ENA__THERM_TRIGGER_CLR__SHIFT);
 
-	reg = soc15_get_register_offset(THM_HWID, 0, mmTHM_THERMAL_INT_ENA_BASE_IDX, mmTHM_THERMAL_INT_ENA);
-	cgs_write_register(hwmgr->device, reg, val);
+	WREG32_SOC15(THM, 0, mmTHM_THERMAL_INT_ENA, val);
 
 	return 0;
 }
@@ -231,10 +225,9 @@ static int vega12_thermal_enable_alert(struct pp_hwmgr *hwmgr)
 */
 int vega12_thermal_disable_alert(struct pp_hwmgr *hwmgr)
 {
-	uint32_t reg;
+	struct amdgpu_device *adev = hwmgr->adev;
 
-	reg = soc15_get_register_offset(THM_HWID, 0, mmTHM_THERMAL_INT_ENA_BASE_IDX, mmTHM_THERMAL_INT_ENA);
-	cgs_write_register(hwmgr->device, reg, 0);
+	WREG32_SOC15(THM, 0, mmTHM_THERMAL_INT_ENA, 0);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/amd/powerplay/inc/pp_soc15.h b/drivers/gpu/drm/amd/powerplay/inc/pp_soc15.h
deleted file mode 100644
index 214f370..0000000
--- a/drivers/gpu/drm/amd/powerplay/inc/pp_soc15.h
+++ /dev/null
@@ -1,52 +0,0 @@
-/*
- * Copyright 2016 Advanced Micro Devices, Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- */
-#ifndef PP_SOC15_H
-#define PP_SOC15_H
-
-#include "soc15_hw_ip.h"
-#include "vega10_ip_offset.h"
-
-inline static uint32_t soc15_get_register_offset(
-		uint32_t hw_id,
-		uint32_t inst,
-		uint32_t segment,
-		uint32_t offset)
-{
-	uint32_t reg = 0;
-
-	if (hw_id == THM_HWID)
-		reg = THM_BASE.instance[inst].segment[segment] + offset;
-	else if (hw_id == NBIF_HWID)
-		reg = NBIF_BASE.instance[inst].segment[segment] + offset;
-	else if (hw_id == MP1_HWID)
-		reg = MP1_BASE.instance[inst].segment[segment] + offset;
-	else if (hw_id == DF_HWID)
-		reg = DF_BASE.instance[inst].segment[segment] + offset;
-	else if (hw_id == GC_HWID)
-		reg = GC_BASE.instance[inst].segment[segment] + offset;
-	else if (hw_id == SMUIO_HWID)
-		reg = SMUIO_BASE.instance[inst].segment[segment] + offset;
-	return reg;
-}
-
-#endif
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c
index bc53f2b..9adea72 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c
@@ -23,7 +23,7 @@
 
 #include "smumgr.h"
 #include "smu10_inc.h"
-#include "pp_soc15.h"
+#include "soc15_common.h"
 #include "smu10_smumgr.h"
 #include "ppatomctrl.h"
 #include "rv_ppsmc.h"
@@ -49,48 +49,41 @@
 
 static uint32_t smu10_wait_for_response(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	uint32_t reg;
 
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_90_BASE_IDX, mmMP1_SMN_C2PMSG_90);
+	reg = SOC15_REG_OFFSET(MP1, 0, mmMP1_SMN_C2PMSG_90);
 
 	phm_wait_for_register_unequal(hwmgr, reg,
 			0, MP1_C2PMSG_90__CONTENT_MASK);
 
-	return cgs_read_register(hwmgr->device, reg);
+	return RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90);
 }
 
 static int smu10_send_msg_to_smc_without_waiting(struct pp_hwmgr *hwmgr,
 		uint16_t msg)
 {
-	uint32_t reg;
+	struct amdgpu_device *adev = hwmgr->adev;
 
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_66_BASE_IDX, mmMP1_SMN_C2PMSG_66);
-	cgs_write_register(hwmgr->device, reg, msg);
+	WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_66, msg);
 
 	return 0;
 }
 
 static int smu10_read_arg_from_smc(struct pp_hwmgr *hwmgr)
 {
-	uint32_t reg;
-
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_82_BASE_IDX, mmMP1_SMN_C2PMSG_82);
+	struct amdgpu_device *adev = hwmgr->adev;
 
-	return cgs_read_register(hwmgr->device, reg);
+	return RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_82);
 }
 
 static int smu10_send_msg_to_smc(struct pp_hwmgr *hwmgr, uint16_t msg)
 {
-	uint32_t reg;
+	struct amdgpu_device *adev = hwmgr->adev;
 
 	smu10_wait_for_response(hwmgr);
 
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_90_BASE_IDX, mmMP1_SMN_C2PMSG_90);
-	cgs_write_register(hwmgr->device, reg, 0);
+	WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90, 0);
 
 	smu10_send_msg_to_smc_without_waiting(hwmgr, msg);
 
@@ -104,17 +97,13 @@ static int smu10_send_msg_to_smc(struct pp_hwmgr *hwmgr, uint16_t msg)
 static int smu10_send_msg_to_smc_with_parameter(struct pp_hwmgr *hwmgr,
 		uint16_t msg, uint32_t parameter)
 {
-	uint32_t reg;
+	struct amdgpu_device *adev = hwmgr->adev;
 
 	smu10_wait_for_response(hwmgr);
 
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_90_BASE_IDX, mmMP1_SMN_C2PMSG_90);
-	cgs_write_register(hwmgr->device, reg, 0);
+	WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90, 0);
 
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_82_BASE_IDX, mmMP1_SMN_C2PMSG_82);
-	cgs_write_register(hwmgr->device, reg, parameter);
+	WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_82, parameter);
 
 	smu10_send_msg_to_smc_without_waiting(hwmgr, msg);
 
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c
index 4aafb04..14ac6d1 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c
@@ -23,7 +23,7 @@
 
 #include "smumgr.h"
 #include "vega10_inc.h"
-#include "pp_soc15.h"
+#include "soc15_common.h"
 #include "vega10_smumgr.h"
 #include "vega10_hwmgr.h"
 #include "vega10_ppsmc.h"
@@ -54,18 +54,13 @@
 
 static bool vega10_is_smc_ram_running(struct pp_hwmgr *hwmgr)
 {
-	uint32_t mp1_fw_flags, reg;
-
-	reg = soc15_get_register_offset(NBIF_HWID, 0,
-			mmPCIE_INDEX2_BASE_IDX, mmPCIE_INDEX2);
+	struct amdgpu_device *adev = hwmgr->adev;
+	uint32_t mp1_fw_flags;
 
-	cgs_write_register(hwmgr->device, reg,
+	WREG32_SOC15(NBIF, 0, mmPCIE_INDEX2,
 			(MP1_Public | (smnMP1_FIRMWARE_FLAGS & 0xffffffff)));
 
-	reg = soc15_get_register_offset(NBIF_HWID, 0,
-			mmPCIE_DATA2_BASE_IDX, mmPCIE_DATA2);
-
-	mp1_fw_flags = cgs_read_register(hwmgr->device, reg);
+	mp1_fw_flags = RREG32_SOC15(NBIF, 0, mmPCIE_DATA2);
 
 	if (mp1_fw_flags & MP1_FIRMWARE_FLAGS__INTERRUPTS_ENABLED_MASK)
 		return true;
@@ -81,11 +76,11 @@ static bool vega10_is_smc_ram_running(struct pp_hwmgr *hwmgr)
  */
 static uint32_t vega10_wait_for_response(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	uint32_t reg;
 	uint32_t ret;
 
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_90_BASE_IDX, mmMP1_SMN_C2PMSG_90);
+	reg = SOC15_REG_OFFSET(MP1, 0, mmMP1_SMN_C2PMSG_90);
 
 	ret = phm_wait_for_register_unequal(hwmgr, reg,
 			0, MP1_C2PMSG_90__CONTENT_MASK);
@@ -93,7 +88,7 @@ static uint32_t vega10_wait_for_response(struct pp_hwmgr *hwmgr)
 	if (ret)
 		pr_err("No response from smu\n");
 
-	return cgs_read_register(hwmgr->device, reg);
+	return RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90);
 }
 
 /*
@@ -105,11 +100,9 @@ static uint32_t vega10_wait_for_response(struct pp_hwmgr *hwmgr)
 static int vega10_send_msg_to_smc_without_waiting(struct pp_hwmgr *hwmgr,
 		uint16_t msg)
 {
-	uint32_t reg;
+	struct amdgpu_device *adev = hwmgr->adev;
 
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_66_BASE_IDX, mmMP1_SMN_C2PMSG_66);
-	cgs_write_register(hwmgr->device, reg, msg);
+	WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_66, msg);
 
 	return 0;
 }
@@ -122,14 +115,12 @@ static int vega10_send_msg_to_smc_without_waiting(struct pp_hwmgr *hwmgr,
  */
 static int vega10_send_msg_to_smc(struct pp_hwmgr *hwmgr, uint16_t msg)
 {
-	uint32_t reg;
+	struct amdgpu_device *adev = hwmgr->adev;
 	uint32_t ret;
 
 	vega10_wait_for_response(hwmgr);
 
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_90_BASE_IDX, mmMP1_SMN_C2PMSG_90);
-	cgs_write_register(hwmgr->device, reg, 0);
+	WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90, 0);
 
 	vega10_send_msg_to_smc_without_waiting(hwmgr, msg);
 
@@ -150,18 +141,14 @@ static int vega10_send_msg_to_smc(struct pp_hwmgr *hwmgr, uint16_t msg)
 static int vega10_send_msg_to_smc_with_parameter(struct pp_hwmgr *hwmgr,
 		uint16_t msg, uint32_t parameter)
 {
-	uint32_t reg;
+	struct amdgpu_device *adev = hwmgr->adev;
 	uint32_t ret;
 
 	vega10_wait_for_response(hwmgr);
 
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_90_BASE_IDX, mmMP1_SMN_C2PMSG_90);
-	cgs_write_register(hwmgr->device, reg, 0);
+	WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90, 0);
 
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_82_BASE_IDX, mmMP1_SMN_C2PMSG_82);
-	cgs_write_register(hwmgr->device, reg, parameter);
+	WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_82, parameter);
 
 	vega10_send_msg_to_smc_without_waiting(hwmgr, msg);
 
@@ -174,12 +161,9 @@ static int vega10_send_msg_to_smc_with_parameter(struct pp_hwmgr *hwmgr,
 
 static int vega10_get_argument(struct pp_hwmgr *hwmgr)
 {
-	uint32_t reg;
-
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_82_BASE_IDX, mmMP1_SMN_C2PMSG_82);
+	struct amdgpu_device *adev = hwmgr->adev;
 
-	return cgs_read_register(hwmgr->device, reg);
+	return RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_82);
 }
 
 static int vega10_copy_table_from_smc(struct pp_hwmgr *hwmgr,
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.c
index 651a3f2..7d9b40e 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.c
@@ -23,7 +23,7 @@
 
 #include "smumgr.h"
 #include "vega12_inc.h"
-#include "pp_soc15.h"
+#include "soc15_common.h"
 #include "vega12_smumgr.h"
 #include "vega12_ppsmc.h"
 #include "vega12/smu9_driver_if.h"
@@ -44,18 +44,13 @@
 
 static bool vega12_is_smc_ram_running(struct pp_hwmgr *hwmgr)
 {
-	uint32_t mp1_fw_flags, reg;
+	struct amdgpu_device *adev = hwmgr->adev;
+	uint32_t mp1_fw_flags;
 
-	reg = soc15_get_register_offset(NBIF_HWID, 0,
-			mmPCIE_INDEX2_BASE_IDX, mmPCIE_INDEX2);
-
-	cgs_write_register(hwmgr->device, reg,
+	WREG32_SOC15(NBIF, 0, mmPCIE_INDEX2,
 			(MP1_Public | (smnMP1_FIRMWARE_FLAGS & 0xffffffff)));
 
-	reg = soc15_get_register_offset(NBIF_HWID, 0,
-			mmPCIE_DATA2_BASE_IDX, mmPCIE_DATA2);
-
-	mp1_fw_flags = cgs_read_register(hwmgr->device, reg);
+	mp1_fw_flags = RREG32_SOC15(NBIF, 0, mmPCIE_DATA2);
 
 	if ((mp1_fw_flags & MP1_FIRMWARE_FLAGS__INTERRUPTS_ENABLED_MASK) >>
 				MP1_FIRMWARE_FLAGS__INTERRUPTS_ENABLED__SHIFT)
@@ -72,15 +67,15 @@ static bool vega12_is_smc_ram_running(struct pp_hwmgr *hwmgr)
  */
 static uint32_t vega12_wait_for_response(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	uint32_t reg;
 
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_90_BASE_IDX, mmMP1_SMN_C2PMSG_90);
+	reg = SOC15_REG_OFFSET(MP1, 0, mmMP1_SMN_C2PMSG_90);
 
 	phm_wait_for_register_unequal(hwmgr, reg,
 			0, MP1_C2PMSG_90__CONTENT_MASK);
 
-	return cgs_read_register(hwmgr->device, reg);
+	return RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90);
 }
 
 /*
@@ -92,11 +87,9 @@ static uint32_t vega12_wait_for_response(struct pp_hwmgr *hwmgr)
 int vega12_send_msg_to_smc_without_waiting(struct pp_hwmgr *hwmgr,
 		uint16_t msg)
 {
-	uint32_t reg;
+	struct amdgpu_device *adev = hwmgr->adev;
 
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_66_BASE_IDX, mmMP1_SMN_C2PMSG_66);
-	cgs_write_register(hwmgr->device, reg, msg);
+	WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_66, msg);
 
 	return 0;
 }
@@ -109,13 +102,11 @@ int vega12_send_msg_to_smc_without_waiting(struct pp_hwmgr *hwmgr,
  */
 int vega12_send_msg_to_smc(struct pp_hwmgr *hwmgr, uint16_t msg)
 {
-	uint32_t reg;
+	struct amdgpu_device *adev = hwmgr->adev;
 
 	vega12_wait_for_response(hwmgr);
 
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_90_BASE_IDX, mmMP1_SMN_C2PMSG_90);
-	cgs_write_register(hwmgr->device, reg, 0);
+	WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90, 0);
 
 	vega12_send_msg_to_smc_without_waiting(hwmgr, msg);
 
@@ -135,17 +126,13 @@ int vega12_send_msg_to_smc(struct pp_hwmgr *hwmgr, uint16_t msg)
 int vega12_send_msg_to_smc_with_parameter(struct pp_hwmgr *hwmgr,
 		uint16_t msg, uint32_t parameter)
 {
-	uint32_t reg;
+	struct amdgpu_device *adev = hwmgr->adev;
 
 	vega12_wait_for_response(hwmgr);
 
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_90_BASE_IDX, mmMP1_SMN_C2PMSG_90);
-	cgs_write_register(hwmgr->device, reg, 0);
+	WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90, 0);
 
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_82_BASE_IDX, mmMP1_SMN_C2PMSG_82);
-	cgs_write_register(hwmgr->device, reg, parameter);
+	WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_82, parameter);
 
 	vega12_send_msg_to_smc_without_waiting(hwmgr, msg);
 
@@ -166,11 +153,9 @@ int vega12_send_msg_to_smc_with_parameter(struct pp_hwmgr *hwmgr,
 int vega12_send_msg_to_smc_with_parameter_without_waiting(
 		struct pp_hwmgr *hwmgr, uint16_t msg, uint32_t parameter)
 {
-	uint32_t reg;
+	struct amdgpu_device *adev = hwmgr->adev;
 
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_66_BASE_IDX, mmMP1_SMN_C2PMSG_66);
-	cgs_write_register(hwmgr->device, reg, parameter);
+	WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_66, parameter);
 
 	return vega12_send_msg_to_smc_without_waiting(hwmgr, msg);
 }
@@ -183,12 +168,9 @@ int vega12_send_msg_to_smc_with_parameter_without_waiting(
  */
 int vega12_read_arg_from_smc(struct pp_hwmgr *hwmgr, uint32_t *arg)
 {
-	uint32_t reg;
-
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_82_BASE_IDX, mmMP1_SMN_C2PMSG_82);
+	struct amdgpu_device *adev = hwmgr->adev;
 
-	*arg = cgs_read_register(hwmgr->device, reg);
+	*arg = RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_82);
 
 	return 0;
 }
-- 
cgit v1.1


From c11d8afe10228e4621acfcb8f302255ea8567a1e Mon Sep 17 00:00:00 2001
From: Evan Quan <evan.quan@amd.com>
Date: Tue, 10 Apr 2018 13:05:49 +0800
Subject: drm/amd/pp: fix the wrong readout engine clock in deep sleep

Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c | 13 ++++---------
 drivers/gpu/drm/amd/powerplay/inc/vega10_ppsmc.h   |  1 +
 2 files changed, 5 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
index f6427c8..c90502b 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
@@ -3805,7 +3805,7 @@ static int vega10_read_sensor(struct pp_hwmgr *hwmgr, int idx,
 			      void *value, int *size)
 {
 	struct amdgpu_device *adev = hwmgr->adev;
-	uint32_t sclk_idx, mclk_idx, activity_percent = 0;
+	uint32_t sclk_mhz, mclk_idx, activity_percent = 0;
 	struct vega10_hwmgr *data = hwmgr->backend;
 	struct vega10_dpm_table *dpm_table = &data->dpm_table;
 	int ret = 0;
@@ -3813,14 +3813,9 @@ static int vega10_read_sensor(struct pp_hwmgr *hwmgr, int idx,
 
 	switch (idx) {
 	case AMDGPU_PP_SENSOR_GFX_SCLK:
-		smum_send_msg_to_smc(hwmgr, PPSMC_MSG_GetCurrentGfxclkIndex);
-		sclk_idx = smum_get_argument(hwmgr);
-		if (sclk_idx <  dpm_table->gfx_table.count) {
-			*((uint32_t *)value) = dpm_table->gfx_table.dpm_levels[sclk_idx].value;
-			*size = 4;
-		} else {
-			ret = -EINVAL;
-		}
+		smum_send_msg_to_smc(hwmgr, PPSMC_MSG_GetAverageGfxclkActualFrequency);
+		sclk_mhz = smum_get_argument(hwmgr);
+		*((uint32_t *)value) = sclk_mhz * 100;
 		break;
 	case AMDGPU_PP_SENSOR_GFX_MCLK:
 		smum_send_msg_to_smc(hwmgr, PPSMC_MSG_GetCurrentUclkIndex);
diff --git a/drivers/gpu/drm/amd/powerplay/inc/vega10_ppsmc.h b/drivers/gpu/drm/amd/powerplay/inc/vega10_ppsmc.h
index c3ed737..715b5a1 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/vega10_ppsmc.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/vega10_ppsmc.h
@@ -131,6 +131,7 @@ typedef uint16_t PPSMC_Result;
 #define PPSMC_MSG_RunAcgInOpenLoop               0x5E
 #define PPSMC_MSG_InitializeAcg                  0x5F
 #define PPSMC_MSG_GetCurrPkgPwr                  0x61
+#define PPSMC_MSG_GetAverageGfxclkActualFrequency 0x63
 #define PPSMC_MSG_SetPccThrottleLevel            0x67
 #define PPSMC_MSG_UpdatePkgPwrPidAlpha           0x68
 #define PPSMC_Message_Count                      0x69
-- 
cgit v1.1


From f1018f50d48395b4a189bf8ea9af1e4441209cfd Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Christian=20K=C3=B6nig?= <christian.koenig@amd.com>
Date: Thu, 5 Apr 2018 14:46:41 +0200
Subject: drm/amdgpu: use ctx bytes_moved
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Instead of the global (inaccurate) counter.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 10 +++-------
 1 file changed, 3 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
index d7d7ce1..de69ab1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
@@ -412,7 +412,6 @@ static bool amdgpu_cs_try_evict(struct amdgpu_cs_parser *p,
 		struct amdgpu_bo_list_entry *candidate = p->evictable;
 		struct amdgpu_bo *bo = candidate->robj;
 		struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
-		u64 initial_bytes_moved, bytes_moved;
 		bool update_bytes_moved_vis;
 		uint32_t other;
 
@@ -436,18 +435,15 @@ static bool amdgpu_cs_try_evict(struct amdgpu_cs_parser *p,
 			continue;
 
 		/* Good we can try to move this BO somewhere else */
-		amdgpu_ttm_placement_from_domain(bo, other);
 		update_bytes_moved_vis =
 			adev->gmc.visible_vram_size < adev->gmc.real_vram_size &&
 			bo->tbo.mem.mem_type == TTM_PL_VRAM &&
 			bo->tbo.mem.start < adev->gmc.visible_vram_size >> PAGE_SHIFT;
-		initial_bytes_moved = atomic64_read(&adev->num_bytes_moved);
+		amdgpu_ttm_placement_from_domain(bo, other);
 		r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
-		bytes_moved = atomic64_read(&adev->num_bytes_moved) -
-			initial_bytes_moved;
-		p->bytes_moved += bytes_moved;
+		p->bytes_moved += ctx.bytes_moved;
 		if (update_bytes_moved_vis)
-			p->bytes_moved_vis += bytes_moved;
+			p->bytes_moved_vis += ctx.bytes_moved;
 
 		if (unlikely(r))
 			break;
-- 
cgit v1.1


From 5422a28fe86f9f77480471385e0a416c27a9ca72 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Christian=20K=C3=B6nig?= <christian.koenig@amd.com>
Date: Thu, 5 Apr 2018 16:42:03 +0200
Subject: drm/amdgpu: fix and cleanup cpu visible VRAM handling
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The detection if a BO was placed in CPU visible VRAM was incorrect.

Fix it and merge it with the correct detection in amdgpu_ttm.c

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c     |  6 ++----
 drivers/gpu/drm/amd/amdgpu/amdgpu_object.h | 21 +++++++++++++++++++++
 drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c    | 19 +++----------------
 3 files changed, 26 insertions(+), 20 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
index de69ab1..68af2f8 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
@@ -382,8 +382,7 @@ retry:
 
 	p->bytes_moved += ctx.bytes_moved;
 	if (adev->gmc.visible_vram_size < adev->gmc.real_vram_size &&
-	    bo->tbo.mem.mem_type == TTM_PL_VRAM &&
-	    bo->tbo.mem.start < adev->gmc.visible_vram_size >> PAGE_SHIFT)
+	    amdgpu_bo_in_cpu_visible_vram(bo))
 		p->bytes_moved_vis += ctx.bytes_moved;
 
 	if (unlikely(r == -ENOMEM) && domain != bo->allowed_domains &&
@@ -437,8 +436,7 @@ static bool amdgpu_cs_try_evict(struct amdgpu_cs_parser *p,
 		/* Good we can try to move this BO somewhere else */
 		update_bytes_moved_vis =
 			adev->gmc.visible_vram_size < adev->gmc.real_vram_size &&
-			bo->tbo.mem.mem_type == TTM_PL_VRAM &&
-			bo->tbo.mem.start < adev->gmc.visible_vram_size >> PAGE_SHIFT;
+			amdgpu_bo_in_cpu_visible_vram(bo);
 		amdgpu_ttm_placement_from_domain(bo, other);
 		r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
 		p->bytes_moved += ctx.bytes_moved;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
index 546f77cb7..3bee133 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
@@ -196,6 +196,27 @@ static inline bool amdgpu_bo_gpu_accessible(struct amdgpu_bo *bo)
 }
 
 /**
+ * amdgpu_bo_in_cpu_visible_vram - check if BO is (partly) in visible VRAM
+ */
+static inline bool amdgpu_bo_in_cpu_visible_vram(struct amdgpu_bo *bo)
+{
+	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
+	unsigned fpfn = adev->gmc.visible_vram_size >> PAGE_SHIFT;
+	struct drm_mm_node *node = bo->tbo.mem.mm_node;
+	unsigned long pages_left;
+
+	if (bo->tbo.mem.mem_type != TTM_PL_VRAM)
+		return false;
+
+	for (pages_left = bo->tbo.mem.num_pages; pages_left;
+	     pages_left -= node->size, node++)
+		if (node->start < fpfn)
+			return true;
+
+	return false;
+}
+
+/**
  * amdgpu_bo_explicit_sync - return whether the bo is explicitly synced
  */
 static inline bool amdgpu_bo_explicit_sync(struct amdgpu_bo *bo)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index 205da3f..ab73300 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -223,20 +223,8 @@ static void amdgpu_evict_flags(struct ttm_buffer_object *bo,
 		if (!adev->mman.buffer_funcs_enabled) {
 			amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_CPU);
 		} else if (adev->gmc.visible_vram_size < adev->gmc.real_vram_size &&
-			   !(abo->flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED)) {
-			unsigned fpfn = adev->gmc.visible_vram_size >> PAGE_SHIFT;
-			struct drm_mm_node *node = bo->mem.mm_node;
-			unsigned long pages_left;
-
-			for (pages_left = bo->mem.num_pages;
-			     pages_left;
-			     pages_left -= node->size, node++) {
-				if (node->start < fpfn)
-					break;
-			}
-
-			if (!pages_left)
-				goto gtt;
+			   !(abo->flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED) &&
+			   amdgpu_bo_in_cpu_visible_vram(abo)) {
 
 			/* Try evicting to the CPU inaccessible part of VRAM
 			 * first, but only set GTT as busy placement, so this
@@ -245,12 +233,11 @@ static void amdgpu_evict_flags(struct ttm_buffer_object *bo,
 			 */
 			amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_VRAM |
 							 AMDGPU_GEM_DOMAIN_GTT);
-			abo->placements[0].fpfn = fpfn;
+			abo->placements[0].fpfn = adev->gmc.visible_vram_size >> PAGE_SHIFT;
 			abo->placements[0].lpfn = 0;
 			abo->placement.busy_placement = &abo->placements[1];
 			abo->placement.num_busy_placement = 1;
 		} else {
-gtt:
 			amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_GTT);
 		}
 		break;
-- 
cgit v1.1


From 45a2d58e84e6d28c2d9ae8e68bd815d9a98ad52e Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Tue, 10 Apr 2018 16:08:44 -0400
Subject: drm/amd/display: Fix 64-bit division in hwss_edp_power_control

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
index 52427ae..68a182c 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
@@ -857,7 +857,7 @@ void hwss_edp_power_control(
 					dm_get_elapse_time_in_ns(
 							ctx,
 							current_ts,
-							link->link_trace.time_stamp.edp_poweroff) / 1000000;
+							div64_u64(link->link_trace.time_stamp.edp_poweroff, 1000000));
 			unsigned long long wait_time_ms = 0;
 
 			/* max 500ms from LCDVDD off to on */
-- 
cgit v1.1


From 2fa417324abd635294c298f9f3119743055bf5b9 Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Mon, 9 Apr 2018 14:27:46 -0400
Subject: drm/amd/display: Remove PRE_VEGA flag

We enabled this upstream by default now and no longer need the flag.

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 3 ---
 drivers/gpu/drm/amd/display/Kconfig        | 8 --------
 2 files changed, 11 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index abc3346..62d6505 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -2093,9 +2093,6 @@ bool amdgpu_device_asic_has_dc_support(enum amd_asic_type asic_type)
 	case CHIP_POLARIS12:
 	case CHIP_TONGA:
 	case CHIP_FIJI:
-#if defined(CONFIG_DRM_AMD_DC_PRE_VEGA)
-		return amdgpu_dc != 0;
-#endif
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
 #if defined(CONFIG_DRM_AMD_DC_DCN1_0)
diff --git a/drivers/gpu/drm/amd/display/Kconfig b/drivers/gpu/drm/amd/display/Kconfig
index 5b124a6..d5d4586 100644
--- a/drivers/gpu/drm/amd/display/Kconfig
+++ b/drivers/gpu/drm/amd/display/Kconfig
@@ -9,14 +9,6 @@ config DRM_AMD_DC
 	  support for AMDGPU. This adds required support for Vega and
 	  Raven ASICs.
 
-config DRM_AMD_DC_PRE_VEGA
-	bool "DC support for Polaris and older ASICs"
-	default y
-	help
-	  Choose this option to enable the new DC support for older asics
-	  by default. This includes Polaris, Carrizo, Tonga, Bonaire,
-	  and Hawaii.
-
 config DRM_AMD_DC_FBC
 	bool "AMD FBC - Enable Frame Buffer Compression"
 	depends on DRM_AMD_DC
-- 
cgit v1.1


From db4b37975888cf22e39f2cabc6590167faabaeaa Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Mon, 9 Apr 2018 14:04:56 -0400
Subject: drm/amd/display: Don't spam debug messages

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Leo (Sunpeng) Li <sunpeng.li@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/include/logger_types.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/include/logger_types.h b/drivers/gpu/drm/amd/display/include/logger_types.h
index 4f332e8..b608a08 100644
--- a/drivers/gpu/drm/amd/display/include/logger_types.h
+++ b/drivers/gpu/drm/amd/display/include/logger_types.h
@@ -32,7 +32,7 @@
 
 #define DC_LOG_ERROR(...) DRM_ERROR(__VA_ARGS__)
 #define DC_LOG_WARNING(...) DRM_WARN(__VA_ARGS__)
-#define DC_LOG_DEBUG(...) DRM_INFO(__VA_ARGS__)
+#define DC_LOG_DEBUG(...) DRM_DEBUG_KMS(__VA_ARGS__)
 #define DC_LOG_DC(...) DRM_DEBUG_KMS(__VA_ARGS__)
 #define DC_LOG_DTN(...) DRM_DEBUG_KMS(__VA_ARGS__)
 #define DC_LOG_SURFACE(...) pr_debug("[SURFACE]:"__VA_ARGS__)
-- 
cgit v1.1


From d6014e776ceb5da2d86ce405f692522f2b0370f2 Mon Sep 17 00:00:00 2001
From: Shirish S <shirish.s@amd.com>
Date: Wed, 28 Mar 2018 12:22:22 +0530
Subject: drm/amd/display: remove dummy is_blanked() to optimise boot time

is_blanked() hook is a dummy one for underlay pipe, hence
when called, it loops for ~300ms at boot.

This patch removes this dummy call and adds missing checks.

Signed-off-by: Shirish S <shirish.s@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c             | 3 +++
 drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.c                    | 3 ++-
 drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator_v.c | 7 -------
 3 files changed, 5 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c
index ebc96b7..481f692 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c
@@ -230,6 +230,9 @@ bool hwss_wait_for_blank_complete(
 {
 	int counter;
 
+	/* Not applicable if the pipe is not primary, save 300ms of boot time */
+	if (!tg->funcs->is_blanked)
+		return true;
 	for (counter = 0; counter < 100; counter++) {
 		if (tg->funcs->is_blanked(tg))
 			break;
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.c b/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.c
index 4877243..0275d6d 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.c
@@ -53,7 +53,8 @@ void dce_pipe_control_lock(struct dc *dc,
 	struct dce_hwseq *hws = dc->hwseq;
 
 	/* Not lock pipe when blank */
-	if (lock && pipe->stream_res.tg->funcs->is_blanked(pipe->stream_res.tg))
+	if (lock && pipe->stream_res.tg->funcs->is_blanked &&
+	    pipe->stream_res.tg->funcs->is_blanked(pipe->stream_res.tg))
 		return;
 
 	val = REG_GET_4(BLND_V_UPDATE_LOCK[pipe->stream_res.tg->inst],
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator_v.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator_v.c
index 8ad0481..a3cef60 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator_v.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator_v.c
@@ -648,12 +648,6 @@ static void dce110_timing_generator_v_disable_vga(
 	return;
 }
 
-static bool dce110_tg_v_is_blanked(struct timing_generator *tg)
-{
-	/* Signal comes from the primary pipe, underlay is never blanked. */
-	return false;
-}
-
 /** ********************************************************************************************
  *
  * DCE11 Timing Generator Constructor / Destructor
@@ -670,7 +664,6 @@ static const struct timing_generator_funcs dce110_tg_v_funcs = {
 		.set_early_control = dce110_timing_generator_v_set_early_control,
 		.wait_for_state = dce110_timing_generator_v_wait_for_state,
 		.set_blank = dce110_timing_generator_v_set_blank,
-		.is_blanked = dce110_tg_v_is_blanked,
 		.set_colors = dce110_timing_generator_v_set_colors,
 		.set_overscan_blank_color =
 				dce110_timing_generator_v_set_overscan_color_black,
-- 
cgit v1.1


From 45313e5f1c72962a21df58af52a421e4c076b2d4 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Tue, 10 Apr 2018 10:58:43 +0800
Subject: drm/amd/pp: Move same macro definitions to hwmgr.h

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c        | 4 ----
 drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c      | 4 ----
 drivers/gpu/drm/amd/powerplay/inc/hwmgr.h               | 2 ++
 drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c        | 3 ---
 drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c      | 3 ---
 drivers/gpu/drm/amd/powerplay/smumgr/iceland_smumgr.c   | 3 ---
 drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c | 2 --
 drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c     | 2 --
 drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c     | 4 ----
 drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c    | 2 --
 10 files changed, 2 insertions(+), 27 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index 1433215..21c021b 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -61,10 +61,6 @@
 #define SMC_CG_IND_START            0xc0030000
 #define SMC_CG_IND_END              0xc0040000
 
-#define VOLTAGE_SCALE               4
-#define VOLTAGE_VID_OFFSET_SCALE1   625
-#define VOLTAGE_VID_OFFSET_SCALE2   100
-
 #define MEM_FREQ_LOW_LATENCY        25000
 #define MEM_FREQ_HIGH_LATENCY       80000
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
index c90502b..26c5602 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
@@ -51,10 +51,6 @@
 #include "smuio/smuio_9_0_offset.h"
 #include "smuio/smuio_9_0_sh_mask.h"
 
-#define VOLTAGE_SCALE  4
-#define VOLTAGE_VID_OFFSET_SCALE1   625
-#define VOLTAGE_VID_OFFSET_SCALE2   100
-
 #define HBM_MEMORY_CHANNEL_WIDTH    128
 
 static const uint32_t channel_number[] = {1, 2, 0, 4, 0, 8, 0, 16, 2};
diff --git a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
index e450ec7..9b3dd7d 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
@@ -38,6 +38,8 @@ struct phm_fan_speed_info;
 struct pp_atomctrl_voltage_table;
 
 #define VOLTAGE_SCALE 4
+#define VOLTAGE_VID_OFFSET_SCALE1   625
+#define VOLTAGE_VID_OFFSET_SCALE2   100
 
 enum DISPLAY_GAP {
 	DISPLAY_GAP_VBLANK_OR_WM = 0,   /* Wait for vblank or MCHG watermark. */
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c
index c28b95f..2a93f3a 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c
@@ -61,9 +61,6 @@
 
 #define SMC_RAM_END 0x40000
 
-#define VOLTAGE_SCALE               4
-#define VOLTAGE_VID_OFFSET_SCALE1    625
-#define VOLTAGE_VID_OFFSET_SCALE2    100
 #define CISLAND_MINIMUM_ENGINE_CLOCK 800
 #define CISLAND_MAX_DEEPSLEEP_DIVIDER_ID 5
 
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c
index dae3422..53df940 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c
@@ -53,10 +53,7 @@
 
 #define FIJI_SMC_SIZE 0x20000
 
-#define VOLTAGE_SCALE 4
 #define POWERTUNE_DEFAULT_SET_MAX    1
-#define VOLTAGE_VID_OFFSET_SCALE1   625
-#define VOLTAGE_VID_OFFSET_SCALE2   100
 #define VDDC_VDDCI_DELTA            300
 #define MC_CG_ARB_FREQ_F1           0x0b
 
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/iceland_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/iceland_smumgr.c
index bc05e35..415f691 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/iceland_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/iceland_smumgr.c
@@ -60,10 +60,7 @@
 
 #define ICELAND_SMC_SIZE               0x20000
 
-#define VOLTAGE_SCALE 4
 #define POWERTUNE_DEFAULT_SET_MAX    1
-#define VOLTAGE_VID_OFFSET_SCALE1   625
-#define VOLTAGE_VID_OFFSET_SCALE2   100
 #define MC_CG_ARB_FREQ_F1           0x0b
 #define VDDC_VDDCI_DELTA            200
 
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c
index d919228..a8c6524 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c
@@ -52,8 +52,6 @@
 #include "dce/dce_10_0_sh_mask.h"
 
 #define POLARIS10_SMC_SIZE 0x20000
-#define VOLTAGE_VID_OFFSET_SCALE1   625
-#define VOLTAGE_VID_OFFSET_SCALE2   100
 #define POWERTUNE_DEFAULT_SET_MAX    1
 #define VDDC_VDDCI_DELTA            200
 #define MC_CG_ARB_FREQ_F1           0x0b
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c
index 9adea72..0a563f6 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c
@@ -33,8 +33,6 @@
 #include "pp_debug.h"
 
 
-#define VOLTAGE_SCALE 4
-
 #define BUFFER_SIZE                 80000
 #define MAX_STRING_SIZE             15
 #define BUFFER_SIZETWO              131072
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c
index 94ba304..782b19f 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c
@@ -55,11 +55,7 @@
 #include "dce/dce_10_0_d.h"
 #include "dce/dce_10_0_sh_mask.h"
 
-
-#define VOLTAGE_SCALE 4
 #define POWERTUNE_DEFAULT_SET_MAX    1
-#define VOLTAGE_VID_OFFSET_SCALE1   625
-#define VOLTAGE_VID_OFFSET_SCALE2   100
 #define MC_CG_ARB_FREQ_F1           0x0b
 #define VDDC_VDDCI_DELTA            200
 
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c
index 14ac6d1..e84669c 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c
@@ -35,8 +35,6 @@
 #define AVFS_EN_MSB		1568
 #define AVFS_EN_LSB		1568
 
-#define VOLTAGE_SCALE	4
-
 /* Microcode file is stored in this buffer */
 #define BUFFER_SIZE                 80000
 #define MAX_STRING_SIZE             15
-- 
cgit v1.1


From 29ae1118d85e8435b12fca512410dbd39920cce9 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Tue, 10 Apr 2018 10:58:43 +0800
Subject: drm/amd/pp: Remove unnecessary forward declaration

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c | 84 +++++++++++-----------
 1 file changed, 41 insertions(+), 43 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
index 26c5602..127c550 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
@@ -75,8 +75,6 @@ static const uint32_t channel_number[] = {1, 2, 0, 4, 0, 8, 0, 16, 2};
 #define DF_CS_AON0_DramBaseAddress0__IntLvNumChan_MASK                                                        0x000000F0L
 #define DF_CS_AON0_DramBaseAddress0__IntLvAddrSel_MASK                                                        0x00000700L
 #define DF_CS_AON0_DramBaseAddress0__DramBaseAddr_MASK                                                        0xFFFFF000L
-static int vega10_force_clock_level(struct pp_hwmgr *hwmgr,
-		enum pp_clock_type type, uint32_t mask);
 
 static const ULONG PhwVega10_Magic = (ULONG)(PHM_VIslands_Magic);
 
@@ -4095,6 +4093,47 @@ static void vega10_set_fan_control_mode(struct pp_hwmgr *hwmgr, uint32_t mode)
 	}
 }
 
+static int vega10_force_clock_level(struct pp_hwmgr *hwmgr,
+		enum pp_clock_type type, uint32_t mask)
+{
+	struct vega10_hwmgr *data = hwmgr->backend;
+
+	switch (type) {
+	case PP_SCLK:
+		data->smc_state_table.gfx_boot_level = mask ? (ffs(mask) - 1) : 0;
+		data->smc_state_table.gfx_max_level = mask ? (fls(mask) - 1) : 0;
+
+		PP_ASSERT_WITH_CODE(!vega10_upload_dpm_bootup_level(hwmgr),
+			"Failed to upload boot level to lowest!",
+			return -EINVAL);
+
+		PP_ASSERT_WITH_CODE(!vega10_upload_dpm_max_level(hwmgr),
+			"Failed to upload dpm max level to highest!",
+			return -EINVAL);
+		break;
+
+	case PP_MCLK:
+		data->smc_state_table.mem_boot_level = mask ? (ffs(mask) - 1) : 0;
+		data->smc_state_table.mem_max_level = mask ? (fls(mask) - 1) : 0;
+
+		PP_ASSERT_WITH_CODE(!vega10_upload_dpm_bootup_level(hwmgr),
+			"Failed to upload boot level to lowest!",
+			return -EINVAL);
+
+		PP_ASSERT_WITH_CODE(!vega10_upload_dpm_max_level(hwmgr),
+			"Failed to upload dpm max level to highest!",
+			return -EINVAL);
+
+		break;
+
+	case PP_PCIE:
+	default:
+		break;
+	}
+
+	return 0;
+}
+
 static int vega10_dpm_force_dpm_level(struct pp_hwmgr *hwmgr,
 				enum amd_dpm_forced_level level)
 {
@@ -4381,47 +4420,6 @@ static int vega10_set_watermarks_for_clocks_ranges(struct pp_hwmgr *hwmgr,
 	return result;
 }
 
-static int vega10_force_clock_level(struct pp_hwmgr *hwmgr,
-		enum pp_clock_type type, uint32_t mask)
-{
-	struct vega10_hwmgr *data = hwmgr->backend;
-
-	switch (type) {
-	case PP_SCLK:
-		data->smc_state_table.gfx_boot_level = mask ? (ffs(mask) - 1) : 0;
-		data->smc_state_table.gfx_max_level = mask ? (fls(mask) - 1) : 0;
-
-		PP_ASSERT_WITH_CODE(!vega10_upload_dpm_bootup_level(hwmgr),
-			"Failed to upload boot level to lowest!",
-			return -EINVAL);
-
-		PP_ASSERT_WITH_CODE(!vega10_upload_dpm_max_level(hwmgr),
-			"Failed to upload dpm max level to highest!",
-			return -EINVAL);
-		break;
-
-	case PP_MCLK:
-		data->smc_state_table.mem_boot_level = mask ? (ffs(mask) - 1) : 0;
-		data->smc_state_table.mem_max_level = mask ? (fls(mask) - 1) : 0;
-
-		PP_ASSERT_WITH_CODE(!vega10_upload_dpm_bootup_level(hwmgr),
-			"Failed to upload boot level to lowest!",
-			return -EINVAL);
-
-		PP_ASSERT_WITH_CODE(!vega10_upload_dpm_max_level(hwmgr),
-			"Failed to upload dpm max level to highest!",
-			return -EINVAL);
-
-		break;
-
-	case PP_PCIE:
-	default:
-		break;
-	}
-
-	return 0;
-}
-
 static int vega10_print_clock_levels(struct pp_hwmgr *hwmgr,
 		enum pp_clock_type type, char *buf)
 {
-- 
cgit v1.1


From 819a23f83e3b2513cffbef418458a47ca02c36b3 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Tue, 10 Apr 2018 17:17:22 +0800
Subject: drm/amdgpu: Add APU support in vi_set_uvd_clocks
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

fix the issue set uvd clock failed on CZ/ST
which lead 1s delay when boot up.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Acked-by: Shirish S <shirish.s@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
---
 drivers/gpu/drm/amd/amdgpu/vi.c | 46 +++++++++++++++++++++++++++++++----------
 1 file changed, 35 insertions(+), 11 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c
index 1b4ee24..51acd7c 100644
--- a/drivers/gpu/drm/amd/amdgpu/vi.c
+++ b/drivers/gpu/drm/amd/amdgpu/vi.c
@@ -728,33 +728,57 @@ static int vi_set_uvd_clock(struct amdgpu_device *adev, u32 clock,
 		return r;
 
 	tmp = RREG32_SMC(cntl_reg);
-	tmp &= ~(CG_DCLK_CNTL__DCLK_DIR_CNTL_EN_MASK |
-		CG_DCLK_CNTL__DCLK_DIVIDER_MASK);
+
+	if (adev->flags & AMD_IS_APU)
+		tmp &= ~CG_DCLK_CNTL__DCLK_DIVIDER_MASK;
+	else
+		tmp &= ~(CG_DCLK_CNTL__DCLK_DIR_CNTL_EN_MASK |
+				CG_DCLK_CNTL__DCLK_DIVIDER_MASK);
 	tmp |= dividers.post_divider;
 	WREG32_SMC(cntl_reg, tmp);
 
 	for (i = 0; i < 100; i++) {
-		if (RREG32_SMC(status_reg) & CG_DCLK_STATUS__DCLK_STATUS_MASK)
-			break;
+		tmp = RREG32_SMC(status_reg);
+		if (adev->flags & AMD_IS_APU) {
+			if (tmp & 0x10000)
+				break;
+		} else {
+			if (tmp & CG_DCLK_STATUS__DCLK_STATUS_MASK)
+				break;
+		}
 		mdelay(10);
 	}
 	if (i == 100)
 		return -ETIMEDOUT;
-
 	return 0;
 }
 
+#define ixGNB_CLK1_DFS_CNTL 0xD82200F0
+#define ixGNB_CLK1_STATUS   0xD822010C
+#define ixGNB_CLK2_DFS_CNTL 0xD8220110
+#define ixGNB_CLK2_STATUS   0xD822012C
+
 static int vi_set_uvd_clocks(struct amdgpu_device *adev, u32 vclk, u32 dclk)
 {
 	int r;
 
-	r = vi_set_uvd_clock(adev, vclk, ixCG_VCLK_CNTL, ixCG_VCLK_STATUS);
-	if (r)
-		return r;
+	if (adev->flags & AMD_IS_APU) {
+		r = vi_set_uvd_clock(adev, vclk, ixGNB_CLK2_DFS_CNTL, ixGNB_CLK2_STATUS);
+		if (r)
+			return r;
 
-	r = vi_set_uvd_clock(adev, dclk, ixCG_DCLK_CNTL, ixCG_DCLK_STATUS);
-	if (r)
-		return r;
+		r = vi_set_uvd_clock(adev, dclk, ixGNB_CLK1_DFS_CNTL, ixGNB_CLK1_STATUS);
+		if (r)
+			return r;
+	} else {
+		r = vi_set_uvd_clock(adev, vclk, ixCG_VCLK_CNTL, ixCG_VCLK_STATUS);
+		if (r)
+			return r;
+
+		r = vi_set_uvd_clock(adev, dclk, ixCG_DCLK_CNTL, ixCG_DCLK_STATUS);
+		if (r)
+			return r;
+	}
 
 	return 0;
 }
-- 
cgit v1.1


From 08ebb6e9f4fd7098c28e0ebbb42847cf0488ebb8 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Tue, 10 Apr 2018 17:49:56 +0800
Subject: drm/amdgpu: Add APU support in vi_set_vce_clocks
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

1. fix set vce clocks failed on Cz/St
   which lead 1s delay when boot up.
2. remove the workaround in vce_v3_0.c

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Acked-by: Shirish S <shirish.s@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
---
 drivers/gpu/drm/amd/amdgpu/vce_v3_0.c |  4 ++--
 drivers/gpu/drm/amd/amdgpu/vi.c       | 31 +++++++++++++++++++++++++------
 2 files changed, 27 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
index 428d192..ac96172 100644
--- a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
@@ -467,8 +467,8 @@ static int vce_v3_0_hw_init(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
 	vce_v3_0_override_vce_clock_gating(adev, true);
-	if (!(adev->flags & AMD_IS_APU))
-		amdgpu_asic_set_vce_clocks(adev, 10000, 10000);
+
+	amdgpu_asic_set_vce_clocks(adev, 10000, 10000);
 
 	for (i = 0; i < adev->vce.num_rings; i++)
 		adev->vce.ring[i].ready = false;
diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c
index 51acd7c..4034a28 100644
--- a/drivers/gpu/drm/amd/amdgpu/vi.c
+++ b/drivers/gpu/drm/amd/amdgpu/vi.c
@@ -757,6 +757,8 @@ static int vi_set_uvd_clock(struct amdgpu_device *adev, u32 clock,
 #define ixGNB_CLK1_STATUS   0xD822010C
 #define ixGNB_CLK2_DFS_CNTL 0xD8220110
 #define ixGNB_CLK2_STATUS   0xD822012C
+#define ixGNB_CLK3_DFS_CNTL 0xD8220130
+#define ixGNB_CLK3_STATUS   0xD822014C
 
 static int vi_set_uvd_clocks(struct amdgpu_device *adev, u32 vclk, u32 dclk)
 {
@@ -788,6 +790,22 @@ static int vi_set_vce_clocks(struct amdgpu_device *adev, u32 evclk, u32 ecclk)
 	int r, i;
 	struct atom_clock_dividers dividers;
 	u32 tmp;
+	u32 reg_ctrl;
+	u32 reg_status;
+	u32 status_mask;
+	u32 reg_mask;
+
+	if (adev->flags & AMD_IS_APU) {
+		reg_ctrl = ixGNB_CLK3_DFS_CNTL;
+		reg_status = ixGNB_CLK3_STATUS;
+		status_mask = 0x00010000;
+		reg_mask = CG_ECLK_CNTL__ECLK_DIVIDER_MASK;
+	} else {
+		reg_ctrl = ixCG_ECLK_CNTL;
+		reg_status = ixCG_ECLK_STATUS;
+		status_mask = CG_ECLK_STATUS__ECLK_STATUS_MASK;
+		reg_mask = CG_ECLK_CNTL__ECLK_DIR_CNTL_EN_MASK | CG_ECLK_CNTL__ECLK_DIVIDER_MASK;
+	}
 
 	r = amdgpu_atombios_get_clock_dividers(adev,
 					       COMPUTE_GPUCLK_INPUT_FLAG_DEFAULT_GPUCLK,
@@ -796,24 +814,25 @@ static int vi_set_vce_clocks(struct amdgpu_device *adev, u32 evclk, u32 ecclk)
 		return r;
 
 	for (i = 0; i < 100; i++) {
-		if (RREG32_SMC(ixCG_ECLK_STATUS) & CG_ECLK_STATUS__ECLK_STATUS_MASK)
+		if (RREG32_SMC(reg_status) & status_mask)
 			break;
 		mdelay(10);
 	}
+
 	if (i == 100)
 		return -ETIMEDOUT;
 
-	tmp = RREG32_SMC(ixCG_ECLK_CNTL);
-	tmp &= ~(CG_ECLK_CNTL__ECLK_DIR_CNTL_EN_MASK |
-		CG_ECLK_CNTL__ECLK_DIVIDER_MASK);
+	tmp = RREG32_SMC(reg_ctrl);
+	tmp &= ~reg_mask;
 	tmp |= dividers.post_divider;
-	WREG32_SMC(ixCG_ECLK_CNTL, tmp);
+	WREG32_SMC(reg_ctrl, tmp);
 
 	for (i = 0; i < 100; i++) {
-		if (RREG32_SMC(ixCG_ECLK_STATUS) & CG_ECLK_STATUS__ECLK_STATUS_MASK)
+		if (RREG32_SMC(reg_status) & status_mask)
 			break;
 		mdelay(10);
 	}
+
 	if (i == 100)
 		return -ETIMEDOUT;
 
-- 
cgit v1.1


From 61279073b1d35ea29bf546c7751bda09610ab5ef Mon Sep 17 00:00:00 2001
From: Kenneth Feng <kenneth.feng@amd.com>
Date: Mon, 9 Apr 2018 14:53:51 +0800
Subject: amd/powerplay: implement the vega12_force_clock_level interface

pp_dpm_sclk/pp_dpm_mclk in sysfs implemented to force
gfxclk/uclk dpm level for Vega12

Signed-off-by: Kenneth Feng <kenneth.feng@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c | 42 +++++++++++++++++++++-
 1 file changed, 41 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
index 7dca75c..df234db 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
@@ -991,15 +991,55 @@ static uint32_t vega12_find_highest_dpm_level(
 
 static int vega12_upload_dpm_min_level(struct pp_hwmgr *hwmgr)
 {
+	struct vega12_hwmgr *data = hwmgr->backend;
+	if (data->smc_state_table.gfx_boot_level !=
+			data->dpm_table.gfx_table.dpm_state.soft_min_level) {
+		smum_send_msg_to_smc_with_parameter(hwmgr,
+			PPSMC_MSG_SetSoftMinByFreq,
+			PPCLK_GFXCLK<<16 | data->dpm_table.gfx_table.dpm_levels[data->smc_state_table.gfx_boot_level].value);
+		data->dpm_table.gfx_table.dpm_state.soft_min_level =
+				data->smc_state_table.gfx_boot_level;
+	}
+
+	if (data->smc_state_table.mem_boot_level !=
+			data->dpm_table.mem_table.dpm_state.soft_min_level) {
+		smum_send_msg_to_smc_with_parameter(hwmgr,
+			PPSMC_MSG_SetSoftMinByFreq,
+			PPCLK_UCLK<<16 | data->dpm_table.mem_table.dpm_levels[data->smc_state_table.mem_boot_level].value);
+		data->dpm_table.mem_table.dpm_state.soft_min_level =
+				data->smc_state_table.mem_boot_level;
+	}
+
 	return 0;
+
 }
 
 static int vega12_upload_dpm_max_level(struct pp_hwmgr *hwmgr)
 {
+	struct vega12_hwmgr *data = hwmgr->backend;
+	if (data->smc_state_table.gfx_max_level !=
+		data->dpm_table.gfx_table.dpm_state.soft_max_level) {
+		smum_send_msg_to_smc_with_parameter(hwmgr,
+			PPSMC_MSG_SetSoftMaxByFreq,
+			/* plus the vale by 1 to align the resolution */
+			PPCLK_GFXCLK<<16 | (data->dpm_table.gfx_table.dpm_levels[data->smc_state_table.gfx_max_level].value + 1));
+		data->dpm_table.gfx_table.dpm_state.soft_max_level =
+				data->smc_state_table.gfx_max_level;
+	}
+
+	if (data->smc_state_table.mem_max_level !=
+		data->dpm_table.mem_table.dpm_state.soft_max_level) {
+		smum_send_msg_to_smc_with_parameter(hwmgr,
+			PPSMC_MSG_SetSoftMaxByFreq,
+			/* plus the vale by 1 to align the resolution */
+			PPCLK_UCLK<<16 | (data->dpm_table.mem_table.dpm_levels[data->smc_state_table.mem_max_level].value + 1));
+		data->dpm_table.mem_table.dpm_state.soft_max_level =
+				data->smc_state_table.mem_max_level;
+	}
+
 	return 0;
 }
 
-
 int vega12_enable_disable_vce_dpm(struct pp_hwmgr *hwmgr, bool enable)
 {
 	struct vega12_hwmgr *data =
-- 
cgit v1.1


From 564be2fc2b1ddb6cbef2bd77f83e91c9e4a1063f Mon Sep 17 00:00:00 2001
From: Kenneth Feng <kenneth.feng@amd.com>
Date: Wed, 4 Apr 2018 15:17:22 +0800
Subject: drm/amd/powerplay: Get more than 8 level gfxclk states

To apply on Vega12 for more than 8 gfx dpm levels

Signed-off-by: Kenneth Feng <kenneth.feng@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.h
index bc98b1d..e81ded1 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.h
@@ -33,7 +33,7 @@
 #define WaterMarksExist  1
 #define WaterMarksLoaded 2
 
-#define VG12_PSUEDO_NUM_GFXCLK_DPM_LEVELS   8
+#define VG12_PSUEDO_NUM_GFXCLK_DPM_LEVELS   16
 #define VG12_PSUEDO_NUM_SOCCLK_DPM_LEVELS   8
 #define VG12_PSUEDO_NUM_DCEFCLK_DPM_LEVELS  8
 #define VG12_PSUEDO_NUM_UCLK_DPM_LEVELS     4
-- 
cgit v1.1


From 0bc8f3d29b188b273e92cd895da3b5c31e86434f Mon Sep 17 00:00:00 2001
From: Kenneth Feng <kenneth.feng@amd.com>
Date: Tue, 10 Apr 2018 17:05:36 +0800
Subject: drm/amd/powerplay: initialzie the dpm intial enabled state

To expose the right dpm levels to the sysfs

Signed-off-by: Kenneth Feng <kenneth.feng@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
index df234db..3e1ed0a 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
@@ -545,6 +545,7 @@ static int vega12_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
 			return -EINVAL);
 
 		dpm_table->dpm_levels[i].value = clock;
+		dpm_table->dpm_levels[i].enabled = true;
 	}
 
 	vega12_init_dpm_state(&(dpm_table->dpm_state));
@@ -564,6 +565,7 @@ static int vega12_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
 			return -EINVAL);
 
 		dpm_table->dpm_levels[i].value = clock;
+		dpm_table->dpm_levels[i].enabled = true;
 	}
 
 	vega12_init_dpm_state(&(dpm_table->dpm_state));
@@ -584,6 +586,7 @@ static int vega12_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
 			return -EINVAL);
 
 		dpm_table->dpm_levels[i].value = clock;
+		dpm_table->dpm_levels[i].enabled = true;
 	}
 
 	vega12_init_dpm_state(&(dpm_table->dpm_state));
@@ -604,6 +607,7 @@ static int vega12_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
 		return -EINVAL);
 
 		dpm_table->dpm_levels[i].value = clock;
+		dpm_table->dpm_levels[i].enabled = true;
 	}
 
 	vega12_init_dpm_state(&(dpm_table->dpm_state));
@@ -624,6 +628,7 @@ static int vega12_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
 			return -EINVAL);
 
 		dpm_table->dpm_levels[i].value = clock;
+		dpm_table->dpm_levels[i].enabled = true;
 	}
 
 	vega12_init_dpm_state(&(dpm_table->dpm_state));
@@ -644,6 +649,7 @@ static int vega12_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
 		return -EINVAL);
 
 		dpm_table->dpm_levels[i].value = clock;
+		dpm_table->dpm_levels[i].enabled = true;
 	}
 
 	vega12_init_dpm_state(&(dpm_table->dpm_state));
@@ -665,6 +671,7 @@ static int vega12_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
 			return -EINVAL);
 
 		dpm_table->dpm_levels[i].value = clock;
+		dpm_table->dpm_levels[i].enabled = true;
 	}
 
 	vega12_init_dpm_state(&(dpm_table->dpm_state));
@@ -685,6 +692,7 @@ static int vega12_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
 			return -EINVAL);
 
 		dpm_table->dpm_levels[i].value = clock;
+		dpm_table->dpm_levels[i].enabled = true;
 	}
 
 	vega12_init_dpm_state(&(dpm_table->dpm_state));
@@ -705,6 +713,7 @@ static int vega12_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
 			return -EINVAL);
 
 		dpm_table->dpm_levels[i].value = clock;
+		dpm_table->dpm_levels[i].enabled = true;
 	}
 
 	vega12_init_dpm_state(&(dpm_table->dpm_state));
@@ -725,6 +734,7 @@ static int vega12_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
 			return -EINVAL);
 
 		dpm_table->dpm_levels[i].value = clock;
+		dpm_table->dpm_levels[i].enabled = true;
 	}
 
 	vega12_init_dpm_state(&(dpm_table->dpm_state));
-- 
cgit v1.1


From 18081c2003915dadc3507b79cf6453f997948ded Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Wed, 4 Apr 2018 18:33:15 +0800
Subject: drm/amd/pp: Remove dead function in smu7_smumgr.c

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c | 10 ----------
 drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.h |  1 -
 2 files changed, 11 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
index 41fab2d..8b9518a 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
@@ -231,16 +231,6 @@ int smu7_send_msg_to_smc_offset(struct pp_hwmgr *hwmgr)
 	return 0;
 }
 
-int smu7_wait_for_smc_inactive(struct pp_hwmgr *hwmgr)
-{
-	if (!smu7_is_smc_ram_running(hwmgr))
-		return -EINVAL;
-
-	PHM_WAIT_VFPF_INDIRECT_FIELD(hwmgr, SMC_IND, SMC_SYSCON_CLOCK_CNTL_0, cken, 0);
-	return 0;
-}
-
-
 enum cgs_ucode_id smu7_convert_fw_type_to_cgs(uint32_t fw_type)
 {
 	enum cgs_ucode_id result = CGS_UCODE_ID_MAXIMUM;
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.h b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.h
index 126d300..39c9bfd 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.h
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.h
@@ -67,7 +67,6 @@ int smu7_send_msg_to_smc_with_parameter(struct pp_hwmgr *hwmgr, uint16_t msg,
 int smu7_send_msg_to_smc_with_parameter_without_waiting(struct pp_hwmgr *hwmgr,
 						uint16_t msg, uint32_t parameter);
 int smu7_send_msg_to_smc_offset(struct pp_hwmgr *hwmgr);
-int smu7_wait_for_smc_inactive(struct pp_hwmgr *hwmgr);
 
 enum cgs_ucode_id smu7_convert_fw_type_to_cgs(uint32_t fw_type);
 int smu7_read_smc_sram_dword(struct pp_hwmgr *hwmgr, uint32_t smc_addr,
-- 
cgit v1.1


From 89a111476676add9ded0286fc7606508b5efb101 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Wed, 4 Apr 2018 18:41:08 +0800
Subject: drm/amd/pp: Remove useless smu7 running state check

Only check smc running state before start smu.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c | 8 --------
 1 file changed, 8 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
index 8b9518a..fb32a3f 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
@@ -167,10 +167,6 @@ int smu7_send_msg_to_smc(struct pp_hwmgr *hwmgr, uint16_t msg)
 {
 	int ret;
 
-	if (!smu7_is_smc_ram_running(hwmgr))
-		return -EINVAL;
-
-
 	PHM_WAIT_FIELD_UNEQUAL(hwmgr, SMC_RESP_0, SMC_RESP, 0);
 
 	ret = PHM_READ_FIELD(hwmgr->device, SMC_RESP_0, SMC_RESP);
@@ -199,10 +195,6 @@ int smu7_send_msg_to_smc_without_waiting(struct pp_hwmgr *hwmgr, uint16_t msg)
 
 int smu7_send_msg_to_smc_with_parameter(struct pp_hwmgr *hwmgr, uint16_t msg, uint32_t parameter)
 {
-	if (!smu7_is_smc_ram_running(hwmgr)) {
-		return -EINVAL;
-	}
-
 	PHM_WAIT_FIELD_UNEQUAL(hwmgr, SMC_RESP_0, SMC_RESP, 0);
 
 	cgs_write_register(hwmgr->device, mmSMC_MSG_ARG_0, parameter);
-- 
cgit v1.1


From 5452cf44d691edada697108f883c78edb40dc281 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Christian=20K=C3=B6nig?= <christian.koenig@amd.com>
Date: Fri, 9 Mar 2018 13:39:47 +0100
Subject: drm/ttm: keep a reference to transfer pipelined BOs
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Make sure the transfered BO is never destroy before the transfer BO.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Roger He <Hongbo.He@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/ttm/ttm_bo_util.c | 50 +++++++++++++++++++++++----------------
 1 file changed, 30 insertions(+), 20 deletions(-)

diff --git a/drivers/gpu/drm/ttm/ttm_bo_util.c b/drivers/gpu/drm/ttm/ttm_bo_util.c
index 2ebbae6..f3bf545 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_util.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_util.c
@@ -39,6 +39,11 @@
 #include <linux/module.h>
 #include <linux/reservation.h>
 
+struct ttm_transfer_obj {
+	struct ttm_buffer_object base;
+	struct ttm_buffer_object *bo;
+};
+
 void ttm_bo_free_old_node(struct ttm_buffer_object *bo)
 {
 	ttm_bo_mem_put(bo, &bo->mem);
@@ -454,7 +459,11 @@ EXPORT_SYMBOL(ttm_bo_move_memcpy);
 
 static void ttm_transfered_destroy(struct ttm_buffer_object *bo)
 {
-	kfree(bo);
+	struct ttm_transfer_obj *fbo;
+
+	fbo = container_of(bo, struct ttm_transfer_obj, base);
+	ttm_bo_unref(&fbo->bo);
+	kfree(fbo);
 }
 
 /**
@@ -475,14 +484,15 @@ static void ttm_transfered_destroy(struct ttm_buffer_object *bo)
 static int ttm_buffer_object_transfer(struct ttm_buffer_object *bo,
 				      struct ttm_buffer_object **new_obj)
 {
-	struct ttm_buffer_object *fbo;
+	struct ttm_transfer_obj *fbo;
 	int ret;
 
 	fbo = kmalloc(sizeof(*fbo), GFP_KERNEL);
 	if (!fbo)
 		return -ENOMEM;
 
-	*fbo = *bo;
+	fbo->base = *bo;
+	fbo->bo = ttm_bo_reference(bo);
 
 	/**
 	 * Fix up members that we shouldn't copy directly:
@@ -490,25 +500,25 @@ static int ttm_buffer_object_transfer(struct ttm_buffer_object *bo,
 	 */
 
 	atomic_inc(&bo->bdev->glob->bo_count);
-	INIT_LIST_HEAD(&fbo->ddestroy);
-	INIT_LIST_HEAD(&fbo->lru);
-	INIT_LIST_HEAD(&fbo->swap);
-	INIT_LIST_HEAD(&fbo->io_reserve_lru);
-	mutex_init(&fbo->wu_mutex);
-	fbo->moving = NULL;
-	drm_vma_node_reset(&fbo->vma_node);
-	atomic_set(&fbo->cpu_writers, 0);
-
-	kref_init(&fbo->list_kref);
-	kref_init(&fbo->kref);
-	fbo->destroy = &ttm_transfered_destroy;
-	fbo->acc_size = 0;
-	fbo->resv = &fbo->ttm_resv;
-	reservation_object_init(fbo->resv);
-	ret = reservation_object_trylock(fbo->resv);
+	INIT_LIST_HEAD(&fbo->base.ddestroy);
+	INIT_LIST_HEAD(&fbo->base.lru);
+	INIT_LIST_HEAD(&fbo->base.swap);
+	INIT_LIST_HEAD(&fbo->base.io_reserve_lru);
+	mutex_init(&fbo->base.wu_mutex);
+	fbo->base.moving = NULL;
+	drm_vma_node_reset(&fbo->base.vma_node);
+	atomic_set(&fbo->base.cpu_writers, 0);
+
+	kref_init(&fbo->base.list_kref);
+	kref_init(&fbo->base.kref);
+	fbo->base.destroy = &ttm_transfered_destroy;
+	fbo->base.acc_size = 0;
+	fbo->base.resv = &fbo->base.ttm_resv;
+	reservation_object_init(fbo->base.resv);
+	ret = reservation_object_trylock(fbo->base.resv);
 	WARN_ON(!ret);
 
-	*new_obj = fbo;
+	*new_obj = &fbo->base;
 	return 0;
 }
 
-- 
cgit v1.1


From 5eeae247d227c448d4db8f60ce184ddb0e0feca0 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Tue, 10 Apr 2018 10:15:26 -0500
Subject: drm/amdgpu/gfx9: cache DB_DEBUG2 and make it available to userspace
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Userspace needs to query this value to work around a hw bug in
certain cases.

Acked-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu.h   | 2 ++
 drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 1 +
 drivers/gpu/drm/amd/amdgpu/soc15.c    | 3 +++
 3 files changed, 6 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index 7e5defb..7eb0e48 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -867,6 +867,8 @@ struct amdgpu_gfx_config {
 
 	/* gfx configure feature */
 	uint32_t double_offchip_lds_buf;
+	/* cached value of DB_DEBUG2 */
+	uint32_t db_debug2;
 };
 
 struct amdgpu_cu_info {
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index 9d39fd5..66bd6c1 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -1600,6 +1600,7 @@ static void gfx_v9_0_gpu_init(struct amdgpu_device *adev)
 
 	gfx_v9_0_setup_rb(adev);
 	gfx_v9_0_get_cu_info(adev, &adev->gfx.cu_info);
+	adev->gfx.config.db_debug2 = RREG32_SOC15(GC, 0, mmDB_DEBUG2);
 
 	/* XXX SH_MEM regs */
 	/* where to put LDS, scratch, GPUVM in FSA64 space */
diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index 2e9ebe8..65e781f 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -287,6 +287,7 @@ static struct soc15_allowed_register_entry soc15_allowed_read_registers[] = {
 	{ SOC15_REG_ENTRY(GC, 0, mmCP_CPC_STALLED_STAT1)},
 	{ SOC15_REG_ENTRY(GC, 0, mmCP_CPC_STATUS)},
 	{ SOC15_REG_ENTRY(GC, 0, mmGB_ADDR_CONFIG)},
+	{ SOC15_REG_ENTRY(GC, 0, mmDB_DEBUG2)},
 };
 
 static uint32_t soc15_read_indexed_register(struct amdgpu_device *adev, u32 se_num,
@@ -315,6 +316,8 @@ static uint32_t soc15_get_register_value(struct amdgpu_device *adev,
 	} else {
 		if (reg_offset == SOC15_REG_OFFSET(GC, 0, mmGB_ADDR_CONFIG))
 			return adev->gfx.config.gb_addr_config;
+		else if (reg_offset == SOC15_REG_OFFSET(GC, 0, mmDB_DEBUG2))
+			return adev->gfx.config.db_debug2;
 		return RREG32(reg_offset);
 	}
 }
-- 
cgit v1.1


From 642ad57058baaa2c105925a75c153bb486877513 Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Thu, 12 Apr 2018 10:51:51 -0400
Subject: Revert "drm/amd/display: fix dereferencing possible ERR_PTR()"
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This reverts commit cd2d6c92a8e39d7e50a5af9fcc67d07e6a89e91d.

Cc: Shirish S <shirish.s@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 265f0166..0c29f3b 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -4941,9 +4941,6 @@ static int dm_atomic_check_plane_state_fb(struct drm_atomic_state *state,
 			return -EDEADLK;
 
 		crtc_state = drm_atomic_get_crtc_state(plane_state->state, crtc);
-		if (IS_ERR(crtc_state))
-			return PTR_ERR(crtc_state);
-
 		if (crtc->primary == plane && crtc_state->active) {
 			if (!plane_state->fb)
 				return -EINVAL;
-- 
cgit v1.1


From 23b9ad21b262b9a85e9b85813e4adfcfb0dd96b3 Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Thu, 12 Apr 2018 10:51:52 -0400
Subject: Revert "drm/amd/display: disable CRTCs with NULL FB on their primary
 plane (V2)"
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This seems to cause flickering and lock-ups for a wide range of users.
Revert until we've found a proper fix for the flickering and lock-ups.

This reverts commit 36cc549d59864b7161f0e23d710c1c4d1b9cf022.

Cc: Shirish S <shirish.s@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 28 -----------------------
 1 file changed, 28 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 0c29f3b..2368ade 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -4925,30 +4925,6 @@ static int dm_update_planes_state(struct dc *dc,
 	return ret;
 }
 
-static int dm_atomic_check_plane_state_fb(struct drm_atomic_state *state,
-					  struct drm_crtc *crtc)
-{
-	struct drm_plane *plane;
-	struct drm_crtc_state *crtc_state;
-
-	WARN_ON(!drm_atomic_get_new_crtc_state(state, crtc));
-
-	drm_for_each_plane_mask(plane, state->dev, crtc->state->plane_mask) {
-		struct drm_plane_state *plane_state =
-			drm_atomic_get_plane_state(state, plane);
-
-		if (IS_ERR(plane_state))
-			return -EDEADLK;
-
-		crtc_state = drm_atomic_get_crtc_state(plane_state->state, crtc);
-		if (crtc->primary == plane && crtc_state->active) {
-			if (!plane_state->fb)
-				return -EINVAL;
-		}
-	}
-	return 0;
-}
-
 static int amdgpu_dm_atomic_check(struct drm_device *dev,
 				  struct drm_atomic_state *state)
 {
@@ -4972,10 +4948,6 @@ static int amdgpu_dm_atomic_check(struct drm_device *dev,
 		goto fail;
 
 	for_each_oldnew_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) {
-		ret = dm_atomic_check_plane_state_fb(state, crtc);
-		if (ret)
-			goto fail;
-
 		if (!drm_atomic_crtc_needs_modeset(new_crtc_state) &&
 		    !new_crtc_state->color_mgmt_changed)
 			continue;
-- 
cgit v1.1


From c73a3626619018adfa2bb0fa1e64310be8e73152 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Wed, 11 Apr 2018 17:57:13 -0500
Subject: drm/amdgpu/powerplay: fix smu7_get_memory_type for fiji

Fiji uses a different register than other smu7 asics, but
we already have this info in the base driver so just
use that.

Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c | 8 ++------
 1 file changed, 2 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index 21c021b..97b7c23 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -4150,13 +4150,9 @@ static int smu7_read_clock_registers(struct pp_hwmgr *hwmgr)
 static int smu7_get_memory_type(struct pp_hwmgr *hwmgr)
 {
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
-	uint32_t temp;
-
-	temp = cgs_read_register(hwmgr->device, mmMC_SEQ_MISC0);
+	struct amdgpu_device *adev = hwmgr->adev;
 
-	data->is_memory_gddr5 = (MC_SEQ_MISC0_GDDR5_VALUE ==
-			((temp & MC_SEQ_MISC0_GDDR5_MASK) >>
-			 MC_SEQ_MISC0_GDDR5_SHIFT));
+	data->is_memory_gddr5 = (adev->gmc.vram_type == AMDGPU_VRAM_TYPE_GDDR5);
 
 	return 0;
 }
-- 
cgit v1.1


From 9da00630188da6e8ad1596c2b58809c833b16154 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Wed, 11 Apr 2018 18:09:39 -0500
Subject: drm/amdgpu/powerplay: rename smu7_upload_mc_firmware

It doesn't actually upload any firmware is just
checks the version.  The actual upload happens in
the gmc modules.

Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index 97b7c23..ed43dd3 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -4071,7 +4071,7 @@ static int smu7_check_states_equal(struct pp_hwmgr *hwmgr,
 	return 0;
 }
 
-static int smu7_upload_mc_firmware(struct pp_hwmgr *hwmgr)
+static int smu7_check_mc_firmware(struct pp_hwmgr *hwmgr)
 {
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
 
@@ -4200,7 +4200,7 @@ static int smu7_setup_asic_task(struct pp_hwmgr *hwmgr)
 {
 	int tmp_result, result = 0;
 
-	smu7_upload_mc_firmware(hwmgr);
+	smu7_check_mc_firmware(hwmgr);
 
 	tmp_result = smu7_read_clock_registers(hwmgr);
 	PP_ASSERT_WITH_CODE((0 == tmp_result),
-- 
cgit v1.1


From 828536385ab0d25b5ddd7153347df04ea3a6961d Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Tue, 27 Mar 2018 11:58:14 -0500
Subject: drm/amdgpu: add emit_reg_write_reg_wait ring callback
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This callback writes a value to a register and then reads
back another register and waits for a value in a single
operation.

Provide a helper function using two operations for engines
that don't support this opertion.

Reviewed-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu.h      |  1 +
 drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c | 20 ++++++++++++++++++++
 drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h |  7 +++++++
 3 files changed, 28 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index 7eb0e48..c25ee75 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -1806,6 +1806,7 @@ amdgpu_get_sdma_instance(struct amdgpu_ring *ring)
 #define amdgpu_ring_emit_rreg(r, d) (r)->funcs->emit_rreg((r), (d))
 #define amdgpu_ring_emit_wreg(r, d, v) (r)->funcs->emit_wreg((r), (d), (v))
 #define amdgpu_ring_emit_reg_wait(r, d, v, m) (r)->funcs->emit_reg_wait((r), (d), (v), (m))
+#define amdgpu_ring_emit_reg_write_reg_wait(r, d0, d1, v, m) (r)->funcs->emit_reg_write_reg_wait((r), (d0), (d1), (v), (m))
 #define amdgpu_ring_emit_tmz(r, b) (r)->funcs->emit_tmz((r), (b))
 #define amdgpu_ring_pad_ib(r, ib) ((r)->funcs->pad_ib((r), (ib)))
 #define amdgpu_ring_init_cond_exec(r) (r)->funcs->init_cond_exec((r))
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
index d5f526f..49cad08 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
@@ -459,6 +459,26 @@ void amdgpu_ring_lru_touch(struct amdgpu_device *adev, struct amdgpu_ring *ring)
 	spin_unlock(&adev->ring_lru_list_lock);
 }
 
+/**
+ * amdgpu_ring_emit_reg_write_reg_wait_helper - ring helper
+ *
+ * @adev: amdgpu_device pointer
+ * @reg0: register to write
+ * @reg1: register to wait on
+ * @ref: reference value to write/wait on
+ * @mask: mask to wait on
+ *
+ * Helper for rings that don't support write and wait in a
+ * single oneshot packet.
+ */
+void amdgpu_ring_emit_reg_write_reg_wait_helper(struct amdgpu_ring *ring,
+						uint32_t reg0, uint32_t reg1,
+						uint32_t ref, uint32_t mask)
+{
+	amdgpu_ring_emit_wreg(ring, reg0, ref);
+	amdgpu_ring_emit_reg_wait(ring, reg1, mask, mask);
+}
+
 /*
  * Debugfs info
  */
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h
index 1a59118..08fcdf6 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h
@@ -154,6 +154,9 @@ struct amdgpu_ring_funcs {
 	void (*emit_wreg)(struct amdgpu_ring *ring, uint32_t reg, uint32_t val);
 	void (*emit_reg_wait)(struct amdgpu_ring *ring, uint32_t reg,
 			      uint32_t val, uint32_t mask);
+	void (*emit_reg_write_reg_wait)(struct amdgpu_ring *ring,
+					uint32_t reg0, uint32_t reg1,
+					uint32_t ref, uint32_t mask);
 	void (*emit_tmz)(struct amdgpu_ring *ring, bool start);
 	/* priority functions */
 	void (*set_priority) (struct amdgpu_ring *ring,
@@ -228,6 +231,10 @@ int amdgpu_ring_lru_get(struct amdgpu_device *adev, int type,
 			int *blacklist, int num_blacklist,
 			bool lru_pipe_order, struct amdgpu_ring **ring);
 void amdgpu_ring_lru_touch(struct amdgpu_device *adev, struct amdgpu_ring *ring);
+void amdgpu_ring_emit_reg_write_reg_wait_helper(struct amdgpu_ring *ring,
+						uint32_t reg0, uint32_t val0,
+						uint32_t reg1, uint32_t val1);
+
 static inline void amdgpu_ring_clear_ring(struct amdgpu_ring *ring)
 {
 	int i = 0;
-- 
cgit v1.1


From 10ed3c3190d38f189ed6857cecca1a2eb6de33a3 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Tue, 27 Mar 2018 15:07:50 -0500
Subject: drm/amdgpu/gfx9: add emit_reg_write_reg_wait ring callback (v2)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This adds support for writing and reading back in a single
oneshot packet.  This is needed to send a tlb invalidation
and wait for ack in a single operation.

v2: squash the gfx ring stall fix

Reviewed-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Emily Deng <Emily.Deng@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index 66bd6c1..583f6f6 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -4138,6 +4138,15 @@ static void gfx_v9_0_ring_emit_reg_wait(struct amdgpu_ring *ring, uint32_t reg,
 	gfx_v9_0_wait_reg_mem(ring, 0, 0, 0, reg, 0, val, mask, 0x20);
 }
 
+static void gfx_v9_0_ring_emit_reg_write_reg_wait(struct amdgpu_ring *ring,
+						  uint32_t reg0, uint32_t reg1,
+						  uint32_t ref, uint32_t mask)
+{
+	int usepfp = (ring->funcs->type == AMDGPU_RING_TYPE_GFX);
+
+	gfx_v9_0_wait_reg_mem(ring, usepfp, 0, 1, reg0, reg1, ref, mask, 0x20);
+}
+
 static void gfx_v9_0_set_gfx_eop_interrupt_state(struct amdgpu_device *adev,
 						 enum amdgpu_interrupt_state state)
 {
@@ -4459,6 +4468,7 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_gfx = {
 	.emit_tmz = gfx_v9_0_ring_emit_tmz,
 	.emit_wreg = gfx_v9_0_ring_emit_wreg,
 	.emit_reg_wait = gfx_v9_0_ring_emit_reg_wait,
+	.emit_reg_write_reg_wait = gfx_v9_0_ring_emit_reg_write_reg_wait,
 };
 
 static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_compute = {
@@ -4493,6 +4503,7 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_compute = {
 	.set_priority = gfx_v9_0_ring_set_priority_compute,
 	.emit_wreg = gfx_v9_0_ring_emit_wreg,
 	.emit_reg_wait = gfx_v9_0_ring_emit_reg_wait,
+	.emit_reg_write_reg_wait = gfx_v9_0_ring_emit_reg_write_reg_wait,
 };
 
 static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_kiq = {
@@ -4523,6 +4534,7 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_kiq = {
 	.emit_rreg = gfx_v9_0_ring_emit_rreg,
 	.emit_wreg = gfx_v9_0_ring_emit_wreg,
 	.emit_reg_wait = gfx_v9_0_ring_emit_reg_wait,
+	.emit_reg_write_reg_wait = gfx_v9_0_ring_emit_reg_write_reg_wait,
 };
 
 static void gfx_v9_0_set_ring_funcs(struct amdgpu_device *adev)
-- 
cgit v1.1


From 4dfe7d7b4e3ba16fc377a48a221bfe8172bc50e1 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Tue, 27 Mar 2018 16:51:41 -0500
Subject: drm/amdgpu/sdma4: add emit_reg_write_reg_wait ring callback (v2)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This adds support for writing and reading back in a single
oneshot packet.  This is needed to send a tlb invalidation
and wait for ack in a single operation.

v2: squash sdma hang fix into this patch

Reviewed-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Emily Deng <Emily.Deng@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
index 2c618a1..03a36cb 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
@@ -1611,6 +1611,7 @@ static const struct amdgpu_ring_funcs sdma_v4_0_ring_funcs = {
 	.pad_ib = sdma_v4_0_ring_pad_ib,
 	.emit_wreg = sdma_v4_0_ring_emit_wreg,
 	.emit_reg_wait = sdma_v4_0_ring_emit_reg_wait,
+	.emit_reg_write_reg_wait = amdgpu_ring_emit_reg_write_reg_wait_helper,
 };
 
 static void sdma_v4_0_set_ring_funcs(struct amdgpu_device *adev)
-- 
cgit v1.1


From 1ab0c9a75f66293a8ea719cc96ae4141218eb0e4 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Tue, 27 Mar 2018 17:05:19 -0500
Subject: drm/amdgpu/uvd7: add emit_reg_write_reg_wait ring callback
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This adds support for writing and reading back using the
helper since the engines doesn't have a oneshot packet.

Reviewed-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
index eddc57f..280c082 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
@@ -1702,6 +1702,7 @@ static const struct amdgpu_ring_funcs uvd_v7_0_enc_ring_vm_funcs = {
 	.end_use = amdgpu_uvd_ring_end_use,
 	.emit_wreg = uvd_v7_0_enc_ring_emit_wreg,
 	.emit_reg_wait = uvd_v7_0_enc_ring_emit_reg_wait,
+	.emit_reg_write_reg_wait = amdgpu_ring_emit_reg_write_reg_wait_helper,
 };
 
 static void uvd_v7_0_set_ring_funcs(struct amdgpu_device *adev)
-- 
cgit v1.1


From 3fa0b1cbc0a57a21c1688601f6b9c340441ba3b6 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Tue, 27 Mar 2018 17:06:33 -0500
Subject: drm/amdgpu/vce4: add emit_reg_write_reg_wait ring callback
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This adds support for writing and reading back using the
helper since the engines doesn't have a oneshot packet.

Reviewed-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/vce_v4_0.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c
index 73fd48d..8fd1b74 100755
--- a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c
@@ -1081,6 +1081,7 @@ static const struct amdgpu_ring_funcs vce_v4_0_ring_vm_funcs = {
 	.end_use = amdgpu_vce_ring_end_use,
 	.emit_wreg = vce_v4_0_emit_wreg,
 	.emit_reg_wait = vce_v4_0_emit_reg_wait,
+	.emit_reg_write_reg_wait = amdgpu_ring_emit_reg_write_reg_wait_helper,
 };
 
 static void vce_v4_0_set_ring_funcs(struct amdgpu_device *adev)
-- 
cgit v1.1


From f58b85e3ec0e3d3ddeff6eb16ace23a42516ae70 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Tue, 27 Mar 2018 17:06:52 -0500
Subject: drm/amdgpu/vcn1: add emit_reg_write_reg_wait ring callback
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This adds support for writing and reading back using the
helper since the engines doesn't have a oneshot packet.

Reviewed-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
index 8c13267..d9a1533 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
@@ -1139,6 +1139,7 @@ static const struct amdgpu_ring_funcs vcn_v1_0_enc_ring_vm_funcs = {
 	.end_use = amdgpu_vcn_ring_end_use,
 	.emit_wreg = vcn_v1_0_enc_ring_emit_wreg,
 	.emit_reg_wait = vcn_v1_0_enc_ring_emit_reg_wait,
+	.emit_reg_write_reg_wait = amdgpu_ring_emit_reg_write_reg_wait_helper,
 };
 
 static void vcn_v1_0_set_dec_ring_funcs(struct amdgpu_device *adev)
-- 
cgit v1.1


From f8bc903707ae87342b97528037e27bf190051c11 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Tue, 27 Mar 2018 17:10:56 -0500
Subject: drm/amdgpu/gmc9: use amdgpu_ring_emit_reg_write_reg_wait in gpu tlb
 flush
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Use amdgpu_ring_emit_reg_write_reg_wait.  On engines that support it,
it provides a write and wait in a single packet which avoids a missed
ack if a world switch happens between the request and waiting for the
ack.

Reviewed-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c | 8 +++-----
 1 file changed, 3 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
index 070946e..aeaed7f 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
@@ -385,11 +385,9 @@ static uint64_t gmc_v9_0_emit_flush_gpu_tlb(struct amdgpu_ring *ring,
 	amdgpu_ring_emit_wreg(ring, hub->ctx0_ptb_addr_hi32 + (2 * vmid),
 			      upper_32_bits(pd_addr));
 
-	amdgpu_ring_emit_wreg(ring, hub->vm_inv_eng0_req + eng, req);
-
-	/* wait for the invalidate to complete */
-	amdgpu_ring_emit_reg_wait(ring, hub->vm_inv_eng0_ack + eng,
-				  1 << vmid, 1 << vmid);
+	amdgpu_ring_emit_reg_write_reg_wait(ring, hub->vm_inv_eng0_req + eng,
+					    hub->vm_inv_eng0_ack + eng,
+					    req, 1 << vmid);
 
 	return pd_addr;
 }
-- 
cgit v1.1


From ebdef28ebbcf767d9fa687acb1d02d97d834c628 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexdeucher@gmail.com>
Date: Fri, 6 Apr 2018 14:54:09 -0500
Subject: drm/amdgpu/gmc: steal the appropriate amount of vram for fw hand-over
 (v3)

Steal 9 MB for vga emulation and fb if vga is enabled, otherwise,
steal enough to cover the current display size as set by the vbios.

If no memory is used (e.g., secondary or headless card), skip
stolen memory reserve.

v2: skip reservation if vram is limited, address Christian's comments
v3: squash in fix from Harry

Reviewed-and-Tested-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com> (v2)
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 14 +++++----
 drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c   | 23 ++++++++++++--
 drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c   | 23 ++++++++++++--
 drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c   | 23 ++++++++++++--
 drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c   | 53 +++++++++++++++++++++++++++++----
 5 files changed, 118 insertions(+), 18 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index ab73300..2be04ac 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -1441,12 +1441,14 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 		return r;
 	}
 
-	r = amdgpu_bo_create_kernel(adev, adev->gmc.stolen_size, PAGE_SIZE,
-				    AMDGPU_GEM_DOMAIN_VRAM,
-				    &adev->stolen_vga_memory,
-				    NULL, NULL);
-	if (r)
-		return r;
+	if (adev->gmc.stolen_size) {
+		r = amdgpu_bo_create_kernel(adev, adev->gmc.stolen_size, PAGE_SIZE,
+					    AMDGPU_GEM_DOMAIN_VRAM,
+					    &adev->stolen_vga_memory,
+					    NULL, NULL);
+		if (r)
+			return r;
+	}
 	DRM_INFO("amdgpu: %uM of VRAM memory ready\n",
 		 (unsigned) (adev->gmc.real_vram_size / (1024 * 1024)));
 
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
index 5617cf6..24e1ea3 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
@@ -825,6 +825,25 @@ static int gmc_v6_0_late_init(void *handle)
 		return 0;
 }
 
+static unsigned gmc_v6_0_get_vbios_fb_size(struct amdgpu_device *adev)
+{
+	u32 d1vga_control = RREG32(mmD1VGA_CONTROL);
+	unsigned size;
+
+	if (REG_GET_FIELD(d1vga_control, D1VGA_CONTROL, D1VGA_MODE_ENABLE)) {
+		size = 9 * 1024 * 1024; /* reserve 8MB for vga emulator and 1 MB for FB */
+	} else {
+		u32 viewport = RREG32(mmVIEWPORT_SIZE);
+		size = (REG_GET_FIELD(viewport, VIEWPORT_SIZE, VIEWPORT_HEIGHT) *
+			REG_GET_FIELD(viewport, VIEWPORT_SIZE, VIEWPORT_WIDTH) *
+			4);
+	}
+	/* return 0 if the pre-OS buffer uses up most of vram */
+	if ((adev->gmc.real_vram_size - size) < (8 * 1024 * 1024))
+		return 0;
+	return size;
+}
+
 static int gmc_v6_0_sw_init(void *handle)
 {
 	int r;
@@ -851,8 +870,6 @@ static int gmc_v6_0_sw_init(void *handle)
 
 	adev->gmc.mc_mask = 0xffffffffffULL;
 
-	adev->gmc.stolen_size = 256 * 1024;
-
 	adev->need_dma32 = false;
 	dma_bits = adev->need_dma32 ? 32 : 40;
 	r = pci_set_dma_mask(adev->pdev, DMA_BIT_MASK(dma_bits));
@@ -878,6 +895,8 @@ static int gmc_v6_0_sw_init(void *handle)
 	if (r)
 		return r;
 
+	adev->gmc.stolen_size = gmc_v6_0_get_vbios_fb_size(adev);
+
 	r = amdgpu_bo_init(adev);
 	if (r)
 		return r;
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
index 80054f3..93861f9 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
@@ -964,6 +964,25 @@ static int gmc_v7_0_late_init(void *handle)
 		return 0;
 }
 
+static unsigned gmc_v7_0_get_vbios_fb_size(struct amdgpu_device *adev)
+{
+	u32 d1vga_control = RREG32(mmD1VGA_CONTROL);
+	unsigned size;
+
+	if (REG_GET_FIELD(d1vga_control, D1VGA_CONTROL, D1VGA_MODE_ENABLE)) {
+		size = 9 * 1024 * 1024; /* reserve 8MB for vga emulator and 1 MB for FB */
+	} else {
+		u32 viewport = RREG32(mmVIEWPORT_SIZE);
+		size = (REG_GET_FIELD(viewport, VIEWPORT_SIZE, VIEWPORT_HEIGHT) *
+			REG_GET_FIELD(viewport, VIEWPORT_SIZE, VIEWPORT_WIDTH) *
+			4);
+	}
+	/* return 0 if the pre-OS buffer uses up most of vram */
+	if ((adev->gmc.real_vram_size - size) < (8 * 1024 * 1024))
+		return 0;
+	return size;
+}
+
 static int gmc_v7_0_sw_init(void *handle)
 {
 	int r;
@@ -998,8 +1017,6 @@ static int gmc_v7_0_sw_init(void *handle)
 	 */
 	adev->gmc.mc_mask = 0xffffffffffULL; /* 40 bit MC */
 
-	adev->gmc.stolen_size = 256 * 1024;
-
 	/* set DMA mask + need_dma32 flags.
 	 * PCIE - can handle 40-bits.
 	 * IGP - can handle 40-bits
@@ -1030,6 +1047,8 @@ static int gmc_v7_0_sw_init(void *handle)
 	if (r)
 		return r;
 
+	adev->gmc.stolen_size = gmc_v7_0_get_vbios_fb_size(adev);
+
 	/* Memory manager */
 	r = amdgpu_bo_init(adev);
 	if (r)
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
index d71d4cb..fbd8f56 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
@@ -1055,6 +1055,25 @@ static int gmc_v8_0_late_init(void *handle)
 		return 0;
 }
 
+static unsigned gmc_v8_0_get_vbios_fb_size(struct amdgpu_device *adev)
+{
+	u32 d1vga_control = RREG32(mmD1VGA_CONTROL);
+	unsigned size;
+
+	if (REG_GET_FIELD(d1vga_control, D1VGA_CONTROL, D1VGA_MODE_ENABLE)) {
+		size = 9 * 1024 * 1024; /* reserve 8MB for vga emulator and 1 MB for FB */
+	} else {
+		u32 viewport = RREG32(mmVIEWPORT_SIZE);
+		size = (REG_GET_FIELD(viewport, VIEWPORT_SIZE, VIEWPORT_HEIGHT) *
+			REG_GET_FIELD(viewport, VIEWPORT_SIZE, VIEWPORT_WIDTH) *
+			4);
+	}
+	/* return 0 if the pre-OS buffer uses up most of vram */
+	if ((adev->gmc.real_vram_size - size) < (8 * 1024 * 1024))
+		return 0;
+	return size;
+}
+
 #define mmMC_SEQ_MISC0_FIJI 0xA71
 
 static int gmc_v8_0_sw_init(void *handle)
@@ -1096,8 +1115,6 @@ static int gmc_v8_0_sw_init(void *handle)
 	 */
 	adev->gmc.mc_mask = 0xffffffffffULL; /* 40 bit MC */
 
-	adev->gmc.stolen_size = 256 * 1024;
-
 	/* set DMA mask + need_dma32 flags.
 	 * PCIE - can handle 40-bits.
 	 * IGP - can handle 40-bits
@@ -1128,6 +1145,8 @@ static int gmc_v8_0_sw_init(void *handle)
 	if (r)
 		return r;
 
+	adev->gmc.stolen_size = gmc_v8_0_get_vbios_fb_size(adev);
+
 	/* Memory manager */
 	r = amdgpu_bo_init(adev);
 	if (r)
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
index aeaed7f..3071f51 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
@@ -57,6 +57,14 @@
 #define DF_CS_AON0_DramBaseAddress0__IntLvAddrSel_MASK                                                        0x00000700L
 #define DF_CS_AON0_DramBaseAddress0__DramBaseAddr_MASK                                                        0xFFFFF000L
 
+/* add these here since we already include dce12 headers and these are for DCN */
+#define mmHUBP0_DCSURF_PRI_VIEWPORT_DIMENSION                                                          0x055d
+#define mmHUBP0_DCSURF_PRI_VIEWPORT_DIMENSION_BASE_IDX                                                 2
+#define HUBP0_DCSURF_PRI_VIEWPORT_DIMENSION__PRI_VIEWPORT_WIDTH__SHIFT                                        0x0
+#define HUBP0_DCSURF_PRI_VIEWPORT_DIMENSION__PRI_VIEWPORT_HEIGHT__SHIFT                                       0x10
+#define HUBP0_DCSURF_PRI_VIEWPORT_DIMENSION__PRI_VIEWPORT_WIDTH_MASK                                          0x00003FFFL
+#define HUBP0_DCSURF_PRI_VIEWPORT_DIMENSION__PRI_VIEWPORT_HEIGHT_MASK                                         0x3FFF0000L
+
 /* XXX Move this macro to VEGA10 header file, which is like vid.h for VI.*/
 #define AMDGPU_NUM_OF_VMIDS			8
 
@@ -791,6 +799,43 @@ static int gmc_v9_0_gart_init(struct amdgpu_device *adev)
 	return amdgpu_gart_table_vram_alloc(adev);
 }
 
+static unsigned gmc_v9_0_get_vbios_fb_size(struct amdgpu_device *adev)
+{
+#if 0
+	u32 d1vga_control = RREG32_SOC15(DCE, 0, mmD1VGA_CONTROL);
+#endif
+	unsigned size;
+
+	if (REG_GET_FIELD(d1vga_control, D1VGA_CONTROL, D1VGA_MODE_ENABLE)) {
+		size = 9 * 1024 * 1024; /* reserve 8MB for vga emulator and 1 MB for FB */
+	} else {
+		u32 viewport;
+
+		switch (adev->asic_type) {
+		case CHIP_RAVEN:
+			viewport = RREG32_SOC15(DCE, 0, mmHUBP0_DCSURF_PRI_VIEWPORT_DIMENSION);
+			size = (REG_GET_FIELD(viewport,
+					      HUBP0_DCSURF_PRI_VIEWPORT_DIMENSION, PRI_VIEWPORT_HEIGHT) *
+				REG_GET_FIELD(viewport,
+					      HUBP0_DCSURF_PRI_VIEWPORT_DIMENSION, PRI_VIEWPORT_WIDTH) *
+				4);
+			break;
+		case CHIP_VEGA10:
+		case CHIP_VEGA12:
+		default:
+			viewport = RREG32_SOC15(DCE, 0, mmSCL0_VIEWPORT_SIZE);
+			size = (REG_GET_FIELD(viewport, SCL0_VIEWPORT_SIZE, VIEWPORT_HEIGHT) *
+				REG_GET_FIELD(viewport, SCL0_VIEWPORT_SIZE, VIEWPORT_WIDTH) *
+				4);
+			break;
+		}
+	}
+	/* return 0 if the pre-OS buffer uses up most of vram */
+	if ((adev->gmc.real_vram_size - size) < (8 * 1024 * 1024))
+		return 0;
+	return size;
+}
+
 static int gmc_v9_0_sw_init(void *handle)
 {
 	int r;
@@ -842,12 +887,6 @@ static int gmc_v9_0_sw_init(void *handle)
 	 */
 	adev->gmc.mc_mask = 0xffffffffffffULL; /* 48 bit MC */
 
-	/*
-	 * It needs to reserve 8M stolen memory for vega10
-	 * TODO: Figure out how to avoid that...
-	 */
-	adev->gmc.stolen_size = 8 * 1024 * 1024;
-
 	/* set DMA mask + need_dma32 flags.
 	 * PCIE - can handle 44-bits.
 	 * IGP - can handle 44-bits
@@ -872,6 +911,8 @@ static int gmc_v9_0_sw_init(void *handle)
 	if (r)
 		return r;
 
+	adev->gmc.stolen_size = gmc_v9_0_get_vbios_fb_size(adev);
+
 	/* Memory manager */
 	r = amdgpu_bo_init(adev);
 	if (r)
-- 
cgit v1.1


From 6f752ec2c20c6a575da29d5b297980f376830e6b Mon Sep 17 00:00:00 2001
From: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Date: Fri, 6 Apr 2018 14:54:10 -0500
Subject: drm/amdgpu: Free VGA stolen memory as soon as possible.

Reserved VRAM is used to avoid overriding pre OS FB.
Once our display stack takes over we don't need the reserved
VRAM anymore.

v2:
Remove comment, we know actually why we need to reserve the stolen VRAM.
Fix return type for amdgpu_ttm_late_init.
v3:
Return 0 in amdgpu_bo_late_init, rebase on changes to previous patch
v4: rebase
v5:
For GMC9 reserve always just 9M and keep the stolem memory around
until GART table curruption on S3 resume is resolved.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_object.c |  7 +++++++
 drivers/gpu/drm/amd/amdgpu/amdgpu_object.h |  1 +
 drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c    |  6 +++++-
 drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h    |  1 +
 drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c      |  2 ++
 drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c      |  2 ++
 drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c      |  2 ++
 drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c      | 26 ++++++++++++++++++++++++++
 8 files changed, 46 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
index 9e23d6f..a160ef0 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
@@ -852,6 +852,13 @@ int amdgpu_bo_init(struct amdgpu_device *adev)
 	return amdgpu_ttm_init(adev);
 }
 
+int amdgpu_bo_late_init(struct amdgpu_device *adev)
+{
+	amdgpu_ttm_late_init(adev);
+
+	return 0;
+}
+
 void amdgpu_bo_fini(struct amdgpu_device *adev)
 {
 	amdgpu_ttm_fini(adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
index 3bee133..1e9fe85 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
@@ -251,6 +251,7 @@ int amdgpu_bo_pin_restricted(struct amdgpu_bo *bo, u32 domain,
 int amdgpu_bo_unpin(struct amdgpu_bo *bo);
 int amdgpu_bo_evict_vram(struct amdgpu_device *adev);
 int amdgpu_bo_init(struct amdgpu_device *adev);
+int amdgpu_bo_late_init(struct amdgpu_device *adev);
 void amdgpu_bo_fini(struct amdgpu_device *adev);
 int amdgpu_bo_fbdev_mmap(struct amdgpu_bo *bo,
 				struct vm_area_struct *vma);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index 2be04ac..29efaac 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -1517,13 +1517,17 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 	return 0;
 }
 
+void amdgpu_ttm_late_init(struct amdgpu_device *adev)
+{
+	amdgpu_bo_free_kernel(&adev->stolen_vga_memory, NULL, NULL);
+}
+
 void amdgpu_ttm_fini(struct amdgpu_device *adev)
 {
 	if (!adev->mman.initialized)
 		return;
 
 	amdgpu_ttm_debugfs_fini(adev);
-	amdgpu_bo_free_kernel(&adev->stolen_vga_memory, NULL, NULL);
 	amdgpu_ttm_fw_reserve_vram_fini(adev);
 	if (adev->mman.aper_base_kaddr)
 		iounmap(adev->mman.aper_base_kaddr);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
index 6ea7de8..e969c87 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
@@ -77,6 +77,7 @@ uint64_t amdgpu_vram_mgr_usage(struct ttm_mem_type_manager *man);
 uint64_t amdgpu_vram_mgr_vis_usage(struct ttm_mem_type_manager *man);
 
 int amdgpu_ttm_init(struct amdgpu_device *adev);
+void amdgpu_ttm_late_init(struct amdgpu_device *adev);
 void amdgpu_ttm_fini(struct amdgpu_device *adev);
 void amdgpu_ttm_set_buffer_funcs_status(struct amdgpu_device *adev,
 					bool enable);
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
index 24e1ea3..79f9ac2 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
@@ -819,6 +819,8 @@ static int gmc_v6_0_late_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
+	amdgpu_bo_late_init(adev);
+
 	if (amdgpu_vm_fault_stop != AMDGPU_VM_FAULT_STOP_ALWAYS)
 		return amdgpu_irq_get(adev, &adev->gmc.vm_fault, 0);
 	else
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
index 93861f9..7147bfe 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
@@ -958,6 +958,8 @@ static int gmc_v7_0_late_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
+	amdgpu_bo_late_init(adev);
+
 	if (amdgpu_vm_fault_stop != AMDGPU_VM_FAULT_STOP_ALWAYS)
 		return amdgpu_irq_get(adev, &adev->gmc.vm_fault, 0);
 	else
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
index fbd8f56..4d970da 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
@@ -1049,6 +1049,8 @@ static int gmc_v8_0_late_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
+	amdgpu_bo_late_init(adev);
+
 	if (amdgpu_vm_fault_stop != AMDGPU_VM_FAULT_STOP_ALWAYS)
 		return amdgpu_irq_get(adev, &adev->gmc.vm_fault, 0);
 	else
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
index 3071f51..e6b00b5 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
@@ -665,6 +665,11 @@ static int gmc_v9_0_late_init(void *handle)
 	unsigned i;
 	int r;
 
+	/*
+	 * TODO - Uncomment once GART corruption issue is fixed.
+	 */
+	/* amdgpu_bo_late_init(adev); */
+
 	for(i = 0; i < adev->num_rings; ++i) {
 		struct amdgpu_ring *ring = adev->rings[i];
 		unsigned vmhub = ring->funcs->vmhub;
@@ -806,6 +811,13 @@ static unsigned gmc_v9_0_get_vbios_fb_size(struct amdgpu_device *adev)
 #endif
 	unsigned size;
 
+	/*
+	 * TODO Remove once GART corruption is resolved
+	 * Check related code in gmc_v9_0_sw_fini
+	 * */
+	size = 9 * 1024 * 1024;
+
+#if 0
 	if (REG_GET_FIELD(d1vga_control, D1VGA_CONTROL, D1VGA_MODE_ENABLE)) {
 		size = 9 * 1024 * 1024; /* reserve 8MB for vga emulator and 1 MB for FB */
 	} else {
@@ -833,6 +845,8 @@ static unsigned gmc_v9_0_get_vbios_fb_size(struct amdgpu_device *adev)
 	/* return 0 if the pre-OS buffer uses up most of vram */
 	if ((adev->gmc.real_vram_size - size) < (8 * 1024 * 1024))
 		return 0;
+
+#endif
 	return size;
 }
 
@@ -956,6 +970,18 @@ static int gmc_v9_0_sw_fini(void *handle)
 	amdgpu_gem_force_release(adev);
 	amdgpu_vm_manager_fini(adev);
 	gmc_v9_0_gart_fini(adev);
+
+	/*
+	* TODO:
+	* Currently there is a bug where some memory client outside
+	* of the driver writes to first 8M of VRAM on S3 resume,
+	* this overrides GART which by default gets placed in first 8M and
+	* causes VM_FAULTS once GTT is accessed.
+	* Keep the stolen memory reservation until the while this is not solved.
+	* Also check code in gmc_v9_0_get_vbios_fb_size and gmc_v9_0_late_init
+	*/
+	amdgpu_bo_free_kernel(&adev->stolen_vga_memory, NULL, NULL);
+
 	amdgpu_bo_fini(adev);
 
 	return 0;
-- 
cgit v1.1


From 8ee3a52e3f35e064a3bf82f21dc74ddaf9843648 Mon Sep 17 00:00:00 2001
From: Emily Deng <Emily.Deng@amd.com>
Date: Mon, 16 Apr 2018 10:07:02 +0800
Subject: drm/gpu-sched: fix force APP kill hang(v4)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

issue:
there are VMC page fault occurred if force APP kill during
3dmark test, the cause is in entity_fini we manually signal
all those jobs in entity's queue which confuse the sync/dep
mechanism:

1)page fault occurred in sdma's clear job which operate on
shadow buffer, and shadow buffer's Gart table is cleaned by
ttm_bo_release since the fence in its reservation was fake signaled
by entity_fini() under the case of SIGKILL received.

2)page fault occurred in gfx' job because during the lifetime
of gfx job we manually fake signal all jobs from its entity
in entity_fini(), thus the unmapping/clear PTE job depend on those
result fence is satisfied and sdma start clearing the PTE and lead
to GFX page fault.

fix:
1)should at least wait all jobs already scheduled complete in entity_fini()
if SIGKILL is the case.

2)if a fence signaled and try to clear some entity's dependency, should
set this entity guilty to prevent its job really run since the dependency
is fake signaled.

v2:
splitting drm_sched_entity_fini() into two functions:
1)The first one is does the waiting, removes the entity from the
runqueue and returns an error when the process was killed.
2)The second one then goes over the entity, install it as
completion signal for the remaining jobs and signals all jobs
with an error code.

v3:
1)Replace the fini1 and fini2 with better name
2)Call the first part before the VM teardown in
amdgpu_driver_postclose_kms() and the second part
after the VM teardown
3)Keep the original function drm_sched_entity_fini to
refine the code.

v4:
1)Rename entity->finished to entity->last_scheduled;
2)Rename drm_sched_entity_fini_job_cb() to
drm_sched_entity_kill_jobs_cb();
3)Pass NULL to drm_sched_entity_fini_job_cb() if -ENOENT;
4)Replace the type of entity->fini_status with "int";
5)Remove the check about entity->finished.

Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Signed-off-by: Emily Deng <Emily.Deng@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu.h       |  2 +
 drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c   | 64 ++++++++++++++++++++++++----
 drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c   |  5 ++-
 drivers/gpu/drm/scheduler/gpu_scheduler.c | 71 ++++++++++++++++++++++++++-----
 include/drm/gpu_scheduler.h               |  7 +++
 5 files changed, 128 insertions(+), 21 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index c25ee75..ea1b285 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -681,6 +681,8 @@ int amdgpu_ctx_ioctl(struct drm_device *dev, void *data,
 int amdgpu_ctx_wait_prev_fence(struct amdgpu_ctx *ctx, unsigned ring_id);
 
 void amdgpu_ctx_mgr_init(struct amdgpu_ctx_mgr *mgr);
+void amdgpu_ctx_mgr_entity_cleanup(struct amdgpu_ctx_mgr *mgr);
+void amdgpu_ctx_mgr_entity_fini(struct amdgpu_ctx_mgr *mgr);
 void amdgpu_ctx_mgr_fini(struct amdgpu_ctx_mgr *mgr);
 
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
index 09d35051..eb80edf 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
@@ -111,8 +111,9 @@ failed:
 	return r;
 }
 
-static void amdgpu_ctx_fini(struct amdgpu_ctx *ctx)
+static void amdgpu_ctx_fini(struct kref *ref)
 {
+	struct amdgpu_ctx *ctx = container_of(ref, struct amdgpu_ctx, refcount);
 	struct amdgpu_device *adev = ctx->adev;
 	unsigned i, j;
 
@@ -125,13 +126,11 @@ static void amdgpu_ctx_fini(struct amdgpu_ctx *ctx)
 	kfree(ctx->fences);
 	ctx->fences = NULL;
 
-	for (i = 0; i < adev->num_rings; i++)
-		drm_sched_entity_fini(&adev->rings[i]->sched,
-				      &ctx->rings[i].entity);
-
 	amdgpu_queue_mgr_fini(adev, &ctx->queue_mgr);
 
 	mutex_destroy(&ctx->lock);
+
+	kfree(ctx);
 }
 
 static int amdgpu_ctx_alloc(struct amdgpu_device *adev,
@@ -170,12 +169,15 @@ static int amdgpu_ctx_alloc(struct amdgpu_device *adev,
 static void amdgpu_ctx_do_release(struct kref *ref)
 {
 	struct amdgpu_ctx *ctx;
+	u32 i;
 
 	ctx = container_of(ref, struct amdgpu_ctx, refcount);
 
-	amdgpu_ctx_fini(ctx);
+	for (i = 0; i < ctx->adev->num_rings; i++)
+		drm_sched_entity_fini(&ctx->adev->rings[i]->sched,
+			&ctx->rings[i].entity);
 
-	kfree(ctx);
+	amdgpu_ctx_fini(ref);
 }
 
 static int amdgpu_ctx_free(struct amdgpu_fpriv *fpriv, uint32_t id)
@@ -435,16 +437,62 @@ void amdgpu_ctx_mgr_init(struct amdgpu_ctx_mgr *mgr)
 	idr_init(&mgr->ctx_handles);
 }
 
+void amdgpu_ctx_mgr_entity_fini(struct amdgpu_ctx_mgr *mgr)
+{
+	struct amdgpu_ctx *ctx;
+	struct idr *idp;
+	uint32_t id, i;
+
+	idp = &mgr->ctx_handles;
+
+	idr_for_each_entry(idp, ctx, id) {
+
+		if (!ctx->adev)
+			return;
+
+		for (i = 0; i < ctx->adev->num_rings; i++)
+			if (kref_read(&ctx->refcount) == 1)
+				drm_sched_entity_do_release(&ctx->adev->rings[i]->sched,
+						  &ctx->rings[i].entity);
+			else
+				DRM_ERROR("ctx %p is still alive\n", ctx);
+	}
+}
+
+void amdgpu_ctx_mgr_entity_cleanup(struct amdgpu_ctx_mgr *mgr)
+{
+	struct amdgpu_ctx *ctx;
+	struct idr *idp;
+	uint32_t id, i;
+
+	idp = &mgr->ctx_handles;
+
+	idr_for_each_entry(idp, ctx, id) {
+
+		if (!ctx->adev)
+			return;
+
+		for (i = 0; i < ctx->adev->num_rings; i++)
+			if (kref_read(&ctx->refcount) == 1)
+				drm_sched_entity_cleanup(&ctx->adev->rings[i]->sched,
+					&ctx->rings[i].entity);
+			else
+				DRM_ERROR("ctx %p is still alive\n", ctx);
+	}
+}
+
 void amdgpu_ctx_mgr_fini(struct amdgpu_ctx_mgr *mgr)
 {
 	struct amdgpu_ctx *ctx;
 	struct idr *idp;
 	uint32_t id;
 
+	amdgpu_ctx_mgr_entity_cleanup(mgr);
+
 	idp = &mgr->ctx_handles;
 
 	idr_for_each_entry(idp, ctx, id) {
-		if (kref_put(&ctx->refcount, amdgpu_ctx_do_release) != 1)
+		if (kref_put(&ctx->refcount, amdgpu_ctx_fini) != 1)
 			DRM_ERROR("ctx %p is still alive\n", ctx);
 	}
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
index bd9e723..1ed3795 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
@@ -913,8 +913,7 @@ void amdgpu_driver_postclose_kms(struct drm_device *dev,
 		return;
 
 	pm_runtime_get_sync(dev->dev);
-
-	amdgpu_ctx_mgr_fini(&fpriv->ctx_mgr);
+	amdgpu_ctx_mgr_entity_fini(&fpriv->ctx_mgr);
 
 	if (adev->asic_type != CHIP_RAVEN) {
 		amdgpu_uvd_free_handles(adev, file_priv);
@@ -935,6 +934,8 @@ void amdgpu_driver_postclose_kms(struct drm_device *dev,
 	pd = amdgpu_bo_ref(fpriv->vm.root.base.bo);
 
 	amdgpu_vm_fini(adev, &fpriv->vm);
+	amdgpu_ctx_mgr_fini(&fpriv->ctx_mgr);
+
 	if (pasid)
 		amdgpu_pasid_free_delayed(pd->tbo.resv, pasid);
 	amdgpu_bo_unref(&pd);
diff --git a/drivers/gpu/drm/scheduler/gpu_scheduler.c b/drivers/gpu/drm/scheduler/gpu_scheduler.c
index 310275e..44d2198 100644
--- a/drivers/gpu/drm/scheduler/gpu_scheduler.c
+++ b/drivers/gpu/drm/scheduler/gpu_scheduler.c
@@ -136,6 +136,8 @@ int drm_sched_entity_init(struct drm_gpu_scheduler *sched,
 	entity->rq = rq;
 	entity->sched = sched;
 	entity->guilty = guilty;
+	entity->fini_status = 0;
+	entity->last_scheduled = NULL;
 
 	spin_lock_init(&entity->rq_lock);
 	spin_lock_init(&entity->queue_lock);
@@ -197,19 +199,30 @@ static bool drm_sched_entity_is_ready(struct drm_sched_entity *entity)
 	return true;
 }
 
+static void drm_sched_entity_kill_jobs_cb(struct dma_fence *f,
+				    struct dma_fence_cb *cb)
+{
+	struct drm_sched_job *job = container_of(cb, struct drm_sched_job,
+						 finish_cb);
+	drm_sched_fence_finished(job->s_fence);
+	WARN_ON(job->s_fence->parent);
+	dma_fence_put(&job->s_fence->finished);
+	job->sched->ops->free_job(job);
+}
+
+
 /**
  * Destroy a context entity
  *
  * @sched       Pointer to scheduler instance
  * @entity	The pointer to a valid scheduler entity
  *
- * Cleanup and free the allocated resources.
+ * Splitting drm_sched_entity_fini() into two functions, The first one is does the waiting,
+ * removes the entity from the runqueue and returns an error when the process was killed.
  */
-void drm_sched_entity_fini(struct drm_gpu_scheduler *sched,
+void drm_sched_entity_do_release(struct drm_gpu_scheduler *sched,
 			   struct drm_sched_entity *entity)
 {
-	int r;
-
 	if (!drm_sched_entity_is_initialized(sched, entity))
 		return;
 	/**
@@ -217,13 +230,28 @@ void drm_sched_entity_fini(struct drm_gpu_scheduler *sched,
 	 * queued IBs or discard them on SIGKILL
 	*/
 	if ((current->flags & PF_SIGNALED) && current->exit_code == SIGKILL)
-		r = -ERESTARTSYS;
+		entity->fini_status = -ERESTARTSYS;
 	else
-		r = wait_event_killable(sched->job_scheduled,
+		entity->fini_status = wait_event_killable(sched->job_scheduled,
 					drm_sched_entity_is_idle(entity));
 	drm_sched_entity_set_rq(entity, NULL);
-	if (r) {
+}
+EXPORT_SYMBOL(drm_sched_entity_do_release);
+
+/**
+ * Destroy a context entity
+ *
+ * @sched       Pointer to scheduler instance
+ * @entity	The pointer to a valid scheduler entity
+ *
+ * The second one then goes over the entity and signals all jobs with an error code.
+ */
+void drm_sched_entity_cleanup(struct drm_gpu_scheduler *sched,
+			   struct drm_sched_entity *entity)
+{
+	if (entity->fini_status) {
 		struct drm_sched_job *job;
+		int r;
 
 		/* Park the kernel for a moment to make sure it isn't processing
 		 * our enity.
@@ -241,13 +269,26 @@ void drm_sched_entity_fini(struct drm_gpu_scheduler *sched,
 			struct drm_sched_fence *s_fence = job->s_fence;
 			drm_sched_fence_scheduled(s_fence);
 			dma_fence_set_error(&s_fence->finished, -ESRCH);
-			drm_sched_fence_finished(s_fence);
-			WARN_ON(s_fence->parent);
-			dma_fence_put(&s_fence->finished);
-			sched->ops->free_job(job);
+			r = dma_fence_add_callback(entity->last_scheduled, &job->finish_cb,
+							drm_sched_entity_kill_jobs_cb);
+			if (r == -ENOENT)
+				drm_sched_entity_kill_jobs_cb(NULL, &job->finish_cb);
+			else if (r)
+				DRM_ERROR("fence add callback failed (%d)\n", r);
 		}
+
+		dma_fence_put(entity->last_scheduled);
+		entity->last_scheduled = NULL;
 	}
 }
+EXPORT_SYMBOL(drm_sched_entity_cleanup);
+
+void drm_sched_entity_fini(struct drm_gpu_scheduler *sched,
+				struct drm_sched_entity *entity)
+{
+	drm_sched_entity_do_release(sched, entity);
+	drm_sched_entity_cleanup(sched, entity);
+}
 EXPORT_SYMBOL(drm_sched_entity_fini);
 
 static void drm_sched_entity_wakeup(struct dma_fence *f, struct dma_fence_cb *cb)
@@ -530,6 +571,10 @@ void drm_sched_job_recovery(struct drm_gpu_scheduler *sched)
 		spin_unlock(&sched->job_list_lock);
 		fence = sched->ops->run_job(s_job);
 		atomic_inc(&sched->hw_rq_count);
+
+		dma_fence_put(s_job->entity->last_scheduled);
+		s_job->entity->last_scheduled = dma_fence_get(&s_fence->finished);
+
 		if (fence) {
 			s_fence->parent = dma_fence_get(fence);
 			r = dma_fence_add_callback(fence, &s_fence->cb,
@@ -556,6 +601,7 @@ int drm_sched_job_init(struct drm_sched_job *job,
 		       void *owner)
 {
 	job->sched = sched;
+	job->entity = entity;
 	job->s_priority = entity->rq - sched->sched_rq;
 	job->s_fence = drm_sched_fence_create(entity, owner);
 	if (!job->s_fence)
@@ -669,6 +715,9 @@ static int drm_sched_main(void *param)
 		fence = sched->ops->run_job(sched_job);
 		drm_sched_fence_scheduled(s_fence);
 
+		dma_fence_put(entity->last_scheduled);
+		entity->last_scheduled = dma_fence_get(&s_fence->finished);
+
 		if (fence) {
 			s_fence->parent = dma_fence_get(fence);
 			r = dma_fence_add_callback(fence, &s_fence->cb,
diff --git a/include/drm/gpu_scheduler.h b/include/drm/gpu_scheduler.h
index c053a32..350a62c 100644
--- a/include/drm/gpu_scheduler.h
+++ b/include/drm/gpu_scheduler.h
@@ -65,6 +65,8 @@ struct drm_sched_entity {
 	struct dma_fence		*dependency;
 	struct dma_fence_cb		cb;
 	atomic_t			*guilty; /* points to ctx's guilty */
+	int            fini_status;
+	struct dma_fence    *last_scheduled;
 };
 
 /**
@@ -119,6 +121,7 @@ struct drm_sched_job {
 	uint64_t			id;
 	atomic_t			karma;
 	enum drm_sched_priority		s_priority;
+	struct drm_sched_entity  *entity;
 };
 
 static inline bool drm_sched_invalidate_job(struct drm_sched_job *s_job,
@@ -186,6 +189,10 @@ int drm_sched_entity_init(struct drm_gpu_scheduler *sched,
 			  struct drm_sched_entity *entity,
 			  struct drm_sched_rq *rq,
 			  uint32_t jobs, atomic_t *guilty);
+void drm_sched_entity_do_release(struct drm_gpu_scheduler *sched,
+			   struct drm_sched_entity *entity);
+void drm_sched_entity_cleanup(struct drm_gpu_scheduler *sched,
+			   struct drm_sched_entity *entity);
 void drm_sched_entity_fini(struct drm_gpu_scheduler *sched,
 			   struct drm_sched_entity *entity);
 void drm_sched_entity_push_job(struct drm_sched_job *sched_job,
-- 
cgit v1.1


From a0701722b68e69443dd3dd7970a9f343b7560a2c Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Wed, 4 Apr 2018 14:11:45 +0800
Subject: Revert "drm/amd/powerply: fix power reading on Fiji"

we don't have limit of [50ms, 4sec] sampling period.
smu calculate average gpu power in real time.
we can read average gpu power through smu message or
read special register.

This reverts commit 462d8dcc9fec0d89f1ff6a1f93f1d4f670878c71.

Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index ed43dd3..5bccf89 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -3364,8 +3364,7 @@ static int smu7_get_gpu_power(struct pp_hwmgr *hwmgr,
 			"Failed to start pm status log!",
 			return -1);
 
-	/* Sampling period from 50ms to 4sec */
-	msleep_interruptible(200);
+	msleep_interruptible(20);
 
 	PP_ASSERT_WITH_CODE(!smum_send_msg_to_smc(hwmgr,
 			PPSMC_MSG_PmStatusLogSample),
-- 
cgit v1.1


From b89c71d1eb1c43c6c61f6d74d7454702d367f18b Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Wed, 4 Apr 2018 14:17:09 +0800
Subject: drm/amd/pp: Refine get_gpu_power for VI

pkgpwr is the average gpu power of 100ms. it is calculated by
firmware in real time.

1. we can send smu message PPSMC_MSG_GetCurrPkgPwr to read currentpkgpwr directly.

2. On Fiji/tonga/bonaire/hawwii, without PPSMC_MSG_GetCurrPkgPwr support.
   Send PPSMC_MSG_PmStatusLogStart/Sample to let smu write currentpkgpwr
   to ixSMU_PM_STATUS_94. driver can read pkgpwr from ixSMU_PM_STATUS_94.

Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c   | 51 ++++++++++++----------
 drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c | 10 +++--
 2 files changed, 34 insertions(+), 27 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index 5bccf89..51867c7 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -3359,30 +3359,33 @@ static int smu7_get_pp_table_entry(struct pp_hwmgr *hwmgr,
 static int smu7_get_gpu_power(struct pp_hwmgr *hwmgr,
 		struct pp_gpu_power *query)
 {
-	PP_ASSERT_WITH_CODE(!smum_send_msg_to_smc(hwmgr,
-			PPSMC_MSG_PmStatusLogStart),
-			"Failed to start pm status log!",
-			return -1);
-
-	msleep_interruptible(20);
-
-	PP_ASSERT_WITH_CODE(!smum_send_msg_to_smc(hwmgr,
-			PPSMC_MSG_PmStatusLogSample),
-			"Failed to sample pm status log!",
-			return -1);
-
-	query->vddc_power = cgs_read_ind_register(hwmgr->device,
-			CGS_IND_REG__SMC,
-			ixSMU_PM_STATUS_40);
-	query->vddci_power = cgs_read_ind_register(hwmgr->device,
-			CGS_IND_REG__SMC,
-			ixSMU_PM_STATUS_49);
-	query->max_gpu_power = cgs_read_ind_register(hwmgr->device,
-			CGS_IND_REG__SMC,
-			ixSMU_PM_STATUS_94);
-	query->average_gpu_power = cgs_read_ind_register(hwmgr->device,
-			CGS_IND_REG__SMC,
-			ixSMU_PM_STATUS_95);
+	int i;
+
+	if (!query)
+		return -EINVAL;
+
+
+	memset(query, 0, sizeof *query);
+
+	smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_GetCurrPkgPwr, 0);
+	query->average_gpu_power = cgs_read_register(hwmgr->device, mmSMC_MSG_ARG_0);
+
+	if (query->average_gpu_power != 0)
+		return 0;
+
+	smum_send_msg_to_smc(hwmgr, PPSMC_MSG_PmStatusLogStart);
+	cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+							ixSMU_PM_STATUS_94, 0);
+
+	for (i = 0; i < 20; i++) {
+		mdelay(1);
+		smum_send_msg_to_smc(hwmgr, PPSMC_MSG_PmStatusLogSample);
+		query->average_gpu_power = cgs_read_ind_register(hwmgr->device,
+						CGS_IND_REG__SMC,
+						ixSMU_PM_STATUS_94);
+		if (query->average_gpu_power != 0)
+			break;
+	}
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
index fb32a3f..10a1123 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
@@ -171,8 +171,10 @@ int smu7_send_msg_to_smc(struct pp_hwmgr *hwmgr, uint16_t msg)
 
 	ret = PHM_READ_FIELD(hwmgr->device, SMC_RESP_0, SMC_RESP);
 
-	if (ret != 1)
-		pr_info("\n failed to send pre message %x ret is %d \n",  msg, ret);
+	if (ret == 0xFE)
+		pr_debug("last message was not supported\n");
+	else if (ret != 1)
+		pr_info("\n last message was failed ret is %d\n", ret);
 
 	cgs_write_register(hwmgr->device, mmSMC_MESSAGE_0, msg);
 
@@ -180,7 +182,9 @@ int smu7_send_msg_to_smc(struct pp_hwmgr *hwmgr, uint16_t msg)
 
 	ret = PHM_READ_FIELD(hwmgr->device, SMC_RESP_0, SMC_RESP);
 
-	if (ret != 1)
+	if (ret == 0xFE)
+		pr_debug("message %x was not supported\n", msg);
+	else if (ret != 1)
 		pr_info("\n failed to send message %x ret is %d \n",  msg, ret);
 
 	return 0;
-- 
cgit v1.1


From 5b79d0482f3c1e8d5d78bd573a41e91dd9f0a5a1 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Wed, 4 Apr 2018 15:37:35 +0800
Subject: drm/amd/pp: Remove struct pp_gpu_power

Currently smu only calculate average gpu power in real time.

for vddc/vddci/max power,
User need to set start time and end time, firmware can calculate
the average vddc/vddci/max power. but the type of return values
is not unified. For Vi, return type is uint.
For vega, return type is float.

so this struct can't be suitable for all asics.

Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c            |  7 ++-----
 drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c             | 22 +++++++--------------
 drivers/gpu/drm/amd/include/kgd_pp_interface.h     |  7 -------
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c   | 23 +++++++++-------------
 drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c | 17 +++++++---------
 drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c | 13 ++++--------
 6 files changed, 29 insertions(+), 60 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
index 1ed3795..efff211 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
@@ -701,9 +701,6 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 		}
 	}
 	case AMDGPU_INFO_SENSOR: {
-		struct pp_gpu_power query = {0};
-		int query_size = sizeof(query);
-
 		if (!adev->pm.dpm_enabled)
 			return -ENOENT;
 
@@ -746,10 +743,10 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 			/* get average GPU power */
 			if (amdgpu_dpm_read_sensor(adev,
 						   AMDGPU_PP_SENSOR_GPU_POWER,
-						   (void *)&query, &query_size)) {
+						   (void *)&ui32, &ui32_size)) {
 				return -EINVAL;
 			}
-			ui32 = query.average_gpu_power >> 8;
+			ui32 >>= 8;
 			break;
 		case AMDGPU_INFO_SENSOR_VDDNB:
 			/* get VDDNB in millivolts */
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
index e5f60fc..744f105 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
@@ -1020,8 +1020,8 @@ static ssize_t amdgpu_hwmon_show_power_avg(struct device *dev,
 {
 	struct amdgpu_device *adev = dev_get_drvdata(dev);
 	struct drm_device *ddev = adev->ddev;
-	struct pp_gpu_power query = {0};
-	int r, size = sizeof(query);
+	u32 query = 0;
+	int r, size = sizeof(u32);
 	unsigned uw;
 
 	/* Can't get power when the card is off */
@@ -1041,7 +1041,7 @@ static ssize_t amdgpu_hwmon_show_power_avg(struct device *dev,
 		return r;
 
 	/* convert to microwatts */
-	uw = (query.average_gpu_power >> 8) * 1000000;
+	uw = (query >> 8) * 1000000 + (query & 0xff) * 1000;
 
 	return snprintf(buf, PAGE_SIZE, "%u\n", uw);
 }
@@ -1752,7 +1752,7 @@ void amdgpu_pm_compute_clocks(struct amdgpu_device *adev)
 static int amdgpu_debugfs_pm_info_pp(struct seq_file *m, struct amdgpu_device *adev)
 {
 	uint32_t value;
-	struct pp_gpu_power query = {0};
+	uint32_t query = 0;
 	int size;
 
 	/* sanity check PP is enabled */
@@ -1775,17 +1775,9 @@ static int amdgpu_debugfs_pm_info_pp(struct seq_file *m, struct amdgpu_device *a
 		seq_printf(m, "\t%u mV (VDDGFX)\n", value);
 	if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_VDDNB, (void *)&value, &size))
 		seq_printf(m, "\t%u mV (VDDNB)\n", value);
-	size = sizeof(query);
-	if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GPU_POWER, (void *)&query, &size)) {
-		seq_printf(m, "\t%u.%u W (VDDC)\n", query.vddc_power >> 8,
-				query.vddc_power & 0xff);
-		seq_printf(m, "\t%u.%u W (VDDCI)\n", query.vddci_power >> 8,
-				query.vddci_power & 0xff);
-		seq_printf(m, "\t%u.%u W (max GPU)\n", query.max_gpu_power >> 8,
-				query.max_gpu_power & 0xff);
-		seq_printf(m, "\t%u.%u W (average GPU)\n", query.average_gpu_power >> 8,
-				query.average_gpu_power & 0xff);
-	}
+	size = sizeof(uint32_t);
+	if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GPU_POWER, (void *)&query, &size))
+		seq_printf(m, "\t%u.%u W (average GPU)\n", query >> 8, query & 0xff);
 	size = sizeof(value);
 	seq_printf(m, "\n");
 
diff --git a/drivers/gpu/drm/amd/include/kgd_pp_interface.h b/drivers/gpu/drm/amd/include/kgd_pp_interface.h
index 5c840c0..1bec907 100644
--- a/drivers/gpu/drm/amd/include/kgd_pp_interface.h
+++ b/drivers/gpu/drm/amd/include/kgd_pp_interface.h
@@ -149,13 +149,6 @@ struct pp_states_info {
 	uint32_t states[16];
 };
 
-struct pp_gpu_power {
-	uint32_t vddc_power;
-	uint32_t vddci_power;
-	uint32_t max_gpu_power;
-	uint32_t average_gpu_power;
-};
-
 #define PP_GROUP_MASK        0xF0000000
 #define PP_GROUP_SHIFT       28
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index 51867c7..f5b3617 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -3356,36 +3356,34 @@ static int smu7_get_pp_table_entry(struct pp_hwmgr *hwmgr,
 	return 0;
 }
 
-static int smu7_get_gpu_power(struct pp_hwmgr *hwmgr,
-		struct pp_gpu_power *query)
+static int smu7_get_gpu_power(struct pp_hwmgr *hwmgr, u32 *query)
 {
 	int i;
+	u32 tmp = 0;
 
 	if (!query)
 		return -EINVAL;
 
-
-	memset(query, 0, sizeof *query);
-
 	smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_GetCurrPkgPwr, 0);
-	query->average_gpu_power = cgs_read_register(hwmgr->device, mmSMC_MSG_ARG_0);
+	tmp = cgs_read_register(hwmgr->device, mmSMC_MSG_ARG_0);
 
-	if (query->average_gpu_power != 0)
+	if (tmp != 0)
 		return 0;
 
 	smum_send_msg_to_smc(hwmgr, PPSMC_MSG_PmStatusLogStart);
 	cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
 							ixSMU_PM_STATUS_94, 0);
 
-	for (i = 0; i < 20; i++) {
+	for (i = 0; i < 10; i++) {
 		mdelay(1);
 		smum_send_msg_to_smc(hwmgr, PPSMC_MSG_PmStatusLogSample);
-		query->average_gpu_power = cgs_read_ind_register(hwmgr->device,
+		tmp = cgs_read_ind_register(hwmgr->device,
 						CGS_IND_REG__SMC,
 						ixSMU_PM_STATUS_94);
-		if (query->average_gpu_power != 0)
+		if (tmp != 0)
 			break;
 	}
+	*query = tmp;
 
 	return 0;
 }
@@ -3438,10 +3436,7 @@ static int smu7_read_sensor(struct pp_hwmgr *hwmgr, int idx,
 		*size = 4;
 		return 0;
 	case AMDGPU_PP_SENSOR_GPU_POWER:
-		if (*size < sizeof(struct pp_gpu_power))
-			return -EINVAL;
-		*size = sizeof(struct pp_gpu_power);
-		return smu7_get_gpu_power(hwmgr, (struct pp_gpu_power *)value);
+		return smu7_get_gpu_power(hwmgr, (uint32_t *)value);
 	case AMDGPU_PP_SENSOR_VDDGFX:
 		if ((data->vr_config & 0xff) == 0x2)
 			val_vid = PHM_READ_INDIRECT_FIELD(hwmgr->device,
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
index 127c550..0bbc564 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
@@ -3781,16 +3781,18 @@ static uint32_t vega10_dpm_get_mclk(struct pp_hwmgr *hwmgr, bool low)
 }
 
 static int vega10_get_gpu_power(struct pp_hwmgr *hwmgr,
-		struct pp_gpu_power *query)
+		uint32_t *query)
 {
 	uint32_t value;
 
+	if (!query)
+		return -EINVAL;
+
 	smum_send_msg_to_smc(hwmgr, PPSMC_MSG_GetCurrPkgPwr);
 	value = smum_get_argument(hwmgr);
 
-	/* power value is an integer */
-	memset(query, 0, sizeof *query);
-	query->average_gpu_power = value << 8;
+	/* SMC returning actual watts, keep consistent with legacy asics, low 8 bit as 8 fractional bits */
+	*query = value << 8;
 
 	return 0;
 }
@@ -3840,12 +3842,7 @@ static int vega10_read_sensor(struct pp_hwmgr *hwmgr, int idx,
 		*size = 4;
 		break;
 	case AMDGPU_PP_SENSOR_GPU_POWER:
-		if (*size < sizeof(struct pp_gpu_power))
-			ret = -EINVAL;
-		else {
-			*size = sizeof(struct pp_gpu_power);
-			ret = vega10_get_gpu_power(hwmgr, (struct pp_gpu_power *)value);
-		}
+		ret = vega10_get_gpu_power(hwmgr, (uint32_t *)value);
 		break;
 	case AMDGPU_PP_SENSOR_VDDGFX:
 		val_vid = (RREG32_SOC15(SMUIO, 0, mmSMUSVI0_PLANE0_CURRENTVID) &
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
index 3e1ed0a..782e209 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
@@ -1113,8 +1113,7 @@ static uint32_t vega12_dpm_get_mclk(struct pp_hwmgr *hwmgr, bool low)
 	return (mem_clk * 100);
 }
 
-static int vega12_get_gpu_power(struct pp_hwmgr *hwmgr,
-		struct pp_gpu_power *query)
+static int vega12_get_gpu_power(struct pp_hwmgr *hwmgr, uint32_t *query)
 {
 #if 0
 	uint32_t value;
@@ -1126,7 +1125,7 @@ static int vega12_get_gpu_power(struct pp_hwmgr *hwmgr,
 
 	vega12_read_arg_from_smc(hwmgr, &value);
 	/* power value is an integer */
-	query->average_gpu_power = value << 8;
+	*query = value << 8;
 #endif
 	return 0;
 }
@@ -1235,12 +1234,8 @@ static int vega12_read_sensor(struct pp_hwmgr *hwmgr, int idx,
 		*size = 4;
 		break;
 	case AMDGPU_PP_SENSOR_GPU_POWER:
-		if (*size < sizeof(struct pp_gpu_power))
-			ret = -EINVAL;
-		else {
-			*size = sizeof(struct pp_gpu_power);
-			ret = vega12_get_gpu_power(hwmgr, (struct pp_gpu_power *)value);
-		}
+		ret = vega12_get_gpu_power(hwmgr, (uint32_t *)value);
+
 		break;
 	default:
 		ret = -EINVAL;
-- 
cgit v1.1


From 8db42a701326c8872d8634c7b4c0d045bf95f394 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Wed, 11 Apr 2018 18:11:49 +0800
Subject: drm/amd/pp: Clear smu response register before send smu message

smu firmware do not update response register immediately under
some delay tasks, we may read out the original value.

so need to clear the register before send smu message.

Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c   | 4 +---
 drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c | 1 +
 2 files changed, 2 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c
index 2a93f3a..2d4ec8a 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c
@@ -208,9 +208,7 @@ static int ci_send_msg_to_smc(struct pp_hwmgr *hwmgr, uint16_t msg)
 {
 	int ret;
 
-	if (!ci_is_smc_ram_running(hwmgr))
-		return -EINVAL;
-
+	cgs_write_register(hwmgr->device, mmSMC_RESP_0, 0);
 	cgs_write_register(hwmgr->device, mmSMC_MESSAGE_0, msg);
 
 	PHM_WAIT_FIELD_UNEQUAL(hwmgr, SMC_RESP_0, SMC_RESP, 0);
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
index 10a1123..64d33b7 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
@@ -176,6 +176,7 @@ int smu7_send_msg_to_smc(struct pp_hwmgr *hwmgr, uint16_t msg)
 	else if (ret != 1)
 		pr_info("\n last message was failed ret is %d\n", ret);
 
+	cgs_write_register(hwmgr->device, mmSMC_RESP_0, 0);
 	cgs_write_register(hwmgr->device, mmSMC_MESSAGE_0, msg);
 
 	PHM_WAIT_FIELD_UNEQUAL(hwmgr, SMC_RESP_0, SMC_RESP, 0);
-- 
cgit v1.1


From 63c2f7ed7bb3e98b4b22d5b136f4749706f17d36 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Sun, 8 Apr 2018 16:57:55 +0800
Subject: drm/amd/pp: Move common code to smu_helper.c

Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c   | 30 +-------
 drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.c   | 82 ++++++++++++++++++++++
 drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.h   | 24 +++++++
 drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c | 43 +-----------
 4 files changed, 109 insertions(+), 70 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index f5b3617..68aae09 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -793,32 +793,6 @@ static int smu7_setup_dpm_tables_v1(struct pp_hwmgr *hwmgr)
 	return 0;
 }
 
-static int smu7_get_voltage_dependency_table(
-			const struct phm_ppt_v1_clock_voltage_dependency_table *allowed_dep_table,
-			struct phm_ppt_v1_clock_voltage_dependency_table *dep_table)
-{
-	uint8_t i = 0;
-	PP_ASSERT_WITH_CODE((0 != allowed_dep_table->count),
-				"Voltage Lookup Table empty",
-				return -EINVAL);
-
-	dep_table->count = allowed_dep_table->count;
-	for (i=0; i<dep_table->count; i++) {
-		dep_table->entries[i].clk = allowed_dep_table->entries[i].clk;
-		dep_table->entries[i].vddInd = allowed_dep_table->entries[i].vddInd;
-		dep_table->entries[i].vdd_offset = allowed_dep_table->entries[i].vdd_offset;
-		dep_table->entries[i].vddc = allowed_dep_table->entries[i].vddc;
-		dep_table->entries[i].vddgfx = allowed_dep_table->entries[i].vddgfx;
-		dep_table->entries[i].vddci = allowed_dep_table->entries[i].vddci;
-		dep_table->entries[i].mvdd = allowed_dep_table->entries[i].mvdd;
-		dep_table->entries[i].phases = allowed_dep_table->entries[i].phases;
-		dep_table->entries[i].cks_enable = allowed_dep_table->entries[i].cks_enable;
-		dep_table->entries[i].cks_voffset = allowed_dep_table->entries[i].cks_voffset;
-	}
-
-	return 0;
-}
-
 static int smu7_odn_initial_default_setting(struct pp_hwmgr *hwmgr)
 {
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
@@ -846,7 +820,7 @@ static int smu7_odn_initial_default_setting(struct pp_hwmgr *hwmgr)
 		entries[i].vddc = dep_sclk_table->entries[i].vddc;
 	}
 
-	smu7_get_voltage_dependency_table(dep_sclk_table,
+	smu_get_voltage_dependency_table_ppt_v1(dep_sclk_table,
 		(struct phm_ppt_v1_clock_voltage_dependency_table *)&(odn_table->vdd_dependency_on_sclk));
 
 	odn_table->odn_memory_clock_dpm_levels.num_of_pl =
@@ -858,7 +832,7 @@ static int smu7_odn_initial_default_setting(struct pp_hwmgr *hwmgr)
 		entries[i].vddc = dep_mclk_table->entries[i].vddc;
 	}
 
-	smu7_get_voltage_dependency_table(dep_mclk_table,
+	smu_get_voltage_dependency_table_ppt_v1(dep_mclk_table,
 		(struct phm_ppt_v1_clock_voltage_dependency_table *)&(odn_table->vdd_dependency_on_mclk));
 
 	return 0;
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.c
index 529be3c..7c23741 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.c
@@ -624,3 +624,85 @@ void *smu_atom_get_data_table(void *dev, uint32_t table, uint16_t *size,
 
 	return NULL;
 }
+
+int smu_get_voltage_dependency_table_ppt_v1(
+			const struct phm_ppt_v1_clock_voltage_dependency_table *allowed_dep_table,
+			struct phm_ppt_v1_clock_voltage_dependency_table *dep_table)
+{
+	uint8_t i = 0;
+	PP_ASSERT_WITH_CODE((0 != allowed_dep_table->count),
+				"Voltage Lookup Table empty",
+				return -EINVAL);
+
+	dep_table->count = allowed_dep_table->count;
+	for (i=0; i<dep_table->count; i++) {
+		dep_table->entries[i].clk = allowed_dep_table->entries[i].clk;
+		dep_table->entries[i].vddInd = allowed_dep_table->entries[i].vddInd;
+		dep_table->entries[i].vdd_offset = allowed_dep_table->entries[i].vdd_offset;
+		dep_table->entries[i].vddc = allowed_dep_table->entries[i].vddc;
+		dep_table->entries[i].vddgfx = allowed_dep_table->entries[i].vddgfx;
+		dep_table->entries[i].vddci = allowed_dep_table->entries[i].vddci;
+		dep_table->entries[i].mvdd = allowed_dep_table->entries[i].mvdd;
+		dep_table->entries[i].phases = allowed_dep_table->entries[i].phases;
+		dep_table->entries[i].cks_enable = allowed_dep_table->entries[i].cks_enable;
+		dep_table->entries[i].cks_voffset = allowed_dep_table->entries[i].cks_voffset;
+	}
+
+	return 0;
+}
+
+int smu_set_watermarks_for_clocks_ranges(void *wt_table,
+		struct pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges)
+{
+	uint32_t i;
+	struct watermarks *table = wt_table;
+
+	if (!table || wm_with_clock_ranges)
+		return -EINVAL;
+
+	if (wm_with_clock_ranges->num_wm_sets_dmif > 4 || wm_with_clock_ranges->num_wm_sets_mcif > 4)
+		return -EINVAL;
+
+	for (i = 0; i < wm_with_clock_ranges->num_wm_sets_dmif; i++) {
+		table->WatermarkRow[1][i].MinClock =
+			cpu_to_le16((uint16_t)
+			(wm_with_clock_ranges->wm_sets_dmif[i].wm_min_dcefclk_in_khz) /
+			100);
+		table->WatermarkRow[1][i].MaxClock =
+			cpu_to_le16((uint16_t)
+			(wm_with_clock_ranges->wm_sets_dmif[i].wm_max_dcefclk_in_khz) /
+			100);
+		table->WatermarkRow[1][i].MinUclk =
+			cpu_to_le16((uint16_t)
+			(wm_with_clock_ranges->wm_sets_dmif[i].wm_min_memclk_in_khz) /
+			100);
+		table->WatermarkRow[1][i].MaxUclk =
+			cpu_to_le16((uint16_t)
+			(wm_with_clock_ranges->wm_sets_dmif[i].wm_max_memclk_in_khz) /
+			100);
+		table->WatermarkRow[1][i].WmSetting = (uint8_t)
+				wm_with_clock_ranges->wm_sets_dmif[i].wm_set_id;
+	}
+
+	for (i = 0; i < wm_with_clock_ranges->num_wm_sets_mcif; i++) {
+		table->WatermarkRow[0][i].MinClock =
+			cpu_to_le16((uint16_t)
+			(wm_with_clock_ranges->wm_sets_mcif[i].wm_min_socclk_in_khz) /
+			100);
+		table->WatermarkRow[0][i].MaxClock =
+			cpu_to_le16((uint16_t)
+			(wm_with_clock_ranges->wm_sets_mcif[i].wm_max_socclk_in_khz) /
+			100);
+		table->WatermarkRow[0][i].MinUclk =
+			cpu_to_le16((uint16_t)
+			(wm_with_clock_ranges->wm_sets_mcif[i].wm_min_memclk_in_khz) /
+			100);
+		table->WatermarkRow[0][i].MaxUclk =
+			cpu_to_le16((uint16_t)
+			(wm_with_clock_ranges->wm_sets_mcif[i].wm_max_memclk_in_khz) /
+			100);
+		table->WatermarkRow[0][i].WmSetting = (uint8_t)
+				wm_with_clock_ranges->wm_sets_mcif[i].wm_set_id;
+	}
+	return 0;
+}
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.h b/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.h
index 14ee162..916cc01 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.h
@@ -26,10 +26,27 @@
 struct pp_atomctrl_voltage_table;
 struct pp_hwmgr;
 struct phm_ppt_v1_voltage_lookup_table;
+struct Watermarks_t;
+struct pp_wm_sets_with_clock_ranges_soc15;
 
 uint8_t convert_to_vid(uint16_t vddc);
 uint16_t convert_to_vddc(uint8_t vid);
 
+struct watermark_row_generic_t {
+	uint16_t MinClock;
+	uint16_t MaxClock;
+	uint16_t MinUclk;
+	uint16_t MaxUclk;
+
+	uint8_t  WmSetting;
+	uint8_t  Padding[3];
+};
+
+struct watermarks {
+	struct watermark_row_generic_t WatermarkRow[2][4];
+	uint32_t     padding[7];
+};
+
 extern int phm_wait_for_register_unequal(struct pp_hwmgr *hwmgr,
 					uint32_t index,
 					uint32_t value, uint32_t mask);
@@ -85,6 +102,13 @@ int smu9_register_irq_handlers(struct pp_hwmgr *hwmgr);
 void *smu_atom_get_data_table(void *dev, uint32_t table, uint16_t *size,
 						uint8_t *frev, uint8_t *crev);
 
+int smu_get_voltage_dependency_table_ppt_v1(
+	const struct phm_ppt_v1_clock_voltage_dependency_table *allowed_dep_table,
+		struct phm_ppt_v1_clock_voltage_dependency_table *dep_table);
+
+int smu_set_watermarks_for_clocks_ranges(void *wt_table,
+		struct pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges);
+
 #define PHM_FIELD_SHIFT(reg, field) reg##__##field##__SHIFT
 #define PHM_FIELD_MASK(reg, field) reg##__##field##_MASK
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
index 0bbc564..384aa07 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
@@ -4367,50 +4367,9 @@ static int vega10_set_watermarks_for_clocks_ranges(struct pp_hwmgr *hwmgr,
 	struct vega10_hwmgr *data = hwmgr->backend;
 	Watermarks_t *table = &(data->smc_state_table.water_marks_table);
 	int result = 0;
-	uint32_t i;
 
 	if (!data->registry_data.disable_water_mark) {
-		for (i = 0; i < wm_with_clock_ranges->num_wm_sets_dmif; i++) {
-			table->WatermarkRow[WM_DCEFCLK][i].MinClock =
-				cpu_to_le16((uint16_t)
-				(wm_with_clock_ranges->wm_sets_dmif[i].wm_min_dcefclk_in_khz) /
-				100);
-			table->WatermarkRow[WM_DCEFCLK][i].MaxClock =
-				cpu_to_le16((uint16_t)
-				(wm_with_clock_ranges->wm_sets_dmif[i].wm_max_dcefclk_in_khz) /
-				100);
-			table->WatermarkRow[WM_DCEFCLK][i].MinUclk =
-				cpu_to_le16((uint16_t)
-				(wm_with_clock_ranges->wm_sets_dmif[i].wm_min_memclk_in_khz) /
-				100);
-			table->WatermarkRow[WM_DCEFCLK][i].MaxUclk =
-				cpu_to_le16((uint16_t)
-				(wm_with_clock_ranges->wm_sets_dmif[i].wm_max_memclk_in_khz) /
-				100);
-			table->WatermarkRow[WM_DCEFCLK][i].WmSetting = (uint8_t)
-					wm_with_clock_ranges->wm_sets_dmif[i].wm_set_id;
-		}
-
-		for (i = 0; i < wm_with_clock_ranges->num_wm_sets_mcif; i++) {
-			table->WatermarkRow[WM_SOCCLK][i].MinClock =
-				cpu_to_le16((uint16_t)
-				(wm_with_clock_ranges->wm_sets_mcif[i].wm_min_socclk_in_khz) /
-				100);
-			table->WatermarkRow[WM_SOCCLK][i].MaxClock =
-				cpu_to_le16((uint16_t)
-				(wm_with_clock_ranges->wm_sets_mcif[i].wm_max_socclk_in_khz) /
-				100);
-			table->WatermarkRow[WM_SOCCLK][i].MinUclk =
-				cpu_to_le16((uint16_t)
-				(wm_with_clock_ranges->wm_sets_mcif[i].wm_min_memclk_in_khz) /
-				100);
-			table->WatermarkRow[WM_SOCCLK][i].MaxUclk =
-				cpu_to_le16((uint16_t)
-				(wm_with_clock_ranges->wm_sets_mcif[i].wm_max_memclk_in_khz) /
-				100);
-			table->WatermarkRow[WM_SOCCLK][i].WmSetting = (uint8_t)
-					wm_with_clock_ranges->wm_sets_mcif[i].wm_set_id;
-		}
+		smu_set_watermarks_for_clocks_ranges(table, wm_with_clock_ranges);
 		data->water_marks_bitmap = WaterMarksExist;
 	}
 
-- 
cgit v1.1


From 1afd30efeddbb1b32cf35d3bf6477b35690eeca6 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Christian=20K=C3=B6nig?= <christian.koenig@amd.com>
Date: Tue, 10 Apr 2018 13:42:29 +0200
Subject: drm/amdgpu: revert "add new bo flag that indicates BOs don't need
 fallback (v2)"
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This reverts commit 6f51d28bfe8e1a676de5cd877639245bed3cc818.

Makes fallback handling to complicated. This is just a feature for the
GEM interface and shouldn't leak into the core BO create function.

Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c     | 3 +--
 drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 5 +----
 include/uapi/drm/amdgpu_drm.h              | 2 --
 3 files changed, 2 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
index 68af2f8..e1756b6 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
@@ -385,8 +385,7 @@ retry:
 	    amdgpu_bo_in_cpu_visible_vram(bo))
 		p->bytes_moved_vis += ctx.bytes_moved;
 
-	if (unlikely(r == -ENOMEM) && domain != bo->allowed_domains &&
-	    !(bo->flags & AMDGPU_GEM_CREATE_NO_FALLBACK)) {
+	if (unlikely(r == -ENOMEM) && domain != bo->allowed_domains) {
 		domain = bo->allowed_domains;
 		goto retry;
 	}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
index a160ef0..1de6864 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
@@ -388,8 +388,6 @@ retry:
 	drm_gem_private_object_init(adev->ddev, &bo->gem_base, size);
 	INIT_LIST_HEAD(&bo->shadow_list);
 	INIT_LIST_HEAD(&bo->va);
-	bo->preferred_domains = preferred_domains;
-	bo->allowed_domains = allowed_domains;
 
 	bo->flags = flags;
 
@@ -426,8 +424,7 @@ retry:
 	r = ttm_bo_init_reserved(&adev->mman.bdev, &bo->tbo, size, type,
 				 &bo->placement, page_align, &ctx, acc_size,
 				 NULL, resv, &amdgpu_ttm_bo_destroy);
-	if (unlikely(r && r != -ERESTARTSYS) && type == ttm_bo_type_device &&
-	    !(flags & AMDGPU_GEM_CREATE_NO_FALLBACK)) {
+	if (unlikely(r && r != -ERESTARTSYS) && type == ttm_bo_type_device) {
 		if (flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED) {
 			flags &= ~AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED;
 			goto retry;
diff --git a/include/uapi/drm/amdgpu_drm.h b/include/uapi/drm/amdgpu_drm.h
index 4f5a27d..c363b67 100644
--- a/include/uapi/drm/amdgpu_drm.h
+++ b/include/uapi/drm/amdgpu_drm.h
@@ -95,8 +95,6 @@ extern "C" {
 #define AMDGPU_GEM_CREATE_VM_ALWAYS_VALID	(1 << 6)
 /* Flag that BO sharing will be explicitly synchronized */
 #define AMDGPU_GEM_CREATE_EXPLICIT_SYNC		(1 << 7)
-/* Flag that BO doesn't need fallback */
-#define AMDGPU_GEM_CREATE_NO_FALLBACK		(1 << 8)
 
 struct drm_amdgpu_gem_create_in  {
 	/** the requested memory size */
-- 
cgit v1.1


From 0808210478c76606c12bb475b3272b7780240812 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Christian=20K=C3=B6nig?= <christian.koenig@amd.com>
Date: Tue, 10 Apr 2018 13:42:38 +0200
Subject: drm/amdgpu: revert "Don't change preferred domian when fallback GTT
 v6"
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This reverts commit 7d1ca1325260a9e9329b10a21e3692e6f188936f.

Makes fallback handling to complicated. This is just a feature for the
GEM interface and shouldn't leak into the core BO create function.

The intended change to preserve the preferred domains is implemented in
a follow up patch.

Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c    | 16 +++++++++++--
 drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 37 +++++++++++-------------------
 2 files changed, 27 insertions(+), 26 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
index 28c2706..46b9ea4 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
@@ -56,11 +56,23 @@ int amdgpu_gem_object_create(struct amdgpu_device *adev, unsigned long size,
 		alignment = PAGE_SIZE;
 	}
 
+retry:
 	r = amdgpu_bo_create(adev, size, alignment, initial_domain,
 			     flags, type, resv, &bo);
 	if (r) {
-		DRM_DEBUG("Failed to allocate GEM object (%ld, %d, %u, %d)\n",
-			  size, initial_domain, alignment, r);
+		if (r != -ERESTARTSYS) {
+			if (flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED) {
+				flags &= ~AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED;
+				goto retry;
+			}
+
+			if (initial_domain == AMDGPU_GEM_DOMAIN_VRAM) {
+				initial_domain |= AMDGPU_GEM_DOMAIN_GTT;
+				goto retry;
+			}
+			DRM_DEBUG("Failed to allocate GEM object (%ld, %d, %u, %d)\n",
+				  size, initial_domain, alignment, r);
+		}
 		return r;
 	}
 	*obj = &bo->gem_base;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
index 1de6864..24f582c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
@@ -356,7 +356,6 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev, unsigned long size,
 	struct amdgpu_bo *bo;
 	unsigned long page_align;
 	size_t acc_size;
-	u32 domains, preferred_domains, allowed_domains;
 	int r;
 
 	page_align = roundup(byte_align, PAGE_SIZE) >> PAGE_SHIFT;
@@ -370,24 +369,22 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev, unsigned long size,
 	acc_size = ttm_bo_dma_acc_size(&adev->mman.bdev, size,
 				       sizeof(struct amdgpu_bo));
 
-	preferred_domains = domain & (AMDGPU_GEM_DOMAIN_VRAM |
-				      AMDGPU_GEM_DOMAIN_GTT |
-				      AMDGPU_GEM_DOMAIN_CPU |
-				      AMDGPU_GEM_DOMAIN_GDS |
-				      AMDGPU_GEM_DOMAIN_GWS |
-				      AMDGPU_GEM_DOMAIN_OA);
-	allowed_domains = preferred_domains;
-	if (type != ttm_bo_type_kernel &&
-	    allowed_domains == AMDGPU_GEM_DOMAIN_VRAM)
-		allowed_domains |= AMDGPU_GEM_DOMAIN_GTT;
-	domains = preferred_domains;
-retry:
 	bo = kzalloc(sizeof(struct amdgpu_bo), GFP_KERNEL);
 	if (bo == NULL)
 		return -ENOMEM;
 	drm_gem_private_object_init(adev->ddev, &bo->gem_base, size);
 	INIT_LIST_HEAD(&bo->shadow_list);
 	INIT_LIST_HEAD(&bo->va);
+	bo->preferred_domains = domain & (AMDGPU_GEM_DOMAIN_VRAM |
+					 AMDGPU_GEM_DOMAIN_GTT |
+					 AMDGPU_GEM_DOMAIN_CPU |
+					 AMDGPU_GEM_DOMAIN_GDS |
+					 AMDGPU_GEM_DOMAIN_GWS |
+					 AMDGPU_GEM_DOMAIN_OA);
+	bo->allowed_domains = bo->preferred_domains;
+	if (type != ttm_bo_type_kernel &&
+	    bo->allowed_domains == AMDGPU_GEM_DOMAIN_VRAM)
+		bo->allowed_domains |= AMDGPU_GEM_DOMAIN_GTT;
 
 	bo->flags = flags;
 
@@ -420,20 +417,12 @@ retry:
 #endif
 
 	bo->tbo.bdev = &adev->mman.bdev;
-	amdgpu_ttm_placement_from_domain(bo, domains);
+	amdgpu_ttm_placement_from_domain(bo, domain);
+
 	r = ttm_bo_init_reserved(&adev->mman.bdev, &bo->tbo, size, type,
 				 &bo->placement, page_align, &ctx, acc_size,
 				 NULL, resv, &amdgpu_ttm_bo_destroy);
-	if (unlikely(r && r != -ERESTARTSYS) && type == ttm_bo_type_device) {
-		if (flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED) {
-			flags &= ~AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED;
-			goto retry;
-		} else if (domains != allowed_domains) {
-			domains = allowed_domains;
-			goto retry;
-		}
-	}
-	if (unlikely(r))
+	if (unlikely(r != 0))
 		return r;
 
 	if (adev->gmc.visible_vram_size < adev->gmc.real_vram_size &&
-- 
cgit v1.1


From 361883649221f975d915e4bc79907da71017f38f Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Christian=20K=C3=B6nig?= <christian.koenig@amd.com>
Date: Mon, 19 Mar 2018 11:49:14 +0100
Subject: drm/amdgpu: re-validate per VM BOs if required v2
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

If a per VM BO ends up in a allowed domain it never moves back into the
prefered domain.

v2: move the extra handling into amdgpu_vm_bo_update when we exit the
    state machine. Make memory type handling generic.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 15 ++++++++++++++-
 1 file changed, 14 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index da55a78..f0fbc33 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -1556,7 +1556,20 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev,
 	}
 
 	spin_lock(&vm->status_lock);
-	list_del_init(&bo_va->base.vm_status);
+	if (bo && bo->tbo.resv == vm->root.base.bo->tbo.resv) {
+		unsigned mem_type = bo->tbo.mem.mem_type;
+
+		/* If the BO is not in its preferred location add it back to
+		 * the evicted list so that it gets validated again on the
+		 * next command submission.
+		 */
+		if (!(bo->preferred_domains & amdgpu_mem_type_to_domain(mem_type)))
+			list_add_tail(&bo_va->base.vm_status, &vm->evicted);
+		else
+			list_del_init(&bo_va->base.vm_status);
+	} else {
+		list_del_init(&bo_va->base.vm_status);
+	}
 	spin_unlock(&vm->status_lock);
 
 	list_splice_init(&bo_va->invalids, &bo_va->valids);
-- 
cgit v1.1


From 03a27de648d8a2b2bf59a7f467855fac2d850350 Mon Sep 17 00:00:00 2001
From: Mikita Lipski <mikita.lipski@amd.com>
Date: Tue, 10 Apr 2018 13:45:00 -0400
Subject: drm/amd/pp: Adding set_watermarks_for_clocks_ranges for SMU10

The function is never implemented for raven on linux.
It follows similair implementation as on windows.

SMU still needs to notify SMC and copy WM table, which is added
here. But on other Asics such as Vega this step is not implemented.

Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c | 13 +++++++++++++
 drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.h |  1 +
 2 files changed, 14 insertions(+)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
index 6ba3b1f..b712d16 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
@@ -992,6 +992,18 @@ static int smu10_read_sensor(struct pp_hwmgr *hwmgr, int idx,
 	return ret;
 }
 
+static int smu10_set_watermarks_for_clocks_ranges(struct pp_hwmgr *hwmgr,
+		struct pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges)
+{
+	struct smu10_hwmgr *data = hwmgr->backend;
+	Watermarks_t *table = &(data->water_marks_table);
+	int result = 0;
+
+	smu_set_watermarks_for_clocks_ranges(table,wm_with_clock_ranges);
+	smum_smc_table_manager(hwmgr, (uint8_t *)table, (uint16_t)SMU10_WMTABLE, false);
+	data->water_marks_exist = true;
+	return result;
+}
 static int smu10_set_mmhub_powergating_by_smu(struct pp_hwmgr *hwmgr)
 {
 	return smum_send_msg_to_smc(hwmgr, PPSMC_MSG_PowerGateMmHub);
@@ -1021,6 +1033,7 @@ static const struct pp_hwmgr_func smu10_hwmgr_funcs = {
 	.get_current_shallow_sleep_clocks = smu10_get_current_shallow_sleep_clocks,
 	.get_clock_by_type_with_latency = smu10_get_clock_by_type_with_latency,
 	.get_clock_by_type_with_voltage = smu10_get_clock_by_type_with_voltage,
+	.set_watermarks_for_clocks_ranges = smu10_set_watermarks_for_clocks_ranges,
 	.get_max_high_clocks = smu10_get_max_high_clocks,
 	.read_sensor = smu10_read_sensor,
 	.set_active_display_count = smu10_set_active_display_count,
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.h
index 175c3a5..f68b218 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.h
@@ -290,6 +290,7 @@ struct smu10_hwmgr {
 	bool                           vcn_dpg_mode;
 
 	bool                           gfx_off_controled_by_driver;
+	bool                           water_marks_exist;
 	Watermarks_t                      water_marks_table;
 	struct smu10_clock_voltage_information   clock_vol_info;
 	DpmClocks_t                       clock_table;
-- 
cgit v1.1


From 5c3517d0c2ead443f378173c698f3bd09cb89d72 Mon Sep 17 00:00:00 2001
From: Mikita Lipski <mikita.lipski@amd.com>
Date: Wed, 11 Apr 2018 16:25:26 -0400
Subject: drm/amd/pp: Adding a function to store cc6 data in SMU10

Filling the smu10_store_cc6_data based on the implementation
of Windows Powerplay.

There is an uncertainty with one of the parameters passed to the function
pstate_switch_disable - is not a part of smu10 private data structure.
So in the function its just ignored.

Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
index b712d16..0f25226 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
@@ -699,6 +699,16 @@ static int smu10_set_cpu_power_state(struct pp_hwmgr *hwmgr)
 static int smu10_store_cc6_data(struct pp_hwmgr *hwmgr, uint32_t separation_time,
 			bool cc6_disable, bool pstate_disable, bool pstate_switch_disable)
 {
+	struct smu10_hwmgr *data = (struct smu10_hwmgr *)(hwmgr->backend);
+
+	if (separation_time != data->separation_time ||
+			cc6_disable != data->cc6_disable ||
+			pstate_disable != data->pstate_disable) {
+		data->separation_time = separation_time;
+		data->cc6_disable = cc6_disable;
+		data->pstate_disable = pstate_disable;
+		data->cc6_setting_changed = true;
+	}
 	return 0;
 }
 
-- 
cgit v1.1


From 2c773de2ecb8c327f2448bd1eecad224e9227087 Mon Sep 17 00:00:00 2001
From: Shirish S <shirish.s@amd.com>
Date: Mon, 16 Apr 2018 12:17:57 +0530
Subject: drm/amdgpu: defer test IBs on the rings at boot (V3)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

amdgpu_ib_ring_tests() runs test IB's on rings at boot
contributes to ~500 ms of amdgpu driver's boot time.

This patch defers it and ensures that its executed
in amdgpu_info_ioctl() if it wasn't scheduled.

V2: Use queue_delayed_work() & flush_delayed_work().
V3: removed usage of separate wq, ensure ib tests is
    run before enabling clockgating.

Signed-off-by: Shirish S <shirish.s@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 17 ++++++-----------
 drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c    |  3 +++
 2 files changed, 9 insertions(+), 11 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 62d6505..d7f2bbd 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -1656,6 +1656,10 @@ static int amdgpu_device_ip_late_set_cg_state(struct amdgpu_device *adev)
 	if (amdgpu_emu_mode == 1)
 		return 0;
 
+	r = amdgpu_ib_ring_tests(adev);
+	if (r)
+		DRM_ERROR("ib ring test failed (%d).\n", r);
+
 	for (i = 0; i < adev->num_ip_blocks; i++) {
 		if (!adev->ip_blocks[i].status.valid)
 			continue;
@@ -1706,8 +1710,8 @@ static int amdgpu_device_ip_late_init(struct amdgpu_device *adev)
 		}
 	}
 
-	mod_delayed_work(system_wq, &adev->late_init_work,
-			msecs_to_jiffies(AMDGPU_RESUME_MS));
+	queue_delayed_work(system_wq, &adev->late_init_work,
+			   msecs_to_jiffies(AMDGPU_RESUME_MS));
 
 	amdgpu_device_fill_reset_magic(adev);
 
@@ -2374,10 +2378,6 @@ fence_driver_init:
 		goto failed;
 	}
 
-	r = amdgpu_ib_ring_tests(adev);
-	if (r)
-		DRM_ERROR("ib ring test failed (%d).\n", r);
-
 	if (amdgpu_sriov_vf(adev))
 		amdgpu_virt_init_data_exchange(adev);
 
@@ -2639,11 +2639,6 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon)
 	}
 	amdgpu_fence_driver_resume(adev);
 
-	if (resume) {
-		r = amdgpu_ib_ring_tests(adev);
-		if (r)
-			DRM_ERROR("ib ring test failed (%d).\n", r);
-	}
 
 	r = amdgpu_device_ip_late_init(adev);
 	if (r)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
index efff211..4e15b6f 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
@@ -279,6 +279,9 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 	if (!info->return_size || !info->return_pointer)
 		return -EINVAL;
 
+	/* Ensure IB tests are run on ring */
+	flush_delayed_work(&adev->late_init_work);
+
 	switch (info->query) {
 	case AMDGPU_INFO_ACCEL_WORKING:
 		ui32 = adev->accel_working;
-- 
cgit v1.1


From c2f84e03a01ad09f18f9f132f8b1e78f699a5494 Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Thu, 12 Apr 2018 16:37:09 -0400
Subject: drm/amd/display: Don't program bypass on linear regamma LUT

Even though this is required for degamma since DCE HW only supports a
couple predefined LUTs we can just program the LUT directly for regamma.

This fixes dark screens which occurs when we program regamma to bypass
while degamma is using srgb LUT.

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Leo Li <sunpeng.li@amd.com>
Cc: stable@vger.kernel.org
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c | 7 -------
 1 file changed, 7 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
index ef5fad8..e3d90e9 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
@@ -139,13 +139,6 @@ int amdgpu_dm_set_regamma_lut(struct dm_crtc_state *crtc)
 	lut = (struct drm_color_lut *)blob->data;
 	lut_size = blob->length / sizeof(struct drm_color_lut);
 
-	if (__is_lut_linear(lut, lut_size)) {
-		/* Set to bypass if lut is set to linear */
-		stream->out_transfer_func->type = TF_TYPE_BYPASS;
-		stream->out_transfer_func->tf = TRANSFER_FUNCTION_LINEAR;
-		return 0;
-	}
-
 	gamma = dc_create_gamma();
 	if (!gamma)
 		return -ENOMEM;
-- 
cgit v1.1


From c74db7e42d9b538d2fa582cf0efe5640b25e950d Mon Sep 17 00:00:00 2001
From: Eric Yang <Eric.Yang2@amd.com>
Date: Tue, 3 Apr 2018 11:36:14 -0400
Subject: drm/amd/display: dal 3.1.42

Signed-off-by: Eric Yang <Eric.Yang2@amd.com>
Reviewed-by: Anthony Koo <Anthony.Koo@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dc.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index 0f566a1..7ac8a1b 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -38,7 +38,7 @@
 #include "inc/compressor.h"
 #include "dml/display_mode_lib.h"
 
-#define DC_VER "3.1.41"
+#define DC_VER "3.1.42"
 
 #define MAX_SURFACES 3
 #define MAX_STREAMS 6
-- 
cgit v1.1


From ab892598d033d1943e1dcb0326f2622d6026f524 Mon Sep 17 00:00:00 2001
From: Roman Li <Roman.Li@amd.com>
Date: Thu, 29 Mar 2018 10:56:17 -0400
Subject: drm/amd/display: fix brightness level after resume from suspend

Adding missing call to cache current backlight values.
Otherwise the brightness resets to default value on resume.

Signed-off-by: Roman Li <Roman.Li@amd.com>
Reviewed-by: Charlene Liu <Charlene.Liu@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_link.c               | 13 +++++++++++++
 drivers/gpu/drm/amd/display/dc/dc_link.h                    |  2 ++
 drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c |  4 +++-
 3 files changed, 18 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
index 0cd286f..b44cf52 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
@@ -2018,6 +2018,19 @@ bool dc_link_set_backlight_level(const struct dc_link *link, uint32_t level,
 	return true;
 }
 
+bool dc_link_set_abm_disable(const struct dc_link *link)
+{
+	struct dc  *core_dc = link->ctx->dc;
+	struct abm *abm = core_dc->res_pool->abm;
+
+	if ((abm == NULL) || (abm->funcs->set_backlight_level == NULL))
+		return false;
+
+	abm->funcs->set_abm_immediate_disable(abm);
+
+	return true;
+}
+
 bool dc_link_set_psr_enable(const struct dc_link *link, bool enable, bool wait)
 {
 	struct dc  *core_dc = link->ctx->dc;
diff --git a/drivers/gpu/drm/amd/display/dc/dc_link.h b/drivers/gpu/drm/amd/display/dc/dc_link.h
index eeff987..8a716baa 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_link.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_link.h
@@ -141,6 +141,8 @@ static inline struct dc_link *dc_get_link_at_index(struct dc *dc, uint32_t link_
 bool dc_link_set_backlight_level(const struct dc_link *dc_link, uint32_t level,
 		uint32_t frame_ramp, const struct dc_stream_state *stream);
 
+bool dc_link_set_abm_disable(const struct dc_link *dc_link);
+
 bool dc_link_set_psr_enable(const struct dc_link *dc_link, bool enable, bool wait);
 
 bool dc_link_get_psr_state(const struct dc_link *dc_link, uint32_t *psr_state);
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
index 68a182c..15897f0 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
@@ -1046,8 +1046,10 @@ void dce110_blank_stream(struct pipe_ctx *pipe_ctx)
 	struct dc_stream_state *stream = pipe_ctx->stream;
 	struct dc_link *link = stream->sink->link;
 
-	if (link->local_sink && link->local_sink->sink_signal == SIGNAL_TYPE_EDP)
+	if (link->local_sink && link->local_sink->sink_signal == SIGNAL_TYPE_EDP) {
 		link->dc->hwss.edp_backlight_control(link, false);
+		dc_link_set_abm_disable(link);
+	}
 
 	if (dc_is_dp_signal(pipe_ctx->stream->signal))
 		pipe_ctx->stream_res.stream_enc->funcs->dp_blank(pipe_ctx->stream_res.stream_enc);
-- 
cgit v1.1


From fcb2008a70c8dffc9179ce41838496ba816e14a1 Mon Sep 17 00:00:00 2001
From: Eric Bernstein <eric.bernstein@amd.com>
Date: Tue, 3 Apr 2018 11:23:11 -0400
Subject: drm/amd/display: Move dp_pixel_encoding_type to stream_encoder
 include

Signed-off-by: Eric Bernstein <eric.bernstein@amd.com>
Reviewed-by: Nikola Cornij <Nikola.Cornij@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h     | 17 -----------------
 .../gpu/drm/amd/display/dc/inc/hw/stream_encoder.h    | 19 +++++++++++++++++++
 2 files changed, 19 insertions(+), 17 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h b/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h
index 9fe7302..cf7433e 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h
@@ -186,23 +186,6 @@ enum controller_dp_test_pattern {
 	CONTROLLER_DP_TEST_PATTERN_COLORSQUARES_CEA
 };
 
-enum dp_pixel_encoding_type {
-	DP_PIXEL_ENCODING_TYPE_RGB444		= 0x00000000,
-	DP_PIXEL_ENCODING_TYPE_YCBCR422		= 0x00000001,
-	DP_PIXEL_ENCODING_TYPE_YCBCR444		= 0x00000002,
-	DP_PIXEL_ENCODING_TYPE_RGB_WIDE_GAMUT	= 0x00000003,
-	DP_PIXEL_ENCODING_TYPE_Y_ONLY		= 0x00000004,
-	DP_PIXEL_ENCODING_TYPE_YCBCR420		= 0x00000005
-};
-
-enum dp_component_depth {
-	DP_COMPONENT_PIXEL_DEPTH_6BPC		= 0x00000000,
-	DP_COMPONENT_PIXEL_DEPTH_8BPC		= 0x00000001,
-	DP_COMPONENT_PIXEL_DEPTH_10BPC		= 0x00000002,
-	DP_COMPONENT_PIXEL_DEPTH_12BPC		= 0x00000003,
-	DP_COMPONENT_PIXEL_DEPTH_16BPC		= 0x00000004
-};
-
 enum dc_lut_mode {
 	LUT_BYPASS,
 	LUT_RAM_A,
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h b/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h
index 5c21336..cfa7ec9 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h
@@ -29,11 +29,29 @@
 #define STREAM_ENCODER_H_
 
 #include "audio_types.h"
+#include "hw_shared.h"
 
 struct dc_bios;
 struct dc_context;
 struct dc_crtc_timing;
 
+enum dp_pixel_encoding_type {
+	DP_PIXEL_ENCODING_TYPE_RGB444		= 0x00000000,
+	DP_PIXEL_ENCODING_TYPE_YCBCR422		= 0x00000001,
+	DP_PIXEL_ENCODING_TYPE_YCBCR444		= 0x00000002,
+	DP_PIXEL_ENCODING_TYPE_RGB_WIDE_GAMUT	= 0x00000003,
+	DP_PIXEL_ENCODING_TYPE_Y_ONLY		= 0x00000004,
+	DP_PIXEL_ENCODING_TYPE_YCBCR420		= 0x00000005
+};
+
+enum dp_component_depth {
+	DP_COMPONENT_PIXEL_DEPTH_6BPC		= 0x00000000,
+	DP_COMPONENT_PIXEL_DEPTH_8BPC		= 0x00000001,
+	DP_COMPONENT_PIXEL_DEPTH_10BPC		= 0x00000002,
+	DP_COMPONENT_PIXEL_DEPTH_12BPC		= 0x00000003,
+	DP_COMPONENT_PIXEL_DEPTH_16BPC		= 0x00000004
+};
+
 struct encoder_info_frame {
 	/* auxiliary video information */
 	struct dc_info_packet avi;
@@ -138,6 +156,7 @@ struct stream_encoder_funcs {
 
 	void (*set_avmute)(
 		struct stream_encoder *enc, bool enable);
+
 };
 
 #endif /* STREAM_ENCODER_H_ */
-- 
cgit v1.1


From fc6de1c565e03f492a3d9725b93092dac0cc1845 Mon Sep 17 00:00:00 2001
From: "Leo (Sunpeng) Li" <sunpeng.li@amd.com>
Date: Tue, 3 Apr 2018 16:07:16 -0400
Subject: drm/amd/display: Fix regamma not affecting full-intensity color
 values

Hardware understands the regamma LUT as a piecewise linear function,
with points spaced exponentially along the range. We previously
programmed the LUT for range [2^-10, 2^0). This causes (normalized)
color values of 1 (=2^0) to miss the programmed LUT, and fall onto the
end region.

For DCE, the end region is extrapolated using a single (base, slope)
pair, using the max y-value from the last point in the curve as base.
This presents a problem, since this value affects all three color
channels. Scaling down the intensity of say - the blue regamma curve -
will not affect it's end region. This is especially noticiable when
using RedShift. It scales down the blue and green channels, but leaves
full-intensity colors unshifted.

Therefore, extend the range to cover [2^-10, 2^1) by programming another
hardware segment, containing only one point. That way, we won't be
hitting the end region.

Note that things are a bit different for DCN, since the end region can
be set per-channel.

Signed-off-by: Leo (Sunpeng) Li <sunpeng.li@amd.com>
Reviewed-by: Krunoslav Kovac <Krunoslav.Kovac@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
index 15897f0..1b5c11c 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
@@ -456,10 +456,13 @@ dce110_translate_regamma_to_hw_format(const struct dc_transfer_func *output_tf,
 
 	} else {
 		/* 10 segments
-		 * segment is from 2^-10 to 2^0
+		 * segment is from 2^-10 to 2^1
+		 * We include an extra segment for range [2^0, 2^1). This is to
+		 * ensure that colors with normalized values of 1 don't miss the
+		 * LUT.
 		 */
 		region_start = -10;
-		region_end = 0;
+		region_end = 1;
 
 		seg_distr[0] = 4;
 		seg_distr[1] = 4;
@@ -471,7 +474,7 @@ dce110_translate_regamma_to_hw_format(const struct dc_transfer_func *output_tf,
 		seg_distr[7] = 4;
 		seg_distr[8] = 4;
 		seg_distr[9] = 4;
-		seg_distr[10] = -1;
+		seg_distr[10] = 0;
 		seg_distr[11] = -1;
 		seg_distr[12] = -1;
 		seg_distr[13] = -1;
-- 
cgit v1.1


From c5b38aec266deade4067ddc606634ace68d2da8c Mon Sep 17 00:00:00 2001
From: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Date: Thu, 29 Mar 2018 16:39:10 -0400
Subject: drm/amd/display: fix segfault on insufficient TG during validation

Signed-off-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Reviewed-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_resource.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
index 8d7bc1f..d7a92ec 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
@@ -1700,7 +1700,7 @@ enum dc_status resource_map_pool_resources(
 		pipe_idx = acquire_first_split_pipe(&context->res_ctx, pool, stream);
 #endif
 
-	if (pipe_idx < 0)
+	if (pipe_idx < 0 || context->res_ctx.pipe_ctx[pipe_idx].stream_res.tg == NULL)
 		return DC_NO_CONTROLLER_RESOURCE;
 
 	pipe_ctx = &context->res_ctx.pipe_ctx[pipe_idx];
-- 
cgit v1.1


From d0f6f1c0319d39b792a7969bf511d5b1870f1f0e Mon Sep 17 00:00:00 2001
From: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Date: Fri, 23 Mar 2018 15:25:43 -0400
Subject: drm/amd/display: change dml init to use default structs

Signed-off-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Reviewed-by: Eric Bernstein <Eric.Bernstein@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../gpu/drm/amd/display/dc/dml/display_mode_lib.c  | 138 ++++++++++++---------
 1 file changed, 76 insertions(+), 62 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.c b/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.c
index c109b2c..fd9d97a 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.c
@@ -26,75 +26,89 @@
 #include "display_mode_lib.h"
 #include "dc_features.h"
 
+static const struct _vcs_dpi_ip_params_st dcn1_0_ip = {
+	.rob_buffer_size_kbytes = 64,
+	.det_buffer_size_kbytes = 164,
+	.dpte_buffer_size_in_pte_reqs = 42,
+	.dpp_output_buffer_pixels = 2560,
+	.opp_output_buffer_lines = 1,
+	.pixel_chunk_size_kbytes = 8,
+	.pte_enable = 1,
+	.pte_chunk_size_kbytes = 2,
+	.meta_chunk_size_kbytes = 2,
+	.writeback_chunk_size_kbytes = 2,
+	.line_buffer_size_bits = 589824,
+	.max_line_buffer_lines = 12,
+	.IsLineBufferBppFixed = 0,
+	.LineBufferFixedBpp = -1,
+	.writeback_luma_buffer_size_kbytes = 12,
+	.writeback_chroma_buffer_size_kbytes = 8,
+	.max_num_dpp = 4,
+	.max_num_wb = 2,
+	.max_dchub_pscl_bw_pix_per_clk = 4,
+	.max_pscl_lb_bw_pix_per_clk = 2,
+	.max_lb_vscl_bw_pix_per_clk = 4,
+	.max_vscl_hscl_bw_pix_per_clk = 4,
+	.max_hscl_ratio = 4,
+	.max_vscl_ratio = 4,
+	.hscl_mults = 4,
+	.vscl_mults = 4,
+	.max_hscl_taps = 8,
+	.max_vscl_taps = 8,
+	.dispclk_ramp_margin_percent = 1,
+	.underscan_factor = 1.10,
+	.min_vblank_lines = 14,
+	.dppclk_delay_subtotal = 90,
+	.dispclk_delay_subtotal = 42,
+	.dcfclk_cstate_latency = 10,
+	.max_inter_dcn_tile_repeaters = 8,
+	.can_vstartup_lines_exceed_vsync_plus_back_porch_lines_minus_one = 0,
+	.bug_forcing_LC_req_same_size_fixed = 0,
+};
+
+static const struct _vcs_dpi_soc_bounding_box_st dcn1_0_soc = {
+	.sr_exit_time_us = 9.0,
+	.sr_enter_plus_exit_time_us = 11.0,
+	.urgent_latency_us = 4.0,
+	.writeback_latency_us = 12.0,
+	.ideal_dram_bw_after_urgent_percent = 80.0,
+	.max_request_size_bytes = 256,
+	.downspread_percent = 0.5,
+	.dram_page_open_time_ns = 50.0,
+	.dram_rw_turnaround_time_ns = 17.5,
+	.dram_return_buffer_per_channel_bytes = 8192,
+	.round_trip_ping_latency_dcfclk_cycles = 128,
+	.urgent_out_of_order_return_per_channel_bytes = 256,
+	.channel_interleave_bytes = 256,
+	.num_banks = 8,
+	.num_chans = 2,
+	.vmm_page_size_bytes = 4096,
+	.dram_clock_change_latency_us = 17.0,
+	.writeback_dram_clock_change_latency_us = 23.0,
+	.return_bus_width_bytes = 64,
+};
+
 static void set_soc_bounding_box(struct _vcs_dpi_soc_bounding_box_st *soc, enum dml_project project)
 {
-	if (project == DML_PROJECT_RAVEN1) {
-		soc->sr_exit_time_us = 9.0;
-		soc->sr_enter_plus_exit_time_us = 11.0;
-		soc->urgent_latency_us = 4.0;
-		soc->writeback_latency_us = 12.0;
-		soc->ideal_dram_bw_after_urgent_percent = 80.0;
-		soc->max_request_size_bytes = 256;
-		soc->downspread_percent = 0.5;
-		soc->dram_page_open_time_ns = 50.0;
-		soc->dram_rw_turnaround_time_ns = 17.5;
-		soc->dram_return_buffer_per_channel_bytes = 8192;
-		soc->round_trip_ping_latency_dcfclk_cycles = 128;
-		soc->urgent_out_of_order_return_per_channel_bytes = 256;
-		soc->channel_interleave_bytes = 256;
-		soc->num_banks = 8;
-		soc->num_chans = 2;
-		soc->vmm_page_size_bytes = 4096;
-		soc->dram_clock_change_latency_us = 17.0;
-		soc->writeback_dram_clock_change_latency_us = 23.0;
-		soc->return_bus_width_bytes = 64;
-	} else {
-		BREAK_TO_DEBUGGER(); /* Invalid Project Specified */
+	switch (project) {
+	case DML_PROJECT_RAVEN1:
+		*soc = dcn1_0_soc;
+		break;
+	default:
+		ASSERT(0);
+		break;
 	}
 }
 
 static void set_ip_params(struct _vcs_dpi_ip_params_st *ip, enum dml_project project)
 {
-	if (project == DML_PROJECT_RAVEN1) {
-		ip->rob_buffer_size_kbytes = 64;
-		ip->det_buffer_size_kbytes = 164;
-		ip->dpte_buffer_size_in_pte_reqs = 42;
-		ip->dpp_output_buffer_pixels = 2560;
-		ip->opp_output_buffer_lines = 1;
-		ip->pixel_chunk_size_kbytes = 8;
-		ip->pte_enable = 1;
-		ip->pte_chunk_size_kbytes = 2;
-		ip->meta_chunk_size_kbytes = 2;
-		ip->writeback_chunk_size_kbytes = 2;
-		ip->line_buffer_size_bits = 589824;
-		ip->max_line_buffer_lines = 12;
-		ip->IsLineBufferBppFixed = 0;
-		ip->LineBufferFixedBpp = -1;
-		ip->writeback_luma_buffer_size_kbytes = 12;
-		ip->writeback_chroma_buffer_size_kbytes = 8;
-		ip->max_num_dpp = 4;
-		ip->max_num_wb = 2;
-		ip->max_dchub_pscl_bw_pix_per_clk = 4;
-		ip->max_pscl_lb_bw_pix_per_clk = 2;
-		ip->max_lb_vscl_bw_pix_per_clk = 4;
-		ip->max_vscl_hscl_bw_pix_per_clk = 4;
-		ip->max_hscl_ratio = 4;
-		ip->max_vscl_ratio = 4;
-		ip->hscl_mults = 4;
-		ip->vscl_mults = 4;
-		ip->max_hscl_taps = 8;
-		ip->max_vscl_taps = 8;
-		ip->dispclk_ramp_margin_percent = 1;
-		ip->underscan_factor = 1.10;
-		ip->min_vblank_lines = 14;
-		ip->dppclk_delay_subtotal = 90;
-		ip->dispclk_delay_subtotal = 42;
-		ip->dcfclk_cstate_latency = 10;
-		ip->max_inter_dcn_tile_repeaters = 8;
-		ip->can_vstartup_lines_exceed_vsync_plus_back_porch_lines_minus_one = 0;
-		ip->bug_forcing_LC_req_same_size_fixed = 0;
-	} else {
-		BREAK_TO_DEBUGGER(); /* Invalid Project Specified */
+	switch (project) {
+	case DML_PROJECT_RAVEN1:
+		*ip = dcn1_0_ip;
+		break;
+	default:
+		ASSERT(0);
+		break;
 	}
 }
 
-- 
cgit v1.1


From 339cc82ae67700cb25a5bb10842cca5b09a79afe Mon Sep 17 00:00:00 2001
From: Yongqiang Sun <yongqiang.sun@amd.com>
Date: Wed, 4 Apr 2018 17:27:18 -0400
Subject: drm/amd/display: Check lid state to determine fast boot optimization.

For legacy enable boot up with lid closed, eDP information couldn't be
read correctly via SBIOS_SCRATCH_3 results in eDP cannot be light up
properly when open lid.
Check lid state instead can resolve the issue.

Signed-off-by: Yongqiang Sun <yongqiang.sun@amd.com>
Reviewed-by: Eric Yang <eric.yang2@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dc_stream.h         |  1 +
 .../amd/display/dc/dce110/dce110_hw_sequencer.c    | 24 ++++++++++++++--------
 2 files changed, 17 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dc_stream.h b/drivers/gpu/drm/amd/display/dc/dc_stream.h
index d7e6d53..11b3433 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_stream.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_stream.h
@@ -92,6 +92,7 @@ struct dc_stream_state {
 	int phy_pix_clk;
 	enum signal_type signal;
 	bool dpms_off;
+	bool lid_state_closed;
 
 	struct dc_stream_status status;
 
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
index 1b5c11c..4a4b3bc 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
@@ -1471,6 +1471,17 @@ static void disable_vga_and_power_gate_all_controllers(
 	}
 }
 
+static bool is_eDP_lid_closed(struct dc_state *context)
+{
+	int i;
+
+	for (i = 0; i < context->stream_count; i++) {
+		if (context->streams[i]->signal == SIGNAL_TYPE_EDP)
+			return context->streams[i]->lid_state_closed;
+	}
+	return false;
+}
+
 static struct dc_link *get_link_for_edp_not_in_use(
 		struct dc *dc,
 		struct dc_state *context)
@@ -1505,20 +1516,17 @@ static struct dc_link *get_link_for_edp_not_in_use(
  */
 void dce110_enable_accelerated_mode(struct dc *dc, struct dc_state *context)
 {
-	struct dc_bios *dcb = dc->ctx->dc_bios;
-
-	/* vbios already light up eDP, so we can leverage vbios and skip eDP
+	/* check eDP lid state:
+	 * If lid is open, vbios already light up eDP, so we can leverage vbios and skip eDP
 	 * programming
 	 */
-	bool can_eDP_fast_boot_optimize =
-			(dcb->funcs->get_vga_enabled_displays(dc->ctx->dc_bios) == ATOM_DISPLAY_LCD1_ACTIVE);
-
-	/* if OS doesn't light up eDP and eDP link is available, we want to disable */
+	bool lid_state_closed = is_eDP_lid_closed(context);
 	struct dc_link *edp_link_to_turnoff = NULL;
 
-	if (can_eDP_fast_boot_optimize) {
+	if (!lid_state_closed) {
 		edp_link_to_turnoff = get_link_for_edp_not_in_use(dc, context);
 
+		/* if OS doesn't light up eDP and eDP link is available, we want to disable */
 		if (!edp_link_to_turnoff)
 			dc->apply_edp_fast_boot_optimization = true;
 	}
-- 
cgit v1.1


From c4b0faae71f33377a11fe19dadcce6deb86f5037 Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Fri, 6 Apr 2018 12:07:19 -0400
Subject: drm/amd/display: Do not create memory allocation if stats not enabled

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/modules/stats/stats.c | 26 +++++++++++++----------
 1 file changed, 15 insertions(+), 11 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/modules/stats/stats.c b/drivers/gpu/drm/amd/display/modules/stats/stats.c
index ed5f680..48e0219 100644
--- a/drivers/gpu/drm/amd/display/modules/stats/stats.c
+++ b/drivers/gpu/drm/amd/display/modules/stats/stats.c
@@ -115,18 +115,22 @@ struct mod_stats *mod_stats_create(struct dc *dc)
 			&reg_data, sizeof(unsigned int), &flag))
 		core_stats->enabled = reg_data;
 
-	core_stats->entries = DAL_STATS_ENTRIES_REGKEY_DEFAULT;
-	if (dm_read_persistent_data(dc->ctx, NULL, NULL,
-			DAL_STATS_ENTRIES_REGKEY,
-			&reg_data, sizeof(unsigned int), &flag)) {
-		if (reg_data > DAL_STATS_ENTRIES_REGKEY_MAX)
-			core_stats->entries = DAL_STATS_ENTRIES_REGKEY_MAX;
-		else
-			core_stats->entries = reg_data;
-	}
+	if (core_stats->enabled) {
+		core_stats->entries = DAL_STATS_ENTRIES_REGKEY_DEFAULT;
+		if (dm_read_persistent_data(dc->ctx, NULL, NULL,
+				DAL_STATS_ENTRIES_REGKEY,
+				&reg_data, sizeof(unsigned int), &flag)) {
+			if (reg_data > DAL_STATS_ENTRIES_REGKEY_MAX)
+				core_stats->entries = DAL_STATS_ENTRIES_REGKEY_MAX;
+			else
+				core_stats->entries = reg_data;
+		}
 
-	core_stats->time = kzalloc(sizeof(struct stats_time_cache) * core_stats->entries,
-					GFP_KERNEL);
+		core_stats->time = kzalloc(sizeof(struct stats_time_cache) * core_stats->entries,
+						GFP_KERNEL);
+	} else {
+		core_stats->entries = 0;
+	}
 
 	if (core_stats->time == NULL)
 		goto fail_construct;
-- 
cgit v1.1


From 5ebfb7a5996ea1dceeb2a392d7e46357042e4506 Mon Sep 17 00:00:00 2001
From: Eric Bernstein <eric.bernstein@amd.com>
Date: Thu, 5 Apr 2018 17:09:20 -0400
Subject: drm/amd/display: Move DCC support functions into dchubbub

Added dchububu.h header file for common enum/struct definitions.
Added new interface functions get_dcc_compression_cap,
dcc_support_swizzle, dcc_support_pixel_format.

Signed-off-by: Eric Bernstein <eric.bernstein@amd.com>
Reviewed-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c    | 221 +++++++++++++++++++-
 .../gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h    |   7 +-
 .../gpu/drm/amd/display/dc/dcn10/dcn10_resource.c  | 231 +--------------------
 drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h   |  64 ++++++
 4 files changed, 291 insertions(+), 232 deletions(-)
 create mode 100644 drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c
index 738f67f..b9fb14a 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c
@@ -476,8 +476,227 @@ void hubbub1_toggle_watermark_change_req(struct hubbub *hubbub)
 			DCHUBBUB_ARB_WATERMARK_CHANGE_REQUEST, watermark_change_req);
 }
 
+static bool hubbub1_dcc_support_swizzle(
+		enum swizzle_mode_values swizzle,
+		unsigned int bytes_per_element,
+		enum segment_order *segment_order_horz,
+		enum segment_order *segment_order_vert)
+{
+	bool standard_swizzle = false;
+	bool display_swizzle = false;
+
+	switch (swizzle) {
+	case DC_SW_4KB_S:
+	case DC_SW_64KB_S:
+	case DC_SW_VAR_S:
+	case DC_SW_4KB_S_X:
+	case DC_SW_64KB_S_X:
+	case DC_SW_VAR_S_X:
+		standard_swizzle = true;
+		break;
+	case DC_SW_4KB_D:
+	case DC_SW_64KB_D:
+	case DC_SW_VAR_D:
+	case DC_SW_4KB_D_X:
+	case DC_SW_64KB_D_X:
+	case DC_SW_VAR_D_X:
+		display_swizzle = true;
+		break;
+	default:
+		break;
+	}
+
+	if (bytes_per_element == 1 && standard_swizzle) {
+		*segment_order_horz = segment_order__contiguous;
+		*segment_order_vert = segment_order__na;
+		return true;
+	}
+	if (bytes_per_element == 2 && standard_swizzle) {
+		*segment_order_horz = segment_order__non_contiguous;
+		*segment_order_vert = segment_order__contiguous;
+		return true;
+	}
+	if (bytes_per_element == 4 && standard_swizzle) {
+		*segment_order_horz = segment_order__non_contiguous;
+		*segment_order_vert = segment_order__contiguous;
+		return true;
+	}
+	if (bytes_per_element == 8 && standard_swizzle) {
+		*segment_order_horz = segment_order__na;
+		*segment_order_vert = segment_order__contiguous;
+		return true;
+	}
+	if (bytes_per_element == 8 && display_swizzle) {
+		*segment_order_horz = segment_order__contiguous;
+		*segment_order_vert = segment_order__non_contiguous;
+		return true;
+	}
+
+	return false;
+}
+
+static bool hubbub1_dcc_support_pixel_format(
+		enum surface_pixel_format format,
+		unsigned int *bytes_per_element)
+{
+	/* DML: get_bytes_per_element */
+	switch (format) {
+	case SURFACE_PIXEL_FORMAT_GRPH_ARGB1555:
+	case SURFACE_PIXEL_FORMAT_GRPH_RGB565:
+		*bytes_per_element = 2;
+		return true;
+	case SURFACE_PIXEL_FORMAT_GRPH_ARGB8888:
+	case SURFACE_PIXEL_FORMAT_GRPH_ABGR8888:
+	case SURFACE_PIXEL_FORMAT_GRPH_ARGB2101010:
+	case SURFACE_PIXEL_FORMAT_GRPH_ABGR2101010:
+		*bytes_per_element = 4;
+		return true;
+	case SURFACE_PIXEL_FORMAT_GRPH_ARGB16161616:
+	case SURFACE_PIXEL_FORMAT_GRPH_ARGB16161616F:
+	case SURFACE_PIXEL_FORMAT_GRPH_ABGR16161616F:
+		*bytes_per_element = 8;
+		return true;
+	default:
+		return false;
+	}
+}
+
+static void hubbub1_get_blk256_size(unsigned int *blk256_width, unsigned int *blk256_height,
+		unsigned int bytes_per_element)
+{
+	/* copied from DML.  might want to refactor DML to leverage from DML */
+	/* DML : get_blk256_size */
+	if (bytes_per_element == 1) {
+		*blk256_width = 16;
+		*blk256_height = 16;
+	} else if (bytes_per_element == 2) {
+		*blk256_width = 16;
+		*blk256_height = 8;
+	} else if (bytes_per_element == 4) {
+		*blk256_width = 8;
+		*blk256_height = 8;
+	} else if (bytes_per_element == 8) {
+		*blk256_width = 8;
+		*blk256_height = 4;
+	}
+}
+
+static void hubbub1_det_request_size(
+		unsigned int height,
+		unsigned int width,
+		unsigned int bpe,
+		bool *req128_horz_wc,
+		bool *req128_vert_wc)
+{
+	unsigned int detile_buf_size = 164 * 1024;  /* 164KB for DCN1.0 */
+
+	unsigned int blk256_height = 0;
+	unsigned int blk256_width = 0;
+	unsigned int swath_bytes_horz_wc, swath_bytes_vert_wc;
+
+	hubbub1_get_blk256_size(&blk256_width, &blk256_height, bpe);
+
+	swath_bytes_horz_wc = height * blk256_height * bpe;
+	swath_bytes_vert_wc = width * blk256_width * bpe;
+
+	*req128_horz_wc = (2 * swath_bytes_horz_wc <= detile_buf_size) ?
+			false : /* full 256B request */
+			true; /* half 128b request */
+
+	*req128_vert_wc = (2 * swath_bytes_vert_wc <= detile_buf_size) ?
+			false : /* full 256B request */
+			true; /* half 128b request */
+}
+
+static bool hubbub1_get_dcc_compression_cap(struct hubbub *hubbub,
+		const struct dc_dcc_surface_param *input,
+		struct dc_surface_dcc_cap *output)
+{
+	struct dc *dc = hubbub->ctx->dc;
+	/* implement section 1.6.2.1 of DCN1_Programming_Guide.docx */
+	enum dcc_control dcc_control;
+	unsigned int bpe;
+	enum segment_order segment_order_horz, segment_order_vert;
+	bool req128_horz_wc, req128_vert_wc;
+
+	memset(output, 0, sizeof(*output));
+
+	if (dc->debug.disable_dcc == DCC_DISABLE)
+		return false;
+
+	if (!hubbub->funcs->dcc_support_pixel_format(input->format, &bpe))
+		return false;
+
+	if (!hubbub->funcs->dcc_support_swizzle(input->swizzle_mode, bpe,
+			&segment_order_horz, &segment_order_vert))
+		return false;
+
+	hubbub1_det_request_size(input->surface_size.height,  input->surface_size.width,
+			bpe, &req128_horz_wc, &req128_vert_wc);
+
+	if (!req128_horz_wc && !req128_vert_wc) {
+		dcc_control = dcc_control__256_256_xxx;
+	} else if (input->scan == SCAN_DIRECTION_HORIZONTAL) {
+		if (!req128_horz_wc)
+			dcc_control = dcc_control__256_256_xxx;
+		else if (segment_order_horz == segment_order__contiguous)
+			dcc_control = dcc_control__128_128_xxx;
+		else
+			dcc_control = dcc_control__256_64_64;
+	} else if (input->scan == SCAN_DIRECTION_VERTICAL) {
+		if (!req128_vert_wc)
+			dcc_control = dcc_control__256_256_xxx;
+		else if (segment_order_vert == segment_order__contiguous)
+			dcc_control = dcc_control__128_128_xxx;
+		else
+			dcc_control = dcc_control__256_64_64;
+	} else {
+		if ((req128_horz_wc &&
+			segment_order_horz == segment_order__non_contiguous) ||
+			(req128_vert_wc &&
+			segment_order_vert == segment_order__non_contiguous))
+			/* access_dir not known, must use most constraining */
+			dcc_control = dcc_control__256_64_64;
+		else
+			/* reg128 is true for either horz and vert
+			 * but segment_order is contiguous
+			 */
+			dcc_control = dcc_control__128_128_xxx;
+	}
+
+	if (dc->debug.disable_dcc == DCC_HALF_REQ_DISALBE &&
+		dcc_control != dcc_control__256_256_xxx)
+		return false;
+
+	switch (dcc_control) {
+	case dcc_control__256_256_xxx:
+		output->grph.rgb.max_uncompressed_blk_size = 256;
+		output->grph.rgb.max_compressed_blk_size = 256;
+		output->grph.rgb.independent_64b_blks = false;
+		break;
+	case dcc_control__128_128_xxx:
+		output->grph.rgb.max_uncompressed_blk_size = 128;
+		output->grph.rgb.max_compressed_blk_size = 128;
+		output->grph.rgb.independent_64b_blks = false;
+		break;
+	case dcc_control__256_64_64:
+		output->grph.rgb.max_uncompressed_blk_size = 256;
+		output->grph.rgb.max_compressed_blk_size = 64;
+		output->grph.rgb.independent_64b_blks = true;
+		break;
+	}
+
+	output->capable = true;
+	output->const_color_support = false;
+
+	return true;
+}
+
 static const struct hubbub_funcs hubbub1_funcs = {
-	.update_dchub = hubbub1_update_dchub
+	.update_dchub = hubbub1_update_dchub,
+	.dcc_support_swizzle = hubbub1_dcc_support_swizzle,
+	.dcc_support_pixel_format = hubbub1_dcc_support_pixel_format,
+	.get_dcc_compression_cap = hubbub1_get_dcc_compression_cap,
 };
 
 void hubbub1_construct(struct hubbub *hubbub,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h
index a16e908..f479f54 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h
@@ -27,6 +27,7 @@
 #define __DC_HUBBUB_DCN10_H__
 
 #include "core_types.h"
+#include "dchubbub.h"
 
 #define HUBHUB_REG_LIST_DCN()\
 	SR(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_A),\
@@ -173,12 +174,6 @@ struct dcn_hubbub_wm {
 	struct dcn_hubbub_wm_set sets[4];
 };
 
-struct hubbub_funcs {
-	void (*update_dchub)(
-			struct hubbub *hubbub,
-			struct dchub_init_data *dh_data);
-};
-
 struct hubbub {
 	const struct hubbub_funcs *funcs;
 	struct dc_context *ctx;
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
index f305f65..2c0a315 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
@@ -937,235 +937,16 @@ static struct pipe_ctx *dcn10_acquire_idle_pipe_for_layer(
 	return idle_pipe;
 }
 
-enum dcc_control {
-	dcc_control__256_256_xxx,
-	dcc_control__128_128_xxx,
-	dcc_control__256_64_64,
-};
-
-enum segment_order {
-	segment_order__na,
-	segment_order__contiguous,
-	segment_order__non_contiguous,
-};
-
-static bool dcc_support_pixel_format(
-		enum surface_pixel_format format,
-		unsigned int *bytes_per_element)
-{
-	/* DML: get_bytes_per_element */
-	switch (format) {
-	case SURFACE_PIXEL_FORMAT_GRPH_ARGB1555:
-	case SURFACE_PIXEL_FORMAT_GRPH_RGB565:
-		*bytes_per_element = 2;
-		return true;
-	case SURFACE_PIXEL_FORMAT_GRPH_ARGB8888:
-	case SURFACE_PIXEL_FORMAT_GRPH_ABGR8888:
-	case SURFACE_PIXEL_FORMAT_GRPH_ARGB2101010:
-	case SURFACE_PIXEL_FORMAT_GRPH_ABGR2101010:
-		*bytes_per_element = 4;
-		return true;
-	case SURFACE_PIXEL_FORMAT_GRPH_ARGB16161616:
-	case SURFACE_PIXEL_FORMAT_GRPH_ARGB16161616F:
-	case SURFACE_PIXEL_FORMAT_GRPH_ABGR16161616F:
-		*bytes_per_element = 8;
-		return true;
-	default:
-		return false;
-	}
-}
-
-static bool dcc_support_swizzle(
-		enum swizzle_mode_values swizzle,
-		unsigned int bytes_per_element,
-		enum segment_order *segment_order_horz,
-		enum segment_order *segment_order_vert)
-{
-	bool standard_swizzle = false;
-	bool display_swizzle = false;
-
-	switch (swizzle) {
-	case DC_SW_4KB_S:
-	case DC_SW_64KB_S:
-	case DC_SW_VAR_S:
-	case DC_SW_4KB_S_X:
-	case DC_SW_64KB_S_X:
-	case DC_SW_VAR_S_X:
-		standard_swizzle = true;
-		break;
-	case DC_SW_4KB_D:
-	case DC_SW_64KB_D:
-	case DC_SW_VAR_D:
-	case DC_SW_4KB_D_X:
-	case DC_SW_64KB_D_X:
-	case DC_SW_VAR_D_X:
-		display_swizzle = true;
-		break;
-	default:
-		break;
-	}
-
-	if (bytes_per_element == 1 && standard_swizzle) {
-		*segment_order_horz = segment_order__contiguous;
-		*segment_order_vert = segment_order__na;
-		return true;
-	}
-	if (bytes_per_element == 2 && standard_swizzle) {
-		*segment_order_horz = segment_order__non_contiguous;
-		*segment_order_vert = segment_order__contiguous;
-		return true;
-	}
-	if (bytes_per_element == 4 && standard_swizzle) {
-		*segment_order_horz = segment_order__non_contiguous;
-		*segment_order_vert = segment_order__contiguous;
-		return true;
-	}
-	if (bytes_per_element == 8 && standard_swizzle) {
-		*segment_order_horz = segment_order__na;
-		*segment_order_vert = segment_order__contiguous;
-		return true;
-	}
-	if (bytes_per_element == 8 && display_swizzle) {
-		*segment_order_horz = segment_order__contiguous;
-		*segment_order_vert = segment_order__non_contiguous;
-		return true;
-	}
-
-	return false;
-}
-
-static void get_blk256_size(unsigned int *blk256_width, unsigned int *blk256_height,
-		unsigned int bytes_per_element)
-{
-	/* copied from DML.  might want to refactor DML to leverage from DML */
-	/* DML : get_blk256_size */
-	if (bytes_per_element == 1) {
-		*blk256_width = 16;
-		*blk256_height = 16;
-	} else if (bytes_per_element == 2) {
-		*blk256_width = 16;
-		*blk256_height = 8;
-	} else if (bytes_per_element == 4) {
-		*blk256_width = 8;
-		*blk256_height = 8;
-	} else if (bytes_per_element == 8) {
-		*blk256_width = 8;
-		*blk256_height = 4;
-	}
-}
-
-static void det_request_size(
-		unsigned int height,
-		unsigned int width,
-		unsigned int bpe,
-		bool *req128_horz_wc,
-		bool *req128_vert_wc)
-{
-	unsigned int detile_buf_size = 164 * 1024;  /* 164KB for DCN1.0 */
-
-	unsigned int blk256_height = 0;
-	unsigned int blk256_width = 0;
-	unsigned int swath_bytes_horz_wc, swath_bytes_vert_wc;
-
-	get_blk256_size(&blk256_width, &blk256_height, bpe);
-
-	swath_bytes_horz_wc = height * blk256_height * bpe;
-	swath_bytes_vert_wc = width * blk256_width * bpe;
-
-	*req128_horz_wc = (2 * swath_bytes_horz_wc <= detile_buf_size) ?
-			false : /* full 256B request */
-			true; /* half 128b request */
-
-	*req128_vert_wc = (2 * swath_bytes_vert_wc <= detile_buf_size) ?
-			false : /* full 256B request */
-			true; /* half 128b request */
-}
-
-static bool get_dcc_compression_cap(const struct dc *dc,
+static bool dcn10_get_dcc_compression_cap(const struct dc *dc,
 		const struct dc_dcc_surface_param *input,
 		struct dc_surface_dcc_cap *output)
 {
-	/* implement section 1.6.2.1 of DCN1_Programming_Guide.docx */
-	enum dcc_control dcc_control;
-	unsigned int bpe;
-	enum segment_order segment_order_horz, segment_order_vert;
-	bool req128_horz_wc, req128_vert_wc;
-
-	memset(output, 0, sizeof(*output));
-
-	if (dc->debug.disable_dcc == DCC_DISABLE)
-		return false;
-
-	if (!dcc_support_pixel_format(input->format,
-			&bpe))
-		return false;
-
-	if (!dcc_support_swizzle(input->swizzle_mode, bpe,
-			&segment_order_horz, &segment_order_vert))
-		return false;
-
-	det_request_size(input->surface_size.height,  input->surface_size.width,
-			bpe, &req128_horz_wc, &req128_vert_wc);
-
-	if (!req128_horz_wc && !req128_vert_wc) {
-		dcc_control = dcc_control__256_256_xxx;
-	} else if (input->scan == SCAN_DIRECTION_HORIZONTAL) {
-		if (!req128_horz_wc)
-			dcc_control = dcc_control__256_256_xxx;
-		else if (segment_order_horz == segment_order__contiguous)
-			dcc_control = dcc_control__128_128_xxx;
-		else
-			dcc_control = dcc_control__256_64_64;
-	} else if (input->scan == SCAN_DIRECTION_VERTICAL) {
-		if (!req128_vert_wc)
-			dcc_control = dcc_control__256_256_xxx;
-		else if (segment_order_vert == segment_order__contiguous)
-			dcc_control = dcc_control__128_128_xxx;
-		else
-			dcc_control = dcc_control__256_64_64;
-	} else {
-		if ((req128_horz_wc &&
-			segment_order_horz == segment_order__non_contiguous) ||
-			(req128_vert_wc &&
-			segment_order_vert == segment_order__non_contiguous))
-			/* access_dir not known, must use most constraining */
-			dcc_control = dcc_control__256_64_64;
-		else
-			/* reg128 is true for either horz and vert
-			 * but segment_order is contiguous
-			 */
-			dcc_control = dcc_control__128_128_xxx;
-	}
-
-	if (dc->debug.disable_dcc == DCC_HALF_REQ_DISALBE &&
-		dcc_control != dcc_control__256_256_xxx)
-		return false;
-
-	switch (dcc_control) {
-	case dcc_control__256_256_xxx:
-		output->grph.rgb.max_uncompressed_blk_size = 256;
-		output->grph.rgb.max_compressed_blk_size = 256;
-		output->grph.rgb.independent_64b_blks = false;
-		break;
-	case dcc_control__128_128_xxx:
-		output->grph.rgb.max_uncompressed_blk_size = 128;
-		output->grph.rgb.max_compressed_blk_size = 128;
-		output->grph.rgb.independent_64b_blks = false;
-		break;
-	case dcc_control__256_64_64:
-		output->grph.rgb.max_uncompressed_blk_size = 256;
-		output->grph.rgb.max_compressed_blk_size = 64;
-		output->grph.rgb.independent_64b_blks = true;
-		break;
-	}
-
-	output->capable = true;
-	output->const_color_support = false;
-
-	return true;
+	return dc->res_pool->hubbub->funcs->get_dcc_compression_cap(
+			dc->res_pool->hubbub,
+			input,
+			output);
 }
 
-
 static void dcn10_destroy_resource_pool(struct resource_pool **pool)
 {
 	struct dcn10_resource_pool *dcn10_pool = TO_DCN10_RES_POOL(*pool);
@@ -1186,7 +967,7 @@ static enum dc_status dcn10_validate_plane(const struct dc_plane_state *plane_st
 }
 
 static struct dc_cap_funcs cap_funcs = {
-	.get_dcc_compression_cap = get_dcc_compression_cap
+	.get_dcc_compression_cap = dcn10_get_dcc_compression_cap
 };
 
 static struct resource_funcs dcn10_res_pool_funcs = {
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h
new file mode 100644
index 0000000..02f757d
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h
@@ -0,0 +1,64 @@
+/*
+ * Copyright 2012-15 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef __DAL_DCHUBBUB_H__
+#define __DAL_DCHUBBUB_H__
+
+
+enum dcc_control {
+	dcc_control__256_256_xxx,
+	dcc_control__128_128_xxx,
+	dcc_control__256_64_64,
+};
+
+enum segment_order {
+	segment_order__na,
+	segment_order__contiguous,
+	segment_order__non_contiguous,
+};
+
+
+struct hubbub_funcs {
+	void (*update_dchub)(
+			struct hubbub *hubbub,
+			struct dchub_init_data *dh_data);
+
+	bool (*get_dcc_compression_cap)(struct hubbub *hubbub,
+			const struct dc_dcc_surface_param *input,
+			struct dc_surface_dcc_cap *output);
+
+	bool (*dcc_support_swizzle)(
+			enum swizzle_mode_values swizzle,
+			unsigned int bytes_per_element,
+			enum segment_order *segment_order_horz,
+			enum segment_order *segment_order_vert);
+
+	bool (*dcc_support_pixel_format)(
+			enum surface_pixel_format format,
+			unsigned int *bytes_per_element);
+};
+
+
+#endif
-- 
cgit v1.1


From 7ac897b5afb98369a4edd71950921026c3029d5f Mon Sep 17 00:00:00 2001
From: Charlene Liu <charlene.liu@amd.com>
Date: Fri, 6 Apr 2018 23:03:12 -0400
Subject: drm/amd/display: HDMI has no sound after Panel power off/on

Signed-off-by: Charlene Liu <charlene.liu@amd.com>
Reviewed-by: Krunoslav Kovac <Krunoslav.Kovac@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Cc: stable@vger.kernel.org
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
index 07c3242..84e26c8 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
@@ -718,6 +718,8 @@ static void dce110_stream_encoder_update_hdmi_info_packets(
 		if (info_frame->avi.valid) {
 			const uint32_t *content =
 				(const uint32_t *) &info_frame->avi.sb[0];
+			/*we need turn on clock before programming AFMT block*/
+			REG_UPDATE(AFMT_CNTL, AFMT_AUDIO_CLOCK_EN, 1);
 
 			REG_WRITE(AFMT_AVI_INFO0, content[0]);
 
-- 
cgit v1.1


From 2c37e49a6bcd5e0c66963301e9feab63b5f928f3 Mon Sep 17 00:00:00 2001
From: Yongqiang Sun <yongqiang.sun@amd.com>
Date: Fri, 6 Apr 2018 21:38:10 -0400
Subject: drm/amd/display: Check SCRATCH reg to determine S3 resume.

Use lid state only to determine fast boot optimization is not enough.
For S3/Resume, due to bios isn't involved in boot, eDP wasn't
light up, while lid state is open, if do fast boot optimization,
eDP panel will skip enable link and result in black screen after boot.
And becasue of bios isn't involved, no matter UEFI or Legacy boot,
BIOS_SCRATCH_3 value should be 0, use this to determine the case.

Signed-off-by: Yongqiang Sun <yongqiang.sun@amd.com>
Reviewed-by: Charlene Liu <Charlene.Liu@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../amd/display/dc/dce110/dce110_hw_sequencer.c    | 33 ++++++++++++++++++----
 1 file changed, 28 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
index 4a4b3bc..bd34193 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
@@ -1516,18 +1516,41 @@ static struct dc_link *get_link_for_edp_not_in_use(
  */
 void dce110_enable_accelerated_mode(struct dc *dc, struct dc_state *context)
 {
-	/* check eDP lid state:
-	 * If lid is open, vbios already light up eDP, so we can leverage vbios and skip eDP
-	 * programming
+	/* check eDP lid state and BIOS_SCRATCH_3 to determine fast boot optimization
+	 * UEFI boot
+	 *				edp_active_status_from_scratch		fast boot optimization
+	 * S4/S5 resume:
+	 * Lid Open		true								true
+	 * Lid Close	false								false
+	 *
+	 * S3/ resume:
+	 * Lid Open		false								false
+	 * Lid Close	false								false
+	 *
+	 * Legacy boot:
+	 *				edp_active_status_from_scratch		fast boot optimization
+	 * S4/S resume:
+	 * Lid Open		true								true
+	 * Lid Close	true								false
+	 *
+	 * S3/ resume:
+	 * Lid Open		false								false
+	 * Lid Close	false								false
 	 */
+	struct dc_bios *dcb = dc->ctx->dc_bios;
 	bool lid_state_closed = is_eDP_lid_closed(context);
 	struct dc_link *edp_link_to_turnoff = NULL;
+	bool edp_active_status_from_scratch =
+			(dcb->funcs->get_vga_enabled_displays(dc->ctx->dc_bios) == ATOM_DISPLAY_LCD1_ACTIVE);
 
+	/*Lid open*/
 	if (!lid_state_closed) {
 		edp_link_to_turnoff = get_link_for_edp_not_in_use(dc, context);
 
-		/* if OS doesn't light up eDP and eDP link is available, we want to disable */
-		if (!edp_link_to_turnoff)
+		/* if OS doesn't light up eDP and eDP link is available, we want to disable
+		 * If resume from S4/S5, should optimization.
+		 */
+		if (!edp_link_to_turnoff && edp_active_status_from_scratch)
 			dc->apply_edp_fast_boot_optimization = true;
 	}
 
-- 
cgit v1.1


From 0a93dc7f595f43b621277ecfc05a44ed0c719a5f Mon Sep 17 00:00:00 2001
From: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Date: Thu, 29 Mar 2018 08:43:02 -0400
Subject: drm/amd/display: add rq/dlg/ttu to dtn log

Signed-off-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Reviewed-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dc_helper.c         |  59 ++++++++
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c  | 153 ++++++++++++++++++++-
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h  |  19 +--
 .../drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c  | 114 ++++++++++++++-
 drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h       |  20 +++
 drivers/gpu/drm/amd/display/dc/inc/reg_helper.h    |  56 ++++++++
 6 files changed, 401 insertions(+), 20 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dc_helper.c b/drivers/gpu/drm/amd/display/dc/dc_helper.c
index 48e1fcf5..bd0fda0 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_helper.c
+++ b/drivers/gpu/drm/amd/display/dc/dc_helper.c
@@ -117,6 +117,65 @@ uint32_t generic_reg_get5(const struct dc_context *ctx, uint32_t addr,
 	return reg_val;
 }
 
+uint32_t generic_reg_get6(const struct dc_context *ctx, uint32_t addr,
+		uint8_t shift1, uint32_t mask1, uint32_t *field_value1,
+		uint8_t shift2, uint32_t mask2, uint32_t *field_value2,
+		uint8_t shift3, uint32_t mask3, uint32_t *field_value3,
+		uint8_t shift4, uint32_t mask4, uint32_t *field_value4,
+		uint8_t shift5, uint32_t mask5, uint32_t *field_value5,
+		uint8_t shift6, uint32_t mask6, uint32_t *field_value6)
+{
+	uint32_t reg_val = dm_read_reg(ctx, addr);
+	*field_value1 = get_reg_field_value_ex(reg_val, mask1, shift1);
+	*field_value2 = get_reg_field_value_ex(reg_val, mask2, shift2);
+	*field_value3 = get_reg_field_value_ex(reg_val, mask3, shift3);
+	*field_value4 = get_reg_field_value_ex(reg_val, mask4, shift4);
+	*field_value5 = get_reg_field_value_ex(reg_val, mask5, shift5);
+	*field_value6 = get_reg_field_value_ex(reg_val, mask6, shift6);
+	return reg_val;
+}
+
+uint32_t generic_reg_get7(const struct dc_context *ctx, uint32_t addr,
+		uint8_t shift1, uint32_t mask1, uint32_t *field_value1,
+		uint8_t shift2, uint32_t mask2, uint32_t *field_value2,
+		uint8_t shift3, uint32_t mask3, uint32_t *field_value3,
+		uint8_t shift4, uint32_t mask4, uint32_t *field_value4,
+		uint8_t shift5, uint32_t mask5, uint32_t *field_value5,
+		uint8_t shift6, uint32_t mask6, uint32_t *field_value6,
+		uint8_t shift7, uint32_t mask7, uint32_t *field_value7)
+{
+	uint32_t reg_val = dm_read_reg(ctx, addr);
+	*field_value1 = get_reg_field_value_ex(reg_val, mask1, shift1);
+	*field_value2 = get_reg_field_value_ex(reg_val, mask2, shift2);
+	*field_value3 = get_reg_field_value_ex(reg_val, mask3, shift3);
+	*field_value4 = get_reg_field_value_ex(reg_val, mask4, shift4);
+	*field_value5 = get_reg_field_value_ex(reg_val, mask5, shift5);
+	*field_value6 = get_reg_field_value_ex(reg_val, mask6, shift6);
+	*field_value7 = get_reg_field_value_ex(reg_val, mask7, shift7);
+	return reg_val;
+}
+
+uint32_t generic_reg_get8(const struct dc_context *ctx, uint32_t addr,
+		uint8_t shift1, uint32_t mask1, uint32_t *field_value1,
+		uint8_t shift2, uint32_t mask2, uint32_t *field_value2,
+		uint8_t shift3, uint32_t mask3, uint32_t *field_value3,
+		uint8_t shift4, uint32_t mask4, uint32_t *field_value4,
+		uint8_t shift5, uint32_t mask5, uint32_t *field_value5,
+		uint8_t shift6, uint32_t mask6, uint32_t *field_value6,
+		uint8_t shift7, uint32_t mask7, uint32_t *field_value7,
+		uint8_t shift8, uint32_t mask8, uint32_t *field_value8)
+{
+	uint32_t reg_val = dm_read_reg(ctx, addr);
+	*field_value1 = get_reg_field_value_ex(reg_val, mask1, shift1);
+	*field_value2 = get_reg_field_value_ex(reg_val, mask2, shift2);
+	*field_value3 = get_reg_field_value_ex(reg_val, mask3, shift3);
+	*field_value4 = get_reg_field_value_ex(reg_val, mask4, shift4);
+	*field_value5 = get_reg_field_value_ex(reg_val, mask5, shift5);
+	*field_value6 = get_reg_field_value_ex(reg_val, mask6, shift6);
+	*field_value7 = get_reg_field_value_ex(reg_val, mask7, shift7);
+	*field_value8 = get_reg_field_value_ex(reg_val, mask8, shift8);
+	return reg_val;
+}
 /* note:  va version of this is pretty bad idea, since there is a output parameter pass by pointer
  * compiler won't be able to check for size match and is prone to stack corruption type of bugs
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
index 4ca9b6e..5806217 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
@@ -756,9 +756,159 @@ void min_set_viewport(
 		  PRI_VIEWPORT_Y_START_C, viewport_c->y);
 }
 
-void hubp1_read_state(struct dcn10_hubp *hubp1,
+void hubp1_read_state(struct hubp *hubp,
 		struct dcn_hubp_state *s)
 {
+	struct dcn10_hubp *hubp1 = TO_DCN10_HUBP(hubp);
+	struct _vcs_dpi_display_dlg_regs_st *dlg_attr = &s->dlg_attr;
+	struct _vcs_dpi_display_ttu_regs_st *ttu_attr = &s->ttu_attr;
+	struct _vcs_dpi_display_rq_regs_st *rq_regs = &s->rq_regs;
+
+	/* Requester */
+	REG_GET(HUBPRET_CONTROL,
+			DET_BUF_PLANE1_BASE_ADDRESS, &rq_regs->plane1_base_address);
+	REG_GET_4(DCN_EXPANSION_MODE,
+			DRQ_EXPANSION_MODE, &rq_regs->drq_expansion_mode,
+			PRQ_EXPANSION_MODE, &rq_regs->prq_expansion_mode,
+			MRQ_EXPANSION_MODE, &rq_regs->mrq_expansion_mode,
+			CRQ_EXPANSION_MODE, &rq_regs->crq_expansion_mode);
+	REG_GET_8(DCHUBP_REQ_SIZE_CONFIG,
+		CHUNK_SIZE, &rq_regs->rq_regs_l.chunk_size,
+		MIN_CHUNK_SIZE, &rq_regs->rq_regs_l.min_chunk_size,
+		META_CHUNK_SIZE, &rq_regs->rq_regs_l.meta_chunk_size,
+		MIN_META_CHUNK_SIZE, &rq_regs->rq_regs_l.min_meta_chunk_size,
+		DPTE_GROUP_SIZE, &rq_regs->rq_regs_l.dpte_group_size,
+		MPTE_GROUP_SIZE, &rq_regs->rq_regs_l.mpte_group_size,
+		SWATH_HEIGHT, &rq_regs->rq_regs_l.swath_height,
+		PTE_ROW_HEIGHT_LINEAR, &rq_regs->rq_regs_l.pte_row_height_linear);
+	REG_GET_8(DCHUBP_REQ_SIZE_CONFIG_C,
+		CHUNK_SIZE_C, &rq_regs->rq_regs_c.chunk_size,
+		MIN_CHUNK_SIZE_C, &rq_regs->rq_regs_c.min_chunk_size,
+		META_CHUNK_SIZE_C, &rq_regs->rq_regs_c.meta_chunk_size,
+		MIN_META_CHUNK_SIZE_C, &rq_regs->rq_regs_c.min_meta_chunk_size,
+		DPTE_GROUP_SIZE_C, &rq_regs->rq_regs_c.dpte_group_size,
+		MPTE_GROUP_SIZE_C, &rq_regs->rq_regs_c.mpte_group_size,
+		SWATH_HEIGHT_C, &rq_regs->rq_regs_c.swath_height,
+		PTE_ROW_HEIGHT_LINEAR_C, &rq_regs->rq_regs_c.pte_row_height_linear);
+
+	/* DLG - Per hubp */
+	REG_GET_2(BLANK_OFFSET_0,
+		REFCYC_H_BLANK_END, &dlg_attr->refcyc_h_blank_end,
+		DLG_V_BLANK_END, &dlg_attr->dlg_vblank_end);
+
+	REG_GET(BLANK_OFFSET_1,
+		MIN_DST_Y_NEXT_START, &dlg_attr->min_dst_y_next_start);
+
+	REG_GET(DST_DIMENSIONS,
+		REFCYC_PER_HTOTAL, &dlg_attr->refcyc_per_htotal);
+
+	REG_GET_2(DST_AFTER_SCALER,
+		REFCYC_X_AFTER_SCALER, &dlg_attr->refcyc_x_after_scaler,
+		DST_Y_AFTER_SCALER, &dlg_attr->dst_y_after_scaler);
+
+	if (REG(PREFETCH_SETTINS))
+		REG_GET_2(PREFETCH_SETTINS,
+			DST_Y_PREFETCH, &dlg_attr->dst_y_prefetch,
+			VRATIO_PREFETCH, &dlg_attr->vratio_prefetch);
+	else
+		REG_GET_2(PREFETCH_SETTINGS,
+			DST_Y_PREFETCH, &dlg_attr->dst_y_prefetch,
+			VRATIO_PREFETCH, &dlg_attr->vratio_prefetch);
+
+	REG_GET_2(VBLANK_PARAMETERS_0,
+		DST_Y_PER_VM_VBLANK, &dlg_attr->dst_y_per_vm_vblank,
+		DST_Y_PER_ROW_VBLANK, &dlg_attr->dst_y_per_row_vblank);
+
+	REG_GET(REF_FREQ_TO_PIX_FREQ,
+		REF_FREQ_TO_PIX_FREQ, &dlg_attr->ref_freq_to_pix_freq);
+
+	/* DLG - Per luma/chroma */
+	REG_GET(VBLANK_PARAMETERS_1,
+		REFCYC_PER_PTE_GROUP_VBLANK_L, &dlg_attr->refcyc_per_pte_group_vblank_l);
+
+	REG_GET(VBLANK_PARAMETERS_3,
+		REFCYC_PER_META_CHUNK_VBLANK_L, &dlg_attr->refcyc_per_meta_chunk_vblank_l);
+
+	if (REG(NOM_PARAMETERS_0))
+		REG_GET(NOM_PARAMETERS_0,
+			DST_Y_PER_PTE_ROW_NOM_L, &dlg_attr->dst_y_per_pte_row_nom_l);
+
+	if (REG(NOM_PARAMETERS_1))
+		REG_GET(NOM_PARAMETERS_1,
+			REFCYC_PER_PTE_GROUP_NOM_L, &dlg_attr->refcyc_per_pte_group_nom_l);
+
+	REG_GET(NOM_PARAMETERS_4,
+		DST_Y_PER_META_ROW_NOM_L, &dlg_attr->dst_y_per_meta_row_nom_l);
+
+	REG_GET(NOM_PARAMETERS_5,
+		REFCYC_PER_META_CHUNK_NOM_L, &dlg_attr->refcyc_per_meta_chunk_nom_l);
+
+	REG_GET_2(PER_LINE_DELIVERY_PRE,
+		REFCYC_PER_LINE_DELIVERY_PRE_L, &dlg_attr->refcyc_per_line_delivery_pre_l,
+		REFCYC_PER_LINE_DELIVERY_PRE_C, &dlg_attr->refcyc_per_line_delivery_pre_c);
+
+	REG_GET_2(PER_LINE_DELIVERY,
+		REFCYC_PER_LINE_DELIVERY_L, &dlg_attr->refcyc_per_line_delivery_l,
+		REFCYC_PER_LINE_DELIVERY_C, &dlg_attr->refcyc_per_line_delivery_c);
+
+	if (REG(PREFETCH_SETTINS_C))
+		REG_GET(PREFETCH_SETTINS_C,
+			VRATIO_PREFETCH_C, &dlg_attr->vratio_prefetch_c);
+	else
+		REG_GET(PREFETCH_SETTINGS_C,
+			VRATIO_PREFETCH_C, &dlg_attr->vratio_prefetch_c);
+
+	REG_GET(VBLANK_PARAMETERS_2,
+		REFCYC_PER_PTE_GROUP_VBLANK_C, &dlg_attr->refcyc_per_pte_group_vblank_c);
+
+	REG_GET(VBLANK_PARAMETERS_4,
+		REFCYC_PER_META_CHUNK_VBLANK_C, &dlg_attr->refcyc_per_meta_chunk_vblank_c);
+
+	if (REG(NOM_PARAMETERS_2))
+		REG_GET(NOM_PARAMETERS_2,
+			DST_Y_PER_PTE_ROW_NOM_C, &dlg_attr->dst_y_per_pte_row_nom_c);
+
+	if (REG(NOM_PARAMETERS_3))
+		REG_GET(NOM_PARAMETERS_3,
+			REFCYC_PER_PTE_GROUP_NOM_C, &dlg_attr->refcyc_per_pte_group_nom_c);
+
+	REG_GET(NOM_PARAMETERS_6,
+		DST_Y_PER_META_ROW_NOM_C, &dlg_attr->dst_y_per_meta_row_nom_c);
+
+	REG_GET(NOM_PARAMETERS_7,
+		REFCYC_PER_META_CHUNK_NOM_C, &dlg_attr->refcyc_per_meta_chunk_nom_c);
+
+	/* TTU - per hubp */
+	REG_GET_2(DCN_TTU_QOS_WM,
+		QoS_LEVEL_LOW_WM, &ttu_attr->qos_level_low_wm,
+		QoS_LEVEL_HIGH_WM, &ttu_attr->qos_level_high_wm);
+
+	REG_GET_2(DCN_GLOBAL_TTU_CNTL,
+		MIN_TTU_VBLANK, &ttu_attr->min_ttu_vblank,
+		QoS_LEVEL_FLIP, &ttu_attr->qos_level_flip);
+
+	/* TTU - per luma/chroma */
+	/* Assumed surf0 is luma and 1 is chroma */
+
+	REG_GET_3(DCN_SURF0_TTU_CNTL0,
+		REFCYC_PER_REQ_DELIVERY, &ttu_attr->refcyc_per_req_delivery_l,
+		QoS_LEVEL_FIXED, &ttu_attr->qos_level_fixed_l,
+		QoS_RAMP_DISABLE, &ttu_attr->qos_ramp_disable_l);
+
+	REG_GET(DCN_SURF0_TTU_CNTL1,
+		REFCYC_PER_REQ_DELIVERY_PRE,
+		&ttu_attr->refcyc_per_req_delivery_pre_l);
+
+	REG_GET_3(DCN_SURF1_TTU_CNTL0,
+		REFCYC_PER_REQ_DELIVERY, &ttu_attr->refcyc_per_req_delivery_c,
+		QoS_LEVEL_FIXED, &ttu_attr->qos_level_fixed_c,
+		QoS_RAMP_DISABLE, &ttu_attr->qos_ramp_disable_c);
+
+	REG_GET(DCN_SURF1_TTU_CNTL1,
+		REFCYC_PER_REQ_DELIVERY_PRE,
+		&ttu_attr->refcyc_per_req_delivery_pre_c);
+
+	/* Rest of hubp */
 	REG_GET(DCSURF_SURFACE_CONFIG,
 			SURFACE_PIXEL_FORMAT, &s->pixel_format);
 
@@ -956,6 +1106,7 @@ static struct hubp_funcs dcn10_hubp_funcs = {
 	.hubp_disconnect = hubp1_disconnect,
 	.hubp_clk_cntl = hubp1_clk_cntl,
 	.hubp_vtg_sel = hubp1_vtg_sel,
+	.hubp_read_state = hubp1_read_state,
 };
 
 /*****************************************/
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
index e0d6d32..920ae3a 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
@@ -690,24 +690,7 @@ void dcn10_hubp_construct(
 	const struct dcn_mi_shift *hubp_shift,
 	const struct dcn_mi_mask *hubp_mask);
 
-
-struct dcn_hubp_state {
-	uint32_t pixel_format;
-	uint32_t inuse_addr_hi;
-	uint32_t viewport_width;
-	uint32_t viewport_height;
-	uint32_t rotation_angle;
-	uint32_t h_mirror_en;
-	uint32_t sw_mode;
-	uint32_t dcc_en;
-	uint32_t blank_en;
-	uint32_t underflow_status;
-	uint32_t ttu_disable;
-	uint32_t min_ttu_vblank;
-	uint32_t qos_level_low_wm;
-	uint32_t qos_level_high_wm;
-};
-void hubp1_read_state(struct dcn10_hubp *hubp1,
+void hubp1_read_state(struct hubp *hubp,
 		struct dcn_hubp_state *s);
 
 enum cursor_pitch hubp1_get_cursor_pitch(unsigned int pitch);
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
index a6cf9ad..7dd130d 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -112,6 +112,104 @@ void dcn10_log_hubbub_state(struct dc *dc)
 	DTN_INFO("\n");
 }
 
+static void print_rq_dlg_ttu_regs(struct dc_context *dc_ctx, struct dcn_hubp_state *s)
+{
+	struct _vcs_dpi_display_dlg_regs_st *dlg_regs = &s->dlg_attr;
+	struct _vcs_dpi_display_ttu_regs_st *ttu_regs = &s->ttu_attr;
+	struct _vcs_dpi_display_rq_regs_st *rq_regs = &s->rq_regs;
+
+	DTN_INFO("========Requester========\n");
+	DTN_INFO("drq_expansion_mode      = 0x%0x\n", rq_regs->drq_expansion_mode);
+	DTN_INFO("prq_expansion_mode      = 0x%0x\n", rq_regs->prq_expansion_mode);
+	DTN_INFO("mrq_expansion_mode      = 0x%0x\n", rq_regs->mrq_expansion_mode);
+	DTN_INFO("crq_expansion_mode      = 0x%0x\n", rq_regs->crq_expansion_mode);
+	DTN_INFO("plane1_base_address     = 0x%0x\n", rq_regs->plane1_base_address);
+	DTN_INFO("==<LUMA>==\n");
+	DTN_INFO("chunk_size              = 0x%0x\n", rq_regs->rq_regs_l.chunk_size);
+	DTN_INFO("min_chunk_size          = 0x%0x\n", rq_regs->rq_regs_l.min_chunk_size);
+	DTN_INFO("meta_chunk_size         = 0x%0x\n", rq_regs->rq_regs_l.meta_chunk_size);
+	DTN_INFO("min_meta_chunk_size     = 0x%0x\n", rq_regs->rq_regs_l.min_meta_chunk_size);
+	DTN_INFO("dpte_group_size         = 0x%0x\n", rq_regs->rq_regs_l.dpte_group_size);
+	DTN_INFO("mpte_group_size         = 0x%0x\n", rq_regs->rq_regs_l.mpte_group_size);
+	DTN_INFO("swath_height            = 0x%0x\n", rq_regs->rq_regs_l.swath_height);
+	DTN_INFO("pte_row_height_linear   = 0x%0x\n", rq_regs->rq_regs_l.pte_row_height_linear);
+	DTN_INFO("==<CHROMA>==\n");
+	DTN_INFO("chunk_size              = 0x%0x\n", rq_regs->rq_regs_c.chunk_size);
+	DTN_INFO("min_chunk_size          = 0x%0x\n", rq_regs->rq_regs_c.min_chunk_size);
+	DTN_INFO("meta_chunk_size         = 0x%0x\n", rq_regs->rq_regs_c.meta_chunk_size);
+	DTN_INFO("min_meta_chunk_size     = 0x%0x\n", rq_regs->rq_regs_c.min_meta_chunk_size);
+	DTN_INFO("dpte_group_size         = 0x%0x\n", rq_regs->rq_regs_c.dpte_group_size);
+	DTN_INFO("mpte_group_size         = 0x%0x\n", rq_regs->rq_regs_c.mpte_group_size);
+	DTN_INFO("swath_height            = 0x%0x\n", rq_regs->rq_regs_c.swath_height);
+	DTN_INFO("pte_row_height_linear   = 0x%0x\n", rq_regs->rq_regs_c.pte_row_height_linear);
+
+	DTN_INFO("========DLG========\n");
+	DTN_INFO("refcyc_h_blank_end                  = 0x%0x\n", dlg_regs->refcyc_h_blank_end);
+	DTN_INFO("dlg_vblank_end                      = 0x%0x\n", dlg_regs->dlg_vblank_end);
+	DTN_INFO("min_dst_y_next_start                = 0x%0x\n", dlg_regs->min_dst_y_next_start);
+	DTN_INFO("refcyc_per_htotal                   = 0x%0x\n", dlg_regs->refcyc_per_htotal);
+	DTN_INFO("refcyc_x_after_scaler               = 0x%0x\n", dlg_regs->refcyc_x_after_scaler);
+	DTN_INFO("dst_y_after_scaler                  = 0x%0x\n", dlg_regs->dst_y_after_scaler);
+	DTN_INFO("dst_y_prefetch                      = 0x%0x\n", dlg_regs->dst_y_prefetch);
+	DTN_INFO("dst_y_per_vm_vblank                 = 0x%0x\n", dlg_regs->dst_y_per_vm_vblank);
+	DTN_INFO("dst_y_per_row_vblank                = 0x%0x\n", dlg_regs->dst_y_per_row_vblank);
+	DTN_INFO("dst_y_per_vm_flip                   = 0x%0x\n", dlg_regs->dst_y_per_vm_flip);
+	DTN_INFO("dst_y_per_row_flip                  = 0x%0x\n", dlg_regs->dst_y_per_row_flip);
+	DTN_INFO("ref_freq_to_pix_freq                = 0x%0x\n", dlg_regs->ref_freq_to_pix_freq);
+	DTN_INFO("vratio_prefetch                     = 0x%0x\n", dlg_regs->vratio_prefetch);
+	DTN_INFO("vratio_prefetch_c                   = 0x%0x\n", dlg_regs->vratio_prefetch_c);
+	DTN_INFO("refcyc_per_pte_group_vblank_l       = 0x%0x\n", dlg_regs->refcyc_per_pte_group_vblank_l);
+	DTN_INFO("refcyc_per_pte_group_vblank_c       = 0x%0x\n", dlg_regs->refcyc_per_pte_group_vblank_c);
+	DTN_INFO("refcyc_per_meta_chunk_vblank_l      = 0x%0x\n", dlg_regs->refcyc_per_meta_chunk_vblank_l);
+	DTN_INFO("refcyc_per_meta_chunk_vblank_c      = 0x%0x\n", dlg_regs->refcyc_per_meta_chunk_vblank_c);
+	DTN_INFO("refcyc_per_pte_group_flip_l         = 0x%0x\n", dlg_regs->refcyc_per_pte_group_flip_l);
+	DTN_INFO("refcyc_per_pte_group_flip_c         = 0x%0x\n", dlg_regs->refcyc_per_pte_group_flip_c);
+	DTN_INFO("refcyc_per_meta_chunk_flip_l        = 0x%0x\n", dlg_regs->refcyc_per_meta_chunk_flip_l);
+	DTN_INFO("refcyc_per_meta_chunk_flip_c        = 0x%0x\n", dlg_regs->refcyc_per_meta_chunk_flip_c);
+	DTN_INFO("dst_y_per_pte_row_nom_l             = 0x%0x\n", dlg_regs->dst_y_per_pte_row_nom_l);
+	DTN_INFO("dst_y_per_pte_row_nom_c             = 0x%0x\n", dlg_regs->dst_y_per_pte_row_nom_c);
+	DTN_INFO("refcyc_per_pte_group_nom_l          = 0x%0x\n", dlg_regs->refcyc_per_pte_group_nom_l);
+	DTN_INFO("refcyc_per_pte_group_nom_c          = 0x%0x\n", dlg_regs->refcyc_per_pte_group_nom_c);
+	DTN_INFO("dst_y_per_meta_row_nom_l            = 0x%0x\n", dlg_regs->dst_y_per_meta_row_nom_l);
+	DTN_INFO("dst_y_per_meta_row_nom_c            = 0x%0x\n", dlg_regs->dst_y_per_meta_row_nom_c);
+	DTN_INFO("refcyc_per_meta_chunk_nom_l         = 0x%0x\n", dlg_regs->refcyc_per_meta_chunk_nom_l);
+	DTN_INFO("refcyc_per_meta_chunk_nom_c         = 0x%0x\n", dlg_regs->refcyc_per_meta_chunk_nom_c);
+	DTN_INFO("refcyc_per_line_delivery_pre_l      = 0x%0x\n", dlg_regs->refcyc_per_line_delivery_pre_l);
+	DTN_INFO("refcyc_per_line_delivery_pre_c      = 0x%0x\n", dlg_regs->refcyc_per_line_delivery_pre_c);
+	DTN_INFO("refcyc_per_line_delivery_l          = 0x%0x\n", dlg_regs->refcyc_per_line_delivery_l);
+	DTN_INFO("refcyc_per_line_delivery_c          = 0x%0x\n", dlg_regs->refcyc_per_line_delivery_c);
+	DTN_INFO("chunk_hdl_adjust_cur0               = 0x%0x\n", dlg_regs->chunk_hdl_adjust_cur0);
+	DTN_INFO("dst_y_offset_cur1                   = 0x%0x\n", dlg_regs->dst_y_offset_cur1);
+	DTN_INFO("chunk_hdl_adjust_cur1               = 0x%0x\n", dlg_regs->chunk_hdl_adjust_cur1);
+	DTN_INFO("vready_after_vcount0                = 0x%0x\n", dlg_regs->vready_after_vcount0);
+	DTN_INFO("dst_y_delta_drq_limit               = 0x%0x\n", dlg_regs->dst_y_delta_drq_limit);
+	DTN_INFO("xfc_reg_transfer_delay              = 0x%0x\n", dlg_regs->xfc_reg_transfer_delay);
+	DTN_INFO("xfc_reg_precharge_delay             = 0x%0x\n", dlg_regs->xfc_reg_precharge_delay);
+	DTN_INFO("xfc_reg_remote_surface_flip_latency = 0x%0x\n", dlg_regs->xfc_reg_remote_surface_flip_latency);
+
+	DTN_INFO("========TTU========\n");
+	DTN_INFO("qos_level_low_wm                  = 0x%0x\n", ttu_regs->qos_level_low_wm);
+	DTN_INFO("qos_level_high_wm                 = 0x%0x\n", ttu_regs->qos_level_high_wm);
+	DTN_INFO("min_ttu_vblank                    = 0x%0x\n", ttu_regs->min_ttu_vblank);
+	DTN_INFO("qos_level_flip                    = 0x%0x\n", ttu_regs->qos_level_flip);
+	DTN_INFO("refcyc_per_req_delivery_pre_l     = 0x%0x\n", ttu_regs->refcyc_per_req_delivery_pre_l);
+	DTN_INFO("refcyc_per_req_delivery_l         = 0x%0x\n", ttu_regs->refcyc_per_req_delivery_l);
+	DTN_INFO("refcyc_per_req_delivery_pre_c     = 0x%0x\n", ttu_regs->refcyc_per_req_delivery_pre_c);
+	DTN_INFO("refcyc_per_req_delivery_c         = 0x%0x\n", ttu_regs->refcyc_per_req_delivery_c);
+	DTN_INFO("refcyc_per_req_delivery_cur0      = 0x%0x\n", ttu_regs->refcyc_per_req_delivery_cur0);
+	DTN_INFO("refcyc_per_req_delivery_pre_cur0  = 0x%0x\n", ttu_regs->refcyc_per_req_delivery_pre_cur0);
+	DTN_INFO("refcyc_per_req_delivery_cur1      = 0x%0x\n", ttu_regs->refcyc_per_req_delivery_cur1);
+	DTN_INFO("refcyc_per_req_delivery_pre_cur1  = 0x%0x\n", ttu_regs->refcyc_per_req_delivery_pre_cur1);
+	DTN_INFO("qos_level_fixed_l                 = 0x%0x\n", ttu_regs->qos_level_fixed_l);
+	DTN_INFO("qos_ramp_disable_l                = 0x%0x\n", ttu_regs->qos_ramp_disable_l);
+	DTN_INFO("qos_level_fixed_c                 = 0x%0x\n", ttu_regs->qos_level_fixed_c);
+	DTN_INFO("qos_ramp_disable_c                = 0x%0x\n", ttu_regs->qos_ramp_disable_c);
+	DTN_INFO("qos_level_fixed_cur0              = 0x%0x\n", ttu_regs->qos_level_fixed_cur0);
+	DTN_INFO("qos_ramp_disable_cur0             = 0x%0x\n", ttu_regs->qos_ramp_disable_cur0);
+	DTN_INFO("qos_level_fixed_cur1              = 0x%0x\n", ttu_regs->qos_level_fixed_cur1);
+	DTN_INFO("qos_ramp_disable_cur1             = 0x%0x\n", ttu_regs->qos_ramp_disable_cur1);
+}
+
 void dcn10_log_hw_state(struct dc *dc)
 {
 	struct dc_context *dc_ctx = dc->ctx;
@@ -129,7 +227,7 @@ void dcn10_log_hw_state(struct dc *dc)
 		struct hubp *hubp = pool->hubps[i];
 		struct dcn_hubp_state s;
 
-		hubp1_read_state(TO_DCN10_HUBP(hubp), &s);
+		hubp->funcs->hubp_read_state(hubp, &s);
 
 		DTN_INFO("[%2d]:  %5xh  %6xh  %5d  %6d  %2xh  %2xh  %6xh"
 				"  %6d  %8d  %7d  %8xh",
@@ -201,6 +299,20 @@ void dcn10_log_hw_state(struct dc *dc)
 	}
 	DTN_INFO("\n");
 
+	for (i = 0; i < pool->pipe_count; i++) {
+		struct hubp *hubp = pool->hubps[i];
+		struct dcn_hubp_state s = {0};
+
+		if (!dc->current_state->res_ctx.pipe_ctx[i].stream)
+			continue;
+
+		hubp->funcs->hubp_read_state(hubp, &s);
+		DTN_INFO("RQ-DLG-TTU registers for HUBP%d:\n", i);
+		print_rq_dlg_ttu_regs(dc_ctx, &s);
+		DTN_INFO("\n");
+	}
+	DTN_INFO("\n");
+
 	log_mpc_crc(dc);
 
 	DTN_INFO_END();
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h b/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h
index 9ced254..3866147 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h
@@ -56,6 +56,25 @@ struct hubp {
 	bool power_gated;
 };
 
+struct dcn_hubp_state {
+	struct _vcs_dpi_display_dlg_regs_st dlg_attr;
+	struct _vcs_dpi_display_ttu_regs_st ttu_attr;
+	struct _vcs_dpi_display_rq_regs_st rq_regs;
+	uint32_t pixel_format;
+	uint32_t inuse_addr_hi;
+	uint32_t viewport_width;
+	uint32_t viewport_height;
+	uint32_t rotation_angle;
+	uint32_t h_mirror_en;
+	uint32_t sw_mode;
+	uint32_t dcc_en;
+	uint32_t blank_en;
+	uint32_t underflow_status;
+	uint32_t ttu_disable;
+	uint32_t min_ttu_vblank;
+	uint32_t qos_level_low_wm;
+	uint32_t qos_level_high_wm;
+};
 
 struct hubp_funcs {
 	void (*hubp_setup)(
@@ -121,6 +140,7 @@ struct hubp_funcs {
 
 	void (*hubp_clk_cntl)(struct hubp *hubp, bool enable);
 	void (*hubp_vtg_sel)(struct hubp *hubp, uint32_t otg_inst);
+	void (*hubp_read_state)(struct hubp *hubp, struct dcn_hubp_state *s);
 
 };
 
diff --git a/drivers/gpu/drm/amd/display/dc/inc/reg_helper.h b/drivers/gpu/drm/amd/display/dc/inc/reg_helper.h
index 77eb728..3306e7b 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/reg_helper.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/reg_helper.h
@@ -183,6 +183,36 @@
 				FN(reg_name, f4), v4, \
 				FN(reg_name, f5), v5)
 
+#define REG_GET_6(reg_name, f1, v1, f2, v2, f3, v3, f4, v4, f5, v5, f6, v6)	\
+		generic_reg_get6(CTX, REG(reg_name), \
+				FN(reg_name, f1), v1, \
+				FN(reg_name, f2), v2, \
+				FN(reg_name, f3), v3, \
+				FN(reg_name, f4), v4, \
+				FN(reg_name, f5), v5, \
+				FN(reg_name, f6), v6)
+
+#define REG_GET_7(reg_name, f1, v1, f2, v2, f3, v3, f4, v4, f5, v5, f6, v6, f7, v7)	\
+		generic_reg_get7(CTX, REG(reg_name), \
+				FN(reg_name, f1), v1, \
+				FN(reg_name, f2), v2, \
+				FN(reg_name, f3), v3, \
+				FN(reg_name, f4), v4, \
+				FN(reg_name, f5), v5, \
+				FN(reg_name, f6), v6, \
+				FN(reg_name, f7), v7)
+
+#define REG_GET_8(reg_name, f1, v1, f2, v2, f3, v3, f4, v4, f5, v5, f6, v6, f7, v7, f8, v8)	\
+		generic_reg_get8(CTX, REG(reg_name), \
+				FN(reg_name, f1), v1, \
+				FN(reg_name, f2), v2, \
+				FN(reg_name, f3), v3, \
+				FN(reg_name, f4), v4, \
+				FN(reg_name, f5), v5, \
+				FN(reg_name, f6), v6, \
+				FN(reg_name, f7), v7, \
+				FN(reg_name, f8), v8)
+
 /* macro to poll and wait for a register field to read back given value */
 
 #define REG_WAIT(reg_name, field, val, delay_between_poll_us, max_try)	\
@@ -389,4 +419,30 @@ uint32_t generic_reg_get5(const struct dc_context *ctx, uint32_t addr,
 		uint8_t shift4, uint32_t mask4, uint32_t *field_value4,
 		uint8_t shift5, uint32_t mask5, uint32_t *field_value5);
 
+uint32_t generic_reg_get6(const struct dc_context *ctx, uint32_t addr,
+		uint8_t shift1, uint32_t mask1, uint32_t *field_value1,
+		uint8_t shift2, uint32_t mask2, uint32_t *field_value2,
+		uint8_t shift3, uint32_t mask3, uint32_t *field_value3,
+		uint8_t shift4, uint32_t mask4, uint32_t *field_value4,
+		uint8_t shift5, uint32_t mask5, uint32_t *field_value5,
+		uint8_t shift6, uint32_t mask6, uint32_t *field_value6);
+
+uint32_t generic_reg_get7(const struct dc_context *ctx, uint32_t addr,
+		uint8_t shift1, uint32_t mask1, uint32_t *field_value1,
+		uint8_t shift2, uint32_t mask2, uint32_t *field_value2,
+		uint8_t shift3, uint32_t mask3, uint32_t *field_value3,
+		uint8_t shift4, uint32_t mask4, uint32_t *field_value4,
+		uint8_t shift5, uint32_t mask5, uint32_t *field_value5,
+		uint8_t shift6, uint32_t mask6, uint32_t *field_value6,
+		uint8_t shift7, uint32_t mask7, uint32_t *field_value7);
+
+uint32_t generic_reg_get8(const struct dc_context *ctx, uint32_t addr,
+		uint8_t shift1, uint32_t mask1, uint32_t *field_value1,
+		uint8_t shift2, uint32_t mask2, uint32_t *field_value2,
+		uint8_t shift3, uint32_t mask3, uint32_t *field_value3,
+		uint8_t shift4, uint32_t mask4, uint32_t *field_value4,
+		uint8_t shift5, uint32_t mask5, uint32_t *field_value5,
+		uint8_t shift6, uint32_t mask6, uint32_t *field_value6,
+		uint8_t shift7, uint32_t mask7, uint32_t *field_value7,
+		uint8_t shift8, uint32_t mask8, uint32_t *field_value8);
 #endif /* DRIVERS_GPU_DRM_AMD_DC_DEV_DC_INC_REG_HELPER_H_ */
-- 
cgit v1.1


From a47654633596a63f14a9035b9c762f8aaf1e00a3 Mon Sep 17 00:00:00 2001
From: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Date: Wed, 4 Apr 2018 16:03:38 -0400
Subject: drm/amd/display: add calculated clock logging to DTN

Signed-off-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Reviewed-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c | 11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
index 7dd130d..e547f46 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -311,7 +311,16 @@ void dcn10_log_hw_state(struct dc *dc)
 		print_rq_dlg_ttu_regs(dc_ctx, &s);
 		DTN_INFO("\n");
 	}
-	DTN_INFO("\n");
+
+	DTN_INFO("\nCALCULATED Clocks: dcfclk_khz:%d  dcfclk_deep_sleep_khz:%d  dispclk_khz:%d\n"
+		"dppclk_khz:%d  max_supported_dppclk_khz:%d  fclk_khz:%d  socclk_khz:%d\n\n",
+			dc->current_state->bw.dcn.calc_clk.dcfclk_khz,
+			dc->current_state->bw.dcn.calc_clk.dcfclk_deep_sleep_khz,
+			dc->current_state->bw.dcn.calc_clk.dispclk_khz,
+			dc->current_state->bw.dcn.calc_clk.dppclk_khz,
+			dc->current_state->bw.dcn.calc_clk.max_supported_dppclk_khz,
+			dc->current_state->bw.dcn.calc_clk.fclk_khz,
+			dc->current_state->bw.dcn.calc_clk.socclk_khz);
 
 	log_mpc_crc(dc);
 
-- 
cgit v1.1


From ad019f7b6db893271d13148d6d80001d0c23cdf9 Mon Sep 17 00:00:00 2001
From: Yue Hin Lau <Yuehin.Lau@amd.com>
Date: Mon, 9 Apr 2018 14:46:32 -0400
Subject: drm/amd/display: add missing colorspace for set black color

Signed-off-by: Yue Hin Lau <Yuehin.Lau@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../gpu/drm/amd/display/dc/core/dc_hw_sequencer.c   | 21 ++++++++++++++++++++-
 1 file changed, 20 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c
index 481f692..83d1215 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c
@@ -208,6 +208,7 @@ void color_space_to_black_color(
 	case COLOR_SPACE_YCBCR709:
 	case COLOR_SPACE_YCBCR601_LIMITED:
 	case COLOR_SPACE_YCBCR709_LIMITED:
+	case COLOR_SPACE_2020_YCBCR:
 		*black_color = black_color_format[BLACK_COLOR_FORMAT_YUV_CV];
 		break;
 
@@ -216,7 +217,25 @@ void color_space_to_black_color(
 			black_color_format[BLACK_COLOR_FORMAT_RGB_LIMITED];
 		break;
 
-	default:
+	/**
+	 * Remove default and add case for all color space
+	 * so when we forget to add new color space
+	 * compiler will give a warning
+	 */
+	case COLOR_SPACE_UNKNOWN:
+	case COLOR_SPACE_SRGB:
+	case COLOR_SPACE_XR_RGB:
+	case COLOR_SPACE_MSREF_SCRGB:
+	case COLOR_SPACE_XV_YCC_709:
+	case COLOR_SPACE_XV_YCC_601:
+	case COLOR_SPACE_2020_RGB_FULLRANGE:
+	case COLOR_SPACE_2020_RGB_LIMITEDRANGE:
+	case COLOR_SPACE_ADOBERGB:
+	case COLOR_SPACE_DCIP3:
+	case COLOR_SPACE_DISPLAYNATIVE:
+	case COLOR_SPACE_DOLBYVISION:
+	case COLOR_SPACE_APPCTRL:
+	case COLOR_SPACE_CUSTOMPOINTS:
 		/* fefault is sRGB black (full range). */
 		*black_color =
 			black_color_format[BLACK_COLOR_FORMAT_RGB_FULLRANGE];
-- 
cgit v1.1


From f0c0761b38ac30b04d4fed436ff10e894ec0e525 Mon Sep 17 00:00:00 2001
From: Yongqiang Sun <yongqiang.sun@amd.com>
Date: Mon, 9 Apr 2018 16:15:20 -0400
Subject: drm/amd/display: Use dig enable to determine fast boot optimization.

Linux doesn't know lid state, better to check dig enable
value from register.

Signed-off-by: Yongqiang Sun <yongqiang.sun@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dc_stream.h         |  1 -
 .../gpu/drm/amd/display/dc/dce/dce_link_encoder.c  |  6 ++-
 .../gpu/drm/amd/display/dc/dce/dce_link_encoder.h  |  2 +
 .../amd/display/dc/dce110/dce110_hw_sequencer.c    | 47 +++++++---------------
 .../gpu/drm/amd/display/dc/inc/hw/link_encoder.h   |  1 +
 5 files changed, 21 insertions(+), 36 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dc_stream.h b/drivers/gpu/drm/amd/display/dc/dc_stream.h
index 11b3433..d7e6d53 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_stream.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_stream.h
@@ -92,7 +92,6 @@ struct dc_stream_state {
 	int phy_pix_clk;
 	enum signal_type signal;
 	bool dpms_off;
-	bool lid_state_closed;
 
 	struct dc_stream_status status;
 
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_link_encoder.c b/drivers/gpu/drm/amd/display/dc/dce/dce_link_encoder.c
index 8167cad..dbe3b26 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_link_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_link_encoder.c
@@ -113,6 +113,7 @@ static const struct link_encoder_funcs dce110_lnk_enc_funcs = {
 	.connect_dig_be_to_fe = dce110_link_encoder_connect_dig_be_to_fe,
 	.enable_hpd = dce110_link_encoder_enable_hpd,
 	.disable_hpd = dce110_link_encoder_disable_hpd,
+	.is_dig_enabled = dce110_is_dig_enabled,
 	.destroy = dce110_link_encoder_destroy
 };
 
@@ -535,8 +536,9 @@ void dce110_psr_program_secondary_packet(struct link_encoder *enc,
 		DP_SEC_GSP0_PRIORITY, 1);
 }
 
-static bool is_dig_enabled(const struct dce110_link_encoder *enc110)
+bool dce110_is_dig_enabled(struct link_encoder *enc)
 {
+	struct dce110_link_encoder *enc110 = TO_DCE110_LINK_ENC(enc);
 	uint32_t value;
 
 	REG_GET(DIG_BE_EN_CNTL, DIG_ENABLE, &value);
@@ -1031,7 +1033,7 @@ void dce110_link_encoder_disable_output(
 	struct bp_transmitter_control cntl = { 0 };
 	enum bp_result result;
 
-	if (!is_dig_enabled(enc110)) {
+	if (!dce110_is_dig_enabled(enc)) {
 		/* OF_SKIP_POWER_DOWN_INACTIVE_ENCODER */
 		return;
 	}
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_link_encoder.h b/drivers/gpu/drm/amd/display/dc/dce/dce_link_encoder.h
index 0ec3433..3470694 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_link_encoder.h
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_link_encoder.h
@@ -263,4 +263,6 @@ void dce110_psr_program_dp_dphy_fast_training(struct link_encoder *enc,
 void dce110_psr_program_secondary_packet(struct link_encoder *enc,
 			unsigned int sdp_transmit_line_num_deadline);
 
+bool dce110_is_dig_enabled(struct link_encoder *enc);
+
 #endif /* __DC_LINK_ENCODER__DCE110_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
index bd34193..e70ccb9 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
@@ -1471,15 +1471,15 @@ static void disable_vga_and_power_gate_all_controllers(
 	}
 }
 
-static bool is_eDP_lid_closed(struct dc_state *context)
+static struct dc_link *get_link_for_edp(struct dc *dc)
 {
 	int i;
 
-	for (i = 0; i < context->stream_count; i++) {
-		if (context->streams[i]->signal == SIGNAL_TYPE_EDP)
-			return context->streams[i]->lid_state_closed;
+	for (i = 0; i < dc->link_count; i++) {
+		if (dc->links[i]->connector_signal == SIGNAL_TYPE_EDP)
+			return dc->links[i];
 	}
-	return false;
+	return NULL;
 }
 
 static struct dc_link *get_link_for_edp_not_in_use(
@@ -1516,41 +1516,22 @@ static struct dc_link *get_link_for_edp_not_in_use(
  */
 void dce110_enable_accelerated_mode(struct dc *dc, struct dc_state *context)
 {
-	/* check eDP lid state and BIOS_SCRATCH_3 to determine fast boot optimization
-	 * UEFI boot
-	 *				edp_active_status_from_scratch		fast boot optimization
-	 * S4/S5 resume:
-	 * Lid Open		true								true
-	 * Lid Close	false								false
-	 *
-	 * S3/ resume:
-	 * Lid Open		false								false
-	 * Lid Close	false								false
-	 *
-	 * Legacy boot:
-	 *				edp_active_status_from_scratch		fast boot optimization
-	 * S4/S resume:
-	 * Lid Open		true								true
-	 * Lid Close	true								false
-	 *
-	 * S3/ resume:
-	 * Lid Open		false								false
-	 * Lid Close	false								false
-	 */
-	struct dc_bios *dcb = dc->ctx->dc_bios;
-	bool lid_state_closed = is_eDP_lid_closed(context);
 	struct dc_link *edp_link_to_turnoff = NULL;
-	bool edp_active_status_from_scratch =
-			(dcb->funcs->get_vga_enabled_displays(dc->ctx->dc_bios) == ATOM_DISPLAY_LCD1_ACTIVE);
+	struct dc_link *edp_link = get_link_for_edp(dc);
+	bool can_eDP_fast_boot_optimize = false;
+
+	if (edp_link) {
+		can_eDP_fast_boot_optimize =
+				edp_link->link_enc->funcs->is_dig_enabled(edp_link->link_enc);
+	}
 
-	/*Lid open*/
-	if (!lid_state_closed) {
+	if (can_eDP_fast_boot_optimize) {
 		edp_link_to_turnoff = get_link_for_edp_not_in_use(dc, context);
 
 		/* if OS doesn't light up eDP and eDP link is available, we want to disable
 		 * If resume from S4/S5, should optimization.
 		 */
-		if (!edp_link_to_turnoff && edp_active_status_from_scratch)
+		if (!edp_link_to_turnoff)
 			dc->apply_edp_fast_boot_optimization = true;
 	}
 
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/link_encoder.h b/drivers/gpu/drm/amd/display/dc/inc/hw/link_encoder.h
index 54d8a13..cf6df2e 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/link_encoder.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/link_encoder.h
@@ -149,6 +149,7 @@ struct link_encoder_funcs {
 		bool connect);
 	void (*enable_hpd)(struct link_encoder *enc);
 	void (*disable_hpd)(struct link_encoder *enc);
+	bool (*is_dig_enabled)(struct link_encoder *enc);
 	void (*destroy)(struct link_encoder **enc);
 };
 
-- 
cgit v1.1


From a906dbb1e20f5791d728c7d9e2366b8acb4f1bb2 Mon Sep 17 00:00:00 2001
From: Chunming Zhou <david1.zhou@amd.com>
Date: Mon, 16 Apr 2018 17:57:19 +0800
Subject: drm/amdgpu: add amdgpu_bo_param
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

amdgpu_bo_create has too many parameters, and used in
too many places. Collect them to one structure.

Signed-off-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 75 +++++++++++++++++-------------
 drivers/gpu/drm/amd/amdgpu/amdgpu_object.h |  9 ++++
 2 files changed, 51 insertions(+), 33 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
index 24f582c..b33a7fd 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
@@ -341,27 +341,25 @@ fail:
 	return false;
 }
 
-static int amdgpu_bo_do_create(struct amdgpu_device *adev, unsigned long size,
-			       int byte_align, u32 domain,
-			       u64 flags, enum ttm_bo_type type,
-			       struct reservation_object *resv,
+static int amdgpu_bo_do_create(struct amdgpu_device *adev,
+			       struct amdgpu_bo_param *bp,
 			       struct amdgpu_bo **bo_ptr)
 {
 	struct ttm_operation_ctx ctx = {
-		.interruptible = (type != ttm_bo_type_kernel),
+		.interruptible = (bp->type != ttm_bo_type_kernel),
 		.no_wait_gpu = false,
-		.resv = resv,
+		.resv = bp->resv,
 		.flags = TTM_OPT_FLAG_ALLOW_RES_EVICT
 	};
 	struct amdgpu_bo *bo;
-	unsigned long page_align;
+	unsigned long page_align, size = bp->size;
 	size_t acc_size;
 	int r;
 
-	page_align = roundup(byte_align, PAGE_SIZE) >> PAGE_SHIFT;
+	page_align = roundup(bp->byte_align, PAGE_SIZE) >> PAGE_SHIFT;
 	size = ALIGN(size, PAGE_SIZE);
 
-	if (!amdgpu_bo_validate_size(adev, size, domain))
+	if (!amdgpu_bo_validate_size(adev, size, bp->domain))
 		return -ENOMEM;
 
 	*bo_ptr = NULL;
@@ -375,18 +373,18 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev, unsigned long size,
 	drm_gem_private_object_init(adev->ddev, &bo->gem_base, size);
 	INIT_LIST_HEAD(&bo->shadow_list);
 	INIT_LIST_HEAD(&bo->va);
-	bo->preferred_domains = domain & (AMDGPU_GEM_DOMAIN_VRAM |
-					 AMDGPU_GEM_DOMAIN_GTT |
-					 AMDGPU_GEM_DOMAIN_CPU |
-					 AMDGPU_GEM_DOMAIN_GDS |
-					 AMDGPU_GEM_DOMAIN_GWS |
-					 AMDGPU_GEM_DOMAIN_OA);
+	bo->preferred_domains = bp->domain & (AMDGPU_GEM_DOMAIN_VRAM |
+					      AMDGPU_GEM_DOMAIN_GTT |
+					      AMDGPU_GEM_DOMAIN_CPU |
+					      AMDGPU_GEM_DOMAIN_GDS |
+					      AMDGPU_GEM_DOMAIN_GWS |
+					      AMDGPU_GEM_DOMAIN_OA);
 	bo->allowed_domains = bo->preferred_domains;
-	if (type != ttm_bo_type_kernel &&
+	if (bp->type != ttm_bo_type_kernel &&
 	    bo->allowed_domains == AMDGPU_GEM_DOMAIN_VRAM)
 		bo->allowed_domains |= AMDGPU_GEM_DOMAIN_GTT;
 
-	bo->flags = flags;
+	bo->flags = bp->flags;
 
 #ifdef CONFIG_X86_32
 	/* XXX: Write-combined CPU mappings of GTT seem broken on 32-bit
@@ -417,11 +415,11 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev, unsigned long size,
 #endif
 
 	bo->tbo.bdev = &adev->mman.bdev;
-	amdgpu_ttm_placement_from_domain(bo, domain);
+	amdgpu_ttm_placement_from_domain(bo, bp->domain);
 
-	r = ttm_bo_init_reserved(&adev->mman.bdev, &bo->tbo, size, type,
+	r = ttm_bo_init_reserved(&adev->mman.bdev, &bo->tbo, size, bp->type,
 				 &bo->placement, page_align, &ctx, acc_size,
-				 NULL, resv, &amdgpu_ttm_bo_destroy);
+				 NULL, bp->resv, &amdgpu_ttm_bo_destroy);
 	if (unlikely(r != 0))
 		return r;
 
@@ -433,10 +431,10 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev, unsigned long size,
 	else
 		amdgpu_cs_report_moved_bytes(adev, ctx.bytes_moved, 0);
 
-	if (type == ttm_bo_type_kernel)
+	if (bp->type == ttm_bo_type_kernel)
 		bo->tbo.priority = 1;
 
-	if (flags & AMDGPU_GEM_CREATE_VRAM_CLEARED &&
+	if (bp->flags & AMDGPU_GEM_CREATE_VRAM_CLEARED &&
 	    bo->tbo.mem.placement & TTM_PL_FLAG_VRAM) {
 		struct dma_fence *fence;
 
@@ -449,20 +447,20 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev, unsigned long size,
 		bo->tbo.moving = dma_fence_get(fence);
 		dma_fence_put(fence);
 	}
-	if (!resv)
+	if (!bp->resv)
 		amdgpu_bo_unreserve(bo);
 	*bo_ptr = bo;
 
 	trace_amdgpu_bo_create(bo);
 
 	/* Treat CPU_ACCESS_REQUIRED only as a hint if given by UMD */
-	if (type == ttm_bo_type_device)
+	if (bp->type == ttm_bo_type_device)
 		bo->flags &= ~AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED;
 
 	return 0;
 
 fail_unreserve:
-	if (!resv)
+	if (!bp->resv)
 		ww_mutex_unlock(&bo->tbo.resv->lock);
 	amdgpu_bo_unref(&bo);
 	return r;
@@ -472,16 +470,21 @@ static int amdgpu_bo_create_shadow(struct amdgpu_device *adev,
 				   unsigned long size, int byte_align,
 				   struct amdgpu_bo *bo)
 {
+	struct amdgpu_bo_param bp = {
+		.size = size,
+		.byte_align = byte_align,
+		.domain = AMDGPU_GEM_DOMAIN_GTT,
+		.flags = AMDGPU_GEM_CREATE_CPU_GTT_USWC |
+			AMDGPU_GEM_CREATE_SHADOW,
+		.type = ttm_bo_type_kernel,
+		.resv = bo->tbo.resv
+	};
 	int r;
 
 	if (bo->shadow)
 		return 0;
 
-	r = amdgpu_bo_do_create(adev, size, byte_align, AMDGPU_GEM_DOMAIN_GTT,
-				AMDGPU_GEM_CREATE_CPU_GTT_USWC |
-				AMDGPU_GEM_CREATE_SHADOW,
-				ttm_bo_type_kernel,
-				bo->tbo.resv, &bo->shadow);
+	r = amdgpu_bo_do_create(adev, &bp, &bo->shadow);
 	if (!r) {
 		bo->shadow->parent = amdgpu_bo_ref(bo);
 		mutex_lock(&adev->shadow_list_lock);
@@ -498,11 +501,17 @@ int amdgpu_bo_create(struct amdgpu_device *adev, unsigned long size,
 		     struct reservation_object *resv,
 		     struct amdgpu_bo **bo_ptr)
 {
-	uint64_t parent_flags = flags & ~AMDGPU_GEM_CREATE_SHADOW;
+	struct amdgpu_bo_param bp = {
+		.size = size,
+		.byte_align = byte_align,
+		.domain = domain,
+		.flags = flags & ~AMDGPU_GEM_CREATE_SHADOW,
+		.type = type,
+		.resv = resv
+	};
 	int r;
 
-	r = amdgpu_bo_do_create(adev, size, byte_align, domain,
-				parent_flags, type, resv, bo_ptr);
+	r = amdgpu_bo_do_create(adev, &bp, bo_ptr);
 	if (r)
 		return r;
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
index 1e9fe85..4bb6f0a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
@@ -33,6 +33,15 @@
 
 #define AMDGPU_BO_INVALID_OFFSET	LONG_MAX
 
+struct amdgpu_bo_param {
+	unsigned long			size;
+	int				byte_align;
+	u32				domain;
+	u64				flags;
+	enum ttm_bo_type		type;
+	struct reservation_object	*resv;
+};
+
 /* bo virtual addresses in a vm */
 struct amdgpu_bo_va_mapping {
 	struct amdgpu_bo_va		*bo_va;
-- 
cgit v1.1


From 3216c6b71d1e6a7dce2fd29c531e8c99c1b88c95 Mon Sep 17 00:00:00 2001
From: Chunming Zhou <david1.zhou@amd.com>
Date: Mon, 16 Apr 2018 18:27:50 +0800
Subject: drm/amdgpu: use amdgpu_bo_param for amdgpu_bo_create v2
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

After that, we can easily add new parameter when need.

v2:
a) rebase.
b) Initialize struct amdgpu_bo_param, future new
member could only be used in some one case, but all member
should have its own initial value.

Signed-off-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com> (v1)
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Cc: christian.koenig@amd.com
Cc: Felix.Kuehling@amd.com
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c       | 12 +++--
 drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c | 11 ++++-
 drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c    | 15 ++++--
 drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c         | 17 ++++---
 drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c          | 11 ++++-
 drivers/gpu/drm/amd/amdgpu/amdgpu_object.c       | 58 ++++++++++++------------
 drivers/gpu/drm/amd/amdgpu/amdgpu_object.h       |  6 +--
 drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c        | 12 +++--
 drivers/gpu/drm/amd/amdgpu/amdgpu_test.c         | 18 +++++---
 drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c          | 15 ++++--
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c           | 26 ++++++++---
 11 files changed, 130 insertions(+), 71 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
index 4d36203..887702c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
@@ -217,13 +217,19 @@ int alloc_gtt_mem(struct kgd_dev *kgd, size_t size,
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)kgd;
 	struct amdgpu_bo *bo = NULL;
+	struct amdgpu_bo_param bp;
 	int r;
 	uint64_t gpu_addr_tmp = 0;
 	void *cpu_ptr_tmp = NULL;
 
-	r = amdgpu_bo_create(adev, size, PAGE_SIZE, AMDGPU_GEM_DOMAIN_GTT,
-			     AMDGPU_GEM_CREATE_CPU_GTT_USWC, ttm_bo_type_kernel,
-			     NULL, &bo);
+	memset(&bp, 0, sizeof(bp));
+	bp.size = size;
+	bp.byte_align = PAGE_SIZE;
+	bp.domain = AMDGPU_GEM_DOMAIN_GTT;
+	bp.flags = AMDGPU_GEM_CREATE_CPU_GTT_USWC;
+	bp.type = ttm_bo_type_kernel;
+	bp.resv = NULL;
+	r = amdgpu_bo_create(adev, &bp, &bo);
 	if (r) {
 		dev_err(adev->dev,
 			"failed to allocate BO for amdkfd (%d)\n", r);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
index 1d6e147..c1b0cdb 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
@@ -1004,6 +1004,7 @@ int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu(
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
 	struct amdgpu_vm *avm = (struct amdgpu_vm *)vm;
 	struct amdgpu_bo *bo;
+	struct amdgpu_bo_param bp;
 	int byte_align;
 	u32 alloc_domain;
 	u64 alloc_flags;
@@ -1069,8 +1070,14 @@ int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu(
 	pr_debug("\tcreate BO VA 0x%llx size 0x%llx domain %s\n",
 			va, size, domain_string(alloc_domain));
 
-	ret = amdgpu_bo_create(adev, size, byte_align,
-				alloc_domain, alloc_flags, ttm_bo_type_device, NULL, &bo);
+	memset(&bp, 0, sizeof(bp));
+	bp.size = size;
+	bp.byte_align = byte_align;
+	bp.domain = alloc_domain;
+	bp.flags = alloc_flags;
+	bp.type = ttm_bo_type_device;
+	bp.resv = NULL;
+	ret = amdgpu_bo_create(adev, &bp, &bo);
 	if (ret) {
 		pr_debug("Failed to create BO on domain %s. ret %d\n",
 				domain_string(alloc_domain), ret);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c
index 02b849b..19cfff3 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c
@@ -75,13 +75,20 @@ static void amdgpu_benchmark_move(struct amdgpu_device *adev, unsigned size,
 {
 	struct amdgpu_bo *dobj = NULL;
 	struct amdgpu_bo *sobj = NULL;
+	struct amdgpu_bo_param bp;
 	uint64_t saddr, daddr;
 	int r, n;
 	int time;
 
+	memset(&bp, 0, sizeof(bp));
+	bp.size = size;
+	bp.byte_align = PAGE_SIZE;
+	bp.domain = sdomain;
+	bp.flags = 0;
+	bp.type = ttm_bo_type_kernel;
+	bp.resv = NULL;
 	n = AMDGPU_BENCHMARK_ITERATIONS;
-	r = amdgpu_bo_create(adev, size, PAGE_SIZE,sdomain, 0,
-			     ttm_bo_type_kernel, NULL, &sobj);
+	r = amdgpu_bo_create(adev, &bp, &sobj);
 	if (r) {
 		goto out_cleanup;
 	}
@@ -93,8 +100,8 @@ static void amdgpu_benchmark_move(struct amdgpu_device *adev, unsigned size,
 	if (r) {
 		goto out_cleanup;
 	}
-	r = amdgpu_bo_create(adev, size, PAGE_SIZE, ddomain, 0,
-			     ttm_bo_type_kernel, NULL, &dobj);
+	bp.domain = ddomain;
+	r = amdgpu_bo_create(adev, &bp, &dobj);
 	if (r) {
 		goto out_cleanup;
 	}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c
index cf0f186..17d6b9f 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c
@@ -113,12 +113,17 @@ int amdgpu_gart_table_vram_alloc(struct amdgpu_device *adev)
 	int r;
 
 	if (adev->gart.robj == NULL) {
-		r = amdgpu_bo_create(adev, adev->gart.table_size, PAGE_SIZE,
-				     AMDGPU_GEM_DOMAIN_VRAM,
-				     AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED |
-				     AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS,
-				     ttm_bo_type_kernel, NULL,
-				     &adev->gart.robj);
+		struct amdgpu_bo_param bp;
+
+		memset(&bp, 0, sizeof(bp));
+		bp.size = adev->gart.table_size;
+		bp.byte_align = PAGE_SIZE;
+		bp.domain = AMDGPU_GEM_DOMAIN_VRAM;
+		bp.flags = AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED |
+			AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS;
+		bp.type = ttm_bo_type_kernel;
+		bp.resv = NULL;
+		r = amdgpu_bo_create(adev, &bp, &adev->gart.robj);
 		if (r) {
 			return r;
 		}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
index 46b9ea4..1200c5b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
@@ -48,17 +48,24 @@ int amdgpu_gem_object_create(struct amdgpu_device *adev, unsigned long size,
 			     struct drm_gem_object **obj)
 {
 	struct amdgpu_bo *bo;
+	struct amdgpu_bo_param bp;
 	int r;
 
+	memset(&bp, 0, sizeof(bp));
 	*obj = NULL;
 	/* At least align on page size */
 	if (alignment < PAGE_SIZE) {
 		alignment = PAGE_SIZE;
 	}
 
+	bp.size = size;
+	bp.byte_align = alignment;
+	bp.type = type;
+	bp.resv = resv;
 retry:
-	r = amdgpu_bo_create(adev, size, alignment, initial_domain,
-			     flags, type, resv, &bo);
+	bp.flags = flags;
+	bp.domain = initial_domain;
+	r = amdgpu_bo_create(adev, &bp, &bo);
 	if (r) {
 		if (r != -ERESTARTSYS) {
 			if (flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED) {
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
index b33a7fd..cac65e3 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
@@ -191,14 +191,21 @@ int amdgpu_bo_create_reserved(struct amdgpu_device *adev,
 			      u32 domain, struct amdgpu_bo **bo_ptr,
 			      u64 *gpu_addr, void **cpu_addr)
 {
+	struct amdgpu_bo_param bp;
 	bool free = false;
 	int r;
 
+	memset(&bp, 0, sizeof(bp));
+	bp.size = size;
+	bp.byte_align = align;
+	bp.domain = domain;
+	bp.flags = AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED |
+		AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS;
+	bp.type = ttm_bo_type_kernel;
+	bp.resv = NULL;
+
 	if (!*bo_ptr) {
-		r = amdgpu_bo_create(adev, size, align, domain,
-				     AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED |
-				     AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS,
-				     ttm_bo_type_kernel, NULL, bo_ptr);
+		r = amdgpu_bo_create(adev, &bp, bo_ptr);
 		if (r) {
 			dev_err(adev->dev, "(%d) failed to allocate kernel bo\n",
 				r);
@@ -470,20 +477,21 @@ static int amdgpu_bo_create_shadow(struct amdgpu_device *adev,
 				   unsigned long size, int byte_align,
 				   struct amdgpu_bo *bo)
 {
-	struct amdgpu_bo_param bp = {
-		.size = size,
-		.byte_align = byte_align,
-		.domain = AMDGPU_GEM_DOMAIN_GTT,
-		.flags = AMDGPU_GEM_CREATE_CPU_GTT_USWC |
-			AMDGPU_GEM_CREATE_SHADOW,
-		.type = ttm_bo_type_kernel,
-		.resv = bo->tbo.resv
-	};
+	struct amdgpu_bo_param bp;
 	int r;
 
 	if (bo->shadow)
 		return 0;
 
+	memset(&bp, 0, sizeof(bp));
+	bp.size = size;
+	bp.byte_align = byte_align;
+	bp.domain = AMDGPU_GEM_DOMAIN_GTT;
+	bp.flags = AMDGPU_GEM_CREATE_CPU_GTT_USWC |
+		AMDGPU_GEM_CREATE_SHADOW;
+	bp.type = ttm_bo_type_kernel;
+	bp.resv = bo->tbo.resv;
+
 	r = amdgpu_bo_do_create(adev, &bp, &bo->shadow);
 	if (!r) {
 		bo->shadow->parent = amdgpu_bo_ref(bo);
@@ -495,34 +503,26 @@ static int amdgpu_bo_create_shadow(struct amdgpu_device *adev,
 	return r;
 }
 
-int amdgpu_bo_create(struct amdgpu_device *adev, unsigned long size,
-		     int byte_align, u32 domain,
-		     u64 flags, enum ttm_bo_type type,
-		     struct reservation_object *resv,
+int amdgpu_bo_create(struct amdgpu_device *adev,
+		     struct amdgpu_bo_param *bp,
 		     struct amdgpu_bo **bo_ptr)
 {
-	struct amdgpu_bo_param bp = {
-		.size = size,
-		.byte_align = byte_align,
-		.domain = domain,
-		.flags = flags & ~AMDGPU_GEM_CREATE_SHADOW,
-		.type = type,
-		.resv = resv
-	};
+	u64 flags = bp->flags;
 	int r;
 
-	r = amdgpu_bo_do_create(adev, &bp, bo_ptr);
+	bp->flags = bp->flags & ~AMDGPU_GEM_CREATE_SHADOW;
+	r = amdgpu_bo_do_create(adev, bp, bo_ptr);
 	if (r)
 		return r;
 
 	if ((flags & AMDGPU_GEM_CREATE_SHADOW) && amdgpu_need_backup(adev)) {
-		if (!resv)
+		if (!bp->resv)
 			WARN_ON(reservation_object_lock((*bo_ptr)->tbo.resv,
 							NULL));
 
-		r = amdgpu_bo_create_shadow(adev, size, byte_align, (*bo_ptr));
+		r = amdgpu_bo_create_shadow(adev, bp->size, bp->byte_align, (*bo_ptr));
 
-		if (!resv)
+		if (!bp->resv)
 			reservation_object_unlock((*bo_ptr)->tbo.resv);
 
 		if (r)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
index 4bb6f0a..e9a21d9 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
@@ -233,10 +233,8 @@ static inline bool amdgpu_bo_explicit_sync(struct amdgpu_bo *bo)
 	return bo->flags & AMDGPU_GEM_CREATE_EXPLICIT_SYNC;
 }
 
-int amdgpu_bo_create(struct amdgpu_device *adev, unsigned long size,
-		     int byte_align, u32 domain,
-		     u64 flags, enum ttm_bo_type type,
-		     struct reservation_object *resv,
+int amdgpu_bo_create(struct amdgpu_device *adev,
+		     struct amdgpu_bo_param *bp,
 		     struct amdgpu_bo **bo_ptr);
 int amdgpu_bo_create_reserved(struct amdgpu_device *adev,
 			      unsigned long size, int align,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c
index 4b584cb7..713417b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c
@@ -102,12 +102,18 @@ amdgpu_gem_prime_import_sg_table(struct drm_device *dev,
 	struct reservation_object *resv = attach->dmabuf->resv;
 	struct amdgpu_device *adev = dev->dev_private;
 	struct amdgpu_bo *bo;
+	struct amdgpu_bo_param bp;
 	int ret;
 
+	memset(&bp, 0, sizeof(bp));
+	bp.size = attach->dmabuf->size;
+	bp.byte_align = PAGE_SIZE;
+	bp.domain = AMDGPU_GEM_DOMAIN_CPU;
+	bp.flags = 0;
+	bp.type = ttm_bo_type_sg;
+	bp.resv = resv;
 	ww_mutex_lock(&resv->lock, NULL);
-	ret = amdgpu_bo_create(adev, attach->dmabuf->size, PAGE_SIZE,
-			       AMDGPU_GEM_DOMAIN_CPU, 0, ttm_bo_type_sg,
-			       resv, &bo);
+	ret = amdgpu_bo_create(adev, &bp, &bo);
 	if (ret)
 		goto error;
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_test.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_test.c
index 2dbe875..d167e8a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_test.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_test.c
@@ -33,6 +33,7 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev)
 	struct amdgpu_ring *ring = adev->mman.buffer_funcs_ring;
 	struct amdgpu_bo *vram_obj = NULL;
 	struct amdgpu_bo **gtt_obj = NULL;
+	struct amdgpu_bo_param bp;
 	uint64_t gart_addr, vram_addr;
 	unsigned n, size;
 	int i, r;
@@ -58,9 +59,15 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev)
 		r = 1;
 		goto out_cleanup;
 	}
-
-	r = amdgpu_bo_create(adev, size, PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM, 0,
-			     ttm_bo_type_kernel, NULL, &vram_obj);
+	memset(&bp, 0, sizeof(bp));
+	bp.size = size;
+	bp.byte_align = PAGE_SIZE;
+	bp.domain = AMDGPU_GEM_DOMAIN_VRAM;
+	bp.flags = 0;
+	bp.type = ttm_bo_type_kernel;
+	bp.resv = NULL;
+
+	r = amdgpu_bo_create(adev, &bp, &vram_obj);
 	if (r) {
 		DRM_ERROR("Failed to create VRAM object\n");
 		goto out_cleanup;
@@ -79,9 +86,8 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev)
 		void **vram_start, **vram_end;
 		struct dma_fence *fence = NULL;
 
-		r = amdgpu_bo_create(adev, size, PAGE_SIZE,
-				     AMDGPU_GEM_DOMAIN_GTT, 0,
-				     ttm_bo_type_kernel, NULL, gtt_obj + i);
+		bp.domain = AMDGPU_GEM_DOMAIN_GTT;
+		r = amdgpu_bo_create(adev, &bp, gtt_obj + i);
 		if (r) {
 			DRM_ERROR("Failed to create GTT object %d\n", i);
 			goto out_lclean;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index 29efaac..dfd22db 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -1316,6 +1316,7 @@ static void amdgpu_ttm_fw_reserve_vram_fini(struct amdgpu_device *adev)
 static int amdgpu_ttm_fw_reserve_vram_init(struct amdgpu_device *adev)
 {
 	struct ttm_operation_ctx ctx = { false, false };
+	struct amdgpu_bo_param bp;
 	int r = 0;
 	int i;
 	u64 vram_size = adev->gmc.visible_vram_size;
@@ -1323,17 +1324,21 @@ static int amdgpu_ttm_fw_reserve_vram_init(struct amdgpu_device *adev)
 	u64 size = adev->fw_vram_usage.size;
 	struct amdgpu_bo *bo;
 
+	memset(&bp, 0, sizeof(bp));
+	bp.size = adev->fw_vram_usage.size;
+	bp.byte_align = PAGE_SIZE;
+	bp.domain = AMDGPU_GEM_DOMAIN_VRAM;
+	bp.flags = AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED |
+		AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS;
+	bp.type = ttm_bo_type_kernel;
+	bp.resv = NULL;
 	adev->fw_vram_usage.va = NULL;
 	adev->fw_vram_usage.reserved_bo = NULL;
 
 	if (adev->fw_vram_usage.size > 0 &&
 		adev->fw_vram_usage.size <= vram_size) {
 
-		r = amdgpu_bo_create(adev, adev->fw_vram_usage.size, PAGE_SIZE,
-				     AMDGPU_GEM_DOMAIN_VRAM,
-				     AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED |
-				     AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS,
-				     ttm_bo_type_kernel, NULL,
+		r = amdgpu_bo_create(adev, &bp,
 				     &adev->fw_vram_usage.reserved_bo);
 		if (r)
 			goto error_create;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index f0fbc33..9ec7c10 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -412,11 +412,16 @@ static int amdgpu_vm_alloc_levels(struct amdgpu_device *adev,
 		struct amdgpu_bo *pt;
 
 		if (!entry->base.bo) {
-			r = amdgpu_bo_create(adev,
-					     amdgpu_vm_bo_size(adev, level),
-					     AMDGPU_GPU_PAGE_SIZE,
-					     AMDGPU_GEM_DOMAIN_VRAM, flags,
-					     ttm_bo_type_kernel, resv, &pt);
+			struct amdgpu_bo_param bp;
+
+			memset(&bp, 0, sizeof(bp));
+			bp.size = amdgpu_vm_bo_size(adev, level);
+			bp.byte_align = AMDGPU_GPU_PAGE_SIZE;
+			bp.domain = AMDGPU_GEM_DOMAIN_VRAM;
+			bp.flags = flags;
+			bp.type = ttm_bo_type_kernel;
+			bp.resv = resv;
+			r = amdgpu_bo_create(adev, &bp, &pt);
 			if (r)
 				return r;
 
@@ -2368,6 +2373,7 @@ void amdgpu_vm_adjust_size(struct amdgpu_device *adev, uint32_t vm_size,
 int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 		   int vm_context, unsigned int pasid)
 {
+	struct amdgpu_bo_param bp;
 	const unsigned align = min(AMDGPU_VM_PTB_ALIGN_SIZE,
 		AMDGPU_VM_PTE_COUNT(adev) * 8);
 	unsigned ring_instance;
@@ -2422,8 +2428,14 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 		flags |= AMDGPU_GEM_CREATE_SHADOW;
 
 	size = amdgpu_vm_bo_size(adev, adev->vm_manager.root_level);
-	r = amdgpu_bo_create(adev, size, align, AMDGPU_GEM_DOMAIN_VRAM, flags,
-			     ttm_bo_type_kernel, NULL, &vm->root.base.bo);
+	memset(&bp, 0, sizeof(bp));
+	bp.size = size;
+	bp.byte_align = align;
+	bp.domain = AMDGPU_GEM_DOMAIN_VRAM;
+	bp.flags = flags;
+	bp.type = ttm_bo_type_kernel;
+	bp.resv = NULL;
+	r = amdgpu_bo_create(adev, &bp, &vm->root.base.bo);
 	if (r)
 		goto error_free_sched_entity;
 
-- 
cgit v1.1


From 7951e376704773134cefcf0751e9042368226f15 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Fri, 13 Apr 2018 16:13:41 +0800
Subject: drm/amdgpu: Reserved vram for smu to save debug info.

v2: check reserved vram size before allocate.

Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu.h           |  1 +
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c    | 44 +++++++++++++++++++++++++++
 drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h       |  2 ++
 drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c       |  6 ++++
 drivers/gpu/drm/amd/powerplay/amd_powerplay.c | 35 +++++++++++++++++++++
 5 files changed, 88 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index ea1b285..d64ef30 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -129,6 +129,7 @@ extern int amdgpu_lbpw;
 extern int amdgpu_compute_multipipe;
 extern int amdgpu_gpu_recovery;
 extern int amdgpu_emu_mode;
+extern uint amdgpu_smu_memory_pool_size;
 
 #ifdef CONFIG_DRM_AMDGPU_SI
 extern int amdgpu_si_support;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index d7f2bbd..5958e81 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -690,6 +690,8 @@ void amdgpu_device_gart_location(struct amdgpu_device *adev,
 {
 	u64 size_af, size_bf;
 
+	mc->gart_size += adev->pm.smu_prv_buffer_size;
+
 	size_af = adev->gmc.mc_mask - mc->vram_end;
 	size_bf = mc->vram_start;
 	if (size_bf > size_af) {
@@ -907,6 +909,46 @@ static void amdgpu_device_check_vm_size(struct amdgpu_device *adev)
 	}
 }
 
+static void amdgpu_device_check_smu_prv_buffer_size(struct amdgpu_device *adev)
+{
+	struct sysinfo si;
+	bool is_os_64 = (sizeof(void *) == 8) ? true : false;
+	uint64_t total_memory;
+	uint64_t dram_size_seven_GB = 0x1B8000000;
+	uint64_t dram_size_three_GB = 0xB8000000;
+
+	if (amdgpu_smu_memory_pool_size == 0)
+		return;
+
+	if (!is_os_64) {
+		DRM_WARN("Not 64-bit OS, feature not supported\n");
+		goto def_value;
+	}
+	si_meminfo(&si);
+	total_memory = (uint64_t)si.totalram * si.mem_unit;
+
+	if ((amdgpu_smu_memory_pool_size == 1) ||
+		(amdgpu_smu_memory_pool_size == 2)) {
+		if (total_memory < dram_size_three_GB)
+			goto def_value1;
+	} else if ((amdgpu_smu_memory_pool_size == 4) ||
+		(amdgpu_smu_memory_pool_size == 8)) {
+		if (total_memory < dram_size_seven_GB)
+			goto def_value1;
+	} else {
+		DRM_WARN("Smu memory pool size not supported\n");
+		goto def_value;
+	}
+	adev->pm.smu_prv_buffer_size = amdgpu_smu_memory_pool_size << 28;
+
+	return;
+
+def_value1:
+	DRM_WARN("No enough system memory\n");
+def_value:
+	adev->pm.smu_prv_buffer_size = 0;
+}
+
 /**
  * amdgpu_device_check_arguments - validate module params
  *
@@ -948,6 +990,8 @@ static void amdgpu_device_check_arguments(struct amdgpu_device *adev)
 		amdgpu_vm_fragment_size = -1;
 	}
 
+	amdgpu_device_check_smu_prv_buffer_size(adev);
+
 	amdgpu_device_check_vm_size(adev);
 
 	amdgpu_device_check_block_size(adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
index b8c5177..19d8bf5 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
@@ -445,6 +445,8 @@ struct amdgpu_pm {
 	uint32_t                pcie_gen_mask;
 	uint32_t                pcie_mlw_mask;
 	struct amd_pp_display_configuration pm_display_cfg;/* set by dc */
+	uint32_t                smu_prv_buffer_size;
+	struct amdgpu_bo        *smu_prv_buffer;
 };
 
 #define R600_SSTU_DFLT                               0
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
index 0b19482..5c0567a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
@@ -132,6 +132,7 @@ int amdgpu_lbpw = -1;
 int amdgpu_compute_multipipe = -1;
 int amdgpu_gpu_recovery = -1; /* auto */
 int amdgpu_emu_mode = 0;
+uint amdgpu_smu_memory_pool_size = 0;
 
 MODULE_PARM_DESC(vramlimit, "Restrict VRAM for testing, in megabytes");
 module_param_named(vramlimit, amdgpu_vram_limit, int, 0600);
@@ -316,6 +317,11 @@ MODULE_PARM_DESC(cik_support, "CIK support (1 = enabled (default), 0 = disabled)
 module_param_named(cik_support, amdgpu_cik_support, int, 0444);
 #endif
 
+MODULE_PARM_DESC(smu_memory_pool_size,
+	"reserve gtt for smu debug usage, 0 = disable,"
+		"0x1 = 256Mbyte, 0x2 = 512Mbyte, 0x4 = 1 Gbyte, 0x8 = 2GByte");
+module_param_named(smu_memory_pool_size, amdgpu_smu_memory_pool_size, uint, 0444);
+
 static const struct pci_device_id pciidlist[] = {
 #ifdef  CONFIG_DRM_AMDGPU_SI
 	{0x1002, 0x6780, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI},
diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
index 66c49b8..6c81914 100644
--- a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
+++ b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
@@ -145,6 +145,37 @@ static int pp_hw_fini(void *handle)
 	return 0;
 }
 
+static void pp_reserve_vram_for_smu(struct amdgpu_device *adev)
+{
+	int r = -EINVAL;
+	void *cpu_ptr = NULL;
+	uint64_t gpu_addr;
+	struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle;
+
+	if (amdgpu_bo_create_kernel(adev, adev->pm.smu_prv_buffer_size,
+						PAGE_SIZE, AMDGPU_GEM_DOMAIN_GTT,
+						&adev->pm.smu_prv_buffer,
+						&gpu_addr,
+						&cpu_ptr)) {
+		DRM_ERROR("amdgpu: failed to create smu prv buffer\n");
+		return;
+	}
+
+	if (hwmgr->hwmgr_func->notify_cac_buffer_info)
+		r = hwmgr->hwmgr_func->notify_cac_buffer_info(hwmgr,
+					lower_32_bits((unsigned long)cpu_ptr),
+					upper_32_bits((unsigned long)cpu_ptr),
+					lower_32_bits(gpu_addr),
+					upper_32_bits(gpu_addr),
+					adev->pm.smu_prv_buffer_size);
+
+	if (r) {
+		amdgpu_bo_free_kernel(&adev->pm.smu_prv_buffer, NULL, NULL);
+		adev->pm.smu_prv_buffer = NULL;
+		DRM_ERROR("amdgpu: failed to notify SMU buffer address\n");
+	}
+}
+
 static int pp_late_init(void *handle)
 {
 	struct amdgpu_device *adev = handle;
@@ -156,6 +187,8 @@ static int pp_late_init(void *handle)
 					AMD_PP_TASK_COMPLETE_INIT, NULL);
 		mutex_unlock(&hwmgr->smu_lock);
 	}
+	if (adev->pm.smu_prv_buffer_size != 0)
+		pp_reserve_vram_for_smu(adev);
 	return 0;
 }
 
@@ -163,6 +196,8 @@ static void pp_late_fini(void *handle)
 {
 	struct amdgpu_device *adev = handle;
 
+	if (adev->pm.smu_prv_buffer)
+		amdgpu_bo_free_kernel(&adev->pm.smu_prv_buffer, NULL, NULL);
 	amd_powerplay_destroy(adev);
 }
 
-- 
cgit v1.1


From a0d454a67737162b0e4b1cc91612d7b25d5681b0 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Fri, 13 Apr 2018 16:16:49 +0800
Subject: drm/amd/pp: Remove dead interface

Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h        |  6 ------
 drivers/gpu/drm/amd/include/kgd_pp_interface.h |  5 -----
 drivers/gpu/drm/amd/powerplay/amd_powerplay.c  | 30 --------------------------
 3 files changed, 41 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
index 19d8bf5..354c6dc 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
@@ -349,12 +349,6 @@ enum amdgpu_pcie_gen {
 		((adev)->powerplay.pp_funcs->set_clockgating_by_smu(\
 			(adev)->powerplay.pp_handle, msg_id))
 
-#define amdgpu_dpm_notify_smu_memory_info(adev, virtual_addr_low, \
-			virtual_addr_hi, mc_addr_low, mc_addr_hi, size) \
-		((adev)->powerplay.pp_funcs->notify_smu_memory_info)( \
-			(adev)->powerplay.pp_handle, virtual_addr_low, \
-			virtual_addr_hi, mc_addr_low, mc_addr_hi, size)
-
 #define amdgpu_dpm_get_power_profile_mode(adev, buf) \
 		((adev)->powerplay.pp_funcs->get_power_profile_mode(\
 			(adev)->powerplay.pp_handle, buf))
diff --git a/drivers/gpu/drm/amd/include/kgd_pp_interface.h b/drivers/gpu/drm/amd/include/kgd_pp_interface.h
index 1bec907..01969b1 100644
--- a/drivers/gpu/drm/amd/include/kgd_pp_interface.h
+++ b/drivers/gpu/drm/amd/include/kgd_pp_interface.h
@@ -239,11 +239,6 @@ struct amd_pm_funcs {
 	int (*load_firmware)(void *handle);
 	int (*wait_for_fw_loading_complete)(void *handle);
 	int (*set_clockgating_by_smu)(void *handle, uint32_t msg_id);
-	int (*notify_smu_memory_info)(void *handle, uint32_t virtual_addr_low,
-					uint32_t virtual_addr_hi,
-					uint32_t mc_addr_low,
-					uint32_t mc_addr_hi,
-					uint32_t size);
 	int (*set_power_limit)(void *handle, uint32_t n);
 	int (*get_power_limit)(void *handle, uint32_t *limit, bool default_limit);
 /* export to DC */
diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
index 6c81914..bd0d387 100644
--- a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
+++ b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
@@ -934,35 +934,6 @@ static int pp_dpm_switch_power_profile(void *handle,
 	return 0;
 }
 
-static int pp_dpm_notify_smu_memory_info(void *handle,
-					uint32_t virtual_addr_low,
-					uint32_t virtual_addr_hi,
-					uint32_t mc_addr_low,
-					uint32_t mc_addr_hi,
-					uint32_t size)
-{
-	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
-
-	if (!hwmgr || !hwmgr->pm_en)
-		return -EINVAL;
-
-	if (hwmgr->hwmgr_func->notify_cac_buffer_info == NULL) {
-		pr_info("%s was not implemented.\n", __func__);
-		return -EINVAL;
-	}
-
-	mutex_lock(&hwmgr->smu_lock);
-
-	ret = hwmgr->hwmgr_func->notify_cac_buffer_info(hwmgr, virtual_addr_low,
-					virtual_addr_hi, mc_addr_low, mc_addr_hi,
-					size);
-
-	mutex_unlock(&hwmgr->smu_lock);
-
-	return ret;
-}
-
 static int pp_set_power_limit(void *handle, uint32_t limit)
 {
 	struct pp_hwmgr *hwmgr = handle;
@@ -1229,7 +1200,6 @@ static const struct amd_pm_funcs pp_dpm_funcs = {
 	.get_vce_clock_state = pp_dpm_get_vce_clock_state,
 	.switch_power_profile = pp_dpm_switch_power_profile,
 	.set_clockgating_by_smu = pp_set_clockgating_by_smu,
-	.notify_smu_memory_info = pp_dpm_notify_smu_memory_info,
 	.get_power_profile_mode = pp_get_power_profile_mode,
 	.set_power_profile_mode = pp_set_power_profile_mode,
 	.odn_edit_dpm_table = pp_odn_edit_dpm_table,
-- 
cgit v1.1


From 8d80fada066bec682f1b7e9015b8412e3460c1b3 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Tue, 17 Apr 2018 17:26:26 +0800
Subject: drm/amd/pp: Fix bug voltage can't be OD separately on VI

Make sure to update the MCLK and SCLK flags when setting the VDDC
flags due to dependencies.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c | 16 ++++++++++------
 1 file changed, 10 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index 68aae09..720ac47 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -4679,23 +4679,27 @@ static void smu7_check_dpm_table_updated(struct pp_hwmgr *hwmgr)
 
 	for (i=0; i < dep_table->count; i++) {
 		if (dep_table->entries[i].vddc != odn_dep_table->entries[i].vddc) {
-			data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_VDDC;
-			break;
+			data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_VDDC | DPMTABLE_OD_UPDATE_MCLK;
+			return;
 		}
 	}
-	if (i == dep_table->count)
+	if (i == dep_table->count && data->need_update_smu7_dpm_table & DPMTABLE_OD_UPDATE_VDDC) {
 		data->need_update_smu7_dpm_table &= ~DPMTABLE_OD_UPDATE_VDDC;
+		data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_MCLK;
+	}
 
 	dep_table = table_info->vdd_dep_on_sclk;
 	odn_dep_table = (struct phm_ppt_v1_clock_voltage_dependency_table *)&(odn_table->vdd_dependency_on_sclk);
 	for (i=0; i < dep_table->count; i++) {
 		if (dep_table->entries[i].vddc != odn_dep_table->entries[i].vddc) {
-			data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_VDDC;
-			break;
+			data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_VDDC | DPMTABLE_OD_UPDATE_SCLK;
+			return;
 		}
 	}
-	if (i == dep_table->count)
+	if (i == dep_table->count && data->need_update_smu7_dpm_table & DPMTABLE_OD_UPDATE_VDDC) {
 		data->need_update_smu7_dpm_table &= ~DPMTABLE_OD_UPDATE_VDDC;
+		data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_SCLK;
+	}
 }
 
 static int smu7_odn_edit_dpm_table(struct pp_hwmgr *hwmgr,
-- 
cgit v1.1


From 32d8c6620d49779600714f197611856ed503a7a5 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Tue, 17 Apr 2018 08:55:44 -0500
Subject: drm/amdgpu: print the vbios version in the debugfs firmware info

Useful for info gathering about what firmwares are in use in
the driver.

Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
index 4e15b6f..d602f8b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
@@ -31,6 +31,7 @@
 #include "amdgpu_sched.h"
 #include "amdgpu_uvd.h"
 #include "amdgpu_vce.h"
+#include "atom.h"
 
 #include <linux/vga_switcheroo.h>
 #include <linux/slab.h>
@@ -1089,6 +1090,7 @@ static int amdgpu_debugfs_firmware_info(struct seq_file *m, void *data)
 	struct amdgpu_device *adev = dev->dev_private;
 	struct drm_amdgpu_info_firmware fw_info;
 	struct drm_amdgpu_query_fw query_fw;
+	struct atom_context *ctx = adev->mode_info.atom_context;
 	int ret, i;
 
 	/* VCE */
@@ -1211,6 +1213,9 @@ static int amdgpu_debugfs_firmware_info(struct seq_file *m, void *data)
 	seq_printf(m, "VCN feature version: %u, firmware version: 0x%08x\n",
 		   fw_info.feature, fw_info.ver);
 
+
+	seq_printf(m, "VBIOS version: %s\n", ctx->vbios_version);
+
 	return 0;
 }
 
-- 
cgit v1.1


From a4b3996aeebbaafd2682f4db06bb5659e1653da7 Mon Sep 17 00:00:00 2001
From: Pixel Ding <Pixel.Ding@amd.com>
Date: Wed, 18 Apr 2018 04:33:26 -0400
Subject: drm/scheduler: always put last_sched fence in entity_fini
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Fix the potential memleak since scheduler main thread always
hold one last_sched fence.

Signed-off-by: Pixel Ding <Pixel.Ding@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/scheduler/gpu_scheduler.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/scheduler/gpu_scheduler.c b/drivers/gpu/drm/scheduler/gpu_scheduler.c
index 44d2198..4968867 100644
--- a/drivers/gpu/drm/scheduler/gpu_scheduler.c
+++ b/drivers/gpu/drm/scheduler/gpu_scheduler.c
@@ -276,10 +276,10 @@ void drm_sched_entity_cleanup(struct drm_gpu_scheduler *sched,
 			else if (r)
 				DRM_ERROR("fence add callback failed (%d)\n", r);
 		}
-
-		dma_fence_put(entity->last_scheduled);
-		entity->last_scheduled = NULL;
 	}
+
+	dma_fence_put(entity->last_scheduled);
+	entity->last_scheduled = NULL;
 }
 EXPORT_SYMBOL(drm_sched_entity_cleanup);
 
-- 
cgit v1.1


From b5b4ea4d98b42f94442e5d46d5942f392ed8af56 Mon Sep 17 00:00:00 2001
From: Pixel Ding <Pixel.Ding@amd.com>
Date: Wed, 18 Apr 2018 04:37:40 -0400
Subject: drm/scheduler: move last_sched fence updating prior to job popping
 (v2)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Make sure main thread won't update last_sched fence when entity
is cleanup.

Fix a racing issue which is caused by putting last_sched fence
twice. Running vulkaninfo in tight loop can produce this issue
as seeing wild fence pointer.

v2: squash in build fix (Christian)

Signed-off-by: Pixel Ding <Pixel.Ding@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Monk Liu <Monk.Liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/scheduler/gpu_scheduler.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/scheduler/gpu_scheduler.c b/drivers/gpu/drm/scheduler/gpu_scheduler.c
index 4968867..088ff2b 100644
--- a/drivers/gpu/drm/scheduler/gpu_scheduler.c
+++ b/drivers/gpu/drm/scheduler/gpu_scheduler.c
@@ -402,6 +402,9 @@ drm_sched_entity_pop_job(struct drm_sched_entity *entity)
 	if (entity->guilty && atomic_read(entity->guilty))
 		dma_fence_set_error(&sched_job->s_fence->finished, -ECANCELED);
 
+	dma_fence_put(entity->last_scheduled);
+	entity->last_scheduled = dma_fence_get(&sched_job->s_fence->finished);
+
 	spsc_queue_pop(&entity->job_queue);
 	return sched_job;
 }
@@ -715,9 +718,6 @@ static int drm_sched_main(void *param)
 		fence = sched->ops->run_job(sched_job);
 		drm_sched_fence_scheduled(s_fence);
 
-		dma_fence_put(entity->last_scheduled);
-		entity->last_scheduled = dma_fence_get(&s_fence->finished);
-
 		if (fence) {
 			s_fence->parent = dma_fence_get(fence);
 			r = dma_fence_add_callback(fence, &s_fence->cb,
-- 
cgit v1.1


From 58cd8fbc64b03d0e9961d627526bd07edbea00b9 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Christian=20K=C3=B6nig?= <christian.koenig@amd.com>
Date: Tue, 17 Apr 2018 14:47:42 +0200
Subject: drm/amdgpu: limit reg_write_reg_wait workaround to SRIOV v2
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Turned out that this locks up some bare metal Vega10.

v2: fix stupid typo

Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index 583f6f6..6a19e03 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -4144,7 +4144,12 @@ static void gfx_v9_0_ring_emit_reg_write_reg_wait(struct amdgpu_ring *ring,
 {
 	int usepfp = (ring->funcs->type == AMDGPU_RING_TYPE_GFX);
 
-	gfx_v9_0_wait_reg_mem(ring, usepfp, 0, 1, reg0, reg1, ref, mask, 0x20);
+	if (amdgpu_sriov_vf(ring->adev))
+		gfx_v9_0_wait_reg_mem(ring, usepfp, 0, 1, reg0, reg1,
+				      ref, mask, 0x20);
+	else
+		amdgpu_ring_emit_reg_write_reg_wait_helper(ring, reg0, reg1,
+							   ref, mask);
 }
 
 static void gfx_v9_0_set_gfx_eop_interrupt_state(struct amdgpu_device *adev,
-- 
cgit v1.1


From aa2b2e2822831d78a283edb12cf8b7da21bdd0ed Mon Sep 17 00:00:00 2001
From: Chunming Zhou <david1.zhou@amd.com>
Date: Tue, 17 Apr 2018 11:52:53 +0800
Subject: drm/amdgpu: set preferred_domain independent of fallback handling
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

When GEM needs to fallback to GTT for VRAM BOs we still want the
preferred domain to be untouched so that the BO has a cance to move back
to VRAM in the future.

Signed-off-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c    |  1 +
 drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 15 +++++++++------
 drivers/gpu/drm/amd/amdgpu/amdgpu_object.h |  1 +
 3 files changed, 11 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
index 1200c5b..ff606ce 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
@@ -62,6 +62,7 @@ int amdgpu_gem_object_create(struct amdgpu_device *adev, unsigned long size,
 	bp.byte_align = alignment;
 	bp.type = type;
 	bp.resv = resv;
+	bp.preferred_domain = initial_domain;
 retry:
 	bp.flags = flags;
 	bp.domain = initial_domain;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
index cac65e3..9258f06 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
@@ -360,6 +360,7 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev,
 	};
 	struct amdgpu_bo *bo;
 	unsigned long page_align, size = bp->size;
+	u32 preferred_domains;
 	size_t acc_size;
 	int r;
 
@@ -380,12 +381,14 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev,
 	drm_gem_private_object_init(adev->ddev, &bo->gem_base, size);
 	INIT_LIST_HEAD(&bo->shadow_list);
 	INIT_LIST_HEAD(&bo->va);
-	bo->preferred_domains = bp->domain & (AMDGPU_GEM_DOMAIN_VRAM |
-					      AMDGPU_GEM_DOMAIN_GTT |
-					      AMDGPU_GEM_DOMAIN_CPU |
-					      AMDGPU_GEM_DOMAIN_GDS |
-					      AMDGPU_GEM_DOMAIN_GWS |
-					      AMDGPU_GEM_DOMAIN_OA);
+	preferred_domains = bp->preferred_domain ? bp->preferred_domain :
+		bp->domain;
+	bo->preferred_domains = preferred_domains & (AMDGPU_GEM_DOMAIN_VRAM |
+						     AMDGPU_GEM_DOMAIN_GTT |
+						     AMDGPU_GEM_DOMAIN_CPU |
+						     AMDGPU_GEM_DOMAIN_GDS |
+						     AMDGPU_GEM_DOMAIN_GWS |
+						     AMDGPU_GEM_DOMAIN_OA);
 	bo->allowed_domains = bo->preferred_domains;
 	if (bp->type != ttm_bo_type_kernel &&
 	    bo->allowed_domains == AMDGPU_GEM_DOMAIN_VRAM)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
index e9a21d9..540e03f 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
@@ -37,6 +37,7 @@ struct amdgpu_bo_param {
 	unsigned long			size;
 	int				byte_align;
 	u32				domain;
+	u32				preferred_domain;
 	u64				flags;
 	enum ttm_bo_type		type;
 	struct reservation_object	*resv;
-- 
cgit v1.1


From 3f188453faf7ba5b59e8064df4afffbc946e25ec Mon Sep 17 00:00:00 2001
From: Chunming Zhou <david1.zhou@amd.com>
Date: Tue, 17 Apr 2018 18:34:40 +0800
Subject: drm/amdgpu: handle domain mask checking v2
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

if domain is illegal, we should return error.
v2:
  remove duplicated domain checking.

Signed-off-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c    | 7 +------
 drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 9 +--------
 include/uapi/drm/amdgpu_drm.h              | 6 ++++++
 3 files changed, 8 insertions(+), 14 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
index ff606ce..c62c3dd 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
@@ -229,12 +229,7 @@ int amdgpu_gem_create_ioctl(struct drm_device *dev, void *data,
 		return -EINVAL;
 
 	/* reject invalid gem domains */
-	if (args->in.domains & ~(AMDGPU_GEM_DOMAIN_CPU |
-				 AMDGPU_GEM_DOMAIN_GTT |
-				 AMDGPU_GEM_DOMAIN_VRAM |
-				 AMDGPU_GEM_DOMAIN_GDS |
-				 AMDGPU_GEM_DOMAIN_GWS |
-				 AMDGPU_GEM_DOMAIN_OA))
+	if (args->in.domains & ~AMDGPU_GEM_DOMAIN_MASK)
 		return -EINVAL;
 
 	/* create a gem object to contain this object in */
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
index 9258f06..feece0a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
@@ -360,7 +360,6 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev,
 	};
 	struct amdgpu_bo *bo;
 	unsigned long page_align, size = bp->size;
-	u32 preferred_domains;
 	size_t acc_size;
 	int r;
 
@@ -381,14 +380,8 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev,
 	drm_gem_private_object_init(adev->ddev, &bo->gem_base, size);
 	INIT_LIST_HEAD(&bo->shadow_list);
 	INIT_LIST_HEAD(&bo->va);
-	preferred_domains = bp->preferred_domain ? bp->preferred_domain :
+	bo->preferred_domains = bp->preferred_domain ? bp->preferred_domain :
 		bp->domain;
-	bo->preferred_domains = preferred_domains & (AMDGPU_GEM_DOMAIN_VRAM |
-						     AMDGPU_GEM_DOMAIN_GTT |
-						     AMDGPU_GEM_DOMAIN_CPU |
-						     AMDGPU_GEM_DOMAIN_GDS |
-						     AMDGPU_GEM_DOMAIN_GWS |
-						     AMDGPU_GEM_DOMAIN_OA);
 	bo->allowed_domains = bo->preferred_domains;
 	if (bp->type != ttm_bo_type_kernel &&
 	    bo->allowed_domains == AMDGPU_GEM_DOMAIN_VRAM)
diff --git a/include/uapi/drm/amdgpu_drm.h b/include/uapi/drm/amdgpu_drm.h
index c363b67..b193e95f1 100644
--- a/include/uapi/drm/amdgpu_drm.h
+++ b/include/uapi/drm/amdgpu_drm.h
@@ -78,6 +78,12 @@ extern "C" {
 #define AMDGPU_GEM_DOMAIN_GDS		0x8
 #define AMDGPU_GEM_DOMAIN_GWS		0x10
 #define AMDGPU_GEM_DOMAIN_OA		0x20
+#define AMDGPU_GEM_DOMAIN_MASK		(AMDGPU_GEM_DOMAIN_CPU | \
+					 AMDGPU_GEM_DOMAIN_GTT | \
+					 AMDGPU_GEM_DOMAIN_VRAM | \
+					 AMDGPU_GEM_DOMAIN_GDS | \
+					 AMDGPU_GEM_DOMAIN_GWS | \
+					 AMDGPU_GEM_DOMAIN_OA)
 
 /* Flag that CPU access will be required for the case of VRAM domain */
 #define AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED	(1 << 0)
-- 
cgit v1.1


From d240cd9eddd943dbe0267d081697195ff1e90b65 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Marek=20Ol=C5=A1=C3=A1k?= <marek.olsak@amd.com>
Date: Tue, 3 Apr 2018 13:05:03 -0400
Subject: drm/amdgpu: optionally do a writeback but don't invalidate TC for IB
 fences
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

There is a new IB flag that enables this new behavior.
Full invalidation is unnecessary for RELEASE_MEM and doesn't make sense
when draw calls from two adjacent gfx IBs run in parallel. This will be
the new default for Mesa.

v2: bump the version

Signed-off-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c   |  3 ++-
 drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c |  5 +++--
 drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c    |  8 ++++++--
 drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h  |  4 +++-
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c    |  2 +-
 drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c     | 11 +++++++----
 drivers/gpu/drm/amd/amdgpu/soc15d.h       |  1 +
 include/uapi/drm/amdgpu_drm.h             |  4 ++++
 8 files changed, 27 insertions(+), 11 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
index 5c0567a..7c17a0b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
@@ -75,9 +75,10 @@
  * - 3.23.0 - Add query for VRAM lost counter
  * - 3.24.0 - Add high priority compute support for gfx9
  * - 3.25.0 - Add support for sensor query info (stable pstate sclk/mclk).
+ * - 3.26.0 - GFX9: Process AMDGPU_IB_FLAG_TC_WB_NOT_INVALIDATE.
  */
 #define KMS_DRIVER_MAJOR	3
-#define KMS_DRIVER_MINOR	25
+#define KMS_DRIVER_MINOR	26
 #define KMS_DRIVER_PATCHLEVEL	0
 
 int amdgpu_vram_limit = 0;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
index 97449e0..d09fcab 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
@@ -131,7 +131,8 @@ static u32 amdgpu_fence_read(struct amdgpu_ring *ring)
  * Emits a fence command on the requested ring (all asics).
  * Returns 0 on success, -ENOMEM on failure.
  */
-int amdgpu_fence_emit(struct amdgpu_ring *ring, struct dma_fence **f)
+int amdgpu_fence_emit(struct amdgpu_ring *ring, struct dma_fence **f,
+		      unsigned flags)
 {
 	struct amdgpu_device *adev = ring->adev;
 	struct amdgpu_fence *fence;
@@ -149,7 +150,7 @@ int amdgpu_fence_emit(struct amdgpu_ring *ring, struct dma_fence **f)
 		       adev->fence_context + ring->idx,
 		       seq);
 	amdgpu_ring_emit_fence(ring, ring->fence_drv.gpu_addr,
-			       seq, AMDGPU_FENCE_FLAG_INT);
+			       seq, flags | AMDGPU_FENCE_FLAG_INT);
 
 	ptr = &ring->fence_drv.fences[seq & ring->fence_drv.num_fences_mask];
 	/* This function can't be called concurrently anyway, otherwise
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
index 311589e..f70eeed 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
@@ -127,6 +127,7 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs,
 	struct amdgpu_vm *vm;
 	uint64_t fence_ctx;
 	uint32_t status = 0, alloc_size;
+	unsigned fence_flags = 0;
 
 	unsigned i;
 	int r = 0;
@@ -227,7 +228,10 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs,
 #endif
 		amdgpu_asic_invalidate_hdp(adev, ring);
 
-	r = amdgpu_fence_emit(ring, f);
+	if (ib->flags & AMDGPU_IB_FLAG_TC_WB_NOT_INVALIDATE)
+		fence_flags |= AMDGPU_FENCE_FLAG_TC_WB_ONLY;
+
+	r = amdgpu_fence_emit(ring, f, fence_flags);
 	if (r) {
 		dev_err(adev->dev, "failed to emit fence (%d)\n", r);
 		if (job && job->vmid)
@@ -242,7 +246,7 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs,
 	/* wrap the last IB with fence */
 	if (job && job->uf_addr) {
 		amdgpu_ring_emit_fence(ring, job->uf_addr, job->uf_sequence,
-				       AMDGPU_FENCE_FLAG_64BIT);
+				       fence_flags | AMDGPU_FENCE_FLAG_64BIT);
 	}
 
 	if (patch_offset != ~0 && ring->funcs->patch_cond_exec)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h
index 08fcdf6..4f8dac2 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h
@@ -42,6 +42,7 @@
 
 #define AMDGPU_FENCE_FLAG_64BIT         (1 << 0)
 #define AMDGPU_FENCE_FLAG_INT           (1 << 1)
+#define AMDGPU_FENCE_FLAG_TC_WB_ONLY    (1 << 2)
 
 enum amdgpu_ring_type {
 	AMDGPU_RING_TYPE_GFX,
@@ -90,7 +91,8 @@ int amdgpu_fence_driver_start_ring(struct amdgpu_ring *ring,
 				   unsigned irq_type);
 void amdgpu_fence_driver_suspend(struct amdgpu_device *adev);
 void amdgpu_fence_driver_resume(struct amdgpu_device *adev);
-int amdgpu_fence_emit(struct amdgpu_ring *ring, struct dma_fence **fence);
+int amdgpu_fence_emit(struct amdgpu_ring *ring, struct dma_fence **fence,
+		      unsigned flags);
 int amdgpu_fence_emit_polling(struct amdgpu_ring *ring, uint32_t *s);
 void amdgpu_fence_process(struct amdgpu_ring *ring);
 int amdgpu_fence_wait_empty(struct amdgpu_ring *ring);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index 9ec7c10..9c2195a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -633,7 +633,7 @@ int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job, bool need_
 		amdgpu_gmc_emit_pasid_mapping(ring, job->vmid, job->pasid);
 
 	if (vm_flush_needed || pasid_mapping_needed) {
-		r = amdgpu_fence_emit(ring, &fence);
+		r = amdgpu_fence_emit(ring, &fence, 0);
 		if (r)
 			return r;
 	}
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index 6a19e03..05b2d34 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -3775,13 +3775,16 @@ static void gfx_v9_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr,
 {
 	bool write64bit = flags & AMDGPU_FENCE_FLAG_64BIT;
 	bool int_sel = flags & AMDGPU_FENCE_FLAG_INT;
+	bool writeback = flags & AMDGPU_FENCE_FLAG_TC_WB_ONLY;
 
 	/* RELEASE_MEM - flush caches, send int */
 	amdgpu_ring_write(ring, PACKET3(PACKET3_RELEASE_MEM, 6));
-	amdgpu_ring_write(ring, (EOP_TCL1_ACTION_EN |
-				 EOP_TC_ACTION_EN |
-				 EOP_TC_WB_ACTION_EN |
-				 EOP_TC_MD_ACTION_EN |
+	amdgpu_ring_write(ring, ((writeback ? (EOP_TC_WB_ACTION_EN |
+					       EOP_TC_NC_ACTION_EN) :
+					      (EOP_TCL1_ACTION_EN |
+					       EOP_TC_ACTION_EN |
+					       EOP_TC_WB_ACTION_EN |
+					       EOP_TC_MD_ACTION_EN)) |
 				 EVENT_TYPE(CACHE_FLUSH_AND_INV_TS_EVENT) |
 				 EVENT_INDEX(5)));
 	amdgpu_ring_write(ring, DATA_SEL(write64bit ? 2 : 1) | INT_SEL(int_sel ? 2 : 0));
diff --git a/drivers/gpu/drm/amd/amdgpu/soc15d.h b/drivers/gpu/drm/amd/amdgpu/soc15d.h
index 7f408f8..839a144 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15d.h
+++ b/drivers/gpu/drm/amd/amdgpu/soc15d.h
@@ -159,6 +159,7 @@
 #define		EOP_TC_WB_ACTION_EN                     (1 << 15) /* L2 */
 #define		EOP_TCL1_ACTION_EN                      (1 << 16)
 #define		EOP_TC_ACTION_EN                        (1 << 17) /* L2 */
+#define		EOP_TC_NC_ACTION_EN			(1 << 19)
 #define		EOP_TC_MD_ACTION_EN			(1 << 21) /* L2 metadata */
 
 #define		DATA_SEL(x)                             ((x) << 29)
diff --git a/include/uapi/drm/amdgpu_drm.h b/include/uapi/drm/amdgpu_drm.h
index b193e95f1..78fe828 100644
--- a/include/uapi/drm/amdgpu_drm.h
+++ b/include/uapi/drm/amdgpu_drm.h
@@ -526,6 +526,10 @@ union drm_amdgpu_cs {
 /* Preempt flag, IB should set Pre_enb bit if PREEMPT flag detected */
 #define AMDGPU_IB_FLAG_PREEMPT (1<<2)
 
+/* The IB fence should do the L2 writeback but not invalidate any shader
+ * caches (L2/vL1/sL1/I$). */
+#define AMDGPU_IB_FLAG_TC_WB_NOT_INVALIDATE (1 << 3)
+
 struct drm_amdgpu_cs_chunk_ib {
 	__u32 _pad;
 	/** AMDGPU_IB_FLAG_* */
-- 
cgit v1.1


From 7fd645f258711a4ea4d777188949494f9e68b787 Mon Sep 17 00:00:00 2001
From: Chunming Zhou <david1.zhou@amd.com>
Date: Wed, 18 Apr 2018 18:35:09 +0800
Subject: drm/amdgpu: fix list not initialized
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Otherwise, cpu stuck for 22s with kernel panic.

Signed-off-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index 9c2195a..8c34060 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -1568,10 +1568,9 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev,
 		 * the evicted list so that it gets validated again on the
 		 * next command submission.
 		 */
+		list_del_init(&bo_va->base.vm_status);
 		if (!(bo->preferred_domains & amdgpu_mem_type_to_domain(mem_type)))
 			list_add_tail(&bo_va->base.vm_status, &vm->evicted);
-		else
-			list_del_init(&bo_va->base.vm_status);
 	} else {
 		list_del_init(&bo_va->base.vm_status);
 	}
-- 
cgit v1.1


From 6197ae28911841369ff61ebbdf9d732ff6069138 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Thu, 19 Apr 2018 12:40:15 +0800
Subject: drm/amd/pp: Fix NULL point check error in
 smu_set_watermarks_for_clocks_ranges

It is caused by
'commit d6c9a7dc86cd ("drm/amd/pp: Move common code to smu_helper.c")'

Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.c
index 7c23741..93a3d02 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.c
@@ -657,7 +657,7 @@ int smu_set_watermarks_for_clocks_ranges(void *wt_table,
 	uint32_t i;
 	struct watermarks *table = wt_table;
 
-	if (!table || wm_with_clock_ranges)
+	if (!table || !wm_with_clock_ranges)
 		return -EINVAL;
 
 	if (wm_with_clock_ranges->num_wm_sets_dmif > 4 || wm_with_clock_ranges->num_wm_sets_mcif > 4)
-- 
cgit v1.1


From bfa8eea29b12e403b391820b7ef5cf5c77ab0afe Mon Sep 17 00:00:00 2001
From: Flora Cui <Flora.Cui@amd.com>
Date: Wed, 18 Apr 2018 17:12:19 +0800
Subject: drm/amdgpu: init gfx9 aperture settings

fix settings.

Signed-off-by: Flora Cui <Flora.Cui@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 5 ++++-
 drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c | 3 +--
 2 files changed, 5 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index 05b2d34..587a873 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -1617,7 +1617,10 @@ static void gfx_v9_0_gpu_init(struct amdgpu_device *adev)
 			tmp = REG_SET_FIELD(0, SH_MEM_CONFIG, ALIGNMENT_MODE,
 					    SH_MEM_ALIGNMENT_MODE_UNALIGNED);
 			WREG32_SOC15(GC, 0, mmSH_MEM_CONFIG, tmp);
-			tmp = adev->gmc.shared_aperture_start >> 48;
+			tmp = REG_SET_FIELD(0, SH_MEM_BASES, PRIVATE_BASE,
+				(adev->gmc.private_aperture_start >> 48));
+			tmp = REG_SET_FIELD(tmp, SH_MEM_BASES, SHARED_BASE,
+				(adev->gmc.shared_aperture_start >> 48));
 			WREG32_SOC15(GC, 0, mmSH_MEM_BASES, tmp);
 		}
 	}
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
index e6b00b5..6c9f7f9 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
@@ -562,8 +562,7 @@ static int gmc_v9_0_early_init(void *handle)
 	adev->gmc.shared_aperture_start = 0x2000000000000000ULL;
 	adev->gmc.shared_aperture_end =
 		adev->gmc.shared_aperture_start + (4ULL << 30) - 1;
-	adev->gmc.private_aperture_start =
-		adev->gmc.shared_aperture_end + 1;
+	adev->gmc.private_aperture_start = 0x1000000000000000ULL;
 	adev->gmc.private_aperture_end =
 		adev->gmc.private_aperture_start + (4ULL << 30) - 1;
 
-- 
cgit v1.1


From bb475839eca7e3990f59a3b4e9e810635ef0ac4a Mon Sep 17 00:00:00 2001
From: Junwei Zhang <Jerry.Zhang@amd.com>
Date: Thu, 19 Apr 2018 13:17:26 +0800
Subject: drm/amdgpu: simplify bo_va list when vm bo update (v2)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

v2: fix compiling warning

Signed-off-by: Junwei Zhang <Jerry.Zhang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 22 ++++++++++------------
 1 file changed, 10 insertions(+), 12 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index 8c34060..6a372ca 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -1509,6 +1509,7 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev,
 	struct drm_mm_node *nodes;
 	struct dma_fence *exclusive, **last_update;
 	uint64_t flags;
+	uint32_t mem_type;
 	int r;
 
 	if (clear || !bo_va->base.bo) {
@@ -1561,19 +1562,16 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev,
 	}
 
 	spin_lock(&vm->status_lock);
-	if (bo && bo->tbo.resv == vm->root.base.bo->tbo.resv) {
-		unsigned mem_type = bo->tbo.mem.mem_type;
+	list_del_init(&bo_va->base.vm_status);
 
-		/* If the BO is not in its preferred location add it back to
-		 * the evicted list so that it gets validated again on the
-		 * next command submission.
-		 */
-		list_del_init(&bo_va->base.vm_status);
-		if (!(bo->preferred_domains & amdgpu_mem_type_to_domain(mem_type)))
-			list_add_tail(&bo_va->base.vm_status, &vm->evicted);
-	} else {
-		list_del_init(&bo_va->base.vm_status);
-	}
+	/* If the BO is not in its preferred location add it back to
+	 * the evicted list so that it gets validated again on the
+	 * next command submission.
+	 */
+	mem_type = bo->tbo.mem.mem_type;
+	if (bo && bo->tbo.resv == vm->root.base.bo->tbo.resv &&
+	    !(bo->preferred_domains & amdgpu_mem_type_to_domain(mem_type)))
+		list_add_tail(&bo_va->base.vm_status, &vm->evicted);
 	spin_unlock(&vm->status_lock);
 
 	list_splice_init(&bo_va->invalids, &bo_va->valids);
-- 
cgit v1.1


From 1a3132a1cc03abcf153d08f4eb471cd7d396f2a3 Mon Sep 17 00:00:00 2001
From: Kenneth Feng <kenneth.feng@amd.com>
Date: Tue, 17 Apr 2018 21:49:51 +0800
Subject: drm/amd/powerplay: header file interface to SMU update

update vega12 smu interface.

Signed-off-by: Kenneth Feng <kenneth.feng@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/inc/vega12/smu9_driver_if.h | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/powerplay/inc/vega12/smu9_driver_if.h b/drivers/gpu/drm/amd/powerplay/inc/vega12/smu9_driver_if.h
index fb696e3..2f8a3b9 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/vega12/smu9_driver_if.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/vega12/smu9_driver_if.h
@@ -412,8 +412,10 @@ typedef struct {
   QuadraticInt_t    ReservedEquation2;
   QuadraticInt_t    ReservedEquation3;
 
+	uint16_t     MinVoltageUlvGfx;
+	uint16_t     MinVoltageUlvSoc;
 
-  uint32_t     Reserved[15];
+	uint32_t     Reserved[14];
 
 
 
-- 
cgit v1.1


From b1f223c02a3a2b41847f48f75797eba5979ea25d Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Christian=20K=C3=B6nig?= <christian.koenig@amd.com>
Date: Sun, 25 Mar 2018 10:10:25 +0200
Subject: drm/amdgpu: print DMA-buf status in debugfs
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Just note if a BO was imported/exported.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c | 11 +++++++++++
 1 file changed, 11 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
index c62c3dd..7d3dc22 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
@@ -780,6 +780,8 @@ static int amdgpu_debugfs_gem_bo_info(int id, void *ptr, void *data)
 	struct amdgpu_bo *bo = gem_to_amdgpu_bo(gobj);
 	struct seq_file *m = data;
 
+	struct dma_buf_attachment *attachment;
+	struct dma_buf *dma_buf;
 	unsigned domain;
 	const char *placement;
 	unsigned pin_count;
@@ -808,6 +810,15 @@ static int amdgpu_debugfs_gem_bo_info(int id, void *ptr, void *data)
 	pin_count = READ_ONCE(bo->pin_count);
 	if (pin_count)
 		seq_printf(m, " pin count %d", pin_count);
+
+	dma_buf = READ_ONCE(bo->gem_base.dma_buf);
+	attachment = READ_ONCE(bo->gem_base.import_attach);
+
+	if (attachment)
+		seq_printf(m, " imported from %p", dma_buf);
+	else if (dma_buf)
+		seq_printf(m, " exported as %p", dma_buf);
+
 	seq_printf(m, "\n");
 
 	return 0;
-- 
cgit v1.1


From e0e93d03efa1c53012cc609fd48112df3e06da69 Mon Sep 17 00:00:00 2001
From: Kenneth Feng <kenneth.feng@amd.com>
Date: Fri, 20 Apr 2018 13:55:39 +0800
Subject: drm/amd/powerplay: add registry key to disable ACG

For the dummy ACG fuses,need to disable ACG, otherwise
corruption will be caused.

Signed-off-by: Kenneth Feng <kenneth.feng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/vega12_processpptables.c | 5 +++++
 drivers/gpu/drm/amd/powerplay/inc/hwmgr.h                    | 1 +
 2 files changed, 6 insertions(+)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_processpptables.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_processpptables.c
index 7fa1ba8..888ddca 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_processpptables.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_processpptables.c
@@ -224,6 +224,11 @@ static int append_vbios_pptable(struct pp_hwmgr *hwmgr, PPTable_t *ppsmc_pptable
 	ppsmc_pptable->AcgGfxclkSpreadPercent = smc_dpm_table.acggfxclkspreadpercent;
 	ppsmc_pptable->AcgGfxclkSpreadFreq = smc_dpm_table.acggfxclkspreadfreq;
 
+	/* 0xFFFF will disable the ACG feature */
+	if (!(hwmgr->feature_mask & PP_ACG_MASK)) {
+		ppsmc_pptable->AcgThresholdFreqHigh = 0xFFFF;
+		ppsmc_pptable->AcgThresholdFreqLow = 0xFFFF;
+	}
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
index 9b3dd7d..2f203ec 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
@@ -82,6 +82,7 @@ enum PP_FEATURE_MASK {
 	PP_SOCCLK_DPM_MASK = 0x1000,
 	PP_DCEFCLK_DPM_MASK = 0x2000,
 	PP_OVERDRIVE_MASK = 0x4000,
+	PP_ACG_MASK = 0x10000,
 };
 
 enum PHM_BackEnd_Magic {
-- 
cgit v1.1


From cf671071334ebbf6c960f88383b35b99d5d53212 Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Tue, 5 Dec 2017 18:48:48 +0800
Subject: drm/amdgpu: update psp gfx if header
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Signed-off-by: Huang Rui <ray.huang@amd.com>
Acked-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/psp_gfx_if.h | 67 ++++++++++++++++++++++-----------
 1 file changed, 46 insertions(+), 21 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/psp_gfx_if.h b/drivers/gpu/drm/amd/amdgpu/psp_gfx_if.h
index 8da6da9..0cf48d2 100644
--- a/drivers/gpu/drm/amd/amdgpu/psp_gfx_if.h
+++ b/drivers/gpu/drm/amd/amdgpu/psp_gfx_if.h
@@ -40,11 +40,20 @@ enum psp_gfx_crtl_cmd_id
     GFX_CTRL_CMD_ID_INIT_GPCOM_RING = 0x00020000,   /* initialize GPCOM ring */
     GFX_CTRL_CMD_ID_DESTROY_RINGS   = 0x00030000,   /* destroy rings */
     GFX_CTRL_CMD_ID_CAN_INIT_RINGS  = 0x00040000,   /* is it allowed to initialized the rings */
+    GFX_CTRL_CMD_ID_ENABLE_INT      = 0x00050000,   /* enable PSP-to-Gfx interrupt */
+    GFX_CTRL_CMD_ID_DISABLE_INT     = 0x00060000,   /* disable PSP-to-Gfx interrupt */
+    GFX_CTRL_CMD_ID_MODE1_RST       = 0x00070000,   /* trigger the Mode 1 reset */
 
     GFX_CTRL_CMD_ID_MAX             = 0x000F0000,   /* max command ID */
 };
 
 
+/*-----------------------------------------------------------------------------
+    NOTE:   All physical addresses used in this interface are actually
+            GPU Virtual Addresses.
+*/
+
+
 /* Control registers of the TEE Gfx interface. These are located in
 *  SRBM-to-PSP mailbox registers (total 8 registers).
 */
@@ -55,8 +64,8 @@ struct psp_gfx_ctrl
     volatile uint32_t   rbi_rptr;         /* +8   Read pointer (index) of RBI ring */
     volatile uint32_t   gpcom_wptr;       /* +12  Write pointer (index) of GPCOM ring */
     volatile uint32_t   gpcom_rptr;       /* +16  Read pointer (index) of GPCOM ring */
-    volatile uint32_t   ring_addr_lo;     /* +20  bits [31:0] of physical address of ring buffer */
-    volatile uint32_t   ring_addr_hi;     /* +24  bits [63:32] of physical address of ring buffer */
+    volatile uint32_t   ring_addr_lo;     /* +20  bits [31:0] of GPU Virtual of ring buffer (VMID=0)*/
+    volatile uint32_t   ring_addr_hi;     /* +24  bits [63:32] of GPU Virtual of ring buffer (VMID=0) */
     volatile uint32_t   ring_buf_size;    /* +28  Ring buffer size (in bytes) */
 
 };
@@ -78,6 +87,8 @@ enum psp_gfx_cmd_id
     GFX_CMD_ID_LOAD_ASD     = 0x00000004,   /* load ASD Driver */
     GFX_CMD_ID_SETUP_TMR    = 0x00000005,   /* setup TMR region */
     GFX_CMD_ID_LOAD_IP_FW   = 0x00000006,   /* load HW IP FW */
+    GFX_CMD_ID_DESTROY_TMR  = 0x00000007,   /* destroy TMR region */
+    GFX_CMD_ID_SAVE_RESTORE = 0x00000008,   /* save/restore HW IP FW */
 
 };
 
@@ -85,11 +96,11 @@ enum psp_gfx_cmd_id
 /* Command to load Trusted Application binary into PSP OS. */
 struct psp_gfx_cmd_load_ta
 {
-    uint32_t        app_phy_addr_lo;        /* bits [31:0] of the physical address of the TA binary (must be 4 KB aligned) */
-    uint32_t        app_phy_addr_hi;        /* bits [63:32] of the physical address of the TA binary */
+    uint32_t        app_phy_addr_lo;        /* bits [31:0] of the GPU Virtual address of the TA binary (must be 4 KB aligned) */
+    uint32_t        app_phy_addr_hi;        /* bits [63:32] of the GPU Virtual address of the TA binary */
     uint32_t        app_len;                /* length of the TA binary in bytes */
-    uint32_t        cmd_buf_phy_addr_lo;    /* bits [31:0] of the physical address of CMD buffer (must be 4 KB aligned) */
-    uint32_t        cmd_buf_phy_addr_hi;    /* bits [63:32] of the physical address of CMD buffer */
+    uint32_t        cmd_buf_phy_addr_lo;    /* bits [31:0] of the GPU Virtual address of CMD buffer (must be 4 KB aligned) */
+    uint32_t        cmd_buf_phy_addr_hi;    /* bits [63:32] of the GPU Virtual address of CMD buffer */
     uint32_t        cmd_buf_len;            /* length of the CMD buffer in bytes; must be multiple of 4 KB */
 
     /* Note: CmdBufLen can be set to 0. In this case no persistent CMD buffer is provided
@@ -111,8 +122,8 @@ struct psp_gfx_cmd_unload_ta
 */
 struct psp_gfx_buf_desc
 {
-    uint32_t        buf_phy_addr_lo;       /* bits [31:0] of physical address of the buffer (must be 4 KB aligned) */
-    uint32_t        buf_phy_addr_hi;       /* bits [63:32] of physical address of the buffer */
+    uint32_t        buf_phy_addr_lo;       /* bits [31:0] of GPU Virtual address of the buffer (must be 4 KB aligned) */
+    uint32_t        buf_phy_addr_hi;       /* bits [63:32] of GPU Virtual address of the buffer */
     uint32_t        buf_size;              /* buffer size in bytes (must be multiple of 4 KB and no bigger than 64 MB) */
 
 };
@@ -145,8 +156,8 @@ struct psp_gfx_cmd_invoke_cmd
 /* Command to setup TMR region. */
 struct psp_gfx_cmd_setup_tmr
 {
-    uint32_t        buf_phy_addr_lo;       /* bits [31:0] of physical address of TMR buffer (must be 4 KB aligned) */
-    uint32_t        buf_phy_addr_hi;       /* bits [63:32] of physical address of TMR buffer */
+    uint32_t        buf_phy_addr_lo;       /* bits [31:0] of GPU Virtual address of TMR buffer (must be 4 KB aligned) */
+    uint32_t        buf_phy_addr_hi;       /* bits [63:32] of GPU Virtual address of TMR buffer */
     uint32_t        buf_size;              /* buffer size in bytes (must be multiple of 4 KB) */
 
 };
@@ -174,18 +185,32 @@ enum psp_gfx_fw_type
     GFX_FW_TYPE_ISP         = 16,
     GFX_FW_TYPE_ACP         = 17,
     GFX_FW_TYPE_SMU         = 18,
+    GFX_FW_TYPE_MMSCH       = 19,
+    GFX_FW_TYPE_RLC_RESTORE_LIST_GPM_MEM        = 20,
+    GFX_FW_TYPE_RLC_RESTORE_LIST_SRM_MEM        = 21,
+    GFX_FW_TYPE_RLC_RESTORE_LIST_CNTL           = 22,
+    GFX_FW_TYPE_MAX         = 23
 };
 
 /* Command to load HW IP FW. */
 struct psp_gfx_cmd_load_ip_fw
 {
-    uint32_t                fw_phy_addr_lo;    /* bits [31:0] of physical address of FW location (must be 4 KB aligned) */
-    uint32_t                fw_phy_addr_hi;    /* bits [63:32] of physical address of FW location */
+    uint32_t                fw_phy_addr_lo;    /* bits [31:0] of GPU Virtual address of FW location (must be 4 KB aligned) */
+    uint32_t                fw_phy_addr_hi;    /* bits [63:32] of GPU Virtual address of FW location */
     uint32_t                fw_size;           /* FW buffer size in bytes */
     enum psp_gfx_fw_type    fw_type;           /* FW type */
 
 };
 
+/* Command to save/restore HW IP FW. */
+struct psp_gfx_cmd_save_restore_ip_fw
+{
+    uint32_t                save_fw;              /* if set, command is used for saving fw otherwise for resetoring*/
+    uint32_t                save_restore_addr_lo; /* bits [31:0] of FB address of GART memory used as save/restore buffer (must be 4 KB aligned) */
+    uint32_t                save_restore_addr_hi; /* bits [63:32] of FB address of GART memory used as save/restore buffer */
+    uint32_t                buf_size;             /* Size of the save/restore buffer in bytes */
+    enum psp_gfx_fw_type    fw_type;              /* FW type */
+};
 
 /* All GFX ring buffer commands. */
 union psp_gfx_commands
@@ -195,7 +220,7 @@ union psp_gfx_commands
     struct psp_gfx_cmd_invoke_cmd       cmd_invoke_cmd;
     struct psp_gfx_cmd_setup_tmr        cmd_setup_tmr;
     struct psp_gfx_cmd_load_ip_fw       cmd_load_ip_fw;
-
+    struct psp_gfx_cmd_save_restore_ip_fw cmd_save_restore_ip_fw;
 };
 
 
@@ -226,8 +251,8 @@ struct psp_gfx_cmd_resp
 
     /* These fields are used for RBI only. They are all 0 in GPCOM commands
     */
-    uint32_t        resp_buf_addr_lo;   /* +12 bits [31:0] of physical address of response buffer (must be 4 KB aligned) */
-    uint32_t        resp_buf_addr_hi;   /* +16 bits [63:32] of physical address of response buffer */
+    uint32_t        resp_buf_addr_lo;   /* +12 bits [31:0] of GPU Virtual address of response buffer (must be 4 KB aligned) */
+    uint32_t        resp_buf_addr_hi;   /* +16 bits [63:32] of GPU Virtual address of response buffer */
     uint32_t        resp_offset;        /* +20 offset within response buffer */
     uint32_t        resp_buf_size;      /* +24 total size of the response buffer in bytes */
 
@@ -251,19 +276,19 @@ struct psp_gfx_cmd_resp
 /* Structure of the Ring Buffer Frame */
 struct psp_gfx_rb_frame
 {
-    uint32_t    cmd_buf_addr_lo;    /* +0  bits [31:0] of physical address of command buffer (must be 4 KB aligned) */
-    uint32_t    cmd_buf_addr_hi;    /* +4  bits [63:32] of physical address of command buffer */
+    uint32_t    cmd_buf_addr_lo;    /* +0  bits [31:0] of GPU Virtual address of command buffer (must be 4 KB aligned) */
+    uint32_t    cmd_buf_addr_hi;    /* +4  bits [63:32] of GPU Virtual address of command buffer */
     uint32_t    cmd_buf_size;       /* +8  command buffer size in bytes */
-    uint32_t    fence_addr_lo;      /* +12 bits [31:0] of physical address of Fence for this frame */
-    uint32_t    fence_addr_hi;      /* +16 bits [63:32] of physical address of Fence for this frame */
+    uint32_t    fence_addr_lo;      /* +12 bits [31:0] of GPU Virtual address of Fence for this frame */
+    uint32_t    fence_addr_hi;      /* +16 bits [63:32] of GPU Virtual address of Fence for this frame */
     uint32_t    fence_value;        /* +20 Fence value */
     uint32_t    sid_lo;             /* +24 bits [31:0] of SID value (used only for RBI frames) */
     uint32_t    sid_hi;             /* +28 bits [63:32] of SID value (used only for RBI frames) */
     uint8_t     vmid;               /* +32 VMID value used for mapping of all addresses for this frame */
     uint8_t     frame_type;         /* +33 1: destory context frame, 0: all other frames; used only for RBI frames */
     uint8_t     reserved1[2];       /* +34 reserved, must be 0 */
-    uint32_t    reserved2[7];       /* +40 reserved, must be 0 */
-    /* total 64 bytes */
+    uint32_t    reserved2[7];       /* +36 reserved, must be 0 */
+                /* total 64 bytes */
 };
 
 #endif /* _PSP_TEE_GFX_IF_H_ */
-- 
cgit v1.1


From d40e9b13c8bad15e56f2e8c9572f62c1229833a6 Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Mon, 22 Jan 2018 17:51:35 +0800
Subject: drm/amdgpu: add new rlc firmware header format v2.1
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Signed-off-by: Huang Rui <ray.huang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c | 34 +++++++++++++++++++++++++++++--
 drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h | 19 +++++++++++++++++
 2 files changed, 51 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
index dd6f989..84d6525 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
@@ -161,8 +161,38 @@ void amdgpu_ucode_print_rlc_hdr(const struct common_firmware_header *hdr)
 			  le32_to_cpu(rlc_hdr->reg_list_format_separate_array_offset_bytes));
 		DRM_DEBUG("reg_list_separate_size_bytes: %u\n",
 			  le32_to_cpu(rlc_hdr->reg_list_separate_size_bytes));
-		DRM_DEBUG("reg_list_separate_size_bytes: %u\n",
-			  le32_to_cpu(rlc_hdr->reg_list_separate_size_bytes));
+		DRM_DEBUG("reg_list_separate_array_offset_bytes: %u\n",
+			  le32_to_cpu(rlc_hdr->reg_list_separate_array_offset_bytes));
+		if (version_minor == 1) {
+			const struct rlc_firmware_header_v2_1 *v2_1 =
+				container_of(rlc_hdr, struct rlc_firmware_header_v2_1, v2_0);
+			DRM_DEBUG("reg_list_format_direct_reg_list_length: %u\n",
+				  le32_to_cpu(v2_1->reg_list_format_direct_reg_list_length));
+			DRM_DEBUG("save_restore_list_cntl_ucode_ver: %u\n",
+				  le32_to_cpu(v2_1->save_restore_list_cntl_ucode_ver));
+			DRM_DEBUG("save_restore_list_cntl_feature_ver: %u\n",
+				  le32_to_cpu(v2_1->save_restore_list_cntl_feature_ver));
+			DRM_DEBUG("save_restore_list_cntl_size_bytes %u\n",
+				  le32_to_cpu(v2_1->save_restore_list_cntl_size_bytes));
+			DRM_DEBUG("save_restore_list_cntl_offset_bytes: %u\n",
+				  le32_to_cpu(v2_1->save_restore_list_cntl_offset_bytes));
+			DRM_DEBUG("save_restore_list_gpm_ucode_ver: %u\n",
+				  le32_to_cpu(v2_1->save_restore_list_gpm_ucode_ver));
+			DRM_DEBUG("save_restore_list_gpm_feature_ver: %u\n",
+				  le32_to_cpu(v2_1->save_restore_list_gpm_feature_ver));
+			DRM_DEBUG("save_restore_list_gpm_size_bytes %u\n",
+				  le32_to_cpu(v2_1->save_restore_list_gpm_size_bytes));
+			DRM_DEBUG("save_restore_list_gpm_offset_bytes: %u\n",
+				  le32_to_cpu(v2_1->save_restore_list_gpm_offset_bytes));
+			DRM_DEBUG("save_restore_list_srm_ucode_ver: %u\n",
+				  le32_to_cpu(v2_1->save_restore_list_srm_ucode_ver));
+			DRM_DEBUG("save_restore_list_srm_feature_ver: %u\n",
+				  le32_to_cpu(v2_1->save_restore_list_srm_feature_ver));
+			DRM_DEBUG("save_restore_list_srm_size_bytes %u\n",
+				  le32_to_cpu(v2_1->save_restore_list_srm_size_bytes));
+			DRM_DEBUG("save_restore_list_srm_offset_bytes: %u\n",
+				  le32_to_cpu(v2_1->save_restore_list_srm_offset_bytes));
+		}
 	} else {
 		DRM_ERROR("Unknown RLC ucode version: %u.%u\n", version_major, version_minor);
 	}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h
index 30b5500..0b262f4 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h
@@ -98,6 +98,24 @@ struct rlc_firmware_header_v2_0 {
 	uint32_t reg_list_separate_array_offset_bytes; /* payload offset from the start of the header */
 };
 
+/* version_major=2, version_minor=1 */
+struct rlc_firmware_header_v2_1 {
+	struct rlc_firmware_header_v2_0 v2_0;
+	uint32_t reg_list_format_direct_reg_list_length; /* length of direct reg list format array */
+	uint32_t save_restore_list_cntl_ucode_ver;
+	uint32_t save_restore_list_cntl_feature_ver;
+	uint32_t save_restore_list_cntl_size_bytes;
+	uint32_t save_restore_list_cntl_offset_bytes;
+	uint32_t save_restore_list_gpm_ucode_ver;
+	uint32_t save_restore_list_gpm_feature_ver;
+	uint32_t save_restore_list_gpm_size_bytes;
+	uint32_t save_restore_list_gpm_offset_bytes;
+	uint32_t save_restore_list_srm_ucode_ver;
+	uint32_t save_restore_list_srm_feature_ver;
+	uint32_t save_restore_list_srm_size_bytes;
+	uint32_t save_restore_list_srm_offset_bytes;
+};
+
 /* version_major=1, version_minor=0 */
 struct sdma_firmware_header_v1_0 {
 	struct common_firmware_header header;
@@ -148,6 +166,7 @@ union amdgpu_firmware_header {
 	struct gfx_firmware_header_v1_0 gfx;
 	struct rlc_firmware_header_v1_0 rlc;
 	struct rlc_firmware_header_v2_0 rlc_v2_0;
+	struct rlc_firmware_header_v2_1 rlc_v2_1;
 	struct sdma_firmware_header_v1_0 sdma;
 	struct sdma_firmware_header_v1_1 sdma_v1_1;
 	struct gpu_info_firmware_header_v1_0 gpu_info;
-- 
cgit v1.1


From 621a6318adea69b08a3652c64bc7cc0cb4dacfb4 Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Mon, 22 Jan 2018 20:48:14 +0800
Subject: drm/amdgpu: add save restore list cntl gpm and srm firmware support
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

RLC save/restore list cntl/gpm_mem/srm_mem ucodes are used for CGPG and gfxoff
function.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu.h       | 15 +++++++++
 drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c   | 36 ++++++++++++++++++++
 drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c | 17 +++++++++-
 drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h |  3 ++
 drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c     | 55 +++++++++++++++++++++++++++++--
 drivers/gpu/drm/amd/amdgpu/psp_v10_0.c    |  9 +++++
 include/uapi/drm/amdgpu_drm.h             |  6 ++++
 7 files changed, 138 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index d64ef30..5ad8939 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -774,9 +774,18 @@ struct amdgpu_rlc {
 	u32 starting_offsets_start;
 	u32 reg_list_format_size_bytes;
 	u32 reg_list_size_bytes;
+	u32 reg_list_format_direct_reg_list_length;
+	u32 save_restore_list_cntl_size_bytes;
+	u32 save_restore_list_gpm_size_bytes;
+	u32 save_restore_list_srm_size_bytes;
 
 	u32 *register_list_format;
 	u32 *register_restore;
+	u8 *save_restore_list_cntl;
+	u8 *save_restore_list_gpm;
+	u8 *save_restore_list_srm;
+
+	bool is_rlc_v2_1;
 };
 
 #define AMDGPU_MAX_COMPUTE_QUEUES KGD_MAX_QUEUES
@@ -943,6 +952,12 @@ struct amdgpu_gfx {
 	uint32_t			ce_feature_version;
 	uint32_t			pfp_feature_version;
 	uint32_t			rlc_feature_version;
+	uint32_t			rlc_srlc_fw_version;
+	uint32_t			rlc_srlc_feature_version;
+	uint32_t			rlc_srlg_fw_version;
+	uint32_t			rlc_srlg_feature_version;
+	uint32_t			rlc_srls_fw_version;
+	uint32_t			rlc_srls_feature_version;
 	uint32_t			mec_feature_version;
 	uint32_t			mec2_feature_version;
 	struct amdgpu_ring		gfx_ring[AMDGPU_MAX_GFX_RINGS];
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
index d602f8b..eb4785e 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
@@ -215,6 +215,18 @@ static int amdgpu_firmware_info(struct drm_amdgpu_info_firmware *fw_info,
 		fw_info->ver = adev->gfx.rlc_fw_version;
 		fw_info->feature = adev->gfx.rlc_feature_version;
 		break;
+	case AMDGPU_INFO_FW_GFX_RLC_RESTORE_LIST_CNTL:
+		fw_info->ver = adev->gfx.rlc_srlc_fw_version;
+		fw_info->feature = adev->gfx.rlc_srlc_feature_version;
+		break;
+	case AMDGPU_INFO_FW_GFX_RLC_RESTORE_LIST_GPM_MEM:
+		fw_info->ver = adev->gfx.rlc_srlg_fw_version;
+		fw_info->feature = adev->gfx.rlc_srlg_feature_version;
+		break;
+	case AMDGPU_INFO_FW_GFX_RLC_RESTORE_LIST_SRM_MEM:
+		fw_info->ver = adev->gfx.rlc_srls_fw_version;
+		fw_info->feature = adev->gfx.rlc_srls_feature_version;
+		break;
 	case AMDGPU_INFO_FW_GFX_MEC:
 		if (query_fw->index == 0) {
 			fw_info->ver = adev->gfx.mec_fw_version;
@@ -1149,6 +1161,30 @@ static int amdgpu_debugfs_firmware_info(struct seq_file *m, void *data)
 	seq_printf(m, "RLC feature version: %u, firmware version: 0x%08x\n",
 		   fw_info.feature, fw_info.ver);
 
+	/* RLC SAVE RESTORE LIST CNTL */
+	query_fw.fw_type = AMDGPU_INFO_FW_GFX_RLC_RESTORE_LIST_CNTL;
+	ret = amdgpu_firmware_info(&fw_info, &query_fw, adev);
+	if (ret)
+		return ret;
+	seq_printf(m, "RLC SRLC feature version: %u, firmware version: 0x%08x\n",
+		   fw_info.feature, fw_info.ver);
+
+	/* RLC SAVE RESTORE LIST GPM MEM */
+	query_fw.fw_type = AMDGPU_INFO_FW_GFX_RLC_RESTORE_LIST_GPM_MEM;
+	ret = amdgpu_firmware_info(&fw_info, &query_fw, adev);
+	if (ret)
+		return ret;
+	seq_printf(m, "RLC SRLG feature version: %u, firmware version: 0x%08x\n",
+		   fw_info.feature, fw_info.ver);
+
+	/* RLC SAVE RESTORE LIST SRM MEM */
+	query_fw.fw_type = AMDGPU_INFO_FW_GFX_RLC_RESTORE_LIST_SRM_MEM;
+	ret = amdgpu_firmware_info(&fw_info, &query_fw, adev);
+	if (ret)
+		return ret;
+	seq_printf(m, "RLC SRLS feature version: %u, firmware version: 0x%08x\n",
+		   fw_info.feature, fw_info.ver);
+
 	/* MEC */
 	query_fw.fw_type = AMDGPU_INFO_FW_GFX_MEC;
 	query_fw.index = 0;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
index 84d6525..0c74c09 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
@@ -337,7 +337,10 @@ static int amdgpu_ucode_init_single_fw(struct amdgpu_device *adev,
 	    (ucode->ucode_id != AMDGPU_UCODE_ID_CP_MEC1 &&
 	     ucode->ucode_id != AMDGPU_UCODE_ID_CP_MEC2 &&
 	     ucode->ucode_id != AMDGPU_UCODE_ID_CP_MEC1_JT &&
-	     ucode->ucode_id != AMDGPU_UCODE_ID_CP_MEC2_JT)) {
+	     ucode->ucode_id != AMDGPU_UCODE_ID_CP_MEC2_JT &&
+	     ucode->ucode_id != AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL &&
+	     ucode->ucode_id != AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM &&
+	     ucode->ucode_id != AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM)) {
 		ucode->ucode_size = le32_to_cpu(header->ucode_size_bytes);
 
 		memcpy(ucode->kaddr, (void *)((uint8_t *)ucode->fw->data +
@@ -359,6 +362,18 @@ static int amdgpu_ucode_init_single_fw(struct amdgpu_device *adev,
 					      le32_to_cpu(header->ucode_array_offset_bytes) +
 					      le32_to_cpu(cp_hdr->jt_offset) * 4),
 		       ucode->ucode_size);
+	} else if (ucode->ucode_id == AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL) {
+		ucode->ucode_size = adev->gfx.rlc.save_restore_list_cntl_size_bytes;
+		memcpy(ucode->kaddr, adev->gfx.rlc.save_restore_list_cntl,
+		       ucode->ucode_size);
+	} else if (ucode->ucode_id == AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM) {
+		ucode->ucode_size = adev->gfx.rlc.save_restore_list_gpm_size_bytes;
+		memcpy(ucode->kaddr, adev->gfx.rlc.save_restore_list_gpm,
+		       ucode->ucode_size);
+	} else if (ucode->ucode_id == AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM) {
+		ucode->ucode_size = adev->gfx.rlc.save_restore_list_srm_size_bytes;
+		memcpy(ucode->kaddr, adev->gfx.rlc.save_restore_list_srm,
+		       ucode->ucode_size);
 	}
 
 	return 0;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h
index 0b262f4..08e3857 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h
@@ -187,6 +187,9 @@ enum AMDGPU_UCODE_ID {
 	AMDGPU_UCODE_ID_CP_MEC2,
 	AMDGPU_UCODE_ID_CP_MEC2_JT,
 	AMDGPU_UCODE_ID_RLC_G,
+	AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL,
+	AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM,
+	AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM,
 	AMDGPU_UCODE_ID_STORAGE,
 	AMDGPU_UCODE_ID_SMC,
 	AMDGPU_UCODE_ID_UVD,
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index 587a873..73b76fa 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -41,7 +41,6 @@
 #define GFX9_MEC_HPD_SIZE 2048
 #define RLCG_UCODE_LOADING_START_ADDRESS 0x00002000L
 #define RLC_SAVE_RESTORE_ADDR_STARTING_OFFSET 0x00000000L
-#define GFX9_RLC_FORMAT_DIRECT_REG_LIST_LENGTH 34
 
 #define mmPWR_MISC_CNTL_STATUS					0x0183
 #define mmPWR_MISC_CNTL_STATUS_BASE_IDX				0
@@ -401,6 +400,27 @@ static void gfx_v9_0_free_microcode(struct amdgpu_device *adev)
 	kfree(adev->gfx.rlc.register_list_format);
 }
 
+static void gfx_v9_0_init_rlc_ext_microcode(struct amdgpu_device *adev)
+{
+	const struct rlc_firmware_header_v2_1 *rlc_hdr;
+
+	rlc_hdr = (const struct rlc_firmware_header_v2_1 *)adev->gfx.rlc_fw->data;
+	adev->gfx.rlc_srlc_fw_version = le32_to_cpu(rlc_hdr->save_restore_list_cntl_ucode_ver);
+	adev->gfx.rlc_srlc_feature_version = le32_to_cpu(rlc_hdr->save_restore_list_cntl_feature_ver);
+	adev->gfx.rlc.save_restore_list_cntl_size_bytes = le32_to_cpu(rlc_hdr->save_restore_list_cntl_size_bytes);
+	adev->gfx.rlc.save_restore_list_cntl = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->save_restore_list_cntl_offset_bytes);
+	adev->gfx.rlc_srlg_fw_version = le32_to_cpu(rlc_hdr->save_restore_list_gpm_ucode_ver);
+	adev->gfx.rlc_srlg_feature_version = le32_to_cpu(rlc_hdr->save_restore_list_gpm_feature_ver);
+	adev->gfx.rlc.save_restore_list_gpm_size_bytes = le32_to_cpu(rlc_hdr->save_restore_list_gpm_size_bytes);
+	adev->gfx.rlc.save_restore_list_gpm = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->save_restore_list_gpm_offset_bytes);
+	adev->gfx.rlc_srls_fw_version = le32_to_cpu(rlc_hdr->save_restore_list_srm_ucode_ver);
+	adev->gfx.rlc_srls_feature_version = le32_to_cpu(rlc_hdr->save_restore_list_srm_feature_ver);
+	adev->gfx.rlc.save_restore_list_srm_size_bytes = le32_to_cpu(rlc_hdr->save_restore_list_srm_size_bytes);
+	adev->gfx.rlc.save_restore_list_srm = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->save_restore_list_srm_offset_bytes);
+	adev->gfx.rlc.reg_list_format_direct_reg_list_length =
+			le32_to_cpu(rlc_hdr->reg_list_format_direct_reg_list_length);
+}
+
 static int gfx_v9_0_init_microcode(struct amdgpu_device *adev)
 {
 	const char *chip_name;
@@ -412,6 +432,8 @@ static int gfx_v9_0_init_microcode(struct amdgpu_device *adev)
 	const struct rlc_firmware_header_v2_0 *rlc_hdr;
 	unsigned int *tmp = NULL;
 	unsigned int i = 0;
+	uint16_t version_major;
+	uint16_t version_minor;
 
 	DRM_DEBUG("\n");
 
@@ -468,6 +490,12 @@ static int gfx_v9_0_init_microcode(struct amdgpu_device *adev)
 		goto out;
 	err = amdgpu_ucode_validate(adev->gfx.rlc_fw);
 	rlc_hdr = (const struct rlc_firmware_header_v2_0 *)adev->gfx.rlc_fw->data;
+
+	version_major = le16_to_cpu(rlc_hdr->header.header_version_major);
+	version_minor = le16_to_cpu(rlc_hdr->header.header_version_minor);
+	if (version_major == 2 && version_minor == 1)
+		adev->gfx.rlc.is_rlc_v2_1 = true;
+
 	adev->gfx.rlc_fw_version = le32_to_cpu(rlc_hdr->header.ucode_version);
 	adev->gfx.rlc_feature_version = le32_to_cpu(rlc_hdr->ucode_feature_version);
 	adev->gfx.rlc.save_and_restore_offset =
@@ -508,6 +536,9 @@ static int gfx_v9_0_init_microcode(struct amdgpu_device *adev)
 	for (i = 0 ; i < (rlc_hdr->reg_list_size_bytes >> 2); i++)
 		adev->gfx.rlc.register_restore[i] = le32_to_cpu(tmp[i]);
 
+	if (adev->gfx.rlc.is_rlc_v2_1)
+		gfx_v9_0_init_rlc_ext_microcode(adev);
+
 	snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_mec.bin", chip_name);
 	err = request_firmware(&adev->gfx.mec_fw, fw_name, adev->dev);
 	if (err)
@@ -566,6 +597,26 @@ static int gfx_v9_0_init_microcode(struct amdgpu_device *adev)
 		adev->firmware.fw_size +=
 			ALIGN(le32_to_cpu(header->ucode_size_bytes), PAGE_SIZE);
 
+		if (adev->gfx.rlc.is_rlc_v2_1) {
+			info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL];
+			info->ucode_id = AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL;
+			info->fw = adev->gfx.rlc_fw;
+			adev->firmware.fw_size +=
+				ALIGN(adev->gfx.rlc.save_restore_list_cntl_size_bytes, PAGE_SIZE);
+
+			info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM];
+			info->ucode_id = AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM;
+			info->fw = adev->gfx.rlc_fw;
+			adev->firmware.fw_size +=
+				ALIGN(adev->gfx.rlc.save_restore_list_gpm_size_bytes, PAGE_SIZE);
+
+			info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM];
+			info->ucode_id = AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM;
+			info->fw = adev->gfx.rlc_fw;
+			adev->firmware.fw_size +=
+				ALIGN(adev->gfx.rlc.save_restore_list_srm_size_bytes, PAGE_SIZE);
+		}
+
 		info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_MEC1];
 		info->ucode_id = AMDGPU_UCODE_ID_CP_MEC1;
 		info->fw = adev->gfx.mec_fw;
@@ -1781,7 +1832,7 @@ static int gfx_v9_0_init_rlc_save_restore_list(struct amdgpu_device *adev)
 
 	/* setup unique_indirect_regs array and indirect_start_offsets array */
 	gfx_v9_0_parse_ind_reg_list(register_list_format,
-				GFX9_RLC_FORMAT_DIRECT_REG_LIST_LENGTH,
+				adev->gfx.rlc.reg_list_format_direct_reg_list_length,
 				adev->gfx.rlc.reg_list_format_size_bytes >> 2,
 				unique_indirect_regs,
 				&unique_indirect_reg_count,
diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c
index 8873d83..0ff136d 100644
--- a/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c
@@ -70,6 +70,15 @@ psp_v10_0_get_fw_type(struct amdgpu_firmware_info *ucode, enum psp_gfx_fw_type *
 	case AMDGPU_UCODE_ID_RLC_G:
 		*type = GFX_FW_TYPE_RLC_G;
 		break;
+	case AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL:
+		*type = GFX_FW_TYPE_RLC_RESTORE_LIST_CNTL;
+		break;
+	case AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM:
+		*type = GFX_FW_TYPE_RLC_RESTORE_LIST_GPM_MEM;
+		break;
+	case AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM:
+		*type = GFX_FW_TYPE_RLC_RESTORE_LIST_SRM_MEM;
+		break;
 	case AMDGPU_UCODE_ID_SMC:
 		*type = GFX_FW_TYPE_SMU;
 		break;
diff --git a/include/uapi/drm/amdgpu_drm.h b/include/uapi/drm/amdgpu_drm.h
index 78fe828..081d256 100644
--- a/include/uapi/drm/amdgpu_drm.h
+++ b/include/uapi/drm/amdgpu_drm.h
@@ -630,6 +630,12 @@ struct drm_amdgpu_cs_chunk_data {
 	#define AMDGPU_INFO_FW_ASD		0x0d
 	/* Subquery id: Query VCN firmware version */
 	#define AMDGPU_INFO_FW_VCN		0x0e
+	/* Subquery id: Query GFX RLC SRLC firmware version */
+	#define AMDGPU_INFO_FW_GFX_RLC_RESTORE_LIST_CNTL 0x0f
+	/* Subquery id: Query GFX RLC SRLG firmware version */
+	#define AMDGPU_INFO_FW_GFX_RLC_RESTORE_LIST_GPM_MEM 0x10
+	/* Subquery id: Query GFX RLC SRLS firmware version */
+	#define AMDGPU_INFO_FW_GFX_RLC_RESTORE_LIST_SRM_MEM 0x11
 /* number of bytes moved for TTM migration */
 #define AMDGPU_INFO_NUM_BYTES_MOVED		0x0f
 /* the used VRAM size */
-- 
cgit v1.1


From 72408a41d0d78dbbd7fe7e24849c683596c8b79a Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Thu, 21 Dec 2017 15:03:31 +0800
Subject: drm/amdgpu: enter rlc safe mode before set cgpg
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Signed-off-by: Huang Rui <ray.huang@amd.com>
Acked-by: Hawking Zhang <Hawking.Zhang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index 73b76fa..69370f0 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -3394,8 +3394,7 @@ static void gfx_v9_0_exit_rlc_safe_mode(struct amdgpu_device *adev)
 static void gfx_v9_0_update_gfx_cg_power_gating(struct amdgpu_device *adev,
 						bool enable)
 {
-	/* TODO: double check if we need to perform under safe mdoe */
-	/* gfx_v9_0_enter_rlc_safe_mode(adev); */
+	gfx_v9_0_enter_rlc_safe_mode(adev);
 
 	if ((adev->pg_flags & AMD_PG_SUPPORT_GFX_PG) && enable) {
 		gfx_v9_0_enable_gfx_cg_power_gating(adev, true);
@@ -3406,7 +3405,7 @@ static void gfx_v9_0_update_gfx_cg_power_gating(struct amdgpu_device *adev,
 		gfx_v9_0_enable_gfx_pipeline_powergating(adev, false);
 	}
 
-	/* gfx_v9_0_exit_rlc_safe_mode(adev); */
+	gfx_v9_0_exit_rlc_safe_mode(adev);
 }
 
 static void gfx_v9_0_update_gfx_mg_power_gating(struct amdgpu_device *adev,
@@ -3797,7 +3796,7 @@ static void gfx_v9_0_ring_emit_ib_gfx(struct amdgpu_ring *ring,
 	}
 
 	amdgpu_ring_write(ring, header);
-BUG_ON(ib->gpu_addr & 0x3); /* Dword align */
+	BUG_ON(ib->gpu_addr & 0x3); /* Dword align */
 	amdgpu_ring_write(ring,
 #ifdef __BIG_ENDIAN
 		(2 << 0) |
-- 
cgit v1.1


From a5acf930269e71c76a7e7ad6819a86919c752fb1 Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Thu, 21 Dec 2017 15:48:27 +0800
Subject: drm/amdgpu: cleanup init power gating function
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Remove gfx_v9_0_enable_sck_slow_down_on_power_up/down and CP power gating
enabling functions because they only need to be called on setting power gating
behavior. We keep it in set_powergating callback to enable/disable PG in
late_init.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Acked-by: Hawking Zhang <Hawking.Zhang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 24 ++++++------------------
 1 file changed, 6 insertions(+), 18 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index 69370f0..eff1fd1 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -2065,6 +2065,9 @@ static void gfx_v9_0_enable_gfx_dynamic_mg_power_gating(struct amdgpu_device *ad
 
 static void gfx_v9_0_init_pg(struct amdgpu_device *adev)
 {
+	if (!adev->gfx.rlc.is_rlc_v2_1)
+		return;
+
 	if (adev->pg_flags & (AMD_PG_SUPPORT_GFX_PG |
 			      AMD_PG_SUPPORT_GFX_SMG |
 			      AMD_PG_SUPPORT_GFX_DMG |
@@ -2075,24 +2078,9 @@ static void gfx_v9_0_init_pg(struct amdgpu_device *adev)
 		gfx_v9_0_init_rlc_save_restore_list(adev);
 		gfx_v9_0_enable_save_restore_machine(adev);
 
-		if (adev->asic_type == CHIP_RAVEN) {
-			WREG32(mmRLC_JUMP_TABLE_RESTORE,
-				adev->gfx.rlc.cp_table_gpu_addr >> 8);
-			gfx_v9_0_init_gfx_power_gating(adev);
-
-			if (adev->pg_flags & AMD_PG_SUPPORT_RLC_SMU_HS) {
-				gfx_v9_0_enable_sck_slow_down_on_power_up(adev, true);
-				gfx_v9_0_enable_sck_slow_down_on_power_down(adev, true);
-			} else {
-				gfx_v9_0_enable_sck_slow_down_on_power_up(adev, false);
-				gfx_v9_0_enable_sck_slow_down_on_power_down(adev, false);
-			}
-
-			if (adev->pg_flags & AMD_PG_SUPPORT_CP)
-				gfx_v9_0_enable_cp_power_gating(adev, true);
-			else
-				gfx_v9_0_enable_cp_power_gating(adev, false);
-		}
+		WREG32(mmRLC_JUMP_TABLE_RESTORE,
+		       adev->gfx.rlc.cp_table_gpu_addr >> 8);
+		gfx_v9_0_init_gfx_power_gating(adev);
 	}
 }
 
-- 
cgit v1.1


From 727b888f6643b69db2cad1a9f0ae5f8804fa12cd Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Thu, 21 Dec 2017 16:13:02 +0800
Subject: drm/amdgpu: revise init_rlc_save_restore_list behavior to support
 latest register_list_format/register_restore table
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

RLC save/restore list will be used on CGPG and GFXOFF function, it loads two bin
table of register_list_format/register_restore in RLC firmware.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Acked-by: Hawking Zhang <Hawking.Zhang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 141 +++++++++++++++++++++-------------
 1 file changed, 87 insertions(+), 54 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index eff1fd1..3abd91f 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -184,6 +184,30 @@ static const struct soc15_reg_golden golden_settings_gc_9_2_1_vg12[] =
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTD_CNTL, 0x01bd9f33, 0x01000000)
 };
 
+static const u32 GFX_RLC_SRM_INDEX_CNTL_ADDR_OFFSETS[] =
+{
+	mmRLC_SRM_INDEX_CNTL_ADDR_0 - mmRLC_SRM_INDEX_CNTL_ADDR_0,
+	mmRLC_SRM_INDEX_CNTL_ADDR_1 - mmRLC_SRM_INDEX_CNTL_ADDR_0,
+	mmRLC_SRM_INDEX_CNTL_ADDR_2 - mmRLC_SRM_INDEX_CNTL_ADDR_0,
+	mmRLC_SRM_INDEX_CNTL_ADDR_3 - mmRLC_SRM_INDEX_CNTL_ADDR_0,
+	mmRLC_SRM_INDEX_CNTL_ADDR_4 - mmRLC_SRM_INDEX_CNTL_ADDR_0,
+	mmRLC_SRM_INDEX_CNTL_ADDR_5 - mmRLC_SRM_INDEX_CNTL_ADDR_0,
+	mmRLC_SRM_INDEX_CNTL_ADDR_6 - mmRLC_SRM_INDEX_CNTL_ADDR_0,
+	mmRLC_SRM_INDEX_CNTL_ADDR_7 - mmRLC_SRM_INDEX_CNTL_ADDR_0,
+};
+
+static const u32 GFX_RLC_SRM_INDEX_CNTL_DATA_OFFSETS[] =
+{
+	mmRLC_SRM_INDEX_CNTL_DATA_0 - mmRLC_SRM_INDEX_CNTL_DATA_0,
+	mmRLC_SRM_INDEX_CNTL_DATA_1 - mmRLC_SRM_INDEX_CNTL_DATA_0,
+	mmRLC_SRM_INDEX_CNTL_DATA_2 - mmRLC_SRM_INDEX_CNTL_DATA_0,
+	mmRLC_SRM_INDEX_CNTL_DATA_3 - mmRLC_SRM_INDEX_CNTL_DATA_0,
+	mmRLC_SRM_INDEX_CNTL_DATA_4 - mmRLC_SRM_INDEX_CNTL_DATA_0,
+	mmRLC_SRM_INDEX_CNTL_DATA_5 - mmRLC_SRM_INDEX_CNTL_DATA_0,
+	mmRLC_SRM_INDEX_CNTL_DATA_6 - mmRLC_SRM_INDEX_CNTL_DATA_0,
+	mmRLC_SRM_INDEX_CNTL_DATA_7 - mmRLC_SRM_INDEX_CNTL_DATA_0,
+};
+
 #define VEGA10_GB_ADDR_CONFIG_GOLDEN 0x2a114042
 #define VEGA12_GB_ADDR_CONFIG_GOLDEN 0x24104041
 #define RAVEN_GB_ADDR_CONFIG_GOLDEN 0x24000042
@@ -1763,55 +1787,42 @@ static void gfx_v9_0_init_csb(struct amdgpu_device *adev)
 			adev->gfx.rlc.clear_state_size);
 }
 
-static void gfx_v9_0_parse_ind_reg_list(int *register_list_format,
+static void gfx_v9_1_parse_ind_reg_list(int *register_list_format,
 				int indirect_offset,
 				int list_size,
 				int *unique_indirect_regs,
 				int *unique_indirect_reg_count,
-				int max_indirect_reg_count,
 				int *indirect_start_offsets,
-				int *indirect_start_offsets_count,
-				int max_indirect_start_offsets_count)
+				int *indirect_start_offsets_count)
 {
 	int idx;
-	bool new_entry = true;
 
 	for (; indirect_offset < list_size; indirect_offset++) {
+		indirect_start_offsets[*indirect_start_offsets_count] = indirect_offset;
+		*indirect_start_offsets_count = *indirect_start_offsets_count + 1;
 
-		if (new_entry) {
-			new_entry = false;
-			indirect_start_offsets[*indirect_start_offsets_count] = indirect_offset;
-			*indirect_start_offsets_count = *indirect_start_offsets_count + 1;
-			BUG_ON(*indirect_start_offsets_count >= max_indirect_start_offsets_count);
-		}
+		while (register_list_format[indirect_offset] != 0xFFFFFFFF) {
+			indirect_offset += 2;
 
-		if (register_list_format[indirect_offset] == 0xFFFFFFFF) {
-			new_entry = true;
-			continue;
-		}
+			/* look for the matching indice */
+			for (idx = 0; idx < *unique_indirect_reg_count; idx++) {
+				if (unique_indirect_regs[idx] ==
+					register_list_format[indirect_offset] ||
+					!unique_indirect_regs[idx])
+					break;
+			}
 
-		indirect_offset += 2;
+			BUG_ON(idx >= *unique_indirect_reg_count);
 
-		/* look for the matching indice */
-		for (idx = 0; idx < *unique_indirect_reg_count; idx++) {
-			if (unique_indirect_regs[idx] ==
-				register_list_format[indirect_offset])
-				break;
-		}
+			if (!unique_indirect_regs[idx])
+				unique_indirect_regs[idx] = register_list_format[indirect_offset];
 
-		if (idx >= *unique_indirect_reg_count) {
-			unique_indirect_regs[*unique_indirect_reg_count] =
-				register_list_format[indirect_offset];
-			idx = *unique_indirect_reg_count;
-			*unique_indirect_reg_count = *unique_indirect_reg_count + 1;
-			BUG_ON(*unique_indirect_reg_count >= max_indirect_reg_count);
+			indirect_offset++;
 		}
-
-		register_list_format[indirect_offset] = idx;
 	}
 }
 
-static int gfx_v9_0_init_rlc_save_restore_list(struct amdgpu_device *adev)
+static int gfx_v9_1_init_rlc_save_restore_list(struct amdgpu_device *adev)
 {
 	int unique_indirect_regs[] = {0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0};
 	int unique_indirect_reg_count = 0;
@@ -1820,7 +1831,7 @@ static int gfx_v9_0_init_rlc_save_restore_list(struct amdgpu_device *adev)
 	int indirect_start_offsets_count = 0;
 
 	int list_size = 0;
-	int i = 0;
+	int i = 0, j = 0;
 	u32 tmp = 0;
 
 	u32 *register_list_format =
@@ -1831,15 +1842,14 @@ static int gfx_v9_0_init_rlc_save_restore_list(struct amdgpu_device *adev)
 		adev->gfx.rlc.reg_list_format_size_bytes);
 
 	/* setup unique_indirect_regs array and indirect_start_offsets array */
-	gfx_v9_0_parse_ind_reg_list(register_list_format,
-				adev->gfx.rlc.reg_list_format_direct_reg_list_length,
-				adev->gfx.rlc.reg_list_format_size_bytes >> 2,
-				unique_indirect_regs,
-				&unique_indirect_reg_count,
-				ARRAY_SIZE(unique_indirect_regs),
-				indirect_start_offsets,
-				&indirect_start_offsets_count,
-				ARRAY_SIZE(indirect_start_offsets));
+	unique_indirect_reg_count = ARRAY_SIZE(unique_indirect_regs);
+	gfx_v9_1_parse_ind_reg_list(register_list_format,
+				    adev->gfx.rlc.reg_list_format_direct_reg_list_length,
+				    adev->gfx.rlc.reg_list_format_size_bytes >> 2,
+				    unique_indirect_regs,
+				    &unique_indirect_reg_count,
+				    indirect_start_offsets,
+				    &indirect_start_offsets_count);
 
 	/* enable auto inc in case it is disabled */
 	tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_SRM_CNTL));
@@ -1853,19 +1863,37 @@ static int gfx_v9_0_init_rlc_save_restore_list(struct amdgpu_device *adev)
 		WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_SRM_ARAM_DATA),
 			adev->gfx.rlc.register_restore[i]);
 
-	/* load direct register */
-	WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_SRM_ARAM_ADDR), 0);
-	for (i = 0; i < adev->gfx.rlc.reg_list_size_bytes >> 2; i++)
-		WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_SRM_ARAM_DATA),
-			adev->gfx.rlc.register_restore[i]);
-
 	/* load indirect register */
 	WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_GPM_SCRATCH_ADDR),
 		adev->gfx.rlc.reg_list_format_start);
-	for (i = 0; i < adev->gfx.rlc.reg_list_format_size_bytes >> 2; i++)
+
+	/* direct register portion */
+	for (i = 0; i < adev->gfx.rlc.reg_list_format_direct_reg_list_length; i++)
 		WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_GPM_SCRATCH_DATA),
 			register_list_format[i]);
 
+	/* indirect register portion */
+	while (i < (adev->gfx.rlc.reg_list_format_size_bytes >> 2)) {
+		if (register_list_format[i] == 0xFFFFFFFF) {
+			WREG32_SOC15(GC, 0, mmRLC_GPM_SCRATCH_DATA, register_list_format[i++]);
+			continue;
+		}
+
+		WREG32_SOC15(GC, 0, mmRLC_GPM_SCRATCH_DATA, register_list_format[i++]);
+		WREG32_SOC15(GC, 0, mmRLC_GPM_SCRATCH_DATA, register_list_format[i++]);
+
+		for (j = 0; j < unique_indirect_reg_count; j++) {
+			if (register_list_format[i] == unique_indirect_regs[j]) {
+				WREG32_SOC15(GC, 0, mmRLC_GPM_SCRATCH_DATA, j);
+				break;
+			}
+		}
+
+		BUG_ON(j >= unique_indirect_reg_count);
+
+		i++;
+	}
+
 	/* set save/restore list size */
 	list_size = adev->gfx.rlc.reg_list_size_bytes >> 2;
 	list_size = list_size >> 1;
@@ -1878,14 +1906,19 @@ static int gfx_v9_0_init_rlc_save_restore_list(struct amdgpu_device *adev)
 		adev->gfx.rlc.starting_offsets_start);
 	for (i = 0; i < ARRAY_SIZE(indirect_start_offsets); i++)
 		WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_GPM_SCRATCH_DATA),
-			indirect_start_offsets[i]);
+		       indirect_start_offsets[i]);
 
 	/* load unique indirect regs*/
 	for (i = 0; i < ARRAY_SIZE(unique_indirect_regs); i++) {
-		WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_SRM_INDEX_CNTL_ADDR_0) + i,
-			unique_indirect_regs[i] & 0x3FFFF);
-		WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_SRM_INDEX_CNTL_DATA_0) + i,
-			unique_indirect_regs[i] >> 20);
+		if (unique_indirect_regs[i] != 0) {
+			WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_SRM_INDEX_CNTL_ADDR_0)
+			       + GFX_RLC_SRM_INDEX_CNTL_ADDR_OFFSETS[i],
+			       unique_indirect_regs[i] & 0x3FFFF);
+
+			WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_SRM_INDEX_CNTL_DATA_0)
+			       + GFX_RLC_SRM_INDEX_CNTL_DATA_OFFSETS[i],
+			       unique_indirect_regs[i] >> 20);
+		}
 	}
 
 	kfree(register_list_format);
@@ -2075,7 +2108,7 @@ static void gfx_v9_0_init_pg(struct amdgpu_device *adev)
 			      AMD_PG_SUPPORT_GDS |
 			      AMD_PG_SUPPORT_RLC_SMU_HS)) {
 		gfx_v9_0_init_csb(adev);
-		gfx_v9_0_init_rlc_save_restore_list(adev);
+		gfx_v9_1_init_rlc_save_restore_list(adev);
 		gfx_v9_0_enable_save_restore_machine(adev);
 
 		WREG32(mmRLC_JUMP_TABLE_RESTORE,
-- 
cgit v1.1


From 0df3e67d343b6af7eb71f6353f93e4d0a5e952a7 Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Wed, 6 Dec 2017 09:23:50 +0800
Subject: drm/amdgpu: add setting powergating method for gfx9
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Signed-off-by: Huang Rui <ray.huang@amd.com>
Acked-by: Hawking Zhang <Hawking.Zhang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index 3abd91f..8d54207 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -3355,6 +3355,11 @@ static int gfx_v9_0_late_init(void *handle)
 	if (r)
 		return r;
 
+	r = amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_GFX,
+						   AMD_PG_STATE_GATE);
+	if (r)
+		return r;
+
 	return 0;
 }
 
-- 
cgit v1.1


From af15890df97d09e2faba2199b36f5e69bf129342 Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Thu, 14 Dec 2017 13:38:13 +0800
Subject: drm/amd/powerplay: send CGPG smc message if PG is enabled for raven
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Signed-off-by: Huang Rui <ray.huang@amd.com>
Acked-by: Hawking Zhang <Hawking.Zhang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c | 8 +++++++-
 drivers/gpu/drm/amd/powerplay/inc/rv_ppsmc.h      | 1 +
 2 files changed, 8 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
index 0f25226..f0727b4 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
@@ -206,12 +206,18 @@ static int smu10_set_power_state_tasks(struct pp_hwmgr *hwmgr, const void *input
 static int smu10_init_power_gate_state(struct pp_hwmgr *hwmgr)
 {
 	struct smu10_hwmgr *smu10_data = (struct smu10_hwmgr *)(hwmgr->backend);
+	struct amdgpu_device *adev = hwmgr->adev;
 
 	smu10_data->vcn_power_gated = true;
 	smu10_data->isp_tileA_power_gated = true;
 	smu10_data->isp_tileB_power_gated = true;
 
-	return 0;
+	if (adev->pg_flags & AMD_PG_SUPPORT_GFX_PG)
+		return smum_send_msg_to_smc_with_parameter(hwmgr,
+							   PPSMC_MSG_SetGfxCGPG,
+							   true);
+	else
+		return 0;
 }
 
 
diff --git a/drivers/gpu/drm/amd/powerplay/inc/rv_ppsmc.h b/drivers/gpu/drm/amd/powerplay/inc/rv_ppsmc.h
index 426bff2..5d07b6e 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/rv_ppsmc.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/rv_ppsmc.h
@@ -75,6 +75,7 @@
 #define PPSMC_MSG_GetMinGfxclkFrequency         0x2C
 #define PPSMC_MSG_GetMaxGfxclkFrequency         0x2D
 #define PPSMC_MSG_SoftReset                     0x2E
+#define PPSMC_MSG_SetGfxCGPG			0x2F
 #define PPSMC_MSG_SetSoftMaxGfxClk              0x30
 #define PPSMC_MSG_SetHardMinGfxClk              0x31
 #define PPSMC_MSG_SetSoftMaxSocclkByFreq        0x32
-- 
cgit v1.1


From fa7bd27d7352bfd57aed60a7e1b678bc1f475fc4 Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Tue, 13 Mar 2018 15:13:46 +0800
Subject: drm/amdgpu: move PP_FEATURE_MASK to amd_shared header
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

It will be used not only for powerplay but also on amdgpu part in future
patches. So move it into amd_shared header file.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h   |  2 --
 drivers/gpu/drm/amd/amdgpu/ci_dpm.c       |  2 +-
 drivers/gpu/drm/amd/amdgpu/kv_dpm.c       |  2 +-
 drivers/gpu/drm/amd/include/amd_shared.h  | 19 +++++++++++++++++++
 drivers/gpu/drm/amd/powerplay/inc/hwmgr.h | 19 -------------------
 5 files changed, 21 insertions(+), 23 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
index 354c6dc..dd6203a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
@@ -52,8 +52,6 @@ enum amdgpu_dpm_event_src {
 	AMDGPU_DPM_EVENT_SRC_DIGIAL_OR_EXTERNAL = 4
 };
 
-#define SCLK_DEEP_SLEEP_MASK 0x8
-
 struct amdgpu_ps {
 	u32 caps; /* vbios flags */
 	u32 class; /* vbios flags */
diff --git a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c
index be6b199..f48168f 100644
--- a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c
@@ -5903,7 +5903,7 @@ static int ci_dpm_init(struct amdgpu_device *adev)
 	pi->pcie_dpm_key_disabled = 0;
 	pi->thermal_sclk_dpm_enabled = 0;
 
-	if (amdgpu_pp_feature_mask & SCLK_DEEP_SLEEP_MASK)
+	if (amdgpu_pp_feature_mask & PP_SCLK_DEEP_SLEEP_MASK)
 		pi->caps_sclk_ds = true;
 	else
 		pi->caps_sclk_ds = false;
diff --git a/drivers/gpu/drm/amd/amdgpu/kv_dpm.c b/drivers/gpu/drm/amd/amdgpu/kv_dpm.c
index bc1720e..ef668a3 100644
--- a/drivers/gpu/drm/amd/amdgpu/kv_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/kv_dpm.c
@@ -2817,7 +2817,7 @@ static int kv_dpm_init(struct amdgpu_device *adev)
 		pi->caps_tcp_ramping = true;
 	}
 
-	if (amdgpu_pp_feature_mask & SCLK_DEEP_SLEEP_MASK)
+	if (amdgpu_pp_feature_mask & PP_SCLK_DEEP_SLEEP_MASK)
 		pi->caps_sclk_ds = true;
 	else
 		pi->caps_sclk_ds = false;
diff --git a/drivers/gpu/drm/amd/include/amd_shared.h b/drivers/gpu/drm/amd/include/amd_shared.h
index 9fa3aae..efeea9a 100644
--- a/drivers/gpu/drm/amd/include/amd_shared.h
+++ b/drivers/gpu/drm/amd/include/amd_shared.h
@@ -109,6 +109,25 @@ enum amd_powergating_state {
 #define AMD_PG_SUPPORT_GFX_PIPELINE		(1 << 12)
 #define AMD_PG_SUPPORT_MMHUB			(1 << 13)
 
+enum PP_FEATURE_MASK {
+	PP_SCLK_DPM_MASK = 0x1,
+	PP_MCLK_DPM_MASK = 0x2,
+	PP_PCIE_DPM_MASK = 0x4,
+	PP_SCLK_DEEP_SLEEP_MASK = 0x8,
+	PP_POWER_CONTAINMENT_MASK = 0x10,
+	PP_UVD_HANDSHAKE_MASK = 0x20,
+	PP_SMC_VOLTAGE_CONTROL_MASK = 0x40,
+	PP_VBI_TIME_SUPPORT_MASK = 0x80,
+	PP_ULV_MASK = 0x100,
+	PP_ENABLE_GFX_CG_THRU_SMU = 0x200,
+	PP_CLOCK_STRETCH_MASK = 0x400,
+	PP_OD_FUZZY_FAN_CONTROL_MASK = 0x800,
+	PP_SOCCLK_DPM_MASK = 0x1000,
+	PP_DCEFCLK_DPM_MASK = 0x2000,
+	PP_OVERDRIVE_MASK = 0x4000,
+	PP_ACG_MASK = 0x10000,
+};
+
 struct amd_ip_funcs {
 	/* Name of IP block */
 	char *name;
diff --git a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
index 2f203ec..0d2b3ce 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
@@ -66,25 +66,6 @@ struct vi_dpm_table {
 #define PCIE_PERF_REQ_GEN2         3
 #define PCIE_PERF_REQ_GEN3         4
 
-enum PP_FEATURE_MASK {
-	PP_SCLK_DPM_MASK = 0x1,
-	PP_MCLK_DPM_MASK = 0x2,
-	PP_PCIE_DPM_MASK = 0x4,
-	PP_SCLK_DEEP_SLEEP_MASK = 0x8,
-	PP_POWER_CONTAINMENT_MASK = 0x10,
-	PP_UVD_HANDSHAKE_MASK = 0x20,
-	PP_SMC_VOLTAGE_CONTROL_MASK = 0x40,
-	PP_VBI_TIME_SUPPORT_MASK = 0x80,
-	PP_ULV_MASK = 0x100,
-	PP_ENABLE_GFX_CG_THRU_SMU = 0x200,
-	PP_CLOCK_STRETCH_MASK = 0x400,
-	PP_OD_FUZZY_FAN_CONTROL_MASK = 0x800,
-	PP_SOCCLK_DPM_MASK = 0x1000,
-	PP_DCEFCLK_DPM_MASK = 0x2000,
-	PP_OVERDRIVE_MASK = 0x4000,
-	PP_ACG_MASK = 0x10000,
-};
-
 enum PHM_BackEnd_Magic {
 	PHM_Dummy_Magic       = 0xAA5555AA,
 	PHM_RV770_Magic       = 0xDCBAABCD,
-- 
cgit v1.1


From 6f92ad2a1772ebaa5eb3d27c9c8dd8caf2e3cbdb Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Fri, 2 Mar 2018 14:16:06 +0800
Subject: drm/amdgpu: add gfxoff feature mask
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/include/amd_shared.h | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/include/amd_shared.h b/drivers/gpu/drm/amd/include/amd_shared.h
index efeea9a..33de330 100644
--- a/drivers/gpu/drm/amd/include/amd_shared.h
+++ b/drivers/gpu/drm/amd/include/amd_shared.h
@@ -125,6 +125,7 @@ enum PP_FEATURE_MASK {
 	PP_SOCCLK_DPM_MASK = 0x1000,
 	PP_DCEFCLK_DPM_MASK = 0x2000,
 	PP_OVERDRIVE_MASK = 0x4000,
+	PP_GFXOFF_MASK = 0x8000,
 	PP_ACG_MASK = 0x10000,
 };
 
-- 
cgit v1.1


From 917d8614c4cdddfb257229e0fb3077b8842dd9e0 Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Fri, 2 Mar 2018 14:40:53 +0800
Subject: drm/amdgpu: set gfxoff disabled by default
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
index 7c17a0b..998ba8e 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
@@ -122,7 +122,7 @@ uint amdgpu_pg_mask = 0xffffffff;
 uint amdgpu_sdma_phase_quantum = 32;
 char *amdgpu_disable_cu = NULL;
 char *amdgpu_virtual_display = NULL;
-uint amdgpu_pp_feature_mask = 0xffffbfff;
+uint amdgpu_pp_feature_mask = 0xffff3fff; /* gfxoff (bit 15) disabled by default */
 int amdgpu_ngg = 0;
 int amdgpu_prim_buf_per_se = 0;
 int amdgpu_pos_buf_per_se = 0;
-- 
cgit v1.1


From 9c82214160ee5e2a1e4137612822a35dc0cc064b Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Fri, 2 Mar 2018 15:10:52 +0800
Subject: drm/amd/powerplay: add gfx off control function
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

gfx_off_control is used to be called for sending enabling/disabling gfxoff
message.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c | 36 ++++++++++++++++++++++-
 drivers/gpu/drm/amd/powerplay/inc/hwmgr.h         |  1 +
 2 files changed, 36 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
index f0727b4..fde1e5c 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
@@ -42,6 +42,13 @@
 #define SMU10_DISPCLK_BYPASS_THRESHOLD     10000 /* 100Mhz */
 #define SMC_RAM_END                     0x40000
 
+#define mmPWR_MISC_CNTL_STATUS					0x0183
+#define mmPWR_MISC_CNTL_STATUS_BASE_IDX				0
+#define PWR_MISC_CNTL_STATUS__PWR_GFX_RLC_CGPG_EN__SHIFT	0x0
+#define PWR_MISC_CNTL_STATUS__PWR_GFXOFF_STATUS__SHIFT		0x1
+#define PWR_MISC_CNTL_STATUS__PWR_GFX_RLC_CGPG_EN_MASK		0x00000001L
+#define PWR_MISC_CNTL_STATUS__PWR_GFXOFF_STATUS_MASK		0x00000006L
+
 static const unsigned long SMU10_Magic = (unsigned long) PHM_Rv_Magic;
 
 
@@ -243,13 +250,31 @@ static int smu10_power_off_asic(struct pp_hwmgr *hwmgr)
 	return smu10_reset_cc6_data(hwmgr);
 }
 
+static bool smu10_is_gfx_on(struct pp_hwmgr *hwmgr)
+{
+	uint32_t reg;
+	struct amdgpu_device *adev = hwmgr->adev;
+
+	reg = RREG32_SOC15(PWR, 0, mmPWR_MISC_CNTL_STATUS);
+	if ((reg & PWR_MISC_CNTL_STATUS__PWR_GFXOFF_STATUS_MASK) ==
+	    (0x2 << PWR_MISC_CNTL_STATUS__PWR_GFXOFF_STATUS__SHIFT))
+		return true;
+
+	return false;
+}
+
 static int smu10_disable_gfx_off(struct pp_hwmgr *hwmgr)
 {
 	struct smu10_hwmgr *smu10_data = (struct smu10_hwmgr *)(hwmgr->backend);
 
-	if (smu10_data->gfx_off_controled_by_driver)
+	if (smu10_data->gfx_off_controled_by_driver) {
 		smum_send_msg_to_smc(hwmgr, PPSMC_MSG_DisableGfxOff);
 
+		/* confirm gfx is back to "on" state */
+		while (!smu10_is_gfx_on(hwmgr))
+			msleep(1);
+	}
+
 	return 0;
 }
 
@@ -273,6 +298,14 @@ static int smu10_enable_dpm_tasks(struct pp_hwmgr *hwmgr)
 	return smu10_enable_gfx_off(hwmgr);
 }
 
+static int smu10_gfx_off_control(struct pp_hwmgr *hwmgr, bool enable)
+{
+	if (enable)
+		return smu10_enable_gfx_off(hwmgr);
+	else
+		return smu10_disable_gfx_off(hwmgr);
+}
+
 static int smu10_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 				struct pp_power_state  *prequest_ps,
 			const struct pp_power_state *pcurrent_ps)
@@ -1060,6 +1093,7 @@ static const struct pp_hwmgr_func smu10_hwmgr_funcs = {
 	.power_state_set = smu10_set_power_state_tasks,
 	.dynamic_state_management_disable = smu10_disable_dpm_tasks,
 	.set_mmhub_powergating_by_smu = smu10_set_mmhub_powergating_by_smu,
+	.gfx_off_control = smu10_gfx_off_control,
 };
 
 int smu10_init_function_pointers(struct pp_hwmgr *hwmgr)
diff --git a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
index 0d2b3ce..3d9743f 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
@@ -296,6 +296,7 @@ struct pp_hwmgr_func {
 	int (*display_clock_voltage_request)(struct pp_hwmgr *hwmgr,
 			struct pp_display_clock_request *clock);
 	int (*get_max_high_clocks)(struct pp_hwmgr *hwmgr, struct amd_pp_simple_clock_info *clocks);
+	int (*gfx_off_control)(struct pp_hwmgr *hwmgr, bool enable);
 	int (*power_off_asic)(struct pp_hwmgr *hwmgr);
 	int (*force_clock_level)(struct pp_hwmgr *hwmgr, enum pp_clock_type type, uint32_t mask);
 	int (*print_clock_levels)(struct pp_hwmgr *hwmgr, enum pp_clock_type type, char *buf);
-- 
cgit v1.1


From 775b0c11e27fce7d204d2911220fd7eebcc074d0 Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Fri, 2 Mar 2018 15:18:54 +0800
Subject: drm/amd/powerplay: enable/disable gfxoff through smu
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/amd_powerplay.c | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
index bd0d387..6976596 100644
--- a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
+++ b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
@@ -222,10 +222,19 @@ static int pp_set_powergating_state(void *handle,
 {
 	struct amdgpu_device *adev = handle;
 	struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle;
+	int ret;
 
 	if (!hwmgr || !hwmgr->pm_en)
 		return 0;
 
+	if (hwmgr->hwmgr_func->gfx_off_control) {
+		/* Enable/disable GFX off through SMU */
+		ret = hwmgr->hwmgr_func->gfx_off_control(hwmgr,
+							 state == AMD_PG_STATE_GATE);
+		if (ret)
+			pr_err("gfx off control failed!\n");
+	}
+
 	if (hwmgr->hwmgr_func->enable_per_cu_power_gating == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
 		return 0;
-- 
cgit v1.1


From 00f54b97d7de97c41cffaad83d32a9bf03edad89 Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Tue, 27 Feb 2018 21:53:00 +0800
Subject: drm/amdgpu: use pp_feature member to store the mask
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu.h           | 1 +
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c    | 2 ++
 drivers/gpu/drm/amd/amdgpu/ci_dpm.c           | 2 +-
 drivers/gpu/drm/amd/amdgpu/kv_dpm.c           | 2 +-
 drivers/gpu/drm/amd/powerplay/amd_powerplay.c | 2 +-
 5 files changed, 6 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index 5ad8939..7570055 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -1430,6 +1430,7 @@ enum amd_hw_ip_block_type {
 struct amd_powerplay {
 	void *pp_handle;
 	const struct amd_pm_funcs *pp_funcs;
+	uint32_t pp_feature;
 };
 
 #define AMDGPU_RESET_MAGIC_NUM 64
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 5958e81..e8b57cf 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -1545,6 +1545,8 @@ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev)
 			return -EAGAIN;
 	}
 
+	adev->powerplay.pp_feature = amdgpu_pp_feature_mask;
+
 	for (i = 0; i < adev->num_ip_blocks; i++) {
 		if ((amdgpu_ip_block_mask & (1 << i)) == 0) {
 			DRM_ERROR("disabled ip block: %d <%s>\n",
diff --git a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c
index f48168f..a266dcf 100644
--- a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c
@@ -5903,7 +5903,7 @@ static int ci_dpm_init(struct amdgpu_device *adev)
 	pi->pcie_dpm_key_disabled = 0;
 	pi->thermal_sclk_dpm_enabled = 0;
 
-	if (amdgpu_pp_feature_mask & PP_SCLK_DEEP_SLEEP_MASK)
+	if (adev->powerplay.pp_feature & PP_SCLK_DEEP_SLEEP_MASK)
 		pi->caps_sclk_ds = true;
 	else
 		pi->caps_sclk_ds = false;
diff --git a/drivers/gpu/drm/amd/amdgpu/kv_dpm.c b/drivers/gpu/drm/amd/amdgpu/kv_dpm.c
index ef668a3..17f7f07 100644
--- a/drivers/gpu/drm/amd/amdgpu/kv_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/kv_dpm.c
@@ -2817,7 +2817,7 @@ static int kv_dpm_init(struct amdgpu_device *adev)
 		pi->caps_tcp_ramping = true;
 	}
 
-	if (amdgpu_pp_feature_mask & PP_SCLK_DEEP_SLEEP_MASK)
+	if (adev->powerplay.pp_feature & PP_SCLK_DEEP_SLEEP_MASK)
 		pi->caps_sclk_ds = true;
 	else
 		pi->caps_sclk_ds = false;
diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
index 6976596..246f8e9 100644
--- a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
+++ b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
@@ -53,7 +53,7 @@ static int amd_powerplay_create(struct amdgpu_device *adev)
 	mutex_init(&hwmgr->smu_lock);
 	hwmgr->chip_family = adev->family;
 	hwmgr->chip_id = adev->asic_type;
-	hwmgr->feature_mask = amdgpu_pp_feature_mask;
+	hwmgr->feature_mask = adev->powerplay.pp_feature;
 	hwmgr->display_config = &adev->pm.pm_display_cfg;
 	adev->powerplay.pp_handle = hwmgr;
 	adev->powerplay.pp_funcs = &pp_dpm_funcs;
-- 
cgit v1.1


From 1dedc62338accff01ce4d56302e1c55a6b43b3d6 Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Tue, 13 Mar 2018 17:59:12 +0800
Subject: drm/amdgpu: clear gfxoff feature mask if the asic is not raven
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
index bca67df..d1052b5 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
@@ -95,7 +95,8 @@ int hwmgr_early_init(struct pp_hwmgr *hwmgr)
 		hwmgr->smumgr_funcs = &ci_smu_funcs;
 		ci_set_asic_special_caps(hwmgr);
 		hwmgr->feature_mask &= ~(PP_VBI_TIME_SUPPORT_MASK |
-					PP_ENABLE_GFX_CG_THRU_SMU);
+					 PP_ENABLE_GFX_CG_THRU_SMU |
+					 PP_GFXOFF_MASK);
 		hwmgr->pp_table_version = PP_TABLE_V0;
 		hwmgr->od_enabled = false;
 		smu7_init_function_pointers(hwmgr);
@@ -103,9 +104,11 @@ int hwmgr_early_init(struct pp_hwmgr *hwmgr)
 	case AMDGPU_FAMILY_CZ:
 		hwmgr->od_enabled = false;
 		hwmgr->smumgr_funcs = &smu8_smu_funcs;
+		hwmgr->feature_mask &= ~PP_GFXOFF_MASK;
 		smu8_init_function_pointers(hwmgr);
 		break;
 	case AMDGPU_FAMILY_VI:
+		hwmgr->feature_mask &= ~PP_GFXOFF_MASK;
 		switch (hwmgr->chip_id) {
 		case CHIP_TOPAZ:
 			hwmgr->smumgr_funcs = &iceland_smu_funcs;
@@ -139,6 +142,7 @@ int hwmgr_early_init(struct pp_hwmgr *hwmgr)
 		smu7_init_function_pointers(hwmgr);
 		break;
 	case AMDGPU_FAMILY_AI:
+		hwmgr->feature_mask &= ~PP_GFXOFF_MASK;
 		switch (hwmgr->chip_id) {
 		case CHIP_VEGA10:
 			hwmgr->smumgr_funcs = &vega10_smu_funcs;
-- 
cgit v1.1


From 9667849bbb8d8a2b97798ba0972fe25d13ea8acf Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Tue, 13 Mar 2018 18:32:39 +0800
Subject: drm/amd/powerplay: add control gfxoff enabling in late init
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/amd_powerplay.c | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
index 246f8e9..b493369 100644
--- a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
+++ b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
@@ -180,6 +180,7 @@ static int pp_late_init(void *handle)
 {
 	struct amdgpu_device *adev = handle;
 	struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle;
+	int ret;
 
 	if (hwmgr && hwmgr->pm_en) {
 		mutex_lock(&hwmgr->smu_lock);
@@ -189,6 +190,14 @@ static int pp_late_init(void *handle)
 	}
 	if (adev->pm.smu_prv_buffer_size != 0)
 		pp_reserve_vram_for_smu(adev);
+
+	if (hwmgr->hwmgr_func->gfx_off_control &&
+	    (hwmgr->feature_mask & PP_GFXOFF_MASK)) {
+		ret = hwmgr->hwmgr_func->gfx_off_control(hwmgr, true);
+		if (ret)
+			pr_err("gfx off enabling failed!\n");
+	}
+
 	return 0;
 }
 
-- 
cgit v1.1


From b083369621e84dc0c8ec1ae7191d009f6f1c4d75 Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Tue, 13 Mar 2018 18:39:48 +0800
Subject: drm/amdgpu: it should disable gfxoff when system is going to suspend
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index e8b57cf..9e917f5 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -1902,6 +1902,12 @@ int amdgpu_device_ip_suspend(struct amdgpu_device *adev)
 	if (amdgpu_sriov_vf(adev))
 		amdgpu_virt_request_full_gpu(adev, false);
 
+	/* ungate SMC block powergating */
+	if (adev->powerplay.pp_feature & PP_GFXOFF_MASK)
+		amdgpu_device_ip_set_powergating_state(adev,
+						       AMD_IP_BLOCK_TYPE_SMC,
+						       AMD_CG_STATE_UNGATE);
+
 	/* ungate SMC block first */
 	r = amdgpu_device_ip_set_clockgating_state(adev, AMD_IP_BLOCK_TYPE_SMC,
 						   AMD_CG_STATE_UNGATE);
-- 
cgit v1.1


From 151b5d7fd35876120dc744f93865e4c7dc2c1f36 Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Tue, 27 Feb 2018 13:43:59 +0800
Subject: drm/amdgpu: fix to disable powergating in hw_fini
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

We need enable CGPG and GFXOFF together. If only enable one of them, this system
will get hang after startx (do draw command). So when gfxoff is disabled, it
also need disable CGPG after that.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index 8d54207..2c5e2a4 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -3137,6 +3137,9 @@ static int gfx_v9_0_hw_fini(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 	int i;
 
+	amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_GFX,
+					       AMD_PG_STATE_UNGATE);
+
 	amdgpu_irq_put(adev, &adev->gfx.priv_reg_irq, 0);
 	amdgpu_irq_put(adev, &adev->gfx.priv_inst_irq, 0);
 
-- 
cgit v1.1


From 9ac4b0d95a7a554bb60d97fbee5fbfd1b73df50a Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Fri, 15 Dec 2017 14:34:57 +0800
Subject: drm/amdgpu: set CGPG if gfxoff is enabled for raven
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/soc15.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index 65e781f..9006576 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -682,6 +682,11 @@ static int soc15_common_early_init(void *handle)
 			AMD_CG_SUPPORT_SDMA_LS;
 		adev->pg_flags = AMD_PG_SUPPORT_SDMA;
 
+		if (adev->powerplay.pp_feature & PP_GFXOFF_MASK)
+			adev->pg_flags |= AMD_PG_SUPPORT_GFX_PG |
+				AMD_PG_SUPPORT_CP |
+				AMD_PG_SUPPORT_RLC_SMU_HS;
+
 		adev->external_rev_id = 0x1;
 		break;
 	default:
-- 
cgit v1.1


From f5264548008a5cde7090c2b6b85c8d65cb86d2f7 Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Thu, 14 Dec 2017 15:33:53 +0800
Subject: drm/amd/powerplay: use the flag to decide whether send gfxoff smc
 message
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
index fde1e5c..7712eb6 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
@@ -81,11 +81,15 @@ static int smu10_initialize_dpm_defaults(struct pp_hwmgr *hwmgr)
 	smu10_data->thermal_auto_throttling_treshold = 0;
 	smu10_data->is_nb_dpm_enabled = 1;
 	smu10_data->dpm_flags = 1;
-	smu10_data->gfx_off_controled_by_driver = false;
 	smu10_data->need_min_deep_sleep_dcefclk = true;
 	smu10_data->num_active_display = 0;
 	smu10_data->deep_sleep_dcefclk = 0;
 
+	if (hwmgr->feature_mask & PP_GFXOFF_MASK)
+		smu10_data->gfx_off_controled_by_driver = true;
+	else
+		smu10_data->gfx_off_controled_by_driver = false;
+
 	phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
 					PHM_PlatformCaps_SclkDeepSleep);
 
-- 
cgit v1.1


From 1d2361e5a6c60d7b142d19555c3e6240ffe93731 Mon Sep 17 00:00:00 2001
From: Samuel Li <Samuel.Li@amd.com>
Date: Wed, 18 Apr 2018 15:06:02 -0400
Subject: drm/amdgpu: Rename amdgpu_display_framebuffer_domains()
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

It returns supported domains for display, and domains actually used are to be
decided later when pinned.

Signed-off-by: Samuel Li <Samuel.Li@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_display.c       | 4 ++--
 drivers/gpu/drm/amd/amdgpu/amdgpu_display.h       | 2 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c            | 2 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c         | 2 +-
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 3 +--
 5 files changed, 6 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c
index b83ae99..76ee8e0 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c
@@ -189,7 +189,7 @@ int amdgpu_display_crtc_page_flip_target(struct drm_crtc *crtc,
 		goto cleanup;
 	}
 
-	r = amdgpu_bo_pin(new_abo, amdgpu_display_framebuffer_domains(adev), &base);
+	r = amdgpu_bo_pin(new_abo, amdgpu_display_supported_domains(adev), &base);
 	if (unlikely(r != 0)) {
 		DRM_ERROR("failed to pin new abo buffer before flip\n");
 		goto unreserve;
@@ -484,7 +484,7 @@ static const struct drm_framebuffer_funcs amdgpu_fb_funcs = {
 	.create_handle = drm_gem_fb_create_handle,
 };
 
-uint32_t amdgpu_display_framebuffer_domains(struct amdgpu_device *adev)
+uint32_t amdgpu_display_supported_domains(struct amdgpu_device *adev)
 {
 	uint32_t domain = AMDGPU_GEM_DOMAIN_VRAM;
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.h
index 2b11d80..f66e3e3 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.h
@@ -23,7 +23,7 @@
 #ifndef __AMDGPU_DISPLAY_H__
 #define __AMDGPU_DISPLAY_H__
 
-uint32_t amdgpu_display_framebuffer_domains(struct amdgpu_device *adev);
+uint32_t amdgpu_display_supported_domains(struct amdgpu_device *adev);
 struct drm_framebuffer *
 amdgpu_display_user_framebuffer_create(struct drm_device *dev,
 				       struct drm_file *file_priv,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c
index ff89e84..bc5fd8e 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c
@@ -137,7 +137,7 @@ static int amdgpufb_create_pinned_object(struct amdgpu_fbdev *rfbdev,
 	/* need to align pitch with crtc limits */
 	mode_cmd->pitches[0] = amdgpu_align_pitch(adev, mode_cmd->width, cpp,
 						  fb_tiled);
-	domain = amdgpu_display_framebuffer_domains(adev);
+	domain = amdgpu_display_supported_domains(adev);
 
 	height = ALIGN(mode_cmd->height, 8);
 	size = mode_cmd->pitches[0] * height;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c
index 713417b..4683626 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c
@@ -215,7 +215,7 @@ static int amdgpu_gem_begin_cpu_access(struct dma_buf *dma_buf,
 	struct amdgpu_bo *bo = gem_to_amdgpu_bo(dma_buf->priv);
 	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
 	struct ttm_operation_ctx ctx = { true, false };
-	u32 domain = amdgpu_display_framebuffer_domains(adev);
+	u32 domain = amdgpu_display_supported_domains(adev);
 	int ret;
 	bool reads = (direction == DMA_BIDIRECTIONAL ||
 		      direction == DMA_FROM_DEVICE);
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 2368ade..28d8c08 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -3049,12 +3049,11 @@ static int dm_plane_helper_prepare_fb(struct drm_plane *plane,
 		return r;
 
 	if (plane->type != DRM_PLANE_TYPE_CURSOR)
-		domain = amdgpu_display_framebuffer_domains(adev);
+		domain = amdgpu_display_supported_domains(adev);
 	else
 		domain = AMDGPU_GEM_DOMAIN_VRAM;
 
 	r = amdgpu_bo_pin(rbo, domain, &afb->address);
-
 	amdgpu_bo_unreserve(rbo);
 
 	if (unlikely(r != 0)) {
-- 
cgit v1.1


From 9b3f217faf48603c91d4ca44a18e6ff74c3c1c0c Mon Sep 17 00:00:00 2001
From: Samuel Li <Samuel.Li@amd.com>
Date: Wed, 18 Apr 2018 16:26:18 -0400
Subject: drm/amdgpu: Remove VRAM from shared bo domains.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This fixes an issue introduced by change "allow framebuffer in GART
memory as well" which could lead to a shared buffer ending up
pinned in vram.  Use GTT if it is included in the domain, otherwise
return an error.

Signed-off-by: Samuel Li <Samuel.Li@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
index feece0a..1985c08 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
@@ -694,8 +694,12 @@ int amdgpu_bo_pin_restricted(struct amdgpu_bo *bo, u32 domain,
 		return -EINVAL;
 
 	/* A shared bo cannot be migrated to VRAM */
-	if (bo->prime_shared_count && (domain == AMDGPU_GEM_DOMAIN_VRAM))
-		return -EINVAL;
+	if (bo->prime_shared_count) {
+		if (domain & AMDGPU_GEM_DOMAIN_GTT)
+			domain = AMDGPU_GEM_DOMAIN_GTT;
+		else
+			return -EINVAL;
+	}
 
 	if (bo->pin_count) {
 		uint32_t mem_type = bo->tbo.mem.mem_type;
-- 
cgit v1.1


From 8567f68147de1f09cc868b52b02a0c11dc048206 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Thu, 19 Apr 2018 13:46:03 -0500
Subject: drm/amdgpu/pm: document power_dpm_force_performance_level

Provide documentation for power_dpm_force_performance_level
which is used to adjust things related to GPU power states.

Acked-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c | 53 ++++++++++++++++++++++++++++++++++
 1 file changed, 53 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
index 744f105..ee11e92 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
@@ -131,6 +131,59 @@ fail:
 	return count;
 }
 
+
+/**
+ * DOC: power_dpm_force_performance_level
+ *
+ * The amdgpu driver provides a sysfs API for adjusting certain power
+ * related parameters.  The file power_dpm_force_performance_level is
+ * used for this.  It accepts the following arguments:
+ * - auto
+ * - low
+ * - high
+ * - manual
+ * - GPU fan
+ * - profile_standard
+ * - profile_min_sclk
+ * - profile_min_mclk
+ * - profile_peak
+ *
+ * auto
+ *
+ * When auto is selected, the driver will attempt to dynamically select
+ * the optimal power profile for current conditions in the driver.
+ *
+ * low
+ *
+ * When low is selected, the clocks are forced to the lowest power state.
+ *
+ * high
+ *
+ * When high is selected, the clocks are forced to the highest power state.
+ *
+ * manual
+ *
+ * When manual is selected, the user can manually adjust which power states
+ * are enabled for each clock domain via the sysfs pp_dpm_mclk, pp_dpm_sclk,
+ * and pp_dpm_pcie files and adjust the power state transition heuristics
+ * via the pp_power_profile_mode sysfs file.
+ *
+ * profile_standard
+ * profile_min_sclk
+ * profile_min_mclk
+ * profile_peak
+ *
+ * When the profiling modes are selected, clock and power gating are
+ * disabled and the clocks are set for different profiling cases. This
+ * mode is recommended for profiling specific work loads where you do
+ * not want clock or power gating for clock fluctuation to interfere
+ * with your results. profile_standard sets the clocks to a fixed clock
+ * level which varies from asic to asic.  profile_min_sclk forces the sclk
+ * to the lowest level.  profile_min_mclk forces the mclk to the lowest level.
+ * profile_peak sets all clocks (mclk, sclk, pcie) to the highest levels.
+ *
+ */
+
 static ssize_t amdgpu_get_dpm_forced_performance_level(struct device *dev,
 						struct device_attribute *attr,
 								char *buf)
-- 
cgit v1.1


From ca8d40ca194dfb573e59a5e42b88da83e63a6630 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Thu, 19 Apr 2018 13:56:41 -0500
Subject: drm/amdgpu/pm: document power_dpm_state

This is a legacy file and is only provided for
backwards compatibility.

Acked-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c | 31 +++++++++++++++++++++++++++++++
 1 file changed, 31 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
index ee11e92..e33e0f4 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
@@ -77,6 +77,37 @@ void amdgpu_pm_acpi_event_handler(struct amdgpu_device *adev)
 	}
 }
 
+/**
+ * DOC: power_dpm_state
+ *
+ * This is a legacy interface and is only provided for backwards compatibility.
+ * The amdgpu driver provides a sysfs API for adjusting certain power
+ * related parameters.  The file power_dpm_state is used for this.
+ * It accepts the following arguments:
+ * - battery
+ * - balanced
+ * - performance
+ *
+ * battery
+ *
+ * On older GPUs, the vbios provided a special power state for battery
+ * operation.  Selecting battery switched to this state.  This is no
+ * longer provided on newer GPUs so the option does nothing in that case.
+ *
+ * balanced
+ *
+ * On older GPUs, the vbios provided a special power state for balanced
+ * operation.  Selecting balanced switched to this state.  This is no
+ * longer provided on newer GPUs so the option does nothing in that case.
+ *
+ * performance
+ *
+ * On older GPUs, the vbios provided a special power state for performance
+ * operation.  Selecting performance switched to this state.  This is no
+ * longer provided on newer GPUs so the option does nothing in that case.
+ *
+ */
+
 static ssize_t amdgpu_get_dpm_state(struct device *dev,
 				    struct device_attribute *attr,
 				    char *buf)
-- 
cgit v1.1


From d54bb40f607d40fca60da0613c65005086653300 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Thu, 19 Apr 2018 14:02:52 -0500
Subject: drm/amdgpu/pm: document pp_table

This file is for uploading new powerplay tables.

Acked-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c | 11 +++++++++++
 1 file changed, 11 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
index e33e0f4..9982f1b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
@@ -408,6 +408,17 @@ fail:
 	return count;
 }
 
+/**
+ * DOC: pp_table
+ *
+ * The amdgpu driver provides a sysfs API for uploading new powerplay
+ * tables.  The file pp_table is used for this.  Reading the file
+ * will dump the current power play table.  Writing to the file
+ * will attempt to upload a new powerplay table and re-initialize
+ * powerplay using that new table.
+ *
+ */
+
 static ssize_t amdgpu_get_pp_table(struct device *dev,
 		struct device_attribute *attr,
 		char *buf)
-- 
cgit v1.1


From 271dc908732b72bb9b1ad22b7cd14e75df3612c5 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Thu, 19 Apr 2018 14:22:24 -0500
Subject: drm/amdgpu/pm: document pp_dpm_sclk pp_dpm_mclk pp_dpm_pcie (v2)

Used for manually masking dpm states.

v2: drop comment about current state (Rex)

Acked-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
index 9982f1b..07f2e96 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
@@ -539,6 +539,23 @@ static ssize_t amdgpu_get_pp_od_clk_voltage(struct device *dev,
 
 }
 
+/**
+ * DOC: pp_dpm_sclk pp_dpm_mclk pp_dpm_pcie
+ *
+ * The amdgpu driver provides a sysfs API for adjusting what power levels
+ * are enabled for a given power state.  The files pp_dpm_sclk, pp_dpm_mclk,
+ * and pp_dpm_pcie are used for this.
+ *
+ * Reading back the files will show you the available power levels within
+ * the power state and the clock information for those levels.
+ *
+ * To manually adjust these states, first select manual using
+ * power_dpm_force_performance_level.  Writing a string of the level
+ * numbers to the file will select which levels you want to enable.
+ * E.g., writing 456 to the file will enable levels 4, 5, and 6.
+ *
+ */
+
 static ssize_t amdgpu_get_pp_dpm_sclk(struct device *dev,
 		struct device_attribute *attr,
 		char *buf)
-- 
cgit v1.1


From 6b2576f5bddae4c89f29481f387735ac99e256d5 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Thu, 19 Apr 2018 14:38:31 -0500
Subject: drm/amdgpu/pm: document pp_power_profile_mode

sysfs file for adjusting power level heuristics.

Acked-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c | 20 ++++++++++++++++++++
 1 file changed, 20 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
index 07f2e96..d6e6641 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
@@ -780,6 +780,26 @@ fail:
 	return count;
 }
 
+/**
+ * DOC: pp_power_profile_mode
+ *
+ * The amdgpu driver provides a sysfs API for adjusting the heuristics
+ * related to switching between power levels in a power state.  The file
+ * pp_power_profile_mode is used for this.
+ *
+ * Reading this file outputs a list of all of the predefined power profiles
+ * and the relevant heuristics settings for that profile.
+ *
+ * To select a profile or create a custom profile, first select manual using
+ * power_dpm_force_performance_level.  Writing the number of a predefined
+ * profile to pp_power_profile_mode will enable those heuristics.  To
+ * create a custom set of heuristics, write a string of numbers to the file
+ * starting with the number of the custom profile along with a setting
+ * for each heuristic parameter.  Due to differences across asic families
+ * the heuristic parameters vary from family to family.
+ *
+ */
+
 static ssize_t amdgpu_get_pp_power_profile_mode(struct device *dev,
 		struct device_attribute *attr,
 		char *buf)
-- 
cgit v1.1


From 4e418c3401867cccc3ba67973d1e03510da7c92d Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Thu, 19 Apr 2018 14:59:55 -0500
Subject: drm/amdgpu/pm: document pp_od_clk_voltage

sysfs interface for fine grained clock and voltage control.

Acked-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c | 23 +++++++++++++++++++++++
 1 file changed, 23 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
index d6e6641..ce8be46 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
@@ -455,6 +455,29 @@ static ssize_t amdgpu_set_pp_table(struct device *dev,
 	return count;
 }
 
+/**
+ * DOC: pp_od_clk_voltage
+ *
+ * The amdgpu driver provides a sysfs API for adjusting the clocks and voltages
+ * in each power level within a power state.  The pp_od_clk_voltage is used for
+ * this.
+ *
+ * Reading the file will display:
+ * - a list of engine clock levels and voltages labeled OD_SCLK
+ * - a list of memory clock levels and voltages labeled OD_MCLK
+ * - a list of valid ranges for sclk, mclk, and voltage labeled OD_RANGE
+ *
+ * To manually adjust these settings, first select manual using
+ * power_dpm_force_performance_level. Enter a new value for each
+ * level by writing a string that contains "s/m level clock voltage" to
+ * the file.  E.g., "s 1 500 820" will update sclk level 1 to be 500 MHz
+ * at 820 mV; "m 0 350 810" will update mclk level 0 to be 350 MHz at
+ * 810 mV.  When you have edited all of the states as needed, write
+ * "c" (commit) to the file to commit your changes.  If you want to reset to the
+ * default power levels, write "r" (reset) to the file to reset them.
+ *
+ */
+
 static ssize_t amdgpu_set_pp_od_clk_voltage(struct device *dev,
 		struct device_attribute *attr,
 		const char *buf,
-- 
cgit v1.1


From d10fb4a6f382474025f326bf90ee3b64396486ea Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Fri, 20 Apr 2018 12:57:10 +0800
Subject: drm/amd/pp: Change pstate_clk frequency unit to 10KHz on Rv

to keep consistent with other asics

Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
index 7712eb6..ef09073 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
@@ -479,8 +479,8 @@ static int smu10_hwmgr_backend_init(struct pp_hwmgr *hwmgr)
 
 	hwmgr->platform_descriptor.minimumClocksReductionPercentage = 50;
 
-	hwmgr->pstate_sclk = SMU10_UMD_PSTATE_GFXCLK;
-	hwmgr->pstate_mclk = SMU10_UMD_PSTATE_FCLK;
+	hwmgr->pstate_sclk = SMU10_UMD_PSTATE_GFXCLK * 100;
+	hwmgr->pstate_mclk = SMU10_UMD_PSTATE_FCLK * 100;
 
 	return result;
 }
-- 
cgit v1.1


From 21c77de35661152e118908a081b8a51e7bca7bb4 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Fri, 20 Apr 2018 13:03:15 +0800
Subject: drm/amd/pp: Use dynamic gfx_clk rather than hardcoded values

Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c | 47 ++++++++++++++---------
 drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.h |  2 -
 2 files changed, 29 insertions(+), 20 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
index ef09073..be6d6e2 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
@@ -383,7 +383,7 @@ static int smu10_get_clock_voltage_dependency_table(struct pp_hwmgr *hwmgr,
 
 static int smu10_populate_clock_table(struct pp_hwmgr *hwmgr)
 {
-	int result;
+	uint32_t result;
 
 	struct smu10_hwmgr *smu10_data = (struct smu10_hwmgr *)(hwmgr->backend);
 	DpmClocks_t  *table = &(smu10_data->clock_table);
@@ -429,11 +429,11 @@ static int smu10_populate_clock_table(struct pp_hwmgr *hwmgr)
 
 	smum_send_msg_to_smc(hwmgr, PPSMC_MSG_GetMinGfxclkFrequency);
 	result = smum_get_argument(hwmgr);
-	smu10_data->gfx_min_freq_limit = result * 100;
+	smu10_data->gfx_min_freq_limit = result / 10 * 1000;
 
 	smum_send_msg_to_smc(hwmgr, PPSMC_MSG_GetMaxGfxclkFrequency);
 	result = smum_get_argument(hwmgr);
-	smu10_data->gfx_max_freq_limit = result * 100;
+	smu10_data->gfx_max_freq_limit = result / 10 * 1000;
 
 	return 0;
 }
@@ -515,6 +515,8 @@ static int smu10_hwmgr_backend_fini(struct pp_hwmgr *hwmgr)
 static int smu10_dpm_force_dpm_level(struct pp_hwmgr *hwmgr,
 				enum amd_dpm_forced_level level)
 {
+	struct smu10_hwmgr *data = hwmgr->backend;
+
 	if (hwmgr->smu_version < 0x1E3700) {
 		pr_info("smu firmware version too old, can not set dpm level\n");
 		return 0;
@@ -525,7 +527,7 @@ static int smu10_dpm_force_dpm_level(struct pp_hwmgr *hwmgr,
 	case AMD_DPM_FORCED_LEVEL_PROFILE_PEAK:
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetHardMinGfxClk,
-						SMU10_UMD_PSTATE_PEAK_GFXCLK);
+						data->gfx_max_freq_limit/100);
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetHardMinFclkByFreq,
 						SMU10_UMD_PSTATE_PEAK_FCLK);
@@ -538,7 +540,7 @@ static int smu10_dpm_force_dpm_level(struct pp_hwmgr *hwmgr,
 
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetSoftMaxGfxClk,
-						SMU10_UMD_PSTATE_PEAK_GFXCLK);
+						data->gfx_max_freq_limit/100);
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetSoftMaxFclkByFreq,
 						SMU10_UMD_PSTATE_PEAK_FCLK);
@@ -552,10 +554,10 @@ static int smu10_dpm_force_dpm_level(struct pp_hwmgr *hwmgr,
 	case AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK:
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetHardMinGfxClk,
-						SMU10_UMD_PSTATE_MIN_GFXCLK);
+						data->gfx_min_freq_limit/100);
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetSoftMaxGfxClk,
-						SMU10_UMD_PSTATE_MIN_GFXCLK);
+						data->gfx_min_freq_limit/100);
 		break;
 	case AMD_DPM_FORCED_LEVEL_PROFILE_MIN_MCLK:
 		smum_send_msg_to_smc_with_parameter(hwmgr,
@@ -595,7 +597,7 @@ static int smu10_dpm_force_dpm_level(struct pp_hwmgr *hwmgr,
 	case AMD_DPM_FORCED_LEVEL_AUTO:
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetHardMinGfxClk,
-						SMU10_UMD_PSTATE_MIN_GFXCLK);
+						data->gfx_min_freq_limit/100);
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetHardMinFclkByFreq,
 						SMU10_UMD_PSTATE_MIN_FCLK);
@@ -608,7 +610,7 @@ static int smu10_dpm_force_dpm_level(struct pp_hwmgr *hwmgr,
 
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetSoftMaxGfxClk,
-						SMU10_UMD_PSTATE_PEAK_GFXCLK);
+						data->gfx_max_freq_limit/100);
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetSoftMaxFclkByFreq,
 						SMU10_UMD_PSTATE_PEAK_FCLK);
@@ -622,10 +624,10 @@ static int smu10_dpm_force_dpm_level(struct pp_hwmgr *hwmgr,
 	case AMD_DPM_FORCED_LEVEL_LOW:
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetHardMinGfxClk,
-						SMU10_UMD_PSTATE_MIN_GFXCLK);
+						data->gfx_min_freq_limit/100);
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetSoftMaxGfxClk,
-						SMU10_UMD_PSTATE_MIN_GFXCLK);
+						data->gfx_min_freq_limit/100);
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetHardMinFclkByFreq,
 						SMU10_UMD_PSTATE_MIN_FCLK);
@@ -773,21 +775,30 @@ static int smu10_print_clock_levels(struct pp_hwmgr *hwmgr,
 	struct smu10_hwmgr *data = (struct smu10_hwmgr *)(hwmgr->backend);
 	struct smu10_voltage_dependency_table *mclk_table =
 			data->clock_vol_info.vdd_dep_on_fclk;
-	int i, now, size = 0;
+	uint32_t i, now, size = 0;
 
 	switch (type) {
 	case PP_SCLK:
 		smum_send_msg_to_smc(hwmgr, PPSMC_MSG_GetGfxclkFrequency);
 		now = smum_get_argument(hwmgr);
 
+	/* driver only know min/max gfx_clk, Add level 1 for all other gfx clks */
+		if (now == data->gfx_max_freq_limit/100)
+			i = 2;
+		else if (now == data->gfx_min_freq_limit/100)
+			i = 0;
+		else
+			i = 1;
+
 		size += sprintf(buf + size, "0: %uMhz %s\n",
-				data->gfx_min_freq_limit / 100,
-				((data->gfx_min_freq_limit / 100)
-				 == now) ? "*" : "");
+					data->gfx_min_freq_limit/100,
+					i == 0 ? "*" : "");
 		size += sprintf(buf + size, "1: %uMhz %s\n",
-				data->gfx_max_freq_limit / 100,
-				((data->gfx_max_freq_limit / 100)
-				 == now) ? "*" : "");
+					i == 1 ? now : SMU10_UMD_PSTATE_GFXCLK,
+					i == 1 ? "*" : "");
+		size += sprintf(buf + size, "2: %uMhz %s\n",
+					data->gfx_max_freq_limit/100,
+					i == 2 ? "*" : "");
 		break;
 	case PP_MCLK:
 		smum_send_msg_to_smc(hwmgr, PPSMC_MSG_GetFclkFrequency);
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.h
index f68b218..1fb296a 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.h
@@ -311,11 +311,9 @@ int smu10_init_function_pointers(struct pp_hwmgr *hwmgr);
 #define SMU10_UMD_PSTATE_FCLK                   933
 #define SMU10_UMD_PSTATE_VCE                    0x03C00320
 
-#define SMU10_UMD_PSTATE_PEAK_GFXCLK            1100
 #define SMU10_UMD_PSTATE_PEAK_SOCCLK            757
 #define SMU10_UMD_PSTATE_PEAK_FCLK              1200
 
-#define SMU10_UMD_PSTATE_MIN_GFXCLK             200
 #define SMU10_UMD_PSTATE_MIN_FCLK               400
 #define SMU10_UMD_PSTATE_MIN_SOCCLK             200
 #define SMU10_UMD_PSTATE_MIN_VCE                0x0190012C
-- 
cgit v1.1


From ca6e0c5bdc44a2cd7152002191a8107fc566084f Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Wed, 18 Apr 2018 18:43:19 +0800
Subject: drm/amd/pp: Refine the OD state checking code in smu7

if vddc restore to default value, driver clear the
bit of DPMTABLE_OD_UPDATE_VDDC and need to repopulate sclk
and mclk table.

1. Remove variable i checking code.
2. move clear DPMTABLE_OD_UPDATE_VDDC bit to the end of the
   function to avoid sclk table will not be updated.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c | 8 ++------
 1 file changed, 2 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index 720ac47..9654593 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -4683,10 +4683,6 @@ static void smu7_check_dpm_table_updated(struct pp_hwmgr *hwmgr)
 			return;
 		}
 	}
-	if (i == dep_table->count && data->need_update_smu7_dpm_table & DPMTABLE_OD_UPDATE_VDDC) {
-		data->need_update_smu7_dpm_table &= ~DPMTABLE_OD_UPDATE_VDDC;
-		data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_MCLK;
-	}
 
 	dep_table = table_info->vdd_dep_on_sclk;
 	odn_dep_table = (struct phm_ppt_v1_clock_voltage_dependency_table *)&(odn_table->vdd_dependency_on_sclk);
@@ -4696,9 +4692,9 @@ static void smu7_check_dpm_table_updated(struct pp_hwmgr *hwmgr)
 			return;
 		}
 	}
-	if (i == dep_table->count && data->need_update_smu7_dpm_table & DPMTABLE_OD_UPDATE_VDDC) {
+	if (data->need_update_smu7_dpm_table & DPMTABLE_OD_UPDATE_VDDC) {
 		data->need_update_smu7_dpm_table &= ~DPMTABLE_OD_UPDATE_VDDC;
-		data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_SCLK;
+		data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_SCLK | DPMTABLE_OD_UPDATE_MCLK;
 	}
 }
 
-- 
cgit v1.1


From d389d607e60809726fe818113c80f5fc3aac4675 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Wed, 18 Apr 2018 21:09:35 +0800
Subject: drm/amd/pp: Change voltage/clk range for OD feature on VI

read vddc range from vbios.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c | 28 +++++++++++
 drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h |  3 ++
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c | 60 ++++++++++++++++--------
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h |  2 +
 4 files changed, 73 insertions(+), 20 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c
index 971fb5d..d58be7e 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c
@@ -1505,3 +1505,31 @@ int atomctrl_get_leakage_vddc_base_on_leakage(struct pp_hwmgr *hwmgr,
 
 	return 0;
 }
+
+void atomctrl_get_voltage_range(struct pp_hwmgr *hwmgr, uint32_t *max_vddc,
+							uint32_t *min_vddc)
+{
+	void *profile;
+
+	profile = smu_atom_get_data_table(hwmgr->adev,
+					GetIndexIntoMasterTable(DATA, ASIC_ProfilingInfo),
+					NULL, NULL, NULL);
+
+	if (profile) {
+		switch (hwmgr->chip_id) {
+		case CHIP_TONGA:
+		case CHIP_FIJI:
+			*max_vddc = le32_to_cpu(((ATOM_ASIC_PROFILING_INFO_V3_3 *)profile)->ulMaxVddc/4);
+			*min_vddc = le32_to_cpu(((ATOM_ASIC_PROFILING_INFO_V3_3 *)profile)->ulMinVddc/4);
+			break;
+		case CHIP_POLARIS11:
+		case CHIP_POLARIS10:
+		case CHIP_POLARIS12:
+			*max_vddc = le32_to_cpu(((ATOM_ASIC_PROFILING_INFO_V3_6 *)profile)->ulMaxVddc/100);
+			*min_vddc = le32_to_cpu(((ATOM_ASIC_PROFILING_INFO_V3_6 *)profile)->ulMinVddc/100);
+			break;
+		default:
+			return;
+		}
+	}
+}
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h
index c672a50..e1b5d6b 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h
@@ -320,5 +320,8 @@ extern int atomctrl_get_leakage_vddc_base_on_leakage(struct pp_hwmgr *hwmgr,
 					uint16_t virtual_voltage_id,
 					uint16_t efuse_voltage_id);
 extern int atomctrl_get_leakage_id_from_efuse(struct pp_hwmgr *hwmgr, uint16_t *virtual_voltage_id);
+
+extern void atomctrl_get_voltage_range(struct pp_hwmgr *hwmgr, uint32_t *max_vddc,
+							uint32_t *min_vddc);
 #endif
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index 9654593..e119637 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -838,6 +838,33 @@ static int smu7_odn_initial_default_setting(struct pp_hwmgr *hwmgr)
 	return 0;
 }
 
+static void smu7_setup_voltage_range_from_vbios(struct pp_hwmgr *hwmgr)
+{
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct phm_ppt_v1_clock_voltage_dependency_table *dep_sclk_table;
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	uint32_t min_vddc, max_vddc;
+
+	if (!table_info)
+		return;
+
+	dep_sclk_table = table_info->vdd_dep_on_sclk;
+
+	atomctrl_get_voltage_range(hwmgr, &max_vddc, &min_vddc);
+
+	if (min_vddc == 0 || min_vddc > 2000
+		|| min_vddc > dep_sclk_table->entries[0].vddc)
+		min_vddc = dep_sclk_table->entries[0].vddc;
+
+	if (max_vddc == 0 || max_vddc > 2000
+		|| max_vddc < dep_sclk_table->entries[dep_sclk_table->count-1].vddc)
+		max_vddc = dep_sclk_table->entries[dep_sclk_table->count-1].vddc;
+
+	data->odn_dpm_table.min_vddc = min_vddc;
+	data->odn_dpm_table.max_vddc = max_vddc;
+}
+
 static int smu7_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
 {
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
@@ -856,8 +883,10 @@ static int smu7_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
 			sizeof(struct smu7_dpm_table));
 
 	/* initialize ODN table */
-	if (hwmgr->od_enabled)
+	if (hwmgr->od_enabled) {
+		smu7_setup_voltage_range_from_vbios(hwmgr);
 		smu7_odn_initial_default_setting(hwmgr);
+	}
 
 	return 0;
 }
@@ -4605,36 +4634,27 @@ static bool smu7_check_clk_voltage_valid(struct pp_hwmgr *hwmgr,
 {
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
 
-	struct phm_ppt_v1_information *table_info =
-			(struct phm_ppt_v1_information *)(hwmgr->pptable);
-	uint32_t min_vddc;
-	struct phm_ppt_v1_clock_voltage_dependency_table *dep_sclk_table;
-
-	if (table_info == NULL)
-		return false;
-
-	dep_sclk_table = table_info->vdd_dep_on_sclk;
-	min_vddc = dep_sclk_table->entries[0].vddc;
-
-	if (voltage < min_vddc || voltage > 2000) {
-		pr_info("OD voltage is out of range [%d - 2000] mV\n", min_vddc);
+	if (voltage < data->odn_dpm_table.min_vddc || voltage > data->odn_dpm_table.max_vddc) {
+		pr_info("OD voltage is out of range [%d - %d] mV\n",
+						data->odn_dpm_table.min_vddc,
+						data->odn_dpm_table.max_vddc);
 		return false;
 	}
 
 	if (type == PP_OD_EDIT_SCLK_VDDC_TABLE) {
-		if (data->vbios_boot_state.sclk_bootup_value > clk ||
+		if (data->golden_dpm_table.sclk_table.dpm_levels[0].value > clk ||
 			hwmgr->platform_descriptor.overdriveLimit.engineClock < clk) {
 			pr_info("OD engine clock is out of range [%d - %d] MHz\n",
-				data->vbios_boot_state.sclk_bootup_value,
-				hwmgr->platform_descriptor.overdriveLimit.engineClock / 100);
+				data->golden_dpm_table.sclk_table.dpm_levels[0].value/100,
+				hwmgr->platform_descriptor.overdriveLimit.engineClock/100);
 			return false;
 		}
 	} else if (type == PP_OD_EDIT_MCLK_VDDC_TABLE) {
-		if (data->vbios_boot_state.mclk_bootup_value > clk ||
+		if (data->golden_dpm_table.mclk_table.dpm_levels[0].value > clk ||
 			hwmgr->platform_descriptor.overdriveLimit.memoryClock < clk) {
 			pr_info("OD memory clock is out of range [%d - %d] MHz\n",
-				data->vbios_boot_state.mclk_bootup_value/100,
-				hwmgr->platform_descriptor.overdriveLimit.memoryClock / 100);
+				data->golden_dpm_table.mclk_table.dpm_levels[0].value/100,
+				hwmgr->platform_descriptor.overdriveLimit.memoryClock/100);
 			return false;
 		}
 	} else {
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h
index f40179c..51a776e 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h
@@ -184,6 +184,8 @@ struct smu7_odn_dpm_table {
 	struct smu7_odn_clock_voltage_dependency_table	vdd_dependency_on_sclk;
 	struct smu7_odn_clock_voltage_dependency_table	vdd_dependency_on_mclk;
 	uint32_t					odn_mclk_min_limit;
+	uint32_t min_vddc;
+	uint32_t max_vddc;
 };
 
 struct profile_mode_setting {
-- 
cgit v1.1


From a3c991f922f99160cb695f9d28e04cd8e818d6f9 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Thu, 19 Apr 2018 10:39:17 +0800
Subject: drm/amd/pp: Print out voltage/clock range in sysfs

when user cat pp_od_clk_voltage
add display info about the sclk/mclk/vddc range that user can overdrive
output as:
OD_SCLK:
0:        300MHz        900mV
1:        400MHz        912mV
2:        500MHz        925mV
3:        600MHz        937mV
4:        700MHz        950mV
5:        800MHz        975mV
6:        900MHz        987mV
7:       1000MHz       1000mV
OD_MCLK:
0:        300MHz        900mV
1:       1500MHz        912mV
OD_RANGE:
SCLK:     300MHz       1200MHz
MCLK:     300MHz       1500MHz
VDDC:     700mV        1200mV

also
1. remove unnecessary whitespace before a quoted newline
2. change unit of frequency Mhz to MHz

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c           |  1 +
 drivers/gpu/drm/amd/include/kgd_pp_interface.h   |  1 +
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c | 26 ++++++++++++++++++------
 3 files changed, 22 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
index ce8be46..27d8dd7 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
@@ -555,6 +555,7 @@ static ssize_t amdgpu_get_pp_od_clk_voltage(struct device *dev,
 	if (adev->powerplay.pp_funcs->print_clock_levels) {
 		size = amdgpu_dpm_print_clock_levels(adev, OD_SCLK, buf);
 		size += amdgpu_dpm_print_clock_levels(adev, OD_MCLK, buf+size);
+		size += amdgpu_dpm_print_clock_levels(adev, OD_RANGE, buf+size);
 		return size;
 	} else {
 		return snprintf(buf, PAGE_SIZE, "\n");
diff --git a/drivers/gpu/drm/amd/include/kgd_pp_interface.h b/drivers/gpu/drm/amd/include/kgd_pp_interface.h
index 01969b1..06f08f3 100644
--- a/drivers/gpu/drm/amd/include/kgd_pp_interface.h
+++ b/drivers/gpu/drm/amd/include/kgd_pp_interface.h
@@ -94,6 +94,7 @@ enum pp_clock_type {
 	PP_PCIE,
 	OD_SCLK,
 	OD_MCLK,
+	OD_RANGE,
 };
 
 enum amd_pp_sensors {
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index e119637..232ea6f 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -4335,22 +4335,36 @@ static int smu7_print_clock_levels(struct pp_hwmgr *hwmgr,
 		break;
 	case OD_SCLK:
 		if (hwmgr->od_enabled) {
-			size = sprintf(buf, "%s: \n", "OD_SCLK");
+			size = sprintf(buf, "%s:\n", "OD_SCLK");
 			for (i = 0; i < odn_sclk_table->num_of_pl; i++)
-				size += sprintf(buf + size, "%d: %10uMhz %10u mV\n",
-					i, odn_sclk_table->entries[i].clock / 100,
+				size += sprintf(buf + size, "%d: %10uMHz %10umV\n",
+					i, odn_sclk_table->entries[i].clock/100,
 					odn_sclk_table->entries[i].vddc);
 		}
 		break;
 	case OD_MCLK:
 		if (hwmgr->od_enabled) {
-			size = sprintf(buf, "%s: \n", "OD_MCLK");
+			size = sprintf(buf, "%s:\n", "OD_MCLK");
 			for (i = 0; i < odn_mclk_table->num_of_pl; i++)
-				size += sprintf(buf + size, "%d: %10uMhz %10u mV\n",
-					i, odn_mclk_table->entries[i].clock / 100,
+				size += sprintf(buf + size, "%d: %10uMHz %10umV\n",
+					i, odn_mclk_table->entries[i].clock/100,
 					odn_mclk_table->entries[i].vddc);
 		}
 		break;
+	case OD_RANGE:
+		if (hwmgr->od_enabled) {
+			size = sprintf(buf, "%s:\n", "OD_RANGE");
+			size += sprintf(buf + size, "SCLK: %7uMHz %10uMHz\n",
+				data->golden_dpm_table.sclk_table.dpm_levels[0].value/100,
+				hwmgr->platform_descriptor.overdriveLimit.engineClock/100);
+			size += sprintf(buf + size, "MCLK: %7uMHz %10uMHz\n",
+				data->golden_dpm_table.mclk_table.dpm_levels[0].value/100,
+				hwmgr->platform_descriptor.overdriveLimit.memoryClock/100);
+			size += sprintf(buf + size, "VDDC: %7umV %11umV\n",
+				data->odn_dpm_table.min_vddc,
+				data->odn_dpm_table.max_vddc);
+		}
+		break;
 	default:
 		break;
 	}
-- 
cgit v1.1


From 9e70b539292652d31091568f89e73b54e3a4f79d Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Sat, 21 Apr 2018 14:09:59 -0500
Subject: drm/amdgpu/powerplay: actually return the power with the new query
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Set query to the power value so we actually return it.  Fixes
no power value returned on asics with the new query.

Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index 232ea6f..c9dd0be 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -3369,6 +3369,7 @@ static int smu7_get_gpu_power(struct pp_hwmgr *hwmgr, u32 *query)
 
 	smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_GetCurrPkgPwr, 0);
 	tmp = cgs_read_register(hwmgr->device, mmSMC_MSG_ARG_0);
+	*query = tmp;
 
 	if (tmp != 0)
 		return 0;
-- 
cgit v1.1


From 6c8d74caa2fa33908ecd07fb1cf1b7bc629b367a Mon Sep 17 00:00:00 2001
From: Samuel Li <Samuel.Li@amd.com>
Date: Wed, 18 Apr 2018 16:15:52 -0400
Subject: drm/amdgpu: Enable scatter gather display support

Enables sg display if vram size <= THRESHOLD(256M); otherwise
still use vram as display buffer.
This patch fixed some potention issues introduced by change
"allow framebuffer in GART memory as well" due to CZ/ST hardware
limitation.

v2: Change default setting to auto.
v3: Move some logic from amdgpu_display_framebuffer_domains()
    to pin function, suggested by Christian.
v4: Split into several patches.
v5: Drop module parameter for now.

Signed-off-by: Samuel Li <Samuel.Li@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu.h        | 1 +
 drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 9 +++++++++
 2 files changed, 10 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index 7570055..03a2c0be 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -138,6 +138,7 @@ extern int amdgpu_si_support;
 extern int amdgpu_cik_support;
 #endif
 
+#define AMDGPU_SG_THRESHOLD			(256*1024*1024)
 #define AMDGPU_DEFAULT_GTT_SIZE_MB		3072ULL /* 3GB by default */
 #define AMDGPU_WAIT_IDLE_TIMEOUT_IN_MS	        3000
 #define AMDGPU_MAX_USEC_TIMEOUT			100000	/* 100 ms */
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
index 1985c08..e62153a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
@@ -701,6 +701,15 @@ int amdgpu_bo_pin_restricted(struct amdgpu_bo *bo, u32 domain,
 			return -EINVAL;
 	}
 
+	/* This assumes only APU display buffers are pinned with (VRAM|GTT).
+	 * See function amdgpu_display_supported_domains()
+	 */
+	if (domain == (AMDGPU_GEM_DOMAIN_VRAM | AMDGPU_GEM_DOMAIN_GTT)) {
+		domain = AMDGPU_GEM_DOMAIN_VRAM;
+		if (adev->gmc.real_vram_size <= AMDGPU_SG_THRESHOLD)
+			domain = AMDGPU_GEM_DOMAIN_GTT;
+	}
+
 	if (bo->pin_count) {
 		uint32_t mem_type = bo->tbo.mem.mem_type;
 
-- 
cgit v1.1


From 8239f57ac3e9bf9ad0cf4d396ebfa721e91ac611 Mon Sep 17 00:00:00 2001
From: Junwei Zhang <Jerry.Zhang@amd.com>
Date: Mon, 23 Apr 2018 17:21:21 +0800
Subject: drm/amdgpu: bo could be null when access in vm bo update

Signed-off-by: Junwei Zhang <Jerry.Zhang@amd.com>
Reviewed-by: David Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index 6a372ca..1c00f1a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -1509,7 +1509,6 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev,
 	struct drm_mm_node *nodes;
 	struct dma_fence *exclusive, **last_update;
 	uint64_t flags;
-	uint32_t mem_type;
 	int r;
 
 	if (clear || !bo_va->base.bo) {
@@ -1568,9 +1567,9 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev,
 	 * the evicted list so that it gets validated again on the
 	 * next command submission.
 	 */
-	mem_type = bo->tbo.mem.mem_type;
 	if (bo && bo->tbo.resv == vm->root.base.bo->tbo.resv &&
-	    !(bo->preferred_domains & amdgpu_mem_type_to_domain(mem_type)))
+	    !(bo->preferred_domains &
+	    amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type)))
 		list_add_tail(&bo_va->base.vm_status, &vm->evicted);
 	spin_unlock(&vm->status_lock);
 
-- 
cgit v1.1


From 38610f15a7ad7a914e4fd0a9a5a6c386700b8ba0 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Nicolai=20H=C3=A4hnle?= <nicolai.haehnle@amd.com>
Date: Thu, 12 Apr 2018 16:34:19 +0200
Subject: drm/amdgpu: set COMPUTE_PGM_RSRC1 for SGPR/VGPR clearing shaders
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Otherwise, the SQ may skip some of the register writes, or shader waves may
be allocated where we don't expect them, so that as a result we don't actually
reset all of the register SRAMs. This can lead to spurious ECC errors later on
if a shader uses an uninitialized register.

Signed-off-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
---
 drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
index b0e591e..e14263f 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
@@ -1459,10 +1459,11 @@ static const u32 sgpr_init_compute_shader[] =
 static const u32 vgpr_init_regs[] =
 {
 	mmCOMPUTE_STATIC_THREAD_MGMT_SE0, 0xffffffff,
-	mmCOMPUTE_RESOURCE_LIMITS, 0,
+	mmCOMPUTE_RESOURCE_LIMITS, 0x1000000, /* CU_GROUP_COUNT=1 */
 	mmCOMPUTE_NUM_THREAD_X, 256*4,
 	mmCOMPUTE_NUM_THREAD_Y, 1,
 	mmCOMPUTE_NUM_THREAD_Z, 1,
+	mmCOMPUTE_PGM_RSRC1, 0x100004f, /* VGPRS=15 (64 logical VGPRs), SGPRS=1 (16 SGPRs), BULKY=1 */
 	mmCOMPUTE_PGM_RSRC2, 20,
 	mmCOMPUTE_USER_DATA_0, 0xedcedc00,
 	mmCOMPUTE_USER_DATA_1, 0xedcedc01,
@@ -1479,10 +1480,11 @@ static const u32 vgpr_init_regs[] =
 static const u32 sgpr1_init_regs[] =
 {
 	mmCOMPUTE_STATIC_THREAD_MGMT_SE0, 0x0f,
-	mmCOMPUTE_RESOURCE_LIMITS, 0x1000000,
+	mmCOMPUTE_RESOURCE_LIMITS, 0x1000000, /* CU_GROUP_COUNT=1 */
 	mmCOMPUTE_NUM_THREAD_X, 256*5,
 	mmCOMPUTE_NUM_THREAD_Y, 1,
 	mmCOMPUTE_NUM_THREAD_Z, 1,
+	mmCOMPUTE_PGM_RSRC1, 0x240, /* SGPRS=9 (80 GPRS) */
 	mmCOMPUTE_PGM_RSRC2, 20,
 	mmCOMPUTE_USER_DATA_0, 0xedcedc00,
 	mmCOMPUTE_USER_DATA_1, 0xedcedc01,
@@ -1503,6 +1505,7 @@ static const u32 sgpr2_init_regs[] =
 	mmCOMPUTE_NUM_THREAD_X, 256*5,
 	mmCOMPUTE_NUM_THREAD_Y, 1,
 	mmCOMPUTE_NUM_THREAD_Z, 1,
+	mmCOMPUTE_PGM_RSRC1, 0x240, /* SGPRS=9 (80 GPRS) */
 	mmCOMPUTE_PGM_RSRC2, 20,
 	mmCOMPUTE_USER_DATA_0, 0xedcedc00,
 	mmCOMPUTE_USER_DATA_1, 0xedcedc01,
-- 
cgit v1.1


From 48edde3959e2a538ff963e6dbdc9c9adca8b159b Mon Sep 17 00:00:00 2001
From: welu <wei.lu2@amd.com>
Date: Tue, 24 Apr 2018 09:13:20 -0400
Subject: drm/amdgpu: change pp_dpm clk/mclk/pcie input format.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

1. support more than 8 values when setting get_pp_dpm_mclk/
sclk/pcie, the former design just parse command format like
"echo xxxx >　pp_dpm_sclk" and current can parse "echo xx xxx
 xxxx > pp_dpm_sclk" whose operation is more user-friendly
and convinent and can offer more values;
2. be compatible with former design like "xx".
3. add DOC: pp_dpm_sclk pp_dpm_mclk pp_dpm_pcie
Bug:KFD-385

Signed-off-by: welu <wei.lu2@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c | 103 +++++++++++++++++++--------------
 1 file changed, 59 insertions(+), 44 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
index 27d8dd7..d9802d9 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
@@ -574,10 +574,10 @@ static ssize_t amdgpu_get_pp_od_clk_voltage(struct device *dev,
  * the power state and the clock information for those levels.
  *
  * To manually adjust these states, first select manual using
- * power_dpm_force_performance_level.  Writing a string of the level
- * numbers to the file will select which levels you want to enable.
- * E.g., writing 456 to the file will enable levels 4, 5, and 6.
- *
+ * power_dpm_force_performance_level.
+ * Secondly,Enter a new value for each level by inputing a string that
+ * contains " echo xx xx xx > pp_dpm_sclk/mclk/pcie"
+ * E.g., echo 4 5 6 to > pp_dpm_sclk will enable sclk levels 4, 5, and 6.
  */
 
 static ssize_t amdgpu_get_pp_dpm_sclk(struct device *dev,
@@ -602,23 +602,27 @@ static ssize_t amdgpu_set_pp_dpm_sclk(struct device *dev,
 	struct amdgpu_device *adev = ddev->dev_private;
 	int ret;
 	long level;
-	uint32_t i, mask = 0;
-	char sub_str[2];
+	uint32_t mask = 0;
+	char *sub_str = NULL;
+	char *tmp;
+	char buf_cpy[count];
+	const char delimiter[3] = {' ', '\n', '\0'};
 
-	for (i = 0; i < strlen(buf); i++) {
-		if (*(buf + i) == '\n')
-			continue;
-		sub_str[0] = *(buf + i);
-		sub_str[1] = '\0';
-		ret = kstrtol(sub_str, 0, &level);
+	memcpy(buf_cpy, buf, count+1);
+	tmp = buf_cpy;
+	while (tmp[0]) {
+		sub_str =  strsep(&tmp, delimiter);
+		if (strlen(sub_str)) {
+			ret = kstrtol(sub_str, 0, &level);
 
-		if (ret) {
-			count = -EINVAL;
-			goto fail;
-		}
-		mask |= 1 << level;
+			if (ret) {
+				count = -EINVAL;
+				goto fail;
+			}
+			mask |= 1 << level;
+		} else
+			break;
 	}
-
 	if (adev->powerplay.pp_funcs->force_clock_level)
 		amdgpu_dpm_force_clock_level(adev, PP_SCLK, mask);
 
@@ -648,21 +652,26 @@ static ssize_t amdgpu_set_pp_dpm_mclk(struct device *dev,
 	struct amdgpu_device *adev = ddev->dev_private;
 	int ret;
 	long level;
-	uint32_t i, mask = 0;
-	char sub_str[2];
+	uint32_t mask = 0;
+	char *sub_str = NULL;
+	char *tmp;
+	char buf_cpy[count];
+	const char delimiter[3] = {' ', '\n', '\0'};
 
-	for (i = 0; i < strlen(buf); i++) {
-		if (*(buf + i) == '\n')
-			continue;
-		sub_str[0] = *(buf + i);
-		sub_str[1] = '\0';
-		ret = kstrtol(sub_str, 0, &level);
+	memcpy(buf_cpy, buf, count+1);
+	tmp = buf_cpy;
+	while (tmp[0]) {
+		sub_str =  strsep(&tmp, delimiter);
+		if (strlen(sub_str)) {
+			ret = kstrtol(sub_str, 0, &level);
 
-		if (ret) {
-			count = -EINVAL;
-			goto fail;
-		}
-		mask |= 1 << level;
+			if (ret) {
+				count = -EINVAL;
+				goto fail;
+			}
+			mask |= 1 << level;
+		} else
+			break;
 	}
 	if (adev->powerplay.pp_funcs->force_clock_level)
 		amdgpu_dpm_force_clock_level(adev, PP_MCLK, mask);
@@ -693,21 +702,27 @@ static ssize_t amdgpu_set_pp_dpm_pcie(struct device *dev,
 	struct amdgpu_device *adev = ddev->dev_private;
 	int ret;
 	long level;
-	uint32_t i, mask = 0;
-	char sub_str[2];
+	uint32_t mask = 0;
+	char *sub_str = NULL;
+	char *tmp;
+	char buf_cpy[count];
+	const char delimiter[3] = {' ', '\n', '\0'};
 
-	for (i = 0; i < strlen(buf); i++) {
-		if (*(buf + i) == '\n')
-			continue;
-		sub_str[0] = *(buf + i);
-		sub_str[1] = '\0';
-		ret = kstrtol(sub_str, 0, &level);
+	memcpy(buf_cpy, buf, count+1);
+	tmp = buf_cpy;
 
-		if (ret) {
-			count = -EINVAL;
-			goto fail;
-		}
-		mask |= 1 << level;
+	while (tmp[0]) {
+		sub_str =  strsep(&tmp, delimiter);
+		if (strlen(sub_str)) {
+			ret = kstrtol(sub_str, 0, &level);
+
+			if (ret) {
+				count = -EINVAL;
+				goto fail;
+			}
+			mask |= 1 << level;
+		} else
+			break;
 	}
 	if (adev->powerplay.pp_funcs->force_clock_level)
 		amdgpu_dpm_force_clock_level(adev, PP_PCIE, mask);
-- 
cgit v1.1


From 09daf474d27aeb9fbd2f665b613d98c76f1e84f0 Mon Sep 17 00:00:00 2001
From: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Date: Tue, 24 Apr 2018 15:15:34 +0200
Subject: drm/amdgpu: fix amdgpu_atpx_get_client_id()'s return type

The method struct vga_switcheroo_handler::get_client_id() is defined
as returning an 'enum vga_switcheroo_client_id' but the implementation
in this driver, amdgpu_atpx_get_client_id(), returns an 'int'.

Fix this by returning 'enum vga_switcheroo_client_id' in this driver too.

Signed-off-by: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c
index 1ae5ae8..1bcb2b2 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c
@@ -550,7 +550,7 @@ static int amdgpu_atpx_init(void)
  * look up whether we are the integrated or discrete GPU (all asics).
  * Returns the client id.
  */
-static int amdgpu_atpx_get_client_id(struct pci_dev *pdev)
+static enum vga_switcheroo_client_id amdgpu_atpx_get_client_id(struct pci_dev *pdev)
 {
 	if (amdgpu_atpx_priv.dhandle == ACPI_HANDLE(&pdev->dev))
 		return VGA_SWITCHEROO_IGD;
-- 
cgit v1.1


From 4a8f264a8a8756bb7d3a478d08e449c67d291ab9 Mon Sep 17 00:00:00 2001
From: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Date: Tue, 24 Apr 2018 15:15:41 +0200
Subject: drm/radeon: fix radeon_atpx_get_client_id()'s return type

The method struct vga_switcheroo_handler::get_client_id() is defined
as returning an 'enum vga_switcheroo_client_id' but the implementation
in this driver, radeon_atpx_get_client_id(), returns an 'int'.

Fix this by returning 'enum vga_switcheroo_client_id' in this driver too.

Signed-off-by: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/radeon/radeon_atpx_handler.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/radeon/radeon_atpx_handler.c b/drivers/gpu/drm/radeon/radeon_atpx_handler.c
index 40be406..fa5fada 100644
--- a/drivers/gpu/drm/radeon/radeon_atpx_handler.c
+++ b/drivers/gpu/drm/radeon/radeon_atpx_handler.c
@@ -526,7 +526,7 @@ static int radeon_atpx_init(void)
  * look up whether we are the integrated or discrete GPU (all asics).
  * Returns the client id.
  */
-static int radeon_atpx_get_client_id(struct pci_dev *pdev)
+static enum vga_switcheroo_client_id radeon_atpx_get_client_id(struct pci_dev *pdev)
 {
 	if (radeon_atpx_priv.dhandle == ACPI_HANDLE(&pdev->dev))
 		return VGA_SWITCHEROO_IGD;
-- 
cgit v1.1


From 7a47f20eb1fb8fa8d7a8fe3a4fd8c721f04c2174 Mon Sep 17 00:00:00 2001
From: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Date: Tue, 24 Apr 2018 15:15:13 +0200
Subject: drm/radeon: fix mode_valid's return type

The method struct drm_connector_helper_funcs::mode_valid is defined
as returning an 'enum drm_mode_status' but the driver implementation
for this method uses an 'int' for it.

Fix this by using 'enum drm_mode_status' in the driver too.

Signed-off-by: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/radeon/radeon_connectors.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/radeon/radeon_connectors.c b/drivers/gpu/drm/radeon/radeon_connectors.c
index df9469a..2aea2bd 100644
--- a/drivers/gpu/drm/radeon/radeon_connectors.c
+++ b/drivers/gpu/drm/radeon/radeon_connectors.c
@@ -852,7 +852,7 @@ static int radeon_lvds_get_modes(struct drm_connector *connector)
 	return ret;
 }
 
-static int radeon_lvds_mode_valid(struct drm_connector *connector,
+static enum drm_mode_status radeon_lvds_mode_valid(struct drm_connector *connector,
 				  struct drm_display_mode *mode)
 {
 	struct drm_encoder *encoder = radeon_best_single_encoder(connector);
@@ -1012,7 +1012,7 @@ static int radeon_vga_get_modes(struct drm_connector *connector)
 	return ret;
 }
 
-static int radeon_vga_mode_valid(struct drm_connector *connector,
+static enum drm_mode_status radeon_vga_mode_valid(struct drm_connector *connector,
 				  struct drm_display_mode *mode)
 {
 	struct drm_device *dev = connector->dev;
@@ -1156,7 +1156,7 @@ static int radeon_tv_get_modes(struct drm_connector *connector)
 	return 1;
 }
 
-static int radeon_tv_mode_valid(struct drm_connector *connector,
+static enum drm_mode_status radeon_tv_mode_valid(struct drm_connector *connector,
 				struct drm_display_mode *mode)
 {
 	if ((mode->hdisplay > 1024) || (mode->vdisplay > 768))
@@ -1498,7 +1498,7 @@ static void radeon_dvi_force(struct drm_connector *connector)
 		radeon_connector->use_digital = true;
 }
 
-static int radeon_dvi_mode_valid(struct drm_connector *connector,
+static enum drm_mode_status radeon_dvi_mode_valid(struct drm_connector *connector,
 				  struct drm_display_mode *mode)
 {
 	struct drm_device *dev = connector->dev;
@@ -1800,7 +1800,7 @@ out:
 	return ret;
 }
 
-static int radeon_dp_mode_valid(struct drm_connector *connector,
+static enum drm_mode_status radeon_dp_mode_valid(struct drm_connector *connector,
 				  struct drm_display_mode *mode)
 {
 	struct drm_device *dev = connector->dev;
-- 
cgit v1.1


From ba9ca0886dc0541ac1a716b3cbd43f640a1ce8c4 Mon Sep 17 00:00:00 2001
From: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Date: Tue, 24 Apr 2018 15:14:18 +0200
Subject: drm/admgpu: fix mode_valid's return type

The method struct drm_connector_helper_funcs::mode_valid is defined
as returning an 'enum drm_mode_status' but the driver implementation
for this method uses an 'int' for it.

Fix this by using 'enum drm_mode_status' in the driver too.

Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c    | 8 ++++----
 drivers/gpu/drm/amd/amdgpu/dce_virtual.c          | 2 +-
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 2 +-
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h | 2 +-
 4 files changed, 7 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c
index 96501ff..8e66851 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c
@@ -691,7 +691,7 @@ static int amdgpu_connector_lvds_get_modes(struct drm_connector *connector)
 	return ret;
 }
 
-static int amdgpu_connector_lvds_mode_valid(struct drm_connector *connector,
+static enum drm_mode_status amdgpu_connector_lvds_mode_valid(struct drm_connector *connector,
 					     struct drm_display_mode *mode)
 {
 	struct drm_encoder *encoder = amdgpu_connector_best_single_encoder(connector);
@@ -843,7 +843,7 @@ static int amdgpu_connector_vga_get_modes(struct drm_connector *connector)
 	return ret;
 }
 
-static int amdgpu_connector_vga_mode_valid(struct drm_connector *connector,
+static enum drm_mode_status amdgpu_connector_vga_mode_valid(struct drm_connector *connector,
 					    struct drm_display_mode *mode)
 {
 	struct drm_device *dev = connector->dev;
@@ -1172,7 +1172,7 @@ static void amdgpu_connector_dvi_force(struct drm_connector *connector)
 		amdgpu_connector->use_digital = true;
 }
 
-static int amdgpu_connector_dvi_mode_valid(struct drm_connector *connector,
+static enum drm_mode_status amdgpu_connector_dvi_mode_valid(struct drm_connector *connector,
 					    struct drm_display_mode *mode)
 {
 	struct drm_device *dev = connector->dev;
@@ -1448,7 +1448,7 @@ out:
 	return ret;
 }
 
-static int amdgpu_connector_dp_mode_valid(struct drm_connector *connector,
+static enum drm_mode_status amdgpu_connector_dp_mode_valid(struct drm_connector *connector,
 					   struct drm_display_mode *mode)
 {
 	struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector);
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
index 89b2286..6454cc3 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
@@ -327,7 +327,7 @@ static int dce_virtual_get_modes(struct drm_connector *connector)
 	return 0;
 }
 
-static int dce_virtual_mode_valid(struct drm_connector *connector,
+static enum drm_mode_status dce_virtual_mode_valid(struct drm_connector *connector,
 				  struct drm_display_mode *mode)
 {
 	return MODE_OK;
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 28d8c08..656a018 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -2838,7 +2838,7 @@ static void handle_edid_mgmt(struct amdgpu_dm_connector *aconnector)
 	create_eml_sink(aconnector);
 }
 
-int amdgpu_dm_connector_mode_valid(struct drm_connector *connector,
+enum drm_mode_status amdgpu_dm_connector_mode_valid(struct drm_connector *connector,
 				   struct drm_display_mode *mode)
 {
 	int result = MODE_ERROR;
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
index 005cf0d..d5aa89a 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
@@ -247,7 +247,7 @@ void amdgpu_dm_connector_init_helper(struct amdgpu_display_manager *dm,
 				     struct dc_link *link,
 				     int link_index);
 
-int amdgpu_dm_connector_mode_valid(struct drm_connector *connector,
+enum drm_mode_status amdgpu_dm_connector_mode_valid(struct drm_connector *connector,
 				   struct drm_display_mode *mode);
 
 void dm_restore_drm_connector_state(struct drm_device *dev,
-- 
cgit v1.1


From c5a4484941be553b37facd681daf990d040cce81 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Wed, 18 Apr 2018 18:46:07 +0800
Subject: drm/amd/pp: Add OVERDRIVE support on Vega10 (v2)

when bit14 in module parameter ppfeaturemask was set.
od feature will be enabled on Vega10 except vbios not support.

user can read od range by reading sysfs pp_od_clk_voltage,
cat pp_od_clk_voltage
OD_SCLK:
0:        852Mhz        800mV
1:        991Mhz        900mV
2:       1138Mhz        950mV
3:       1269Mhz       1000mV
4:       1348Mhz       1050mV
5:       1399Mhz       1100mV
6:       1440Mhz       1150mV
7:       1500Mhz       1200mV
OD_MCLK:
0:        167Mhz        800mV
1:        500Mhz        800mV
2:        800Mhz        950mV
3:        945Mhz       1000mV
OD_RANGE:
SCLK:     852MHz       2200MHz
MCLK:     167MHz       1500MHz
VDDC:     800mV        1200mV

and can configure the clock/voltage by writing pp_od_clk_voltage
for example:

echo "s 0 900 820">pp_od_clk_voltage to change the sclk/vddc
to 900MHz and 820 mV in dpm level0.

echo "r" to change the clk/voltage to  default value.

echo "c">pp_od_clk_voltage
to commit the change

v2: squash in warning fix (Alex)

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c | 719 +++++++++++----------
 drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.h |  26 +-
 .../gpu/drm/amd/powerplay/inc/hardwaremanager.h    |   6 +-
 3 files changed, 390 insertions(+), 361 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
index 384aa07..7486120 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
@@ -285,6 +285,48 @@ static int vega10_set_features_platform_caps(struct pp_hwmgr *hwmgr)
 	return 0;
 }
 
+static int vega10_odn_initial_default_setting(struct pp_hwmgr *hwmgr)
+{
+	struct vega10_hwmgr *data = hwmgr->backend;
+	struct phm_ppt_v2_information *table_info =
+			(struct phm_ppt_v2_information *)(hwmgr->pptable);
+	struct vega10_odn_dpm_table *odn_table = &(data->odn_dpm_table);
+	struct vega10_odn_vddc_lookup_table *od_lookup_table;
+	struct phm_ppt_v1_voltage_lookup_table *vddc_lookup_table;
+	struct phm_ppt_v1_clock_voltage_dependency_table *dep_table[3];
+	struct phm_ppt_v1_clock_voltage_dependency_table *od_table[3];
+	uint32_t i;
+
+	od_lookup_table = &odn_table->vddc_lookup_table;
+	vddc_lookup_table = table_info->vddc_lookup_table;
+
+	for (i = 0; i < vddc_lookup_table->count; i++)
+		od_lookup_table->entries[i].us_vdd = vddc_lookup_table->entries[i].us_vdd;
+
+	od_lookup_table->count = vddc_lookup_table->count;
+
+	dep_table[0] = table_info->vdd_dep_on_sclk;
+	dep_table[1] = table_info->vdd_dep_on_mclk;
+	dep_table[2] = table_info->vdd_dep_on_socclk;
+	od_table[0] = (struct phm_ppt_v1_clock_voltage_dependency_table *)&odn_table->vdd_dep_on_sclk;
+	od_table[1] = (struct phm_ppt_v1_clock_voltage_dependency_table *)&odn_table->vdd_dep_on_mclk;
+	od_table[2] = (struct phm_ppt_v1_clock_voltage_dependency_table *)&odn_table->vdd_dep_on_socclk;
+
+	for (i = 0; i < 3; i++)
+		smu_get_voltage_dependency_table_ppt_v1(dep_table[i], od_table[i]);
+
+	if (odn_table->max_vddc == 0 || odn_table->max_vddc > 2000)
+		odn_table->max_vddc = dep_table[0]->entries[dep_table[0]->count - 1].vddc;
+	if (odn_table->min_vddc == 0 || odn_table->min_vddc > 2000)
+		odn_table->min_vddc = dep_table[0]->entries[0].vddc;
+
+	i = od_table[2]->count - 1;
+	od_table[2]->entries[i].clk = hwmgr->platform_descriptor.overdriveLimit.memoryClock;
+	od_table[2]->entries[i].vddc = odn_table->max_vddc;
+
+	return 0;
+}
+
 static void vega10_init_dpm_defaults(struct pp_hwmgr *hwmgr)
 {
 	struct vega10_hwmgr *data = hwmgr->backend;
@@ -421,7 +463,6 @@ static void vega10_init_dpm_defaults(struct pp_hwmgr *hwmgr)
 		/* ACG firmware has major version 5 */
 	if ((hwmgr->smu_version & 0xff000000) == 0x5000000)
 		data->smu_features[GNLD_ACG].supported = true;
-
 	if (data->registry_data.didt_support)
 		data->smu_features[GNLD_DIDT].supported = true;
 
@@ -1360,48 +1401,6 @@ static int vega10_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
 	memcpy(&(data->golden_dpm_table), &(data->dpm_table),
 			sizeof(struct vega10_dpm_table));
 
-	if (PP_CAP(PHM_PlatformCaps_ODNinACSupport) ||
-	    PP_CAP(PHM_PlatformCaps_ODNinDCSupport)) {
-		data->odn_dpm_table.odn_core_clock_dpm_levels.num_of_pl =
-						data->dpm_table.gfx_table.count;
-		for (i = 0; i < data->dpm_table.gfx_table.count; i++) {
-			data->odn_dpm_table.odn_core_clock_dpm_levels.entries[i].clock =
-					data->dpm_table.gfx_table.dpm_levels[i].value;
-			data->odn_dpm_table.odn_core_clock_dpm_levels.entries[i].enabled = true;
-		}
-
-		data->odn_dpm_table.vdd_dependency_on_sclk.count =
-				dep_gfx_table->count;
-		for (i = 0; i < dep_gfx_table->count; i++) {
-			data->odn_dpm_table.vdd_dependency_on_sclk.entries[i].clk =
-					dep_gfx_table->entries[i].clk;
-			data->odn_dpm_table.vdd_dependency_on_sclk.entries[i].vddInd =
-					dep_gfx_table->entries[i].vddInd;
-			data->odn_dpm_table.vdd_dependency_on_sclk.entries[i].cks_enable =
-					dep_gfx_table->entries[i].cks_enable;
-			data->odn_dpm_table.vdd_dependency_on_sclk.entries[i].cks_voffset =
-					dep_gfx_table->entries[i].cks_voffset;
-		}
-
-		data->odn_dpm_table.odn_memory_clock_dpm_levels.num_of_pl =
-						data->dpm_table.mem_table.count;
-		for (i = 0; i < data->dpm_table.mem_table.count; i++) {
-			data->odn_dpm_table.odn_memory_clock_dpm_levels.entries[i].clock =
-					data->dpm_table.mem_table.dpm_levels[i].value;
-			data->odn_dpm_table.odn_memory_clock_dpm_levels.entries[i].enabled = true;
-		}
-
-		data->odn_dpm_table.vdd_dependency_on_mclk.count = dep_mclk_table->count;
-		for (i = 0; i < dep_mclk_table->count; i++) {
-			data->odn_dpm_table.vdd_dependency_on_mclk.entries[i].clk =
-					dep_mclk_table->entries[i].clk;
-			data->odn_dpm_table.vdd_dependency_on_mclk.entries[i].vddInd =
-					dep_mclk_table->entries[i].vddInd;
-			data->odn_dpm_table.vdd_dependency_on_mclk.entries[i].vddci =
-					dep_mclk_table->entries[i].vddci;
-		}
-	}
-
 	return 0;
 }
 
@@ -1504,18 +1503,18 @@ static int vega10_populate_single_gfx_level(struct pp_hwmgr *hwmgr,
 {
 	struct phm_ppt_v2_information *table_info =
 			(struct phm_ppt_v2_information *)(hwmgr->pptable);
-	struct phm_ppt_v1_clock_voltage_dependency_table *dep_on_sclk =
-			table_info->vdd_dep_on_sclk;
+	struct phm_ppt_v1_clock_voltage_dependency_table *dep_on_sclk;
 	struct vega10_hwmgr *data = hwmgr->backend;
 	struct pp_atomfwctrl_clock_dividers_soc15 dividers;
 	uint32_t gfx_max_clock =
 			hwmgr->platform_descriptor.overdriveLimit.engineClock;
 	uint32_t i = 0;
 
-	if (data->apply_overdrive_next_settings_mask &
-			DPMTABLE_OD_UPDATE_VDDC)
+	if (hwmgr->od_enabled)
 		dep_on_sclk = (struct phm_ppt_v1_clock_voltage_dependency_table *)
-						&(data->odn_dpm_table.vdd_dependency_on_sclk);
+						&(data->odn_dpm_table.vdd_dep_on_sclk);
+	else
+		dep_on_sclk = table_info->vdd_dep_on_sclk;
 
 	PP_ASSERT_WITH_CODE(dep_on_sclk,
 			"Invalid SOC_VDD-GFX_CLK Dependency Table!",
@@ -1567,23 +1566,32 @@ static int vega10_populate_single_soc_level(struct pp_hwmgr *hwmgr,
 		uint32_t soc_clock, uint8_t *current_soc_did,
 		uint8_t *current_vol_index)
 {
+	struct vega10_hwmgr *data = hwmgr->backend;
 	struct phm_ppt_v2_information *table_info =
 			(struct phm_ppt_v2_information *)(hwmgr->pptable);
-	struct phm_ppt_v1_clock_voltage_dependency_table *dep_on_soc =
-			table_info->vdd_dep_on_socclk;
+	struct phm_ppt_v1_clock_voltage_dependency_table *dep_on_soc;
 	struct pp_atomfwctrl_clock_dividers_soc15 dividers;
 	uint32_t i;
 
-	PP_ASSERT_WITH_CODE(dep_on_soc,
-			"Invalid SOC_VDD-SOC_CLK Dependency Table!",
-			return -EINVAL);
-	for (i = 0; i < dep_on_soc->count; i++) {
-		if (dep_on_soc->entries[i].clk == soc_clock)
-			break;
+	if (hwmgr->od_enabled) {
+		dep_on_soc = (struct phm_ppt_v1_clock_voltage_dependency_table *)
+						&data->odn_dpm_table.vdd_dep_on_socclk;
+		for (i = 0; i < dep_on_soc->count; i++) {
+			if (dep_on_soc->entries[i].clk >= soc_clock)
+				break;
+		}
+	} else {
+		dep_on_soc = table_info->vdd_dep_on_socclk;
+		for (i = 0; i < dep_on_soc->count; i++) {
+			if (dep_on_soc->entries[i].clk == soc_clock)
+				break;
+		}
 	}
+
 	PP_ASSERT_WITH_CODE(dep_on_soc->count > i,
 			"Cannot find SOC_CLK in SOC_VDD-SOC_CLK Dependency Table",
 			return -EINVAL);
+
 	PP_ASSERT_WITH_CODE(!pp_atomfwctrl_get_gpu_pll_dividers_vega10(hwmgr,
 			COMPUTE_GPUCLK_INPUT_FLAG_DEFAULT_GPUCLK,
 			soc_clock, &dividers),
@@ -1592,22 +1600,6 @@ static int vega10_populate_single_soc_level(struct pp_hwmgr *hwmgr,
 
 	*current_soc_did = (uint8_t)dividers.ulDid;
 	*current_vol_index = (uint8_t)(dep_on_soc->entries[i].vddInd);
-
-	return 0;
-}
-
-uint16_t vega10_locate_vddc_given_clock(struct pp_hwmgr *hwmgr,
-		uint32_t clk,
-		struct phm_ppt_v1_clock_voltage_dependency_table *dep_table)
-{
-	uint16_t i;
-
-	for (i = 0; i < dep_table->count; i++) {
-		if (dep_table->entries[i].clk == clk)
-			return dep_table->entries[i].vddc;
-	}
-
-	pr_info("[LocateVddcGivenClock] Cannot locate SOC Vddc for this clock!");
 	return 0;
 }
 
@@ -1621,8 +1613,6 @@ static int vega10_populate_all_graphic_levels(struct pp_hwmgr *hwmgr)
 	struct vega10_hwmgr *data = hwmgr->backend;
 	struct phm_ppt_v2_information *table_info =
 			(struct phm_ppt_v2_information *)(hwmgr->pptable);
-	struct phm_ppt_v1_clock_voltage_dependency_table *dep_table =
-			table_info->vdd_dep_on_socclk;
 	PPTable_t *pp_table = &(data->smc_state_table.pp_table);
 	struct vega10_single_dpm_table *dpm_table = &(data->dpm_table.gfx_table);
 	int result = 0;
@@ -1653,11 +1643,6 @@ static int vega10_populate_all_graphic_levels(struct pp_hwmgr *hwmgr)
 
 	dpm_table = &(data->dpm_table.soc_table);
 	for (i = 0; i < dpm_table->count; i++) {
-		pp_table->SocVid[i] =
-				(uint8_t)convert_to_vid(
-				vega10_locate_vddc_given_clock(hwmgr,
-						dpm_table->dpm_levels[i].value,
-						dep_table));
 		result = vega10_populate_single_soc_level(hwmgr,
 				dpm_table->dpm_levels[i].value,
 				&(pp_table->SocclkDid[i]),
@@ -1668,7 +1653,6 @@ static int vega10_populate_all_graphic_levels(struct pp_hwmgr *hwmgr)
 
 	j = i - 1;
 	while (i < NUM_SOCCLK_DPM_LEVELS) {
-		pp_table->SocVid[i] = pp_table->SocVid[j];
 		result = vega10_populate_single_soc_level(hwmgr,
 				dpm_table->dpm_levels[j].value,
 				&(pp_table->SocclkDid[i]),
@@ -1681,6 +1665,32 @@ static int vega10_populate_all_graphic_levels(struct pp_hwmgr *hwmgr)
 	return result;
 }
 
+static void vega10_populate_vddc_soc_levels(struct pp_hwmgr *hwmgr)
+{
+	struct vega10_hwmgr *data = hwmgr->backend;
+	PPTable_t *pp_table = &(data->smc_state_table.pp_table);
+	struct phm_ppt_v2_information *table_info = hwmgr->pptable;
+	struct phm_ppt_v1_voltage_lookup_table *vddc_lookup_table;
+
+	uint8_t soc_vid = 0;
+	uint32_t i, max_vddc_level;
+
+	if (hwmgr->od_enabled)
+		vddc_lookup_table = (struct phm_ppt_v1_voltage_lookup_table *)&data->odn_dpm_table.vddc_lookup_table;
+	else
+		vddc_lookup_table = table_info->vddc_lookup_table;
+
+	max_vddc_level = vddc_lookup_table->count;
+	for (i = 0; i < max_vddc_level; i++) {
+		soc_vid = (uint8_t)convert_to_vid(vddc_lookup_table->entries[i].us_vdd);
+		pp_table->SocVid[i] = soc_vid;
+	}
+	while (i < MAX_REGULAR_DPM_NUMBER) {
+		pp_table->SocVid[i] = soc_vid;
+		i++;
+	}
+}
+
 /**
  * @brief Populates single SMC GFXCLK structure using the provided clock.
  *
@@ -1695,25 +1705,25 @@ static int vega10_populate_single_memory_level(struct pp_hwmgr *hwmgr,
 	struct vega10_hwmgr *data = hwmgr->backend;
 	struct phm_ppt_v2_information *table_info =
 			(struct phm_ppt_v2_information *)(hwmgr->pptable);
-	struct phm_ppt_v1_clock_voltage_dependency_table *dep_on_mclk =
-			table_info->vdd_dep_on_mclk;
+	struct phm_ppt_v1_clock_voltage_dependency_table *dep_on_mclk;
 	struct pp_atomfwctrl_clock_dividers_soc15 dividers;
 	uint32_t mem_max_clock =
 			hwmgr->platform_descriptor.overdriveLimit.memoryClock;
 	uint32_t i = 0;
 
-	if (data->apply_overdrive_next_settings_mask &
-			DPMTABLE_OD_UPDATE_VDDC)
+	if (hwmgr->od_enabled)
 		dep_on_mclk = (struct phm_ppt_v1_clock_voltage_dependency_table *)
-					&data->odn_dpm_table.vdd_dependency_on_mclk;
+					&data->odn_dpm_table.vdd_dep_on_mclk;
+	else
+		dep_on_mclk = table_info->vdd_dep_on_mclk;
 
 	PP_ASSERT_WITH_CODE(dep_on_mclk,
 			"Invalid SOC_VDD-UCLK Dependency Table!",
 			return -EINVAL);
 
-	if (data->need_update_dpm_table & DPMTABLE_OD_UPDATE_MCLK)
+	if (data->need_update_dpm_table & DPMTABLE_OD_UPDATE_MCLK) {
 		mem_clock = mem_clock > mem_max_clock ? mem_max_clock : mem_clock;
-	else {
+	} else {
 		for (i = 0; i < dep_on_mclk->count; i++) {
 			if (dep_on_mclk->entries[i].clk == mem_clock)
 				break;
@@ -2057,6 +2067,9 @@ static int vega10_populate_avfs_parameters(struct pp_hwmgr *hwmgr)
 	if (data->smu_features[GNLD_AVFS].supported) {
 		result = pp_atomfwctrl_get_avfs_information(hwmgr, &avfs_params);
 		if (!result) {
+			data->odn_dpm_table.max_vddc = avfs_params.ulMaxVddc;
+			data->odn_dpm_table.min_vddc = avfs_params.ulMinVddc;
+
 			pp_table->MinVoltageVid = (uint8_t)
 					convert_to_vid((uint16_t)(avfs_params.ulMinVddc));
 			pp_table->MaxVoltageVid = (uint8_t)
@@ -2335,6 +2348,22 @@ static int vega10_avfs_enable(struct pp_hwmgr *hwmgr, bool enable)
 	return 0;
 }
 
+static int vega10_update_avfs(struct pp_hwmgr *hwmgr)
+{
+	struct vega10_hwmgr *data = hwmgr->backend;
+
+	if (data->need_update_dpm_table & DPMTABLE_OD_UPDATE_VDDC) {
+		vega10_avfs_enable(hwmgr, false);
+	} else if (data->need_update_dpm_table) {
+		vega10_avfs_enable(hwmgr, false);
+		vega10_avfs_enable(hwmgr, true);
+	} else {
+		vega10_avfs_enable(hwmgr, true);
+	}
+
+	return 0;
+}
+
 static int vega10_populate_and_upload_avfs_fuse_override(struct pp_hwmgr *hwmgr)
 {
 	int result = 0;
@@ -2396,6 +2425,10 @@ static int vega10_init_smc_table(struct pp_hwmgr *hwmgr)
 			"Failed to setup default DPM tables!",
 			return result);
 
+	/* initialize ODN table */
+	if (hwmgr->od_enabled)
+		vega10_odn_initial_default_setting(hwmgr);
+
 	pp_atomfwctrl_get_voltage_table_v4(hwmgr, VOLTAGE_TYPE_VDDC,
 			VOLTAGE_OBJ_SVID2,  &voltage_table);
 	pp_table->MaxVidStep = voltage_table.max_vid_step;
@@ -2442,6 +2475,8 @@ static int vega10_init_smc_table(struct pp_hwmgr *hwmgr)
 			"Failed to initialize Memory Level!",
 			return result);
 
+	vega10_populate_vddc_soc_levels(hwmgr);
+
 	result = vega10_populate_all_display_clock_levels(hwmgr);
 	PP_ASSERT_WITH_CODE(!result,
 			"Failed to initialize Display Level!",
@@ -3164,82 +3199,11 @@ static int vega10_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 
 static int vega10_find_dpm_states_clocks_in_dpm_table(struct pp_hwmgr *hwmgr, const void *input)
 {
-	const struct phm_set_power_state_input *states =
-			(const struct phm_set_power_state_input *)input;
-	const struct vega10_power_state *vega10_ps =
-			cast_const_phw_vega10_power_state(states->pnew_state);
 	struct vega10_hwmgr *data = hwmgr->backend;
-	struct vega10_single_dpm_table *sclk_table =
-			&(data->dpm_table.gfx_table);
-	uint32_t sclk = vega10_ps->performance_levels
-			[vega10_ps->performance_level_count - 1].gfx_clock;
-	struct vega10_single_dpm_table *mclk_table =
-			&(data->dpm_table.mem_table);
-	uint32_t mclk = vega10_ps->performance_levels
-			[vega10_ps->performance_level_count - 1].mem_clock;
-	struct PP_Clocks min_clocks = {0};
-	uint32_t i;
-
-	data->need_update_dpm_table = 0;
-
-	if (PP_CAP(PHM_PlatformCaps_ODNinACSupport) ||
-	    PP_CAP(PHM_PlatformCaps_ODNinDCSupport)) {
-		for (i = 0; i < sclk_table->count; i++) {
-			if (sclk == sclk_table->dpm_levels[i].value)
-				break;
-		}
-
-		if (!(data->apply_overdrive_next_settings_mask &
-				DPMTABLE_OD_UPDATE_SCLK) && i >= sclk_table->count) {
-			/* Check SCLK in DAL's minimum clocks
-			 * in case DeepSleep divider update is required.
-			 */
-			if (data->display_timing.min_clock_in_sr !=
-					min_clocks.engineClockInSR &&
-				(min_clocks.engineClockInSR >=
-						VEGA10_MINIMUM_ENGINE_CLOCK ||
-					data->display_timing.min_clock_in_sr >=
-						VEGA10_MINIMUM_ENGINE_CLOCK))
-				data->need_update_dpm_table |= DPMTABLE_UPDATE_SCLK;
-		}
-
-		if (data->display_timing.num_existing_displays !=
-				hwmgr->display_config->num_display)
-			data->need_update_dpm_table |= DPMTABLE_UPDATE_MCLK;
-	} else {
-		for (i = 0; i < sclk_table->count; i++) {
-			if (sclk == sclk_table->dpm_levels[i].value)
-				break;
-		}
-
-		if (i >= sclk_table->count)
-			data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_SCLK;
-		else {
-			/* Check SCLK in DAL's minimum clocks
-			 * in case DeepSleep divider update is required.
-			 */
-			if (data->display_timing.min_clock_in_sr !=
-					min_clocks.engineClockInSR &&
-				(min_clocks.engineClockInSR >=
-						VEGA10_MINIMUM_ENGINE_CLOCK ||
-					data->display_timing.min_clock_in_sr >=
-						VEGA10_MINIMUM_ENGINE_CLOCK))
-				data->need_update_dpm_table |= DPMTABLE_UPDATE_SCLK;
-		}
 
-		for (i = 0; i < mclk_table->count; i++) {
-			if (mclk == mclk_table->dpm_levels[i].value)
-				break;
-		}
-
-		if (i >= mclk_table->count)
-			data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_MCLK;
+	if (data->display_timing.num_existing_displays != hwmgr->display_config->num_display)
+		data->need_update_dpm_table |= DPMTABLE_UPDATE_MCLK;
 
-		if (data->display_timing.num_existing_displays !=
-				hwmgr->display_config->num_display ||
-				i >= mclk_table->count)
-			data->need_update_dpm_table |= DPMTABLE_UPDATE_MCLK;
-	}
 	return 0;
 }
 
@@ -3247,194 +3211,29 @@ static int vega10_populate_and_upload_sclk_mclk_dpm_levels(
 		struct pp_hwmgr *hwmgr, const void *input)
 {
 	int result = 0;
-	const struct phm_set_power_state_input *states =
-			(const struct phm_set_power_state_input *)input;
-	const struct vega10_power_state *vega10_ps =
-			cast_const_phw_vega10_power_state(states->pnew_state);
 	struct vega10_hwmgr *data = hwmgr->backend;
-	uint32_t sclk = vega10_ps->performance_levels
-			[vega10_ps->performance_level_count - 1].gfx_clock;
-	uint32_t mclk = vega10_ps->performance_levels
-			[vega10_ps->performance_level_count - 1].mem_clock;
-	struct vega10_dpm_table *dpm_table = &data->dpm_table;
-	struct vega10_dpm_table *golden_dpm_table =
-			&data->golden_dpm_table;
-	uint32_t dpm_count, clock_percent;
-	uint32_t i;
-
-	if (PP_CAP(PHM_PlatformCaps_ODNinACSupport) ||
-	    PP_CAP(PHM_PlatformCaps_ODNinDCSupport)) {
-
-		if (!data->need_update_dpm_table &&
-			!data->apply_optimized_settings &&
-			!data->apply_overdrive_next_settings_mask)
-			return 0;
 
-		if (data->apply_overdrive_next_settings_mask &
-				DPMTABLE_OD_UPDATE_SCLK) {
-			for (dpm_count = 0;
-					dpm_count < dpm_table->gfx_table.count;
-					dpm_count++) {
-				dpm_table->gfx_table.dpm_levels[dpm_count].enabled =
-					data->odn_dpm_table.odn_core_clock_dpm_levels.entries[dpm_count].enabled;
-				dpm_table->gfx_table.dpm_levels[dpm_count].value =
-					data->odn_dpm_table.odn_core_clock_dpm_levels.entries[dpm_count].clock;
-			}
-		}
-
-		if (data->apply_overdrive_next_settings_mask &
-				DPMTABLE_OD_UPDATE_MCLK) {
-			for (dpm_count = 0;
-					dpm_count < dpm_table->mem_table.count;
-					dpm_count++) {
-				dpm_table->mem_table.dpm_levels[dpm_count].enabled =
-					data->odn_dpm_table.odn_memory_clock_dpm_levels.entries[dpm_count].enabled;
-				dpm_table->mem_table.dpm_levels[dpm_count].value =
-					data->odn_dpm_table.odn_memory_clock_dpm_levels.entries[dpm_count].clock;
-			}
-		}
-
-		if ((data->need_update_dpm_table & DPMTABLE_UPDATE_SCLK) ||
-			data->apply_optimized_settings ||
-			(data->apply_overdrive_next_settings_mask &
-					DPMTABLE_OD_UPDATE_SCLK)) {
-			result = vega10_populate_all_graphic_levels(hwmgr);
-			PP_ASSERT_WITH_CODE(!result,
-					"Failed to populate SCLK during PopulateNewDPMClocksStates Function!",
-					return result);
-		}
-
-		if ((data->need_update_dpm_table & DPMTABLE_UPDATE_MCLK) ||
-			(data->apply_overdrive_next_settings_mask &
-					DPMTABLE_OD_UPDATE_MCLK)){
-			result = vega10_populate_all_memory_levels(hwmgr);
-			PP_ASSERT_WITH_CODE(!result,
-					"Failed to populate MCLK during PopulateNewDPMClocksStates Function!",
-					return result);
-		}
-	} else {
-		if (!data->need_update_dpm_table &&
-				!data->apply_optimized_settings)
-			return 0;
-
-		if (data->need_update_dpm_table & DPMTABLE_OD_UPDATE_SCLK &&
-				data->smu_features[GNLD_DPM_GFXCLK].supported) {
-				dpm_table->
-				gfx_table.dpm_levels[dpm_table->gfx_table.count - 1].
-				value = sclk;
-				if (hwmgr->od_enabled) {
-					/* Need to do calculation based on the golden DPM table
-					 * as the Heatmap GPU Clock axis is also based on
-					 * the default values
-					 */
-					PP_ASSERT_WITH_CODE(
-							golden_dpm_table->gfx_table.dpm_levels
-							[golden_dpm_table->gfx_table.count - 1].value,
-							"Divide by 0!",
-							return -1);
-
-					dpm_count = dpm_table->gfx_table.count < 2 ?
-							0 : dpm_table->gfx_table.count - 2;
-					for (i = dpm_count; i > 1; i--) {
-						if (sclk > golden_dpm_table->gfx_table.dpm_levels
-							[golden_dpm_table->gfx_table.count - 1].value) {
-							clock_percent =
-								((sclk - golden_dpm_table->gfx_table.dpm_levels
-								[golden_dpm_table->gfx_table.count - 1].value) *
-								100) /
-								golden_dpm_table->gfx_table.dpm_levels
-								[golden_dpm_table->gfx_table.count - 1].value;
-
-							dpm_table->gfx_table.dpm_levels[i].value =
-								golden_dpm_table->gfx_table.dpm_levels[i].value +
-								(golden_dpm_table->gfx_table.dpm_levels[i].value *
-								clock_percent) / 100;
-						} else if (golden_dpm_table->
-								gfx_table.dpm_levels[dpm_table->gfx_table.count-1].value >
-								sclk) {
-							clock_percent =
-								((golden_dpm_table->gfx_table.dpm_levels
-								[golden_dpm_table->gfx_table.count - 1].value -
-								sclk) *	100) /
-								golden_dpm_table->gfx_table.dpm_levels
-								[golden_dpm_table->gfx_table.count-1].value;
-
-							dpm_table->gfx_table.dpm_levels[i].value =
-								golden_dpm_table->gfx_table.dpm_levels[i].value -
-								(golden_dpm_table->gfx_table.dpm_levels[i].value *
-								clock_percent) / 100;
-						} else
-							dpm_table->gfx_table.dpm_levels[i].value =
-								golden_dpm_table->gfx_table.dpm_levels[i].value;
-					}
-				}
-			}
-
-		if (data->need_update_dpm_table & DPMTABLE_OD_UPDATE_MCLK &&
-				data->smu_features[GNLD_DPM_UCLK].supported) {
-			dpm_table->
-			mem_table.dpm_levels[dpm_table->mem_table.count - 1].
-			value = mclk;
+	if (!data->need_update_dpm_table)
+		return 0;
 
-			if (hwmgr->od_enabled) {
-				PP_ASSERT_WITH_CODE(
-					golden_dpm_table->mem_table.dpm_levels
-					[golden_dpm_table->mem_table.count - 1].value,
-					"Divide by 0!",
-					return -1);
+	if (data->need_update_dpm_table &
+			(DPMTABLE_OD_UPDATE_SCLK + DPMTABLE_UPDATE_SCLK + DPMTABLE_UPDATE_SOCCLK)) {
+		result = vega10_populate_all_graphic_levels(hwmgr);
+		PP_ASSERT_WITH_CODE((0 == result),
+				"Failed to populate SCLK during PopulateNewDPMClocksStates Function!",
+				return result);
+	}
 
-				dpm_count = dpm_table->mem_table.count < 2 ?
-						0 : dpm_table->mem_table.count - 2;
-				for (i = dpm_count; i > 1; i--) {
-					if (mclk > golden_dpm_table->mem_table.dpm_levels
-						[golden_dpm_table->mem_table.count-1].value) {
-						clock_percent = ((mclk -
-							golden_dpm_table->mem_table.dpm_levels
-							[golden_dpm_table->mem_table.count-1].value) *
-							100) /
-							golden_dpm_table->mem_table.dpm_levels
-							[golden_dpm_table->mem_table.count-1].value;
-
-						dpm_table->mem_table.dpm_levels[i].value =
-							golden_dpm_table->mem_table.dpm_levels[i].value +
-							(golden_dpm_table->mem_table.dpm_levels[i].value *
-							clock_percent) / 100;
-					} else if (golden_dpm_table->mem_table.dpm_levels
-							[dpm_table->mem_table.count-1].value > mclk) {
-						clock_percent = ((golden_dpm_table->mem_table.dpm_levels
-							[golden_dpm_table->mem_table.count-1].value - mclk) *
-							100) /
-							golden_dpm_table->mem_table.dpm_levels
-							[golden_dpm_table->mem_table.count-1].value;
-
-						dpm_table->mem_table.dpm_levels[i].value =
-							golden_dpm_table->mem_table.dpm_levels[i].value -
-							(golden_dpm_table->mem_table.dpm_levels[i].value *
-							clock_percent) / 100;
-					} else
-						dpm_table->mem_table.dpm_levels[i].value =
-							golden_dpm_table->mem_table.dpm_levels[i].value;
-				}
-			}
-		}
+	if (data->need_update_dpm_table &
+			(DPMTABLE_OD_UPDATE_MCLK + DPMTABLE_UPDATE_MCLK)) {
+		result = vega10_populate_all_memory_levels(hwmgr);
+		PP_ASSERT_WITH_CODE((0 == result),
+				"Failed to populate MCLK during PopulateNewDPMClocksStates Function!",
+				return result);
+	}
 
-		if ((data->need_update_dpm_table &
-			(DPMTABLE_OD_UPDATE_SCLK + DPMTABLE_UPDATE_SCLK)) ||
-			data->apply_optimized_settings) {
-			result = vega10_populate_all_graphic_levels(hwmgr);
-			PP_ASSERT_WITH_CODE(!result,
-					"Failed to populate SCLK during PopulateNewDPMClocksStates Function!",
-					return result);
-		}
+	vega10_populate_vddc_soc_levels(hwmgr);
 
-		if (data->need_update_dpm_table &
-				(DPMTABLE_OD_UPDATE_MCLK + DPMTABLE_UPDATE_MCLK)) {
-			result = vega10_populate_all_memory_levels(hwmgr);
-			PP_ASSERT_WITH_CODE(!result,
-					"Failed to populate MCLK during PopulateNewDPMClocksStates Function!",
-					return result);
-		}
-	}
 	return result;
 }
 
@@ -3730,8 +3529,9 @@ static int vega10_set_power_state_tasks(struct pp_hwmgr *hwmgr,
 	PP_ASSERT_WITH_CODE(!result,
 			"Failed to upload PPtable!", return result);
 
-	data->apply_optimized_settings = false;
-	data->apply_overdrive_next_settings_mask = 0;
+	vega10_update_avfs(hwmgr);
+
+	data->need_update_dpm_table &= DPMTABLE_OD_UPDATE_VDDC;
 
 	return 0;
 }
@@ -4383,6 +4183,8 @@ static int vega10_print_clock_levels(struct pp_hwmgr *hwmgr,
 	struct vega10_single_dpm_table *sclk_table = &(data->dpm_table.gfx_table);
 	struct vega10_single_dpm_table *mclk_table = &(data->dpm_table.mem_table);
 	struct vega10_pcie_table *pcie_table = &(data->dpm_table.pcie_table);
+	struct vega10_odn_clock_voltage_dependency_table *podn_vdd_dep = NULL;
+
 	int i, now, size = 0;
 
 	switch (type) {
@@ -4421,6 +4223,40 @@ static int vega10_print_clock_levels(struct pp_hwmgr *hwmgr,
 					(pcie_table->pcie_gen[i] == 2) ? "8.0GT/s, x16" : "",
 					(i == now) ? "*" : "");
 		break;
+	case OD_SCLK:
+		if (hwmgr->od_enabled) {
+			size = sprintf(buf, "%s:\n", "OD_SCLK");
+			podn_vdd_dep = &data->odn_dpm_table.vdd_dep_on_sclk;
+			for (i = 0; i < podn_vdd_dep->count; i++)
+				size += sprintf(buf + size, "%d: %10uMhz %10umV\n",
+					i, podn_vdd_dep->entries[i].clk / 100,
+						podn_vdd_dep->entries[i].vddc);
+		}
+		break;
+	case OD_MCLK:
+		if (hwmgr->od_enabled) {
+			size = sprintf(buf, "%s:\n", "OD_MCLK");
+			podn_vdd_dep = &data->odn_dpm_table.vdd_dep_on_mclk;
+			for (i = 0; i < podn_vdd_dep->count; i++)
+				size += sprintf(buf + size, "%d: %10uMhz %10umV\n",
+					i, podn_vdd_dep->entries[i].clk/100,
+						podn_vdd_dep->entries[i].vddc);
+		}
+		break;
+	case OD_RANGE:
+		if (hwmgr->od_enabled) {
+			size = sprintf(buf, "%s:\n", "OD_RANGE");
+			size += sprintf(buf + size, "SCLK: %7uMHz %10uMHz\n",
+				data->golden_dpm_table.gfx_table.dpm_levels[0].value/100,
+				hwmgr->platform_descriptor.overdriveLimit.engineClock/100);
+			size += sprintf(buf + size, "MCLK: %7uMHz %10uMHz\n",
+				data->golden_dpm_table.mem_table.dpm_levels[0].value/100,
+				hwmgr->platform_descriptor.overdriveLimit.memoryClock/100);
+			size += sprintf(buf + size, "VDDC: %7umV %11umV\n",
+				data->odn_dpm_table.min_vddc,
+				data->odn_dpm_table.max_vddc);
+		}
+		break;
 	default:
 		break;
 	}
@@ -4808,6 +4644,200 @@ static int vega10_set_power_profile_mode(struct pp_hwmgr *hwmgr, long *input, ui
 	return 0;
 }
 
+
+static bool vega10_check_clk_voltage_valid(struct pp_hwmgr *hwmgr,
+					enum PP_OD_DPM_TABLE_COMMAND type,
+					uint32_t clk,
+					uint32_t voltage)
+{
+	struct vega10_hwmgr *data = hwmgr->backend;
+	struct vega10_odn_dpm_table *odn_table = &(data->odn_dpm_table);
+	struct vega10_single_dpm_table *golden_table;
+
+	if (voltage < odn_table->min_vddc || voltage > odn_table->max_vddc) {
+		pr_info("OD voltage is out of range [%d - %d] mV\n", odn_table->min_vddc, odn_table->max_vddc);
+		return false;
+	}
+
+	if (type == PP_OD_EDIT_SCLK_VDDC_TABLE) {
+		golden_table = &(data->golden_dpm_table.gfx_table);
+		if (golden_table->dpm_levels[0].value > clk ||
+			hwmgr->platform_descriptor.overdriveLimit.engineClock < clk) {
+			pr_info("OD engine clock is out of range [%d - %d] MHz\n",
+				golden_table->dpm_levels[0].value/100,
+				hwmgr->platform_descriptor.overdriveLimit.engineClock/100);
+			return false;
+		}
+	} else if (type == PP_OD_EDIT_MCLK_VDDC_TABLE) {
+		golden_table = &(data->golden_dpm_table.mem_table);
+		if (golden_table->dpm_levels[0].value > clk ||
+			hwmgr->platform_descriptor.overdriveLimit.memoryClock < clk) {
+			pr_info("OD memory clock is out of range [%d - %d] MHz\n",
+				golden_table->dpm_levels[0].value/100,
+				hwmgr->platform_descriptor.overdriveLimit.memoryClock/100);
+			return false;
+		}
+	} else {
+		return false;
+	}
+
+	return true;
+}
+
+static void vega10_check_dpm_table_updated(struct pp_hwmgr *hwmgr)
+{
+	struct vega10_hwmgr *data = hwmgr->backend;
+	struct vega10_odn_dpm_table *odn_table = &(data->odn_dpm_table);
+	struct phm_ppt_v2_information *table_info = hwmgr->pptable;
+	struct phm_ppt_v1_clock_voltage_dependency_table *dep_table;
+	struct phm_ppt_v1_clock_voltage_dependency_table *odn_dep_table;
+	uint32_t i;
+
+	dep_table = table_info->vdd_dep_on_mclk;
+	odn_dep_table = (struct phm_ppt_v1_clock_voltage_dependency_table *)&(odn_table->vdd_dep_on_mclk);
+
+	for (i = 0; i < dep_table->count; i++) {
+		if (dep_table->entries[i].vddc != odn_dep_table->entries[i].vddc) {
+			data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_VDDC | DPMTABLE_OD_UPDATE_MCLK;
+			return;
+		}
+	}
+
+	dep_table = table_info->vdd_dep_on_sclk;
+	odn_dep_table = (struct phm_ppt_v1_clock_voltage_dependency_table *)&(odn_table->vdd_dep_on_sclk);
+	for (i = 0; i < dep_table->count; i++) {
+		if (dep_table->entries[i].vddc != odn_dep_table->entries[i].vddc) {
+			data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_VDDC | DPMTABLE_OD_UPDATE_SCLK;
+			return;
+		}
+	}
+
+	if (data->need_update_dpm_table & DPMTABLE_OD_UPDATE_VDDC) {
+		data->need_update_dpm_table &= ~DPMTABLE_OD_UPDATE_VDDC;
+		data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_SCLK | DPMTABLE_OD_UPDATE_MCLK;
+	}
+}
+
+static void vega10_odn_update_soc_table(struct pp_hwmgr *hwmgr,
+						enum PP_OD_DPM_TABLE_COMMAND type)
+{
+	struct vega10_hwmgr *data = hwmgr->backend;
+	struct phm_ppt_v2_information *table_info = hwmgr->pptable;
+	struct phm_ppt_v1_clock_voltage_dependency_table *dep_table = table_info->vdd_dep_on_socclk;
+	struct vega10_single_dpm_table *dpm_table = &data->golden_dpm_table.soc_table;
+
+	struct vega10_odn_clock_voltage_dependency_table *podn_vdd_dep_on_socclk =
+							&data->odn_dpm_table.vdd_dep_on_socclk;
+	struct vega10_odn_vddc_lookup_table *od_vddc_lookup_table = &data->odn_dpm_table.vddc_lookup_table;
+
+	struct vega10_odn_clock_voltage_dependency_table *podn_vdd_dep;
+	uint8_t i, j;
+
+	if (type == PP_OD_EDIT_SCLK_VDDC_TABLE) {
+		podn_vdd_dep = &data->odn_dpm_table.vdd_dep_on_sclk;
+		for (i = 0; i < podn_vdd_dep->count - 1; i++)
+			od_vddc_lookup_table->entries[i].us_vdd = podn_vdd_dep->entries[i].vddc;
+		if (od_vddc_lookup_table->entries[i].us_vdd < podn_vdd_dep->entries[i].vddc)
+			od_vddc_lookup_table->entries[i].us_vdd = podn_vdd_dep->entries[i].vddc;
+	} else if (type == PP_OD_EDIT_MCLK_VDDC_TABLE) {
+		podn_vdd_dep = &data->odn_dpm_table.vdd_dep_on_mclk;
+		for (i = 0; i < dpm_table->count; i++) {
+			for (j = 0; j < od_vddc_lookup_table->count; j++) {
+				if (od_vddc_lookup_table->entries[j].us_vdd >
+					podn_vdd_dep->entries[i].vddc)
+					break;
+			}
+			if (j == od_vddc_lookup_table->count) {
+				od_vddc_lookup_table->entries[j-1].us_vdd =
+					podn_vdd_dep->entries[i].vddc;
+				data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_VDDC;
+			}
+			podn_vdd_dep->entries[i].vddInd = j;
+		}
+		dpm_table = &data->dpm_table.soc_table;
+		for (i = 0; i < dep_table->count; i++) {
+			if (dep_table->entries[i].vddInd == podn_vdd_dep->entries[dep_table->count-1].vddInd &&
+					dep_table->entries[i].clk < podn_vdd_dep->entries[dep_table->count-1].clk) {
+				data->need_update_dpm_table |= DPMTABLE_UPDATE_SOCCLK;
+				podn_vdd_dep_on_socclk->entries[i].clk = podn_vdd_dep->entries[dep_table->count-1].clk;
+				dpm_table->dpm_levels[i].value = podn_vdd_dep_on_socclk->entries[i].clk;
+			}
+		}
+		if (podn_vdd_dep_on_socclk->entries[podn_vdd_dep_on_socclk->count - 1].clk <
+					podn_vdd_dep->entries[dep_table->count-1].clk) {
+			data->need_update_dpm_table |= DPMTABLE_UPDATE_SOCCLK;
+			podn_vdd_dep_on_socclk->entries[podn_vdd_dep_on_socclk->count - 1].clk = podn_vdd_dep->entries[dep_table->count-1].clk;
+			dpm_table->dpm_levels[podn_vdd_dep_on_socclk->count - 1].value = podn_vdd_dep->entries[dep_table->count-1].clk;
+		}
+		if (podn_vdd_dep_on_socclk->entries[podn_vdd_dep_on_socclk->count - 1].vddInd <
+					podn_vdd_dep->entries[dep_table->count-1].vddInd) {
+			data->need_update_dpm_table |= DPMTABLE_UPDATE_SOCCLK;
+			podn_vdd_dep_on_socclk->entries[podn_vdd_dep_on_socclk->count - 1].vddInd = podn_vdd_dep->entries[dep_table->count-1].vddInd;
+		}
+	}
+}
+
+static int vega10_odn_edit_dpm_table(struct pp_hwmgr *hwmgr,
+					enum PP_OD_DPM_TABLE_COMMAND type,
+					long *input, uint32_t size)
+{
+	struct vega10_hwmgr *data = hwmgr->backend;
+	struct vega10_odn_clock_voltage_dependency_table *podn_vdd_dep_table;
+	struct vega10_single_dpm_table *dpm_table;
+
+	uint32_t input_clk;
+	uint32_t input_vol;
+	uint32_t input_level;
+	uint32_t i;
+
+	PP_ASSERT_WITH_CODE(input, "NULL user input for clock and voltage",
+				return -EINVAL);
+
+	if (!hwmgr->od_enabled) {
+		pr_info("OverDrive feature not enabled\n");
+		return -EINVAL;
+	}
+
+	if (PP_OD_EDIT_SCLK_VDDC_TABLE == type) {
+		dpm_table = &data->dpm_table.gfx_table;
+		podn_vdd_dep_table = &data->odn_dpm_table.vdd_dep_on_sclk;
+		data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_SCLK;
+	} else if (PP_OD_EDIT_MCLK_VDDC_TABLE == type) {
+		dpm_table = &data->dpm_table.mem_table;
+		podn_vdd_dep_table = &data->odn_dpm_table.vdd_dep_on_mclk;
+		data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_MCLK;
+	} else if (PP_OD_RESTORE_DEFAULT_TABLE == type) {
+		memcpy(&(data->dpm_table), &(data->golden_dpm_table), sizeof(struct vega10_dpm_table));
+		vega10_odn_initial_default_setting(hwmgr);
+		return 0;
+	} else if (PP_OD_COMMIT_DPM_TABLE == type) {
+		vega10_check_dpm_table_updated(hwmgr);
+		return 0;
+	} else {
+		return -EINVAL;
+	}
+
+	for (i = 0; i < size; i += 3) {
+		if (i + 3 > size || input[i] >= podn_vdd_dep_table->count) {
+			pr_info("invalid clock voltage input\n");
+			return 0;
+		}
+		input_level = input[i];
+		input_clk = input[i+1] * 100;
+		input_vol = input[i+2];
+
+		if (vega10_check_clk_voltage_valid(hwmgr, type, input_clk, input_vol)) {
+			dpm_table->dpm_levels[input_level].value = input_clk;
+			podn_vdd_dep_table->entries[input_level].clk = input_clk;
+			podn_vdd_dep_table->entries[input_level].vddc = input_vol;
+		} else {
+			return -EINVAL;
+		}
+	}
+	vega10_odn_update_soc_table(hwmgr, type);
+	return 0;
+}
+
 static const struct pp_hwmgr_func vega10_hwmgr_funcs = {
 	.backend_init = vega10_hwmgr_backend_init,
 	.backend_fini = vega10_hwmgr_backend_fini,
@@ -4866,6 +4896,7 @@ static const struct pp_hwmgr_func vega10_hwmgr_funcs = {
 	.get_power_profile_mode = vega10_get_power_profile_mode,
 	.set_power_profile_mode = vega10_set_power_profile_mode,
 	.set_power_limit = vega10_set_power_limit,
+	.odn_edit_dpm_table = vega10_odn_edit_dpm_table,
 };
 
 int vega10_enable_smc_features(struct pp_hwmgr *hwmgr,
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.h
index 5339ea1..aadd6cb 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.h
@@ -282,15 +282,21 @@ struct vega10_registry_data {
 
 struct vega10_odn_clock_voltage_dependency_table {
 	uint32_t count;
-	struct phm_ppt_v1_clock_voltage_dependency_record
-		entries[MAX_REGULAR_DPM_NUMBER];
+	struct phm_ppt_v1_clock_voltage_dependency_record entries[MAX_REGULAR_DPM_NUMBER];
+};
+
+struct vega10_odn_vddc_lookup_table {
+	uint32_t count;
+	struct phm_ppt_v1_voltage_lookup_record entries[MAX_REGULAR_DPM_NUMBER];
 };
 
 struct vega10_odn_dpm_table {
-	struct phm_odn_clock_levels		odn_core_clock_dpm_levels;
-	struct phm_odn_clock_levels		odn_memory_clock_dpm_levels;
-	struct vega10_odn_clock_voltage_dependency_table		vdd_dependency_on_sclk;
-	struct vega10_odn_clock_voltage_dependency_table		vdd_dependency_on_mclk;
+	struct vega10_odn_clock_voltage_dependency_table vdd_dep_on_sclk;
+	struct vega10_odn_clock_voltage_dependency_table vdd_dep_on_mclk;
+	struct vega10_odn_clock_voltage_dependency_table vdd_dep_on_socclk;
+	struct vega10_odn_vddc_lookup_table vddc_lookup_table;
+	uint32_t max_vddc;
+	uint32_t min_vddc;
 };
 
 struct vega10_odn_fan_table {
@@ -301,8 +307,8 @@ struct vega10_odn_fan_table {
 };
 
 struct vega10_hwmgr {
-	struct vega10_dpm_table			dpm_table;
-	struct vega10_dpm_table			golden_dpm_table;
+	struct vega10_dpm_table          dpm_table;
+	struct vega10_dpm_table          golden_dpm_table;
 	struct vega10_registry_data      registry_data;
 	struct vega10_vbios_boot_state   vbios_boot_state;
 	struct vega10_mclk_latency_table mclk_latency_table;
@@ -368,12 +374,8 @@ struct vega10_hwmgr {
 	bool                           need_long_memory_training;
 
 	/* Internal settings to apply the application power optimization parameters */
-	bool                           apply_optimized_settings;
 	uint32_t                       disable_dpm_mask;
 
-	/* ---- Overdrive next setting ---- */
-	uint32_t                       apply_overdrive_next_settings_mask;
-
 	/* ---- SMU9 ---- */
 	struct smu_features            smu_features[GNLD_FEATURES_MAX];
 	struct vega10_smc_state_table  smc_state_table;
diff --git a/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h b/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h
index 8b78bbe..9bb8785 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h
@@ -377,11 +377,7 @@ struct phm_clocks {
 #define DPMTABLE_UPDATE_SCLK        0x00000004
 #define DPMTABLE_UPDATE_MCLK        0x00000008
 #define DPMTABLE_OD_UPDATE_VDDC     0x00000010
-
-/* To determine if sclk and mclk are in overdrive state */
-#define SCLK_OVERDRIVE_ENABLED           0x00000001
-#define MCLK_OVERDRIVE_ENABLED           0x00000002
-#define VDDC_OVERDRIVE_ENABLED           0x00000010
+#define DPMTABLE_UPDATE_SOCCLK      0x00000020
 
 struct phm_odn_performance_level {
 	uint32_t clock;
-- 
cgit v1.1


From 037d1a66ae640ca2723f47c0115ffa9e603699b3 Mon Sep 17 00:00:00 2001
From: Mathieu Malaterre <malat@debian.org>
Date: Tue, 24 Apr 2018 21:55:11 +0200
Subject: drm/radeon: Change the default to PCI on PowerPC
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

AGP mode is unstable on PowerPC. Symptoms are generally of the form:

[ 1228.795711] radeon 0000:00:10.0: ring 0 stalled for more than 10240msec

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Mathieu Malaterre <malat@debian.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/radeon/radeon_drv.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c
index b28288a..2a7977a 100644
--- a/drivers/gpu/drm/radeon/radeon_drv.c
+++ b/drivers/gpu/drm/radeon/radeon_drv.c
@@ -168,7 +168,12 @@ int radeon_no_wb;
 int radeon_modeset = -1;
 int radeon_dynclks = -1;
 int radeon_r4xx_atom = 0;
+#ifdef __powerpc__
+/* Default to PCI on PowerPC (fdo #95017) */
+int radeon_agpmode = -1;
+#else
 int radeon_agpmode = 0;
+#endif
 int radeon_vram_limit = 0;
 int radeon_gart_size = -1; /* auto */
 int radeon_benchmarking = 0;
-- 
cgit v1.1


From 48ff108d9dc42bf92256484c50cdb3697f5ccb04 Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Thu, 9 Nov 2017 13:18:24 -0500
Subject: drm/amdgpu: add VEGAM ASIC type

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 1 +
 include/drm/amd_asic_type.h                | 1 +
 2 files changed, 2 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 9e917f5..8ce60e6 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -83,6 +83,7 @@ static const char *amdgpu_asic_name[] = {
 	"POLARIS10",
 	"POLARIS11",
 	"POLARIS12",
+	"VEGAM",
 	"VEGA10",
 	"VEGA12",
 	"RAVEN",
diff --git a/include/drm/amd_asic_type.h b/include/drm/amd_asic_type.h
index 6c731c5..695bde7 100644
--- a/include/drm/amd_asic_type.h
+++ b/include/drm/amd_asic_type.h
@@ -44,6 +44,7 @@ enum amd_asic_type {
 	CHIP_POLARIS10,
 	CHIP_POLARIS11,
 	CHIP_POLARIS12,
+	CHIP_VEGAM,
 	CHIP_VEGA10,
 	CHIP_VEGA12,
 	CHIP_RAVEN,
-- 
cgit v1.1


From cc07f18ddb618af5ad28669dcb32b27e2f2312af Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Thu, 9 Nov 2017 13:19:58 -0500
Subject: drm/amdgpu: bypass GPU info firmware load for VEGAM

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 8ce60e6..47b65f3 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -1367,9 +1367,10 @@ static int amdgpu_device_parse_gpu_info_fw(struct amdgpu_device *adev)
 	case CHIP_TOPAZ:
 	case CHIP_TONGA:
 	case CHIP_FIJI:
-	case CHIP_POLARIS11:
 	case CHIP_POLARIS10:
+	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
+	case CHIP_VEGAM:
 	case CHIP_CARRIZO:
 	case CHIP_STONEY:
 #ifdef CONFIG_DRM_AMDGPU_SI
-- 
cgit v1.1


From 32cc7e536a546e4e2ad9ac75d02ce07d9d2327f2 Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Thu, 9 Nov 2017 13:22:54 -0500
Subject: drm/amdgpu: set VEGAM to ASIC family and ip blocks

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 47b65f3..7929ff8 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -1476,9 +1476,10 @@ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev)
 	case CHIP_TOPAZ:
 	case CHIP_TONGA:
 	case CHIP_FIJI:
-	case CHIP_POLARIS11:
 	case CHIP_POLARIS10:
+	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
+	case CHIP_VEGAM:
 	case CHIP_CARRIZO:
 	case CHIP_STONEY:
 		if (adev->asic_type == CHIP_CARRIZO || adev->asic_type == CHIP_STONEY)
-- 
cgit v1.1


From 34fd54bc0891b0d835de73978ba5277665814be3 Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Thu, 9 Nov 2017 13:26:54 -0500
Subject: drm/amdgpu: specify VEGAM ucode SMU load method

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
index 0c74c09..ee71c40 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
@@ -295,6 +295,7 @@ amdgpu_ucode_get_load_type(struct amdgpu_device *adev, int load_type)
 	case CHIP_POLARIS10:
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
+	case CHIP_VEGAM:
 		if (!load_type)
 			return AMDGPU_FW_LOAD_DIRECT;
 		else
-- 
cgit v1.1


From 5830bb986dcd6aea290ef54446e077c09cc8498e Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Thu, 9 Nov 2017 13:24:47 -0500
Subject: drm/amdgpu: add VEGAM SMU firmware support

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c       | 3 +++
 drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c | 1 +
 2 files changed, 4 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
index a8a942c..5b3d3bf 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
@@ -385,6 +385,9 @@ static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device,
 			case CHIP_POLARIS12:
 				strcpy(fw_name, "amdgpu/polaris12_smc.bin");
 				break;
+			case CHIP_VEGAM:
+				strcpy(fw_name, "amdgpu/vegam_smc.bin");
+				break;
 			case CHIP_VEGA10:
 				if ((adev->pdev->device == 0x687f) &&
 					((adev->pdev->revision == 0xc0) ||
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c
index c28b60a..ee236df 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c
@@ -41,6 +41,7 @@ MODULE_FIRMWARE("amdgpu/polaris11_smc.bin");
 MODULE_FIRMWARE("amdgpu/polaris11_smc_sk.bin");
 MODULE_FIRMWARE("amdgpu/polaris11_k_smc.bin");
 MODULE_FIRMWARE("amdgpu/polaris12_smc.bin");
+MODULE_FIRMWARE("amdgpu/vegam_smc.bin");
 MODULE_FIRMWARE("amdgpu/vega10_smc.bin");
 MODULE_FIRMWARE("amdgpu/vega10_acg_smc.bin");
 MODULE_FIRMWARE("amdgpu/vega12_smc.bin");
-- 
cgit v1.1


From be2c8cde0b867033914fc48d51b0cca0481b39b6 Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Fri, 3 Nov 2017 14:22:16 -0400
Subject: drm/amdgpu/virtual_dce: add VEGAM support

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/dce_virtual.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
index 6454cc3..de7be3d 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
@@ -460,8 +460,9 @@ static int dce_virtual_hw_init(void *handle)
 		break;
 	case CHIP_CARRIZO:
 	case CHIP_STONEY:
-	case CHIP_POLARIS11:
 	case CHIP_POLARIS10:
+	case CHIP_POLARIS11:
+	case CHIP_VEGAM:
 		dce_v11_0_disable_dce(adev);
 		break;
 	case CHIP_TOPAZ:
-- 
cgit v1.1


From 675fd32b2730f362b425a65f99fcc1eae8898fc5 Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Wed, 8 Nov 2017 18:07:12 -0500
Subject: drm/amdgpu: add VEGAM dc support check

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 7929ff8..e6657ec 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -2147,9 +2147,10 @@ bool amdgpu_device_asic_has_dc_support(enum amd_asic_type asic_type)
 	case CHIP_MULLINS:
 	case CHIP_CARRIZO:
 	case CHIP_STONEY:
-	case CHIP_POLARIS11:
 	case CHIP_POLARIS10:
+	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
+	case CHIP_VEGAM:
 	case CHIP_TONGA:
 	case CHIP_FIJI:
 	case CHIP_VEGA10:
-- 
cgit v1.1


From 589ecd753aa9e69ea40e307d2a0c013b03e418f1 Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Thu, 16 Nov 2017 13:15:12 -0500
Subject: drm/amdgpu: skip VEGAM MC firmware load

Directly loaded by VBIOS

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
index 4d970da..97fcca8 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
@@ -231,6 +231,7 @@ static int gmc_v8_0_init_microcode(struct amdgpu_device *adev)
 	case CHIP_FIJI:
 	case CHIP_CARRIZO:
 	case CHIP_STONEY:
+	case CHIP_VEGAM:
 		return 0;
 	default: BUG();
 	}
-- 
cgit v1.1


From 13b75aac5dd9a6448417769c43d21b2343ce1cc8 Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Wed, 11 Apr 2018 15:18:20 -0500
Subject: drm/amdgpu: add VEGAM GMC golden settings

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
index 97fcca8..6721b04 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
@@ -138,6 +138,7 @@ static void gmc_v8_0_init_golden_registers(struct amdgpu_device *adev)
 		break;
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
+	case CHIP_VEGAM:
 		amdgpu_device_program_register_sequence(adev,
 							golden_settings_polaris11_a11,
 							ARRAY_SIZE(golden_settings_polaris11_a11));
-- 
cgit v1.1


From f43c72ba03152920c52f1921e45100c6c090faef Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Wed, 11 Apr 2018 15:20:35 -0500
Subject: drm/amdgpu: initialize VEGAM GMC (v2)

v2: use proper register rather than hardcoding.

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
index 6721b04..1edbe6b 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
@@ -569,9 +569,10 @@ static int gmc_v8_0_mc_init(struct amdgpu_device *adev)
 	/* set the gart size */
 	if (amdgpu_gart_size == -1) {
 		switch (adev->asic_type) {
-		case CHIP_POLARIS11: /* all engines support GPUVM */
 		case CHIP_POLARIS10: /* all engines support GPUVM */
+		case CHIP_POLARIS11: /* all engines support GPUVM */
 		case CHIP_POLARIS12: /* all engines support GPUVM */
+		case CHIP_VEGAM:     /* all engines support GPUVM */
 		default:
 			adev->gmc.gart_size = 256ULL << 20;
 			break;
@@ -1091,7 +1092,8 @@ static int gmc_v8_0_sw_init(void *handle)
 	} else {
 		u32 tmp;
 
-		if (adev->asic_type == CHIP_FIJI)
+		if ((adev->asic_type == CHIP_FIJI) ||
+		    (adev->asic_type == CHIP_VEGAM))
 			tmp = RREG32(mmMC_SEQ_MISC0_FIJI);
 		else
 			tmp = RREG32(mmMC_SEQ_MISC0);
-- 
cgit v1.1


From 2267e26241d6cd0c8d92614a4a70562b009354c9 Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Thu, 9 Nov 2017 13:56:12 -0500
Subject: drm/amdgpu: add VEGAM SDMA firmware support

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c | 11 ++++++++---
 1 file changed, 8 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c
index be20a38..add0b80 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c
@@ -62,6 +62,8 @@ MODULE_FIRMWARE("amdgpu/polaris11_sdma.bin");
 MODULE_FIRMWARE("amdgpu/polaris11_sdma1.bin");
 MODULE_FIRMWARE("amdgpu/polaris12_sdma.bin");
 MODULE_FIRMWARE("amdgpu/polaris12_sdma1.bin");
+MODULE_FIRMWARE("amdgpu/vegam_sdma.bin");
+MODULE_FIRMWARE("amdgpu/vegam_sdma1.bin");
 
 
 static const u32 sdma_offsets[SDMA_MAX_INSTANCE] =
@@ -275,15 +277,18 @@ static int sdma_v3_0_init_microcode(struct amdgpu_device *adev)
 	case CHIP_FIJI:
 		chip_name = "fiji";
 		break;
-	case CHIP_POLARIS11:
-		chip_name = "polaris11";
-		break;
 	case CHIP_POLARIS10:
 		chip_name = "polaris10";
 		break;
+	case CHIP_POLARIS11:
+		chip_name = "polaris11";
+		break;
 	case CHIP_POLARIS12:
 		chip_name = "polaris12";
 		break;
+	case CHIP_VEGAM:
+		chip_name = "vegam";
+		break;
 	case CHIP_CARRIZO:
 		chip_name = "carrizo";
 		break;
-- 
cgit v1.1


From c3f27c08ec15b61bed2a1af592ab5bdc89fe7dee Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Wed, 11 Apr 2018 15:22:20 -0500
Subject: drm/amdgpu: add VEGAM SDMA golden settings

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c
index add0b80..aa9ab29 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c
@@ -211,6 +211,7 @@ static void sdma_v3_0_init_golden_registers(struct amdgpu_device *adev)
 		break;
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
+	case CHIP_VEGAM:
 		amdgpu_device_program_register_sequence(adev,
 							golden_settings_polaris11_a11,
 							ARRAY_SIZE(golden_settings_polaris11_a11));
-- 
cgit v1.1


From 62aac2010de1d233739c30a168d0bbff31b3cb43 Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Fri, 10 Nov 2017 11:04:09 -0500
Subject: drm/amdgpu: add VEGAM GFX firmware support

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c | 42 ++++++++++++++++++++++-------------
 1 file changed, 26 insertions(+), 16 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
index e14263f..2be2870 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
@@ -125,18 +125,6 @@ MODULE_FIRMWARE("amdgpu/fiji_mec.bin");
 MODULE_FIRMWARE("amdgpu/fiji_mec2.bin");
 MODULE_FIRMWARE("amdgpu/fiji_rlc.bin");
 
-MODULE_FIRMWARE("amdgpu/polaris11_ce.bin");
-MODULE_FIRMWARE("amdgpu/polaris11_ce_2.bin");
-MODULE_FIRMWARE("amdgpu/polaris11_pfp.bin");
-MODULE_FIRMWARE("amdgpu/polaris11_pfp_2.bin");
-MODULE_FIRMWARE("amdgpu/polaris11_me.bin");
-MODULE_FIRMWARE("amdgpu/polaris11_me_2.bin");
-MODULE_FIRMWARE("amdgpu/polaris11_mec.bin");
-MODULE_FIRMWARE("amdgpu/polaris11_mec_2.bin");
-MODULE_FIRMWARE("amdgpu/polaris11_mec2.bin");
-MODULE_FIRMWARE("amdgpu/polaris11_mec2_2.bin");
-MODULE_FIRMWARE("amdgpu/polaris11_rlc.bin");
-
 MODULE_FIRMWARE("amdgpu/polaris10_ce.bin");
 MODULE_FIRMWARE("amdgpu/polaris10_ce_2.bin");
 MODULE_FIRMWARE("amdgpu/polaris10_pfp.bin");
@@ -149,6 +137,18 @@ MODULE_FIRMWARE("amdgpu/polaris10_mec2.bin");
 MODULE_FIRMWARE("amdgpu/polaris10_mec2_2.bin");
 MODULE_FIRMWARE("amdgpu/polaris10_rlc.bin");
 
+MODULE_FIRMWARE("amdgpu/polaris11_ce.bin");
+MODULE_FIRMWARE("amdgpu/polaris11_ce_2.bin");
+MODULE_FIRMWARE("amdgpu/polaris11_pfp.bin");
+MODULE_FIRMWARE("amdgpu/polaris11_pfp_2.bin");
+MODULE_FIRMWARE("amdgpu/polaris11_me.bin");
+MODULE_FIRMWARE("amdgpu/polaris11_me_2.bin");
+MODULE_FIRMWARE("amdgpu/polaris11_mec.bin");
+MODULE_FIRMWARE("amdgpu/polaris11_mec_2.bin");
+MODULE_FIRMWARE("amdgpu/polaris11_mec2.bin");
+MODULE_FIRMWARE("amdgpu/polaris11_mec2_2.bin");
+MODULE_FIRMWARE("amdgpu/polaris11_rlc.bin");
+
 MODULE_FIRMWARE("amdgpu/polaris12_ce.bin");
 MODULE_FIRMWARE("amdgpu/polaris12_ce_2.bin");
 MODULE_FIRMWARE("amdgpu/polaris12_pfp.bin");
@@ -161,6 +161,13 @@ MODULE_FIRMWARE("amdgpu/polaris12_mec2.bin");
 MODULE_FIRMWARE("amdgpu/polaris12_mec2_2.bin");
 MODULE_FIRMWARE("amdgpu/polaris12_rlc.bin");
 
+MODULE_FIRMWARE("amdgpu/vegam_ce.bin");
+MODULE_FIRMWARE("amdgpu/vegam_pfp.bin");
+MODULE_FIRMWARE("amdgpu/vegam_me.bin");
+MODULE_FIRMWARE("amdgpu/vegam_mec.bin");
+MODULE_FIRMWARE("amdgpu/vegam_mec2.bin");
+MODULE_FIRMWARE("amdgpu/vegam_rlc.bin");
+
 static const struct amdgpu_gds_reg_offset amdgpu_gds_reg_offset[] =
 {
 	{mmGDS_VMID0_BASE, mmGDS_VMID0_SIZE, mmGDS_GWS_VMID0, mmGDS_OA_VMID0},
@@ -918,17 +925,20 @@ static int gfx_v8_0_init_microcode(struct amdgpu_device *adev)
 	case CHIP_FIJI:
 		chip_name = "fiji";
 		break;
-	case CHIP_POLARIS11:
-		chip_name = "polaris11";
+	case CHIP_STONEY:
+		chip_name = "stoney";
 		break;
 	case CHIP_POLARIS10:
 		chip_name = "polaris10";
 		break;
+	case CHIP_POLARIS11:
+		chip_name = "polaris11";
+		break;
 	case CHIP_POLARIS12:
 		chip_name = "polaris12";
 		break;
-	case CHIP_STONEY:
-		chip_name = "stoney";
+	case CHIP_VEGAM:
+		chip_name = "vegam";
 		break;
 	default:
 		BUG();
-- 
cgit v1.1


From aefbbd6cc55cba823fecd0231116a8e1073e4892 Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Thu, 16 Nov 2017 13:41:03 -0500
Subject: drm/amdgpu: add VEGAM GFX golden settings

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c | 39 +++++++++++++++++++++++++++++++++++
 1 file changed, 39 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
index 2be2870..d789723 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
@@ -299,6 +299,37 @@ static const u32 tonga_mgcg_cgcg_init[] =
 	mmCP_MEM_SLP_CNTL, 0x00000001, 0x00000001,
 };
 
+static const u32 golden_settings_vegam_a11[] =
+{
+	mmCB_HW_CONTROL, 0x0001f3cf, 0x00007208,
+	mmCB_HW_CONTROL_2, 0x0f000000, 0x0d000000,
+	mmCB_HW_CONTROL_3, 0x000001ff, 0x00000040,
+	mmDB_DEBUG2, 0xf00fffff, 0x00000400,
+	mmPA_SC_ENHANCE, 0xffffffff, 0x20000001,
+	mmPA_SC_LINE_STIPPLE_STATE, 0x0000ff0f, 0x00000000,
+	mmPA_SC_RASTER_CONFIG, 0x3f3fffff, 0x3a00161a,
+	mmPA_SC_RASTER_CONFIG_1, 0x0000003f, 0x0000002e,
+	mmRLC_CGCG_CGLS_CTRL, 0x00000003, 0x0001003c,
+	mmRLC_CGCG_CGLS_CTRL_3D, 0xffffffff, 0x0001003c,
+	mmSQ_CONFIG, 0x07f80000, 0x01180000,
+	mmTA_CNTL_AUX, 0x000f000f, 0x000b0000,
+	mmTCC_CTRL, 0x00100000, 0xf31fff7f,
+	mmTCP_ADDR_CONFIG, 0x000003ff, 0x000000f7,
+	mmTCP_CHAN_STEER_HI, 0xffffffff, 0x00000000,
+	mmTCP_CHAN_STEER_LO, 0xffffffff, 0x32761054,
+	mmVGT_RESET_DEBUG, 0x00000004, 0x00000004,
+};
+
+static const u32 vegam_golden_common_all[] =
+{
+	mmGRBM_GFX_INDEX, 0xffffffff, 0xe0000000,
+	mmGB_ADDR_CONFIG, 0xffffffff, 0x22011003,
+	mmSPI_RESOURCE_RESERVE_CU_0, 0xffffffff, 0x00000800,
+	mmSPI_RESOURCE_RESERVE_CU_1, 0xffffffff, 0x00000800,
+	mmSPI_RESOURCE_RESERVE_EN_CU_0, 0xffffffff, 0x00FF7FBF,
+	mmSPI_RESOURCE_RESERVE_EN_CU_1, 0xffffffff, 0x00FF7FAF,
+};
+
 static const u32 golden_settings_polaris11_a11[] =
 {
 	mmCB_HW_CONTROL, 0x0000f3cf, 0x00007208,
@@ -719,6 +750,14 @@ static void gfx_v8_0_init_golden_registers(struct amdgpu_device *adev)
 							tonga_golden_common_all,
 							ARRAY_SIZE(tonga_golden_common_all));
 		break;
+	case CHIP_VEGAM:
+		amdgpu_device_program_register_sequence(adev,
+							golden_settings_vegam_a11,
+							ARRAY_SIZE(golden_settings_vegam_a11));
+		amdgpu_device_program_register_sequence(adev,
+							vegam_golden_common_all,
+							ARRAY_SIZE(vegam_golden_common_all));
+		break;
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
 		amdgpu_device_program_register_sequence(adev,
-- 
cgit v1.1


From 7176546958ddd7d4732d2a19692a3b14e3519caa Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Thu, 16 Nov 2017 13:49:56 -0500
Subject: drm/amdgpu: initialize VEGAM GFX

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c | 21 +++++++++++++++------
 1 file changed, 15 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
index d789723..818874b 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
@@ -1819,6 +1819,7 @@ static int gfx_v8_0_gpu_early_init(struct amdgpu_device *adev)
 		gb_addr_config = POLARIS11_GB_ADDR_CONFIG_GOLDEN;
 		break;
 	case CHIP_POLARIS10:
+	case CHIP_VEGAM:
 		ret = amdgpu_atombios_get_gfx_info(adev);
 		if (ret)
 			return ret;
@@ -2006,12 +2007,13 @@ static int gfx_v8_0_sw_init(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
 	switch (adev->asic_type) {
-	case CHIP_FIJI:
 	case CHIP_TONGA:
+	case CHIP_CARRIZO:
+	case CHIP_FIJI:
+	case CHIP_POLARIS10:
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
-	case CHIP_POLARIS10:
-	case CHIP_CARRIZO:
+	case CHIP_VEGAM:
 		adev->gfx.mec.num_mec = 2;
 		break;
 	case CHIP_TOPAZ:
@@ -2372,6 +2374,7 @@ static void gfx_v8_0_tiling_mode_table_init(struct amdgpu_device *adev)
 
 		break;
 	case CHIP_FIJI:
+	case CHIP_VEGAM:
 		modearray[0] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
 				PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
 				TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
@@ -3553,6 +3556,7 @@ gfx_v8_0_raster_config(struct amdgpu_device *adev, u32 *rconf, u32 *rconf1)
 {
 	switch (adev->asic_type) {
 	case CHIP_FIJI:
+	case CHIP_VEGAM:
 		*rconf |= RB_MAP_PKR0(2) | RB_MAP_PKR1(2) |
 			  RB_XSEL2(1) | PKR_MAP(2) |
 			  PKR_XSEL(1) | PKR_YSEL(1) |
@@ -4120,7 +4124,8 @@ static void gfx_v8_0_init_pg(struct amdgpu_device *adev)
 		gfx_v8_0_init_power_gating(adev);
 		WREG32(mmRLC_PG_ALWAYS_ON_CU_MASK, adev->gfx.cu_info.ao_cu_mask);
 	} else if ((adev->asic_type == CHIP_POLARIS11) ||
-		   (adev->asic_type == CHIP_POLARIS12)) {
+		   (adev->asic_type == CHIP_POLARIS12) ||
+		   (adev->asic_type == CHIP_VEGAM)) {
 		gfx_v8_0_init_csb(adev);
 		gfx_v8_0_init_save_restore_list(adev);
 		gfx_v8_0_enable_save_restore_machine(adev);
@@ -4195,7 +4200,8 @@ static int gfx_v8_0_rlc_resume(struct amdgpu_device *adev)
 	WREG32(mmRLC_CGCG_CGLS_CTRL, tmp);
 	if (adev->asic_type == CHIP_POLARIS11 ||
 	    adev->asic_type == CHIP_POLARIS10 ||
-	    adev->asic_type == CHIP_POLARIS12) {
+	    adev->asic_type == CHIP_POLARIS12 ||
+	    adev->asic_type == CHIP_VEGAM) {
 		tmp = RREG32(mmRLC_CGCG_CGLS_CTRL_3D);
 		tmp &= ~0x3;
 		WREG32(mmRLC_CGCG_CGLS_CTRL_3D, tmp);
@@ -5547,7 +5553,8 @@ static void gfx_v8_0_enable_gfx_static_mg_power_gating(struct amdgpu_device *ade
 						       bool enable)
 {
 	if ((adev->asic_type == CHIP_POLARIS11) ||
-	    (adev->asic_type == CHIP_POLARIS12))
+	    (adev->asic_type == CHIP_POLARIS12) ||
+	    (adev->asic_type == CHIP_VEGAM))
 		/* Send msg to SMU via Powerplay */
 		amdgpu_device_ip_set_powergating_state(adev,
 						       AMD_IP_BLOCK_TYPE_SMC,
@@ -5637,6 +5644,7 @@ static int gfx_v8_0_set_powergating_state(void *handle,
 		break;
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
+	case CHIP_VEGAM:
 		if ((adev->pg_flags & AMD_PG_SUPPORT_GFX_SMG) && enable)
 			gfx_v8_0_enable_gfx_static_mg_power_gating(adev, true);
 		else
@@ -6203,6 +6211,7 @@ static int gfx_v8_0_set_clockgating_state(void *handle,
 	case CHIP_POLARIS10:
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
+	case CHIP_VEGAM:
 		gfx_v8_0_polaris_update_gfx_clock_gating(adev, state);
 		break;
 	default:
-- 
cgit v1.1


From ba8f7ad0e5b25851299cd45a63b57d843e50b577 Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Fri, 10 Nov 2017 12:27:40 -0500
Subject: drm/amdgpu: add VEGAM UVD firmware support

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
index 627542b..d8dd402 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
@@ -66,6 +66,7 @@
 #define FIRMWARE_POLARIS10	"amdgpu/polaris10_uvd.bin"
 #define FIRMWARE_POLARIS11	"amdgpu/polaris11_uvd.bin"
 #define FIRMWARE_POLARIS12	"amdgpu/polaris12_uvd.bin"
+#define FIRMWARE_VEGAM		"amdgpu/vegam_uvd.bin"
 
 #define FIRMWARE_VEGA10		"amdgpu/vega10_uvd.bin"
 #define FIRMWARE_VEGA12		"amdgpu/vega12_uvd.bin"
@@ -109,6 +110,7 @@ MODULE_FIRMWARE(FIRMWARE_STONEY);
 MODULE_FIRMWARE(FIRMWARE_POLARIS10);
 MODULE_FIRMWARE(FIRMWARE_POLARIS11);
 MODULE_FIRMWARE(FIRMWARE_POLARIS12);
+MODULE_FIRMWARE(FIRMWARE_VEGAM);
 
 MODULE_FIRMWARE(FIRMWARE_VEGA10);
 MODULE_FIRMWARE(FIRMWARE_VEGA12);
@@ -172,6 +174,9 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
 	case CHIP_VEGA12:
 		fw_name = FIRMWARE_VEGA12;
 		break;
+	case CHIP_VEGAM:
+		fw_name = FIRMWARE_VEGAM;
+		break;
 	default:
 		return -EINVAL;
 	}
-- 
cgit v1.1


From 136b10ad9b515a7ffdfbf4df01941856682bf94e Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Wed, 11 Apr 2018 15:24:01 -0500
Subject: drm/amdgpu: add VEGAM UVD encode support

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
index f26f515..6d33598 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
@@ -62,7 +62,7 @@ static void uvd_v6_0_enable_mgcg(struct amdgpu_device *adev,
 static inline bool uvd_v6_0_enc_support(struct amdgpu_device *adev)
 {
 	return ((adev->asic_type >= CHIP_POLARIS10) &&
-			(adev->asic_type <= CHIP_POLARIS12) &&
+			(adev->asic_type <= CHIP_VEGAM) &&
 			(!adev->uvd.fw_version || adev->uvd.fw_version >= FW_1_130_16));
 }
 
-- 
cgit v1.1


From f11ded5ec23602d651cab3381243c527ad8c55a9 Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Wed, 11 Apr 2018 15:25:57 -0500
Subject: drm/amdgpu: add VEGAM VCE firmware support

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
index d7261e0..e2186ed 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
@@ -51,8 +51,9 @@
 #define FIRMWARE_FIJI		"amdgpu/fiji_vce.bin"
 #define FIRMWARE_STONEY		"amdgpu/stoney_vce.bin"
 #define FIRMWARE_POLARIS10	"amdgpu/polaris10_vce.bin"
-#define FIRMWARE_POLARIS11         "amdgpu/polaris11_vce.bin"
-#define FIRMWARE_POLARIS12         "amdgpu/polaris12_vce.bin"
+#define FIRMWARE_POLARIS11	"amdgpu/polaris11_vce.bin"
+#define FIRMWARE_POLARIS12	"amdgpu/polaris12_vce.bin"
+#define FIRMWARE_VEGAM		"amdgpu/vegam_vce.bin"
 
 #define FIRMWARE_VEGA10		"amdgpu/vega10_vce.bin"
 #define FIRMWARE_VEGA12		"amdgpu/vega12_vce.bin"
@@ -71,6 +72,7 @@ MODULE_FIRMWARE(FIRMWARE_STONEY);
 MODULE_FIRMWARE(FIRMWARE_POLARIS10);
 MODULE_FIRMWARE(FIRMWARE_POLARIS11);
 MODULE_FIRMWARE(FIRMWARE_POLARIS12);
+MODULE_FIRMWARE(FIRMWARE_VEGAM);
 
 MODULE_FIRMWARE(FIRMWARE_VEGA10);
 MODULE_FIRMWARE(FIRMWARE_VEGA12);
@@ -132,6 +134,9 @@ int amdgpu_vce_sw_init(struct amdgpu_device *adev, unsigned long size)
 	case CHIP_POLARIS12:
 		fw_name = FIRMWARE_POLARIS12;
 		break;
+	case CHIP_VEGAM:
+		fw_name = FIRMWARE_VEGAM;
+		break;
 	case CHIP_VEGA10:
 		fw_name = FIRMWARE_VEGA10;
 		break;
-- 
cgit v1.1


From a771289786824f15d4d4307242389d0499e83e59 Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Fri, 10 Nov 2017 12:32:04 -0500
Subject: drm/amdgpu: add VEGAM to VCE harvest config

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/vce_v3_0.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
index ac96172..0999c84 100644
--- a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
@@ -388,7 +388,8 @@ static unsigned vce_v3_0_get_harvest_config(struct amdgpu_device *adev)
 	default:
 		if ((adev->asic_type == CHIP_POLARIS10) ||
 		    (adev->asic_type == CHIP_POLARIS11) ||
-		    (adev->asic_type == CHIP_POLARIS12))
+		    (adev->asic_type == CHIP_POLARIS12) ||
+		    (adev->asic_type == CHIP_VEGAM))
 			return AMDGPU_VCE_HARVEST_VCE1;
 
 		return 0;
-- 
cgit v1.1


From b51c5194a5b8d781e45a86776f2eec234f7567fe Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Wed, 11 Apr 2018 15:28:28 -0500
Subject: drm/amdgpu: add VEGAM support to vi

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/vi.c | 31 +++++++++++++++++++++++++++++--
 1 file changed, 29 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c
index 4034a28..4ac1288 100644
--- a/drivers/gpu/drm/amd/amdgpu/vi.c
+++ b/drivers/gpu/drm/amd/amdgpu/vi.c
@@ -305,9 +305,10 @@ static void vi_init_golden_registers(struct amdgpu_device *adev)
 							stoney_mgcg_cgcg_init,
 							ARRAY_SIZE(stoney_mgcg_cgcg_init));
 		break;
-	case CHIP_POLARIS11:
 	case CHIP_POLARIS10:
+	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
+	case CHIP_VEGAM:
 	default:
 		break;
 	}
@@ -1096,6 +1097,30 @@ static int vi_common_early_init(void *handle)
 		adev->pg_flags = 0;
 		adev->external_rev_id = adev->rev_id + 0x64;
 		break;
+	case CHIP_VEGAM:
+		adev->cg_flags = 0;
+			/*AMD_CG_SUPPORT_GFX_MGCG |
+			AMD_CG_SUPPORT_GFX_RLC_LS |
+			AMD_CG_SUPPORT_GFX_CP_LS |
+			AMD_CG_SUPPORT_GFX_CGCG |
+			AMD_CG_SUPPORT_GFX_CGLS |
+			AMD_CG_SUPPORT_GFX_3D_CGCG |
+			AMD_CG_SUPPORT_GFX_3D_CGLS |
+			AMD_CG_SUPPORT_SDMA_MGCG |
+			AMD_CG_SUPPORT_SDMA_LS |
+			AMD_CG_SUPPORT_BIF_MGCG |
+			AMD_CG_SUPPORT_BIF_LS |
+			AMD_CG_SUPPORT_HDP_MGCG |
+			AMD_CG_SUPPORT_HDP_LS |
+			AMD_CG_SUPPORT_ROM_MGCG |
+			AMD_CG_SUPPORT_MC_MGCG |
+			AMD_CG_SUPPORT_MC_LS |
+			AMD_CG_SUPPORT_DRM_LS |
+			AMD_CG_SUPPORT_UVD_MGCG |
+			AMD_CG_SUPPORT_VCE_MGCG;*/
+		adev->pg_flags = 0;
+		adev->external_rev_id = adev->rev_id + 0x6E;
+		break;
 	case CHIP_CARRIZO:
 		adev->cg_flags = AMD_CG_SUPPORT_UVD_MGCG |
 			AMD_CG_SUPPORT_GFX_MGCG |
@@ -1487,6 +1512,7 @@ static int vi_common_set_clockgating_state(void *handle,
 	case CHIP_POLARIS10:
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
+	case CHIP_VEGAM:
 		vi_common_set_clockgating_state_by_smu(adev, state);
 	default:
 		break;
@@ -1616,9 +1642,10 @@ int vi_set_ip_blocks(struct amdgpu_device *adev)
 			amdgpu_device_ip_block_add(adev, &vce_v3_0_ip_block);
 		}
 		break;
-	case CHIP_POLARIS11:
 	case CHIP_POLARIS10:
+	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
+	case CHIP_VEGAM:
 		amdgpu_device_ip_block_add(adev, &vi_common_ip_block);
 		amdgpu_device_ip_block_add(adev, &gmc_v8_1_ip_block);
 		amdgpu_device_ip_block_add(adev, &tonga_ih_ip_block);
-- 
cgit v1.1


From e930793280799e66c3197e2ee6e70b1129f8aa12 Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Thu, 9 Nov 2017 13:25:31 -0500
Subject: drm/amdgpu: add VEGAM pci ids

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
index 998ba8e..739e7e0 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
@@ -541,6 +541,9 @@ static const struct pci_device_id pciidlist[] = {
 	{0x1002, 0x6995, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_POLARIS12},
 	{0x1002, 0x6997, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_POLARIS12},
 	{0x1002, 0x699F, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_POLARIS12},
+	/* VEGAM */
+	{0x1002, 0x694C, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGAM},
+	{0x1002, 0x694E, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGAM},
 	/* Vega 10 */
 	{0x1002, 0x6860, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10},
 	{0x1002, 0x6861, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10},
-- 
cgit v1.1


From 0c75d5acc80dc5247962370c9f555922197b1ec3 Mon Sep 17 00:00:00 2001
From: "Jerry (Fangzhi) Zuo" <Jerry.Zuo@amd.com>
Date: Wed, 11 Apr 2018 15:39:35 -0500
Subject: drm/amd/display: Implement VEGAM device IDs in DC

Implement device IDs for VEGAM

Signed-off-by: Jerry (Fangzhi) Zuo <Jerry.Zuo@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/Kconfig                         | 6 ++++++
 drivers/gpu/drm/amd/display/dc/bios/command_table_helper.c  | 3 +++
 drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.c | 3 +++
 drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c            | 9 +++++++++
 drivers/gpu/drm/amd/display/dc/core/dc_resource.c           | 7 +++++++
 drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c       | 6 ++++++
 drivers/gpu/drm/amd/display/dc/gpio/hw_factory.c            | 3 +++
 drivers/gpu/drm/amd/display/dc/gpio/hw_translate.c          | 3 +++
 drivers/gpu/drm/amd/display/dc/i2caux/i2caux.c              | 3 +++
 drivers/gpu/drm/amd/display/dc/inc/dce_calcs.h              | 3 +++
 drivers/gpu/drm/amd/display/include/dal_asic_id.h           | 7 +++++++
 drivers/gpu/drm/amd/display/include/dal_types.h             | 3 +++
 12 files changed, 56 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/Kconfig b/drivers/gpu/drm/amd/display/Kconfig
index d5d4586..e6ca72c 100644
--- a/drivers/gpu/drm/amd/display/Kconfig
+++ b/drivers/gpu/drm/amd/display/Kconfig
@@ -34,4 +34,10 @@ config DEBUG_KERNEL_DC
 	  if you want to hit
 	  kdgb_break in assert.
 
+config DRM_AMD_DC_VEGAM
+        bool "VEGAM support"
+        depends on DRM_AMD_DC
+        help
+         Choose this option if you want to have
+         VEGAM support for display engine
 endmenu
diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper.c b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper.c
index 2979358..be066c4 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper.c
@@ -51,6 +51,9 @@ bool dal_bios_parser_init_cmd_tbl_helper(
 		return true;
 
 	case DCE_VERSION_11_2:
+#if defined(CONFIG_DRM_AMD_DC_VEGAM)
+	case DCE_VERSION_11_22:
+#endif
 		*h = dal_cmd_tbl_helper_dce112_get_table();
 		return true;
 
diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.c b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.c
index 9a4d30d..9b9e069 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.c
@@ -52,6 +52,9 @@ bool dal_bios_parser_init_cmd_tbl_helper2(
 		return true;
 
 	case DCE_VERSION_11_2:
+#if defined(CONFIG_DRM_AMD_DC_VEGAM)
+	case DCE_VERSION_11_22:
+#endif
 		*h = dal_cmd_tbl_helper_dce112_get_table2();
 		return true;
 #if defined(CONFIG_DRM_AMD_DC_DCN1_0)
diff --git a/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c b/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c
index 56f46a0..4ee3c26 100644
--- a/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c
+++ b/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c
@@ -59,6 +59,10 @@ static enum bw_calcs_version bw_calcs_version_from_asic_id(struct hw_asic_id asi
 			return BW_CALCS_VERSION_POLARIS10;
 		if (ASIC_REV_IS_POLARIS11_M(asic_id.hw_internal_rev))
 			return BW_CALCS_VERSION_POLARIS11;
+#if defined(CONFIG_DRM_AMD_DC_VEGAM)
+		if (ASIC_REV_IS_VEGAM(asic_id.hw_internal_rev))
+			return BW_CALCS_VERSION_VEGAM;
+#endif
 		return BW_CALCS_VERSION_INVALID;
 
 	case FAMILY_AI:
@@ -2147,6 +2151,11 @@ void bw_calcs_init(struct bw_calcs_dceip *bw_dceip,
 		dceip.mcifwr_all_surfaces_burst_time = bw_int_to_fixed(0); /* todo: this is a bug*/
 		break;
 	case BW_CALCS_VERSION_POLARIS10:
+#if defined(CONFIG_DRM_AMD_DC_VEGAM)
+		/* TODO: Treat VEGAM the same as P10 for now
+		 * Need to tune the para for VEGAM if needed */
+	case BW_CALCS_VERSION_VEGAM:
+#endif
 		vbios.memory_type = bw_def_gddr5;
 		vbios.dram_channel_width_in_bits = 32;
 		vbios.number_of_dram_channels = asic_id.vram_width / vbios.dram_channel_width_in_bits;
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
index d7a92ec..447729c 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
@@ -79,6 +79,10 @@ enum dce_version resource_parse_asic_id(struct hw_asic_id asic_id)
 				ASIC_REV_IS_POLARIS12_V(asic_id.hw_internal_rev)) {
 			dc_version = DCE_VERSION_11_2;
 		}
+#if defined(CONFIG_DRM_AMD_DC_VEGAM)
+		if (ASIC_REV_IS_VEGAM(asic_id.hw_internal_rev))
+			dc_version = DCE_VERSION_11_22;
+#endif
 		break;
 	case FAMILY_AI:
 		dc_version = DCE_VERSION_12_0;
@@ -125,6 +129,9 @@ struct resource_pool *dc_create_resource_pool(
 			num_virtual_links, dc, asic_id);
 		break;
 	case DCE_VERSION_11_2:
+#if defined(CONFIG_DRM_AMD_DC_VEGAM)
+	case DCE_VERSION_11_22:
+#endif
 		res_pool = dce112_create_resource_pool(
 			num_virtual_links, dc);
 		break;
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c
index 67dad7f..223db98 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c
@@ -590,6 +590,9 @@ static uint32_t dce110_get_pix_clk_dividers(
 			pll_settings, pix_clk_params);
 		break;
 	case DCE_VERSION_11_2:
+#if defined(CONFIG_DRM_AMD_DC_VEGAM)
+	case DCE_VERSION_11_22:
+#endif
 	case DCE_VERSION_12_0:
 #if defined(CONFIG_DRM_AMD_DC_DCN1_0)
 	case DCN_VERSION_1_0:
@@ -979,6 +982,9 @@ static bool dce110_program_pix_clk(
 
 		break;
 	case DCE_VERSION_11_2:
+#if defined(CONFIG_DRM_AMD_DC_VEGAM)
+	case DCE_VERSION_11_22:
+#endif
 	case DCE_VERSION_12_0:
 #if defined(CONFIG_DRM_AMD_DC_DCN1_0)
 	case DCN_VERSION_1_0:
diff --git a/drivers/gpu/drm/amd/display/dc/gpio/hw_factory.c b/drivers/gpu/drm/amd/display/dc/gpio/hw_factory.c
index 87b580f..61fe484 100644
--- a/drivers/gpu/drm/amd/display/dc/gpio/hw_factory.c
+++ b/drivers/gpu/drm/amd/display/dc/gpio/hw_factory.c
@@ -75,6 +75,9 @@ bool dal_hw_factory_init(
 		return true;
 	case DCE_VERSION_11_0:
 	case DCE_VERSION_11_2:
+#if defined(CONFIG_DRM_AMD_DC_VEGAM)
+	case DCE_VERSION_11_22:
+#endif
 		dal_hw_factory_dce110_init(factory);
 		return true;
 	case DCE_VERSION_12_0:
diff --git a/drivers/gpu/drm/amd/display/dc/gpio/hw_translate.c b/drivers/gpu/drm/amd/display/dc/gpio/hw_translate.c
index 0ae8ace..910ae2b7 100644
--- a/drivers/gpu/drm/amd/display/dc/gpio/hw_translate.c
+++ b/drivers/gpu/drm/amd/display/dc/gpio/hw_translate.c
@@ -72,6 +72,9 @@ bool dal_hw_translate_init(
 	case DCE_VERSION_10_0:
 	case DCE_VERSION_11_0:
 	case DCE_VERSION_11_2:
+#if defined(CONFIG_DRM_AMD_DC_VEGAM)
+	case DCE_VERSION_11_22:
+#endif
 		dal_hw_translate_dce110_init(translate);
 		return true;
 	case DCE_VERSION_12_0:
diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/i2caux.c b/drivers/gpu/drm/amd/display/dc/i2caux/i2caux.c
index 5cbf662..c3d7c32 100644
--- a/drivers/gpu/drm/amd/display/dc/i2caux/i2caux.c
+++ b/drivers/gpu/drm/amd/display/dc/i2caux/i2caux.c
@@ -83,6 +83,9 @@ struct i2caux *dal_i2caux_create(
 	case DCE_VERSION_8_3:
 		return dal_i2caux_dce80_create(ctx);
 	case DCE_VERSION_11_2:
+#if defined(CONFIG_DRM_AMD_DC_VEGAM)
+	case DCE_VERSION_11_22:
+#endif
 		return dal_i2caux_dce112_create(ctx);
 	case DCE_VERSION_11_0:
 		return dal_i2caux_dce110_create(ctx);
diff --git a/drivers/gpu/drm/amd/display/dc/inc/dce_calcs.h b/drivers/gpu/drm/amd/display/dc/inc/dce_calcs.h
index 0bd87f2..933ea7a 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/dce_calcs.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/dce_calcs.h
@@ -43,6 +43,9 @@ enum bw_calcs_version {
 	BW_CALCS_VERSION_POLARIS10,
 	BW_CALCS_VERSION_POLARIS11,
 	BW_CALCS_VERSION_POLARIS12,
+#if defined(CONFIG_DRM_AMD_DC_VEGAM)
+	BW_CALCS_VERSION_VEGAM,
+#endif
 	BW_CALCS_VERSION_STONEY,
 	BW_CALCS_VERSION_VEGA10
 };
diff --git a/drivers/gpu/drm/amd/display/include/dal_asic_id.h b/drivers/gpu/drm/amd/display/include/dal_asic_id.h
index 9831cb5..3e8e535 100644
--- a/drivers/gpu/drm/amd/display/include/dal_asic_id.h
+++ b/drivers/gpu/drm/amd/display/include/dal_asic_id.h
@@ -98,7 +98,14 @@
 		(eChipRev < VI_POLARIS11_M_A0))
 #define ASIC_REV_IS_POLARIS11_M(eChipRev) ((eChipRev >= VI_POLARIS11_M_A0) &&  \
 		(eChipRev < VI_POLARIS12_V_A0))
+#if defined(CONFIG_DRM_AMD_DC_VEGAM)
+#define VI_VEGAM_A0 110
+#define ASIC_REV_IS_POLARIS12_V(eChipRev) ((eChipRev >= VI_POLARIS12_V_A0) && \
+		(eChipRev < VI_VEGAM_A0))
+#define ASIC_REV_IS_VEGAM(eChipRev) (eChipRev >= VI_VEGAM_A0)
+#else
 #define ASIC_REV_IS_POLARIS12_V(eChipRev) (eChipRev >= VI_POLARIS12_V_A0)
+#endif
 
 /* DCE11 */
 #define CZ_CARRIZO_A0 0x01
diff --git a/drivers/gpu/drm/amd/display/include/dal_types.h b/drivers/gpu/drm/amd/display/include/dal_types.h
index fa54396..5b1f8ce 100644
--- a/drivers/gpu/drm/amd/display/include/dal_types.h
+++ b/drivers/gpu/drm/amd/display/include/dal_types.h
@@ -40,6 +40,9 @@ enum dce_version {
 	DCE_VERSION_10_0,
 	DCE_VERSION_11_0,
 	DCE_VERSION_11_2,
+#if defined(CONFIG_DRM_AMD_DC_VEGAM)
+	DCE_VERSION_11_22,
+#endif
 	DCE_VERSION_12_0,
 	DCE_VERSION_MAX,
 	DCN_VERSION_1_0,
-- 
cgit v1.1


From 7737de91633b1cd6b3a0b15347a633667a9bc2fc Mon Sep 17 00:00:00 2001
From: "Jerry (Fangzhi) Zuo" <Jerry.Zuo@amd.com>
Date: Thu, 9 Nov 2017 11:51:13 -0500
Subject: drm/amd/display: Implement VEGAM device IDs in DM

Add CHIP_VEGAM

Signed-off-by: Jerry (Fangzhi) Zuo <Jerry.Zuo@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 656a018..8379a37 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -1524,6 +1524,9 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS10:
 	case CHIP_POLARIS12:
+#if defined(CONFIG_DRM_AMD_DC_VEGAM)
+	case CHIP_VEGAM:
+#endif
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
 		if (dce110_register_irq_handlers(dm->adev)) {
@@ -1716,6 +1719,9 @@ static int dm_early_init(void *handle)
 		adev->mode_info.plane_type = dm_plane_type_default;
 		break;
 	case CHIP_POLARIS10:
+#if defined(CONFIG_DRM_AMD_DC_VEGAM)
+	case CHIP_VEGAM:
+#endif
 		adev->mode_info.num_crtc = 6;
 		adev->mode_info.num_hpd = 6;
 		adev->mode_info.num_dig = 6;
-- 
cgit v1.1


From 221adb2172f10ebc3a1f86c18923692a58cff1de Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Thu, 19 Apr 2018 16:38:46 -0500
Subject: drm/amdgpu: Add VEGAM support to the legacy DCE 11 module

DC is preferred.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/dce_v11_0.c | 13 ++++++++++---
 1 file changed, 10 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
index d3ae508..a5b96ea 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
@@ -173,6 +173,7 @@ static void dce_v11_0_init_golden_registers(struct amdgpu_device *adev)
 							ARRAY_SIZE(polaris11_golden_settings_a11));
 		break;
 	case CHIP_POLARIS10:
+	case CHIP_VEGAM:
 		amdgpu_device_program_register_sequence(adev,
 							polaris10_golden_settings_a11,
 							ARRAY_SIZE(polaris10_golden_settings_a11));
@@ -473,6 +474,7 @@ static int dce_v11_0_get_num_crtc (struct amdgpu_device *adev)
 		num_crtc = 2;
 		break;
 	case CHIP_POLARIS10:
+	case CHIP_VEGAM:
 		num_crtc = 6;
 		break;
 	case CHIP_POLARIS11:
@@ -1445,6 +1447,7 @@ static int dce_v11_0_audio_init(struct amdgpu_device *adev)
 		adev->mode_info.audio.num_pins = 7;
 		break;
 	case CHIP_POLARIS10:
+	case CHIP_VEGAM:
 		adev->mode_info.audio.num_pins = 8;
 		break;
 	case CHIP_POLARIS11:
@@ -2248,7 +2251,8 @@ static u32 dce_v11_0_pick_pll(struct drm_crtc *crtc)
 
 	if ((adev->asic_type == CHIP_POLARIS10) ||
 	    (adev->asic_type == CHIP_POLARIS11) ||
-	    (adev->asic_type == CHIP_POLARIS12)) {
+	    (adev->asic_type == CHIP_POLARIS12) ||
+	    (adev->asic_type == CHIP_VEGAM)) {
 		struct amdgpu_encoder *amdgpu_encoder =
 			to_amdgpu_encoder(amdgpu_crtc->encoder);
 		struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv;
@@ -2666,7 +2670,8 @@ static int dce_v11_0_crtc_mode_set(struct drm_crtc *crtc,
 
 	if ((adev->asic_type == CHIP_POLARIS10) ||
 	    (adev->asic_type == CHIP_POLARIS11) ||
-	    (adev->asic_type == CHIP_POLARIS12)) {
+	    (adev->asic_type == CHIP_POLARIS12) ||
+	    (adev->asic_type == CHIP_VEGAM)) {
 		struct amdgpu_encoder *amdgpu_encoder =
 			to_amdgpu_encoder(amdgpu_crtc->encoder);
 		int encoder_mode =
@@ -2823,6 +2828,7 @@ static int dce_v11_0_early_init(void *handle)
 		adev->mode_info.num_dig = 9;
 		break;
 	case CHIP_POLARIS10:
+	case CHIP_VEGAM:
 		adev->mode_info.num_hpd = 6;
 		adev->mode_info.num_dig = 6;
 		break;
@@ -2942,7 +2948,8 @@ static int dce_v11_0_hw_init(void *handle)
 	amdgpu_atombios_encoder_init_dig(adev);
 	if ((adev->asic_type == CHIP_POLARIS10) ||
 	    (adev->asic_type == CHIP_POLARIS11) ||
-	    (adev->asic_type == CHIP_POLARIS12)) {
+	    (adev->asic_type == CHIP_POLARIS12) ||
+	    (adev->asic_type == CHIP_VEGAM)) {
 		amdgpu_atombios_crtc_set_dce_clock(adev, adev->clock.default_dispclk,
 						   DCE_CLOCK_TYPE_DISPCLK, ATOM_GCK_DFS);
 		amdgpu_atombios_crtc_set_dce_clock(adev, 0,
-- 
cgit v1.1


From 48231fd51667a89514d0eaba893ae0743fd0877d Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Tue, 21 Nov 2017 13:34:48 -0500
Subject: drm/amd/display: Use HBR2 if eDP monitor it doesn't advertise link
 rate

Some eDP displays use the extra link rate table to advertise link rate
support. If they do that they don't need to provide link rate through
the usual registers. Since we don't currently have support for the extra
link rate table default to HBR2 for the display in this.

Note that this is a HACK. Ultimately we need to teach DC to use the
extra link rate table.

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
index 07cc438..0a190c2 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
@@ -2393,6 +2393,10 @@ bool detect_dp_sink_caps(struct dc_link *link)
 void detect_edp_sink_caps(struct dc_link *link)
 {
 	retrieve_link_cap(link);
+
+	if (link->reported_link_cap.link_rate == LINK_RATE_UNKNOWN)
+		link->reported_link_cap.link_rate = LINK_RATE_HIGH2;
+
 	link->verified_link_cap = link->reported_link_cap;
 }
 
-- 
cgit v1.1


From f4ad6fa99772969c16c3fc8877e450b48e93e102 Mon Sep 17 00:00:00 2001
From: Eric Huang <JinHuiEric.Huang@amd.com>
Date: Thu, 9 Nov 2017 16:29:28 -0500
Subject: drm/amd/powerplay: add smu75 header files

Signed-off-by: Eric Huang <JinHuiEric.Huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/inc/smu75.h          | 760 ++++++++++++++++++
 drivers/gpu/drm/amd/powerplay/inc/smu75_discrete.h | 886 +++++++++++++++++++++
 2 files changed, 1646 insertions(+)
 create mode 100644 drivers/gpu/drm/amd/powerplay/inc/smu75.h
 create mode 100644 drivers/gpu/drm/amd/powerplay/inc/smu75_discrete.h

diff --git a/drivers/gpu/drm/amd/powerplay/inc/smu75.h b/drivers/gpu/drm/amd/powerplay/inc/smu75.h
new file mode 100644
index 0000000..7715230
--- /dev/null
+++ b/drivers/gpu/drm/amd/powerplay/inc/smu75.h
@@ -0,0 +1,760 @@
+/*
+ * Copyright 2017 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#ifndef SMU75_H
+#define SMU75_H
+
+#pragma pack(push, 1)
+
+typedef struct {
+	uint32_t high;
+	uint32_t low;
+} data_64_t;
+
+typedef struct {
+	data_64_t high;
+	data_64_t low;
+} data_128_t;
+
+#define SMU__DGPU_ONLY
+
+#define SMU__NUM_SCLK_DPM_STATE  8
+#define SMU__NUM_MCLK_DPM_LEVELS 4
+#define SMU__NUM_LCLK_DPM_LEVELS 8
+#define SMU__NUM_PCIE_DPM_LEVELS 8
+
+#define SMU7_CONTEXT_ID_SMC        1
+#define SMU7_CONTEXT_ID_VBIOS      2
+
+#define SMU75_MAX_LEVELS_VDDC            16
+#define SMU75_MAX_LEVELS_VDDGFX          16
+#define SMU75_MAX_LEVELS_VDDCI           8
+#define SMU75_MAX_LEVELS_MVDD            4
+
+#define SMU_MAX_SMIO_LEVELS              4
+
+#define SMU75_MAX_LEVELS_GRAPHICS        SMU__NUM_SCLK_DPM_STATE
+#define SMU75_MAX_LEVELS_MEMORY          SMU__NUM_MCLK_DPM_LEVELS
+#define SMU75_MAX_LEVELS_GIO             SMU__NUM_LCLK_DPM_LEVELS
+#define SMU75_MAX_LEVELS_LINK            SMU__NUM_PCIE_DPM_LEVELS
+#define SMU75_MAX_LEVELS_UVD             8
+#define SMU75_MAX_LEVELS_VCE             8
+#define SMU75_MAX_LEVELS_ACP             8
+#define SMU75_MAX_LEVELS_SAMU            8
+#define SMU75_MAX_ENTRIES_SMIO           32
+
+#define DPM_NO_LIMIT 0
+#define DPM_NO_UP 1
+#define DPM_GO_DOWN 2
+#define DPM_GO_UP 3
+
+#define SMU7_FIRST_DPM_GRAPHICS_LEVEL    0
+#define SMU7_FIRST_DPM_MEMORY_LEVEL      0
+
+#define GPIO_CLAMP_MODE_VRHOT      1
+#define GPIO_CLAMP_MODE_THERM      2
+#define GPIO_CLAMP_MODE_DC         4
+
+#define SCRATCH_B_TARG_PCIE_INDEX_SHIFT 0
+#define SCRATCH_B_TARG_PCIE_INDEX_MASK  (0x7<<SCRATCH_B_TARG_PCIE_INDEX_SHIFT)
+#define SCRATCH_B_CURR_PCIE_INDEX_SHIFT 3
+#define SCRATCH_B_CURR_PCIE_INDEX_MASK  (0x7<<SCRATCH_B_CURR_PCIE_INDEX_SHIFT)
+#define SCRATCH_B_TARG_UVD_INDEX_SHIFT  6
+#define SCRATCH_B_TARG_UVD_INDEX_MASK   (0x7<<SCRATCH_B_TARG_UVD_INDEX_SHIFT)
+#define SCRATCH_B_CURR_UVD_INDEX_SHIFT  9
+#define SCRATCH_B_CURR_UVD_INDEX_MASK   (0x7<<SCRATCH_B_CURR_UVD_INDEX_SHIFT)
+#define SCRATCH_B_TARG_VCE_INDEX_SHIFT  12
+#define SCRATCH_B_TARG_VCE_INDEX_MASK   (0x7<<SCRATCH_B_TARG_VCE_INDEX_SHIFT)
+#define SCRATCH_B_CURR_VCE_INDEX_SHIFT  15
+#define SCRATCH_B_CURR_VCE_INDEX_MASK   (0x7<<SCRATCH_B_CURR_VCE_INDEX_SHIFT)
+#define SCRATCH_B_TARG_ACP_INDEX_SHIFT  18
+#define SCRATCH_B_TARG_ACP_INDEX_MASK   (0x7<<SCRATCH_B_TARG_ACP_INDEX_SHIFT)
+#define SCRATCH_B_CURR_ACP_INDEX_SHIFT  21
+#define SCRATCH_B_CURR_ACP_INDEX_MASK   (0x7<<SCRATCH_B_CURR_ACP_INDEX_SHIFT)
+#define SCRATCH_B_TARG_SAMU_INDEX_SHIFT 24
+#define SCRATCH_B_TARG_SAMU_INDEX_MASK  (0x7<<SCRATCH_B_TARG_SAMU_INDEX_SHIFT)
+#define SCRATCH_B_CURR_SAMU_INDEX_SHIFT 27
+#define SCRATCH_B_CURR_SAMU_INDEX_MASK  (0x7<<SCRATCH_B_CURR_SAMU_INDEX_SHIFT)
+
+/* Virtualization Defines */
+#define CG_XDMA_MASK  0x1
+#define CG_XDMA_SHIFT 0
+#define CG_UVD_MASK   0x2
+#define CG_UVD_SHIFT  1
+#define CG_VCE_MASK   0x4
+#define CG_VCE_SHIFT  2
+#define CG_SAMU_MASK  0x8
+#define CG_SAMU_SHIFT 3
+#define CG_GFX_MASK   0x10
+#define CG_GFX_SHIFT  4
+#define CG_SDMA_MASK  0x20
+#define CG_SDMA_SHIFT 5
+#define CG_HDP_MASK   0x40
+#define CG_HDP_SHIFT  6
+#define CG_MC_MASK    0x80
+#define CG_MC_SHIFT   7
+#define CG_DRM_MASK   0x100
+#define CG_DRM_SHIFT  8
+#define CG_ROM_MASK   0x200
+#define CG_ROM_SHIFT  9
+#define CG_BIF_MASK   0x400
+#define CG_BIF_SHIFT  10
+
+#if defined SMU__DGPU_ONLY
+#define SMU75_DTE_ITERATIONS 5
+#define SMU75_DTE_SOURCES 3
+#define SMU75_DTE_SINKS 1
+#define SMU75_NUM_CPU_TES 0
+#define SMU75_NUM_GPU_TES 1
+#define SMU75_NUM_NON_TES 2
+#define SMU75_DTE_FAN_SCALAR_MIN 0x100
+#define SMU75_DTE_FAN_SCALAR_MAX 0x166
+#define SMU75_DTE_FAN_TEMP_MAX 93
+#define SMU75_DTE_FAN_TEMP_MIN 83
+#endif
+#define SMU75_THERMAL_INPUT_LOOP_COUNT 2
+#define SMU75_THERMAL_CLAMP_MODE_COUNT 2
+
+#define EXP_M1_1  93
+#define EXP_M2_1  195759
+#define EXP_B_1   111176531
+
+#define EXP_M1_2  67
+#define EXP_M2_2  153720
+#define EXP_B_2   94415767
+
+#define EXP_M1_3  48
+#define EXP_M2_3  119796
+#define EXP_B_3   79195279
+
+#define EXP_M1_4  550
+#define EXP_M2_4  1484190
+#define EXP_B_4   1051432828
+
+#define EXP_M1_5  394
+#define EXP_M2_5  1143049
+#define EXP_B_5   864288432
+
+struct SMU7_HystController_Data {
+	uint16_t waterfall_up;
+	uint16_t waterfall_down;
+	uint16_t waterfall_limit;
+	uint16_t release_cnt;
+	uint16_t release_limit;
+	uint16_t spare;
+};
+
+typedef struct SMU7_HystController_Data SMU7_HystController_Data;
+
+struct SMU75_PIDController {
+	uint32_t Ki;
+	int32_t LFWindupUpperLim;
+	int32_t LFWindupLowerLim;
+	uint32_t StatePrecision;
+	uint32_t LfPrecision;
+	uint32_t LfOffset;
+	uint32_t MaxState;
+	uint32_t MaxLfFraction;
+	uint32_t StateShift;
+};
+
+typedef struct SMU75_PIDController SMU75_PIDController;
+
+struct SMU7_LocalDpmScoreboard {
+	uint32_t PercentageBusy;
+
+	int32_t  PIDError;
+	int32_t  PIDIntegral;
+	int32_t  PIDOutput;
+
+	uint32_t SigmaDeltaAccum;
+	uint32_t SigmaDeltaOutput;
+	uint32_t SigmaDeltaLevel;
+
+	uint32_t UtilizationSetpoint;
+
+	uint8_t  TdpClampMode;
+	uint8_t  TdcClampMode;
+	uint8_t  ThermClampMode;
+	uint8_t  VoltageBusy;
+
+	int8_t   CurrLevel;
+	int8_t   TargLevel;
+	uint8_t  LevelChangeInProgress;
+	uint8_t  UpHyst;
+
+	uint8_t  DownHyst;
+	uint8_t  VoltageDownHyst;
+	uint8_t  DpmEnable;
+	uint8_t  DpmRunning;
+
+	uint8_t  DpmForce;
+	uint8_t  DpmForceLevel;
+	uint8_t  DisplayWatermark;
+	uint8_t  McArbIndex;
+
+	uint32_t MinimumPerfSclk;
+
+	uint8_t  AcpiReq;
+	uint8_t  AcpiAck;
+	uint8_t  GfxClkSlow;
+	uint8_t  GpioClampMode;
+
+	uint8_t  EnableModeSwitchRLCNotification;
+	uint8_t  EnabledLevelsChange;
+	uint8_t  DteClampMode;
+	uint8_t  FpsClampMode;
+
+	uint16_t LevelResidencyCounters [SMU75_MAX_LEVELS_GRAPHICS];
+	uint16_t LevelSwitchCounters [SMU75_MAX_LEVELS_GRAPHICS];
+
+	void     (*TargetStateCalculator)(uint8_t);
+	void     (*SavedTargetStateCalculator)(uint8_t);
+
+	uint16_t AutoDpmInterval;
+	uint16_t AutoDpmRange;
+
+	uint8_t  FpsEnabled;
+	uint8_t  MaxPerfLevel;
+	uint8_t  AllowLowClkInterruptToHost;
+	uint8_t  FpsRunning;
+
+	uint32_t MaxAllowedFrequency;
+
+	uint32_t FilteredSclkFrequency;
+	uint32_t LastSclkFrequency;
+	uint32_t FilteredSclkFrequencyCnt;
+
+	uint8_t MinPerfLevel;
+#ifdef SMU__FIRMWARE_SCKS_PRESENT__1
+	uint8_t ScksClampMode;
+	uint8_t padding[2];
+#else
+	uint8_t padding[3];
+#endif
+
+	uint16_t FpsAlpha;
+	uint16_t DeltaTime;
+	uint32_t CurrentFps;
+	uint32_t FilteredFps;
+	uint32_t FrameCount;
+	uint32_t FrameCountLast;
+	uint16_t FpsTargetScalar;
+	uint16_t FpsWaterfallLimitScalar;
+	uint16_t FpsAlphaScalar;
+	uint16_t spare8;
+	SMU7_HystController_Data HystControllerData;
+};
+
+typedef struct SMU7_LocalDpmScoreboard SMU7_LocalDpmScoreboard;
+
+#define SMU7_MAX_VOLTAGE_CLIENTS 12
+
+typedef uint8_t (*VoltageChangeHandler_t)(uint16_t, uint8_t);
+
+#define VDDC_MASK    0x00007FFF
+#define VDDC_SHIFT   0
+#define VDDCI_MASK   0x3FFF8000
+#define VDDCI_SHIFT  15
+#define PHASES_MASK  0xC0000000
+#define PHASES_SHIFT 30
+
+typedef uint32_t SMU_VoltageLevel;
+
+struct SMU7_VoltageScoreboard {
+	SMU_VoltageLevel TargetVoltage;
+	uint16_t MaxVid;
+	uint8_t  HighestVidOffset;
+	uint8_t  CurrentVidOffset;
+
+	uint16_t CurrentVddc;
+	uint16_t CurrentVddci;
+
+	uint8_t  ControllerBusy;
+	uint8_t  CurrentVid;
+	uint8_t  CurrentVddciVid;
+	uint8_t  padding;
+
+	SMU_VoltageLevel RequestedVoltage[SMU7_MAX_VOLTAGE_CLIENTS];
+	SMU_VoltageLevel TargetVoltageState;
+	uint8_t  EnabledRequest[SMU7_MAX_VOLTAGE_CLIENTS];
+
+	uint8_t  padding2;
+	uint8_t  padding3;
+	uint8_t  ControllerEnable;
+	uint8_t  ControllerRunning;
+	uint16_t CurrentStdVoltageHiSidd;
+	uint16_t CurrentStdVoltageLoSidd;
+	uint8_t  OverrideVoltage;
+	uint8_t  padding4;
+	uint8_t  padding5;
+	uint8_t  CurrentPhases;
+
+	VoltageChangeHandler_t ChangeVddc;
+	VoltageChangeHandler_t ChangeVddci;
+	VoltageChangeHandler_t ChangePhase;
+	VoltageChangeHandler_t ChangeMvdd;
+
+	VoltageChangeHandler_t functionLinks[6];
+
+	uint16_t * VddcFollower1;
+	int16_t  Driver_OD_RequestedVidOffset1;
+	int16_t  Driver_OD_RequestedVidOffset2;
+};
+
+typedef struct SMU7_VoltageScoreboard SMU7_VoltageScoreboard;
+
+#define SMU7_MAX_PCIE_LINK_SPEEDS 3
+
+struct SMU7_PCIeLinkSpeedScoreboard {
+	uint8_t     DpmEnable;
+	uint8_t     DpmRunning;
+	uint8_t     DpmForce;
+	uint8_t     DpmForceLevel;
+
+	uint8_t     CurrentLinkSpeed;
+	uint8_t     EnabledLevelsChange;
+	uint16_t    AutoDpmInterval;
+
+	uint16_t    AutoDpmRange;
+	uint16_t    AutoDpmCount;
+
+	uint8_t     DpmMode;
+	uint8_t     AcpiReq;
+	uint8_t     AcpiAck;
+	uint8_t     CurrentLinkLevel;
+};
+
+typedef struct SMU7_PCIeLinkSpeedScoreboard SMU7_PCIeLinkSpeedScoreboard;
+
+#define SMU7_LKGE_LUT_NUM_OF_TEMP_ENTRIES 16
+#define SMU7_LKGE_LUT_NUM_OF_VOLT_ENTRIES 16
+
+#define SMU7_SCALE_I  7
+#define SMU7_SCALE_R 12
+
+struct SMU7_PowerScoreboard {
+	uint32_t GpuPower;
+
+	uint32_t VddcPower;
+	uint32_t VddcVoltage;
+	uint32_t VddcCurrent;
+
+	uint32_t VddciPower;
+	uint32_t VddciVoltage;
+	uint32_t VddciCurrent;
+
+	uint32_t RocPower;
+
+	uint16_t Telemetry_1_slope;
+	uint16_t Telemetry_2_slope;
+	int32_t  Telemetry_1_offset;
+	int32_t  Telemetry_2_offset;
+
+	uint8_t MCLK_patch_flag;
+	uint8_t reserved[3];
+};
+
+typedef struct SMU7_PowerScoreboard SMU7_PowerScoreboard;
+
+#define SMU7_SCLK_DPM_CONFIG_MASK                        0x01
+#define SMU7_VOLTAGE_CONTROLLER_CONFIG_MASK              0x02
+#define SMU7_THERMAL_CONTROLLER_CONFIG_MASK              0x04
+#define SMU7_MCLK_DPM_CONFIG_MASK                        0x08
+#define SMU7_UVD_DPM_CONFIG_MASK                         0x10
+#define SMU7_VCE_DPM_CONFIG_MASK                         0x20
+#define SMU7_ACP_DPM_CONFIG_MASK                         0x40
+#define SMU7_SAMU_DPM_CONFIG_MASK                        0x80
+#define SMU7_PCIEGEN_DPM_CONFIG_MASK                    0x100
+
+#define SMU7_ACP_MCLK_HANDSHAKE_DISABLE                  0x00000001
+#define SMU7_ACP_SCLK_HANDSHAKE_DISABLE                  0x00000002
+#define SMU7_UVD_MCLK_HANDSHAKE_DISABLE                  0x00000100
+#define SMU7_UVD_SCLK_HANDSHAKE_DISABLE                  0x00000200
+#define SMU7_VCE_MCLK_HANDSHAKE_DISABLE                  0x00010000
+#define SMU7_VCE_SCLK_HANDSHAKE_DISABLE                  0x00020000
+
+struct SMU75_SoftRegisters {
+	uint32_t        RefClockFrequency;
+	uint32_t        PmTimerPeriod;
+	uint32_t        FeatureEnables;
+#if defined (SMU__DGPU_ONLY)
+	uint32_t        PreVBlankGap;
+	uint32_t        VBlankTimeout;
+	uint32_t        TrainTimeGap;
+	uint32_t        MvddSwitchTime;
+	uint32_t        LongestAcpiTrainTime;
+	uint32_t        AcpiDelay;
+	uint32_t        G5TrainTime;
+	uint32_t        DelayMpllPwron;
+	uint32_t        VoltageChangeTimeout;
+#endif
+	uint32_t        HandshakeDisables;
+
+	uint8_t         DisplayPhy1Config;
+	uint8_t         DisplayPhy2Config;
+	uint8_t         DisplayPhy3Config;
+	uint8_t         DisplayPhy4Config;
+
+	uint8_t         DisplayPhy5Config;
+	uint8_t         DisplayPhy6Config;
+	uint8_t         DisplayPhy7Config;
+	uint8_t         DisplayPhy8Config;
+
+	uint32_t        AverageGraphicsActivity;
+	uint32_t        AverageMemoryActivity;
+	uint32_t        AverageGioActivity;
+
+	uint8_t         SClkDpmEnabledLevels;
+	uint8_t         MClkDpmEnabledLevels;
+	uint8_t         LClkDpmEnabledLevels;
+	uint8_t         PCIeDpmEnabledLevels;
+
+	uint8_t         UVDDpmEnabledLevels;
+	uint8_t         SAMUDpmEnabledLevels;
+	uint8_t         ACPDpmEnabledLevels;
+	uint8_t         VCEDpmEnabledLevels;
+
+	uint32_t        DRAM_LOG_ADDR_H;
+	uint32_t        DRAM_LOG_ADDR_L;
+	uint32_t        DRAM_LOG_PHY_ADDR_H;
+	uint32_t        DRAM_LOG_PHY_ADDR_L;
+	uint32_t        DRAM_LOG_BUFF_SIZE;
+	uint32_t        UlvEnterCount;
+	uint32_t        UlvTime;
+	uint32_t        UcodeLoadStatus;
+	uint32_t        AllowMvddSwitch;
+	uint8_t         Activity_Weight;
+	uint8_t         Reserved8[3];
+};
+
+typedef struct SMU75_SoftRegisters SMU75_SoftRegisters;
+
+struct SMU75_Firmware_Header {
+	uint32_t Digest[5];
+	uint32_t Version;
+	uint32_t HeaderSize;
+	uint32_t Flags;
+	uint32_t EntryPoint;
+	uint32_t CodeSize;
+	uint32_t ImageSize;
+
+	uint32_t Rtos;
+	uint32_t SoftRegisters;
+	uint32_t DpmTable;
+	uint32_t FanTable;
+	uint32_t CacConfigTable;
+	uint32_t CacStatusTable;
+	uint32_t mcRegisterTable;
+	uint32_t mcArbDramTimingTable;
+	uint32_t PmFuseTable;
+	uint32_t Globals;
+	uint32_t ClockStretcherTable;
+	uint32_t VftTable;
+	uint32_t Reserved1;
+	uint32_t AvfsCksOff_AvfsGbvTable;
+	uint32_t AvfsCksOff_BtcGbvTable;
+	uint32_t MM_AvfsTable;
+	uint32_t PowerSharingTable;
+	uint32_t AvfsTable;
+	uint32_t AvfsCksOffGbvTable;
+	uint32_t AvfsMeanNSigma;
+	uint32_t AvfsSclkOffsetTable;
+	uint32_t Reserved[12];
+	uint32_t Signature;
+};
+
+typedef struct SMU75_Firmware_Header SMU75_Firmware_Header;
+
+#define SMU7_FIRMWARE_HEADER_LOCATION 0x20000
+
+enum  DisplayConfig {
+	PowerDown = 1,
+	DP54x4,
+	DP54x2,
+	DP54x1,
+	DP27x4,
+	DP27x2,
+	DP27x1,
+	HDMI297,
+	HDMI162,
+	LVDS,
+	DP324x4,
+	DP324x2,
+	DP324x1
+};
+
+#define MC_BLOCK_COUNT 1
+#define CPL_BLOCK_COUNT 5
+#define SE_BLOCK_COUNT 15
+#define GC_BLOCK_COUNT 24
+
+struct SMU7_Local_Cac {
+	uint8_t BlockId;
+	uint8_t SignalId;
+	uint8_t Threshold;
+	uint8_t Padding;
+};
+
+typedef struct SMU7_Local_Cac SMU7_Local_Cac;
+
+struct SMU7_Local_Cac_Table {
+	SMU7_Local_Cac CplLocalCac[CPL_BLOCK_COUNT];
+	SMU7_Local_Cac McLocalCac[MC_BLOCK_COUNT];
+	SMU7_Local_Cac SeLocalCac[SE_BLOCK_COUNT];
+	SMU7_Local_Cac GcLocalCac[GC_BLOCK_COUNT];
+};
+
+typedef struct SMU7_Local_Cac_Table SMU7_Local_Cac_Table;
+
+#pragma pack(pop)
+
+#define CG_SYS_BITMASK_FIRST_BIT      0
+#define CG_SYS_BITMASK_LAST_BIT       10
+#define CG_SYS_BIF_MGLS_SHIFT         0
+#define CG_SYS_ROM_SHIFT              1
+#define CG_SYS_MC_MGCG_SHIFT          2
+#define CG_SYS_MC_MGLS_SHIFT          3
+#define CG_SYS_SDMA_MGCG_SHIFT        4
+#define CG_SYS_SDMA_MGLS_SHIFT        5
+#define CG_SYS_DRM_MGCG_SHIFT         6
+#define CG_SYS_HDP_MGCG_SHIFT         7
+#define CG_SYS_HDP_MGLS_SHIFT         8
+#define CG_SYS_DRM_MGLS_SHIFT         9
+#define CG_SYS_BIF_MGCG_SHIFT         10
+
+#define CG_SYS_BIF_MGLS_MASK          0x1
+#define CG_SYS_ROM_MASK               0x2
+#define CG_SYS_MC_MGCG_MASK           0x4
+#define CG_SYS_MC_MGLS_MASK           0x8
+#define CG_SYS_SDMA_MGCG_MASK         0x10
+#define CG_SYS_SDMA_MGLS_MASK         0x20
+#define CG_SYS_DRM_MGCG_MASK          0x40
+#define CG_SYS_HDP_MGCG_MASK          0x80
+#define CG_SYS_HDP_MGLS_MASK          0x100
+#define CG_SYS_DRM_MGLS_MASK          0x200
+#define CG_SYS_BIF_MGCG_MASK          0x400
+
+#define CG_GFX_BITMASK_FIRST_BIT      16
+#define CG_GFX_BITMASK_LAST_BIT       24
+
+#define CG_GFX_CGCG_SHIFT             16
+#define CG_GFX_CGLS_SHIFT             17
+#define CG_CPF_MGCG_SHIFT             18
+#define CG_RLC_MGCG_SHIFT             19
+#define CG_GFX_OTHERS_MGCG_SHIFT      20
+#define CG_GFX_3DCG_SHIFT             21
+#define CG_GFX_3DLS_SHIFT             22
+#define CG_GFX_RLC_LS_SHIFT           23
+#define CG_GFX_CP_LS_SHIFT            24
+
+#define CG_GFX_CGCG_MASK              0x00010000
+#define CG_GFX_CGLS_MASK              0x00020000
+#define CG_CPF_MGCG_MASK              0x00040000
+#define CG_RLC_MGCG_MASK              0x00080000
+#define CG_GFX_OTHERS_MGCG_MASK       0x00100000
+#define CG_GFX_3DCG_MASK              0x00200000
+#define CG_GFX_3DLS_MASK              0x00400000
+#define CG_GFX_RLC_LS_MASK            0x00800000
+#define CG_GFX_CP_LS_MASK             0x01000000
+
+
+#define VRCONF_VDDC_MASK         0x000000FF
+#define VRCONF_VDDC_SHIFT        0
+#define VRCONF_VDDGFX_MASK       0x0000FF00
+#define VRCONF_VDDGFX_SHIFT      8
+#define VRCONF_VDDCI_MASK        0x00FF0000
+#define VRCONF_VDDCI_SHIFT       16
+#define VRCONF_MVDD_MASK         0xFF000000
+#define VRCONF_MVDD_SHIFT        24
+
+#define VR_MERGED_WITH_VDDC      0
+#define VR_SVI2_PLANE_1          1
+#define VR_SVI2_PLANE_2          2
+#define VR_SMIO_PATTERN_1        3
+#define VR_SMIO_PATTERN_2        4
+#define VR_STATIC_VOLTAGE        5
+
+#define CLOCK_STRETCHER_MAX_ENTRIES 0x4
+#define CKS_LOOKUPTable_MAX_ENTRIES 0x4
+
+#define CLOCK_STRETCHER_SETTING_DDT_MASK             0x01
+#define CLOCK_STRETCHER_SETTING_DDT_SHIFT            0x0
+#define CLOCK_STRETCHER_SETTING_STRETCH_AMOUNT_MASK  0x1E
+#define CLOCK_STRETCHER_SETTING_STRETCH_AMOUNT_SHIFT 0x1
+#define CLOCK_STRETCHER_SETTING_ENABLE_MASK          0x80
+#define CLOCK_STRETCHER_SETTING_ENABLE_SHIFT         0x7
+
+struct SMU_ClockStretcherDataTableEntry {
+	uint8_t minVID;
+	uint8_t maxVID;
+
+	uint16_t setting;
+};
+typedef struct SMU_ClockStretcherDataTableEntry SMU_ClockStretcherDataTableEntry;
+
+struct SMU_ClockStretcherDataTable {
+	SMU_ClockStretcherDataTableEntry ClockStretcherDataTableEntry[CLOCK_STRETCHER_MAX_ENTRIES];
+};
+typedef struct SMU_ClockStretcherDataTable SMU_ClockStretcherDataTable;
+
+struct SMU_CKS_LOOKUPTableEntry {
+	uint16_t minFreq;
+	uint16_t maxFreq;
+
+	uint8_t setting;
+	uint8_t padding[3];
+};
+typedef struct SMU_CKS_LOOKUPTableEntry SMU_CKS_LOOKUPTableEntry;
+
+struct SMU_CKS_LOOKUPTable {
+	SMU_CKS_LOOKUPTableEntry CKS_LOOKUPTableEntry[CKS_LOOKUPTable_MAX_ENTRIES];
+};
+typedef struct SMU_CKS_LOOKUPTable SMU_CKS_LOOKUPTable;
+
+struct AgmAvfsData_t {
+	uint16_t avgPsmCount[28];
+	uint16_t minPsmCount[28];
+};
+typedef struct AgmAvfsData_t AgmAvfsData_t;
+
+enum VFT_COLUMNS {
+	SCLK0,
+	SCLK1,
+	SCLK2,
+	SCLK3,
+	SCLK4,
+	SCLK5,
+	SCLK6,
+	SCLK7,
+
+	NUM_VFT_COLUMNS
+};
+enum {
+  SCS_FUSE_T0,
+  SCS_FUSE_T1,
+  NUM_SCS_FUSE_TEMPERATURE
+};
+enum {
+  SCKS_ON,
+  SCKS_OFF,
+  NUM_SCKS_STATE_TYPES
+};
+
+#define VFT_TABLE_DEFINED
+
+#define TEMP_RANGE_MAXSTEPS 12
+struct VFT_CELL_t {
+	uint16_t Voltage;
+};
+
+typedef struct VFT_CELL_t VFT_CELL_t;
+#ifdef SMU__FIRMWARE_SCKS_PRESENT__1
+struct SCS_CELL_t {
+	uint16_t PsmCnt[NUM_SCKS_STATE_TYPES];
+};
+typedef struct SCS_CELL_t SCS_CELL_t;
+#endif
+
+struct VFT_TABLE_t {
+	VFT_CELL_t    Cell[TEMP_RANGE_MAXSTEPS][NUM_VFT_COLUMNS];
+	uint16_t      AvfsGbv [NUM_VFT_COLUMNS];
+	uint16_t      BtcGbv  [NUM_VFT_COLUMNS];
+	int16_t       Temperature [TEMP_RANGE_MAXSTEPS];
+
+#ifdef SMU__FIRMWARE_SCKS_PRESENT__1
+	SCS_CELL_t    ScksCell[TEMP_RANGE_MAXSTEPS][NUM_VFT_COLUMNS];
+#endif
+
+	uint8_t       NumTemperatureSteps;
+	uint8_t       padding[3];
+};
+typedef struct VFT_TABLE_t VFT_TABLE_t;
+
+#define BTCGB_VDROOP_TABLE_MAX_ENTRIES 2
+#define AVFSGB_VDROOP_TABLE_MAX_ENTRIES 2
+
+struct GB_VDROOP_TABLE_t {
+	int32_t a0;
+	int32_t a1;
+	int32_t a2;
+	uint32_t spare;
+};
+typedef struct GB_VDROOP_TABLE_t GB_VDROOP_TABLE_t;
+
+struct SMU_QuadraticCoeffs {
+	int32_t m1;
+	int32_t b;
+
+	int16_t m2;
+	uint8_t m1_shift;
+	uint8_t m2_shift;
+};
+typedef struct SMU_QuadraticCoeffs SMU_QuadraticCoeffs;
+
+struct AVFS_Margin_t {
+	VFT_CELL_t Cell[NUM_VFT_COLUMNS];
+};
+typedef struct AVFS_Margin_t AVFS_Margin_t;
+
+struct AVFS_CksOff_Gbv_t {
+	VFT_CELL_t Cell[NUM_VFT_COLUMNS];
+};
+typedef struct AVFS_CksOff_Gbv_t AVFS_CksOff_Gbv_t;
+
+struct AVFS_CksOff_AvfsGbv_t {
+	VFT_CELL_t Cell[NUM_VFT_COLUMNS];
+};
+typedef struct AVFS_CksOff_AvfsGbv_t AVFS_CksOff_AvfsGbv_t;
+
+struct AVFS_CksOff_BtcGbv_t {
+	VFT_CELL_t Cell[NUM_VFT_COLUMNS];
+};
+typedef struct AVFS_CksOff_BtcGbv_t AVFS_CksOff_BtcGbv_t;
+
+struct AVFS_meanNsigma_t {
+	uint32_t Aconstant[3];
+	uint16_t DC_tol_sigma;
+	uint16_t Platform_mean;
+	uint16_t Platform_sigma;
+	uint16_t PSM_Age_CompFactor;
+	uint8_t  Static_Voltage_Offset[NUM_VFT_COLUMNS];
+};
+typedef struct AVFS_meanNsigma_t AVFS_meanNsigma_t;
+
+struct AVFS_Sclk_Offset_t {
+	uint16_t Sclk_Offset[8];
+};
+typedef struct AVFS_Sclk_Offset_t AVFS_Sclk_Offset_t;
+
+struct Power_Sharing_t {
+	uint32_t EnergyCounter;
+	uint32_t EngeryThreshold;
+	uint64_t AM_SCLK_CNT;
+	uint64_t AM_0_BUSY_CNT;
+};
+typedef struct Power_Sharing_t  Power_Sharing_t;
+
+
+#endif
+
+
diff --git a/drivers/gpu/drm/amd/powerplay/inc/smu75_discrete.h b/drivers/gpu/drm/amd/powerplay/inc/smu75_discrete.h
new file mode 100644
index 0000000..b64e58a2
--- /dev/null
+++ b/drivers/gpu/drm/amd/powerplay/inc/smu75_discrete.h
@@ -0,0 +1,886 @@
+/*
+ * Copyright 2017 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef SMU75_DISCRETE_H
+#define SMU75_DISCRETE_H
+
+#include "smu75.h"
+
+#pragma pack(push, 1)
+
+#define NUM_SCLK_RANGE 8
+
+#define VCO_3_6 1
+#define VCO_2_4 3
+
+#define POSTDIV_DIV_BY_1  0
+#define POSTDIV_DIV_BY_2  1
+#define POSTDIV_DIV_BY_4  2
+#define POSTDIV_DIV_BY_8  3
+#define POSTDIV_DIV_BY_16 4
+
+struct sclkFcwRange_t {
+	uint8_t  vco_setting; /* 1: 3-6GHz, 3: 2-4GHz */
+	uint8_t  postdiv;     /* divide by 2^n */
+	uint16_t fcw_pcc;
+	uint16_t fcw_trans_upper;
+	uint16_t fcw_trans_lower;
+};
+typedef struct sclkFcwRange_t sclkFcwRange_t;
+
+struct SMIO_Pattern {
+	uint16_t Voltage;
+	uint8_t  Smio;
+	uint8_t  padding;
+};
+
+typedef struct SMIO_Pattern SMIO_Pattern;
+
+struct SMIO_Table {
+	SMIO_Pattern Pattern[SMU_MAX_SMIO_LEVELS];
+};
+
+typedef struct SMIO_Table SMIO_Table;
+
+struct SMU_SclkSetting {
+	uint32_t    SclkFrequency;
+	uint16_t    Fcw_int;
+	uint16_t    Fcw_frac;
+	uint16_t    Pcc_fcw_int;
+	uint8_t     PllRange;
+	uint8_t     SSc_En;
+	uint16_t    Sclk_slew_rate;
+	uint16_t    Pcc_up_slew_rate;
+	uint16_t    Pcc_down_slew_rate;
+	uint16_t    Fcw1_int;
+	uint16_t    Fcw1_frac;
+	uint16_t    Sclk_ss_slew_rate;
+};
+typedef struct SMU_SclkSetting SMU_SclkSetting;
+
+struct SMU75_Discrete_GraphicsLevel {
+	SMU_VoltageLevel MinVoltage;
+
+	uint8_t     pcieDpmLevel;
+	uint8_t     DeepSleepDivId;
+	uint16_t    ActivityLevel;
+
+	uint32_t    CgSpllFuncCntl3;
+	uint32_t    CgSpllFuncCntl4;
+	uint32_t    CcPwrDynRm;
+	uint32_t    CcPwrDynRm1;
+
+	uint8_t     SclkDid;
+	uint8_t     padding;
+	uint8_t     EnabledForActivity;
+	uint8_t     EnabledForThrottle;
+	uint8_t     UpHyst;
+	uint8_t     DownHyst;
+	uint8_t     VoltageDownHyst;
+	uint8_t     PowerThrottle;
+
+	SMU_SclkSetting SclkSetting;
+
+	uint8_t  ScksStretchThreshVid[NUM_SCKS_STATE_TYPES];
+	uint16_t Padding;
+};
+
+typedef struct SMU75_Discrete_GraphicsLevel SMU75_Discrete_GraphicsLevel;
+
+struct SMU75_Discrete_ACPILevel {
+	uint32_t    Flags;
+	SMU_VoltageLevel MinVoltage;
+	uint32_t    SclkFrequency;
+	uint8_t     SclkDid;
+	uint8_t     DisplayWatermark;
+	uint8_t     DeepSleepDivId;
+	uint8_t     padding;
+	uint32_t    CcPwrDynRm;
+	uint32_t    CcPwrDynRm1;
+
+	SMU_SclkSetting SclkSetting;
+};
+
+typedef struct SMU75_Discrete_ACPILevel SMU75_Discrete_ACPILevel;
+
+struct SMU75_Discrete_Ulv {
+	uint32_t    CcPwrDynRm;
+	uint32_t    CcPwrDynRm1;
+	uint16_t    VddcOffset;
+	uint8_t     VddcOffsetVid;
+	uint8_t     VddcPhase;
+	uint16_t    BifSclkDfs;
+	uint16_t    Reserved;
+};
+
+typedef struct SMU75_Discrete_Ulv SMU75_Discrete_Ulv;
+
+struct SMU75_Discrete_MemoryLevel {
+	SMU_VoltageLevel MinVoltage;
+	uint32_t    MinMvdd;
+
+	uint32_t    MclkFrequency;
+
+	uint8_t     StutterEnable;
+	uint8_t     EnabledForThrottle;
+	uint8_t     EnabledForActivity;
+	uint8_t     padding_0;
+
+	uint8_t     UpHyst;
+	uint8_t     DownHyst;
+	uint8_t     VoltageDownHyst;
+	uint8_t     padding_1;
+
+	uint16_t    ActivityLevel;
+	uint8_t     DisplayWatermark;
+	uint8_t     padding_2;
+
+	uint16_t    Fcw_int;
+	uint16_t    Fcw_frac;
+	uint8_t     Postdiv;
+	uint8_t     padding_3[3];
+};
+
+typedef struct SMU75_Discrete_MemoryLevel SMU75_Discrete_MemoryLevel;
+
+struct SMU75_Discrete_LinkLevel {
+	uint8_t     PcieGenSpeed;
+	uint8_t     PcieLaneCount;
+	uint8_t     EnabledForActivity;
+	uint8_t     SPC;
+	uint32_t    DownThreshold;
+	uint32_t    UpThreshold;
+	uint16_t    BifSclkDfs;
+	uint16_t    Reserved;
+};
+
+typedef struct SMU75_Discrete_LinkLevel SMU75_Discrete_LinkLevel;
+
+
+/* MC ARB DRAM Timing registers. */
+struct SMU75_Discrete_MCArbDramTimingTableEntry {
+	uint32_t McArbDramTiming;
+	uint32_t McArbDramTiming2;
+	uint32_t McArbBurstTime;
+	uint32_t McArbRfshRate;
+	uint32_t McArbMisc3;
+};
+
+typedef struct SMU75_Discrete_MCArbDramTimingTableEntry SMU75_Discrete_MCArbDramTimingTableEntry;
+
+struct SMU75_Discrete_MCArbDramTimingTable {
+	SMU75_Discrete_MCArbDramTimingTableEntry entries[SMU__NUM_SCLK_DPM_STATE][SMU__NUM_MCLK_DPM_LEVELS];
+};
+
+typedef struct SMU75_Discrete_MCArbDramTimingTable SMU75_Discrete_MCArbDramTimingTable;
+
+/* UVD VCLK/DCLK state (level) definition. */
+struct SMU75_Discrete_UvdLevel {
+	uint32_t VclkFrequency;
+	uint32_t DclkFrequency;
+	SMU_VoltageLevel MinVoltage;
+	uint8_t  VclkDivider;
+	uint8_t  DclkDivider;
+	uint8_t  padding[2];
+};
+
+typedef struct SMU75_Discrete_UvdLevel SMU75_Discrete_UvdLevel;
+
+/* Clocks for other external blocks (VCE, ACP, SAMU). */
+struct SMU75_Discrete_ExtClkLevel {
+	uint32_t Frequency;
+	SMU_VoltageLevel MinVoltage;
+	uint8_t  Divider;
+	uint8_t  padding[3];
+};
+
+typedef struct SMU75_Discrete_ExtClkLevel SMU75_Discrete_ExtClkLevel;
+
+struct SMU75_Discrete_StateInfo {
+	uint32_t SclkFrequency;
+	uint32_t MclkFrequency;
+	uint32_t VclkFrequency;
+	uint32_t DclkFrequency;
+	uint32_t SamclkFrequency;
+	uint32_t AclkFrequency;
+	uint32_t EclkFrequency;
+	uint16_t MvddVoltage;
+	uint16_t padding16;
+	uint8_t  DisplayWatermark;
+	uint8_t  McArbIndex;
+	uint8_t  McRegIndex;
+	uint8_t  SeqIndex;
+	uint8_t  SclkDid;
+	int8_t   SclkIndex;
+	int8_t   MclkIndex;
+	uint8_t  PCIeGen;
+};
+
+typedef struct SMU75_Discrete_StateInfo SMU75_Discrete_StateInfo;
+
+struct SMU75_Discrete_DpmTable {
+	SMU75_PIDController                  GraphicsPIDController;
+	SMU75_PIDController                  MemoryPIDController;
+	SMU75_PIDController                  LinkPIDController;
+
+	uint32_t                            SystemFlags;
+
+	uint32_t                            VRConfig;
+	uint32_t                            SmioMask1;
+	uint32_t                            SmioMask2;
+	SMIO_Table                          SmioTable1;
+	SMIO_Table                          SmioTable2;
+
+	uint32_t                            MvddLevelCount;
+
+	uint8_t                             BapmVddcVidHiSidd        [SMU75_MAX_LEVELS_VDDC];
+	uint8_t                             BapmVddcVidLoSidd        [SMU75_MAX_LEVELS_VDDC];
+	uint8_t                             BapmVddcVidHiSidd2       [SMU75_MAX_LEVELS_VDDC];
+
+	uint8_t                             GraphicsDpmLevelCount;
+	uint8_t                             MemoryDpmLevelCount;
+	uint8_t                             LinkLevelCount;
+	uint8_t                             MasterDeepSleepControl;
+
+	uint8_t                             UvdLevelCount;
+	uint8_t                             VceLevelCount;
+	uint8_t                             AcpLevelCount;
+	uint8_t                             SamuLevelCount;
+
+	uint8_t                             ThermOutGpio;
+	uint8_t                             ThermOutPolarity;
+	uint8_t                             ThermOutMode;
+	uint8_t                             BootPhases;
+
+	uint8_t                             VRHotLevel;
+	uint8_t                             LdoRefSel;
+
+	uint8_t                             Reserved1[2];
+
+	uint16_t                            FanStartTemperature;
+	uint16_t                            FanStopTemperature;
+
+	uint16_t                            MaxVoltage;
+	uint16_t                            Reserved2;
+	uint32_t                            Reserved;
+
+	SMU75_Discrete_GraphicsLevel        GraphicsLevel           [SMU75_MAX_LEVELS_GRAPHICS];
+	SMU75_Discrete_MemoryLevel          MemoryACPILevel;
+	SMU75_Discrete_MemoryLevel          MemoryLevel             [SMU75_MAX_LEVELS_MEMORY];
+	SMU75_Discrete_LinkLevel            LinkLevel               [SMU75_MAX_LEVELS_LINK];
+	SMU75_Discrete_ACPILevel            ACPILevel;
+	SMU75_Discrete_UvdLevel             UvdLevel                [SMU75_MAX_LEVELS_UVD];
+	SMU75_Discrete_ExtClkLevel          VceLevel                [SMU75_MAX_LEVELS_VCE];
+	SMU75_Discrete_ExtClkLevel          AcpLevel                [SMU75_MAX_LEVELS_ACP];
+	SMU75_Discrete_ExtClkLevel          SamuLevel               [SMU75_MAX_LEVELS_SAMU];
+	SMU75_Discrete_Ulv                  Ulv;
+
+	uint8_t                             DisplayWatermark        [SMU75_MAX_LEVELS_MEMORY][SMU75_MAX_LEVELS_GRAPHICS];
+
+	uint32_t                            SclkStepSize;
+	uint32_t                            Smio                    [SMU75_MAX_ENTRIES_SMIO];
+
+	uint8_t                             UvdBootLevel;
+	uint8_t                             VceBootLevel;
+	uint8_t                             AcpBootLevel;
+	uint8_t                             SamuBootLevel;
+
+	uint8_t                             GraphicsBootLevel;
+	uint8_t                             GraphicsVoltageChangeEnable;
+	uint8_t                             GraphicsThermThrottleEnable;
+	uint8_t                             GraphicsInterval;
+
+	uint8_t                             VoltageInterval;
+	uint8_t                             ThermalInterval;
+	uint16_t                            TemperatureLimitHigh;
+
+	uint16_t                            TemperatureLimitLow;
+	uint8_t                             MemoryBootLevel;
+	uint8_t                             MemoryVoltageChangeEnable;
+
+	uint16_t                            BootMVdd;
+	uint8_t                             MemoryInterval;
+	uint8_t                             MemoryThermThrottleEnable;
+
+	uint16_t                            VoltageResponseTime;
+	uint16_t                            PhaseResponseTime;
+
+	uint8_t                             PCIeBootLinkLevel;
+	uint8_t                             PCIeGenInterval;
+	uint8_t                             DTEInterval;
+	uint8_t                             DTEMode;
+
+	uint8_t                             SVI2Enable;
+	uint8_t                             VRHotGpio;
+	uint8_t                             AcDcGpio;
+	uint8_t                             ThermGpio;
+
+	uint16_t                            PPM_PkgPwrLimit;
+	uint16_t                            PPM_TemperatureLimit;
+
+	uint16_t                            DefaultTdp;
+	uint16_t                            TargetTdp;
+
+	uint16_t                            FpsHighThreshold;
+	uint16_t                            FpsLowThreshold;
+
+	uint16_t                            BAPMTI_R  [SMU75_DTE_ITERATIONS][SMU75_DTE_SOURCES][SMU75_DTE_SINKS];
+	uint16_t                            BAPMTI_RC [SMU75_DTE_ITERATIONS][SMU75_DTE_SOURCES][SMU75_DTE_SINKS];
+
+	uint16_t                            TemperatureLimitEdge;
+	uint16_t                            TemperatureLimitHotspot;
+
+	uint16_t                            BootVddc;
+	uint16_t                            BootVddci;
+
+	uint16_t                            FanGainEdge;
+	uint16_t                            FanGainHotspot;
+
+	uint32_t                            LowSclkInterruptThreshold;
+	uint32_t                            VddGfxReChkWait;
+
+	uint8_t                             ClockStretcherAmount;
+	uint8_t                             Sclk_CKS_masterEn0_7;
+	uint8_t                             Sclk_CKS_masterEn8_15;
+	uint8_t                             DPMFreezeAndForced;
+
+	uint8_t                             Sclk_voltageOffset[8];
+
+	SMU_ClockStretcherDataTable         ClockStretcherDataTable;
+	SMU_CKS_LOOKUPTable                 CKS_LOOKUPTable;
+
+	uint32_t                            CurrSclkPllRange;
+	sclkFcwRange_t                      SclkFcwRangeTable[NUM_SCLK_RANGE];
+
+	GB_VDROOP_TABLE_t                   BTCGB_VDROOP_TABLE[BTCGB_VDROOP_TABLE_MAX_ENTRIES];
+	SMU_QuadraticCoeffs                 AVFSGB_FUSE_TABLE[AVFSGB_VDROOP_TABLE_MAX_ENTRIES];
+};
+
+typedef struct SMU75_Discrete_DpmTable SMU75_Discrete_DpmTable;
+
+struct SMU75_Discrete_FanTable {
+	uint16_t FdoMode;
+	int16_t  TempMin;
+	int16_t  TempMed;
+	int16_t  TempMax;
+	int16_t  Slope1;
+	int16_t  Slope2;
+	int16_t  FdoMin;
+	int16_t  HystUp;
+	int16_t  HystDown;
+	int16_t  HystSlope;
+	int16_t  TempRespLim;
+	int16_t  TempCurr;
+	int16_t  SlopeCurr;
+	int16_t  PwmCurr;
+	uint32_t RefreshPeriod;
+	int16_t  FdoMax;
+	uint8_t  TempSrc;
+	int8_t   Padding;
+};
+
+typedef struct SMU75_Discrete_FanTable SMU75_Discrete_FanTable;
+
+#define SMU7_DISCRETE_GPIO_SCLK_DEBUG             4
+#define SMU7_DISCRETE_GPIO_SCLK_DEBUG_BIT         (0x1 << SMU7_DISCRETE_GPIO_SCLK_DEBUG)
+
+
+
+struct SMU7_MclkDpmScoreboard {
+	uint32_t PercentageBusy;
+
+	int32_t  PIDError;
+	int32_t  PIDIntegral;
+	int32_t  PIDOutput;
+
+	uint32_t SigmaDeltaAccum;
+	uint32_t SigmaDeltaOutput;
+	uint32_t SigmaDeltaLevel;
+
+	uint32_t UtilizationSetpoint;
+
+	uint8_t  TdpClampMode;
+	uint8_t  TdcClampMode;
+	uint8_t  ThermClampMode;
+	uint8_t  VoltageBusy;
+
+	int8_t   CurrLevel;
+	int8_t   TargLevel;
+	uint8_t  LevelChangeInProgress;
+	uint8_t  UpHyst;
+
+	uint8_t  DownHyst;
+	uint8_t  VoltageDownHyst;
+	uint8_t  DpmEnable;
+	uint8_t  DpmRunning;
+
+	uint8_t  DpmForce;
+	uint8_t  DpmForceLevel;
+	uint8_t  padding2;
+	uint8_t  McArbIndex;
+
+	uint32_t MinimumPerfMclk;
+
+	uint8_t  AcpiReq;
+	uint8_t  AcpiAck;
+	uint8_t  MclkSwitchInProgress;
+	uint8_t  MclkSwitchCritical;
+
+	uint8_t  IgnoreVBlank;
+	uint8_t  TargetMclkIndex;
+	uint8_t  TargetMvddIndex;
+	uint8_t  MclkSwitchResult;
+
+	uint16_t VbiFailureCount;
+	uint8_t  VbiWaitCounter;
+	uint8_t  EnabledLevelsChange;
+
+	uint16_t LevelResidencyCounters [SMU75_MAX_LEVELS_MEMORY];
+	uint16_t LevelSwitchCounters [SMU75_MAX_LEVELS_MEMORY];
+
+	void     (*TargetStateCalculator)(uint8_t);
+	void     (*SavedTargetStateCalculator)(uint8_t);
+
+	uint16_t AutoDpmInterval;
+	uint16_t AutoDpmRange;
+
+	uint16_t VbiTimeoutCount;
+	uint16_t MclkSwitchingTime;
+
+	uint8_t  fastSwitch;
+	uint8_t  Save_PIC_VDDGFX_EXIT;
+	uint8_t  Save_PIC_VDDGFX_ENTER;
+	uint8_t  VbiTimeout;
+
+	uint32_t HbmTempRegBackup;
+};
+
+typedef struct SMU7_MclkDpmScoreboard SMU7_MclkDpmScoreboard;
+
+struct SMU7_UlvScoreboard {
+	uint8_t     EnterUlv;
+	uint8_t     ExitUlv;
+	uint8_t     UlvActive;
+	uint8_t     WaitingForUlv;
+	uint8_t     UlvEnable;
+	uint8_t     UlvRunning;
+	uint8_t     UlvMasterEnable;
+	uint8_t     padding;
+	uint32_t    UlvAbortedCount;
+	uint32_t    UlvTimeStamp;
+};
+
+typedef struct SMU7_UlvScoreboard SMU7_UlvScoreboard;
+
+struct VddgfxSavedRegisters {
+	uint32_t GPU_DBG[3];
+	uint32_t MEC_BaseAddress_Hi;
+	uint32_t MEC_BaseAddress_Lo;
+	uint32_t THM_TMON0_CTRL2__RDIR_PRESENT;
+	uint32_t THM_TMON1_CTRL2__RDIR_PRESENT;
+	uint32_t CP_INT_CNTL;
+};
+
+typedef struct VddgfxSavedRegisters VddgfxSavedRegisters;
+
+struct SMU7_VddGfxScoreboard {
+	uint8_t     VddGfxEnable;
+	uint8_t     VddGfxActive;
+	uint8_t     VPUResetOccured;
+	uint8_t     padding;
+
+	uint32_t    VddGfxEnteredCount;
+	uint32_t    VddGfxAbortedCount;
+
+	uint32_t    VddGfxVid;
+
+	VddgfxSavedRegisters SavedRegisters;
+};
+
+typedef struct SMU7_VddGfxScoreboard SMU7_VddGfxScoreboard;
+
+struct SMU7_TdcLimitScoreboard {
+	uint8_t  Enable;
+	uint8_t  Running;
+	uint16_t Alpha;
+	uint32_t FilteredIddc;
+	uint32_t IddcLimit;
+	uint32_t IddcHyst;
+	SMU7_HystController_Data HystControllerData;
+};
+
+typedef struct SMU7_TdcLimitScoreboard SMU7_TdcLimitScoreboard;
+
+struct SMU7_PkgPwrLimitScoreboard {
+	uint8_t  Enable;
+	uint8_t  Running;
+	uint16_t Alpha;
+	uint32_t FilteredPkgPwr;
+	uint32_t Limit;
+	uint32_t Hyst;
+	uint32_t LimitFromDriver;
+	uint8_t PowerSharingEnabled;
+	uint8_t PowerSharingCounter;
+	uint8_t PowerSharingINTEnabled;
+	uint8_t GFXActivityCounterEnabled;
+	uint32_t EnergyCount;
+	uint32_t PSACTCount;
+	uint8_t RollOverRequired;
+	uint8_t RollOverCount;
+	uint8_t padding[2];
+	SMU7_HystController_Data HystControllerData;
+};
+
+typedef struct SMU7_PkgPwrLimitScoreboard SMU7_PkgPwrLimitScoreboard;
+
+struct SMU7_BapmScoreboard {
+	uint32_t source_powers[SMU75_DTE_SOURCES];
+	uint32_t source_powers_last[SMU75_DTE_SOURCES];
+	int32_t entity_temperatures[SMU75_NUM_GPU_TES];
+	int32_t initial_entity_temperatures[SMU75_NUM_GPU_TES];
+	int32_t Limit;
+	int32_t Hyst;
+	int32_t therm_influence_coeff_table[SMU75_DTE_ITERATIONS * SMU75_DTE_SOURCES * SMU75_DTE_SINKS * 2];
+	int32_t therm_node_table[SMU75_DTE_ITERATIONS * SMU75_DTE_SOURCES * SMU75_DTE_SINKS];
+	uint16_t ConfigTDPPowerScalar;
+	uint16_t FanSpeedPowerScalar;
+	uint16_t OverDrivePowerScalar;
+	uint16_t OverDriveLimitScalar;
+	uint16_t FinalPowerScalar;
+	uint8_t VariantID;
+	uint8_t spare997;
+
+	SMU7_HystController_Data HystControllerData;
+
+	int32_t temperature_gradient_slope;
+	int32_t temperature_gradient;
+	uint32_t measured_temperature;
+};
+
+
+typedef struct SMU7_BapmScoreboard SMU7_BapmScoreboard;
+
+struct SMU7_AcpiScoreboard {
+	uint32_t SavedInterruptMask[2];
+	uint8_t LastACPIRequest;
+	uint8_t CgBifResp;
+	uint8_t RequestType;
+	uint8_t Padding;
+	SMU75_Discrete_ACPILevel D0Level;
+};
+
+typedef struct SMU7_AcpiScoreboard SMU7_AcpiScoreboard;
+
+struct SMU75_Discrete_PmFuses {
+	uint8_t BapmVddCVidHiSidd[8];
+
+	uint8_t BapmVddCVidLoSidd[8];
+
+	uint8_t VddCVid[8];
+
+	uint8_t SviLoadLineEn;
+	uint8_t SviLoadLineVddC;
+	uint8_t SviLoadLineTrimVddC;
+	uint8_t SviLoadLineOffsetVddC;
+
+	uint16_t TDC_VDDC_PkgLimit;
+	uint8_t TDC_VDDC_ThrottleReleaseLimitPerc;
+	uint8_t TDC_MAWt;
+
+	uint8_t TdcWaterfallCtl;
+	uint8_t LPMLTemperatureMin;
+	uint8_t LPMLTemperatureMax;
+	uint8_t Reserved;
+
+	uint8_t LPMLTemperatureScaler[16];
+
+	int16_t FuzzyFan_ErrorSetDelta;
+	int16_t FuzzyFan_ErrorRateSetDelta;
+	int16_t FuzzyFan_PwmSetDelta;
+	uint16_t Reserved6;
+
+	uint8_t GnbLPML[16];
+
+	uint8_t GnbLPMLMaxVid;
+	uint8_t GnbLPMLMinVid;
+	uint8_t Reserved1[2];
+
+	uint16_t BapmVddCBaseLeakageHiSidd;
+	uint16_t BapmVddCBaseLeakageLoSidd;
+
+	uint16_t  VFT_Temp[3];
+	uint8_t   Version;
+	uint8_t   padding;
+
+	SMU_QuadraticCoeffs VFT_ATE[3];
+
+	SMU_QuadraticCoeffs AVFS_GB;
+	SMU_QuadraticCoeffs ATE_ACBTC_GB;
+
+	SMU_QuadraticCoeffs P2V;
+
+	uint32_t PsmCharzFreq;
+
+	uint16_t InversionVoltage;
+	uint16_t PsmCharzTemp;
+
+	uint32_t EnabledAvfsModules;
+
+	SMU_QuadraticCoeffs BtcGbv_CksOff;
+};
+
+typedef struct SMU75_Discrete_PmFuses SMU75_Discrete_PmFuses;
+
+struct SMU7_Discrete_Log_Header_Table {
+	uint32_t    version;
+	uint32_t    asic_id;
+	uint16_t    flags;
+	uint16_t    entry_size;
+	uint32_t    total_size;
+	uint32_t    num_of_entries;
+	uint8_t     type;
+	uint8_t     mode;
+	uint8_t     filler_0[2];
+	uint32_t    filler_1[2];
+};
+
+typedef struct SMU7_Discrete_Log_Header_Table SMU7_Discrete_Log_Header_Table;
+
+struct SMU7_Discrete_Log_Cntl {
+	uint8_t             Enabled;
+	uint8_t             Type;
+	uint8_t             padding[2];
+	uint32_t            BufferSize;
+	uint32_t            SamplesLogged;
+	uint32_t            SampleSize;
+	uint32_t            AddrL;
+	uint32_t            AddrH;
+};
+
+typedef struct SMU7_Discrete_Log_Cntl SMU7_Discrete_Log_Cntl;
+
+#if defined SMU__DGPU_ONLY
+#define CAC_ACC_NW_NUM_OF_SIGNALS 87
+#endif
+
+
+struct SMU7_Discrete_Cac_Collection_Table {
+	uint32_t temperature;
+	uint32_t cac_acc_nw[CAC_ACC_NW_NUM_OF_SIGNALS];
+};
+
+typedef struct SMU7_Discrete_Cac_Collection_Table SMU7_Discrete_Cac_Collection_Table;
+
+struct SMU7_Discrete_Cac_Verification_Table {
+	uint32_t VddcTotalPower;
+	uint32_t VddcLeakagePower;
+	uint32_t VddcConstantPower;
+	uint32_t VddcGfxDynamicPower;
+	uint32_t VddcUvdDynamicPower;
+	uint32_t VddcVceDynamicPower;
+	uint32_t VddcAcpDynamicPower;
+	uint32_t VddcPcieDynamicPower;
+	uint32_t VddcDceDynamicPower;
+	uint32_t VddcCurrent;
+	uint32_t VddcVoltage;
+	uint32_t VddciTotalPower;
+	uint32_t VddciLeakagePower;
+	uint32_t VddciConstantPower;
+	uint32_t VddciDynamicPower;
+	uint32_t Vddr1TotalPower;
+	uint32_t Vddr1LeakagePower;
+	uint32_t Vddr1ConstantPower;
+	uint32_t Vddr1DynamicPower;
+	uint32_t spare[4];
+	uint32_t temperature;
+};
+
+typedef struct SMU7_Discrete_Cac_Verification_Table SMU7_Discrete_Cac_Verification_Table;
+
+struct SMU7_Discrete_Pm_Status_Table {
+	int32_t T_meas_max[SMU75_THERMAL_INPUT_LOOP_COUNT];
+	int32_t T_meas_acc[SMU75_THERMAL_INPUT_LOOP_COUNT];
+
+	uint32_t I_calc_max;
+	uint32_t I_calc_acc;
+	uint32_t P_meas_acc;
+	uint32_t V_meas_load_acc;
+	uint32_t I_meas_acc;
+	uint32_t P_meas_acc_vddci;
+	uint32_t V_meas_load_acc_vddci;
+	uint32_t I_meas_acc_vddci;
+
+	uint16_t Sclk_dpm_residency[8];
+	uint16_t Uvd_dpm_residency[8];
+	uint16_t Vce_dpm_residency[8];
+	uint16_t Mclk_dpm_residency[4];
+
+	uint32_t P_roc_acc;
+	uint32_t PkgPwr_max;
+	uint32_t PkgPwr_acc;
+	uint32_t MclkSwitchingTime_max;
+	uint32_t MclkSwitchingTime_acc;
+	uint32_t FanPwm_acc;
+	uint32_t FanRpm_acc;
+	uint32_t Gfx_busy_acc;
+	uint32_t Mc_busy_acc;
+	uint32_t Fps_acc;
+
+	uint32_t AccCnt;
+};
+
+typedef struct SMU7_Discrete_Pm_Status_Table SMU7_Discrete_Pm_Status_Table;
+
+struct SMU7_Discrete_AutoWattMan_Status_Table {
+	int32_t T_meas_acc[SMU75_THERMAL_INPUT_LOOP_COUNT];
+	uint16_t Sclk_dpm_residency[8];
+	uint16_t Mclk_dpm_residency[4];
+	uint32_t TgpPwr_acc;
+	uint32_t Gfx_busy_acc;
+	uint32_t Mc_busy_acc;
+	uint32_t AccCnt;
+};
+
+typedef struct SMU7_Discrete_AutoWattMan_Status_Table SMU7_Discrete_AutoWattMan_Status_Table;
+
+#define SMU7_MAX_GFX_CU_COUNT 24
+#define SMU7_MIN_GFX_CU_COUNT  8
+#define SMU7_GFX_CU_PG_ENABLE_DC_MAX_CU_SHIFT 0
+#define SMU7_GFX_CU_PG_ENABLE_DC_MAX_CU_MASK  (0xFFFF << SMU7_GFX_CU_PG_ENABLE_DC_MAX_CU_SHIFT)
+#define SMU7_GFX_CU_PG_ENABLE_AC_MAX_CU_SHIFT 16
+#define SMU7_GFX_CU_PG_ENABLE_AC_MAX_CU_MASK  (0xFFFF << SMU7_GFX_CU_PG_ENABLE_AC_MAX_CU_SHIFT)
+
+struct SMU7_GfxCuPgScoreboard {
+	uint8_t Enabled;
+	uint8_t WaterfallUp;
+	uint8_t WaterfallDown;
+	uint8_t WaterfallLimit;
+	uint8_t CurrMaxCu;
+	uint8_t TargMaxCu;
+	uint8_t ClampMode;
+	uint8_t Active;
+	uint8_t MaxSupportedCu;
+	uint8_t MinSupportedCu;
+	uint8_t PendingGfxCuHostInterrupt;
+	uint8_t LastFilteredMaxCuInteger;
+	uint16_t FilteredMaxCu;
+	uint16_t FilteredMaxCuAlpha;
+	uint16_t FilterResetCount;
+	uint16_t FilterResetCountLimit;
+	uint8_t ForceCu;
+	uint8_t ForceCuCount;
+	uint8_t AcModeMaxCu;
+	uint8_t DcModeMaxCu;
+};
+
+typedef struct SMU7_GfxCuPgScoreboard SMU7_GfxCuPgScoreboard;
+
+#define SMU7_SCLK_CAC 0x561
+#define SMU7_MCLK_CAC 0xF9
+#define SMU7_VCLK_CAC 0x2DE
+#define SMU7_DCLK_CAC 0x2DE
+#define SMU7_ECLK_CAC 0x25E
+#define SMU7_ACLK_CAC 0x25E
+#define SMU7_SAMCLK_CAC 0x25E
+#define SMU7_DISPCLK_CAC 0x100
+#define SMU7_CAC_CONSTANT 0x2EE3430
+#define SMU7_CAC_CONSTANT_SHIFT 18
+
+#define SMU7_VDDCI_MCLK_CONST        1765
+#define SMU7_VDDCI_MCLK_CONST_SHIFT  16
+#define SMU7_VDDCI_VDDCI_CONST       50958
+#define SMU7_VDDCI_VDDCI_CONST_SHIFT 14
+#define SMU7_VDDCI_CONST             11781
+#define SMU7_VDDCI_STROBE_PWR        1331
+
+#define SMU7_VDDR1_CONST            693
+#define SMU7_VDDR1_CAC_WEIGHT       20
+#define SMU7_VDDR1_CAC_WEIGHT_SHIFT 19
+#define SMU7_VDDR1_STROBE_PWR       512
+
+#define SMU7_AREA_COEFF_UVD 0xA78
+#define SMU7_AREA_COEFF_VCE 0x190A
+#define SMU7_AREA_COEFF_ACP 0x22D1
+#define SMU7_AREA_COEFF_SAMU 0x534
+
+#define SMU7_THERM_OUT_MODE_DISABLE       0x0
+#define SMU7_THERM_OUT_MODE_THERM_ONLY    0x1
+#define SMU7_THERM_OUT_MODE_THERM_VRHOT   0x2
+
+#define SQ_Enable_MASK 0x1
+#define SQ_IR_MASK 0x2
+#define SQ_PCC_MASK 0x4
+#define SQ_EDC_MASK 0x8
+
+#define TCP_Enable_MASK 0x100
+#define TCP_IR_MASK 0x200
+#define TCP_PCC_MASK 0x400
+#define TCP_EDC_MASK 0x800
+
+#define TD_Enable_MASK 0x10000
+#define TD_IR_MASK 0x20000
+#define TD_PCC_MASK 0x40000
+#define TD_EDC_MASK 0x80000
+
+#define DB_Enable_MASK 0x1000000
+#define DB_IR_MASK 0x2000000
+#define DB_PCC_MASK 0x4000000
+#define DB_EDC_MASK 0x8000000
+
+#define SQ_Enable_SHIFT 0
+#define SQ_IR_SHIFT 1
+#define SQ_PCC_SHIFT 2
+#define SQ_EDC_SHIFT 3
+
+#define TCP_Enable_SHIFT 8
+#define TCP_IR_SHIFT 9
+#define TCP_PCC_SHIFT 10
+#define TCP_EDC_SHIFT 11
+
+#define TD_Enable_SHIFT 16
+#define TD_IR_SHIFT 17
+#define TD_PCC_SHIFT 18
+#define TD_EDC_SHIFT 19
+
+#define DB_Enable_SHIFT 24
+#define DB_IR_SHIFT 25
+#define DB_PCC_SHIFT 26
+#define DB_EDC_SHIFT 27
+
+#define PMFUSES_AVFSSIZE 104
+
+#define BTCGB0_Vdroop_Enable_MASK  0x1
+#define BTCGB1_Vdroop_Enable_MASK  0x2
+#define AVFSGB0_Vdroop_Enable_MASK 0x4
+#define AVFSGB1_Vdroop_Enable_MASK 0x8
+
+#define BTCGB0_Vdroop_Enable_SHIFT  0
+#define BTCGB1_Vdroop_Enable_SHIFT  1
+#define AVFSGB0_Vdroop_Enable_SHIFT 2
+#define AVFSGB1_Vdroop_Enable_SHIFT 3
+
+#pragma pack(pop)
+
+
+#endif
+
-- 
cgit v1.1


From ee38fbf3bfb58c427dc3670a70cc6df4ebe1c785 Mon Sep 17 00:00:00 2001
From: Eric Huang <JinHuiEric.Huang@amd.com>
Date: Fri, 17 Nov 2017 11:17:48 -0500
Subject: drm/amd: add a new struct in atombios.h

Signed-off-by: Eric Huang <JinHuiEric.Huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/include/atombios.h | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/drivers/gpu/drm/amd/include/atombios.h b/drivers/gpu/drm/amd/include/atombios.h
index f696bbb..7931502 100644
--- a/drivers/gpu/drm/amd/include/atombios.h
+++ b/drivers/gpu/drm/amd/include/atombios.h
@@ -632,6 +632,13 @@ typedef struct _COMPUTE_MEMORY_CLOCK_PARAM_PARAMETERS_V2_2
   ULONG ulReserved;
 }COMPUTE_MEMORY_CLOCK_PARAM_PARAMETERS_V2_2;
 
+typedef struct _COMPUTE_MEMORY_CLOCK_PARAM_PARAMETERS_V2_3
+{
+  COMPUTE_MEMORY_ENGINE_PLL_PARAMETERS_V4 ulClock;
+  USHORT  usMclk_fcw_frac;                  //fractional divider of fcw = usSclk_fcw_frac/65536
+  USHORT  usMclk_fcw_int;                   //integer divider of fcwc
+}COMPUTE_MEMORY_CLOCK_PARAM_PARAMETERS_V2_3;
+
 //Input parameter of DynamicMemorySettingsTable
 //when ATOM_COMPUTE_CLOCK_FREQ.ulComputeClockFlag = COMPUTE_MEMORY_PLL_PARAM
 typedef struct _DYNAMICE_MEMORY_SETTINGS_PARAMETER
-- 
cgit v1.1


From 4eeed17e713b9e6494a08ab37623283723596b5a Mon Sep 17 00:00:00 2001
From: Eric Huang <JinHuiEric.Huang@amd.com>
Date: Fri, 17 Nov 2017 11:21:02 -0500
Subject: drm/amd/powerplay: update ppatomctrl.c (v2)

used for calculating memory clocks in powerplay.

v2: handle endian swapping of atom data (Alex)

Signed-off-by: Eric Huang <JinHuiEric.Huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c | 31 ++++++++++++++++++++++++
 drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h | 10 ++++++++
 2 files changed, 41 insertions(+)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c
index d58be7e..cf99c5e 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c
@@ -23,6 +23,7 @@
 #include "pp_debug.h"
 #include <linux/module.h>
 #include <linux/slab.h>
+#include <linux/delay.h>
 #include "atom.h"
 #include "ppatomctrl.h"
 #include "atombios.h"
@@ -314,6 +315,36 @@ int atomctrl_get_memory_pll_dividers_vi(struct pp_hwmgr *hwmgr,
 	return result;
 }
 
+int atomctrl_get_memory_pll_dividers_ai(struct pp_hwmgr *hwmgr,
+					uint32_t clock_value,
+					pp_atomctrl_memory_clock_param_ai *mpll_param)
+{
+	struct amdgpu_device *adev = hwmgr->adev;
+	COMPUTE_MEMORY_CLOCK_PARAM_PARAMETERS_V2_3 mpll_parameters = {0};
+	int result;
+
+	mpll_parameters.ulClock.ulClock = cpu_to_le32(clock_value);
+
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
+			GetIndexIntoMasterTable(COMMAND, ComputeMemoryClockParam),
+			(uint32_t *)&mpll_parameters);
+
+	/* VEGAM's mpll takes sometime to finish computing */
+	udelay(10);
+
+	if (!result) {
+		mpll_param->ulMclk_fcw_int =
+			le16_to_cpu(mpll_parameters.usMclk_fcw_int);
+		mpll_param->ulMclk_fcw_frac =
+			le16_to_cpu(mpll_parameters.usMclk_fcw_frac);
+		mpll_param->ulClock =
+			le32_to_cpu(mpll_parameters.ulClock.ulClock);
+		mpll_param->ulPostDiv = mpll_parameters.ulClock.ucPostDiv;
+	}
+
+	return result;
+}
+
 int atomctrl_get_engine_pll_dividers_kong(struct pp_hwmgr *hwmgr,
 					  uint32_t clock_value,
 					  pp_atomctrl_clock_dividers_kong *dividers)
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h
index e1b5d6b..3ee54f1 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h
@@ -146,6 +146,14 @@ struct pp_atomctrl_memory_clock_param {
 };
 typedef struct pp_atomctrl_memory_clock_param pp_atomctrl_memory_clock_param;
 
+struct pp_atomctrl_memory_clock_param_ai {
+	uint32_t ulClock;
+	uint32_t ulPostDiv;
+	uint16_t ulMclk_fcw_frac;
+	uint16_t ulMclk_fcw_int;
+};
+typedef struct pp_atomctrl_memory_clock_param_ai pp_atomctrl_memory_clock_param_ai;
+
 struct pp_atomctrl_internal_ss_info {
 	uint32_t speed_spectrum_percentage;                      /* in 1/100 percentage */
 	uint32_t speed_spectrum_rate;                            /* in KHz */
@@ -295,6 +303,8 @@ extern bool atomctrl_is_voltage_controlled_by_gpio_v3(struct pp_hwmgr *hwmgr, ui
 extern int atomctrl_get_voltage_table_v3(struct pp_hwmgr *hwmgr, uint8_t voltage_type, uint8_t voltage_mode, pp_atomctrl_voltage_table *voltage_table);
 extern int atomctrl_get_memory_pll_dividers_vi(struct pp_hwmgr *hwmgr,
 		uint32_t clock_value, pp_atomctrl_memory_clock_param *mpll_param);
+extern int atomctrl_get_memory_pll_dividers_ai(struct pp_hwmgr *hwmgr,
+		uint32_t clock_value, pp_atomctrl_memory_clock_param_ai *mpll_param);
 extern int atomctrl_get_engine_pll_dividers_kong(struct pp_hwmgr *hwmgr,
 						 uint32_t clock_value,
 						 pp_atomctrl_clock_dividers_kong *dividers);
-- 
cgit v1.1


From 4dc1a2d9288dbba903696d2dd5d83b5311f2d026 Mon Sep 17 00:00:00 2001
From: Eric Huang <JinHuiEric.Huang@amd.com>
Date: Fri, 17 Nov 2017 11:31:09 -0500
Subject: drm/amd/powerplay: update process pptables

Add functionality to fetch gpio table from vbios.

Signed-off-by: Eric Huang <JinHuiEric.Huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../amd/powerplay/hwmgr/process_pptables_v1_0.c    | 37 ++++++++++++++++++++++
 1 file changed, 37 insertions(+)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.c b/drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.c
index 8516516..f0d48b1 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.c
@@ -728,6 +728,32 @@ static int get_mm_clock_voltage_table(
 	return 0;
 }
 
+static int get_gpio_table(struct pp_hwmgr *hwmgr,
+		struct phm_ppt_v1_gpio_table **pp_tonga_gpio_table,
+		const ATOM_Tonga_GPIO_Table *atom_gpio_table)
+{
+	uint32_t table_size;
+	struct phm_ppt_v1_gpio_table *pp_gpio_table;
+	struct phm_ppt_v1_information *pp_table_information =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+
+	table_size = sizeof(struct phm_ppt_v1_gpio_table);
+	pp_gpio_table = kzalloc(table_size, GFP_KERNEL);
+	if (!pp_gpio_table)
+		return -ENOMEM;
+
+	if (pp_table_information->vdd_dep_on_sclk->count <
+			atom_gpio_table->ucVRHotTriggeredSclkDpmIndex)
+		PP_ASSERT_WITH_CODE(false,
+				"SCLK DPM index for VRHot cannot exceed the total sclk level count!",);
+	else
+		pp_gpio_table->vrhot_triggered_sclk_dpm_index =
+				atom_gpio_table->ucVRHotTriggeredSclkDpmIndex;
+
+	*pp_tonga_gpio_table = pp_gpio_table;
+
+	return 0;
+}
 /**
  * Private Function used during initialization.
  * Initialize clock voltage dependency
@@ -761,11 +787,15 @@ static int init_clock_voltage_dependency(
 	const PPTable_Generic_SubTable_Header *pcie_table =
 		(const PPTable_Generic_SubTable_Header *)(((unsigned long) powerplay_table) +
 		le16_to_cpu(powerplay_table->usPCIETableOffset));
+	const ATOM_Tonga_GPIO_Table *gpio_table =
+		(const ATOM_Tonga_GPIO_Table *)(((unsigned long) powerplay_table) +
+		le16_to_cpu(powerplay_table->usGPIOTableOffset));
 
 	pp_table_information->vdd_dep_on_sclk = NULL;
 	pp_table_information->vdd_dep_on_mclk = NULL;
 	pp_table_information->mm_dep_table = NULL;
 	pp_table_information->pcie_table = NULL;
+	pp_table_information->gpio_table = NULL;
 
 	if (powerplay_table->usMMDependencyTableOffset != 0)
 		result = get_mm_clock_voltage_table(hwmgr,
@@ -810,6 +840,10 @@ static int init_clock_voltage_dependency(
 		result = get_valid_clk(hwmgr, &pp_table_information->valid_sclk_values,
 		pp_table_information->vdd_dep_on_sclk);
 
+	if (!result && gpio_table)
+		result = get_gpio_table(hwmgr, &pp_table_information->gpio_table,
+				gpio_table);
+
 	return result;
 }
 
@@ -1116,6 +1150,9 @@ static int pp_tables_v1_0_uninitialize(struct pp_hwmgr *hwmgr)
 	kfree(pp_table_information->pcie_table);
 	pp_table_information->pcie_table = NULL;
 
+	kfree(pp_table_information->gpio_table);
+	pp_table_information->gpio_table = NULL;
+
 	kfree(hwmgr->pptable);
 	hwmgr->pptable = NULL;
 
-- 
cgit v1.1


From ac7822b0026fbc33f82023b155542426b1bd211b Mon Sep 17 00:00:00 2001
From: Eric Huang <JinHuiEric.Huang@amd.com>
Date: Wed, 11 Apr 2018 15:32:58 -0500
Subject: drm/amd/powerplay: add smumgr support for VEGAM (v2)

The smumgr handles communication between the driver
and the SMU for power management.

v2: fix typo (Alex)

Signed-off-by: Eric Huang <JinHuiEric.Huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c        |    6 +
 drivers/gpu/drm/amd/powerplay/smumgr/Makefile      |    2 +-
 .../gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c    | 2382 ++++++++++++++++++++
 .../gpu/drm/amd/powerplay/smumgr/vegam_smumgr.h    |   75 +
 4 files changed, 2464 insertions(+), 1 deletion(-)
 create mode 100644 drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c
 create mode 100644 drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.h

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
index d1052b5..eecb118 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
@@ -40,6 +40,7 @@ extern const struct pp_smumgr_func iceland_smu_funcs;
 extern const struct pp_smumgr_func tonga_smu_funcs;
 extern const struct pp_smumgr_func fiji_smu_funcs;
 extern const struct pp_smumgr_func polaris10_smu_funcs;
+extern const struct pp_smumgr_func vegam_smu_funcs;
 extern const struct pp_smumgr_func vega10_smu_funcs;
 extern const struct pp_smumgr_func vega12_smu_funcs;
 extern const struct pp_smumgr_func smu10_smu_funcs;
@@ -136,6 +137,11 @@ int hwmgr_early_init(struct pp_hwmgr *hwmgr)
 			polaris_set_asic_special_caps(hwmgr);
 			hwmgr->feature_mask &= ~(PP_UVD_HANDSHAKE_MASK);
 			break;
+		case CHIP_VEGAM:
+			hwmgr->smumgr_funcs = &vegam_smu_funcs;
+			polaris_set_asic_special_caps(hwmgr);
+			hwmgr->feature_mask &= ~(PP_UVD_HANDSHAKE_MASK);
+			break;
 		default:
 			return -EINVAL;
 		}
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/Makefile b/drivers/gpu/drm/amd/powerplay/smumgr/Makefile
index 9587550..0a20040 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/Makefile
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/Makefile
@@ -26,7 +26,7 @@
 SMU_MGR = smumgr.o smu8_smumgr.o tonga_smumgr.o fiji_smumgr.o \
 	  polaris10_smumgr.o iceland_smumgr.o \
 	  smu7_smumgr.o vega10_smumgr.o smu10_smumgr.o ci_smumgr.o \
-	  vega12_smumgr.o
+	  vega12_smumgr.o vegam_smumgr.o
 
 AMD_PP_SMUMGR = $(addprefix $(AMD_PP_PATH)/smumgr/,$(SMU_MGR))
 
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c
new file mode 100644
index 0000000..c9a5633
--- /dev/null
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c
@@ -0,0 +1,2382 @@
+/*
+ * Copyright 2017 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#include "pp_debug.h"
+#include "smumgr.h"
+#include "smu_ucode_xfer_vi.h"
+#include "vegam_smumgr.h"
+#include "smu/smu_7_1_3_d.h"
+#include "smu/smu_7_1_3_sh_mask.h"
+#include "gmc/gmc_8_1_d.h"
+#include "gmc/gmc_8_1_sh_mask.h"
+#include "oss/oss_3_0_d.h"
+#include "gca/gfx_8_0_d.h"
+#include "bif/bif_5_0_d.h"
+#include "bif/bif_5_0_sh_mask.h"
+#include "ppatomctrl.h"
+#include "cgs_common.h"
+#include "smu7_ppsmc.h"
+
+#include "smu7_dyn_defaults.h"
+
+#include "smu7_hwmgr.h"
+#include "hardwaremanager.h"
+#include "ppatomctrl.h"
+#include "atombios.h"
+#include "pppcielanes.h"
+
+#include "dce/dce_11_2_d.h"
+#include "dce/dce_11_2_sh_mask.h"
+
+#define PPVEGAM_TARGETACTIVITY_DFLT                     50
+
+#define VOLTAGE_VID_OFFSET_SCALE1   625
+#define VOLTAGE_VID_OFFSET_SCALE2   100
+#define POWERTUNE_DEFAULT_SET_MAX    1
+#define VDDC_VDDCI_DELTA            200
+#define MC_CG_ARB_FREQ_F1           0x0b
+
+#define STRAP_ASIC_RO_LSB    2168
+#define STRAP_ASIC_RO_MSB    2175
+
+#define PPSMC_MSG_ApplyAvfsCksOffVoltage      ((uint16_t) 0x415)
+#define PPSMC_MSG_EnableModeSwitchRLCNotification  ((uint16_t) 0x305)
+
+static const struct vegam_pt_defaults
+vegam_power_tune_data_set_array[POWERTUNE_DEFAULT_SET_MAX] = {
+	/* sviLoadLIneEn, SviLoadLineVddC, TDC_VDDC_ThrottleReleaseLimitPerc, TDC_MAWt,
+	 * TdcWaterfallCtl, DTEAmbientTempBase, DisplayCac, BAPM_TEMP_GRADIENT */
+	{ 1, 0xF, 0xFD, 0x19, 5, 45, 0, 0xB0000,
+	{ 0x79, 0x253, 0x25D, 0xAE, 0x72, 0x80, 0x83, 0x86, 0x6F, 0xC8, 0xC9, 0xC9, 0x2F, 0x4D, 0x61},
+	{ 0x17C, 0x172, 0x180, 0x1BC, 0x1B3, 0x1BD, 0x206, 0x200, 0x203, 0x25D, 0x25A, 0x255, 0x2C3, 0x2C5, 0x2B4 } },
+};
+
+static const sclkFcwRange_t Range_Table[NUM_SCLK_RANGE] = {
+			{VCO_2_4, POSTDIV_DIV_BY_16,  75, 160, 112},
+			{VCO_3_6, POSTDIV_DIV_BY_16, 112, 224, 160},
+			{VCO_2_4, POSTDIV_DIV_BY_8,   75, 160, 112},
+			{VCO_3_6, POSTDIV_DIV_BY_8,  112, 224, 160},
+			{VCO_2_4, POSTDIV_DIV_BY_4,   75, 160, 112},
+			{VCO_3_6, POSTDIV_DIV_BY_4,  112, 216, 160},
+			{VCO_2_4, POSTDIV_DIV_BY_2,   75, 160, 108},
+			{VCO_3_6, POSTDIV_DIV_BY_2,  112, 216, 160} };
+
+static int vegam_smu_init(struct pp_hwmgr *hwmgr)
+{
+	struct vegam_smumgr *smu_data;
+
+	smu_data = kzalloc(sizeof(struct vegam_smumgr), GFP_KERNEL);
+	if (smu_data == NULL)
+		return -ENOMEM;
+
+	hwmgr->smu_backend = smu_data;
+
+	if (smu7_init(hwmgr)) {
+		kfree(smu_data);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int vegam_start_smu_in_protection_mode(struct pp_hwmgr *hwmgr)
+{
+	int result = 0;
+
+	/* Wait for smc boot up */
+	/* PHM_WAIT_VFPF_INDIRECT_FIELD_UNEQUAL(smumgr, SMC_IND, RCU_UC_EVENTS, boot_seq_done, 0) */
+
+	/* Assert reset */
+	PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
+					SMC_SYSCON_RESET_CNTL, rst_reg, 1);
+
+	result = smu7_upload_smu_firmware_image(hwmgr);
+	if (result != 0)
+		return result;
+
+	/* Clear status */
+	cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixSMU_STATUS, 0);
+
+	PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
+					SMC_SYSCON_CLOCK_CNTL_0, ck_disable, 0);
+
+	/* De-assert reset */
+	PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
+					SMC_SYSCON_RESET_CNTL, rst_reg, 0);
+
+
+	PHM_WAIT_VFPF_INDIRECT_FIELD(hwmgr, SMC_IND, RCU_UC_EVENTS, INTERRUPTS_ENABLED, 1);
+
+
+	/* Call Test SMU message with 0x20000 offset to trigger SMU start */
+	smu7_send_msg_to_smc_offset(hwmgr);
+
+	/* Wait done bit to be set */
+	/* Check pass/failed indicator */
+
+	PHM_WAIT_VFPF_INDIRECT_FIELD_UNEQUAL(hwmgr, SMC_IND, SMU_STATUS, SMU_DONE, 0);
+
+	if (1 != PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
+						SMU_STATUS, SMU_PASS))
+		PP_ASSERT_WITH_CODE(false, "SMU Firmware start failed!", return -1);
+
+	cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixFIRMWARE_FLAGS, 0);
+
+	PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
+					SMC_SYSCON_RESET_CNTL, rst_reg, 1);
+
+	PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
+					SMC_SYSCON_RESET_CNTL, rst_reg, 0);
+
+	/* Wait for firmware to initialize */
+	PHM_WAIT_VFPF_INDIRECT_FIELD(hwmgr, SMC_IND, FIRMWARE_FLAGS, INTERRUPTS_ENABLED, 1);
+
+	return result;
+}
+
+static int vegam_start_smu_in_non_protection_mode(struct pp_hwmgr *hwmgr)
+{
+	int result = 0;
+
+	/* wait for smc boot up */
+	PHM_WAIT_VFPF_INDIRECT_FIELD_UNEQUAL(hwmgr, SMC_IND, RCU_UC_EVENTS, boot_seq_done, 0);
+
+	/* Clear firmware interrupt enable flag */
+	/* PHM_WRITE_VFPF_INDIRECT_FIELD(pSmuMgr, SMC_IND, SMC_SYSCON_MISC_CNTL, pre_fetcher_en, 1); */
+	cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+				ixFIRMWARE_FLAGS, 0);
+
+	PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
+					SMC_SYSCON_RESET_CNTL,
+					rst_reg, 1);
+
+	result = smu7_upload_smu_firmware_image(hwmgr);
+	if (result != 0)
+		return result;
+
+	/* Set smc instruct start point at 0x0 */
+	smu7_program_jump_on_start(hwmgr);
+
+	PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
+					SMC_SYSCON_CLOCK_CNTL_0, ck_disable, 0);
+
+	PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
+					SMC_SYSCON_RESET_CNTL, rst_reg, 0);
+
+	/* Wait for firmware to initialize */
+
+	PHM_WAIT_VFPF_INDIRECT_FIELD(hwmgr, SMC_IND,
+					FIRMWARE_FLAGS, INTERRUPTS_ENABLED, 1);
+
+	return result;
+}
+
+static int vegam_start_smu(struct pp_hwmgr *hwmgr)
+{
+	int result = 0;
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+
+	/* Only start SMC if SMC RAM is not running */
+	if (!smu7_is_smc_ram_running(hwmgr) && hwmgr->not_vf) {
+		smu_data->protected_mode = (uint8_t)(PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device,
+				CGS_IND_REG__SMC, SMU_FIRMWARE, SMU_MODE));
+		smu_data->smu7_data.security_hard_key = (uint8_t)(PHM_READ_VFPF_INDIRECT_FIELD(
+				hwmgr->device, CGS_IND_REG__SMC, SMU_FIRMWARE, SMU_SEL));
+
+		/* Check if SMU is running in protected mode */
+		if (smu_data->protected_mode == 0)
+			result = vegam_start_smu_in_non_protection_mode(hwmgr);
+		else
+			result = vegam_start_smu_in_protection_mode(hwmgr);
+
+		if (result != 0)
+			PP_ASSERT_WITH_CODE(0, "Failed to load SMU ucode.", return result);
+	}
+
+	/* Setup SoftRegsStart here for register lookup in case DummyBackEnd is used and ProcessFirmwareHeader is not executed */
+	smu7_read_smc_sram_dword(hwmgr,
+			SMU7_FIRMWARE_HEADER_LOCATION + offsetof(SMU75_Firmware_Header, SoftRegisters),
+			&(smu_data->smu7_data.soft_regs_start),
+			0x40000);
+
+	result = smu7_request_smu_load_fw(hwmgr);
+
+	return result;
+}
+
+static int vegam_process_firmware_header(struct pp_hwmgr *hwmgr)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	uint32_t tmp;
+	int result;
+	bool error = false;
+
+	result = smu7_read_smc_sram_dword(hwmgr,
+			SMU7_FIRMWARE_HEADER_LOCATION +
+			offsetof(SMU75_Firmware_Header, DpmTable),
+			&tmp, SMC_RAM_END);
+
+	if (0 == result)
+		smu_data->smu7_data.dpm_table_start = tmp;
+
+	error |= (0 != result);
+
+	result = smu7_read_smc_sram_dword(hwmgr,
+			SMU7_FIRMWARE_HEADER_LOCATION +
+			offsetof(SMU75_Firmware_Header, SoftRegisters),
+			&tmp, SMC_RAM_END);
+
+	if (!result) {
+		data->soft_regs_start = tmp;
+		smu_data->smu7_data.soft_regs_start = tmp;
+	}
+
+	error |= (0 != result);
+
+	result = smu7_read_smc_sram_dword(hwmgr,
+			SMU7_FIRMWARE_HEADER_LOCATION +
+			offsetof(SMU75_Firmware_Header, mcRegisterTable),
+			&tmp, SMC_RAM_END);
+
+	if (!result)
+		smu_data->smu7_data.mc_reg_table_start = tmp;
+
+	result = smu7_read_smc_sram_dword(hwmgr,
+			SMU7_FIRMWARE_HEADER_LOCATION +
+			offsetof(SMU75_Firmware_Header, FanTable),
+			&tmp, SMC_RAM_END);
+
+	if (!result)
+		smu_data->smu7_data.fan_table_start = tmp;
+
+	error |= (0 != result);
+
+	result = smu7_read_smc_sram_dword(hwmgr,
+			SMU7_FIRMWARE_HEADER_LOCATION +
+			offsetof(SMU75_Firmware_Header, mcArbDramTimingTable),
+			&tmp, SMC_RAM_END);
+
+	if (!result)
+		smu_data->smu7_data.arb_table_start = tmp;
+
+	error |= (0 != result);
+
+	result = smu7_read_smc_sram_dword(hwmgr,
+			SMU7_FIRMWARE_HEADER_LOCATION +
+			offsetof(SMU75_Firmware_Header, Version),
+			&tmp, SMC_RAM_END);
+
+	if (!result)
+		hwmgr->microcode_version_info.SMC = tmp;
+
+	error |= (0 != result);
+
+	return error ? -1 : 0;
+}
+
+static bool vegam_is_dpm_running(struct pp_hwmgr *hwmgr)
+{
+	return (1 == PHM_READ_INDIRECT_FIELD(hwmgr->device,
+			CGS_IND_REG__SMC, FEATURE_STATUS, VOLTAGE_CONTROLLER_ON))
+			? true : false;
+}
+
+static uint32_t vegam_get_mac_definition(uint32_t value)
+{
+	switch (value) {
+	case SMU_MAX_LEVELS_GRAPHICS:
+		return SMU75_MAX_LEVELS_GRAPHICS;
+	case SMU_MAX_LEVELS_MEMORY:
+		return SMU75_MAX_LEVELS_MEMORY;
+	case SMU_MAX_LEVELS_LINK:
+		return SMU75_MAX_LEVELS_LINK;
+	case SMU_MAX_ENTRIES_SMIO:
+		return SMU75_MAX_ENTRIES_SMIO;
+	case SMU_MAX_LEVELS_VDDC:
+		return SMU75_MAX_LEVELS_VDDC;
+	case SMU_MAX_LEVELS_VDDGFX:
+		return SMU75_MAX_LEVELS_VDDGFX;
+	case SMU_MAX_LEVELS_VDDCI:
+		return SMU75_MAX_LEVELS_VDDCI;
+	case SMU_MAX_LEVELS_MVDD:
+		return SMU75_MAX_LEVELS_MVDD;
+	case SMU_UVD_MCLK_HANDSHAKE_DISABLE:
+		return SMU7_UVD_MCLK_HANDSHAKE_DISABLE |
+				SMU7_VCE_MCLK_HANDSHAKE_DISABLE;
+	}
+
+	pr_warn("can't get the mac of %x\n", value);
+	return 0;
+}
+
+static int vegam_update_uvd_smc_table(struct pp_hwmgr *hwmgr)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	uint32_t mm_boot_level_offset, mm_boot_level_value;
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+
+	smu_data->smc_state_table.UvdBootLevel = 0;
+	if (table_info->mm_dep_table->count > 0)
+		smu_data->smc_state_table.UvdBootLevel =
+				(uint8_t) (table_info->mm_dep_table->count - 1);
+	mm_boot_level_offset = smu_data->smu7_data.dpm_table_start + offsetof(SMU75_Discrete_DpmTable,
+						UvdBootLevel);
+	mm_boot_level_offset /= 4;
+	mm_boot_level_offset *= 4;
+	mm_boot_level_value = cgs_read_ind_register(hwmgr->device,
+			CGS_IND_REG__SMC, mm_boot_level_offset);
+	mm_boot_level_value &= 0x00FFFFFF;
+	mm_boot_level_value |= smu_data->smc_state_table.UvdBootLevel << 24;
+	cgs_write_ind_register(hwmgr->device,
+			CGS_IND_REG__SMC, mm_boot_level_offset, mm_boot_level_value);
+
+	if (!phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+			PHM_PlatformCaps_UVDDPM) ||
+		phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+			PHM_PlatformCaps_StablePState))
+		smum_send_msg_to_smc_with_parameter(hwmgr,
+				PPSMC_MSG_UVDDPM_SetEnabledMask,
+				(uint32_t)(1 << smu_data->smc_state_table.UvdBootLevel));
+	return 0;
+}
+
+static int vegam_update_vce_smc_table(struct pp_hwmgr *hwmgr)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	uint32_t mm_boot_level_offset, mm_boot_level_value;
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+
+	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+					PHM_PlatformCaps_StablePState))
+		smu_data->smc_state_table.VceBootLevel =
+			(uint8_t) (table_info->mm_dep_table->count - 1);
+	else
+		smu_data->smc_state_table.VceBootLevel = 0;
+
+	mm_boot_level_offset = smu_data->smu7_data.dpm_table_start +
+					offsetof(SMU75_Discrete_DpmTable, VceBootLevel);
+	mm_boot_level_offset /= 4;
+	mm_boot_level_offset *= 4;
+	mm_boot_level_value = cgs_read_ind_register(hwmgr->device,
+			CGS_IND_REG__SMC, mm_boot_level_offset);
+	mm_boot_level_value &= 0xFF00FFFF;
+	mm_boot_level_value |= smu_data->smc_state_table.VceBootLevel << 16;
+	cgs_write_ind_register(hwmgr->device,
+			CGS_IND_REG__SMC, mm_boot_level_offset, mm_boot_level_value);
+
+	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_StablePState))
+		smum_send_msg_to_smc_with_parameter(hwmgr,
+				PPSMC_MSG_VCEDPM_SetEnabledMask,
+				(uint32_t)1 << smu_data->smc_state_table.VceBootLevel);
+	return 0;
+}
+
+static int vegam_update_samu_smc_table(struct pp_hwmgr *hwmgr)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	uint32_t mm_boot_level_offset, mm_boot_level_value;
+
+
+	smu_data->smc_state_table.SamuBootLevel = 0;
+	mm_boot_level_offset = smu_data->smu7_data.dpm_table_start +
+				offsetof(SMU75_Discrete_DpmTable, SamuBootLevel);
+
+	mm_boot_level_offset /= 4;
+	mm_boot_level_offset *= 4;
+	mm_boot_level_value = cgs_read_ind_register(hwmgr->device,
+			CGS_IND_REG__SMC, mm_boot_level_offset);
+	mm_boot_level_value &= 0xFFFFFF00;
+	mm_boot_level_value |= smu_data->smc_state_table.SamuBootLevel << 0;
+	cgs_write_ind_register(hwmgr->device,
+			CGS_IND_REG__SMC, mm_boot_level_offset, mm_boot_level_value);
+
+	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+			PHM_PlatformCaps_StablePState))
+		smum_send_msg_to_smc_with_parameter(hwmgr,
+				PPSMC_MSG_SAMUDPM_SetEnabledMask,
+				(uint32_t)(1 << smu_data->smc_state_table.SamuBootLevel));
+	return 0;
+}
+
+
+static int vegam_update_bif_smc_table(struct pp_hwmgr *hwmgr)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	struct phm_ppt_v1_pcie_table *pcie_table = table_info->pcie_table;
+	int max_entry, i;
+
+	max_entry = (SMU75_MAX_LEVELS_LINK < pcie_table->count) ?
+						SMU75_MAX_LEVELS_LINK :
+						pcie_table->count;
+	/* Setup BIF_SCLK levels */
+	for (i = 0; i < max_entry; i++)
+		smu_data->bif_sclk_table[i] = pcie_table->entries[i].pcie_sclk;
+	return 0;
+}
+
+static int vegam_update_smc_table(struct pp_hwmgr *hwmgr, uint32_t type)
+{
+	switch (type) {
+	case SMU_UVD_TABLE:
+		vegam_update_uvd_smc_table(hwmgr);
+		break;
+	case SMU_VCE_TABLE:
+		vegam_update_vce_smc_table(hwmgr);
+		break;
+	case SMU_SAMU_TABLE:
+		vegam_update_samu_smc_table(hwmgr);
+		break;
+	case SMU_BIF_TABLE:
+		vegam_update_bif_smc_table(hwmgr);
+		break;
+	default:
+		break;
+	}
+	return 0;
+}
+
+static void vegam_initialize_power_tune_defaults(struct pp_hwmgr *hwmgr)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	struct  phm_ppt_v1_information *table_info =
+			(struct  phm_ppt_v1_information *)(hwmgr->pptable);
+
+	if (table_info &&
+			table_info->cac_dtp_table->usPowerTuneDataSetID <= POWERTUNE_DEFAULT_SET_MAX &&
+			table_info->cac_dtp_table->usPowerTuneDataSetID)
+		smu_data->power_tune_defaults =
+				&vegam_power_tune_data_set_array
+				[table_info->cac_dtp_table->usPowerTuneDataSetID - 1];
+	else
+		smu_data->power_tune_defaults = &vegam_power_tune_data_set_array[0];
+
+}
+
+static int vegam_populate_smc_mvdd_table(struct pp_hwmgr *hwmgr,
+			SMU75_Discrete_DpmTable *table)
+{
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	uint32_t count, level;
+
+	if (SMU7_VOLTAGE_CONTROL_BY_GPIO == data->mvdd_control) {
+		count = data->mvdd_voltage_table.count;
+		if (count > SMU_MAX_SMIO_LEVELS)
+			count = SMU_MAX_SMIO_LEVELS;
+		for (level = 0; level < count; level++) {
+			table->SmioTable2.Pattern[level].Voltage = PP_HOST_TO_SMC_US(
+					data->mvdd_voltage_table.entries[count].value * VOLTAGE_SCALE);
+			/* Index into DpmTable.Smio. Drive bits from Smio entry to get this voltage level.*/
+			table->SmioTable2.Pattern[level].Smio =
+				(uint8_t) level;
+			table->Smio[level] |=
+				data->mvdd_voltage_table.entries[level].smio_low;
+		}
+		table->SmioMask2 = data->mvdd_voltage_table.mask_low;
+
+		table->MvddLevelCount = (uint32_t) PP_HOST_TO_SMC_UL(count);
+	}
+
+	return 0;
+}
+
+static int vegam_populate_smc_vddci_table(struct pp_hwmgr *hwmgr,
+					struct SMU75_Discrete_DpmTable *table)
+{
+	uint32_t count, level;
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+
+	count = data->vddci_voltage_table.count;
+
+	if (SMU7_VOLTAGE_CONTROL_BY_GPIO == data->vddci_control) {
+		if (count > SMU_MAX_SMIO_LEVELS)
+			count = SMU_MAX_SMIO_LEVELS;
+		for (level = 0; level < count; ++level) {
+			table->SmioTable1.Pattern[level].Voltage = PP_HOST_TO_SMC_US(
+					data->vddci_voltage_table.entries[level].value * VOLTAGE_SCALE);
+			table->SmioTable1.Pattern[level].Smio = (uint8_t) level;
+
+			table->Smio[level] |= data->vddci_voltage_table.entries[level].smio_low;
+		}
+	}
+
+	table->SmioMask1 = data->vddci_voltage_table.mask_low;
+
+	return 0;
+}
+
+static int vegam_populate_cac_table(struct pp_hwmgr *hwmgr,
+		struct SMU75_Discrete_DpmTable *table)
+{
+	uint32_t count;
+	uint8_t index;
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	struct phm_ppt_v1_voltage_lookup_table *lookup_table =
+			table_info->vddc_lookup_table;
+	/* tables is already swapped, so in order to use the value from it,
+	 * we need to swap it back.
+	 * We are populating vddc CAC data to BapmVddc table
+	 * in split and merged mode
+	 */
+	for (count = 0; count < lookup_table->count; count++) {
+		index = phm_get_voltage_index(lookup_table,
+				data->vddc_voltage_table.entries[count].value);
+		table->BapmVddcVidLoSidd[count] =
+				convert_to_vid(lookup_table->entries[index].us_cac_low);
+		table->BapmVddcVidHiSidd[count] =
+				convert_to_vid(lookup_table->entries[index].us_cac_mid);
+		table->BapmVddcVidHiSidd2[count] =
+				convert_to_vid(lookup_table->entries[index].us_cac_high);
+	}
+
+	return 0;
+}
+
+static int vegam_populate_smc_voltage_tables(struct pp_hwmgr *hwmgr,
+		struct SMU75_Discrete_DpmTable *table)
+{
+	vegam_populate_smc_vddci_table(hwmgr, table);
+	vegam_populate_smc_mvdd_table(hwmgr, table);
+	vegam_populate_cac_table(hwmgr, table);
+
+	return 0;
+}
+
+static int vegam_populate_ulv_level(struct pp_hwmgr *hwmgr,
+		struct SMU75_Discrete_Ulv *state)
+{
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+
+	state->CcPwrDynRm = 0;
+	state->CcPwrDynRm1 = 0;
+
+	state->VddcOffset = (uint16_t) table_info->us_ulv_voltage_offset;
+	state->VddcOffsetVid = (uint8_t)(table_info->us_ulv_voltage_offset *
+			VOLTAGE_VID_OFFSET_SCALE2 / VOLTAGE_VID_OFFSET_SCALE1);
+
+	state->VddcPhase = data->vddc_phase_shed_control ^ 0x3;
+
+	CONVERT_FROM_HOST_TO_SMC_UL(state->CcPwrDynRm);
+	CONVERT_FROM_HOST_TO_SMC_UL(state->CcPwrDynRm1);
+	CONVERT_FROM_HOST_TO_SMC_US(state->VddcOffset);
+
+	return 0;
+}
+
+static int vegam_populate_ulv_state(struct pp_hwmgr *hwmgr,
+		struct SMU75_Discrete_DpmTable *table)
+{
+	return vegam_populate_ulv_level(hwmgr, &table->Ulv);
+}
+
+static int vegam_populate_smc_link_level(struct pp_hwmgr *hwmgr,
+		struct SMU75_Discrete_DpmTable *table)
+{
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct vegam_smumgr *smu_data =
+			(struct vegam_smumgr *)(hwmgr->smu_backend);
+	struct smu7_dpm_table *dpm_table = &data->dpm_table;
+	int i;
+
+	/* Index (dpm_table->pcie_speed_table.count)
+	 * is reserved for PCIE boot level. */
+	for (i = 0; i <= dpm_table->pcie_speed_table.count; i++) {
+		table->LinkLevel[i].PcieGenSpeed  =
+				(uint8_t)dpm_table->pcie_speed_table.dpm_levels[i].value;
+		table->LinkLevel[i].PcieLaneCount = (uint8_t)encode_pcie_lane_width(
+				dpm_table->pcie_speed_table.dpm_levels[i].param1);
+		table->LinkLevel[i].EnabledForActivity = 1;
+		table->LinkLevel[i].SPC = (uint8_t)(data->pcie_spc_cap & 0xff);
+		table->LinkLevel[i].DownThreshold = PP_HOST_TO_SMC_UL(5);
+		table->LinkLevel[i].UpThreshold = PP_HOST_TO_SMC_UL(30);
+	}
+
+	smu_data->smc_state_table.LinkLevelCount =
+			(uint8_t)dpm_table->pcie_speed_table.count;
+
+/* To Do move to hwmgr */
+	data->dpm_level_enable_mask.pcie_dpm_enable_mask =
+			phm_get_dpm_level_enable_mask_value(&dpm_table->pcie_speed_table);
+
+	return 0;
+}
+
+static int vegam_get_dependency_volt_by_clk(struct pp_hwmgr *hwmgr,
+		struct phm_ppt_v1_clock_voltage_dependency_table *dep_table,
+		uint32_t clock, SMU_VoltageLevel *voltage, uint32_t *mvdd)
+{
+	uint32_t i;
+	uint16_t vddci;
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+
+	*voltage = *mvdd = 0;
+
+	/* clock - voltage dependency table is empty table */
+	if (dep_table->count == 0)
+		return -EINVAL;
+
+	for (i = 0; i < dep_table->count; i++) {
+		/* find first sclk bigger than request */
+		if (dep_table->entries[i].clk >= clock) {
+			*voltage |= (dep_table->entries[i].vddc *
+					VOLTAGE_SCALE) << VDDC_SHIFT;
+			if (SMU7_VOLTAGE_CONTROL_NONE == data->vddci_control)
+				*voltage |= (data->vbios_boot_state.vddci_bootup_value *
+						VOLTAGE_SCALE) << VDDCI_SHIFT;
+			else if (dep_table->entries[i].vddci)
+				*voltage |= (dep_table->entries[i].vddci *
+						VOLTAGE_SCALE) << VDDCI_SHIFT;
+			else {
+				vddci = phm_find_closest_vddci(&(data->vddci_voltage_table),
+						(dep_table->entries[i].vddc -
+								(uint16_t)VDDC_VDDCI_DELTA));
+				*voltage |= (vddci * VOLTAGE_SCALE) << VDDCI_SHIFT;
+			}
+
+			if (SMU7_VOLTAGE_CONTROL_NONE == data->mvdd_control)
+				*mvdd = data->vbios_boot_state.mvdd_bootup_value *
+					VOLTAGE_SCALE;
+			else if (dep_table->entries[i].mvdd)
+				*mvdd = (uint32_t) dep_table->entries[i].mvdd *
+					VOLTAGE_SCALE;
+
+			*voltage |= 1 << PHASES_SHIFT;
+			return 0;
+		}
+	}
+
+	/* sclk is bigger than max sclk in the dependence table */
+	*voltage |= (dep_table->entries[i - 1].vddc * VOLTAGE_SCALE) << VDDC_SHIFT;
+	vddci = phm_find_closest_vddci(&(data->vddci_voltage_table),
+			(dep_table->entries[i - 1].vddc -
+					(uint16_t)VDDC_VDDCI_DELTA));
+
+	if (SMU7_VOLTAGE_CONTROL_NONE == data->vddci_control)
+		*voltage |= (data->vbios_boot_state.vddci_bootup_value *
+				VOLTAGE_SCALE) << VDDCI_SHIFT;
+	else if (dep_table->entries[i - 1].vddci)
+		*voltage |= (dep_table->entries[i - 1].vddci *
+				VOLTAGE_SCALE) << VDDC_SHIFT;
+	else
+		*voltage |= (vddci * VOLTAGE_SCALE) << VDDCI_SHIFT;
+
+	if (SMU7_VOLTAGE_CONTROL_NONE == data->mvdd_control)
+		*mvdd = data->vbios_boot_state.mvdd_bootup_value * VOLTAGE_SCALE;
+	else if (dep_table->entries[i].mvdd)
+		*mvdd = (uint32_t) dep_table->entries[i - 1].mvdd * VOLTAGE_SCALE;
+
+	return 0;
+}
+
+static void vegam_get_sclk_range_table(struct pp_hwmgr *hwmgr,
+				   SMU75_Discrete_DpmTable  *table)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	uint32_t i, ref_clk;
+
+	struct pp_atom_ctrl_sclk_range_table range_table_from_vbios = { { {0} } };
+
+	ref_clk = amdgpu_asic_get_xclk((struct amdgpu_device *)hwmgr->adev);
+
+	if (0 == atomctrl_get_smc_sclk_range_table(hwmgr, &range_table_from_vbios)) {
+		for (i = 0; i < NUM_SCLK_RANGE; i++) {
+			table->SclkFcwRangeTable[i].vco_setting =
+					range_table_from_vbios.entry[i].ucVco_setting;
+			table->SclkFcwRangeTable[i].postdiv =
+					range_table_from_vbios.entry[i].ucPostdiv;
+			table->SclkFcwRangeTable[i].fcw_pcc =
+					range_table_from_vbios.entry[i].usFcw_pcc;
+
+			table->SclkFcwRangeTable[i].fcw_trans_upper =
+					range_table_from_vbios.entry[i].usFcw_trans_upper;
+			table->SclkFcwRangeTable[i].fcw_trans_lower =
+					range_table_from_vbios.entry[i].usRcw_trans_lower;
+
+			CONVERT_FROM_HOST_TO_SMC_US(table->SclkFcwRangeTable[i].fcw_pcc);
+			CONVERT_FROM_HOST_TO_SMC_US(table->SclkFcwRangeTable[i].fcw_trans_upper);
+			CONVERT_FROM_HOST_TO_SMC_US(table->SclkFcwRangeTable[i].fcw_trans_lower);
+		}
+		return;
+	}
+
+	for (i = 0; i < NUM_SCLK_RANGE; i++) {
+		smu_data->range_table[i].trans_lower_frequency =
+				(ref_clk * Range_Table[i].fcw_trans_lower) >> Range_Table[i].postdiv;
+		smu_data->range_table[i].trans_upper_frequency =
+				(ref_clk * Range_Table[i].fcw_trans_upper) >> Range_Table[i].postdiv;
+
+		table->SclkFcwRangeTable[i].vco_setting = Range_Table[i].vco_setting;
+		table->SclkFcwRangeTable[i].postdiv = Range_Table[i].postdiv;
+		table->SclkFcwRangeTable[i].fcw_pcc = Range_Table[i].fcw_pcc;
+
+		table->SclkFcwRangeTable[i].fcw_trans_upper = Range_Table[i].fcw_trans_upper;
+		table->SclkFcwRangeTable[i].fcw_trans_lower = Range_Table[i].fcw_trans_lower;
+
+		CONVERT_FROM_HOST_TO_SMC_US(table->SclkFcwRangeTable[i].fcw_pcc);
+		CONVERT_FROM_HOST_TO_SMC_US(table->SclkFcwRangeTable[i].fcw_trans_upper);
+		CONVERT_FROM_HOST_TO_SMC_US(table->SclkFcwRangeTable[i].fcw_trans_lower);
+	}
+}
+
+static int vegam_calculate_sclk_params(struct pp_hwmgr *hwmgr,
+		uint32_t clock, SMU_SclkSetting *sclk_setting)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	const SMU75_Discrete_DpmTable *table = &(smu_data->smc_state_table);
+	struct pp_atomctrl_clock_dividers_ai dividers;
+	uint32_t ref_clock;
+	uint32_t pcc_target_percent, pcc_target_freq, ss_target_percent, ss_target_freq;
+	uint8_t i;
+	int result;
+	uint64_t temp;
+
+	sclk_setting->SclkFrequency = clock;
+	/* get the engine clock dividers for this clock value */
+	result = atomctrl_get_engine_pll_dividers_ai(hwmgr, clock,  &dividers);
+	if (result == 0) {
+		sclk_setting->Fcw_int = dividers.usSclk_fcw_int;
+		sclk_setting->Fcw_frac = dividers.usSclk_fcw_frac;
+		sclk_setting->Pcc_fcw_int = dividers.usPcc_fcw_int;
+		sclk_setting->PllRange = dividers.ucSclkPllRange;
+		sclk_setting->Sclk_slew_rate = 0x400;
+		sclk_setting->Pcc_up_slew_rate = dividers.usPcc_fcw_slew_frac;
+		sclk_setting->Pcc_down_slew_rate = 0xffff;
+		sclk_setting->SSc_En = dividers.ucSscEnable;
+		sclk_setting->Fcw1_int = dividers.usSsc_fcw1_int;
+		sclk_setting->Fcw1_frac = dividers.usSsc_fcw1_frac;
+		sclk_setting->Sclk_ss_slew_rate = dividers.usSsc_fcw_slew_frac;
+		return result;
+	}
+
+	ref_clock = amdgpu_asic_get_xclk((struct amdgpu_device *)hwmgr->adev);
+
+	for (i = 0; i < NUM_SCLK_RANGE; i++) {
+		if (clock > smu_data->range_table[i].trans_lower_frequency
+		&& clock <= smu_data->range_table[i].trans_upper_frequency) {
+			sclk_setting->PllRange = i;
+			break;
+		}
+	}
+
+	sclk_setting->Fcw_int = (uint16_t)
+			((clock << table->SclkFcwRangeTable[sclk_setting->PllRange].postdiv) /
+					ref_clock);
+	temp = clock << table->SclkFcwRangeTable[sclk_setting->PllRange].postdiv;
+	temp <<= 0x10;
+	do_div(temp, ref_clock);
+	sclk_setting->Fcw_frac = temp & 0xffff;
+
+	pcc_target_percent = 10; /*  Hardcode 10% for now. */
+	pcc_target_freq = clock - (clock * pcc_target_percent / 100);
+	sclk_setting->Pcc_fcw_int = (uint16_t)
+			((pcc_target_freq << table->SclkFcwRangeTable[sclk_setting->PllRange].postdiv) /
+					ref_clock);
+
+	ss_target_percent = 2; /*  Hardcode 2% for now. */
+	sclk_setting->SSc_En = 0;
+	if (ss_target_percent) {
+		sclk_setting->SSc_En = 1;
+		ss_target_freq = clock - (clock * ss_target_percent / 100);
+		sclk_setting->Fcw1_int = (uint16_t)
+				((ss_target_freq << table->SclkFcwRangeTable[sclk_setting->PllRange].postdiv) /
+						ref_clock);
+		temp = ss_target_freq << table->SclkFcwRangeTable[sclk_setting->PllRange].postdiv;
+		temp <<= 0x10;
+		do_div(temp, ref_clock);
+		sclk_setting->Fcw1_frac = temp & 0xffff;
+	}
+
+	return 0;
+}
+
+static uint8_t vegam_get_sleep_divider_id_from_clock(uint32_t clock,
+		uint32_t clock_insr)
+{
+	uint8_t i;
+	uint32_t temp;
+	uint32_t min = max(clock_insr, (uint32_t)SMU7_MINIMUM_ENGINE_CLOCK);
+
+	PP_ASSERT_WITH_CODE((clock >= min),
+			"Engine clock can't satisfy stutter requirement!",
+			return 0);
+	for (i = 31;  ; i--) {
+		temp = clock / (i + 1);
+
+		if (temp >= min || i == 0)
+			break;
+	}
+	return i;
+}
+
+static int vegam_populate_single_graphic_level(struct pp_hwmgr *hwmgr,
+		uint32_t clock, struct SMU75_Discrete_GraphicsLevel *level)
+{
+	int result;
+	/* PP_Clocks minClocks; */
+	uint32_t mvdd;
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	SMU_SclkSetting curr_sclk_setting = { 0 };
+
+	result = vegam_calculate_sclk_params(hwmgr, clock, &curr_sclk_setting);
+
+	/* populate graphics levels */
+	result = vegam_get_dependency_volt_by_clk(hwmgr,
+			table_info->vdd_dep_on_sclk, clock,
+			&level->MinVoltage, &mvdd);
+
+	PP_ASSERT_WITH_CODE((0 == result),
+			"can not find VDDC voltage value for "
+			"VDDC engine clock dependency table",
+			return result);
+	level->ActivityLevel = (uint16_t)(SclkDPMTuning_VEGAM >> DPMTuning_Activity_Shift);
+
+	level->CcPwrDynRm = 0;
+	level->CcPwrDynRm1 = 0;
+	level->EnabledForActivity = 0;
+	level->EnabledForThrottle = 1;
+	level->VoltageDownHyst = 0;
+	level->PowerThrottle = 0;
+	data->display_timing.min_clock_in_sr = hwmgr->display_config->min_core_set_clock_in_sr;
+
+	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_SclkDeepSleep))
+		level->DeepSleepDivId = vegam_get_sleep_divider_id_from_clock(clock,
+								hwmgr->display_config->min_core_set_clock_in_sr);
+
+	level->SclkSetting = curr_sclk_setting;
+
+	CONVERT_FROM_HOST_TO_SMC_UL(level->MinVoltage);
+	CONVERT_FROM_HOST_TO_SMC_UL(level->CcPwrDynRm);
+	CONVERT_FROM_HOST_TO_SMC_UL(level->CcPwrDynRm1);
+	CONVERT_FROM_HOST_TO_SMC_US(level->ActivityLevel);
+	CONVERT_FROM_HOST_TO_SMC_UL(level->SclkSetting.SclkFrequency);
+	CONVERT_FROM_HOST_TO_SMC_US(level->SclkSetting.Fcw_int);
+	CONVERT_FROM_HOST_TO_SMC_US(level->SclkSetting.Fcw_frac);
+	CONVERT_FROM_HOST_TO_SMC_US(level->SclkSetting.Pcc_fcw_int);
+	CONVERT_FROM_HOST_TO_SMC_US(level->SclkSetting.Sclk_slew_rate);
+	CONVERT_FROM_HOST_TO_SMC_US(level->SclkSetting.Pcc_up_slew_rate);
+	CONVERT_FROM_HOST_TO_SMC_US(level->SclkSetting.Pcc_down_slew_rate);
+	CONVERT_FROM_HOST_TO_SMC_US(level->SclkSetting.Fcw1_int);
+	CONVERT_FROM_HOST_TO_SMC_US(level->SclkSetting.Fcw1_frac);
+	CONVERT_FROM_HOST_TO_SMC_US(level->SclkSetting.Sclk_ss_slew_rate);
+	return 0;
+}
+
+static int vegam_populate_all_graphic_levels(struct pp_hwmgr *hwmgr)
+{
+	struct smu7_hwmgr *hw_data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	struct smu7_dpm_table *dpm_table = &hw_data->dpm_table;
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	struct phm_ppt_v1_pcie_table *pcie_table = table_info->pcie_table;
+	uint8_t pcie_entry_cnt = (uint8_t) hw_data->dpm_table.pcie_speed_table.count;
+	int result = 0;
+	uint32_t array = smu_data->smu7_data.dpm_table_start +
+			offsetof(SMU75_Discrete_DpmTable, GraphicsLevel);
+	uint32_t array_size = sizeof(struct SMU75_Discrete_GraphicsLevel) *
+			SMU75_MAX_LEVELS_GRAPHICS;
+	struct SMU75_Discrete_GraphicsLevel *levels =
+			smu_data->smc_state_table.GraphicsLevel;
+	uint32_t i, max_entry;
+	uint8_t hightest_pcie_level_enabled = 0,
+		lowest_pcie_level_enabled = 0,
+		mid_pcie_level_enabled = 0,
+		count = 0;
+
+	vegam_get_sclk_range_table(hwmgr, &(smu_data->smc_state_table));
+
+	for (i = 0; i < dpm_table->sclk_table.count; i++) {
+
+		result = vegam_populate_single_graphic_level(hwmgr,
+				dpm_table->sclk_table.dpm_levels[i].value,
+				&(smu_data->smc_state_table.GraphicsLevel[i]));
+		if (result)
+			return result;
+
+		levels[i].UpHyst = (uint8_t)
+				(SclkDPMTuning_VEGAM >> DPMTuning_Uphyst_Shift);
+		levels[i].DownHyst = (uint8_t)
+				(SclkDPMTuning_VEGAM >> DPMTuning_Downhyst_Shift);
+		/* Making sure only DPM level 0-1 have Deep Sleep Div ID populated. */
+		if (i > 1)
+			levels[i].DeepSleepDivId = 0;
+	}
+	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+					PHM_PlatformCaps_SPLLShutdownSupport))
+		smu_data->smc_state_table.GraphicsLevel[0].SclkSetting.SSc_En = 0;
+
+	smu_data->smc_state_table.GraphicsDpmLevelCount =
+			(uint8_t)dpm_table->sclk_table.count;
+	hw_data->dpm_level_enable_mask.sclk_dpm_enable_mask =
+			phm_get_dpm_level_enable_mask_value(&dpm_table->sclk_table);
+
+	for (i = 0; i < dpm_table->sclk_table.count; i++)
+		levels[i].EnabledForActivity =
+				(hw_data->dpm_level_enable_mask.sclk_dpm_enable_mask >> i) & 0x1;
+
+	if (pcie_table != NULL) {
+		PP_ASSERT_WITH_CODE((1 <= pcie_entry_cnt),
+				"There must be 1 or more PCIE levels defined in PPTable.",
+				return -EINVAL);
+		max_entry = pcie_entry_cnt - 1;
+		for (i = 0; i < dpm_table->sclk_table.count; i++)
+			levels[i].pcieDpmLevel =
+					(uint8_t) ((i < max_entry) ? i : max_entry);
+	} else {
+		while (hw_data->dpm_level_enable_mask.pcie_dpm_enable_mask &&
+				((hw_data->dpm_level_enable_mask.pcie_dpm_enable_mask &
+						(1 << (hightest_pcie_level_enabled + 1))) != 0))
+			hightest_pcie_level_enabled++;
+
+		while (hw_data->dpm_level_enable_mask.pcie_dpm_enable_mask &&
+				((hw_data->dpm_level_enable_mask.pcie_dpm_enable_mask &
+						(1 << lowest_pcie_level_enabled)) == 0))
+			lowest_pcie_level_enabled++;
+
+		while ((count < hightest_pcie_level_enabled) &&
+				((hw_data->dpm_level_enable_mask.pcie_dpm_enable_mask &
+						(1 << (lowest_pcie_level_enabled + 1 + count))) == 0))
+			count++;
+
+		mid_pcie_level_enabled = (lowest_pcie_level_enabled + 1 + count) <
+				hightest_pcie_level_enabled ?
+						(lowest_pcie_level_enabled + 1 + count) :
+						hightest_pcie_level_enabled;
+
+		/* set pcieDpmLevel to hightest_pcie_level_enabled */
+		for (i = 2; i < dpm_table->sclk_table.count; i++)
+			levels[i].pcieDpmLevel = hightest_pcie_level_enabled;
+
+		/* set pcieDpmLevel to lowest_pcie_level_enabled */
+		levels[0].pcieDpmLevel = lowest_pcie_level_enabled;
+
+		/* set pcieDpmLevel to mid_pcie_level_enabled */
+		levels[1].pcieDpmLevel = mid_pcie_level_enabled;
+	}
+	/* level count will send to smc once at init smc table and never change */
+	result = smu7_copy_bytes_to_smc(hwmgr, array, (uint8_t *)levels,
+			(uint32_t)array_size, SMC_RAM_END);
+
+	return result;
+}
+
+static int vegam_calculate_mclk_params(struct pp_hwmgr *hwmgr,
+		uint32_t clock, struct SMU75_Discrete_MemoryLevel *mem_level)
+{
+	struct pp_atomctrl_memory_clock_param_ai mpll_param;
+
+	PP_ASSERT_WITH_CODE(!atomctrl_get_memory_pll_dividers_ai(hwmgr,
+			clock, &mpll_param),
+			"Failed to retrieve memory pll parameter.",
+			return -EINVAL);
+
+	mem_level->MclkFrequency = (uint32_t)mpll_param.ulClock;
+	mem_level->Fcw_int = (uint16_t)mpll_param.ulMclk_fcw_int;
+	mem_level->Fcw_frac = (uint16_t)mpll_param.ulMclk_fcw_frac;
+	mem_level->Postdiv = (uint8_t)mpll_param.ulPostDiv;
+
+	return 0;
+}
+
+static int vegam_populate_single_memory_level(struct pp_hwmgr *hwmgr,
+		uint32_t clock, struct SMU75_Discrete_MemoryLevel *mem_level)
+{
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	int result = 0;
+	uint32_t mclk_stutter_mode_threshold = 60000;
+
+
+	if (table_info->vdd_dep_on_mclk) {
+		result = vegam_get_dependency_volt_by_clk(hwmgr,
+				table_info->vdd_dep_on_mclk, clock,
+				&mem_level->MinVoltage, &mem_level->MinMvdd);
+		PP_ASSERT_WITH_CODE(!result,
+				"can not find MinVddc voltage value from memory "
+				"VDDC voltage dependency table", return result);
+	}
+
+	result = vegam_calculate_mclk_params(hwmgr, clock, mem_level);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to calculate mclk params.",
+			return -EINVAL);
+
+	mem_level->EnabledForThrottle = 1;
+	mem_level->EnabledForActivity = 0;
+	mem_level->VoltageDownHyst = 0;
+	mem_level->ActivityLevel = (uint16_t)
+			(MemoryDPMTuning_VEGAM >> DPMTuning_Activity_Shift);
+	mem_level->StutterEnable = false;
+	mem_level->DisplayWatermark = PPSMC_DISPLAY_WATERMARK_LOW;
+
+	data->display_timing.num_existing_displays = hwmgr->display_config->num_display;
+
+	if (mclk_stutter_mode_threshold &&
+		(clock <= mclk_stutter_mode_threshold) &&
+		(PHM_READ_FIELD(hwmgr->device, DPG_PIPE_STUTTER_CONTROL,
+				STUTTER_ENABLE) & 0x1))
+		mem_level->StutterEnable = true;
+
+	if (!result) {
+		CONVERT_FROM_HOST_TO_SMC_UL(mem_level->MinMvdd);
+		CONVERT_FROM_HOST_TO_SMC_UL(mem_level->MclkFrequency);
+		CONVERT_FROM_HOST_TO_SMC_US(mem_level->Fcw_int);
+		CONVERT_FROM_HOST_TO_SMC_US(mem_level->Fcw_frac);
+		CONVERT_FROM_HOST_TO_SMC_US(mem_level->ActivityLevel);
+		CONVERT_FROM_HOST_TO_SMC_UL(mem_level->MinVoltage);
+	}
+
+	return result;
+}
+
+static int vegam_populate_all_memory_levels(struct pp_hwmgr *hwmgr)
+{
+	struct smu7_hwmgr *hw_data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	struct smu7_dpm_table *dpm_table = &hw_data->dpm_table;
+	int result;
+	/* populate MCLK dpm table to SMU7 */
+	uint32_t array = smu_data->smu7_data.dpm_table_start +
+			offsetof(SMU75_Discrete_DpmTable, MemoryLevel);
+	uint32_t array_size = sizeof(SMU75_Discrete_MemoryLevel) *
+			SMU75_MAX_LEVELS_MEMORY;
+	struct SMU75_Discrete_MemoryLevel *levels =
+			smu_data->smc_state_table.MemoryLevel;
+	uint32_t i;
+
+	for (i = 0; i < dpm_table->mclk_table.count; i++) {
+		PP_ASSERT_WITH_CODE((0 != dpm_table->mclk_table.dpm_levels[i].value),
+				"can not populate memory level as memory clock is zero",
+				return -EINVAL);
+		result = vegam_populate_single_memory_level(hwmgr,
+				dpm_table->mclk_table.dpm_levels[i].value,
+				&levels[i]);
+
+		if (result)
+			return result;
+
+		levels[i].UpHyst = (uint8_t)
+				(MemoryDPMTuning_VEGAM >> DPMTuning_Uphyst_Shift);
+		levels[i].DownHyst = (uint8_t)
+				(MemoryDPMTuning_VEGAM >> DPMTuning_Downhyst_Shift);
+	}
+
+	smu_data->smc_state_table.MemoryDpmLevelCount =
+			(uint8_t)dpm_table->mclk_table.count;
+	hw_data->dpm_level_enable_mask.mclk_dpm_enable_mask =
+			phm_get_dpm_level_enable_mask_value(&dpm_table->mclk_table);
+
+	for (i = 0; i < dpm_table->mclk_table.count; i++)
+		levels[i].EnabledForActivity =
+				(hw_data->dpm_level_enable_mask.mclk_dpm_enable_mask >> i) & 0x1;
+
+	levels[dpm_table->mclk_table.count - 1].DisplayWatermark =
+			PPSMC_DISPLAY_WATERMARK_HIGH;
+
+	/* level count will send to smc once at init smc table and never change */
+	result = smu7_copy_bytes_to_smc(hwmgr, array, (uint8_t *)levels,
+			(uint32_t)array_size, SMC_RAM_END);
+
+	return result;
+}
+
+static int vegam_populate_mvdd_value(struct pp_hwmgr *hwmgr,
+		uint32_t mclk, SMIO_Pattern *smio_pat)
+{
+	const struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	uint32_t i = 0;
+
+	if (SMU7_VOLTAGE_CONTROL_NONE != data->mvdd_control) {
+		/* find mvdd value which clock is more than request */
+		for (i = 0; i < table_info->vdd_dep_on_mclk->count; i++) {
+			if (mclk <= table_info->vdd_dep_on_mclk->entries[i].clk) {
+				smio_pat->Voltage = data->mvdd_voltage_table.entries[i].value;
+				break;
+			}
+		}
+		PP_ASSERT_WITH_CODE(i < table_info->vdd_dep_on_mclk->count,
+				"MVDD Voltage is outside the supported range.",
+				return -EINVAL);
+	} else
+		return -EINVAL;
+
+	return 0;
+}
+
+static int vegam_populate_smc_acpi_level(struct pp_hwmgr *hwmgr,
+		SMU75_Discrete_DpmTable *table)
+{
+	int result = 0;
+	uint32_t sclk_frequency;
+	const struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	SMIO_Pattern vol_level;
+	uint32_t mvdd;
+	uint16_t us_mvdd;
+
+	table->ACPILevel.Flags &= ~PPSMC_SWSTATE_FLAG_DC;
+
+	/* Get MinVoltage and Frequency from DPM0,
+	 * already converted to SMC_UL */
+	sclk_frequency = data->vbios_boot_state.sclk_bootup_value;
+	result = vegam_get_dependency_volt_by_clk(hwmgr,
+			table_info->vdd_dep_on_sclk,
+			sclk_frequency,
+			&table->ACPILevel.MinVoltage, &mvdd);
+	PP_ASSERT_WITH_CODE(!result,
+			"Cannot find ACPI VDDC voltage value "
+			"in Clock Dependency Table",
+			);
+
+	result = vegam_calculate_sclk_params(hwmgr, sclk_frequency,
+			&(table->ACPILevel.SclkSetting));
+	PP_ASSERT_WITH_CODE(!result,
+			"Error retrieving Engine Clock dividers from VBIOS.",
+			return result);
+
+	table->ACPILevel.DeepSleepDivId = 0;
+	table->ACPILevel.CcPwrDynRm = 0;
+	table->ACPILevel.CcPwrDynRm1 = 0;
+
+	CONVERT_FROM_HOST_TO_SMC_UL(table->ACPILevel.Flags);
+	CONVERT_FROM_HOST_TO_SMC_UL(table->ACPILevel.MinVoltage);
+	CONVERT_FROM_HOST_TO_SMC_UL(table->ACPILevel.CcPwrDynRm);
+	CONVERT_FROM_HOST_TO_SMC_UL(table->ACPILevel.CcPwrDynRm1);
+
+	CONVERT_FROM_HOST_TO_SMC_UL(table->ACPILevel.SclkSetting.SclkFrequency);
+	CONVERT_FROM_HOST_TO_SMC_US(table->ACPILevel.SclkSetting.Fcw_int);
+	CONVERT_FROM_HOST_TO_SMC_US(table->ACPILevel.SclkSetting.Fcw_frac);
+	CONVERT_FROM_HOST_TO_SMC_US(table->ACPILevel.SclkSetting.Pcc_fcw_int);
+	CONVERT_FROM_HOST_TO_SMC_US(table->ACPILevel.SclkSetting.Sclk_slew_rate);
+	CONVERT_FROM_HOST_TO_SMC_US(table->ACPILevel.SclkSetting.Pcc_up_slew_rate);
+	CONVERT_FROM_HOST_TO_SMC_US(table->ACPILevel.SclkSetting.Pcc_down_slew_rate);
+	CONVERT_FROM_HOST_TO_SMC_US(table->ACPILevel.SclkSetting.Fcw1_int);
+	CONVERT_FROM_HOST_TO_SMC_US(table->ACPILevel.SclkSetting.Fcw1_frac);
+	CONVERT_FROM_HOST_TO_SMC_US(table->ACPILevel.SclkSetting.Sclk_ss_slew_rate);
+
+
+	/* Get MinVoltage and Frequency from DPM0, already converted to SMC_UL */
+	table->MemoryACPILevel.MclkFrequency = data->vbios_boot_state.mclk_bootup_value;
+	result = vegam_get_dependency_volt_by_clk(hwmgr,
+			table_info->vdd_dep_on_mclk,
+			table->MemoryACPILevel.MclkFrequency,
+			&table->MemoryACPILevel.MinVoltage, &mvdd);
+	PP_ASSERT_WITH_CODE((0 == result),
+			"Cannot find ACPI VDDCI voltage value "
+			"in Clock Dependency Table",
+			);
+
+	us_mvdd = 0;
+	if ((SMU7_VOLTAGE_CONTROL_NONE == data->mvdd_control) ||
+			(data->mclk_dpm_key_disabled))
+		us_mvdd = data->vbios_boot_state.mvdd_bootup_value;
+	else {
+		if (!vegam_populate_mvdd_value(hwmgr,
+				data->dpm_table.mclk_table.dpm_levels[0].value,
+				&vol_level))
+			us_mvdd = vol_level.Voltage;
+	}
+
+	if (!vegam_populate_mvdd_value(hwmgr, 0, &vol_level))
+		table->MemoryACPILevel.MinMvdd = PP_HOST_TO_SMC_UL(vol_level.Voltage);
+	else
+		table->MemoryACPILevel.MinMvdd = 0;
+
+	table->MemoryACPILevel.StutterEnable = false;
+
+	table->MemoryACPILevel.EnabledForThrottle = 0;
+	table->MemoryACPILevel.EnabledForActivity = 0;
+	table->MemoryACPILevel.UpHyst = 0;
+	table->MemoryACPILevel.DownHyst = 100;
+	table->MemoryACPILevel.VoltageDownHyst = 0;
+	table->MemoryACPILevel.ActivityLevel =
+		PP_HOST_TO_SMC_US(data->current_profile_setting.mclk_activity);
+
+	CONVERT_FROM_HOST_TO_SMC_UL(table->MemoryACPILevel.MclkFrequency);
+	CONVERT_FROM_HOST_TO_SMC_UL(table->MemoryACPILevel.MinVoltage);
+
+	return result;
+}
+
+static int vegam_populate_smc_vce_level(struct pp_hwmgr *hwmgr,
+		SMU75_Discrete_DpmTable *table)
+{
+	int result = -EINVAL;
+	uint8_t count;
+	struct pp_atomctrl_clock_dividers_vi dividers;
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	struct phm_ppt_v1_mm_clock_voltage_dependency_table *mm_table =
+			table_info->mm_dep_table;
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	uint32_t vddci;
+
+	table->VceLevelCount = (uint8_t)(mm_table->count);
+	table->VceBootLevel = 0;
+
+	for (count = 0; count < table->VceLevelCount; count++) {
+		table->VceLevel[count].Frequency = mm_table->entries[count].eclk;
+		table->VceLevel[count].MinVoltage = 0;
+		table->VceLevel[count].MinVoltage |=
+				(mm_table->entries[count].vddc * VOLTAGE_SCALE) << VDDC_SHIFT;
+
+		if (SMU7_VOLTAGE_CONTROL_BY_GPIO == data->vddci_control)
+			vddci = (uint32_t)phm_find_closest_vddci(&(data->vddci_voltage_table),
+						mm_table->entries[count].vddc - VDDC_VDDCI_DELTA);
+		else if (SMU7_VOLTAGE_CONTROL_BY_SVID2 == data->vddci_control)
+			vddci = mm_table->entries[count].vddc - VDDC_VDDCI_DELTA;
+		else
+			vddci = (data->vbios_boot_state.vddci_bootup_value * VOLTAGE_SCALE) << VDDCI_SHIFT;
+
+
+		table->VceLevel[count].MinVoltage |=
+				(vddci * VOLTAGE_SCALE) << VDDCI_SHIFT;
+		table->VceLevel[count].MinVoltage |= 1 << PHASES_SHIFT;
+
+		/*retrieve divider value for VBIOS */
+		result = atomctrl_get_dfs_pll_dividers_vi(hwmgr,
+				table->VceLevel[count].Frequency, &dividers);
+		PP_ASSERT_WITH_CODE((0 == result),
+				"can not find divide id for VCE engine clock",
+				return result);
+
+		table->VceLevel[count].Divider = (uint8_t)dividers.pll_post_divider;
+
+		CONVERT_FROM_HOST_TO_SMC_UL(table->VceLevel[count].Frequency);
+		CONVERT_FROM_HOST_TO_SMC_UL(table->VceLevel[count].MinVoltage);
+	}
+	return result;
+}
+
+static int vegam_populate_smc_samu_level(struct pp_hwmgr *hwmgr,
+		SMU75_Discrete_DpmTable *table)
+{
+	int result = -EINVAL;
+	uint8_t count;
+	struct pp_atomctrl_clock_dividers_vi dividers;
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	struct phm_ppt_v1_mm_clock_voltage_dependency_table *mm_table =
+			table_info->mm_dep_table;
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	uint32_t vddci;
+
+	table->SamuBootLevel = 0;
+	table->SamuLevelCount = (uint8_t)(mm_table->count);
+
+	for (count = 0; count < table->SamuLevelCount; count++) {
+		/* not sure whether we need evclk or not */
+		table->SamuLevel[count].MinVoltage = 0;
+		table->SamuLevel[count].Frequency = mm_table->entries[count].samclock;
+		table->SamuLevel[count].MinVoltage |= (mm_table->entries[count].vddc *
+				VOLTAGE_SCALE) << VDDC_SHIFT;
+
+		if (SMU7_VOLTAGE_CONTROL_BY_GPIO == data->vddci_control)
+			vddci = (uint32_t)phm_find_closest_vddci(&(data->vddci_voltage_table),
+						mm_table->entries[count].vddc - VDDC_VDDCI_DELTA);
+		else if (SMU7_VOLTAGE_CONTROL_BY_SVID2 == data->vddci_control)
+			vddci = mm_table->entries[count].vddc - VDDC_VDDCI_DELTA;
+		else
+			vddci = (data->vbios_boot_state.vddci_bootup_value * VOLTAGE_SCALE) << VDDCI_SHIFT;
+
+		table->SamuLevel[count].MinVoltage |= (vddci * VOLTAGE_SCALE) << VDDCI_SHIFT;
+		table->SamuLevel[count].MinVoltage |= 1 << PHASES_SHIFT;
+
+		/* retrieve divider value for VBIOS */
+		result = atomctrl_get_dfs_pll_dividers_vi(hwmgr,
+				table->SamuLevel[count].Frequency, &dividers);
+		PP_ASSERT_WITH_CODE((0 == result),
+				"can not find divide id for samu clock", return result);
+
+		table->SamuLevel[count].Divider = (uint8_t)dividers.pll_post_divider;
+
+		CONVERT_FROM_HOST_TO_SMC_UL(table->SamuLevel[count].Frequency);
+		CONVERT_FROM_HOST_TO_SMC_UL(table->SamuLevel[count].MinVoltage);
+	}
+	return result;
+}
+
+static int vegam_populate_memory_timing_parameters(struct pp_hwmgr *hwmgr,
+		int32_t eng_clock, int32_t mem_clock,
+		SMU75_Discrete_MCArbDramTimingTableEntry *arb_regs)
+{
+	uint32_t dram_timing;
+	uint32_t dram_timing2;
+	uint32_t burst_time;
+	uint32_t rfsh_rate;
+	uint32_t misc3;
+
+	int result;
+
+	result = atomctrl_set_engine_dram_timings_rv770(hwmgr,
+			eng_clock, mem_clock);
+	PP_ASSERT_WITH_CODE(result == 0,
+			"Error calling VBIOS to set DRAM_TIMING.",
+			return result);
+
+	dram_timing = cgs_read_register(hwmgr->device, mmMC_ARB_DRAM_TIMING);
+	dram_timing2 = cgs_read_register(hwmgr->device, mmMC_ARB_DRAM_TIMING2);
+	burst_time = cgs_read_register(hwmgr->device, mmMC_ARB_BURST_TIME);
+	rfsh_rate = cgs_read_register(hwmgr->device, mmMC_ARB_RFSH_RATE);
+	misc3 = cgs_read_register(hwmgr->device, mmMC_ARB_MISC3);
+
+	arb_regs->McArbDramTiming  = PP_HOST_TO_SMC_UL(dram_timing);
+	arb_regs->McArbDramTiming2 = PP_HOST_TO_SMC_UL(dram_timing2);
+	arb_regs->McArbBurstTime   = PP_HOST_TO_SMC_UL(burst_time);
+	arb_regs->McArbRfshRate = PP_HOST_TO_SMC_UL(rfsh_rate);
+	arb_regs->McArbMisc3 = PP_HOST_TO_SMC_UL(misc3);
+
+	return 0;
+}
+
+static int vegam_program_memory_timing_parameters(struct pp_hwmgr *hwmgr)
+{
+	struct smu7_hwmgr *hw_data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	struct SMU75_Discrete_MCArbDramTimingTable arb_regs = {0};
+	uint32_t i, j;
+	int result = 0;
+
+	for (i = 0; i < hw_data->dpm_table.sclk_table.count; i++) {
+		for (j = 0; j < hw_data->dpm_table.mclk_table.count; j++) {
+			result = vegam_populate_memory_timing_parameters(hwmgr,
+					hw_data->dpm_table.sclk_table.dpm_levels[i].value,
+					hw_data->dpm_table.mclk_table.dpm_levels[j].value,
+					&arb_regs.entries[i][j]);
+			if (result)
+				return result;
+		}
+	}
+
+	result = smu7_copy_bytes_to_smc(
+			hwmgr,
+			smu_data->smu7_data.arb_table_start,
+			(uint8_t *)&arb_regs,
+			sizeof(SMU75_Discrete_MCArbDramTimingTable),
+			SMC_RAM_END);
+	return result;
+}
+
+static int vegam_populate_smc_uvd_level(struct pp_hwmgr *hwmgr,
+		struct SMU75_Discrete_DpmTable *table)
+{
+	int result = -EINVAL;
+	uint8_t count;
+	struct pp_atomctrl_clock_dividers_vi dividers;
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	struct phm_ppt_v1_mm_clock_voltage_dependency_table *mm_table =
+			table_info->mm_dep_table;
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	uint32_t vddci;
+
+	table->UvdLevelCount = (uint8_t)(mm_table->count);
+	table->UvdBootLevel = 0;
+
+	for (count = 0; count < table->UvdLevelCount; count++) {
+		table->UvdLevel[count].MinVoltage = 0;
+		table->UvdLevel[count].VclkFrequency = mm_table->entries[count].vclk;
+		table->UvdLevel[count].DclkFrequency = mm_table->entries[count].dclk;
+		table->UvdLevel[count].MinVoltage |=
+				(mm_table->entries[count].vddc * VOLTAGE_SCALE) << VDDC_SHIFT;
+
+		if (SMU7_VOLTAGE_CONTROL_BY_GPIO == data->vddci_control)
+			vddci = (uint32_t)phm_find_closest_vddci(&(data->vddci_voltage_table),
+						mm_table->entries[count].vddc - VDDC_VDDCI_DELTA);
+		else if (SMU7_VOLTAGE_CONTROL_BY_SVID2 == data->vddci_control)
+			vddci = mm_table->entries[count].vddc - VDDC_VDDCI_DELTA;
+		else
+			vddci = (data->vbios_boot_state.vddci_bootup_value * VOLTAGE_SCALE) << VDDCI_SHIFT;
+
+		table->UvdLevel[count].MinVoltage |= (vddci * VOLTAGE_SCALE) << VDDCI_SHIFT;
+		table->UvdLevel[count].MinVoltage |= 1 << PHASES_SHIFT;
+
+		/* retrieve divider value for VBIOS */
+		result = atomctrl_get_dfs_pll_dividers_vi(hwmgr,
+				table->UvdLevel[count].VclkFrequency, &dividers);
+		PP_ASSERT_WITH_CODE((0 == result),
+				"can not find divide id for Vclk clock", return result);
+
+		table->UvdLevel[count].VclkDivider = (uint8_t)dividers.pll_post_divider;
+
+		result = atomctrl_get_dfs_pll_dividers_vi(hwmgr,
+				table->UvdLevel[count].DclkFrequency, &dividers);
+		PP_ASSERT_WITH_CODE((0 == result),
+				"can not find divide id for Dclk clock", return result);
+
+		table->UvdLevel[count].DclkDivider = (uint8_t)dividers.pll_post_divider;
+
+		CONVERT_FROM_HOST_TO_SMC_UL(table->UvdLevel[count].VclkFrequency);
+		CONVERT_FROM_HOST_TO_SMC_UL(table->UvdLevel[count].DclkFrequency);
+		CONVERT_FROM_HOST_TO_SMC_UL(table->UvdLevel[count].MinVoltage);
+	}
+
+	return result;
+}
+
+static int vegam_populate_smc_boot_level(struct pp_hwmgr *hwmgr,
+		struct SMU75_Discrete_DpmTable *table)
+{
+	int result = 0;
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+
+	table->GraphicsBootLevel = 0;
+	table->MemoryBootLevel = 0;
+
+	/* find boot level from dpm table */
+	result = phm_find_boot_level(&(data->dpm_table.sclk_table),
+			data->vbios_boot_state.sclk_bootup_value,
+			(uint32_t *)&(table->GraphicsBootLevel));
+
+	result = phm_find_boot_level(&(data->dpm_table.mclk_table),
+			data->vbios_boot_state.mclk_bootup_value,
+			(uint32_t *)&(table->MemoryBootLevel));
+
+	table->BootVddc  = data->vbios_boot_state.vddc_bootup_value *
+			VOLTAGE_SCALE;
+	table->BootVddci = data->vbios_boot_state.vddci_bootup_value *
+			VOLTAGE_SCALE;
+	table->BootMVdd  = data->vbios_boot_state.mvdd_bootup_value *
+			VOLTAGE_SCALE;
+
+	CONVERT_FROM_HOST_TO_SMC_US(table->BootVddc);
+	CONVERT_FROM_HOST_TO_SMC_US(table->BootVddci);
+	CONVERT_FROM_HOST_TO_SMC_US(table->BootMVdd);
+
+	return 0;
+}
+
+static int vegam_populate_smc_initial_state(struct pp_hwmgr *hwmgr)
+{
+	struct smu7_hwmgr *hw_data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	uint8_t count, level;
+
+	count = (uint8_t)(table_info->vdd_dep_on_sclk->count);
+
+	for (level = 0; level < count; level++) {
+		if (table_info->vdd_dep_on_sclk->entries[level].clk >=
+				hw_data->vbios_boot_state.sclk_bootup_value) {
+			smu_data->smc_state_table.GraphicsBootLevel = level;
+			break;
+		}
+	}
+
+	count = (uint8_t)(table_info->vdd_dep_on_mclk->count);
+	for (level = 0; level < count; level++) {
+		if (table_info->vdd_dep_on_mclk->entries[level].clk >=
+				hw_data->vbios_boot_state.mclk_bootup_value) {
+			smu_data->smc_state_table.MemoryBootLevel = level;
+			break;
+		}
+	}
+
+	return 0;
+}
+
+static uint16_t scale_fan_gain_settings(uint16_t raw_setting)
+{
+	uint32_t tmp;
+	tmp = raw_setting * 4096 / 100;
+	return (uint16_t)tmp;
+}
+
+static int vegam_populate_bapm_parameters_in_dpm_table(struct pp_hwmgr *hwmgr)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+
+	const struct vegam_pt_defaults *defaults = smu_data->power_tune_defaults;
+	SMU75_Discrete_DpmTable  *table = &(smu_data->smc_state_table);
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	struct phm_cac_tdp_table *cac_dtp_table = table_info->cac_dtp_table;
+	struct pp_advance_fan_control_parameters *fan_table =
+			&hwmgr->thermal_controller.advanceFanControlParameters;
+	int i, j, k;
+	const uint16_t *pdef1;
+	const uint16_t *pdef2;
+
+	table->DefaultTdp = PP_HOST_TO_SMC_US((uint16_t)(cac_dtp_table->usTDP * 128));
+	table->TargetTdp  = PP_HOST_TO_SMC_US((uint16_t)(cac_dtp_table->usTDP * 128));
+
+	PP_ASSERT_WITH_CODE(cac_dtp_table->usTargetOperatingTemp <= 255,
+				"Target Operating Temp is out of Range!",
+				);
+
+	table->TemperatureLimitEdge = PP_HOST_TO_SMC_US(
+			cac_dtp_table->usTargetOperatingTemp * 256);
+	table->TemperatureLimitHotspot = PP_HOST_TO_SMC_US(
+			cac_dtp_table->usTemperatureLimitHotspot * 256);
+	table->FanGainEdge = PP_HOST_TO_SMC_US(
+			scale_fan_gain_settings(fan_table->usFanGainEdge));
+	table->FanGainHotspot = PP_HOST_TO_SMC_US(
+			scale_fan_gain_settings(fan_table->usFanGainHotspot));
+
+	pdef1 = defaults->BAPMTI_R;
+	pdef2 = defaults->BAPMTI_RC;
+
+	for (i = 0; i < SMU75_DTE_ITERATIONS; i++) {
+		for (j = 0; j < SMU75_DTE_SOURCES; j++) {
+			for (k = 0; k < SMU75_DTE_SINKS; k++) {
+				table->BAPMTI_R[i][j][k] = PP_HOST_TO_SMC_US(*pdef1);
+				table->BAPMTI_RC[i][j][k] = PP_HOST_TO_SMC_US(*pdef2);
+				pdef1++;
+				pdef2++;
+			}
+		}
+	}
+
+	return 0;
+}
+
+static int vegam_populate_clock_stretcher_data_table(struct pp_hwmgr *hwmgr)
+{
+	uint32_t ro, efuse, volt_without_cks, volt_with_cks, value, max, min;
+	struct vegam_smumgr *smu_data =
+			(struct vegam_smumgr *)(hwmgr->smu_backend);
+
+	uint8_t i, stretch_amount, stretch_amount2, volt_offset = 0;
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	struct phm_ppt_v1_clock_voltage_dependency_table *sclk_table =
+			table_info->vdd_dep_on_sclk;
+	uint32_t mask = (1 << ((STRAP_ASIC_RO_MSB - STRAP_ASIC_RO_LSB) + 1)) - 1;
+
+	stretch_amount = (uint8_t)table_info->cac_dtp_table->usClockStretchAmount;
+
+	atomctrl_read_efuse(hwmgr, STRAP_ASIC_RO_LSB, STRAP_ASIC_RO_MSB,
+			mask, &efuse);
+
+	min = 1200;
+	max = 2500;
+
+	ro = efuse * (max - min) / 255 + min;
+
+	/* Populate Sclk_CKS_masterEn0_7 and Sclk_voltageOffset */
+	for (i = 0; i < sclk_table->count; i++) {
+		smu_data->smc_state_table.Sclk_CKS_masterEn0_7 |=
+				sclk_table->entries[i].cks_enable << i;
+		volt_without_cks = (uint32_t)((2753594000U + (sclk_table->entries[i].clk/100) *
+				136418 - (ro - 70) * 1000000) /
+				(2424180 - (sclk_table->entries[i].clk/100) * 1132925/1000));
+		volt_with_cks = (uint32_t)((2797202000U + sclk_table->entries[i].clk/100 *
+				3232 - (ro - 65) * 1000000) /
+				(2522480 - sclk_table->entries[i].clk/100 * 115764/100));
+
+		if (volt_without_cks >= volt_with_cks)
+			volt_offset = (uint8_t)(((volt_without_cks - volt_with_cks +
+					sclk_table->entries[i].cks_voffset) * 100 + 624) / 625);
+
+		smu_data->smc_state_table.Sclk_voltageOffset[i] = volt_offset;
+	}
+
+	smu_data->smc_state_table.LdoRefSel =
+			(table_info->cac_dtp_table->ucCKS_LDO_REFSEL != 0) ?
+			table_info->cac_dtp_table->ucCKS_LDO_REFSEL : 5;
+	/* Populate CKS Lookup Table */
+	if (stretch_amount == 1 || stretch_amount == 2 || stretch_amount == 5)
+		stretch_amount2 = 0;
+	else if (stretch_amount == 3 || stretch_amount == 4)
+		stretch_amount2 = 1;
+	else {
+		phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+				PHM_PlatformCaps_ClockStretcher);
+		PP_ASSERT_WITH_CODE(false,
+				"Stretch Amount in PPTable not supported\n",
+				return -EINVAL);
+	}
+
+	value = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixPWR_CKS_CNTL);
+	value &= 0xFFFFFFFE;
+	cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixPWR_CKS_CNTL, value);
+
+	return 0;
+}
+
+static bool vegam_is_hw_avfs_present(struct pp_hwmgr *hwmgr)
+{
+	uint32_t efuse;
+
+	efuse = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+			ixSMU_EFUSE_0 + (49 * 4));
+	efuse &= 0x00000001;
+
+	if (efuse)
+		return true;
+
+	return false;
+}
+
+static int vegam_populate_avfs_parameters(struct pp_hwmgr *hwmgr)
+{
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+
+	SMU75_Discrete_DpmTable  *table = &(smu_data->smc_state_table);
+	int result = 0;
+	struct pp_atom_ctrl__avfs_parameters avfs_params = {0};
+	AVFS_meanNsigma_t AVFS_meanNsigma = { {0} };
+	AVFS_Sclk_Offset_t AVFS_SclkOffset = { {0} };
+	uint32_t tmp, i;
+
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)hwmgr->pptable;
+	struct phm_ppt_v1_clock_voltage_dependency_table *sclk_table =
+			table_info->vdd_dep_on_sclk;
+
+	if (!hwmgr->avfs_supported)
+		return 0;
+
+	result = atomctrl_get_avfs_information(hwmgr, &avfs_params);
+
+	if (0 == result) {
+		table->BTCGB_VDROOP_TABLE[0].a0 =
+				PP_HOST_TO_SMC_UL(avfs_params.ulGB_VDROOP_TABLE_CKSON_a0);
+		table->BTCGB_VDROOP_TABLE[0].a1 =
+				PP_HOST_TO_SMC_UL(avfs_params.ulGB_VDROOP_TABLE_CKSON_a1);
+		table->BTCGB_VDROOP_TABLE[0].a2 =
+				PP_HOST_TO_SMC_UL(avfs_params.ulGB_VDROOP_TABLE_CKSON_a2);
+		table->BTCGB_VDROOP_TABLE[1].a0 =
+				PP_HOST_TO_SMC_UL(avfs_params.ulGB_VDROOP_TABLE_CKSOFF_a0);
+		table->BTCGB_VDROOP_TABLE[1].a1 =
+				PP_HOST_TO_SMC_UL(avfs_params.ulGB_VDROOP_TABLE_CKSOFF_a1);
+		table->BTCGB_VDROOP_TABLE[1].a2 =
+				PP_HOST_TO_SMC_UL(avfs_params.ulGB_VDROOP_TABLE_CKSOFF_a2);
+		table->AVFSGB_FUSE_TABLE[0].m1 =
+				PP_HOST_TO_SMC_UL(avfs_params.ulAVFSGB_FUSE_TABLE_CKSON_m1);
+		table->AVFSGB_FUSE_TABLE[0].m2 =
+				PP_HOST_TO_SMC_US(avfs_params.usAVFSGB_FUSE_TABLE_CKSON_m2);
+		table->AVFSGB_FUSE_TABLE[0].b =
+				PP_HOST_TO_SMC_UL(avfs_params.ulAVFSGB_FUSE_TABLE_CKSON_b);
+		table->AVFSGB_FUSE_TABLE[0].m1_shift = 24;
+		table->AVFSGB_FUSE_TABLE[0].m2_shift = 12;
+		table->AVFSGB_FUSE_TABLE[1].m1 =
+				PP_HOST_TO_SMC_UL(avfs_params.ulAVFSGB_FUSE_TABLE_CKSOFF_m1);
+		table->AVFSGB_FUSE_TABLE[1].m2 =
+				PP_HOST_TO_SMC_US(avfs_params.usAVFSGB_FUSE_TABLE_CKSOFF_m2);
+		table->AVFSGB_FUSE_TABLE[1].b =
+				PP_HOST_TO_SMC_UL(avfs_params.ulAVFSGB_FUSE_TABLE_CKSOFF_b);
+		table->AVFSGB_FUSE_TABLE[1].m1_shift = 24;
+		table->AVFSGB_FUSE_TABLE[1].m2_shift = 12;
+		table->MaxVoltage = PP_HOST_TO_SMC_US(avfs_params.usMaxVoltage_0_25mv);
+		AVFS_meanNsigma.Aconstant[0] =
+				PP_HOST_TO_SMC_UL(avfs_params.ulAVFS_meanNsigma_Acontant0);
+		AVFS_meanNsigma.Aconstant[1] =
+				PP_HOST_TO_SMC_UL(avfs_params.ulAVFS_meanNsigma_Acontant1);
+		AVFS_meanNsigma.Aconstant[2] =
+				PP_HOST_TO_SMC_UL(avfs_params.ulAVFS_meanNsigma_Acontant2);
+		AVFS_meanNsigma.DC_tol_sigma =
+				PP_HOST_TO_SMC_US(avfs_params.usAVFS_meanNsigma_DC_tol_sigma);
+		AVFS_meanNsigma.Platform_mean =
+				PP_HOST_TO_SMC_US(avfs_params.usAVFS_meanNsigma_Platform_mean);
+		AVFS_meanNsigma.PSM_Age_CompFactor =
+				PP_HOST_TO_SMC_US(avfs_params.usPSM_Age_ComFactor);
+		AVFS_meanNsigma.Platform_sigma =
+				PP_HOST_TO_SMC_US(avfs_params.usAVFS_meanNsigma_Platform_sigma);
+
+		for (i = 0; i < sclk_table->count; i++) {
+			AVFS_meanNsigma.Static_Voltage_Offset[i] =
+					(uint8_t)(sclk_table->entries[i].cks_voffset * 100 / 625);
+			AVFS_SclkOffset.Sclk_Offset[i] =
+					PP_HOST_TO_SMC_US((uint16_t)
+							(sclk_table->entries[i].sclk_offset) / 100);
+		}
+
+		result = smu7_read_smc_sram_dword(hwmgr,
+				SMU7_FIRMWARE_HEADER_LOCATION +
+				offsetof(SMU75_Firmware_Header, AvfsMeanNSigma),
+				&tmp, SMC_RAM_END);
+		smu7_copy_bytes_to_smc(hwmgr,
+					tmp,
+					(uint8_t *)&AVFS_meanNsigma,
+					sizeof(AVFS_meanNsigma_t),
+					SMC_RAM_END);
+
+		result = smu7_read_smc_sram_dword(hwmgr,
+				SMU7_FIRMWARE_HEADER_LOCATION +
+				offsetof(SMU75_Firmware_Header, AvfsSclkOffsetTable),
+				&tmp, SMC_RAM_END);
+		smu7_copy_bytes_to_smc(hwmgr,
+					tmp,
+					(uint8_t *)&AVFS_SclkOffset,
+					sizeof(AVFS_Sclk_Offset_t),
+					SMC_RAM_END);
+
+		data->avfs_vdroop_override_setting =
+				(avfs_params.ucEnableGB_VDROOP_TABLE_CKSON << BTCGB0_Vdroop_Enable_SHIFT) |
+				(avfs_params.ucEnableGB_VDROOP_TABLE_CKSOFF << BTCGB1_Vdroop_Enable_SHIFT) |
+				(avfs_params.ucEnableGB_FUSE_TABLE_CKSON << AVFSGB0_Vdroop_Enable_SHIFT) |
+				(avfs_params.ucEnableGB_FUSE_TABLE_CKSOFF << AVFSGB1_Vdroop_Enable_SHIFT);
+		data->apply_avfs_cks_off_voltage =
+				(avfs_params.ucEnableApplyAVFS_CKS_OFF_Voltage == 1) ? true : false;
+	}
+	return result;
+}
+
+static int vegam_populate_vr_config(struct pp_hwmgr *hwmgr,
+		struct SMU75_Discrete_DpmTable *table)
+{
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct vegam_smumgr *smu_data =
+			(struct vegam_smumgr *)(hwmgr->smu_backend);
+	uint16_t config;
+
+	config = VR_MERGED_WITH_VDDC;
+	table->VRConfig |= (config << VRCONF_VDDGFX_SHIFT);
+
+	/* Set Vddc Voltage Controller */
+	if (SMU7_VOLTAGE_CONTROL_BY_SVID2 == data->voltage_control) {
+		config = VR_SVI2_PLANE_1;
+		table->VRConfig |= config;
+	} else {
+		PP_ASSERT_WITH_CODE(false,
+				"VDDC should be on SVI2 control in merged mode!",
+				);
+	}
+	/* Set Vddci Voltage Controller */
+	if (SMU7_VOLTAGE_CONTROL_BY_SVID2 == data->vddci_control) {
+		config = VR_SVI2_PLANE_2;  /* only in merged mode */
+		table->VRConfig |= (config << VRCONF_VDDCI_SHIFT);
+	} else if (SMU7_VOLTAGE_CONTROL_BY_GPIO == data->vddci_control) {
+		config = VR_SMIO_PATTERN_1;
+		table->VRConfig |= (config << VRCONF_VDDCI_SHIFT);
+	} else {
+		config = VR_STATIC_VOLTAGE;
+		table->VRConfig |= (config << VRCONF_VDDCI_SHIFT);
+	}
+	/* Set Mvdd Voltage Controller */
+	if (SMU7_VOLTAGE_CONTROL_BY_SVID2 == data->mvdd_control) {
+		if (config != VR_SVI2_PLANE_2) {
+			config = VR_SVI2_PLANE_2;
+			table->VRConfig |= (config << VRCONF_MVDD_SHIFT);
+			cgs_write_ind_register(hwmgr->device,
+					CGS_IND_REG__SMC,
+					smu_data->smu7_data.soft_regs_start +
+					offsetof(SMU75_SoftRegisters, AllowMvddSwitch),
+					0x1);
+		} else {
+			PP_ASSERT_WITH_CODE(false,
+					"SVI2 Plane 2 is already taken, set MVDD as Static",);
+			config = VR_STATIC_VOLTAGE;
+			table->VRConfig = (config << VRCONF_MVDD_SHIFT);
+		}
+	} else if (SMU7_VOLTAGE_CONTROL_BY_GPIO == data->mvdd_control) {
+		config = VR_SMIO_PATTERN_2;
+		table->VRConfig = (config << VRCONF_MVDD_SHIFT);
+		cgs_write_ind_register(hwmgr->device,
+				CGS_IND_REG__SMC,
+				smu_data->smu7_data.soft_regs_start +
+				offsetof(SMU75_SoftRegisters, AllowMvddSwitch),
+				0x1);
+	} else {
+		config = VR_STATIC_VOLTAGE;
+		table->VRConfig |= (config << VRCONF_MVDD_SHIFT);
+	}
+
+	return 0;
+}
+
+static int vegam_populate_svi_load_line(struct pp_hwmgr *hwmgr)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	const struct vegam_pt_defaults *defaults = smu_data->power_tune_defaults;
+
+	smu_data->power_tune_table.SviLoadLineEn = defaults->SviLoadLineEn;
+	smu_data->power_tune_table.SviLoadLineVddC = defaults->SviLoadLineVddC;
+	smu_data->power_tune_table.SviLoadLineTrimVddC = 3;
+	smu_data->power_tune_table.SviLoadLineOffsetVddC = 0;
+
+	return 0;
+}
+
+static int vegam_populate_tdc_limit(struct pp_hwmgr *hwmgr)
+{
+	uint16_t tdc_limit;
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	const struct vegam_pt_defaults *defaults = smu_data->power_tune_defaults;
+
+	tdc_limit = (uint16_t)(table_info->cac_dtp_table->usTDC * 128);
+	smu_data->power_tune_table.TDC_VDDC_PkgLimit =
+			CONVERT_FROM_HOST_TO_SMC_US(tdc_limit);
+	smu_data->power_tune_table.TDC_VDDC_ThrottleReleaseLimitPerc =
+			defaults->TDC_VDDC_ThrottleReleaseLimitPerc;
+	smu_data->power_tune_table.TDC_MAWt = defaults->TDC_MAWt;
+
+	return 0;
+}
+
+static int vegam_populate_dw8(struct pp_hwmgr *hwmgr, uint32_t fuse_table_offset)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	const struct vegam_pt_defaults *defaults = smu_data->power_tune_defaults;
+	uint32_t temp;
+
+	if (smu7_read_smc_sram_dword(hwmgr,
+			fuse_table_offset +
+			offsetof(SMU75_Discrete_PmFuses, TdcWaterfallCtl),
+			(uint32_t *)&temp, SMC_RAM_END))
+		PP_ASSERT_WITH_CODE(false,
+				"Attempt to read PmFuses.DW6 (SviLoadLineEn) from SMC Failed!",
+				return -EINVAL);
+	else {
+		smu_data->power_tune_table.TdcWaterfallCtl = defaults->TdcWaterfallCtl;
+		smu_data->power_tune_table.LPMLTemperatureMin =
+				(uint8_t)((temp >> 16) & 0xff);
+		smu_data->power_tune_table.LPMLTemperatureMax =
+				(uint8_t)((temp >> 8) & 0xff);
+		smu_data->power_tune_table.Reserved = (uint8_t)(temp & 0xff);
+	}
+	return 0;
+}
+
+static int vegam_populate_temperature_scaler(struct pp_hwmgr *hwmgr)
+{
+	int i;
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+
+	/* Currently not used. Set all to zero. */
+	for (i = 0; i < 16; i++)
+		smu_data->power_tune_table.LPMLTemperatureScaler[i] = 0;
+
+	return 0;
+}
+
+static int vegam_populate_fuzzy_fan(struct pp_hwmgr *hwmgr)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+
+/* TO DO move to hwmgr */
+	if ((hwmgr->thermal_controller.advanceFanControlParameters.usFanOutputSensitivity & (1 << 15))
+		|| 0 == hwmgr->thermal_controller.advanceFanControlParameters.usFanOutputSensitivity)
+		hwmgr->thermal_controller.advanceFanControlParameters.usFanOutputSensitivity =
+			hwmgr->thermal_controller.advanceFanControlParameters.usDefaultFanOutputSensitivity;
+
+	smu_data->power_tune_table.FuzzyFan_PwmSetDelta = PP_HOST_TO_SMC_US(
+				hwmgr->thermal_controller.advanceFanControlParameters.usFanOutputSensitivity);
+	return 0;
+}
+
+static int vegam_populate_gnb_lpml(struct pp_hwmgr *hwmgr)
+{
+	int i;
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+
+	/* Currently not used. Set all to zero. */
+	for (i = 0; i < 16; i++)
+		smu_data->power_tune_table.GnbLPML[i] = 0;
+
+	return 0;
+}
+
+static int vegam_populate_bapm_vddc_base_leakage_sidd(struct pp_hwmgr *hwmgr)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	uint16_t hi_sidd = smu_data->power_tune_table.BapmVddCBaseLeakageHiSidd;
+	uint16_t lo_sidd = smu_data->power_tune_table.BapmVddCBaseLeakageLoSidd;
+	struct phm_cac_tdp_table *cac_table = table_info->cac_dtp_table;
+
+	hi_sidd = (uint16_t)(cac_table->usHighCACLeakage / 100 * 256);
+	lo_sidd = (uint16_t)(cac_table->usLowCACLeakage / 100 * 256);
+
+	smu_data->power_tune_table.BapmVddCBaseLeakageHiSidd =
+			CONVERT_FROM_HOST_TO_SMC_US(hi_sidd);
+	smu_data->power_tune_table.BapmVddCBaseLeakageLoSidd =
+			CONVERT_FROM_HOST_TO_SMC_US(lo_sidd);
+
+	return 0;
+}
+
+static int vegam_populate_pm_fuses(struct pp_hwmgr *hwmgr)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	uint32_t pm_fuse_table_offset;
+
+	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+			PHM_PlatformCaps_PowerContainment)) {
+		if (smu7_read_smc_sram_dword(hwmgr,
+				SMU7_FIRMWARE_HEADER_LOCATION +
+				offsetof(SMU75_Firmware_Header, PmFuseTable),
+				&pm_fuse_table_offset, SMC_RAM_END))
+			PP_ASSERT_WITH_CODE(false,
+					"Attempt to get pm_fuse_table_offset Failed!",
+					return -EINVAL);
+
+		if (vegam_populate_svi_load_line(hwmgr))
+			PP_ASSERT_WITH_CODE(false,
+					"Attempt to populate SviLoadLine Failed!",
+					return -EINVAL);
+
+		if (vegam_populate_tdc_limit(hwmgr))
+			PP_ASSERT_WITH_CODE(false,
+					"Attempt to populate TDCLimit Failed!", return -EINVAL);
+
+		if (vegam_populate_dw8(hwmgr, pm_fuse_table_offset))
+			PP_ASSERT_WITH_CODE(false,
+					"Attempt to populate TdcWaterfallCtl, "
+					"LPMLTemperature Min and Max Failed!",
+					return -EINVAL);
+
+		if (0 != vegam_populate_temperature_scaler(hwmgr))
+			PP_ASSERT_WITH_CODE(false,
+					"Attempt to populate LPMLTemperatureScaler Failed!",
+					return -EINVAL);
+
+		if (vegam_populate_fuzzy_fan(hwmgr))
+			PP_ASSERT_WITH_CODE(false,
+					"Attempt to populate Fuzzy Fan Control parameters Failed!",
+					return -EINVAL);
+
+		if (vegam_populate_gnb_lpml(hwmgr))
+			PP_ASSERT_WITH_CODE(false,
+					"Attempt to populate GnbLPML Failed!",
+					return -EINVAL);
+
+		if (vegam_populate_bapm_vddc_base_leakage_sidd(hwmgr))
+			PP_ASSERT_WITH_CODE(false,
+					"Attempt to populate BapmVddCBaseLeakage Hi and Lo "
+					"Sidd Failed!", return -EINVAL);
+
+		if (smu7_copy_bytes_to_smc(hwmgr, pm_fuse_table_offset,
+				(uint8_t *)&smu_data->power_tune_table,
+				(sizeof(struct SMU75_Discrete_PmFuses) - PMFUSES_AVFSSIZE),
+				SMC_RAM_END))
+			PP_ASSERT_WITH_CODE(false,
+					"Attempt to download PmFuseTable Failed!",
+					return -EINVAL);
+	}
+	return 0;
+}
+
+static int vegam_enable_reconfig_cus(struct pp_hwmgr *hwmgr)
+{
+	struct amdgpu_device *adev = hwmgr->adev;
+
+	smum_send_msg_to_smc_with_parameter(hwmgr,
+					    PPSMC_MSG_EnableModeSwitchRLCNotification,
+					    adev->gfx.cu_info.number);
+
+	return 0;
+}
+
+static int vegam_init_smc_table(struct pp_hwmgr *hwmgr)
+{
+	int result;
+	struct smu7_hwmgr *hw_data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	struct SMU75_Discrete_DpmTable *table = &(smu_data->smc_state_table);
+	uint8_t i;
+	struct pp_atomctrl_gpio_pin_assignment gpio_pin;
+	struct phm_ppt_v1_gpio_table *gpio_table =
+			(struct phm_ppt_v1_gpio_table *)table_info->gpio_table;
+	pp_atomctrl_clock_dividers_vi dividers;
+
+	phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+			PHM_PlatformCaps_AutomaticDCTransition);
+
+	vegam_initialize_power_tune_defaults(hwmgr);
+
+	if (SMU7_VOLTAGE_CONTROL_NONE != hw_data->voltage_control)
+		vegam_populate_smc_voltage_tables(hwmgr, table);
+
+	table->SystemFlags = 0;
+	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+			PHM_PlatformCaps_AutomaticDCTransition))
+		table->SystemFlags |= PPSMC_SYSTEMFLAG_GPIO_DC;
+
+	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+			PHM_PlatformCaps_StepVddc))
+		table->SystemFlags |= PPSMC_SYSTEMFLAG_STEPVDDC;
+
+	if (hw_data->is_memory_gddr5)
+		table->SystemFlags |= PPSMC_SYSTEMFLAG_GDDR5;
+
+	if (hw_data->ulv_supported && table_info->us_ulv_voltage_offset) {
+		result = vegam_populate_ulv_state(hwmgr, table);
+		PP_ASSERT_WITH_CODE(!result,
+				"Failed to initialize ULV state!", return result);
+		cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+				ixCG_ULV_PARAMETER, SMU7_CGULVPARAMETER_DFLT);
+	}
+
+	result = vegam_populate_smc_link_level(hwmgr, table);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to initialize Link Level!", return result);
+
+	result = vegam_populate_all_graphic_levels(hwmgr);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to initialize Graphics Level!", return result);
+
+	result = vegam_populate_all_memory_levels(hwmgr);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to initialize Memory Level!", return result);
+
+	result = vegam_populate_smc_acpi_level(hwmgr, table);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to initialize ACPI Level!", return result);
+
+	result = vegam_populate_smc_vce_level(hwmgr, table);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to initialize VCE Level!", return result);
+
+	result = vegam_populate_smc_samu_level(hwmgr, table);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to initialize SAMU Level!", return result);
+
+	/* Since only the initial state is completely set up at this point
+	 * (the other states are just copies of the boot state) we only
+	 * need to populate the  ARB settings for the initial state.
+	 */
+	result = vegam_program_memory_timing_parameters(hwmgr);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to Write ARB settings for the initial state.", return result);
+
+	result = vegam_populate_smc_uvd_level(hwmgr, table);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to initialize UVD Level!", return result);
+
+	result = vegam_populate_smc_boot_level(hwmgr, table);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to initialize Boot Level!", return result);
+
+	result = vegam_populate_smc_initial_state(hwmgr);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to initialize Boot State!", return result);
+
+	result = vegam_populate_bapm_parameters_in_dpm_table(hwmgr);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to populate BAPM Parameters!", return result);
+
+	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+			PHM_PlatformCaps_ClockStretcher)) {
+		result = vegam_populate_clock_stretcher_data_table(hwmgr);
+		PP_ASSERT_WITH_CODE(!result,
+				"Failed to populate Clock Stretcher Data Table!",
+				return result);
+	}
+
+	result = vegam_populate_avfs_parameters(hwmgr);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to populate AVFS Parameters!", return result;);
+
+	table->CurrSclkPllRange = 0xff;
+	table->GraphicsVoltageChangeEnable  = 1;
+	table->GraphicsThermThrottleEnable  = 1;
+	table->GraphicsInterval = 1;
+	table->VoltageInterval  = 1;
+	table->ThermalInterval  = 1;
+	table->TemperatureLimitHigh =
+			table_info->cac_dtp_table->usTargetOperatingTemp *
+			SMU7_Q88_FORMAT_CONVERSION_UNIT;
+	table->TemperatureLimitLow  =
+			(table_info->cac_dtp_table->usTargetOperatingTemp - 1) *
+			SMU7_Q88_FORMAT_CONVERSION_UNIT;
+	table->MemoryVoltageChangeEnable = 1;
+	table->MemoryInterval = 1;
+	table->VoltageResponseTime = 0;
+	table->PhaseResponseTime = 0;
+	table->MemoryThermThrottleEnable = 1;
+
+	PP_ASSERT_WITH_CODE(hw_data->dpm_table.pcie_speed_table.count >= 1,
+			"There must be 1 or more PCIE levels defined in PPTable.",
+			return -EINVAL);
+	table->PCIeBootLinkLevel =
+			hw_data->dpm_table.pcie_speed_table.count;
+	table->PCIeGenInterval = 1;
+	table->VRConfig = 0;
+
+	result = vegam_populate_vr_config(hwmgr, table);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to populate VRConfig setting!", return result);
+
+	table->ThermGpio = 17;
+	table->SclkStepSize = 0x4000;
+
+	if (atomctrl_get_pp_assign_pin(hwmgr,
+			VDDC_VRHOT_GPIO_PINID, &gpio_pin)) {
+		table->VRHotGpio = gpio_pin.uc_gpio_pin_bit_shift;
+		if (gpio_table)
+			table->VRHotLevel =
+					table_info->gpio_table->vrhot_triggered_sclk_dpm_index;
+	} else {
+		table->VRHotGpio = SMU7_UNUSED_GPIO_PIN;
+		phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+				PHM_PlatformCaps_RegulatorHot);
+	}
+
+	if (atomctrl_get_pp_assign_pin(hwmgr,
+			PP_AC_DC_SWITCH_GPIO_PINID,	&gpio_pin)) {
+		table->AcDcGpio = gpio_pin.uc_gpio_pin_bit_shift;
+		if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+				PHM_PlatformCaps_AutomaticDCTransition) &&
+				!smum_send_msg_to_smc(hwmgr, PPSMC_MSG_UseNewGPIOScheme))
+			phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+					PHM_PlatformCaps_SMCtoPPLIBAcdcGpioScheme);
+	} else {
+		table->AcDcGpio = SMU7_UNUSED_GPIO_PIN;
+		phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+				PHM_PlatformCaps_AutomaticDCTransition);
+	}
+
+	/* Thermal Output GPIO */
+	if (atomctrl_get_pp_assign_pin(hwmgr,
+			THERMAL_INT_OUTPUT_GPIO_PINID, &gpio_pin)) {
+		table->ThermOutGpio = gpio_pin.uc_gpio_pin_bit_shift;
+
+		/* For porlarity read GPIOPAD_A with assigned Gpio pin
+		 * since VBIOS will program this register to set 'inactive state',
+		 * driver can then determine 'active state' from this and
+		 * program SMU with correct polarity
+		 */
+		table->ThermOutPolarity =
+				(0 == (cgs_read_register(hwmgr->device, mmGPIOPAD_A) &
+				(1 << gpio_pin.uc_gpio_pin_bit_shift))) ? 1:0;
+		table->ThermOutMode = SMU7_THERM_OUT_MODE_THERM_ONLY;
+
+		/* if required, combine VRHot/PCC with thermal out GPIO */
+		if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+				PHM_PlatformCaps_RegulatorHot) &&
+			phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+				PHM_PlatformCaps_CombinePCCWithThermalSignal))
+			table->ThermOutMode = SMU7_THERM_OUT_MODE_THERM_VRHOT;
+	} else {
+		table->ThermOutGpio = 17;
+		table->ThermOutPolarity = 1;
+		table->ThermOutMode = SMU7_THERM_OUT_MODE_DISABLE;
+	}
+
+	/* Populate BIF_SCLK levels into SMC DPM table */
+	for (i = 0; i <= hw_data->dpm_table.pcie_speed_table.count; i++) {
+		result = atomctrl_get_dfs_pll_dividers_vi(hwmgr,
+				smu_data->bif_sclk_table[i], &dividers);
+		PP_ASSERT_WITH_CODE(!result,
+				"Can not find DFS divide id for Sclk",
+				return result);
+
+		if (i == 0)
+			table->Ulv.BifSclkDfs =
+					PP_HOST_TO_SMC_US((uint16_t)(dividers.pll_post_divider));
+		else
+			table->LinkLevel[i - 1].BifSclkDfs =
+					PP_HOST_TO_SMC_US((uint16_t)(dividers.pll_post_divider));
+	}
+
+	for (i = 0; i < SMU75_MAX_ENTRIES_SMIO; i++)
+		table->Smio[i] = PP_HOST_TO_SMC_UL(table->Smio[i]);
+
+	CONVERT_FROM_HOST_TO_SMC_UL(table->SystemFlags);
+	CONVERT_FROM_HOST_TO_SMC_UL(table->VRConfig);
+	CONVERT_FROM_HOST_TO_SMC_UL(table->SmioMask1);
+	CONVERT_FROM_HOST_TO_SMC_UL(table->SmioMask2);
+	CONVERT_FROM_HOST_TO_SMC_UL(table->SclkStepSize);
+	CONVERT_FROM_HOST_TO_SMC_UL(table->CurrSclkPllRange);
+	CONVERT_FROM_HOST_TO_SMC_US(table->TemperatureLimitHigh);
+	CONVERT_FROM_HOST_TO_SMC_US(table->TemperatureLimitLow);
+	CONVERT_FROM_HOST_TO_SMC_US(table->VoltageResponseTime);
+	CONVERT_FROM_HOST_TO_SMC_US(table->PhaseResponseTime);
+
+	/* Upload all dpm data to SMC memory.(dpm level, dpm level count etc) */
+	result = smu7_copy_bytes_to_smc(hwmgr,
+			smu_data->smu7_data.dpm_table_start +
+			offsetof(SMU75_Discrete_DpmTable, SystemFlags),
+			(uint8_t *)&(table->SystemFlags),
+			sizeof(SMU75_Discrete_DpmTable) - 3 * sizeof(SMU75_PIDController),
+			SMC_RAM_END);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to upload dpm data to SMC memory!", return result);
+
+	result = vegam_populate_pm_fuses(hwmgr);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to  populate PM fuses to SMC memory!", return result);
+
+	result = vegam_enable_reconfig_cus(hwmgr);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to enable reconfigurable CUs!", return result);
+
+	return 0;
+}
+
+static uint32_t vegam_get_offsetof(uint32_t type, uint32_t member)
+{
+	switch (type) {
+	case SMU_SoftRegisters:
+		switch (member) {
+		case HandshakeDisables:
+			return offsetof(SMU75_SoftRegisters, HandshakeDisables);
+		case VoltageChangeTimeout:
+			return offsetof(SMU75_SoftRegisters, VoltageChangeTimeout);
+		case AverageGraphicsActivity:
+			return offsetof(SMU75_SoftRegisters, AverageGraphicsActivity);
+		case PreVBlankGap:
+			return offsetof(SMU75_SoftRegisters, PreVBlankGap);
+		case VBlankTimeout:
+			return offsetof(SMU75_SoftRegisters, VBlankTimeout);
+		case UcodeLoadStatus:
+			return offsetof(SMU75_SoftRegisters, UcodeLoadStatus);
+		case DRAM_LOG_ADDR_H:
+			return offsetof(SMU75_SoftRegisters, DRAM_LOG_ADDR_H);
+		case DRAM_LOG_ADDR_L:
+			return offsetof(SMU75_SoftRegisters, DRAM_LOG_ADDR_L);
+		case DRAM_LOG_PHY_ADDR_H:
+			return offsetof(SMU75_SoftRegisters, DRAM_LOG_PHY_ADDR_H);
+		case DRAM_LOG_PHY_ADDR_L:
+			return offsetof(SMU75_SoftRegisters, DRAM_LOG_PHY_ADDR_L);
+		case DRAM_LOG_BUFF_SIZE:
+			return offsetof(SMU75_SoftRegisters, DRAM_LOG_BUFF_SIZE);
+		}
+	case SMU_Discrete_DpmTable:
+		switch (member) {
+		case UvdBootLevel:
+			return offsetof(SMU75_Discrete_DpmTable, UvdBootLevel);
+		case VceBootLevel:
+			return offsetof(SMU75_Discrete_DpmTable, VceBootLevel);
+		case SamuBootLevel:
+			return offsetof(SMU75_Discrete_DpmTable, SamuBootLevel);
+		case LowSclkInterruptThreshold:
+			return offsetof(SMU75_Discrete_DpmTable, LowSclkInterruptThreshold);
+		}
+	}
+	pr_warn("can't get the offset of type %x member %x\n", type, member);
+	return 0;
+}
+
+static int vegam_program_mem_timing_parameters(struct pp_hwmgr *hwmgr)
+{
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+
+	if (data->need_update_smu7_dpm_table &
+		(DPMTABLE_OD_UPDATE_SCLK +
+		DPMTABLE_UPDATE_SCLK +
+		DPMTABLE_UPDATE_MCLK))
+		return vegam_program_memory_timing_parameters(hwmgr);
+
+	return 0;
+}
+
+static int vegam_update_sclk_threshold(struct pp_hwmgr *hwmgr)
+{
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct vegam_smumgr *smu_data =
+			(struct vegam_smumgr *)(hwmgr->smu_backend);
+	int result = 0;
+	uint32_t low_sclk_interrupt_threshold = 0;
+
+	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+			PHM_PlatformCaps_SclkThrottleLowNotification)
+	    && (data->low_sclk_interrupt_threshold != 0)) {
+		low_sclk_interrupt_threshold =
+				data->low_sclk_interrupt_threshold;
+
+		CONVERT_FROM_HOST_TO_SMC_UL(low_sclk_interrupt_threshold);
+
+		result = smu7_copy_bytes_to_smc(
+				hwmgr,
+				smu_data->smu7_data.dpm_table_start +
+				offsetof(SMU75_Discrete_DpmTable,
+					LowSclkInterruptThreshold),
+				(uint8_t *)&low_sclk_interrupt_threshold,
+				sizeof(uint32_t),
+				SMC_RAM_END);
+	}
+	PP_ASSERT_WITH_CODE((result == 0),
+			"Failed to update SCLK threshold!", return result);
+
+	result = vegam_program_mem_timing_parameters(hwmgr);
+	PP_ASSERT_WITH_CODE((result == 0),
+			"Failed to program memory timing parameters!",
+			);
+
+	return result;
+}
+
+int vegam_thermal_avfs_enable(struct pp_hwmgr *hwmgr)
+{
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	int ret;
+
+	if (!hwmgr->avfs_supported)
+		return 0;
+
+	ret = smum_send_msg_to_smc(hwmgr, PPSMC_MSG_EnableAvfs);
+	if (!ret) {
+		if (data->apply_avfs_cks_off_voltage)
+			ret = smum_send_msg_to_smc(hwmgr, PPSMC_MSG_ApplyAvfsCksOffVoltage);
+	}
+
+	return ret;
+}
+
+static int vegam_thermal_setup_fan_table(struct pp_hwmgr *hwmgr)
+{
+	PP_ASSERT_WITH_CODE(hwmgr->thermal_controller.fanInfo.bNoFan,
+			"VBIOS fan info is not correct!",
+			);
+	phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+			PHM_PlatformCaps_MicrocodeFanControl);
+	return 0;
+}
+
+const struct pp_smumgr_func vegam_smu_funcs = {
+	.smu_init = vegam_smu_init,
+	.smu_fini = smu7_smu_fini,
+	.start_smu = vegam_start_smu,
+	.check_fw_load_finish = smu7_check_fw_load_finish,
+	.request_smu_load_fw = smu7_reload_firmware,
+	.request_smu_load_specific_fw = NULL,
+	.send_msg_to_smc = smu7_send_msg_to_smc,
+	.send_msg_to_smc_with_parameter = smu7_send_msg_to_smc_with_parameter,
+	.process_firmware_header = vegam_process_firmware_header,
+	.is_dpm_running = vegam_is_dpm_running,
+	.get_mac_definition = vegam_get_mac_definition,
+	.update_smc_table = vegam_update_smc_table,
+	.init_smc_table = vegam_init_smc_table,
+	.get_offsetof = vegam_get_offsetof,
+	.populate_all_graphic_levels = vegam_populate_all_graphic_levels,
+	.populate_all_memory_levels = vegam_populate_all_memory_levels,
+	.update_sclk_threshold = vegam_update_sclk_threshold,
+	.is_hw_avfs_present = vegam_is_hw_avfs_present,
+	.thermal_avfs_enable = vegam_thermal_avfs_enable,
+	.is_dpm_running = vegam_is_dpm_running,
+	.thermal_setup_fan_table = vegam_thermal_setup_fan_table,
+};
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.h b/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.h
new file mode 100644
index 0000000..2b65582
--- /dev/null
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.h
@@ -0,0 +1,75 @@
+/*
+ * Copyright 2017 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef _VEGAM_SMUMANAGER_H
+#define _VEGAM_SMUMANAGER_H
+
+
+#include <pp_endian.h>
+#include "smu75_discrete.h"
+#include "smu7_smumgr.h"
+
+#define SMC_RAM_END 0x40000
+
+#define DPMTuning_Uphyst_Shift    0
+#define DPMTuning_Downhyst_Shift  8
+#define DPMTuning_Activity_Shift  16
+
+#define GraphicsDPMTuning_VEGAM    0x001e6400
+#define MemoryDPMTuning_VEGAM      0x000f3c0a
+#define SclkDPMTuning_VEGAM        0x002d000a
+#define MclkDPMTuning_VEGAM        0x001f100a
+
+
+struct vegam_pt_defaults {
+	uint8_t   SviLoadLineEn;
+	uint8_t   SviLoadLineVddC;
+	uint8_t   TDC_VDDC_ThrottleReleaseLimitPerc;
+	uint8_t   TDC_MAWt;
+	uint8_t   TdcWaterfallCtl;
+	uint8_t   DTEAmbientTempBase;
+
+	uint32_t  DisplayCac;
+	uint32_t  BAPM_TEMP_GRADIENT;
+	uint16_t  BAPMTI_R[SMU75_DTE_ITERATIONS * SMU75_DTE_SOURCES * SMU75_DTE_SINKS];
+	uint16_t  BAPMTI_RC[SMU75_DTE_ITERATIONS * SMU75_DTE_SOURCES * SMU75_DTE_SINKS];
+};
+
+struct vegam_range_table {
+	uint32_t trans_lower_frequency; /* in 10khz */
+	uint32_t trans_upper_frequency;
+};
+
+struct vegam_smumgr {
+	struct smu7_smumgr smu7_data;
+	uint8_t protected_mode;
+	SMU75_Discrete_DpmTable              smc_state_table;
+	struct SMU75_Discrete_Ulv            ulv_setting;
+	struct SMU75_Discrete_PmFuses  power_tune_table;
+	struct vegam_range_table                range_table[NUM_SCLK_RANGE];
+	const struct vegam_pt_defaults       *power_tune_defaults;
+	uint32_t               bif_sclk_table[SMU75_MAX_LEVELS_LINK];
+};
+
+
+#endif
-- 
cgit v1.1


From 0c24e7ef233b528699798a3db3ab57ee0317f2f0 Mon Sep 17 00:00:00 2001
From: Eric Huang <JinHuiEric.Huang@amd.com>
Date: Wed, 11 Apr 2018 15:38:11 -0500
Subject: drm/amd/powerplay: add specific changes for VEGAM in smu7_hwmgr.c

VEGAM specific changes for smu7:
1. add avfs control.
2. add a smc message defferent as smu7.
3. don't switch mc arb memory timing.
4. update LCAC_MC0/1_CNTL value.

Signed-off-by: Eric Huang <JinHuiEric.Huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c | 71 ++++++++++++++++++++----
 1 file changed, 61 insertions(+), 10 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index c9dd0be..4c94e7a 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -83,6 +83,14 @@ static const struct profile_mode_setting smu7_profiling[5] =
 					 {1, 0, 5, 30, 0, 0, 0, 0},
 					};
 
+#define PPSMC_MSG_SetVBITimeout_VEGAM    ((uint16_t) 0x310)
+
+#define ixPWR_SVI2_PLANE1_LOAD                     0xC0200280
+#define PWR_SVI2_PLANE1_LOAD__PSI1_MASK                    0x00000020L
+#define PWR_SVI2_PLANE1_LOAD__PSI0_EN_MASK                 0x00000040L
+#define PWR_SVI2_PLANE1_LOAD__PSI1__SHIFT                  0x00000005
+#define PWR_SVI2_PLANE1_LOAD__PSI0_EN__SHIFT               0x00000006
+
 /** Values for the CG_THERMAL_CTRL::DPM_EVENT_SRC field. */
 enum DPM_EVENT_SRC {
 	DPM_EVENT_SRC_ANALOG = 0,
@@ -164,6 +172,13 @@ static int smu7_get_current_pcie_lane_number(struct pp_hwmgr *hwmgr)
 */
 static int smu7_enable_smc_voltage_controller(struct pp_hwmgr *hwmgr)
 {
+	if (hwmgr->chip_id == CHIP_VEGAM) {
+		PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device,
+				CGS_IND_REG__SMC, PWR_SVI2_PLANE1_LOAD, PSI1, 0);
+		PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device,
+				CGS_IND_REG__SMC, PWR_SVI2_PLANE1_LOAD, PSI0_EN, 0);
+	}
+
 	if (hwmgr->feature_mask & PP_SMC_VOLTAGE_CONTROL_MASK)
 		smum_send_msg_to_smc(hwmgr, PPSMC_MSG_Voltage_Cntl_Enable);
 
@@ -964,6 +979,22 @@ static int smu7_disable_deep_sleep_master_switch(struct pp_hwmgr *hwmgr)
 	return 0;
 }
 
+static int smu7_disable_sclk_vce_handshake(struct pp_hwmgr *hwmgr)
+{
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	uint32_t soft_register_value = 0;
+	uint32_t handshake_disables_offset = data->soft_regs_start
+				+ smum_get_offsetof(hwmgr,
+					SMU_SoftRegisters, HandshakeDisables);
+
+	soft_register_value = cgs_read_ind_register(hwmgr->device,
+				CGS_IND_REG__SMC, handshake_disables_offset);
+	soft_register_value |= SMU7_VCE_SCLK_HANDSHAKE_DISABLE;
+	cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+			handshake_disables_offset, soft_register_value);
+	return 0;
+}
+
 static int smu7_disable_handshake_uvd(struct pp_hwmgr *hwmgr)
 {
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
@@ -987,6 +1018,9 @@ static int smu7_enable_sclk_mclk_dpm(struct pp_hwmgr *hwmgr)
 
 	/* enable SCLK dpm */
 	if (!data->sclk_dpm_key_disabled)
+		if (hwmgr->chip_id == CHIP_VEGAM)
+			smu7_disable_sclk_vce_handshake(hwmgr);
+
 		PP_ASSERT_WITH_CODE(
 		(0 == smum_send_msg_to_smc(hwmgr, PPSMC_MSG_DPM_Enable)),
 		"Failed to enable SCLK DPM during DPM Start Function!",
@@ -996,13 +1030,15 @@ static int smu7_enable_sclk_mclk_dpm(struct pp_hwmgr *hwmgr)
 	if (0 == data->mclk_dpm_key_disabled) {
 		if (!(hwmgr->feature_mask & PP_UVD_HANDSHAKE_MASK))
 			smu7_disable_handshake_uvd(hwmgr);
+
 		PP_ASSERT_WITH_CODE(
 				(0 == smum_send_msg_to_smc(hwmgr,
 						PPSMC_MSG_MCLKDPM_Enable)),
 				"Failed to enable MCLK DPM during DPM Start Function!",
 				return -EINVAL);
 
-		PHM_WRITE_FIELD(hwmgr->device, MC_SEQ_CNTL_3, CAC_EN, 0x1);
+		if (hwmgr->chip_family != CHIP_VEGAM)
+			PHM_WRITE_FIELD(hwmgr->device, MC_SEQ_CNTL_3, CAC_EN, 0x1);
 
 
 		if (hwmgr->chip_family == AMDGPU_FAMILY_CI) {
@@ -1018,8 +1054,13 @@ static int smu7_enable_sclk_mclk_dpm(struct pp_hwmgr *hwmgr)
 			cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixLCAC_MC1_CNTL, 0x5);
 			cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixLCAC_CPL_CNTL, 0x100005);
 			udelay(10);
-			cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixLCAC_MC0_CNTL, 0x400005);
-			cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixLCAC_MC1_CNTL, 0x400005);
+			if (hwmgr->chip_id == CHIP_VEGAM) {
+				cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixLCAC_MC0_CNTL, 0x400009);
+				cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixLCAC_MC1_CNTL, 0x400009);
+			} else {
+				cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixLCAC_MC0_CNTL, 0x400005);
+				cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixLCAC_MC1_CNTL, 0x400005);
+			}
 			cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixLCAC_CPL_CNTL, 0x500005);
 		}
 	}
@@ -1260,10 +1301,12 @@ static int smu7_enable_dpm_tasks(struct pp_hwmgr *hwmgr)
 	PP_ASSERT_WITH_CODE((0 == tmp_result),
 			"Failed to process firmware header!", result = tmp_result);
 
-	tmp_result = smu7_initial_switch_from_arbf0_to_f1(hwmgr);
-	PP_ASSERT_WITH_CODE((0 == tmp_result),
-			"Failed to initialize switch from ArbF0 to F1!",
-			result = tmp_result);
+	if (hwmgr->chip_id != CHIP_VEGAM) {
+		tmp_result = smu7_initial_switch_from_arbf0_to_f1(hwmgr);
+		PP_ASSERT_WITH_CODE((0 == tmp_result),
+				"Failed to initialize switch from ArbF0 to F1!",
+				result = tmp_result);
+	}
 
 	result = smu7_setup_default_dpm_tables(hwmgr);
 	PP_ASSERT_WITH_CODE(0 == result,
@@ -2753,6 +2796,9 @@ static int smu7_vblank_too_short(struct pp_hwmgr *hwmgr,
 	case CHIP_POLARIS12:
 		switch_limit_us = data->is_memory_gddr5 ? 190 : 150;
 		break;
+	case CHIP_VEGAM:
+		switch_limit_us = 30;
+		break;
 	default:
 		switch_limit_us = data->is_memory_gddr5 ? 450 : 150;
 		break;
@@ -3801,9 +3847,14 @@ static int smu7_notify_smc_display(struct pp_hwmgr *hwmgr)
 {
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
 
-	if (hwmgr->feature_mask & PP_VBI_TIME_SUPPORT_MASK)
-		smum_send_msg_to_smc_with_parameter(hwmgr,
-			(PPSMC_Msg)PPSMC_MSG_SetVBITimeout, data->frame_time_x2);
+	if (hwmgr->feature_mask & PP_VBI_TIME_SUPPORT_MASK) {
+		if (hwmgr->chip_id == CHIP_VEGAM)
+			smum_send_msg_to_smc_with_parameter(hwmgr,
+					(PPSMC_Msg)PPSMC_MSG_SetVBITimeout_VEGAM, data->frame_time_x2);
+		else
+			smum_send_msg_to_smc_with_parameter(hwmgr,
+					(PPSMC_Msg)PPSMC_MSG_SetVBITimeout, data->frame_time_x2);
+	}
 	return (smum_send_msg_to_smc(hwmgr, (PPSMC_Msg)PPSMC_HasDisplay) == 0) ?  0 : -EINVAL;
 }
 
-- 
cgit v1.1


From eda8377d1628caa07b642fce8fcd1938010e949f Mon Sep 17 00:00:00 2001
From: Eric Huang <JinHuiEric.Huang@amd.com>
Date: Wed, 11 Apr 2018 18:23:54 -0500
Subject: drm/powerplay: Add powertune table for VEGAM

Add the powertune table for VEGAM.

Signed-off-by: Eric Huang <JinHuiEric.Huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c   | 189 +++++++++++++++++++++
 1 file changed, 189 insertions(+)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c
index a55ee16..a264e0c 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c
@@ -623,6 +623,190 @@ static const struct gpu_pt_config_reg DIDTConfig_Polaris11_Kicker[] =
 	{   0xFFFFFFFF  }  /* End of list */
 };
 
+static const struct gpu_pt_config_reg GCCACConfig_VegaM[] =
+{
+// ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
+//      Offset                             Mask                                                Shift                                               Value       Type
+// ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
+    // DIDT_SQ
+    //
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x00060013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x00860013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x01060013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x01860013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x02060013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x02860013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x03060013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x03860013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x04060013, GPU_CONFIGREG_GC_CAC_IND },
+
+    // DIDT_TD
+    //
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x000E0013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x008E0013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x010E0013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x018E0013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x020E0013, GPU_CONFIGREG_GC_CAC_IND },
+
+    // DIDT_TCP
+    //
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x00100013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x00900013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x01100013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x01900013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x02100013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x02900013, GPU_CONFIGREG_GC_CAC_IND },
+
+    {   0xFFFFFFFF  }  // End of list
+};
+
+static const struct gpu_pt_config_reg DIDTConfig_VegaM[] =
+{
+// ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
+//      Offset                             Mask                                                Shift                                               Value       Type
+// ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
+    // DIDT_SQ
+    //
+    {   ixDIDT_SQ_WEIGHT0_3,               DIDT_SQ_WEIGHT0_3__WEIGHT0_MASK,                    DIDT_SQ_WEIGHT0_3__WEIGHT0__SHIFT,                  0x0073,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_WEIGHT0_3,               DIDT_SQ_WEIGHT0_3__WEIGHT1_MASK,                    DIDT_SQ_WEIGHT0_3__WEIGHT1__SHIFT,                  0x00ab,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_WEIGHT0_3,               DIDT_SQ_WEIGHT0_3__WEIGHT2_MASK,                    DIDT_SQ_WEIGHT0_3__WEIGHT2__SHIFT,                  0x0084,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_WEIGHT0_3,               DIDT_SQ_WEIGHT0_3__WEIGHT3_MASK,                    DIDT_SQ_WEIGHT0_3__WEIGHT3__SHIFT,                  0x005a,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_SQ_WEIGHT4_7,               DIDT_SQ_WEIGHT4_7__WEIGHT4_MASK,                    DIDT_SQ_WEIGHT4_7__WEIGHT4__SHIFT,                  0x0067,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_WEIGHT4_7,               DIDT_SQ_WEIGHT4_7__WEIGHT5_MASK,                    DIDT_SQ_WEIGHT4_7__WEIGHT5__SHIFT,                  0x0084,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_WEIGHT4_7,               DIDT_SQ_WEIGHT4_7__WEIGHT6_MASK,                    DIDT_SQ_WEIGHT4_7__WEIGHT6__SHIFT,                  0x0027,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_WEIGHT4_7,               DIDT_SQ_WEIGHT4_7__WEIGHT7_MASK,                    DIDT_SQ_WEIGHT4_7__WEIGHT7__SHIFT,                  0x0046,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_SQ_WEIGHT8_11,              DIDT_SQ_WEIGHT8_11__WEIGHT8_MASK,                   DIDT_SQ_WEIGHT8_11__WEIGHT8__SHIFT,                 0x00aa,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_WEIGHT8_11,              DIDT_SQ_WEIGHT8_11__WEIGHT9_MASK,                   DIDT_SQ_WEIGHT8_11__WEIGHT9__SHIFT,                 0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_WEIGHT8_11,              DIDT_SQ_WEIGHT8_11__WEIGHT10_MASK,                  DIDT_SQ_WEIGHT8_11__WEIGHT10__SHIFT,                0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_WEIGHT8_11,              DIDT_SQ_WEIGHT8_11__WEIGHT11_MASK,                  DIDT_SQ_WEIGHT8_11__WEIGHT11__SHIFT,                0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_SQ_CTRL1,                   DIDT_SQ_CTRL1__MIN_POWER_MASK,                      DIDT_SQ_CTRL1__MIN_POWER__SHIFT,                    0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL1,                   DIDT_SQ_CTRL1__MAX_POWER_MASK,                      DIDT_SQ_CTRL1__MAX_POWER__SHIFT,                    0xffff,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_SQ_CTRL_OCP,                DIDT_SQ_CTRL_OCP__UNUSED_0_MASK,                    DIDT_SQ_CTRL_OCP__UNUSED_0__SHIFT,                  0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL_OCP,                DIDT_SQ_CTRL_OCP__OCP_MAX_POWER_MASK,               DIDT_SQ_CTRL_OCP__OCP_MAX_POWER__SHIFT,             0xffff,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_SQ_CTRL2,                   DIDT_SQ_CTRL2__MAX_POWER_DELTA_MASK,                DIDT_SQ_CTRL2__MAX_POWER_DELTA__SHIFT,              0x3853,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL2,                   DIDT_SQ_CTRL2__UNUSED_0_MASK,                       DIDT_SQ_CTRL2__UNUSED_0__SHIFT,                     0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL2,                   DIDT_SQ_CTRL2__SHORT_TERM_INTERVAL_SIZE_MASK,       DIDT_SQ_CTRL2__SHORT_TERM_INTERVAL_SIZE__SHIFT,     0x005a,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL2,                   DIDT_SQ_CTRL2__UNUSED_1_MASK,                       DIDT_SQ_CTRL2__UNUSED_1__SHIFT,                     0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL2,                   DIDT_SQ_CTRL2__LONG_TERM_INTERVAL_RATIO_MASK,       DIDT_SQ_CTRL2__LONG_TERM_INTERVAL_RATIO__SHIFT,     0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL2,                   DIDT_SQ_CTRL2__UNUSED_2_MASK,                       DIDT_SQ_CTRL2__UNUSED_2__SHIFT,                     0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_SQ_STALL_CTRL,              DIDT_SQ_STALL_CTRL__DIDT_STALL_CTRL_ENABLE_MASK,    DIDT_SQ_STALL_CTRL__DIDT_STALL_CTRL_ENABLE__SHIFT,  0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_STALL_CTRL,              DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_HI_MASK,       DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_HI__SHIFT,     0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_STALL_CTRL,              DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_LO_MASK,       DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_LO__SHIFT,     0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_STALL_CTRL,              DIDT_SQ_STALL_CTRL__DIDT_HI_POWER_THRESHOLD_MASK,   DIDT_SQ_STALL_CTRL__DIDT_HI_POWER_THRESHOLD__SHIFT, 0x0ebb,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_STALL_CTRL,              DIDT_SQ_STALL_CTRL__UNUSED_0_MASK,                  DIDT_SQ_STALL_CTRL__UNUSED_0__SHIFT,                0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_SQ_TUNING_CTRL,             DIDT_SQ_TUNING_CTRL__DIDT_TUNING_ENABLE_MASK,       DIDT_SQ_TUNING_CTRL__DIDT_TUNING_ENABLE__SHIFT,     0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_TUNING_CTRL,             DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_HI_MASK,       DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_HI__SHIFT,     0x3853,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_TUNING_CTRL,             DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_LO_MASK,       DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_LO__SHIFT,     0x3153,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_TUNING_CTRL,             DIDT_SQ_TUNING_CTRL__UNUSED_0_MASK,                 DIDT_SQ_TUNING_CTRL__UNUSED_0__SHIFT,               0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_SQ_CTRL0,                   DIDT_SQ_CTRL0__DIDT_CTRL_EN_MASK,                   DIDT_SQ_CTRL0__DIDT_CTRL_EN__SHIFT,                 0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL0,                   DIDT_SQ_CTRL0__USE_REF_CLOCK_MASK,                  DIDT_SQ_CTRL0__USE_REF_CLOCK__SHIFT,                0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL0,                   DIDT_SQ_CTRL0__PHASE_OFFSET_MASK,                   DIDT_SQ_CTRL0__PHASE_OFFSET__SHIFT,                 0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL0,                   DIDT_SQ_CTRL0__DIDT_CTRL_RST_MASK,                  DIDT_SQ_CTRL0__DIDT_CTRL_RST__SHIFT,                0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL0,                   DIDT_SQ_CTRL0__DIDT_CLK_EN_OVERRIDE_MASK,           DIDT_SQ_CTRL0__DIDT_CLK_EN_OVERRIDE__SHIFT,         0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL0,                   DIDT_SQ_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI_MASK,     DIDT_SQ_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI__SHIFT,   0x0010,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL0,                   DIDT_SQ_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO_MASK,     DIDT_SQ_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO__SHIFT,   0x0010,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL0,                   DIDT_SQ_CTRL0__UNUSED_0_MASK,                       DIDT_SQ_CTRL0__UNUSED_0__SHIFT,                     0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    // DIDT_TD
+    //
+    {   ixDIDT_TD_WEIGHT0_3,               DIDT_TD_WEIGHT0_3__WEIGHT0_MASK,                    DIDT_TD_WEIGHT0_3__WEIGHT0__SHIFT,                  0x000a,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_WEIGHT0_3,               DIDT_TD_WEIGHT0_3__WEIGHT1_MASK,                    DIDT_TD_WEIGHT0_3__WEIGHT1__SHIFT,                  0x0010,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_WEIGHT0_3,               DIDT_TD_WEIGHT0_3__WEIGHT2_MASK,                    DIDT_TD_WEIGHT0_3__WEIGHT2__SHIFT,                  0x0017,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_WEIGHT0_3,               DIDT_TD_WEIGHT0_3__WEIGHT3_MASK,                    DIDT_TD_WEIGHT0_3__WEIGHT3__SHIFT,                  0x002f,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TD_WEIGHT4_7,               DIDT_TD_WEIGHT4_7__WEIGHT4_MASK,                    DIDT_TD_WEIGHT4_7__WEIGHT4__SHIFT,                  0x0046,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_WEIGHT4_7,               DIDT_TD_WEIGHT4_7__WEIGHT5_MASK,                    DIDT_TD_WEIGHT4_7__WEIGHT5__SHIFT,                  0x005d,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_WEIGHT4_7,               DIDT_TD_WEIGHT4_7__WEIGHT6_MASK,                    DIDT_TD_WEIGHT4_7__WEIGHT6__SHIFT,                  0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_WEIGHT4_7,               DIDT_TD_WEIGHT4_7__WEIGHT7_MASK,                    DIDT_TD_WEIGHT4_7__WEIGHT7__SHIFT,                  0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TD_CTRL1,                   DIDT_TD_CTRL1__MIN_POWER_MASK,                      DIDT_TD_CTRL1__MIN_POWER__SHIFT,                    0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL1,                   DIDT_TD_CTRL1__MAX_POWER_MASK,                      DIDT_TD_CTRL1__MAX_POWER__SHIFT,                    0xffff,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TD_CTRL_OCP,                DIDT_TD_CTRL_OCP__UNUSED_0_MASK,                    DIDT_TD_CTRL_OCP__UNUSED_0__SHIFT,                  0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL_OCP,                DIDT_TD_CTRL_OCP__OCP_MAX_POWER_MASK,               DIDT_TD_CTRL_OCP__OCP_MAX_POWER__SHIFT,             0x00ff,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TD_CTRL2,                   DIDT_TD_CTRL2__MAX_POWER_DELTA_MASK,                DIDT_TD_CTRL2__MAX_POWER_DELTA__SHIFT,              0x3fff,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL2,                   DIDT_TD_CTRL2__UNUSED_0_MASK,                       DIDT_TD_CTRL2__UNUSED_0__SHIFT,                     0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL2,                   DIDT_TD_CTRL2__SHORT_TERM_INTERVAL_SIZE_MASK,       DIDT_TD_CTRL2__SHORT_TERM_INTERVAL_SIZE__SHIFT,     0x000f,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL2,                   DIDT_TD_CTRL2__UNUSED_1_MASK,                       DIDT_TD_CTRL2__UNUSED_1__SHIFT,                     0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL2,                   DIDT_TD_CTRL2__LONG_TERM_INTERVAL_RATIO_MASK,       DIDT_TD_CTRL2__LONG_TERM_INTERVAL_RATIO__SHIFT,     0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL2,                   DIDT_TD_CTRL2__UNUSED_2_MASK,                       DIDT_TD_CTRL2__UNUSED_2__SHIFT,                     0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TD_STALL_CTRL,              DIDT_TD_STALL_CTRL__DIDT_STALL_CTRL_ENABLE_MASK,    DIDT_TD_STALL_CTRL__DIDT_STALL_CTRL_ENABLE__SHIFT,  0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_STALL_CTRL,              DIDT_TD_STALL_CTRL__DIDT_STALL_DELAY_HI_MASK,       DIDT_TD_STALL_CTRL__DIDT_STALL_DELAY_HI__SHIFT,     0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_STALL_CTRL,              DIDT_TD_STALL_CTRL__DIDT_STALL_DELAY_LO_MASK,       DIDT_TD_STALL_CTRL__DIDT_STALL_DELAY_LO__SHIFT,     0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_STALL_CTRL,              DIDT_TD_STALL_CTRL__DIDT_HI_POWER_THRESHOLD_MASK,   DIDT_TD_STALL_CTRL__DIDT_HI_POWER_THRESHOLD__SHIFT, 0x01aa,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_STALL_CTRL,              DIDT_TD_STALL_CTRL__UNUSED_0_MASK,                  DIDT_TD_STALL_CTRL__UNUSED_0__SHIFT,                0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TD_TUNING_CTRL,             DIDT_TD_TUNING_CTRL__DIDT_TUNING_ENABLE_MASK,       DIDT_TD_TUNING_CTRL__DIDT_TUNING_ENABLE__SHIFT,     0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_TUNING_CTRL,             DIDT_TD_TUNING_CTRL__MAX_POWER_DELTA_HI_MASK,       DIDT_TD_TUNING_CTRL__MAX_POWER_DELTA_HI__SHIFT,     0x0dde,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_TUNING_CTRL,             DIDT_TD_TUNING_CTRL__MAX_POWER_DELTA_LO_MASK,       DIDT_TD_TUNING_CTRL__MAX_POWER_DELTA_LO__SHIFT,     0x0dde,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_TUNING_CTRL,             DIDT_TD_TUNING_CTRL__UNUSED_0_MASK,                 DIDT_TD_TUNING_CTRL__UNUSED_0__SHIFT,               0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TD_CTRL0,                   DIDT_TD_CTRL0__DIDT_CTRL_EN_MASK,                   DIDT_TD_CTRL0__DIDT_CTRL_EN__SHIFT,                 0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL0,                   DIDT_TD_CTRL0__USE_REF_CLOCK_MASK,                  DIDT_TD_CTRL0__USE_REF_CLOCK__SHIFT,                0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL0,                   DIDT_TD_CTRL0__PHASE_OFFSET_MASK,                   DIDT_TD_CTRL0__PHASE_OFFSET__SHIFT,                 0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL0,                   DIDT_TD_CTRL0__DIDT_CTRL_RST_MASK,                  DIDT_TD_CTRL0__DIDT_CTRL_RST__SHIFT,                0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL0,                   DIDT_TD_CTRL0__DIDT_CLK_EN_OVERRIDE_MASK,           DIDT_TD_CTRL0__DIDT_CLK_EN_OVERRIDE__SHIFT,         0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL0,                   DIDT_TD_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI_MASK,     DIDT_TD_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI__SHIFT,   0x0009,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL0,                   DIDT_TD_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO_MASK,     DIDT_TD_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO__SHIFT,   0x0009,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL0,                   DIDT_TD_CTRL0__UNUSED_0_MASK,                       DIDT_TD_CTRL0__UNUSED_0__SHIFT,                     0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    // DIDT_TCP
+    //
+    {   ixDIDT_TCP_WEIGHT0_3,              DIDT_TCP_WEIGHT0_3__WEIGHT0_MASK,                   DIDT_TCP_WEIGHT0_3__WEIGHT0__SHIFT,                 0x0004,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_WEIGHT0_3,              DIDT_TCP_WEIGHT0_3__WEIGHT1_MASK,                   DIDT_TCP_WEIGHT0_3__WEIGHT1__SHIFT,                 0x0037,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_WEIGHT0_3,              DIDT_TCP_WEIGHT0_3__WEIGHT2_MASK,                   DIDT_TCP_WEIGHT0_3__WEIGHT2__SHIFT,                 0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_WEIGHT0_3,              DIDT_TCP_WEIGHT0_3__WEIGHT3_MASK,                   DIDT_TCP_WEIGHT0_3__WEIGHT3__SHIFT,                 0x00ff,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TCP_WEIGHT4_7,              DIDT_TCP_WEIGHT4_7__WEIGHT4_MASK,                   DIDT_TCP_WEIGHT4_7__WEIGHT4__SHIFT,                 0x0054,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_WEIGHT4_7,              DIDT_TCP_WEIGHT4_7__WEIGHT5_MASK,                   DIDT_TCP_WEIGHT4_7__WEIGHT5__SHIFT,                 0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_WEIGHT4_7,              DIDT_TCP_WEIGHT4_7__WEIGHT6_MASK,                   DIDT_TCP_WEIGHT4_7__WEIGHT6__SHIFT,                 0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_WEIGHT4_7,              DIDT_TCP_WEIGHT4_7__WEIGHT7_MASK,                   DIDT_TCP_WEIGHT4_7__WEIGHT7__SHIFT,                 0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TCP_CTRL1,                  DIDT_TCP_CTRL1__MIN_POWER_MASK,                     DIDT_TCP_CTRL1__MIN_POWER__SHIFT,                   0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL1,                  DIDT_TCP_CTRL1__MAX_POWER_MASK,                     DIDT_TCP_CTRL1__MAX_POWER__SHIFT,                   0xffff,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TCP_CTRL_OCP,               DIDT_TCP_CTRL_OCP__UNUSED_0_MASK,                   DIDT_TCP_CTRL_OCP__UNUSED_0__SHIFT,                 0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL_OCP,               DIDT_TCP_CTRL_OCP__OCP_MAX_POWER_MASK,              DIDT_TCP_CTRL_OCP__OCP_MAX_POWER__SHIFT,            0xffff,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TCP_CTRL2,                  DIDT_TCP_CTRL2__MAX_POWER_DELTA_MASK,               DIDT_TCP_CTRL2__MAX_POWER_DELTA__SHIFT,             0x3dde,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL2,                  DIDT_TCP_CTRL2__UNUSED_0_MASK,                      DIDT_TCP_CTRL2__UNUSED_0__SHIFT,                    0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL2,                  DIDT_TCP_CTRL2__SHORT_TERM_INTERVAL_SIZE_MASK,      DIDT_TCP_CTRL2__SHORT_TERM_INTERVAL_SIZE__SHIFT,    0x0032,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL2,                  DIDT_TCP_CTRL2__UNUSED_1_MASK,                      DIDT_TCP_CTRL2__UNUSED_1__SHIFT,                    0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL2,                  DIDT_TCP_CTRL2__LONG_TERM_INTERVAL_RATIO_MASK,      DIDT_TCP_CTRL2__LONG_TERM_INTERVAL_RATIO__SHIFT,    0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL2,                  DIDT_TCP_CTRL2__UNUSED_2_MASK,                      DIDT_TCP_CTRL2__UNUSED_2__SHIFT,                    0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TCP_STALL_CTRL,             DIDT_TCP_STALL_CTRL__DIDT_STALL_CTRL_ENABLE_MASK,   DIDT_TCP_STALL_CTRL__DIDT_STALL_CTRL_ENABLE__SHIFT, 0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_STALL_CTRL,             DIDT_TCP_STALL_CTRL__DIDT_STALL_DELAY_HI_MASK,      DIDT_TCP_STALL_CTRL__DIDT_STALL_DELAY_HI__SHIFT,    0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_STALL_CTRL,             DIDT_TCP_STALL_CTRL__DIDT_STALL_DELAY_LO_MASK,      DIDT_TCP_STALL_CTRL__DIDT_STALL_DELAY_LO__SHIFT,    0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_STALL_CTRL,             DIDT_TCP_STALL_CTRL__DIDT_HI_POWER_THRESHOLD_MASK,  DIDT_TCP_STALL_CTRL__DIDT_HI_POWER_THRESHOLD__SHIFT,0x01aa,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_STALL_CTRL,             DIDT_TCP_STALL_CTRL__UNUSED_0_MASK,                 DIDT_TCP_STALL_CTRL__UNUSED_0__SHIFT,               0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TCP_TUNING_CTRL,            DIDT_TCP_TUNING_CTRL__DIDT_TUNING_ENABLE_MASK,      DIDT_TCP_TUNING_CTRL__DIDT_TUNING_ENABLE__SHIFT,    0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_TUNING_CTRL,            DIDT_TCP_TUNING_CTRL__MAX_POWER_DELTA_HI_MASK,      DIDT_TCP_TUNING_CTRL__MAX_POWER_DELTA_HI__SHIFT,    0x3dde,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_TUNING_CTRL,            DIDT_TCP_TUNING_CTRL__MAX_POWER_DELTA_LO_MASK,      DIDT_TCP_TUNING_CTRL__MAX_POWER_DELTA_LO__SHIFT,    0x3dde,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_TUNING_CTRL,            DIDT_TCP_TUNING_CTRL__UNUSED_0_MASK,                DIDT_TCP_TUNING_CTRL__UNUSED_0__SHIFT,              0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TCP_CTRL0,                   DIDT_TCP_CTRL0__DIDT_CTRL_EN_MASK,                   DIDT_TCP_CTRL0__DIDT_CTRL_EN__SHIFT,                 0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL0,                   DIDT_TCP_CTRL0__USE_REF_CLOCK_MASK,                  DIDT_TCP_CTRL0__USE_REF_CLOCK__SHIFT,                0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL0,                   DIDT_TCP_CTRL0__PHASE_OFFSET_MASK,                   DIDT_TCP_CTRL0__PHASE_OFFSET__SHIFT,                 0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL0,                   DIDT_TCP_CTRL0__DIDT_CTRL_RST_MASK,                  DIDT_TCP_CTRL0__DIDT_CTRL_RST__SHIFT,                0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL0,                   DIDT_TCP_CTRL0__DIDT_CLK_EN_OVERRIDE_MASK,           DIDT_TCP_CTRL0__DIDT_CLK_EN_OVERRIDE__SHIFT,         0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL0,                   DIDT_TCP_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI_MASK,     DIDT_TCP_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI__SHIFT,   0x0010,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL0,                   DIDT_TCP_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO_MASK,     DIDT_TCP_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO__SHIFT,   0x0010,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL0,                   DIDT_TCP_CTRL0__UNUSED_0_MASK,                       DIDT_TCP_CTRL0__UNUSED_0__SHIFT,                     0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   0xFFFFFFFF  }  // End of list
+};
 static int smu7_enable_didt(struct pp_hwmgr *hwmgr, const bool enable)
 {
 	uint32_t en = enable ? 1 : 0;
@@ -768,6 +952,11 @@ int smu7_enable_didt_config(struct pp_hwmgr *hwmgr)
 				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", return result);
 				result = smu7_program_pt_config_registers(hwmgr, DIDTConfig_Polaris12);
 				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", return result);
+			} else if (hwmgr->chip_id == CHIP_VEGAM) {
+				result = smu7_program_pt_config_registers(hwmgr, GCCACConfig_VegaM);
+				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", return result);
+				result = smu7_program_pt_config_registers(hwmgr, DIDTConfig_VegaM);
+				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", return result);
 			}
 		}
 		cgs_write_register(hwmgr->device, mmGRBM_GFX_INDEX, value2);
-- 
cgit v1.1


From 5dd3f9efd4199f0d9e8244322934494ebd140dfd Mon Sep 17 00:00:00 2001
From: Pixel Ding <Pixel.Ding@amd.com>
Date: Tue, 24 Apr 2018 22:52:45 -0400
Subject: drm/scheduler: don't update last scheduled fence in TDR

The current sequence in scheduler thread is:
1. update last sched fence
2. job begin (adding to mirror list)
3. job finish (remove from mirror list)
4. back to 1

Since we update last sched prior to joining mirror list, the jobs
in mirror list already pass the last sched fence. TDR just run
the jobs in mirror list, so we should not update the last sched
fences in TDR.

Signed-off-by: Pixel Ding <Pixel.Ding@amd.com>
Reviewed-by: Monk Liu <monk.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/scheduler/gpu_scheduler.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/drivers/gpu/drm/scheduler/gpu_scheduler.c b/drivers/gpu/drm/scheduler/gpu_scheduler.c
index 088ff2b..1f1dd70 100644
--- a/drivers/gpu/drm/scheduler/gpu_scheduler.c
+++ b/drivers/gpu/drm/scheduler/gpu_scheduler.c
@@ -575,9 +575,6 @@ void drm_sched_job_recovery(struct drm_gpu_scheduler *sched)
 		fence = sched->ops->run_job(s_job);
 		atomic_inc(&sched->hw_rq_count);
 
-		dma_fence_put(s_job->entity->last_scheduled);
-		s_job->entity->last_scheduled = dma_fence_get(&s_fence->finished);
-
 		if (fence) {
 			s_fence->parent = dma_fence_get(fence);
 			r = dma_fence_add_callback(fence, &s_fence->cb,
-- 
cgit v1.1


From abc342538cc4670a107e45037c39d1dda8174563 Mon Sep 17 00:00:00 2001
From: Emily Deng <Emily.Deng@amd.com>
Date: Thu, 26 Apr 2018 18:02:55 +0800
Subject: drm/amdgpu: For sriov reset, move IB test into exclusive mode

When put the IB test out of exclusive mode, and do sriov reset,
the IB test will randomly fail. As out of exclusive mode it uses
kiq to do read and write registers, but as it has world switch,
the kiq read and write time will be random, sometimes it will
beyond the MAX_KIQ_REG_WAIT and then the read or write register
will fail, which will result the IB test fail.

Signed-off-by: Emily Deng <Emily.Deng@amd.com>
Reviewed-by: Monk Liu <monk.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index e6657ec..9fb20a5 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -3143,20 +3143,19 @@ static int amdgpu_device_reset_sriov(struct amdgpu_device *adev,
 
 	/* now we are okay to resume SMC/CP/SDMA */
 	r = amdgpu_device_ip_reinit_late_sriov(adev);
-	amdgpu_virt_release_full_gpu(adev, true);
 	if (r)
 		goto error;
 
 	amdgpu_irq_gpu_reset_resume_helper(adev);
 	r = amdgpu_ib_ring_tests(adev);
 
+error:
+	amdgpu_virt_release_full_gpu(adev, true);
 	if (!r && adev->virt.gim_feature & AMDGIM_FEATURE_GIM_FLR_VRAMLOST) {
 		atomic_inc(&adev->vram_lost_counter);
 		r = amdgpu_device_handle_vram_lost(adev);
 	}
 
-error:
-
 	return r;
 }
 
-- 
cgit v1.1


From 6e9c2b88eb42bdda6ba1f1a39238c446782d443e Mon Sep 17 00:00:00 2001
From: Emily Deng <Emily.Deng@amd.com>
Date: Thu, 26 Apr 2018 18:02:14 +0800
Subject: drm/amdgpu/sriov: Need to set in_gpu_reset flag to back after gpu
 reset

After host os reset gpu reset, need to set flag in_gpu_reset to
zero.

Signed-off-by: Emily Deng <Emily.Deng@amd.com>
Reviewed-by: Monk Liu <monk.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c
index 4933486..078f70f 100644
--- a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c
+++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c
@@ -260,8 +260,10 @@ static void xgpu_ai_mailbox_flr_work(struct work_struct *work)
 	} while (timeout > 1);
 
 flr_done:
-	if (locked)
+	if (locked) {
+		adev->in_gpu_reset = 0;
 		mutex_unlock(&adev->lock_reset);
+	}
 
 	/* Trigger recovery for world switch failure if no TDR */
 	if (amdgpu_lockup_timeout == 0)
-- 
cgit v1.1


From 40c21ed6b372c58fb214307f8186f7a0c1885bfc Mon Sep 17 00:00:00 2001
From: Mikita Lipski <mikita.lipski@amd.com>
Date: Wed, 10 Jan 2018 10:01:38 -0500
Subject: drm/amd/display: Fix deadlock when flushing irq

Lock irq table when reading a work in queue,
unlock to flush the work, lock again till all tasks
are cleared

Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c
index 490017d..4be21bf 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c
@@ -329,14 +329,15 @@ void amdgpu_dm_irq_fini(struct amdgpu_device *adev)
 {
 	int src;
 	struct irq_list_head *lh;
+	unsigned long irq_table_flags;
 	DRM_DEBUG_KMS("DM_IRQ: releasing resources.\n");
-
 	for (src = 0; src < DAL_IRQ_SOURCES_NUMBER; src++) {
-
+		DM_IRQ_TABLE_LOCK(adev, irq_table_flags);
 		/* The handler was removed from the table,
 		 * it means it is safe to flush all the 'work'
 		 * (because no code can schedule a new one). */
 		lh = &adev->dm.irq_handler_list_low_tab[src];
+		DM_IRQ_TABLE_UNLOCK(adev, irq_table_flags);
 		flush_work(&lh->work);
 	}
 }
-- 
cgit v1.1


From a80aa93de1a0e69fdb83e04a9aca7c33bfb18941 Mon Sep 17 00:00:00 2001
From: Mikita Lipski <mikita.lipski@amd.com>
Date: Sat, 3 Feb 2018 14:18:07 -0500
Subject: drm/amd/display: Unify dm resume sequence into a single call

Merge amdgpu_dm_display_resume function into dm_resume,
as it is not called anywhere else anymore.

Initially the call was broken down into 2 functions for cursor corruption
issue. Now the issue is not visible, hence the dm_resume will perform
dm_display_resume in it.

Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 31 ++++++++---------------
 1 file changed, 10 insertions(+), 21 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 8379a37..cc105f1 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -644,18 +644,6 @@ amdgpu_dm_find_first_crtc_matching_connector(struct drm_atomic_state *state,
 static int dm_resume(void *handle)
 {
 	struct amdgpu_device *adev = handle;
-	struct amdgpu_display_manager *dm = &adev->dm;
-	int ret = 0;
-
-	/* power on hardware */
-	dc_set_power_state(dm->dc, DC_ACPI_CM_POWER_STATE_D0);
-
-	ret = amdgpu_dm_display_resume(adev);
-	return ret;
-}
-
-int amdgpu_dm_display_resume(struct amdgpu_device *adev)
-{
 	struct drm_device *ddev = adev->ddev;
 	struct amdgpu_display_manager *dm = &adev->dm;
 	struct amdgpu_dm_connector *aconnector;
@@ -666,10 +654,12 @@ int amdgpu_dm_display_resume(struct amdgpu_device *adev)
 	struct drm_plane *plane;
 	struct drm_plane_state *new_plane_state;
 	struct dm_plane_state *dm_new_plane_state;
-
-	int ret = 0;
+	int ret;
 	int i;
 
+	/* power on hardware */
+	dc_set_power_state(dm->dc, DC_ACPI_CM_POWER_STATE_D0);
+
 	/* program HPD filter */
 	dc_resume(dm->dc);
 
@@ -683,8 +673,7 @@ int amdgpu_dm_display_resume(struct amdgpu_device *adev)
 	amdgpu_dm_irq_resume_early(adev);
 
 	/* Do detection*/
-	list_for_each_entry(connector,
-			&ddev->mode_config.connector_list, head) {
+	list_for_each_entry(connector, &ddev->mode_config.connector_list, head) {
 		aconnector = to_amdgpu_dm_connector(connector);
 
 		/*
@@ -706,7 +695,7 @@ int amdgpu_dm_display_resume(struct amdgpu_device *adev)
 	}
 
 	/* Force mode set in atomic comit */
-	for_each_new_crtc_in_state(adev->dm.cached_state, crtc, new_crtc_state, i)
+	for_each_new_crtc_in_state(dm->cached_state, crtc, new_crtc_state, i)
 		new_crtc_state->active_changed = true;
 
 	/*
@@ -714,7 +703,7 @@ int amdgpu_dm_display_resume(struct amdgpu_device *adev)
 	 * them here, since they were duplicated as part of the suspend
 	 * procedure.
 	 */
-	for_each_new_crtc_in_state(adev->dm.cached_state, crtc, new_crtc_state, i) {
+	for_each_new_crtc_in_state(dm->cached_state, crtc, new_crtc_state, i) {
 		dm_new_crtc_state = to_dm_crtc_state(new_crtc_state);
 		if (dm_new_crtc_state->stream) {
 			WARN_ON(kref_read(&dm_new_crtc_state->stream->refcount) > 1);
@@ -723,7 +712,7 @@ int amdgpu_dm_display_resume(struct amdgpu_device *adev)
 		}
 	}
 
-	for_each_new_plane_in_state(adev->dm.cached_state, plane, new_plane_state, i) {
+	for_each_new_plane_in_state(dm->cached_state, plane, new_plane_state, i) {
 		dm_new_plane_state = to_dm_plane_state(new_plane_state);
 		if (dm_new_plane_state->dc_state) {
 			WARN_ON(kref_read(&dm_new_plane_state->dc_state->refcount) > 1);
@@ -732,9 +721,9 @@ int amdgpu_dm_display_resume(struct amdgpu_device *adev)
 		}
 	}
 
-	ret = drm_atomic_helper_resume(ddev, adev->dm.cached_state);
+	ret = drm_atomic_helper_resume(ddev, dm->cached_state);
 
-	adev->dm.cached_state = NULL;
+	dm->cached_state = NULL;
 
 	amdgpu_dm_irq_resume_late(adev);
 
-- 
cgit v1.1


From 3d777c82bda9a82cc304de5ba51122d248f7a584 Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Mon, 16 Apr 2018 17:28:11 -0400
Subject: drm/amd/display: Disallow enabling CRTC without primary plane with FB
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The below commit

    "drm/atomic: Try to preserve the crtc enabled state in drm_atomic_remove_fb, v2"

introduces a slight behavioral change to rmfb. Instead of disabling a crtc
when the primary plane is disabled, it now preserves it.

Since DC is currently not equipped to handle this we need to fail such
a commit, otherwise we might see a corrupted screen.

This is based on Shirish's previous approach but avoids adding all
planes to the new atomic state which leads to a full update in DC for
any commit, and is not what we intend.

Theoretically DM should be able to deal with states with fully populated planes,
even for simple updates, such as cursor updates. This should still be
addressed in the future.

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Tested-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Cc: stable@vger.kernel.org
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index cc105f1..96a57be 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -4628,6 +4628,7 @@ static int dm_update_crtcs_state(struct dc *dc,
 		struct amdgpu_dm_connector *aconnector = NULL;
 		struct drm_connector_state *new_con_state = NULL;
 		struct dm_connector_state *dm_conn_state = NULL;
+		struct drm_plane_state *new_plane_state = NULL;
 
 		new_stream = NULL;
 
@@ -4635,6 +4636,13 @@ static int dm_update_crtcs_state(struct dc *dc,
 		dm_new_crtc_state = to_dm_crtc_state(new_crtc_state);
 		acrtc = to_amdgpu_crtc(crtc);
 
+		new_plane_state = drm_atomic_get_new_plane_state(state, new_crtc_state->crtc->primary);
+
+		if (new_crtc_state->enable && new_plane_state && !new_plane_state->fb) {
+			ret = -EINVAL;
+			goto fail;
+		}
+
 		aconnector = amdgpu_dm_find_first_crtc_matching_connector(state, crtc);
 
 		/* TODO This hack should go away */
@@ -4831,7 +4839,7 @@ static int dm_update_planes_state(struct dc *dc,
 			if (!dm_old_crtc_state->stream)
 				continue;
 
-			DRM_DEBUG_DRIVER("Disabling DRM plane: %d on DRM crtc %d\n",
+			DRM_DEBUG_ATOMIC("Disabling DRM plane: %d on DRM crtc %d\n",
 					plane->base.id, old_plane_crtc->base.id);
 
 			if (!dc_remove_plane_from_context(
-- 
cgit v1.1


From 388277b17023e69fdbf6ed68cec42fa7616ce454 Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Wed, 11 Apr 2018 13:19:56 -0400
Subject: drm/amd/display: fix issue related to infopacket was not transmitted

Check in code was incorrect, and infopacket is only transmitted after update
function is called multiple times.
Purpose of the function was to check if infopackets are being enabled, and
then enable global control. Fix the code to do this.

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../drm/amd/display/dc/dce/dce_stream_encoder.c    | 25 ++++++----------------
 .../amd/display/dc/dcn10/dcn10_stream_encoder.c    | 11 ++++++----
 2 files changed, 13 insertions(+), 23 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
index 84e26c8..e265a0a 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
@@ -819,7 +819,7 @@ static void dce110_stream_encoder_update_dp_info_packets(
 	const struct encoder_info_frame *info_frame)
 {
 	struct dce110_stream_encoder *enc110 = DCE110STRENC_FROM_STRENC(enc);
-	uint32_t value = REG_READ(DP_SEC_CNTL);
+	uint32_t value = 0;
 
 	if (info_frame->vsc.valid)
 		dce110_update_generic_info_packet(
@@ -853,6 +853,7 @@ static void dce110_stream_encoder_update_dp_info_packets(
 	* Therefore we need to enable master bit
 	* if at least on of the fields is not 0
 	*/
+	value = REG_READ(DP_SEC_CNTL);
 	if (value)
 		REG_UPDATE(DP_SEC_CNTL, DP_SEC_STREAM_ENABLE, 1);
 }
@@ -862,7 +863,7 @@ static void dce110_stream_encoder_stop_dp_info_packets(
 {
 	/* stop generic packets on DP */
 	struct dce110_stream_encoder *enc110 = DCE110STRENC_FROM_STRENC(enc);
-	uint32_t value = REG_READ(DP_SEC_CNTL);
+	uint32_t value = 0;
 
 	if (enc110->se_mask->DP_SEC_AVI_ENABLE) {
 		REG_SET_7(DP_SEC_CNTL, 0,
@@ -875,25 +876,10 @@ static void dce110_stream_encoder_stop_dp_info_packets(
 			DP_SEC_STREAM_ENABLE, 0);
 	}
 
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
-	if (enc110->se_mask->DP_SEC_GSP7_ENABLE) {
-		REG_SET_10(DP_SEC_CNTL, 0,
-			DP_SEC_GSP0_ENABLE, 0,
-			DP_SEC_GSP1_ENABLE, 0,
-			DP_SEC_GSP2_ENABLE, 0,
-			DP_SEC_GSP3_ENABLE, 0,
-			DP_SEC_GSP4_ENABLE, 0,
-			DP_SEC_GSP5_ENABLE, 0,
-			DP_SEC_GSP6_ENABLE, 0,
-			DP_SEC_GSP7_ENABLE, 0,
-			DP_SEC_MPG_ENABLE, 0,
-			DP_SEC_STREAM_ENABLE, 0);
-	}
-#endif
 	/* this register shared with audio info frame.
 	 * therefore we need to keep master enabled
 	 * if at least one of the fields is not 0 */
-
+	value = REG_READ(DP_SEC_CNTL);
 	if (value)
 		REG_UPDATE(DP_SEC_CNTL, DP_SEC_STREAM_ENABLE, 1);
 
@@ -1496,7 +1482,7 @@ static void dce110_se_disable_dp_audio(
 	struct stream_encoder *enc)
 {
 	struct dce110_stream_encoder *enc110 = DCE110STRENC_FROM_STRENC(enc);
-	uint32_t value = REG_READ(DP_SEC_CNTL);
+	uint32_t value = 0;
 
 	/* Disable Audio packets */
 	REG_UPDATE_5(DP_SEC_CNTL,
@@ -1508,6 +1494,7 @@ static void dce110_se_disable_dp_audio(
 
 	/* This register shared with encoder info frame. Therefore we need to
 	keep master enabled if at least on of the fields is not 0 */
+	value = REG_READ(DP_SEC_CNTL);
 	if (value != 0)
 		REG_UPDATE(DP_SEC_CNTL, DP_SEC_STREAM_ENABLE, 1);
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c
index 9ec46f8..befd863 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c
@@ -686,7 +686,7 @@ void enc1_stream_encoder_update_dp_info_packets(
 	const struct encoder_info_frame *info_frame)
 {
 	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
-	uint32_t value = REG_READ(DP_SEC_CNTL);
+	uint32_t value = 0;
 
 	if (info_frame->vsc.valid)
 		enc1_update_generic_info_packet(
@@ -713,6 +713,7 @@ void enc1_stream_encoder_update_dp_info_packets(
 	REG_UPDATE(DP_SEC_CNTL, DP_SEC_GSP2_ENABLE, info_frame->spd.valid);
 	REG_UPDATE(DP_SEC_CNTL, DP_SEC_GSP3_ENABLE, info_frame->hdrsmd.valid);
 
+
 	/* This bit is the master enable bit.
 	 * When enabling secondary stream engine,
 	 * this master bit must also be set.
@@ -720,6 +721,7 @@ void enc1_stream_encoder_update_dp_info_packets(
 	 * Therefore we need to enable master bit
 	 * if at least on of the fields is not 0
 	 */
+	value = REG_READ(DP_SEC_CNTL);
 	if (value)
 		REG_UPDATE(DP_SEC_CNTL, DP_SEC_STREAM_ENABLE, 1);
 }
@@ -729,7 +731,7 @@ void enc1_stream_encoder_stop_dp_info_packets(
 {
 	/* stop generic packets on DP */
 	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
-	uint32_t value = REG_READ(DP_SEC_CNTL);
+	uint32_t value = 0;
 
 	REG_SET_10(DP_SEC_CNTL, 0,
 		DP_SEC_GSP0_ENABLE, 0,
@@ -746,7 +748,7 @@ void enc1_stream_encoder_stop_dp_info_packets(
 	/* this register shared with audio info frame.
 	 * therefore we need to keep master enabled
 	 * if at least one of the fields is not 0 */
-
+	value = REG_READ(DP_SEC_CNTL);
 	if (value)
 		REG_UPDATE(DP_SEC_CNTL, DP_SEC_STREAM_ENABLE, 1);
 
@@ -1356,7 +1358,7 @@ static void enc1_se_disable_dp_audio(
 	struct stream_encoder *enc)
 {
 	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
-	uint32_t value = REG_READ(DP_SEC_CNTL);
+	uint32_t value = 0;
 
 	/* Disable Audio packets */
 	REG_UPDATE_5(DP_SEC_CNTL,
@@ -1369,6 +1371,7 @@ static void enc1_se_disable_dp_audio(
 	/* This register shared with encoder info frame. Therefore we need to
 	 * keep master enabled if at least on of the fields is not 0
 	 */
+	value = REG_READ(DP_SEC_CNTL);
 	if (value != 0)
 		REG_UPDATE(DP_SEC_CNTL, DP_SEC_STREAM_ENABLE, 1);
 
-- 
cgit v1.1


From 8e357610ca0cc44a875df68f608a756fa56b1797 Mon Sep 17 00:00:00 2001
From: Eric Bernstein <eric.bernstein@amd.com>
Date: Mon, 9 Apr 2018 15:47:42 -0400
Subject: drm/amd/display: Make program_output_csc HWSS interface function

Signed-off-by: Eric Bernstein <eric.bernstein@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c | 5 +++--
 drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h         | 7 +++++++
 2 files changed, 10 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
index e547f46..7cecab0 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -1564,7 +1564,7 @@ static void program_csc_matrix(struct pipe_ctx *pipe_ctx,
 	}
 }
 
-static void program_output_csc(struct dc *dc,
+static void dcn10_program_output_csc(struct dc *dc,
 		struct pipe_ctx *pipe_ctx,
 		enum dc_color_space colorspace,
 		uint16_t *matrix,
@@ -1917,7 +1917,7 @@ static void update_dchubp_dpp(
 		/*gamut remap*/
 		program_gamut_remap(pipe_ctx);
 
-		program_output_csc(dc,
+		dc->hwss.program_output_csc(dc,
 				pipe_ctx,
 				pipe_ctx->stream->output_color_space,
 				pipe_ctx->stream->csc_color_matrix.matrix,
@@ -2667,6 +2667,7 @@ static const struct hw_sequencer_funcs dcn10_funcs = {
 	.update_pending_status = dcn10_update_pending_status,
 	.set_input_transfer_func = dcn10_set_input_transfer_func,
 	.set_output_transfer_func = dcn10_set_output_transfer_func,
+	.program_output_csc = dcn10_program_output_csc,
 	.power_down = dce110_power_down,
 	.enable_accelerated_mode = dce110_enable_accelerated_mode,
 	.enable_timing_synchronization = dcn10_enable_timing_synchronization,
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
index f54d478..be6cf7e 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
@@ -95,6 +95,12 @@ struct hw_sequencer_funcs {
 			enum dc_color_space colorspace,
 			uint16_t *matrix);
 
+	void (*program_output_csc)(struct dc *dc,
+			struct pipe_ctx *pipe_ctx,
+			enum dc_color_space colorspace,
+			uint16_t *matrix,
+			int opp_id);
+
 	void (*update_plane_addr)(
 		const struct dc *dc,
 		struct pipe_ctx *pipe_ctx);
@@ -203,6 +209,7 @@ struct hw_sequencer_funcs {
 
 	void (*set_cursor_position)(struct pipe_ctx *pipe);
 	void (*set_cursor_attribute)(struct pipe_ctx *pipe);
+
 };
 
 void color_space_to_black_color(
-- 
cgit v1.1


From 3158223efde597521505b586a88a6d43c8f2324f Mon Sep 17 00:00:00 2001
From: Eric Bernstein <eric.bernstein@amd.com>
Date: Mon, 9 Apr 2018 17:19:27 -0400
Subject: drm/amd/display: Refactor otg_blank sequence

Also rename otg_blank to blank_pixel_data.

Signed-off-by: Eric Bernstein <eric.bernstein@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../amd/display/dc/dce110/dce110_hw_sequencer.c    |  6 ++--
 .../drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c  | 35 +++++++++++-----------
 drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h  |  8 ++++-
 3 files changed, 28 insertions(+), 21 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
index e70ccb9..51c6c70 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
@@ -1233,7 +1233,7 @@ static void program_scaler(const struct dc *dc,
 		&pipe_ctx->plane_res.scl_data);
 }
 
-static enum dc_status dce110_prog_pixclk_crtc_otg(
+static enum dc_status dce110_enable_stream_timing(
 		struct pipe_ctx *pipe_ctx,
 		struct dc_state *context,
 		struct dc *dc)
@@ -1299,7 +1299,7 @@ static enum dc_status apply_single_controller_ctx_to_hw(
 			pipe_ctx[pipe_ctx->pipe_idx];
 
 	/*  */
-	dc->hwss.prog_pixclk_crtc_otg(pipe_ctx, context, dc);
+	dc->hwss.enable_stream_timing(pipe_ctx, context, dc);
 
 	/* FPGA does not program backend */
 	if (IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) {
@@ -3041,7 +3041,7 @@ static const struct hw_sequencer_funcs dce110_funcs = {
 	.get_position = get_position,
 	.set_static_screen_control = set_static_screen_control,
 	.reset_hw_ctx_wrap = dce110_reset_hw_ctx_wrap,
-	.prog_pixclk_crtc_otg = dce110_prog_pixclk_crtc_otg,
+	.enable_stream_timing = dce110_enable_stream_timing,
 	.setup_stereo = NULL,
 	.set_avmute = dce110_set_avmute,
 	.wait_for_mpcc_disconnect = dce110_wait_for_mpcc_disconnect,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
index 7cecab0..8eea38b 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -593,7 +593,7 @@ static void false_optc_underflow_wa(
 		tg->funcs->clear_optc_underflow(tg);
 }
 
-static enum dc_status dcn10_prog_pixclk_crtc_otg(
+static enum dc_status dcn10_enable_stream_timing(
 		struct pipe_ctx *pipe_ctx,
 		struct dc_state *context,
 		struct dc *dc)
@@ -1950,9 +1950,9 @@ static void update_dchubp_dpp(
 		hubp->funcs->set_blank(hubp, false);
 }
 
-static void dcn10_otg_blank(
+static void dcn10_blank_pixel_data(
 		struct dc *dc,
-		struct stream_resource stream_res,
+		struct stream_resource *stream_res,
 		struct dc_stream_state *stream,
 		bool blank)
 {
@@ -1963,21 +1963,21 @@ static void dcn10_otg_blank(
 	color_space = stream->output_color_space;
 	color_space_to_black_color(dc, color_space, &black_color);
 
-	if (stream_res.tg->funcs->set_blank_color)
-		stream_res.tg->funcs->set_blank_color(
-				stream_res.tg,
+	if (stream_res->tg->funcs->set_blank_color)
+		stream_res->tg->funcs->set_blank_color(
+				stream_res->tg,
 				&black_color);
 
 	if (!blank) {
-		if (stream_res.tg->funcs->set_blank)
-			stream_res.tg->funcs->set_blank(stream_res.tg, blank);
-		if (stream_res.abm)
-			stream_res.abm->funcs->set_abm_level(stream_res.abm, stream->abm_level);
+		if (stream_res->tg->funcs->set_blank)
+			stream_res->tg->funcs->set_blank(stream_res->tg, blank);
+		if (stream_res->abm)
+			stream_res->abm->funcs->set_abm_level(stream_res->abm, stream->abm_level);
 	} else if (blank) {
-		if (stream_res.abm)
-			stream_res.abm->funcs->set_abm_immediate_disable(stream_res.abm);
-		if (stream_res.tg->funcs->set_blank)
-			stream_res.tg->funcs->set_blank(stream_res.tg, blank);
+		if (stream_res->abm)
+			stream_res->abm->funcs->set_abm_immediate_disable(stream_res->abm);
+		if (stream_res->tg->funcs->set_blank)
+			stream_res->tg->funcs->set_blank(stream_res->tg, blank);
 	}
 }
 
@@ -2016,7 +2016,7 @@ static void program_all_pipe_in_tree(
 		pipe_ctx->stream_res.tg->funcs->program_global_sync(
 				pipe_ctx->stream_res.tg);
 
-		dcn10_otg_blank(dc, pipe_ctx->stream_res,
+		dc->hwss.blank_pixel_data(dc, &pipe_ctx->stream_res,
 				pipe_ctx->stream, blank);
 	}
 
@@ -2136,7 +2136,7 @@ static void dcn10_apply_ctx_for_surface(
 
 	if (num_planes == 0) {
 		/* OTG blank before remove all front end */
-		dcn10_otg_blank(dc, top_pipe_to_program->stream_res, top_pipe_to_program->stream, true);
+		dc->hwss.blank_pixel_data(dc, &top_pipe_to_program->stream_res, top_pipe_to_program->stream, true);
 	}
 
 	/* Disconnect unused mpcc */
@@ -2679,10 +2679,11 @@ static const struct hw_sequencer_funcs dcn10_funcs = {
 	.blank_stream = dce110_blank_stream,
 	.enable_display_power_gating = dcn10_dummy_display_power_gating,
 	.disable_plane = dcn10_disable_plane,
+	.blank_pixel_data = dcn10_blank_pixel_data,
 	.pipe_control_lock = dcn10_pipe_control_lock,
 	.set_bandwidth = dcn10_set_bandwidth,
 	.reset_hw_ctx_wrap = reset_hw_ctx_wrap,
-	.prog_pixclk_crtc_otg = dcn10_prog_pixclk_crtc_otg,
+	.enable_stream_timing = dcn10_enable_stream_timing,
 	.set_drr = set_drr,
 	.get_position = get_position,
 	.set_static_screen_control = set_static_screen_control,
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
index be6cf7e..29abf3e 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
@@ -65,6 +65,7 @@ struct dchub_init_data;
 struct dc_static_screen_events;
 struct resource_pool;
 struct resource_context;
+struct stream_resource;
 
 struct hw_sequencer_funcs {
 
@@ -162,6 +163,11 @@ struct hw_sequencer_funcs {
 				struct dc *dc,
 				struct pipe_ctx *pipe,
 				bool lock);
+	void (*blank_pixel_data)(
+			struct dc *dc,
+			struct stream_resource *stream_res,
+			struct dc_stream_state *stream,
+			bool blank);
 
 	void (*set_bandwidth)(
 			struct dc *dc,
@@ -177,7 +183,7 @@ struct hw_sequencer_funcs {
 	void (*set_static_screen_control)(struct pipe_ctx **pipe_ctx,
 			int num_pipes, const struct dc_static_screen_events *events);
 
-	enum dc_status (*prog_pixclk_crtc_otg)(
+	enum dc_status (*enable_stream_timing)(
 			struct pipe_ctx *pipe_ctx,
 			struct dc_state *context,
 			struct dc *dc);
-- 
cgit v1.1


From 50834eb488a30026de040ab5d209ca9f980ae14b Mon Sep 17 00:00:00 2001
From: Hersen Wu <hersenxs.wu@amd.com>
Date: Wed, 11 Apr 2018 15:22:10 -0400
Subject: drm/amd/display: DP link validation bug for YUV422

remove limit YUV422 color depth to 24bits which is
workaround for old ASIC

Signed-off-by: Hersen Wu <hersenxs.wu@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c | 47 +++++++++++-------------
 1 file changed, 21 insertions(+), 26 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
index 0a190c2..7d609c7 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
@@ -1379,34 +1379,29 @@ static uint32_t bandwidth_in_kbps_from_timing(
 	uint32_t bits_per_channel = 0;
 	uint32_t kbps;
 
-	if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR422)
+	switch (timing->display_color_depth) {
+	case COLOR_DEPTH_666:
+		bits_per_channel = 6;
+		break;
+	case COLOR_DEPTH_888:
+		bits_per_channel = 8;
+		break;
+	case COLOR_DEPTH_101010:
+		bits_per_channel = 10;
+		break;
+	case COLOR_DEPTH_121212:
 		bits_per_channel = 12;
-	else{
-
-		switch (timing->display_color_depth) {
-
-		case COLOR_DEPTH_666:
-			bits_per_channel = 6;
-			break;
-		case COLOR_DEPTH_888:
-			bits_per_channel = 8;
-			break;
-		case COLOR_DEPTH_101010:
-			bits_per_channel = 10;
-			break;
-		case COLOR_DEPTH_121212:
-			bits_per_channel = 12;
-			break;
-		case COLOR_DEPTH_141414:
-			bits_per_channel = 14;
-			break;
-		case COLOR_DEPTH_161616:
-			bits_per_channel = 16;
-			break;
-		default:
-			break;
-		}
+		break;
+	case COLOR_DEPTH_141414:
+		bits_per_channel = 14;
+		break;
+	case COLOR_DEPTH_161616:
+		bits_per_channel = 16;
+		break;
+	default:
+		break;
 	}
+
 	ASSERT(bits_per_channel != 0);
 
 	kbps = timing->pix_clk_khz;
-- 
cgit v1.1


From 8a79593d77de17619e99c23495ac243759704b87 Mon Sep 17 00:00:00 2001
From: Tony Cheng <tony.cheng@amd.com>
Date: Mon, 16 Apr 2018 13:30:02 -0400
Subject: drm/amd/display: dal 3.1.43

Signed-off-by: Tony Cheng <tony.cheng@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dc.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index 7ac8a1b..9215298 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -38,7 +38,7 @@
 #include "inc/compressor.h"
 #include "dml/display_mode_lib.h"
 
-#define DC_VER "3.1.42"
+#define DC_VER "3.1.43"
 
 #define MAX_SURFACES 3
 #define MAX_STREAMS 6
-- 
cgit v1.1


From 55a01d4023ce7249eed361731b373c78e62b73e0 Mon Sep 17 00:00:00 2001
From: Krunoslav Kovac <Krunoslav.Kovac@amd.com>
Date: Fri, 13 Apr 2018 16:06:24 -0400
Subject: drm/amd/display: Add user_regamma to color module

Signed-off-by: Krunoslav Kovac <Krunoslav.Kovac@amd.com>
Reviewed-by: Anthony Koo <Anthony.Koo@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../drm/amd/display/modules/color/color_gamma.c    | 314 ++++++++++++++++++++-
 .../drm/amd/display/modules/color/color_gamma.h    |  48 +++-
 2 files changed, 348 insertions(+), 14 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c
index e7e374f..ad0ff50 100644
--- a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c
+++ b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c
@@ -185,14 +185,14 @@ struct dividers {
 
 static void build_coefficients(struct gamma_coefficients *coefficients, bool is_2_4)
 {
-		static const int32_t numerator01[] = { 31308, 180000};
-		static const int32_t numerator02[] = { 12920, 4500};
-		static const int32_t numerator03[] = { 55, 99};
-		static const int32_t numerator04[] = { 55, 99};
-		static const int32_t numerator05[] = { 2400, 2200};
+	static const int32_t numerator01[] = { 31308, 180000};
+	static const int32_t numerator02[] = { 12920, 4500};
+	static const int32_t numerator03[] = { 55, 99};
+	static const int32_t numerator04[] = { 55, 99};
+	static const int32_t numerator05[] = { 2400, 2200};
 
-		uint32_t i = 0;
-		uint32_t index = is_2_4 == true ? 0:1;
+	uint32_t i = 0;
+	uint32_t index = is_2_4 == true ? 0:1;
 
 	do {
 		coefficients->a0[i] = dal_fixed31_32_from_fraction(
@@ -691,7 +691,7 @@ static void build_degamma(struct pwl_float_data_ex *curve,
 	}
 }
 
-static bool scale_gamma(struct pwl_float_data *pwl_rgb,
+static void scale_gamma(struct pwl_float_data *pwl_rgb,
 		const struct dc_gamma *ramp,
 		struct dividers dividers)
 {
@@ -752,11 +752,9 @@ static bool scale_gamma(struct pwl_float_data *pwl_rgb,
 			dividers.divider3);
 	rgb->b = dal_fixed31_32_mul(rgb_last->b,
 			dividers.divider3);
-
-	return true;
 }
 
-static bool scale_gamma_dx(struct pwl_float_data *pwl_rgb,
+static void scale_gamma_dx(struct pwl_float_data *pwl_rgb,
 		const struct dc_gamma *ramp,
 		struct dividers dividers)
 {
@@ -818,8 +816,71 @@ static bool scale_gamma_dx(struct pwl_float_data *pwl_rgb,
 				pwl_rgb[i-1].g, 2), pwl_rgb[i-2].g);
 	pwl_rgb[i].b =  dal_fixed31_32_sub(dal_fixed31_32_mul_int(
 				pwl_rgb[i-1].b, 2), pwl_rgb[i-2].b);
+}
 
-	return true;
+/* todo: all these scale_gamma functions are inherently the same but
+ *  take different structures as params or different format for ramp
+ *  values. We could probably implement it in a more generic fashion
+ */
+static void scale_user_regamma_ramp(struct pwl_float_data *pwl_rgb,
+		const struct regamma_ramp *ramp,
+		struct dividers dividers)
+{
+	unsigned short max_driver = 0xFFFF;
+	unsigned short max_os = 0xFF00;
+	unsigned short scaler = max_os;
+	uint32_t i;
+	struct pwl_float_data *rgb = pwl_rgb;
+	struct pwl_float_data *rgb_last = rgb + GAMMA_RGB_256_ENTRIES - 1;
+
+	i = 0;
+	do {
+		if (ramp->gamma[i] > max_os ||
+				ramp->gamma[i + 256] > max_os ||
+				ramp->gamma[i + 512] > max_os) {
+			scaler = max_driver;
+			break;
+		}
+		i++;
+	} while (i != GAMMA_RGB_256_ENTRIES);
+
+	i = 0;
+	do {
+		rgb->r = dal_fixed31_32_from_fraction(
+				ramp->gamma[i], scaler);
+		rgb->g = dal_fixed31_32_from_fraction(
+				ramp->gamma[i + 256], scaler);
+		rgb->b = dal_fixed31_32_from_fraction(
+				ramp->gamma[i + 512], scaler);
+
+		++rgb;
+		++i;
+	} while (i != GAMMA_RGB_256_ENTRIES);
+
+	rgb->r = dal_fixed31_32_mul(rgb_last->r,
+			dividers.divider1);
+	rgb->g = dal_fixed31_32_mul(rgb_last->g,
+			dividers.divider1);
+	rgb->b = dal_fixed31_32_mul(rgb_last->b,
+			dividers.divider1);
+
+	++rgb;
+
+	rgb->r = dal_fixed31_32_mul(rgb_last->r,
+			dividers.divider2);
+	rgb->g = dal_fixed31_32_mul(rgb_last->g,
+			dividers.divider2);
+	rgb->b = dal_fixed31_32_mul(rgb_last->b,
+			dividers.divider2);
+
+	++rgb;
+
+	rgb->r = dal_fixed31_32_mul(rgb_last->r,
+			dividers.divider3);
+	rgb->g = dal_fixed31_32_mul(rgb_last->g,
+			dividers.divider3);
+	rgb->b = dal_fixed31_32_mul(rgb_last->b,
+			dividers.divider3);
 }
 
 /*
@@ -949,7 +1010,7 @@ static inline void copy_rgb_regamma_to_coordinates_x(
 	uint32_t i = 0;
 	const struct pwl_float_data_ex *rgb_regamma = rgb_ex;
 
-	while (i <= hw_points_num) {
+	while (i <= hw_points_num + 1) {
 		coords->regamma_y_red = rgb_regamma->r;
 		coords->regamma_y_green = rgb_regamma->g;
 		coords->regamma_y_blue = rgb_regamma->b;
@@ -1002,6 +1063,102 @@ static bool calculate_interpolated_hardware_curve(
 	return true;
 }
 
+/* The "old" interpolation uses a complicated scheme to build an array of
+ * coefficients while also using an array of 0-255 normalized to 0-1
+ * Then there's another loop using both of the above + new scaled user ramp
+ * and we concatenate them. It also searches for points of interpolation and
+ * uses enums for positions.
+ *
+ * This function uses a different approach:
+ * user ramp is always applied on X with 0/255, 1/255, 2/255, ..., 255/255
+ * To find index for hwX , we notice the following:
+ * i/255 <= hwX < (i+1)/255  <=> i <= 255*hwX < i+1
+ * See apply_lut_1d which is the same principle, but on 4K entry 1D LUT
+ *
+ * Once the index is known, combined Y is simply:
+ * user_ramp(index) + (hwX-index/255)*(user_ramp(index+1) - user_ramp(index)
+ *
+ * We should switch to this method in all cases, it's simpler and faster
+ * ToDo one day - for now this only applies to ADL regamma to avoid regression
+ * for regular use cases (sRGB and PQ)
+ */
+static void interpolate_user_regamma(uint32_t hw_points_num,
+		struct pwl_float_data *rgb_user,
+		bool apply_degamma,
+		struct dc_transfer_func_distributed_points *tf_pts)
+{
+	uint32_t i;
+	uint32_t color = 0;
+	int32_t index;
+	int32_t index_next;
+	struct fixed31_32 *tf_point;
+	struct fixed31_32 hw_x;
+	struct fixed31_32 norm_factor =
+			dal_fixed31_32_from_int_nonconst(255);
+	struct fixed31_32 norm_x;
+	struct fixed31_32 index_f;
+	struct fixed31_32 lut1;
+	struct fixed31_32 lut2;
+	struct fixed31_32 delta_lut;
+	struct fixed31_32 delta_index;
+
+	i = 0;
+	/* fixed_pt library has problems handling too small values */
+	while (i != 32) {
+		tf_pts->red[i] = dal_fixed31_32_zero;
+		tf_pts->green[i] = dal_fixed31_32_zero;
+		tf_pts->blue[i] = dal_fixed31_32_zero;
+		++i;
+	}
+	while (i <= hw_points_num + 1) {
+		for (color = 0; color < 3; color++) {
+			if (color == 0)
+				tf_point = &tf_pts->red[i];
+			else if (color == 1)
+				tf_point = &tf_pts->green[i];
+			else
+				tf_point = &tf_pts->blue[i];
+
+			if (apply_degamma) {
+				if (color == 0)
+					hw_x = coordinates_x[i].regamma_y_red;
+				else if (color == 1)
+					hw_x = coordinates_x[i].regamma_y_green;
+				else
+					hw_x = coordinates_x[i].regamma_y_blue;
+			} else
+				hw_x = coordinates_x[i].x;
+
+			norm_x = dal_fixed31_32_mul(norm_factor, hw_x);
+			index = dal_fixed31_32_floor(norm_x);
+			if (index < 0 || index > 255)
+				continue;
+
+			index_f = dal_fixed31_32_from_int_nonconst(index);
+			index_next = (index == 255) ? index : index + 1;
+
+			if (color == 0) {
+				lut1 = rgb_user[index].r;
+				lut2 = rgb_user[index_next].r;
+			} else if (color == 1) {
+				lut1 = rgb_user[index].g;
+				lut2 = rgb_user[index_next].g;
+			} else {
+				lut1 = rgb_user[index].b;
+				lut2 = rgb_user[index_next].b;
+			}
+
+			// we have everything now, so interpolate
+			delta_lut = dal_fixed31_32_sub(lut2, lut1);
+			delta_index = dal_fixed31_32_sub(norm_x, index_f);
+
+			*tf_point = dal_fixed31_32_add(lut1,
+				dal_fixed31_32_mul(delta_index, delta_lut));
+		}
+		++i;
+	}
+}
+
 static void build_new_custom_resulted_curve(
 	uint32_t hw_points_num,
 	struct dc_transfer_func_distributed_points *tf_pts)
@@ -1025,6 +1182,29 @@ static void build_new_custom_resulted_curve(
 	}
 }
 
+static void apply_degamma_for_user_regamma(struct pwl_float_data_ex *rgb_regamma,
+		uint32_t hw_points_num)
+{
+	uint32_t i;
+
+	struct gamma_coefficients coeff;
+	struct pwl_float_data_ex *rgb = rgb_regamma;
+	const struct hw_x_point *coord_x = coordinates_x;
+
+	build_coefficients(&coeff, true);
+
+	i = 0;
+	while (i != hw_points_num + 1) {
+		rgb->r = translate_from_linear_space_ex(
+				coord_x->x, &coeff, 0);
+		rgb->g = rgb->r;
+		rgb->b = rgb->r;
+		++coord_x;
+		++rgb;
+		++i;
+	}
+}
+
 static bool map_regamma_hw_to_x_user(
 	const struct dc_gamma *ramp,
 	struct pixel_gamma_point *coeff128,
@@ -1062,6 +1242,7 @@ static bool map_regamma_hw_to_x_user(
 		}
 	}
 
+	/* this should be named differently, all it does is clamp to 0-1 */
 	build_new_custom_resulted_curve(hw_points_num, tf_pts);
 
 	return true;
@@ -1168,6 +1349,113 @@ rgb_user_alloc_fail:
 	return ret;
 }
 
+bool calculate_user_regamma_coeff(struct dc_transfer_func *output_tf,
+		const struct regamma_lut *regamma)
+{
+	struct gamma_coefficients coeff;
+	const struct hw_x_point *coord_x = coordinates_x;
+	uint32_t i = 0;
+
+	do {
+		coeff.a0[i] = dal_fixed31_32_from_fraction(
+				regamma->coeff.A0[i], 10000000);
+		coeff.a1[i] = dal_fixed31_32_from_fraction(
+				regamma->coeff.A1[i], 1000);
+		coeff.a2[i] = dal_fixed31_32_from_fraction(
+				regamma->coeff.A2[i], 1000);
+		coeff.a3[i] = dal_fixed31_32_from_fraction(
+				regamma->coeff.A3[i], 1000);
+		coeff.user_gamma[i] = dal_fixed31_32_from_fraction(
+				regamma->coeff.gamma[i], 1000);
+
+		++i;
+	} while (i != 3);
+
+	i = 0;
+	/* fixed_pt library has problems handling too small values */
+	while (i != 32) {
+		output_tf->tf_pts.red[i] = dal_fixed31_32_zero;
+		output_tf->tf_pts.green[i] = dal_fixed31_32_zero;
+		output_tf->tf_pts.blue[i] = dal_fixed31_32_zero;
+		++coord_x;
+		++i;
+	}
+	while (i != MAX_HW_POINTS + 1) {
+		output_tf->tf_pts.red[i] = translate_from_linear_space_ex(
+				coord_x->x, &coeff, 0);
+		output_tf->tf_pts.green[i] = translate_from_linear_space_ex(
+				coord_x->x, &coeff, 1);
+		output_tf->tf_pts.blue[i] = translate_from_linear_space_ex(
+				coord_x->x, &coeff, 2);
+		++coord_x;
+		++i;
+	}
+
+	// this function just clamps output to 0-1
+	build_new_custom_resulted_curve(MAX_HW_POINTS, &output_tf->tf_pts);
+	output_tf->type = TF_TYPE_DISTRIBUTED_POINTS;
+
+	return true;
+}
+
+bool calculate_user_regamma_ramp(struct dc_transfer_func *output_tf,
+		const struct regamma_lut *regamma)
+{
+	struct dc_transfer_func_distributed_points *tf_pts = &output_tf->tf_pts;
+	struct dividers dividers;
+
+	struct pwl_float_data *rgb_user = NULL;
+	struct pwl_float_data_ex *rgb_regamma = NULL;
+	bool ret = false;
+
+	if (regamma == NULL)
+		return false;
+
+	output_tf->type = TF_TYPE_DISTRIBUTED_POINTS;
+
+	rgb_user = kzalloc(sizeof(*rgb_user) * (GAMMA_RGB_256_ENTRIES + _EXTRA_POINTS),
+			GFP_KERNEL);
+	if (!rgb_user)
+		goto rgb_user_alloc_fail;
+
+	rgb_regamma = kzalloc(sizeof(*rgb_regamma) * (MAX_HW_POINTS + _EXTRA_POINTS),
+			GFP_KERNEL);
+	if (!rgb_regamma)
+		goto rgb_regamma_alloc_fail;
+
+	dividers.divider1 = dal_fixed31_32_from_fraction(3, 2);
+	dividers.divider2 = dal_fixed31_32_from_int(2);
+	dividers.divider3 = dal_fixed31_32_from_fraction(5, 2);
+
+	scale_user_regamma_ramp(rgb_user, &regamma->ramp, dividers);
+
+	if (regamma->flags.bits.applyDegamma == 1) {
+		apply_degamma_for_user_regamma(rgb_regamma, MAX_HW_POINTS);
+		copy_rgb_regamma_to_coordinates_x(coordinates_x,
+				MAX_HW_POINTS, rgb_regamma);
+	}
+
+	interpolate_user_regamma(MAX_HW_POINTS, rgb_user,
+			regamma->flags.bits.applyDegamma, tf_pts);
+
+	// no custom HDR curves!
+	tf_pts->end_exponent = 0;
+	tf_pts->x_point_at_y1_red = 1;
+	tf_pts->x_point_at_y1_green = 1;
+	tf_pts->x_point_at_y1_blue = 1;
+
+	// this function just clamps output to 0-1
+	build_new_custom_resulted_curve(MAX_HW_POINTS, tf_pts);
+
+	ret = true;
+
+	kfree(rgb_regamma);
+rgb_regamma_alloc_fail:
+	kfree(rgb_user);
+rgb_user_alloc_fail:
+	return ret;
+}
+
 bool mod_color_calculate_degamma_params(struct dc_transfer_func *input_tf,
 		const struct dc_gamma *ramp, bool mapUserRamp)
 {
diff --git a/drivers/gpu/drm/amd/display/modules/color/color_gamma.h b/drivers/gpu/drm/amd/display/modules/color/color_gamma.h
index b7f9bc2..b6404899 100644
--- a/drivers/gpu/drm/amd/display/modules/color/color_gamma.h
+++ b/drivers/gpu/drm/amd/display/modules/color/color_gamma.h
@@ -32,6 +32,47 @@ struct dc_transfer_func_distributed_points;
 struct dc_rgb_fixed;
 enum dc_transfer_func_predefined;
 
+/* For SetRegamma ADL interface support
+ * Must match escape type
+ */
+union regamma_flags {
+	unsigned int raw;
+	struct {
+		unsigned int gammaRampArray       :1;    // RegammaRamp is in use
+		unsigned int gammaFromEdid        :1;    //gamma from edid is in use
+		unsigned int gammaFromEdidEx      :1;    //gamma from edid is in use , but only for Display Id 1.2
+		unsigned int gammaFromUser        :1;    //user custom gamma is used
+		unsigned int coeffFromUser        :1;    //coeff. A0-A3 from user is in use
+		unsigned int coeffFromEdid        :1;    //coeff. A0-A3 from edid is in use
+		unsigned int applyDegamma         :1;    //flag for additional degamma correction in driver
+		unsigned int gammaPredefinedSRGB  :1;    //flag for SRGB gamma
+		unsigned int gammaPredefinedPQ    :1;    //flag for PQ gamma
+		unsigned int gammaPredefinedPQ2084Interim :1;    //flag for PQ gamma, lower max nits
+		unsigned int gammaPredefined36    :1;    //flag for 3.6 gamma
+		unsigned int gammaPredefinedReset :1;    //flag to return to previous gamma
+	} bits;
+};
+
+struct regamma_ramp {
+	unsigned short gamma[256*3];  // gamma ramp packed  in same way as OS windows ,r , g & b
+};
+
+struct regamma_coeff {
+	int    gamma[3];
+	int    A0[3];
+	int    A1[3];
+	int    A2[3];
+	int    A3[3];
+};
+
+struct regamma_lut {
+	union regamma_flags flags;
+	union {
+		struct regamma_ramp ramp;
+		struct regamma_coeff coeff;
+	};
+};
+
 void setup_x_points_distribution(void);
 void precompute_pq(void);
 void precompute_de_pq(void);
@@ -45,9 +86,14 @@ bool mod_color_calculate_degamma_params(struct dc_transfer_func *output_tf,
 bool mod_color_calculate_curve(enum dc_transfer_func_predefined  trans,
 		struct dc_transfer_func_distributed_points *points);
 
-bool  mod_color_calculate_degamma_curve(enum dc_transfer_func_predefined trans,
+bool mod_color_calculate_degamma_curve(enum dc_transfer_func_predefined trans,
 				struct dc_transfer_func_distributed_points *points);
 
+bool calculate_user_regamma_coeff(struct dc_transfer_func *output_tf,
+		const struct regamma_lut *regamma);
+
+bool calculate_user_regamma_ramp(struct dc_transfer_func *output_tf,
+		const struct regamma_lut *regamma);
 
 
 #endif /* COLOR_MOD_COLOR_GAMMA_H_ */
-- 
cgit v1.1


From c0aceb7d6303ca138b0def39f25d432057548f43 Mon Sep 17 00:00:00 2001
From: Charlene Liu <charlene.liu@amd.com>
Date: Mon, 16 Apr 2018 15:14:15 -0400
Subject: drm/amd/display: add cursor TTU CRQ related

Signed-off-by: Charlene Liu <charlene.liu@amd.com>
Reviewed-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c |  7 +++++++
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h | 10 +++++++++-
 2 files changed, 16 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
index 5806217..759fcd1 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
@@ -613,6 +613,13 @@ void hubp1_program_deadline(
 	REG_SET(DCN_SURF1_TTU_CNTL1, 0,
 		REFCYC_PER_REQ_DELIVERY_PRE,
 		ttu_attr->refcyc_per_req_delivery_pre_c);
+
+	REG_SET_3(DCN_CUR0_TTU_CNTL0, 0,
+		REFCYC_PER_REQ_DELIVERY, ttu_attr->refcyc_per_req_delivery_cur0,
+		QoS_LEVEL_FIXED, ttu_attr->qos_level_fixed_cur0,
+		QoS_RAMP_DISABLE, ttu_attr->qos_ramp_disable_cur0);
+	REG_SET(DCN_CUR0_TTU_CNTL1, 0,
+		REFCYC_PER_REQ_DELIVERY_PRE, ttu_attr->refcyc_per_req_delivery_pre_cur0);
 }
 
 static void hubp1_setup(
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
index 920ae3a..02045a8 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
@@ -93,6 +93,8 @@
 	SRI(DCN_SURF0_TTU_CNTL1, HUBPREQ, id),\
 	SRI(DCN_SURF1_TTU_CNTL0, HUBPREQ, id),\
 	SRI(DCN_SURF1_TTU_CNTL1, HUBPREQ, id),\
+	SRI(DCN_CUR0_TTU_CNTL0, HUBPREQ, id),\
+	SRI(DCN_CUR0_TTU_CNTL1, HUBPREQ, id),\
 	SRI(HUBP_CLK_CNTL, HUBP, id)
 
 /* Register address initialization macro for ASICs with VM */
@@ -203,6 +205,8 @@
 	uint32_t DCN_SURF0_TTU_CNTL1; \
 	uint32_t DCN_SURF1_TTU_CNTL0; \
 	uint32_t DCN_SURF1_TTU_CNTL1; \
+	uint32_t DCN_CUR0_TTU_CNTL0; \
+	uint32_t DCN_CUR0_TTU_CNTL1; \
 	uint32_t DCN_VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_MSB; \
 	uint32_t DCN_VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LSB; \
 	uint32_t DCN_VM_CONTEXT0_PAGE_TABLE_START_ADDR_MSB; \
@@ -368,7 +372,11 @@
 	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_2, DST_Y_PER_PTE_ROW_NOM_C, mask_sh),\
 	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_3, REFCYC_PER_PTE_GROUP_NOM_C, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCN_VM_MX_L1_TLB_CNTL, ENABLE_L1_TLB, mask_sh),\
-	HUBP_SF(HUBPREQ0_DCN_VM_MX_L1_TLB_CNTL, SYSTEM_ACCESS_MODE, mask_sh)
+	HUBP_SF(HUBPREQ0_DCN_VM_MX_L1_TLB_CNTL, SYSTEM_ACCESS_MODE, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCN_CUR0_TTU_CNTL0, REFCYC_PER_REQ_DELIVERY, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCN_CUR0_TTU_CNTL0, QoS_LEVEL_FIXED, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCN_CUR0_TTU_CNTL0, QoS_RAMP_DISABLE, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCN_CUR0_TTU_CNTL1, REFCYC_PER_REQ_DELIVERY_PRE, mask_sh)
 
 #define HUBP_MASK_SH_LIST_DCN10(mask_sh)\
 	HUBP_MASK_SH_LIST_DCN(mask_sh),\
-- 
cgit v1.1


From 7c91bd434e5765dc5dbcf155253f2b8c740fbef9 Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Fri, 13 Apr 2018 09:40:21 -0400
Subject: drm/amd/display: add some DTN logs for input and output tf

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c   | 25 +++++++++++++
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h   | 27 ++++++++++++++
 .../drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c  | 41 ++++++++++++++++++++++
 drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h        | 16 +++++++++
 4 files changed, 109 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
index c008a71..8c4d9e5 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
@@ -98,6 +98,30 @@ enum gamut_remap_select {
 	GAMUT_REMAP_COMB_COEFF
 };
 
+void dpp_read_state(struct dpp *dpp_base,
+		struct dcn_dpp_state *s)
+{
+	struct dcn10_dpp *dpp = TO_DCN10_DPP(dpp_base);
+
+	REG_GET(CM_IGAM_CONTROL,
+			CM_IGAM_LUT_MODE, &s->igam_lut_mode);
+	REG_GET(CM_IGAM_CONTROL,
+			CM_IGAM_INPUT_FORMAT, &s->igam_input_format);
+	REG_GET(CM_DGAM_CONTROL,
+			CM_DGAM_LUT_MODE, &s->dgam_lut_mode);
+	REG_GET(CM_RGAM_CONTROL,
+			CM_RGAM_LUT_MODE, &s->rgam_lut_mode);
+	REG_GET(CM_GAMUT_REMAP_CONTROL,
+			CM_GAMUT_REMAP_MODE, &s->gamut_remap_mode);
+
+	s->gamut_remap_c11_c12 = REG_READ(CM_GAMUT_REMAP_C11_C12);
+	s->gamut_remap_c13_c14 = REG_READ(CM_GAMUT_REMAP_C13_C14);
+	s->gamut_remap_c21_c22 = REG_READ(CM_GAMUT_REMAP_C21_C22);
+	s->gamut_remap_c23_c24 = REG_READ(CM_GAMUT_REMAP_C23_C24);
+	s->gamut_remap_c31_c32 = REG_READ(CM_GAMUT_REMAP_C31_C32);
+	s->gamut_remap_c33_c34 = REG_READ(CM_GAMUT_REMAP_C33_C34);
+}
+
 /* Program gamut remap in bypass mode */
 void dpp_set_gamut_remap_bypass(struct dcn10_dpp *dpp)
 {
@@ -450,6 +474,7 @@ void dpp1_dppclk_control(
 }
 
 static const struct dpp_funcs dcn10_dpp_funcs = {
+		.dpp_read_state = dpp_read_state,
 		.dpp_reset = dpp_reset,
 		.dpp_set_scaler = dpp1_dscl_set_scaler_manual_scale,
 		.dpp_get_optimal_number_of_taps = dpp_get_optimal_number_of_taps,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h
index 3fccf99..5944a3b 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h
@@ -44,6 +44,10 @@
 #define TF_REG_LIST_DCN(id) \
 	SRI(CM_GAMUT_REMAP_CONTROL, CM, id),\
 	SRI(CM_GAMUT_REMAP_C11_C12, CM, id),\
+	SRI(CM_GAMUT_REMAP_C13_C14, CM, id),\
+	SRI(CM_GAMUT_REMAP_C21_C22, CM, id),\
+	SRI(CM_GAMUT_REMAP_C23_C24, CM, id),\
+	SRI(CM_GAMUT_REMAP_C31_C32, CM, id),\
 	SRI(CM_GAMUT_REMAP_C33_C34, CM, id),\
 	SRI(DSCL_EXT_OVERSCAN_LEFT_RIGHT, DSCL, id), \
 	SRI(DSCL_EXT_OVERSCAN_TOP_BOTTOM, DSCL, id), \
@@ -177,6 +181,14 @@
 	TF_SF(CM0_CM_GAMUT_REMAP_CONTROL, CM_GAMUT_REMAP_MODE, mask_sh),\
 	TF_SF(CM0_CM_GAMUT_REMAP_C11_C12, CM_GAMUT_REMAP_C11, mask_sh),\
 	TF_SF(CM0_CM_GAMUT_REMAP_C11_C12, CM_GAMUT_REMAP_C12, mask_sh),\
+	TF_SF(CM0_CM_GAMUT_REMAP_C13_C14, CM_GAMUT_REMAP_C13, mask_sh),\
+	TF_SF(CM0_CM_GAMUT_REMAP_C13_C14, CM_GAMUT_REMAP_C14, mask_sh),\
+	TF_SF(CM0_CM_GAMUT_REMAP_C21_C22, CM_GAMUT_REMAP_C21, mask_sh),\
+	TF_SF(CM0_CM_GAMUT_REMAP_C21_C22, CM_GAMUT_REMAP_C22, mask_sh),\
+	TF_SF(CM0_CM_GAMUT_REMAP_C23_C24, CM_GAMUT_REMAP_C23, mask_sh),\
+	TF_SF(CM0_CM_GAMUT_REMAP_C23_C24, CM_GAMUT_REMAP_C24, mask_sh),\
+	TF_SF(CM0_CM_GAMUT_REMAP_C31_C32, CM_GAMUT_REMAP_C31, mask_sh),\
+	TF_SF(CM0_CM_GAMUT_REMAP_C31_C32, CM_GAMUT_REMAP_C32, mask_sh),\
 	TF_SF(CM0_CM_GAMUT_REMAP_C33_C34, CM_GAMUT_REMAP_C33, mask_sh),\
 	TF_SF(CM0_CM_GAMUT_REMAP_C33_C34, CM_GAMUT_REMAP_C34, mask_sh),\
 	TF_SF(DSCL0_DSCL_EXT_OVERSCAN_LEFT_RIGHT, EXT_OVERSCAN_LEFT, mask_sh),\
@@ -524,6 +536,14 @@
 	type CM_GAMUT_REMAP_MODE; \
 	type CM_GAMUT_REMAP_C11; \
 	type CM_GAMUT_REMAP_C12; \
+	type CM_GAMUT_REMAP_C13; \
+	type CM_GAMUT_REMAP_C14; \
+	type CM_GAMUT_REMAP_C21; \
+	type CM_GAMUT_REMAP_C22; \
+	type CM_GAMUT_REMAP_C23; \
+	type CM_GAMUT_REMAP_C24; \
+	type CM_GAMUT_REMAP_C31; \
+	type CM_GAMUT_REMAP_C32; \
 	type CM_GAMUT_REMAP_C33; \
 	type CM_GAMUT_REMAP_C34; \
 	type CM_COMA_C11; \
@@ -1095,6 +1115,10 @@ struct dcn_dpp_mask {
 	uint32_t RECOUT_SIZE; \
 	uint32_t CM_GAMUT_REMAP_CONTROL; \
 	uint32_t CM_GAMUT_REMAP_C11_C12; \
+	uint32_t CM_GAMUT_REMAP_C13_C14; \
+	uint32_t CM_GAMUT_REMAP_C21_C22; \
+	uint32_t CM_GAMUT_REMAP_C23_C24; \
+	uint32_t CM_GAMUT_REMAP_C31_C32; \
 	uint32_t CM_GAMUT_REMAP_C33_C34; \
 	uint32_t CM_COMA_C11_C12; \
 	uint32_t CM_COMA_C33_C34; \
@@ -1407,6 +1431,9 @@ bool dpp_get_optimal_number_of_taps(
 		struct scaler_data *scl_data,
 		const struct scaling_taps *in_taps);
 
+void dpp_read_state(struct dpp *dpp_base,
+		struct dcn_dpp_state *s);
+
 void dpp_reset(struct dpp *dpp_base);
 
 void dpp1_cm_program_regamma_lut(
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
index 8eea38b..9a64211 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -250,6 +250,47 @@ void dcn10_log_hw_state(struct dc *dc)
 	}
 	DTN_INFO("\n");
 
+	DTN_INFO("DPP:    IGAM format  IGAM mode    DGAM mode    RGAM mode"
+			"  GAMUT mode  C11 C12   C13 C14   C21 C22   C23 C24   "
+			"C31 C32   C33 C34\n");
+	for (i = 0; i < pool->pipe_count; i++) {
+		struct dpp *dpp = pool->dpps[i];
+		struct dcn_dpp_state s;
+
+		dpp->funcs->dpp_read_state(dpp, &s);
+
+		DTN_INFO("[%2d]:  %11xh  %-11s  %-11s  %-11s"
+				"%08xh   %08xh %08xh %08xh %08xh %08xh %08xh",
+				dpp->inst,
+				s.igam_input_format,
+				(s.igam_lut_mode == 0) ? "BypassFixed" :
+					((s.igam_lut_mode == 1) ? "BypassFloat" :
+					((s.igam_lut_mode == 2) ? "RAM" :
+					((s.igam_lut_mode == 3) ? "RAM" :
+								 "Unknown"))),
+				(s.dgam_lut_mode == 0) ? "Bypass" :
+					((s.dgam_lut_mode == 1) ? "sRGB" :
+					((s.dgam_lut_mode == 2) ? "Ycc" :
+					((s.dgam_lut_mode == 3) ? "RAM" :
+					((s.dgam_lut_mode == 4) ? "RAM" :
+								 "Unknown")))),
+				(s.rgam_lut_mode == 0) ? "Bypass" :
+					((s.rgam_lut_mode == 1) ? "sRGB" :
+					((s.rgam_lut_mode == 2) ? "Ycc" :
+					((s.rgam_lut_mode == 3) ? "RAM" :
+					((s.rgam_lut_mode == 4) ? "RAM" :
+								 "Unknown")))),
+				s.gamut_remap_mode,
+				s.gamut_remap_c11_c12,
+				s.gamut_remap_c13_c14,
+				s.gamut_remap_c21_c22,
+				s.gamut_remap_c23_c24,
+				s.gamut_remap_c31_c32,
+				s.gamut_remap_c33_c34);
+		DTN_INFO("\n");
+	}
+	DTN_INFO("\n");
+
 	DTN_INFO("MPCC:  OPP  DPP  MPCCBOT  MODE  ALPHA_MODE  PREMULT  OVERLAP_ONLY  IDLE\n");
 	for (i = 0; i < pool->pipe_count; i++) {
 		struct mpcc_state s = {0};
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h
index bb7af1b..582458f 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h
@@ -44,7 +44,23 @@ struct dpp_grph_csc_adjustment {
 	enum graphics_gamut_adjust_type gamut_adjust_type;
 };
 
+struct dcn_dpp_state {
+	uint32_t igam_lut_mode;
+	uint32_t igam_input_format;
+	uint32_t dgam_lut_mode;
+	uint32_t rgam_lut_mode;
+	uint32_t gamut_remap_mode;
+	uint32_t gamut_remap_c11_c12;
+	uint32_t gamut_remap_c13_c14;
+	uint32_t gamut_remap_c21_c22;
+	uint32_t gamut_remap_c23_c24;
+	uint32_t gamut_remap_c31_c32;
+	uint32_t gamut_remap_c33_c34;
+};
+
 struct dpp_funcs {
+	void (*dpp_read_state)(struct dpp *dpp, struct dcn_dpp_state *s);
+
 	void (*dpp_reset)(struct dpp *dpp);
 
 	void (*dpp_set_scaler)(struct dpp *dpp,
-- 
cgit v1.1


From 7b265fd96cbfa03630a2db90b3891b8397bf2208 Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Tue, 17 Apr 2018 12:12:56 -0400
Subject: drm/amd/display: update dtn logging and goldens

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
index 9a64211..574d37c 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -260,7 +260,7 @@ void dcn10_log_hw_state(struct dc *dc)
 		dpp->funcs->dpp_read_state(dpp, &s);
 
 		DTN_INFO("[%2d]:  %11xh  %-11s  %-11s  %-11s"
-				"%08xh   %08xh %08xh %08xh %08xh %08xh %08xh",
+				"%8x    %08xh %08xh %08xh %08xh %08xh %08xh",
 				dpp->inst,
 				s.igam_input_format,
 				(s.igam_lut_mode == 0) ? "BypassFixed" :
-- 
cgit v1.1


From 34cb6b3860a4aecafaae0df8fa84b6fc784f507c Mon Sep 17 00:00:00 2001
From: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Date: Wed, 11 Apr 2018 11:51:32 -0400
Subject: drm/amd/display: compact the rq/dlg/ttu log

Signed-off-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c  |   4 +-
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h  |  24 ++-
 .../drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c  | 235 +++++++++------------
 drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h       |  22 +-
 4 files changed, 128 insertions(+), 157 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
index 759fcd1..159bebc 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
@@ -763,10 +763,10 @@ void min_set_viewport(
 		  PRI_VIEWPORT_Y_START_C, viewport_c->y);
 }
 
-void hubp1_read_state(struct hubp *hubp,
-		struct dcn_hubp_state *s)
+void hubp1_read_state(struct hubp *hubp)
 {
 	struct dcn10_hubp *hubp1 = TO_DCN10_HUBP(hubp);
+	struct dcn_hubp_state *s = &hubp1->state;
 	struct _vcs_dpi_display_dlg_regs_st *dlg_attr = &s->dlg_attr;
 	struct _vcs_dpi_display_ttu_regs_st *ttu_attr = &s->ttu_attr;
 	struct _vcs_dpi_display_rq_regs_st *rq_regs = &s->rq_regs;
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
index 02045a8..fe9b8c4 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
@@ -619,8 +619,29 @@ struct dcn_mi_mask {
 	DCN_HUBP_REG_FIELD_LIST(uint32_t);
 };
 
+struct dcn_hubp_state {
+	struct _vcs_dpi_display_dlg_regs_st dlg_attr;
+	struct _vcs_dpi_display_ttu_regs_st ttu_attr;
+	struct _vcs_dpi_display_rq_regs_st rq_regs;
+	uint32_t pixel_format;
+	uint32_t inuse_addr_hi;
+	uint32_t viewport_width;
+	uint32_t viewport_height;
+	uint32_t rotation_angle;
+	uint32_t h_mirror_en;
+	uint32_t sw_mode;
+	uint32_t dcc_en;
+	uint32_t blank_en;
+	uint32_t underflow_status;
+	uint32_t ttu_disable;
+	uint32_t min_ttu_vblank;
+	uint32_t qos_level_low_wm;
+	uint32_t qos_level_high_wm;
+};
+
 struct dcn10_hubp {
 	struct hubp base;
+	struct dcn_hubp_state state;
 	const struct dcn_mi_registers *hubp_regs;
 	const struct dcn_mi_shift *hubp_shift;
 	const struct dcn_mi_mask *hubp_mask;
@@ -698,8 +719,7 @@ void dcn10_hubp_construct(
 	const struct dcn_mi_shift *hubp_shift,
 	const struct dcn_mi_mask *hubp_mask);
 
-void hubp1_read_state(struct hubp *hubp,
-		struct dcn_hubp_state *s);
+void hubp1_read_state(struct hubp *hubp);
 
 enum cursor_pitch hubp1_get_cursor_pitch(unsigned int pitch);
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
index 574d37c..572fa60 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -112,143 +112,127 @@ void dcn10_log_hubbub_state(struct dc *dc)
 	DTN_INFO("\n");
 }
 
-static void print_rq_dlg_ttu_regs(struct dc_context *dc_ctx, struct dcn_hubp_state *s)
-{
-	struct _vcs_dpi_display_dlg_regs_st *dlg_regs = &s->dlg_attr;
-	struct _vcs_dpi_display_ttu_regs_st *ttu_regs = &s->ttu_attr;
-	struct _vcs_dpi_display_rq_regs_st *rq_regs = &s->rq_regs;
-
-	DTN_INFO("========Requester========\n");
-	DTN_INFO("drq_expansion_mode      = 0x%0x\n", rq_regs->drq_expansion_mode);
-	DTN_INFO("prq_expansion_mode      = 0x%0x\n", rq_regs->prq_expansion_mode);
-	DTN_INFO("mrq_expansion_mode      = 0x%0x\n", rq_regs->mrq_expansion_mode);
-	DTN_INFO("crq_expansion_mode      = 0x%0x\n", rq_regs->crq_expansion_mode);
-	DTN_INFO("plane1_base_address     = 0x%0x\n", rq_regs->plane1_base_address);
-	DTN_INFO("==<LUMA>==\n");
-	DTN_INFO("chunk_size              = 0x%0x\n", rq_regs->rq_regs_l.chunk_size);
-	DTN_INFO("min_chunk_size          = 0x%0x\n", rq_regs->rq_regs_l.min_chunk_size);
-	DTN_INFO("meta_chunk_size         = 0x%0x\n", rq_regs->rq_regs_l.meta_chunk_size);
-	DTN_INFO("min_meta_chunk_size     = 0x%0x\n", rq_regs->rq_regs_l.min_meta_chunk_size);
-	DTN_INFO("dpte_group_size         = 0x%0x\n", rq_regs->rq_regs_l.dpte_group_size);
-	DTN_INFO("mpte_group_size         = 0x%0x\n", rq_regs->rq_regs_l.mpte_group_size);
-	DTN_INFO("swath_height            = 0x%0x\n", rq_regs->rq_regs_l.swath_height);
-	DTN_INFO("pte_row_height_linear   = 0x%0x\n", rq_regs->rq_regs_l.pte_row_height_linear);
-	DTN_INFO("==<CHROMA>==\n");
-	DTN_INFO("chunk_size              = 0x%0x\n", rq_regs->rq_regs_c.chunk_size);
-	DTN_INFO("min_chunk_size          = 0x%0x\n", rq_regs->rq_regs_c.min_chunk_size);
-	DTN_INFO("meta_chunk_size         = 0x%0x\n", rq_regs->rq_regs_c.meta_chunk_size);
-	DTN_INFO("min_meta_chunk_size     = 0x%0x\n", rq_regs->rq_regs_c.min_meta_chunk_size);
-	DTN_INFO("dpte_group_size         = 0x%0x\n", rq_regs->rq_regs_c.dpte_group_size);
-	DTN_INFO("mpte_group_size         = 0x%0x\n", rq_regs->rq_regs_c.mpte_group_size);
-	DTN_INFO("swath_height            = 0x%0x\n", rq_regs->rq_regs_c.swath_height);
-	DTN_INFO("pte_row_height_linear   = 0x%0x\n", rq_regs->rq_regs_c.pte_row_height_linear);
-
-	DTN_INFO("========DLG========\n");
-	DTN_INFO("refcyc_h_blank_end                  = 0x%0x\n", dlg_regs->refcyc_h_blank_end);
-	DTN_INFO("dlg_vblank_end                      = 0x%0x\n", dlg_regs->dlg_vblank_end);
-	DTN_INFO("min_dst_y_next_start                = 0x%0x\n", dlg_regs->min_dst_y_next_start);
-	DTN_INFO("refcyc_per_htotal                   = 0x%0x\n", dlg_regs->refcyc_per_htotal);
-	DTN_INFO("refcyc_x_after_scaler               = 0x%0x\n", dlg_regs->refcyc_x_after_scaler);
-	DTN_INFO("dst_y_after_scaler                  = 0x%0x\n", dlg_regs->dst_y_after_scaler);
-	DTN_INFO("dst_y_prefetch                      = 0x%0x\n", dlg_regs->dst_y_prefetch);
-	DTN_INFO("dst_y_per_vm_vblank                 = 0x%0x\n", dlg_regs->dst_y_per_vm_vblank);
-	DTN_INFO("dst_y_per_row_vblank                = 0x%0x\n", dlg_regs->dst_y_per_row_vblank);
-	DTN_INFO("dst_y_per_vm_flip                   = 0x%0x\n", dlg_regs->dst_y_per_vm_flip);
-	DTN_INFO("dst_y_per_row_flip                  = 0x%0x\n", dlg_regs->dst_y_per_row_flip);
-	DTN_INFO("ref_freq_to_pix_freq                = 0x%0x\n", dlg_regs->ref_freq_to_pix_freq);
-	DTN_INFO("vratio_prefetch                     = 0x%0x\n", dlg_regs->vratio_prefetch);
-	DTN_INFO("vratio_prefetch_c                   = 0x%0x\n", dlg_regs->vratio_prefetch_c);
-	DTN_INFO("refcyc_per_pte_group_vblank_l       = 0x%0x\n", dlg_regs->refcyc_per_pte_group_vblank_l);
-	DTN_INFO("refcyc_per_pte_group_vblank_c       = 0x%0x\n", dlg_regs->refcyc_per_pte_group_vblank_c);
-	DTN_INFO("refcyc_per_meta_chunk_vblank_l      = 0x%0x\n", dlg_regs->refcyc_per_meta_chunk_vblank_l);
-	DTN_INFO("refcyc_per_meta_chunk_vblank_c      = 0x%0x\n", dlg_regs->refcyc_per_meta_chunk_vblank_c);
-	DTN_INFO("refcyc_per_pte_group_flip_l         = 0x%0x\n", dlg_regs->refcyc_per_pte_group_flip_l);
-	DTN_INFO("refcyc_per_pte_group_flip_c         = 0x%0x\n", dlg_regs->refcyc_per_pte_group_flip_c);
-	DTN_INFO("refcyc_per_meta_chunk_flip_l        = 0x%0x\n", dlg_regs->refcyc_per_meta_chunk_flip_l);
-	DTN_INFO("refcyc_per_meta_chunk_flip_c        = 0x%0x\n", dlg_regs->refcyc_per_meta_chunk_flip_c);
-	DTN_INFO("dst_y_per_pte_row_nom_l             = 0x%0x\n", dlg_regs->dst_y_per_pte_row_nom_l);
-	DTN_INFO("dst_y_per_pte_row_nom_c             = 0x%0x\n", dlg_regs->dst_y_per_pte_row_nom_c);
-	DTN_INFO("refcyc_per_pte_group_nom_l          = 0x%0x\n", dlg_regs->refcyc_per_pte_group_nom_l);
-	DTN_INFO("refcyc_per_pte_group_nom_c          = 0x%0x\n", dlg_regs->refcyc_per_pte_group_nom_c);
-	DTN_INFO("dst_y_per_meta_row_nom_l            = 0x%0x\n", dlg_regs->dst_y_per_meta_row_nom_l);
-	DTN_INFO("dst_y_per_meta_row_nom_c            = 0x%0x\n", dlg_regs->dst_y_per_meta_row_nom_c);
-	DTN_INFO("refcyc_per_meta_chunk_nom_l         = 0x%0x\n", dlg_regs->refcyc_per_meta_chunk_nom_l);
-	DTN_INFO("refcyc_per_meta_chunk_nom_c         = 0x%0x\n", dlg_regs->refcyc_per_meta_chunk_nom_c);
-	DTN_INFO("refcyc_per_line_delivery_pre_l      = 0x%0x\n", dlg_regs->refcyc_per_line_delivery_pre_l);
-	DTN_INFO("refcyc_per_line_delivery_pre_c      = 0x%0x\n", dlg_regs->refcyc_per_line_delivery_pre_c);
-	DTN_INFO("refcyc_per_line_delivery_l          = 0x%0x\n", dlg_regs->refcyc_per_line_delivery_l);
-	DTN_INFO("refcyc_per_line_delivery_c          = 0x%0x\n", dlg_regs->refcyc_per_line_delivery_c);
-	DTN_INFO("chunk_hdl_adjust_cur0               = 0x%0x\n", dlg_regs->chunk_hdl_adjust_cur0);
-	DTN_INFO("dst_y_offset_cur1                   = 0x%0x\n", dlg_regs->dst_y_offset_cur1);
-	DTN_INFO("chunk_hdl_adjust_cur1               = 0x%0x\n", dlg_regs->chunk_hdl_adjust_cur1);
-	DTN_INFO("vready_after_vcount0                = 0x%0x\n", dlg_regs->vready_after_vcount0);
-	DTN_INFO("dst_y_delta_drq_limit               = 0x%0x\n", dlg_regs->dst_y_delta_drq_limit);
-	DTN_INFO("xfc_reg_transfer_delay              = 0x%0x\n", dlg_regs->xfc_reg_transfer_delay);
-	DTN_INFO("xfc_reg_precharge_delay             = 0x%0x\n", dlg_regs->xfc_reg_precharge_delay);
-	DTN_INFO("xfc_reg_remote_surface_flip_latency = 0x%0x\n", dlg_regs->xfc_reg_remote_surface_flip_latency);
-
-	DTN_INFO("========TTU========\n");
-	DTN_INFO("qos_level_low_wm                  = 0x%0x\n", ttu_regs->qos_level_low_wm);
-	DTN_INFO("qos_level_high_wm                 = 0x%0x\n", ttu_regs->qos_level_high_wm);
-	DTN_INFO("min_ttu_vblank                    = 0x%0x\n", ttu_regs->min_ttu_vblank);
-	DTN_INFO("qos_level_flip                    = 0x%0x\n", ttu_regs->qos_level_flip);
-	DTN_INFO("refcyc_per_req_delivery_pre_l     = 0x%0x\n", ttu_regs->refcyc_per_req_delivery_pre_l);
-	DTN_INFO("refcyc_per_req_delivery_l         = 0x%0x\n", ttu_regs->refcyc_per_req_delivery_l);
-	DTN_INFO("refcyc_per_req_delivery_pre_c     = 0x%0x\n", ttu_regs->refcyc_per_req_delivery_pre_c);
-	DTN_INFO("refcyc_per_req_delivery_c         = 0x%0x\n", ttu_regs->refcyc_per_req_delivery_c);
-	DTN_INFO("refcyc_per_req_delivery_cur0      = 0x%0x\n", ttu_regs->refcyc_per_req_delivery_cur0);
-	DTN_INFO("refcyc_per_req_delivery_pre_cur0  = 0x%0x\n", ttu_regs->refcyc_per_req_delivery_pre_cur0);
-	DTN_INFO("refcyc_per_req_delivery_cur1      = 0x%0x\n", ttu_regs->refcyc_per_req_delivery_cur1);
-	DTN_INFO("refcyc_per_req_delivery_pre_cur1  = 0x%0x\n", ttu_regs->refcyc_per_req_delivery_pre_cur1);
-	DTN_INFO("qos_level_fixed_l                 = 0x%0x\n", ttu_regs->qos_level_fixed_l);
-	DTN_INFO("qos_ramp_disable_l                = 0x%0x\n", ttu_regs->qos_ramp_disable_l);
-	DTN_INFO("qos_level_fixed_c                 = 0x%0x\n", ttu_regs->qos_level_fixed_c);
-	DTN_INFO("qos_ramp_disable_c                = 0x%0x\n", ttu_regs->qos_ramp_disable_c);
-	DTN_INFO("qos_level_fixed_cur0              = 0x%0x\n", ttu_regs->qos_level_fixed_cur0);
-	DTN_INFO("qos_ramp_disable_cur0             = 0x%0x\n", ttu_regs->qos_ramp_disable_cur0);
-	DTN_INFO("qos_level_fixed_cur1              = 0x%0x\n", ttu_regs->qos_level_fixed_cur1);
-	DTN_INFO("qos_ramp_disable_cur1             = 0x%0x\n", ttu_regs->qos_ramp_disable_cur1);
-}
-
-void dcn10_log_hw_state(struct dc *dc)
+static void dcn10_log_hubp_states(struct dc *dc)
 {
 	struct dc_context *dc_ctx = dc->ctx;
 	struct resource_pool *pool = dc->res_pool;
 	int i;
 
-	DTN_INFO_BEGIN();
-
-	dcn10_log_hubbub_state(dc);
-
 	DTN_INFO("HUBP:  format  addr_hi  width  height"
 			"  rot  mir  sw_mode  dcc_en  blank_en  ttu_dis  underflow"
 			"   min_ttu_vblank       qos_low_wm      qos_high_wm\n");
 	for (i = 0; i < pool->pipe_count; i++) {
 		struct hubp *hubp = pool->hubps[i];
-		struct dcn_hubp_state s;
+		struct dcn_hubp_state *s = &(TO_DCN10_HUBP(hubp)->state);
 
-		hubp->funcs->hubp_read_state(hubp, &s);
+		hubp->funcs->hubp_read_state(hubp);
 
 		DTN_INFO("[%2d]:  %5xh  %6xh  %5d  %6d  %2xh  %2xh  %6xh"
 				"  %6d  %8d  %7d  %8xh",
 				hubp->inst,
-				s.pixel_format,
-				s.inuse_addr_hi,
-				s.viewport_width,
-				s.viewport_height,
-				s.rotation_angle,
-				s.h_mirror_en,
-				s.sw_mode,
-				s.dcc_en,
-				s.blank_en,
-				s.ttu_disable,
-				s.underflow_status);
-		DTN_INFO_MICRO_SEC(s.min_ttu_vblank);
-		DTN_INFO_MICRO_SEC(s.qos_level_low_wm);
-		DTN_INFO_MICRO_SEC(s.qos_level_high_wm);
+				s->pixel_format,
+				s->inuse_addr_hi,
+				s->viewport_width,
+				s->viewport_height,
+				s->rotation_angle,
+				s->h_mirror_en,
+				s->sw_mode,
+				s->dcc_en,
+				s->blank_en,
+				s->ttu_disable,
+				s->underflow_status);
+		DTN_INFO_MICRO_SEC(s->min_ttu_vblank);
+		DTN_INFO_MICRO_SEC(s->qos_level_low_wm);
+		DTN_INFO_MICRO_SEC(s->qos_level_high_wm);
 		DTN_INFO("\n");
 	}
+
+	DTN_INFO("\n=========RQ========\n");
+	DTN_INFO("HUBP:  drq_exp_m  prq_exp_m  mrq_exp_m  crq_exp_m  plane1_ba  L:chunk_s  min_chu_s  meta_ch_s"
+		"  min_m_c_s  dpte_gr_s  mpte_gr_s  swath_hei  pte_row_h  C:chunk_s  min_chu_s  meta_ch_s"
+		"  min_m_c_s  dpte_gr_s  mpte_gr_s  swath_hei  pte_row_h\n");
+	for (i = 0; i < pool->pipe_count; i++) {
+		struct dcn_hubp_state *s = &(TO_DCN10_HUBP(pool->hubps[i])->state);
+		struct _vcs_dpi_display_rq_regs_st *rq_regs = &s->rq_regs;
+
+		DTN_INFO("[%2d]:  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh\n",
+			i, rq_regs->drq_expansion_mode, rq_regs->prq_expansion_mode, rq_regs->mrq_expansion_mode,
+			rq_regs->crq_expansion_mode, rq_regs->plane1_base_address, rq_regs->rq_regs_l.chunk_size,
+			rq_regs->rq_regs_l.min_chunk_size, rq_regs->rq_regs_l.meta_chunk_size,
+			rq_regs->rq_regs_l.min_meta_chunk_size, rq_regs->rq_regs_l.dpte_group_size,
+			rq_regs->rq_regs_l.mpte_group_size, rq_regs->rq_regs_l.swath_height,
+			rq_regs->rq_regs_l.pte_row_height_linear, rq_regs->rq_regs_c.chunk_size, rq_regs->rq_regs_c.min_chunk_size,
+			rq_regs->rq_regs_c.meta_chunk_size, rq_regs->rq_regs_c.min_meta_chunk_size,
+			rq_regs->rq_regs_c.dpte_group_size, rq_regs->rq_regs_c.mpte_group_size,
+			rq_regs->rq_regs_c.swath_height, rq_regs->rq_regs_c.pte_row_height_linear);
+	}
+
+	DTN_INFO("========DLG========\n");
+	DTN_INFO("HUBP:  rc_hbe     dlg_vbe    min_d_y_n  rc_per_ht  rc_x_a_s "
+			"  dst_y_a_s  dst_y_pf   dst_y_vvb  dst_y_rvb  dst_y_vfl  dst_y_rfl  rf_pix_fq"
+			"  vratio_pf  vrat_pf_c  rc_pg_vbl  rc_pg_vbc  rc_mc_vbl  rc_mc_vbc  rc_pg_fll"
+			"  rc_pg_flc  rc_mc_fll  rc_mc_flc  pr_nom_l   pr_nom_c   rc_pg_nl   rc_pg_nc "
+			"  mr_nom_l   mr_nom_c   rc_mc_nl   rc_mc_nc   rc_ld_pl   rc_ld_pc   rc_ld_l  "
+			"  rc_ld_c    cha_cur0   ofst_cur1  cha_cur1   vr_af_vc0  ddrq_limt  x_rt_dlay"
+			"  x_rp_dlay  x_rr_sfl\n");
+	for (i = 0; i < pool->pipe_count; i++) {
+		struct dcn_hubp_state *s = &(TO_DCN10_HUBP(pool->hubps[i])->state);
+		struct _vcs_dpi_display_dlg_regs_st *dlg_regs = &s->dlg_attr;
+
+		DTN_INFO("[%2d]:  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh"
+			"%  8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh"
+			"  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh\n",
+			i, dlg_regs->refcyc_h_blank_end, dlg_regs->dlg_vblank_end, dlg_regs->min_dst_y_next_start,
+			dlg_regs->refcyc_per_htotal, dlg_regs->refcyc_x_after_scaler, dlg_regs->dst_y_after_scaler,
+			dlg_regs->dst_y_prefetch, dlg_regs->dst_y_per_vm_vblank, dlg_regs->dst_y_per_row_vblank,
+			dlg_regs->dst_y_per_vm_flip, dlg_regs->dst_y_per_row_flip, dlg_regs->ref_freq_to_pix_freq,
+			dlg_regs->vratio_prefetch, dlg_regs->vratio_prefetch_c, dlg_regs->refcyc_per_pte_group_vblank_l,
+			dlg_regs->refcyc_per_pte_group_vblank_c, dlg_regs->refcyc_per_meta_chunk_vblank_l,
+			dlg_regs->refcyc_per_meta_chunk_vblank_c, dlg_regs->refcyc_per_pte_group_flip_l,
+			dlg_regs->refcyc_per_pte_group_flip_c, dlg_regs->refcyc_per_meta_chunk_flip_l,
+			dlg_regs->refcyc_per_meta_chunk_flip_c, dlg_regs->dst_y_per_pte_row_nom_l,
+			dlg_regs->dst_y_per_pte_row_nom_c, dlg_regs->refcyc_per_pte_group_nom_l,
+			dlg_regs->refcyc_per_pte_group_nom_c, dlg_regs->dst_y_per_meta_row_nom_l,
+			dlg_regs->dst_y_per_meta_row_nom_c, dlg_regs->refcyc_per_meta_chunk_nom_l,
+			dlg_regs->refcyc_per_meta_chunk_nom_c, dlg_regs->refcyc_per_line_delivery_pre_l,
+			dlg_regs->refcyc_per_line_delivery_pre_c, dlg_regs->refcyc_per_line_delivery_l,
+			dlg_regs->refcyc_per_line_delivery_c, dlg_regs->chunk_hdl_adjust_cur0, dlg_regs->dst_y_offset_cur1,
+			dlg_regs->chunk_hdl_adjust_cur1, dlg_regs->vready_after_vcount0, dlg_regs->dst_y_delta_drq_limit,
+			dlg_regs->xfc_reg_transfer_delay, dlg_regs->xfc_reg_precharge_delay,
+			dlg_regs->xfc_reg_remote_surface_flip_latency);
+	}
+
+	DTN_INFO("========TTU========\n");
+	DTN_INFO("HUBP:  qos_ll_wm  qos_lh_wm  mn_ttu_vb  qos_l_flp  rc_rd_p_l  rc_rd_l    rc_rd_p_c"
+			"  rc_rd_c    rc_rd_c0   rc_rd_pc0  rc_rd_c1   rc_rd_pc1  qos_lf_l   qos_rds_l"
+			"  qos_lf_c   qos_rds_c  qos_lf_c0  qos_rds_c0 qos_lf_c1  qos_rds_c1\n");
+	for (i = 0; i < pool->pipe_count; i++) {
+		struct dcn_hubp_state *s = &(TO_DCN10_HUBP(pool->hubps[i])->state);
+		struct _vcs_dpi_display_ttu_regs_st *ttu_regs = &s->ttu_attr;
+
+		DTN_INFO("[%2d]:  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh\n",
+			i, ttu_regs->qos_level_low_wm, ttu_regs->qos_level_high_wm, ttu_regs->min_ttu_vblank,
+			ttu_regs->qos_level_flip, ttu_regs->refcyc_per_req_delivery_pre_l, ttu_regs->refcyc_per_req_delivery_l,
+			ttu_regs->refcyc_per_req_delivery_pre_c, ttu_regs->refcyc_per_req_delivery_c, ttu_regs->refcyc_per_req_delivery_cur0,
+			ttu_regs->refcyc_per_req_delivery_pre_cur0, ttu_regs->refcyc_per_req_delivery_cur1,
+			ttu_regs->refcyc_per_req_delivery_pre_cur1, ttu_regs->qos_level_fixed_l, ttu_regs->qos_ramp_disable_l,
+			ttu_regs->qos_level_fixed_c, ttu_regs->qos_ramp_disable_c, ttu_regs->qos_level_fixed_cur0,
+			ttu_regs->qos_ramp_disable_cur0, ttu_regs->qos_level_fixed_cur1, ttu_regs->qos_ramp_disable_cur1);
+	}
 	DTN_INFO("\n");
+}
+
+void dcn10_log_hw_state(struct dc *dc)
+{
+	struct dc_context *dc_ctx = dc->ctx;
+	struct resource_pool *pool = dc->res_pool;
+	int i;
+
+	DTN_INFO_BEGIN();
+
+	dcn10_log_hubbub_state(dc);
+
+	dcn10_log_hubp_states(dc);
 
 	DTN_INFO("DPP:    IGAM format  IGAM mode    DGAM mode    RGAM mode"
 			"  GAMUT mode  C11 C12   C13 C14   C21 C22   C23 C24   "
@@ -340,19 +324,6 @@ void dcn10_log_hw_state(struct dc *dc)
 	}
 	DTN_INFO("\n");
 
-	for (i = 0; i < pool->pipe_count; i++) {
-		struct hubp *hubp = pool->hubps[i];
-		struct dcn_hubp_state s = {0};
-
-		if (!dc->current_state->res_ctx.pipe_ctx[i].stream)
-			continue;
-
-		hubp->funcs->hubp_read_state(hubp, &s);
-		DTN_INFO("RQ-DLG-TTU registers for HUBP%d:\n", i);
-		print_rq_dlg_ttu_regs(dc_ctx, &s);
-		DTN_INFO("\n");
-	}
-
 	DTN_INFO("\nCALCULATED Clocks: dcfclk_khz:%d  dcfclk_deep_sleep_khz:%d  dispclk_khz:%d\n"
 		"dppclk_khz:%d  max_supported_dppclk_khz:%d  fclk_khz:%d  socclk_khz:%d\n\n",
 			dc->current_state->bw.dcn.calc_clk.dcfclk_khz,
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h b/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h
index 3866147..331f8ff 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h
@@ -56,26 +56,6 @@ struct hubp {
 	bool power_gated;
 };
 
-struct dcn_hubp_state {
-	struct _vcs_dpi_display_dlg_regs_st dlg_attr;
-	struct _vcs_dpi_display_ttu_regs_st ttu_attr;
-	struct _vcs_dpi_display_rq_regs_st rq_regs;
-	uint32_t pixel_format;
-	uint32_t inuse_addr_hi;
-	uint32_t viewport_width;
-	uint32_t viewport_height;
-	uint32_t rotation_angle;
-	uint32_t h_mirror_en;
-	uint32_t sw_mode;
-	uint32_t dcc_en;
-	uint32_t blank_en;
-	uint32_t underflow_status;
-	uint32_t ttu_disable;
-	uint32_t min_ttu_vblank;
-	uint32_t qos_level_low_wm;
-	uint32_t qos_level_high_wm;
-};
-
 struct hubp_funcs {
 	void (*hubp_setup)(
 			struct hubp *hubp,
@@ -140,7 +120,7 @@ struct hubp_funcs {
 
 	void (*hubp_clk_cntl)(struct hubp *hubp, bool enable);
 	void (*hubp_vtg_sel)(struct hubp *hubp, uint32_t otg_inst);
-	void (*hubp_read_state)(struct hubp *hubp, struct dcn_hubp_state *s);
+	void (*hubp_read_state)(struct hubp *hubp);
 
 };
 
-- 
cgit v1.1


From 8ec06a179613f7b7379daf78844afe1cc301e2c7 Mon Sep 17 00:00:00 2001
From: Julian Parkin <jparkin@amd.com>
Date: Fri, 13 Apr 2018 13:23:02 -0400
Subject: drm/amd/display: Add assert that chroma pitch is non zero

Signed-off-by: Julian Parkin <jparkin@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
index 159bebc..0cbc83e 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
@@ -146,6 +146,9 @@ void hubp1_program_size_and_rotation(
 	 * 444 or 420 luma
 	 */
 	if (format >= SURFACE_PIXEL_FORMAT_VIDEO_BEGIN) {
+		ASSERT(plane_size->video.chroma_pitch != 0);
+		/* Chroma pitch zero can cause system hang! */
+
 		pitch = plane_size->video.luma_pitch - 1;
 		meta_pitch = dcc->video.meta_pitch_l - 1;
 		pitch_c = plane_size->video.chroma_pitch - 1;
-- 
cgit v1.1


From 7ab3fdde04218c4733e96712b651751c413d51c3 Mon Sep 17 00:00:00 2001
From: "Jerry (Fangzhi) Zuo" <Jerry.Zuo@amd.com>
Date: Tue, 17 Apr 2018 13:49:48 -0400
Subject: drm/amd/display: Update MST edid property every time

Extended fix to: "Don't read EDID in atomic_check"

Fix display property not observed in GUI display after hot plug.

Call drm_mode_connector_update_edid_property every time in
.get_modes hook, due to the fact that edid property is getting
removed from usermode ioctl DRM_IOCTL_MODE_GETCONNECTOR each time
in hot unplug.

Signed-off-by: Jerry (Fangzhi) Zuo <Jerry.Zuo@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
index 305292a..8c1d084 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
@@ -253,11 +253,11 @@ static int dm_dp_mst_get_modes(struct drm_connector *connector)
 		if (aconnector->dc_sink)
 			amdgpu_dm_add_sink_to_freesync_module(
 					connector, edid);
-
-		drm_mode_connector_update_edid_property(
-						&aconnector->base, edid);
 	}
 
+	drm_mode_connector_update_edid_property(
+					&aconnector->base, aconnector->edid);
+
 	ret = drm_add_edid_modes(connector, aconnector->edid);
 
 	return ret;
-- 
cgit v1.1


From 6b622181230e7c9286f594e4e51266b6f019b031 Mon Sep 17 00:00:00 2001
From: Julian Parkin <jparkin@amd.com>
Date: Tue, 17 Apr 2018 11:49:06 -0400
Subject: drm/amd/display: reprogram infoframe during apply_ctx_to_hw

To ensure the infoframe gets updated during an SDR/HDR switch
this change adds a new function to to check if the HDR static
metadata has changed and adds it to is_timing_changed and
pipe_need_reprogram checks

Signed-off-by: Julian Parkin <jparkin@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_resource.c | 18 ++++++++++++++++++
 1 file changed, 18 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
index 447729c..4de1b44 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
@@ -1310,6 +1310,19 @@ bool dc_add_all_planes_for_stream(
 }
 
 
+static bool is_hdr_static_meta_changed(struct dc_stream_state *cur_stream,
+	struct dc_stream_state *new_stream)
+{
+	if (cur_stream == NULL)
+		return true;
+
+	if (memcmp(&cur_stream->hdr_static_metadata,
+			&new_stream->hdr_static_metadata,
+			sizeof(struct dc_info_packet)) != 0)
+		return true;
+
+	return false;
+}
 
 static bool is_timing_changed(struct dc_stream_state *cur_stream,
 		struct dc_stream_state *new_stream)
@@ -1345,6 +1358,9 @@ static bool are_stream_backends_same(
 	if (is_timing_changed(stream_a, stream_b))
 		return false;
 
+	if (is_hdr_static_meta_changed(stream_a, stream_b))
+		return false;
+
 	return true;
 }
 
@@ -2548,6 +2564,8 @@ bool pipe_need_reprogram(
 	if (is_timing_changed(pipe_ctx_old->stream, pipe_ctx->stream))
 		return true;
 
+	if (is_hdr_static_meta_changed(pipe_ctx_old->stream, pipe_ctx->stream))
+		return true;
 
 	return false;
 }
-- 
cgit v1.1


From 70ee2def89e678940190b2c2f7d65fdef5647e07 Mon Sep 17 00:00:00 2001
From: "Jerry (Fangzhi) Zuo" <Jerry.Zuo@amd.com>
Date: Wed, 9 May 2018 14:15:16 -0500
Subject: drm/amd/display: Check dc_sink every time in MST hotplug

Extended fix to: "Don't read EDID in atomic_check"

Fix issue of missing dc_sink in .mode_valid in hot plug routine.

Need to check dc_sink everytime in .get_modes hook after checking
edid, since edid is not getting removed in hot unplug but dc_sink
doesn't.

Signed-off-by: Jerry (Fangzhi) Zuo <Jerry.Zuo@amd.com>
Reviewed-by: Roman Li <Roman.Li@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c  | 16 +++++++++-------
 1 file changed, 9 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
index 8c1d084..ace9ad5 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
@@ -226,10 +226,6 @@ static int dm_dp_mst_get_modes(struct drm_connector *connector)
 
 	if (!aconnector->edid) {
 		struct edid *edid;
-		struct dc_sink *dc_sink;
-		struct dc_sink_init_data init_params = {
-				.link = aconnector->dc_link,
-				.sink_signal = SIGNAL_TYPE_DISPLAY_PORT_MST };
 		edid = drm_dp_mst_get_edid(connector, &aconnector->mst_port->mst_mgr, aconnector->port);
 
 		if (!edid) {
@@ -240,11 +236,17 @@ static int dm_dp_mst_get_modes(struct drm_connector *connector)
 		}
 
 		aconnector->edid = edid;
+	}
 
+	if (!aconnector->dc_sink) {
+		struct dc_sink *dc_sink;
+		struct dc_sink_init_data init_params = {
+				.link = aconnector->dc_link,
+				.sink_signal = SIGNAL_TYPE_DISPLAY_PORT_MST };
 		dc_sink = dc_link_add_remote_sink(
 			aconnector->dc_link,
-			(uint8_t *)edid,
-			(edid->extensions + 1) * EDID_LENGTH,
+			(uint8_t *)aconnector->edid,
+			(aconnector->edid->extensions + 1) * EDID_LENGTH,
 			&init_params);
 
 		dc_sink->priv = aconnector;
@@ -252,7 +254,7 @@ static int dm_dp_mst_get_modes(struct drm_connector *connector)
 
 		if (aconnector->dc_sink)
 			amdgpu_dm_add_sink_to_freesync_module(
-					connector, edid);
+					connector, aconnector->edid);
 	}
 
 	drm_mode_connector_update_edid_property(
-- 
cgit v1.1


From 3bff0796e90960b0abfea21c7f2b50d2d6a19bd9 Mon Sep 17 00:00:00 2001
From: Martin Tsai <Martin.Tsai@amd.com>
Date: Tue, 17 Apr 2018 17:20:06 -0400
Subject: drm/amd/display: to synchronize the hubp and dpp programming in
 cursor control

Signed-off-by: Martin Tsai <Martin.Tsai@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_stream.c | 18 ++++++++++++++++++
 1 file changed, 18 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c
index 3b2ddbd..3732a1d 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c
@@ -178,6 +178,7 @@ bool dc_stream_set_cursor_attributes(
 	int i;
 	struct dc  *core_dc;
 	struct resource_context *res_ctx;
+	struct pipe_ctx *pipe_to_program = NULL;
 
 	if (NULL == stream) {
 		dm_error("DC: dc_stream is NULL!\n");
@@ -205,9 +206,17 @@ bool dc_stream_set_cursor_attributes(
 		if (pipe_ctx->top_pipe && pipe_ctx->plane_state != pipe_ctx->top_pipe->plane_state)
 			continue;
 
+		if (!pipe_to_program) {
+			pipe_to_program = pipe_ctx;
+			core_dc->hwss.pipe_control_lock(core_dc, pipe_to_program, true);
+		}
 
 		core_dc->hwss.set_cursor_attribute(pipe_ctx);
 	}
+
+	if (pipe_to_program)
+		core_dc->hwss.pipe_control_lock(core_dc, pipe_to_program, false);
+
 	return true;
 }
 
@@ -218,6 +227,7 @@ bool dc_stream_set_cursor_position(
 	int i;
 	struct dc  *core_dc;
 	struct resource_context *res_ctx;
+	struct pipe_ctx *pipe_to_program = NULL;
 
 	if (NULL == stream) {
 		dm_error("DC: dc_stream is NULL!\n");
@@ -243,9 +253,17 @@ bool dc_stream_set_cursor_position(
 				!pipe_ctx->plane_res.ipp)
 			continue;
 
+		if (!pipe_to_program) {
+			pipe_to_program = pipe_ctx;
+			core_dc->hwss.pipe_control_lock(core_dc, pipe_to_program, true);
+		}
+
 		core_dc->hwss.set_cursor_position(pipe_ctx);
 	}
 
+	if (pipe_to_program)
+		core_dc->hwss.pipe_control_lock(core_dc, pipe_to_program, false);
+
 	return true;
 }
 
-- 
cgit v1.1


From 34ff937029edf708d7db0d450b5f505969b68950 Mon Sep 17 00:00:00 2001
From: Tony Cheng <tony.cheng@amd.com>
Date: Mon, 16 Apr 2018 13:30:41 -0400
Subject: drm/amd/display: dal 3.1.44

Signed-off-by: Tony Cheng <tony.cheng@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dc.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index 9215298..cd4f434 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -38,7 +38,7 @@
 #include "inc/compressor.h"
 #include "dml/display_mode_lib.h"
 
-#define DC_VER "3.1.43"
+#define DC_VER "3.1.44"
 
 #define MAX_SURFACES 3
 #define MAX_STREAMS 6
-- 
cgit v1.1


From 403dc5e8aab7b3d18a625e516698ea49d2be432e Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Tue, 10 Apr 2018 16:06:34 -0400
Subject: drm/amd/display: Use int for calculating vline start

We are not sure these calculations will never need negative numbers. Use
signed integers and warn and cap at 0 if this ever happens.

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c
index 2c5dbec..c734b7f 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c
@@ -96,10 +96,10 @@ static void optc1_disable_stereo(struct timing_generator *optc)
 static uint32_t get_start_vline(struct timing_generator *optc, const struct dc_crtc_timing *dc_crtc_timing)
 {
 	struct dc_crtc_timing patched_crtc_timing;
-	uint32_t vesa_sync_start;
-	uint32_t asic_blank_end;
-	uint32_t interlace_factor;
-	uint32_t vertical_line_start;
+	int vesa_sync_start;
+	int asic_blank_end;
+	int interlace_factor;
+	int vertical_line_start;
 
 	patched_crtc_timing = *dc_crtc_timing;
 	optc1_apply_front_porch_workaround(optc, &patched_crtc_timing);
-- 
cgit v1.1


From ada8ce1530a7d1bf2e37560afa57bad911d36e81 Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Fri, 20 Apr 2018 10:53:50 -0400
Subject: drm/amd/display: Couple formatting fixes

Things such as mis-indent, and space at beginning of line.

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c       | 3 ++-
 drivers/gpu/drm/amd/display/dc/core/dc_resource.c       | 6 +++---
 drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c | 2 +-
 3 files changed, 6 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 96a57be..3be17e2 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -2692,6 +2692,7 @@ static void amdgpu_dm_connector_destroy(struct drm_connector *connector)
 	const struct dc_link *link = aconnector->dc_link;
 	struct amdgpu_device *adev = connector->dev->dev_private;
 	struct amdgpu_display_manager *dm = &adev->dm;
+
 #if defined(CONFIG_BACKLIGHT_CLASS_DEVICE) ||\
 	defined(CONFIG_BACKLIGHT_CLASS_DEVICE_MODULE)
 
@@ -4649,7 +4650,7 @@ static int dm_update_crtcs_state(struct dc *dc,
 		if (aconnector && enable) {
 			// Make sure fake sink is created in plug-in scenario
 			new_con_state = drm_atomic_get_connector_state(state,
- 								    &aconnector->base);
+								    &aconnector->base);
 
 			if (IS_ERR(new_con_state)) {
 				ret = PTR_ERR_OR_ZERO(new_con_state);
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
index 4de1b44..9eb731f 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
@@ -1798,9 +1798,9 @@ enum dc_status dc_validate_global_state(
 		return DC_ERROR_UNEXPECTED;
 
 	if (dc->res_pool->funcs->validate_global) {
-			result = dc->res_pool->funcs->validate_global(dc, new_ctx);
-			if (result != DC_OK)
-				return result;
+		result = dc->res_pool->funcs->validate_global(dc, new_ctx);
+		if (result != DC_OK)
+			return result;
 	}
 
 	for (i = 0; i < new_ctx->stream_count; i++) {
diff --git a/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c b/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c
index 0a47663..00c0a1e 100644
--- a/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c
@@ -430,7 +430,7 @@ static struct stream_encoder *dce112_stream_encoder_create(
 
 	if (!enc110)
 		return NULL;
-	
+
 	dce110_stream_encoder_construct(enc110, ctx, ctx->dc_bios, eng_id,
 					&stream_enc_regs[eng_id],
 					&se_shift, &se_mask);
-- 
cgit v1.1


From 868b83511ba92b24ce5c5d852cf16b1bc07e13a7 Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Fri, 20 Apr 2018 11:05:07 -0400
Subject: drm/amd/display: Add VG12 ASIC IDs

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/include/dal_asic_id.h | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/include/dal_asic_id.h b/drivers/gpu/drm/amd/display/include/dal_asic_id.h
index 3e8e535..1b987b6 100644
--- a/drivers/gpu/drm/amd/display/include/dal_asic_id.h
+++ b/drivers/gpu/drm/amd/display/include/dal_asic_id.h
@@ -120,9 +120,14 @@
 
 #define AI_GREENLAND_P_A0 1
 #define AI_GREENLAND_P_A1 2
+#define AI_UNKNOWN 0xFF
 
-#define ASICREV_IS_GREENLAND_M(eChipRev)  (eChipRev < AI_UNKNOWN)
-#define ASICREV_IS_GREENLAND_P(eChipRev)  (eChipRev < AI_UNKNOWN)
+#define AI_VEGA12_P_A0 20
+#define ASICREV_IS_GREENLAND_M(eChipRev)  (eChipRev < AI_VEGA12_P_A0)
+#define ASICREV_IS_GREENLAND_P(eChipRev)  (eChipRev < AI_VEGA12_P_A0)
+
+#define ASICREV_IS_VEGA12_P(eChipRev) ((eChipRev >= AI_VEGA12_P_A0) && (eChipRev < AI_UNKNOWN))
+#define ASICREV_IS_VEGA12_p(eChipRev) ((eChipRev >= AI_VEGA12_P_A0) && (eChipRev < AI_UNKNOWN))
 
 /* DCN1_0 */
 #define INTERNAL_REV_RAVEN_A0             0x00    /* First spin of Raven */
-- 
cgit v1.1


From 3aabfcd70077743a5943acd86e70604945e384c1 Mon Sep 17 00:00:00 2001
From: "Jerry (Fangzhi) Zuo" <Jerry.Zuo@amd.com>
Date: Mon, 5 Mar 2018 14:59:57 -0500
Subject: drm/amd: Add BIOS smu_info v3_3 required struct def.

Signed-off-by: Jerry (Fangzhi) Zuo <Jerry.Zuo@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/include/atomfirmware.h | 170 ++++++++++++++++++++++++++++-
 1 file changed, 168 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/include/atomfirmware.h b/drivers/gpu/drm/amd/include/atomfirmware.h
index 0f5ad54..de177ce 100644
--- a/drivers/gpu/drm/amd/include/atomfirmware.h
+++ b/drivers/gpu/drm/amd/include/atomfirmware.h
@@ -501,6 +501,32 @@ enum atom_cooling_solution_id{
   LIQUID_COOLING = 0x01
 };
 
+struct atom_firmware_info_v3_2 {
+  struct atom_common_table_header table_header;
+  uint32_t firmware_revision;
+  uint32_t bootup_sclk_in10khz;
+  uint32_t bootup_mclk_in10khz;
+  uint32_t firmware_capability;             // enum atombios_firmware_capability
+  uint32_t main_call_parser_entry;          /* direct address of main parser call in VBIOS binary. */
+  uint32_t bios_scratch_reg_startaddr;      // 1st bios scratch register dword address
+  uint16_t bootup_vddc_mv;
+  uint16_t bootup_vddci_mv;
+  uint16_t bootup_mvddc_mv;
+  uint16_t bootup_vddgfx_mv;
+  uint8_t  mem_module_id;
+  uint8_t  coolingsolution_id;              /*0: Air cooling; 1: Liquid cooling ... */
+  uint8_t  reserved1[2];
+  uint32_t mc_baseaddr_high;
+  uint32_t mc_baseaddr_low;
+  uint8_t  board_i2c_feature_id;            // enum of atom_board_i2c_feature_id_def
+  uint8_t  board_i2c_feature_gpio_id;       // i2c id find in gpio_lut data table gpio_id
+  uint8_t  board_i2c_feature_slave_addr;
+  uint8_t  reserved3;
+  uint16_t bootup_mvddq_mv;
+  uint16_t bootup_mvpp_mv;
+  uint32_t zfbstartaddrin16mb;
+  uint32_t reserved2[3];
+};
 
 /* 
   ***************************************************************************
@@ -1169,7 +1195,29 @@ struct  atom_gfx_info_v2_2
   uint32_t rlc_gpu_timer_refclk; 
 };
 
-
+struct  atom_gfx_info_v2_3 {
+  struct  atom_common_table_header  table_header;
+  uint8_t gfxip_min_ver;
+  uint8_t gfxip_max_ver;
+  uint8_t max_shader_engines;
+  uint8_t max_tile_pipes;
+  uint8_t max_cu_per_sh;
+  uint8_t max_sh_per_se;
+  uint8_t max_backends_per_se;
+  uint8_t max_texture_channel_caches;
+  uint32_t regaddr_cp_dma_src_addr;
+  uint32_t regaddr_cp_dma_src_addr_hi;
+  uint32_t regaddr_cp_dma_dst_addr;
+  uint32_t regaddr_cp_dma_dst_addr_hi;
+  uint32_t regaddr_cp_dma_command;
+  uint32_t regaddr_cp_status;
+  uint32_t regaddr_rlc_gpu_clock_32;
+  uint32_t rlc_gpu_timer_refclk;
+  uint8_t active_cu_per_sh;
+  uint8_t active_rb_per_se;
+  uint16_t gcgoldenoffset;
+  uint32_t rm21_sram_vmin_value;
+};
 
 /* 
   ***************************************************************************
@@ -1198,6 +1246,76 @@ struct atom_smu_info_v3_1
   uint8_t  fw_ctf_polarity;         // GPIO polarity for CTF
 };
 
+struct atom_smu_info_v3_2 {
+  struct   atom_common_table_header  table_header;
+  uint8_t  smuip_min_ver;
+  uint8_t  smuip_max_ver;
+  uint8_t  smu_rsd1;
+  uint8_t  gpuclk_ss_mode;
+  uint16_t sclk_ss_percentage;
+  uint16_t sclk_ss_rate_10hz;
+  uint16_t gpuclk_ss_percentage;    // in unit of 0.001%
+  uint16_t gpuclk_ss_rate_10hz;
+  uint32_t core_refclk_10khz;
+  uint8_t  ac_dc_gpio_bit;          // GPIO bit shift in SMU_GPIOPAD_A  configured for AC/DC switching, =0xff means invalid
+  uint8_t  ac_dc_polarity;          // GPIO polarity for AC/DC switching
+  uint8_t  vr0hot_gpio_bit;         // GPIO bit shift in SMU_GPIOPAD_A  configured for VR0 HOT event, =0xff means invalid
+  uint8_t  vr0hot_polarity;         // GPIO polarity for VR0 HOT event
+  uint8_t  vr1hot_gpio_bit;         // GPIO bit shift in SMU_GPIOPAD_A configured for VR1 HOT event , =0xff means invalid
+  uint8_t  vr1hot_polarity;         // GPIO polarity for VR1 HOT event
+  uint8_t  fw_ctf_gpio_bit;         // GPIO bit shift in SMU_GPIOPAD_A configured for CTF, =0xff means invalid
+  uint8_t  fw_ctf_polarity;         // GPIO polarity for CTF
+  uint8_t  pcc_gpio_bit;            // GPIO bit shift in SMU_GPIOPAD_A configured for PCC, =0xff means invalid
+  uint8_t  pcc_gpio_polarity;       // GPIO polarity for CTF
+  uint16_t smugoldenoffset;
+  uint32_t gpupll_vco_freq_10khz;
+  uint32_t bootup_smnclk_10khz;
+  uint32_t bootup_socclk_10khz;
+  uint32_t bootup_mp0clk_10khz;
+  uint32_t bootup_mp1clk_10khz;
+  uint32_t bootup_lclk_10khz;
+  uint32_t bootup_dcefclk_10khz;
+  uint32_t ctf_threshold_override_value;
+  uint32_t reserved[5];
+};
+
+struct atom_smu_info_v3_3 {
+  struct   atom_common_table_header  table_header;
+  uint8_t  smuip_min_ver;
+  uint8_t  smuip_max_ver;
+  uint8_t  smu_rsd1;
+  uint8_t  gpuclk_ss_mode;
+  uint16_t sclk_ss_percentage;
+  uint16_t sclk_ss_rate_10hz;
+  uint16_t gpuclk_ss_percentage;    // in unit of 0.001%
+  uint16_t gpuclk_ss_rate_10hz;
+  uint32_t core_refclk_10khz;
+  uint8_t  ac_dc_gpio_bit;          // GPIO bit shift in SMU_GPIOPAD_A  configured for AC/DC switching, =0xff means invalid
+  uint8_t  ac_dc_polarity;          // GPIO polarity for AC/DC switching
+  uint8_t  vr0hot_gpio_bit;         // GPIO bit shift in SMU_GPIOPAD_A  configured for VR0 HOT event, =0xff means invalid
+  uint8_t  vr0hot_polarity;         // GPIO polarity for VR0 HOT event
+  uint8_t  vr1hot_gpio_bit;         // GPIO bit shift in SMU_GPIOPAD_A configured for VR1 HOT event , =0xff means invalid
+  uint8_t  vr1hot_polarity;         // GPIO polarity for VR1 HOT event
+  uint8_t  fw_ctf_gpio_bit;         // GPIO bit shift in SMU_GPIOPAD_A configured for CTF, =0xff means invalid
+  uint8_t  fw_ctf_polarity;         // GPIO polarity for CTF
+  uint8_t  pcc_gpio_bit;            // GPIO bit shift in SMU_GPIOPAD_A configured for PCC, =0xff means invalid
+  uint8_t  pcc_gpio_polarity;       // GPIO polarity for CTF
+  uint16_t smugoldenoffset;
+  uint32_t gpupll_vco_freq_10khz;
+  uint32_t bootup_smnclk_10khz;
+  uint32_t bootup_socclk_10khz;
+  uint32_t bootup_mp0clk_10khz;
+  uint32_t bootup_mp1clk_10khz;
+  uint32_t bootup_lclk_10khz;
+  uint32_t bootup_dcefclk_10khz;
+  uint32_t ctf_threshold_override_value;
+  uint32_t syspll3_0_vco_freq_10khz;
+  uint32_t syspll3_1_vco_freq_10khz;
+  uint32_t bootup_fclk_10khz;
+  uint32_t bootup_waflclk_10khz;
+  uint32_t reserved[3];
+};
+
 /*
  ***************************************************************************
    Data Table smc_dpm_info  structure
@@ -1283,7 +1401,6 @@ struct atom_smc_dpm_info_v4_1
 	uint32_t boardreserved[10];
 };
 
-
 /* 
   ***************************************************************************
     Data Table asic_profiling_info  structure
@@ -1864,6 +1981,55 @@ enum atom_smu9_syspll0_clock_id
   SMU9_SYSPLL0_DISPCLK_ID  = 11,      //	DISPCLK
 };
 
+enum atom_smu11_syspll_id {
+  SMU11_SYSPLL0_ID            = 0,
+  SMU11_SYSPLL1_0_ID          = 1,
+  SMU11_SYSPLL1_1_ID          = 2,
+  SMU11_SYSPLL1_2_ID          = 3,
+  SMU11_SYSPLL2_ID            = 4,
+  SMU11_SYSPLL3_0_ID          = 5,
+  SMU11_SYSPLL3_1_ID          = 6,
+};
+
+
+enum atom_smu11_syspll0_clock_id {
+  SMU11_SYSPLL0_SOCCLK_ID   = 0,       //	SOCCLK
+  SMU11_SYSPLL0_MP0CLK_ID   = 1,       //	MP0CLK
+  SMU11_SYSPLL0_DCLK_ID     = 2,       //	DCLK
+  SMU11_SYSPLL0_VCLK_ID     = 3,       //	VCLK
+  SMU11_SYSPLL0_ECLK_ID     = 4,       //	ECLK
+  SMU11_SYSPLL0_DCEFCLK_ID  = 5,       //	DCEFCLK
+};
+
+
+enum atom_smu11_syspll1_0_clock_id {
+  SMU11_SYSPLL1_0_UCLKA_ID   = 0,       // UCLK_a
+};
+
+enum atom_smu11_syspll1_1_clock_id {
+  SMU11_SYSPLL1_0_UCLKB_ID   = 0,       // UCLK_b
+};
+
+enum atom_smu11_syspll1_2_clock_id {
+  SMU11_SYSPLL1_0_FCLK_ID   = 0,        // FCLK
+};
+
+enum atom_smu11_syspll2_clock_id {
+  SMU11_SYSPLL2_GFXCLK_ID   = 0,        // GFXCLK
+};
+
+enum atom_smu11_syspll3_0_clock_id {
+  SMU11_SYSPLL3_0_WAFCLK_ID = 0,       //	WAFCLK
+  SMU11_SYSPLL3_0_DISPCLK_ID = 1,      //	DISPCLK
+  SMU11_SYSPLL3_0_DPREFCLK_ID = 2,     //	DPREFCLK
+};
+
+enum atom_smu11_syspll3_1_clock_id {
+  SMU11_SYSPLL3_1_MP1CLK_ID = 0,       //	MP1CLK
+  SMU11_SYSPLL3_1_SMNCLK_ID = 1,       //	SMNCLK
+  SMU11_SYSPLL3_1_LCLK_ID = 2,         //	LCLK
+};
+
 struct  atom_get_smu_clock_info_output_parameters_v3_1
 {
   union {
-- 
cgit v1.1


From d66057830c12855253a5c2d246ef26a66b43e2d5 Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Fri, 20 Apr 2018 10:56:18 -0400
Subject: drm/amd/display: Add get_firmware_info_v3_2 for VG12

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c | 86 +++++++++++++++++++++-
 1 file changed, 85 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
index 985fe8c..10a5807 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
@@ -70,6 +70,10 @@ static enum bp_result get_firmware_info_v3_1(
 	struct bios_parser *bp,
 	struct dc_firmware_info *info);
 
+static enum bp_result get_firmware_info_v3_2(
+	struct bios_parser *bp,
+	struct dc_firmware_info *info);
+
 static struct atom_hpd_int_record *get_hpd_record(struct bios_parser *bp,
 		struct atom_display_object_path_v2 *object);
 
@@ -1321,9 +1325,11 @@ static enum bp_result bios_parser_get_firmware_info(
 		case 3:
 			switch (revision.minor) {
 			case 1:
-			case 2:
 				result = get_firmware_info_v3_1(bp, info);
 				break;
+			case 2:
+				result = get_firmware_info_v3_2(bp, info);
+				break;
 			default:
 				break;
 			}
@@ -1383,6 +1389,84 @@ static enum bp_result get_firmware_info_v3_1(
 	return BP_RESULT_OK;
 }
 
+static enum bp_result get_firmware_info_v3_2(
+	struct bios_parser *bp,
+	struct dc_firmware_info *info)
+{
+	struct atom_firmware_info_v3_2 *firmware_info;
+	struct atom_display_controller_info_v4_1 *dce_info = NULL;
+	struct atom_common_table_header *header;
+	struct atom_data_revision revision;
+	struct atom_smu_info_v3_2 *smu_info_v3_2 = NULL;
+	struct atom_smu_info_v3_3 *smu_info_v3_3 = NULL;
+
+	if (!info)
+		return BP_RESULT_BADINPUT;
+
+	firmware_info = GET_IMAGE(struct atom_firmware_info_v3_2,
+			DATA_TABLES(firmwareinfo));
+
+	dce_info = GET_IMAGE(struct atom_display_controller_info_v4_1,
+			DATA_TABLES(dce_info));
+
+	if (!firmware_info || !dce_info)
+		return BP_RESULT_BADBIOSTABLE;
+
+	memset(info, 0, sizeof(*info));
+
+	header = GET_IMAGE(struct atom_common_table_header,
+					DATA_TABLES(smu_info));
+	get_atom_data_table_revision(header, &revision);
+
+	if (revision.minor == 2) {
+		/* Vega12 */
+		smu_info_v3_2 = GET_IMAGE(struct atom_smu_info_v3_2,
+							DATA_TABLES(smu_info));
+
+		if (!smu_info_v3_2)
+			return BP_RESULT_BADBIOSTABLE;
+
+		info->default_engine_clk = smu_info_v3_2->bootup_dcefclk_10khz * 10;
+	} else if (revision.minor == 3) {
+		/* Vega20 */
+		smu_info_v3_3 = GET_IMAGE(struct atom_smu_info_v3_3,
+							DATA_TABLES(smu_info));
+
+		if (!smu_info_v3_3)
+			return BP_RESULT_BADBIOSTABLE;
+
+		info->default_engine_clk = smu_info_v3_3->bootup_dcefclk_10khz * 10;
+	}
+
+	 // We need to convert from 10KHz units into KHz units.
+	info->default_memory_clk = firmware_info->bootup_mclk_in10khz * 10;
+
+	 /* 27MHz for Vega10 & Vega12; 100MHz for Vega20 */
+	info->pll_info.crystal_frequency = dce_info->dce_refclk_10khz * 10;
+	/* Hardcode frequency if BIOS gives no DCE Ref Clk */
+	if (info->pll_info.crystal_frequency == 0) {
+		if (revision.minor == 2)
+			info->pll_info.crystal_frequency = 27000;
+		else if (revision.minor == 3)
+			info->pll_info.crystal_frequency = 100000;
+	}
+	/*dp_phy_ref_clk is not correct for atom_display_controller_info_v4_2, but we don't use it*/
+	info->dp_phy_ref_clk     = dce_info->dpphy_refclk_10khz * 10;
+	info->i2c_engine_ref_clk = dce_info->i2c_engine_refclk_10khz * 10;
+
+	/* Get GPU PLL VCO Clock */
+	if (bp->cmd_tbl.get_smu_clock_info != NULL) {
+		if (revision.minor == 2)
+			info->smu_gpu_pll_output_freq =
+					bp->cmd_tbl.get_smu_clock_info(bp, SMU9_SYSPLL0_ID) * 10;
+		else if (revision.minor == 3)
+			info->smu_gpu_pll_output_freq =
+					bp->cmd_tbl.get_smu_clock_info(bp, SMU11_SYSPLL3_0_ID) * 10;
+	}
+
+	return BP_RESULT_OK;
+}
+
 static enum bp_result bios_parser_get_encoder_cap_info(
 	struct dc_bios *dcb,
 	struct graphics_object_id object_id,
-- 
cgit v1.1


From 8a61bc085ffab3071c59efcbeff4044c034e7490 Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Tue, 24 Apr 2018 10:49:20 -0400
Subject: drm/amd/display: Don't return ddc result and read_bytes in same
 return value

The two ranges overlap.

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c  | 20 ++++++++++++--------
 drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c    | 10 +++++++---
 drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h     |  5 +++--
 3 files changed, 22 insertions(+), 13 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
index ace9ad5..4304d9e 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
@@ -83,21 +83,22 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux,
 	enum i2c_mot_mode mot = (msg->request & DP_AUX_I2C_MOT) ?
 		I2C_MOT_TRUE : I2C_MOT_FALSE;
 	enum ddc_result res;
-	ssize_t read_bytes;
+	uint32_t read_bytes = msg->size;
 
 	if (WARN_ON(msg->size > 16))
 		return -E2BIG;
 
 	switch (msg->request & ~DP_AUX_I2C_MOT) {
 	case DP_AUX_NATIVE_READ:
-		read_bytes = dal_ddc_service_read_dpcd_data(
+		res = dal_ddc_service_read_dpcd_data(
 				TO_DM_AUX(aux)->ddc_service,
 				false,
 				I2C_MOT_UNDEF,
 				msg->address,
 				msg->buffer,
-				msg->size);
-		return read_bytes;
+				msg->size,
+				&read_bytes);
+		break;
 	case DP_AUX_NATIVE_WRITE:
 		res = dal_ddc_service_write_dpcd_data(
 				TO_DM_AUX(aux)->ddc_service,
@@ -108,14 +109,15 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux,
 				msg->size);
 		break;
 	case DP_AUX_I2C_READ:
-		read_bytes = dal_ddc_service_read_dpcd_data(
+		res = dal_ddc_service_read_dpcd_data(
 				TO_DM_AUX(aux)->ddc_service,
 				true,
 				mot,
 				msg->address,
 				msg->buffer,
-				msg->size);
-		return read_bytes;
+				msg->size,
+				&read_bytes);
+		break;
 	case DP_AUX_I2C_WRITE:
 		res = dal_ddc_service_write_dpcd_data(
 				TO_DM_AUX(aux)->ddc_service,
@@ -137,7 +139,9 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux,
 		 r == DDC_RESULT_SUCESSFULL);
 #endif
 
-	return msg->size;
+	if (res != DDC_RESULT_SUCESSFULL)
+		return -EIO;
+	return read_bytes;
 }
 
 static enum drm_connector_status
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c
index 49c2fac..ae48d60 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c
@@ -629,13 +629,14 @@ bool dal_ddc_service_query_ddc_data(
 	return ret;
 }
 
-ssize_t dal_ddc_service_read_dpcd_data(
+enum ddc_result dal_ddc_service_read_dpcd_data(
 	struct ddc_service *ddc,
 	bool i2c,
 	enum i2c_mot_mode mot,
 	uint32_t address,
 	uint8_t *data,
-	uint32_t len)
+	uint32_t len,
+	uint32_t *read)
 {
 	struct aux_payload read_payload = {
 		.i2c_over_aux = i2c,
@@ -652,6 +653,8 @@ ssize_t dal_ddc_service_read_dpcd_data(
 		.mot = mot
 	};
 
+	*read = 0;
+
 	if (len > DEFAULT_AUX_MAX_DATA_SIZE) {
 		BREAK_TO_DEBUGGER();
 		return DDC_RESULT_FAILED_INVALID_OPERATION;
@@ -661,7 +664,8 @@ ssize_t dal_ddc_service_read_dpcd_data(
 		ddc->ctx->i2caux,
 		ddc->ddc_pin,
 		&command)) {
-		return (ssize_t)command.payloads->length;
+		*read = command.payloads->length;
+		return DDC_RESULT_SUCESSFULL;
 	}
 
 	return DDC_RESULT_FAILED_OPERATION;
diff --git a/drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h b/drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h
index 090b7a8..30b3a08 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h
@@ -102,13 +102,14 @@ bool dal_ddc_service_query_ddc_data(
 		uint8_t *read_buf,
 		uint32_t read_size);
 
-ssize_t dal_ddc_service_read_dpcd_data(
+enum ddc_result dal_ddc_service_read_dpcd_data(
 		struct ddc_service *ddc,
 		bool i2c,
 		enum i2c_mot_mode mot,
 		uint32_t address,
 		uint8_t *data,
-		uint32_t len);
+		uint32_t len,
+		uint32_t *read);
 
 enum ddc_result dal_ddc_service_write_dpcd_data(
 		struct ddc_service *ddc,
-- 
cgit v1.1


From f3336254263c3b6a7734379084f4873a335f1577 Mon Sep 17 00:00:00 2001
From: Xiaojie Yuan <Xiaojie.Yuan@amd.com>
Date: Thu, 26 Apr 2018 18:50:26 +0800
Subject: drm/amdgpu/uvd7: add emit_reg_write_reg_wait ring callback

Fix the NULL pointer dereference while running amdgpu_test:

[   54.972246] BUG: unable to handle kernel NULL pointer dereference at 0000000000000000
[   54.972265] IP:           (null)
[   54.972273] PGD 0 P4D 0
[   54.972280] Oops: 0010 [#1] SMP PTI
[   54.972288] Modules linked in: amdkfd amd_iommu_v2 amdgpu(OE) chash gpu_sched ttm drm_kms_helper drm i2c_algo_bit fb_sys_fops syscopyarea sysfillrect sysimgblt snd_hda_codec_realtek snd_hda_codec_generic snd_hda_codec_hdmi snd_hda_intel snd_hda_codec snd_hda_core snd_hwdep intel_rapl snd_pcm snd_seq_midi snd_seq_midi_event snd_rawmidi x86_pkg_temp_thermal intel_powerclamp coretemp kvm_intel snd_seq snd_seq_device kvm irqbypass snd_timer crct10dif_pclmul crc32_pclmul ghash_clmulni_intel pcbc snd soundcore joydev input_leds aesni_intel aes_x86_64 crypto_simd glue_helper cryptd idma64 virt_dma mei_me intel_lpss_pci serio_raw intel_cstate intel_rapl_perf shpchp intel_pch_thermal mei mac_hid intel_lpss acpi_pad parport_pc ppdev nfsd lp auth_rpcgss nfs_acl lockd grace sunrpc parport autofs4 hid_generic
[   54.972434]  usbhid mxm_wmi e1000e psmouse ahci hid libahci wmi pinctrl_sunrisepoint video pinctrl_intel
[   54.972457] CPU: 6 PID: 1393 Comm: uvd Tainted: G           OE    4.16.0-rc7-27fb84fda777 #1
[   54.972473] Hardware name: MSI MS-7984/Z170 KRAIT GAMING (MS-7984), BIOS B.80 05/11/2016
[   54.972489] RIP: 0010:          (null)
[   54.972497] RSP: 0018:ffffaea002c8bcc0 EFLAGS: 00010202
[   54.972508] RAX: 0000000000000000 RBX: ffff9d30d3c56f60 RCX: 00000000007c0002
[   54.972522] RDX: 000000000001a6fb RSI: 000000000001a6e9 RDI: ffff9d30d3c56f60
[   54.972536] RBP: ffffaea002c8bd10 R08: 0000000000000002 R09: ffffffffc06977d0
[   54.972550] R10: 0000000000000040 R11: 0000000000000000 R12: 0000000000000002
[   54.972564] R13: ffff9d30d3c5001c R14: ffff9d30d3c50000 R15: 0000000000000006
[   54.972579] FS:  0000000000000000(0000) GS:ffff9d30eed80000(0000) knlGS:0000000000000000
[   54.972594] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[   54.972606] CR2: 0000000000000000 CR3: 00000002dbc0a001 CR4: 00000000003606e0
[   54.972620] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[   54.972634] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[   54.972648] Call Trace:
[   54.972685]  ? gmc_v9_0_emit_flush_gpu_tlb+0x111/0x140 [amdgpu]
[   54.972721]  uvd_v7_0_ring_emit_vm_flush+0x31/0x70 [amdgpu]
[   54.972751]  amdgpu_vm_flush+0x5dc/0x6c0 [amdgpu]
[   54.972787]  ? pp_dpm_powergate_uvd+0x50/0x80 [amdgpu]
[   54.972816]  amdgpu_ib_schedule+0x120/0x4e0 [amdgpu]
[   54.972850]  amdgpu_job_run+0x17b/0x1c0 [amdgpu]
[   54.972861]  drm_sched_main+0x2cc/0x490 [gpu_sched]
[   54.972873]  ? wait_woken+0x80/0x80
[   54.972882]  kthread+0x121/0x140
[   54.972891]  ? drm_sched_job_finish+0xf0/0xf0 [gpu_sched]
[   54.972902]  ? kthread_create_worker_on_cpu+0x70/0x70
[   54.972914]  ret_from_fork+0x35/0x40
[   54.972922] Code:  Bad RIP value.
[   54.972932] RIP:           (null) RSP: ffffaea002c8bcc0
[   54.972943] CR2: 0000000000000000
[   54.972951] ---[ end trace 5feb349263bbf633 ]---

Signed-off-by: Xiaojie Yuan <Xiaojie.Yuan@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
index 280c082..2251db4 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
@@ -1671,6 +1671,7 @@ static const struct amdgpu_ring_funcs uvd_v7_0_ring_vm_funcs = {
 	.end_use = amdgpu_uvd_ring_end_use,
 	.emit_wreg = uvd_v7_0_ring_emit_wreg,
 	.emit_reg_wait = uvd_v7_0_ring_emit_reg_wait,
+	.emit_reg_write_reg_wait = amdgpu_ring_emit_reg_write_reg_wait_helper,
 };
 
 static const struct amdgpu_ring_funcs uvd_v7_0_enc_ring_vm_funcs = {
-- 
cgit v1.1


From f7dbe9186d85dd63df7868d408fea6859281c446 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Michel=20D=C3=A4nzer?= <michel.daenzer@amd.com>
Date: Tue, 17 Apr 2018 12:25:22 +0200
Subject: drm/amd/display: Use kvzalloc for potentially large allocations
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Allocating up to 32 physically contiguous pages can easily fail (and has
failed for me), and isn't necessary anyway.

Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_surface.c   | 14 ++---
 .../drm/amd/display/modules/color/color_gamma.c    | 72 ++++++++++++----------
 2 files changed, 45 insertions(+), 41 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_surface.c b/drivers/gpu/drm/amd/display/dc/core/dc_surface.c
index 9593877..68a71ad 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_surface.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_surface.c
@@ -72,8 +72,8 @@ struct dc_plane_state *dc_create_plane_state(struct dc *dc)
 {
 	struct dc *core_dc = dc;
 
-	struct dc_plane_state *plane_state = kzalloc(sizeof(*plane_state),
-						     GFP_KERNEL);
+	struct dc_plane_state *plane_state = kvzalloc(sizeof(*plane_state),
+						      GFP_KERNEL);
 
 	if (NULL == plane_state)
 		return NULL;
@@ -126,7 +126,7 @@ static void dc_plane_state_free(struct kref *kref)
 {
 	struct dc_plane_state *plane_state = container_of(kref, struct dc_plane_state, refcount);
 	destruct(plane_state);
-	kfree(plane_state);
+	kvfree(plane_state);
 }
 
 void dc_plane_state_release(struct dc_plane_state *plane_state)
@@ -142,7 +142,7 @@ void dc_gamma_retain(struct dc_gamma *gamma)
 static void dc_gamma_free(struct kref *kref)
 {
 	struct dc_gamma *gamma = container_of(kref, struct dc_gamma, refcount);
-	kfree(gamma);
+	kvfree(gamma);
 }
 
 void dc_gamma_release(struct dc_gamma **gamma)
@@ -153,7 +153,7 @@ void dc_gamma_release(struct dc_gamma **gamma)
 
 struct dc_gamma *dc_create_gamma(void)
 {
-	struct dc_gamma *gamma = kzalloc(sizeof(*gamma), GFP_KERNEL);
+	struct dc_gamma *gamma = kvzalloc(sizeof(*gamma), GFP_KERNEL);
 
 	if (gamma == NULL)
 		goto alloc_fail;
@@ -173,7 +173,7 @@ void dc_transfer_func_retain(struct dc_transfer_func *tf)
 static void dc_transfer_func_free(struct kref *kref)
 {
 	struct dc_transfer_func *tf = container_of(kref, struct dc_transfer_func, refcount);
-	kfree(tf);
+	kvfree(tf);
 }
 
 void dc_transfer_func_release(struct dc_transfer_func *tf)
@@ -183,7 +183,7 @@ void dc_transfer_func_release(struct dc_transfer_func *tf)
 
 struct dc_transfer_func *dc_create_transfer_func()
 {
-	struct dc_transfer_func *tf = kzalloc(sizeof(*tf), GFP_KERNEL);
+	struct dc_transfer_func *tf = kvzalloc(sizeof(*tf), GFP_KERNEL);
 
 	if (tf == NULL)
 		goto alloc_fail;
diff --git a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c
index ad0ff50..15e5b72 100644
--- a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c
+++ b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c
@@ -1274,19 +1274,19 @@ bool mod_color_calculate_regamma_params(struct dc_transfer_func *output_tf,
 
 	output_tf->type = TF_TYPE_DISTRIBUTED_POINTS;
 
-	rgb_user = kzalloc(sizeof(*rgb_user) * (ramp->num_entries + _EXTRA_POINTS),
-			   GFP_KERNEL);
+	rgb_user = kvzalloc(sizeof(*rgb_user) * (ramp->num_entries + _EXTRA_POINTS),
+			    GFP_KERNEL);
 	if (!rgb_user)
 		goto rgb_user_alloc_fail;
-	rgb_regamma = kzalloc(sizeof(*rgb_regamma) * (MAX_HW_POINTS + _EXTRA_POINTS),
-			GFP_KERNEL);
+	rgb_regamma = kvzalloc(sizeof(*rgb_regamma) * (MAX_HW_POINTS + _EXTRA_POINTS),
+			       GFP_KERNEL);
 	if (!rgb_regamma)
 		goto rgb_regamma_alloc_fail;
-	axix_x = kzalloc(sizeof(*axix_x) * (ramp->num_entries + 3),
-			 GFP_KERNEL);
+	axix_x = kvzalloc(sizeof(*axix_x) * (ramp->num_entries + 3),
+			  GFP_KERNEL);
 	if (!axix_x)
 		goto axix_x_alloc_fail;
-	coeff = kzalloc(sizeof(*coeff) * (MAX_HW_POINTS + _EXTRA_POINTS), GFP_KERNEL);
+	coeff = kvzalloc(sizeof(*coeff) * (MAX_HW_POINTS + _EXTRA_POINTS), GFP_KERNEL);
 	if (!coeff)
 		goto coeff_alloc_fail;
 
@@ -1338,13 +1338,13 @@ bool mod_color_calculate_regamma_params(struct dc_transfer_func *output_tf,
 
 	ret = true;
 
-	kfree(coeff);
+	kvfree(coeff);
 coeff_alloc_fail:
-	kfree(axix_x);
+	kvfree(axix_x);
 axix_x_alloc_fail:
-	kfree(rgb_regamma);
+	kvfree(rgb_regamma);
 rgb_regamma_alloc_fail:
-	kfree(rgb_user);
+	kvfree(rgb_user);
 rgb_user_alloc_fail:
 	return ret;
 }
@@ -1480,19 +1480,19 @@ bool mod_color_calculate_degamma_params(struct dc_transfer_func *input_tf,
 
 	input_tf->type = TF_TYPE_DISTRIBUTED_POINTS;
 
-	rgb_user = kzalloc(sizeof(*rgb_user) * (ramp->num_entries + _EXTRA_POINTS),
-			   GFP_KERNEL);
+	rgb_user = kvzalloc(sizeof(*rgb_user) * (ramp->num_entries + _EXTRA_POINTS),
+			    GFP_KERNEL);
 	if (!rgb_user)
 		goto rgb_user_alloc_fail;
-	curve = kzalloc(sizeof(*curve) * (MAX_HW_POINTS + _EXTRA_POINTS),
-			GFP_KERNEL);
+	curve = kvzalloc(sizeof(*curve) * (MAX_HW_POINTS + _EXTRA_POINTS),
+			 GFP_KERNEL);
 	if (!curve)
 		goto curve_alloc_fail;
-	axix_x = kzalloc(sizeof(*axix_x) * (ramp->num_entries + _EXTRA_POINTS),
-			 GFP_KERNEL);
+	axix_x = kvzalloc(sizeof(*axix_x) * (ramp->num_entries + _EXTRA_POINTS),
+			  GFP_KERNEL);
 	if (!axix_x)
 		goto axix_x_alloc_fail;
-	coeff = kzalloc(sizeof(*coeff) * (MAX_HW_POINTS + _EXTRA_POINTS), GFP_KERNEL);
+	coeff = kvzalloc(sizeof(*coeff) * (MAX_HW_POINTS + _EXTRA_POINTS), GFP_KERNEL);
 	if (!coeff)
 		goto coeff_alloc_fail;
 
@@ -1534,13 +1534,13 @@ bool mod_color_calculate_degamma_params(struct dc_transfer_func *input_tf,
 
 	ret = true;
 
-	kfree(coeff);
+	kvfree(coeff);
 coeff_alloc_fail:
-	kfree(axix_x);
+	kvfree(axix_x);
 axix_x_alloc_fail:
-	kfree(curve);
+	kvfree(curve);
 curve_alloc_fail:
-	kfree(rgb_user);
+	kvfree(rgb_user);
 rgb_user_alloc_fail:
 
 	return ret;
@@ -1569,8 +1569,9 @@ bool  mod_color_calculate_curve(enum dc_transfer_func_predefined trans,
 		}
 		ret = true;
 	} else if (trans == TRANSFER_FUNCTION_PQ) {
-		rgb_regamma = kzalloc(sizeof(*rgb_regamma) * (MAX_HW_POINTS +
-						_EXTRA_POINTS), GFP_KERNEL);
+		rgb_regamma = kvzalloc(sizeof(*rgb_regamma) *
+				       (MAX_HW_POINTS + _EXTRA_POINTS),
+				       GFP_KERNEL);
 		if (!rgb_regamma)
 			goto rgb_regamma_alloc_fail;
 		points->end_exponent = 7;
@@ -1590,11 +1591,12 @@ bool  mod_color_calculate_curve(enum dc_transfer_func_predefined trans,
 		}
 		ret = true;
 
-		kfree(rgb_regamma);
+		kvfree(rgb_regamma);
 	} else if (trans == TRANSFER_FUNCTION_SRGB ||
 			  trans == TRANSFER_FUNCTION_BT709) {
-		rgb_regamma = kzalloc(sizeof(*rgb_regamma) * (MAX_HW_POINTS +
-						_EXTRA_POINTS), GFP_KERNEL);
+		rgb_regamma = kvzalloc(sizeof(*rgb_regamma) *
+				       (MAX_HW_POINTS + _EXTRA_POINTS),
+				       GFP_KERNEL);
 		if (!rgb_regamma)
 			goto rgb_regamma_alloc_fail;
 		points->end_exponent = 0;
@@ -1612,7 +1614,7 @@ bool  mod_color_calculate_curve(enum dc_transfer_func_predefined trans,
 		}
 		ret = true;
 
-		kfree(rgb_regamma);
+		kvfree(rgb_regamma);
 	}
 rgb_regamma_alloc_fail:
 	return ret;
@@ -1636,8 +1638,9 @@ bool  mod_color_calculate_degamma_curve(enum dc_transfer_func_predefined trans,
 		}
 		ret = true;
 	} else if (trans == TRANSFER_FUNCTION_PQ) {
-		rgb_degamma = kzalloc(sizeof(*rgb_degamma) * (MAX_HW_POINTS +
-						_EXTRA_POINTS), GFP_KERNEL);
+		rgb_degamma = kvzalloc(sizeof(*rgb_degamma) *
+				       (MAX_HW_POINTS +	_EXTRA_POINTS),
+				       GFP_KERNEL);
 		if (!rgb_degamma)
 			goto rgb_degamma_alloc_fail;
 
@@ -1652,11 +1655,12 @@ bool  mod_color_calculate_degamma_curve(enum dc_transfer_func_predefined trans,
 		}
 		ret = true;
 
-		kfree(rgb_degamma);
+		kvfree(rgb_degamma);
 	} else if (trans == TRANSFER_FUNCTION_SRGB ||
 			  trans == TRANSFER_FUNCTION_BT709) {
-		rgb_degamma = kzalloc(sizeof(*rgb_degamma) * (MAX_HW_POINTS +
-						_EXTRA_POINTS), GFP_KERNEL);
+		rgb_degamma = kvzalloc(sizeof(*rgb_degamma) *
+				       (MAX_HW_POINTS + _EXTRA_POINTS),
+				       GFP_KERNEL);
 		if (!rgb_degamma)
 			goto rgb_degamma_alloc_fail;
 
@@ -1670,7 +1674,7 @@ bool  mod_color_calculate_degamma_curve(enum dc_transfer_func_predefined trans,
 		}
 		ret = true;
 
-		kfree(rgb_degamma);
+		kvfree(rgb_degamma);
 	}
 	points->end_exponent = 0;
 	points->x_point_at_y1_red = 1;
-- 
cgit v1.1


From 87ac8fb08bc7c69a39842c73f3a9d06eb73f02cc Mon Sep 17 00:00:00 2001
From: Shirish S <shirish.s@amd.com>
Date: Wed, 25 Apr 2018 14:42:28 +0530
Subject: drm/amd/display: disable FBC on underlay pipe

FBC is not applicable for the underlay pipe,
hence disallow enabling and disabling of the same.

This also fixes the BUG hit of calling sleep in
atomic context.

Signed-off-by: Shirish S <shirish.s@amd.com>
Reviewed-by: Roman Li <Roman.Li@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
index 51c6c70..2288d0a 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
@@ -2747,6 +2747,9 @@ static void dce110_program_front_end_for_pipe(
 	struct dc_plane_state *plane_state = pipe_ctx->plane_state;
 	struct xfm_grph_csc_adjustment adjust;
 	struct out_csc_color_matrix tbl_entry;
+#if defined(CONFIG_DRM_AMD_DC_FBC)
+	unsigned int underlay_idx = dc->res_pool->underlay_pipe_index;
+#endif
 	unsigned int i;
 	DC_LOGGER_INIT();
 	memset(&tbl_entry, 0, sizeof(tbl_entry));
@@ -2788,7 +2791,9 @@ static void dce110_program_front_end_for_pipe(
 	program_scaler(dc, pipe_ctx);
 
 #if defined(CONFIG_DRM_AMD_DC_FBC)
-	if (dc->fbc_compressor && old_pipe->stream) {
+	/* fbc not applicable on Underlay pipe */
+	if (dc->fbc_compressor && old_pipe->stream &&
+	    pipe_ctx->pipe_idx != underlay_idx) {
 		if (plane_state->tiling_info.gfx8.array_mode == DC_ARRAY_LINEAR_GENERAL)
 			dc->fbc_compressor->funcs->disable_fbc(dc->fbc_compressor);
 		else
-- 
cgit v1.1


From 00862edba135c30d65b9af863163fdfa1f502c61 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Michel=20D=C3=A4nzer?= <michel.daenzer@amd.com>
Date: Wed, 25 Apr 2018 17:32:10 +0200
Subject: drm/ttm: Use GFP_TRANSHUGE_LIGHT for allocating huge pages
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

GFP_TRANSHUGE tries very hard to allocate huge pages, which can result
in long delays with high memory pressure. I have observed firefox
freezing for up to around a minute due to this while restic was taking
a full system backup.

Since we don't really need huge pages, use GFP_TRANSHUGE_LIGHT |
__GFP_NORETRY instead, in order to fail quickly when there are no huge
pages available.

Set __GFP_KSWAPD_RECLAIM as well, in order for huge pages to be freed
up in the background if necessary.

With these changes, I'm no longer seeing freezes during a restic backup.

Cc: stable@vger.kernel.org
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/ttm/ttm_page_alloc.c     | 11 ++++++++---
 drivers/gpu/drm/ttm/ttm_page_alloc_dma.c |  3 ++-
 2 files changed, 10 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/ttm/ttm_page_alloc.c b/drivers/gpu/drm/ttm/ttm_page_alloc.c
index f0481b7..06c94e3 100644
--- a/drivers/gpu/drm/ttm/ttm_page_alloc.c
+++ b/drivers/gpu/drm/ttm/ttm_page_alloc.c
@@ -910,7 +910,8 @@ static int ttm_get_pages(struct page **pages, unsigned npages, int flags,
 			while (npages >= HPAGE_PMD_NR) {
 				gfp_t huge_flags = gfp_flags;
 
-				huge_flags |= GFP_TRANSHUGE;
+				huge_flags |= GFP_TRANSHUGE_LIGHT | __GFP_NORETRY |
+					__GFP_KSWAPD_RECLAIM;
 				huge_flags &= ~__GFP_MOVABLE;
 				huge_flags &= ~__GFP_COMP;
 				p = alloc_pages(huge_flags, HPAGE_PMD_ORDER);
@@ -1027,11 +1028,15 @@ int ttm_page_alloc_init(struct ttm_mem_global *glob, unsigned max_pages)
 				  GFP_USER | GFP_DMA32, "uc dma", 0);
 
 	ttm_page_pool_init_locked(&_manager->wc_pool_huge,
-				  GFP_TRANSHUGE	& ~(__GFP_MOVABLE | __GFP_COMP),
+				  (GFP_TRANSHUGE_LIGHT | __GFP_NORETRY |
+				   __GFP_KSWAPD_RECLAIM) &
+				  ~(__GFP_MOVABLE | __GFP_COMP),
 				  "wc huge", order);
 
 	ttm_page_pool_init_locked(&_manager->uc_pool_huge,
-				  GFP_TRANSHUGE	& ~(__GFP_MOVABLE | __GFP_COMP)
+				  (GFP_TRANSHUGE_LIGHT | __GFP_NORETRY |
+				   __GFP_KSWAPD_RECLAIM) &
+				  ~(__GFP_MOVABLE | __GFP_COMP)
 				  , "uc huge", order);
 
 	_manager->options.max_size = max_pages;
diff --git a/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c b/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c
index 8a25d19..f63d99c 100644
--- a/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c
+++ b/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c
@@ -910,7 +910,8 @@ static gfp_t ttm_dma_pool_gfp_flags(struct ttm_dma_tt *ttm_dma, bool huge)
 		gfp_flags |= __GFP_ZERO;
 
 	if (huge) {
-		gfp_flags |= GFP_TRANSHUGE;
+		gfp_flags |= GFP_TRANSHUGE_LIGHT | __GFP_NORETRY |
+			__GFP_KSWAPD_RECLAIM;
 		gfp_flags &= ~__GFP_MOVABLE;
 		gfp_flags &= ~__GFP_COMP;
 	}
-- 
cgit v1.1


From 719a39a1e9b2dfbfb86f17a8da696b714a3b885d Mon Sep 17 00:00:00 2001
From: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Date: Mon, 30 Apr 2018 10:04:42 -0400
Subject: drm/amdgpu: Switch to interruptable wait to recover from ring hang.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

v2:
Use dma_fence_wait instead of dma_fence_wait_timeout(...,MAX_SCHEDULE_TIMEOUT)
Avoid printing error message for ERESTARTSYS

Originally-by: David Panariti <David.Panariti@amd.com>
Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
index eb80edf..6741a62 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
@@ -421,9 +421,11 @@ int amdgpu_ctx_wait_prev_fence(struct amdgpu_ctx *ctx, unsigned ring_id)
 
 	if (other) {
 		signed long r;
-		r = dma_fence_wait_timeout(other, false, MAX_SCHEDULE_TIMEOUT);
+		r = dma_fence_wait(other, true);
 		if (r < 0) {
-			DRM_ERROR("Error (%ld) waiting for fence!\n", r);
+			if (r != -ERESTARTSYS)
+				DRM_ERROR("Error (%ld) waiting for fence!\n", r);
+
 			return r;
 		}
 	}
-- 
cgit v1.1


From fc5a136ddad944d2f909d3ffcde924b7afa792f4 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Fri, 27 Apr 2018 13:46:08 +0800
Subject: drm/amd/pp: Skip fan attributes if fan not present

With powerplay enabled, also need to skip fan attributes
if no fan present.

Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c      | 21 ++++++++-------------
 drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c |  3 +++
 2 files changed, 11 insertions(+), 13 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
index d9802d9..2c82126 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
@@ -1364,19 +1364,14 @@ static umode_t hwmon_attributes_visible(struct kobject *kobj,
 	struct amdgpu_device *adev = dev_get_drvdata(dev);
 	umode_t effective_mode = attr->mode;
 
-	/* handle non-powerplay limitations */
-	if (!adev->powerplay.pp_handle) {
-		/* Skip fan attributes if fan is not present */
-		if (adev->pm.no_fan &&
-		    (attr == &sensor_dev_attr_pwm1.dev_attr.attr ||
-		     attr == &sensor_dev_attr_pwm1_enable.dev_attr.attr ||
-		     attr == &sensor_dev_attr_pwm1_max.dev_attr.attr ||
-		     attr == &sensor_dev_attr_pwm1_min.dev_attr.attr))
-			return 0;
-		/* requires powerplay */
-		if (attr == &sensor_dev_attr_fan1_input.dev_attr.attr)
-			return 0;
-	}
+
+	/* Skip fan attributes if fan is not present */
+	if (adev->pm.no_fan && (attr == &sensor_dev_attr_pwm1.dev_attr.attr ||
+	    attr == &sensor_dev_attr_pwm1_enable.dev_attr.attr ||
+	    attr == &sensor_dev_attr_pwm1_max.dev_attr.attr ||
+	    attr == &sensor_dev_attr_pwm1_min.dev_attr.attr ||
+	    attr == &sensor_dev_attr_fan1_input.dev_attr.attr))
+		return 0;
 
 	/* Skip limit attributes if DPM is not enabled */
 	if (!adev->pm.dpm_enabled &&
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
index eecb118..71b4233 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
@@ -229,6 +229,9 @@ int hwmgr_hw_init(struct pp_hwmgr *hwmgr)
 	if (ret)
 		goto err;
 
+	((struct amdgpu_device *)hwmgr->adev)->pm.no_fan =
+				hwmgr->thermal_controller.fanInfo.bNoFan;
+
 	ret = hwmgr->hwmgr_func->backend_init(hwmgr);
 	if (ret)
 		goto err1;
-- 
cgit v1.1


From 51d45cbc9196b07f3fc66df5dafd3010c04913a3 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Fri, 27 Apr 2018 14:09:30 +0800
Subject: drm/amdgpu: Fix display corruption on CI with dpm enabled

with dpm enabled, need to get active crtcs in dc/no-dc mode.

caused by
'commit ebb649667a31 ("drm/amdgpu: Set pm_display_cfg in non-dc mode")'

Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c | 32 ++++++++++++++++----------------
 1 file changed, 16 insertions(+), 16 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
index 2c82126..b455da4 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
@@ -1878,26 +1878,26 @@ void amdgpu_pm_compute_clocks(struct amdgpu_device *adev)
 			amdgpu_fence_wait_empty(ring);
 	}
 
-	if (!amdgpu_device_has_dc_support(adev)) {
-		mutex_lock(&adev->pm.mutex);
-		amdgpu_dpm_get_active_displays(adev);
-		adev->pm.pm_display_cfg.num_display = adev->pm.dpm.new_active_crtcs;
-		adev->pm.pm_display_cfg.vrefresh = amdgpu_dpm_get_vrefresh(adev);
-		adev->pm.pm_display_cfg.min_vblank_time = amdgpu_dpm_get_vblank_time(adev);
-		/* we have issues with mclk switching with refresh rates over 120 hz on the non-DC code. */
-		if (adev->pm.pm_display_cfg.vrefresh > 120)
-			adev->pm.pm_display_cfg.min_vblank_time = 0;
-		if (adev->powerplay.pp_funcs->display_configuration_change)
-			adev->powerplay.pp_funcs->display_configuration_change(
-							adev->powerplay.pp_handle,
-							&adev->pm.pm_display_cfg);
-		mutex_unlock(&adev->pm.mutex);
-	}
-
 	if (adev->powerplay.pp_funcs->dispatch_tasks) {
+		if (!amdgpu_device_has_dc_support(adev)) {
+			mutex_lock(&adev->pm.mutex);
+			amdgpu_dpm_get_active_displays(adev);
+			adev->pm.pm_display_cfg.num_display = adev->pm.dpm.new_active_crtcs;
+			adev->pm.pm_display_cfg.vrefresh = amdgpu_dpm_get_vrefresh(adev);
+			adev->pm.pm_display_cfg.min_vblank_time = amdgpu_dpm_get_vblank_time(adev);
+			/* we have issues with mclk switching with refresh rates over 120 hz on the non-DC code. */
+			if (adev->pm.pm_display_cfg.vrefresh > 120)
+				adev->pm.pm_display_cfg.min_vblank_time = 0;
+			if (adev->powerplay.pp_funcs->display_configuration_change)
+				adev->powerplay.pp_funcs->display_configuration_change(
+								adev->powerplay.pp_handle,
+								&adev->pm.pm_display_cfg);
+			mutex_unlock(&adev->pm.mutex);
+		}
 		amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_DISPLAY_CONFIG_CHANGE, NULL);
 	} else {
 		mutex_lock(&adev->pm.mutex);
+		amdgpu_dpm_get_active_displays(adev);
 		/* update battery/ac status */
 		if (power_supply_is_system_supplied() > 0)
 			adev->pm.dpm.ac_power = true;
-- 
cgit v1.1


From dfe8a0187c8dde66b3bc52882826b1e53920ad56 Mon Sep 17 00:00:00 2001
From: Tom St Denis <tom.stdenis@amd.com>
Date: Tue, 1 May 2018 10:15:16 -0400
Subject: drm/amd/amdgpu: vcn10 Add callback for emit_reg_write_reg_wait

The callback .emit_reg_write_reg_wait was missing for vcn decode
which resulted in a kernel oops.

Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
index d9a1533..0501746b 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
@@ -1109,6 +1109,7 @@ static const struct amdgpu_ring_funcs vcn_v1_0_dec_ring_vm_funcs = {
 	.end_use = amdgpu_vcn_ring_end_use,
 	.emit_wreg = vcn_v1_0_dec_ring_emit_wreg,
 	.emit_reg_wait = vcn_v1_0_dec_ring_emit_reg_wait,
+	.emit_reg_write_reg_wait = amdgpu_ring_emit_reg_write_reg_wait_helper,
 };
 
 static const struct amdgpu_ring_funcs vcn_v1_0_enc_ring_vm_funcs = {
-- 
cgit v1.1


From 7e4237dbe494f9721463fd1f2d3b9e52ec74930e Mon Sep 17 00:00:00 2001
From: Tom St Denis <tom.stdenis@amd.com>
Date: Wed, 2 May 2018 13:01:36 -0400
Subject: drm/amd/amdgpu: Add some documentation to the debugfs entries

Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c | 193 +++++++++++++++++++++++++++-
 1 file changed, 189 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
index c98e597..f5fb937 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
@@ -28,8 +28,13 @@
 #include <linux/debugfs.h>
 #include "amdgpu.h"
 
-/*
- * Debugfs
+/**
+ * amdgpu_debugfs_add_files - Add simple debugfs entries
+ *
+ * @adev:  Device to attach debugfs entries to
+ * @files:  Array of function callbacks that respond to reads
+ * @nfiles: Number of callbacks to register
+ *
  */
 int amdgpu_debugfs_add_files(struct amdgpu_device *adev,
 			     const struct drm_info_list *files,
@@ -64,7 +69,33 @@ int amdgpu_debugfs_add_files(struct amdgpu_device *adev,
 
 #if defined(CONFIG_DEBUG_FS)
 
-
+/**
+ * amdgpu_debugfs_process_reg_op - Handle MMIO register reads/writes
+ *
+ * @read: True if reading
+ * @f: open file handle
+ * @buf: User buffer to write/read to
+ * @size: Number of bytes to write/read
+ * @pos:  Offset to seek to
+ *
+ * This debugfs entry has special meaning on the offset being sought.
+ * Various bits have different meanings:
+ *
+ * Bit 62:  Indicates a GRBM bank switch is needed
+ * Bit 61:  Indicates a SRBM bank switch is needed (implies bit 62 is
+ * 			zero)
+ * Bits 24..33: The SE or ME selector if needed
+ * Bits 34..43: The SH (or SA) or PIPE selector if needed
+ * Bits 44..53: The INSTANCE (or CU/WGP) or QUEUE selector if needed
+ *
+ * Bit 23:  Indicates that the PM power gating lock should be held
+ * 			This is necessary to read registers that might be
+ * 			unreliable during a power gating transistion.
+ *
+ * The lower bits are the BYTE offset of the register to read.  This
+ * allows reading multiple registers in a single call and having
+ * the returned size reflect that.
+ */
 static int  amdgpu_debugfs_process_reg_op(bool read, struct file *f,
 		char __user *buf, size_t size, loff_t *pos)
 {
@@ -164,19 +195,37 @@ end:
 	return result;
 }
 
-
+/**
+ * amdgpu_debugfs_regs_read - Callback for reading MMIO registers
+ */
 static ssize_t amdgpu_debugfs_regs_read(struct file *f, char __user *buf,
 					size_t size, loff_t *pos)
 {
 	return amdgpu_debugfs_process_reg_op(true, f, buf, size, pos);
 }
 
+/**
+ * amdgpu_debugfs_regs_write - Callback for writing MMIO registers
+ */
 static ssize_t amdgpu_debugfs_regs_write(struct file *f, const char __user *buf,
 					 size_t size, loff_t *pos)
 {
 	return amdgpu_debugfs_process_reg_op(false, f, (char __user *)buf, size, pos);
 }
 
+
+/**
+ * amdgpu_debugfs_regs_pcie_read - Read from a PCIE register
+ *
+ * @f: open file handle
+ * @buf: User buffer to store read data in
+ * @size: Number of bytes to read
+ * @pos:  Offset to seek to
+ *
+ * The lower bits are the BYTE offset of the register to read.  This
+ * allows reading multiple registers in a single call and having
+ * the returned size reflect that.
+ */
 static ssize_t amdgpu_debugfs_regs_pcie_read(struct file *f, char __user *buf,
 					size_t size, loff_t *pos)
 {
@@ -204,6 +253,18 @@ static ssize_t amdgpu_debugfs_regs_pcie_read(struct file *f, char __user *buf,
 	return result;
 }
 
+/**
+ * amdgpu_debugfs_regs_pcie_write - Write to a PCIE register
+ *
+ * @f: open file handle
+ * @buf: User buffer to write data from
+ * @size: Number of bytes to write
+ * @pos:  Offset to seek to
+ *
+ * The lower bits are the BYTE offset of the register to write.  This
+ * allows writing multiple registers in a single call and having
+ * the returned size reflect that.
+ */
 static ssize_t amdgpu_debugfs_regs_pcie_write(struct file *f, const char __user *buf,
 					 size_t size, loff_t *pos)
 {
@@ -232,6 +293,18 @@ static ssize_t amdgpu_debugfs_regs_pcie_write(struct file *f, const char __user
 	return result;
 }
 
+/**
+ * amdgpu_debugfs_regs_didt_read - Read from a DIDT register
+ *
+ * @f: open file handle
+ * @buf: User buffer to store read data in
+ * @size: Number of bytes to read
+ * @pos:  Offset to seek to
+ *
+ * The lower bits are the BYTE offset of the register to read.  This
+ * allows reading multiple registers in a single call and having
+ * the returned size reflect that.
+ */
 static ssize_t amdgpu_debugfs_regs_didt_read(struct file *f, char __user *buf,
 					size_t size, loff_t *pos)
 {
@@ -259,6 +332,18 @@ static ssize_t amdgpu_debugfs_regs_didt_read(struct file *f, char __user *buf,
 	return result;
 }
 
+/**
+ * amdgpu_debugfs_regs_didt_write - Write to a DIDT register
+ *
+ * @f: open file handle
+ * @buf: User buffer to write data from
+ * @size: Number of bytes to write
+ * @pos:  Offset to seek to
+ *
+ * The lower bits are the BYTE offset of the register to write.  This
+ * allows writing multiple registers in a single call and having
+ * the returned size reflect that.
+ */
 static ssize_t amdgpu_debugfs_regs_didt_write(struct file *f, const char __user *buf,
 					 size_t size, loff_t *pos)
 {
@@ -287,6 +372,18 @@ static ssize_t amdgpu_debugfs_regs_didt_write(struct file *f, const char __user
 	return result;
 }
 
+/**
+ * amdgpu_debugfs_regs_smc_read - Read from a SMC register
+ *
+ * @f: open file handle
+ * @buf: User buffer to store read data in
+ * @size: Number of bytes to read
+ * @pos:  Offset to seek to
+ *
+ * The lower bits are the BYTE offset of the register to read.  This
+ * allows reading multiple registers in a single call and having
+ * the returned size reflect that.
+ */
 static ssize_t amdgpu_debugfs_regs_smc_read(struct file *f, char __user *buf,
 					size_t size, loff_t *pos)
 {
@@ -314,6 +411,18 @@ static ssize_t amdgpu_debugfs_regs_smc_read(struct file *f, char __user *buf,
 	return result;
 }
 
+/**
+ * amdgpu_debugfs_regs_smc_write - Write to a SMC register
+ *
+ * @f: open file handle
+ * @buf: User buffer to write data from
+ * @size: Number of bytes to write
+ * @pos:  Offset to seek to
+ *
+ * The lower bits are the BYTE offset of the register to write.  This
+ * allows writing multiple registers in a single call and having
+ * the returned size reflect that.
+ */
 static ssize_t amdgpu_debugfs_regs_smc_write(struct file *f, const char __user *buf,
 					 size_t size, loff_t *pos)
 {
@@ -342,6 +451,20 @@ static ssize_t amdgpu_debugfs_regs_smc_write(struct file *f, const char __user *
 	return result;
 }
 
+/**
+ * amdgpu_debugfs_gca_config_read - Read from gfx config data
+ *
+ * @f: open file handle
+ * @buf: User buffer to store read data in
+ * @size: Number of bytes to read
+ * @pos:  Offset to seek to
+ *
+ * This file is used to access configuration data in a somewhat
+ * stable fashion.  The format is a series of DWORDs with the first
+ * indicating which revision it is.  New content is appended to the
+ * end so that older software can still read the data.
+ */
+
 static ssize_t amdgpu_debugfs_gca_config_read(struct file *f, char __user *buf,
 					size_t size, loff_t *pos)
 {
@@ -418,6 +541,19 @@ static ssize_t amdgpu_debugfs_gca_config_read(struct file *f, char __user *buf,
 	return result;
 }
 
+/**
+ * amdgpu_debugfs_sensor_read - Read from the powerplay sensors
+ *
+ * @f: open file handle
+ * @buf: User buffer to store read data in
+ * @size: Number of bytes to read
+ * @pos:  Offset to seek to
+ *
+ * The offset is treated as the BYTE address of one of the sensors
+ * enumerated in amd/include/kgd_pp_interface.h under the
+ * 'amd_pp_sensors' enumeration.  For instance to read the UVD VCLK
+ * you would use the offset 3 * 4 = 12.
+ */
 static ssize_t amdgpu_debugfs_sensor_read(struct file *f, char __user *buf,
 					size_t size, loff_t *pos)
 {
@@ -457,6 +593,27 @@ static ssize_t amdgpu_debugfs_sensor_read(struct file *f, char __user *buf,
 	return !r ? outsize : r;
 }
 
+/** amdgpu_debugfs_wave_read - Read WAVE STATUS data
+ *
+ * @f: open file handle
+ * @buf: User buffer to store read data in
+ * @size: Number of bytes to read
+ * @pos:  Offset to seek to
+ *
+ * The offset being sought changes which wave that the status data
+ * will be returned for.  The bits are used as follows:
+ *
+ * Bits 0..6: 	Byte offset into data
+ * Bits 7..14:	SE selector
+ * Bits 15..22:	SH/SA selector
+ * Bits 23..30: CU/{WGP+SIMD} selector
+ * Bits 31..36: WAVE ID selector
+ * Bits 37..44: SIMD ID selector
+ *
+ * The returned data begins with one DWORD of version information
+ * Followed by WAVE STATUS registers relevant to the GFX IP version
+ * being used.  See gfx_v8_0_read_wave_data() for an example output.
+ */
 static ssize_t amdgpu_debugfs_wave_read(struct file *f, char __user *buf,
 					size_t size, loff_t *pos)
 {
@@ -507,6 +664,28 @@ static ssize_t amdgpu_debugfs_wave_read(struct file *f, char __user *buf,
 	return result;
 }
 
+/** amdgpu_debugfs_gpr_read - Read wave gprs
+ *
+ * @f: open file handle
+ * @buf: User buffer to store read data in
+ * @size: Number of bytes to read
+ * @pos:  Offset to seek to
+ *
+ * The offset being sought changes which wave that the status data
+ * will be returned for.  The bits are used as follows:
+ *
+ * Bits 0..11:	Byte offset into data
+ * Bits 12..19:	SE selector
+ * Bits 20..27:	SH/SA selector
+ * Bits 28..35: CU/{WGP+SIMD} selector
+ * Bits 36..43: WAVE ID selector
+ * Bits 37..44: SIMD ID selector
+ * Bits 52..59: Thread selector
+ * Bits 60..61: Bank selector (VGPR=0,SGPR=1)
+ *
+ * The return data comes from the SGPR or VGPR register bank for
+ * the selected operational unit.
+ */
 static ssize_t amdgpu_debugfs_gpr_read(struct file *f, char __user *buf,
 					size_t size, loff_t *pos)
 {
@@ -637,6 +816,12 @@ static const char *debugfs_regs_names[] = {
 	"amdgpu_gpr",
 };
 
+/**
+ * amdgpu_debugfs_regs_init -	Initialize debugfs entries that provide
+ * 								register access.
+ *
+ * @adev: The device to attach the debugfs entries to
+ */
 int amdgpu_debugfs_regs_init(struct amdgpu_device *adev)
 {
 	struct drm_minor *minor = adev->ddev->primary;
-- 
cgit v1.1


From 3f4299bee6eda852489ce4fd307dd709a09f5d8f Mon Sep 17 00:00:00 2001
From: Chunming Zhou <david1.zhou@amd.com>
Date: Tue, 24 Apr 2018 12:14:39 +0800
Subject: drm/amdgpu: abstract bo_base init function
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Signed-off-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 73 ++++++++++++++++++----------------
 1 file changed, 38 insertions(+), 35 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index 1c00f1a..71dcdef 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -94,6 +94,36 @@ struct amdgpu_prt_cb {
 	struct dma_fence_cb cb;
 };
 
+static void amdgpu_vm_bo_base_init(struct amdgpu_vm_bo_base *base,
+				   struct amdgpu_vm *vm,
+				   struct amdgpu_bo *bo)
+{
+	base->vm = vm;
+	base->bo = bo;
+	INIT_LIST_HEAD(&base->bo_list);
+	INIT_LIST_HEAD(&base->vm_status);
+
+	if (!bo)
+		return;
+	list_add_tail(&base->bo_list, &bo->va);
+
+	if (bo->tbo.resv != vm->root.base.bo->tbo.resv)
+		return;
+
+	if (bo->preferred_domains &
+	    amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type))
+		return;
+
+	/*
+	 * we checked all the prerequisites, but it looks like this per vm bo
+	 * is currently evicted. add the bo to the evicted list to make sure it
+	 * is validated on next vm use to avoid fault.
+	 * */
+	spin_lock(&vm->status_lock);
+	list_move_tail(&base->vm_status, &vm->evicted);
+	spin_unlock(&vm->status_lock);
+}
+
 /**
  * amdgpu_vm_level_shift - return the addr shift for each level
  *
@@ -446,11 +476,9 @@ static int amdgpu_vm_alloc_levels(struct amdgpu_device *adev,
 			*/
 			pt->parent = amdgpu_bo_ref(parent->base.bo);
 
-			entry->base.vm = vm;
-			entry->base.bo = pt;
-			list_add_tail(&entry->base.bo_list, &pt->va);
+			amdgpu_vm_bo_base_init(&entry->base, vm, pt);
 			spin_lock(&vm->status_lock);
-			list_add(&entry->base.vm_status, &vm->relocated);
+			list_move(&entry->base.vm_status, &vm->relocated);
 			spin_unlock(&vm->status_lock);
 		}
 
@@ -1841,36 +1869,12 @@ struct amdgpu_bo_va *amdgpu_vm_bo_add(struct amdgpu_device *adev,
 	if (bo_va == NULL) {
 		return NULL;
 	}
-	bo_va->base.vm = vm;
-	bo_va->base.bo = bo;
-	INIT_LIST_HEAD(&bo_va->base.bo_list);
-	INIT_LIST_HEAD(&bo_va->base.vm_status);
+	amdgpu_vm_bo_base_init(&bo_va->base, vm, bo);
 
 	bo_va->ref_count = 1;
 	INIT_LIST_HEAD(&bo_va->valids);
 	INIT_LIST_HEAD(&bo_va->invalids);
 
-	if (!bo)
-		return bo_va;
-
-	list_add_tail(&bo_va->base.bo_list, &bo->va);
-
-	if (bo->tbo.resv != vm->root.base.bo->tbo.resv)
-		return bo_va;
-
-	if (bo->preferred_domains &
-	    amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type))
-		return bo_va;
-
-	/*
-	 * We checked all the prerequisites, but it looks like this per VM BO
-	 * is currently evicted. add the BO to the evicted list to make sure it
-	 * is validated on next VM use to avoid fault.
-	 * */
-	spin_lock(&vm->status_lock);
-	list_move_tail(&bo_va->base.vm_status, &vm->evicted);
-	spin_unlock(&vm->status_lock);
-
 	return bo_va;
 }
 
@@ -2370,6 +2374,7 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 		   int vm_context, unsigned int pasid)
 {
 	struct amdgpu_bo_param bp;
+	struct amdgpu_bo *root;
 	const unsigned align = min(AMDGPU_VM_PTB_ALIGN_SIZE,
 		AMDGPU_VM_PTE_COUNT(adev) * 8);
 	unsigned ring_instance;
@@ -2431,23 +2436,21 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 	bp.flags = flags;
 	bp.type = ttm_bo_type_kernel;
 	bp.resv = NULL;
-	r = amdgpu_bo_create(adev, &bp, &vm->root.base.bo);
+	r = amdgpu_bo_create(adev, &bp, &root);
 	if (r)
 		goto error_free_sched_entity;
 
-	r = amdgpu_bo_reserve(vm->root.base.bo, true);
+	r = amdgpu_bo_reserve(root, true);
 	if (r)
 		goto error_free_root;
 
-	r = amdgpu_vm_clear_bo(adev, vm, vm->root.base.bo,
+	r = amdgpu_vm_clear_bo(adev, vm, root,
 			       adev->vm_manager.root_level,
 			       vm->pte_support_ats);
 	if (r)
 		goto error_unreserve;
 
-	vm->root.base.vm = vm;
-	list_add_tail(&vm->root.base.bo_list, &vm->root.base.bo->va);
-	list_add_tail(&vm->root.base.vm_status, &vm->evicted);
+	amdgpu_vm_bo_base_init(&vm->root.base, vm, root);
 	amdgpu_bo_unreserve(vm->root.base.bo);
 
 	if (pasid) {
-- 
cgit v1.1


From 4bebcceededa794a26827d40ab52555c2ec37deb Mon Sep 17 00:00:00 2001
From: Chunming Zhou <david1.zhou@amd.com>
Date: Tue, 24 Apr 2018 13:54:10 +0800
Subject: drm/amdgpu: invalidate parent bo when shadow bo was invalidated
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Shadow BO is located on GTT and its parent (PT and PD) BO could located on VRAM.
In some case, the BO on GTT could be evicted but the parent did not. This may
cause the shadow BO not be put in the evict list and could not be invalidate
correctly.
v2: suggested by Christian

Signed-off-by: Chunming Zhou <david1.zhou@amd.com>
Reported-by: Shaoyun Liu <Shaoyun.Liu@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index 71dcdef..8e71d39 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -2252,6 +2252,10 @@ void amdgpu_vm_bo_invalidate(struct amdgpu_device *adev,
 {
 	struct amdgpu_vm_bo_base *bo_base;
 
+	/* shadow bo doesn't have bo base, its validation needs its parent */
+	if (bo->parent && bo->parent->shadow == bo)
+		bo = bo->parent;
+
 	list_for_each_entry(bo_base, &bo->va, bo_list) {
 		struct amdgpu_vm *vm = bo_base->vm;
 
-- 
cgit v1.1


From 4aa8c41bfb772cd86a5726e374b13d2f31cd9f4c Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Mon, 7 May 2018 14:23:04 +0800
Subject: drm/amd/pp: Refine the output of pp_power_profile_mode on VI
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

In order to keep consist with Vega,
the output format of the pp_power_profile_mode would be
<integer><mode name string>< “*” for current profile>:"detail settings"
and remove the "CURRENT" mode line.

for example:
NUM        MODE_NAME     SCLK_UP_HYST   SCLK_DOWN_HYST SCLK_ACTIVE_LEVEL     MCLK_UP_HYST   MCLK_DOWN_HYST MCLK_ACTIVE_LEVEL
  0   3D_FULL_SCREEN:        0              100               30                0              100               10
  1     POWER_SAVING:       10                0               30                -                -                -
  2            VIDEO:        -                -                -               10               16               31
  3               VR:        0               11               50                0              100               10
  4          COMPUTE:        0                5               30                -                -                -
  5         CUSTOM *:        0                5               30                0              100               10
NUM        MODE_NAME     SCLK_UP_HYST   SCLK_DOWN_HYST SCLK_ACTIVE_LEVEL     MCLK_UP_HYST   MCLK_DOWN_HYST MCLK_ACTIVE_LEVEL
  0   3D_FULL_SCREEN:        0              100               30                0              100               10
  1   POWER_SAVING *:       10                0               30                0              100               10
  2            VIDEO:        -                -                -               10               16               31
  3               VR:        0               11               50                0              100               10
  4          COMPUTE:        0                5               30                -                -                -
  5           CUSTOM:        -                -                -                -                -                -

Reviewed-by: Evan Quan <evan.quan@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c | 52 +++++++++++-------------
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h |  1 -
 2 files changed, 23 insertions(+), 30 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index 4c94e7a..39e49ce 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -75,12 +75,13 @@
 #define PCIE_BUS_CLK                10000
 #define TCLK                        (PCIE_BUS_CLK / 10)
 
-static const struct profile_mode_setting smu7_profiling[5] =
+static const struct profile_mode_setting smu7_profiling[6] =
 					{{1, 0, 100, 30, 1, 0, 100, 10},
 					 {1, 10, 0, 30, 0, 0, 0, 0},
 					 {0, 0, 0, 0, 1, 10, 16, 31},
 					 {1, 0, 11, 50, 1, 0, 100, 10},
 					 {1, 0, 5, 30, 0, 0, 0, 0},
+					 {0, 0, 0, 0, 0, 0, 0, 0},
 					};
 
 #define PPSMC_MSG_SetVBITimeout_VEGAM    ((uint16_t) 0x310)
@@ -4882,6 +4883,17 @@ static int smu7_get_power_profile_mode(struct pp_hwmgr *hwmgr, char *buf)
 	len = sizeof(smu7_profiling) / sizeof(struct profile_mode_setting);
 
 	for (i = 0; i < len; i++) {
+		if (i == hwmgr->power_profile_mode) {
+			size += sprintf(buf + size, "%3d %14s %s: %8d %16d %16d %16d %16d %16d\n",
+			i, profile_name[i], "*",
+			data->current_profile_setting.sclk_up_hyst,
+			data->current_profile_setting.sclk_down_hyst,
+			data->current_profile_setting.sclk_activity,
+			data->current_profile_setting.mclk_up_hyst,
+			data->current_profile_setting.mclk_down_hyst,
+			data->current_profile_setting.mclk_activity);
+			continue;
+		}
 		if (smu7_profiling[i].bupdate_sclk)
 			size += sprintf(buf + size, "%3d %16s: %8d %16d %16d ",
 			i, profile_name[i], smu7_profiling[i].sclk_up_hyst,
@@ -4901,24 +4913,6 @@ static int smu7_get_power_profile_mode(struct pp_hwmgr *hwmgr, char *buf)
 			"-", "-", "-");
 	}
 
-	size += sprintf(buf + size, "%3d %16s: %8d %16d %16d %16d %16d %16d\n",
-			i, profile_name[i],
-			data->custom_profile_setting.sclk_up_hyst,
-			data->custom_profile_setting.sclk_down_hyst,
-			data->custom_profile_setting.sclk_activity,
-			data->custom_profile_setting.mclk_up_hyst,
-			data->custom_profile_setting.mclk_down_hyst,
-			data->custom_profile_setting.mclk_activity);
-
-	size += sprintf(buf + size, "%3s %16s: %8d %16d %16d %16d %16d %16d\n",
-			"*", "CURRENT",
-			data->current_profile_setting.sclk_up_hyst,
-			data->current_profile_setting.sclk_down_hyst,
-			data->current_profile_setting.sclk_activity,
-			data->current_profile_setting.mclk_up_hyst,
-			data->current_profile_setting.mclk_down_hyst,
-			data->current_profile_setting.mclk_activity);
-
 	return size;
 }
 
@@ -4957,16 +4951,16 @@ static int smu7_set_power_profile_mode(struct pp_hwmgr *hwmgr, long *input, uint
 		if (size < 8)
 			return -EINVAL;
 
-		data->custom_profile_setting.bupdate_sclk = input[0];
-		data->custom_profile_setting.sclk_up_hyst = input[1];
-		data->custom_profile_setting.sclk_down_hyst = input[2];
-		data->custom_profile_setting.sclk_activity = input[3];
-		data->custom_profile_setting.bupdate_mclk = input[4];
-		data->custom_profile_setting.mclk_up_hyst = input[5];
-		data->custom_profile_setting.mclk_down_hyst = input[6];
-		data->custom_profile_setting.mclk_activity = input[7];
-		if (!smum_update_dpm_settings(hwmgr, &data->custom_profile_setting)) {
-			memcpy(&data->current_profile_setting, &data->custom_profile_setting, sizeof(struct profile_mode_setting));
+		tmp.bupdate_sclk = input[0];
+		tmp.sclk_up_hyst = input[1];
+		tmp.sclk_down_hyst = input[2];
+		tmp.sclk_activity = input[3];
+		tmp.bupdate_mclk = input[4];
+		tmp.mclk_up_hyst = input[5];
+		tmp.mclk_down_hyst = input[6];
+		tmp.mclk_activity = input[7];
+		if (!smum_update_dpm_settings(hwmgr, &tmp)) {
+			memcpy(&data->current_profile_setting, &tmp, sizeof(struct profile_mode_setting));
 			hwmgr->power_profile_mode = mode;
 		}
 		break;
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h
index 51a776e..c91e75d 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h
@@ -327,7 +327,6 @@ struct smu7_hwmgr {
 	uint16_t                              mem_latency_high;
 	uint16_t                              mem_latency_low;
 	uint32_t                              vr_config;
-	struct profile_mode_setting           custom_profile_setting;
 	struct profile_mode_setting           current_profile_setting;
 };
 
-- 
cgit v1.1


From 3d3c4f1b4d37412b2e3f1dd5e95293e61efa83df Mon Sep 17 00:00:00 2001
From: Colin Ian King <colin.king@canonical.com>
Date: Sat, 28 Apr 2018 23:21:55 +0100
Subject: drm/amd/powerplay: fix spelling mistake: "contruct" -> "construct"

Trivial fix to spelling mistake in PP_ASSERT_WITH_CODE message text

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c   | 2 +-
 drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index 39e49ce..8eb3f51 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -1270,7 +1270,7 @@ static int smu7_enable_dpm_tasks(struct pp_hwmgr *hwmgr)
 
 		tmp_result = smu7_construct_voltage_tables(hwmgr);
 		PP_ASSERT_WITH_CODE((0 == tmp_result),
-				"Failed to contruct voltage tables!",
+				"Failed to construct voltage tables!",
 				result = tmp_result);
 	}
 	smum_initialize_mc_reg_table(hwmgr);
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
index 7486120..d156b7b 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
@@ -2860,7 +2860,7 @@ static int vega10_enable_dpm_tasks(struct pp_hwmgr *hwmgr)
 
 	tmp_result = vega10_construct_voltage_tables(hwmgr);
 	PP_ASSERT_WITH_CODE(!tmp_result,
-			"Failed to contruct voltage tables!",
+			"Failed to construct voltage tables!",
 			result = tmp_result);
 
 	tmp_result = vega10_init_smc_table(hwmgr);
-- 
cgit v1.1


From f4c2cc43218150da670f526aba1eeb3bcec9e3d2 Mon Sep 17 00:00:00 2001
From: Colin Ian King <colin.king@canonical.com>
Date: Wed, 2 May 2018 15:43:16 +0100
Subject: drm/amd/display: clean up assignment of amdgpu_crtc

The declaration of pointer amdgpu_crtc has a redundant assignment to
amdgpu_crtc. Clean this up by removing it.

Detected by CoverityScan, CID#1460299 ("Evaluation order violation")

Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 3be17e2..3e0f385 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -3748,7 +3748,7 @@ static void remove_stream(struct amdgpu_device *adev,
 static int get_cursor_position(struct drm_plane *plane, struct drm_crtc *crtc,
 			       struct dc_cursor_position *position)
 {
-	struct amdgpu_crtc *amdgpu_crtc = amdgpu_crtc = to_amdgpu_crtc(crtc);
+	struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
 	int x, y;
 	int xorigin = 0, yorigin = 0;
 
-- 
cgit v1.1


From 267256b5d884b1494cf30636c66cd95eeb25f41f Mon Sep 17 00:00:00 2001
From: Junwei Zhang <Jerry.Zhang@amd.com>
Date: Wed, 9 May 2018 17:17:58 +0800
Subject: drm/amd/powerplay: add PME smu message for raven

Used for working around an audio bug on some platforms.

Signed-off-by: Junwei Zhang <Jerry.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/inc/rv_ppsmc.h | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/powerplay/inc/rv_ppsmc.h b/drivers/gpu/drm/amd/powerplay/inc/rv_ppsmc.h
index 5d07b6e..a2991fa 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/rv_ppsmc.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/rv_ppsmc.h
@@ -82,7 +82,8 @@
 #define PPSMC_MSG_SetSoftMaxFclkByFreq          0x33
 #define PPSMC_MSG_SetSoftMaxVcn                 0x34
 #define PPSMC_MSG_PowerGateMmHub                0x35
-#define PPSMC_Message_Count                     0x36
+#define PPSMC_MSG_SetRccPfcPmeRestoreRegister   0x36
+#define PPSMC_Message_Count                     0x37
 
 
 typedef uint16_t PPSMC_Result;
-- 
cgit v1.1


From 4ccd2d931c4bbebbca5a5e233f0d28ed57482e90 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Tue, 8 May 2018 14:20:25 +0800
Subject: drm/amd/pp: Implement force_clock_level for RV

under manual dpm mode, user can set gfx/mem clock
through sysfs pp_dpm_sclk/mclk on Rv.

Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c | 45 +++++++++++++++++++++++
 1 file changed, 45 insertions(+)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
index be6d6e2..8b75f52 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
@@ -766,6 +766,51 @@ static int smu10_get_dal_power_level(struct pp_hwmgr *hwmgr,
 static int smu10_force_clock_level(struct pp_hwmgr *hwmgr,
 		enum pp_clock_type type, uint32_t mask)
 {
+	struct smu10_hwmgr *data = hwmgr->backend;
+	struct smu10_voltage_dependency_table *mclk_table =
+					data->clock_vol_info.vdd_dep_on_fclk;
+	uint32_t low, high;
+
+	low = mask ? (ffs(mask) - 1) : 0;
+	high = mask ? (fls(mask) - 1) : 0;
+
+	switch (type) {
+	case PP_SCLK:
+		if (low > 2 || high > 2) {
+			pr_info("Currently sclk only support 3 levels on RV\n");
+			return -EINVAL;
+		}
+
+		smum_send_msg_to_smc_with_parameter(hwmgr,
+						PPSMC_MSG_SetHardMinGfxClk,
+						low == 2 ? data->gfx_max_freq_limit/100 :
+						low == 1 ? SMU10_UMD_PSTATE_GFXCLK :
+						data->gfx_min_freq_limit/100);
+
+		smum_send_msg_to_smc_with_parameter(hwmgr,
+						PPSMC_MSG_SetSoftMaxGfxClk,
+						high == 0 ? data->gfx_min_freq_limit/100 :
+						high == 1 ? SMU10_UMD_PSTATE_GFXCLK :
+						data->gfx_max_freq_limit/100);
+		break;
+
+	case PP_MCLK:
+		if (low > mclk_table->count - 1 || high > mclk_table->count - 1)
+			return -EINVAL;
+
+		smum_send_msg_to_smc_with_parameter(hwmgr,
+						PPSMC_MSG_SetHardMinFclkByFreq,
+						mclk_table->entries[low].clk/100);
+
+		smum_send_msg_to_smc_with_parameter(hwmgr,
+						PPSMC_MSG_SetSoftMaxFclkByFreq,
+						mclk_table->entries[high].clk/100);
+		break;
+
+	case PP_PCIE:
+	default:
+		break;
+	}
 	return 0;
 }
 
-- 
cgit v1.1


From 9164e8b7b32edeea75bf713a61f8bd1701b9a61b Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Thu, 10 May 2018 19:51:09 +0800
Subject: drm/amd/pp: Fix performance drop on Fiji

The performance drop if the default TDP more than 256 Watt

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c
index a264e0c..99b29ff 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c
@@ -1042,12 +1042,10 @@ int smu7_set_power_limit(struct pp_hwmgr *hwmgr, uint32_t n)
 {
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
 
-	n = (n & 0xff) << 8;
-
 	if (data->power_containment_features &
 			POWERCONTAINMENT_FEATURE_PkgPwrLimit)
 		return smum_send_msg_to_smc_with_parameter(hwmgr,
-				PPSMC_MSG_PkgPwrSetLimit, n);
+				PPSMC_MSG_PkgPwrSetLimit, n<<8);
 	return 0;
 }
 
-- 
cgit v1.1


From 967c650d4973dd87355c8540a18e6fa58844e547 Mon Sep 17 00:00:00 2001
From: Junwei Zhang <Jerry.Zhang@amd.com>
Date: Fri, 11 May 2018 10:54:40 +0800
Subject: drm/ttm: remove priority hard code when initializing ttm bo
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Then priority could be set before initialization.
By default, it requires to kzalloc ttm bo. In fact, we always do so.

Signed-off-by: Junwei Zhang <Jerry.Zhang@amd.com>
Reviewed-by: David Zhou <david1.zhou@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/ttm/ttm_bo.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c
index 98e06f8..cba5015 100644
--- a/drivers/gpu/drm/ttm/ttm_bo.c
+++ b/drivers/gpu/drm/ttm/ttm_bo.c
@@ -1175,7 +1175,6 @@ int ttm_bo_init_reserved(struct ttm_bo_device *bdev,
 	reservation_object_init(&bo->ttm_resv);
 	atomic_inc(&bo->bdev->glob->bo_count);
 	drm_vma_node_reset(&bo->vma_node);
-	bo->priority = 0;
 
 	/*
 	 * For ttm_bo_type_device buffers, allocate
-- 
cgit v1.1


From a50cb94819f81bfafb5dc5a605baba9b40ba3243 Mon Sep 17 00:00:00 2001
From: Junwei Zhang <Jerry.Zhang@amd.com>
Date: Fri, 11 May 2018 11:02:23 +0800
Subject: drm/amdgpu: set ttm bo priority before initialization
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Signed-off-by: Junwei Zhang <Jerry.Zhang@amd.com>
Reviewed-by: David Zhou <david1.zhou@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
index e62153a..6a9e46a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
@@ -419,6 +419,8 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev,
 
 	bo->tbo.bdev = &adev->mman.bdev;
 	amdgpu_ttm_placement_from_domain(bo, bp->domain);
+	if (bp->type == ttm_bo_type_kernel)
+		bo->tbo.priority = 1;
 
 	r = ttm_bo_init_reserved(&adev->mman.bdev, &bo->tbo, size, bp->type,
 				 &bo->placement, page_align, &ctx, acc_size,
@@ -434,9 +436,6 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev,
 	else
 		amdgpu_cs_report_moved_bytes(adev, ctx.bytes_moved, 0);
 
-	if (bp->type == ttm_bo_type_kernel)
-		bo->tbo.priority = 1;
-
 	if (bp->flags & AMDGPU_GEM_CREATE_VRAM_CLEARED &&
 	    bo->tbo.mem.placement & TTM_PL_FLAG_VRAM) {
 		struct dma_fence *fence;
-- 
cgit v1.1


From 996cab955384122848d8132554de43dce0d3c8a2 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Christian=20K=C3=B6nig?= <christian.koenig@amd.com>
Date: Tue, 8 May 2018 12:26:52 +0200
Subject: drm/amdgpu: add HDP flush dummy for UVD 6/7
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The UVD firmware doesn't seem to like the HDP flush here.

This worked for years without HDP flush, so just skip it.

Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c | 16 ++++++++++++++--
 drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c | 13 ++++++++++++-
 2 files changed, 26 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
index 6d33598..8041b26 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
@@ -964,6 +964,16 @@ static void uvd_v6_0_enc_ring_emit_fence(struct amdgpu_ring *ring, u64 addr,
 }
 
 /**
+ * uvd_v6_0_ring_emit_hdp_flush - skip HDP flushing
+ *
+ * @ring: amdgpu_ring pointer
+ */
+static void uvd_v6_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
+{
+	/* The firmware doesn't seem to like touching registers at this point. */
+}
+
+/**
  * uvd_v6_0_ring_test_ring - register write test
  *
  * @ring: amdgpu_ring pointer
@@ -1528,12 +1538,13 @@ static const struct amdgpu_ring_funcs uvd_v6_0_ring_phys_funcs = {
 	.set_wptr = uvd_v6_0_ring_set_wptr,
 	.parse_cs = amdgpu_uvd_ring_parse_cs,
 	.emit_frame_size =
-		6 + 6 + /* hdp flush / invalidate */
+		6 + /* hdp invalidate */
 		10 + /* uvd_v6_0_ring_emit_pipeline_sync */
 		14, /* uvd_v6_0_ring_emit_fence x1 no user fence */
 	.emit_ib_size = 8, /* uvd_v6_0_ring_emit_ib */
 	.emit_ib = uvd_v6_0_ring_emit_ib,
 	.emit_fence = uvd_v6_0_ring_emit_fence,
+	.emit_hdp_flush = uvd_v6_0_ring_emit_hdp_flush,
 	.test_ring = uvd_v6_0_ring_test_ring,
 	.test_ib = amdgpu_uvd_ring_test_ib,
 	.insert_nop = amdgpu_ring_insert_nop,
@@ -1552,7 +1563,7 @@ static const struct amdgpu_ring_funcs uvd_v6_0_ring_vm_funcs = {
 	.get_wptr = uvd_v6_0_ring_get_wptr,
 	.set_wptr = uvd_v6_0_ring_set_wptr,
 	.emit_frame_size =
-		6 + 6 + /* hdp flush / invalidate */
+		6 + /* hdp invalidate */
 		10 + /* uvd_v6_0_ring_emit_pipeline_sync */
 		VI_FLUSH_GPU_TLB_NUM_WREG * 6 + 8 + /* uvd_v6_0_ring_emit_vm_flush */
 		14 + 14, /* uvd_v6_0_ring_emit_fence x2 vm fence */
@@ -1561,6 +1572,7 @@ static const struct amdgpu_ring_funcs uvd_v6_0_ring_vm_funcs = {
 	.emit_fence = uvd_v6_0_ring_emit_fence,
 	.emit_vm_flush = uvd_v6_0_ring_emit_vm_flush,
 	.emit_pipeline_sync = uvd_v6_0_ring_emit_pipeline_sync,
+	.emit_hdp_flush = uvd_v6_0_ring_emit_hdp_flush,
 	.test_ring = uvd_v6_0_ring_test_ring,
 	.test_ib = amdgpu_uvd_ring_test_ib,
 	.insert_nop = amdgpu_ring_insert_nop,
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
index 2251db4..b0de1e0 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
@@ -1136,6 +1136,16 @@ static void uvd_v7_0_enc_ring_emit_fence(struct amdgpu_ring *ring, u64 addr,
 }
 
 /**
+ * uvd_v7_0_ring_emit_hdp_flush - skip HDP flushing
+ *
+ * @ring: amdgpu_ring pointer
+ */
+static void uvd_v7_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
+{
+	/* The firmware doesn't seem to like touching registers at this point. */
+}
+
+/**
  * uvd_v7_0_ring_test_ring - register write test
  *
  * @ring: amdgpu_ring pointer
@@ -1654,7 +1664,7 @@ static const struct amdgpu_ring_funcs uvd_v7_0_ring_vm_funcs = {
 	.get_wptr = uvd_v7_0_ring_get_wptr,
 	.set_wptr = uvd_v7_0_ring_set_wptr,
 	.emit_frame_size =
-		6 + 6 + /* hdp flush / invalidate */
+		6 + /* hdp invalidate */
 		SOC15_FLUSH_GPU_TLB_NUM_WREG * 6 +
 		SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 8 +
 		8 + /* uvd_v7_0_ring_emit_vm_flush */
@@ -1663,6 +1673,7 @@ static const struct amdgpu_ring_funcs uvd_v7_0_ring_vm_funcs = {
 	.emit_ib = uvd_v7_0_ring_emit_ib,
 	.emit_fence = uvd_v7_0_ring_emit_fence,
 	.emit_vm_flush = uvd_v7_0_ring_emit_vm_flush,
+	.emit_hdp_flush = uvd_v7_0_ring_emit_hdp_flush,
 	.test_ring = uvd_v7_0_ring_test_ring,
 	.test_ib = amdgpu_uvd_ring_test_ib,
 	.insert_nop = uvd_v7_0_ring_insert_nop,
-- 
cgit v1.1


From 323a9dbc452da5c155e5c17fe91c07093824fe27 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Thu, 10 May 2018 15:10:14 -0500
Subject: drm/amdgpu/gmc9: remove unused register defs
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

These got moved to the new df module so no longer
used in this file.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c | 14 --------------
 1 file changed, 14 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
index 6c9f7f9..6cccf0e 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
@@ -43,20 +43,6 @@
 #include "gfxhub_v1_0.h"
 #include "mmhub_v1_0.h"
 
-#define mmDF_CS_AON0_DramBaseAddress0                                                                  0x0044
-#define mmDF_CS_AON0_DramBaseAddress0_BASE_IDX                                                         0
-//DF_CS_AON0_DramBaseAddress0
-#define DF_CS_AON0_DramBaseAddress0__AddrRngVal__SHIFT                                                        0x0
-#define DF_CS_AON0_DramBaseAddress0__LgcyMmioHoleEn__SHIFT                                                    0x1
-#define DF_CS_AON0_DramBaseAddress0__IntLvNumChan__SHIFT                                                      0x4
-#define DF_CS_AON0_DramBaseAddress0__IntLvAddrSel__SHIFT                                                      0x8
-#define DF_CS_AON0_DramBaseAddress0__DramBaseAddr__SHIFT                                                      0xc
-#define DF_CS_AON0_DramBaseAddress0__AddrRngVal_MASK                                                          0x00000001L
-#define DF_CS_AON0_DramBaseAddress0__LgcyMmioHoleEn_MASK                                                      0x00000002L
-#define DF_CS_AON0_DramBaseAddress0__IntLvNumChan_MASK                                                        0x000000F0L
-#define DF_CS_AON0_DramBaseAddress0__IntLvAddrSel_MASK                                                        0x00000700L
-#define DF_CS_AON0_DramBaseAddress0__DramBaseAddr_MASK                                                        0xFFFFF000L
-
 /* add these here since we already include dce12 headers and these are for DCN */
 #define mmHUBP0_DCSURF_PRI_VIEWPORT_DIMENSION                                                          0x055d
 #define mmHUBP0_DCSURF_PRI_VIEWPORT_DIMENSION_BASE_IDX                                                 2
-- 
cgit v1.1


From c430bc977059a73758f666da545bc16c759fc165 Mon Sep 17 00:00:00 2001
From: Junwei Zhang <Jerry.Zhang@amd.com>
Date: Fri, 11 May 2018 14:54:31 +0800
Subject: drm/amdgpu: fix null pointer for bo unmap trace function
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

fix crash in trace.

Signed-off-by: Junwei Zhang <Jerry.Zhang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h
index 532263a..e96e26d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h
@@ -275,7 +275,7 @@ TRACE_EVENT(amdgpu_vm_bo_unmap,
 			     ),
 
 	    TP_fast_assign(
-			   __entry->bo = bo_va->base.bo;
+			   __entry->bo = bo_va ? bo_va->base.bo : NULL;
 			   __entry->start = mapping->start;
 			   __entry->last = mapping->last;
 			   __entry->offset = mapping->offset;
-- 
cgit v1.1


From a2a330ad66313084c9432b32862aa7e1255da9b4 Mon Sep 17 00:00:00 2001
From: Shirish S <shirish.s@amd.com>
Date: Fri, 27 Apr 2018 15:47:21 +0530
Subject: drm/amd/display: remove need of modeset flag for overlay planes (V2)

This patch is in continuation to the
"843e3c7 drm/amd/display: defer modeset check in dm_update_planes_state"
where we started to eliminate the dependency on
DRM_MODE_ATOMIC_ALLOW_MODESET to be set by the user space,
which as such is not mandatory.

After deferring, this patch eliminates the dependency on the flag
for overlay planes.

This has to be done in stages as its a pretty complex and requires thorough
testing before we free primary planes as well from dependency on modeset
flag.

V2: Simplified the plane type check.

Signed-off-by: Shirish S <shirish.s@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 3e0f385..f2f54a9 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -4104,7 +4104,7 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
 		}
 		spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
 
-		if (!pflip_needed) {
+		if (!pflip_needed || plane->type == DRM_PLANE_TYPE_OVERLAY) {
 			WARN_ON(!dm_new_plane_state->dc_state);
 
 			plane_states_constructed[planes_count] = dm_new_plane_state->dc_state;
@@ -4827,7 +4827,8 @@ static int dm_update_planes_state(struct dc *dc,
 
 		/* Remove any changed/removed planes */
 		if (!enable) {
-			if (pflip_needed)
+			if (pflip_needed &&
+			    plane->type != DRM_PLANE_TYPE_OVERLAY)
 				continue;
 
 			if (!old_plane_crtc)
@@ -4874,7 +4875,8 @@ static int dm_update_planes_state(struct dc *dc,
 			if (!dm_new_crtc_state->stream)
 				continue;
 
-			if (pflip_needed)
+			if (pflip_needed &&
+			    plane->type != DRM_PLANE_TYPE_OVERLAY)
 				continue;
 
 			WARN_ON(dm_new_plane_state->dc_state);
-- 
cgit v1.1


From 8eb77198131bab4417b711f899473f4ee6b8ad55 Mon Sep 17 00:00:00 2001
From: Mikita Lipski <mikita.lipski@amd.com>
Date: Mon, 14 May 2018 10:13:57 -0400
Subject: drm/amd/powerplay: Add notify PWE function to SMU10

Functionality to message smc to enable pwe after gpu suspense.
It is used in case when display resumes from S3 and wants to start
audio driver by enabling pwe.

Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c | 8 ++++++++
 drivers/gpu/drm/amd/powerplay/inc/hwmgr.h         | 1 +
 2 files changed, 9 insertions(+)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
index 8b75f52..2f69bfa 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
@@ -1113,6 +1113,13 @@ static int smu10_set_watermarks_for_clocks_ranges(struct pp_hwmgr *hwmgr,
 	data->water_marks_exist = true;
 	return result;
 }
+
+static int smu10_smus_notify_pwe(struct pp_hwmgr *hwmgr)
+{
+
+	return smum_send_msg_to_smc(hwmgr, PPSMC_MSG_SetRccPfcPmeRestoreRegister);
+}
+
 static int smu10_set_mmhub_powergating_by_smu(struct pp_hwmgr *hwmgr)
 {
 	return smum_send_msg_to_smc(hwmgr, PPSMC_MSG_PowerGateMmHub);
@@ -1153,6 +1160,7 @@ static const struct pp_hwmgr_func smu10_hwmgr_funcs = {
 	.power_state_set = smu10_set_power_state_tasks,
 	.dynamic_state_management_disable = smu10_disable_dpm_tasks,
 	.set_mmhub_powergating_by_smu = smu10_set_mmhub_powergating_by_smu,
+	.smus_notify_pwe = smu10_smus_notify_pwe,
 	.gfx_off_control = smu10_gfx_off_control,
 };
 
diff --git a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
index 3d9743f..3c321c7 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
@@ -326,6 +326,7 @@ struct pp_hwmgr_func {
 					long *input, uint32_t size);
 	int (*set_power_limit)(struct pp_hwmgr *hwmgr, uint32_t n);
 	int (*set_mmhub_powergating_by_smu)(struct pp_hwmgr *hwmgr);
+	int (*smus_notify_pwe)(struct pp_hwmgr *hwmgr);
 };
 
 struct pp_table_func {
-- 
cgit v1.1


From 959a2091fae0fa498c79e095a4f6cbbb202a1194 Mon Sep 17 00:00:00 2001
From: Yong Zhao <yong.zhao@amd.com>
Date: Mon, 14 May 2018 12:15:27 -0400
Subject: drm/amdgpu: Add support to change mtype for 2nd part of gart BOs on
 GFX9
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This change prepares for a workaround in amdkfd for a GFX9 HW bug. It
requires the control stack memory of compute queues, which is allocated
from the second page of MQD gart BOs, to have mtype NC, rather than
the default UC.

Signed-off-by: Yong Zhao <yong.zhao@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 54 +++++++++++++++++++++++++++------
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h  |  5 +--
 include/uapi/drm/amdgpu_drm.h           |  4 +++
 3 files changed, 51 insertions(+), 12 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index dfd22db..cc3b067 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -834,6 +834,45 @@ static void amdgpu_ttm_tt_unpin_userptr(struct ttm_tt *ttm)
 	sg_free_table(ttm->sg);
 }
 
+int amdgpu_ttm_gart_bind(struct amdgpu_device *adev,
+				struct ttm_buffer_object *tbo,
+				uint64_t flags)
+{
+	struct amdgpu_bo *abo = ttm_to_amdgpu_bo(tbo);
+	struct ttm_tt *ttm = tbo->ttm;
+	struct amdgpu_ttm_tt *gtt = (void *)ttm;
+	int r;
+
+	if (abo->flags & AMDGPU_GEM_CREATE_MQD_GFX9) {
+		uint64_t page_idx = 1;
+
+		r = amdgpu_gart_bind(adev, gtt->offset, page_idx,
+				ttm->pages, gtt->ttm.dma_address, flags);
+		if (r)
+			goto gart_bind_fail;
+
+		/* Patch mtype of the second part BO */
+		flags &=  ~AMDGPU_PTE_MTYPE_MASK;
+		flags |= AMDGPU_PTE_MTYPE(AMDGPU_MTYPE_NC);
+
+		r = amdgpu_gart_bind(adev,
+				gtt->offset + (page_idx << PAGE_SHIFT),
+				ttm->num_pages - page_idx,
+				&ttm->pages[page_idx],
+				&(gtt->ttm.dma_address[page_idx]), flags);
+	} else {
+		r = amdgpu_gart_bind(adev, gtt->offset, ttm->num_pages,
+				     ttm->pages, gtt->ttm.dma_address, flags);
+	}
+
+gart_bind_fail:
+	if (r)
+		DRM_ERROR("failed to bind %lu pages at 0x%08llX\n",
+			  ttm->num_pages, gtt->offset);
+
+	return r;
+}
+
 static int amdgpu_ttm_backend_bind(struct ttm_tt *ttm,
 				   struct ttm_mem_reg *bo_mem)
 {
@@ -907,8 +946,7 @@ int amdgpu_ttm_alloc_gart(struct ttm_buffer_object *bo)
 
 	flags = amdgpu_ttm_tt_pte_flags(adev, bo->ttm, &tmp);
 	gtt->offset = (u64)tmp.start << PAGE_SHIFT;
-	r = amdgpu_gart_bind(adev, gtt->offset, bo->ttm->num_pages,
-			     bo->ttm->pages, gtt->ttm.dma_address, flags);
+	r = amdgpu_ttm_gart_bind(adev, bo, flags);
 	if (unlikely(r)) {
 		ttm_bo_mem_put(bo, &tmp);
 		return r;
@@ -925,19 +963,15 @@ int amdgpu_ttm_alloc_gart(struct ttm_buffer_object *bo)
 int amdgpu_ttm_recover_gart(struct ttm_buffer_object *tbo)
 {
 	struct amdgpu_device *adev = amdgpu_ttm_adev(tbo->bdev);
-	struct amdgpu_ttm_tt *gtt = (void *)tbo->ttm;
 	uint64_t flags;
 	int r;
 
-	if (!gtt)
+	if (!tbo->ttm)
 		return 0;
 
-	flags = amdgpu_ttm_tt_pte_flags(adev, &gtt->ttm.ttm, &tbo->mem);
-	r = amdgpu_gart_bind(adev, gtt->offset, gtt->ttm.ttm.num_pages,
-			     gtt->ttm.ttm.pages, gtt->ttm.dma_address, flags);
-	if (r)
-		DRM_ERROR("failed to bind %lu pages at 0x%08llX\n",
-			  gtt->ttm.ttm.num_pages, gtt->offset);
+	flags = amdgpu_ttm_tt_pte_flags(adev, tbo->ttm, &tbo->mem);
+	r = amdgpu_ttm_gart_bind(adev, tbo, flags);
+
 	return r;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
index 30f0803..4cf6786 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
@@ -75,11 +75,12 @@ struct amdgpu_bo_list_entry;
 /* PDE Block Fragment Size for VEGA10 */
 #define AMDGPU_PDE_BFS(a)	((uint64_t)a << 59)
 
-/* VEGA10 only */
+
+/* For GFX9 */
 #define AMDGPU_PTE_MTYPE(a)    ((uint64_t)a << 57)
 #define AMDGPU_PTE_MTYPE_MASK	AMDGPU_PTE_MTYPE(3ULL)
 
-/* For Raven */
+#define AMDGPU_MTYPE_NC 0
 #define AMDGPU_MTYPE_CC 2
 
 #define AMDGPU_PTE_DEFAULT_ATC  (AMDGPU_PTE_SYSTEM      \
diff --git a/include/uapi/drm/amdgpu_drm.h b/include/uapi/drm/amdgpu_drm.h
index 081d256..78b4dd8 100644
--- a/include/uapi/drm/amdgpu_drm.h
+++ b/include/uapi/drm/amdgpu_drm.h
@@ -101,6 +101,10 @@ extern "C" {
 #define AMDGPU_GEM_CREATE_VM_ALWAYS_VALID	(1 << 6)
 /* Flag that BO sharing will be explicitly synchronized */
 #define AMDGPU_GEM_CREATE_EXPLICIT_SYNC		(1 << 7)
+/* Flag that indicates allocating MQD gart on GFX9, where the mtype
+ * for the second page onward should be set to NC.
+ */
+#define AMDGPU_GEM_CREATE_MQD_GFX9		(1 << 8)
 
 struct drm_amdgpu_gem_create_in  {
 	/** the requested memory size */
-- 
cgit v1.1


From c7535379f660b721998ad6ab397809b0cbeb66d7 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Christian=20K=C3=B6nig?= <christian.koenig@amd.com>
Date: Fri, 11 May 2018 23:13:39 +0800
Subject: drm/amdgpu: drop printing the BO offset in the gem debugfs (v2)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

It is meaningless anyway.

v2: remove unused variable (Alex)

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c | 5 -----
 1 file changed, 5 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
index 7d3dc22..f79bbf8 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
@@ -785,7 +785,6 @@ static int amdgpu_debugfs_gem_bo_info(int id, void *ptr, void *data)
 	unsigned domain;
 	const char *placement;
 	unsigned pin_count;
-	uint64_t offset;
 
 	domain = amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type);
 	switch (domain) {
@@ -803,10 +802,6 @@ static int amdgpu_debugfs_gem_bo_info(int id, void *ptr, void *data)
 	seq_printf(m, "\t0x%08x: %12ld byte %s",
 		   id, amdgpu_bo_size(bo), placement);
 
-	offset = READ_ONCE(bo->tbo.mem.start);
-	if (offset != AMDGPU_BO_INVALID_OFFSET)
-		seq_printf(m, " @ 0x%010Lx", offset);
-
 	pin_count = READ_ONCE(bo->pin_count);
 	if (pin_count)
 		seq_printf(m, " pin count %d", pin_count);
-- 
cgit v1.1


From 6b155d6af03a053a7de2a72255563d7ef40c9644 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Christian=20K=C3=B6nig?= <christian.koenig@amd.com>
Date: Fri, 11 May 2018 23:14:29 +0800
Subject: drm/amdgpu: print the BO flags in the gem debugfs entry
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Quite useful to know.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c | 15 +++++++++++++++
 1 file changed, 15 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
index f79bbf8..2c8e273 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
@@ -774,6 +774,12 @@ int amdgpu_mode_dumb_create(struct drm_file *file_priv,
 }
 
 #if defined(CONFIG_DEBUG_FS)
+
+#define amdgpu_debugfs_gem_bo_print_flag(m, bo, flag)	\
+	if (bo->flags & (AMDGPU_GEM_CREATE_ ## flag)) {	\
+		seq_printf((m), " " #flag);		\
+	}
+
 static int amdgpu_debugfs_gem_bo_info(int id, void *ptr, void *data)
 {
 	struct drm_gem_object *gobj = ptr;
@@ -814,6 +820,15 @@ static int amdgpu_debugfs_gem_bo_info(int id, void *ptr, void *data)
 	else if (dma_buf)
 		seq_printf(m, " exported as %p", dma_buf);
 
+	amdgpu_debugfs_gem_bo_print_flag(m, bo, CPU_ACCESS_REQUIRED);
+	amdgpu_debugfs_gem_bo_print_flag(m, bo, NO_CPU_ACCESS);
+	amdgpu_debugfs_gem_bo_print_flag(m, bo, CPU_GTT_USWC);
+	amdgpu_debugfs_gem_bo_print_flag(m, bo, VRAM_CLEARED);
+	amdgpu_debugfs_gem_bo_print_flag(m, bo, SHADOW);
+	amdgpu_debugfs_gem_bo_print_flag(m, bo, VRAM_CONTIGUOUS);
+	amdgpu_debugfs_gem_bo_print_flag(m, bo, VM_ALWAYS_VALID);
+	amdgpu_debugfs_gem_bo_print_flag(m, bo, EXPLICIT_SYNC);
+
 	seq_printf(m, "\n");
 
 	return 0;
-- 
cgit v1.1


From 1297bf2e916d2012995b642dd6851332a73126c2 Mon Sep 17 00:00:00 2001
From: Dirk Hohndel <dirk@hohndel.org>
Date: Wed, 2 May 2018 15:46:21 +0200
Subject: Add SPDX idenitifier and clarify license
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This is dual licensed under GPL-2.0 or MIT.

Signed-off-by: Dirk Hohndel (VMware) <dirk@hohndel.org>
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/ttm/ttm_agp_backend.c  | 1 +
 drivers/gpu/drm/ttm/ttm_bo.c           | 1 +
 drivers/gpu/drm/ttm/ttm_bo_manager.c   | 1 +
 drivers/gpu/drm/ttm/ttm_bo_util.c      | 1 +
 drivers/gpu/drm/ttm/ttm_bo_vm.c        | 1 +
 drivers/gpu/drm/ttm/ttm_execbuf_util.c | 1 +
 drivers/gpu/drm/ttm/ttm_lock.c         | 1 +
 drivers/gpu/drm/ttm/ttm_memory.c       | 1 +
 drivers/gpu/drm/ttm/ttm_module.c       | 1 +
 drivers/gpu/drm/ttm/ttm_object.c       | 1 +
 drivers/gpu/drm/ttm/ttm_tt.c           | 1 +
 11 files changed, 11 insertions(+)

diff --git a/drivers/gpu/drm/ttm/ttm_agp_backend.c b/drivers/gpu/drm/ttm/ttm_agp_backend.c
index 7c2485f..ea4d59e 100644
--- a/drivers/gpu/drm/ttm/ttm_agp_backend.c
+++ b/drivers/gpu/drm/ttm/ttm_agp_backend.c
@@ -1,3 +1,4 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
  * Copyright (c) 2006-2009 VMware, Inc., Palo Alto, CA., USA
diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c
index cba5015..5d8688e52 100644
--- a/drivers/gpu/drm/ttm/ttm_bo.c
+++ b/drivers/gpu/drm/ttm/ttm_bo.c
@@ -1,3 +1,4 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
  * Copyright (c) 2006-2009 VMware, Inc., Palo Alto, CA., USA
diff --git a/drivers/gpu/drm/ttm/ttm_bo_manager.c b/drivers/gpu/drm/ttm/ttm_bo_manager.c
index a7c232d..18d3deb 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_manager.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_manager.c
@@ -1,3 +1,4 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
  * Copyright (c) 2007-2010 VMware, Inc., Palo Alto, CA., USA
diff --git a/drivers/gpu/drm/ttm/ttm_bo_util.c b/drivers/gpu/drm/ttm/ttm_bo_util.c
index f3bf545..f2c1677 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_util.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_util.c
@@ -1,3 +1,4 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
  * Copyright (c) 2007-2009 VMware, Inc., Palo Alto, CA., USA
diff --git a/drivers/gpu/drm/ttm/ttm_bo_vm.c b/drivers/gpu/drm/ttm/ttm_bo_vm.c
index 8eba95b..c7ece76 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_vm.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_vm.c
@@ -1,3 +1,4 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
  * Copyright (c) 2006-2009 VMware, Inc., Palo Alto, CA., USA
diff --git a/drivers/gpu/drm/ttm/ttm_execbuf_util.c b/drivers/gpu/drm/ttm/ttm_execbuf_util.c
index 3dca206e..e73ae0d 100644
--- a/drivers/gpu/drm/ttm/ttm_execbuf_util.c
+++ b/drivers/gpu/drm/ttm/ttm_execbuf_util.c
@@ -1,3 +1,4 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
  * Copyright (c) 2006-2009 VMware, Inc., Palo Alto, CA., USA
diff --git a/drivers/gpu/drm/ttm/ttm_lock.c b/drivers/gpu/drm/ttm/ttm_lock.c
index 913f431..20694b8 100644
--- a/drivers/gpu/drm/ttm/ttm_lock.c
+++ b/drivers/gpu/drm/ttm/ttm_lock.c
@@ -1,3 +1,4 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
  * Copyright (c) 2007-2009 VMware, Inc., Palo Alto, CA., USA
diff --git a/drivers/gpu/drm/ttm/ttm_memory.c b/drivers/gpu/drm/ttm/ttm_memory.c
index 27856c5..450387c 100644
--- a/drivers/gpu/drm/ttm/ttm_memory.c
+++ b/drivers/gpu/drm/ttm/ttm_memory.c
@@ -1,3 +1,4 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
  * Copyright (c) 2006-2009 VMware, Inc., Palo Alto, CA., USA
diff --git a/drivers/gpu/drm/ttm/ttm_module.c b/drivers/gpu/drm/ttm/ttm_module.c
index 66fc639..6ff40c0 100644
--- a/drivers/gpu/drm/ttm/ttm_module.c
+++ b/drivers/gpu/drm/ttm/ttm_module.c
@@ -1,3 +1,4 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
  * Copyright (c) 2006-2009 VMware, Inc., Palo Alto, CA., USA
diff --git a/drivers/gpu/drm/ttm/ttm_object.c b/drivers/gpu/drm/ttm/ttm_object.c
index 1aa2baa..74f1b1e 100644
--- a/drivers/gpu/drm/ttm/ttm_object.c
+++ b/drivers/gpu/drm/ttm/ttm_object.c
@@ -1,3 +1,4 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
  * Copyright (c) 2009-2013 VMware, Inc., Palo Alto, CA., USA
diff --git a/drivers/gpu/drm/ttm/ttm_tt.c b/drivers/gpu/drm/ttm/ttm_tt.c
index 7e672be..a1e5439 100644
--- a/drivers/gpu/drm/ttm/ttm_tt.c
+++ b/drivers/gpu/drm/ttm/ttm_tt.c
@@ -1,3 +1,4 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
  * Copyright (c) 2006-2009 VMware, Inc., Palo Alto, CA., USA
-- 
cgit v1.1


From 8344c53f57057b42a5da87e9557c40fcda18fb7a Mon Sep 17 00:00:00 2001
From: Nayan Deshmukh <nayan26deshmukh@gmail.com>
Date: Thu, 29 Mar 2018 22:36:32 +0530
Subject: drm/scheduler: remove unused parameter
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

this patch also effect the amdgpu and etnaviv drivers which
use the function drm_sched_entity_init

Signed-off-by: Nayan Deshmukh <nayan26deshmukh@gmail.com>
Suggested-by: Christian König <christian.koenig@amd.com>
Acked-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c   | 2 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c   | 2 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c   | 2 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c   | 2 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c   | 4 ++--
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c    | 2 +-
 drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c     | 2 +-
 drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c     | 2 +-
 drivers/gpu/drm/etnaviv/etnaviv_drv.c     | 2 +-
 drivers/gpu/drm/scheduler/gpu_scheduler.c | 3 +--
 include/drm/gpu_scheduler.h               | 2 +-
 11 files changed, 12 insertions(+), 13 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
index 6741a62..a8e531d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
@@ -91,7 +91,7 @@ static int amdgpu_ctx_init(struct amdgpu_device *adev,
 			continue;
 
 		r = drm_sched_entity_init(&ring->sched, &ctx->rings[i].entity,
-					  rq, amdgpu_sched_jobs, &ctx->guilty);
+					  rq, &ctx->guilty);
 		if (r)
 			goto failed;
 	}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index cc3b067..5e9fd25 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -111,7 +111,7 @@ static int amdgpu_ttm_global_init(struct amdgpu_device *adev)
 	ring = adev->mman.buffer_funcs_ring;
 	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_KERNEL];
 	r = drm_sched_entity_init(&ring->sched, &adev->mman.entity,
-				  rq, amdgpu_sched_jobs, NULL);
+				  rq, NULL);
 	if (r) {
 		DRM_ERROR("Failed setting up TTM BO move run queue.\n");
 		goto error_entity;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
index d8dd402..de4d77a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
@@ -242,7 +242,7 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
 	ring = &adev->uvd.ring;
 	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
 	r = drm_sched_entity_init(&ring->sched, &adev->uvd.entity,
-				  rq, amdgpu_sched_jobs, NULL);
+				  rq, NULL);
 	if (r != 0) {
 		DRM_ERROR("Failed setting up UVD run queue.\n");
 		return r;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
index e2186ed..a86322f 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
@@ -186,7 +186,7 @@ int amdgpu_vce_sw_init(struct amdgpu_device *adev, unsigned long size)
 	ring = &adev->vce.ring[0];
 	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
 	r = drm_sched_entity_init(&ring->sched, &adev->vce.entity,
-				  rq, amdgpu_sched_jobs, NULL);
+				  rq, NULL);
 	if (r != 0) {
 		DRM_ERROR("Failed setting up VCE run queue.\n");
 		return r;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
index 58e4953..e5d234c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
@@ -105,7 +105,7 @@ int amdgpu_vcn_sw_init(struct amdgpu_device *adev)
 	ring = &adev->vcn.ring_dec;
 	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
 	r = drm_sched_entity_init(&ring->sched, &adev->vcn.entity_dec,
-				  rq, amdgpu_sched_jobs, NULL);
+				  rq, NULL);
 	if (r != 0) {
 		DRM_ERROR("Failed setting up VCN dec run queue.\n");
 		return r;
@@ -114,7 +114,7 @@ int amdgpu_vcn_sw_init(struct amdgpu_device *adev)
 	ring = &adev->vcn.ring_enc[0];
 	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
 	r = drm_sched_entity_init(&ring->sched, &adev->vcn.entity_enc,
-				  rq, amdgpu_sched_jobs, NULL);
+				  rq, NULL);
 	if (r != 0) {
 		DRM_ERROR("Failed setting up VCN enc run queue.\n");
 		return r;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index 8e71d39..1a8f4e0 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -2404,7 +2404,7 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 	ring = adev->vm_manager.vm_pte_rings[ring_instance];
 	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_KERNEL];
 	r = drm_sched_entity_init(&ring->sched, &vm->entity,
-				  rq, amdgpu_sched_jobs, NULL);
+				  rq, NULL);
 	if (r)
 		return r;
 
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
index 8041b26..ca6ab56 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
@@ -429,7 +429,7 @@ static int uvd_v6_0_sw_init(void *handle)
 		ring = &adev->uvd.ring_enc[0];
 		rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
 		r = drm_sched_entity_init(&ring->sched, &adev->uvd.entity_enc,
-					  rq, amdgpu_sched_jobs, NULL);
+					  rq, NULL);
 		if (r) {
 			DRM_ERROR("Failed setting up UVD ENC run queue.\n");
 			return r;
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
index b0de1e0..0ca63d5 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
@@ -418,7 +418,7 @@ static int uvd_v7_0_sw_init(void *handle)
 	ring = &adev->uvd.ring_enc[0];
 	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
 	r = drm_sched_entity_init(&ring->sched, &adev->uvd.entity_enc,
-				  rq, amdgpu_sched_jobs, NULL);
+				  rq, NULL);
 	if (r) {
 		DRM_ERROR("Failed setting up UVD ENC run queue.\n");
 		return r;
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_drv.c b/drivers/gpu/drm/etnaviv/etnaviv_drv.c
index ab50090..23e73c2 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_drv.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_drv.c
@@ -116,7 +116,7 @@ static int etnaviv_open(struct drm_device *dev, struct drm_file *file)
 			drm_sched_entity_init(&gpu->sched,
 				&ctx->sched_entity[i],
 				&gpu->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL],
-				32, NULL);
+				NULL);
 			}
 	}
 
diff --git a/drivers/gpu/drm/scheduler/gpu_scheduler.c b/drivers/gpu/drm/scheduler/gpu_scheduler.c
index 1f1dd70..a364fc0 100644
--- a/drivers/gpu/drm/scheduler/gpu_scheduler.c
+++ b/drivers/gpu/drm/scheduler/gpu_scheduler.c
@@ -117,7 +117,6 @@ drm_sched_rq_select_entity(struct drm_sched_rq *rq)
  * @sched	The pointer to the scheduler
  * @entity	The pointer to a valid drm_sched_entity
  * @rq		The run queue this entity belongs
- * @jobs	The max number of jobs in the job queue
  * @guilty      atomic_t set to 1 when a job on this queue
  *              is found to be guilty causing a timeout
  *
@@ -126,7 +125,7 @@ drm_sched_rq_select_entity(struct drm_sched_rq *rq)
 int drm_sched_entity_init(struct drm_gpu_scheduler *sched,
 			  struct drm_sched_entity *entity,
 			  struct drm_sched_rq *rq,
-			  uint32_t jobs, atomic_t *guilty)
+			  atomic_t *guilty)
 {
 	if (!(sched && entity && rq))
 		return -EINVAL;
diff --git a/include/drm/gpu_scheduler.h b/include/drm/gpu_scheduler.h
index 350a62c..5238006 100644
--- a/include/drm/gpu_scheduler.h
+++ b/include/drm/gpu_scheduler.h
@@ -188,7 +188,7 @@ void drm_sched_fini(struct drm_gpu_scheduler *sched);
 int drm_sched_entity_init(struct drm_gpu_scheduler *sched,
 			  struct drm_sched_entity *entity,
 			  struct drm_sched_rq *rq,
-			  uint32_t jobs, atomic_t *guilty);
+			  atomic_t *guilty);
 void drm_sched_entity_do_release(struct drm_gpu_scheduler *sched,
 			   struct drm_sched_entity *entity);
 void drm_sched_entity_cleanup(struct drm_gpu_scheduler *sched,
-- 
cgit v1.1


From 548da31da9805645b1e8043da5081b9745545248 Mon Sep 17 00:00:00 2001
From: Stephen Rothwell <sfr@canb.auug.org.au>
Date: Wed, 16 May 2018 16:43:34 +1000
Subject: drm/amdgpu: include pagemap.h for release_pages()

Fixes: 5ae0283e831a ("drm/amdgpu: Add userptr support for KFD"
Cc: Felix Kuehling <Felix.Kuehling@amd.com>
Cc: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Dave Airlie <airlied@redhat.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
index 72ab2b1..ff8fd75 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
@@ -23,6 +23,7 @@
 #define pr_fmt(fmt) "kfd2kgd: " fmt
 
 #include <linux/list.h>
+#include <linux/pagemap.h>
 #include <linux/sched/mm.h>
 #include <drm/drmP.h>
 #include "amdgpu_object.h"
-- 
cgit v1.1


From dd856d924b2471bf28e7c60df881529525a1192c Mon Sep 17 00:00:00 2001
From: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Date: Tue, 15 May 2018 18:57:36 +0300
Subject: drm: rcar-du: Fix rcar_du_of_init() stub

The rcar_du_of_init() function is supposed to be defined as a stub when
CONFIG_DRM_RCAR_LVDS is disabled as the rcar_du_of.c file isn't compiled
in that case. However, a bug in the configuration option check makes it
a stub when CONFIG_DRM_RCAR_LVDS=m as well, which prevents legacy DTs
from being fixed at boot time. Fix the configuration option check by
using IS_ENABLED.

Fixes: 81c0e3dd8292 ("drm: rcar-du: Fix legacy DT to create LVDS encoder nodes")
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180515155736.3379-1-laurent.pinchart+renesas@ideasonboard.com
---
 drivers/gpu/drm/rcar-du/rcar_du_of.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/rcar-du/rcar_du_of.h b/drivers/gpu/drm/rcar-du/rcar_du_of.h
index c2e65a7..8dd3fbe 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_of.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_of.h
@@ -11,7 +11,7 @@
 
 struct of_device_id;
 
-#ifdef CONFIG_DRM_RCAR_LVDS
+#if IS_ENABLED(CONFIG_DRM_RCAR_LVDS)
 void __init rcar_du_of_init(const struct of_device_id *of_ids);
 #else
 static inline void rcar_du_of_init(const struct of_device_id *of_ids) { }
-- 
cgit v1.1


From 315852b422972e6ebb1dfddaadada09e46a2681a Mon Sep 17 00:00:00 2001
From: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Date: Tue, 15 May 2018 20:47:52 +0300
Subject: drm: rcar-du: Fix build failure

Commit 75a07f399cd4 ("drm: rcar-du: Zero-out sg_tables when duplicating
plane state") introduced a reference to the alpha field of struct
rcar_du_vsp_plane_state that got removed in commit 301a9b8d5456
("drm/rcar-du: Convert to the new generic alpha property"). The issue
stems from the merge of the two commits through separate branches and
breaks compilation of the driver. Fix it.

Fixes: 75a07f399cd4 ("drm: rcar-du: Zero-out sg_tables when duplicating plane state")
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Reviewed-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Tested-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180515174752.28954-1-laurent.pinchart+renesas@ideasonboard.com
---
 drivers/gpu/drm/rcar-du/rcar_du_vsp.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/drivers/gpu/drm/rcar-du/rcar_du_vsp.c b/drivers/gpu/drm/rcar-du/rcar_du_vsp.c
index 27a4408..c59f0cf 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_vsp.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_vsp.c
@@ -314,7 +314,6 @@ rcar_du_vsp_plane_atomic_duplicate_state(struct drm_plane *plane)
 		return NULL;
 
 	__drm_atomic_helper_plane_duplicate_state(plane, &copy->state);
-	copy->alpha = to_rcar_vsp_plane_state(plane->state)->alpha;
 
 	return &copy->state;
 }
-- 
cgit v1.1


From 230630bd3834af0ea6ec75354ec21819de148ee1 Mon Sep 17 00:00:00 2001
From: Thierry Reding <treding@nvidia.com>
Date: Fri, 4 May 2018 15:08:49 +0200
Subject: drm/tegra: gr3d: Properly clean up resources

Failure to register the Tegra DRM client would leak the resources. Move
cleanup code to error unwinding gotos to fix that and share the cleanup
code with the other error paths.

Reviewed-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/gr3d.c | 28 ++++++++++++++++++++++------
 1 file changed, 22 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/tegra/gr3d.c b/drivers/gpu/drm/tegra/gr3d.c
index ce51206..9303278e 100644
--- a/drivers/gpu/drm/tegra/gr3d.c
+++ b/drivers/gpu/drm/tegra/gr3d.c
@@ -52,8 +52,9 @@ static int gr3d_init(struct host1x_client *client)
 
 	client->syncpts[0] = host1x_syncpt_request(client, flags);
 	if (!client->syncpts[0]) {
-		host1x_channel_put(gr3d->channel);
-		return -ENOMEM;
+		err = -ENOMEM;
+		dev_err(client->dev, "failed to request syncpoint: %d\n", err);
+		goto put;
 	}
 
 	if (tegra->domain) {
@@ -65,15 +66,30 @@ static int gr3d_init(struct host1x_client *client)
 				dev_err(client->dev,
 					"failed to attach to domain: %d\n",
 					err);
-				host1x_syncpt_free(client->syncpts[0]);
-				host1x_channel_put(gr3d->channel);
 				iommu_group_put(gr3d->group);
-				return err;
+				goto free;
 			}
 		}
 	}
 
-	return tegra_drm_register_client(dev->dev_private, drm);
+	err = tegra_drm_register_client(dev->dev_private, drm);
+	if (err < 0) {
+		dev_err(client->dev, "failed to register client: %d\n", err);
+		goto detach;
+	}
+
+	return 0;
+
+detach:
+	if (gr3d->group) {
+		iommu_detach_group(tegra->domain, gr3d->group);
+		iommu_group_put(gr3d->group);
+	}
+free:
+	host1x_syncpt_free(client->syncpts[0]);
+put:
+	host1x_channel_put(gr3d->channel);
+	return err;
 }
 
 static int gr3d_exit(struct host1x_client *client)
-- 
cgit v1.1


From 0c407de5ed1a329468122cbf4f3e727e0c1e3f36 Mon Sep 17 00:00:00 2001
From: Thierry Reding <treding@nvidia.com>
Date: Fri, 4 May 2018 15:02:24 +0200
Subject: drm/tegra: Refactor IOMMU attach/detach

Attaching to and detaching from an IOMMU uses the same code sequence in
every driver, so factor it out into separate helpers.

Reviewed-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/dc.c   | 42 +++++++---------------------------------
 drivers/gpu/drm/tegra/drm.c  | 46 ++++++++++++++++++++++++++++++++++++++++++++
 drivers/gpu/drm/tegra/drm.h  |  4 ++++
 drivers/gpu/drm/tegra/gr2d.c | 32 ++++++++----------------------
 drivers/gpu/drm/tegra/gr3d.c | 31 +++++++----------------------
 5 files changed, 72 insertions(+), 83 deletions(-)

diff --git a/drivers/gpu/drm/tegra/dc.c b/drivers/gpu/drm/tegra/dc.c
index c843f11..3e7ec39 100644
--- a/drivers/gpu/drm/tegra/dc.c
+++ b/drivers/gpu/drm/tegra/dc.c
@@ -1837,21 +1837,11 @@ static int tegra_dc_init(struct host1x_client *client)
 	if (!dc->syncpt)
 		dev_warn(dc->dev, "failed to allocate syncpoint\n");
 
-	if (tegra->domain) {
-		dc->group = iommu_group_get(client->dev);
-
-		if (dc->group && dc->group != tegra->group) {
-			err = iommu_attach_group(tegra->domain, dc->group);
-			if (err < 0) {
-				dev_err(dc->dev,
-					"failed to attach to domain: %d\n",
-					err);
-				iommu_group_put(dc->group);
-				return err;
-			}
-
-			tegra->group = dc->group;
-		}
+	dc->group = host1x_client_iommu_attach(client, true);
+	if (IS_ERR(dc->group)) {
+		err = PTR_ERR(dc->group);
+		dev_err(client->dev, "failed to attach to domain: %d\n", err);
+		return err;
 	}
 
 	if (dc->soc->wgrps)
@@ -1916,15 +1906,7 @@ cleanup:
 	if (!IS_ERR(primary))
 		drm_plane_cleanup(primary);
 
-	if (dc->group) {
-		if (dc->group == tegra->group) {
-			iommu_detach_group(tegra->domain, dc->group);
-			tegra->group = NULL;
-		}
-
-		iommu_group_put(dc->group);
-	}
-
+	host1x_client_iommu_detach(client, dc->group);
 	host1x_syncpt_free(dc->syncpt);
 
 	return err;
@@ -1932,9 +1914,7 @@ cleanup:
 
 static int tegra_dc_exit(struct host1x_client *client)
 {
-	struct drm_device *drm = dev_get_drvdata(client->parent);
 	struct tegra_dc *dc = host1x_client_to_dc(client);
-	struct tegra_drm *tegra = drm->dev_private;
 	int err;
 
 	devm_free_irq(dc->dev, dc->irq, dc);
@@ -1945,15 +1925,7 @@ static int tegra_dc_exit(struct host1x_client *client)
 		return err;
 	}
 
-	if (dc->group) {
-		if (dc->group == tegra->group) {
-			iommu_detach_group(tegra->domain, dc->group);
-			tegra->group = NULL;
-		}
-
-		iommu_group_put(dc->group);
-	}
-
+	host1x_client_iommu_detach(client, dc->group);
 	host1x_syncpt_free(dc->syncpt);
 
 	return 0;
diff --git a/drivers/gpu/drm/tegra/drm.c b/drivers/gpu/drm/tegra/drm.c
index 7afe2f6..181e82c 100644
--- a/drivers/gpu/drm/tegra/drm.c
+++ b/drivers/gpu/drm/tegra/drm.c
@@ -1114,6 +1114,52 @@ int tegra_drm_unregister_client(struct tegra_drm *tegra,
 	return 0;
 }
 
+struct iommu_group *host1x_client_iommu_attach(struct host1x_client *client,
+					       bool shared)
+{
+	struct drm_device *drm = dev_get_drvdata(client->parent);
+	struct tegra_drm *tegra = drm->dev_private;
+	struct iommu_group *group = NULL;
+	int err;
+
+	if (tegra->domain) {
+		group = iommu_group_get(client->dev);
+		if (!group) {
+			dev_err(client->dev, "failed to get IOMMU group\n");
+			return ERR_PTR(-ENODEV);
+		}
+
+		if (!shared || (shared && (group != tegra->group))) {
+			err = iommu_attach_group(tegra->domain, group);
+			if (err < 0) {
+				iommu_group_put(group);
+				return ERR_PTR(err);
+			}
+
+			if (shared && !tegra->group)
+				tegra->group = group;
+		}
+	}
+
+	return group;
+}
+
+void host1x_client_iommu_detach(struct host1x_client *client,
+				struct iommu_group *group)
+{
+	struct drm_device *drm = dev_get_drvdata(client->parent);
+	struct tegra_drm *tegra = drm->dev_private;
+
+	if (group) {
+		if (group == tegra->group) {
+			iommu_detach_group(tegra->domain, group);
+			tegra->group = NULL;
+		}
+
+		iommu_group_put(group);
+	}
+}
+
 void *tegra_drm_alloc(struct tegra_drm *tegra, size_t size, dma_addr_t *dma)
 {
 	struct iova *alloc;
diff --git a/drivers/gpu/drm/tegra/drm.h b/drivers/gpu/drm/tegra/drm.h
index 4f41aae..fe263cf 100644
--- a/drivers/gpu/drm/tegra/drm.h
+++ b/drivers/gpu/drm/tegra/drm.h
@@ -110,6 +110,10 @@ int tegra_drm_register_client(struct tegra_drm *tegra,
 			      struct tegra_drm_client *client);
 int tegra_drm_unregister_client(struct tegra_drm *tegra,
 				struct tegra_drm_client *client);
+struct iommu_group *host1x_client_iommu_attach(struct host1x_client *client,
+					       bool shared);
+void host1x_client_iommu_detach(struct host1x_client *client,
+				struct iommu_group *group);
 
 int tegra_drm_init(struct tegra_drm *tegra, struct drm_device *drm);
 int tegra_drm_exit(struct tegra_drm *tegra);
diff --git a/drivers/gpu/drm/tegra/gr2d.c b/drivers/gpu/drm/tegra/gr2d.c
index 0b42e99..2cd0f66 100644
--- a/drivers/gpu/drm/tegra/gr2d.c
+++ b/drivers/gpu/drm/tegra/gr2d.c
@@ -32,7 +32,6 @@ static int gr2d_init(struct host1x_client *client)
 	struct tegra_drm_client *drm = host1x_to_drm_client(client);
 	struct drm_device *dev = dev_get_drvdata(client->parent);
 	unsigned long flags = HOST1X_SYNCPT_HAS_BASE;
-	struct tegra_drm *tegra = dev->dev_private;
 	struct gr2d *gr2d = to_gr2d(drm);
 	int err;
 
@@ -47,22 +46,14 @@ static int gr2d_init(struct host1x_client *client)
 		goto put;
 	}
 
-	if (tegra->domain) {
-		gr2d->group = iommu_group_get(client->dev);
-
-		if (gr2d->group) {
-			err = iommu_attach_group(tegra->domain, gr2d->group);
-			if (err < 0) {
-				dev_err(client->dev,
-					"failed to attach to domain: %d\n",
-					err);
-				iommu_group_put(gr2d->group);
-				goto free;
-			}
-		}
+	gr2d->group = host1x_client_iommu_attach(client, false);
+	if (IS_ERR(gr2d->group)) {
+		err = PTR_ERR(gr2d->group);
+		dev_err(client->dev, "failed to attach to domain: %d\n", err);
+		goto free;
 	}
 
-	err = tegra_drm_register_client(tegra, drm);
+	err = tegra_drm_register_client(dev->dev_private, drm);
 	if (err < 0) {
 		dev_err(client->dev, "failed to register client: %d\n", err);
 		goto detach;
@@ -71,10 +62,7 @@ static int gr2d_init(struct host1x_client *client)
 	return 0;
 
 detach:
-	if (gr2d->group) {
-		iommu_detach_group(tegra->domain, gr2d->group);
-		iommu_group_put(gr2d->group);
-	}
+	host1x_client_iommu_detach(client, gr2d->group);
 free:
 	host1x_syncpt_free(client->syncpts[0]);
 put:
@@ -94,14 +82,10 @@ static int gr2d_exit(struct host1x_client *client)
 	if (err < 0)
 		return err;
 
+	host1x_client_iommu_detach(client, gr2d->group);
 	host1x_syncpt_free(client->syncpts[0]);
 	host1x_channel_put(gr2d->channel);
 
-	if (gr2d->group) {
-		iommu_detach_group(tegra->domain, gr2d->group);
-		iommu_group_put(gr2d->group);
-	}
-
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/tegra/gr3d.c b/drivers/gpu/drm/tegra/gr3d.c
index 9303278e..b00002f 100644
--- a/drivers/gpu/drm/tegra/gr3d.c
+++ b/drivers/gpu/drm/tegra/gr3d.c
@@ -42,7 +42,6 @@ static int gr3d_init(struct host1x_client *client)
 	struct tegra_drm_client *drm = host1x_to_drm_client(client);
 	struct drm_device *dev = dev_get_drvdata(client->parent);
 	unsigned long flags = HOST1X_SYNCPT_HAS_BASE;
-	struct tegra_drm *tegra = dev->dev_private;
 	struct gr3d *gr3d = to_gr3d(drm);
 	int err;
 
@@ -57,19 +56,11 @@ static int gr3d_init(struct host1x_client *client)
 		goto put;
 	}
 
-	if (tegra->domain) {
-		gr3d->group = iommu_group_get(client->dev);
-
-		if (gr3d->group) {
-			err = iommu_attach_group(tegra->domain, gr3d->group);
-			if (err < 0) {
-				dev_err(client->dev,
-					"failed to attach to domain: %d\n",
-					err);
-				iommu_group_put(gr3d->group);
-				goto free;
-			}
-		}
+	gr3d->group = host1x_client_iommu_attach(client, false);
+	if (IS_ERR(gr3d->group)) {
+		err = PTR_ERR(gr3d->group);
+		dev_err(client->dev, "failed to attach to domain: %d\n", err);
+		goto free;
 	}
 
 	err = tegra_drm_register_client(dev->dev_private, drm);
@@ -81,10 +72,7 @@ static int gr3d_init(struct host1x_client *client)
 	return 0;
 
 detach:
-	if (gr3d->group) {
-		iommu_detach_group(tegra->domain, gr3d->group);
-		iommu_group_put(gr3d->group);
-	}
+	host1x_client_iommu_detach(client, gr3d->group);
 free:
 	host1x_syncpt_free(client->syncpts[0]);
 put:
@@ -96,7 +84,6 @@ static int gr3d_exit(struct host1x_client *client)
 {
 	struct tegra_drm_client *drm = host1x_to_drm_client(client);
 	struct drm_device *dev = dev_get_drvdata(client->parent);
-	struct tegra_drm *tegra = dev->dev_private;
 	struct gr3d *gr3d = to_gr3d(drm);
 	int err;
 
@@ -104,14 +91,10 @@ static int gr3d_exit(struct host1x_client *client)
 	if (err < 0)
 		return err;
 
+	host1x_client_iommu_detach(client, gr3d->group);
 	host1x_syncpt_free(client->syncpts[0]);
 	host1x_channel_put(gr3d->channel);
 
-	if (gr3d->group) {
-		iommu_detach_group(tegra->domain, gr3d->group);
-		iommu_group_put(gr3d->group);
-	}
-
 	return 0;
 }
 
-- 
cgit v1.1


From acc6a3a9afdd4e0537342012656cdb5c4a3127c5 Mon Sep 17 00:00:00 2001
From: Dmitry Osipenko <digetx@gmail.com>
Date: Fri, 4 May 2018 17:39:58 +0300
Subject: drm/tegra: dc: Enable plane scaling filters

Currently resized plane produces a "pixelated" image which doesn't look
nice, especially in a case of a video overlay. Enable scaling filters that
significantly improve image quality of a scaled overlay.

Signed-off-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/dc.c | 81 ++++++++++++++++++++++++++++++++++++++++++++++
 drivers/gpu/drm/tegra/dc.h |  7 ++++
 2 files changed, 88 insertions(+)

diff --git a/drivers/gpu/drm/tegra/dc.c b/drivers/gpu/drm/tegra/dc.c
index 3e7ec39..a4dd866 100644
--- a/drivers/gpu/drm/tegra/dc.c
+++ b/drivers/gpu/drm/tegra/dc.c
@@ -224,6 +224,39 @@ static void tegra_plane_setup_blending(struct tegra_plane *plane,
 	tegra_plane_writel(plane, value, DC_WIN_BLEND_LAYER_CONTROL);
 }
 
+static bool
+tegra_plane_use_horizontal_filtering(struct tegra_plane *plane,
+				     const struct tegra_dc_window *window)
+{
+	struct tegra_dc *dc = plane->dc;
+
+	if (window->src.w == window->dst.w)
+		return false;
+
+	if (plane->index == 0 && dc->soc->has_win_a_without_filters)
+		return false;
+
+	return true;
+}
+
+static bool
+tegra_plane_use_vertical_filtering(struct tegra_plane *plane,
+				   const struct tegra_dc_window *window)
+{
+	struct tegra_dc *dc = plane->dc;
+
+	if (window->src.h == window->dst.h)
+		return false;
+
+	if (plane->index == 0 && dc->soc->has_win_a_without_filters)
+		return false;
+
+	if (plane->index == 2 && dc->soc->has_win_c_without_vert_filter)
+		return false;
+
+	return true;
+}
+
 static void tegra_dc_setup_window(struct tegra_plane *plane,
 				  const struct tegra_dc_window *window)
 {
@@ -361,6 +394,44 @@ static void tegra_dc_setup_window(struct tegra_plane *plane,
 	if (window->bottom_up)
 		value |= V_DIRECTION;
 
+	if (tegra_plane_use_horizontal_filtering(plane, window)) {
+		/*
+		 * Enable horizontal 6-tap filter and set filtering
+		 * coefficients to the default values defined in TRM.
+		 */
+		tegra_plane_writel(plane, 0x00008000, DC_WIN_H_FILTER_P(0));
+		tegra_plane_writel(plane, 0x3e087ce1, DC_WIN_H_FILTER_P(1));
+		tegra_plane_writel(plane, 0x3b117ac1, DC_WIN_H_FILTER_P(2));
+		tegra_plane_writel(plane, 0x591b73aa, DC_WIN_H_FILTER_P(3));
+		tegra_plane_writel(plane, 0x57256d9a, DC_WIN_H_FILTER_P(4));
+		tegra_plane_writel(plane, 0x552f668b, DC_WIN_H_FILTER_P(5));
+		tegra_plane_writel(plane, 0x73385e8b, DC_WIN_H_FILTER_P(6));
+		tegra_plane_writel(plane, 0x72435583, DC_WIN_H_FILTER_P(7));
+		tegra_plane_writel(plane, 0x714c4c8b, DC_WIN_H_FILTER_P(8));
+		tegra_plane_writel(plane, 0x70554393, DC_WIN_H_FILTER_P(9));
+		tegra_plane_writel(plane, 0x715e389b, DC_WIN_H_FILTER_P(10));
+		tegra_plane_writel(plane, 0x71662faa, DC_WIN_H_FILTER_P(11));
+		tegra_plane_writel(plane, 0x536d25ba, DC_WIN_H_FILTER_P(12));
+		tegra_plane_writel(plane, 0x55731bca, DC_WIN_H_FILTER_P(13));
+		tegra_plane_writel(plane, 0x387a11d9, DC_WIN_H_FILTER_P(14));
+		tegra_plane_writel(plane, 0x3c7c08f1, DC_WIN_H_FILTER_P(15));
+
+		value |= H_FILTER;
+	}
+
+	if (tegra_plane_use_vertical_filtering(plane, window)) {
+		unsigned int i, k;
+
+		/*
+		 * Enable vertical 2-tap filter and set filtering
+		 * coefficients to the default values defined in TRM.
+		 */
+		for (i = 0, k = 128; i < 16; i++, k -= 8)
+			tegra_plane_writel(plane, k, DC_WIN_V_FILTER_P(i));
+
+		value |= V_FILTER;
+	}
+
 	tegra_plane_writel(plane, value, DC_WIN_WIN_OPTIONS);
 
 	if (dc->soc->supports_blending)
@@ -1951,6 +2022,8 @@ static const struct tegra_dc_soc_info tegra20_dc_soc_info = {
 	.num_overlay_formats = ARRAY_SIZE(tegra20_overlay_formats),
 	.overlay_formats = tegra20_overlay_formats,
 	.modifiers = tegra20_modifiers,
+	.has_win_a_without_filters = true,
+	.has_win_c_without_vert_filter = true,
 };
 
 static const struct tegra_dc_soc_info tegra30_dc_soc_info = {
@@ -1968,6 +2041,8 @@ static const struct tegra_dc_soc_info tegra30_dc_soc_info = {
 	.num_overlay_formats = ARRAY_SIZE(tegra20_overlay_formats),
 	.overlay_formats = tegra20_overlay_formats,
 	.modifiers = tegra20_modifiers,
+	.has_win_a_without_filters = false,
+	.has_win_c_without_vert_filter = false,
 };
 
 static const struct tegra_dc_soc_info tegra114_dc_soc_info = {
@@ -1985,6 +2060,8 @@ static const struct tegra_dc_soc_info tegra114_dc_soc_info = {
 	.num_overlay_formats = ARRAY_SIZE(tegra114_overlay_formats),
 	.overlay_formats = tegra114_overlay_formats,
 	.modifiers = tegra20_modifiers,
+	.has_win_a_without_filters = false,
+	.has_win_c_without_vert_filter = false,
 };
 
 static const struct tegra_dc_soc_info tegra124_dc_soc_info = {
@@ -2002,6 +2079,8 @@ static const struct tegra_dc_soc_info tegra124_dc_soc_info = {
 	.num_overlay_formats = ARRAY_SIZE(tegra124_overlay_formats),
 	.overlay_formats = tegra124_overlay_formats,
 	.modifiers = tegra124_modifiers,
+	.has_win_a_without_filters = false,
+	.has_win_c_without_vert_filter = false,
 };
 
 static const struct tegra_dc_soc_info tegra210_dc_soc_info = {
@@ -2019,6 +2098,8 @@ static const struct tegra_dc_soc_info tegra210_dc_soc_info = {
 	.num_overlay_formats = ARRAY_SIZE(tegra114_overlay_formats),
 	.overlay_formats = tegra114_overlay_formats,
 	.modifiers = tegra124_modifiers,
+	.has_win_a_without_filters = false,
+	.has_win_c_without_vert_filter = false,
 };
 
 static const struct tegra_windowgroup_soc tegra186_dc_wgrps[] = {
diff --git a/drivers/gpu/drm/tegra/dc.h b/drivers/gpu/drm/tegra/dc.h
index 7be786f..556321f 100644
--- a/drivers/gpu/drm/tegra/dc.h
+++ b/drivers/gpu/drm/tegra/dc.h
@@ -67,6 +67,8 @@ struct tegra_dc_soc_info {
 	const u32 *overlay_formats;
 	unsigned int num_overlay_formats;
 	const u64 *modifiers;
+	bool has_win_a_without_filters;
+	bool has_win_c_without_vert_filter;
 };
 
 struct tegra_dc {
@@ -553,6 +555,9 @@ int tegra_dc_rgb_exit(struct tegra_dc *dc);
 #define  THREAD_NUM(x) (((x) & 0x1f) << 1)
 #define  THREAD_GROUP_ENABLE (1 << 0)
 
+#define DC_WIN_H_FILTER_P(p)			(0x601 + (p))
+#define DC_WIN_V_FILTER_P(p)			(0x619 + (p))
+
 #define DC_WIN_CSC_YOF				0x611
 #define DC_WIN_CSC_KYRGB			0x612
 #define DC_WIN_CSC_KUR				0x613
@@ -566,6 +571,8 @@ int tegra_dc_rgb_exit(struct tegra_dc *dc);
 #define H_DIRECTION  (1 <<  0)
 #define V_DIRECTION  (1 <<  2)
 #define COLOR_EXPAND (1 <<  6)
+#define H_FILTER     (1 <<  8)
+#define V_FILTER     (1 << 10)
 #define CSC_ENABLE   (1 << 18)
 #define WIN_ENABLE   (1 << 30)
 
-- 
cgit v1.1


From 3dae08bc076b93487ed2df50bcfa892113e89d9d Mon Sep 17 00:00:00 2001
From: Dmitry Osipenko <digetx@gmail.com>
Date: Fri, 4 May 2018 17:39:59 +0300
Subject: drm/tegra: plane: Implement zpos plane property for older Tegras

Older Tegra's do not support plane's Z position handling in hardware,
but the hardware provides knobs to implement it in software.

Signed-off-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/dc.c    | 134 +++++++++++++++++++++--------
 drivers/gpu/drm/tegra/plane.c | 193 ++++++++++++++++++++++++++++++------------
 drivers/gpu/drm/tegra/plane.h |  13 +--
 3 files changed, 244 insertions(+), 96 deletions(-)

diff --git a/drivers/gpu/drm/tegra/dc.c b/drivers/gpu/drm/tegra/dc.c
index a4dd866..51581d9 100644
--- a/drivers/gpu/drm/tegra/dc.c
+++ b/drivers/gpu/drm/tegra/dc.c
@@ -163,28 +163,89 @@ static void tegra_plane_setup_blending_legacy(struct tegra_plane *plane)
 			 BLEND_COLOR_KEY_NONE;
 	u32 blendnokey = BLEND_WEIGHT1(255) | BLEND_WEIGHT0(255);
 	struct tegra_plane_state *state;
+	u32 blending[2];
 	unsigned int i;
 
+	/* disable blending for non-overlapping case */
+	tegra_plane_writel(plane, blendnokey, DC_WIN_BLEND_NOKEY);
+	tegra_plane_writel(plane, foreground, DC_WIN_BLEND_1WIN);
+
 	state = to_tegra_plane_state(plane->base.state);
 
-	/* alpha contribution is 1 minus sum of overlapping windows */
-	for (i = 0; i < 3; i++) {
-		if (state->dependent[i])
-			background[i] |= BLEND_CONTROL_DEPENDENT;
-	}
+	if (state->opaque) {
+		/*
+		 * Since custom fix-weight blending isn't utilized and weight
+		 * of top window is set to max, we can enforce dependent
+		 * blending which in this case results in transparent bottom
+		 * window if top window is opaque and if top window enables
+		 * alpha blending, then bottom window is getting alpha value
+		 * of 1 minus the sum of alpha components of the overlapping
+		 * plane.
+		 */
+		background[0] |= BLEND_CONTROL_DEPENDENT;
+		background[1] |= BLEND_CONTROL_DEPENDENT;
 
-	/* enable alpha blending if pixel format has an alpha component */
-	if (!state->opaque)
+		/*
+		 * The region where three windows overlap is the intersection
+		 * of the two regions where two windows overlap. It contributes
+		 * to the area if all of the windows on top of it have an alpha
+		 * component.
+		 */
+		switch (state->base.normalized_zpos) {
+		case 0:
+			if (state->blending[0].alpha &&
+			    state->blending[1].alpha)
+				background[2] |= BLEND_CONTROL_DEPENDENT;
+			break;
+
+		case 1:
+			background[2] |= BLEND_CONTROL_DEPENDENT;
+			break;
+		}
+	} else {
+		/*
+		 * Enable alpha blending if pixel format has an alpha
+		 * component.
+		 */
 		foreground |= BLEND_CONTROL_ALPHA;
 
-	/*
-	 * Disable blending and assume Window A is the bottom-most window,
-	 * Window C is the top-most window and Window B is in the middle.
-	 */
-	tegra_plane_writel(plane, blendnokey, DC_WIN_BLEND_NOKEY);
-	tegra_plane_writel(plane, foreground, DC_WIN_BLEND_1WIN);
+		/*
+		 * If any of the windows on top of this window is opaque, it
+		 * will completely conceal this window within that area. If
+		 * top window has an alpha component, it is blended over the
+		 * bottom window.
+		 */
+		for (i = 0; i < 2; i++) {
+			if (state->blending[i].alpha &&
+			    state->blending[i].top)
+				background[i] |= BLEND_CONTROL_DEPENDENT;
+		}
 
-	switch (plane->index) {
+		switch (state->base.normalized_zpos) {
+		case 0:
+			if (state->blending[0].alpha &&
+			    state->blending[1].alpha)
+				background[2] |= BLEND_CONTROL_DEPENDENT;
+			break;
+
+		case 1:
+			/*
+			 * When both middle and topmost windows have an alpha,
+			 * these windows a mixed together and then the result
+			 * is blended over the bottom window.
+			 */
+			if (state->blending[0].alpha &&
+			    state->blending[0].top)
+				background[2] |= BLEND_CONTROL_ALPHA;
+
+			if (state->blending[1].alpha &&
+			    state->blending[1].top)
+				background[2] |= BLEND_CONTROL_ALPHA;
+			break;
+		}
+	}
+
+	switch (state->base.normalized_zpos) {
 	case 0:
 		tegra_plane_writel(plane, background[0], DC_WIN_BLEND_2WIN_X);
 		tegra_plane_writel(plane, background[1], DC_WIN_BLEND_2WIN_Y);
@@ -192,8 +253,21 @@ static void tegra_plane_setup_blending_legacy(struct tegra_plane *plane)
 		break;
 
 	case 1:
-		tegra_plane_writel(plane, foreground, DC_WIN_BLEND_2WIN_X);
-		tegra_plane_writel(plane, background[1], DC_WIN_BLEND_2WIN_Y);
+		/*
+		 * If window B / C is topmost, then X / Y registers are
+		 * matching the order of blending[...] state indices,
+		 * otherwise a swap is required.
+		 */
+		if (!state->blending[0].top && state->blending[1].top) {
+			blending[0] = foreground;
+			blending[1] = background[1];
+		} else {
+			blending[0] = background[0];
+			blending[1] = foreground;
+		}
+
+		tegra_plane_writel(plane, blending[0], DC_WIN_BLEND_2WIN_X);
+		tegra_plane_writel(plane, blending[1], DC_WIN_BLEND_2WIN_Y);
 		tegra_plane_writel(plane, background[2], DC_WIN_BLEND_3WIN_XY);
 		break;
 
@@ -525,14 +599,14 @@ static int tegra_plane_atomic_check(struct drm_plane *plane,
 	struct tegra_bo_tiling *tiling = &plane_state->tiling;
 	struct tegra_plane *tegra = to_tegra_plane(plane);
 	struct tegra_dc *dc = to_tegra_dc(state->crtc);
-	unsigned int format;
 	int err;
 
 	/* no need for further checks if the plane is being disabled */
 	if (!state->crtc)
 		return 0;
 
-	err = tegra_plane_format(state->fb->format->format, &format,
+	err = tegra_plane_format(state->fb->format->format,
+				 &plane_state->format,
 				 &plane_state->swap);
 	if (err < 0)
 		return err;
@@ -544,21 +618,11 @@ static int tegra_plane_atomic_check(struct drm_plane *plane,
 	 * be emulated by disabling alpha blending for the plane.
 	 */
 	if (!dc->soc->supports_blending) {
-		if (!tegra_plane_format_has_alpha(format)) {
-			err = tegra_plane_format_get_alpha(format, &format);
-			if (err < 0)
-				return err;
-
-			plane_state->opaque = true;
-		} else {
-			plane_state->opaque = false;
-		}
-
-		tegra_plane_check_dependent(tegra, plane_state);
+		err = tegra_plane_setup_legacy_state(tegra, plane_state);
+		if (err < 0)
+			return err;
 	}
 
-	plane_state->format = format;
-
 	err = tegra_fb_get_tiling(state->fb, tiling);
 	if (err < 0)
 		return err;
@@ -710,9 +774,7 @@ static struct drm_plane *tegra_primary_plane_create(struct drm_device *drm,
 	}
 
 	drm_plane_helper_add(&plane->base, &tegra_plane_helper_funcs);
-
-	if (dc->soc->supports_blending)
-		drm_plane_create_zpos_property(&plane->base, 0, 0, 255);
+	drm_plane_create_zpos_property(&plane->base, plane->index, 0, 255);
 
 	return &plane->base;
 }
@@ -989,9 +1051,7 @@ static struct drm_plane *tegra_dc_overlay_plane_create(struct drm_device *drm,
 	}
 
 	drm_plane_helper_add(&plane->base, &tegra_plane_helper_funcs);
-
-	if (dc->soc->supports_blending)
-		drm_plane_create_zpos_property(&plane->base, 0, 0, 255);
+	drm_plane_create_zpos_property(&plane->base, plane->index, 0, 255);
 
 	return &plane->base;
 }
diff --git a/drivers/gpu/drm/tegra/plane.c b/drivers/gpu/drm/tegra/plane.c
index 176ef46..0406c2e 100644
--- a/drivers/gpu/drm/tegra/plane.c
+++ b/drivers/gpu/drm/tegra/plane.c
@@ -23,6 +23,7 @@ static void tegra_plane_destroy(struct drm_plane *plane)
 
 static void tegra_plane_reset(struct drm_plane *plane)
 {
+	struct tegra_plane *p = to_tegra_plane(plane);
 	struct tegra_plane_state *state;
 
 	if (plane->state)
@@ -35,6 +36,8 @@ static void tegra_plane_reset(struct drm_plane *plane)
 	if (state) {
 		plane->state = &state->base;
 		plane->state->plane = plane;
+		plane->state->zpos = p->index;
+		plane->state->normalized_zpos = p->index;
 	}
 }
 
@@ -55,8 +58,8 @@ tegra_plane_atomic_duplicate_state(struct drm_plane *plane)
 	copy->swap = state->swap;
 	copy->opaque = state->opaque;
 
-	for (i = 0; i < 3; i++)
-		copy->dependent[i] = state->dependent[i];
+	for (i = 0; i < 2; i++)
+		copy->blending[i] = state->blending[i];
 
 	return &copy->base;
 }
@@ -267,24 +270,8 @@ static bool __drm_format_has_alpha(u32 format)
 	return false;
 }
 
-/*
- * This is applicable to Tegra20 and Tegra30 only where the opaque formats can
- * be emulated using the alpha formats and alpha blending disabled.
- */
-bool tegra_plane_format_has_alpha(unsigned int format)
-{
-	switch (format) {
-	case WIN_COLOR_DEPTH_B5G5R5A1:
-	case WIN_COLOR_DEPTH_A1B5G5R5:
-	case WIN_COLOR_DEPTH_R8G8B8A8:
-	case WIN_COLOR_DEPTH_B8G8R8A8:
-		return true;
-	}
-
-	return false;
-}
-
-int tegra_plane_format_get_alpha(unsigned int opaque, unsigned int *alpha)
+static int tegra_plane_format_get_alpha(unsigned int opaque,
+					unsigned int *alpha)
 {
 	if (tegra_plane_format_is_yuv(opaque, NULL)) {
 		*alpha = opaque;
@@ -316,6 +303,67 @@ int tegra_plane_format_get_alpha(unsigned int opaque, unsigned int *alpha)
 	return -EINVAL;
 }
 
+/*
+ * This is applicable to Tegra20 and Tegra30 only where the opaque formats can
+ * be emulated using the alpha formats and alpha blending disabled.
+ */
+static int tegra_plane_setup_opacity(struct tegra_plane *tegra,
+				     struct tegra_plane_state *state)
+{
+	unsigned int format;
+	int err;
+
+	switch (state->format) {
+	case WIN_COLOR_DEPTH_B5G5R5A1:
+	case WIN_COLOR_DEPTH_A1B5G5R5:
+	case WIN_COLOR_DEPTH_R8G8B8A8:
+	case WIN_COLOR_DEPTH_B8G8R8A8:
+		state->opaque = false;
+		break;
+
+	default:
+		err = tegra_plane_format_get_alpha(state->format, &format);
+		if (err < 0)
+			return err;
+
+		state->format = format;
+		state->opaque = true;
+		break;
+	}
+
+	return 0;
+}
+
+static int tegra_plane_check_transparency(struct tegra_plane *tegra,
+					  struct tegra_plane_state *state)
+{
+	struct drm_plane_state *old, *plane_state;
+	struct drm_plane *plane;
+
+	old = drm_atomic_get_old_plane_state(state->base.state, &tegra->base);
+
+	/* check if zpos / transparency changed */
+	if (old->normalized_zpos == state->base.normalized_zpos &&
+	    to_tegra_plane_state(old)->opaque == state->opaque)
+		return 0;
+
+	/* include all sibling planes into this commit */
+	drm_for_each_plane(plane, tegra->base.dev) {
+		struct tegra_plane *p = to_tegra_plane(plane);
+
+		/* skip this plane and planes on different CRTCs */
+		if (p == tegra || p->dc != tegra->dc)
+			continue;
+
+		plane_state = drm_atomic_get_plane_state(state->base.state,
+							 plane);
+		if (IS_ERR(plane_state))
+			return PTR_ERR(plane_state);
+	}
+
+	return 1;
+}
+
 static unsigned int tegra_plane_get_overlap_index(struct tegra_plane *plane,
 						  struct tegra_plane *other)
 {
@@ -336,61 +384,98 @@ static unsigned int tegra_plane_get_overlap_index(struct tegra_plane *plane,
 	return index;
 }
 
-void tegra_plane_check_dependent(struct tegra_plane *tegra,
-				 struct tegra_plane_state *state)
+static void tegra_plane_update_transparency(struct tegra_plane *tegra,
+					    struct tegra_plane_state *state)
 {
-	struct drm_plane_state *old, *new;
+	struct drm_plane_state *new;
 	struct drm_plane *plane;
-	unsigned int zpos[2];
 	unsigned int i;
 
-	for (i = 0; i < 2; i++)
-		zpos[i] = 0;
-
-	for_each_oldnew_plane_in_state(state->base.state, plane, old, new, i) {
+	for_each_new_plane_in_state(state->base.state, plane, new, i) {
 		struct tegra_plane *p = to_tegra_plane(plane);
 		unsigned index;
 
 		/* skip this plane and planes on different CRTCs */
-		if (p == tegra || new->crtc != state->base.crtc)
+		if (p == tegra || p->dc != tegra->dc)
 			continue;
 
 		index = tegra_plane_get_overlap_index(tegra, p);
 
-		state->dependent[index] = false;
+		if (new->fb && __drm_format_has_alpha(new->fb->format->format))
+			state->blending[index].alpha = true;
+		else
+			state->blending[index].alpha = false;
+
+		if (new->normalized_zpos > state->base.normalized_zpos)
+			state->blending[index].top = true;
+		else
+			state->blending[index].top = false;
 
 		/*
-		 * If any of the other planes is on top of this plane and uses
-		 * a format with an alpha component, mark this plane as being
-		 * dependent, meaning it's alpha value will be 1 minus the sum
-		 * of alpha components of the overlapping planes.
+		 * Missing framebuffer means that plane is disabled, in this
+		 * case mark B / C window as top to be able to differentiate
+		 * windows indices order in regards to zPos for the middle
+		 * window X / Y registers programming.
 		 */
-		if (p->index > tegra->index) {
-			if (__drm_format_has_alpha(new->fb->format->format))
-				state->dependent[index] = true;
-
-			/* keep track of the Z position */
-			zpos[index] = p->index;
-		}
+		if (!new->fb)
+			state->blending[index].top = (index == 1);
 	}
+}
+
+static int tegra_plane_setup_transparency(struct tegra_plane *tegra,
+					  struct tegra_plane_state *state)
+{
+	struct tegra_plane_state *tegra_state;
+	struct drm_plane_state *new;
+	struct drm_plane *plane;
+	int err;
 
 	/*
-	 * The region where three windows overlap is the intersection of the
-	 * two regions where two windows overlap. It contributes to the area
-	 * if any of the windows on top of it have an alpha component.
+	 * If planes zpos / transparency changed, sibling planes blending
+	 * state may require adjustment and in this case they will be included
+	 * into this atom commit, otherwise blending state is unchanged.
 	 */
-	for (i = 0; i < 2; i++)
-		state->dependent[2] = state->dependent[2] ||
-				      state->dependent[i];
+	err = tegra_plane_check_transparency(tegra, state);
+	if (err <= 0)
+		return err;
 
 	/*
-	 * However, if any of the windows on top of this window is opaque, it
-	 * will completely conceal this window within that area, so avoid the
-	 * window from contributing to the area.
+	 * All planes are now in the atomic state, walk them up and update
+	 * transparency state for each plane.
 	 */
-	for (i = 0; i < 2; i++) {
-		if (zpos[i] > tegra->index)
-			state->dependent[2] = state->dependent[2] &&
-					      state->dependent[i];
+	drm_for_each_plane(plane, tegra->base.dev) {
+		struct tegra_plane *p = to_tegra_plane(plane);
+
+		/* skip planes on different CRTCs */
+		if (p->dc != tegra->dc)
+			continue;
+
+		new = drm_atomic_get_new_plane_state(state->base.state, plane);
+		tegra_state = to_tegra_plane_state(new);
+
+		/*
+		 * There is no need to update blending state for the disabled
+		 * plane.
+		 */
+		if (new->fb)
+			tegra_plane_update_transparency(p, tegra_state);
 	}
+
+	return 0;
+}
+
+int tegra_plane_setup_legacy_state(struct tegra_plane *tegra,
+				   struct tegra_plane_state *state)
+{
+	int err;
+
+	err = tegra_plane_setup_opacity(tegra, state);
+	if (err < 0)
+		return err;
+
+	err = tegra_plane_setup_transparency(tegra, state);
+	if (err < 0)
+		return err;
+
+	return 0;
 }
diff --git a/drivers/gpu/drm/tegra/plane.h b/drivers/gpu/drm/tegra/plane.h
index 6938719..7360ddf 100644
--- a/drivers/gpu/drm/tegra/plane.h
+++ b/drivers/gpu/drm/tegra/plane.h
@@ -34,6 +34,11 @@ static inline struct tegra_plane *to_tegra_plane(struct drm_plane *plane)
 	return container_of(plane, struct tegra_plane, base);
 }
 
+struct tegra_plane_legacy_blending_state {
+	bool alpha;
+	bool top;
+};
+
 struct tegra_plane_state {
 	struct drm_plane_state base;
 
@@ -42,8 +47,8 @@ struct tegra_plane_state {
 	u32 swap;
 
 	/* used for legacy blending support only */
+	struct tegra_plane_legacy_blending_state blending[2];
 	bool opaque;
-	bool dependent[3];
 };
 
 static inline struct tegra_plane_state *
@@ -62,9 +67,7 @@ int tegra_plane_state_add(struct tegra_plane *plane,
 
 int tegra_plane_format(u32 fourcc, u32 *format, u32 *swap);
 bool tegra_plane_format_is_yuv(unsigned int format, bool *planar);
-bool tegra_plane_format_has_alpha(unsigned int format);
-int tegra_plane_format_get_alpha(unsigned int opaque, unsigned int *alpha);
-void tegra_plane_check_dependent(struct tegra_plane *tegra,
-				 struct tegra_plane_state *state);
+int tegra_plane_setup_legacy_state(struct tegra_plane *tegra,
+				   struct tegra_plane_state *state);
 
 #endif /* TEGRA_PLANE_H */
-- 
cgit v1.1


From a43d0a00ea58a665905d94e8ab469ff888b1b0e1 Mon Sep 17 00:00:00 2001
From: Dmitry Osipenko <digetx@gmail.com>
Date: Fri, 4 May 2018 17:40:00 +0300
Subject: drm/tegra: dc: Rename supports_blending to has_legacy_blending

Older Tegra chips do support blending as well. Rename the SoC info entry
.supports_blending to .has_legacy_blending to eliminate the confusion.

Signed-off-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/dc.c | 20 ++++++++++----------
 drivers/gpu/drm/tegra/dc.h |  2 +-
 2 files changed, 11 insertions(+), 11 deletions(-)

diff --git a/drivers/gpu/drm/tegra/dc.c b/drivers/gpu/drm/tegra/dc.c
index 51581d9..31e12a9 100644
--- a/drivers/gpu/drm/tegra/dc.c
+++ b/drivers/gpu/drm/tegra/dc.c
@@ -508,10 +508,10 @@ static void tegra_dc_setup_window(struct tegra_plane *plane,
 
 	tegra_plane_writel(plane, value, DC_WIN_WIN_OPTIONS);
 
-	if (dc->soc->supports_blending)
-		tegra_plane_setup_blending(plane, window);
-	else
+	if (dc->soc->has_legacy_blending)
 		tegra_plane_setup_blending_legacy(plane);
+	else
+		tegra_plane_setup_blending(plane, window);
 }
 
 static const u32 tegra20_primary_formats[] = {
@@ -617,7 +617,7 @@ static int tegra_plane_atomic_check(struct drm_plane *plane,
 	 * the corresponding opaque formats. However, the opaque formats can
 	 * be emulated by disabling alpha blending for the plane.
 	 */
-	if (!dc->soc->supports_blending) {
+	if (dc->soc->has_legacy_blending) {
 		err = tegra_plane_setup_legacy_state(tegra, plane_state);
 		if (err < 0)
 			return err;
@@ -2072,7 +2072,7 @@ static const struct tegra_dc_soc_info tegra20_dc_soc_info = {
 	.supports_interlacing = false,
 	.supports_cursor = false,
 	.supports_block_linear = false,
-	.supports_blending = false,
+	.has_legacy_blending = true,
 	.pitch_align = 8,
 	.has_powergate = false,
 	.coupled_pm = true,
@@ -2091,7 +2091,7 @@ static const struct tegra_dc_soc_info tegra30_dc_soc_info = {
 	.supports_interlacing = false,
 	.supports_cursor = false,
 	.supports_block_linear = false,
-	.supports_blending = false,
+	.has_legacy_blending = true,
 	.pitch_align = 8,
 	.has_powergate = false,
 	.coupled_pm = false,
@@ -2110,7 +2110,7 @@ static const struct tegra_dc_soc_info tegra114_dc_soc_info = {
 	.supports_interlacing = false,
 	.supports_cursor = false,
 	.supports_block_linear = false,
-	.supports_blending = false,
+	.has_legacy_blending = true,
 	.pitch_align = 64,
 	.has_powergate = true,
 	.coupled_pm = false,
@@ -2129,7 +2129,7 @@ static const struct tegra_dc_soc_info tegra124_dc_soc_info = {
 	.supports_interlacing = true,
 	.supports_cursor = true,
 	.supports_block_linear = true,
-	.supports_blending = true,
+	.has_legacy_blending = false,
 	.pitch_align = 64,
 	.has_powergate = true,
 	.coupled_pm = false,
@@ -2148,7 +2148,7 @@ static const struct tegra_dc_soc_info tegra210_dc_soc_info = {
 	.supports_interlacing = true,
 	.supports_cursor = true,
 	.supports_block_linear = true,
-	.supports_blending = true,
+	.has_legacy_blending = false,
 	.pitch_align = 64,
 	.has_powergate = true,
 	.coupled_pm = false,
@@ -2201,7 +2201,7 @@ static const struct tegra_dc_soc_info tegra186_dc_soc_info = {
 	.supports_interlacing = true,
 	.supports_cursor = true,
 	.supports_block_linear = true,
-	.supports_blending = true,
+	.has_legacy_blending = false,
 	.pitch_align = 64,
 	.has_powergate = false,
 	.coupled_pm = false,
diff --git a/drivers/gpu/drm/tegra/dc.h b/drivers/gpu/drm/tegra/dc.h
index 556321f..e96f582 100644
--- a/drivers/gpu/drm/tegra/dc.h
+++ b/drivers/gpu/drm/tegra/dc.h
@@ -55,7 +55,7 @@ struct tegra_dc_soc_info {
 	bool supports_interlacing;
 	bool supports_cursor;
 	bool supports_block_linear;
-	bool supports_blending;
+	bool has_legacy_blending;
 	unsigned int pitch_align;
 	bool has_powergate;
 	bool coupled_pm;
-- 
cgit v1.1


From c5fb5426dda897fbfeb3ddba81c9811f1178132c Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Sat, 10 Mar 2018 05:15:18 +0800
Subject: drm/amdgpu/gfx9: Update golden setting for gfx9_0.

Update golden_settings_gc_9_0[].

Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 17 +++++------------
 1 file changed, 5 insertions(+), 12 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index fc19118..b05b7ae 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -72,29 +72,22 @@ MODULE_FIRMWARE("amdgpu/raven_rlc.bin");
 
 static const struct soc15_reg_golden golden_settings_gc_9_0[] =
 {
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCPC_UTCL1_CNTL, 0x08000000, 0x08000080),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCPF_UTCL1_CNTL, 0x08000000, 0x08000080),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCPG_UTCL1_CNTL, 0x08000000, 0x08000080),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmDB_DEBUG2, 0xf00fffff, 0x00000420),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmGB_GPU_ID, 0x0000000f, 0x00000000),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmIA_UTCL1_CNTL, 0x08000000, 0x08000080),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_BINNER_EVENT_CNTL_3, 0x00000003, 0x82400024),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_ENHANCE, 0x3fffffff, 0x00000001),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_LINE_STIPPLE_STATE, 0x0000ff0f, 0x00000000),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_GPM_UTCL1_CNTL_0, 0x08000000, 0x08000080),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_GPM_UTCL1_CNTL_1, 0x08000000, 0x08000080),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_GPM_UTCL1_CNTL_2, 0x08000000, 0x08000080),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_PREWALKER_UTCL1_CNTL, 0x08000000, 0x08000080),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_SPM_UTCL1_CNTL, 0x08000000, 0x08000080),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmSH_MEM_CONFIG, 0x00001000, 0x00001000),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_CONFIG_CNTL_1, 0x0000000f, 0x01000107),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_RESOURCE_RESERVE_CU_0, 0x0007ffff, 0x00000800),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_RESOURCE_RESERVE_CU_1, 0x0007ffff, 0x00000800),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_RESOURCE_RESERVE_EN_CU_0, 0x01ffffff, 0x0000ff87),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_RESOURCE_RESERVE_EN_CU_1, 0x01ffffff, 0x0000ff8f),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmSQC_CONFIG, 0x03000000, 0x020a2000),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTA_CNTL_AUX, 0xfffffeef, 0x010b0000),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTCP_CHAN_STEER_HI, 0xffffffff, 0x4a2c0e68),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTCP_CHAN_STEER_LO, 0xffffffff, 0xb5d3f197),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmVGT_CACHE_INVALIDATION, 0x3fff3af3, 0x19200000),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmVGT_GS_MAX_WAVE_ID, 0x00000fff, 0x000003ff),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmWD_UTCL1_CNTL, 0x08000000, 0x08000080)
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmVGT_GS_MAX_WAVE_ID, 0x00000fff, 0x000003ff)
 };
 
 static const struct soc15_reg_golden golden_settings_gc_9_0_vg10[] =
-- 
cgit v1.1


From 73aa1b9af5947f103913124f93ca19e6f3af1c1b Mon Sep 17 00:00:00 2001
From: Evan Quan <evan.quan@amd.com>
Date: Wed, 9 May 2018 10:57:53 +0800
Subject: drm/amd/powerplay: new framework to honour DAL clock limits

This is needed for vega12 and vega20 which do not support legacy
powerstate. With this new framework, the DAL clocks limits can also
be honored on these asics.

Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c | 9 +++++++++
 drivers/gpu/drm/amd/powerplay/hwmgr/pp_psm.c          | 7 +++++++
 drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h   | 2 ++
 drivers/gpu/drm/amd/powerplay/inc/hwmgr.h             | 2 ++
 4 files changed, 20 insertions(+)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c
index e411012..f5571e9f 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c
@@ -132,6 +132,15 @@ int phm_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 	return 0;
 }
 
+int phm_apply_clock_adjust_rules(struct pp_hwmgr *hwmgr)
+{
+	PHM_FUNC_CHECK(hwmgr);
+
+	if (hwmgr->hwmgr_func->apply_clocks_adjust_rules != NULL)
+		return hwmgr->hwmgr_func->apply_clocks_adjust_rules(hwmgr);
+	return 0;
+}
+
 int phm_powerdown_uvd(struct pp_hwmgr *hwmgr)
 {
 	PHM_FUNC_CHECK(hwmgr);
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/pp_psm.c b/drivers/gpu/drm/amd/powerplay/hwmgr/pp_psm.c
index 308bff2..2a2955c 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/pp_psm.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/pp_psm.c
@@ -265,6 +265,13 @@ int psm_adjust_power_state_dynamic(struct pp_hwmgr *hwmgr, bool skip,
 	if (skip)
 		return 0;
 
+	if (!hwmgr->ps)
+		/*
+		 * for vega12/vega20 which does not support power state manager
+		 * DAL clock limits should also be honoured
+		 */
+		phm_apply_clock_adjust_rules(hwmgr);
+
 	phm_display_configuration_changed(hwmgr);
 
 	if (hwmgr->ps)
diff --git a/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h b/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h
index 9bb8785..e029555 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h
@@ -410,6 +410,8 @@ extern int phm_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 				   struct pp_power_state *adjusted_ps,
 			     const struct pp_power_state *current_ps);
 
+extern int phm_apply_clock_adjust_rules(struct pp_hwmgr *hwmgr);
+
 extern int phm_force_dpm_levels(struct pp_hwmgr *hwmgr, enum amd_dpm_forced_level level);
 extern int phm_display_configuration_changed(struct pp_hwmgr *hwmgr);
 extern int phm_notify_smc_display_config_after_ps_adjustment(struct pp_hwmgr *hwmgr);
diff --git a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
index 3c321c7..9b6c6af 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
@@ -229,6 +229,8 @@ struct pp_hwmgr_func {
 				struct pp_power_state  *prequest_ps,
 			const struct pp_power_state *pcurrent_ps);
 
+	int (*apply_clocks_adjust_rules)(struct pp_hwmgr *hwmgr);
+
 	int (*force_dpm_level)(struct pp_hwmgr *hw_mgr,
 					enum amd_dpm_forced_level level);
 
-- 
cgit v1.1


From 11a89b431e41dcfaa4e7b9806233f60de905287b Mon Sep 17 00:00:00 2001
From: Evan Quan <evan.quan@amd.com>
Date: Wed, 9 May 2018 11:08:29 +0800
Subject: drm/amd/powerplay: add a framework for perfroming pre display
 configuration change settings

Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c | 10 ++++++++++
 drivers/gpu/drm/amd/powerplay/hwmgr/pp_psm.c          |  2 ++
 drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h   |  1 +
 drivers/gpu/drm/amd/powerplay/inc/hwmgr.h             |  1 +
 4 files changed, 14 insertions(+)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c
index f5571e9f..a0bb921 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c
@@ -170,6 +170,16 @@ int phm_disable_clock_power_gatings(struct pp_hwmgr *hwmgr)
 	return 0;
 }
 
+int phm_pre_display_configuration_changed(struct pp_hwmgr *hwmgr)
+{
+	PHM_FUNC_CHECK(hwmgr);
+
+	if (NULL != hwmgr->hwmgr_func->pre_display_config_changed)
+		hwmgr->hwmgr_func->pre_display_config_changed(hwmgr);
+
+	return 0;
+
+}
 
 int phm_display_configuration_changed(struct pp_hwmgr *hwmgr)
 {
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/pp_psm.c b/drivers/gpu/drm/amd/powerplay/hwmgr/pp_psm.c
index 2a2955c..0af13c1 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/pp_psm.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/pp_psm.c
@@ -272,6 +272,8 @@ int psm_adjust_power_state_dynamic(struct pp_hwmgr *hwmgr, bool skip,
 		 */
 		phm_apply_clock_adjust_rules(hwmgr);
 
+	phm_pre_display_configuration_changed(hwmgr);
+
 	phm_display_configuration_changed(hwmgr);
 
 	if (hwmgr->ps)
diff --git a/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h b/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h
index e029555..a202247 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h
@@ -413,6 +413,7 @@ extern int phm_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 extern int phm_apply_clock_adjust_rules(struct pp_hwmgr *hwmgr);
 
 extern int phm_force_dpm_levels(struct pp_hwmgr *hwmgr, enum amd_dpm_forced_level level);
+extern int phm_pre_display_configuration_changed(struct pp_hwmgr *hwmgr);
 extern int phm_display_configuration_changed(struct pp_hwmgr *hwmgr);
 extern int phm_notify_smc_display_config_after_ps_adjustment(struct pp_hwmgr *hwmgr);
 extern int phm_register_irq_handlers(struct pp_hwmgr *hwmgr);
diff --git a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
index 9b6c6af..b99fb8a 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
@@ -254,6 +254,7 @@ struct pp_hwmgr_func {
 						const void *state);
 	int (*enable_clock_power_gating)(struct pp_hwmgr *hwmgr);
 	int (*notify_smc_display_config_after_ps_adjustment)(struct pp_hwmgr *hwmgr);
+	int (*pre_display_config_changed)(struct pp_hwmgr *hwmgr);
 	int (*display_config_changed)(struct pp_hwmgr *hwmgr);
 	int (*disable_clock_power_gating)(struct pp_hwmgr *hwmgr);
 	int (*update_clock_gatings)(struct pp_hwmgr *hwmgr,
-- 
cgit v1.1


From cc3a98cc6efd0e60e0ed547f9f76f5d4e23fb758 Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Fri, 11 May 2018 14:41:40 +0800
Subject: drm/amdgpu: Drop the unused header files in soc15.c.

Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/soc15.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index 9006576..f31df18 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -41,8 +41,6 @@
 #include "sdma1/sdma1_4_0_offset.h"
 #include "hdp/hdp_4_0_offset.h"
 #include "hdp/hdp_4_0_sh_mask.h"
-#include "mp/mp_9_0_offset.h"
-#include "mp/mp_9_0_sh_mask.h"
 #include "smuio/smuio_9_0_offset.h"
 #include "smuio/smuio_9_0_sh_mask.h"
 
-- 
cgit v1.1


From b6110c00ced26b66999eb00b90c35b767cd45da4 Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Tue, 6 Feb 2018 12:29:23 +0800
Subject: drm/amdgpu: Fix hardcoded base offset of vram pages

In gmc_v9_0_vram_gtt_location(),the vram_base_offset is hardcoded
to 0 in dGPU. Fix it by reading mmMC_VM_FB_OFFSET or return
zfb_phys_addr if ZFB is enabled.

Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
index 6cccf0e..7343069 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
@@ -693,10 +693,7 @@ static void gmc_v9_0_vram_gtt_location(struct amdgpu_device *adev,
 	amdgpu_device_vram_location(adev, &adev->gmc, base);
 	amdgpu_device_gart_location(adev, mc);
 	/* base offset of vram pages */
-	if (adev->flags & AMD_IS_APU)
-		adev->vm_manager.vram_base_offset = gfxhub_v1_0_get_mc_fb_offset(adev);
-	else
-		adev->vm_manager.vram_base_offset = 0;
+	adev->vm_manager.vram_base_offset = gfxhub_v1_0_get_mc_fb_offset(adev);
 }
 
 /**
-- 
cgit v1.1


From 5eb26e7ae16b8da302a361824b9c4a53a6f3ee0f Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Tue, 30 Jan 2018 10:59:23 +0800
Subject: drm/amd: Add vega20_ip_offset.h headerfile for vega20. (v2)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This headerfile contains vega20's ip base addresses.

v2: squash in MP1_BASE fix

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/include/vega20_ip_offset.h | 1050 ++++++++++++++++++++++++
 1 file changed, 1050 insertions(+)
 create mode 100644 drivers/gpu/drm/amd/include/vega20_ip_offset.h

diff --git a/drivers/gpu/drm/amd/include/vega20_ip_offset.h b/drivers/gpu/drm/amd/include/vega20_ip_offset.h
new file mode 100644
index 0000000..97db93c
--- /dev/null
+++ b/drivers/gpu/drm/amd/include/vega20_ip_offset.h
@@ -0,0 +1,1050 @@
+/*
+ * Copyright (C) 2018  Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included
+ * in all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
+ * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN
+ * AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ */
+#ifndef _vega20_ip_offset_HEADER
+#define _vega20_ip_offset_HEADER
+
+#define MAX_INSTANCE                                       6
+#define MAX_SEGMENT                                        6
+
+
+struct IP_BASE_INSTANCE
+{
+    unsigned int segment[MAX_SEGMENT];
+};
+
+struct IP_BASE
+{
+    struct IP_BASE_INSTANCE instance[MAX_INSTANCE];
+};
+
+
+static const struct IP_BASE ATHUB_BASE            ={ { { { 0x00000C20, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE CLK_BASE            ={ { { { 0x00016C00, 0x00016E00, 0x00017000, 0x00017200, 0x0001B000, 0x0001B200 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE DCE_BASE            ={ { { { 0x00000012, 0x000000C0, 0x000034C0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE DF_BASE            ={ { { { 0x00007000, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE FUSE_BASE            ={ { { { 0x00017400, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE GC_BASE            ={ { { { 0x00002000, 0x0000A000, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE HDP_BASE            ={ { { { 0x00000F20, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE MMHUB_BASE            ={ { { { 0x0001A000, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE MP0_BASE            ={ { { { 0x00016000, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE MP1_BASE            ={ { { { 0x00016000, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE NBIO_BASE            ={ { { { 0x00000000, 0x00000014, 0x00000D20, 0x00010400, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE OSSSYS_BASE            ={ { { { 0x000010A0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE SDMA0_BASE            ={ { { { 0x00001260, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE SDMA1_BASE            ={ { { { 0x00001860, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE SMUIO_BASE            ={ { { { 0x00016800, 0x00016A00, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE THM_BASE            ={ { { { 0x00016600, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE UMC_BASE            ={ { { { 0x00014000, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE UVD_BASE            ={ { { { 0x00007800, 0x00007E00, 0, 0, 0, 0 } },
+                                        { { 0, 0x00009000, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE VCE_BASE            ={ { { { 0x00008800, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE XDMA_BASE            ={ { { { 0x00003400, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE RSMU_BASE            ={ { { { 0x00012000, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+
+
+#define ATHUB_BASE__INST0_SEG0                     0x00000C20
+#define ATHUB_BASE__INST0_SEG1                     0
+#define ATHUB_BASE__INST0_SEG2                     0
+#define ATHUB_BASE__INST0_SEG3                     0
+#define ATHUB_BASE__INST0_SEG4                     0
+#define ATHUB_BASE__INST0_SEG5                     0
+
+#define ATHUB_BASE__INST1_SEG0                     0
+#define ATHUB_BASE__INST1_SEG1                     0
+#define ATHUB_BASE__INST1_SEG2                     0
+#define ATHUB_BASE__INST1_SEG3                     0
+#define ATHUB_BASE__INST1_SEG4                     0
+#define ATHUB_BASE__INST1_SEG5                     0
+
+#define ATHUB_BASE__INST2_SEG0                     0
+#define ATHUB_BASE__INST2_SEG1                     0
+#define ATHUB_BASE__INST2_SEG2                     0
+#define ATHUB_BASE__INST2_SEG3                     0
+#define ATHUB_BASE__INST2_SEG4                     0
+#define ATHUB_BASE__INST2_SEG5                     0
+
+#define ATHUB_BASE__INST3_SEG0                     0
+#define ATHUB_BASE__INST3_SEG1                     0
+#define ATHUB_BASE__INST3_SEG2                     0
+#define ATHUB_BASE__INST3_SEG3                     0
+#define ATHUB_BASE__INST3_SEG4                     0
+#define ATHUB_BASE__INST3_SEG5                     0
+
+#define ATHUB_BASE__INST4_SEG0                     0
+#define ATHUB_BASE__INST4_SEG1                     0
+#define ATHUB_BASE__INST4_SEG2                     0
+#define ATHUB_BASE__INST4_SEG3                     0
+#define ATHUB_BASE__INST4_SEG4                     0
+#define ATHUB_BASE__INST4_SEG5                     0
+
+#define ATHUB_BASE__INST5_SEG0                     0
+#define ATHUB_BASE__INST5_SEG1                     0
+#define ATHUB_BASE__INST5_SEG2                     0
+#define ATHUB_BASE__INST5_SEG3                     0
+#define ATHUB_BASE__INST5_SEG4                     0
+#define ATHUB_BASE__INST5_SEG5                     0
+
+#define CLK_BASE__INST0_SEG0                       0x00016C00
+#define CLK_BASE__INST0_SEG1                       0x00016E00
+#define CLK_BASE__INST0_SEG2                       0x00017000
+#define CLK_BASE__INST0_SEG3                       0x00017200
+#define CLK_BASE__INST0_SEG4                       0x0001B000
+#define CLK_BASE__INST0_SEG5                       0x0001B200
+
+#define CLK_BASE__INST1_SEG0                       0
+#define CLK_BASE__INST1_SEG1                       0
+#define CLK_BASE__INST1_SEG2                       0
+#define CLK_BASE__INST1_SEG3                       0
+#define CLK_BASE__INST1_SEG4                       0
+#define CLK_BASE__INST1_SEG5                       0
+
+#define CLK_BASE__INST2_SEG0                       0
+#define CLK_BASE__INST2_SEG1                       0
+#define CLK_BASE__INST2_SEG2                       0
+#define CLK_BASE__INST2_SEG3                       0
+#define CLK_BASE__INST2_SEG4                       0
+#define CLK_BASE__INST2_SEG5                       0
+
+#define CLK_BASE__INST3_SEG0                       0
+#define CLK_BASE__INST3_SEG1                       0
+#define CLK_BASE__INST3_SEG2                       0
+#define CLK_BASE__INST3_SEG3                       0
+#define CLK_BASE__INST3_SEG4                       0
+#define CLK_BASE__INST3_SEG5                       0
+
+#define CLK_BASE__INST4_SEG0                       0
+#define CLK_BASE__INST4_SEG1                       0
+#define CLK_BASE__INST4_SEG2                       0
+#define CLK_BASE__INST4_SEG3                       0
+#define CLK_BASE__INST4_SEG4                       0
+#define CLK_BASE__INST4_SEG5                       0
+
+#define CLK_BASE__INST5_SEG0                       0
+#define CLK_BASE__INST5_SEG1                       0
+#define CLK_BASE__INST5_SEG2                       0
+#define CLK_BASE__INST5_SEG3                       0
+#define CLK_BASE__INST5_SEG4                       0
+#define CLK_BASE__INST5_SEG5                       0
+
+#define DCE_BASE__INST0_SEG0                       0x00000012
+#define DCE_BASE__INST0_SEG1                       0x000000C0
+#define DCE_BASE__INST0_SEG2                       0x000034C0
+#define DCE_BASE__INST0_SEG3                       0
+#define DCE_BASE__INST0_SEG4                       0
+#define DCE_BASE__INST0_SEG5                       0
+
+#define DCE_BASE__INST1_SEG0                       0
+#define DCE_BASE__INST1_SEG1                       0
+#define DCE_BASE__INST1_SEG2                       0
+#define DCE_BASE__INST1_SEG3                       0
+#define DCE_BASE__INST1_SEG4                       0
+#define DCE_BASE__INST1_SEG5                       0
+
+#define DCE_BASE__INST2_SEG0                       0
+#define DCE_BASE__INST2_SEG1                       0
+#define DCE_BASE__INST2_SEG2                       0
+#define DCE_BASE__INST2_SEG3                       0
+#define DCE_BASE__INST2_SEG4                       0
+#define DCE_BASE__INST2_SEG5                       0
+
+#define DCE_BASE__INST3_SEG0                       0
+#define DCE_BASE__INST3_SEG1                       0
+#define DCE_BASE__INST3_SEG2                       0
+#define DCE_BASE__INST3_SEG3                       0
+#define DCE_BASE__INST3_SEG4                       0
+#define DCE_BASE__INST3_SEG5                       0
+
+#define DCE_BASE__INST4_SEG0                       0
+#define DCE_BASE__INST4_SEG1                       0
+#define DCE_BASE__INST4_SEG2                       0
+#define DCE_BASE__INST4_SEG3                       0
+#define DCE_BASE__INST4_SEG4                       0
+#define DCE_BASE__INST4_SEG5                       0
+
+#define DCE_BASE__INST5_SEG0                       0
+#define DCE_BASE__INST5_SEG1                       0
+#define DCE_BASE__INST5_SEG2                       0
+#define DCE_BASE__INST5_SEG3                       0
+#define DCE_BASE__INST5_SEG4                       0
+#define DCE_BASE__INST5_SEG5                       0
+
+#define DF_BASE__INST0_SEG0                        0x00007000
+#define DF_BASE__INST0_SEG1                        0
+#define DF_BASE__INST0_SEG2                        0
+#define DF_BASE__INST0_SEG3                        0
+#define DF_BASE__INST0_SEG4                        0
+#define DF_BASE__INST0_SEG5                        0
+
+#define DF_BASE__INST1_SEG0                        0
+#define DF_BASE__INST1_SEG1                        0
+#define DF_BASE__INST1_SEG2                        0
+#define DF_BASE__INST1_SEG3                        0
+#define DF_BASE__INST1_SEG4                        0
+#define DF_BASE__INST1_SEG5                        0
+
+#define DF_BASE__INST2_SEG0                        0
+#define DF_BASE__INST2_SEG1                        0
+#define DF_BASE__INST2_SEG2                        0
+#define DF_BASE__INST2_SEG3                        0
+#define DF_BASE__INST2_SEG4                        0
+#define DF_BASE__INST2_SEG5                        0
+
+#define DF_BASE__INST3_SEG0                        0
+#define DF_BASE__INST3_SEG1                        0
+#define DF_BASE__INST3_SEG2                        0
+#define DF_BASE__INST3_SEG3                        0
+#define DF_BASE__INST3_SEG4                        0
+#define DF_BASE__INST3_SEG5                        0
+
+#define DF_BASE__INST4_SEG0                        0
+#define DF_BASE__INST4_SEG1                        0
+#define DF_BASE__INST4_SEG2                        0
+#define DF_BASE__INST4_SEG3                        0
+#define DF_BASE__INST4_SEG4                        0
+#define DF_BASE__INST4_SEG5                        0
+
+#define DF_BASE__INST5_SEG0                        0
+#define DF_BASE__INST5_SEG1                        0
+#define DF_BASE__INST5_SEG2                        0
+#define DF_BASE__INST5_SEG3                        0
+#define DF_BASE__INST5_SEG4                        0
+#define DF_BASE__INST5_SEG5                        0
+
+#define FUSE_BASE__INST0_SEG0                      0x00017400
+#define FUSE_BASE__INST0_SEG1                      0
+#define FUSE_BASE__INST0_SEG2                      0
+#define FUSE_BASE__INST0_SEG3                      0
+#define FUSE_BASE__INST0_SEG4                      0
+#define FUSE_BASE__INST0_SEG5                      0
+
+#define FUSE_BASE__INST1_SEG0                      0
+#define FUSE_BASE__INST1_SEG1                      0
+#define FUSE_BASE__INST1_SEG2                      0
+#define FUSE_BASE__INST1_SEG3                      0
+#define FUSE_BASE__INST1_SEG4                      0
+#define FUSE_BASE__INST1_SEG5                      0
+
+#define FUSE_BASE__INST2_SEG0                      0
+#define FUSE_BASE__INST2_SEG1                      0
+#define FUSE_BASE__INST2_SEG2                      0
+#define FUSE_BASE__INST2_SEG3                      0
+#define FUSE_BASE__INST2_SEG4                      0
+#define FUSE_BASE__INST2_SEG5                      0
+
+#define FUSE_BASE__INST3_SEG0                      0
+#define FUSE_BASE__INST3_SEG1                      0
+#define FUSE_BASE__INST3_SEG2                      0
+#define FUSE_BASE__INST3_SEG3                      0
+#define FUSE_BASE__INST3_SEG4                      0
+#define FUSE_BASE__INST3_SEG5                      0
+
+#define FUSE_BASE__INST4_SEG0                      0
+#define FUSE_BASE__INST4_SEG1                      0
+#define FUSE_BASE__INST4_SEG2                      0
+#define FUSE_BASE__INST4_SEG3                      0
+#define FUSE_BASE__INST4_SEG4                      0
+#define FUSE_BASE__INST4_SEG5                      0
+
+#define FUSE_BASE__INST5_SEG0                      0
+#define FUSE_BASE__INST5_SEG1                      0
+#define FUSE_BASE__INST5_SEG2                      0
+#define FUSE_BASE__INST5_SEG3                      0
+#define FUSE_BASE__INST5_SEG4                      0
+#define FUSE_BASE__INST5_SEG5                      0
+
+#define GC_BASE__INST0_SEG0                        0x00002000
+#define GC_BASE__INST0_SEG1                        0x0000A000
+#define GC_BASE__INST0_SEG2                        0
+#define GC_BASE__INST0_SEG3                        0
+#define GC_BASE__INST0_SEG4                        0
+#define GC_BASE__INST0_SEG5                        0
+
+#define GC_BASE__INST1_SEG0                        0
+#define GC_BASE__INST1_SEG1                        0
+#define GC_BASE__INST1_SEG2                        0
+#define GC_BASE__INST1_SEG3                        0
+#define GC_BASE__INST1_SEG4                        0
+#define GC_BASE__INST1_SEG5                        0
+
+#define GC_BASE__INST2_SEG0                        0
+#define GC_BASE__INST2_SEG1                        0
+#define GC_BASE__INST2_SEG2                        0
+#define GC_BASE__INST2_SEG3                        0
+#define GC_BASE__INST2_SEG4                        0
+#define GC_BASE__INST2_SEG5                        0
+
+#define GC_BASE__INST3_SEG0                        0
+#define GC_BASE__INST3_SEG1                        0
+#define GC_BASE__INST3_SEG2                        0
+#define GC_BASE__INST3_SEG3                        0
+#define GC_BASE__INST3_SEG4                        0
+#define GC_BASE__INST3_SEG5                        0
+
+#define GC_BASE__INST4_SEG0                        0
+#define GC_BASE__INST4_SEG1                        0
+#define GC_BASE__INST4_SEG2                        0
+#define GC_BASE__INST4_SEG3                        0
+#define GC_BASE__INST4_SEG4                        0
+#define GC_BASE__INST4_SEG5                        0
+
+#define GC_BASE__INST5_SEG0                        0
+#define GC_BASE__INST5_SEG1                        0
+#define GC_BASE__INST5_SEG2                        0
+#define GC_BASE__INST5_SEG3                        0
+#define GC_BASE__INST5_SEG4                        0
+#define GC_BASE__INST5_SEG5                        0
+
+#define HDP_BASE__INST0_SEG0                       0x00000F20
+#define HDP_BASE__INST0_SEG1                       0
+#define HDP_BASE__INST0_SEG2                       0
+#define HDP_BASE__INST0_SEG3                       0
+#define HDP_BASE__INST0_SEG4                       0
+#define HDP_BASE__INST0_SEG5                       0
+
+#define HDP_BASE__INST1_SEG0                       0
+#define HDP_BASE__INST1_SEG1                       0
+#define HDP_BASE__INST1_SEG2                       0
+#define HDP_BASE__INST1_SEG3                       0
+#define HDP_BASE__INST1_SEG4                       0
+#define HDP_BASE__INST1_SEG5                       0
+
+#define HDP_BASE__INST2_SEG0                       0
+#define HDP_BASE__INST2_SEG1                       0
+#define HDP_BASE__INST2_SEG2                       0
+#define HDP_BASE__INST2_SEG3                       0
+#define HDP_BASE__INST2_SEG4                       0
+#define HDP_BASE__INST2_SEG5                       0
+
+#define HDP_BASE__INST3_SEG0                       0
+#define HDP_BASE__INST3_SEG1                       0
+#define HDP_BASE__INST3_SEG2                       0
+#define HDP_BASE__INST3_SEG3                       0
+#define HDP_BASE__INST3_SEG4                       0
+#define HDP_BASE__INST3_SEG5                       0
+
+#define HDP_BASE__INST4_SEG0                       0
+#define HDP_BASE__INST4_SEG1                       0
+#define HDP_BASE__INST4_SEG2                       0
+#define HDP_BASE__INST4_SEG3                       0
+#define HDP_BASE__INST4_SEG4                       0
+#define HDP_BASE__INST4_SEG5                       0
+
+#define HDP_BASE__INST5_SEG0                       0
+#define HDP_BASE__INST5_SEG1                       0
+#define HDP_BASE__INST5_SEG2                       0
+#define HDP_BASE__INST5_SEG3                       0
+#define HDP_BASE__INST5_SEG4                       0
+#define HDP_BASE__INST5_SEG5                       0
+
+#define MMHUB_BASE__INST0_SEG0                     0x0001A000
+#define MMHUB_BASE__INST0_SEG1                     0
+#define MMHUB_BASE__INST0_SEG2                     0
+#define MMHUB_BASE__INST0_SEG3                     0
+#define MMHUB_BASE__INST0_SEG4                     0
+#define MMHUB_BASE__INST0_SEG5                     0
+
+#define MMHUB_BASE__INST1_SEG0                     0
+#define MMHUB_BASE__INST1_SEG1                     0
+#define MMHUB_BASE__INST1_SEG2                     0
+#define MMHUB_BASE__INST1_SEG3                     0
+#define MMHUB_BASE__INST1_SEG4                     0
+#define MMHUB_BASE__INST1_SEG5                     0
+
+#define MMHUB_BASE__INST2_SEG0                     0
+#define MMHUB_BASE__INST2_SEG1                     0
+#define MMHUB_BASE__INST2_SEG2                     0
+#define MMHUB_BASE__INST2_SEG3                     0
+#define MMHUB_BASE__INST2_SEG4                     0
+#define MMHUB_BASE__INST2_SEG5                     0
+
+#define MMHUB_BASE__INST3_SEG0                     0
+#define MMHUB_BASE__INST3_SEG1                     0
+#define MMHUB_BASE__INST3_SEG2                     0
+#define MMHUB_BASE__INST3_SEG3                     0
+#define MMHUB_BASE__INST3_SEG4                     0
+#define MMHUB_BASE__INST3_SEG5                     0
+
+#define MMHUB_BASE__INST4_SEG0                     0
+#define MMHUB_BASE__INST4_SEG1                     0
+#define MMHUB_BASE__INST4_SEG2                     0
+#define MMHUB_BASE__INST4_SEG3                     0
+#define MMHUB_BASE__INST4_SEG4                     0
+#define MMHUB_BASE__INST4_SEG5                     0
+
+#define MMHUB_BASE__INST5_SEG0                     0
+#define MMHUB_BASE__INST5_SEG1                     0
+#define MMHUB_BASE__INST5_SEG2                     0
+#define MMHUB_BASE__INST5_SEG3                     0
+#define MMHUB_BASE__INST5_SEG4                     0
+#define MMHUB_BASE__INST5_SEG5                     0
+
+#define MP0_BASE__INST0_SEG0                       0x00016000
+#define MP0_BASE__INST0_SEG1                       0
+#define MP0_BASE__INST0_SEG2                       0
+#define MP0_BASE__INST0_SEG3                       0
+#define MP0_BASE__INST0_SEG4                       0
+#define MP0_BASE__INST0_SEG5                       0
+
+#define MP0_BASE__INST1_SEG0                       0
+#define MP0_BASE__INST1_SEG1                       0
+#define MP0_BASE__INST1_SEG2                       0
+#define MP0_BASE__INST1_SEG3                       0
+#define MP0_BASE__INST1_SEG4                       0
+#define MP0_BASE__INST1_SEG5                       0
+
+#define MP0_BASE__INST2_SEG0                       0
+#define MP0_BASE__INST2_SEG1                       0
+#define MP0_BASE__INST2_SEG2                       0
+#define MP0_BASE__INST2_SEG3                       0
+#define MP0_BASE__INST2_SEG4                       0
+#define MP0_BASE__INST2_SEG5                       0
+
+#define MP0_BASE__INST3_SEG0                       0
+#define MP0_BASE__INST3_SEG1                       0
+#define MP0_BASE__INST3_SEG2                       0
+#define MP0_BASE__INST3_SEG3                       0
+#define MP0_BASE__INST3_SEG4                       0
+#define MP0_BASE__INST3_SEG5                       0
+
+#define MP0_BASE__INST4_SEG0                       0
+#define MP0_BASE__INST4_SEG1                       0
+#define MP0_BASE__INST4_SEG2                       0
+#define MP0_BASE__INST4_SEG3                       0
+#define MP0_BASE__INST4_SEG4                       0
+#define MP0_BASE__INST4_SEG5                       0
+
+#define MP0_BASE__INST5_SEG0                       0
+#define MP0_BASE__INST5_SEG1                       0
+#define MP0_BASE__INST5_SEG2                       0
+#define MP0_BASE__INST5_SEG3                       0
+#define MP0_BASE__INST5_SEG4                       0
+#define MP0_BASE__INST5_SEG5                       0
+
+#define MP1_BASE__INST0_SEG0                       0x00016000
+#define MP1_BASE__INST0_SEG1                       0
+#define MP1_BASE__INST0_SEG2                       0
+#define MP1_BASE__INST0_SEG3                       0
+#define MP1_BASE__INST0_SEG4                       0
+#define MP1_BASE__INST0_SEG5                       0
+
+#define MP1_BASE__INST1_SEG0                       0
+#define MP1_BASE__INST1_SEG1                       0
+#define MP1_BASE__INST1_SEG2                       0
+#define MP1_BASE__INST1_SEG3                       0
+#define MP1_BASE__INST1_SEG4                       0
+#define MP1_BASE__INST1_SEG5                       0
+
+#define MP1_BASE__INST2_SEG0                       0
+#define MP1_BASE__INST2_SEG1                       0
+#define MP1_BASE__INST2_SEG2                       0
+#define MP1_BASE__INST2_SEG3                       0
+#define MP1_BASE__INST2_SEG4                       0
+#define MP1_BASE__INST2_SEG5                       0
+
+#define MP1_BASE__INST3_SEG0                       0
+#define MP1_BASE__INST3_SEG1                       0
+#define MP1_BASE__INST3_SEG2                       0
+#define MP1_BASE__INST3_SEG3                       0
+#define MP1_BASE__INST3_SEG4                       0
+#define MP1_BASE__INST3_SEG5                       0
+
+#define MP1_BASE__INST4_SEG0                       0
+#define MP1_BASE__INST4_SEG1                       0
+#define MP1_BASE__INST4_SEG2                       0
+#define MP1_BASE__INST4_SEG3                       0
+#define MP1_BASE__INST4_SEG4                       0
+#define MP1_BASE__INST4_SEG5                       0
+
+#define MP1_BASE__INST5_SEG0                       0
+#define MP1_BASE__INST5_SEG1                       0
+#define MP1_BASE__INST5_SEG2                       0
+#define MP1_BASE__INST5_SEG3                       0
+#define MP1_BASE__INST5_SEG4                       0
+#define MP1_BASE__INST5_SEG5                       0
+
+#define NBIO_BASE__INST0_SEG0                      0x00000000
+#define NBIO_BASE__INST0_SEG1                      0x00000014
+#define NBIO_BASE__INST0_SEG2                      0x00000D20
+#define NBIO_BASE__INST0_SEG3                      0x00010400
+#define NBIO_BASE__INST0_SEG4                      0
+#define NBIO_BASE__INST0_SEG5                      0
+
+#define NBIO_BASE__INST1_SEG0                      0
+#define NBIO_BASE__INST1_SEG1                      0
+#define NBIO_BASE__INST1_SEG2                      0
+#define NBIO_BASE__INST1_SEG3                      0
+#define NBIO_BASE__INST1_SEG4                      0
+#define NBIO_BASE__INST1_SEG5                      0
+
+#define NBIO_BASE__INST2_SEG0                      0
+#define NBIO_BASE__INST2_SEG1                      0
+#define NBIO_BASE__INST2_SEG2                      0
+#define NBIO_BASE__INST2_SEG3                      0
+#define NBIO_BASE__INST2_SEG4                      0
+#define NBIO_BASE__INST2_SEG5                      0
+
+#define NBIO_BASE__INST3_SEG0                      0
+#define NBIO_BASE__INST3_SEG1                      0
+#define NBIO_BASE__INST3_SEG2                      0
+#define NBIO_BASE__INST3_SEG3                      0
+#define NBIO_BASE__INST3_SEG4                      0
+#define NBIO_BASE__INST3_SEG5                      0
+
+#define NBIO_BASE__INST4_SEG0                      0
+#define NBIO_BASE__INST4_SEG1                      0
+#define NBIO_BASE__INST4_SEG2                      0
+#define NBIO_BASE__INST4_SEG3                      0
+#define NBIO_BASE__INST4_SEG4                      0
+#define NBIO_BASE__INST4_SEG5                      0
+
+#define NBIO_BASE__INST5_SEG0                      0
+#define NBIO_BASE__INST5_SEG1                      0
+#define NBIO_BASE__INST5_SEG2                      0
+#define NBIO_BASE__INST5_SEG3                      0
+#define NBIO_BASE__INST5_SEG4                      0
+#define NBIO_BASE__INST5_SEG5                      0
+
+#define OSSSYS_BASE__INST0_SEG0                    0x000010A0
+#define OSSSYS_BASE__INST0_SEG1                    0
+#define OSSSYS_BASE__INST0_SEG2                    0
+#define OSSSYS_BASE__INST0_SEG3                    0
+#define OSSSYS_BASE__INST0_SEG4                    0
+#define OSSSYS_BASE__INST0_SEG5                    0
+
+#define OSSSYS_BASE__INST1_SEG0                    0
+#define OSSSYS_BASE__INST1_SEG1                    0
+#define OSSSYS_BASE__INST1_SEG2                    0
+#define OSSSYS_BASE__INST1_SEG3                    0
+#define OSSSYS_BASE__INST1_SEG4                    0
+#define OSSSYS_BASE__INST1_SEG5                    0
+
+#define OSSSYS_BASE__INST2_SEG0                    0
+#define OSSSYS_BASE__INST2_SEG1                    0
+#define OSSSYS_BASE__INST2_SEG2                    0
+#define OSSSYS_BASE__INST2_SEG3                    0
+#define OSSSYS_BASE__INST2_SEG4                    0
+#define OSSSYS_BASE__INST2_SEG5                    0
+
+#define OSSSYS_BASE__INST3_SEG0                    0
+#define OSSSYS_BASE__INST3_SEG1                    0
+#define OSSSYS_BASE__INST3_SEG2                    0
+#define OSSSYS_BASE__INST3_SEG3                    0
+#define OSSSYS_BASE__INST3_SEG4                    0
+#define OSSSYS_BASE__INST3_SEG5                    0
+
+#define OSSSYS_BASE__INST4_SEG0                    0
+#define OSSSYS_BASE__INST4_SEG1                    0
+#define OSSSYS_BASE__INST4_SEG2                    0
+#define OSSSYS_BASE__INST4_SEG3                    0
+#define OSSSYS_BASE__INST4_SEG4                    0
+#define OSSSYS_BASE__INST4_SEG5                    0
+
+#define OSSSYS_BASE__INST5_SEG0                    0
+#define OSSSYS_BASE__INST5_SEG1                    0
+#define OSSSYS_BASE__INST5_SEG2                    0
+#define OSSSYS_BASE__INST5_SEG3                    0
+#define OSSSYS_BASE__INST5_SEG4                    0
+#define OSSSYS_BASE__INST5_SEG5                    0
+
+#define SDMA0_BASE__INST0_SEG0                     0x00001260
+#define SDMA0_BASE__INST0_SEG1                     0
+#define SDMA0_BASE__INST0_SEG2                     0
+#define SDMA0_BASE__INST0_SEG3                     0
+#define SDMA0_BASE__INST0_SEG4                     0
+#define SDMA0_BASE__INST0_SEG5                     0
+
+#define SDMA0_BASE__INST1_SEG0                     0
+#define SDMA0_BASE__INST1_SEG1                     0
+#define SDMA0_BASE__INST1_SEG2                     0
+#define SDMA0_BASE__INST1_SEG3                     0
+#define SDMA0_BASE__INST1_SEG4                     0
+#define SDMA0_BASE__INST1_SEG5                     0
+
+#define SDMA0_BASE__INST2_SEG0                     0
+#define SDMA0_BASE__INST2_SEG1                     0
+#define SDMA0_BASE__INST2_SEG2                     0
+#define SDMA0_BASE__INST2_SEG3                     0
+#define SDMA0_BASE__INST2_SEG4                     0
+#define SDMA0_BASE__INST2_SEG5                     0
+
+#define SDMA0_BASE__INST3_SEG0                     0
+#define SDMA0_BASE__INST3_SEG1                     0
+#define SDMA0_BASE__INST3_SEG2                     0
+#define SDMA0_BASE__INST3_SEG3                     0
+#define SDMA0_BASE__INST3_SEG4                     0
+#define SDMA0_BASE__INST3_SEG5                     0
+
+#define SDMA0_BASE__INST4_SEG0                     0
+#define SDMA0_BASE__INST4_SEG1                     0
+#define SDMA0_BASE__INST4_SEG2                     0
+#define SDMA0_BASE__INST4_SEG3                     0
+#define SDMA0_BASE__INST4_SEG4                     0
+#define SDMA0_BASE__INST4_SEG5                     0
+
+#define SDMA0_BASE__INST5_SEG0                     0
+#define SDMA0_BASE__INST5_SEG1                     0
+#define SDMA0_BASE__INST5_SEG2                     0
+#define SDMA0_BASE__INST5_SEG3                     0
+#define SDMA0_BASE__INST5_SEG4                     0
+#define SDMA0_BASE__INST5_SEG5                     0
+
+#define SDMA1_BASE__INST0_SEG0                     0x00001860
+#define SDMA1_BASE__INST0_SEG1                     0
+#define SDMA1_BASE__INST0_SEG2                     0
+#define SDMA1_BASE__INST0_SEG3                     0
+#define SDMA1_BASE__INST0_SEG4                     0
+#define SDMA1_BASE__INST0_SEG5                     0
+
+#define SDMA1_BASE__INST1_SEG0                     0
+#define SDMA1_BASE__INST1_SEG1                     0
+#define SDMA1_BASE__INST1_SEG2                     0
+#define SDMA1_BASE__INST1_SEG3                     0
+#define SDMA1_BASE__INST1_SEG4                     0
+#define SDMA1_BASE__INST1_SEG5                     0
+
+#define SDMA1_BASE__INST2_SEG0                     0
+#define SDMA1_BASE__INST2_SEG1                     0
+#define SDMA1_BASE__INST2_SEG2                     0
+#define SDMA1_BASE__INST2_SEG3                     0
+#define SDMA1_BASE__INST2_SEG4                     0
+#define SDMA1_BASE__INST2_SEG5                     0
+
+#define SDMA1_BASE__INST3_SEG0                     0
+#define SDMA1_BASE__INST3_SEG1                     0
+#define SDMA1_BASE__INST3_SEG2                     0
+#define SDMA1_BASE__INST3_SEG3                     0
+#define SDMA1_BASE__INST3_SEG4                     0
+#define SDMA1_BASE__INST3_SEG5                     0
+
+#define SDMA1_BASE__INST4_SEG0                     0
+#define SDMA1_BASE__INST4_SEG1                     0
+#define SDMA1_BASE__INST4_SEG2                     0
+#define SDMA1_BASE__INST4_SEG3                     0
+#define SDMA1_BASE__INST4_SEG4                     0
+#define SDMA1_BASE__INST4_SEG5                     0
+
+#define SDMA1_BASE__INST5_SEG0                     0
+#define SDMA1_BASE__INST5_SEG1                     0
+#define SDMA1_BASE__INST5_SEG2                     0
+#define SDMA1_BASE__INST5_SEG3                     0
+#define SDMA1_BASE__INST5_SEG4                     0
+#define SDMA1_BASE__INST5_SEG5                     0
+
+#define SMUIO_BASE__INST0_SEG0                     0x00016800
+#define SMUIO_BASE__INST0_SEG1                     0x00016A00
+#define SMUIO_BASE__INST0_SEG2                     0
+#define SMUIO_BASE__INST0_SEG3                     0
+#define SMUIO_BASE__INST0_SEG4                     0
+#define SMUIO_BASE__INST0_SEG5                     0
+
+#define SMUIO_BASE__INST1_SEG0                     0
+#define SMUIO_BASE__INST1_SEG1                     0
+#define SMUIO_BASE__INST1_SEG2                     0
+#define SMUIO_BASE__INST1_SEG3                     0
+#define SMUIO_BASE__INST1_SEG4                     0
+#define SMUIO_BASE__INST1_SEG5                     0
+
+#define SMUIO_BASE__INST2_SEG0                     0
+#define SMUIO_BASE__INST2_SEG1                     0
+#define SMUIO_BASE__INST2_SEG2                     0
+#define SMUIO_BASE__INST2_SEG3                     0
+#define SMUIO_BASE__INST2_SEG4                     0
+#define SMUIO_BASE__INST2_SEG5                     0
+
+#define SMUIO_BASE__INST3_SEG0                     0
+#define SMUIO_BASE__INST3_SEG1                     0
+#define SMUIO_BASE__INST3_SEG2                     0
+#define SMUIO_BASE__INST3_SEG3                     0
+#define SMUIO_BASE__INST3_SEG4                     0
+#define SMUIO_BASE__INST3_SEG5                     0
+
+#define SMUIO_BASE__INST4_SEG0                     0
+#define SMUIO_BASE__INST4_SEG1                     0
+#define SMUIO_BASE__INST4_SEG2                     0
+#define SMUIO_BASE__INST4_SEG3                     0
+#define SMUIO_BASE__INST4_SEG4                     0
+#define SMUIO_BASE__INST4_SEG5                     0
+
+#define SMUIO_BASE__INST5_SEG0                     0
+#define SMUIO_BASE__INST5_SEG1                     0
+#define SMUIO_BASE__INST5_SEG2                     0
+#define SMUIO_BASE__INST5_SEG3                     0
+#define SMUIO_BASE__INST5_SEG4                     0
+#define SMUIO_BASE__INST5_SEG5                     0
+
+#define THM_BASE__INST0_SEG0                       0x00016600
+#define THM_BASE__INST0_SEG1                       0
+#define THM_BASE__INST0_SEG2                       0
+#define THM_BASE__INST0_SEG3                       0
+#define THM_BASE__INST0_SEG4                       0
+#define THM_BASE__INST0_SEG5                       0
+
+#define THM_BASE__INST1_SEG0                       0
+#define THM_BASE__INST1_SEG1                       0
+#define THM_BASE__INST1_SEG2                       0
+#define THM_BASE__INST1_SEG3                       0
+#define THM_BASE__INST1_SEG4                       0
+#define THM_BASE__INST1_SEG5                       0
+
+#define THM_BASE__INST2_SEG0                       0
+#define THM_BASE__INST2_SEG1                       0
+#define THM_BASE__INST2_SEG2                       0
+#define THM_BASE__INST2_SEG3                       0
+#define THM_BASE__INST2_SEG4                       0
+#define THM_BASE__INST2_SEG5                       0
+
+#define THM_BASE__INST3_SEG0                       0
+#define THM_BASE__INST3_SEG1                       0
+#define THM_BASE__INST3_SEG2                       0
+#define THM_BASE__INST3_SEG3                       0
+#define THM_BASE__INST3_SEG4                       0
+#define THM_BASE__INST3_SEG5                       0
+
+#define THM_BASE__INST4_SEG0                       0
+#define THM_BASE__INST4_SEG1                       0
+#define THM_BASE__INST4_SEG2                       0
+#define THM_BASE__INST4_SEG3                       0
+#define THM_BASE__INST4_SEG4                       0
+#define THM_BASE__INST4_SEG5                       0
+
+#define THM_BASE__INST5_SEG0                       0
+#define THM_BASE__INST5_SEG1                       0
+#define THM_BASE__INST5_SEG2                       0
+#define THM_BASE__INST5_SEG3                       0
+#define THM_BASE__INST5_SEG4                       0
+#define THM_BASE__INST5_SEG5                       0
+
+#define UMC_BASE__INST0_SEG0                       0x00014000
+#define UMC_BASE__INST0_SEG1                       0
+#define UMC_BASE__INST0_SEG2                       0
+#define UMC_BASE__INST0_SEG3                       0
+#define UMC_BASE__INST0_SEG4                       0
+#define UMC_BASE__INST0_SEG5                       0
+
+#define UMC_BASE__INST1_SEG0                       0
+#define UMC_BASE__INST1_SEG1                       0
+#define UMC_BASE__INST1_SEG2                       0
+#define UMC_BASE__INST1_SEG3                       0
+#define UMC_BASE__INST1_SEG4                       0
+#define UMC_BASE__INST1_SEG5                       0
+
+#define UMC_BASE__INST2_SEG0                       0
+#define UMC_BASE__INST2_SEG1                       0
+#define UMC_BASE__INST2_SEG2                       0
+#define UMC_BASE__INST2_SEG3                       0
+#define UMC_BASE__INST2_SEG4                       0
+#define UMC_BASE__INST2_SEG5                       0
+
+#define UMC_BASE__INST3_SEG0                       0
+#define UMC_BASE__INST3_SEG1                       0
+#define UMC_BASE__INST3_SEG2                       0
+#define UMC_BASE__INST3_SEG3                       0
+#define UMC_BASE__INST3_SEG4                       0
+#define UMC_BASE__INST3_SEG5                       0
+
+#define UMC_BASE__INST4_SEG0                       0
+#define UMC_BASE__INST4_SEG1                       0
+#define UMC_BASE__INST4_SEG2                       0
+#define UMC_BASE__INST4_SEG3                       0
+#define UMC_BASE__INST4_SEG4                       0
+#define UMC_BASE__INST4_SEG5                       0
+
+#define UMC_BASE__INST5_SEG0                       0
+#define UMC_BASE__INST5_SEG1                       0
+#define UMC_BASE__INST5_SEG2                       0
+#define UMC_BASE__INST5_SEG3                       0
+#define UMC_BASE__INST5_SEG4                       0
+#define UMC_BASE__INST5_SEG5                       0
+
+#define UVD_BASE__INST0_SEG0                       0x00007800
+#define UVD_BASE__INST0_SEG1                       0x00007E00
+#define UVD_BASE__INST0_SEG2                       0
+#define UVD_BASE__INST0_SEG3                       0
+#define UVD_BASE__INST0_SEG4                       0
+#define UVD_BASE__INST0_SEG5                       0
+
+#define UVD_BASE__INST1_SEG0                       0
+#define UVD_BASE__INST1_SEG1                       0x00009000
+#define UVD_BASE__INST1_SEG2                       0
+#define UVD_BASE__INST1_SEG3                       0
+#define UVD_BASE__INST1_SEG4                       0
+#define UVD_BASE__INST1_SEG5                       0
+
+#define UVD_BASE__INST2_SEG0                       0
+#define UVD_BASE__INST2_SEG1                       0
+#define UVD_BASE__INST2_SEG2                       0
+#define UVD_BASE__INST2_SEG3                       0
+#define UVD_BASE__INST2_SEG4                       0
+#define UVD_BASE__INST2_SEG5                       0
+
+#define UVD_BASE__INST3_SEG0                       0
+#define UVD_BASE__INST3_SEG1                       0
+#define UVD_BASE__INST3_SEG2                       0
+#define UVD_BASE__INST3_SEG3                       0
+#define UVD_BASE__INST3_SEG4                       0
+#define UVD_BASE__INST3_SEG5                       0
+
+#define UVD_BASE__INST4_SEG0                       0
+#define UVD_BASE__INST4_SEG1                       0
+#define UVD_BASE__INST4_SEG2                       0
+#define UVD_BASE__INST4_SEG3                       0
+#define UVD_BASE__INST4_SEG4                       0
+#define UVD_BASE__INST4_SEG5                       0
+
+#define UVD_BASE__INST5_SEG0                       0
+#define UVD_BASE__INST5_SEG1                       0
+#define UVD_BASE__INST5_SEG2                       0
+#define UVD_BASE__INST5_SEG3                       0
+#define UVD_BASE__INST5_SEG4                       0
+#define UVD_BASE__INST5_SEG5                       0
+
+#define VCE_BASE__INST0_SEG0                       0x00008800
+#define VCE_BASE__INST0_SEG1                       0
+#define VCE_BASE__INST0_SEG2                       0
+#define VCE_BASE__INST0_SEG3                       0
+#define VCE_BASE__INST0_SEG4                       0
+#define VCE_BASE__INST0_SEG5                       0
+
+#define VCE_BASE__INST1_SEG0                       0
+#define VCE_BASE__INST1_SEG1                       0
+#define VCE_BASE__INST1_SEG2                       0
+#define VCE_BASE__INST1_SEG3                       0
+#define VCE_BASE__INST1_SEG4                       0
+#define VCE_BASE__INST1_SEG5                       0
+
+#define VCE_BASE__INST2_SEG0                       0
+#define VCE_BASE__INST2_SEG1                       0
+#define VCE_BASE__INST2_SEG2                       0
+#define VCE_BASE__INST2_SEG3                       0
+#define VCE_BASE__INST2_SEG4                       0
+#define VCE_BASE__INST2_SEG5                       0
+
+#define VCE_BASE__INST3_SEG0                       0
+#define VCE_BASE__INST3_SEG1                       0
+#define VCE_BASE__INST3_SEG2                       0
+#define VCE_BASE__INST3_SEG3                       0
+#define VCE_BASE__INST3_SEG4                       0
+#define VCE_BASE__INST3_SEG5                       0
+
+#define VCE_BASE__INST4_SEG0                       0
+#define VCE_BASE__INST4_SEG1                       0
+#define VCE_BASE__INST4_SEG2                       0
+#define VCE_BASE__INST4_SEG3                       0
+#define VCE_BASE__INST4_SEG4                       0
+#define VCE_BASE__INST4_SEG5                       0
+
+#define VCE_BASE__INST5_SEG0                       0
+#define VCE_BASE__INST5_SEG1                       0
+#define VCE_BASE__INST5_SEG2                       0
+#define VCE_BASE__INST5_SEG3                       0
+#define VCE_BASE__INST5_SEG4                       0
+#define VCE_BASE__INST5_SEG5                       0
+
+#define XDMA_BASE__INST0_SEG0                      0x00003400
+#define XDMA_BASE__INST0_SEG1                      0
+#define XDMA_BASE__INST0_SEG2                      0
+#define XDMA_BASE__INST0_SEG3                      0
+#define XDMA_BASE__INST0_SEG4                      0
+#define XDMA_BASE__INST0_SEG5                      0
+
+#define XDMA_BASE__INST1_SEG0                      0
+#define XDMA_BASE__INST1_SEG1                      0
+#define XDMA_BASE__INST1_SEG2                      0
+#define XDMA_BASE__INST1_SEG3                      0
+#define XDMA_BASE__INST1_SEG4                      0
+#define XDMA_BASE__INST1_SEG5                      0
+
+#define XDMA_BASE__INST2_SEG0                      0
+#define XDMA_BASE__INST2_SEG1                      0
+#define XDMA_BASE__INST2_SEG2                      0
+#define XDMA_BASE__INST2_SEG3                      0
+#define XDMA_BASE__INST2_SEG4                      0
+#define XDMA_BASE__INST2_SEG5                      0
+
+#define XDMA_BASE__INST3_SEG0                      0
+#define XDMA_BASE__INST3_SEG1                      0
+#define XDMA_BASE__INST3_SEG2                      0
+#define XDMA_BASE__INST3_SEG3                      0
+#define XDMA_BASE__INST3_SEG4                      0
+#define XDMA_BASE__INST3_SEG5                      0
+
+#define XDMA_BASE__INST4_SEG0                      0
+#define XDMA_BASE__INST4_SEG1                      0
+#define XDMA_BASE__INST4_SEG2                      0
+#define XDMA_BASE__INST4_SEG3                      0
+#define XDMA_BASE__INST4_SEG4                      0
+#define XDMA_BASE__INST4_SEG5                      0
+
+#define XDMA_BASE__INST5_SEG0                      0
+#define XDMA_BASE__INST5_SEG1                      0
+#define XDMA_BASE__INST5_SEG2                      0
+#define XDMA_BASE__INST5_SEG3                      0
+#define XDMA_BASE__INST5_SEG4                      0
+#define XDMA_BASE__INST5_SEG5                      0
+
+#define RSMU_BASE__INST0_SEG0                      0x00012000
+#define RSMU_BASE__INST0_SEG1                      0
+#define RSMU_BASE__INST0_SEG2                      0
+#define RSMU_BASE__INST0_SEG3                      0
+#define RSMU_BASE__INST0_SEG4                      0
+#define RSMU_BASE__INST0_SEG5                      0
+
+#define RSMU_BASE__INST1_SEG0                      0
+#define RSMU_BASE__INST1_SEG1                      0
+#define RSMU_BASE__INST1_SEG2                      0
+#define RSMU_BASE__INST1_SEG3                      0
+#define RSMU_BASE__INST1_SEG4                      0
+#define RSMU_BASE__INST1_SEG5                      0
+
+#define RSMU_BASE__INST2_SEG0                      0
+#define RSMU_BASE__INST2_SEG1                      0
+#define RSMU_BASE__INST2_SEG2                      0
+#define RSMU_BASE__INST2_SEG3                      0
+#define RSMU_BASE__INST2_SEG4                      0
+#define RSMU_BASE__INST2_SEG5                      0
+
+#define RSMU_BASE__INST3_SEG0                      0
+#define RSMU_BASE__INST3_SEG1                      0
+#define RSMU_BASE__INST3_SEG2                      0
+#define RSMU_BASE__INST3_SEG3                      0
+#define RSMU_BASE__INST3_SEG4                      0
+#define RSMU_BASE__INST3_SEG5                      0
+
+#define RSMU_BASE__INST4_SEG0                      0
+#define RSMU_BASE__INST4_SEG1                      0
+#define RSMU_BASE__INST4_SEG2                      0
+#define RSMU_BASE__INST4_SEG3                      0
+#define RSMU_BASE__INST4_SEG4                      0
+#define RSMU_BASE__INST4_SEG5                      0
+
+#define RSMU_BASE__INST5_SEG0                      0
+#define RSMU_BASE__INST5_SEG1                      0
+#define RSMU_BASE__INST5_SEG2                      0
+#define RSMU_BASE__INST5_SEG3                      0
+#define RSMU_BASE__INST5_SEG4                      0
+#define RSMU_BASE__INST5_SEG5                      0
+
+#endif
+
-- 
cgit v1.1


From b2f87c9182deaf495ec4fefde89584910ec137d8 Mon Sep 17 00:00:00 2001
From: James Zhu <James.Zhu@amd.com>
Date: Tue, 17 Apr 2018 16:25:58 -0400
Subject: drm/amd/include/vg20: adjust VCE_BASE to reuse vce 4.0 header files

Vega20 uses vce 4.1 engine, all the registers have the
same absolute offset with vce 4.0. By adjusting vega20
VCE_BASE, vce 4.1 can reuse vce 4.0 header files.

Signed-off-by: James Zhu <James.Zhu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/include/vega20_ip_offset.h | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/include/vega20_ip_offset.h b/drivers/gpu/drm/amd/include/vega20_ip_offset.h
index 97db93c..2a2a9cc 100644
--- a/drivers/gpu/drm/amd/include/vega20_ip_offset.h
+++ b/drivers/gpu/drm/amd/include/vega20_ip_offset.h
@@ -144,7 +144,8 @@ static const struct IP_BASE UVD_BASE            ={ { { { 0x00007800, 0x00007E00,
                                         { { 0, 0, 0, 0, 0, 0 } },
                                         { { 0, 0, 0, 0, 0, 0 } },
                                         { { 0, 0, 0, 0, 0, 0 } } } };
-static const struct IP_BASE VCE_BASE            ={ { { { 0x00008800, 0, 0, 0, 0, 0 } },
+/* Adjust VCE_BASE to make vce_4_1 use vce_4_0 offset header files*/
+static const struct IP_BASE VCE_BASE            ={ { { { 0x00007E00/* 0x00008800 */, 0, 0, 0, 0, 0 } },
                                         { { 0, 0, 0, 0, 0, 0 } },
                                         { { 0, 0, 0, 0, 0, 0 } },
                                         { { 0, 0, 0, 0, 0, 0 } },
-- 
cgit v1.1


From 956fcddc0b2a7430b6ee4783827f57cb7c823c7d Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Fri, 20 Apr 2018 12:27:54 +0800
Subject: drm/amdgpu: Add vega20 to asic_type enum.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Add vega20 to amd_asic_type enum and amdgpu_asic_name[].

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 1 +
 include/drm/amd_asic_type.h                | 1 +
 2 files changed, 2 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 9fb20a5..f84fc56 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -86,6 +86,7 @@ static const char *amdgpu_asic_name[] = {
 	"VEGAM",
 	"VEGA10",
 	"VEGA12",
+	"VEGA20",
 	"RAVEN",
 	"LAST",
 };
diff --git a/include/drm/amd_asic_type.h b/include/drm/amd_asic_type.h
index 695bde7..dd63d08 100644
--- a/include/drm/amd_asic_type.h
+++ b/include/drm/amd_asic_type.h
@@ -47,6 +47,7 @@ enum amd_asic_type {
 	CHIP_VEGAM,
 	CHIP_VEGA10,
 	CHIP_VEGA12,
+	CHIP_VEGA20,
 	CHIP_RAVEN,
 	CHIP_LAST,
 };
-- 
cgit v1.1


From 27c0bc7163ae8484d3a15324122774b240fadd21 Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Thu, 17 May 2018 10:01:19 -0500
Subject: drm/amdgpu: Add gpu_info firmware for vega20. (v2)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

vega20_gpu_info firmware stores gpu configuration for vega20.

v2: drop gpu info firmware for vega20

Squash of:
drm/amdgpu: Add gpu_info firmware for vega20.
drm/amdgpu: drop gpu_info firmware for vega20

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index f84fc56..3a8d4bc 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -1388,6 +1388,7 @@ static int amdgpu_device_parse_gpu_info_fw(struct amdgpu_device *adev)
 	case CHIP_KABINI:
 	case CHIP_MULLINS:
 #endif
+	case CHIP_VEGA20:
 	default:
 		return 0;
 	case CHIP_VEGA10:
-- 
cgit v1.1


From e4bd8170407dc54bc3f4b0e140816e51f13f3e71 Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Fri, 20 Apr 2018 12:33:33 +0800
Subject: drm/amdgpu: set asic family for vega20.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 3a8d4bc..2d46ad7 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -1523,6 +1523,7 @@ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev)
 #endif
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 	case CHIP_RAVEN:
 		if (adev->asic_type == CHIP_RAVEN)
 			adev->family = AMDGPU_FAMILY_RV;
-- 
cgit v1.1


From a167ae2509132e97c94d66cf1ce15ba2fa620248 Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Fri, 20 Apr 2018 12:46:21 +0800
Subject: drm/amdgpu: Add smu firmware support for vega20
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c       | 3 +++
 drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c | 1 +
 2 files changed, 4 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
index 5b3d3bf..e950730 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
@@ -400,6 +400,9 @@ static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device,
 			case CHIP_VEGA12:
 				strcpy(fw_name, "amdgpu/vega12_smc.bin");
 				break;
+			case CHIP_VEGA20:
+				strcpy(fw_name, "amdgpu/vega20_smc.bin");
+				break;
 			default:
 				DRM_ERROR("SMC firmware not supported\n");
 				return -EINVAL;
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c
index ee236df..c983793 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c
@@ -45,6 +45,7 @@ MODULE_FIRMWARE("amdgpu/vegam_smc.bin");
 MODULE_FIRMWARE("amdgpu/vega10_smc.bin");
 MODULE_FIRMWARE("amdgpu/vega10_acg_smc.bin");
 MODULE_FIRMWARE("amdgpu/vega12_smc.bin");
+MODULE_FIRMWARE("amdgpu/vega20_smc.bin");
 
 int smum_thermal_avfs_enable(struct pp_hwmgr *hwmgr)
 {
-- 
cgit v1.1


From d3bfb6647cc66664f1e09706690444d2d09a56a8 Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Fri, 20 Apr 2018 13:32:46 +0800
Subject: drm/amdgpu/powerplay: Add initial vega20 support v2
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Initial powerplay support the same as vega10 for now.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
index 71b4233..e63bc47 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
@@ -151,6 +151,7 @@ int hwmgr_early_init(struct pp_hwmgr *hwmgr)
 		hwmgr->feature_mask &= ~PP_GFXOFF_MASK;
 		switch (hwmgr->chip_id) {
 		case CHIP_VEGA10:
+		case CHIP_VEGA20:
 			hwmgr->smumgr_funcs = &vega10_smu_funcs;
 			vega10_hwmgr_init(hwmgr);
 			break;
-- 
cgit v1.1


From 8fd2d849da98924e1e021314de289d4a3a31d07f Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Fri, 20 Apr 2018 13:36:54 +0800
Subject: drm/amdgpu/psp: Add initial psp support for vega20
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The same as vega10 for now.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c | 1 +
 drivers/gpu/drm/amd/amdgpu/psp_v3_1.c   | 3 +++
 2 files changed, 4 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
index c7d43e0..9f1a5bd 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
@@ -52,6 +52,7 @@ static int psp_sw_init(void *handle)
 	switch (adev->asic_type) {
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 		psp_v3_1_set_psp_funcs(psp);
 		break;
 	case CHIP_RAVEN:
diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c
index 196e75d..0c768e3 100644
--- a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c
+++ b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c
@@ -41,6 +41,9 @@ MODULE_FIRMWARE("amdgpu/vega10_sos.bin");
 MODULE_FIRMWARE("amdgpu/vega10_asd.bin");
 MODULE_FIRMWARE("amdgpu/vega12_sos.bin");
 MODULE_FIRMWARE("amdgpu/vega12_asd.bin");
+MODULE_FIRMWARE("amdgpu/vega20_sos.bin");
+MODULE_FIRMWARE("amdgpu/vega20_asd.bin");
+
 
 #define smnMP1_FIRMWARE_FLAGS 0x3010028
 
-- 
cgit v1.1


From 4b1f540ae1a9eba826538cb37f6791729e2bcec8 Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Fri, 20 Apr 2018 13:38:24 +0800
Subject: drm/amdgpu: Add vega20 ucode loading method
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The same as vega10.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
index 75592bd..b419d6e 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
@@ -303,6 +303,7 @@ amdgpu_ucode_get_load_type(struct amdgpu_device *adev, int load_type)
 	case CHIP_VEGA10:
 	case CHIP_RAVEN:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 		if (!load_type)
 			return AMDGPU_FW_LOAD_DIRECT;
 		else
-- 
cgit v1.1


From cac18c82e0c5b39b69648942576dbd1d6f9d056e Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Fri, 11 May 2018 13:44:09 -0500
Subject: drm/amdgpu: Specify vega20 uvd firmware
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
index de4d77a..fd1e9cd 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
@@ -70,6 +70,7 @@
 
 #define FIRMWARE_VEGA10		"amdgpu/vega10_uvd.bin"
 #define FIRMWARE_VEGA12		"amdgpu/vega12_uvd.bin"
+#define FIRMWARE_VEGA20		"amdgpu/vega20_uvd.bin"
 
 #define mmUVD_GPCOM_VCPU_DATA0_VEGA10 (0x03c4 + 0x7e00)
 #define mmUVD_GPCOM_VCPU_DATA1_VEGA10 (0x03c5 + 0x7e00)
@@ -114,6 +115,7 @@ MODULE_FIRMWARE(FIRMWARE_VEGAM);
 
 MODULE_FIRMWARE(FIRMWARE_VEGA10);
 MODULE_FIRMWARE(FIRMWARE_VEGA12);
+MODULE_FIRMWARE(FIRMWARE_VEGA20);
 
 static void amdgpu_uvd_idle_work_handler(struct work_struct *work);
 
@@ -177,6 +179,9 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
 	case CHIP_VEGAM:
 		fw_name = FIRMWARE_VEGAM;
 		break;
+	case CHIP_VEGA20:
+		fw_name = FIRMWARE_VEGA20;
+		break;
 	default:
 		return -EINVAL;
 	}
-- 
cgit v1.1


From 341b4ce2330b0af3fa09db545dc2d552a99dbdec Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Fri, 20 Apr 2018 13:46:49 +0800
Subject: drm/amdgpu: Specify vega20 vce firmware
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
index a86322f..23d960e 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
@@ -57,6 +57,7 @@
 
 #define FIRMWARE_VEGA10		"amdgpu/vega10_vce.bin"
 #define FIRMWARE_VEGA12		"amdgpu/vega12_vce.bin"
+#define FIRMWARE_VEGA20		"amdgpu/vega20_vce.bin"
 
 #ifdef CONFIG_DRM_AMDGPU_CIK
 MODULE_FIRMWARE(FIRMWARE_BONAIRE);
@@ -76,6 +77,7 @@ MODULE_FIRMWARE(FIRMWARE_VEGAM);
 
 MODULE_FIRMWARE(FIRMWARE_VEGA10);
 MODULE_FIRMWARE(FIRMWARE_VEGA12);
+MODULE_FIRMWARE(FIRMWARE_VEGA20);
 
 static void amdgpu_vce_idle_work_handler(struct work_struct *work);
 
@@ -143,6 +145,9 @@ int amdgpu_vce_sw_init(struct amdgpu_device *adev, unsigned long size)
 	case CHIP_VEGA12:
 		fw_name = FIRMWARE_VEGA12;
 		break;
+	case CHIP_VEGA20:
+		fw_name = FIRMWARE_VEGA20;
+		break;
 
 	default:
 		return -EINVAL;
-- 
cgit v1.1


From a2c319b63ea377bce4f278d4ca1cb4d6da31e4fb Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Fri, 20 Apr 2018 13:48:23 +0800
Subject: drm/amdgpu/virtual_dce: Add vega20 support
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/dce_virtual.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
index de7be3d..dbf2ccd 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
@@ -473,6 +473,7 @@ static int dce_virtual_hw_init(void *handle)
 		break;
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 		break;
 	default:
 		DRM_ERROR("Virtual display unsupported ASIC type: 0x%X\n", adev->asic_type);
-- 
cgit v1.1


From d96b428c3cea9ed12d03635a02fbf8644e315bc0 Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Fri, 20 Apr 2018 13:56:43 +0800
Subject: drm/amdgpu/gmc9: Add vega20 support
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
index 7343069..b60ed28 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
@@ -752,6 +752,7 @@ static int gmc_v9_0_mc_init(struct amdgpu_device *adev)
 		switch (adev->asic_type) {
 		case CHIP_VEGA10:  /* all engines support GPUVM */
 		case CHIP_VEGA12:  /* all engines support GPUVM */
+		case CHIP_VEGA20:
 		default:
 			adev->gmc.gart_size = 512ULL << 20;
 			break;
@@ -857,6 +858,7 @@ static int gmc_v9_0_sw_init(void *handle)
 		break;
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 		/*
 		 * To fulfill 4-level page support,
 		 * vm size is 256TB (48bit), maximum size of Vega10,
@@ -974,6 +976,7 @@ static void gmc_v9_0_init_golden_registers(struct amdgpu_device *adev)
 
 	switch (adev->asic_type) {
 	case CHIP_VEGA10:
+	case CHIP_VEGA20:
 		soc15_program_register_sequence(adev,
 						golden_settings_mmhub_1_0_0,
 						ARRAY_SIZE(golden_settings_mmhub_1_0_0));
-- 
cgit v1.1


From c2d7fd2baeba4c65a3cf7f61d6d54c205e4608f8 Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Fri, 20 Apr 2018 13:58:09 +0800
Subject: drm/amdgpu/mmhub: Add clockgating support for vega20
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c
index 43f9257..3d53c44 100644
--- a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c
@@ -734,6 +734,7 @@ int mmhub_v1_0_set_clockgating(struct amdgpu_device *adev,
 	switch (adev->asic_type) {
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 	case CHIP_RAVEN:
 		mmhub_v1_0_update_medium_grain_clock_gating(adev,
 				state == AMD_CG_STATE_GATE ? true : false);
-- 
cgit v1.1


From 54a29ef758f6cc6b66b5f27dbfd90c9683920fab Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Fri, 20 Apr 2018 14:00:02 +0800
Subject: drm/amdgpu/sdma4: Specify vega20 firmware
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
index 03a36cb..79b3a45 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
@@ -42,6 +42,8 @@ MODULE_FIRMWARE("amdgpu/vega10_sdma.bin");
 MODULE_FIRMWARE("amdgpu/vega10_sdma1.bin");
 MODULE_FIRMWARE("amdgpu/vega12_sdma.bin");
 MODULE_FIRMWARE("amdgpu/vega12_sdma1.bin");
+MODULE_FIRMWARE("amdgpu/vega20_sdma.bin");
+MODULE_FIRMWARE("amdgpu/vega20_sdma1.bin");
 MODULE_FIRMWARE("amdgpu/raven_sdma.bin");
 
 #define SDMA0_POWER_CNTL__ON_OFF_CONDITION_HOLD_TIME_MASK  0x000000F8L
@@ -182,6 +184,9 @@ static int sdma_v4_0_init_microcode(struct amdgpu_device *adev)
 	case CHIP_VEGA12:
 		chip_name = "vega12";
 		break;
+	case CHIP_VEGA20:
+		chip_name = "vega20";
+		break;
 	case CHIP_RAVEN:
 		chip_name = "raven";
 		break;
-- 
cgit v1.1


From 84f50e9c80a74f9f8cac819c7a4b7ca220945b6d Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Tue, 23 Jan 2018 11:13:02 +0800
Subject: drm/amdgpu/sdma4: Add vega20 golden settings (v3)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

v2: squash in updates (Alex)
v3: squash in more updates (Alex)

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c | 27 +++++++++++++++++++++++++++
 1 file changed, 27 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
index 79b3a45..dc12c36 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
@@ -109,6 +109,28 @@ static const struct soc15_reg_golden golden_settings_sdma_4_1[] =
 	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_UTCL1_PAGE, 0x000003ff, 0x000003c0)
 };
 
+static const struct soc15_reg_golden golden_settings_sdma_4_2[] =
+{
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_CHICKEN_BITS, 0xfe931f07, 0x02831d07),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_CLK_CTRL, 0xffffffff, 0x3f000100),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_GB_ADDR_CONFIG, 0x0000773f, 0x00004002),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_GB_ADDR_CONFIG_READ, 0x0000773f, 0x00004002),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_GFX_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_PAGE_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_RLC0_RB_WPTR_POLL_CNTL, 0xfffffff0, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_RLC1_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_UTCL1_PAGE, 0x000003ff, 0x000003c0),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_CHICKEN_BITS, 0xfe931f07, 0x02831d07),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_CLK_CTRL, 0xffffffff, 0x3f000100),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_GB_ADDR_CONFIG, 0x0000773f, 0x00004002),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_GB_ADDR_CONFIG_READ, 0x0000773f, 0x00004002),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_GFX_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_PAGE_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_RLC0_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_RLC1_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_UTCL1_PAGE, 0x000003ff, 0x000003c0)
+};
+
 static const struct soc15_reg_golden golden_settings_sdma_rv1[] =
 {
 	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_GB_ADDR_CONFIG, 0x0018773f, 0x00000002),
@@ -141,6 +163,11 @@ static void sdma_v4_0_init_golden_registers(struct amdgpu_device *adev)
 						golden_settings_sdma_vg12,
 						ARRAY_SIZE(golden_settings_sdma_vg12));
 		break;
+	case CHIP_VEGA20:
+		soc15_program_register_sequence(adev,
+						golden_settings_sdma_4_2,
+						ARRAY_SIZE(golden_settings_sdma_4_2));
+		break;
 	case CHIP_RAVEN:
 		soc15_program_register_sequence(adev,
 						 golden_settings_sdma_4_1,
-- 
cgit v1.1


From 7eb32a7012ee592d6567e133a0d9c8c26e2590bf Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Tue, 23 Jan 2018 11:16:16 +0800
Subject: drm/amdgpu/sdma4: Add clockgating support for vega20
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
index dc12c36..ca53b3f 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
@@ -1548,6 +1548,7 @@ static int sdma_v4_0_set_clockgating_state(void *handle,
 	switch (adev->asic_type) {
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 	case CHIP_RAVEN:
 		sdma_v4_0_update_medium_grain_clock_gating(adev,
 				state == AMD_CG_STATE_GATE ? true : false);
-- 
cgit v1.1


From 940328fe35ab6e9f0eb1118f3cf91a22f97da298 Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Fri, 20 Apr 2018 14:22:48 +0800
Subject: drm/amdgpu/gfx9: Add support for vega20 firmware
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index b05b7ae..6976317 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -63,6 +63,13 @@ MODULE_FIRMWARE("amdgpu/vega12_mec.bin");
 MODULE_FIRMWARE("amdgpu/vega12_mec2.bin");
 MODULE_FIRMWARE("amdgpu/vega12_rlc.bin");
 
+MODULE_FIRMWARE("amdgpu/vega20_ce.bin");
+MODULE_FIRMWARE("amdgpu/vega20_pfp.bin");
+MODULE_FIRMWARE("amdgpu/vega20_me.bin");
+MODULE_FIRMWARE("amdgpu/vega20_mec.bin");
+MODULE_FIRMWARE("amdgpu/vega20_mec2.bin");
+MODULE_FIRMWARE("amdgpu/vega20_rlc.bin");
+
 MODULE_FIRMWARE("amdgpu/raven_ce.bin");
 MODULE_FIRMWARE("amdgpu/raven_pfp.bin");
 MODULE_FIRMWARE("amdgpu/raven_me.bin");
@@ -461,6 +468,9 @@ static int gfx_v9_0_init_microcode(struct amdgpu_device *adev)
 	case CHIP_VEGA12:
 		chip_name = "vega12";
 		break;
+	case CHIP_VEGA20:
+		chip_name = "vega20";
+		break;
 	case CHIP_RAVEN:
 		chip_name = "raven";
 		break;
-- 
cgit v1.1


From bb5368aac5b83c1fbb39ccd0d4a89af4465c84e2 Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Tue, 23 Jan 2018 14:47:26 +0800
Subject: drm/amdgpu/gfx9: Add vega20 golden settings (v3)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

v2: squash in updates (Alex)
v3: squash in more updates (Alex)

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 22 ++++++++++++++++++++++
 1 file changed, 22 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index 6976317..3749279 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -108,6 +108,20 @@ static const struct soc15_reg_golden golden_settings_gc_9_0_vg10[] =
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTD_CNTL, 0x00001800, 0x00000800)
 };
 
+static const struct soc15_reg_golden golden_settings_gc_9_0_vg20[] =
+{
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_HW_CONTROL_2, 0x0f000000, 0x0a000000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_HW_CONTROL_3, 0x30000000, 0x10000000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmGB_ADDR_CONFIG, 0xf3e777ff, 0x22014042),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmGB_ADDR_CONFIG_READ, 0xf3e777ff, 0x22014042),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmDB_DEBUG2, 0x00003e00, 0x00000400),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_ENHANCE_1, 0xff840000, 0x04040000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRMI_UTCL1_CNTL2, 0x00030000, 0x00030000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_CONFIG_CNTL_1, 0xffff010f, 0x01000107),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTA_CNTL_AUX, 0x000b0000, 0x000b0000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTD_CNTL, 0x01000000, 0x01000000)
+};
+
 static const struct soc15_reg_golden golden_settings_gc_9_1[] =
 {
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_HW_CONTROL, 0xfffdf3cf, 0x00014104),
@@ -241,6 +255,14 @@ static void gfx_v9_0_init_golden_registers(struct amdgpu_device *adev)
 						golden_settings_gc_9_2_1_vg12,
 						ARRAY_SIZE(golden_settings_gc_9_2_1_vg12));
 		break;
+	case CHIP_VEGA20:
+		soc15_program_register_sequence(adev,
+						golden_settings_gc_9_0,
+						ARRAY_SIZE(golden_settings_gc_9_0));
+		soc15_program_register_sequence(adev,
+						golden_settings_gc_9_0_vg20,
+						ARRAY_SIZE(golden_settings_gc_9_0_vg20));
+		break;
 	case CHIP_RAVEN:
 		soc15_program_register_sequence(adev,
 						 golden_settings_gc_9_1,
-- 
cgit v1.1


From d3adedb4559c01d18a934250e41a4660b4d89ac3 Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Fri, 20 Apr 2018 14:40:11 +0800
Subject: drm/amdgpu/gfx9: Add gfx config for vega20. (v4)

v2: clean up (Alex)
v3: additional cleanups (Alex)
v4: drop leftover TODO (Alex)

Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index 3749279..8335d98a 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -1137,6 +1137,16 @@ static void gfx_v9_0_gpu_early_init(struct amdgpu_device *adev)
 		gb_addr_config = VEGA12_GB_ADDR_CONFIG_GOLDEN;
 		DRM_INFO("fix gfx.config for vega12\n");
 		break;
+	case CHIP_VEGA20:
+		adev->gfx.config.max_hw_contexts = 8;
+		adev->gfx.config.sc_prim_fifo_size_frontend = 0x20;
+		adev->gfx.config.sc_prim_fifo_size_backend = 0x100;
+		adev->gfx.config.sc_hiz_tile_fifo_size = 0x30;
+		adev->gfx.config.sc_earlyz_tile_fifo_size = 0x4C0;
+		gb_addr_config = RREG32_SOC15(GC, 0, mmGB_ADDR_CONFIG);
+		gb_addr_config &= ~0xf3e777ff;
+		gb_addr_config |= 0x22014042;
+		break;
 	case CHIP_RAVEN:
 		adev->gfx.config.max_hw_contexts = 8;
 		adev->gfx.config.sc_prim_fifo_size_frontend = 0x20;
-- 
cgit v1.1


From 61324ddc5b7a43c3b989fbbb2ac5d99009a04d4b Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Fri, 20 Apr 2018 15:51:26 +0800
Subject: drm/amdgpu/gfx9: Add support for vega20
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index 8335d98a..92ed268 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -1429,6 +1429,7 @@ static int gfx_v9_0_sw_init(void *handle)
 	switch (adev->asic_type) {
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 	case CHIP_RAVEN:
 		adev->gfx.mec.num_mec = 2;
 		break;
@@ -4715,6 +4716,7 @@ static void gfx_v9_0_set_rlc_funcs(struct amdgpu_device *adev)
 	switch (adev->asic_type) {
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 	case CHIP_RAVEN:
 		adev->gfx.rlc.funcs = &gfx_v9_0_rlc_funcs;
 		break;
-- 
cgit v1.1


From 28b576b27a7acb29ce5b64da69d3855f6302350d Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Tue, 23 Jan 2018 15:03:36 +0800
Subject: drm/amdgpu/gfx9: Add clockgatting support for vega20
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index 92ed268..13253e0 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -3724,6 +3724,7 @@ static int gfx_v9_0_set_clockgating_state(void *handle,
 	switch (adev->asic_type) {
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 	case CHIP_RAVEN:
 		gfx_v9_0_update_gfx_clock_gating(adev,
 						 state == AMD_CG_STATE_GATE ? true : false);
-- 
cgit v1.1


From 935be7a0ce4e181a23fc840861088e79dcb3dc08 Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Fri, 26 Jan 2018 15:06:22 +0800
Subject: drm/amdgpu/soc15:Add vega20 soc15_common_early_init support
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Set external_rev_id and disable cg,pg for now.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/soc15.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index f31df18..f45bea8 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -658,6 +658,11 @@ static int soc15_common_early_init(void *handle)
 		adev->pg_flags = 0;
 		adev->external_rev_id = adev->rev_id + 0x14;
 		break;
+	case CHIP_VEGA20:
+		adev->cg_flags = 0;
+		adev->pg_flags = 0;
+		adev->external_rev_id = adev->rev_id + 0x28;
+		break;
 	case CHIP_RAVEN:
 		adev->cg_flags = AMD_CG_SUPPORT_GFX_MGCG |
 			AMD_CG_SUPPORT_GFX_MGLS |
-- 
cgit v1.1


From f980d127dba80214b4d793942492d3a4e6c46be0 Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Fri, 26 Jan 2018 15:10:55 +0800
Subject: drm/amdgpu/soc15: Set common clockgating for vega20.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Same as vega10 for now.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/soc15.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index f45bea8..1fd75f5 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -875,6 +875,7 @@ static int soc15_common_set_clockgating_state(void *handle,
 	switch (adev->asic_type) {
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 		adev->nbio_funcs->update_medium_grain_clock_gating(adev,
 				state == AMD_CG_STATE_GATE ? true : false);
 		adev->nbio_funcs->update_medium_grain_light_sleep(adev,
-- 
cgit v1.1


From 8ee273e516a096ee00b3be7cc15c8924aa3b1ef1 Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Fri, 23 Mar 2018 14:42:28 -0500
Subject: drm/amdgpu/soc15: dynamic initialize ip offset for vega20
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Vega20 need a seperate vega20_reg_init.c due to ip base
offset difference.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/Makefile          |  3 +-
 drivers/gpu/drm/amd/amdgpu/soc15.c           |  3 ++
 drivers/gpu/drm/amd/amdgpu/soc15.h           |  1 +
 drivers/gpu/drm/amd/amdgpu/vega20_reg_init.c | 53 ++++++++++++++++++++++++++++
 4 files changed, 59 insertions(+), 1 deletion(-)
 create mode 100644 drivers/gpu/drm/amd/amdgpu/vega20_reg_init.c

diff --git a/drivers/gpu/drm/amd/amdgpu/Makefile b/drivers/gpu/drm/amd/amdgpu/Makefile
index 68e9f58..012ea37 100644
--- a/drivers/gpu/drm/amd/amdgpu/Makefile
+++ b/drivers/gpu/drm/amd/amdgpu/Makefile
@@ -62,7 +62,8 @@ amdgpu-$(CONFIG_DRM_AMDGPU_CIK)+= cik.o cik_ih.o kv_smc.o kv_dpm.o \
 amdgpu-$(CONFIG_DRM_AMDGPU_SI)+= si.o gmc_v6_0.o gfx_v6_0.o si_ih.o si_dma.o dce_v6_0.o si_dpm.o si_smc.o
 
 amdgpu-y += \
-	vi.o mxgpu_vi.o nbio_v6_1.o soc15.o emu_soc.o mxgpu_ai.o nbio_v7_0.o vega10_reg_init.o
+	vi.o mxgpu_vi.o nbio_v6_1.o soc15.o emu_soc.o mxgpu_ai.o nbio_v7_0.o vega10_reg_init.o \
+	vega20_reg_init.o
 
 # add DF block
 amdgpu-y += \
diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index 1fd75f5..c3133d1 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -487,6 +487,9 @@ int soc15_set_ip_blocks(struct amdgpu_device *adev)
 	case CHIP_RAVEN:
 		vega10_reg_base_init(adev);
 		break;
+	case CHIP_VEGA20:
+		vega20_reg_base_init(adev);
+		break;
 	default:
 		return -EINVAL;
 	}
diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.h b/drivers/gpu/drm/amd/amdgpu/soc15.h
index f70da8a..1f714b7 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.h
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.h
@@ -55,5 +55,6 @@ void soc15_program_register_sequence(struct amdgpu_device *adev,
 					     const u32 array_size);
 
 int vega10_reg_base_init(struct amdgpu_device *adev);
+int vega20_reg_base_init(struct amdgpu_device *adev);
 
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/vega20_reg_init.c b/drivers/gpu/drm/amd/amdgpu/vega20_reg_init.c
new file mode 100644
index 0000000..52778de
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/vega20_reg_init.c
@@ -0,0 +1,53 @@
+/*
+ * Copyright 2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#include "amdgpu.h"
+#include "soc15.h"
+
+#include "soc15_common.h"
+#include "soc15_hw_ip.h"
+#include "vega20_ip_offset.h"
+
+int vega20_reg_base_init(struct amdgpu_device *adev)
+{
+	/* HW has more IP blocks,  only initialized the blocke beend by our driver  */
+	uint32_t i;
+	for (i = 0 ; i < MAX_INSTANCE ; ++i) {
+		adev->reg_offset[GC_HWIP][i] = (uint32_t *)(&(GC_BASE.instance[i]));
+		adev->reg_offset[HDP_HWIP][i] = (uint32_t *)(&(HDP_BASE.instance[i]));
+		adev->reg_offset[MMHUB_HWIP][i] = (uint32_t *)(&(MMHUB_BASE.instance[i]));
+		adev->reg_offset[ATHUB_HWIP][i] = (uint32_t *)(&(ATHUB_BASE.instance[i]));
+		adev->reg_offset[NBIO_HWIP][i] = (uint32_t *)(&(NBIO_BASE.instance[i]));
+		adev->reg_offset[MP0_HWIP][i] = (uint32_t *)(&(MP0_BASE.instance[i]));
+		adev->reg_offset[UVD_HWIP][i] = (uint32_t *)(&(UVD_BASE.instance[i]));
+		adev->reg_offset[VCE_HWIP][i] = (uint32_t *)(&(VCE_BASE.instance[i]));
+		adev->reg_offset[DF_HWIP][i] = (uint32_t *)(&(DF_BASE.instance[i]));
+		adev->reg_offset[DCE_HWIP][i] = (uint32_t *)(&(DCE_BASE.instance[i]));
+		adev->reg_offset[OSSSYS_HWIP][i] = (uint32_t *)(&(OSSSYS_BASE.instance[i]));
+		adev->reg_offset[SDMA0_HWIP][i] = (uint32_t *)(&(SDMA0_BASE.instance[i]));
+		adev->reg_offset[SDMA1_HWIP][i] = (uint32_t *)(&(SDMA1_BASE.instance[i]));
+		adev->reg_offset[SMUIO_HWIP][i] = (uint32_t *)(&(SMUIO_BASE.instance[i]));
+	}
+	return 0;
+}
+
+
-- 
cgit v1.1


From 7c7af6c10d5dc733c2f181f653cb0a5b64e372a5 Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Fri, 20 Apr 2018 18:35:42 +0800
Subject: drm/amdgpu/soc15: Add ip blocks for vega20 (v2)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Same as vega10 now.

v2: squash in typo fix

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/soc15.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index c3133d1..10337fb 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -508,6 +508,7 @@ int soc15_set_ip_blocks(struct amdgpu_device *adev)
 	switch (adev->asic_type) {
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 		amdgpu_device_ip_block_add(adev, &vega10_common_ip_block);
 		amdgpu_device_ip_block_add(adev, &gmc_v9_0_ip_block);
 		amdgpu_device_ip_block_add(adev, &vega10_ih_ip_block);
-- 
cgit v1.1


From a95d89e2d8e268d90d0f97c9c57d61006eec78c3 Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Fri, 23 Mar 2018 14:44:28 -0500
Subject: drm/amdgpu: Add nbio support for vega20 (v2)

Some register offset in nbio v7.4 are different with v7.0.

v2: Use nbio7.0 for now.

TODO: add a new nbio 7.4 module (Alex)

Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c | 18 +++++++++++++++++-
 drivers/gpu/drm/amd/amdgpu/soc15.c     |  2 ++
 2 files changed, 19 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c b/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c
index df34dc7..365517c 100644
--- a/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c
@@ -34,10 +34,19 @@
 #define smnCPM_CONTROL                                                                                  0x11180460
 #define smnPCIE_CNTL2                                                                                   0x11180070
 
+/* vega20 */
+#define mmRCC_DEV0_EPF0_STRAP0_VG20                                                                         0x0011
+#define mmRCC_DEV0_EPF0_STRAP0_VG20_BASE_IDX                                                                2
+
 static u32 nbio_v7_0_get_rev_id(struct amdgpu_device *adev)
 {
         u32 tmp = RREG32_SOC15(NBIO, 0, mmRCC_DEV0_EPF0_STRAP0);
 
+	if (adev->asic_type == CHIP_VEGA20)
+		tmp = RREG32_SOC15(NBIO, 0, mmRCC_DEV0_EPF0_STRAP0_VG20);
+	else
+		tmp = RREG32_SOC15(NBIO, 0, mmRCC_DEV0_EPF0_STRAP0);
+
 	tmp &= RCC_DEV0_EPF0_STRAP0__STRAP_ATI_REV_ID_DEV0_F0_MASK;
 	tmp >>= RCC_DEV0_EPF0_STRAP0__STRAP_ATI_REV_ID_DEV0_F0__SHIFT;
 
@@ -75,10 +84,14 @@ static void nbio_v7_0_sdma_doorbell_range(struct amdgpu_device *adev, int instan
 			SOC15_REG_OFFSET(NBIO, 0, mmBIF_SDMA1_DOORBELL_RANGE);
 
 	u32 doorbell_range = RREG32(reg);
+	u32 range = 2;
+
+	if (adev->asic_type == CHIP_VEGA20)
+		range = 8;
 
 	if (use_doorbell) {
 		doorbell_range = REG_SET_FIELD(doorbell_range, BIF_SDMA0_DOORBELL_RANGE, OFFSET, doorbell_index);
-		doorbell_range = REG_SET_FIELD(doorbell_range, BIF_SDMA0_DOORBELL_RANGE, SIZE, 2);
+		doorbell_range = REG_SET_FIELD(doorbell_range, BIF_SDMA0_DOORBELL_RANGE, SIZE, range);
 	} else
 		doorbell_range = REG_SET_FIELD(doorbell_range, BIF_SDMA0_DOORBELL_RANGE, SIZE, 0);
 
@@ -133,6 +146,9 @@ static void nbio_v7_0_update_medium_grain_clock_gating(struct amdgpu_device *ade
 {
 	uint32_t def, data;
 
+	if (adev->asic_type == CHIP_VEGA20)
+		return;
+
 	/* NBIF_MGCG_CTRL_LCLK */
 	def = data = RREG32_PCIE(smnNBIF_MGCG_CTRL_LCLK);
 
diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index 10337fb..4e065c6 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -496,6 +496,8 @@ int soc15_set_ip_blocks(struct amdgpu_device *adev)
 
 	if (adev->flags & AMD_IS_APU)
 		adev->nbio_funcs = &nbio_v7_0_funcs;
+	else if (adev->asic_type == CHIP_VEGA20)
+		adev->nbio_funcs = &nbio_v7_0_funcs;
 	else
 		adev->nbio_funcs = &nbio_v6_1_funcs;
 
-- 
cgit v1.1


From 1fe6bf2f33fe6728cfb206e2ce476cb2d1dae406 Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Fri, 20 Apr 2018 19:50:01 +0800
Subject: drm/amd/display/dm: Add vega20 support
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index f2f54a9..6f5cb26 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -1115,6 +1115,7 @@ static int dce110_register_irq_handlers(struct amdgpu_device *adev)
 
 	if (adev->asic_type == CHIP_VEGA10 ||
 	    adev->asic_type == CHIP_VEGA12 ||
+	    adev->asic_type == CHIP_VEGA20 ||
 	    adev->asic_type == CHIP_RAVEN)
 		client_id = SOC15_IH_CLIENTID_DCE;
 
@@ -1518,6 +1519,7 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
 #endif
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 		if (dce110_register_irq_handlers(dm->adev)) {
 			DRM_ERROR("DM: Failed to initialize IRQ\n");
 			goto fail;
@@ -1718,6 +1720,7 @@ static int dm_early_init(void *handle)
 		break;
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 		adev->mode_info.num_crtc = 6;
 		adev->mode_info.num_hpd = 6;
 		adev->mode_info.num_dig = 6;
@@ -1966,6 +1969,7 @@ static int fill_plane_attributes_from_fb(struct amdgpu_device *adev,
 
 	if (adev->asic_type == CHIP_VEGA10 ||
 	    adev->asic_type == CHIP_VEGA12 ||
+	    adev->asic_type == CHIP_VEGA20 ||
 	    adev->asic_type == CHIP_RAVEN) {
 		/* Fill GFX9 params */
 		plane_state->tiling_info.gfx9.num_pipes =
-- 
cgit v1.1


From c6034aa2c4fc54bbe429cc6414f83a25bb4913f7 Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Sat, 3 Feb 2018 12:19:46 +0800
Subject: drm/amdgpu: Add vega20 to dc support check (v2)

v2: fix whitespace

Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 2d46ad7..0e3f69d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -2158,6 +2158,7 @@ bool amdgpu_device_asic_has_dc_support(enum amd_asic_type asic_type)
 	case CHIP_FIJI:
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 #if defined(CONFIG_DRM_AMD_DC_DCN1_0)
 	case CHIP_RAVEN:
 #endif
-- 
cgit v1.1


From d82420b56a17d5b39579bc46f8dad757be684f94 Mon Sep 17 00:00:00 2001
From: Roman Li <Roman.Li@amd.com>
Date: Wed, 14 Feb 2018 17:20:54 -0500
Subject: drm/amd: Add dce-12.1 gpio aux registers (v2)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Updating dce12 register headers by adding dc registers
required for potential DP LTTPR support.

v2: fix mode change

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Roman Li <Roman.Li@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../drm/amd/include/asic_reg/dce/dce_12_0_offset.h |  12 ++
 .../amd/include/asic_reg/dce/dce_12_0_sh_mask.h    | 152 +++++++++++++++++++++
 2 files changed, 164 insertions(+)

diff --git a/drivers/gpu/drm/amd/include/asic_reg/dce/dce_12_0_offset.h b/drivers/gpu/drm/amd/include/asic_reg/dce/dce_12_0_offset.h
index f730d06..b6f74bf 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/dce/dce_12_0_offset.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/dce/dce_12_0_offset.h
@@ -2095,6 +2095,18 @@
 #define mmDC_GPIO_AUX_CTRL_2_BASE_IDX                                                                  2
 #define mmDC_GPIO_RXEN                                                                                 0x212f
 #define mmDC_GPIO_RXEN_BASE_IDX                                                                        2
+#define mmDC_GPIO_AUX_CTRL_3                                                                           0x2130
+#define mmDC_GPIO_AUX_CTRL_3_BASE_IDX                                                                  2
+#define mmDC_GPIO_AUX_CTRL_4                                                                           0x2131
+#define mmDC_GPIO_AUX_CTRL_4_BASE_IDX                                                                  2
+#define mmDC_GPIO_AUX_CTRL_5                                                                           0x2132
+#define mmDC_GPIO_AUX_CTRL_5_BASE_IDX                                                                  2
+#define mmAUXI2C_PAD_ALL_PWR_OK                                                                        0x2133
+#define mmAUXI2C_PAD_ALL_PWR_OK_BASE_IDX                                                               2
+#define mmDC_GPIO_PULLUPEN                                                                             0x2134
+#define mmDC_GPIO_PULLUPEN_BASE_IDX                                                                    2
+#define mmDC_GPIO_AUX_CTRL_6                                                                           0x2135
+#define mmDC_GPIO_AUX_CTRL_6_BASE_IDX                                                                  2
 #define mmBPHYC_DAC_MACRO_CNTL                                                                         0x2136
 #define mmBPHYC_DAC_MACRO_CNTL_BASE_IDX                                                                2
 #define mmDAC_MACRO_CNTL_RESERVED0                                                                     0x2136
diff --git a/drivers/gpu/drm/amd/include/asic_reg/dce/dce_12_0_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/dce/dce_12_0_sh_mask.h
index 6d3162c..bcd190a 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/dce/dce_12_0_sh_mask.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/dce/dce_12_0_sh_mask.h
@@ -10971,6 +10971,158 @@
 #define DC_GPIO_RXEN__DC_GPIO_BLON_RXEN_MASK                                                                  0x00100000L
 #define DC_GPIO_RXEN__DC_GPIO_DIGON_RXEN_MASK                                                                 0x00200000L
 #define DC_GPIO_RXEN__DC_GPIO_ENA_BL_RXEN_MASK                                                                0x00400000L
+//DC_GPIO_AUX_CTRL_3
+#define DC_GPIO_AUX_CTRL_3__AUX1_NEN_RTERM__SHIFT                                                             0x0
+#define DC_GPIO_AUX_CTRL_3__AUX2_NEN_RTERM__SHIFT                                                             0x1
+#define DC_GPIO_AUX_CTRL_3__AUX3_NEN_RTERM__SHIFT                                                             0x2
+#define DC_GPIO_AUX_CTRL_3__AUX4_NEN_RTERM__SHIFT                                                             0x3
+#define DC_GPIO_AUX_CTRL_3__AUX5_NEN_RTERM__SHIFT                                                             0x4
+#define DC_GPIO_AUX_CTRL_3__AUX6_NEN_RTERM__SHIFT                                                             0x5
+#define DC_GPIO_AUX_CTRL_3__AUX1_DP_DN_SWAP__SHIFT                                                            0x8
+#define DC_GPIO_AUX_CTRL_3__AUX2_DP_DN_SWAP__SHIFT                                                            0x9
+#define DC_GPIO_AUX_CTRL_3__AUX3_DP_DN_SWAP__SHIFT                                                            0xa
+#define DC_GPIO_AUX_CTRL_3__AUX4_DP_DN_SWAP__SHIFT                                                            0xb
+#define DC_GPIO_AUX_CTRL_3__AUX5_DP_DN_SWAP__SHIFT                                                            0xc
+#define DC_GPIO_AUX_CTRL_3__AUX6_DP_DN_SWAP__SHIFT                                                            0xd
+#define DC_GPIO_AUX_CTRL_3__AUX1_HYS_TUNE__SHIFT                                                              0x10
+#define DC_GPIO_AUX_CTRL_3__AUX2_HYS_TUNE__SHIFT                                                              0x12
+#define DC_GPIO_AUX_CTRL_3__AUX3_HYS_TUNE__SHIFT                                                              0x14
+#define DC_GPIO_AUX_CTRL_3__AUX4_HYS_TUNE__SHIFT                                                              0x16
+#define DC_GPIO_AUX_CTRL_3__AUX5_HYS_TUNE__SHIFT                                                              0x18
+#define DC_GPIO_AUX_CTRL_3__AUX6_HYS_TUNE__SHIFT                                                              0x1a
+#define DC_GPIO_AUX_CTRL_3__AUX1_NEN_RTERM_MASK                                                               0x00000001L
+#define DC_GPIO_AUX_CTRL_3__AUX2_NEN_RTERM_MASK                                                               0x00000002L
+#define DC_GPIO_AUX_CTRL_3__AUX3_NEN_RTERM_MASK                                                               0x00000004L
+#define DC_GPIO_AUX_CTRL_3__AUX4_NEN_RTERM_MASK                                                               0x00000008L
+#define DC_GPIO_AUX_CTRL_3__AUX5_NEN_RTERM_MASK                                                               0x00000010L
+#define DC_GPIO_AUX_CTRL_3__AUX6_NEN_RTERM_MASK                                                               0x00000020L
+#define DC_GPIO_AUX_CTRL_3__AUX1_DP_DN_SWAP_MASK                                                              0x00000100L
+#define DC_GPIO_AUX_CTRL_3__AUX2_DP_DN_SWAP_MASK                                                              0x00000200L
+#define DC_GPIO_AUX_CTRL_3__AUX3_DP_DN_SWAP_MASK                                                              0x00000400L
+#define DC_GPIO_AUX_CTRL_3__AUX4_DP_DN_SWAP_MASK                                                              0x00000800L
+#define DC_GPIO_AUX_CTRL_3__AUX5_DP_DN_SWAP_MASK                                                              0x00001000L
+#define DC_GPIO_AUX_CTRL_3__AUX6_DP_DN_SWAP_MASK                                                              0x00002000L
+#define DC_GPIO_AUX_CTRL_3__AUX1_HYS_TUNE_MASK                                                                0x00030000L
+#define DC_GPIO_AUX_CTRL_3__AUX2_HYS_TUNE_MASK                                                                0x000C0000L
+#define DC_GPIO_AUX_CTRL_3__AUX3_HYS_TUNE_MASK                                                                0x00300000L
+#define DC_GPIO_AUX_CTRL_3__AUX4_HYS_TUNE_MASK                                                                0x00C00000L
+#define DC_GPIO_AUX_CTRL_3__AUX5_HYS_TUNE_MASK                                                                0x03000000L
+#define DC_GPIO_AUX_CTRL_3__AUX6_HYS_TUNE_MASK                                                                0x0C000000L
+//DC_GPIO_AUX_CTRL_4
+#define DC_GPIO_AUX_CTRL_4__AUX1_AUX_CTRL__SHIFT                                                              0x0
+#define DC_GPIO_AUX_CTRL_4__AUX2_AUX_CTRL__SHIFT                                                              0x4
+#define DC_GPIO_AUX_CTRL_4__AUX3_AUX_CTRL__SHIFT                                                              0x8
+#define DC_GPIO_AUX_CTRL_4__AUX4_AUX_CTRL__SHIFT                                                              0xc
+#define DC_GPIO_AUX_CTRL_4__AUX5_AUX_CTRL__SHIFT                                                              0x10
+#define DC_GPIO_AUX_CTRL_4__AUX6_AUX_CTRL__SHIFT                                                              0x14
+#define DC_GPIO_AUX_CTRL_4__AUX1_AUX_CTRL_MASK                                                                0x0000000FL
+#define DC_GPIO_AUX_CTRL_4__AUX2_AUX_CTRL_MASK                                                                0x000000F0L
+#define DC_GPIO_AUX_CTRL_4__AUX3_AUX_CTRL_MASK                                                                0x00000F00L
+#define DC_GPIO_AUX_CTRL_4__AUX4_AUX_CTRL_MASK                                                                0x0000F000L
+#define DC_GPIO_AUX_CTRL_4__AUX5_AUX_CTRL_MASK                                                                0x000F0000L
+#define DC_GPIO_AUX_CTRL_4__AUX6_AUX_CTRL_MASK                                                                0x00F00000L
+//DC_GPIO_AUX_CTRL_5
+#define DC_GPIO_AUX_CTRL_5__AUX1_VOD_TUNE__SHIFT                                                              0x0
+#define DC_GPIO_AUX_CTRL_5__AUX2_VOD_TUNE__SHIFT                                                              0x2
+#define DC_GPIO_AUX_CTRL_5__AUX3_VOD_TUNE__SHIFT                                                              0x4
+#define DC_GPIO_AUX_CTRL_5__AUX4_VOD_TUNE__SHIFT                                                              0x6
+#define DC_GPIO_AUX_CTRL_5__AUX5_VOD_TUNE__SHIFT                                                              0x8
+#define DC_GPIO_AUX_CTRL_5__AUX6_VOD_TUNE__SHIFT                                                              0xa
+#define DC_GPIO_AUX_CTRL_5__DDC_PAD1_I2CMODE__SHIFT                                                           0xc
+#define DC_GPIO_AUX_CTRL_5__DDC_PAD2_I2CMODE__SHIFT                                                           0xd
+#define DC_GPIO_AUX_CTRL_5__DDC_PAD3_I2CMODE__SHIFT                                                           0xe
+#define DC_GPIO_AUX_CTRL_5__DDC_PAD4_I2CMODE__SHIFT                                                           0xf
+#define DC_GPIO_AUX_CTRL_5__DDC_PAD5_I2CMODE__SHIFT                                                           0x10
+#define DC_GPIO_AUX_CTRL_5__DDC_PAD6_I2CMODE__SHIFT                                                           0x11
+#define DC_GPIO_AUX_CTRL_5__DDC1_I2C_VPH_1V2_EN__SHIFT                                                        0x12
+#define DC_GPIO_AUX_CTRL_5__DDC2_I2C_VPH_1V2_EN__SHIFT                                                        0x13
+#define DC_GPIO_AUX_CTRL_5__DDC3_I2C_VPH_1V2_EN__SHIFT                                                        0x14
+#define DC_GPIO_AUX_CTRL_5__DDC4_I2C_VPH_1V2_EN__SHIFT                                                        0x15
+#define DC_GPIO_AUX_CTRL_5__DDC5_I2C_VPH_1V2_EN__SHIFT                                                        0x16
+#define DC_GPIO_AUX_CTRL_5__DDC6_I2C_VPH_1V2_EN__SHIFT                                                        0x17
+#define DC_GPIO_AUX_CTRL_5__DDC1_PAD_I2C_CTRL__SHIFT                                                          0x18
+#define DC_GPIO_AUX_CTRL_5__DDC2_PAD_I2C_CTRL__SHIFT                                                          0x19
+#define DC_GPIO_AUX_CTRL_5__DDC3_PAD_I2C_CTRL__SHIFT                                                          0x1a
+#define DC_GPIO_AUX_CTRL_5__DDC4_PAD_I2C_CTRL__SHIFT                                                          0x1b
+#define DC_GPIO_AUX_CTRL_5__DDC5_PAD_I2C_CTRL__SHIFT                                                          0x1c
+#define DC_GPIO_AUX_CTRL_5__DDC6_PAD_I2C_CTRL__SHIFT                                                          0x1d
+#define DC_GPIO_AUX_CTRL_5__AUX1_VOD_TUNE_MASK                                                                0x00000003L
+#define DC_GPIO_AUX_CTRL_5__AUX2_VOD_TUNE_MASK                                                                0x0000000CL
+#define DC_GPIO_AUX_CTRL_5__AUX3_VOD_TUNE_MASK                                                                0x00000030L
+#define DC_GPIO_AUX_CTRL_5__AUX4_VOD_TUNE_MASK                                                                0x000000C0L
+#define DC_GPIO_AUX_CTRL_5__AUX5_VOD_TUNE_MASK                                                                0x00000300L
+#define DC_GPIO_AUX_CTRL_5__AUX6_VOD_TUNE_MASK                                                                0x00000C00L
+#define DC_GPIO_AUX_CTRL_5__DDC_PAD1_I2CMODE_MASK                                                             0x00001000L
+#define DC_GPIO_AUX_CTRL_5__DDC_PAD2_I2CMODE_MASK                                                             0x00002000L
+#define DC_GPIO_AUX_CTRL_5__DDC_PAD3_I2CMODE_MASK                                                             0x00004000L
+#define DC_GPIO_AUX_CTRL_5__DDC_PAD4_I2CMODE_MASK                                                             0x00008000L
+#define DC_GPIO_AUX_CTRL_5__DDC_PAD5_I2CMODE_MASK                                                             0x00010000L
+#define DC_GPIO_AUX_CTRL_5__DDC_PAD6_I2CMODE_MASK                                                             0x00020000L
+#define DC_GPIO_AUX_CTRL_5__DDC1_I2C_VPH_1V2_EN_MASK                                                          0x00040000L
+#define DC_GPIO_AUX_CTRL_5__DDC2_I2C_VPH_1V2_EN_MASK                                                          0x00080000L
+#define DC_GPIO_AUX_CTRL_5__DDC3_I2C_VPH_1V2_EN_MASK                                                          0x00100000L
+#define DC_GPIO_AUX_CTRL_5__DDC4_I2C_VPH_1V2_EN_MASK                                                          0x00200000L
+#define DC_GPIO_AUX_CTRL_5__DDC5_I2C_VPH_1V2_EN_MASK                                                          0x00400000L
+#define DC_GPIO_AUX_CTRL_5__DDC6_I2C_VPH_1V2_EN_MASK                                                          0x00800000L
+#define DC_GPIO_AUX_CTRL_5__DDC1_PAD_I2C_CTRL_MASK                                                            0x01000000L
+#define DC_GPIO_AUX_CTRL_5__DDC2_PAD_I2C_CTRL_MASK                                                            0x02000000L
+#define DC_GPIO_AUX_CTRL_5__DDC3_PAD_I2C_CTRL_MASK                                                            0x04000000L
+#define DC_GPIO_AUX_CTRL_5__DDC4_PAD_I2C_CTRL_MASK                                                            0x08000000L
+#define DC_GPIO_AUX_CTRL_5__DDC5_PAD_I2C_CTRL_MASK                                                            0x10000000L
+#define DC_GPIO_AUX_CTRL_5__DDC6_PAD_I2C_CTRL_MASK                                                            0x20000000L
+//AUXI2C_PAD_ALL_PWR_OK
+#define AUXI2C_PAD_ALL_PWR_OK__AUXI2C_PHY1_ALL_PWR_OK__SHIFT                                                  0x0
+#define AUXI2C_PAD_ALL_PWR_OK__AUXI2C_PHY2_ALL_PWR_OK__SHIFT                                                  0x1
+#define AUXI2C_PAD_ALL_PWR_OK__AUXI2C_PHY3_ALL_PWR_OK__SHIFT                                                  0x2
+#define AUXI2C_PAD_ALL_PWR_OK__AUXI2C_PHY4_ALL_PWR_OK__SHIFT                                                  0x3
+#define AUXI2C_PAD_ALL_PWR_OK__AUXI2C_PHY5_ALL_PWR_OK__SHIFT                                                  0x4
+#define AUXI2C_PAD_ALL_PWR_OK__AUXI2C_PHY6_ALL_PWR_OK__SHIFT                                                  0x5
+#define AUXI2C_PAD_ALL_PWR_OK__AUXI2C_PHY1_ALL_PWR_OK_MASK                                                    0x00000001L
+#define AUXI2C_PAD_ALL_PWR_OK__AUXI2C_PHY2_ALL_PWR_OK_MASK                                                    0x00000002L
+#define AUXI2C_PAD_ALL_PWR_OK__AUXI2C_PHY3_ALL_PWR_OK_MASK                                                    0x00000004L
+#define AUXI2C_PAD_ALL_PWR_OK__AUXI2C_PHY4_ALL_PWR_OK_MASK                                                    0x00000008L
+#define AUXI2C_PAD_ALL_PWR_OK__AUXI2C_PHY5_ALL_PWR_OK_MASK                                                    0x00000010L
+#define AUXI2C_PAD_ALL_PWR_OK__AUXI2C_PHY6_ALL_PWR_OK_MASK                                                    0x00000020L
+//DC_GPIO_PULLUPEN
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICA_PU_EN__SHIFT                                                       0x0
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICB_PU_EN__SHIFT                                                       0x1
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICC_PU_EN__SHIFT                                                       0x2
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICD_PU_EN__SHIFT                                                       0x3
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICE_PU_EN__SHIFT                                                       0x4
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICF_PU_EN__SHIFT                                                       0x5
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICG_PU_EN__SHIFT                                                       0x6
+#define DC_GPIO_PULLUPEN__DC_GPIO_HSYNCA_PU_EN__SHIFT                                                         0x8
+#define DC_GPIO_PULLUPEN__DC_GPIO_VSYNCA_PU_EN__SHIFT                                                         0x9
+#define DC_GPIO_PULLUPEN__DC_GPIO_HPD1_PU_EN__SHIFT                                                           0xe
+#define DC_GPIO_PULLUPEN__DC_GPIO_BLON_PU_EN__SHIFT                                                           0x14
+#define DC_GPIO_PULLUPEN__DC_GPIO_DIGON_PU_EN__SHIFT                                                          0x15
+#define DC_GPIO_PULLUPEN__DC_GPIO_ENA_BL_PU_EN__SHIFT                                                         0x16
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICA_PU_EN_MASK                                                         0x00000001L
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICB_PU_EN_MASK                                                         0x00000002L
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICC_PU_EN_MASK                                                         0x00000004L
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICD_PU_EN_MASK                                                         0x00000008L
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICE_PU_EN_MASK                                                         0x00000010L
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICF_PU_EN_MASK                                                         0x00000020L
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICG_PU_EN_MASK                                                         0x00000040L
+#define DC_GPIO_PULLUPEN__DC_GPIO_HSYNCA_PU_EN_MASK                                                           0x00000100L
+#define DC_GPIO_PULLUPEN__DC_GPIO_VSYNCA_PU_EN_MASK                                                           0x00000200L
+#define DC_GPIO_PULLUPEN__DC_GPIO_HPD1_PU_EN_MASK                                                             0x00004000L
+#define DC_GPIO_PULLUPEN__DC_GPIO_BLON_PU_EN_MASK                                                             0x00100000L
+#define DC_GPIO_PULLUPEN__DC_GPIO_DIGON_PU_EN_MASK                                                            0x00200000L
+#define DC_GPIO_PULLUPEN__DC_GPIO_ENA_BL_PU_EN_MASK                                                           0x00400000L
+//DC_GPIO_AUX_CTRL_6
+#define DC_GPIO_AUX_CTRL_6__AUX1_PAD_RXSEL__SHIFT                                                             0x0
+#define DC_GPIO_AUX_CTRL_6__AUX2_PAD_RXSEL__SHIFT                                                             0x2
+#define DC_GPIO_AUX_CTRL_6__AUX3_PAD_RXSEL__SHIFT                                                             0x4
+#define DC_GPIO_AUX_CTRL_6__AUX4_PAD_RXSEL__SHIFT                                                             0x6
+#define DC_GPIO_AUX_CTRL_6__AUX5_PAD_RXSEL__SHIFT                                                             0x8
+#define DC_GPIO_AUX_CTRL_6__AUX6_PAD_RXSEL__SHIFT                                                             0xa
+#define DC_GPIO_AUX_CTRL_6__AUX1_PAD_RXSEL_MASK                                                               0x00000003L
+#define DC_GPIO_AUX_CTRL_6__AUX2_PAD_RXSEL_MASK                                                               0x0000000CL
+#define DC_GPIO_AUX_CTRL_6__AUX3_PAD_RXSEL_MASK                                                               0x00000030L
+#define DC_GPIO_AUX_CTRL_6__AUX4_PAD_RXSEL_MASK                                                               0x000000C0L
+#define DC_GPIO_AUX_CTRL_6__AUX5_PAD_RXSEL_MASK                                                               0x00000300L
+#define DC_GPIO_AUX_CTRL_6__AUX6_PAD_RXSEL_MASK                                                               0x00000C00L
 //BPHYC_DAC_MACRO_CNTL
 #define BPHYC_DAC_MACRO_CNTL__BPHYC_DAC_WHITE_LEVEL__SHIFT                                                    0x0
 #define BPHYC_DAC_MACRO_CNTL__BPHYC_DAC_WHITE_FINE_CONTROL__SHIFT                                             0x8
-- 
cgit v1.1


From 138bc36051f817ce5bee33b0e7a4873bb04f1eb4 Mon Sep 17 00:00:00 2001
From: "Jerry (Fangzhi) Zuo" <Jerry.Zuo@amd.com>
Date: Fri, 11 May 2018 13:46:19 -0500
Subject: drm/amd/display: Add Vega20 config. support
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Jerry (Fangzhi) Zuo <Jerry.Zuo@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/Kconfig | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/Kconfig b/drivers/gpu/drm/amd/display/Kconfig
index e6ca72c..6dcec9c 100644
--- a/drivers/gpu/drm/amd/display/Kconfig
+++ b/drivers/gpu/drm/amd/display/Kconfig
@@ -40,4 +40,13 @@ config DRM_AMD_DC_VEGAM
         help
          Choose this option if you want to have
          VEGAM support for display engine
+
+config DRM_AMD_DC_VG20
+	bool "Vega20 support"
+	depends on DRM_AMD_DC
+	help
+		Choose this option if you want to have
+		Vega20 support for display engine
+
+
 endmenu
-- 
cgit v1.1


From 14a13a0ef0665924a5e87947309b6c9abfb41903 Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Fri, 20 Apr 2018 21:03:10 +0800
Subject: drm/amd/display: Remove COMBO_DISPLAY_PLL0 from Vega20

Signed-off-by: Jerry (Fangzhi) Zuo <Jerry.Zuo@amd.com>
Reviewed-by: Hersen Wu <hersenxs.wu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c   | 11 ++++++++++-
 drivers/gpu/drm/amd/display/include/dal_asic_id.h |  6 ++++++
 2 files changed, 16 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c b/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c
index 78e6beb..aa4cf30 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c
@@ -35,7 +35,7 @@
 #endif
 #include "core_types.h"
 #include "dc_types.h"
-
+#include "dal_asic_id.h"
 
 #define TO_DCE_CLOCKS(clocks)\
 	container_of(clocks, struct dce_disp_clk, base)
@@ -413,9 +413,18 @@ static int dce112_set_clock(
 	/*VBIOS will determine DPREFCLK frequency, so we don't set it*/
 	dce_clk_params.target_clock_frequency = 0;
 	dce_clk_params.clock_type = DCECLOCK_TYPE_DPREFCLK;
+#ifndef CONFIG_DRM_AMD_DC_VG20
 	dce_clk_params.flags.USE_GENLOCK_AS_SOURCE_FOR_DPREFCLK =
 			(dce_clk_params.pll_id ==
 					CLOCK_SOURCE_COMBO_DISPLAY_PLL0);
+#else
+	if (!ASICREV_IS_VEGA20_P(clk->ctx->asic_id.hw_internal_rev))
+		dce_clk_params.flags.USE_GENLOCK_AS_SOURCE_FOR_DPREFCLK =
+			(dce_clk_params.pll_id ==
+					CLOCK_SOURCE_COMBO_DISPLAY_PLL0);
+	else
+		dce_clk_params.flags.USE_GENLOCK_AS_SOURCE_FOR_DPREFCLK = false;
+#endif
 
 	bp->funcs->set_dce_clock(bp, &dce_clk_params);
 
diff --git a/drivers/gpu/drm/amd/display/include/dal_asic_id.h b/drivers/gpu/drm/amd/display/include/dal_asic_id.h
index 1b987b6..77d2856 100644
--- a/drivers/gpu/drm/amd/display/include/dal_asic_id.h
+++ b/drivers/gpu/drm/amd/display/include/dal_asic_id.h
@@ -117,6 +117,12 @@
 	((rev >= STONEY_A0) && (rev < CZ_UNKNOWN))
 
 /* DCE12 */
+#define AI_UNKNOWN 0xFF
+
+#ifdef CONFIG_DRM_AMD_DC_VG20
+#define AI_VEGA20_P_A0 40
+#define ASICREV_IS_VEGA20_P(eChipRev) ((eChipRev >= AI_VEGA20_P_A0) && (eChipRev < AI_UNKNOWN))
+#endif
 
 #define AI_GREENLAND_P_A0 1
 #define AI_GREENLAND_P_A1 2
-- 
cgit v1.1


From 1edb2c8a32160c00273485efea8d18080e31cc09 Mon Sep 17 00:00:00 2001
From: "Jerry (Fangzhi) Zuo" <Jerry.Zuo@amd.com>
Date: Fri, 11 May 2018 13:51:43 -0500
Subject: drm/amd/display: Add BIOS smu_info v3_3 support for Vega20
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Jerry (Fangzhi) Zuo <Jerry.Zuo@amd.com>
Reviewed-by: Hersen Wu <hersenxs.wu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
---
 drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
index 10a5807..4561673 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
@@ -1330,6 +1330,11 @@ static enum bp_result bios_parser_get_firmware_info(
 			case 2:
 				result = get_firmware_info_v3_2(bp, info);
 				break;
+			case 3:
+#ifdef CONFIG_DRM_AMD_DC_VG20
+				result = get_firmware_info_v3_2(bp, info);
+#endif
+				break;
 			default:
 				break;
 			}
-- 
cgit v1.1


From 8ad63122f9f22dde172b98fe9c75818831e57f4b Mon Sep 17 00:00:00 2001
From: "Jerry (Fangzhi) Zuo" <Jerry.Zuo@amd.com>
Date: Mon, 5 Mar 2018 16:12:23 -0500
Subject: drm/amd/display: Add harvest IP support for Vega20
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Retrieve fuses to determine the availability of pipes, and
eliminate pipes that cannot be used.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Jerry (Fangzhi) Zuo <Jerry.Zuo@amd.com>
Reviewed-by: Hersen Wu <hersenxs.wu@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
---
 .../drm/amd/display/dc/dce120/dce120_resource.c    | 208 +++++++++++++++++++++
 1 file changed, 208 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c b/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c
index fda0157..545f35f 100644
--- a/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c
@@ -814,6 +814,213 @@ static void bw_calcs_data_update_from_pplib(struct dc *dc)
 	dm_pp_notify_wm_clock_changes(dc->ctx, &clk_ranges);
 }
 
+#ifdef CONFIG_DRM_AMD_DC_VG20
+static uint32_t read_pipe_fuses(struct dc_context *ctx)
+{
+	uint32_t value = dm_read_reg_soc15(ctx, mmCC_DC_PIPE_DIS, 0);
+	/* VG20 support max 6 pipes */
+	value = value & 0x3f;
+	return value;
+}
+
+static bool construct(
+	uint8_t num_virtual_links,
+	struct dc *dc,
+	struct dce110_resource_pool *pool)
+{
+	unsigned int i;
+	int j;
+	struct dc_context *ctx = dc->ctx;
+	struct irq_service_init_data irq_init_data;
+	bool harvest_enabled = ASICREV_IS_VEGA20_P(ctx->asic_id.hw_internal_rev);
+	uint32_t pipe_fuses;
+
+	ctx->dc_bios->regs = &bios_regs;
+
+	pool->base.res_cap = &res_cap;
+	pool->base.funcs = &dce120_res_pool_funcs;
+
+	/* TODO: Fill more data from GreenlandAsicCapability.cpp */
+	pool->base.pipe_count = res_cap.num_timing_generator;
+	pool->base.timing_generator_count = pool->base.res_cap->num_timing_generator;
+	pool->base.underlay_pipe_index = NO_UNDERLAY_PIPE;
+
+	dc->caps.max_downscale_ratio = 200;
+	dc->caps.i2c_speed_in_khz = 100;
+	dc->caps.max_cursor_size = 128;
+	dc->caps.dual_link_dvi = true;
+
+	dc->debug = debug_defaults;
+
+	/*************************************************
+	 *  Create resources                             *
+	 *************************************************/
+
+	pool->base.clock_sources[DCE120_CLK_SRC_PLL0] =
+			dce120_clock_source_create(ctx, ctx->dc_bios,
+				CLOCK_SOURCE_COMBO_PHY_PLL0,
+				&clk_src_regs[0], false);
+	pool->base.clock_sources[DCE120_CLK_SRC_PLL1] =
+			dce120_clock_source_create(ctx, ctx->dc_bios,
+				CLOCK_SOURCE_COMBO_PHY_PLL1,
+				&clk_src_regs[1], false);
+	pool->base.clock_sources[DCE120_CLK_SRC_PLL2] =
+			dce120_clock_source_create(ctx, ctx->dc_bios,
+				CLOCK_SOURCE_COMBO_PHY_PLL2,
+				&clk_src_regs[2], false);
+	pool->base.clock_sources[DCE120_CLK_SRC_PLL3] =
+			dce120_clock_source_create(ctx, ctx->dc_bios,
+				CLOCK_SOURCE_COMBO_PHY_PLL3,
+				&clk_src_regs[3], false);
+	pool->base.clock_sources[DCE120_CLK_SRC_PLL4] =
+			dce120_clock_source_create(ctx, ctx->dc_bios,
+				CLOCK_SOURCE_COMBO_PHY_PLL4,
+				&clk_src_regs[4], false);
+	pool->base.clock_sources[DCE120_CLK_SRC_PLL5] =
+			dce120_clock_source_create(ctx, ctx->dc_bios,
+				CLOCK_SOURCE_COMBO_PHY_PLL5,
+				&clk_src_regs[5], false);
+	pool->base.clk_src_count = DCE120_CLK_SRC_TOTAL;
+
+	pool->base.dp_clock_source =
+			dce120_clock_source_create(ctx, ctx->dc_bios,
+				CLOCK_SOURCE_ID_DP_DTO,
+				&clk_src_regs[0], true);
+
+	for (i = 0; i < pool->base.clk_src_count; i++) {
+		if (pool->base.clock_sources[i] == NULL) {
+			dm_error("DC: failed to create clock sources!\n");
+			BREAK_TO_DEBUGGER();
+			goto clk_src_create_fail;
+		}
+	}
+
+	pool->base.display_clock = dce120_disp_clk_create(ctx);
+	if (pool->base.display_clock == NULL) {
+		dm_error("DC: failed to create display clock!\n");
+		BREAK_TO_DEBUGGER();
+		goto disp_clk_create_fail;
+	}
+
+	pool->base.dmcu = dce_dmcu_create(ctx,
+			&dmcu_regs,
+			&dmcu_shift,
+			&dmcu_mask);
+	if (pool->base.dmcu == NULL) {
+		dm_error("DC: failed to create dmcu!\n");
+		BREAK_TO_DEBUGGER();
+		goto res_create_fail;
+	}
+
+	pool->base.abm = dce_abm_create(ctx,
+			&abm_regs,
+			&abm_shift,
+			&abm_mask);
+	if (pool->base.abm == NULL) {
+		dm_error("DC: failed to create abm!\n");
+		BREAK_TO_DEBUGGER();
+		goto res_create_fail;
+	}
+
+	irq_init_data.ctx = dc->ctx;
+	pool->base.irqs = dal_irq_service_dce120_create(&irq_init_data);
+	if (!pool->base.irqs)
+		goto irqs_create_fail;
+
+	/* retrieve valid pipe fuses */
+	if (harvest_enabled)
+		pipe_fuses = read_pipe_fuses(ctx);
+
+	/* index to valid pipe resource */
+	j = 0;
+	for (i = 0; i < pool->base.pipe_count; i++) {
+		if (harvest_enabled) {
+			if ((pipe_fuses & (1 << i)) != 0) {
+				dm_error("DC: skip invalid pipe %d!\n", i);
+				continue;
+			}
+		}
+
+		pool->base.timing_generators[j] =
+				dce120_timing_generator_create(
+					ctx,
+					i,
+					&dce120_tg_offsets[i]);
+		if (pool->base.timing_generators[j] == NULL) {
+			BREAK_TO_DEBUGGER();
+			dm_error("DC: failed to create tg!\n");
+			goto controller_create_fail;
+		}
+
+		pool->base.mis[j] = dce120_mem_input_create(ctx, i);
+
+		if (pool->base.mis[j] == NULL) {
+			BREAK_TO_DEBUGGER();
+			dm_error(
+				"DC: failed to create memory input!\n");
+			goto controller_create_fail;
+		}
+
+		pool->base.ipps[j] = dce120_ipp_create(ctx, i);
+		if (pool->base.ipps[i] == NULL) {
+			BREAK_TO_DEBUGGER();
+			dm_error(
+				"DC: failed to create input pixel processor!\n");
+			goto controller_create_fail;
+		}
+
+		pool->base.transforms[j] = dce120_transform_create(ctx, i);
+		if (pool->base.transforms[i] == NULL) {
+			BREAK_TO_DEBUGGER();
+			dm_error(
+				"DC: failed to create transform!\n");
+			goto res_create_fail;
+		}
+
+		pool->base.opps[j] = dce120_opp_create(
+			ctx,
+			i);
+		if (pool->base.opps[j] == NULL) {
+			BREAK_TO_DEBUGGER();
+			dm_error(
+				"DC: failed to create output pixel processor!\n");
+		}
+
+		/* check next valid pipe */
+		j++;
+	}
+
+	/* valid pipe num */
+	pool->base.pipe_count = j;
+	pool->base.timing_generator_count = j;
+
+	if (!resource_construct(num_virtual_links, dc, &pool->base,
+			 &res_create_funcs))
+		goto res_create_fail;
+
+	/* Create hardware sequencer */
+	if (!dce120_hw_sequencer_create(dc))
+		goto controller_create_fail;
+
+	dc->caps.max_planes =  pool->base.pipe_count;
+
+	bw_calcs_init(dc->bw_dceip, dc->bw_vbios, dc->ctx->asic_id);
+
+	bw_calcs_data_update_from_pplib(dc);
+
+	return true;
+
+irqs_create_fail:
+controller_create_fail:
+disp_clk_create_fail:
+clk_src_create_fail:
+res_create_fail:
+
+	destruct(pool);
+
+	return false;
+}
+#else
 static bool construct(
 	uint8_t num_virtual_links,
 	struct dc *dc,
@@ -988,6 +1195,7 @@ res_create_fail:
 
 	return false;
 }
+#endif
 
 struct resource_pool *dce120_create_resource_pool(
 	uint8_t num_virtual_links,
-- 
cgit v1.1


From 6f68711dd63522aab34c3e9513fa42a7586a95e5 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Thu, 15 Mar 2018 21:32:27 -0500
Subject: drm/amdgpu/atomfirmware: add new gfx_info data table v2.4 (v2)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Adds additional gfx configuration data.

v2: fix typo

Acked-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/include/atomfirmware.h | 34 ++++++++++++++++++++++++++++++
 1 file changed, 34 insertions(+)

diff --git a/drivers/gpu/drm/amd/include/atomfirmware.h b/drivers/gpu/drm/amd/include/atomfirmware.h
index de177ce..fd5e80c 100644
--- a/drivers/gpu/drm/amd/include/atomfirmware.h
+++ b/drivers/gpu/drm/amd/include/atomfirmware.h
@@ -1219,6 +1219,40 @@ struct  atom_gfx_info_v2_3 {
   uint32_t rm21_sram_vmin_value;
 };
 
+struct  atom_gfx_info_v2_4 {
+  struct  atom_common_table_header  table_header;
+  uint8_t gfxip_min_ver;
+  uint8_t gfxip_max_ver;
+  uint8_t gc_num_se;
+  uint8_t max_tile_pipes;
+  uint8_t gc_num_cu_per_sh;
+  uint8_t gc_num_sh_per_se;
+  uint8_t gc_num_rb_per_se;
+  uint8_t gc_num_tccs;
+  uint32_t regaddr_cp_dma_src_addr;
+  uint32_t regaddr_cp_dma_src_addr_hi;
+  uint32_t regaddr_cp_dma_dst_addr;
+  uint32_t regaddr_cp_dma_dst_addr_hi;
+  uint32_t regaddr_cp_dma_command;
+  uint32_t regaddr_cp_status;
+  uint32_t regaddr_rlc_gpu_clock_32;
+  uint32_t rlc_gpu_timer_refclk;
+  uint8_t active_cu_per_sh;
+  uint8_t active_rb_per_se;
+  uint16_t gcgoldenoffset;
+  uint32_t rm21_sram_vmin_value;
+  uint16_t gc_num_gprs;
+  uint16_t gc_gsprim_buff_depth;
+  uint16_t gc_parameter_cache_depth;
+  uint16_t gc_wave_size;
+  uint16_t gc_max_waves_per_simd;
+  uint16_t gc_lds_size;
+  uint8_t gc_num_max_gs_thds;
+  uint8_t gc_gs_table_depth;
+  uint8_t gc_double_offchip_lds_buffer;
+  uint8_t gc_max_scratch_slots_per_cu;
+};
+
 /* 
   ***************************************************************************
     Data Table smu_info  structure
-- 
cgit v1.1


From 59b0b509f1ae0c7ca54607f2770a1aec6e55d8dc Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Tue, 20 Mar 2018 12:24:03 -0500
Subject: drm/amdgpu/atomfirmware: add parser for gfx_info table
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Add support for the gfx_info table on boards that use atomfirmware.

Acked-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c | 46 ++++++++++++++++++++++++
 drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h |  1 +
 2 files changed, 47 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c
index a0f48cb..7014d58 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c
@@ -322,3 +322,49 @@ int amdgpu_atomfirmware_get_clock_info(struct amdgpu_device *adev)
 
 	return ret;
 }
+
+union gfx_info {
+	struct  atom_gfx_info_v2_4 v24;
+};
+
+int amdgpu_atomfirmware_get_gfx_info(struct amdgpu_device *adev)
+{
+	struct amdgpu_mode_info *mode_info = &adev->mode_info;
+	int index;
+	uint8_t frev, crev;
+	uint16_t data_offset;
+
+	index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1,
+					    gfx_info);
+	if (amdgpu_atom_parse_data_header(mode_info->atom_context, index, NULL,
+				   &frev, &crev, &data_offset)) {
+		union gfx_info *gfx_info = (union gfx_info *)
+			(mode_info->atom_context->bios + data_offset);
+		switch (crev) {
+		case 4:
+			adev->gfx.config.max_shader_engines = gfx_info->v24.gc_num_se;
+			adev->gfx.config.max_cu_per_sh = gfx_info->v24.gc_num_cu_per_sh;
+			adev->gfx.config.max_sh_per_se = gfx_info->v24.gc_num_sh_per_se;
+			adev->gfx.config.max_backends_per_se = gfx_info->v24.gc_num_rb_per_se;
+			adev->gfx.config.max_texture_channel_caches = gfx_info->v24.gc_num_tccs;
+			adev->gfx.config.max_gprs = le16_to_cpu(gfx_info->v24.gc_num_gprs);
+			adev->gfx.config.max_gs_threads = gfx_info->v24.gc_num_max_gs_thds;
+			adev->gfx.config.gs_vgt_table_depth = gfx_info->v24.gc_gs_table_depth;
+			adev->gfx.config.gs_prim_buffer_depth =
+				le16_to_cpu(gfx_info->v24.gc_gsprim_buff_depth);
+			adev->gfx.config.double_offchip_lds_buf =
+				gfx_info->v24.gc_double_offchip_lds_buffer;
+			adev->gfx.cu_info.wave_front_size = gfx_info->v24.gc_wave_size;
+			adev->gfx.cu_info.max_waves_per_simd =
+				le16_to_cpu(gfx_info->v24.gc_max_waves_per_simd);
+			adev->gfx.cu_info.max_scratch_slots_per_cu =
+				gfx_info->v24.gc_max_scratch_slots_per_cu;
+			adev->gfx.cu_info.lds_size = le16_to_cpu(gfx_info->v24.gc_lds_size);
+			return 0;
+		default:
+			return -EINVAL;
+		}
+
+	}
+	return -EINVAL;
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h
index 7689c96..20f158f 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h
@@ -30,5 +30,6 @@ int amdgpu_atomfirmware_allocate_fb_scratch(struct amdgpu_device *adev);
 int amdgpu_atomfirmware_get_vram_width(struct amdgpu_device *adev);
 int amdgpu_atomfirmware_get_vram_type(struct amdgpu_device *adev);
 int amdgpu_atomfirmware_get_clock_info(struct amdgpu_device *adev);
+int amdgpu_atomfirmware_get_gfx_info(struct amdgpu_device *adev);
 
 #endif
-- 
cgit v1.1


From 3251c0438a1efcc51c357f7014b33b9e02b129cd Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Fri, 20 Apr 2018 12:31:04 +0800
Subject: drm/amdgpu: Use vbios table for gpu info on vega20

Use the vbios table rather than gpu info firmware.

Squash of the following patches:
drm/amdgpu/vg20: fallback to vbios table if gpu info fw is not available (v2)
drm/amdgpu: drop gpu_info firmware for vega20

Reviewed-by: Amber Lin <Amber.Lin@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 14 ++++++++++++--
 1 file changed, 12 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index 13253e0..d7530fd 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -27,6 +27,7 @@
 #include "amdgpu_gfx.h"
 #include "soc15.h"
 #include "soc15d.h"
+#include "amdgpu_atomfirmware.h"
 
 #include "gc/gc_9_0_offset.h"
 #include "gc/gc_9_0_sh_mask.h"
@@ -1113,9 +1114,10 @@ static const struct amdgpu_gfx_funcs gfx_v9_0_gfx_funcs = {
 	.select_me_pipe_q = &gfx_v9_0_select_me_pipe_q
 };
 
-static void gfx_v9_0_gpu_early_init(struct amdgpu_device *adev)
+static int gfx_v9_0_gpu_early_init(struct amdgpu_device *adev)
 {
 	u32 gb_addr_config;
+	int err;
 
 	adev->gfx.funcs = &gfx_v9_0_gfx_funcs;
 
@@ -1146,6 +1148,10 @@ static void gfx_v9_0_gpu_early_init(struct amdgpu_device *adev)
 		gb_addr_config = RREG32_SOC15(GC, 0, mmGB_ADDR_CONFIG);
 		gb_addr_config &= ~0xf3e777ff;
 		gb_addr_config |= 0x22014042;
+		/* check vbios table if gpu info is not available */
+		err = amdgpu_atomfirmware_get_gfx_info(adev);
+		if (err)
+			return err;
 		break;
 	case CHIP_RAVEN:
 		adev->gfx.config.max_hw_contexts = 8;
@@ -1196,6 +1202,8 @@ static void gfx_v9_0_gpu_early_init(struct amdgpu_device *adev)
 					adev->gfx.config.gb_addr_config,
 					GB_ADDR_CONFIG,
 					PIPE_INTERLEAVE_SIZE));
+
+	return 0;
 }
 
 static int gfx_v9_0_ngg_create_buf(struct amdgpu_device *adev,
@@ -1557,7 +1565,9 @@ static int gfx_v9_0_sw_init(void *handle)
 
 	adev->gfx.ce_ram_size = 0x8000;
 
-	gfx_v9_0_gpu_early_init(adev);
+	r = gfx_v9_0_gpu_early_init(adev);
+	if (r)
+		return r;
 
 	r = gfx_v9_0_ngg_init(adev);
 	if (r)
-- 
cgit v1.1


From 24e6bc784363ee4056d81c8990a0127891678b43 Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Tue, 24 Apr 2018 11:11:16 +0800
Subject: drm/amdgpu: Set vega20 load_type to AMDGPU_FW_LOAD_DIRECT.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Please revert this patch when psp load fw is enabled.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
index b419d6e..f55f72a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
@@ -303,11 +303,12 @@ amdgpu_ucode_get_load_type(struct amdgpu_device *adev, int load_type)
 	case CHIP_VEGA10:
 	case CHIP_RAVEN:
 	case CHIP_VEGA12:
-	case CHIP_VEGA20:
 		if (!load_type)
 			return AMDGPU_FW_LOAD_DIRECT;
 		else
 			return AMDGPU_FW_LOAD_PSP;
+	case CHIP_VEGA20:
+		return AMDGPU_FW_LOAD_DIRECT;
 	default:
 		DRM_ERROR("Unknown firmware load type\n");
 	}
-- 
cgit v1.1


From 27db6a0073f162cdb15975c9d29d159d772b1ec0 Mon Sep 17 00:00:00 2001
From: Dmitry Osipenko <digetx@gmail.com>
Date: Mon, 23 Apr 2018 12:54:56 +0300
Subject: gpu: host1x: Fix dma_free_wc() argument in the error path

If IOVA allocation or IOMMU mapping fails, dma_free_wc() is invoked with
size=0 because of a typo, that triggers "kernel BUG at mm/vmalloc.c:124!".

Signed-off-by: Dmitry Osipenko <digetx@gmail.com>
Reviewed-by: Mikko Perttunen <mperttunen@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/host1x/cdma.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/host1x/cdma.c b/drivers/gpu/host1x/cdma.c
index 28541b2..cf6caa9 100644
--- a/drivers/gpu/host1x/cdma.c
+++ b/drivers/gpu/host1x/cdma.c
@@ -127,7 +127,7 @@ static int host1x_pushbuffer_init(struct push_buffer *pb)
 iommu_free_iova:
 	__free_iova(&host1x->iova, alloc);
 iommu_free_mem:
-	dma_free_wc(host1x->dev, pb->alloc_size, pb->mapped, pb->phys);
+	dma_free_wc(host1x->dev, size, pb->mapped, pb->phys);
 
 	return err;
 }
-- 
cgit v1.1


From 5f43ac8d80e4c768380e86d312a591472d080eeb Mon Sep 17 00:00:00 2001
From: Thierry Reding <treding@nvidia.com>
Date: Mon, 23 Apr 2018 08:57:44 +0200
Subject: drm/tegra: Fix order of teardown in IOMMU case

The original code works fine, this is merely a cosmetic change to make
the teardown order the reverse of the setup order.

Reviewed-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/drm.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/tegra/drm.c b/drivers/gpu/drm/tegra/drm.c
index 181e82c..7b9f73b 100644
--- a/drivers/gpu/drm/tegra/drm.c
+++ b/drivers/gpu/drm/tegra/drm.c
@@ -204,10 +204,10 @@ config:
 	drm_mode_config_cleanup(drm);
 
 	if (tegra->domain) {
-		iommu_domain_free(tegra->domain);
-		drm_mm_takedown(&tegra->mm);
 		mutex_destroy(&tegra->mm_lock);
+		drm_mm_takedown(&tegra->mm);
 		put_iova_domain(&tegra->carveout.domain);
+		iommu_domain_free(tegra->domain);
 	}
 free:
 	kfree(tegra);
@@ -230,10 +230,10 @@ static void tegra_drm_unload(struct drm_device *drm)
 		return;
 
 	if (tegra->domain) {
-		iommu_domain_free(tegra->domain);
-		drm_mm_takedown(&tegra->mm);
 		mutex_destroy(&tegra->mm_lock);
+		drm_mm_takedown(&tegra->mm);
 		put_iova_domain(&tegra->carveout.domain);
+		iommu_domain_free(tegra->domain);
 	}
 
 	kfree(tegra);
-- 
cgit v1.1


From 24cfdc1ac7d4260aa8416505b9cb6316c9e89021 Mon Sep 17 00:00:00 2001
From: Thierry Reding <treding@nvidia.com>
Date: Mon, 23 Apr 2018 08:57:45 +0200
Subject: drm/tegra: Acquire a reference to the IOVA cache

The IOVA API uses a memory cache to allocate IOVA nodes from. To make
sure that this cache is available, obtain a reference to it and release
the reference when the cache is no longer needed.

On 64-bit ARM this is hidden by the fact that the DMA mapping API gets
that reference and never releases it. On 32-bit ARM, however, the DMA
mapping API doesn't do that, so allocation of IOVA nodes fails.

Fixes: ad92601521ea ("drm/tegra: Add Tegra DRM allocation API")
Reviewed-by: Dmitry Osipenko <digetx@gmail.com>
Tested-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/drm.c | 10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/tegra/drm.c b/drivers/gpu/drm/tegra/drm.c
index 7b9f73b..3cdef65 100644
--- a/drivers/gpu/drm/tegra/drm.c
+++ b/drivers/gpu/drm/tegra/drm.c
@@ -113,6 +113,10 @@ static int tegra_drm_load(struct drm_device *drm, unsigned long flags)
 			goto free;
 		}
 
+		err = iova_cache_get();
+		if (err < 0)
+			goto domain;
+
 		geometry = &tegra->domain->geometry;
 		gem_start = geometry->aperture_start;
 		gem_end = geometry->aperture_end - CARVEOUT_SZ;
@@ -207,8 +211,11 @@ config:
 		mutex_destroy(&tegra->mm_lock);
 		drm_mm_takedown(&tegra->mm);
 		put_iova_domain(&tegra->carveout.domain);
-		iommu_domain_free(tegra->domain);
+		iova_cache_put();
 	}
+domain:
+	if (tegra->domain)
+		iommu_domain_free(tegra->domain);
 free:
 	kfree(tegra);
 	return err;
@@ -233,6 +240,7 @@ static void tegra_drm_unload(struct drm_device *drm)
 		mutex_destroy(&tegra->mm_lock);
 		drm_mm_takedown(&tegra->mm);
 		put_iova_domain(&tegra->carveout.domain);
+		iova_cache_put();
 		iommu_domain_free(tegra->domain);
 	}
 
-- 
cgit v1.1


From f40e1590c5270e5559fb95a5a0a7c1f5266a522d Mon Sep 17 00:00:00 2001
From: Thierry Reding <treding@nvidia.com>
Date: Mon, 14 May 2018 11:14:00 +0200
Subject: gpu: host1x: Acquire a reference to the IOVA cache

The IOVA API uses a memory cache to allocate IOVA nodes from. To make
sure that this cache is available, obtain a reference to it and release
the reference when the cache is no longer needed.

On 64-bit ARM this is hidden by the fact that the DMA mapping API gets
that reference and never releases it. On 32-bit ARM, this is papered
over by the Tegra DRM driver (the sole user of the host1x API requiring
the cache) acquiring a reference to the IOVA cache for its own purposes.
However, there may be additional users of this API in the future, so fix
this upfront to avoid surprises.

Fixes: 404bfb78daf3 ("gpu: host1x: Add IOMMU support")
Reviewed-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/host1x/dev.c | 11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/host1x/dev.c b/drivers/gpu/host1x/dev.c
index 03db711..f1d5f76 100644
--- a/drivers/gpu/host1x/dev.c
+++ b/drivers/gpu/host1x/dev.c
@@ -223,10 +223,14 @@ static int host1x_probe(struct platform_device *pdev)
 		struct iommu_domain_geometry *geometry;
 		unsigned long order;
 
+		err = iova_cache_get();
+		if (err < 0)
+			goto put_group;
+
 		host->domain = iommu_domain_alloc(&platform_bus_type);
 		if (!host->domain) {
 			err = -ENOMEM;
-			goto put_group;
+			goto put_cache;
 		}
 
 		err = iommu_attach_group(host->domain, host->group);
@@ -234,6 +238,7 @@ static int host1x_probe(struct platform_device *pdev)
 			if (err == -ENODEV) {
 				iommu_domain_free(host->domain);
 				host->domain = NULL;
+				iova_cache_put();
 				iommu_group_put(host->group);
 				host->group = NULL;
 				goto skip_iommu;
@@ -308,6 +313,9 @@ fail_detach_device:
 fail_free_domain:
 	if (host->domain)
 		iommu_domain_free(host->domain);
+put_cache:
+	if (host->group)
+		iova_cache_put();
 put_group:
 	iommu_group_put(host->group);
 
@@ -328,6 +336,7 @@ static int host1x_remove(struct platform_device *pdev)
 		put_iova_domain(&host->iova);
 		iommu_detach_group(host->domain, host->group);
 		iommu_domain_free(host->domain);
+		iova_cache_put();
 		iommu_group_put(host->group);
 	}
 
-- 
cgit v1.1


From cc7add70cad12054e096b034578827d7065f64bb Mon Sep 17 00:00:00 2001
From: Souptick Joarder <jrdr.linux@gmail.com>
Date: Tue, 17 Apr 2018 19:17:55 +0530
Subject: drm/tegra: Adding new typedef vm_fault_t

Use new return type vm_fault_t for fault handler. For now, this is just
documenting that the function returns a VM_FAULT value rather than an
errno. Once all instances are converted, vm_fault_t will become a
distinct type.

Reference id -> 1c8f422059ae ("mm: change return type to vm_fault_t")

Previously vm_insert_page() returns err which driver mapped into
VM_FAULT_* type. The new function vmf_insert_page() will replace this
inefficiency by returning VM_FAULT_* type.

Signed-off-by: Souptick Joarder <jrdr.linux@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/gem.c | 18 ++----------------
 1 file changed, 2 insertions(+), 16 deletions(-)

diff --git a/drivers/gpu/drm/tegra/gem.c b/drivers/gpu/drm/tegra/gem.c
index 8b0b4ff..1c40117 100644
--- a/drivers/gpu/drm/tegra/gem.c
+++ b/drivers/gpu/drm/tegra/gem.c
@@ -422,14 +422,13 @@ int tegra_bo_dumb_create(struct drm_file *file, struct drm_device *drm,
 	return 0;
 }
 
-static int tegra_bo_fault(struct vm_fault *vmf)
+static vm_fault_t tegra_bo_fault(struct vm_fault *vmf)
 {
 	struct vm_area_struct *vma = vmf->vma;
 	struct drm_gem_object *gem = vma->vm_private_data;
 	struct tegra_bo *bo = to_tegra_bo(gem);
 	struct page *page;
 	pgoff_t offset;
-	int err;
 
 	if (!bo->pages)
 		return VM_FAULT_SIGBUS;
@@ -437,20 +436,7 @@ static int tegra_bo_fault(struct vm_fault *vmf)
 	offset = (vmf->address - vma->vm_start) >> PAGE_SHIFT;
 	page = bo->pages[offset];
 
-	err = vm_insert_page(vma, vmf->address, page);
-	switch (err) {
-	case -EAGAIN:
-	case 0:
-	case -ERESTARTSYS:
-	case -EINTR:
-	case -EBUSY:
-		return VM_FAULT_NOPAGE;
-
-	case -ENOMEM:
-		return VM_FAULT_OOM;
-	}
-
-	return VM_FAULT_SIGBUS;
+	return vmf_insert_page(vma, vmf->address, page);
 }
 
 const struct vm_operations_struct tegra_bo_vm_ops = {
-- 
cgit v1.1


From e1189921b5ff9dcfec52b21cf12bb52c5dccd34d Mon Sep 17 00:00:00 2001
From: Daniel Stone <daniels@collabora.com>
Date: Fri, 30 Mar 2018 15:11:25 +0100
Subject: drm/tegra: Remove duplicate framebuffer num_planes

drm_framebuffer already stores num_planes for us.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Cc: Thierry Reding <thierry.reding@gmail.com>
Cc: linux-tegra@vger.kernel.org
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/drm.h | 1 -
 drivers/gpu/drm/tegra/fb.c  | 6 ++----
 2 files changed, 2 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/tegra/drm.h b/drivers/gpu/drm/tegra/drm.h
index fe263cf..61a4657 100644
--- a/drivers/gpu/drm/tegra/drm.h
+++ b/drivers/gpu/drm/tegra/drm.h
@@ -32,7 +32,6 @@ struct reset_control;
 struct tegra_fb {
 	struct drm_framebuffer base;
 	struct tegra_bo **planes;
-	unsigned int num_planes;
 };
 
 #ifdef CONFIG_DRM_FBDEV_EMULATION
diff --git a/drivers/gpu/drm/tegra/fb.c b/drivers/gpu/drm/tegra/fb.c
index e694349..75badf3 100644
--- a/drivers/gpu/drm/tegra/fb.c
+++ b/drivers/gpu/drm/tegra/fb.c
@@ -107,7 +107,7 @@ static void tegra_fb_destroy(struct drm_framebuffer *framebuffer)
 	struct tegra_fb *fb = to_tegra_fb(framebuffer);
 	unsigned int i;
 
-	for (i = 0; i < fb->num_planes; i++) {
+	for (i = 0; i < framebuffer->format->num_planes; i++) {
 		struct tegra_bo *bo = fb->planes[i];
 
 		if (bo) {
@@ -155,11 +155,9 @@ static struct tegra_fb *tegra_fb_alloc(struct drm_device *drm,
 		return ERR_PTR(-ENOMEM);
 	}
 
-	fb->num_planes = num_planes;
-
 	drm_helper_mode_fill_fb_struct(drm, &fb->base, mode_cmd);
 
-	for (i = 0; i < fb->num_planes; i++)
+	for (i = 0; i < fb->base.format->num_planes; i++)
 		fb->planes[i] = planes[i];
 
 	err = drm_framebuffer_init(drm, &fb->base, &tegra_fb_funcs);
-- 
cgit v1.1


From 0bc6af006f5ce7fb92d41dc8e01b621bd8d2226b Mon Sep 17 00:00:00 2001
From: Daniel Stone <daniels@collabora.com>
Date: Fri, 30 Mar 2018 15:11:26 +0100
Subject: drm/tegra: Move GEM BOs to drm_framebuffer

Since drm_framebuffer can now store GEM objects directly, place them
there rather than in our own subclass. As this makes the framebuffer
create_handle function the same as the GEM framebuffer helper, we
can reuse that.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Cc: Thierry Reding <thierry.reding@gmail.com>
Cc: linux-tegra@vger.kernel.org
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/drm.h |  1 -
 drivers/gpu/drm/tegra/fb.c  | 37 ++++++++-----------------------------
 2 files changed, 8 insertions(+), 30 deletions(-)

diff --git a/drivers/gpu/drm/tegra/drm.h b/drivers/gpu/drm/tegra/drm.h
index 61a4657..7f9810f 100644
--- a/drivers/gpu/drm/tegra/drm.h
+++ b/drivers/gpu/drm/tegra/drm.h
@@ -31,7 +31,6 @@ struct reset_control;
 
 struct tegra_fb {
 	struct drm_framebuffer base;
-	struct tegra_bo **planes;
 };
 
 #ifdef CONFIG_DRM_FBDEV_EMULATION
diff --git a/drivers/gpu/drm/tegra/fb.c b/drivers/gpu/drm/tegra/fb.c
index 75badf3..5bc8f96 100644
--- a/drivers/gpu/drm/tegra/fb.c
+++ b/drivers/gpu/drm/tegra/fb.c
@@ -14,6 +14,7 @@
 
 #include "drm.h"
 #include "gem.h"
+#include <drm/drm_gem_framebuffer_helper.h>
 
 static inline struct tegra_fb *to_tegra_fb(struct drm_framebuffer *fb)
 {
@@ -30,19 +31,14 @@ static inline struct tegra_fbdev *to_tegra_fbdev(struct drm_fb_helper *helper)
 struct tegra_bo *tegra_fb_get_plane(struct drm_framebuffer *framebuffer,
 				    unsigned int index)
 {
-	struct tegra_fb *fb = to_tegra_fb(framebuffer);
-
-	if (index >= framebuffer->format->num_planes)
-		return NULL;
-
-	return fb->planes[index];
+	return to_tegra_bo(drm_gem_fb_get_obj(framebuffer, index));
 }
 
 bool tegra_fb_is_bottom_up(struct drm_framebuffer *framebuffer)
 {
-	struct tegra_fb *fb = to_tegra_fb(framebuffer);
+	struct tegra_bo *bo = tegra_fb_get_plane(framebuffer, 0);
 
-	if (fb->planes[0]->flags & TEGRA_BO_BOTTOM_UP)
+	if (bo->flags & TEGRA_BO_BOTTOM_UP)
 		return true;
 
 	return false;
@@ -51,8 +47,7 @@ bool tegra_fb_is_bottom_up(struct drm_framebuffer *framebuffer)
 int tegra_fb_get_tiling(struct drm_framebuffer *framebuffer,
 			struct tegra_bo_tiling *tiling)
 {
-	struct tegra_fb *fb = to_tegra_fb(framebuffer);
-	uint64_t modifier = fb->base.modifier;
+	uint64_t modifier = framebuffer->modifier;
 
 	switch (modifier) {
 	case DRM_FORMAT_MOD_LINEAR:
@@ -108,7 +103,7 @@ static void tegra_fb_destroy(struct drm_framebuffer *framebuffer)
 	unsigned int i;
 
 	for (i = 0; i < framebuffer->format->num_planes; i++) {
-		struct tegra_bo *bo = fb->planes[i];
+		struct tegra_bo *bo = tegra_fb_get_plane(framebuffer, i);
 
 		if (bo) {
 			if (bo->pages)
@@ -119,21 +114,12 @@ static void tegra_fb_destroy(struct drm_framebuffer *framebuffer)
 	}
 
 	drm_framebuffer_cleanup(framebuffer);
-	kfree(fb->planes);
 	kfree(fb);
 }
 
-static int tegra_fb_create_handle(struct drm_framebuffer *framebuffer,
-				  struct drm_file *file, unsigned int *handle)
-{
-	struct tegra_fb *fb = to_tegra_fb(framebuffer);
-
-	return drm_gem_handle_create(file, &fb->planes[0]->gem, handle);
-}
-
 static const struct drm_framebuffer_funcs tegra_fb_funcs = {
 	.destroy = tegra_fb_destroy,
-	.create_handle = tegra_fb_create_handle,
+	.create_handle = drm_gem_fb_create_handle,
 };
 
 static struct tegra_fb *tegra_fb_alloc(struct drm_device *drm,
@@ -149,22 +135,15 @@ static struct tegra_fb *tegra_fb_alloc(struct drm_device *drm,
 	if (!fb)
 		return ERR_PTR(-ENOMEM);
 
-	fb->planes = kzalloc(num_planes * sizeof(*planes), GFP_KERNEL);
-	if (!fb->planes) {
-		kfree(fb);
-		return ERR_PTR(-ENOMEM);
-	}
-
 	drm_helper_mode_fill_fb_struct(drm, &fb->base, mode_cmd);
 
 	for (i = 0; i < fb->base.format->num_planes; i++)
-		fb->planes[i] = planes[i];
+		fb->base.obj[i] = &planes[i]->gem;
 
 	err = drm_framebuffer_init(drm, &fb->base, &tegra_fb_funcs);
 	if (err < 0) {
 		dev_err(drm->dev, "failed to initialize framebuffer: %d\n",
 			err);
-		kfree(fb->planes);
 		kfree(fb);
 		return ERR_PTR(err);
 	}
-- 
cgit v1.1


From dbc33c7d65536bce447057dc6f882decc515047d Mon Sep 17 00:00:00 2001
From: Daniel Stone <daniels@collabora.com>
Date: Fri, 30 Mar 2018 15:11:27 +0100
Subject: drm/tegra: tegra_fb -> drm_framebuffer

Since tegra_fb is now the same as drm_framebuffer, we can just replace
the type completely.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Cc: Thierry Reding <thierry.reding@gmail.com>
Cc: linux-tegra@vger.kernel.org
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/drm.h |  6 +-----
 drivers/gpu/drm/tegra/fb.c  | 34 ++++++++++++++--------------------
 2 files changed, 15 insertions(+), 25 deletions(-)

diff --git a/drivers/gpu/drm/tegra/drm.h b/drivers/gpu/drm/tegra/drm.h
index 7f9810f..f47a605 100644
--- a/drivers/gpu/drm/tegra/drm.h
+++ b/drivers/gpu/drm/tegra/drm.h
@@ -29,14 +29,10 @@
 
 struct reset_control;
 
-struct tegra_fb {
-	struct drm_framebuffer base;
-};
-
 #ifdef CONFIG_DRM_FBDEV_EMULATION
 struct tegra_fbdev {
 	struct drm_fb_helper base;
-	struct tegra_fb *fb;
+	struct drm_framebuffer *fb;
 };
 #endif
 
diff --git a/drivers/gpu/drm/tegra/fb.c b/drivers/gpu/drm/tegra/fb.c
index 5bc8f96..57da968 100644
--- a/drivers/gpu/drm/tegra/fb.c
+++ b/drivers/gpu/drm/tegra/fb.c
@@ -16,11 +16,6 @@
 #include "gem.h"
 #include <drm/drm_gem_framebuffer_helper.h>
 
-static inline struct tegra_fb *to_tegra_fb(struct drm_framebuffer *fb)
-{
-	return container_of(fb, struct tegra_fb, base);
-}
-
 #ifdef CONFIG_DRM_FBDEV_EMULATION
 static inline struct tegra_fbdev *to_tegra_fbdev(struct drm_fb_helper *helper)
 {
@@ -99,7 +94,6 @@ int tegra_fb_get_tiling(struct drm_framebuffer *framebuffer,
 
 static void tegra_fb_destroy(struct drm_framebuffer *framebuffer)
 {
-	struct tegra_fb *fb = to_tegra_fb(framebuffer);
 	unsigned int i;
 
 	for (i = 0; i < framebuffer->format->num_planes; i++) {
@@ -114,7 +108,7 @@ static void tegra_fb_destroy(struct drm_framebuffer *framebuffer)
 	}
 
 	drm_framebuffer_cleanup(framebuffer);
-	kfree(fb);
+	kfree(framebuffer);
 }
 
 static const struct drm_framebuffer_funcs tegra_fb_funcs = {
@@ -122,12 +116,12 @@ static const struct drm_framebuffer_funcs tegra_fb_funcs = {
 	.create_handle = drm_gem_fb_create_handle,
 };
 
-static struct tegra_fb *tegra_fb_alloc(struct drm_device *drm,
-				       const struct drm_mode_fb_cmd2 *mode_cmd,
-				       struct tegra_bo **planes,
-				       unsigned int num_planes)
+static struct drm_framebuffer *tegra_fb_alloc(struct drm_device *drm,
+					      const struct drm_mode_fb_cmd2 *mode_cmd,
+					      struct tegra_bo **planes,
+					      unsigned int num_planes)
 {
-	struct tegra_fb *fb;
+	struct drm_framebuffer *fb;
 	unsigned int i;
 	int err;
 
@@ -135,12 +129,12 @@ static struct tegra_fb *tegra_fb_alloc(struct drm_device *drm,
 	if (!fb)
 		return ERR_PTR(-ENOMEM);
 
-	drm_helper_mode_fill_fb_struct(drm, &fb->base, mode_cmd);
+	drm_helper_mode_fill_fb_struct(drm, fb, mode_cmd);
 
-	for (i = 0; i < fb->base.format->num_planes; i++)
-		fb->base.obj[i] = &planes[i]->gem;
+	for (i = 0; i < fb->format->num_planes; i++)
+		fb->obj[i] = &planes[i]->gem;
 
-	err = drm_framebuffer_init(drm, &fb->base, &tegra_fb_funcs);
+	err = drm_framebuffer_init(drm, fb, &tegra_fb_funcs);
 	if (err < 0) {
 		dev_err(drm->dev, "failed to initialize framebuffer: %d\n",
 			err);
@@ -158,7 +152,7 @@ struct drm_framebuffer *tegra_fb_create(struct drm_device *drm,
 	unsigned int hsub, vsub, i;
 	struct tegra_bo *planes[4];
 	struct drm_gem_object *gem;
-	struct tegra_fb *fb;
+	struct drm_framebuffer *fb;
 	int err;
 
 	hsub = drm_format_horz_chroma_subsampling(cmd->pixel_format);
@@ -194,7 +188,7 @@ struct drm_framebuffer *tegra_fb_create(struct drm_device *drm,
 		goto unreference;
 	}
 
-	return &fb->base;
+	return fb;
 
 unreference:
 	while (i--)
@@ -275,7 +269,7 @@ static int tegra_fbdev_probe(struct drm_fb_helper *helper,
 		return PTR_ERR(fbdev->fb);
 	}
 
-	fb = &fbdev->fb->base;
+	fb = fbdev->fb;
 	helper->fb = fb;
 	helper->fbdev = info;
 
@@ -376,7 +370,7 @@ static void tegra_fbdev_exit(struct tegra_fbdev *fbdev)
 	drm_fb_helper_unregister_fbi(&fbdev->base);
 
 	if (fbdev->fb)
-		drm_framebuffer_remove(&fbdev->fb->base);
+		drm_framebuffer_remove(fbdev->fb);
 
 	drm_fb_helper_fini(&fbdev->base);
 	tegra_fbdev_free(fbdev);
-- 
cgit v1.1


From c34a997d033df6bbeaf5c06e9124f89bc0ecac8d Mon Sep 17 00:00:00 2001
From: Daniel Stone <daniels@collabora.com>
Date: Fri, 30 Mar 2018 15:11:28 +0100
Subject: drm/tegra: Move fbdev unmap special case

User framebuffers are created with either bo->pages or bo->vaddr set,
depending on whether or not an IOMMU is present. On the other hand, the
framebuffer created for fbdev emulation has a vaddr mapping made if
bo->pages is set after creation. This is set up in fbdev probe.

Remove the special case unmapping from the general-purpose framebuffer
destroy, and move it to fbdev teardown.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Cc: Thierry Reding <thierry.reding@gmail.com>
Cc: linux-tegra@vger.kernel.org
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/fb.c | 17 +++++++++++------
 1 file changed, 11 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/tegra/fb.c b/drivers/gpu/drm/tegra/fb.c
index 57da968..709aa6e 100644
--- a/drivers/gpu/drm/tegra/fb.c
+++ b/drivers/gpu/drm/tegra/fb.c
@@ -99,12 +99,8 @@ static void tegra_fb_destroy(struct drm_framebuffer *framebuffer)
 	for (i = 0; i < framebuffer->format->num_planes; i++) {
 		struct tegra_bo *bo = tegra_fb_get_plane(framebuffer, i);
 
-		if (bo) {
-			if (bo->pages)
-				vunmap(bo->vaddr);
-
+		if (bo)
 			drm_gem_object_put_unlocked(&bo->gem);
-		}
 	}
 
 	drm_framebuffer_cleanup(framebuffer);
@@ -369,8 +365,17 @@ static void tegra_fbdev_exit(struct tegra_fbdev *fbdev)
 {
 	drm_fb_helper_unregister_fbi(&fbdev->base);
 
-	if (fbdev->fb)
+	if (fbdev->fb) {
+		struct tegra_bo *bo = tegra_fb_get_plane(fbdev->fb, 0);
+
+		/* Undo the special mapping we made in fbdev probe. */
+		if (bo && bo->pages) {
+			vunmap(bo->vaddr);
+			bo->vaddr = 0;
+		}
+
 		drm_framebuffer_remove(fbdev->fb);
+	}
 
 	drm_fb_helper_fini(&fbdev->base);
 	tegra_fbdev_free(fbdev);
-- 
cgit v1.1


From 5cb8b9969be6f14ac3b7ba90de8f7a65f68e46fe Mon Sep 17 00:00:00 2001
From: Daniel Stone <daniels@collabora.com>
Date: Fri, 30 Mar 2018 15:11:29 +0100
Subject: drm/tegra: Use drm_gem_fb_destroy

Now that our destroy function is the same as the helper, use that
directly.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Cc: Thierry Reding <thierry.reding@gmail.com>
Cc: linux-tegra@vger.kernel.org
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/fb.c | 17 +----------------
 1 file changed, 1 insertion(+), 16 deletions(-)

diff --git a/drivers/gpu/drm/tegra/fb.c b/drivers/gpu/drm/tegra/fb.c
index 709aa6e..4c22cdd 100644
--- a/drivers/gpu/drm/tegra/fb.c
+++ b/drivers/gpu/drm/tegra/fb.c
@@ -92,23 +92,8 @@ int tegra_fb_get_tiling(struct drm_framebuffer *framebuffer,
 	return 0;
 }
 
-static void tegra_fb_destroy(struct drm_framebuffer *framebuffer)
-{
-	unsigned int i;
-
-	for (i = 0; i < framebuffer->format->num_planes; i++) {
-		struct tegra_bo *bo = tegra_fb_get_plane(framebuffer, i);
-
-		if (bo)
-			drm_gem_object_put_unlocked(&bo->gem);
-	}
-
-	drm_framebuffer_cleanup(framebuffer);
-	kfree(framebuffer);
-}
-
 static const struct drm_framebuffer_funcs tegra_fb_funcs = {
-	.destroy = tegra_fb_destroy,
+	.destroy = drm_gem_fb_destroy,
 	.create_handle = drm_gem_fb_create_handle,
 };
 
-- 
cgit v1.1


From 9dfbd7319909a948146f5c3438f7bd86c2c53cb6 Mon Sep 17 00:00:00 2001
From: Arnd Bergmann <arnd@arndb.de>
Date: Fri, 2 Feb 2018 16:27:31 +0100
Subject: drm/nouveau: nouveau: use larger buffer in nvif_vmm_map

gcc points out a buffer that is clearly too small to be used
in a meaningful way, as the 'sizeof(*args) + argc > sizeof(stack)'
will always fail:

In function 'memcpy',
    inlined from 'nvif_vmm_map' at drivers/gpu/drm/nouveau/nvif/vmm.c:55:2:
include/linux/string.h:353:9: error: '__builtin_memcpy' offset 40 is out of the bounds [0, 16] of object 'stack' with type 'u8[16]' {aka 'unsigned char[16]'} [-Werror=array-bounds]
  return __builtin_memcpy(p, q, size);
         ^~~~~~~~~~~~~~~~~~~~~~~~~~~~
drivers/gpu/drm/nouveau/nvif/vmm.c: In function 'nvif_vmm_map':
drivers/gpu/drm/nouveau/nvif/vmm.c:40:5: note: 'stack' declared here

This makes the buffer large enough so it should serve the purpose
that the author presumably had in mind. Alternatively we could
just get rid of it completely and simplify the code at the cost
of always doing the kmalloc (as we do in the current version).

Fixes: 920d2b5ef215 ("drm/nouveau/mmu: define user interfaces to mmu vmm opertaions")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvif/vmm.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/nouveau/nvif/vmm.c b/drivers/gpu/drm/nouveau/nvif/vmm.c
index 31cdb2d..191832b 100644
--- a/drivers/gpu/drm/nouveau/nvif/vmm.c
+++ b/drivers/gpu/drm/nouveau/nvif/vmm.c
@@ -37,7 +37,7 @@ nvif_vmm_map(struct nvif_vmm *vmm, u64 addr, u64 size, void *argv, u32 argc,
 	     struct nvif_mem *mem, u64 offset)
 {
 	struct nvif_vmm_map_v0 *args;
-	u8 stack[16];
+	u8 stack[48];
 	int ret;
 
 	if (sizeof(*args) + argc > sizeof(stack)) {
-- 
cgit v1.1


From 7bf5b70befd7817b9e42acbd2291b2042ea1bf81 Mon Sep 17 00:00:00 2001
From: "Gustavo A. R. Silva" <gustavo@embeddedor.com>
Date: Tue, 13 Mar 2018 11:24:11 -0500
Subject: drm/nouveau/secboot: remove VLA usage

In preparation to enabling -Wvla, remove VLA. In this particular
case directly use macro NVKM_MSGQUEUE_CMDLINE_SIZE instead of local
variable cmdline_size. Also, remove cmdline_size as it is not
actually useful anymore.

The use of stack Variable Length Arrays needs to be avoided, as they
can be a vector for stack exhaustion, which can be both a runtime bug
or a security flaw. Also, in general, as code evolves it is easy to
lose track of how big a VLA can get. Thus, we can end up having runtime
failures that are hard to debug.

Also, fixed as part of the directive to remove all VLAs from
the kernel: https://lkml.org/lkml/2018/3/7/621

Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Reviewed-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode_msgqueue.c | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode_msgqueue.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode_msgqueue.c
index 6f10b09..1e1f1c6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode_msgqueue.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode_msgqueue.c
@@ -80,12 +80,11 @@ acr_ls_msgqueue_post_run(struct nvkm_msgqueue *queue,
 			 struct nvkm_falcon *falcon, u32 addr_args)
 {
 	struct nvkm_device *device = falcon->owner->device;
-	u32 cmdline_size = NVKM_MSGQUEUE_CMDLINE_SIZE;
-	u8 buf[cmdline_size];
+	u8 buf[NVKM_MSGQUEUE_CMDLINE_SIZE];
 
-	memset(buf, 0, cmdline_size);
+	memset(buf, 0, sizeof(buf));
 	nvkm_msgqueue_write_cmdline(queue, buf);
-	nvkm_falcon_load_dmem(falcon, buf, addr_args, cmdline_size, 0);
+	nvkm_falcon_load_dmem(falcon, buf, addr_args, sizeof(buf), 0);
 	/* rearm the queue so it will wait for the init message */
 	nvkm_msgqueue_reinit(queue);
 
-- 
cgit v1.1


From 1ce466894b532dc26b02a334fc609378b9231ea8 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/core: define FAULT subdev

This will be responsible for the handling of MMU fault buffers on GPUs
that support them.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/include/nvkm/core/device.h  | 3 +++
 drivers/gpu/drm/nouveau/include/nvkm/subdev/fault.h | 8 ++++++++
 drivers/gpu/drm/nouveau/nvkm/core/subdev.c          | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c   | 2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/device/priv.h   | 1 +
 drivers/gpu/drm/nouveau/nvkm/subdev/Kbuild          | 1 +
 drivers/gpu/drm/nouveau/nvkm/subdev/fault/Kbuild    | 0
 7 files changed, 16 insertions(+)
 create mode 100644 drivers/gpu/drm/nouveau/include/nvkm/subdev/fault.h
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/subdev/fault/Kbuild

diff --git a/drivers/gpu/drm/nouveau/include/nvkm/core/device.h b/drivers/gpu/drm/nouveau/include/nvkm/core/device.h
index 560265b..f2f9b9e 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/core/device.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/core/device.h
@@ -22,6 +22,7 @@ enum nvkm_devidx {
 	NVKM_SUBDEV_LTC,
 	NVKM_SUBDEV_MMU,
 	NVKM_SUBDEV_BAR,
+	NVKM_SUBDEV_FAULT,
 	NVKM_SUBDEV_PMU,
 	NVKM_SUBDEV_VOLT,
 	NVKM_SUBDEV_ICCSENSE,
@@ -123,6 +124,7 @@ struct nvkm_device {
 	struct nvkm_bus *bus;
 	struct nvkm_clk *clk;
 	struct nvkm_devinit *devinit;
+	struct nvkm_fault *fault;
 	struct nvkm_fb *fb;
 	struct nvkm_fuse *fuse;
 	struct nvkm_gpio *gpio;
@@ -194,6 +196,7 @@ struct nvkm_device_chip {
 	int (*bus     )(struct nvkm_device *, int idx, struct nvkm_bus **);
 	int (*clk     )(struct nvkm_device *, int idx, struct nvkm_clk **);
 	int (*devinit )(struct nvkm_device *, int idx, struct nvkm_devinit **);
+	int (*fault   )(struct nvkm_device *, int idx, struct nvkm_fault **);
 	int (*fb      )(struct nvkm_device *, int idx, struct nvkm_fb **);
 	int (*fuse    )(struct nvkm_device *, int idx, struct nvkm_fuse **);
 	int (*gpio    )(struct nvkm_device *, int idx, struct nvkm_gpio **);
diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/fault.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/fault.h
new file mode 100644
index 0000000..35a9cfc
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/fault.h
@@ -0,0 +1,8 @@
+#ifndef __NVKM_FAULT_H__
+#define __NVKM_FAULT_H__
+#include <core/subdev.h>
+
+struct nvkm_fault {
+	struct nvkm_subdev subdev;
+};
+#endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/core/subdev.c b/drivers/gpu/drm/nouveau/nvkm/core/subdev.c
index a134d225..c707306 100644
--- a/drivers/gpu/drm/nouveau/nvkm/core/subdev.c
+++ b/drivers/gpu/drm/nouveau/nvkm/core/subdev.c
@@ -35,6 +35,7 @@ nvkm_subdev_name[NVKM_SUBDEV_NR] = {
 	[NVKM_SUBDEV_BUS     ] = "bus",
 	[NVKM_SUBDEV_CLK     ] = "clk",
 	[NVKM_SUBDEV_DEVINIT ] = "devinit",
+	[NVKM_SUBDEV_FAULT   ] = "fault",
 	[NVKM_SUBDEV_FB      ] = "fb",
 	[NVKM_SUBDEV_FUSE    ] = "fuse",
 	[NVKM_SUBDEV_GPIO    ] = "gpio",
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 05cd674..70f3cc0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2420,6 +2420,7 @@ nvkm_device_subdev(struct nvkm_device *device, int index)
 	_(BUS     , device->bus     , &device->bus->subdev);
 	_(CLK     , device->clk     , &device->clk->subdev);
 	_(DEVINIT , device->devinit , &device->devinit->subdev);
+	_(FAULT   , device->fault   , &device->fault->subdev);
 	_(FB      , device->fb      , &device->fb->subdev);
 	_(FUSE    , device->fuse    , &device->fuse->subdev);
 	_(GPIO    , device->gpio    , &device->gpio->subdev);
@@ -2891,6 +2892,7 @@ nvkm_device_ctor(const struct nvkm_device_func *func,
 		_(NVKM_SUBDEV_BUS     ,      bus);
 		_(NVKM_SUBDEV_CLK     ,      clk);
 		_(NVKM_SUBDEV_DEVINIT ,  devinit);
+		_(NVKM_SUBDEV_FAULT   ,    fault);
 		_(NVKM_SUBDEV_FB      ,       fb);
 		_(NVKM_SUBDEV_FUSE    ,     fuse);
 		_(NVKM_SUBDEV_GPIO    ,     gpio);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/priv.h b/drivers/gpu/drm/nouveau/nvkm/engine/device/priv.h
index 08d0bf6..253ab91 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/priv.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/priv.h
@@ -8,6 +8,7 @@
 #include <subdev/bus.h>
 #include <subdev/clk.h>
 #include <subdev/devinit.h>
+#include <subdev/fault.h>
 #include <subdev/fb.h>
 #include <subdev/fuse.h>
 #include <subdev/gpio.h>
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/Kbuild
index 3f5d38d..cfdffef 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/Kbuild
@@ -3,6 +3,7 @@ include $(src)/nvkm/subdev/bios/Kbuild
 include $(src)/nvkm/subdev/bus/Kbuild
 include $(src)/nvkm/subdev/clk/Kbuild
 include $(src)/nvkm/subdev/devinit/Kbuild
+include $(src)/nvkm/subdev/fault/Kbuild
 include $(src)/nvkm/subdev/fb/Kbuild
 include $(src)/nvkm/subdev/fuse/Kbuild
 include $(src)/nvkm/subdev/gpio/Kbuild
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fault/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/Kbuild
new file mode 100644
index 0000000..e69de29
-- 
cgit v1.1


From 2f68234fb3e7d0b123a8166ba46bd5c3e577b270 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/mc/gp100-: route fault buffer interrupts to FAULT

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp100.c | 20 +++++++++++++++++++-
 drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp10b.c |  2 +-
 drivers/gpu/drm/nouveau/nvkm/subdev/mc/priv.h  |  2 ++
 3 files changed, 22 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp100.c
index 7321ad3..43db245 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp100.c
@@ -75,10 +75,28 @@ gp100_mc_intr_mask(struct nvkm_mc *base, u32 mask, u32 intr)
 	spin_unlock_irqrestore(&mc->lock, flags);
 }
 
+const struct nvkm_mc_map
+gp100_mc_intr[] = {
+	{ 0x04000000, NVKM_ENGINE_DISP },
+	{ 0x00000100, NVKM_ENGINE_FIFO },
+	{ 0x00000200, NVKM_SUBDEV_FAULT },
+	{ 0x40000000, NVKM_SUBDEV_IBUS },
+	{ 0x10000000, NVKM_SUBDEV_BUS },
+	{ 0x08000000, NVKM_SUBDEV_FB },
+	{ 0x02000000, NVKM_SUBDEV_LTC },
+	{ 0x01000000, NVKM_SUBDEV_PMU },
+	{ 0x00200000, NVKM_SUBDEV_GPIO },
+	{ 0x00200000, NVKM_SUBDEV_I2C },
+	{ 0x00100000, NVKM_SUBDEV_TIMER },
+	{ 0x00040000, NVKM_SUBDEV_THERM },
+	{ 0x00002000, NVKM_SUBDEV_FB },
+	{},
+};
+
 static const struct nvkm_mc_func
 gp100_mc = {
 	.init = nv50_mc_init,
-	.intr = gk104_mc_intr,
+	.intr = gp100_mc_intr,
 	.intr_unarm = gp100_mc_intr_unarm,
 	.intr_rearm = gp100_mc_intr_rearm,
 	.intr_mask = gp100_mc_intr_mask,
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp10b.c
index 2283e3b..ff8629d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp10b.c
@@ -34,7 +34,7 @@ gp10b_mc_init(struct nvkm_mc *mc)
 static const struct nvkm_mc_func
 gp10b_mc = {
 	.init = gp10b_mc_init,
-	.intr = gk104_mc_intr,
+	.intr = gp100_mc_intr,
 	.intr_unarm = gp100_mc_intr_unarm,
 	.intr_rearm = gp100_mc_intr_rearm,
 	.intr_mask = gp100_mc_intr_mask,
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/priv.h b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/priv.h
index 8869d79c..d9e3691 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/priv.h
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/priv.h
@@ -57,4 +57,6 @@ int gp100_mc_new_(const struct nvkm_mc_func *, struct nvkm_device *, int,
 
 extern const struct nvkm_mc_map gk104_mc_intr[];
 extern const struct nvkm_mc_map gk104_mc_reset[];
+
+extern const struct nvkm_mc_map gp100_mc_intr[];
 #endif
-- 
cgit v1.1


From 36780d7eee827047bd9e736f6e2c0be650f30b4e Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/fault: add infrastructure to support fault buffers

GPU-specific support will be added separately.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 .../gpu/drm/nouveau/include/nvkm/subdev/fault.h    |   6 +
 drivers/gpu/drm/nouveau/nvkm/subdev/fault/Kbuild   |   1 +
 drivers/gpu/drm/nouveau/nvkm/subdev/fault/base.c   | 179 +++++++++++++++++++++
 drivers/gpu/drm/nouveau/nvkm/subdev/fault/priv.h   |  34 ++++
 4 files changed, 220 insertions(+)
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/subdev/fault/base.c
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/subdev/fault/priv.h

diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/fault.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/fault.h
index 35a9cfc..6ba9c17 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/fault.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/fault.h
@@ -3,6 +3,12 @@
 #include <core/subdev.h>
 
 struct nvkm_fault {
+	const struct nvkm_fault_func *func;
 	struct nvkm_subdev subdev;
+
+	struct nvkm_fault_buffer *buffer[1];
+	int buffer_nr;
+
+	struct nvkm_event event;
 };
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fault/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/Kbuild
index e69de29..2e4c226 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/fault/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/Kbuild
@@ -0,0 +1 @@
+nvkm-y += nvkm/subdev/fault/base.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fault/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/base.c
new file mode 100644
index 0000000..007bf4a
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/base.c
@@ -0,0 +1,179 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "priv.h"
+
+#include <core/memory.h>
+#include <core/notify.h>
+#include <subdev/bar.h>
+#include <subdev/mmu.h>
+
+static void
+nvkm_fault_ntfy_fini(struct nvkm_event *event, int type, int index)
+{
+	struct nvkm_fault *fault = container_of(event, typeof(*fault), event);
+	fault->func->buffer.fini(fault->buffer[index]);
+}
+
+static void
+nvkm_fault_ntfy_init(struct nvkm_event *event, int type, int index)
+{
+	struct nvkm_fault *fault = container_of(event, typeof(*fault), event);
+	fault->func->buffer.init(fault->buffer[index]);
+}
+
+static int
+nvkm_fault_ntfy_ctor(struct nvkm_object *object, void *argv, u32 argc,
+		     struct nvkm_notify *notify)
+{
+	struct nvkm_fault_buffer *buffer = nvkm_fault_buffer(object);
+	if (argc == 0) {
+		notify->size  = 0;
+		notify->types = 1;
+		notify->index = buffer->id;
+		return 0;
+	}
+	return -ENOSYS;
+}
+
+static const struct nvkm_event_func
+nvkm_fault_ntfy = {
+	.ctor = nvkm_fault_ntfy_ctor,
+	.init = nvkm_fault_ntfy_init,
+	.fini = nvkm_fault_ntfy_fini,
+};
+
+static void
+nvkm_fault_intr(struct nvkm_subdev *subdev)
+{
+	struct nvkm_fault *fault = nvkm_fault(subdev);
+	return fault->func->intr(fault);
+}
+
+static int
+nvkm_fault_fini(struct nvkm_subdev *subdev, bool suspend)
+{
+	struct nvkm_fault *fault = nvkm_fault(subdev);
+	if (fault->func->fini)
+		fault->func->fini(fault);
+	return 0;
+}
+
+static int
+nvkm_fault_init(struct nvkm_subdev *subdev)
+{
+	struct nvkm_fault *fault = nvkm_fault(subdev);
+	if (fault->func->init)
+		fault->func->init(fault);
+	return 0;
+}
+
+static int
+nvkm_fault_oneinit_buffer(struct nvkm_fault *fault, int id)
+{
+	struct nvkm_subdev *subdev = &fault->subdev;
+	struct nvkm_device *device = subdev->device;
+	struct nvkm_vmm *bar2 = nvkm_bar_bar2_vmm(device);
+	struct nvkm_fault_buffer *buffer;
+	int ret;
+
+	if (!(buffer = kzalloc(sizeof(*buffer), GFP_KERNEL)))
+		return -ENOMEM;
+	buffer->fault = fault;
+	buffer->id = id;
+	buffer->entries = fault->func->buffer.entries(buffer);
+	fault->buffer[id] = buffer;
+
+	nvkm_debug(subdev, "buffer %d: %d entries\n", id, buffer->entries);
+
+	ret = nvkm_memory_new(device, NVKM_MEM_TARGET_INST, buffer->entries *
+			      fault->func->buffer.entry_size, 0x1000, true,
+			      &buffer->mem);
+	if (ret)
+		return ret;
+
+	ret = nvkm_vmm_get(bar2, 12, nvkm_memory_size(buffer->mem),
+			   &buffer->vma);
+	if (ret)
+		return ret;
+
+	return nvkm_memory_map(buffer->mem, 0, bar2, buffer->vma, NULL, 0);
+}
+
+static int
+nvkm_fault_oneinit(struct nvkm_subdev *subdev)
+{
+	struct nvkm_fault *fault = nvkm_fault(subdev);
+	int ret, i;
+
+	for (i = 0; i < ARRAY_SIZE(fault->buffer); i++) {
+		if (i < fault->func->buffer.nr) {
+			ret = nvkm_fault_oneinit_buffer(fault, i);
+			if (ret)
+				return ret;
+			fault->buffer_nr = i + 1;
+		}
+	}
+
+	return nvkm_event_init(&nvkm_fault_ntfy, 1, fault->buffer_nr,
+			       &fault->event);
+}
+
+static void *
+nvkm_fault_dtor(struct nvkm_subdev *subdev)
+{
+	struct nvkm_vmm *bar2 = nvkm_bar_bar2_vmm(subdev->device);
+	struct nvkm_fault *fault = nvkm_fault(subdev);
+	int i;
+
+	nvkm_event_fini(&fault->event);
+
+	for (i = 0; i < fault->buffer_nr; i++) {
+		if (fault->buffer[i]) {
+			nvkm_vmm_put(bar2, &fault->buffer[i]->vma);
+			nvkm_memory_unref(&fault->buffer[i]->mem);
+			kfree(fault->buffer[i]);
+		}
+	}
+
+	return fault;
+}
+
+static const struct nvkm_subdev_func
+nvkm_fault = {
+	.dtor = nvkm_fault_dtor,
+	.oneinit = nvkm_fault_oneinit,
+	.init = nvkm_fault_init,
+	.fini = nvkm_fault_fini,
+	.intr = nvkm_fault_intr,
+};
+
+int
+nvkm_fault_new_(const struct nvkm_fault_func *func, struct nvkm_device *device,
+		int index, struct nvkm_fault **pfault)
+{
+	struct nvkm_fault *fault;
+	if (!(fault = *pfault = kzalloc(sizeof(*fault), GFP_KERNEL)))
+		return -ENOMEM;
+	nvkm_subdev_ctor(&nvkm_fault, device, index, &fault->subdev);
+	fault->func = func;
+	return 0;
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fault/priv.h b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/priv.h
new file mode 100644
index 0000000..44843ec
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/priv.h
@@ -0,0 +1,34 @@
+#ifndef __NVKM_FAULT_PRIV_H__
+#define __NVKM_FAULT_PRIV_H__
+#define nvkm_fault_buffer(p) container_of((p), struct nvkm_fault_buffer, object)
+#define nvkm_fault(p) container_of((p), struct nvkm_fault, subdev)
+#include <subdev/fault.h>
+
+#include <core/event.h>
+#include <core/object.h>
+
+struct nvkm_fault_buffer {
+	struct nvkm_object object;
+	struct nvkm_fault *fault;
+	int id;
+	int entries;
+	struct nvkm_memory *mem;
+	struct nvkm_vma *vma;
+};
+
+int nvkm_fault_new_(const struct nvkm_fault_func *, struct nvkm_device *,
+		    int index, struct nvkm_fault **);
+
+struct nvkm_fault_func {
+	void (*init)(struct nvkm_fault *);
+	void (*fini)(struct nvkm_fault *);
+	void (*intr)(struct nvkm_fault *);
+	struct {
+		int nr;
+		u32 entry_size;
+		u32 (*entries)(struct nvkm_fault_buffer *);
+		void (*init)(struct nvkm_fault_buffer *);
+		void (*fini)(struct nvkm_fault_buffer *);
+	} buffer;
+};
+#endif
-- 
cgit v1.1


From d0e9351e420695907e28e921d3786265253787c1 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/fault/gp100: implement replayable fault buffer
 initialisation

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 .../gpu/drm/nouveau/include/nvkm/subdev/fault.h    |  2 +
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c  |  7 +++
 drivers/gpu/drm/nouveau/nvkm/subdev/fault/Kbuild   |  1 +
 drivers/gpu/drm/nouveau/nvkm/subdev/fault/gp100.c  | 69 ++++++++++++++++++++++
 4 files changed, 79 insertions(+)
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/subdev/fault/gp100.c

diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/fault.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/fault.h
index 6ba9c17..8ca66e5 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/fault.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/fault.h
@@ -11,4 +11,6 @@ struct nvkm_fault {
 
 	struct nvkm_event event;
 };
+
+int gp100_fault_new(struct nvkm_device *, int, struct nvkm_fault **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 70f3cc0..379e701 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2161,6 +2161,7 @@ nv130_chipset = {
 	.bios = nvkm_bios_new,
 	.bus = gf100_bus_new,
 	.devinit = gm200_devinit_new,
+	.fault = gp100_fault_new,
 	.fb = gp100_fb_new,
 	.fuse = gm107_fuse_new,
 	.gpio = gk104_gpio_new,
@@ -2196,6 +2197,7 @@ nv132_chipset = {
 	.bios = nvkm_bios_new,
 	.bus = gf100_bus_new,
 	.devinit = gm200_devinit_new,
+	.fault = gp100_fault_new,
 	.fb = gp102_fb_new,
 	.fuse = gm107_fuse_new,
 	.gpio = gk104_gpio_new,
@@ -2231,6 +2233,7 @@ nv134_chipset = {
 	.bios = nvkm_bios_new,
 	.bus = gf100_bus_new,
 	.devinit = gm200_devinit_new,
+	.fault = gp100_fault_new,
 	.fb = gp102_fb_new,
 	.fuse = gm107_fuse_new,
 	.gpio = gk104_gpio_new,
@@ -2266,6 +2269,7 @@ nv136_chipset = {
 	.bios = nvkm_bios_new,
 	.bus = gf100_bus_new,
 	.devinit = gm200_devinit_new,
+	.fault = gp100_fault_new,
 	.fb = gp102_fb_new,
 	.fuse = gm107_fuse_new,
 	.gpio = gk104_gpio_new,
@@ -2301,6 +2305,7 @@ nv137_chipset = {
 	.bios = nvkm_bios_new,
 	.bus = gf100_bus_new,
 	.devinit = gm200_devinit_new,
+	.fault = gp100_fault_new,
 	.fb = gp102_fb_new,
 	.fuse = gm107_fuse_new,
 	.gpio = gk104_gpio_new,
@@ -2336,6 +2341,7 @@ nv138_chipset = {
 	.bios = nvkm_bios_new,
 	.bus = gf100_bus_new,
 	.devinit = gm200_devinit_new,
+	.fault = gp100_fault_new,
 	.fb = gp102_fb_new,
 	.fuse = gm107_fuse_new,
 	.gpio = gk104_gpio_new,
@@ -2369,6 +2375,7 @@ nv13b_chipset = {
 	.name = "GP10B",
 	.bar = gm20b_bar_new,
 	.bus = gf100_bus_new,
+	.fault = gp100_fault_new,
 	.fb = gp10b_fb_new,
 	.fuse = gm107_fuse_new,
 	.ibus = gp10b_ibus_new,
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fault/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/Kbuild
index 2e4c226..807ea40 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/fault/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/Kbuild
@@ -1 +1,2 @@
 nvkm-y += nvkm/subdev/fault/base.o
+nvkm-y += nvkm/subdev/fault/gp100.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fault/gp100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/gp100.c
new file mode 100644
index 0000000..5e71db2
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/gp100.c
@@ -0,0 +1,69 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "priv.h"
+
+#include <subdev/mmu.h>
+
+static void
+gp100_fault_buffer_fini(struct nvkm_fault_buffer *buffer)
+{
+	struct nvkm_device *device = buffer->fault->subdev.device;
+	nvkm_mask(device, 0x002a70, 0x00000001, 0x00000000);
+}
+
+static void
+gp100_fault_buffer_init(struct nvkm_fault_buffer *buffer)
+{
+	struct nvkm_device *device = buffer->fault->subdev.device;
+	nvkm_wr32(device, 0x002a74, upper_32_bits(buffer->vma->addr));
+	nvkm_wr32(device, 0x002a70, lower_32_bits(buffer->vma->addr));
+	nvkm_mask(device, 0x002a70, 0x00000001, 0x00000001);
+}
+
+static u32
+gp100_fault_buffer_entries(struct nvkm_fault_buffer *buffer)
+{
+	return nvkm_rd32(buffer->fault->subdev.device, 0x002a78);
+}
+
+static void
+gp100_fault_intr(struct nvkm_fault *fault)
+{
+	nvkm_event_send(&fault->event, 1, 0, NULL, 0);
+}
+
+static const struct nvkm_fault_func
+gp100_fault = {
+	.intr = gp100_fault_intr,
+	.buffer.nr = 1,
+	.buffer.entry_size = 32,
+	.buffer.entries = gp100_fault_buffer_entries,
+	.buffer.init = gp100_fault_buffer_init,
+	.buffer.fini = gp100_fault_buffer_fini,
+};
+
+int
+gp100_fault_new(struct nvkm_device *device, int index,
+		struct nvkm_fault **pfault)
+{
+	return nvkm_fault_new_(&gp100_fault, device, index, pfault);
+}
-- 
cgit v1.1


From d1ea77ab5f7cf378864255fa90f1ab70676a2a10 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/fb/gf100-: bump size of mmu debug buffers to match big
 page size

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/subdev/fb/gf100.c | 4 ++--
 drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp100.c | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gf100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gf100.c
index cdc4e0a2..e8dc4e9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gf100.c
@@ -46,10 +46,10 @@ gf100_fb_oneinit(struct nvkm_fb *base)
 {
 	struct gf100_fb *fb = gf100_fb(base);
 	struct nvkm_device *device = fb->base.subdev.device;
-	int ret, size = 0x1000;
+	int ret, size = 1 << (fb->base.page ? fb->base.page : 17);
 
 	size = nvkm_longopt(device->cfgopt, "MmuDebugBufferSize", size);
-	size = min(size, 0x1000);
+	size = max(size, 0x1000);
 
 	ret = nvkm_memory_new(device, NVKM_MEM_TARGET_INST, size, 0x1000,
 			      true, &fb->base.mmu_rd);
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp100.c
index 147f69b..d0a47b9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp100.c
@@ -48,7 +48,7 @@ gp100_fb_init(struct nvkm_fb *base)
 	nvkm_wr32(device, 0x100cc8, nvkm_memory_addr(fb->base.mmu_wr) >> 8);
 	nvkm_wr32(device, 0x100ccc, nvkm_memory_addr(fb->base.mmu_rd) >> 8);
 	nvkm_mask(device, 0x100cc4, 0x00060000,
-		  max(nvkm_memory_size(fb->base.mmu_rd) >> 16, (u64)2) << 17);
+		  min(nvkm_memory_size(fb->base.mmu_rd) >> 16, (u64)2) << 17);
 }
 
 static const struct nvkm_fb_func
-- 
cgit v1.1


From 85a3b9c8398b2c4c3698a9d851165acf4ffc8d26 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/fb/gm200-: fix overwriting of big page setting

Likely a rebase bug.  Should have no impact in default configuration due
to using per-instance setting by default.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/subdev/fb/gm200.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gm200.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gm200.c
index 8137e19..d3b8c33 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gm200.c
@@ -49,8 +49,6 @@ gm200_fb_init(struct nvkm_fb *base)
 	if (fb->r100c10_page)
 		nvkm_wr32(device, 0x100c10, fb->r100c10 >> 8);
 
-	nvkm_mask(device, 0x100c80, 0x00000001, 0x00000000); /* 128KiB lpg */
-
 	nvkm_wr32(device, 0x100cc8, nvkm_memory_addr(fb->base.mmu_wr) >> 8);
 	nvkm_wr32(device, 0x100ccc, nvkm_memory_addr(fb->base.mmu_rd) >> 8);
 	nvkm_mask(device, 0x100cc4, 0x00060000,
-- 
cgit v1.1


From 3b9ba66ab0e027e6d7a2b9b62cbb5cd547421ebd Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/disp/nv50-: delay subunit construction until oneinit

We should be reading registers to determine which subunits are really
present on a given board, and this needs to be done after DEVINIT.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/disp/base.c |  6 +++
 drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c | 66 +++++++++++++++----------
 drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h |  4 ++
 drivers/gpu/drm/nouveau/nvkm/engine/disp/priv.h |  1 +
 4 files changed, 51 insertions(+), 26 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/base.c
index 93a75e5..5b9d9c6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/base.c
@@ -376,6 +376,12 @@ nvkm_disp_oneinit(struct nvkm_engine *engine)
 	if (ret)
 		return ret;
 
+	if (disp->func->oneinit) {
+		ret = disp->func->oneinit(disp);
+		if (ret)
+			return ret;
+	}
+
 	i = 0;
 	list_for_each_entry(head, &disp->head, head)
 		i = max(i, head->id + 1);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
index 0c570db..43373b3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
@@ -59,9 +59,44 @@ nv50_disp_dtor_(struct nvkm_disp *base)
 	return disp;
 }
 
+static int
+nv50_disp_oneinit_(struct nvkm_disp *base)
+{
+	struct nv50_disp *disp = nv50_disp(base);
+	const struct nv50_disp_func *func = disp->func;
+	int ret, i;
+
+	for (i = 0; func->head.new && i < disp->head.nr; i++) {
+		ret = func->head.new(&disp->base, i);
+		if (ret)
+			return ret;
+	}
+
+	for (i = 0; func->dac.new && i < func->dac.nr; i++) {
+		ret = func->dac.new(&disp->base, i);
+		if (ret)
+			return ret;
+	}
+
+	for (i = 0; func->pior.new && i < func->pior.nr; i++) {
+		ret = func->pior.new(&disp->base, i);
+		if (ret)
+			return ret;
+	}
+
+	for (i = 0; func->sor.new && i < func->sor.nr; i++) {
+		ret = func->sor.new(&disp->base, i);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
 static const struct nvkm_disp_func
 nv50_disp_ = {
 	.dtor = nv50_disp_dtor_,
+	.oneinit = nv50_disp_oneinit_,
 	.intr = nv50_disp_intr_,
 	.root = nv50_disp_root_,
 };
@@ -71,7 +106,7 @@ nv50_disp_new_(const struct nv50_disp_func *func, struct nvkm_device *device,
 	       int index, int heads, struct nvkm_disp **pdisp)
 {
 	struct nv50_disp *disp;
-	int ret, i;
+	int ret;
 
 	if (!(disp = kzalloc(sizeof(*disp), GFP_KERNEL)))
 		return -ENOMEM;
@@ -85,33 +120,12 @@ nv50_disp_new_(const struct nv50_disp_func *func, struct nvkm_device *device,
 	disp->wq = create_singlethread_workqueue("nvkm-disp");
 	if (!disp->wq)
 		return -ENOMEM;
-	INIT_WORK(&disp->supervisor, func->super);
-
-	for (i = 0; func->head.new && i < heads; i++) {
-		ret = func->head.new(&disp->base, i);
-		if (ret)
-			return ret;
-	}
 
-	for (i = 0; func->dac.new && i < func->dac.nr; i++) {
-		ret = func->dac.new(&disp->base, i);
-		if (ret)
-			return ret;
-	}
-
-	for (i = 0; func->pior.new && i < func->pior.nr; i++) {
-		ret = func->pior.new(&disp->base, i);
-		if (ret)
-			return ret;
-	}
-
-	for (i = 0; func->sor.new && i < func->sor.nr; i++) {
-		ret = func->sor.new(&disp->base, i);
-		if (ret)
-			return ret;
-	}
+	INIT_WORK(&disp->supervisor, func->super);
+	disp->head.nr = heads;
 
-	return nvkm_event_init(func->uevent, 1, 1 + (heads * 4), &disp->uevent);
+	return nvkm_event_init(func->uevent, 1, ARRAY_SIZE(disp->chan),
+			       &disp->uevent);
 }
 
 static u32
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
index eb0b8ac..0ae32cd 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
@@ -16,6 +16,10 @@ struct nv50_disp {
 	struct nvkm_event uevent;
 
 	struct {
+		int nr;
+	} head;
+
+	struct {
 		u32 lvdsconf;
 	} sor;
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/priv.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/priv.h
index 6c9bfff..c614351 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/priv.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/priv.h
@@ -12,6 +12,7 @@ void nvkm_disp_vblank(struct nvkm_disp *, int head);
 
 struct nvkm_disp_func {
 	void *(*dtor)(struct nvkm_disp *);
+	int (*oneinit)(struct nvkm_disp *);
 	void (*intr)(struct nvkm_disp *);
 
 	const struct nvkm_disp_oclass *(*root)(struct nvkm_disp *);
-- 
cgit v1.1


From f7b2ece37fce822692d3d6e616e0d0f3df9d4f49 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/disp/nv50-: fetch mask of available heads during oneinit

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c       |  4 ++--
 drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c       |  4 ++--
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c     | 12 ++----------
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c     |  4 ++--
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c     |  4 ++--
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c     |  4 ++--
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c     |  4 ++--
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gp100.c     |  4 ++--
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gp102.c     |  4 ++--
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c     |  4 ++--
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c     |  4 ++--
 drivers/gpu/drm/nouveau/nvkm/engine/disp/head.h      |  4 ++++
 drivers/gpu/drm/nouveau/nvkm/engine/disp/headgf119.c | 11 ++++++++---
 drivers/gpu/drm/nouveau/nvkm/engine/disp/headnv50.c  |  7 +++++++
 drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c     |  4 ++--
 drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c     |  4 ++--
 drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c      | 13 ++++++++-----
 drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h      |  6 +++---
 18 files changed, 56 insertions(+), 45 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c
index 842e1b7..a1741a8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c
@@ -32,7 +32,7 @@ g84_disp = {
 	.uevent = &nv50_disp_chan_uevent,
 	.super = nv50_disp_super,
 	.root = &g84_disp_root_oclass,
-	.head.new = nv50_head_new,
+	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
 	.dac = { .nr = 3, .new = nv50_dac_new },
 	.sor = { .nr = 2, .new = g84_sor_new },
 	.pior = { .nr = 3, .new = nv50_pior_new },
@@ -41,5 +41,5 @@ g84_disp = {
 int
 g84_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return nv50_disp_new_(&g84_disp, device, index, 2, pdisp);
+	return nv50_disp_new_(&g84_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c
index d184e6a..b7febc4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c
@@ -32,7 +32,7 @@ g94_disp = {
 	.uevent = &nv50_disp_chan_uevent,
 	.super = nv50_disp_super,
 	.root = &g94_disp_root_oclass,
-	.head.new = nv50_head_new,
+	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
 	.dac = { .nr = 3, .new = nv50_dac_new },
 	.sor = { .nr = 4, .new = g94_sor_new },
 	.pior = { .nr = 3, .new = nv50_pior_new },
@@ -41,5 +41,5 @@ g94_disp = {
 int
 g94_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return nv50_disp_new_(&g94_disp, device, index, 2, pdisp);
+	return nv50_disp_new_(&g94_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c
index d8765b5..b580ca9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c
@@ -164,14 +164,6 @@ gf119_disp_intr(struct nv50_disp *disp)
 	}
 }
 
-int
-gf119_disp_new_(const struct nv50_disp_func *func, struct nvkm_device *device,
-		int index, struct nvkm_disp **pdisp)
-{
-	u32 heads = nvkm_rd32(device, 0x022448);
-	return nv50_disp_new_(func, device, index, heads, pdisp);
-}
-
 static const struct nv50_disp_func
 gf119_disp = {
 	.intr = gf119_disp_intr,
@@ -179,7 +171,7 @@ gf119_disp = {
 	.uevent = &gf119_disp_chan_uevent,
 	.super = gf119_disp_super,
 	.root = &gf119_disp_root_oclass,
-	.head.new = gf119_head_new,
+	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
 	.dac = { .nr = 3, .new = gf119_dac_new },
 	.sor = { .nr = 4, .new = gf119_sor_new },
 };
@@ -187,5 +179,5 @@ gf119_disp = {
 int
 gf119_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return gf119_disp_new_(&gf119_disp, device, index, pdisp);
+	return nv50_disp_new_(&gf119_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c
index e8fe9f3..b3ee5b1 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c
@@ -33,7 +33,7 @@ gk104_disp = {
 	.uevent = &gf119_disp_chan_uevent,
 	.super = gf119_disp_super,
 	.root = &gk104_disp_root_oclass,
-	.head.new = gf119_head_new,
+	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
 	.dac = { .nr = 3, .new = gf119_dac_new },
 	.sor = { .nr = 4, .new = gk104_sor_new },
 };
@@ -41,5 +41,5 @@ gk104_disp = {
 int
 gk104_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return gf119_disp_new_(&gk104_disp, device, index, pdisp);
+	return nv50_disp_new_(&gk104_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c
index 7696875..22533ab 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c
@@ -33,7 +33,7 @@ gk110_disp = {
 	.uevent = &gf119_disp_chan_uevent,
 	.super = gf119_disp_super,
 	.root = &gk110_disp_root_oclass,
-	.head.new = gf119_head_new,
+	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
 	.dac = { .nr = 3, .new = gf119_dac_new },
 	.sor = { .nr = 4, .new = gk104_sor_new },
 };
@@ -41,5 +41,5 @@ gk110_disp = {
 int
 gk110_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return gf119_disp_new_(&gk110_disp, device, index, pdisp);
+	return nv50_disp_new_(&gk110_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c
index ede70e5..85e6021 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c
@@ -33,7 +33,7 @@ gm107_disp = {
 	.uevent = &gf119_disp_chan_uevent,
 	.super = gf119_disp_super,
 	.root = &gm107_disp_root_oclass,
-	.head.new = gf119_head_new,
+	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
 	.dac = { .nr = 3, .new = gf119_dac_new },
 	.sor = { .nr = 4, .new = gm107_sor_new },
 };
@@ -41,5 +41,5 @@ gm107_disp = {
 int
 gm107_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return gf119_disp_new_(&gm107_disp, device, index, pdisp);
+	return nv50_disp_new_(&gm107_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c
index 292d3b5..48ab65d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c
@@ -33,7 +33,7 @@ gm200_disp = {
 	.uevent = &gf119_disp_chan_uevent,
 	.super = gf119_disp_super,
 	.root = &gm200_disp_root_oclass,
-	.head.new = gf119_head_new,
+	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
 	.dac = { .nr = 3, .new = gf119_dac_new },
 	.sor = { .nr = 4, .new = gm200_sor_new },
 };
@@ -41,5 +41,5 @@ gm200_disp = {
 int
 gm200_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return gf119_disp_new_(&gm200_disp, device, index, pdisp);
+	return nv50_disp_new_(&gm200_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp100.c
index 39eb98b..190d030 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp100.c
@@ -33,12 +33,12 @@ gp100_disp = {
 	.uevent = &gf119_disp_chan_uevent,
 	.super = gf119_disp_super,
 	.root = &gp100_disp_root_oclass,
-	.head.new = gf119_head_new,
+	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
 	.sor = { .nr = 4, .new = gm200_sor_new },
 };
 
 int
 gp100_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return gf119_disp_new_(&gp100_disp, device, index, pdisp);
+	return nv50_disp_new_(&gp100_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp102.c
index 91d70fe1..a059c65 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp102.c
@@ -59,12 +59,12 @@ gp102_disp = {
 	.uevent = &gf119_disp_chan_uevent,
 	.super = gf119_disp_super,
 	.root = &gp102_disp_root_oclass,
-	.head.new = gf119_head_new,
+	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
 	.sor = { .nr = 4, .new = gm200_sor_new },
 };
 
 int
 gp102_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return gf119_disp_new_(&gp102_disp, device, index, pdisp);
+	return nv50_disp_new_(&gp102_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c
index bf00c4e..7554c73 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c
@@ -32,7 +32,7 @@ gt200_disp = {
 	.uevent = &nv50_disp_chan_uevent,
 	.super = nv50_disp_super,
 	.root = &gt200_disp_root_oclass,
-	.head.new = nv50_head_new,
+	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
 	.dac = { .nr = 3, .new = nv50_dac_new },
 	.sor = { .nr = 2, .new = g84_sor_new },
 	.pior = { .nr = 3, .new = nv50_pior_new },
@@ -41,5 +41,5 @@ gt200_disp = {
 int
 gt200_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return nv50_disp_new_(&gt200_disp, device, index, 2, pdisp);
+	return nv50_disp_new_(&gt200_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c
index 2cdd4d7..6a878a7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c
@@ -32,7 +32,7 @@ gt215_disp = {
 	.uevent = &nv50_disp_chan_uevent,
 	.super = nv50_disp_super,
 	.root = &gt215_disp_root_oclass,
-	.head.new = nv50_head_new,
+	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
 	.dac = { .nr = 3, .new = nv50_dac_new },
 	.sor = { .nr = 4, .new = gt215_sor_new },
 	.pior = { .nr = 3, .new = nv50_pior_new },
@@ -41,5 +41,5 @@ gt215_disp = {
 int
 gt215_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return nv50_disp_new_(&gt215_disp, device, index, 2, pdisp);
+	return nv50_disp_new_(&gt215_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/head.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/head.h
index 57030b3..4a5d789 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/head.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/head.h
@@ -52,6 +52,10 @@ void nv50_head_rgpos(struct nvkm_head *, u16 *, u16 *);
 #define HEAD_DBG(h,f,a...) HEAD_MSG((h), debug, f, ##a)
 
 int nv04_head_new(struct nvkm_disp *, int id);
+
+int nv50_head_cnt(struct nvkm_disp *, unsigned long *);
 int nv50_head_new(struct nvkm_disp *, int id);
+
+int gf119_head_cnt(struct nvkm_disp *, unsigned long *);
 int gf119_head_new(struct nvkm_disp *, int id);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/headgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/headgf119.c
index 9fd7ae3..bcbdaaf 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/headgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/headgf119.c
@@ -92,8 +92,13 @@ gf119_head = {
 int
 gf119_head_new(struct nvkm_disp *disp, int id)
 {
-	struct nvkm_device *device = disp->engine.subdev.device;
-	if (!(nvkm_rd32(device, 0x612004) & (0x00000001 << id)))
-		return 0;
 	return nvkm_head_new_(&gf119_head, disp, id);
 }
+
+int
+gf119_head_cnt(struct nvkm_disp *disp, unsigned long *pmask)
+{
+	struct nvkm_device *device = disp->engine.subdev.device;
+	*pmask = nvkm_rd32(device, 0x612004) & 0x0000000f;
+	return nvkm_rd32(device, 0x022448);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/headnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/headnv50.c
index c80d06d..e7d5c39 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/headnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/headnv50.c
@@ -90,3 +90,10 @@ nv50_head_new(struct nvkm_disp *disp, int id)
 {
 	return nvkm_head_new_(&nv50_head, disp, id);
 }
+
+int
+nv50_head_cnt(struct nvkm_disp *disp, unsigned long *pmask)
+{
+	*pmask = 3;
+	return 2;
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c
index d7e0fbb..9fc7507 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c
@@ -30,7 +30,7 @@ mcp77_disp = {
 	.uevent = &nv50_disp_chan_uevent,
 	.super = nv50_disp_super,
 	.root = &g94_disp_root_oclass,
-	.head.new = nv50_head_new,
+	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
 	.dac = { .nr = 3, .new = nv50_dac_new },
 	.sor = { .nr = 4, .new = mcp77_sor_new },
 	.pior = { .nr = 3, .new = nv50_pior_new },
@@ -39,5 +39,5 @@ mcp77_disp = {
 int
 mcp77_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return nv50_disp_new_(&mcp77_disp, device, index, 2, pdisp);
+	return nv50_disp_new_(&mcp77_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c
index 7b75c57..28647d3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c
@@ -30,7 +30,7 @@ mcp89_disp = {
 	.uevent = &nv50_disp_chan_uevent,
 	.super = nv50_disp_super,
 	.root = &gt215_disp_root_oclass,
-	.head.new = nv50_head_new,
+	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
 	.dac = { .nr = 3, .new = nv50_dac_new },
 	.sor = { .nr = 4, .new = mcp89_sor_new },
 	.pior = { .nr = 3, .new = nv50_pior_new },
@@ -39,5 +39,5 @@ mcp89_disp = {
 int
 mcp89_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return nv50_disp_new_(&mcp89_disp, device, index, 2, pdisp);
+	return nv50_disp_new_(&mcp89_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
index 43373b3..bf50b7b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
@@ -64,9 +64,13 @@ nv50_disp_oneinit_(struct nvkm_disp *base)
 {
 	struct nv50_disp *disp = nv50_disp(base);
 	const struct nv50_disp_func *func = disp->func;
+	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
 	int ret, i;
 
-	for (i = 0; func->head.new && i < disp->head.nr; i++) {
+	disp->head.nr = func->head.cnt(&disp->base, &disp->head.mask);
+	nvkm_debug(subdev, "  Head(s): %d (%02lx)\n",
+		   disp->head.nr, disp->head.mask);
+	for_each_set_bit(i, &disp->head.mask, disp->head.nr) {
 		ret = func->head.new(&disp->base, i);
 		if (ret)
 			return ret;
@@ -103,7 +107,7 @@ nv50_disp_ = {
 
 int
 nv50_disp_new_(const struct nv50_disp_func *func, struct nvkm_device *device,
-	       int index, int heads, struct nvkm_disp **pdisp)
+	       int index, struct nvkm_disp **pdisp)
 {
 	struct nv50_disp *disp;
 	int ret;
@@ -122,7 +126,6 @@ nv50_disp_new_(const struct nv50_disp_func *func, struct nvkm_device *device,
 		return -ENOMEM;
 
 	INIT_WORK(&disp->supervisor, func->super);
-	disp->head.nr = heads;
 
 	return nvkm_event_init(func->uevent, 1, ARRAY_SIZE(disp->chan),
 			       &disp->uevent);
@@ -633,7 +636,7 @@ nv50_disp = {
 	.uevent = &nv50_disp_chan_uevent,
 	.super = nv50_disp_super,
 	.root = &nv50_disp_root_oclass,
-	.head.new = nv50_head_new,
+	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
 	.dac = { .nr = 3, .new = nv50_dac_new },
 	.sor = { .nr = 2, .new = nv50_sor_new },
 	.pior = { .nr = 3, .new = nv50_pior_new },
@@ -642,5 +645,5 @@ nv50_disp = {
 int
 nv50_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return nv50_disp_new_(&nv50_disp, device, index, 2, pdisp);
+	return nv50_disp_new_(&nv50_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
index 0ae32cd..c7fe29c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
@@ -16,6 +16,7 @@ struct nv50_disp {
 	struct nvkm_event uevent;
 
 	struct {
+		unsigned long mask;
 		int nr;
 	} head;
 
@@ -38,9 +39,7 @@ void nv50_disp_super_2_2(struct nv50_disp *, struct nvkm_head *);
 void nv50_disp_super_3_0(struct nv50_disp *, struct nvkm_head *);
 
 int nv50_disp_new_(const struct nv50_disp_func *, struct nvkm_device *,
-		   int index, int heads, struct nvkm_disp **);
-int gf119_disp_new_(const struct nv50_disp_func *, struct nvkm_device *,
-		    int index, struct nvkm_disp **);
+		   int index, struct nvkm_disp **);
 
 struct nv50_disp_func {
 	void (*intr)(struct nv50_disp *);
@@ -52,6 +51,7 @@ struct nv50_disp_func {
 	const struct nvkm_disp_oclass *root;
 
 	struct {
+		int (*cnt)(struct nvkm_disp *, unsigned long *mask);
 		int (*new)(struct nvkm_disp *, int id);
 	} head;
 
-- 
cgit v1.1


From bf5d1a6b6a4489b7887589fca6321d4024da71c8 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/disp/nv50-: fetch mask of available dacs during oneinit

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/disp/dacgf119.c  | 11 ++++++++---
 drivers/gpu/drm/nouveau/nvkm/engine/disp/dacnv50.c   | 11 ++++++++---
 drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c       |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c       |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c     |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c     |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c     |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c     |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c     |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c     |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c     |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/ior.h       |  3 +++
 drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c     |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c     |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c      | 15 ++++++++++-----
 drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h      |  9 ++-------
 drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c  |  2 +-
 18 files changed, 44 insertions(+), 31 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dacgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dacgf119.c
index dbd032e..71a9477 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dacgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dacgf119.c
@@ -58,8 +58,13 @@ gf119_dac = {
 int
 gf119_dac_new(struct nvkm_disp *disp, int id)
 {
-	struct nvkm_device *device = disp->engine.subdev.device;
-	if (!(nvkm_rd32(device, 0x612004) & (0x00000010 << id)))
-		return 0;
 	return nvkm_ior_new_(&gf119_dac, disp, DAC, id);
 }
+
+int
+gf119_dac_cnt(struct nvkm_disp *disp, unsigned long *pmask)
+{
+	struct nvkm_device *device = disp->engine.subdev.device;
+	*pmask = (nvkm_rd32(device, 0x612004) & 0x000000f0) >> 4;
+	return 4;
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dacnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dacnv50.c
index 85e692b..558012d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dacnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dacnv50.c
@@ -109,8 +109,13 @@ nv50_dac = {
 int
 nv50_dac_new(struct nvkm_disp *disp, int id)
 {
-	struct nvkm_device *device = disp->engine.subdev.device;
-	if (!(nvkm_rd32(device, 0x610184) & (0x00100000 << id)))
-		return 0;
 	return nvkm_ior_new_(&nv50_dac, disp, DAC, id);
 }
+
+int
+nv50_dac_cnt(struct nvkm_disp *disp, unsigned long *pmask)
+{
+	struct nvkm_device *device = disp->engine.subdev.device;
+	*pmask = (nvkm_rd32(device, 0x610184) & 0x00700000) >> 20;
+	return 3;
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c
index a1741a8..9f31f04 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c
@@ -33,7 +33,7 @@ g84_disp = {
 	.super = nv50_disp_super,
 	.root = &g84_disp_root_oclass,
 	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
-	.dac = { .nr = 3, .new = nv50_dac_new },
+	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
 	.sor = { .nr = 2, .new = g84_sor_new },
 	.pior = { .nr = 3, .new = nv50_pior_new },
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c
index b7febc4..71fe26b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c
@@ -33,7 +33,7 @@ g94_disp = {
 	.super = nv50_disp_super,
 	.root = &g94_disp_root_oclass,
 	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
-	.dac = { .nr = 3, .new = nv50_dac_new },
+	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
 	.sor = { .nr = 4, .new = g94_sor_new },
 	.pior = { .nr = 3, .new = nv50_pior_new },
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c
index b580ca9..c8495bc 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c
@@ -172,7 +172,7 @@ gf119_disp = {
 	.super = gf119_disp_super,
 	.root = &gf119_disp_root_oclass,
 	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
-	.dac = { .nr = 3, .new = gf119_dac_new },
+	.dac = { .cnt = gf119_dac_cnt, .new = gf119_dac_new },
 	.sor = { .nr = 4, .new = gf119_sor_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c
index b3ee5b1..b948619 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c
@@ -34,7 +34,7 @@ gk104_disp = {
 	.super = gf119_disp_super,
 	.root = &gk104_disp_root_oclass,
 	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
-	.dac = { .nr = 3, .new = gf119_dac_new },
+	.dac = { .cnt = gf119_dac_cnt, .new = gf119_dac_new },
 	.sor = { .nr = 4, .new = gk104_sor_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c
index 22533ab..0b49457 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c
@@ -34,7 +34,7 @@ gk110_disp = {
 	.super = gf119_disp_super,
 	.root = &gk110_disp_root_oclass,
 	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
-	.dac = { .nr = 3, .new = gf119_dac_new },
+	.dac = { .cnt = gf119_dac_cnt, .new = gf119_dac_new },
 	.sor = { .nr = 4, .new = gk104_sor_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c
index 85e6021..35ad965 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c
@@ -34,7 +34,7 @@ gm107_disp = {
 	.super = gf119_disp_super,
 	.root = &gm107_disp_root_oclass,
 	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
-	.dac = { .nr = 3, .new = gf119_dac_new },
+	.dac = { .cnt = gf119_dac_cnt, .new = gf119_dac_new },
 	.sor = { .nr = 4, .new = gm107_sor_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c
index 48ab65d..b2a5d36 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c
@@ -34,7 +34,7 @@ gm200_disp = {
 	.super = gf119_disp_super,
 	.root = &gm200_disp_root_oclass,
 	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
-	.dac = { .nr = 3, .new = gf119_dac_new },
+	.dac = { .cnt = gf119_dac_cnt, .new = gf119_dac_new },
 	.sor = { .nr = 4, .new = gm200_sor_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c
index 7554c73..88b6a34 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c
@@ -33,7 +33,7 @@ gt200_disp = {
 	.super = nv50_disp_super,
 	.root = &gt200_disp_root_oclass,
 	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
-	.dac = { .nr = 3, .new = nv50_dac_new },
+	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
 	.sor = { .nr = 2, .new = g84_sor_new },
 	.pior = { .nr = 3, .new = nv50_pior_new },
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c
index 6a878a7..1519ca5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c
@@ -33,7 +33,7 @@ gt215_disp = {
 	.super = nv50_disp_super,
 	.root = &gt215_disp_root_oclass,
 	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
-	.dac = { .nr = 3, .new = nv50_dac_new },
+	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
 	.sor = { .nr = 4, .new = gt215_sor_new },
 	.pior = { .nr = 3, .new = nv50_pior_new },
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ior.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ior.h
index 4548c03..89d9ab1 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ior.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ior.h
@@ -153,7 +153,10 @@ void gf119_hda_eld(struct nvkm_ior *, u8 *, u8);
 #define IOR_WARN(i,f,a...) IOR_MSG((i), warn, f, ##a)
 #define IOR_DBG(i,f,a...) IOR_MSG((i), debug, f, ##a)
 
+int nv50_dac_cnt(struct nvkm_disp *, unsigned long *);
 int nv50_dac_new(struct nvkm_disp *, int);
+
+int gf119_dac_cnt(struct nvkm_disp *, unsigned long *);
 int gf119_dac_new(struct nvkm_disp *, int);
 
 int nv50_pior_new(struct nvkm_disp *, int);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c
index 9fc7507..fb18854 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c
@@ -31,7 +31,7 @@ mcp77_disp = {
 	.super = nv50_disp_super,
 	.root = &g94_disp_root_oclass,
 	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
-	.dac = { .nr = 3, .new = nv50_dac_new },
+	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
 	.sor = { .nr = 4, .new = mcp77_sor_new },
 	.pior = { .nr = 3, .new = nv50_pior_new },
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c
index 28647d3..ae7fb96 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c
@@ -31,7 +31,7 @@ mcp89_disp = {
 	.super = nv50_disp_super,
 	.root = &gt215_disp_root_oclass,
 	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
-	.dac = { .nr = 3, .new = nv50_dac_new },
+	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
 	.sor = { .nr = 4, .new = mcp89_sor_new },
 	.pior = { .nr = 3, .new = nv50_pior_new },
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
index bf50b7b..500ef8d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
@@ -76,10 +76,15 @@ nv50_disp_oneinit_(struct nvkm_disp *base)
 			return ret;
 	}
 
-	for (i = 0; func->dac.new && i < func->dac.nr; i++) {
-		ret = func->dac.new(&disp->base, i);
-		if (ret)
-			return ret;
+	if (func->dac.cnt) {
+		disp->dac.nr = func->dac.cnt(&disp->base, &disp->dac.mask);
+		nvkm_debug(subdev, "   DAC(s): %d (%02lx)\n",
+			   disp->dac.nr, disp->dac.mask);
+		for_each_set_bit(i, &disp->dac.mask, disp->dac.nr) {
+			ret = func->dac.new(&disp->base, i);
+			if (ret)
+				return ret;
+		}
 	}
 
 	for (i = 0; func->pior.new && i < func->pior.nr; i++) {
@@ -637,7 +642,7 @@ nv50_disp = {
 	.super = nv50_disp_super,
 	.root = &nv50_disp_root_oclass,
 	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
-	.dac = { .nr = 3, .new = nv50_dac_new },
+	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
 	.sor = { .nr = 2, .new = nv50_sor_new },
 	.pior = { .nr = 3, .new = nv50_pior_new },
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
index c7fe29c..3415a52 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
@@ -18,7 +18,7 @@ struct nv50_disp {
 	struct {
 		unsigned long mask;
 		int nr;
-	} head;
+	} head, dac;
 
 	struct {
 		u32 lvdsconf;
@@ -53,12 +53,7 @@ struct nv50_disp_func {
 	struct {
 		int (*cnt)(struct nvkm_disp *, unsigned long *mask);
 		int (*new)(struct nvkm_disp *, int id);
-	} head;
-
-	struct {
-		int nr;
-		int (*new)(struct nvkm_disp *, int id);
-	} dac;
+	} head, dac;
 
 	struct {
 		int nr;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
index 333c842..7f22b87 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
@@ -64,7 +64,7 @@ gf119_disp_root_init(struct nv50_disp_root *root)
 	}
 
 	/* ... DAC caps */
-	for (i = 0; i < disp->func->dac.nr; i++) {
+	for (i = 0; i < disp->dac.nr; i++) {
 		tmp = nvkm_rd32(device, 0x61a000 + (i * 0x800));
 		nvkm_wr32(device, 0x6101c0 + (i * 0x800), tmp);
 	}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
index 1208524..3fbaccd9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
@@ -427,7 +427,7 @@ nv50_disp_root_init(struct nv50_disp_root *root)
 	}
 
 	/* ... DAC caps */
-	for (i = 0; i < disp->func->dac.nr; i++) {
+	for (i = 0; i < disp->dac.nr; i++) {
 		tmp = nvkm_rd32(device, 0x61a000 + (i * 0x800));
 		nvkm_wr32(device, 0x6101d0 + (i * 0x04), tmp);
 	}
-- 
cgit v1.1


From 9fe4e177045f4b5af25d25859e30450ff1f18be9 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/disp/nv50-: fetch mask of available sors during oneinit

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c       |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c       |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c     |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c     |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c     |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c     |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c     |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gp100.c     |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gp102.c     |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c     |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c     |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/ior.h       | 10 ++++++++--
 drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c     |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c     |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c      |  7 +++++--
 drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h      |  9 +++------
 drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c  |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/sorg84.c    |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/sorg94.c    | 10 +++++++++-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgf119.c  | 19 +++++++++----------
 drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgk104.c  |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm107.c  |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm200.c  |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgt215.c  |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/sormcp77.c  |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/sormcp89.c  |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/sornv50.c   | 19 +++++++++----------
 28 files changed, 65 insertions(+), 53 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c
index 9f31f04..287ac5e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c
@@ -34,7 +34,7 @@ g84_disp = {
 	.root = &g84_disp_root_oclass,
 	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
 	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
-	.sor = { .nr = 2, .new = g84_sor_new },
+	.sor = { .cnt = nv50_sor_cnt, .new = g84_sor_new },
 	.pior = { .nr = 3, .new = nv50_pior_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c
index 71fe26b..e21f76d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c
@@ -34,7 +34,7 @@ g94_disp = {
 	.root = &g94_disp_root_oclass,
 	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
 	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
-	.sor = { .nr = 4, .new = g94_sor_new },
+	.sor = { .cnt = g94_sor_cnt, .new = g94_sor_new },
 	.pior = { .nr = 3, .new = nv50_pior_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c
index c8495bc..0139d14 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c
@@ -173,7 +173,7 @@ gf119_disp = {
 	.root = &gf119_disp_root_oclass,
 	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
 	.dac = { .cnt = gf119_dac_cnt, .new = gf119_dac_new },
-	.sor = { .nr = 4, .new = gf119_sor_new },
+	.sor = { .cnt = gf119_sor_cnt, .new = gf119_sor_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c
index b948619..6a59a52 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c
@@ -35,7 +35,7 @@ gk104_disp = {
 	.root = &gk104_disp_root_oclass,
 	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
 	.dac = { .cnt = gf119_dac_cnt, .new = gf119_dac_new },
-	.sor = { .nr = 4, .new = gk104_sor_new },
+	.sor = { .cnt = gf119_sor_cnt, .new = gk104_sor_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c
index 0b49457..f3b10dc4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c
@@ -35,7 +35,7 @@ gk110_disp = {
 	.root = &gk110_disp_root_oclass,
 	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
 	.dac = { .cnt = gf119_dac_cnt, .new = gf119_dac_new },
-	.sor = { .nr = 4, .new = gk104_sor_new },
+	.sor = { .cnt = gf119_sor_cnt, .new = gk104_sor_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c
index 35ad965..068c595 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c
@@ -35,7 +35,7 @@ gm107_disp = {
 	.root = &gm107_disp_root_oclass,
 	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
 	.dac = { .cnt = gf119_dac_cnt, .new = gf119_dac_new },
-	.sor = { .nr = 4, .new = gm107_sor_new },
+	.sor = { .cnt = gf119_sor_cnt, .new = gm107_sor_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c
index b2a5d36..1c27dbe 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c
@@ -35,7 +35,7 @@ gm200_disp = {
 	.root = &gm200_disp_root_oclass,
 	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
 	.dac = { .cnt = gf119_dac_cnt, .new = gf119_dac_new },
-	.sor = { .nr = 4, .new = gm200_sor_new },
+	.sor = { .cnt = gf119_sor_cnt, .new = gm200_sor_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp100.c
index 190d030..84933b6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp100.c
@@ -34,7 +34,7 @@ gp100_disp = {
 	.super = gf119_disp_super,
 	.root = &gp100_disp_root_oclass,
 	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
-	.sor = { .nr = 4, .new = gm200_sor_new },
+	.sor = { .cnt = gf119_sor_cnt, .new = gm200_sor_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp102.c
index a059c65..b36d926 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp102.c
@@ -60,7 +60,7 @@ gp102_disp = {
 	.super = gf119_disp_super,
 	.root = &gp102_disp_root_oclass,
 	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
-	.sor = { .nr = 4, .new = gm200_sor_new },
+	.sor = { .cnt = gf119_sor_cnt, .new = gm200_sor_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c
index 88b6a34..2ca92a2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c
@@ -34,7 +34,7 @@ gt200_disp = {
 	.root = &gt200_disp_root_oclass,
 	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
 	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
-	.sor = { .nr = 2, .new = g84_sor_new },
+	.sor = { .cnt = nv50_sor_cnt, .new = g84_sor_new },
 	.pior = { .nr = 3, .new = nv50_pior_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c
index 1519ca5..73b0390 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c
@@ -34,7 +34,7 @@ gt215_disp = {
 	.root = &gt215_disp_root_oclass,
 	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
 	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
-	.sor = { .nr = 4, .new = gt215_sor_new },
+	.sor = { .cnt = g94_sor_cnt, .new = gt215_sor_new },
 	.pior = { .nr = 3, .new = nv50_pior_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ior.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ior.h
index 89d9ab1..6432e06 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ior.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ior.h
@@ -106,7 +106,6 @@ nv50_sor_link(struct nvkm_ior *ior)
 	return nv50_ior_base(ior) + ((ior->asy.link == 2) * 0x80);
 }
 
-int nv50_sor_new_(const struct nvkm_ior_func *, struct nvkm_disp *, int id);
 void nv50_sor_state(struct nvkm_ior *, struct nvkm_ior_state *);
 void nv50_sor_power(struct nvkm_ior *, bool, bool, bool, bool, bool);
 void nv50_sor_clock(struct nvkm_ior *);
@@ -122,7 +121,6 @@ void g94_sor_dp_watermark(struct nvkm_ior *, int, u8);
 
 void gt215_sor_dp_audio(struct nvkm_ior *, int, bool);
 
-int gf119_sor_new_(const struct nvkm_ior_func *, struct nvkm_disp *, int id);
 void gf119_sor_state(struct nvkm_ior *, struct nvkm_ior_state *);
 void gf119_sor_clock(struct nvkm_ior *);
 int gf119_sor_dp_links(struct nvkm_ior *, struct nvkm_i2c_aux *);
@@ -161,13 +159,21 @@ int gf119_dac_new(struct nvkm_disp *, int);
 
 int nv50_pior_new(struct nvkm_disp *, int);
 
+int nv50_sor_cnt(struct nvkm_disp *, unsigned long *);
 int nv50_sor_new(struct nvkm_disp *, int);
+
 int g84_sor_new(struct nvkm_disp *, int);
+
+int g94_sor_cnt(struct nvkm_disp *, unsigned long *);
 int g94_sor_new(struct nvkm_disp *, int);
+
 int mcp77_sor_new(struct nvkm_disp *, int);
 int gt215_sor_new(struct nvkm_disp *, int);
 int mcp89_sor_new(struct nvkm_disp *, int);
+
+int gf119_sor_cnt(struct nvkm_disp *, unsigned long *);
 int gf119_sor_new(struct nvkm_disp *, int);
+
 int gk104_sor_new(struct nvkm_disp *, int);
 int gm107_sor_new(struct nvkm_disp *, int);
 int gm200_sor_new(struct nvkm_disp *, int);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c
index fb18854..c3f6504 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c
@@ -32,7 +32,7 @@ mcp77_disp = {
 	.root = &g94_disp_root_oclass,
 	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
 	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
-	.sor = { .nr = 4, .new = mcp77_sor_new },
+	.sor = { .cnt = g94_sor_cnt, .new = mcp77_sor_new },
 	.pior = { .nr = 3, .new = nv50_pior_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c
index ae7fb96..2d1d921 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c
@@ -32,7 +32,7 @@ mcp89_disp = {
 	.root = &gt215_disp_root_oclass,
 	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
 	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
-	.sor = { .nr = 4, .new = mcp89_sor_new },
+	.sor = { .cnt = g94_sor_cnt, .new = mcp89_sor_new },
 	.pior = { .nr = 3, .new = nv50_pior_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
index 500ef8d..1e32814 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
@@ -93,7 +93,10 @@ nv50_disp_oneinit_(struct nvkm_disp *base)
 			return ret;
 	}
 
-	for (i = 0; func->sor.new && i < func->sor.nr; i++) {
+	disp->sor.nr = func->sor.cnt(&disp->base, &disp->sor.mask);
+	nvkm_debug(subdev, "   SOR(s): %d (%02lx)\n",
+		   disp->sor.nr, disp->sor.mask);
+	for_each_set_bit(i, &disp->sor.mask, disp->sor.nr) {
 		ret = func->sor.new(&disp->base, i);
 		if (ret)
 			return ret;
@@ -643,7 +646,7 @@ nv50_disp = {
 	.root = &nv50_disp_root_oclass,
 	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
 	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
-	.sor = { .nr = 2, .new = nv50_sor_new },
+	.sor = { .cnt = nv50_sor_cnt, .new = nv50_sor_new },
 	.pior = { .nr = 3, .new = nv50_pior_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
index 3415a52..ee2968b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
@@ -21,6 +21,8 @@ struct nv50_disp {
 	} head, dac;
 
 	struct {
+		unsigned long mask;
+		int nr;
 		u32 lvdsconf;
 	} sor;
 
@@ -53,12 +55,7 @@ struct nv50_disp_func {
 	struct {
 		int (*cnt)(struct nvkm_disp *, unsigned long *mask);
 		int (*new)(struct nvkm_disp *, int id);
-	} head, dac;
-
-	struct {
-		int nr;
-		int (*new)(struct nvkm_disp *, int id);
-	} sor;
+	} head, dac, sor;
 
 	struct {
 		int nr;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
index 7f22b87..4ba2d80 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
@@ -70,7 +70,7 @@ gf119_disp_root_init(struct nv50_disp_root *root)
 	}
 
 	/* ... SOR caps */
-	for (i = 0; i < disp->func->sor.nr; i++) {
+	for (i = 0; i < disp->sor.nr; i++) {
 		tmp = nvkm_rd32(device, 0x61c000 + (i * 0x800));
 		nvkm_wr32(device, 0x6301c4 + (i * 0x800), tmp);
 	}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
index 3fbaccd9..9d0b5b7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
@@ -433,7 +433,7 @@ nv50_disp_root_init(struct nv50_disp_root *root)
 	}
 
 	/* ... SOR caps */
-	for (i = 0; i < disp->func->sor.nr; i++) {
+	for (i = 0; i < disp->sor.nr; i++) {
 		tmp = nvkm_rd32(device, 0x61c000 + (i * 0x800));
 		nvkm_wr32(device, 0x6101e0 + (i * 0x04), tmp);
 	}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorg84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorg84.c
index f40b909..ec3a7db 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorg84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorg84.c
@@ -34,5 +34,5 @@ g84_sor = {
 int
 g84_sor_new(struct nvkm_disp *disp, int id)
 {
-	return nv50_sor_new_(&g84_sor, disp, id);
+	return nvkm_ior_new_(&g84_sor, disp, SOR, id);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorg94.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorg94.c
index 49aeafd..4d59d02 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorg94.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorg94.c
@@ -279,5 +279,13 @@ g94_sor = {
 int
 g94_sor_new(struct nvkm_disp *disp, int id)
 {
-	return nv50_sor_new_(&g94_sor, disp, id);
+	return nvkm_ior_new_(&g94_sor, disp, SOR, id);
+}
+
+int
+g94_sor_cnt(struct nvkm_disp *disp, unsigned long *pmask)
+{
+	struct nvkm_device *device = disp->engine.subdev.device;
+	*pmask = (nvkm_rd32(device, 0x610184) & 0x0f000000) >> 24;
+	return 4;
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgf119.c
index 700fc75..e6e6dfb 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgf119.c
@@ -152,15 +152,6 @@ gf119_sor_state(struct nvkm_ior *sor, struct nvkm_ior_state *state)
 	state->head = ctrl & 0x0000000f;
 }
 
-int
-gf119_sor_new_(const struct nvkm_ior_func *func, struct nvkm_disp *disp, int id)
-{
-	struct nvkm_device *device = disp->engine.subdev.device;
-	if (!(nvkm_rd32(device, 0x612004) & (0x00000100 << id)))
-		return 0;
-	return nvkm_ior_new_(func, disp, SOR, id);
-}
-
 static const struct nvkm_ior_func
 gf119_sor = {
 	.state = gf119_sor_state,
@@ -189,5 +180,13 @@ gf119_sor = {
 int
 gf119_sor_new(struct nvkm_disp *disp, int id)
 {
-	return gf119_sor_new_(&gf119_sor, disp, id);
+	return nvkm_ior_new_(&gf119_sor, disp, SOR, id);
+}
+
+int
+gf119_sor_cnt(struct nvkm_disp *disp, unsigned long *pmask)
+{
+	struct nvkm_device *device = disp->engine.subdev.device;
+	*pmask = (nvkm_rd32(device, 0x612004) & 0x0000ff00) >> 8;
+	return 8;
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgk104.c
index a1547bd..b94090e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgk104.c
@@ -49,5 +49,5 @@ gk104_sor = {
 int
 gk104_sor_new(struct nvkm_disp *disp, int id)
 {
-	return gf119_sor_new_(&gk104_sor, disp, id);
+	return nvkm_ior_new_(&gk104_sor, disp, SOR, id);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm107.c
index 6023095..e6965de 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm107.c
@@ -63,5 +63,5 @@ gm107_sor = {
 int
 gm107_sor_new(struct nvkm_disp *disp, int id)
 {
-	return gf119_sor_new_(&gm107_sor, disp, id);
+	return nvkm_ior_new_(&gm107_sor, disp, SOR, id);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm200.c
index f9b8107..8bc019b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm200.c
@@ -120,5 +120,5 @@ gm200_sor = {
 int
 gm200_sor_new(struct nvkm_disp *disp, int id)
 {
-	return gf119_sor_new_(&gm200_sor, disp, id);
+	return nvkm_ior_new_(&gm200_sor, disp, SOR, id);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgt215.c
index da228b5..54d134d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgt215.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgt215.c
@@ -65,5 +65,5 @@ gt215_sor = {
 int
 gt215_sor_new(struct nvkm_disp *disp, int id)
 {
-	return nv50_sor_new_(&gt215_sor, disp, id);
+	return nvkm_ior_new_(&gt215_sor, disp, SOR, id);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sormcp77.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sormcp77.c
index c0179cc..8a70dd25 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sormcp77.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sormcp77.c
@@ -44,5 +44,5 @@ mcp77_sor = {
 int
 mcp77_sor_new(struct nvkm_disp *disp, int id)
 {
-	return nv50_sor_new_(&mcp77_sor, disp, id);
+	return nvkm_ior_new_(&mcp77_sor, disp, SOR, id);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sormcp89.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sormcp89.c
index 9bb01cd..eac9c5be 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sormcp89.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sormcp89.c
@@ -49,5 +49,5 @@ mcp89_sor = {
 int
 mcp89_sor_new(struct nvkm_disp *disp, int id)
 {
-	return nv50_sor_new_(&mcp89_sor, disp, id);
+	return nvkm_ior_new_(&mcp89_sor, disp, SOR, id);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sornv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sornv50.c
index f3ebd0c..b4729f8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sornv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sornv50.c
@@ -84,15 +84,6 @@ nv50_sor_state(struct nvkm_ior *sor, struct nvkm_ior_state *state)
 	state->head = ctrl & 0x00000003;
 }
 
-int
-nv50_sor_new_(const struct nvkm_ior_func *func, struct nvkm_disp *disp, int id)
-{
-	struct nvkm_device *device = disp->engine.subdev.device;
-	if (!(nvkm_rd32(device, 0x610184) & (0x01000000 << id)))
-		return 0;
-	return nvkm_ior_new_(func, disp, SOR, id);
-}
-
 static const struct nvkm_ior_func
 nv50_sor = {
 	.state = nv50_sor_state,
@@ -103,5 +94,13 @@ nv50_sor = {
 int
 nv50_sor_new(struct nvkm_disp *disp, int id)
 {
-	return nv50_sor_new_(&nv50_sor, disp, id);
+	return nvkm_ior_new_(&nv50_sor, disp, SOR, id);
+}
+
+int
+nv50_sor_cnt(struct nvkm_disp *disp, unsigned long *pmask)
+{
+	struct nvkm_device *device = disp->engine.subdev.device;
+	*pmask = (nvkm_rd32(device, 0x610184) & 0x03000000) >> 24;
+	return 2;
 }
-- 
cgit v1.1


From f5e088d6f0b4d969b2e7d8f931af082ba2527a56 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/disp/nv50-: fetch mask of available piors during oneinit

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c      |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c      |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c    |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c    |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/ior.h      |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c    |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c    |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c     | 15 ++++++++++-----
 drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h     |  9 +++------
 drivers/gpu/drm/nouveau/nvkm/engine/disp/piornv50.c | 11 ++++++++---
 drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c |  2 +-
 11 files changed, 29 insertions(+), 21 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c
index 287ac5e..1ec81f3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c
@@ -35,7 +35,7 @@ g84_disp = {
 	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
 	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
 	.sor = { .cnt = nv50_sor_cnt, .new = g84_sor_new },
-	.pior = { .nr = 3, .new = nv50_pior_new },
+	.pior = { .cnt = nv50_pior_cnt, .new = nv50_pior_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c
index e21f76d..791c2cd 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c
@@ -35,7 +35,7 @@ g94_disp = {
 	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
 	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
 	.sor = { .cnt = g94_sor_cnt, .new = g94_sor_new },
-	.pior = { .nr = 3, .new = nv50_pior_new },
+	.pior = { .cnt = nv50_pior_cnt, .new = nv50_pior_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c
index 2ca92a2..62e721d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c
@@ -35,7 +35,7 @@ gt200_disp = {
 	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
 	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
 	.sor = { .cnt = nv50_sor_cnt, .new = g84_sor_new },
-	.pior = { .nr = 3, .new = nv50_pior_new },
+	.pior = { .cnt = nv50_pior_cnt, .new = nv50_pior_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c
index 73b0390..a5b1b14 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c
@@ -35,7 +35,7 @@ gt215_disp = {
 	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
 	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
 	.sor = { .cnt = g94_sor_cnt, .new = gt215_sor_new },
-	.pior = { .nr = 3, .new = nv50_pior_new },
+	.pior = { .cnt = nv50_pior_cnt, .new = nv50_pior_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ior.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ior.h
index 6432e06..9d43ab2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ior.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ior.h
@@ -157,6 +157,7 @@ int nv50_dac_new(struct nvkm_disp *, int);
 int gf119_dac_cnt(struct nvkm_disp *, unsigned long *);
 int gf119_dac_new(struct nvkm_disp *, int);
 
+int nv50_pior_cnt(struct nvkm_disp *, unsigned long *);
 int nv50_pior_new(struct nvkm_disp *, int);
 
 int nv50_sor_cnt(struct nvkm_disp *, unsigned long *);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c
index c3f6504..ff49040 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c
@@ -33,7 +33,7 @@ mcp77_disp = {
 	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
 	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
 	.sor = { .cnt = g94_sor_cnt, .new = mcp77_sor_new },
-	.pior = { .nr = 3, .new = nv50_pior_new },
+	.pior = { .cnt = nv50_pior_cnt, .new = nv50_pior_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c
index 2d1d921..0cf968d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c
@@ -33,7 +33,7 @@ mcp89_disp = {
 	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
 	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
 	.sor = { .cnt = g94_sor_cnt, .new = mcp89_sor_new },
-	.pior = { .nr = 3, .new = nv50_pior_new },
+	.pior = { .cnt = nv50_pior_cnt, .new = nv50_pior_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
index 1e32814..c0faa39 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
@@ -87,10 +87,15 @@ nv50_disp_oneinit_(struct nvkm_disp *base)
 		}
 	}
 
-	for (i = 0; func->pior.new && i < func->pior.nr; i++) {
-		ret = func->pior.new(&disp->base, i);
-		if (ret)
-			return ret;
+	if (func->pior.cnt) {
+		disp->pior.nr = func->pior.cnt(&disp->base, &disp->pior.mask);
+		nvkm_debug(subdev, "  PIOR(s): %d (%02lx)\n",
+			   disp->pior.nr, disp->pior.mask);
+		for_each_set_bit(i, &disp->pior.mask, disp->pior.nr) {
+			ret = func->pior.new(&disp->base, i);
+			if (ret)
+				return ret;
+		}
 	}
 
 	disp->sor.nr = func->sor.cnt(&disp->base, &disp->sor.mask);
@@ -647,7 +652,7 @@ nv50_disp = {
 	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
 	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
 	.sor = { .cnt = nv50_sor_cnt, .new = nv50_sor_new },
-	.pior = { .nr = 3, .new = nv50_pior_new },
+	.pior = { .cnt = nv50_pior_cnt, .new = nv50_pior_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
index ee2968b..a29bcf7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
@@ -27,6 +27,8 @@ struct nv50_disp {
 	} sor;
 
 	struct {
+		unsigned long mask;
+		int nr;
 		u8 type[3];
 	} pior;
 
@@ -55,12 +57,7 @@ struct nv50_disp_func {
 	struct {
 		int (*cnt)(struct nvkm_disp *, unsigned long *mask);
 		int (*new)(struct nvkm_disp *, int id);
-	} head, dac, sor;
-
-	struct {
-		int nr;
-		int (*new)(struct nvkm_disp *, int id);
-	} pior;
+	} head, dac, sor, pior;
 };
 
 void nv50_disp_intr(struct nv50_disp *);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/piornv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/piornv50.c
index 99b3b90..e997a20 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/piornv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/piornv50.c
@@ -127,8 +127,13 @@ nv50_pior = {
 int
 nv50_pior_new(struct nvkm_disp *disp, int id)
 {
-	struct nvkm_device *device = disp->engine.subdev.device;
-	if (!(nvkm_rd32(device, 0x610184) & (0x10000000 << id)))
-		return 0;
 	return nvkm_ior_new_(&nv50_pior, disp, PIOR, id);
 }
+
+int
+nv50_pior_cnt(struct nvkm_disp *disp, unsigned long *pmask)
+{
+	struct nvkm_device *device = disp->engine.subdev.device;
+	*pmask = (nvkm_rd32(device, 0x610184) & 0x70000000) >> 28;
+	return 3;
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
index 9d0b5b7..c8379bf 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
@@ -439,7 +439,7 @@ nv50_disp_root_init(struct nv50_disp_root *root)
 	}
 
 	/* ... PIOR caps */
-	for (i = 0; i < disp->func->pior.nr; i++) {
+	for (i = 0; i < disp->pior.nr; i++) {
 		tmp = nvkm_rd32(device, 0x61e000 + (i * 0x800));
 		nvkm_wr32(device, 0x6101f0 + (i * 0x04), tmp);
 	}
-- 
cgit v1.1


From bb3b0a422074606400e6aff216300bb4f012b22a Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/disp/nv50-: initialise from the engine, rather than the
 user object

Engines are initialised on an as-needed basis, so this results in the
same behaviour, whilst allowing us to simplify things a bit.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/disp/base.c    |  17 +++
 .../gpu/drm/nouveau/nvkm/engine/disp/channv50.h    |   1 +
 .../gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c   |   2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c    |  12 +--
 drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c     |   2 +
 drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c     |   2 +
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c   |  82 ++++++++++++++
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c   |   2 +
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c   |   2 +
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c   |   2 +
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c   |   2 +
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gp100.c   |   2 +
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gp102.c   |   2 +
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c   |   2 +
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c   |   2 +
 drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c   |   2 +
 drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c   |   2 +
 drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c    | 107 +++++++++++++++++-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h    |   9 ++
 drivers/gpu/drm/nouveau/nvkm/engine/disp/priv.h    |   2 +
 drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c |   2 -
 drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c |   2 -
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c   |  84 ---------------
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c   |   2 -
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c   |   2 -
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c   |   2 -
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c   |   2 -
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c   |   2 -
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c   |   2 -
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c   |   2 -
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c   |   2 -
 .../gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c    | 120 ---------------------
 .../gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h    |  10 --
 33 files changed, 247 insertions(+), 243 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/base.c
index 5b9d9c6..32fa94a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/base.c
@@ -220,6 +220,9 @@ nvkm_disp_fini(struct nvkm_engine *engine, bool suspend)
 	struct nvkm_conn *conn;
 	struct nvkm_outp *outp;
 
+	if (disp->func->fini)
+		disp->func->fini(disp);
+
 	list_for_each_entry(outp, &disp->outp, head) {
 		nvkm_outp_fini(outp);
 	}
@@ -237,6 +240,7 @@ nvkm_disp_init(struct nvkm_engine *engine)
 	struct nvkm_disp *disp = nvkm_disp(engine);
 	struct nvkm_conn *conn;
 	struct nvkm_outp *outp;
+	struct nvkm_ior *ior;
 
 	list_for_each_entry(conn, &disp->conn, head) {
 		nvkm_conn_init(conn);
@@ -246,6 +250,19 @@ nvkm_disp_init(struct nvkm_engine *engine)
 		nvkm_outp_init(outp);
 	}
 
+	if (disp->func->init) {
+		int ret = disp->func->init(disp);
+		if (ret)
+			return ret;
+	}
+
+	/* Set 'normal' (ie. when it's attached to a head) state for
+	 * each output resource to 'fully enabled'.
+	 */
+	list_for_each_entry(ior, &disp->ior, head) {
+		ior->func->power(ior, true, true, true, true, true);
+	}
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
index 40681db..b518585 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
@@ -4,6 +4,7 @@
 #define nv50_disp_chan(p) container_of((p), struct nv50_disp_chan, object)
 #include <core/object.h>
 #include "nv50.h"
+struct nv50_disp_root;
 
 struct nv50_disp_chan {
 	const struct nv50_disp_chan_func *func;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c
index ce7cd74..6680ff8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c
@@ -31,7 +31,7 @@ int
 gf119_disp_dmac_bind(struct nv50_disp_dmac *chan,
 		     struct nvkm_object *object, u32 handle)
 {
-	return nvkm_ramht_insert(chan->base.root->ramht, object,
+	return nvkm_ramht_insert(chan->base.root->disp->ramht, object,
 				 chan->base.chid.user, -9, handle,
 				 chan->base.chid.user << 27 | 0x00000001);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c
index 070ec5e..c80d047 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c
@@ -33,7 +33,7 @@
 
 struct nv50_disp_dmac_object {
 	struct nvkm_oproxy oproxy;
-	struct nv50_disp_root *root;
+	struct nv50_disp *disp;
 	int hash;
 };
 
@@ -42,7 +42,7 @@ nv50_disp_dmac_child_del_(struct nvkm_oproxy *base)
 {
 	struct nv50_disp_dmac_object *object =
 		container_of(base, typeof(*object), oproxy);
-	nvkm_ramht_remove(object->root->ramht, object->hash);
+	nvkm_ramht_remove(object->disp->ramht, object->hash);
 }
 
 static const struct nvkm_oproxy_func
@@ -56,8 +56,8 @@ nv50_disp_dmac_child_new_(struct nv50_disp_chan *base,
 			  void *data, u32 size, struct nvkm_object **pobject)
 {
 	struct nv50_disp_dmac *chan = nv50_disp_dmac(base);
-	struct nv50_disp_root *root = chan->base.root;
-	struct nvkm_device *device = root->disp->base.engine.subdev.device;
+	struct nv50_disp *disp = chan->base.root->disp;
+	struct nvkm_device *device = disp->base.engine.subdev.device;
 	const struct nvkm_device_oclass *sclass = oclass->priv;
 	struct nv50_disp_dmac_object *object;
 	int ret;
@@ -65,7 +65,7 @@ nv50_disp_dmac_child_new_(struct nv50_disp_chan *base,
 	if (!(object = kzalloc(sizeof(*object), GFP_KERNEL)))
 		return -ENOMEM;
 	nvkm_oproxy_ctor(&nv50_disp_dmac_child_func_, oclass, &object->oproxy);
-	object->root = root;
+	object->disp = disp;
 	*pobject = &object->oproxy.base;
 
 	ret = sclass->ctor(device, oclass, data, size, &object->oproxy.object);
@@ -177,7 +177,7 @@ int
 nv50_disp_dmac_bind(struct nv50_disp_dmac *chan,
 		    struct nvkm_object *object, u32 handle)
 {
-	return nvkm_ramht_insert(chan->base.root->ramht, object,
+	return nvkm_ramht_insert(chan->base.root->disp->ramht, object,
 				 chan->base.chid.user, -10, handle,
 				 chan->base.chid.user << 28 |
 				 chan->base.chid.user);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c
index 1ec81f3..731f188 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c
@@ -28,6 +28,8 @@
 
 static const struct nv50_disp_func
 g84_disp = {
+	.init = nv50_disp_init,
+	.fini = nv50_disp_fini,
 	.intr = nv50_disp_intr,
 	.uevent = &nv50_disp_chan_uevent,
 	.super = nv50_disp_super,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c
index 791c2cd..def54fe 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c
@@ -28,6 +28,8 @@
 
 static const struct nv50_disp_func
 g94_disp = {
+	.init = nv50_disp_init,
+	.fini = nv50_disp_fini,
 	.intr = nv50_disp_intr,
 	.uevent = &nv50_disp_chan_uevent,
 	.super = nv50_disp_super,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c
index 0139d14..382e6a6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c
@@ -26,6 +26,9 @@
 #include "ior.h"
 #include "rootnv50.h"
 
+#include <core/ramht.h>
+#include <subdev/timer.h>
+
 void
 gf119_disp_super(struct work_struct *work)
 {
@@ -164,8 +167,87 @@ gf119_disp_intr(struct nv50_disp *disp)
 	}
 }
 
+void
+gf119_disp_fini(struct nv50_disp *disp)
+{
+	struct nvkm_device *device = disp->base.engine.subdev.device;
+	/* disable all interrupts */
+	nvkm_wr32(device, 0x6100b0, 0x00000000);
+}
+
+int
+gf119_disp_init(struct nv50_disp *disp)
+{
+	struct nvkm_device *device = disp->base.engine.subdev.device;
+	struct nvkm_head *head;
+	u32 tmp;
+	int i;
+
+	/* The below segments of code copying values from one register to
+	 * another appear to inform EVO of the display capabilities or
+	 * something similar.
+	 */
+
+	/* ... CRTC caps */
+	list_for_each_entry(head, &disp->base.head, head) {
+		const u32 hoff = head->id * 0x800;
+		tmp = nvkm_rd32(device, 0x616104 + hoff);
+		nvkm_wr32(device, 0x6101b4 + hoff, tmp);
+		tmp = nvkm_rd32(device, 0x616108 + hoff);
+		nvkm_wr32(device, 0x6101b8 + hoff, tmp);
+		tmp = nvkm_rd32(device, 0x61610c + hoff);
+		nvkm_wr32(device, 0x6101bc + hoff, tmp);
+	}
+
+	/* ... DAC caps */
+	for (i = 0; i < disp->dac.nr; i++) {
+		tmp = nvkm_rd32(device, 0x61a000 + (i * 0x800));
+		nvkm_wr32(device, 0x6101c0 + (i * 0x800), tmp);
+	}
+
+	/* ... SOR caps */
+	for (i = 0; i < disp->sor.nr; i++) {
+		tmp = nvkm_rd32(device, 0x61c000 + (i * 0x800));
+		nvkm_wr32(device, 0x6301c4 + (i * 0x800), tmp);
+	}
+
+	/* steal display away from vbios, or something like that */
+	if (nvkm_rd32(device, 0x6100ac) & 0x00000100) {
+		nvkm_wr32(device, 0x6100ac, 0x00000100);
+		nvkm_mask(device, 0x6194e8, 0x00000001, 0x00000000);
+		if (nvkm_msec(device, 2000,
+			if (!(nvkm_rd32(device, 0x6194e8) & 0x00000002))
+				break;
+		) < 0)
+			return -EBUSY;
+	}
+
+	/* point at display engine memory area (hash table, objects) */
+	nvkm_wr32(device, 0x610010, (disp->inst->addr >> 8) | 9);
+
+	/* enable supervisor interrupts, disable everything else */
+	nvkm_wr32(device, 0x610090, 0x00000000);
+	nvkm_wr32(device, 0x6100a0, 0x00000000);
+	nvkm_wr32(device, 0x6100b0, 0x00000307);
+
+	/* disable underflow reporting, preventing an intermittent issue
+	 * on some gk104 boards where the production vbios left this
+	 * setting enabled by default.
+	 *
+	 * ftp://download.nvidia.com/open-gpu-doc/gk104-disable-underflow-reporting/1/gk104-disable-underflow-reporting.txt
+	 */
+	list_for_each_entry(head, &disp->base.head, head) {
+		const u32 hoff = head->id * 0x800;
+		nvkm_mask(device, 0x616308 + hoff, 0x00000111, 0x00000010);
+	}
+
+	return 0;
+}
+
 static const struct nv50_disp_func
 gf119_disp = {
+	.init = gf119_disp_init,
+	.fini = gf119_disp_fini,
 	.intr = gf119_disp_intr,
 	.intr_error = gf119_disp_intr_error,
 	.uevent = &gf119_disp_chan_uevent,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c
index 6a59a52..4c3439b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c
@@ -28,6 +28,8 @@
 
 static const struct nv50_disp_func
 gk104_disp = {
+	.init = gf119_disp_init,
+	.fini = gf119_disp_fini,
 	.intr = gf119_disp_intr,
 	.intr_error = gf119_disp_intr_error,
 	.uevent = &gf119_disp_chan_uevent,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c
index f3b10dc4..bc6f475 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c
@@ -28,6 +28,8 @@
 
 static const struct nv50_disp_func
 gk110_disp = {
+	.init = gf119_disp_init,
+	.fini = gf119_disp_fini,
 	.intr = gf119_disp_intr,
 	.intr_error = gf119_disp_intr_error,
 	.uevent = &gf119_disp_chan_uevent,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c
index 068c595..031cf6b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c
@@ -28,6 +28,8 @@
 
 static const struct nv50_disp_func
 gm107_disp = {
+	.init = gf119_disp_init,
+	.fini = gf119_disp_fini,
 	.intr = gf119_disp_intr,
 	.intr_error = gf119_disp_intr_error,
 	.uevent = &gf119_disp_chan_uevent,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c
index 1c27dbe..ec9c33a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c
@@ -28,6 +28,8 @@
 
 static const struct nv50_disp_func
 gm200_disp = {
+	.init = gf119_disp_init,
+	.fini = gf119_disp_fini,
 	.intr = gf119_disp_intr,
 	.intr_error = gf119_disp_intr_error,
 	.uevent = &gf119_disp_chan_uevent,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp100.c
index 84933b6..fd62166 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp100.c
@@ -28,6 +28,8 @@
 
 static const struct nv50_disp_func
 gp100_disp = {
+	.init = gf119_disp_init,
+	.fini = gf119_disp_fini,
 	.intr = gf119_disp_intr,
 	.intr_error = gf119_disp_intr_error,
 	.uevent = &gf119_disp_chan_uevent,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp102.c
index b36d926..0a2c5b5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp102.c
@@ -54,6 +54,8 @@ gp102_disp_intr_error(struct nv50_disp *disp, int chid)
 
 static const struct nv50_disp_func
 gp102_disp = {
+	.init = gf119_disp_init,
+	.fini = gf119_disp_fini,
 	.intr = gf119_disp_intr,
 	.intr_error = gp102_disp_intr_error,
 	.uevent = &gf119_disp_chan_uevent,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c
index 62e721d..f801837 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c
@@ -28,6 +28,8 @@
 
 static const struct nv50_disp_func
 gt200_disp = {
+	.init = nv50_disp_init,
+	.fini = nv50_disp_fini,
 	.intr = nv50_disp_intr,
 	.uevent = &nv50_disp_chan_uevent,
 	.super = nv50_disp_super,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c
index a5b1b14..7581efc 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c
@@ -28,6 +28,8 @@
 
 static const struct nv50_disp_func
 gt215_disp = {
+	.init = nv50_disp_init,
+	.fini = nv50_disp_fini,
 	.intr = nv50_disp_intr,
 	.uevent = &nv50_disp_chan_uevent,
 	.super = nv50_disp_super,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c
index ff49040..cfdce23 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c
@@ -26,6 +26,8 @@
 
 static const struct nv50_disp_func
 mcp77_disp = {
+	.init = nv50_disp_init,
+	.fini = nv50_disp_fini,
 	.intr = nv50_disp_intr,
 	.uevent = &nv50_disp_chan_uevent,
 	.super = nv50_disp_super,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c
index 0cf968d..85d9329 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c
@@ -26,6 +26,8 @@
 
 static const struct nv50_disp_func
 mcp89_disp = {
+	.init = nv50_disp_init,
+	.fini = nv50_disp_fini,
 	.intr = nv50_disp_intr,
 	.uevent = &nv50_disp_chan_uevent,
 	.super = nv50_disp_super,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
index c0faa39..1d2280a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
@@ -28,7 +28,7 @@
 
 #include <core/client.h>
 #include <core/enum.h>
-#include <core/gpuobj.h>
+#include <core/ramht.h>
 #include <subdev/bios.h>
 #include <subdev/bios/disp.h>
 #include <subdev/bios/init.h>
@@ -49,13 +49,32 @@ nv50_disp_intr_(struct nvkm_disp *base)
 	disp->func->intr(disp);
 }
 
+static void
+nv50_disp_fini_(struct nvkm_disp *base)
+{
+	struct nv50_disp *disp = nv50_disp(base);
+	disp->func->fini(disp);
+}
+
+static int
+nv50_disp_init_(struct nvkm_disp *base)
+{
+	struct nv50_disp *disp = nv50_disp(base);
+	return disp->func->init(disp);
+}
+
 static void *
 nv50_disp_dtor_(struct nvkm_disp *base)
 {
 	struct nv50_disp *disp = nv50_disp(base);
+
+	nvkm_ramht_del(&disp->ramht);
+	nvkm_gpuobj_del(&disp->inst);
+
 	nvkm_event_fini(&disp->uevent);
 	if (disp->wq)
 		destroy_workqueue(disp->wq);
+
 	return disp;
 }
 
@@ -65,6 +84,7 @@ nv50_disp_oneinit_(struct nvkm_disp *base)
 	struct nv50_disp *disp = nv50_disp(base);
 	const struct nv50_disp_func *func = disp->func;
 	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
 	int ret, i;
 
 	disp->head.nr = func->head.cnt(&disp->base, &disp->head.mask);
@@ -107,13 +127,20 @@ nv50_disp_oneinit_(struct nvkm_disp *base)
 			return ret;
 	}
 
-	return 0;
+	ret = nvkm_gpuobj_new(device, 0x10000, 0x10000, false, NULL,
+			      &disp->inst);
+	if (ret)
+		return ret;
+
+	return nvkm_ramht_new(device, 0x1000, 0, disp->inst, &disp->ramht);
 }
 
 static const struct nvkm_disp_func
 nv50_disp_ = {
 	.dtor = nv50_disp_dtor_,
 	.oneinit = nv50_disp_oneinit_,
+	.init = nv50_disp_init_,
+	.fini = nv50_disp_fini_,
 	.intr = nv50_disp_intr_,
 	.root = nv50_disp_root_,
 };
@@ -643,8 +670,84 @@ nv50_disp_intr(struct nv50_disp *disp)
 	}
 }
 
+void
+nv50_disp_fini(struct nv50_disp *disp)
+{
+	struct nvkm_device *device = disp->base.engine.subdev.device;
+	/* disable all interrupts */
+	nvkm_wr32(device, 0x610024, 0x00000000);
+	nvkm_wr32(device, 0x610020, 0x00000000);
+}
+
+int
+nv50_disp_init(struct nv50_disp *disp)
+{
+	struct nvkm_device *device = disp->base.engine.subdev.device;
+	struct nvkm_head *head;
+	u32 tmp;
+	int i;
+
+	/* The below segments of code copying values from one register to
+	 * another appear to inform EVO of the display capabilities or
+	 * something similar.  NFI what the 0x614004 caps are for..
+	 */
+	tmp = nvkm_rd32(device, 0x614004);
+	nvkm_wr32(device, 0x610184, tmp);
+
+	/* ... CRTC caps */
+	list_for_each_entry(head, &disp->base.head, head) {
+		tmp = nvkm_rd32(device, 0x616100 + (head->id * 0x800));
+		nvkm_wr32(device, 0x610190 + (head->id * 0x10), tmp);
+		tmp = nvkm_rd32(device, 0x616104 + (head->id * 0x800));
+		nvkm_wr32(device, 0x610194 + (head->id * 0x10), tmp);
+		tmp = nvkm_rd32(device, 0x616108 + (head->id * 0x800));
+		nvkm_wr32(device, 0x610198 + (head->id * 0x10), tmp);
+		tmp = nvkm_rd32(device, 0x61610c + (head->id * 0x800));
+		nvkm_wr32(device, 0x61019c + (head->id * 0x10), tmp);
+	}
+
+	/* ... DAC caps */
+	for (i = 0; i < disp->dac.nr; i++) {
+		tmp = nvkm_rd32(device, 0x61a000 + (i * 0x800));
+		nvkm_wr32(device, 0x6101d0 + (i * 0x04), tmp);
+	}
+
+	/* ... SOR caps */
+	for (i = 0; i < disp->sor.nr; i++) {
+		tmp = nvkm_rd32(device, 0x61c000 + (i * 0x800));
+		nvkm_wr32(device, 0x6101e0 + (i * 0x04), tmp);
+	}
+
+	/* ... PIOR caps */
+	for (i = 0; i < disp->pior.nr; i++) {
+		tmp = nvkm_rd32(device, 0x61e000 + (i * 0x800));
+		nvkm_wr32(device, 0x6101f0 + (i * 0x04), tmp);
+	}
+
+	/* steal display away from vbios, or something like that */
+	if (nvkm_rd32(device, 0x610024) & 0x00000100) {
+		nvkm_wr32(device, 0x610024, 0x00000100);
+		nvkm_mask(device, 0x6194e8, 0x00000001, 0x00000000);
+		if (nvkm_msec(device, 2000,
+			if (!(nvkm_rd32(device, 0x6194e8) & 0x00000002))
+				break;
+		) < 0)
+			return -EBUSY;
+	}
+
+	/* point at display engine memory area (hash table, objects) */
+	nvkm_wr32(device, 0x610010, (disp->inst->addr >> 8) | 9);
+
+	/* enable supervisor interrupts, disable everything else */
+	nvkm_wr32(device, 0x61002c, 0x00000370);
+	nvkm_wr32(device, 0x610028, 0x00000000);
+	return 0;
+}
+
 static const struct nv50_disp_func
 nv50_disp = {
+	.init = nv50_disp_init,
+	.fini = nv50_disp_fini,
 	.intr = nv50_disp_intr,
 	.uevent = &nv50_disp_chan_uevent,
 	.super = nv50_disp_super,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
index a29bcf7..bb622d0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
@@ -32,6 +32,9 @@ struct nv50_disp {
 		u8 type[3];
 	} pior;
 
+	struct nvkm_gpuobj *inst;
+	struct nvkm_ramht *ramht;
+
 	struct nv50_disp_chan *chan[21];
 };
 
@@ -46,6 +49,8 @@ int nv50_disp_new_(const struct nv50_disp_func *, struct nvkm_device *,
 		   int index, struct nvkm_disp **);
 
 struct nv50_disp_func {
+	int (*init)(struct nv50_disp *);
+	void (*fini)(struct nv50_disp *);
 	void (*intr)(struct nv50_disp *);
 	void (*intr_error)(struct nv50_disp *, int chid);
 
@@ -60,9 +65,13 @@ struct nv50_disp_func {
 	} head, dac, sor, pior;
 };
 
+int nv50_disp_init(struct nv50_disp *);
+void nv50_disp_fini(struct nv50_disp *);
 void nv50_disp_intr(struct nv50_disp *);
 void nv50_disp_super(struct work_struct *);
 
+int gf119_disp_init(struct nv50_disp *);
+void gf119_disp_fini(struct nv50_disp *);
 void gf119_disp_intr(struct nv50_disp *);
 void gf119_disp_super(struct work_struct *);
 void gf119_disp_intr_error(struct nv50_disp *, int);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/priv.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/priv.h
index c614351..ef66c5f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/priv.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/priv.h
@@ -13,6 +13,8 @@ void nvkm_disp_vblank(struct nvkm_disp *, int head);
 struct nvkm_disp_func {
 	void *(*dtor)(struct nvkm_disp *);
 	int (*oneinit)(struct nvkm_disp *);
+	int (*init)(struct nvkm_disp *);
+	void (*fini)(struct nvkm_disp *);
 	void (*intr)(struct nvkm_disp *);
 
 	const struct nvkm_disp_oclass *(*root)(struct nvkm_disp *);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c
index 721e4f7..36ac0d42 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c
@@ -28,8 +28,6 @@
 
 static const struct nv50_disp_root_func
 g84_disp_root = {
-	.init = nv50_disp_root_init,
-	.fini = nv50_disp_root_fini,
 	.dmac = {
 		&g84_disp_core_oclass,
 		&g84_disp_base_oclass,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c
index 9493f6e..18b87b3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c
@@ -28,8 +28,6 @@
 
 static const struct nv50_disp_root_func
 g94_disp_root = {
-	.init = nv50_disp_root_init,
-	.fini = nv50_disp_root_fini,
 	.dmac = {
 		&g94_disp_core_oclass,
 		&gt200_disp_base_oclass,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
index 4ba2d80..7c5701f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
@@ -22,96 +22,12 @@
  * Authors: Ben Skeggs
  */
 #include "rootnv50.h"
-#include "head.h"
 #include "dmacnv50.h"
 
-#include <core/ramht.h>
-#include <subdev/timer.h>
-
 #include <nvif/class.h>
 
-void
-gf119_disp_root_fini(struct nv50_disp_root *root)
-{
-	struct nvkm_device *device = root->disp->base.engine.subdev.device;
-	/* disable all interrupts */
-	nvkm_wr32(device, 0x6100b0, 0x00000000);
-}
-
-int
-gf119_disp_root_init(struct nv50_disp_root *root)
-{
-	struct nv50_disp *disp = root->disp;
-	struct nvkm_head *head;
-	struct nvkm_device *device = disp->base.engine.subdev.device;
-	u32 tmp;
-	int i;
-
-	/* The below segments of code copying values from one register to
-	 * another appear to inform EVO of the display capabilities or
-	 * something similar.
-	 */
-
-	/* ... CRTC caps */
-	list_for_each_entry(head, &disp->base.head, head) {
-		const u32 hoff = head->id * 0x800;
-		tmp = nvkm_rd32(device, 0x616104 + hoff);
-		nvkm_wr32(device, 0x6101b4 + hoff, tmp);
-		tmp = nvkm_rd32(device, 0x616108 + hoff);
-		nvkm_wr32(device, 0x6101b8 + hoff, tmp);
-		tmp = nvkm_rd32(device, 0x61610c + hoff);
-		nvkm_wr32(device, 0x6101bc + hoff, tmp);
-	}
-
-	/* ... DAC caps */
-	for (i = 0; i < disp->dac.nr; i++) {
-		tmp = nvkm_rd32(device, 0x61a000 + (i * 0x800));
-		nvkm_wr32(device, 0x6101c0 + (i * 0x800), tmp);
-	}
-
-	/* ... SOR caps */
-	for (i = 0; i < disp->sor.nr; i++) {
-		tmp = nvkm_rd32(device, 0x61c000 + (i * 0x800));
-		nvkm_wr32(device, 0x6301c4 + (i * 0x800), tmp);
-	}
-
-	/* steal display away from vbios, or something like that */
-	if (nvkm_rd32(device, 0x6100ac) & 0x00000100) {
-		nvkm_wr32(device, 0x6100ac, 0x00000100);
-		nvkm_mask(device, 0x6194e8, 0x00000001, 0x00000000);
-		if (nvkm_msec(device, 2000,
-			if (!(nvkm_rd32(device, 0x6194e8) & 0x00000002))
-				break;
-		) < 0)
-			return -EBUSY;
-	}
-
-	/* point at display engine memory area (hash table, objects) */
-	nvkm_wr32(device, 0x610010, (root->instmem->addr >> 8) | 9);
-
-	/* enable supervisor interrupts, disable everything else */
-	nvkm_wr32(device, 0x610090, 0x00000000);
-	nvkm_wr32(device, 0x6100a0, 0x00000000);
-	nvkm_wr32(device, 0x6100b0, 0x00000307);
-
-	/* disable underflow reporting, preventing an intermittent issue
-	 * on some gk104 boards where the production vbios left this
-	 * setting enabled by default.
-	 *
-	 * ftp://download.nvidia.com/open-gpu-doc/gk104-disable-underflow-reporting/1/gk104-disable-underflow-reporting.txt
-	 */
-	list_for_each_entry(head, &disp->base.head, head) {
-		const u32 hoff = head->id * 0x800;
-		nvkm_mask(device, 0x616308 + hoff, 0x00000111, 0x00000010);
-	}
-
-	return 0;
-}
-
 static const struct nv50_disp_root_func
 gf119_disp_root = {
-	.init = gf119_disp_root_init,
-	.fini = gf119_disp_root_fini,
 	.dmac = {
 		&gf119_disp_core_oclass,
 		&gf119_disp_base_oclass,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c
index 0bfdb1d..c0946a6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c
@@ -28,8 +28,6 @@
 
 static const struct nv50_disp_root_func
 gk104_disp_root = {
-	.init = gf119_disp_root_init,
-	.fini = gf119_disp_root_fini,
 	.dmac = {
 		&gk104_disp_core_oclass,
 		&gk104_disp_base_oclass,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c
index 1e8dbed..2ebc166 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c
@@ -28,8 +28,6 @@
 
 static const struct nv50_disp_root_func
 gk110_disp_root = {
-	.init = gf119_disp_root_init,
-	.fini = gf119_disp_root_fini,
 	.dmac = {
 		&gk110_disp_core_oclass,
 		&gk110_disp_base_oclass,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c
index 44c55be..5a62c9e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c
@@ -28,8 +28,6 @@
 
 static const struct nv50_disp_root_func
 gm107_disp_root = {
-	.init = gf119_disp_root_init,
-	.fini = gf119_disp_root_fini,
 	.dmac = {
 		&gm107_disp_core_oclass,
 		&gk110_disp_base_oclass,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c
index 38f5ee1..2634e06 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c
@@ -28,8 +28,6 @@
 
 static const struct nv50_disp_root_func
 gm200_disp_root = {
-	.init = gf119_disp_root_init,
-	.fini = gf119_disp_root_fini,
 	.dmac = {
 		&gm200_disp_core_oclass,
 		&gk110_disp_base_oclass,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c
index ac8fdd7..7847235 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c
@@ -28,8 +28,6 @@
 
 static const struct nv50_disp_root_func
 gp100_disp_root = {
-	.init = gf119_disp_root_init,
-	.fini = gf119_disp_root_fini,
 	.dmac = {
 		&gp100_disp_core_oclass,
 		&gk110_disp_base_oclass,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c
index 37122ca..2fdfa8d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c
@@ -28,8 +28,6 @@
 
 static const struct nv50_disp_root_func
 gp102_disp_root = {
-	.init = gf119_disp_root_init,
-	.fini = gf119_disp_root_fini,
 	.dmac = {
 		&gp102_disp_core_oclass,
 		&gp102_disp_base_oclass,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c
index 124a0c2..facad27 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c
@@ -28,8 +28,6 @@
 
 static const struct nv50_disp_root_func
 gt200_disp_root = {
-	.init = nv50_disp_root_init,
-	.fini = nv50_disp_root_fini,
 	.dmac = {
 		&gt200_disp_core_oclass,
 		&gt200_disp_base_oclass,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c
index dff52f3..3e93db5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c
@@ -28,8 +28,6 @@
 
 static const struct nv50_disp_root_func
 gt215_disp_root = {
-	.init = nv50_disp_root_init,
-	.fini = nv50_disp_root_fini,
 	.dmac = {
 		&gt215_disp_core_oclass,
 		&gt215_disp_base_oclass,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
index c8379bf..072c8c0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
@@ -28,8 +28,6 @@
 #include "ior.h"
 
 #include <core/client.h>
-#include <core/ramht.h>
-#include <subdev/timer.h>
 
 #include <nvif/class.h>
 #include <nvif/cl5070.h>
@@ -315,49 +313,16 @@ nv50_disp_root_child_get_(struct nvkm_object *object, int index,
 	return -EINVAL;
 }
 
-static int
-nv50_disp_root_fini_(struct nvkm_object *object, bool suspend)
-{
-	struct nv50_disp_root *root = nv50_disp_root(object);
-	root->func->fini(root);
-	return 0;
-}
-
-static int
-nv50_disp_root_init_(struct nvkm_object *object)
-{
-	struct nv50_disp_root *root = nv50_disp_root(object);
-	struct nvkm_ior *ior;
-	int ret;
-
-	ret = root->func->init(root);
-	if (ret)
-		return ret;
-
-	/* Set 'normal' (ie. when it's attached to a head) state for
-	 * each output resource to 'fully enabled'.
-	 */
-	list_for_each_entry(ior, &root->disp->base.ior, head) {
-		ior->func->power(ior, true, true, true, true, true);
-	}
-
-	return 0;
-}
-
 static void *
 nv50_disp_root_dtor_(struct nvkm_object *object)
 {
 	struct nv50_disp_root *root = nv50_disp_root(object);
-	nvkm_ramht_del(&root->ramht);
-	nvkm_gpuobj_del(&root->instmem);
 	return root;
 }
 
 static const struct nvkm_object_func
 nv50_disp_root_ = {
 	.dtor = nv50_disp_root_dtor_,
-	.init = nv50_disp_root_init_,
-	.fini = nv50_disp_root_fini_,
 	.mthd = nv50_disp_root_mthd_,
 	.ntfy = nvkm_disp_ntfy,
 	.sclass = nv50_disp_root_child_get_,
@@ -370,8 +335,6 @@ nv50_disp_root_new_(const struct nv50_disp_root_func *func,
 {
 	struct nv50_disp *disp = nv50_disp(base);
 	struct nv50_disp_root *root;
-	struct nvkm_device *device = disp->base.engine.subdev.device;
-	int ret;
 
 	if (!(root = kzalloc(sizeof(*root), GFP_KERNEL)))
 		return -ENOMEM;
@@ -380,94 +343,11 @@ nv50_disp_root_new_(const struct nv50_disp_root_func *func,
 	nvkm_object_ctor(&nv50_disp_root_, oclass, &root->object);
 	root->func = func;
 	root->disp = disp;
-
-	ret = nvkm_gpuobj_new(disp->base.engine.subdev.device, 0x10000, 0x10000,
-			      false, NULL, &root->instmem);
-	if (ret)
-		return ret;
-
-	return nvkm_ramht_new(device, 0x1000, 0, root->instmem, &root->ramht);
-}
-
-void
-nv50_disp_root_fini(struct nv50_disp_root *root)
-{
-	struct nvkm_device *device = root->disp->base.engine.subdev.device;
-	/* disable all interrupts */
-	nvkm_wr32(device, 0x610024, 0x00000000);
-	nvkm_wr32(device, 0x610020, 0x00000000);
-}
-
-int
-nv50_disp_root_init(struct nv50_disp_root *root)
-{
-	struct nv50_disp *disp = root->disp;
-	struct nvkm_head *head;
-	struct nvkm_device *device = disp->base.engine.subdev.device;
-	u32 tmp;
-	int i;
-
-	/* The below segments of code copying values from one register to
-	 * another appear to inform EVO of the display capabilities or
-	 * something similar.  NFI what the 0x614004 caps are for..
-	 */
-	tmp = nvkm_rd32(device, 0x614004);
-	nvkm_wr32(device, 0x610184, tmp);
-
-	/* ... CRTC caps */
-	list_for_each_entry(head, &disp->base.head, head) {
-		tmp = nvkm_rd32(device, 0x616100 + (head->id * 0x800));
-		nvkm_wr32(device, 0x610190 + (head->id * 0x10), tmp);
-		tmp = nvkm_rd32(device, 0x616104 + (head->id * 0x800));
-		nvkm_wr32(device, 0x610194 + (head->id * 0x10), tmp);
-		tmp = nvkm_rd32(device, 0x616108 + (head->id * 0x800));
-		nvkm_wr32(device, 0x610198 + (head->id * 0x10), tmp);
-		tmp = nvkm_rd32(device, 0x61610c + (head->id * 0x800));
-		nvkm_wr32(device, 0x61019c + (head->id * 0x10), tmp);
-	}
-
-	/* ... DAC caps */
-	for (i = 0; i < disp->dac.nr; i++) {
-		tmp = nvkm_rd32(device, 0x61a000 + (i * 0x800));
-		nvkm_wr32(device, 0x6101d0 + (i * 0x04), tmp);
-	}
-
-	/* ... SOR caps */
-	for (i = 0; i < disp->sor.nr; i++) {
-		tmp = nvkm_rd32(device, 0x61c000 + (i * 0x800));
-		nvkm_wr32(device, 0x6101e0 + (i * 0x04), tmp);
-	}
-
-	/* ... PIOR caps */
-	for (i = 0; i < disp->pior.nr; i++) {
-		tmp = nvkm_rd32(device, 0x61e000 + (i * 0x800));
-		nvkm_wr32(device, 0x6101f0 + (i * 0x04), tmp);
-	}
-
-	/* steal display away from vbios, or something like that */
-	if (nvkm_rd32(device, 0x610024) & 0x00000100) {
-		nvkm_wr32(device, 0x610024, 0x00000100);
-		nvkm_mask(device, 0x6194e8, 0x00000001, 0x00000000);
-		if (nvkm_msec(device, 2000,
-			if (!(nvkm_rd32(device, 0x6194e8) & 0x00000002))
-				break;
-		) < 0)
-			return -EBUSY;
-	}
-
-	/* point at display engine memory area (hash table, objects) */
-	nvkm_wr32(device, 0x610010, (root->instmem->addr >> 8) | 9);
-
-	/* enable supervisor interrupts, disable everything else */
-	nvkm_wr32(device, 0x61002c, 0x00000370);
-	nvkm_wr32(device, 0x610028, 0x00000000);
 	return 0;
 }
 
 static const struct nv50_disp_root_func
 nv50_disp_root = {
-	.init = nv50_disp_root_init,
-	.fini = nv50_disp_root_fini,
 	.dmac = {
 		&nv50_disp_core_oclass,
 		&nv50_disp_base_oclass,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
index 4818fa6..06b554b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
@@ -10,14 +10,9 @@ struct nv50_disp_root {
 	const struct nv50_disp_root_func *func;
 	struct nv50_disp *disp;
 	struct nvkm_object object;
-
-	struct nvkm_gpuobj *instmem;
-	struct nvkm_ramht *ramht;
 };
 
 struct nv50_disp_root_func {
-	int (*init)(struct nv50_disp_root *);
-	void (*fini)(struct nv50_disp_root *);
 	const struct nv50_disp_dmac_oclass *dmac[3];
 	const struct nv50_disp_pioc_oclass *pioc[2];
 };
@@ -25,11 +20,6 @@ struct nv50_disp_root_func {
 int  nv50_disp_root_new_(const struct nv50_disp_root_func *, struct nvkm_disp *,
 			 const struct nvkm_oclass *, void *data, u32 size,
 			 struct nvkm_object **);
-int  nv50_disp_root_init(struct nv50_disp_root *);
-void nv50_disp_root_fini(struct nv50_disp_root *);
-
-int  gf119_disp_root_init(struct nv50_disp_root *);
-void gf119_disp_root_fini(struct nv50_disp_root *);
 
 extern const struct nvkm_disp_oclass nv50_disp_root_oclass;
 extern const struct nvkm_disp_oclass g84_disp_root_oclass;
-- 
cgit v1.1


From abc1d4379bafc504b05039db2336b3955b17ffdb Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/disp/nv50-: replace user object with engine pointer in
 channels

More simplification.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 .../gpu/drm/nouveau/nvkm/engine/disp/basenv50.c    |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/channv50.c    | 22 ++++++++++------------
 .../gpu/drm/nouveau/nvkm/engine/disp/channv50.h    |  6 +++---
 .../gpu/drm/nouveau/nvkm/engine/disp/coregf119.c   |  4 ++--
 .../gpu/drm/nouveau/nvkm/engine/disp/coregp102.c   |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/corenv50.c    |  6 +++---
 .../gpu/drm/nouveau/nvkm/engine/disp/cursnv50.c    |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c   |  6 +++---
 .../gpu/drm/nouveau/nvkm/engine/disp/dmacgp102.c   |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c    | 14 +++++++-------
 .../gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h    |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/oimmnv50.c    |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/ovlynv50.c    |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/piocgf119.c   |  4 ++--
 .../gpu/drm/nouveau/nvkm/engine/disp/piocnv50.c    |  4 ++--
 15 files changed, 39 insertions(+), 41 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basenv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basenv50.c
index f1d6b82..418741a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basenv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basenv50.c
@@ -58,7 +58,7 @@ nv50_disp_base_new(const struct nv50_disp_dmac_func *func,
 	} else
 		return ret;
 
-	return nv50_disp_dmac_new_(func, mthd, root, chid + head,
+	return nv50_disp_dmac_new_(func, mthd, disp, chid + head,
 				   head, push, oclass, pobject);
 }
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c
index 723dcbd..96b732d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c
@@ -65,7 +65,7 @@ nv50_disp_mthd_list(struct nv50_disp *disp, int debug, u32 base, int c,
 void
 nv50_disp_chan_mthd(struct nv50_disp_chan *chan, int debug)
 {
-	struct nv50_disp *disp = chan->root->disp;
+	struct nv50_disp *disp = chan->disp;
 	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
 	const struct nv50_disp_chan_mthd *mthd = chan->mthd;
 	const struct nv50_disp_mthd_list *list;
@@ -158,7 +158,7 @@ static int
 nv50_disp_chan_rd32(struct nvkm_object *object, u64 addr, u32 *data)
 {
 	struct nv50_disp_chan *chan = nv50_disp_chan(object);
-	struct nv50_disp *disp = chan->root->disp;
+	struct nv50_disp *disp = chan->disp;
 	struct nvkm_device *device = disp->base.engine.subdev.device;
 	*data = nvkm_rd32(device, 0x640000 + (chan->chid.user * 0x1000) + addr);
 	return 0;
@@ -168,7 +168,7 @@ static int
 nv50_disp_chan_wr32(struct nvkm_object *object, u64 addr, u32 data)
 {
 	struct nv50_disp_chan *chan = nv50_disp_chan(object);
-	struct nv50_disp *disp = chan->root->disp;
+	struct nv50_disp *disp = chan->disp;
 	struct nvkm_device *device = disp->base.engine.subdev.device;
 	nvkm_wr32(device, 0x640000 + (chan->chid.user * 0x1000) + addr, data);
 	return 0;
@@ -179,7 +179,7 @@ nv50_disp_chan_ntfy(struct nvkm_object *object, u32 type,
 		    struct nvkm_event **pevent)
 {
 	struct nv50_disp_chan *chan = nv50_disp_chan(object);
-	struct nv50_disp *disp = chan->root->disp;
+	struct nv50_disp *disp = chan->disp;
 	switch (type) {
 	case NV50_DISP_CORE_CHANNEL_DMA_V0_NTFY_UEVENT:
 		*pevent = &disp->uevent;
@@ -195,7 +195,7 @@ nv50_disp_chan_map(struct nvkm_object *object, void *argv, u32 argc,
 		   enum nvkm_object_map *type, u64 *addr, u64 *size)
 {
 	struct nv50_disp_chan *chan = nv50_disp_chan(object);
-	struct nv50_disp *disp = chan->root->disp;
+	struct nv50_disp *disp = chan->disp;
 	struct nvkm_device *device = disp->base.engine.subdev.device;
 	*type = NVKM_OBJECT_MAP_IO;
 	*addr = device->func->resource_addr(device, 0) +
@@ -245,7 +245,7 @@ static void *
 nv50_disp_chan_dtor(struct nvkm_object *object)
 {
 	struct nv50_disp_chan *chan = nv50_disp_chan(object);
-	struct nv50_disp *disp = chan->root->disp;
+	struct nv50_disp *disp = chan->disp;
 	if (chan->chid.user >= 0)
 		disp->chan[chan->chid.user] = NULL;
 	return chan->func->dtor ? chan->func->dtor(chan) : chan;
@@ -266,16 +266,14 @@ nv50_disp_chan = {
 int
 nv50_disp_chan_ctor(const struct nv50_disp_chan_func *func,
 		    const struct nv50_disp_chan_mthd *mthd,
-		    struct nv50_disp_root *root, int ctrl, int user, int head,
+		    struct nv50_disp *disp, int ctrl, int user, int head,
 		    const struct nvkm_oclass *oclass,
 		    struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = root->disp;
-
 	nvkm_object_ctor(&nv50_disp_chan, oclass, &chan->object);
 	chan->func = func;
 	chan->mthd = mthd;
-	chan->root = root;
+	chan->disp = disp;
 	chan->chid.ctrl = ctrl;
 	chan->chid.user = user;
 	chan->head = head;
@@ -291,7 +289,7 @@ nv50_disp_chan_ctor(const struct nv50_disp_chan_func *func,
 int
 nv50_disp_chan_new_(const struct nv50_disp_chan_func *func,
 		    const struct nv50_disp_chan_mthd *mthd,
-		    struct nv50_disp_root *root, int ctrl, int user, int head,
+		    struct nv50_disp *disp, int ctrl, int user, int head,
 		    const struct nvkm_oclass *oclass,
 		    struct nvkm_object **pobject)
 {
@@ -301,6 +299,6 @@ nv50_disp_chan_new_(const struct nv50_disp_chan_func *func,
 		return -ENOMEM;
 	*pobject = &chan->object;
 
-	return nv50_disp_chan_ctor(func, mthd, root, ctrl, user,
+	return nv50_disp_chan_ctor(func, mthd, disp, ctrl, user,
 				   head, oclass, chan);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
index b518585..c9dc58c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
@@ -9,7 +9,7 @@ struct nv50_disp_root;
 struct nv50_disp_chan {
 	const struct nv50_disp_chan_func *func;
 	const struct nv50_disp_chan_mthd *mthd;
-	struct nv50_disp_root *root;
+	struct nv50_disp *disp;
 
 	struct {
 		int ctrl;
@@ -32,11 +32,11 @@ struct nv50_disp_chan_func {
 
 int nv50_disp_chan_ctor(const struct nv50_disp_chan_func *,
 			const struct nv50_disp_chan_mthd *,
-			struct nv50_disp_root *, int ctrl, int user, int head,
+			struct nv50_disp *, int ctrl, int user, int head,
 			const struct nvkm_oclass *, struct nv50_disp_chan *);
 int nv50_disp_chan_new_(const struct nv50_disp_chan_func *,
 			const struct nv50_disp_chan_mthd *,
-			struct nv50_disp_root *, int ctrl, int user, int head,
+			struct nv50_disp *, int ctrl, int user, int head,
 			const struct nvkm_oclass *, struct nvkm_object **);
 
 extern const struct nv50_disp_chan_func nv50_disp_pioc_func;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c
index 21fbf89..200dd90 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c
@@ -174,7 +174,7 @@ gf119_disp_core_chan_mthd = {
 void
 gf119_disp_core_fini(struct nv50_disp_dmac *chan)
 {
-	struct nv50_disp *disp = chan->base.root->disp;
+	struct nv50_disp *disp = chan->base.disp;
 	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 
@@ -197,7 +197,7 @@ gf119_disp_core_fini(struct nv50_disp_dmac *chan)
 static int
 gf119_disp_core_init(struct nv50_disp_dmac *chan)
 {
-	struct nv50_disp *disp = chan->base.root->disp;
+	struct nv50_disp *disp = chan->base.disp;
 	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c
index b0df4b7..6ad5f2f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c
@@ -31,7 +31,7 @@
 static int
 gp102_disp_core_init(struct nv50_disp_dmac *chan)
 {
-	struct nv50_disp *disp = chan->base.root->disp;
+	struct nv50_disp *disp = chan->base.disp;
 	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c
index b547c8b..839cbc8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c
@@ -54,7 +54,7 @@ nv50_disp_core_new(const struct nv50_disp_dmac_func *func,
 	} else
 		return ret;
 
-	return nv50_disp_dmac_new_(func, mthd, root, chid, 0,
+	return nv50_disp_dmac_new_(func, mthd, root->disp, chid, 0,
 				   push, oclass, pobject);
 }
 
@@ -168,7 +168,7 @@ nv50_disp_core_chan_mthd = {
 static void
 nv50_disp_core_fini(struct nv50_disp_dmac *chan)
 {
-	struct nv50_disp *disp = chan->base.root->disp;
+	struct nv50_disp *disp = chan->base.disp;
 	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 
@@ -190,7 +190,7 @@ nv50_disp_core_fini(struct nv50_disp_dmac *chan)
 static int
 nv50_disp_core_init(struct nv50_disp_dmac *chan)
 {
-	struct nv50_disp *disp = chan->base.root->disp;
+	struct nv50_disp *disp = chan->base.disp;
 	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursnv50.c
index ab51121..1f9a6c3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursnv50.c
@@ -55,7 +55,7 @@ nv50_disp_curs_new(const struct nv50_disp_chan_func *func,
 	} else
 		return ret;
 
-	return nv50_disp_chan_new_(func, mthd, root, ctrl + head, user + head,
+	return nv50_disp_chan_new_(func, mthd, disp, ctrl + head, user + head,
 				   head, oclass, pobject);
 }
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c
index 6680ff8..b73bcc3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c
@@ -31,7 +31,7 @@ int
 gf119_disp_dmac_bind(struct nv50_disp_dmac *chan,
 		     struct nvkm_object *object, u32 handle)
 {
-	return nvkm_ramht_insert(chan->base.root->disp->ramht, object,
+	return nvkm_ramht_insert(chan->base.disp->ramht, object,
 				 chan->base.chid.user, -9, handle,
 				 chan->base.chid.user << 27 | 0x00000001);
 }
@@ -39,7 +39,7 @@ gf119_disp_dmac_bind(struct nv50_disp_dmac *chan,
 void
 gf119_disp_dmac_fini(struct nv50_disp_dmac *chan)
 {
-	struct nv50_disp *disp = chan->base.root->disp;
+	struct nv50_disp *disp = chan->base.disp;
 	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 	int ctrl = chan->base.chid.ctrl;
@@ -64,7 +64,7 @@ gf119_disp_dmac_fini(struct nv50_disp_dmac *chan)
 static int
 gf119_disp_dmac_init(struct nv50_disp_dmac *chan)
 {
-	struct nv50_disp *disp = chan->base.root->disp;
+	struct nv50_disp *disp = chan->base.disp;
 	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 	int ctrl = chan->base.chid.ctrl;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgp102.c
index cdead95..62e9b84 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgp102.c
@@ -29,7 +29,7 @@
 static int
 gp102_disp_dmac_init(struct nv50_disp_dmac *chan)
 {
-	struct nv50_disp *disp = chan->base.root->disp;
+	struct nv50_disp *disp = chan->base.disp;
 	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 	int ctrl = chan->base.chid.ctrl;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c
index c80d047..d081947 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c
@@ -56,7 +56,7 @@ nv50_disp_dmac_child_new_(struct nv50_disp_chan *base,
 			  void *data, u32 size, struct nvkm_object **pobject)
 {
 	struct nv50_disp_dmac *chan = nv50_disp_dmac(base);
-	struct nv50_disp *disp = chan->base.root->disp;
+	struct nv50_disp *disp = chan->base.disp;
 	struct nvkm_device *device = disp->base.engine.subdev.device;
 	const struct nvkm_device_oclass *sclass = oclass->priv;
 	struct nv50_disp_dmac_object *object;
@@ -85,7 +85,7 @@ nv50_disp_dmac_child_get_(struct nv50_disp_chan *base, int index,
 			  struct nvkm_oclass *sclass)
 {
 	struct nv50_disp_dmac *chan = nv50_disp_dmac(base);
-	struct nv50_disp *disp = chan->base.root->disp;
+	struct nv50_disp *disp = chan->base.disp;
 	struct nvkm_device *device = disp->base.engine.subdev.device;
 	const struct nvkm_device_oclass *oclass = NULL;
 
@@ -133,7 +133,7 @@ nv50_disp_dmac_func_ = {
 int
 nv50_disp_dmac_new_(const struct nv50_disp_dmac_func *func,
 		    const struct nv50_disp_chan_mthd *mthd,
-		    struct nv50_disp_root *root, int chid, int head, u64 push,
+		    struct nv50_disp *disp, int chid, int head, u64 push,
 		    const struct nvkm_oclass *oclass,
 		    struct nvkm_object **pobject)
 {
@@ -147,7 +147,7 @@ nv50_disp_dmac_new_(const struct nv50_disp_dmac_func *func,
 	*pobject = &chan->base.object;
 	chan->func = func;
 
-	ret = nv50_disp_chan_ctor(&nv50_disp_dmac_func_, mthd, root,
+	ret = nv50_disp_chan_ctor(&nv50_disp_dmac_func_, mthd, disp,
 				  chid, chid, head, oclass, &chan->base);
 	if (ret)
 		return ret;
@@ -177,7 +177,7 @@ int
 nv50_disp_dmac_bind(struct nv50_disp_dmac *chan,
 		    struct nvkm_object *object, u32 handle)
 {
-	return nvkm_ramht_insert(chan->base.root->disp->ramht, object,
+	return nvkm_ramht_insert(chan->base.disp->ramht, object,
 				 chan->base.chid.user, -10, handle,
 				 chan->base.chid.user << 28 |
 				 chan->base.chid.user);
@@ -186,7 +186,7 @@ nv50_disp_dmac_bind(struct nv50_disp_dmac *chan,
 static void
 nv50_disp_dmac_fini(struct nv50_disp_dmac *chan)
 {
-	struct nv50_disp *disp = chan->base.root->disp;
+	struct nv50_disp *disp = chan->base.disp;
 	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 	int ctrl = chan->base.chid.ctrl;
@@ -210,7 +210,7 @@ nv50_disp_dmac_fini(struct nv50_disp_dmac *chan)
 static int
 nv50_disp_dmac_init(struct nv50_disp_dmac *chan)
 {
-	struct nv50_disp *disp = chan->base.root->disp;
+	struct nv50_disp *disp = chan->base.disp;
 	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 	int ctrl = chan->base.chid.ctrl;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h
index f9b9821..ca32386 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h
@@ -18,7 +18,7 @@ struct nv50_disp_dmac_func {
 
 int nv50_disp_dmac_new_(const struct nv50_disp_dmac_func *,
 			const struct nv50_disp_chan_mthd *,
-			struct nv50_disp_root *, int chid, int head, u64 push,
+			struct nv50_disp *, int chid, int head, u64 push,
 			const struct nvkm_oclass *, struct nvkm_object **);
 
 extern const struct nv50_disp_dmac_func nv50_disp_dmac_func;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmnv50.c
index f3b0fa2..7e013cd 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmnv50.c
@@ -55,7 +55,7 @@ nv50_disp_oimm_new(const struct nv50_disp_chan_func *func,
 	} else
 		return ret;
 
-	return nv50_disp_chan_new_(func, mthd, root, ctrl + head, user + head,
+	return nv50_disp_chan_new_(func, mthd, disp, ctrl + head, user + head,
 				   head, oclass, pobject);
 }
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlynv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlynv50.c
index 9ebaaa6..d0e5546 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlynv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlynv50.c
@@ -58,7 +58,7 @@ nv50_disp_ovly_new(const struct nv50_disp_dmac_func *func,
 	} else
 		return ret;
 
-	return nv50_disp_dmac_new_(func, mthd, root, chid + head,
+	return nv50_disp_dmac_new_(func, mthd, disp, chid + head,
 				   head, push, oclass, pobject);
 }
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocgf119.c
index 0abaa64..7b1e9bf 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocgf119.c
@@ -29,7 +29,7 @@
 static void
 gf119_disp_pioc_fini(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->root->disp;
+	struct nv50_disp *disp = chan->disp;
 	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 	int ctrl = chan->chid.ctrl;
@@ -52,7 +52,7 @@ gf119_disp_pioc_fini(struct nv50_disp_chan *chan)
 static int
 gf119_disp_pioc_init(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->root->disp;
+	struct nv50_disp *disp = chan->disp;
 	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 	int ctrl = chan->chid.ctrl;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocnv50.c
index 0211e0e..60c2012 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocnv50.c
@@ -29,7 +29,7 @@
 static void
 nv50_disp_pioc_fini(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->root->disp;
+	struct nv50_disp *disp = chan->disp;
 	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 	int ctrl = chan->chid.ctrl;
@@ -48,7 +48,7 @@ nv50_disp_pioc_fini(struct nv50_disp_chan *chan)
 static int
 nv50_disp_pioc_init(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->root->disp;
+	struct nv50_disp *disp = chan->disp;
 	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 	int ctrl = chan->chid.ctrl;
-- 
cgit v1.1


From 46f74a8ad79c4da47190df8492f0534fe8c02652 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/disp/nv50-: simplify definition of overlay channels

Introduces a new method of defining channels available from the display,
common to all channel types, allowing for more flexibility in available
channel types/counts, and reducing the amount of boiler-plate required.

This will be required to support Volta.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild    |  1 -
 .../gpu/drm/nouveau/nvkm/engine/disp/channv50.h    | 28 ++++++++++++++--
 .../gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h    | 12 -------
 drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlyg84.c | 24 +++++---------
 .../gpu/drm/nouveau/nvkm/engine/disp/ovlygf119.c   | 22 +++++--------
 .../gpu/drm/nouveau/nvkm/engine/disp/ovlygk104.c   | 22 +++++--------
 .../gpu/drm/nouveau/nvkm/engine/disp/ovlygp102.c   | 20 ++++--------
 .../gpu/drm/nouveau/nvkm/engine/disp/ovlygt200.c   | 22 +++++--------
 .../gpu/drm/nouveau/nvkm/engine/disp/ovlygt215.c   | 38 ----------------------
 .../gpu/drm/nouveau/nvkm/engine/disp/ovlynv50.c    | 38 +++++++++-------------
 drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c |  5 ++-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c |  5 ++-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c   |  5 ++-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c   |  5 ++-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c   |  5 ++-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c   |  5 ++-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c   |  5 ++-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c   |  5 ++-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c   |  5 ++-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c   |  5 ++-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c   |  5 ++-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c    | 23 ++++++++++++-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h    |  7 +++-
 23 files changed, 154 insertions(+), 158 deletions(-)
 delete mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygt215.c

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild b/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild
index 48ce669..67861f2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild
@@ -96,7 +96,6 @@ nvkm-y += nvkm/engine/disp/coregp102.o
 nvkm-y += nvkm/engine/disp/ovlynv50.o
 nvkm-y += nvkm/engine/disp/ovlyg84.o
 nvkm-y += nvkm/engine/disp/ovlygt200.o
-nvkm-y += nvkm/engine/disp/ovlygt215.o
 nvkm-y += nvkm/engine/disp/ovlygf119.o
 nvkm-y += nvkm/engine/disp/ovlygk104.o
 nvkm-y += nvkm/engine/disp/ovlygp102.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
index c9dc58c..b222a1d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
@@ -5,6 +5,7 @@
 #include <core/object.h>
 #include "nv50.h"
 struct nv50_disp_root;
+struct nv50_disp_dmac_func;
 
 struct nv50_disp_chan {
 	const struct nv50_disp_chan_func *func;
@@ -49,6 +50,30 @@ void nv50_disp_chan_uevent_send(struct nv50_disp *, int);
 
 extern const struct nvkm_event_func gf119_disp_chan_uevent;
 
+int nv50_disp_ovly_new_(const struct nv50_disp_dmac_func *,
+			const struct nv50_disp_chan_mthd *,
+			struct nv50_disp *, int chid,
+			const struct nvkm_oclass *, void *argv, u32 argc,
+			struct nvkm_object **);
+
+int nv50_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
+		       struct nv50_disp *, struct nvkm_object **);
+
+int g84_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
+		      struct nv50_disp *, struct nvkm_object **);
+
+int gt200_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
+
+int gf119_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
+
+int gk104_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
+
+int gp102_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
+
 struct nv50_disp_mthd_list {
 	u32 mthd;
 	u32 addr;
@@ -81,7 +106,6 @@ extern const struct nv50_disp_chan_mthd g84_disp_core_chan_mthd;
 extern const struct nv50_disp_mthd_list g84_disp_core_mthd_dac;
 extern const struct nv50_disp_mthd_list g84_disp_core_mthd_head;
 extern const struct nv50_disp_chan_mthd g84_disp_base_chan_mthd;
-extern const struct nv50_disp_chan_mthd g84_disp_ovly_chan_mthd;
 
 extern const struct nv50_disp_chan_mthd g94_disp_core_chan_mthd;
 
@@ -92,7 +116,7 @@ extern const struct nv50_disp_mthd_list gf119_disp_core_mthd_pior;
 extern const struct nv50_disp_chan_mthd gf119_disp_base_chan_mthd;
 
 extern const struct nv50_disp_chan_mthd gk104_disp_core_chan_mthd;
-extern const struct nv50_disp_chan_mthd gk104_disp_ovly_chan_mthd;
+extern const struct nv50_disp_chan_mthd gk104_disp_ovly_mthd;
 
 struct nv50_disp_pioc_oclass {
 	int (*ctor)(const struct nv50_disp_chan_func *,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h
index ca32386..45caaee 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h
@@ -55,37 +55,26 @@ int nv50_disp_base_new(const struct nv50_disp_dmac_func *,
 		       struct nv50_disp_root *, int chid,
 		       const struct nvkm_oclass *oclass, void *data, u32 size,
 		       struct nvkm_object **);
-int nv50_disp_ovly_new(const struct nv50_disp_dmac_func *,
-		       const struct nv50_disp_chan_mthd *,
-		       struct nv50_disp_root *, int chid,
-		       const struct nvkm_oclass *oclass, void *data, u32 size,
-		       struct nvkm_object **);
 
 extern const struct nv50_disp_dmac_oclass nv50_disp_core_oclass;
 extern const struct nv50_disp_dmac_oclass nv50_disp_base_oclass;
-extern const struct nv50_disp_dmac_oclass nv50_disp_ovly_oclass;
 
 extern const struct nv50_disp_dmac_oclass g84_disp_core_oclass;
 extern const struct nv50_disp_dmac_oclass g84_disp_base_oclass;
-extern const struct nv50_disp_dmac_oclass g84_disp_ovly_oclass;
 
 extern const struct nv50_disp_dmac_oclass g94_disp_core_oclass;
 
 extern const struct nv50_disp_dmac_oclass gt200_disp_core_oclass;
 extern const struct nv50_disp_dmac_oclass gt200_disp_base_oclass;
-extern const struct nv50_disp_dmac_oclass gt200_disp_ovly_oclass;
 
 extern const struct nv50_disp_dmac_oclass gt215_disp_core_oclass;
 extern const struct nv50_disp_dmac_oclass gt215_disp_base_oclass;
-extern const struct nv50_disp_dmac_oclass gt215_disp_ovly_oclass;
 
 extern const struct nv50_disp_dmac_oclass gf119_disp_core_oclass;
 extern const struct nv50_disp_dmac_oclass gf119_disp_base_oclass;
-extern const struct nv50_disp_dmac_oclass gf119_disp_ovly_oclass;
 
 extern const struct nv50_disp_dmac_oclass gk104_disp_core_oclass;
 extern const struct nv50_disp_dmac_oclass gk104_disp_base_oclass;
-extern const struct nv50_disp_dmac_oclass gk104_disp_ovly_oclass;
 
 extern const struct nv50_disp_dmac_oclass gk110_disp_core_oclass;
 extern const struct nv50_disp_dmac_oclass gk110_disp_base_oclass;
@@ -98,5 +87,4 @@ extern const struct nv50_disp_dmac_oclass gp100_disp_core_oclass;
 
 extern const struct nv50_disp_dmac_oclass gp102_disp_core_oclass;
 extern const struct nv50_disp_dmac_oclass gp102_disp_base_oclass;
-extern const struct nv50_disp_dmac_oclass gp102_disp_ovly_oclass;
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlyg84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlyg84.c
index db6234e..6b55cf4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlyg84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlyg84.c
@@ -22,9 +22,6 @@
  * Authors: Ben Skeggs
  */
 #include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
 
 static const struct nv50_disp_mthd_list
 g84_disp_ovly_mthd_base = {
@@ -54,8 +51,8 @@ g84_disp_ovly_mthd_base = {
 	}
 };
 
-const struct nv50_disp_chan_mthd
-g84_disp_ovly_chan_mthd = {
+static const struct nv50_disp_chan_mthd
+g84_disp_ovly_mthd = {
 	.name = "Overlay",
 	.addr = 0x000540,
 	.prev = 0x000004,
@@ -65,13 +62,10 @@ g84_disp_ovly_chan_mthd = {
 	}
 };
 
-const struct nv50_disp_dmac_oclass
-g84_disp_ovly_oclass = {
-	.base.oclass = G82_DISP_OVERLAY_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_ovly_new,
-	.func = &nv50_disp_dmac_func,
-	.mthd = &g84_disp_ovly_chan_mthd,
-	.chid = 3,
-};
+int
+g84_disp_ovly_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		  struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_ovly_new_(&nv50_disp_dmac_func, &g84_disp_ovly_mthd,
+				   disp, 3, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygf119.c
index 5985879..30901ca 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygf119.c
@@ -22,9 +22,6 @@
  * Authors: Ben Skeggs
  */
 #include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
 
 static const struct nv50_disp_mthd_list
 gf119_disp_ovly_mthd_base = {
@@ -79,7 +76,7 @@ gf119_disp_ovly_mthd_base = {
 };
 
 static const struct nv50_disp_chan_mthd
-gf119_disp_ovly_chan_mthd = {
+gf119_disp_ovly_mthd = {
 	.name = "Overlay",
 	.addr = 0x001000,
 	.prev = -0x020000,
@@ -89,13 +86,10 @@ gf119_disp_ovly_chan_mthd = {
 	}
 };
 
-const struct nv50_disp_dmac_oclass
-gf119_disp_ovly_oclass = {
-	.base.oclass = GF110_DISP_OVERLAY_CONTROL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_ovly_new,
-	.func = &gf119_disp_dmac_func,
-	.mthd = &gf119_disp_ovly_chan_mthd,
-	.chid = 5,
-};
+int
+gf119_disp_ovly_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_ovly_new_(&gf119_disp_dmac_func, &gf119_disp_ovly_mthd,
+				   disp, 5, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygk104.c
index 2f0220b..682c146 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygk104.c
@@ -22,9 +22,6 @@
  * Authors: Ben Skeggs
  */
 #include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
 
 static const struct nv50_disp_mthd_list
 gk104_disp_ovly_mthd_base = {
@@ -81,7 +78,7 @@ gk104_disp_ovly_mthd_base = {
 };
 
 const struct nv50_disp_chan_mthd
-gk104_disp_ovly_chan_mthd = {
+gk104_disp_ovly_mthd = {
 	.name = "Overlay",
 	.addr = 0x001000,
 	.prev = -0x020000,
@@ -91,13 +88,10 @@ gk104_disp_ovly_chan_mthd = {
 	}
 };
 
-const struct nv50_disp_dmac_oclass
-gk104_disp_ovly_oclass = {
-	.base.oclass = GK104_DISP_OVERLAY_CONTROL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_ovly_new,
-	.func = &gf119_disp_dmac_func,
-	.mthd = &gk104_disp_ovly_chan_mthd,
-	.chid = 5,
-};
+int
+gk104_disp_ovly_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_ovly_new_(&gf119_disp_dmac_func, &gk104_disp_ovly_mthd,
+				   disp, 5, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygp102.c
index 589bd2f..bcc5ac4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygp102.c
@@ -22,17 +22,11 @@
  * Authors: Ben Skeggs
  */
 #include "dmacnv50.h"
-#include "rootnv50.h"
 
-#include <nvif/class.h>
-
-const struct nv50_disp_dmac_oclass
-gp102_disp_ovly_oclass = {
-	.base.oclass = GK104_DISP_OVERLAY_CONTROL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_ovly_new,
-	.func = &gp102_disp_dmac_func,
-	.mthd = &gk104_disp_ovly_chan_mthd,
-	.chid = 5,
-};
+int
+gp102_disp_ovly_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_ovly_new_(&gp102_disp_dmac_func, &gk104_disp_ovly_mthd,
+				   disp, 5, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygt200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygt200.c
index f858053..655deb0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygt200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygt200.c
@@ -22,9 +22,6 @@
  * Authors: Ben Skeggs
  */
 #include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
 
 static const struct nv50_disp_mthd_list
 gt200_disp_ovly_mthd_base = {
@@ -58,7 +55,7 @@ gt200_disp_ovly_mthd_base = {
 };
 
 static const struct nv50_disp_chan_mthd
-gt200_disp_ovly_chan_mthd = {
+gt200_disp_ovly_mthd = {
 	.name = "Overlay",
 	.addr = 0x000540,
 	.prev = 0x000004,
@@ -68,13 +65,10 @@ gt200_disp_ovly_chan_mthd = {
 	}
 };
 
-const struct nv50_disp_dmac_oclass
-gt200_disp_ovly_oclass = {
-	.base.oclass = GT200_DISP_OVERLAY_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_ovly_new,
-	.func = &nv50_disp_dmac_func,
-	.mthd = &gt200_disp_ovly_chan_mthd,
-	.chid = 3,
-};
+int
+gt200_disp_ovly_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_ovly_new_(&nv50_disp_dmac_func, &gt200_disp_ovly_mthd,
+				   disp, 3, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygt215.c
deleted file mode 100644
index c947e1e..0000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygt215.c
+++ /dev/null
@@ -1,38 +0,0 @@
-/*
- * Copyright 2015 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
- */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_dmac_oclass
-gt215_disp_ovly_oclass = {
-	.base.oclass = GT214_DISP_OVERLAY_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_ovly_new,
-	.func = &nv50_disp_dmac_func,
-	.mthd = &g84_disp_ovly_chan_mthd,
-	.chid = 3,
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlynv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlynv50.c
index d0e5546..46f5df0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlynv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlynv50.c
@@ -23,31 +23,28 @@
  */
 #include "dmacnv50.h"
 #include "head.h"
-#include "rootnv50.h"
 
 #include <core/client.h>
 
-#include <nvif/class.h>
 #include <nvif/cl507e.h>
 #include <nvif/unpack.h>
 
 int
-nv50_disp_ovly_new(const struct nv50_disp_dmac_func *func,
-		   const struct nv50_disp_chan_mthd *mthd,
-		   struct nv50_disp_root *root, int chid,
-		   const struct nvkm_oclass *oclass, void *data, u32 size,
-		   struct nvkm_object **pobject)
+nv50_disp_ovly_new_(const struct nv50_disp_dmac_func *func,
+		    const struct nv50_disp_chan_mthd *mthd,
+		    struct nv50_disp *disp, int chid,
+		    const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nvkm_object **pobject)
 {
 	union {
 		struct nv50_disp_overlay_channel_dma_v0 v0;
-	} *args = data;
+	} *args = argv;
 	struct nvkm_object *parent = oclass->parent;
-	struct nv50_disp *disp = root->disp;
 	int head, ret = -ENOSYS;
 	u64 push;
 
-	nvif_ioctl(parent, "create disp overlay channel dma size %d\n", size);
-	if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false))) {
+	nvif_ioctl(parent, "create disp overlay channel dma size %d\n", argc);
+	if (!(ret = nvif_unpack(ret, &argv, &argc, args->v0, 0, 0, false))) {
 		nvif_ioctl(parent, "create disp overlay channel dma vers %d "
 				   "pushbuf %016llx head %d\n",
 			   args->v0.version, args->v0.pushbuf, args->v0.head);
@@ -91,7 +88,7 @@ nv50_disp_ovly_mthd_base = {
 };
 
 static const struct nv50_disp_chan_mthd
-nv50_disp_ovly_chan_mthd = {
+nv50_disp_ovly_mthd = {
 	.name = "Overlay",
 	.addr = 0x000540,
 	.prev = 0x000004,
@@ -101,13 +98,10 @@ nv50_disp_ovly_chan_mthd = {
 	}
 };
 
-const struct nv50_disp_dmac_oclass
-nv50_disp_ovly_oclass = {
-	.base.oclass = NV50_DISP_OVERLAY_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_ovly_new,
-	.func = &nv50_disp_dmac_func,
-	.mthd = &nv50_disp_ovly_chan_mthd,
-	.chid = 3,
-};
+int
+nv50_disp_ovly_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		   struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_ovly_new_(&nv50_disp_dmac_func, &nv50_disp_ovly_mthd,
+				   disp, 3, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c
index 36ac0d42..ebb0803 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c
@@ -31,12 +31,15 @@ g84_disp_root = {
 	.dmac = {
 		&g84_disp_core_oclass,
 		&g84_disp_base_oclass,
-		&g84_disp_ovly_oclass,
 	},
 	.pioc = {
 		&g84_disp_oimm_oclass,
 		&g84_disp_curs_oclass,
 	},
+	.user = {
+		{{0,0,G82_DISP_OVERLAY_CHANNEL_DMA},  g84_disp_ovly_new },
+		{}
+	},
 };
 
 static int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c
index 18b87b3..6949cf0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c
@@ -31,12 +31,15 @@ g94_disp_root = {
 	.dmac = {
 		&g94_disp_core_oclass,
 		&gt200_disp_base_oclass,
-		&gt200_disp_ovly_oclass,
 	},
 	.pioc = {
 		&g84_disp_oimm_oclass,
 		&g84_disp_curs_oclass,
 	},
+	.user = {
+		{{0,0,GT200_DISP_OVERLAY_CHANNEL_DMA}, gt200_disp_ovly_new },
+		{}
+	},
 };
 
 static int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
index 7c5701f..acc897f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
@@ -31,12 +31,15 @@ gf119_disp_root = {
 	.dmac = {
 		&gf119_disp_core_oclass,
 		&gf119_disp_base_oclass,
-		&gf119_disp_ovly_oclass,
 	},
 	.pioc = {
 		&gf119_disp_oimm_oclass,
 		&gf119_disp_curs_oclass,
 	},
+	.user = {
+		{{0,0,GF110_DISP_OVERLAY_CONTROL_DMA}, gf119_disp_ovly_new },
+		{}
+	},
 };
 
 static int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c
index c0946a6..2d19a06 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c
@@ -31,12 +31,15 @@ gk104_disp_root = {
 	.dmac = {
 		&gk104_disp_core_oclass,
 		&gk104_disp_base_oclass,
-		&gk104_disp_ovly_oclass,
 	},
 	.pioc = {
 		&gk104_disp_oimm_oclass,
 		&gk104_disp_curs_oclass,
 	},
+	.user = {
+		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
+		{}
+	},
 };
 
 static int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c
index 2ebc166..d7e2249 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c
@@ -31,12 +31,15 @@ gk110_disp_root = {
 	.dmac = {
 		&gk110_disp_core_oclass,
 		&gk110_disp_base_oclass,
-		&gk104_disp_ovly_oclass,
 	},
 	.pioc = {
 		&gk104_disp_oimm_oclass,
 		&gk104_disp_curs_oclass,
 	},
+	.user = {
+		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
+		{}
+	},
 };
 
 static int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c
index 5a62c9e..de87b27 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c
@@ -31,12 +31,15 @@ gm107_disp_root = {
 	.dmac = {
 		&gm107_disp_core_oclass,
 		&gk110_disp_base_oclass,
-		&gk104_disp_ovly_oclass,
 	},
 	.pioc = {
 		&gk104_disp_oimm_oclass,
 		&gk104_disp_curs_oclass,
 	},
+	.user = {
+		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
+		{}
+	},
 };
 
 static int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c
index 2634e06..9e97c1e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c
@@ -31,12 +31,15 @@ gm200_disp_root = {
 	.dmac = {
 		&gm200_disp_core_oclass,
 		&gk110_disp_base_oclass,
-		&gk104_disp_ovly_oclass,
 	},
 	.pioc = {
 		&gk104_disp_oimm_oclass,
 		&gk104_disp_curs_oclass,
 	},
+	.user = {
+		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
+		{}
+	},
 };
 
 static int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c
index 7847235..af7031f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c
@@ -31,12 +31,15 @@ gp100_disp_root = {
 	.dmac = {
 		&gp100_disp_core_oclass,
 		&gk110_disp_base_oclass,
-		&gk104_disp_ovly_oclass,
 	},
 	.pioc = {
 		&gk104_disp_oimm_oclass,
 		&gk104_disp_curs_oclass,
 	},
+	.user = {
+		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
+		{}
+	},
 };
 
 static int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c
index 2fdfa8d..7603a41 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c
@@ -31,12 +31,15 @@ gp102_disp_root = {
 	.dmac = {
 		&gp102_disp_core_oclass,
 		&gp102_disp_base_oclass,
-		&gp102_disp_ovly_oclass,
 	},
 	.pioc = {
 		&gp102_disp_oimm_oclass,
 		&gp102_disp_curs_oclass,
 	},
+	.user = {
+		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gp102_disp_ovly_new },
+		{}
+	},
 };
 
 static int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c
index facad27..8ef149d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c
@@ -31,12 +31,15 @@ gt200_disp_root = {
 	.dmac = {
 		&gt200_disp_core_oclass,
 		&gt200_disp_base_oclass,
-		&gt200_disp_ovly_oclass,
 	},
 	.pioc = {
 		&g84_disp_oimm_oclass,
 		&g84_disp_curs_oclass,
 	},
+	.user = {
+		{{0,0,GT200_DISP_OVERLAY_CHANNEL_DMA}, gt200_disp_ovly_new },
+		{}
+	},
 };
 
 static int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c
index 3e93db5..07c8013 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c
@@ -31,12 +31,15 @@ gt215_disp_root = {
 	.dmac = {
 		&gt215_disp_core_oclass,
 		&gt215_disp_base_oclass,
-		&gt215_disp_ovly_oclass,
 	},
 	.pioc = {
 		&gt215_disp_oimm_oclass,
 		&gt215_disp_curs_oclass,
 	},
+	.user = {
+		{{0,0,GT214_DISP_OVERLAY_CHANNEL_DMA},   g84_disp_ovly_new },
+		{}
+	},
 };
 
 static int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
index 072c8c0..f02368f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
@@ -289,6 +289,15 @@ nv50_disp_root_pioc_new_(const struct nvkm_oclass *oclass,
 }
 
 static int
+nv50_disp_root_child_new_(const struct nvkm_oclass *oclass,
+			  void *argv, u32 argc, struct nvkm_object **pobject)
+{
+	struct nv50_disp *disp = nv50_disp_root(oclass->parent)->disp;
+	const struct nv50_disp_user *user = oclass->priv;
+	return user->ctor(oclass, argv, argc, disp, pobject);
+}
+
+static int
 nv50_disp_root_child_get_(struct nvkm_object *object, int index,
 			  struct nvkm_oclass *sclass)
 {
@@ -310,6 +319,15 @@ nv50_disp_root_child_get_(struct nvkm_object *object, int index,
 		return 0;
 	}
 
+	index -= ARRAY_SIZE(root->func->pioc);
+
+	if (root->func->user[index].ctor) {
+		sclass->base = root->func->user[index].base;
+		sclass->priv = root->func->user + index;
+		sclass->ctor = nv50_disp_root_child_new_;
+		return 0;
+	}
+
 	return -EINVAL;
 }
 
@@ -351,12 +369,15 @@ nv50_disp_root = {
 	.dmac = {
 		&nv50_disp_core_oclass,
 		&nv50_disp_base_oclass,
-		&nv50_disp_ovly_oclass,
 	},
 	.pioc = {
 		&nv50_disp_oimm_oclass,
 		&nv50_disp_curs_oclass,
 	},
+	.user = {
+		{{0,0,NV50_DISP_OVERLAY_CHANNEL_DMA}, nv50_disp_ovly_new },
+		{}
+	},
 };
 
 static int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
index 06b554b..1c4985a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
@@ -13,8 +13,13 @@ struct nv50_disp_root {
 };
 
 struct nv50_disp_root_func {
-	const struct nv50_disp_dmac_oclass *dmac[3];
+	const struct nv50_disp_dmac_oclass *dmac[2];
 	const struct nv50_disp_pioc_oclass *pioc[2];
+	struct nv50_disp_user {
+		struct nvkm_sclass base;
+		int (*ctor)(const struct nvkm_oclass *, void *argv, u32 argc,
+			    struct nv50_disp *, struct nvkm_object **);
+	} user[];
 };
 
 int  nv50_disp_root_new_(const struct nv50_disp_root_func *, struct nvkm_disp *,
-- 
cgit v1.1


From c2c3a00310df71e1f92d99ec3d5818d152f12bc8 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/disp/nv50-: simplify definition of overlay immediate
 channels

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild    |  3 --
 .../gpu/drm/nouveau/nvkm/engine/disp/channv50.h    | 21 ++++++------
 drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmg84.c | 37 ----------------------
 .../gpu/drm/nouveau/nvkm/engine/disp/oimmgf119.c   | 19 ++++-------
 .../gpu/drm/nouveau/nvkm/engine/disp/oimmgk104.c   | 37 ----------------------
 .../gpu/drm/nouveau/nvkm/engine/disp/oimmgp102.c   | 19 ++++-------
 .../gpu/drm/nouveau/nvkm/engine/disp/oimmgt215.c   | 37 ----------------------
 .../gpu/drm/nouveau/nvkm/engine/disp/oimmnv50.c    | 36 +++++++++------------
 drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c   |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c   |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c   |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c   |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c   |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c   |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c   |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c   |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c   |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c    |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h    |  2 +-
 21 files changed, 52 insertions(+), 183 deletions(-)
 delete mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmg84.c
 delete mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgk104.c
 delete mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgt215.c

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild b/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild
index 67861f2..b53a0e2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild
@@ -111,8 +111,5 @@ nvkm-y += nvkm/engine/disp/cursgk104.o
 nvkm-y += nvkm/engine/disp/cursgp102.o
 
 nvkm-y += nvkm/engine/disp/oimmnv50.o
-nvkm-y += nvkm/engine/disp/oimmg84.o
-nvkm-y += nvkm/engine/disp/oimmgt215.o
 nvkm-y += nvkm/engine/disp/oimmgf119.o
-nvkm-y += nvkm/engine/disp/oimmgk104.o
 nvkm-y += nvkm/engine/disp/oimmgp102.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
index b222a1d..ba30766 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
@@ -50,12 +50,18 @@ void nv50_disp_chan_uevent_send(struct nv50_disp *, int);
 
 extern const struct nvkm_event_func gf119_disp_chan_uevent;
 
+int nv50_disp_oimm_new_(const struct nv50_disp_chan_func *,
+			struct nv50_disp *, int ctrl, int user,
+			const struct nvkm_oclass *, void *argv, u32 argc,
+			struct nvkm_object **);
 int nv50_disp_ovly_new_(const struct nv50_disp_dmac_func *,
 			const struct nv50_disp_chan_mthd *,
 			struct nv50_disp *, int chid,
 			const struct nvkm_oclass *, void *argv, u32 argc,
 			struct nvkm_object **);
 
+int nv50_disp_oimm_new(const struct nvkm_oclass *, void *, u32,
+		       struct nv50_disp *, struct nvkm_object **);
 int nv50_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
 		       struct nv50_disp *, struct nvkm_object **);
 
@@ -65,12 +71,16 @@ int g84_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
 int gt200_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
 			struct nv50_disp *, struct nvkm_object **);
 
+int gf119_disp_oimm_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
 int gf119_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
 			struct nv50_disp *, struct nvkm_object **);
 
 int gk104_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
 			struct nv50_disp *, struct nvkm_object **);
 
+int gp102_disp_oimm_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
 int gp102_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
 			struct nv50_disp *, struct nvkm_object **);
 
@@ -133,22 +143,16 @@ struct nv50_disp_pioc_oclass {
 	} chid;
 };
 
-extern const struct nv50_disp_pioc_oclass nv50_disp_oimm_oclass;
 extern const struct nv50_disp_pioc_oclass nv50_disp_curs_oclass;
 
-extern const struct nv50_disp_pioc_oclass g84_disp_oimm_oclass;
 extern const struct nv50_disp_pioc_oclass g84_disp_curs_oclass;
 
-extern const struct nv50_disp_pioc_oclass gt215_disp_oimm_oclass;
 extern const struct nv50_disp_pioc_oclass gt215_disp_curs_oclass;
 
-extern const struct nv50_disp_pioc_oclass gf119_disp_oimm_oclass;
 extern const struct nv50_disp_pioc_oclass gf119_disp_curs_oclass;
 
-extern const struct nv50_disp_pioc_oclass gk104_disp_oimm_oclass;
 extern const struct nv50_disp_pioc_oclass gk104_disp_curs_oclass;
 
-extern const struct nv50_disp_pioc_oclass gp102_disp_oimm_oclass;
 extern const struct nv50_disp_pioc_oclass gp102_disp_curs_oclass;
 
 int nv50_disp_curs_new(const struct nv50_disp_chan_func *,
@@ -156,9 +160,4 @@ int nv50_disp_curs_new(const struct nv50_disp_chan_func *,
 		       struct nv50_disp_root *, int ctrl, int user,
 		       const struct nvkm_oclass *, void *data, u32 size,
 		       struct nvkm_object **);
-int nv50_disp_oimm_new(const struct nv50_disp_chan_func *,
-		       const struct nv50_disp_chan_mthd *,
-		       struct nv50_disp_root *, int ctrl, int user,
-		       const struct nvkm_oclass *, void *data, u32 size,
-		       struct nvkm_object **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmg84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmg84.c
deleted file mode 100644
index 5ad5d0f..0000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmg84.c
+++ /dev/null
@@ -1,37 +0,0 @@
-/*
- * Copyright 2012 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs
- */
-#include "channv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_pioc_oclass
-g84_disp_oimm_oclass = {
-	.base.oclass = G82_DISP_OVERLAY,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_oimm_new,
-	.func = &nv50_disp_pioc_func,
-	.chid = { 5, 5 },
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgf119.c
index 1f9fd34..1ae0bcf 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgf119.c
@@ -22,16 +22,11 @@
  * Authors: Ben Skeggs
  */
 #include "channv50.h"
-#include "rootnv50.h"
 
-#include <nvif/class.h>
-
-const struct nv50_disp_pioc_oclass
-gf119_disp_oimm_oclass = {
-	.base.oclass = GF110_DISP_OVERLAY,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_oimm_new,
-	.func = &gf119_disp_pioc_func,
-	.chid = { 9, 9 },
-};
+int
+gf119_disp_oimm_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_oimm_new_(&gf119_disp_pioc_func, disp, 9, 9,
+				   oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgk104.c
deleted file mode 100644
index 0c09fe8..0000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgk104.c
+++ /dev/null
@@ -1,37 +0,0 @@
-/*
- * Copyright 2012 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs
- */
-#include "channv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_pioc_oclass
-gk104_disp_oimm_oclass = {
-	.base.oclass = GK104_DISP_OVERLAY,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_oimm_new,
-	.func = &gf119_disp_pioc_func,
-	.chid = { 9, 9 },
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgp102.c
index abf8236..30ffb10 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgp102.c
@@ -22,16 +22,11 @@
  * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
 #include "channv50.h"
-#include "rootnv50.h"
 
-#include <nvif/class.h>
-
-const struct nv50_disp_pioc_oclass
-gp102_disp_oimm_oclass = {
-	.base.oclass = GK104_DISP_OVERLAY,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_oimm_new,
-	.func = &gf119_disp_pioc_func,
-	.chid = { 9, 13 },
-};
+int
+gp102_disp_oimm_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_oimm_new_(&gf119_disp_pioc_func, disp, 9, 13,
+				   oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgt215.c
deleted file mode 100644
index 1281db2..0000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgt215.c
+++ /dev/null
@@ -1,37 +0,0 @@
-/*
- * Copyright 2012 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs
- */
-#include "channv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_pioc_oclass
-gt215_disp_oimm_oclass = {
-	.base.oclass = GT214_DISP_OVERLAY,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_oimm_new,
-	.func = &nv50_disp_pioc_func,
-	.chid = { 5, 5 },
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmnv50.c
index 7e013cd..0db99bf 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmnv50.c
@@ -23,30 +23,26 @@
  */
 #include "channv50.h"
 #include "head.h"
-#include "rootnv50.h"
 
 #include <core/client.h>
 
-#include <nvif/class.h>
 #include <nvif/cl507b.h>
 #include <nvif/unpack.h>
 
 int
-nv50_disp_oimm_new(const struct nv50_disp_chan_func *func,
-		   const struct nv50_disp_chan_mthd *mthd,
-		   struct nv50_disp_root *root, int ctrl, int user,
-		   const struct nvkm_oclass *oclass, void *data, u32 size,
-		   struct nvkm_object **pobject)
+nv50_disp_oimm_new_(const struct nv50_disp_chan_func *func,
+		    struct nv50_disp *disp, int ctrl, int user,
+		    const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nvkm_object **pobject)
 {
 	union {
 		struct nv50_disp_overlay_v0 v0;
-	} *args = data;
+	} *args = argv;
 	struct nvkm_object *parent = oclass->parent;
-	struct nv50_disp *disp = root->disp;
 	int head, ret = -ENOSYS;
 
-	nvif_ioctl(parent, "create disp overlay size %d\n", size);
-	if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false))) {
+	nvif_ioctl(parent, "create disp overlay size %d\n", argc);
+	if (!(ret = nvif_unpack(ret, &argv, &argc, args->v0, 0, 0, false))) {
 		nvif_ioctl(parent, "create disp overlay vers %d head %d\n",
 			   args->v0.version, args->v0.head);
 		if (!nvkm_head_find(&disp->base, args->v0.head))
@@ -55,16 +51,14 @@ nv50_disp_oimm_new(const struct nv50_disp_chan_func *func,
 	} else
 		return ret;
 
-	return nv50_disp_chan_new_(func, mthd, disp, ctrl + head, user + head,
+	return nv50_disp_chan_new_(func, NULL, disp, ctrl + head, user + head,
 				   head, oclass, pobject);
 }
 
-const struct nv50_disp_pioc_oclass
-nv50_disp_oimm_oclass = {
-	.base.oclass = NV50_DISP_OVERLAY,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_oimm_new,
-	.func = &nv50_disp_pioc_func,
-	.chid = { 5, 5 },
-};
+int
+nv50_disp_oimm_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		   struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_oimm_new_(&nv50_disp_pioc_func, disp, 5, 5,
+				   oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c
index ebb0803..650ed0d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c
@@ -33,10 +33,10 @@ g84_disp_root = {
 		&g84_disp_base_oclass,
 	},
 	.pioc = {
-		&g84_disp_oimm_oclass,
 		&g84_disp_curs_oclass,
 	},
 	.user = {
+		{{0,0,G82_DISP_OVERLAY            }, nv50_disp_oimm_new },
 		{{0,0,G82_DISP_OVERLAY_CHANNEL_DMA},  g84_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c
index 6949cf0..19d23e0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c
@@ -33,10 +33,10 @@ g94_disp_root = {
 		&gt200_disp_base_oclass,
 	},
 	.pioc = {
-		&g84_disp_oimm_oclass,
 		&g84_disp_curs_oclass,
 	},
 	.user = {
+		{{0,0,  G82_DISP_OVERLAY            },  nv50_disp_oimm_new },
 		{{0,0,GT200_DISP_OVERLAY_CHANNEL_DMA}, gt200_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
index acc897f..ef8be6a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
@@ -33,10 +33,10 @@ gf119_disp_root = {
 		&gf119_disp_base_oclass,
 	},
 	.pioc = {
-		&gf119_disp_oimm_oclass,
 		&gf119_disp_curs_oclass,
 	},
 	.user = {
+		{{0,0,GF110_DISP_OVERLAY            }, gf119_disp_oimm_new },
 		{{0,0,GF110_DISP_OVERLAY_CONTROL_DMA}, gf119_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c
index 2d19a06..67002c0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c
@@ -33,10 +33,10 @@ gk104_disp_root = {
 		&gk104_disp_base_oclass,
 	},
 	.pioc = {
-		&gk104_disp_oimm_oclass,
 		&gk104_disp_curs_oclass,
 	},
 	.user = {
+		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
 		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c
index d7e2249..f82cf9c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c
@@ -33,10 +33,10 @@ gk110_disp_root = {
 		&gk110_disp_base_oclass,
 	},
 	.pioc = {
-		&gk104_disp_oimm_oclass,
 		&gk104_disp_curs_oclass,
 	},
 	.user = {
+		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
 		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c
index de87b27..170961e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c
@@ -33,10 +33,10 @@ gm107_disp_root = {
 		&gk110_disp_base_oclass,
 	},
 	.pioc = {
-		&gk104_disp_oimm_oclass,
 		&gk104_disp_curs_oclass,
 	},
 	.user = {
+		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
 		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c
index 9e97c1e..3f77682 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c
@@ -33,10 +33,10 @@ gm200_disp_root = {
 		&gk110_disp_base_oclass,
 	},
 	.pioc = {
-		&gk104_disp_oimm_oclass,
 		&gk104_disp_curs_oclass,
 	},
 	.user = {
+		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
 		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c
index af7031f..c87b1d2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c
@@ -33,10 +33,10 @@ gp100_disp_root = {
 		&gk110_disp_base_oclass,
 	},
 	.pioc = {
-		&gk104_disp_oimm_oclass,
 		&gk104_disp_curs_oclass,
 	},
 	.user = {
+		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
 		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c
index 7603a41..16516a4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c
@@ -33,10 +33,10 @@ gp102_disp_root = {
 		&gp102_disp_base_oclass,
 	},
 	.pioc = {
-		&gp102_disp_oimm_oclass,
 		&gp102_disp_curs_oclass,
 	},
 	.user = {
+		{{0,0,GK104_DISP_OVERLAY            }, gp102_disp_oimm_new },
 		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gp102_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c
index 8ef149d..6d46bf6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c
@@ -33,10 +33,10 @@ gt200_disp_root = {
 		&gt200_disp_base_oclass,
 	},
 	.pioc = {
-		&g84_disp_oimm_oclass,
 		&g84_disp_curs_oclass,
 	},
 	.user = {
+		{{0,0,  G82_DISP_OVERLAY            },  nv50_disp_oimm_new },
 		{{0,0,GT200_DISP_OVERLAY_CHANNEL_DMA}, gt200_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c
index 07c8013..6863c94 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c
@@ -33,10 +33,10 @@ gt215_disp_root = {
 		&gt215_disp_base_oclass,
 	},
 	.pioc = {
-		&gt215_disp_oimm_oclass,
 		&gt215_disp_curs_oclass,
 	},
 	.user = {
+		{{0,0,GT214_DISP_OVERLAY            },  nv50_disp_oimm_new },
 		{{0,0,GT214_DISP_OVERLAY_CHANNEL_DMA},   g84_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
index f02368f..41219c2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
@@ -371,10 +371,10 @@ nv50_disp_root = {
 		&nv50_disp_base_oclass,
 	},
 	.pioc = {
-		&nv50_disp_oimm_oclass,
 		&nv50_disp_curs_oclass,
 	},
 	.user = {
+		{{0,0,NV50_DISP_OVERLAY            }, nv50_disp_oimm_new },
 		{{0,0,NV50_DISP_OVERLAY_CHANNEL_DMA}, nv50_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
index 1c4985a..67f9518 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
@@ -14,7 +14,7 @@ struct nv50_disp_root {
 
 struct nv50_disp_root_func {
 	const struct nv50_disp_dmac_oclass *dmac[2];
-	const struct nv50_disp_pioc_oclass *pioc[2];
+	const struct nv50_disp_pioc_oclass *pioc[1];
 	struct nv50_disp_user {
 		struct nvkm_sclass base;
 		int (*ctor)(const struct nvkm_oclass *, void *argv, u32 argc,
-- 
cgit v1.1


From 3ceeef9c03fc9ed6adbb0646b4b89096ca568670 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/disp/nv50-: simplify definition of base channels

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild    |  4 ---
 drivers/gpu/drm/nouveau/nvkm/engine/disp/baseg84.c | 24 +++++---------
 .../gpu/drm/nouveau/nvkm/engine/disp/basegf119.c   | 22 +++++--------
 .../gpu/drm/nouveau/nvkm/engine/disp/basegk104.c   | 38 ----------------------
 .../gpu/drm/nouveau/nvkm/engine/disp/basegk110.c   | 38 ----------------------
 .../gpu/drm/nouveau/nvkm/engine/disp/basegp102.c   | 20 ++++--------
 .../gpu/drm/nouveau/nvkm/engine/disp/basegt200.c   | 38 ----------------------
 .../gpu/drm/nouveau/nvkm/engine/disp/basegt215.c   | 38 ----------------------
 .../gpu/drm/nouveau/nvkm/engine/disp/basenv50.c    | 38 +++++++++-------------
 .../gpu/drm/nouveau/nvkm/engine/disp/channv50.h    | 16 +++++++--
 .../gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h    | 13 --------
 drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c   |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c   |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c   |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c   |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c   |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c   |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c   |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c   |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c   |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c    |  2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h    |  2 +-
 24 files changed, 67 insertions(+), 248 deletions(-)
 delete mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/basegk104.c
 delete mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/basegk110.c
 delete mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/basegt200.c
 delete mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/basegt215.c

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild b/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild
index b53a0e2..c2d56bb 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild
@@ -73,11 +73,7 @@ nvkm-y += nvkm/engine/disp/dmacgp102.o
 
 nvkm-y += nvkm/engine/disp/basenv50.o
 nvkm-y += nvkm/engine/disp/baseg84.o
-nvkm-y += nvkm/engine/disp/basegt200.o
-nvkm-y += nvkm/engine/disp/basegt215.o
 nvkm-y += nvkm/engine/disp/basegf119.o
-nvkm-y += nvkm/engine/disp/basegk104.o
-nvkm-y += nvkm/engine/disp/basegk110.o
 nvkm-y += nvkm/engine/disp/basegp102.o
 
 nvkm-y += nvkm/engine/disp/corenv50.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/baseg84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/baseg84.c
index 6d17630..03ec508 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/baseg84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/baseg84.c
@@ -22,9 +22,6 @@
  * Authors: Ben Skeggs
  */
 #include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
 
 static const struct nv50_disp_mthd_list
 g84_disp_base_mthd_base = {
@@ -56,8 +53,8 @@ g84_disp_base_mthd_base = {
 	}
 };
 
-const struct nv50_disp_chan_mthd
-g84_disp_base_chan_mthd = {
+static const struct nv50_disp_chan_mthd
+g84_disp_base_mthd = {
 	.name = "Base",
 	.addr = 0x000540,
 	.prev = 0x000004,
@@ -68,13 +65,10 @@ g84_disp_base_chan_mthd = {
 	}
 };
 
-const struct nv50_disp_dmac_oclass
-g84_disp_base_oclass = {
-	.base.oclass = G82_DISP_BASE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_base_new,
-	.func = &nv50_disp_dmac_func,
-	.mthd = &g84_disp_base_chan_mthd,
-	.chid = 1,
-};
+int
+g84_disp_base_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		  struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_base_new_(&nv50_disp_dmac_func, &g84_disp_base_mthd,
+				   disp, 1, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegf119.c
index ebcb925..4c372dc 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegf119.c
@@ -22,9 +22,6 @@
  * Authors: Ben Skeggs
  */
 #include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
 
 static const struct nv50_disp_mthd_list
 gf119_disp_base_mthd_base = {
@@ -91,7 +88,7 @@ gf119_disp_base_mthd_image = {
 };
 
 const struct nv50_disp_chan_mthd
-gf119_disp_base_chan_mthd = {
+gf119_disp_base_mthd = {
 	.name = "Base",
 	.addr = 0x001000,
 	.prev = -0x020000,
@@ -102,13 +99,10 @@ gf119_disp_base_chan_mthd = {
 	}
 };
 
-const struct nv50_disp_dmac_oclass
-gf119_disp_base_oclass = {
-	.base.oclass = GF110_DISP_BASE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_base_new,
-	.func = &gf119_disp_dmac_func,
-	.mthd = &gf119_disp_base_chan_mthd,
-	.chid = 1,
-};
+int
+gf119_disp_base_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_base_new_(&gf119_disp_dmac_func, &gf119_disp_base_mthd,
+				   disp, 1, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegk104.c
deleted file mode 100644
index 780a1d9..0000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegk104.c
+++ /dev/null
@@ -1,38 +0,0 @@
-/*
- * Copyright 2015 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
- */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_dmac_oclass
-gk104_disp_base_oclass = {
-	.base.oclass = GK104_DISP_BASE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_base_new,
-	.func = &gf119_disp_dmac_func,
-	.mthd = &gf119_disp_base_chan_mthd,
-	.chid = 1,
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegk110.c
deleted file mode 100644
index d8bdd24..0000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegk110.c
+++ /dev/null
@@ -1,38 +0,0 @@
-/*
- * Copyright 2015 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
- */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_dmac_oclass
-gk110_disp_base_oclass = {
-	.base.oclass = GK110_DISP_BASE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_base_new,
-	.func = &gf119_disp_dmac_func,
-	.mthd = &gf119_disp_base_chan_mthd,
-	.chid = 1,
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegp102.c
index 8a3cdeef..3a25259 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegp102.c
@@ -22,17 +22,11 @@
  * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
 #include "dmacnv50.h"
-#include "rootnv50.h"
 
-#include <nvif/class.h>
-
-const struct nv50_disp_dmac_oclass
-gp102_disp_base_oclass = {
-	.base.oclass = GK110_DISP_BASE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_base_new,
-	.func = &gp102_disp_dmac_func,
-	.mthd = &gf119_disp_base_chan_mthd,
-	.chid = 1,
-};
+int
+gp102_disp_base_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_base_new_(&gp102_disp_dmac_func, &gf119_disp_base_mthd,
+				   disp, 1, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegt200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegt200.c
deleted file mode 100644
index 93451e4..0000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegt200.c
+++ /dev/null
@@ -1,38 +0,0 @@
-/*
- * Copyright 2015 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
- */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_dmac_oclass
-gt200_disp_base_oclass = {
-	.base.oclass = GT200_DISP_BASE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_base_new,
-	.func = &nv50_disp_dmac_func,
-	.mthd = &g84_disp_base_chan_mthd,
-	.chid = 1,
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegt215.c
deleted file mode 100644
index 08e2b1f..0000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegt215.c
+++ /dev/null
@@ -1,38 +0,0 @@
-/*
- * Copyright 2015 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
- */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_dmac_oclass
-gt215_disp_base_oclass = {
-	.base.oclass = GT214_DISP_BASE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_base_new,
-	.func = &nv50_disp_dmac_func,
-	.mthd = &g84_disp_base_chan_mthd,
-	.chid = 1,
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basenv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basenv50.c
index 418741a..11639e2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basenv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basenv50.c
@@ -23,31 +23,28 @@
  */
 #include "dmacnv50.h"
 #include "head.h"
-#include "rootnv50.h"
 
 #include <core/client.h>
 
-#include <nvif/class.h>
 #include <nvif/cl507c.h>
 #include <nvif/unpack.h>
 
 int
-nv50_disp_base_new(const struct nv50_disp_dmac_func *func,
-		   const struct nv50_disp_chan_mthd *mthd,
-		   struct nv50_disp_root *root, int chid,
-		   const struct nvkm_oclass *oclass, void *data, u32 size,
-		   struct nvkm_object **pobject)
+nv50_disp_base_new_(const struct nv50_disp_dmac_func *func,
+		    const struct nv50_disp_chan_mthd *mthd,
+		    struct nv50_disp *disp, int chid,
+		    const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nvkm_object **pobject)
 {
 	union {
 		struct nv50_disp_base_channel_dma_v0 v0;
-	} *args = data;
+	} *args = argv;
 	struct nvkm_object *parent = oclass->parent;
-	struct nv50_disp *disp = root->disp;
 	int head, ret = -ENOSYS;
 	u64 push;
 
-	nvif_ioctl(parent, "create disp base channel dma size %d\n", size);
-	if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false))) {
+	nvif_ioctl(parent, "create disp base channel dma size %d\n", argc);
+	if (!(ret = nvif_unpack(ret, &argv, &argc, args->v0, 0, 0, false))) {
 		nvif_ioctl(parent, "create disp base channel dma vers %d "
 				   "pushbuf %016llx head %d\n",
 			   args->v0.version, args->v0.pushbuf, args->v0.head);
@@ -102,7 +99,7 @@ nv50_disp_base_mthd_image = {
 };
 
 static const struct nv50_disp_chan_mthd
-nv50_disp_base_chan_mthd = {
+nv50_disp_base_mthd = {
 	.name = "Base",
 	.addr = 0x000540,
 	.prev = 0x000004,
@@ -113,13 +110,10 @@ nv50_disp_base_chan_mthd = {
 	}
 };
 
-const struct nv50_disp_dmac_oclass
-nv50_disp_base_oclass = {
-	.base.oclass = NV50_DISP_BASE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_base_new,
-	.func = &nv50_disp_dmac_func,
-	.mthd = &nv50_disp_base_chan_mthd,
-	.chid = 1,
-};
+int
+nv50_disp_base_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		   struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_base_new_(&nv50_disp_dmac_func, &nv50_disp_base_mthd,
+				   disp, 1, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
index ba30766..5d16277 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
@@ -54,6 +54,11 @@ int nv50_disp_oimm_new_(const struct nv50_disp_chan_func *,
 			struct nv50_disp *, int ctrl, int user,
 			const struct nvkm_oclass *, void *argv, u32 argc,
 			struct nvkm_object **);
+int nv50_disp_base_new_(const struct nv50_disp_dmac_func *,
+			const struct nv50_disp_chan_mthd *,
+			struct nv50_disp *, int chid,
+			const struct nvkm_oclass *, void *argv, u32 argc,
+			struct nvkm_object **);
 int nv50_disp_ovly_new_(const struct nv50_disp_dmac_func *,
 			const struct nv50_disp_chan_mthd *,
 			struct nv50_disp *, int chid,
@@ -62,9 +67,13 @@ int nv50_disp_ovly_new_(const struct nv50_disp_dmac_func *,
 
 int nv50_disp_oimm_new(const struct nvkm_oclass *, void *, u32,
 		       struct nv50_disp *, struct nvkm_object **);
+int nv50_disp_base_new(const struct nvkm_oclass *, void *, u32,
+		       struct nv50_disp *, struct nvkm_object **);
 int nv50_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
 		       struct nv50_disp *, struct nvkm_object **);
 
+int g84_disp_base_new(const struct nvkm_oclass *, void *, u32,
+		      struct nv50_disp *, struct nvkm_object **);
 int g84_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
 		      struct nv50_disp *, struct nvkm_object **);
 
@@ -73,6 +82,8 @@ int gt200_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
 
 int gf119_disp_oimm_new(const struct nvkm_oclass *, void *, u32,
 			struct nv50_disp *, struct nvkm_object **);
+int gf119_disp_base_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
 int gf119_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
 			struct nv50_disp *, struct nvkm_object **);
 
@@ -81,6 +92,8 @@ int gk104_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
 
 int gp102_disp_oimm_new(const struct nvkm_oclass *, void *, u32,
 			struct nv50_disp *, struct nvkm_object **);
+int gp102_disp_base_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
 int gp102_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
 			struct nv50_disp *, struct nvkm_object **);
 
@@ -115,7 +128,6 @@ extern const struct nv50_disp_mthd_list nv50_disp_base_mthd_image;
 extern const struct nv50_disp_chan_mthd g84_disp_core_chan_mthd;
 extern const struct nv50_disp_mthd_list g84_disp_core_mthd_dac;
 extern const struct nv50_disp_mthd_list g84_disp_core_mthd_head;
-extern const struct nv50_disp_chan_mthd g84_disp_base_chan_mthd;
 
 extern const struct nv50_disp_chan_mthd g94_disp_core_chan_mthd;
 
@@ -123,7 +135,7 @@ extern const struct nv50_disp_mthd_list gf119_disp_core_mthd_base;
 extern const struct nv50_disp_mthd_list gf119_disp_core_mthd_dac;
 extern const struct nv50_disp_mthd_list gf119_disp_core_mthd_sor;
 extern const struct nv50_disp_mthd_list gf119_disp_core_mthd_pior;
-extern const struct nv50_disp_chan_mthd gf119_disp_base_chan_mthd;
+extern const struct nv50_disp_chan_mthd gf119_disp_base_mthd;
 
 extern const struct nv50_disp_chan_mthd gk104_disp_core_chan_mthd;
 extern const struct nv50_disp_chan_mthd gk104_disp_ovly_mthd;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h
index 45caaee..ae094c7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h
@@ -50,34 +50,22 @@ int nv50_disp_core_new(const struct nv50_disp_dmac_func *,
 		       struct nv50_disp_root *, int chid,
 		       const struct nvkm_oclass *oclass, void *data, u32 size,
 		       struct nvkm_object **);
-int nv50_disp_base_new(const struct nv50_disp_dmac_func *,
-		       const struct nv50_disp_chan_mthd *,
-		       struct nv50_disp_root *, int chid,
-		       const struct nvkm_oclass *oclass, void *data, u32 size,
-		       struct nvkm_object **);
 
 extern const struct nv50_disp_dmac_oclass nv50_disp_core_oclass;
-extern const struct nv50_disp_dmac_oclass nv50_disp_base_oclass;
 
 extern const struct nv50_disp_dmac_oclass g84_disp_core_oclass;
-extern const struct nv50_disp_dmac_oclass g84_disp_base_oclass;
 
 extern const struct nv50_disp_dmac_oclass g94_disp_core_oclass;
 
 extern const struct nv50_disp_dmac_oclass gt200_disp_core_oclass;
-extern const struct nv50_disp_dmac_oclass gt200_disp_base_oclass;
 
 extern const struct nv50_disp_dmac_oclass gt215_disp_core_oclass;
-extern const struct nv50_disp_dmac_oclass gt215_disp_base_oclass;
 
 extern const struct nv50_disp_dmac_oclass gf119_disp_core_oclass;
-extern const struct nv50_disp_dmac_oclass gf119_disp_base_oclass;
 
 extern const struct nv50_disp_dmac_oclass gk104_disp_core_oclass;
-extern const struct nv50_disp_dmac_oclass gk104_disp_base_oclass;
 
 extern const struct nv50_disp_dmac_oclass gk110_disp_core_oclass;
-extern const struct nv50_disp_dmac_oclass gk110_disp_base_oclass;
 
 extern const struct nv50_disp_dmac_oclass gm107_disp_core_oclass;
 
@@ -86,5 +74,4 @@ extern const struct nv50_disp_dmac_oclass gm200_disp_core_oclass;
 extern const struct nv50_disp_dmac_oclass gp100_disp_core_oclass;
 
 extern const struct nv50_disp_dmac_oclass gp102_disp_core_oclass;
-extern const struct nv50_disp_dmac_oclass gp102_disp_base_oclass;
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c
index 650ed0d..fa9076b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c
@@ -30,13 +30,13 @@ static const struct nv50_disp_root_func
 g84_disp_root = {
 	.dmac = {
 		&g84_disp_core_oclass,
-		&g84_disp_base_oclass,
 	},
 	.pioc = {
 		&g84_disp_curs_oclass,
 	},
 	.user = {
 		{{0,0,G82_DISP_OVERLAY            }, nv50_disp_oimm_new },
+		{{0,0,G82_DISP_BASE_CHANNEL_DMA   },  g84_disp_base_new },
 		{{0,0,G82_DISP_OVERLAY_CHANNEL_DMA},  g84_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c
index 19d23e0..6aeab0f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c
@@ -30,13 +30,13 @@ static const struct nv50_disp_root_func
 g94_disp_root = {
 	.dmac = {
 		&g94_disp_core_oclass,
-		&gt200_disp_base_oclass,
 	},
 	.pioc = {
 		&g84_disp_curs_oclass,
 	},
 	.user = {
 		{{0,0,  G82_DISP_OVERLAY            },  nv50_disp_oimm_new },
+		{{0,0,GT200_DISP_BASE_CHANNEL_DMA   },   g84_disp_base_new },
 		{{0,0,GT200_DISP_OVERLAY_CHANNEL_DMA}, gt200_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
index ef8be6a..b44b145 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
@@ -30,13 +30,13 @@ static const struct nv50_disp_root_func
 gf119_disp_root = {
 	.dmac = {
 		&gf119_disp_core_oclass,
-		&gf119_disp_base_oclass,
 	},
 	.pioc = {
 		&gf119_disp_curs_oclass,
 	},
 	.user = {
 		{{0,0,GF110_DISP_OVERLAY            }, gf119_disp_oimm_new },
+		{{0,0,GF110_DISP_BASE_CHANNEL_DMA   }, gf119_disp_base_new },
 		{{0,0,GF110_DISP_OVERLAY_CONTROL_DMA}, gf119_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c
index 67002c0..c43eab9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c
@@ -30,13 +30,13 @@ static const struct nv50_disp_root_func
 gk104_disp_root = {
 	.dmac = {
 		&gk104_disp_core_oclass,
-		&gk104_disp_base_oclass,
 	},
 	.pioc = {
 		&gk104_disp_curs_oclass,
 	},
 	.user = {
 		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
+		{{0,0,GK104_DISP_BASE_CHANNEL_DMA   }, gf119_disp_base_new },
 		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c
index f82cf9c..2d48e73 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c
@@ -30,13 +30,13 @@ static const struct nv50_disp_root_func
 gk110_disp_root = {
 	.dmac = {
 		&gk110_disp_core_oclass,
-		&gk110_disp_base_oclass,
 	},
 	.pioc = {
 		&gk104_disp_curs_oclass,
 	},
 	.user = {
 		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
+		{{0,0,GK110_DISP_BASE_CHANNEL_DMA   }, gf119_disp_base_new },
 		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c
index 170961e..904125e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c
@@ -30,13 +30,13 @@ static const struct nv50_disp_root_func
 gm107_disp_root = {
 	.dmac = {
 		&gm107_disp_core_oclass,
-		&gk110_disp_base_oclass,
 	},
 	.pioc = {
 		&gk104_disp_curs_oclass,
 	},
 	.user = {
 		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
+		{{0,0,GK110_DISP_BASE_CHANNEL_DMA   }, gf119_disp_base_new },
 		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c
index 3f77682..52e9cca 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c
@@ -30,13 +30,13 @@ static const struct nv50_disp_root_func
 gm200_disp_root = {
 	.dmac = {
 		&gm200_disp_core_oclass,
-		&gk110_disp_base_oclass,
 	},
 	.pioc = {
 		&gk104_disp_curs_oclass,
 	},
 	.user = {
 		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
+		{{0,0,GK110_DISP_BASE_CHANNEL_DMA   }, gf119_disp_base_new },
 		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c
index c87b1d2..57f40d4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c
@@ -30,13 +30,13 @@ static const struct nv50_disp_root_func
 gp100_disp_root = {
 	.dmac = {
 		&gp100_disp_core_oclass,
-		&gk110_disp_base_oclass,
 	},
 	.pioc = {
 		&gk104_disp_curs_oclass,
 	},
 	.user = {
 		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
+		{{0,0,GK110_DISP_BASE_CHANNEL_DMA   }, gf119_disp_base_new },
 		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c
index 16516a4..f66d7fc 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c
@@ -30,13 +30,13 @@ static const struct nv50_disp_root_func
 gp102_disp_root = {
 	.dmac = {
 		&gp102_disp_core_oclass,
-		&gp102_disp_base_oclass,
 	},
 	.pioc = {
 		&gp102_disp_curs_oclass,
 	},
 	.user = {
 		{{0,0,GK104_DISP_OVERLAY            }, gp102_disp_oimm_new },
+		{{0,0,GK110_DISP_BASE_CHANNEL_DMA   }, gp102_disp_base_new },
 		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gp102_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c
index 6d46bf6..c280179 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c
@@ -30,13 +30,13 @@ static const struct nv50_disp_root_func
 gt200_disp_root = {
 	.dmac = {
 		&gt200_disp_core_oclass,
-		&gt200_disp_base_oclass,
 	},
 	.pioc = {
 		&g84_disp_curs_oclass,
 	},
 	.user = {
 		{{0,0,  G82_DISP_OVERLAY            },  nv50_disp_oimm_new },
+		{{0,0,GT200_DISP_BASE_CHANNEL_DMA   },   g84_disp_base_new },
 		{{0,0,GT200_DISP_OVERLAY_CHANNEL_DMA}, gt200_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c
index 6863c94..b9a6a32 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c
@@ -30,13 +30,13 @@ static const struct nv50_disp_root_func
 gt215_disp_root = {
 	.dmac = {
 		&gt215_disp_core_oclass,
-		&gt215_disp_base_oclass,
 	},
 	.pioc = {
 		&gt215_disp_curs_oclass,
 	},
 	.user = {
 		{{0,0,GT214_DISP_OVERLAY            },  nv50_disp_oimm_new },
+		{{0,0,GT214_DISP_BASE_CHANNEL_DMA   },   g84_disp_base_new },
 		{{0,0,GT214_DISP_OVERLAY_CHANNEL_DMA},   g84_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
index 41219c2..a87531d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
@@ -368,13 +368,13 @@ static const struct nv50_disp_root_func
 nv50_disp_root = {
 	.dmac = {
 		&nv50_disp_core_oclass,
-		&nv50_disp_base_oclass,
 	},
 	.pioc = {
 		&nv50_disp_curs_oclass,
 	},
 	.user = {
 		{{0,0,NV50_DISP_OVERLAY            }, nv50_disp_oimm_new },
+		{{0,0,NV50_DISP_BASE_CHANNEL_DMA   }, nv50_disp_base_new },
 		{{0,0,NV50_DISP_OVERLAY_CHANNEL_DMA}, nv50_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
index 67f9518..371c6ee 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
@@ -13,7 +13,7 @@ struct nv50_disp_root {
 };
 
 struct nv50_disp_root_func {
-	const struct nv50_disp_dmac_oclass *dmac[2];
+	const struct nv50_disp_dmac_oclass *dmac[1];
 	const struct nv50_disp_pioc_oclass *pioc[1];
 	struct nv50_disp_user {
 		struct nvkm_sclass base;
-- 
cgit v1.1


From 6d41a7536f8cff35be9b3c4ccb94e55f1553a7a1 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/disp/nv50-: simplify definition of cursor channels

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild    |  3 --
 .../gpu/drm/nouveau/nvkm/engine/disp/channv50.h    | 43 +++++-----------------
 drivers/gpu/drm/nouveau/nvkm/engine/disp/cursg84.c | 37 -------------------
 .../gpu/drm/nouveau/nvkm/engine/disp/cursgf119.c   | 19 ++++------
 .../gpu/drm/nouveau/nvkm/engine/disp/cursgk104.c   | 37 -------------------
 .../gpu/drm/nouveau/nvkm/engine/disp/cursgp102.c   | 19 ++++------
 .../gpu/drm/nouveau/nvkm/engine/disp/cursgt215.c   | 37 -------------------
 .../gpu/drm/nouveau/nvkm/engine/disp/cursnv50.c    | 36 ++++++++----------
 drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c |  4 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c |  4 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c   |  4 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c   |  4 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c   |  4 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c   |  4 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c   |  4 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c   |  4 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c   |  4 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c   |  4 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c   |  4 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c    | 23 +-----------
 .../gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h    |  2 -
 21 files changed, 51 insertions(+), 249 deletions(-)
 delete mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/cursg84.c
 delete mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgk104.c
 delete mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgt215.c

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild b/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild
index c2d56bb..ee41fb7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild
@@ -100,10 +100,7 @@ nvkm-y += nvkm/engine/disp/piocnv50.o
 nvkm-y += nvkm/engine/disp/piocgf119.o
 
 nvkm-y += nvkm/engine/disp/cursnv50.o
-nvkm-y += nvkm/engine/disp/cursg84.o
-nvkm-y += nvkm/engine/disp/cursgt215.o
 nvkm-y += nvkm/engine/disp/cursgf119.o
-nvkm-y += nvkm/engine/disp/cursgk104.o
 nvkm-y += nvkm/engine/disp/cursgp102.o
 
 nvkm-y += nvkm/engine/disp/oimmnv50.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
index 5d16277..d52420f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
@@ -50,6 +50,10 @@ void nv50_disp_chan_uevent_send(struct nv50_disp *, int);
 
 extern const struct nvkm_event_func gf119_disp_chan_uevent;
 
+int nv50_disp_curs_new_(const struct nv50_disp_chan_func *,
+			struct nv50_disp *, int ctrl, int user,
+			const struct nvkm_oclass *, void *argv, u32 argc,
+			struct nvkm_object **);
 int nv50_disp_oimm_new_(const struct nv50_disp_chan_func *,
 			struct nv50_disp *, int ctrl, int user,
 			const struct nvkm_oclass *, void *argv, u32 argc,
@@ -65,6 +69,8 @@ int nv50_disp_ovly_new_(const struct nv50_disp_dmac_func *,
 			const struct nvkm_oclass *, void *argv, u32 argc,
 			struct nvkm_object **);
 
+int nv50_disp_curs_new(const struct nvkm_oclass *, void *, u32,
+		       struct nv50_disp *, struct nvkm_object **);
 int nv50_disp_oimm_new(const struct nvkm_oclass *, void *, u32,
 		       struct nv50_disp *, struct nvkm_object **);
 int nv50_disp_base_new(const struct nvkm_oclass *, void *, u32,
@@ -80,6 +86,8 @@ int g84_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
 int gt200_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
 			struct nv50_disp *, struct nvkm_object **);
 
+int gf119_disp_curs_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
 int gf119_disp_oimm_new(const struct nvkm_oclass *, void *, u32,
 			struct nv50_disp *, struct nvkm_object **);
 int gf119_disp_base_new(const struct nvkm_oclass *, void *, u32,
@@ -90,6 +98,8 @@ int gf119_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
 int gk104_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
 			struct nv50_disp *, struct nvkm_object **);
 
+int gp102_disp_curs_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
 int gp102_disp_oimm_new(const struct nvkm_oclass *, void *, u32,
 			struct nv50_disp *, struct nvkm_object **);
 int gp102_disp_base_new(const struct nvkm_oclass *, void *, u32,
@@ -139,37 +149,4 @@ extern const struct nv50_disp_chan_mthd gf119_disp_base_mthd;
 
 extern const struct nv50_disp_chan_mthd gk104_disp_core_chan_mthd;
 extern const struct nv50_disp_chan_mthd gk104_disp_ovly_mthd;
-
-struct nv50_disp_pioc_oclass {
-	int (*ctor)(const struct nv50_disp_chan_func *,
-		    const struct nv50_disp_chan_mthd *,
-		    struct nv50_disp_root *, int ctrl, int user,
-		    const struct nvkm_oclass *, void *data, u32 size,
-		    struct nvkm_object **);
-	struct nvkm_sclass base;
-	const struct nv50_disp_chan_func *func;
-	const struct nv50_disp_chan_mthd *mthd;
-	struct {
-		int ctrl;
-		int user;
-	} chid;
-};
-
-extern const struct nv50_disp_pioc_oclass nv50_disp_curs_oclass;
-
-extern const struct nv50_disp_pioc_oclass g84_disp_curs_oclass;
-
-extern const struct nv50_disp_pioc_oclass gt215_disp_curs_oclass;
-
-extern const struct nv50_disp_pioc_oclass gf119_disp_curs_oclass;
-
-extern const struct nv50_disp_pioc_oclass gk104_disp_curs_oclass;
-
-extern const struct nv50_disp_pioc_oclass gp102_disp_curs_oclass;
-
-int nv50_disp_curs_new(const struct nv50_disp_chan_func *,
-		       const struct nv50_disp_chan_mthd *,
-		       struct nv50_disp_root *, int ctrl, int user,
-		       const struct nvkm_oclass *, void *data, u32 size,
-		       struct nvkm_object **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursg84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursg84.c
deleted file mode 100644
index fa781b5..0000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursg84.c
+++ /dev/null
@@ -1,37 +0,0 @@
-/*
- * Copyright 2015 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
- */
-#include "channv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_pioc_oclass
-g84_disp_curs_oclass = {
-	.base.oclass = G82_DISP_CURSOR,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_curs_new,
-	.func = &nv50_disp_pioc_func,
-	.chid = { 7, 7 },
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgf119.c
index 2be6fb0..cdda365 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgf119.c
@@ -22,16 +22,11 @@
  * Authors: Ben Skeggs
  */
 #include "channv50.h"
-#include "rootnv50.h"
 
-#include <nvif/class.h>
-
-const struct nv50_disp_pioc_oclass
-gf119_disp_curs_oclass = {
-	.base.oclass = GF110_DISP_CURSOR,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_curs_new,
-	.func = &gf119_disp_pioc_func,
-	.chid = { 13, 13 },
-};
+int
+gf119_disp_curs_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_curs_new_(&gf119_disp_pioc_func, disp, 13, 13,
+				   oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgk104.c
deleted file mode 100644
index 2a99db4..0000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgk104.c
+++ /dev/null
@@ -1,37 +0,0 @@
-/*
- * Copyright 2012 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs
- */
-#include "channv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_pioc_oclass
-gk104_disp_curs_oclass = {
-	.base.oclass = GK104_DISP_CURSOR,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_curs_new,
-	.func = &gf119_disp_pioc_func,
-	.chid = { 13, 13 },
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgp102.c
index e958210..1a4601f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgp102.c
@@ -22,16 +22,11 @@
  * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
 #include "channv50.h"
-#include "rootnv50.h"
 
-#include <nvif/class.h>
-
-const struct nv50_disp_pioc_oclass
-gp102_disp_curs_oclass = {
-	.base.oclass = GK104_DISP_CURSOR,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_curs_new,
-	.func = &gf119_disp_pioc_func,
-	.chid = { 13, 17 },
-};
+int
+gp102_disp_curs_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_curs_new_(&gf119_disp_pioc_func, disp, 13, 17,
+				   oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgt215.c
deleted file mode 100644
index 00a7f35..0000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgt215.c
+++ /dev/null
@@ -1,37 +0,0 @@
-/*
- * Copyright 2015 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
- */
-#include "channv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_pioc_oclass
-gt215_disp_curs_oclass = {
-	.base.oclass = GT214_DISP_CURSOR,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_curs_new,
-	.func = &nv50_disp_pioc_func,
-	.chid = { 7, 7 },
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursnv50.c
index 1f9a6c3..d297585 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursnv50.c
@@ -23,30 +23,26 @@
  */
 #include "channv50.h"
 #include "head.h"
-#include "rootnv50.h"
 
 #include <core/client.h>
 
-#include <nvif/class.h>
 #include <nvif/cl507a.h>
 #include <nvif/unpack.h>
 
 int
-nv50_disp_curs_new(const struct nv50_disp_chan_func *func,
-		   const struct nv50_disp_chan_mthd *mthd,
-		   struct nv50_disp_root *root, int ctrl, int user,
-		   const struct nvkm_oclass *oclass, void *data, u32 size,
-		   struct nvkm_object **pobject)
+nv50_disp_curs_new_(const struct nv50_disp_chan_func *func,
+		    struct nv50_disp *disp, int ctrl, int user,
+		    const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nvkm_object **pobject)
 {
 	union {
 		struct nv50_disp_cursor_v0 v0;
-	} *args = data;
+	} *args = argv;
 	struct nvkm_object *parent = oclass->parent;
-	struct nv50_disp *disp = root->disp;
 	int head, ret = -ENOSYS;
 
-	nvif_ioctl(parent, "create disp cursor size %d\n", size);
-	if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false))) {
+	nvif_ioctl(parent, "create disp cursor size %d\n", argc);
+	if (!(ret = nvif_unpack(ret, &argv, &argc, args->v0, 0, 0, false))) {
 		nvif_ioctl(parent, "create disp cursor vers %d head %d\n",
 			   args->v0.version, args->v0.head);
 		if (!nvkm_head_find(&disp->base, args->v0.head))
@@ -55,16 +51,14 @@ nv50_disp_curs_new(const struct nv50_disp_chan_func *func,
 	} else
 		return ret;
 
-	return nv50_disp_chan_new_(func, mthd, disp, ctrl + head, user + head,
+	return nv50_disp_chan_new_(func, NULL, disp, ctrl + head, user + head,
 				   head, oclass, pobject);
 }
 
-const struct nv50_disp_pioc_oclass
-nv50_disp_curs_oclass = {
-	.base.oclass = NV50_DISP_CURSOR,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_curs_new,
-	.func = &nv50_disp_pioc_func,
-	.chid = { 7, 7 },
-};
+int
+nv50_disp_curs_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		   struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_curs_new_(&nv50_disp_pioc_func, disp, 7, 7,
+				   oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c
index fa9076b..0f7e662 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c
@@ -31,10 +31,8 @@ g84_disp_root = {
 	.dmac = {
 		&g84_disp_core_oclass,
 	},
-	.pioc = {
-		&g84_disp_curs_oclass,
-	},
 	.user = {
+		{{0,0,G82_DISP_CURSOR             }, nv50_disp_curs_new },
 		{{0,0,G82_DISP_OVERLAY            }, nv50_disp_oimm_new },
 		{{0,0,G82_DISP_BASE_CHANNEL_DMA   },  g84_disp_base_new },
 		{{0,0,G82_DISP_OVERLAY_CHANNEL_DMA},  g84_disp_ovly_new },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c
index 6aeab0f..20709d6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c
@@ -31,10 +31,8 @@ g94_disp_root = {
 	.dmac = {
 		&g94_disp_core_oclass,
 	},
-	.pioc = {
-		&g84_disp_curs_oclass,
-	},
 	.user = {
+		{{0,0,  G82_DISP_CURSOR             },  nv50_disp_curs_new },
 		{{0,0,  G82_DISP_OVERLAY            },  nv50_disp_oimm_new },
 		{{0,0,GT200_DISP_BASE_CHANNEL_DMA   },   g84_disp_base_new },
 		{{0,0,GT200_DISP_OVERLAY_CHANNEL_DMA}, gt200_disp_ovly_new },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
index b44b145..1161698 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
@@ -31,10 +31,8 @@ gf119_disp_root = {
 	.dmac = {
 		&gf119_disp_core_oclass,
 	},
-	.pioc = {
-		&gf119_disp_curs_oclass,
-	},
 	.user = {
+		{{0,0,GF110_DISP_CURSOR             }, gf119_disp_curs_new },
 		{{0,0,GF110_DISP_OVERLAY            }, gf119_disp_oimm_new },
 		{{0,0,GF110_DISP_BASE_CHANNEL_DMA   }, gf119_disp_base_new },
 		{{0,0,GF110_DISP_OVERLAY_CONTROL_DMA}, gf119_disp_ovly_new },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c
index c43eab9..15379ff 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c
@@ -31,10 +31,8 @@ gk104_disp_root = {
 	.dmac = {
 		&gk104_disp_core_oclass,
 	},
-	.pioc = {
-		&gk104_disp_curs_oclass,
-	},
 	.user = {
+		{{0,0,GK104_DISP_CURSOR             }, gf119_disp_curs_new },
 		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
 		{{0,0,GK104_DISP_BASE_CHANNEL_DMA   }, gf119_disp_base_new },
 		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c
index 2d48e73..0a47674f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c
@@ -31,10 +31,8 @@ gk110_disp_root = {
 	.dmac = {
 		&gk110_disp_core_oclass,
 	},
-	.pioc = {
-		&gk104_disp_curs_oclass,
-	},
 	.user = {
+		{{0,0,GK104_DISP_CURSOR             }, gf119_disp_curs_new },
 		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
 		{{0,0,GK110_DISP_BASE_CHANNEL_DMA   }, gf119_disp_base_new },
 		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c
index 904125e..819521e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c
@@ -31,10 +31,8 @@ gm107_disp_root = {
 	.dmac = {
 		&gm107_disp_core_oclass,
 	},
-	.pioc = {
-		&gk104_disp_curs_oclass,
-	},
 	.user = {
+		{{0,0,GK104_DISP_CURSOR             }, gf119_disp_curs_new },
 		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
 		{{0,0,GK110_DISP_BASE_CHANNEL_DMA   }, gf119_disp_base_new },
 		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c
index 52e9cca..0dca177 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c
@@ -31,10 +31,8 @@ gm200_disp_root = {
 	.dmac = {
 		&gm200_disp_core_oclass,
 	},
-	.pioc = {
-		&gk104_disp_curs_oclass,
-	},
 	.user = {
+		{{0,0,GK104_DISP_CURSOR             }, gf119_disp_curs_new },
 		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
 		{{0,0,GK110_DISP_BASE_CHANNEL_DMA   }, gf119_disp_base_new },
 		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c
index 57f40d4..6fcf858 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c
@@ -31,10 +31,8 @@ gp100_disp_root = {
 	.dmac = {
 		&gp100_disp_core_oclass,
 	},
-	.pioc = {
-		&gk104_disp_curs_oclass,
-	},
 	.user = {
+		{{0,0,GK104_DISP_CURSOR             }, gf119_disp_curs_new },
 		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
 		{{0,0,GK110_DISP_BASE_CHANNEL_DMA   }, gf119_disp_base_new },
 		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c
index f66d7fc..bf5cbc3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c
@@ -31,10 +31,8 @@ gp102_disp_root = {
 	.dmac = {
 		&gp102_disp_core_oclass,
 	},
-	.pioc = {
-		&gp102_disp_curs_oclass,
-	},
 	.user = {
+		{{0,0,GK104_DISP_CURSOR             }, gp102_disp_curs_new },
 		{{0,0,GK104_DISP_OVERLAY            }, gp102_disp_oimm_new },
 		{{0,0,GK110_DISP_BASE_CHANNEL_DMA   }, gp102_disp_base_new },
 		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gp102_disp_ovly_new },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c
index c280179..2bd3f36 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c
@@ -31,10 +31,8 @@ gt200_disp_root = {
 	.dmac = {
 		&gt200_disp_core_oclass,
 	},
-	.pioc = {
-		&g84_disp_curs_oclass,
-	},
 	.user = {
+		{{0,0,  G82_DISP_CURSOR             },  nv50_disp_curs_new },
 		{{0,0,  G82_DISP_OVERLAY            },  nv50_disp_oimm_new },
 		{{0,0,GT200_DISP_BASE_CHANNEL_DMA   },   g84_disp_base_new },
 		{{0,0,GT200_DISP_OVERLAY_CHANNEL_DMA}, gt200_disp_ovly_new },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c
index b9a6a32..aa4c523 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c
@@ -31,10 +31,8 @@ gt215_disp_root = {
 	.dmac = {
 		&gt215_disp_core_oclass,
 	},
-	.pioc = {
-		&gt215_disp_curs_oclass,
-	},
 	.user = {
+		{{0,0,GT214_DISP_CURSOR             },  nv50_disp_curs_new },
 		{{0,0,GT214_DISP_OVERLAY            },  nv50_disp_oimm_new },
 		{{0,0,GT214_DISP_BASE_CHANNEL_DMA   },   g84_disp_base_new },
 		{{0,0,GT214_DISP_OVERLAY_CHANNEL_DMA},   g84_disp_ovly_new },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
index a87531d..24e8b41 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
@@ -279,16 +279,6 @@ nv50_disp_root_dmac_new_(const struct nvkm_oclass *oclass,
 }
 
 static int
-nv50_disp_root_pioc_new_(const struct nvkm_oclass *oclass,
-			 void *data, u32 size, struct nvkm_object **pobject)
-{
-	const struct nv50_disp_pioc_oclass *sclass = oclass->priv;
-	struct nv50_disp_root *root = nv50_disp_root(oclass->parent);
-	return sclass->ctor(sclass->func, sclass->mthd, root, sclass->chid.ctrl,
-			    sclass->chid.user, oclass, data, size, pobject);
-}
-
-static int
 nv50_disp_root_child_new_(const struct nvkm_oclass *oclass,
 			  void *argv, u32 argc, struct nvkm_object **pobject)
 {
@@ -312,15 +302,6 @@ nv50_disp_root_child_get_(struct nvkm_object *object, int index,
 
 	index -= ARRAY_SIZE(root->func->dmac);
 
-	if (index < ARRAY_SIZE(root->func->pioc)) {
-		sclass->base = root->func->pioc[index]->base;
-		sclass->priv = root->func->pioc[index];
-		sclass->ctor = nv50_disp_root_pioc_new_;
-		return 0;
-	}
-
-	index -= ARRAY_SIZE(root->func->pioc);
-
 	if (root->func->user[index].ctor) {
 		sclass->base = root->func->user[index].base;
 		sclass->priv = root->func->user + index;
@@ -369,10 +350,8 @@ nv50_disp_root = {
 	.dmac = {
 		&nv50_disp_core_oclass,
 	},
-	.pioc = {
-		&nv50_disp_curs_oclass,
-	},
 	.user = {
+		{{0,0,NV50_DISP_CURSOR             }, nv50_disp_curs_new },
 		{{0,0,NV50_DISP_OVERLAY            }, nv50_disp_oimm_new },
 		{{0,0,NV50_DISP_BASE_CHANNEL_DMA   }, nv50_disp_base_new },
 		{{0,0,NV50_DISP_OVERLAY_CHANNEL_DMA}, nv50_disp_ovly_new },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
index 371c6ee..484868a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
@@ -3,7 +3,6 @@
 #define __NV50_DISP_ROOT_H__
 #define nv50_disp_root(p) container_of((p), struct nv50_disp_root, object)
 #include "nv50.h"
-#include "channv50.h"
 #include "dmacnv50.h"
 
 struct nv50_disp_root {
@@ -14,7 +13,6 @@ struct nv50_disp_root {
 
 struct nv50_disp_root_func {
 	const struct nv50_disp_dmac_oclass *dmac[1];
-	const struct nv50_disp_pioc_oclass *pioc[1];
 	struct nv50_disp_user {
 		struct nvkm_sclass base;
 		int (*ctor)(const struct nvkm_oclass *, void *argv, u32 argc,
-- 
cgit v1.1


From 9b096283bf78f659e1286ef9b783b27ecf5a9977 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/disp/nv50-: simplify definiton of core channels

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild    |  6 ----
 .../gpu/drm/nouveau/nvkm/engine/disp/channv50.h    | 24 +++++++++++--
 drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg84.c | 22 +++++-------
 drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg94.c | 22 +++++-------
 .../gpu/drm/nouveau/nvkm/engine/disp/coregf119.c   | 24 +++++--------
 .../gpu/drm/nouveau/nvkm/engine/disp/coregk104.c   | 22 +++++-------
 .../gpu/drm/nouveau/nvkm/engine/disp/coregk110.c   | 38 --------------------
 .../gpu/drm/nouveau/nvkm/engine/disp/coregm107.c   | 38 --------------------
 .../gpu/drm/nouveau/nvkm/engine/disp/coregm200.c   | 38 --------------------
 .../gpu/drm/nouveau/nvkm/engine/disp/coregp100.c   | 38 --------------------
 .../gpu/drm/nouveau/nvkm/engine/disp/coregp102.c   | 20 ++++-------
 .../gpu/drm/nouveau/nvkm/engine/disp/coregt200.c   | 38 --------------------
 .../gpu/drm/nouveau/nvkm/engine/disp/coregt215.c   | 38 --------------------
 .../gpu/drm/nouveau/nvkm/engine/disp/corenv50.c    | 39 +++++++++-----------
 .../gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h    | 42 ----------------------
 drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c    |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c |  4 +--
 drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c |  4 +--
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c   |  4 +--
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c   |  4 +--
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c   |  4 +--
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c   |  4 +--
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c   |  4 +--
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c   |  4 +--
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c   |  4 +--
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c   |  4 +--
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c   |  4 +--
 .../gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c    | 23 +-----------
 .../gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h    |  2 +-
 29 files changed, 91 insertions(+), 428 deletions(-)
 delete mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/coregk110.c
 delete mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/coregm107.c
 delete mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/coregm200.c
 delete mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp100.c
 delete mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/coregt200.c
 delete mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/coregt215.c

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild b/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild
index ee41fb7..b580581 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild
@@ -79,14 +79,8 @@ nvkm-y += nvkm/engine/disp/basegp102.o
 nvkm-y += nvkm/engine/disp/corenv50.o
 nvkm-y += nvkm/engine/disp/coreg84.o
 nvkm-y += nvkm/engine/disp/coreg94.o
-nvkm-y += nvkm/engine/disp/coregt200.o
-nvkm-y += nvkm/engine/disp/coregt215.o
 nvkm-y += nvkm/engine/disp/coregf119.o
 nvkm-y += nvkm/engine/disp/coregk104.o
-nvkm-y += nvkm/engine/disp/coregk110.o
-nvkm-y += nvkm/engine/disp/coregm107.o
-nvkm-y += nvkm/engine/disp/coregm200.o
-nvkm-y += nvkm/engine/disp/coregp100.o
 nvkm-y += nvkm/engine/disp/coregp102.o
 
 nvkm-y += nvkm/engine/disp/ovlynv50.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
index d52420f..10ce217 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
@@ -63,6 +63,11 @@ int nv50_disp_base_new_(const struct nv50_disp_dmac_func *,
 			struct nv50_disp *, int chid,
 			const struct nvkm_oclass *, void *argv, u32 argc,
 			struct nvkm_object **);
+int nv50_disp_core_new_(const struct nv50_disp_dmac_func *,
+			const struct nv50_disp_chan_mthd *,
+			struct nv50_disp *, int chid,
+			const struct nvkm_oclass *oclass, void *argv, u32 argc,
+			struct nvkm_object **);
 int nv50_disp_ovly_new_(const struct nv50_disp_dmac_func *,
 			const struct nv50_disp_chan_mthd *,
 			struct nv50_disp *, int chid,
@@ -75,14 +80,21 @@ int nv50_disp_oimm_new(const struct nvkm_oclass *, void *, u32,
 		       struct nv50_disp *, struct nvkm_object **);
 int nv50_disp_base_new(const struct nvkm_oclass *, void *, u32,
 		       struct nv50_disp *, struct nvkm_object **);
+int nv50_disp_core_new(const struct nvkm_oclass *, void *, u32,
+		       struct nv50_disp *, struct nvkm_object **);
 int nv50_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
 		       struct nv50_disp *, struct nvkm_object **);
 
 int g84_disp_base_new(const struct nvkm_oclass *, void *, u32,
 		      struct nv50_disp *, struct nvkm_object **);
+int g84_disp_core_new(const struct nvkm_oclass *, void *, u32,
+		      struct nv50_disp *, struct nvkm_object **);
 int g84_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
 		      struct nv50_disp *, struct nvkm_object **);
 
+int g94_disp_core_new(const struct nvkm_oclass *, void *, u32,
+		      struct nv50_disp *, struct nvkm_object **);
+
 int gt200_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
 			struct nv50_disp *, struct nvkm_object **);
 
@@ -92,9 +104,13 @@ int gf119_disp_oimm_new(const struct nvkm_oclass *, void *, u32,
 			struct nv50_disp *, struct nvkm_object **);
 int gf119_disp_base_new(const struct nvkm_oclass *, void *, u32,
 			struct nv50_disp *, struct nvkm_object **);
+int gf119_disp_core_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
 int gf119_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
 			struct nv50_disp *, struct nvkm_object **);
 
+int gk104_disp_core_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
 int gk104_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
 			struct nv50_disp *, struct nvkm_object **);
 
@@ -104,6 +120,8 @@ int gp102_disp_oimm_new(const struct nvkm_oclass *, void *, u32,
 			struct nv50_disp *, struct nvkm_object **);
 int gp102_disp_base_new(const struct nvkm_oclass *, void *, u32,
 			struct nv50_disp *, struct nvkm_object **);
+int gp102_disp_core_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
 int gp102_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
 			struct nv50_disp *, struct nvkm_object **);
 
@@ -135,11 +153,11 @@ extern const struct nv50_disp_mthd_list nv50_disp_core_mthd_sor;
 extern const struct nv50_disp_mthd_list nv50_disp_core_mthd_pior;
 extern const struct nv50_disp_mthd_list nv50_disp_base_mthd_image;
 
-extern const struct nv50_disp_chan_mthd g84_disp_core_chan_mthd;
+extern const struct nv50_disp_chan_mthd g84_disp_core_mthd;
 extern const struct nv50_disp_mthd_list g84_disp_core_mthd_dac;
 extern const struct nv50_disp_mthd_list g84_disp_core_mthd_head;
 
-extern const struct nv50_disp_chan_mthd g94_disp_core_chan_mthd;
+extern const struct nv50_disp_chan_mthd g94_disp_core_mthd;
 
 extern const struct nv50_disp_mthd_list gf119_disp_core_mthd_base;
 extern const struct nv50_disp_mthd_list gf119_disp_core_mthd_dac;
@@ -147,6 +165,6 @@ extern const struct nv50_disp_mthd_list gf119_disp_core_mthd_sor;
 extern const struct nv50_disp_mthd_list gf119_disp_core_mthd_pior;
 extern const struct nv50_disp_chan_mthd gf119_disp_base_mthd;
 
-extern const struct nv50_disp_chan_mthd gk104_disp_core_chan_mthd;
+extern const struct nv50_disp_chan_mthd gk104_disp_core_mthd;
 extern const struct nv50_disp_chan_mthd gk104_disp_ovly_mthd;
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg84.c
index 1baa5c3..b16857f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg84.c
@@ -22,9 +22,6 @@
  * Authors: Ben Skeggs
  */
 #include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
 
 const struct nv50_disp_mthd_list
 g84_disp_core_mthd_dac = {
@@ -91,7 +88,7 @@ g84_disp_core_mthd_head = {
 };
 
 const struct nv50_disp_chan_mthd
-g84_disp_core_chan_mthd = {
+g84_disp_core_mthd = {
 	.name = "Core",
 	.addr = 0x000000,
 	.prev = 0x000004,
@@ -105,13 +102,10 @@ g84_disp_core_chan_mthd = {
 	}
 };
 
-const struct nv50_disp_dmac_oclass
-g84_disp_core_oclass = {
-	.base.oclass = G82_DISP_CORE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_core_new,
-	.func = &nv50_disp_core_func,
-	.mthd = &g84_disp_core_chan_mthd,
-	.chid = 0,
-};
+int
+g84_disp_core_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		  struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_core_new_(&nv50_disp_core_func, &g84_disp_core_mthd,
+				   disp, 0, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg94.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg94.c
index c65c9f3..ea5f489 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg94.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg94.c
@@ -22,9 +22,6 @@
  * Authors: Ben Skeggs
  */
 #include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
 
 static const struct nv50_disp_mthd_list
 g94_disp_core_mthd_sor = {
@@ -37,7 +34,7 @@ g94_disp_core_mthd_sor = {
 };
 
 const struct nv50_disp_chan_mthd
-g94_disp_core_chan_mthd = {
+g94_disp_core_mthd = {
 	.name = "Core",
 	.addr = 0x000000,
 	.prev = 0x000004,
@@ -51,13 +48,10 @@ g94_disp_core_chan_mthd = {
 	}
 };
 
-const struct nv50_disp_dmac_oclass
-g94_disp_core_oclass = {
-	.base.oclass = GT206_DISP_CORE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_core_new,
-	.func = &nv50_disp_core_func,
-	.mthd = &g94_disp_core_chan_mthd,
-	.chid = 0,
-};
+int
+g94_disp_core_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		  struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_core_new_(&nv50_disp_core_func, &g94_disp_core_mthd,
+				   disp, 0, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c
index 200dd90..9e48cc3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c
@@ -22,14 +22,9 @@
  * Authors: Ben Skeggs
  */
 #include "dmacnv50.h"
-#include "rootnv50.h"
 
-#include <core/client.h>
 #include <subdev/timer.h>
 
-#include <nvif/class.h>
-#include <nvif/unpack.h>
-
 const struct nv50_disp_mthd_list
 gf119_disp_core_mthd_base = {
 	.mthd = 0x0000,
@@ -157,7 +152,7 @@ gf119_disp_core_mthd_head = {
 };
 
 static const struct nv50_disp_chan_mthd
-gf119_disp_core_chan_mthd = {
+gf119_disp_core_mthd = {
 	.name = "Core",
 	.addr = 0x000000,
 	.prev = -0x020000,
@@ -232,13 +227,10 @@ gf119_disp_core_func = {
 	.bind = gf119_disp_dmac_bind,
 };
 
-const struct nv50_disp_dmac_oclass
-gf119_disp_core_oclass = {
-	.base.oclass = GF110_DISP_CORE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_core_new,
-	.func = &gf119_disp_core_func,
-	.mthd = &gf119_disp_core_chan_mthd,
-	.chid = 0,
-};
+int
+gf119_disp_core_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_core_new_(&gf119_disp_core_func, &gf119_disp_core_mthd,
+				   disp, 0, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregk104.c
index 088ab22..ca09595 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregk104.c
@@ -22,9 +22,6 @@
  * Authors: Ben Skeggs
  */
 #include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
 
 static const struct nv50_disp_mthd_list
 gk104_disp_core_mthd_head = {
@@ -106,7 +103,7 @@ gk104_disp_core_mthd_head = {
 };
 
 const struct nv50_disp_chan_mthd
-gk104_disp_core_chan_mthd = {
+gk104_disp_core_mthd = {
 	.name = "Core",
 	.addr = 0x000000,
 	.prev = -0x020000,
@@ -120,13 +117,10 @@ gk104_disp_core_chan_mthd = {
 	}
 };
 
-const struct nv50_disp_dmac_oclass
-gk104_disp_core_oclass = {
-	.base.oclass = GK104_DISP_CORE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_core_new,
-	.func = &gf119_disp_core_func,
-	.mthd = &gk104_disp_core_chan_mthd,
-	.chid = 0,
-};
+int
+gk104_disp_core_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_core_new_(&gf119_disp_core_func, &gk104_disp_core_mthd,
+				   disp, 0, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregk110.c
deleted file mode 100644
index df0f45c..0000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregk110.c
+++ /dev/null
@@ -1,38 +0,0 @@
-/*
- * Copyright 2015 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
- */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_dmac_oclass
-gk110_disp_core_oclass = {
-	.base.oclass = GK110_DISP_CORE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_core_new,
-	.func = &gf119_disp_core_func,
-	.mthd = &gk104_disp_core_chan_mthd,
-	.chid = 0,
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregm107.c
deleted file mode 100644
index 9e27f8f..0000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregm107.c
+++ /dev/null
@@ -1,38 +0,0 @@
-/*
- * Copyright 2015 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
- */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_dmac_oclass
-gm107_disp_core_oclass = {
-	.base.oclass = GM107_DISP_CORE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_core_new,
-	.func = &gf119_disp_core_func,
-	.mthd = &gk104_disp_core_chan_mthd,
-	.chid = 0,
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregm200.c
deleted file mode 100644
index bb23a86..0000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregm200.c
+++ /dev/null
@@ -1,38 +0,0 @@
-/*
- * Copyright 2015 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
- */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_dmac_oclass
-gm200_disp_core_oclass = {
-	.base.oclass = GM200_DISP_CORE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_core_new,
-	.func = &gf119_disp_core_func,
-	.mthd = &gk104_disp_core_chan_mthd,
-	.chid = 0,
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp100.c
deleted file mode 100644
index d5dff66..0000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp100.c
+++ /dev/null
@@ -1,38 +0,0 @@
-/*
- * Copyright 2015 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
- */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_dmac_oclass
-gp100_disp_core_oclass = {
-	.base.oclass = GP100_DISP_CORE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_core_new,
-	.func = &gf119_disp_core_func,
-	.mthd = &gk104_disp_core_chan_mthd,
-	.chid = 0,
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c
index 6ad5f2f..3ec353e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c
@@ -22,12 +22,9 @@
  * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
 #include "dmacnv50.h"
-#include "rootnv50.h"
 
 #include <subdev/timer.h>
 
-#include <nvif/class.h>
-
 static int
 gp102_disp_core_init(struct nv50_disp_dmac *chan)
 {
@@ -66,13 +63,10 @@ gp102_disp_core_func = {
 	.bind = gf119_disp_dmac_bind,
 };
 
-const struct nv50_disp_dmac_oclass
-gp102_disp_core_oclass = {
-	.base.oclass = GP102_DISP_CORE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_core_new,
-	.func = &gp102_disp_core_func,
-	.mthd = &gk104_disp_core_chan_mthd,
-	.chid = 0,
-};
+int
+gp102_disp_core_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_core_new_(&gp102_disp_core_func, &gk104_disp_core_mthd,
+				   disp, 0, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregt200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregt200.c
deleted file mode 100644
index b234547..0000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregt200.c
+++ /dev/null
@@ -1,38 +0,0 @@
-/*
- * Copyright 2015 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
- */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_dmac_oclass
-gt200_disp_core_oclass = {
-	.base.oclass = GT200_DISP_CORE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_core_new,
-	.func = &nv50_disp_core_func,
-	.mthd = &g84_disp_core_chan_mthd,
-	.chid = 0,
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregt215.c
deleted file mode 100644
index 8f5ba20..0000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregt215.c
+++ /dev/null
@@ -1,38 +0,0 @@
-/*
- * Copyright 2015 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
- */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_dmac_oclass
-gt215_disp_core_oclass = {
-	.base.oclass = GT214_DISP_CORE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_core_new,
-	.func = &nv50_disp_core_func,
-	.mthd = &g94_disp_core_chan_mthd,
-	.chid = 0,
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c
index 839cbc8..8cdcf5b5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c
@@ -22,31 +22,29 @@
  * Authors: Ben Skeggs
  */
 #include "dmacnv50.h"
-#include "rootnv50.h"
 
 #include <core/client.h>
 #include <subdev/timer.h>
 
-#include <nvif/class.h>
 #include <nvif/cl507d.h>
 #include <nvif/unpack.h>
 
 int
-nv50_disp_core_new(const struct nv50_disp_dmac_func *func,
-		   const struct nv50_disp_chan_mthd *mthd,
-		   struct nv50_disp_root *root, int chid,
-		   const struct nvkm_oclass *oclass, void *data, u32 size,
-		   struct nvkm_object **pobject)
+nv50_disp_core_new_(const struct nv50_disp_dmac_func *func,
+		    const struct nv50_disp_chan_mthd *mthd,
+		    struct nv50_disp *disp, int chid,
+		    const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nvkm_object **pobject)
 {
 	union {
 		struct nv50_disp_core_channel_dma_v0 v0;
-	} *args = data;
+	} *args = argv;
 	struct nvkm_object *parent = oclass->parent;
 	u64 push;
 	int ret = -ENOSYS;
 
-	nvif_ioctl(parent, "create disp core channel dma size %d\n", size);
-	if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false))) {
+	nvif_ioctl(parent, "create disp core channel dma size %d\n", argc);
+	if (!(ret = nvif_unpack(ret, &argv, &argc, args->v0, 0, 0, false))) {
 		nvif_ioctl(parent, "create disp core channel dma vers %d "
 				   "pushbuf %016llx\n",
 			   args->v0.version, args->v0.pushbuf);
@@ -54,7 +52,7 @@ nv50_disp_core_new(const struct nv50_disp_dmac_func *func,
 	} else
 		return ret;
 
-	return nv50_disp_dmac_new_(func, mthd, root->disp, chid, 0,
+	return nv50_disp_dmac_new_(func, mthd, disp, chid, 0,
 				   push, oclass, pobject);
 }
 
@@ -151,7 +149,7 @@ nv50_disp_core_mthd_head = {
 };
 
 static const struct nv50_disp_chan_mthd
-nv50_disp_core_chan_mthd = {
+nv50_disp_core_mthd = {
 	.name = "Core",
 	.addr = 0x000000,
 	.prev = 0x000004,
@@ -231,13 +229,10 @@ nv50_disp_core_func = {
 	.bind = nv50_disp_dmac_bind,
 };
 
-const struct nv50_disp_dmac_oclass
-nv50_disp_core_oclass = {
-	.base.oclass = NV50_DISP_CORE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_core_new,
-	.func = &nv50_disp_core_func,
-	.mthd = &nv50_disp_core_chan_mthd,
-	.chid = 0,
-};
+int
+nv50_disp_core_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		   struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_core_new_(&nv50_disp_core_func, &nv50_disp_core_mthd,
+				   disp, 0, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h
index ae094c7..feeb588 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h
@@ -32,46 +32,4 @@ extern const struct nv50_disp_dmac_func gf119_disp_core_func;
 void gf119_disp_core_fini(struct nv50_disp_dmac *);
 
 extern const struct nv50_disp_dmac_func gp102_disp_dmac_func;
-
-struct nv50_disp_dmac_oclass {
-	int (*ctor)(const struct nv50_disp_dmac_func *,
-		    const struct nv50_disp_chan_mthd *,
-		    struct nv50_disp_root *, int chid,
-		    const struct nvkm_oclass *, void *data, u32 size,
-		    struct nvkm_object **);
-	struct nvkm_sclass base;
-	const struct nv50_disp_dmac_func *func;
-	const struct nv50_disp_chan_mthd *mthd;
-	int chid;
-};
-
-int nv50_disp_core_new(const struct nv50_disp_dmac_func *,
-		       const struct nv50_disp_chan_mthd *,
-		       struct nv50_disp_root *, int chid,
-		       const struct nvkm_oclass *oclass, void *data, u32 size,
-		       struct nvkm_object **);
-
-extern const struct nv50_disp_dmac_oclass nv50_disp_core_oclass;
-
-extern const struct nv50_disp_dmac_oclass g84_disp_core_oclass;
-
-extern const struct nv50_disp_dmac_oclass g94_disp_core_oclass;
-
-extern const struct nv50_disp_dmac_oclass gt200_disp_core_oclass;
-
-extern const struct nv50_disp_dmac_oclass gt215_disp_core_oclass;
-
-extern const struct nv50_disp_dmac_oclass gf119_disp_core_oclass;
-
-extern const struct nv50_disp_dmac_oclass gk104_disp_core_oclass;
-
-extern const struct nv50_disp_dmac_oclass gk110_disp_core_oclass;
-
-extern const struct nv50_disp_dmac_oclass gm107_disp_core_oclass;
-
-extern const struct nv50_disp_dmac_oclass gm200_disp_core_oclass;
-
-extern const struct nv50_disp_dmac_oclass gp100_disp_core_oclass;
-
-extern const struct nv50_disp_dmac_oclass gp102_disp_core_oclass;
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
index 1d2280a..4a37c44 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
@@ -24,6 +24,7 @@
 #include "nv50.h"
 #include "head.h"
 #include "ior.h"
+#include "channv50.h"
 #include "rootnv50.h"
 
 #include <core/client.h>
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c
index 0f7e662..e51f89f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c
@@ -28,13 +28,11 @@
 
 static const struct nv50_disp_root_func
 g84_disp_root = {
-	.dmac = {
-		&g84_disp_core_oclass,
-	},
 	.user = {
 		{{0,0,G82_DISP_CURSOR             }, nv50_disp_curs_new },
 		{{0,0,G82_DISP_OVERLAY            }, nv50_disp_oimm_new },
 		{{0,0,G82_DISP_BASE_CHANNEL_DMA   },  g84_disp_base_new },
+		{{0,0,G82_DISP_CORE_CHANNEL_DMA   },  g84_disp_core_new },
 		{{0,0,G82_DISP_OVERLAY_CHANNEL_DMA},  g84_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c
index 20709d6..ed7838e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c
@@ -28,13 +28,11 @@
 
 static const struct nv50_disp_root_func
 g94_disp_root = {
-	.dmac = {
-		&g94_disp_core_oclass,
-	},
 	.user = {
 		{{0,0,  G82_DISP_CURSOR             },  nv50_disp_curs_new },
 		{{0,0,  G82_DISP_OVERLAY            },  nv50_disp_oimm_new },
 		{{0,0,GT200_DISP_BASE_CHANNEL_DMA   },   g84_disp_base_new },
+		{{0,0,GT206_DISP_CORE_CHANNEL_DMA   },   g94_disp_core_new },
 		{{0,0,GT200_DISP_OVERLAY_CHANNEL_DMA}, gt200_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
index 1161698..ac92e65 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
@@ -28,13 +28,11 @@
 
 static const struct nv50_disp_root_func
 gf119_disp_root = {
-	.dmac = {
-		&gf119_disp_core_oclass,
-	},
 	.user = {
 		{{0,0,GF110_DISP_CURSOR             }, gf119_disp_curs_new },
 		{{0,0,GF110_DISP_OVERLAY            }, gf119_disp_oimm_new },
 		{{0,0,GF110_DISP_BASE_CHANNEL_DMA   }, gf119_disp_base_new },
+		{{0,0,GF110_DISP_CORE_CHANNEL_DMA   }, gf119_disp_core_new },
 		{{0,0,GF110_DISP_OVERLAY_CONTROL_DMA}, gf119_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c
index 15379ff..3bb6d60 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c
@@ -28,13 +28,11 @@
 
 static const struct nv50_disp_root_func
 gk104_disp_root = {
-	.dmac = {
-		&gk104_disp_core_oclass,
-	},
 	.user = {
 		{{0,0,GK104_DISP_CURSOR             }, gf119_disp_curs_new },
 		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
 		{{0,0,GK104_DISP_BASE_CHANNEL_DMA   }, gf119_disp_base_new },
+		{{0,0,GK104_DISP_CORE_CHANNEL_DMA   }, gk104_disp_core_new },
 		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c
index 0a47674f..336419815 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c
@@ -28,13 +28,11 @@
 
 static const struct nv50_disp_root_func
 gk110_disp_root = {
-	.dmac = {
-		&gk110_disp_core_oclass,
-	},
 	.user = {
 		{{0,0,GK104_DISP_CURSOR             }, gf119_disp_curs_new },
 		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
 		{{0,0,GK110_DISP_BASE_CHANNEL_DMA   }, gf119_disp_base_new },
+		{{0,0,GK110_DISP_CORE_CHANNEL_DMA   }, gk104_disp_core_new },
 		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c
index 819521e..c53e71e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c
@@ -28,13 +28,11 @@
 
 static const struct nv50_disp_root_func
 gm107_disp_root = {
-	.dmac = {
-		&gm107_disp_core_oclass,
-	},
 	.user = {
 		{{0,0,GK104_DISP_CURSOR             }, gf119_disp_curs_new },
 		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
 		{{0,0,GK110_DISP_BASE_CHANNEL_DMA   }, gf119_disp_base_new },
+		{{0,0,GM107_DISP_CORE_CHANNEL_DMA   }, gk104_disp_core_new },
 		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c
index 0dca177..85409d1 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c
@@ -28,13 +28,11 @@
 
 static const struct nv50_disp_root_func
 gm200_disp_root = {
-	.dmac = {
-		&gm200_disp_core_oclass,
-	},
 	.user = {
 		{{0,0,GK104_DISP_CURSOR             }, gf119_disp_curs_new },
 		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
 		{{0,0,GK110_DISP_BASE_CHANNEL_DMA   }, gf119_disp_base_new },
+		{{0,0,GM200_DISP_CORE_CHANNEL_DMA   }, gk104_disp_core_new },
 		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c
index 6fcf858..ebfd245 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c
@@ -28,13 +28,11 @@
 
 static const struct nv50_disp_root_func
 gp100_disp_root = {
-	.dmac = {
-		&gp100_disp_core_oclass,
-	},
 	.user = {
 		{{0,0,GK104_DISP_CURSOR             }, gf119_disp_curs_new },
 		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
 		{{0,0,GK110_DISP_BASE_CHANNEL_DMA   }, gf119_disp_base_new },
+		{{0,0,GP100_DISP_CORE_CHANNEL_DMA   }, gk104_disp_core_new },
 		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c
index bf5cbc3..54b5fda 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c
@@ -28,13 +28,11 @@
 
 static const struct nv50_disp_root_func
 gp102_disp_root = {
-	.dmac = {
-		&gp102_disp_core_oclass,
-	},
 	.user = {
 		{{0,0,GK104_DISP_CURSOR             }, gp102_disp_curs_new },
 		{{0,0,GK104_DISP_OVERLAY            }, gp102_disp_oimm_new },
 		{{0,0,GK110_DISP_BASE_CHANNEL_DMA   }, gp102_disp_base_new },
+		{{0,0,GP102_DISP_CORE_CHANNEL_DMA   }, gp102_disp_core_new },
 		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gp102_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c
index 2bd3f36..14ac83b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c
@@ -28,13 +28,11 @@
 
 static const struct nv50_disp_root_func
 gt200_disp_root = {
-	.dmac = {
-		&gt200_disp_core_oclass,
-	},
 	.user = {
 		{{0,0,  G82_DISP_CURSOR             },  nv50_disp_curs_new },
 		{{0,0,  G82_DISP_OVERLAY            },  nv50_disp_oimm_new },
 		{{0,0,GT200_DISP_BASE_CHANNEL_DMA   },   g84_disp_base_new },
+		{{0,0,GT200_DISP_CORE_CHANNEL_DMA   },   g84_disp_core_new },
 		{{0,0,GT200_DISP_OVERLAY_CHANNEL_DMA}, gt200_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c
index aa4c523..cb1a208 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c
@@ -28,13 +28,11 @@
 
 static const struct nv50_disp_root_func
 gt215_disp_root = {
-	.dmac = {
-		&gt215_disp_core_oclass,
-	},
 	.user = {
 		{{0,0,GT214_DISP_CURSOR             },  nv50_disp_curs_new },
 		{{0,0,GT214_DISP_OVERLAY            },  nv50_disp_oimm_new },
 		{{0,0,GT214_DISP_BASE_CHANNEL_DMA   },   g84_disp_base_new },
+		{{0,0,GT214_DISP_CORE_CHANNEL_DMA   },   g94_disp_core_new },
 		{{0,0,GT214_DISP_OVERLAY_CHANNEL_DMA},   g84_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
index 24e8b41..9d716ee7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
@@ -269,16 +269,6 @@ nv50_disp_root_mthd_(struct nvkm_object *object, u32 mthd, void *data, u32 size)
 }
 
 static int
-nv50_disp_root_dmac_new_(const struct nvkm_oclass *oclass,
-			 void *data, u32 size, struct nvkm_object **pobject)
-{
-	const struct nv50_disp_dmac_oclass *sclass = oclass->priv;
-	struct nv50_disp_root *root = nv50_disp_root(oclass->parent);
-	return sclass->ctor(sclass->func, sclass->mthd, root, sclass->chid,
-			    oclass, data, size, pobject);
-}
-
-static int
 nv50_disp_root_child_new_(const struct nvkm_oclass *oclass,
 			  void *argv, u32 argc, struct nvkm_object **pobject)
 {
@@ -293,15 +283,6 @@ nv50_disp_root_child_get_(struct nvkm_object *object, int index,
 {
 	struct nv50_disp_root *root = nv50_disp_root(object);
 
-	if (index < ARRAY_SIZE(root->func->dmac)) {
-		sclass->base = root->func->dmac[index]->base;
-		sclass->priv = root->func->dmac[index];
-		sclass->ctor = nv50_disp_root_dmac_new_;
-		return 0;
-	}
-
-	index -= ARRAY_SIZE(root->func->dmac);
-
 	if (root->func->user[index].ctor) {
 		sclass->base = root->func->user[index].base;
 		sclass->priv = root->func->user + index;
@@ -347,13 +328,11 @@ nv50_disp_root_new_(const struct nv50_disp_root_func *func,
 
 static const struct nv50_disp_root_func
 nv50_disp_root = {
-	.dmac = {
-		&nv50_disp_core_oclass,
-	},
 	.user = {
 		{{0,0,NV50_DISP_CURSOR             }, nv50_disp_curs_new },
 		{{0,0,NV50_DISP_OVERLAY            }, nv50_disp_oimm_new },
 		{{0,0,NV50_DISP_BASE_CHANNEL_DMA   }, nv50_disp_base_new },
+		{{0,0,NV50_DISP_CORE_CHANNEL_DMA   }, nv50_disp_core_new },
 		{{0,0,NV50_DISP_OVERLAY_CHANNEL_DMA}, nv50_disp_ovly_new },
 		{}
 	},
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
index 484868a..12fc13b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
@@ -12,7 +12,7 @@ struct nv50_disp_root {
 };
 
 struct nv50_disp_root_func {
-	const struct nv50_disp_dmac_oclass *dmac[1];
+	int blah;
 	struct nv50_disp_user {
 		struct nvkm_sclass base;
 		int (*ctor)(const struct nvkm_oclass *, void *argv, u32 argc,
-- 
cgit v1.1


From 8531f57027136fa63ddae91821ca89b32b571fe2 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/disp/nv50-: merge handling of pio and dma channels

Unnecessarily complicated, and a barrier to cleanly supporting Volta.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/disp/baseg84.c |   2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/basegf119.c   |   2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/basegp102.c   |   2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/basenv50.c    |   4 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/channv50.c    |  97 ++++++++++----
 .../gpu/drm/nouveau/nvkm/engine/disp/channv50.h    |  39 +++---
 drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg84.c |   2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg94.c |   2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/coregf119.c   |  14 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/coregk104.c   |   2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/coregp102.c   |   9 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/corenv50.c    |  16 +--
 .../gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c   |  31 ++---
 .../gpu/drm/nouveau/nvkm/engine/disp/dmacgp102.c   |  14 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c    | 147 +++------------------
 .../gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h    |  35 -----
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c   |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gp102.c   |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h    |   7 +
 drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlyg84.c |   2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/ovlygf119.c   |   2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/ovlygk104.c   |   2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/ovlygp102.c   |   2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/ovlygt200.c   |   2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/ovlynv50.c    |   4 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c |   2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c |   2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c   |   2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c   |   2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c   |   2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c   |   2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c   |   2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c   |   2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c   |   2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c   |   2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c   |   2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c    |   2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h    |   2 +-
 38 files changed, 183 insertions(+), 284 deletions(-)
 delete mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/baseg84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/baseg84.c
index 03ec508..01253f4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/baseg84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/baseg84.c
@@ -21,7 +21,7 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
+#include "channv50.h"
 
 static const struct nv50_disp_mthd_list
 g84_disp_base_mthd_base = {
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegf119.c
index 4c372dc..389e19d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegf119.c
@@ -21,7 +21,7 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
+#include "channv50.h"
 
 static const struct nv50_disp_mthd_list
 gf119_disp_base_mthd_base = {
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegp102.c
index 3a25259..0cb23d6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegp102.c
@@ -21,7 +21,7 @@
  *
  * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
-#include "dmacnv50.h"
+#include "channv50.h"
 
 int
 gp102_disp_base_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basenv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basenv50.c
index 11639e2..19eb7dd 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basenv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basenv50.c
@@ -21,7 +21,7 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
+#include "channv50.h"
 #include "head.h"
 
 #include <core/client.h>
@@ -30,7 +30,7 @@
 #include <nvif/unpack.h>
 
 int
-nv50_disp_base_new_(const struct nv50_disp_dmac_func *func,
+nv50_disp_base_new_(const struct nv50_disp_chan_func *func,
 		    const struct nv50_disp_chan_mthd *mthd,
 		    struct nv50_disp *disp, int chid,
 		    const struct nvkm_oclass *oclass, void *argv, u32 argc,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c
index 96b732d..91a2f0c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c
@@ -26,6 +26,7 @@
 
 #include <core/client.h>
 #include <core/notify.h>
+#include <core/oproxy.h>
 #include <core/ramht.h>
 #include <engine/dma.h>
 
@@ -204,25 +205,76 @@ nv50_disp_chan_map(struct nvkm_object *object, void *argv, u32 argc,
 	return 0;
 }
 
+struct nv50_disp_chan_object {
+	struct nvkm_oproxy oproxy;
+	struct nv50_disp *disp;
+	int hash;
+};
+
+static void
+nv50_disp_chan_child_del_(struct nvkm_oproxy *base)
+{
+	struct nv50_disp_chan_object *object =
+		container_of(base, typeof(*object), oproxy);
+	nvkm_ramht_remove(object->disp->ramht, object->hash);
+}
+
+static const struct nvkm_oproxy_func
+nv50_disp_chan_child_func_ = {
+	.dtor[0] = nv50_disp_chan_child_del_,
+};
+
 static int
 nv50_disp_chan_child_new(const struct nvkm_oclass *oclass,
-			 void *data, u32 size, struct nvkm_object **pobject)
+			 void *argv, u32 argc, struct nvkm_object **pobject)
 {
 	struct nv50_disp_chan *chan = nv50_disp_chan(oclass->parent);
-	return chan->func->child_new(chan, oclass, data, size, pobject);
+	struct nv50_disp *disp = chan->disp;
+	struct nvkm_device *device = disp->base.engine.subdev.device;
+	const struct nvkm_device_oclass *sclass = oclass->priv;
+	struct nv50_disp_chan_object *object;
+	int ret;
+
+	if (!(object = kzalloc(sizeof(*object), GFP_KERNEL)))
+		return -ENOMEM;
+	nvkm_oproxy_ctor(&nv50_disp_chan_child_func_, oclass, &object->oproxy);
+	object->disp = disp;
+	*pobject = &object->oproxy.base;
+
+	ret = sclass->ctor(device, oclass, argv, argc, &object->oproxy.object);
+	if (ret)
+		return ret;
+
+	object->hash = chan->func->bind(chan, object->oproxy.object,
+					      oclass->handle);
+	if (object->hash < 0)
+		return object->hash;
+
+	return 0;
 }
 
 static int
 nv50_disp_chan_child_get(struct nvkm_object *object, int index,
-			 struct nvkm_oclass *oclass)
+			 struct nvkm_oclass *sclass)
 {
 	struct nv50_disp_chan *chan = nv50_disp_chan(object);
-	if (chan->func->child_get) {
-		int ret = chan->func->child_get(chan, index, oclass);
-		if (ret == 0)
-			oclass->ctor = nv50_disp_chan_child_new;
-		return ret;
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	const struct nvkm_device_oclass *oclass = NULL;
+
+	if (chan->func->bind)
+		sclass->engine = nvkm_device_engine(device, NVKM_ENGINE_DMAOBJ);
+	else
+		sclass->engine = NULL;
+
+	if (sclass->engine && sclass->engine->func->base.sclass) {
+		sclass->engine->func->base.sclass(sclass, index, &oclass);
+		if (oclass) {
+			sclass->ctor = nv50_disp_chan_child_new,
+			sclass->priv = oclass;
+			return 0;
+		}
 	}
+
 	return -EINVAL;
 }
 
@@ -248,7 +300,7 @@ nv50_disp_chan_dtor(struct nvkm_object *object)
 	struct nv50_disp *disp = chan->disp;
 	if (chan->chid.user >= 0)
 		disp->chan[chan->chid.user] = NULL;
-	return chan->func->dtor ? chan->func->dtor(chan) : chan;
+	return chan;
 }
 
 static const struct nvkm_object_func
@@ -264,12 +316,18 @@ nv50_disp_chan = {
 };
 
 int
-nv50_disp_chan_ctor(const struct nv50_disp_chan_func *func,
+nv50_disp_chan_new_(const struct nv50_disp_chan_func *func,
 		    const struct nv50_disp_chan_mthd *mthd,
 		    struct nv50_disp *disp, int ctrl, int user, int head,
 		    const struct nvkm_oclass *oclass,
-		    struct nv50_disp_chan *chan)
+		    struct nvkm_object **pobject)
 {
+	struct nv50_disp_chan *chan;
+
+	if (!(chan = kzalloc(sizeof(*chan), GFP_KERNEL)))
+		return -ENOMEM;
+	*pobject = &chan->object;
+
 	nvkm_object_ctor(&nv50_disp_chan, oclass, &chan->object);
 	chan->func = func;
 	chan->mthd = mthd;
@@ -285,20 +343,3 @@ nv50_disp_chan_ctor(const struct nv50_disp_chan_func *func,
 	disp->chan[chan->chid.user] = chan;
 	return 0;
 }
-
-int
-nv50_disp_chan_new_(const struct nv50_disp_chan_func *func,
-		    const struct nv50_disp_chan_mthd *mthd,
-		    struct nv50_disp *disp, int ctrl, int user, int head,
-		    const struct nvkm_oclass *oclass,
-		    struct nvkm_object **pobject)
-{
-	struct nv50_disp_chan *chan;
-
-	if (!(chan = kzalloc(sizeof(*chan), GFP_KERNEL)))
-		return -ENOMEM;
-	*pobject = &chan->object;
-
-	return nv50_disp_chan_ctor(func, mthd, disp, ctrl, user,
-				   head, oclass, chan);
-}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
index 10ce217..feb8a56 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
@@ -5,7 +5,6 @@
 #include <core/object.h>
 #include "nv50.h"
 struct nv50_disp_root;
-struct nv50_disp_dmac_func;
 
 struct nv50_disp_chan {
 	const struct nv50_disp_chan_func *func;
@@ -19,36 +18,38 @@ struct nv50_disp_chan {
 	int head;
 
 	struct nvkm_object object;
+
+	u64 push;
 };
 
 struct nv50_disp_chan_func {
-	void *(*dtor)(struct nv50_disp_chan *);
 	int (*init)(struct nv50_disp_chan *);
 	void (*fini)(struct nv50_disp_chan *);
-	int (*child_get)(struct nv50_disp_chan *, int index,
-			 struct nvkm_oclass *);
-	int (*child_new)(struct nv50_disp_chan *, const struct nvkm_oclass *,
-			 void *data, u32 size, struct nvkm_object **);
+	int (*bind)(struct nv50_disp_chan *, struct nvkm_object *, u32 handle);
 };
 
-int nv50_disp_chan_ctor(const struct nv50_disp_chan_func *,
-			const struct nv50_disp_chan_mthd *,
-			struct nv50_disp *, int ctrl, int user, int head,
-			const struct nvkm_oclass *, struct nv50_disp_chan *);
 int nv50_disp_chan_new_(const struct nv50_disp_chan_func *,
 			const struct nv50_disp_chan_mthd *,
 			struct nv50_disp *, int ctrl, int user, int head,
 			const struct nvkm_oclass *, struct nvkm_object **);
+int nv50_disp_dmac_new_(const struct nv50_disp_chan_func *,
+			const struct nv50_disp_chan_mthd *,
+			struct nv50_disp *, int chid, int head, u64 push,
+			const struct nvkm_oclass *, struct nvkm_object **);
 
 extern const struct nv50_disp_chan_func nv50_disp_pioc_func;
-extern const struct nv50_disp_chan_func gf119_disp_pioc_func;
+extern const struct nv50_disp_chan_func nv50_disp_dmac_func;
+int nv50_disp_dmac_bind(struct nv50_disp_chan *, struct nvkm_object *, u32);
+extern const struct nv50_disp_chan_func nv50_disp_core_func;
 
-extern const struct nvkm_event_func nv50_disp_chan_uevent;
-int  nv50_disp_chan_uevent_ctor(struct nvkm_object *, void *, u32,
-				struct nvkm_notify *);
-void nv50_disp_chan_uevent_send(struct nv50_disp *, int);
+extern const struct nv50_disp_chan_func gf119_disp_pioc_func;
+extern const struct nv50_disp_chan_func gf119_disp_dmac_func;
+void gf119_disp_dmac_fini(struct nv50_disp_chan *);
+int gf119_disp_dmac_bind(struct nv50_disp_chan *, struct nvkm_object *, u32);
+extern const struct nv50_disp_chan_func gf119_disp_core_func;
+void gf119_disp_core_fini(struct nv50_disp_chan *);
 
-extern const struct nvkm_event_func gf119_disp_chan_uevent;
+extern const struct nv50_disp_chan_func gp102_disp_dmac_func;
 
 int nv50_disp_curs_new_(const struct nv50_disp_chan_func *,
 			struct nv50_disp *, int ctrl, int user,
@@ -58,17 +59,17 @@ int nv50_disp_oimm_new_(const struct nv50_disp_chan_func *,
 			struct nv50_disp *, int ctrl, int user,
 			const struct nvkm_oclass *, void *argv, u32 argc,
 			struct nvkm_object **);
-int nv50_disp_base_new_(const struct nv50_disp_dmac_func *,
+int nv50_disp_base_new_(const struct nv50_disp_chan_func *,
 			const struct nv50_disp_chan_mthd *,
 			struct nv50_disp *, int chid,
 			const struct nvkm_oclass *, void *argv, u32 argc,
 			struct nvkm_object **);
-int nv50_disp_core_new_(const struct nv50_disp_dmac_func *,
+int nv50_disp_core_new_(const struct nv50_disp_chan_func *,
 			const struct nv50_disp_chan_mthd *,
 			struct nv50_disp *, int chid,
 			const struct nvkm_oclass *oclass, void *argv, u32 argc,
 			struct nvkm_object **);
-int nv50_disp_ovly_new_(const struct nv50_disp_dmac_func *,
+int nv50_disp_ovly_new_(const struct nv50_disp_chan_func *,
 			const struct nv50_disp_chan_mthd *,
 			struct nv50_disp *, int chid,
 			const struct nvkm_oclass *, void *argv, u32 argc,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg84.c
index b16857f..cfc54aa 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg84.c
@@ -21,7 +21,7 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
+#include "channv50.h"
 
 const struct nv50_disp_mthd_list
 g84_disp_core_mthd_dac = {
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg94.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg94.c
index ea5f489..e911925 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg94.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg94.c
@@ -21,7 +21,7 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
+#include "channv50.h"
 
 static const struct nv50_disp_mthd_list
 g94_disp_core_mthd_sor = {
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c
index 9e48cc3..17c6616 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c
@@ -21,7 +21,7 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
+#include "channv50.h"
 
 #include <subdev/timer.h>
 
@@ -167,10 +167,9 @@ gf119_disp_core_mthd = {
 };
 
 void
-gf119_disp_core_fini(struct nv50_disp_dmac *chan)
+gf119_disp_core_fini(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->base.disp;
-	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 
 	/* deactivate channel */
@@ -190,10 +189,9 @@ gf119_disp_core_fini(struct nv50_disp_dmac *chan)
 }
 
 static int
-gf119_disp_core_init(struct nv50_disp_dmac *chan)
+gf119_disp_core_init(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->base.disp;
-	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 
 	/* enable error reporting */
@@ -220,7 +218,7 @@ gf119_disp_core_init(struct nv50_disp_dmac *chan)
 	return 0;
 }
 
-const struct nv50_disp_dmac_func
+const struct nv50_disp_chan_func
 gf119_disp_core_func = {
 	.init = gf119_disp_core_init,
 	.fini = gf119_disp_core_fini,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregk104.c
index ca09595..5c80017 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregk104.c
@@ -21,7 +21,7 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
+#include "channv50.h"
 
 static const struct nv50_disp_mthd_list
 gk104_disp_core_mthd_head = {
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c
index 3ec353e..57d2605 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c
@@ -21,15 +21,14 @@
  *
  * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
-#include "dmacnv50.h"
+#include "channv50.h"
 
 #include <subdev/timer.h>
 
 static int
-gp102_disp_core_init(struct nv50_disp_dmac *chan)
+gp102_disp_core_init(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->base.disp;
-	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 
 	/* enable error reporting */
@@ -56,7 +55,7 @@ gp102_disp_core_init(struct nv50_disp_dmac *chan)
 	return 0;
 }
 
-static const struct nv50_disp_dmac_func
+static const struct nv50_disp_chan_func
 gp102_disp_core_func = {
 	.init = gp102_disp_core_init,
 	.fini = gf119_disp_core_fini,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c
index 8cdcf5b5..d648c4d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c
@@ -21,7 +21,7 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
+#include "channv50.h"
 
 #include <core/client.h>
 #include <subdev/timer.h>
@@ -30,7 +30,7 @@
 #include <nvif/unpack.h>
 
 int
-nv50_disp_core_new_(const struct nv50_disp_dmac_func *func,
+nv50_disp_core_new_(const struct nv50_disp_chan_func *func,
 		    const struct nv50_disp_chan_mthd *mthd,
 		    struct nv50_disp *disp, int chid,
 		    const struct nvkm_oclass *oclass, void *argv, u32 argc,
@@ -164,10 +164,9 @@ nv50_disp_core_mthd = {
 };
 
 static void
-nv50_disp_core_fini(struct nv50_disp_dmac *chan)
+nv50_disp_core_fini(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->base.disp;
-	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 
 	/* deactivate channel */
@@ -186,10 +185,9 @@ nv50_disp_core_fini(struct nv50_disp_dmac *chan)
 }
 
 static int
-nv50_disp_core_init(struct nv50_disp_dmac *chan)
+nv50_disp_core_init(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->base.disp;
-	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 
 	/* enable error reporting */
@@ -222,7 +220,7 @@ nv50_disp_core_init(struct nv50_disp_dmac *chan)
 	return 0;
 }
 
-const struct nv50_disp_dmac_func
+const struct nv50_disp_chan_func
 nv50_disp_core_func = {
 	.init = nv50_disp_core_init,
 	.fini = nv50_disp_core_fini,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c
index b73bcc3..f69749a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c
@@ -21,29 +21,27 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
-#include "rootnv50.h"
+#include "channv50.h"
 
 #include <core/ramht.h>
 #include <subdev/timer.h>
 
 int
-gf119_disp_dmac_bind(struct nv50_disp_dmac *chan,
+gf119_disp_dmac_bind(struct nv50_disp_chan *chan,
 		     struct nvkm_object *object, u32 handle)
 {
-	return nvkm_ramht_insert(chan->base.disp->ramht, object,
-				 chan->base.chid.user, -9, handle,
-				 chan->base.chid.user << 27 | 0x00000001);
+	return nvkm_ramht_insert(chan->disp->ramht, object,
+				 chan->chid.user, -9, handle,
+				 chan->chid.user << 27 | 0x00000001);
 }
 
 void
-gf119_disp_dmac_fini(struct nv50_disp_dmac *chan)
+gf119_disp_dmac_fini(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->base.disp;
-	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
-	int ctrl = chan->base.chid.ctrl;
-	int user = chan->base.chid.user;
+	int ctrl = chan->chid.ctrl;
+	int user = chan->chid.user;
 
 	/* deactivate channel */
 	nvkm_mask(device, 0x610490 + (ctrl * 0x0010), 0x00001010, 0x00001000);
@@ -62,13 +60,12 @@ gf119_disp_dmac_fini(struct nv50_disp_dmac *chan)
 }
 
 static int
-gf119_disp_dmac_init(struct nv50_disp_dmac *chan)
+gf119_disp_dmac_init(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->base.disp;
-	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
-	int ctrl = chan->base.chid.ctrl;
-	int user = chan->base.chid.user;
+	int ctrl = chan->chid.ctrl;
+	int user = chan->chid.user;
 
 	/* enable error reporting */
 	nvkm_mask(device, 0x6100a0, 0x00000001 << user, 0x00000001 << user);
@@ -94,7 +91,7 @@ gf119_disp_dmac_init(struct nv50_disp_dmac *chan)
 	return 0;
 }
 
-const struct nv50_disp_dmac_func
+const struct nv50_disp_chan_func
 gf119_disp_dmac_func = {
 	.init = gf119_disp_dmac_init,
 	.fini = gf119_disp_dmac_fini,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgp102.c
index 62e9b84..22fa592 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgp102.c
@@ -21,19 +21,17 @@
  *
  * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
-#include "dmacnv50.h"
-#include "rootnv50.h"
+#include "channv50.h"
 
 #include <subdev/timer.h>
 
 static int
-gp102_disp_dmac_init(struct nv50_disp_dmac *chan)
+gp102_disp_dmac_init(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->base.disp;
-	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
-	int ctrl = chan->base.chid.ctrl;
-	int user = chan->base.chid.user;
+	int ctrl = chan->chid.ctrl;
+	int user = chan->chid.user;
 
 	/* enable error reporting */
 	nvkm_mask(device, 0x6100a0, 0x00000001 << user, 0x00000001 << user);
@@ -59,7 +57,7 @@ gp102_disp_dmac_init(struct nv50_disp_dmac *chan)
 	return 0;
 }
 
-const struct nv50_disp_dmac_func
+const struct nv50_disp_chan_func
 gp102_disp_dmac_func = {
 	.init = gp102_disp_dmac_init,
 	.fini = gf119_disp_dmac_fini,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c
index d081947..5cd08ca 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c
@@ -21,117 +21,16 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
-#include "rootnv50.h"
+#include "channv50.h"
 
 #include <core/client.h>
-#include <core/oproxy.h>
 #include <core/ramht.h>
 #include <subdev/fb.h>
 #include <subdev/timer.h>
 #include <engine/dma.h>
 
-struct nv50_disp_dmac_object {
-	struct nvkm_oproxy oproxy;
-	struct nv50_disp *disp;
-	int hash;
-};
-
-static void
-nv50_disp_dmac_child_del_(struct nvkm_oproxy *base)
-{
-	struct nv50_disp_dmac_object *object =
-		container_of(base, typeof(*object), oproxy);
-	nvkm_ramht_remove(object->disp->ramht, object->hash);
-}
-
-static const struct nvkm_oproxy_func
-nv50_disp_dmac_child_func_ = {
-	.dtor[0] = nv50_disp_dmac_child_del_,
-};
-
-static int
-nv50_disp_dmac_child_new_(struct nv50_disp_chan *base,
-			  const struct nvkm_oclass *oclass,
-			  void *data, u32 size, struct nvkm_object **pobject)
-{
-	struct nv50_disp_dmac *chan = nv50_disp_dmac(base);
-	struct nv50_disp *disp = chan->base.disp;
-	struct nvkm_device *device = disp->base.engine.subdev.device;
-	const struct nvkm_device_oclass *sclass = oclass->priv;
-	struct nv50_disp_dmac_object *object;
-	int ret;
-
-	if (!(object = kzalloc(sizeof(*object), GFP_KERNEL)))
-		return -ENOMEM;
-	nvkm_oproxy_ctor(&nv50_disp_dmac_child_func_, oclass, &object->oproxy);
-	object->disp = disp;
-	*pobject = &object->oproxy.base;
-
-	ret = sclass->ctor(device, oclass, data, size, &object->oproxy.object);
-	if (ret)
-		return ret;
-
-	object->hash = chan->func->bind(chan, object->oproxy.object,
-					      oclass->handle);
-	if (object->hash < 0)
-		return object->hash;
-
-	return 0;
-}
-
-static int
-nv50_disp_dmac_child_get_(struct nv50_disp_chan *base, int index,
-			  struct nvkm_oclass *sclass)
-{
-	struct nv50_disp_dmac *chan = nv50_disp_dmac(base);
-	struct nv50_disp *disp = chan->base.disp;
-	struct nvkm_device *device = disp->base.engine.subdev.device;
-	const struct nvkm_device_oclass *oclass = NULL;
-
-	sclass->engine = nvkm_device_engine(device, NVKM_ENGINE_DMAOBJ);
-	if (sclass->engine && sclass->engine->func->base.sclass) {
-		sclass->engine->func->base.sclass(sclass, index, &oclass);
-		if (oclass) {
-			sclass->priv = oclass;
-			return 0;
-		}
-	}
-
-	return -EINVAL;
-}
-
-static void
-nv50_disp_dmac_fini_(struct nv50_disp_chan *base)
-{
-	struct nv50_disp_dmac *chan = nv50_disp_dmac(base);
-	chan->func->fini(chan);
-}
-
-static int
-nv50_disp_dmac_init_(struct nv50_disp_chan *base)
-{
-	struct nv50_disp_dmac *chan = nv50_disp_dmac(base);
-	return chan->func->init(chan);
-}
-
-static void *
-nv50_disp_dmac_dtor_(struct nv50_disp_chan *base)
-{
-	return nv50_disp_dmac(base);
-}
-
-static const struct nv50_disp_chan_func
-nv50_disp_dmac_func_ = {
-	.dtor = nv50_disp_dmac_dtor_,
-	.init = nv50_disp_dmac_init_,
-	.fini = nv50_disp_dmac_fini_,
-	.child_get = nv50_disp_dmac_child_get_,
-	.child_new = nv50_disp_dmac_child_new_,
-};
-
 int
-nv50_disp_dmac_new_(const struct nv50_disp_dmac_func *func,
+nv50_disp_dmac_new_(const struct nv50_disp_chan_func *func,
 		    const struct nv50_disp_chan_mthd *mthd,
 		    struct nv50_disp *disp, int chid, int head, u64 push,
 		    const struct nvkm_oclass *oclass,
@@ -139,16 +38,12 @@ nv50_disp_dmac_new_(const struct nv50_disp_dmac_func *func,
 {
 	struct nvkm_client *client = oclass->client;
 	struct nvkm_dmaobj *dmaobj;
-	struct nv50_disp_dmac *chan;
+	struct nv50_disp_chan *chan;
 	int ret;
 
-	if (!(chan = kzalloc(sizeof(*chan), GFP_KERNEL)))
-		return -ENOMEM;
-	*pobject = &chan->base.object;
-	chan->func = func;
-
-	ret = nv50_disp_chan_ctor(&nv50_disp_dmac_func_, mthd, disp,
-				  chid, chid, head, oclass, &chan->base);
+	ret = nv50_disp_chan_new_(func, mthd, disp, chid, chid, head, oclass,
+				  pobject);
+	chan = nv50_disp_chan(*pobject);
 	if (ret)
 		return ret;
 
@@ -174,23 +69,22 @@ nv50_disp_dmac_new_(const struct nv50_disp_dmac_func *func,
 }
 
 int
-nv50_disp_dmac_bind(struct nv50_disp_dmac *chan,
+nv50_disp_dmac_bind(struct nv50_disp_chan *chan,
 		    struct nvkm_object *object, u32 handle)
 {
-	return nvkm_ramht_insert(chan->base.disp->ramht, object,
-				 chan->base.chid.user, -10, handle,
-				 chan->base.chid.user << 28 |
-				 chan->base.chid.user);
+	return nvkm_ramht_insert(chan->disp->ramht, object,
+				 chan->chid.user, -10, handle,
+				 chan->chid.user << 28 |
+				 chan->chid.user);
 }
 
 static void
-nv50_disp_dmac_fini(struct nv50_disp_dmac *chan)
+nv50_disp_dmac_fini(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->base.disp;
-	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
-	int ctrl = chan->base.chid.ctrl;
-	int user = chan->base.chid.user;
+	int ctrl = chan->chid.ctrl;
+	int user = chan->chid.user;
 
 	/* deactivate channel */
 	nvkm_mask(device, 0x610200 + (ctrl * 0x0010), 0x00001010, 0x00001000);
@@ -208,13 +102,12 @@ nv50_disp_dmac_fini(struct nv50_disp_dmac *chan)
 }
 
 static int
-nv50_disp_dmac_init(struct nv50_disp_dmac *chan)
+nv50_disp_dmac_init(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->base.disp;
-	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
-	int ctrl = chan->base.chid.ctrl;
-	int user = chan->base.chid.user;
+	int ctrl = chan->chid.ctrl;
+	int user = chan->chid.user;
 
 	/* enable error reporting */
 	nvkm_mask(device, 0x610028, 0x00010000 << user, 0x00010000 << user);
@@ -240,7 +133,7 @@ nv50_disp_dmac_init(struct nv50_disp_dmac *chan)
 	return 0;
 }
 
-const struct nv50_disp_dmac_func
+const struct nv50_disp_chan_func
 nv50_disp_dmac_func = {
 	.init = nv50_disp_dmac_init,
 	.fini = nv50_disp_dmac_fini,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h
deleted file mode 100644
index feeb588..0000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h
+++ /dev/null
@@ -1,35 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-#ifndef __NV50_DISP_DMAC_H__
-#define __NV50_DISP_DMAC_H__
-#define nv50_disp_dmac(p) container_of((p), struct nv50_disp_dmac, base)
-#include "channv50.h"
-
-struct nv50_disp_dmac {
-	const struct nv50_disp_dmac_func *func;
-	struct nv50_disp_chan base;
-	u32 push;
-};
-
-struct nv50_disp_dmac_func {
-	int  (*init)(struct nv50_disp_dmac *);
-	void (*fini)(struct nv50_disp_dmac *);
-	int  (*bind)(struct nv50_disp_dmac *, struct nvkm_object *, u32 handle);
-};
-
-int nv50_disp_dmac_new_(const struct nv50_disp_dmac_func *,
-			const struct nv50_disp_chan_mthd *,
-			struct nv50_disp *, int chid, int head, u64 push,
-			const struct nvkm_oclass *, struct nvkm_object **);
-
-extern const struct nv50_disp_dmac_func nv50_disp_dmac_func;
-int nv50_disp_dmac_bind(struct nv50_disp_dmac *, struct nvkm_object *, u32);
-extern const struct nv50_disp_dmac_func nv50_disp_core_func;
-
-extern const struct nv50_disp_dmac_func gf119_disp_dmac_func;
-void gf119_disp_dmac_fini(struct nv50_disp_dmac *);
-int gf119_disp_dmac_bind(struct nv50_disp_dmac *, struct nvkm_object *, u32);
-extern const struct nv50_disp_dmac_func gf119_disp_core_func;
-void gf119_disp_core_fini(struct nv50_disp_dmac *);
-
-extern const struct nv50_disp_dmac_func gp102_disp_dmac_func;
-#endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c
index 382e6a6..794e909 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c
@@ -24,6 +24,7 @@
 #include "nv50.h"
 #include "head.h"
 #include "ior.h"
+#include "channv50.h"
 #include "rootnv50.h"
 
 #include <core/ramht.h>
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp102.c
index 0a2c5b5..3468dde 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp102.c
@@ -24,6 +24,7 @@
 #include "nv50.h"
 #include "head.h"
 #include "ior.h"
+#include "channv50.h"
 #include "rootnv50.h"
 
 static void
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
index bb622d0..77aa2c8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
@@ -79,4 +79,11 @@ void gf119_disp_intr_error(struct nv50_disp *, int);
 void nv50_disp_dptmds_war_2(struct nv50_disp *, struct dcb_output *);
 void nv50_disp_dptmds_war_3(struct nv50_disp *, struct dcb_output *);
 void nv50_disp_update_sppll1(struct nv50_disp *);
+
+extern const struct nvkm_event_func nv50_disp_chan_uevent;
+int  nv50_disp_chan_uevent_ctor(struct nvkm_object *, void *, u32,
+				struct nvkm_notify *);
+void nv50_disp_chan_uevent_send(struct nv50_disp *, int);
+
+extern const struct nvkm_event_func gf119_disp_chan_uevent;
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlyg84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlyg84.c
index 6b55cf4..31b915d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlyg84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlyg84.c
@@ -21,7 +21,7 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
+#include "channv50.h"
 
 static const struct nv50_disp_mthd_list
 g84_disp_ovly_mthd_base = {
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygf119.c
index 30901ca..83fd534 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygf119.c
@@ -21,7 +21,7 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
+#include "channv50.h"
 
 static const struct nv50_disp_mthd_list
 gf119_disp_ovly_mthd_base = {
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygk104.c
index 682c146..a7acacb 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygk104.c
@@ -21,7 +21,7 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
+#include "channv50.h"
 
 static const struct nv50_disp_mthd_list
 gk104_disp_ovly_mthd_base = {
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygp102.c
index bcc5ac4..e0eca6e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygp102.c
@@ -21,7 +21,7 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
+#include "channv50.h"
 
 int
 gp102_disp_ovly_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygt200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygt200.c
index 655deb0..dc60cd0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygt200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygt200.c
@@ -21,7 +21,7 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
+#include "channv50.h"
 
 static const struct nv50_disp_mthd_list
 gt200_disp_ovly_mthd_base = {
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlynv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlynv50.c
index 46f5df0..6974c12 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlynv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlynv50.c
@@ -21,7 +21,7 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
+#include "channv50.h"
 #include "head.h"
 
 #include <core/client.h>
@@ -30,7 +30,7 @@
 #include <nvif/unpack.h>
 
 int
-nv50_disp_ovly_new_(const struct nv50_disp_dmac_func *func,
+nv50_disp_ovly_new_(const struct nv50_disp_chan_func *func,
 		    const struct nv50_disp_chan_mthd *mthd,
 		    struct nv50_disp *disp, int chid,
 		    const struct nvkm_oclass *oclass, void *argv, u32 argc,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c
index e51f89f..1ed371f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c
@@ -22,7 +22,7 @@
  * Authors: Ben Skeggs
  */
 #include "rootnv50.h"
-#include "dmacnv50.h"
+#include "channv50.h"
 
 #include <nvif/class.h>
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c
index ed7838e..ef579eb 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c
@@ -22,7 +22,7 @@
  * Authors: Ben Skeggs
  */
 #include "rootnv50.h"
-#include "dmacnv50.h"
+#include "channv50.h"
 
 #include <nvif/class.h>
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
index ac92e65..fe01116 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
@@ -22,7 +22,7 @@
  * Authors: Ben Skeggs
  */
 #include "rootnv50.h"
-#include "dmacnv50.h"
+#include "channv50.h"
 
 #include <nvif/class.h>
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c
index 3bb6d60..9e8ffd3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c
@@ -22,7 +22,7 @@
  * Authors: Ben Skeggs
  */
 #include "rootnv50.h"
-#include "dmacnv50.h"
+#include "channv50.h"
 
 #include <nvif/class.h>
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c
index 336419815..dc85cc1 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c
@@ -22,7 +22,7 @@
  * Authors: Ben Skeggs
  */
 #include "rootnv50.h"
-#include "dmacnv50.h"
+#include "channv50.h"
 
 #include <nvif/class.h>
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c
index c53e71e..e0181ca 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c
@@ -22,7 +22,7 @@
  * Authors: Ben Skeggs
  */
 #include "rootnv50.h"
-#include "dmacnv50.h"
+#include "channv50.h"
 
 #include <nvif/class.h>
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c
index 85409d1..e5e590e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c
@@ -22,7 +22,7 @@
  * Authors: Ben Skeggs
  */
 #include "rootnv50.h"
-#include "dmacnv50.h"
+#include "channv50.h"
 
 #include <nvif/class.h>
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c
index ebfd245..762a1a9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c
@@ -22,7 +22,7 @@
  * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
 #include "rootnv50.h"
-#include "dmacnv50.h"
+#include "channv50.h"
 
 #include <nvif/class.h>
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c
index 54b5fda..c7f0094 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c
@@ -22,7 +22,7 @@
  * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
 #include "rootnv50.h"
-#include "dmacnv50.h"
+#include "channv50.h"
 
 #include <nvif/class.h>
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c
index 14ac83b..a6963654 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c
@@ -22,7 +22,7 @@
  * Authors: Ben Skeggs
  */
 #include "rootnv50.h"
-#include "dmacnv50.h"
+#include "channv50.h"
 
 #include <nvif/class.h>
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c
index cb1a208..4fe0a3a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c
@@ -22,7 +22,7 @@
  * Authors: Ben Skeggs
  */
 #include "rootnv50.h"
-#include "dmacnv50.h"
+#include "channv50.h"
 
 #include <nvif/class.h>
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
index 9d716ee7..3aa5a28 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
@@ -22,7 +22,7 @@
  * Authors: Ben Skeggs
  */
 #include "rootnv50.h"
-#include "dmacnv50.h"
+#include "channv50.h"
 #include "dp.h"
 #include "head.h"
 #include "ior.h"
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
index 12fc13b..9983a42 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
@@ -2,8 +2,8 @@
 #ifndef __NV50_DISP_ROOT_H__
 #define __NV50_DISP_ROOT_H__
 #define nv50_disp_root(p) container_of((p), struct nv50_disp_root, object)
+#include <core/object.h>
 #include "nv50.h"
-#include "dmacnv50.h"
 
 struct nv50_disp_root {
 	const struct nv50_disp_root_func *func;
-- 
cgit v1.1


From 4a8621a24a8f68ecba6e59dccad2b252fa90ba59 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/disp/nv50-: add channel interfaces to determine the user
 area

This will be required to support Volta.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 .../gpu/drm/nouveau/nvkm/engine/disp/channv50.c    | 27 +++++++++++++---------
 .../gpu/drm/nouveau/nvkm/engine/disp/channv50.h    |  2 ++
 .../gpu/drm/nouveau/nvkm/engine/disp/coregf119.c   |  1 +
 .../gpu/drm/nouveau/nvkm/engine/disp/coregp102.c   |  1 +
 .../gpu/drm/nouveau/nvkm/engine/disp/corenv50.c    |  1 +
 .../gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c   |  1 +
 .../gpu/drm/nouveau/nvkm/engine/disp/dmacgp102.c   |  1 +
 .../gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c    |  1 +
 .../gpu/drm/nouveau/nvkm/engine/disp/piocgf119.c   |  1 +
 .../gpu/drm/nouveau/nvkm/engine/disp/piocnv50.c    |  1 +
 10 files changed, 26 insertions(+), 11 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c
index 91a2f0c..8e79aa5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c
@@ -155,13 +155,20 @@ nv50_disp_chan_uevent = {
 	.fini = nv50_disp_chan_uevent_fini,
 };
 
+u64
+nv50_disp_chan_user(struct nv50_disp_chan *chan, u64 *psize)
+{
+	*psize = 0x1000;
+	return 0x640000 + (chan->chid.user * 0x1000);
+}
+
 static int
 nv50_disp_chan_rd32(struct nvkm_object *object, u64 addr, u32 *data)
 {
 	struct nv50_disp_chan *chan = nv50_disp_chan(object);
-	struct nv50_disp *disp = chan->disp;
-	struct nvkm_device *device = disp->base.engine.subdev.device;
-	*data = nvkm_rd32(device, 0x640000 + (chan->chid.user * 0x1000) + addr);
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	u64 size, base = chan->func->user(chan, &size);
+	*data = nvkm_rd32(device, base + addr);
 	return 0;
 }
 
@@ -169,9 +176,9 @@ static int
 nv50_disp_chan_wr32(struct nvkm_object *object, u64 addr, u32 data)
 {
 	struct nv50_disp_chan *chan = nv50_disp_chan(object);
-	struct nv50_disp *disp = chan->disp;
-	struct nvkm_device *device = disp->base.engine.subdev.device;
-	nvkm_wr32(device, 0x640000 + (chan->chid.user * 0x1000) + addr, data);
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	u64 size, base = chan->func->user(chan, &size);
+	nvkm_wr32(device, base + addr, data);
 	return 0;
 }
 
@@ -196,12 +203,10 @@ nv50_disp_chan_map(struct nvkm_object *object, void *argv, u32 argc,
 		   enum nvkm_object_map *type, u64 *addr, u64 *size)
 {
 	struct nv50_disp_chan *chan = nv50_disp_chan(object);
-	struct nv50_disp *disp = chan->disp;
-	struct nvkm_device *device = disp->base.engine.subdev.device;
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	const u64 base = device->func->resource_addr(device, 0);
 	*type = NVKM_OBJECT_MAP_IO;
-	*addr = device->func->resource_addr(device, 0) +
-		0x640000 + (chan->chid.user * 0x1000);
-	*size = 0x001000;
+	*addr = base + chan->func->user(chan, size);
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
index feb8a56..75ae181 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
@@ -25,6 +25,7 @@ struct nv50_disp_chan {
 struct nv50_disp_chan_func {
 	int (*init)(struct nv50_disp_chan *);
 	void (*fini)(struct nv50_disp_chan *);
+	u64 (*user)(struct nv50_disp_chan *, u64 *size);
 	int (*bind)(struct nv50_disp_chan *, struct nvkm_object *, u32 handle);
 };
 
@@ -37,6 +38,7 @@ int nv50_disp_dmac_new_(const struct nv50_disp_chan_func *,
 			struct nv50_disp *, int chid, int head, u64 push,
 			const struct nvkm_oclass *, struct nvkm_object **);
 
+u64 nv50_disp_chan_user(struct nv50_disp_chan *, u64 *);
 extern const struct nv50_disp_chan_func nv50_disp_pioc_func;
 extern const struct nv50_disp_chan_func nv50_disp_dmac_func;
 int nv50_disp_dmac_bind(struct nv50_disp_chan *, struct nvkm_object *, u32);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c
index 17c6616..9ba4a8c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c
@@ -222,6 +222,7 @@ const struct nv50_disp_chan_func
 gf119_disp_core_func = {
 	.init = gf119_disp_core_init,
 	.fini = gf119_disp_core_fini,
+	.user = nv50_disp_chan_user,
 	.bind = gf119_disp_dmac_bind,
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c
index 57d2605..aae5db4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c
@@ -59,6 +59,7 @@ static const struct nv50_disp_chan_func
 gp102_disp_core_func = {
 	.init = gp102_disp_core_init,
 	.fini = gf119_disp_core_fini,
+	.user = nv50_disp_chan_user,
 	.bind = gf119_disp_dmac_bind,
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c
index d648c4d..5fd449d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c
@@ -224,6 +224,7 @@ const struct nv50_disp_chan_func
 nv50_disp_core_func = {
 	.init = nv50_disp_core_init,
 	.fini = nv50_disp_core_fini,
+	.user = nv50_disp_chan_user,
 	.bind = nv50_disp_dmac_bind,
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c
index f69749a..2a6d072 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c
@@ -95,5 +95,6 @@ const struct nv50_disp_chan_func
 gf119_disp_dmac_func = {
 	.init = gf119_disp_dmac_init,
 	.fini = gf119_disp_dmac_fini,
+	.user = nv50_disp_chan_user,
 	.bind = gf119_disp_dmac_bind,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgp102.c
index 22fa592..7e6b308 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgp102.c
@@ -61,5 +61,6 @@ const struct nv50_disp_chan_func
 gp102_disp_dmac_func = {
 	.init = gp102_disp_dmac_init,
 	.fini = gf119_disp_dmac_fini,
+	.user = nv50_disp_chan_user,
 	.bind = gf119_disp_dmac_bind,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c
index 5cd08ca..5db26e3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c
@@ -137,5 +137,6 @@ const struct nv50_disp_chan_func
 nv50_disp_dmac_func = {
 	.init = nv50_disp_dmac_init,
 	.fini = nv50_disp_dmac_fini,
+	.user = nv50_disp_chan_user,
 	.bind = nv50_disp_dmac_bind,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocgf119.c
index 7b1e9bf..5970e40 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocgf119.c
@@ -80,4 +80,5 @@ const struct nv50_disp_chan_func
 gf119_disp_pioc_func = {
 	.init = gf119_disp_pioc_init,
 	.fini = gf119_disp_pioc_fini,
+	.user = nv50_disp_chan_user,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocnv50.c
index 60c2012..0a76bda 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocnv50.c
@@ -82,4 +82,5 @@ const struct nv50_disp_chan_func
 nv50_disp_pioc_func = {
 	.init = nv50_disp_pioc_init,
 	.fini = nv50_disp_pioc_fini,
+	.user = nv50_disp_chan_user,
 };
-- 
cgit v1.1


From a9c44a88ca2f957c755bcb2ce8b9d2e031d65f64 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/disp/nv50-: add channel interfaces to control error
 interrupts

This will be required to support Volta, but also allows us to remove code
that's duplicated for each channel type already.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/disp/changf119.c | 13 +++++++++++++
 drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c  | 11 +++++++++++
 drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h  |  3 +++
 drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c |  8 +-------
 drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c |  4 +---
 drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c  |  7 +------
 drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c |  8 +-------
 drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgp102.c |  4 +---
 drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c  |  7 +------
 drivers/gpu/drm/nouveau/nvkm/engine/disp/piocgf119.c |  8 +-------
 drivers/gpu/drm/nouveau/nvkm/engine/disp/piocnv50.c  |  1 +
 11 files changed, 35 insertions(+), 39 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/changf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/changf119.c
index 17a3d83..29e6dd5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/changf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/changf119.c
@@ -47,3 +47,16 @@ gf119_disp_chan_uevent = {
 	.init = gf119_disp_chan_uevent_init,
 	.fini = gf119_disp_chan_uevent_fini,
 };
+
+void
+gf119_disp_chan_intr(struct nv50_disp_chan *chan, bool en)
+{
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	const u64 mask = 0x00000001 << chan->chid.user;
+	if (!en) {
+		nvkm_mask(device, 0x610090, mask, 0x00000000);
+		nvkm_mask(device, 0x6100a0, mask, 0x00000000);
+	} else {
+		nvkm_mask(device, 0x6100a0, mask, mask);
+	}
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c
index 8e79aa5..53c3ed6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c
@@ -162,6 +162,15 @@ nv50_disp_chan_user(struct nv50_disp_chan *chan, u64 *psize)
 	return 0x640000 + (chan->chid.user * 0x1000);
 }
 
+void
+nv50_disp_chan_intr(struct nv50_disp_chan *chan, bool en)
+{
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	const u64 mask = 0x00010001 << chan->chid.user;
+	const u64 data = en ? 0x00010000 : 0x00000000;
+	nvkm_mask(device, 0x610028, mask, data);
+}
+
 static int
 nv50_disp_chan_rd32(struct nvkm_object *object, u64 addr, u32 *data)
 {
@@ -288,6 +297,7 @@ nv50_disp_chan_fini(struct nvkm_object *object, bool suspend)
 {
 	struct nv50_disp_chan *chan = nv50_disp_chan(object);
 	chan->func->fini(chan);
+	chan->func->intr(chan, false);
 	return 0;
 }
 
@@ -295,6 +305,7 @@ static int
 nv50_disp_chan_init(struct nvkm_object *object)
 {
 	struct nv50_disp_chan *chan = nv50_disp_chan(object);
+	chan->func->intr(chan, true);
 	return chan->func->init(chan);
 }
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
index 75ae181..2a48243 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
@@ -25,6 +25,7 @@ struct nv50_disp_chan {
 struct nv50_disp_chan_func {
 	int (*init)(struct nv50_disp_chan *);
 	void (*fini)(struct nv50_disp_chan *);
+	void (*intr)(struct nv50_disp_chan *, bool en);
 	u64 (*user)(struct nv50_disp_chan *, u64 *size);
 	int (*bind)(struct nv50_disp_chan *, struct nvkm_object *, u32 handle);
 };
@@ -38,12 +39,14 @@ int nv50_disp_dmac_new_(const struct nv50_disp_chan_func *,
 			struct nv50_disp *, int chid, int head, u64 push,
 			const struct nvkm_oclass *, struct nvkm_object **);
 
+void nv50_disp_chan_intr(struct nv50_disp_chan *, bool);
 u64 nv50_disp_chan_user(struct nv50_disp_chan *, u64 *);
 extern const struct nv50_disp_chan_func nv50_disp_pioc_func;
 extern const struct nv50_disp_chan_func nv50_disp_dmac_func;
 int nv50_disp_dmac_bind(struct nv50_disp_chan *, struct nvkm_object *, u32);
 extern const struct nv50_disp_chan_func nv50_disp_core_func;
 
+void gf119_disp_chan_intr(struct nv50_disp_chan *, bool);
 extern const struct nv50_disp_chan_func gf119_disp_pioc_func;
 extern const struct nv50_disp_chan_func gf119_disp_dmac_func;
 void gf119_disp_dmac_fini(struct nv50_disp_chan *);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c
index 9ba4a8c..d162b9cf4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c
@@ -182,10 +182,6 @@ gf119_disp_core_fini(struct nv50_disp_chan *chan)
 		nvkm_error(subdev, "core fini: %08x\n",
 			   nvkm_rd32(device, 0x610490));
 	}
-
-	/* disable error reporting and completion notification */
-	nvkm_mask(device, 0x610090, 0x00000001, 0x00000000);
-	nvkm_mask(device, 0x6100a0, 0x00000001, 0x00000000);
 }
 
 static int
@@ -194,9 +190,6 @@ gf119_disp_core_init(struct nv50_disp_chan *chan)
 	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 
-	/* enable error reporting */
-	nvkm_mask(device, 0x6100a0, 0x00000001, 0x00000001);
-
 	/* initialise channel for dma command submission */
 	nvkm_wr32(device, 0x610494, chan->push);
 	nvkm_wr32(device, 0x610498, 0x00010000);
@@ -222,6 +215,7 @@ const struct nv50_disp_chan_func
 gf119_disp_core_func = {
 	.init = gf119_disp_core_init,
 	.fini = gf119_disp_core_fini,
+	.intr = gf119_disp_chan_intr,
 	.user = nv50_disp_chan_user,
 	.bind = gf119_disp_dmac_bind,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c
index aae5db4..5b7f993 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c
@@ -31,9 +31,6 @@ gp102_disp_core_init(struct nv50_disp_chan *chan)
 	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 
-	/* enable error reporting */
-	nvkm_mask(device, 0x6100a0, 0x00000001, 0x00000001);
-
 	/* initialise channel for dma command submission */
 	nvkm_wr32(device, 0x611494, chan->push);
 	nvkm_wr32(device, 0x611498, 0x00010000);
@@ -59,6 +56,7 @@ static const struct nv50_disp_chan_func
 gp102_disp_core_func = {
 	.init = gp102_disp_core_init,
 	.fini = gf119_disp_core_fini,
+	.intr = gf119_disp_chan_intr,
 	.user = nv50_disp_chan_user,
 	.bind = gf119_disp_dmac_bind,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c
index 5fd449d..55db9a2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c
@@ -179,9 +179,6 @@ nv50_disp_core_fini(struct nv50_disp_chan *chan)
 		nvkm_error(subdev, "core fini: %08x\n",
 			   nvkm_rd32(device, 0x610200));
 	}
-
-	/* disable error reporting and completion notifications */
-	nvkm_mask(device, 0x610028, 0x00010001, 0x00000000);
 }
 
 static int
@@ -190,9 +187,6 @@ nv50_disp_core_init(struct nv50_disp_chan *chan)
 	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 
-	/* enable error reporting */
-	nvkm_mask(device, 0x610028, 0x00010000, 0x00010000);
-
 	/* attempt to unstick channel from some unknown state */
 	if ((nvkm_rd32(device, 0x610200) & 0x009f0000) == 0x00020000)
 		nvkm_mask(device, 0x610200, 0x00800000, 0x00800000);
@@ -224,6 +218,7 @@ const struct nv50_disp_chan_func
 nv50_disp_core_func = {
 	.init = nv50_disp_core_init,
 	.fini = nv50_disp_core_fini,
+	.intr = nv50_disp_chan_intr,
 	.user = nv50_disp_chan_user,
 	.bind = nv50_disp_dmac_bind,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c
index 2a6d072..edf7dd0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c
@@ -53,10 +53,6 @@ gf119_disp_dmac_fini(struct nv50_disp_chan *chan)
 		nvkm_error(subdev, "ch %d fini: %08x\n", user,
 			   nvkm_rd32(device, 0x610490 + (ctrl * 0x10)));
 	}
-
-	/* disable error reporting and completion notification */
-	nvkm_mask(device, 0x610090, 0x00000001 << user, 0x00000000);
-	nvkm_mask(device, 0x6100a0, 0x00000001 << user, 0x00000000);
 }
 
 static int
@@ -67,9 +63,6 @@ gf119_disp_dmac_init(struct nv50_disp_chan *chan)
 	int ctrl = chan->chid.ctrl;
 	int user = chan->chid.user;
 
-	/* enable error reporting */
-	nvkm_mask(device, 0x6100a0, 0x00000001 << user, 0x00000001 << user);
-
 	/* initialise channel for dma command submission */
 	nvkm_wr32(device, 0x610494 + (ctrl * 0x0010), chan->push);
 	nvkm_wr32(device, 0x610498 + (ctrl * 0x0010), 0x00010000);
@@ -95,6 +88,7 @@ const struct nv50_disp_chan_func
 gf119_disp_dmac_func = {
 	.init = gf119_disp_dmac_init,
 	.fini = gf119_disp_dmac_fini,
+	.intr = gf119_disp_chan_intr,
 	.user = nv50_disp_chan_user,
 	.bind = gf119_disp_dmac_bind,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgp102.c
index 7e6b308..f21a433 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgp102.c
@@ -33,9 +33,6 @@ gp102_disp_dmac_init(struct nv50_disp_chan *chan)
 	int ctrl = chan->chid.ctrl;
 	int user = chan->chid.user;
 
-	/* enable error reporting */
-	nvkm_mask(device, 0x6100a0, 0x00000001 << user, 0x00000001 << user);
-
 	/* initialise channel for dma command submission */
 	nvkm_wr32(device, 0x611494 + (ctrl * 0x0010), chan->push);
 	nvkm_wr32(device, 0x611498 + (ctrl * 0x0010), 0x00010000);
@@ -61,6 +58,7 @@ const struct nv50_disp_chan_func
 gp102_disp_dmac_func = {
 	.init = gp102_disp_dmac_init,
 	.fini = gf119_disp_dmac_fini,
+	.intr = gf119_disp_chan_intr,
 	.user = nv50_disp_chan_user,
 	.bind = gf119_disp_dmac_bind,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c
index 5db26e3..981b98d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c
@@ -96,9 +96,6 @@ nv50_disp_dmac_fini(struct nv50_disp_chan *chan)
 		nvkm_error(subdev, "ch %d fini timeout, %08x\n", user,
 			   nvkm_rd32(device, 0x610200 + (ctrl * 0x10)));
 	}
-
-	/* disable error reporting and completion notifications */
-	nvkm_mask(device, 0x610028, 0x00010001 << user, 0x00000000 << user);
 }
 
 static int
@@ -109,9 +106,6 @@ nv50_disp_dmac_init(struct nv50_disp_chan *chan)
 	int ctrl = chan->chid.ctrl;
 	int user = chan->chid.user;
 
-	/* enable error reporting */
-	nvkm_mask(device, 0x610028, 0x00010000 << user, 0x00010000 << user);
-
 	/* initialise channel for dma command submission */
 	nvkm_wr32(device, 0x610204 + (ctrl * 0x0010), chan->push);
 	nvkm_wr32(device, 0x610208 + (ctrl * 0x0010), 0x00010000);
@@ -137,6 +131,7 @@ const struct nv50_disp_chan_func
 nv50_disp_dmac_func = {
 	.init = nv50_disp_dmac_init,
 	.fini = nv50_disp_dmac_fini,
+	.intr = nv50_disp_chan_intr,
 	.user = nv50_disp_chan_user,
 	.bind = nv50_disp_dmac_bind,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocgf119.c
index 5970e40..5296e7b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocgf119.c
@@ -43,10 +43,6 @@ gf119_disp_pioc_fini(struct nv50_disp_chan *chan)
 		nvkm_error(subdev, "ch %d fini: %08x\n", user,
 			   nvkm_rd32(device, 0x610490 + (ctrl * 0x10)));
 	}
-
-	/* disable error reporting and completion notification */
-	nvkm_mask(device, 0x610090, 0x00000001 << user, 0x00000000);
-	nvkm_mask(device, 0x6100a0, 0x00000001 << user, 0x00000000);
 }
 
 static int
@@ -58,9 +54,6 @@ gf119_disp_pioc_init(struct nv50_disp_chan *chan)
 	int ctrl = chan->chid.ctrl;
 	int user = chan->chid.user;
 
-	/* enable error reporting */
-	nvkm_mask(device, 0x6100a0, 0x00000001 << user, 0x00000001 << user);
-
 	/* activate channel */
 	nvkm_wr32(device, 0x610490 + (ctrl * 0x10), 0x00000001);
 	if (nvkm_msec(device, 2000,
@@ -80,5 +73,6 @@ const struct nv50_disp_chan_func
 gf119_disp_pioc_func = {
 	.init = gf119_disp_pioc_init,
 	.fini = gf119_disp_pioc_fini,
+	.intr = gf119_disp_chan_intr,
 	.user = nv50_disp_chan_user,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocnv50.c
index 0a76bda..4faed6f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocnv50.c
@@ -82,5 +82,6 @@ const struct nv50_disp_chan_func
 nv50_disp_pioc_func = {
 	.init = nv50_disp_pioc_init,
 	.fini = nv50_disp_pioc_fini,
+	.intr = nv50_disp_chan_intr,
 	.user = nv50_disp_chan_user,
 };
-- 
cgit v1.1


From f5650478ab07c0921127a6a0735253b64073e978 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/disp/nv50-: pass nvkm_memory objects for channel push
 buffers

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/include/nvif/mem.h         |  2 +
 drivers/gpu/drm/nouveau/include/nvif/mmu.h         |  1 +
 drivers/gpu/drm/nouveau/nv50_display.c             | 48 ++++++++--------------
 drivers/gpu/drm/nouveau/nvif/mem.c                 | 13 ++++++
 drivers/gpu/drm/nouveau/nvif/mmu.c                 | 11 +++++
 .../gpu/drm/nouveau/nvkm/engine/disp/channv50.c    |  1 +
 .../gpu/drm/nouveau/nvkm/engine/disp/channv50.h    |  1 +
 .../gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c    | 22 +++++-----
 8 files changed, 55 insertions(+), 44 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/include/nvif/mem.h b/drivers/gpu/drm/nouveau/include/nvif/mem.h
index b542fe3..80ee4ab 100644
--- a/drivers/gpu/drm/nouveau/include/nvif/mem.h
+++ b/drivers/gpu/drm/nouveau/include/nvif/mem.h
@@ -15,4 +15,6 @@ int nvif_mem_init_type(struct nvif_mmu *mmu, s32 oclass, int type, u8 page,
 int nvif_mem_init(struct nvif_mmu *mmu, s32 oclass, u8 type, u8 page,
 		  u64 size, void *argv, u32 argc, struct nvif_mem *);
 void nvif_mem_fini(struct nvif_mem *);
+
+int nvif_mem_init_map(struct nvif_mmu *, u8 type, u64 size, struct nvif_mem *);
 #endif
diff --git a/drivers/gpu/drm/nouveau/include/nvif/mmu.h b/drivers/gpu/drm/nouveau/include/nvif/mmu.h
index c8cd5b5..747ecf6 100644
--- a/drivers/gpu/drm/nouveau/include/nvif/mmu.h
+++ b/drivers/gpu/drm/nouveau/include/nvif/mmu.h
@@ -8,6 +8,7 @@ struct nvif_mmu {
 	u8  heap_nr;
 	u8  type_nr;
 	u16 kind_nr;
+	s32 mem;
 
 	struct {
 		u64 size;
diff --git a/drivers/gpu/drm/nouveau/nv50_display.c b/drivers/gpu/drm/nouveau/nv50_display.c
index 2b3ccd8..e90330e 100644
--- a/drivers/gpu/drm/nouveau/nv50_display.c
+++ b/drivers/gpu/drm/nouveau/nv50_display.c
@@ -34,6 +34,8 @@
 #include <drm/drm_plane_helper.h>
 #include <drm/drm_edid.h>
 
+#include <nvif/mem.h>
+
 #include <nvif/class.h>
 #include <nvif/cl0002.h>
 #include <nvif/cl5070.h>
@@ -400,7 +402,8 @@ struct nv50_dmac_ctxdma {
 
 struct nv50_dmac {
 	struct nv50_chan base;
-	dma_addr_t handle;
+
+	struct nvif_mem push;
 	u32 *ptr;
 
 	struct nvif_object sync;
@@ -482,9 +485,8 @@ nv50_dmac_ctxdma_new(struct nv50_dmac *dmac, struct nouveau_framebuffer *fb)
 }
 
 static void
-nv50_dmac_destroy(struct nv50_dmac *dmac, struct nvif_object *disp)
+nv50_dmac_destroy(struct nv50_dmac *dmac)
 {
-	struct nvif_device *device = dmac->base.device;
 	struct nv50_dmac_ctxdma *ctxdma, *ctxtmp;
 
 	list_for_each_entry_safe(ctxdma, ctxtmp, &dmac->ctxdma, head) {
@@ -496,10 +498,7 @@ nv50_dmac_destroy(struct nv50_dmac *dmac, struct nvif_object *disp)
 
 	nv50_chan_destroy(&dmac->base);
 
-	if (dmac->ptr) {
-		struct device *dev = nvxx_device(device)->dev;
-		dma_free_coherent(dev, PAGE_SIZE, dmac->ptr, dmac->handle);
-	}
+	nvif_mem_fini(&dmac->push);
 }
 
 static int
@@ -507,33 +506,24 @@ nv50_dmac_create(struct nvif_device *device, struct nvif_object *disp,
 		 const s32 *oclass, u8 head, void *data, u32 size, u64 syncbuf,
 		 struct nv50_dmac *dmac)
 {
+	struct nouveau_cli *cli = (void *)device->object.client;
 	struct nv50_disp_core_channel_dma_v0 *args = data;
-	struct nvif_object pushbuf;
 	int ret;
 
 	mutex_init(&dmac->lock);
 	INIT_LIST_HEAD(&dmac->ctxdma);
 
-	dmac->ptr = dma_alloc_coherent(nvxx_device(device)->dev, PAGE_SIZE,
-				       &dmac->handle, GFP_KERNEL);
-	if (!dmac->ptr)
-		return -ENOMEM;
-
-	ret = nvif_object_init(&device->object, 0, NV_DMA_FROM_MEMORY,
-			       &(struct nv_dma_v0) {
-					.target = NV_DMA_V0_TARGET_PCI_US,
-					.access = NV_DMA_V0_ACCESS_RD,
-					.start = dmac->handle + 0x0000,
-					.limit = dmac->handle + 0x0fff,
-			       }, sizeof(struct nv_dma_v0), &pushbuf);
+	ret = nvif_mem_init_map(&cli->mmu, NVIF_MEM_COHERENT, 0x1000,
+				&dmac->push);
 	if (ret)
 		return ret;
 
-	args->pushbuf = nvif_handle(&pushbuf);
+	dmac->ptr = dmac->push.object.map.ptr;
+
+	args->pushbuf = nvif_handle(&dmac->push.object);
 
 	ret = nv50_chan_create(device, disp, oclass, head, data, size,
 			       &dmac->base);
-	nvif_object_fini(&pushbuf);
 	if (ret)
 		return ret;
 
@@ -574,9 +564,7 @@ static int
 nv50_core_create(struct nvif_device *device, struct nvif_object *disp,
 		 u64 syncbuf, struct nv50_mast *core)
 {
-	struct nv50_disp_core_channel_dma_v0 args = {
-		.pushbuf = 0xb0007d00,
-	};
+	struct nv50_disp_core_channel_dma_v0 args = {};
 	static const s32 oclass[] = {
 		GP102_DISP_CORE_CHANNEL_DMA,
 		GP100_DISP_CORE_CHANNEL_DMA,
@@ -612,7 +600,6 @@ nv50_base_create(struct nvif_device *device, struct nvif_object *disp,
 		 int head, u64 syncbuf, struct nv50_sync *base)
 {
 	struct nv50_disp_base_channel_dma_v0 args = {
-		.pushbuf = 0xb0007c00 | head,
 		.head = head,
 	};
 	static const s32 oclass[] = {
@@ -643,7 +630,6 @@ nv50_ovly_create(struct nvif_device *device, struct nvif_object *disp,
 		 int head, u64 syncbuf, struct nv50_ovly *ovly)
 {
 	struct nv50_disp_overlay_channel_dma_v0 args = {
-		.pushbuf = 0xb0007e00 | head,
 		.head = head,
 	};
 	static const s32 oclass[] = {
@@ -1472,9 +1458,8 @@ nv50_base_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
 static void *
 nv50_base_dtor(struct nv50_wndw *wndw)
 {
-	struct nv50_disp *disp = nv50_disp(wndw->plane.dev);
 	struct nv50_base *base = nv50_base(wndw);
-	nv50_dmac_destroy(&base->chan.base, disp->disp);
+	nv50_dmac_destroy(&base->chan.base);
 	return base;
 }
 
@@ -2354,11 +2339,10 @@ nv50_head_reset(struct drm_crtc *crtc)
 static void
 nv50_head_destroy(struct drm_crtc *crtc)
 {
-	struct nv50_disp *disp = nv50_disp(crtc->dev);
 	struct nv50_head *head = nv50_head(crtc);
 	int i;
 
-	nv50_dmac_destroy(&head->ovly.base, disp->disp);
+	nv50_dmac_destroy(&head->ovly.base);
 	nv50_pioc_destroy(&head->oimm.base);
 
 	for (i = 0; i < ARRAY_SIZE(head->lut.nvbo); i++)
@@ -4430,7 +4414,7 @@ nv50_display_destroy(struct drm_device *dev)
 {
 	struct nv50_disp *disp = nv50_disp(dev);
 
-	nv50_dmac_destroy(&disp->mast.base, disp->disp);
+	nv50_dmac_destroy(&disp->mast.base);
 
 	nouveau_bo_unmap(disp->sync);
 	if (disp->sync)
diff --git a/drivers/gpu/drm/nouveau/nvif/mem.c b/drivers/gpu/drm/nouveau/nvif/mem.c
index 0f9382c..b6ebb3b 100644
--- a/drivers/gpu/drm/nouveau/nvif/mem.c
+++ b/drivers/gpu/drm/nouveau/nvif/mem.c
@@ -24,6 +24,19 @@
 
 #include <nvif/if000a.h>
 
+int
+nvif_mem_init_map(struct nvif_mmu *mmu, u8 type, u64 size, struct nvif_mem *mem)
+{
+	int ret = nvif_mem_init(mmu, mmu->mem, NVIF_MEM_MAPPABLE | type, 0,
+				size, NULL, 0, mem);
+	if (ret == 0) {
+		ret = nvif_object_map(&mem->object, NULL, 0);
+		if (ret)
+			nvif_mem_fini(mem);
+	}
+	return ret;
+}
+
 void
 nvif_mem_fini(struct nvif_mem *mem)
 {
diff --git a/drivers/gpu/drm/nouveau/nvif/mmu.c b/drivers/gpu/drm/nouveau/nvif/mmu.c
index 15d0dcb..358ac4f 100644
--- a/drivers/gpu/drm/nouveau/nvif/mmu.c
+++ b/drivers/gpu/drm/nouveau/nvif/mmu.c
@@ -36,6 +36,12 @@ nvif_mmu_fini(struct nvif_mmu *mmu)
 int
 nvif_mmu_init(struct nvif_object *parent, s32 oclass, struct nvif_mmu *mmu)
 {
+	static const struct nvif_mclass mems[] = {
+		{ NVIF_CLASS_MEM_GF100, -1 },
+		{ NVIF_CLASS_MEM_NV50 , -1 },
+		{ NVIF_CLASS_MEM_NV04 , -1 },
+		{}
+	};
 	struct nvif_mmu_v0 args;
 	int ret, i;
 
@@ -54,6 +60,11 @@ nvif_mmu_init(struct nvif_object *parent, s32 oclass, struct nvif_mmu *mmu)
 	mmu->type_nr = args.type_nr;
 	mmu->kind_nr = args.kind_nr;
 
+	ret = nvif_mclass(&mmu->object, mems);
+	if (ret < 0)
+		goto done;
+	mmu->mem = mems[ret].oclass;
+
 	mmu->heap = kmalloc(sizeof(*mmu->heap) * mmu->heap_nr, GFP_KERNEL);
 	mmu->type = kmalloc(sizeof(*mmu->type) * mmu->type_nr, GFP_KERNEL);
 	if (ret = -ENOMEM, !mmu->heap || !mmu->type)
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c
index 53c3ed6..57719f6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c
@@ -316,6 +316,7 @@ nv50_disp_chan_dtor(struct nvkm_object *object)
 	struct nv50_disp *disp = chan->disp;
 	if (chan->chid.user >= 0)
 		disp->chan[chan->chid.user] = NULL;
+	nvkm_memory_unref(&chan->memory);
 	return chan;
 }
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
index 2a48243..391b007 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
@@ -19,6 +19,7 @@ struct nv50_disp_chan {
 
 	struct nvkm_object object;
 
+	struct nvkm_memory *memory;
 	u64 push;
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c
index 981b98d..9e8a9d7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c
@@ -26,6 +26,7 @@
 #include <core/client.h>
 #include <core/ramht.h>
 #include <subdev/fb.h>
+#include <subdev/mmu.h>
 #include <subdev/timer.h>
 #include <engine/dma.h>
 
@@ -37,7 +38,6 @@ nv50_disp_dmac_new_(const struct nv50_disp_chan_func *func,
 		    struct nvkm_object **pobject)
 {
 	struct nvkm_client *client = oclass->client;
-	struct nvkm_dmaobj *dmaobj;
 	struct nv50_disp_chan *chan;
 	int ret;
 
@@ -47,24 +47,22 @@ nv50_disp_dmac_new_(const struct nv50_disp_chan_func *func,
 	if (ret)
 		return ret;
 
-	dmaobj = nvkm_dmaobj_search(client, push);
-	if (IS_ERR(dmaobj))
-		return PTR_ERR(dmaobj);
+	chan->memory = nvkm_umem_search(client, push);
+	if (IS_ERR(chan->memory))
+		return PTR_ERR(chan->memory);
 
-	if (dmaobj->limit - dmaobj->start != 0xfff)
+	if (nvkm_memory_size(chan->memory) < 0x1000)
 		return -EINVAL;
 
-	switch (dmaobj->target) {
-	case NV_MEM_TARGET_VRAM:
-		chan->push = 0x00000001 | dmaobj->start >> 8;
-		break;
-	case NV_MEM_TARGET_PCI_NOSNOOP:
-		chan->push = 0x00000003 | dmaobj->start >> 8;
-		break;
+	switch (nvkm_memory_target(chan->memory)) {
+	case NVKM_MEM_TARGET_VRAM: chan->push = 0x00000001; break;
+	case NVKM_MEM_TARGET_NCOH: chan->push = 0x00000002; break;
+	case NVKM_MEM_TARGET_HOST: chan->push = 0x00000003; break;
 	default:
 		return -EINVAL;
 	}
 
+	chan->push |= nvkm_memory_addr(chan->memory) >> 8;
 	return 0;
 }
 
-- 
cgit v1.1


From c5c9127b25b2946369877d16b3c208cf54d4bf54 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/device: implement a generic method to query
 device-specific properties

We have a need to fetch data from GPU-specific sub-devices that is not
tied to any particular engine object.

This commit provides the framework to support such queries.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/include/nvif/cl0080.h      | 16 ++++++++
 drivers/gpu/drm/nouveau/include/nvkm/core/engine.h |  1 +
 drivers/gpu/drm/nouveau/include/nvkm/core/subdev.h |  2 +
 drivers/gpu/drm/nouveau/nvkm/core/engine.c         | 15 +++++++
 drivers/gpu/drm/nouveau/nvkm/core/subdev.c         |  8 ++++
 drivers/gpu/drm/nouveau/nvkm/engine/device/user.c  | 47 +++++++++++++++++++++-
 6 files changed, 88 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/nouveau/include/nvif/cl0080.h b/drivers/gpu/drm/nouveau/include/nvif/cl0080.h
index 2740278..51a4af6 100644
--- a/drivers/gpu/drm/nouveau/include/nvif/cl0080.h
+++ b/drivers/gpu/drm/nouveau/include/nvif/cl0080.h
@@ -39,9 +39,25 @@ struct nv_device_info_v0 {
 	char  name[64];
 };
 
+struct nv_device_info_v1 {
+	__u8  version;
+	__u8  count;
+	__u8  pad02[6];
+	struct nv_device_info_v1_data {
+		__u64 mthd; /* NV_DEVICE_INFO_* (see below). */
+		__u64 data;
+	} data[];
+};
+
 struct nv_device_time_v0 {
 	__u8  version;
 	__u8  pad01[7];
 	__u64 time;
 };
+
+#define NV_DEVICE_INFO_UNIT                               (0xffffffffULL << 32)
+#define NV_DEVICE_INFO(n)                          ((n) | (0x00000000ULL << 32))
+
+/* This will be returned for unsupported queries. */
+#define NV_DEVICE_INFO_INVALID                                           ~0ULL
 #endif
diff --git a/drivers/gpu/drm/nouveau/include/nvkm/core/engine.h b/drivers/gpu/drm/nouveau/include/nvkm/core/engine.h
index ebf8473..8a2be5b 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/core/engine.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/core/engine.h
@@ -18,6 +18,7 @@ struct nvkm_engine_func {
 	void *(*dtor)(struct nvkm_engine *);
 	void (*preinit)(struct nvkm_engine *);
 	int (*oneinit)(struct nvkm_engine *);
+	int (*info)(struct nvkm_engine *, u64 mthd, u64 *data);
 	int (*init)(struct nvkm_engine *);
 	int (*fini)(struct nvkm_engine *, bool suspend);
 	void (*intr)(struct nvkm_engine *);
diff --git a/drivers/gpu/drm/nouveau/include/nvkm/core/subdev.h b/drivers/gpu/drm/nouveau/include/nvkm/core/subdev.h
index 63df229..85a0777 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/core/subdev.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/core/subdev.h
@@ -17,6 +17,7 @@ struct nvkm_subdev_func {
 	void *(*dtor)(struct nvkm_subdev *);
 	int (*preinit)(struct nvkm_subdev *);
 	int (*oneinit)(struct nvkm_subdev *);
+	int (*info)(struct nvkm_subdev *, u64 mthd, u64 *data);
 	int (*init)(struct nvkm_subdev *);
 	int (*fini)(struct nvkm_subdev *, bool suspend);
 	void (*intr)(struct nvkm_subdev *);
@@ -29,6 +30,7 @@ void nvkm_subdev_del(struct nvkm_subdev **);
 int  nvkm_subdev_preinit(struct nvkm_subdev *);
 int  nvkm_subdev_init(struct nvkm_subdev *);
 int  nvkm_subdev_fini(struct nvkm_subdev *, bool suspend);
+int  nvkm_subdev_info(struct nvkm_subdev *, u64, u64 *);
 void nvkm_subdev_intr(struct nvkm_subdev *);
 
 /* subdev logging */
diff --git a/drivers/gpu/drm/nouveau/nvkm/core/engine.c b/drivers/gpu/drm/nouveau/nvkm/core/engine.c
index 657231c..d0322ce 100644
--- a/drivers/gpu/drm/nouveau/nvkm/core/engine.c
+++ b/drivers/gpu/drm/nouveau/nvkm/core/engine.c
@@ -83,6 +83,20 @@ nvkm_engine_intr(struct nvkm_subdev *subdev)
 }
 
 static int
+nvkm_engine_info(struct nvkm_subdev *subdev, u64 mthd, u64 *data)
+{
+	struct nvkm_engine *engine = nvkm_engine(subdev);
+	if (engine->func->info) {
+		if ((engine = nvkm_engine_ref(engine))) {
+			int ret = engine->func->info(engine, mthd, data);
+			nvkm_engine_unref(&engine);
+			return ret;
+		}
+	}
+	return -ENOSYS;
+}
+
+static int
 nvkm_engine_fini(struct nvkm_subdev *subdev, bool suspend)
 {
 	struct nvkm_engine *engine = nvkm_engine(subdev);
@@ -150,6 +164,7 @@ nvkm_engine_func = {
 	.preinit = nvkm_engine_preinit,
 	.init = nvkm_engine_init,
 	.fini = nvkm_engine_fini,
+	.info = nvkm_engine_info,
 	.intr = nvkm_engine_intr,
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/core/subdev.c b/drivers/gpu/drm/nouveau/nvkm/core/subdev.c
index c707306..b96f9e2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/core/subdev.c
+++ b/drivers/gpu/drm/nouveau/nvkm/core/subdev.c
@@ -93,6 +93,14 @@ nvkm_subdev_intr(struct nvkm_subdev *subdev)
 }
 
 int
+nvkm_subdev_info(struct nvkm_subdev *subdev, u64 mthd, u64 *data)
+{
+	if (subdev->func->info)
+		return subdev->func->info(subdev, mthd, data);
+	return -ENOSYS;
+}
+
+int
 nvkm_subdev_fini(struct nvkm_subdev *subdev, bool suspend)
 {
 	struct nvkm_device *device = subdev->device;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c
index 17adcb4..3526516 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c
@@ -40,6 +40,40 @@ struct nvkm_udevice {
 };
 
 static int
+nvkm_udevice_info_subdev(struct nvkm_device *device, u64 mthd, u64 *data)
+{
+	struct nvkm_subdev *subdev;
+	enum nvkm_devidx subidx;
+
+	switch (mthd & NV_DEVICE_INFO_UNIT) {
+	default:
+		return -EINVAL;
+	}
+
+	subdev = nvkm_device_subdev(device, subidx);
+	if (subdev)
+		return nvkm_subdev_info(subdev, mthd, data);
+	return -ENODEV;
+}
+
+static void
+nvkm_udevice_info_v1(struct nvkm_device *device,
+		     struct nv_device_info_v1_data *args)
+{
+	if (args->mthd & NV_DEVICE_INFO_UNIT) {
+		if (nvkm_udevice_info_subdev(device, args->mthd, &args->data))
+			args->mthd = NV_DEVICE_INFO_INVALID;
+		return;
+	}
+
+	switch (args->mthd) {
+	default:
+		args->mthd = NV_DEVICE_INFO_INVALID;
+		break;
+	}
+}
+
+static int
 nvkm_udevice_info(struct nvkm_udevice *udev, void *data, u32 size)
 {
 	struct nvkm_object *object = &udev->object;
@@ -48,10 +82,21 @@ nvkm_udevice_info(struct nvkm_udevice *udev, void *data, u32 size)
 	struct nvkm_instmem *imem = device->imem;
 	union {
 		struct nv_device_info_v0 v0;
+		struct nv_device_info_v1 v1;
 	} *args = data;
-	int ret = -ENOSYS;
+	int ret = -ENOSYS, i;
 
 	nvif_ioctl(object, "device info size %d\n", size);
+	if (!(ret = nvif_unpack(ret, &data, &size, args->v1, 1, 1, true))) {
+		nvif_ioctl(object, "device info vers %d count %d\n",
+			   args->v1.version, args->v1.count);
+		if (args->v1.count * sizeof(args->v1.data[0]) == size) {
+			for (i = 0; i < args->v1.count; i++)
+				nvkm_udevice_info_v1(device, &args->v1.data[i]);
+			return 0;
+		}
+		return -EINVAL;
+	} else
 	if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false))) {
 		nvif_ioctl(object, "device info vers %d\n", args->v0.version);
 	} else
-- 
cgit v1.1


From 6eb01aa8988873167adc5285f4afef310d01b8fb Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/device: support querying available engines of a specific
 type

Will be used for fifo runlist selection.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/include/nvif/cl0080.h     | 19 +++++++++++++++++
 drivers/gpu/drm/nouveau/nvkm/engine/device/user.c | 25 +++++++++++++++++++++++
 2 files changed, 44 insertions(+)

diff --git a/drivers/gpu/drm/nouveau/include/nvif/cl0080.h b/drivers/gpu/drm/nouveau/include/nvif/cl0080.h
index 51a4af6..6a54cda 100644
--- a/drivers/gpu/drm/nouveau/include/nvif/cl0080.h
+++ b/drivers/gpu/drm/nouveau/include/nvif/cl0080.h
@@ -60,4 +60,23 @@ struct nv_device_time_v0 {
 
 /* This will be returned for unsupported queries. */
 #define NV_DEVICE_INFO_INVALID                                           ~0ULL
+
+/* These return a mask of available engines of particular type. */
+#define NV_DEVICE_INFO_ENGINE_SW                     NV_DEVICE_INFO(0x00000000)
+#define NV_DEVICE_INFO_ENGINE_GR                     NV_DEVICE_INFO(0x00000001)
+#define NV_DEVICE_INFO_ENGINE_MPEG                   NV_DEVICE_INFO(0x00000002)
+#define NV_DEVICE_INFO_ENGINE_ME                     NV_DEVICE_INFO(0x00000003)
+#define NV_DEVICE_INFO_ENGINE_CIPHER                 NV_DEVICE_INFO(0x00000004)
+#define NV_DEVICE_INFO_ENGINE_BSP                    NV_DEVICE_INFO(0x00000005)
+#define NV_DEVICE_INFO_ENGINE_VP                     NV_DEVICE_INFO(0x00000006)
+#define NV_DEVICE_INFO_ENGINE_CE                     NV_DEVICE_INFO(0x00000007)
+#define NV_DEVICE_INFO_ENGINE_SEC                    NV_DEVICE_INFO(0x00000008)
+#define NV_DEVICE_INFO_ENGINE_MSVLD                  NV_DEVICE_INFO(0x00000009)
+#define NV_DEVICE_INFO_ENGINE_MSPDEC                 NV_DEVICE_INFO(0x0000000a)
+#define NV_DEVICE_INFO_ENGINE_MSPPP                  NV_DEVICE_INFO(0x0000000b)
+#define NV_DEVICE_INFO_ENGINE_MSENC                  NV_DEVICE_INFO(0x0000000c)
+#define NV_DEVICE_INFO_ENGINE_VIC                    NV_DEVICE_INFO(0x0000000d)
+#define NV_DEVICE_INFO_ENGINE_SEC2                   NV_DEVICE_INFO(0x0000000e)
+#define NV_DEVICE_INFO_ENGINE_NVDEC                  NV_DEVICE_INFO(0x0000000f)
+#define NV_DEVICE_INFO_ENGINE_NVENC                  NV_DEVICE_INFO(0x00000010)
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c
index 3526516..42a552d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c
@@ -67,6 +67,31 @@ nvkm_udevice_info_v1(struct nvkm_device *device,
 	}
 
 	switch (args->mthd) {
+#define ENGINE__(A,B,C) NV_DEVICE_INFO_ENGINE_##A: { int _i;                   \
+	for (_i = (B), args->data = 0ULL; _i <= (C); _i++) {                   \
+		if (nvkm_device_engine(device, _i))                            \
+			args->data |= BIT_ULL(_i);                             \
+	}                                                                      \
+}
+#define ENGINE_A(A) ENGINE__(A, NVKM_ENGINE_##A   , NVKM_ENGINE_##A)
+#define ENGINE_B(A) ENGINE__(A, NVKM_ENGINE_##A##0, NVKM_ENGINE_##A##_LAST)
+	case ENGINE_A(SW    ); break;
+	case ENGINE_A(GR    ); break;
+	case ENGINE_A(MPEG  ); break;
+	case ENGINE_A(ME    ); break;
+	case ENGINE_A(CIPHER); break;
+	case ENGINE_A(BSP   ); break;
+	case ENGINE_A(VP    ); break;
+	case ENGINE_B(CE    ); break;
+	case ENGINE_A(SEC   ); break;
+	case ENGINE_A(MSVLD ); break;
+	case ENGINE_A(MSPDEC); break;
+	case ENGINE_A(MSPPP ); break;
+	case ENGINE_A(MSENC ); break;
+	case ENGINE_A(VIC   ); break;
+	case ENGINE_A(SEC2  ); break;
+	case ENGINE_A(NVDEC ); break;
+	case ENGINE_B(NVENC ); break;
 	default:
 		args->mthd = NV_DEVICE_INFO_INVALID;
 		break;
-- 
cgit v1.1


From eb47db4f3bb58b0143a911b29417e89f28e1b0c8 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/fifo: support channel count query

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/include/nvif/cl0080.h     |  4 ++++
 drivers/gpu/drm/nouveau/include/nvif/device.h     |  1 -
 drivers/gpu/drm/nouveau/nouveau_chan.c            | 25 +++++++++++++++++++++++
 drivers/gpu/drm/nouveau/nouveau_chan.h            |  1 +
 drivers/gpu/drm/nouveau/nouveau_drm.c             |  4 ++++
 drivers/gpu/drm/nouveau/nouveau_drv.h             |  6 ++++++
 drivers/gpu/drm/nouveau/nouveau_fence.c           | 11 +++++-----
 drivers/gpu/drm/nouveau/nouveau_fence.h           |  2 --
 drivers/gpu/drm/nouveau/nv04_fence.c              |  2 --
 drivers/gpu/drm/nouveau/nv10_fence.c              |  2 --
 drivers/gpu/drm/nouveau/nv17_fence.c              |  2 --
 drivers/gpu/drm/nouveau/nv50_fence.c              |  2 --
 drivers/gpu/drm/nouveau/nv84_fence.c              | 11 ++++------
 drivers/gpu/drm/nouveau/nvkm/engine/device/user.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c   | 14 +++++++++++++
 15 files changed, 64 insertions(+), 24 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/include/nvif/cl0080.h b/drivers/gpu/drm/nouveau/include/nvif/cl0080.h
index 6a54cda..5af610e 100644
--- a/drivers/gpu/drm/nouveau/include/nvif/cl0080.h
+++ b/drivers/gpu/drm/nouveau/include/nvif/cl0080.h
@@ -57,6 +57,7 @@ struct nv_device_time_v0 {
 
 #define NV_DEVICE_INFO_UNIT                               (0xffffffffULL << 32)
 #define NV_DEVICE_INFO(n)                          ((n) | (0x00000000ULL << 32))
+#define NV_DEVICE_FIFO(n)                          ((n) | (0x00000001ULL << 32))
 
 /* This will be returned for unsupported queries. */
 #define NV_DEVICE_INFO_INVALID                                           ~0ULL
@@ -79,4 +80,7 @@ struct nv_device_time_v0 {
 #define NV_DEVICE_INFO_ENGINE_SEC2                   NV_DEVICE_INFO(0x0000000e)
 #define NV_DEVICE_INFO_ENGINE_NVDEC                  NV_DEVICE_INFO(0x0000000f)
 #define NV_DEVICE_INFO_ENGINE_NVENC                  NV_DEVICE_INFO(0x00000010)
+
+/* Returns the number of available channels. */
+#define NV_DEVICE_FIFO_CHANNELS                      NV_DEVICE_FIFO(0x00000000)
 #endif
diff --git a/drivers/gpu/drm/nouveau/include/nvif/device.h b/drivers/gpu/drm/nouveau/include/nvif/device.h
index 6edb626..216dbd9 100644
--- a/drivers/gpu/drm/nouveau/include/nvif/device.h
+++ b/drivers/gpu/drm/nouveau/include/nvif/device.h
@@ -67,6 +67,5 @@ u64  nvif_device_time(struct nvif_device *);
 #include <engine/fifo.h>
 #include <engine/gr.h>
 
-#define nvxx_fifo(a) nvxx_device(a)->fifo
 #define nvxx_gr(a) nvxx_device(a)->gr
 #endif
diff --git a/drivers/gpu/drm/nouveau/nouveau_chan.c b/drivers/gpu/drm/nouveau/nouveau_chan.c
index af11166..db69d13 100644
--- a/drivers/gpu/drm/nouveau/nouveau_chan.c
+++ b/drivers/gpu/drm/nouveau/nouveau_chan.c
@@ -474,3 +474,28 @@ done:
 	cli->base.super = super;
 	return ret;
 }
+
+int
+nouveau_channels_init(struct nouveau_drm *drm)
+{
+	struct {
+		struct nv_device_info_v1 m;
+		struct {
+			struct nv_device_info_v1_data channels;
+		} v;
+	} args = {
+		.m.version = 1,
+		.m.count = sizeof(args.v) / sizeof(args.v.channels),
+		.v.channels.mthd = NV_DEVICE_FIFO_CHANNELS,
+	};
+	struct nvif_object *device = &drm->client.device.object;
+	int ret;
+
+	ret = nvif_object_mthd(device, NV_DEVICE_V0_INFO, &args, sizeof(args));
+	if (ret || args.v.channels.mthd == NV_DEVICE_INFO_INVALID)
+		return -ENODEV;
+
+	drm->chan.nr = args.v.channels.data;
+	drm->chan.context_base = dma_fence_context_alloc(drm->chan.nr);
+	return 0;
+}
diff --git a/drivers/gpu/drm/nouveau/nouveau_chan.h b/drivers/gpu/drm/nouveau/nouveau_chan.h
index 14607c1..64454c2 100644
--- a/drivers/gpu/drm/nouveau/nouveau_chan.h
+++ b/drivers/gpu/drm/nouveau/nouveau_chan.h
@@ -45,6 +45,7 @@ struct nouveau_channel {
 	atomic_t killed;
 };
 
+int nouveau_channels_init(struct nouveau_drm *);
 
 int  nouveau_channel_new(struct nouveau_drm *, struct nvif_device *,
 			 u32 arg0, u32 arg1, struct nouveau_channel **);
diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c
index bbbf353..dddd425 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_drm.c
@@ -307,6 +307,10 @@ nouveau_accel_init(struct nouveau_drm *drm)
 	if (nouveau_noaccel)
 		return;
 
+	ret = nouveau_channels_init(drm);
+	if (ret)
+		return;
+
 	/* initialise synchronisation routines */
 	/*XXX: this is crap, but the fence/channel stuff is a little
 	 *     backwards in some places.  this will be fixed.
diff --git a/drivers/gpu/drm/nouveau/nouveau_drv.h b/drivers/gpu/drm/nouveau/nouveau_drv.h
index 881b44b..6e1acae 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drv.h
+++ b/drivers/gpu/drm/nouveau/nouveau_drv.h
@@ -170,6 +170,12 @@ struct nouveau_drm {
 	/* synchronisation */
 	void *fence;
 
+	/* Global channel management. */
+	struct {
+		int nr;
+		u64 context_base;
+	} chan;
+
 	/* context for accelerated drm-internal operations */
 	struct nouveau_channel *cechan;
 	struct nouveau_channel *channel;
diff --git a/drivers/gpu/drm/nouveau/nouveau_fence.c b/drivers/gpu/drm/nouveau/nouveau_fence.c
index 503fa94..412d49b 100644
--- a/drivers/gpu/drm/nouveau/nouveau_fence.c
+++ b/drivers/gpu/drm/nouveau/nouveau_fence.c
@@ -74,15 +74,14 @@ nouveau_fence_signal(struct nouveau_fence *fence)
 }
 
 static struct nouveau_fence *
-nouveau_local_fence(struct dma_fence *fence, struct nouveau_drm *drm) {
-	struct nouveau_fence_priv *priv = (void*)drm->fence;
-
+nouveau_local_fence(struct dma_fence *fence, struct nouveau_drm *drm)
+{
 	if (fence->ops != &nouveau_fence_ops_legacy &&
 	    fence->ops != &nouveau_fence_ops_uevent)
 		return NULL;
 
-	if (fence->context < priv->context_base ||
-	    fence->context >= priv->context_base + priv->contexts)
+	if (fence->context < drm->chan.context_base ||
+	    fence->context >= drm->chan.context_base + drm->chan.nr)
 		return NULL;
 
 	return from_fence(fence);
@@ -176,7 +175,7 @@ nouveau_fence_context_new(struct nouveau_channel *chan, struct nouveau_fence_cha
 	INIT_LIST_HEAD(&fctx->flip);
 	INIT_LIST_HEAD(&fctx->pending);
 	spin_lock_init(&fctx->lock);
-	fctx->context = priv->context_base + chan->chid;
+	fctx->context = chan->drm->chan.context_base + chan->chid;
 
 	if (chan == chan->drm->cechan)
 		strcpy(fctx->name, "copy engine channel");
diff --git a/drivers/gpu/drm/nouveau/nouveau_fence.h b/drivers/gpu/drm/nouveau/nouveau_fence.h
index 5bd8d30..b999e60 100644
--- a/drivers/gpu/drm/nouveau/nouveau_fence.h
+++ b/drivers/gpu/drm/nouveau/nouveau_fence.h
@@ -55,8 +55,6 @@ struct nouveau_fence_priv {
 	int  (*context_new)(struct nouveau_channel *);
 	void (*context_del)(struct nouveau_channel *);
 
-	u32 contexts;
-	u64 context_base;
 	bool uevent;
 };
 
diff --git a/drivers/gpu/drm/nouveau/nv04_fence.c b/drivers/gpu/drm/nouveau/nv04_fence.c
index fa8f237..c41e82b 100644
--- a/drivers/gpu/drm/nouveau/nv04_fence.c
+++ b/drivers/gpu/drm/nouveau/nv04_fence.c
@@ -109,7 +109,5 @@ nv04_fence_create(struct nouveau_drm *drm)
 	priv->base.dtor = nv04_fence_destroy;
 	priv->base.context_new = nv04_fence_context_new;
 	priv->base.context_del = nv04_fence_context_del;
-	priv->base.contexts = 15;
-	priv->base.context_base = dma_fence_context_alloc(priv->base.contexts);
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/nv10_fence.c b/drivers/gpu/drm/nouveau/nv10_fence.c
index 2998bde..4476b71 100644
--- a/drivers/gpu/drm/nouveau/nv10_fence.c
+++ b/drivers/gpu/drm/nouveau/nv10_fence.c
@@ -103,8 +103,6 @@ nv10_fence_create(struct nouveau_drm *drm)
 	priv->base.dtor = nv10_fence_destroy;
 	priv->base.context_new = nv10_fence_context_new;
 	priv->base.context_del = nv10_fence_context_del;
-	priv->base.contexts = 31;
-	priv->base.context_base = dma_fence_context_alloc(priv->base.contexts);
 	spin_lock_init(&priv->lock);
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/nv17_fence.c b/drivers/gpu/drm/nouveau/nv17_fence.c
index 6477b70..5d613d4 100644
--- a/drivers/gpu/drm/nouveau/nv17_fence.c
+++ b/drivers/gpu/drm/nouveau/nv17_fence.c
@@ -125,8 +125,6 @@ nv17_fence_create(struct nouveau_drm *drm)
 	priv->base.resume = nv17_fence_resume;
 	priv->base.context_new = nv17_fence_context_new;
 	priv->base.context_del = nv10_fence_context_del;
-	priv->base.contexts = 31;
-	priv->base.context_base = dma_fence_context_alloc(priv->base.contexts);
 	spin_lock_init(&priv->lock);
 
 	ret = nouveau_bo_new(&drm->client, 4096, 0x1000, TTM_PL_FLAG_VRAM,
diff --git a/drivers/gpu/drm/nouveau/nv50_fence.c b/drivers/gpu/drm/nouveau/nv50_fence.c
index a369d97..a00ecc3de 100644
--- a/drivers/gpu/drm/nouveau/nv50_fence.c
+++ b/drivers/gpu/drm/nouveau/nv50_fence.c
@@ -78,8 +78,6 @@ nv50_fence_create(struct nouveau_drm *drm)
 	priv->base.resume = nv17_fence_resume;
 	priv->base.context_new = nv50_fence_context_new;
 	priv->base.context_del = nv10_fence_context_del;
-	priv->base.contexts = 127;
-	priv->base.context_base = dma_fence_context_alloc(priv->base.contexts);
 	spin_lock_init(&priv->lock);
 
 	ret = nouveau_bo_new(&drm->client, 4096, 0x1000, TTM_PL_FLAG_VRAM,
diff --git a/drivers/gpu/drm/nouveau/nv84_fence.c b/drivers/gpu/drm/nouveau/nv84_fence.c
index 5f0c0c2..0906648 100644
--- a/drivers/gpu/drm/nouveau/nv84_fence.c
+++ b/drivers/gpu/drm/nouveau/nv84_fence.c
@@ -141,9 +141,9 @@ nv84_fence_suspend(struct nouveau_drm *drm)
 	struct nv84_fence_priv *priv = drm->fence;
 	int i;
 
-	priv->suspend = vmalloc(priv->base.contexts * sizeof(u32));
+	priv->suspend = vmalloc(drm->chan.nr * sizeof(u32));
 	if (priv->suspend) {
-		for (i = 0; i < priv->base.contexts; i++)
+		for (i = 0; i < drm->chan.nr; i++)
 			priv->suspend[i] = nouveau_bo_rd32(priv->bo, i*4);
 	}
 
@@ -157,7 +157,7 @@ nv84_fence_resume(struct nouveau_drm *drm)
 	int i;
 
 	if (priv->suspend) {
-		for (i = 0; i < priv->base.contexts; i++)
+		for (i = 0; i < drm->chan.nr; i++)
 			nouveau_bo_wr32(priv->bo, i*4, priv->suspend[i]);
 		vfree(priv->suspend);
 		priv->suspend = NULL;
@@ -179,7 +179,6 @@ nv84_fence_destroy(struct nouveau_drm *drm)
 int
 nv84_fence_create(struct nouveau_drm *drm)
 {
-	struct nvkm_fifo *fifo = nvxx_fifo(&drm->client.device);
 	struct nv84_fence_priv *priv;
 	u32 domain;
 	int ret;
@@ -194,8 +193,6 @@ nv84_fence_create(struct nouveau_drm *drm)
 	priv->base.context_new = nv84_fence_context_new;
 	priv->base.context_del = nv84_fence_context_del;
 
-	priv->base.contexts = fifo->nr;
-	priv->base.context_base = dma_fence_context_alloc(priv->base.contexts);
 	priv->base.uevent = true;
 
 	mutex_init(&priv->mutex);
@@ -207,7 +204,7 @@ nv84_fence_create(struct nouveau_drm *drm)
 			  * will lose CPU/GPU coherency!
 			  */
 			 TTM_PL_FLAG_TT | TTM_PL_FLAG_UNCACHED;
-	ret = nouveau_bo_new(&drm->client, 16 * priv->base.contexts, 0,
+	ret = nouveau_bo_new(&drm->client, 16 * drm->chan.nr, 0,
 			     domain, 0, 0, NULL, NULL, &priv->bo);
 	if (ret == 0) {
 		ret = nouveau_bo_pin(priv->bo, domain, false);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c
index 42a552d..600bdb8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c
@@ -46,6 +46,7 @@ nvkm_udevice_info_subdev(struct nvkm_device *device, u64 mthd, u64 *data)
 	enum nvkm_devidx subidx;
 
 	switch (mthd & NV_DEVICE_INFO_UNIT) {
+	case NV_DEVICE_FIFO(0): subidx = NVKM_ENGINE_FIFO; break;
 	default:
 		return -EINVAL;
 	}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c
index 64f6b76..49b37a8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c
@@ -30,6 +30,7 @@
 #include <subdev/mc.h>
 
 #include <nvif/event.h>
+#include <nvif/cl0080.h>
 #include <nvif/unpack.h>
 
 void
@@ -271,6 +272,18 @@ nvkm_fifo_fini(struct nvkm_engine *engine, bool suspend)
 }
 
 static int
+nvkm_fifo_info(struct nvkm_engine *engine, u64 mthd, u64 *data)
+{
+	struct nvkm_fifo *fifo = nvkm_fifo(engine);
+	switch (mthd) {
+	case NV_DEVICE_FIFO_CHANNELS: *data = fifo->nr; return 0;
+	default:
+		break;
+	}
+	return -ENOSYS;
+}
+
+static int
 nvkm_fifo_oneinit(struct nvkm_engine *engine)
 {
 	struct nvkm_fifo *fifo = nvkm_fifo(engine);
@@ -311,6 +324,7 @@ nvkm_fifo = {
 	.dtor = nvkm_fifo_dtor,
 	.preinit = nvkm_fifo_preinit,
 	.oneinit = nvkm_fifo_oneinit,
+	.info = nvkm_fifo_info,
 	.init = nvkm_fifo_init,
 	.fini = nvkm_fifo_fini,
 	.intr = nvkm_fifo_intr,
-- 
cgit v1.1


From 55b8e85b0b83150a20b17068e3229e64bd174d45 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/fifo/gk104-: accept engine contexts for CE3 and up

These can exist on GP100 and newer.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c
index 80c8752..6846199 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c
@@ -62,9 +62,8 @@ gk104_fifo_gpfifo_engine_addr(struct nvkm_engine *engine)
 {
 	switch (engine->subdev.index) {
 	case NVKM_ENGINE_SW    :
-	case NVKM_ENGINE_CE0   :
-	case NVKM_ENGINE_CE1   :
-	case NVKM_ENGINE_CE2   : return 0x0000;
+	case NVKM_ENGINE_CE0...NVKM_ENGINE_CE_LAST:
+		return 0;
 	case NVKM_ENGINE_GR    : return 0x0210;
 	case NVKM_ENGINE_SEC   : return 0x0220;
 	case NVKM_ENGINE_MSPDEC: return 0x0250;
-- 
cgit v1.1


From ddc669e25645b32158cc2b6da26faf5dac92fffc Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/fifo/gk104-: allow fault recovery code to be called by
 other subdevs

This will be required to support Volta.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h |   2 +
 .../gpu/drm/nouveau/include/nvkm/subdev/fault.h    |  13 ++
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c    |   6 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c   | 191 ++++++++++++---------
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h   |  21 +--
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk110.c   |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c   |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c   |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c   |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c   |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c   |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c   |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c   |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h    |   1 +
 14 files changed, 141 insertions(+), 101 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h b/drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h
index c17b3a9..0d96ede 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h
@@ -4,6 +4,7 @@
 #include <core/engine.h>
 #include <core/object.h>
 #include <core/event.h>
+struct nvkm_fault_data;
 
 #define NVKM_FIFO_CHID_NR 4096
 
@@ -45,6 +46,7 @@ struct nvkm_fifo {
 	struct nvkm_event kevent; /* channel killed */
 };
 
+void nvkm_fifo_fault(struct nvkm_fifo *, struct nvkm_fault_data *);
 void nvkm_fifo_pause(struct nvkm_fifo *, unsigned long *);
 void nvkm_fifo_start(struct nvkm_fifo *, unsigned long *);
 
diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/fault.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/fault.h
index 8ca66e5..8e9bc30 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/fault.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/fault.h
@@ -12,5 +12,18 @@ struct nvkm_fault {
 	struct nvkm_event event;
 };
 
+struct nvkm_fault_data {
+	u64  addr;
+	u64  inst;
+	u64  time;
+	u8 engine;
+	u8  valid;
+	u8    gpc;
+	u8    hub;
+	u8 access;
+	u8 client;
+	u8 reason;
+};
+
 int gp100_fault_new(struct nvkm_device *, int, struct nvkm_fault **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c
index 49b37a8..ed56087 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c
@@ -57,6 +57,12 @@ nvkm_fifo_start(struct nvkm_fifo *fifo, unsigned long *flags)
 }
 
 void
+nvkm_fifo_fault(struct nvkm_fifo *fifo, struct nvkm_fault_data *info)
+{
+	return fifo->func->fault(fifo, info);
+}
+
+void
 nvkm_fifo_chan_put(struct nvkm_fifo *fifo, unsigned long flags,
 		   struct nvkm_fifo_chan **pchan)
 {
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
index 84bd703..b5706b1 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
@@ -27,6 +27,7 @@
 #include <core/client.h>
 #include <core/gpuobj.h>
 #include <subdev/bar.h>
+#include <subdev/fault.h>
 #include <subdev/timer.h>
 #include <subdev/top.h>
 #include <engine/sw.h>
@@ -347,6 +348,90 @@ gk104_fifo_recover_engn(struct gk104_fifo *fifo, int engn)
 	schedule_work(&fifo->recover.work);
 }
 
+static void
+gk104_fifo_fault(struct nvkm_fifo *base, struct nvkm_fault_data *info)
+{
+	struct gk104_fifo *fifo = gk104_fifo(base);
+	struct nvkm_subdev *subdev = &fifo->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	const struct nvkm_enum *er, *ee, *ec, *ea;
+	struct nvkm_engine *engine = NULL;
+	struct nvkm_fifo_chan *chan;
+	unsigned long flags;
+	char ct[8] = "HUB/", en[16] = "";
+	int engn;
+
+	er = nvkm_enum_find(fifo->func->fault.reason, info->reason);
+	ee = nvkm_enum_find(fifo->func->fault.engine, info->engine);
+	if (info->hub) {
+		ec = nvkm_enum_find(fifo->func->fault.hubclient, info->client);
+	} else {
+		ec = nvkm_enum_find(fifo->func->fault.gpcclient, info->client);
+		snprintf(ct, sizeof(ct), "GPC%d/", info->gpc);
+	}
+	ea = nvkm_enum_find(fifo->func->fault.access, info->access);
+
+	if (ee && ee->data2) {
+		switch (ee->data2) {
+		case NVKM_SUBDEV_BAR:
+			nvkm_mask(device, 0x001704, 0x00000000, 0x00000000);
+			break;
+		case NVKM_SUBDEV_INSTMEM:
+			nvkm_mask(device, 0x001714, 0x00000000, 0x00000000);
+			break;
+		case NVKM_ENGINE_IFB:
+			nvkm_mask(device, 0x001718, 0x00000000, 0x00000000);
+			break;
+		default:
+			engine = nvkm_device_engine(device, ee->data2);
+			break;
+		}
+	}
+
+	if (ee == NULL) {
+		enum nvkm_devidx engidx = nvkm_top_fault(device, info->engine);
+		if (engidx < NVKM_SUBDEV_NR) {
+			const char *src = nvkm_subdev_name[engidx];
+			char *dst = en;
+			do {
+				*dst++ = toupper(*src++);
+			} while(*src);
+			engine = nvkm_device_engine(device, engidx);
+		}
+	} else {
+		snprintf(en, sizeof(en), "%s", ee->name);
+	}
+
+	spin_lock_irqsave(&fifo->base.lock, flags);
+	chan = nvkm_fifo_chan_inst_locked(&fifo->base, info->inst);
+
+	nvkm_error(subdev,
+		   "fault %02x [%s] at %016llx engine %02x [%s] client %02x "
+		   "[%s%s] reason %02x [%s] on channel %d [%010llx %s]\n",
+		   info->access, ea ? ea->name : "", info->addr,
+		   info->engine, ee ? ee->name : en,
+		   info->client, ct, ec ? ec->name : "",
+		   info->reason, er ? er->name : "", chan ? chan->chid : -1,
+		   info->inst, chan ? chan->object.client->name : "unknown");
+
+	/* Kill the channel that caused the fault. */
+	if (chan)
+		gk104_fifo_recover_chan(&fifo->base, chan->chid);
+
+	/* Channel recovery will probably have already done this for the
+	 * correct engine(s), but just in case we can't find the channel
+	 * information...
+	 */
+	for (engn = 0; engn < fifo->engine_nr && engine; engn++) {
+		if (fifo->engine[engn].engine == engine) {
+			gk104_fifo_recover_engn(fifo, engn);
+			break;
+		}
+	}
+
+	spin_unlock_irqrestore(&fifo->base.lock, flags);
+}
+
 static const struct nvkm_enum
 gk104_fifo_bind_reason[] = {
 	{ 0x01, "BIND_NOT_UNBOUND" },
@@ -456,88 +541,21 @@ gk104_fifo_intr_fault(struct gk104_fifo *fifo, int unit)
 	u32 inst = nvkm_rd32(device, 0x002800 + (unit * 0x10));
 	u32 valo = nvkm_rd32(device, 0x002804 + (unit * 0x10));
 	u32 vahi = nvkm_rd32(device, 0x002808 + (unit * 0x10));
-	u32 stat = nvkm_rd32(device, 0x00280c + (unit * 0x10));
-	u32 gpc    = (stat & 0x1f000000) >> 24;
-	u32 client = (stat & 0x00001f00) >> 8;
-	u32 write  = (stat & 0x00000080);
-	u32 hub    = (stat & 0x00000040);
-	u32 reason = (stat & 0x0000000f);
-	const struct nvkm_enum *er, *eu, *ec;
-	struct nvkm_engine *engine = NULL;
-	struct nvkm_fifo_chan *chan;
-	unsigned long flags;
-	char gpcid[8] = "", en[16] = "";
-	int engn;
-
-	er = nvkm_enum_find(fifo->func->fault.reason, reason);
-	eu = nvkm_enum_find(fifo->func->fault.engine, unit);
-	if (hub) {
-		ec = nvkm_enum_find(fifo->func->fault.hubclient, client);
-	} else {
-		ec = nvkm_enum_find(fifo->func->fault.gpcclient, client);
-		snprintf(gpcid, sizeof(gpcid), "GPC%d/", gpc);
-	}
-
-	if (eu && eu->data2) {
-		switch (eu->data2) {
-		case NVKM_SUBDEV_BAR:
-			nvkm_mask(device, 0x001704, 0x00000000, 0x00000000);
-			break;
-		case NVKM_SUBDEV_INSTMEM:
-			nvkm_mask(device, 0x001714, 0x00000000, 0x00000000);
-			break;
-		case NVKM_ENGINE_IFB:
-			nvkm_mask(device, 0x001718, 0x00000000, 0x00000000);
-			break;
-		default:
-			engine = nvkm_device_engine(device, eu->data2);
-			break;
-		}
-	}
-
-	if (eu == NULL) {
-		enum nvkm_devidx engidx = nvkm_top_fault(device, unit);
-		if (engidx < NVKM_SUBDEV_NR) {
-			const char *src = nvkm_subdev_name[engidx];
-			char *dst = en;
-			do {
-				*dst++ = toupper(*src++);
-			} while(*src);
-			engine = nvkm_device_engine(device, engidx);
-		}
-	} else {
-		snprintf(en, sizeof(en), "%s", eu->name);
-	}
-
-	spin_lock_irqsave(&fifo->base.lock, flags);
-	chan = nvkm_fifo_chan_inst_locked(&fifo->base, (u64)inst << 12);
-
-	nvkm_error(subdev,
-		   "%s fault at %010llx engine %02x [%s] client %02x [%s%s] "
-		   "reason %02x [%s] on channel %d [%010llx %s]\n",
-		   write ? "write" : "read", (u64)vahi << 32 | valo,
-		   unit, en, client, gpcid, ec ? ec->name : "",
-		   reason, er ? er->name : "", chan ? chan->chid : -1,
-		   (u64)inst << 12,
-		   chan ? chan->object.client->name : "unknown");
-
-
-	/* Kill the channel that caused the fault. */
-	if (chan)
-		gk104_fifo_recover_chan(&fifo->base, chan->chid);
-
-	/* Channel recovery will probably have already done this for the
-	 * correct engine(s), but just in case we can't find the channel
-	 * information...
-	 */
-	for (engn = 0; engn < fifo->engine_nr && engine; engn++) {
-		if (fifo->engine[engn].engine == engine) {
-			gk104_fifo_recover_engn(fifo, engn);
-			break;
-		}
-	}
-
-	spin_unlock_irqrestore(&fifo->base.lock, flags);
+	u32 type = nvkm_rd32(device, 0x00280c + (unit * 0x10));
+	struct nvkm_fault_data info;
+
+	info.inst   =  (u64)inst << 12;
+	info.addr   = ((u64)vahi << 32) | valo;
+	info.time   = 0;
+	info.engine = unit;
+	info.valid  = 1;
+	info.gpc    = (type & 0x1f000000) >> 24;
+	info.client = (type & 0x00001f00) >> 8;
+	info.access = (type & 0x00000080) >> 7;
+	info.hub    = (type & 0x00000040) >> 6;
+	info.reason = (type & 0x000000ff);
+
+	nvkm_fifo_fault(&fifo->base, &info);
 }
 
 static const struct nvkm_bitfield gk104_fifo_pbdma_intr_0[] = {
@@ -897,6 +915,7 @@ gk104_fifo_ = {
 	.init = gk104_fifo_init,
 	.fini = gk104_fifo_fini,
 	.intr = gk104_fifo_intr,
+	.fault = gk104_fifo_fault,
 	.uevent_init = gk104_fifo_uevent_init,
 	.uevent_fini = gk104_fifo_uevent_fini,
 	.recover_chan = gk104_fifo_recover_chan,
@@ -919,6 +938,13 @@ gk104_fifo_new_(const struct gk104_fifo_func *func, struct nvkm_device *device,
 }
 
 const struct nvkm_enum
+gk104_fifo_fault_access[] = {
+	{ 0x0, "READ" },
+	{ 0x1, "WRITE" },
+	{}
+};
+
+const struct nvkm_enum
 gk104_fifo_fault_engine[] = {
 	{ 0x00, "GR", NULL, NVKM_ENGINE_GR },
 	{ 0x01, "DISPLAY" },
@@ -1035,6 +1061,7 @@ gk104_fifo_fault_gpcclient[] = {
 
 static const struct gk104_fifo_func
 gk104_fifo = {
+	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gk104_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
index 1579785..0e8b572 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
@@ -44,6 +44,7 @@ struct gk104_fifo {
 
 struct gk104_fifo_func {
 	struct {
+		const struct nvkm_enum *access;
 		const struct nvkm_enum *engine;
 		const struct nvkm_enum *reason;
 		const struct nvkm_enum *hubclient;
@@ -59,25 +60,7 @@ void gk104_fifo_runlist_insert(struct gk104_fifo *, struct gk104_fifo_chan *);
 void gk104_fifo_runlist_remove(struct gk104_fifo *, struct gk104_fifo_chan *);
 void gk104_fifo_runlist_commit(struct gk104_fifo *, int runl);
 
-static inline u64
-gk104_fifo_engine_subdev(int engine)
-{
-	switch (engine) {
-	case 0: return (1ULL << NVKM_ENGINE_GR) |
-		       (1ULL << NVKM_ENGINE_SW) |
-		       (1ULL << NVKM_ENGINE_CE2);
-	case 1: return (1ULL << NVKM_ENGINE_MSPDEC);
-	case 2: return (1ULL << NVKM_ENGINE_MSPPP);
-	case 3: return (1ULL << NVKM_ENGINE_MSVLD);
-	case 4: return (1ULL << NVKM_ENGINE_CE0);
-	case 5: return (1ULL << NVKM_ENGINE_CE1);
-	case 6: return (1ULL << NVKM_ENGINE_MSENC);
-	default:
-		WARN_ON(1);
-		return 0;
-	}
-}
-
+extern const struct nvkm_enum gk104_fifo_fault_access[];
 extern const struct nvkm_enum gk104_fifo_fault_engine[];
 extern const struct nvkm_enum gk104_fifo_fault_reason[];
 extern const struct nvkm_enum gk104_fifo_fault_hubclient[];
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk110.c
index b2f8ab7..ad792b6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk110.c
@@ -26,6 +26,7 @@
 
 static const struct gk104_fifo_func
 gk110_fifo = {
+	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gk104_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c
index 160617d..5402d22 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c
@@ -26,6 +26,7 @@
 
 static const struct gk104_fifo_func
 gk208_fifo = {
+	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gk104_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c
index be9f5c1..0d7f9f5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c
@@ -24,6 +24,7 @@
 
 static const struct gk104_fifo_func
 gk20a_fifo = {
+	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gk104_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c
index 29c0806..58a46ee 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c
@@ -49,6 +49,7 @@ gm107_fifo_fault_engine[] = {
 
 static const struct gk104_fifo_func
 gm107_fifo = {
+	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gm107_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c
index b069f78..f84d539 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c
@@ -26,6 +26,7 @@
 
 static const struct gk104_fifo_func
 gm200_fifo = {
+	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gm107_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c
index 2ed87c2..66399b9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c
@@ -24,6 +24,7 @@
 
 static const struct gk104_fifo_func
 gm20b_fifo = {
+	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gm107_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c
index 41f16cf..2b8a6cf 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c
@@ -50,6 +50,7 @@ gp100_fifo_fault_engine[] = {
 
 static const struct gk104_fifo_func
 gp100_fifo = {
+	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gp100_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c
index 4af96c3..71b8d93 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c
@@ -24,6 +24,7 @@
 
 static const struct gk104_fifo_func
 gp10b_fifo = {
+	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gp100_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h
index ae76b1a..df74b54 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h
@@ -21,6 +21,7 @@ struct nvkm_fifo_func {
 	void (*init)(struct nvkm_fifo *);
 	void (*fini)(struct nvkm_fifo *);
 	void (*intr)(struct nvkm_fifo *);
+	void (*fault)(struct nvkm_fifo *, struct nvkm_fault_data *);
 	void (*pause)(struct nvkm_fifo *, unsigned long *);
 	void (*start)(struct nvkm_fifo *, unsigned long *);
 	void (*uevent_init)(struct nvkm_fifo *);
-- 
cgit v1.1


From cc36205085bb6e3a4eed1edbe413fd2235cadb27 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/fifo/gk104-: support querying engines available on each
 runlist

Will be used to improve channel runlist selection.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/include/nvif/cl0080.h    |  7 ++
 drivers/gpu/drm/nouveau/include/nvif/device.h    |  5 ++
 drivers/gpu/drm/nouveau/include/nvif/fifo.h      | 18 +++++
 drivers/gpu/drm/nouveau/nvif/Kbuild              |  1 +
 drivers/gpu/drm/nouveau/nvif/device.c            |  3 +
 drivers/gpu/drm/nouveau/nvif/fifo.c              | 99 ++++++++++++++++++++++++
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c  |  2 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c | 30 +++++++
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h  |  1 +
 9 files changed, 166 insertions(+)
 create mode 100644 drivers/gpu/drm/nouveau/include/nvif/fifo.h
 create mode 100644 drivers/gpu/drm/nouveau/nvif/fifo.c

diff --git a/drivers/gpu/drm/nouveau/include/nvif/cl0080.h b/drivers/gpu/drm/nouveau/include/nvif/cl0080.h
index 5af610e..49c1c90 100644
--- a/drivers/gpu/drm/nouveau/include/nvif/cl0080.h
+++ b/drivers/gpu/drm/nouveau/include/nvif/cl0080.h
@@ -83,4 +83,11 @@ struct nv_device_time_v0 {
 
 /* Returns the number of available channels. */
 #define NV_DEVICE_FIFO_CHANNELS                      NV_DEVICE_FIFO(0x00000000)
+
+/* Returns a mask of available runlists. */
+#define NV_DEVICE_FIFO_RUNLISTS                      NV_DEVICE_FIFO(0x00000001)
+
+/* These return a mask of engines available on a particular runlist. */
+#define NV_DEVICE_FIFO_RUNLIST_ENGINES(n)     ((n) + NV_DEVICE_FIFO(0x00000010))
+#define NV_DEVICE_FIFO_RUNLIST_ENGINES__SIZE                                64
 #endif
diff --git a/drivers/gpu/drm/nouveau/include/nvif/device.h b/drivers/gpu/drm/nouveau/include/nvif/device.h
index 216dbd9..76fe21e 100644
--- a/drivers/gpu/drm/nouveau/include/nvif/device.h
+++ b/drivers/gpu/drm/nouveau/include/nvif/device.h
@@ -8,6 +8,11 @@
 struct nvif_device {
 	struct nvif_object object;
 	struct nv_device_info_v0 info;
+
+	struct nvif_fifo_runlist {
+		u64 engines;
+	} *runlist;
+	int runlists;
 };
 
 int  nvif_device_init(struct nvif_object *, u32 handle, s32 oclass, void *, u32,
diff --git a/drivers/gpu/drm/nouveau/include/nvif/fifo.h b/drivers/gpu/drm/nouveau/include/nvif/fifo.h
new file mode 100644
index 0000000..e9468c9
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/include/nvif/fifo.h
@@ -0,0 +1,18 @@
+#ifndef __NVIF_FIFO_H__
+#define __NVIF_FIFO_H__
+#include <nvif/device.h>
+
+/* Returns mask of runlists that support a NV_DEVICE_INFO_ENGINE_* type. */
+u64 nvif_fifo_runlist(struct nvif_device *, u64 engine);
+
+/* CE-supporting runlists (excluding GRCE, if others exist). */
+static inline u64
+nvif_fifo_runlist_ce(struct nvif_device *device)
+{
+	u64 runmgr = nvif_fifo_runlist(device, NV_DEVICE_INFO_ENGINE_GR);
+	u64 runmce = nvif_fifo_runlist(device, NV_DEVICE_INFO_ENGINE_CE);
+	if (runmce && !(runmce &= ~runmgr))
+		runmce = runmgr;
+	return runmce;
+}
+#endif
diff --git a/drivers/gpu/drm/nouveau/nvif/Kbuild b/drivers/gpu/drm/nouveau/nvif/Kbuild
index f1675a4..c817b02 100644
--- a/drivers/gpu/drm/nouveau/nvif/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvif/Kbuild
@@ -2,6 +2,7 @@ nvif-y := nvif/object.o
 nvif-y += nvif/client.o
 nvif-y += nvif/device.o
 nvif-y += nvif/driver.o
+nvif-y += nvif/fifo.o
 nvif-y += nvif/mem.o
 nvif-y += nvif/mmu.o
 nvif-y += nvif/notify.o
diff --git a/drivers/gpu/drm/nouveau/nvif/device.c b/drivers/gpu/drm/nouveau/nvif/device.c
index 252d8c3..ca5eb3d 100644
--- a/drivers/gpu/drm/nouveau/nvif/device.c
+++ b/drivers/gpu/drm/nouveau/nvif/device.c
@@ -37,6 +37,8 @@ nvif_device_time(struct nvif_device *device)
 void
 nvif_device_fini(struct nvif_device *device)
 {
+	kfree(device->runlist);
+	device->runlist = NULL;
 	nvif_object_fini(&device->object);
 }
 
@@ -46,6 +48,7 @@ nvif_device_init(struct nvif_object *parent, u32 handle, s32 oclass,
 {
 	int ret = nvif_object_init(parent, handle, oclass, data, size,
 				   &device->object);
+	device->runlist = NULL;
 	if (ret == 0) {
 		device->info.version = 0;
 		ret = nvif_object_mthd(&device->object, NV_DEVICE_V0_INFO,
diff --git a/drivers/gpu/drm/nouveau/nvif/fifo.c b/drivers/gpu/drm/nouveau/nvif/fifo.c
new file mode 100644
index 0000000..99d4fd1
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvif/fifo.c
@@ -0,0 +1,99 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include <nvif/fifo.h>
+
+static int
+nvif_fifo_runlists(struct nvif_device *device)
+{
+	struct nvif_object *object = &device->object;
+	struct {
+		struct nv_device_info_v1 m;
+		struct {
+			struct nv_device_info_v1_data runlists;
+			struct nv_device_info_v1_data runlist[64];
+		} v;
+	} *a;
+	int ret, i;
+
+	if (device->runlist)
+		return 0;
+
+	if (!(a = kmalloc(sizeof(*a), GFP_KERNEL)))
+		return -ENOMEM;
+	a->m.version = 1;
+	a->m.count = sizeof(a->v) / sizeof(a->v.runlists);
+	a->v.runlists.mthd = NV_DEVICE_FIFO_RUNLISTS;
+	for (i = 0; i < ARRAY_SIZE(a->v.runlist); i++)
+		a->v.runlist[i].mthd = NV_DEVICE_FIFO_RUNLIST_ENGINES(i);
+
+	ret = nvif_object_mthd(object, NV_DEVICE_V0_INFO, a, sizeof(*a));
+	if (ret)
+		goto done;
+
+	device->runlists = fls64(a->v.runlists.data);
+	device->runlist = kzalloc(sizeof(*device->runlist) *
+				  device->runlists, GFP_KERNEL);
+	if (!device->runlist) {
+		ret = -ENOMEM;
+		goto done;
+	}
+
+	for (i = 0; i < device->runlists; i++) {
+		if (a->v.runlists.data & BIT_ULL(i))
+			device->runlist[i].engines = a->v.runlist[i].data;
+	}
+
+done:
+	kfree(a);
+	return ret;
+}
+
+u64
+nvif_fifo_runlist(struct nvif_device *device, u64 engine)
+{
+	struct nvif_object *object = &device->object;
+	struct {
+		struct nv_device_info_v1 m;
+		struct {
+			struct nv_device_info_v1_data engine;
+		} v;
+	} a = {
+		.m.version = 1,
+		.m.count = sizeof(a.v) / sizeof(a.v.engine),
+		.v.engine.mthd = engine,
+	};
+	u64 runm = 0;
+	int ret, i;
+
+	if ((ret = nvif_fifo_runlists(device)))
+		return runm;
+
+	ret = nvif_object_mthd(object, NV_DEVICE_V0_INFO, &a, sizeof(a));
+	if (ret == 0) {
+		for (i = 0; i < device->runlists; i++) {
+			if (device->runlist[i].engines & a.v.engine.data)
+				runm |= BIT_ULL(i);
+		}
+	}
+
+	return runm;
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c
index ed56087..1642d8e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c
@@ -284,6 +284,8 @@ nvkm_fifo_info(struct nvkm_engine *engine, u64 mthd, u64 *data)
 	switch (mthd) {
 	case NV_DEVICE_FIFO_CHANNELS: *data = fifo->nr; return 0;
 	default:
+		if (fifo->func->info)
+			return fifo->func->info(fifo, mthd, data);
 		break;
 	}
 	return -ENOSYS;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
index b5706b1..b6c23cf 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
@@ -33,6 +33,7 @@
 #include <engine/sw.h>
 
 #include <nvif/class.h>
+#include <nvif/cl0080.h>
 
 struct gk104_fifo_engine_status {
 	bool busy;
@@ -784,6 +785,34 @@ gk104_fifo_fini(struct nvkm_fifo *base)
 }
 
 static int
+gk104_fifo_info(struct nvkm_fifo *base, u64 mthd, u64 *data)
+{
+	struct gk104_fifo *fifo = gk104_fifo(base);
+	switch (mthd) {
+	case NV_DEVICE_FIFO_RUNLISTS:
+		*data = (1ULL << fifo->runlist_nr) - 1;
+		return 0;
+	case NV_DEVICE_FIFO_RUNLIST_ENGINES(0)...
+	     NV_DEVICE_FIFO_RUNLIST_ENGINES(63): {
+		int runl = mthd - NV_DEVICE_FIFO_RUNLIST_ENGINES(0), engn;
+		if (runl < fifo->runlist_nr) {
+			unsigned long engm = fifo->runlist[runl].engm;
+			struct nvkm_engine *engine;
+			*data = 0;
+			for_each_set_bit(engn, &engm, fifo->engine_nr) {
+				if ((engine = fifo->engine[engn].engine))
+					*data |= BIT_ULL(engine->subdev.index);
+			}
+			return 0;
+		}
+	}
+		return -EINVAL;
+	default:
+		return -EINVAL;
+	}
+}
+
+static int
 gk104_fifo_oneinit(struct nvkm_fifo *base)
 {
 	struct gk104_fifo *fifo = gk104_fifo(base);
@@ -912,6 +941,7 @@ static const struct nvkm_fifo_func
 gk104_fifo_ = {
 	.dtor = gk104_fifo_dtor,
 	.oneinit = gk104_fifo_oneinit,
+	.info = gk104_fifo_info,
 	.init = gk104_fifo_init,
 	.fini = gk104_fifo_fini,
 	.intr = gk104_fifo_intr,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h
index df74b54..f9b4e9d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h
@@ -18,6 +18,7 @@ struct nvkm_fifo_chan_oclass;
 struct nvkm_fifo_func {
 	void *(*dtor)(struct nvkm_fifo *);
 	int (*oneinit)(struct nvkm_fifo *);
+	int (*info)(struct nvkm_fifo *, u64 mthd, u64 *data);
 	void (*init)(struct nvkm_fifo *);
 	void (*fini)(struct nvkm_fifo *);
 	void (*intr)(struct nvkm_fifo *);
-- 
cgit v1.1


From a7cf01809bf23b95413d8047bd91cdc3cedd1ca1 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/fifo/gk104-: require explicit runlist selection for
 channel allocation

We didn't used to be aware that runlist/engine IDs weren't the same thing,
or that there was such variability in configuration between GPUs.

By exposing this information to a client, and giving it explicit control
of which runlist it's allocating a channel on, we're able to make better
choices.

The immediate effect of this is that on GPUs where CE0 is the "GRCE", we
will now be allocating a copy engine running asynchronously to GR for BO
migrations - as intended.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/include/nvif/cla06f.h      | 18 +----
 drivers/gpu/drm/nouveau/nouveau_abi16.c            | 35 ++++-----
 drivers/gpu/drm/nouveau/nouveau_chan.c             |  4 +-
 drivers/gpu/drm/nouveau/nouveau_drm.c              |  8 +--
 .../gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c | 83 +++++-----------------
 5 files changed, 43 insertions(+), 105 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/include/nvif/cla06f.h b/drivers/gpu/drm/nouveau/include/nvif/cla06f.h
index 56f5bd8..fbfcffc 100644
--- a/drivers/gpu/drm/nouveau/include/nvif/cla06f.h
+++ b/drivers/gpu/drm/nouveau/include/nvif/cla06f.h
@@ -4,25 +4,11 @@
 
 struct kepler_channel_gpfifo_a_v0 {
 	__u8  version;
-	__u8  pad01[5];
+	__u8  pad01[1];
 	__u16 chid;
-#define NVA06F_V0_ENGINE_SW                                          0x00000001
-#define NVA06F_V0_ENGINE_GR                                          0x00000002
-#define NVA06F_V0_ENGINE_SEC                                         0x00000004
-#define NVA06F_V0_ENGINE_MSVLD                                       0x00000010
-#define NVA06F_V0_ENGINE_MSPDEC                                      0x00000020
-#define NVA06F_V0_ENGINE_MSPPP                                       0x00000040
-#define NVA06F_V0_ENGINE_MSENC                                       0x00000080
-#define NVA06F_V0_ENGINE_VIC                                         0x00000100
-#define NVA06F_V0_ENGINE_NVDEC                                       0x00000200
-#define NVA06F_V0_ENGINE_NVENC0                                      0x00000400
-#define NVA06F_V0_ENGINE_NVENC1                                      0x00000800
-#define NVA06F_V0_ENGINE_CE0                                         0x00010000
-#define NVA06F_V0_ENGINE_CE1                                         0x00020000
-#define NVA06F_V0_ENGINE_CE2                                         0x00040000
-	__u32 engines;
 	__u32 ilength;
 	__u64 ioffset;
+	__u64 runlist;
 	__u64 vmm;
 };
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_abi16.c b/drivers/gpu/drm/nouveau/nouveau_abi16.c
index ece650a..ea24727 100644
--- a/drivers/gpu/drm/nouveau/nouveau_abi16.c
+++ b/drivers/gpu/drm/nouveau/nouveau_abi16.c
@@ -23,6 +23,7 @@
 
 #include <nvif/client.h>
 #include <nvif/driver.h>
+#include <nvif/fifo.h>
 #include <nvif/ioctl.h>
 #include <nvif/class.h>
 #include <nvif/cl0002.h>
@@ -256,6 +257,7 @@ nouveau_abi16_ioctl_channel_alloc(ABI16_IOCTL_ARGS)
 	struct nouveau_abi16 *abi16 = nouveau_abi16_get(file_priv);
 	struct nouveau_abi16_chan *chan;
 	struct nvif_device *device;
+	u64 engine;
 	int ret;
 
 	if (unlikely(!abi16))
@@ -268,25 +270,26 @@ nouveau_abi16_ioctl_channel_alloc(ABI16_IOCTL_ARGS)
 
 	/* hack to allow channel engine type specification on kepler */
 	if (device->info.family >= NV_DEVICE_INFO_V0_KEPLER) {
-		if (init->fb_ctxdma_handle != ~0)
-			init->fb_ctxdma_handle = NVA06F_V0_ENGINE_GR;
-		else {
-			init->fb_ctxdma_handle = 0;
-#define _(A,B) if (init->tt_ctxdma_handle & (A)) init->fb_ctxdma_handle |= (B)
-			_(0x01, NVA06F_V0_ENGINE_GR);
-			_(0x02, NVA06F_V0_ENGINE_MSPDEC);
-			_(0x04, NVA06F_V0_ENGINE_MSPPP);
-			_(0x08, NVA06F_V0_ENGINE_MSVLD);
-			_(0x10, NVA06F_V0_ENGINE_CE0);
-			_(0x20, NVA06F_V0_ENGINE_CE1);
-			_(0x40, NVA06F_V0_ENGINE_MSENC);
-#undef _
+		if (init->fb_ctxdma_handle == ~0) {
+			switch (init->tt_ctxdma_handle) {
+			case 0x01: engine = NV_DEVICE_INFO_ENGINE_GR    ; break;
+			case 0x02: engine = NV_DEVICE_INFO_ENGINE_MSPDEC; break;
+			case 0x04: engine = NV_DEVICE_INFO_ENGINE_MSPPP ; break;
+			case 0x08: engine = NV_DEVICE_INFO_ENGINE_MSVLD ; break;
+			case 0x30: engine = NV_DEVICE_INFO_ENGINE_CE    ; break;
+			default:
+				return nouveau_abi16_put(abi16, -ENOSYS);
+			}
+		} else {
+			engine = NV_DEVICE_INFO_ENGINE_GR;
 		}
 
-		/* allow flips to be executed if this is a graphics channel */
+		if (engine != NV_DEVICE_INFO_ENGINE_CE)
+			engine = nvif_fifo_runlist(device, engine);
+		else
+			engine = nvif_fifo_runlist_ce(device);
+		init->fb_ctxdma_handle = engine;
 		init->tt_ctxdma_handle = 0;
-		if (init->fb_ctxdma_handle == NVA06F_V0_ENGINE_GR)
-			init->tt_ctxdma_handle = 1;
 	}
 
 	if (init->fb_ctxdma_handle == ~0 || init->tt_ctxdma_handle == ~0)
diff --git a/drivers/gpu/drm/nouveau/nouveau_chan.c b/drivers/gpu/drm/nouveau/nouveau_chan.c
index db69d13..67950a5c5 100644
--- a/drivers/gpu/drm/nouveau/nouveau_chan.c
+++ b/drivers/gpu/drm/nouveau/nouveau_chan.c
@@ -214,7 +214,7 @@ nouveau_channel_prep(struct nouveau_drm *drm, struct nvif_device *device,
 
 static int
 nouveau_channel_ind(struct nouveau_drm *drm, struct nvif_device *device,
-		    u32 engine, struct nouveau_channel **pchan)
+		    u64 runlist, struct nouveau_channel **pchan)
 {
 	struct nouveau_cli *cli = (void *)device->object.client;
 	static const u16 oclasses[] = { PASCAL_CHANNEL_GPFIFO_A,
@@ -245,9 +245,9 @@ nouveau_channel_ind(struct nouveau_drm *drm, struct nvif_device *device,
 	do {
 		if (oclass[0] >= KEPLER_CHANNEL_GPFIFO_A) {
 			args.kepler.version = 0;
-			args.kepler.engines = engine;
 			args.kepler.ilength = 0x02000;
 			args.kepler.ioffset = 0x10000 + chan->push.addr;
+			args.kepler.runlist = runlist;
 			args.kepler.vmm = nvif_handle(&cli->vmm.vmm.object);
 			size = sizeof(args.kepler);
 		} else
diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c
index dddd425..6caece4 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_drm.c
@@ -38,6 +38,7 @@
 #include <core/tegra.h>
 
 #include <nvif/driver.h>
+#include <nvif/fifo.h>
 
 #include <nvif/class.h>
 #include <nvif/cl0002.h>
@@ -358,13 +359,12 @@ nouveau_accel_init(struct nouveau_drm *drm)
 
 	if (device->info.family >= NV_DEVICE_INFO_V0_KEPLER) {
 		ret = nouveau_channel_new(drm, &drm->client.device,
-					  NVA06F_V0_ENGINE_CE0 |
-					  NVA06F_V0_ENGINE_CE1,
-					  0, &drm->cechan);
+					  nvif_fifo_runlist_ce(device), 0,
+					  &drm->cechan);
 		if (ret)
 			NV_ERROR(drm, "failed to create ce channel, %d\n", ret);
 
-		arg0 = NVA06F_V0_ENGINE_GR;
+		arg0 = nvif_fifo_runlist(device, NV_DEVICE_INFO_ENGINE_GR);
 		arg1 = 1;
 	} else
 	if (device->info.chipset >= 0xa3 &&
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c
index 6846199..e331ab1 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c
@@ -222,62 +222,30 @@ gk104_fifo_gpfifo_func = {
 	.engine_fini = gk104_fifo_gpfifo_engine_fini,
 };
 
-struct gk104_fifo_chan_func {
-	u32 engine;
-	u64 subdev;
-};
-
 static int
-gk104_fifo_gpfifo_new_(const struct gk104_fifo_chan_func *func,
-		       struct gk104_fifo *fifo, u32 *engmask, u16 *chid,
+gk104_fifo_gpfifo_new_(struct gk104_fifo *fifo, u64 *runlists, u16 *chid,
 		       u64 vmm, u64 ioffset, u64 ilength,
 		       const struct nvkm_oclass *oclass,
 		       struct nvkm_object **pobject)
 {
 	struct gk104_fifo_chan *chan;
-	int runlist = -1, ret = -ENOSYS, i, j;
-	u32 engines = 0, present = 0;
+	int runlist = ffs(*runlists) -1, ret, i;
+	unsigned long engm;
 	u64 subdevs = 0;
 	u64 usermem;
 
-	if (!vmm)
+	if (!vmm || runlist < 0 || runlist >= fifo->runlist_nr)
 		return -EINVAL;
+	*runlists = BIT_ULL(runlist);
 
-	/* Determine which downstream engines are present */
-	for (i = 0; i < fifo->engine_nr; i++) {
-		struct nvkm_engine *engine = fifo->engine[i].engine;
-		if (engine) {
-			u64 submask = BIT_ULL(engine->subdev.index);
-			for (j = 0; func[j].subdev; j++) {
-				if (func[j].subdev & submask) {
-					present |= func[j].engine;
-					break;
-				}
-			}
-
-			if (!func[j].subdev)
-				continue;
-
-			if (runlist < 0 && (*engmask & present))
-				runlist = fifo->engine[i].runl;
-			if (runlist == fifo->engine[i].runl) {
-				engines |= func[j].engine;
-				subdevs |= func[j].subdev;
-			}
-		}
+	engm = fifo->runlist[runlist].engm;
+	for_each_set_bit(i, &engm, fifo->engine_nr) {
+		if (fifo->engine[i].engine)
+			subdevs |= BIT_ULL(fifo->engine[i].engine->subdev.index);
 	}
 
-	/* Just an engine mask query?  All done here! */
-	if (!*engmask) {
-		*engmask = present;
-		return nvkm_object_new(oclass, NULL, 0, pobject);
-	}
-
-	/* No runlist?  No supported engines. */
-	*engmask = present;
-	if (runlist < 0)
-		return -ENODEV;
-	*engmask = engines;
+	if (subdevs & BIT_ULL(NVKM_ENGINE_GR))
+		subdevs |= BIT_ULL(NVKM_ENGINE_SW);
 
 	/* Allocate the channel. */
 	if (!(chan = kzalloc(sizeof(*chan), GFP_KERNEL)))
@@ -327,26 +295,6 @@ gk104_fifo_gpfifo_new_(const struct gk104_fifo_chan_func *func,
 	return 0;
 }
 
-static const struct gk104_fifo_chan_func
-gk104_fifo_gpfifo[] = {
-	{ NVA06F_V0_ENGINE_SW | NVA06F_V0_ENGINE_GR,
-		BIT_ULL(NVKM_ENGINE_SW) | BIT_ULL(NVKM_ENGINE_GR)
-	},
-	{ NVA06F_V0_ENGINE_SEC   , BIT_ULL(NVKM_ENGINE_SEC   ) },
-	{ NVA06F_V0_ENGINE_MSVLD , BIT_ULL(NVKM_ENGINE_MSVLD ) },
-	{ NVA06F_V0_ENGINE_MSPDEC, BIT_ULL(NVKM_ENGINE_MSPDEC) },
-	{ NVA06F_V0_ENGINE_MSPPP , BIT_ULL(NVKM_ENGINE_MSPPP ) },
-	{ NVA06F_V0_ENGINE_MSENC , BIT_ULL(NVKM_ENGINE_MSENC ) },
-	{ NVA06F_V0_ENGINE_VIC   , BIT_ULL(NVKM_ENGINE_VIC   ) },
-	{ NVA06F_V0_ENGINE_NVDEC , BIT_ULL(NVKM_ENGINE_NVDEC ) },
-	{ NVA06F_V0_ENGINE_NVENC0, BIT_ULL(NVKM_ENGINE_NVENC0) },
-	{ NVA06F_V0_ENGINE_NVENC1, BIT_ULL(NVKM_ENGINE_NVENC1) },
-	{ NVA06F_V0_ENGINE_CE0   , BIT_ULL(NVKM_ENGINE_CE0   ) },
-	{ NVA06F_V0_ENGINE_CE1   , BIT_ULL(NVKM_ENGINE_CE1   ) },
-	{ NVA06F_V0_ENGINE_CE2   , BIT_ULL(NVKM_ENGINE_CE2   ) },
-	{}
-};
-
 int
 gk104_fifo_gpfifo_new(struct nvkm_fifo *base, const struct nvkm_oclass *oclass,
 		      void *data, u32 size, struct nvkm_object **pobject)
@@ -361,11 +309,12 @@ gk104_fifo_gpfifo_new(struct nvkm_fifo *base, const struct nvkm_oclass *oclass,
 	nvif_ioctl(parent, "create channel gpfifo size %d\n", size);
 	if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false))) {
 		nvif_ioctl(parent, "create channel gpfifo vers %d vmm %llx "
-				   "ioffset %016llx ilength %08x engine %08x\n",
+				   "ioffset %016llx ilength %08x "
+				   "runlist %016llx\n",
 			   args->v0.version, args->v0.vmm, args->v0.ioffset,
-			   args->v0.ilength, args->v0.engines);
-		return gk104_fifo_gpfifo_new_(gk104_fifo_gpfifo, fifo,
-					      &args->v0.engines,
+			   args->v0.ilength, args->v0.runlist);
+		return gk104_fifo_gpfifo_new_(fifo,
+					      &args->v0.runlist,
 					      &args->v0.chid,
 					       args->v0.vmm,
 					       args->v0.ioffset,
-- 
cgit v1.1


From f9360c3aa61f792de3c839c63bfadf8640255d8c Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/fifo/gk104-: simplify definition of channel classes

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/Kbuild    |  3 --
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c    | 24 ++++++++++-----
 .../gpu/drm/nouveau/nvkm/engine/fifo/changk104.h   |  7 +----
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c   | 28 +++++++++++++-----
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h   |  6 +++-
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk110.c   |  7 ++---
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c   |  7 ++---
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c   |  7 ++---
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c   |  7 ++---
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c   |  7 ++---
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c   |  7 ++---
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c   |  7 ++---
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c   |  7 ++---
 .../gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c | 11 +------
 .../gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk110.c | 34 ----------------------
 .../gpu/drm/nouveau/nvkm/engine/fifo/gpfifogm200.c | 34 ----------------------
 .../gpu/drm/nouveau/nvkm/engine/fifo/gpfifogp100.c | 34 ----------------------
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h    |  5 ++--
 18 files changed, 71 insertions(+), 171 deletions(-)
 delete mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk110.c
 delete mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogm200.c
 delete mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogp100.c

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/Kbuild b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/Kbuild
index 64e5183..b888ea6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/Kbuild
@@ -31,6 +31,3 @@ nvkm-y += nvkm/engine/fifo/gpfifonv50.o
 nvkm-y += nvkm/engine/fifo/gpfifog84.o
 nvkm-y += nvkm/engine/fifo/gpfifogf100.o
 nvkm-y += nvkm/engine/fifo/gpfifogk104.o
-nvkm-y += nvkm/engine/fifo/gpfifogk110.o
-nvkm-y += nvkm/engine/fifo/gpfifogm200.o
-nvkm-y += nvkm/engine/fifo/gpfifogp100.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c
index 1642d8e..c773caf 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c
@@ -216,6 +216,20 @@ nvkm_fifo_uevent(struct nvkm_fifo *fifo)
 }
 
 static int
+nvkm_fifo_class_new_(struct nvkm_device *device,
+		     const struct nvkm_oclass *oclass, void *data, u32 size,
+		     struct nvkm_object **pobject)
+{
+	struct nvkm_fifo *fifo = nvkm_fifo(oclass->engine);
+	return fifo->func->class_new(fifo, oclass, data, size, pobject);
+}
+
+static const struct nvkm_device_oclass
+nvkm_fifo_class_ = {
+	.ctor = nvkm_fifo_class_new_,
+};
+
+static int
 nvkm_fifo_class_new(struct nvkm_device *device,
 		    const struct nvkm_oclass *oclass, void *data, u32 size,
 		    struct nvkm_object **pobject)
@@ -239,13 +253,9 @@ nvkm_fifo_class_get(struct nvkm_oclass *oclass, int index,
 	int c = 0;
 
 	if (fifo->func->class_get) {
-		int ret = fifo->func->class_get(fifo, index, &sclass);
-		if (ret == 0) {
-			oclass->base = sclass->base;
-			oclass->engn = sclass;
-			*class = &nvkm_fifo_class;
-			return 0;
-		}
+		int ret = fifo->func->class_get(fifo, index, oclass);
+		if (ret == 0)
+			*class = &nvkm_fifo_class_;
 		return ret;
 	}
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/changk104.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/changk104.h
index 1208e3d..08b4415f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/changk104.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/changk104.h
@@ -19,11 +19,6 @@ struct gk104_fifo_chan {
 	} engn[NVKM_SUBDEV_NR];
 };
 
-int gk104_fifo_gpfifo_new(struct nvkm_fifo *, const struct nvkm_oclass *,
+int gk104_fifo_gpfifo_new(struct gk104_fifo *, const struct nvkm_oclass *,
 			  void *data, u32 size, struct nvkm_object **);
-
-extern const struct nvkm_fifo_chan_oclass gk104_fifo_gpfifo_oclass;
-extern const struct nvkm_fifo_chan_oclass gk110_fifo_gpfifo_oclass;
-extern const struct nvkm_fifo_chan_oclass gm200_fifo_gpfifo_oclass;
-extern const struct nvkm_fifo_chan_oclass gp100_fifo_gpfifo_oclass;
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
index b6c23cf..316b2b1 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
@@ -95,15 +95,29 @@ gk104_fifo_engine_status(struct gk104_fifo *fifo, int engn,
 }
 
 static int
+gk104_fifo_class_new(struct nvkm_fifo *base, const struct nvkm_oclass *oclass,
+		     void *argv, u32 argc, struct nvkm_object **pobject)
+{
+	struct gk104_fifo *fifo = gk104_fifo(base);
+	if (oclass->engn == &fifo->func->chan) {
+		const struct gk104_fifo_chan_user *user = oclass->engn;
+		return user->ctor(fifo, oclass, argv, argc, pobject);
+	}
+	WARN_ON(1);
+	return -EINVAL;
+}
+
+static int
 gk104_fifo_class_get(struct nvkm_fifo *base, int index,
-		     const struct nvkm_fifo_chan_oclass **psclass)
+		     struct nvkm_oclass *oclass)
 {
 	struct gk104_fifo *fifo = gk104_fifo(base);
 	int c = 0;
 
-	while ((*psclass = fifo->func->chan[c])) {
-		if (c++ == index)
-			return 0;
+	if (fifo->func->chan.ctor && c++ == index) {
+		oclass->base =  fifo->func->chan.user;
+		oclass->engn = &fifo->func->chan;
+		return 0;
 	}
 
 	return c;
@@ -950,6 +964,7 @@ gk104_fifo_ = {
 	.uevent_fini = gk104_fifo_uevent_fini,
 	.recover_chan = gk104_fifo_recover_chan,
 	.class_get = gk104_fifo_class_get,
+	.class_new = gk104_fifo_class_new,
 };
 
 int
@@ -1096,10 +1111,7 @@ gk104_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.chan = {
-		&gk104_fifo_gpfifo_oclass,
-		NULL
-	},
+	.chan = {{0,0,KEPLER_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
index 0e8b572..41f1f36 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
@@ -51,7 +51,11 @@ struct gk104_fifo_func {
 		const struct nvkm_enum *gpcclient;
 	} fault;
 
-	const struct nvkm_fifo_chan_oclass *chan[];
+	struct gk104_fifo_chan_user {
+		struct nvkm_sclass user;
+		int (*ctor)(struct gk104_fifo *, const struct nvkm_oclass *,
+			    void *, u32, struct nvkm_object **);
+	} chan;
 };
 
 int gk104_fifo_new_(const struct gk104_fifo_func *, struct nvkm_device *,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk110.c
index ad792b6..9611bf1 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk110.c
@@ -24,6 +24,8 @@
 #include "gk104.h"
 #include "changk104.h"
 
+#include <nvif/class.h>
+
 static const struct gk104_fifo_func
 gk110_fifo = {
 	.fault.access = gk104_fifo_fault_access,
@@ -31,10 +33,7 @@ gk110_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.chan = {
-		&gk110_fifo_gpfifo_oclass,
-		NULL
-	},
+	.chan = {{0,0,KEPLER_CHANNEL_GPFIFO_B}, gk104_fifo_gpfifo_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c
index 5402d22..fabc690 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c
@@ -24,6 +24,8 @@
 #include "gk104.h"
 #include "changk104.h"
 
+#include <nvif/class.h>
+
 static const struct gk104_fifo_func
 gk208_fifo = {
 	.fault.access = gk104_fifo_fault_access,
@@ -31,10 +33,7 @@ gk208_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.chan = {
-		&gk104_fifo_gpfifo_oclass,
-		NULL
-	},
+	.chan = {{0,0,KEPLER_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c
index 0d7f9f5..dee1bd7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c
@@ -22,6 +22,8 @@
 #include "gk104.h"
 #include "changk104.h"
 
+#include <nvif/class.h>
+
 static const struct gk104_fifo_func
 gk20a_fifo = {
 	.fault.access = gk104_fifo_fault_access,
@@ -29,10 +31,7 @@ gk20a_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.chan = {
-		&gk104_fifo_gpfifo_oclass,
-		NULL
-	},
+	.chan = {{0,0,KEPLER_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c
index 58a46ee..09f9782 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c
@@ -24,6 +24,8 @@
 #include "gk104.h"
 #include "changk104.h"
 
+#include <nvif/class.h>
+
 const struct nvkm_enum
 gm107_fifo_fault_engine[] = {
 	{ 0x01, "DISPLAY" },
@@ -54,10 +56,7 @@ gm107_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.chan = {
-		&gk110_fifo_gpfifo_oclass,
-		NULL
-	},
+	.chan = {{0,0,KEPLER_CHANNEL_GPFIFO_B}, gk104_fifo_gpfifo_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c
index f84d539..052b7c2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c
@@ -24,6 +24,8 @@
 #include "gk104.h"
 #include "changk104.h"
 
+#include <nvif/class.h>
+
 static const struct gk104_fifo_func
 gm200_fifo = {
 	.fault.access = gk104_fifo_fault_access,
@@ -31,10 +33,7 @@ gm200_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.chan = {
-		&gm200_fifo_gpfifo_oclass,
-		NULL
-	},
+	.chan = {{0,0,MAXWELL_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c
index 66399b9..844787c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c
@@ -22,6 +22,8 @@
 #include "gk104.h"
 #include "changk104.h"
 
+#include <nvif/class.h>
+
 static const struct gk104_fifo_func
 gm20b_fifo = {
 	.fault.access = gk104_fifo_fault_access,
@@ -29,10 +31,7 @@ gm20b_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.chan = {
-		&gm200_fifo_gpfifo_oclass,
-		NULL
-	},
+	.chan = {{0,0,MAXWELL_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c
index 2b8a6cf..2e31847 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c
@@ -24,6 +24,8 @@
 #include "gk104.h"
 #include "changk104.h"
 
+#include <nvif/class.h>
+
 const struct nvkm_enum
 gp100_fifo_fault_engine[] = {
 	{ 0x01, "DISPLAY" },
@@ -55,10 +57,7 @@ gp100_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.chan = {
-		&gp100_fifo_gpfifo_oclass,
-		NULL
-	},
+	.chan = {{0,0,PASCAL_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c
index 71b8d93..38ab6e1 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c
@@ -22,6 +22,8 @@
 #include "gk104.h"
 #include "changk104.h"
 
+#include <nvif/class.h>
+
 static const struct gk104_fifo_func
 gp10b_fifo = {
 	.fault.access = gk104_fifo_fault_access,
@@ -29,10 +31,7 @@ gp10b_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.chan = {
-		&gp100_fifo_gpfifo_oclass,
-		NULL
-	},
+	.chan = {{0,0,PASCAL_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c
index e331ab1..f7a4e0e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c
@@ -296,14 +296,13 @@ gk104_fifo_gpfifo_new_(struct gk104_fifo *fifo, u64 *runlists, u16 *chid,
 }
 
 int
-gk104_fifo_gpfifo_new(struct nvkm_fifo *base, const struct nvkm_oclass *oclass,
+gk104_fifo_gpfifo_new(struct gk104_fifo *fifo, const struct nvkm_oclass *oclass,
 		      void *data, u32 size, struct nvkm_object **pobject)
 {
 	struct nvkm_object *parent = oclass->parent;
 	union {
 		struct kepler_channel_gpfifo_a_v0 v0;
 	} *args = data;
-	struct gk104_fifo *fifo = gk104_fifo(base);
 	int ret = -ENOSYS;
 
 	nvif_ioctl(parent, "create channel gpfifo size %d\n", size);
@@ -324,11 +323,3 @@ gk104_fifo_gpfifo_new(struct nvkm_fifo *base, const struct nvkm_oclass *oclass,
 
 	return ret;
 }
-
-const struct nvkm_fifo_chan_oclass
-gk104_fifo_gpfifo_oclass = {
-	.base.oclass = KEPLER_CHANNEL_GPFIFO_A,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = gk104_fifo_gpfifo_new,
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk110.c
deleted file mode 100644
index a9aa69c..0000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk110.c
+++ /dev/null
@@ -1,34 +0,0 @@
-/*
- * Copyright 2016 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs
- */
-#include "changk104.h"
-
-#include <nvif/class.h>
-
-const struct nvkm_fifo_chan_oclass
-gk110_fifo_gpfifo_oclass = {
-	.base.oclass = KEPLER_CHANNEL_GPFIFO_B,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = gk104_fifo_gpfifo_new,
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogm200.c
deleted file mode 100644
index a133151..0000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogm200.c
+++ /dev/null
@@ -1,34 +0,0 @@
-/*
- * Copyright 2015 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs
- */
-#include "changk104.h"
-
-#include <nvif/class.h>
-
-const struct nvkm_fifo_chan_oclass
-gm200_fifo_gpfifo_oclass = {
-	.base.oclass = MAXWELL_CHANNEL_GPFIFO_A,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = gk104_fifo_gpfifo_new,
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogp100.c
deleted file mode 100644
index 1530a92..0000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogp100.c
+++ /dev/null
@@ -1,34 +0,0 @@
-/*
- * Copyright 2016 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs
- */
-#include "changk104.h"
-
-#include <nvif/class.h>
-
-const struct nvkm_fifo_chan_oclass
-gp100_fifo_gpfifo_oclass = {
-	.base.oclass = PASCAL_CHANNEL_GPFIFO_A,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = gk104_fifo_gpfifo_new,
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h
index f9b4e9d..d5acbba 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h
@@ -28,8 +28,9 @@ struct nvkm_fifo_func {
 	void (*uevent_init)(struct nvkm_fifo *);
 	void (*uevent_fini)(struct nvkm_fifo *);
 	void (*recover_chan)(struct nvkm_fifo *, int chid);
-	int (*class_get)(struct nvkm_fifo *, int index,
-			 const struct nvkm_fifo_chan_oclass **);
+	int (*class_get)(struct nvkm_fifo *, int index, struct nvkm_oclass *);
+	int (*class_new)(struct nvkm_fifo *, const struct nvkm_oclass *,
+			 void *, u32, struct nvkm_object **);
 	const struct nvkm_fifo_chan_oclass *chan[];
 };
 
-- 
cgit v1.1


From 665870837a764fc7ba6f0e1291291e86f909c29b Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/fifo/gk104-: add interfaces to support different runlist
 layouts

This will be required to support features on newer hardware.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c | 40 +++++++++++++++---------
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h | 10 ++++++
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk110.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c |  1 +
 10 files changed, 44 insertions(+), 14 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
index 316b2b1..afb3ed0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
@@ -140,6 +140,7 @@ gk104_fifo_uevent_init(struct nvkm_fifo *fifo)
 void
 gk104_fifo_runlist_commit(struct gk104_fifo *fifo, int runl)
 {
+	const struct gk104_fifo_runlist_func *func = fifo->func->runlist;
 	struct gk104_fifo_chan *chan;
 	struct nvkm_subdev *subdev = &fifo->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
@@ -153,9 +154,7 @@ gk104_fifo_runlist_commit(struct gk104_fifo *fifo, int runl)
 
 	nvkm_kmap(mem);
 	list_for_each_entry(chan, &fifo->runlist[runl].chan, head) {
-		nvkm_wo32(mem, (nr * 8) + 0, chan->base.chid);
-		nvkm_wo32(mem, (nr * 8) + 4, 0x00000000);
-		nr++;
+		func->chan(chan, mem, nr++ * func->size);
 	}
 	nvkm_done(mem);
 
@@ -196,6 +195,20 @@ gk104_fifo_runlist_insert(struct gk104_fifo *fifo, struct gk104_fifo_chan *chan)
 	mutex_unlock(&fifo->base.engine.subdev.mutex);
 }
 
+void
+gk104_fifo_runlist_chan(struct gk104_fifo_chan *chan,
+			struct nvkm_memory *memory, u32 offset)
+{
+	nvkm_wo32(memory, offset + 0, chan->base.chid);
+	nvkm_wo32(memory, offset + 4, 0x00000000);
+}
+
+const struct gk104_fifo_runlist_func
+gk104_fifo_runlist = {
+	.size = 8,
+	.chan = gk104_fifo_runlist_chan,
+};
+
 static void
 gk104_fifo_recover_work(struct work_struct *w)
 {
@@ -874,17 +887,15 @@ gk104_fifo_oneinit(struct nvkm_fifo *base)
 	kfree(map);
 
 	for (i = 0; i < fifo->runlist_nr; i++) {
-		ret = nvkm_memory_new(device, NVKM_MEM_TARGET_INST,
-				      0x8000, 0x1000, false,
-				      &fifo->runlist[i].mem[0]);
-		if (ret)
-			return ret;
-
-		ret = nvkm_memory_new(device, NVKM_MEM_TARGET_INST,
-				      0x8000, 0x1000, false,
-				      &fifo->runlist[i].mem[1]);
-		if (ret)
-			return ret;
+		for (j = 0; j < ARRAY_SIZE(fifo->runlist[i].mem); j++) {
+			ret = nvkm_memory_new(device, NVKM_MEM_TARGET_INST,
+					      fifo->base.nr * 2/* TSG+chan */ *
+					      fifo->func->runlist->size,
+					      0x1000, false,
+					      &fifo->runlist[i].mem[j]);
+			if (ret)
+				return ret;
+		}
 
 		init_waitqueue_head(&fifo->runlist[i].wait);
 		INIT_LIST_HEAD(&fifo->runlist[i].chan);
@@ -1111,6 +1122,7 @@ gk104_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
+	.runlist = &gk104_fifo_runlist,
 	.chan = {{0,0,KEPLER_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
index 41f1f36..c32ea45 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
@@ -51,6 +51,12 @@ struct gk104_fifo_func {
 		const struct nvkm_enum *gpcclient;
 	} fault;
 
+	const struct gk104_fifo_runlist_func {
+		u8 size;
+		void (*chan)(struct gk104_fifo_chan *,
+			     struct nvkm_memory *, u32 offset);
+	} *runlist;
+
 	struct gk104_fifo_chan_user {
 		struct nvkm_sclass user;
 		int (*ctor)(struct gk104_fifo *, const struct nvkm_oclass *,
@@ -65,10 +71,14 @@ void gk104_fifo_runlist_remove(struct gk104_fifo *, struct gk104_fifo_chan *);
 void gk104_fifo_runlist_commit(struct gk104_fifo *, int runl);
 
 extern const struct nvkm_enum gk104_fifo_fault_access[];
+
 extern const struct nvkm_enum gk104_fifo_fault_engine[];
 extern const struct nvkm_enum gk104_fifo_fault_reason[];
 extern const struct nvkm_enum gk104_fifo_fault_hubclient[];
 extern const struct nvkm_enum gk104_fifo_fault_gpcclient[];
+extern const struct gk104_fifo_runlist_func gk104_fifo_runlist;
+void gk104_fifo_runlist_chan(struct gk104_fifo_chan *,
+			     struct nvkm_memory *, u32);
 
 extern const struct nvkm_enum gm107_fifo_fault_engine[];
 extern const struct nvkm_enum gp100_fifo_fault_engine[];
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk110.c
index 9611bf1..f7e1604 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk110.c
@@ -33,6 +33,7 @@ gk110_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
+	.runlist = &gk104_fifo_runlist,
 	.chan = {{0,0,KEPLER_CHANNEL_GPFIFO_B}, gk104_fifo_gpfifo_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c
index fabc690..ff936ab 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c
@@ -33,6 +33,7 @@ gk208_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
+	.runlist = &gk104_fifo_runlist,
 	.chan = {{0,0,KEPLER_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c
index dee1bd7..cb78d4d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c
@@ -31,6 +31,7 @@ gk20a_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
+	.runlist = &gk104_fifo_runlist,
 	.chan = {{0,0,KEPLER_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c
index 09f9782..f749d93 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c
@@ -56,6 +56,7 @@ gm107_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
+	.runlist = &gk104_fifo_runlist,
 	.chan = {{0,0,KEPLER_CHANNEL_GPFIFO_B}, gk104_fifo_gpfifo_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c
index 052b7c2..3b97bd4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c
@@ -33,6 +33,7 @@ gm200_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
+	.runlist = &gk104_fifo_runlist,
 	.chan = {{0,0,MAXWELL_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c
index 844787c..c7e11a4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c
@@ -31,6 +31,7 @@ gm20b_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
+	.runlist = &gk104_fifo_runlist,
 	.chan = {{0,0,MAXWELL_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c
index 2e31847..5c0cc40 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c
@@ -57,6 +57,7 @@ gp100_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
+	.runlist = &gk104_fifo_runlist,
 	.chan = {{0,0,PASCAL_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c
index 38ab6e1..fe3a8fa 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c
@@ -31,6 +31,7 @@ gp10b_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
+	.runlist = &gk104_fifo_runlist,
 	.chan = {{0,0,PASCAL_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
-- 
cgit v1.1


From 4f2fc25c0f8bcc8db1b8a7b21e88c3d7f35c5acb Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/fifo/gk104-: poll for runlist update completion

Newer HW doesn't appear to send this event, which will cause long delays
in runlist updates if they don't complete immediately.

RM doesn't use these events anywhere, and an NVGPU commit message notes
that polling is the preferred method even on HW that supports the event.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
index afb3ed0..2b8e0d3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
@@ -170,10 +170,10 @@ gk104_fifo_runlist_commit(struct gk104_fifo *fifo, int runl)
 				    (target << 28));
 	nvkm_wr32(device, 0x002274, (runl << 20) | nr);
 
-	if (wait_event_timeout(fifo->runlist[runl].wait,
-			       !(nvkm_rd32(device, 0x002284 + (runl * 0x08))
-				       & 0x00100000),
-			       msecs_to_jiffies(2000)) == 0)
+	if (nvkm_msec(device, 2000,
+		if (!(nvkm_rd32(device, 0x002284 + (runl * 0x08)) & 0x00100000))
+			break;
+	) < 0)
 		nvkm_error(subdev, "runlist %d update timeout\n", runl);
 unlock:
 	mutex_unlock(&subdev->mutex);
-- 
cgit v1.1


From 8c4e9f9dffb96a891d31e108b47f081233cb3e81 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/fifo/gk110-: support writing channel group runlist
 entries

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/cgrp.h    | 11 +++++++++
 .../gpu/drm/nouveau/nvkm/engine/fifo/changk104.h   |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c   | 26 ++++++++++++++++++++--
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h   |  9 +++++++-
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk110.c   | 21 ++++++++++++++++-
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c   |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c   |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c   |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c   |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c   |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c   |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c   |  2 +-
 12 files changed, 71 insertions(+), 11 deletions(-)
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/fifo/cgrp.h

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/cgrp.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/cgrp.h
new file mode 100644
index 0000000..d0ac60b
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/cgrp.h
@@ -0,0 +1,11 @@
+#ifndef __NVKM_FIFO_CGRP_H__
+#define __NVKM_FIFO_CGRP_H__
+#include "priv.h"
+
+struct nvkm_fifo_cgrp {
+	int id;
+	struct list_head head;
+	struct list_head chan;
+	int chan_nr;
+};
+#endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/changk104.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/changk104.h
index 08b4415f..391e864 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/changk104.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/changk104.h
@@ -10,6 +10,7 @@ struct gk104_fifo_chan {
 	struct gk104_fifo *fifo;
 	int runl;
 
+	struct nvkm_fifo_cgrp *cgrp;
 	struct list_head head;
 	bool killed;
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
index 2b8e0d3..bc87f18 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
@@ -22,6 +22,7 @@
  * Authors: Ben Skeggs
  */
 #include "gk104.h"
+#include "cgrp.h"
 #include "changk104.h"
 
 #include <core/client.h>
@@ -145,6 +146,7 @@ gk104_fifo_runlist_commit(struct gk104_fifo *fifo, int runl)
 	struct nvkm_subdev *subdev = &fifo->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 	struct nvkm_memory *mem;
+	struct nvkm_fifo_cgrp *cgrp;
 	int nr = 0;
 	int target;
 
@@ -156,6 +158,13 @@ gk104_fifo_runlist_commit(struct gk104_fifo *fifo, int runl)
 	list_for_each_entry(chan, &fifo->runlist[runl].chan, head) {
 		func->chan(chan, mem, nr++ * func->size);
 	}
+
+	list_for_each_entry(cgrp, &fifo->runlist[runl].cgrp, head) {
+		func->cgrp(cgrp, mem, nr++ * func->size);
+		list_for_each_entry(chan, &cgrp->chan, head) {
+			func->chan(chan, mem, nr++ * func->size);
+		}
+	}
 	nvkm_done(mem);
 
 	switch (nvkm_memory_target(mem)) {
@@ -182,16 +191,28 @@ unlock:
 void
 gk104_fifo_runlist_remove(struct gk104_fifo *fifo, struct gk104_fifo_chan *chan)
 {
+	struct nvkm_fifo_cgrp *cgrp = chan->cgrp;
 	mutex_lock(&fifo->base.engine.subdev.mutex);
-	list_del_init(&chan->head);
+	if (!list_empty(&chan->head)) {
+		list_del_init(&chan->head);
+		if (cgrp && !--cgrp->chan_nr)
+			list_del_init(&cgrp->head);
+	}
 	mutex_unlock(&fifo->base.engine.subdev.mutex);
 }
 
 void
 gk104_fifo_runlist_insert(struct gk104_fifo *fifo, struct gk104_fifo_chan *chan)
 {
+	struct nvkm_fifo_cgrp *cgrp = chan->cgrp;
 	mutex_lock(&fifo->base.engine.subdev.mutex);
-	list_add_tail(&chan->head, &fifo->runlist[chan->runl].chan);
+	if (cgrp) {
+		if (!cgrp->chan_nr++)
+			list_add_tail(&cgrp->head, &fifo->runlist[chan->runl].cgrp);
+		list_add_tail(&chan->head, &cgrp->chan);
+	} else {
+		list_add_tail(&chan->head, &fifo->runlist[chan->runl].chan);
+	}
 	mutex_unlock(&fifo->base.engine.subdev.mutex);
 }
 
@@ -898,6 +919,7 @@ gk104_fifo_oneinit(struct nvkm_fifo *base)
 		}
 
 		init_waitqueue_head(&fifo->runlist[i].wait);
+		INIT_LIST_HEAD(&fifo->runlist[i].cgrp);
 		INIT_LIST_HEAD(&fifo->runlist[i].chan);
 	}
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
index c32ea45..dfb3da84 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
@@ -3,6 +3,7 @@
 #define __GK104_FIFO_H__
 #define gk104_fifo(p) container_of((p), struct gk104_fifo, base)
 #include "priv.h"
+struct nvkm_fifo_cgrp;
 
 #include <core/enum.h>
 #include <subdev/mmu.h>
@@ -31,6 +32,7 @@ struct gk104_fifo {
 		struct nvkm_memory *mem[2];
 		int next;
 		wait_queue_head_t wait;
+		struct list_head cgrp;
 		struct list_head chan;
 		u32 engm;
 	} runlist[16];
@@ -53,6 +55,8 @@ struct gk104_fifo_func {
 
 	const struct gk104_fifo_runlist_func {
 		u8 size;
+		void (*cgrp)(struct nvkm_fifo_cgrp *,
+			     struct nvkm_memory *, u32 offset);
 		void (*chan)(struct gk104_fifo_chan *,
 			     struct nvkm_memory *, u32 offset);
 	} *runlist;
@@ -71,7 +75,6 @@ void gk104_fifo_runlist_remove(struct gk104_fifo *, struct gk104_fifo_chan *);
 void gk104_fifo_runlist_commit(struct gk104_fifo *, int runl);
 
 extern const struct nvkm_enum gk104_fifo_fault_access[];
-
 extern const struct nvkm_enum gk104_fifo_fault_engine[];
 extern const struct nvkm_enum gk104_fifo_fault_reason[];
 extern const struct nvkm_enum gk104_fifo_fault_hubclient[];
@@ -80,6 +83,10 @@ extern const struct gk104_fifo_runlist_func gk104_fifo_runlist;
 void gk104_fifo_runlist_chan(struct gk104_fifo_chan *,
 			     struct nvkm_memory *, u32);
 
+extern const struct gk104_fifo_runlist_func gk110_fifo_runlist;
+void gk110_fifo_runlist_cgrp(struct nvkm_fifo_cgrp *,
+			     struct nvkm_memory *, u32);
+
 extern const struct nvkm_enum gm107_fifo_fault_engine[];
 extern const struct nvkm_enum gp100_fifo_fault_engine[];
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk110.c
index f7e1604..ac7655a1 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk110.c
@@ -22,10 +22,29 @@
  * Authors: Ben Skeggs
  */
 #include "gk104.h"
+#include "cgrp.h"
 #include "changk104.h"
 
+#include <core/memory.h>
+
 #include <nvif/class.h>
 
+void
+gk110_fifo_runlist_cgrp(struct nvkm_fifo_cgrp *cgrp,
+			struct nvkm_memory *memory, u32 offset)
+{
+	nvkm_wo32(memory, offset + 0, (cgrp->chan_nr << 26) | (128 << 18) |
+				      (3 << 14) | 0x00002000 | cgrp->id);
+	nvkm_wo32(memory, offset + 4, 0x00000000);
+}
+
+const struct gk104_fifo_runlist_func
+gk110_fifo_runlist = {
+	.size = 8,
+	.cgrp = gk110_fifo_runlist_cgrp,
+	.chan = gk104_fifo_runlist_chan,
+};
+
 static const struct gk104_fifo_func
 gk110_fifo = {
 	.fault.access = gk104_fifo_fault_access,
@@ -33,7 +52,7 @@ gk110_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.runlist = &gk104_fifo_runlist,
+	.runlist = &gk110_fifo_runlist,
 	.chan = {{0,0,KEPLER_CHANNEL_GPFIFO_B}, gk104_fifo_gpfifo_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c
index ff936ab..b7385cd 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c
@@ -33,7 +33,7 @@ gk208_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.runlist = &gk104_fifo_runlist,
+	.runlist = &gk110_fifo_runlist,
 	.chan = {{0,0,KEPLER_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c
index cb78d4d..15cc80c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c
@@ -31,7 +31,7 @@ gk20a_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.runlist = &gk104_fifo_runlist,
+	.runlist = &gk110_fifo_runlist,
 	.chan = {{0,0,KEPLER_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c
index f749d93..99b89f6a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c
@@ -56,7 +56,7 @@ gm107_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.runlist = &gk104_fifo_runlist,
+	.runlist = &gk110_fifo_runlist,
 	.chan = {{0,0,KEPLER_CHANNEL_GPFIFO_B}, gk104_fifo_gpfifo_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c
index 3b97bd4..16fd8de 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c
@@ -33,7 +33,7 @@ gm200_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.runlist = &gk104_fifo_runlist,
+	.runlist = &gk110_fifo_runlist,
 	.chan = {{0,0,MAXWELL_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c
index c7e11a4..bfa5fa5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c
@@ -31,7 +31,7 @@ gm20b_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.runlist = &gk104_fifo_runlist,
+	.runlist = &gk110_fifo_runlist,
 	.chan = {{0,0,MAXWELL_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c
index 5c0cc40..c285294 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c
@@ -57,7 +57,7 @@ gp100_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.runlist = &gk104_fifo_runlist,
+	.runlist = &gk110_fifo_runlist,
 	.chan = {{0,0,PASCAL_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c
index fe3a8fa..29a4029 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c
@@ -31,7 +31,7 @@ gp10b_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.runlist = &gk104_fifo_runlist,
+	.runlist = &gk110_fifo_runlist,
 	.chan = {{0,0,PASCAL_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
-- 
cgit v1.1


From 79bb4b617f965736d2e1c616235302c1d0e823b2 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/fifo/gk208-: write pbdma timeout regs during
 initialisation

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c |  3 +++
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h |  4 ++++
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c | 11 +++++++++++
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c |  1 +
 9 files changed, 24 insertions(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
index bc87f18..5c22ce9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
@@ -962,6 +962,9 @@ gk104_fifo_init(struct nvkm_fifo *base)
 
 	nvkm_wr32(device, 0x002254, 0x10000000 | fifo->user.bar->addr >> 12);
 
+	if (fifo->func->init_pbdma_timeout)
+		fifo->func->init_pbdma_timeout(fifo);
+
 	nvkm_wr32(device, 0x002100, 0xffffffff);
 	nvkm_wr32(device, 0x002140, 0x7fffffff);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
index dfb3da84..e232cee 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
@@ -45,6 +45,8 @@ struct gk104_fifo {
 };
 
 struct gk104_fifo_func {
+	void (*init_pbdma_timeout)(struct gk104_fifo *);
+
 	struct {
 		const struct nvkm_enum *access;
 		const struct nvkm_enum *engine;
@@ -87,6 +89,8 @@ extern const struct gk104_fifo_runlist_func gk110_fifo_runlist;
 void gk110_fifo_runlist_cgrp(struct nvkm_fifo_cgrp *,
 			     struct nvkm_memory *, u32);
 
+void gk208_fifo_init_pbdma_timeout(struct gk104_fifo *);
+
 extern const struct nvkm_enum gm107_fifo_fault_engine[];
 extern const struct nvkm_enum gp100_fifo_fault_engine[];
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c
index b7385cd..5ea7e45 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c
@@ -26,8 +26,19 @@
 
 #include <nvif/class.h>
 
+void
+gk208_fifo_init_pbdma_timeout(struct gk104_fifo *fifo)
+{
+	struct nvkm_device *device = fifo->base.engine.subdev.device;
+	int i;
+
+	for (i = 0; i < fifo->pbdma_nr; i++)
+		nvkm_wr32(device, 0x04012c + (i * 0x2000), 0x0000ffff);
+}
+
 static const struct gk104_fifo_func
 gk208_fifo = {
+	.init_pbdma_timeout = gk208_fifo_init_pbdma_timeout,
 	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gk104_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c
index 15cc80c..535a0eb 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c
@@ -26,6 +26,7 @@
 
 static const struct gk104_fifo_func
 gk20a_fifo = {
+	.init_pbdma_timeout = gk208_fifo_init_pbdma_timeout,
 	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gk104_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c
index 99b89f6a..a28f5d2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c
@@ -51,6 +51,7 @@ gm107_fifo_fault_engine[] = {
 
 static const struct gk104_fifo_func
 gm107_fifo = {
+	.init_pbdma_timeout = gk208_fifo_init_pbdma_timeout,
 	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gm107_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c
index 16fd8de..0bd87a0b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c
@@ -28,6 +28,7 @@
 
 static const struct gk104_fifo_func
 gm200_fifo = {
+	.init_pbdma_timeout = gk208_fifo_init_pbdma_timeout,
 	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gm107_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c
index bfa5fa5..32a8e3d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c
@@ -26,6 +26,7 @@
 
 static const struct gk104_fifo_func
 gm20b_fifo = {
+	.init_pbdma_timeout = gk208_fifo_init_pbdma_timeout,
 	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gm107_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c
index c285294..69b201f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c
@@ -52,6 +52,7 @@ gp100_fifo_fault_engine[] = {
 
 static const struct gk104_fifo_func
 gp100_fifo = {
+	.init_pbdma_timeout = gk208_fifo_init_pbdma_timeout,
 	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gp100_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c
index 29a4029..fac7be5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c
@@ -26,6 +26,7 @@
 
 static const struct gk104_fifo_func
 gp10b_fifo = {
+	.init_pbdma_timeout = gk208_fifo_init_pbdma_timeout,
 	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gp100_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
-- 
cgit v1.1


From eda12417d3daf6cb37f41e9b1c46854e7dd75e91 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/fifo/gm107-: write instance address in channel runlist
 entry

RM does this for some reason.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c | 19 ++++++++++++++++++-
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c |  2 +-
 6 files changed, 24 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
index e232cee..1d70a5d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
@@ -92,5 +92,7 @@ void gk110_fifo_runlist_cgrp(struct nvkm_fifo_cgrp *,
 void gk208_fifo_init_pbdma_timeout(struct gk104_fifo *);
 
 extern const struct nvkm_enum gm107_fifo_fault_engine[];
+extern const struct gk104_fifo_runlist_func gm107_fifo_runlist;
+
 extern const struct nvkm_enum gp100_fifo_fault_engine[];
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c
index a28f5d2..79ae19b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c
@@ -24,8 +24,25 @@
 #include "gk104.h"
 #include "changk104.h"
 
+#include <core/gpuobj.h>
+
 #include <nvif/class.h>
 
+static void
+gm107_fifo_runlist_chan(struct gk104_fifo_chan *chan,
+			struct nvkm_memory *memory, u32 offset)
+{
+	nvkm_wo32(memory, offset + 0, chan->base.chid);
+	nvkm_wo32(memory, offset + 4, chan->base.inst->addr >> 12);
+}
+
+const struct gk104_fifo_runlist_func
+gm107_fifo_runlist = {
+	.size = 8,
+	.cgrp = gk110_fifo_runlist_cgrp,
+	.chan = gm107_fifo_runlist_chan,
+};
+
 const struct nvkm_enum
 gm107_fifo_fault_engine[] = {
 	{ 0x01, "DISPLAY" },
@@ -57,7 +74,7 @@ gm107_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.runlist = &gk110_fifo_runlist,
+	.runlist = &gm107_fifo_runlist,
 	.chan = {{0,0,KEPLER_CHANNEL_GPFIFO_B}, gk104_fifo_gpfifo_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c
index 0bd87a0b..49565fa 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c
@@ -34,7 +34,7 @@ gm200_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.runlist = &gk110_fifo_runlist,
+	.runlist = &gm107_fifo_runlist,
 	.chan = {{0,0,MAXWELL_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c
index 32a8e3d..46736513 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c
@@ -32,7 +32,7 @@ gm20b_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.runlist = &gk110_fifo_runlist,
+	.runlist = &gm107_fifo_runlist,
 	.chan = {{0,0,MAXWELL_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c
index 69b201f..f137bae 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c
@@ -58,7 +58,7 @@ gp100_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.runlist = &gk110_fifo_runlist,
+	.runlist = &gm107_fifo_runlist,
 	.chan = {{0,0,PASCAL_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c
index fac7be5..787e911 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c
@@ -32,7 +32,7 @@ gp10b_fifo = {
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.runlist = &gk110_fifo_runlist,
+	.runlist = &gm107_fifo_runlist,
 	.chan = {{0,0,PASCAL_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
-- 
cgit v1.1


From 334cc26d4db10ae7d8f18de27869b95fe84c7d28 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/fifo/gp100-: force individual channels into a channel
 group

RM does this for some reason, and is enforced in HW on Volta.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c   | 37 ++++++++++++++++++----
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h   |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c   |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c   |  1 +
 .../gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c | 28 +++++++++++++---
 5 files changed, 57 insertions(+), 11 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
index 5c22ce9..767e0ab 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
@@ -285,6 +285,32 @@ gk104_fifo_recover_runl(struct gk104_fifo *fifo, int runl)
 	schedule_work(&fifo->recover.work);
 }
 
+static struct gk104_fifo_chan *
+gk104_fifo_recover_chid(struct gk104_fifo *fifo, int runl, int chid)
+{
+	struct gk104_fifo_chan *chan;
+	struct nvkm_fifo_cgrp *cgrp;
+
+	list_for_each_entry(chan, &fifo->runlist[runl].chan, head) {
+		if (chan->base.chid == chid) {
+			list_del_init(&chan->head);
+			return chan;
+		}
+	}
+
+	list_for_each_entry(cgrp, &fifo->runlist[runl].cgrp, head) {
+		if (cgrp->id == chid) {
+			chan = list_first_entry(&cgrp->chan, typeof(*chan), head);
+			list_del_init(&chan->head);
+			if (!--cgrp->chan_nr)
+				list_del_init(&cgrp->head);
+			return chan;
+		}
+	}
+
+	return NULL;
+}
+
 static void
 gk104_fifo_recover_chan(struct nvkm_fifo *base, int chid)
 {
@@ -302,13 +328,10 @@ gk104_fifo_recover_chan(struct nvkm_fifo *base, int chid)
 		return;
 
 	/* Lookup SW state for channel, and mark it as dead. */
-	list_for_each_entry(chan, &fifo->runlist[runl].chan, head) {
-		if (chan->base.chid == chid) {
-			list_del_init(&chan->head);
-			chan->killed = true;
-			nvkm_fifo_kevent(&fifo->base, chid);
-			break;
-		}
+	chan = gk104_fifo_recover_chid(fifo, runl, chid);
+	if (chan) {
+		chan->killed = true;
+		nvkm_fifo_kevent(&fifo->base, chid);
 	}
 
 	/* Disable channel. */
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
index 1d70a5d..1d182d8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
@@ -68,6 +68,7 @@ struct gk104_fifo_func {
 		int (*ctor)(struct gk104_fifo *, const struct nvkm_oclass *,
 			    void *, u32, struct nvkm_object **);
 	} chan;
+	bool cgrp_force;
 };
 
 int gk104_fifo_new_(const struct gk104_fifo_func *, struct nvkm_device *,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c
index f137bae..e2f8f90 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c
@@ -60,6 +60,7 @@ gp100_fifo = {
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
 	.runlist = &gm107_fifo_runlist,
 	.chan = {{0,0,PASCAL_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
+	.cgrp_force = true,
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c
index 787e911..7733bf7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c
@@ -34,6 +34,7 @@ gp10b_fifo = {
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
 	.runlist = &gm107_fifo_runlist,
 	.chan = {{0,0,PASCAL_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
+	.cgrp_force = true,
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c
index f7a4e0e..60e7d72 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c
@@ -22,6 +22,7 @@
  * Authors: Ben Skeggs
  */
 #include "changk104.h"
+#include "cgrp.h"
 
 #include <core/client.h>
 #include <core/gpuobj.h>
@@ -40,16 +41,21 @@ gk104_fifo_gpfifo_kick(struct gk104_fifo_chan *chan)
 	struct nvkm_subdev *subdev = &fifo->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 	struct nvkm_client *client = chan->base.object.client;
+	struct nvkm_fifo_cgrp *cgrp = chan->cgrp;
 	int ret = 0;
 
 	mutex_lock(&subdev->mutex);
-	nvkm_wr32(device, 0x002634, chan->base.chid);
+	if (cgrp)
+		nvkm_wr32(device, 0x002634, cgrp->id | 0x01000000);
+	else
+		nvkm_wr32(device, 0x002634, chan->base.chid);
 	if (nvkm_msec(device, 2000,
 		if (!(nvkm_rd32(device, 0x002634) & 0x00100000))
 			break;
 	) < 0) {
-		nvkm_error(subdev, "channel %d [%s] kick timeout\n",
-			   chan->base.chid, client->name);
+		nvkm_error(subdev, "%s %d [%s] kick timeout\n",
+			   cgrp ? "tsg" : "channel",
+			   cgrp ? cgrp->id : chan->base.chid, client->name);
 		nvkm_fifo_recover_chan(&fifo->base, chan->base.chid);
 		ret = -ETIMEDOUT;
 	}
@@ -207,7 +213,9 @@ gk104_fifo_gpfifo_init(struct nvkm_fifo_chan *base)
 static void *
 gk104_fifo_gpfifo_dtor(struct nvkm_fifo_chan *base)
 {
-	return gk104_fifo_chan(base);
+	struct gk104_fifo_chan *chan = gk104_fifo_chan(base);
+	kfree(chan->cgrp);
+	return chan;
 }
 
 static const struct nvkm_fifo_chan_func
@@ -264,6 +272,18 @@ gk104_fifo_gpfifo_new_(struct gk104_fifo *fifo, u64 *runlists, u16 *chid,
 
 	*chid = chan->base.chid;
 
+	/* Hack to support GPUs where even individual channels should be
+	 * part of a channel group.
+	 */
+	if (fifo->func->cgrp_force) {
+		if (!(chan->cgrp = kmalloc(sizeof(*chan->cgrp), GFP_KERNEL)))
+			return -ENOMEM;
+		chan->cgrp->id = chan->base.chid;
+		INIT_LIST_HEAD(&chan->cgrp->head);
+		INIT_LIST_HEAD(&chan->cgrp->chan);
+		chan->cgrp->chan_nr = 0;
+	}
+
 	/* Clear channel control registers. */
 	usermem = chan->base.chid * 0x200;
 	ilength = order_base_2(ilength / 8);
-- 
cgit v1.1


From 1246f1dc224a2c4ab61a1454cff669918b92e9da Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise init_gpc_mmu + apply fixes from
 traces

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c | 21 +++++++++++++++------
 2 files changed, 16 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 2f8dc10..5be4111 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -1920,7 +1920,7 @@ gf100_gr_init_gpc_mmu(struct gf100_gr *gr)
 	struct nvkm_fb *fb = device->fb;
 
 	nvkm_wr32(device, 0x418880, nvkm_rd32(device, 0x100c80) & 0x00000001);
-	nvkm_wr32(device, 0x4188a4, 0x00000000);
+	nvkm_wr32(device, 0x4188a4, 0x03000000);
 	nvkm_wr32(device, 0x418888, 0x00000000);
 	nvkm_wr32(device, 0x41888c, 0x00000000);
 	nvkm_wr32(device, 0x418890, 0x00000000);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
index 2c67fac..5ea74f6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
@@ -308,22 +308,30 @@ gm107_gr_init_bios(struct gf100_gr *gr)
 	}
 }
 
+static void
+gm107_gr_init_gpc_mmu(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	struct nvkm_fb *fb = device->fb;
+
+	nvkm_wr32(device, GPC_BCAST(0x0880), 0x00000000);
+	nvkm_wr32(device, GPC_BCAST(0x0890), 0x00000000);
+	nvkm_wr32(device, GPC_BCAST(0x0894), 0x00000000);
+	nvkm_wr32(device, GPC_BCAST(0x08b4), nvkm_memory_addr(fb->mmu_wr) >> 8);
+	nvkm_wr32(device, GPC_BCAST(0x08b8), nvkm_memory_addr(fb->mmu_rd) >> 8);
+}
+
 static int
 gm107_gr_init(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
-	struct nvkm_fb *fb = device->fb;
 	const u32 magicgpc918 = DIV_ROUND_UP(0x00800000, gr->tpc_total);
 	u32 data[TPC_MAX / 8] = {};
 	u8  tpcnr[GPC_MAX];
 	int gpc, tpc, rop;
 	int i;
 
-	nvkm_wr32(device, GPC_BCAST(0x0880), 0x00000000);
-	nvkm_wr32(device, GPC_BCAST(0x0890), 0x00000000);
-	nvkm_wr32(device, GPC_BCAST(0x0894), 0x00000000);
-	nvkm_wr32(device, GPC_BCAST(0x08b4), nvkm_memory_addr(fb->mmu_wr) >> 8);
-	nvkm_wr32(device, GPC_BCAST(0x08b8), nvkm_memory_addr(fb->mmu_rd) >> 8);
+	gr->func->init_gpc_mmu(gr);
 
 	gf100_gr_mmio(gr, gr->func->mmio);
 
@@ -442,6 +450,7 @@ gm107_gr_gpccs_ucode = {
 static const struct gf100_gr_func
 gm107_gr = {
 	.init = gm107_gr_init,
+	.init_gpc_mmu = gm107_gr_init_gpc_mmu,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gm107_gr_pack_mmio,
-- 
cgit v1.1


From 6f63a5fb1ec37cd55d7c886ac7d76f95a4cea2ce Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: support firmware-provided sw_nonctx everywhere

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 5be4111..eb50cfc 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -1941,7 +1941,10 @@ gf100_gr_init(struct gf100_gr *gr)
 
 	gr->func->init_gpc_mmu(gr);
 
-	gf100_gr_mmio(gr, gr->func->mmio);
+	if (gr->fuc_sw_nonctx)
+		gf100_gr_mmio(gr, gr->fuc_sw_nonctx);
+	else
+		gf100_gr_mmio(gr, gr->func->mmio);
 
 	nvkm_mask(device, TPC_UNIT(0, 0, 0x05c), 0x00000001, 0x00000001);
 
-- 
cgit v1.1


From 8b058ca5186535163bdcc55d81a9f341c71139f6 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise r405a14

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c | 3 +++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c | 7 +++++++
 3 files changed, 11 insertions(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index eb50cfc..d127c6a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -1946,6 +1946,9 @@ gf100_gr_init(struct gf100_gr *gr)
 	else
 		gf100_gr_mmio(gr, gr->func->mmio);
 
+	if (gr->func->init_r405a14)
+		gr->func->init_r405a14(gr);
+
 	nvkm_mask(device, TPC_UNIT(0, 0, 0x05c), 0x00000001, 0x00000001);
 
 	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index c8ec3fd..d940a1a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -122,6 +122,7 @@ struct gf100_gr_func {
 	void (*dtor)(struct gf100_gr *);
 	int (*init)(struct gf100_gr *);
 	void (*init_gpc_mmu)(struct gf100_gr *);
+	void (*init_r405a14)(struct gf100_gr *);
 	void (*init_rop_active_fbps)(struct gf100_gr *);
 	void (*init_ppc_exceptions)(struct gf100_gr *);
 	void (*init_swdx_pes_mask)(struct gf100_gr *);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
index cc152eb..3a59dea 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
@@ -103,10 +103,17 @@ gf108_gr_pack_mmio[] = {
  * PGRAPH engine/subdev functions
  ******************************************************************************/
 
+static void
+gf108_gr_init_r405a14(struct gf100_gr *gr)
+{
+	nvkm_wr32(gr->base.engine.subdev.device, 0x405a14, 0x80000000);
+}
+
 static const struct gf100_gr_func
 gf108_gr = {
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
+	.init_r405a14 = gf108_gr_init_r405a14,
 	.mmio = gf108_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
-- 
cgit v1.1


From cd9662f89e2cd953ce9b2fcf02fdaae847592bd1 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: support clkgate_pack everywhere

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index d127c6a..eaf32f7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -32,6 +32,7 @@
 #include <subdev/fb.h>
 #include <subdev/mc.h>
 #include <subdev/pmu.h>
+#include <subdev/therm.h>
 #include <subdev/timer.h>
 #include <engine/fifo.h>
 
@@ -1949,6 +1950,9 @@ gf100_gr_init(struct gf100_gr *gr)
 	if (gr->func->init_r405a14)
 		gr->func->init_r405a14(gr);
 
+	if (gr->func->clkgate_pack)
+		nvkm_therm_clkgate_init(device->therm, gr->func->clkgate_pack);
+
 	nvkm_mask(device, TPC_UNIT(0, 0, 0x05c), 0x00000001, 0x00000001);
 
 	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-- 
cgit v1.1


From a37279e94c91e3b30ee50dfd96f8f33f3be43f17 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise init_bios

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c | 3 +++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h | 2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c | 3 ++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c | 5 +++--
 4 files changed, 9 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index eaf32f7..6ecb4a0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -1953,6 +1953,9 @@ gf100_gr_init(struct gf100_gr *gr)
 	if (gr->func->clkgate_pack)
 		nvkm_therm_clkgate_init(device->therm, gr->func->clkgate_pack);
 
+	if (gr->func->init_bios)
+		gr->func->init_bios(gr);
+
 	nvkm_mask(device, TPC_UNIT(0, 0, 0x05c), 0x00000001, 0x00000001);
 
 	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index d940a1a..1320bef 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -123,6 +123,7 @@ struct gf100_gr_func {
 	int (*init)(struct gf100_gr *);
 	void (*init_gpc_mmu)(struct gf100_gr *);
 	void (*init_r405a14)(struct gf100_gr *);
+	void (*init_bios)(struct gf100_gr *);
 	void (*init_rop_active_fbps)(struct gf100_gr *);
 	void (*init_ppc_exceptions)(struct gf100_gr *);
 	void (*init_swdx_pes_mask)(struct gf100_gr *);
@@ -151,7 +152,6 @@ void gk104_gr_init_ppc_exceptions(struct gf100_gr *);
 
 int gk20a_gr_init(struct gf100_gr *);
 
-int gm200_gr_init(struct gf100_gr *);
 int gm200_gr_rops(struct gf100_gr *);
 
 int gp100_gr_init(struct gf100_gr *);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
index 5ea74f6..d5629cd 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
@@ -335,7 +335,7 @@ gm107_gr_init(struct gf100_gr *gr)
 
 	gf100_gr_mmio(gr, gr->func->mmio);
 
-	gm107_gr_init_bios(gr);
+	gr->func->init_bios(gr);
 
 	nvkm_wr32(device, GPC_UNIT(0, 0x3018), 0x00000001);
 
@@ -451,6 +451,7 @@ static const struct gf100_gr_func
 gm107_gr = {
 	.init = gm107_gr_init,
 	.init_gpc_mmu = gm107_gr_init_gpc_mmu,
+	.init_bios = gm107_gr_init_bios,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gm107_gr_pack_mmio,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
index 6435f12..7a793a0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
@@ -61,7 +61,7 @@ gm200_gr_init_rop_active_fbps(struct gf100_gr *gr)
 	nvkm_mask(device, 0x408958, 0x0000000f, fbp_count); /* crop */
 }
 
-int
+static int
 gm200_gr_init(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -75,7 +75,7 @@ gm200_gr_init(struct gf100_gr *gr)
 
 	gf100_gr_mmio(gr, gr->fuc_sw_nonctx);
 
-	gm107_gr_init_bios(gr);
+	gr->func->init_bios(gr);
 
 	nvkm_wr32(device, GPC_UNIT(0, 0x3018), 0x00000001);
 
@@ -210,6 +210,7 @@ static const struct gf100_gr_func
 gm200_gr = {
 	.init = gm200_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
+	.init_bios = gm107_gr_init_bios,
 	.init_rop_active_fbps = gm200_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
-- 
cgit v1.1


From 2fe5ff6371350ce224dc2cc1da0e01888a1f9999 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise init_vsc_stream_master

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c  | 10 +++++++++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h  |  5 ++++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c  | 10 +++++++++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c  |  3 ++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c  |  3 ++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c  |  3 ++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c  |  1 +
 17 files changed, 39 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 6ecb4a0..0c92b8c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -1930,6 +1930,13 @@ gf100_gr_init_gpc_mmu(struct gf100_gr *gr)
 	nvkm_wr32(device, 0x4188b8, nvkm_memory_addr(fb->mmu_rd) >> 8);
 }
 
+void
+gf100_gr_init_vsc_stream_master(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, TPC_UNIT(0, 0, 0x05c), 0x00000001, 0x00000001);
+}
+
 int
 gf100_gr_init(struct gf100_gr *gr)
 {
@@ -1956,7 +1963,7 @@ gf100_gr_init(struct gf100_gr *gr)
 	if (gr->func->init_bios)
 		gr->func->init_bios(gr);
 
-	nvkm_mask(device, TPC_UNIT(0, 0, 0x05c), 0x00000001, 0x00000001);
+	gr->func->init_vsc_stream_master(gr);
 
 	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
 	for (i = 0, gpc = -1; i < gr->tpc_total; i++) {
@@ -2068,6 +2075,7 @@ static const struct gf100_gr_func
 gf100_gr = {
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
 	.mmio = gf100_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index 1320bef..45792a3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -124,6 +124,7 @@ struct gf100_gr_func {
 	void (*init_gpc_mmu)(struct gf100_gr *);
 	void (*init_r405a14)(struct gf100_gr *);
 	void (*init_bios)(struct gf100_gr *);
+	void (*init_vsc_stream_master)(struct gf100_gr *);
 	void (*init_rop_active_fbps)(struct gf100_gr *);
 	void (*init_ppc_exceptions)(struct gf100_gr *);
 	void (*init_swdx_pes_mask)(struct gf100_gr *);
@@ -143,10 +144,12 @@ struct gf100_gr_func {
 	struct nvkm_sclass sclass[];
 };
 
-int gf100_gr_init(struct gf100_gr *);
 int gf100_gr_rops(struct gf100_gr *);
+int gf100_gr_init(struct gf100_gr *);
+void gf100_gr_init_vsc_stream_master(struct gf100_gr *);
 
 int gk104_gr_init(struct gf100_gr *);
+void gk104_gr_init_vsc_stream_master(struct gf100_gr *);
 void gk104_gr_init_rop_active_fbps(struct gf100_gr *);
 void gk104_gr_init_ppc_exceptions(struct gf100_gr *);
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
index ec0f119..61b27b1 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
@@ -116,6 +116,7 @@ static const struct gf100_gr_func
 gf104_gr = {
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
 	.mmio = gf104_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
index 3a59dea..992727f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
@@ -114,6 +114,7 @@ gf108_gr = {
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_r405a14 = gf108_gr_init_r405a14,
+	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
 	.mmio = gf108_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
index 10d2d73..07a57cc 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
@@ -88,6 +88,7 @@ static const struct gf100_gr_func
 gf110_gr = {
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
 	.mmio = gf110_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
index ac09a07..29ab01b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
@@ -124,6 +124,7 @@ static const struct gf100_gr_func
 gf117_gr = {
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
 	.mmio = gf117_gr_pack_mmio,
 	.fecs.ucode = &gf117_gr_fecs_ucode,
 	.gpccs.ucode = &gf117_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
index 7f449ec..d3b06dd7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
@@ -179,6 +179,7 @@ static const struct gf100_gr_func
 gf119_gr = {
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
 	.mmio = gf119_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
index 1b52fcb..835f498 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
@@ -404,6 +404,13 @@ gk104_gr_init_ppc_exceptions(struct gf100_gr *gr)
 	}
 }
 
+void
+gk104_gr_init_vsc_stream_master(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, GPC_UNIT(0, 0x3018), 0x00000001);
+}
+
 int
 gk104_gr_init(struct gf100_gr *gr)
 {
@@ -421,7 +428,7 @@ gk104_gr_init(struct gf100_gr *gr)
 		nvkm_therm_clkgate_init(gr->base.engine.subdev.device->therm,
 					gr->func->clkgate_pack);
 
-	nvkm_wr32(device, GPC_UNIT(0, 0x3018), 0x00000001);
+	gr->func->init_vsc_stream_master(gr);
 
 	memset(data, 0x00, sizeof(data));
 	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
@@ -536,6 +543,7 @@ static const struct gf100_gr_func
 gk104_gr = {
 	.init = gk104_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk104_gr_pack_mmio,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
index 4da916a..eef7476e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
@@ -338,6 +338,7 @@ static const struct gf100_gr_func
 gk110_gr = {
 	.init = gk104_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk110_gr_pack_mmio,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
index 1912c0b..ed4361a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
@@ -104,6 +104,7 @@ static const struct gf100_gr_func
 gk110b_gr = {
 	.init = gk104_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk110b_gr_pack_mmio,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
index 1fc2581..4c8beb9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
@@ -163,6 +163,7 @@ static const struct gf100_gr_func
 gk208_gr = {
 	.init = gk104_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk208_gr_pack_mmio,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
index d5629cd..58c03d6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
@@ -337,7 +337,7 @@ gm107_gr_init(struct gf100_gr *gr)
 
 	gr->func->init_bios(gr);
 
-	nvkm_wr32(device, GPC_UNIT(0, 0x3018), 0x00000001);
+	gr->func->init_vsc_stream_master(gr);
 
 	memset(data, 0x00, sizeof(data));
 	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
@@ -452,6 +452,7 @@ gm107_gr = {
 	.init = gm107_gr_init,
 	.init_gpc_mmu = gm107_gr_init_gpc_mmu,
 	.init_bios = gm107_gr_init_bios,
+	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gm107_gr_pack_mmio,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
index 7a793a0..8583aad 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
@@ -77,7 +77,7 @@ gm200_gr_init(struct gf100_gr *gr)
 
 	gr->func->init_bios(gr);
 
-	nvkm_wr32(device, GPC_UNIT(0, 0x3018), 0x00000001);
+	gr->func->init_vsc_stream_master(gr);
 
 	memset(data, 0x00, sizeof(data));
 	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
@@ -211,6 +211,7 @@ gm200_gr = {
 	.init = gm200_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_bios = gm107_gr_init_bios,
+	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_rop_active_fbps = gm200_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
index 867a5f7..edaaebb 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
@@ -63,7 +63,7 @@ gp100_gr_init(struct gf100_gr *gr)
 
 	gf100_gr_mmio(gr, gr->fuc_sw_nonctx);
 
-	nvkm_wr32(device, GPC_UNIT(0, 0x3018), 0x00000001);
+	gr->func->init_vsc_stream_master(gr);
 
 	memset(data, 0x00, sizeof(data));
 	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
@@ -160,6 +160,7 @@ static const struct gf100_gr_func
 gp100_gr = {
 	.init = gp100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_num_active_ltcs = gp100_gr_init_num_active_ltcs,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
index 61e3a0b..821a6c4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
@@ -44,6 +44,7 @@ static const struct gf100_gr_func
 gp102_gr = {
 	.init = gp100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_swdx_pes_mask = gp102_gr_init_swdx_pes_mask,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
index f727232..e2e1297 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
@@ -30,6 +30,7 @@ static const struct gf100_gr_func
 gp107_gr = {
 	.init = gp100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_swdx_pes_mask = gp102_gr_init_swdx_pes_mask,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
index 5f3d161..e90cc3c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
@@ -37,6 +37,7 @@ static const struct gf100_gr_func
 gp10b_gr = {
 	.init = gp100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_num_active_ltcs = gp10b_gr_init_num_active_ltcs,
-- 
cgit v1.1


From 02917aa39d56f504b47354135120000da1efa760 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise init_zcull

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c  | 62 ++++++++++++-------------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h  |  4 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c  | 29 ++++++++++++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c  | 31 +------------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c  | 33 ++-----------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c  | 31 +------------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c  | 31 +------------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c  | 31 +------------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c  |  1 +
 19 files changed, 85 insertions(+), 178 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 0c92b8c..56d9ead 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -1931,6 +1931,34 @@ gf100_gr_init_gpc_mmu(struct gf100_gr *gr)
 }
 
 void
+gf100_gr_init_zcull(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	const u32 magicgpc918 = DIV_ROUND_UP(0x00800000, gr->tpc_total);
+	u32 data[TPC_MAX / 8] = {};
+	u8  tpcnr[GPC_MAX];
+	int gpc, tpc;
+	int i;
+
+	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
+	for (i = 0, gpc = -1; i < gr->tpc_total; i++) {
+		do {
+			gpc = (gpc + 1) % gr->gpc_nr;
+		} while (!tpcnr[gpc]);
+		tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--;
+
+		data[i / 8] |= tpc << ((i % 8) * 4);
+	}
+
+	nvkm_wr32(device, GPC_BCAST(0x0980), data[0]);
+	nvkm_wr32(device, GPC_BCAST(0x0984), data[1]);
+	nvkm_wr32(device, GPC_BCAST(0x0988), data[2]);
+	nvkm_wr32(device, GPC_BCAST(0x098c), data[3]);
+
+	nvkm_wr32(device, GPC_BCAST(0x1bd4), magicgpc918);
+}
+
+void
 gf100_gr_init_vsc_stream_master(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -1941,11 +1969,7 @@ int
 gf100_gr_init(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
-	const u32 magicgpc918 = DIV_ROUND_UP(0x00800000, gr->tpc_total);
-	u32 data[TPC_MAX / 8] = {};
-	u8  tpcnr[GPC_MAX];
 	int gpc, tpc, rop;
-	int i;
 
 	gr->func->init_gpc_mmu(gr);
 
@@ -1964,34 +1988,7 @@ gf100_gr_init(struct gf100_gr *gr)
 		gr->func->init_bios(gr);
 
 	gr->func->init_vsc_stream_master(gr);
-
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-	for (i = 0, gpc = -1; i < gr->tpc_total; i++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while (!tpcnr[gpc]);
-		tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--;
-
-		data[i / 8] |= tpc << ((i % 8) * 4);
-	}
-
-	nvkm_wr32(device, GPC_BCAST(0x0980), data[0]);
-	nvkm_wr32(device, GPC_BCAST(0x0984), data[1]);
-	nvkm_wr32(device, GPC_BCAST(0x0988), data[2]);
-	nvkm_wr32(device, GPC_BCAST(0x098c), data[3]);
-
-	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0914),
-			  gr->screen_tile_row_offset << 8 | gr->tpc_nr[gpc]);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0910), 0x00040000 |
-							 gr->tpc_total);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0918), magicgpc918);
-	}
-
-	if (device->chipset != 0xd7)
-		nvkm_wr32(device, GPC_BCAST(0x1bd4), magicgpc918);
-	else
-		nvkm_wr32(device, GPC_BCAST(0x3fd4), magicgpc918);
+	gr->func->init_zcull(gr);
 
 	nvkm_wr32(device, GPC_BCAST(0x08ac), nvkm_rd32(device, 0x100800));
 
@@ -2076,6 +2073,7 @@ gf100_gr = {
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
+	.init_zcull = gf100_gr_init_zcull,
 	.mmio = gf100_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index 45792a3..b322dc4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -125,6 +125,7 @@ struct gf100_gr_func {
 	void (*init_r405a14)(struct gf100_gr *);
 	void (*init_bios)(struct gf100_gr *);
 	void (*init_vsc_stream_master)(struct gf100_gr *);
+	void (*init_zcull)(struct gf100_gr *);
 	void (*init_rop_active_fbps)(struct gf100_gr *);
 	void (*init_ppc_exceptions)(struct gf100_gr *);
 	void (*init_swdx_pes_mask)(struct gf100_gr *);
@@ -147,6 +148,9 @@ struct gf100_gr_func {
 int gf100_gr_rops(struct gf100_gr *);
 int gf100_gr_init(struct gf100_gr *);
 void gf100_gr_init_vsc_stream_master(struct gf100_gr *);
+void gf100_gr_init_zcull(struct gf100_gr *);
+
+void gf117_gr_init_zcull(struct gf100_gr *);
 
 int gk104_gr_init(struct gf100_gr *);
 void gk104_gr_init_vsc_stream_master(struct gf100_gr *);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
index 61b27b1..23c2613 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
@@ -117,6 +117,7 @@ gf104_gr = {
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
+	.init_zcull = gf100_gr_init_zcull,
 	.mmio = gf104_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
index 992727f..a79a786 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
@@ -115,6 +115,7 @@ gf108_gr = {
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_r405a14 = gf108_gr_init_r405a14,
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
+	.init_zcull = gf100_gr_init_zcull,
 	.mmio = gf108_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
index 07a57cc..fd8e137 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
@@ -89,6 +89,7 @@ gf110_gr = {
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
+	.init_zcull = gf100_gr_init_zcull,
 	.mmio = gf110_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
index 29ab01b..01e213e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
@@ -120,11 +120,40 @@ gf117_gr_gpccs_ucode = {
 	.data.size = sizeof(gf117_grgpc_data),
 };
 
+void
+gf117_gr_init_zcull(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	const u32 magicgpc918 = DIV_ROUND_UP(0x00800000, gr->tpc_total);
+	u32 data[TPC_MAX / 8] = {};
+	u8  tpcnr[GPC_MAX];
+	int gpc, tpc;
+	int i;
+
+	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
+	for (i = 0, gpc = -1; i < gr->tpc_total; i++) {
+		do {
+			gpc = (gpc + 1) % gr->gpc_nr;
+		} while (!tpcnr[gpc]);
+		tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--;
+
+		data[i / 8] |= tpc << ((i % 8) * 4);
+	}
+
+	nvkm_wr32(device, GPC_BCAST(0x0980), data[0]);
+	nvkm_wr32(device, GPC_BCAST(0x0984), data[1]);
+	nvkm_wr32(device, GPC_BCAST(0x0988), data[2]);
+	nvkm_wr32(device, GPC_BCAST(0x098c), data[3]);
+
+	nvkm_wr32(device, GPC_BCAST(0x3fd4), magicgpc918);
+}
+
 static const struct gf100_gr_func
 gf117_gr = {
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
 	.mmio = gf117_gr_pack_mmio,
 	.fecs.ucode = &gf117_gr_fecs_ucode,
 	.gpccs.ucode = &gf117_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
index d3b06dd7..269922e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
@@ -180,6 +180,7 @@ gf119_gr = {
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
+	.init_zcull = gf100_gr_init_zcull,
 	.mmio = gf119_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
index 835f498..d1b7fb8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
@@ -415,11 +415,7 @@ int
 gk104_gr_init(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
-	const u32 magicgpc918 = DIV_ROUND_UP(0x00800000, gr->tpc_total);
-	u32 data[TPC_MAX / 8] = {};
-	u8  tpcnr[GPC_MAX];
 	int gpc, tpc, rop;
-	int i;
 
 	gr->func->init_gpc_mmu(gr);
 
@@ -429,32 +425,8 @@ gk104_gr_init(struct gf100_gr *gr)
 					gr->func->clkgate_pack);
 
 	gr->func->init_vsc_stream_master(gr);
+	gr->func->init_zcull(gr);
 
-	memset(data, 0x00, sizeof(data));
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-	for (i = 0, gpc = -1; i < gr->tpc_total; i++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while (!tpcnr[gpc]);
-		tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--;
-
-		data[i / 8] |= tpc << ((i % 8) * 4);
-	}
-
-	nvkm_wr32(device, GPC_BCAST(0x0980), data[0]);
-	nvkm_wr32(device, GPC_BCAST(0x0984), data[1]);
-	nvkm_wr32(device, GPC_BCAST(0x0988), data[2]);
-	nvkm_wr32(device, GPC_BCAST(0x098c), data[3]);
-
-	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0914),
-			  gr->screen_tile_row_offset << 8 | gr->tpc_nr[gpc]);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0910), 0x00040000 |
-							 gr->tpc_total);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0918), magicgpc918);
-	}
-
-	nvkm_wr32(device, GPC_BCAST(0x3fd4), magicgpc918);
 	nvkm_wr32(device, GPC_BCAST(0x08ac), nvkm_rd32(device, 0x100800));
 
 	gr->func->init_rop_active_fbps(gr);
@@ -544,6 +516,7 @@ gk104_gr = {
 	.init = gk104_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk104_gr_pack_mmio,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
index eef7476e..0eba041 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
@@ -339,6 +339,7 @@ gk110_gr = {
 	.init = gk104_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk110_gr_pack_mmio,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
index ed4361a..84a1664 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
@@ -105,6 +105,7 @@ gk110b_gr = {
 	.init = gk104_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk110b_gr_pack_mmio,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
index 4c8beb9..701199a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
@@ -164,6 +164,7 @@ gk208_gr = {
 	.init = gk104_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk208_gr_pack_mmio,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c
index de8b806..a806643 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c
@@ -219,11 +219,7 @@ int
 gk20a_gr_init(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
-	const u32 magicgpc918 = DIV_ROUND_UP(0x00800000, gr->tpc_total);
-	u32 data[TPC_MAX / 8] = {};
-	u8  tpcnr[GPC_MAX];
-	int gpc, tpc;
-	int ret, i;
+	int ret;
 
 	/* Clear SCC RAM */
 	nvkm_wr32(device, 0x40802c, 0x1);
@@ -246,31 +242,7 @@ gk20a_gr_init(struct gf100_gr *gr)
 	nvkm_mask(device, 0x503018, 0x1, 0x1);
 
 	/* Zcull init */
-	memset(data, 0x00, sizeof(data));
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-	for (i = 0, gpc = -1; i < gr->tpc_total; i++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while (!tpcnr[gpc]);
-		tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--;
-
-		data[i / 8] |= tpc << ((i % 8) * 4);
-	}
-
-	nvkm_wr32(device, GPC_BCAST(0x0980), data[0]);
-	nvkm_wr32(device, GPC_BCAST(0x0984), data[1]);
-	nvkm_wr32(device, GPC_BCAST(0x0988), data[2]);
-	nvkm_wr32(device, GPC_BCAST(0x098c), data[3]);
-
-	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0914),
-			  gr->screen_tile_row_offset << 8 | gr->tpc_nr[gpc]);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0910), 0x00040000 |
-			  gr->tpc_total);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0918), magicgpc918);
-	}
-
-	nvkm_wr32(device, GPC_BCAST(0x3fd4), magicgpc918);
+	gr->func->init_zcull(gr);
 
 	gr->func->init_rop_active_fbps(gr);
 
@@ -311,6 +283,7 @@ gk20a_gr_init(struct gf100_gr *gr)
 static const struct gf100_gr_func
 gk20a_gr = {
 	.init = gk20a_gr_init,
+	.init_zcull = gf117_gr_init_zcull,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.set_hww_esr_report_mask = gk20a_gr_set_hww_esr_report_mask,
 	.rops = gf100_gr_rops,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
index 58c03d6..a5f5e05 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
@@ -325,11 +325,7 @@ static int
 gm107_gr_init(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
-	const u32 magicgpc918 = DIV_ROUND_UP(0x00800000, gr->tpc_total);
-	u32 data[TPC_MAX / 8] = {};
-	u8  tpcnr[GPC_MAX];
 	int gpc, tpc, rop;
-	int i;
 
 	gr->func->init_gpc_mmu(gr);
 
@@ -338,32 +334,8 @@ gm107_gr_init(struct gf100_gr *gr)
 	gr->func->init_bios(gr);
 
 	gr->func->init_vsc_stream_master(gr);
+	gr->func->init_zcull(gr);
 
-	memset(data, 0x00, sizeof(data));
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-	for (i = 0, gpc = -1; i < gr->tpc_total; i++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while (!tpcnr[gpc]);
-		tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--;
-
-		data[i / 8] |= tpc << ((i % 8) * 4);
-	}
-
-	nvkm_wr32(device, GPC_BCAST(0x0980), data[0]);
-	nvkm_wr32(device, GPC_BCAST(0x0984), data[1]);
-	nvkm_wr32(device, GPC_BCAST(0x0988), data[2]);
-	nvkm_wr32(device, GPC_BCAST(0x098c), data[3]);
-
-	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0914),
-			  gr->screen_tile_row_offset << 8 | gr->tpc_nr[gpc]);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0910), 0x00040000 |
-							 gr->tpc_total);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0918), magicgpc918);
-	}
-
-	nvkm_wr32(device, GPC_BCAST(0x3fd4), magicgpc918);
 	nvkm_wr32(device, GPC_BCAST(0x08ac), nvkm_rd32(device, 0x100800));
 
 	gr->func->init_rop_active_fbps(gr);
@@ -453,6 +425,7 @@ gm107_gr = {
 	.init_gpc_mmu = gm107_gr_init_gpc_mmu,
 	.init_bios = gm107_gr_init_bios,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gm107_gr_pack_mmio,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
index 8583aad..2ceef17 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
@@ -65,11 +65,7 @@ static int
 gm200_gr_init(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
-	const u32 magicgpc918 = DIV_ROUND_UP(0x00800000, gr->tpc_total);
-	u32 data[TPC_MAX / 8] = {};
-	u8  tpcnr[GPC_MAX];
 	int gpc, tpc, rop;
-	int i;
 
 	gr->func->init_gpc_mmu(gr);
 
@@ -78,32 +74,8 @@ gm200_gr_init(struct gf100_gr *gr)
 	gr->func->init_bios(gr);
 
 	gr->func->init_vsc_stream_master(gr);
+	gr->func->init_zcull(gr);
 
-	memset(data, 0x00, sizeof(data));
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-	for (i = 0, gpc = -1; i < gr->tpc_total; i++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while (!tpcnr[gpc]);
-		tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--;
-
-		data[i / 8] |= tpc << ((i % 8) * 4);
-	}
-
-	nvkm_wr32(device, GPC_BCAST(0x0980), data[0]);
-	nvkm_wr32(device, GPC_BCAST(0x0984), data[1]);
-	nvkm_wr32(device, GPC_BCAST(0x0988), data[2]);
-	nvkm_wr32(device, GPC_BCAST(0x098c), data[3]);
-
-	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0914),
-			  gr->screen_tile_row_offset << 8 | gr->tpc_nr[gpc]);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0910), 0x00040000 |
-							 gr->tpc_total);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0918), magicgpc918);
-	}
-
-	nvkm_wr32(device, GPC_BCAST(0x3fd4), magicgpc918);
 	nvkm_wr32(device, GPC_BCAST(0x08ac), nvkm_rd32(device, 0x100800));
 	nvkm_wr32(device, GPC_BCAST(0x033c), nvkm_rd32(device, 0x100804));
 
@@ -212,6 +184,7 @@ gm200_gr = {
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_bios = gm107_gr_init_bios,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
 	.init_rop_active_fbps = gm200_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c
index 69479af..fcf86d5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c
@@ -65,6 +65,7 @@ gm20b_gr_set_hww_esr_report_mask(struct gf100_gr *gr)
 static const struct gf100_gr_func
 gm20b_gr = {
 	.init = gk20a_gr_init,
+	.init_zcull = gf117_gr_init_zcull,
 	.init_gpc_mmu = gm20b_gr_init_gpc_mmu,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.set_hww_esr_report_mask = gm20b_gr_set_hww_esr_report_mask,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
index edaaebb..7d3317e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
@@ -53,43 +53,15 @@ int
 gp100_gr_init(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
-	const u32 magicgpc918 = DIV_ROUND_UP(0x00800000, gr->tpc_total);
-	u32 data[TPC_MAX / 8] = {};
-	u8  tpcnr[GPC_MAX];
 	int gpc, tpc, rop;
-	int i;
 
 	gr->func->init_gpc_mmu(gr);
 
 	gf100_gr_mmio(gr, gr->fuc_sw_nonctx);
 
 	gr->func->init_vsc_stream_master(gr);
+	gr->func->init_zcull(gr);
 
-	memset(data, 0x00, sizeof(data));
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-	for (i = 0, gpc = -1; i < gr->tpc_total; i++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while (!tpcnr[gpc]);
-		tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--;
-
-		data[i / 8] |= tpc << ((i % 8) * 4);
-	}
-
-	nvkm_wr32(device, GPC_BCAST(0x0980), data[0]);
-	nvkm_wr32(device, GPC_BCAST(0x0984), data[1]);
-	nvkm_wr32(device, GPC_BCAST(0x0988), data[2]);
-	nvkm_wr32(device, GPC_BCAST(0x098c), data[3]);
-
-	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0914),
-			  gr->screen_tile_row_offset << 8 | gr->tpc_nr[gpc]);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0910), 0x00040000 |
-							 gr->tpc_total);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0918), magicgpc918);
-	}
-
-	nvkm_wr32(device, GPC_BCAST(0x3fd4), magicgpc918);
 	gr->func->init_num_active_ltcs(gr);
 
 	gr->func->init_rop_active_fbps(gr);
@@ -161,6 +133,7 @@ gp100_gr = {
 	.init = gp100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_num_active_ltcs = gp100_gr_init_num_active_ltcs,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
index 821a6c4..0d9bf6c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
@@ -45,6 +45,7 @@ gp102_gr = {
 	.init = gp100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_swdx_pes_mask = gp102_gr_init_swdx_pes_mask,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
index e2e1297..67d567e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
@@ -31,6 +31,7 @@ gp107_gr = {
 	.init = gp100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_swdx_pes_mask = gp102_gr_init_swdx_pes_mask,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
index e90cc3c..af06b10 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
@@ -38,6 +38,7 @@ gp10b_gr = {
 	.init = gp100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_num_active_ltcs = gp10b_gr_init_num_active_ltcs,
-- 
cgit v1.1


From bfd27f39b5419724883bbd04910c4c35e57b6154 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise init_num_active_ltcs

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c  | 11 +++++++++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h  |  6 +++---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c  |  4 ++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c  |  4 ++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c  | 13 ++++++++++---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c  | 11 +----------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c  |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c  |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c  | 10 +---------
 17 files changed, 38 insertions(+), 33 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 56d9ead..d33e531 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -1931,6 +1931,13 @@ gf100_gr_init_gpc_mmu(struct gf100_gr *gr)
 }
 
 void
+gf100_gr_init_num_active_ltcs(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, GPC_BCAST(0x08ac), nvkm_rd32(device, 0x100800));
+}
+
+void
 gf100_gr_init_zcull(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -1989,8 +1996,7 @@ gf100_gr_init(struct gf100_gr *gr)
 
 	gr->func->init_vsc_stream_master(gr);
 	gr->func->init_zcull(gr);
-
-	nvkm_wr32(device, GPC_BCAST(0x08ac), nvkm_rd32(device, 0x100800));
+	gr->func->init_num_active_ltcs(gr);
 
 	nvkm_wr32(device, 0x400500, 0x00010001);
 
@@ -2074,6 +2080,7 @@ gf100_gr = {
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
 	.init_zcull = gf100_gr_init_zcull,
+	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.mmio = gf100_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index b322dc4..7f6aed8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -126,10 +126,10 @@ struct gf100_gr_func {
 	void (*init_bios)(struct gf100_gr *);
 	void (*init_vsc_stream_master)(struct gf100_gr *);
 	void (*init_zcull)(struct gf100_gr *);
+	void (*init_num_active_ltcs)(struct gf100_gr *);
 	void (*init_rop_active_fbps)(struct gf100_gr *);
 	void (*init_ppc_exceptions)(struct gf100_gr *);
 	void (*init_swdx_pes_mask)(struct gf100_gr *);
-	void (*init_num_active_ltcs)(struct gf100_gr *);
 	void (*set_hww_esr_report_mask)(struct gf100_gr *);
 	const struct gf100_gr_pack *mmio;
 	struct {
@@ -149,6 +149,7 @@ int gf100_gr_rops(struct gf100_gr *);
 int gf100_gr_init(struct gf100_gr *);
 void gf100_gr_init_vsc_stream_master(struct gf100_gr *);
 void gf100_gr_init_zcull(struct gf100_gr *);
+void gf100_gr_init_num_active_ltcs(struct gf100_gr *);
 
 void gf117_gr_init_zcull(struct gf100_gr *);
 
@@ -160,6 +161,7 @@ void gk104_gr_init_ppc_exceptions(struct gf100_gr *);
 int gk20a_gr_init(struct gf100_gr *);
 
 int gm200_gr_rops(struct gf100_gr *);
+void gm200_gr_init_num_active_ltcs(struct gf100_gr *);
 
 int gp100_gr_init(struct gf100_gr *);
 void gp100_gr_init_rop_active_fbps(struct gf100_gr *);
@@ -315,7 +317,5 @@ void gm107_gr_init_bios(struct gf100_gr *);
 
 void gm200_gr_init_gpc_mmu(struct gf100_gr *);
 
-void gp100_gr_init_num_active_ltcs(struct gf100_gr *gr);
-
 void gp102_gr_init_swdx_pes_mask(struct gf100_gr *);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
index 23c2613..b6d67f5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
@@ -118,6 +118,7 @@ gf104_gr = {
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
 	.init_zcull = gf100_gr_init_zcull,
+	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.mmio = gf104_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
index a79a786..9b5796b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
@@ -116,6 +116,7 @@ gf108_gr = {
 	.init_r405a14 = gf108_gr_init_r405a14,
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
 	.init_zcull = gf100_gr_init_zcull,
+	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.mmio = gf108_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
index fd8e137..7e346a0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
@@ -90,6 +90,7 @@ gf110_gr = {
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
 	.init_zcull = gf100_gr_init_zcull,
+	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.mmio = gf110_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
index 01e213e..5c68c5f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
@@ -154,6 +154,7 @@ gf117_gr = {
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
 	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.mmio = gf117_gr_pack_mmio,
 	.fecs.ucode = &gf117_gr_fecs_ucode,
 	.gpccs.ucode = &gf117_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
index 269922e..b642bd2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
@@ -181,6 +181,7 @@ gf119_gr = {
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
 	.init_zcull = gf100_gr_init_zcull,
+	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.mmio = gf119_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
index d1b7fb8..3f19de2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
@@ -426,8 +426,7 @@ gk104_gr_init(struct gf100_gr *gr)
 
 	gr->func->init_vsc_stream_master(gr);
 	gr->func->init_zcull(gr);
-
-	nvkm_wr32(device, GPC_BCAST(0x08ac), nvkm_rd32(device, 0x100800));
+	gr->func->init_num_active_ltcs(gr);
 
 	gr->func->init_rop_active_fbps(gr);
 
@@ -517,6 +516,7 @@ gk104_gr = {
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk104_gr_pack_mmio,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
index 0eba041..9a8428c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
@@ -340,6 +340,7 @@ gk110_gr = {
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk110_gr_pack_mmio,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
index 84a1664..08f3f8c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
@@ -106,6 +106,7 @@ gk110b_gr = {
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk110b_gr_pack_mmio,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
index 701199a..49ee549 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
@@ -165,6 +165,7 @@ gk208_gr = {
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk208_gr_pack_mmio,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
index a5f5e05..bbb0a28 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
@@ -335,8 +335,7 @@ gm107_gr_init(struct gf100_gr *gr)
 
 	gr->func->init_vsc_stream_master(gr);
 	gr->func->init_zcull(gr);
-
-	nvkm_wr32(device, GPC_BCAST(0x08ac), nvkm_rd32(device, 0x100800));
+	gr->func->init_num_active_ltcs(gr);
 
 	gr->func->init_rop_active_fbps(gr);
 
@@ -426,6 +425,7 @@ gm107_gr = {
 	.init_bios = gm107_gr_init_bios,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gm107_gr_pack_mmio,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
index 2ceef17..5007435 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
@@ -39,6 +39,14 @@ gm200_gr_rops(struct gf100_gr *gr)
 }
 
 void
+gm200_gr_init_num_active_ltcs(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, GPC_BCAST(0x08ac), nvkm_rd32(device, 0x100800));
+	nvkm_wr32(device, GPC_BCAST(0x033c), nvkm_rd32(device, 0x100804));
+}
+
+void
 gm200_gr_init_gpc_mmu(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -75,9 +83,7 @@ gm200_gr_init(struct gf100_gr *gr)
 
 	gr->func->init_vsc_stream_master(gr);
 	gr->func->init_zcull(gr);
-
-	nvkm_wr32(device, GPC_BCAST(0x08ac), nvkm_rd32(device, 0x100800));
-	nvkm_wr32(device, GPC_BCAST(0x033c), nvkm_rd32(device, 0x100804));
+	gr->func->init_num_active_ltcs(gr);
 
 	gr->func->init_rop_active_fbps(gr);
 
@@ -185,6 +191,7 @@ gm200_gr = {
 	.init_bios = gm107_gr_init_bios,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gm200_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gm200_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
index 7d3317e..1e67959 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
@@ -40,15 +40,6 @@ gp100_gr_init_rop_active_fbps(struct gf100_gr *gr)
 	nvkm_mask(device, 0x408958, 0x0000000f, fbp_count); /* crop */
 }
 
-void
-gp100_gr_init_num_active_ltcs(struct gf100_gr *gr)
-{
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-
-	nvkm_wr32(device, GPC_BCAST(0x08ac), nvkm_rd32(device, 0x100800));
-	nvkm_wr32(device, GPC_BCAST(0x033c), nvkm_rd32(device, 0x100804));
-}
-
 int
 gp100_gr_init(struct gf100_gr *gr)
 {
@@ -134,9 +125,9 @@ gp100_gr = {
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gm200_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
-	.init_num_active_ltcs = gp100_gr_init_num_active_ltcs,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 2,
 	.grctx = &gp100_grctx,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
index 0d9bf6c..17bd872 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
@@ -46,10 +46,10 @@ gp102_gr = {
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gm200_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_swdx_pes_mask = gp102_gr_init_swdx_pes_mask,
-	.init_num_active_ltcs = gp100_gr_init_num_active_ltcs,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 3,
 	.grctx = &gp102_grctx,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
index 67d567e..3fba3af 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
@@ -32,10 +32,10 @@ gp107_gr = {
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gm200_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_swdx_pes_mask = gp102_gr_init_swdx_pes_mask,
-	.init_num_active_ltcs = gp100_gr_init_num_active_ltcs,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 1,
 	.grctx = &gp107_grctx,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
index af06b10..97f8667 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
@@ -25,23 +25,15 @@
 
 #include <nvif/class.h>
 
-static void
-gp10b_gr_init_num_active_ltcs(struct gf100_gr *gr)
-{
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-
-	nvkm_wr32(device, GPC_BCAST(0x08ac), nvkm_rd32(device, 0x100800));
-}
-
 static const struct gf100_gr_func
 gp10b_gr = {
 	.init = gp100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
-	.init_num_active_ltcs = gp10b_gr_init_num_active_ltcs,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 1,
 	.grctx = &gp102_grctx,
-- 
cgit v1.1


From 429412e231a27d48cb492dc1c647e857677240b3 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise init_rop_active_fbps

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index d33e531..6912eaa 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -1997,6 +1997,8 @@ gf100_gr_init(struct gf100_gr *gr)
 	gr->func->init_vsc_stream_master(gr);
 	gr->func->init_zcull(gr);
 	gr->func->init_num_active_ltcs(gr);
+	if (gr->func->init_rop_active_fbps)
+		gr->func->init_rop_active_fbps(gr);
 
 	nvkm_wr32(device, 0x400500, 0x00010001);
 
-- 
cgit v1.1


From 0f78acc86bbfc60cdaffd4eb03e4a35b28397cbb Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: implement another chunk of bios-provided init

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c | 28 ++++++++++++++++++++++++++
 3 files changed, 31 insertions(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 6912eaa..01b903a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -1999,6 +1999,8 @@ gf100_gr_init(struct gf100_gr *gr)
 	gr->func->init_num_active_ltcs(gr);
 	if (gr->func->init_rop_active_fbps)
 		gr->func->init_rop_active_fbps(gr);
+	if (gr->func->init_bios_2)
+		gr->func->init_bios_2(gr);
 
 	nvkm_wr32(device, 0x400500, 0x00010001);
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index 7f6aed8..11cf2df 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -128,6 +128,7 @@ struct gf100_gr_func {
 	void (*init_zcull)(struct gf100_gr *);
 	void (*init_num_active_ltcs)(struct gf100_gr *);
 	void (*init_rop_active_fbps)(struct gf100_gr *);
+	void (*init_bios_2)(struct gf100_gr *);
 	void (*init_ppc_exceptions)(struct gf100_gr *);
 	void (*init_swdx_pes_mask)(struct gf100_gr *);
 	void (*set_hww_esr_report_mask)(struct gf100_gr *);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
index bbb0a28..c653df3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
@@ -25,6 +25,8 @@
 #include "ctxgf100.h"
 
 #include <subdev/bios.h>
+#include <subdev/bios/bit.h>
+#include <subdev/bios/init.h>
 #include <subdev/bios/P0260.h>
 #include <subdev/fb.h>
 
@@ -279,6 +281,31 @@ gm107_gr_pack_mmio[] = {
  * PGRAPH engine/subdev functions
  ******************************************************************************/
 
+static void
+gm107_gr_init_bios_2(struct gf100_gr *gr)
+{
+	struct nvkm_subdev *subdev = &gr->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	struct nvkm_bios *bios = device->bios;
+	struct bit_entry bit_P;
+	if (!bit_entry(bios, 'P', &bit_P) &&
+	    bit_P.version == 2 && bit_P.length >= 0x2c) {
+		u32 data = nvbios_rd32(bios, bit_P.offset + 0x28);
+		if (data) {
+			u8 ver = nvbios_rd08(bios, data + 0x00);
+			u8 hdr = nvbios_rd08(bios, data + 0x01);
+			if (ver == 0x20 && hdr >= 8) {
+				data = nvbios_rd32(bios, data + 0x04);
+				if (data) {
+					u32 save = nvkm_rd32(device, 0x619444);
+					nvbios_init(subdev, data);
+					nvkm_wr32(device, 0x619444, save);
+				}
+			}
+		}
+	}
+}
+
 void
 gm107_gr_init_bios(struct gf100_gr *gr)
 {
@@ -427,6 +454,7 @@ gm107_gr = {
 	.init_zcull = gf117_gr_init_zcull,
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
+	.init_bios_2 = gm107_gr_init_bios_2,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gm107_gr_pack_mmio,
 	.fecs.ucode = &gm107_gr_fecs_ucode,
-- 
cgit v1.1


From dff30dbd1d9336687ae1aa0b13e326c44f879c4e Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise init_swdx_pes_mask

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c | 2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h | 6 +++---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c | 2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c | 2 +-
 4 files changed, 7 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 01b903a..7ad6ea0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -2001,6 +2001,8 @@ gf100_gr_init(struct gf100_gr *gr)
 		gr->func->init_rop_active_fbps(gr);
 	if (gr->func->init_bios_2)
 		gr->func->init_bios_2(gr);
+	if (gr->func->init_swdx_pes_mask)
+		gr->func->init_swdx_pes_mask(gr);
 
 	nvkm_wr32(device, 0x400500, 0x00010001);
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index 11cf2df..858024b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -129,8 +129,8 @@ struct gf100_gr_func {
 	void (*init_num_active_ltcs)(struct gf100_gr *);
 	void (*init_rop_active_fbps)(struct gf100_gr *);
 	void (*init_bios_2)(struct gf100_gr *);
-	void (*init_ppc_exceptions)(struct gf100_gr *);
 	void (*init_swdx_pes_mask)(struct gf100_gr *);
+	void (*init_ppc_exceptions)(struct gf100_gr *);
 	void (*set_hww_esr_report_mask)(struct gf100_gr *);
 	const struct gf100_gr_pack *mmio;
 	struct {
@@ -167,6 +167,8 @@ void gm200_gr_init_num_active_ltcs(struct gf100_gr *);
 int gp100_gr_init(struct gf100_gr *);
 void gp100_gr_init_rop_active_fbps(struct gf100_gr *);
 
+void gp102_gr_init_swdx_pes_mask(struct gf100_gr *);
+
 #define gf100_gr_chan(p) container_of((p), struct gf100_gr_chan, object)
 #include <core/object.h>
 
@@ -317,6 +319,4 @@ extern const struct gf100_gr_init gm107_gr_init_cbm_0[];
 void gm107_gr_init_bios(struct gf100_gr *);
 
 void gm200_gr_init_gpc_mmu(struct gf100_gr *);
-
-void gp102_gr_init_swdx_pes_mask(struct gf100_gr *);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
index 17bd872..860a789 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
@@ -48,8 +48,8 @@ gp102_gr = {
 	.init_zcull = gf117_gr_init_zcull,
 	.init_num_active_ltcs = gm200_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
-	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_swdx_pes_mask = gp102_gr_init_swdx_pes_mask,
+	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 3,
 	.grctx = &gp102_grctx,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
index 3fba3af..03f2264 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
@@ -34,8 +34,8 @@ gp107_gr = {
 	.init_zcull = gf117_gr_init_zcull,
 	.init_num_active_ltcs = gm200_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
-	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_swdx_pes_mask = gp102_gr_init_swdx_pes_mask,
+	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 1,
 	.grctx = &gp107_grctx,
-- 
cgit v1.1


From 7c76ebb65a9c8d29780e7324c8ae067c5cd980ca Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100: write 0x400124 during init

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c | 1 +
 2 files changed, 2 insertions(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 7ad6ea0..0c53d45 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -2008,6 +2008,7 @@ gf100_gr_init(struct gf100_gr *gr)
 
 	nvkm_wr32(device, 0x400100, 0xffffffff);
 	nvkm_wr32(device, 0x40013c, 0xffffffff);
+	nvkm_wr32(device, 0x400124, 0x00000002);
 
 	nvkm_wr32(device, 0x409c24, 0x000f0000);
 	nvkm_wr32(device, 0x404000, 0xc0000000);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
index 3f19de2..703aed4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
@@ -434,6 +434,7 @@ gk104_gr_init(struct gf100_gr *gr)
 
 	nvkm_wr32(device, 0x400100, 0xffffffff);
 	nvkm_wr32(device, 0x40013c, 0xffffffff);
+	nvkm_wr32(device, 0x400124, 0x00000002);
 
 	nvkm_wr32(device, 0x409ffc, 0x00000000);
 	nvkm_wr32(device, 0x409c14, 0x00003e3e);
-- 
cgit v1.1


From 2585a1b1312e96c6a28f3008029408b5feca3ff4 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise init_fecs_exceptions + apply fixes
 from traces

The value for GF100 has changed here, but it matches RM now.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c  | 11 ++++++++++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h  |  3 +++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c  | 14 +++++++++++---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c  |  3 ++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c  |  3 ++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c  |  9 ++++++++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c  |  1 +
 17 files changed, 47 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 0c53d45..0493483 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -1915,6 +1915,13 @@ gf100_gr_new_(const struct gf100_gr_func *func, struct nvkm_device *device,
 }
 
 void
+gf100_gr_init_fecs_exceptions(struct gf100_gr *gr)
+{
+	const u32 data = gr->firmware ? 0x000e0000 : 0x000e0001;
+	nvkm_wr32(gr->base.engine.subdev.device, 0x409c24, data);
+}
+
+void
 gf100_gr_init_gpc_mmu(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -2010,7 +2017,8 @@ gf100_gr_init(struct gf100_gr *gr)
 	nvkm_wr32(device, 0x40013c, 0xffffffff);
 	nvkm_wr32(device, 0x400124, 0x00000002);
 
-	nvkm_wr32(device, 0x409c24, 0x000f0000);
+	gr->func->init_fecs_exceptions(gr);
+
 	nvkm_wr32(device, 0x404000, 0xc0000000);
 	nvkm_wr32(device, 0x404600, 0xc0000000);
 	nvkm_wr32(device, 0x408030, 0xc0000000);
@@ -2088,6 +2096,7 @@ gf100_gr = {
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
 	.init_zcull = gf100_gr_init_zcull,
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
+	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.mmio = gf100_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index 858024b..387938f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -130,6 +130,7 @@ struct gf100_gr_func {
 	void (*init_rop_active_fbps)(struct gf100_gr *);
 	void (*init_bios_2)(struct gf100_gr *);
 	void (*init_swdx_pes_mask)(struct gf100_gr *);
+	void (*init_fecs_exceptions)(struct gf100_gr *);
 	void (*init_ppc_exceptions)(struct gf100_gr *);
 	void (*set_hww_esr_report_mask)(struct gf100_gr *);
 	const struct gf100_gr_pack *mmio;
@@ -151,6 +152,7 @@ int gf100_gr_init(struct gf100_gr *);
 void gf100_gr_init_vsc_stream_master(struct gf100_gr *);
 void gf100_gr_init_zcull(struct gf100_gr *);
 void gf100_gr_init_num_active_ltcs(struct gf100_gr *);
+void gf100_gr_init_fecs_exceptions(struct gf100_gr *);
 
 void gf117_gr_init_zcull(struct gf100_gr *);
 
@@ -166,6 +168,7 @@ void gm200_gr_init_num_active_ltcs(struct gf100_gr *);
 
 int gp100_gr_init(struct gf100_gr *);
 void gp100_gr_init_rop_active_fbps(struct gf100_gr *);
+void gp100_gr_init_fecs_exceptions(struct gf100_gr *);
 
 void gp102_gr_init_swdx_pes_mask(struct gf100_gr *);
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
index b6d67f5..f76995b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
@@ -119,6 +119,7 @@ gf104_gr = {
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
 	.init_zcull = gf100_gr_init_zcull,
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
+	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.mmio = gf104_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
index 9b5796b..ada2697 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
@@ -117,6 +117,7 @@ gf108_gr = {
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
 	.init_zcull = gf100_gr_init_zcull,
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
+	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.mmio = gf108_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
index 7e346a0..80ced8f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
@@ -91,6 +91,7 @@ gf110_gr = {
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
 	.init_zcull = gf100_gr_init_zcull,
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
+	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.mmio = gf110_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
index 5c68c5f..37ca121 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
@@ -155,6 +155,7 @@ gf117_gr = {
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
 	.init_zcull = gf117_gr_init_zcull,
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
+	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.mmio = gf117_gr_pack_mmio,
 	.fecs.ucode = &gf117_gr_fecs_ucode,
 	.gpccs.ucode = &gf117_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
index b642bd2..ddf05c5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
@@ -182,6 +182,7 @@ gf119_gr = {
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
 	.init_zcull = gf100_gr_init_zcull,
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
+	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.mmio = gf119_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
index 703aed4..746ad3d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
@@ -380,6 +380,15 @@ gk104_clkgate_pack[] = {
  * PGRAPH engine/subdev functions
  ******************************************************************************/
 
+static void
+gk104_gr_init_fecs_exceptions(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, 0x409ffc, 0x00000000);
+	nvkm_wr32(device, 0x409c14, 0x00003e3e);
+	nvkm_wr32(device, 0x409c24, 0x000f0001);
+}
+
 void
 gk104_gr_init_rop_active_fbps(struct gf100_gr *gr)
 {
@@ -436,9 +445,7 @@ gk104_gr_init(struct gf100_gr *gr)
 	nvkm_wr32(device, 0x40013c, 0xffffffff);
 	nvkm_wr32(device, 0x400124, 0x00000002);
 
-	nvkm_wr32(device, 0x409ffc, 0x00000000);
-	nvkm_wr32(device, 0x409c14, 0x00003e3e);
-	nvkm_wr32(device, 0x409c24, 0x000f0001);
+	gr->func->init_fecs_exceptions(gr);
 	nvkm_wr32(device, 0x404000, 0xc0000000);
 	nvkm_wr32(device, 0x404600, 0xc0000000);
 	nvkm_wr32(device, 0x408030, 0xc0000000);
@@ -519,6 +526,7 @@ gk104_gr = {
 	.init_zcull = gf117_gr_init_zcull,
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
+	.init_fecs_exceptions = gk104_gr_init_fecs_exceptions,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk104_gr_pack_mmio,
 	.fecs.ucode = &gk104_gr_fecs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
index 9a8428c..8a6340d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
@@ -342,6 +342,7 @@ gk110_gr = {
 	.init_zcull = gf117_gr_init_zcull,
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
+	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk110_gr_pack_mmio,
 	.fecs.ucode = &gk110_gr_fecs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
index 08f3f8c..b50e681 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
@@ -108,6 +108,7 @@ gk110b_gr = {
 	.init_zcull = gf117_gr_init_zcull,
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
+	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk110b_gr_pack_mmio,
 	.fecs.ucode = &gk110_gr_fecs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
index 49ee549..7a938bb 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
@@ -167,6 +167,7 @@ gk208_gr = {
 	.init_zcull = gf117_gr_init_zcull,
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
+	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk208_gr_pack_mmio,
 	.fecs.ucode = &gk208_gr_fecs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
index c653df3..9f2df29 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
@@ -371,7 +371,7 @@ gm107_gr_init(struct gf100_gr *gr)
 	nvkm_wr32(device, 0x400100, 0xffffffff);
 	nvkm_wr32(device, 0x40013c, 0xffffffff);
 	nvkm_wr32(device, 0x400124, 0x00000002);
-	nvkm_wr32(device, 0x409c24, 0x000e0000);
+	gr->func->init_fecs_exceptions(gr);
 
 	nvkm_wr32(device, 0x404000, 0xc0000000);
 	nvkm_wr32(device, 0x404600, 0xc0000000);
@@ -455,6 +455,7 @@ gm107_gr = {
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_bios_2 = gm107_gr_init_bios_2,
+	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gm107_gr_pack_mmio,
 	.fecs.ucode = &gm107_gr_fecs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
index 5007435..a957993 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
@@ -91,7 +91,7 @@ gm200_gr_init(struct gf100_gr *gr)
 	nvkm_wr32(device, 0x400100, 0xffffffff);
 	nvkm_wr32(device, 0x40013c, 0xffffffff);
 	nvkm_wr32(device, 0x400124, 0x00000002);
-	nvkm_wr32(device, 0x409c24, 0x000e0000);
+	gr->func->init_fecs_exceptions(gr);
 	nvkm_wr32(device, 0x405848, 0xc0000000);
 	nvkm_wr32(device, 0x40584c, 0x00000001);
 	nvkm_wr32(device, 0x404000, 0xc0000000);
@@ -193,6 +193,7 @@ gm200_gr = {
 	.init_zcull = gf117_gr_init_zcull,
 	.init_num_active_ltcs = gm200_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gm200_gr_init_rop_active_fbps,
+	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 2,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
index 1e67959..564b5b1 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
@@ -31,6 +31,12 @@
  ******************************************************************************/
 
 void
+gp100_gr_init_fecs_exceptions(struct gf100_gr *gr)
+{
+	nvkm_wr32(gr->base.engine.subdev.device, 0x409c24, 0x000f0002);
+}
+
+void
 gp100_gr_init_rop_active_fbps(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -63,7 +69,7 @@ gp100_gr_init(struct gf100_gr *gr)
 	nvkm_wr32(device, 0x400100, 0xffffffff);
 	nvkm_wr32(device, 0x40013c, 0xffffffff);
 	nvkm_wr32(device, 0x400124, 0x00000002);
-	nvkm_wr32(device, 0x409c24, 0x000f0002);
+	gr->func->init_fecs_exceptions(gr);
 	nvkm_wr32(device, 0x405848, 0xc0000000);
 	nvkm_mask(device, 0x40584c, 0x00000000, 0x00000001);
 	nvkm_wr32(device, 0x404000, 0xc0000000);
@@ -127,6 +133,7 @@ gp100_gr = {
 	.init_zcull = gf117_gr_init_zcull,
 	.init_num_active_ltcs = gm200_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
+	.init_fecs_exceptions = gp100_gr_init_fecs_exceptions,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 2,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
index 860a789..04803fa 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
@@ -49,6 +49,7 @@ gp102_gr = {
 	.init_num_active_ltcs = gm200_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
 	.init_swdx_pes_mask = gp102_gr_init_swdx_pes_mask,
+	.init_fecs_exceptions = gp100_gr_init_fecs_exceptions,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 3,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
index 03f2264..c21cb8a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
@@ -35,6 +35,7 @@ gp107_gr = {
 	.init_num_active_ltcs = gm200_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
 	.init_swdx_pes_mask = gp102_gr_init_swdx_pes_mask,
+	.init_fecs_exceptions = gp100_gr_init_fecs_exceptions,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 1,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
index 97f8667..222b5b0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
@@ -33,6 +33,7 @@ gp10b_gr = {
 	.init_zcull = gf117_gr_init_zcull,
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
+	.init_fecs_exceptions = gp100_gr_init_fecs_exceptions,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 1,
-- 
cgit v1.1


From 3ac72e98b40ead6225eb38bcf78ec540357106c0 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise init_ds_hww_esr_2

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c | 12 ++++++++++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c |  4 ++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c |  1 +
 7 files changed, 19 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 0493483..1f764df 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -2018,6 +2018,8 @@ gf100_gr_init(struct gf100_gr *gr)
 	nvkm_wr32(device, 0x400124, 0x00000002);
 
 	gr->func->init_fecs_exceptions(gr);
+	if (gr->func->init_ds_hww_esr_2)
+		gr->func->init_ds_hww_esr_2(gr);
 
 	nvkm_wr32(device, 0x404000, 0xc0000000);
 	nvkm_wr32(device, 0x404600, 0xc0000000);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index 387938f..ff3e265 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -131,6 +131,7 @@ struct gf100_gr_func {
 	void (*init_bios_2)(struct gf100_gr *);
 	void (*init_swdx_pes_mask)(struct gf100_gr *);
 	void (*init_fecs_exceptions)(struct gf100_gr *);
+	void (*init_ds_hww_esr_2)(struct gf100_gr *);
 	void (*init_ppc_exceptions)(struct gf100_gr *);
 	void (*set_hww_esr_report_mask)(struct gf100_gr *);
 	const struct gf100_gr_pack *mmio;
@@ -165,6 +166,7 @@ int gk20a_gr_init(struct gf100_gr *);
 
 int gm200_gr_rops(struct gf100_gr *);
 void gm200_gr_init_num_active_ltcs(struct gf100_gr *);
+void gm200_gr_init_ds_hww_esr_2(struct gf100_gr *);
 
 int gp100_gr_init(struct gf100_gr *);
 void gp100_gr_init_rop_active_fbps(struct gf100_gr *);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
index a957993..b5994dc 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
@@ -39,6 +39,14 @@ gm200_gr_rops(struct gf100_gr *gr)
 }
 
 void
+gm200_gr_init_ds_hww_esr_2(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, 0x405848, 0xc0000000);
+	nvkm_mask(device, 0x40584c, 0x00000001, 0x00000001);
+}
+
+void
 gm200_gr_init_num_active_ltcs(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -92,8 +100,7 @@ gm200_gr_init(struct gf100_gr *gr)
 	nvkm_wr32(device, 0x40013c, 0xffffffff);
 	nvkm_wr32(device, 0x400124, 0x00000002);
 	gr->func->init_fecs_exceptions(gr);
-	nvkm_wr32(device, 0x405848, 0xc0000000);
-	nvkm_wr32(device, 0x40584c, 0x00000001);
+	gr->func->init_ds_hww_esr_2(gr);
 	nvkm_wr32(device, 0x404000, 0xc0000000);
 	nvkm_wr32(device, 0x404600, 0xc0000000);
 	nvkm_wr32(device, 0x408030, 0xc0000000);
@@ -194,6 +201,7 @@ gm200_gr = {
 	.init_num_active_ltcs = gm200_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gm200_gr_init_rop_active_fbps,
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
+	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 2,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
index 564b5b1..676f58a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
@@ -70,8 +70,7 @@ gp100_gr_init(struct gf100_gr *gr)
 	nvkm_wr32(device, 0x40013c, 0xffffffff);
 	nvkm_wr32(device, 0x400124, 0x00000002);
 	gr->func->init_fecs_exceptions(gr);
-	nvkm_wr32(device, 0x405848, 0xc0000000);
-	nvkm_mask(device, 0x40584c, 0x00000000, 0x00000001);
+	gr->func->init_ds_hww_esr_2(gr);
 	nvkm_wr32(device, 0x404000, 0xc0000000);
 	nvkm_wr32(device, 0x404600, 0xc0000000);
 	nvkm_wr32(device, 0x408030, 0xc0000000);
@@ -134,6 +133,7 @@ gp100_gr = {
 	.init_num_active_ltcs = gm200_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
 	.init_fecs_exceptions = gp100_gr_init_fecs_exceptions,
+	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 2,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
index 04803fa..3694687 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
@@ -50,6 +50,7 @@ gp102_gr = {
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
 	.init_swdx_pes_mask = gp102_gr_init_swdx_pes_mask,
 	.init_fecs_exceptions = gp100_gr_init_fecs_exceptions,
+	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 3,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
index c21cb8a..c83ad01 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
@@ -36,6 +36,7 @@ gp107_gr = {
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
 	.init_swdx_pes_mask = gp102_gr_init_swdx_pes_mask,
 	.init_fecs_exceptions = gp100_gr_init_fecs_exceptions,
+	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 1,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
index 222b5b0..8fef3b5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
@@ -34,6 +34,7 @@ gp10b_gr = {
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
 	.init_fecs_exceptions = gp100_gr_init_fecs_exceptions,
+	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 1,
-- 
cgit v1.1


From 2b297b0d6d33aee99254b43c3e41100fc75ea4ab Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise init_40601c

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c | 12 +++++++++++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c |  1 +
 7 files changed, 18 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 1f764df..1b067e6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -1915,6 +1915,12 @@ gf100_gr_new_(const struct gf100_gr_func *func, struct nvkm_device *device,
 }
 
 void
+gf100_gr_init_40601c(struct gf100_gr *gr)
+{
+	nvkm_wr32(gr->base.engine.subdev.device, 0x40601c, 0xc0000000);
+}
+
+void
 gf100_gr_init_fecs_exceptions(struct gf100_gr *gr)
 {
 	const u32 data = gr->firmware ? 0x000e0000 : 0x000e0001;
@@ -2024,7 +2030,10 @@ gf100_gr_init(struct gf100_gr *gr)
 	nvkm_wr32(device, 0x404000, 0xc0000000);
 	nvkm_wr32(device, 0x404600, 0xc0000000);
 	nvkm_wr32(device, 0x408030, 0xc0000000);
-	nvkm_wr32(device, 0x40601c, 0xc0000000);
+
+	if (gr->func->init_40601c)
+		gr->func->init_40601c(gr);
+
 	nvkm_wr32(device, 0x404490, 0xc0000000);
 	nvkm_wr32(device, 0x406018, 0xc0000000);
 	nvkm_wr32(device, 0x405840, 0xc0000000);
@@ -2099,6 +2108,7 @@ gf100_gr = {
 	.init_zcull = gf100_gr_init_zcull,
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
+	.init_40601c = gf100_gr_init_40601c,
 	.mmio = gf100_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index ff3e265..8c42a7a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -132,6 +132,7 @@ struct gf100_gr_func {
 	void (*init_swdx_pes_mask)(struct gf100_gr *);
 	void (*init_fecs_exceptions)(struct gf100_gr *);
 	void (*init_ds_hww_esr_2)(struct gf100_gr *);
+	void (*init_40601c)(struct gf100_gr *);
 	void (*init_ppc_exceptions)(struct gf100_gr *);
 	void (*set_hww_esr_report_mask)(struct gf100_gr *);
 	const struct gf100_gr_pack *mmio;
@@ -154,6 +155,7 @@ void gf100_gr_init_vsc_stream_master(struct gf100_gr *);
 void gf100_gr_init_zcull(struct gf100_gr *);
 void gf100_gr_init_num_active_ltcs(struct gf100_gr *);
 void gf100_gr_init_fecs_exceptions(struct gf100_gr *);
+void gf100_gr_init_40601c(struct gf100_gr *);
 
 void gf117_gr_init_zcull(struct gf100_gr *);
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
index f76995b..61d6eef 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
@@ -120,6 +120,7 @@ gf104_gr = {
 	.init_zcull = gf100_gr_init_zcull,
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
+	.init_40601c = gf100_gr_init_40601c,
 	.mmio = gf104_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
index ada2697..d4f712e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
@@ -118,6 +118,7 @@ gf108_gr = {
 	.init_zcull = gf100_gr_init_zcull,
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
+	.init_40601c = gf100_gr_init_40601c,
 	.mmio = gf108_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
index 80ced8f..1b6c2f3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
@@ -92,6 +92,7 @@ gf110_gr = {
 	.init_zcull = gf100_gr_init_zcull,
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
+	.init_40601c = gf100_gr_init_40601c,
 	.mmio = gf110_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
index 37ca121..ae76e818 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
@@ -156,6 +156,7 @@ gf117_gr = {
 	.init_zcull = gf117_gr_init_zcull,
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
+	.init_40601c = gf100_gr_init_40601c,
 	.mmio = gf117_gr_pack_mmio,
 	.fecs.ucode = &gf117_gr_fecs_ucode,
 	.gpccs.ucode = &gf117_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
index ddf05c5..a3970c3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
@@ -183,6 +183,7 @@ gf119_gr = {
 	.init_zcull = gf100_gr_init_zcull,
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
+	.init_40601c = gf100_gr_init_40601c,
 	.mmio = gf119_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
-- 
cgit v1.1


From 0a5b97304b9e2cd07c78a399c5395d5fb0118341 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise init_sked_hww_esr

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c  | 4 ++++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h  | 2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c  | 9 ++++++++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c  | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c  | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c  | 3 ++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c  | 3 ++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c  | 3 ++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c  | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c  | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c  | 1 +
 12 files changed, 26 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 1b067e6..7ccd8f2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -2036,6 +2036,10 @@ gf100_gr_init(struct gf100_gr *gr)
 
 	nvkm_wr32(device, 0x404490, 0xc0000000);
 	nvkm_wr32(device, 0x406018, 0xc0000000);
+
+	if (gr->func->init_sked_hww_esr)
+		gr->func->init_sked_hww_esr(gr);
+
 	nvkm_wr32(device, 0x405840, 0xc0000000);
 	nvkm_wr32(device, 0x405844, 0x00ffffff);
 	nvkm_mask(device, 0x419cc0, 0x00000008, 0x00000008);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index 8c42a7a..c292cf3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -133,6 +133,7 @@ struct gf100_gr_func {
 	void (*init_fecs_exceptions)(struct gf100_gr *);
 	void (*init_ds_hww_esr_2)(struct gf100_gr *);
 	void (*init_40601c)(struct gf100_gr *);
+	void (*init_sked_hww_esr)(struct gf100_gr *);
 	void (*init_ppc_exceptions)(struct gf100_gr *);
 	void (*set_hww_esr_report_mask)(struct gf100_gr *);
 	const struct gf100_gr_pack *mmio;
@@ -163,6 +164,7 @@ int gk104_gr_init(struct gf100_gr *);
 void gk104_gr_init_vsc_stream_master(struct gf100_gr *);
 void gk104_gr_init_rop_active_fbps(struct gf100_gr *);
 void gk104_gr_init_ppc_exceptions(struct gf100_gr *);
+void gk104_gr_init_sked_hww_esr(struct gf100_gr *);
 
 int gk20a_gr_init(struct gf100_gr *);
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
index 746ad3d..6ba604e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
@@ -380,6 +380,12 @@ gk104_clkgate_pack[] = {
  * PGRAPH engine/subdev functions
  ******************************************************************************/
 
+void
+gk104_gr_init_sked_hww_esr(struct gf100_gr *gr)
+{
+	nvkm_wr32(gr->base.engine.subdev.device, 0x407020, 0x40000000);
+}
+
 static void
 gk104_gr_init_fecs_exceptions(struct gf100_gr *gr)
 {
@@ -451,7 +457,7 @@ gk104_gr_init(struct gf100_gr *gr)
 	nvkm_wr32(device, 0x408030, 0xc0000000);
 	nvkm_wr32(device, 0x404490, 0xc0000000);
 	nvkm_wr32(device, 0x406018, 0xc0000000);
-	nvkm_wr32(device, 0x407020, 0x40000000);
+	gr->func->init_sked_hww_esr(gr);
 	nvkm_wr32(device, 0x405840, 0xc0000000);
 	nvkm_wr32(device, 0x405844, 0x00ffffff);
 	nvkm_mask(device, 0x419cc0, 0x00000008, 0x00000008);
@@ -527,6 +533,7 @@ gk104_gr = {
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_fecs_exceptions = gk104_gr_init_fecs_exceptions,
+	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk104_gr_pack_mmio,
 	.fecs.ucode = &gk104_gr_fecs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
index 8a6340d..7a07d24 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
@@ -343,6 +343,7 @@ gk110_gr = {
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
+	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk110_gr_pack_mmio,
 	.fecs.ucode = &gk110_gr_fecs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
index b50e681..1c9f59c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
@@ -109,6 +109,7 @@ gk110b_gr = {
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
+	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk110b_gr_pack_mmio,
 	.fecs.ucode = &gk110_gr_fecs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
index 7a938bb..40c8724 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
@@ -168,6 +168,7 @@ gk208_gr = {
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
+	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk208_gr_pack_mmio,
 	.fecs.ucode = &gk208_gr_fecs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
index 9f2df29..3d180ed 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
@@ -378,7 +378,7 @@ gm107_gr_init(struct gf100_gr *gr)
 	nvkm_wr32(device, 0x408030, 0xc0000000);
 	nvkm_wr32(device, 0x404490, 0xc0000000);
 	nvkm_wr32(device, 0x406018, 0xc0000000);
-	nvkm_wr32(device, 0x407020, 0x40000000);
+	gr->func->init_sked_hww_esr(gr);
 	nvkm_wr32(device, 0x405840, 0xc0000000);
 	nvkm_wr32(device, 0x405844, 0x00ffffff);
 	nvkm_mask(device, 0x419cc0, 0x00000008, 0x00000008);
@@ -456,6 +456,7 @@ gm107_gr = {
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_bios_2 = gm107_gr_init_bios_2,
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
+	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gm107_gr_pack_mmio,
 	.fecs.ucode = &gm107_gr_fecs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
index b5994dc..9436ab3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
@@ -106,7 +106,7 @@ gm200_gr_init(struct gf100_gr *gr)
 	nvkm_wr32(device, 0x408030, 0xc0000000);
 	nvkm_wr32(device, 0x404490, 0xc0000000);
 	nvkm_wr32(device, 0x406018, 0xc0000000);
-	nvkm_wr32(device, 0x407020, 0x40000000);
+	gr->func->init_sked_hww_esr(gr);
 	nvkm_wr32(device, 0x405840, 0xc0000000);
 	nvkm_wr32(device, 0x405844, 0x00ffffff);
 	nvkm_mask(device, 0x419cc0, 0x00000008, 0x00000008);
@@ -202,6 +202,7 @@ gm200_gr = {
 	.init_rop_active_fbps = gm200_gr_init_rop_active_fbps,
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
+	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 2,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
index 676f58a..72ea16e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
@@ -76,7 +76,7 @@ gp100_gr_init(struct gf100_gr *gr)
 	nvkm_wr32(device, 0x408030, 0xc0000000);
 	nvkm_wr32(device, 0x404490, 0xc0000000);
 	nvkm_wr32(device, 0x406018, 0xc0000000);
-	nvkm_wr32(device, 0x407020, 0x40000000);
+	gr->func->init_sked_hww_esr(gr);
 	nvkm_wr32(device, 0x405840, 0xc0000000);
 	nvkm_wr32(device, 0x405844, 0x00ffffff);
 	nvkm_mask(device, 0x419cc0, 0x00000008, 0x00000008);
@@ -134,6 +134,7 @@ gp100_gr = {
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
 	.init_fecs_exceptions = gp100_gr_init_fecs_exceptions,
 	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
+	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 2,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
index 3694687..309815b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
@@ -51,6 +51,7 @@ gp102_gr = {
 	.init_swdx_pes_mask = gp102_gr_init_swdx_pes_mask,
 	.init_fecs_exceptions = gp100_gr_init_fecs_exceptions,
 	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
+	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 3,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
index c83ad01..1ae9e7d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
@@ -37,6 +37,7 @@ gp107_gr = {
 	.init_swdx_pes_mask = gp102_gr_init_swdx_pes_mask,
 	.init_fecs_exceptions = gp100_gr_init_fecs_exceptions,
 	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
+	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 1,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
index 8fef3b5..68e2128 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
@@ -35,6 +35,7 @@ gp10b_gr = {
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
 	.init_fecs_exceptions = gp100_gr_init_fecs_exceptions,
 	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
+	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 1,
-- 
cgit v1.1


From 0feab0250d34c7114b442f49b1ce18a9906b543d Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise init_419cc0 + apply fixes from
 traces

Pulled some init out of main per-GPC/TPC loops to match RM.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c  | 21 +++++++++++++++++++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h  |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c  |  4 ++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c  |  4 ++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c  |  4 ++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c  |  4 ++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c  |  1 +
 17 files changed, 40 insertions(+), 10 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 7ccd8f2..ad18ef9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -1915,6 +1915,20 @@ gf100_gr_new_(const struct gf100_gr_func *func, struct nvkm_device *device,
 }
 
 void
+gf100_gr_init_419cc0(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	int gpc, tpc;
+
+	nvkm_mask(device, 0x419cc0, 0x00000008, 0x00000008);
+
+	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
+		for (tpc = 0; tpc < gr->tpc_nr[gpc]; tpc++)
+			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x48c), 0xc0000000);
+	}
+}
+
+void
 gf100_gr_init_40601c(struct gf100_gr *gr)
 {
 	nvkm_wr32(gr->base.engine.subdev.device, 0x40601c, 0xc0000000);
@@ -2042,7 +2056,10 @@ gf100_gr_init(struct gf100_gr *gr)
 
 	nvkm_wr32(device, 0x405840, 0xc0000000);
 	nvkm_wr32(device, 0x405844, 0x00ffffff);
-	nvkm_mask(device, 0x419cc0, 0x00000008, 0x00000008);
+
+	if (gr->func->init_419cc0)
+		gr->func->init_419cc0(gr);
+
 	nvkm_mask(device, 0x419eb4, 0x00001000, 0x00001000);
 
 	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
@@ -2054,7 +2071,6 @@ gf100_gr_init(struct gf100_gr *gr)
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x508), 0xffffffff);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x224), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x48c), 0xc0000000);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x001ffffe);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x0000000f);
@@ -2113,6 +2129,7 @@ gf100_gr = {
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_40601c = gf100_gr_init_40601c,
+	.init_419cc0 = gf100_gr_init_419cc0,
 	.mmio = gf100_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index c292cf3..cab0948 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -134,6 +134,7 @@ struct gf100_gr_func {
 	void (*init_ds_hww_esr_2)(struct gf100_gr *);
 	void (*init_40601c)(struct gf100_gr *);
 	void (*init_sked_hww_esr)(struct gf100_gr *);
+	void (*init_419cc0)(struct gf100_gr *);
 	void (*init_ppc_exceptions)(struct gf100_gr *);
 	void (*set_hww_esr_report_mask)(struct gf100_gr *);
 	const struct gf100_gr_pack *mmio;
@@ -157,6 +158,7 @@ void gf100_gr_init_zcull(struct gf100_gr *);
 void gf100_gr_init_num_active_ltcs(struct gf100_gr *);
 void gf100_gr_init_fecs_exceptions(struct gf100_gr *);
 void gf100_gr_init_40601c(struct gf100_gr *);
+void gf100_gr_init_419cc0(struct gf100_gr *);
 
 void gf117_gr_init_zcull(struct gf100_gr *);
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
index 61d6eef..8598e15 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
@@ -121,6 +121,7 @@ gf104_gr = {
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_40601c = gf100_gr_init_40601c,
+	.init_419cc0 = gf100_gr_init_419cc0,
 	.mmio = gf104_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
index d4f712e..83c71ff 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
@@ -119,6 +119,7 @@ gf108_gr = {
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_40601c = gf100_gr_init_40601c,
+	.init_419cc0 = gf100_gr_init_419cc0,
 	.mmio = gf108_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
index 1b6c2f3..b2b4489 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
@@ -93,6 +93,7 @@ gf110_gr = {
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_40601c = gf100_gr_init_40601c,
+	.init_419cc0 = gf100_gr_init_419cc0,
 	.mmio = gf110_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
index ae76e818..ec91a59 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
@@ -157,6 +157,7 @@ gf117_gr = {
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_40601c = gf100_gr_init_40601c,
+	.init_419cc0 = gf100_gr_init_419cc0,
 	.mmio = gf117_gr_pack_mmio,
 	.fecs.ucode = &gf117_gr_fecs_ucode,
 	.gpccs.ucode = &gf117_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
index a3970c3..570527f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
@@ -184,6 +184,7 @@ gf119_gr = {
 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_40601c = gf100_gr_init_40601c,
+	.init_419cc0 = gf100_gr_init_419cc0,
 	.mmio = gf119_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
index 6ba604e..dcd59af 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
@@ -460,7 +460,7 @@ gk104_gr_init(struct gf100_gr *gr)
 	gr->func->init_sked_hww_esr(gr);
 	nvkm_wr32(device, 0x405840, 0xc0000000);
 	nvkm_wr32(device, 0x405844, 0x00ffffff);
-	nvkm_mask(device, 0x419cc0, 0x00000008, 0x00000008);
+	gr->func->init_419cc0(gr);
 	nvkm_mask(device, 0x419eb4, 0x00001000, 0x00001000);
 
 	gr->func->init_ppc_exceptions(gr);
@@ -474,7 +474,6 @@ gk104_gr_init(struct gf100_gr *gr)
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x508), 0xffffffff);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x224), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x48c), 0xc0000000);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x001ffffe);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x0000000f);
@@ -534,6 +533,7 @@ gk104_gr = {
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_fecs_exceptions = gk104_gr_init_fecs_exceptions,
 	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
+	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk104_gr_pack_mmio,
 	.fecs.ucode = &gk104_gr_fecs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
index 7a07d24..bea0cd3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
@@ -344,6 +344,7 @@ gk110_gr = {
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
+	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk110_gr_pack_mmio,
 	.fecs.ucode = &gk110_gr_fecs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
index 1c9f59c..e4ae88d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
@@ -110,6 +110,7 @@ gk110b_gr = {
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
+	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk110b_gr_pack_mmio,
 	.fecs.ucode = &gk110_gr_fecs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
index 40c8724..1e214d9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
@@ -169,6 +169,7 @@ gk208_gr = {
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
+	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk208_gr_pack_mmio,
 	.fecs.ucode = &gk208_gr_fecs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
index 3d180ed..df668b9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
@@ -381,7 +381,7 @@ gm107_gr_init(struct gf100_gr *gr)
 	gr->func->init_sked_hww_esr(gr);
 	nvkm_wr32(device, 0x405840, 0xc0000000);
 	nvkm_wr32(device, 0x405844, 0x00ffffff);
-	nvkm_mask(device, 0x419cc0, 0x00000008, 0x00000008);
+	gr->func->init_419cc0(gr);
 
 	gr->func->init_ppc_exceptions(gr);
 
@@ -394,7 +394,6 @@ gm107_gr_init(struct gf100_gr *gr)
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x508), 0xffffffff);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x224), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x48c), 0xc0000000);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x430), 0xc0000000);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x00dffffe);
@@ -457,6 +456,7 @@ gm107_gr = {
 	.init_bios_2 = gm107_gr_init_bios_2,
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
+	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gm107_gr_pack_mmio,
 	.fecs.ucode = &gm107_gr_fecs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
index 9436ab3..9f8171a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
@@ -109,7 +109,7 @@ gm200_gr_init(struct gf100_gr *gr)
 	gr->func->init_sked_hww_esr(gr);
 	nvkm_wr32(device, 0x405840, 0xc0000000);
 	nvkm_wr32(device, 0x405844, 0x00ffffff);
-	nvkm_mask(device, 0x419cc0, 0x00000008, 0x00000008);
+	gr->func->init_419cc0(gr);
 
 	gr->func->init_ppc_exceptions(gr);
 
@@ -122,7 +122,6 @@ gm200_gr_init(struct gf100_gr *gr)
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x508), 0xffffffff);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x224), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x48c), 0xc0000000);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x430), 0xc0000000);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x00dffffe);
@@ -203,6 +202,7 @@ gm200_gr = {
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
 	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
+	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 2,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
index 72ea16e..7518d24 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
@@ -79,7 +79,7 @@ gp100_gr_init(struct gf100_gr *gr)
 	gr->func->init_sked_hww_esr(gr);
 	nvkm_wr32(device, 0x405840, 0xc0000000);
 	nvkm_wr32(device, 0x405844, 0x00ffffff);
-	nvkm_mask(device, 0x419cc0, 0x00000008, 0x00000008);
+	gr->func->init_419cc0(gr);
 
 	nvkm_mask(device, 0x419c9c, 0x00010000, 0x00010000);
 	nvkm_mask(device, 0x419c9c, 0x00020000, 0x00020000);
@@ -95,7 +95,6 @@ gp100_gr_init(struct gf100_gr *gr)
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x508), 0xffffffff);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x224), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x48c), 0xc0000000);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x430), 0xc0000000);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x00dffffe);
@@ -135,6 +134,7 @@ gp100_gr = {
 	.init_fecs_exceptions = gp100_gr_init_fecs_exceptions,
 	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
 	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
+	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 2,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
index 309815b..230e2ee 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
@@ -52,6 +52,7 @@ gp102_gr = {
 	.init_fecs_exceptions = gp100_gr_init_fecs_exceptions,
 	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
 	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
+	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 3,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
index 1ae9e7d..55b7a7e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
@@ -38,6 +38,7 @@ gp107_gr = {
 	.init_fecs_exceptions = gp100_gr_init_fecs_exceptions,
 	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
 	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
+	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 1,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
index 68e2128..b13a48b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
@@ -36,6 +36,7 @@ gp10b_gr = {
 	.init_fecs_exceptions = gp100_gr_init_fecs_exceptions,
 	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
 	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
+	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 1,
-- 
cgit v1.1


From 0a84a51334b5d75decd23b735aab00ff4698eeb2 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise init_419eb4 + apply fixes from
 traces

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c  | 12 ++++++++++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h  |  4 ++++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c  |  3 ++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c  | 15 +++++++++++++++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c |  1 +
 10 files changed, 37 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index ad18ef9..9472782 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -1915,6 +1915,13 @@ gf100_gr_new_(const struct gf100_gr_func *func, struct nvkm_device *device,
 }
 
 void
+gf100_gr_init_419eb4(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x419eb4, 0x00001000, 0x00001000);
+}
+
+void
 gf100_gr_init_419cc0(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -2059,8 +2066,8 @@ gf100_gr_init(struct gf100_gr *gr)
 
 	if (gr->func->init_419cc0)
 		gr->func->init_419cc0(gr);
-
-	nvkm_mask(device, 0x419eb4, 0x00001000, 0x00001000);
+	if (gr->func->init_419eb4)
+		gr->func->init_419eb4(gr);
 
 	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
 		nvkm_wr32(device, GPC_UNIT(gpc, 0x0420), 0xc0000000);
@@ -2130,6 +2137,7 @@ gf100_gr = {
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_40601c = gf100_gr_init_40601c,
 	.init_419cc0 = gf100_gr_init_419cc0,
+	.init_419eb4 = gf100_gr_init_419eb4,
 	.mmio = gf100_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index cab0948..1dae373 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -135,6 +135,7 @@ struct gf100_gr_func {
 	void (*init_40601c)(struct gf100_gr *);
 	void (*init_sked_hww_esr)(struct gf100_gr *);
 	void (*init_419cc0)(struct gf100_gr *);
+	void (*init_419eb4)(struct gf100_gr *);
 	void (*init_ppc_exceptions)(struct gf100_gr *);
 	void (*set_hww_esr_report_mask)(struct gf100_gr *);
 	const struct gf100_gr_pack *mmio;
@@ -159,6 +160,7 @@ void gf100_gr_init_num_active_ltcs(struct gf100_gr *);
 void gf100_gr_init_fecs_exceptions(struct gf100_gr *);
 void gf100_gr_init_40601c(struct gf100_gr *);
 void gf100_gr_init_419cc0(struct gf100_gr *);
+void gf100_gr_init_419eb4(struct gf100_gr *);
 
 void gf117_gr_init_zcull(struct gf100_gr *);
 
@@ -168,6 +170,8 @@ void gk104_gr_init_rop_active_fbps(struct gf100_gr *);
 void gk104_gr_init_ppc_exceptions(struct gf100_gr *);
 void gk104_gr_init_sked_hww_esr(struct gf100_gr *);
 
+void gk110_gr_init_419eb4(struct gf100_gr *);
+
 int gk20a_gr_init(struct gf100_gr *);
 
 int gm200_gr_rops(struct gf100_gr *);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
index 8598e15..59d4dac 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
@@ -122,6 +122,7 @@ gf104_gr = {
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_40601c = gf100_gr_init_40601c,
 	.init_419cc0 = gf100_gr_init_419cc0,
+	.init_419eb4 = gf100_gr_init_419eb4,
 	.mmio = gf104_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
index 83c71ff..73cb23a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
@@ -120,6 +120,7 @@ gf108_gr = {
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_40601c = gf100_gr_init_40601c,
 	.init_419cc0 = gf100_gr_init_419cc0,
+	.init_419eb4 = gf100_gr_init_419eb4,
 	.mmio = gf108_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
index b2b4489..1d7188e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
@@ -94,6 +94,7 @@ gf110_gr = {
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_40601c = gf100_gr_init_40601c,
 	.init_419cc0 = gf100_gr_init_419cc0,
+	.init_419eb4 = gf100_gr_init_419eb4,
 	.mmio = gf110_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
index ec91a59..3465567 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
@@ -158,6 +158,7 @@ gf117_gr = {
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_40601c = gf100_gr_init_40601c,
 	.init_419cc0 = gf100_gr_init_419cc0,
+	.init_419eb4 = gf100_gr_init_419eb4,
 	.mmio = gf117_gr_pack_mmio,
 	.fecs.ucode = &gf117_gr_fecs_ucode,
 	.gpccs.ucode = &gf117_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
index 570527f..28c00cb 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
@@ -185,6 +185,7 @@ gf119_gr = {
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_40601c = gf100_gr_init_40601c,
 	.init_419cc0 = gf100_gr_init_419cc0,
+	.init_419eb4 = gf100_gr_init_419eb4,
 	.mmio = gf119_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
index dcd59af..775b474 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
@@ -461,7 +461,7 @@ gk104_gr_init(struct gf100_gr *gr)
 	nvkm_wr32(device, 0x405840, 0xc0000000);
 	nvkm_wr32(device, 0x405844, 0x00ffffff);
 	gr->func->init_419cc0(gr);
-	nvkm_mask(device, 0x419eb4, 0x00001000, 0x00001000);
+	gr->func->init_419eb4(gr);
 
 	gr->func->init_ppc_exceptions(gr);
 
@@ -534,6 +534,7 @@ gk104_gr = {
 	.init_fecs_exceptions = gk104_gr_init_fecs_exceptions,
 	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
 	.init_419cc0 = gf100_gr_init_419cc0,
+	.init_419eb4 = gf100_gr_init_419eb4,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk104_gr_pack_mmio,
 	.fecs.ucode = &gk104_gr_fecs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
index bea0cd3..daaa449 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
@@ -334,6 +334,20 @@ gk110_gr_gpccs_ucode = {
 	.data.size = sizeof(gk110_grgpc_data),
 };
 
+void
+gk110_gr_init_419eb4(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x419eb4, 0x00001000, 0x00001000);
+	nvkm_mask(device, 0x419eb4, 0x00002000, 0x00002000);
+	nvkm_mask(device, 0x419eb4, 0x00004000, 0x00004000);
+	nvkm_mask(device, 0x419eb4, 0x00008000, 0x00008000);
+	nvkm_mask(device, 0x419eb4, 0x00001000, 0x00000000);
+	nvkm_mask(device, 0x419eb4, 0x00002000, 0x00000000);
+	nvkm_mask(device, 0x419eb4, 0x00004000, 0x00000000);
+	nvkm_mask(device, 0x419eb4, 0x00008000, 0x00000000);
+}
+
 static const struct gf100_gr_func
 gk110_gr = {
 	.init = gk104_gr_init,
@@ -345,6 +359,7 @@ gk110_gr = {
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
 	.init_419cc0 = gf100_gr_init_419cc0,
+	.init_419eb4 = gk110_gr_init_419eb4,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk110_gr_pack_mmio,
 	.fecs.ucode = &gk110_gr_fecs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
index e4ae88d..ee56b00 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
@@ -111,6 +111,7 @@ gk110b_gr = {
 	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
 	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
 	.init_419cc0 = gf100_gr_init_419cc0,
+	.init_419eb4 = gk110_gr_init_419eb4,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.mmio = gk110b_gr_pack_mmio,
 	.fecs.ucode = &gk110_gr_fecs_ucode,
-- 
cgit v1.1


From 778f18c607e30206c1a791a4d356f1ed32bc1947 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise init_419c9c + apply fixes from
 traces

Deliberately removed from non-GP100, as RM doesn't touch it.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c | 14 +++++++++++---
 3 files changed, 14 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 9472782..a379794 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -2068,6 +2068,8 @@ gf100_gr_init(struct gf100_gr *gr)
 		gr->func->init_419cc0(gr);
 	if (gr->func->init_419eb4)
 		gr->func->init_419eb4(gr);
+	if (gr->func->init_419c9c)
+		gr->func->init_419c9c(gr);
 
 	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
 		nvkm_wr32(device, GPC_UNIT(gpc, 0x0420), 0xc0000000);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index 1dae373..e328957 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -136,6 +136,7 @@ struct gf100_gr_func {
 	void (*init_sked_hww_esr)(struct gf100_gr *);
 	void (*init_419cc0)(struct gf100_gr *);
 	void (*init_419eb4)(struct gf100_gr *);
+	void (*init_419c9c)(struct gf100_gr *);
 	void (*init_ppc_exceptions)(struct gf100_gr *);
 	void (*set_hww_esr_report_mask)(struct gf100_gr *);
 	const struct gf100_gr_pack *mmio;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
index 7518d24..3fcbe1f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
@@ -30,6 +30,14 @@
  * PGRAPH engine/subdev functions
  ******************************************************************************/
 
+static void
+gp100_gr_init_419c9c(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x419c9c, 0x00010000, 0x00010000);
+	nvkm_mask(device, 0x419c9c, 0x00020000, 0x00020000);
+}
+
 void
 gp100_gr_init_fecs_exceptions(struct gf100_gr *gr)
 {
@@ -80,9 +88,8 @@ gp100_gr_init(struct gf100_gr *gr)
 	nvkm_wr32(device, 0x405840, 0xc0000000);
 	nvkm_wr32(device, 0x405844, 0x00ffffff);
 	gr->func->init_419cc0(gr);
-
-	nvkm_mask(device, 0x419c9c, 0x00010000, 0x00010000);
-	nvkm_mask(device, 0x419c9c, 0x00020000, 0x00020000);
+	if (gr->func->init_419c9c)
+		gr->func->init_419c9c(gr);
 
 	gr->func->init_ppc_exceptions(gr);
 
@@ -135,6 +142,7 @@ gp100_gr = {
 	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
 	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
 	.init_419cc0 = gf100_gr_init_419cc0,
+	.init_419c9c = gp100_gr_init_419c9c,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 2,
-- 
cgit v1.1


From 70d2148209abd851768396cb4719710de9aeddc0 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise init_ppc_exceptions

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index a379794..7764425 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -2071,6 +2071,9 @@ gf100_gr_init(struct gf100_gr *gr)
 	if (gr->func->init_419c9c)
 		gr->func->init_419c9c(gr);
 
+	if (gr->func->init_ppc_exceptions)
+		gr->func->init_ppc_exceptions(gr);
+
 	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
 		nvkm_wr32(device, GPC_UNIT(gpc, 0x0420), 0xc0000000);
 		nvkm_wr32(device, GPC_UNIT(gpc, 0x0900), 0xc0000000);
-- 
cgit v1.1


From f3ef80c0c491bd1a5ae4c02acbdabb8c9cedb315 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise init_tex_hww_esr

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c  | 11 ++++++++++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h  |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c  |  3 ++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c  |  3 ++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c  |  3 ++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c  |  3 ++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c  |  1 +
 17 files changed, 31 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 7764425..25f64ec 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -1915,6 +1915,13 @@ gf100_gr_new_(const struct gf100_gr_func *func, struct nvkm_device *device,
 }
 
 void
+gf100_gr_init_tex_hww_esr(struct gf100_gr *gr, int tpc, int gpc)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x224), 0xc0000000);
+}
+
+void
 gf100_gr_init_419eb4(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -2082,7 +2089,8 @@ gf100_gr_init(struct gf100_gr *gr)
 		for (tpc = 0; tpc < gr->tpc_nr[gpc]; tpc++) {
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x508), 0xffffffff);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x224), 0xc0000000);
+			if (gr->func->init_tex_hww_esr)
+				gr->func->init_tex_hww_esr(gr, gpc, tpc);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x001ffffe);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x0000000f);
@@ -2143,6 +2151,7 @@ gf100_gr = {
 	.init_40601c = gf100_gr_init_40601c,
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_419eb4 = gf100_gr_init_419eb4,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.mmio = gf100_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index e328957..0bbe11d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -138,6 +138,7 @@ struct gf100_gr_func {
 	void (*init_419eb4)(struct gf100_gr *);
 	void (*init_419c9c)(struct gf100_gr *);
 	void (*init_ppc_exceptions)(struct gf100_gr *);
+	void (*init_tex_hww_esr)(struct gf100_gr *, int gpc, int tpc);
 	void (*set_hww_esr_report_mask)(struct gf100_gr *);
 	const struct gf100_gr_pack *mmio;
 	struct {
@@ -162,6 +163,7 @@ void gf100_gr_init_fecs_exceptions(struct gf100_gr *);
 void gf100_gr_init_40601c(struct gf100_gr *);
 void gf100_gr_init_419cc0(struct gf100_gr *);
 void gf100_gr_init_419eb4(struct gf100_gr *);
+void gf100_gr_init_tex_hww_esr(struct gf100_gr *, int, int);
 
 void gf117_gr_init_zcull(struct gf100_gr *);
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
index 59d4dac..ff03f13 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
@@ -123,6 +123,7 @@ gf104_gr = {
 	.init_40601c = gf100_gr_init_40601c,
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_419eb4 = gf100_gr_init_419eb4,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.mmio = gf104_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
index 73cb23a..ce60f5a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
@@ -121,6 +121,7 @@ gf108_gr = {
 	.init_40601c = gf100_gr_init_40601c,
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_419eb4 = gf100_gr_init_419eb4,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.mmio = gf108_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
index 1d7188e..4c21cbe 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
@@ -95,6 +95,7 @@ gf110_gr = {
 	.init_40601c = gf100_gr_init_40601c,
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_419eb4 = gf100_gr_init_419eb4,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.mmio = gf110_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
index 3465567..10340e9f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
@@ -159,6 +159,7 @@ gf117_gr = {
 	.init_40601c = gf100_gr_init_40601c,
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_419eb4 = gf100_gr_init_419eb4,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.mmio = gf117_gr_pack_mmio,
 	.fecs.ucode = &gf117_gr_fecs_ucode,
 	.gpccs.ucode = &gf117_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
index 28c00cb..5147c59 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
@@ -186,6 +186,7 @@ gf119_gr = {
 	.init_40601c = gf100_gr_init_40601c,
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_419eb4 = gf100_gr_init_419eb4,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.mmio = gf119_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
index 775b474..18051963 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
@@ -473,7 +473,7 @@ gk104_gr_init(struct gf100_gr *gr)
 		for (tpc = 0; tpc < gr->tpc_nr[gpc]; tpc++) {
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x508), 0xffffffff);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x224), 0xc0000000);
+			gr->func->init_tex_hww_esr(gr, gpc, tpc);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x001ffffe);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x0000000f);
@@ -536,6 +536,7 @@ gk104_gr = {
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_419eb4 = gf100_gr_init_419eb4,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.mmio = gk104_gr_pack_mmio,
 	.fecs.ucode = &gk104_gr_fecs_ucode,
 	.gpccs.ucode = &gk104_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
index daaa449..f4c73a9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
@@ -361,6 +361,7 @@ gk110_gr = {
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_419eb4 = gk110_gr_init_419eb4,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.mmio = gk110_gr_pack_mmio,
 	.fecs.ucode = &gk110_gr_fecs_ucode,
 	.gpccs.ucode = &gk110_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
index ee56b00..b585ab2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
@@ -113,6 +113,7 @@ gk110b_gr = {
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_419eb4 = gk110_gr_init_419eb4,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.mmio = gk110b_gr_pack_mmio,
 	.fecs.ucode = &gk110_gr_fecs_ucode,
 	.gpccs.ucode = &gk110_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
index 1e214d9..7d1f585 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
@@ -171,6 +171,7 @@ gk208_gr = {
 	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.mmio = gk208_gr_pack_mmio,
 	.fecs.ucode = &gk208_gr_fecs_ucode,
 	.gpccs.ucode = &gk208_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
index df668b9..8ee7723 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
@@ -393,7 +393,7 @@ gm107_gr_init(struct gf100_gr *gr)
 		for (tpc = 0; tpc < gr->tpc_nr[gpc]; tpc++) {
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x508), 0xffffffff);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x224), 0xc0000000);
+			gr->func->init_tex_hww_esr(gr, gpc, tpc);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x430), 0xc0000000);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x00dffffe);
@@ -458,6 +458,7 @@ gm107_gr = {
 	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.mmio = gm107_gr_pack_mmio,
 	.fecs.ucode = &gm107_gr_fecs_ucode,
 	.gpccs.ucode = &gm107_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
index 9f8171a..0d02009 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
@@ -121,7 +121,7 @@ gm200_gr_init(struct gf100_gr *gr)
 		for (tpc = 0; tpc < gr->tpc_nr[gpc]; tpc++) {
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x508), 0xffffffff);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x224), 0xc0000000);
+			gr->func->init_tex_hww_esr(gr, gpc, tpc);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x430), 0xc0000000);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x00dffffe);
@@ -204,6 +204,7 @@ gm200_gr = {
 	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 2,
 	.grctx = &gm200_grctx,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
index 3fcbe1f..c798608 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
@@ -101,7 +101,7 @@ gp100_gr_init(struct gf100_gr *gr)
 		for (tpc = 0; tpc < gr->tpc_nr[gpc]; tpc++) {
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x508), 0xffffffff);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x224), 0xc0000000);
+			gr->func->init_tex_hww_esr(gr, gpc, tpc);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x430), 0xc0000000);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x00dffffe);
@@ -144,6 +144,7 @@ gp100_gr = {
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_419c9c = gp100_gr_init_419c9c,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 2,
 	.grctx = &gp100_grctx,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
index 230e2ee..a7fe2d9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
@@ -54,6 +54,7 @@ gp102_gr = {
 	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 3,
 	.grctx = &gp102_grctx,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
index 55b7a7e..f085d01 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
@@ -40,6 +40,7 @@ gp107_gr = {
 	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 1,
 	.grctx = &gp107_grctx,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
index b13a48b..9530edc 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
@@ -38,6 +38,7 @@ gp10b_gr = {
 	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 1,
 	.grctx = &gp102_grctx,
-- 
cgit v1.1


From ab4d49a349653dcd902be8974c4f7927cd49b11d Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise init_504430

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h |  3 +++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c | 10 +++++++++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c |  3 ++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c |  3 ++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c |  1 +
 8 files changed, 21 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 25f64ec..274ff01 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -2092,6 +2092,8 @@ gf100_gr_init(struct gf100_gr *gr)
 			if (gr->func->init_tex_hww_esr)
 				gr->func->init_tex_hww_esr(gr, gpc, tpc);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
+			if (gr->func->init_504430)
+				gr->func->init_504430(gr, gpc, tpc);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x001ffffe);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x0000000f);
 		}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index 0bbe11d..81d5931 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -139,6 +139,7 @@ struct gf100_gr_func {
 	void (*init_419c9c)(struct gf100_gr *);
 	void (*init_ppc_exceptions)(struct gf100_gr *);
 	void (*init_tex_hww_esr)(struct gf100_gr *, int gpc, int tpc);
+	void (*init_504430)(struct gf100_gr *, int gpc, int tpc);
 	void (*set_hww_esr_report_mask)(struct gf100_gr *);
 	const struct gf100_gr_pack *mmio;
 	struct {
@@ -175,6 +176,8 @@ void gk104_gr_init_sked_hww_esr(struct gf100_gr *);
 
 void gk110_gr_init_419eb4(struct gf100_gr *);
 
+void gm107_gr_init_504430(struct gf100_gr *, int, int);
+
 int gk20a_gr_init(struct gf100_gr *);
 
 int gm200_gr_rops(struct gf100_gr *);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
index 8ee7723..4c8c256 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
@@ -281,6 +281,13 @@ gm107_gr_pack_mmio[] = {
  * PGRAPH engine/subdev functions
  ******************************************************************************/
 
+void
+gm107_gr_init_504430(struct gf100_gr *gr, int gpc, int tpc)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x430), 0xc0000000);
+}
+
 static void
 gm107_gr_init_bios_2(struct gf100_gr *gr)
 {
@@ -395,7 +402,7 @@ gm107_gr_init(struct gf100_gr *gr)
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff);
 			gr->func->init_tex_hww_esr(gr, gpc, tpc);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x430), 0xc0000000);
+			gr->func->init_504430(gr, gpc, tpc);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x00dffffe);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x00000005);
 		}
@@ -459,6 +466,7 @@ gm107_gr = {
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_504430 = gm107_gr_init_504430,
 	.mmio = gm107_gr_pack_mmio,
 	.fecs.ucode = &gm107_gr_fecs_ucode,
 	.gpccs.ucode = &gm107_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
index 0d02009..69f191e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
@@ -123,7 +123,7 @@ gm200_gr_init(struct gf100_gr *gr)
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff);
 			gr->func->init_tex_hww_esr(gr, gpc, tpc);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x430), 0xc0000000);
+			gr->func->init_504430(gr, gpc, tpc);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x00dffffe);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x00000005);
 		}
@@ -205,6 +205,7 @@ gm200_gr = {
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_504430 = gm107_gr_init_504430,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 2,
 	.grctx = &gm200_grctx,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
index c798608..76a9416 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
@@ -103,7 +103,7 @@ gp100_gr_init(struct gf100_gr *gr)
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff);
 			gr->func->init_tex_hww_esr(gr, gpc, tpc);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x430), 0xc0000000);
+			gr->func->init_504430(gr, gpc, gpc);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x00dffffe);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x00000105);
 		}
@@ -145,6 +145,7 @@ gp100_gr = {
 	.init_419c9c = gp100_gr_init_419c9c,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_504430 = gm107_gr_init_504430,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 2,
 	.grctx = &gp100_grctx,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
index a7fe2d9..7072578 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
@@ -55,6 +55,7 @@ gp102_gr = {
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_504430 = gm107_gr_init_504430,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 3,
 	.grctx = &gp102_grctx,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
index f085d01..f23f70f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
@@ -41,6 +41,7 @@ gp107_gr = {
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_504430 = gm107_gr_init_504430,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 1,
 	.grctx = &gp107_grctx,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
index 9530edc..0ff1759 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
@@ -39,6 +39,7 @@ gp10b_gr = {
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_504430 = gm107_gr_init_504430,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 1,
 	.grctx = &gp102_grctx,
-- 
cgit v1.1


From 4615e9b4387f142e4ff495dc61525249a6926e91 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise init_shader_exceptions

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c  | 14 +++++++++++---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h  |  4 ++++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c  |  4 ++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c  | 12 ++++++++++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c  |  4 ++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c  | 12 ++++++++++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c  |  1 +
 17 files changed, 50 insertions(+), 11 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 274ff01..071b3c1 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -1915,7 +1915,15 @@ gf100_gr_new_(const struct gf100_gr_func *func, struct nvkm_device *device,
 }
 
 void
-gf100_gr_init_tex_hww_esr(struct gf100_gr *gr, int tpc, int gpc)
+gf100_gr_init_shader_exceptions(struct gf100_gr *gr, int gpc, int tpc)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x001ffffe);
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x0000000f);
+}
+
+void
+gf100_gr_init_tex_hww_esr(struct gf100_gr *gr, int gpc, int tpc)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
 	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x224), 0xc0000000);
@@ -2094,8 +2102,7 @@ gf100_gr_init(struct gf100_gr *gr)
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
 			if (gr->func->init_504430)
 				gr->func->init_504430(gr, gpc, tpc);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x001ffffe);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x0000000f);
+			gr->func->init_shader_exceptions(gr, gpc, tpc);
 		}
 		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c90), 0xffffffff);
 		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c94), 0xffffffff);
@@ -2154,6 +2161,7 @@ gf100_gr = {
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_419eb4 = gf100_gr_init_419eb4,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
 	.mmio = gf100_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index 81d5931..02505c0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -140,6 +140,7 @@ struct gf100_gr_func {
 	void (*init_ppc_exceptions)(struct gf100_gr *);
 	void (*init_tex_hww_esr)(struct gf100_gr *, int gpc, int tpc);
 	void (*init_504430)(struct gf100_gr *, int gpc, int tpc);
+	void (*init_shader_exceptions)(struct gf100_gr *, int gpc, int tpc);
 	void (*set_hww_esr_report_mask)(struct gf100_gr *);
 	const struct gf100_gr_pack *mmio;
 	struct {
@@ -165,6 +166,7 @@ void gf100_gr_init_40601c(struct gf100_gr *);
 void gf100_gr_init_419cc0(struct gf100_gr *);
 void gf100_gr_init_419eb4(struct gf100_gr *);
 void gf100_gr_init_tex_hww_esr(struct gf100_gr *, int, int);
+void gf100_gr_init_shader_exceptions(struct gf100_gr *, int, int);
 
 void gf117_gr_init_zcull(struct gf100_gr *);
 
@@ -177,6 +179,7 @@ void gk104_gr_init_sked_hww_esr(struct gf100_gr *);
 void gk110_gr_init_419eb4(struct gf100_gr *);
 
 void gm107_gr_init_504430(struct gf100_gr *, int, int);
+void gm107_gr_init_shader_exceptions(struct gf100_gr *, int, int);
 
 int gk20a_gr_init(struct gf100_gr *);
 
@@ -187,6 +190,7 @@ void gm200_gr_init_ds_hww_esr_2(struct gf100_gr *);
 int gp100_gr_init(struct gf100_gr *);
 void gp100_gr_init_rop_active_fbps(struct gf100_gr *);
 void gp100_gr_init_fecs_exceptions(struct gf100_gr *);
+void gp100_gr_init_shader_exceptions(struct gf100_gr *, int, int);
 
 void gp102_gr_init_swdx_pes_mask(struct gf100_gr *);
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
index ff03f13..c615444 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
@@ -124,6 +124,7 @@ gf104_gr = {
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_419eb4 = gf100_gr_init_419eb4,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
 	.mmio = gf104_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
index ce60f5a..8e8a99e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
@@ -122,6 +122,7 @@ gf108_gr = {
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_419eb4 = gf100_gr_init_419eb4,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
 	.mmio = gf108_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
index 4c21cbe..6496411 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
@@ -96,6 +96,7 @@ gf110_gr = {
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_419eb4 = gf100_gr_init_419eb4,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
 	.mmio = gf110_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
index 10340e9f..21aa8e2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
@@ -160,6 +160,7 @@ gf117_gr = {
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_419eb4 = gf100_gr_init_419eb4,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
 	.mmio = gf117_gr_pack_mmio,
 	.fecs.ucode = &gf117_gr_fecs_ucode,
 	.gpccs.ucode = &gf117_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
index 5147c59..7d73a9f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
@@ -187,6 +187,7 @@ gf119_gr = {
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_419eb4 = gf100_gr_init_419eb4,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
 	.mmio = gf119_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
index 18051963..2073531 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
@@ -475,8 +475,7 @@ gk104_gr_init(struct gf100_gr *gr)
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff);
 			gr->func->init_tex_hww_esr(gr, gpc, tpc);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x001ffffe);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x0000000f);
+			gr->func->init_shader_exceptions(gr, gpc, tpc);
 		}
 		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c90), 0xffffffff);
 		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c94), 0xffffffff);
@@ -537,6 +536,7 @@ gk104_gr = {
 	.init_419eb4 = gf100_gr_init_419eb4,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
 	.mmio = gk104_gr_pack_mmio,
 	.fecs.ucode = &gk104_gr_fecs_ucode,
 	.gpccs.ucode = &gk104_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
index f4c73a9..e52cc24 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
@@ -362,6 +362,7 @@ gk110_gr = {
 	.init_419eb4 = gk110_gr_init_419eb4,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
 	.mmio = gk110_gr_pack_mmio,
 	.fecs.ucode = &gk110_gr_fecs_ucode,
 	.gpccs.ucode = &gk110_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
index b585ab2..3530381 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
@@ -114,6 +114,7 @@ gk110b_gr = {
 	.init_419eb4 = gk110_gr_init_419eb4,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
 	.mmio = gk110b_gr_pack_mmio,
 	.fecs.ucode = &gk110_gr_fecs_ucode,
 	.gpccs.ucode = &gk110_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
index 7d1f585..4e5cd54 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
@@ -172,6 +172,7 @@ gk208_gr = {
 	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
 	.mmio = gk208_gr_pack_mmio,
 	.fecs.ucode = &gk208_gr_fecs_ucode,
 	.gpccs.ucode = &gk208_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
index 4c8c256..3b8ae13 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
@@ -282,6 +282,14 @@ gm107_gr_pack_mmio[] = {
  ******************************************************************************/
 
 void
+gm107_gr_init_shader_exceptions(struct gf100_gr *gr, int gpc, int tpc)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x00dffffe);
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x00000005);
+}
+
+void
 gm107_gr_init_504430(struct gf100_gr *gr, int gpc, int tpc)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -403,8 +411,7 @@ gm107_gr_init(struct gf100_gr *gr)
 			gr->func->init_tex_hww_esr(gr, gpc, tpc);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
 			gr->func->init_504430(gr, gpc, tpc);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x00dffffe);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x00000005);
+			gr->func->init_shader_exceptions(gr, gpc, tpc);
 		}
 		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c90), 0xffffffff);
 		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c94), 0xffffffff);
@@ -467,6 +474,7 @@ gm107_gr = {
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_504430 = gm107_gr_init_504430,
+	.init_shader_exceptions = gm107_gr_init_shader_exceptions,
 	.mmio = gm107_gr_pack_mmio,
 	.fecs.ucode = &gm107_gr_fecs_ucode,
 	.gpccs.ucode = &gm107_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
index 69f191e..3e2a78d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
@@ -124,8 +124,7 @@ gm200_gr_init(struct gf100_gr *gr)
 			gr->func->init_tex_hww_esr(gr, gpc, tpc);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
 			gr->func->init_504430(gr, gpc, tpc);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x00dffffe);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x00000005);
+			gr->func->init_shader_exceptions(gr, gpc, tpc);
 		}
 		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c90), 0xffffffff);
 		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c94), 0xffffffff);
@@ -206,6 +205,7 @@ gm200_gr = {
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_504430 = gm107_gr_init_504430,
+	.init_shader_exceptions = gm107_gr_init_shader_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 2,
 	.grctx = &gm200_grctx,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
index 76a9416..48a0e14 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
@@ -30,6 +30,14 @@
  * PGRAPH engine/subdev functions
  ******************************************************************************/
 
+void
+gp100_gr_init_shader_exceptions(struct gf100_gr *gr, int gpc, int tpc)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x00dffffe);
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x00000105);
+}
+
 static void
 gp100_gr_init_419c9c(struct gf100_gr *gr)
 {
@@ -104,8 +112,7 @@ gp100_gr_init(struct gf100_gr *gr)
 			gr->func->init_tex_hww_esr(gr, gpc, tpc);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
 			gr->func->init_504430(gr, gpc, gpc);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x00dffffe);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x00000105);
+			gr->func->init_shader_exceptions(gr, gpc, tpc);
 		}
 		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c90), 0xffffffff);
 		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c94), 0xffffffff);
@@ -146,6 +153,7 @@ gp100_gr = {
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_504430 = gm107_gr_init_504430,
+	.init_shader_exceptions = gp100_gr_init_shader_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 2,
 	.grctx = &gp100_grctx,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
index 7072578..7ce0652 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
@@ -56,6 +56,7 @@ gp102_gr = {
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_504430 = gm107_gr_init_504430,
+	.init_shader_exceptions = gp100_gr_init_shader_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 3,
 	.grctx = &gp102_grctx,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
index f23f70f..dbb2d4e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
@@ -42,6 +42,7 @@ gp107_gr = {
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_504430 = gm107_gr_init_504430,
+	.init_shader_exceptions = gp100_gr_init_shader_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 1,
 	.grctx = &gp107_grctx,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
index 0ff1759..7f23d8b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
@@ -40,6 +40,7 @@ gp10b_gr = {
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_504430 = gm107_gr_init_504430,
+	.init_shader_exceptions = gp100_gr_init_shader_exceptions,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 1,
 	.grctx = &gp102_grctx,
-- 
cgit v1.1


From 6df6d2b95e9411f499b7d71a6cc495cff7548d6e Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: apply be exception fixes from traces

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c | 4 ++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c | 4 ++--
 2 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 071b3c1..452b527 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -2109,8 +2109,8 @@ gf100_gr_init(struct gf100_gr *gr)
 	}
 
 	for (rop = 0; rop < gr->rop_nr; rop++) {
-		nvkm_wr32(device, ROP_UNIT(rop, 0x144), 0xc0000000);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x070), 0xc0000000);
+		nvkm_wr32(device, ROP_UNIT(rop, 0x144), 0x40000000);
+		nvkm_wr32(device, ROP_UNIT(rop, 0x070), 0x40000000);
 		nvkm_wr32(device, ROP_UNIT(rop, 0x204), 0xffffffff);
 		nvkm_wr32(device, ROP_UNIT(rop, 0x208), 0xffffffff);
 	}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
index 2073531..1822509 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
@@ -482,8 +482,8 @@ gk104_gr_init(struct gf100_gr *gr)
 	}
 
 	for (rop = 0; rop < gr->rop_nr; rop++) {
-		nvkm_wr32(device, ROP_UNIT(rop, 0x144), 0xc0000000);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x070), 0xc0000000);
+		nvkm_wr32(device, ROP_UNIT(rop, 0x144), 0x40000000);
+		nvkm_wr32(device, ROP_UNIT(rop, 0x070), 0x40000000);
 		nvkm_wr32(device, ROP_UNIT(rop, 0x204), 0xffffffff);
 		nvkm_wr32(device, ROP_UNIT(rop, 0x208), 0xffffffff);
 	}
-- 
cgit v1.1


From 04547482aed8c77b823de9427c3f0a7b481a351a Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise init_400054

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c  | 10 +++++++++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h  |  3 +++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c  |  3 ++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c  |  9 ++++++++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c  |  3 ++-
 13 files changed, 32 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 452b527..ac4c0d8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -1915,6 +1915,12 @@ gf100_gr_new_(const struct gf100_gr_func *func, struct nvkm_device *device,
 }
 
 void
+gf100_gr_init_400054(struct gf100_gr *gr)
+{
+	nvkm_wr32(gr->base.engine.subdev.device, 0x400054, 0x34ce3464);
+}
+
+void
 gf100_gr_init_shader_exceptions(struct gf100_gr *gr, int gpc, int tpc)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -2122,7 +2128,8 @@ gf100_gr_init(struct gf100_gr *gr)
 	nvkm_wr32(device, 0x40011c, 0xffffffff);
 	nvkm_wr32(device, 0x400134, 0xffffffff);
 
-	nvkm_wr32(device, 0x400054, 0x34ce3464);
+	if (gr->func->init_400054)
+		gr->func->init_400054(gr);
 
 	gf100_gr_zbc_init(gr);
 
@@ -2162,6 +2169,7 @@ gf100_gr = {
 	.init_419eb4 = gf100_gr_init_419eb4,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
+	.init_400054 = gf100_gr_init_400054,
 	.mmio = gf100_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index 02505c0..9a57f1a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -141,6 +141,7 @@ struct gf100_gr_func {
 	void (*init_tex_hww_esr)(struct gf100_gr *, int gpc, int tpc);
 	void (*init_504430)(struct gf100_gr *, int gpc, int tpc);
 	void (*init_shader_exceptions)(struct gf100_gr *, int gpc, int tpc);
+	void (*init_400054)(struct gf100_gr *);
 	void (*set_hww_esr_report_mask)(struct gf100_gr *);
 	const struct gf100_gr_pack *mmio;
 	struct {
@@ -167,6 +168,7 @@ void gf100_gr_init_419cc0(struct gf100_gr *);
 void gf100_gr_init_419eb4(struct gf100_gr *);
 void gf100_gr_init_tex_hww_esr(struct gf100_gr *, int, int);
 void gf100_gr_init_shader_exceptions(struct gf100_gr *, int, int);
+void gf100_gr_init_400054(struct gf100_gr *);
 
 void gf117_gr_init_zcull(struct gf100_gr *);
 
@@ -180,6 +182,7 @@ void gk110_gr_init_419eb4(struct gf100_gr *);
 
 void gm107_gr_init_504430(struct gf100_gr *, int, int);
 void gm107_gr_init_shader_exceptions(struct gf100_gr *, int, int);
+void gm107_gr_init_400054(struct gf100_gr *);
 
 int gk20a_gr_init(struct gf100_gr *);
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
index c615444..d5276ca 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
@@ -125,6 +125,7 @@ gf104_gr = {
 	.init_419eb4 = gf100_gr_init_419eb4,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
+	.init_400054 = gf100_gr_init_400054,
 	.mmio = gf104_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
index 8e8a99e..8f22a31 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
@@ -123,6 +123,7 @@ gf108_gr = {
 	.init_419eb4 = gf100_gr_init_419eb4,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
+	.init_400054 = gf100_gr_init_400054,
 	.mmio = gf108_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
index 6496411..fcbfdc7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
@@ -97,6 +97,7 @@ gf110_gr = {
 	.init_419eb4 = gf100_gr_init_419eb4,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
+	.init_400054 = gf100_gr_init_400054,
 	.mmio = gf110_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
index 21aa8e2..f526ccd 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
@@ -161,6 +161,7 @@ gf117_gr = {
 	.init_419eb4 = gf100_gr_init_419eb4,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
+	.init_400054 = gf100_gr_init_400054,
 	.mmio = gf117_gr_pack_mmio,
 	.fecs.ucode = &gf117_gr_fecs_ucode,
 	.gpccs.ucode = &gf117_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
index 7d73a9f..5d3d22f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
@@ -188,6 +188,7 @@ gf119_gr = {
 	.init_419eb4 = gf100_gr_init_419eb4,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
+	.init_400054 = gf100_gr_init_400054,
 	.mmio = gf119_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
index 1822509..36f32e6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
@@ -495,7 +495,7 @@ gk104_gr_init(struct gf100_gr *gr)
 	nvkm_wr32(device, 0x40011c, 0xffffffff);
 	nvkm_wr32(device, 0x400134, 0xffffffff);
 
-	nvkm_wr32(device, 0x400054, 0x34ce3464);
+	gr->func->init_400054(gr);
 
 	gf100_gr_zbc_init(gr);
 
@@ -537,6 +537,7 @@ gk104_gr = {
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
+	.init_400054 = gf100_gr_init_400054,
 	.mmio = gk104_gr_pack_mmio,
 	.fecs.ucode = &gk104_gr_fecs_ucode,
 	.gpccs.ucode = &gk104_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
index e52cc24..7bcea81 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
@@ -363,6 +363,7 @@ gk110_gr = {
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
+	.init_400054 = gf100_gr_init_400054,
 	.mmio = gk110_gr_pack_mmio,
 	.fecs.ucode = &gk110_gr_fecs_ucode,
 	.gpccs.ucode = &gk110_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
index 3530381..5af2a54 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
@@ -115,6 +115,7 @@ gk110b_gr = {
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
+	.init_400054 = gf100_gr_init_400054,
 	.mmio = gk110b_gr_pack_mmio,
 	.fecs.ucode = &gk110_gr_fecs_ucode,
 	.gpccs.ucode = &gk110_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
index 4e5cd54..dd08373 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
@@ -173,6 +173,7 @@ gk208_gr = {
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
+	.init_400054 = gf100_gr_init_400054,
 	.mmio = gk208_gr_pack_mmio,
 	.fecs.ucode = &gk208_gr_fecs_ucode,
 	.gpccs.ucode = &gk208_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
index 3b8ae13..4680d99 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
@@ -282,6 +282,12 @@ gm107_gr_pack_mmio[] = {
  ******************************************************************************/
 
 void
+gm107_gr_init_400054(struct gf100_gr *gr)
+{
+	nvkm_wr32(gr->base.engine.subdev.device, 0x400054, 0x2c350f63);
+}
+
+void
 gm107_gr_init_shader_exceptions(struct gf100_gr *gr, int gpc, int tpc)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -431,7 +437,7 @@ gm107_gr_init(struct gf100_gr *gr)
 	nvkm_wr32(device, 0x40011c, 0xffffffff);
 	nvkm_wr32(device, 0x400134, 0xffffffff);
 
-	nvkm_wr32(device, 0x400054, 0x2c350f63);
+	gr->func->init_400054(gr);
 
 	gf100_gr_zbc_init(gr);
 
@@ -475,6 +481,7 @@ gm107_gr = {
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_504430 = gm107_gr_init_504430,
 	.init_shader_exceptions = gm107_gr_init_shader_exceptions,
+	.init_400054 = gm107_gr_init_400054,
 	.mmio = gm107_gr_pack_mmio,
 	.fecs.ucode = &gm107_gr_fecs_ucode,
 	.gpccs.ucode = &gm107_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
index 3e2a78d..bbe1887 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
@@ -144,7 +144,7 @@ gm200_gr_init(struct gf100_gr *gr)
 	nvkm_wr32(device, 0x40011c, 0xffffffff);
 	nvkm_wr32(device, 0x400134, 0xffffffff);
 
-	nvkm_wr32(device, 0x400054, 0x2c350f63);
+	gr->func->init_400054(gr);
 
 	gf100_gr_zbc_init(gr);
 
@@ -206,6 +206,7 @@ gm200_gr = {
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_504430 = gm107_gr_init_504430,
 	.init_shader_exceptions = gm107_gr_init_shader_exceptions,
+	.init_400054 = gm107_gr_init_400054,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 2,
 	.grctx = &gm200_grctx,
-- 
cgit v1.1


From 525230cb204db5edb0ffc42e324612809c663c75 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: delete duplicated init code

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h  |  2 -
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c  | 78 +------------------------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c  |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c  |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c  | 77 +-----------------------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c  | 76 +-----------------------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c  | 77 +-----------------------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c  |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c  |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c  |  2 +-
 11 files changed, 10 insertions(+), 312 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index 9a57f1a..9ed51dc 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -172,7 +172,6 @@ void gf100_gr_init_400054(struct gf100_gr *);
 
 void gf117_gr_init_zcull(struct gf100_gr *);
 
-int gk104_gr_init(struct gf100_gr *);
 void gk104_gr_init_vsc_stream_master(struct gf100_gr *);
 void gk104_gr_init_rop_active_fbps(struct gf100_gr *);
 void gk104_gr_init_ppc_exceptions(struct gf100_gr *);
@@ -190,7 +189,6 @@ int gm200_gr_rops(struct gf100_gr *);
 void gm200_gr_init_num_active_ltcs(struct gf100_gr *);
 void gm200_gr_init_ds_hww_esr_2(struct gf100_gr *);
 
-int gp100_gr_init(struct gf100_gr *);
 void gp100_gr_init_rop_active_fbps(struct gf100_gr *);
 void gp100_gr_init_fecs_exceptions(struct gf100_gr *);
 void gp100_gr_init_shader_exceptions(struct gf100_gr *, int, int);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
index 36f32e6..2e5c48b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
@@ -426,82 +426,6 @@ gk104_gr_init_vsc_stream_master(struct gf100_gr *gr)
 	nvkm_wr32(device, GPC_UNIT(0, 0x3018), 0x00000001);
 }
 
-int
-gk104_gr_init(struct gf100_gr *gr)
-{
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-	int gpc, tpc, rop;
-
-	gr->func->init_gpc_mmu(gr);
-
-	gf100_gr_mmio(gr, gr->func->mmio);
-	if (gr->func->clkgate_pack)
-		nvkm_therm_clkgate_init(gr->base.engine.subdev.device->therm,
-					gr->func->clkgate_pack);
-
-	gr->func->init_vsc_stream_master(gr);
-	gr->func->init_zcull(gr);
-	gr->func->init_num_active_ltcs(gr);
-
-	gr->func->init_rop_active_fbps(gr);
-
-	nvkm_wr32(device, 0x400500, 0x00010001);
-
-	nvkm_wr32(device, 0x400100, 0xffffffff);
-	nvkm_wr32(device, 0x40013c, 0xffffffff);
-	nvkm_wr32(device, 0x400124, 0x00000002);
-
-	gr->func->init_fecs_exceptions(gr);
-	nvkm_wr32(device, 0x404000, 0xc0000000);
-	nvkm_wr32(device, 0x404600, 0xc0000000);
-	nvkm_wr32(device, 0x408030, 0xc0000000);
-	nvkm_wr32(device, 0x404490, 0xc0000000);
-	nvkm_wr32(device, 0x406018, 0xc0000000);
-	gr->func->init_sked_hww_esr(gr);
-	nvkm_wr32(device, 0x405840, 0xc0000000);
-	nvkm_wr32(device, 0x405844, 0x00ffffff);
-	gr->func->init_419cc0(gr);
-	gr->func->init_419eb4(gr);
-
-	gr->func->init_ppc_exceptions(gr);
-
-	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0420), 0xc0000000);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0900), 0xc0000000);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x1028), 0xc0000000);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0824), 0xc0000000);
-		for (tpc = 0; tpc < gr->tpc_nr[gpc]; tpc++) {
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x508), 0xffffffff);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff);
-			gr->func->init_tex_hww_esr(gr, gpc, tpc);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
-			gr->func->init_shader_exceptions(gr, gpc, tpc);
-		}
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c90), 0xffffffff);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c94), 0xffffffff);
-	}
-
-	for (rop = 0; rop < gr->rop_nr; rop++) {
-		nvkm_wr32(device, ROP_UNIT(rop, 0x144), 0x40000000);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x070), 0x40000000);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x204), 0xffffffff);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x208), 0xffffffff);
-	}
-
-	nvkm_wr32(device, 0x400108, 0xffffffff);
-	nvkm_wr32(device, 0x400138, 0xffffffff);
-	nvkm_wr32(device, 0x400118, 0xffffffff);
-	nvkm_wr32(device, 0x400130, 0xffffffff);
-	nvkm_wr32(device, 0x40011c, 0xffffffff);
-	nvkm_wr32(device, 0x400134, 0xffffffff);
-
-	gr->func->init_400054(gr);
-
-	gf100_gr_zbc_init(gr);
-
-	return gf100_gr_init_ctxctl(gr);
-}
-
 #include "fuc/hubgk104.fuc3.h"
 
 static struct gf100_gr_ucode
@@ -524,7 +448,7 @@ gk104_gr_gpccs_ucode = {
 
 static const struct gf100_gr_func
 gk104_gr = {
-	.init = gk104_gr_init,
+	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_zcull = gf117_gr_init_zcull,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
index 7bcea81..9adb55f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
@@ -350,7 +350,7 @@ gk110_gr_init_419eb4(struct gf100_gr *gr)
 
 static const struct gf100_gr_func
 gk110_gr = {
-	.init = gk104_gr_init,
+	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_zcull = gf117_gr_init_zcull,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
index 5af2a54..f848f15 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
@@ -102,7 +102,7 @@ gk110b_gr_pack_mmio[] = {
 
 static const struct gf100_gr_func
 gk110b_gr = {
-	.init = gk104_gr_init,
+	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_zcull = gf117_gr_init_zcull,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
index dd08373..9c678f1 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
@@ -161,7 +161,7 @@ gk208_gr_gpccs_ucode = {
 
 static const struct gf100_gr_func
 gk208_gr = {
-	.init = gk104_gr_init,
+	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_zcull = gf117_gr_init_zcull,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
index 4680d99..c598fa5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
@@ -369,81 +369,6 @@ gm107_gr_init_gpc_mmu(struct gf100_gr *gr)
 	nvkm_wr32(device, GPC_BCAST(0x08b8), nvkm_memory_addr(fb->mmu_rd) >> 8);
 }
 
-static int
-gm107_gr_init(struct gf100_gr *gr)
-{
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-	int gpc, tpc, rop;
-
-	gr->func->init_gpc_mmu(gr);
-
-	gf100_gr_mmio(gr, gr->func->mmio);
-
-	gr->func->init_bios(gr);
-
-	gr->func->init_vsc_stream_master(gr);
-	gr->func->init_zcull(gr);
-	gr->func->init_num_active_ltcs(gr);
-
-	gr->func->init_rop_active_fbps(gr);
-
-	nvkm_wr32(device, 0x400500, 0x00010001);
-
-	nvkm_wr32(device, 0x400100, 0xffffffff);
-	nvkm_wr32(device, 0x40013c, 0xffffffff);
-	nvkm_wr32(device, 0x400124, 0x00000002);
-	gr->func->init_fecs_exceptions(gr);
-
-	nvkm_wr32(device, 0x404000, 0xc0000000);
-	nvkm_wr32(device, 0x404600, 0xc0000000);
-	nvkm_wr32(device, 0x408030, 0xc0000000);
-	nvkm_wr32(device, 0x404490, 0xc0000000);
-	nvkm_wr32(device, 0x406018, 0xc0000000);
-	gr->func->init_sked_hww_esr(gr);
-	nvkm_wr32(device, 0x405840, 0xc0000000);
-	nvkm_wr32(device, 0x405844, 0x00ffffff);
-	gr->func->init_419cc0(gr);
-
-	gr->func->init_ppc_exceptions(gr);
-
-	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0420), 0xc0000000);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0900), 0xc0000000);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x1028), 0xc0000000);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0824), 0xc0000000);
-		for (tpc = 0; tpc < gr->tpc_nr[gpc]; tpc++) {
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x508), 0xffffffff);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff);
-			gr->func->init_tex_hww_esr(gr, gpc, tpc);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
-			gr->func->init_504430(gr, gpc, tpc);
-			gr->func->init_shader_exceptions(gr, gpc, tpc);
-		}
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c90), 0xffffffff);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c94), 0xffffffff);
-	}
-
-	for (rop = 0; rop < gr->rop_nr; rop++) {
-		nvkm_wr32(device, ROP_UNIT(rop, 0x144), 0x40000000);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x070), 0x40000000);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x204), 0xffffffff);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x208), 0xffffffff);
-	}
-
-	nvkm_wr32(device, 0x400108, 0xffffffff);
-	nvkm_wr32(device, 0x400138, 0xffffffff);
-	nvkm_wr32(device, 0x400118, 0xffffffff);
-	nvkm_wr32(device, 0x400130, 0xffffffff);
-	nvkm_wr32(device, 0x40011c, 0xffffffff);
-	nvkm_wr32(device, 0x400134, 0xffffffff);
-
-	gr->func->init_400054(gr);
-
-	gf100_gr_zbc_init(gr);
-
-	return gf100_gr_init_ctxctl(gr);
-}
-
 #include "fuc/hubgm107.fuc5.h"
 
 static struct gf100_gr_ucode
@@ -466,7 +391,7 @@ gm107_gr_gpccs_ucode = {
 
 static const struct gf100_gr_func
 gm107_gr = {
-	.init = gm107_gr_init,
+	.init = gf100_gr_init,
 	.init_gpc_mmu = gm107_gr_init_gpc_mmu,
 	.init_bios = gm107_gr_init_bios,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
index bbe1887..4dcb56b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
@@ -77,80 +77,6 @@ gm200_gr_init_rop_active_fbps(struct gf100_gr *gr)
 	nvkm_mask(device, 0x408958, 0x0000000f, fbp_count); /* crop */
 }
 
-static int
-gm200_gr_init(struct gf100_gr *gr)
-{
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-	int gpc, tpc, rop;
-
-	gr->func->init_gpc_mmu(gr);
-
-	gf100_gr_mmio(gr, gr->fuc_sw_nonctx);
-
-	gr->func->init_bios(gr);
-
-	gr->func->init_vsc_stream_master(gr);
-	gr->func->init_zcull(gr);
-	gr->func->init_num_active_ltcs(gr);
-
-	gr->func->init_rop_active_fbps(gr);
-
-	nvkm_wr32(device, 0x400500, 0x00010001);
-	nvkm_wr32(device, 0x400100, 0xffffffff);
-	nvkm_wr32(device, 0x40013c, 0xffffffff);
-	nvkm_wr32(device, 0x400124, 0x00000002);
-	gr->func->init_fecs_exceptions(gr);
-	gr->func->init_ds_hww_esr_2(gr);
-	nvkm_wr32(device, 0x404000, 0xc0000000);
-	nvkm_wr32(device, 0x404600, 0xc0000000);
-	nvkm_wr32(device, 0x408030, 0xc0000000);
-	nvkm_wr32(device, 0x404490, 0xc0000000);
-	nvkm_wr32(device, 0x406018, 0xc0000000);
-	gr->func->init_sked_hww_esr(gr);
-	nvkm_wr32(device, 0x405840, 0xc0000000);
-	nvkm_wr32(device, 0x405844, 0x00ffffff);
-	gr->func->init_419cc0(gr);
-
-	gr->func->init_ppc_exceptions(gr);
-
-	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0420), 0xc0000000);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0900), 0xc0000000);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x1028), 0xc0000000);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0824), 0xc0000000);
-		for (tpc = 0; tpc < gr->tpc_nr[gpc]; tpc++) {
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x508), 0xffffffff);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff);
-			gr->func->init_tex_hww_esr(gr, gpc, tpc);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
-			gr->func->init_504430(gr, gpc, tpc);
-			gr->func->init_shader_exceptions(gr, gpc, tpc);
-		}
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c90), 0xffffffff);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c94), 0xffffffff);
-	}
-
-	for (rop = 0; rop < gr->rop_nr; rop++) {
-		nvkm_wr32(device, ROP_UNIT(rop, 0x144), 0x40000000);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x070), 0x40000000);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x204), 0xffffffff);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x208), 0xffffffff);
-	}
-
-	nvkm_wr32(device, 0x400108, 0xffffffff);
-	nvkm_wr32(device, 0x400138, 0xffffffff);
-	nvkm_wr32(device, 0x400118, 0xffffffff);
-	nvkm_wr32(device, 0x400130, 0xffffffff);
-	nvkm_wr32(device, 0x40011c, 0xffffffff);
-	nvkm_wr32(device, 0x400134, 0xffffffff);
-
-	gr->func->init_400054(gr);
-
-	gf100_gr_zbc_init(gr);
-
-	return gf100_gr_init_ctxctl(gr);
-}
-
 int
 gm200_gr_new_(const struct gf100_gr_func *func, struct nvkm_device *device,
 	      int index, struct nvkm_gr **pgr)
@@ -191,7 +117,7 @@ gm200_gr_new_(const struct gf100_gr_func *func, struct nvkm_device *device,
 
 static const struct gf100_gr_func
 gm200_gr = {
-	.init = gm200_gr_init,
+	.init = gf100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_bios = gm107_gr_init_bios,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
index 48a0e14..cc507e8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
@@ -62,84 +62,9 @@ gp100_gr_init_rop_active_fbps(struct gf100_gr *gr)
 	nvkm_mask(device, 0x408958, 0x0000000f, fbp_count); /* crop */
 }
 
-int
-gp100_gr_init(struct gf100_gr *gr)
-{
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-	int gpc, tpc, rop;
-
-	gr->func->init_gpc_mmu(gr);
-
-	gf100_gr_mmio(gr, gr->fuc_sw_nonctx);
-
-	gr->func->init_vsc_stream_master(gr);
-	gr->func->init_zcull(gr);
-
-	gr->func->init_num_active_ltcs(gr);
-
-	gr->func->init_rop_active_fbps(gr);
-	if (gr->func->init_swdx_pes_mask)
-		gr->func->init_swdx_pes_mask(gr);
-
-	nvkm_wr32(device, 0x400500, 0x00010001);
-	nvkm_wr32(device, 0x400100, 0xffffffff);
-	nvkm_wr32(device, 0x40013c, 0xffffffff);
-	nvkm_wr32(device, 0x400124, 0x00000002);
-	gr->func->init_fecs_exceptions(gr);
-	gr->func->init_ds_hww_esr_2(gr);
-	nvkm_wr32(device, 0x404000, 0xc0000000);
-	nvkm_wr32(device, 0x404600, 0xc0000000);
-	nvkm_wr32(device, 0x408030, 0xc0000000);
-	nvkm_wr32(device, 0x404490, 0xc0000000);
-	nvkm_wr32(device, 0x406018, 0xc0000000);
-	gr->func->init_sked_hww_esr(gr);
-	nvkm_wr32(device, 0x405840, 0xc0000000);
-	nvkm_wr32(device, 0x405844, 0x00ffffff);
-	gr->func->init_419cc0(gr);
-	if (gr->func->init_419c9c)
-		gr->func->init_419c9c(gr);
-
-	gr->func->init_ppc_exceptions(gr);
-
-	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0420), 0xc0000000);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0900), 0xc0000000);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x1028), 0xc0000000);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0824), 0xc0000000);
-		for (tpc = 0; tpc < gr->tpc_nr[gpc]; tpc++) {
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x508), 0xffffffff);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff);
-			gr->func->init_tex_hww_esr(gr, gpc, tpc);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
-			gr->func->init_504430(gr, gpc, gpc);
-			gr->func->init_shader_exceptions(gr, gpc, tpc);
-		}
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c90), 0xffffffff);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c94), 0xffffffff);
-	}
-
-	for (rop = 0; rop < gr->rop_nr; rop++) {
-		nvkm_wr32(device, ROP_UNIT(rop, 0x144), 0x40000000);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x070), 0x40000000);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x204), 0xffffffff);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x208), 0xffffffff);
-	}
-
-	nvkm_wr32(device, 0x400108, 0xffffffff);
-	nvkm_wr32(device, 0x400138, 0xffffffff);
-	nvkm_wr32(device, 0x400118, 0xffffffff);
-	nvkm_wr32(device, 0x400130, 0xffffffff);
-	nvkm_wr32(device, 0x40011c, 0xffffffff);
-	nvkm_wr32(device, 0x400134, 0xffffffff);
-
-	gf100_gr_zbc_init(gr);
-
-	return gf100_gr_init_ctxctl(gr);
-}
-
 static const struct gf100_gr_func
 gp100_gr = {
-	.init = gp100_gr_init,
+	.init = gf100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_zcull = gf117_gr_init_zcull,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
index 7ce0652..86d1ff7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
@@ -42,7 +42,7 @@ gp102_gr_init_swdx_pes_mask(struct gf100_gr *gr)
 
 static const struct gf100_gr_func
 gp102_gr = {
-	.init = gp100_gr_init,
+	.init = gf100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_zcull = gf117_gr_init_zcull,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
index dbb2d4e..14007b5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
@@ -28,7 +28,7 @@
 
 static const struct gf100_gr_func
 gp107_gr = {
-	.init = gp100_gr_init,
+	.init = gf100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_zcull = gf117_gr_init_zcull,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
index 7f23d8b..450a96d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
@@ -27,7 +27,7 @@
 
 static const struct gf100_gr_func
 gp10b_gr = {
-	.init = gp100_gr_init,
+	.init = gf100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
 	.init_zcull = gf117_gr_init_zcull,
-- 
cgit v1.1


From aedc49fd0ebc2aaca2176aae27f170224b139b15 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: support firmware-provided sw_ctx everywhere

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c | 14 +++++++++-----
 1 file changed, 9 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index 8810150..e2c8077 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1239,11 +1239,15 @@ gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	nvkm_mc_unk260(device, 0);
 
-	gf100_gr_mmio(gr, grctx->hub);
-	gf100_gr_mmio(gr, grctx->gpc);
-	gf100_gr_mmio(gr, grctx->zcull);
-	gf100_gr_mmio(gr, grctx->tpc);
-	gf100_gr_mmio(gr, grctx->ppc);
+	if (!gr->fuc_sw_ctx) {
+		gf100_gr_mmio(gr, grctx->hub);
+		gf100_gr_mmio(gr, grctx->gpc);
+		gf100_gr_mmio(gr, grctx->zcull);
+		gf100_gr_mmio(gr, grctx->tpc);
+		gf100_gr_mmio(gr, grctx->ppc);
+	} else {
+		gf100_gr_mmio(gr, gr->fuc_sw_ctx);
+	}
 
 	idle_timeout = nvkm_mask(device, 0x404154, 0xffffffff, 0x00000000);
 
-- 
cgit v1.1


From ea4a2bb530464f9bbd2728a3c3c58dd758fb36c4 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise patch_ltc, noting missing init

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c  |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h  |  3 +++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c  | 12 ++++++++++++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c  |  1 +
 6 files changed, 20 insertions(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index e2c8077..4500599 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1254,6 +1254,8 @@ gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->bundle(info);
 	grctx->pagepool(info);
 	grctx->attrib(info);
+	if (grctx->patch_ltc)
+		grctx->patch_ltc(info);
 	grctx->unkn(gr);
 
 	gf100_grctx_generate_tpcid(gr);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
index 5199e5a..94612ca 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
@@ -48,6 +48,8 @@ struct gf100_grctx_func {
 	u32 attrib_nr;
 	u32 alpha_nr_max;
 	u32 alpha_nr;
+	/* other patch buffer stuff */
+	void (*patch_ltc)(struct gf100_grctx *);
 };
 
 extern const struct gf100_grctx_func gf100_grctx;
@@ -80,6 +82,7 @@ extern const struct gf100_grctx_func gk20a_grctx;
 void gk104_grctx_generate_main(struct gf100_gr *, struct gf100_grctx *);
 void gk104_grctx_generate_bundle(struct gf100_grctx *);
 void gk104_grctx_generate_pagepool(struct gf100_grctx *);
+void gk104_grctx_generate_patch_ltc(struct gf100_grctx *);
 void gk104_grctx_generate_unkn(struct gf100_gr *);
 void gk104_grctx_generate_r418bb8(struct gf100_gr *);
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
index 825c8fd..da019c4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
@@ -841,6 +841,17 @@ gk104_grctx_pack_ppc[] = {
  ******************************************************************************/
 
 void
+gk104_grctx_generate_patch_ltc(struct gf100_grctx *info)
+{
+	struct nvkm_device *device = info->gr->base.engine.subdev.device;
+	u32 data0 = nvkm_rd32(device, 0x17e91c);
+	u32 data1 = nvkm_rd32(device, 0x17e920);
+	/*XXX: Figure out how to modify this correctly! */
+	mmio_wr32(info, 0x17e91c, data0);
+	mmio_wr32(info, 0x17e920, data1);
+}
+
+void
 gk104_grctx_generate_bundle(struct gf100_grctx *info)
 {
 	const struct gf100_grctx_func *grctx = info->gr->func->grctx;
@@ -1005,4 +1016,5 @@ gk104_grctx = {
 	.attrib_nr = 0x218,
 	.alpha_nr_max = 0x7ff,
 	.alpha_nr = 0x648,
+	.patch_ltc = gk104_grctx_generate_patch_ltc,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
index 7b95ec2..3ad98cd 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
@@ -830,4 +830,5 @@ gk110_grctx = {
 	.attrib_nr = 0x218,
 	.alpha_nr_max = 0x7ff,
 	.alpha_nr = 0x648,
+	.patch_ltc = gk104_grctx_generate_patch_ltc,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
index 048b115..47a4a07 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
@@ -91,4 +91,5 @@ gk110b_grctx = {
 	.attrib_nr = 0x218,
 	.alpha_nr_max = 0x7ff,
 	.alpha_nr = 0x648,
+	.patch_ltc = gk104_grctx_generate_patch_ltc,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
index 67b7a1b..ba04c86 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
@@ -552,4 +552,5 @@ gk208_grctx = {
 	.attrib_nr = 0x218,
 	.alpha_nr_max = 0x7ff,
 	.alpha_nr = 0x648,
+	.patch_ltc = gk104_grctx_generate_patch_ltc,
 };
-- 
cgit v1.1


From fc740f545d912b32e26f12e240270e1dc36fa26e Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise sm_id/tpc_nr

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c  | 60 +++++++++++++---------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h  |  9 +++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c  |  2 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c  |  2 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c  |  2 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c  |  4 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c  |  2 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c  |  4 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c  |  2 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c |  2 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c  |  2 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk20a.c  |  4 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c  | 26 +++-------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c  | 21 +-------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm20b.c  |  3 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c  |  3 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c     |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h     |  1 +
 20 files changed, 85 insertions(+), 67 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index 4500599..14e6bf0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1080,28 +1080,6 @@ gf100_grctx_generate_unkn(struct gf100_gr *gr)
 }
 
 void
-gf100_grctx_generate_tpcid(struct gf100_gr *gr)
-{
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-	int gpc, tpc, id;
-
-	for (tpc = 0, id = 0; tpc < 4; tpc++) {
-		for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-			if (tpc < gr->tpc_nr[gpc]) {
-				nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x698), id);
-				nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x4e8), id);
-				nvkm_wr32(device, GPC_UNIT(gpc, 0x0c10 + tpc * 4), id);
-				nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x088), id);
-				id++;
-			}
-
-			nvkm_wr32(device, GPC_UNIT(gpc, 0x0c08), gr->tpc_nr[gpc]);
-			nvkm_wr32(device, GPC_UNIT(gpc, 0x0c8c), gr->tpc_nr[gpc]);
-		}
-	}
-}
-
-void
 gf100_grctx_generate_r406028(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -1231,6 +1209,40 @@ gf100_grctx_generate_r406800(struct gf100_gr *gr)
 }
 
 void
+gf100_grctx_generate_tpc_nr(struct gf100_gr *gr, int gpc)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, GPC_UNIT(gpc, 0x0c08), gr->tpc_nr[gpc]);
+	nvkm_wr32(device, GPC_UNIT(gpc, 0x0c8c), gr->tpc_nr[gpc]);
+}
+
+void
+gf100_grctx_generate_sm_id(struct gf100_gr *gr, int gpc, int tpc, int sm)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x698), sm);
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x4e8), sm);
+	nvkm_wr32(device, GPC_UNIT(gpc, 0x0c10 + tpc * 4), sm);
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x088), sm);
+}
+
+void
+gf100_grctx_generate_floorsweep(struct gf100_gr *gr)
+{
+	const struct gf100_grctx_func *func = gr->func->grctx;
+	int tpc, gpc, sm;
+
+	for (tpc = 0, sm = 0; tpc < gr->tpc_max; tpc++) {
+		for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
+			if (tpc < gr->tpc_nr[gpc])
+				func->sm_id(gr, gpc, tpc, sm++);
+			if (func->tpc_nr)
+				func->tpc_nr(gr, gpc);
+		}
+	}
+}
+
+void
 gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -1258,7 +1270,7 @@ gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 		grctx->patch_ltc(info);
 	grctx->unkn(gr);
 
-	gf100_grctx_generate_tpcid(gr);
+	gf100_grctx_generate_floorsweep(gr);
 	gf100_grctx_generate_r406028(gr);
 	gf100_grctx_generate_r4060a8(gr);
 	gf100_grctx_generate_r418bb8(gr);
@@ -1410,4 +1422,6 @@ gf100_grctx = {
 	.attrib = gf100_grctx_generate_attrib,
 	.attrib_nr_max = 0x324,
 	.attrib_nr = 0x218,
+	.sm_id = gf100_grctx_generate_sm_id,
+	.tpc_nr = gf100_grctx_generate_tpc_nr,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
index 94612ca..6fa9b62 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
@@ -50,6 +50,9 @@ struct gf100_grctx_func {
 	u32 alpha_nr;
 	/* other patch buffer stuff */
 	void (*patch_ltc)(struct gf100_grctx *);
+	/* floorsweeping */
+	void (*sm_id)(struct gf100_gr *, int gpc, int tpc, int sm);
+	void (*tpc_nr)(struct gf100_gr *, int gpc);
 };
 
 extern const struct gf100_grctx_func gf100_grctx;
@@ -59,11 +62,13 @@ void gf100_grctx_generate_bundle(struct gf100_grctx *);
 void gf100_grctx_generate_pagepool(struct gf100_grctx *);
 void gf100_grctx_generate_attrib(struct gf100_grctx *);
 void gf100_grctx_generate_unkn(struct gf100_gr *);
-void gf100_grctx_generate_tpcid(struct gf100_gr *);
+void gf100_grctx_generate_floorsweep(struct gf100_gr *);
 void gf100_grctx_generate_r406028(struct gf100_gr *);
 void gf100_grctx_generate_r4060a8(struct gf100_gr *);
 void gf100_grctx_generate_r418bb8(struct gf100_gr *);
 void gf100_grctx_generate_r406800(struct gf100_gr *);
+void gf100_grctx_generate_sm_id(struct gf100_gr *, int, int, int);
+void gf100_grctx_generate_tpc_nr(struct gf100_gr *, int);
 
 extern const struct gf100_grctx_func gf108_grctx;
 void gf108_grctx_generate_attrib(struct gf100_grctx *);
@@ -89,6 +94,7 @@ void gk104_grctx_generate_r418bb8(struct gf100_gr *);
 void gm107_grctx_generate_bundle(struct gf100_grctx *);
 void gm107_grctx_generate_pagepool(struct gf100_grctx *);
 void gm107_grctx_generate_attrib(struct gf100_grctx *);
+void gm107_grctx_generate_sm_id(struct gf100_gr *, int, int, int);
 
 extern const struct gf100_grctx_func gk110_grctx;
 extern const struct gf100_grctx_func gk110b_grctx;
@@ -100,7 +106,6 @@ void gm107_grctx_generate_pagepool(struct gf100_grctx *);
 void gm107_grctx_generate_attrib(struct gf100_grctx *);
 
 extern const struct gf100_grctx_func gm200_grctx;
-void gm200_grctx_generate_tpcid(struct gf100_gr *);
 void gm200_grctx_generate_405b60(struct gf100_gr *);
 
 extern const struct gf100_grctx_func gm20b_grctx;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c
index 54fd74e..19cc84e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c
@@ -96,4 +96,6 @@ gf104_grctx = {
 	.attrib = gf100_grctx_generate_attrib,
 	.attrib_nr_max = 0x324,
 	.attrib_nr = 0x218,
+	.sm_id = gf100_grctx_generate_sm_id,
+	.tpc_nr = gf100_grctx_generate_tpc_nr,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c
index 82f71b1..3703731 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c
@@ -794,4 +794,6 @@ gf108_grctx = {
 	.attrib_nr = 0x218,
 	.alpha_nr_max = 0x324,
 	.alpha_nr = 0x218,
+	.sm_id = gf100_grctx_generate_sm_id,
+	.tpc_nr = gf100_grctx_generate_tpc_nr,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c
index 7df398b..6d72ae0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c
@@ -347,4 +347,6 @@ gf110_grctx = {
 	.attrib = gf100_grctx_generate_attrib,
 	.attrib_nr_max = 0x324,
 	.attrib_nr = 0x218,
+	.sm_id = gf100_grctx_generate_sm_id,
+	.tpc_nr = gf100_grctx_generate_tpc_nr,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
index 19301d8..806102a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
@@ -240,7 +240,7 @@ gf117_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->attrib(info);
 	grctx->unkn(gr);
 
-	gf100_grctx_generate_tpcid(gr);
+	gf100_grctx_generate_floorsweep(gr);
 	gf100_grctx_generate_r406028(gr);
 	gf100_grctx_generate_r4060a8(gr);
 	gk104_grctx_generate_r418bb8(gr);
@@ -275,4 +275,6 @@ gf117_grctx = {
 	.attrib_nr = 0x218,
 	.alpha_nr_max = 0x7ff,
 	.alpha_nr = 0x324,
+	.sm_id = gf100_grctx_generate_sm_id,
+	.tpc_nr = gf100_grctx_generate_tpc_nr,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c
index 605185b..784ab94 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c
@@ -517,4 +517,6 @@ gf119_grctx = {
 	.attrib_nr = 0x218,
 	.alpha_nr_max = 0x324,
 	.alpha_nr = 0x218,
+	.sm_id = gf100_grctx_generate_sm_id,
+	.tpc_nr = gf100_grctx_generate_tpc_nr,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
index da019c4..2be72fd 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
@@ -974,7 +974,7 @@ gk104_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->attrib(info);
 	grctx->unkn(gr);
 
-	gf100_grctx_generate_tpcid(gr);
+	gf100_grctx_generate_floorsweep(gr);
 	gf100_grctx_generate_r406028(gr);
 	gk104_grctx_generate_r418bb8(gr);
 	gf100_grctx_generate_r406800(gr);
@@ -1017,4 +1017,6 @@ gk104_grctx = {
 	.alpha_nr_max = 0x7ff,
 	.alpha_nr = 0x648,
 	.patch_ltc = gk104_grctx_generate_patch_ltc,
+	.sm_id = gf100_grctx_generate_sm_id,
+	.tpc_nr = gf100_grctx_generate_tpc_nr,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
index 3ad98cd..5013fc5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
@@ -831,4 +831,6 @@ gk110_grctx = {
 	.alpha_nr_max = 0x7ff,
 	.alpha_nr = 0x648,
 	.patch_ltc = gk104_grctx_generate_patch_ltc,
+	.sm_id = gf100_grctx_generate_sm_id,
+	.tpc_nr = gf100_grctx_generate_tpc_nr,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
index 47a4a07..1dd5742 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
@@ -92,4 +92,6 @@ gk110b_grctx = {
 	.alpha_nr_max = 0x7ff,
 	.alpha_nr = 0x648,
 	.patch_ltc = gk104_grctx_generate_patch_ltc,
+	.sm_id = gf100_grctx_generate_sm_id,
+	.tpc_nr = gf100_grctx_generate_tpc_nr,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
index ba04c86..214f4da 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
@@ -553,4 +553,6 @@ gk208_grctx = {
 	.alpha_nr_max = 0x7ff,
 	.alpha_nr = 0x648,
 	.patch_ltc = gk104_grctx_generate_patch_ltc,
+	.sm_id = gf100_grctx_generate_sm_id,
+	.tpc_nr = gf100_grctx_generate_tpc_nr,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk20a.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk20a.c
index da7c35a..faa5f18 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk20a.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk20a.c
@@ -42,7 +42,7 @@ gk20a_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	grctx->unkn(gr);
 
-	gf100_grctx_generate_tpcid(gr);
+	gf100_grctx_generate_floorsweep(gr);
 	gf100_grctx_generate_r406028(gr);
 	gk104_grctx_generate_r418bb8(gr);
 	gf100_grctx_generate_r406800(gr);
@@ -82,4 +82,6 @@ gk20a_grctx = {
 	.attrib_nr = 0x240,
 	.alpha_nr_max = 0x648 + (0x648 / 2),
 	.alpha_nr = 0x648,
+	.sm_id = gf100_grctx_generate_sm_id,
+	.tpc_nr = gf100_grctx_generate_tpc_nr,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
index 9b43d4c..e5c0273 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
@@ -930,25 +930,13 @@ gm107_grctx_generate_attrib(struct gf100_grctx *info)
 	}
 }
 
-static void
-gm107_grctx_generate_tpcid(struct gf100_gr *gr)
+void
+gm107_grctx_generate_sm_id(struct gf100_gr *gr, int gpc, int tpc, int sm)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
-	int gpc, tpc, id;
-
-	for (tpc = 0, id = 0; tpc < 4; tpc++) {
-		for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-			if (tpc < gr->tpc_nr[gpc]) {
-				nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x698), id);
-				nvkm_wr32(device, GPC_UNIT(gpc, 0x0c10 + tpc * 4), id);
-				nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x088), id);
-				id++;
-			}
-
-			nvkm_wr32(device, GPC_UNIT(gpc, 0x0c08), gr->tpc_nr[gpc]);
-			nvkm_wr32(device, GPC_UNIT(gpc, 0x0c8c), gr->tpc_nr[gpc]);
-		}
-	}
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x698), sm);
+	nvkm_wr32(device, GPC_UNIT(gpc, 0x0c10 + tpc * 4), sm);
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x088), sm);
 }
 
 static void
@@ -972,7 +960,7 @@ gm107_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->attrib(info);
 	grctx->unkn(gr);
 
-	gm107_grctx_generate_tpcid(gr);
+	gf100_grctx_generate_floorsweep(gr);
 	gf100_grctx_generate_r406028(gr);
 	gk104_grctx_generate_r418bb8(gr);
 	gf100_grctx_generate_r406800(gr);
@@ -1016,4 +1004,6 @@ gm107_grctx = {
 	.attrib_nr = 0xaa0,
 	.alpha_nr_max = 0x1800,
 	.alpha_nr = 0x1000,
+	.sm_id = gm107_grctx_generate_sm_id,
+	.tpc_nr = gf100_grctx_generate_tpc_nr,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
index db209d3..be5e25a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
@@ -28,24 +28,6 @@
  ******************************************************************************/
 
 void
-gm200_grctx_generate_tpcid(struct gf100_gr *gr)
-{
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-	int gpc, tpc, id;
-
-	for (tpc = 0, id = 0; tpc < TPC_MAX_PER_GPC; tpc++) {
-		for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-			if (tpc < gr->tpc_nr[gpc]) {
-				nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x698), id);
-				nvkm_wr32(device, GPC_UNIT(gpc, 0x0c10 + tpc * 4), id);
-				nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x088), id);
-				id++;
-			}
-		}
-	}
-}
-
-void
 gm200_grctx_generate_405b60(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -94,7 +76,7 @@ gm200_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->attrib(info);
 	grctx->unkn(gr);
 
-	gm200_grctx_generate_tpcid(gr);
+	gf100_grctx_generate_floorsweep(gr);
 	gf100_grctx_generate_r406028(gr);
 	gk104_grctx_generate_r418bb8(gr);
 
@@ -133,4 +115,5 @@ gm200_grctx = {
 	.attrib_nr = 0x400,
 	.alpha_nr_max = 0x1800,
 	.alpha_nr = 0x1000,
+	.sm_id = gm107_grctx_generate_sm_id,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm20b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm20b.c
index e5702e3..363d198 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm20b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm20b.c
@@ -53,7 +53,7 @@ gm20b_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	grctx->unkn(gr);
 
-	gm200_grctx_generate_tpcid(gr);
+	gf100_grctx_generate_floorsweep(gr);
 	gm20b_grctx_generate_r406028(gr);
 	gk104_grctx_generate_r418bb8(gr);
 
@@ -98,4 +98,5 @@ gm20b_grctx = {
 	.attrib_nr = 0x400,
 	.alpha_nr_max = 0xc00,
 	.alpha_nr = 0x800,
+	.sm_id = gm107_grctx_generate_sm_id,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
index 88ea322..ac8618f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
@@ -138,7 +138,7 @@ gp100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->attrib(info);
 	grctx->unkn(gr);
 
-	gm200_grctx_generate_tpcid(gr);
+	gf100_grctx_generate_floorsweep(gr);
 	gf100_grctx_generate_r406028(gr);
 	gk104_grctx_generate_r418bb8(gr);
 
@@ -174,4 +174,5 @@ gp100_grctx = {
 	.attrib_nr = 0x440,
 	.alpha_nr_max = 0xc00,
 	.alpha_nr = 0x800,
+	.sm_id = gm107_grctx_generate_sm_id,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
index 7a66b4c..a267abc 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
@@ -94,4 +94,5 @@ gp102_grctx = {
 	.attrib_nr = 0x320,
 	.alpha_nr_max = 0xc00,
 	.alpha_nr = 0x800,
+	.sm_id = gm107_grctx_generate_sm_id,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
index 8da91a0..77345b2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
@@ -44,4 +44,5 @@ gp107_grctx = {
 	.attrib_nr = 0x540,
 	.alpha_nr_max = 0xc00,
 	.alpha_nr = 0x800,
+	.sm_id = gm107_grctx_generate_sm_id,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index ac4c0d8..b9519fa 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -1675,6 +1675,7 @@ gf100_gr_oneinit(struct nvkm_gr *base)
 	gr->gpc_nr = nvkm_rd32(device, 0x409604) & 0x0000001f;
 	for (i = 0; i < gr->gpc_nr; i++) {
 		gr->tpc_nr[i]  = nvkm_rd32(device, GPC_UNIT(i, 0x2608));
+		gr->tpc_max = max(gr->tpc_max, gr->tpc_nr[i]);
 		gr->tpc_total += gr->tpc_nr[i];
 		gr->ppc_nr[i]  = gr->func->ppc_nr;
 		for (j = 0; j < gr->ppc_nr[i]; j++) {
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index 9ed51dc..ad352ee 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -99,6 +99,7 @@ struct gf100_gr {
 	u8 rop_nr;
 	u8 gpc_nr;
 	u8 tpc_nr[GPC_MAX];
+	u8 tpc_max;
 	u8 tpc_total;
 	u8 ppc_nr[GPC_MAX];
 	u8 ppc_mask[GPC_MAX];
-- 
cgit v1.1


From e51f75d5012e88c90b3a05b6706475d83cb7a6eb Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise tpc_per_gpc

GM20B now also shares the same code, as NVGPU shows it doesn't need
special treatment.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c | 25 +++++++++--------------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h |  1 -
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c |  1 -
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c |  1 -
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk20a.c |  1 -
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c |  1 -
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c |  1 -
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm20b.c | 15 --------------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c |  1 -
 9 files changed, 10 insertions(+), 37 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index 14e6bf0..a4865f7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1080,19 +1080,6 @@ gf100_grctx_generate_unkn(struct gf100_gr *gr)
 }
 
 void
-gf100_grctx_generate_r406028(struct gf100_gr *gr)
-{
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-	u32 tmp[GPC_MAX / 8] = {}, i = 0;
-	for (i = 0; i < gr->gpc_nr; i++)
-		tmp[i / 8] |= gr->tpc_nr[i] << ((i % 8) * 4);
-	for (i = 0; i < 4; i++) {
-		nvkm_wr32(device, 0x406028 + (i * 4), tmp[i]);
-		nvkm_wr32(device, 0x405870 + (i * 4), tmp[i]);
-	}
-}
-
-void
 gf100_grctx_generate_r4060a8(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -1229,8 +1216,10 @@ gf100_grctx_generate_sm_id(struct gf100_gr *gr, int gpc, int tpc, int sm)
 void
 gf100_grctx_generate_floorsweep(struct gf100_gr *gr)
 {
+	struct nvkm_device *device = gr->base.engine.subdev.device;
 	const struct gf100_grctx_func *func = gr->func->grctx;
-	int tpc, gpc, sm;
+	int tpc, gpc, sm, i, j;
+	u32 data;
 
 	for (tpc = 0, sm = 0; tpc < gr->tpc_max; tpc++) {
 		for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
@@ -1240,6 +1229,13 @@ gf100_grctx_generate_floorsweep(struct gf100_gr *gr)
 				func->tpc_nr(gr, gpc);
 		}
 	}
+
+	for (gpc = 0, i = 0; i < 4; i++) {
+		for (data = 0, j = 0; j < 8 && gpc < gr->gpc_nr; j++, gpc++)
+			data |= gr->tpc_nr[gpc] << (j * 4);
+		nvkm_wr32(device, 0x406028 + (i * 4), data);
+		nvkm_wr32(device, 0x405870 + (i * 4), data);
+	}
 }
 
 void
@@ -1271,7 +1267,6 @@ gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->unkn(gr);
 
 	gf100_grctx_generate_floorsweep(gr);
-	gf100_grctx_generate_r406028(gr);
 	gf100_grctx_generate_r4060a8(gr);
 	gf100_grctx_generate_r418bb8(gr);
 	gf100_grctx_generate_r406800(gr);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
index 6fa9b62..b9e2879 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
@@ -63,7 +63,6 @@ void gf100_grctx_generate_pagepool(struct gf100_grctx *);
 void gf100_grctx_generate_attrib(struct gf100_grctx *);
 void gf100_grctx_generate_unkn(struct gf100_gr *);
 void gf100_grctx_generate_floorsweep(struct gf100_gr *);
-void gf100_grctx_generate_r406028(struct gf100_gr *);
 void gf100_grctx_generate_r4060a8(struct gf100_gr *);
 void gf100_grctx_generate_r418bb8(struct gf100_gr *);
 void gf100_grctx_generate_r406800(struct gf100_gr *);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
index 806102a..a091485 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
@@ -241,7 +241,6 @@ gf117_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->unkn(gr);
 
 	gf100_grctx_generate_floorsweep(gr);
-	gf100_grctx_generate_r406028(gr);
 	gf100_grctx_generate_r4060a8(gr);
 	gk104_grctx_generate_r418bb8(gr);
 	gf100_grctx_generate_r406800(gr);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
index 2be72fd..cb6c156 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
@@ -975,7 +975,6 @@ gk104_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->unkn(gr);
 
 	gf100_grctx_generate_floorsweep(gr);
-	gf100_grctx_generate_r406028(gr);
 	gk104_grctx_generate_r418bb8(gr);
 	gf100_grctx_generate_r406800(gr);
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk20a.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk20a.c
index faa5f18..dda3051 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk20a.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk20a.c
@@ -43,7 +43,6 @@ gk20a_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->unkn(gr);
 
 	gf100_grctx_generate_floorsweep(gr);
-	gf100_grctx_generate_r406028(gr);
 	gk104_grctx_generate_r418bb8(gr);
 	gf100_grctx_generate_r406800(gr);
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
index e5c0273..77cce9b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
@@ -961,7 +961,6 @@ gm107_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->unkn(gr);
 
 	gf100_grctx_generate_floorsweep(gr);
-	gf100_grctx_generate_r406028(gr);
 	gk104_grctx_generate_r418bb8(gr);
 	gf100_grctx_generate_r406800(gr);
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
index be5e25a..ac6724b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
@@ -77,7 +77,6 @@ gm200_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->unkn(gr);
 
 	gf100_grctx_generate_floorsweep(gr);
-	gf100_grctx_generate_r406028(gr);
 	gk104_grctx_generate_r418bb8(gr);
 
 	for (i = 0; i < 8; i++)
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm20b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm20b.c
index 363d198..6420d13 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm20b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm20b.c
@@ -22,20 +22,6 @@
 #include "ctxgf100.h"
 
 static void
-gm20b_grctx_generate_r406028(struct gf100_gr *gr)
-{
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-	u32 tpc_per_gpc = 0;
-	int i;
-
-	for (i = 0; i < gr->gpc_nr; i++)
-		tpc_per_gpc |= gr->tpc_nr[i] << (4 * i);
-
-	nvkm_wr32(device, 0x406028, tpc_per_gpc);
-	nvkm_wr32(device, 0x405870, tpc_per_gpc);
-}
-
-static void
 gm20b_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -54,7 +40,6 @@ gm20b_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->unkn(gr);
 
 	gf100_grctx_generate_floorsweep(gr);
-	gm20b_grctx_generate_r406028(gr);
 	gk104_grctx_generate_r418bb8(gr);
 
 	for (i = 0; i < 8; i++)
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
index ac8618f..701341d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
@@ -139,7 +139,6 @@ gp100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->unkn(gr);
 
 	gf100_grctx_generate_floorsweep(gr);
-	gf100_grctx_generate_r406028(gr);
 	gk104_grctx_generate_r418bb8(gr);
 
 	for (i = 0; i < 8; i++)
-- 
cgit v1.1


From 9d8a80df73b58c700e36a0051b2fb44f252693e2 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise r4060a8 + apply fixes from traces

Also fixes some GPUs where we write too many registers.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c | 11 ++++++++---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h |  3 ++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c |  1 +
 7 files changed, 15 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index a4865f7..aededb2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1083,7 +1083,9 @@ void
 gf100_grctx_generate_r4060a8(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
-	u8  tpcnr[GPC_MAX], data[TPC_MAX];
+	const u8 gpcmax = nvkm_rd32(device, 0x022430);
+	const u8 tpcmax = nvkm_rd32(device, 0x022434) * gpcmax;
+	u8 tpcnr[GPC_MAX], data[TPC_MAX];
 	int gpc, tpc, i;
 
 	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
@@ -1098,7 +1100,7 @@ gf100_grctx_generate_r4060a8(struct gf100_gr *gr)
 		data[tpc] = gpc;
 	}
 
-	for (i = 0; i < 4; i++)
+	for (i = 0; i < DIV_ROUND_UP(tpcmax, 4); i++)
 		nvkm_wr32(device, 0x4060a8 + (i * 4), ((u32 *)data)[i]);
 }
 
@@ -1236,6 +1238,9 @@ gf100_grctx_generate_floorsweep(struct gf100_gr *gr)
 		nvkm_wr32(device, 0x406028 + (i * 4), data);
 		nvkm_wr32(device, 0x405870 + (i * 4), data);
 	}
+
+	if (func->r4060a8)
+		func->r4060a8(gr);
 }
 
 void
@@ -1267,7 +1272,6 @@ gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->unkn(gr);
 
 	gf100_grctx_generate_floorsweep(gr);
-	gf100_grctx_generate_r4060a8(gr);
 	gf100_grctx_generate_r418bb8(gr);
 	gf100_grctx_generate_r406800(gr);
 
@@ -1419,4 +1423,5 @@ gf100_grctx = {
 	.attrib_nr = 0x218,
 	.sm_id = gf100_grctx_generate_sm_id,
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
+	.r4060a8 = gf100_grctx_generate_r4060a8,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
index b9e2879..8430d82 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
@@ -53,6 +53,7 @@ struct gf100_grctx_func {
 	/* floorsweeping */
 	void (*sm_id)(struct gf100_gr *, int gpc, int tpc, int sm);
 	void (*tpc_nr)(struct gf100_gr *, int gpc);
+	void (*r4060a8)(struct gf100_gr *);
 };
 
 extern const struct gf100_grctx_func gf100_grctx;
@@ -63,11 +64,11 @@ void gf100_grctx_generate_pagepool(struct gf100_grctx *);
 void gf100_grctx_generate_attrib(struct gf100_grctx *);
 void gf100_grctx_generate_unkn(struct gf100_gr *);
 void gf100_grctx_generate_floorsweep(struct gf100_gr *);
-void gf100_grctx_generate_r4060a8(struct gf100_gr *);
 void gf100_grctx_generate_r418bb8(struct gf100_gr *);
 void gf100_grctx_generate_r406800(struct gf100_gr *);
 void gf100_grctx_generate_sm_id(struct gf100_gr *, int, int, int);
 void gf100_grctx_generate_tpc_nr(struct gf100_gr *, int);
+void gf100_grctx_generate_r4060a8(struct gf100_gr *);
 
 extern const struct gf100_grctx_func gf108_grctx;
 void gf108_grctx_generate_attrib(struct gf100_grctx *);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c
index 19cc84e..471fa9e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c
@@ -98,4 +98,5 @@ gf104_grctx = {
 	.attrib_nr = 0x218,
 	.sm_id = gf100_grctx_generate_sm_id,
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
+	.r4060a8 = gf100_grctx_generate_r4060a8,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c
index 3703731..b472a3e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c
@@ -796,4 +796,5 @@ gf108_grctx = {
 	.alpha_nr = 0x218,
 	.sm_id = gf100_grctx_generate_sm_id,
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
+	.r4060a8 = gf100_grctx_generate_r4060a8,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c
index 6d72ae0..12a98f6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c
@@ -349,4 +349,5 @@ gf110_grctx = {
 	.attrib_nr = 0x218,
 	.sm_id = gf100_grctx_generate_sm_id,
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
+	.r4060a8 = gf100_grctx_generate_r4060a8,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
index a091485..5c4d6d9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
@@ -241,7 +241,6 @@ gf117_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->unkn(gr);
 
 	gf100_grctx_generate_floorsweep(gr);
-	gf100_grctx_generate_r4060a8(gr);
 	gk104_grctx_generate_r418bb8(gr);
 	gf100_grctx_generate_r406800(gr);
 
@@ -276,4 +275,5 @@ gf117_grctx = {
 	.alpha_nr = 0x324,
 	.sm_id = gf100_grctx_generate_sm_id,
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
+	.r4060a8 = gf100_grctx_generate_r4060a8,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c
index 784ab94..cbf6c6a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c
@@ -519,4 +519,5 @@ gf119_grctx = {
 	.alpha_nr = 0x218,
 	.sm_id = gf100_grctx_generate_sm_id,
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
+	.r4060a8 = gf100_grctx_generate_r4060a8,
 };
-- 
cgit v1.1


From ff209c235de9c3437e131b39eb976ff4bcc4c516 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise rop_mapping

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c  |  6 ++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h  |  5 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c  | 61 +++++++++++++++++++++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c  | 61 +---------------------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk20a.c  |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c  |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c  |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm20b.c  |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c  |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c  |  1 +
 18 files changed, 82 insertions(+), 70 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index aededb2..d35711b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1105,7 +1105,7 @@ gf100_grctx_generate_r4060a8(struct gf100_gr *gr)
 }
 
 void
-gf100_grctx_generate_r418bb8(struct gf100_gr *gr)
+gf100_grctx_generate_rop_mapping(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
 	u32 data[6] = {}, data2[2] = {};
@@ -1241,6 +1241,8 @@ gf100_grctx_generate_floorsweep(struct gf100_gr *gr)
 
 	if (func->r4060a8)
 		func->r4060a8(gr);
+
+	func->rop_mapping(gr);
 }
 
 void
@@ -1272,7 +1274,6 @@ gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->unkn(gr);
 
 	gf100_grctx_generate_floorsweep(gr);
-	gf100_grctx_generate_r418bb8(gr);
 	gf100_grctx_generate_r406800(gr);
 
 	gf100_gr_icmd(gr, grctx->icmd);
@@ -1424,4 +1425,5 @@ gf100_grctx = {
 	.sm_id = gf100_grctx_generate_sm_id,
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
 	.r4060a8 = gf100_grctx_generate_r4060a8,
+	.rop_mapping = gf100_grctx_generate_rop_mapping,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
index 8430d82..0c0d2a5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
@@ -54,6 +54,7 @@ struct gf100_grctx_func {
 	void (*sm_id)(struct gf100_gr *, int gpc, int tpc, int sm);
 	void (*tpc_nr)(struct gf100_gr *, int gpc);
 	void (*r4060a8)(struct gf100_gr *);
+	void (*rop_mapping)(struct gf100_gr *);
 };
 
 extern const struct gf100_grctx_func gf100_grctx;
@@ -64,11 +65,11 @@ void gf100_grctx_generate_pagepool(struct gf100_grctx *);
 void gf100_grctx_generate_attrib(struct gf100_grctx *);
 void gf100_grctx_generate_unkn(struct gf100_gr *);
 void gf100_grctx_generate_floorsweep(struct gf100_gr *);
-void gf100_grctx_generate_r418bb8(struct gf100_gr *);
 void gf100_grctx_generate_r406800(struct gf100_gr *);
 void gf100_grctx_generate_sm_id(struct gf100_gr *, int, int, int);
 void gf100_grctx_generate_tpc_nr(struct gf100_gr *, int);
 void gf100_grctx_generate_r4060a8(struct gf100_gr *);
+void gf100_grctx_generate_rop_mapping(struct gf100_gr *);
 
 extern const struct gf100_grctx_func gf108_grctx;
 void gf108_grctx_generate_attrib(struct gf100_grctx *);
@@ -79,6 +80,7 @@ extern const struct gf100_grctx_func gf110_grctx;
 
 extern const struct gf100_grctx_func gf117_grctx;
 void gf117_grctx_generate_attrib(struct gf100_grctx *);
+void gf117_grctx_generate_rop_mapping(struct gf100_gr *);
 
 extern const struct gf100_grctx_func gf119_grctx;
 
@@ -89,7 +91,6 @@ void gk104_grctx_generate_bundle(struct gf100_grctx *);
 void gk104_grctx_generate_pagepool(struct gf100_grctx *);
 void gk104_grctx_generate_patch_ltc(struct gf100_grctx *);
 void gk104_grctx_generate_unkn(struct gf100_gr *);
-void gk104_grctx_generate_r418bb8(struct gf100_gr *);
 
 void gm107_grctx_generate_bundle(struct gf100_grctx *);
 void gm107_grctx_generate_pagepool(struct gf100_grctx *);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c
index 471fa9e..eb0d377 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c
@@ -99,4 +99,5 @@ gf104_grctx = {
 	.sm_id = gf100_grctx_generate_sm_id,
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
 	.r4060a8 = gf100_grctx_generate_r4060a8,
+	.rop_mapping = gf100_grctx_generate_rop_mapping,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c
index b472a3e..f218d98 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c
@@ -797,4 +797,5 @@ gf108_grctx = {
 	.sm_id = gf100_grctx_generate_sm_id,
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
 	.r4060a8 = gf100_grctx_generate_r4060a8,
+	.rop_mapping = gf100_grctx_generate_rop_mapping,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c
index 12a98f6..35c1638 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c
@@ -350,4 +350,5 @@ gf110_grctx = {
 	.sm_id = gf100_grctx_generate_sm_id,
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
 	.r4060a8 = gf100_grctx_generate_r4060a8,
+	.rop_mapping = gf100_grctx_generate_rop_mapping,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
index 5c4d6d9..7aaf8a2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
@@ -180,6 +180,65 @@ gf117_grctx_pack_ppc[] = {
  ******************************************************************************/
 
 void
+gf117_grctx_generate_rop_mapping(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	u32 data[6] = {}, data2[2] = {};
+	u8  tpcnr[GPC_MAX];
+	u8  shift, ntpcv;
+	int gpc, tpc, i;
+
+	/* calculate first set of magics */
+	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
+
+	gpc = -1;
+	for (tpc = 0; tpc < gr->tpc_total; tpc++) {
+		do {
+			gpc = (gpc + 1) % gr->gpc_nr;
+		} while (!tpcnr[gpc]);
+		tpcnr[gpc]--;
+
+		data[tpc / 6] |= gpc << ((tpc % 6) * 5);
+	}
+
+	for (; tpc < 32; tpc++)
+		data[tpc / 6] |= 7 << ((tpc % 6) * 5);
+
+	/* and the second... */
+	shift = 0;
+	ntpcv = gr->tpc_total;
+	while (!(ntpcv & (1 << 4))) {
+		ntpcv <<= 1;
+		shift++;
+	}
+
+	data2[0]  = (ntpcv << 16);
+	data2[0] |= (shift << 21);
+	data2[0] |= (((1 << (0 + 5)) % ntpcv) << 24);
+	for (i = 1; i < 7; i++)
+		data2[1] |= ((1 << (i + 5)) % ntpcv) << ((i - 1) * 5);
+
+	/* GPC_BROADCAST */
+	nvkm_wr32(device, 0x418bb8, (gr->tpc_total << 8) |
+				     gr->screen_tile_row_offset);
+	for (i = 0; i < 6; i++)
+		nvkm_wr32(device, 0x418b08 + (i * 4), data[i]);
+
+	/* GPC_BROADCAST.TP_BROADCAST */
+	nvkm_wr32(device, 0x41bfd0, (gr->tpc_total << 8) |
+				     gr->screen_tile_row_offset | data2[0]);
+	nvkm_wr32(device, 0x41bfe4, data2[1]);
+	for (i = 0; i < 6; i++)
+		nvkm_wr32(device, 0x41bf00 + (i * 4), data[i]);
+
+	/* UNK78xx */
+	nvkm_wr32(device, 0x4078bc, (gr->tpc_total << 8) |
+				     gr->screen_tile_row_offset);
+	for (i = 0; i < 6; i++)
+		nvkm_wr32(device, 0x40780c + (i * 4), data[i]);
+}
+
+void
 gf117_grctx_generate_attrib(struct gf100_grctx *info)
 {
 	struct gf100_gr *gr = info->gr;
@@ -241,7 +300,6 @@ gf117_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->unkn(gr);
 
 	gf100_grctx_generate_floorsweep(gr);
-	gk104_grctx_generate_r418bb8(gr);
 	gf100_grctx_generate_r406800(gr);
 
 	for (i = 0; i < 8; i++)
@@ -276,4 +334,5 @@ gf117_grctx = {
 	.sm_id = gf100_grctx_generate_sm_id,
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
 	.r4060a8 = gf100_grctx_generate_r4060a8,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c
index cbf6c6a..4cd5d86 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c
@@ -520,4 +520,5 @@ gf119_grctx = {
 	.sm_id = gf100_grctx_generate_sm_id,
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
 	.r4060a8 = gf100_grctx_generate_r4060a8,
+	.rop_mapping = gf100_grctx_generate_rop_mapping,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
index cb6c156..25a5209 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
@@ -893,65 +893,6 @@ gk104_grctx_generate_unkn(struct gf100_gr *gr)
 }
 
 void
-gk104_grctx_generate_r418bb8(struct gf100_gr *gr)
-{
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-	u32 data[6] = {}, data2[2] = {};
-	u8  tpcnr[GPC_MAX];
-	u8  shift, ntpcv;
-	int gpc, tpc, i;
-
-	/* calculate first set of magics */
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-
-	gpc = -1;
-	for (tpc = 0; tpc < gr->tpc_total; tpc++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while (!tpcnr[gpc]);
-		tpcnr[gpc]--;
-
-		data[tpc / 6] |= gpc << ((tpc % 6) * 5);
-	}
-
-	for (; tpc < 32; tpc++)
-		data[tpc / 6] |= 7 << ((tpc % 6) * 5);
-
-	/* and the second... */
-	shift = 0;
-	ntpcv = gr->tpc_total;
-	while (!(ntpcv & (1 << 4))) {
-		ntpcv <<= 1;
-		shift++;
-	}
-
-	data2[0]  = (ntpcv << 16);
-	data2[0] |= (shift << 21);
-	data2[0] |= (((1 << (0 + 5)) % ntpcv) << 24);
-	for (i = 1; i < 7; i++)
-		data2[1] |= ((1 << (i + 5)) % ntpcv) << ((i - 1) * 5);
-
-	/* GPC_BROADCAST */
-	nvkm_wr32(device, 0x418bb8, (gr->tpc_total << 8) |
-				     gr->screen_tile_row_offset);
-	for (i = 0; i < 6; i++)
-		nvkm_wr32(device, 0x418b08 + (i * 4), data[i]);
-
-	/* GPC_BROADCAST.TP_BROADCAST */
-	nvkm_wr32(device, 0x41bfd0, (gr->tpc_total << 8) |
-				     gr->screen_tile_row_offset | data2[0]);
-	nvkm_wr32(device, 0x41bfe4, data2[1]);
-	for (i = 0; i < 6; i++)
-		nvkm_wr32(device, 0x41bf00 + (i * 4), data[i]);
-
-	/* UNK78xx */
-	nvkm_wr32(device, 0x4078bc, (gr->tpc_total << 8) |
-				     gr->screen_tile_row_offset);
-	for (i = 0; i < 6; i++)
-		nvkm_wr32(device, 0x40780c + (i * 4), data[i]);
-}
-
-void
 gk104_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -975,7 +916,6 @@ gk104_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->unkn(gr);
 
 	gf100_grctx_generate_floorsweep(gr);
-	gk104_grctx_generate_r418bb8(gr);
 	gf100_grctx_generate_r406800(gr);
 
 	for (i = 0; i < 8; i++)
@@ -1018,4 +958,5 @@ gk104_grctx = {
 	.patch_ltc = gk104_grctx_generate_patch_ltc,
 	.sm_id = gf100_grctx_generate_sm_id,
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
index 5013fc5..038cc47 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
@@ -833,4 +833,5 @@ gk110_grctx = {
 	.patch_ltc = gk104_grctx_generate_patch_ltc,
 	.sm_id = gf100_grctx_generate_sm_id,
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
index 1dd5742..8b025e2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
@@ -94,4 +94,5 @@ gk110b_grctx = {
 	.patch_ltc = gk104_grctx_generate_patch_ltc,
 	.sm_id = gf100_grctx_generate_sm_id,
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
index 214f4da..9479b404 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
@@ -555,4 +555,5 @@ gk208_grctx = {
 	.patch_ltc = gk104_grctx_generate_patch_ltc,
 	.sm_id = gf100_grctx_generate_sm_id,
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk20a.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk20a.c
index dda3051..23abbfa 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk20a.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk20a.c
@@ -43,7 +43,6 @@ gk20a_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->unkn(gr);
 
 	gf100_grctx_generate_floorsweep(gr);
-	gk104_grctx_generate_r418bb8(gr);
 	gf100_grctx_generate_r406800(gr);
 
 	for (i = 0; i < 8; i++)
@@ -83,4 +82,5 @@ gk20a_grctx = {
 	.alpha_nr = 0x648,
 	.sm_id = gf100_grctx_generate_sm_id,
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
index 77cce9b..fef6652 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
@@ -961,7 +961,6 @@ gm107_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->unkn(gr);
 
 	gf100_grctx_generate_floorsweep(gr);
-	gk104_grctx_generate_r418bb8(gr);
 	gf100_grctx_generate_r406800(gr);
 
 	nvkm_wr32(device, 0x4064d0, 0x00000001);
@@ -1005,4 +1004,5 @@ gm107_grctx = {
 	.alpha_nr = 0x1000,
 	.sm_id = gm107_grctx_generate_sm_id,
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
index ac6724b..cfccd75 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
@@ -77,7 +77,6 @@ gm200_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->unkn(gr);
 
 	gf100_grctx_generate_floorsweep(gr);
-	gk104_grctx_generate_r418bb8(gr);
 
 	for (i = 0; i < 8; i++)
 		nvkm_wr32(device, 0x4064d0 + (i * 0x04), 0x00000000);
@@ -115,4 +114,5 @@ gm200_grctx = {
 	.alpha_nr_max = 0x1800,
 	.alpha_nr = 0x1000,
 	.sm_id = gm107_grctx_generate_sm_id,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm20b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm20b.c
index 6420d13..3dd4e18 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm20b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm20b.c
@@ -40,7 +40,6 @@ gm20b_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->unkn(gr);
 
 	gf100_grctx_generate_floorsweep(gr);
-	gk104_grctx_generate_r418bb8(gr);
 
 	for (i = 0; i < 8; i++)
 		nvkm_wr32(device, 0x4064d0 + (i * 0x04), 0x00000000);
@@ -84,4 +83,5 @@ gm20b_grctx = {
 	.alpha_nr_max = 0xc00,
 	.alpha_nr = 0x800,
 	.sm_id = gm107_grctx_generate_sm_id,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
index 701341d..e099907 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
@@ -139,7 +139,6 @@ gp100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->unkn(gr);
 
 	gf100_grctx_generate_floorsweep(gr);
-	gk104_grctx_generate_r418bb8(gr);
 
 	for (i = 0; i < 8; i++)
 		nvkm_wr32(device, 0x4064d0 + (i * 0x04), 0x00000000);
@@ -174,4 +173,5 @@ gp100_grctx = {
 	.alpha_nr_max = 0xc00,
 	.alpha_nr = 0x800,
 	.sm_id = gm107_grctx_generate_sm_id,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
index a267abc..553a609 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
@@ -95,4 +95,5 @@ gp102_grctx = {
 	.alpha_nr_max = 0xc00,
 	.alpha_nr = 0x800,
 	.sm_id = gm107_grctx_generate_sm_id,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
index 77345b2..db3fff8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
@@ -45,4 +45,5 @@ gp107_grctx = {
 	.alpha_nr_max = 0xc00,
 	.alpha_nr = 0x800,
 	.sm_id = gm107_grctx_generate_sm_id,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
 };
-- 
cgit v1.1


From 43952c6f43106c88b4dcdc99285d92172d8c57cd Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise alpha_beta_tables + improve
 algorithms

I haven't yet been able to find a fully programatic way of calculating the
same mapping as NVIDIA for GF100-GF119, so the algorithm partially depends
on data tables for specific configurations.

I couldn't find traces for every possibility, so the algorithm will switch
to a mapping similar to what GK104-GM10x use if it encounters one.  We did
the wrong thing before anyway, so shouldn't matter too much.

The algorithm used in the GK104 implementation was ported from NVGPU.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c  | 158 +++++++++++++++++----
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h  |   5 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c  |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c  |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c  |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c  |   2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c  |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c  |  49 ++++++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c  |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c  |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk20a.c  |   2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c  |   2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c     |  10 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h     |   1 +
 15 files changed, 201 insertions(+), 35 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index d35711b..7a22614 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1163,37 +1163,140 @@ gf100_grctx_generate_rop_mapping(struct gf100_gr *gr)
 		nvkm_wr32(device, 0x40780c + (i * 4), data[i]);
 }
 
+static const u32
+gf100_grctx_alpha_beta_map[17][32] = {
+	[1] = {
+		1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
+		1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
+	},
+	[2] = {
+		1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
+		1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
+	},
+	//XXX: 3
+	[4] = {
+		1, 1, 1, 1, 1, 1, 1, 1,
+		2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+		3, 3, 3, 3, 3, 3, 3, 3,
+	},
+	//XXX: 5
+	//XXX: 6
+	[7] = {
+		1, 1, 1, 1,
+		2, 2, 2, 2, 2, 2,
+		3, 3, 3, 3, 3, 3,
+		4, 4, 4, 4, 4, 4,
+		5, 5, 5, 5, 5, 5,
+		6, 6, 6, 6,
+	},
+	[8] = {
+		1, 1, 1,
+		2, 2, 2, 2, 2,
+		3, 3, 3, 3, 3,
+		4, 4, 4, 4, 4, 4,
+		5, 5, 5, 5, 5,
+		6, 6, 6, 6, 6,
+		7, 7, 7,
+	},
+	//XXX: 9
+	//XXX: 10
+	[11] = {
+		1, 1,
+		2, 2, 2, 2,
+		3, 3, 3,
+		4, 4, 4, 4,
+		5, 5, 5,
+		6, 6, 6,
+		7, 7, 7, 7,
+		8, 8, 8,
+		9, 9, 9, 9,
+		10, 10,
+	},
+	//XXX: 12
+	//XXX: 13
+	[14] = {
+		1, 1,
+		2, 2,
+		3, 3, 3,
+		4, 4, 4,
+		5, 5,
+		6, 6, 6,
+		7, 7,
+		8, 8, 8,
+		9, 9,
+		10, 10, 10,
+		11, 11, 11,
+		12, 12,
+		13, 13,
+	},
+	[15] = {
+		1, 1,
+		2, 2,
+		3, 3,
+		4, 4, 4,
+		5, 5,
+		6, 6, 6,
+		7, 7,
+		8, 8,
+		9, 9, 9,
+		10, 10,
+		11, 11, 11,
+		12, 12,
+		13, 13,
+		14, 14,
+	},
+	[16] = {
+		1, 1,
+		2, 2,
+		3, 3,
+		4, 4,
+		5, 5,
+		6, 6, 6,
+		7, 7,
+		8, 8,
+		9, 9,
+		10, 10, 10,
+		11, 11,
+		12, 12,
+		13, 13,
+		14, 14,
+		15, 15,
+	},
+};
+
 void
-gf100_grctx_generate_r406800(struct gf100_gr *gr)
+gf100_grctx_generate_alpha_beta_tables(struct gf100_gr *gr)
 {
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-	u64 tpc_mask = 0, tpc_set = 0;
-	u8  tpcnr[GPC_MAX];
-	int gpc, tpc;
-	int i, a, b;
+	struct nvkm_subdev *subdev = &gr->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	int i, gpc;
 
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-	for (gpc = 0; gpc < gr->gpc_nr; gpc++)
-		tpc_mask |= ((1ULL << gr->tpc_nr[gpc]) - 1) << (gpc * 8);
-
-	for (i = 0, gpc = -1, b = -1; i < 32; i++) {
-		a = (i * (gr->tpc_total - 1)) / 32;
-		if (a != b) {
-			b = a;
-			do {
-				gpc = (gpc + 1) % gr->gpc_nr;
-			} while (!tpcnr[gpc]);
-			tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--;
-
-			tpc_set |= 1ULL << ((gpc * 8) + tpc);
+	for (i = 0; i < 32; i++) {
+		u32 atarget = gf100_grctx_alpha_beta_map[gr->tpc_total][i];
+		u32 abits[GPC_MAX] = {}, amask = 0, bmask = 0;
+
+		if (!atarget) {
+			nvkm_warn(subdev, "missing alpha/beta mapping table\n");
+			atarget = max_t(u32, gr->tpc_total * i / 32, 1);
 		}
 
-		nvkm_wr32(device, 0x406800 + (i * 0x20), lower_32_bits(tpc_set));
-		nvkm_wr32(device, 0x406c00 + (i * 0x20), lower_32_bits(tpc_set ^ tpc_mask));
-		if (gr->gpc_nr > 4) {
-			nvkm_wr32(device, 0x406804 + (i * 0x20), upper_32_bits(tpc_set));
-			nvkm_wr32(device, 0x406c04 + (i * 0x20), upper_32_bits(tpc_set ^ tpc_mask));
+		while (atarget) {
+			for (gpc = 0; atarget && gpc < gr->gpc_nr; gpc++) {
+				if (abits[gpc] < gr->tpc_nr[gpc]) {
+					abits[gpc]++;
+					atarget--;
+				}
+			}
 		}
+
+		for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
+			u32 bbits = gr->tpc_nr[gpc] - abits[gpc];
+			amask |= ((1 << abits[gpc]) - 1) << (gpc * 8);
+			bmask |= ((1 << bbits) - 1) << abits[gpc] << (gpc * 8);
+		}
+
+		nvkm_wr32(device, 0x406800 + (i * 0x20), amask);
+		nvkm_wr32(device, 0x406c00 + (i * 0x20), bmask);
 	}
 }
 
@@ -1243,6 +1346,9 @@ gf100_grctx_generate_floorsweep(struct gf100_gr *gr)
 		func->r4060a8(gr);
 
 	func->rop_mapping(gr);
+
+	if (func->alpha_beta_tables)
+		func->alpha_beta_tables(gr);
 }
 
 void
@@ -1274,7 +1380,6 @@ gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->unkn(gr);
 
 	gf100_grctx_generate_floorsweep(gr);
-	gf100_grctx_generate_r406800(gr);
 
 	gf100_gr_icmd(gr, grctx->icmd);
 	nvkm_wr32(device, 0x404154, idle_timeout);
@@ -1426,4 +1531,5 @@ gf100_grctx = {
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
 	.r4060a8 = gf100_grctx_generate_r4060a8,
 	.rop_mapping = gf100_grctx_generate_rop_mapping,
+	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
index 0c0d2a5..a531076 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
@@ -55,6 +55,7 @@ struct gf100_grctx_func {
 	void (*tpc_nr)(struct gf100_gr *, int gpc);
 	void (*r4060a8)(struct gf100_gr *);
 	void (*rop_mapping)(struct gf100_gr *);
+	void (*alpha_beta_tables)(struct gf100_gr *);
 };
 
 extern const struct gf100_grctx_func gf100_grctx;
@@ -65,11 +66,11 @@ void gf100_grctx_generate_pagepool(struct gf100_grctx *);
 void gf100_grctx_generate_attrib(struct gf100_grctx *);
 void gf100_grctx_generate_unkn(struct gf100_gr *);
 void gf100_grctx_generate_floorsweep(struct gf100_gr *);
-void gf100_grctx_generate_r406800(struct gf100_gr *);
 void gf100_grctx_generate_sm_id(struct gf100_gr *, int, int, int);
 void gf100_grctx_generate_tpc_nr(struct gf100_gr *, int);
 void gf100_grctx_generate_r4060a8(struct gf100_gr *);
 void gf100_grctx_generate_rop_mapping(struct gf100_gr *);
+void gf100_grctx_generate_alpha_beta_tables(struct gf100_gr *);
 
 extern const struct gf100_grctx_func gf108_grctx;
 void gf108_grctx_generate_attrib(struct gf100_grctx *);
@@ -85,6 +86,8 @@ void gf117_grctx_generate_rop_mapping(struct gf100_gr *);
 extern const struct gf100_grctx_func gf119_grctx;
 
 extern const struct gf100_grctx_func gk104_grctx;
+void gk104_grctx_generate_alpha_beta_tables(struct gf100_gr *);
+
 extern const struct gf100_grctx_func gk20a_grctx;
 void gk104_grctx_generate_main(struct gf100_gr *, struct gf100_grctx *);
 void gk104_grctx_generate_bundle(struct gf100_grctx *);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c
index eb0d377..edeb369 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c
@@ -100,4 +100,5 @@ gf104_grctx = {
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
 	.r4060a8 = gf100_grctx_generate_r4060a8,
 	.rop_mapping = gf100_grctx_generate_rop_mapping,
+	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c
index f218d98..267cfdf 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c
@@ -798,4 +798,5 @@ gf108_grctx = {
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
 	.r4060a8 = gf100_grctx_generate_r4060a8,
 	.rop_mapping = gf100_grctx_generate_rop_mapping,
+	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c
index 35c1638..1fb934f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c
@@ -351,4 +351,5 @@ gf110_grctx = {
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
 	.r4060a8 = gf100_grctx_generate_r4060a8,
 	.rop_mapping = gf100_grctx_generate_rop_mapping,
+	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
index 7aaf8a2..f159b55 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
@@ -300,7 +300,6 @@ gf117_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->unkn(gr);
 
 	gf100_grctx_generate_floorsweep(gr);
-	gf100_grctx_generate_r406800(gr);
 
 	for (i = 0; i < 8; i++)
 		nvkm_wr32(device, 0x4064d0 + (i * 0x04), 0x00000000);
@@ -335,4 +334,5 @@ gf117_grctx = {
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
 	.r4060a8 = gf100_grctx_generate_r4060a8,
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
+	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c
index 4cd5d86..3720afd 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c
@@ -521,4 +521,5 @@ gf119_grctx = {
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
 	.r4060a8 = gf100_grctx_generate_r4060a8,
 	.rop_mapping = gf100_grctx_generate_rop_mapping,
+	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
index 25a5209..25576c1 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
@@ -916,7 +916,6 @@ gk104_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->unkn(gr);
 
 	gf100_grctx_generate_floorsweep(gr);
-	gf100_grctx_generate_r406800(gr);
 
 	for (i = 0; i < 8; i++)
 		nvkm_wr32(device, 0x4064d0 + (i * 0x04), 0x00000000);
@@ -933,6 +932,53 @@ gk104_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	nvkm_mask(device, 0x41be10, 0x00800000, 0x00800000);
 }
 
+void
+gk104_grctx_generate_alpha_beta_tables(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	int i, j, gpc, ppc;
+
+	for (i = 0; i < 32; i++) {
+		u32 atarget = max_t(u32, gr->tpc_total * i / 32, 1);
+		u32 btarget = gr->tpc_total - atarget;
+		bool alpha = atarget < btarget;
+		u64 amask = 0, bmask = 0;
+
+		for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
+			for (ppc = 0; ppc < gr->func->ppc_nr; ppc++) {
+				u32 ppc_tpcs = gr->ppc_tpc_nr[gpc][ppc];
+				u32 abits, bbits, pmask;
+
+				if (alpha) {
+					abits = atarget ? ppc_tpcs : 0;
+					bbits = ppc_tpcs - abits;
+				} else {
+					bbits = btarget ? ppc_tpcs : 0;
+					abits = ppc_tpcs - bbits;
+				}
+
+				pmask = gr->ppc_tpc_mask[gpc][ppc];
+				while (ppc_tpcs-- > abits)
+					pmask &= pmask - 1;
+				amask |= (u64)pmask << (gpc * 8);
+
+				pmask ^= gr->ppc_tpc_mask[gpc][ppc];
+				bmask |= (u64)pmask << (gpc * 8);
+
+				atarget -= min(abits, atarget);
+				btarget -= min(bbits, btarget);
+				if ((abits > 0) || (bbits > 0))
+					alpha = !alpha;
+			}
+		}
+
+		for (j = 0; j < gr->gpc_nr; j += 4, amask >>= 32, bmask >>= 32) {
+			nvkm_wr32(device, 0x406800 + (i * 0x20) + j, amask);
+			nvkm_wr32(device, 0x406c00 + (i * 0x20) + j, bmask);
+		}
+	}
+}
+
 const struct gf100_grctx_func
 gk104_grctx = {
 	.main  = gk104_grctx_generate_main,
@@ -959,4 +1005,5 @@ gk104_grctx = {
 	.sm_id = gf100_grctx_generate_sm_id,
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
+	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
index 038cc47..284570a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
@@ -834,4 +834,5 @@ gk110_grctx = {
 	.sm_id = gf100_grctx_generate_sm_id,
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
+	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
index 8b025e2..ffd8cf9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
@@ -95,4 +95,5 @@ gk110b_grctx = {
 	.sm_id = gf100_grctx_generate_sm_id,
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
+	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
index 9479b404..e5e4d4dc 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
@@ -556,4 +556,5 @@ gk208_grctx = {
 	.sm_id = gf100_grctx_generate_sm_id,
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
+	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk20a.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk20a.c
index 23abbfa..896d473 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk20a.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk20a.c
@@ -43,7 +43,6 @@ gk20a_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->unkn(gr);
 
 	gf100_grctx_generate_floorsweep(gr);
-	gf100_grctx_generate_r406800(gr);
 
 	for (i = 0; i < 8; i++)
 		nvkm_wr32(device, 0x4064d0 + (i * 0x04), 0x00000000);
@@ -83,4 +82,5 @@ gk20a_grctx = {
 	.sm_id = gf100_grctx_generate_sm_id,
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
+	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
index fef6652..c209bf3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
@@ -961,7 +961,6 @@ gm107_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	grctx->unkn(gr);
 
 	gf100_grctx_generate_floorsweep(gr);
-	gf100_grctx_generate_r406800(gr);
 
 	nvkm_wr32(device, 0x4064d0, 0x00000001);
 	for (i = 1; i < 8; i++)
@@ -1005,4 +1004,5 @@ gm107_grctx = {
 	.sm_id = gm107_grctx_generate_sm_id,
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
+	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index b9519fa..fe3b44d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -1679,10 +1679,12 @@ gf100_gr_oneinit(struct nvkm_gr *base)
 		gr->tpc_total += gr->tpc_nr[i];
 		gr->ppc_nr[i]  = gr->func->ppc_nr;
 		for (j = 0; j < gr->ppc_nr[i]; j++) {
-			u8 mask = nvkm_rd32(device, GPC_UNIT(i, 0x0c30 + (j * 4)));
-			if (mask)
-				gr->ppc_mask[i] |= (1 << j);
-			gr->ppc_tpc_nr[i][j] = hweight8(mask);
+			gr->ppc_tpc_mask[i][j] =
+				nvkm_rd32(device, GPC_UNIT(i, 0x0c30 + (j * 4)));
+			if (gr->ppc_tpc_mask[i][j] == 0)
+				continue;
+			gr->ppc_mask[i] |= (1 << j);
+			gr->ppc_tpc_nr[i][j] = hweight8(gr->ppc_tpc_mask[i][j]);
 		}
 	}
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index ad352ee..6f7a786 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -103,6 +103,7 @@ struct gf100_gr {
 	u8 tpc_total;
 	u8 ppc_nr[GPC_MAX];
 	u8 ppc_mask[GPC_MAX];
+	u8 ppc_tpc_mask[GPC_MAX][4];
 	u8 ppc_tpc_nr[GPC_MAX][4];
 
 	struct gf100_gr_data mmio_data[4];
-- 
cgit v1.1


From c4a2b6385dd4a32759acf8e7884acd1115054887 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-gf119: modify max_ways_evict where required

I don't think this is done after Fermi, NVGPU used to do it but removed
the code, and I've not seen RM traces touching it either.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c | 12 ++++++++++++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c |  1 +
 7 files changed, 19 insertions(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index 7a22614..cdf74f3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1163,6 +1163,15 @@ gf100_grctx_generate_rop_mapping(struct gf100_gr *gr)
 		nvkm_wr32(device, 0x40780c + (i * 4), data[i]);
 }
 
+void
+gf100_grctx_generate_max_ways_evict(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	u32 fbps = nvkm_rd32(device, 0x121c74);
+	if (fbps == 1)
+		nvkm_mask(device, 0x17e91c, 0x001f0000, 0x00090000);
+}
+
 static const u32
 gf100_grctx_alpha_beta_map[17][32] = {
 	[1] = {
@@ -1349,6 +1358,8 @@ gf100_grctx_generate_floorsweep(struct gf100_gr *gr)
 
 	if (func->alpha_beta_tables)
 		func->alpha_beta_tables(gr);
+	if (func->max_ways_evict)
+		func->max_ways_evict(gr);
 }
 
 void
@@ -1532,4 +1543,5 @@ gf100_grctx = {
 	.r4060a8 = gf100_grctx_generate_r4060a8,
 	.rop_mapping = gf100_grctx_generate_rop_mapping,
 	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
+	.max_ways_evict = gf100_grctx_generate_max_ways_evict,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
index a531076..41cb875 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
@@ -56,6 +56,7 @@ struct gf100_grctx_func {
 	void (*r4060a8)(struct gf100_gr *);
 	void (*rop_mapping)(struct gf100_gr *);
 	void (*alpha_beta_tables)(struct gf100_gr *);
+	void (*max_ways_evict)(struct gf100_gr *);
 };
 
 extern const struct gf100_grctx_func gf100_grctx;
@@ -71,6 +72,7 @@ void gf100_grctx_generate_tpc_nr(struct gf100_gr *, int);
 void gf100_grctx_generate_r4060a8(struct gf100_gr *);
 void gf100_grctx_generate_rop_mapping(struct gf100_gr *);
 void gf100_grctx_generate_alpha_beta_tables(struct gf100_gr *);
+void gf100_grctx_generate_max_ways_evict(struct gf100_gr *);
 
 extern const struct gf100_grctx_func gf108_grctx;
 void gf108_grctx_generate_attrib(struct gf100_grctx *);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c
index edeb369..d246bb6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c
@@ -101,4 +101,5 @@ gf104_grctx = {
 	.r4060a8 = gf100_grctx_generate_r4060a8,
 	.rop_mapping = gf100_grctx_generate_rop_mapping,
 	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
+	.max_ways_evict = gf100_grctx_generate_max_ways_evict,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c
index 267cfdf..3b6e645 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c
@@ -799,4 +799,5 @@ gf108_grctx = {
 	.r4060a8 = gf100_grctx_generate_r4060a8,
 	.rop_mapping = gf100_grctx_generate_rop_mapping,
 	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
+	.max_ways_evict = gf100_grctx_generate_max_ways_evict,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c
index 1fb934f..b65a1f3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c
@@ -352,4 +352,5 @@ gf110_grctx = {
 	.r4060a8 = gf100_grctx_generate_r4060a8,
 	.rop_mapping = gf100_grctx_generate_rop_mapping,
 	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
+	.max_ways_evict = gf100_grctx_generate_max_ways_evict,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
index f159b55..423b097 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
@@ -335,4 +335,5 @@ gf117_grctx = {
 	.r4060a8 = gf100_grctx_generate_r4060a8,
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
 	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
+	.max_ways_evict = gf100_grctx_generate_max_ways_evict,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c
index 3720afd..6b2fcfe 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c
@@ -522,4 +522,5 @@ gf119_grctx = {
 	.r4060a8 = gf100_grctx_generate_r4060a8,
 	.rop_mapping = gf100_grctx_generate_rop_mapping,
 	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
+	.max_ways_evict = gf100_grctx_generate_max_ways_evict,
 };
-- 
cgit v1.1


From 60770fa28bd7d69097d3a186fe8cfa1ec21c9c1d Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise dist_skip_table + improve algorithm

The algorithm for GM200 and newer matches RM for all the boards I have, but
I don't have enough data to try and figure something out for earlier boards,
so these will still write zeroes to the table as we did before.

The code in NVGPU isn't helpful here, it appears to handle specific cases.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c  |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h  |  3 +++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c  | 15 +++++++++----
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c  |  5 +----
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c  |  5 +----
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c  | 25 ++++++++++++++++++++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c  |  3 +--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c     |  3 +++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h     |  1 +
 14 files changed, 51 insertions(+), 16 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index cdf74f3..176be71 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1360,6 +1360,8 @@ gf100_grctx_generate_floorsweep(struct gf100_gr *gr)
 		func->alpha_beta_tables(gr);
 	if (func->max_ways_evict)
 		func->max_ways_evict(gr);
+	if (func->dist_skip_table)
+		func->dist_skip_table(gr);
 }
 
 void
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
index 41cb875..dd1c73b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
@@ -57,6 +57,7 @@ struct gf100_grctx_func {
 	void (*rop_mapping)(struct gf100_gr *);
 	void (*alpha_beta_tables)(struct gf100_gr *);
 	void (*max_ways_evict)(struct gf100_gr *);
+	void (*dist_skip_table)(struct gf100_gr *);
 };
 
 extern const struct gf100_grctx_func gf100_grctx;
@@ -84,6 +85,7 @@ extern const struct gf100_grctx_func gf110_grctx;
 extern const struct gf100_grctx_func gf117_grctx;
 void gf117_grctx_generate_attrib(struct gf100_grctx *);
 void gf117_grctx_generate_rop_mapping(struct gf100_gr *);
+void gf117_grctx_generate_dist_skip_table(struct gf100_gr *);
 
 extern const struct gf100_grctx_func gf119_grctx;
 
@@ -112,6 +114,7 @@ void gm107_grctx_generate_pagepool(struct gf100_grctx *);
 void gm107_grctx_generate_attrib(struct gf100_grctx *);
 
 extern const struct gf100_grctx_func gm200_grctx;
+void gm200_grctx_generate_dist_skip_table(struct gf100_gr *);
 void gm200_grctx_generate_405b60(struct gf100_gr *);
 
 extern const struct gf100_grctx_func gm20b_grctx;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
index 423b097..b3f4127 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
@@ -180,6 +180,16 @@ gf117_grctx_pack_ppc[] = {
  ******************************************************************************/
 
 void
+gf117_grctx_generate_dist_skip_table(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	int i;
+
+	for (i = 0; i < 8; i++)
+		nvkm_wr32(device, 0x4064d0 + (i * 0x04), 0x00000000);
+}
+
+void
 gf117_grctx_generate_rop_mapping(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -282,7 +292,6 @@ gf117_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	struct nvkm_device *device = gr->base.engine.subdev.device;
 	const struct gf100_grctx_func *grctx = gr->func->grctx;
 	u32 idle_timeout;
-	int i;
 
 	nvkm_mc_unk260(device, 0);
 
@@ -301,9 +310,6 @@ gf117_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	gf100_grctx_generate_floorsweep(gr);
 
-	for (i = 0; i < 8; i++)
-		nvkm_wr32(device, 0x4064d0 + (i * 0x04), 0x00000000);
-
 	gf100_gr_icmd(gr, grctx->icmd);
 	nvkm_wr32(device, 0x404154, idle_timeout);
 	gf100_gr_mthd(gr, grctx->mthd);
@@ -336,4 +342,5 @@ gf117_grctx = {
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
 	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
 	.max_ways_evict = gf100_grctx_generate_max_ways_evict,
+	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
index 25576c1..1216931 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
@@ -898,7 +898,6 @@ gk104_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	struct nvkm_device *device = gr->base.engine.subdev.device;
 	const struct gf100_grctx_func *grctx = gr->func->grctx;
 	u32 idle_timeout;
-	int i;
 
 	nvkm_mc_unk260(device, 0);
 
@@ -917,9 +916,6 @@ gk104_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	gf100_grctx_generate_floorsweep(gr);
 
-	for (i = 0; i < 8; i++)
-		nvkm_wr32(device, 0x4064d0 + (i * 0x04), 0x00000000);
-
 	nvkm_wr32(device, 0x405b00, (gr->tpc_total << 8) | gr->gpc_nr);
 	nvkm_mask(device, 0x419f78, 0x00000001, 0x00000000);
 
@@ -1006,4 +1002,5 @@ gk104_grctx = {
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
 	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
+	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
index 284570a..e6a54dc 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
@@ -835,4 +835,5 @@ gk110_grctx = {
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
 	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
+	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
index ffd8cf9..ef82ebe 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
@@ -96,4 +96,5 @@ gk110b_grctx = {
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
 	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
+	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
index e5e4d4dc..226f8aa 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
@@ -557,4 +557,5 @@ gk208_grctx = {
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
 	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
+	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
index c209bf3..cdf9d60 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
@@ -945,7 +945,6 @@ gm107_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	struct nvkm_device *device = gr->base.engine.subdev.device;
 	const struct gf100_grctx_func *grctx = gr->func->grctx;
 	u32 idle_timeout;
-	int i;
 
 	gf100_gr_mmio(gr, grctx->hub);
 	gf100_gr_mmio(gr, grctx->gpc);
@@ -962,9 +961,6 @@ gm107_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	gf100_grctx_generate_floorsweep(gr);
 
-	nvkm_wr32(device, 0x4064d0, 0x00000001);
-	for (i = 1; i < 8; i++)
-		nvkm_wr32(device, 0x4064d0 + (i * 0x04), 0x00000000);
 	nvkm_wr32(device, 0x406500, 0x00000001);
 
 	nvkm_wr32(device, 0x405b00, (gr->tpc_total << 8) | gr->gpc_nr);
@@ -1005,4 +1001,5 @@ gm107_grctx = {
 	.tpc_nr = gf100_grctx_generate_tpc_nr,
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
 	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
+	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
index cfccd75..6891206 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
@@ -78,8 +78,6 @@ gm200_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	gf100_grctx_generate_floorsweep(gr);
 
-	for (i = 0; i < 8; i++)
-		nvkm_wr32(device, 0x4064d0 + (i * 0x04), 0x00000000);
 	nvkm_wr32(device, 0x406500, 0x00000000);
 
 	nvkm_wr32(device, 0x405b00, (gr->tpc_total << 8) | gr->gpc_nr);
@@ -98,6 +96,28 @@ gm200_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	nvkm_mask(device, 0x418e4c, 0xffffffff, 0x70000000);
 }
 
+void
+gm200_grctx_generate_dist_skip_table(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	u32 data[8] = {};
+	int gpc, ppc, i;
+
+	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
+		for (ppc = 0; ppc < gr->ppc_nr[gpc]; ppc++) {
+			u8 ppc_tpcs = gr->ppc_tpc_nr[gpc][ppc];
+			u8 ppc_tpcm = gr->ppc_tpc_mask[gpc][ppc];
+			while (ppc_tpcs-- > gr->ppc_tpc_min)
+				ppc_tpcm &= ppc_tpcm - 1;
+			ppc_tpcm ^= gr->ppc_tpc_mask[gpc][ppc];
+			((u8 *)data)[gpc] |= ppc_tpcm;
+		}
+	}
+
+	for (i = 0; i < ARRAY_SIZE(data); i++)
+		nvkm_wr32(device, 0x4064d0 + (i * 0x04), data[i]);
+}
+
 const struct gf100_grctx_func
 gm200_grctx = {
 	.main  = gm200_grctx_generate_main,
@@ -115,4 +135,5 @@ gm200_grctx = {
 	.alpha_nr = 0x1000,
 	.sm_id = gm107_grctx_generate_sm_id,
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
+	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
index e099907..1a3d0c5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
@@ -140,8 +140,6 @@ gp100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	gf100_grctx_generate_floorsweep(gr);
 
-	for (i = 0; i < 8; i++)
-		nvkm_wr32(device, 0x4064d0 + (i * 0x04), 0x00000000);
 	nvkm_wr32(device, 0x406500, 0x00000000);
 
 	nvkm_wr32(device, 0x405b00, (gr->tpc_total << 8) | gr->gpc_nr);
@@ -174,4 +172,5 @@ gp100_grctx = {
 	.alpha_nr = 0x800,
 	.sm_id = gm107_grctx_generate_sm_id,
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
+	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
index 553a609..2aeabb3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
@@ -96,4 +96,5 @@ gp102_grctx = {
 	.alpha_nr = 0x800,
 	.sm_id = gm107_grctx_generate_sm_id,
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
+	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
index db3fff8..4aea2f6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
@@ -46,4 +46,5 @@ gp107_grctx = {
 	.alpha_nr = 0x800,
 	.sm_id = gm107_grctx_generate_sm_id,
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
+	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index fe3b44d..dd4a410 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -1685,6 +1685,9 @@ gf100_gr_oneinit(struct nvkm_gr *base)
 				continue;
 			gr->ppc_mask[i] |= (1 << j);
 			gr->ppc_tpc_nr[i][j] = hweight8(gr->ppc_tpc_mask[i][j]);
+			if (gr->ppc_tpc_min == 0 ||
+			    gr->ppc_tpc_min > gr->ppc_tpc_nr[i][j])
+				gr->ppc_tpc_min = gr->ppc_tpc_nr[i][j];
 		}
 	}
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index 6f7a786..c2a1b2a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -105,6 +105,7 @@ struct gf100_gr {
 	u8 ppc_mask[GPC_MAX];
 	u8 ppc_tpc_mask[GPC_MAX][4];
 	u8 ppc_tpc_nr[GPC_MAX][4];
+	u8 ppc_tpc_min;
 
 	struct gf100_gr_data mmio_data[4];
 	struct gf100_gr_mmio mmio_list[4096/8];
-- 
cgit v1.1


From e7163b192226206bc350a09a52603bd103a5ff6a Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise r406500

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c | 2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h | 2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c | 9 +++++++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c | 9 +++++++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c | 3 +--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c | 1 +
 7 files changed, 21 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index 176be71..7bd28c1 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1362,6 +1362,8 @@ gf100_grctx_generate_floorsweep(struct gf100_gr *gr)
 		func->max_ways_evict(gr);
 	if (func->dist_skip_table)
 		func->dist_skip_table(gr);
+	if (func->r406500)
+		func->r406500(gr);
 }
 
 void
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
index dd1c73b..7d949a5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
@@ -58,6 +58,7 @@ struct gf100_grctx_func {
 	void (*alpha_beta_tables)(struct gf100_gr *);
 	void (*max_ways_evict)(struct gf100_gr *);
 	void (*dist_skip_table)(struct gf100_gr *);
+	void (*r406500)(struct gf100_gr *);
 };
 
 extern const struct gf100_grctx_func gf100_grctx;
@@ -115,6 +116,7 @@ void gm107_grctx_generate_attrib(struct gf100_grctx *);
 
 extern const struct gf100_grctx_func gm200_grctx;
 void gm200_grctx_generate_dist_skip_table(struct gf100_gr *);
+void gm200_grctx_generate_r406500(struct gf100_gr *);
 void gm200_grctx_generate_405b60(struct gf100_gr *);
 
 extern const struct gf100_grctx_func gm20b_grctx;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
index cdf9d60..aa5fff3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
@@ -930,6 +930,12 @@ gm107_grctx_generate_attrib(struct gf100_grctx *info)
 	}
 }
 
+static void
+gm107_grctx_generate_r406500(struct gf100_gr *gr)
+{
+	nvkm_wr32(gr->base.engine.subdev.device, 0x406500, 0x00000001);
+}
+
 void
 gm107_grctx_generate_sm_id(struct gf100_gr *gr, int gpc, int tpc, int sm)
 {
@@ -961,8 +967,6 @@ gm107_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	gf100_grctx_generate_floorsweep(gr);
 
-	nvkm_wr32(device, 0x406500, 0x00000001);
-
 	nvkm_wr32(device, 0x405b00, (gr->tpc_total << 8) | gr->gpc_nr);
 
 	gf100_gr_icmd(gr, grctx->icmd);
@@ -1002,4 +1006,5 @@ gm107_grctx = {
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
 	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
 	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
+	.r406500 = gm107_grctx_generate_r406500,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
index 6891206..9c4db04 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
@@ -59,6 +59,12 @@ gm200_grctx_generate_405b60(struct gf100_gr *gr)
 		nvkm_wr32(device, 0x405ba0 + (i * 4), gpcs[i]);
 }
 
+void
+gm200_grctx_generate_r406500(struct gf100_gr *gr)
+{
+	nvkm_wr32(gr->base.engine.subdev.device, 0x406500, 0x00000000);
+}
+
 static void
 gm200_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 {
@@ -78,8 +84,6 @@ gm200_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	gf100_grctx_generate_floorsweep(gr);
 
-	nvkm_wr32(device, 0x406500, 0x00000000);
-
 	nvkm_wr32(device, 0x405b00, (gr->tpc_total << 8) | gr->gpc_nr);
 
 	for (tmp = 0, i = 0; i < gr->gpc_nr; i++)
@@ -136,4 +140,5 @@ gm200_grctx = {
 	.sm_id = gm107_grctx_generate_sm_id,
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
 	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
+	.r406500 = gm200_grctx_generate_r406500,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
index 1a3d0c5..3bad1a5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
@@ -140,8 +140,6 @@ gp100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	gf100_grctx_generate_floorsweep(gr);
 
-	nvkm_wr32(device, 0x406500, 0x00000000);
-
 	nvkm_wr32(device, 0x405b00, (gr->tpc_total << 8) | gr->gpc_nr);
 
 	for (tmp = 0, i = 0; i < gr->gpc_nr; i++)
@@ -173,4 +171,5 @@ gp100_grctx = {
 	.sm_id = gm107_grctx_generate_sm_id,
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
 	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
+	.r406500 = gm200_grctx_generate_r406500,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
index 2aeabb3..dea0097 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
@@ -97,4 +97,5 @@ gp102_grctx = {
 	.sm_id = gm107_grctx_generate_sm_id,
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
 	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
+	.r406500 = gm200_grctx_generate_r406500,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
index 4aea2f6..a97c4b0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
@@ -47,4 +47,5 @@ gp107_grctx = {
 	.sm_id = gm107_grctx_generate_sm_id,
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
 	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
+	.r406500 = gm200_grctx_generate_r406500,
 };
-- 
cgit v1.1


From 60c0264a667fe80ac48d746d073e9d869a5d52f0 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise gpc_tpc_nr

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c  | 2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h  | 2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c  | 9 ++++++++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c  | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c  | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c  | 3 +--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c  | 3 +--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c  | 3 +--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c  | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c  | 1 +
 11 files changed, 20 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index 7bd28c1..c1c83e5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1364,6 +1364,8 @@ gf100_grctx_generate_floorsweep(struct gf100_gr *gr)
 		func->dist_skip_table(gr);
 	if (func->r406500)
 		func->r406500(gr);
+	if (func->gpc_tpc_nr)
+		func->gpc_tpc_nr(gr);
 }
 
 void
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
index 7d949a5..c91904d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
@@ -59,6 +59,7 @@ struct gf100_grctx_func {
 	void (*max_ways_evict)(struct gf100_gr *);
 	void (*dist_skip_table)(struct gf100_gr *);
 	void (*r406500)(struct gf100_gr *);
+	void (*gpc_tpc_nr)(struct gf100_gr *);
 };
 
 extern const struct gf100_grctx_func gf100_grctx;
@@ -92,6 +93,7 @@ extern const struct gf100_grctx_func gf119_grctx;
 
 extern const struct gf100_grctx_func gk104_grctx;
 void gk104_grctx_generate_alpha_beta_tables(struct gf100_gr *);
+void gk104_grctx_generate_gpc_tpc_nr(struct gf100_gr *);
 
 extern const struct gf100_grctx_func gk20a_grctx;
 void gk104_grctx_generate_main(struct gf100_gr *, struct gf100_grctx *);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
index 1216931..302b8c7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
@@ -893,6 +893,13 @@ gk104_grctx_generate_unkn(struct gf100_gr *gr)
 }
 
 void
+gk104_grctx_generate_gpc_tpc_nr(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, 0x405b00, (gr->tpc_total << 8) | gr->gpc_nr);
+}
+
+void
 gk104_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -916,7 +923,6 @@ gk104_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	gf100_grctx_generate_floorsweep(gr);
 
-	nvkm_wr32(device, 0x405b00, (gr->tpc_total << 8) | gr->gpc_nr);
 	nvkm_mask(device, 0x419f78, 0x00000001, 0x00000000);
 
 	gf100_gr_icmd(gr, grctx->icmd);
@@ -1003,4 +1009,5 @@ gk104_grctx = {
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
 	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
 	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
+	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
index e6a54dc..64e00d7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
@@ -836,4 +836,5 @@ gk110_grctx = {
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
 	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
 	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
+	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
index ef82ebe..f800ed5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
@@ -97,4 +97,5 @@ gk110b_grctx = {
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
 	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
 	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
+	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
index 226f8aa..494d9a9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
@@ -558,4 +558,5 @@ gk208_grctx = {
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
 	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
 	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
+	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
index aa5fff3..6b279bb 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
@@ -967,8 +967,6 @@ gm107_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	gf100_grctx_generate_floorsweep(gr);
 
-	nvkm_wr32(device, 0x405b00, (gr->tpc_total << 8) | gr->gpc_nr);
-
 	gf100_gr_icmd(gr, grctx->icmd);
 	nvkm_wr32(device, 0x404154, idle_timeout);
 	gf100_gr_mthd(gr, grctx->mthd);
@@ -1007,4 +1005,5 @@ gm107_grctx = {
 	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
 	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
 	.r406500 = gm107_grctx_generate_r406500,
+	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
index 9c4db04..3ba5e95 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
@@ -84,8 +84,6 @@ gm200_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	gf100_grctx_generate_floorsweep(gr);
 
-	nvkm_wr32(device, 0x405b00, (gr->tpc_total << 8) | gr->gpc_nr);
-
 	for (tmp = 0, i = 0; i < gr->gpc_nr; i++)
 		tmp |= ((1 << gr->tpc_nr[i]) - 1) << (i * 4);
 	nvkm_wr32(device, 0x4041c4, tmp);
@@ -141,4 +139,5 @@ gm200_grctx = {
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
 	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
 	.r406500 = gm200_grctx_generate_r406500,
+	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
index 3bad1a5..c48617b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
@@ -140,8 +140,6 @@ gp100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	gf100_grctx_generate_floorsweep(gr);
 
-	nvkm_wr32(device, 0x405b00, (gr->tpc_total << 8) | gr->gpc_nr);
-
 	for (tmp = 0, i = 0; i < gr->gpc_nr; i++)
 		tmp |= ((1 << gr->tpc_nr[i]) - 1) << (i * 5);
 	nvkm_wr32(device, 0x4041c4, tmp);
@@ -172,4 +170,5 @@ gp100_grctx = {
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
 	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
 	.r406500 = gm200_grctx_generate_r406500,
+	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
index dea0097..ec4fbe8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
@@ -98,4 +98,5 @@ gp102_grctx = {
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
 	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
 	.r406500 = gm200_grctx_generate_r406500,
+	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
index a97c4b0..84c98cd 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
@@ -48,4 +48,5 @@ gp107_grctx = {
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
 	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
 	.r406500 = gm200_grctx_generate_r406500,
+	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
 };
-- 
cgit v1.1


From aa5e38dc9fdf0a11724561777d712bfdf0d6ad99 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise r419f78 + apply fixes from traces

Removed from GK110[B]/GK208 as RM traces show it not being touched.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c | 10 ++++++++--
 3 files changed, 11 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index c1c83e5..3793d48 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1366,6 +1366,8 @@ gf100_grctx_generate_floorsweep(struct gf100_gr *gr)
 		func->r406500(gr);
 	if (func->gpc_tpc_nr)
 		func->gpc_tpc_nr(gr);
+	if (func->r419f78)
+		func->r419f78(gr);
 }
 
 void
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
index c91904d..e84b46f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
@@ -60,6 +60,7 @@ struct gf100_grctx_func {
 	void (*dist_skip_table)(struct gf100_gr *);
 	void (*r406500)(struct gf100_gr *);
 	void (*gpc_tpc_nr)(struct gf100_gr *);
+	void (*r419f78)(struct gf100_gr *);
 };
 
 extern const struct gf100_grctx_func gf100_grctx;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
index 302b8c7..f527bca 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
@@ -892,6 +892,13 @@ gk104_grctx_generate_unkn(struct gf100_gr *gr)
 	nvkm_mask(device, 0x419c00, 0x00000008, 0x00000008);
 }
 
+static void
+gk104_grctx_generate_r419f78(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x419f78, 0x00000001, 0x00000000);
+}
+
 void
 gk104_grctx_generate_gpc_tpc_nr(struct gf100_gr *gr)
 {
@@ -923,8 +930,6 @@ gk104_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	gf100_grctx_generate_floorsweep(gr);
 
-	nvkm_mask(device, 0x419f78, 0x00000001, 0x00000000);
-
 	gf100_gr_icmd(gr, grctx->icmd);
 	nvkm_wr32(device, 0x404154, idle_timeout);
 	gf100_gr_mthd(gr, grctx->mthd);
@@ -1010,4 +1015,5 @@ gk104_grctx = {
 	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
 	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
 	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
+	.r419f78 = gk104_grctx_generate_r419f78,
 };
-- 
cgit v1.1


From fc36076441bae141893bd79899d19aa1b5fdf524 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:46 +1000
Subject: drm/nouveau/gr/gf100-: virtualise tpc_mask + apply fixes from traces

We weren't placing higher TPC IDs in the right place on some configurations.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c |  4 ++++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h | 12 ++++++------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c | 22 +++++++++++++---------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm20b.c |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c | 22 ++++++++--------------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h    |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c    |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c    |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c    |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c    |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c    |  2 ++
 13 files changed, 47 insertions(+), 30 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index 3793d48..a52f27f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1368,6 +1368,10 @@ gf100_grctx_generate_floorsweep(struct gf100_gr *gr)
 		func->gpc_tpc_nr(gr);
 	if (func->r419f78)
 		func->r419f78(gr);
+	if (func->tpc_mask)
+		func->tpc_mask(gr);
+	if (func->smid_config)
+		func->smid_config(gr);
 }
 
 void
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
index e84b46f..d319e76 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
@@ -61,6 +61,8 @@ struct gf100_grctx_func {
 	void (*r406500)(struct gf100_gr *);
 	void (*gpc_tpc_nr)(struct gf100_gr *);
 	void (*r419f78)(struct gf100_gr *);
+	void (*tpc_mask)(struct gf100_gr *);
+	void (*smid_config)(struct gf100_gr *);
 };
 
 extern const struct gf100_grctx_func gf100_grctx;
@@ -103,11 +105,6 @@ void gk104_grctx_generate_pagepool(struct gf100_grctx *);
 void gk104_grctx_generate_patch_ltc(struct gf100_grctx *);
 void gk104_grctx_generate_unkn(struct gf100_gr *);
 
-void gm107_grctx_generate_bundle(struct gf100_grctx *);
-void gm107_grctx_generate_pagepool(struct gf100_grctx *);
-void gm107_grctx_generate_attrib(struct gf100_grctx *);
-void gm107_grctx_generate_sm_id(struct gf100_gr *, int, int, int);
-
 extern const struct gf100_grctx_func gk110_grctx;
 extern const struct gf100_grctx_func gk110b_grctx;
 extern const struct gf100_grctx_func gk208_grctx;
@@ -116,17 +113,20 @@ extern const struct gf100_grctx_func gm107_grctx;
 void gm107_grctx_generate_bundle(struct gf100_grctx *);
 void gm107_grctx_generate_pagepool(struct gf100_grctx *);
 void gm107_grctx_generate_attrib(struct gf100_grctx *);
+void gm107_grctx_generate_sm_id(struct gf100_gr *, int, int, int);
 
 extern const struct gf100_grctx_func gm200_grctx;
 void gm200_grctx_generate_dist_skip_table(struct gf100_gr *);
 void gm200_grctx_generate_r406500(struct gf100_gr *);
-void gm200_grctx_generate_405b60(struct gf100_gr *);
+void gm200_grctx_generate_tpc_mask(struct gf100_gr *);
+void gm200_grctx_generate_smid_config(struct gf100_gr *);
 
 extern const struct gf100_grctx_func gm20b_grctx;
 
 extern const struct gf100_grctx_func gp100_grctx;
 void gp100_grctx_generate_main(struct gf100_gr *, struct gf100_grctx *);
 void gp100_grctx_generate_pagepool(struct gf100_grctx *);
+void gp100_grctx_generate_smid_config(struct gf100_gr *);
 
 extern const struct gf100_grctx_func gp102_grctx;
 void gp102_grctx_generate_attrib(struct gf100_grctx *);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
index 3ba5e95..f1e87b9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
@@ -28,7 +28,7 @@
  ******************************************************************************/
 
 void
-gm200_grctx_generate_405b60(struct gf100_gr *gr)
+gm200_grctx_generate_smid_config(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
 	const u32 dist_nr = DIV_ROUND_UP(gr->tpc_total, 4);
@@ -60,6 +60,15 @@ gm200_grctx_generate_405b60(struct gf100_gr *gr)
 }
 
 void
+gm200_grctx_generate_tpc_mask(struct gf100_gr *gr)
+{
+	u32 tmp, i;
+	for (tmp = 0, i = 0; i < gr->gpc_nr; i++)
+		tmp |= ((1 << gr->tpc_nr[i]) - 1) << (i * gr->func->tpc_nr);
+	nvkm_wr32(gr->base.engine.subdev.device, 0x4041c4, tmp);
+}
+
+void
 gm200_grctx_generate_r406500(struct gf100_gr *gr)
 {
 	nvkm_wr32(gr->base.engine.subdev.device, 0x406500, 0x00000000);
@@ -70,8 +79,7 @@ gm200_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
 	const struct gf100_grctx_func *grctx = gr->func->grctx;
-	u32 idle_timeout, tmp;
-	int i;
+	u32 idle_timeout;
 
 	gf100_gr_mmio(gr, gr->fuc_sw_ctx);
 
@@ -84,12 +92,6 @@ gm200_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	gf100_grctx_generate_floorsweep(gr);
 
-	for (tmp = 0, i = 0; i < gr->gpc_nr; i++)
-		tmp |= ((1 << gr->tpc_nr[i]) - 1) << (i * 4);
-	nvkm_wr32(device, 0x4041c4, tmp);
-
-	gm200_grctx_generate_405b60(gr);
-
 	gf100_gr_icmd(gr, gr->fuc_bundle);
 	nvkm_wr32(device, 0x404154, idle_timeout);
 	gf100_gr_mthd(gr, gr->fuc_method);
@@ -140,4 +142,6 @@ gm200_grctx = {
 	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
 	.r406500 = gm200_grctx_generate_r406500,
 	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
+	.tpc_mask = gm200_grctx_generate_tpc_mask,
+	.smid_config = gm200_grctx_generate_smid_config,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm20b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm20b.c
index 3dd4e18..a1d9e11 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm20b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm20b.c
@@ -52,7 +52,7 @@ gm20b_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 		tmp |= ((1 << gr->tpc_nr[i]) - 1) << (i * 4);
 	nvkm_wr32(device, 0x4041c4, tmp);
 
-	gm200_grctx_generate_405b60(gr);
+	gm200_grctx_generate_smid_config(gr);
 
 	gf100_gr_wait_idle(gr);
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
index c48617b..821219a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
@@ -89,13 +89,12 @@ gp100_grctx_generate_attrib(struct gf100_grctx *info)
 	mmio_wr32(info, 0x41befc, 0x00000000);
 }
 
-static void
-gp100_grctx_generate_405b60(struct gf100_gr *gr)
+void
+gp100_grctx_generate_smid_config(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
 	const u32 dist_nr = DIV_ROUND_UP(gr->tpc_total, 4);
-	u32 dist[TPC_MAX / 4] = {};
-	u32 gpcs[GPC_MAX * 2] = {};
+	u32 dist[TPC_MAX / 4] = {}, gpcs[16] = {};
 	u8  tpcnr[GPC_MAX];
 	int tpc, gpc, i;
 
@@ -112,12 +111,12 @@ gp100_grctx_generate_405b60(struct gf100_gr *gr)
 		tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--;
 
 		dist[i / 4] |= ((gpc << 4) | tpc) << ((i % 4) * 8);
-		gpcs[gpc + (gr->gpc_nr * (tpc / 4))] |= i << (tpc * 8);
+		gpcs[gpc + (gr->func->gpc_nr * (tpc / 4))] |= i << (tpc * 8);
 	}
 
 	for (i = 0; i < dist_nr; i++)
 		nvkm_wr32(device, 0x405b60 + (i * 4), dist[i]);
-	for (i = 0; i < gr->gpc_nr * 2; i++)
+	for (i = 0; i < ARRAY_SIZE(gpcs); i++)
 		nvkm_wr32(device, 0x405ba0 + (i * 4), gpcs[i]);
 }
 
@@ -126,8 +125,7 @@ gp100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
 	const struct gf100_grctx_func *grctx = gr->func->grctx;
-	u32 idle_timeout, tmp;
-	int i;
+	u32 idle_timeout;
 
 	gf100_gr_mmio(gr, gr->fuc_sw_ctx);
 
@@ -140,12 +138,6 @@ gp100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	gf100_grctx_generate_floorsweep(gr);
 
-	for (tmp = 0, i = 0; i < gr->gpc_nr; i++)
-		tmp |= ((1 << gr->tpc_nr[i]) - 1) << (i * 5);
-	nvkm_wr32(device, 0x4041c4, tmp);
-
-	gp100_grctx_generate_405b60(gr);
-
 	gf100_gr_icmd(gr, gr->fuc_bundle);
 	nvkm_wr32(device, 0x404154, idle_timeout);
 	gf100_gr_mthd(gr, gr->fuc_method);
@@ -171,4 +163,6 @@ gp100_grctx = {
 	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
 	.r406500 = gm200_grctx_generate_r406500,
 	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
+	.tpc_mask = gm200_grctx_generate_tpc_mask,
+	.smid_config = gp100_grctx_generate_smid_config,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
index ec4fbe8..611819f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
@@ -99,4 +99,6 @@ gp102_grctx = {
 	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
 	.r406500 = gm200_grctx_generate_r406500,
 	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
+	.tpc_mask = gm200_grctx_generate_tpc_mask,
+	.smid_config = gp100_grctx_generate_smid_config,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
index 84c98cd..d908317 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
@@ -49,4 +49,6 @@ gp107_grctx = {
 	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
 	.r406500 = gm200_grctx_generate_r406500,
 	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
+	.tpc_mask = gm200_grctx_generate_tpc_mask,
+	.smid_config = gp100_grctx_generate_smid_config,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index c2a1b2a..31109ce 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -154,6 +154,8 @@ struct gf100_gr_func {
 		struct gf100_gr_ucode *ucode;
 	} gpccs;
 	int (*rops)(struct gf100_gr *);
+	int gpc_nr;
+	int tpc_nr;
 	int ppc_nr;
 	const struct gf100_grctx_func *grctx;
 	const struct nvkm_therm_clkgate_pack *clkgate_pack;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
index 4dcb56b..ae0eaf8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
@@ -134,6 +134,7 @@ gm200_gr = {
 	.init_shader_exceptions = gm107_gr_init_shader_exceptions,
 	.init_400054 = gm107_gr_init_400054,
 	.rops = gm200_gr_rops,
+	.tpc_nr = 4,
 	.ppc_nr = 2,
 	.grctx = &gm200_grctx,
 	.sclass = {
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
index cc507e8..3addbc1 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
@@ -80,6 +80,8 @@ gp100_gr = {
 	.init_504430 = gm107_gr_init_504430,
 	.init_shader_exceptions = gp100_gr_init_shader_exceptions,
 	.rops = gm200_gr_rops,
+	.gpc_nr = 6,
+	.tpc_nr = 5,
 	.ppc_nr = 2,
 	.grctx = &gp100_grctx,
 	.sclass = {
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
index 86d1ff7..ea99c15 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
@@ -58,6 +58,8 @@ gp102_gr = {
 	.init_504430 = gm107_gr_init_504430,
 	.init_shader_exceptions = gp100_gr_init_shader_exceptions,
 	.rops = gm200_gr_rops,
+	.gpc_nr = 6,
+	.tpc_nr = 5,
 	.ppc_nr = 3,
 	.grctx = &gp102_grctx,
 	.sclass = {
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
index 14007b5..09cba53 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
@@ -44,6 +44,8 @@ gp107_gr = {
 	.init_504430 = gm107_gr_init_504430,
 	.init_shader_exceptions = gp100_gr_init_shader_exceptions,
 	.rops = gm200_gr_rops,
+	.gpc_nr = 2,
+	.tpc_nr = 3,
 	.ppc_nr = 1,
 	.grctx = &gp107_grctx,
 	.sclass = {
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
index 450a96d..4972bf8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
@@ -42,6 +42,8 @@ gp10b_gr = {
 	.init_504430 = gm107_gr_init_504430,
 	.init_shader_exceptions = gp100_gr_init_shader_exceptions,
 	.rops = gm200_gr_rops,
+	.gpc_nr = 1,
+	.tpc_nr = 2,
 	.ppc_nr = 1,
 	.grctx = &gp102_grctx,
 	.sclass = {
-- 
cgit v1.1


From 0e5a5e86f3edb0845f6caf8f9819eebb26ec040a Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/gr/gf100-: support firmware-provided bundle/method
 everywhere

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c | 12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index a52f27f..5350ab0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1404,9 +1404,17 @@ gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	gf100_grctx_generate_floorsweep(gr);
 
-	gf100_gr_icmd(gr, grctx->icmd);
+	if (gr->fuc_bundle)
+		gf100_gr_icmd(gr, gr->fuc_bundle);
+	else
+		gf100_gr_icmd(gr, grctx->icmd);
+
 	nvkm_wr32(device, 0x404154, idle_timeout);
-	gf100_gr_mthd(gr, grctx->mthd);
+
+	if (gr->fuc_method)
+		gf100_gr_mthd(gr, gr->fuc_method);
+	else
+		gf100_gr_mthd(gr, grctx->mthd);
 	nvkm_mc_unk260(device, 1);
 }
 
-- 
cgit v1.1


From 99a3c67e84c955736a109e645371c7132c1188f2 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/gr/gf100-gf119: update 419cb8 where required

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c | 11 +++++++++++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h |  3 +++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c |  1 +
 7 files changed, 19 insertions(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index 5350ab0..e29ac88 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1025,6 +1025,13 @@ gf100_grctx_mmio_item(struct gf100_grctx *info, u32 addr, u32 data,
 }
 
 void
+gf100_grctx_generate_r419cb8(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x419cb8, 0x00007c00, 0x00000000);
+}
+
+void
 gf100_grctx_generate_bundle(struct gf100_grctx *info)
 {
 	const struct gf100_grctx_func *grctx = info->gr->func->grctx;
@@ -1416,6 +1423,9 @@ gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	else
 		gf100_gr_mthd(gr, grctx->mthd);
 	nvkm_mc_unk260(device, 1);
+
+	if (grctx->r419cb8)
+		grctx->r419cb8(gr);
 }
 
 #define CB_RESERVED 0x80000
@@ -1564,4 +1574,5 @@ gf100_grctx = {
 	.rop_mapping = gf100_grctx_generate_rop_mapping,
 	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
 	.max_ways_evict = gf100_grctx_generate_max_ways_evict,
+	.r419cb8 = gf100_grctx_generate_r419cb8,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
index d319e76..67628b9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
@@ -63,6 +63,8 @@ struct gf100_grctx_func {
 	void (*r419f78)(struct gf100_gr *);
 	void (*tpc_mask)(struct gf100_gr *);
 	void (*smid_config)(struct gf100_gr *);
+	/* misc other things */
+	void (*r419cb8)(struct gf100_gr *);
 };
 
 extern const struct gf100_grctx_func gf100_grctx;
@@ -79,6 +81,7 @@ void gf100_grctx_generate_r4060a8(struct gf100_gr *);
 void gf100_grctx_generate_rop_mapping(struct gf100_gr *);
 void gf100_grctx_generate_alpha_beta_tables(struct gf100_gr *);
 void gf100_grctx_generate_max_ways_evict(struct gf100_gr *);
+void gf100_grctx_generate_r419cb8(struct gf100_gr *);
 
 extern const struct gf100_grctx_func gf108_grctx;
 void gf108_grctx_generate_attrib(struct gf100_grctx *);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c
index d246bb6..7f3b928 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c
@@ -102,4 +102,5 @@ gf104_grctx = {
 	.rop_mapping = gf100_grctx_generate_rop_mapping,
 	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
 	.max_ways_evict = gf100_grctx_generate_max_ways_evict,
+	.r419cb8 = gf100_grctx_generate_r419cb8,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c
index 3b6e645..369d64f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c
@@ -800,4 +800,5 @@ gf108_grctx = {
 	.rop_mapping = gf100_grctx_generate_rop_mapping,
 	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
 	.max_ways_evict = gf100_grctx_generate_max_ways_evict,
+	.r419cb8 = gf100_grctx_generate_r419cb8,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c
index b65a1f3..d59c248 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c
@@ -353,4 +353,5 @@ gf110_grctx = {
 	.rop_mapping = gf100_grctx_generate_rop_mapping,
 	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
 	.max_ways_evict = gf100_grctx_generate_max_ways_evict,
+	.r419cb8 = gf100_grctx_generate_r419cb8,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
index b3f4127..e922e3e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
@@ -343,4 +343,5 @@ gf117_grctx = {
 	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
 	.max_ways_evict = gf100_grctx_generate_max_ways_evict,
 	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
+	.r419cb8 = gf100_grctx_generate_r419cb8,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c
index 6b2fcfe..cc1a935 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c
@@ -523,4 +523,5 @@ gf119_grctx = {
 	.rop_mapping = gf100_grctx_generate_rop_mapping,
 	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
 	.max_ways_evict = gf100_grctx_generate_max_ways_evict,
+	.r419cb8 = gf100_grctx_generate_r419cb8,
 };
-- 
cgit v1.1


From 5b54b5b92543cac6b9f3728935869fafdb3bf0d9 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/gr/gf100-: note missing 418800 modifications

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c  |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h  |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c  | 19 +++++++++++++++++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c  |  1 +
 6 files changed, 24 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index e29ac88..0dcb227 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1426,6 +1426,8 @@ gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	if (grctx->r419cb8)
 		grctx->r419cb8(gr);
+	if (grctx->r418800)
+		grctx->r418800(gr);
 }
 
 #define CB_RESERVED 0x80000
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
index 67628b9..0d05664 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
@@ -65,6 +65,7 @@ struct gf100_grctx_func {
 	void (*smid_config)(struct gf100_gr *);
 	/* misc other things */
 	void (*r419cb8)(struct gf100_gr *);
+	void (*r418800)(struct gf100_gr *);
 };
 
 extern const struct gf100_grctx_func gf100_grctx;
@@ -107,6 +108,7 @@ void gk104_grctx_generate_bundle(struct gf100_grctx *);
 void gk104_grctx_generate_pagepool(struct gf100_grctx *);
 void gk104_grctx_generate_patch_ltc(struct gf100_grctx *);
 void gk104_grctx_generate_unkn(struct gf100_gr *);
+void gk104_grctx_generate_r418800(struct gf100_gr *);
 
 extern const struct gf100_grctx_func gk110_grctx;
 extern const struct gf100_grctx_func gk110b_grctx;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
index f527bca..a4fe36c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
@@ -841,6 +841,21 @@ gk104_grctx_pack_ppc[] = {
  ******************************************************************************/
 
 void
+gk104_grctx_generate_r418800(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	/*XXX: Not real sure where to apply these, there doesn't seem
+	 *     to be any pattern to which chipsets it's done on.
+	 *
+	 *     Perhaps a VBIOS tweak?
+	 */
+	if (0) {
+		nvkm_mask(device, 0x418800, 0x00200000, 0x00200000);
+		nvkm_mask(device, 0x41be10, 0x00800000, 0x00800000);
+	}
+}
+
+void
 gk104_grctx_generate_patch_ltc(struct gf100_grctx *info)
 {
 	struct nvkm_device *device = info->gr->base.engine.subdev.device;
@@ -935,8 +950,7 @@ gk104_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	gf100_gr_mthd(gr, grctx->mthd);
 	nvkm_mc_unk260(device, 1);
 
-	nvkm_mask(device, 0x418800, 0x00200000, 0x00200000);
-	nvkm_mask(device, 0x41be10, 0x00800000, 0x00800000);
+	grctx->r418800(gr);
 }
 
 void
@@ -1016,4 +1030,5 @@ gk104_grctx = {
 	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
 	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
 	.r419f78 = gk104_grctx_generate_r419f78,
+	.r418800 = gk104_grctx_generate_r418800,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
index 64e00d7..7102a24 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
@@ -837,4 +837,5 @@ gk110_grctx = {
 	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
 	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
 	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
+	.r418800 = gk104_grctx_generate_r418800,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
index f800ed5..049de07 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
@@ -98,4 +98,5 @@ gk110b_grctx = {
 	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
 	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
 	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
+	.r418800 = gk104_grctx_generate_r418800,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
index 494d9a9..c69494f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
@@ -559,4 +559,5 @@ gk208_grctx = {
 	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
 	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
 	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
+	.r418800 = gk104_grctx_generate_r418800,
 };
-- 
cgit v1.1


From ad45a92b9a104285dd3c95b6a8a4d5e0b50b6929 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/gr/gf100-: update 419eb0 where required

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c  | 2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h  | 3 +++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c  | 8 ++++++++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c | 1 +
 4 files changed, 14 insertions(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index 0dcb227..e8be2fb 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1428,6 +1428,8 @@ gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 		grctx->r419cb8(gr);
 	if (grctx->r418800)
 		grctx->r418800(gr);
+	if (grctx->r419eb0)
+		grctx->r419eb0(gr);
 }
 
 #define CB_RESERVED 0x80000
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
index 0d05664..4a6d470 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
@@ -66,6 +66,7 @@ struct gf100_grctx_func {
 	/* misc other things */
 	void (*r419cb8)(struct gf100_gr *);
 	void (*r418800)(struct gf100_gr *);
+	void (*r419eb0)(struct gf100_gr *);
 };
 
 extern const struct gf100_grctx_func gf100_grctx;
@@ -111,6 +112,8 @@ void gk104_grctx_generate_unkn(struct gf100_gr *);
 void gk104_grctx_generate_r418800(struct gf100_gr *);
 
 extern const struct gf100_grctx_func gk110_grctx;
+void gk110_grctx_generate_r419eb0(struct gf100_gr *);
+
 extern const struct gf100_grctx_func gk110b_grctx;
 extern const struct gf100_grctx_func gk208_grctx;
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
index 7102a24..f00e7af 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
@@ -808,6 +808,13 @@ gk110_grctx_pack_ppc[] = {
  * PGRAPH context implementation
  ******************************************************************************/
 
+void
+gk110_grctx_generate_r419eb0(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x419eb0, 0x00001000, 0x00001000);
+}
+
 const struct gf100_grctx_func
 gk110_grctx = {
 	.main  = gk104_grctx_generate_main,
@@ -838,4 +845,5 @@ gk110_grctx = {
 	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
 	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
 	.r418800 = gk104_grctx_generate_r418800,
+	.r419eb0 = gk110_grctx_generate_r419eb0,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
index 049de07..ed09cfd 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
@@ -99,4 +99,5 @@ gk110b_grctx = {
 	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
 	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
 	.r418800 = gk104_grctx_generate_r418800,
+	.r419eb0 = gk110_grctx_generate_r419eb0,
 };
-- 
cgit v1.1


From 18d17221dd58741a8590ba0a40a9ded82aa5d619 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/gr/gf100-: virtualise r419e00

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c | 16 ++++++++++++----
 3 files changed, 15 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index e8be2fb..d2b78ed 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1430,6 +1430,8 @@ gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 		grctx->r418800(gr);
 	if (grctx->r419eb0)
 		grctx->r419eb0(gr);
+	if (grctx->r419e00)
+		grctx->r419e00(gr);
 }
 
 #define CB_RESERVED 0x80000
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
index 4a6d470..474b5dd 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
@@ -67,6 +67,7 @@ struct gf100_grctx_func {
 	void (*r419cb8)(struct gf100_gr *);
 	void (*r418800)(struct gf100_gr *);
 	void (*r419eb0)(struct gf100_gr *);
+	void (*r419e00)(struct gf100_gr *);
 };
 
 extern const struct gf100_grctx_func gf100_grctx;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
index 6b279bb..12a9431 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
@@ -860,6 +860,16 @@ gm107_grctx_pack_ppc[] = {
  * PGRAPH context implementation
  ******************************************************************************/
 
+static void
+gm107_grctx_generate_r419e00(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x419e00, 0x00808080, 0x00808080);
+	nvkm_mask(device, 0x419ccc, 0x80000000, 0x80000000);
+	nvkm_mask(device, 0x419f80, 0x80000000, 0x80000000);
+	nvkm_mask(device, 0x419f88, 0x80000000, 0x80000000);
+}
+
 void
 gm107_grctx_generate_bundle(struct gf100_grctx *info)
 {
@@ -971,10 +981,7 @@ gm107_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	nvkm_wr32(device, 0x404154, idle_timeout);
 	gf100_gr_mthd(gr, grctx->mthd);
 
-	nvkm_mask(device, 0x419e00, 0x00808080, 0x00808080);
-	nvkm_mask(device, 0x419ccc, 0x80000000, 0x80000000);
-	nvkm_mask(device, 0x419f80, 0x80000000, 0x80000000);
-	nvkm_mask(device, 0x419f88, 0x80000000, 0x80000000);
+	grctx->r419e00(gr);
 }
 
 const struct gf100_grctx_func
@@ -1006,4 +1013,5 @@ gm107_grctx = {
 	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
 	.r406500 = gm107_grctx_generate_r406500,
 	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
+	.r419e00 = gm107_grctx_generate_r419e00,
 };
-- 
cgit v1.1


From c2592adea7a81857bf27f5b820640e67a0c6b664 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/gr/gf100-: virtualise r418e94

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c | 12 ++++++++++--
 3 files changed, 13 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index d2b78ed..24792be 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1432,6 +1432,8 @@ gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 		grctx->r419eb0(gr);
 	if (grctx->r419e00)
 		grctx->r419e00(gr);
+	if (grctx->r418e94)
+		grctx->r418e94(gr);
 }
 
 #define CB_RESERVED 0x80000
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
index 474b5dd..ac7f4f5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
@@ -68,6 +68,7 @@ struct gf100_grctx_func {
 	void (*r418800)(struct gf100_gr *);
 	void (*r419eb0)(struct gf100_gr *);
 	void (*r419e00)(struct gf100_gr *);
+	void (*r418e94)(struct gf100_gr *);
 };
 
 extern const struct gf100_grctx_func gf100_grctx;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
index f1e87b9..13951aa 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
@@ -27,6 +27,14 @@
  * PGRAPH context implementation
  ******************************************************************************/
 
+static void
+gm200_grctx_generate_r418e94(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x418e94, 0xffffffff, 0xc4230000);
+	nvkm_mask(device, 0x418e4c, 0xffffffff, 0x70000000);
+}
+
 void
 gm200_grctx_generate_smid_config(struct gf100_gr *gr)
 {
@@ -96,8 +104,7 @@ gm200_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 	nvkm_wr32(device, 0x404154, idle_timeout);
 	gf100_gr_mthd(gr, gr->fuc_method);
 
-	nvkm_mask(device, 0x418e94, 0xffffffff, 0xc4230000);
-	nvkm_mask(device, 0x418e4c, 0xffffffff, 0x70000000);
+	grctx->r418e94(gr);
 }
 
 void
@@ -144,4 +151,5 @@ gm200_grctx = {
 	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
 	.tpc_mask = gm200_grctx_generate_tpc_mask,
 	.smid_config = gm200_grctx_generate_smid_config,
+	.r418e94 = gm200_grctx_generate_r418e94,
 };
-- 
cgit v1.1


From 8d56fc48d3563cb3767c1d3dd243974168067f99 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/gr/gf100-: update 419a3c where required

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c | 2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h | 2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c | 8 ++++++++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c | 1 +
 6 files changed, 15 insertions(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index 24792be..1d1d0cc 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1434,6 +1434,8 @@ gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 		grctx->r419e00(gr);
 	if (grctx->r418e94)
 		grctx->r418e94(gr);
+	if (grctx->r419a3c)
+		grctx->r419a3c(gr);
 }
 
 #define CB_RESERVED 0x80000
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
index ac7f4f5..bff24e6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
@@ -69,6 +69,7 @@ struct gf100_grctx_func {
 	void (*r419eb0)(struct gf100_gr *);
 	void (*r419e00)(struct gf100_gr *);
 	void (*r418e94)(struct gf100_gr *);
+	void (*r419a3c)(struct gf100_gr *);
 };
 
 extern const struct gf100_grctx_func gf100_grctx;
@@ -130,6 +131,7 @@ void gm200_grctx_generate_dist_skip_table(struct gf100_gr *);
 void gm200_grctx_generate_r406500(struct gf100_gr *);
 void gm200_grctx_generate_tpc_mask(struct gf100_gr *);
 void gm200_grctx_generate_smid_config(struct gf100_gr *);
+void gm200_grctx_generate_r419a3c(struct gf100_gr *);
 
 extern const struct gf100_grctx_func gm20b_grctx;
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
index 13951aa..7a9be04 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
@@ -27,6 +27,13 @@
  * PGRAPH context implementation
  ******************************************************************************/
 
+void
+gm200_grctx_generate_r419a3c(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x419a3c, 0x00000014, 0x00000000);
+}
+
 static void
 gm200_grctx_generate_r418e94(struct gf100_gr *gr)
 {
@@ -152,4 +159,5 @@ gm200_grctx = {
 	.tpc_mask = gm200_grctx_generate_tpc_mask,
 	.smid_config = gm200_grctx_generate_smid_config,
 	.r418e94 = gm200_grctx_generate_r418e94,
+	.r419a3c = gm200_grctx_generate_r419a3c,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
index 821219a..c60f924 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
@@ -165,4 +165,5 @@ gp100_grctx = {
 	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
 	.tpc_mask = gm200_grctx_generate_tpc_mask,
 	.smid_config = gp100_grctx_generate_smid_config,
+	.r419a3c = gm200_grctx_generate_r419a3c,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
index 611819f..3af42c1 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
@@ -101,4 +101,5 @@ gp102_grctx = {
 	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
 	.tpc_mask = gm200_grctx_generate_tpc_mask,
 	.smid_config = gp100_grctx_generate_smid_config,
+	.r419a3c = gm200_grctx_generate_r419a3c,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
index d908317..1864674 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
@@ -51,4 +51,5 @@ gp107_grctx = {
 	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
 	.tpc_mask = gm200_grctx_generate_tpc_mask,
 	.smid_config = gp100_grctx_generate_smid_config,
+	.r419a3c = gm200_grctx_generate_r419a3c,
 };
-- 
cgit v1.1


From a5537f980e4aba64ce1a0b14ee8fb27d0fd10362 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/gr/gf100-: update r408840 where required

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/include/nvkm/engine/gr.h  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c |  4 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/Kbuild     |  2 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c |  2 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h |  3 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c |  8 +++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp104.c | 47 +++++++++++++++++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c    | 62 +++++++++++++++++++++++
 8 files changed, 127 insertions(+), 2 deletions(-)
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp104.c
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c

diff --git a/drivers/gpu/drm/nouveau/include/nvkm/engine/gr.h b/drivers/gpu/drm/nouveau/include/nvkm/engine/gr.h
index fb18f10..33b2f2e 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/engine/gr.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/engine/gr.h
@@ -45,6 +45,7 @@ int gm200_gr_new(struct nvkm_device *, int, struct nvkm_gr **);
 int gm20b_gr_new(struct nvkm_device *, int, struct nvkm_gr **);
 int gp100_gr_new(struct nvkm_device *, int, struct nvkm_gr **);
 int gp102_gr_new(struct nvkm_device *, int, struct nvkm_gr **);
+int gp104_gr_new(struct nvkm_device *, int, struct nvkm_gr **);
 int gp107_gr_new(struct nvkm_device *, int, struct nvkm_gr **);
 int gp10b_gr_new(struct nvkm_device *, int, struct nvkm_gr **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 379e701..16e8090 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2256,7 +2256,7 @@ nv134_chipset = {
 	.disp = gp102_disp_new,
 	.dma = gf119_dma_new,
 	.fifo = gp100_fifo_new,
-	.gr = gp102_gr_new,
+	.gr = gp104_gr_new,
 	.nvdec = gp102_nvdec_new,
 	.sec2 = gp102_sec2_new,
 	.sw = gf100_sw_new,
@@ -2292,7 +2292,7 @@ nv136_chipset = {
 	.disp = gp102_disp_new,
 	.dma = gf119_dma_new,
 	.fifo = gp100_fifo_new,
-	.gr = gp102_gr_new,
+	.gr = gp104_gr_new,
 	.nvdec = gp102_nvdec_new,
 	.sec2 = gp102_sec2_new,
 	.sw = gf100_sw_new,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/Kbuild b/drivers/gpu/drm/nouveau/nvkm/engine/gr/Kbuild
index 8a22558..42342b4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/Kbuild
@@ -33,6 +33,7 @@ nvkm-y += nvkm/engine/gr/gm200.o
 nvkm-y += nvkm/engine/gr/gm20b.o
 nvkm-y += nvkm/engine/gr/gp100.o
 nvkm-y += nvkm/engine/gr/gp102.o
+nvkm-y += nvkm/engine/gr/gp104.o
 nvkm-y += nvkm/engine/gr/gp107.o
 nvkm-y += nvkm/engine/gr/gp10b.o
 
@@ -54,4 +55,5 @@ nvkm-y += nvkm/engine/gr/ctxgm200.o
 nvkm-y += nvkm/engine/gr/ctxgm20b.o
 nvkm-y += nvkm/engine/gr/ctxgp100.o
 nvkm-y += nvkm/engine/gr/ctxgp102.o
+nvkm-y += nvkm/engine/gr/ctxgp104.o
 nvkm-y += nvkm/engine/gr/ctxgp107.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index 1d1d0cc..5560376 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1436,6 +1436,8 @@ gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 		grctx->r418e94(gr);
 	if (grctx->r419a3c)
 		grctx->r419a3c(gr);
+	if (grctx->r408840)
+		grctx->r408840(gr);
 }
 
 #define CB_RESERVED 0x80000
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
index bff24e6..668b2c7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
@@ -70,6 +70,7 @@ struct gf100_grctx_func {
 	void (*r419e00)(struct gf100_gr *);
 	void (*r418e94)(struct gf100_gr *);
 	void (*r419a3c)(struct gf100_gr *);
+	void (*r408840)(struct gf100_gr *);
 };
 
 extern const struct gf100_grctx_func gf100_grctx;
@@ -143,6 +144,8 @@ void gp100_grctx_generate_smid_config(struct gf100_gr *);
 extern const struct gf100_grctx_func gp102_grctx;
 void gp102_grctx_generate_attrib(struct gf100_grctx *);
 
+extern const struct gf100_grctx_func gp104_grctx;
+
 extern const struct gf100_grctx_func gp107_grctx;
 
 /* context init value lists */
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
index 3af42c1..3c78a6d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
@@ -29,6 +29,13 @@
  * PGRAPH context implementation
  ******************************************************************************/
 
+static void
+gp102_grctx_generate_r408840(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x408840, 0x00000003, 0x00000000);
+}
+
 void
 gp102_grctx_generate_attrib(struct gf100_grctx *info)
 {
@@ -102,4 +109,5 @@ gp102_grctx = {
 	.tpc_mask = gm200_grctx_generate_tpc_mask,
 	.smid_config = gp100_grctx_generate_smid_config,
 	.r419a3c = gm200_grctx_generate_r419a3c,
+	.r408840 = gp102_grctx_generate_r408840,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp104.c
new file mode 100644
index 0000000..020cb04
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp104.c
@@ -0,0 +1,47 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "ctxgf100.h"
+
+const struct gf100_grctx_func
+gp104_grctx = {
+	.main = gp100_grctx_generate_main,
+	.unkn = gk104_grctx_generate_unkn,
+	.bundle = gm107_grctx_generate_bundle,
+	.bundle_size = 0x3000,
+	.bundle_min_gpm_fifo_depth = 0x180,
+	.bundle_token_limit = 0x900,
+	.pagepool = gp100_grctx_generate_pagepool,
+	.pagepool_size = 0x20000,
+	.attrib = gp102_grctx_generate_attrib,
+	.attrib_nr_max = 0x5d4,
+	.attrib_nr = 0x320,
+	.alpha_nr_max = 0xc00,
+	.alpha_nr = 0x800,
+	.sm_id = gm107_grctx_generate_sm_id,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
+	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
+	.r406500 = gm200_grctx_generate_r406500,
+	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
+	.tpc_mask = gm200_grctx_generate_tpc_mask,
+	.smid_config = gp100_grctx_generate_smid_config,
+	.r419a3c = gm200_grctx_generate_r419a3c,
+};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c
new file mode 100644
index 0000000..289d8b2
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c
@@ -0,0 +1,62 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "gf100.h"
+#include "ctxgf100.h"
+
+#include <nvif/class.h>
+
+static const struct gf100_gr_func
+gp104_gr = {
+	.init = gf100_gr_init,
+	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gm200_gr_init_num_active_ltcs,
+	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
+	.init_swdx_pes_mask = gp102_gr_init_swdx_pes_mask,
+	.init_fecs_exceptions = gp100_gr_init_fecs_exceptions,
+	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
+	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
+	.init_419cc0 = gf100_gr_init_419cc0,
+	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_504430 = gm107_gr_init_504430,
+	.init_shader_exceptions = gp100_gr_init_shader_exceptions,
+	.rops = gm200_gr_rops,
+	.gpc_nr = 6,
+	.tpc_nr = 5,
+	.ppc_nr = 3,
+	.grctx = &gp104_grctx,
+	.sclass = {
+		{ -1, -1, FERMI_TWOD_A },
+		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
+		{ -1, -1, PASCAL_B, &gf100_fermi },
+		{ -1, -1, PASCAL_COMPUTE_B },
+		{}
+	}
+};
+
+int
+gp104_gr_new(struct nvkm_device *device, int index, struct nvkm_gr **pgr)
+{
+	return gm200_gr_new_(&gp104_gr, device, index, pgr);
+}
-- 
cgit v1.1


From 201ed6f651e72d886d9066237a830aa49d2ffcd1 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/gr/gf100-: delete duplicated grctx init code

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h  |  2 --
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c  | 32 +-------------------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c  | 34 +---------------------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c  |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c  |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c  | 31 +-------------------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c  | 27 +----------------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c  | 25 +---------------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c  |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp104.c  |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c  |  2 +-
 12 files changed, 11 insertions(+), 152 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
index 668b2c7..be57ff0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
@@ -108,7 +108,6 @@ void gk104_grctx_generate_alpha_beta_tables(struct gf100_gr *);
 void gk104_grctx_generate_gpc_tpc_nr(struct gf100_gr *);
 
 extern const struct gf100_grctx_func gk20a_grctx;
-void gk104_grctx_generate_main(struct gf100_gr *, struct gf100_grctx *);
 void gk104_grctx_generate_bundle(struct gf100_grctx *);
 void gk104_grctx_generate_pagepool(struct gf100_grctx *);
 void gk104_grctx_generate_patch_ltc(struct gf100_grctx *);
@@ -137,7 +136,6 @@ void gm200_grctx_generate_r419a3c(struct gf100_gr *);
 extern const struct gf100_grctx_func gm20b_grctx;
 
 extern const struct gf100_grctx_func gp100_grctx;
-void gp100_grctx_generate_main(struct gf100_gr *, struct gf100_grctx *);
 void gp100_grctx_generate_pagepool(struct gf100_grctx *);
 void gp100_grctx_generate_smid_config(struct gf100_gr *);
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
index e922e3e..4b7b4f8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
@@ -286,39 +286,9 @@ gf117_grctx_generate_attrib(struct gf100_grctx *info)
 	}
 }
 
-static void
-gf117_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
-{
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-	const struct gf100_grctx_func *grctx = gr->func->grctx;
-	u32 idle_timeout;
-
-	nvkm_mc_unk260(device, 0);
-
-	gf100_gr_mmio(gr, grctx->hub);
-	gf100_gr_mmio(gr, grctx->gpc);
-	gf100_gr_mmio(gr, grctx->zcull);
-	gf100_gr_mmio(gr, grctx->tpc);
-	gf100_gr_mmio(gr, grctx->ppc);
-
-	idle_timeout = nvkm_mask(device, 0x404154, 0xffffffff, 0x00000000);
-
-	grctx->bundle(info);
-	grctx->pagepool(info);
-	grctx->attrib(info);
-	grctx->unkn(gr);
-
-	gf100_grctx_generate_floorsweep(gr);
-
-	gf100_gr_icmd(gr, grctx->icmd);
-	nvkm_wr32(device, 0x404154, idle_timeout);
-	gf100_gr_mthd(gr, grctx->mthd);
-	nvkm_mc_unk260(device, 1);
-}
-
 const struct gf100_grctx_func
 gf117_grctx = {
-	.main  = gf117_grctx_generate_main,
+	.main  = gf100_grctx_generate_main,
 	.unkn  = gk104_grctx_generate_unkn,
 	.hub   = gf117_grctx_pack_hub,
 	.gpc   = gf117_grctx_pack_gpc,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
index a4fe36c..bdf2a1e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
@@ -922,38 +922,6 @@ gk104_grctx_generate_gpc_tpc_nr(struct gf100_gr *gr)
 }
 
 void
-gk104_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
-{
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-	const struct gf100_grctx_func *grctx = gr->func->grctx;
-	u32 idle_timeout;
-
-	nvkm_mc_unk260(device, 0);
-
-	gf100_gr_mmio(gr, grctx->hub);
-	gf100_gr_mmio(gr, grctx->gpc);
-	gf100_gr_mmio(gr, grctx->zcull);
-	gf100_gr_mmio(gr, grctx->tpc);
-	gf100_gr_mmio(gr, grctx->ppc);
-
-	idle_timeout = nvkm_mask(device, 0x404154, 0xffffffff, 0x00000000);
-
-	grctx->bundle(info);
-	grctx->pagepool(info);
-	grctx->attrib(info);
-	grctx->unkn(gr);
-
-	gf100_grctx_generate_floorsweep(gr);
-
-	gf100_gr_icmd(gr, grctx->icmd);
-	nvkm_wr32(device, 0x404154, idle_timeout);
-	gf100_gr_mthd(gr, grctx->mthd);
-	nvkm_mc_unk260(device, 1);
-
-	grctx->r418800(gr);
-}
-
-void
 gk104_grctx_generate_alpha_beta_tables(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -1002,7 +970,7 @@ gk104_grctx_generate_alpha_beta_tables(struct gf100_gr *gr)
 
 const struct gf100_grctx_func
 gk104_grctx = {
-	.main  = gk104_grctx_generate_main,
+	.main  = gf100_grctx_generate_main,
 	.unkn  = gk104_grctx_generate_unkn,
 	.hub   = gk104_grctx_pack_hub,
 	.gpc   = gk104_grctx_pack_gpc,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
index f00e7af..2ba35d7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
@@ -817,7 +817,7 @@ gk110_grctx_generate_r419eb0(struct gf100_gr *gr)
 
 const struct gf100_grctx_func
 gk110_grctx = {
-	.main  = gk104_grctx_generate_main,
+	.main  = gf100_grctx_generate_main,
 	.unkn  = gk104_grctx_generate_unkn,
 	.hub   = gk110_grctx_pack_hub,
 	.gpc   = gk110_grctx_pack_gpc,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
index ed09cfd..1112f8d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
@@ -71,7 +71,7 @@ gk110b_grctx_pack_tpc[] = {
 
 const struct gf100_grctx_func
 gk110b_grctx = {
-	.main  = gk104_grctx_generate_main,
+	.main  = gf100_grctx_generate_main,
 	.unkn  = gk104_grctx_generate_unkn,
 	.hub   = gk110_grctx_pack_hub,
 	.gpc   = gk110_grctx_pack_gpc,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
index c69494f..613c5cf8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
@@ -532,7 +532,7 @@ gk208_grctx_pack_ppc[] = {
 
 const struct gf100_grctx_func
 gk208_grctx = {
-	.main  = gk104_grctx_generate_main,
+	.main  = gf100_grctx_generate_main,
 	.unkn  = gk104_grctx_generate_unkn,
 	.hub   = gk208_grctx_pack_hub,
 	.gpc   = gk208_grctx_pack_gpc,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
index 12a9431..7816dcb 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
@@ -955,38 +955,9 @@ gm107_grctx_generate_sm_id(struct gf100_gr *gr, int gpc, int tpc, int sm)
 	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x088), sm);
 }
 
-static void
-gm107_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
-{
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-	const struct gf100_grctx_func *grctx = gr->func->grctx;
-	u32 idle_timeout;
-
-	gf100_gr_mmio(gr, grctx->hub);
-	gf100_gr_mmio(gr, grctx->gpc);
-	gf100_gr_mmio(gr, grctx->zcull);
-	gf100_gr_mmio(gr, grctx->tpc);
-	gf100_gr_mmio(gr, grctx->ppc);
-
-	idle_timeout = nvkm_mask(device, 0x404154, 0xffffffff, 0x00000000);
-
-	grctx->bundle(info);
-	grctx->pagepool(info);
-	grctx->attrib(info);
-	grctx->unkn(gr);
-
-	gf100_grctx_generate_floorsweep(gr);
-
-	gf100_gr_icmd(gr, grctx->icmd);
-	nvkm_wr32(device, 0x404154, idle_timeout);
-	gf100_gr_mthd(gr, grctx->mthd);
-
-	grctx->r419e00(gr);
-}
-
 const struct gf100_grctx_func
 gm107_grctx = {
-	.main  = gm107_grctx_generate_main,
+	.main  = gf100_grctx_generate_main,
 	.unkn  = gk104_grctx_generate_unkn,
 	.hub   = gm107_grctx_pack_hub,
 	.gpc   = gm107_grctx_pack_gpc,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
index 7a9be04..7107ec4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
@@ -89,31 +89,6 @@ gm200_grctx_generate_r406500(struct gf100_gr *gr)
 	nvkm_wr32(gr->base.engine.subdev.device, 0x406500, 0x00000000);
 }
 
-static void
-gm200_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
-{
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-	const struct gf100_grctx_func *grctx = gr->func->grctx;
-	u32 idle_timeout;
-
-	gf100_gr_mmio(gr, gr->fuc_sw_ctx);
-
-	idle_timeout = nvkm_mask(device, 0x404154, 0xffffffff, 0x00000000);
-
-	grctx->bundle(info);
-	grctx->pagepool(info);
-	grctx->attrib(info);
-	grctx->unkn(gr);
-
-	gf100_grctx_generate_floorsweep(gr);
-
-	gf100_gr_icmd(gr, gr->fuc_bundle);
-	nvkm_wr32(device, 0x404154, idle_timeout);
-	gf100_gr_mthd(gr, gr->fuc_method);
-
-	grctx->r418e94(gr);
-}
-
 void
 gm200_grctx_generate_dist_skip_table(struct gf100_gr *gr)
 {
@@ -138,7 +113,7 @@ gm200_grctx_generate_dist_skip_table(struct gf100_gr *gr)
 
 const struct gf100_grctx_func
 gm200_grctx = {
-	.main  = gm200_grctx_generate_main,
+	.main  = gf100_grctx_generate_main,
 	.unkn  = gk104_grctx_generate_unkn,
 	.bundle = gm107_grctx_generate_bundle,
 	.bundle_size = 0x3000,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
index c60f924..af6330c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
@@ -120,32 +120,9 @@ gp100_grctx_generate_smid_config(struct gf100_gr *gr)
 		nvkm_wr32(device, 0x405ba0 + (i * 4), gpcs[i]);
 }
 
-void
-gp100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
-{
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-	const struct gf100_grctx_func *grctx = gr->func->grctx;
-	u32 idle_timeout;
-
-	gf100_gr_mmio(gr, gr->fuc_sw_ctx);
-
-	idle_timeout = nvkm_mask(device, 0x404154, 0xffffffff, 0x00000000);
-
-	grctx->pagepool(info);
-	grctx->bundle(info);
-	grctx->attrib(info);
-	grctx->unkn(gr);
-
-	gf100_grctx_generate_floorsweep(gr);
-
-	gf100_gr_icmd(gr, gr->fuc_bundle);
-	nvkm_wr32(device, 0x404154, idle_timeout);
-	gf100_gr_mthd(gr, gr->fuc_method);
-}
-
 const struct gf100_grctx_func
 gp100_grctx = {
-	.main  = gp100_grctx_generate_main,
+	.main  = gf100_grctx_generate_main,
 	.unkn  = gk104_grctx_generate_unkn,
 	.bundle = gm107_grctx_generate_bundle,
 	.bundle_size = 0x3000,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
index 3c78a6d..8a438c2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
@@ -88,7 +88,7 @@ gp102_grctx_generate_attrib(struct gf100_grctx *info)
 
 const struct gf100_grctx_func
 gp102_grctx = {
-	.main = gp100_grctx_generate_main,
+	.main = gf100_grctx_generate_main,
 	.unkn = gk104_grctx_generate_unkn,
 	.bundle = gm107_grctx_generate_bundle,
 	.bundle_size = 0x3000,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp104.c
index 020cb04..5f799c7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp104.c
@@ -23,7 +23,7 @@
 
 const struct gf100_grctx_func
 gp104_grctx = {
-	.main = gp100_grctx_generate_main,
+	.main = gf100_grctx_generate_main,
 	.unkn = gk104_grctx_generate_unkn,
 	.bundle = gm107_grctx_generate_bundle,
 	.bundle_size = 0x3000,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
index 1864674..a69e824 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
@@ -31,7 +31,7 @@
 
 const struct gf100_grctx_func
 gp107_grctx = {
-	.main = gp100_grctx_generate_main,
+	.main = gf100_grctx_generate_main,
 	.unkn = gk104_grctx_generate_unkn,
 	.bundle = gm107_grctx_generate_bundle,
 	.bundle_size = 0x3000,
-- 
cgit v1.1


From 74b6068bd660a806e801ae039dbab58dc284301e Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/gr/gf100-: add missing reset sequence before golden
 context init

RM and NVGPU both have a variant of this, we probably should too.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c | 20 ++++++++++++++++++++
 1 file changed, 20 insertions(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index 5560376..4c25389 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1456,6 +1456,26 @@ gf100_grctx_generate(struct gf100_gr *gr)
 	int ret, i;
 	u64 addr;
 
+	/* NV_PGRAPH_FE_PWR_MODE_FORCE_ON. */
+	nvkm_wr32(device, 0x404170, 0x00000012);
+	nvkm_msec(device, 2000,
+		if (!(nvkm_rd32(device, 0x404170) & 0x00000010))
+			break;
+	);
+
+	/* Reset FECS. */
+	nvkm_wr32(device, 0x409614, 0x00000070);
+	nvkm_usec(device, 10, NVKM_DELAY);
+	nvkm_mask(device, 0x409614, 0x00000700, 0x00000700);
+	nvkm_usec(device, 10, NVKM_DELAY);
+	nvkm_rd32(device, 0x409614);
+
+	/* NV_PGRAPH_FE_PWR_MODE_AUTO. */
+	nvkm_wr32(device, 0x404170, 0x00000010);
+
+	/* Init SCC RAM. */
+	nvkm_wr32(device, 0x40802c, 0x00000001);
+
 	/* Allocate memory to for a "channel", which we'll use to generate
 	 * the default context values.
 	 */
-- 
cgit v1.1


From 5c05a589856ad5f79c22b0500340291c591c3050 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/gr/gf100-: virtualise trap_mp

Required to support Volta.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c  | 5 +++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h  | 2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c  | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c  | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c  | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c  | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c  | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c  | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c  | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c  | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c  | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c  | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c  | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c  | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c  | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c  | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c  | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c  | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c  | 1 +
 20 files changed, 23 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index dd4a410..f05d9d4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -1010,7 +1010,7 @@ static const struct nvkm_bitfield gf100_mp_global_error[] = {
 	{}
 };
 
-static void
+void
 gf100_gr_trap_mp(struct gf100_gr *gr, int gpc, int tpc)
 {
 	struct nvkm_subdev *subdev = &gr->base.engine.subdev;
@@ -1046,7 +1046,7 @@ gf100_gr_trap_tpc(struct gf100_gr *gr, int gpc, int tpc)
 	}
 
 	if (stat & 0x00000002) {
-		gf100_gr_trap_mp(gr, gpc, tpc);
+		gr->func->trap_mp(gr, gpc, tpc);
 		stat &= ~0x00000002;
 	}
 
@@ -2176,6 +2176,7 @@ gf100_gr = {
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
 	.init_400054 = gf100_gr_init_400054,
+	.trap_mp = gf100_gr_trap_mp,
 	.mmio = gf100_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index 31109ce..c25b93a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -145,6 +145,7 @@ struct gf100_gr_func {
 	void (*init_504430)(struct gf100_gr *, int gpc, int tpc);
 	void (*init_shader_exceptions)(struct gf100_gr *, int gpc, int tpc);
 	void (*init_400054)(struct gf100_gr *);
+	void (*trap_mp)(struct gf100_gr *, int gpc, int tpc);
 	void (*set_hww_esr_report_mask)(struct gf100_gr *);
 	const struct gf100_gr_pack *mmio;
 	struct {
@@ -301,6 +302,7 @@ extern const struct gf100_gr_init gf100_gr_init_be_0[];
 extern const struct gf100_gr_init gf100_gr_init_fe_1[];
 extern const struct gf100_gr_init gf100_gr_init_pe_1[];
 void gf100_gr_init_gpc_mmu(struct gf100_gr *);
+void gf100_gr_trap_mp(struct gf100_gr *, int, int);
 
 extern const struct gf100_gr_init gf104_gr_init_ds_0[];
 extern const struct gf100_gr_init gf104_gr_init_tex_0[];
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
index d5276ca..df9cbed 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
@@ -126,6 +126,7 @@ gf104_gr = {
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
 	.init_400054 = gf100_gr_init_400054,
+	.trap_mp = gf100_gr_trap_mp,
 	.mmio = gf104_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
index 8f22a31..8ffa0fd 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
@@ -124,6 +124,7 @@ gf108_gr = {
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
 	.init_400054 = gf100_gr_init_400054,
+	.trap_mp = gf100_gr_trap_mp,
 	.mmio = gf108_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
index fcbfdc7..0d4293e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
@@ -98,6 +98,7 @@ gf110_gr = {
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
 	.init_400054 = gf100_gr_init_400054,
+	.trap_mp = gf100_gr_trap_mp,
 	.mmio = gf110_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
index f526ccd..e3c1dbb 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
@@ -162,6 +162,7 @@ gf117_gr = {
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
 	.init_400054 = gf100_gr_init_400054,
+	.trap_mp = gf100_gr_trap_mp,
 	.mmio = gf117_gr_pack_mmio,
 	.fecs.ucode = &gf117_gr_fecs_ucode,
 	.gpccs.ucode = &gf117_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
index 5d3d22f..1ed70b9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
@@ -189,6 +189,7 @@ gf119_gr = {
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
 	.init_400054 = gf100_gr_init_400054,
+	.trap_mp = gf100_gr_trap_mp,
 	.mmio = gf119_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
index 2e5c48b..86819ab 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
@@ -462,6 +462,7 @@ gk104_gr = {
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
 	.init_400054 = gf100_gr_init_400054,
+	.trap_mp = gf100_gr_trap_mp,
 	.mmio = gk104_gr_pack_mmio,
 	.fecs.ucode = &gk104_gr_fecs_ucode,
 	.gpccs.ucode = &gk104_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
index 9adb55f..e30d94f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
@@ -364,6 +364,7 @@ gk110_gr = {
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
 	.init_400054 = gf100_gr_init_400054,
+	.trap_mp = gf100_gr_trap_mp,
 	.mmio = gk110_gr_pack_mmio,
 	.fecs.ucode = &gk110_gr_fecs_ucode,
 	.gpccs.ucode = &gk110_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
index f848f15..253b981 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
@@ -116,6 +116,7 @@ gk110b_gr = {
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
 	.init_400054 = gf100_gr_init_400054,
+	.trap_mp = gf100_gr_trap_mp,
 	.mmio = gk110b_gr_pack_mmio,
 	.fecs.ucode = &gk110_gr_fecs_ucode,
 	.gpccs.ucode = &gk110_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
index 9c678f1..702e909 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
@@ -174,6 +174,7 @@ gk208_gr = {
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
 	.init_400054 = gf100_gr_init_400054,
+	.trap_mp = gf100_gr_trap_mp,
 	.mmio = gk208_gr_pack_mmio,
 	.fecs.ucode = &gk208_gr_fecs_ucode,
 	.gpccs.ucode = &gk208_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c
index a806643..95f7d85 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c
@@ -285,6 +285,7 @@ gk20a_gr = {
 	.init = gk20a_gr_init,
 	.init_zcull = gf117_gr_init_zcull,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
+	.trap_mp = gf100_gr_trap_mp,
 	.set_hww_esr_report_mask = gk20a_gr_set_hww_esr_report_mask,
 	.rops = gf100_gr_rops,
 	.ppc_nr = 1,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
index c598fa5..d67bf94 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
@@ -407,6 +407,7 @@ gm107_gr = {
 	.init_504430 = gm107_gr_init_504430,
 	.init_shader_exceptions = gm107_gr_init_shader_exceptions,
 	.init_400054 = gm107_gr_init_400054,
+	.trap_mp = gf100_gr_trap_mp,
 	.mmio = gm107_gr_pack_mmio,
 	.fecs.ucode = &gm107_gr_fecs_ucode,
 	.gpccs.ucode = &gm107_gr_gpccs_ucode,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
index ae0eaf8..03b255e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
@@ -133,6 +133,7 @@ gm200_gr = {
 	.init_504430 = gm107_gr_init_504430,
 	.init_shader_exceptions = gm107_gr_init_shader_exceptions,
 	.init_400054 = gm107_gr_init_400054,
+	.trap_mp = gf100_gr_trap_mp,
 	.rops = gm200_gr_rops,
 	.tpc_nr = 4,
 	.ppc_nr = 2,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c
index fcf86d5..d2f9c7b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c
@@ -68,6 +68,7 @@ gm20b_gr = {
 	.init_zcull = gf117_gr_init_zcull,
 	.init_gpc_mmu = gm20b_gr_init_gpc_mmu,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
+	.trap_mp = gf100_gr_trap_mp,
 	.set_hww_esr_report_mask = gm20b_gr_set_hww_esr_report_mask,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 1,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
index 3addbc1..e5f941f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
@@ -79,6 +79,7 @@ gp100_gr = {
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_504430 = gm107_gr_init_504430,
 	.init_shader_exceptions = gp100_gr_init_shader_exceptions,
+	.trap_mp = gf100_gr_trap_mp,
 	.rops = gm200_gr_rops,
 	.gpc_nr = 6,
 	.tpc_nr = 5,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
index ea99c15..09e2665 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
@@ -57,6 +57,7 @@ gp102_gr = {
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_504430 = gm107_gr_init_504430,
 	.init_shader_exceptions = gp100_gr_init_shader_exceptions,
+	.trap_mp = gf100_gr_trap_mp,
 	.rops = gm200_gr_rops,
 	.gpc_nr = 6,
 	.tpc_nr = 5,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c
index 289d8b2..844fc9d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c
@@ -41,6 +41,7 @@ gp104_gr = {
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_504430 = gm107_gr_init_504430,
 	.init_shader_exceptions = gp100_gr_init_shader_exceptions,
+	.trap_mp = gf100_gr_trap_mp,
 	.rops = gm200_gr_rops,
 	.gpc_nr = 6,
 	.tpc_nr = 5,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
index 09cba53..674385d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
@@ -43,6 +43,7 @@ gp107_gr = {
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_504430 = gm107_gr_init_504430,
 	.init_shader_exceptions = gp100_gr_init_shader_exceptions,
+	.trap_mp = gf100_gr_trap_mp,
 	.rops = gm200_gr_rops,
 	.gpc_nr = 2,
 	.tpc_nr = 3,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
index 4972bf8..6103186 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
@@ -41,6 +41,7 @@ gp10b_gr = {
 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
 	.init_504430 = gm107_gr_init_504430,
 	.init_shader_exceptions = gp100_gr_init_shader_exceptions,
+	.trap_mp = gf100_gr_trap_mp,
 	.rops = gm200_gr_rops,
 	.gpc_nr = 1,
 	.tpc_nr = 2,
-- 
cgit v1.1


From 5f6474a4e6ce3291abb1843b279a23a0bb050d37 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/gr/gf100-: port tile mapping calculations from NVGPU

There's also a couple of hardcoded tables for a couple of very specific
configurations that NVGPU's algorithm didn't work for.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c |  23 +----
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c |  23 +----
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c    | 118 +++++++++++++++-------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h    |   8 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c    |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c    |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c    |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c    |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c    |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c    |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c    |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c   |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c    |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c    |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c    |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c    |  41 ++++++++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c    |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c    |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c    |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c    |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c    |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c    |   1 +
 22 files changed, 153 insertions(+), 77 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index 4c25389..949e121 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1116,27 +1116,14 @@ gf100_grctx_generate_rop_mapping(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
 	u32 data[6] = {}, data2[2] = {};
-	u8  tpcnr[GPC_MAX];
 	u8  shift, ntpcv;
-	int gpc, tpc, i;
-
-	/* calculate first set of magics */
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-
-	gpc = -1;
-	for (tpc = 0; tpc < gr->tpc_total; tpc++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while (!tpcnr[gpc]);
-		tpcnr[gpc]--;
-
-		data[tpc / 6] |= gpc << ((tpc % 6) * 5);
-	}
+	int i;
 
-	for (; tpc < 32; tpc++)
-		data[tpc / 6] |= 7 << ((tpc % 6) * 5);
+	/* Pack tile map into register format. */
+	for (i = 0; i < 32; i++)
+		data[i / 6] |= (gr->tile[i] & 0x07) << ((i % 6) * 5);
 
-	/* and the second... */
+	/* Magic. */
 	shift = 0;
 	ntpcv = gr->tpc_total;
 	while (!(ntpcv & (1 << 4))) {
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
index 4b7b4f8..bc4e86b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
@@ -194,27 +194,14 @@ gf117_grctx_generate_rop_mapping(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
 	u32 data[6] = {}, data2[2] = {};
-	u8  tpcnr[GPC_MAX];
 	u8  shift, ntpcv;
-	int gpc, tpc, i;
-
-	/* calculate first set of magics */
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-
-	gpc = -1;
-	for (tpc = 0; tpc < gr->tpc_total; tpc++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while (!tpcnr[gpc]);
-		tpcnr[gpc]--;
-
-		data[tpc / 6] |= gpc << ((tpc % 6) * 5);
-	}
+	int i;
 
-	for (; tpc < 32; tpc++)
-		data[tpc / 6] |= 7 << ((tpc % 6) * 5);
+	/* Pack tile map into register format. */
+	for (i = 0; i < 32; i++)
+		data[i / 6] |= (gr->tile[i] & 0x07) << ((i % 6) * 5);
 
-	/* and the second... */
+	/* Magic. */
 	shift = 0;
 	ntpcv = gr->tpc_total;
 	while (!(ntpcv & (1 << 4))) {
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index f05d9d4..519b109 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -1652,6 +1652,82 @@ gf100_gr_init_ctxctl(struct gf100_gr *gr)
 	return ret;
 }
 
+void
+gf100_gr_oneinit_tiles(struct gf100_gr *gr)
+{
+	static const u8 primes[] = {
+		3, 5, 7, 11, 13, 17, 19, 23, 29, 31, 37, 41, 43, 47, 53, 59, 61
+	};
+	int init_frac[GPC_MAX], init_err[GPC_MAX], run_err[GPC_MAX], i, j;
+	u32 mul_factor, comm_denom;
+	u8  gpc_map[GPC_MAX];
+	bool sorted;
+
+	switch (gr->tpc_total) {
+	case 15: gr->screen_tile_row_offset = 0x06; break;
+	case 14: gr->screen_tile_row_offset = 0x05; break;
+	case 13: gr->screen_tile_row_offset = 0x02; break;
+	case 11: gr->screen_tile_row_offset = 0x07; break;
+	case 10: gr->screen_tile_row_offset = 0x06; break;
+	case  7:
+	case  5: gr->screen_tile_row_offset = 0x01; break;
+	case  3: gr->screen_tile_row_offset = 0x02; break;
+	case  2:
+	case  1: gr->screen_tile_row_offset = 0x01; break;
+	default: gr->screen_tile_row_offset = 0x03;
+		for (i = 0; i < ARRAY_SIZE(primes); i++) {
+			if (gr->tpc_total % primes[i]) {
+				gr->screen_tile_row_offset = primes[i];
+				break;
+			}
+		}
+		break;
+	}
+
+	/* Sort GPCs by TPC count, highest-to-lowest. */
+	for (i = 0; i < gr->gpc_nr; i++)
+		gpc_map[i] = i;
+	sorted = false;
+
+	while (!sorted) {
+		for (sorted = true, i = 0; i < gr->gpc_nr - 1; i++) {
+			if (gr->tpc_nr[gpc_map[i + 1]] >
+			    gr->tpc_nr[gpc_map[i + 0]]) {
+				u8 swap = gpc_map[i];
+				gpc_map[i + 0] = gpc_map[i + 1];
+				gpc_map[i + 1] = swap;
+				sorted = false;
+			}
+		}
+	}
+
+	/* Determine tile->GPC mapping */
+	mul_factor = gr->gpc_nr * gr->tpc_max;
+	if (mul_factor & 1)
+		mul_factor = 2;
+	else
+		mul_factor = 1;
+
+	comm_denom = gr->gpc_nr * gr->tpc_max * mul_factor;
+
+	for (i = 0; i < gr->gpc_nr; i++) {
+		init_frac[i] = gr->tpc_nr[gpc_map[i]] * gr->gpc_nr * mul_factor;
+		 init_err[i] = i * gr->tpc_max * mul_factor - comm_denom/2;
+		  run_err[i] = init_frac[i] + init_err[i];
+	}
+
+	for (i = 0; i < gr->tpc_total;) {
+		for (j = 0; j < gr->gpc_nr; j++) {
+			if ((run_err[j] * 2) >= comm_denom) {
+				gr->tile[i++] = gpc_map[j];
+				run_err[j] += init_frac[j] - comm_denom;
+			} else {
+				run_err[j] += init_frac[j];
+			}
+		}
+	}
+}
+
 static int
 gf100_gr_oneinit(struct nvkm_gr *base)
 {
@@ -1691,45 +1767,8 @@ gf100_gr_oneinit(struct nvkm_gr *base)
 		}
 	}
 
-	/*XXX: these need figuring out... though it might not even matter */
-	switch (device->chipset) {
-	case 0xc0:
-		if (gr->tpc_total == 11) { /* 465, 3/4/4/0, 4 */
-			gr->screen_tile_row_offset = 0x07;
-		} else
-		if (gr->tpc_total == 14) { /* 470, 3/3/4/4, 5 */
-			gr->screen_tile_row_offset = 0x05;
-		} else
-		if (gr->tpc_total == 15) { /* 480, 3/4/4/4, 6 */
-			gr->screen_tile_row_offset = 0x06;
-		}
-		break;
-	case 0xc3: /* 450, 4/0/0/0, 2 */
-		gr->screen_tile_row_offset = 0x03;
-		break;
-	case 0xc4: /* 460, 3/4/0/0, 4 */
-		gr->screen_tile_row_offset = 0x01;
-		break;
-	case 0xc1: /* 2/0/0/0, 1 */
-		gr->screen_tile_row_offset = 0x01;
-		break;
-	case 0xc8: /* 4/4/3/4, 5 */
-		gr->screen_tile_row_offset = 0x06;
-		break;
-	case 0xce: /* 4/4/0/0, 4 */
-		gr->screen_tile_row_offset = 0x03;
-		break;
-	case 0xcf: /* 4/0/0/0, 3 */
-		gr->screen_tile_row_offset = 0x03;
-		break;
-	case 0xd7:
-	case 0xd9: /* 1/0/0/0, 1 */
-	case 0xea: /* gk20a */
-	case 0x12b: /* gm20b */
-		gr->screen_tile_row_offset = 0x01;
-		break;
-	}
-
+	memset(gr->tile, 0xff, sizeof(gr->tile));
+	gr->func->oneinit_tiles(gr);
 	return 0;
 }
 
@@ -2164,6 +2203,7 @@ gf100_gr_gpccs_ucode = {
 
 static const struct gf100_gr_func
 gf100_gr = {
+	.oneinit_tiles = gf100_gr_oneinit_tiles,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index c25b93a..53a173e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -107,12 +107,13 @@ struct gf100_gr {
 	u8 ppc_tpc_nr[GPC_MAX][4];
 	u8 ppc_tpc_min;
 
+	u8 screen_tile_row_offset;
+	u8 tile[TPC_MAX];
+
 	struct gf100_gr_data mmio_data[4];
 	struct gf100_gr_mmio mmio_list[4096/8];
 	u32  size;
 	u32 *data;
-
-	u8 screen_tile_row_offset;
 };
 
 int gf100_gr_ctor(const struct gf100_gr_func *, struct nvkm_device *,
@@ -123,6 +124,7 @@ void *gf100_gr_dtor(struct nvkm_gr *);
 
 struct gf100_gr_func {
 	void (*dtor)(struct gf100_gr *);
+	void (*oneinit_tiles)(struct gf100_gr *);
 	int (*init)(struct gf100_gr *);
 	void (*init_gpc_mmu)(struct gf100_gr *);
 	void (*init_r405a14)(struct gf100_gr *);
@@ -164,6 +166,7 @@ struct gf100_gr_func {
 };
 
 int gf100_gr_rops(struct gf100_gr *);
+void gf100_gr_oneinit_tiles(struct gf100_gr *);
 int gf100_gr_init(struct gf100_gr *);
 void gf100_gr_init_vsc_stream_master(struct gf100_gr *);
 void gf100_gr_init_zcull(struct gf100_gr *);
@@ -191,6 +194,7 @@ void gm107_gr_init_400054(struct gf100_gr *);
 
 int gk20a_gr_init(struct gf100_gr *);
 
+void gm200_gr_oneinit_tiles(struct gf100_gr *);
 int gm200_gr_rops(struct gf100_gr *);
 void gm200_gr_init_num_active_ltcs(struct gf100_gr *);
 void gm200_gr_init_ds_hww_esr_2(struct gf100_gr *);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
index df9cbed..8b49b8f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
@@ -114,6 +114,7 @@ gf104_gr_pack_mmio[] = {
 
 static const struct gf100_gr_func
 gf104_gr = {
+	.oneinit_tiles = gf100_gr_oneinit_tiles,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
index 8ffa0fd..6432aeb 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
@@ -111,6 +111,7 @@ gf108_gr_init_r405a14(struct gf100_gr *gr)
 
 static const struct gf100_gr_func
 gf108_gr = {
+	.oneinit_tiles = gf100_gr_oneinit_tiles,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_r405a14 = gf108_gr_init_r405a14,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
index 0d4293e..4e007c9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
@@ -86,6 +86,7 @@ gf110_gr_pack_mmio[] = {
 
 static const struct gf100_gr_func
 gf110_gr = {
+	.oneinit_tiles = gf100_gr_oneinit_tiles,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
index e3c1dbb..2ddb728 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
@@ -150,6 +150,7 @@ gf117_gr_init_zcull(struct gf100_gr *gr)
 
 static const struct gf100_gr_func
 gf117_gr = {
+	.oneinit_tiles = gf100_gr_oneinit_tiles,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
index 1ed70b9..f0f10a4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
@@ -177,6 +177,7 @@ gf119_gr_pack_mmio[] = {
 
 static const struct gf100_gr_func
 gf119_gr = {
+	.oneinit_tiles = gf100_gr_oneinit_tiles,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
index 86819ab..d57fb5f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
@@ -448,6 +448,7 @@ gk104_gr_gpccs_ucode = {
 
 static const struct gf100_gr_func
 gk104_gr = {
+	.oneinit_tiles = gf100_gr_oneinit_tiles,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
index e30d94f..41997eb 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
@@ -350,6 +350,7 @@ gk110_gr_init_419eb4(struct gf100_gr *gr)
 
 static const struct gf100_gr_func
 gk110_gr = {
+	.oneinit_tiles = gf100_gr_oneinit_tiles,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
index 253b981..b7a6479 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
@@ -102,6 +102,7 @@ gk110b_gr_pack_mmio[] = {
 
 static const struct gf100_gr_func
 gk110b_gr = {
+	.oneinit_tiles = gf100_gr_oneinit_tiles,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
index 702e909..5f1e71a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
@@ -161,6 +161,7 @@ gk208_gr_gpccs_ucode = {
 
 static const struct gf100_gr_func
 gk208_gr = {
+	.oneinit_tiles = gf100_gr_oneinit_tiles,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c
index 95f7d85..ab4e538 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c
@@ -282,6 +282,7 @@ gk20a_gr_init(struct gf100_gr *gr)
 
 static const struct gf100_gr_func
 gk20a_gr = {
+	.oneinit_tiles = gf100_gr_oneinit_tiles,
 	.init = gk20a_gr_init,
 	.init_zcull = gf117_gr_init_zcull,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
index d67bf94..98f74fe 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
@@ -391,6 +391,7 @@ gm107_gr_gpccs_ucode = {
 
 static const struct gf100_gr_func
 gm107_gr = {
+	.oneinit_tiles = gf100_gr_oneinit_tiles,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gm107_gr_init_gpc_mmu,
 	.init_bios = gm107_gr_init_bios,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
index 03b255e..3e017fd 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
@@ -77,6 +77,46 @@ gm200_gr_init_rop_active_fbps(struct gf100_gr *gr)
 	nvkm_mask(device, 0x408958, 0x0000000f, fbp_count); /* crop */
 }
 
+static u8
+gm200_gr_tile_map_6_24[] = {
+	0, 1, 2, 3, 4, 5, 3, 4, 5, 0, 1, 2, 0, 1, 2, 3, 4, 5, 3, 4, 5, 0, 1, 2,
+};
+
+static u8
+gm200_gr_tile_map_4_16[] = {
+	0, 1, 2, 3, 2, 3, 0, 1, 3, 0, 1, 2, 1, 2, 3, 0,
+};
+
+static u8
+gm200_gr_tile_map_2_8[] = {
+	0, 1, 1, 0, 0, 1, 1, 0,
+};
+
+void
+gm200_gr_oneinit_tiles(struct gf100_gr *gr)
+{
+	/*XXX: Not sure what this is about.  The algorithm from NVGPU
+	 *     seems to work for all boards I tried from earlier (and
+	 *     later) GPUs except in these specific configurations.
+	 *
+	 *     Let's just hardcode them for now.
+	 */
+	if (gr->gpc_nr == 2 && gr->tpc_total == 8) {
+		memcpy(gr->tile, gm200_gr_tile_map_2_8, gr->tpc_total);
+		gr->screen_tile_row_offset = 1;
+	} else
+	if (gr->gpc_nr == 4 && gr->tpc_total == 16) {
+		memcpy(gr->tile, gm200_gr_tile_map_4_16, gr->tpc_total);
+		gr->screen_tile_row_offset = 4;
+	} else
+	if (gr->gpc_nr == 6 && gr->tpc_total == 24) {
+		memcpy(gr->tile, gm200_gr_tile_map_6_24, gr->tpc_total);
+		gr->screen_tile_row_offset = 5;
+	} else {
+		gf100_gr_oneinit_tiles(gr);
+	}
+}
+
 int
 gm200_gr_new_(const struct gf100_gr_func *func, struct nvkm_device *device,
 	      int index, struct nvkm_gr **pgr)
@@ -117,6 +157,7 @@ gm200_gr_new_(const struct gf100_gr_func *func, struct nvkm_device *device,
 
 static const struct gf100_gr_func
 gm200_gr = {
+	.oneinit_tiles = gm200_gr_oneinit_tiles,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_bios = gm107_gr_init_bios,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c
index d2f9c7b..29d3b94 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c
@@ -64,6 +64,7 @@ gm20b_gr_set_hww_esr_report_mask(struct gf100_gr *gr)
 
 static const struct gf100_gr_func
 gm20b_gr = {
+	.oneinit_tiles = gm200_gr_oneinit_tiles,
 	.init = gk20a_gr_init,
 	.init_zcull = gf117_gr_init_zcull,
 	.init_gpc_mmu = gm20b_gr_init_gpc_mmu,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
index e5f941f..70d0aa2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
@@ -64,6 +64,7 @@ gp100_gr_init_rop_active_fbps(struct gf100_gr *gr)
 
 static const struct gf100_gr_func
 gp100_gr = {
+	.oneinit_tiles = gm200_gr_oneinit_tiles,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
index 09e2665..71bfe2d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
@@ -42,6 +42,7 @@ gp102_gr_init_swdx_pes_mask(struct gf100_gr *gr)
 
 static const struct gf100_gr_func
 gp102_gr = {
+	.oneinit_tiles = gm200_gr_oneinit_tiles,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c
index 844fc9d..234c970 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c
@@ -26,6 +26,7 @@
 
 static const struct gf100_gr_func
 gp104_gr = {
+	.oneinit_tiles = gm200_gr_oneinit_tiles,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
index 674385d..6c57240 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
@@ -28,6 +28,7 @@
 
 static const struct gf100_gr_func
 gp107_gr = {
+	.oneinit_tiles = gm200_gr_oneinit_tiles,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
index 6103186..aaaa284 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
@@ -27,6 +27,7 @@
 
 static const struct gf100_gr_func
 gp10b_gr = {
+	.oneinit_tiles = gm200_gr_oneinit_tiles,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
-- 
cgit v1.1


From d00ffc0c403784c9f88d8da357f9f33f855289a4 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/gr/gf100-: port zcull tile mapping calculations from
 NVGPU

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c | 34 +++++++++++++-------------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c | 34 +++++++++++++-------------
 2 files changed, 34 insertions(+), 34 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 519b109..f6d8841 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -2042,25 +2042,25 @@ gf100_gr_init_zcull(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
 	const u32 magicgpc918 = DIV_ROUND_UP(0x00800000, gr->tpc_total);
-	u32 data[TPC_MAX / 8] = {};
-	u8  tpcnr[GPC_MAX];
-	int gpc, tpc;
-	int i;
-
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-	for (i = 0, gpc = -1; i < gr->tpc_total; i++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while (!tpcnr[gpc]);
-		tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--;
-
-		data[i / 8] |= tpc << ((i % 8) * 4);
+	const u8 tile_nr = ALIGN(gr->tpc_total, 32);
+	u8 bank[GPC_MAX] = {}, gpc, i, j;
+	u32 data;
+
+	for (i = 0; i < tile_nr; i += 8) {
+		for (data = 0, j = 0; j < 8 && i + j < gr->tpc_total; j++) {
+			data |= bank[gr->tile[i + j]] << (j * 4);
+			bank[gr->tile[i + j]]++;
+		}
+		nvkm_wr32(device, GPC_BCAST(0x0980 + ((i / 8) * 4)), data);
 	}
 
-	nvkm_wr32(device, GPC_BCAST(0x0980), data[0]);
-	nvkm_wr32(device, GPC_BCAST(0x0984), data[1]);
-	nvkm_wr32(device, GPC_BCAST(0x0988), data[2]);
-	nvkm_wr32(device, GPC_BCAST(0x098c), data[3]);
+	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
+		nvkm_wr32(device, GPC_UNIT(gpc, 0x0914),
+			  gr->screen_tile_row_offset << 8 | gr->tpc_nr[gpc]);
+		nvkm_wr32(device, GPC_UNIT(gpc, 0x0910), 0x00040000 |
+							 gr->tpc_total);
+		nvkm_wr32(device, GPC_UNIT(gpc, 0x0918), magicgpc918);
+	}
 
 	nvkm_wr32(device, GPC_BCAST(0x1bd4), magicgpc918);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
index 2ddb728..d6831b4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
@@ -125,25 +125,25 @@ gf117_gr_init_zcull(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
 	const u32 magicgpc918 = DIV_ROUND_UP(0x00800000, gr->tpc_total);
-	u32 data[TPC_MAX / 8] = {};
-	u8  tpcnr[GPC_MAX];
-	int gpc, tpc;
-	int i;
-
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-	for (i = 0, gpc = -1; i < gr->tpc_total; i++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while (!tpcnr[gpc]);
-		tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--;
-
-		data[i / 8] |= tpc << ((i % 8) * 4);
+	const u8 tile_nr = ALIGN(gr->tpc_total, 32);
+	u8 bank[GPC_MAX] = {}, gpc, i, j;
+	u32 data;
+
+	for (i = 0; i < tile_nr; i += 8) {
+		for (data = 0, j = 0; j < 8 && i + j < gr->tpc_total; j++) {
+			data |= bank[gr->tile[i + j]] << (j * 4);
+			bank[gr->tile[i + j]]++;
+		}
+		nvkm_wr32(device, GPC_BCAST(0x0980 + ((i / 8) * 4)), data);
 	}
 
-	nvkm_wr32(device, GPC_BCAST(0x0980), data[0]);
-	nvkm_wr32(device, GPC_BCAST(0x0984), data[1]);
-	nvkm_wr32(device, GPC_BCAST(0x0988), data[2]);
-	nvkm_wr32(device, GPC_BCAST(0x098c), data[3]);
+	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
+		nvkm_wr32(device, GPC_UNIT(gpc, 0x0914),
+			  gr->screen_tile_row_offset << 8 | gr->tpc_nr[gpc]);
+		nvkm_wr32(device, GPC_UNIT(gpc, 0x0910), 0x00040000 |
+							 gr->tpc_total);
+		nvkm_wr32(device, GPC_UNIT(gpc, 0x0918), magicgpc918);
+	}
 
 	nvkm_wr32(device, GPC_BCAST(0x3fd4), magicgpc918);
 }
-- 
cgit v1.1


From 068cae743c2ad08a082d6fef007e6b38f5fb3b16 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/gr/gf100-: calculate and use sm mapping table

There's a number of places that require this data, so let's separate out
the calculations to ensure they remain consistent.

This is incorrect for GM200 and newer, but will produce the same results
as we did before.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c | 40 +++++++++--------------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c | 22 ++++---------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c | 22 ++++---------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c    | 17 ++++++++++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h    |  9 +++++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c    |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c    |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c    |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c    |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c    |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c    |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c    |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c   |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c    |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c    |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c    |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c    |  8 +++++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c    |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c    |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c    |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c    |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c    |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c    |  1 +
 23 files changed, 79 insertions(+), 56 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index 949e121..c2dcc01 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1092,23 +1092,18 @@ gf100_grctx_generate_r4060a8(struct gf100_gr *gr)
 	struct nvkm_device *device = gr->base.engine.subdev.device;
 	const u8 gpcmax = nvkm_rd32(device, 0x022430);
 	const u8 tpcmax = nvkm_rd32(device, 0x022434) * gpcmax;
-	u8 tpcnr[GPC_MAX], data[TPC_MAX];
-	int gpc, tpc, i;
-
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-	memset(data, 0x1f, sizeof(data));
-
-	gpc = -1;
-	for (tpc = 0; tpc < gr->tpc_total; tpc++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while (!tpcnr[gpc]);
-		tpcnr[gpc]--;
-		data[tpc] = gpc;
-	}
+	int i, j, sm = 0;
+	u32 data;
 
-	for (i = 0; i < DIV_ROUND_UP(tpcmax, 4); i++)
-		nvkm_wr32(device, 0x4060a8 + (i * 4), ((u32 *)data)[i]);
+	for (i = 0; i < DIV_ROUND_UP(tpcmax, 4); i++) {
+		for (data = 0, j = 0; j < 4; j++) {
+			if (sm < gr->sm_nr)
+				data |= gr->sm[sm++].gpc << (j * 8);
+			else
+				data |= 0x1f << (j * 8);
+		}
+		nvkm_wr32(device, 0x4060a8 + (i * 4), data);
+	}
 }
 
 void
@@ -1326,16 +1321,13 @@ gf100_grctx_generate_floorsweep(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
 	const struct gf100_grctx_func *func = gr->func->grctx;
-	int tpc, gpc, sm, i, j;
+	int gpc, sm, i, j;
 	u32 data;
 
-	for (tpc = 0, sm = 0; tpc < gr->tpc_max; tpc++) {
-		for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-			if (tpc < gr->tpc_nr[gpc])
-				func->sm_id(gr, gpc, tpc, sm++);
-			if (func->tpc_nr)
-				func->tpc_nr(gr, gpc);
-		}
+	for (sm = 0; sm < gr->sm_nr; sm++) {
+		func->sm_id(gr, gr->sm[sm].gpc, gr->sm[sm].tpc, sm);
+		if (func->tpc_nr)
+			func->tpc_nr(gr, gr->sm[sm].gpc);
 	}
 
 	for (gpc = 0, i = 0; i < 4; i++) {
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
index 7107ec4..013d05a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
@@ -49,23 +49,13 @@ gm200_grctx_generate_smid_config(struct gf100_gr *gr)
 	const u32 dist_nr = DIV_ROUND_UP(gr->tpc_total, 4);
 	u32 dist[TPC_MAX / 4] = {};
 	u32 gpcs[GPC_MAX] = {};
-	u8  tpcnr[GPC_MAX];
-	int tpc, gpc, i;
+	u8  sm, i;
 
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-
-	/* won't result in the same distribution as the binary driver where
-	 * some of the gpcs have more tpcs than others, but this shall do
-	 * for the moment.  the code for earlier gpus has this issue too.
-	 */
-	for (gpc = -1, i = 0; i < gr->tpc_total; i++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while(!tpcnr[gpc]);
-		tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--;
-
-		dist[i / 4] |= ((gpc << 4) | tpc) << ((i % 4) * 8);
-		gpcs[gpc] |= i << (tpc * 8);
+	for (sm = 0; sm < gr->sm_nr; sm++) {
+		const u8 gpc = gr->sm[sm].gpc;
+		const u8 tpc = gr->sm[sm].tpc;
+		dist[sm / 4] |= ((gpc << 4) | tpc) << ((sm % 4) * 8);
+		gpcs[gpc] |= sm << (tpc * 8);
 	}
 
 	for (i = 0; i < dist_nr; i++)
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
index af6330c..2344fd8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
@@ -95,23 +95,13 @@ gp100_grctx_generate_smid_config(struct gf100_gr *gr)
 	struct nvkm_device *device = gr->base.engine.subdev.device;
 	const u32 dist_nr = DIV_ROUND_UP(gr->tpc_total, 4);
 	u32 dist[TPC_MAX / 4] = {}, gpcs[16] = {};
-	u8  tpcnr[GPC_MAX];
-	int tpc, gpc, i;
+	u8  sm, i;
 
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-
-	/* won't result in the same distribution as the binary driver where
-	 * some of the gpcs have more tpcs than others, but this shall do
-	 * for the moment.  the code for earlier gpus has this issue too.
-	 */
-	for (gpc = -1, i = 0; i < gr->tpc_total; i++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while(!tpcnr[gpc]);
-		tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--;
-
-		dist[i / 4] |= ((gpc << 4) | tpc) << ((i % 4) * 8);
-		gpcs[gpc + (gr->func->gpc_nr * (tpc / 4))] |= i << (tpc * 8);
+	for (sm = 0; sm < gr->sm_nr; sm++) {
+		const u8 gpc = gr->sm[sm].gpc;
+		const u8 tpc = gr->sm[sm].tpc;
+		dist[sm / 4] |= ((gpc << 4) | tpc) << ((sm % 4) * 8);
+		gpcs[gpc + (gr->func->gpc_nr * (tpc / 4))] |= sm << ((tpc % 4) * 8);
 	}
 
 	for (i = 0; i < dist_nr; i++)
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index f6d8841..b2070c8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -1653,6 +1653,21 @@ gf100_gr_init_ctxctl(struct gf100_gr *gr)
 }
 
 void
+gf100_gr_oneinit_sm_id(struct gf100_gr *gr)
+{
+	int tpc, gpc;
+	for (tpc = 0; tpc < gr->tpc_max; tpc++) {
+		for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
+			if (tpc < gr->tpc_nr[gpc]) {
+				gr->sm[gr->sm_nr].gpc = gpc;
+				gr->sm[gr->sm_nr].tpc = tpc;
+				gr->sm_nr++;
+			}
+		}
+	}
+}
+
+void
 gf100_gr_oneinit_tiles(struct gf100_gr *gr)
 {
 	static const u8 primes[] = {
@@ -1769,6 +1784,7 @@ gf100_gr_oneinit(struct nvkm_gr *base)
 
 	memset(gr->tile, 0xff, sizeof(gr->tile));
 	gr->func->oneinit_tiles(gr);
+	gr->func->oneinit_sm_id(gr);
 	return 0;
 }
 
@@ -2204,6 +2220,7 @@ gf100_gr_gpccs_ucode = {
 static const struct gf100_gr_func
 gf100_gr = {
 	.oneinit_tiles = gf100_gr_oneinit_tiles,
+	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index 53a173e..55dedd8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -110,6 +110,12 @@ struct gf100_gr {
 	u8 screen_tile_row_offset;
 	u8 tile[TPC_MAX];
 
+	struct {
+		u8 gpc;
+		u8 tpc;
+	} sm[TPC_MAX];
+	u8 sm_nr;
+
 	struct gf100_gr_data mmio_data[4];
 	struct gf100_gr_mmio mmio_list[4096/8];
 	u32  size;
@@ -125,6 +131,7 @@ void *gf100_gr_dtor(struct nvkm_gr *);
 struct gf100_gr_func {
 	void (*dtor)(struct gf100_gr *);
 	void (*oneinit_tiles)(struct gf100_gr *);
+	void (*oneinit_sm_id)(struct gf100_gr *);
 	int (*init)(struct gf100_gr *);
 	void (*init_gpc_mmu)(struct gf100_gr *);
 	void (*init_r405a14)(struct gf100_gr *);
@@ -167,6 +174,7 @@ struct gf100_gr_func {
 
 int gf100_gr_rops(struct gf100_gr *);
 void gf100_gr_oneinit_tiles(struct gf100_gr *);
+void gf100_gr_oneinit_sm_id(struct gf100_gr *);
 int gf100_gr_init(struct gf100_gr *);
 void gf100_gr_init_vsc_stream_master(struct gf100_gr *);
 void gf100_gr_init_zcull(struct gf100_gr *);
@@ -195,6 +203,7 @@ void gm107_gr_init_400054(struct gf100_gr *);
 int gk20a_gr_init(struct gf100_gr *);
 
 void gm200_gr_oneinit_tiles(struct gf100_gr *);
+void gm200_gr_oneinit_sm_id(struct gf100_gr *);
 int gm200_gr_rops(struct gf100_gr *);
 void gm200_gr_init_num_active_ltcs(struct gf100_gr *);
 void gm200_gr_init_ds_hww_esr_2(struct gf100_gr *);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
index 8b49b8f..1d8e16a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
@@ -115,6 +115,7 @@ gf104_gr_pack_mmio[] = {
 static const struct gf100_gr_func
 gf104_gr = {
 	.oneinit_tiles = gf100_gr_oneinit_tiles,
+	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
index 6432aeb..a5a74df 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
@@ -112,6 +112,7 @@ gf108_gr_init_r405a14(struct gf100_gr *gr)
 static const struct gf100_gr_func
 gf108_gr = {
 	.oneinit_tiles = gf100_gr_oneinit_tiles,
+	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_r405a14 = gf108_gr_init_r405a14,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
index 4e007c9..45fada0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
@@ -87,6 +87,7 @@ gf110_gr_pack_mmio[] = {
 static const struct gf100_gr_func
 gf110_gr = {
 	.oneinit_tiles = gf100_gr_oneinit_tiles,
+	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
index d6831b4..5ee167d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
@@ -151,6 +151,7 @@ gf117_gr_init_zcull(struct gf100_gr *gr)
 static const struct gf100_gr_func
 gf117_gr = {
 	.oneinit_tiles = gf100_gr_oneinit_tiles,
+	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
index f0f10a4..2096552 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
@@ -178,6 +178,7 @@ gf119_gr_pack_mmio[] = {
 static const struct gf100_gr_func
 gf119_gr = {
 	.oneinit_tiles = gf100_gr_oneinit_tiles,
+	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
index d57fb5f..cce250a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
@@ -449,6 +449,7 @@ gk104_gr_gpccs_ucode = {
 static const struct gf100_gr_func
 gk104_gr = {
 	.oneinit_tiles = gf100_gr_oneinit_tiles,
+	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
index 41997eb..558b497 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
@@ -351,6 +351,7 @@ gk110_gr_init_419eb4(struct gf100_gr *gr)
 static const struct gf100_gr_func
 gk110_gr = {
 	.oneinit_tiles = gf100_gr_oneinit_tiles,
+	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
index b7a6479..c8c48a2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
@@ -103,6 +103,7 @@ gk110b_gr_pack_mmio[] = {
 static const struct gf100_gr_func
 gk110b_gr = {
 	.oneinit_tiles = gf100_gr_oneinit_tiles,
+	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
index 5f1e71a..477a7de 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
@@ -162,6 +162,7 @@ gk208_gr_gpccs_ucode = {
 static const struct gf100_gr_func
 gk208_gr = {
 	.oneinit_tiles = gf100_gr_oneinit_tiles,
+	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c
index ab4e538..11a32fa 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c
@@ -283,6 +283,7 @@ gk20a_gr_init(struct gf100_gr *gr)
 static const struct gf100_gr_func
 gk20a_gr = {
 	.oneinit_tiles = gf100_gr_oneinit_tiles,
+	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
 	.init = gk20a_gr_init,
 	.init_zcull = gf117_gr_init_zcull,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
index 98f74fe..a6937f8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
@@ -392,6 +392,7 @@ gm107_gr_gpccs_ucode = {
 static const struct gf100_gr_func
 gm107_gr = {
 	.oneinit_tiles = gf100_gr_oneinit_tiles,
+	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gm107_gr_init_gpc_mmu,
 	.init_bios = gm107_gr_init_bios,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
index 3e017fd..8966d2a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
@@ -93,6 +93,13 @@ gm200_gr_tile_map_2_8[] = {
 };
 
 void
+gm200_gr_oneinit_sm_id(struct gf100_gr *gr)
+{
+	/*XXX: There's a different algorithm here I've not yet figured out. */
+	gf100_gr_oneinit_sm_id(gr);
+}
+
+void
 gm200_gr_oneinit_tiles(struct gf100_gr *gr)
 {
 	/*XXX: Not sure what this is about.  The algorithm from NVGPU
@@ -158,6 +165,7 @@ gm200_gr_new_(const struct gf100_gr_func *func, struct nvkm_device *device,
 static const struct gf100_gr_func
 gm200_gr = {
 	.oneinit_tiles = gm200_gr_oneinit_tiles,
+	.oneinit_sm_id = gm200_gr_oneinit_sm_id,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_bios = gm107_gr_init_bios,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c
index 29d3b94..afa1c6e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c
@@ -65,6 +65,7 @@ gm20b_gr_set_hww_esr_report_mask(struct gf100_gr *gr)
 static const struct gf100_gr_func
 gm20b_gr = {
 	.oneinit_tiles = gm200_gr_oneinit_tiles,
+	.oneinit_sm_id = gm200_gr_oneinit_sm_id,
 	.init = gk20a_gr_init,
 	.init_zcull = gf117_gr_init_zcull,
 	.init_gpc_mmu = gm20b_gr_init_gpc_mmu,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
index 70d0aa2..1d9d876 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
@@ -65,6 +65,7 @@ gp100_gr_init_rop_active_fbps(struct gf100_gr *gr)
 static const struct gf100_gr_func
 gp100_gr = {
 	.oneinit_tiles = gm200_gr_oneinit_tiles,
+	.oneinit_sm_id = gm200_gr_oneinit_sm_id,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
index 71bfe2d..2d9a2c3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
@@ -43,6 +43,7 @@ gp102_gr_init_swdx_pes_mask(struct gf100_gr *gr)
 static const struct gf100_gr_func
 gp102_gr = {
 	.oneinit_tiles = gm200_gr_oneinit_tiles,
+	.oneinit_sm_id = gm200_gr_oneinit_sm_id,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c
index 234c970..e466ae4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c
@@ -27,6 +27,7 @@
 static const struct gf100_gr_func
 gp104_gr = {
 	.oneinit_tiles = gm200_gr_oneinit_tiles,
+	.oneinit_sm_id = gm200_gr_oneinit_sm_id,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
index 6c57240..2fa046a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
@@ -29,6 +29,7 @@
 static const struct gf100_gr_func
 gp107_gr = {
 	.oneinit_tiles = gm200_gr_oneinit_tiles,
+	.oneinit_sm_id = gm200_gr_oneinit_sm_id,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
index aaaa284..0a01a30 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
@@ -28,6 +28,7 @@
 static const struct gf100_gr_func
 gp10b_gr = {
 	.oneinit_tiles = gm200_gr_oneinit_tiles,
+	.oneinit_sm_id = gm200_gr_oneinit_sm_id,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
-- 
cgit v1.1


From 6f0233329bacd8de54959e0c0b9b6c46bf5118b8 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/gr/gf100-: swap bundle and pagepool

Makes it easier to diff against RM traces.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index c2dcc01..4096e2d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1381,8 +1381,8 @@ gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	idle_timeout = nvkm_mask(device, 0x404154, 0xffffffff, 0x00000000);
 
-	grctx->bundle(info);
 	grctx->pagepool(info);
+	grctx->bundle(info);
 	grctx->attrib(info);
 	if (grctx->patch_ltc)
 		grctx->patch_ltc(info);
-- 
cgit v1.1


From 191e323278dd7025ecd5fef02fbb984cfc91eebb Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/gr/gf100-gm10x: update register lists

There are differences on GM200 and newer too, but we can't fix them there
as they come from firmware packages.

A request has been made to NVIDIA to release updated firmware.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c  | 13 ++++++++++---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h  | 12 ++++++++----
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c  |  3 ++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c  | 10 ++++++++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c  |  8 +++-----
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c  | 10 ++++++++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c  |  8 +++-----
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c  | 12 +++++++++---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c  | 10 ++++++++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c |  3 ++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c  | 10 ++++++++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c  | 10 ++++++++--
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c     |  3 ++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h     |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c     |  7 +++++++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c     |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c    |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c     |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c     |  9 ++++++---
 19 files changed, 96 insertions(+), 36 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index 4096e2d..1ed63ed 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -850,12 +850,17 @@ gf100_grctx_init_gcc_0[] = {
 };
 
 const struct gf100_gr_pack
-gf100_grctx_pack_gpc[] = {
+gf100_grctx_pack_gpc_0[] = {
 	{ gf100_grctx_init_gpc_unk_0 },
 	{ gf100_grctx_init_prop_0 },
 	{ gf100_grctx_init_gpc_unk_1 },
 	{ gf100_grctx_init_setup_0 },
 	{ gf100_grctx_init_zcull_0 },
+	{}
+};
+
+const struct gf100_gr_pack
+gf100_grctx_pack_gpc_1[] = {
 	{ gf100_grctx_init_crstr_0 },
 	{ gf100_grctx_init_gpm_0 },
 	{ gf100_grctx_init_gcc_0 },
@@ -1371,8 +1376,9 @@ gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	if (!gr->fuc_sw_ctx) {
 		gf100_gr_mmio(gr, grctx->hub);
-		gf100_gr_mmio(gr, grctx->gpc);
+		gf100_gr_mmio(gr, grctx->gpc_0);
 		gf100_gr_mmio(gr, grctx->zcull);
+		gf100_gr_mmio(gr, grctx->gpc_1);
 		gf100_gr_mmio(gr, grctx->tpc);
 		gf100_gr_mmio(gr, grctx->ppc);
 	} else {
@@ -1567,7 +1573,8 @@ gf100_grctx = {
 	.main  = gf100_grctx_generate_main,
 	.unkn  = gf100_grctx_generate_unkn,
 	.hub   = gf100_grctx_pack_hub,
-	.gpc   = gf100_grctx_pack_gpc,
+	.gpc_0 = gf100_grctx_pack_gpc_0,
+	.gpc_1 = gf100_grctx_pack_gpc_1,
 	.zcull = gf100_grctx_pack_zcull,
 	.tpc   = gf100_grctx_pack_tpc,
 	.icmd  = gf100_grctx_pack_icmd,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
index be57ff0..1c06c67 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
@@ -27,7 +27,8 @@ struct gf100_grctx_func {
 	void  (*unkn)(struct gf100_gr *);
 	/* mmio context data */
 	const struct gf100_gr_pack *hub;
-	const struct gf100_gr_pack *gpc;
+	const struct gf100_gr_pack *gpc_0;
+	const struct gf100_gr_pack *gpc_1;
 	const struct gf100_gr_pack *zcull;
 	const struct gf100_gr_pack *tpc;
 	const struct gf100_gr_pack *ppc;
@@ -163,7 +164,8 @@ extern const struct gf100_gr_init gf100_grctx_init_memfmt_0[];
 extern const struct gf100_gr_init gf100_grctx_init_rstr2d_0[];
 extern const struct gf100_gr_init gf100_grctx_init_scc_0[];
 
-extern const struct gf100_gr_pack gf100_grctx_pack_gpc[];
+extern const struct gf100_gr_pack gf100_grctx_pack_gpc_0[];
+extern const struct gf100_gr_pack gf100_grctx_pack_gpc_1[];
 extern const struct gf100_gr_init gf100_grctx_init_gpc_unk_0[];
 extern const struct gf100_gr_init gf100_grctx_init_prop_0[];
 extern const struct gf100_gr_init gf100_grctx_init_gpc_unk_1[];
@@ -212,6 +214,8 @@ extern const struct gf100_gr_init gf117_grctx_init_pe_0[];
 
 extern const struct gf100_gr_init gf117_grctx_init_wwdx_0[];
 
+extern const struct gf100_gr_pack gf117_grctx_pack_gpc_1[];
+
 extern const struct gf100_gr_init gk104_grctx_init_memfmt_0[];
 extern const struct gf100_gr_init gk104_grctx_init_ds_0[];
 extern const struct gf100_gr_init gk104_grctx_init_scc_0[];
@@ -221,7 +225,6 @@ extern const struct gf100_gr_init gk104_grctx_init_gpm_0[];
 extern const struct gf100_gr_init gk104_grctx_init_pes_0[];
 
 extern const struct gf100_gr_pack gk104_grctx_pack_hub[];
-extern const struct gf100_gr_pack gk104_grctx_pack_gpc[];
 extern const struct gf100_gr_pack gk104_grctx_pack_tpc[];
 extern const struct gf100_gr_pack gk104_grctx_pack_ppc[];
 extern const struct gf100_gr_pack gk104_grctx_pack_icmd[];
@@ -235,7 +238,8 @@ extern const struct gf100_gr_pack gk110_grctx_pack_hub[];
 extern const struct gf100_gr_init gk110_grctx_init_pri_0[];
 extern const struct gf100_gr_init gk110_grctx_init_cwd_0[];
 
-extern const struct gf100_gr_pack gk110_grctx_pack_gpc[];
+extern const struct gf100_gr_pack gk110_grctx_pack_gpc_0[];
+extern const struct gf100_gr_pack gk110_grctx_pack_gpc_1[];
 extern const struct gf100_gr_init gk110_grctx_init_gpc_unk_2[];
 
 extern const struct gf100_gr_init gk110_grctx_init_tex_0[];
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c
index 7f3b928..7a0564b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c
@@ -84,7 +84,8 @@ gf104_grctx = {
 	.main  = gf100_grctx_generate_main,
 	.unkn  = gf100_grctx_generate_unkn,
 	.hub   = gf100_grctx_pack_hub,
-	.gpc   = gf100_grctx_pack_gpc,
+	.gpc_0 = gf100_grctx_pack_gpc_0,
+	.gpc_1 = gf100_grctx_pack_gpc_1,
 	.zcull = gf100_grctx_pack_zcull,
 	.tpc   = gf104_grctx_pack_tpc,
 	.icmd  = gf100_grctx_pack_icmd,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c
index 369d64f..dda2c32 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c
@@ -667,12 +667,17 @@ gf108_grctx_init_gpm_0[] = {
 };
 
 static const struct gf100_gr_pack
-gf108_grctx_pack_gpc[] = {
+gf108_grctx_pack_gpc_0[] = {
 	{ gf100_grctx_init_gpc_unk_0 },
 	{ gf100_grctx_init_prop_0 },
 	{ gf100_grctx_init_gpc_unk_1 },
 	{ gf108_grctx_init_setup_0 },
 	{ gf100_grctx_init_zcull_0 },
+	{}
+};
+
+static const struct gf100_gr_pack
+gf108_grctx_pack_gpc_1[] = {
 	{ gf100_grctx_init_crstr_0 },
 	{ gf108_grctx_init_gpm_0 },
 	{ gf100_grctx_init_gcc_0 },
@@ -780,7 +785,8 @@ gf108_grctx = {
 	.main  = gf100_grctx_generate_main,
 	.unkn  = gf108_grctx_generate_unkn,
 	.hub   = gf108_grctx_pack_hub,
-	.gpc   = gf108_grctx_pack_gpc,
+	.gpc_0 = gf108_grctx_pack_gpc_0,
+	.gpc_1 = gf108_grctx_pack_gpc_1,
 	.zcull = gf100_grctx_pack_zcull,
 	.tpc   = gf108_grctx_pack_tpc,
 	.icmd  = gf108_grctx_pack_icmd,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c
index d59c248..f5cca5e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c
@@ -314,15 +314,12 @@ gf110_grctx_init_setup_0[] = {
 };
 
 static const struct gf100_gr_pack
-gf110_grctx_pack_gpc[] = {
+gf110_grctx_pack_gpc_0[] = {
 	{ gf100_grctx_init_gpc_unk_0 },
 	{ gf100_grctx_init_prop_0 },
 	{ gf100_grctx_init_gpc_unk_1 },
 	{ gf110_grctx_init_setup_0 },
 	{ gf100_grctx_init_zcull_0 },
-	{ gf100_grctx_init_crstr_0 },
-	{ gf100_grctx_init_gpm_0 },
-	{ gf100_grctx_init_gcc_0 },
 	{}
 };
 
@@ -335,7 +332,8 @@ gf110_grctx = {
 	.main  = gf100_grctx_generate_main,
 	.unkn  = gf100_grctx_generate_unkn,
 	.hub   = gf100_grctx_pack_hub,
-	.gpc   = gf110_grctx_pack_gpc,
+	.gpc_0 = gf110_grctx_pack_gpc_0,
+	.gpc_1 = gf100_grctx_pack_gpc_1,
 	.zcull = gf100_grctx_pack_zcull,
 	.tpc   = gf100_grctx_pack_tpc,
 	.icmd  = gf110_grctx_pack_icmd,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
index bc4e86b..276c282 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
@@ -84,12 +84,17 @@ gf117_grctx_init_setup_0[] = {
 };
 
 static const struct gf100_gr_pack
-gf117_grctx_pack_gpc[] = {
+gf117_grctx_pack_gpc_0[] = {
 	{ gf100_grctx_init_gpc_unk_0 },
 	{ gf119_grctx_init_prop_0 },
 	{ gf119_grctx_init_gpc_unk_1 },
 	{ gf117_grctx_init_setup_0 },
 	{ gf100_grctx_init_zcull_0 },
+	{}
+};
+
+const struct gf100_gr_pack
+gf117_grctx_pack_gpc_1[] = {
 	{ gf119_grctx_init_crstr_0 },
 	{ gf108_grctx_init_gpm_0 },
 	{ gf100_grctx_init_gcc_0 },
@@ -278,7 +283,8 @@ gf117_grctx = {
 	.main  = gf100_grctx_generate_main,
 	.unkn  = gk104_grctx_generate_unkn,
 	.hub   = gf117_grctx_pack_hub,
-	.gpc   = gf117_grctx_pack_gpc,
+	.gpc_0 = gf117_grctx_pack_gpc_0,
+	.gpc_1 = gf117_grctx_pack_gpc_1,
 	.zcull = gf100_grctx_pack_zcull,
 	.tpc   = gf117_grctx_pack_tpc,
 	.ppc   = gf117_grctx_pack_ppc,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c
index cc1a935..0cfe463 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c
@@ -431,15 +431,12 @@ gf119_grctx_init_crstr_0[] = {
 };
 
 static const struct gf100_gr_pack
-gf119_grctx_pack_gpc[] = {
+gf119_grctx_pack_gpc_0[] = {
 	{ gf100_grctx_init_gpc_unk_0 },
 	{ gf119_grctx_init_prop_0 },
 	{ gf119_grctx_init_gpc_unk_1 },
 	{ gf119_grctx_init_setup_0 },
 	{ gf100_grctx_init_zcull_0 },
-	{ gf119_grctx_init_crstr_0 },
-	{ gf108_grctx_init_gpm_0 },
-	{ gf100_grctx_init_gcc_0 },
 	{}
 };
 
@@ -503,7 +500,8 @@ gf119_grctx = {
 	.main  = gf100_grctx_generate_main,
 	.unkn  = gf108_grctx_generate_unkn,
 	.hub   = gf119_grctx_pack_hub,
-	.gpc   = gf119_grctx_pack_gpc,
+	.gpc_0 = gf119_grctx_pack_gpc_0,
+	.gpc_1 = gf117_grctx_pack_gpc_1,
 	.zcull = gf100_grctx_pack_zcull,
 	.tpc   = gf119_grctx_pack_tpc,
 	.icmd  = gf119_grctx_pack_icmd,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
index bdf2a1e..304e9d2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
@@ -739,13 +739,18 @@ gk104_grctx_init_gpm_0[] = {
 	{}
 };
 
-const struct gf100_gr_pack
-gk104_grctx_pack_gpc[] = {
+static const struct gf100_gr_pack
+gk104_grctx_pack_gpc_0[] = {
 	{ gf100_grctx_init_gpc_unk_0 },
 	{ gf119_grctx_init_prop_0 },
 	{ gf119_grctx_init_gpc_unk_1 },
 	{ gk104_grctx_init_setup_0 },
 	{ gf100_grctx_init_zcull_0 },
+	{}
+};
+
+static const struct gf100_gr_pack
+gk104_grctx_pack_gpc_1[] = {
 	{ gf119_grctx_init_crstr_0 },
 	{ gk104_grctx_init_gpm_0 },
 	{ gf100_grctx_init_gcc_0 },
@@ -973,7 +978,8 @@ gk104_grctx = {
 	.main  = gf100_grctx_generate_main,
 	.unkn  = gk104_grctx_generate_unkn,
 	.hub   = gk104_grctx_pack_hub,
-	.gpc   = gk104_grctx_pack_gpc,
+	.gpc_0 = gk104_grctx_pack_gpc_0,
+	.gpc_1 = gk104_grctx_pack_gpc_1,
 	.zcull = gf100_grctx_pack_zcull,
 	.tpc   = gk104_grctx_pack_tpc,
 	.ppc   = gk104_grctx_pack_ppc,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
index 2ba35d7..86547cfc 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
@@ -704,12 +704,17 @@ gk110_grctx_init_gpc_unk_2[] = {
 };
 
 const struct gf100_gr_pack
-gk110_grctx_pack_gpc[] = {
+gk110_grctx_pack_gpc_0[] = {
 	{ gf100_grctx_init_gpc_unk_0 },
 	{ gf119_grctx_init_prop_0 },
 	{ gf119_grctx_init_gpc_unk_1 },
 	{ gk110_grctx_init_setup_0 },
 	{ gf100_grctx_init_zcull_0 },
+	{}
+};
+
+const struct gf100_gr_pack
+gk110_grctx_pack_gpc_1[] = {
 	{ gf119_grctx_init_crstr_0 },
 	{ gk104_grctx_init_gpm_0 },
 	{ gk110_grctx_init_gpc_unk_2 },
@@ -820,7 +825,8 @@ gk110_grctx = {
 	.main  = gf100_grctx_generate_main,
 	.unkn  = gk104_grctx_generate_unkn,
 	.hub   = gk110_grctx_pack_hub,
-	.gpc   = gk110_grctx_pack_gpc,
+	.gpc_0 = gk110_grctx_pack_gpc_0,
+	.gpc_1 = gk110_grctx_pack_gpc_1,
 	.zcull = gf100_grctx_pack_zcull,
 	.tpc   = gk110_grctx_pack_tpc,
 	.ppc   = gk110_grctx_pack_ppc,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
index 1112f8d..ebb947b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
@@ -74,7 +74,8 @@ gk110b_grctx = {
 	.main  = gf100_grctx_generate_main,
 	.unkn  = gk104_grctx_generate_unkn,
 	.hub   = gk110_grctx_pack_hub,
-	.gpc   = gk110_grctx_pack_gpc,
+	.gpc_0 = gk110_grctx_pack_gpc_0,
+	.gpc_1 = gk110_grctx_pack_gpc_1,
 	.zcull = gf100_grctx_pack_zcull,
 	.tpc   = gk110b_grctx_pack_tpc,
 	.ppc   = gk110_grctx_pack_ppc,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
index 613c5cf8..4d40512 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
@@ -443,12 +443,17 @@ gk208_grctx_init_gpm_0[] = {
 };
 
 static const struct gf100_gr_pack
-gk208_grctx_pack_gpc[] = {
+gk208_grctx_pack_gpc_0[] = {
 	{ gf100_grctx_init_gpc_unk_0 },
 	{ gk208_grctx_init_prop_0 },
 	{ gk208_grctx_init_gpc_unk_1 },
 	{ gk208_grctx_init_setup_0 },
 	{ gf100_grctx_init_zcull_0 },
+	{}
+};
+
+static const struct gf100_gr_pack
+gk208_grctx_pack_gpc_1[] = {
 	{ gk208_grctx_init_crstr_0 },
 	{ gk208_grctx_init_gpm_0 },
 	{ gk110_grctx_init_gpc_unk_2 },
@@ -535,7 +540,8 @@ gk208_grctx = {
 	.main  = gf100_grctx_generate_main,
 	.unkn  = gk104_grctx_generate_unkn,
 	.hub   = gk208_grctx_pack_hub,
-	.gpc   = gk208_grctx_pack_gpc,
+	.gpc_0 = gk208_grctx_pack_gpc_0,
+	.gpc_1 = gk208_grctx_pack_gpc_1,
 	.zcull = gf100_grctx_pack_zcull,
 	.tpc   = gk208_grctx_pack_tpc,
 	.ppc   = gk208_grctx_pack_ppc,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
index 7816dcb..0b3964e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
@@ -744,12 +744,17 @@ gm107_grctx_init_gpc_unk_2[] = {
 };
 
 static const struct gf100_gr_pack
-gm107_grctx_pack_gpc[] = {
+gm107_grctx_pack_gpc_0[] = {
 	{ gm107_grctx_init_gpc_unk_0 },
 	{ gk208_grctx_init_prop_0 },
 	{ gm107_grctx_init_gpc_unk_1 },
 	{ gm107_grctx_init_setup_0 },
 	{ gf100_grctx_init_zcull_0 },
+	{}
+};
+
+static const struct gf100_gr_pack
+gm107_grctx_pack_gpc_1[] = {
 	{ gk208_grctx_init_crstr_0 },
 	{ gk104_grctx_init_gpm_0 },
 	{ gm107_grctx_init_gpc_unk_2 },
@@ -960,7 +965,8 @@ gm107_grctx = {
 	.main  = gf100_grctx_generate_main,
 	.unkn  = gk104_grctx_generate_unkn,
 	.hub   = gm107_grctx_pack_hub,
-	.gpc   = gm107_grctx_pack_gpc,
+	.gpc_0 = gm107_grctx_pack_gpc_0,
+	.gpc_1 = gm107_grctx_pack_gpc_1,
 	.zcull = gf100_grctx_pack_zcull,
 	.tpc   = gm107_grctx_pack_tpc,
 	.ppc   = gm107_grctx_pack_ppc,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index b2070c8..6941861 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -1612,7 +1612,8 @@ gf100_gr_init_ctxctl_int(struct gf100_gr *gr)
 
 	/* load register lists */
 	gf100_gr_init_csdata(gr, grctx->hub, 0x409000, 0x000, 0x000000);
-	gf100_gr_init_csdata(gr, grctx->gpc, 0x41a000, 0x000, 0x418000);
+	gf100_gr_init_csdata(gr, grctx->gpc_0, 0x41a000, 0x000, 0x418000);
+	gf100_gr_init_csdata(gr, grctx->gpc_1, 0x41a000, 0x000, 0x418000);
 	gf100_gr_init_csdata(gr, grctx->tpc, 0x41a000, 0x004, 0x419800);
 	gf100_gr_init_csdata(gr, grctx->ppc, 0x41a000, 0x008, 0x41be00);
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index 55dedd8..9e608df 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -338,6 +338,7 @@ extern const struct gf100_gr_init gf117_gr_init_wwdx_0[];
 extern const struct gf100_gr_init gf117_gr_init_cbm_0[];
 
 extern const struct gf100_gr_init gk104_gr_init_main_0[];
+extern const struct gf100_gr_init gk104_gr_init_gpc_unk_2[];
 extern const struct gf100_gr_init gk104_gr_init_tpccs_0[];
 extern const struct gf100_gr_init gk104_gr_init_pe_0[];
 extern const struct gf100_gr_init gk104_gr_init_be_0[];
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
index cce250a..9abacb2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
@@ -83,6 +83,12 @@ gk104_gr_init_gpc_unk_1[] = {
 };
 
 const struct gf100_gr_init
+gk104_gr_init_gpc_unk_2[] = {
+	{ 0x418884,   1, 0x04, 0x00000000 },
+	{}
+};
+
+const struct gf100_gr_init
 gk104_gr_init_tpccs_0[] = {
 	{ 0x419d0c,   1, 0x04, 0x00000000 },
 	{ 0x419d10,   1, 0x04, 0x00000014 },
@@ -160,6 +166,7 @@ gk104_gr_pack_mmio[] = {
 	{ gf119_gr_init_gpm_0 },
 	{ gk104_gr_init_gpc_unk_1 },
 	{ gf100_gr_init_gcc_0 },
+	{ gk104_gr_init_gpc_unk_2 },
 	{ gk104_gr_init_tpccs_0 },
 	{ gf119_gr_init_tex_0 },
 	{ gk104_gr_init_pe_0 },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
index 558b497..dba70d5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
@@ -143,6 +143,7 @@ gk110_gr_pack_mmio[] = {
 	{ gf119_gr_init_gpm_0 },
 	{ gk110_gr_init_gpc_unk_1 },
 	{ gf100_gr_init_gcc_0 },
+	{ gk104_gr_init_gpc_unk_2 },
 	{ gk104_gr_init_tpccs_0 },
 	{ gk110_gr_init_tex_0 },
 	{ gk104_gr_init_pe_0 },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
index c8c48a2..48bc8d8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
@@ -82,6 +82,7 @@ gk110b_gr_pack_mmio[] = {
 	{ gf119_gr_init_gpm_0 },
 	{ gk110_gr_init_gpc_unk_1 },
 	{ gf100_gr_init_gcc_0 },
+	{ gk104_gr_init_gpc_unk_2 },
 	{ gk104_gr_init_tpccs_0 },
 	{ gk110_gr_init_tex_0 },
 	{ gk104_gr_init_pe_0 },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
index 477a7de..7f45b12 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
@@ -121,6 +121,7 @@ gk208_gr_pack_mmio[] = {
 	{ gf119_gr_init_gpm_0 },
 	{ gk110_gr_init_gpc_unk_1 },
 	{ gf100_gr_init_gcc_0 },
+	{ gk104_gr_init_gpc_unk_2 },
 	{ gk104_gr_init_tpccs_0 },
 	{ gk208_gr_init_tex_0 },
 	{ gk104_gr_init_pe_0 },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
index a6937f8..f5411af 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
@@ -38,6 +38,10 @@
 
 static const struct gf100_gr_init
 gm107_gr_init_main_0[] = {
+	{ 0x40880c,   1, 0x04, 0x00000000 },
+	{ 0x408910,   1, 0x04, 0x00000000 },
+	{ 0x408984,   1, 0x04, 0x00000000 },
+	{ 0x41a8a0,   1, 0x04, 0x00000000 },
 	{ 0x400080,   1, 0x04, 0x003003c2 },
 	{ 0x400088,   1, 0x04, 0x0001bfe7 },
 	{ 0x40008c,   1, 0x04, 0x00060000 },
@@ -212,14 +216,13 @@ gm107_gr_init_cbm_0[] = {
 static const struct gf100_gr_init
 gm107_gr_init_be_0[] = {
 	{ 0x408890,   1, 0x04, 0x000000ff },
-	{ 0x40880c,   1, 0x04, 0x00000000 },
 	{ 0x408850,   1, 0x04, 0x00000004 },
 	{ 0x408878,   1, 0x04, 0x00c81603 },
 	{ 0x40887c,   1, 0x04, 0x80543432 },
 	{ 0x408880,   1, 0x04, 0x0010581e },
 	{ 0x408884,   1, 0x04, 0x00001205 },
 	{ 0x408974,   1, 0x04, 0x000000ff },
-	{ 0x408910,   9, 0x04, 0x00000000 },
+	{ 0x408914,   8, 0x04, 0x00000000 },
 	{ 0x408950,   1, 0x04, 0x00000000 },
 	{ 0x408954,   1, 0x04, 0x0000ffff },
 	{ 0x408958,   1, 0x04, 0x00000034 },
@@ -229,7 +232,6 @@ gm107_gr_init_be_0[] = {
 	{ 0x408968,   1, 0x04, 0x02808833 },
 	{ 0x40896c,   1, 0x04, 0x01f02438 },
 	{ 0x408970,   1, 0x04, 0x00012c00 },
-	{ 0x408984,   1, 0x04, 0x00000000 },
 	{ 0x408988,   1, 0x04, 0x08040201 },
 	{ 0x40898c,   1, 0x04, 0x80402010 },
 	{}
@@ -262,6 +264,7 @@ gm107_gr_pack_mmio[] = {
 	{ gf100_gr_init_gpm_0 },
 	{ gm107_gr_init_gpc_unk_1 },
 	{ gf100_gr_init_gcc_0 },
+	{ gk104_gr_init_gpc_unk_2 },
 	{ gm107_gr_init_tpccs_0 },
 	{ gm107_gr_init_tex_0 },
 	{ gm107_gr_init_pe_0 },
-- 
cgit v1.1


From 17f2d4df32104b327178771fb323391fbf53921c Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/gr/gp100-: fix pagepool setup

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
index 2344fd8..c51e3a2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
@@ -36,7 +36,7 @@ gp100_grctx_generate_pagepool(struct gf100_grctx *info)
 	const int s = 8;
 	const int b = mmio_vram(info, grctx->pagepool_size, (1 << s), true);
 	mmio_refn(info, 0x40800c, 0x00000000, s, b);
-	mmio_wr32(info, 0x408010, 0x80000000);
+	mmio_wr32(info, 0x408010, 0x8007d800);
 	mmio_refn(info, 0x419004, 0x00000000, s, b);
 	mmio_wr32(info, 0x419008, 0x00000000);
 }
-- 
cgit v1.1


From 7a058a900ccb010c32ca1f29f6f9728a3654a519 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/gr/gp100-: fix attrib cb setup

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c | 17 +++++++++-------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c | 24 ++++++++++++++---------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp104.c |  3 ++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c    |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h    |  1 +
 7 files changed, 32 insertions(+), 17 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
index 1c06c67..9ce3d00 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
@@ -49,6 +49,7 @@ struct gf100_grctx_func {
 	u32 attrib_nr;
 	u32 alpha_nr_max;
 	u32 alpha_nr;
+	u32 gfxp_nr;
 	/* other patch buffer stuff */
 	void (*patch_ltc)(struct gf100_grctx *);
 	/* floorsweeping */
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
index c51e3a2..0b33262 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
@@ -48,14 +48,17 @@ gp100_grctx_generate_attrib(struct gf100_grctx *info)
 	const struct gf100_grctx_func *grctx = gr->func->grctx;
 	const u32  alpha = grctx->alpha_nr;
 	const u32 attrib = grctx->attrib_nr;
-	const u32 pertpc = 0x20 * (grctx->attrib_nr_max + grctx->alpha_nr_max);
-	const u32   size = roundup(gr->tpc_total * pertpc, 0x80);
 	const int s = 12;
-	const int b = mmio_vram(info, size, (1 << s), false);
 	const int max_batches = 0xffff;
+	u32 size = grctx->alpha_nr_max * gr->tpc_total;
 	u32 ao = 0;
-	u32 bo = ao + grctx->alpha_nr_max * gr->tpc_total;
-	int gpc, ppc, n = 0;
+	u32 bo = ao + size;
+	int gpc, ppc, b, n = 0;
+
+	for (gpc = 0; gpc < gr->gpc_nr; gpc++)
+		size += grctx->attrib_nr_max * gr->ppc_nr[gpc] * gr->ppc_tpc_max;
+	size = ((size * 0x20) + 128) & ~127;
+	b = mmio_vram(info, size, (1 << s), false);
 
 	mmio_refn(info, 0x418810, 0x80000000, s, b);
 	mmio_refn(info, 0x419848, 0x10000000, s, b);
@@ -69,7 +72,7 @@ gp100_grctx_generate_attrib(struct gf100_grctx *info)
 	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
 		for (ppc = 0; ppc < gr->ppc_nr[gpc]; ppc++, n++) {
 			const u32 as =  alpha * gr->ppc_tpc_nr[gpc][ppc];
-			const u32 bs = attrib * gr->ppc_tpc_nr[gpc][ppc];
+			const u32 bs = attrib * gr->ppc_tpc_max;
 			const u32 u = 0x418ea0 + (n * 0x04);
 			const u32 o = PPC_UNIT(gpc, ppc, 0);
 			if (!(gr->ppc_mask[gpc] & (1 << ppc)))
@@ -77,7 +80,7 @@ gp100_grctx_generate_attrib(struct gf100_grctx *info)
 			mmio_wr32(info, o + 0xc0, bs);
 			mmio_wr32(info, o + 0xf4, bo);
 			mmio_wr32(info, o + 0xf0, bs);
-			bo += grctx->attrib_nr_max * gr->ppc_tpc_nr[gpc][ppc];
+			bo += grctx->attrib_nr_max * gr->ppc_tpc_max;
 			mmio_wr32(info, o + 0xe4, as);
 			mmio_wr32(info, o + 0xf8, ao);
 			ao += grctx->alpha_nr_max * gr->ppc_tpc_nr[gpc][ppc];
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
index 8a438c2..daee17b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
@@ -43,14 +43,18 @@ gp102_grctx_generate_attrib(struct gf100_grctx *info)
 	const struct gf100_grctx_func *grctx = gr->func->grctx;
 	const u32  alpha = grctx->alpha_nr;
 	const u32 attrib = grctx->attrib_nr;
-	const u32 pertpc = 0x20 * (grctx->attrib_nr_max + grctx->alpha_nr_max);
-	const u32   size = roundup(gr->tpc_total * pertpc, 0x80);
+	const u32   gfxp = grctx->gfxp_nr;
 	const int s = 12;
-	const int b = mmio_vram(info, size, (1 << s), false);
 	const int max_batches = 0xffff;
+	u32 size = grctx->alpha_nr_max * gr->tpc_total;
 	u32 ao = 0;
-	u32 bo = ao + grctx->alpha_nr_max * gr->tpc_total;
-	int gpc, ppc, n = 0;
+	u32 bo = ao + size;
+	int gpc, ppc, b, n = 0;
+
+	for (gpc = 0; gpc < gr->gpc_nr; gpc++)
+		size += grctx->gfxp_nr * gr->ppc_nr[gpc] * gr->ppc_tpc_max;
+	size = ((size * 0x20) + 128) & ~127;
+	b = mmio_vram(info, size, (1 << s), false);
 
 	mmio_refn(info, 0x418810, 0x80000000, s, b);
 	mmio_refn(info, 0x419848, 0x10000000, s, b);
@@ -64,17 +68,18 @@ gp102_grctx_generate_attrib(struct gf100_grctx *info)
 	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
 		for (ppc = 0; ppc < gr->ppc_nr[gpc]; ppc++, n++) {
 			const u32 as =  alpha * gr->ppc_tpc_nr[gpc][ppc];
-			const u32 bs = attrib * gr->ppc_tpc_nr[gpc][ppc];
+			const u32 bs = attrib * gr->ppc_tpc_max;
+			const u32 gs =   gfxp * gr->ppc_tpc_max;
 			const u32 u = 0x418ea0 + (n * 0x04);
 			const u32 o = PPC_UNIT(gpc, ppc, 0);
 			const u32 p = GPC_UNIT(gpc, 0xc44 + (ppc * 4));
 			if (!(gr->ppc_mask[gpc] & (1 << ppc)))
 				continue;
-			mmio_wr32(info, o + 0xc0, bs);
+			mmio_wr32(info, o + 0xc0, gs);
 			mmio_wr32(info, p, bs);
 			mmio_wr32(info, o + 0xf4, bo);
 			mmio_wr32(info, o + 0xf0, bs);
-			bo += grctx->attrib_nr_max * gr->ppc_tpc_nr[gpc][ppc];
+			bo += gs;
 			mmio_wr32(info, o + 0xe4, as);
 			mmio_wr32(info, o + 0xf8, ao);
 			ao += grctx->alpha_nr_max * gr->ppc_tpc_nr[gpc][ppc];
@@ -97,10 +102,11 @@ gp102_grctx = {
 	.pagepool = gp100_grctx_generate_pagepool,
 	.pagepool_size = 0x20000,
 	.attrib = gp102_grctx_generate_attrib,
-	.attrib_nr_max = 0x5d4,
+	.attrib_nr_max = 0x4b0,
 	.attrib_nr = 0x320,
 	.alpha_nr_max = 0xc00,
 	.alpha_nr = 0x800,
+	.gfxp_nr = 0xba8,
 	.sm_id = gm107_grctx_generate_sm_id,
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
 	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp104.c
index 5f799c7..3b85e3d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp104.c
@@ -32,10 +32,11 @@ gp104_grctx = {
 	.pagepool = gp100_grctx_generate_pagepool,
 	.pagepool_size = 0x20000,
 	.attrib = gp102_grctx_generate_attrib,
-	.attrib_nr_max = 0x5d4,
+	.attrib_nr_max = 0x4b0,
 	.attrib_nr = 0x320,
 	.alpha_nr_max = 0xc00,
 	.alpha_nr = 0x800,
+	.gfxp_nr = 0xba8,
 	.sm_id = gm107_grctx_generate_sm_id,
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
 	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
index a69e824..5060c5e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
@@ -44,6 +44,7 @@ gp107_grctx = {
 	.attrib_nr = 0x540,
 	.alpha_nr_max = 0xc00,
 	.alpha_nr = 0x800,
+	.gfxp_nr = 0xe94,
 	.sm_id = gm107_grctx_generate_sm_id,
 	.rop_mapping = gf117_grctx_generate_rop_mapping,
 	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 6941861..084a5d1 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -1780,6 +1780,8 @@ gf100_gr_oneinit(struct nvkm_gr *base)
 			if (gr->ppc_tpc_min == 0 ||
 			    gr->ppc_tpc_min > gr->ppc_tpc_nr[i][j])
 				gr->ppc_tpc_min = gr->ppc_tpc_nr[i][j];
+			if (gr->ppc_tpc_max < gr->ppc_tpc_nr[i][j])
+				gr->ppc_tpc_max = gr->ppc_tpc_nr[i][j];
 		}
 	}
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index 9e608df..390bcc1 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -106,6 +106,7 @@ struct gf100_gr {
 	u8 ppc_tpc_mask[GPC_MAX][4];
 	u8 ppc_tpc_nr[GPC_MAX][4];
 	u8 ppc_tpc_min;
+	u8 ppc_tpc_max;
 
 	u8 screen_tile_row_offset;
 	u8 tile[TPC_MAX];
-- 
cgit v1.1


From e9d03335f604a1123b8de3103ce8e06db4ad777a Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/gr/gp100-: use correct registers for zbc colour/depth
 setup

These were missed the first time around due to the driver version I traced
using the older registers still.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c  | 37 +++++++++++++----------
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h  | 10 +++++++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c  | 39 +++++++++++++++++++++++++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c  |  1 +
 20 files changed, 88 insertions(+), 15 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 084a5d1..0cffafb 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -92,7 +92,7 @@ gf100_gr_zbc_color_get(struct gf100_gr *gr, int format,
 	memcpy(gr->zbc_color[zbc].l2, l2, sizeof(gr->zbc_color[zbc].l2));
 	gr->zbc_color[zbc].format = format;
 	nvkm_ltc_zbc_color_get(ltc, zbc, l2);
-	gf100_gr_zbc_clear_color(gr, zbc);
+	gr->func->zbc->clear_color(gr, zbc);
 	return zbc;
 }
 
@@ -137,10 +137,16 @@ gf100_gr_zbc_depth_get(struct gf100_gr *gr, int format,
 	gr->zbc_depth[zbc].ds = ds;
 	gr->zbc_depth[zbc].l2 = l2;
 	nvkm_ltc_zbc_depth_get(ltc, zbc, l2);
-	gf100_gr_zbc_clear_depth(gr, zbc);
+	gr->func->zbc->clear_depth(gr, zbc);
 	return zbc;
 }
 
+const struct gf100_gr_func_zbc
+gf100_gr_zbc = {
+	.clear_color = gf100_gr_zbc_clear_color,
+	.clear_depth = gf100_gr_zbc_clear_depth,
+};
+
 /*******************************************************************************
  * Graphics object classes
  ******************************************************************************/
@@ -744,21 +750,21 @@ gf100_gr_zbc_init(struct gf100_gr *gr)
 	const u32 f32_1[] = { 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000,
 			      0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000 };
 	struct nvkm_ltc *ltc = gr->base.engine.subdev.device->ltc;
-	int index;
+	int index, c = ltc->zbc_min, d = ltc->zbc_min;
 
 	if (!gr->zbc_color[0].format) {
-		gf100_gr_zbc_color_get(gr, 1,  & zero[0],   &zero[4]);
-		gf100_gr_zbc_color_get(gr, 2,  &  one[0],    &one[4]);
-		gf100_gr_zbc_color_get(gr, 4,  &f32_0[0],  &f32_0[4]);
-		gf100_gr_zbc_color_get(gr, 4,  &f32_1[0],  &f32_1[4]);
-		gf100_gr_zbc_depth_get(gr, 1, 0x00000000, 0x00000000);
-		gf100_gr_zbc_depth_get(gr, 1, 0x3f800000, 0x3f800000);
-	}
-
-	for (index = ltc->zbc_min; index <= ltc->zbc_max; index++)
-		gf100_gr_zbc_clear_color(gr, index);
-	for (index = ltc->zbc_min; index <= ltc->zbc_max; index++)
-		gf100_gr_zbc_clear_depth(gr, index);
+		gf100_gr_zbc_color_get(gr, 1,  & zero[0],   &zero[4]); c++;
+		gf100_gr_zbc_color_get(gr, 2,  &  one[0],    &one[4]); c++;
+		gf100_gr_zbc_color_get(gr, 4,  &f32_0[0],  &f32_0[4]); c++;
+		gf100_gr_zbc_color_get(gr, 4,  &f32_1[0],  &f32_1[4]); c++;
+		gf100_gr_zbc_depth_get(gr, 1, 0x00000000, 0x00000000); d++;
+		gf100_gr_zbc_depth_get(gr, 1, 0x3f800000, 0x3f800000); d++;
+	}
+
+	for (index = c; index <= ltc->zbc_max; index++)
+		gr->func->zbc->clear_color(gr, index);
+	for (index = d; index <= ltc->zbc_max; index++)
+		gr->func->zbc->clear_depth(gr, index);
 }
 
 /**
@@ -2242,6 +2248,7 @@ gf100_gr = {
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
 	.rops = gf100_gr_rops,
 	.grctx = &gf100_grctx,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, FERMI_MEMORY_TO_MEMORY_FORMAT_A },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index 390bcc1..d82951a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -129,6 +129,11 @@ int gf100_gr_new_(const struct gf100_gr_func *, struct nvkm_device *,
 		  int, struct nvkm_gr **);
 void *gf100_gr_dtor(struct nvkm_gr *);
 
+struct gf100_gr_func_zbc {
+	void (*clear_color)(struct gf100_gr *, int zbc);
+	void (*clear_depth)(struct gf100_gr *, int zbc);
+};
+
 struct gf100_gr_func {
 	void (*dtor)(struct gf100_gr *);
 	void (*oneinit_tiles)(struct gf100_gr *);
@@ -170,6 +175,7 @@ struct gf100_gr_func {
 	int ppc_nr;
 	const struct gf100_grctx_func *grctx;
 	const struct nvkm_therm_clkgate_pack *clkgate_pack;
+	const struct gf100_gr_func_zbc *zbc;
 	struct nvkm_sclass sclass[];
 };
 
@@ -187,6 +193,7 @@ void gf100_gr_init_419eb4(struct gf100_gr *);
 void gf100_gr_init_tex_hww_esr(struct gf100_gr *, int, int);
 void gf100_gr_init_shader_exceptions(struct gf100_gr *, int, int);
 void gf100_gr_init_400054(struct gf100_gr *);
+extern const struct gf100_gr_func_zbc gf100_gr_zbc;
 
 void gf117_gr_init_zcull(struct gf100_gr *);
 
@@ -212,6 +219,9 @@ void gm200_gr_init_ds_hww_esr_2(struct gf100_gr *);
 void gp100_gr_init_rop_active_fbps(struct gf100_gr *);
 void gp100_gr_init_fecs_exceptions(struct gf100_gr *);
 void gp100_gr_init_shader_exceptions(struct gf100_gr *, int, int);
+extern const struct gf100_gr_func_zbc gp100_gr_zbc;
+void gp100_gr_zbc_clear_color(struct gf100_gr *, int);
+void gp100_gr_zbc_clear_depth(struct gf100_gr *, int);
 
 void gp102_gr_init_swdx_pes_mask(struct gf100_gr *);
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
index 1d8e16a..42c2fd9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
@@ -134,6 +134,7 @@ gf104_gr = {
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
 	.rops = gf100_gr_rops,
 	.grctx = &gf104_grctx,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, FERMI_MEMORY_TO_MEMORY_FORMAT_A },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
index a5a74df..4731a46 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
@@ -132,6 +132,7 @@ gf108_gr = {
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
 	.rops = gf100_gr_rops,
 	.grctx = &gf108_grctx,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, FERMI_MEMORY_TO_MEMORY_FORMAT_A },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
index 45fada0..cdf759c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
@@ -106,6 +106,7 @@ gf110_gr = {
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
 	.rops = gf100_gr_rops,
 	.grctx = &gf110_grctx,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, FERMI_MEMORY_TO_MEMORY_FORMAT_A },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
index 5ee167d..a4158f8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
@@ -171,6 +171,7 @@ gf117_gr = {
 	.rops = gf100_gr_rops,
 	.ppc_nr = 1,
 	.grctx = &gf117_grctx,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, FERMI_MEMORY_TO_MEMORY_FORMAT_A },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
index 2096552..4197844 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
@@ -197,6 +197,7 @@ gf119_gr = {
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
 	.rops = gf100_gr_rops,
 	.grctx = &gf119_grctx,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, FERMI_MEMORY_TO_MEMORY_FORMAT_A },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
index 9abacb2..477fee3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
@@ -479,6 +479,7 @@ gk104_gr = {
 	.ppc_nr = 1,
 	.grctx = &gk104_grctx,
 	.clkgate_pack = gk104_clkgate_pack,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_A },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
index dba70d5..7cd628c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
@@ -375,6 +375,7 @@ gk110_gr = {
 	.ppc_nr = 2,
 	.grctx = &gk110_grctx,
 	.clkgate_pack = gk110_clkgate_pack,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
index 48bc8d8..a38faa2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
@@ -126,6 +126,7 @@ gk110b_gr = {
 	.rops = gf100_gr_rops,
 	.ppc_nr = 2,
 	.grctx = &gk110b_grctx,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
index 7f45b12..5845666 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
@@ -184,6 +184,7 @@ gk208_gr = {
 	.rops = gf100_gr_rops,
 	.ppc_nr = 1,
 	.grctx = &gk208_grctx,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c
index 11a32fa..500cb08 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c
@@ -292,6 +292,7 @@ gk20a_gr = {
 	.rops = gf100_gr_rops,
 	.ppc_nr = 1,
 	.grctx = &gk20a_grctx,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_A },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
index f5411af..92e31d3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
@@ -419,6 +419,7 @@ gm107_gr = {
 	.rops = gf100_gr_rops,
 	.ppc_nr = 2,
 	.grctx = &gm107_grctx,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
index 8966d2a..eff3066 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
@@ -187,6 +187,7 @@ gm200_gr = {
 	.tpc_nr = 4,
 	.ppc_nr = 2,
 	.grctx = &gm200_grctx,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c
index afa1c6e..a667770 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c
@@ -75,6 +75,7 @@ gm20b_gr = {
 	.rops = gm200_gr_rops,
 	.ppc_nr = 1,
 	.grctx = &gm20b_grctx,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
index 1d9d876..ef16fee6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
@@ -29,6 +29,44 @@
 /*******************************************************************************
  * PGRAPH engine/subdev functions
  ******************************************************************************/
+void
+gp100_gr_zbc_clear_color(struct gf100_gr *gr, int zbc)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	const int znum =  zbc - 1;
+	const u32 zoff = znum * 4;
+
+	if (gr->zbc_color[zbc].format) {
+		nvkm_wr32(device, 0x418010 + zoff, gr->zbc_color[zbc].ds[0]);
+		nvkm_wr32(device, 0x41804c + zoff, gr->zbc_color[zbc].ds[1]);
+		nvkm_wr32(device, 0x418088 + zoff, gr->zbc_color[zbc].ds[2]);
+		nvkm_wr32(device, 0x4180c4 + zoff, gr->zbc_color[zbc].ds[3]);
+	}
+
+	nvkm_mask(device, 0x418100 + ((znum / 4) * 4),
+			  0x0000007f << ((znum % 4) * 7),
+			  gr->zbc_color[zbc].format << ((znum % 4) * 7));
+}
+
+void
+gp100_gr_zbc_clear_depth(struct gf100_gr *gr, int zbc)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	const int znum =  zbc - 1;
+	const u32 zoff = znum * 4;
+
+	if (gr->zbc_depth[zbc].format)
+		nvkm_wr32(device, 0x418110 + zoff, gr->zbc_depth[zbc].ds);
+	nvkm_mask(device, 0x41814c + ((znum / 4) * 4),
+			  0x0000007f << ((znum % 4) * 7),
+			  gr->zbc_depth[zbc].format << ((znum % 4) * 7));
+}
+
+const struct gf100_gr_func_zbc
+gp100_gr_zbc = {
+	.clear_color = gp100_gr_zbc_clear_color,
+	.clear_depth = gp100_gr_zbc_clear_depth,
+};
 
 void
 gp100_gr_init_shader_exceptions(struct gf100_gr *gr, int gpc, int tpc)
@@ -87,6 +125,7 @@ gp100_gr = {
 	.tpc_nr = 5,
 	.ppc_nr = 2,
 	.grctx = &gp100_grctx,
+	.zbc = &gp100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
index 2d9a2c3..8fc95a0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
@@ -65,6 +65,7 @@ gp102_gr = {
 	.tpc_nr = 5,
 	.ppc_nr = 3,
 	.grctx = &gp102_grctx,
+	.zbc = &gp100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c
index e466ae4..56f92c7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c
@@ -49,6 +49,7 @@ gp104_gr = {
 	.tpc_nr = 5,
 	.ppc_nr = 3,
 	.grctx = &gp104_grctx,
+	.zbc = &gp100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
index 2fa046a..4bb0340 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
@@ -51,6 +51,7 @@ gp107_gr = {
 	.tpc_nr = 3,
 	.ppc_nr = 1,
 	.grctx = &gp107_grctx,
+	.zbc = &gp100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
index 0a01a30..b2a39ad 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
@@ -49,6 +49,7 @@ gp10b_gr = {
 	.tpc_nr = 2,
 	.ppc_nr = 1,
 	.grctx = &gp102_grctx,
+	.zbc = &gp100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
-- 
cgit v1.1


From 4b2c71edf0d7832ef4d2fe5b17402d1130b415dc Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/gr/gp102-: setup stencil zbc

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/include/nvkm/subdev/ltc.h |  3 ++
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c | 12 ++---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c    | 12 ++++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h    | 12 ++++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c    |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c    | 58 ++++++++++++++++++++++-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c    |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c    |  2 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c    |  2 +-
 drivers/gpu/drm/nouveau/nvkm/subdev/ltc/Kbuild    |  1 +
 drivers/gpu/drm/nouveau/nvkm/subdev/ltc/base.c    | 10 ++++
 drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gp100.c   |  6 +--
 drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gp102.c   | 51 ++++++++++++++++++++
 drivers/gpu/drm/nouveau/nvkm/subdev/ltc/priv.h    |  5 ++
 14 files changed, 162 insertions(+), 16 deletions(-)
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gp102.c

diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/ltc.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/ltc.h
index 95b6115..9db5f82 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/ltc.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/ltc.h
@@ -21,12 +21,14 @@ struct nvkm_ltc {
 	int zbc_max;
 	u32 zbc_color[NVKM_LTC_MAX_ZBC_CNT][4];
 	u32 zbc_depth[NVKM_LTC_MAX_ZBC_CNT];
+	u32 zbc_stencil[NVKM_LTC_MAX_ZBC_CNT];
 };
 
 void nvkm_ltc_tags_clear(struct nvkm_device *, u32 first, u32 count);
 
 int nvkm_ltc_zbc_color_get(struct nvkm_ltc *, int index, const u32[4]);
 int nvkm_ltc_zbc_depth_get(struct nvkm_ltc *, int index, const u32);
+int nvkm_ltc_zbc_stencil_get(struct nvkm_ltc *, int index, const u32);
 
 void nvkm_ltc_invalidate(struct nvkm_ltc *);
 void nvkm_ltc_flush(struct nvkm_ltc *);
@@ -37,4 +39,5 @@ int gk20a_ltc_new(struct nvkm_device *, int, struct nvkm_ltc **);
 int gm107_ltc_new(struct nvkm_device *, int, struct nvkm_ltc **);
 int gm200_ltc_new(struct nvkm_device *, int, struct nvkm_ltc **);
 int gp100_ltc_new(struct nvkm_device *, int, struct nvkm_ltc **);
+int gp102_ltc_new(struct nvkm_device *, int, struct nvkm_ltc **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 16e8090..5c79c79 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2204,7 +2204,7 @@ nv132_chipset = {
 	.i2c = gm200_i2c_new,
 	.ibus = gm200_ibus_new,
 	.imem = nv50_instmem_new,
-	.ltc = gp100_ltc_new,
+	.ltc = gp102_ltc_new,
 	.mc = gp100_mc_new,
 	.mmu = gp100_mmu_new,
 	.therm = gp100_therm_new,
@@ -2240,7 +2240,7 @@ nv134_chipset = {
 	.i2c = gm200_i2c_new,
 	.ibus = gm200_ibus_new,
 	.imem = nv50_instmem_new,
-	.ltc = gp100_ltc_new,
+	.ltc = gp102_ltc_new,
 	.mc = gp100_mc_new,
 	.mmu = gp100_mmu_new,
 	.therm = gp100_therm_new,
@@ -2276,7 +2276,7 @@ nv136_chipset = {
 	.i2c = gm200_i2c_new,
 	.ibus = gm200_ibus_new,
 	.imem = nv50_instmem_new,
-	.ltc = gp100_ltc_new,
+	.ltc = gp102_ltc_new,
 	.mc = gp100_mc_new,
 	.mmu = gp100_mmu_new,
 	.therm = gp100_therm_new,
@@ -2312,7 +2312,7 @@ nv137_chipset = {
 	.i2c = gm200_i2c_new,
 	.ibus = gm200_ibus_new,
 	.imem = nv50_instmem_new,
-	.ltc = gp100_ltc_new,
+	.ltc = gp102_ltc_new,
 	.mc = gp100_mc_new,
 	.mmu = gp100_mmu_new,
 	.therm = gp100_therm_new,
@@ -2348,7 +2348,7 @@ nv138_chipset = {
 	.i2c = gm200_i2c_new,
 	.ibus = gm200_ibus_new,
 	.imem = nv50_instmem_new,
-	.ltc = gp100_ltc_new,
+	.ltc = gp102_ltc_new,
 	.mc = gp100_mc_new,
 	.mmu = gp100_mmu_new,
 	.therm = gp100_therm_new,
@@ -2380,7 +2380,7 @@ nv13b_chipset = {
 	.fuse = gm107_fuse_new,
 	.ibus = gp10b_ibus_new,
 	.imem = gk20a_instmem_new,
-	.ltc = gp100_ltc_new,
+	.ltc = gp102_ltc_new,
 	.mc = gp10b_mc_new,
 	.mmu = gp10b_mmu_new,
 	.secboot = gp10b_secboot_new,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 0cffafb..86ae5c7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -750,7 +750,7 @@ gf100_gr_zbc_init(struct gf100_gr *gr)
 	const u32 f32_1[] = { 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000,
 			      0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000 };
 	struct nvkm_ltc *ltc = gr->base.engine.subdev.device->ltc;
-	int index, c = ltc->zbc_min, d = ltc->zbc_min;
+	int index, c = ltc->zbc_min, d = ltc->zbc_min, s = ltc->zbc_min;
 
 	if (!gr->zbc_color[0].format) {
 		gf100_gr_zbc_color_get(gr, 1,  & zero[0],   &zero[4]); c++;
@@ -759,12 +759,22 @@ gf100_gr_zbc_init(struct gf100_gr *gr)
 		gf100_gr_zbc_color_get(gr, 4,  &f32_1[0],  &f32_1[4]); c++;
 		gf100_gr_zbc_depth_get(gr, 1, 0x00000000, 0x00000000); d++;
 		gf100_gr_zbc_depth_get(gr, 1, 0x3f800000, 0x3f800000); d++;
+		if (gr->func->zbc->stencil_get) {
+			gr->func->zbc->stencil_get(gr, 1, 0x00, 0x00); s++;
+			gr->func->zbc->stencil_get(gr, 1, 0x01, 0x01); s++;
+			gr->func->zbc->stencil_get(gr, 1, 0xff, 0xff); s++;
+		}
 	}
 
 	for (index = c; index <= ltc->zbc_max; index++)
 		gr->func->zbc->clear_color(gr, index);
 	for (index = d; index <= ltc->zbc_max; index++)
 		gr->func->zbc->clear_depth(gr, index);
+
+	if (gr->func->zbc->clear_stencil) {
+		for (index = s; index <= ltc->zbc_max; index++)
+			gr->func->zbc->clear_stencil(gr, index);
+	}
 }
 
 /**
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index d82951a..edf6eda 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -72,6 +72,12 @@ struct gf100_gr_zbc_depth {
 	u32 l2;
 };
 
+struct gf100_gr_zbc_stencil {
+	u32 format;
+	u32 ds;
+	u32 l2;
+};
+
 struct gf100_gr {
 	const struct gf100_gr_func *func;
 	struct nvkm_gr base;
@@ -95,6 +101,7 @@ struct gf100_gr {
 
 	struct gf100_gr_zbc_color zbc_color[NVKM_LTC_MAX_ZBC_CNT];
 	struct gf100_gr_zbc_depth zbc_depth[NVKM_LTC_MAX_ZBC_CNT];
+	struct gf100_gr_zbc_stencil zbc_stencil[NVKM_LTC_MAX_ZBC_CNT];
 
 	u8 rop_nr;
 	u8 gpc_nr;
@@ -132,6 +139,9 @@ void *gf100_gr_dtor(struct nvkm_gr *);
 struct gf100_gr_func_zbc {
 	void (*clear_color)(struct gf100_gr *, int zbc);
 	void (*clear_depth)(struct gf100_gr *, int zbc);
+	int (*stencil_get)(struct gf100_gr *, int format,
+			   const u32 ds, const u32 l2);
+	void (*clear_stencil)(struct gf100_gr *, int zbc);
 };
 
 struct gf100_gr_func {
@@ -219,11 +229,11 @@ void gm200_gr_init_ds_hww_esr_2(struct gf100_gr *);
 void gp100_gr_init_rop_active_fbps(struct gf100_gr *);
 void gp100_gr_init_fecs_exceptions(struct gf100_gr *);
 void gp100_gr_init_shader_exceptions(struct gf100_gr *, int, int);
-extern const struct gf100_gr_func_zbc gp100_gr_zbc;
 void gp100_gr_zbc_clear_color(struct gf100_gr *, int);
 void gp100_gr_zbc_clear_depth(struct gf100_gr *, int);
 
 void gp102_gr_init_swdx_pes_mask(struct gf100_gr *);
+extern const struct gf100_gr_func_zbc gp102_gr_zbc;
 
 #define gf100_gr_chan(p) container_of((p), struct gf100_gr_chan, object)
 #include <core/object.h>
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
index ef16fee6..9d0521c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
@@ -62,7 +62,7 @@ gp100_gr_zbc_clear_depth(struct gf100_gr *gr, int zbc)
 			  gr->zbc_depth[zbc].format << ((znum % 4) * 7));
 }
 
-const struct gf100_gr_func_zbc
+static const struct gf100_gr_func_zbc
 gp100_gr_zbc = {
 	.clear_color = gp100_gr_zbc_clear_color,
 	.clear_depth = gp100_gr_zbc_clear_depth,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
index 8fc95a0..37f7d73 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
@@ -26,6 +26,62 @@
 
 #include <nvif/class.h>
 
+static void
+gp102_gr_zbc_clear_stencil(struct gf100_gr *gr, int zbc)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	const int znum =  zbc - 1;
+	const u32 zoff = znum * 4;
+
+	if (gr->zbc_stencil[zbc].format)
+		nvkm_wr32(device, 0x41815c + zoff, gr->zbc_stencil[zbc].ds);
+	nvkm_mask(device, 0x418198 + ((znum / 4) * 4),
+			  0x0000007f << ((znum % 4) * 7),
+			  gr->zbc_stencil[zbc].format << ((znum % 4) * 7));
+}
+
+static int
+gp102_gr_zbc_stencil_get(struct gf100_gr *gr, int format,
+			 const u32 ds, const u32 l2)
+{
+	struct nvkm_ltc *ltc = gr->base.engine.subdev.device->ltc;
+	int zbc = -ENOSPC, i;
+
+	for (i = ltc->zbc_min; i <= ltc->zbc_max; i++) {
+		if (gr->zbc_stencil[i].format) {
+			if (gr->zbc_stencil[i].format != format)
+				continue;
+			if (gr->zbc_stencil[i].ds != ds)
+				continue;
+			if (gr->zbc_stencil[i].l2 != l2) {
+				WARN_ON(1);
+				return -EINVAL;
+			}
+			return i;
+		} else {
+			zbc = (zbc < 0) ? i : zbc;
+		}
+	}
+
+	if (zbc < 0)
+		return zbc;
+
+	gr->zbc_stencil[zbc].format = format;
+	gr->zbc_stencil[zbc].ds = ds;
+	gr->zbc_stencil[zbc].l2 = l2;
+	nvkm_ltc_zbc_stencil_get(ltc, zbc, l2);
+	gr->func->zbc->clear_stencil(gr, zbc);
+	return zbc;
+}
+
+const struct gf100_gr_func_zbc
+gp102_gr_zbc = {
+	.clear_color = gp100_gr_zbc_clear_color,
+	.clear_depth = gp100_gr_zbc_clear_depth,
+	.stencil_get = gp102_gr_zbc_stencil_get,
+	.clear_stencil = gp102_gr_zbc_clear_stencil,
+};
+
 void
 gp102_gr_init_swdx_pes_mask(struct gf100_gr *gr)
 {
@@ -65,7 +121,7 @@ gp102_gr = {
 	.tpc_nr = 5,
 	.ppc_nr = 3,
 	.grctx = &gp102_grctx,
-	.zbc = &gp100_gr_zbc,
+	.zbc = &gp102_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c
index 56f92c7..4573c91 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c
@@ -49,7 +49,7 @@ gp104_gr = {
 	.tpc_nr = 5,
 	.ppc_nr = 3,
 	.grctx = &gp104_grctx,
-	.zbc = &gp100_gr_zbc,
+	.zbc = &gp102_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
index 4bb0340..812aba9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
@@ -51,7 +51,7 @@ gp107_gr = {
 	.tpc_nr = 3,
 	.ppc_nr = 1,
 	.grctx = &gp107_grctx,
-	.zbc = &gp100_gr_zbc,
+	.zbc = &gp102_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
index b2a39ad..303dced 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
@@ -49,7 +49,7 @@ gp10b_gr = {
 	.tpc_nr = 2,
 	.ppc_nr = 1,
 	.grctx = &gp102_grctx,
-	.zbc = &gp100_gr_zbc,
+	.zbc = &gp102_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/Kbuild
index 12d6f4f..290ff1c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/Kbuild
@@ -4,3 +4,4 @@ nvkm-y += nvkm/subdev/ltc/gk104.o
 nvkm-y += nvkm/subdev/ltc/gm107.o
 nvkm-y += nvkm/subdev/ltc/gm200.o
 nvkm-y += nvkm/subdev/ltc/gp100.o
+nvkm-y += nvkm/subdev/ltc/gp102.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/base.c
index 1f18527..2324217 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/base.c
@@ -55,6 +55,14 @@ nvkm_ltc_zbc_depth_get(struct nvkm_ltc *ltc, int index, const u32 depth)
 	return index;
 }
 
+int
+nvkm_ltc_zbc_stencil_get(struct nvkm_ltc *ltc, int index, const u32 stencil)
+{
+	ltc->zbc_stencil[index] = stencil;
+	ltc->func->zbc_clear_stencil(ltc, index, stencil);
+	return index;
+}
+
 void
 nvkm_ltc_invalidate(struct nvkm_ltc *ltc)
 {
@@ -92,6 +100,8 @@ nvkm_ltc_init(struct nvkm_subdev *subdev)
 	for (i = ltc->zbc_min; i <= ltc->zbc_max; i++) {
 		ltc->func->zbc_clear_color(ltc, i, ltc->zbc_color[i]);
 		ltc->func->zbc_clear_depth(ltc, i, ltc->zbc_depth[i]);
+		if (ltc->func->zbc_clear_stencil)
+			ltc->func->zbc_clear_stencil(ltc, i, ltc->zbc_stencil[i]);
 	}
 
 	ltc->func->init(ltc);
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gp100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gp100.c
index e34d421..e923ed7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gp100.c
@@ -23,7 +23,7 @@
  */
 #include "priv.h"
 
-static void
+void
 gp100_ltc_intr(struct nvkm_ltc *ltc)
 {
 	struct nvkm_device *device = ltc->subdev.device;
@@ -38,7 +38,7 @@ gp100_ltc_intr(struct nvkm_ltc *ltc)
 	}
 }
 
-static int
+int
 gp100_ltc_oneinit(struct nvkm_ltc *ltc)
 {
 	struct nvkm_device *device = ltc->subdev.device;
@@ -48,7 +48,7 @@ gp100_ltc_oneinit(struct nvkm_ltc *ltc)
 	return 0;
 }
 
-static void
+void
 gp100_ltc_init(struct nvkm_ltc *ltc)
 {
 	/*XXX: PMU LS call to setup tagram address */
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gp102.c b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gp102.c
new file mode 100644
index 0000000..601747a
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gp102.c
@@ -0,0 +1,51 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "priv.h"
+
+void
+gp102_ltc_zbc_clear_stencil(struct nvkm_ltc *ltc, int i, const u32 stencil)
+{
+	struct nvkm_device *device = ltc->subdev.device;
+	nvkm_mask(device, 0x17e338, 0x0000000f, i);
+	nvkm_wr32(device, 0x17e204, stencil);
+}
+
+static const struct nvkm_ltc_func
+gp102_ltc = {
+	.oneinit = gp100_ltc_oneinit,
+	.init = gp100_ltc_init,
+	.intr = gp100_ltc_intr,
+	.cbc_clear = gm107_ltc_cbc_clear,
+	.cbc_wait = gm107_ltc_cbc_wait,
+	.zbc = 16,
+	.zbc_clear_color = gm107_ltc_zbc_clear_color,
+	.zbc_clear_depth = gm107_ltc_zbc_clear_depth,
+	.zbc_clear_stencil = gp102_ltc_zbc_clear_stencil,
+	.invalidate = gf100_ltc_invalidate,
+	.flush = gf100_ltc_flush,
+};
+
+int
+gp102_ltc_new(struct nvkm_device *device, int index, struct nvkm_ltc **pltc)
+{
+	return nvkm_ltc_new_(&gp102_ltc, device, index, pltc);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/priv.h b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/priv.h
index e71cc25..9dcde43 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/priv.h
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/priv.h
@@ -19,6 +19,7 @@ struct nvkm_ltc_func {
 	int zbc;
 	void (*zbc_clear_color)(struct nvkm_ltc *, int, const u32[4]);
 	void (*zbc_clear_depth)(struct nvkm_ltc *, int, const u32);
+	void (*zbc_clear_stencil)(struct nvkm_ltc *, int, const u32);
 
 	void (*invalidate)(struct nvkm_ltc *);
 	void (*flush)(struct nvkm_ltc *);
@@ -41,4 +42,8 @@ void gm107_ltc_cbc_clear(struct nvkm_ltc *, u32, u32);
 void gm107_ltc_cbc_wait(struct nvkm_ltc *);
 void gm107_ltc_zbc_clear_color(struct nvkm_ltc *, int, const u32[4]);
 void gm107_ltc_zbc_clear_depth(struct nvkm_ltc *, int, const u32);
+
+int gp100_ltc_oneinit(struct nvkm_ltc *);
+void gp100_ltc_init(struct nvkm_ltc *);
+void gp100_ltc_intr(struct nvkm_ltc *);
 #endif
-- 
cgit v1.1


From 19ca10d82e33bcfe92412c461fc3534ec1e14747 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/gem: lookup VMAs for buffers referenced by pushbuf ioctl

We previously only did this for push buffers, but an upcoming patch will
need to attach fences to all VMAs to resolve another issue.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nouveau_dma.c | 10 +---------
 drivers/gpu/drm/nouveau/nouveau_dma.h |  5 ++---
 drivers/gpu/drm/nouveau/nouveau_gem.c | 19 ++++++++++++++++---
 3 files changed, 19 insertions(+), 15 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nouveau_dma.c b/drivers/gpu/drm/nouveau/nouveau_dma.c
index 10e84f6..e0664d2 100644
--- a/drivers/gpu/drm/nouveau/nouveau_dma.c
+++ b/drivers/gpu/drm/nouveau/nouveau_dma.c
@@ -80,18 +80,10 @@ READ_GET(struct nouveau_channel *chan, uint64_t *prev_get, int *timeout)
 }
 
 void
-nv50_dma_push(struct nouveau_channel *chan, struct nouveau_bo *bo,
-	      int delta, int length)
+nv50_dma_push(struct nouveau_channel *chan, u64 offset, int length)
 {
-	struct nouveau_cli *cli = (void *)chan->user.client;
 	struct nouveau_bo *pb = chan->push.buffer;
-	struct nouveau_vma *vma;
 	int ip = (chan->dma.ib_put * 2) + chan->dma.ib_base;
-	u64 offset;
-
-	vma = nouveau_vma_find(bo, &cli->vmm);
-	BUG_ON(!vma);
-	offset = vma->addr + delta;
 
 	BUG_ON(chan->dma.ib_free < 1);
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_dma.h b/drivers/gpu/drm/nouveau/nouveau_dma.h
index 74e10b1..89c8711 100644
--- a/drivers/gpu/drm/nouveau/nouveau_dma.h
+++ b/drivers/gpu/drm/nouveau/nouveau_dma.h
@@ -31,8 +31,7 @@
 #include "nouveau_chan.h"
 
 int nouveau_dma_wait(struct nouveau_channel *, int slots, int size);
-void nv50_dma_push(struct nouveau_channel *, struct nouveau_bo *,
-		   int delta, int length);
+void nv50_dma_push(struct nouveau_channel *, u64 addr, int length);
 
 /*
  * There's a hw race condition where you can't jump to your PUT offset,
@@ -151,7 +150,7 @@ FIRE_RING(struct nouveau_channel *chan)
 	chan->accel_done = true;
 
 	if (chan->dma.ib_max) {
-		nv50_dma_push(chan, chan->push.buffer, chan->dma.put << 2,
+		nv50_dma_push(chan, chan->push.addr + (chan->dma.put << 2),
 			      (chan->dma.cur - chan->dma.put) << 2);
 	} else {
 		WRITE_PUT(chan->dma.cur);
diff --git a/drivers/gpu/drm/nouveau/nouveau_gem.c b/drivers/gpu/drm/nouveau/nouveau_gem.c
index e72a7e3..707e02c 100644
--- a/drivers/gpu/drm/nouveau/nouveau_gem.c
+++ b/drivers/gpu/drm/nouveau/nouveau_gem.c
@@ -432,7 +432,20 @@ retry:
 			}
 		}
 
-		b->user_priv = (uint64_t)(unsigned long)nvbo;
+		if (cli->vmm.vmm.object.oclass >= NVIF_CLASS_VMM_NV50) {
+			struct nouveau_vmm *vmm = &cli->vmm;
+			struct nouveau_vma *vma = nouveau_vma_find(nvbo, vmm);
+			if (!vma) {
+				NV_PRINTK(err, cli, "vma not found!\n");
+				ret = -EINVAL;
+				break;
+			}
+
+			b->user_priv = (uint64_t)(unsigned long)vma;
+		} else {
+			b->user_priv = (uint64_t)(unsigned long)nvbo;
+		}
+
 		nvbo->reserved_by = file_priv;
 		nvbo->pbbo_index = i;
 		if ((b->valid_domains & NOUVEAU_GEM_DOMAIN_VRAM) &&
@@ -763,10 +776,10 @@ nouveau_gem_ioctl_pushbuf(struct drm_device *dev, void *data,
 		}
 
 		for (i = 0; i < req->nr_push; i++) {
-			struct nouveau_bo *nvbo = (void *)(unsigned long)
+			struct nouveau_vma *vma = (void *)(unsigned long)
 				bo[push[i].bo_index].user_priv;
 
-			nv50_dma_push(chan, nvbo, push[i].offset,
+			nv50_dma_push(chan, vma->addr + push[i].offset,
 				      push[i].length);
 		}
 	} else
-- 
cgit v1.1


From 0db912af8f5ad4fa4dc08a9c8e411a10953c5403 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/gem: attach fences to VMAs to track GPU usage

An upcoming patch will use these to fix issues related to the deferred
unmapping of GEM objects.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nouveau_gem.c | 14 +++++++++++++-
 drivers/gpu/drm/nouveau/nouveau_vmm.c |  1 +
 drivers/gpu/drm/nouveau/nouveau_vmm.h |  2 ++
 3 files changed, 16 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/nouveau/nouveau_gem.c b/drivers/gpu/drm/nouveau/nouveau_gem.c
index 707e02c..2016d9e 100644
--- a/drivers/gpu/drm/nouveau/nouveau_gem.c
+++ b/drivers/gpu/drm/nouveau/nouveau_gem.c
@@ -99,6 +99,7 @@ struct nouveau_gem_object_unmap {
 static void
 nouveau_gem_object_delete(struct nouveau_vma *vma)
 {
+	nouveau_fence_unref(&vma->fence);
 	nouveau_vma_del(&vma);
 }
 
@@ -344,9 +345,20 @@ validate_fini_no_ticket(struct validate_op *op, struct nouveau_fence *fence,
 		nvbo = list_entry(op->list.next, struct nouveau_bo, entry);
 		b = &pbbo[nvbo->pbbo_index];
 
-		if (likely(fence))
+		if (likely(fence)) {
+			struct nouveau_drm *drm = nouveau_bdev(nvbo->bo.bdev);
+			struct nouveau_vma *vma;
+
 			nouveau_bo_fence(nvbo, fence, !!b->write_domains);
 
+			if (drm->client.vmm.vmm.object.oclass >= NVIF_CLASS_VMM_NV50) {
+				vma = (void *)(unsigned long)b->user_priv;
+				nouveau_fence_unref(&vma->fence);
+				dma_fence_get(&fence->base);
+				vma->fence = fence;
+			}
+		}
+
 		if (unlikely(nvbo->validate_mapped)) {
 			ttm_bo_kunmap(&nvbo->kmap);
 			nvbo->validate_mapped = false;
diff --git a/drivers/gpu/drm/nouveau/nouveau_vmm.c b/drivers/gpu/drm/nouveau/nouveau_vmm.c
index f5371d9..2032c3e 100644
--- a/drivers/gpu/drm/nouveau/nouveau_vmm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_vmm.c
@@ -92,6 +92,7 @@ nouveau_vma_new(struct nouveau_bo *nvbo, struct nouveau_vmm *vmm,
 	vma->refs = 1;
 	vma->addr = ~0ULL;
 	vma->mem = NULL;
+	vma->fence = NULL;
 	list_add_tail(&vma->head, &nvbo->vma_list);
 
 	if (nvbo->bo.mem.mem_type != TTM_PL_SYSTEM &&
diff --git a/drivers/gpu/drm/nouveau/nouveau_vmm.h b/drivers/gpu/drm/nouveau/nouveau_vmm.h
index 5c31f43..7e3b118 100644
--- a/drivers/gpu/drm/nouveau/nouveau_vmm.h
+++ b/drivers/gpu/drm/nouveau/nouveau_vmm.h
@@ -11,6 +11,8 @@ struct nouveau_vma {
 	u64 addr;
 
 	struct nouveau_mem *mem;
+
+	struct nouveau_fence *fence;
 };
 
 struct nouveau_vma *nouveau_vma_find(struct nouveau_bo *, struct nouveau_vmm *);
-- 
cgit v1.1


From 470db8b78186efe840b6452c6c4934178058059e Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/gem: tie deferred unmapping of buffers to VMA fence
 completion

As VMAs are per-client, unlike buffers, this allows us to avoid referencing
foreign fences (those that belong to another client/driver) from the client
deferred work handler, and prevent some not-fun race conditions that can be
triggered when a fence stalls.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nouveau_gem.c | 17 ++---------------
 1 file changed, 2 insertions(+), 15 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nouveau_gem.c b/drivers/gpu/drm/nouveau/nouveau_gem.c
index 2016d9e..300daee 100644
--- a/drivers/gpu/drm/nouveau/nouveau_gem.c
+++ b/drivers/gpu/drm/nouveau/nouveau_gem.c
@@ -115,25 +115,12 @@ nouveau_gem_object_delete_work(struct nouveau_cli_work *w)
 static void
 nouveau_gem_object_unmap(struct nouveau_bo *nvbo, struct nouveau_vma *vma)
 {
-	const bool mapped = nvbo->bo.mem.mem_type != TTM_PL_SYSTEM;
-	struct reservation_object *resv = nvbo->bo.resv;
-	struct reservation_object_list *fobj;
+	struct dma_fence *fence = vma->fence ? &vma->fence->base : NULL;
 	struct nouveau_gem_object_unmap *work;
-	struct dma_fence *fence = NULL;
-
-	fobj = reservation_object_get_list(resv);
 
 	list_del_init(&vma->head);
 
-	if (fobj && fobj->shared_count > 1)
-		ttm_bo_wait(&nvbo->bo, false, false);
-	else if (fobj && fobj->shared_count == 1)
-		fence = rcu_dereference_protected(fobj->shared[0],
-						reservation_object_held(resv));
-	else
-		fence = reservation_object_get_excl(nvbo->bo.resv);
-
-	if (!fence || !mapped) {
+	if (!fence) {
 		nouveau_gem_object_delete(vma);
 		return;
 	}
-- 
cgit v1.1


From 11e451e74050d9e9030581ce40337838acfcea5b Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau: remove fence wait code from deferred client work handler

Fences attached to deferred client work items now originate from channels
belonging to the client, meaning we can be certain they've been signalled
before we destroy a client.

This closes a race that could happen if the dma_fence_wait_timeout() call
didn't succeed.  When the fence was later signalled, a use-after-free was
possible.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nouveau_drm.c | 30 ++++++++++++++----------------
 1 file changed, 14 insertions(+), 16 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c
index 6caece4..64b8fd0 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_drm.c
@@ -113,24 +113,22 @@ nouveau_name(struct drm_device *dev)
 }
 
 static inline bool
-nouveau_cli_work_ready(struct dma_fence *fence, bool wait)
+nouveau_cli_work_ready(struct dma_fence *fence)
 {
-	if (!dma_fence_is_signaled(fence)) {
-		if (!wait)
-			return false;
-		WARN_ON(dma_fence_wait_timeout(fence, false, 2 * HZ) <= 0);
-	}
+	if (!dma_fence_is_signaled(fence))
+		return false;
 	dma_fence_put(fence);
 	return true;
 }
 
 static void
-nouveau_cli_work_flush(struct nouveau_cli *cli, bool wait)
+nouveau_cli_work(struct work_struct *w)
 {
+	struct nouveau_cli *cli = container_of(w, typeof(*cli), work);
 	struct nouveau_cli_work *work, *wtmp;
 	mutex_lock(&cli->lock);
 	list_for_each_entry_safe(work, wtmp, &cli->worker, head) {
-		if (!work->fence || nouveau_cli_work_ready(work->fence, wait)) {
+		if (!work->fence || nouveau_cli_work_ready(work->fence)) {
 			list_del(&work->head);
 			work->func(work);
 		}
@@ -159,16 +157,16 @@ nouveau_cli_work_queue(struct nouveau_cli *cli, struct dma_fence *fence,
 }
 
 static void
-nouveau_cli_work(struct work_struct *w)
-{
-	struct nouveau_cli *cli = container_of(w, typeof(*cli), work);
-	nouveau_cli_work_flush(cli, false);
-}
-
-static void
 nouveau_cli_fini(struct nouveau_cli *cli)
 {
-	nouveau_cli_work_flush(cli, true);
+	/* All our channels are dead now, which means all the fences they
+	 * own are signalled, and all callback functions have been called.
+	 *
+	 * So, after flushing the workqueue, there should be nothing left.
+	 */
+	flush_work(&cli->work);
+	WARN_ON(!list_empty(&cli->worker));
+
 	usif_client_fini(cli);
 	nouveau_vmm_fini(&cli->vmm);
 	nvif_mmu_fini(&cli->mmu);
-- 
cgit v1.1


From 92b4eaaf9a84a7bd35db6f903c0ecbda4f9594ee Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau: no need to create ctxdma for push buffers on fermi and
 up

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nouveau_chan.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nouveau_chan.c b/drivers/gpu/drm/nouveau/nouveau_chan.c
index 67950a5c5..97900e9 100644
--- a/drivers/gpu/drm/nouveau/nouveau_chan.c
+++ b/drivers/gpu/drm/nouveau/nouveau_chan.c
@@ -163,12 +163,15 @@ nouveau_channel_prep(struct nouveau_drm *drm, struct nvif_device *device,
 			return ret;
 		}
 
+		chan->push.addr = chan->push.vma->addr;
+
+		if (device->info.family >= NV_DEVICE_INFO_V0_FERMI)
+			return 0;
+
 		args.target = NV_DMA_V0_TARGET_VM;
 		args.access = NV_DMA_V0_ACCESS_VM;
 		args.start = 0;
 		args.limit = cli->vmm.vmm.limit - 1;
-
-		chan->push.addr = chan->push.vma->addr;
 	} else
 	if (chan->push.buffer->bo.mem.mem_type == TTM_PL_VRAM) {
 		if (device->info.family == NV_DEVICE_INFO_V0_TNT) {
-- 
cgit v1.1


From 512fa0b8a398539c3c2db251f6c40da4ef065d09 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/drm/nv50-: remove allocation of sw class

Hasn't been required for a long time.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nouveau_dma.h |  1 -
 drivers/gpu/drm/nouveau/nouveau_drm.c | 40 +++++++++++++++++------------------
 2 files changed, 19 insertions(+), 22 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nouveau_dma.h b/drivers/gpu/drm/nouveau/nouveau_dma.h
index 89c8711..fc5e3f4 100644
--- a/drivers/gpu/drm/nouveau/nouveau_dma.h
+++ b/drivers/gpu/drm/nouveau/nouveau_dma.h
@@ -54,7 +54,6 @@ enum {
 
 	NvSub2D		= 3, /* DO NOT CHANGE - hardcoded for kepler gr fifo */
 	NvSubCopy	= 4, /* DO NOT CHANGE - hardcoded for kepler gr fifo */
-	FermiSw		= 5, /* DO NOT CHANGE (well.. 6/7 will work...) */
 };
 
 /* Object handles - for stuff that's doesn't use handle == oclass. */
diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c
index 64b8fd0..8e506c5 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_drm.c
@@ -388,38 +388,36 @@ nouveau_accel_init(struct nouveau_drm *drm)
 		return;
 	}
 
-	ret = nvif_object_init(&drm->channel->user, NVDRM_NVSW,
-			       nouveau_abi16_swclass(drm), NULL, 0, &drm->nvsw);
-	if (ret == 0) {
-		ret = RING_SPACE(drm->channel, 2);
+	if (device->info.family < NV_DEVICE_INFO_V0_TESLA) {
+		ret = nvif_object_init(&drm->channel->user, NVDRM_NVSW,
+				       nouveau_abi16_swclass(drm), NULL, 0,
+				       &drm->nvsw);
 		if (ret == 0) {
-			if (device->info.family < NV_DEVICE_INFO_V0_FERMI) {
+			ret = RING_SPACE(drm->channel, 2);
+			if (ret == 0) {
 				BEGIN_NV04(drm->channel, NvSubSw, 0, 1);
-				OUT_RING  (drm->channel, NVDRM_NVSW);
-			} else
-			if (device->info.family < NV_DEVICE_INFO_V0_KEPLER) {
-				BEGIN_NVC0(drm->channel, FermiSw, 0, 1);
-				OUT_RING  (drm->channel, 0x001f0000);
+				OUT_RING  (drm->channel, drm->nvsw.handle);
+			}
+
+			ret = nvif_notify_init(&drm->nvsw,
+					       nouveau_flip_complete,
+					       false, NV04_NVSW_NTFY_UEVENT,
+					       NULL, 0, 0, &drm->flip);
+			if (ret == 0)
+				ret = nvif_notify_get(&drm->flip);
+			if (ret) {
+				nouveau_accel_fini(drm);
+				return;
 			}
 		}
 
-		ret = nvif_notify_init(&drm->nvsw, nouveau_flip_complete,
-				       false, NV04_NVSW_NTFY_UEVENT,
-				       NULL, 0, 0, &drm->flip);
-		if (ret == 0)
-			ret = nvif_notify_get(&drm->flip);
 		if (ret) {
+			NV_ERROR(drm, "failed to allocate sw class, %d\n", ret);
 			nouveau_accel_fini(drm);
 			return;
 		}
 	}
 
-	if (ret) {
-		NV_ERROR(drm, "failed to allocate software object, %d\n", ret);
-		nouveau_accel_fini(drm);
-		return;
-	}
-
 	if (device->info.family < NV_DEVICE_INFO_V0_FERMI) {
 		ret = nvkm_gpuobj_new(nvxx_device(&drm->client.device), 32, 0,
 				      false, NULL, &drm->notify);
-- 
cgit v1.1


From 0d4a2c5767dc6136079b11ed45934143d309026e Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms: move display class instantiation to library

This function is useful outside of DRM code.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/include/nvif/disp.h   | 12 ++++++
 drivers/gpu/drm/nouveau/include/nvif/object.h | 16 ++++++++
 drivers/gpu/drm/nouveau/nouveau_connector.c   | 14 +++----
 drivers/gpu/drm/nouveau/nouveau_display.c     | 35 ++++------------
 drivers/gpu/drm/nouveau/nouveau_display.h     |  3 +-
 drivers/gpu/drm/nouveau/nv50_display.c        | 50 +++++++++++------------
 drivers/gpu/drm/nouveau/nvif/Kbuild           |  1 +
 drivers/gpu/drm/nouveau/nvif/disp.c           | 59 +++++++++++++++++++++++++++
 8 files changed, 129 insertions(+), 61 deletions(-)
 create mode 100644 drivers/gpu/drm/nouveau/include/nvif/disp.h
 create mode 100644 drivers/gpu/drm/nouveau/nvif/disp.c

diff --git a/drivers/gpu/drm/nouveau/include/nvif/disp.h b/drivers/gpu/drm/nouveau/include/nvif/disp.h
new file mode 100644
index 0000000..7c0eda3
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/include/nvif/disp.h
@@ -0,0 +1,12 @@
+#ifndef __NVIF_DISP_H__
+#define __NVIF_DISP_H__
+#include <nvif/object.h>
+struct nvif_device;
+
+struct nvif_disp {
+	struct nvif_object object;
+};
+
+int nvif_disp_ctor(struct nvif_device *, s32 oclass, struct nvif_disp *);
+void nvif_disp_dtor(struct nvif_disp *);
+#endif
diff --git a/drivers/gpu/drm/nouveau/include/nvif/object.h b/drivers/gpu/drm/nouveau/include/nvif/object.h
index a2d5244..20754d9 100644
--- a/drivers/gpu/drm/nouveau/include/nvif/object.h
+++ b/drivers/gpu/drm/nouveau/include/nvif/object.h
@@ -99,6 +99,22 @@ struct nvif_mclass {
 	ret;                                                                   \
 })
 
+#define nvif_sclass(o,m,u) ({                                                  \
+	const typeof(m[0]) *_mclass = (m);                                     \
+	s32 _oclass = (u);                                                     \
+	int _cid;                                                              \
+	if (_oclass) {                                                         \
+		for (_cid = 0; _mclass[_cid].oclass; _cid++) {                 \
+			if (_mclass[_cid].oclass == _oclass)                   \
+				break;                                         \
+		}                                                              \
+		_cid = _mclass[_cid].oclass ? _cid : -ENOSYS;                  \
+	} else {                                                               \
+		_cid = nvif_mclass((o), _mclass);                              \
+	}                                                                      \
+	_cid;                                                                  \
+})
+
 /*XXX*/
 #include <core/object.h>
 #define nvxx_object(a) ({                                                      \
diff --git a/drivers/gpu/drm/nouveau/nouveau_connector.c b/drivers/gpu/drm/nouveau/nouveau_connector.c
index 6ed9cb0..18e3239 100644
--- a/drivers/gpu/drm/nouveau/nouveau_connector.c
+++ b/drivers/gpu/drm/nouveau/nouveau_connector.c
@@ -151,7 +151,7 @@ nouveau_conn_atomic_set_property(struct drm_connector *connector,
 				/* ... except prior to G80, where the code
 				 * doesn't support such things.
 				 */
-				if (disp->disp.oclass < NV50_DISP)
+				if (disp->disp.object.oclass < NV50_DISP)
 					return -EINVAL;
 				break;
 			default:
@@ -260,7 +260,7 @@ nouveau_conn_reset(struct drm_connector *connector)
 	asyc->procamp.color_vibrance = 150;
 	asyc->procamp.vibrant_hue = 90;
 
-	if (nouveau_display(connector->dev)->disp.oclass < NV50_DISP) {
+	if (nouveau_display(connector->dev)->disp.object.oclass < NV50_DISP) {
 		switch (connector->connector_type) {
 		case DRM_MODE_CONNECTOR_LVDS:
 			/* See note in nouveau_conn_atomic_set_property(). */
@@ -314,7 +314,7 @@ nouveau_conn_attach_properties(struct drm_connector *connector)
 	case DRM_MODE_CONNECTOR_TV:
 		break;
 	case DRM_MODE_CONNECTOR_VGA:
-		if (disp->disp.oclass < NV50_DISP)
+		if (disp->disp.object.oclass < NV50_DISP)
 			break; /* Can only scale on DFPs. */
 		/* Fall-through. */
 	default:
@@ -1321,7 +1321,7 @@ nouveau_connector_create(struct drm_device *dev, int index)
 	}
 
 	/* HDMI 3D support */
-	if ((disp->disp.oclass >= G82_DISP)
+	if ((disp->disp.object.oclass >= G82_DISP)
 	    && ((type == DRM_MODE_CONNECTOR_DisplayPort)
 		|| (type == DRM_MODE_CONNECTOR_eDP)
 		|| (type == DRM_MODE_CONNECTOR_HDMIA)))
@@ -1343,7 +1343,7 @@ nouveau_connector_create(struct drm_device *dev, int index)
 	case DCB_CONNECTOR_LVDS_SPWG:
 	case DCB_CONNECTOR_eDP:
 		/* see note in nouveau_connector_set_property() */
-		if (disp->disp.oclass < NV50_DISP) {
+		if (disp->disp.object.oclass < NV50_DISP) {
 			nv_connector->scaling_mode = DRM_MODE_SCALE_FULLSCREEN;
 			break;
 		}
@@ -1366,8 +1366,8 @@ nouveau_connector_create(struct drm_device *dev, int index)
 		break;
 	}
 
-	ret = nvif_notify_init(&disp->disp, nouveau_connector_hotplug, true,
-			       NV04_DISP_NTFY_CONN,
+	ret = nvif_notify_init(&disp->disp.object, nouveau_connector_hotplug,
+			       true, NV04_DISP_NTFY_CONN,
 			       &(struct nvif_notify_conn_req_v0) {
 				.mask = NVIF_NOTIFY_CONN_V0_ANY,
 				.conn = index,
diff --git a/drivers/gpu/drm/nouveau/nouveau_display.c b/drivers/gpu/drm/nouveau/nouveau_display.c
index 7d0bec8d..774b429 100644
--- a/drivers/gpu/drm/nouveau/nouveau_display.c
+++ b/drivers/gpu/drm/nouveau/nouveau_display.c
@@ -116,7 +116,7 @@ nouveau_display_scanoutpos_head(struct drm_crtc *crtc, int *vpos, int *hpos,
 	bool ret = false;
 
 	do {
-		ret = nvif_mthd(&disp->disp, 0, &args, sizeof(args));
+		ret = nvif_mthd(&disp->disp.object, 0, &args, sizeof(args));
 		if (ret != 0)
 			return false;
 
@@ -175,7 +175,7 @@ nouveau_display_vblank_init(struct drm_device *dev)
 
 	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
 		struct nouveau_crtc *nv_crtc = nouveau_crtc(crtc);
-		ret = nvif_notify_init(&disp->disp,
+		ret = nvif_notify_init(&disp->disp.object,
 				       nouveau_display_vblank_handler, false,
 				       NV04_DISP_NTFY_VBLANK,
 				       &(struct nvif_notify_head_req_v0) {
@@ -454,10 +454,10 @@ nouveau_display_create_properties(struct drm_device *dev)
 	struct nouveau_display *disp = nouveau_display(dev);
 	int gen;
 
-	if (disp->disp.oclass < NV50_DISP)
+	if (disp->disp.object.oclass < NV50_DISP)
 		gen = 0;
 	else
-	if (disp->disp.oclass < GF110_DISP)
+	if (disp->disp.object.oclass < GF110_DISP)
 		gen = 1;
 	else
 		gen = 2;
@@ -533,31 +533,10 @@ nouveau_display_create(struct drm_device *dev)
 	drm_kms_helper_poll_disable(dev);
 
 	if (nouveau_modeset != 2 && drm->vbios.dcb.entries) {
-		static const u16 oclass[] = {
-			GP102_DISP,
-			GP100_DISP,
-			GM200_DISP,
-			GM107_DISP,
-			GK110_DISP,
-			GK104_DISP,
-			GF110_DISP,
-			GT214_DISP,
-			GT206_DISP,
-			GT200_DISP,
-			G82_DISP,
-			NV50_DISP,
-			NV04_DISP,
-		};
-		int i;
-
-		for (i = 0, ret = -ENODEV; ret && i < ARRAY_SIZE(oclass); i++) {
-			ret = nvif_object_init(&drm->client.device.object, 0,
-					       oclass[i], NULL, 0, &disp->disp);
-		}
-
+		ret = nvif_disp_ctor(&drm->client.device, 0, &disp->disp);
 		if (ret == 0) {
 			nouveau_display_create_properties(dev);
-			if (disp->disp.oclass < NV50_DISP)
+			if (disp->disp.object.oclass < NV50_DISP)
 				ret = nv04_display_create(dev);
 			else
 				ret = nv50_display_create(dev);
@@ -611,7 +590,7 @@ nouveau_display_destroy(struct drm_device *dev)
 	if (disp->dtor)
 		disp->dtor(dev);
 
-	nvif_object_fini(&disp->disp);
+	nvif_disp_dtor(&disp->disp);
 
 	nouveau_drm(dev)->display = NULL;
 	kfree(disp);
diff --git a/drivers/gpu/drm/nouveau/nouveau_display.h b/drivers/gpu/drm/nouveau/nouveau_display.h
index 270ba56..54aa7c3 100644
--- a/drivers/gpu/drm/nouveau/nouveau_display.h
+++ b/drivers/gpu/drm/nouveau/nouveau_display.h
@@ -2,6 +2,7 @@
 #ifndef __NOUVEAU_DISPLAY_H__
 #define __NOUVEAU_DISPLAY_H__
 #include "nouveau_drv.h"
+#include <nvif/disp.h>
 
 struct nouveau_framebuffer {
 	struct drm_framebuffer base;
@@ -38,7 +39,7 @@ struct nouveau_display {
 	int  (*init)(struct drm_device *);
 	void (*fini)(struct drm_device *);
 
-	struct nvif_object disp;
+	struct nvif_disp disp;
 
 	struct drm_property *dithering_mode;
 	struct drm_property *dithering_depth;
diff --git a/drivers/gpu/drm/nouveau/nv50_display.c b/drivers/gpu/drm/nouveau/nv50_display.c
index e90330e..f2156c8 100644
--- a/drivers/gpu/drm/nouveau/nv50_display.c
+++ b/drivers/gpu/drm/nouveau/nv50_display.c
@@ -663,7 +663,7 @@ struct nv50_head {
 #define nv50_vers(c) nv50_chan(c)->user.oclass
 
 struct nv50_disp {
-	struct nvif_object *disp;
+	struct nvif_disp *disp;
 	struct nv50_mast mast;
 
 	struct nouveau_bo *sync;
@@ -1201,7 +1201,7 @@ nv50_curs_new(struct nouveau_drm *drm, struct nv50_head *head,
 	struct nv50_curs *curs;
 	int cid, ret;
 
-	cid = nvif_mclass(disp->disp, curses);
+	cid = nvif_mclass(&disp->disp->object, curses);
 	if (cid < 0) {
 		NV_ERROR(drm, "No supported cursor immediate class\n");
 		return cid;
@@ -1219,8 +1219,8 @@ nv50_curs_new(struct nouveau_drm *drm, struct nv50_head *head,
 		return ret;
 	}
 
-	ret = nvif_object_init(disp->disp, 0, curses[cid].oclass, &args,
-			       sizeof(args), &curs->chan);
+	ret = nvif_object_init(&disp->disp->object, 0, curses[cid].oclass,
+			       &args, sizeof(args), &curs->chan);
 	if (ret) {
 		NV_ERROR(drm, "curs%04x allocation failed: %d\n",
 			 curses[cid].oclass, ret);
@@ -1517,8 +1517,8 @@ nv50_base_new(struct nouveau_drm *drm, struct nv50_head *head,
 		return ret;
 	}
 
-	ret = nv50_base_create(&drm->client.device, disp->disp, base->id,
-			       disp->sync->bo.offset, &base->chan);
+	ret = nv50_base_create(&drm->client.device, &disp->disp->object,
+			       base->id, disp->sync->bo.offset, &base->chan);
 	if (ret)
 		return ret;
 
@@ -2105,7 +2105,7 @@ nv50_head_atomic_check_lut(struct nv50_head *head,
 		return;
 	}
 
-	if (disp->disp->oclass < GF110_DISP) {
+	if (disp->disp->object.oclass < GF110_DISP) {
 		asyh->lut.mode = (asyh->base.cpp == 1) ? 0 : 1;
 		asyh->set.ilut = true;
 	} else {
@@ -2404,12 +2404,12 @@ nv50_head_create(struct drm_device *dev, int index)
 	}
 
 	/* allocate overlay resources */
-	ret = nv50_oimm_create(device, disp->disp, index, &head->oimm);
+	ret = nv50_oimm_create(device, &disp->disp->object, index, &head->oimm);
 	if (ret)
 		goto out;
 
-	ret = nv50_ovly_create(device, disp->disp, index, disp->sync->bo.offset,
-			       &head->ovly);
+	ret = nv50_ovly_create(device, &disp->disp->object, index,
+			       disp->sync->bo.offset, &head->ovly);
 	if (ret)
 		goto out;
 
@@ -2435,7 +2435,7 @@ nv50_outp_release(struct nouveau_encoder *nv_encoder)
 		.base.hashm  = nv_encoder->dcb->hashm,
 	};
 
-	nvif_mthd(disp->disp, 0, &args, sizeof(args));
+	nvif_mthd(&disp->disp->object, 0, &args, sizeof(args));
 	nv_encoder->or = -1;
 	nv_encoder->link = 0;
 }
@@ -2456,7 +2456,7 @@ nv50_outp_acquire(struct nouveau_encoder *nv_encoder)
 	};
 	int ret;
 
-	ret = nvif_mthd(disp->disp, 0, &args, sizeof(args));
+	ret = nvif_mthd(&disp->disp->object, 0, &args, sizeof(args));
 	if (ret) {
 		NV_ERROR(drm, "error acquiring output path: %d\n", ret);
 		return ret;
@@ -2618,7 +2618,7 @@ nv50_dac_detect(struct drm_encoder *encoder, struct drm_connector *connector)
 	if (args.load.data == 0)
 		args.load.data = 340;
 
-	ret = nvif_mthd(disp->disp, 0, &args, sizeof(args));
+	ret = nvif_mthd(&disp->disp->object, 0, &args, sizeof(args));
 	if (ret || !args.load.load)
 		return connector_status_disconnected;
 
@@ -2694,7 +2694,7 @@ nv50_audio_disable(struct drm_encoder *encoder, struct nouveau_crtc *nv_crtc)
 				(0x0100 << nv_crtc->index),
 	};
 
-	nvif_mthd(disp->disp, 0, &args, sizeof(args));
+	nvif_mthd(&disp->disp->object, 0, &args, sizeof(args));
 }
 
 static void
@@ -2724,7 +2724,7 @@ nv50_audio_enable(struct drm_encoder *encoder, struct drm_display_mode *mode)
 
 	memcpy(args.data, nv_connector->base.eld, sizeof(args.data));
 
-	nvif_mthd(disp->disp, 0, &args,
+	nvif_mthd(&disp->disp->object, 0, &args,
 		  sizeof(args.base) + drm_eld_size(args.data));
 }
 
@@ -2747,7 +2747,7 @@ nv50_hdmi_disable(struct drm_encoder *encoder, struct nouveau_crtc *nv_crtc)
 			       (0x0100 << nv_crtc->index),
 	};
 
-	nvif_mthd(disp->disp, 0, &args, sizeof(args));
+	nvif_mthd(&disp->disp->object, 0, &args, sizeof(args));
 }
 
 static void
@@ -2808,7 +2808,7 @@ nv50_hdmi_enable(struct drm_encoder *encoder, struct drm_display_mode *mode)
 		+ sizeof(args.pwr)
 		+ args.pwr.avi_infoframe_length
 		+ args.pwr.vendor_infoframe_length;
-	nvif_mthd(disp->disp, 0, &args, size);
+	nvif_mthd(&disp->disp->object, 0, &args, size);
 	nv50_audio_enable(encoder, mode);
 }
 
@@ -2923,7 +2923,7 @@ nv50_msto_prepare(struct nv50_msto *msto)
 		  msto->encoder.name, msto->head->base.base.name,
 		  args.vcpi.start_slot, args.vcpi.num_slots,
 		  args.vcpi.pbn, args.vcpi.aligned_pbn);
-	nvif_mthd(&drm->display->disp, 0, &args, sizeof(args));
+	nvif_mthd(&drm->display->disp.object, 0, &args, sizeof(args));
 }
 
 static int
@@ -3341,7 +3341,7 @@ nv50_mstm_enable(struct nv50_mstm *mstm, u8 dpcd, int state)
 		.mst.state = state,
 	};
 	struct nouveau_drm *drm = nouveau_drm(outp->base.base.dev);
-	struct nvif_object *disp = &drm->display->disp;
+	struct nvif_object *disp = &drm->display->disp.object;
 	int ret;
 
 	if (dpcd >= 0x12) {
@@ -3610,7 +3610,7 @@ nv50_sor_enable(struct drm_encoder *encoder)
 				lvds.lvds.script |= 0x0200;
 		}
 
-		nvif_mthd(disp->disp, 0, &lvds, sizeof(lvds));
+		nvif_mthd(&disp->disp->object, 0, &lvds, sizeof(lvds));
 		break;
 	case DCB_OUTPUT_DP:
 		if (nv_connector->base.display_info.bpc == 6)
@@ -3696,7 +3696,7 @@ nv50_sor_create(struct drm_connector *connector, struct dcb_output *dcbe)
 		struct nvkm_i2c_aux *aux =
 			nvkm_i2c_aux_find(i2c, dcbe->i2c_index);
 		if (aux) {
-			if (disp->disp->oclass < GF110_DISP) {
+			if (disp->disp->object.oclass < GF110_DISP) {
 				/* HW has no support for address-only
 				 * transactions, so we're required to
 				 * use custom I2C-over-AUX code.
@@ -3709,7 +3709,7 @@ nv50_sor_create(struct drm_connector *connector, struct dcb_output *dcbe)
 		}
 
 		/*TODO: Use DP Info Table to check for support. */
-		if (disp->disp->oclass >= GF110_DISP) {
+		if (disp->disp->object.oclass >= GF110_DISP) {
 			ret = nv50_mstm_new(nv_encoder, &nv_connector->aux, 16,
 					    nv_connector->base.base.id,
 					    &nv_encoder->dp.mstm);
@@ -4474,13 +4474,13 @@ nv50_display_create(struct drm_device *dev)
 		goto out;
 
 	/* allocate master evo channel */
-	ret = nv50_core_create(device, disp->disp, disp->sync->bo.offset,
-			      &disp->mast);
+	ret = nv50_core_create(device, &disp->disp->object,
+			       disp->sync->bo.offset, &disp->mast);
 	if (ret)
 		goto out;
 
 	/* create crtc objects to represent the hw heads */
-	if (disp->disp->oclass >= GF110_DISP)
+	if (disp->disp->object.oclass >= GF110_DISP)
 		crtcs = nvif_rd32(&device->object, 0x612004) & 0xf;
 	else
 		crtcs = 0x3;
diff --git a/drivers/gpu/drm/nouveau/nvif/Kbuild b/drivers/gpu/drm/nouveau/nvif/Kbuild
index c817b02..3db1250 100644
--- a/drivers/gpu/drm/nouveau/nvif/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvif/Kbuild
@@ -1,6 +1,7 @@
 nvif-y := nvif/object.o
 nvif-y += nvif/client.o
 nvif-y += nvif/device.o
+nvif-y += nvif/disp.o
 nvif-y += nvif/driver.o
 nvif-y += nvif/fifo.o
 nvif-y += nvif/mem.o
diff --git a/drivers/gpu/drm/nouveau/nvif/disp.c b/drivers/gpu/drm/nouveau/nvif/disp.c
new file mode 100644
index 0000000..7006482
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvif/disp.c
@@ -0,0 +1,59 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include <nvif/disp.h>
+#include <nvif/device.h>
+
+#include <nvif/class.h>
+
+void
+nvif_disp_dtor(struct nvif_disp *disp)
+{
+	nvif_object_fini(&disp->object);
+}
+
+int
+nvif_disp_ctor(struct nvif_device *device, s32 oclass, struct nvif_disp *disp)
+{
+	static const struct nvif_mclass disps[] = {
+		{ GP102_DISP, -1 },
+		{ GP100_DISP, -1 },
+		{ GM200_DISP, -1 },
+		{ GM107_DISP, -1 },
+		{ GK110_DISP, -1 },
+		{ GK104_DISP, -1 },
+		{ GF110_DISP, -1 },
+		{ GT214_DISP, -1 },
+		{ GT206_DISP, -1 },
+		{ GT200_DISP, -1 },
+		{   G82_DISP, -1 },
+		{  NV50_DISP, -1 },
+		{  NV04_DISP, -1 },
+		{}
+	};
+	int cid = nvif_sclass(&device->object, disps, oclass);
+	disp->object.client = NULL;
+	if (cid < 0)
+		return cid;
+
+	return nvif_object_init(&device->object, 0, disps[cid].oclass,
+				NULL, 0, &disp->object);
+}
-- 
cgit v1.1


From 30ed49b55b6e44e004c3095671e74fea93ee84cb Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50-: move code underneath dispnv50/

The code is about to be split up, and this matches dispnv04.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/Kbuild          |    8 +-
 drivers/gpu/drm/nouveau/dispnv50/Kbuild |    1 +
 drivers/gpu/drm/nouveau/dispnv50/disp.c | 4542 +++++++++++++++++++++++++++++++
 drivers/gpu/drm/nouveau/nv50_display.c  | 4542 -------------------------------
 4 files changed, 4546 insertions(+), 4547 deletions(-)
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/Kbuild
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/disp.c
 delete mode 100644 drivers/gpu/drm/nouveau/nv50_display.c

diff --git a/drivers/gpu/drm/nouveau/Kbuild b/drivers/gpu/drm/nouveau/Kbuild
index 9c0c650..b17843d 100644
--- a/drivers/gpu/drm/nouveau/Kbuild
+++ b/drivers/gpu/drm/nouveau/Kbuild
@@ -38,14 +38,16 @@ nouveau-y += nouveau_vmm.o
 
 # DRM - modesetting
 nouveau-$(CONFIG_DRM_NOUVEAU_BACKLIGHT) += nouveau_backlight.o
+nouveau-y += nouveau_bios.o
 nouveau-y += nouveau_connector.o
 nouveau-y += nouveau_display.o
-nouveau-y += nv50_display.o
 nouveau-y += nouveau_dp.o
 nouveau-y += nouveau_fbcon.o
 nouveau-y += nv04_fbcon.o
 nouveau-y += nv50_fbcon.o
 nouveau-y += nvc0_fbcon.o
+include $(src)/dispnv04/Kbuild
+include $(src)/dispnv50/Kbuild
 
 # DRM - command submission
 nouveau-y += nouveau_abi16.o
@@ -59,8 +61,4 @@ nouveau-y += nv50_fence.o
 nouveau-y += nv84_fence.o
 nouveau-y += nvc0_fence.o
 
-# DRM - prehistoric modesetting (NV04-G7x)
-nouveau-y += nouveau_bios.o
-include $(src)/dispnv04/Kbuild
-
 obj-$(CONFIG_DRM_NOUVEAU) += nouveau.o
diff --git a/drivers/gpu/drm/nouveau/dispnv50/Kbuild b/drivers/gpu/drm/nouveau/dispnv50/Kbuild
new file mode 100644
index 0000000..43fc8be
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/Kbuild
@@ -0,0 +1 @@
+nouveau-y += dispnv50/disp.o
diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c
new file mode 100644
index 0000000..f2156c8
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c
@@ -0,0 +1,4542 @@
+/*
+ * Copyright 2011 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include <linux/dma-mapping.h>
+#include <linux/hdmi.h>
+
+#include <drm/drmP.h>
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_dp_helper.h>
+#include <drm/drm_fb_helper.h>
+#include <drm/drm_plane_helper.h>
+#include <drm/drm_edid.h>
+
+#include <nvif/mem.h>
+
+#include <nvif/class.h>
+#include <nvif/cl0002.h>
+#include <nvif/cl5070.h>
+#include <nvif/cl507a.h>
+#include <nvif/cl507b.h>
+#include <nvif/cl507c.h>
+#include <nvif/cl507d.h>
+#include <nvif/cl507e.h>
+#include <nvif/event.h>
+
+#include "nouveau_drv.h"
+#include "nouveau_dma.h"
+#include "nouveau_gem.h"
+#include "nouveau_connector.h"
+#include "nouveau_encoder.h"
+#include "nouveau_crtc.h"
+#include "nouveau_fence.h"
+#include "nouveau_fbcon.h"
+#include "nv50_display.h"
+
+#define EVO_DMA_NR 9
+
+#define EVO_MASTER  (0x00)
+#define EVO_FLIP(c) (0x01 + (c))
+#define EVO_OVLY(c) (0x05 + (c))
+#define EVO_OIMM(c) (0x09 + (c))
+#define EVO_CURS(c) (0x0d + (c))
+
+/* offsets in shared sync bo of various structures */
+#define EVO_SYNC(c, o) ((c) * 0x0100 + (o))
+#define EVO_MAST_NTFY     EVO_SYNC(      0, 0x00)
+#define EVO_FLIP_SEM0(c)  EVO_SYNC((c) + 1, 0x00)
+#define EVO_FLIP_SEM1(c)  EVO_SYNC((c) + 1, 0x10)
+#define EVO_FLIP_NTFY0(c) EVO_SYNC((c) + 1, 0x20)
+#define EVO_FLIP_NTFY1(c) EVO_SYNC((c) + 1, 0x30)
+
+/******************************************************************************
+ * Atomic state
+ *****************************************************************************/
+#define nv50_atom(p) container_of((p), struct nv50_atom, state)
+
+struct nv50_atom {
+	struct drm_atomic_state state;
+
+	struct list_head outp;
+	bool lock_core;
+	bool flush_disable;
+};
+
+struct nv50_outp_atom {
+	struct list_head head;
+
+	struct drm_encoder *encoder;
+	bool flush_disable;
+
+	union {
+		struct {
+			bool ctrl:1;
+		};
+		u8 mask;
+	} clr;
+
+	union {
+		struct {
+			bool ctrl:1;
+		};
+		u8 mask;
+	} set;
+};
+
+#define nv50_head_atom(p) container_of((p), struct nv50_head_atom, state)
+
+struct nv50_head_atom {
+	struct drm_crtc_state state;
+
+	struct {
+		u16 iW;
+		u16 iH;
+		u16 oW;
+		u16 oH;
+	} view;
+
+	struct nv50_head_mode {
+		bool interlace;
+		u32 clock;
+		struct {
+			u16 active;
+			u16 synce;
+			u16 blanke;
+			u16 blanks;
+		} h;
+		struct {
+			u32 active;
+			u16 synce;
+			u16 blanke;
+			u16 blanks;
+			u16 blank2s;
+			u16 blank2e;
+			u16 blankus;
+		} v;
+	} mode;
+
+	struct {
+		bool visible;
+		u32 handle;
+		u64 offset:40;
+		u8  mode:4;
+	} lut;
+
+	struct {
+		bool visible;
+		u32 handle;
+		u64 offset:40;
+		u8  format;
+		u8  kind:7;
+		u8  layout:1;
+		u8  block:4;
+		u32 pitch:20;
+		u16 x;
+		u16 y;
+		u16 w;
+		u16 h;
+	} core;
+
+	struct {
+		bool visible;
+		u32 handle;
+		u64 offset:40;
+		u8  layout:1;
+		u8  format:1;
+	} curs;
+
+	struct {
+		u8  depth;
+		u8  cpp;
+		u16 x;
+		u16 y;
+		u16 w;
+		u16 h;
+	} base;
+
+	struct {
+		u8 cpp;
+	} ovly;
+
+	struct {
+		bool enable:1;
+		u8 bits:2;
+		u8 mode:4;
+	} dither;
+
+	struct {
+		struct {
+			u16 cos:12;
+			u16 sin:12;
+		} sat;
+	} procamp;
+
+	union {
+		struct {
+			bool ilut:1;
+			bool core:1;
+			bool curs:1;
+		};
+		u8 mask;
+	} clr;
+
+	union {
+		struct {
+			bool ilut:1;
+			bool core:1;
+			bool curs:1;
+			bool view:1;
+			bool mode:1;
+			bool base:1;
+			bool ovly:1;
+			bool dither:1;
+			bool procamp:1;
+		};
+		u16 mask;
+	} set;
+};
+
+static inline struct nv50_head_atom *
+nv50_head_atom_get(struct drm_atomic_state *state, struct drm_crtc *crtc)
+{
+	struct drm_crtc_state *statec = drm_atomic_get_crtc_state(state, crtc);
+	if (IS_ERR(statec))
+		return (void *)statec;
+	return nv50_head_atom(statec);
+}
+
+#define nv50_wndw_atom(p) container_of((p), struct nv50_wndw_atom, state)
+
+struct nv50_wndw_atom {
+	struct drm_plane_state state;
+	u8 interval;
+
+	struct {
+		u32  handle;
+		u16  offset:12;
+		bool awaken:1;
+	} ntfy;
+
+	struct {
+		u32 handle;
+		u16 offset:12;
+		u32 acquire;
+		u32 release;
+	} sema;
+
+	struct {
+		u8 enable:2;
+	} lut;
+
+	struct {
+		u8  mode:2;
+		u8  interval:4;
+
+		u8  format;
+		u8  kind:7;
+		u8  layout:1;
+		u8  block:4;
+		u32 pitch:20;
+		u16 w;
+		u16 h;
+
+		u32 handle;
+		u64 offset;
+	} image;
+
+	struct {
+		u16 x;
+		u16 y;
+	} point;
+
+	union {
+		struct {
+			bool ntfy:1;
+			bool sema:1;
+			bool image:1;
+		};
+		u8 mask;
+	} clr;
+
+	union {
+		struct {
+			bool ntfy:1;
+			bool sema:1;
+			bool image:1;
+			bool lut:1;
+			bool point:1;
+		};
+		u8 mask;
+	} set;
+};
+
+/******************************************************************************
+ * EVO channel
+ *****************************************************************************/
+
+struct nv50_chan {
+	struct nvif_object user;
+	struct nvif_device *device;
+};
+
+static int
+nv50_chan_create(struct nvif_device *device, struct nvif_object *disp,
+		 const s32 *oclass, u8 head, void *data, u32 size,
+		 struct nv50_chan *chan)
+{
+	struct nvif_sclass *sclass;
+	int ret, i, n;
+
+	chan->device = device;
+
+	ret = n = nvif_object_sclass_get(disp, &sclass);
+	if (ret < 0)
+		return ret;
+
+	while (oclass[0]) {
+		for (i = 0; i < n; i++) {
+			if (sclass[i].oclass == oclass[0]) {
+				ret = nvif_object_init(disp, 0, oclass[0],
+						       data, size, &chan->user);
+				if (ret == 0)
+					nvif_object_map(&chan->user, NULL, 0);
+				nvif_object_sclass_put(&sclass);
+				return ret;
+			}
+		}
+		oclass++;
+	}
+
+	nvif_object_sclass_put(&sclass);
+	return -ENOSYS;
+}
+
+static void
+nv50_chan_destroy(struct nv50_chan *chan)
+{
+	nvif_object_fini(&chan->user);
+}
+
+/******************************************************************************
+ * PIO EVO channel
+ *****************************************************************************/
+
+struct nv50_pioc {
+	struct nv50_chan base;
+};
+
+static void
+nv50_pioc_destroy(struct nv50_pioc *pioc)
+{
+	nv50_chan_destroy(&pioc->base);
+}
+
+static int
+nv50_pioc_create(struct nvif_device *device, struct nvif_object *disp,
+		 const s32 *oclass, u8 head, void *data, u32 size,
+		 struct nv50_pioc *pioc)
+{
+	return nv50_chan_create(device, disp, oclass, head, data, size,
+				&pioc->base);
+}
+
+/******************************************************************************
+ * Overlay Immediate
+ *****************************************************************************/
+
+struct nv50_oimm {
+	struct nv50_pioc base;
+};
+
+static int
+nv50_oimm_create(struct nvif_device *device, struct nvif_object *disp,
+		 int head, struct nv50_oimm *oimm)
+{
+	struct nv50_disp_cursor_v0 args = {
+		.head = head,
+	};
+	static const s32 oclass[] = {
+		GK104_DISP_OVERLAY,
+		GF110_DISP_OVERLAY,
+		GT214_DISP_OVERLAY,
+		G82_DISP_OVERLAY,
+		NV50_DISP_OVERLAY,
+		0
+	};
+
+	return nv50_pioc_create(device, disp, oclass, head, &args, sizeof(args),
+				&oimm->base);
+}
+
+/******************************************************************************
+ * DMA EVO channel
+ *****************************************************************************/
+
+struct nv50_dmac_ctxdma {
+	struct list_head head;
+	struct nvif_object object;
+};
+
+struct nv50_dmac {
+	struct nv50_chan base;
+
+	struct nvif_mem push;
+	u32 *ptr;
+
+	struct nvif_object sync;
+	struct nvif_object vram;
+	struct list_head ctxdma;
+
+	/* Protects against concurrent pushbuf access to this channel, lock is
+	 * grabbed by evo_wait (if the pushbuf reservation is successful) and
+	 * dropped again by evo_kick. */
+	struct mutex lock;
+};
+
+static void
+nv50_dmac_ctxdma_del(struct nv50_dmac_ctxdma *ctxdma)
+{
+	nvif_object_fini(&ctxdma->object);
+	list_del(&ctxdma->head);
+	kfree(ctxdma);
+}
+
+static struct nv50_dmac_ctxdma *
+nv50_dmac_ctxdma_new(struct nv50_dmac *dmac, struct nouveau_framebuffer *fb)
+{
+	struct nouveau_drm *drm = nouveau_drm(fb->base.dev);
+	struct nv50_dmac_ctxdma *ctxdma;
+	const u8    kind = fb->nvbo->kind;
+	const u32 handle = 0xfb000000 | kind;
+	struct {
+		struct nv_dma_v0 base;
+		union {
+			struct nv50_dma_v0 nv50;
+			struct gf100_dma_v0 gf100;
+			struct gf119_dma_v0 gf119;
+		};
+	} args = {};
+	u32 argc = sizeof(args.base);
+	int ret;
+
+	list_for_each_entry(ctxdma, &dmac->ctxdma, head) {
+		if (ctxdma->object.handle == handle)
+			return ctxdma;
+	}
+
+	if (!(ctxdma = kzalloc(sizeof(*ctxdma), GFP_KERNEL)))
+		return ERR_PTR(-ENOMEM);
+	list_add(&ctxdma->head, &dmac->ctxdma);
+
+	args.base.target = NV_DMA_V0_TARGET_VRAM;
+	args.base.access = NV_DMA_V0_ACCESS_RDWR;
+	args.base.start  = 0;
+	args.base.limit  = drm->client.device.info.ram_user - 1;
+
+	if (drm->client.device.info.chipset < 0x80) {
+		args.nv50.part = NV50_DMA_V0_PART_256;
+		argc += sizeof(args.nv50);
+	} else
+	if (drm->client.device.info.chipset < 0xc0) {
+		args.nv50.part = NV50_DMA_V0_PART_256;
+		args.nv50.kind = kind;
+		argc += sizeof(args.nv50);
+	} else
+	if (drm->client.device.info.chipset < 0xd0) {
+		args.gf100.kind = kind;
+		argc += sizeof(args.gf100);
+	} else {
+		args.gf119.page = GF119_DMA_V0_PAGE_LP;
+		args.gf119.kind = kind;
+		argc += sizeof(args.gf119);
+	}
+
+	ret = nvif_object_init(&dmac->base.user, handle, NV_DMA_IN_MEMORY,
+			       &args, argc, &ctxdma->object);
+	if (ret) {
+		nv50_dmac_ctxdma_del(ctxdma);
+		return ERR_PTR(ret);
+	}
+
+	return ctxdma;
+}
+
+static void
+nv50_dmac_destroy(struct nv50_dmac *dmac)
+{
+	struct nv50_dmac_ctxdma *ctxdma, *ctxtmp;
+
+	list_for_each_entry_safe(ctxdma, ctxtmp, &dmac->ctxdma, head) {
+		nv50_dmac_ctxdma_del(ctxdma);
+	}
+
+	nvif_object_fini(&dmac->vram);
+	nvif_object_fini(&dmac->sync);
+
+	nv50_chan_destroy(&dmac->base);
+
+	nvif_mem_fini(&dmac->push);
+}
+
+static int
+nv50_dmac_create(struct nvif_device *device, struct nvif_object *disp,
+		 const s32 *oclass, u8 head, void *data, u32 size, u64 syncbuf,
+		 struct nv50_dmac *dmac)
+{
+	struct nouveau_cli *cli = (void *)device->object.client;
+	struct nv50_disp_core_channel_dma_v0 *args = data;
+	int ret;
+
+	mutex_init(&dmac->lock);
+	INIT_LIST_HEAD(&dmac->ctxdma);
+
+	ret = nvif_mem_init_map(&cli->mmu, NVIF_MEM_COHERENT, 0x1000,
+				&dmac->push);
+	if (ret)
+		return ret;
+
+	dmac->ptr = dmac->push.object.map.ptr;
+
+	args->pushbuf = nvif_handle(&dmac->push.object);
+
+	ret = nv50_chan_create(device, disp, oclass, head, data, size,
+			       &dmac->base);
+	if (ret)
+		return ret;
+
+	ret = nvif_object_init(&dmac->base.user, 0xf0000000, NV_DMA_IN_MEMORY,
+			       &(struct nv_dma_v0) {
+					.target = NV_DMA_V0_TARGET_VRAM,
+					.access = NV_DMA_V0_ACCESS_RDWR,
+					.start = syncbuf + 0x0000,
+					.limit = syncbuf + 0x0fff,
+			       }, sizeof(struct nv_dma_v0),
+			       &dmac->sync);
+	if (ret)
+		return ret;
+
+	ret = nvif_object_init(&dmac->base.user, 0xf0000001, NV_DMA_IN_MEMORY,
+			       &(struct nv_dma_v0) {
+					.target = NV_DMA_V0_TARGET_VRAM,
+					.access = NV_DMA_V0_ACCESS_RDWR,
+					.start = 0,
+					.limit = device->info.ram_user - 1,
+			       }, sizeof(struct nv_dma_v0),
+			       &dmac->vram);
+	if (ret)
+		return ret;
+
+	return ret;
+}
+
+/******************************************************************************
+ * Core
+ *****************************************************************************/
+
+struct nv50_mast {
+	struct nv50_dmac base;
+};
+
+static int
+nv50_core_create(struct nvif_device *device, struct nvif_object *disp,
+		 u64 syncbuf, struct nv50_mast *core)
+{
+	struct nv50_disp_core_channel_dma_v0 args = {};
+	static const s32 oclass[] = {
+		GP102_DISP_CORE_CHANNEL_DMA,
+		GP100_DISP_CORE_CHANNEL_DMA,
+		GM200_DISP_CORE_CHANNEL_DMA,
+		GM107_DISP_CORE_CHANNEL_DMA,
+		GK110_DISP_CORE_CHANNEL_DMA,
+		GK104_DISP_CORE_CHANNEL_DMA,
+		GF110_DISP_CORE_CHANNEL_DMA,
+		GT214_DISP_CORE_CHANNEL_DMA,
+		GT206_DISP_CORE_CHANNEL_DMA,
+		GT200_DISP_CORE_CHANNEL_DMA,
+		G82_DISP_CORE_CHANNEL_DMA,
+		NV50_DISP_CORE_CHANNEL_DMA,
+		0
+	};
+
+	return nv50_dmac_create(device, disp, oclass, 0, &args, sizeof(args),
+				syncbuf, &core->base);
+}
+
+/******************************************************************************
+ * Base
+ *****************************************************************************/
+
+struct nv50_sync {
+	struct nv50_dmac base;
+	u32 addr;
+	u32 data;
+};
+
+static int
+nv50_base_create(struct nvif_device *device, struct nvif_object *disp,
+		 int head, u64 syncbuf, struct nv50_sync *base)
+{
+	struct nv50_disp_base_channel_dma_v0 args = {
+		.head = head,
+	};
+	static const s32 oclass[] = {
+		GK110_DISP_BASE_CHANNEL_DMA,
+		GK104_DISP_BASE_CHANNEL_DMA,
+		GF110_DISP_BASE_CHANNEL_DMA,
+		GT214_DISP_BASE_CHANNEL_DMA,
+		GT200_DISP_BASE_CHANNEL_DMA,
+		G82_DISP_BASE_CHANNEL_DMA,
+		NV50_DISP_BASE_CHANNEL_DMA,
+		0
+	};
+
+	return nv50_dmac_create(device, disp, oclass, head, &args, sizeof(args),
+				syncbuf, &base->base);
+}
+
+/******************************************************************************
+ * Overlay
+ *****************************************************************************/
+
+struct nv50_ovly {
+	struct nv50_dmac base;
+};
+
+static int
+nv50_ovly_create(struct nvif_device *device, struct nvif_object *disp,
+		 int head, u64 syncbuf, struct nv50_ovly *ovly)
+{
+	struct nv50_disp_overlay_channel_dma_v0 args = {
+		.head = head,
+	};
+	static const s32 oclass[] = {
+		GK104_DISP_OVERLAY_CONTROL_DMA,
+		GF110_DISP_OVERLAY_CONTROL_DMA,
+		GT214_DISP_OVERLAY_CHANNEL_DMA,
+		GT200_DISP_OVERLAY_CHANNEL_DMA,
+		G82_DISP_OVERLAY_CHANNEL_DMA,
+		NV50_DISP_OVERLAY_CHANNEL_DMA,
+		0
+	};
+
+	return nv50_dmac_create(device, disp, oclass, head, &args, sizeof(args),
+				syncbuf, &ovly->base);
+}
+
+struct nv50_head {
+	struct nouveau_crtc base;
+	struct {
+		struct nouveau_bo *nvbo[2];
+		int next;
+	} lut;
+	struct nv50_ovly ovly;
+	struct nv50_oimm oimm;
+};
+
+#define nv50_head(c) ((struct nv50_head *)nouveau_crtc(c))
+#define nv50_ovly(c) (&nv50_head(c)->ovly)
+#define nv50_oimm(c) (&nv50_head(c)->oimm)
+#define nv50_chan(c) (&(c)->base.base)
+#define nv50_vers(c) nv50_chan(c)->user.oclass
+
+struct nv50_disp {
+	struct nvif_disp *disp;
+	struct nv50_mast mast;
+
+	struct nouveau_bo *sync;
+
+	struct mutex mutex;
+};
+
+static struct nv50_disp *
+nv50_disp(struct drm_device *dev)
+{
+	return nouveau_display(dev)->priv;
+}
+
+#define nv50_mast(d) (&nv50_disp(d)->mast)
+
+/******************************************************************************
+ * EVO channel helpers
+ *****************************************************************************/
+static u32 *
+evo_wait(void *evoc, int nr)
+{
+	struct nv50_dmac *dmac = evoc;
+	struct nvif_device *device = dmac->base.device;
+	u32 put = nvif_rd32(&dmac->base.user, 0x0000) / 4;
+
+	mutex_lock(&dmac->lock);
+	if (put + nr >= (PAGE_SIZE / 4) - 8) {
+		dmac->ptr[put] = 0x20000000;
+
+		nvif_wr32(&dmac->base.user, 0x0000, 0x00000000);
+		if (nvif_msec(device, 2000,
+			if (!nvif_rd32(&dmac->base.user, 0x0004))
+				break;
+		) < 0) {
+			mutex_unlock(&dmac->lock);
+			pr_err("nouveau: evo channel stalled\n");
+			return NULL;
+		}
+
+		put = 0;
+	}
+
+	return dmac->ptr + put;
+}
+
+static void
+evo_kick(u32 *push, void *evoc)
+{
+	struct nv50_dmac *dmac = evoc;
+	nvif_wr32(&dmac->base.user, 0x0000, (push - dmac->ptr) << 2);
+	mutex_unlock(&dmac->lock);
+}
+
+#define evo_mthd(p, m, s) do {						\
+	const u32 _m = (m), _s = (s);					\
+	if (drm_debug & DRM_UT_KMS)					\
+		pr_err("%04x %d %s\n", _m, _s, __func__);		\
+	*((p)++) = ((_s << 18) | _m);					\
+} while(0)
+
+#define evo_data(p, d) do {						\
+	const u32 _d = (d);						\
+	if (drm_debug & DRM_UT_KMS)					\
+		pr_err("\t%08x\n", _d);					\
+	*((p)++) = _d;							\
+} while(0)
+
+/******************************************************************************
+ * Plane
+ *****************************************************************************/
+#define nv50_wndw(p) container_of((p), struct nv50_wndw, plane)
+
+struct nv50_wndw {
+	const struct nv50_wndw_func *func;
+	struct nv50_dmac *dmac;
+
+	struct drm_plane plane;
+
+	struct nvif_notify notify;
+	u16 ntfy;
+	u16 sema;
+	u32 data;
+};
+
+struct nv50_wndw_func {
+	void *(*dtor)(struct nv50_wndw *);
+	int (*acquire)(struct nv50_wndw *, struct nv50_wndw_atom *asyw,
+		       struct nv50_head_atom *asyh);
+	void (*release)(struct nv50_wndw *, struct nv50_wndw_atom *asyw,
+			struct nv50_head_atom *asyh);
+	void (*prepare)(struct nv50_wndw *, struct nv50_head_atom *asyh,
+			struct nv50_wndw_atom *asyw);
+
+	void (*sema_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
+	void (*sema_clr)(struct nv50_wndw *);
+	void (*ntfy_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
+	void (*ntfy_clr)(struct nv50_wndw *);
+	int (*ntfy_wait_begun)(struct nv50_wndw *, struct nv50_wndw_atom *);
+	void (*image_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
+	void (*image_clr)(struct nv50_wndw *);
+	void (*lut)(struct nv50_wndw *, struct nv50_wndw_atom *);
+	void (*point)(struct nv50_wndw *, struct nv50_wndw_atom *);
+
+	u32 (*update)(struct nv50_wndw *, u32 interlock);
+};
+
+static int
+nv50_wndw_wait_armed(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	if (asyw->set.ntfy)
+		return wndw->func->ntfy_wait_begun(wndw, asyw);
+	return 0;
+}
+
+static u32
+nv50_wndw_flush_clr(struct nv50_wndw *wndw, u32 interlock, bool flush,
+		    struct nv50_wndw_atom *asyw)
+{
+	if (asyw->clr.sema && (!asyw->set.sema || flush))
+		wndw->func->sema_clr(wndw);
+	if (asyw->clr.ntfy && (!asyw->set.ntfy || flush))
+		wndw->func->ntfy_clr(wndw);
+	if (asyw->clr.image && (!asyw->set.image || flush))
+		wndw->func->image_clr(wndw);
+
+	return flush ? wndw->func->update(wndw, interlock) : 0;
+}
+
+static u32
+nv50_wndw_flush_set(struct nv50_wndw *wndw, u32 interlock,
+		    struct nv50_wndw_atom *asyw)
+{
+	if (interlock) {
+		asyw->image.mode = 0;
+		asyw->image.interval = 1;
+	}
+
+	if (asyw->set.sema ) wndw->func->sema_set (wndw, asyw);
+	if (asyw->set.ntfy ) wndw->func->ntfy_set (wndw, asyw);
+	if (asyw->set.image) wndw->func->image_set(wndw, asyw);
+	if (asyw->set.lut  ) wndw->func->lut      (wndw, asyw);
+	if (asyw->set.point) wndw->func->point    (wndw, asyw);
+
+	return wndw->func->update(wndw, interlock);
+}
+
+static void
+nv50_wndw_atomic_check_release(struct nv50_wndw *wndw,
+			       struct nv50_wndw_atom *asyw,
+			       struct nv50_head_atom *asyh)
+{
+	struct nouveau_drm *drm = nouveau_drm(wndw->plane.dev);
+	NV_ATOMIC(drm, "%s release\n", wndw->plane.name);
+	wndw->func->release(wndw, asyw, asyh);
+	asyw->ntfy.handle = 0;
+	asyw->sema.handle = 0;
+}
+
+static int
+nv50_wndw_atomic_check_acquire(struct nv50_wndw *wndw,
+			       struct nv50_wndw_atom *asyw,
+			       struct nv50_head_atom *asyh)
+{
+	struct nouveau_framebuffer *fb = nouveau_framebuffer(asyw->state.fb);
+	struct nouveau_drm *drm = nouveau_drm(wndw->plane.dev);
+	int ret;
+
+	NV_ATOMIC(drm, "%s acquire\n", wndw->plane.name);
+
+	asyw->image.w = fb->base.width;
+	asyw->image.h = fb->base.height;
+	asyw->image.kind = fb->nvbo->kind;
+
+	if (asyh->state.pageflip_flags & DRM_MODE_PAGE_FLIP_ASYNC)
+		asyw->interval = 0;
+	else
+		asyw->interval = 1;
+
+	if (asyw->image.kind) {
+		asyw->image.layout = 0;
+		if (drm->client.device.info.chipset >= 0xc0)
+			asyw->image.block = fb->nvbo->mode >> 4;
+		else
+			asyw->image.block = fb->nvbo->mode;
+		asyw->image.pitch = (fb->base.pitches[0] / 4) << 4;
+	} else {
+		asyw->image.layout = 1;
+		asyw->image.block  = 0;
+		asyw->image.pitch  = fb->base.pitches[0];
+	}
+
+	ret = wndw->func->acquire(wndw, asyw, asyh);
+	if (ret)
+		return ret;
+
+	if (asyw->set.image) {
+		if (!(asyw->image.mode = asyw->interval ? 0 : 1))
+			asyw->image.interval = asyw->interval;
+		else
+			asyw->image.interval = 0;
+	}
+
+	return 0;
+}
+
+static int
+nv50_wndw_atomic_check(struct drm_plane *plane, struct drm_plane_state *state)
+{
+	struct nouveau_drm *drm = nouveau_drm(plane->dev);
+	struct nv50_wndw *wndw = nv50_wndw(plane);
+	struct nv50_wndw_atom *armw = nv50_wndw_atom(wndw->plane.state);
+	struct nv50_wndw_atom *asyw = nv50_wndw_atom(state);
+	struct nv50_head_atom *harm = NULL, *asyh = NULL;
+	bool varm = false, asyv = false, asym = false;
+	int ret;
+
+	NV_ATOMIC(drm, "%s atomic_check\n", plane->name);
+	if (asyw->state.crtc) {
+		asyh = nv50_head_atom_get(asyw->state.state, asyw->state.crtc);
+		if (IS_ERR(asyh))
+			return PTR_ERR(asyh);
+		asym = drm_atomic_crtc_needs_modeset(&asyh->state);
+		asyv = asyh->state.active;
+	}
+
+	if (armw->state.crtc) {
+		harm = nv50_head_atom_get(asyw->state.state, armw->state.crtc);
+		if (IS_ERR(harm))
+			return PTR_ERR(harm);
+		varm = harm->state.crtc->state->active;
+	}
+
+	if (asyv) {
+		asyw->point.x = asyw->state.crtc_x;
+		asyw->point.y = asyw->state.crtc_y;
+		if (memcmp(&armw->point, &asyw->point, sizeof(asyw->point)))
+			asyw->set.point = true;
+
+		ret = nv50_wndw_atomic_check_acquire(wndw, asyw, asyh);
+		if (ret)
+			return ret;
+	} else
+	if (varm) {
+		nv50_wndw_atomic_check_release(wndw, asyw, harm);
+	} else {
+		return 0;
+	}
+
+	if (!asyv || asym) {
+		asyw->clr.ntfy = armw->ntfy.handle != 0;
+		asyw->clr.sema = armw->sema.handle != 0;
+		if (wndw->func->image_clr)
+			asyw->clr.image = armw->image.handle != 0;
+		asyw->set.lut = wndw->func->lut && asyv;
+	}
+
+	return 0;
+}
+
+static void
+nv50_wndw_cleanup_fb(struct drm_plane *plane, struct drm_plane_state *old_state)
+{
+	struct nouveau_framebuffer *fb = nouveau_framebuffer(old_state->fb);
+	struct nouveau_drm *drm = nouveau_drm(plane->dev);
+
+	NV_ATOMIC(drm, "%s cleanup: %p\n", plane->name, old_state->fb);
+	if (!old_state->fb)
+		return;
+
+	nouveau_bo_unpin(fb->nvbo);
+}
+
+static int
+nv50_wndw_prepare_fb(struct drm_plane *plane, struct drm_plane_state *state)
+{
+	struct nouveau_framebuffer *fb = nouveau_framebuffer(state->fb);
+	struct nouveau_drm *drm = nouveau_drm(plane->dev);
+	struct nv50_wndw *wndw = nv50_wndw(plane);
+	struct nv50_wndw_atom *asyw = nv50_wndw_atom(state);
+	struct nv50_head_atom *asyh;
+	struct nv50_dmac_ctxdma *ctxdma;
+	int ret;
+
+	NV_ATOMIC(drm, "%s prepare: %p\n", plane->name, state->fb);
+	if (!asyw->state.fb)
+		return 0;
+
+	ret = nouveau_bo_pin(fb->nvbo, TTM_PL_FLAG_VRAM, true);
+	if (ret)
+		return ret;
+
+	ctxdma = nv50_dmac_ctxdma_new(wndw->dmac, fb);
+	if (IS_ERR(ctxdma)) {
+		nouveau_bo_unpin(fb->nvbo);
+		return PTR_ERR(ctxdma);
+	}
+
+	asyw->state.fence = reservation_object_get_excl_rcu(fb->nvbo->bo.resv);
+	asyw->image.handle = ctxdma->object.handle;
+	asyw->image.offset = fb->nvbo->bo.offset;
+
+	if (wndw->func->prepare) {
+		asyh = nv50_head_atom_get(asyw->state.state, asyw->state.crtc);
+		if (IS_ERR(asyh))
+			return PTR_ERR(asyh);
+
+		wndw->func->prepare(wndw, asyh, asyw);
+	}
+
+	return 0;
+}
+
+static const struct drm_plane_helper_funcs
+nv50_wndw_helper = {
+	.prepare_fb = nv50_wndw_prepare_fb,
+	.cleanup_fb = nv50_wndw_cleanup_fb,
+	.atomic_check = nv50_wndw_atomic_check,
+};
+
+static void
+nv50_wndw_atomic_destroy_state(struct drm_plane *plane,
+			       struct drm_plane_state *state)
+{
+	struct nv50_wndw_atom *asyw = nv50_wndw_atom(state);
+	__drm_atomic_helper_plane_destroy_state(&asyw->state);
+	kfree(asyw);
+}
+
+static struct drm_plane_state *
+nv50_wndw_atomic_duplicate_state(struct drm_plane *plane)
+{
+	struct nv50_wndw_atom *armw = nv50_wndw_atom(plane->state);
+	struct nv50_wndw_atom *asyw;
+	if (!(asyw = kmalloc(sizeof(*asyw), GFP_KERNEL)))
+		return NULL;
+	__drm_atomic_helper_plane_duplicate_state(plane, &asyw->state);
+	asyw->interval = 1;
+	asyw->sema = armw->sema;
+	asyw->ntfy = armw->ntfy;
+	asyw->image = armw->image;
+	asyw->point = armw->point;
+	asyw->lut = armw->lut;
+	asyw->clr.mask = 0;
+	asyw->set.mask = 0;
+	return &asyw->state;
+}
+
+static void
+nv50_wndw_reset(struct drm_plane *plane)
+{
+	struct nv50_wndw_atom *asyw;
+
+	if (WARN_ON(!(asyw = kzalloc(sizeof(*asyw), GFP_KERNEL))))
+		return;
+
+	if (plane->state)
+		plane->funcs->atomic_destroy_state(plane, plane->state);
+	plane->state = &asyw->state;
+	plane->state->plane = plane;
+	plane->state->rotation = DRM_MODE_ROTATE_0;
+}
+
+static void
+nv50_wndw_destroy(struct drm_plane *plane)
+{
+	struct nv50_wndw *wndw = nv50_wndw(plane);
+	void *data;
+	nvif_notify_fini(&wndw->notify);
+	data = wndw->func->dtor(wndw);
+	drm_plane_cleanup(&wndw->plane);
+	kfree(data);
+}
+
+static const struct drm_plane_funcs
+nv50_wndw = {
+	.update_plane = drm_atomic_helper_update_plane,
+	.disable_plane = drm_atomic_helper_disable_plane,
+	.destroy = nv50_wndw_destroy,
+	.reset = nv50_wndw_reset,
+	.atomic_duplicate_state = nv50_wndw_atomic_duplicate_state,
+	.atomic_destroy_state = nv50_wndw_atomic_destroy_state,
+};
+
+static void
+nv50_wndw_fini(struct nv50_wndw *wndw)
+{
+	nvif_notify_put(&wndw->notify);
+}
+
+static void
+nv50_wndw_init(struct nv50_wndw *wndw)
+{
+	nvif_notify_get(&wndw->notify);
+}
+
+static int
+nv50_wndw_ctor(const struct nv50_wndw_func *func, struct drm_device *dev,
+	       enum drm_plane_type type, const char *name, int index,
+	       struct nv50_dmac *dmac, const u32 *format, int nformat,
+	       struct nv50_wndw *wndw)
+{
+	int ret;
+
+	wndw->func = func;
+	wndw->dmac = dmac;
+
+	ret = drm_universal_plane_init(dev, &wndw->plane, 0, &nv50_wndw,
+				       format, nformat, NULL,
+				       type, "%s-%d", name, index);
+	if (ret)
+		return ret;
+
+	drm_plane_helper_add(&wndw->plane, &nv50_wndw_helper);
+	return 0;
+}
+
+/******************************************************************************
+ * Cursor plane
+ *****************************************************************************/
+#define nv50_curs(p) container_of((p), struct nv50_curs, wndw)
+
+struct nv50_curs {
+	struct nv50_wndw wndw;
+	struct nvif_object chan;
+};
+
+static u32
+nv50_curs_update(struct nv50_wndw *wndw, u32 interlock)
+{
+	struct nv50_curs *curs = nv50_curs(wndw);
+	nvif_wr32(&curs->chan, 0x0080, 0x00000000);
+	return 0;
+}
+
+static void
+nv50_curs_point(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	struct nv50_curs *curs = nv50_curs(wndw);
+	nvif_wr32(&curs->chan, 0x0084, (asyw->point.y << 16) | asyw->point.x);
+}
+
+static void
+nv50_curs_prepare(struct nv50_wndw *wndw, struct nv50_head_atom *asyh,
+		  struct nv50_wndw_atom *asyw)
+{
+	u32 handle = nv50_disp(wndw->plane.dev)->mast.base.vram.handle;
+	u32 offset = asyw->image.offset;
+	if (asyh->curs.handle != handle || asyh->curs.offset != offset) {
+		asyh->curs.handle = handle;
+		asyh->curs.offset = offset;
+		asyh->set.curs = asyh->curs.visible;
+	}
+}
+
+static void
+nv50_curs_release(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
+		  struct nv50_head_atom *asyh)
+{
+	asyh->curs.visible = false;
+}
+
+static int
+nv50_curs_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
+		  struct nv50_head_atom *asyh)
+{
+	int ret;
+
+	ret = drm_atomic_helper_check_plane_state(&asyw->state, &asyh->state,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  true, true);
+	asyh->curs.visible = asyw->state.visible;
+	if (ret || !asyh->curs.visible)
+		return ret;
+
+	switch (asyw->state.fb->width) {
+	case 32: asyh->curs.layout = 0; break;
+	case 64: asyh->curs.layout = 1; break;
+	default:
+		return -EINVAL;
+	}
+
+	if (asyw->state.fb->width != asyw->state.fb->height)
+		return -EINVAL;
+
+	switch (asyw->state.fb->format->format) {
+	case DRM_FORMAT_ARGB8888: asyh->curs.format = 1; break;
+	default:
+		WARN_ON(1);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static void *
+nv50_curs_dtor(struct nv50_wndw *wndw)
+{
+	struct nv50_curs *curs = nv50_curs(wndw);
+	nvif_object_fini(&curs->chan);
+	return curs;
+}
+
+static const u32
+nv50_curs_format[] = {
+	DRM_FORMAT_ARGB8888,
+};
+
+static const struct nv50_wndw_func
+nv50_curs = {
+	.dtor = nv50_curs_dtor,
+	.acquire = nv50_curs_acquire,
+	.release = nv50_curs_release,
+	.prepare = nv50_curs_prepare,
+	.point = nv50_curs_point,
+	.update = nv50_curs_update,
+};
+
+static int
+nv50_curs_new(struct nouveau_drm *drm, struct nv50_head *head,
+	      struct nv50_curs **pcurs)
+{
+	static const struct nvif_mclass curses[] = {
+		{ GK104_DISP_CURSOR, 0 },
+		{ GF110_DISP_CURSOR, 0 },
+		{ GT214_DISP_CURSOR, 0 },
+		{   G82_DISP_CURSOR, 0 },
+		{  NV50_DISP_CURSOR, 0 },
+		{}
+	};
+	struct nv50_disp_cursor_v0 args = {
+		.head = head->base.index,
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	struct nv50_curs *curs;
+	int cid, ret;
+
+	cid = nvif_mclass(&disp->disp->object, curses);
+	if (cid < 0) {
+		NV_ERROR(drm, "No supported cursor immediate class\n");
+		return cid;
+	}
+
+	if (!(curs = *pcurs = kzalloc(sizeof(*curs), GFP_KERNEL)))
+		return -ENOMEM;
+
+	ret = nv50_wndw_ctor(&nv50_curs, drm->dev, DRM_PLANE_TYPE_CURSOR,
+			     "curs", head->base.index, &disp->mast.base,
+			     nv50_curs_format, ARRAY_SIZE(nv50_curs_format),
+			     &curs->wndw);
+	if (ret) {
+		kfree(curs);
+		return ret;
+	}
+
+	ret = nvif_object_init(&disp->disp->object, 0, curses[cid].oclass,
+			       &args, sizeof(args), &curs->chan);
+	if (ret) {
+		NV_ERROR(drm, "curs%04x allocation failed: %d\n",
+			 curses[cid].oclass, ret);
+		return ret;
+	}
+
+	return 0;
+}
+
+/******************************************************************************
+ * Primary plane
+ *****************************************************************************/
+#define nv50_base(p) container_of((p), struct nv50_base, wndw)
+
+struct nv50_base {
+	struct nv50_wndw wndw;
+	struct nv50_sync chan;
+	int id;
+};
+
+static int
+nv50_base_notify(struct nvif_notify *notify)
+{
+	return NVIF_NOTIFY_KEEP;
+}
+
+static void
+nv50_base_lut(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	struct nv50_base *base = nv50_base(wndw);
+	u32 *push;
+	if ((push = evo_wait(&base->chan, 2))) {
+		evo_mthd(push, 0x00e0, 1);
+		evo_data(push, asyw->lut.enable << 30);
+		evo_kick(push, &base->chan);
+	}
+}
+
+static void
+nv50_base_image_clr(struct nv50_wndw *wndw)
+{
+	struct nv50_base *base = nv50_base(wndw);
+	u32 *push;
+	if ((push = evo_wait(&base->chan, 4))) {
+		evo_mthd(push, 0x0084, 1);
+		evo_data(push, 0x00000000);
+		evo_mthd(push, 0x00c0, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &base->chan);
+	}
+}
+
+static void
+nv50_base_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	struct nv50_base *base = nv50_base(wndw);
+	const s32 oclass = base->chan.base.base.user.oclass;
+	u32 *push;
+	if ((push = evo_wait(&base->chan, 10))) {
+		evo_mthd(push, 0x0084, 1);
+		evo_data(push, (asyw->image.mode << 8) |
+			       (asyw->image.interval << 4));
+		evo_mthd(push, 0x00c0, 1);
+		evo_data(push, asyw->image.handle);
+		if (oclass < G82_DISP_BASE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0800, 5);
+			evo_data(push, asyw->image.offset >> 8);
+			evo_data(push, 0x00000000);
+			evo_data(push, (asyw->image.h << 16) | asyw->image.w);
+			evo_data(push, (asyw->image.layout << 20) |
+					asyw->image.pitch |
+					asyw->image.block);
+			evo_data(push, (asyw->image.kind << 16) |
+				       (asyw->image.format << 8));
+		} else
+		if (oclass < GF110_DISP_BASE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0800, 5);
+			evo_data(push, asyw->image.offset >> 8);
+			evo_data(push, 0x00000000);
+			evo_data(push, (asyw->image.h << 16) | asyw->image.w);
+			evo_data(push, (asyw->image.layout << 20) |
+					asyw->image.pitch |
+					asyw->image.block);
+			evo_data(push, asyw->image.format << 8);
+		} else {
+			evo_mthd(push, 0x0400, 5);
+			evo_data(push, asyw->image.offset >> 8);
+			evo_data(push, 0x00000000);
+			evo_data(push, (asyw->image.h << 16) | asyw->image.w);
+			evo_data(push, (asyw->image.layout << 24) |
+					asyw->image.pitch |
+					asyw->image.block);
+			evo_data(push, asyw->image.format << 8);
+		}
+		evo_kick(push, &base->chan);
+	}
+}
+
+static void
+nv50_base_ntfy_clr(struct nv50_wndw *wndw)
+{
+	struct nv50_base *base = nv50_base(wndw);
+	u32 *push;
+	if ((push = evo_wait(&base->chan, 2))) {
+		evo_mthd(push, 0x00a4, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &base->chan);
+	}
+}
+
+static void
+nv50_base_ntfy_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	struct nv50_base *base = nv50_base(wndw);
+	u32 *push;
+	if ((push = evo_wait(&base->chan, 3))) {
+		evo_mthd(push, 0x00a0, 2);
+		evo_data(push, (asyw->ntfy.awaken << 30) | asyw->ntfy.offset);
+		evo_data(push, asyw->ntfy.handle);
+		evo_kick(push, &base->chan);
+	}
+}
+
+static void
+nv50_base_sema_clr(struct nv50_wndw *wndw)
+{
+	struct nv50_base *base = nv50_base(wndw);
+	u32 *push;
+	if ((push = evo_wait(&base->chan, 2))) {
+		evo_mthd(push, 0x0094, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &base->chan);
+	}
+}
+
+static void
+nv50_base_sema_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	struct nv50_base *base = nv50_base(wndw);
+	u32 *push;
+	if ((push = evo_wait(&base->chan, 5))) {
+		evo_mthd(push, 0x0088, 4);
+		evo_data(push, asyw->sema.offset);
+		evo_data(push, asyw->sema.acquire);
+		evo_data(push, asyw->sema.release);
+		evo_data(push, asyw->sema.handle);
+		evo_kick(push, &base->chan);
+	}
+}
+
+static u32
+nv50_base_update(struct nv50_wndw *wndw, u32 interlock)
+{
+	struct nv50_base *base = nv50_base(wndw);
+	u32 *push;
+
+	if (!(push = evo_wait(&base->chan, 2)))
+		return 0;
+	evo_mthd(push, 0x0080, 1);
+	evo_data(push, interlock);
+	evo_kick(push, &base->chan);
+
+	if (base->chan.base.base.user.oclass < GF110_DISP_BASE_CHANNEL_DMA)
+		return interlock ? 2 << (base->id * 8) : 0;
+	return interlock ? 2 << (base->id * 4) : 0;
+}
+
+static int
+nv50_base_ntfy_wait_begun(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	struct nouveau_drm *drm = nouveau_drm(wndw->plane.dev);
+	struct nv50_disp *disp = nv50_disp(wndw->plane.dev);
+	if (nvif_msec(&drm->client.device, 2000ULL,
+		u32 data = nouveau_bo_rd32(disp->sync, asyw->ntfy.offset / 4);
+		if ((data & 0xc0000000) == 0x40000000)
+			break;
+		usleep_range(1, 2);
+	) < 0)
+		return -ETIMEDOUT;
+	return 0;
+}
+
+static void
+nv50_base_release(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
+		  struct nv50_head_atom *asyh)
+{
+	asyh->base.cpp = 0;
+}
+
+static int
+nv50_base_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
+		  struct nv50_head_atom *asyh)
+{
+	const struct drm_framebuffer *fb = asyw->state.fb;
+	int ret;
+
+	if (!fb->format->depth)
+		return -EINVAL;
+
+	ret = drm_atomic_helper_check_plane_state(&asyw->state, &asyh->state,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  false, true);
+	if (ret)
+		return ret;
+
+	asyh->base.depth = fb->format->depth;
+	asyh->base.cpp = fb->format->cpp[0];
+	asyh->base.x = asyw->state.src.x1 >> 16;
+	asyh->base.y = asyw->state.src.y1 >> 16;
+	asyh->base.w = asyw->state.fb->width;
+	asyh->base.h = asyw->state.fb->height;
+
+	switch (fb->format->format) {
+	case DRM_FORMAT_C8         : asyw->image.format = 0x1e; break;
+	case DRM_FORMAT_RGB565     : asyw->image.format = 0xe8; break;
+	case DRM_FORMAT_XRGB1555   :
+	case DRM_FORMAT_ARGB1555   : asyw->image.format = 0xe9; break;
+	case DRM_FORMAT_XRGB8888   :
+	case DRM_FORMAT_ARGB8888   : asyw->image.format = 0xcf; break;
+	case DRM_FORMAT_XBGR2101010:
+	case DRM_FORMAT_ABGR2101010: asyw->image.format = 0xd1; break;
+	case DRM_FORMAT_XBGR8888   :
+	case DRM_FORMAT_ABGR8888   : asyw->image.format = 0xd5; break;
+	default:
+		WARN_ON(1);
+		return -EINVAL;
+	}
+
+	asyw->lut.enable = 1;
+	asyw->set.image = true;
+	return 0;
+}
+
+static void *
+nv50_base_dtor(struct nv50_wndw *wndw)
+{
+	struct nv50_base *base = nv50_base(wndw);
+	nv50_dmac_destroy(&base->chan.base);
+	return base;
+}
+
+static const u32
+nv50_base_format[] = {
+	DRM_FORMAT_C8,
+	DRM_FORMAT_RGB565,
+	DRM_FORMAT_XRGB1555,
+	DRM_FORMAT_ARGB1555,
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_XBGR2101010,
+	DRM_FORMAT_ABGR2101010,
+	DRM_FORMAT_XBGR8888,
+	DRM_FORMAT_ABGR8888,
+};
+
+static const struct nv50_wndw_func
+nv50_base = {
+	.dtor = nv50_base_dtor,
+	.acquire = nv50_base_acquire,
+	.release = nv50_base_release,
+	.sema_set = nv50_base_sema_set,
+	.sema_clr = nv50_base_sema_clr,
+	.ntfy_set = nv50_base_ntfy_set,
+	.ntfy_clr = nv50_base_ntfy_clr,
+	.ntfy_wait_begun = nv50_base_ntfy_wait_begun,
+	.image_set = nv50_base_image_set,
+	.image_clr = nv50_base_image_clr,
+	.lut = nv50_base_lut,
+	.update = nv50_base_update,
+};
+
+static int
+nv50_base_new(struct nouveau_drm *drm, struct nv50_head *head,
+	      struct nv50_base **pbase)
+{
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	struct nv50_base *base;
+	int ret;
+
+	if (!(base = *pbase = kzalloc(sizeof(*base), GFP_KERNEL)))
+		return -ENOMEM;
+	base->id = head->base.index;
+	base->wndw.ntfy = EVO_FLIP_NTFY0(base->id);
+	base->wndw.sema = EVO_FLIP_SEM0(base->id);
+	base->wndw.data = 0x00000000;
+
+	ret = nv50_wndw_ctor(&nv50_base, drm->dev, DRM_PLANE_TYPE_PRIMARY,
+			     "base", base->id, &base->chan.base,
+			     nv50_base_format, ARRAY_SIZE(nv50_base_format),
+			     &base->wndw);
+	if (ret) {
+		kfree(base);
+		return ret;
+	}
+
+	ret = nv50_base_create(&drm->client.device, &disp->disp->object,
+			       base->id, disp->sync->bo.offset, &base->chan);
+	if (ret)
+		return ret;
+
+	return nvif_notify_init(&base->chan.base.base.user, nv50_base_notify,
+				false,
+				NV50_DISP_BASE_CHANNEL_DMA_V0_NTFY_UEVENT,
+				&(struct nvif_notify_uevent_req) {},
+				sizeof(struct nvif_notify_uevent_req),
+				sizeof(struct nvif_notify_uevent_rep),
+				&base->wndw.notify);
+}
+
+/******************************************************************************
+ * Head
+ *****************************************************************************/
+static void
+nv50_head_procamp(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
+			evo_mthd(push, 0x08a8 + (head->base.index * 0x400), 1);
+		else
+			evo_mthd(push, 0x0498 + (head->base.index * 0x300), 1);
+		evo_data(push, (asyh->procamp.sat.sin << 20) |
+			       (asyh->procamp.sat.cos << 8));
+		evo_kick(push, core);
+	}
+}
+
+static void
+nv50_head_dither(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
+			evo_mthd(push, 0x08a0 + (head->base.index * 0x0400), 1);
+		else
+		if (core->base.user.oclass < GK104_DISP_CORE_CHANNEL_DMA)
+			evo_mthd(push, 0x0490 + (head->base.index * 0x0300), 1);
+		else
+			evo_mthd(push, 0x04a0 + (head->base.index * 0x0300), 1);
+		evo_data(push, (asyh->dither.mode << 3) |
+			       (asyh->dither.bits << 1) |
+			        asyh->dither.enable);
+		evo_kick(push, core);
+	}
+}
+
+static void
+nv50_head_ovly(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
+	u32 bounds = 0;
+	u32 *push;
+
+	if (asyh->base.cpp) {
+		switch (asyh->base.cpp) {
+		case 8: bounds |= 0x00000500; break;
+		case 4: bounds |= 0x00000300; break;
+		case 2: bounds |= 0x00000100; break;
+		default:
+			WARN_ON(1);
+			break;
+		}
+		bounds |= 0x00000001;
+	}
+
+	if ((push = evo_wait(core, 2))) {
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
+			evo_mthd(push, 0x0904 + head->base.index * 0x400, 1);
+		else
+			evo_mthd(push, 0x04d4 + head->base.index * 0x300, 1);
+		evo_data(push, bounds);
+		evo_kick(push, core);
+	}
+}
+
+static void
+nv50_head_base(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
+	u32 bounds = 0;
+	u32 *push;
+
+	if (asyh->base.cpp) {
+		switch (asyh->base.cpp) {
+		case 8: bounds |= 0x00000500; break;
+		case 4: bounds |= 0x00000300; break;
+		case 2: bounds |= 0x00000100; break;
+		case 1: bounds |= 0x00000000; break;
+		default:
+			WARN_ON(1);
+			break;
+		}
+		bounds |= 0x00000001;
+	}
+
+	if ((push = evo_wait(core, 2))) {
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
+			evo_mthd(push, 0x0900 + head->base.index * 0x400, 1);
+		else
+			evo_mthd(push, 0x04d0 + head->base.index * 0x300, 1);
+		evo_data(push, bounds);
+		evo_kick(push, core);
+	}
+}
+
+static void
+nv50_head_curs_clr(struct nv50_head *head)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
+	u32 *push;
+	if ((push = evo_wait(core, 4))) {
+		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0880 + head->base.index * 0x400, 1);
+			evo_data(push, 0x05000000);
+		} else
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0880 + head->base.index * 0x400, 1);
+			evo_data(push, 0x05000000);
+			evo_mthd(push, 0x089c + head->base.index * 0x400, 1);
+			evo_data(push, 0x00000000);
+		} else {
+			evo_mthd(push, 0x0480 + head->base.index * 0x300, 1);
+			evo_data(push, 0x05000000);
+			evo_mthd(push, 0x048c + head->base.index * 0x300, 1);
+			evo_data(push, 0x00000000);
+		}
+		evo_kick(push, core);
+	}
+}
+
+static void
+nv50_head_curs_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
+	u32 *push;
+	if ((push = evo_wait(core, 5))) {
+		if (core->base.user.oclass < G82_DISP_BASE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0880 + head->base.index * 0x400, 2);
+			evo_data(push, 0x80000000 | (asyh->curs.layout << 26) |
+						    (asyh->curs.format << 24));
+			evo_data(push, asyh->curs.offset >> 8);
+		} else
+		if (core->base.user.oclass < GF110_DISP_BASE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0880 + head->base.index * 0x400, 2);
+			evo_data(push, 0x80000000 | (asyh->curs.layout << 26) |
+						    (asyh->curs.format << 24));
+			evo_data(push, asyh->curs.offset >> 8);
+			evo_mthd(push, 0x089c + head->base.index * 0x400, 1);
+			evo_data(push, asyh->curs.handle);
+		} else {
+			evo_mthd(push, 0x0480 + head->base.index * 0x300, 2);
+			evo_data(push, 0x80000000 | (asyh->curs.layout << 26) |
+						    (asyh->curs.format << 24));
+			evo_data(push, asyh->curs.offset >> 8);
+			evo_mthd(push, 0x048c + head->base.index * 0x300, 1);
+			evo_data(push, asyh->curs.handle);
+		}
+		evo_kick(push, core);
+	}
+}
+
+static void
+nv50_head_core_clr(struct nv50_head *head)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
+			evo_mthd(push, 0x0874 + head->base.index * 0x400, 1);
+		else
+			evo_mthd(push, 0x0474 + head->base.index * 0x300, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, core);
+	}
+}
+
+static void
+nv50_head_core_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
+	u32 *push;
+	if ((push = evo_wait(core, 9))) {
+		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0860 + head->base.index * 0x400, 1);
+			evo_data(push, asyh->core.offset >> 8);
+			evo_mthd(push, 0x0868 + head->base.index * 0x400, 4);
+			evo_data(push, (asyh->core.h << 16) | asyh->core.w);
+			evo_data(push, asyh->core.layout << 20 |
+				       (asyh->core.pitch >> 8) << 8 |
+				       asyh->core.block);
+			evo_data(push, asyh->core.kind << 16 |
+				       asyh->core.format << 8);
+			evo_data(push, asyh->core.handle);
+			evo_mthd(push, 0x08c0 + head->base.index * 0x400, 1);
+			evo_data(push, (asyh->core.y << 16) | asyh->core.x);
+			/* EVO will complain with INVALID_STATE if we have an
+			 * active cursor and (re)specify HeadSetContextDmaIso
+			 * without also updating HeadSetOffsetCursor.
+			 */
+			asyh->set.curs = asyh->curs.visible;
+		} else
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0860 + head->base.index * 0x400, 1);
+			evo_data(push, asyh->core.offset >> 8);
+			evo_mthd(push, 0x0868 + head->base.index * 0x400, 4);
+			evo_data(push, (asyh->core.h << 16) | asyh->core.w);
+			evo_data(push, asyh->core.layout << 20 |
+				       (asyh->core.pitch >> 8) << 8 |
+				       asyh->core.block);
+			evo_data(push, asyh->core.format << 8);
+			evo_data(push, asyh->core.handle);
+			evo_mthd(push, 0x08c0 + head->base.index * 0x400, 1);
+			evo_data(push, (asyh->core.y << 16) | asyh->core.x);
+		} else {
+			evo_mthd(push, 0x0460 + head->base.index * 0x300, 1);
+			evo_data(push, asyh->core.offset >> 8);
+			evo_mthd(push, 0x0468 + head->base.index * 0x300, 4);
+			evo_data(push, (asyh->core.h << 16) | asyh->core.w);
+			evo_data(push, asyh->core.layout << 24 |
+				       (asyh->core.pitch >> 8) << 8 |
+				       asyh->core.block);
+			evo_data(push, asyh->core.format << 8);
+			evo_data(push, asyh->core.handle);
+			evo_mthd(push, 0x04b0 + head->base.index * 0x300, 1);
+			evo_data(push, (asyh->core.y << 16) | asyh->core.x);
+		}
+		evo_kick(push, core);
+	}
+}
+
+static void
+nv50_head_lut_clr(struct nv50_head *head)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
+	u32 *push;
+	if ((push = evo_wait(core, 4))) {
+		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0840 + (head->base.index * 0x400), 1);
+			evo_data(push, 0x40000000);
+		} else
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0840 + (head->base.index * 0x400), 1);
+			evo_data(push, 0x40000000);
+			evo_mthd(push, 0x085c + (head->base.index * 0x400), 1);
+			evo_data(push, 0x00000000);
+		} else {
+			evo_mthd(push, 0x0440 + (head->base.index * 0x300), 1);
+			evo_data(push, 0x03000000);
+			evo_mthd(push, 0x045c + (head->base.index * 0x300), 1);
+			evo_data(push, 0x00000000);
+		}
+		evo_kick(push, core);
+	}
+}
+
+static void
+nv50_head_lut_load(struct drm_property_blob *blob, int mode,
+		   struct nouveau_bo *nvbo)
+{
+	struct drm_color_lut *in = (struct drm_color_lut *)blob->data;
+	void __iomem *lut = (u8 *)nvbo_kmap_obj_iovirtual(nvbo);
+	const int size = blob->length / sizeof(*in);
+	int bits, shift, i;
+	u16 zero, r, g, b;
+
+	/* This can't happen.. But it shuts the compiler up. */
+	if (WARN_ON(size != 256))
+		return;
+
+	switch (mode) {
+	case 0: /* LORES. */
+	case 1: /* HIRES. */
+		bits = 11;
+		shift = 3;
+		zero = 0x0000;
+		break;
+	case 7: /* INTERPOLATE_257_UNITY_RANGE. */
+		bits = 14;
+		shift = 0;
+		zero = 0x6000;
+		break;
+	default:
+		WARN_ON(1);
+		return;
+	}
+
+	for (i = 0; i < size; i++) {
+		r = (drm_color_lut_extract(in[i].  red, bits) + zero) << shift;
+		g = (drm_color_lut_extract(in[i].green, bits) + zero) << shift;
+		b = (drm_color_lut_extract(in[i]. blue, bits) + zero) << shift;
+		writew(r, lut + (i * 0x08) + 0);
+		writew(g, lut + (i * 0x08) + 2);
+		writew(b, lut + (i * 0x08) + 4);
+	}
+
+	/* INTERPOLATE modes require a "next" entry to interpolate with,
+	 * so we replicate the last entry to deal with this for now.
+	 */
+	writew(r, lut + (i * 0x08) + 0);
+	writew(g, lut + (i * 0x08) + 2);
+	writew(b, lut + (i * 0x08) + 4);
+}
+
+static void
+nv50_head_lut_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
+	u32 *push;
+	if ((push = evo_wait(core, 7))) {
+		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0840 + (head->base.index * 0x400), 2);
+			evo_data(push, 0x80000000 | asyh->lut.mode << 30);
+			evo_data(push, asyh->lut.offset >> 8);
+		} else
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0840 + (head->base.index * 0x400), 2);
+			evo_data(push, 0x80000000 | asyh->lut.mode << 30);
+			evo_data(push, asyh->lut.offset >> 8);
+			evo_mthd(push, 0x085c + (head->base.index * 0x400), 1);
+			evo_data(push, asyh->lut.handle);
+		} else {
+			evo_mthd(push, 0x0440 + (head->base.index * 0x300), 4);
+			evo_data(push, 0x80000000 | asyh->lut.mode << 24);
+			evo_data(push, asyh->lut.offset >> 8);
+			evo_data(push, 0x00000000);
+			evo_data(push, 0x00000000);
+			evo_mthd(push, 0x045c + (head->base.index * 0x300), 1);
+			evo_data(push, asyh->lut.handle);
+		}
+		evo_kick(push, core);
+	}
+}
+
+static void
+nv50_head_mode(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
+	struct nv50_head_mode *m = &asyh->mode;
+	u32 *push;
+	if ((push = evo_wait(core, 14))) {
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0804 + (head->base.index * 0x400), 2);
+			evo_data(push, 0x00800000 | m->clock);
+			evo_data(push, m->interlace ? 0x00000002 : 0x00000000);
+			evo_mthd(push, 0x0810 + (head->base.index * 0x400), 7);
+			evo_data(push, 0x00000000);
+			evo_data(push, (m->v.active  << 16) | m->h.active );
+			evo_data(push, (m->v.synce   << 16) | m->h.synce  );
+			evo_data(push, (m->v.blanke  << 16) | m->h.blanke );
+			evo_data(push, (m->v.blanks  << 16) | m->h.blanks );
+			evo_data(push, (m->v.blank2e << 16) | m->v.blank2s);
+			evo_data(push, asyh->mode.v.blankus);
+			evo_mthd(push, 0x082c + (head->base.index * 0x400), 1);
+			evo_data(push, 0x00000000);
+		} else {
+			evo_mthd(push, 0x0410 + (head->base.index * 0x300), 6);
+			evo_data(push, 0x00000000);
+			evo_data(push, (m->v.active  << 16) | m->h.active );
+			evo_data(push, (m->v.synce   << 16) | m->h.synce  );
+			evo_data(push, (m->v.blanke  << 16) | m->h.blanke );
+			evo_data(push, (m->v.blanks  << 16) | m->h.blanks );
+			evo_data(push, (m->v.blank2e << 16) | m->v.blank2s);
+			evo_mthd(push, 0x042c + (head->base.index * 0x300), 2);
+			evo_data(push, 0x00000000); /* ??? */
+			evo_data(push, 0xffffff00);
+			evo_mthd(push, 0x0450 + (head->base.index * 0x300), 3);
+			evo_data(push, m->clock * 1000);
+			evo_data(push, 0x00200000); /* ??? */
+			evo_data(push, m->clock * 1000);
+		}
+		evo_kick(push, core);
+	}
+}
+
+static void
+nv50_head_view(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
+	u32 *push;
+	if ((push = evo_wait(core, 10))) {
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
+			evo_mthd(push, 0x08a4 + (head->base.index * 0x400), 1);
+			evo_data(push, 0x00000000);
+			evo_mthd(push, 0x08c8 + (head->base.index * 0x400), 1);
+			evo_data(push, (asyh->view.iH << 16) | asyh->view.iW);
+			evo_mthd(push, 0x08d8 + (head->base.index * 0x400), 2);
+			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
+			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
+		} else {
+			evo_mthd(push, 0x0494 + (head->base.index * 0x300), 1);
+			evo_data(push, 0x00000000);
+			evo_mthd(push, 0x04b8 + (head->base.index * 0x300), 1);
+			evo_data(push, (asyh->view.iH << 16) | asyh->view.iW);
+			evo_mthd(push, 0x04c0 + (head->base.index * 0x300), 3);
+			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
+			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
+			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
+		}
+		evo_kick(push, core);
+	}
+}
+
+static void
+nv50_head_flush_clr(struct nv50_head *head, struct nv50_head_atom *asyh, bool y)
+{
+	if (asyh->clr.ilut && (!asyh->set.ilut || y))
+		nv50_head_lut_clr(head);
+	if (asyh->clr.core && (!asyh->set.core || y))
+		nv50_head_core_clr(head);
+	if (asyh->clr.curs && (!asyh->set.curs || y))
+		nv50_head_curs_clr(head);
+}
+
+static void
+nv50_head_flush_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	if (asyh->set.view   ) nv50_head_view    (head, asyh);
+	if (asyh->set.mode   ) nv50_head_mode    (head, asyh);
+	if (asyh->set.ilut   ) {
+		struct nouveau_bo *nvbo = head->lut.nvbo[head->lut.next];
+		struct drm_property_blob *blob = asyh->state.gamma_lut;
+		if (blob)
+			nv50_head_lut_load(blob, asyh->lut.mode, nvbo);
+		asyh->lut.offset = nvbo->bo.offset;
+		head->lut.next ^= 1;
+		nv50_head_lut_set(head, asyh);
+	}
+	if (asyh->set.core   ) nv50_head_core_set(head, asyh);
+	if (asyh->set.curs   ) nv50_head_curs_set(head, asyh);
+	if (asyh->set.base   ) nv50_head_base    (head, asyh);
+	if (asyh->set.ovly   ) nv50_head_ovly    (head, asyh);
+	if (asyh->set.dither ) nv50_head_dither  (head, asyh);
+	if (asyh->set.procamp) nv50_head_procamp (head, asyh);
+}
+
+static void
+nv50_head_atomic_check_procamp(struct nv50_head_atom *armh,
+			       struct nv50_head_atom *asyh,
+			       struct nouveau_conn_atom *asyc)
+{
+	const int vib = asyc->procamp.color_vibrance - 100;
+	const int hue = asyc->procamp.vibrant_hue - 90;
+	const int adj = (vib > 0) ? 50 : 0;
+	asyh->procamp.sat.cos = ((vib * 2047 + adj) / 100) & 0xfff;
+	asyh->procamp.sat.sin = ((hue * 2047) / 100) & 0xfff;
+	asyh->set.procamp = true;
+}
+
+static void
+nv50_head_atomic_check_dither(struct nv50_head_atom *armh,
+			      struct nv50_head_atom *asyh,
+			      struct nouveau_conn_atom *asyc)
+{
+	struct drm_connector *connector = asyc->state.connector;
+	u32 mode = 0x00;
+
+	if (asyc->dither.mode == DITHERING_MODE_AUTO) {
+		if (asyh->base.depth > connector->display_info.bpc * 3)
+			mode = DITHERING_MODE_DYNAMIC2X2;
+	} else {
+		mode = asyc->dither.mode;
+	}
+
+	if (asyc->dither.depth == DITHERING_DEPTH_AUTO) {
+		if (connector->display_info.bpc >= 8)
+			mode |= DITHERING_DEPTH_8BPC;
+	} else {
+		mode |= asyc->dither.depth;
+	}
+
+	asyh->dither.enable = mode;
+	asyh->dither.bits = mode >> 1;
+	asyh->dither.mode = mode >> 3;
+	asyh->set.dither = true;
+}
+
+static void
+nv50_head_atomic_check_view(struct nv50_head_atom *armh,
+			    struct nv50_head_atom *asyh,
+			    struct nouveau_conn_atom *asyc)
+{
+	struct drm_connector *connector = asyc->state.connector;
+	struct drm_display_mode *omode = &asyh->state.adjusted_mode;
+	struct drm_display_mode *umode = &asyh->state.mode;
+	int mode = asyc->scaler.mode;
+	struct edid *edid;
+	int umode_vdisplay, omode_hdisplay, omode_vdisplay;
+
+	if (connector->edid_blob_ptr)
+		edid = (struct edid *)connector->edid_blob_ptr->data;
+	else
+		edid = NULL;
+
+	if (!asyc->scaler.full) {
+		if (mode == DRM_MODE_SCALE_NONE)
+			omode = umode;
+	} else {
+		/* Non-EDID LVDS/eDP mode. */
+		mode = DRM_MODE_SCALE_FULLSCREEN;
+	}
+
+	/* For the user-specified mode, we must ignore doublescan and
+	 * the like, but honor frame packing.
+	 */
+	umode_vdisplay = umode->vdisplay;
+	if ((umode->flags & DRM_MODE_FLAG_3D_MASK) == DRM_MODE_FLAG_3D_FRAME_PACKING)
+		umode_vdisplay += umode->vtotal;
+	asyh->view.iW = umode->hdisplay;
+	asyh->view.iH = umode_vdisplay;
+	/* For the output mode, we can just use the stock helper. */
+	drm_mode_get_hv_timing(omode, &omode_hdisplay, &omode_vdisplay);
+	asyh->view.oW = omode_hdisplay;
+	asyh->view.oH = omode_vdisplay;
+
+	/* Add overscan compensation if necessary, will keep the aspect
+	 * ratio the same as the backend mode unless overridden by the
+	 * user setting both hborder and vborder properties.
+	 */
+	if ((asyc->scaler.underscan.mode == UNDERSCAN_ON ||
+	    (asyc->scaler.underscan.mode == UNDERSCAN_AUTO &&
+	     drm_detect_hdmi_monitor(edid)))) {
+		u32 bX = asyc->scaler.underscan.hborder;
+		u32 bY = asyc->scaler.underscan.vborder;
+		u32 r = (asyh->view.oH << 19) / asyh->view.oW;
+
+		if (bX) {
+			asyh->view.oW -= (bX * 2);
+			if (bY) asyh->view.oH -= (bY * 2);
+			else    asyh->view.oH  = ((asyh->view.oW * r) + (r / 2)) >> 19;
+		} else {
+			asyh->view.oW -= (asyh->view.oW >> 4) + 32;
+			if (bY) asyh->view.oH -= (bY * 2);
+			else    asyh->view.oH  = ((asyh->view.oW * r) + (r / 2)) >> 19;
+		}
+	}
+
+	/* Handle CENTER/ASPECT scaling, taking into account the areas
+	 * removed already for overscan compensation.
+	 */
+	switch (mode) {
+	case DRM_MODE_SCALE_CENTER:
+		asyh->view.oW = min((u16)umode->hdisplay, asyh->view.oW);
+		asyh->view.oH = min((u16)umode_vdisplay, asyh->view.oH);
+		/* fall-through */
+	case DRM_MODE_SCALE_ASPECT:
+		if (asyh->view.oH < asyh->view.oW) {
+			u32 r = (asyh->view.iW << 19) / asyh->view.iH;
+			asyh->view.oW = ((asyh->view.oH * r) + (r / 2)) >> 19;
+		} else {
+			u32 r = (asyh->view.iH << 19) / asyh->view.iW;
+			asyh->view.oH = ((asyh->view.oW * r) + (r / 2)) >> 19;
+		}
+		break;
+	default:
+		break;
+	}
+
+	asyh->set.view = true;
+}
+
+static void
+nv50_head_atomic_check_lut(struct nv50_head *head,
+			   struct nv50_head_atom *armh,
+			   struct nv50_head_atom *asyh)
+{
+	struct nv50_disp *disp = nv50_disp(head->base.base.dev);
+
+	/* An I8 surface without an input LUT makes no sense, and
+	 * EVO will throw an error if you try.
+	 *
+	 * Legacy clients actually cause this due to the order in
+	 * which they call ioctls, so we will enable the LUT with
+	 * whatever contents the buffer already contains to avoid
+	 * triggering the error check.
+	 */
+	if (!asyh->state.gamma_lut && asyh->base.cpp != 1) {
+		asyh->lut.handle = 0;
+		asyh->clr.ilut = armh->lut.visible;
+		return;
+	}
+
+	if (disp->disp->object.oclass < GF110_DISP) {
+		asyh->lut.mode = (asyh->base.cpp == 1) ? 0 : 1;
+		asyh->set.ilut = true;
+	} else {
+		asyh->lut.mode = 7;
+		asyh->set.ilut = asyh->state.color_mgmt_changed;
+	}
+	asyh->lut.handle = disp->mast.base.vram.handle;
+}
+
+static void
+nv50_head_atomic_check_mode(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct drm_display_mode *mode = &asyh->state.adjusted_mode;
+	struct nv50_head_mode *m = &asyh->mode;
+	u32 blankus;
+
+	drm_mode_set_crtcinfo(mode, CRTC_INTERLACE_HALVE_V | CRTC_STEREO_DOUBLE);
+
+	/*
+	 * DRM modes are defined in terms of a repeating interval
+	 * starting with the active display area.  The hardware modes
+	 * are defined in terms of a repeating interval starting one
+	 * unit (pixel or line) into the sync pulse.  So, add bias.
+	 */
+
+	m->h.active = mode->crtc_htotal;
+	m->h.synce  = mode->crtc_hsync_end - mode->crtc_hsync_start - 1;
+	m->h.blanke = mode->crtc_hblank_end - mode->crtc_hsync_start - 1;
+	m->h.blanks = m->h.blanke + mode->crtc_hdisplay;
+
+	m->v.active = mode->crtc_vtotal;
+	m->v.synce  = mode->crtc_vsync_end - mode->crtc_vsync_start - 1;
+	m->v.blanke = mode->crtc_vblank_end - mode->crtc_vsync_start - 1;
+	m->v.blanks = m->v.blanke + mode->crtc_vdisplay;
+
+	/*XXX: Safe underestimate, even "0" works */
+	blankus = (m->v.active - mode->crtc_vdisplay - 2) * m->h.active;
+	blankus *= 1000;
+	blankus /= mode->crtc_clock;
+	m->v.blankus = blankus;
+
+	if (mode->flags & DRM_MODE_FLAG_INTERLACE) {
+		m->v.blank2e =  m->v.active + m->v.blanke;
+		m->v.blank2s =  m->v.blank2e + mode->crtc_vdisplay;
+		m->v.active  = (m->v.active * 2) + 1;
+		m->interlace = true;
+	} else {
+		m->v.blank2e = 0;
+		m->v.blank2s = 1;
+		m->interlace = false;
+	}
+	m->clock = mode->crtc_clock;
+
+	asyh->set.mode = true;
+}
+
+static int
+nv50_head_atomic_check(struct drm_crtc *crtc, struct drm_crtc_state *state)
+{
+	struct nouveau_drm *drm = nouveau_drm(crtc->dev);
+	struct nv50_disp *disp = nv50_disp(crtc->dev);
+	struct nv50_head *head = nv50_head(crtc);
+	struct nv50_head_atom *armh = nv50_head_atom(crtc->state);
+	struct nv50_head_atom *asyh = nv50_head_atom(state);
+	struct nouveau_conn_atom *asyc = NULL;
+	struct drm_connector_state *conns;
+	struct drm_connector *conn;
+	int i;
+
+	NV_ATOMIC(drm, "%s atomic_check %d\n", crtc->name, asyh->state.active);
+	if (asyh->state.active) {
+		for_each_new_connector_in_state(asyh->state.state, conn, conns, i) {
+			if (conns->crtc == crtc) {
+				asyc = nouveau_conn_atom(conns);
+				break;
+			}
+		}
+
+		if (armh->state.active) {
+			if (asyc) {
+				if (asyh->state.mode_changed)
+					asyc->set.scaler = true;
+				if (armh->base.depth != asyh->base.depth)
+					asyc->set.dither = true;
+			}
+		} else {
+			if (asyc)
+				asyc->set.mask = ~0;
+			asyh->set.mask = ~0;
+		}
+
+		if (asyh->state.mode_changed)
+			nv50_head_atomic_check_mode(head, asyh);
+
+		if (asyh->state.color_mgmt_changed ||
+		    asyh->base.cpp != armh->base.cpp)
+			nv50_head_atomic_check_lut(head, armh, asyh);
+		asyh->lut.visible = asyh->lut.handle != 0;
+
+		if (asyc) {
+			if (asyc->set.scaler)
+				nv50_head_atomic_check_view(armh, asyh, asyc);
+			if (asyc->set.dither)
+				nv50_head_atomic_check_dither(armh, asyh, asyc);
+			if (asyc->set.procamp)
+				nv50_head_atomic_check_procamp(armh, asyh, asyc);
+		}
+
+		if ((asyh->core.visible = (asyh->base.cpp != 0))) {
+			asyh->core.x = asyh->base.x;
+			asyh->core.y = asyh->base.y;
+			asyh->core.w = asyh->base.w;
+			asyh->core.h = asyh->base.h;
+		} else
+		if ((asyh->core.visible = asyh->curs.visible) ||
+		    (asyh->core.visible = asyh->lut.visible)) {
+			/*XXX: We need to either find some way of having the
+			 *     primary base layer appear black, while still
+			 *     being able to display the other layers, or we
+			 *     need to allocate a dummy black surface here.
+			 */
+			asyh->core.x = 0;
+			asyh->core.y = 0;
+			asyh->core.w = asyh->state.mode.hdisplay;
+			asyh->core.h = asyh->state.mode.vdisplay;
+		}
+		asyh->core.handle = disp->mast.base.vram.handle;
+		asyh->core.offset = 0;
+		asyh->core.format = 0xcf;
+		asyh->core.kind = 0;
+		asyh->core.layout = 1;
+		asyh->core.block = 0;
+		asyh->core.pitch = ALIGN(asyh->core.w, 64) * 4;
+		asyh->set.base = armh->base.cpp != asyh->base.cpp;
+		asyh->set.ovly = armh->ovly.cpp != asyh->ovly.cpp;
+	} else {
+		asyh->lut.visible = false;
+		asyh->core.visible = false;
+		asyh->curs.visible = false;
+		asyh->base.cpp = 0;
+		asyh->ovly.cpp = 0;
+	}
+
+	if (!drm_atomic_crtc_needs_modeset(&asyh->state)) {
+		if (asyh->core.visible) {
+			if (memcmp(&armh->core, &asyh->core, sizeof(asyh->core)))
+				asyh->set.core = true;
+		} else
+		if (armh->core.visible) {
+			asyh->clr.core = true;
+		}
+
+		if (asyh->curs.visible) {
+			if (memcmp(&armh->curs, &asyh->curs, sizeof(asyh->curs)))
+				asyh->set.curs = true;
+		} else
+		if (armh->curs.visible) {
+			asyh->clr.curs = true;
+		}
+	} else {
+		asyh->clr.ilut = armh->lut.visible;
+		asyh->clr.core = armh->core.visible;
+		asyh->clr.curs = armh->curs.visible;
+		asyh->set.ilut = asyh->lut.visible;
+		asyh->set.core = asyh->core.visible;
+		asyh->set.curs = asyh->curs.visible;
+	}
+
+	if (asyh->clr.mask || asyh->set.mask)
+		nv50_atom(asyh->state.state)->lock_core = true;
+	return 0;
+}
+
+static const struct drm_crtc_helper_funcs
+nv50_head_help = {
+	.atomic_check = nv50_head_atomic_check,
+};
+
+static void
+nv50_head_atomic_destroy_state(struct drm_crtc *crtc,
+			       struct drm_crtc_state *state)
+{
+	struct nv50_head_atom *asyh = nv50_head_atom(state);
+	__drm_atomic_helper_crtc_destroy_state(&asyh->state);
+	kfree(asyh);
+}
+
+static struct drm_crtc_state *
+nv50_head_atomic_duplicate_state(struct drm_crtc *crtc)
+{
+	struct nv50_head_atom *armh = nv50_head_atom(crtc->state);
+	struct nv50_head_atom *asyh;
+	if (!(asyh = kmalloc(sizeof(*asyh), GFP_KERNEL)))
+		return NULL;
+	__drm_atomic_helper_crtc_duplicate_state(crtc, &asyh->state);
+	asyh->view = armh->view;
+	asyh->mode = armh->mode;
+	asyh->lut  = armh->lut;
+	asyh->core = armh->core;
+	asyh->curs = armh->curs;
+	asyh->base = armh->base;
+	asyh->ovly = armh->ovly;
+	asyh->dither = armh->dither;
+	asyh->procamp = armh->procamp;
+	asyh->clr.mask = 0;
+	asyh->set.mask = 0;
+	return &asyh->state;
+}
+
+static void
+__drm_atomic_helper_crtc_reset(struct drm_crtc *crtc,
+			       struct drm_crtc_state *state)
+{
+	if (crtc->state)
+		crtc->funcs->atomic_destroy_state(crtc, crtc->state);
+	crtc->state = state;
+	crtc->state->crtc = crtc;
+}
+
+static void
+nv50_head_reset(struct drm_crtc *crtc)
+{
+	struct nv50_head_atom *asyh;
+
+	if (WARN_ON(!(asyh = kzalloc(sizeof(*asyh), GFP_KERNEL))))
+		return;
+
+	__drm_atomic_helper_crtc_reset(crtc, &asyh->state);
+}
+
+static void
+nv50_head_destroy(struct drm_crtc *crtc)
+{
+	struct nv50_head *head = nv50_head(crtc);
+	int i;
+
+	nv50_dmac_destroy(&head->ovly.base);
+	nv50_pioc_destroy(&head->oimm.base);
+
+	for (i = 0; i < ARRAY_SIZE(head->lut.nvbo); i++)
+		nouveau_bo_unmap_unpin_unref(&head->lut.nvbo[i]);
+
+	drm_crtc_cleanup(crtc);
+	kfree(crtc);
+}
+
+static const struct drm_crtc_funcs
+nv50_head_func = {
+	.reset = nv50_head_reset,
+	.gamma_set = drm_atomic_helper_legacy_gamma_set,
+	.destroy = nv50_head_destroy,
+	.set_config = drm_atomic_helper_set_config,
+	.page_flip = drm_atomic_helper_page_flip,
+	.atomic_duplicate_state = nv50_head_atomic_duplicate_state,
+	.atomic_destroy_state = nv50_head_atomic_destroy_state,
+};
+
+static int
+nv50_head_create(struct drm_device *dev, int index)
+{
+	struct nouveau_drm *drm = nouveau_drm(dev);
+	struct nvif_device *device = &drm->client.device;
+	struct nv50_disp *disp = nv50_disp(dev);
+	struct nv50_head *head;
+	struct nv50_base *base;
+	struct nv50_curs *curs;
+	struct drm_crtc *crtc;
+	int ret, i;
+
+	head = kzalloc(sizeof(*head), GFP_KERNEL);
+	if (!head)
+		return -ENOMEM;
+
+	head->base.index = index;
+	ret = nv50_base_new(drm, head, &base);
+	if (ret == 0)
+		ret = nv50_curs_new(drm, head, &curs);
+	if (ret) {
+		kfree(head);
+		return ret;
+	}
+
+	crtc = &head->base.base;
+	drm_crtc_init_with_planes(dev, crtc, &base->wndw.plane,
+				  &curs->wndw.plane, &nv50_head_func,
+				  "head-%d", head->base.index);
+	drm_crtc_helper_add(crtc, &nv50_head_help);
+	drm_mode_crtc_set_gamma_size(crtc, 256);
+
+	for (i = 0; i < ARRAY_SIZE(head->lut.nvbo); i++) {
+		ret = nouveau_bo_new_pin_map(&drm->client, 1025 * 8, 0x100,
+					     TTM_PL_FLAG_VRAM,
+					     &head->lut.nvbo[i]);
+		if (ret)
+			goto out;
+	}
+
+	/* allocate overlay resources */
+	ret = nv50_oimm_create(device, &disp->disp->object, index, &head->oimm);
+	if (ret)
+		goto out;
+
+	ret = nv50_ovly_create(device, &disp->disp->object, index,
+			       disp->sync->bo.offset, &head->ovly);
+	if (ret)
+		goto out;
+
+out:
+	if (ret)
+		nv50_head_destroy(crtc);
+	return ret;
+}
+
+/******************************************************************************
+ * Output path helpers
+ *****************************************************************************/
+static void
+nv50_outp_release(struct nouveau_encoder *nv_encoder)
+{
+	struct nv50_disp *disp = nv50_disp(nv_encoder->base.base.dev);
+	struct {
+		struct nv50_disp_mthd_v1 base;
+	} args = {
+		.base.version = 1,
+		.base.method = NV50_DISP_MTHD_V1_RELEASE,
+		.base.hasht  = nv_encoder->dcb->hasht,
+		.base.hashm  = nv_encoder->dcb->hashm,
+	};
+
+	nvif_mthd(&disp->disp->object, 0, &args, sizeof(args));
+	nv_encoder->or = -1;
+	nv_encoder->link = 0;
+}
+
+static int
+nv50_outp_acquire(struct nouveau_encoder *nv_encoder)
+{
+	struct nouveau_drm *drm = nouveau_drm(nv_encoder->base.base.dev);
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	struct {
+		struct nv50_disp_mthd_v1 base;
+		struct nv50_disp_acquire_v0 info;
+	} args = {
+		.base.version = 1,
+		.base.method = NV50_DISP_MTHD_V1_ACQUIRE,
+		.base.hasht  = nv_encoder->dcb->hasht,
+		.base.hashm  = nv_encoder->dcb->hashm,
+	};
+	int ret;
+
+	ret = nvif_mthd(&disp->disp->object, 0, &args, sizeof(args));
+	if (ret) {
+		NV_ERROR(drm, "error acquiring output path: %d\n", ret);
+		return ret;
+	}
+
+	nv_encoder->or = args.info.or;
+	nv_encoder->link = args.info.link;
+	return 0;
+}
+
+static int
+nv50_outp_atomic_check_view(struct drm_encoder *encoder,
+			    struct drm_crtc_state *crtc_state,
+			    struct drm_connector_state *conn_state,
+			    struct drm_display_mode *native_mode)
+{
+	struct drm_display_mode *adjusted_mode = &crtc_state->adjusted_mode;
+	struct drm_display_mode *mode = &crtc_state->mode;
+	struct drm_connector *connector = conn_state->connector;
+	struct nouveau_conn_atom *asyc = nouveau_conn_atom(conn_state);
+	struct nouveau_drm *drm = nouveau_drm(encoder->dev);
+
+	NV_ATOMIC(drm, "%s atomic_check\n", encoder->name);
+	asyc->scaler.full = false;
+	if (!native_mode)
+		return 0;
+
+	if (asyc->scaler.mode == DRM_MODE_SCALE_NONE) {
+		switch (connector->connector_type) {
+		case DRM_MODE_CONNECTOR_LVDS:
+		case DRM_MODE_CONNECTOR_eDP:
+			/* Force use of scaler for non-EDID modes. */
+			if (adjusted_mode->type & DRM_MODE_TYPE_DRIVER)
+				break;
+			mode = native_mode;
+			asyc->scaler.full = true;
+			break;
+		default:
+			break;
+		}
+	} else {
+		mode = native_mode;
+	}
+
+	if (!drm_mode_equal(adjusted_mode, mode)) {
+		drm_mode_copy(adjusted_mode, mode);
+		crtc_state->mode_changed = true;
+	}
+
+	return 0;
+}
+
+static int
+nv50_outp_atomic_check(struct drm_encoder *encoder,
+		       struct drm_crtc_state *crtc_state,
+		       struct drm_connector_state *conn_state)
+{
+	struct nouveau_connector *nv_connector =
+		nouveau_connector(conn_state->connector);
+	return nv50_outp_atomic_check_view(encoder, crtc_state, conn_state,
+					   nv_connector->native_mode);
+}
+
+/******************************************************************************
+ * DAC
+ *****************************************************************************/
+static void
+nv50_dac_disable(struct drm_encoder *encoder)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nv50_mast *mast = nv50_mast(encoder->dev);
+	const int or = nv_encoder->or;
+	u32 *push;
+
+	if (nv_encoder->crtc) {
+		push = evo_wait(mast, 4);
+		if (push) {
+			if (nv50_vers(mast) < GF110_DISP_CORE_CHANNEL_DMA) {
+				evo_mthd(push, 0x0400 + (or * 0x080), 1);
+				evo_data(push, 0x00000000);
+			} else {
+				evo_mthd(push, 0x0180 + (or * 0x020), 1);
+				evo_data(push, 0x00000000);
+			}
+			evo_kick(push, mast);
+		}
+	}
+
+	nv_encoder->crtc = NULL;
+	nv50_outp_release(nv_encoder);
+}
+
+static void
+nv50_dac_enable(struct drm_encoder *encoder)
+{
+	struct nv50_mast *mast = nv50_mast(encoder->dev);
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
+	struct drm_display_mode *mode = &nv_crtc->base.state->adjusted_mode;
+	u32 *push;
+
+	nv50_outp_acquire(nv_encoder);
+
+	push = evo_wait(mast, 8);
+	if (push) {
+		if (nv50_vers(mast) < GF110_DISP_CORE_CHANNEL_DMA) {
+			u32 syncs = 0x00000000;
+
+			if (mode->flags & DRM_MODE_FLAG_NHSYNC)
+				syncs |= 0x00000001;
+			if (mode->flags & DRM_MODE_FLAG_NVSYNC)
+				syncs |= 0x00000002;
+
+			evo_mthd(push, 0x0400 + (nv_encoder->or * 0x080), 2);
+			evo_data(push, 1 << nv_crtc->index);
+			evo_data(push, syncs);
+		} else {
+			u32 magic = 0x31ec6000 | (nv_crtc->index << 25);
+			u32 syncs = 0x00000001;
+
+			if (mode->flags & DRM_MODE_FLAG_NHSYNC)
+				syncs |= 0x00000008;
+			if (mode->flags & DRM_MODE_FLAG_NVSYNC)
+				syncs |= 0x00000010;
+
+			if (mode->flags & DRM_MODE_FLAG_INTERLACE)
+				magic |= 0x00000001;
+
+			evo_mthd(push, 0x0404 + (nv_crtc->index * 0x300), 2);
+			evo_data(push, syncs);
+			evo_data(push, magic);
+			evo_mthd(push, 0x0180 + (nv_encoder->or * 0x020), 1);
+			evo_data(push, 1 << nv_crtc->index);
+		}
+
+		evo_kick(push, mast);
+	}
+
+	nv_encoder->crtc = encoder->crtc;
+}
+
+static enum drm_connector_status
+nv50_dac_detect(struct drm_encoder *encoder, struct drm_connector *connector)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nv50_disp *disp = nv50_disp(encoder->dev);
+	struct {
+		struct nv50_disp_mthd_v1 base;
+		struct nv50_disp_dac_load_v0 load;
+	} args = {
+		.base.version = 1,
+		.base.method = NV50_DISP_MTHD_V1_DAC_LOAD,
+		.base.hasht  = nv_encoder->dcb->hasht,
+		.base.hashm  = nv_encoder->dcb->hashm,
+	};
+	int ret;
+
+	args.load.data = nouveau_drm(encoder->dev)->vbios.dactestval;
+	if (args.load.data == 0)
+		args.load.data = 340;
+
+	ret = nvif_mthd(&disp->disp->object, 0, &args, sizeof(args));
+	if (ret || !args.load.load)
+		return connector_status_disconnected;
+
+	return connector_status_connected;
+}
+
+static const struct drm_encoder_helper_funcs
+nv50_dac_help = {
+	.atomic_check = nv50_outp_atomic_check,
+	.enable = nv50_dac_enable,
+	.disable = nv50_dac_disable,
+	.detect = nv50_dac_detect
+};
+
+static void
+nv50_dac_destroy(struct drm_encoder *encoder)
+{
+	drm_encoder_cleanup(encoder);
+	kfree(encoder);
+}
+
+static const struct drm_encoder_funcs
+nv50_dac_func = {
+	.destroy = nv50_dac_destroy,
+};
+
+static int
+nv50_dac_create(struct drm_connector *connector, struct dcb_output *dcbe)
+{
+	struct nouveau_drm *drm = nouveau_drm(connector->dev);
+	struct nvkm_i2c *i2c = nvxx_i2c(&drm->client.device);
+	struct nvkm_i2c_bus *bus;
+	struct nouveau_encoder *nv_encoder;
+	struct drm_encoder *encoder;
+	int type = DRM_MODE_ENCODER_DAC;
+
+	nv_encoder = kzalloc(sizeof(*nv_encoder), GFP_KERNEL);
+	if (!nv_encoder)
+		return -ENOMEM;
+	nv_encoder->dcb = dcbe;
+
+	bus = nvkm_i2c_bus_find(i2c, dcbe->i2c_index);
+	if (bus)
+		nv_encoder->i2c = &bus->i2c;
+
+	encoder = to_drm_encoder(nv_encoder);
+	encoder->possible_crtcs = dcbe->heads;
+	encoder->possible_clones = 0;
+	drm_encoder_init(connector->dev, encoder, &nv50_dac_func, type,
+			 "dac-%04x-%04x", dcbe->hasht, dcbe->hashm);
+	drm_encoder_helper_add(encoder, &nv50_dac_help);
+
+	drm_mode_connector_attach_encoder(connector, encoder);
+	return 0;
+}
+
+/******************************************************************************
+ * Audio
+ *****************************************************************************/
+static void
+nv50_audio_disable(struct drm_encoder *encoder, struct nouveau_crtc *nv_crtc)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nv50_disp *disp = nv50_disp(encoder->dev);
+	struct {
+		struct nv50_disp_mthd_v1 base;
+		struct nv50_disp_sor_hda_eld_v0 eld;
+	} args = {
+		.base.version = 1,
+		.base.method  = NV50_DISP_MTHD_V1_SOR_HDA_ELD,
+		.base.hasht   = nv_encoder->dcb->hasht,
+		.base.hashm   = (0xf0ff & nv_encoder->dcb->hashm) |
+				(0x0100 << nv_crtc->index),
+	};
+
+	nvif_mthd(&disp->disp->object, 0, &args, sizeof(args));
+}
+
+static void
+nv50_audio_enable(struct drm_encoder *encoder, struct drm_display_mode *mode)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
+	struct nouveau_connector *nv_connector;
+	struct nv50_disp *disp = nv50_disp(encoder->dev);
+	struct __packed {
+		struct {
+			struct nv50_disp_mthd_v1 mthd;
+			struct nv50_disp_sor_hda_eld_v0 eld;
+		} base;
+		u8 data[sizeof(nv_connector->base.eld)];
+	} args = {
+		.base.mthd.version = 1,
+		.base.mthd.method  = NV50_DISP_MTHD_V1_SOR_HDA_ELD,
+		.base.mthd.hasht   = nv_encoder->dcb->hasht,
+		.base.mthd.hashm   = (0xf0ff & nv_encoder->dcb->hashm) |
+				     (0x0100 << nv_crtc->index),
+	};
+
+	nv_connector = nouveau_encoder_connector_get(nv_encoder);
+	if (!drm_detect_monitor_audio(nv_connector->edid))
+		return;
+
+	memcpy(args.data, nv_connector->base.eld, sizeof(args.data));
+
+	nvif_mthd(&disp->disp->object, 0, &args,
+		  sizeof(args.base) + drm_eld_size(args.data));
+}
+
+/******************************************************************************
+ * HDMI
+ *****************************************************************************/
+static void
+nv50_hdmi_disable(struct drm_encoder *encoder, struct nouveau_crtc *nv_crtc)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nv50_disp *disp = nv50_disp(encoder->dev);
+	struct {
+		struct nv50_disp_mthd_v1 base;
+		struct nv50_disp_sor_hdmi_pwr_v0 pwr;
+	} args = {
+		.base.version = 1,
+		.base.method = NV50_DISP_MTHD_V1_SOR_HDMI_PWR,
+		.base.hasht  = nv_encoder->dcb->hasht,
+		.base.hashm  = (0xf0ff & nv_encoder->dcb->hashm) |
+			       (0x0100 << nv_crtc->index),
+	};
+
+	nvif_mthd(&disp->disp->object, 0, &args, sizeof(args));
+}
+
+static void
+nv50_hdmi_enable(struct drm_encoder *encoder, struct drm_display_mode *mode)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
+	struct nv50_disp *disp = nv50_disp(encoder->dev);
+	struct {
+		struct nv50_disp_mthd_v1 base;
+		struct nv50_disp_sor_hdmi_pwr_v0 pwr;
+		u8 infoframes[2 * 17]; /* two frames, up to 17 bytes each */
+	} args = {
+		.base.version = 1,
+		.base.method = NV50_DISP_MTHD_V1_SOR_HDMI_PWR,
+		.base.hasht  = nv_encoder->dcb->hasht,
+		.base.hashm  = (0xf0ff & nv_encoder->dcb->hashm) |
+			       (0x0100 << nv_crtc->index),
+		.pwr.state = 1,
+		.pwr.rekey = 56, /* binary driver, and tegra, constant */
+	};
+	struct nouveau_connector *nv_connector;
+	u32 max_ac_packet;
+	union hdmi_infoframe avi_frame;
+	union hdmi_infoframe vendor_frame;
+	int ret;
+	int size;
+
+	nv_connector = nouveau_encoder_connector_get(nv_encoder);
+	if (!drm_detect_hdmi_monitor(nv_connector->edid))
+		return;
+
+	ret = drm_hdmi_avi_infoframe_from_display_mode(&avi_frame.avi, mode,
+						       false);
+	if (!ret) {
+		/* We have an AVI InfoFrame, populate it to the display */
+		args.pwr.avi_infoframe_length
+			= hdmi_infoframe_pack(&avi_frame, args.infoframes, 17);
+	}
+
+	ret = drm_hdmi_vendor_infoframe_from_display_mode(&vendor_frame.vendor.hdmi,
+							  &nv_connector->base, mode);
+	if (!ret) {
+		/* We have a Vendor InfoFrame, populate it to the display */
+		args.pwr.vendor_infoframe_length
+			= hdmi_infoframe_pack(&vendor_frame,
+					      args.infoframes
+					      + args.pwr.avi_infoframe_length,
+					      17);
+	}
+
+	max_ac_packet  = mode->htotal - mode->hdisplay;
+	max_ac_packet -= args.pwr.rekey;
+	max_ac_packet -= 18; /* constant from tegra */
+	args.pwr.max_ac_packet = max_ac_packet / 32;
+
+	size = sizeof(args.base)
+		+ sizeof(args.pwr)
+		+ args.pwr.avi_infoframe_length
+		+ args.pwr.vendor_infoframe_length;
+	nvif_mthd(&disp->disp->object, 0, &args, size);
+	nv50_audio_enable(encoder, mode);
+}
+
+/******************************************************************************
+ * MST
+ *****************************************************************************/
+#define nv50_mstm(p) container_of((p), struct nv50_mstm, mgr)
+#define nv50_mstc(p) container_of((p), struct nv50_mstc, connector)
+#define nv50_msto(p) container_of((p), struct nv50_msto, encoder)
+
+struct nv50_mstm {
+	struct nouveau_encoder *outp;
+
+	struct drm_dp_mst_topology_mgr mgr;
+	struct nv50_msto *msto[4];
+
+	bool modified;
+	bool disabled;
+	int links;
+};
+
+struct nv50_mstc {
+	struct nv50_mstm *mstm;
+	struct drm_dp_mst_port *port;
+	struct drm_connector connector;
+
+	struct drm_display_mode *native;
+	struct edid *edid;
+
+	int pbn;
+};
+
+struct nv50_msto {
+	struct drm_encoder encoder;
+
+	struct nv50_head *head;
+	struct nv50_mstc *mstc;
+	bool disabled;
+};
+
+static struct drm_dp_payload *
+nv50_msto_payload(struct nv50_msto *msto)
+{
+	struct nouveau_drm *drm = nouveau_drm(msto->encoder.dev);
+	struct nv50_mstc *mstc = msto->mstc;
+	struct nv50_mstm *mstm = mstc->mstm;
+	int vcpi = mstc->port->vcpi.vcpi, i;
+
+	NV_ATOMIC(drm, "%s: vcpi %d\n", msto->encoder.name, vcpi);
+	for (i = 0; i < mstm->mgr.max_payloads; i++) {
+		struct drm_dp_payload *payload = &mstm->mgr.payloads[i];
+		NV_ATOMIC(drm, "%s: %d: vcpi %d start 0x%02x slots 0x%02x\n",
+			  mstm->outp->base.base.name, i, payload->vcpi,
+			  payload->start_slot, payload->num_slots);
+	}
+
+	for (i = 0; i < mstm->mgr.max_payloads; i++) {
+		struct drm_dp_payload *payload = &mstm->mgr.payloads[i];
+		if (payload->vcpi == vcpi)
+			return payload;
+	}
+
+	return NULL;
+}
+
+static void
+nv50_msto_cleanup(struct nv50_msto *msto)
+{
+	struct nouveau_drm *drm = nouveau_drm(msto->encoder.dev);
+	struct nv50_mstc *mstc = msto->mstc;
+	struct nv50_mstm *mstm = mstc->mstm;
+
+	NV_ATOMIC(drm, "%s: msto cleanup\n", msto->encoder.name);
+	if (mstc->port && mstc->port->vcpi.vcpi > 0 && !nv50_msto_payload(msto))
+		drm_dp_mst_deallocate_vcpi(&mstm->mgr, mstc->port);
+	if (msto->disabled) {
+		msto->mstc = NULL;
+		msto->head = NULL;
+		msto->disabled = false;
+	}
+}
+
+static void
+nv50_msto_prepare(struct nv50_msto *msto)
+{
+	struct nouveau_drm *drm = nouveau_drm(msto->encoder.dev);
+	struct nv50_mstc *mstc = msto->mstc;
+	struct nv50_mstm *mstm = mstc->mstm;
+	struct {
+		struct nv50_disp_mthd_v1 base;
+		struct nv50_disp_sor_dp_mst_vcpi_v0 vcpi;
+	} args = {
+		.base.version = 1,
+		.base.method = NV50_DISP_MTHD_V1_SOR_DP_MST_VCPI,
+		.base.hasht  = mstm->outp->dcb->hasht,
+		.base.hashm  = (0xf0ff & mstm->outp->dcb->hashm) |
+			       (0x0100 << msto->head->base.index),
+	};
+
+	NV_ATOMIC(drm, "%s: msto prepare\n", msto->encoder.name);
+	if (mstc->port && mstc->port->vcpi.vcpi > 0) {
+		struct drm_dp_payload *payload = nv50_msto_payload(msto);
+		if (payload) {
+			args.vcpi.start_slot = payload->start_slot;
+			args.vcpi.num_slots = payload->num_slots;
+			args.vcpi.pbn = mstc->port->vcpi.pbn;
+			args.vcpi.aligned_pbn = mstc->port->vcpi.aligned_pbn;
+		}
+	}
+
+	NV_ATOMIC(drm, "%s: %s: %02x %02x %04x %04x\n",
+		  msto->encoder.name, msto->head->base.base.name,
+		  args.vcpi.start_slot, args.vcpi.num_slots,
+		  args.vcpi.pbn, args.vcpi.aligned_pbn);
+	nvif_mthd(&drm->display->disp.object, 0, &args, sizeof(args));
+}
+
+static int
+nv50_msto_atomic_check(struct drm_encoder *encoder,
+		       struct drm_crtc_state *crtc_state,
+		       struct drm_connector_state *conn_state)
+{
+	struct nv50_mstc *mstc = nv50_mstc(conn_state->connector);
+	struct nv50_mstm *mstm = mstc->mstm;
+	int bpp = conn_state->connector->display_info.bpc * 3;
+	int slots;
+
+	mstc->pbn = drm_dp_calc_pbn_mode(crtc_state->adjusted_mode.clock, bpp);
+
+	slots = drm_dp_find_vcpi_slots(&mstm->mgr, mstc->pbn);
+	if (slots < 0)
+		return slots;
+
+	return nv50_outp_atomic_check_view(encoder, crtc_state, conn_state,
+					   mstc->native);
+}
+
+static void
+nv50_msto_enable(struct drm_encoder *encoder)
+{
+	struct nv50_head *head = nv50_head(encoder->crtc);
+	struct nv50_msto *msto = nv50_msto(encoder);
+	struct nv50_mstc *mstc = NULL;
+	struct nv50_mstm *mstm = NULL;
+	struct drm_connector *connector;
+	struct drm_connector_list_iter conn_iter;
+	u8 proto, depth;
+	int slots;
+	bool r;
+
+	drm_connector_list_iter_begin(encoder->dev, &conn_iter);
+	drm_for_each_connector_iter(connector, &conn_iter) {
+		if (connector->state->best_encoder == &msto->encoder) {
+			mstc = nv50_mstc(connector);
+			mstm = mstc->mstm;
+			break;
+		}
+	}
+	drm_connector_list_iter_end(&conn_iter);
+
+	if (WARN_ON(!mstc))
+		return;
+
+	slots = drm_dp_find_vcpi_slots(&mstm->mgr, mstc->pbn);
+	r = drm_dp_mst_allocate_vcpi(&mstm->mgr, mstc->port, mstc->pbn, slots);
+	WARN_ON(!r);
+
+	if (!mstm->links++)
+		nv50_outp_acquire(mstm->outp);
+
+	if (mstm->outp->link & 1)
+		proto = 0x8;
+	else
+		proto = 0x9;
+
+	switch (mstc->connector.display_info.bpc) {
+	case  6: depth = 0x2; break;
+	case  8: depth = 0x5; break;
+	case 10:
+	default: depth = 0x6; break;
+	}
+
+	mstm->outp->update(mstm->outp, head->base.index,
+			   &head->base.base.state->adjusted_mode, proto, depth);
+
+	msto->head = head;
+	msto->mstc = mstc;
+	mstm->modified = true;
+}
+
+static void
+nv50_msto_disable(struct drm_encoder *encoder)
+{
+	struct nv50_msto *msto = nv50_msto(encoder);
+	struct nv50_mstc *mstc = msto->mstc;
+	struct nv50_mstm *mstm = mstc->mstm;
+
+	if (mstc->port)
+		drm_dp_mst_reset_vcpi_slots(&mstm->mgr, mstc->port);
+
+	mstm->outp->update(mstm->outp, msto->head->base.index, NULL, 0, 0);
+	mstm->modified = true;
+	if (!--mstm->links)
+		mstm->disabled = true;
+	msto->disabled = true;
+}
+
+static const struct drm_encoder_helper_funcs
+nv50_msto_help = {
+	.disable = nv50_msto_disable,
+	.enable = nv50_msto_enable,
+	.atomic_check = nv50_msto_atomic_check,
+};
+
+static void
+nv50_msto_destroy(struct drm_encoder *encoder)
+{
+	struct nv50_msto *msto = nv50_msto(encoder);
+	drm_encoder_cleanup(&msto->encoder);
+	kfree(msto);
+}
+
+static const struct drm_encoder_funcs
+nv50_msto = {
+	.destroy = nv50_msto_destroy,
+};
+
+static int
+nv50_msto_new(struct drm_device *dev, u32 heads, const char *name, int id,
+	      struct nv50_msto **pmsto)
+{
+	struct nv50_msto *msto;
+	int ret;
+
+	if (!(msto = *pmsto = kzalloc(sizeof(*msto), GFP_KERNEL)))
+		return -ENOMEM;
+
+	ret = drm_encoder_init(dev, &msto->encoder, &nv50_msto,
+			       DRM_MODE_ENCODER_DPMST, "%s-mst-%d", name, id);
+	if (ret) {
+		kfree(*pmsto);
+		*pmsto = NULL;
+		return ret;
+	}
+
+	drm_encoder_helper_add(&msto->encoder, &nv50_msto_help);
+	msto->encoder.possible_crtcs = heads;
+	return 0;
+}
+
+static struct drm_encoder *
+nv50_mstc_atomic_best_encoder(struct drm_connector *connector,
+			      struct drm_connector_state *connector_state)
+{
+	struct nv50_head *head = nv50_head(connector_state->crtc);
+	struct nv50_mstc *mstc = nv50_mstc(connector);
+	if (mstc->port) {
+		struct nv50_mstm *mstm = mstc->mstm;
+		return &mstm->msto[head->base.index]->encoder;
+	}
+	return NULL;
+}
+
+static struct drm_encoder *
+nv50_mstc_best_encoder(struct drm_connector *connector)
+{
+	struct nv50_mstc *mstc = nv50_mstc(connector);
+	if (mstc->port) {
+		struct nv50_mstm *mstm = mstc->mstm;
+		return &mstm->msto[0]->encoder;
+	}
+	return NULL;
+}
+
+static enum drm_mode_status
+nv50_mstc_mode_valid(struct drm_connector *connector,
+		     struct drm_display_mode *mode)
+{
+	return MODE_OK;
+}
+
+static int
+nv50_mstc_get_modes(struct drm_connector *connector)
+{
+	struct nv50_mstc *mstc = nv50_mstc(connector);
+	int ret = 0;
+
+	mstc->edid = drm_dp_mst_get_edid(&mstc->connector, mstc->port->mgr, mstc->port);
+	drm_mode_connector_update_edid_property(&mstc->connector, mstc->edid);
+	if (mstc->edid)
+		ret = drm_add_edid_modes(&mstc->connector, mstc->edid);
+
+	if (!mstc->connector.display_info.bpc)
+		mstc->connector.display_info.bpc = 8;
+
+	if (mstc->native)
+		drm_mode_destroy(mstc->connector.dev, mstc->native);
+	mstc->native = nouveau_conn_native_mode(&mstc->connector);
+	return ret;
+}
+
+static const struct drm_connector_helper_funcs
+nv50_mstc_help = {
+	.get_modes = nv50_mstc_get_modes,
+	.mode_valid = nv50_mstc_mode_valid,
+	.best_encoder = nv50_mstc_best_encoder,
+	.atomic_best_encoder = nv50_mstc_atomic_best_encoder,
+};
+
+static enum drm_connector_status
+nv50_mstc_detect(struct drm_connector *connector, bool force)
+{
+	struct nv50_mstc *mstc = nv50_mstc(connector);
+	if (!mstc->port)
+		return connector_status_disconnected;
+	return drm_dp_mst_detect_port(connector, mstc->port->mgr, mstc->port);
+}
+
+static void
+nv50_mstc_destroy(struct drm_connector *connector)
+{
+	struct nv50_mstc *mstc = nv50_mstc(connector);
+	drm_connector_cleanup(&mstc->connector);
+	kfree(mstc);
+}
+
+static const struct drm_connector_funcs
+nv50_mstc = {
+	.reset = nouveau_conn_reset,
+	.detect = nv50_mstc_detect,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.destroy = nv50_mstc_destroy,
+	.atomic_duplicate_state = nouveau_conn_atomic_duplicate_state,
+	.atomic_destroy_state = nouveau_conn_atomic_destroy_state,
+	.atomic_set_property = nouveau_conn_atomic_set_property,
+	.atomic_get_property = nouveau_conn_atomic_get_property,
+};
+
+static int
+nv50_mstc_new(struct nv50_mstm *mstm, struct drm_dp_mst_port *port,
+	      const char *path, struct nv50_mstc **pmstc)
+{
+	struct drm_device *dev = mstm->outp->base.base.dev;
+	struct nv50_mstc *mstc;
+	int ret, i;
+
+	if (!(mstc = *pmstc = kzalloc(sizeof(*mstc), GFP_KERNEL)))
+		return -ENOMEM;
+	mstc->mstm = mstm;
+	mstc->port = port;
+
+	ret = drm_connector_init(dev, &mstc->connector, &nv50_mstc,
+				 DRM_MODE_CONNECTOR_DisplayPort);
+	if (ret) {
+		kfree(*pmstc);
+		*pmstc = NULL;
+		return ret;
+	}
+
+	drm_connector_helper_add(&mstc->connector, &nv50_mstc_help);
+
+	mstc->connector.funcs->reset(&mstc->connector);
+	nouveau_conn_attach_properties(&mstc->connector);
+
+	for (i = 0; i < ARRAY_SIZE(mstm->msto) && mstm->msto[i]; i++)
+		drm_mode_connector_attach_encoder(&mstc->connector, &mstm->msto[i]->encoder);
+
+	drm_object_attach_property(&mstc->connector.base, dev->mode_config.path_property, 0);
+	drm_object_attach_property(&mstc->connector.base, dev->mode_config.tile_property, 0);
+	drm_mode_connector_set_path_property(&mstc->connector, path);
+	return 0;
+}
+
+static void
+nv50_mstm_cleanup(struct nv50_mstm *mstm)
+{
+	struct nouveau_drm *drm = nouveau_drm(mstm->outp->base.base.dev);
+	struct drm_encoder *encoder;
+	int ret;
+
+	NV_ATOMIC(drm, "%s: mstm cleanup\n", mstm->outp->base.base.name);
+	ret = drm_dp_check_act_status(&mstm->mgr);
+
+	ret = drm_dp_update_payload_part2(&mstm->mgr);
+
+	drm_for_each_encoder(encoder, mstm->outp->base.base.dev) {
+		if (encoder->encoder_type == DRM_MODE_ENCODER_DPMST) {
+			struct nv50_msto *msto = nv50_msto(encoder);
+			struct nv50_mstc *mstc = msto->mstc;
+			if (mstc && mstc->mstm == mstm)
+				nv50_msto_cleanup(msto);
+		}
+	}
+
+	mstm->modified = false;
+}
+
+static void
+nv50_mstm_prepare(struct nv50_mstm *mstm)
+{
+	struct nouveau_drm *drm = nouveau_drm(mstm->outp->base.base.dev);
+	struct drm_encoder *encoder;
+	int ret;
+
+	NV_ATOMIC(drm, "%s: mstm prepare\n", mstm->outp->base.base.name);
+	ret = drm_dp_update_payload_part1(&mstm->mgr);
+
+	drm_for_each_encoder(encoder, mstm->outp->base.base.dev) {
+		if (encoder->encoder_type == DRM_MODE_ENCODER_DPMST) {
+			struct nv50_msto *msto = nv50_msto(encoder);
+			struct nv50_mstc *mstc = msto->mstc;
+			if (mstc && mstc->mstm == mstm)
+				nv50_msto_prepare(msto);
+		}
+	}
+
+	if (mstm->disabled) {
+		if (!mstm->links)
+			nv50_outp_release(mstm->outp);
+		mstm->disabled = false;
+	}
+}
+
+static void
+nv50_mstm_hotplug(struct drm_dp_mst_topology_mgr *mgr)
+{
+	struct nv50_mstm *mstm = nv50_mstm(mgr);
+	drm_kms_helper_hotplug_event(mstm->outp->base.base.dev);
+}
+
+static void
+nv50_mstm_destroy_connector(struct drm_dp_mst_topology_mgr *mgr,
+			    struct drm_connector *connector)
+{
+	struct nouveau_drm *drm = nouveau_drm(connector->dev);
+	struct nv50_mstc *mstc = nv50_mstc(connector);
+
+	drm_connector_unregister(&mstc->connector);
+
+	drm_fb_helper_remove_one_connector(&drm->fbcon->helper, &mstc->connector);
+
+	drm_modeset_lock(&drm->dev->mode_config.connection_mutex, NULL);
+	mstc->port = NULL;
+	drm_modeset_unlock(&drm->dev->mode_config.connection_mutex);
+
+	drm_connector_unreference(&mstc->connector);
+}
+
+static void
+nv50_mstm_register_connector(struct drm_connector *connector)
+{
+	struct nouveau_drm *drm = nouveau_drm(connector->dev);
+
+	drm_fb_helper_add_one_connector(&drm->fbcon->helper, connector);
+
+	drm_connector_register(connector);
+}
+
+static struct drm_connector *
+nv50_mstm_add_connector(struct drm_dp_mst_topology_mgr *mgr,
+			struct drm_dp_mst_port *port, const char *path)
+{
+	struct nv50_mstm *mstm = nv50_mstm(mgr);
+	struct nv50_mstc *mstc;
+	int ret;
+
+	ret = nv50_mstc_new(mstm, port, path, &mstc);
+	if (ret) {
+		if (mstc)
+			mstc->connector.funcs->destroy(&mstc->connector);
+		return NULL;
+	}
+
+	return &mstc->connector;
+}
+
+static const struct drm_dp_mst_topology_cbs
+nv50_mstm = {
+	.add_connector = nv50_mstm_add_connector,
+	.register_connector = nv50_mstm_register_connector,
+	.destroy_connector = nv50_mstm_destroy_connector,
+	.hotplug = nv50_mstm_hotplug,
+};
+
+void
+nv50_mstm_service(struct nv50_mstm *mstm)
+{
+	struct drm_dp_aux *aux = mstm ? mstm->mgr.aux : NULL;
+	bool handled = true;
+	int ret;
+	u8 esi[8] = {};
+
+	if (!aux)
+		return;
+
+	while (handled) {
+		ret = drm_dp_dpcd_read(aux, DP_SINK_COUNT_ESI, esi, 8);
+		if (ret != 8) {
+			drm_dp_mst_topology_mgr_set_mst(&mstm->mgr, false);
+			return;
+		}
+
+		drm_dp_mst_hpd_irq(&mstm->mgr, esi, &handled);
+		if (!handled)
+			break;
+
+		drm_dp_dpcd_write(aux, DP_SINK_COUNT_ESI + 1, &esi[1], 3);
+	}
+}
+
+void
+nv50_mstm_remove(struct nv50_mstm *mstm)
+{
+	if (mstm)
+		drm_dp_mst_topology_mgr_set_mst(&mstm->mgr, false);
+}
+
+static int
+nv50_mstm_enable(struct nv50_mstm *mstm, u8 dpcd, int state)
+{
+	struct nouveau_encoder *outp = mstm->outp;
+	struct {
+		struct nv50_disp_mthd_v1 base;
+		struct nv50_disp_sor_dp_mst_link_v0 mst;
+	} args = {
+		.base.version = 1,
+		.base.method = NV50_DISP_MTHD_V1_SOR_DP_MST_LINK,
+		.base.hasht = outp->dcb->hasht,
+		.base.hashm = outp->dcb->hashm,
+		.mst.state = state,
+	};
+	struct nouveau_drm *drm = nouveau_drm(outp->base.base.dev);
+	struct nvif_object *disp = &drm->display->disp.object;
+	int ret;
+
+	if (dpcd >= 0x12) {
+		ret = drm_dp_dpcd_readb(mstm->mgr.aux, DP_MSTM_CTRL, &dpcd);
+		if (ret < 0)
+			return ret;
+
+		dpcd &= ~DP_MST_EN;
+		if (state)
+			dpcd |= DP_MST_EN;
+
+		ret = drm_dp_dpcd_writeb(mstm->mgr.aux, DP_MSTM_CTRL, dpcd);
+		if (ret < 0)
+			return ret;
+	}
+
+	return nvif_mthd(disp, 0, &args, sizeof(args));
+}
+
+int
+nv50_mstm_detect(struct nv50_mstm *mstm, u8 dpcd[8], int allow)
+{
+	int ret, state = 0;
+
+	if (!mstm)
+		return 0;
+
+	if (dpcd[0] >= 0x12) {
+		ret = drm_dp_dpcd_readb(mstm->mgr.aux, DP_MSTM_CAP, &dpcd[1]);
+		if (ret < 0)
+			return ret;
+
+		if (!(dpcd[1] & DP_MST_CAP))
+			dpcd[0] = 0x11;
+		else
+			state = allow;
+	}
+
+	ret = nv50_mstm_enable(mstm, dpcd[0], state);
+	if (ret)
+		return ret;
+
+	ret = drm_dp_mst_topology_mgr_set_mst(&mstm->mgr, state);
+	if (ret)
+		return nv50_mstm_enable(mstm, dpcd[0], 0);
+
+	return mstm->mgr.mst_state;
+}
+
+static void
+nv50_mstm_fini(struct nv50_mstm *mstm)
+{
+	if (mstm && mstm->mgr.mst_state)
+		drm_dp_mst_topology_mgr_suspend(&mstm->mgr);
+}
+
+static void
+nv50_mstm_init(struct nv50_mstm *mstm)
+{
+	if (mstm && mstm->mgr.mst_state)
+		drm_dp_mst_topology_mgr_resume(&mstm->mgr);
+}
+
+static void
+nv50_mstm_del(struct nv50_mstm **pmstm)
+{
+	struct nv50_mstm *mstm = *pmstm;
+	if (mstm) {
+		kfree(*pmstm);
+		*pmstm = NULL;
+	}
+}
+
+static int
+nv50_mstm_new(struct nouveau_encoder *outp, struct drm_dp_aux *aux, int aux_max,
+	      int conn_base_id, struct nv50_mstm **pmstm)
+{
+	const int max_payloads = hweight8(outp->dcb->heads);
+	struct drm_device *dev = outp->base.base.dev;
+	struct nv50_mstm *mstm;
+	int ret, i;
+	u8 dpcd;
+
+	/* This is a workaround for some monitors not functioning
+	 * correctly in MST mode on initial module load.  I think
+	 * some bad interaction with the VBIOS may be responsible.
+	 *
+	 * A good ol' off and on again seems to work here ;)
+	 */
+	ret = drm_dp_dpcd_readb(aux, DP_DPCD_REV, &dpcd);
+	if (ret >= 0 && dpcd >= 0x12)
+		drm_dp_dpcd_writeb(aux, DP_MSTM_CTRL, 0);
+
+	if (!(mstm = *pmstm = kzalloc(sizeof(*mstm), GFP_KERNEL)))
+		return -ENOMEM;
+	mstm->outp = outp;
+	mstm->mgr.cbs = &nv50_mstm;
+
+	ret = drm_dp_mst_topology_mgr_init(&mstm->mgr, dev, aux, aux_max,
+					   max_payloads, conn_base_id);
+	if (ret)
+		return ret;
+
+	for (i = 0; i < max_payloads; i++) {
+		ret = nv50_msto_new(dev, outp->dcb->heads, outp->base.base.name,
+				    i, &mstm->msto[i]);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
+/******************************************************************************
+ * SOR
+ *****************************************************************************/
+static void
+nv50_sor_update(struct nouveau_encoder *nv_encoder, u8 head,
+		struct drm_display_mode *mode, u8 proto, u8 depth)
+{
+	struct nv50_dmac *core = &nv50_mast(nv_encoder->base.base.dev)->base;
+	u32 *push;
+
+	if (!mode) {
+		nv_encoder->ctrl &= ~BIT(head);
+		if (!(nv_encoder->ctrl & 0x0000000f))
+			nv_encoder->ctrl = 0;
+	} else {
+		nv_encoder->ctrl |= proto << 8;
+		nv_encoder->ctrl |= BIT(head);
+	}
+
+	if ((push = evo_wait(core, 6))) {
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
+			if (mode) {
+				if (mode->flags & DRM_MODE_FLAG_NHSYNC)
+					nv_encoder->ctrl |= 0x00001000;
+				if (mode->flags & DRM_MODE_FLAG_NVSYNC)
+					nv_encoder->ctrl |= 0x00002000;
+				nv_encoder->ctrl |= depth << 16;
+			}
+			evo_mthd(push, 0x0600 + (nv_encoder->or * 0x40), 1);
+		} else {
+			if (mode) {
+				u32 magic = 0x31ec6000 | (head << 25);
+				u32 syncs = 0x00000001;
+				if (mode->flags & DRM_MODE_FLAG_NHSYNC)
+					syncs |= 0x00000008;
+				if (mode->flags & DRM_MODE_FLAG_NVSYNC)
+					syncs |= 0x00000010;
+				if (mode->flags & DRM_MODE_FLAG_INTERLACE)
+					magic |= 0x00000001;
+
+				evo_mthd(push, 0x0404 + (head * 0x300), 2);
+				evo_data(push, syncs | (depth << 6));
+				evo_data(push, magic);
+			}
+			evo_mthd(push, 0x0200 + (nv_encoder->or * 0x20), 1);
+		}
+		evo_data(push, nv_encoder->ctrl);
+		evo_kick(push, core);
+	}
+}
+
+static void
+nv50_sor_disable(struct drm_encoder *encoder)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(nv_encoder->crtc);
+
+	nv_encoder->crtc = NULL;
+
+	if (nv_crtc) {
+		struct nvkm_i2c_aux *aux = nv_encoder->aux;
+		u8 pwr;
+
+		if (aux) {
+			int ret = nvkm_rdaux(aux, DP_SET_POWER, &pwr, 1);
+			if (ret == 0) {
+				pwr &= ~DP_SET_POWER_MASK;
+				pwr |=  DP_SET_POWER_D3;
+				nvkm_wraux(aux, DP_SET_POWER, &pwr, 1);
+			}
+		}
+
+		nv_encoder->update(nv_encoder, nv_crtc->index, NULL, 0, 0);
+		nv50_audio_disable(encoder, nv_crtc);
+		nv50_hdmi_disable(&nv_encoder->base.base, nv_crtc);
+		nv50_outp_release(nv_encoder);
+	}
+}
+
+static void
+nv50_sor_enable(struct drm_encoder *encoder)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
+	struct drm_display_mode *mode = &nv_crtc->base.state->adjusted_mode;
+	struct {
+		struct nv50_disp_mthd_v1 base;
+		struct nv50_disp_sor_lvds_script_v0 lvds;
+	} lvds = {
+		.base.version = 1,
+		.base.method  = NV50_DISP_MTHD_V1_SOR_LVDS_SCRIPT,
+		.base.hasht   = nv_encoder->dcb->hasht,
+		.base.hashm   = nv_encoder->dcb->hashm,
+	};
+	struct nv50_disp *disp = nv50_disp(encoder->dev);
+	struct drm_device *dev = encoder->dev;
+	struct nouveau_drm *drm = nouveau_drm(dev);
+	struct nouveau_connector *nv_connector;
+	struct nvbios *bios = &drm->vbios;
+	u8 proto = 0xf;
+	u8 depth = 0x0;
+
+	nv_connector = nouveau_encoder_connector_get(nv_encoder);
+	nv_encoder->crtc = encoder->crtc;
+	nv50_outp_acquire(nv_encoder);
+
+	switch (nv_encoder->dcb->type) {
+	case DCB_OUTPUT_TMDS:
+		if (nv_encoder->link & 1) {
+			proto = 0x1;
+			/* Only enable dual-link if:
+			 *  - Need to (i.e. rate > 165MHz)
+			 *  - DCB says we can
+			 *  - Not an HDMI monitor, since there's no dual-link
+			 *    on HDMI.
+			 */
+			if (mode->clock >= 165000 &&
+			    nv_encoder->dcb->duallink_possible &&
+			    !drm_detect_hdmi_monitor(nv_connector->edid))
+				proto |= 0x4;
+		} else {
+			proto = 0x2;
+		}
+
+		nv50_hdmi_enable(&nv_encoder->base.base, mode);
+		break;
+	case DCB_OUTPUT_LVDS:
+		proto = 0x0;
+
+		if (bios->fp_no_ddc) {
+			if (bios->fp.dual_link)
+				lvds.lvds.script |= 0x0100;
+			if (bios->fp.if_is_24bit)
+				lvds.lvds.script |= 0x0200;
+		} else {
+			if (nv_connector->type == DCB_CONNECTOR_LVDS_SPWG) {
+				if (((u8 *)nv_connector->edid)[121] == 2)
+					lvds.lvds.script |= 0x0100;
+			} else
+			if (mode->clock >= bios->fp.duallink_transition_clk) {
+				lvds.lvds.script |= 0x0100;
+			}
+
+			if (lvds.lvds.script & 0x0100) {
+				if (bios->fp.strapless_is_24bit & 2)
+					lvds.lvds.script |= 0x0200;
+			} else {
+				if (bios->fp.strapless_is_24bit & 1)
+					lvds.lvds.script |= 0x0200;
+			}
+
+			if (nv_connector->base.display_info.bpc == 8)
+				lvds.lvds.script |= 0x0200;
+		}
+
+		nvif_mthd(&disp->disp->object, 0, &lvds, sizeof(lvds));
+		break;
+	case DCB_OUTPUT_DP:
+		if (nv_connector->base.display_info.bpc == 6)
+			depth = 0x2;
+		else
+		if (nv_connector->base.display_info.bpc == 8)
+			depth = 0x5;
+		else
+			depth = 0x6;
+
+		if (nv_encoder->link & 1)
+			proto = 0x8;
+		else
+			proto = 0x9;
+
+		nv50_audio_enable(encoder, mode);
+		break;
+	default:
+		BUG();
+		break;
+	}
+
+	nv_encoder->update(nv_encoder, nv_crtc->index, mode, proto, depth);
+}
+
+static const struct drm_encoder_helper_funcs
+nv50_sor_help = {
+	.atomic_check = nv50_outp_atomic_check,
+	.enable = nv50_sor_enable,
+	.disable = nv50_sor_disable,
+};
+
+static void
+nv50_sor_destroy(struct drm_encoder *encoder)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	nv50_mstm_del(&nv_encoder->dp.mstm);
+	drm_encoder_cleanup(encoder);
+	kfree(encoder);
+}
+
+static const struct drm_encoder_funcs
+nv50_sor_func = {
+	.destroy = nv50_sor_destroy,
+};
+
+static int
+nv50_sor_create(struct drm_connector *connector, struct dcb_output *dcbe)
+{
+	struct nouveau_connector *nv_connector = nouveau_connector(connector);
+	struct nouveau_drm *drm = nouveau_drm(connector->dev);
+	struct nvkm_i2c *i2c = nvxx_i2c(&drm->client.device);
+	struct nouveau_encoder *nv_encoder;
+	struct drm_encoder *encoder;
+	int type, ret;
+
+	switch (dcbe->type) {
+	case DCB_OUTPUT_LVDS: type = DRM_MODE_ENCODER_LVDS; break;
+	case DCB_OUTPUT_TMDS:
+	case DCB_OUTPUT_DP:
+	default:
+		type = DRM_MODE_ENCODER_TMDS;
+		break;
+	}
+
+	nv_encoder = kzalloc(sizeof(*nv_encoder), GFP_KERNEL);
+	if (!nv_encoder)
+		return -ENOMEM;
+	nv_encoder->dcb = dcbe;
+	nv_encoder->update = nv50_sor_update;
+
+	encoder = to_drm_encoder(nv_encoder);
+	encoder->possible_crtcs = dcbe->heads;
+	encoder->possible_clones = 0;
+	drm_encoder_init(connector->dev, encoder, &nv50_sor_func, type,
+			 "sor-%04x-%04x", dcbe->hasht, dcbe->hashm);
+	drm_encoder_helper_add(encoder, &nv50_sor_help);
+
+	drm_mode_connector_attach_encoder(connector, encoder);
+
+	if (dcbe->type == DCB_OUTPUT_DP) {
+		struct nv50_disp *disp = nv50_disp(encoder->dev);
+		struct nvkm_i2c_aux *aux =
+			nvkm_i2c_aux_find(i2c, dcbe->i2c_index);
+		if (aux) {
+			if (disp->disp->object.oclass < GF110_DISP) {
+				/* HW has no support for address-only
+				 * transactions, so we're required to
+				 * use custom I2C-over-AUX code.
+				 */
+				nv_encoder->i2c = &aux->i2c;
+			} else {
+				nv_encoder->i2c = &nv_connector->aux.ddc;
+			}
+			nv_encoder->aux = aux;
+		}
+
+		/*TODO: Use DP Info Table to check for support. */
+		if (disp->disp->object.oclass >= GF110_DISP) {
+			ret = nv50_mstm_new(nv_encoder, &nv_connector->aux, 16,
+					    nv_connector->base.base.id,
+					    &nv_encoder->dp.mstm);
+			if (ret)
+				return ret;
+		}
+	} else {
+		struct nvkm_i2c_bus *bus =
+			nvkm_i2c_bus_find(i2c, dcbe->i2c_index);
+		if (bus)
+			nv_encoder->i2c = &bus->i2c;
+	}
+
+	return 0;
+}
+
+/******************************************************************************
+ * PIOR
+ *****************************************************************************/
+static int
+nv50_pior_atomic_check(struct drm_encoder *encoder,
+		       struct drm_crtc_state *crtc_state,
+		       struct drm_connector_state *conn_state)
+{
+	int ret = nv50_outp_atomic_check(encoder, crtc_state, conn_state);
+	if (ret)
+		return ret;
+	crtc_state->adjusted_mode.clock *= 2;
+	return 0;
+}
+
+static void
+nv50_pior_disable(struct drm_encoder *encoder)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nv50_mast *mast = nv50_mast(encoder->dev);
+	const int or = nv_encoder->or;
+	u32 *push;
+
+	if (nv_encoder->crtc) {
+		push = evo_wait(mast, 4);
+		if (push) {
+			if (nv50_vers(mast) < GF110_DISP_CORE_CHANNEL_DMA) {
+				evo_mthd(push, 0x0700 + (or * 0x040), 1);
+				evo_data(push, 0x00000000);
+			}
+			evo_kick(push, mast);
+		}
+	}
+
+	nv_encoder->crtc = NULL;
+	nv50_outp_release(nv_encoder);
+}
+
+static void
+nv50_pior_enable(struct drm_encoder *encoder)
+{
+	struct nv50_mast *mast = nv50_mast(encoder->dev);
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
+	struct nouveau_connector *nv_connector;
+	struct drm_display_mode *mode = &nv_crtc->base.state->adjusted_mode;
+	u8 owner = 1 << nv_crtc->index;
+	u8 proto, depth;
+	u32 *push;
+
+	nv50_outp_acquire(nv_encoder);
+
+	nv_connector = nouveau_encoder_connector_get(nv_encoder);
+	switch (nv_connector->base.display_info.bpc) {
+	case 10: depth = 0x6; break;
+	case  8: depth = 0x5; break;
+	case  6: depth = 0x2; break;
+	default: depth = 0x0; break;
+	}
+
+	switch (nv_encoder->dcb->type) {
+	case DCB_OUTPUT_TMDS:
+	case DCB_OUTPUT_DP:
+		proto = 0x0;
+		break;
+	default:
+		BUG();
+		break;
+	}
+
+	push = evo_wait(mast, 8);
+	if (push) {
+		if (nv50_vers(mast) < GF110_DISP_CORE_CHANNEL_DMA) {
+			u32 ctrl = (depth << 16) | (proto << 8) | owner;
+			if (mode->flags & DRM_MODE_FLAG_NHSYNC)
+				ctrl |= 0x00001000;
+			if (mode->flags & DRM_MODE_FLAG_NVSYNC)
+				ctrl |= 0x00002000;
+			evo_mthd(push, 0x0700 + (nv_encoder->or * 0x040), 1);
+			evo_data(push, ctrl);
+		}
+
+		evo_kick(push, mast);
+	}
+
+	nv_encoder->crtc = encoder->crtc;
+}
+
+static const struct drm_encoder_helper_funcs
+nv50_pior_help = {
+	.atomic_check = nv50_pior_atomic_check,
+	.enable = nv50_pior_enable,
+	.disable = nv50_pior_disable,
+};
+
+static void
+nv50_pior_destroy(struct drm_encoder *encoder)
+{
+	drm_encoder_cleanup(encoder);
+	kfree(encoder);
+}
+
+static const struct drm_encoder_funcs
+nv50_pior_func = {
+	.destroy = nv50_pior_destroy,
+};
+
+static int
+nv50_pior_create(struct drm_connector *connector, struct dcb_output *dcbe)
+{
+	struct nouveau_connector *nv_connector = nouveau_connector(connector);
+	struct nouveau_drm *drm = nouveau_drm(connector->dev);
+	struct nvkm_i2c *i2c = nvxx_i2c(&drm->client.device);
+	struct nvkm_i2c_bus *bus = NULL;
+	struct nvkm_i2c_aux *aux = NULL;
+	struct i2c_adapter *ddc;
+	struct nouveau_encoder *nv_encoder;
+	struct drm_encoder *encoder;
+	int type;
+
+	switch (dcbe->type) {
+	case DCB_OUTPUT_TMDS:
+		bus  = nvkm_i2c_bus_find(i2c, NVKM_I2C_BUS_EXT(dcbe->extdev));
+		ddc  = bus ? &bus->i2c : NULL;
+		type = DRM_MODE_ENCODER_TMDS;
+		break;
+	case DCB_OUTPUT_DP:
+		aux  = nvkm_i2c_aux_find(i2c, NVKM_I2C_AUX_EXT(dcbe->extdev));
+		ddc  = aux ? &nv_connector->aux.ddc : NULL;
+		type = DRM_MODE_ENCODER_TMDS;
+		break;
+	default:
+		return -ENODEV;
+	}
+
+	nv_encoder = kzalloc(sizeof(*nv_encoder), GFP_KERNEL);
+	if (!nv_encoder)
+		return -ENOMEM;
+	nv_encoder->dcb = dcbe;
+	nv_encoder->i2c = ddc;
+	nv_encoder->aux = aux;
+
+	encoder = to_drm_encoder(nv_encoder);
+	encoder->possible_crtcs = dcbe->heads;
+	encoder->possible_clones = 0;
+	drm_encoder_init(connector->dev, encoder, &nv50_pior_func, type,
+			 "pior-%04x-%04x", dcbe->hasht, dcbe->hashm);
+	drm_encoder_helper_add(encoder, &nv50_pior_help);
+
+	drm_mode_connector_attach_encoder(connector, encoder);
+	return 0;
+}
+
+/******************************************************************************
+ * Atomic
+ *****************************************************************************/
+
+static void
+nv50_disp_atomic_commit_core(struct nouveau_drm *drm, u32 interlock)
+{
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	struct nv50_dmac *core = &disp->mast.base;
+	struct nv50_mstm *mstm;
+	struct drm_encoder *encoder;
+	u32 *push;
+
+	NV_ATOMIC(drm, "commit core %08x\n", interlock);
+
+	drm_for_each_encoder(encoder, drm->dev) {
+		if (encoder->encoder_type != DRM_MODE_ENCODER_DPMST) {
+			mstm = nouveau_encoder(encoder)->dp.mstm;
+			if (mstm && mstm->modified)
+				nv50_mstm_prepare(mstm);
+		}
+	}
+
+	if ((push = evo_wait(core, 5))) {
+		evo_mthd(push, 0x0084, 1);
+		evo_data(push, 0x80000000);
+		evo_mthd(push, 0x0080, 2);
+		evo_data(push, interlock);
+		evo_data(push, 0x00000000);
+		nouveau_bo_wr32(disp->sync, 0, 0x00000000);
+		evo_kick(push, core);
+		if (nvif_msec(&drm->client.device, 2000ULL,
+			if (nouveau_bo_rd32(disp->sync, 0))
+				break;
+			usleep_range(1, 2);
+		) < 0)
+			NV_ERROR(drm, "EVO timeout\n");
+	}
+
+	drm_for_each_encoder(encoder, drm->dev) {
+		if (encoder->encoder_type != DRM_MODE_ENCODER_DPMST) {
+			mstm = nouveau_encoder(encoder)->dp.mstm;
+			if (mstm && mstm->modified)
+				nv50_mstm_cleanup(mstm);
+		}
+	}
+}
+
+static void
+nv50_disp_atomic_commit_tail(struct drm_atomic_state *state)
+{
+	struct drm_device *dev = state->dev;
+	struct drm_crtc_state *new_crtc_state, *old_crtc_state;
+	struct drm_crtc *crtc;
+	struct drm_plane_state *new_plane_state;
+	struct drm_plane *plane;
+	struct nouveau_drm *drm = nouveau_drm(dev);
+	struct nv50_disp *disp = nv50_disp(dev);
+	struct nv50_atom *atom = nv50_atom(state);
+	struct nv50_outp_atom *outp, *outt;
+	u32 interlock_core = 0;
+	u32 interlock_chan = 0;
+	int i;
+
+	NV_ATOMIC(drm, "commit %d %d\n", atom->lock_core, atom->flush_disable);
+	drm_atomic_helper_wait_for_fences(dev, state, false);
+	drm_atomic_helper_wait_for_dependencies(state);
+	drm_atomic_helper_update_legacy_modeset_state(dev, state);
+
+	if (atom->lock_core)
+		mutex_lock(&disp->mutex);
+
+	/* Disable head(s). */
+	for_each_oldnew_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) {
+		struct nv50_head_atom *asyh = nv50_head_atom(new_crtc_state);
+		struct nv50_head *head = nv50_head(crtc);
+
+		NV_ATOMIC(drm, "%s: clr %04x (set %04x)\n", crtc->name,
+			  asyh->clr.mask, asyh->set.mask);
+		if (old_crtc_state->active && !new_crtc_state->active)
+			drm_crtc_vblank_off(crtc);
+
+		if (asyh->clr.mask) {
+			nv50_head_flush_clr(head, asyh, atom->flush_disable);
+			interlock_core |= 1;
+		}
+	}
+
+	/* Disable plane(s). */
+	for_each_new_plane_in_state(state, plane, new_plane_state, i) {
+		struct nv50_wndw_atom *asyw = nv50_wndw_atom(new_plane_state);
+		struct nv50_wndw *wndw = nv50_wndw(plane);
+
+		NV_ATOMIC(drm, "%s: clr %02x (set %02x)\n", plane->name,
+			  asyw->clr.mask, asyw->set.mask);
+		if (!asyw->clr.mask)
+			continue;
+
+		interlock_chan |= nv50_wndw_flush_clr(wndw, interlock_core,
+						      atom->flush_disable,
+						      asyw);
+	}
+
+	/* Disable output path(s). */
+	list_for_each_entry(outp, &atom->outp, head) {
+		const struct drm_encoder_helper_funcs *help;
+		struct drm_encoder *encoder;
+
+		encoder = outp->encoder;
+		help = encoder->helper_private;
+
+		NV_ATOMIC(drm, "%s: clr %02x (set %02x)\n", encoder->name,
+			  outp->clr.mask, outp->set.mask);
+
+		if (outp->clr.mask) {
+			help->disable(encoder);
+			interlock_core |= 1;
+			if (outp->flush_disable) {
+				nv50_disp_atomic_commit_core(drm, interlock_chan);
+				interlock_core = 0;
+				interlock_chan = 0;
+			}
+		}
+	}
+
+	/* Flush disable. */
+	if (interlock_core) {
+		if (atom->flush_disable) {
+			nv50_disp_atomic_commit_core(drm, interlock_chan);
+			interlock_core = 0;
+			interlock_chan = 0;
+		}
+	}
+
+	/* Update output path(s). */
+	list_for_each_entry_safe(outp, outt, &atom->outp, head) {
+		const struct drm_encoder_helper_funcs *help;
+		struct drm_encoder *encoder;
+
+		encoder = outp->encoder;
+		help = encoder->helper_private;
+
+		NV_ATOMIC(drm, "%s: set %02x (clr %02x)\n", encoder->name,
+			  outp->set.mask, outp->clr.mask);
+
+		if (outp->set.mask) {
+			help->enable(encoder);
+			interlock_core = 1;
+		}
+
+		list_del(&outp->head);
+		kfree(outp);
+	}
+
+	/* Update head(s). */
+	for_each_oldnew_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) {
+		struct nv50_head_atom *asyh = nv50_head_atom(new_crtc_state);
+		struct nv50_head *head = nv50_head(crtc);
+
+		NV_ATOMIC(drm, "%s: set %04x (clr %04x)\n", crtc->name,
+			  asyh->set.mask, asyh->clr.mask);
+
+		if (asyh->set.mask) {
+			nv50_head_flush_set(head, asyh);
+			interlock_core = 1;
+		}
+
+		if (new_crtc_state->active) {
+			if (!old_crtc_state->active)
+				drm_crtc_vblank_on(crtc);
+			if (new_crtc_state->event)
+				drm_crtc_vblank_get(crtc);
+		}
+	}
+
+	/* Update plane(s). */
+	for_each_new_plane_in_state(state, plane, new_plane_state, i) {
+		struct nv50_wndw_atom *asyw = nv50_wndw_atom(new_plane_state);
+		struct nv50_wndw *wndw = nv50_wndw(plane);
+
+		NV_ATOMIC(drm, "%s: set %02x (clr %02x)\n", plane->name,
+			  asyw->set.mask, asyw->clr.mask);
+		if ( !asyw->set.mask &&
+		    (!asyw->clr.mask || atom->flush_disable))
+			continue;
+
+		interlock_chan |= nv50_wndw_flush_set(wndw, interlock_core, asyw);
+	}
+
+	/* Flush update. */
+	if (interlock_core) {
+		if (!interlock_chan && atom->state.legacy_cursor_update) {
+			u32 *push = evo_wait(&disp->mast, 2);
+			if (push) {
+				evo_mthd(push, 0x0080, 1);
+				evo_data(push, 0x00000000);
+				evo_kick(push, &disp->mast);
+			}
+		} else {
+			nv50_disp_atomic_commit_core(drm, interlock_chan);
+		}
+	}
+
+	if (atom->lock_core)
+		mutex_unlock(&disp->mutex);
+
+	/* Wait for HW to signal completion. */
+	for_each_new_plane_in_state(state, plane, new_plane_state, i) {
+		struct nv50_wndw_atom *asyw = nv50_wndw_atom(new_plane_state);
+		struct nv50_wndw *wndw = nv50_wndw(plane);
+		int ret = nv50_wndw_wait_armed(wndw, asyw);
+		if (ret)
+			NV_ERROR(drm, "%s: timeout\n", plane->name);
+	}
+
+	for_each_new_crtc_in_state(state, crtc, new_crtc_state, i) {
+		if (new_crtc_state->event) {
+			unsigned long flags;
+			/* Get correct count/ts if racing with vblank irq */
+			if (new_crtc_state->active)
+				drm_crtc_accurate_vblank_count(crtc);
+			spin_lock_irqsave(&crtc->dev->event_lock, flags);
+			drm_crtc_send_vblank_event(crtc, new_crtc_state->event);
+			spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
+
+			new_crtc_state->event = NULL;
+			if (new_crtc_state->active)
+				drm_crtc_vblank_put(crtc);
+		}
+	}
+
+	drm_atomic_helper_commit_hw_done(state);
+	drm_atomic_helper_cleanup_planes(dev, state);
+	drm_atomic_helper_commit_cleanup_done(state);
+	drm_atomic_state_put(state);
+}
+
+static void
+nv50_disp_atomic_commit_work(struct work_struct *work)
+{
+	struct drm_atomic_state *state =
+		container_of(work, typeof(*state), commit_work);
+	nv50_disp_atomic_commit_tail(state);
+}
+
+static int
+nv50_disp_atomic_commit(struct drm_device *dev,
+			struct drm_atomic_state *state, bool nonblock)
+{
+	struct nouveau_drm *drm = nouveau_drm(dev);
+	struct nv50_disp *disp = nv50_disp(dev);
+	struct drm_plane_state *new_plane_state;
+	struct drm_plane *plane;
+	struct drm_crtc *crtc;
+	bool active = false;
+	int ret, i;
+
+	ret = pm_runtime_get_sync(dev->dev);
+	if (ret < 0 && ret != -EACCES)
+		return ret;
+
+	ret = drm_atomic_helper_setup_commit(state, nonblock);
+	if (ret)
+		goto done;
+
+	INIT_WORK(&state->commit_work, nv50_disp_atomic_commit_work);
+
+	ret = drm_atomic_helper_prepare_planes(dev, state);
+	if (ret)
+		goto done;
+
+	if (!nonblock) {
+		ret = drm_atomic_helper_wait_for_fences(dev, state, true);
+		if (ret)
+			goto err_cleanup;
+	}
+
+	ret = drm_atomic_helper_swap_state(state, true);
+	if (ret)
+		goto err_cleanup;
+
+	for_each_new_plane_in_state(state, plane, new_plane_state, i) {
+		struct nv50_wndw_atom *asyw = nv50_wndw_atom(new_plane_state);
+		struct nv50_wndw *wndw = nv50_wndw(plane);
+
+		if (asyw->set.image) {
+			asyw->ntfy.handle = wndw->dmac->sync.handle;
+			asyw->ntfy.offset = wndw->ntfy;
+			asyw->ntfy.awaken = false;
+			asyw->set.ntfy = true;
+			nouveau_bo_wr32(disp->sync, wndw->ntfy / 4, 0x00000000);
+			wndw->ntfy ^= 0x10;
+		}
+	}
+
+	drm_atomic_state_get(state);
+
+	if (nonblock)
+		queue_work(system_unbound_wq, &state->commit_work);
+	else
+		nv50_disp_atomic_commit_tail(state);
+
+	drm_for_each_crtc(crtc, dev) {
+		if (crtc->state->enable) {
+			if (!drm->have_disp_power_ref) {
+				drm->have_disp_power_ref = true;
+				return 0;
+			}
+			active = true;
+			break;
+		}
+	}
+
+	if (!active && drm->have_disp_power_ref) {
+		pm_runtime_put_autosuspend(dev->dev);
+		drm->have_disp_power_ref = false;
+	}
+
+err_cleanup:
+	if (ret)
+		drm_atomic_helper_cleanup_planes(dev, state);
+done:
+	pm_runtime_put_autosuspend(dev->dev);
+	return ret;
+}
+
+static struct nv50_outp_atom *
+nv50_disp_outp_atomic_add(struct nv50_atom *atom, struct drm_encoder *encoder)
+{
+	struct nv50_outp_atom *outp;
+
+	list_for_each_entry(outp, &atom->outp, head) {
+		if (outp->encoder == encoder)
+			return outp;
+	}
+
+	outp = kzalloc(sizeof(*outp), GFP_KERNEL);
+	if (!outp)
+		return ERR_PTR(-ENOMEM);
+
+	list_add(&outp->head, &atom->outp);
+	outp->encoder = encoder;
+	return outp;
+}
+
+static int
+nv50_disp_outp_atomic_check_clr(struct nv50_atom *atom,
+				struct drm_connector_state *old_connector_state)
+{
+	struct drm_encoder *encoder = old_connector_state->best_encoder;
+	struct drm_crtc_state *old_crtc_state, *new_crtc_state;
+	struct drm_crtc *crtc;
+	struct nv50_outp_atom *outp;
+
+	if (!(crtc = old_connector_state->crtc))
+		return 0;
+
+	old_crtc_state = drm_atomic_get_old_crtc_state(&atom->state, crtc);
+	new_crtc_state = drm_atomic_get_new_crtc_state(&atom->state, crtc);
+	if (old_crtc_state->active && drm_atomic_crtc_needs_modeset(new_crtc_state)) {
+		outp = nv50_disp_outp_atomic_add(atom, encoder);
+		if (IS_ERR(outp))
+			return PTR_ERR(outp);
+
+		if (outp->encoder->encoder_type == DRM_MODE_ENCODER_DPMST) {
+			outp->flush_disable = true;
+			atom->flush_disable = true;
+		}
+		outp->clr.ctrl = true;
+		atom->lock_core = true;
+	}
+
+	return 0;
+}
+
+static int
+nv50_disp_outp_atomic_check_set(struct nv50_atom *atom,
+				struct drm_connector_state *connector_state)
+{
+	struct drm_encoder *encoder = connector_state->best_encoder;
+	struct drm_crtc_state *new_crtc_state;
+	struct drm_crtc *crtc;
+	struct nv50_outp_atom *outp;
+
+	if (!(crtc = connector_state->crtc))
+		return 0;
+
+	new_crtc_state = drm_atomic_get_new_crtc_state(&atom->state, crtc);
+	if (new_crtc_state->active && drm_atomic_crtc_needs_modeset(new_crtc_state)) {
+		outp = nv50_disp_outp_atomic_add(atom, encoder);
+		if (IS_ERR(outp))
+			return PTR_ERR(outp);
+
+		outp->set.ctrl = true;
+		atom->lock_core = true;
+	}
+
+	return 0;
+}
+
+static int
+nv50_disp_atomic_check(struct drm_device *dev, struct drm_atomic_state *state)
+{
+	struct nv50_atom *atom = nv50_atom(state);
+	struct drm_connector_state *old_connector_state, *new_connector_state;
+	struct drm_connector *connector;
+	int ret, i;
+
+	ret = drm_atomic_helper_check(dev, state);
+	if (ret)
+		return ret;
+
+	for_each_oldnew_connector_in_state(state, connector, old_connector_state, new_connector_state, i) {
+		ret = nv50_disp_outp_atomic_check_clr(atom, old_connector_state);
+		if (ret)
+			return ret;
+
+		ret = nv50_disp_outp_atomic_check_set(atom, new_connector_state);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
+static void
+nv50_disp_atomic_state_clear(struct drm_atomic_state *state)
+{
+	struct nv50_atom *atom = nv50_atom(state);
+	struct nv50_outp_atom *outp, *outt;
+
+	list_for_each_entry_safe(outp, outt, &atom->outp, head) {
+		list_del(&outp->head);
+		kfree(outp);
+	}
+
+	drm_atomic_state_default_clear(state);
+}
+
+static void
+nv50_disp_atomic_state_free(struct drm_atomic_state *state)
+{
+	struct nv50_atom *atom = nv50_atom(state);
+	drm_atomic_state_default_release(&atom->state);
+	kfree(atom);
+}
+
+static struct drm_atomic_state *
+nv50_disp_atomic_state_alloc(struct drm_device *dev)
+{
+	struct nv50_atom *atom;
+	if (!(atom = kzalloc(sizeof(*atom), GFP_KERNEL)) ||
+	    drm_atomic_state_init(dev, &atom->state) < 0) {
+		kfree(atom);
+		return NULL;
+	}
+	INIT_LIST_HEAD(&atom->outp);
+	return &atom->state;
+}
+
+static const struct drm_mode_config_funcs
+nv50_disp_func = {
+	.fb_create = nouveau_user_framebuffer_create,
+	.output_poll_changed = drm_fb_helper_output_poll_changed,
+	.atomic_check = nv50_disp_atomic_check,
+	.atomic_commit = nv50_disp_atomic_commit,
+	.atomic_state_alloc = nv50_disp_atomic_state_alloc,
+	.atomic_state_clear = nv50_disp_atomic_state_clear,
+	.atomic_state_free = nv50_disp_atomic_state_free,
+};
+
+/******************************************************************************
+ * Init
+ *****************************************************************************/
+
+void
+nv50_display_fini(struct drm_device *dev)
+{
+	struct nouveau_encoder *nv_encoder;
+	struct drm_encoder *encoder;
+	struct drm_plane *plane;
+
+	drm_for_each_plane(plane, dev) {
+		struct nv50_wndw *wndw = nv50_wndw(plane);
+		if (plane->funcs != &nv50_wndw)
+			continue;
+		nv50_wndw_fini(wndw);
+	}
+
+	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
+		if (encoder->encoder_type != DRM_MODE_ENCODER_DPMST) {
+			nv_encoder = nouveau_encoder(encoder);
+			nv50_mstm_fini(nv_encoder->dp.mstm);
+		}
+	}
+}
+
+int
+nv50_display_init(struct drm_device *dev)
+{
+	struct drm_encoder *encoder;
+	struct drm_plane *plane;
+	u32 *push;
+
+	push = evo_wait(nv50_mast(dev), 32);
+	if (!push)
+		return -EBUSY;
+
+	evo_mthd(push, 0x0088, 1);
+	evo_data(push, nv50_mast(dev)->base.sync.handle);
+	evo_kick(push, nv50_mast(dev));
+
+	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
+		if (encoder->encoder_type != DRM_MODE_ENCODER_DPMST) {
+			struct nouveau_encoder *nv_encoder =
+				nouveau_encoder(encoder);
+			nv50_mstm_init(nv_encoder->dp.mstm);
+		}
+	}
+
+	drm_for_each_plane(plane, dev) {
+		struct nv50_wndw *wndw = nv50_wndw(plane);
+		if (plane->funcs != &nv50_wndw)
+			continue;
+		nv50_wndw_init(wndw);
+	}
+
+	return 0;
+}
+
+void
+nv50_display_destroy(struct drm_device *dev)
+{
+	struct nv50_disp *disp = nv50_disp(dev);
+
+	nv50_dmac_destroy(&disp->mast.base);
+
+	nouveau_bo_unmap(disp->sync);
+	if (disp->sync)
+		nouveau_bo_unpin(disp->sync);
+	nouveau_bo_ref(NULL, &disp->sync);
+
+	nouveau_display(dev)->priv = NULL;
+	kfree(disp);
+}
+
+MODULE_PARM_DESC(atomic, "Expose atomic ioctl (default: disabled)");
+static int nouveau_atomic = 0;
+module_param_named(atomic, nouveau_atomic, int, 0400);
+
+int
+nv50_display_create(struct drm_device *dev)
+{
+	struct nvif_device *device = &nouveau_drm(dev)->client.device;
+	struct nouveau_drm *drm = nouveau_drm(dev);
+	struct dcb_table *dcb = &drm->vbios.dcb;
+	struct drm_connector *connector, *tmp;
+	struct nv50_disp *disp;
+	struct dcb_output *dcbe;
+	int crtcs, ret, i;
+
+	disp = kzalloc(sizeof(*disp), GFP_KERNEL);
+	if (!disp)
+		return -ENOMEM;
+
+	mutex_init(&disp->mutex);
+
+	nouveau_display(dev)->priv = disp;
+	nouveau_display(dev)->dtor = nv50_display_destroy;
+	nouveau_display(dev)->init = nv50_display_init;
+	nouveau_display(dev)->fini = nv50_display_fini;
+	disp->disp = &nouveau_display(dev)->disp;
+	dev->mode_config.funcs = &nv50_disp_func;
+	dev->driver->driver_features |= DRIVER_PREFER_XBGR_30BPP;
+	if (nouveau_atomic)
+		dev->driver->driver_features |= DRIVER_ATOMIC;
+
+	/* small shared memory area we use for notifiers and semaphores */
+	ret = nouveau_bo_new(&drm->client, 4096, 0x1000, TTM_PL_FLAG_VRAM,
+			     0, 0x0000, NULL, NULL, &disp->sync);
+	if (!ret) {
+		ret = nouveau_bo_pin(disp->sync, TTM_PL_FLAG_VRAM, true);
+		if (!ret) {
+			ret = nouveau_bo_map(disp->sync);
+			if (ret)
+				nouveau_bo_unpin(disp->sync);
+		}
+		if (ret)
+			nouveau_bo_ref(NULL, &disp->sync);
+	}
+
+	if (ret)
+		goto out;
+
+	/* allocate master evo channel */
+	ret = nv50_core_create(device, &disp->disp->object,
+			       disp->sync->bo.offset, &disp->mast);
+	if (ret)
+		goto out;
+
+	/* create crtc objects to represent the hw heads */
+	if (disp->disp->object.oclass >= GF110_DISP)
+		crtcs = nvif_rd32(&device->object, 0x612004) & 0xf;
+	else
+		crtcs = 0x3;
+
+	for (i = 0; i < fls(crtcs); i++) {
+		if (!(crtcs & (1 << i)))
+			continue;
+		ret = nv50_head_create(dev, i);
+		if (ret)
+			goto out;
+	}
+
+	/* create encoder/connector objects based on VBIOS DCB table */
+	for (i = 0, dcbe = &dcb->entry[0]; i < dcb->entries; i++, dcbe++) {
+		connector = nouveau_connector_create(dev, dcbe->connector);
+		if (IS_ERR(connector))
+			continue;
+
+		if (dcbe->location == DCB_LOC_ON_CHIP) {
+			switch (dcbe->type) {
+			case DCB_OUTPUT_TMDS:
+			case DCB_OUTPUT_LVDS:
+			case DCB_OUTPUT_DP:
+				ret = nv50_sor_create(connector, dcbe);
+				break;
+			case DCB_OUTPUT_ANALOG:
+				ret = nv50_dac_create(connector, dcbe);
+				break;
+			default:
+				ret = -ENODEV;
+				break;
+			}
+		} else {
+			ret = nv50_pior_create(connector, dcbe);
+		}
+
+		if (ret) {
+			NV_WARN(drm, "failed to create encoder %d/%d/%d: %d\n",
+				     dcbe->location, dcbe->type,
+				     ffs(dcbe->or) - 1, ret);
+			ret = 0;
+		}
+	}
+
+	/* cull any connectors we created that don't have an encoder */
+	list_for_each_entry_safe(connector, tmp, &dev->mode_config.connector_list, head) {
+		if (connector->encoder_ids[0])
+			continue;
+
+		NV_WARN(drm, "%s has no encoders, removing\n",
+			connector->name);
+		connector->funcs->destroy(connector);
+	}
+
+out:
+	if (ret)
+		nv50_display_destroy(dev);
+	return ret;
+}
diff --git a/drivers/gpu/drm/nouveau/nv50_display.c b/drivers/gpu/drm/nouveau/nv50_display.c
deleted file mode 100644
index f2156c8..0000000
--- a/drivers/gpu/drm/nouveau/nv50_display.c
+++ /dev/null
@@ -1,4542 +0,0 @@
-/*
- * Copyright 2011 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs
- */
-
-#include <linux/dma-mapping.h>
-#include <linux/hdmi.h>
-
-#include <drm/drmP.h>
-#include <drm/drm_atomic.h>
-#include <drm/drm_atomic_helper.h>
-#include <drm/drm_crtc_helper.h>
-#include <drm/drm_dp_helper.h>
-#include <drm/drm_fb_helper.h>
-#include <drm/drm_plane_helper.h>
-#include <drm/drm_edid.h>
-
-#include <nvif/mem.h>
-
-#include <nvif/class.h>
-#include <nvif/cl0002.h>
-#include <nvif/cl5070.h>
-#include <nvif/cl507a.h>
-#include <nvif/cl507b.h>
-#include <nvif/cl507c.h>
-#include <nvif/cl507d.h>
-#include <nvif/cl507e.h>
-#include <nvif/event.h>
-
-#include "nouveau_drv.h"
-#include "nouveau_dma.h"
-#include "nouveau_gem.h"
-#include "nouveau_connector.h"
-#include "nouveau_encoder.h"
-#include "nouveau_crtc.h"
-#include "nouveau_fence.h"
-#include "nouveau_fbcon.h"
-#include "nv50_display.h"
-
-#define EVO_DMA_NR 9
-
-#define EVO_MASTER  (0x00)
-#define EVO_FLIP(c) (0x01 + (c))
-#define EVO_OVLY(c) (0x05 + (c))
-#define EVO_OIMM(c) (0x09 + (c))
-#define EVO_CURS(c) (0x0d + (c))
-
-/* offsets in shared sync bo of various structures */
-#define EVO_SYNC(c, o) ((c) * 0x0100 + (o))
-#define EVO_MAST_NTFY     EVO_SYNC(      0, 0x00)
-#define EVO_FLIP_SEM0(c)  EVO_SYNC((c) + 1, 0x00)
-#define EVO_FLIP_SEM1(c)  EVO_SYNC((c) + 1, 0x10)
-#define EVO_FLIP_NTFY0(c) EVO_SYNC((c) + 1, 0x20)
-#define EVO_FLIP_NTFY1(c) EVO_SYNC((c) + 1, 0x30)
-
-/******************************************************************************
- * Atomic state
- *****************************************************************************/
-#define nv50_atom(p) container_of((p), struct nv50_atom, state)
-
-struct nv50_atom {
-	struct drm_atomic_state state;
-
-	struct list_head outp;
-	bool lock_core;
-	bool flush_disable;
-};
-
-struct nv50_outp_atom {
-	struct list_head head;
-
-	struct drm_encoder *encoder;
-	bool flush_disable;
-
-	union {
-		struct {
-			bool ctrl:1;
-		};
-		u8 mask;
-	} clr;
-
-	union {
-		struct {
-			bool ctrl:1;
-		};
-		u8 mask;
-	} set;
-};
-
-#define nv50_head_atom(p) container_of((p), struct nv50_head_atom, state)
-
-struct nv50_head_atom {
-	struct drm_crtc_state state;
-
-	struct {
-		u16 iW;
-		u16 iH;
-		u16 oW;
-		u16 oH;
-	} view;
-
-	struct nv50_head_mode {
-		bool interlace;
-		u32 clock;
-		struct {
-			u16 active;
-			u16 synce;
-			u16 blanke;
-			u16 blanks;
-		} h;
-		struct {
-			u32 active;
-			u16 synce;
-			u16 blanke;
-			u16 blanks;
-			u16 blank2s;
-			u16 blank2e;
-			u16 blankus;
-		} v;
-	} mode;
-
-	struct {
-		bool visible;
-		u32 handle;
-		u64 offset:40;
-		u8  mode:4;
-	} lut;
-
-	struct {
-		bool visible;
-		u32 handle;
-		u64 offset:40;
-		u8  format;
-		u8  kind:7;
-		u8  layout:1;
-		u8  block:4;
-		u32 pitch:20;
-		u16 x;
-		u16 y;
-		u16 w;
-		u16 h;
-	} core;
-
-	struct {
-		bool visible;
-		u32 handle;
-		u64 offset:40;
-		u8  layout:1;
-		u8  format:1;
-	} curs;
-
-	struct {
-		u8  depth;
-		u8  cpp;
-		u16 x;
-		u16 y;
-		u16 w;
-		u16 h;
-	} base;
-
-	struct {
-		u8 cpp;
-	} ovly;
-
-	struct {
-		bool enable:1;
-		u8 bits:2;
-		u8 mode:4;
-	} dither;
-
-	struct {
-		struct {
-			u16 cos:12;
-			u16 sin:12;
-		} sat;
-	} procamp;
-
-	union {
-		struct {
-			bool ilut:1;
-			bool core:1;
-			bool curs:1;
-		};
-		u8 mask;
-	} clr;
-
-	union {
-		struct {
-			bool ilut:1;
-			bool core:1;
-			bool curs:1;
-			bool view:1;
-			bool mode:1;
-			bool base:1;
-			bool ovly:1;
-			bool dither:1;
-			bool procamp:1;
-		};
-		u16 mask;
-	} set;
-};
-
-static inline struct nv50_head_atom *
-nv50_head_atom_get(struct drm_atomic_state *state, struct drm_crtc *crtc)
-{
-	struct drm_crtc_state *statec = drm_atomic_get_crtc_state(state, crtc);
-	if (IS_ERR(statec))
-		return (void *)statec;
-	return nv50_head_atom(statec);
-}
-
-#define nv50_wndw_atom(p) container_of((p), struct nv50_wndw_atom, state)
-
-struct nv50_wndw_atom {
-	struct drm_plane_state state;
-	u8 interval;
-
-	struct {
-		u32  handle;
-		u16  offset:12;
-		bool awaken:1;
-	} ntfy;
-
-	struct {
-		u32 handle;
-		u16 offset:12;
-		u32 acquire;
-		u32 release;
-	} sema;
-
-	struct {
-		u8 enable:2;
-	} lut;
-
-	struct {
-		u8  mode:2;
-		u8  interval:4;
-
-		u8  format;
-		u8  kind:7;
-		u8  layout:1;
-		u8  block:4;
-		u32 pitch:20;
-		u16 w;
-		u16 h;
-
-		u32 handle;
-		u64 offset;
-	} image;
-
-	struct {
-		u16 x;
-		u16 y;
-	} point;
-
-	union {
-		struct {
-			bool ntfy:1;
-			bool sema:1;
-			bool image:1;
-		};
-		u8 mask;
-	} clr;
-
-	union {
-		struct {
-			bool ntfy:1;
-			bool sema:1;
-			bool image:1;
-			bool lut:1;
-			bool point:1;
-		};
-		u8 mask;
-	} set;
-};
-
-/******************************************************************************
- * EVO channel
- *****************************************************************************/
-
-struct nv50_chan {
-	struct nvif_object user;
-	struct nvif_device *device;
-};
-
-static int
-nv50_chan_create(struct nvif_device *device, struct nvif_object *disp,
-		 const s32 *oclass, u8 head, void *data, u32 size,
-		 struct nv50_chan *chan)
-{
-	struct nvif_sclass *sclass;
-	int ret, i, n;
-
-	chan->device = device;
-
-	ret = n = nvif_object_sclass_get(disp, &sclass);
-	if (ret < 0)
-		return ret;
-
-	while (oclass[0]) {
-		for (i = 0; i < n; i++) {
-			if (sclass[i].oclass == oclass[0]) {
-				ret = nvif_object_init(disp, 0, oclass[0],
-						       data, size, &chan->user);
-				if (ret == 0)
-					nvif_object_map(&chan->user, NULL, 0);
-				nvif_object_sclass_put(&sclass);
-				return ret;
-			}
-		}
-		oclass++;
-	}
-
-	nvif_object_sclass_put(&sclass);
-	return -ENOSYS;
-}
-
-static void
-nv50_chan_destroy(struct nv50_chan *chan)
-{
-	nvif_object_fini(&chan->user);
-}
-
-/******************************************************************************
- * PIO EVO channel
- *****************************************************************************/
-
-struct nv50_pioc {
-	struct nv50_chan base;
-};
-
-static void
-nv50_pioc_destroy(struct nv50_pioc *pioc)
-{
-	nv50_chan_destroy(&pioc->base);
-}
-
-static int
-nv50_pioc_create(struct nvif_device *device, struct nvif_object *disp,
-		 const s32 *oclass, u8 head, void *data, u32 size,
-		 struct nv50_pioc *pioc)
-{
-	return nv50_chan_create(device, disp, oclass, head, data, size,
-				&pioc->base);
-}
-
-/******************************************************************************
- * Overlay Immediate
- *****************************************************************************/
-
-struct nv50_oimm {
-	struct nv50_pioc base;
-};
-
-static int
-nv50_oimm_create(struct nvif_device *device, struct nvif_object *disp,
-		 int head, struct nv50_oimm *oimm)
-{
-	struct nv50_disp_cursor_v0 args = {
-		.head = head,
-	};
-	static const s32 oclass[] = {
-		GK104_DISP_OVERLAY,
-		GF110_DISP_OVERLAY,
-		GT214_DISP_OVERLAY,
-		G82_DISP_OVERLAY,
-		NV50_DISP_OVERLAY,
-		0
-	};
-
-	return nv50_pioc_create(device, disp, oclass, head, &args, sizeof(args),
-				&oimm->base);
-}
-
-/******************************************************************************
- * DMA EVO channel
- *****************************************************************************/
-
-struct nv50_dmac_ctxdma {
-	struct list_head head;
-	struct nvif_object object;
-};
-
-struct nv50_dmac {
-	struct nv50_chan base;
-
-	struct nvif_mem push;
-	u32 *ptr;
-
-	struct nvif_object sync;
-	struct nvif_object vram;
-	struct list_head ctxdma;
-
-	/* Protects against concurrent pushbuf access to this channel, lock is
-	 * grabbed by evo_wait (if the pushbuf reservation is successful) and
-	 * dropped again by evo_kick. */
-	struct mutex lock;
-};
-
-static void
-nv50_dmac_ctxdma_del(struct nv50_dmac_ctxdma *ctxdma)
-{
-	nvif_object_fini(&ctxdma->object);
-	list_del(&ctxdma->head);
-	kfree(ctxdma);
-}
-
-static struct nv50_dmac_ctxdma *
-nv50_dmac_ctxdma_new(struct nv50_dmac *dmac, struct nouveau_framebuffer *fb)
-{
-	struct nouveau_drm *drm = nouveau_drm(fb->base.dev);
-	struct nv50_dmac_ctxdma *ctxdma;
-	const u8    kind = fb->nvbo->kind;
-	const u32 handle = 0xfb000000 | kind;
-	struct {
-		struct nv_dma_v0 base;
-		union {
-			struct nv50_dma_v0 nv50;
-			struct gf100_dma_v0 gf100;
-			struct gf119_dma_v0 gf119;
-		};
-	} args = {};
-	u32 argc = sizeof(args.base);
-	int ret;
-
-	list_for_each_entry(ctxdma, &dmac->ctxdma, head) {
-		if (ctxdma->object.handle == handle)
-			return ctxdma;
-	}
-
-	if (!(ctxdma = kzalloc(sizeof(*ctxdma), GFP_KERNEL)))
-		return ERR_PTR(-ENOMEM);
-	list_add(&ctxdma->head, &dmac->ctxdma);
-
-	args.base.target = NV_DMA_V0_TARGET_VRAM;
-	args.base.access = NV_DMA_V0_ACCESS_RDWR;
-	args.base.start  = 0;
-	args.base.limit  = drm->client.device.info.ram_user - 1;
-
-	if (drm->client.device.info.chipset < 0x80) {
-		args.nv50.part = NV50_DMA_V0_PART_256;
-		argc += sizeof(args.nv50);
-	} else
-	if (drm->client.device.info.chipset < 0xc0) {
-		args.nv50.part = NV50_DMA_V0_PART_256;
-		args.nv50.kind = kind;
-		argc += sizeof(args.nv50);
-	} else
-	if (drm->client.device.info.chipset < 0xd0) {
-		args.gf100.kind = kind;
-		argc += sizeof(args.gf100);
-	} else {
-		args.gf119.page = GF119_DMA_V0_PAGE_LP;
-		args.gf119.kind = kind;
-		argc += sizeof(args.gf119);
-	}
-
-	ret = nvif_object_init(&dmac->base.user, handle, NV_DMA_IN_MEMORY,
-			       &args, argc, &ctxdma->object);
-	if (ret) {
-		nv50_dmac_ctxdma_del(ctxdma);
-		return ERR_PTR(ret);
-	}
-
-	return ctxdma;
-}
-
-static void
-nv50_dmac_destroy(struct nv50_dmac *dmac)
-{
-	struct nv50_dmac_ctxdma *ctxdma, *ctxtmp;
-
-	list_for_each_entry_safe(ctxdma, ctxtmp, &dmac->ctxdma, head) {
-		nv50_dmac_ctxdma_del(ctxdma);
-	}
-
-	nvif_object_fini(&dmac->vram);
-	nvif_object_fini(&dmac->sync);
-
-	nv50_chan_destroy(&dmac->base);
-
-	nvif_mem_fini(&dmac->push);
-}
-
-static int
-nv50_dmac_create(struct nvif_device *device, struct nvif_object *disp,
-		 const s32 *oclass, u8 head, void *data, u32 size, u64 syncbuf,
-		 struct nv50_dmac *dmac)
-{
-	struct nouveau_cli *cli = (void *)device->object.client;
-	struct nv50_disp_core_channel_dma_v0 *args = data;
-	int ret;
-
-	mutex_init(&dmac->lock);
-	INIT_LIST_HEAD(&dmac->ctxdma);
-
-	ret = nvif_mem_init_map(&cli->mmu, NVIF_MEM_COHERENT, 0x1000,
-				&dmac->push);
-	if (ret)
-		return ret;
-
-	dmac->ptr = dmac->push.object.map.ptr;
-
-	args->pushbuf = nvif_handle(&dmac->push.object);
-
-	ret = nv50_chan_create(device, disp, oclass, head, data, size,
-			       &dmac->base);
-	if (ret)
-		return ret;
-
-	ret = nvif_object_init(&dmac->base.user, 0xf0000000, NV_DMA_IN_MEMORY,
-			       &(struct nv_dma_v0) {
-					.target = NV_DMA_V0_TARGET_VRAM,
-					.access = NV_DMA_V0_ACCESS_RDWR,
-					.start = syncbuf + 0x0000,
-					.limit = syncbuf + 0x0fff,
-			       }, sizeof(struct nv_dma_v0),
-			       &dmac->sync);
-	if (ret)
-		return ret;
-
-	ret = nvif_object_init(&dmac->base.user, 0xf0000001, NV_DMA_IN_MEMORY,
-			       &(struct nv_dma_v0) {
-					.target = NV_DMA_V0_TARGET_VRAM,
-					.access = NV_DMA_V0_ACCESS_RDWR,
-					.start = 0,
-					.limit = device->info.ram_user - 1,
-			       }, sizeof(struct nv_dma_v0),
-			       &dmac->vram);
-	if (ret)
-		return ret;
-
-	return ret;
-}
-
-/******************************************************************************
- * Core
- *****************************************************************************/
-
-struct nv50_mast {
-	struct nv50_dmac base;
-};
-
-static int
-nv50_core_create(struct nvif_device *device, struct nvif_object *disp,
-		 u64 syncbuf, struct nv50_mast *core)
-{
-	struct nv50_disp_core_channel_dma_v0 args = {};
-	static const s32 oclass[] = {
-		GP102_DISP_CORE_CHANNEL_DMA,
-		GP100_DISP_CORE_CHANNEL_DMA,
-		GM200_DISP_CORE_CHANNEL_DMA,
-		GM107_DISP_CORE_CHANNEL_DMA,
-		GK110_DISP_CORE_CHANNEL_DMA,
-		GK104_DISP_CORE_CHANNEL_DMA,
-		GF110_DISP_CORE_CHANNEL_DMA,
-		GT214_DISP_CORE_CHANNEL_DMA,
-		GT206_DISP_CORE_CHANNEL_DMA,
-		GT200_DISP_CORE_CHANNEL_DMA,
-		G82_DISP_CORE_CHANNEL_DMA,
-		NV50_DISP_CORE_CHANNEL_DMA,
-		0
-	};
-
-	return nv50_dmac_create(device, disp, oclass, 0, &args, sizeof(args),
-				syncbuf, &core->base);
-}
-
-/******************************************************************************
- * Base
- *****************************************************************************/
-
-struct nv50_sync {
-	struct nv50_dmac base;
-	u32 addr;
-	u32 data;
-};
-
-static int
-nv50_base_create(struct nvif_device *device, struct nvif_object *disp,
-		 int head, u64 syncbuf, struct nv50_sync *base)
-{
-	struct nv50_disp_base_channel_dma_v0 args = {
-		.head = head,
-	};
-	static const s32 oclass[] = {
-		GK110_DISP_BASE_CHANNEL_DMA,
-		GK104_DISP_BASE_CHANNEL_DMA,
-		GF110_DISP_BASE_CHANNEL_DMA,
-		GT214_DISP_BASE_CHANNEL_DMA,
-		GT200_DISP_BASE_CHANNEL_DMA,
-		G82_DISP_BASE_CHANNEL_DMA,
-		NV50_DISP_BASE_CHANNEL_DMA,
-		0
-	};
-
-	return nv50_dmac_create(device, disp, oclass, head, &args, sizeof(args),
-				syncbuf, &base->base);
-}
-
-/******************************************************************************
- * Overlay
- *****************************************************************************/
-
-struct nv50_ovly {
-	struct nv50_dmac base;
-};
-
-static int
-nv50_ovly_create(struct nvif_device *device, struct nvif_object *disp,
-		 int head, u64 syncbuf, struct nv50_ovly *ovly)
-{
-	struct nv50_disp_overlay_channel_dma_v0 args = {
-		.head = head,
-	};
-	static const s32 oclass[] = {
-		GK104_DISP_OVERLAY_CONTROL_DMA,
-		GF110_DISP_OVERLAY_CONTROL_DMA,
-		GT214_DISP_OVERLAY_CHANNEL_DMA,
-		GT200_DISP_OVERLAY_CHANNEL_DMA,
-		G82_DISP_OVERLAY_CHANNEL_DMA,
-		NV50_DISP_OVERLAY_CHANNEL_DMA,
-		0
-	};
-
-	return nv50_dmac_create(device, disp, oclass, head, &args, sizeof(args),
-				syncbuf, &ovly->base);
-}
-
-struct nv50_head {
-	struct nouveau_crtc base;
-	struct {
-		struct nouveau_bo *nvbo[2];
-		int next;
-	} lut;
-	struct nv50_ovly ovly;
-	struct nv50_oimm oimm;
-};
-
-#define nv50_head(c) ((struct nv50_head *)nouveau_crtc(c))
-#define nv50_ovly(c) (&nv50_head(c)->ovly)
-#define nv50_oimm(c) (&nv50_head(c)->oimm)
-#define nv50_chan(c) (&(c)->base.base)
-#define nv50_vers(c) nv50_chan(c)->user.oclass
-
-struct nv50_disp {
-	struct nvif_disp *disp;
-	struct nv50_mast mast;
-
-	struct nouveau_bo *sync;
-
-	struct mutex mutex;
-};
-
-static struct nv50_disp *
-nv50_disp(struct drm_device *dev)
-{
-	return nouveau_display(dev)->priv;
-}
-
-#define nv50_mast(d) (&nv50_disp(d)->mast)
-
-/******************************************************************************
- * EVO channel helpers
- *****************************************************************************/
-static u32 *
-evo_wait(void *evoc, int nr)
-{
-	struct nv50_dmac *dmac = evoc;
-	struct nvif_device *device = dmac->base.device;
-	u32 put = nvif_rd32(&dmac->base.user, 0x0000) / 4;
-
-	mutex_lock(&dmac->lock);
-	if (put + nr >= (PAGE_SIZE / 4) - 8) {
-		dmac->ptr[put] = 0x20000000;
-
-		nvif_wr32(&dmac->base.user, 0x0000, 0x00000000);
-		if (nvif_msec(device, 2000,
-			if (!nvif_rd32(&dmac->base.user, 0x0004))
-				break;
-		) < 0) {
-			mutex_unlock(&dmac->lock);
-			pr_err("nouveau: evo channel stalled\n");
-			return NULL;
-		}
-
-		put = 0;
-	}
-
-	return dmac->ptr + put;
-}
-
-static void
-evo_kick(u32 *push, void *evoc)
-{
-	struct nv50_dmac *dmac = evoc;
-	nvif_wr32(&dmac->base.user, 0x0000, (push - dmac->ptr) << 2);
-	mutex_unlock(&dmac->lock);
-}
-
-#define evo_mthd(p, m, s) do {						\
-	const u32 _m = (m), _s = (s);					\
-	if (drm_debug & DRM_UT_KMS)					\
-		pr_err("%04x %d %s\n", _m, _s, __func__);		\
-	*((p)++) = ((_s << 18) | _m);					\
-} while(0)
-
-#define evo_data(p, d) do {						\
-	const u32 _d = (d);						\
-	if (drm_debug & DRM_UT_KMS)					\
-		pr_err("\t%08x\n", _d);					\
-	*((p)++) = _d;							\
-} while(0)
-
-/******************************************************************************
- * Plane
- *****************************************************************************/
-#define nv50_wndw(p) container_of((p), struct nv50_wndw, plane)
-
-struct nv50_wndw {
-	const struct nv50_wndw_func *func;
-	struct nv50_dmac *dmac;
-
-	struct drm_plane plane;
-
-	struct nvif_notify notify;
-	u16 ntfy;
-	u16 sema;
-	u32 data;
-};
-
-struct nv50_wndw_func {
-	void *(*dtor)(struct nv50_wndw *);
-	int (*acquire)(struct nv50_wndw *, struct nv50_wndw_atom *asyw,
-		       struct nv50_head_atom *asyh);
-	void (*release)(struct nv50_wndw *, struct nv50_wndw_atom *asyw,
-			struct nv50_head_atom *asyh);
-	void (*prepare)(struct nv50_wndw *, struct nv50_head_atom *asyh,
-			struct nv50_wndw_atom *asyw);
-
-	void (*sema_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
-	void (*sema_clr)(struct nv50_wndw *);
-	void (*ntfy_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
-	void (*ntfy_clr)(struct nv50_wndw *);
-	int (*ntfy_wait_begun)(struct nv50_wndw *, struct nv50_wndw_atom *);
-	void (*image_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
-	void (*image_clr)(struct nv50_wndw *);
-	void (*lut)(struct nv50_wndw *, struct nv50_wndw_atom *);
-	void (*point)(struct nv50_wndw *, struct nv50_wndw_atom *);
-
-	u32 (*update)(struct nv50_wndw *, u32 interlock);
-};
-
-static int
-nv50_wndw_wait_armed(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
-{
-	if (asyw->set.ntfy)
-		return wndw->func->ntfy_wait_begun(wndw, asyw);
-	return 0;
-}
-
-static u32
-nv50_wndw_flush_clr(struct nv50_wndw *wndw, u32 interlock, bool flush,
-		    struct nv50_wndw_atom *asyw)
-{
-	if (asyw->clr.sema && (!asyw->set.sema || flush))
-		wndw->func->sema_clr(wndw);
-	if (asyw->clr.ntfy && (!asyw->set.ntfy || flush))
-		wndw->func->ntfy_clr(wndw);
-	if (asyw->clr.image && (!asyw->set.image || flush))
-		wndw->func->image_clr(wndw);
-
-	return flush ? wndw->func->update(wndw, interlock) : 0;
-}
-
-static u32
-nv50_wndw_flush_set(struct nv50_wndw *wndw, u32 interlock,
-		    struct nv50_wndw_atom *asyw)
-{
-	if (interlock) {
-		asyw->image.mode = 0;
-		asyw->image.interval = 1;
-	}
-
-	if (asyw->set.sema ) wndw->func->sema_set (wndw, asyw);
-	if (asyw->set.ntfy ) wndw->func->ntfy_set (wndw, asyw);
-	if (asyw->set.image) wndw->func->image_set(wndw, asyw);
-	if (asyw->set.lut  ) wndw->func->lut      (wndw, asyw);
-	if (asyw->set.point) wndw->func->point    (wndw, asyw);
-
-	return wndw->func->update(wndw, interlock);
-}
-
-static void
-nv50_wndw_atomic_check_release(struct nv50_wndw *wndw,
-			       struct nv50_wndw_atom *asyw,
-			       struct nv50_head_atom *asyh)
-{
-	struct nouveau_drm *drm = nouveau_drm(wndw->plane.dev);
-	NV_ATOMIC(drm, "%s release\n", wndw->plane.name);
-	wndw->func->release(wndw, asyw, asyh);
-	asyw->ntfy.handle = 0;
-	asyw->sema.handle = 0;
-}
-
-static int
-nv50_wndw_atomic_check_acquire(struct nv50_wndw *wndw,
-			       struct nv50_wndw_atom *asyw,
-			       struct nv50_head_atom *asyh)
-{
-	struct nouveau_framebuffer *fb = nouveau_framebuffer(asyw->state.fb);
-	struct nouveau_drm *drm = nouveau_drm(wndw->plane.dev);
-	int ret;
-
-	NV_ATOMIC(drm, "%s acquire\n", wndw->plane.name);
-
-	asyw->image.w = fb->base.width;
-	asyw->image.h = fb->base.height;
-	asyw->image.kind = fb->nvbo->kind;
-
-	if (asyh->state.pageflip_flags & DRM_MODE_PAGE_FLIP_ASYNC)
-		asyw->interval = 0;
-	else
-		asyw->interval = 1;
-
-	if (asyw->image.kind) {
-		asyw->image.layout = 0;
-		if (drm->client.device.info.chipset >= 0xc0)
-			asyw->image.block = fb->nvbo->mode >> 4;
-		else
-			asyw->image.block = fb->nvbo->mode;
-		asyw->image.pitch = (fb->base.pitches[0] / 4) << 4;
-	} else {
-		asyw->image.layout = 1;
-		asyw->image.block  = 0;
-		asyw->image.pitch  = fb->base.pitches[0];
-	}
-
-	ret = wndw->func->acquire(wndw, asyw, asyh);
-	if (ret)
-		return ret;
-
-	if (asyw->set.image) {
-		if (!(asyw->image.mode = asyw->interval ? 0 : 1))
-			asyw->image.interval = asyw->interval;
-		else
-			asyw->image.interval = 0;
-	}
-
-	return 0;
-}
-
-static int
-nv50_wndw_atomic_check(struct drm_plane *plane, struct drm_plane_state *state)
-{
-	struct nouveau_drm *drm = nouveau_drm(plane->dev);
-	struct nv50_wndw *wndw = nv50_wndw(plane);
-	struct nv50_wndw_atom *armw = nv50_wndw_atom(wndw->plane.state);
-	struct nv50_wndw_atom *asyw = nv50_wndw_atom(state);
-	struct nv50_head_atom *harm = NULL, *asyh = NULL;
-	bool varm = false, asyv = false, asym = false;
-	int ret;
-
-	NV_ATOMIC(drm, "%s atomic_check\n", plane->name);
-	if (asyw->state.crtc) {
-		asyh = nv50_head_atom_get(asyw->state.state, asyw->state.crtc);
-		if (IS_ERR(asyh))
-			return PTR_ERR(asyh);
-		asym = drm_atomic_crtc_needs_modeset(&asyh->state);
-		asyv = asyh->state.active;
-	}
-
-	if (armw->state.crtc) {
-		harm = nv50_head_atom_get(asyw->state.state, armw->state.crtc);
-		if (IS_ERR(harm))
-			return PTR_ERR(harm);
-		varm = harm->state.crtc->state->active;
-	}
-
-	if (asyv) {
-		asyw->point.x = asyw->state.crtc_x;
-		asyw->point.y = asyw->state.crtc_y;
-		if (memcmp(&armw->point, &asyw->point, sizeof(asyw->point)))
-			asyw->set.point = true;
-
-		ret = nv50_wndw_atomic_check_acquire(wndw, asyw, asyh);
-		if (ret)
-			return ret;
-	} else
-	if (varm) {
-		nv50_wndw_atomic_check_release(wndw, asyw, harm);
-	} else {
-		return 0;
-	}
-
-	if (!asyv || asym) {
-		asyw->clr.ntfy = armw->ntfy.handle != 0;
-		asyw->clr.sema = armw->sema.handle != 0;
-		if (wndw->func->image_clr)
-			asyw->clr.image = armw->image.handle != 0;
-		asyw->set.lut = wndw->func->lut && asyv;
-	}
-
-	return 0;
-}
-
-static void
-nv50_wndw_cleanup_fb(struct drm_plane *plane, struct drm_plane_state *old_state)
-{
-	struct nouveau_framebuffer *fb = nouveau_framebuffer(old_state->fb);
-	struct nouveau_drm *drm = nouveau_drm(plane->dev);
-
-	NV_ATOMIC(drm, "%s cleanup: %p\n", plane->name, old_state->fb);
-	if (!old_state->fb)
-		return;
-
-	nouveau_bo_unpin(fb->nvbo);
-}
-
-static int
-nv50_wndw_prepare_fb(struct drm_plane *plane, struct drm_plane_state *state)
-{
-	struct nouveau_framebuffer *fb = nouveau_framebuffer(state->fb);
-	struct nouveau_drm *drm = nouveau_drm(plane->dev);
-	struct nv50_wndw *wndw = nv50_wndw(plane);
-	struct nv50_wndw_atom *asyw = nv50_wndw_atom(state);
-	struct nv50_head_atom *asyh;
-	struct nv50_dmac_ctxdma *ctxdma;
-	int ret;
-
-	NV_ATOMIC(drm, "%s prepare: %p\n", plane->name, state->fb);
-	if (!asyw->state.fb)
-		return 0;
-
-	ret = nouveau_bo_pin(fb->nvbo, TTM_PL_FLAG_VRAM, true);
-	if (ret)
-		return ret;
-
-	ctxdma = nv50_dmac_ctxdma_new(wndw->dmac, fb);
-	if (IS_ERR(ctxdma)) {
-		nouveau_bo_unpin(fb->nvbo);
-		return PTR_ERR(ctxdma);
-	}
-
-	asyw->state.fence = reservation_object_get_excl_rcu(fb->nvbo->bo.resv);
-	asyw->image.handle = ctxdma->object.handle;
-	asyw->image.offset = fb->nvbo->bo.offset;
-
-	if (wndw->func->prepare) {
-		asyh = nv50_head_atom_get(asyw->state.state, asyw->state.crtc);
-		if (IS_ERR(asyh))
-			return PTR_ERR(asyh);
-
-		wndw->func->prepare(wndw, asyh, asyw);
-	}
-
-	return 0;
-}
-
-static const struct drm_plane_helper_funcs
-nv50_wndw_helper = {
-	.prepare_fb = nv50_wndw_prepare_fb,
-	.cleanup_fb = nv50_wndw_cleanup_fb,
-	.atomic_check = nv50_wndw_atomic_check,
-};
-
-static void
-nv50_wndw_atomic_destroy_state(struct drm_plane *plane,
-			       struct drm_plane_state *state)
-{
-	struct nv50_wndw_atom *asyw = nv50_wndw_atom(state);
-	__drm_atomic_helper_plane_destroy_state(&asyw->state);
-	kfree(asyw);
-}
-
-static struct drm_plane_state *
-nv50_wndw_atomic_duplicate_state(struct drm_plane *plane)
-{
-	struct nv50_wndw_atom *armw = nv50_wndw_atom(plane->state);
-	struct nv50_wndw_atom *asyw;
-	if (!(asyw = kmalloc(sizeof(*asyw), GFP_KERNEL)))
-		return NULL;
-	__drm_atomic_helper_plane_duplicate_state(plane, &asyw->state);
-	asyw->interval = 1;
-	asyw->sema = armw->sema;
-	asyw->ntfy = armw->ntfy;
-	asyw->image = armw->image;
-	asyw->point = armw->point;
-	asyw->lut = armw->lut;
-	asyw->clr.mask = 0;
-	asyw->set.mask = 0;
-	return &asyw->state;
-}
-
-static void
-nv50_wndw_reset(struct drm_plane *plane)
-{
-	struct nv50_wndw_atom *asyw;
-
-	if (WARN_ON(!(asyw = kzalloc(sizeof(*asyw), GFP_KERNEL))))
-		return;
-
-	if (plane->state)
-		plane->funcs->atomic_destroy_state(plane, plane->state);
-	plane->state = &asyw->state;
-	plane->state->plane = plane;
-	plane->state->rotation = DRM_MODE_ROTATE_0;
-}
-
-static void
-nv50_wndw_destroy(struct drm_plane *plane)
-{
-	struct nv50_wndw *wndw = nv50_wndw(plane);
-	void *data;
-	nvif_notify_fini(&wndw->notify);
-	data = wndw->func->dtor(wndw);
-	drm_plane_cleanup(&wndw->plane);
-	kfree(data);
-}
-
-static const struct drm_plane_funcs
-nv50_wndw = {
-	.update_plane = drm_atomic_helper_update_plane,
-	.disable_plane = drm_atomic_helper_disable_plane,
-	.destroy = nv50_wndw_destroy,
-	.reset = nv50_wndw_reset,
-	.atomic_duplicate_state = nv50_wndw_atomic_duplicate_state,
-	.atomic_destroy_state = nv50_wndw_atomic_destroy_state,
-};
-
-static void
-nv50_wndw_fini(struct nv50_wndw *wndw)
-{
-	nvif_notify_put(&wndw->notify);
-}
-
-static void
-nv50_wndw_init(struct nv50_wndw *wndw)
-{
-	nvif_notify_get(&wndw->notify);
-}
-
-static int
-nv50_wndw_ctor(const struct nv50_wndw_func *func, struct drm_device *dev,
-	       enum drm_plane_type type, const char *name, int index,
-	       struct nv50_dmac *dmac, const u32 *format, int nformat,
-	       struct nv50_wndw *wndw)
-{
-	int ret;
-
-	wndw->func = func;
-	wndw->dmac = dmac;
-
-	ret = drm_universal_plane_init(dev, &wndw->plane, 0, &nv50_wndw,
-				       format, nformat, NULL,
-				       type, "%s-%d", name, index);
-	if (ret)
-		return ret;
-
-	drm_plane_helper_add(&wndw->plane, &nv50_wndw_helper);
-	return 0;
-}
-
-/******************************************************************************
- * Cursor plane
- *****************************************************************************/
-#define nv50_curs(p) container_of((p), struct nv50_curs, wndw)
-
-struct nv50_curs {
-	struct nv50_wndw wndw;
-	struct nvif_object chan;
-};
-
-static u32
-nv50_curs_update(struct nv50_wndw *wndw, u32 interlock)
-{
-	struct nv50_curs *curs = nv50_curs(wndw);
-	nvif_wr32(&curs->chan, 0x0080, 0x00000000);
-	return 0;
-}
-
-static void
-nv50_curs_point(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
-{
-	struct nv50_curs *curs = nv50_curs(wndw);
-	nvif_wr32(&curs->chan, 0x0084, (asyw->point.y << 16) | asyw->point.x);
-}
-
-static void
-nv50_curs_prepare(struct nv50_wndw *wndw, struct nv50_head_atom *asyh,
-		  struct nv50_wndw_atom *asyw)
-{
-	u32 handle = nv50_disp(wndw->plane.dev)->mast.base.vram.handle;
-	u32 offset = asyw->image.offset;
-	if (asyh->curs.handle != handle || asyh->curs.offset != offset) {
-		asyh->curs.handle = handle;
-		asyh->curs.offset = offset;
-		asyh->set.curs = asyh->curs.visible;
-	}
-}
-
-static void
-nv50_curs_release(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
-		  struct nv50_head_atom *asyh)
-{
-	asyh->curs.visible = false;
-}
-
-static int
-nv50_curs_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
-		  struct nv50_head_atom *asyh)
-{
-	int ret;
-
-	ret = drm_atomic_helper_check_plane_state(&asyw->state, &asyh->state,
-						  DRM_PLANE_HELPER_NO_SCALING,
-						  DRM_PLANE_HELPER_NO_SCALING,
-						  true, true);
-	asyh->curs.visible = asyw->state.visible;
-	if (ret || !asyh->curs.visible)
-		return ret;
-
-	switch (asyw->state.fb->width) {
-	case 32: asyh->curs.layout = 0; break;
-	case 64: asyh->curs.layout = 1; break;
-	default:
-		return -EINVAL;
-	}
-
-	if (asyw->state.fb->width != asyw->state.fb->height)
-		return -EINVAL;
-
-	switch (asyw->state.fb->format->format) {
-	case DRM_FORMAT_ARGB8888: asyh->curs.format = 1; break;
-	default:
-		WARN_ON(1);
-		return -EINVAL;
-	}
-
-	return 0;
-}
-
-static void *
-nv50_curs_dtor(struct nv50_wndw *wndw)
-{
-	struct nv50_curs *curs = nv50_curs(wndw);
-	nvif_object_fini(&curs->chan);
-	return curs;
-}
-
-static const u32
-nv50_curs_format[] = {
-	DRM_FORMAT_ARGB8888,
-};
-
-static const struct nv50_wndw_func
-nv50_curs = {
-	.dtor = nv50_curs_dtor,
-	.acquire = nv50_curs_acquire,
-	.release = nv50_curs_release,
-	.prepare = nv50_curs_prepare,
-	.point = nv50_curs_point,
-	.update = nv50_curs_update,
-};
-
-static int
-nv50_curs_new(struct nouveau_drm *drm, struct nv50_head *head,
-	      struct nv50_curs **pcurs)
-{
-	static const struct nvif_mclass curses[] = {
-		{ GK104_DISP_CURSOR, 0 },
-		{ GF110_DISP_CURSOR, 0 },
-		{ GT214_DISP_CURSOR, 0 },
-		{   G82_DISP_CURSOR, 0 },
-		{  NV50_DISP_CURSOR, 0 },
-		{}
-	};
-	struct nv50_disp_cursor_v0 args = {
-		.head = head->base.index,
-	};
-	struct nv50_disp *disp = nv50_disp(drm->dev);
-	struct nv50_curs *curs;
-	int cid, ret;
-
-	cid = nvif_mclass(&disp->disp->object, curses);
-	if (cid < 0) {
-		NV_ERROR(drm, "No supported cursor immediate class\n");
-		return cid;
-	}
-
-	if (!(curs = *pcurs = kzalloc(sizeof(*curs), GFP_KERNEL)))
-		return -ENOMEM;
-
-	ret = nv50_wndw_ctor(&nv50_curs, drm->dev, DRM_PLANE_TYPE_CURSOR,
-			     "curs", head->base.index, &disp->mast.base,
-			     nv50_curs_format, ARRAY_SIZE(nv50_curs_format),
-			     &curs->wndw);
-	if (ret) {
-		kfree(curs);
-		return ret;
-	}
-
-	ret = nvif_object_init(&disp->disp->object, 0, curses[cid].oclass,
-			       &args, sizeof(args), &curs->chan);
-	if (ret) {
-		NV_ERROR(drm, "curs%04x allocation failed: %d\n",
-			 curses[cid].oclass, ret);
-		return ret;
-	}
-
-	return 0;
-}
-
-/******************************************************************************
- * Primary plane
- *****************************************************************************/
-#define nv50_base(p) container_of((p), struct nv50_base, wndw)
-
-struct nv50_base {
-	struct nv50_wndw wndw;
-	struct nv50_sync chan;
-	int id;
-};
-
-static int
-nv50_base_notify(struct nvif_notify *notify)
-{
-	return NVIF_NOTIFY_KEEP;
-}
-
-static void
-nv50_base_lut(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
-{
-	struct nv50_base *base = nv50_base(wndw);
-	u32 *push;
-	if ((push = evo_wait(&base->chan, 2))) {
-		evo_mthd(push, 0x00e0, 1);
-		evo_data(push, asyw->lut.enable << 30);
-		evo_kick(push, &base->chan);
-	}
-}
-
-static void
-nv50_base_image_clr(struct nv50_wndw *wndw)
-{
-	struct nv50_base *base = nv50_base(wndw);
-	u32 *push;
-	if ((push = evo_wait(&base->chan, 4))) {
-		evo_mthd(push, 0x0084, 1);
-		evo_data(push, 0x00000000);
-		evo_mthd(push, 0x00c0, 1);
-		evo_data(push, 0x00000000);
-		evo_kick(push, &base->chan);
-	}
-}
-
-static void
-nv50_base_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
-{
-	struct nv50_base *base = nv50_base(wndw);
-	const s32 oclass = base->chan.base.base.user.oclass;
-	u32 *push;
-	if ((push = evo_wait(&base->chan, 10))) {
-		evo_mthd(push, 0x0084, 1);
-		evo_data(push, (asyw->image.mode << 8) |
-			       (asyw->image.interval << 4));
-		evo_mthd(push, 0x00c0, 1);
-		evo_data(push, asyw->image.handle);
-		if (oclass < G82_DISP_BASE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0800, 5);
-			evo_data(push, asyw->image.offset >> 8);
-			evo_data(push, 0x00000000);
-			evo_data(push, (asyw->image.h << 16) | asyw->image.w);
-			evo_data(push, (asyw->image.layout << 20) |
-					asyw->image.pitch |
-					asyw->image.block);
-			evo_data(push, (asyw->image.kind << 16) |
-				       (asyw->image.format << 8));
-		} else
-		if (oclass < GF110_DISP_BASE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0800, 5);
-			evo_data(push, asyw->image.offset >> 8);
-			evo_data(push, 0x00000000);
-			evo_data(push, (asyw->image.h << 16) | asyw->image.w);
-			evo_data(push, (asyw->image.layout << 20) |
-					asyw->image.pitch |
-					asyw->image.block);
-			evo_data(push, asyw->image.format << 8);
-		} else {
-			evo_mthd(push, 0x0400, 5);
-			evo_data(push, asyw->image.offset >> 8);
-			evo_data(push, 0x00000000);
-			evo_data(push, (asyw->image.h << 16) | asyw->image.w);
-			evo_data(push, (asyw->image.layout << 24) |
-					asyw->image.pitch |
-					asyw->image.block);
-			evo_data(push, asyw->image.format << 8);
-		}
-		evo_kick(push, &base->chan);
-	}
-}
-
-static void
-nv50_base_ntfy_clr(struct nv50_wndw *wndw)
-{
-	struct nv50_base *base = nv50_base(wndw);
-	u32 *push;
-	if ((push = evo_wait(&base->chan, 2))) {
-		evo_mthd(push, 0x00a4, 1);
-		evo_data(push, 0x00000000);
-		evo_kick(push, &base->chan);
-	}
-}
-
-static void
-nv50_base_ntfy_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
-{
-	struct nv50_base *base = nv50_base(wndw);
-	u32 *push;
-	if ((push = evo_wait(&base->chan, 3))) {
-		evo_mthd(push, 0x00a0, 2);
-		evo_data(push, (asyw->ntfy.awaken << 30) | asyw->ntfy.offset);
-		evo_data(push, asyw->ntfy.handle);
-		evo_kick(push, &base->chan);
-	}
-}
-
-static void
-nv50_base_sema_clr(struct nv50_wndw *wndw)
-{
-	struct nv50_base *base = nv50_base(wndw);
-	u32 *push;
-	if ((push = evo_wait(&base->chan, 2))) {
-		evo_mthd(push, 0x0094, 1);
-		evo_data(push, 0x00000000);
-		evo_kick(push, &base->chan);
-	}
-}
-
-static void
-nv50_base_sema_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
-{
-	struct nv50_base *base = nv50_base(wndw);
-	u32 *push;
-	if ((push = evo_wait(&base->chan, 5))) {
-		evo_mthd(push, 0x0088, 4);
-		evo_data(push, asyw->sema.offset);
-		evo_data(push, asyw->sema.acquire);
-		evo_data(push, asyw->sema.release);
-		evo_data(push, asyw->sema.handle);
-		evo_kick(push, &base->chan);
-	}
-}
-
-static u32
-nv50_base_update(struct nv50_wndw *wndw, u32 interlock)
-{
-	struct nv50_base *base = nv50_base(wndw);
-	u32 *push;
-
-	if (!(push = evo_wait(&base->chan, 2)))
-		return 0;
-	evo_mthd(push, 0x0080, 1);
-	evo_data(push, interlock);
-	evo_kick(push, &base->chan);
-
-	if (base->chan.base.base.user.oclass < GF110_DISP_BASE_CHANNEL_DMA)
-		return interlock ? 2 << (base->id * 8) : 0;
-	return interlock ? 2 << (base->id * 4) : 0;
-}
-
-static int
-nv50_base_ntfy_wait_begun(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
-{
-	struct nouveau_drm *drm = nouveau_drm(wndw->plane.dev);
-	struct nv50_disp *disp = nv50_disp(wndw->plane.dev);
-	if (nvif_msec(&drm->client.device, 2000ULL,
-		u32 data = nouveau_bo_rd32(disp->sync, asyw->ntfy.offset / 4);
-		if ((data & 0xc0000000) == 0x40000000)
-			break;
-		usleep_range(1, 2);
-	) < 0)
-		return -ETIMEDOUT;
-	return 0;
-}
-
-static void
-nv50_base_release(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
-		  struct nv50_head_atom *asyh)
-{
-	asyh->base.cpp = 0;
-}
-
-static int
-nv50_base_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
-		  struct nv50_head_atom *asyh)
-{
-	const struct drm_framebuffer *fb = asyw->state.fb;
-	int ret;
-
-	if (!fb->format->depth)
-		return -EINVAL;
-
-	ret = drm_atomic_helper_check_plane_state(&asyw->state, &asyh->state,
-						  DRM_PLANE_HELPER_NO_SCALING,
-						  DRM_PLANE_HELPER_NO_SCALING,
-						  false, true);
-	if (ret)
-		return ret;
-
-	asyh->base.depth = fb->format->depth;
-	asyh->base.cpp = fb->format->cpp[0];
-	asyh->base.x = asyw->state.src.x1 >> 16;
-	asyh->base.y = asyw->state.src.y1 >> 16;
-	asyh->base.w = asyw->state.fb->width;
-	asyh->base.h = asyw->state.fb->height;
-
-	switch (fb->format->format) {
-	case DRM_FORMAT_C8         : asyw->image.format = 0x1e; break;
-	case DRM_FORMAT_RGB565     : asyw->image.format = 0xe8; break;
-	case DRM_FORMAT_XRGB1555   :
-	case DRM_FORMAT_ARGB1555   : asyw->image.format = 0xe9; break;
-	case DRM_FORMAT_XRGB8888   :
-	case DRM_FORMAT_ARGB8888   : asyw->image.format = 0xcf; break;
-	case DRM_FORMAT_XBGR2101010:
-	case DRM_FORMAT_ABGR2101010: asyw->image.format = 0xd1; break;
-	case DRM_FORMAT_XBGR8888   :
-	case DRM_FORMAT_ABGR8888   : asyw->image.format = 0xd5; break;
-	default:
-		WARN_ON(1);
-		return -EINVAL;
-	}
-
-	asyw->lut.enable = 1;
-	asyw->set.image = true;
-	return 0;
-}
-
-static void *
-nv50_base_dtor(struct nv50_wndw *wndw)
-{
-	struct nv50_base *base = nv50_base(wndw);
-	nv50_dmac_destroy(&base->chan.base);
-	return base;
-}
-
-static const u32
-nv50_base_format[] = {
-	DRM_FORMAT_C8,
-	DRM_FORMAT_RGB565,
-	DRM_FORMAT_XRGB1555,
-	DRM_FORMAT_ARGB1555,
-	DRM_FORMAT_XRGB8888,
-	DRM_FORMAT_ARGB8888,
-	DRM_FORMAT_XBGR2101010,
-	DRM_FORMAT_ABGR2101010,
-	DRM_FORMAT_XBGR8888,
-	DRM_FORMAT_ABGR8888,
-};
-
-static const struct nv50_wndw_func
-nv50_base = {
-	.dtor = nv50_base_dtor,
-	.acquire = nv50_base_acquire,
-	.release = nv50_base_release,
-	.sema_set = nv50_base_sema_set,
-	.sema_clr = nv50_base_sema_clr,
-	.ntfy_set = nv50_base_ntfy_set,
-	.ntfy_clr = nv50_base_ntfy_clr,
-	.ntfy_wait_begun = nv50_base_ntfy_wait_begun,
-	.image_set = nv50_base_image_set,
-	.image_clr = nv50_base_image_clr,
-	.lut = nv50_base_lut,
-	.update = nv50_base_update,
-};
-
-static int
-nv50_base_new(struct nouveau_drm *drm, struct nv50_head *head,
-	      struct nv50_base **pbase)
-{
-	struct nv50_disp *disp = nv50_disp(drm->dev);
-	struct nv50_base *base;
-	int ret;
-
-	if (!(base = *pbase = kzalloc(sizeof(*base), GFP_KERNEL)))
-		return -ENOMEM;
-	base->id = head->base.index;
-	base->wndw.ntfy = EVO_FLIP_NTFY0(base->id);
-	base->wndw.sema = EVO_FLIP_SEM0(base->id);
-	base->wndw.data = 0x00000000;
-
-	ret = nv50_wndw_ctor(&nv50_base, drm->dev, DRM_PLANE_TYPE_PRIMARY,
-			     "base", base->id, &base->chan.base,
-			     nv50_base_format, ARRAY_SIZE(nv50_base_format),
-			     &base->wndw);
-	if (ret) {
-		kfree(base);
-		return ret;
-	}
-
-	ret = nv50_base_create(&drm->client.device, &disp->disp->object,
-			       base->id, disp->sync->bo.offset, &base->chan);
-	if (ret)
-		return ret;
-
-	return nvif_notify_init(&base->chan.base.base.user, nv50_base_notify,
-				false,
-				NV50_DISP_BASE_CHANNEL_DMA_V0_NTFY_UEVENT,
-				&(struct nvif_notify_uevent_req) {},
-				sizeof(struct nvif_notify_uevent_req),
-				sizeof(struct nvif_notify_uevent_rep),
-				&base->wndw.notify);
-}
-
-/******************************************************************************
- * Head
- *****************************************************************************/
-static void
-nv50_head_procamp(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
-	u32 *push;
-	if ((push = evo_wait(core, 2))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
-			evo_mthd(push, 0x08a8 + (head->base.index * 0x400), 1);
-		else
-			evo_mthd(push, 0x0498 + (head->base.index * 0x300), 1);
-		evo_data(push, (asyh->procamp.sat.sin << 20) |
-			       (asyh->procamp.sat.cos << 8));
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_dither(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
-	u32 *push;
-	if ((push = evo_wait(core, 2))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
-			evo_mthd(push, 0x08a0 + (head->base.index * 0x0400), 1);
-		else
-		if (core->base.user.oclass < GK104_DISP_CORE_CHANNEL_DMA)
-			evo_mthd(push, 0x0490 + (head->base.index * 0x0300), 1);
-		else
-			evo_mthd(push, 0x04a0 + (head->base.index * 0x0300), 1);
-		evo_data(push, (asyh->dither.mode << 3) |
-			       (asyh->dither.bits << 1) |
-			        asyh->dither.enable);
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_ovly(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
-	u32 bounds = 0;
-	u32 *push;
-
-	if (asyh->base.cpp) {
-		switch (asyh->base.cpp) {
-		case 8: bounds |= 0x00000500; break;
-		case 4: bounds |= 0x00000300; break;
-		case 2: bounds |= 0x00000100; break;
-		default:
-			WARN_ON(1);
-			break;
-		}
-		bounds |= 0x00000001;
-	}
-
-	if ((push = evo_wait(core, 2))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
-			evo_mthd(push, 0x0904 + head->base.index * 0x400, 1);
-		else
-			evo_mthd(push, 0x04d4 + head->base.index * 0x300, 1);
-		evo_data(push, bounds);
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_base(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
-	u32 bounds = 0;
-	u32 *push;
-
-	if (asyh->base.cpp) {
-		switch (asyh->base.cpp) {
-		case 8: bounds |= 0x00000500; break;
-		case 4: bounds |= 0x00000300; break;
-		case 2: bounds |= 0x00000100; break;
-		case 1: bounds |= 0x00000000; break;
-		default:
-			WARN_ON(1);
-			break;
-		}
-		bounds |= 0x00000001;
-	}
-
-	if ((push = evo_wait(core, 2))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
-			evo_mthd(push, 0x0900 + head->base.index * 0x400, 1);
-		else
-			evo_mthd(push, 0x04d0 + head->base.index * 0x300, 1);
-		evo_data(push, bounds);
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_curs_clr(struct nv50_head *head)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
-	u32 *push;
-	if ((push = evo_wait(core, 4))) {
-		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0880 + head->base.index * 0x400, 1);
-			evo_data(push, 0x05000000);
-		} else
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0880 + head->base.index * 0x400, 1);
-			evo_data(push, 0x05000000);
-			evo_mthd(push, 0x089c + head->base.index * 0x400, 1);
-			evo_data(push, 0x00000000);
-		} else {
-			evo_mthd(push, 0x0480 + head->base.index * 0x300, 1);
-			evo_data(push, 0x05000000);
-			evo_mthd(push, 0x048c + head->base.index * 0x300, 1);
-			evo_data(push, 0x00000000);
-		}
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_curs_set(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
-	u32 *push;
-	if ((push = evo_wait(core, 5))) {
-		if (core->base.user.oclass < G82_DISP_BASE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0880 + head->base.index * 0x400, 2);
-			evo_data(push, 0x80000000 | (asyh->curs.layout << 26) |
-						    (asyh->curs.format << 24));
-			evo_data(push, asyh->curs.offset >> 8);
-		} else
-		if (core->base.user.oclass < GF110_DISP_BASE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0880 + head->base.index * 0x400, 2);
-			evo_data(push, 0x80000000 | (asyh->curs.layout << 26) |
-						    (asyh->curs.format << 24));
-			evo_data(push, asyh->curs.offset >> 8);
-			evo_mthd(push, 0x089c + head->base.index * 0x400, 1);
-			evo_data(push, asyh->curs.handle);
-		} else {
-			evo_mthd(push, 0x0480 + head->base.index * 0x300, 2);
-			evo_data(push, 0x80000000 | (asyh->curs.layout << 26) |
-						    (asyh->curs.format << 24));
-			evo_data(push, asyh->curs.offset >> 8);
-			evo_mthd(push, 0x048c + head->base.index * 0x300, 1);
-			evo_data(push, asyh->curs.handle);
-		}
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_core_clr(struct nv50_head *head)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
-	u32 *push;
-	if ((push = evo_wait(core, 2))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
-			evo_mthd(push, 0x0874 + head->base.index * 0x400, 1);
-		else
-			evo_mthd(push, 0x0474 + head->base.index * 0x300, 1);
-		evo_data(push, 0x00000000);
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_core_set(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
-	u32 *push;
-	if ((push = evo_wait(core, 9))) {
-		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0860 + head->base.index * 0x400, 1);
-			evo_data(push, asyh->core.offset >> 8);
-			evo_mthd(push, 0x0868 + head->base.index * 0x400, 4);
-			evo_data(push, (asyh->core.h << 16) | asyh->core.w);
-			evo_data(push, asyh->core.layout << 20 |
-				       (asyh->core.pitch >> 8) << 8 |
-				       asyh->core.block);
-			evo_data(push, asyh->core.kind << 16 |
-				       asyh->core.format << 8);
-			evo_data(push, asyh->core.handle);
-			evo_mthd(push, 0x08c0 + head->base.index * 0x400, 1);
-			evo_data(push, (asyh->core.y << 16) | asyh->core.x);
-			/* EVO will complain with INVALID_STATE if we have an
-			 * active cursor and (re)specify HeadSetContextDmaIso
-			 * without also updating HeadSetOffsetCursor.
-			 */
-			asyh->set.curs = asyh->curs.visible;
-		} else
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0860 + head->base.index * 0x400, 1);
-			evo_data(push, asyh->core.offset >> 8);
-			evo_mthd(push, 0x0868 + head->base.index * 0x400, 4);
-			evo_data(push, (asyh->core.h << 16) | asyh->core.w);
-			evo_data(push, asyh->core.layout << 20 |
-				       (asyh->core.pitch >> 8) << 8 |
-				       asyh->core.block);
-			evo_data(push, asyh->core.format << 8);
-			evo_data(push, asyh->core.handle);
-			evo_mthd(push, 0x08c0 + head->base.index * 0x400, 1);
-			evo_data(push, (asyh->core.y << 16) | asyh->core.x);
-		} else {
-			evo_mthd(push, 0x0460 + head->base.index * 0x300, 1);
-			evo_data(push, asyh->core.offset >> 8);
-			evo_mthd(push, 0x0468 + head->base.index * 0x300, 4);
-			evo_data(push, (asyh->core.h << 16) | asyh->core.w);
-			evo_data(push, asyh->core.layout << 24 |
-				       (asyh->core.pitch >> 8) << 8 |
-				       asyh->core.block);
-			evo_data(push, asyh->core.format << 8);
-			evo_data(push, asyh->core.handle);
-			evo_mthd(push, 0x04b0 + head->base.index * 0x300, 1);
-			evo_data(push, (asyh->core.y << 16) | asyh->core.x);
-		}
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_lut_clr(struct nv50_head *head)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
-	u32 *push;
-	if ((push = evo_wait(core, 4))) {
-		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0840 + (head->base.index * 0x400), 1);
-			evo_data(push, 0x40000000);
-		} else
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0840 + (head->base.index * 0x400), 1);
-			evo_data(push, 0x40000000);
-			evo_mthd(push, 0x085c + (head->base.index * 0x400), 1);
-			evo_data(push, 0x00000000);
-		} else {
-			evo_mthd(push, 0x0440 + (head->base.index * 0x300), 1);
-			evo_data(push, 0x03000000);
-			evo_mthd(push, 0x045c + (head->base.index * 0x300), 1);
-			evo_data(push, 0x00000000);
-		}
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_lut_load(struct drm_property_blob *blob, int mode,
-		   struct nouveau_bo *nvbo)
-{
-	struct drm_color_lut *in = (struct drm_color_lut *)blob->data;
-	void __iomem *lut = (u8 *)nvbo_kmap_obj_iovirtual(nvbo);
-	const int size = blob->length / sizeof(*in);
-	int bits, shift, i;
-	u16 zero, r, g, b;
-
-	/* This can't happen.. But it shuts the compiler up. */
-	if (WARN_ON(size != 256))
-		return;
-
-	switch (mode) {
-	case 0: /* LORES. */
-	case 1: /* HIRES. */
-		bits = 11;
-		shift = 3;
-		zero = 0x0000;
-		break;
-	case 7: /* INTERPOLATE_257_UNITY_RANGE. */
-		bits = 14;
-		shift = 0;
-		zero = 0x6000;
-		break;
-	default:
-		WARN_ON(1);
-		return;
-	}
-
-	for (i = 0; i < size; i++) {
-		r = (drm_color_lut_extract(in[i].  red, bits) + zero) << shift;
-		g = (drm_color_lut_extract(in[i].green, bits) + zero) << shift;
-		b = (drm_color_lut_extract(in[i]. blue, bits) + zero) << shift;
-		writew(r, lut + (i * 0x08) + 0);
-		writew(g, lut + (i * 0x08) + 2);
-		writew(b, lut + (i * 0x08) + 4);
-	}
-
-	/* INTERPOLATE modes require a "next" entry to interpolate with,
-	 * so we replicate the last entry to deal with this for now.
-	 */
-	writew(r, lut + (i * 0x08) + 0);
-	writew(g, lut + (i * 0x08) + 2);
-	writew(b, lut + (i * 0x08) + 4);
-}
-
-static void
-nv50_head_lut_set(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
-	u32 *push;
-	if ((push = evo_wait(core, 7))) {
-		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0840 + (head->base.index * 0x400), 2);
-			evo_data(push, 0x80000000 | asyh->lut.mode << 30);
-			evo_data(push, asyh->lut.offset >> 8);
-		} else
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0840 + (head->base.index * 0x400), 2);
-			evo_data(push, 0x80000000 | asyh->lut.mode << 30);
-			evo_data(push, asyh->lut.offset >> 8);
-			evo_mthd(push, 0x085c + (head->base.index * 0x400), 1);
-			evo_data(push, asyh->lut.handle);
-		} else {
-			evo_mthd(push, 0x0440 + (head->base.index * 0x300), 4);
-			evo_data(push, 0x80000000 | asyh->lut.mode << 24);
-			evo_data(push, asyh->lut.offset >> 8);
-			evo_data(push, 0x00000000);
-			evo_data(push, 0x00000000);
-			evo_mthd(push, 0x045c + (head->base.index * 0x300), 1);
-			evo_data(push, asyh->lut.handle);
-		}
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_mode(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
-	struct nv50_head_mode *m = &asyh->mode;
-	u32 *push;
-	if ((push = evo_wait(core, 14))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0804 + (head->base.index * 0x400), 2);
-			evo_data(push, 0x00800000 | m->clock);
-			evo_data(push, m->interlace ? 0x00000002 : 0x00000000);
-			evo_mthd(push, 0x0810 + (head->base.index * 0x400), 7);
-			evo_data(push, 0x00000000);
-			evo_data(push, (m->v.active  << 16) | m->h.active );
-			evo_data(push, (m->v.synce   << 16) | m->h.synce  );
-			evo_data(push, (m->v.blanke  << 16) | m->h.blanke );
-			evo_data(push, (m->v.blanks  << 16) | m->h.blanks );
-			evo_data(push, (m->v.blank2e << 16) | m->v.blank2s);
-			evo_data(push, asyh->mode.v.blankus);
-			evo_mthd(push, 0x082c + (head->base.index * 0x400), 1);
-			evo_data(push, 0x00000000);
-		} else {
-			evo_mthd(push, 0x0410 + (head->base.index * 0x300), 6);
-			evo_data(push, 0x00000000);
-			evo_data(push, (m->v.active  << 16) | m->h.active );
-			evo_data(push, (m->v.synce   << 16) | m->h.synce  );
-			evo_data(push, (m->v.blanke  << 16) | m->h.blanke );
-			evo_data(push, (m->v.blanks  << 16) | m->h.blanks );
-			evo_data(push, (m->v.blank2e << 16) | m->v.blank2s);
-			evo_mthd(push, 0x042c + (head->base.index * 0x300), 2);
-			evo_data(push, 0x00000000); /* ??? */
-			evo_data(push, 0xffffff00);
-			evo_mthd(push, 0x0450 + (head->base.index * 0x300), 3);
-			evo_data(push, m->clock * 1000);
-			evo_data(push, 0x00200000); /* ??? */
-			evo_data(push, m->clock * 1000);
-		}
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_view(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
-	u32 *push;
-	if ((push = evo_wait(core, 10))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x08a4 + (head->base.index * 0x400), 1);
-			evo_data(push, 0x00000000);
-			evo_mthd(push, 0x08c8 + (head->base.index * 0x400), 1);
-			evo_data(push, (asyh->view.iH << 16) | asyh->view.iW);
-			evo_mthd(push, 0x08d8 + (head->base.index * 0x400), 2);
-			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
-			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
-		} else {
-			evo_mthd(push, 0x0494 + (head->base.index * 0x300), 1);
-			evo_data(push, 0x00000000);
-			evo_mthd(push, 0x04b8 + (head->base.index * 0x300), 1);
-			evo_data(push, (asyh->view.iH << 16) | asyh->view.iW);
-			evo_mthd(push, 0x04c0 + (head->base.index * 0x300), 3);
-			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
-			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
-			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
-		}
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_flush_clr(struct nv50_head *head, struct nv50_head_atom *asyh, bool y)
-{
-	if (asyh->clr.ilut && (!asyh->set.ilut || y))
-		nv50_head_lut_clr(head);
-	if (asyh->clr.core && (!asyh->set.core || y))
-		nv50_head_core_clr(head);
-	if (asyh->clr.curs && (!asyh->set.curs || y))
-		nv50_head_curs_clr(head);
-}
-
-static void
-nv50_head_flush_set(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	if (asyh->set.view   ) nv50_head_view    (head, asyh);
-	if (asyh->set.mode   ) nv50_head_mode    (head, asyh);
-	if (asyh->set.ilut   ) {
-		struct nouveau_bo *nvbo = head->lut.nvbo[head->lut.next];
-		struct drm_property_blob *blob = asyh->state.gamma_lut;
-		if (blob)
-			nv50_head_lut_load(blob, asyh->lut.mode, nvbo);
-		asyh->lut.offset = nvbo->bo.offset;
-		head->lut.next ^= 1;
-		nv50_head_lut_set(head, asyh);
-	}
-	if (asyh->set.core   ) nv50_head_core_set(head, asyh);
-	if (asyh->set.curs   ) nv50_head_curs_set(head, asyh);
-	if (asyh->set.base   ) nv50_head_base    (head, asyh);
-	if (asyh->set.ovly   ) nv50_head_ovly    (head, asyh);
-	if (asyh->set.dither ) nv50_head_dither  (head, asyh);
-	if (asyh->set.procamp) nv50_head_procamp (head, asyh);
-}
-
-static void
-nv50_head_atomic_check_procamp(struct nv50_head_atom *armh,
-			       struct nv50_head_atom *asyh,
-			       struct nouveau_conn_atom *asyc)
-{
-	const int vib = asyc->procamp.color_vibrance - 100;
-	const int hue = asyc->procamp.vibrant_hue - 90;
-	const int adj = (vib > 0) ? 50 : 0;
-	asyh->procamp.sat.cos = ((vib * 2047 + adj) / 100) & 0xfff;
-	asyh->procamp.sat.sin = ((hue * 2047) / 100) & 0xfff;
-	asyh->set.procamp = true;
-}
-
-static void
-nv50_head_atomic_check_dither(struct nv50_head_atom *armh,
-			      struct nv50_head_atom *asyh,
-			      struct nouveau_conn_atom *asyc)
-{
-	struct drm_connector *connector = asyc->state.connector;
-	u32 mode = 0x00;
-
-	if (asyc->dither.mode == DITHERING_MODE_AUTO) {
-		if (asyh->base.depth > connector->display_info.bpc * 3)
-			mode = DITHERING_MODE_DYNAMIC2X2;
-	} else {
-		mode = asyc->dither.mode;
-	}
-
-	if (asyc->dither.depth == DITHERING_DEPTH_AUTO) {
-		if (connector->display_info.bpc >= 8)
-			mode |= DITHERING_DEPTH_8BPC;
-	} else {
-		mode |= asyc->dither.depth;
-	}
-
-	asyh->dither.enable = mode;
-	asyh->dither.bits = mode >> 1;
-	asyh->dither.mode = mode >> 3;
-	asyh->set.dither = true;
-}
-
-static void
-nv50_head_atomic_check_view(struct nv50_head_atom *armh,
-			    struct nv50_head_atom *asyh,
-			    struct nouveau_conn_atom *asyc)
-{
-	struct drm_connector *connector = asyc->state.connector;
-	struct drm_display_mode *omode = &asyh->state.adjusted_mode;
-	struct drm_display_mode *umode = &asyh->state.mode;
-	int mode = asyc->scaler.mode;
-	struct edid *edid;
-	int umode_vdisplay, omode_hdisplay, omode_vdisplay;
-
-	if (connector->edid_blob_ptr)
-		edid = (struct edid *)connector->edid_blob_ptr->data;
-	else
-		edid = NULL;
-
-	if (!asyc->scaler.full) {
-		if (mode == DRM_MODE_SCALE_NONE)
-			omode = umode;
-	} else {
-		/* Non-EDID LVDS/eDP mode. */
-		mode = DRM_MODE_SCALE_FULLSCREEN;
-	}
-
-	/* For the user-specified mode, we must ignore doublescan and
-	 * the like, but honor frame packing.
-	 */
-	umode_vdisplay = umode->vdisplay;
-	if ((umode->flags & DRM_MODE_FLAG_3D_MASK) == DRM_MODE_FLAG_3D_FRAME_PACKING)
-		umode_vdisplay += umode->vtotal;
-	asyh->view.iW = umode->hdisplay;
-	asyh->view.iH = umode_vdisplay;
-	/* For the output mode, we can just use the stock helper. */
-	drm_mode_get_hv_timing(omode, &omode_hdisplay, &omode_vdisplay);
-	asyh->view.oW = omode_hdisplay;
-	asyh->view.oH = omode_vdisplay;
-
-	/* Add overscan compensation if necessary, will keep the aspect
-	 * ratio the same as the backend mode unless overridden by the
-	 * user setting both hborder and vborder properties.
-	 */
-	if ((asyc->scaler.underscan.mode == UNDERSCAN_ON ||
-	    (asyc->scaler.underscan.mode == UNDERSCAN_AUTO &&
-	     drm_detect_hdmi_monitor(edid)))) {
-		u32 bX = asyc->scaler.underscan.hborder;
-		u32 bY = asyc->scaler.underscan.vborder;
-		u32 r = (asyh->view.oH << 19) / asyh->view.oW;
-
-		if (bX) {
-			asyh->view.oW -= (bX * 2);
-			if (bY) asyh->view.oH -= (bY * 2);
-			else    asyh->view.oH  = ((asyh->view.oW * r) + (r / 2)) >> 19;
-		} else {
-			asyh->view.oW -= (asyh->view.oW >> 4) + 32;
-			if (bY) asyh->view.oH -= (bY * 2);
-			else    asyh->view.oH  = ((asyh->view.oW * r) + (r / 2)) >> 19;
-		}
-	}
-
-	/* Handle CENTER/ASPECT scaling, taking into account the areas
-	 * removed already for overscan compensation.
-	 */
-	switch (mode) {
-	case DRM_MODE_SCALE_CENTER:
-		asyh->view.oW = min((u16)umode->hdisplay, asyh->view.oW);
-		asyh->view.oH = min((u16)umode_vdisplay, asyh->view.oH);
-		/* fall-through */
-	case DRM_MODE_SCALE_ASPECT:
-		if (asyh->view.oH < asyh->view.oW) {
-			u32 r = (asyh->view.iW << 19) / asyh->view.iH;
-			asyh->view.oW = ((asyh->view.oH * r) + (r / 2)) >> 19;
-		} else {
-			u32 r = (asyh->view.iH << 19) / asyh->view.iW;
-			asyh->view.oH = ((asyh->view.oW * r) + (r / 2)) >> 19;
-		}
-		break;
-	default:
-		break;
-	}
-
-	asyh->set.view = true;
-}
-
-static void
-nv50_head_atomic_check_lut(struct nv50_head *head,
-			   struct nv50_head_atom *armh,
-			   struct nv50_head_atom *asyh)
-{
-	struct nv50_disp *disp = nv50_disp(head->base.base.dev);
-
-	/* An I8 surface without an input LUT makes no sense, and
-	 * EVO will throw an error if you try.
-	 *
-	 * Legacy clients actually cause this due to the order in
-	 * which they call ioctls, so we will enable the LUT with
-	 * whatever contents the buffer already contains to avoid
-	 * triggering the error check.
-	 */
-	if (!asyh->state.gamma_lut && asyh->base.cpp != 1) {
-		asyh->lut.handle = 0;
-		asyh->clr.ilut = armh->lut.visible;
-		return;
-	}
-
-	if (disp->disp->object.oclass < GF110_DISP) {
-		asyh->lut.mode = (asyh->base.cpp == 1) ? 0 : 1;
-		asyh->set.ilut = true;
-	} else {
-		asyh->lut.mode = 7;
-		asyh->set.ilut = asyh->state.color_mgmt_changed;
-	}
-	asyh->lut.handle = disp->mast.base.vram.handle;
-}
-
-static void
-nv50_head_atomic_check_mode(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct drm_display_mode *mode = &asyh->state.adjusted_mode;
-	struct nv50_head_mode *m = &asyh->mode;
-	u32 blankus;
-
-	drm_mode_set_crtcinfo(mode, CRTC_INTERLACE_HALVE_V | CRTC_STEREO_DOUBLE);
-
-	/*
-	 * DRM modes are defined in terms of a repeating interval
-	 * starting with the active display area.  The hardware modes
-	 * are defined in terms of a repeating interval starting one
-	 * unit (pixel or line) into the sync pulse.  So, add bias.
-	 */
-
-	m->h.active = mode->crtc_htotal;
-	m->h.synce  = mode->crtc_hsync_end - mode->crtc_hsync_start - 1;
-	m->h.blanke = mode->crtc_hblank_end - mode->crtc_hsync_start - 1;
-	m->h.blanks = m->h.blanke + mode->crtc_hdisplay;
-
-	m->v.active = mode->crtc_vtotal;
-	m->v.synce  = mode->crtc_vsync_end - mode->crtc_vsync_start - 1;
-	m->v.blanke = mode->crtc_vblank_end - mode->crtc_vsync_start - 1;
-	m->v.blanks = m->v.blanke + mode->crtc_vdisplay;
-
-	/*XXX: Safe underestimate, even "0" works */
-	blankus = (m->v.active - mode->crtc_vdisplay - 2) * m->h.active;
-	blankus *= 1000;
-	blankus /= mode->crtc_clock;
-	m->v.blankus = blankus;
-
-	if (mode->flags & DRM_MODE_FLAG_INTERLACE) {
-		m->v.blank2e =  m->v.active + m->v.blanke;
-		m->v.blank2s =  m->v.blank2e + mode->crtc_vdisplay;
-		m->v.active  = (m->v.active * 2) + 1;
-		m->interlace = true;
-	} else {
-		m->v.blank2e = 0;
-		m->v.blank2s = 1;
-		m->interlace = false;
-	}
-	m->clock = mode->crtc_clock;
-
-	asyh->set.mode = true;
-}
-
-static int
-nv50_head_atomic_check(struct drm_crtc *crtc, struct drm_crtc_state *state)
-{
-	struct nouveau_drm *drm = nouveau_drm(crtc->dev);
-	struct nv50_disp *disp = nv50_disp(crtc->dev);
-	struct nv50_head *head = nv50_head(crtc);
-	struct nv50_head_atom *armh = nv50_head_atom(crtc->state);
-	struct nv50_head_atom *asyh = nv50_head_atom(state);
-	struct nouveau_conn_atom *asyc = NULL;
-	struct drm_connector_state *conns;
-	struct drm_connector *conn;
-	int i;
-
-	NV_ATOMIC(drm, "%s atomic_check %d\n", crtc->name, asyh->state.active);
-	if (asyh->state.active) {
-		for_each_new_connector_in_state(asyh->state.state, conn, conns, i) {
-			if (conns->crtc == crtc) {
-				asyc = nouveau_conn_atom(conns);
-				break;
-			}
-		}
-
-		if (armh->state.active) {
-			if (asyc) {
-				if (asyh->state.mode_changed)
-					asyc->set.scaler = true;
-				if (armh->base.depth != asyh->base.depth)
-					asyc->set.dither = true;
-			}
-		} else {
-			if (asyc)
-				asyc->set.mask = ~0;
-			asyh->set.mask = ~0;
-		}
-
-		if (asyh->state.mode_changed)
-			nv50_head_atomic_check_mode(head, asyh);
-
-		if (asyh->state.color_mgmt_changed ||
-		    asyh->base.cpp != armh->base.cpp)
-			nv50_head_atomic_check_lut(head, armh, asyh);
-		asyh->lut.visible = asyh->lut.handle != 0;
-
-		if (asyc) {
-			if (asyc->set.scaler)
-				nv50_head_atomic_check_view(armh, asyh, asyc);
-			if (asyc->set.dither)
-				nv50_head_atomic_check_dither(armh, asyh, asyc);
-			if (asyc->set.procamp)
-				nv50_head_atomic_check_procamp(armh, asyh, asyc);
-		}
-
-		if ((asyh->core.visible = (asyh->base.cpp != 0))) {
-			asyh->core.x = asyh->base.x;
-			asyh->core.y = asyh->base.y;
-			asyh->core.w = asyh->base.w;
-			asyh->core.h = asyh->base.h;
-		} else
-		if ((asyh->core.visible = asyh->curs.visible) ||
-		    (asyh->core.visible = asyh->lut.visible)) {
-			/*XXX: We need to either find some way of having the
-			 *     primary base layer appear black, while still
-			 *     being able to display the other layers, or we
-			 *     need to allocate a dummy black surface here.
-			 */
-			asyh->core.x = 0;
-			asyh->core.y = 0;
-			asyh->core.w = asyh->state.mode.hdisplay;
-			asyh->core.h = asyh->state.mode.vdisplay;
-		}
-		asyh->core.handle = disp->mast.base.vram.handle;
-		asyh->core.offset = 0;
-		asyh->core.format = 0xcf;
-		asyh->core.kind = 0;
-		asyh->core.layout = 1;
-		asyh->core.block = 0;
-		asyh->core.pitch = ALIGN(asyh->core.w, 64) * 4;
-		asyh->set.base = armh->base.cpp != asyh->base.cpp;
-		asyh->set.ovly = armh->ovly.cpp != asyh->ovly.cpp;
-	} else {
-		asyh->lut.visible = false;
-		asyh->core.visible = false;
-		asyh->curs.visible = false;
-		asyh->base.cpp = 0;
-		asyh->ovly.cpp = 0;
-	}
-
-	if (!drm_atomic_crtc_needs_modeset(&asyh->state)) {
-		if (asyh->core.visible) {
-			if (memcmp(&armh->core, &asyh->core, sizeof(asyh->core)))
-				asyh->set.core = true;
-		} else
-		if (armh->core.visible) {
-			asyh->clr.core = true;
-		}
-
-		if (asyh->curs.visible) {
-			if (memcmp(&armh->curs, &asyh->curs, sizeof(asyh->curs)))
-				asyh->set.curs = true;
-		} else
-		if (armh->curs.visible) {
-			asyh->clr.curs = true;
-		}
-	} else {
-		asyh->clr.ilut = armh->lut.visible;
-		asyh->clr.core = armh->core.visible;
-		asyh->clr.curs = armh->curs.visible;
-		asyh->set.ilut = asyh->lut.visible;
-		asyh->set.core = asyh->core.visible;
-		asyh->set.curs = asyh->curs.visible;
-	}
-
-	if (asyh->clr.mask || asyh->set.mask)
-		nv50_atom(asyh->state.state)->lock_core = true;
-	return 0;
-}
-
-static const struct drm_crtc_helper_funcs
-nv50_head_help = {
-	.atomic_check = nv50_head_atomic_check,
-};
-
-static void
-nv50_head_atomic_destroy_state(struct drm_crtc *crtc,
-			       struct drm_crtc_state *state)
-{
-	struct nv50_head_atom *asyh = nv50_head_atom(state);
-	__drm_atomic_helper_crtc_destroy_state(&asyh->state);
-	kfree(asyh);
-}
-
-static struct drm_crtc_state *
-nv50_head_atomic_duplicate_state(struct drm_crtc *crtc)
-{
-	struct nv50_head_atom *armh = nv50_head_atom(crtc->state);
-	struct nv50_head_atom *asyh;
-	if (!(asyh = kmalloc(sizeof(*asyh), GFP_KERNEL)))
-		return NULL;
-	__drm_atomic_helper_crtc_duplicate_state(crtc, &asyh->state);
-	asyh->view = armh->view;
-	asyh->mode = armh->mode;
-	asyh->lut  = armh->lut;
-	asyh->core = armh->core;
-	asyh->curs = armh->curs;
-	asyh->base = armh->base;
-	asyh->ovly = armh->ovly;
-	asyh->dither = armh->dither;
-	asyh->procamp = armh->procamp;
-	asyh->clr.mask = 0;
-	asyh->set.mask = 0;
-	return &asyh->state;
-}
-
-static void
-__drm_atomic_helper_crtc_reset(struct drm_crtc *crtc,
-			       struct drm_crtc_state *state)
-{
-	if (crtc->state)
-		crtc->funcs->atomic_destroy_state(crtc, crtc->state);
-	crtc->state = state;
-	crtc->state->crtc = crtc;
-}
-
-static void
-nv50_head_reset(struct drm_crtc *crtc)
-{
-	struct nv50_head_atom *asyh;
-
-	if (WARN_ON(!(asyh = kzalloc(sizeof(*asyh), GFP_KERNEL))))
-		return;
-
-	__drm_atomic_helper_crtc_reset(crtc, &asyh->state);
-}
-
-static void
-nv50_head_destroy(struct drm_crtc *crtc)
-{
-	struct nv50_head *head = nv50_head(crtc);
-	int i;
-
-	nv50_dmac_destroy(&head->ovly.base);
-	nv50_pioc_destroy(&head->oimm.base);
-
-	for (i = 0; i < ARRAY_SIZE(head->lut.nvbo); i++)
-		nouveau_bo_unmap_unpin_unref(&head->lut.nvbo[i]);
-
-	drm_crtc_cleanup(crtc);
-	kfree(crtc);
-}
-
-static const struct drm_crtc_funcs
-nv50_head_func = {
-	.reset = nv50_head_reset,
-	.gamma_set = drm_atomic_helper_legacy_gamma_set,
-	.destroy = nv50_head_destroy,
-	.set_config = drm_atomic_helper_set_config,
-	.page_flip = drm_atomic_helper_page_flip,
-	.atomic_duplicate_state = nv50_head_atomic_duplicate_state,
-	.atomic_destroy_state = nv50_head_atomic_destroy_state,
-};
-
-static int
-nv50_head_create(struct drm_device *dev, int index)
-{
-	struct nouveau_drm *drm = nouveau_drm(dev);
-	struct nvif_device *device = &drm->client.device;
-	struct nv50_disp *disp = nv50_disp(dev);
-	struct nv50_head *head;
-	struct nv50_base *base;
-	struct nv50_curs *curs;
-	struct drm_crtc *crtc;
-	int ret, i;
-
-	head = kzalloc(sizeof(*head), GFP_KERNEL);
-	if (!head)
-		return -ENOMEM;
-
-	head->base.index = index;
-	ret = nv50_base_new(drm, head, &base);
-	if (ret == 0)
-		ret = nv50_curs_new(drm, head, &curs);
-	if (ret) {
-		kfree(head);
-		return ret;
-	}
-
-	crtc = &head->base.base;
-	drm_crtc_init_with_planes(dev, crtc, &base->wndw.plane,
-				  &curs->wndw.plane, &nv50_head_func,
-				  "head-%d", head->base.index);
-	drm_crtc_helper_add(crtc, &nv50_head_help);
-	drm_mode_crtc_set_gamma_size(crtc, 256);
-
-	for (i = 0; i < ARRAY_SIZE(head->lut.nvbo); i++) {
-		ret = nouveau_bo_new_pin_map(&drm->client, 1025 * 8, 0x100,
-					     TTM_PL_FLAG_VRAM,
-					     &head->lut.nvbo[i]);
-		if (ret)
-			goto out;
-	}
-
-	/* allocate overlay resources */
-	ret = nv50_oimm_create(device, &disp->disp->object, index, &head->oimm);
-	if (ret)
-		goto out;
-
-	ret = nv50_ovly_create(device, &disp->disp->object, index,
-			       disp->sync->bo.offset, &head->ovly);
-	if (ret)
-		goto out;
-
-out:
-	if (ret)
-		nv50_head_destroy(crtc);
-	return ret;
-}
-
-/******************************************************************************
- * Output path helpers
- *****************************************************************************/
-static void
-nv50_outp_release(struct nouveau_encoder *nv_encoder)
-{
-	struct nv50_disp *disp = nv50_disp(nv_encoder->base.base.dev);
-	struct {
-		struct nv50_disp_mthd_v1 base;
-	} args = {
-		.base.version = 1,
-		.base.method = NV50_DISP_MTHD_V1_RELEASE,
-		.base.hasht  = nv_encoder->dcb->hasht,
-		.base.hashm  = nv_encoder->dcb->hashm,
-	};
-
-	nvif_mthd(&disp->disp->object, 0, &args, sizeof(args));
-	nv_encoder->or = -1;
-	nv_encoder->link = 0;
-}
-
-static int
-nv50_outp_acquire(struct nouveau_encoder *nv_encoder)
-{
-	struct nouveau_drm *drm = nouveau_drm(nv_encoder->base.base.dev);
-	struct nv50_disp *disp = nv50_disp(drm->dev);
-	struct {
-		struct nv50_disp_mthd_v1 base;
-		struct nv50_disp_acquire_v0 info;
-	} args = {
-		.base.version = 1,
-		.base.method = NV50_DISP_MTHD_V1_ACQUIRE,
-		.base.hasht  = nv_encoder->dcb->hasht,
-		.base.hashm  = nv_encoder->dcb->hashm,
-	};
-	int ret;
-
-	ret = nvif_mthd(&disp->disp->object, 0, &args, sizeof(args));
-	if (ret) {
-		NV_ERROR(drm, "error acquiring output path: %d\n", ret);
-		return ret;
-	}
-
-	nv_encoder->or = args.info.or;
-	nv_encoder->link = args.info.link;
-	return 0;
-}
-
-static int
-nv50_outp_atomic_check_view(struct drm_encoder *encoder,
-			    struct drm_crtc_state *crtc_state,
-			    struct drm_connector_state *conn_state,
-			    struct drm_display_mode *native_mode)
-{
-	struct drm_display_mode *adjusted_mode = &crtc_state->adjusted_mode;
-	struct drm_display_mode *mode = &crtc_state->mode;
-	struct drm_connector *connector = conn_state->connector;
-	struct nouveau_conn_atom *asyc = nouveau_conn_atom(conn_state);
-	struct nouveau_drm *drm = nouveau_drm(encoder->dev);
-
-	NV_ATOMIC(drm, "%s atomic_check\n", encoder->name);
-	asyc->scaler.full = false;
-	if (!native_mode)
-		return 0;
-
-	if (asyc->scaler.mode == DRM_MODE_SCALE_NONE) {
-		switch (connector->connector_type) {
-		case DRM_MODE_CONNECTOR_LVDS:
-		case DRM_MODE_CONNECTOR_eDP:
-			/* Force use of scaler for non-EDID modes. */
-			if (adjusted_mode->type & DRM_MODE_TYPE_DRIVER)
-				break;
-			mode = native_mode;
-			asyc->scaler.full = true;
-			break;
-		default:
-			break;
-		}
-	} else {
-		mode = native_mode;
-	}
-
-	if (!drm_mode_equal(adjusted_mode, mode)) {
-		drm_mode_copy(adjusted_mode, mode);
-		crtc_state->mode_changed = true;
-	}
-
-	return 0;
-}
-
-static int
-nv50_outp_atomic_check(struct drm_encoder *encoder,
-		       struct drm_crtc_state *crtc_state,
-		       struct drm_connector_state *conn_state)
-{
-	struct nouveau_connector *nv_connector =
-		nouveau_connector(conn_state->connector);
-	return nv50_outp_atomic_check_view(encoder, crtc_state, conn_state,
-					   nv_connector->native_mode);
-}
-
-/******************************************************************************
- * DAC
- *****************************************************************************/
-static void
-nv50_dac_disable(struct drm_encoder *encoder)
-{
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nv50_mast *mast = nv50_mast(encoder->dev);
-	const int or = nv_encoder->or;
-	u32 *push;
-
-	if (nv_encoder->crtc) {
-		push = evo_wait(mast, 4);
-		if (push) {
-			if (nv50_vers(mast) < GF110_DISP_CORE_CHANNEL_DMA) {
-				evo_mthd(push, 0x0400 + (or * 0x080), 1);
-				evo_data(push, 0x00000000);
-			} else {
-				evo_mthd(push, 0x0180 + (or * 0x020), 1);
-				evo_data(push, 0x00000000);
-			}
-			evo_kick(push, mast);
-		}
-	}
-
-	nv_encoder->crtc = NULL;
-	nv50_outp_release(nv_encoder);
-}
-
-static void
-nv50_dac_enable(struct drm_encoder *encoder)
-{
-	struct nv50_mast *mast = nv50_mast(encoder->dev);
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
-	struct drm_display_mode *mode = &nv_crtc->base.state->adjusted_mode;
-	u32 *push;
-
-	nv50_outp_acquire(nv_encoder);
-
-	push = evo_wait(mast, 8);
-	if (push) {
-		if (nv50_vers(mast) < GF110_DISP_CORE_CHANNEL_DMA) {
-			u32 syncs = 0x00000000;
-
-			if (mode->flags & DRM_MODE_FLAG_NHSYNC)
-				syncs |= 0x00000001;
-			if (mode->flags & DRM_MODE_FLAG_NVSYNC)
-				syncs |= 0x00000002;
-
-			evo_mthd(push, 0x0400 + (nv_encoder->or * 0x080), 2);
-			evo_data(push, 1 << nv_crtc->index);
-			evo_data(push, syncs);
-		} else {
-			u32 magic = 0x31ec6000 | (nv_crtc->index << 25);
-			u32 syncs = 0x00000001;
-
-			if (mode->flags & DRM_MODE_FLAG_NHSYNC)
-				syncs |= 0x00000008;
-			if (mode->flags & DRM_MODE_FLAG_NVSYNC)
-				syncs |= 0x00000010;
-
-			if (mode->flags & DRM_MODE_FLAG_INTERLACE)
-				magic |= 0x00000001;
-
-			evo_mthd(push, 0x0404 + (nv_crtc->index * 0x300), 2);
-			evo_data(push, syncs);
-			evo_data(push, magic);
-			evo_mthd(push, 0x0180 + (nv_encoder->or * 0x020), 1);
-			evo_data(push, 1 << nv_crtc->index);
-		}
-
-		evo_kick(push, mast);
-	}
-
-	nv_encoder->crtc = encoder->crtc;
-}
-
-static enum drm_connector_status
-nv50_dac_detect(struct drm_encoder *encoder, struct drm_connector *connector)
-{
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nv50_disp *disp = nv50_disp(encoder->dev);
-	struct {
-		struct nv50_disp_mthd_v1 base;
-		struct nv50_disp_dac_load_v0 load;
-	} args = {
-		.base.version = 1,
-		.base.method = NV50_DISP_MTHD_V1_DAC_LOAD,
-		.base.hasht  = nv_encoder->dcb->hasht,
-		.base.hashm  = nv_encoder->dcb->hashm,
-	};
-	int ret;
-
-	args.load.data = nouveau_drm(encoder->dev)->vbios.dactestval;
-	if (args.load.data == 0)
-		args.load.data = 340;
-
-	ret = nvif_mthd(&disp->disp->object, 0, &args, sizeof(args));
-	if (ret || !args.load.load)
-		return connector_status_disconnected;
-
-	return connector_status_connected;
-}
-
-static const struct drm_encoder_helper_funcs
-nv50_dac_help = {
-	.atomic_check = nv50_outp_atomic_check,
-	.enable = nv50_dac_enable,
-	.disable = nv50_dac_disable,
-	.detect = nv50_dac_detect
-};
-
-static void
-nv50_dac_destroy(struct drm_encoder *encoder)
-{
-	drm_encoder_cleanup(encoder);
-	kfree(encoder);
-}
-
-static const struct drm_encoder_funcs
-nv50_dac_func = {
-	.destroy = nv50_dac_destroy,
-};
-
-static int
-nv50_dac_create(struct drm_connector *connector, struct dcb_output *dcbe)
-{
-	struct nouveau_drm *drm = nouveau_drm(connector->dev);
-	struct nvkm_i2c *i2c = nvxx_i2c(&drm->client.device);
-	struct nvkm_i2c_bus *bus;
-	struct nouveau_encoder *nv_encoder;
-	struct drm_encoder *encoder;
-	int type = DRM_MODE_ENCODER_DAC;
-
-	nv_encoder = kzalloc(sizeof(*nv_encoder), GFP_KERNEL);
-	if (!nv_encoder)
-		return -ENOMEM;
-	nv_encoder->dcb = dcbe;
-
-	bus = nvkm_i2c_bus_find(i2c, dcbe->i2c_index);
-	if (bus)
-		nv_encoder->i2c = &bus->i2c;
-
-	encoder = to_drm_encoder(nv_encoder);
-	encoder->possible_crtcs = dcbe->heads;
-	encoder->possible_clones = 0;
-	drm_encoder_init(connector->dev, encoder, &nv50_dac_func, type,
-			 "dac-%04x-%04x", dcbe->hasht, dcbe->hashm);
-	drm_encoder_helper_add(encoder, &nv50_dac_help);
-
-	drm_mode_connector_attach_encoder(connector, encoder);
-	return 0;
-}
-
-/******************************************************************************
- * Audio
- *****************************************************************************/
-static void
-nv50_audio_disable(struct drm_encoder *encoder, struct nouveau_crtc *nv_crtc)
-{
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nv50_disp *disp = nv50_disp(encoder->dev);
-	struct {
-		struct nv50_disp_mthd_v1 base;
-		struct nv50_disp_sor_hda_eld_v0 eld;
-	} args = {
-		.base.version = 1,
-		.base.method  = NV50_DISP_MTHD_V1_SOR_HDA_ELD,
-		.base.hasht   = nv_encoder->dcb->hasht,
-		.base.hashm   = (0xf0ff & nv_encoder->dcb->hashm) |
-				(0x0100 << nv_crtc->index),
-	};
-
-	nvif_mthd(&disp->disp->object, 0, &args, sizeof(args));
-}
-
-static void
-nv50_audio_enable(struct drm_encoder *encoder, struct drm_display_mode *mode)
-{
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
-	struct nouveau_connector *nv_connector;
-	struct nv50_disp *disp = nv50_disp(encoder->dev);
-	struct __packed {
-		struct {
-			struct nv50_disp_mthd_v1 mthd;
-			struct nv50_disp_sor_hda_eld_v0 eld;
-		} base;
-		u8 data[sizeof(nv_connector->base.eld)];
-	} args = {
-		.base.mthd.version = 1,
-		.base.mthd.method  = NV50_DISP_MTHD_V1_SOR_HDA_ELD,
-		.base.mthd.hasht   = nv_encoder->dcb->hasht,
-		.base.mthd.hashm   = (0xf0ff & nv_encoder->dcb->hashm) |
-				     (0x0100 << nv_crtc->index),
-	};
-
-	nv_connector = nouveau_encoder_connector_get(nv_encoder);
-	if (!drm_detect_monitor_audio(nv_connector->edid))
-		return;
-
-	memcpy(args.data, nv_connector->base.eld, sizeof(args.data));
-
-	nvif_mthd(&disp->disp->object, 0, &args,
-		  sizeof(args.base) + drm_eld_size(args.data));
-}
-
-/******************************************************************************
- * HDMI
- *****************************************************************************/
-static void
-nv50_hdmi_disable(struct drm_encoder *encoder, struct nouveau_crtc *nv_crtc)
-{
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nv50_disp *disp = nv50_disp(encoder->dev);
-	struct {
-		struct nv50_disp_mthd_v1 base;
-		struct nv50_disp_sor_hdmi_pwr_v0 pwr;
-	} args = {
-		.base.version = 1,
-		.base.method = NV50_DISP_MTHD_V1_SOR_HDMI_PWR,
-		.base.hasht  = nv_encoder->dcb->hasht,
-		.base.hashm  = (0xf0ff & nv_encoder->dcb->hashm) |
-			       (0x0100 << nv_crtc->index),
-	};
-
-	nvif_mthd(&disp->disp->object, 0, &args, sizeof(args));
-}
-
-static void
-nv50_hdmi_enable(struct drm_encoder *encoder, struct drm_display_mode *mode)
-{
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
-	struct nv50_disp *disp = nv50_disp(encoder->dev);
-	struct {
-		struct nv50_disp_mthd_v1 base;
-		struct nv50_disp_sor_hdmi_pwr_v0 pwr;
-		u8 infoframes[2 * 17]; /* two frames, up to 17 bytes each */
-	} args = {
-		.base.version = 1,
-		.base.method = NV50_DISP_MTHD_V1_SOR_HDMI_PWR,
-		.base.hasht  = nv_encoder->dcb->hasht,
-		.base.hashm  = (0xf0ff & nv_encoder->dcb->hashm) |
-			       (0x0100 << nv_crtc->index),
-		.pwr.state = 1,
-		.pwr.rekey = 56, /* binary driver, and tegra, constant */
-	};
-	struct nouveau_connector *nv_connector;
-	u32 max_ac_packet;
-	union hdmi_infoframe avi_frame;
-	union hdmi_infoframe vendor_frame;
-	int ret;
-	int size;
-
-	nv_connector = nouveau_encoder_connector_get(nv_encoder);
-	if (!drm_detect_hdmi_monitor(nv_connector->edid))
-		return;
-
-	ret = drm_hdmi_avi_infoframe_from_display_mode(&avi_frame.avi, mode,
-						       false);
-	if (!ret) {
-		/* We have an AVI InfoFrame, populate it to the display */
-		args.pwr.avi_infoframe_length
-			= hdmi_infoframe_pack(&avi_frame, args.infoframes, 17);
-	}
-
-	ret = drm_hdmi_vendor_infoframe_from_display_mode(&vendor_frame.vendor.hdmi,
-							  &nv_connector->base, mode);
-	if (!ret) {
-		/* We have a Vendor InfoFrame, populate it to the display */
-		args.pwr.vendor_infoframe_length
-			= hdmi_infoframe_pack(&vendor_frame,
-					      args.infoframes
-					      + args.pwr.avi_infoframe_length,
-					      17);
-	}
-
-	max_ac_packet  = mode->htotal - mode->hdisplay;
-	max_ac_packet -= args.pwr.rekey;
-	max_ac_packet -= 18; /* constant from tegra */
-	args.pwr.max_ac_packet = max_ac_packet / 32;
-
-	size = sizeof(args.base)
-		+ sizeof(args.pwr)
-		+ args.pwr.avi_infoframe_length
-		+ args.pwr.vendor_infoframe_length;
-	nvif_mthd(&disp->disp->object, 0, &args, size);
-	nv50_audio_enable(encoder, mode);
-}
-
-/******************************************************************************
- * MST
- *****************************************************************************/
-#define nv50_mstm(p) container_of((p), struct nv50_mstm, mgr)
-#define nv50_mstc(p) container_of((p), struct nv50_mstc, connector)
-#define nv50_msto(p) container_of((p), struct nv50_msto, encoder)
-
-struct nv50_mstm {
-	struct nouveau_encoder *outp;
-
-	struct drm_dp_mst_topology_mgr mgr;
-	struct nv50_msto *msto[4];
-
-	bool modified;
-	bool disabled;
-	int links;
-};
-
-struct nv50_mstc {
-	struct nv50_mstm *mstm;
-	struct drm_dp_mst_port *port;
-	struct drm_connector connector;
-
-	struct drm_display_mode *native;
-	struct edid *edid;
-
-	int pbn;
-};
-
-struct nv50_msto {
-	struct drm_encoder encoder;
-
-	struct nv50_head *head;
-	struct nv50_mstc *mstc;
-	bool disabled;
-};
-
-static struct drm_dp_payload *
-nv50_msto_payload(struct nv50_msto *msto)
-{
-	struct nouveau_drm *drm = nouveau_drm(msto->encoder.dev);
-	struct nv50_mstc *mstc = msto->mstc;
-	struct nv50_mstm *mstm = mstc->mstm;
-	int vcpi = mstc->port->vcpi.vcpi, i;
-
-	NV_ATOMIC(drm, "%s: vcpi %d\n", msto->encoder.name, vcpi);
-	for (i = 0; i < mstm->mgr.max_payloads; i++) {
-		struct drm_dp_payload *payload = &mstm->mgr.payloads[i];
-		NV_ATOMIC(drm, "%s: %d: vcpi %d start 0x%02x slots 0x%02x\n",
-			  mstm->outp->base.base.name, i, payload->vcpi,
-			  payload->start_slot, payload->num_slots);
-	}
-
-	for (i = 0; i < mstm->mgr.max_payloads; i++) {
-		struct drm_dp_payload *payload = &mstm->mgr.payloads[i];
-		if (payload->vcpi == vcpi)
-			return payload;
-	}
-
-	return NULL;
-}
-
-static void
-nv50_msto_cleanup(struct nv50_msto *msto)
-{
-	struct nouveau_drm *drm = nouveau_drm(msto->encoder.dev);
-	struct nv50_mstc *mstc = msto->mstc;
-	struct nv50_mstm *mstm = mstc->mstm;
-
-	NV_ATOMIC(drm, "%s: msto cleanup\n", msto->encoder.name);
-	if (mstc->port && mstc->port->vcpi.vcpi > 0 && !nv50_msto_payload(msto))
-		drm_dp_mst_deallocate_vcpi(&mstm->mgr, mstc->port);
-	if (msto->disabled) {
-		msto->mstc = NULL;
-		msto->head = NULL;
-		msto->disabled = false;
-	}
-}
-
-static void
-nv50_msto_prepare(struct nv50_msto *msto)
-{
-	struct nouveau_drm *drm = nouveau_drm(msto->encoder.dev);
-	struct nv50_mstc *mstc = msto->mstc;
-	struct nv50_mstm *mstm = mstc->mstm;
-	struct {
-		struct nv50_disp_mthd_v1 base;
-		struct nv50_disp_sor_dp_mst_vcpi_v0 vcpi;
-	} args = {
-		.base.version = 1,
-		.base.method = NV50_DISP_MTHD_V1_SOR_DP_MST_VCPI,
-		.base.hasht  = mstm->outp->dcb->hasht,
-		.base.hashm  = (0xf0ff & mstm->outp->dcb->hashm) |
-			       (0x0100 << msto->head->base.index),
-	};
-
-	NV_ATOMIC(drm, "%s: msto prepare\n", msto->encoder.name);
-	if (mstc->port && mstc->port->vcpi.vcpi > 0) {
-		struct drm_dp_payload *payload = nv50_msto_payload(msto);
-		if (payload) {
-			args.vcpi.start_slot = payload->start_slot;
-			args.vcpi.num_slots = payload->num_slots;
-			args.vcpi.pbn = mstc->port->vcpi.pbn;
-			args.vcpi.aligned_pbn = mstc->port->vcpi.aligned_pbn;
-		}
-	}
-
-	NV_ATOMIC(drm, "%s: %s: %02x %02x %04x %04x\n",
-		  msto->encoder.name, msto->head->base.base.name,
-		  args.vcpi.start_slot, args.vcpi.num_slots,
-		  args.vcpi.pbn, args.vcpi.aligned_pbn);
-	nvif_mthd(&drm->display->disp.object, 0, &args, sizeof(args));
-}
-
-static int
-nv50_msto_atomic_check(struct drm_encoder *encoder,
-		       struct drm_crtc_state *crtc_state,
-		       struct drm_connector_state *conn_state)
-{
-	struct nv50_mstc *mstc = nv50_mstc(conn_state->connector);
-	struct nv50_mstm *mstm = mstc->mstm;
-	int bpp = conn_state->connector->display_info.bpc * 3;
-	int slots;
-
-	mstc->pbn = drm_dp_calc_pbn_mode(crtc_state->adjusted_mode.clock, bpp);
-
-	slots = drm_dp_find_vcpi_slots(&mstm->mgr, mstc->pbn);
-	if (slots < 0)
-		return slots;
-
-	return nv50_outp_atomic_check_view(encoder, crtc_state, conn_state,
-					   mstc->native);
-}
-
-static void
-nv50_msto_enable(struct drm_encoder *encoder)
-{
-	struct nv50_head *head = nv50_head(encoder->crtc);
-	struct nv50_msto *msto = nv50_msto(encoder);
-	struct nv50_mstc *mstc = NULL;
-	struct nv50_mstm *mstm = NULL;
-	struct drm_connector *connector;
-	struct drm_connector_list_iter conn_iter;
-	u8 proto, depth;
-	int slots;
-	bool r;
-
-	drm_connector_list_iter_begin(encoder->dev, &conn_iter);
-	drm_for_each_connector_iter(connector, &conn_iter) {
-		if (connector->state->best_encoder == &msto->encoder) {
-			mstc = nv50_mstc(connector);
-			mstm = mstc->mstm;
-			break;
-		}
-	}
-	drm_connector_list_iter_end(&conn_iter);
-
-	if (WARN_ON(!mstc))
-		return;
-
-	slots = drm_dp_find_vcpi_slots(&mstm->mgr, mstc->pbn);
-	r = drm_dp_mst_allocate_vcpi(&mstm->mgr, mstc->port, mstc->pbn, slots);
-	WARN_ON(!r);
-
-	if (!mstm->links++)
-		nv50_outp_acquire(mstm->outp);
-
-	if (mstm->outp->link & 1)
-		proto = 0x8;
-	else
-		proto = 0x9;
-
-	switch (mstc->connector.display_info.bpc) {
-	case  6: depth = 0x2; break;
-	case  8: depth = 0x5; break;
-	case 10:
-	default: depth = 0x6; break;
-	}
-
-	mstm->outp->update(mstm->outp, head->base.index,
-			   &head->base.base.state->adjusted_mode, proto, depth);
-
-	msto->head = head;
-	msto->mstc = mstc;
-	mstm->modified = true;
-}
-
-static void
-nv50_msto_disable(struct drm_encoder *encoder)
-{
-	struct nv50_msto *msto = nv50_msto(encoder);
-	struct nv50_mstc *mstc = msto->mstc;
-	struct nv50_mstm *mstm = mstc->mstm;
-
-	if (mstc->port)
-		drm_dp_mst_reset_vcpi_slots(&mstm->mgr, mstc->port);
-
-	mstm->outp->update(mstm->outp, msto->head->base.index, NULL, 0, 0);
-	mstm->modified = true;
-	if (!--mstm->links)
-		mstm->disabled = true;
-	msto->disabled = true;
-}
-
-static const struct drm_encoder_helper_funcs
-nv50_msto_help = {
-	.disable = nv50_msto_disable,
-	.enable = nv50_msto_enable,
-	.atomic_check = nv50_msto_atomic_check,
-};
-
-static void
-nv50_msto_destroy(struct drm_encoder *encoder)
-{
-	struct nv50_msto *msto = nv50_msto(encoder);
-	drm_encoder_cleanup(&msto->encoder);
-	kfree(msto);
-}
-
-static const struct drm_encoder_funcs
-nv50_msto = {
-	.destroy = nv50_msto_destroy,
-};
-
-static int
-nv50_msto_new(struct drm_device *dev, u32 heads, const char *name, int id,
-	      struct nv50_msto **pmsto)
-{
-	struct nv50_msto *msto;
-	int ret;
-
-	if (!(msto = *pmsto = kzalloc(sizeof(*msto), GFP_KERNEL)))
-		return -ENOMEM;
-
-	ret = drm_encoder_init(dev, &msto->encoder, &nv50_msto,
-			       DRM_MODE_ENCODER_DPMST, "%s-mst-%d", name, id);
-	if (ret) {
-		kfree(*pmsto);
-		*pmsto = NULL;
-		return ret;
-	}
-
-	drm_encoder_helper_add(&msto->encoder, &nv50_msto_help);
-	msto->encoder.possible_crtcs = heads;
-	return 0;
-}
-
-static struct drm_encoder *
-nv50_mstc_atomic_best_encoder(struct drm_connector *connector,
-			      struct drm_connector_state *connector_state)
-{
-	struct nv50_head *head = nv50_head(connector_state->crtc);
-	struct nv50_mstc *mstc = nv50_mstc(connector);
-	if (mstc->port) {
-		struct nv50_mstm *mstm = mstc->mstm;
-		return &mstm->msto[head->base.index]->encoder;
-	}
-	return NULL;
-}
-
-static struct drm_encoder *
-nv50_mstc_best_encoder(struct drm_connector *connector)
-{
-	struct nv50_mstc *mstc = nv50_mstc(connector);
-	if (mstc->port) {
-		struct nv50_mstm *mstm = mstc->mstm;
-		return &mstm->msto[0]->encoder;
-	}
-	return NULL;
-}
-
-static enum drm_mode_status
-nv50_mstc_mode_valid(struct drm_connector *connector,
-		     struct drm_display_mode *mode)
-{
-	return MODE_OK;
-}
-
-static int
-nv50_mstc_get_modes(struct drm_connector *connector)
-{
-	struct nv50_mstc *mstc = nv50_mstc(connector);
-	int ret = 0;
-
-	mstc->edid = drm_dp_mst_get_edid(&mstc->connector, mstc->port->mgr, mstc->port);
-	drm_mode_connector_update_edid_property(&mstc->connector, mstc->edid);
-	if (mstc->edid)
-		ret = drm_add_edid_modes(&mstc->connector, mstc->edid);
-
-	if (!mstc->connector.display_info.bpc)
-		mstc->connector.display_info.bpc = 8;
-
-	if (mstc->native)
-		drm_mode_destroy(mstc->connector.dev, mstc->native);
-	mstc->native = nouveau_conn_native_mode(&mstc->connector);
-	return ret;
-}
-
-static const struct drm_connector_helper_funcs
-nv50_mstc_help = {
-	.get_modes = nv50_mstc_get_modes,
-	.mode_valid = nv50_mstc_mode_valid,
-	.best_encoder = nv50_mstc_best_encoder,
-	.atomic_best_encoder = nv50_mstc_atomic_best_encoder,
-};
-
-static enum drm_connector_status
-nv50_mstc_detect(struct drm_connector *connector, bool force)
-{
-	struct nv50_mstc *mstc = nv50_mstc(connector);
-	if (!mstc->port)
-		return connector_status_disconnected;
-	return drm_dp_mst_detect_port(connector, mstc->port->mgr, mstc->port);
-}
-
-static void
-nv50_mstc_destroy(struct drm_connector *connector)
-{
-	struct nv50_mstc *mstc = nv50_mstc(connector);
-	drm_connector_cleanup(&mstc->connector);
-	kfree(mstc);
-}
-
-static const struct drm_connector_funcs
-nv50_mstc = {
-	.reset = nouveau_conn_reset,
-	.detect = nv50_mstc_detect,
-	.fill_modes = drm_helper_probe_single_connector_modes,
-	.destroy = nv50_mstc_destroy,
-	.atomic_duplicate_state = nouveau_conn_atomic_duplicate_state,
-	.atomic_destroy_state = nouveau_conn_atomic_destroy_state,
-	.atomic_set_property = nouveau_conn_atomic_set_property,
-	.atomic_get_property = nouveau_conn_atomic_get_property,
-};
-
-static int
-nv50_mstc_new(struct nv50_mstm *mstm, struct drm_dp_mst_port *port,
-	      const char *path, struct nv50_mstc **pmstc)
-{
-	struct drm_device *dev = mstm->outp->base.base.dev;
-	struct nv50_mstc *mstc;
-	int ret, i;
-
-	if (!(mstc = *pmstc = kzalloc(sizeof(*mstc), GFP_KERNEL)))
-		return -ENOMEM;
-	mstc->mstm = mstm;
-	mstc->port = port;
-
-	ret = drm_connector_init(dev, &mstc->connector, &nv50_mstc,
-				 DRM_MODE_CONNECTOR_DisplayPort);
-	if (ret) {
-		kfree(*pmstc);
-		*pmstc = NULL;
-		return ret;
-	}
-
-	drm_connector_helper_add(&mstc->connector, &nv50_mstc_help);
-
-	mstc->connector.funcs->reset(&mstc->connector);
-	nouveau_conn_attach_properties(&mstc->connector);
-
-	for (i = 0; i < ARRAY_SIZE(mstm->msto) && mstm->msto[i]; i++)
-		drm_mode_connector_attach_encoder(&mstc->connector, &mstm->msto[i]->encoder);
-
-	drm_object_attach_property(&mstc->connector.base, dev->mode_config.path_property, 0);
-	drm_object_attach_property(&mstc->connector.base, dev->mode_config.tile_property, 0);
-	drm_mode_connector_set_path_property(&mstc->connector, path);
-	return 0;
-}
-
-static void
-nv50_mstm_cleanup(struct nv50_mstm *mstm)
-{
-	struct nouveau_drm *drm = nouveau_drm(mstm->outp->base.base.dev);
-	struct drm_encoder *encoder;
-	int ret;
-
-	NV_ATOMIC(drm, "%s: mstm cleanup\n", mstm->outp->base.base.name);
-	ret = drm_dp_check_act_status(&mstm->mgr);
-
-	ret = drm_dp_update_payload_part2(&mstm->mgr);
-
-	drm_for_each_encoder(encoder, mstm->outp->base.base.dev) {
-		if (encoder->encoder_type == DRM_MODE_ENCODER_DPMST) {
-			struct nv50_msto *msto = nv50_msto(encoder);
-			struct nv50_mstc *mstc = msto->mstc;
-			if (mstc && mstc->mstm == mstm)
-				nv50_msto_cleanup(msto);
-		}
-	}
-
-	mstm->modified = false;
-}
-
-static void
-nv50_mstm_prepare(struct nv50_mstm *mstm)
-{
-	struct nouveau_drm *drm = nouveau_drm(mstm->outp->base.base.dev);
-	struct drm_encoder *encoder;
-	int ret;
-
-	NV_ATOMIC(drm, "%s: mstm prepare\n", mstm->outp->base.base.name);
-	ret = drm_dp_update_payload_part1(&mstm->mgr);
-
-	drm_for_each_encoder(encoder, mstm->outp->base.base.dev) {
-		if (encoder->encoder_type == DRM_MODE_ENCODER_DPMST) {
-			struct nv50_msto *msto = nv50_msto(encoder);
-			struct nv50_mstc *mstc = msto->mstc;
-			if (mstc && mstc->mstm == mstm)
-				nv50_msto_prepare(msto);
-		}
-	}
-
-	if (mstm->disabled) {
-		if (!mstm->links)
-			nv50_outp_release(mstm->outp);
-		mstm->disabled = false;
-	}
-}
-
-static void
-nv50_mstm_hotplug(struct drm_dp_mst_topology_mgr *mgr)
-{
-	struct nv50_mstm *mstm = nv50_mstm(mgr);
-	drm_kms_helper_hotplug_event(mstm->outp->base.base.dev);
-}
-
-static void
-nv50_mstm_destroy_connector(struct drm_dp_mst_topology_mgr *mgr,
-			    struct drm_connector *connector)
-{
-	struct nouveau_drm *drm = nouveau_drm(connector->dev);
-	struct nv50_mstc *mstc = nv50_mstc(connector);
-
-	drm_connector_unregister(&mstc->connector);
-
-	drm_fb_helper_remove_one_connector(&drm->fbcon->helper, &mstc->connector);
-
-	drm_modeset_lock(&drm->dev->mode_config.connection_mutex, NULL);
-	mstc->port = NULL;
-	drm_modeset_unlock(&drm->dev->mode_config.connection_mutex);
-
-	drm_connector_unreference(&mstc->connector);
-}
-
-static void
-nv50_mstm_register_connector(struct drm_connector *connector)
-{
-	struct nouveau_drm *drm = nouveau_drm(connector->dev);
-
-	drm_fb_helper_add_one_connector(&drm->fbcon->helper, connector);
-
-	drm_connector_register(connector);
-}
-
-static struct drm_connector *
-nv50_mstm_add_connector(struct drm_dp_mst_topology_mgr *mgr,
-			struct drm_dp_mst_port *port, const char *path)
-{
-	struct nv50_mstm *mstm = nv50_mstm(mgr);
-	struct nv50_mstc *mstc;
-	int ret;
-
-	ret = nv50_mstc_new(mstm, port, path, &mstc);
-	if (ret) {
-		if (mstc)
-			mstc->connector.funcs->destroy(&mstc->connector);
-		return NULL;
-	}
-
-	return &mstc->connector;
-}
-
-static const struct drm_dp_mst_topology_cbs
-nv50_mstm = {
-	.add_connector = nv50_mstm_add_connector,
-	.register_connector = nv50_mstm_register_connector,
-	.destroy_connector = nv50_mstm_destroy_connector,
-	.hotplug = nv50_mstm_hotplug,
-};
-
-void
-nv50_mstm_service(struct nv50_mstm *mstm)
-{
-	struct drm_dp_aux *aux = mstm ? mstm->mgr.aux : NULL;
-	bool handled = true;
-	int ret;
-	u8 esi[8] = {};
-
-	if (!aux)
-		return;
-
-	while (handled) {
-		ret = drm_dp_dpcd_read(aux, DP_SINK_COUNT_ESI, esi, 8);
-		if (ret != 8) {
-			drm_dp_mst_topology_mgr_set_mst(&mstm->mgr, false);
-			return;
-		}
-
-		drm_dp_mst_hpd_irq(&mstm->mgr, esi, &handled);
-		if (!handled)
-			break;
-
-		drm_dp_dpcd_write(aux, DP_SINK_COUNT_ESI + 1, &esi[1], 3);
-	}
-}
-
-void
-nv50_mstm_remove(struct nv50_mstm *mstm)
-{
-	if (mstm)
-		drm_dp_mst_topology_mgr_set_mst(&mstm->mgr, false);
-}
-
-static int
-nv50_mstm_enable(struct nv50_mstm *mstm, u8 dpcd, int state)
-{
-	struct nouveau_encoder *outp = mstm->outp;
-	struct {
-		struct nv50_disp_mthd_v1 base;
-		struct nv50_disp_sor_dp_mst_link_v0 mst;
-	} args = {
-		.base.version = 1,
-		.base.method = NV50_DISP_MTHD_V1_SOR_DP_MST_LINK,
-		.base.hasht = outp->dcb->hasht,
-		.base.hashm = outp->dcb->hashm,
-		.mst.state = state,
-	};
-	struct nouveau_drm *drm = nouveau_drm(outp->base.base.dev);
-	struct nvif_object *disp = &drm->display->disp.object;
-	int ret;
-
-	if (dpcd >= 0x12) {
-		ret = drm_dp_dpcd_readb(mstm->mgr.aux, DP_MSTM_CTRL, &dpcd);
-		if (ret < 0)
-			return ret;
-
-		dpcd &= ~DP_MST_EN;
-		if (state)
-			dpcd |= DP_MST_EN;
-
-		ret = drm_dp_dpcd_writeb(mstm->mgr.aux, DP_MSTM_CTRL, dpcd);
-		if (ret < 0)
-			return ret;
-	}
-
-	return nvif_mthd(disp, 0, &args, sizeof(args));
-}
-
-int
-nv50_mstm_detect(struct nv50_mstm *mstm, u8 dpcd[8], int allow)
-{
-	int ret, state = 0;
-
-	if (!mstm)
-		return 0;
-
-	if (dpcd[0] >= 0x12) {
-		ret = drm_dp_dpcd_readb(mstm->mgr.aux, DP_MSTM_CAP, &dpcd[1]);
-		if (ret < 0)
-			return ret;
-
-		if (!(dpcd[1] & DP_MST_CAP))
-			dpcd[0] = 0x11;
-		else
-			state = allow;
-	}
-
-	ret = nv50_mstm_enable(mstm, dpcd[0], state);
-	if (ret)
-		return ret;
-
-	ret = drm_dp_mst_topology_mgr_set_mst(&mstm->mgr, state);
-	if (ret)
-		return nv50_mstm_enable(mstm, dpcd[0], 0);
-
-	return mstm->mgr.mst_state;
-}
-
-static void
-nv50_mstm_fini(struct nv50_mstm *mstm)
-{
-	if (mstm && mstm->mgr.mst_state)
-		drm_dp_mst_topology_mgr_suspend(&mstm->mgr);
-}
-
-static void
-nv50_mstm_init(struct nv50_mstm *mstm)
-{
-	if (mstm && mstm->mgr.mst_state)
-		drm_dp_mst_topology_mgr_resume(&mstm->mgr);
-}
-
-static void
-nv50_mstm_del(struct nv50_mstm **pmstm)
-{
-	struct nv50_mstm *mstm = *pmstm;
-	if (mstm) {
-		kfree(*pmstm);
-		*pmstm = NULL;
-	}
-}
-
-static int
-nv50_mstm_new(struct nouveau_encoder *outp, struct drm_dp_aux *aux, int aux_max,
-	      int conn_base_id, struct nv50_mstm **pmstm)
-{
-	const int max_payloads = hweight8(outp->dcb->heads);
-	struct drm_device *dev = outp->base.base.dev;
-	struct nv50_mstm *mstm;
-	int ret, i;
-	u8 dpcd;
-
-	/* This is a workaround for some monitors not functioning
-	 * correctly in MST mode on initial module load.  I think
-	 * some bad interaction with the VBIOS may be responsible.
-	 *
-	 * A good ol' off and on again seems to work here ;)
-	 */
-	ret = drm_dp_dpcd_readb(aux, DP_DPCD_REV, &dpcd);
-	if (ret >= 0 && dpcd >= 0x12)
-		drm_dp_dpcd_writeb(aux, DP_MSTM_CTRL, 0);
-
-	if (!(mstm = *pmstm = kzalloc(sizeof(*mstm), GFP_KERNEL)))
-		return -ENOMEM;
-	mstm->outp = outp;
-	mstm->mgr.cbs = &nv50_mstm;
-
-	ret = drm_dp_mst_topology_mgr_init(&mstm->mgr, dev, aux, aux_max,
-					   max_payloads, conn_base_id);
-	if (ret)
-		return ret;
-
-	for (i = 0; i < max_payloads; i++) {
-		ret = nv50_msto_new(dev, outp->dcb->heads, outp->base.base.name,
-				    i, &mstm->msto[i]);
-		if (ret)
-			return ret;
-	}
-
-	return 0;
-}
-
-/******************************************************************************
- * SOR
- *****************************************************************************/
-static void
-nv50_sor_update(struct nouveau_encoder *nv_encoder, u8 head,
-		struct drm_display_mode *mode, u8 proto, u8 depth)
-{
-	struct nv50_dmac *core = &nv50_mast(nv_encoder->base.base.dev)->base;
-	u32 *push;
-
-	if (!mode) {
-		nv_encoder->ctrl &= ~BIT(head);
-		if (!(nv_encoder->ctrl & 0x0000000f))
-			nv_encoder->ctrl = 0;
-	} else {
-		nv_encoder->ctrl |= proto << 8;
-		nv_encoder->ctrl |= BIT(head);
-	}
-
-	if ((push = evo_wait(core, 6))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			if (mode) {
-				if (mode->flags & DRM_MODE_FLAG_NHSYNC)
-					nv_encoder->ctrl |= 0x00001000;
-				if (mode->flags & DRM_MODE_FLAG_NVSYNC)
-					nv_encoder->ctrl |= 0x00002000;
-				nv_encoder->ctrl |= depth << 16;
-			}
-			evo_mthd(push, 0x0600 + (nv_encoder->or * 0x40), 1);
-		} else {
-			if (mode) {
-				u32 magic = 0x31ec6000 | (head << 25);
-				u32 syncs = 0x00000001;
-				if (mode->flags & DRM_MODE_FLAG_NHSYNC)
-					syncs |= 0x00000008;
-				if (mode->flags & DRM_MODE_FLAG_NVSYNC)
-					syncs |= 0x00000010;
-				if (mode->flags & DRM_MODE_FLAG_INTERLACE)
-					magic |= 0x00000001;
-
-				evo_mthd(push, 0x0404 + (head * 0x300), 2);
-				evo_data(push, syncs | (depth << 6));
-				evo_data(push, magic);
-			}
-			evo_mthd(push, 0x0200 + (nv_encoder->or * 0x20), 1);
-		}
-		evo_data(push, nv_encoder->ctrl);
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_sor_disable(struct drm_encoder *encoder)
-{
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nouveau_crtc *nv_crtc = nouveau_crtc(nv_encoder->crtc);
-
-	nv_encoder->crtc = NULL;
-
-	if (nv_crtc) {
-		struct nvkm_i2c_aux *aux = nv_encoder->aux;
-		u8 pwr;
-
-		if (aux) {
-			int ret = nvkm_rdaux(aux, DP_SET_POWER, &pwr, 1);
-			if (ret == 0) {
-				pwr &= ~DP_SET_POWER_MASK;
-				pwr |=  DP_SET_POWER_D3;
-				nvkm_wraux(aux, DP_SET_POWER, &pwr, 1);
-			}
-		}
-
-		nv_encoder->update(nv_encoder, nv_crtc->index, NULL, 0, 0);
-		nv50_audio_disable(encoder, nv_crtc);
-		nv50_hdmi_disable(&nv_encoder->base.base, nv_crtc);
-		nv50_outp_release(nv_encoder);
-	}
-}
-
-static void
-nv50_sor_enable(struct drm_encoder *encoder)
-{
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
-	struct drm_display_mode *mode = &nv_crtc->base.state->adjusted_mode;
-	struct {
-		struct nv50_disp_mthd_v1 base;
-		struct nv50_disp_sor_lvds_script_v0 lvds;
-	} lvds = {
-		.base.version = 1,
-		.base.method  = NV50_DISP_MTHD_V1_SOR_LVDS_SCRIPT,
-		.base.hasht   = nv_encoder->dcb->hasht,
-		.base.hashm   = nv_encoder->dcb->hashm,
-	};
-	struct nv50_disp *disp = nv50_disp(encoder->dev);
-	struct drm_device *dev = encoder->dev;
-	struct nouveau_drm *drm = nouveau_drm(dev);
-	struct nouveau_connector *nv_connector;
-	struct nvbios *bios = &drm->vbios;
-	u8 proto = 0xf;
-	u8 depth = 0x0;
-
-	nv_connector = nouveau_encoder_connector_get(nv_encoder);
-	nv_encoder->crtc = encoder->crtc;
-	nv50_outp_acquire(nv_encoder);
-
-	switch (nv_encoder->dcb->type) {
-	case DCB_OUTPUT_TMDS:
-		if (nv_encoder->link & 1) {
-			proto = 0x1;
-			/* Only enable dual-link if:
-			 *  - Need to (i.e. rate > 165MHz)
-			 *  - DCB says we can
-			 *  - Not an HDMI monitor, since there's no dual-link
-			 *    on HDMI.
-			 */
-			if (mode->clock >= 165000 &&
-			    nv_encoder->dcb->duallink_possible &&
-			    !drm_detect_hdmi_monitor(nv_connector->edid))
-				proto |= 0x4;
-		} else {
-			proto = 0x2;
-		}
-
-		nv50_hdmi_enable(&nv_encoder->base.base, mode);
-		break;
-	case DCB_OUTPUT_LVDS:
-		proto = 0x0;
-
-		if (bios->fp_no_ddc) {
-			if (bios->fp.dual_link)
-				lvds.lvds.script |= 0x0100;
-			if (bios->fp.if_is_24bit)
-				lvds.lvds.script |= 0x0200;
-		} else {
-			if (nv_connector->type == DCB_CONNECTOR_LVDS_SPWG) {
-				if (((u8 *)nv_connector->edid)[121] == 2)
-					lvds.lvds.script |= 0x0100;
-			} else
-			if (mode->clock >= bios->fp.duallink_transition_clk) {
-				lvds.lvds.script |= 0x0100;
-			}
-
-			if (lvds.lvds.script & 0x0100) {
-				if (bios->fp.strapless_is_24bit & 2)
-					lvds.lvds.script |= 0x0200;
-			} else {
-				if (bios->fp.strapless_is_24bit & 1)
-					lvds.lvds.script |= 0x0200;
-			}
-
-			if (nv_connector->base.display_info.bpc == 8)
-				lvds.lvds.script |= 0x0200;
-		}
-
-		nvif_mthd(&disp->disp->object, 0, &lvds, sizeof(lvds));
-		break;
-	case DCB_OUTPUT_DP:
-		if (nv_connector->base.display_info.bpc == 6)
-			depth = 0x2;
-		else
-		if (nv_connector->base.display_info.bpc == 8)
-			depth = 0x5;
-		else
-			depth = 0x6;
-
-		if (nv_encoder->link & 1)
-			proto = 0x8;
-		else
-			proto = 0x9;
-
-		nv50_audio_enable(encoder, mode);
-		break;
-	default:
-		BUG();
-		break;
-	}
-
-	nv_encoder->update(nv_encoder, nv_crtc->index, mode, proto, depth);
-}
-
-static const struct drm_encoder_helper_funcs
-nv50_sor_help = {
-	.atomic_check = nv50_outp_atomic_check,
-	.enable = nv50_sor_enable,
-	.disable = nv50_sor_disable,
-};
-
-static void
-nv50_sor_destroy(struct drm_encoder *encoder)
-{
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	nv50_mstm_del(&nv_encoder->dp.mstm);
-	drm_encoder_cleanup(encoder);
-	kfree(encoder);
-}
-
-static const struct drm_encoder_funcs
-nv50_sor_func = {
-	.destroy = nv50_sor_destroy,
-};
-
-static int
-nv50_sor_create(struct drm_connector *connector, struct dcb_output *dcbe)
-{
-	struct nouveau_connector *nv_connector = nouveau_connector(connector);
-	struct nouveau_drm *drm = nouveau_drm(connector->dev);
-	struct nvkm_i2c *i2c = nvxx_i2c(&drm->client.device);
-	struct nouveau_encoder *nv_encoder;
-	struct drm_encoder *encoder;
-	int type, ret;
-
-	switch (dcbe->type) {
-	case DCB_OUTPUT_LVDS: type = DRM_MODE_ENCODER_LVDS; break;
-	case DCB_OUTPUT_TMDS:
-	case DCB_OUTPUT_DP:
-	default:
-		type = DRM_MODE_ENCODER_TMDS;
-		break;
-	}
-
-	nv_encoder = kzalloc(sizeof(*nv_encoder), GFP_KERNEL);
-	if (!nv_encoder)
-		return -ENOMEM;
-	nv_encoder->dcb = dcbe;
-	nv_encoder->update = nv50_sor_update;
-
-	encoder = to_drm_encoder(nv_encoder);
-	encoder->possible_crtcs = dcbe->heads;
-	encoder->possible_clones = 0;
-	drm_encoder_init(connector->dev, encoder, &nv50_sor_func, type,
-			 "sor-%04x-%04x", dcbe->hasht, dcbe->hashm);
-	drm_encoder_helper_add(encoder, &nv50_sor_help);
-
-	drm_mode_connector_attach_encoder(connector, encoder);
-
-	if (dcbe->type == DCB_OUTPUT_DP) {
-		struct nv50_disp *disp = nv50_disp(encoder->dev);
-		struct nvkm_i2c_aux *aux =
-			nvkm_i2c_aux_find(i2c, dcbe->i2c_index);
-		if (aux) {
-			if (disp->disp->object.oclass < GF110_DISP) {
-				/* HW has no support for address-only
-				 * transactions, so we're required to
-				 * use custom I2C-over-AUX code.
-				 */
-				nv_encoder->i2c = &aux->i2c;
-			} else {
-				nv_encoder->i2c = &nv_connector->aux.ddc;
-			}
-			nv_encoder->aux = aux;
-		}
-
-		/*TODO: Use DP Info Table to check for support. */
-		if (disp->disp->object.oclass >= GF110_DISP) {
-			ret = nv50_mstm_new(nv_encoder, &nv_connector->aux, 16,
-					    nv_connector->base.base.id,
-					    &nv_encoder->dp.mstm);
-			if (ret)
-				return ret;
-		}
-	} else {
-		struct nvkm_i2c_bus *bus =
-			nvkm_i2c_bus_find(i2c, dcbe->i2c_index);
-		if (bus)
-			nv_encoder->i2c = &bus->i2c;
-	}
-
-	return 0;
-}
-
-/******************************************************************************
- * PIOR
- *****************************************************************************/
-static int
-nv50_pior_atomic_check(struct drm_encoder *encoder,
-		       struct drm_crtc_state *crtc_state,
-		       struct drm_connector_state *conn_state)
-{
-	int ret = nv50_outp_atomic_check(encoder, crtc_state, conn_state);
-	if (ret)
-		return ret;
-	crtc_state->adjusted_mode.clock *= 2;
-	return 0;
-}
-
-static void
-nv50_pior_disable(struct drm_encoder *encoder)
-{
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nv50_mast *mast = nv50_mast(encoder->dev);
-	const int or = nv_encoder->or;
-	u32 *push;
-
-	if (nv_encoder->crtc) {
-		push = evo_wait(mast, 4);
-		if (push) {
-			if (nv50_vers(mast) < GF110_DISP_CORE_CHANNEL_DMA) {
-				evo_mthd(push, 0x0700 + (or * 0x040), 1);
-				evo_data(push, 0x00000000);
-			}
-			evo_kick(push, mast);
-		}
-	}
-
-	nv_encoder->crtc = NULL;
-	nv50_outp_release(nv_encoder);
-}
-
-static void
-nv50_pior_enable(struct drm_encoder *encoder)
-{
-	struct nv50_mast *mast = nv50_mast(encoder->dev);
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
-	struct nouveau_connector *nv_connector;
-	struct drm_display_mode *mode = &nv_crtc->base.state->adjusted_mode;
-	u8 owner = 1 << nv_crtc->index;
-	u8 proto, depth;
-	u32 *push;
-
-	nv50_outp_acquire(nv_encoder);
-
-	nv_connector = nouveau_encoder_connector_get(nv_encoder);
-	switch (nv_connector->base.display_info.bpc) {
-	case 10: depth = 0x6; break;
-	case  8: depth = 0x5; break;
-	case  6: depth = 0x2; break;
-	default: depth = 0x0; break;
-	}
-
-	switch (nv_encoder->dcb->type) {
-	case DCB_OUTPUT_TMDS:
-	case DCB_OUTPUT_DP:
-		proto = 0x0;
-		break;
-	default:
-		BUG();
-		break;
-	}
-
-	push = evo_wait(mast, 8);
-	if (push) {
-		if (nv50_vers(mast) < GF110_DISP_CORE_CHANNEL_DMA) {
-			u32 ctrl = (depth << 16) | (proto << 8) | owner;
-			if (mode->flags & DRM_MODE_FLAG_NHSYNC)
-				ctrl |= 0x00001000;
-			if (mode->flags & DRM_MODE_FLAG_NVSYNC)
-				ctrl |= 0x00002000;
-			evo_mthd(push, 0x0700 + (nv_encoder->or * 0x040), 1);
-			evo_data(push, ctrl);
-		}
-
-		evo_kick(push, mast);
-	}
-
-	nv_encoder->crtc = encoder->crtc;
-}
-
-static const struct drm_encoder_helper_funcs
-nv50_pior_help = {
-	.atomic_check = nv50_pior_atomic_check,
-	.enable = nv50_pior_enable,
-	.disable = nv50_pior_disable,
-};
-
-static void
-nv50_pior_destroy(struct drm_encoder *encoder)
-{
-	drm_encoder_cleanup(encoder);
-	kfree(encoder);
-}
-
-static const struct drm_encoder_funcs
-nv50_pior_func = {
-	.destroy = nv50_pior_destroy,
-};
-
-static int
-nv50_pior_create(struct drm_connector *connector, struct dcb_output *dcbe)
-{
-	struct nouveau_connector *nv_connector = nouveau_connector(connector);
-	struct nouveau_drm *drm = nouveau_drm(connector->dev);
-	struct nvkm_i2c *i2c = nvxx_i2c(&drm->client.device);
-	struct nvkm_i2c_bus *bus = NULL;
-	struct nvkm_i2c_aux *aux = NULL;
-	struct i2c_adapter *ddc;
-	struct nouveau_encoder *nv_encoder;
-	struct drm_encoder *encoder;
-	int type;
-
-	switch (dcbe->type) {
-	case DCB_OUTPUT_TMDS:
-		bus  = nvkm_i2c_bus_find(i2c, NVKM_I2C_BUS_EXT(dcbe->extdev));
-		ddc  = bus ? &bus->i2c : NULL;
-		type = DRM_MODE_ENCODER_TMDS;
-		break;
-	case DCB_OUTPUT_DP:
-		aux  = nvkm_i2c_aux_find(i2c, NVKM_I2C_AUX_EXT(dcbe->extdev));
-		ddc  = aux ? &nv_connector->aux.ddc : NULL;
-		type = DRM_MODE_ENCODER_TMDS;
-		break;
-	default:
-		return -ENODEV;
-	}
-
-	nv_encoder = kzalloc(sizeof(*nv_encoder), GFP_KERNEL);
-	if (!nv_encoder)
-		return -ENOMEM;
-	nv_encoder->dcb = dcbe;
-	nv_encoder->i2c = ddc;
-	nv_encoder->aux = aux;
-
-	encoder = to_drm_encoder(nv_encoder);
-	encoder->possible_crtcs = dcbe->heads;
-	encoder->possible_clones = 0;
-	drm_encoder_init(connector->dev, encoder, &nv50_pior_func, type,
-			 "pior-%04x-%04x", dcbe->hasht, dcbe->hashm);
-	drm_encoder_helper_add(encoder, &nv50_pior_help);
-
-	drm_mode_connector_attach_encoder(connector, encoder);
-	return 0;
-}
-
-/******************************************************************************
- * Atomic
- *****************************************************************************/
-
-static void
-nv50_disp_atomic_commit_core(struct nouveau_drm *drm, u32 interlock)
-{
-	struct nv50_disp *disp = nv50_disp(drm->dev);
-	struct nv50_dmac *core = &disp->mast.base;
-	struct nv50_mstm *mstm;
-	struct drm_encoder *encoder;
-	u32 *push;
-
-	NV_ATOMIC(drm, "commit core %08x\n", interlock);
-
-	drm_for_each_encoder(encoder, drm->dev) {
-		if (encoder->encoder_type != DRM_MODE_ENCODER_DPMST) {
-			mstm = nouveau_encoder(encoder)->dp.mstm;
-			if (mstm && mstm->modified)
-				nv50_mstm_prepare(mstm);
-		}
-	}
-
-	if ((push = evo_wait(core, 5))) {
-		evo_mthd(push, 0x0084, 1);
-		evo_data(push, 0x80000000);
-		evo_mthd(push, 0x0080, 2);
-		evo_data(push, interlock);
-		evo_data(push, 0x00000000);
-		nouveau_bo_wr32(disp->sync, 0, 0x00000000);
-		evo_kick(push, core);
-		if (nvif_msec(&drm->client.device, 2000ULL,
-			if (nouveau_bo_rd32(disp->sync, 0))
-				break;
-			usleep_range(1, 2);
-		) < 0)
-			NV_ERROR(drm, "EVO timeout\n");
-	}
-
-	drm_for_each_encoder(encoder, drm->dev) {
-		if (encoder->encoder_type != DRM_MODE_ENCODER_DPMST) {
-			mstm = nouveau_encoder(encoder)->dp.mstm;
-			if (mstm && mstm->modified)
-				nv50_mstm_cleanup(mstm);
-		}
-	}
-}
-
-static void
-nv50_disp_atomic_commit_tail(struct drm_atomic_state *state)
-{
-	struct drm_device *dev = state->dev;
-	struct drm_crtc_state *new_crtc_state, *old_crtc_state;
-	struct drm_crtc *crtc;
-	struct drm_plane_state *new_plane_state;
-	struct drm_plane *plane;
-	struct nouveau_drm *drm = nouveau_drm(dev);
-	struct nv50_disp *disp = nv50_disp(dev);
-	struct nv50_atom *atom = nv50_atom(state);
-	struct nv50_outp_atom *outp, *outt;
-	u32 interlock_core = 0;
-	u32 interlock_chan = 0;
-	int i;
-
-	NV_ATOMIC(drm, "commit %d %d\n", atom->lock_core, atom->flush_disable);
-	drm_atomic_helper_wait_for_fences(dev, state, false);
-	drm_atomic_helper_wait_for_dependencies(state);
-	drm_atomic_helper_update_legacy_modeset_state(dev, state);
-
-	if (atom->lock_core)
-		mutex_lock(&disp->mutex);
-
-	/* Disable head(s). */
-	for_each_oldnew_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) {
-		struct nv50_head_atom *asyh = nv50_head_atom(new_crtc_state);
-		struct nv50_head *head = nv50_head(crtc);
-
-		NV_ATOMIC(drm, "%s: clr %04x (set %04x)\n", crtc->name,
-			  asyh->clr.mask, asyh->set.mask);
-		if (old_crtc_state->active && !new_crtc_state->active)
-			drm_crtc_vblank_off(crtc);
-
-		if (asyh->clr.mask) {
-			nv50_head_flush_clr(head, asyh, atom->flush_disable);
-			interlock_core |= 1;
-		}
-	}
-
-	/* Disable plane(s). */
-	for_each_new_plane_in_state(state, plane, new_plane_state, i) {
-		struct nv50_wndw_atom *asyw = nv50_wndw_atom(new_plane_state);
-		struct nv50_wndw *wndw = nv50_wndw(plane);
-
-		NV_ATOMIC(drm, "%s: clr %02x (set %02x)\n", plane->name,
-			  asyw->clr.mask, asyw->set.mask);
-		if (!asyw->clr.mask)
-			continue;
-
-		interlock_chan |= nv50_wndw_flush_clr(wndw, interlock_core,
-						      atom->flush_disable,
-						      asyw);
-	}
-
-	/* Disable output path(s). */
-	list_for_each_entry(outp, &atom->outp, head) {
-		const struct drm_encoder_helper_funcs *help;
-		struct drm_encoder *encoder;
-
-		encoder = outp->encoder;
-		help = encoder->helper_private;
-
-		NV_ATOMIC(drm, "%s: clr %02x (set %02x)\n", encoder->name,
-			  outp->clr.mask, outp->set.mask);
-
-		if (outp->clr.mask) {
-			help->disable(encoder);
-			interlock_core |= 1;
-			if (outp->flush_disable) {
-				nv50_disp_atomic_commit_core(drm, interlock_chan);
-				interlock_core = 0;
-				interlock_chan = 0;
-			}
-		}
-	}
-
-	/* Flush disable. */
-	if (interlock_core) {
-		if (atom->flush_disable) {
-			nv50_disp_atomic_commit_core(drm, interlock_chan);
-			interlock_core = 0;
-			interlock_chan = 0;
-		}
-	}
-
-	/* Update output path(s). */
-	list_for_each_entry_safe(outp, outt, &atom->outp, head) {
-		const struct drm_encoder_helper_funcs *help;
-		struct drm_encoder *encoder;
-
-		encoder = outp->encoder;
-		help = encoder->helper_private;
-
-		NV_ATOMIC(drm, "%s: set %02x (clr %02x)\n", encoder->name,
-			  outp->set.mask, outp->clr.mask);
-
-		if (outp->set.mask) {
-			help->enable(encoder);
-			interlock_core = 1;
-		}
-
-		list_del(&outp->head);
-		kfree(outp);
-	}
-
-	/* Update head(s). */
-	for_each_oldnew_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) {
-		struct nv50_head_atom *asyh = nv50_head_atom(new_crtc_state);
-		struct nv50_head *head = nv50_head(crtc);
-
-		NV_ATOMIC(drm, "%s: set %04x (clr %04x)\n", crtc->name,
-			  asyh->set.mask, asyh->clr.mask);
-
-		if (asyh->set.mask) {
-			nv50_head_flush_set(head, asyh);
-			interlock_core = 1;
-		}
-
-		if (new_crtc_state->active) {
-			if (!old_crtc_state->active)
-				drm_crtc_vblank_on(crtc);
-			if (new_crtc_state->event)
-				drm_crtc_vblank_get(crtc);
-		}
-	}
-
-	/* Update plane(s). */
-	for_each_new_plane_in_state(state, plane, new_plane_state, i) {
-		struct nv50_wndw_atom *asyw = nv50_wndw_atom(new_plane_state);
-		struct nv50_wndw *wndw = nv50_wndw(plane);
-
-		NV_ATOMIC(drm, "%s: set %02x (clr %02x)\n", plane->name,
-			  asyw->set.mask, asyw->clr.mask);
-		if ( !asyw->set.mask &&
-		    (!asyw->clr.mask || atom->flush_disable))
-			continue;
-
-		interlock_chan |= nv50_wndw_flush_set(wndw, interlock_core, asyw);
-	}
-
-	/* Flush update. */
-	if (interlock_core) {
-		if (!interlock_chan && atom->state.legacy_cursor_update) {
-			u32 *push = evo_wait(&disp->mast, 2);
-			if (push) {
-				evo_mthd(push, 0x0080, 1);
-				evo_data(push, 0x00000000);
-				evo_kick(push, &disp->mast);
-			}
-		} else {
-			nv50_disp_atomic_commit_core(drm, interlock_chan);
-		}
-	}
-
-	if (atom->lock_core)
-		mutex_unlock(&disp->mutex);
-
-	/* Wait for HW to signal completion. */
-	for_each_new_plane_in_state(state, plane, new_plane_state, i) {
-		struct nv50_wndw_atom *asyw = nv50_wndw_atom(new_plane_state);
-		struct nv50_wndw *wndw = nv50_wndw(plane);
-		int ret = nv50_wndw_wait_armed(wndw, asyw);
-		if (ret)
-			NV_ERROR(drm, "%s: timeout\n", plane->name);
-	}
-
-	for_each_new_crtc_in_state(state, crtc, new_crtc_state, i) {
-		if (new_crtc_state->event) {
-			unsigned long flags;
-			/* Get correct count/ts if racing with vblank irq */
-			if (new_crtc_state->active)
-				drm_crtc_accurate_vblank_count(crtc);
-			spin_lock_irqsave(&crtc->dev->event_lock, flags);
-			drm_crtc_send_vblank_event(crtc, new_crtc_state->event);
-			spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
-
-			new_crtc_state->event = NULL;
-			if (new_crtc_state->active)
-				drm_crtc_vblank_put(crtc);
-		}
-	}
-
-	drm_atomic_helper_commit_hw_done(state);
-	drm_atomic_helper_cleanup_planes(dev, state);
-	drm_atomic_helper_commit_cleanup_done(state);
-	drm_atomic_state_put(state);
-}
-
-static void
-nv50_disp_atomic_commit_work(struct work_struct *work)
-{
-	struct drm_atomic_state *state =
-		container_of(work, typeof(*state), commit_work);
-	nv50_disp_atomic_commit_tail(state);
-}
-
-static int
-nv50_disp_atomic_commit(struct drm_device *dev,
-			struct drm_atomic_state *state, bool nonblock)
-{
-	struct nouveau_drm *drm = nouveau_drm(dev);
-	struct nv50_disp *disp = nv50_disp(dev);
-	struct drm_plane_state *new_plane_state;
-	struct drm_plane *plane;
-	struct drm_crtc *crtc;
-	bool active = false;
-	int ret, i;
-
-	ret = pm_runtime_get_sync(dev->dev);
-	if (ret < 0 && ret != -EACCES)
-		return ret;
-
-	ret = drm_atomic_helper_setup_commit(state, nonblock);
-	if (ret)
-		goto done;
-
-	INIT_WORK(&state->commit_work, nv50_disp_atomic_commit_work);
-
-	ret = drm_atomic_helper_prepare_planes(dev, state);
-	if (ret)
-		goto done;
-
-	if (!nonblock) {
-		ret = drm_atomic_helper_wait_for_fences(dev, state, true);
-		if (ret)
-			goto err_cleanup;
-	}
-
-	ret = drm_atomic_helper_swap_state(state, true);
-	if (ret)
-		goto err_cleanup;
-
-	for_each_new_plane_in_state(state, plane, new_plane_state, i) {
-		struct nv50_wndw_atom *asyw = nv50_wndw_atom(new_plane_state);
-		struct nv50_wndw *wndw = nv50_wndw(plane);
-
-		if (asyw->set.image) {
-			asyw->ntfy.handle = wndw->dmac->sync.handle;
-			asyw->ntfy.offset = wndw->ntfy;
-			asyw->ntfy.awaken = false;
-			asyw->set.ntfy = true;
-			nouveau_bo_wr32(disp->sync, wndw->ntfy / 4, 0x00000000);
-			wndw->ntfy ^= 0x10;
-		}
-	}
-
-	drm_atomic_state_get(state);
-
-	if (nonblock)
-		queue_work(system_unbound_wq, &state->commit_work);
-	else
-		nv50_disp_atomic_commit_tail(state);
-
-	drm_for_each_crtc(crtc, dev) {
-		if (crtc->state->enable) {
-			if (!drm->have_disp_power_ref) {
-				drm->have_disp_power_ref = true;
-				return 0;
-			}
-			active = true;
-			break;
-		}
-	}
-
-	if (!active && drm->have_disp_power_ref) {
-		pm_runtime_put_autosuspend(dev->dev);
-		drm->have_disp_power_ref = false;
-	}
-
-err_cleanup:
-	if (ret)
-		drm_atomic_helper_cleanup_planes(dev, state);
-done:
-	pm_runtime_put_autosuspend(dev->dev);
-	return ret;
-}
-
-static struct nv50_outp_atom *
-nv50_disp_outp_atomic_add(struct nv50_atom *atom, struct drm_encoder *encoder)
-{
-	struct nv50_outp_atom *outp;
-
-	list_for_each_entry(outp, &atom->outp, head) {
-		if (outp->encoder == encoder)
-			return outp;
-	}
-
-	outp = kzalloc(sizeof(*outp), GFP_KERNEL);
-	if (!outp)
-		return ERR_PTR(-ENOMEM);
-
-	list_add(&outp->head, &atom->outp);
-	outp->encoder = encoder;
-	return outp;
-}
-
-static int
-nv50_disp_outp_atomic_check_clr(struct nv50_atom *atom,
-				struct drm_connector_state *old_connector_state)
-{
-	struct drm_encoder *encoder = old_connector_state->best_encoder;
-	struct drm_crtc_state *old_crtc_state, *new_crtc_state;
-	struct drm_crtc *crtc;
-	struct nv50_outp_atom *outp;
-
-	if (!(crtc = old_connector_state->crtc))
-		return 0;
-
-	old_crtc_state = drm_atomic_get_old_crtc_state(&atom->state, crtc);
-	new_crtc_state = drm_atomic_get_new_crtc_state(&atom->state, crtc);
-	if (old_crtc_state->active && drm_atomic_crtc_needs_modeset(new_crtc_state)) {
-		outp = nv50_disp_outp_atomic_add(atom, encoder);
-		if (IS_ERR(outp))
-			return PTR_ERR(outp);
-
-		if (outp->encoder->encoder_type == DRM_MODE_ENCODER_DPMST) {
-			outp->flush_disable = true;
-			atom->flush_disable = true;
-		}
-		outp->clr.ctrl = true;
-		atom->lock_core = true;
-	}
-
-	return 0;
-}
-
-static int
-nv50_disp_outp_atomic_check_set(struct nv50_atom *atom,
-				struct drm_connector_state *connector_state)
-{
-	struct drm_encoder *encoder = connector_state->best_encoder;
-	struct drm_crtc_state *new_crtc_state;
-	struct drm_crtc *crtc;
-	struct nv50_outp_atom *outp;
-
-	if (!(crtc = connector_state->crtc))
-		return 0;
-
-	new_crtc_state = drm_atomic_get_new_crtc_state(&atom->state, crtc);
-	if (new_crtc_state->active && drm_atomic_crtc_needs_modeset(new_crtc_state)) {
-		outp = nv50_disp_outp_atomic_add(atom, encoder);
-		if (IS_ERR(outp))
-			return PTR_ERR(outp);
-
-		outp->set.ctrl = true;
-		atom->lock_core = true;
-	}
-
-	return 0;
-}
-
-static int
-nv50_disp_atomic_check(struct drm_device *dev, struct drm_atomic_state *state)
-{
-	struct nv50_atom *atom = nv50_atom(state);
-	struct drm_connector_state *old_connector_state, *new_connector_state;
-	struct drm_connector *connector;
-	int ret, i;
-
-	ret = drm_atomic_helper_check(dev, state);
-	if (ret)
-		return ret;
-
-	for_each_oldnew_connector_in_state(state, connector, old_connector_state, new_connector_state, i) {
-		ret = nv50_disp_outp_atomic_check_clr(atom, old_connector_state);
-		if (ret)
-			return ret;
-
-		ret = nv50_disp_outp_atomic_check_set(atom, new_connector_state);
-		if (ret)
-			return ret;
-	}
-
-	return 0;
-}
-
-static void
-nv50_disp_atomic_state_clear(struct drm_atomic_state *state)
-{
-	struct nv50_atom *atom = nv50_atom(state);
-	struct nv50_outp_atom *outp, *outt;
-
-	list_for_each_entry_safe(outp, outt, &atom->outp, head) {
-		list_del(&outp->head);
-		kfree(outp);
-	}
-
-	drm_atomic_state_default_clear(state);
-}
-
-static void
-nv50_disp_atomic_state_free(struct drm_atomic_state *state)
-{
-	struct nv50_atom *atom = nv50_atom(state);
-	drm_atomic_state_default_release(&atom->state);
-	kfree(atom);
-}
-
-static struct drm_atomic_state *
-nv50_disp_atomic_state_alloc(struct drm_device *dev)
-{
-	struct nv50_atom *atom;
-	if (!(atom = kzalloc(sizeof(*atom), GFP_KERNEL)) ||
-	    drm_atomic_state_init(dev, &atom->state) < 0) {
-		kfree(atom);
-		return NULL;
-	}
-	INIT_LIST_HEAD(&atom->outp);
-	return &atom->state;
-}
-
-static const struct drm_mode_config_funcs
-nv50_disp_func = {
-	.fb_create = nouveau_user_framebuffer_create,
-	.output_poll_changed = drm_fb_helper_output_poll_changed,
-	.atomic_check = nv50_disp_atomic_check,
-	.atomic_commit = nv50_disp_atomic_commit,
-	.atomic_state_alloc = nv50_disp_atomic_state_alloc,
-	.atomic_state_clear = nv50_disp_atomic_state_clear,
-	.atomic_state_free = nv50_disp_atomic_state_free,
-};
-
-/******************************************************************************
- * Init
- *****************************************************************************/
-
-void
-nv50_display_fini(struct drm_device *dev)
-{
-	struct nouveau_encoder *nv_encoder;
-	struct drm_encoder *encoder;
-	struct drm_plane *plane;
-
-	drm_for_each_plane(plane, dev) {
-		struct nv50_wndw *wndw = nv50_wndw(plane);
-		if (plane->funcs != &nv50_wndw)
-			continue;
-		nv50_wndw_fini(wndw);
-	}
-
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
-		if (encoder->encoder_type != DRM_MODE_ENCODER_DPMST) {
-			nv_encoder = nouveau_encoder(encoder);
-			nv50_mstm_fini(nv_encoder->dp.mstm);
-		}
-	}
-}
-
-int
-nv50_display_init(struct drm_device *dev)
-{
-	struct drm_encoder *encoder;
-	struct drm_plane *plane;
-	u32 *push;
-
-	push = evo_wait(nv50_mast(dev), 32);
-	if (!push)
-		return -EBUSY;
-
-	evo_mthd(push, 0x0088, 1);
-	evo_data(push, nv50_mast(dev)->base.sync.handle);
-	evo_kick(push, nv50_mast(dev));
-
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
-		if (encoder->encoder_type != DRM_MODE_ENCODER_DPMST) {
-			struct nouveau_encoder *nv_encoder =
-				nouveau_encoder(encoder);
-			nv50_mstm_init(nv_encoder->dp.mstm);
-		}
-	}
-
-	drm_for_each_plane(plane, dev) {
-		struct nv50_wndw *wndw = nv50_wndw(plane);
-		if (plane->funcs != &nv50_wndw)
-			continue;
-		nv50_wndw_init(wndw);
-	}
-
-	return 0;
-}
-
-void
-nv50_display_destroy(struct drm_device *dev)
-{
-	struct nv50_disp *disp = nv50_disp(dev);
-
-	nv50_dmac_destroy(&disp->mast.base);
-
-	nouveau_bo_unmap(disp->sync);
-	if (disp->sync)
-		nouveau_bo_unpin(disp->sync);
-	nouveau_bo_ref(NULL, &disp->sync);
-
-	nouveau_display(dev)->priv = NULL;
-	kfree(disp);
-}
-
-MODULE_PARM_DESC(atomic, "Expose atomic ioctl (default: disabled)");
-static int nouveau_atomic = 0;
-module_param_named(atomic, nouveau_atomic, int, 0400);
-
-int
-nv50_display_create(struct drm_device *dev)
-{
-	struct nvif_device *device = &nouveau_drm(dev)->client.device;
-	struct nouveau_drm *drm = nouveau_drm(dev);
-	struct dcb_table *dcb = &drm->vbios.dcb;
-	struct drm_connector *connector, *tmp;
-	struct nv50_disp *disp;
-	struct dcb_output *dcbe;
-	int crtcs, ret, i;
-
-	disp = kzalloc(sizeof(*disp), GFP_KERNEL);
-	if (!disp)
-		return -ENOMEM;
-
-	mutex_init(&disp->mutex);
-
-	nouveau_display(dev)->priv = disp;
-	nouveau_display(dev)->dtor = nv50_display_destroy;
-	nouveau_display(dev)->init = nv50_display_init;
-	nouveau_display(dev)->fini = nv50_display_fini;
-	disp->disp = &nouveau_display(dev)->disp;
-	dev->mode_config.funcs = &nv50_disp_func;
-	dev->driver->driver_features |= DRIVER_PREFER_XBGR_30BPP;
-	if (nouveau_atomic)
-		dev->driver->driver_features |= DRIVER_ATOMIC;
-
-	/* small shared memory area we use for notifiers and semaphores */
-	ret = nouveau_bo_new(&drm->client, 4096, 0x1000, TTM_PL_FLAG_VRAM,
-			     0, 0x0000, NULL, NULL, &disp->sync);
-	if (!ret) {
-		ret = nouveau_bo_pin(disp->sync, TTM_PL_FLAG_VRAM, true);
-		if (!ret) {
-			ret = nouveau_bo_map(disp->sync);
-			if (ret)
-				nouveau_bo_unpin(disp->sync);
-		}
-		if (ret)
-			nouveau_bo_ref(NULL, &disp->sync);
-	}
-
-	if (ret)
-		goto out;
-
-	/* allocate master evo channel */
-	ret = nv50_core_create(device, &disp->disp->object,
-			       disp->sync->bo.offset, &disp->mast);
-	if (ret)
-		goto out;
-
-	/* create crtc objects to represent the hw heads */
-	if (disp->disp->object.oclass >= GF110_DISP)
-		crtcs = nvif_rd32(&device->object, 0x612004) & 0xf;
-	else
-		crtcs = 0x3;
-
-	for (i = 0; i < fls(crtcs); i++) {
-		if (!(crtcs & (1 << i)))
-			continue;
-		ret = nv50_head_create(dev, i);
-		if (ret)
-			goto out;
-	}
-
-	/* create encoder/connector objects based on VBIOS DCB table */
-	for (i = 0, dcbe = &dcb->entry[0]; i < dcb->entries; i++, dcbe++) {
-		connector = nouveau_connector_create(dev, dcbe->connector);
-		if (IS_ERR(connector))
-			continue;
-
-		if (dcbe->location == DCB_LOC_ON_CHIP) {
-			switch (dcbe->type) {
-			case DCB_OUTPUT_TMDS:
-			case DCB_OUTPUT_LVDS:
-			case DCB_OUTPUT_DP:
-				ret = nv50_sor_create(connector, dcbe);
-				break;
-			case DCB_OUTPUT_ANALOG:
-				ret = nv50_dac_create(connector, dcbe);
-				break;
-			default:
-				ret = -ENODEV;
-				break;
-			}
-		} else {
-			ret = nv50_pior_create(connector, dcbe);
-		}
-
-		if (ret) {
-			NV_WARN(drm, "failed to create encoder %d/%d/%d: %d\n",
-				     dcbe->location, dcbe->type,
-				     ffs(dcbe->or) - 1, ret);
-			ret = 0;
-		}
-	}
-
-	/* cull any connectors we created that don't have an encoder */
-	list_for_each_entry_safe(connector, tmp, &dev->mode_config.connector_list, head) {
-		if (connector->encoder_ids[0])
-			continue;
-
-		NV_WARN(drm, "%s has no encoders, removing\n",
-			connector->name);
-		connector->funcs->destroy(connector);
-	}
-
-out:
-	if (ret)
-		nv50_display_destroy(dev);
-	return ret;
-}
-- 
cgit v1.1


From 62b290fc7b36e8fec2a370b946d7117c1899b6c1 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50-: fix i2c-over-aux on anx9805

We don't support address-only transactions there.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/disp.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c
index f2156c8..9aa1750 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/disp.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c
@@ -3836,7 +3836,6 @@ nv50_pior_func = {
 static int
 nv50_pior_create(struct drm_connector *connector, struct dcb_output *dcbe)
 {
-	struct nouveau_connector *nv_connector = nouveau_connector(connector);
 	struct nouveau_drm *drm = nouveau_drm(connector->dev);
 	struct nvkm_i2c *i2c = nvxx_i2c(&drm->client.device);
 	struct nvkm_i2c_bus *bus = NULL;
@@ -3854,7 +3853,7 @@ nv50_pior_create(struct drm_connector *connector, struct dcb_output *dcbe)
 		break;
 	case DCB_OUTPUT_DP:
 		aux  = nvkm_i2c_aux_find(i2c, NVKM_I2C_AUX_EXT(dcbe->extdev));
-		ddc  = aux ? &nv_connector->aux.ddc : NULL;
+		ddc  = aux ? &aux->i2c : NULL;
 		type = DRM_MODE_ENCODER_TMDS;
 		break;
 	default:
-- 
cgit v1.1


From 5bca1621c07c3ad37b5a4943450a892e18984df0 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50-: move fb ctxdma tracking into windows

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/disp.c | 163 ++++++++++++++++----------------
 1 file changed, 84 insertions(+), 79 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c
index 9aa1750..fc3055d 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/disp.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c
@@ -395,7 +395,7 @@ nv50_oimm_create(struct nvif_device *device, struct nvif_object *disp,
  * DMA EVO channel
  *****************************************************************************/
 
-struct nv50_dmac_ctxdma {
+struct nv50_wndw_ctxdma {
 	struct list_head head;
 	struct nvif_object object;
 };
@@ -408,7 +408,6 @@ struct nv50_dmac {
 
 	struct nvif_object sync;
 	struct nvif_object vram;
-	struct list_head ctxdma;
 
 	/* Protects against concurrent pushbuf access to this channel, lock is
 	 * grabbed by evo_wait (if the pushbuf reservation is successful) and
@@ -417,82 +416,8 @@ struct nv50_dmac {
 };
 
 static void
-nv50_dmac_ctxdma_del(struct nv50_dmac_ctxdma *ctxdma)
-{
-	nvif_object_fini(&ctxdma->object);
-	list_del(&ctxdma->head);
-	kfree(ctxdma);
-}
-
-static struct nv50_dmac_ctxdma *
-nv50_dmac_ctxdma_new(struct nv50_dmac *dmac, struct nouveau_framebuffer *fb)
-{
-	struct nouveau_drm *drm = nouveau_drm(fb->base.dev);
-	struct nv50_dmac_ctxdma *ctxdma;
-	const u8    kind = fb->nvbo->kind;
-	const u32 handle = 0xfb000000 | kind;
-	struct {
-		struct nv_dma_v0 base;
-		union {
-			struct nv50_dma_v0 nv50;
-			struct gf100_dma_v0 gf100;
-			struct gf119_dma_v0 gf119;
-		};
-	} args = {};
-	u32 argc = sizeof(args.base);
-	int ret;
-
-	list_for_each_entry(ctxdma, &dmac->ctxdma, head) {
-		if (ctxdma->object.handle == handle)
-			return ctxdma;
-	}
-
-	if (!(ctxdma = kzalloc(sizeof(*ctxdma), GFP_KERNEL)))
-		return ERR_PTR(-ENOMEM);
-	list_add(&ctxdma->head, &dmac->ctxdma);
-
-	args.base.target = NV_DMA_V0_TARGET_VRAM;
-	args.base.access = NV_DMA_V0_ACCESS_RDWR;
-	args.base.start  = 0;
-	args.base.limit  = drm->client.device.info.ram_user - 1;
-
-	if (drm->client.device.info.chipset < 0x80) {
-		args.nv50.part = NV50_DMA_V0_PART_256;
-		argc += sizeof(args.nv50);
-	} else
-	if (drm->client.device.info.chipset < 0xc0) {
-		args.nv50.part = NV50_DMA_V0_PART_256;
-		args.nv50.kind = kind;
-		argc += sizeof(args.nv50);
-	} else
-	if (drm->client.device.info.chipset < 0xd0) {
-		args.gf100.kind = kind;
-		argc += sizeof(args.gf100);
-	} else {
-		args.gf119.page = GF119_DMA_V0_PAGE_LP;
-		args.gf119.kind = kind;
-		argc += sizeof(args.gf119);
-	}
-
-	ret = nvif_object_init(&dmac->base.user, handle, NV_DMA_IN_MEMORY,
-			       &args, argc, &ctxdma->object);
-	if (ret) {
-		nv50_dmac_ctxdma_del(ctxdma);
-		return ERR_PTR(ret);
-	}
-
-	return ctxdma;
-}
-
-static void
 nv50_dmac_destroy(struct nv50_dmac *dmac)
 {
-	struct nv50_dmac_ctxdma *ctxdma, *ctxtmp;
-
-	list_for_each_entry_safe(ctxdma, ctxtmp, &dmac->ctxdma, head) {
-		nv50_dmac_ctxdma_del(ctxdma);
-	}
-
 	nvif_object_fini(&dmac->vram);
 	nvif_object_fini(&dmac->sync);
 
@@ -511,7 +436,6 @@ nv50_dmac_create(struct nvif_device *device, struct nvif_object *disp,
 	int ret;
 
 	mutex_init(&dmac->lock);
-	INIT_LIST_HEAD(&dmac->ctxdma);
 
 	ret = nvif_mem_init_map(&cli->mmu, NVIF_MEM_COHERENT, 0x1000,
 				&dmac->push);
@@ -740,6 +664,11 @@ struct nv50_wndw {
 	const struct nv50_wndw_func *func;
 	struct nv50_dmac *dmac;
 
+	struct {
+		struct nvif_object *parent;
+		struct list_head list;
+	} ctxdma;
+
 	struct drm_plane plane;
 
 	struct nvif_notify notify;
@@ -770,6 +699,74 @@ struct nv50_wndw_func {
 	u32 (*update)(struct nv50_wndw *, u32 interlock);
 };
 
+static void
+nv50_wndw_ctxdma_del(struct nv50_wndw_ctxdma *ctxdma)
+{
+	nvif_object_fini(&ctxdma->object);
+	list_del(&ctxdma->head);
+	kfree(ctxdma);
+}
+
+static struct nv50_wndw_ctxdma *
+nv50_wndw_ctxdma_new(struct nv50_wndw *wndw, struct nouveau_framebuffer *fb)
+{
+	struct nouveau_drm *drm = nouveau_drm(fb->base.dev);
+	struct nv50_wndw_ctxdma *ctxdma;
+	const u8    kind = fb->nvbo->kind;
+	const u32 handle = 0xfb000000 | kind;
+	struct {
+		struct nv_dma_v0 base;
+		union {
+			struct nv50_dma_v0 nv50;
+			struct gf100_dma_v0 gf100;
+			struct gf119_dma_v0 gf119;
+		};
+	} args = {};
+	u32 argc = sizeof(args.base);
+	int ret;
+
+	list_for_each_entry(ctxdma, &wndw->ctxdma.list, head) {
+		if (ctxdma->object.handle == handle)
+			return ctxdma;
+	}
+
+	if (!(ctxdma = kzalloc(sizeof(*ctxdma), GFP_KERNEL)))
+		return ERR_PTR(-ENOMEM);
+	list_add(&ctxdma->head, &wndw->ctxdma.list);
+
+	args.base.target = NV_DMA_V0_TARGET_VRAM;
+	args.base.access = NV_DMA_V0_ACCESS_RDWR;
+	args.base.start  = 0;
+	args.base.limit  = drm->client.device.info.ram_user - 1;
+
+	if (drm->client.device.info.chipset < 0x80) {
+		args.nv50.part = NV50_DMA_V0_PART_256;
+		argc += sizeof(args.nv50);
+	} else
+	if (drm->client.device.info.chipset < 0xc0) {
+		args.nv50.part = NV50_DMA_V0_PART_256;
+		args.nv50.kind = kind;
+		argc += sizeof(args.nv50);
+	} else
+	if (drm->client.device.info.chipset < 0xd0) {
+		args.gf100.kind = kind;
+		argc += sizeof(args.gf100);
+	} else {
+		args.gf119.page = GF119_DMA_V0_PAGE_LP;
+		args.gf119.kind = kind;
+		argc += sizeof(args.gf119);
+	}
+
+	ret = nvif_object_init(wndw->ctxdma.parent, handle, NV_DMA_IN_MEMORY,
+			       &args, argc, &ctxdma->object);
+	if (ret) {
+		nv50_wndw_ctxdma_del(ctxdma);
+		return ERR_PTR(ret);
+	}
+
+	return ctxdma;
+}
+
 static int
 nv50_wndw_wait_armed(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 {
@@ -944,7 +941,7 @@ nv50_wndw_prepare_fb(struct drm_plane *plane, struct drm_plane_state *state)
 	struct nv50_wndw *wndw = nv50_wndw(plane);
 	struct nv50_wndw_atom *asyw = nv50_wndw_atom(state);
 	struct nv50_head_atom *asyh;
-	struct nv50_dmac_ctxdma *ctxdma;
+	struct nv50_wndw_ctxdma *ctxdma;
 	int ret;
 
 	NV_ATOMIC(drm, "%s prepare: %p\n", plane->name, state->fb);
@@ -955,7 +952,7 @@ nv50_wndw_prepare_fb(struct drm_plane *plane, struct drm_plane_state *state)
 	if (ret)
 		return ret;
 
-	ctxdma = nv50_dmac_ctxdma_new(wndw->dmac, fb);
+	ctxdma = nv50_wndw_ctxdma_new(wndw, fb);
 	if (IS_ERR(ctxdma)) {
 		nouveau_bo_unpin(fb->nvbo);
 		return PTR_ERR(ctxdma);
@@ -1030,7 +1027,13 @@ static void
 nv50_wndw_destroy(struct drm_plane *plane)
 {
 	struct nv50_wndw *wndw = nv50_wndw(plane);
+	struct nv50_wndw_ctxdma *ctxdma, *ctxtmp;
 	void *data;
+
+	list_for_each_entry_safe(ctxdma, ctxtmp, &wndw->ctxdma.list, head) {
+		nv50_wndw_ctxdma_del(ctxdma);
+	}
+
 	nvif_notify_fini(&wndw->notify);
 	data = wndw->func->dtor(wndw);
 	drm_plane_cleanup(&wndw->plane);
@@ -1069,6 +1072,7 @@ nv50_wndw_ctor(const struct nv50_wndw_func *func, struct drm_device *dev,
 
 	wndw->func = func;
 	wndw->dmac = dmac;
+	wndw->ctxdma.parent = &dmac->base.user;
 
 	ret = drm_universal_plane_init(dev, &wndw->plane, 0, &nv50_wndw,
 				       format, nformat, NULL,
@@ -1077,6 +1081,7 @@ nv50_wndw_ctor(const struct nv50_wndw_func *func, struct drm_device *dev,
 		return ret;
 
 	drm_plane_helper_add(&wndw->plane, &nv50_wndw_helper);
+	INIT_LIST_HEAD(&wndw->ctxdma.list);
 	return 0;
 }
 
-- 
cgit v1.1


From a97c530eb968bad8d945d4f64fb550fa37a9d362 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50-: modify overlay allocation so the code can be
 split

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/disp.c | 277 +++++++++++++++++++++-----------
 1 file changed, 179 insertions(+), 98 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c
index fc3055d..26caca2 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/disp.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c
@@ -341,57 +341,6 @@ nv50_chan_destroy(struct nv50_chan *chan)
 }
 
 /******************************************************************************
- * PIO EVO channel
- *****************************************************************************/
-
-struct nv50_pioc {
-	struct nv50_chan base;
-};
-
-static void
-nv50_pioc_destroy(struct nv50_pioc *pioc)
-{
-	nv50_chan_destroy(&pioc->base);
-}
-
-static int
-nv50_pioc_create(struct nvif_device *device, struct nvif_object *disp,
-		 const s32 *oclass, u8 head, void *data, u32 size,
-		 struct nv50_pioc *pioc)
-{
-	return nv50_chan_create(device, disp, oclass, head, data, size,
-				&pioc->base);
-}
-
-/******************************************************************************
- * Overlay Immediate
- *****************************************************************************/
-
-struct nv50_oimm {
-	struct nv50_pioc base;
-};
-
-static int
-nv50_oimm_create(struct nvif_device *device, struct nvif_object *disp,
-		 int head, struct nv50_oimm *oimm)
-{
-	struct nv50_disp_cursor_v0 args = {
-		.head = head,
-	};
-	static const s32 oclass[] = {
-		GK104_DISP_OVERLAY,
-		GF110_DISP_OVERLAY,
-		GT214_DISP_OVERLAY,
-		G82_DISP_OVERLAY,
-		NV50_DISP_OVERLAY,
-		0
-	};
-
-	return nv50_pioc_create(device, disp, oclass, head, &args, sizeof(args),
-				&oimm->base);
-}
-
-/******************************************************************************
  * DMA EVO channel
  *****************************************************************************/
 
@@ -541,43 +490,12 @@ nv50_base_create(struct nvif_device *device, struct nvif_object *disp,
 				syncbuf, &base->base);
 }
 
-/******************************************************************************
- * Overlay
- *****************************************************************************/
-
-struct nv50_ovly {
-	struct nv50_dmac base;
-};
-
-static int
-nv50_ovly_create(struct nvif_device *device, struct nvif_object *disp,
-		 int head, u64 syncbuf, struct nv50_ovly *ovly)
-{
-	struct nv50_disp_overlay_channel_dma_v0 args = {
-		.head = head,
-	};
-	static const s32 oclass[] = {
-		GK104_DISP_OVERLAY_CONTROL_DMA,
-		GF110_DISP_OVERLAY_CONTROL_DMA,
-		GT214_DISP_OVERLAY_CHANNEL_DMA,
-		GT200_DISP_OVERLAY_CHANNEL_DMA,
-		G82_DISP_OVERLAY_CHANNEL_DMA,
-		NV50_DISP_OVERLAY_CHANNEL_DMA,
-		0
-	};
-
-	return nv50_dmac_create(device, disp, oclass, head, &args, sizeof(args),
-				syncbuf, &ovly->base);
-}
-
 struct nv50_head {
 	struct nouveau_crtc base;
 	struct {
 		struct nouveau_bo *nvbo[2];
 		int next;
 	} lut;
-	struct nv50_ovly ovly;
-	struct nv50_oimm oimm;
 };
 
 #define nv50_head(c) ((struct nv50_head *)nouveau_crtc(c))
@@ -662,7 +580,9 @@ evo_kick(u32 *push, void *evoc)
 
 struct nv50_wndw {
 	const struct nv50_wndw_func *func;
+	const struct nv50_wimm_func *immd;
 	struct nv50_dmac *dmac;
+	int id;
 
 	struct {
 		struct nvif_object *parent;
@@ -671,6 +591,9 @@ struct nv50_wndw {
 
 	struct drm_plane plane;
 
+	struct nv50_dmac wndw;
+	struct nv50_dmac wimm;
+
 	struct nvif_notify notify;
 	u16 ntfy;
 	u16 sema;
@@ -699,6 +622,9 @@ struct nv50_wndw_func {
 	u32 (*update)(struct nv50_wndw *, u32 interlock);
 };
 
+struct nv50_wimm_func {
+};
+
 static void
 nv50_wndw_ctxdma_del(struct nv50_wndw_ctxdma *ctxdma)
 {
@@ -1028,14 +954,17 @@ nv50_wndw_destroy(struct drm_plane *plane)
 {
 	struct nv50_wndw *wndw = nv50_wndw(plane);
 	struct nv50_wndw_ctxdma *ctxdma, *ctxtmp;
-	void *data;
+	void *data = wndw;
 
 	list_for_each_entry_safe(ctxdma, ctxtmp, &wndw->ctxdma.list, head) {
 		nv50_wndw_ctxdma_del(ctxdma);
 	}
 
 	nvif_notify_fini(&wndw->notify);
-	data = wndw->func->dtor(wndw);
+	if (wndw->func->dtor)
+		data = wndw->func->dtor(wndw);
+	nv50_dmac_destroy(&wndw->wimm);
+	nv50_dmac_destroy(&wndw->wndw);
 	drm_plane_cleanup(&wndw->plane);
 	kfree(data);
 }
@@ -1085,6 +1014,170 @@ nv50_wndw_ctor(const struct nv50_wndw_func *func, struct drm_device *dev,
 	return 0;
 }
 
+static int
+nv50_wndw_new_(const struct nv50_wndw_func *func, struct drm_device *dev,
+	       enum drm_plane_type type, const char *name, int index,
+	       const u32 *format, struct nv50_wndw **pwndw)
+{
+	struct nv50_wndw *wndw;
+	int nformat;
+	int ret;
+
+	if (!(wndw = *pwndw = kzalloc(sizeof(*wndw), GFP_KERNEL)))
+		return -ENOMEM;
+	wndw->id = index;
+
+	for (nformat = 0; format[nformat]; nformat++);
+
+	ret = nv50_wndw_ctor(func, dev, type, name, index,
+			     &wndw->wndw, format, nformat, wndw);
+	if (ret) {
+		kfree(*pwndw);
+		*pwndw = NULL;
+	}
+
+	return ret;
+}
+
+/******************************************************************************
+ * Overlay
+ *****************************************************************************/
+
+static const struct nv50_wimm_func
+oimm507b = {
+};
+
+static int
+oimm507b_init_(const struct nv50_wimm_func *func, struct nouveau_drm *drm,
+	       s32 oclass, struct nv50_wndw *wndw)
+{
+	struct nv50_disp_overlay_v0 args = {
+		.head = wndw->id,
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	int ret;
+
+	ret = nvif_object_init(&disp->disp->object, 0, oclass, &args,
+			       sizeof(args), &wndw->wimm.base.user);
+	if (ret) {
+		NV_ERROR(drm, "oimm%04x allocation failed: %d\n", oclass, ret);
+		return ret;
+	}
+
+	nvif_object_map(&wndw->wimm.base.user, NULL, 0);
+	wndw->immd = func;
+	return 0;
+}
+
+static int
+oimm507b_init(struct nouveau_drm *drm, s32 oclass, struct nv50_wndw *wndw)
+{
+	return oimm507b_init_(&oimm507b, drm, oclass, wndw);
+}
+
+static int
+nv50_oimm_init(struct nouveau_drm *drm, struct nv50_wndw *wndw)
+{
+	static const struct {
+		s32 oclass;
+		int version;
+		int (*init)(struct nouveau_drm *, s32, struct nv50_wndw *);
+	} oimms[] = {
+		{ GK104_DISP_OVERLAY, 0, oimm507b_init },
+		{ GF110_DISP_OVERLAY, 0, oimm507b_init },
+		{ GT214_DISP_OVERLAY, 0, oimm507b_init },
+		{   G82_DISP_OVERLAY, 0, oimm507b_init },
+		{  NV50_DISP_OVERLAY, 0, oimm507b_init },
+		{}
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	int cid;
+
+	cid = nvif_mclass(&disp->disp->object, oimms);
+	if (cid < 0) {
+		NV_ERROR(drm, "No supported overlay immediate class\n");
+		return cid;
+	}
+
+	return oimms[cid].init(drm, oimms[cid].oclass, wndw);
+}
+
+static const struct nv50_wndw_func
+ovly507e = {
+};
+
+static const u32
+ovly507e_format[] = {
+	0
+};
+
+static int
+ovly507e_new_(const struct nv50_wndw_func *func, const u32 *format,
+	      struct nouveau_drm *drm, int head, s32 oclass,
+	      struct nv50_wndw **pwndw)
+{
+	struct nv50_disp_overlay_channel_dma_v0 args = {
+		.head = head,
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	struct nv50_wndw *wndw;
+	int ret;
+
+	ret = nv50_wndw_new_(func, drm->dev, DRM_PLANE_TYPE_OVERLAY,
+			     "ovly", head, format, &wndw);
+	if (*pwndw = wndw, ret)
+		return ret;
+
+	ret = nv50_dmac_create(&drm->client.device, &disp->disp->object,
+			       &oclass, 0, &args, sizeof(args),
+			       disp->sync->bo.offset, &wndw->wndw);
+	if (ret) {
+		NV_ERROR(drm, "ovly%04x allocation failed: %d\n", oclass, ret);
+		return ret;
+	}
+
+	return 0;
+}
+
+static int
+ovly507e_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return ovly507e_new_(&ovly507e, ovly507e_format, drm, head, oclass, pwndw);
+}
+
+static int
+nv50_ovly_new(struct nouveau_drm *drm, int head, struct nv50_wndw **pwndw)
+{
+	static const struct {
+		s32 oclass;
+		int version;
+		int (*new)(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+	} ovlys[] = {
+		{ GK104_DISP_OVERLAY_CONTROL_DMA, 0, ovly507e_new },
+		{ GF110_DISP_OVERLAY_CONTROL_DMA, 0, ovly507e_new },
+		{ GT214_DISP_OVERLAY_CHANNEL_DMA, 0, ovly507e_new },
+		{ GT200_DISP_OVERLAY_CHANNEL_DMA, 0, ovly507e_new },
+		{   G82_DISP_OVERLAY_CHANNEL_DMA, 0, ovly507e_new },
+		{  NV50_DISP_OVERLAY_CHANNEL_DMA, 0, ovly507e_new },
+		{}
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	int cid, ret;
+
+	cid = nvif_mclass(&disp->disp->object, ovlys);
+	if (cid < 0) {
+		NV_ERROR(drm, "No supported overlay class\n");
+		return cid;
+	}
+
+	ret = ovlys[cid].new(drm, head, ovlys[cid].oclass, pwndw);
+	if (ret)
+		return ret;
+
+	return nv50_oimm_init(drm, *pwndw);
+}
+
 /******************************************************************************
  * Cursor plane
  *****************************************************************************/
@@ -2347,9 +2440,6 @@ nv50_head_destroy(struct drm_crtc *crtc)
 	struct nv50_head *head = nv50_head(crtc);
 	int i;
 
-	nv50_dmac_destroy(&head->ovly.base);
-	nv50_pioc_destroy(&head->oimm.base);
-
 	for (i = 0; i < ARRAY_SIZE(head->lut.nvbo); i++)
 		nouveau_bo_unmap_unpin_unref(&head->lut.nvbo[i]);
 
@@ -2372,11 +2462,10 @@ static int
 nv50_head_create(struct drm_device *dev, int index)
 {
 	struct nouveau_drm *drm = nouveau_drm(dev);
-	struct nvif_device *device = &drm->client.device;
-	struct nv50_disp *disp = nv50_disp(dev);
 	struct nv50_head *head;
 	struct nv50_base *base;
 	struct nv50_curs *curs;
+	struct nv50_wndw *wndw;
 	struct drm_crtc *crtc;
 	int ret, i;
 
@@ -2409,15 +2498,7 @@ nv50_head_create(struct drm_device *dev, int index)
 	}
 
 	/* allocate overlay resources */
-	ret = nv50_oimm_create(device, &disp->disp->object, index, &head->oimm);
-	if (ret)
-		goto out;
-
-	ret = nv50_ovly_create(device, &disp->disp->object, index,
-			       disp->sync->bo.offset, &head->ovly);
-	if (ret)
-		goto out;
-
+	ret = nv50_ovly_new(drm, head->base.index, &wndw);
 out:
 	if (ret)
 		nv50_head_destroy(crtc);
-- 
cgit v1.1


From b97ace4072267ea44a254ef2c3b001d2122313dc Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50-: modify cursor allocation so the code can be
 split

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/disp.c | 136 +++++++++++++++++---------------
 1 file changed, 71 insertions(+), 65 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c
index 26caca2..33cb358 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/disp.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c
@@ -617,12 +617,14 @@ struct nv50_wndw_func {
 	void (*image_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
 	void (*image_clr)(struct nv50_wndw *);
 	void (*lut)(struct nv50_wndw *, struct nv50_wndw_atom *);
-	void (*point)(struct nv50_wndw *, struct nv50_wndw_atom *);
 
 	u32 (*update)(struct nv50_wndw *, u32 interlock);
 };
 
 struct nv50_wimm_func {
+	void (*point)(struct nv50_wndw *, struct nv50_wndw_atom *);
+
+	u32 (*update)(struct nv50_wndw *, u32 interlock);
 };
 
 static void
@@ -728,9 +730,12 @@ nv50_wndw_flush_set(struct nv50_wndw *wndw, u32 interlock,
 	if (asyw->set.ntfy ) wndw->func->ntfy_set (wndw, asyw);
 	if (asyw->set.image) wndw->func->image_set(wndw, asyw);
 	if (asyw->set.lut  ) wndw->func->lut      (wndw, asyw);
-	if (asyw->set.point) wndw->func->point    (wndw, asyw);
+	if (asyw->set.point) {
+		wndw->immd->point(wndw, asyw);
+		wndw->immd->update(wndw, interlock);
+	}
 
-	return wndw->func->update(wndw, interlock);
+	return wndw->func->update ? wndw->func->update(wndw, interlock) : 0;
 }
 
 static void
@@ -1181,28 +1186,26 @@ nv50_ovly_new(struct nouveau_drm *drm, int head, struct nv50_wndw **pwndw)
 /******************************************************************************
  * Cursor plane
  *****************************************************************************/
-#define nv50_curs(p) container_of((p), struct nv50_curs, wndw)
-
-struct nv50_curs {
-	struct nv50_wndw wndw;
-	struct nvif_object chan;
-};
-
 static u32
 nv50_curs_update(struct nv50_wndw *wndw, u32 interlock)
 {
-	struct nv50_curs *curs = nv50_curs(wndw);
-	nvif_wr32(&curs->chan, 0x0080, 0x00000000);
+	nvif_wr32(&wndw->wimm.base.user, 0x0080, 0x00000000);
 	return 0;
 }
 
 static void
 nv50_curs_point(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 {
-	struct nv50_curs *curs = nv50_curs(wndw);
-	nvif_wr32(&curs->chan, 0x0084, (asyw->point.y << 16) | asyw->point.x);
+	nvif_wr32(&wndw->wimm.base.user, 0x0084, (asyw->point.y << 16) |
+						  asyw->point.x);
 }
 
+static const struct nv50_wimm_func
+curs507a = {
+	.point = nv50_curs_point,
+	.update = nv50_curs_update,
+};
+
 static void
 nv50_curs_prepare(struct nv50_wndw *wndw, struct nv50_head_atom *asyh,
 		  struct nv50_wndw_atom *asyw)
@@ -1257,77 +1260,82 @@ nv50_curs_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
 	return 0;
 }
 
-static void *
-nv50_curs_dtor(struct nv50_wndw *wndw)
-{
-	struct nv50_curs *curs = nv50_curs(wndw);
-	nvif_object_fini(&curs->chan);
-	return curs;
-}
-
 static const u32
 nv50_curs_format[] = {
 	DRM_FORMAT_ARGB8888,
+	0
 };
 
 static const struct nv50_wndw_func
 nv50_curs = {
-	.dtor = nv50_curs_dtor,
 	.acquire = nv50_curs_acquire,
 	.release = nv50_curs_release,
 	.prepare = nv50_curs_prepare,
-	.point = nv50_curs_point,
-	.update = nv50_curs_update,
 };
 
 static int
-nv50_curs_new(struct nouveau_drm *drm, struct nv50_head *head,
-	      struct nv50_curs **pcurs)
-{
-	static const struct nvif_mclass curses[] = {
-		{ GK104_DISP_CURSOR, 0 },
-		{ GF110_DISP_CURSOR, 0 },
-		{ GT214_DISP_CURSOR, 0 },
-		{   G82_DISP_CURSOR, 0 },
-		{  NV50_DISP_CURSOR, 0 },
-		{}
-	};
+curs507a_new_(const struct nv50_wimm_func *func, struct nouveau_drm *drm,
+	      int head, s32 oclass, struct nv50_wndw **pwndw)
+{
 	struct nv50_disp_cursor_v0 args = {
-		.head = head->base.index,
+		.head = head,
 	};
 	struct nv50_disp *disp = nv50_disp(drm->dev);
-	struct nv50_curs *curs;
-	int cid, ret;
-
-	cid = nvif_mclass(&disp->disp->object, curses);
-	if (cid < 0) {
-		NV_ERROR(drm, "No supported cursor immediate class\n");
-		return cid;
-	}
-
-	if (!(curs = *pcurs = kzalloc(sizeof(*curs), GFP_KERNEL)))
-		return -ENOMEM;
+	struct nv50_wndw *wndw;
+	int ret;
 
-	ret = nv50_wndw_ctor(&nv50_curs, drm->dev, DRM_PLANE_TYPE_CURSOR,
-			     "curs", head->base.index, &disp->mast.base,
-			     nv50_curs_format, ARRAY_SIZE(nv50_curs_format),
-			     &curs->wndw);
-	if (ret) {
-		kfree(curs);
+	ret = nv50_wndw_new_(&nv50_curs, drm->dev, DRM_PLANE_TYPE_CURSOR,
+			     "curs", head, nv50_curs_format, &wndw);
+	if (*pwndw = wndw, ret)
 		return ret;
-	}
 
-	ret = nvif_object_init(&disp->disp->object, 0, curses[cid].oclass,
-			       &args, sizeof(args), &curs->chan);
+	ret = nvif_object_init(&disp->disp->object, 0, oclass, &args,
+			       sizeof(args), &wndw->wimm.base.user);
 	if (ret) {
-		NV_ERROR(drm, "curs%04x allocation failed: %d\n",
-			 curses[cid].oclass, ret);
+		NV_ERROR(drm, "curs%04x allocation failed: %d\n", oclass, ret);
 		return ret;
 	}
 
+	nvif_object_map(&wndw->wimm.base.user, NULL, 0);
+	wndw->immd = func;
+	wndw->ctxdma.parent = &disp->mast.base.base.user;
 	return 0;
 }
 
+static int
+curs507a_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return curs507a_new_(&curs507a, drm, head, oclass, pwndw);
+}
+
+static int
+nv50_curs_new(struct nouveau_drm *drm, int head, struct nv50_wndw **pwndw)
+{
+	struct {
+		s32 oclass;
+		int version;
+		int (*new)(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+	} curses[] = {
+		{ GK104_DISP_CURSOR, 0, curs507a_new },
+		{ GF110_DISP_CURSOR, 0, curs507a_new },
+		{ GT214_DISP_CURSOR, 0, curs507a_new },
+		{   G82_DISP_CURSOR, 0, curs507a_new },
+		{  NV50_DISP_CURSOR, 0, curs507a_new },
+		{}
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	int cid;
+
+	cid = nvif_mclass(&disp->disp->object, curses);
+	if (cid < 0) {
+		NV_ERROR(drm, "No supported cursor immediate class\n");
+		return cid;
+	}
+
+	return curses[cid].new(drm, head, curses[cid].oclass, pwndw);
+}
+
 /******************************************************************************
  * Primary plane
  *****************************************************************************/
@@ -2464,8 +2472,7 @@ nv50_head_create(struct drm_device *dev, int index)
 	struct nouveau_drm *drm = nouveau_drm(dev);
 	struct nv50_head *head;
 	struct nv50_base *base;
-	struct nv50_curs *curs;
-	struct nv50_wndw *wndw;
+	struct nv50_wndw *curs, *wndw;
 	struct drm_crtc *crtc;
 	int ret, i;
 
@@ -2476,16 +2483,15 @@ nv50_head_create(struct drm_device *dev, int index)
 	head->base.index = index;
 	ret = nv50_base_new(drm, head, &base);
 	if (ret == 0)
-		ret = nv50_curs_new(drm, head, &curs);
+		ret = nv50_curs_new(drm, head->base.index, &curs);
 	if (ret) {
 		kfree(head);
 		return ret;
 	}
 
 	crtc = &head->base.base;
-	drm_crtc_init_with_planes(dev, crtc, &base->wndw.plane,
-				  &curs->wndw.plane, &nv50_head_func,
-				  "head-%d", head->base.index);
+	drm_crtc_init_with_planes(dev, crtc, &base->wndw.plane, &curs->plane,
+				  &nv50_head_func, "head-%d", head->base.index);
 	drm_crtc_helper_add(crtc, &nv50_head_help);
 	drm_mode_crtc_set_gamma_size(crtc, 256);
 
-- 
cgit v1.1


From d7c6e97a32329032ba7af1f53cab2767832fed77 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50-: modify base allocation so the code can be
 split

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/disp.c | 238 ++++++++++++++------------------
 1 file changed, 105 insertions(+), 133 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c
index 33cb358..f0edf9b 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/disp.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c
@@ -468,28 +468,6 @@ struct nv50_sync {
 	u32 data;
 };
 
-static int
-nv50_base_create(struct nvif_device *device, struct nvif_object *disp,
-		 int head, u64 syncbuf, struct nv50_sync *base)
-{
-	struct nv50_disp_base_channel_dma_v0 args = {
-		.head = head,
-	};
-	static const s32 oclass[] = {
-		GK110_DISP_BASE_CHANNEL_DMA,
-		GK104_DISP_BASE_CHANNEL_DMA,
-		GF110_DISP_BASE_CHANNEL_DMA,
-		GT214_DISP_BASE_CHANNEL_DMA,
-		GT200_DISP_BASE_CHANNEL_DMA,
-		G82_DISP_BASE_CHANNEL_DMA,
-		NV50_DISP_BASE_CHANNEL_DMA,
-		0
-	};
-
-	return nv50_dmac_create(device, disp, oclass, head, &args, sizeof(args),
-				syncbuf, &base->base);
-}
-
 struct nv50_head {
 	struct nouveau_crtc base;
 	struct {
@@ -581,7 +559,6 @@ evo_kick(u32 *push, void *evoc)
 struct nv50_wndw {
 	const struct nv50_wndw_func *func;
 	const struct nv50_wimm_func *immd;
-	struct nv50_dmac *dmac;
 	int id;
 
 	struct {
@@ -601,7 +578,6 @@ struct nv50_wndw {
 };
 
 struct nv50_wndw_func {
-	void *(*dtor)(struct nv50_wndw *);
 	int (*acquire)(struct nv50_wndw *, struct nv50_wndw_atom *asyw,
 		       struct nv50_head_atom *asyh);
 	void (*release)(struct nv50_wndw *, struct nv50_wndw_atom *asyw,
@@ -959,19 +935,16 @@ nv50_wndw_destroy(struct drm_plane *plane)
 {
 	struct nv50_wndw *wndw = nv50_wndw(plane);
 	struct nv50_wndw_ctxdma *ctxdma, *ctxtmp;
-	void *data = wndw;
 
 	list_for_each_entry_safe(ctxdma, ctxtmp, &wndw->ctxdma.list, head) {
 		nv50_wndw_ctxdma_del(ctxdma);
 	}
 
 	nvif_notify_fini(&wndw->notify);
-	if (wndw->func->dtor)
-		data = wndw->func->dtor(wndw);
 	nv50_dmac_destroy(&wndw->wimm);
 	nv50_dmac_destroy(&wndw->wndw);
 	drm_plane_cleanup(&wndw->plane);
-	kfree(data);
+	kfree(wndw);
 }
 
 static const struct drm_plane_funcs
@@ -984,6 +957,12 @@ nv50_wndw = {
 	.atomic_destroy_state = nv50_wndw_atomic_destroy_state,
 };
 
+static int
+nv50_wndw_notify(struct nvif_notify *notify)
+{
+	return NVIF_NOTIFY_KEEP;
+}
+
 static void
 nv50_wndw_fini(struct nv50_wndw *wndw)
 {
@@ -997,29 +976,6 @@ nv50_wndw_init(struct nv50_wndw *wndw)
 }
 
 static int
-nv50_wndw_ctor(const struct nv50_wndw_func *func, struct drm_device *dev,
-	       enum drm_plane_type type, const char *name, int index,
-	       struct nv50_dmac *dmac, const u32 *format, int nformat,
-	       struct nv50_wndw *wndw)
-{
-	int ret;
-
-	wndw->func = func;
-	wndw->dmac = dmac;
-	wndw->ctxdma.parent = &dmac->base.user;
-
-	ret = drm_universal_plane_init(dev, &wndw->plane, 0, &nv50_wndw,
-				       format, nformat, NULL,
-				       type, "%s-%d", name, index);
-	if (ret)
-		return ret;
-
-	drm_plane_helper_add(&wndw->plane, &nv50_wndw_helper);
-	INIT_LIST_HEAD(&wndw->ctxdma.list);
-	return 0;
-}
-
-static int
 nv50_wndw_new_(const struct nv50_wndw_func *func, struct drm_device *dev,
 	       enum drm_plane_type type, const char *name, int index,
 	       const u32 *format, struct nv50_wndw **pwndw)
@@ -1030,18 +986,27 @@ nv50_wndw_new_(const struct nv50_wndw_func *func, struct drm_device *dev,
 
 	if (!(wndw = *pwndw = kzalloc(sizeof(*wndw), GFP_KERNEL)))
 		return -ENOMEM;
+	wndw->func = func;
 	wndw->id = index;
 
+	wndw->ctxdma.parent = &wndw->wndw.base.user;
+	INIT_LIST_HEAD(&wndw->ctxdma.list);
+
 	for (nformat = 0; format[nformat]; nformat++);
 
-	ret = nv50_wndw_ctor(func, dev, type, name, index,
-			     &wndw->wndw, format, nformat, wndw);
+	ret = drm_universal_plane_init(dev, &wndw->plane, 0, &nv50_wndw,
+				       format, nformat, NULL,
+				       type, "%s-%d", name, index);
 	if (ret) {
 		kfree(*pwndw);
 		*pwndw = NULL;
+		return ret;
 	}
 
-	return ret;
+	drm_plane_helper_add(&wndw->plane, &nv50_wndw_helper);
+
+	wndw->notify.func = nv50_wndw_notify;
+	return 0;
 }
 
 /******************************************************************************
@@ -1339,53 +1304,36 @@ nv50_curs_new(struct nouveau_drm *drm, int head, struct nv50_wndw **pwndw)
 /******************************************************************************
  * Primary plane
  *****************************************************************************/
-#define nv50_base(p) container_of((p), struct nv50_base, wndw)
-
-struct nv50_base {
-	struct nv50_wndw wndw;
-	struct nv50_sync chan;
-	int id;
-};
-
-static int
-nv50_base_notify(struct nvif_notify *notify)
-{
-	return NVIF_NOTIFY_KEEP;
-}
-
 static void
 nv50_base_lut(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 {
-	struct nv50_base *base = nv50_base(wndw);
 	u32 *push;
-	if ((push = evo_wait(&base->chan, 2))) {
+	if ((push = evo_wait(&wndw->wndw, 2))) {
 		evo_mthd(push, 0x00e0, 1);
 		evo_data(push, asyw->lut.enable << 30);
-		evo_kick(push, &base->chan);
+		evo_kick(push, &wndw->wndw);
 	}
 }
 
 static void
 nv50_base_image_clr(struct nv50_wndw *wndw)
 {
-	struct nv50_base *base = nv50_base(wndw);
 	u32 *push;
-	if ((push = evo_wait(&base->chan, 4))) {
+	if ((push = evo_wait(&wndw->wndw, 4))) {
 		evo_mthd(push, 0x0084, 1);
 		evo_data(push, 0x00000000);
 		evo_mthd(push, 0x00c0, 1);
 		evo_data(push, 0x00000000);
-		evo_kick(push, &base->chan);
+		evo_kick(push, &wndw->wndw);
 	}
 }
 
 static void
 nv50_base_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 {
-	struct nv50_base *base = nv50_base(wndw);
-	const s32 oclass = base->chan.base.base.user.oclass;
+	const s32 oclass = wndw->wndw.base.user.oclass;
 	u32 *push;
-	if ((push = evo_wait(&base->chan, 10))) {
+	if ((push = evo_wait(&wndw->wndw, 10))) {
 		evo_mthd(push, 0x0084, 1);
 		evo_data(push, (asyw->image.mode << 8) |
 			       (asyw->image.interval << 4));
@@ -1421,77 +1369,72 @@ nv50_base_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 					asyw->image.block);
 			evo_data(push, asyw->image.format << 8);
 		}
-		evo_kick(push, &base->chan);
+		evo_kick(push, &wndw->wndw);
 	}
 }
 
 static void
 nv50_base_ntfy_clr(struct nv50_wndw *wndw)
 {
-	struct nv50_base *base = nv50_base(wndw);
 	u32 *push;
-	if ((push = evo_wait(&base->chan, 2))) {
+	if ((push = evo_wait(&wndw->wndw, 2))) {
 		evo_mthd(push, 0x00a4, 1);
 		evo_data(push, 0x00000000);
-		evo_kick(push, &base->chan);
+		evo_kick(push, &wndw->wndw);
 	}
 }
 
 static void
 nv50_base_ntfy_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 {
-	struct nv50_base *base = nv50_base(wndw);
 	u32 *push;
-	if ((push = evo_wait(&base->chan, 3))) {
+	if ((push = evo_wait(&wndw->wndw, 3))) {
 		evo_mthd(push, 0x00a0, 2);
 		evo_data(push, (asyw->ntfy.awaken << 30) | asyw->ntfy.offset);
 		evo_data(push, asyw->ntfy.handle);
-		evo_kick(push, &base->chan);
+		evo_kick(push, &wndw->wndw);
 	}
 }
 
 static void
 nv50_base_sema_clr(struct nv50_wndw *wndw)
 {
-	struct nv50_base *base = nv50_base(wndw);
 	u32 *push;
-	if ((push = evo_wait(&base->chan, 2))) {
+	if ((push = evo_wait(&wndw->wndw, 2))) {
 		evo_mthd(push, 0x0094, 1);
 		evo_data(push, 0x00000000);
-		evo_kick(push, &base->chan);
+		evo_kick(push, &wndw->wndw);
 	}
 }
 
 static void
 nv50_base_sema_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 {
-	struct nv50_base *base = nv50_base(wndw);
 	u32 *push;
-	if ((push = evo_wait(&base->chan, 5))) {
+	if ((push = evo_wait(&wndw->wndw, 5))) {
 		evo_mthd(push, 0x0088, 4);
 		evo_data(push, asyw->sema.offset);
 		evo_data(push, asyw->sema.acquire);
 		evo_data(push, asyw->sema.release);
 		evo_data(push, asyw->sema.handle);
-		evo_kick(push, &base->chan);
+		evo_kick(push, &wndw->wndw);
 	}
 }
 
 static u32
 nv50_base_update(struct nv50_wndw *wndw, u32 interlock)
 {
-	struct nv50_base *base = nv50_base(wndw);
 	u32 *push;
 
-	if (!(push = evo_wait(&base->chan, 2)))
+	if (!(push = evo_wait(&wndw->wndw, 2)))
 		return 0;
 	evo_mthd(push, 0x0080, 1);
 	evo_data(push, interlock);
-	evo_kick(push, &base->chan);
+	evo_kick(push, &wndw->wndw);
 
-	if (base->chan.base.base.user.oclass < GF110_DISP_BASE_CHANNEL_DMA)
-		return interlock ? 2 << (base->id * 8) : 0;
-	return interlock ? 2 << (base->id * 4) : 0;
+	if (wndw->wndw.base.user.oclass < GF110_DISP_BASE_CHANNEL_DMA)
+		return interlock ? 2 << (wndw->id * 8) : 0;
+	return interlock ? 2 << (wndw->id * 4) : 0;
 }
 
 static int
@@ -1561,14 +1504,6 @@ nv50_base_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
 	return 0;
 }
 
-static void *
-nv50_base_dtor(struct nv50_wndw *wndw)
-{
-	struct nv50_base *base = nv50_base(wndw);
-	nv50_dmac_destroy(&base->chan.base);
-	return base;
-}
-
 static const u32
 nv50_base_format[] = {
 	DRM_FORMAT_C8,
@@ -1581,11 +1516,11 @@ nv50_base_format[] = {
 	DRM_FORMAT_ABGR2101010,
 	DRM_FORMAT_XBGR8888,
 	DRM_FORMAT_ABGR8888,
+	0
 };
 
 static const struct nv50_wndw_func
 nv50_base = {
-	.dtor = nv50_base_dtor,
 	.acquire = nv50_base_acquire,
 	.release = nv50_base_release,
 	.sema_set = nv50_base_sema_set,
@@ -1600,41 +1535,79 @@ nv50_base = {
 };
 
 static int
-nv50_base_new(struct nouveau_drm *drm, struct nv50_head *head,
-	      struct nv50_base **pbase)
+base507c_new_(const struct nv50_wndw_func *func, const u32 *format,
+	      struct nouveau_drm *drm, int head, s32 oclass,
+	      struct nv50_wndw **pwndw)
 {
+	struct nv50_disp_base_channel_dma_v0 args = {
+		.head = head,
+	};
 	struct nv50_disp *disp = nv50_disp(drm->dev);
-	struct nv50_base *base;
+	struct nv50_wndw *wndw;
 	int ret;
 
-	if (!(base = *pbase = kzalloc(sizeof(*base), GFP_KERNEL)))
-		return -ENOMEM;
-	base->id = head->base.index;
-	base->wndw.ntfy = EVO_FLIP_NTFY0(base->id);
-	base->wndw.sema = EVO_FLIP_SEM0(base->id);
-	base->wndw.data = 0x00000000;
-
-	ret = nv50_wndw_ctor(&nv50_base, drm->dev, DRM_PLANE_TYPE_PRIMARY,
-			     "base", base->id, &base->chan.base,
-			     nv50_base_format, ARRAY_SIZE(nv50_base_format),
-			     &base->wndw);
+	ret = nv50_wndw_new_(func, drm->dev, DRM_PLANE_TYPE_PRIMARY,
+			     "base", head, format, &wndw);
+	if (*pwndw = wndw, ret)
+		return ret;
+
+	ret = nv50_dmac_create(&drm->client.device, &disp->disp->object,
+			       &oclass, head, &args, sizeof(args),
+			       disp->sync->bo.offset, &wndw->wndw);
 	if (ret) {
-		kfree(base);
+		NV_ERROR(drm, "base%04x allocation failed: %d\n", oclass, ret);
 		return ret;
 	}
 
-	ret = nv50_base_create(&drm->client.device, &disp->disp->object,
-			       base->id, disp->sync->bo.offset, &base->chan);
+	ret = nvif_notify_init(&wndw->wndw.base.user, wndw->notify.func,
+			       false, NV50_DISP_BASE_CHANNEL_DMA_V0_NTFY_UEVENT,
+			       &(struct nvif_notify_uevent_req) {},
+			       sizeof(struct nvif_notify_uevent_req),
+			       sizeof(struct nvif_notify_uevent_rep),
+			       &wndw->notify);
 	if (ret)
 		return ret;
 
-	return nvif_notify_init(&base->chan.base.base.user, nv50_base_notify,
-				false,
-				NV50_DISP_BASE_CHANNEL_DMA_V0_NTFY_UEVENT,
-				&(struct nvif_notify_uevent_req) {},
-				sizeof(struct nvif_notify_uevent_req),
-				sizeof(struct nvif_notify_uevent_rep),
-				&base->wndw.notify);
+	wndw->ntfy = EVO_FLIP_NTFY0(wndw->id);
+	wndw->sema = EVO_FLIP_SEM0(wndw->id);
+	wndw->data = 0x00000000;
+	return 0;
+}
+
+static int
+base507c_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return base507c_new_(&nv50_base, nv50_base_format, drm, head, oclass, pwndw);
+}
+
+static int
+nv50_base_new(struct nouveau_drm *drm, int head, struct nv50_wndw **pwndw)
+{
+	struct {
+		s32 oclass;
+		int version;
+		int (*new)(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+	} bases[] = {
+		{ GK110_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
+		{ GK104_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
+		{ GF110_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
+		{ GT214_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
+		{ GT200_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
+		{   G82_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
+		{  NV50_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
+		{}
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	int cid;
+
+	cid = nvif_mclass(&disp->disp->object, bases);
+	if (cid < 0) {
+		NV_ERROR(drm, "No supported base class\n");
+		return cid;
+	}
+
+	return bases[cid].new(drm, head, bases[cid].oclass, pwndw);
 }
 
 /******************************************************************************
@@ -2471,7 +2444,6 @@ nv50_head_create(struct drm_device *dev, int index)
 {
 	struct nouveau_drm *drm = nouveau_drm(dev);
 	struct nv50_head *head;
-	struct nv50_base *base;
 	struct nv50_wndw *curs, *wndw;
 	struct drm_crtc *crtc;
 	int ret, i;
@@ -2481,7 +2453,7 @@ nv50_head_create(struct drm_device *dev, int index)
 		return -ENOMEM;
 
 	head->base.index = index;
-	ret = nv50_base_new(drm, head, &base);
+	ret = nv50_base_new(drm, head->base.index, &wndw);
 	if (ret == 0)
 		ret = nv50_curs_new(drm, head->base.index, &curs);
 	if (ret) {
@@ -2490,7 +2462,7 @@ nv50_head_create(struct drm_device *dev, int index)
 	}
 
 	crtc = &head->base.base;
-	drm_crtc_init_with_planes(dev, crtc, &base->wndw.plane, &curs->plane,
+	drm_crtc_init_with_planes(dev, crtc, &wndw->plane, &curs->plane,
 				  &nv50_head_func, "head-%d", head->base.index);
 	drm_crtc_helper_add(crtc, &nv50_head_help);
 	drm_mode_crtc_set_gamma_size(crtc, 256);
@@ -4256,7 +4228,7 @@ nv50_disp_atomic_commit(struct drm_device *dev,
 		struct nv50_wndw *wndw = nv50_wndw(plane);
 
 		if (asyw->set.image) {
-			asyw->ntfy.handle = wndw->dmac->sync.handle;
+			asyw->ntfy.handle = wndw->wndw.sync.handle;
 			asyw->ntfy.offset = wndw->ntfy;
 			asyw->ntfy.awaken = false;
 			asyw->set.ntfy = true;
-- 
cgit v1.1


From 9ca6f1ebba10240ad02f7c659481899a28220fbc Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50: modify core allocation so the code can be split

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/disp.c | 215 ++++++++++++++++++++------------
 1 file changed, 134 insertions(+), 81 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c
index f0edf9b..abdf39e 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/disp.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c
@@ -426,39 +426,6 @@ nv50_dmac_create(struct nvif_device *device, struct nvif_object *disp,
 }
 
 /******************************************************************************
- * Core
- *****************************************************************************/
-
-struct nv50_mast {
-	struct nv50_dmac base;
-};
-
-static int
-nv50_core_create(struct nvif_device *device, struct nvif_object *disp,
-		 u64 syncbuf, struct nv50_mast *core)
-{
-	struct nv50_disp_core_channel_dma_v0 args = {};
-	static const s32 oclass[] = {
-		GP102_DISP_CORE_CHANNEL_DMA,
-		GP100_DISP_CORE_CHANNEL_DMA,
-		GM200_DISP_CORE_CHANNEL_DMA,
-		GM107_DISP_CORE_CHANNEL_DMA,
-		GK110_DISP_CORE_CHANNEL_DMA,
-		GK104_DISP_CORE_CHANNEL_DMA,
-		GF110_DISP_CORE_CHANNEL_DMA,
-		GT214_DISP_CORE_CHANNEL_DMA,
-		GT206_DISP_CORE_CHANNEL_DMA,
-		GT200_DISP_CORE_CHANNEL_DMA,
-		G82_DISP_CORE_CHANNEL_DMA,
-		NV50_DISP_CORE_CHANNEL_DMA,
-		0
-	};
-
-	return nv50_dmac_create(device, disp, oclass, 0, &args, sizeof(args),
-				syncbuf, &core->base);
-}
-
-/******************************************************************************
  * Base
  *****************************************************************************/
 
@@ -477,14 +444,10 @@ struct nv50_head {
 };
 
 #define nv50_head(c) ((struct nv50_head *)nouveau_crtc(c))
-#define nv50_ovly(c) (&nv50_head(c)->ovly)
-#define nv50_oimm(c) (&nv50_head(c)->oimm)
-#define nv50_chan(c) (&(c)->base.base)
-#define nv50_vers(c) nv50_chan(c)->user.oclass
 
 struct nv50_disp {
 	struct nvif_disp *disp;
-	struct nv50_mast mast;
+	struct nv50_core *core;
 
 	struct nouveau_bo *sync;
 
@@ -497,7 +460,41 @@ nv50_disp(struct drm_device *dev)
 	return nouveau_display(dev)->priv;
 }
 
-#define nv50_mast(d) (&nv50_disp(d)->mast)
+/******************************************************************************
+ * Core
+ *****************************************************************************/
+
+struct nv50_core {
+	const struct nv50_core_func *func;
+	struct nv50_dmac chan;
+};
+
+struct nv50_core_func {
+};
+
+static int
+core507d_new_(const struct nv50_core_func *func, struct nouveau_drm *drm,
+	      s32 oclass, struct nv50_core **pcore)
+{
+	struct nv50_disp_core_channel_dma_v0 args = {};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	struct nv50_core *core;
+	int ret;
+
+	if (!(core = *pcore = kzalloc(sizeof(*core), GFP_KERNEL)))
+		return -ENOMEM;
+	core->func = func;
+
+	ret = nv50_dmac_create(&drm->client.device, &disp->disp->object,
+			       &oclass, 0, &args, sizeof(args),
+			       disp->sync->bo.offset, &core->chan);
+	if (ret) {
+		NV_ERROR(drm, "core%04x allocation failed: %d\n", oclass, ret);
+		return ret;
+	}
+
+	return 0;
+}
 
 /******************************************************************************
  * EVO channel helpers
@@ -1175,7 +1172,7 @@ static void
 nv50_curs_prepare(struct nv50_wndw *wndw, struct nv50_head_atom *asyh,
 		  struct nv50_wndw_atom *asyw)
 {
-	u32 handle = nv50_disp(wndw->plane.dev)->mast.base.vram.handle;
+	u32 handle = nv50_disp(wndw->plane.dev)->core->chan.vram.handle;
 	u32 offset = asyw->image.offset;
 	if (asyh->curs.handle != handle || asyh->curs.offset != offset) {
 		asyh->curs.handle = handle;
@@ -1263,7 +1260,7 @@ curs507a_new_(const struct nv50_wimm_func *func, struct nouveau_drm *drm,
 
 	nvif_object_map(&wndw->wimm.base.user, NULL, 0);
 	wndw->immd = func;
-	wndw->ctxdma.parent = &disp->mast.base.base.user;
+	wndw->ctxdma.parent = &disp->core->chan.base.user;
 	return 0;
 }
 
@@ -1616,7 +1613,7 @@ nv50_base_new(struct nouveau_drm *drm, int head, struct nv50_wndw **pwndw)
 static void
 nv50_head_procamp(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
 	if ((push = evo_wait(core, 2))) {
 		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
@@ -1632,7 +1629,7 @@ nv50_head_procamp(struct nv50_head *head, struct nv50_head_atom *asyh)
 static void
 nv50_head_dither(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
 	if ((push = evo_wait(core, 2))) {
 		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
@@ -1652,7 +1649,7 @@ nv50_head_dither(struct nv50_head *head, struct nv50_head_atom *asyh)
 static void
 nv50_head_ovly(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 bounds = 0;
 	u32 *push;
 
@@ -1681,7 +1678,7 @@ nv50_head_ovly(struct nv50_head *head, struct nv50_head_atom *asyh)
 static void
 nv50_head_base(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 bounds = 0;
 	u32 *push;
 
@@ -1711,7 +1708,7 @@ nv50_head_base(struct nv50_head *head, struct nv50_head_atom *asyh)
 static void
 nv50_head_curs_clr(struct nv50_head *head)
 {
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
 	if ((push = evo_wait(core, 4))) {
 		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
@@ -1736,7 +1733,7 @@ nv50_head_curs_clr(struct nv50_head *head)
 static void
 nv50_head_curs_set(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
 	if ((push = evo_wait(core, 5))) {
 		if (core->base.user.oclass < G82_DISP_BASE_CHANNEL_DMA) {
@@ -1767,7 +1764,7 @@ nv50_head_curs_set(struct nv50_head *head, struct nv50_head_atom *asyh)
 static void
 nv50_head_core_clr(struct nv50_head *head)
 {
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
 	if ((push = evo_wait(core, 2))) {
 		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
@@ -1782,7 +1779,7 @@ nv50_head_core_clr(struct nv50_head *head)
 static void
 nv50_head_core_set(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
 	if ((push = evo_wait(core, 9))) {
 		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
@@ -1836,7 +1833,7 @@ nv50_head_core_set(struct nv50_head *head, struct nv50_head_atom *asyh)
 static void
 nv50_head_lut_clr(struct nv50_head *head)
 {
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
 	if ((push = evo_wait(core, 4))) {
 		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
@@ -1909,7 +1906,7 @@ nv50_head_lut_load(struct drm_property_blob *blob, int mode,
 static void
 nv50_head_lut_set(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
 	if ((push = evo_wait(core, 7))) {
 		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
@@ -1939,7 +1936,7 @@ nv50_head_lut_set(struct nv50_head *head, struct nv50_head_atom *asyh)
 static void
 nv50_head_mode(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	struct nv50_head_mode *m = &asyh->mode;
 	u32 *push;
 	if ((push = evo_wait(core, 14))) {
@@ -1980,7 +1977,7 @@ nv50_head_mode(struct nv50_head *head, struct nv50_head_atom *asyh)
 static void
 nv50_head_view(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
 	if ((push = evo_wait(core, 10))) {
 		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
@@ -2191,7 +2188,7 @@ nv50_head_atomic_check_lut(struct nv50_head *head,
 		asyh->lut.mode = 7;
 		asyh->set.ilut = asyh->state.color_mgmt_changed;
 	}
-	asyh->lut.handle = disp->mast.base.vram.handle;
+	asyh->lut.handle = disp->core->chan.vram.handle;
 }
 
 static void
@@ -2311,7 +2308,7 @@ nv50_head_atomic_check(struct drm_crtc *crtc, struct drm_crtc_state *state)
 			asyh->core.w = asyh->state.mode.hdisplay;
 			asyh->core.h = asyh->state.mode.vdisplay;
 		}
-		asyh->core.handle = disp->mast.base.vram.handle;
+		asyh->core.handle = disp->core->chan.vram.handle;
 		asyh->core.offset = 0;
 		asyh->core.format = 0xcf;
 		asyh->core.kind = 0;
@@ -2483,6 +2480,61 @@ out:
 	return ret;
 }
 
+static const struct nv50_core_func
+core507d = {
+};
+
+static int
+core507d_new(struct nouveau_drm *drm, s32 oclass, struct nv50_core **pcore)
+{
+	return core507d_new_(&core507d, drm, oclass, pcore);
+}
+
+static void
+nv50_core_del(struct nv50_core **pcore)
+{
+	struct nv50_core *core = *pcore;
+	if (core) {
+		nv50_dmac_destroy(&core->chan);
+		kfree(*pcore);
+		*pcore = NULL;
+	}
+}
+
+static int
+nv50_core_new(struct nouveau_drm *drm, struct nv50_core **pcore)
+{
+	struct {
+		s32 oclass;
+		int version;
+		int (*new)(struct nouveau_drm *, s32, struct nv50_core **);
+	} cores[] = {
+		{ GP102_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{ GP100_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{ GM200_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{ GM107_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{ GK110_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{ GK104_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{ GF110_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{ GT214_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{ GT206_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{ GT200_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{   G82_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{  NV50_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{}
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	int cid;
+
+	cid = nvif_mclass(&disp->disp->object, cores);
+	if (cid < 0) {
+		NV_ERROR(drm, "No supported core channel class\n");
+		return cid;
+	}
+
+	return cores[cid].new(drm, cores[cid].oclass, pcore);
+}
+
 /******************************************************************************
  * Output path helpers
  *****************************************************************************/
@@ -2591,21 +2643,21 @@ static void
 nv50_dac_disable(struct drm_encoder *encoder)
 {
 	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nv50_mast *mast = nv50_mast(encoder->dev);
+	struct nv50_dmac *core = &nv50_disp(encoder->dev)->core->chan;
 	const int or = nv_encoder->or;
 	u32 *push;
 
 	if (nv_encoder->crtc) {
-		push = evo_wait(mast, 4);
+		push = evo_wait(core, 4);
 		if (push) {
-			if (nv50_vers(mast) < GF110_DISP_CORE_CHANNEL_DMA) {
+			if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
 				evo_mthd(push, 0x0400 + (or * 0x080), 1);
 				evo_data(push, 0x00000000);
 			} else {
 				evo_mthd(push, 0x0180 + (or * 0x020), 1);
 				evo_data(push, 0x00000000);
 			}
-			evo_kick(push, mast);
+			evo_kick(push, core);
 		}
 	}
 
@@ -2616,7 +2668,7 @@ nv50_dac_disable(struct drm_encoder *encoder)
 static void
 nv50_dac_enable(struct drm_encoder *encoder)
 {
-	struct nv50_mast *mast = nv50_mast(encoder->dev);
+	struct nv50_dmac *core = &nv50_disp(encoder->dev)->core->chan;
 	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
 	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
 	struct drm_display_mode *mode = &nv_crtc->base.state->adjusted_mode;
@@ -2624,9 +2676,9 @@ nv50_dac_enable(struct drm_encoder *encoder)
 
 	nv50_outp_acquire(nv_encoder);
 
-	push = evo_wait(mast, 8);
+	push = evo_wait(core, 8);
 	if (push) {
-		if (nv50_vers(mast) < GF110_DISP_CORE_CHANNEL_DMA) {
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
 			u32 syncs = 0x00000000;
 
 			if (mode->flags & DRM_MODE_FLAG_NHSYNC)
@@ -2656,7 +2708,7 @@ nv50_dac_enable(struct drm_encoder *encoder)
 			evo_data(push, 1 << nv_crtc->index);
 		}
 
-		evo_kick(push, mast);
+		evo_kick(push, core);
 	}
 
 	nv_encoder->crtc = encoder->crtc;
@@ -3526,7 +3578,8 @@ static void
 nv50_sor_update(struct nouveau_encoder *nv_encoder, u8 head,
 		struct drm_display_mode *mode, u8 proto, u8 depth)
 {
-	struct nv50_dmac *core = &nv50_mast(nv_encoder->base.base.dev)->base;
+	struct nv50_disp *disp = nv50_disp(nv_encoder->base.base.dev);
+	struct nv50_dmac *core = &disp->core->chan;
 	u32 *push;
 
 	if (!mode) {
@@ -3809,18 +3862,18 @@ static void
 nv50_pior_disable(struct drm_encoder *encoder)
 {
 	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nv50_mast *mast = nv50_mast(encoder->dev);
+	struct nv50_dmac *core = &nv50_disp(encoder->dev)->core->chan;
 	const int or = nv_encoder->or;
 	u32 *push;
 
 	if (nv_encoder->crtc) {
-		push = evo_wait(mast, 4);
+		push = evo_wait(core, 4);
 		if (push) {
-			if (nv50_vers(mast) < GF110_DISP_CORE_CHANNEL_DMA) {
+			if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
 				evo_mthd(push, 0x0700 + (or * 0x040), 1);
 				evo_data(push, 0x00000000);
 			}
-			evo_kick(push, mast);
+			evo_kick(push, core);
 		}
 	}
 
@@ -3831,7 +3884,7 @@ nv50_pior_disable(struct drm_encoder *encoder)
 static void
 nv50_pior_enable(struct drm_encoder *encoder)
 {
-	struct nv50_mast *mast = nv50_mast(encoder->dev);
+	struct nv50_dmac *core = &nv50_disp(encoder->dev)->core->chan;
 	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
 	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
 	struct nouveau_connector *nv_connector;
@@ -3860,9 +3913,9 @@ nv50_pior_enable(struct drm_encoder *encoder)
 		break;
 	}
 
-	push = evo_wait(mast, 8);
+	push = evo_wait(core, 8);
 	if (push) {
-		if (nv50_vers(mast) < GF110_DISP_CORE_CHANNEL_DMA) {
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
 			u32 ctrl = (depth << 16) | (proto << 8) | owner;
 			if (mode->flags & DRM_MODE_FLAG_NHSYNC)
 				ctrl |= 0x00001000;
@@ -3872,7 +3925,7 @@ nv50_pior_enable(struct drm_encoder *encoder)
 			evo_data(push, ctrl);
 		}
 
-		evo_kick(push, mast);
+		evo_kick(push, core);
 	}
 
 	nv_encoder->crtc = encoder->crtc;
@@ -3950,7 +4003,7 @@ static void
 nv50_disp_atomic_commit_core(struct nouveau_drm *drm, u32 interlock)
 {
 	struct nv50_disp *disp = nv50_disp(drm->dev);
-	struct nv50_dmac *core = &disp->mast.base;
+	struct nv50_dmac *core = &disp->core->chan;
 	struct nv50_mstm *mstm;
 	struct drm_encoder *encoder;
 	u32 *push;
@@ -4134,11 +4187,11 @@ nv50_disp_atomic_commit_tail(struct drm_atomic_state *state)
 	/* Flush update. */
 	if (interlock_core) {
 		if (!interlock_chan && atom->state.legacy_cursor_update) {
-			u32 *push = evo_wait(&disp->mast, 2);
+			u32 *push = evo_wait(&disp->core->chan, 2);
 			if (push) {
 				evo_mthd(push, 0x0080, 1);
 				evo_data(push, 0x00000000);
-				evo_kick(push, &disp->mast);
+				evo_kick(push, &disp->core->chan);
 			}
 		} else {
 			nv50_disp_atomic_commit_core(drm, interlock_chan);
@@ -4442,17 +4495,18 @@ nv50_display_fini(struct drm_device *dev)
 int
 nv50_display_init(struct drm_device *dev)
 {
+	struct nv50_dmac *core = &nv50_disp(dev)->core->chan;
 	struct drm_encoder *encoder;
 	struct drm_plane *plane;
 	u32 *push;
 
-	push = evo_wait(nv50_mast(dev), 32);
+	push = evo_wait(core, 32);
 	if (!push)
 		return -EBUSY;
 
 	evo_mthd(push, 0x0088, 1);
-	evo_data(push, nv50_mast(dev)->base.sync.handle);
-	evo_kick(push, nv50_mast(dev));
+	evo_data(push, core->sync.handle);
+	evo_kick(push, core);
 
 	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
 		if (encoder->encoder_type != DRM_MODE_ENCODER_DPMST) {
@@ -4477,7 +4531,7 @@ nv50_display_destroy(struct drm_device *dev)
 {
 	struct nv50_disp *disp = nv50_disp(dev);
 
-	nv50_dmac_destroy(&disp->mast.base);
+	nv50_core_del(&disp->core);
 
 	nouveau_bo_unmap(disp->sync);
 	if (disp->sync)
@@ -4537,8 +4591,7 @@ nv50_display_create(struct drm_device *dev)
 		goto out;
 
 	/* allocate master evo channel */
-	ret = nv50_core_create(device, &disp->disp->object,
-			       disp->sync->bo.offset, &disp->mast);
+	ret = nv50_core_new(drm, &disp->core);
 	if (ret)
 		goto out;
 
-- 
cgit v1.1


From 10ffe0fad53308ff54da0c6b1c5befca4e6915a1 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50-: abstract head interfaces so the code can be
 split

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/disp.c | 64 +++++++++++++++++++++++++--------
 1 file changed, 50 insertions(+), 14 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c
index abdf39e..a23a33d 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/disp.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c
@@ -436,6 +436,7 @@ struct nv50_sync {
 };
 
 struct nv50_head {
+	const struct nv50_head_func *func;
 	struct nouveau_crtc base;
 	struct {
 		struct nouveau_bo *nvbo[2];
@@ -443,7 +444,22 @@ struct nv50_head {
 	} lut;
 };
 
-#define nv50_head(c) ((struct nv50_head *)nouveau_crtc(c))
+struct nv50_head_func {
+	void (*view)(struct nv50_head *, struct nv50_head_atom *);
+	void (*mode)(struct nv50_head *, struct nv50_head_atom *);
+	void (*ilut_set)(struct nv50_head *, struct nv50_head_atom *);
+	void (*ilut_clr)(struct nv50_head *);
+	void (*core_set)(struct nv50_head *, struct nv50_head_atom *);
+	void (*core_clr)(struct nv50_head *);
+	void (*curs_set)(struct nv50_head *, struct nv50_head_atom *);
+	void (*curs_clr)(struct nv50_head *);
+	void (*base)(struct nv50_head *, struct nv50_head_atom *);
+	void (*ovly)(struct nv50_head *, struct nv50_head_atom *);
+	void (*dither)(struct nv50_head *, struct nv50_head_atom *);
+	void (*procamp)(struct nv50_head *, struct nv50_head_atom *);
+};
+
+#define nv50_head(c) container_of((c), struct nv50_head, base.base)
 
 struct nv50_disp {
 	struct nvif_disp *disp;
@@ -470,6 +486,7 @@ struct nv50_core {
 };
 
 struct nv50_core_func {
+	const struct nv50_head_func *head;
 };
 
 static int
@@ -2002,22 +2019,38 @@ nv50_head_view(struct nv50_head *head, struct nv50_head_atom *asyh)
 	}
 }
 
+static const struct nv50_head_func
+head507d = {
+	.view = nv50_head_view,
+	.mode = nv50_head_mode,
+	.ilut_set = nv50_head_lut_set,
+	.ilut_clr = nv50_head_lut_clr,
+	.core_set = nv50_head_core_set,
+	.core_clr = nv50_head_core_clr,
+	.curs_set = nv50_head_curs_set,
+	.curs_clr = nv50_head_curs_clr,
+	.base = nv50_head_base,
+	.ovly = nv50_head_ovly,
+	.dither = nv50_head_dither,
+	.procamp = nv50_head_procamp,
+};
+
 static void
 nv50_head_flush_clr(struct nv50_head *head, struct nv50_head_atom *asyh, bool y)
 {
 	if (asyh->clr.ilut && (!asyh->set.ilut || y))
-		nv50_head_lut_clr(head);
+		head->func->ilut_clr(head);
 	if (asyh->clr.core && (!asyh->set.core || y))
-		nv50_head_core_clr(head);
+		head->func->core_clr(head);
 	if (asyh->clr.curs && (!asyh->set.curs || y))
-		nv50_head_curs_clr(head);
+		head->func->curs_clr(head);
 }
 
 static void
 nv50_head_flush_set(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
-	if (asyh->set.view   ) nv50_head_view    (head, asyh);
-	if (asyh->set.mode   ) nv50_head_mode    (head, asyh);
+	if (asyh->set.view   ) head->func->view    (head, asyh);
+	if (asyh->set.mode   ) head->func->mode    (head, asyh);
 	if (asyh->set.ilut   ) {
 		struct nouveau_bo *nvbo = head->lut.nvbo[head->lut.next];
 		struct drm_property_blob *blob = asyh->state.gamma_lut;
@@ -2025,14 +2058,14 @@ nv50_head_flush_set(struct nv50_head *head, struct nv50_head_atom *asyh)
 			nv50_head_lut_load(blob, asyh->lut.mode, nvbo);
 		asyh->lut.offset = nvbo->bo.offset;
 		head->lut.next ^= 1;
-		nv50_head_lut_set(head, asyh);
+		head->func->ilut_set(head, asyh);
 	}
-	if (asyh->set.core   ) nv50_head_core_set(head, asyh);
-	if (asyh->set.curs   ) nv50_head_curs_set(head, asyh);
-	if (asyh->set.base   ) nv50_head_base    (head, asyh);
-	if (asyh->set.ovly   ) nv50_head_ovly    (head, asyh);
-	if (asyh->set.dither ) nv50_head_dither  (head, asyh);
-	if (asyh->set.procamp) nv50_head_procamp (head, asyh);
+	if (asyh->set.core   ) head->func->core_set(head, asyh);
+	if (asyh->set.curs   ) head->func->curs_set(head, asyh);
+	if (asyh->set.base   ) head->func->base    (head, asyh);
+	if (asyh->set.ovly   ) head->func->ovly    (head, asyh);
+	if (asyh->set.dither ) head->func->dither  (head, asyh);
+	if (asyh->set.procamp) head->func->procamp (head, asyh);
 }
 
 static void
@@ -2422,7 +2455,7 @@ nv50_head_destroy(struct drm_crtc *crtc)
 		nouveau_bo_unmap_unpin_unref(&head->lut.nvbo[i]);
 
 	drm_crtc_cleanup(crtc);
-	kfree(crtc);
+	kfree(head);
 }
 
 static const struct drm_crtc_funcs
@@ -2440,6 +2473,7 @@ static int
 nv50_head_create(struct drm_device *dev, int index)
 {
 	struct nouveau_drm *drm = nouveau_drm(dev);
+	struct nv50_disp *disp = nv50_disp(dev);
 	struct nv50_head *head;
 	struct nv50_wndw *curs, *wndw;
 	struct drm_crtc *crtc;
@@ -2449,6 +2483,7 @@ nv50_head_create(struct drm_device *dev, int index)
 	if (!head)
 		return -ENOMEM;
 
+	head->func = disp->core->func->head;
 	head->base.index = index;
 	ret = nv50_base_new(drm, head->base.index, &wndw);
 	if (ret == 0)
@@ -2482,6 +2517,7 @@ out:
 
 static const struct nv50_core_func
 core507d = {
+	.head = &head507d,
 };
 
 static int
-- 
cgit v1.1


From 2ca7fb5c1cc69ee7fc1a3c048c6f2b75cf842df9 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50: handle SetControlOutputResource from head

Removes duplicated code from OR-specific functions.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/disp.c   | 115 ++++++++++++++----------------
 drivers/gpu/drm/nouveau/nouveau_encoder.h |   3 +-
 2 files changed, 57 insertions(+), 61 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c
index a23a33d..f22c637 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/disp.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c
@@ -194,6 +194,12 @@ struct nv50_head_atom {
 		} sat;
 	} procamp;
 
+	struct {
+		u8 nhsync:1;
+		u8 nvsync:1;
+		u8 depth:4;
+	} or;
+
 	union {
 		struct {
 			bool ilut:1;
@@ -214,6 +220,7 @@ struct nv50_head_atom {
 			bool ovly:1;
 			bool dither:1;
 			bool procamp:1;
+			bool or:1;
 		};
 		u16 mask;
 	} set;
@@ -457,6 +464,7 @@ struct nv50_head_func {
 	void (*ovly)(struct nv50_head *, struct nv50_head_atom *);
 	void (*dither)(struct nv50_head *, struct nv50_head_atom *);
 	void (*procamp)(struct nv50_head *, struct nv50_head_atom *);
+	void (*or)(struct nv50_head *, struct nv50_head_atom *);
 };
 
 #define nv50_head(c) container_of((c), struct nv50_head, base.base)
@@ -1628,6 +1636,23 @@ nv50_base_new(struct nouveau_drm *drm, int head, struct nv50_wndw **pwndw)
  * Head
  *****************************************************************************/
 static void
+head907d_or(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if (core->base.user.oclass >= GF110_DISP_CORE_CHANNEL_DMA &&
+	    (push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x0404 + (head->base.index * 0x300), 2);
+		evo_data(push, 0x00000001 | (asyh->or.depth  << 6) |
+					    (asyh->or.nvsync << 4) |
+					    (asyh->or.nhsync << 3));
+		evo_data(push, 0x31ec6000 | (head->base.index << 25) |
+					     asyh->mode.interlace);
+		evo_kick(push, core);
+	}
+}
+
+static void
 nv50_head_procamp(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
 	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
@@ -2033,6 +2058,7 @@ head507d = {
 	.ovly = nv50_head_ovly,
 	.dither = nv50_head_dither,
 	.procamp = nv50_head_procamp,
+	.or = head907d_or,
 };
 
 static void
@@ -2066,6 +2092,7 @@ nv50_head_flush_set(struct nv50_head *head, struct nv50_head_atom *asyh)
 	if (asyh->set.ovly   ) head->func->ovly    (head, asyh);
 	if (asyh->set.dither ) head->func->dither  (head, asyh);
 	if (asyh->set.procamp) head->func->procamp (head, asyh);
+	if (asyh->set.or     ) head->func->or      (head, asyh);
 }
 
 static void
@@ -2268,6 +2295,9 @@ nv50_head_atomic_check_mode(struct nv50_head *head, struct nv50_head_atom *asyh)
 	}
 	m->clock = mode->crtc_clock;
 
+	asyh->or.nhsync = !!(mode->flags & DRM_MODE_FLAG_NHSYNC);
+	asyh->or.nvsync = !!(mode->flags & DRM_MODE_FLAG_NVSYNC);
+	asyh->set.or = head->func->or != NULL;
 	asyh->set.mode = true;
 }
 
@@ -2304,6 +2334,7 @@ nv50_head_atomic_check(struct drm_crtc *crtc, struct drm_crtc_state *state)
 			if (asyc)
 				asyc->set.mask = ~0;
 			asyh->set.mask = ~0;
+			asyh->set.or = head->func->or != NULL;
 		}
 
 		if (asyh->state.mode_changed)
@@ -2707,7 +2738,7 @@ nv50_dac_enable(struct drm_encoder *encoder)
 	struct nv50_dmac *core = &nv50_disp(encoder->dev)->core->chan;
 	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
 	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
-	struct drm_display_mode *mode = &nv_crtc->base.state->adjusted_mode;
+	struct nv50_head_atom *asyh = nv50_head_atom(nv_crtc->base.state);
 	u32 *push;
 
 	nv50_outp_acquire(nv_encoder);
@@ -2715,37 +2746,17 @@ nv50_dac_enable(struct drm_encoder *encoder)
 	push = evo_wait(core, 8);
 	if (push) {
 		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			u32 syncs = 0x00000000;
-
-			if (mode->flags & DRM_MODE_FLAG_NHSYNC)
-				syncs |= 0x00000001;
-			if (mode->flags & DRM_MODE_FLAG_NVSYNC)
-				syncs |= 0x00000002;
-
 			evo_mthd(push, 0x0400 + (nv_encoder->or * 0x080), 2);
 			evo_data(push, 1 << nv_crtc->index);
-			evo_data(push, syncs);
+			evo_data(push, (asyh->or.nvsync << 1) | asyh->or.nhsync);
 		} else {
-			u32 magic = 0x31ec6000 | (nv_crtc->index << 25);
-			u32 syncs = 0x00000001;
-
-			if (mode->flags & DRM_MODE_FLAG_NHSYNC)
-				syncs |= 0x00000008;
-			if (mode->flags & DRM_MODE_FLAG_NVSYNC)
-				syncs |= 0x00000010;
-
-			if (mode->flags & DRM_MODE_FLAG_INTERLACE)
-				magic |= 0x00000001;
-
-			evo_mthd(push, 0x0404 + (nv_crtc->index * 0x300), 2);
-			evo_data(push, syncs);
-			evo_data(push, magic);
 			evo_mthd(push, 0x0180 + (nv_encoder->or * 0x020), 1);
 			evo_data(push, 1 << nv_crtc->index);
 		}
 
 		evo_kick(push, core);
 	}
+	asyh->or.depth = 0;
 
 	nv_encoder->crtc = encoder->crtc;
 }
@@ -3144,7 +3155,7 @@ nv50_msto_enable(struct drm_encoder *encoder)
 	}
 
 	mstm->outp->update(mstm->outp, head->base.index,
-			   &head->base.base.state->adjusted_mode, proto, depth);
+			   nv50_head_atom(head->base.base.state), proto, depth);
 
 	msto->head = head;
 	msto->mstc = mstc;
@@ -3612,46 +3623,31 @@ nv50_mstm_new(struct nouveau_encoder *outp, struct drm_dp_aux *aux, int aux_max,
  *****************************************************************************/
 static void
 nv50_sor_update(struct nouveau_encoder *nv_encoder, u8 head,
-		struct drm_display_mode *mode, u8 proto, u8 depth)
+		struct nv50_head_atom *asyh, u8 proto, u8 depth)
 {
 	struct nv50_disp *disp = nv50_disp(nv_encoder->base.base.dev);
 	struct nv50_dmac *core = &disp->core->chan;
 	u32 *push;
 
-	if (!mode) {
+	if (!asyh) {
 		nv_encoder->ctrl &= ~BIT(head);
 		if (!(nv_encoder->ctrl & 0x0000000f))
 			nv_encoder->ctrl = 0;
 	} else {
 		nv_encoder->ctrl |= proto << 8;
 		nv_encoder->ctrl |= BIT(head);
+		asyh->or.depth = depth;
 	}
 
 	if ((push = evo_wait(core, 6))) {
 		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			if (mode) {
-				if (mode->flags & DRM_MODE_FLAG_NHSYNC)
-					nv_encoder->ctrl |= 0x00001000;
-				if (mode->flags & DRM_MODE_FLAG_NVSYNC)
-					nv_encoder->ctrl |= 0x00002000;
-				nv_encoder->ctrl |= depth << 16;
+			if (asyh) {
+				nv_encoder->ctrl |= asyh->or.depth  << 16 |
+						    asyh->or.nvsync << 13 |
+						    asyh->or.nhsync << 12;
 			}
 			evo_mthd(push, 0x0600 + (nv_encoder->or * 0x40), 1);
 		} else {
-			if (mode) {
-				u32 magic = 0x31ec6000 | (head << 25);
-				u32 syncs = 0x00000001;
-				if (mode->flags & DRM_MODE_FLAG_NHSYNC)
-					syncs |= 0x00000008;
-				if (mode->flags & DRM_MODE_FLAG_NVSYNC)
-					syncs |= 0x00000010;
-				if (mode->flags & DRM_MODE_FLAG_INTERLACE)
-					magic |= 0x00000001;
-
-				evo_mthd(push, 0x0404 + (head * 0x300), 2);
-				evo_data(push, syncs | (depth << 6));
-				evo_data(push, magic);
-			}
 			evo_mthd(push, 0x0200 + (nv_encoder->or * 0x20), 1);
 		}
 		evo_data(push, nv_encoder->ctrl);
@@ -3692,7 +3688,8 @@ nv50_sor_enable(struct drm_encoder *encoder)
 {
 	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
 	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
-	struct drm_display_mode *mode = &nv_crtc->base.state->adjusted_mode;
+	struct nv50_head_atom *asyh = nv50_head_atom(nv_crtc->base.state);
+	struct drm_display_mode *mode = &asyh->state.adjusted_mode;
 	struct {
 		struct nv50_disp_mthd_v1 base;
 		struct nv50_disp_sor_lvds_script_v0 lvds;
@@ -3786,7 +3783,7 @@ nv50_sor_enable(struct drm_encoder *encoder)
 		break;
 	}
 
-	nv_encoder->update(nv_encoder, nv_crtc->index, mode, proto, depth);
+	nv_encoder->update(nv_encoder, nv_crtc->index, asyh, proto, depth);
 }
 
 static const struct drm_encoder_helper_funcs
@@ -3924,19 +3921,19 @@ nv50_pior_enable(struct drm_encoder *encoder)
 	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
 	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
 	struct nouveau_connector *nv_connector;
-	struct drm_display_mode *mode = &nv_crtc->base.state->adjusted_mode;
+	struct nv50_head_atom *asyh = nv50_head_atom(nv_crtc->base.state);
 	u8 owner = 1 << nv_crtc->index;
-	u8 proto, depth;
+	u8 proto;
 	u32 *push;
 
 	nv50_outp_acquire(nv_encoder);
 
 	nv_connector = nouveau_encoder_connector_get(nv_encoder);
 	switch (nv_connector->base.display_info.bpc) {
-	case 10: depth = 0x6; break;
-	case  8: depth = 0x5; break;
-	case  6: depth = 0x2; break;
-	default: depth = 0x0; break;
+	case 10: asyh->or.depth = 0x6; break;
+	case  8: asyh->or.depth = 0x5; break;
+	case  6: asyh->or.depth = 0x2; break;
+	default: asyh->or.depth = 0x0; break;
 	}
 
 	switch (nv_encoder->dcb->type) {
@@ -3952,13 +3949,11 @@ nv50_pior_enable(struct drm_encoder *encoder)
 	push = evo_wait(core, 8);
 	if (push) {
 		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			u32 ctrl = (depth << 16) | (proto << 8) | owner;
-			if (mode->flags & DRM_MODE_FLAG_NHSYNC)
-				ctrl |= 0x00001000;
-			if (mode->flags & DRM_MODE_FLAG_NVSYNC)
-				ctrl |= 0x00002000;
 			evo_mthd(push, 0x0700 + (nv_encoder->or * 0x040), 1);
-			evo_data(push, ctrl);
+			evo_data(push, (asyh->or.depth  << 16) |
+				       (asyh->or.nvsync << 13) |
+				       (asyh->or.nhsync << 12) |
+				       (proto << 8) | owner);
 		}
 
 		evo_kick(push, core);
diff --git a/drivers/gpu/drm/nouveau/nouveau_encoder.h b/drivers/gpu/drm/nouveau/nouveau_encoder.h
index e28d966..3517f92 100644
--- a/drivers/gpu/drm/nouveau/nouveau_encoder.h
+++ b/drivers/gpu/drm/nouveau/nouveau_encoder.h
@@ -32,6 +32,7 @@
 #include <drm/drm_encoder_slave.h>
 #include <drm/drm_dp_mst_helper.h>
 #include "dispnv04/disp.h"
+struct nv50_head_atom;
 
 #define NV_DPMS_CLEARED 0x80
 
@@ -68,7 +69,7 @@ struct nouveau_encoder {
 	void (*enc_save)(struct drm_encoder *encoder);
 	void (*enc_restore)(struct drm_encoder *encoder);
 	void (*update)(struct nouveau_encoder *, u8 head,
-		       struct drm_display_mode *, u8 proto, u8 depth);
+		       struct nv50_head_atom *, u8 proto, u8 depth);
 };
 
 struct nouveau_encoder *
-- 
cgit v1.1


From 0a3687716bb0a53a363b63cb5ba2bddc14c3bd2a Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50: abstract OR interfaces so the code can be split

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/disp.c | 180 ++++++++++++++++++--------------
 1 file changed, 102 insertions(+), 78 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c
index f22c637..995109e 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/disp.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c
@@ -495,6 +495,14 @@ struct nv50_core {
 
 struct nv50_core_func {
 	const struct nv50_head_func *head;
+	const struct nv50_outp_func *dac;
+	const struct nv50_outp_func *sor;
+	const struct nv50_outp_func *pior;
+};
+
+struct nv50_outp_func {
+	void (*ctrl)(struct nv50_core *, int or, u32 ctrl,
+		     struct nv50_head_atom *);
 };
 
 static int
@@ -1641,7 +1649,7 @@ head907d_or(struct nv50_head *head, struct nv50_head_atom *asyh)
 	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
 	if (core->base.user.oclass >= GF110_DISP_CORE_CHANNEL_DMA &&
-	    (push = evo_wait(core, 2))) {
+	    (push = evo_wait(core, 3))) {
 		evo_mthd(push, 0x0404 + (head->base.index * 0x300), 2);
 		evo_data(push, 0x00000001 | (asyh->or.depth  << 6) |
 					    (asyh->or.nvsync << 4) |
@@ -2546,9 +2554,15 @@ out:
 	return ret;
 }
 
+static const struct nv50_outp_func dac507d;
+static const struct nv50_outp_func sor507d;
+static const struct nv50_outp_func pior507d;
 static const struct nv50_core_func
 core507d = {
 	.head = &head507d,
+	.dac = &dac507d,
+	.sor = &sor507d,
+	.pior = &pior507d,
 };
 
 static int
@@ -2707,27 +2721,39 @@ nv50_outp_atomic_check(struct drm_encoder *encoder,
  * DAC
  *****************************************************************************/
 static void
-nv50_dac_disable(struct drm_encoder *encoder)
+dac507d_ctrl(struct nv50_core *core, int or, u32 ctrl,
+	     struct nv50_head_atom *asyh)
 {
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nv50_dmac *core = &nv50_disp(encoder->dev)->core->chan;
-	const int or = nv_encoder->or;
-	u32 *push;
-
-	if (nv_encoder->crtc) {
-		push = evo_wait(core, 4);
-		if (push) {
-			if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-				evo_mthd(push, 0x0400 + (or * 0x080), 1);
-				evo_data(push, 0x00000000);
-			} else {
-				evo_mthd(push, 0x0180 + (or * 0x020), 1);
-				evo_data(push, 0x00000000);
+	u32 *push, sync = 0;
+	if ((push = evo_wait(&core->chan, 3))) {
+		if (core->chan.base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
+			if (asyh) {
+				sync |= asyh->or.nvsync << 1;
+				sync |= asyh->or.nhsync;
 			}
-			evo_kick(push, core);
+			evo_mthd(push, 0x0400 + (or * 0x080), 2);
+			evo_data(push, ctrl);
+			evo_data(push, sync);
+		} else {
+			evo_mthd(push, 0x0180 + (or * 0x020), 1);
+			evo_data(push, ctrl);
 		}
+		evo_kick(push, &core->chan);
 	}
+}
+
+static const struct nv50_outp_func
+dac507d = {
+	.ctrl = dac507d_ctrl,
+};
 
+static void
+nv50_dac_disable(struct drm_encoder *encoder)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nv50_core *core = nv50_disp(encoder->dev)->core;
+	if (nv_encoder->crtc)
+		core->func->dac->ctrl(core, nv_encoder->or, 0x00000000, NULL);
 	nv_encoder->crtc = NULL;
 	nv50_outp_release(nv_encoder);
 }
@@ -2735,27 +2761,14 @@ nv50_dac_disable(struct drm_encoder *encoder)
 static void
 nv50_dac_enable(struct drm_encoder *encoder)
 {
-	struct nv50_dmac *core = &nv50_disp(encoder->dev)->core->chan;
 	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
 	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
 	struct nv50_head_atom *asyh = nv50_head_atom(nv_crtc->base.state);
-	u32 *push;
+	struct nv50_core *core = nv50_disp(encoder->dev)->core;
 
 	nv50_outp_acquire(nv_encoder);
 
-	push = evo_wait(core, 8);
-	if (push) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0400 + (nv_encoder->or * 0x080), 2);
-			evo_data(push, 1 << nv_crtc->index);
-			evo_data(push, (asyh->or.nvsync << 1) | asyh->or.nhsync);
-		} else {
-			evo_mthd(push, 0x0180 + (nv_encoder->or * 0x020), 1);
-			evo_data(push, 1 << nv_crtc->index);
-		}
-
-		evo_kick(push, core);
-	}
+	core->func->dac->ctrl(core, nv_encoder->or, 1 << nv_crtc->index, asyh);
 	asyh->or.depth = 0;
 
 	nv_encoder->crtc = encoder->crtc;
@@ -3622,12 +3635,37 @@ nv50_mstm_new(struct nouveau_encoder *outp, struct drm_dp_aux *aux, int aux_max,
  * SOR
  *****************************************************************************/
 static void
+sor507d_ctrl(struct nv50_core *core, int or, u32 ctrl,
+	     struct nv50_head_atom *asyh)
+{
+	u32 *push;
+	if ((push = evo_wait(&core->chan, 6))) {
+		if (core->chan.base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
+			if (asyh) {
+				ctrl |= asyh->or.depth  << 16;
+				ctrl |= asyh->or.nvsync << 13;
+				ctrl |= asyh->or.nhsync << 12;
+			}
+			evo_mthd(push, 0x0600 + (or * 0x40), 1);
+		} else {
+			evo_mthd(push, 0x0200 + (or * 0x20), 1);
+		}
+		evo_data(push, ctrl);
+		evo_kick(push, &core->chan);
+	}
+}
+
+static const struct nv50_outp_func
+sor507d = {
+	.ctrl = sor507d_ctrl,
+};
+
+static void
 nv50_sor_update(struct nouveau_encoder *nv_encoder, u8 head,
 		struct nv50_head_atom *asyh, u8 proto, u8 depth)
 {
 	struct nv50_disp *disp = nv50_disp(nv_encoder->base.base.dev);
-	struct nv50_dmac *core = &disp->core->chan;
-	u32 *push;
+	struct nv50_core *core = disp->core;
 
 	if (!asyh) {
 		nv_encoder->ctrl &= ~BIT(head);
@@ -3639,20 +3677,7 @@ nv50_sor_update(struct nouveau_encoder *nv_encoder, u8 head,
 		asyh->or.depth = depth;
 	}
 
-	if ((push = evo_wait(core, 6))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			if (asyh) {
-				nv_encoder->ctrl |= asyh->or.depth  << 16 |
-						    asyh->or.nvsync << 13 |
-						    asyh->or.nhsync << 12;
-			}
-			evo_mthd(push, 0x0600 + (nv_encoder->or * 0x40), 1);
-		} else {
-			evo_mthd(push, 0x0200 + (nv_encoder->or * 0x20), 1);
-		}
-		evo_data(push, nv_encoder->ctrl);
-		evo_kick(push, core);
-	}
+	core->func->sor->ctrl(core, nv_encoder->or, nv_encoder->ctrl, asyh);
 }
 
 static void
@@ -3879,6 +3904,30 @@ nv50_sor_create(struct drm_connector *connector, struct dcb_output *dcbe)
 /******************************************************************************
  * PIOR
  *****************************************************************************/
+static void
+pior507d_ctrl(struct nv50_core *core, int or, u32 ctrl,
+	      struct nv50_head_atom *asyh)
+{
+	u32 *push;
+	if ((push = evo_wait(&core->chan, 8))) {
+		if (core->chan.base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
+			if (asyh) {
+				ctrl |= asyh->or.depth  << 16;
+				ctrl |= asyh->or.nvsync << 13;
+				ctrl |= asyh->or.nhsync << 12;
+			}
+			evo_mthd(push, 0x0700 + (or * 0x040), 1);
+			evo_data(push, ctrl);
+		}
+		evo_kick(push, &core->chan);
+	}
+}
+
+static const struct nv50_outp_func
+pior507d = {
+	.ctrl = pior507d_ctrl,
+};
+
 static int
 nv50_pior_atomic_check(struct drm_encoder *encoder,
 		       struct drm_crtc_state *crtc_state,
@@ -3895,21 +3944,9 @@ static void
 nv50_pior_disable(struct drm_encoder *encoder)
 {
 	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nv50_dmac *core = &nv50_disp(encoder->dev)->core->chan;
-	const int or = nv_encoder->or;
-	u32 *push;
-
-	if (nv_encoder->crtc) {
-		push = evo_wait(core, 4);
-		if (push) {
-			if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-				evo_mthd(push, 0x0700 + (or * 0x040), 1);
-				evo_data(push, 0x00000000);
-			}
-			evo_kick(push, core);
-		}
-	}
-
+	struct nv50_core *core = nv50_disp(encoder->dev)->core;
+	if (nv_encoder->crtc)
+		core->func->pior->ctrl(core, nv_encoder->or, 0x00000000, NULL);
 	nv_encoder->crtc = NULL;
 	nv50_outp_release(nv_encoder);
 }
@@ -3917,14 +3954,13 @@ nv50_pior_disable(struct drm_encoder *encoder)
 static void
 nv50_pior_enable(struct drm_encoder *encoder)
 {
-	struct nv50_dmac *core = &nv50_disp(encoder->dev)->core->chan;
 	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
 	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
 	struct nouveau_connector *nv_connector;
 	struct nv50_head_atom *asyh = nv50_head_atom(nv_crtc->base.state);
+	struct nv50_core *core = nv50_disp(encoder->dev)->core;
 	u8 owner = 1 << nv_crtc->index;
 	u8 proto;
-	u32 *push;
 
 	nv50_outp_acquire(nv_encoder);
 
@@ -3946,19 +3982,7 @@ nv50_pior_enable(struct drm_encoder *encoder)
 		break;
 	}
 
-	push = evo_wait(core, 8);
-	if (push) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0700 + (nv_encoder->or * 0x040), 1);
-			evo_data(push, (asyh->or.depth  << 16) |
-				       (asyh->or.nvsync << 13) |
-				       (asyh->or.nhsync << 12) |
-				       (proto << 8) | owner);
-		}
-
-		evo_kick(push, core);
-	}
-
+	core->func->pior->ctrl(core, nv_encoder->or, (proto << 8) | owner, asyh);
 	nv_encoder->crtc = encoder->crtc;
 }
 
-- 
cgit v1.1


From 1590700d94ac53772491ed3103a4e8b8de01640a Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50-: split each resource type into their own source
 files

There should be no code changes here, just shuffling stuff around.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/Kbuild     |   26 +
 drivers/gpu/drm/nouveau/dispnv50/atom.h     |  207 ++
 drivers/gpu/drm/nouveau/dispnv50/base.c     |   53 +
 drivers/gpu/drm/nouveau/dispnv50/base.h     |    8 +
 drivers/gpu/drm/nouveau/dispnv50/base507c.c |  307 +++
 drivers/gpu/drm/nouveau/dispnv50/core.c     |   69 +
 drivers/gpu/drm/nouveau/dispnv50/core.h     |   26 +
 drivers/gpu/drm/nouveau/dispnv50/core507d.c |   65 +
 drivers/gpu/drm/nouveau/dispnv50/curs.c     |   51 +
 drivers/gpu/drm/nouveau/dispnv50/curs.h     |    8 +
 drivers/gpu/drm/nouveau/dispnv50/curs507a.c |  151 ++
 drivers/gpu/drm/nouveau/dispnv50/dac507d.c  |   51 +
 drivers/gpu/drm/nouveau/dispnv50/disp.c     | 2751 ++-------------------------
 drivers/gpu/drm/nouveau/dispnv50/disp.h     |   71 +
 drivers/gpu/drm/nouveau/dispnv50/head.c     |  566 ++++++
 drivers/gpu/drm/nouveau/dispnv50/head.h     |   39 +
 drivers/gpu/drm/nouveau/dispnv50/head507d.c |  403 ++++
 drivers/gpu/drm/nouveau/dispnv50/oimm.c     |   51 +
 drivers/gpu/drm/nouveau/dispnv50/oimm.h     |    8 +
 drivers/gpu/drm/nouveau/dispnv50/oimm507b.c |   56 +
 drivers/gpu/drm/nouveau/dispnv50/ovly.c     |   57 +
 drivers/gpu/drm/nouveau/dispnv50/ovly.h     |    8 +
 drivers/gpu/drm/nouveau/dispnv50/ovly507e.c |   70 +
 drivers/gpu/drm/nouveau/dispnv50/pior507d.c |   48 +
 drivers/gpu/drm/nouveau/dispnv50/sor507d.c  |   50 +
 drivers/gpu/drm/nouveau/dispnv50/wndw.c     |  434 +++++
 drivers/gpu/drm/nouveau/dispnv50/wndw.h     |   73 +
 drivers/gpu/drm/nouveau/nv50_display.h      |    1 -
 28 files changed, 3096 insertions(+), 2612 deletions(-)
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/atom.h
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/base.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/base.h
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/base507c.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/core.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/core.h
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/core507d.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/curs.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/curs.h
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/curs507a.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/dac507d.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/disp.h
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/head.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/head.h
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/head507d.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/oimm.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/oimm.h
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/oimm507b.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/ovly.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/ovly.h
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/ovly507e.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/pior507d.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/sor507d.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/wndw.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/wndw.h

diff --git a/drivers/gpu/drm/nouveau/dispnv50/Kbuild b/drivers/gpu/drm/nouveau/dispnv50/Kbuild
index 43fc8be..f3877d2 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/Kbuild
+++ b/drivers/gpu/drm/nouveau/dispnv50/Kbuild
@@ -1 +1,27 @@
 nouveau-y += dispnv50/disp.o
+
+nouveau-y += dispnv50/core.o
+nouveau-y += dispnv50/core507d.o
+
+nouveau-y += dispnv50/dac507d.o
+
+nouveau-y += dispnv50/pior507d.o
+
+nouveau-y += dispnv50/sor507d.o
+
+nouveau-y += dispnv50/head.o
+nouveau-y += dispnv50/head507d.o
+
+nouveau-y += dispnv50/wndw.o
+
+nouveau-y += dispnv50/base.o
+nouveau-y += dispnv50/base507c.o
+
+nouveau-y += dispnv50/curs.o
+nouveau-y += dispnv50/curs507a.o
+
+nouveau-y += dispnv50/oimm.o
+nouveau-y += dispnv50/oimm507b.o
+
+nouveau-y += dispnv50/ovly.o
+nouveau-y += dispnv50/ovly507e.o
diff --git a/drivers/gpu/drm/nouveau/dispnv50/atom.h b/drivers/gpu/drm/nouveau/dispnv50/atom.h
new file mode 100644
index 0000000..8c97e25
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/atom.h
@@ -0,0 +1,207 @@
+#ifndef __NV50_KMS_ATOM_H__
+#define __NV50_KMS_ATOM_H__
+#define nv50_atom(p) container_of((p), struct nv50_atom, state)
+#include <drm/drm_atomic.h>
+
+struct nv50_atom {
+	struct drm_atomic_state state;
+
+	struct list_head outp;
+	bool lock_core;
+	bool flush_disable;
+};
+
+#define nv50_head_atom(p) container_of((p), struct nv50_head_atom, state)
+
+struct nv50_head_atom {
+	struct drm_crtc_state state;
+
+	struct {
+		u16 iW;
+		u16 iH;
+		u16 oW;
+		u16 oH;
+	} view;
+
+	struct nv50_head_mode {
+		bool interlace;
+		u32 clock;
+		struct {
+			u16 active;
+			u16 synce;
+			u16 blanke;
+			u16 blanks;
+		} h;
+		struct {
+			u32 active;
+			u16 synce;
+			u16 blanke;
+			u16 blanks;
+			u16 blank2s;
+			u16 blank2e;
+			u16 blankus;
+		} v;
+	} mode;
+
+	struct {
+		bool visible;
+		u32 handle;
+		u64 offset:40;
+		u8  mode:4;
+	} ilut;
+
+	struct {
+		bool visible;
+		u32 handle;
+		u64 offset:40;
+		u8  format;
+		u8  kind:7;
+		u8  layout:1;
+		u8  block:4;
+		u32 pitch:20;
+		u16 x;
+		u16 y;
+		u16 w;
+		u16 h;
+	} core;
+
+	struct {
+		bool visible;
+		u32 handle;
+		u64 offset:40;
+		u8  layout:1;
+		u8  format:1;
+	} curs;
+
+	struct {
+		u8  depth;
+		u8  cpp;
+		u16 x;
+		u16 y;
+		u16 w;
+		u16 h;
+	} base;
+
+	struct {
+		u8 cpp;
+	} ovly;
+
+	struct {
+		bool enable:1;
+		u8 bits:2;
+		u8 mode:4;
+	} dither;
+
+	struct {
+		struct {
+			u16 cos:12;
+			u16 sin:12;
+		} sat;
+	} procamp;
+
+	struct {
+		u8 nhsync:1;
+		u8 nvsync:1;
+		u8 depth:4;
+	} or;
+
+	union {
+		struct {
+			bool ilut:1;
+			bool core:1;
+			bool curs:1;
+		};
+		u8 mask;
+	} clr;
+
+	union {
+		struct {
+			bool ilut:1;
+			bool core:1;
+			bool curs:1;
+			bool view:1;
+			bool mode:1;
+			bool base:1;
+			bool ovly:1;
+			bool dither:1;
+			bool procamp:1;
+			bool or:1;
+		};
+		u16 mask;
+	} set;
+};
+
+static inline struct nv50_head_atom *
+nv50_head_atom_get(struct drm_atomic_state *state, struct drm_crtc *crtc)
+{
+	struct drm_crtc_state *statec = drm_atomic_get_crtc_state(state, crtc);
+	if (IS_ERR(statec))
+		return (void *)statec;
+	return nv50_head_atom(statec);
+}
+
+#define nv50_wndw_atom(p) container_of((p), struct nv50_wndw_atom, state)
+
+struct nv50_wndw_atom {
+	struct drm_plane_state state;
+	u8 interval;
+
+	struct {
+		u32  handle;
+		u16  offset:12;
+		bool awaken:1;
+	} ntfy;
+
+	struct {
+		u32 handle;
+		u16 offset:12;
+		u32 acquire;
+		u32 release;
+	} sema;
+
+	struct {
+		u8 enable:2;
+	} lut;
+
+	struct {
+		u8  mode:2;
+		u8  interval:4;
+
+		u8  format;
+		u8  kind:7;
+		u8  layout:1;
+		u8  block:4;
+		u32 pitch:20;
+		u16 w;
+		u16 h;
+
+		u32 handle;
+		u64 offset;
+	} image;
+
+	struct {
+		u16 x;
+		u16 y;
+	} point;
+
+	union {
+		struct {
+			bool ntfy:1;
+			bool sema:1;
+			bool image:1;
+		};
+		u8 mask;
+	} clr;
+
+	union {
+		struct {
+			bool ntfy:1;
+			bool sema:1;
+			bool image:1;
+			bool lut:1;
+			bool point:1;
+		};
+		u8 mask;
+	} set;
+};
+#endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base.c b/drivers/gpu/drm/nouveau/dispnv50/base.c
new file mode 100644
index 0000000..12ca5d7
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/base.c
@@ -0,0 +1,53 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "base.h"
+
+#include <nvif/class.h>
+
+int
+nv50_base_new(struct nouveau_drm *drm, int head, struct nv50_wndw **pwndw)
+{
+	struct {
+		s32 oclass;
+		int version;
+		int (*new)(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+	} bases[] = {
+		{ GK110_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
+		{ GK104_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
+		{ GF110_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
+		{ GT214_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
+		{ GT200_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
+		{   G82_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
+		{  NV50_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
+		{}
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	int cid;
+
+	cid = nvif_mclass(&disp->disp->object, bases);
+	if (cid < 0) {
+		NV_ERROR(drm, "No supported base class\n");
+		return cid;
+	}
+
+	return bases[cid].new(drm, head, bases[cid].oclass, pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base.h b/drivers/gpu/drm/nouveau/dispnv50/base.h
new file mode 100644
index 0000000..1daba73
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/base.h
@@ -0,0 +1,8 @@
+#ifndef __NV50_KMS_BASE_H__
+#define __NV50_KMS_BASE_H__
+#include "wndw.h"
+
+int base507c_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+
+int nv50_base_new(struct nouveau_drm *, int head, struct nv50_wndw **);
+#endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base507c.c b/drivers/gpu/drm/nouveau/dispnv50/base507c.c
new file mode 100644
index 0000000..b73e7b4
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/base507c.c
@@ -0,0 +1,307 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "base.h"
+
+#include <nvif/class.h>
+#include <nvif/cl507c.h>
+#include <nvif/event.h>
+
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_plane_helper.h>
+#include "nouveau_bo.h"
+
+static u32
+base507c_update(struct nv50_wndw *wndw, u32 interlock)
+{
+	u32 *push;
+
+	if (!(push = evo_wait(&wndw->wndw, 2)))
+		return 0;
+	evo_mthd(push, 0x0080, 1);
+	evo_data(push, interlock);
+	evo_kick(push, &wndw->wndw);
+
+	if (wndw->wndw.base.user.oclass < GF110_DISP_BASE_CHANNEL_DMA)
+		return interlock ? 2 << (wndw->id * 8) : 0;
+	return interlock ? 2 << (wndw->id * 4) : 0;
+}
+
+static void
+base507c_lut(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 2))) {
+		evo_mthd(push, 0x00e0, 1);
+		evo_data(push, asyw->lut.enable << 30);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+base507c_image_clr(struct nv50_wndw *wndw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 4))) {
+		evo_mthd(push, 0x0084, 1);
+		evo_data(push, 0x00000000);
+		evo_mthd(push, 0x00c0, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+base507c_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	const s32 oclass = wndw->wndw.base.user.oclass;
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 10))) {
+		evo_mthd(push, 0x0084, 1);
+		evo_data(push, asyw->image.mode << 8 |
+			       asyw->image.interval << 4);
+		evo_mthd(push, 0x00c0, 1);
+		evo_data(push, asyw->image.handle);
+		if (oclass < G82_DISP_BASE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0800, 5);
+			evo_data(push, asyw->image.offset >> 8);
+			evo_data(push, 0x00000000);
+			evo_data(push, (asyw->image.h << 16) | asyw->image.w);
+			evo_data(push, (asyw->image.layout << 20) |
+					asyw->image.pitch |
+					asyw->image.block);
+			evo_data(push, (asyw->image.kind << 16) |
+				       (asyw->image.format << 8));
+		} else
+		if (oclass < GF110_DISP_BASE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0800, 5);
+			evo_data(push, asyw->image.offset >> 8);
+			evo_data(push, 0x00000000);
+			evo_data(push, (asyw->image.h << 16) | asyw->image.w);
+			evo_data(push, (asyw->image.layout << 20) |
+					asyw->image.pitch |
+					asyw->image.block);
+			evo_data(push, asyw->image.format << 8);
+		} else {
+			evo_mthd(push, 0x0400, 5);
+			evo_data(push, asyw->image.offset >> 8);
+			evo_data(push, 0x00000000);
+			evo_data(push, (asyw->image.h << 16) | asyw->image.w);
+			evo_data(push, (asyw->image.layout << 24) |
+					asyw->image.pitch |
+					asyw->image.block);
+			evo_data(push, asyw->image.format << 8);
+		}
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static int
+base507c_ntfy_wait_begun(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	struct nouveau_drm *drm = nouveau_drm(wndw->plane.dev);
+	struct nv50_disp *disp = nv50_disp(wndw->plane.dev);
+	if (nvif_msec(&drm->client.device, 2000ULL,
+		u32 data = nouveau_bo_rd32(disp->sync, asyw->ntfy.offset / 4);
+		if ((data & 0xc0000000) == 0x40000000)
+			break;
+		usleep_range(1, 2);
+	) < 0)
+		return -ETIMEDOUT;
+	return 0;
+}
+
+static void
+base507c_ntfy_clr(struct nv50_wndw *wndw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 2))) {
+		evo_mthd(push, 0x00a4, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+base507c_ntfy_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 3))) {
+		evo_mthd(push, 0x00a0, 2);
+		evo_data(push, asyw->ntfy.awaken << 30 | asyw->ntfy.offset);
+		evo_data(push, asyw->ntfy.handle);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+base507c_sema_clr(struct nv50_wndw *wndw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 2))) {
+		evo_mthd(push, 0x0094, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+base507c_sema_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 5))) {
+		evo_mthd(push, 0x0088, 4);
+		evo_data(push, asyw->sema.offset);
+		evo_data(push, asyw->sema.acquire);
+		evo_data(push, asyw->sema.release);
+		evo_data(push, asyw->sema.handle);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+base507c_release(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
+		 struct nv50_head_atom *asyh)
+{
+	asyh->base.cpp = 0;
+}
+
+static int
+base507c_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
+		 struct nv50_head_atom *asyh)
+{
+	const struct drm_framebuffer *fb = asyw->state.fb;
+	int ret;
+
+	if (!fb->format->depth)
+		return -EINVAL;
+
+	ret = drm_atomic_helper_check_plane_state(&asyw->state, &asyh->state,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  false, true);
+	if (ret)
+		return ret;
+
+	asyh->base.depth = fb->format->depth;
+	asyh->base.cpp = fb->format->cpp[0];
+	asyh->base.x = asyw->state.src.x1 >> 16;
+	asyh->base.y = asyw->state.src.y1 >> 16;
+	asyh->base.w = asyw->state.fb->width;
+	asyh->base.h = asyw->state.fb->height;
+
+	switch (fb->format->format) {
+	case DRM_FORMAT_C8         : asyw->image.format = 0x1e; break;
+	case DRM_FORMAT_RGB565     : asyw->image.format = 0xe8; break;
+	case DRM_FORMAT_XRGB1555   :
+	case DRM_FORMAT_ARGB1555   : asyw->image.format = 0xe9; break;
+	case DRM_FORMAT_XRGB8888   :
+	case DRM_FORMAT_ARGB8888   : asyw->image.format = 0xcf; break;
+	case DRM_FORMAT_XBGR2101010:
+	case DRM_FORMAT_ABGR2101010: asyw->image.format = 0xd1; break;
+	case DRM_FORMAT_XBGR8888   :
+	case DRM_FORMAT_ABGR8888   : asyw->image.format = 0xd5; break;
+	default:
+		WARN_ON(1);
+		return -EINVAL;
+	}
+
+	asyw->lut.enable = 1;
+	asyw->set.image = true;
+	return 0;
+}
+
+static const u32
+base507c_format[] = {
+	DRM_FORMAT_C8,
+	DRM_FORMAT_RGB565,
+	DRM_FORMAT_XRGB1555,
+	DRM_FORMAT_ARGB1555,
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_XBGR2101010,
+	DRM_FORMAT_ABGR2101010,
+	DRM_FORMAT_XBGR8888,
+	DRM_FORMAT_ABGR8888,
+	0
+};
+
+static const struct nv50_wndw_func
+base507c = {
+	.acquire = base507c_acquire,
+	.release = base507c_release,
+	.sema_set = base507c_sema_set,
+	.sema_clr = base507c_sema_clr,
+	.ntfy_set = base507c_ntfy_set,
+	.ntfy_clr = base507c_ntfy_clr,
+	.ntfy_wait_begun = base507c_ntfy_wait_begun,
+	.image_set = base507c_image_set,
+	.image_clr = base507c_image_clr,
+	.lut = base507c_lut,
+	.update = base507c_update,
+};
+
+static int
+base507c_new_(const struct nv50_wndw_func *func, const u32 *format,
+	      struct nouveau_drm *drm, int head, s32 oclass,
+	      struct nv50_wndw **pwndw)
+{
+	struct nv50_disp_base_channel_dma_v0 args = {
+		.head = head,
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	struct nv50_wndw *wndw;
+	int ret;
+
+	ret = nv50_wndw_new_(func, drm->dev, DRM_PLANE_TYPE_PRIMARY,
+			     "base", head, format, &wndw);
+	if (*pwndw = wndw, ret)
+		return ret;
+
+	ret = nv50_dmac_create(&drm->client.device, &disp->disp->object,
+			       &oclass, head, &args, sizeof(args),
+			       disp->sync->bo.offset, &wndw->wndw);
+	if (ret) {
+		NV_ERROR(drm, "base%04x allocation failed: %d\n", oclass, ret);
+		return ret;
+	}
+
+	ret = nvif_notify_init(&wndw->wndw.base.user, wndw->notify.func,
+			       false, NV50_DISP_BASE_CHANNEL_DMA_V0_NTFY_UEVENT,
+			       &(struct nvif_notify_uevent_req) {},
+			       sizeof(struct nvif_notify_uevent_req),
+			       sizeof(struct nvif_notify_uevent_rep),
+			       &wndw->notify);
+	if (ret)
+		return ret;
+
+	wndw->ntfy = NV50_DISP_BASE_NTFY(wndw->id);
+	wndw->sema = NV50_DISP_BASE_SEM0(wndw->id);
+	wndw->data = 0x00000000;
+	return 0;
+}
+
+int
+base507c_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return base507c_new_(&base507c, base507c_format, drm, head, oclass, pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/core.c b/drivers/gpu/drm/nouveau/dispnv50/core.c
new file mode 100644
index 0000000..b12899f
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/core.c
@@ -0,0 +1,69 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "core.h"
+
+#include <nvif/class.h>
+
+void
+nv50_core_del(struct nv50_core **pcore)
+{
+	struct nv50_core *core = *pcore;
+	if (core) {
+		nv50_dmac_destroy(&core->chan);
+		kfree(*pcore);
+		*pcore = NULL;
+	}
+}
+
+int
+nv50_core_new(struct nouveau_drm *drm, struct nv50_core **pcore)
+{
+	struct {
+		s32 oclass;
+		int version;
+		int (*new)(struct nouveau_drm *, s32, struct nv50_core **);
+	} cores[] = {
+		{ GP102_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{ GP100_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{ GM200_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{ GM107_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{ GK110_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{ GK104_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{ GF110_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{ GT214_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{ GT206_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{ GT200_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{   G82_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{  NV50_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{}
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	int cid;
+
+	cid = nvif_mclass(&disp->disp->object, cores);
+	if (cid < 0) {
+		NV_ERROR(drm, "No supported core channel class\n");
+		return cid;
+	}
+
+	return cores[cid].new(drm, cores[cid].oclass, pcore);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/core.h b/drivers/gpu/drm/nouveau/dispnv50/core.h
new file mode 100644
index 0000000..3cd5446
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/core.h
@@ -0,0 +1,26 @@
+#ifndef __NV50_KMS_CORE_H__
+#define __NV50_KMS_CORE_H__
+#include "disp.h"
+#include "atom.h"
+
+struct nv50_core {
+	const struct nv50_core_func *func;
+	struct nv50_dmac chan;
+};
+
+int nv50_core_new(struct nouveau_drm *, struct nv50_core **);
+void nv50_core_del(struct nv50_core **);
+
+struct nv50_core_func {
+	const struct nv50_head_func *head;
+	const struct nv50_outp_func {
+		void (*ctrl)(struct nv50_core *, int or, u32 ctrl,
+			     struct nv50_head_atom *);
+	} *dac, *pior, *sor;
+};
+
+int core507d_new(struct nouveau_drm *, s32, struct nv50_core **);
+extern const struct nv50_outp_func dac507d;
+extern const struct nv50_outp_func sor507d;
+extern const struct nv50_outp_func pior507d;
+#endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/core507d.c b/drivers/gpu/drm/nouveau/dispnv50/core507d.c
new file mode 100644
index 0000000..b0325f6
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/core507d.c
@@ -0,0 +1,65 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "core.h"
+#include "head.h"
+
+#include <nvif/cl507d.h>
+
+#include "nouveau_bo.h"
+
+static const struct nv50_core_func
+core507d = {
+	.head = &head507d,
+	.dac = &dac507d,
+	.sor = &sor507d,
+	.pior = &pior507d,
+};
+
+static int
+core507d_new_(const struct nv50_core_func *func, struct nouveau_drm *drm,
+	      s32 oclass, struct nv50_core **pcore)
+{
+	struct nv50_disp_core_channel_dma_v0 args = {};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	struct nv50_core *core;
+	int ret;
+
+	if (!(core = *pcore = kzalloc(sizeof(*core), GFP_KERNEL)))
+		return -ENOMEM;
+	core->func = func;
+
+	ret = nv50_dmac_create(&drm->client.device, &disp->disp->object,
+			       &oclass, 0, &args, sizeof(args),
+			       disp->sync->bo.offset, &core->chan);
+	if (ret) {
+		NV_ERROR(drm, "core%04x allocation failed: %d\n", oclass, ret);
+		return ret;
+	}
+
+	return 0;
+}
+
+int
+core507d_new(struct nouveau_drm *drm, s32 oclass, struct nv50_core **pcore)
+{
+	return core507d_new_(&core507d, drm, oclass, pcore);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/curs.c b/drivers/gpu/drm/nouveau/dispnv50/curs.c
new file mode 100644
index 0000000..6d60e97
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/curs.c
@@ -0,0 +1,51 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "curs.h"
+
+#include <nvif/class.h>
+
+int
+nv50_curs_new(struct nouveau_drm *drm, int head, struct nv50_wndw **pwndw)
+{
+	struct {
+		s32 oclass;
+		int version;
+		int (*new)(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+	} curses[] = {
+		{ GK104_DISP_CURSOR, 0, curs507a_new },
+		{ GF110_DISP_CURSOR, 0, curs507a_new },
+		{ GT214_DISP_CURSOR, 0, curs507a_new },
+		{   G82_DISP_CURSOR, 0, curs507a_new },
+		{  NV50_DISP_CURSOR, 0, curs507a_new },
+		{}
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	int cid;
+
+	cid = nvif_mclass(&disp->disp->object, curses);
+	if (cid < 0) {
+		NV_ERROR(drm, "No supported cursor immediate class\n");
+		return cid;
+	}
+
+	return curses[cid].new(drm, head, curses[cid].oclass, pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/curs.h b/drivers/gpu/drm/nouveau/dispnv50/curs.h
new file mode 100644
index 0000000..b85ca9f
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/curs.h
@@ -0,0 +1,8 @@
+#ifndef __NV50_KMS_CURS_H__
+#define __NV50_KMS_CURS_H__
+#include "wndw.h"
+
+int curs507a_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+
+int nv50_curs_new(struct nouveau_drm *, int head, struct nv50_wndw **);
+#endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/curs507a.c b/drivers/gpu/drm/nouveau/dispnv50/curs507a.c
new file mode 100644
index 0000000..1a3e199
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/curs507a.c
@@ -0,0 +1,151 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "curs.h"
+#include "core.h"
+
+#include <nvif/cl507a.h>
+
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_plane_helper.h>
+
+static u32
+curs507a_update(struct nv50_wndw *wndw, u32 interlock)
+{
+	nvif_wr32(&wndw->wimm.base.user, 0x0080, 0x00000000);
+	return 0;
+}
+
+static void
+curs507a_point(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	nvif_wr32(&wndw->wimm.base.user, 0x0084, asyw->point.y << 16 |
+						 asyw->point.x);
+}
+
+static const struct nv50_wimm_func
+curs507a = {
+	.point = curs507a_point,
+	.update = curs507a_update,
+};
+
+static void
+curs507a_prepare(struct nv50_wndw *wndw, struct nv50_head_atom *asyh,
+		 struct nv50_wndw_atom *asyw)
+{
+	u32 handle = nv50_disp(wndw->plane.dev)->core->chan.vram.handle;
+	u32 offset = asyw->image.offset;
+	if (asyh->curs.handle != handle || asyh->curs.offset != offset) {
+		asyh->curs.handle = handle;
+		asyh->curs.offset = offset;
+		asyh->set.curs = asyh->curs.visible;
+	}
+}
+
+static void
+curs507a_release(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
+		 struct nv50_head_atom *asyh)
+{
+	asyh->curs.visible = false;
+}
+
+static int
+curs507a_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
+		 struct nv50_head_atom *asyh)
+{
+	int ret;
+
+	ret = drm_atomic_helper_check_plane_state(&asyw->state, &asyh->state,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  true, true);
+	asyh->curs.visible = asyw->state.visible;
+	if (ret || !asyh->curs.visible)
+		return ret;
+
+	switch (asyw->state.fb->width) {
+	case 32: asyh->curs.layout = 0; break;
+	case 64: asyh->curs.layout = 1; break;
+	default:
+		return -EINVAL;
+	}
+
+	if (asyw->state.fb->width != asyw->state.fb->height)
+		return -EINVAL;
+
+	switch (asyw->state.fb->format->format) {
+	case DRM_FORMAT_ARGB8888: asyh->curs.format = 1; break;
+	default:
+		WARN_ON(1);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static const u32
+curs507a_format[] = {
+	DRM_FORMAT_ARGB8888,
+	0
+};
+
+static const struct nv50_wndw_func
+curs507a_wndw = {
+	.acquire = curs507a_acquire,
+	.release = curs507a_release,
+	.prepare = curs507a_prepare,
+};
+
+static int
+curs507a_new_(const struct nv50_wimm_func *func, struct nouveau_drm *drm,
+	      int head, s32 oclass, struct nv50_wndw **pwndw)
+{
+	struct nv50_disp_cursor_v0 args = {
+		.head = head,
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	struct nv50_wndw *wndw;
+	int ret;
+
+	ret = nv50_wndw_new_(&curs507a_wndw, drm->dev, DRM_PLANE_TYPE_CURSOR,
+			     "curs", head, curs507a_format, &wndw);
+	if (*pwndw = wndw, ret)
+		return ret;
+
+	ret = nvif_object_init(&disp->disp->object, 0, oclass, &args,
+			       sizeof(args), &wndw->wimm.base.user);
+	if (ret) {
+		NV_ERROR(drm, "curs%04x allocation failed: %d\n", oclass, ret);
+		return ret;
+	}
+
+	nvif_object_map(&wndw->wimm.base.user, NULL, 0);
+	wndw->immd = func;
+	wndw->ctxdma.parent = &disp->core->chan.base.user;
+	return 0;
+}
+
+int
+curs507a_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return curs507a_new_(&curs507a, drm, head, oclass, pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/dac507d.c b/drivers/gpu/drm/nouveau/dispnv50/dac507d.c
new file mode 100644
index 0000000..28b6025
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/dac507d.c
@@ -0,0 +1,51 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "core.h"
+
+#include <nvif/class.h>
+
+static void
+dac507d_ctrl(struct nv50_core *core, int or, u32 ctrl,
+	     struct nv50_head_atom *asyh)
+{
+	u32 *push, sync = 0;
+	if ((push = evo_wait(&core->chan, 3))) {
+		if (core->chan.base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
+			if (asyh) {
+				sync |= asyh->or.nvsync << 1;
+				sync |= asyh->or.nhsync;
+			}
+			evo_mthd(push, 0x0400 + (or * 0x080), 2);
+			evo_data(push, ctrl);
+			evo_data(push, sync);
+		} else {
+			evo_mthd(push, 0x0180 + (or * 0x020), 1);
+			evo_data(push, ctrl);
+		}
+		evo_kick(push, &core->chan);
+	}
+}
+
+const struct nv50_outp_func
+dac507d = {
+	.ctrl = dac507d_ctrl,
+};
diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c
index 995109e..a8367c5 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/disp.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c
@@ -21,12 +21,16 @@
  *
  * Authors: Ben Skeggs
  */
+#include "disp.h"
+#include "atom.h"
+#include "core.h"
+#include "head.h"
+#include "wndw.h"
 
 #include <linux/dma-mapping.h>
 #include <linux/hdmi.h>
 
 #include <drm/drmP.h>
-#include <drm/drm_atomic.h>
 #include <drm/drm_atomic_helper.h>
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_dp_helper.h>
@@ -34,2586 +38,188 @@
 #include <drm/drm_plane_helper.h>
 #include <drm/drm_edid.h>
 
-#include <nvif/mem.h>
-
-#include <nvif/class.h>
-#include <nvif/cl0002.h>
-#include <nvif/cl5070.h>
-#include <nvif/cl507a.h>
-#include <nvif/cl507b.h>
-#include <nvif/cl507c.h>
-#include <nvif/cl507d.h>
-#include <nvif/cl507e.h>
-#include <nvif/event.h>
-
-#include "nouveau_drv.h"
-#include "nouveau_dma.h"
-#include "nouveau_gem.h"
-#include "nouveau_connector.h"
-#include "nouveau_encoder.h"
-#include "nouveau_crtc.h"
-#include "nouveau_fence.h"
-#include "nouveau_fbcon.h"
-#include "nv50_display.h"
-
-#define EVO_DMA_NR 9
-
-#define EVO_MASTER  (0x00)
-#define EVO_FLIP(c) (0x01 + (c))
-#define EVO_OVLY(c) (0x05 + (c))
-#define EVO_OIMM(c) (0x09 + (c))
-#define EVO_CURS(c) (0x0d + (c))
-
-/* offsets in shared sync bo of various structures */
-#define EVO_SYNC(c, o) ((c) * 0x0100 + (o))
-#define EVO_MAST_NTFY     EVO_SYNC(      0, 0x00)
-#define EVO_FLIP_SEM0(c)  EVO_SYNC((c) + 1, 0x00)
-#define EVO_FLIP_SEM1(c)  EVO_SYNC((c) + 1, 0x10)
-#define EVO_FLIP_NTFY0(c) EVO_SYNC((c) + 1, 0x20)
-#define EVO_FLIP_NTFY1(c) EVO_SYNC((c) + 1, 0x30)
-
-/******************************************************************************
- * Atomic state
- *****************************************************************************/
-#define nv50_atom(p) container_of((p), struct nv50_atom, state)
-
-struct nv50_atom {
-	struct drm_atomic_state state;
-
-	struct list_head outp;
-	bool lock_core;
-	bool flush_disable;
-};
-
-struct nv50_outp_atom {
-	struct list_head head;
-
-	struct drm_encoder *encoder;
-	bool flush_disable;
-
-	union {
-		struct {
-			bool ctrl:1;
-		};
-		u8 mask;
-	} clr;
-
-	union {
-		struct {
-			bool ctrl:1;
-		};
-		u8 mask;
-	} set;
-};
-
-#define nv50_head_atom(p) container_of((p), struct nv50_head_atom, state)
-
-struct nv50_head_atom {
-	struct drm_crtc_state state;
-
-	struct {
-		u16 iW;
-		u16 iH;
-		u16 oW;
-		u16 oH;
-	} view;
-
-	struct nv50_head_mode {
-		bool interlace;
-		u32 clock;
-		struct {
-			u16 active;
-			u16 synce;
-			u16 blanke;
-			u16 blanks;
-		} h;
-		struct {
-			u32 active;
-			u16 synce;
-			u16 blanke;
-			u16 blanks;
-			u16 blank2s;
-			u16 blank2e;
-			u16 blankus;
-		} v;
-	} mode;
-
-	struct {
-		bool visible;
-		u32 handle;
-		u64 offset:40;
-		u8  mode:4;
-	} lut;
-
-	struct {
-		bool visible;
-		u32 handle;
-		u64 offset:40;
-		u8  format;
-		u8  kind:7;
-		u8  layout:1;
-		u8  block:4;
-		u32 pitch:20;
-		u16 x;
-		u16 y;
-		u16 w;
-		u16 h;
-	} core;
-
-	struct {
-		bool visible;
-		u32 handle;
-		u64 offset:40;
-		u8  layout:1;
-		u8  format:1;
-	} curs;
-
-	struct {
-		u8  depth;
-		u8  cpp;
-		u16 x;
-		u16 y;
-		u16 w;
-		u16 h;
-	} base;
-
-	struct {
-		u8 cpp;
-	} ovly;
-
-	struct {
-		bool enable:1;
-		u8 bits:2;
-		u8 mode:4;
-	} dither;
-
-	struct {
-		struct {
-			u16 cos:12;
-			u16 sin:12;
-		} sat;
-	} procamp;
-
-	struct {
-		u8 nhsync:1;
-		u8 nvsync:1;
-		u8 depth:4;
-	} or;
-
-	union {
-		struct {
-			bool ilut:1;
-			bool core:1;
-			bool curs:1;
-		};
-		u8 mask;
-	} clr;
-
-	union {
-		struct {
-			bool ilut:1;
-			bool core:1;
-			bool curs:1;
-			bool view:1;
-			bool mode:1;
-			bool base:1;
-			bool ovly:1;
-			bool dither:1;
-			bool procamp:1;
-			bool or:1;
-		};
-		u16 mask;
-	} set;
-};
-
-static inline struct nv50_head_atom *
-nv50_head_atom_get(struct drm_atomic_state *state, struct drm_crtc *crtc)
-{
-	struct drm_crtc_state *statec = drm_atomic_get_crtc_state(state, crtc);
-	if (IS_ERR(statec))
-		return (void *)statec;
-	return nv50_head_atom(statec);
-}
-
-#define nv50_wndw_atom(p) container_of((p), struct nv50_wndw_atom, state)
-
-struct nv50_wndw_atom {
-	struct drm_plane_state state;
-	u8 interval;
-
-	struct {
-		u32  handle;
-		u16  offset:12;
-		bool awaken:1;
-	} ntfy;
-
-	struct {
-		u32 handle;
-		u16 offset:12;
-		u32 acquire;
-		u32 release;
-	} sema;
-
-	struct {
-		u8 enable:2;
-	} lut;
-
-	struct {
-		u8  mode:2;
-		u8  interval:4;
-
-		u8  format;
-		u8  kind:7;
-		u8  layout:1;
-		u8  block:4;
-		u32 pitch:20;
-		u16 w;
-		u16 h;
-
-		u32 handle;
-		u64 offset;
-	} image;
-
-	struct {
-		u16 x;
-		u16 y;
-	} point;
-
-	union {
-		struct {
-			bool ntfy:1;
-			bool sema:1;
-			bool image:1;
-		};
-		u8 mask;
-	} clr;
-
-	union {
-		struct {
-			bool ntfy:1;
-			bool sema:1;
-			bool image:1;
-			bool lut:1;
-			bool point:1;
-		};
-		u8 mask;
-	} set;
-};
-
-/******************************************************************************
- * EVO channel
- *****************************************************************************/
-
-struct nv50_chan {
-	struct nvif_object user;
-	struct nvif_device *device;
-};
-
-static int
-nv50_chan_create(struct nvif_device *device, struct nvif_object *disp,
-		 const s32 *oclass, u8 head, void *data, u32 size,
-		 struct nv50_chan *chan)
-{
-	struct nvif_sclass *sclass;
-	int ret, i, n;
-
-	chan->device = device;
-
-	ret = n = nvif_object_sclass_get(disp, &sclass);
-	if (ret < 0)
-		return ret;
-
-	while (oclass[0]) {
-		for (i = 0; i < n; i++) {
-			if (sclass[i].oclass == oclass[0]) {
-				ret = nvif_object_init(disp, 0, oclass[0],
-						       data, size, &chan->user);
-				if (ret == 0)
-					nvif_object_map(&chan->user, NULL, 0);
-				nvif_object_sclass_put(&sclass);
-				return ret;
-			}
-		}
-		oclass++;
-	}
-
-	nvif_object_sclass_put(&sclass);
-	return -ENOSYS;
-}
-
-static void
-nv50_chan_destroy(struct nv50_chan *chan)
-{
-	nvif_object_fini(&chan->user);
-}
-
-/******************************************************************************
- * DMA EVO channel
- *****************************************************************************/
-
-struct nv50_wndw_ctxdma {
-	struct list_head head;
-	struct nvif_object object;
-};
-
-struct nv50_dmac {
-	struct nv50_chan base;
-
-	struct nvif_mem push;
-	u32 *ptr;
-
-	struct nvif_object sync;
-	struct nvif_object vram;
-
-	/* Protects against concurrent pushbuf access to this channel, lock is
-	 * grabbed by evo_wait (if the pushbuf reservation is successful) and
-	 * dropped again by evo_kick. */
-	struct mutex lock;
-};
-
-static void
-nv50_dmac_destroy(struct nv50_dmac *dmac)
-{
-	nvif_object_fini(&dmac->vram);
-	nvif_object_fini(&dmac->sync);
-
-	nv50_chan_destroy(&dmac->base);
-
-	nvif_mem_fini(&dmac->push);
-}
-
-static int
-nv50_dmac_create(struct nvif_device *device, struct nvif_object *disp,
-		 const s32 *oclass, u8 head, void *data, u32 size, u64 syncbuf,
-		 struct nv50_dmac *dmac)
-{
-	struct nouveau_cli *cli = (void *)device->object.client;
-	struct nv50_disp_core_channel_dma_v0 *args = data;
-	int ret;
-
-	mutex_init(&dmac->lock);
-
-	ret = nvif_mem_init_map(&cli->mmu, NVIF_MEM_COHERENT, 0x1000,
-				&dmac->push);
-	if (ret)
-		return ret;
-
-	dmac->ptr = dmac->push.object.map.ptr;
-
-	args->pushbuf = nvif_handle(&dmac->push.object);
-
-	ret = nv50_chan_create(device, disp, oclass, head, data, size,
-			       &dmac->base);
-	if (ret)
-		return ret;
-
-	ret = nvif_object_init(&dmac->base.user, 0xf0000000, NV_DMA_IN_MEMORY,
-			       &(struct nv_dma_v0) {
-					.target = NV_DMA_V0_TARGET_VRAM,
-					.access = NV_DMA_V0_ACCESS_RDWR,
-					.start = syncbuf + 0x0000,
-					.limit = syncbuf + 0x0fff,
-			       }, sizeof(struct nv_dma_v0),
-			       &dmac->sync);
-	if (ret)
-		return ret;
-
-	ret = nvif_object_init(&dmac->base.user, 0xf0000001, NV_DMA_IN_MEMORY,
-			       &(struct nv_dma_v0) {
-					.target = NV_DMA_V0_TARGET_VRAM,
-					.access = NV_DMA_V0_ACCESS_RDWR,
-					.start = 0,
-					.limit = device->info.ram_user - 1,
-			       }, sizeof(struct nv_dma_v0),
-			       &dmac->vram);
-	if (ret)
-		return ret;
-
-	return ret;
-}
-
-/******************************************************************************
- * Base
- *****************************************************************************/
-
-struct nv50_sync {
-	struct nv50_dmac base;
-	u32 addr;
-	u32 data;
-};
-
-struct nv50_head {
-	const struct nv50_head_func *func;
-	struct nouveau_crtc base;
-	struct {
-		struct nouveau_bo *nvbo[2];
-		int next;
-	} lut;
-};
-
-struct nv50_head_func {
-	void (*view)(struct nv50_head *, struct nv50_head_atom *);
-	void (*mode)(struct nv50_head *, struct nv50_head_atom *);
-	void (*ilut_set)(struct nv50_head *, struct nv50_head_atom *);
-	void (*ilut_clr)(struct nv50_head *);
-	void (*core_set)(struct nv50_head *, struct nv50_head_atom *);
-	void (*core_clr)(struct nv50_head *);
-	void (*curs_set)(struct nv50_head *, struct nv50_head_atom *);
-	void (*curs_clr)(struct nv50_head *);
-	void (*base)(struct nv50_head *, struct nv50_head_atom *);
-	void (*ovly)(struct nv50_head *, struct nv50_head_atom *);
-	void (*dither)(struct nv50_head *, struct nv50_head_atom *);
-	void (*procamp)(struct nv50_head *, struct nv50_head_atom *);
-	void (*or)(struct nv50_head *, struct nv50_head_atom *);
-};
-
-#define nv50_head(c) container_of((c), struct nv50_head, base.base)
-
-struct nv50_disp {
-	struct nvif_disp *disp;
-	struct nv50_core *core;
-
-	struct nouveau_bo *sync;
-
-	struct mutex mutex;
-};
-
-static struct nv50_disp *
-nv50_disp(struct drm_device *dev)
-{
-	return nouveau_display(dev)->priv;
-}
-
-/******************************************************************************
- * Core
- *****************************************************************************/
-
-struct nv50_core {
-	const struct nv50_core_func *func;
-	struct nv50_dmac chan;
-};
-
-struct nv50_core_func {
-	const struct nv50_head_func *head;
-	const struct nv50_outp_func *dac;
-	const struct nv50_outp_func *sor;
-	const struct nv50_outp_func *pior;
-};
-
-struct nv50_outp_func {
-	void (*ctrl)(struct nv50_core *, int or, u32 ctrl,
-		     struct nv50_head_atom *);
-};
-
-static int
-core507d_new_(const struct nv50_core_func *func, struct nouveau_drm *drm,
-	      s32 oclass, struct nv50_core **pcore)
-{
-	struct nv50_disp_core_channel_dma_v0 args = {};
-	struct nv50_disp *disp = nv50_disp(drm->dev);
-	struct nv50_core *core;
-	int ret;
-
-	if (!(core = *pcore = kzalloc(sizeof(*core), GFP_KERNEL)))
-		return -ENOMEM;
-	core->func = func;
-
-	ret = nv50_dmac_create(&drm->client.device, &disp->disp->object,
-			       &oclass, 0, &args, sizeof(args),
-			       disp->sync->bo.offset, &core->chan);
-	if (ret) {
-		NV_ERROR(drm, "core%04x allocation failed: %d\n", oclass, ret);
-		return ret;
-	}
-
-	return 0;
-}
-
-/******************************************************************************
- * EVO channel helpers
- *****************************************************************************/
-static u32 *
-evo_wait(void *evoc, int nr)
-{
-	struct nv50_dmac *dmac = evoc;
-	struct nvif_device *device = dmac->base.device;
-	u32 put = nvif_rd32(&dmac->base.user, 0x0000) / 4;
-
-	mutex_lock(&dmac->lock);
-	if (put + nr >= (PAGE_SIZE / 4) - 8) {
-		dmac->ptr[put] = 0x20000000;
-
-		nvif_wr32(&dmac->base.user, 0x0000, 0x00000000);
-		if (nvif_msec(device, 2000,
-			if (!nvif_rd32(&dmac->base.user, 0x0004))
-				break;
-		) < 0) {
-			mutex_unlock(&dmac->lock);
-			pr_err("nouveau: evo channel stalled\n");
-			return NULL;
-		}
-
-		put = 0;
-	}
-
-	return dmac->ptr + put;
-}
-
-static void
-evo_kick(u32 *push, void *evoc)
-{
-	struct nv50_dmac *dmac = evoc;
-	nvif_wr32(&dmac->base.user, 0x0000, (push - dmac->ptr) << 2);
-	mutex_unlock(&dmac->lock);
-}
-
-#define evo_mthd(p, m, s) do {						\
-	const u32 _m = (m), _s = (s);					\
-	if (drm_debug & DRM_UT_KMS)					\
-		pr_err("%04x %d %s\n", _m, _s, __func__);		\
-	*((p)++) = ((_s << 18) | _m);					\
-} while(0)
-
-#define evo_data(p, d) do {						\
-	const u32 _d = (d);						\
-	if (drm_debug & DRM_UT_KMS)					\
-		pr_err("\t%08x\n", _d);					\
-	*((p)++) = _d;							\
-} while(0)
-
-/******************************************************************************
- * Plane
- *****************************************************************************/
-#define nv50_wndw(p) container_of((p), struct nv50_wndw, plane)
-
-struct nv50_wndw {
-	const struct nv50_wndw_func *func;
-	const struct nv50_wimm_func *immd;
-	int id;
-
-	struct {
-		struct nvif_object *parent;
-		struct list_head list;
-	} ctxdma;
-
-	struct drm_plane plane;
-
-	struct nv50_dmac wndw;
-	struct nv50_dmac wimm;
-
-	struct nvif_notify notify;
-	u16 ntfy;
-	u16 sema;
-	u32 data;
-};
-
-struct nv50_wndw_func {
-	int (*acquire)(struct nv50_wndw *, struct nv50_wndw_atom *asyw,
-		       struct nv50_head_atom *asyh);
-	void (*release)(struct nv50_wndw *, struct nv50_wndw_atom *asyw,
-			struct nv50_head_atom *asyh);
-	void (*prepare)(struct nv50_wndw *, struct nv50_head_atom *asyh,
-			struct nv50_wndw_atom *asyw);
-
-	void (*sema_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
-	void (*sema_clr)(struct nv50_wndw *);
-	void (*ntfy_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
-	void (*ntfy_clr)(struct nv50_wndw *);
-	int (*ntfy_wait_begun)(struct nv50_wndw *, struct nv50_wndw_atom *);
-	void (*image_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
-	void (*image_clr)(struct nv50_wndw *);
-	void (*lut)(struct nv50_wndw *, struct nv50_wndw_atom *);
-
-	u32 (*update)(struct nv50_wndw *, u32 interlock);
-};
-
-struct nv50_wimm_func {
-	void (*point)(struct nv50_wndw *, struct nv50_wndw_atom *);
-
-	u32 (*update)(struct nv50_wndw *, u32 interlock);
-};
-
-static void
-nv50_wndw_ctxdma_del(struct nv50_wndw_ctxdma *ctxdma)
-{
-	nvif_object_fini(&ctxdma->object);
-	list_del(&ctxdma->head);
-	kfree(ctxdma);
-}
-
-static struct nv50_wndw_ctxdma *
-nv50_wndw_ctxdma_new(struct nv50_wndw *wndw, struct nouveau_framebuffer *fb)
-{
-	struct nouveau_drm *drm = nouveau_drm(fb->base.dev);
-	struct nv50_wndw_ctxdma *ctxdma;
-	const u8    kind = fb->nvbo->kind;
-	const u32 handle = 0xfb000000 | kind;
-	struct {
-		struct nv_dma_v0 base;
-		union {
-			struct nv50_dma_v0 nv50;
-			struct gf100_dma_v0 gf100;
-			struct gf119_dma_v0 gf119;
-		};
-	} args = {};
-	u32 argc = sizeof(args.base);
-	int ret;
-
-	list_for_each_entry(ctxdma, &wndw->ctxdma.list, head) {
-		if (ctxdma->object.handle == handle)
-			return ctxdma;
-	}
-
-	if (!(ctxdma = kzalloc(sizeof(*ctxdma), GFP_KERNEL)))
-		return ERR_PTR(-ENOMEM);
-	list_add(&ctxdma->head, &wndw->ctxdma.list);
-
-	args.base.target = NV_DMA_V0_TARGET_VRAM;
-	args.base.access = NV_DMA_V0_ACCESS_RDWR;
-	args.base.start  = 0;
-	args.base.limit  = drm->client.device.info.ram_user - 1;
-
-	if (drm->client.device.info.chipset < 0x80) {
-		args.nv50.part = NV50_DMA_V0_PART_256;
-		argc += sizeof(args.nv50);
-	} else
-	if (drm->client.device.info.chipset < 0xc0) {
-		args.nv50.part = NV50_DMA_V0_PART_256;
-		args.nv50.kind = kind;
-		argc += sizeof(args.nv50);
-	} else
-	if (drm->client.device.info.chipset < 0xd0) {
-		args.gf100.kind = kind;
-		argc += sizeof(args.gf100);
-	} else {
-		args.gf119.page = GF119_DMA_V0_PAGE_LP;
-		args.gf119.kind = kind;
-		argc += sizeof(args.gf119);
-	}
-
-	ret = nvif_object_init(wndw->ctxdma.parent, handle, NV_DMA_IN_MEMORY,
-			       &args, argc, &ctxdma->object);
-	if (ret) {
-		nv50_wndw_ctxdma_del(ctxdma);
-		return ERR_PTR(ret);
-	}
-
-	return ctxdma;
-}
-
-static int
-nv50_wndw_wait_armed(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
-{
-	if (asyw->set.ntfy)
-		return wndw->func->ntfy_wait_begun(wndw, asyw);
-	return 0;
-}
-
-static u32
-nv50_wndw_flush_clr(struct nv50_wndw *wndw, u32 interlock, bool flush,
-		    struct nv50_wndw_atom *asyw)
-{
-	if (asyw->clr.sema && (!asyw->set.sema || flush))
-		wndw->func->sema_clr(wndw);
-	if (asyw->clr.ntfy && (!asyw->set.ntfy || flush))
-		wndw->func->ntfy_clr(wndw);
-	if (asyw->clr.image && (!asyw->set.image || flush))
-		wndw->func->image_clr(wndw);
-
-	return flush ? wndw->func->update(wndw, interlock) : 0;
-}
-
-static u32
-nv50_wndw_flush_set(struct nv50_wndw *wndw, u32 interlock,
-		    struct nv50_wndw_atom *asyw)
-{
-	if (interlock) {
-		asyw->image.mode = 0;
-		asyw->image.interval = 1;
-	}
-
-	if (asyw->set.sema ) wndw->func->sema_set (wndw, asyw);
-	if (asyw->set.ntfy ) wndw->func->ntfy_set (wndw, asyw);
-	if (asyw->set.image) wndw->func->image_set(wndw, asyw);
-	if (asyw->set.lut  ) wndw->func->lut      (wndw, asyw);
-	if (asyw->set.point) {
-		wndw->immd->point(wndw, asyw);
-		wndw->immd->update(wndw, interlock);
-	}
-
-	return wndw->func->update ? wndw->func->update(wndw, interlock) : 0;
-}
-
-static void
-nv50_wndw_atomic_check_release(struct nv50_wndw *wndw,
-			       struct nv50_wndw_atom *asyw,
-			       struct nv50_head_atom *asyh)
-{
-	struct nouveau_drm *drm = nouveau_drm(wndw->plane.dev);
-	NV_ATOMIC(drm, "%s release\n", wndw->plane.name);
-	wndw->func->release(wndw, asyw, asyh);
-	asyw->ntfy.handle = 0;
-	asyw->sema.handle = 0;
-}
-
-static int
-nv50_wndw_atomic_check_acquire(struct nv50_wndw *wndw,
-			       struct nv50_wndw_atom *asyw,
-			       struct nv50_head_atom *asyh)
-{
-	struct nouveau_framebuffer *fb = nouveau_framebuffer(asyw->state.fb);
-	struct nouveau_drm *drm = nouveau_drm(wndw->plane.dev);
-	int ret;
-
-	NV_ATOMIC(drm, "%s acquire\n", wndw->plane.name);
-
-	asyw->image.w = fb->base.width;
-	asyw->image.h = fb->base.height;
-	asyw->image.kind = fb->nvbo->kind;
-
-	if (asyh->state.pageflip_flags & DRM_MODE_PAGE_FLIP_ASYNC)
-		asyw->interval = 0;
-	else
-		asyw->interval = 1;
-
-	if (asyw->image.kind) {
-		asyw->image.layout = 0;
-		if (drm->client.device.info.chipset >= 0xc0)
-			asyw->image.block = fb->nvbo->mode >> 4;
-		else
-			asyw->image.block = fb->nvbo->mode;
-		asyw->image.pitch = (fb->base.pitches[0] / 4) << 4;
-	} else {
-		asyw->image.layout = 1;
-		asyw->image.block  = 0;
-		asyw->image.pitch  = fb->base.pitches[0];
-	}
-
-	ret = wndw->func->acquire(wndw, asyw, asyh);
-	if (ret)
-		return ret;
-
-	if (asyw->set.image) {
-		if (!(asyw->image.mode = asyw->interval ? 0 : 1))
-			asyw->image.interval = asyw->interval;
-		else
-			asyw->image.interval = 0;
-	}
-
-	return 0;
-}
-
-static int
-nv50_wndw_atomic_check(struct drm_plane *plane, struct drm_plane_state *state)
-{
-	struct nouveau_drm *drm = nouveau_drm(plane->dev);
-	struct nv50_wndw *wndw = nv50_wndw(plane);
-	struct nv50_wndw_atom *armw = nv50_wndw_atom(wndw->plane.state);
-	struct nv50_wndw_atom *asyw = nv50_wndw_atom(state);
-	struct nv50_head_atom *harm = NULL, *asyh = NULL;
-	bool varm = false, asyv = false, asym = false;
-	int ret;
-
-	NV_ATOMIC(drm, "%s atomic_check\n", plane->name);
-	if (asyw->state.crtc) {
-		asyh = nv50_head_atom_get(asyw->state.state, asyw->state.crtc);
-		if (IS_ERR(asyh))
-			return PTR_ERR(asyh);
-		asym = drm_atomic_crtc_needs_modeset(&asyh->state);
-		asyv = asyh->state.active;
-	}
-
-	if (armw->state.crtc) {
-		harm = nv50_head_atom_get(asyw->state.state, armw->state.crtc);
-		if (IS_ERR(harm))
-			return PTR_ERR(harm);
-		varm = harm->state.crtc->state->active;
-	}
-
-	if (asyv) {
-		asyw->point.x = asyw->state.crtc_x;
-		asyw->point.y = asyw->state.crtc_y;
-		if (memcmp(&armw->point, &asyw->point, sizeof(asyw->point)))
-			asyw->set.point = true;
-
-		ret = nv50_wndw_atomic_check_acquire(wndw, asyw, asyh);
-		if (ret)
-			return ret;
-	} else
-	if (varm) {
-		nv50_wndw_atomic_check_release(wndw, asyw, harm);
-	} else {
-		return 0;
-	}
-
-	if (!asyv || asym) {
-		asyw->clr.ntfy = armw->ntfy.handle != 0;
-		asyw->clr.sema = armw->sema.handle != 0;
-		if (wndw->func->image_clr)
-			asyw->clr.image = armw->image.handle != 0;
-		asyw->set.lut = wndw->func->lut && asyv;
-	}
-
-	return 0;
-}
-
-static void
-nv50_wndw_cleanup_fb(struct drm_plane *plane, struct drm_plane_state *old_state)
-{
-	struct nouveau_framebuffer *fb = nouveau_framebuffer(old_state->fb);
-	struct nouveau_drm *drm = nouveau_drm(plane->dev);
-
-	NV_ATOMIC(drm, "%s cleanup: %p\n", plane->name, old_state->fb);
-	if (!old_state->fb)
-		return;
-
-	nouveau_bo_unpin(fb->nvbo);
-}
-
-static int
-nv50_wndw_prepare_fb(struct drm_plane *plane, struct drm_plane_state *state)
-{
-	struct nouveau_framebuffer *fb = nouveau_framebuffer(state->fb);
-	struct nouveau_drm *drm = nouveau_drm(plane->dev);
-	struct nv50_wndw *wndw = nv50_wndw(plane);
-	struct nv50_wndw_atom *asyw = nv50_wndw_atom(state);
-	struct nv50_head_atom *asyh;
-	struct nv50_wndw_ctxdma *ctxdma;
-	int ret;
-
-	NV_ATOMIC(drm, "%s prepare: %p\n", plane->name, state->fb);
-	if (!asyw->state.fb)
-		return 0;
-
-	ret = nouveau_bo_pin(fb->nvbo, TTM_PL_FLAG_VRAM, true);
-	if (ret)
-		return ret;
-
-	ctxdma = nv50_wndw_ctxdma_new(wndw, fb);
-	if (IS_ERR(ctxdma)) {
-		nouveau_bo_unpin(fb->nvbo);
-		return PTR_ERR(ctxdma);
-	}
-
-	asyw->state.fence = reservation_object_get_excl_rcu(fb->nvbo->bo.resv);
-	asyw->image.handle = ctxdma->object.handle;
-	asyw->image.offset = fb->nvbo->bo.offset;
-
-	if (wndw->func->prepare) {
-		asyh = nv50_head_atom_get(asyw->state.state, asyw->state.crtc);
-		if (IS_ERR(asyh))
-			return PTR_ERR(asyh);
-
-		wndw->func->prepare(wndw, asyh, asyw);
-	}
-
-	return 0;
-}
-
-static const struct drm_plane_helper_funcs
-nv50_wndw_helper = {
-	.prepare_fb = nv50_wndw_prepare_fb,
-	.cleanup_fb = nv50_wndw_cleanup_fb,
-	.atomic_check = nv50_wndw_atomic_check,
-};
-
-static void
-nv50_wndw_atomic_destroy_state(struct drm_plane *plane,
-			       struct drm_plane_state *state)
-{
-	struct nv50_wndw_atom *asyw = nv50_wndw_atom(state);
-	__drm_atomic_helper_plane_destroy_state(&asyw->state);
-	kfree(asyw);
-}
-
-static struct drm_plane_state *
-nv50_wndw_atomic_duplicate_state(struct drm_plane *plane)
-{
-	struct nv50_wndw_atom *armw = nv50_wndw_atom(plane->state);
-	struct nv50_wndw_atom *asyw;
-	if (!(asyw = kmalloc(sizeof(*asyw), GFP_KERNEL)))
-		return NULL;
-	__drm_atomic_helper_plane_duplicate_state(plane, &asyw->state);
-	asyw->interval = 1;
-	asyw->sema = armw->sema;
-	asyw->ntfy = armw->ntfy;
-	asyw->image = armw->image;
-	asyw->point = armw->point;
-	asyw->lut = armw->lut;
-	asyw->clr.mask = 0;
-	asyw->set.mask = 0;
-	return &asyw->state;
-}
-
-static void
-nv50_wndw_reset(struct drm_plane *plane)
-{
-	struct nv50_wndw_atom *asyw;
-
-	if (WARN_ON(!(asyw = kzalloc(sizeof(*asyw), GFP_KERNEL))))
-		return;
-
-	if (plane->state)
-		plane->funcs->atomic_destroy_state(plane, plane->state);
-	plane->state = &asyw->state;
-	plane->state->plane = plane;
-	plane->state->rotation = DRM_MODE_ROTATE_0;
-}
-
-static void
-nv50_wndw_destroy(struct drm_plane *plane)
-{
-	struct nv50_wndw *wndw = nv50_wndw(plane);
-	struct nv50_wndw_ctxdma *ctxdma, *ctxtmp;
-
-	list_for_each_entry_safe(ctxdma, ctxtmp, &wndw->ctxdma.list, head) {
-		nv50_wndw_ctxdma_del(ctxdma);
-	}
-
-	nvif_notify_fini(&wndw->notify);
-	nv50_dmac_destroy(&wndw->wimm);
-	nv50_dmac_destroy(&wndw->wndw);
-	drm_plane_cleanup(&wndw->plane);
-	kfree(wndw);
-}
-
-static const struct drm_plane_funcs
-nv50_wndw = {
-	.update_plane = drm_atomic_helper_update_plane,
-	.disable_plane = drm_atomic_helper_disable_plane,
-	.destroy = nv50_wndw_destroy,
-	.reset = nv50_wndw_reset,
-	.atomic_duplicate_state = nv50_wndw_atomic_duplicate_state,
-	.atomic_destroy_state = nv50_wndw_atomic_destroy_state,
-};
-
-static int
-nv50_wndw_notify(struct nvif_notify *notify)
-{
-	return NVIF_NOTIFY_KEEP;
-}
-
-static void
-nv50_wndw_fini(struct nv50_wndw *wndw)
-{
-	nvif_notify_put(&wndw->notify);
-}
-
-static void
-nv50_wndw_init(struct nv50_wndw *wndw)
-{
-	nvif_notify_get(&wndw->notify);
-}
-
-static int
-nv50_wndw_new_(const struct nv50_wndw_func *func, struct drm_device *dev,
-	       enum drm_plane_type type, const char *name, int index,
-	       const u32 *format, struct nv50_wndw **pwndw)
-{
-	struct nv50_wndw *wndw;
-	int nformat;
-	int ret;
-
-	if (!(wndw = *pwndw = kzalloc(sizeof(*wndw), GFP_KERNEL)))
-		return -ENOMEM;
-	wndw->func = func;
-	wndw->id = index;
-
-	wndw->ctxdma.parent = &wndw->wndw.base.user;
-	INIT_LIST_HEAD(&wndw->ctxdma.list);
-
-	for (nformat = 0; format[nformat]; nformat++);
-
-	ret = drm_universal_plane_init(dev, &wndw->plane, 0, &nv50_wndw,
-				       format, nformat, NULL,
-				       type, "%s-%d", name, index);
-	if (ret) {
-		kfree(*pwndw);
-		*pwndw = NULL;
-		return ret;
-	}
-
-	drm_plane_helper_add(&wndw->plane, &nv50_wndw_helper);
-
-	wndw->notify.func = nv50_wndw_notify;
-	return 0;
-}
-
-/******************************************************************************
- * Overlay
- *****************************************************************************/
-
-static const struct nv50_wimm_func
-oimm507b = {
-};
-
-static int
-oimm507b_init_(const struct nv50_wimm_func *func, struct nouveau_drm *drm,
-	       s32 oclass, struct nv50_wndw *wndw)
-{
-	struct nv50_disp_overlay_v0 args = {
-		.head = wndw->id,
-	};
-	struct nv50_disp *disp = nv50_disp(drm->dev);
-	int ret;
-
-	ret = nvif_object_init(&disp->disp->object, 0, oclass, &args,
-			       sizeof(args), &wndw->wimm.base.user);
-	if (ret) {
-		NV_ERROR(drm, "oimm%04x allocation failed: %d\n", oclass, ret);
-		return ret;
-	}
-
-	nvif_object_map(&wndw->wimm.base.user, NULL, 0);
-	wndw->immd = func;
-	return 0;
-}
-
-static int
-oimm507b_init(struct nouveau_drm *drm, s32 oclass, struct nv50_wndw *wndw)
-{
-	return oimm507b_init_(&oimm507b, drm, oclass, wndw);
-}
-
-static int
-nv50_oimm_init(struct nouveau_drm *drm, struct nv50_wndw *wndw)
-{
-	static const struct {
-		s32 oclass;
-		int version;
-		int (*init)(struct nouveau_drm *, s32, struct nv50_wndw *);
-	} oimms[] = {
-		{ GK104_DISP_OVERLAY, 0, oimm507b_init },
-		{ GF110_DISP_OVERLAY, 0, oimm507b_init },
-		{ GT214_DISP_OVERLAY, 0, oimm507b_init },
-		{   G82_DISP_OVERLAY, 0, oimm507b_init },
-		{  NV50_DISP_OVERLAY, 0, oimm507b_init },
-		{}
-	};
-	struct nv50_disp *disp = nv50_disp(drm->dev);
-	int cid;
-
-	cid = nvif_mclass(&disp->disp->object, oimms);
-	if (cid < 0) {
-		NV_ERROR(drm, "No supported overlay immediate class\n");
-		return cid;
-	}
-
-	return oimms[cid].init(drm, oimms[cid].oclass, wndw);
-}
-
-static const struct nv50_wndw_func
-ovly507e = {
-};
-
-static const u32
-ovly507e_format[] = {
-	0
-};
-
-static int
-ovly507e_new_(const struct nv50_wndw_func *func, const u32 *format,
-	      struct nouveau_drm *drm, int head, s32 oclass,
-	      struct nv50_wndw **pwndw)
-{
-	struct nv50_disp_overlay_channel_dma_v0 args = {
-		.head = head,
-	};
-	struct nv50_disp *disp = nv50_disp(drm->dev);
-	struct nv50_wndw *wndw;
-	int ret;
-
-	ret = nv50_wndw_new_(func, drm->dev, DRM_PLANE_TYPE_OVERLAY,
-			     "ovly", head, format, &wndw);
-	if (*pwndw = wndw, ret)
-		return ret;
-
-	ret = nv50_dmac_create(&drm->client.device, &disp->disp->object,
-			       &oclass, 0, &args, sizeof(args),
-			       disp->sync->bo.offset, &wndw->wndw);
-	if (ret) {
-		NV_ERROR(drm, "ovly%04x allocation failed: %d\n", oclass, ret);
-		return ret;
-	}
-
-	return 0;
-}
-
-static int
-ovly507e_new(struct nouveau_drm *drm, int head, s32 oclass,
-	     struct nv50_wndw **pwndw)
-{
-	return ovly507e_new_(&ovly507e, ovly507e_format, drm, head, oclass, pwndw);
-}
-
-static int
-nv50_ovly_new(struct nouveau_drm *drm, int head, struct nv50_wndw **pwndw)
-{
-	static const struct {
-		s32 oclass;
-		int version;
-		int (*new)(struct nouveau_drm *, int, s32, struct nv50_wndw **);
-	} ovlys[] = {
-		{ GK104_DISP_OVERLAY_CONTROL_DMA, 0, ovly507e_new },
-		{ GF110_DISP_OVERLAY_CONTROL_DMA, 0, ovly507e_new },
-		{ GT214_DISP_OVERLAY_CHANNEL_DMA, 0, ovly507e_new },
-		{ GT200_DISP_OVERLAY_CHANNEL_DMA, 0, ovly507e_new },
-		{   G82_DISP_OVERLAY_CHANNEL_DMA, 0, ovly507e_new },
-		{  NV50_DISP_OVERLAY_CHANNEL_DMA, 0, ovly507e_new },
-		{}
-	};
-	struct nv50_disp *disp = nv50_disp(drm->dev);
-	int cid, ret;
-
-	cid = nvif_mclass(&disp->disp->object, ovlys);
-	if (cid < 0) {
-		NV_ERROR(drm, "No supported overlay class\n");
-		return cid;
-	}
-
-	ret = ovlys[cid].new(drm, head, ovlys[cid].oclass, pwndw);
-	if (ret)
-		return ret;
-
-	return nv50_oimm_init(drm, *pwndw);
-}
-
-/******************************************************************************
- * Cursor plane
- *****************************************************************************/
-static u32
-nv50_curs_update(struct nv50_wndw *wndw, u32 interlock)
-{
-	nvif_wr32(&wndw->wimm.base.user, 0x0080, 0x00000000);
-	return 0;
-}
-
-static void
-nv50_curs_point(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
-{
-	nvif_wr32(&wndw->wimm.base.user, 0x0084, (asyw->point.y << 16) |
-						  asyw->point.x);
-}
-
-static const struct nv50_wimm_func
-curs507a = {
-	.point = nv50_curs_point,
-	.update = nv50_curs_update,
-};
-
-static void
-nv50_curs_prepare(struct nv50_wndw *wndw, struct nv50_head_atom *asyh,
-		  struct nv50_wndw_atom *asyw)
-{
-	u32 handle = nv50_disp(wndw->plane.dev)->core->chan.vram.handle;
-	u32 offset = asyw->image.offset;
-	if (asyh->curs.handle != handle || asyh->curs.offset != offset) {
-		asyh->curs.handle = handle;
-		asyh->curs.offset = offset;
-		asyh->set.curs = asyh->curs.visible;
-	}
-}
-
-static void
-nv50_curs_release(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
-		  struct nv50_head_atom *asyh)
-{
-	asyh->curs.visible = false;
-}
-
-static int
-nv50_curs_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
-		  struct nv50_head_atom *asyh)
-{
-	int ret;
-
-	ret = drm_atomic_helper_check_plane_state(&asyw->state, &asyh->state,
-						  DRM_PLANE_HELPER_NO_SCALING,
-						  DRM_PLANE_HELPER_NO_SCALING,
-						  true, true);
-	asyh->curs.visible = asyw->state.visible;
-	if (ret || !asyh->curs.visible)
-		return ret;
-
-	switch (asyw->state.fb->width) {
-	case 32: asyh->curs.layout = 0; break;
-	case 64: asyh->curs.layout = 1; break;
-	default:
-		return -EINVAL;
-	}
-
-	if (asyw->state.fb->width != asyw->state.fb->height)
-		return -EINVAL;
-
-	switch (asyw->state.fb->format->format) {
-	case DRM_FORMAT_ARGB8888: asyh->curs.format = 1; break;
-	default:
-		WARN_ON(1);
-		return -EINVAL;
-	}
-
-	return 0;
-}
-
-static const u32
-nv50_curs_format[] = {
-	DRM_FORMAT_ARGB8888,
-	0
-};
-
-static const struct nv50_wndw_func
-nv50_curs = {
-	.acquire = nv50_curs_acquire,
-	.release = nv50_curs_release,
-	.prepare = nv50_curs_prepare,
-};
-
-static int
-curs507a_new_(const struct nv50_wimm_func *func, struct nouveau_drm *drm,
-	      int head, s32 oclass, struct nv50_wndw **pwndw)
-{
-	struct nv50_disp_cursor_v0 args = {
-		.head = head,
-	};
-	struct nv50_disp *disp = nv50_disp(drm->dev);
-	struct nv50_wndw *wndw;
-	int ret;
-
-	ret = nv50_wndw_new_(&nv50_curs, drm->dev, DRM_PLANE_TYPE_CURSOR,
-			     "curs", head, nv50_curs_format, &wndw);
-	if (*pwndw = wndw, ret)
-		return ret;
-
-	ret = nvif_object_init(&disp->disp->object, 0, oclass, &args,
-			       sizeof(args), &wndw->wimm.base.user);
-	if (ret) {
-		NV_ERROR(drm, "curs%04x allocation failed: %d\n", oclass, ret);
-		return ret;
-	}
-
-	nvif_object_map(&wndw->wimm.base.user, NULL, 0);
-	wndw->immd = func;
-	wndw->ctxdma.parent = &disp->core->chan.base.user;
-	return 0;
-}
-
-static int
-curs507a_new(struct nouveau_drm *drm, int head, s32 oclass,
-	     struct nv50_wndw **pwndw)
-{
-	return curs507a_new_(&curs507a, drm, head, oclass, pwndw);
-}
-
-static int
-nv50_curs_new(struct nouveau_drm *drm, int head, struct nv50_wndw **pwndw)
-{
-	struct {
-		s32 oclass;
-		int version;
-		int (*new)(struct nouveau_drm *, int, s32, struct nv50_wndw **);
-	} curses[] = {
-		{ GK104_DISP_CURSOR, 0, curs507a_new },
-		{ GF110_DISP_CURSOR, 0, curs507a_new },
-		{ GT214_DISP_CURSOR, 0, curs507a_new },
-		{   G82_DISP_CURSOR, 0, curs507a_new },
-		{  NV50_DISP_CURSOR, 0, curs507a_new },
-		{}
-	};
-	struct nv50_disp *disp = nv50_disp(drm->dev);
-	int cid;
-
-	cid = nvif_mclass(&disp->disp->object, curses);
-	if (cid < 0) {
-		NV_ERROR(drm, "No supported cursor immediate class\n");
-		return cid;
-	}
-
-	return curses[cid].new(drm, head, curses[cid].oclass, pwndw);
-}
-
-/******************************************************************************
- * Primary plane
- *****************************************************************************/
-static void
-nv50_base_lut(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
-{
-	u32 *push;
-	if ((push = evo_wait(&wndw->wndw, 2))) {
-		evo_mthd(push, 0x00e0, 1);
-		evo_data(push, asyw->lut.enable << 30);
-		evo_kick(push, &wndw->wndw);
-	}
-}
-
-static void
-nv50_base_image_clr(struct nv50_wndw *wndw)
-{
-	u32 *push;
-	if ((push = evo_wait(&wndw->wndw, 4))) {
-		evo_mthd(push, 0x0084, 1);
-		evo_data(push, 0x00000000);
-		evo_mthd(push, 0x00c0, 1);
-		evo_data(push, 0x00000000);
-		evo_kick(push, &wndw->wndw);
-	}
-}
-
-static void
-nv50_base_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
-{
-	const s32 oclass = wndw->wndw.base.user.oclass;
-	u32 *push;
-	if ((push = evo_wait(&wndw->wndw, 10))) {
-		evo_mthd(push, 0x0084, 1);
-		evo_data(push, (asyw->image.mode << 8) |
-			       (asyw->image.interval << 4));
-		evo_mthd(push, 0x00c0, 1);
-		evo_data(push, asyw->image.handle);
-		if (oclass < G82_DISP_BASE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0800, 5);
-			evo_data(push, asyw->image.offset >> 8);
-			evo_data(push, 0x00000000);
-			evo_data(push, (asyw->image.h << 16) | asyw->image.w);
-			evo_data(push, (asyw->image.layout << 20) |
-					asyw->image.pitch |
-					asyw->image.block);
-			evo_data(push, (asyw->image.kind << 16) |
-				       (asyw->image.format << 8));
-		} else
-		if (oclass < GF110_DISP_BASE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0800, 5);
-			evo_data(push, asyw->image.offset >> 8);
-			evo_data(push, 0x00000000);
-			evo_data(push, (asyw->image.h << 16) | asyw->image.w);
-			evo_data(push, (asyw->image.layout << 20) |
-					asyw->image.pitch |
-					asyw->image.block);
-			evo_data(push, asyw->image.format << 8);
-		} else {
-			evo_mthd(push, 0x0400, 5);
-			evo_data(push, asyw->image.offset >> 8);
-			evo_data(push, 0x00000000);
-			evo_data(push, (asyw->image.h << 16) | asyw->image.w);
-			evo_data(push, (asyw->image.layout << 24) |
-					asyw->image.pitch |
-					asyw->image.block);
-			evo_data(push, asyw->image.format << 8);
-		}
-		evo_kick(push, &wndw->wndw);
-	}
-}
-
-static void
-nv50_base_ntfy_clr(struct nv50_wndw *wndw)
-{
-	u32 *push;
-	if ((push = evo_wait(&wndw->wndw, 2))) {
-		evo_mthd(push, 0x00a4, 1);
-		evo_data(push, 0x00000000);
-		evo_kick(push, &wndw->wndw);
-	}
-}
-
-static void
-nv50_base_ntfy_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
-{
-	u32 *push;
-	if ((push = evo_wait(&wndw->wndw, 3))) {
-		evo_mthd(push, 0x00a0, 2);
-		evo_data(push, (asyw->ntfy.awaken << 30) | asyw->ntfy.offset);
-		evo_data(push, asyw->ntfy.handle);
-		evo_kick(push, &wndw->wndw);
-	}
-}
-
-static void
-nv50_base_sema_clr(struct nv50_wndw *wndw)
-{
-	u32 *push;
-	if ((push = evo_wait(&wndw->wndw, 2))) {
-		evo_mthd(push, 0x0094, 1);
-		evo_data(push, 0x00000000);
-		evo_kick(push, &wndw->wndw);
-	}
-}
-
-static void
-nv50_base_sema_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
-{
-	u32 *push;
-	if ((push = evo_wait(&wndw->wndw, 5))) {
-		evo_mthd(push, 0x0088, 4);
-		evo_data(push, asyw->sema.offset);
-		evo_data(push, asyw->sema.acquire);
-		evo_data(push, asyw->sema.release);
-		evo_data(push, asyw->sema.handle);
-		evo_kick(push, &wndw->wndw);
-	}
-}
-
-static u32
-nv50_base_update(struct nv50_wndw *wndw, u32 interlock)
-{
-	u32 *push;
-
-	if (!(push = evo_wait(&wndw->wndw, 2)))
-		return 0;
-	evo_mthd(push, 0x0080, 1);
-	evo_data(push, interlock);
-	evo_kick(push, &wndw->wndw);
-
-	if (wndw->wndw.base.user.oclass < GF110_DISP_BASE_CHANNEL_DMA)
-		return interlock ? 2 << (wndw->id * 8) : 0;
-	return interlock ? 2 << (wndw->id * 4) : 0;
-}
-
-static int
-nv50_base_ntfy_wait_begun(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
-{
-	struct nouveau_drm *drm = nouveau_drm(wndw->plane.dev);
-	struct nv50_disp *disp = nv50_disp(wndw->plane.dev);
-	if (nvif_msec(&drm->client.device, 2000ULL,
-		u32 data = nouveau_bo_rd32(disp->sync, asyw->ntfy.offset / 4);
-		if ((data & 0xc0000000) == 0x40000000)
-			break;
-		usleep_range(1, 2);
-	) < 0)
-		return -ETIMEDOUT;
-	return 0;
-}
-
-static void
-nv50_base_release(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
-		  struct nv50_head_atom *asyh)
-{
-	asyh->base.cpp = 0;
-}
-
-static int
-nv50_base_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
-		  struct nv50_head_atom *asyh)
-{
-	const struct drm_framebuffer *fb = asyw->state.fb;
-	int ret;
-
-	if (!fb->format->depth)
-		return -EINVAL;
-
-	ret = drm_atomic_helper_check_plane_state(&asyw->state, &asyh->state,
-						  DRM_PLANE_HELPER_NO_SCALING,
-						  DRM_PLANE_HELPER_NO_SCALING,
-						  false, true);
-	if (ret)
-		return ret;
-
-	asyh->base.depth = fb->format->depth;
-	asyh->base.cpp = fb->format->cpp[0];
-	asyh->base.x = asyw->state.src.x1 >> 16;
-	asyh->base.y = asyw->state.src.y1 >> 16;
-	asyh->base.w = asyw->state.fb->width;
-	asyh->base.h = asyw->state.fb->height;
-
-	switch (fb->format->format) {
-	case DRM_FORMAT_C8         : asyw->image.format = 0x1e; break;
-	case DRM_FORMAT_RGB565     : asyw->image.format = 0xe8; break;
-	case DRM_FORMAT_XRGB1555   :
-	case DRM_FORMAT_ARGB1555   : asyw->image.format = 0xe9; break;
-	case DRM_FORMAT_XRGB8888   :
-	case DRM_FORMAT_ARGB8888   : asyw->image.format = 0xcf; break;
-	case DRM_FORMAT_XBGR2101010:
-	case DRM_FORMAT_ABGR2101010: asyw->image.format = 0xd1; break;
-	case DRM_FORMAT_XBGR8888   :
-	case DRM_FORMAT_ABGR8888   : asyw->image.format = 0xd5; break;
-	default:
-		WARN_ON(1);
-		return -EINVAL;
-	}
-
-	asyw->lut.enable = 1;
-	asyw->set.image = true;
-	return 0;
-}
-
-static const u32
-nv50_base_format[] = {
-	DRM_FORMAT_C8,
-	DRM_FORMAT_RGB565,
-	DRM_FORMAT_XRGB1555,
-	DRM_FORMAT_ARGB1555,
-	DRM_FORMAT_XRGB8888,
-	DRM_FORMAT_ARGB8888,
-	DRM_FORMAT_XBGR2101010,
-	DRM_FORMAT_ABGR2101010,
-	DRM_FORMAT_XBGR8888,
-	DRM_FORMAT_ABGR8888,
-	0
-};
-
-static const struct nv50_wndw_func
-nv50_base = {
-	.acquire = nv50_base_acquire,
-	.release = nv50_base_release,
-	.sema_set = nv50_base_sema_set,
-	.sema_clr = nv50_base_sema_clr,
-	.ntfy_set = nv50_base_ntfy_set,
-	.ntfy_clr = nv50_base_ntfy_clr,
-	.ntfy_wait_begun = nv50_base_ntfy_wait_begun,
-	.image_set = nv50_base_image_set,
-	.image_clr = nv50_base_image_clr,
-	.lut = nv50_base_lut,
-	.update = nv50_base_update,
-};
-
-static int
-base507c_new_(const struct nv50_wndw_func *func, const u32 *format,
-	      struct nouveau_drm *drm, int head, s32 oclass,
-	      struct nv50_wndw **pwndw)
-{
-	struct nv50_disp_base_channel_dma_v0 args = {
-		.head = head,
-	};
-	struct nv50_disp *disp = nv50_disp(drm->dev);
-	struct nv50_wndw *wndw;
-	int ret;
-
-	ret = nv50_wndw_new_(func, drm->dev, DRM_PLANE_TYPE_PRIMARY,
-			     "base", head, format, &wndw);
-	if (*pwndw = wndw, ret)
-		return ret;
-
-	ret = nv50_dmac_create(&drm->client.device, &disp->disp->object,
-			       &oclass, head, &args, sizeof(args),
-			       disp->sync->bo.offset, &wndw->wndw);
-	if (ret) {
-		NV_ERROR(drm, "base%04x allocation failed: %d\n", oclass, ret);
-		return ret;
-	}
-
-	ret = nvif_notify_init(&wndw->wndw.base.user, wndw->notify.func,
-			       false, NV50_DISP_BASE_CHANNEL_DMA_V0_NTFY_UEVENT,
-			       &(struct nvif_notify_uevent_req) {},
-			       sizeof(struct nvif_notify_uevent_req),
-			       sizeof(struct nvif_notify_uevent_rep),
-			       &wndw->notify);
-	if (ret)
-		return ret;
-
-	wndw->ntfy = EVO_FLIP_NTFY0(wndw->id);
-	wndw->sema = EVO_FLIP_SEM0(wndw->id);
-	wndw->data = 0x00000000;
-	return 0;
-}
-
-static int
-base507c_new(struct nouveau_drm *drm, int head, s32 oclass,
-	     struct nv50_wndw **pwndw)
-{
-	return base507c_new_(&nv50_base, nv50_base_format, drm, head, oclass, pwndw);
-}
-
-static int
-nv50_base_new(struct nouveau_drm *drm, int head, struct nv50_wndw **pwndw)
-{
-	struct {
-		s32 oclass;
-		int version;
-		int (*new)(struct nouveau_drm *, int, s32, struct nv50_wndw **);
-	} bases[] = {
-		{ GK110_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
-		{ GK104_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
-		{ GF110_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
-		{ GT214_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
-		{ GT200_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
-		{   G82_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
-		{  NV50_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
-		{}
-	};
-	struct nv50_disp *disp = nv50_disp(drm->dev);
-	int cid;
-
-	cid = nvif_mclass(&disp->disp->object, bases);
-	if (cid < 0) {
-		NV_ERROR(drm, "No supported base class\n");
-		return cid;
-	}
-
-	return bases[cid].new(drm, head, bases[cid].oclass, pwndw);
-}
-
-/******************************************************************************
- * Head
- *****************************************************************************/
-static void
-head907d_or(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
-	u32 *push;
-	if (core->base.user.oclass >= GF110_DISP_CORE_CHANNEL_DMA &&
-	    (push = evo_wait(core, 3))) {
-		evo_mthd(push, 0x0404 + (head->base.index * 0x300), 2);
-		evo_data(push, 0x00000001 | (asyh->or.depth  << 6) |
-					    (asyh->or.nvsync << 4) |
-					    (asyh->or.nhsync << 3));
-		evo_data(push, 0x31ec6000 | (head->base.index << 25) |
-					     asyh->mode.interlace);
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_procamp(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
-	u32 *push;
-	if ((push = evo_wait(core, 2))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
-			evo_mthd(push, 0x08a8 + (head->base.index * 0x400), 1);
-		else
-			evo_mthd(push, 0x0498 + (head->base.index * 0x300), 1);
-		evo_data(push, (asyh->procamp.sat.sin << 20) |
-			       (asyh->procamp.sat.cos << 8));
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_dither(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
-	u32 *push;
-	if ((push = evo_wait(core, 2))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
-			evo_mthd(push, 0x08a0 + (head->base.index * 0x0400), 1);
-		else
-		if (core->base.user.oclass < GK104_DISP_CORE_CHANNEL_DMA)
-			evo_mthd(push, 0x0490 + (head->base.index * 0x0300), 1);
-		else
-			evo_mthd(push, 0x04a0 + (head->base.index * 0x0300), 1);
-		evo_data(push, (asyh->dither.mode << 3) |
-			       (asyh->dither.bits << 1) |
-			        asyh->dither.enable);
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_ovly(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
-	u32 bounds = 0;
-	u32 *push;
-
-	if (asyh->base.cpp) {
-		switch (asyh->base.cpp) {
-		case 8: bounds |= 0x00000500; break;
-		case 4: bounds |= 0x00000300; break;
-		case 2: bounds |= 0x00000100; break;
-		default:
-			WARN_ON(1);
-			break;
-		}
-		bounds |= 0x00000001;
-	}
-
-	if ((push = evo_wait(core, 2))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
-			evo_mthd(push, 0x0904 + head->base.index * 0x400, 1);
-		else
-			evo_mthd(push, 0x04d4 + head->base.index * 0x300, 1);
-		evo_data(push, bounds);
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_base(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
-	u32 bounds = 0;
-	u32 *push;
-
-	if (asyh->base.cpp) {
-		switch (asyh->base.cpp) {
-		case 8: bounds |= 0x00000500; break;
-		case 4: bounds |= 0x00000300; break;
-		case 2: bounds |= 0x00000100; break;
-		case 1: bounds |= 0x00000000; break;
-		default:
-			WARN_ON(1);
-			break;
-		}
-		bounds |= 0x00000001;
-	}
-
-	if ((push = evo_wait(core, 2))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
-			evo_mthd(push, 0x0900 + head->base.index * 0x400, 1);
-		else
-			evo_mthd(push, 0x04d0 + head->base.index * 0x300, 1);
-		evo_data(push, bounds);
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_curs_clr(struct nv50_head *head)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
-	u32 *push;
-	if ((push = evo_wait(core, 4))) {
-		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0880 + head->base.index * 0x400, 1);
-			evo_data(push, 0x05000000);
-		} else
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0880 + head->base.index * 0x400, 1);
-			evo_data(push, 0x05000000);
-			evo_mthd(push, 0x089c + head->base.index * 0x400, 1);
-			evo_data(push, 0x00000000);
-		} else {
-			evo_mthd(push, 0x0480 + head->base.index * 0x300, 1);
-			evo_data(push, 0x05000000);
-			evo_mthd(push, 0x048c + head->base.index * 0x300, 1);
-			evo_data(push, 0x00000000);
-		}
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_curs_set(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
-	u32 *push;
-	if ((push = evo_wait(core, 5))) {
-		if (core->base.user.oclass < G82_DISP_BASE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0880 + head->base.index * 0x400, 2);
-			evo_data(push, 0x80000000 | (asyh->curs.layout << 26) |
-						    (asyh->curs.format << 24));
-			evo_data(push, asyh->curs.offset >> 8);
-		} else
-		if (core->base.user.oclass < GF110_DISP_BASE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0880 + head->base.index * 0x400, 2);
-			evo_data(push, 0x80000000 | (asyh->curs.layout << 26) |
-						    (asyh->curs.format << 24));
-			evo_data(push, asyh->curs.offset >> 8);
-			evo_mthd(push, 0x089c + head->base.index * 0x400, 1);
-			evo_data(push, asyh->curs.handle);
-		} else {
-			evo_mthd(push, 0x0480 + head->base.index * 0x300, 2);
-			evo_data(push, 0x80000000 | (asyh->curs.layout << 26) |
-						    (asyh->curs.format << 24));
-			evo_data(push, asyh->curs.offset >> 8);
-			evo_mthd(push, 0x048c + head->base.index * 0x300, 1);
-			evo_data(push, asyh->curs.handle);
-		}
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_core_clr(struct nv50_head *head)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
-	u32 *push;
-	if ((push = evo_wait(core, 2))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
-			evo_mthd(push, 0x0874 + head->base.index * 0x400, 1);
-		else
-			evo_mthd(push, 0x0474 + head->base.index * 0x300, 1);
-		evo_data(push, 0x00000000);
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_core_set(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
-	u32 *push;
-	if ((push = evo_wait(core, 9))) {
-		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0860 + head->base.index * 0x400, 1);
-			evo_data(push, asyh->core.offset >> 8);
-			evo_mthd(push, 0x0868 + head->base.index * 0x400, 4);
-			evo_data(push, (asyh->core.h << 16) | asyh->core.w);
-			evo_data(push, asyh->core.layout << 20 |
-				       (asyh->core.pitch >> 8) << 8 |
-				       asyh->core.block);
-			evo_data(push, asyh->core.kind << 16 |
-				       asyh->core.format << 8);
-			evo_data(push, asyh->core.handle);
-			evo_mthd(push, 0x08c0 + head->base.index * 0x400, 1);
-			evo_data(push, (asyh->core.y << 16) | asyh->core.x);
-			/* EVO will complain with INVALID_STATE if we have an
-			 * active cursor and (re)specify HeadSetContextDmaIso
-			 * without also updating HeadSetOffsetCursor.
-			 */
-			asyh->set.curs = asyh->curs.visible;
-		} else
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0860 + head->base.index * 0x400, 1);
-			evo_data(push, asyh->core.offset >> 8);
-			evo_mthd(push, 0x0868 + head->base.index * 0x400, 4);
-			evo_data(push, (asyh->core.h << 16) | asyh->core.w);
-			evo_data(push, asyh->core.layout << 20 |
-				       (asyh->core.pitch >> 8) << 8 |
-				       asyh->core.block);
-			evo_data(push, asyh->core.format << 8);
-			evo_data(push, asyh->core.handle);
-			evo_mthd(push, 0x08c0 + head->base.index * 0x400, 1);
-			evo_data(push, (asyh->core.y << 16) | asyh->core.x);
-		} else {
-			evo_mthd(push, 0x0460 + head->base.index * 0x300, 1);
-			evo_data(push, asyh->core.offset >> 8);
-			evo_mthd(push, 0x0468 + head->base.index * 0x300, 4);
-			evo_data(push, (asyh->core.h << 16) | asyh->core.w);
-			evo_data(push, asyh->core.layout << 24 |
-				       (asyh->core.pitch >> 8) << 8 |
-				       asyh->core.block);
-			evo_data(push, asyh->core.format << 8);
-			evo_data(push, asyh->core.handle);
-			evo_mthd(push, 0x04b0 + head->base.index * 0x300, 1);
-			evo_data(push, (asyh->core.y << 16) | asyh->core.x);
-		}
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_lut_clr(struct nv50_head *head)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
-	u32 *push;
-	if ((push = evo_wait(core, 4))) {
-		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0840 + (head->base.index * 0x400), 1);
-			evo_data(push, 0x40000000);
-		} else
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0840 + (head->base.index * 0x400), 1);
-			evo_data(push, 0x40000000);
-			evo_mthd(push, 0x085c + (head->base.index * 0x400), 1);
-			evo_data(push, 0x00000000);
-		} else {
-			evo_mthd(push, 0x0440 + (head->base.index * 0x300), 1);
-			evo_data(push, 0x03000000);
-			evo_mthd(push, 0x045c + (head->base.index * 0x300), 1);
-			evo_data(push, 0x00000000);
-		}
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_lut_load(struct drm_property_blob *blob, int mode,
-		   struct nouveau_bo *nvbo)
-{
-	struct drm_color_lut *in = (struct drm_color_lut *)blob->data;
-	void __iomem *lut = (u8 *)nvbo_kmap_obj_iovirtual(nvbo);
-	const int size = blob->length / sizeof(*in);
-	int bits, shift, i;
-	u16 zero, r, g, b;
-
-	/* This can't happen.. But it shuts the compiler up. */
-	if (WARN_ON(size != 256))
-		return;
-
-	switch (mode) {
-	case 0: /* LORES. */
-	case 1: /* HIRES. */
-		bits = 11;
-		shift = 3;
-		zero = 0x0000;
-		break;
-	case 7: /* INTERPOLATE_257_UNITY_RANGE. */
-		bits = 14;
-		shift = 0;
-		zero = 0x6000;
-		break;
-	default:
-		WARN_ON(1);
-		return;
-	}
-
-	for (i = 0; i < size; i++) {
-		r = (drm_color_lut_extract(in[i].  red, bits) + zero) << shift;
-		g = (drm_color_lut_extract(in[i].green, bits) + zero) << shift;
-		b = (drm_color_lut_extract(in[i]. blue, bits) + zero) << shift;
-		writew(r, lut + (i * 0x08) + 0);
-		writew(g, lut + (i * 0x08) + 2);
-		writew(b, lut + (i * 0x08) + 4);
-	}
-
-	/* INTERPOLATE modes require a "next" entry to interpolate with,
-	 * so we replicate the last entry to deal with this for now.
-	 */
-	writew(r, lut + (i * 0x08) + 0);
-	writew(g, lut + (i * 0x08) + 2);
-	writew(b, lut + (i * 0x08) + 4);
-}
-
-static void
-nv50_head_lut_set(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
-	u32 *push;
-	if ((push = evo_wait(core, 7))) {
-		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0840 + (head->base.index * 0x400), 2);
-			evo_data(push, 0x80000000 | asyh->lut.mode << 30);
-			evo_data(push, asyh->lut.offset >> 8);
-		} else
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0840 + (head->base.index * 0x400), 2);
-			evo_data(push, 0x80000000 | asyh->lut.mode << 30);
-			evo_data(push, asyh->lut.offset >> 8);
-			evo_mthd(push, 0x085c + (head->base.index * 0x400), 1);
-			evo_data(push, asyh->lut.handle);
-		} else {
-			evo_mthd(push, 0x0440 + (head->base.index * 0x300), 4);
-			evo_data(push, 0x80000000 | asyh->lut.mode << 24);
-			evo_data(push, asyh->lut.offset >> 8);
-			evo_data(push, 0x00000000);
-			evo_data(push, 0x00000000);
-			evo_mthd(push, 0x045c + (head->base.index * 0x300), 1);
-			evo_data(push, asyh->lut.handle);
-		}
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_mode(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
-	struct nv50_head_mode *m = &asyh->mode;
-	u32 *push;
-	if ((push = evo_wait(core, 14))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0804 + (head->base.index * 0x400), 2);
-			evo_data(push, 0x00800000 | m->clock);
-			evo_data(push, m->interlace ? 0x00000002 : 0x00000000);
-			evo_mthd(push, 0x0810 + (head->base.index * 0x400), 7);
-			evo_data(push, 0x00000000);
-			evo_data(push, (m->v.active  << 16) | m->h.active );
-			evo_data(push, (m->v.synce   << 16) | m->h.synce  );
-			evo_data(push, (m->v.blanke  << 16) | m->h.blanke );
-			evo_data(push, (m->v.blanks  << 16) | m->h.blanks );
-			evo_data(push, (m->v.blank2e << 16) | m->v.blank2s);
-			evo_data(push, asyh->mode.v.blankus);
-			evo_mthd(push, 0x082c + (head->base.index * 0x400), 1);
-			evo_data(push, 0x00000000);
-		} else {
-			evo_mthd(push, 0x0410 + (head->base.index * 0x300), 6);
-			evo_data(push, 0x00000000);
-			evo_data(push, (m->v.active  << 16) | m->h.active );
-			evo_data(push, (m->v.synce   << 16) | m->h.synce  );
-			evo_data(push, (m->v.blanke  << 16) | m->h.blanke );
-			evo_data(push, (m->v.blanks  << 16) | m->h.blanks );
-			evo_data(push, (m->v.blank2e << 16) | m->v.blank2s);
-			evo_mthd(push, 0x042c + (head->base.index * 0x300), 2);
-			evo_data(push, 0x00000000); /* ??? */
-			evo_data(push, 0xffffff00);
-			evo_mthd(push, 0x0450 + (head->base.index * 0x300), 3);
-			evo_data(push, m->clock * 1000);
-			evo_data(push, 0x00200000); /* ??? */
-			evo_data(push, m->clock * 1000);
-		}
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_view(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
-	u32 *push;
-	if ((push = evo_wait(core, 10))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x08a4 + (head->base.index * 0x400), 1);
-			evo_data(push, 0x00000000);
-			evo_mthd(push, 0x08c8 + (head->base.index * 0x400), 1);
-			evo_data(push, (asyh->view.iH << 16) | asyh->view.iW);
-			evo_mthd(push, 0x08d8 + (head->base.index * 0x400), 2);
-			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
-			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
-		} else {
-			evo_mthd(push, 0x0494 + (head->base.index * 0x300), 1);
-			evo_data(push, 0x00000000);
-			evo_mthd(push, 0x04b8 + (head->base.index * 0x300), 1);
-			evo_data(push, (asyh->view.iH << 16) | asyh->view.iW);
-			evo_mthd(push, 0x04c0 + (head->base.index * 0x300), 3);
-			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
-			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
-			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
-		}
-		evo_kick(push, core);
-	}
-}
-
-static const struct nv50_head_func
-head507d = {
-	.view = nv50_head_view,
-	.mode = nv50_head_mode,
-	.ilut_set = nv50_head_lut_set,
-	.ilut_clr = nv50_head_lut_clr,
-	.core_set = nv50_head_core_set,
-	.core_clr = nv50_head_core_clr,
-	.curs_set = nv50_head_curs_set,
-	.curs_clr = nv50_head_curs_clr,
-	.base = nv50_head_base,
-	.ovly = nv50_head_ovly,
-	.dither = nv50_head_dither,
-	.procamp = nv50_head_procamp,
-	.or = head907d_or,
-};
-
-static void
-nv50_head_flush_clr(struct nv50_head *head, struct nv50_head_atom *asyh, bool y)
-{
-	if (asyh->clr.ilut && (!asyh->set.ilut || y))
-		head->func->ilut_clr(head);
-	if (asyh->clr.core && (!asyh->set.core || y))
-		head->func->core_clr(head);
-	if (asyh->clr.curs && (!asyh->set.curs || y))
-		head->func->curs_clr(head);
-}
-
-static void
-nv50_head_flush_set(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	if (asyh->set.view   ) head->func->view    (head, asyh);
-	if (asyh->set.mode   ) head->func->mode    (head, asyh);
-	if (asyh->set.ilut   ) {
-		struct nouveau_bo *nvbo = head->lut.nvbo[head->lut.next];
-		struct drm_property_blob *blob = asyh->state.gamma_lut;
-		if (blob)
-			nv50_head_lut_load(blob, asyh->lut.mode, nvbo);
-		asyh->lut.offset = nvbo->bo.offset;
-		head->lut.next ^= 1;
-		head->func->ilut_set(head, asyh);
-	}
-	if (asyh->set.core   ) head->func->core_set(head, asyh);
-	if (asyh->set.curs   ) head->func->curs_set(head, asyh);
-	if (asyh->set.base   ) head->func->base    (head, asyh);
-	if (asyh->set.ovly   ) head->func->ovly    (head, asyh);
-	if (asyh->set.dither ) head->func->dither  (head, asyh);
-	if (asyh->set.procamp) head->func->procamp (head, asyh);
-	if (asyh->set.or     ) head->func->or      (head, asyh);
-}
-
-static void
-nv50_head_atomic_check_procamp(struct nv50_head_atom *armh,
-			       struct nv50_head_atom *asyh,
-			       struct nouveau_conn_atom *asyc)
-{
-	const int vib = asyc->procamp.color_vibrance - 100;
-	const int hue = asyc->procamp.vibrant_hue - 90;
-	const int adj = (vib > 0) ? 50 : 0;
-	asyh->procamp.sat.cos = ((vib * 2047 + adj) / 100) & 0xfff;
-	asyh->procamp.sat.sin = ((hue * 2047) / 100) & 0xfff;
-	asyh->set.procamp = true;
-}
-
-static void
-nv50_head_atomic_check_dither(struct nv50_head_atom *armh,
-			      struct nv50_head_atom *asyh,
-			      struct nouveau_conn_atom *asyc)
-{
-	struct drm_connector *connector = asyc->state.connector;
-	u32 mode = 0x00;
-
-	if (asyc->dither.mode == DITHERING_MODE_AUTO) {
-		if (asyh->base.depth > connector->display_info.bpc * 3)
-			mode = DITHERING_MODE_DYNAMIC2X2;
-	} else {
-		mode = asyc->dither.mode;
-	}
-
-	if (asyc->dither.depth == DITHERING_DEPTH_AUTO) {
-		if (connector->display_info.bpc >= 8)
-			mode |= DITHERING_DEPTH_8BPC;
-	} else {
-		mode |= asyc->dither.depth;
-	}
-
-	asyh->dither.enable = mode;
-	asyh->dither.bits = mode >> 1;
-	asyh->dither.mode = mode >> 3;
-	asyh->set.dither = true;
-}
-
-static void
-nv50_head_atomic_check_view(struct nv50_head_atom *armh,
-			    struct nv50_head_atom *asyh,
-			    struct nouveau_conn_atom *asyc)
-{
-	struct drm_connector *connector = asyc->state.connector;
-	struct drm_display_mode *omode = &asyh->state.adjusted_mode;
-	struct drm_display_mode *umode = &asyh->state.mode;
-	int mode = asyc->scaler.mode;
-	struct edid *edid;
-	int umode_vdisplay, omode_hdisplay, omode_vdisplay;
-
-	if (connector->edid_blob_ptr)
-		edid = (struct edid *)connector->edid_blob_ptr->data;
-	else
-		edid = NULL;
-
-	if (!asyc->scaler.full) {
-		if (mode == DRM_MODE_SCALE_NONE)
-			omode = umode;
-	} else {
-		/* Non-EDID LVDS/eDP mode. */
-		mode = DRM_MODE_SCALE_FULLSCREEN;
-	}
-
-	/* For the user-specified mode, we must ignore doublescan and
-	 * the like, but honor frame packing.
-	 */
-	umode_vdisplay = umode->vdisplay;
-	if ((umode->flags & DRM_MODE_FLAG_3D_MASK) == DRM_MODE_FLAG_3D_FRAME_PACKING)
-		umode_vdisplay += umode->vtotal;
-	asyh->view.iW = umode->hdisplay;
-	asyh->view.iH = umode_vdisplay;
-	/* For the output mode, we can just use the stock helper. */
-	drm_mode_get_hv_timing(omode, &omode_hdisplay, &omode_vdisplay);
-	asyh->view.oW = omode_hdisplay;
-	asyh->view.oH = omode_vdisplay;
-
-	/* Add overscan compensation if necessary, will keep the aspect
-	 * ratio the same as the backend mode unless overridden by the
-	 * user setting both hborder and vborder properties.
-	 */
-	if ((asyc->scaler.underscan.mode == UNDERSCAN_ON ||
-	    (asyc->scaler.underscan.mode == UNDERSCAN_AUTO &&
-	     drm_detect_hdmi_monitor(edid)))) {
-		u32 bX = asyc->scaler.underscan.hborder;
-		u32 bY = asyc->scaler.underscan.vborder;
-		u32 r = (asyh->view.oH << 19) / asyh->view.oW;
-
-		if (bX) {
-			asyh->view.oW -= (bX * 2);
-			if (bY) asyh->view.oH -= (bY * 2);
-			else    asyh->view.oH  = ((asyh->view.oW * r) + (r / 2)) >> 19;
-		} else {
-			asyh->view.oW -= (asyh->view.oW >> 4) + 32;
-			if (bY) asyh->view.oH -= (bY * 2);
-			else    asyh->view.oH  = ((asyh->view.oW * r) + (r / 2)) >> 19;
-		}
-	}
-
-	/* Handle CENTER/ASPECT scaling, taking into account the areas
-	 * removed already for overscan compensation.
-	 */
-	switch (mode) {
-	case DRM_MODE_SCALE_CENTER:
-		asyh->view.oW = min((u16)umode->hdisplay, asyh->view.oW);
-		asyh->view.oH = min((u16)umode_vdisplay, asyh->view.oH);
-		/* fall-through */
-	case DRM_MODE_SCALE_ASPECT:
-		if (asyh->view.oH < asyh->view.oW) {
-			u32 r = (asyh->view.iW << 19) / asyh->view.iH;
-			asyh->view.oW = ((asyh->view.oH * r) + (r / 2)) >> 19;
-		} else {
-			u32 r = (asyh->view.iH << 19) / asyh->view.iW;
-			asyh->view.oH = ((asyh->view.oW * r) + (r / 2)) >> 19;
-		}
-		break;
-	default:
-		break;
-	}
-
-	asyh->set.view = true;
-}
-
-static void
-nv50_head_atomic_check_lut(struct nv50_head *head,
-			   struct nv50_head_atom *armh,
-			   struct nv50_head_atom *asyh)
-{
-	struct nv50_disp *disp = nv50_disp(head->base.base.dev);
+#include <nvif/class.h>
+#include <nvif/cl0002.h>
+#include <nvif/cl5070.h>
+#include <nvif/cl507d.h>
+#include <nvif/event.h>
 
-	/* An I8 surface without an input LUT makes no sense, and
-	 * EVO will throw an error if you try.
-	 *
-	 * Legacy clients actually cause this due to the order in
-	 * which they call ioctls, so we will enable the LUT with
-	 * whatever contents the buffer already contains to avoid
-	 * triggering the error check.
-	 */
-	if (!asyh->state.gamma_lut && asyh->base.cpp != 1) {
-		asyh->lut.handle = 0;
-		asyh->clr.ilut = armh->lut.visible;
-		return;
-	}
+#include "nouveau_drv.h"
+#include "nouveau_dma.h"
+#include "nouveau_gem.h"
+#include "nouveau_connector.h"
+#include "nouveau_encoder.h"
+#include "nouveau_fence.h"
+#include "nouveau_fbcon.h"
 
-	if (disp->disp->object.oclass < GF110_DISP) {
-		asyh->lut.mode = (asyh->base.cpp == 1) ? 0 : 1;
-		asyh->set.ilut = true;
-	} else {
-		asyh->lut.mode = 7;
-		asyh->set.ilut = asyh->state.color_mgmt_changed;
-	}
-	asyh->lut.handle = disp->core->chan.vram.handle;
-}
+/******************************************************************************
+ * Atomic state
+ *****************************************************************************/
 
-static void
-nv50_head_atomic_check_mode(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct drm_display_mode *mode = &asyh->state.adjusted_mode;
-	struct nv50_head_mode *m = &asyh->mode;
-	u32 blankus;
+struct nv50_outp_atom {
+	struct list_head head;
 
-	drm_mode_set_crtcinfo(mode, CRTC_INTERLACE_HALVE_V | CRTC_STEREO_DOUBLE);
+	struct drm_encoder *encoder;
+	bool flush_disable;
 
-	/*
-	 * DRM modes are defined in terms of a repeating interval
-	 * starting with the active display area.  The hardware modes
-	 * are defined in terms of a repeating interval starting one
-	 * unit (pixel or line) into the sync pulse.  So, add bias.
-	 */
+	union {
+		struct {
+			bool ctrl:1;
+		};
+		u8 mask;
+	} clr;
 
-	m->h.active = mode->crtc_htotal;
-	m->h.synce  = mode->crtc_hsync_end - mode->crtc_hsync_start - 1;
-	m->h.blanke = mode->crtc_hblank_end - mode->crtc_hsync_start - 1;
-	m->h.blanks = m->h.blanke + mode->crtc_hdisplay;
-
-	m->v.active = mode->crtc_vtotal;
-	m->v.synce  = mode->crtc_vsync_end - mode->crtc_vsync_start - 1;
-	m->v.blanke = mode->crtc_vblank_end - mode->crtc_vsync_start - 1;
-	m->v.blanks = m->v.blanke + mode->crtc_vdisplay;
-
-	/*XXX: Safe underestimate, even "0" works */
-	blankus = (m->v.active - mode->crtc_vdisplay - 2) * m->h.active;
-	blankus *= 1000;
-	blankus /= mode->crtc_clock;
-	m->v.blankus = blankus;
-
-	if (mode->flags & DRM_MODE_FLAG_INTERLACE) {
-		m->v.blank2e =  m->v.active + m->v.blanke;
-		m->v.blank2s =  m->v.blank2e + mode->crtc_vdisplay;
-		m->v.active  = (m->v.active * 2) + 1;
-		m->interlace = true;
-	} else {
-		m->v.blank2e = 0;
-		m->v.blank2s = 1;
-		m->interlace = false;
-	}
-	m->clock = mode->crtc_clock;
+	union {
+		struct {
+			bool ctrl:1;
+		};
+		u8 mask;
+	} set;
+};
 
-	asyh->or.nhsync = !!(mode->flags & DRM_MODE_FLAG_NHSYNC);
-	asyh->or.nvsync = !!(mode->flags & DRM_MODE_FLAG_NVSYNC);
-	asyh->set.or = head->func->or != NULL;
-	asyh->set.mode = true;
-}
+/******************************************************************************
+ * EVO channel
+ *****************************************************************************/
 
 static int
-nv50_head_atomic_check(struct drm_crtc *crtc, struct drm_crtc_state *state)
+nv50_chan_create(struct nvif_device *device, struct nvif_object *disp,
+		 const s32 *oclass, u8 head, void *data, u32 size,
+		 struct nv50_chan *chan)
 {
-	struct nouveau_drm *drm = nouveau_drm(crtc->dev);
-	struct nv50_disp *disp = nv50_disp(crtc->dev);
-	struct nv50_head *head = nv50_head(crtc);
-	struct nv50_head_atom *armh = nv50_head_atom(crtc->state);
-	struct nv50_head_atom *asyh = nv50_head_atom(state);
-	struct nouveau_conn_atom *asyc = NULL;
-	struct drm_connector_state *conns;
-	struct drm_connector *conn;
-	int i;
-
-	NV_ATOMIC(drm, "%s atomic_check %d\n", crtc->name, asyh->state.active);
-	if (asyh->state.active) {
-		for_each_new_connector_in_state(asyh->state.state, conn, conns, i) {
-			if (conns->crtc == crtc) {
-				asyc = nouveau_conn_atom(conns);
-				break;
-			}
-		}
-
-		if (armh->state.active) {
-			if (asyc) {
-				if (asyh->state.mode_changed)
-					asyc->set.scaler = true;
-				if (armh->base.depth != asyh->base.depth)
-					asyc->set.dither = true;
-			}
-		} else {
-			if (asyc)
-				asyc->set.mask = ~0;
-			asyh->set.mask = ~0;
-			asyh->set.or = head->func->or != NULL;
-		}
-
-		if (asyh->state.mode_changed)
-			nv50_head_atomic_check_mode(head, asyh);
-
-		if (asyh->state.color_mgmt_changed ||
-		    asyh->base.cpp != armh->base.cpp)
-			nv50_head_atomic_check_lut(head, armh, asyh);
-		asyh->lut.visible = asyh->lut.handle != 0;
-
-		if (asyc) {
-			if (asyc->set.scaler)
-				nv50_head_atomic_check_view(armh, asyh, asyc);
-			if (asyc->set.dither)
-				nv50_head_atomic_check_dither(armh, asyh, asyc);
-			if (asyc->set.procamp)
-				nv50_head_atomic_check_procamp(armh, asyh, asyc);
-		}
+	struct nvif_sclass *sclass;
+	int ret, i, n;
 
-		if ((asyh->core.visible = (asyh->base.cpp != 0))) {
-			asyh->core.x = asyh->base.x;
-			asyh->core.y = asyh->base.y;
-			asyh->core.w = asyh->base.w;
-			asyh->core.h = asyh->base.h;
-		} else
-		if ((asyh->core.visible = asyh->curs.visible) ||
-		    (asyh->core.visible = asyh->lut.visible)) {
-			/*XXX: We need to either find some way of having the
-			 *     primary base layer appear black, while still
-			 *     being able to display the other layers, or we
-			 *     need to allocate a dummy black surface here.
-			 */
-			asyh->core.x = 0;
-			asyh->core.y = 0;
-			asyh->core.w = asyh->state.mode.hdisplay;
-			asyh->core.h = asyh->state.mode.vdisplay;
-		}
-		asyh->core.handle = disp->core->chan.vram.handle;
-		asyh->core.offset = 0;
-		asyh->core.format = 0xcf;
-		asyh->core.kind = 0;
-		asyh->core.layout = 1;
-		asyh->core.block = 0;
-		asyh->core.pitch = ALIGN(asyh->core.w, 64) * 4;
-		asyh->set.base = armh->base.cpp != asyh->base.cpp;
-		asyh->set.ovly = armh->ovly.cpp != asyh->ovly.cpp;
-	} else {
-		asyh->lut.visible = false;
-		asyh->core.visible = false;
-		asyh->curs.visible = false;
-		asyh->base.cpp = 0;
-		asyh->ovly.cpp = 0;
-	}
+	chan->device = device;
 
-	if (!drm_atomic_crtc_needs_modeset(&asyh->state)) {
-		if (asyh->core.visible) {
-			if (memcmp(&armh->core, &asyh->core, sizeof(asyh->core)))
-				asyh->set.core = true;
-		} else
-		if (armh->core.visible) {
-			asyh->clr.core = true;
-		}
+	ret = n = nvif_object_sclass_get(disp, &sclass);
+	if (ret < 0)
+		return ret;
 
-		if (asyh->curs.visible) {
-			if (memcmp(&armh->curs, &asyh->curs, sizeof(asyh->curs)))
-				asyh->set.curs = true;
-		} else
-		if (armh->curs.visible) {
-			asyh->clr.curs = true;
+	while (oclass[0]) {
+		for (i = 0; i < n; i++) {
+			if (sclass[i].oclass == oclass[0]) {
+				ret = nvif_object_init(disp, 0, oclass[0],
+						       data, size, &chan->user);
+				if (ret == 0)
+					nvif_object_map(&chan->user, NULL, 0);
+				nvif_object_sclass_put(&sclass);
+				return ret;
+			}
 		}
-	} else {
-		asyh->clr.ilut = armh->lut.visible;
-		asyh->clr.core = armh->core.visible;
-		asyh->clr.curs = armh->curs.visible;
-		asyh->set.ilut = asyh->lut.visible;
-		asyh->set.core = asyh->core.visible;
-		asyh->set.curs = asyh->curs.visible;
+		oclass++;
 	}
 
-	if (asyh->clr.mask || asyh->set.mask)
-		nv50_atom(asyh->state.state)->lock_core = true;
-	return 0;
+	nvif_object_sclass_put(&sclass);
+	return -ENOSYS;
 }
 
-static const struct drm_crtc_helper_funcs
-nv50_head_help = {
-	.atomic_check = nv50_head_atomic_check,
-};
-
 static void
-nv50_head_atomic_destroy_state(struct drm_crtc *crtc,
-			       struct drm_crtc_state *state)
-{
-	struct nv50_head_atom *asyh = nv50_head_atom(state);
-	__drm_atomic_helper_crtc_destroy_state(&asyh->state);
-	kfree(asyh);
-}
-
-static struct drm_crtc_state *
-nv50_head_atomic_duplicate_state(struct drm_crtc *crtc)
+nv50_chan_destroy(struct nv50_chan *chan)
 {
-	struct nv50_head_atom *armh = nv50_head_atom(crtc->state);
-	struct nv50_head_atom *asyh;
-	if (!(asyh = kmalloc(sizeof(*asyh), GFP_KERNEL)))
-		return NULL;
-	__drm_atomic_helper_crtc_duplicate_state(crtc, &asyh->state);
-	asyh->view = armh->view;
-	asyh->mode = armh->mode;
-	asyh->lut  = armh->lut;
-	asyh->core = armh->core;
-	asyh->curs = armh->curs;
-	asyh->base = armh->base;
-	asyh->ovly = armh->ovly;
-	asyh->dither = armh->dither;
-	asyh->procamp = armh->procamp;
-	asyh->clr.mask = 0;
-	asyh->set.mask = 0;
-	return &asyh->state;
+	nvif_object_fini(&chan->user);
 }
 
-static void
-__drm_atomic_helper_crtc_reset(struct drm_crtc *crtc,
-			       struct drm_crtc_state *state)
-{
-	if (crtc->state)
-		crtc->funcs->atomic_destroy_state(crtc, crtc->state);
-	crtc->state = state;
-	crtc->state->crtc = crtc;
-}
+/******************************************************************************
+ * DMA EVO channel
+ *****************************************************************************/
 
-static void
-nv50_head_reset(struct drm_crtc *crtc)
+void
+nv50_dmac_destroy(struct nv50_dmac *dmac)
 {
-	struct nv50_head_atom *asyh;
+	nvif_object_fini(&dmac->vram);
+	nvif_object_fini(&dmac->sync);
 
-	if (WARN_ON(!(asyh = kzalloc(sizeof(*asyh), GFP_KERNEL))))
-		return;
+	nv50_chan_destroy(&dmac->base);
 
-	__drm_atomic_helper_crtc_reset(crtc, &asyh->state);
+	nvif_mem_fini(&dmac->push);
 }
 
-static void
-nv50_head_destroy(struct drm_crtc *crtc)
+int
+nv50_dmac_create(struct nvif_device *device, struct nvif_object *disp,
+		 const s32 *oclass, u8 head, void *data, u32 size, u64 syncbuf,
+		 struct nv50_dmac *dmac)
 {
-	struct nv50_head *head = nv50_head(crtc);
-	int i;
-
-	for (i = 0; i < ARRAY_SIZE(head->lut.nvbo); i++)
-		nouveau_bo_unmap_unpin_unref(&head->lut.nvbo[i]);
+	struct nouveau_cli *cli = (void *)device->object.client;
+	struct nv50_disp_core_channel_dma_v0 *args = data;
+	int ret;
 
-	drm_crtc_cleanup(crtc);
-	kfree(head);
-}
+	mutex_init(&dmac->lock);
 
-static const struct drm_crtc_funcs
-nv50_head_func = {
-	.reset = nv50_head_reset,
-	.gamma_set = drm_atomic_helper_legacy_gamma_set,
-	.destroy = nv50_head_destroy,
-	.set_config = drm_atomic_helper_set_config,
-	.page_flip = drm_atomic_helper_page_flip,
-	.atomic_duplicate_state = nv50_head_atomic_duplicate_state,
-	.atomic_destroy_state = nv50_head_atomic_destroy_state,
-};
+	ret = nvif_mem_init_map(&cli->mmu, NVIF_MEM_COHERENT, 0x1000,
+				&dmac->push);
+	if (ret)
+		return ret;
 
-static int
-nv50_head_create(struct drm_device *dev, int index)
-{
-	struct nouveau_drm *drm = nouveau_drm(dev);
-	struct nv50_disp *disp = nv50_disp(dev);
-	struct nv50_head *head;
-	struct nv50_wndw *curs, *wndw;
-	struct drm_crtc *crtc;
-	int ret, i;
+	dmac->ptr = dmac->push.object.map.ptr;
 
-	head = kzalloc(sizeof(*head), GFP_KERNEL);
-	if (!head)
-		return -ENOMEM;
+	args->pushbuf = nvif_handle(&dmac->push.object);
 
-	head->func = disp->core->func->head;
-	head->base.index = index;
-	ret = nv50_base_new(drm, head->base.index, &wndw);
-	if (ret == 0)
-		ret = nv50_curs_new(drm, head->base.index, &curs);
-	if (ret) {
-		kfree(head);
+	ret = nv50_chan_create(device, disp, oclass, head, data, size,
+			       &dmac->base);
+	if (ret)
 		return ret;
-	}
 
-	crtc = &head->base.base;
-	drm_crtc_init_with_planes(dev, crtc, &wndw->plane, &curs->plane,
-				  &nv50_head_func, "head-%d", head->base.index);
-	drm_crtc_helper_add(crtc, &nv50_head_help);
-	drm_mode_crtc_set_gamma_size(crtc, 256);
-
-	for (i = 0; i < ARRAY_SIZE(head->lut.nvbo); i++) {
-		ret = nouveau_bo_new_pin_map(&drm->client, 1025 * 8, 0x100,
-					     TTM_PL_FLAG_VRAM,
-					     &head->lut.nvbo[i]);
-		if (ret)
-			goto out;
-	}
+	ret = nvif_object_init(&dmac->base.user, 0xf0000000, NV_DMA_IN_MEMORY,
+			       &(struct nv_dma_v0) {
+					.target = NV_DMA_V0_TARGET_VRAM,
+					.access = NV_DMA_V0_ACCESS_RDWR,
+					.start = syncbuf + 0x0000,
+					.limit = syncbuf + 0x0fff,
+			       }, sizeof(struct nv_dma_v0),
+			       &dmac->sync);
+	if (ret)
+		return ret;
 
-	/* allocate overlay resources */
-	ret = nv50_ovly_new(drm, head->base.index, &wndw);
-out:
+	ret = nvif_object_init(&dmac->base.user, 0xf0000001, NV_DMA_IN_MEMORY,
+			       &(struct nv_dma_v0) {
+					.target = NV_DMA_V0_TARGET_VRAM,
+					.access = NV_DMA_V0_ACCESS_RDWR,
+					.start = 0,
+					.limit = device->info.ram_user - 1,
+			       }, sizeof(struct nv_dma_v0),
+			       &dmac->vram);
 	if (ret)
-		nv50_head_destroy(crtc);
+		return ret;
+
 	return ret;
 }
 
-static const struct nv50_outp_func dac507d;
-static const struct nv50_outp_func sor507d;
-static const struct nv50_outp_func pior507d;
-static const struct nv50_core_func
-core507d = {
-	.head = &head507d,
-	.dac = &dac507d,
-	.sor = &sor507d,
-	.pior = &pior507d,
-};
-
-static int
-core507d_new(struct nouveau_drm *drm, s32 oclass, struct nv50_core **pcore)
+/******************************************************************************
+ * EVO channel helpers
+ *****************************************************************************/
+u32 *
+evo_wait(struct nv50_dmac *evoc, int nr)
 {
-	return core507d_new_(&core507d, drm, oclass, pcore);
-}
+	struct nv50_dmac *dmac = evoc;
+	struct nvif_device *device = dmac->base.device;
+	u32 put = nvif_rd32(&dmac->base.user, 0x0000) / 4;
 
-static void
-nv50_core_del(struct nv50_core **pcore)
-{
-	struct nv50_core *core = *pcore;
-	if (core) {
-		nv50_dmac_destroy(&core->chan);
-		kfree(*pcore);
-		*pcore = NULL;
-	}
-}
+	mutex_lock(&dmac->lock);
+	if (put + nr >= (PAGE_SIZE / 4) - 8) {
+		dmac->ptr[put] = 0x20000000;
 
-static int
-nv50_core_new(struct nouveau_drm *drm, struct nv50_core **pcore)
-{
-	struct {
-		s32 oclass;
-		int version;
-		int (*new)(struct nouveau_drm *, s32, struct nv50_core **);
-	} cores[] = {
-		{ GP102_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
-		{ GP100_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
-		{ GM200_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
-		{ GM107_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
-		{ GK110_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
-		{ GK104_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
-		{ GF110_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
-		{ GT214_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
-		{ GT206_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
-		{ GT200_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
-		{   G82_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
-		{  NV50_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
-		{}
-	};
-	struct nv50_disp *disp = nv50_disp(drm->dev);
-	int cid;
+		nvif_wr32(&dmac->base.user, 0x0000, 0x00000000);
+		if (nvif_msec(device, 2000,
+			if (!nvif_rd32(&dmac->base.user, 0x0004))
+				break;
+		) < 0) {
+			mutex_unlock(&dmac->lock);
+			pr_err("nouveau: evo channel stalled\n");
+			return NULL;
+		}
 
-	cid = nvif_mclass(&disp->disp->object, cores);
-	if (cid < 0) {
-		NV_ERROR(drm, "No supported core channel class\n");
-		return cid;
+		put = 0;
 	}
 
-	return cores[cid].new(drm, cores[cid].oclass, pcore);
+	return dmac->ptr + put;
+}
+
+void
+evo_kick(u32 *push, struct nv50_dmac *evoc)
+{
+	struct nv50_dmac *dmac = evoc;
+	nvif_wr32(&dmac->base.user, 0x0000, (push - dmac->ptr) << 2);
+	mutex_unlock(&dmac->lock);
 }
 
 /******************************************************************************
@@ -2721,33 +327,6 @@ nv50_outp_atomic_check(struct drm_encoder *encoder,
  * DAC
  *****************************************************************************/
 static void
-dac507d_ctrl(struct nv50_core *core, int or, u32 ctrl,
-	     struct nv50_head_atom *asyh)
-{
-	u32 *push, sync = 0;
-	if ((push = evo_wait(&core->chan, 3))) {
-		if (core->chan.base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			if (asyh) {
-				sync |= asyh->or.nvsync << 1;
-				sync |= asyh->or.nhsync;
-			}
-			evo_mthd(push, 0x0400 + (or * 0x080), 2);
-			evo_data(push, ctrl);
-			evo_data(push, sync);
-		} else {
-			evo_mthd(push, 0x0180 + (or * 0x020), 1);
-			evo_data(push, ctrl);
-		}
-		evo_kick(push, &core->chan);
-	}
-}
-
-static const struct nv50_outp_func
-dac507d = {
-	.ctrl = dac507d_ctrl,
-};
-
-static void
 nv50_dac_disable(struct drm_encoder *encoder)
 {
 	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
@@ -3635,32 +1214,6 @@ nv50_mstm_new(struct nouveau_encoder *outp, struct drm_dp_aux *aux, int aux_max,
  * SOR
  *****************************************************************************/
 static void
-sor507d_ctrl(struct nv50_core *core, int or, u32 ctrl,
-	     struct nv50_head_atom *asyh)
-{
-	u32 *push;
-	if ((push = evo_wait(&core->chan, 6))) {
-		if (core->chan.base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			if (asyh) {
-				ctrl |= asyh->or.depth  << 16;
-				ctrl |= asyh->or.nvsync << 13;
-				ctrl |= asyh->or.nhsync << 12;
-			}
-			evo_mthd(push, 0x0600 + (or * 0x40), 1);
-		} else {
-			evo_mthd(push, 0x0200 + (or * 0x20), 1);
-		}
-		evo_data(push, ctrl);
-		evo_kick(push, &core->chan);
-	}
-}
-
-static const struct nv50_outp_func
-sor507d = {
-	.ctrl = sor507d_ctrl,
-};
-
-static void
 nv50_sor_update(struct nouveau_encoder *nv_encoder, u8 head,
 		struct nv50_head_atom *asyh, u8 proto, u8 depth)
 {
@@ -3904,30 +1457,6 @@ nv50_sor_create(struct drm_connector *connector, struct dcb_output *dcbe)
 /******************************************************************************
  * PIOR
  *****************************************************************************/
-static void
-pior507d_ctrl(struct nv50_core *core, int or, u32 ctrl,
-	      struct nv50_head_atom *asyh)
-{
-	u32 *push;
-	if ((push = evo_wait(&core->chan, 8))) {
-		if (core->chan.base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			if (asyh) {
-				ctrl |= asyh->or.depth  << 16;
-				ctrl |= asyh->or.nvsync << 13;
-				ctrl |= asyh->or.nhsync << 12;
-			}
-			evo_mthd(push, 0x0700 + (or * 0x040), 1);
-			evo_data(push, ctrl);
-		}
-		evo_kick(push, &core->chan);
-	}
-}
-
-static const struct nv50_outp_func
-pior507d = {
-	.ctrl = pior507d_ctrl,
-};
-
 static int
 nv50_pior_atomic_check(struct drm_encoder *encoder,
 		       struct drm_crtc_state *crtc_state,
diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.h b/drivers/gpu/drm/nouveau/dispnv50/disp.h
new file mode 100644
index 0000000..7cbd668
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.h
@@ -0,0 +1,71 @@
+#ifndef __NV50_KMS_H__
+#define __NV50_KMS_H__
+#include <nvif/mem.h>
+
+#include "nouveau_display.h"
+
+struct nv50_disp {
+	struct nvif_disp *disp;
+	struct nv50_core *core;
+
+#define NV50_DISP_SYNC(c, o)                                ((c) * 0x040 + (o))
+#define NV50_DISP_CORE_NTFY                       NV50_DISP_SYNC(0      , 0x00)
+#define NV50_DISP_WNDW_SEM0(c)                    NV50_DISP_SYNC(1 + (c), 0x00)
+#define NV50_DISP_WNDW_SEM1(c)                    NV50_DISP_SYNC(1 + (c), 0x10)
+#define NV50_DISP_WNDW_NTFY(c)                    NV50_DISP_SYNC(1 + (c), 0x20)
+#define NV50_DISP_BASE_SEM0(c)                    NV50_DISP_WNDW_SEM0(0 + (c))
+#define NV50_DISP_BASE_SEM1(c)                    NV50_DISP_WNDW_SEM1(0 + (c))
+#define NV50_DISP_BASE_NTFY(c)                    NV50_DISP_WNDW_NTFY(0 + (c))
+	struct nouveau_bo *sync;
+
+	struct mutex mutex;
+};
+
+static inline struct nv50_disp *
+nv50_disp(struct drm_device *dev)
+{
+	return nouveau_display(dev)->priv;
+}
+
+struct nv50_chan {
+	struct nvif_object user;
+	struct nvif_device *device;
+};
+
+struct nv50_dmac {
+	struct nv50_chan base;
+
+	struct nvif_mem push;
+	u32 *ptr;
+
+	struct nvif_object sync;
+	struct nvif_object vram;
+
+	/* Protects against concurrent pushbuf access to this channel, lock is
+	 * grabbed by evo_wait (if the pushbuf reservation is successful) and
+	 * dropped again by evo_kick. */
+	struct mutex lock;
+};
+
+int nv50_dmac_create(struct nvif_device *device, struct nvif_object *disp,
+		     const s32 *oclass, u8 head, void *data, u32 size,
+		     u64 syncbuf, struct nv50_dmac *dmac);
+void nv50_dmac_destroy(struct nv50_dmac *);
+
+u32 *evo_wait(struct nv50_dmac *, int nr);
+void evo_kick(u32 *, struct nv50_dmac *);
+
+#define evo_mthd(p, m, s) do {						\
+	const u32 _m = (m), _s = (s);					\
+	if (drm_debug & DRM_UT_KMS)					\
+		pr_err("%04x %d %s\n", _m, _s, __func__);		\
+	*((p)++) = ((_s << 18) | _m);					\
+} while(0)
+
+#define evo_data(p, d) do {						\
+	const u32 _d = (d);						\
+	if (drm_debug & DRM_UT_KMS)					\
+		pr_err("\t%08x\n", _d);					\
+	*((p)++) = _d;							\
+} while(0)
+#endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head.c b/drivers/gpu/drm/nouveau/dispnv50/head.c
new file mode 100644
index 0000000..6a809ff
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/head.c
@@ -0,0 +1,566 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "head.h"
+#include "base.h"
+#include "core.h"
+#include "curs.h"
+#include "ovly.h"
+
+#include <nvif/class.h>
+
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include "nouveau_connector.h"
+#include "nouveau_bo.h"
+
+static void
+nv50_head_lut_load(struct drm_property_blob *blob, int mode,
+		   struct nouveau_bo *nvbo)
+{
+	struct drm_color_lut *in = (struct drm_color_lut *)blob->data;
+	void __iomem *lut = (u8 *)nvbo_kmap_obj_iovirtual(nvbo);
+	const int size = blob->length / sizeof(*in);
+	int bits, shift, i;
+	u16 zero, r, g, b;
+
+	/* This can't happen.. But it shuts the compiler up. */
+	if (WARN_ON(size != 256))
+		return;
+
+	switch (mode) {
+	case 0: /* LORES. */
+	case 1: /* HIRES. */
+		bits = 11;
+		shift = 3;
+		zero = 0x0000;
+		break;
+	case 7: /* INTERPOLATE_257_UNITY_RANGE. */
+		bits = 14;
+		shift = 0;
+		zero = 0x6000;
+		break;
+	default:
+		WARN_ON(1);
+		return;
+	}
+
+	for (i = 0; i < size; i++) {
+		r = (drm_color_lut_extract(in[i].  red, bits) + zero) << shift;
+		g = (drm_color_lut_extract(in[i].green, bits) + zero) << shift;
+		b = (drm_color_lut_extract(in[i]. blue, bits) + zero) << shift;
+		writew(r, lut + (i * 0x08) + 0);
+		writew(g, lut + (i * 0x08) + 2);
+		writew(b, lut + (i * 0x08) + 4);
+	}
+
+	/* INTERPOLATE modes require a "next" entry to interpolate with,
+	 * so we replicate the last entry to deal with this for now.
+	 */
+	writew(r, lut + (i * 0x08) + 0);
+	writew(g, lut + (i * 0x08) + 2);
+	writew(b, lut + (i * 0x08) + 4);
+}
+
+void
+nv50_head_flush_clr(struct nv50_head *head, struct nv50_head_atom *asyh, bool y)
+{
+	if (asyh->clr.ilut && (!asyh->set.ilut || y))
+		head->func->ilut_clr(head);
+	if (asyh->clr.core && (!asyh->set.core || y))
+		head->func->core_clr(head);
+	if (asyh->clr.curs && (!asyh->set.curs || y))
+		head->func->curs_clr(head);
+}
+
+void
+nv50_head_flush_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	if (asyh->set.view   ) head->func->view    (head, asyh);
+	if (asyh->set.mode   ) head->func->mode    (head, asyh);
+	if (asyh->set.ilut   ) {
+		struct nouveau_bo *nvbo = head->ilut.nvbo[head->ilut.next];
+		struct drm_property_blob *blob = asyh->state.gamma_lut;
+		if (blob)
+			nv50_head_lut_load(blob, asyh->ilut.mode, nvbo);
+		asyh->ilut.offset = nvbo->bo.offset;
+		head->ilut.next ^= 1;
+		head->func->ilut_set(head, asyh);
+	}
+	if (asyh->set.core   ) head->func->core_set(head, asyh);
+	if (asyh->set.curs   ) head->func->curs_set(head, asyh);
+	if (asyh->set.base   ) head->func->base    (head, asyh);
+	if (asyh->set.ovly   ) head->func->ovly    (head, asyh);
+	if (asyh->set.dither ) head->func->dither  (head, asyh);
+	if (asyh->set.procamp) head->func->procamp (head, asyh);
+	if (asyh->set.or     ) head->func->or      (head, asyh);
+}
+
+static void
+nv50_head_atomic_check_procamp(struct nv50_head_atom *armh,
+			       struct nv50_head_atom *asyh,
+			       struct nouveau_conn_atom *asyc)
+{
+	const int vib = asyc->procamp.color_vibrance - 100;
+	const int hue = asyc->procamp.vibrant_hue - 90;
+	const int adj = (vib > 0) ? 50 : 0;
+	asyh->procamp.sat.cos = ((vib * 2047 + adj) / 100) & 0xfff;
+	asyh->procamp.sat.sin = ((hue * 2047) / 100) & 0xfff;
+	asyh->set.procamp = true;
+}
+
+static void
+nv50_head_atomic_check_dither(struct nv50_head_atom *armh,
+			      struct nv50_head_atom *asyh,
+			      struct nouveau_conn_atom *asyc)
+{
+	struct drm_connector *connector = asyc->state.connector;
+	u32 mode = 0x00;
+
+	if (asyc->dither.mode == DITHERING_MODE_AUTO) {
+		if (asyh->base.depth > connector->display_info.bpc * 3)
+			mode = DITHERING_MODE_DYNAMIC2X2;
+	} else {
+		mode = asyc->dither.mode;
+	}
+
+	if (asyc->dither.depth == DITHERING_DEPTH_AUTO) {
+		if (connector->display_info.bpc >= 8)
+			mode |= DITHERING_DEPTH_8BPC;
+	} else {
+		mode |= asyc->dither.depth;
+	}
+
+	asyh->dither.enable = mode;
+	asyh->dither.bits = mode >> 1;
+	asyh->dither.mode = mode >> 3;
+	asyh->set.dither = true;
+}
+
+static void
+nv50_head_atomic_check_view(struct nv50_head_atom *armh,
+			    struct nv50_head_atom *asyh,
+			    struct nouveau_conn_atom *asyc)
+{
+	struct drm_connector *connector = asyc->state.connector;
+	struct drm_display_mode *omode = &asyh->state.adjusted_mode;
+	struct drm_display_mode *umode = &asyh->state.mode;
+	int mode = asyc->scaler.mode;
+	struct edid *edid;
+	int umode_vdisplay, omode_hdisplay, omode_vdisplay;
+
+	if (connector->edid_blob_ptr)
+		edid = (struct edid *)connector->edid_blob_ptr->data;
+	else
+		edid = NULL;
+
+	if (!asyc->scaler.full) {
+		if (mode == DRM_MODE_SCALE_NONE)
+			omode = umode;
+	} else {
+		/* Non-EDID LVDS/eDP mode. */
+		mode = DRM_MODE_SCALE_FULLSCREEN;
+	}
+
+	/* For the user-specified mode, we must ignore doublescan and
+	 * the like, but honor frame packing.
+	 */
+	umode_vdisplay = umode->vdisplay;
+	if ((umode->flags & DRM_MODE_FLAG_3D_MASK) == DRM_MODE_FLAG_3D_FRAME_PACKING)
+		umode_vdisplay += umode->vtotal;
+	asyh->view.iW = umode->hdisplay;
+	asyh->view.iH = umode_vdisplay;
+	/* For the output mode, we can just use the stock helper. */
+	drm_mode_get_hv_timing(omode, &omode_hdisplay, &omode_vdisplay);
+	asyh->view.oW = omode_hdisplay;
+	asyh->view.oH = omode_vdisplay;
+
+	/* Add overscan compensation if necessary, will keep the aspect
+	 * ratio the same as the backend mode unless overridden by the
+	 * user setting both hborder and vborder properties.
+	 */
+	if ((asyc->scaler.underscan.mode == UNDERSCAN_ON ||
+	    (asyc->scaler.underscan.mode == UNDERSCAN_AUTO &&
+	     drm_detect_hdmi_monitor(edid)))) {
+		u32 bX = asyc->scaler.underscan.hborder;
+		u32 bY = asyc->scaler.underscan.vborder;
+		u32 r = (asyh->view.oH << 19) / asyh->view.oW;
+
+		if (bX) {
+			asyh->view.oW -= (bX * 2);
+			if (bY) asyh->view.oH -= (bY * 2);
+			else    asyh->view.oH  = ((asyh->view.oW * r) + (r / 2)) >> 19;
+		} else {
+			asyh->view.oW -= (asyh->view.oW >> 4) + 32;
+			if (bY) asyh->view.oH -= (bY * 2);
+			else    asyh->view.oH  = ((asyh->view.oW * r) + (r / 2)) >> 19;
+		}
+	}
+
+	/* Handle CENTER/ASPECT scaling, taking into account the areas
+	 * removed already for overscan compensation.
+	 */
+	switch (mode) {
+	case DRM_MODE_SCALE_CENTER:
+		asyh->view.oW = min((u16)umode->hdisplay, asyh->view.oW);
+		asyh->view.oH = min((u16)umode_vdisplay, asyh->view.oH);
+		/* fall-through */
+	case DRM_MODE_SCALE_ASPECT:
+		if (asyh->view.oH < asyh->view.oW) {
+			u32 r = (asyh->view.iW << 19) / asyh->view.iH;
+			asyh->view.oW = ((asyh->view.oH * r) + (r / 2)) >> 19;
+		} else {
+			u32 r = (asyh->view.iH << 19) / asyh->view.iW;
+			asyh->view.oH = ((asyh->view.oW * r) + (r / 2)) >> 19;
+		}
+		break;
+	default:
+		break;
+	}
+
+	asyh->set.view = true;
+}
+
+static void
+nv50_head_atomic_check_lut(struct nv50_head *head,
+			   struct nv50_head_atom *armh,
+			   struct nv50_head_atom *asyh)
+{
+	struct nv50_disp *disp = nv50_disp(head->base.base.dev);
+
+	/* An I8 surface without an input LUT makes no sense, and
+	 * EVO will throw an error if you try.
+	 *
+	 * Legacy clients actually cause this due to the order in
+	 * which they call ioctls, so we will enable the LUT with
+	 * whatever contents the buffer already contains to avoid
+	 * triggering the error check.
+	 */
+	if (!asyh->state.gamma_lut && asyh->base.cpp != 1) {
+		asyh->ilut.handle = 0;
+		asyh->clr.ilut = armh->ilut.visible;
+		return;
+	}
+
+	if (disp->disp->object.oclass < GF110_DISP) {
+		asyh->ilut.mode = (asyh->base.cpp == 1) ? 0 : 1;
+		asyh->set.ilut = true;
+	} else {
+		asyh->ilut.mode = 7;
+		asyh->set.ilut = asyh->state.color_mgmt_changed;
+	}
+	asyh->ilut.handle = disp->core->chan.vram.handle;
+}
+
+static void
+nv50_head_atomic_check_mode(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct drm_display_mode *mode = &asyh->state.adjusted_mode;
+	struct nv50_head_mode *m = &asyh->mode;
+	u32 blankus;
+
+	drm_mode_set_crtcinfo(mode, CRTC_INTERLACE_HALVE_V | CRTC_STEREO_DOUBLE);
+
+	/*
+	 * DRM modes are defined in terms of a repeating interval
+	 * starting with the active display area.  The hardware modes
+	 * are defined in terms of a repeating interval starting one
+	 * unit (pixel or line) into the sync pulse.  So, add bias.
+	 */
+
+	m->h.active = mode->crtc_htotal;
+	m->h.synce  = mode->crtc_hsync_end - mode->crtc_hsync_start - 1;
+	m->h.blanke = mode->crtc_hblank_end - mode->crtc_hsync_start - 1;
+	m->h.blanks = m->h.blanke + mode->crtc_hdisplay;
+
+	m->v.active = mode->crtc_vtotal;
+	m->v.synce  = mode->crtc_vsync_end - mode->crtc_vsync_start - 1;
+	m->v.blanke = mode->crtc_vblank_end - mode->crtc_vsync_start - 1;
+	m->v.blanks = m->v.blanke + mode->crtc_vdisplay;
+
+	/*XXX: Safe underestimate, even "0" works */
+	blankus = (m->v.active - mode->crtc_vdisplay - 2) * m->h.active;
+	blankus *= 1000;
+	blankus /= mode->crtc_clock;
+	m->v.blankus = blankus;
+
+	if (mode->flags & DRM_MODE_FLAG_INTERLACE) {
+		m->v.blank2e =  m->v.active + m->v.blanke;
+		m->v.blank2s =  m->v.blank2e + mode->crtc_vdisplay;
+		m->v.active  = (m->v.active * 2) + 1;
+		m->interlace = true;
+	} else {
+		m->v.blank2e = 0;
+		m->v.blank2s = 1;
+		m->interlace = false;
+	}
+	m->clock = mode->crtc_clock;
+
+	asyh->or.nhsync = !!(mode->flags & DRM_MODE_FLAG_NHSYNC);
+	asyh->or.nvsync = !!(mode->flags & DRM_MODE_FLAG_NVSYNC);
+	asyh->set.or = head->func->or != NULL;
+	asyh->set.mode = true;
+}
+
+static int
+nv50_head_atomic_check(struct drm_crtc *crtc, struct drm_crtc_state *state)
+{
+	struct nouveau_drm *drm = nouveau_drm(crtc->dev);
+	struct nv50_disp *disp = nv50_disp(crtc->dev);
+	struct nv50_head *head = nv50_head(crtc);
+	struct nv50_head_atom *armh = nv50_head_atom(crtc->state);
+	struct nv50_head_atom *asyh = nv50_head_atom(state);
+	struct nouveau_conn_atom *asyc = NULL;
+	struct drm_connector_state *conns;
+	struct drm_connector *conn;
+	int i;
+
+	NV_ATOMIC(drm, "%s atomic_check %d\n", crtc->name, asyh->state.active);
+	if (asyh->state.active) {
+		for_each_new_connector_in_state(asyh->state.state, conn, conns, i) {
+			if (conns->crtc == crtc) {
+				asyc = nouveau_conn_atom(conns);
+				break;
+			}
+		}
+
+		if (armh->state.active) {
+			if (asyc) {
+				if (asyh->state.mode_changed)
+					asyc->set.scaler = true;
+				if (armh->base.depth != asyh->base.depth)
+					asyc->set.dither = true;
+			}
+		} else {
+			if (asyc)
+				asyc->set.mask = ~0;
+			asyh->set.mask = ~0;
+			asyh->set.or = head->func->or != NULL;
+		}
+
+		if (asyh->state.mode_changed)
+			nv50_head_atomic_check_mode(head, asyh);
+
+		if (asyh->state.color_mgmt_changed ||
+		    asyh->base.cpp != armh->base.cpp)
+			nv50_head_atomic_check_lut(head, armh, asyh);
+		asyh->ilut.visible = asyh->ilut.handle != 0;
+
+		if (asyc) {
+			if (asyc->set.scaler)
+				nv50_head_atomic_check_view(armh, asyh, asyc);
+			if (asyc->set.dither)
+				nv50_head_atomic_check_dither(armh, asyh, asyc);
+			if (asyc->set.procamp)
+				nv50_head_atomic_check_procamp(armh, asyh, asyc);
+		}
+
+		if ((asyh->core.visible = (asyh->base.cpp != 0))) {
+			asyh->core.x = asyh->base.x;
+			asyh->core.y = asyh->base.y;
+			asyh->core.w = asyh->base.w;
+			asyh->core.h = asyh->base.h;
+		} else
+		if ((asyh->core.visible = asyh->curs.visible) ||
+		    (asyh->core.visible = asyh->ilut.visible)) {
+			/*XXX: We need to either find some way of having the
+			 *     primary base layer appear black, while still
+			 *     being able to display the other layers, or we
+			 *     need to allocate a dummy black surface here.
+			 */
+			asyh->core.x = 0;
+			asyh->core.y = 0;
+			asyh->core.w = asyh->state.mode.hdisplay;
+			asyh->core.h = asyh->state.mode.vdisplay;
+		}
+		asyh->core.handle = disp->core->chan.vram.handle;
+		asyh->core.offset = 0;
+		asyh->core.format = 0xcf;
+		asyh->core.kind = 0;
+		asyh->core.layout = 1;
+		asyh->core.block = 0;
+		asyh->core.pitch = ALIGN(asyh->core.w, 64) * 4;
+		asyh->set.base = armh->base.cpp != asyh->base.cpp;
+		asyh->set.ovly = armh->ovly.cpp != asyh->ovly.cpp;
+	} else {
+		asyh->ilut.visible = false;
+		asyh->core.visible = false;
+		asyh->curs.visible = false;
+		asyh->base.cpp = 0;
+		asyh->ovly.cpp = 0;
+	}
+
+	if (!drm_atomic_crtc_needs_modeset(&asyh->state)) {
+		if (asyh->core.visible) {
+			if (memcmp(&armh->core, &asyh->core, sizeof(asyh->core)))
+				asyh->set.core = true;
+		} else
+		if (armh->core.visible) {
+			asyh->clr.core = true;
+		}
+
+		if (asyh->curs.visible) {
+			if (memcmp(&armh->curs, &asyh->curs, sizeof(asyh->curs)))
+				asyh->set.curs = true;
+		} else
+		if (armh->curs.visible) {
+			asyh->clr.curs = true;
+		}
+	} else {
+		asyh->clr.ilut = armh->ilut.visible;
+		asyh->clr.core = armh->core.visible;
+		asyh->clr.curs = armh->curs.visible;
+		asyh->set.ilut = asyh->ilut.visible;
+		asyh->set.core = asyh->core.visible;
+		asyh->set.curs = asyh->curs.visible;
+	}
+
+	if (asyh->clr.mask || asyh->set.mask)
+		nv50_atom(asyh->state.state)->lock_core = true;
+	return 0;
+}
+
+static const struct drm_crtc_helper_funcs
+nv50_head_help = {
+	.atomic_check = nv50_head_atomic_check,
+};
+
+static void
+nv50_head_atomic_destroy_state(struct drm_crtc *crtc,
+			       struct drm_crtc_state *state)
+{
+	struct nv50_head_atom *asyh = nv50_head_atom(state);
+	__drm_atomic_helper_crtc_destroy_state(&asyh->state);
+	kfree(asyh);
+}
+
+static struct drm_crtc_state *
+nv50_head_atomic_duplicate_state(struct drm_crtc *crtc)
+{
+	struct nv50_head_atom *armh = nv50_head_atom(crtc->state);
+	struct nv50_head_atom *asyh;
+	if (!(asyh = kmalloc(sizeof(*asyh), GFP_KERNEL)))
+		return NULL;
+	__drm_atomic_helper_crtc_duplicate_state(crtc, &asyh->state);
+	asyh->view = armh->view;
+	asyh->mode = armh->mode;
+	asyh->ilut = armh->ilut;
+	asyh->core = armh->core;
+	asyh->curs = armh->curs;
+	asyh->base = armh->base;
+	asyh->ovly = armh->ovly;
+	asyh->dither = armh->dither;
+	asyh->procamp = armh->procamp;
+	asyh->clr.mask = 0;
+	asyh->set.mask = 0;
+	return &asyh->state;
+}
+
+static void
+__drm_atomic_helper_crtc_reset(struct drm_crtc *crtc,
+			       struct drm_crtc_state *state)
+{
+	if (crtc->state)
+		crtc->funcs->atomic_destroy_state(crtc, crtc->state);
+	crtc->state = state;
+	crtc->state->crtc = crtc;
+}
+
+static void
+nv50_head_reset(struct drm_crtc *crtc)
+{
+	struct nv50_head_atom *asyh;
+
+	if (WARN_ON(!(asyh = kzalloc(sizeof(*asyh), GFP_KERNEL))))
+		return;
+
+	__drm_atomic_helper_crtc_reset(crtc, &asyh->state);
+}
+
+static void
+nv50_head_destroy(struct drm_crtc *crtc)
+{
+	struct nv50_head *head = nv50_head(crtc);
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(head->ilut.nvbo); i++)
+		nouveau_bo_unmap_unpin_unref(&head->ilut.nvbo[i]);
+
+	drm_crtc_cleanup(crtc);
+	kfree(head);
+}
+
+static const struct drm_crtc_funcs
+nv50_head_func = {
+	.reset = nv50_head_reset,
+	.gamma_set = drm_atomic_helper_legacy_gamma_set,
+	.destroy = nv50_head_destroy,
+	.set_config = drm_atomic_helper_set_config,
+	.page_flip = drm_atomic_helper_page_flip,
+	.atomic_duplicate_state = nv50_head_atomic_duplicate_state,
+	.atomic_destroy_state = nv50_head_atomic_destroy_state,
+};
+
+int
+nv50_head_create(struct drm_device *dev, int index)
+{
+	struct nouveau_drm *drm = nouveau_drm(dev);
+	struct nv50_disp *disp = nv50_disp(dev);
+	struct nv50_head *head;
+	struct nv50_wndw *curs, *wndw;
+	struct drm_crtc *crtc;
+	int ret, i;
+
+	head = kzalloc(sizeof(*head), GFP_KERNEL);
+	if (!head)
+		return -ENOMEM;
+
+	head->func = disp->core->func->head;
+	head->base.index = index;
+	ret = nv50_base_new(drm, head->base.index, &wndw);
+	if (ret == 0)
+		ret = nv50_curs_new(drm, head->base.index, &curs);
+	if (ret) {
+		kfree(head);
+		return ret;
+	}
+
+	crtc = &head->base.base;
+	drm_crtc_init_with_planes(dev, crtc, &wndw->plane, &curs->plane,
+				  &nv50_head_func, "head-%d", head->base.index);
+	drm_crtc_helper_add(crtc, &nv50_head_help);
+	drm_mode_crtc_set_gamma_size(crtc, 256);
+
+	for (i = 0; i < ARRAY_SIZE(head->ilut.nvbo); i++) {
+		ret = nouveau_bo_new_pin_map(&drm->client, 1025 * 8, 0x100,
+					     TTM_PL_FLAG_VRAM,
+					     &head->ilut.nvbo[i]);
+		if (ret)
+			goto out;
+	}
+
+	/* allocate overlay resources */
+	ret = nv50_ovly_new(drm, head->base.index, &wndw);
+out:
+	if (ret)
+		nv50_head_destroy(crtc);
+	return ret;
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head.h b/drivers/gpu/drm/nouveau/dispnv50/head.h
new file mode 100644
index 0000000..23099a8
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/head.h
@@ -0,0 +1,39 @@
+#ifndef __NV50_KMS_HEAD_H__
+#define __NV50_KMS_HEAD_H__
+#define nv50_head(c) container_of((c), struct nv50_head, base.base)
+#include "disp.h"
+#include "atom.h"
+
+#include "nouveau_crtc.h"
+
+struct nv50_head {
+	const struct nv50_head_func *func;
+	struct nouveau_crtc base;
+	struct {
+		struct nouveau_bo *nvbo[2];
+		int next;
+	} ilut;
+};
+
+int nv50_head_create(struct drm_device *, int index);
+void nv50_head_flush_set(struct nv50_head *, struct nv50_head_atom *);
+void nv50_head_flush_clr(struct nv50_head *, struct nv50_head_atom *, bool y);
+
+struct nv50_head_func {
+	void (*view)(struct nv50_head *, struct nv50_head_atom *);
+	void (*mode)(struct nv50_head *, struct nv50_head_atom *);
+	void (*ilut_set)(struct nv50_head *, struct nv50_head_atom *);
+	void (*ilut_clr)(struct nv50_head *);
+	void (*core_set)(struct nv50_head *, struct nv50_head_atom *);
+	void (*core_clr)(struct nv50_head *);
+	void (*curs_set)(struct nv50_head *, struct nv50_head_atom *);
+	void (*curs_clr)(struct nv50_head *);
+	void (*base)(struct nv50_head *, struct nv50_head_atom *);
+	void (*ovly)(struct nv50_head *, struct nv50_head_atom *);
+	void (*dither)(struct nv50_head *, struct nv50_head_atom *);
+	void (*procamp)(struct nv50_head *, struct nv50_head_atom *);
+	void (*or)(struct nv50_head *, struct nv50_head_atom *);
+};
+
+extern const struct nv50_head_func head507d;
+#endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head507d.c b/drivers/gpu/drm/nouveau/dispnv50/head507d.c
new file mode 100644
index 0000000..92fa249
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/head507d.c
@@ -0,0 +1,403 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "head.h"
+#include "core.h"
+
+#include <nvif/class.h>
+
+static void
+head907d_or(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if (core->base.user.oclass >= GF110_DISP_CORE_CHANNEL_DMA &&
+	    (push = evo_wait(core, 3))) {
+		evo_mthd(push, 0x0404 + (head->base.index * 0x300), 2);
+		evo_data(push, 0x00000001 | (asyh->or.depth  << 6) |
+					    (asyh->or.nvsync << 4) |
+					    (asyh->or.nhsync << 3));
+		evo_data(push, 0x31ec6000 | (head->base.index << 25) |
+					     asyh->mode.interlace);
+		evo_kick(push, core);
+	}
+}
+
+static void
+head507d_procamp(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
+			evo_mthd(push, 0x08a8 + (head->base.index * 0x400), 1);
+		else
+			evo_mthd(push, 0x0498 + (head->base.index * 0x300), 1);
+		evo_data(push, (asyh->procamp.sat.sin << 20) |
+			       (asyh->procamp.sat.cos << 8));
+		evo_kick(push, core);
+	}
+}
+
+static void
+head507d_dither(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
+			evo_mthd(push, 0x08a0 + (head->base.index * 0x0400), 1);
+		else
+		if (core->base.user.oclass < GK104_DISP_CORE_CHANNEL_DMA)
+			evo_mthd(push, 0x0490 + (head->base.index * 0x0300), 1);
+		else
+			evo_mthd(push, 0x04a0 + (head->base.index * 0x0300), 1);
+		evo_data(push, (asyh->dither.mode << 3) |
+			       (asyh->dither.bits << 1) |
+			        asyh->dither.enable);
+		evo_kick(push, core);
+	}
+}
+
+static void
+head507d_ovly(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 bounds = 0;
+	u32 *push;
+
+	if (asyh->ovly.cpp) {
+		switch (asyh->ovly.cpp) {
+		case 8: bounds |= 0x00000500; break;
+		case 4: bounds |= 0x00000300; break;
+		case 2: bounds |= 0x00000100; break;
+		default:
+			WARN_ON(1);
+			break;
+		}
+		bounds |= 0x00000001;
+	}
+
+	if ((push = evo_wait(core, 2))) {
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
+			evo_mthd(push, 0x0904 + head->base.index * 0x400, 1);
+		else
+			evo_mthd(push, 0x04d4 + head->base.index * 0x300, 1);
+		evo_data(push, bounds);
+		evo_kick(push, core);
+	}
+}
+
+static void
+head507d_base(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 bounds = 0;
+	u32 *push;
+
+	if (asyh->base.cpp) {
+		switch (asyh->base.cpp) {
+		case 8: bounds |= 0x00000500; break;
+		case 4: bounds |= 0x00000300; break;
+		case 2: bounds |= 0x00000100; break;
+		case 1: bounds |= 0x00000000; break;
+		default:
+			WARN_ON(1);
+			break;
+		}
+		bounds |= 0x00000001;
+	}
+
+	if ((push = evo_wait(core, 2))) {
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
+			evo_mthd(push, 0x0900 + head->base.index * 0x400, 1);
+		else
+			evo_mthd(push, 0x04d0 + head->base.index * 0x300, 1);
+		evo_data(push, bounds);
+		evo_kick(push, core);
+	}
+}
+
+static void
+head507d_curs_clr(struct nv50_head *head)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 4))) {
+		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0880 + head->base.index * 0x400, 1);
+			evo_data(push, 0x05000000);
+		} else
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0880 + head->base.index * 0x400, 1);
+			evo_data(push, 0x05000000);
+			evo_mthd(push, 0x089c + head->base.index * 0x400, 1);
+			evo_data(push, 0x00000000);
+		} else {
+			evo_mthd(push, 0x0480 + head->base.index * 0x300, 1);
+			evo_data(push, 0x05000000);
+			evo_mthd(push, 0x048c + head->base.index * 0x300, 1);
+			evo_data(push, 0x00000000);
+		}
+		evo_kick(push, core);
+	}
+}
+
+static void
+head507d_curs_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 5))) {
+		if (core->base.user.oclass < G82_DISP_BASE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0880 + head->base.index * 0x400, 2);
+			evo_data(push, 0x80000000 | (asyh->curs.layout << 26) |
+						    (asyh->curs.format << 24));
+			evo_data(push, asyh->curs.offset >> 8);
+		} else
+		if (core->base.user.oclass < GF110_DISP_BASE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0880 + head->base.index * 0x400, 2);
+			evo_data(push, 0x80000000 | (asyh->curs.layout << 26) |
+						    (asyh->curs.format << 24));
+			evo_data(push, asyh->curs.offset >> 8);
+			evo_mthd(push, 0x089c + head->base.index * 0x400, 1);
+			evo_data(push, asyh->curs.handle);
+		} else {
+			evo_mthd(push, 0x0480 + head->base.index * 0x300, 2);
+			evo_data(push, 0x80000000 | (asyh->curs.layout << 26) |
+						    (asyh->curs.format << 24));
+			evo_data(push, asyh->curs.offset >> 8);
+			evo_mthd(push, 0x048c + head->base.index * 0x300, 1);
+			evo_data(push, asyh->curs.handle);
+		}
+		evo_kick(push, core);
+	}
+}
+
+static void
+head507d_core_clr(struct nv50_head *head)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
+			evo_mthd(push, 0x0874 + head->base.index * 0x400, 1);
+		else
+			evo_mthd(push, 0x0474 + head->base.index * 0x300, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, core);
+	}
+}
+
+static void
+head507d_core_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 9))) {
+		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0860 + head->base.index * 0x400, 1);
+			evo_data(push, asyh->core.offset >> 8);
+			evo_mthd(push, 0x0868 + head->base.index * 0x400, 4);
+			evo_data(push, (asyh->core.h << 16) | asyh->core.w);
+			evo_data(push, asyh->core.layout << 20 |
+				       (asyh->core.pitch >> 8) << 8 |
+				       asyh->core.block);
+			evo_data(push, asyh->core.kind << 16 |
+				       asyh->core.format << 8);
+			evo_data(push, asyh->core.handle);
+			evo_mthd(push, 0x08c0 + head->base.index * 0x400, 1);
+			evo_data(push, (asyh->core.y << 16) | asyh->core.x);
+			/* EVO will complain with INVALID_STATE if we have an
+			 * active cursor and (re)specify HeadSetContextDmaIso
+			 * without also updating HeadSetOffsetCursor.
+			 */
+			asyh->set.curs = asyh->curs.visible;
+		} else
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0860 + head->base.index * 0x400, 1);
+			evo_data(push, asyh->core.offset >> 8);
+			evo_mthd(push, 0x0868 + head->base.index * 0x400, 4);
+			evo_data(push, (asyh->core.h << 16) | asyh->core.w);
+			evo_data(push, asyh->core.layout << 20 |
+				       (asyh->core.pitch >> 8) << 8 |
+				       asyh->core.block);
+			evo_data(push, asyh->core.format << 8);
+			evo_data(push, asyh->core.handle);
+			evo_mthd(push, 0x08c0 + head->base.index * 0x400, 1);
+			evo_data(push, (asyh->core.y << 16) | asyh->core.x);
+		} else {
+			evo_mthd(push, 0x0460 + head->base.index * 0x300, 1);
+			evo_data(push, asyh->core.offset >> 8);
+			evo_mthd(push, 0x0468 + head->base.index * 0x300, 4);
+			evo_data(push, (asyh->core.h << 16) | asyh->core.w);
+			evo_data(push, asyh->core.layout << 24 |
+				       (asyh->core.pitch >> 8) << 8 |
+				       asyh->core.block);
+			evo_data(push, asyh->core.format << 8);
+			evo_data(push, asyh->core.handle);
+			evo_mthd(push, 0x04b0 + head->base.index * 0x300, 1);
+			evo_data(push, (asyh->core.y << 16) | asyh->core.x);
+		}
+		evo_kick(push, core);
+	}
+}
+
+static void
+head507d_ilut_clr(struct nv50_head *head)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 4))) {
+		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0840 + (head->base.index * 0x400), 1);
+			evo_data(push, 0x40000000);
+		} else
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0840 + (head->base.index * 0x400), 1);
+			evo_data(push, 0x40000000);
+			evo_mthd(push, 0x085c + (head->base.index * 0x400), 1);
+			evo_data(push, 0x00000000);
+		} else {
+			evo_mthd(push, 0x0440 + (head->base.index * 0x300), 1);
+			evo_data(push, 0x03000000);
+			evo_mthd(push, 0x045c + (head->base.index * 0x300), 1);
+			evo_data(push, 0x00000000);
+		}
+		evo_kick(push, core);
+	}
+}
+
+static void
+head507d_ilut_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 7))) {
+		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0840 + (head->base.index * 0x400), 2);
+			evo_data(push, 0x80000000 | asyh->ilut.mode << 30);
+			evo_data(push, asyh->ilut.offset >> 8);
+		} else
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0840 + (head->base.index * 0x400), 2);
+			evo_data(push, 0x80000000 | asyh->ilut.mode << 30);
+			evo_data(push, asyh->ilut.offset >> 8);
+			evo_mthd(push, 0x085c + (head->base.index * 0x400), 1);
+			evo_data(push, asyh->ilut.handle);
+		} else {
+			evo_mthd(push, 0x0440 + (head->base.index * 0x300), 4);
+			evo_data(push, 0x80000000 | asyh->ilut.mode << 24);
+			evo_data(push, asyh->ilut.offset >> 8);
+			evo_data(push, 0x00000000);
+			evo_data(push, 0x00000000);
+			evo_mthd(push, 0x045c + (head->base.index * 0x300), 1);
+			evo_data(push, asyh->ilut.handle);
+		}
+		evo_kick(push, core);
+	}
+}
+
+static void
+head507d_mode(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	struct nv50_head_mode *m = &asyh->mode;
+	u32 *push;
+	if ((push = evo_wait(core, 14))) {
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
+			evo_mthd(push, 0x0804 + (head->base.index * 0x400), 2);
+			evo_data(push, 0x00800000 | m->clock);
+			evo_data(push, m->interlace ? 0x00000002 : 0x00000000);
+			evo_mthd(push, 0x0810 + (head->base.index * 0x400), 7);
+			evo_data(push, 0x00000000);
+			evo_data(push, (m->v.active  << 16) | m->h.active );
+			evo_data(push, (m->v.synce   << 16) | m->h.synce  );
+			evo_data(push, (m->v.blanke  << 16) | m->h.blanke );
+			evo_data(push, (m->v.blanks  << 16) | m->h.blanks );
+			evo_data(push, (m->v.blank2e << 16) | m->v.blank2s);
+			evo_data(push, asyh->mode.v.blankus);
+			evo_mthd(push, 0x082c + (head->base.index * 0x400), 1);
+			evo_data(push, 0x00000000);
+		} else {
+			evo_mthd(push, 0x0410 + (head->base.index * 0x300), 6);
+			evo_data(push, 0x00000000);
+			evo_data(push, (m->v.active  << 16) | m->h.active );
+			evo_data(push, (m->v.synce   << 16) | m->h.synce  );
+			evo_data(push, (m->v.blanke  << 16) | m->h.blanke );
+			evo_data(push, (m->v.blanks  << 16) | m->h.blanks );
+			evo_data(push, (m->v.blank2e << 16) | m->v.blank2s);
+			evo_mthd(push, 0x042c + (head->base.index * 0x300), 2);
+			evo_data(push, 0x00000000); /* ??? */
+			evo_data(push, 0xffffff00);
+			evo_mthd(push, 0x0450 + (head->base.index * 0x300), 3);
+			evo_data(push, m->clock * 1000);
+			evo_data(push, 0x00200000); /* ??? */
+			evo_data(push, m->clock * 1000);
+		}
+		evo_kick(push, core);
+	}
+}
+
+static void
+head507d_view(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 10))) {
+		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
+			evo_mthd(push, 0x08a4 + (head->base.index * 0x400), 1);
+			evo_data(push, 0x00000000);
+			evo_mthd(push, 0x08c8 + (head->base.index * 0x400), 1);
+			evo_data(push, (asyh->view.iH << 16) | asyh->view.iW);
+			evo_mthd(push, 0x08d8 + (head->base.index * 0x400), 2);
+			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
+			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
+		} else {
+			evo_mthd(push, 0x0494 + (head->base.index * 0x300), 1);
+			evo_data(push, 0x00000000);
+			evo_mthd(push, 0x04b8 + (head->base.index * 0x300), 1);
+			evo_data(push, (asyh->view.iH << 16) | asyh->view.iW);
+			evo_mthd(push, 0x04c0 + (head->base.index * 0x300), 3);
+			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
+			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
+			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
+		}
+		evo_kick(push, core);
+	}
+}
+
+const struct nv50_head_func
+head507d = {
+	.view = head507d_view,
+	.mode = head507d_mode,
+	.ilut_set = head507d_ilut_set,
+	.ilut_clr = head507d_ilut_clr,
+	.core_set = head507d_core_set,
+	.core_clr = head507d_core_clr,
+	.curs_set = head507d_curs_set,
+	.curs_clr = head507d_curs_clr,
+	.base = head507d_base,
+	.ovly = head507d_ovly,
+	.dither = head507d_dither,
+	.procamp = head507d_procamp,
+	.or = head907d_or,
+};
diff --git a/drivers/gpu/drm/nouveau/dispnv50/oimm.c b/drivers/gpu/drm/nouveau/dispnv50/oimm.c
new file mode 100644
index 0000000..2a2841d
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/oimm.c
@@ -0,0 +1,51 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "oimm.h"
+
+#include <nvif/class.h>
+
+int
+nv50_oimm_init(struct nouveau_drm *drm, struct nv50_wndw *wndw)
+{
+	static const struct {
+		s32 oclass;
+		int version;
+		int (*init)(struct nouveau_drm *, s32, struct nv50_wndw *);
+	} oimms[] = {
+		{ GK104_DISP_OVERLAY, 0, oimm507b_init },
+		{ GF110_DISP_OVERLAY, 0, oimm507b_init },
+		{ GT214_DISP_OVERLAY, 0, oimm507b_init },
+		{   G82_DISP_OVERLAY, 0, oimm507b_init },
+		{  NV50_DISP_OVERLAY, 0, oimm507b_init },
+		{}
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	int cid;
+
+	cid = nvif_mclass(&disp->disp->object, oimms);
+	if (cid < 0) {
+		NV_ERROR(drm, "No supported overlay immediate class\n");
+		return cid;
+	}
+
+	return oimms[cid].init(drm, oimms[cid].oclass, wndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/oimm.h b/drivers/gpu/drm/nouveau/dispnv50/oimm.h
new file mode 100644
index 0000000..6fa51f1
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/oimm.h
@@ -0,0 +1,8 @@
+#ifndef __NV50_KMS_OIMM_H__
+#define __NV50_KMS_OIMM_H__
+#include "wndw.h"
+
+int oimm507b_init(struct nouveau_drm *, s32, struct nv50_wndw *);
+
+int nv50_oimm_init(struct nouveau_drm *, struct nv50_wndw *);
+#endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/oimm507b.c b/drivers/gpu/drm/nouveau/dispnv50/oimm507b.c
new file mode 100644
index 0000000..c4baca8
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/oimm507b.c
@@ -0,0 +1,56 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "oimm.h"
+
+#include <nvif/cl507b.h>
+
+static const struct nv50_wimm_func
+oimm507b = {
+};
+
+static int
+oimm507b_init_(const struct nv50_wimm_func *func, struct nouveau_drm *drm,
+	       s32 oclass, struct nv50_wndw *wndw)
+{
+	struct nv50_disp_overlay_v0 args = {
+		.head = wndw->id,
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	int ret;
+
+	ret = nvif_object_init(&disp->disp->object, 0, oclass, &args,
+			       sizeof(args), &wndw->wimm.base.user);
+	if (ret) {
+		NV_ERROR(drm, "oimm%04x allocation failed: %d\n", oclass, ret);
+		return ret;
+	}
+
+	nvif_object_map(&wndw->wimm.base.user, NULL, 0);
+	wndw->immd = func;
+	return 0;
+}
+
+int
+oimm507b_init(struct nouveau_drm *drm, s32 oclass, struct nv50_wndw *wndw)
+{
+	return oimm507b_init_(&oimm507b, drm, oclass, wndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/ovly.c b/drivers/gpu/drm/nouveau/dispnv50/ovly.c
new file mode 100644
index 0000000..ac2d3b6
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/ovly.c
@@ -0,0 +1,57 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "ovly.h"
+#include "oimm.h"
+
+#include <nvif/class.h>
+
+int
+nv50_ovly_new(struct nouveau_drm *drm, int head, struct nv50_wndw **pwndw)
+{
+	static const struct {
+		s32 oclass;
+		int version;
+		int (*new)(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+	} ovlys[] = {
+		{ GK104_DISP_OVERLAY_CONTROL_DMA, 0, ovly507e_new },
+		{ GF110_DISP_OVERLAY_CONTROL_DMA, 0, ovly507e_new },
+		{ GT214_DISP_OVERLAY_CHANNEL_DMA, 0, ovly507e_new },
+		{ GT200_DISP_OVERLAY_CHANNEL_DMA, 0, ovly507e_new },
+		{   G82_DISP_OVERLAY_CHANNEL_DMA, 0, ovly507e_new },
+		{  NV50_DISP_OVERLAY_CHANNEL_DMA, 0, ovly507e_new },
+		{}
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	int cid, ret;
+
+	cid = nvif_mclass(&disp->disp->object, ovlys);
+	if (cid < 0) {
+		NV_ERROR(drm, "No supported overlay class\n");
+		return cid;
+	}
+
+	ret = ovlys[cid].new(drm, head, ovlys[cid].oclass, pwndw);
+	if (ret)
+		return ret;
+
+	return nv50_oimm_init(drm, *pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/ovly.h b/drivers/gpu/drm/nouveau/dispnv50/ovly.h
new file mode 100644
index 0000000..90af1f2
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/ovly.h
@@ -0,0 +1,8 @@
+#ifndef __NV50_KMS_OVLY_H__
+#define __NV50_KMS_OVLY_H__
+#include "wndw.h"
+
+int ovly507e_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+
+int nv50_ovly_new(struct nouveau_drm *, int head, struct nv50_wndw **);
+#endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/ovly507e.c b/drivers/gpu/drm/nouveau/dispnv50/ovly507e.c
new file mode 100644
index 0000000..ceec512
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/ovly507e.c
@@ -0,0 +1,70 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "ovly.h"
+
+#include <nvif/cl507e.h>
+
+#include "nouveau_bo.h"
+
+static const struct nv50_wndw_func
+ovly507e = {
+};
+
+static const u32
+ovly507e_format[] = {
+	0
+};
+
+static int
+ovly507e_new_(const struct nv50_wndw_func *func, const u32 *format,
+	      struct nouveau_drm *drm, int head, s32 oclass,
+	      struct nv50_wndw **pwndw)
+{
+	struct nv50_disp_overlay_channel_dma_v0 args = {
+		.head = head,
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	struct nv50_wndw *wndw;
+	int ret;
+
+	ret = nv50_wndw_new_(func, drm->dev, DRM_PLANE_TYPE_OVERLAY,
+			     "ovly", head, format, &wndw);
+	if (*pwndw = wndw, ret)
+		return ret;
+
+	ret = nv50_dmac_create(&drm->client.device, &disp->disp->object,
+			       &oclass, 0, &args, sizeof(args),
+			       disp->sync->bo.offset, &wndw->wndw);
+	if (ret) {
+		NV_ERROR(drm, "ovly%04x allocation failed: %d\n", oclass, ret);
+		return ret;
+	}
+
+	return 0;
+}
+
+int
+ovly507e_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return ovly507e_new_(&ovly507e, ovly507e_format, drm, head, oclass, pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/pior507d.c b/drivers/gpu/drm/nouveau/dispnv50/pior507d.c
new file mode 100644
index 0000000..a99ba6a
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/pior507d.c
@@ -0,0 +1,48 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "core.h"
+
+#include <nvif/class.h>
+
+static void
+pior507d_ctrl(struct nv50_core *core, int or, u32 ctrl,
+	      struct nv50_head_atom *asyh)
+{
+	u32 *push;
+	if ((push = evo_wait(&core->chan, 8))) {
+		if (core->chan.base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
+			if (asyh) {
+				ctrl |= asyh->or.depth  << 16;
+				ctrl |= asyh->or.nvsync << 13;
+				ctrl |= asyh->or.nhsync << 12;
+			}
+			evo_mthd(push, 0x0700 + (or * 0x040), 1);
+			evo_data(push, ctrl);
+		}
+		evo_kick(push, &core->chan);
+	}
+}
+
+const struct nv50_outp_func
+pior507d = {
+	.ctrl = pior507d_ctrl,
+};
diff --git a/drivers/gpu/drm/nouveau/dispnv50/sor507d.c b/drivers/gpu/drm/nouveau/dispnv50/sor507d.c
new file mode 100644
index 0000000..2d540de
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/sor507d.c
@@ -0,0 +1,50 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "core.h"
+
+#include <nvif/class.h>
+
+static void
+sor507d_ctrl(struct nv50_core *core, int or, u32 ctrl,
+	     struct nv50_head_atom *asyh)
+{
+	u32 *push;
+	if ((push = evo_wait(&core->chan, 6))) {
+		if (core->chan.base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
+			if (asyh) {
+				ctrl |= asyh->or.depth  << 16;
+				ctrl |= asyh->or.nvsync << 13;
+				ctrl |= asyh->or.nhsync << 12;
+			}
+			evo_mthd(push, 0x0600 + (or * 0x40), 1);
+		} else {
+			evo_mthd(push, 0x0200 + (or * 0x20), 1);
+		}
+		evo_data(push, ctrl);
+		evo_kick(push, &core->chan);
+	}
+}
+
+const struct nv50_outp_func
+sor507d = {
+	.ctrl = sor507d_ctrl,
+};
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.c b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
new file mode 100644
index 0000000..71a4c50
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
@@ -0,0 +1,434 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "wndw.h"
+
+#include <nvif/class.h>
+#include <nvif/cl0002.h>
+
+#include <drm/drm_atomic_helper.h>
+#include "nouveau_bo.h"
+
+static void
+nv50_wndw_ctxdma_del(struct nv50_wndw_ctxdma *ctxdma)
+{
+	nvif_object_fini(&ctxdma->object);
+	list_del(&ctxdma->head);
+	kfree(ctxdma);
+}
+
+static struct nv50_wndw_ctxdma *
+nv50_wndw_ctxdma_new(struct nv50_wndw *wndw, struct nouveau_framebuffer *fb)
+{
+	struct nouveau_drm *drm = nouveau_drm(fb->base.dev);
+	struct nv50_wndw_ctxdma *ctxdma;
+	const u8    kind = fb->nvbo->kind;
+	const u32 handle = 0xfb000000 | kind;
+	struct {
+		struct nv_dma_v0 base;
+		union {
+			struct nv50_dma_v0 nv50;
+			struct gf100_dma_v0 gf100;
+			struct gf119_dma_v0 gf119;
+		};
+	} args = {};
+	u32 argc = sizeof(args.base);
+	int ret;
+
+	list_for_each_entry(ctxdma, &wndw->ctxdma.list, head) {
+		if (ctxdma->object.handle == handle)
+			return ctxdma;
+	}
+
+	if (!(ctxdma = kzalloc(sizeof(*ctxdma), GFP_KERNEL)))
+		return ERR_PTR(-ENOMEM);
+	list_add(&ctxdma->head, &wndw->ctxdma.list);
+
+	args.base.target = NV_DMA_V0_TARGET_VRAM;
+	args.base.access = NV_DMA_V0_ACCESS_RDWR;
+	args.base.start  = 0;
+	args.base.limit  = drm->client.device.info.ram_user - 1;
+
+	if (drm->client.device.info.chipset < 0x80) {
+		args.nv50.part = NV50_DMA_V0_PART_256;
+		argc += sizeof(args.nv50);
+	} else
+	if (drm->client.device.info.chipset < 0xc0) {
+		args.nv50.part = NV50_DMA_V0_PART_256;
+		args.nv50.kind = kind;
+		argc += sizeof(args.nv50);
+	} else
+	if (drm->client.device.info.chipset < 0xd0) {
+		args.gf100.kind = kind;
+		argc += sizeof(args.gf100);
+	} else {
+		args.gf119.page = GF119_DMA_V0_PAGE_LP;
+		args.gf119.kind = kind;
+		argc += sizeof(args.gf119);
+	}
+
+	ret = nvif_object_init(wndw->ctxdma.parent, handle, NV_DMA_IN_MEMORY,
+			       &args, argc, &ctxdma->object);
+	if (ret) {
+		nv50_wndw_ctxdma_del(ctxdma);
+		return ERR_PTR(ret);
+	}
+
+	return ctxdma;
+}
+
+int
+nv50_wndw_wait_armed(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	if (asyw->set.ntfy)
+		return wndw->func->ntfy_wait_begun(wndw, asyw);
+	return 0;
+}
+
+u32
+nv50_wndw_flush_clr(struct nv50_wndw *wndw, u32 interlock, bool flush,
+		    struct nv50_wndw_atom *asyw)
+{
+	if (asyw->clr.sema && (!asyw->set.sema || flush))
+		wndw->func->sema_clr(wndw);
+	if (asyw->clr.ntfy && (!asyw->set.ntfy || flush))
+		wndw->func->ntfy_clr(wndw);
+	if (asyw->clr.image && (!asyw->set.image || flush))
+		wndw->func->image_clr(wndw);
+
+	return flush ? wndw->func->update(wndw, interlock) : 0;
+}
+
+u32
+nv50_wndw_flush_set(struct nv50_wndw *wndw, u32 interlock,
+		    struct nv50_wndw_atom *asyw)
+{
+	if (interlock) {
+		asyw->image.mode = 0;
+		asyw->image.interval = 1;
+	}
+
+	if (asyw->set.sema ) wndw->func->sema_set (wndw, asyw);
+	if (asyw->set.ntfy ) wndw->func->ntfy_set (wndw, asyw);
+	if (asyw->set.image) wndw->func->image_set(wndw, asyw);
+	if (asyw->set.lut  ) wndw->func->lut      (wndw, asyw);
+	if (asyw->set.point) {
+		wndw->immd->point(wndw, asyw);
+		wndw->immd->update(wndw, interlock);
+	}
+
+	return wndw->func->update ? wndw->func->update(wndw, interlock) : 0;
+}
+
+static void
+nv50_wndw_atomic_check_release(struct nv50_wndw *wndw,
+			       struct nv50_wndw_atom *asyw,
+			       struct nv50_head_atom *asyh)
+{
+	struct nouveau_drm *drm = nouveau_drm(wndw->plane.dev);
+	NV_ATOMIC(drm, "%s release\n", wndw->plane.name);
+	wndw->func->release(wndw, asyw, asyh);
+	asyw->ntfy.handle = 0;
+	asyw->sema.handle = 0;
+}
+
+static int
+nv50_wndw_atomic_check_acquire(struct nv50_wndw *wndw,
+			       struct nv50_wndw_atom *asyw,
+			       struct nv50_head_atom *asyh)
+{
+	struct nouveau_framebuffer *fb = nouveau_framebuffer(asyw->state.fb);
+	struct nouveau_drm *drm = nouveau_drm(wndw->plane.dev);
+	int ret;
+
+	NV_ATOMIC(drm, "%s acquire\n", wndw->plane.name);
+
+	asyw->image.w = fb->base.width;
+	asyw->image.h = fb->base.height;
+	asyw->image.kind = fb->nvbo->kind;
+
+	if (asyh->state.pageflip_flags & DRM_MODE_PAGE_FLIP_ASYNC)
+		asyw->interval = 0;
+	else
+		asyw->interval = 1;
+
+	if (asyw->image.kind) {
+		asyw->image.layout = 0;
+		if (drm->client.device.info.chipset >= 0xc0)
+			asyw->image.block = fb->nvbo->mode >> 4;
+		else
+			asyw->image.block = fb->nvbo->mode;
+		asyw->image.pitch = (fb->base.pitches[0] / 4) << 4;
+	} else {
+		asyw->image.layout = 1;
+		asyw->image.block  = 0;
+		asyw->image.pitch  = fb->base.pitches[0];
+	}
+
+	ret = wndw->func->acquire(wndw, asyw, asyh);
+	if (ret)
+		return ret;
+
+	if (asyw->set.image) {
+		if (!(asyw->image.mode = asyw->interval ? 0 : 1))
+			asyw->image.interval = asyw->interval;
+		else
+			asyw->image.interval = 0;
+	}
+
+	return 0;
+}
+
+int
+nv50_wndw_atomic_check(struct drm_plane *plane, struct drm_plane_state *state)
+{
+	struct nouveau_drm *drm = nouveau_drm(plane->dev);
+	struct nv50_wndw *wndw = nv50_wndw(plane);
+	struct nv50_wndw_atom *armw = nv50_wndw_atom(wndw->plane.state);
+	struct nv50_wndw_atom *asyw = nv50_wndw_atom(state);
+	struct nv50_head_atom *harm = NULL, *asyh = NULL;
+	bool varm = false, asyv = false, asym = false;
+	int ret;
+
+	NV_ATOMIC(drm, "%s atomic_check\n", plane->name);
+	if (asyw->state.crtc) {
+		asyh = nv50_head_atom_get(asyw->state.state, asyw->state.crtc);
+		if (IS_ERR(asyh))
+			return PTR_ERR(asyh);
+		asym = drm_atomic_crtc_needs_modeset(&asyh->state);
+		asyv = asyh->state.active;
+	}
+
+	if (armw->state.crtc) {
+		harm = nv50_head_atom_get(asyw->state.state, armw->state.crtc);
+		if (IS_ERR(harm))
+			return PTR_ERR(harm);
+		varm = harm->state.crtc->state->active;
+	}
+
+	if (asyv) {
+		asyw->point.x = asyw->state.crtc_x;
+		asyw->point.y = asyw->state.crtc_y;
+		if (memcmp(&armw->point, &asyw->point, sizeof(asyw->point)))
+			asyw->set.point = true;
+
+		ret = nv50_wndw_atomic_check_acquire(wndw, asyw, asyh);
+		if (ret)
+			return ret;
+	} else
+	if (varm) {
+		nv50_wndw_atomic_check_release(wndw, asyw, harm);
+	} else {
+		return 0;
+	}
+
+	if (!asyv || asym) {
+		asyw->clr.ntfy = armw->ntfy.handle != 0;
+		asyw->clr.sema = armw->sema.handle != 0;
+		if (wndw->func->image_clr)
+			asyw->clr.image = armw->image.handle != 0;
+		asyw->set.lut = wndw->func->lut && asyv;
+	}
+
+	return 0;
+}
+
+static void
+nv50_wndw_cleanup_fb(struct drm_plane *plane, struct drm_plane_state *old_state)
+{
+	struct nouveau_framebuffer *fb = nouveau_framebuffer(old_state->fb);
+	struct nouveau_drm *drm = nouveau_drm(plane->dev);
+
+	NV_ATOMIC(drm, "%s cleanup: %p\n", plane->name, old_state->fb);
+	if (!old_state->fb)
+		return;
+
+	nouveau_bo_unpin(fb->nvbo);
+}
+
+static int
+nv50_wndw_prepare_fb(struct drm_plane *plane, struct drm_plane_state *state)
+{
+	struct nouveau_framebuffer *fb = nouveau_framebuffer(state->fb);
+	struct nouveau_drm *drm = nouveau_drm(plane->dev);
+	struct nv50_wndw *wndw = nv50_wndw(plane);
+	struct nv50_wndw_atom *asyw = nv50_wndw_atom(state);
+	struct nv50_head_atom *asyh;
+	struct nv50_wndw_ctxdma *ctxdma;
+	int ret;
+
+	NV_ATOMIC(drm, "%s prepare: %p\n", plane->name, state->fb);
+	if (!asyw->state.fb)
+		return 0;
+
+	ret = nouveau_bo_pin(fb->nvbo, TTM_PL_FLAG_VRAM, true);
+	if (ret)
+		return ret;
+
+	ctxdma = nv50_wndw_ctxdma_new(wndw, fb);
+	if (IS_ERR(ctxdma)) {
+		nouveau_bo_unpin(fb->nvbo);
+		return PTR_ERR(ctxdma);
+	}
+
+	asyw->state.fence = reservation_object_get_excl_rcu(fb->nvbo->bo.resv);
+	asyw->image.handle = ctxdma->object.handle;
+	asyw->image.offset = fb->nvbo->bo.offset;
+
+	if (wndw->func->prepare) {
+		asyh = nv50_head_atom_get(asyw->state.state, asyw->state.crtc);
+		if (IS_ERR(asyh))
+			return PTR_ERR(asyh);
+
+		wndw->func->prepare(wndw, asyh, asyw);
+	}
+
+	return 0;
+}
+
+static const struct drm_plane_helper_funcs
+nv50_wndw_helper = {
+	.prepare_fb = nv50_wndw_prepare_fb,
+	.cleanup_fb = nv50_wndw_cleanup_fb,
+	.atomic_check = nv50_wndw_atomic_check,
+};
+
+static void
+nv50_wndw_atomic_destroy_state(struct drm_plane *plane,
+			       struct drm_plane_state *state)
+{
+	struct nv50_wndw_atom *asyw = nv50_wndw_atom(state);
+	__drm_atomic_helper_plane_destroy_state(&asyw->state);
+	kfree(asyw);
+}
+
+static struct drm_plane_state *
+nv50_wndw_atomic_duplicate_state(struct drm_plane *plane)
+{
+	struct nv50_wndw_atom *armw = nv50_wndw_atom(plane->state);
+	struct nv50_wndw_atom *asyw;
+	if (!(asyw = kmalloc(sizeof(*asyw), GFP_KERNEL)))
+		return NULL;
+	__drm_atomic_helper_plane_duplicate_state(plane, &asyw->state);
+	asyw->interval = 1;
+	asyw->sema = armw->sema;
+	asyw->ntfy = armw->ntfy;
+	asyw->image = armw->image;
+	asyw->point = armw->point;
+	asyw->lut = armw->lut;
+	asyw->clr.mask = 0;
+	asyw->set.mask = 0;
+	return &asyw->state;
+}
+
+static void
+nv50_wndw_reset(struct drm_plane *plane)
+{
+	struct nv50_wndw_atom *asyw;
+
+	if (WARN_ON(!(asyw = kzalloc(sizeof(*asyw), GFP_KERNEL))))
+		return;
+
+	if (plane->state)
+		plane->funcs->atomic_destroy_state(plane, plane->state);
+	plane->state = &asyw->state;
+	plane->state->plane = plane;
+	plane->state->rotation = DRM_MODE_ROTATE_0;
+}
+
+static void
+nv50_wndw_destroy(struct drm_plane *plane)
+{
+	struct nv50_wndw *wndw = nv50_wndw(plane);
+	struct nv50_wndw_ctxdma *ctxdma, *ctxtmp;
+
+	list_for_each_entry_safe(ctxdma, ctxtmp, &wndw->ctxdma.list, head) {
+		nv50_wndw_ctxdma_del(ctxdma);
+	}
+
+	nvif_notify_fini(&wndw->notify);
+	nv50_dmac_destroy(&wndw->wimm);
+	nv50_dmac_destroy(&wndw->wndw);
+	drm_plane_cleanup(&wndw->plane);
+	kfree(wndw);
+}
+
+const struct drm_plane_funcs
+nv50_wndw = {
+	.update_plane = drm_atomic_helper_update_plane,
+	.disable_plane = drm_atomic_helper_disable_plane,
+	.destroy = nv50_wndw_destroy,
+	.reset = nv50_wndw_reset,
+	.atomic_duplicate_state = nv50_wndw_atomic_duplicate_state,
+	.atomic_destroy_state = nv50_wndw_atomic_destroy_state,
+};
+
+static int
+nv50_wndw_notify(struct nvif_notify *notify)
+{
+	return NVIF_NOTIFY_KEEP;
+}
+
+void
+nv50_wndw_fini(struct nv50_wndw *wndw)
+{
+	nvif_notify_put(&wndw->notify);
+}
+
+void
+nv50_wndw_init(struct nv50_wndw *wndw)
+{
+	nvif_notify_get(&wndw->notify);
+}
+
+int
+nv50_wndw_new_(const struct nv50_wndw_func *func, struct drm_device *dev,
+	       enum drm_plane_type type, const char *name, int index,
+	       const u32 *format, struct nv50_wndw **pwndw)
+{
+	struct nv50_wndw *wndw;
+	int nformat;
+	int ret;
+
+	if (!(wndw = *pwndw = kzalloc(sizeof(*wndw), GFP_KERNEL)))
+		return -ENOMEM;
+	wndw->func = func;
+	wndw->id = index;
+
+	wndw->ctxdma.parent = &wndw->wndw.base.user;
+	INIT_LIST_HEAD(&wndw->ctxdma.list);
+
+	for (nformat = 0; format[nformat]; nformat++);
+
+	ret = drm_universal_plane_init(dev, &wndw->plane, 0, &nv50_wndw,
+				       format, nformat, NULL,
+				       type, "%s-%d", name, index);
+	if (ret) {
+		kfree(*pwndw);
+		*pwndw = NULL;
+		return ret;
+	}
+
+	drm_plane_helper_add(&wndw->plane, &nv50_wndw_helper);
+
+	wndw->notify.func = nv50_wndw_notify;
+	return 0;
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.h b/drivers/gpu/drm/nouveau/dispnv50/wndw.h
new file mode 100644
index 0000000..1931e30
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.h
@@ -0,0 +1,73 @@
+#ifndef __NV50_KMS_WNDW_H__
+#define __NV50_KMS_WNDW_H__
+#define nv50_wndw(p) container_of((p), struct nv50_wndw, plane)
+#include "disp.h"
+#include "atom.h"
+
+#include <nvif/notify.h>
+
+struct nv50_wndw_ctxdma {
+	struct list_head head;
+	struct nvif_object object;
+};
+
+struct nv50_wndw {
+	const struct nv50_wndw_func *func;
+	const struct nv50_wimm_func *immd;
+	int id;
+
+	struct {
+		struct nvif_object *parent;
+		struct list_head list;
+	} ctxdma;
+
+	struct drm_plane plane;
+
+	struct nv50_dmac wndw;
+	struct nv50_dmac wimm;
+
+	struct nvif_notify notify;
+	u16 ntfy;
+	u16 sema;
+	u32 data;
+};
+
+int nv50_wndw_new_(const struct nv50_wndw_func *, struct drm_device *,
+		   enum drm_plane_type, const char *name, int index,
+		   const u32 *format, struct nv50_wndw **);
+void nv50_wndw_init(struct nv50_wndw *);
+void nv50_wndw_fini(struct nv50_wndw *);
+u32 nv50_wndw_flush_set(struct nv50_wndw *, u32 interlock,
+			struct nv50_wndw_atom *);
+u32 nv50_wndw_flush_clr(struct nv50_wndw *, u32 interlock, bool flush,
+			struct nv50_wndw_atom *);
+int nv50_wndw_wait_armed(struct nv50_wndw *, struct nv50_wndw_atom *);
+
+struct nv50_wndw_func {
+	int (*acquire)(struct nv50_wndw *, struct nv50_wndw_atom *asyw,
+		       struct nv50_head_atom *asyh);
+	void (*release)(struct nv50_wndw *, struct nv50_wndw_atom *asyw,
+			struct nv50_head_atom *asyh);
+	void (*prepare)(struct nv50_wndw *, struct nv50_head_atom *asyh,
+			struct nv50_wndw_atom *asyw);
+
+	void (*sema_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
+	void (*sema_clr)(struct nv50_wndw *);
+	void (*ntfy_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
+	void (*ntfy_clr)(struct nv50_wndw *);
+	int (*ntfy_wait_begun)(struct nv50_wndw *, struct nv50_wndw_atom *);
+	void (*image_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
+	void (*image_clr)(struct nv50_wndw *);
+	void (*lut)(struct nv50_wndw *, struct nv50_wndw_atom *);
+
+	u32 (*update)(struct nv50_wndw *, u32 interlock);
+};
+
+extern const struct drm_plane_funcs nv50_wndw;
+
+struct nv50_wimm_func {
+	void (*point)(struct nv50_wndw *, struct nv50_wndw_atom *);
+
+	u32 (*update)(struct nv50_wndw *, u32 interlock);
+};
+#endif
diff --git a/drivers/gpu/drm/nouveau/nv50_display.h b/drivers/gpu/drm/nouveau/nv50_display.h
index 918187c..fbd3b15 100644
--- a/drivers/gpu/drm/nouveau/nv50_display.h
+++ b/drivers/gpu/drm/nouveau/nv50_display.h
@@ -28,7 +28,6 @@
 #define __NV50_DISPLAY_H__
 
 #include "nouveau_display.h"
-#include "nouveau_crtc.h"
 #include "nouveau_reg.h"
 
 int  nv50_display_create(struct drm_device *);
-- 
cgit v1.1


From 09e1b78aab5715eacab02e4047c7a47d72f6a1e9 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50-: split core implementation by hardware class

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/Kbuild     |   8 +
 drivers/gpu/drm/nouveau/dispnv50/core.c     |  22 +-
 drivers/gpu/drm/nouveau/dispnv50/core.h     |  20 ++
 drivers/gpu/drm/nouveau/dispnv50/core507d.c |  51 ++++-
 drivers/gpu/drm/nouveau/dispnv50/core827d.c |  41 ++++
 drivers/gpu/drm/nouveau/dispnv50/core907d.c |  40 ++++
 drivers/gpu/drm/nouveau/dispnv50/core917d.c |  40 ++++
 drivers/gpu/drm/nouveau/dispnv50/dac507d.c  |  19 +-
 drivers/gpu/drm/nouveau/dispnv50/dac907d.c  |  39 ++++
 drivers/gpu/drm/nouveau/dispnv50/disp.c     |  46 +---
 drivers/gpu/drm/nouveau/dispnv50/head.c     |  29 +--
 drivers/gpu/drm/nouveau/dispnv50/head.h     |  27 +++
 drivers/gpu/drm/nouveau/dispnv50/head507d.c | 328 +++++++++-------------------
 drivers/gpu/drm/nouveau/dispnv50/head827d.c | 120 ++++++++++
 drivers/gpu/drm/nouveau/dispnv50/head907d.c | 274 +++++++++++++++++++++++
 drivers/gpu/drm/nouveau/dispnv50/head917d.c |  55 +++++
 drivers/gpu/drm/nouveau/dispnv50/pior507d.c |  18 +-
 drivers/gpu/drm/nouveau/dispnv50/sor507d.c  |  18 +-
 drivers/gpu/drm/nouveau/dispnv50/sor907d.c  |  41 ++++
 19 files changed, 903 insertions(+), 333 deletions(-)
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/core827d.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/core907d.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/core917d.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/dac907d.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/head827d.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/head907d.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/head917d.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/sor907d.c

diff --git a/drivers/gpu/drm/nouveau/dispnv50/Kbuild b/drivers/gpu/drm/nouveau/dispnv50/Kbuild
index f3877d2..cde3ae9 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/Kbuild
+++ b/drivers/gpu/drm/nouveau/dispnv50/Kbuild
@@ -2,15 +2,23 @@ nouveau-y += dispnv50/disp.o
 
 nouveau-y += dispnv50/core.o
 nouveau-y += dispnv50/core507d.o
+nouveau-y += dispnv50/core827d.o
+nouveau-y += dispnv50/core907d.o
+nouveau-y += dispnv50/core917d.o
 
 nouveau-y += dispnv50/dac507d.o
+nouveau-y += dispnv50/dac907d.o
 
 nouveau-y += dispnv50/pior507d.o
 
 nouveau-y += dispnv50/sor507d.o
+nouveau-y += dispnv50/sor907d.o
 
 nouveau-y += dispnv50/head.o
 nouveau-y += dispnv50/head507d.o
+nouveau-y += dispnv50/head827d.o
+nouveau-y += dispnv50/head907d.o
+nouveau-y += dispnv50/head917d.o
 
 nouveau-y += dispnv50/wndw.o
 
diff --git a/drivers/gpu/drm/nouveau/dispnv50/core.c b/drivers/gpu/drm/nouveau/dispnv50/core.c
index b12899f..f87cbaa 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/core.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/core.c
@@ -42,17 +42,17 @@ nv50_core_new(struct nouveau_drm *drm, struct nv50_core **pcore)
 		int version;
 		int (*new)(struct nouveau_drm *, s32, struct nv50_core **);
 	} cores[] = {
-		{ GP102_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
-		{ GP100_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
-		{ GM200_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
-		{ GM107_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
-		{ GK110_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
-		{ GK104_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
-		{ GF110_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
-		{ GT214_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
-		{ GT206_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
-		{ GT200_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
-		{   G82_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{ GP102_DISP_CORE_CHANNEL_DMA, 0, core917d_new },
+		{ GP100_DISP_CORE_CHANNEL_DMA, 0, core917d_new },
+		{ GM200_DISP_CORE_CHANNEL_DMA, 0, core917d_new },
+		{ GM107_DISP_CORE_CHANNEL_DMA, 0, core917d_new },
+		{ GK110_DISP_CORE_CHANNEL_DMA, 0, core917d_new },
+		{ GK104_DISP_CORE_CHANNEL_DMA, 0, core917d_new },
+		{ GF110_DISP_CORE_CHANNEL_DMA, 0, core907d_new },
+		{ GT214_DISP_CORE_CHANNEL_DMA, 0, core827d_new },
+		{ GT206_DISP_CORE_CHANNEL_DMA, 0, core827d_new },
+		{ GT200_DISP_CORE_CHANNEL_DMA, 0, core827d_new },
+		{   G82_DISP_CORE_CHANNEL_DMA, 0, core827d_new },
 		{  NV50_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
 		{}
 	};
diff --git a/drivers/gpu/drm/nouveau/dispnv50/core.h b/drivers/gpu/drm/nouveau/dispnv50/core.h
index 3cd5446..5fd7ddd 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/core.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/core.h
@@ -12,6 +12,12 @@ int nv50_core_new(struct nouveau_drm *, struct nv50_core **);
 void nv50_core_del(struct nv50_core **);
 
 struct nv50_core_func {
+	void (*init)(struct nv50_core *);
+	void (*ntfy_init)(struct nouveau_bo *, u32 offset);
+	int (*ntfy_wait_done)(struct nouveau_bo *, u32 offset,
+			      struct nvif_device *);
+	void (*update)(struct nv50_core *, u32 interlock, bool ntfy);
+
 	const struct nv50_head_func *head;
 	const struct nv50_outp_func {
 		void (*ctrl)(struct nv50_core *, int or, u32 ctrl,
@@ -20,7 +26,21 @@ struct nv50_core_func {
 };
 
 int core507d_new(struct nouveau_drm *, s32, struct nv50_core **);
+int core507d_new_(const struct nv50_core_func *, struct nouveau_drm *, s32,
+		  struct nv50_core **);
+void core507d_init(struct nv50_core *);
+void core507d_ntfy_init(struct nouveau_bo *, u32);
+int core507d_ntfy_wait_done(struct nouveau_bo *, u32, struct nvif_device *);
+void core507d_update(struct nv50_core *, u32, bool);
 extern const struct nv50_outp_func dac507d;
 extern const struct nv50_outp_func sor507d;
 extern const struct nv50_outp_func pior507d;
+
+int core827d_new(struct nouveau_drm *, s32, struct nv50_core **);
+
+int core907d_new(struct nouveau_drm *, s32, struct nv50_core **);
+extern const struct nv50_outp_func dac907d;
+extern const struct nv50_outp_func sor907d;
+
+int core917d_new(struct nouveau_drm *, s32, struct nv50_core **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/core507d.c b/drivers/gpu/drm/nouveau/dispnv50/core507d.c
index b0325f6..96d7d8f 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/core507d.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/core507d.c
@@ -26,15 +26,64 @@
 
 #include "nouveau_bo.h"
 
+void
+core507d_update(struct nv50_core *core, u32 interlock, bool ntfy)
+{
+	u32 *push;
+	if ((push = evo_wait(&core->chan, 5))) {
+		if (ntfy) {
+			evo_mthd(push, 0x0084, 1);
+			evo_data(push, 0x80000000 | NV50_DISP_CORE_NTFY);
+		}
+		evo_mthd(push, 0x0080, 2);
+		evo_data(push, interlock);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &core->chan);
+	}
+}
+
+int
+core507d_ntfy_wait_done(struct nouveau_bo *bo, u32 offset,
+			struct nvif_device *device)
+{
+	s64 time = nvif_msec(device, 2000ULL,
+		if (nouveau_bo_rd32(bo, offset / 4))
+			break;
+		usleep_range(1, 2);
+	);
+	return time < 0 ? time : 0;
+}
+
+void
+core507d_ntfy_init(struct nouveau_bo *bo, u32 offset)
+{
+	nouveau_bo_wr32(bo, offset / 4, 0x00000000);
+}
+
+void
+core507d_init(struct nv50_core *core)
+{
+	u32 *push;
+	if ((push = evo_wait(&core->chan, 2))) {
+		evo_mthd(push, 0x0088, 1);
+		evo_data(push, core->chan.sync.handle);
+		evo_kick(push, &core->chan);
+	}
+}
+
 static const struct nv50_core_func
 core507d = {
+	.init = core507d_init,
+	.ntfy_init = core507d_ntfy_init,
+	.ntfy_wait_done = core507d_ntfy_wait_done,
+	.update = core507d_update,
 	.head = &head507d,
 	.dac = &dac507d,
 	.sor = &sor507d,
 	.pior = &pior507d,
 };
 
-static int
+int
 core507d_new_(const struct nv50_core_func *func, struct nouveau_drm *drm,
 	      s32 oclass, struct nv50_core **pcore)
 {
diff --git a/drivers/gpu/drm/nouveau/dispnv50/core827d.c b/drivers/gpu/drm/nouveau/dispnv50/core827d.c
new file mode 100644
index 0000000..6123a06
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/core827d.c
@@ -0,0 +1,41 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "core.h"
+#include "head.h"
+
+static const struct nv50_core_func
+core827d = {
+	.init = core507d_init,
+	.ntfy_init = core507d_ntfy_init,
+	.ntfy_wait_done = core507d_ntfy_wait_done,
+	.update = core507d_update,
+	.head = &head827d,
+	.dac = &dac507d,
+	.sor = &sor507d,
+	.pior = &pior507d,
+};
+
+int
+core827d_new(struct nouveau_drm *drm, s32 oclass, struct nv50_core **pcore)
+{
+	return core507d_new_(&core827d, drm, oclass, pcore);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/core907d.c b/drivers/gpu/drm/nouveau/dispnv50/core907d.c
new file mode 100644
index 0000000..ef822f8
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/core907d.c
@@ -0,0 +1,40 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "core.h"
+#include "head.h"
+
+static const struct nv50_core_func
+core907d = {
+	.init = core507d_init,
+	.ntfy_init = core507d_ntfy_init,
+	.ntfy_wait_done = core507d_ntfy_wait_done,
+	.update = core507d_update,
+	.head = &head907d,
+	.dac = &dac907d,
+	.sor = &sor907d,
+};
+
+int
+core907d_new(struct nouveau_drm *drm, s32 oclass, struct nv50_core **pcore)
+{
+	return core507d_new_(&core907d, drm, oclass, pcore);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/core917d.c b/drivers/gpu/drm/nouveau/dispnv50/core917d.c
new file mode 100644
index 0000000..392338d
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/core917d.c
@@ -0,0 +1,40 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "core.h"
+#include "head.h"
+
+static const struct nv50_core_func
+core917d = {
+	.init = core507d_init,
+	.ntfy_init = core507d_ntfy_init,
+	.ntfy_wait_done = core507d_ntfy_wait_done,
+	.update = core507d_update,
+	.head = &head917d,
+	.dac = &dac907d,
+	.sor = &sor907d,
+};
+
+int
+core917d_new(struct nouveau_drm *drm, s32 oclass, struct nv50_core **pcore)
+{
+	return core507d_new_(&core917d, drm, oclass, pcore);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/dac507d.c b/drivers/gpu/drm/nouveau/dispnv50/dac507d.c
index 28b6025..2a10ef7 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/dac507d.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/dac507d.c
@@ -21,26 +21,19 @@
  */
 #include "core.h"
 
-#include <nvif/class.h>
-
 static void
 dac507d_ctrl(struct nv50_core *core, int or, u32 ctrl,
 	     struct nv50_head_atom *asyh)
 {
 	u32 *push, sync = 0;
 	if ((push = evo_wait(&core->chan, 3))) {
-		if (core->chan.base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			if (asyh) {
-				sync |= asyh->or.nvsync << 1;
-				sync |= asyh->or.nhsync;
-			}
-			evo_mthd(push, 0x0400 + (or * 0x080), 2);
-			evo_data(push, ctrl);
-			evo_data(push, sync);
-		} else {
-			evo_mthd(push, 0x0180 + (or * 0x020), 1);
-			evo_data(push, ctrl);
+		if (asyh) {
+			sync |= asyh->or.nvsync << 1;
+			sync |= asyh->or.nhsync;
 		}
+		evo_mthd(push, 0x0400 + (or * 0x080), 2);
+		evo_data(push, ctrl);
+		evo_data(push, sync);
 		evo_kick(push, &core->chan);
 	}
 }
diff --git a/drivers/gpu/drm/nouveau/dispnv50/dac907d.c b/drivers/gpu/drm/nouveau/dispnv50/dac907d.c
new file mode 100644
index 0000000..11e87fa
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/dac907d.c
@@ -0,0 +1,39 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "core.h"
+
+static void
+dac907d_ctrl(struct nv50_core *core, int or, u32 ctrl,
+	     struct nv50_head_atom *asyh)
+{
+	u32 *push;
+	if ((push = evo_wait(&core->chan, 2))) {
+		evo_mthd(push, 0x0180 + (or * 0x020), 1);
+		evo_data(push, ctrl);
+		evo_kick(push, &core->chan);
+	}
+}
+
+const struct nv50_outp_func
+dac907d = {
+	.ctrl = dac907d_ctrl,
+};
diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c
index a8367c5..6136bee 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/disp.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c
@@ -1587,10 +1587,9 @@ static void
 nv50_disp_atomic_commit_core(struct nouveau_drm *drm, u32 interlock)
 {
 	struct nv50_disp *disp = nv50_disp(drm->dev);
-	struct nv50_dmac *core = &disp->core->chan;
+	struct nv50_core *core = disp->core;
 	struct nv50_mstm *mstm;
 	struct drm_encoder *encoder;
-	u32 *push;
 
 	NV_ATOMIC(drm, "commit core %08x\n", interlock);
 
@@ -1602,21 +1601,11 @@ nv50_disp_atomic_commit_core(struct nouveau_drm *drm, u32 interlock)
 		}
 	}
 
-	if ((push = evo_wait(core, 5))) {
-		evo_mthd(push, 0x0084, 1);
-		evo_data(push, 0x80000000);
-		evo_mthd(push, 0x0080, 2);
-		evo_data(push, interlock);
-		evo_data(push, 0x00000000);
-		nouveau_bo_wr32(disp->sync, 0, 0x00000000);
-		evo_kick(push, core);
-		if (nvif_msec(&drm->client.device, 2000ULL,
-			if (nouveau_bo_rd32(disp->sync, 0))
-				break;
-			usleep_range(1, 2);
-		) < 0)
-			NV_ERROR(drm, "EVO timeout\n");
-	}
+	core->func->ntfy_init(disp->sync, NV50_DISP_CORE_NTFY);
+	core->func->update(core, interlock, true);
+	if (core->func->ntfy_wait_done(disp->sync, NV50_DISP_CORE_NTFY,
+				       disp->core->chan.base.device))
+		NV_ERROR(drm, "core notifier timeout\n");
 
 	drm_for_each_encoder(encoder, drm->dev) {
 		if (encoder->encoder_type != DRM_MODE_ENCODER_DPMST) {
@@ -1770,16 +1759,10 @@ nv50_disp_atomic_commit_tail(struct drm_atomic_state *state)
 
 	/* Flush update. */
 	if (interlock_core) {
-		if (!interlock_chan && atom->state.legacy_cursor_update) {
-			u32 *push = evo_wait(&disp->core->chan, 2);
-			if (push) {
-				evo_mthd(push, 0x0080, 1);
-				evo_data(push, 0x00000000);
-				evo_kick(push, &disp->core->chan);
-			}
-		} else {
+		if (interlock_chan || !atom->state.legacy_cursor_update)
 			nv50_disp_atomic_commit_core(drm, interlock_chan);
-		}
+		else
+			disp->core->func->update(disp->core, 0, false);
 	}
 
 	if (atom->lock_core)
@@ -2079,18 +2062,11 @@ nv50_display_fini(struct drm_device *dev)
 int
 nv50_display_init(struct drm_device *dev)
 {
-	struct nv50_dmac *core = &nv50_disp(dev)->core->chan;
+	struct nv50_core *core = nv50_disp(dev)->core;
 	struct drm_encoder *encoder;
 	struct drm_plane *plane;
-	u32 *push;
-
-	push = evo_wait(core, 32);
-	if (!push)
-		return -EBUSY;
 
-	evo_mthd(push, 0x0088, 1);
-	evo_data(push, core->sync.handle);
-	evo_kick(push, core);
+	core->func->init(core);
 
 	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
 		if (encoder->encoder_type != DRM_MODE_ENCODER_DPMST) {
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head.c b/drivers/gpu/drm/nouveau/dispnv50/head.c
index 6a809ff..1335c00 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/head.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/head.c
@@ -324,7 +324,6 @@ static int
 nv50_head_atomic_check(struct drm_crtc *crtc, struct drm_crtc_state *state)
 {
 	struct nouveau_drm *drm = nouveau_drm(crtc->dev);
-	struct nv50_disp *disp = nv50_disp(crtc->dev);
 	struct nv50_head *head = nv50_head(crtc);
 	struct nv50_head_atom *armh = nv50_head_atom(crtc->state);
 	struct nv50_head_atom *asyh = nv50_head_atom(state);
@@ -373,31 +372,9 @@ nv50_head_atomic_check(struct drm_crtc *crtc, struct drm_crtc_state *state)
 				nv50_head_atomic_check_procamp(armh, asyh, asyc);
 		}
 
-		if ((asyh->core.visible = (asyh->base.cpp != 0))) {
-			asyh->core.x = asyh->base.x;
-			asyh->core.y = asyh->base.y;
-			asyh->core.w = asyh->base.w;
-			asyh->core.h = asyh->base.h;
-		} else
-		if ((asyh->core.visible = asyh->curs.visible) ||
-		    (asyh->core.visible = asyh->ilut.visible)) {
-			/*XXX: We need to either find some way of having the
-			 *     primary base layer appear black, while still
-			 *     being able to display the other layers, or we
-			 *     need to allocate a dummy black surface here.
-			 */
-			asyh->core.x = 0;
-			asyh->core.y = 0;
-			asyh->core.w = asyh->state.mode.hdisplay;
-			asyh->core.h = asyh->state.mode.vdisplay;
-		}
-		asyh->core.handle = disp->core->chan.vram.handle;
-		asyh->core.offset = 0;
-		asyh->core.format = 0xcf;
-		asyh->core.kind = 0;
-		asyh->core.layout = 1;
-		asyh->core.block = 0;
-		asyh->core.pitch = ALIGN(asyh->core.w, 64) * 4;
+		if (head->func->core_calc)
+			head->func->core_calc(head, asyh);
+
 		asyh->set.base = armh->base.cpp != asyh->base.cpp;
 		asyh->set.ovly = armh->ovly.cpp != asyh->ovly.cpp;
 	} else {
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head.h b/drivers/gpu/drm/nouveau/dispnv50/head.h
index 23099a8..d00cebd 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/head.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/head.h
@@ -24,6 +24,7 @@ struct nv50_head_func {
 	void (*mode)(struct nv50_head *, struct nv50_head_atom *);
 	void (*ilut_set)(struct nv50_head *, struct nv50_head_atom *);
 	void (*ilut_clr)(struct nv50_head *);
+	void (*core_calc)(struct nv50_head *, struct nv50_head_atom *);
 	void (*core_set)(struct nv50_head *, struct nv50_head_atom *);
 	void (*core_clr)(struct nv50_head *);
 	void (*curs_set)(struct nv50_head *, struct nv50_head_atom *);
@@ -36,4 +37,30 @@ struct nv50_head_func {
 };
 
 extern const struct nv50_head_func head507d;
+void head507d_view(struct nv50_head *, struct nv50_head_atom *);
+void head507d_mode(struct nv50_head *, struct nv50_head_atom *);
+void head507d_core_calc(struct nv50_head *, struct nv50_head_atom *);
+void head507d_core_clr(struct nv50_head *);
+void head507d_base(struct nv50_head *, struct nv50_head_atom *);
+void head507d_ovly(struct nv50_head *, struct nv50_head_atom *);
+void head507d_dither(struct nv50_head *, struct nv50_head_atom *);
+void head507d_procamp(struct nv50_head *, struct nv50_head_atom *);
+
+extern const struct nv50_head_func head827d;
+
+extern const struct nv50_head_func head907d;
+void head907d_view(struct nv50_head *, struct nv50_head_atom *);
+void head907d_mode(struct nv50_head *, struct nv50_head_atom *);
+void head907d_ilut_set(struct nv50_head *, struct nv50_head_atom *);
+void head907d_ilut_clr(struct nv50_head *);
+void head907d_core_set(struct nv50_head *, struct nv50_head_atom *);
+void head907d_core_clr(struct nv50_head *);
+void head907d_curs_set(struct nv50_head *, struct nv50_head_atom *);
+void head907d_curs_clr(struct nv50_head *);
+void head907d_base(struct nv50_head *, struct nv50_head_atom *);
+void head907d_ovly(struct nv50_head *, struct nv50_head_atom *);
+void head907d_procamp(struct nv50_head *, struct nv50_head_atom *);
+void head907d_or(struct nv50_head *, struct nv50_head_atom *);
+
+extern const struct nv50_head_func head917d;
 #endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head507d.c b/drivers/gpu/drm/nouveau/dispnv50/head507d.c
index 92fa249..5f06fa1 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/head507d.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/head507d.c
@@ -22,62 +22,34 @@
 #include "head.h"
 #include "core.h"
 
-#include <nvif/class.h>
-
-static void
-head907d_or(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
-	u32 *push;
-	if (core->base.user.oclass >= GF110_DISP_CORE_CHANNEL_DMA &&
-	    (push = evo_wait(core, 3))) {
-		evo_mthd(push, 0x0404 + (head->base.index * 0x300), 2);
-		evo_data(push, 0x00000001 | (asyh->or.depth  << 6) |
-					    (asyh->or.nvsync << 4) |
-					    (asyh->or.nhsync << 3));
-		evo_data(push, 0x31ec6000 | (head->base.index << 25) |
-					     asyh->mode.interlace);
-		evo_kick(push, core);
-	}
-}
-
-static void
+void
 head507d_procamp(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
 	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
 	if ((push = evo_wait(core, 2))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
-			evo_mthd(push, 0x08a8 + (head->base.index * 0x400), 1);
-		else
-			evo_mthd(push, 0x0498 + (head->base.index * 0x300), 1);
-		evo_data(push, (asyh->procamp.sat.sin << 20) |
-			       (asyh->procamp.sat.cos << 8));
+		evo_mthd(push, 0x08a8 + (head->base.index * 0x400), 1);
+		evo_data(push, asyh->procamp.sat.sin << 20 |
+			       asyh->procamp.sat.cos << 8);
 		evo_kick(push, core);
 	}
 }
 
-static void
+void
 head507d_dither(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
 	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
 	if ((push = evo_wait(core, 2))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
-			evo_mthd(push, 0x08a0 + (head->base.index * 0x0400), 1);
-		else
-		if (core->base.user.oclass < GK104_DISP_CORE_CHANNEL_DMA)
-			evo_mthd(push, 0x0490 + (head->base.index * 0x0300), 1);
-		else
-			evo_mthd(push, 0x04a0 + (head->base.index * 0x0300), 1);
-		evo_data(push, (asyh->dither.mode << 3) |
-			       (asyh->dither.bits << 1) |
-			        asyh->dither.enable);
+		evo_mthd(push, 0x08a0 + (head->base.index * 0x0400), 1);
+		evo_data(push, asyh->dither.mode << 3 |
+			       asyh->dither.bits << 1 |
+			       asyh->dither.enable);
 		evo_kick(push, core);
 	}
 }
 
-static void
+void
 head507d_ovly(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
 	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
@@ -97,16 +69,13 @@ head507d_ovly(struct nv50_head *head, struct nv50_head_atom *asyh)
 	}
 
 	if ((push = evo_wait(core, 2))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
-			evo_mthd(push, 0x0904 + head->base.index * 0x400, 1);
-		else
-			evo_mthd(push, 0x04d4 + head->base.index * 0x300, 1);
+		evo_mthd(push, 0x0904 + head->base.index * 0x400, 1);
 		evo_data(push, bounds);
 		evo_kick(push, core);
 	}
 }
 
-static void
+void
 head507d_base(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
 	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
@@ -127,10 +96,7 @@ head507d_base(struct nv50_head *head, struct nv50_head_atom *asyh)
 	}
 
 	if ((push = evo_wait(core, 2))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
-			evo_mthd(push, 0x0900 + head->base.index * 0x400, 1);
-		else
-			evo_mthd(push, 0x04d0 + head->base.index * 0x300, 1);
+		evo_mthd(push, 0x0900 + head->base.index * 0x400, 1);
 		evo_data(push, bounds);
 		evo_kick(push, core);
 	}
@@ -141,22 +107,9 @@ head507d_curs_clr(struct nv50_head *head)
 {
 	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
-	if ((push = evo_wait(core, 4))) {
-		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0880 + head->base.index * 0x400, 1);
-			evo_data(push, 0x05000000);
-		} else
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0880 + head->base.index * 0x400, 1);
-			evo_data(push, 0x05000000);
-			evo_mthd(push, 0x089c + head->base.index * 0x400, 1);
-			evo_data(push, 0x00000000);
-		} else {
-			evo_mthd(push, 0x0480 + head->base.index * 0x300, 1);
-			evo_data(push, 0x05000000);
-			evo_mthd(push, 0x048c + head->base.index * 0x300, 1);
-			evo_data(push, 0x00000000);
-		}
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x0880 + head->base.index * 0x400, 1);
+		evo_data(push, 0x05000000);
 		evo_kick(push, core);
 	}
 }
@@ -166,42 +119,22 @@ head507d_curs_set(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
 	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
-	if ((push = evo_wait(core, 5))) {
-		if (core->base.user.oclass < G82_DISP_BASE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0880 + head->base.index * 0x400, 2);
-			evo_data(push, 0x80000000 | (asyh->curs.layout << 26) |
-						    (asyh->curs.format << 24));
-			evo_data(push, asyh->curs.offset >> 8);
-		} else
-		if (core->base.user.oclass < GF110_DISP_BASE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0880 + head->base.index * 0x400, 2);
-			evo_data(push, 0x80000000 | (asyh->curs.layout << 26) |
-						    (asyh->curs.format << 24));
-			evo_data(push, asyh->curs.offset >> 8);
-			evo_mthd(push, 0x089c + head->base.index * 0x400, 1);
-			evo_data(push, asyh->curs.handle);
-		} else {
-			evo_mthd(push, 0x0480 + head->base.index * 0x300, 2);
-			evo_data(push, 0x80000000 | (asyh->curs.layout << 26) |
-						    (asyh->curs.format << 24));
-			evo_data(push, asyh->curs.offset >> 8);
-			evo_mthd(push, 0x048c + head->base.index * 0x300, 1);
-			evo_data(push, asyh->curs.handle);
-		}
+	if ((push = evo_wait(core, 3))) {
+		evo_mthd(push, 0x0880 + head->base.index * 0x400, 2);
+		evo_data(push, 0x80000000 | asyh->curs.layout << 26 |
+					    asyh->curs.format << 24);
+		evo_data(push, asyh->curs.offset >> 8);
 		evo_kick(push, core);
 	}
 }
 
-static void
+void
 head507d_core_clr(struct nv50_head *head)
 {
 	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
 	if ((push = evo_wait(core, 2))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
-			evo_mthd(push, 0x0874 + head->base.index * 0x400, 1);
-		else
-			evo_mthd(push, 0x0474 + head->base.index * 0x300, 1);
+		evo_mthd(push, 0x0874 + head->base.index * 0x400, 1);
 		evo_data(push, 0x00000000);
 		evo_kick(push, core);
 	}
@@ -213,52 +146,57 @@ head507d_core_set(struct nv50_head *head, struct nv50_head_atom *asyh)
 	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
 	if ((push = evo_wait(core, 9))) {
-		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0860 + head->base.index * 0x400, 1);
-			evo_data(push, asyh->core.offset >> 8);
-			evo_mthd(push, 0x0868 + head->base.index * 0x400, 4);
-			evo_data(push, (asyh->core.h << 16) | asyh->core.w);
-			evo_data(push, asyh->core.layout << 20 |
-				       (asyh->core.pitch >> 8) << 8 |
-				       asyh->core.block);
-			evo_data(push, asyh->core.kind << 16 |
-				       asyh->core.format << 8);
-			evo_data(push, asyh->core.handle);
-			evo_mthd(push, 0x08c0 + head->base.index * 0x400, 1);
-			evo_data(push, (asyh->core.y << 16) | asyh->core.x);
-			/* EVO will complain with INVALID_STATE if we have an
-			 * active cursor and (re)specify HeadSetContextDmaIso
-			 * without also updating HeadSetOffsetCursor.
-			 */
-			asyh->set.curs = asyh->curs.visible;
-		} else
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0860 + head->base.index * 0x400, 1);
-			evo_data(push, asyh->core.offset >> 8);
-			evo_mthd(push, 0x0868 + head->base.index * 0x400, 4);
-			evo_data(push, (asyh->core.h << 16) | asyh->core.w);
-			evo_data(push, asyh->core.layout << 20 |
-				       (asyh->core.pitch >> 8) << 8 |
-				       asyh->core.block);
-			evo_data(push, asyh->core.format << 8);
-			evo_data(push, asyh->core.handle);
-			evo_mthd(push, 0x08c0 + head->base.index * 0x400, 1);
-			evo_data(push, (asyh->core.y << 16) | asyh->core.x);
-		} else {
-			evo_mthd(push, 0x0460 + head->base.index * 0x300, 1);
-			evo_data(push, asyh->core.offset >> 8);
-			evo_mthd(push, 0x0468 + head->base.index * 0x300, 4);
-			evo_data(push, (asyh->core.h << 16) | asyh->core.w);
-			evo_data(push, asyh->core.layout << 24 |
-				       (asyh->core.pitch >> 8) << 8 |
-				       asyh->core.block);
-			evo_data(push, asyh->core.format << 8);
-			evo_data(push, asyh->core.handle);
-			evo_mthd(push, 0x04b0 + head->base.index * 0x300, 1);
-			evo_data(push, (asyh->core.y << 16) | asyh->core.x);
-		}
+		evo_mthd(push, 0x0860 + head->base.index * 0x400, 1);
+		evo_data(push, asyh->core.offset >> 8);
+		evo_mthd(push, 0x0868 + head->base.index * 0x400, 4);
+		evo_data(push, asyh->core.h << 16 | asyh->core.w);
+		evo_data(push, asyh->core.layout << 20 |
+			       asyh->core.pitch >> 8 << 8 |
+			       asyh->core.block);
+		evo_data(push, asyh->core.kind << 16 |
+			       asyh->core.format << 8);
+		evo_data(push, asyh->core.handle);
+		evo_mthd(push, 0x08c0 + head->base.index * 0x400, 1);
+		evo_data(push, asyh->core.y << 16 | asyh->core.x);
 		evo_kick(push, core);
+
+		/* EVO will complain with INVALID_STATE if we have an
+		 * active cursor and (re)specify HeadSetContextDmaIso
+		 * without also updating HeadSetOffsetCursor.
+		 */
+		asyh->set.curs = asyh->curs.visible;
+	}
+}
+
+void
+head507d_core_calc(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_disp *disp = nv50_disp(head->base.base.dev);
+	if ((asyh->core.visible = (asyh->base.cpp != 0))) {
+		asyh->core.x = asyh->base.x;
+		asyh->core.y = asyh->base.y;
+		asyh->core.w = asyh->base.w;
+		asyh->core.h = asyh->base.h;
+	} else
+	if ((asyh->core.visible = asyh->curs.visible) ||
+	    (asyh->core.visible = asyh->ilut.visible)) {
+		/*XXX: We need to either find some way of having the
+		 *     primary base layer appear black, while still
+		 *     being able to display the other layers, or we
+		 *     need to allocate a dummy black surface here.
+		 */
+		asyh->core.x = 0;
+		asyh->core.y = 0;
+		asyh->core.w = asyh->state.mode.hdisplay;
+		asyh->core.h = asyh->state.mode.vdisplay;
 	}
+	asyh->core.handle = disp->core->chan.vram.handle;
+	asyh->core.offset = 0;
+	asyh->core.format = 0xcf;
+	asyh->core.kind = 0;
+	asyh->core.layout = 1;
+	asyh->core.block = 0;
+	asyh->core.pitch = ALIGN(asyh->core.w, 64) * 4;
 }
 
 static void
@@ -266,22 +204,9 @@ head507d_ilut_clr(struct nv50_head *head)
 {
 	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
-	if ((push = evo_wait(core, 4))) {
-		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0840 + (head->base.index * 0x400), 1);
-			evo_data(push, 0x40000000);
-		} else
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0840 + (head->base.index * 0x400), 1);
-			evo_data(push, 0x40000000);
-			evo_mthd(push, 0x085c + (head->base.index * 0x400), 1);
-			evo_data(push, 0x00000000);
-		} else {
-			evo_mthd(push, 0x0440 + (head->base.index * 0x300), 1);
-			evo_data(push, 0x03000000);
-			evo_mthd(push, 0x045c + (head->base.index * 0x300), 1);
-			evo_data(push, 0x00000000);
-		}
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x0840 + (head->base.index * 0x400), 1);
+		evo_data(push, 0x40000000);
 		evo_kick(push, core);
 	}
 }
@@ -291,96 +216,51 @@ head507d_ilut_set(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
 	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
-	if ((push = evo_wait(core, 7))) {
-		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0840 + (head->base.index * 0x400), 2);
-			evo_data(push, 0x80000000 | asyh->ilut.mode << 30);
-			evo_data(push, asyh->ilut.offset >> 8);
-		} else
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0840 + (head->base.index * 0x400), 2);
-			evo_data(push, 0x80000000 | asyh->ilut.mode << 30);
-			evo_data(push, asyh->ilut.offset >> 8);
-			evo_mthd(push, 0x085c + (head->base.index * 0x400), 1);
-			evo_data(push, asyh->ilut.handle);
-		} else {
-			evo_mthd(push, 0x0440 + (head->base.index * 0x300), 4);
-			evo_data(push, 0x80000000 | asyh->ilut.mode << 24);
-			evo_data(push, asyh->ilut.offset >> 8);
-			evo_data(push, 0x00000000);
-			evo_data(push, 0x00000000);
-			evo_mthd(push, 0x045c + (head->base.index * 0x300), 1);
-			evo_data(push, asyh->ilut.handle);
-		}
+	if ((push = evo_wait(core, 3))) {
+		evo_mthd(push, 0x0840 + (head->base.index * 0x400), 2);
+		evo_data(push, 0x80000000 | asyh->ilut.mode << 30);
+		evo_data(push, asyh->ilut.offset >> 8);
 		evo_kick(push, core);
 	}
 }
 
-static void
+void
 head507d_mode(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
 	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	struct nv50_head_mode *m = &asyh->mode;
 	u32 *push;
-	if ((push = evo_wait(core, 14))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0804 + (head->base.index * 0x400), 2);
-			evo_data(push, 0x00800000 | m->clock);
-			evo_data(push, m->interlace ? 0x00000002 : 0x00000000);
-			evo_mthd(push, 0x0810 + (head->base.index * 0x400), 7);
-			evo_data(push, 0x00000000);
-			evo_data(push, (m->v.active  << 16) | m->h.active );
-			evo_data(push, (m->v.synce   << 16) | m->h.synce  );
-			evo_data(push, (m->v.blanke  << 16) | m->h.blanke );
-			evo_data(push, (m->v.blanks  << 16) | m->h.blanks );
-			evo_data(push, (m->v.blank2e << 16) | m->v.blank2s);
-			evo_data(push, asyh->mode.v.blankus);
-			evo_mthd(push, 0x082c + (head->base.index * 0x400), 1);
-			evo_data(push, 0x00000000);
-		} else {
-			evo_mthd(push, 0x0410 + (head->base.index * 0x300), 6);
-			evo_data(push, 0x00000000);
-			evo_data(push, (m->v.active  << 16) | m->h.active );
-			evo_data(push, (m->v.synce   << 16) | m->h.synce  );
-			evo_data(push, (m->v.blanke  << 16) | m->h.blanke );
-			evo_data(push, (m->v.blanks  << 16) | m->h.blanks );
-			evo_data(push, (m->v.blank2e << 16) | m->v.blank2s);
-			evo_mthd(push, 0x042c + (head->base.index * 0x300), 2);
-			evo_data(push, 0x00000000); /* ??? */
-			evo_data(push, 0xffffff00);
-			evo_mthd(push, 0x0450 + (head->base.index * 0x300), 3);
-			evo_data(push, m->clock * 1000);
-			evo_data(push, 0x00200000); /* ??? */
-			evo_data(push, m->clock * 1000);
-		}
+	if ((push = evo_wait(core, 13))) {
+		evo_mthd(push, 0x0804 + (head->base.index * 0x400), 2);
+		evo_data(push, 0x00800000 | m->clock);
+		evo_data(push, m->interlace ? 0x00000002 : 0x00000000);
+		evo_mthd(push, 0x0810 + (head->base.index * 0x400), 7);
+		evo_data(push, 0x00000000);
+		evo_data(push, m->v.active  << 16 | m->h.active );
+		evo_data(push, m->v.synce   << 16 | m->h.synce  );
+		evo_data(push, m->v.blanke  << 16 | m->h.blanke );
+		evo_data(push, m->v.blanks  << 16 | m->h.blanks );
+		evo_data(push, m->v.blank2e << 16 | m->v.blank2s);
+		evo_data(push, asyh->mode.v.blankus);
+		evo_mthd(push, 0x082c + (head->base.index * 0x400), 1);
+		evo_data(push, 0x00000000);
 		evo_kick(push, core);
 	}
 }
 
-static void
+void
 head507d_view(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
 	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
-	if ((push = evo_wait(core, 10))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x08a4 + (head->base.index * 0x400), 1);
-			evo_data(push, 0x00000000);
-			evo_mthd(push, 0x08c8 + (head->base.index * 0x400), 1);
-			evo_data(push, (asyh->view.iH << 16) | asyh->view.iW);
-			evo_mthd(push, 0x08d8 + (head->base.index * 0x400), 2);
-			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
-			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
-		} else {
-			evo_mthd(push, 0x0494 + (head->base.index * 0x300), 1);
-			evo_data(push, 0x00000000);
-			evo_mthd(push, 0x04b8 + (head->base.index * 0x300), 1);
-			evo_data(push, (asyh->view.iH << 16) | asyh->view.iW);
-			evo_mthd(push, 0x04c0 + (head->base.index * 0x300), 3);
-			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
-			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
-			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
-		}
+	if ((push = evo_wait(core, 7))) {
+		evo_mthd(push, 0x08a4 + (head->base.index * 0x400), 1);
+		evo_data(push, 0x00000000);
+		evo_mthd(push, 0x08c8 + (head->base.index * 0x400), 1);
+		evo_data(push, asyh->view.iH << 16 | asyh->view.iW);
+		evo_mthd(push, 0x08d8 + (head->base.index * 0x400), 2);
+		evo_data(push, asyh->view.oH << 16 | asyh->view.oW);
+		evo_data(push, asyh->view.oH << 16 | asyh->view.oW);
 		evo_kick(push, core);
 	}
 }
@@ -391,6 +271,7 @@ head507d = {
 	.mode = head507d_mode,
 	.ilut_set = head507d_ilut_set,
 	.ilut_clr = head507d_ilut_clr,
+	.core_calc = head507d_core_calc,
 	.core_set = head507d_core_set,
 	.core_clr = head507d_core_clr,
 	.curs_set = head507d_curs_set,
@@ -399,5 +280,4 @@ head507d = {
 	.ovly = head507d_ovly,
 	.dither = head507d_dither,
 	.procamp = head507d_procamp,
-	.or = head907d_or,
 };
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head827d.c b/drivers/gpu/drm/nouveau/dispnv50/head827d.c
new file mode 100644
index 0000000..84ce595
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/head827d.c
@@ -0,0 +1,120 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "head.h"
+#include "core.h"
+
+static void
+head827d_curs_clr(struct nv50_head *head)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 4))) {
+		evo_mthd(push, 0x0880 + head->base.index * 0x400, 1);
+		evo_data(push, 0x05000000);
+		evo_mthd(push, 0x089c + head->base.index * 0x400, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, core);
+	}
+}
+
+static void
+head827d_curs_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 5))) {
+		evo_mthd(push, 0x0880 + head->base.index * 0x400, 2);
+		evo_data(push, 0x80000000 | asyh->curs.layout << 26 |
+					    asyh->curs.format << 24);
+		evo_data(push, asyh->curs.offset >> 8);
+		evo_mthd(push, 0x089c + head->base.index * 0x400, 1);
+		evo_data(push, asyh->curs.handle);
+		evo_kick(push, core);
+	}
+}
+
+static void
+head827d_core_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 9))) {
+		evo_mthd(push, 0x0860 + head->base.index * 0x400, 1);
+		evo_data(push, asyh->core.offset >> 8);
+		evo_mthd(push, 0x0868 + head->base.index * 0x400, 4);
+		evo_data(push, asyh->core.h << 16 | asyh->core.w);
+		evo_data(push, asyh->core.layout << 20 |
+			       (asyh->core.pitch >> 8) << 8 |
+			       asyh->core.block);
+		evo_data(push, asyh->core.format << 8);
+		evo_data(push, asyh->core.handle);
+		evo_mthd(push, 0x08c0 + head->base.index * 0x400, 1);
+		evo_data(push, asyh->core.y << 16 | asyh->core.x);
+		evo_kick(push, core);
+	}
+}
+
+static void
+head827d_ilut_clr(struct nv50_head *head)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 4))) {
+		evo_mthd(push, 0x0840 + (head->base.index * 0x400), 1);
+		evo_data(push, 0x40000000);
+		evo_mthd(push, 0x085c + (head->base.index * 0x400), 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, core);
+	}
+}
+
+static void
+head827d_ilut_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 5))) {
+		evo_mthd(push, 0x0840 + (head->base.index * 0x400), 2);
+		evo_data(push, 0x80000000 | asyh->ilut.mode << 30);
+		evo_data(push, asyh->ilut.offset >> 8);
+		evo_mthd(push, 0x085c + (head->base.index * 0x400), 1);
+		evo_data(push, asyh->ilut.handle);
+		evo_kick(push, core);
+	}
+}
+
+const struct nv50_head_func
+head827d = {
+	.view = head507d_view,
+	.mode = head507d_mode,
+	.ilut_set = head827d_ilut_set,
+	.ilut_clr = head827d_ilut_clr,
+	.core_calc = head507d_core_calc,
+	.core_set = head827d_core_set,
+	.core_clr = head507d_core_clr,
+	.curs_set = head827d_curs_set,
+	.curs_clr = head827d_curs_clr,
+	.base = head507d_base,
+	.ovly = head507d_ovly,
+	.dither = head507d_dither,
+	.procamp = head507d_procamp,
+};
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head907d.c b/drivers/gpu/drm/nouveau/dispnv50/head907d.c
new file mode 100644
index 0000000..0035ecc
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/head907d.c
@@ -0,0 +1,274 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "head.h"
+#include "core.h"
+
+void
+head907d_or(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 3))) {
+		evo_mthd(push, 0x0404 + (head->base.index * 0x300), 2);
+		evo_data(push, 0x00000001 | asyh->or.depth  << 6 |
+					    asyh->or.nvsync << 4 |
+					    asyh->or.nhsync << 3);
+		evo_data(push, 0x31ec6000 | head->base.index << 25 |
+					    asyh->mode.interlace);
+		evo_kick(push, core);
+	}
+}
+
+void
+head907d_procamp(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x0498 + (head->base.index * 0x300), 1);
+		evo_data(push, asyh->procamp.sat.sin << 20 |
+			       asyh->procamp.sat.cos << 8);
+		evo_kick(push, core);
+	}
+}
+
+static void
+head907d_dither(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x0490 + (head->base.index * 0x0300), 1);
+		evo_data(push, asyh->dither.mode << 3 |
+			       asyh->dither.bits << 1 |
+			       asyh->dither.enable);
+		evo_kick(push, core);
+	}
+}
+
+void
+head907d_ovly(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 bounds = 0;
+	u32 *push;
+
+	if (asyh->ovly.cpp) {
+		switch (asyh->ovly.cpp) {
+		case 8: bounds |= 0x00000500; break;
+		case 4: bounds |= 0x00000300; break;
+		case 2: bounds |= 0x00000100; break;
+		default:
+			WARN_ON(1);
+			break;
+		}
+		bounds |= 0x00000001;
+	}
+
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x04d4 + head->base.index * 0x300, 1);
+		evo_data(push, bounds);
+		evo_kick(push, core);
+	}
+}
+
+void
+head907d_base(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 bounds = 0;
+	u32 *push;
+
+	if (asyh->base.cpp) {
+		switch (asyh->base.cpp) {
+		case 8: bounds |= 0x00000500; break;
+		case 4: bounds |= 0x00000300; break;
+		case 2: bounds |= 0x00000100; break;
+		case 1: bounds |= 0x00000000; break;
+		default:
+			WARN_ON(1);
+			break;
+		}
+		bounds |= 0x00000001;
+	}
+
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x04d0 + head->base.index * 0x300, 1);
+		evo_data(push, bounds);
+		evo_kick(push, core);
+	}
+}
+
+void
+head907d_curs_clr(struct nv50_head *head)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 4))) {
+		evo_mthd(push, 0x0480 + head->base.index * 0x300, 1);
+		evo_data(push, 0x05000000);
+		evo_mthd(push, 0x048c + head->base.index * 0x300, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, core);
+	}
+}
+
+void
+head907d_curs_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 5))) {
+		evo_mthd(push, 0x0480 + head->base.index * 0x300, 2);
+		evo_data(push, 0x80000000 | asyh->curs.layout << 26 |
+					    asyh->curs.format << 24);
+		evo_data(push, asyh->curs.offset >> 8);
+		evo_mthd(push, 0x048c + head->base.index * 0x300, 1);
+		evo_data(push, asyh->curs.handle);
+		evo_kick(push, core);
+	}
+}
+
+void
+head907d_core_clr(struct nv50_head *head)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x0474 + head->base.index * 0x300, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, core);
+	}
+}
+
+void
+head907d_core_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 9))) {
+		evo_mthd(push, 0x0460 + head->base.index * 0x300, 1);
+		evo_data(push, asyh->core.offset >> 8);
+		evo_mthd(push, 0x0468 + head->base.index * 0x300, 4);
+		evo_data(push, asyh->core.h << 16 | asyh->core.w);
+		evo_data(push, asyh->core.layout << 24 |
+			       (asyh->core.pitch >> 8) << 8 |
+			       asyh->core.block);
+		evo_data(push, asyh->core.format << 8);
+		evo_data(push, asyh->core.handle);
+		evo_mthd(push, 0x04b0 + head->base.index * 0x300, 1);
+		evo_data(push, asyh->core.y << 16 | asyh->core.x);
+		evo_kick(push, core);
+	}
+}
+
+void
+head907d_ilut_clr(struct nv50_head *head)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 4))) {
+		evo_mthd(push, 0x0440 + (head->base.index * 0x300), 1);
+		evo_data(push, 0x03000000);
+		evo_mthd(push, 0x045c + (head->base.index * 0x300), 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, core);
+	}
+}
+
+void
+head907d_ilut_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 7))) {
+		evo_mthd(push, 0x0440 + (head->base.index * 0x300), 4);
+		evo_data(push, 0x80000000 | asyh->ilut.mode << 24);
+		evo_data(push, asyh->ilut.offset >> 8);
+		evo_data(push, 0x00000000);
+		evo_data(push, 0x00000000);
+		evo_mthd(push, 0x045c + (head->base.index * 0x300), 1);
+		evo_data(push, asyh->ilut.handle);
+		evo_kick(push, core);
+	}
+}
+
+void
+head907d_mode(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	struct nv50_head_mode *m = &asyh->mode;
+	u32 *push;
+	if ((push = evo_wait(core, 14))) {
+		evo_mthd(push, 0x0410 + (head->base.index * 0x300), 6);
+		evo_data(push, 0x00000000);
+		evo_data(push, m->v.active  << 16 | m->h.active );
+		evo_data(push, m->v.synce   << 16 | m->h.synce  );
+		evo_data(push, m->v.blanke  << 16 | m->h.blanke );
+		evo_data(push, m->v.blanks  << 16 | m->h.blanks );
+		evo_data(push, m->v.blank2e << 16 | m->v.blank2s);
+		evo_mthd(push, 0x042c + (head->base.index * 0x300), 2);
+		evo_data(push, 0x00000000); /* ??? */
+		evo_data(push, 0xffffff00);
+		evo_mthd(push, 0x0450 + (head->base.index * 0x300), 3);
+		evo_data(push, m->clock * 1000);
+		evo_data(push, 0x00200000); /* ??? */
+		evo_data(push, m->clock * 1000);
+		evo_kick(push, core);
+	}
+}
+
+void
+head907d_view(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 8))) {
+		evo_mthd(push, 0x0494 + (head->base.index * 0x300), 1);
+		evo_data(push, 0x00000000);
+		evo_mthd(push, 0x04b8 + (head->base.index * 0x300), 1);
+		evo_data(push, asyh->view.iH << 16 | asyh->view.iW);
+		evo_mthd(push, 0x04c0 + (head->base.index * 0x300), 3);
+		evo_data(push, asyh->view.oH << 16 | asyh->view.oW);
+		evo_data(push, asyh->view.oH << 16 | asyh->view.oW);
+		evo_data(push, asyh->view.oH << 16 | asyh->view.oW);
+		evo_kick(push, core);
+	}
+}
+
+const struct nv50_head_func
+head907d = {
+	.view = head907d_view,
+	.mode = head907d_mode,
+	.ilut_set = head907d_ilut_set,
+	.ilut_clr = head907d_ilut_clr,
+	.core_calc = head507d_core_calc,
+	.core_set = head907d_core_set,
+	.core_clr = head907d_core_clr,
+	.curs_set = head907d_curs_set,
+	.curs_clr = head907d_curs_clr,
+	.base = head907d_base,
+	.ovly = head907d_ovly,
+	.dither = head907d_dither,
+	.procamp = head907d_procamp,
+	.or = head907d_or,
+};
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head917d.c b/drivers/gpu/drm/nouveau/dispnv50/head917d.c
new file mode 100644
index 0000000..5341ea3
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/head917d.c
@@ -0,0 +1,55 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "head.h"
+#include "core.h"
+
+static void
+head917d_dither(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x04a0 + (head->base.index * 0x0300), 1);
+		evo_data(push, asyh->dither.mode << 3 |
+			       asyh->dither.bits << 1 |
+			       asyh->dither.enable);
+		evo_kick(push, core);
+	}
+}
+
+const struct nv50_head_func
+head917d = {
+	.view = head907d_view,
+	.mode = head907d_mode,
+	.ilut_set = head907d_ilut_set,
+	.ilut_clr = head907d_ilut_clr,
+	.core_calc = head507d_core_calc,
+	.core_set = head907d_core_set,
+	.core_clr = head907d_core_clr,
+	.curs_set = head907d_curs_set,
+	.curs_clr = head907d_curs_clr,
+	.base = head907d_base,
+	.ovly = head907d_ovly,
+	.dither = head917d_dither,
+	.procamp = head907d_procamp,
+	.or = head907d_or,
+};
diff --git a/drivers/gpu/drm/nouveau/dispnv50/pior507d.c b/drivers/gpu/drm/nouveau/dispnv50/pior507d.c
index a99ba6a..d2bac6a3 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/pior507d.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/pior507d.c
@@ -21,23 +21,19 @@
  */
 #include "core.h"
 
-#include <nvif/class.h>
-
 static void
 pior507d_ctrl(struct nv50_core *core, int or, u32 ctrl,
 	      struct nv50_head_atom *asyh)
 {
 	u32 *push;
-	if ((push = evo_wait(&core->chan, 8))) {
-		if (core->chan.base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			if (asyh) {
-				ctrl |= asyh->or.depth  << 16;
-				ctrl |= asyh->or.nvsync << 13;
-				ctrl |= asyh->or.nhsync << 12;
-			}
-			evo_mthd(push, 0x0700 + (or * 0x040), 1);
-			evo_data(push, ctrl);
+	if ((push = evo_wait(&core->chan, 2))) {
+		if (asyh) {
+			ctrl |= asyh->or.depth  << 16;
+			ctrl |= asyh->or.nvsync << 13;
+			ctrl |= asyh->or.nhsync << 12;
 		}
+		evo_mthd(push, 0x0700 + (or * 0x040), 1);
+		evo_data(push, ctrl);
 		evo_kick(push, &core->chan);
 	}
 }
diff --git a/drivers/gpu/drm/nouveau/dispnv50/sor507d.c b/drivers/gpu/drm/nouveau/dispnv50/sor507d.c
index 2d540de..5222fe6 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/sor507d.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/sor507d.c
@@ -21,24 +21,18 @@
  */
 #include "core.h"
 
-#include <nvif/class.h>
-
 static void
 sor507d_ctrl(struct nv50_core *core, int or, u32 ctrl,
 	     struct nv50_head_atom *asyh)
 {
 	u32 *push;
-	if ((push = evo_wait(&core->chan, 6))) {
-		if (core->chan.base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			if (asyh) {
-				ctrl |= asyh->or.depth  << 16;
-				ctrl |= asyh->or.nvsync << 13;
-				ctrl |= asyh->or.nhsync << 12;
-			}
-			evo_mthd(push, 0x0600 + (or * 0x40), 1);
-		} else {
-			evo_mthd(push, 0x0200 + (or * 0x20), 1);
+	if ((push = evo_wait(&core->chan, 2))) {
+		if (asyh) {
+			ctrl |= asyh->or.depth  << 16;
+			ctrl |= asyh->or.nvsync << 13;
+			ctrl |= asyh->or.nhsync << 12;
 		}
+		evo_mthd(push, 0x0600 + (or * 0x40), 1);
 		evo_data(push, ctrl);
 		evo_kick(push, &core->chan);
 	}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/sor907d.c b/drivers/gpu/drm/nouveau/dispnv50/sor907d.c
new file mode 100644
index 0000000..b0314ec1
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/sor907d.c
@@ -0,0 +1,41 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "core.h"
+
+#include <nvif/class.h>
+
+static void
+sor907d_ctrl(struct nv50_core *core, int or, u32 ctrl,
+	     struct nv50_head_atom *asyh)
+{
+	u32 *push;
+	if ((push = evo_wait(&core->chan, 2))) {
+		evo_mthd(push, 0x0200 + (or * 0x20), 1);
+		evo_data(push, ctrl);
+		evo_kick(push, &core->chan);
+	}
+}
+
+const struct nv50_outp_func
+sor907d = {
+	.ctrl = sor907d_ctrl,
+};
-- 
cgit v1.1


From ccd27db8c731817ef36e75de2b5fdc2e79550213 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50-: split base implementation by hardware class

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/Kbuild     |   2 +
 drivers/gpu/drm/nouveau/dispnv50/base.c     |  12 ++--
 drivers/gpu/drm/nouveau/dispnv50/base.h     |  21 ++++++
 drivers/gpu/drm/nouveau/dispnv50/base507c.c | 101 +++++++++++-----------------
 drivers/gpu/drm/nouveau/dispnv50/base827c.c |  67 ++++++++++++++++++
 drivers/gpu/drm/nouveau/dispnv50/base907c.c |  80 ++++++++++++++++++++++
 drivers/gpu/drm/nouveau/dispnv50/disp.c     |  11 +--
 drivers/gpu/drm/nouveau/dispnv50/wndw.c     |  22 +++++-
 drivers/gpu/drm/nouveau/dispnv50/wndw.h     |   5 +-
 9 files changed, 242 insertions(+), 79 deletions(-)
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/base827c.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/base907c.c

diff --git a/drivers/gpu/drm/nouveau/dispnv50/Kbuild b/drivers/gpu/drm/nouveau/dispnv50/Kbuild
index cde3ae9..674221d 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/Kbuild
+++ b/drivers/gpu/drm/nouveau/dispnv50/Kbuild
@@ -24,6 +24,8 @@ nouveau-y += dispnv50/wndw.o
 
 nouveau-y += dispnv50/base.o
 nouveau-y += dispnv50/base507c.o
+nouveau-y += dispnv50/base827c.o
+nouveau-y += dispnv50/base907c.o
 
 nouveau-y += dispnv50/curs.o
 nouveau-y += dispnv50/curs507a.o
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base.c b/drivers/gpu/drm/nouveau/dispnv50/base.c
index 12ca5d7..5f184ab 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/base.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/base.c
@@ -31,12 +31,12 @@ nv50_base_new(struct nouveau_drm *drm, int head, struct nv50_wndw **pwndw)
 		int version;
 		int (*new)(struct nouveau_drm *, int, s32, struct nv50_wndw **);
 	} bases[] = {
-		{ GK110_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
-		{ GK104_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
-		{ GF110_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
-		{ GT214_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
-		{ GT200_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
-		{   G82_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
+		{ GK110_DISP_BASE_CHANNEL_DMA, 0, base907c_new },
+		{ GK104_DISP_BASE_CHANNEL_DMA, 0, base907c_new },
+		{ GF110_DISP_BASE_CHANNEL_DMA, 0, base907c_new },
+		{ GT214_DISP_BASE_CHANNEL_DMA, 0, base827c_new },
+		{ GT200_DISP_BASE_CHANNEL_DMA, 0, base827c_new },
+		{   G82_DISP_BASE_CHANNEL_DMA, 0, base827c_new },
 		{  NV50_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
 		{}
 	};
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base.h b/drivers/gpu/drm/nouveau/dispnv50/base.h
index 1daba73..edf96a8 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/base.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/base.h
@@ -3,6 +3,27 @@
 #include "wndw.h"
 
 int base507c_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+int base507c_new_(const struct nv50_wndw_func *, const u32 *format,
+		  struct nouveau_drm *, int head, s32 oclass,
+		  struct nv50_wndw **);
+extern const u32 base507c_format[];
+int base507c_acquire(struct nv50_wndw *, struct nv50_wndw_atom *,
+		     struct nv50_head_atom *);
+void base507c_release(struct nv50_wndw *, struct nv50_wndw_atom *,
+		      struct nv50_head_atom *);
+void base507c_sema_set(struct nv50_wndw *, struct nv50_wndw_atom *);
+void base507c_sema_clr(struct nv50_wndw *);
+void base507c_ntfy_reset(struct nouveau_bo *, u32);
+void base507c_ntfy_set(struct nv50_wndw *, struct nv50_wndw_atom *);
+void base507c_ntfy_clr(struct nv50_wndw *);
+int base507c_ntfy_wait_begun(struct nouveau_bo *, u32, struct nvif_device *);
+void base507c_image_clr(struct nv50_wndw *);
+void base507c_lut(struct nv50_wndw *, struct nv50_wndw_atom *);
+u32 base507c_update(struct nv50_wndw *, u32);
+
+int base827c_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+
+int base907c_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
 
 int nv50_base_new(struct nouveau_drm *, int head, struct nv50_wndw **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base507c.c b/drivers/gpu/drm/nouveau/dispnv50/base507c.c
index b73e7b4..850c8de 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/base507c.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/base507c.c
@@ -21,7 +21,6 @@
  */
 #include "base.h"
 
-#include <nvif/class.h>
 #include <nvif/cl507c.h>
 #include <nvif/event.h>
 
@@ -29,23 +28,20 @@
 #include <drm/drm_plane_helper.h>
 #include "nouveau_bo.h"
 
-static u32
+u32
 base507c_update(struct nv50_wndw *wndw, u32 interlock)
 {
 	u32 *push;
-
-	if (!(push = evo_wait(&wndw->wndw, 2)))
-		return 0;
-	evo_mthd(push, 0x0080, 1);
-	evo_data(push, interlock);
-	evo_kick(push, &wndw->wndw);
-
-	if (wndw->wndw.base.user.oclass < GF110_DISP_BASE_CHANNEL_DMA)
+	if ((push = evo_wait(&wndw->wndw, 2))) {
+		evo_mthd(push, 0x0080, 1);
+		evo_data(push, interlock);
+		evo_kick(push, &wndw->wndw);
 		return interlock ? 2 << (wndw->id * 8) : 0;
-	return interlock ? 2 << (wndw->id * 4) : 0;
+	}
+	return 0;
 }
 
-static void
+void
 base507c_lut(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 {
 	u32 *push;
@@ -56,7 +52,7 @@ base507c_lut(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 	}
 }
 
-static void
+void
 base507c_image_clr(struct nv50_wndw *wndw)
 {
 	u32 *push;
@@ -72,7 +68,6 @@ base507c_image_clr(struct nv50_wndw *wndw)
 static void
 base507c_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 {
-	const s32 oclass = wndw->wndw.base.user.oclass;
 	u32 *push;
 	if ((push = evo_wait(&wndw->wndw, 10))) {
 		evo_mthd(push, 0x0084, 1);
@@ -80,56 +75,33 @@ base507c_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 			       asyw->image.interval << 4);
 		evo_mthd(push, 0x00c0, 1);
 		evo_data(push, asyw->image.handle);
-		if (oclass < G82_DISP_BASE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0800, 5);
-			evo_data(push, asyw->image.offset >> 8);
-			evo_data(push, 0x00000000);
-			evo_data(push, (asyw->image.h << 16) | asyw->image.w);
-			evo_data(push, (asyw->image.layout << 20) |
-					asyw->image.pitch |
-					asyw->image.block);
-			evo_data(push, (asyw->image.kind << 16) |
-				       (asyw->image.format << 8));
-		} else
-		if (oclass < GF110_DISP_BASE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0800, 5);
-			evo_data(push, asyw->image.offset >> 8);
-			evo_data(push, 0x00000000);
-			evo_data(push, (asyw->image.h << 16) | asyw->image.w);
-			evo_data(push, (asyw->image.layout << 20) |
-					asyw->image.pitch |
-					asyw->image.block);
-			evo_data(push, asyw->image.format << 8);
-		} else {
-			evo_mthd(push, 0x0400, 5);
-			evo_data(push, asyw->image.offset >> 8);
-			evo_data(push, 0x00000000);
-			evo_data(push, (asyw->image.h << 16) | asyw->image.w);
-			evo_data(push, (asyw->image.layout << 24) |
-					asyw->image.pitch |
-					asyw->image.block);
-			evo_data(push, asyw->image.format << 8);
-		}
+		evo_mthd(push, 0x0800, 5);
+		evo_data(push, asyw->image.offset >> 8);
+		evo_data(push, 0x00000000);
+		evo_data(push, asyw->image.h << 16 | asyw->image.w);
+		evo_data(push, asyw->image.layout << 20 |
+			       asyw->image.pitch |
+			       asyw->image.block);
+		evo_data(push, asyw->image.kind << 16 |
+			       asyw->image.format << 8);
 		evo_kick(push, &wndw->wndw);
 	}
 }
 
-static int
-base507c_ntfy_wait_begun(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+int
+base507c_ntfy_wait_begun(struct nouveau_bo *bo, u32 offset,
+			 struct nvif_device *device)
 {
-	struct nouveau_drm *drm = nouveau_drm(wndw->plane.dev);
-	struct nv50_disp *disp = nv50_disp(wndw->plane.dev);
-	if (nvif_msec(&drm->client.device, 2000ULL,
-		u32 data = nouveau_bo_rd32(disp->sync, asyw->ntfy.offset / 4);
+	s64 time = nvif_msec(device, 2000ULL,
+		u32 data = nouveau_bo_rd32(bo, offset / 4);
 		if ((data & 0xc0000000) == 0x40000000)
 			break;
 		usleep_range(1, 2);
-	) < 0)
-		return -ETIMEDOUT;
-	return 0;
+	);
+	return time < 0 ? time : 0;
 }
 
-static void
+void
 base507c_ntfy_clr(struct nv50_wndw *wndw)
 {
 	u32 *push;
@@ -140,7 +112,7 @@ base507c_ntfy_clr(struct nv50_wndw *wndw)
 	}
 }
 
-static void
+void
 base507c_ntfy_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 {
 	u32 *push;
@@ -152,7 +124,13 @@ base507c_ntfy_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 	}
 }
 
-static void
+void
+base507c_ntfy_reset(struct nouveau_bo *bo, u32 offset)
+{
+	nouveau_bo_wr32(bo, offset / 4, 0x00000000);
+}
+
+void
 base507c_sema_clr(struct nv50_wndw *wndw)
 {
 	u32 *push;
@@ -163,7 +141,7 @@ base507c_sema_clr(struct nv50_wndw *wndw)
 	}
 }
 
-static void
+void
 base507c_sema_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 {
 	u32 *push;
@@ -177,14 +155,14 @@ base507c_sema_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 	}
 }
 
-static void
+void
 base507c_release(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
 		 struct nv50_head_atom *asyh)
 {
 	asyh->base.cpp = 0;
 }
 
-static int
+int
 base507c_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
 		 struct nv50_head_atom *asyh)
 {
@@ -229,7 +207,7 @@ base507c_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
 	return 0;
 }
 
-static const u32
+const u32
 base507c_format[] = {
 	DRM_FORMAT_C8,
 	DRM_FORMAT_RGB565,
@@ -250,6 +228,7 @@ base507c = {
 	.release = base507c_release,
 	.sema_set = base507c_sema_set,
 	.sema_clr = base507c_sema_clr,
+	.ntfy_reset = base507c_ntfy_reset,
 	.ntfy_set = base507c_ntfy_set,
 	.ntfy_clr = base507c_ntfy_clr,
 	.ntfy_wait_begun = base507c_ntfy_wait_begun,
@@ -259,7 +238,7 @@ base507c = {
 	.update = base507c_update,
 };
 
-static int
+int
 base507c_new_(const struct nv50_wndw_func *func, const u32 *format,
 	      struct nouveau_drm *drm, int head, s32 oclass,
 	      struct nv50_wndw **pwndw)
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base827c.c b/drivers/gpu/drm/nouveau/dispnv50/base827c.c
new file mode 100644
index 0000000..0d356ae
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/base827c.c
@@ -0,0 +1,67 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "base.h"
+
+static void
+base827c_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 10))) {
+		evo_mthd(push, 0x0084, 1);
+		evo_data(push, asyw->image.mode << 8 |
+			       asyw->image.interval << 4);
+		evo_mthd(push, 0x00c0, 1);
+		evo_data(push, asyw->image.handle);
+		evo_mthd(push, 0x0800, 5);
+		evo_data(push, asyw->image.offset >> 8);
+		evo_data(push, 0x00000000);
+		evo_data(push, asyw->image.h << 16 | asyw->image.w);
+		evo_data(push, asyw->image.layout << 20 |
+			       asyw->image.pitch |
+			       asyw->image.block);
+		evo_data(push, asyw->image.format << 8);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static const struct nv50_wndw_func
+base827c = {
+	.acquire = base507c_acquire,
+	.release = base507c_release,
+	.sema_set = base507c_sema_set,
+	.sema_clr = base507c_sema_clr,
+	.ntfy_reset = base507c_ntfy_reset,
+	.ntfy_set = base507c_ntfy_set,
+	.ntfy_clr = base507c_ntfy_clr,
+	.ntfy_wait_begun = base507c_ntfy_wait_begun,
+	.image_set = base827c_image_set,
+	.image_clr = base507c_image_clr,
+	.lut = base507c_lut,
+	.update = base507c_update,
+};
+
+int
+base827c_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return base507c_new_(&base827c, base507c_format, drm, head, oclass, pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base907c.c b/drivers/gpu/drm/nouveau/dispnv50/base907c.c
new file mode 100644
index 0000000..171d978
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/base907c.c
@@ -0,0 +1,80 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "base.h"
+
+static u32
+base907c_update(struct nv50_wndw *wndw, u32 interlock)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 2))) {
+		evo_mthd(push, 0x0080, 1);
+		evo_data(push, interlock);
+		evo_kick(push, &wndw->wndw);
+		return interlock ? 2 << (wndw->id * 4) : 0;
+	}
+	return 0;
+}
+
+static void
+base907c_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 10))) {
+		evo_mthd(push, 0x0084, 1);
+		evo_data(push, asyw->image.mode << 8 |
+			       asyw->image.interval << 4);
+		evo_mthd(push, 0x00c0, 1);
+		evo_data(push, asyw->image.handle);
+		evo_mthd(push, 0x0400, 5);
+		evo_data(push, asyw->image.offset >> 8);
+		evo_data(push, 0x00000000);
+		evo_data(push, asyw->image.h << 16 | asyw->image.w);
+		evo_data(push, asyw->image.layout << 24 |
+			       asyw->image.pitch |
+			       asyw->image.block);
+		evo_data(push, asyw->image.format << 8);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static const struct nv50_wndw_func
+base907c = {
+	.acquire = base507c_acquire,
+	.release = base507c_release,
+	.sema_set = base507c_sema_set,
+	.sema_clr = base507c_sema_clr,
+	.ntfy_reset = base507c_ntfy_reset,
+	.ntfy_set = base507c_ntfy_set,
+	.ntfy_clr = base507c_ntfy_clr,
+	.ntfy_wait_begun = base507c_ntfy_wait_begun,
+	.image_set = base907c_image_set,
+	.image_clr = base507c_image_clr,
+	.lut = base507c_lut,
+	.update = base907c_update,
+};
+
+int
+base907c_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return base507c_new_(&base907c, base507c_format, drm, head, oclass, pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c
index 6136bee..c2b1578 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/disp.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c
@@ -1812,7 +1812,6 @@ nv50_disp_atomic_commit(struct drm_device *dev,
 			struct drm_atomic_state *state, bool nonblock)
 {
 	struct nouveau_drm *drm = nouveau_drm(dev);
-	struct nv50_disp *disp = nv50_disp(dev);
 	struct drm_plane_state *new_plane_state;
 	struct drm_plane *plane;
 	struct drm_crtc *crtc;
@@ -1847,14 +1846,8 @@ nv50_disp_atomic_commit(struct drm_device *dev,
 		struct nv50_wndw_atom *asyw = nv50_wndw_atom(new_plane_state);
 		struct nv50_wndw *wndw = nv50_wndw(plane);
 
-		if (asyw->set.image) {
-			asyw->ntfy.handle = wndw->wndw.sync.handle;
-			asyw->ntfy.offset = wndw->ntfy;
-			asyw->ntfy.awaken = false;
-			asyw->set.ntfy = true;
-			nouveau_bo_wr32(disp->sync, wndw->ntfy / 4, 0x00000000);
-			wndw->ntfy ^= 0x10;
-		}
+		if (asyw->set.image)
+			nv50_wndw_ntfy_enable(wndw, asyw);
 	}
 
 	drm_atomic_state_get(state);
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.c b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
index 71a4c50..a1e53c7 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/wndw.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
@@ -98,8 +98,12 @@ nv50_wndw_ctxdma_new(struct nv50_wndw *wndw, struct nouveau_framebuffer *fb)
 int
 nv50_wndw_wait_armed(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 {
-	if (asyw->set.ntfy)
-		return wndw->func->ntfy_wait_begun(wndw, asyw);
+	struct nv50_disp *disp = nv50_disp(wndw->plane.dev);
+	if (asyw->set.ntfy) {
+		return wndw->func->ntfy_wait_begun(disp->sync,
+						   asyw->ntfy.offset,
+						   wndw->wndw.base.device);
+	}
 	return 0;
 }
 
@@ -138,6 +142,20 @@ nv50_wndw_flush_set(struct nv50_wndw *wndw, u32 interlock,
 	return wndw->func->update ? wndw->func->update(wndw, interlock) : 0;
 }
 
+void
+nv50_wndw_ntfy_enable(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	struct nv50_disp *disp = nv50_disp(wndw->plane.dev);
+
+	asyw->ntfy.handle = wndw->wndw.sync.handle;
+	asyw->ntfy.offset = wndw->ntfy;
+	asyw->ntfy.awaken = false;
+	asyw->set.ntfy = true;
+
+	wndw->func->ntfy_reset(disp->sync, wndw->ntfy);
+	wndw->ntfy ^= 0x10;
+}
+
 static void
 nv50_wndw_atomic_check_release(struct nv50_wndw *wndw,
 			       struct nv50_wndw_atom *asyw,
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.h b/drivers/gpu/drm/nouveau/dispnv50/wndw.h
index 1931e30..7025973 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/wndw.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.h
@@ -41,6 +41,7 @@ u32 nv50_wndw_flush_set(struct nv50_wndw *, u32 interlock,
 			struct nv50_wndw_atom *);
 u32 nv50_wndw_flush_clr(struct nv50_wndw *, u32 interlock, bool flush,
 			struct nv50_wndw_atom *);
+void nv50_wndw_ntfy_enable(struct nv50_wndw *, struct nv50_wndw_atom *);
 int nv50_wndw_wait_armed(struct nv50_wndw *, struct nv50_wndw_atom *);
 
 struct nv50_wndw_func {
@@ -53,9 +54,11 @@ struct nv50_wndw_func {
 
 	void (*sema_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
 	void (*sema_clr)(struct nv50_wndw *);
+	void (*ntfy_reset)(struct nouveau_bo *, u32 offset);
 	void (*ntfy_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
 	void (*ntfy_clr)(struct nv50_wndw *);
-	int (*ntfy_wait_begun)(struct nv50_wndw *, struct nv50_wndw_atom *);
+	int (*ntfy_wait_begun)(struct nouveau_bo *, u32 offset,
+			       struct nvif_device *);
 	void (*image_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
 	void (*image_clr)(struct nv50_wndw *);
 	void (*lut)(struct nv50_wndw *, struct nv50_wndw_atom *);
-- 
cgit v1.1


From 9d6c2fe1917fc5ba6a9e8586ca16d007410baf42 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50-: allow specification of valid heads for a
 window

This will be required to support Volta, where window ID != head.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/base507c.c | 2 +-
 drivers/gpu/drm/nouveau/dispnv50/curs507a.c | 2 +-
 drivers/gpu/drm/nouveau/dispnv50/ovly507e.c | 2 +-
 drivers/gpu/drm/nouveau/dispnv50/wndw.c     | 4 ++--
 drivers/gpu/drm/nouveau/dispnv50/wndw.h     | 2 +-
 5 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/dispnv50/base507c.c b/drivers/gpu/drm/nouveau/dispnv50/base507c.c
index 850c8de..548a6a6 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/base507c.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/base507c.c
@@ -251,7 +251,7 @@ base507c_new_(const struct nv50_wndw_func *func, const u32 *format,
 	int ret;
 
 	ret = nv50_wndw_new_(func, drm->dev, DRM_PLANE_TYPE_PRIMARY,
-			     "base", head, format, &wndw);
+			     "base", head, format, BIT(head), &wndw);
 	if (*pwndw = wndw, ret)
 		return ret;
 
diff --git a/drivers/gpu/drm/nouveau/dispnv50/curs507a.c b/drivers/gpu/drm/nouveau/dispnv50/curs507a.c
index 1a3e199..3e5e98c 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/curs507a.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/curs507a.c
@@ -126,7 +126,7 @@ curs507a_new_(const struct nv50_wimm_func *func, struct nouveau_drm *drm,
 	int ret;
 
 	ret = nv50_wndw_new_(&curs507a_wndw, drm->dev, DRM_PLANE_TYPE_CURSOR,
-			     "curs", head, curs507a_format, &wndw);
+			     "curs", head, curs507a_format, BIT(head), &wndw);
 	if (*pwndw = wndw, ret)
 		return ret;
 
diff --git a/drivers/gpu/drm/nouveau/dispnv50/ovly507e.c b/drivers/gpu/drm/nouveau/dispnv50/ovly507e.c
index ceec512..1b85262 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/ovly507e.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/ovly507e.c
@@ -47,7 +47,7 @@ ovly507e_new_(const struct nv50_wndw_func *func, const u32 *format,
 	int ret;
 
 	ret = nv50_wndw_new_(func, drm->dev, DRM_PLANE_TYPE_OVERLAY,
-			     "ovly", head, format, &wndw);
+			     "ovly", head, format, BIT(head), &wndw);
 	if (*pwndw = wndw, ret)
 		return ret;
 
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.c b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
index a1e53c7..764db73 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/wndw.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
@@ -420,7 +420,7 @@ nv50_wndw_init(struct nv50_wndw *wndw)
 int
 nv50_wndw_new_(const struct nv50_wndw_func *func, struct drm_device *dev,
 	       enum drm_plane_type type, const char *name, int index,
-	       const u32 *format, struct nv50_wndw **pwndw)
+	       const u32 *format, u32 heads, struct nv50_wndw **pwndw)
 {
 	struct nv50_wndw *wndw;
 	int nformat;
@@ -436,7 +436,7 @@ nv50_wndw_new_(const struct nv50_wndw_func *func, struct drm_device *dev,
 
 	for (nformat = 0; format[nformat]; nformat++);
 
-	ret = drm_universal_plane_init(dev, &wndw->plane, 0, &nv50_wndw,
+	ret = drm_universal_plane_init(dev, &wndw->plane, heads, &nv50_wndw,
 				       format, nformat, NULL,
 				       type, "%s-%d", name, index);
 	if (ret) {
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.h b/drivers/gpu/drm/nouveau/dispnv50/wndw.h
index 7025973..8672c28 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/wndw.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.h
@@ -34,7 +34,7 @@ struct nv50_wndw {
 
 int nv50_wndw_new_(const struct nv50_wndw_func *, struct drm_device *,
 		   enum drm_plane_type, const char *name, int index,
-		   const u32 *format, struct nv50_wndw **);
+		   const u32 *format, u32 heads, struct nv50_wndw **);
 void nv50_wndw_init(struct nv50_wndw *);
 void nv50_wndw_fini(struct nv50_wndw *);
 u32 nv50_wndw_flush_set(struct nv50_wndw *, u32 interlock,
-- 
cgit v1.1


From f88bc9d3ecca5ddc29642269f4624d07265c1bf5 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50-: unify set/clr masks

This is a simplification that'll be used to improve interlock handling.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/atom.h | 26 ++++----------------------
 drivers/gpu/drm/nouveau/dispnv50/disp.c | 11 ++---------
 drivers/gpu/drm/nouveau/dispnv50/head.c | 15 ++++++++-------
 drivers/gpu/drm/nouveau/dispnv50/wndw.c | 12 ++++++------
 4 files changed, 20 insertions(+), 44 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/dispnv50/atom.h b/drivers/gpu/drm/nouveau/dispnv50/atom.h
index 8c97e25..8bf1806 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/atom.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/atom.h
@@ -105,16 +105,7 @@ struct nv50_head_atom {
 		u8 depth:4;
 	} or;
 
-	union {
-		struct {
-			bool ilut:1;
-			bool core:1;
-			bool curs:1;
-		};
-		u8 mask;
-	} clr;
-
-	union {
+	union nv50_head_atom_mask {
 		struct {
 			bool ilut:1;
 			bool core:1;
@@ -128,7 +119,7 @@ struct nv50_head_atom {
 			bool or:1;
 		};
 		u16 mask;
-	} set;
+	} set, clr;
 };
 
 static inline struct nv50_head_atom *
@@ -184,16 +175,7 @@ struct nv50_wndw_atom {
 		u16 y;
 	} point;
 
-	union {
-		struct {
-			bool ntfy:1;
-			bool sema:1;
-			bool image:1;
-		};
-		u8 mask;
-	} clr;
-
-	union {
+	union nv50_wndw_atom_mask {
 		struct {
 			bool ntfy:1;
 			bool sema:1;
@@ -202,6 +184,6 @@ struct nv50_wndw_atom {
 			bool point:1;
 		};
 		u8 mask;
-	} set;
+	} set, clr;
 };
 #endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c
index c2b1578..006562f 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/disp.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c
@@ -62,19 +62,12 @@ struct nv50_outp_atom {
 	struct drm_encoder *encoder;
 	bool flush_disable;
 
-	union {
+	union nv50_outp_atom_mask {
 		struct {
 			bool ctrl:1;
 		};
 		u8 mask;
-	} clr;
-
-	union {
-		struct {
-			bool ctrl:1;
-		};
-		u8 mask;
-	} set;
+	} set, clr;
 };
 
 /******************************************************************************
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head.c b/drivers/gpu/drm/nouveau/dispnv50/head.c
index 1335c00..2eb7fdb 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/head.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/head.c
@@ -81,14 +81,15 @@ nv50_head_lut_load(struct drm_property_blob *blob, int mode,
 }
 
 void
-nv50_head_flush_clr(struct nv50_head *head, struct nv50_head_atom *asyh, bool y)
+nv50_head_flush_clr(struct nv50_head *head,
+		    struct nv50_head_atom *asyh, bool flush)
 {
-	if (asyh->clr.ilut && (!asyh->set.ilut || y))
-		head->func->ilut_clr(head);
-	if (asyh->clr.core && (!asyh->set.core || y))
-		head->func->core_clr(head);
-	if (asyh->clr.curs && (!asyh->set.curs || y))
-		head->func->curs_clr(head);
+	union nv50_head_atom_mask clr = {
+		.mask = asyh->clr.mask & ~(flush ? 0 : asyh->set.mask),
+	};
+	if (clr.ilut) head->func->ilut_clr(head);
+	if (clr.core) head->func->core_clr(head);
+	if (clr.curs) head->func->curs_clr(head);
 }
 
 void
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.c b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
index 764db73..8a7636f 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/wndw.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
@@ -111,12 +111,12 @@ u32
 nv50_wndw_flush_clr(struct nv50_wndw *wndw, u32 interlock, bool flush,
 		    struct nv50_wndw_atom *asyw)
 {
-	if (asyw->clr.sema && (!asyw->set.sema || flush))
-		wndw->func->sema_clr(wndw);
-	if (asyw->clr.ntfy && (!asyw->set.ntfy || flush))
-		wndw->func->ntfy_clr(wndw);
-	if (asyw->clr.image && (!asyw->set.image || flush))
-		wndw->func->image_clr(wndw);
+	union nv50_wndw_atom_mask clr = {
+		.mask = asyw->clr.mask & ~(flush ? 0 : asyw->set.mask),
+	};
+	if (clr.sema ) wndw->func-> sema_clr(wndw);
+	if (clr.ntfy ) wndw->func-> ntfy_clr(wndw);
+	if (clr.image) wndw->func->image_clr(wndw);
 
 	return flush ? wndw->func->update(wndw, interlock) : 0;
 }
-- 
cgit v1.1


From 43c181e9deb5f4215d4ef0cb227fde509da7cc5e Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50-: move drm format->hw conversion into common
 code

This will be required to support additional HW features.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/base507c.c | 16 ----------------
 drivers/gpu/drm/nouveau/dispnv50/curs507a.c |  4 ++--
 drivers/gpu/drm/nouveau/dispnv50/wndw.c     | 25 +++++++++++++++++++++++++
 3 files changed, 27 insertions(+), 18 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/dispnv50/base507c.c b/drivers/gpu/drm/nouveau/dispnv50/base507c.c
index 548a6a6..43dcbcd 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/base507c.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/base507c.c
@@ -186,22 +186,6 @@ base507c_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
 	asyh->base.w = asyw->state.fb->width;
 	asyh->base.h = asyw->state.fb->height;
 
-	switch (fb->format->format) {
-	case DRM_FORMAT_C8         : asyw->image.format = 0x1e; break;
-	case DRM_FORMAT_RGB565     : asyw->image.format = 0xe8; break;
-	case DRM_FORMAT_XRGB1555   :
-	case DRM_FORMAT_ARGB1555   : asyw->image.format = 0xe9; break;
-	case DRM_FORMAT_XRGB8888   :
-	case DRM_FORMAT_ARGB8888   : asyw->image.format = 0xcf; break;
-	case DRM_FORMAT_XBGR2101010:
-	case DRM_FORMAT_ABGR2101010: asyw->image.format = 0xd1; break;
-	case DRM_FORMAT_XBGR8888   :
-	case DRM_FORMAT_ABGR8888   : asyw->image.format = 0xd5; break;
-	default:
-		WARN_ON(1);
-		return -EINVAL;
-	}
-
 	asyw->lut.enable = 1;
 	asyw->set.image = true;
 	return 0;
diff --git a/drivers/gpu/drm/nouveau/dispnv50/curs507a.c b/drivers/gpu/drm/nouveau/dispnv50/curs507a.c
index 3e5e98c..f7e56a8 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/curs507a.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/curs507a.c
@@ -91,8 +91,8 @@ curs507a_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
 	if (asyw->state.fb->width != asyw->state.fb->height)
 		return -EINVAL;
 
-	switch (asyw->state.fb->format->format) {
-	case DRM_FORMAT_ARGB8888: asyh->curs.format = 1; break;
+	switch (asyw->image.format) {
+	case 0xcf: asyh->curs.format = 1; break;
 	default:
 		WARN_ON(1);
 		return -EINVAL;
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.c b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
index 8a7636f..cfd998a 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/wndw.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
@@ -169,6 +169,27 @@ nv50_wndw_atomic_check_release(struct nv50_wndw *wndw,
 }
 
 static int
+nv50_wndw_atomic_check_acquire_rgb(struct nv50_wndw_atom *asyw)
+{
+	switch (asyw->state.fb->format->format) {
+	case DRM_FORMAT_C8         : asyw->image.format = 0x1e; break;
+	case DRM_FORMAT_XRGB8888   :
+	case DRM_FORMAT_ARGB8888   : asyw->image.format = 0xcf; break;
+	case DRM_FORMAT_RGB565     : asyw->image.format = 0xe8; break;
+	case DRM_FORMAT_XRGB1555   :
+	case DRM_FORMAT_ARGB1555   : asyw->image.format = 0xe9; break;
+	case DRM_FORMAT_XBGR2101010:
+	case DRM_FORMAT_ABGR2101010: asyw->image.format = 0xd1; break;
+	case DRM_FORMAT_XBGR8888   :
+	case DRM_FORMAT_ABGR8888   : asyw->image.format = 0xd5; break;
+	default:
+		WARN_ON(1);
+		return -EINVAL;
+	}
+	return 0;
+}
+
+static int
 nv50_wndw_atomic_check_acquire(struct nv50_wndw *wndw,
 			       struct nv50_wndw_atom *asyw,
 			       struct nv50_head_atom *asyh)
@@ -183,6 +204,10 @@ nv50_wndw_atomic_check_acquire(struct nv50_wndw *wndw,
 	asyw->image.h = fb->base.height;
 	asyw->image.kind = fb->nvbo->kind;
 
+	ret = nv50_wndw_atomic_check_acquire_rgb(asyw);
+	if (ret)
+		return ret;
+
 	if (asyh->state.pageflip_flags & DRM_MODE_PAGE_FLIP_ASYNC)
 		asyw->interval = 0;
 	else
-- 
cgit v1.1


From 261fcfa96991d6652b061262c1879cc0bdd1aa3a Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50-: extend window image data for stereo/planar
 formats

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/atom.h     |  6 +++---
 drivers/gpu/drm/nouveau/dispnv50/base507c.c |  6 +++---
 drivers/gpu/drm/nouveau/dispnv50/base827c.c |  6 +++---
 drivers/gpu/drm/nouveau/dispnv50/base907c.c |  6 +++---
 drivers/gpu/drm/nouveau/dispnv50/curs507a.c |  2 +-
 drivers/gpu/drm/nouveau/dispnv50/wndw.c     | 10 +++++-----
 6 files changed, 18 insertions(+), 18 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/dispnv50/atom.h b/drivers/gpu/drm/nouveau/dispnv50/atom.h
index 8bf1806..53638ee 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/atom.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/atom.h
@@ -162,12 +162,12 @@ struct nv50_wndw_atom {
 		u8  kind:7;
 		u8  layout:1;
 		u8  block:4;
-		u32 pitch:20;
+		u32 pitch[3];
 		u16 w;
 		u16 h;
 
-		u32 handle;
-		u64 offset;
+		u32 handle[6];
+		u64 offset[6];
 	} image;
 
 	struct {
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base507c.c b/drivers/gpu/drm/nouveau/dispnv50/base507c.c
index 43dcbcd..1c65ddc 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/base507c.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/base507c.c
@@ -74,13 +74,13 @@ base507c_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 		evo_data(push, asyw->image.mode << 8 |
 			       asyw->image.interval << 4);
 		evo_mthd(push, 0x00c0, 1);
-		evo_data(push, asyw->image.handle);
+		evo_data(push, asyw->image.handle[0]);
 		evo_mthd(push, 0x0800, 5);
-		evo_data(push, asyw->image.offset >> 8);
+		evo_data(push, asyw->image.offset[0] >> 8);
 		evo_data(push, 0x00000000);
 		evo_data(push, asyw->image.h << 16 | asyw->image.w);
 		evo_data(push, asyw->image.layout << 20 |
-			       asyw->image.pitch |
+			       asyw->image.pitch[0] |
 			       asyw->image.block);
 		evo_data(push, asyw->image.kind << 16 |
 			       asyw->image.format << 8);
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base827c.c b/drivers/gpu/drm/nouveau/dispnv50/base827c.c
index 0d356ae..9dc968c 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/base827c.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/base827c.c
@@ -30,13 +30,13 @@ base827c_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 		evo_data(push, asyw->image.mode << 8 |
 			       asyw->image.interval << 4);
 		evo_mthd(push, 0x00c0, 1);
-		evo_data(push, asyw->image.handle);
+		evo_data(push, asyw->image.handle[0]);
 		evo_mthd(push, 0x0800, 5);
-		evo_data(push, asyw->image.offset >> 8);
+		evo_data(push, asyw->image.offset[0] >> 8);
 		evo_data(push, 0x00000000);
 		evo_data(push, asyw->image.h << 16 | asyw->image.w);
 		evo_data(push, asyw->image.layout << 20 |
-			       asyw->image.pitch |
+			       asyw->image.pitch[0] |
 			       asyw->image.block);
 		evo_data(push, asyw->image.format << 8);
 		evo_kick(push, &wndw->wndw);
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base907c.c b/drivers/gpu/drm/nouveau/dispnv50/base907c.c
index 171d978..5321c55 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/base907c.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/base907c.c
@@ -43,13 +43,13 @@ base907c_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 		evo_data(push, asyw->image.mode << 8 |
 			       asyw->image.interval << 4);
 		evo_mthd(push, 0x00c0, 1);
-		evo_data(push, asyw->image.handle);
+		evo_data(push, asyw->image.handle[0]);
 		evo_mthd(push, 0x0400, 5);
-		evo_data(push, asyw->image.offset >> 8);
+		evo_data(push, asyw->image.offset[0] >> 8);
 		evo_data(push, 0x00000000);
 		evo_data(push, asyw->image.h << 16 | asyw->image.w);
 		evo_data(push, asyw->image.layout << 24 |
-			       asyw->image.pitch |
+			       asyw->image.pitch[0] |
 			       asyw->image.block);
 		evo_data(push, asyw->image.format << 8);
 		evo_kick(push, &wndw->wndw);
diff --git a/drivers/gpu/drm/nouveau/dispnv50/curs507a.c b/drivers/gpu/drm/nouveau/dispnv50/curs507a.c
index f7e56a8..589c75c 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/curs507a.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/curs507a.c
@@ -52,7 +52,7 @@ curs507a_prepare(struct nv50_wndw *wndw, struct nv50_head_atom *asyh,
 		 struct nv50_wndw_atom *asyw)
 {
 	u32 handle = nv50_disp(wndw->plane.dev)->core->chan.vram.handle;
-	u32 offset = asyw->image.offset;
+	u32 offset = asyw->image.offset[0];
 	if (asyh->curs.handle != handle || asyh->curs.offset != offset) {
 		asyh->curs.handle = handle;
 		asyh->curs.offset = offset;
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.c b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
index cfd998a..4b64f64 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/wndw.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
@@ -219,11 +219,11 @@ nv50_wndw_atomic_check_acquire(struct nv50_wndw *wndw,
 			asyw->image.block = fb->nvbo->mode >> 4;
 		else
 			asyw->image.block = fb->nvbo->mode;
-		asyw->image.pitch = (fb->base.pitches[0] / 4) << 4;
+		asyw->image.pitch[0] = (fb->base.pitches[0] / 4) << 4;
 	} else {
 		asyw->image.layout = 1;
 		asyw->image.block  = 0;
-		asyw->image.pitch  = fb->base.pitches[0];
+		asyw->image.pitch[0] = fb->base.pitches[0];
 	}
 
 	ret = wndw->func->acquire(wndw, asyw, asyh);
@@ -287,7 +287,7 @@ nv50_wndw_atomic_check(struct drm_plane *plane, struct drm_plane_state *state)
 		asyw->clr.ntfy = armw->ntfy.handle != 0;
 		asyw->clr.sema = armw->sema.handle != 0;
 		if (wndw->func->image_clr)
-			asyw->clr.image = armw->image.handle != 0;
+			asyw->clr.image = armw->image.handle[0] != 0;
 		asyw->set.lut = wndw->func->lut && asyv;
 	}
 
@@ -333,8 +333,8 @@ nv50_wndw_prepare_fb(struct drm_plane *plane, struct drm_plane_state *state)
 	}
 
 	asyw->state.fence = reservation_object_get_excl_rcu(fb->nvbo->bo.resv);
-	asyw->image.handle = ctxdma->object.handle;
-	asyw->image.offset = fb->nvbo->bo.offset;
+	asyw->image.handle[0] = ctxdma->object.handle;
+	asyw->image.offset[0] = fb->nvbo->bo.offset;
 
 	if (wndw->func->prepare) {
 		asyh = nv50_head_atom_get(asyw->state.state, asyw->state.crtc);
-- 
cgit v1.1


From 34508f9d260cbd7b91f988c858f50ad956750ee3 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50-: determine MST support from DP Info Table

GV100 doesn't support MST, use the information provided in VBIOS tables to
detect its presence instead.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/disp.c               | 9 +++++++--
 drivers/gpu/drm/nouveau/include/nvkm/subdev/bios/dp.h | 4 ++++
 drivers/gpu/drm/nouveau/nvkm/subdev/bios/dp.c         | 2 +-
 3 files changed, 12 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c
index 006562f..eaa63b4 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/disp.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c
@@ -52,6 +52,8 @@
 #include "nouveau_fence.h"
 #include "nouveau_fbcon.h"
 
+#include <subdev/bios/dp.h>
+
 /******************************************************************************
  * Atomic state
  *****************************************************************************/
@@ -1383,9 +1385,12 @@ nv50_sor_create(struct drm_connector *connector, struct dcb_output *dcbe)
 {
 	struct nouveau_connector *nv_connector = nouveau_connector(connector);
 	struct nouveau_drm *drm = nouveau_drm(connector->dev);
+	struct nvkm_bios *bios = nvxx_bios(&drm->client.device);
 	struct nvkm_i2c *i2c = nvxx_i2c(&drm->client.device);
 	struct nouveau_encoder *nv_encoder;
 	struct drm_encoder *encoder;
+	u8 ver, hdr, cnt, len;
+	u32 data;
 	int type, ret;
 
 	switch (dcbe->type) {
@@ -1429,8 +1434,8 @@ nv50_sor_create(struct drm_connector *connector, struct dcb_output *dcbe)
 			nv_encoder->aux = aux;
 		}
 
-		/*TODO: Use DP Info Table to check for support. */
-		if (disp->disp->object.oclass >= GF110_DISP) {
+		if ((data = nvbios_dp_table(bios, &ver, &hdr, &cnt, &len)) &&
+		    ver >= 0x40 && (nvbios_rd08(bios, data + 0x08) & 0x04)) {
 			ret = nv50_mstm_new(nv_encoder, &nv_connector->aux, 16,
 					    nv_connector->base.base.id,
 					    &nv_encoder->dp.mstm);
diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/bios/dp.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/bios/dp.h
index df34b41..512e25a 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/bios/dp.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/bios/dp.h
@@ -1,6 +1,10 @@
 /* SPDX-License-Identifier: GPL-2.0 */
 #ifndef __NVBIOS_DP_H__
 #define __NVBIOS_DP_H__
+
+u16
+nvbios_dp_table(struct nvkm_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
+
 struct nvbios_dpout {
 	u16 type;
 	u16 mask;
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/dp.c b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/dp.c
index 7c7efa4..3133b28 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/dp.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/dp.c
@@ -25,7 +25,7 @@
 #include <subdev/bios/bit.h>
 #include <subdev/bios/dp.h>
 
-static u16
+u16
 nvbios_dp_table(struct nvkm_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
 {
 	struct bit_entry d;
-- 
cgit v1.1


From 53e0a3e70de69dc9f498d26c6b5495b2771ee374 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50-: simplify tracking of channel interlocks

Instead of windows returning their core channel interlock mask if they
know core has been modified, it's recorded unconditionally and used if
required when update methods are emitted.

This will be required to support Volta.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/Kbuild     |  3 ++
 drivers/gpu/drm/nouveau/dispnv50/base.h     |  4 +--
 drivers/gpu/drm/nouveau/dispnv50/base507c.c | 16 +++++------
 drivers/gpu/drm/nouveau/dispnv50/base827c.c |  3 +-
 drivers/gpu/drm/nouveau/dispnv50/base907c.c | 18 ++----------
 drivers/gpu/drm/nouveau/dispnv50/core.h     |  5 ++--
 drivers/gpu/drm/nouveau/dispnv50/core507d.c |  5 ++--
 drivers/gpu/drm/nouveau/dispnv50/curs.c     |  4 +--
 drivers/gpu/drm/nouveau/dispnv50/curs.h     |  6 ++++
 drivers/gpu/drm/nouveau/dispnv50/curs507a.c | 18 ++++++------
 drivers/gpu/drm/nouveau/dispnv50/curs907a.c | 30 ++++++++++++++++++++
 drivers/gpu/drm/nouveau/dispnv50/disp.c     | 42 +++++++++++++---------------
 drivers/gpu/drm/nouveau/dispnv50/disp.h     | 11 ++++++++
 drivers/gpu/drm/nouveau/dispnv50/ovly.c     | 10 +++----
 drivers/gpu/drm/nouveau/dispnv50/ovly.h     |  8 ++++++
 drivers/gpu/drm/nouveau/dispnv50/ovly507e.c | 11 +++++---
 drivers/gpu/drm/nouveau/dispnv50/ovly827e.c | 43 +++++++++++++++++++++++++++++
 drivers/gpu/drm/nouveau/dispnv50/ovly907e.c | 34 +++++++++++++++++++++++
 drivers/gpu/drm/nouveau/dispnv50/wndw.c     | 23 ++++++++++-----
 drivers/gpu/drm/nouveau/dispnv50/wndw.h     | 16 ++++++-----
 20 files changed, 224 insertions(+), 86 deletions(-)
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/curs907a.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/ovly827e.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/ovly907e.c

diff --git a/drivers/gpu/drm/nouveau/dispnv50/Kbuild b/drivers/gpu/drm/nouveau/dispnv50/Kbuild
index 674221d..3e53484 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/Kbuild
+++ b/drivers/gpu/drm/nouveau/dispnv50/Kbuild
@@ -29,9 +29,12 @@ nouveau-y += dispnv50/base907c.o
 
 nouveau-y += dispnv50/curs.o
 nouveau-y += dispnv50/curs507a.o
+nouveau-y += dispnv50/curs907a.o
 
 nouveau-y += dispnv50/oimm.o
 nouveau-y += dispnv50/oimm507b.o
 
 nouveau-y += dispnv50/ovly.o
 nouveau-y += dispnv50/ovly507e.o
+nouveau-y += dispnv50/ovly827e.o
+nouveau-y += dispnv50/ovly907e.o
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base.h b/drivers/gpu/drm/nouveau/dispnv50/base.h
index edf96a8..71fc103 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/base.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/base.h
@@ -5,7 +5,7 @@
 int base507c_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
 int base507c_new_(const struct nv50_wndw_func *, const u32 *format,
 		  struct nouveau_drm *, int head, s32 oclass,
-		  struct nv50_wndw **);
+		  u32 interlock_data, struct nv50_wndw **);
 extern const u32 base507c_format[];
 int base507c_acquire(struct nv50_wndw *, struct nv50_wndw_atom *,
 		     struct nv50_head_atom *);
@@ -19,7 +19,7 @@ void base507c_ntfy_clr(struct nv50_wndw *);
 int base507c_ntfy_wait_begun(struct nouveau_bo *, u32, struct nvif_device *);
 void base507c_image_clr(struct nv50_wndw *);
 void base507c_lut(struct nv50_wndw *, struct nv50_wndw_atom *);
-u32 base507c_update(struct nv50_wndw *, u32);
+void base507c_update(struct nv50_wndw *, u32 *);
 
 int base827c_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
 
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base507c.c b/drivers/gpu/drm/nouveau/dispnv50/base507c.c
index 1c65ddc..819403f 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/base507c.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/base507c.c
@@ -28,17 +28,15 @@
 #include <drm/drm_plane_helper.h>
 #include "nouveau_bo.h"
 
-u32
-base507c_update(struct nv50_wndw *wndw, u32 interlock)
+void
+base507c_update(struct nv50_wndw *wndw, u32 *interlock)
 {
 	u32 *push;
 	if ((push = evo_wait(&wndw->wndw, 2))) {
 		evo_mthd(push, 0x0080, 1);
-		evo_data(push, interlock);
+		evo_data(push, interlock[NV50_DISP_INTERLOCK_CORE]);
 		evo_kick(push, &wndw->wndw);
-		return interlock ? 2 << (wndw->id * 8) : 0;
 	}
-	return 0;
 }
 
 void
@@ -224,7 +222,7 @@ base507c = {
 
 int
 base507c_new_(const struct nv50_wndw_func *func, const u32 *format,
-	      struct nouveau_drm *drm, int head, s32 oclass,
+	      struct nouveau_drm *drm, int head, s32 oclass, u32 interlock_data,
 	      struct nv50_wndw **pwndw)
 {
 	struct nv50_disp_base_channel_dma_v0 args = {
@@ -235,7 +233,8 @@ base507c_new_(const struct nv50_wndw_func *func, const u32 *format,
 	int ret;
 
 	ret = nv50_wndw_new_(func, drm->dev, DRM_PLANE_TYPE_PRIMARY,
-			     "base", head, format, BIT(head), &wndw);
+			     "base", head, format, BIT(head),
+			     NV50_DISP_INTERLOCK_BASE, interlock_data, &wndw);
 	if (*pwndw = wndw, ret)
 		return ret;
 
@@ -266,5 +265,6 @@ int
 base507c_new(struct nouveau_drm *drm, int head, s32 oclass,
 	     struct nv50_wndw **pwndw)
 {
-	return base507c_new_(&base507c, base507c_format, drm, head, oclass, pwndw);
+	return base507c_new_(&base507c, base507c_format, drm, head, oclass,
+			     0x00000002 << (head * 8), pwndw);
 }
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base827c.c b/drivers/gpu/drm/nouveau/dispnv50/base827c.c
index 9dc968c..240a640 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/base827c.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/base827c.c
@@ -63,5 +63,6 @@ int
 base827c_new(struct nouveau_drm *drm, int head, s32 oclass,
 	     struct nv50_wndw **pwndw)
 {
-	return base507c_new_(&base827c, base507c_format, drm, head, oclass, pwndw);
+	return base507c_new_(&base827c, base507c_format, drm, head, oclass,
+			     0x00000002 << (head * 8), pwndw);
 }
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base907c.c b/drivers/gpu/drm/nouveau/dispnv50/base907c.c
index 5321c55..6c32a4e 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/base907c.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/base907c.c
@@ -21,19 +21,6 @@
  */
 #include "base.h"
 
-static u32
-base907c_update(struct nv50_wndw *wndw, u32 interlock)
-{
-	u32 *push;
-	if ((push = evo_wait(&wndw->wndw, 2))) {
-		evo_mthd(push, 0x0080, 1);
-		evo_data(push, interlock);
-		evo_kick(push, &wndw->wndw);
-		return interlock ? 2 << (wndw->id * 4) : 0;
-	}
-	return 0;
-}
-
 static void
 base907c_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 {
@@ -69,12 +56,13 @@ base907c = {
 	.image_set = base907c_image_set,
 	.image_clr = base507c_image_clr,
 	.lut = base507c_lut,
-	.update = base907c_update,
+	.update = base507c_update,
 };
 
 int
 base907c_new(struct nouveau_drm *drm, int head, s32 oclass,
 	     struct nv50_wndw **pwndw)
 {
-	return base507c_new_(&base907c, base507c_format, drm, head, oclass, pwndw);
+	return base507c_new_(&base907c, base507c_format, drm, head, oclass,
+			     0x00000002 << (head * 4), pwndw);
 }
diff --git a/drivers/gpu/drm/nouveau/dispnv50/core.h b/drivers/gpu/drm/nouveau/dispnv50/core.h
index 5fd7ddd..c490d7d 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/core.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/core.h
@@ -16,7 +16,7 @@ struct nv50_core_func {
 	void (*ntfy_init)(struct nouveau_bo *, u32 offset);
 	int (*ntfy_wait_done)(struct nouveau_bo *, u32 offset,
 			      struct nvif_device *);
-	void (*update)(struct nv50_core *, u32 interlock, bool ntfy);
+	void (*update)(struct nv50_core *, u32 *interlock, bool ntfy);
 
 	const struct nv50_head_func *head;
 	const struct nv50_outp_func {
@@ -31,7 +31,8 @@ int core507d_new_(const struct nv50_core_func *, struct nouveau_drm *, s32,
 void core507d_init(struct nv50_core *);
 void core507d_ntfy_init(struct nouveau_bo *, u32);
 int core507d_ntfy_wait_done(struct nouveau_bo *, u32, struct nvif_device *);
-void core507d_update(struct nv50_core *, u32, bool);
+void core507d_update(struct nv50_core *, u32 *, bool);
+
 extern const struct nv50_outp_func dac507d;
 extern const struct nv50_outp_func sor507d;
 extern const struct nv50_outp_func pior507d;
diff --git a/drivers/gpu/drm/nouveau/dispnv50/core507d.c b/drivers/gpu/drm/nouveau/dispnv50/core507d.c
index 96d7d8f..e7fcfa6 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/core507d.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/core507d.c
@@ -27,7 +27,7 @@
 #include "nouveau_bo.h"
 
 void
-core507d_update(struct nv50_core *core, u32 interlock, bool ntfy)
+core507d_update(struct nv50_core *core, u32 *interlock, bool ntfy)
 {
 	u32 *push;
 	if ((push = evo_wait(&core->chan, 5))) {
@@ -36,7 +36,8 @@ core507d_update(struct nv50_core *core, u32 interlock, bool ntfy)
 			evo_data(push, 0x80000000 | NV50_DISP_CORE_NTFY);
 		}
 		evo_mthd(push, 0x0080, 2);
-		evo_data(push, interlock);
+		evo_data(push, interlock[NV50_DISP_INTERLOCK_BASE] |
+			       interlock[NV50_DISP_INTERLOCK_OVLY]);
 		evo_data(push, 0x00000000);
 		evo_kick(push, &core->chan);
 	}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/curs.c b/drivers/gpu/drm/nouveau/dispnv50/curs.c
index 6d60e97..fb842ed 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/curs.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/curs.c
@@ -31,8 +31,8 @@ nv50_curs_new(struct nouveau_drm *drm, int head, struct nv50_wndw **pwndw)
 		int version;
 		int (*new)(struct nouveau_drm *, int, s32, struct nv50_wndw **);
 	} curses[] = {
-		{ GK104_DISP_CURSOR, 0, curs507a_new },
-		{ GF110_DISP_CURSOR, 0, curs507a_new },
+		{ GK104_DISP_CURSOR, 0, curs907a_new },
+		{ GF110_DISP_CURSOR, 0, curs907a_new },
 		{ GT214_DISP_CURSOR, 0, curs507a_new },
 		{   G82_DISP_CURSOR, 0, curs507a_new },
 		{  NV50_DISP_CURSOR, 0, curs507a_new },
diff --git a/drivers/gpu/drm/nouveau/dispnv50/curs.h b/drivers/gpu/drm/nouveau/dispnv50/curs.h
index b85ca9f..2285247 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/curs.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/curs.h
@@ -3,6 +3,12 @@
 #include "wndw.h"
 
 int curs507a_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+int curs507a_new_(const struct nv50_wimm_func *, struct nouveau_drm *,
+		  int head, s32 oclass, u32 interlock_data,
+		  struct nv50_wndw **);
+extern const struct nv50_wimm_func curs507a;
+
+int curs907a_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
 
 int nv50_curs_new(struct nouveau_drm *, int head, struct nv50_wndw **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/curs507a.c b/drivers/gpu/drm/nouveau/dispnv50/curs507a.c
index 589c75c..ba05bcb1 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/curs507a.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/curs507a.c
@@ -27,11 +27,10 @@
 #include <drm/drm_atomic_helper.h>
 #include <drm/drm_plane_helper.h>
 
-static u32
-curs507a_update(struct nv50_wndw *wndw, u32 interlock)
+static void
+curs507a_update(struct nv50_wndw *wndw, u32 *interlock)
 {
 	nvif_wr32(&wndw->wimm.base.user, 0x0080, 0x00000000);
-	return 0;
 }
 
 static void
@@ -41,7 +40,7 @@ curs507a_point(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 						 asyw->point.x);
 }
 
-static const struct nv50_wimm_func
+const struct nv50_wimm_func
 curs507a = {
 	.point = curs507a_point,
 	.update = curs507a_update,
@@ -114,9 +113,10 @@ curs507a_wndw = {
 	.prepare = curs507a_prepare,
 };
 
-static int
+int
 curs507a_new_(const struct nv50_wimm_func *func, struct nouveau_drm *drm,
-	      int head, s32 oclass, struct nv50_wndw **pwndw)
+	      int head, s32 oclass, u32 interlock_data,
+	      struct nv50_wndw **pwndw)
 {
 	struct nv50_disp_cursor_v0 args = {
 		.head = head,
@@ -126,7 +126,8 @@ curs507a_new_(const struct nv50_wimm_func *func, struct nouveau_drm *drm,
 	int ret;
 
 	ret = nv50_wndw_new_(&curs507a_wndw, drm->dev, DRM_PLANE_TYPE_CURSOR,
-			     "curs", head, curs507a_format, BIT(head), &wndw);
+			     "curs", head, curs507a_format, BIT(head),
+			     NV50_DISP_INTERLOCK_CURS, interlock_data, &wndw);
 	if (*pwndw = wndw, ret)
 		return ret;
 
@@ -147,5 +148,6 @@ int
 curs507a_new(struct nouveau_drm *drm, int head, s32 oclass,
 	     struct nv50_wndw **pwndw)
 {
-	return curs507a_new_(&curs507a, drm, head, oclass, pwndw);
+	return curs507a_new_(&curs507a, drm, head, oclass,
+			     0x00000001 << (head * 8), pwndw);
 }
diff --git a/drivers/gpu/drm/nouveau/dispnv50/curs907a.c b/drivers/gpu/drm/nouveau/dispnv50/curs907a.c
new file mode 100644
index 0000000..d742362
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/curs907a.c
@@ -0,0 +1,30 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "curs.h"
+
+int
+curs907a_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return curs507a_new_(&curs507a, drm, head, oclass,
+			     0x00000001 << (head * 4), pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c
index eaa63b4..e80d11c 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/disp.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c
@@ -1582,14 +1582,14 @@ nv50_pior_create(struct drm_connector *connector, struct dcb_output *dcbe)
  *****************************************************************************/
 
 static void
-nv50_disp_atomic_commit_core(struct nouveau_drm *drm, u32 interlock)
+nv50_disp_atomic_commit_core(struct nouveau_drm *drm, u32 *interlock)
 {
 	struct nv50_disp *disp = nv50_disp(drm->dev);
 	struct nv50_core *core = disp->core;
 	struct nv50_mstm *mstm;
 	struct drm_encoder *encoder;
 
-	NV_ATOMIC(drm, "commit core %08x\n", interlock);
+	NV_ATOMIC(drm, "commit core %08x\n", interlock[NV50_DISP_INTERLOCK_BASE]);
 
 	drm_for_each_encoder(encoder, drm->dev) {
 		if (encoder->encoder_type != DRM_MODE_ENCODER_DPMST) {
@@ -1626,8 +1626,7 @@ nv50_disp_atomic_commit_tail(struct drm_atomic_state *state)
 	struct nv50_disp *disp = nv50_disp(dev);
 	struct nv50_atom *atom = nv50_atom(state);
 	struct nv50_outp_atom *outp, *outt;
-	u32 interlock_core = 0;
-	u32 interlock_chan = 0;
+	u32 interlock[NV50_DISP_INTERLOCK__SIZE] = {};
 	int i;
 
 	NV_ATOMIC(drm, "commit %d %d\n", atom->lock_core, atom->flush_disable);
@@ -1650,7 +1649,7 @@ nv50_disp_atomic_commit_tail(struct drm_atomic_state *state)
 
 		if (asyh->clr.mask) {
 			nv50_head_flush_clr(head, asyh, atom->flush_disable);
-			interlock_core |= 1;
+			interlock[NV50_DISP_INTERLOCK_CORE] |= 1;
 		}
 	}
 
@@ -1664,9 +1663,7 @@ nv50_disp_atomic_commit_tail(struct drm_atomic_state *state)
 		if (!asyw->clr.mask)
 			continue;
 
-		interlock_chan |= nv50_wndw_flush_clr(wndw, interlock_core,
-						      atom->flush_disable,
-						      asyw);
+		nv50_wndw_flush_clr(wndw, interlock, atom->flush_disable, asyw);
 	}
 
 	/* Disable output path(s). */
@@ -1682,21 +1679,19 @@ nv50_disp_atomic_commit_tail(struct drm_atomic_state *state)
 
 		if (outp->clr.mask) {
 			help->disable(encoder);
-			interlock_core |= 1;
+			interlock[NV50_DISP_INTERLOCK_CORE] |= 1;
 			if (outp->flush_disable) {
-				nv50_disp_atomic_commit_core(drm, interlock_chan);
-				interlock_core = 0;
-				interlock_chan = 0;
+				nv50_disp_atomic_commit_core(drm, interlock);
+				memset(interlock, 0x00, sizeof(interlock));
 			}
 		}
 	}
 
 	/* Flush disable. */
-	if (interlock_core) {
+	if (interlock[NV50_DISP_INTERLOCK_CORE]) {
 		if (atom->flush_disable) {
-			nv50_disp_atomic_commit_core(drm, interlock_chan);
-			interlock_core = 0;
-			interlock_chan = 0;
+			nv50_disp_atomic_commit_core(drm, interlock);
+			memset(interlock, 0x00, sizeof(interlock));
 		}
 	}
 
@@ -1713,7 +1708,7 @@ nv50_disp_atomic_commit_tail(struct drm_atomic_state *state)
 
 		if (outp->set.mask) {
 			help->enable(encoder);
-			interlock_core = 1;
+			interlock[NV50_DISP_INTERLOCK_CORE] = 1;
 		}
 
 		list_del(&outp->head);
@@ -1730,7 +1725,7 @@ nv50_disp_atomic_commit_tail(struct drm_atomic_state *state)
 
 		if (asyh->set.mask) {
 			nv50_head_flush_set(head, asyh);
-			interlock_core = 1;
+			interlock[NV50_DISP_INTERLOCK_CORE] = 1;
 		}
 
 		if (new_crtc_state->active) {
@@ -1752,15 +1747,16 @@ nv50_disp_atomic_commit_tail(struct drm_atomic_state *state)
 		    (!asyw->clr.mask || atom->flush_disable))
 			continue;
 
-		interlock_chan |= nv50_wndw_flush_set(wndw, interlock_core, asyw);
+		nv50_wndw_flush_set(wndw, interlock, asyw);
 	}
 
 	/* Flush update. */
-	if (interlock_core) {
-		if (interlock_chan || !atom->state.legacy_cursor_update)
-			nv50_disp_atomic_commit_core(drm, interlock_chan);
+	if (interlock[NV50_DISP_INTERLOCK_CORE]) {
+		if (interlock[NV50_DISP_INTERLOCK_BASE] ||
+		    !atom->state.legacy_cursor_update)
+			nv50_disp_atomic_commit_core(drm, interlock);
 		else
-			disp->core->func->update(disp->core, 0, false);
+			disp->core->func->update(disp->core, interlock, false);
 	}
 
 	if (atom->lock_core)
diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.h b/drivers/gpu/drm/nouveau/dispnv50/disp.h
index 7cbd668..f3a963b 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/disp.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.h
@@ -27,6 +27,17 @@ nv50_disp(struct drm_device *dev)
 	return nouveau_display(dev)->priv;
 }
 
+struct nv50_disp_interlock {
+	enum nv50_disp_interlock_type {
+		NV50_DISP_INTERLOCK_CORE = 0,
+		NV50_DISP_INTERLOCK_CURS,
+		NV50_DISP_INTERLOCK_BASE,
+		NV50_DISP_INTERLOCK_OVLY,
+		NV50_DISP_INTERLOCK__SIZE
+	} type;
+	u32 data;
+};
+
 struct nv50_chan {
 	struct nvif_object user;
 	struct nvif_device *device;
diff --git a/drivers/gpu/drm/nouveau/dispnv50/ovly.c b/drivers/gpu/drm/nouveau/dispnv50/ovly.c
index ac2d3b6..be0f16f 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/ovly.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/ovly.c
@@ -32,11 +32,11 @@ nv50_ovly_new(struct nouveau_drm *drm, int head, struct nv50_wndw **pwndw)
 		int version;
 		int (*new)(struct nouveau_drm *, int, s32, struct nv50_wndw **);
 	} ovlys[] = {
-		{ GK104_DISP_OVERLAY_CONTROL_DMA, 0, ovly507e_new },
-		{ GF110_DISP_OVERLAY_CONTROL_DMA, 0, ovly507e_new },
-		{ GT214_DISP_OVERLAY_CHANNEL_DMA, 0, ovly507e_new },
-		{ GT200_DISP_OVERLAY_CHANNEL_DMA, 0, ovly507e_new },
-		{   G82_DISP_OVERLAY_CHANNEL_DMA, 0, ovly507e_new },
+		{ GK104_DISP_OVERLAY_CONTROL_DMA, 0, ovly907e_new },
+		{ GF110_DISP_OVERLAY_CONTROL_DMA, 0, ovly907e_new },
+		{ GT214_DISP_OVERLAY_CHANNEL_DMA, 0, ovly827e_new },
+		{ GT200_DISP_OVERLAY_CHANNEL_DMA, 0, ovly827e_new },
+		{   G82_DISP_OVERLAY_CHANNEL_DMA, 0, ovly827e_new },
 		{  NV50_DISP_OVERLAY_CHANNEL_DMA, 0, ovly507e_new },
 		{}
 	};
diff --git a/drivers/gpu/drm/nouveau/dispnv50/ovly.h b/drivers/gpu/drm/nouveau/dispnv50/ovly.h
index 90af1f2..d149ef6 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/ovly.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/ovly.h
@@ -3,6 +3,14 @@
 #include "wndw.h"
 
 int ovly507e_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+int ovly507e_new_(const struct nv50_wndw_func *, const u32 *format,
+		  struct nouveau_drm *, int head, s32 oclass,
+		  u32 interlock_data, struct nv50_wndw **);
+
+extern const u32 ovly827e_format[];
+
+int ovly827e_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+int ovly907e_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
 
 int nv50_ovly_new(struct nouveau_drm *, int head, struct nv50_wndw **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/ovly507e.c b/drivers/gpu/drm/nouveau/dispnv50/ovly507e.c
index 1b85262..732eea3 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/ovly507e.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/ovly507e.c
@@ -34,9 +34,9 @@ ovly507e_format[] = {
 	0
 };
 
-static int
+int
 ovly507e_new_(const struct nv50_wndw_func *func, const u32 *format,
-	      struct nouveau_drm *drm, int head, s32 oclass,
+	      struct nouveau_drm *drm, int head, s32 oclass, u32 interlock_data,
 	      struct nv50_wndw **pwndw)
 {
 	struct nv50_disp_overlay_channel_dma_v0 args = {
@@ -47,7 +47,9 @@ ovly507e_new_(const struct nv50_wndw_func *func, const u32 *format,
 	int ret;
 
 	ret = nv50_wndw_new_(func, drm->dev, DRM_PLANE_TYPE_OVERLAY,
-			     "ovly", head, format, BIT(head), &wndw);
+			     "ovly", head, format, BIT(head),
+			     NV50_DISP_INTERLOCK_OVLY, interlock_data,
+			     &wndw);
 	if (*pwndw = wndw, ret)
 		return ret;
 
@@ -66,5 +68,6 @@ int
 ovly507e_new(struct nouveau_drm *drm, int head, s32 oclass,
 	     struct nv50_wndw **pwndw)
 {
-	return ovly507e_new_(&ovly507e, ovly507e_format, drm, head, oclass, pwndw);
+	return ovly507e_new_(&ovly507e, ovly507e_format, drm, head, oclass,
+			     0x00000004 << (head * 8), pwndw);
 }
diff --git a/drivers/gpu/drm/nouveau/dispnv50/ovly827e.c b/drivers/gpu/drm/nouveau/dispnv50/ovly827e.c
new file mode 100644
index 0000000..a8115f1
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/ovly827e.c
@@ -0,0 +1,43 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "ovly.h"
+
+#include <nouveau_bo.h>
+
+#include <nvif/cl507e.h>
+
+static const struct nv50_wndw_func
+ovly827e = {
+};
+
+const u32
+ovly827e_format[] = {
+	0
+};
+
+int
+ovly827e_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return ovly507e_new_(&ovly827e, ovly827e_format, drm, head, oclass,
+			     0x00000004 << (head * 8), pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/ovly907e.c b/drivers/gpu/drm/nouveau/dispnv50/ovly907e.c
new file mode 100644
index 0000000..f50da646
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/ovly907e.c
@@ -0,0 +1,34 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "ovly.h"
+
+static const struct nv50_wndw_func
+ovly907e = {
+};
+
+int
+ovly907e_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return ovly507e_new_(&ovly907e, ovly827e_format, drm, head, oclass,
+			     0x00000004 << (head * 4), pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.c b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
index 4b64f64..8f62c2a 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/wndw.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
@@ -107,8 +107,8 @@ nv50_wndw_wait_armed(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 	return 0;
 }
 
-u32
-nv50_wndw_flush_clr(struct nv50_wndw *wndw, u32 interlock, bool flush,
+void
+nv50_wndw_flush_clr(struct nv50_wndw *wndw, u32 *interlock, bool flush,
 		    struct nv50_wndw_atom *asyw)
 {
 	union nv50_wndw_atom_mask clr = {
@@ -118,11 +118,13 @@ nv50_wndw_flush_clr(struct nv50_wndw *wndw, u32 interlock, bool flush,
 	if (clr.ntfy ) wndw->func-> ntfy_clr(wndw);
 	if (clr.image) wndw->func->image_clr(wndw);
 
-	return flush ? wndw->func->update(wndw, interlock) : 0;
+	interlock[wndw->interlock.type] |= wndw->interlock.data;
+	if (flush)
+		wndw->func->update(wndw, interlock);
 }
 
-u32
-nv50_wndw_flush_set(struct nv50_wndw *wndw, u32 interlock,
+void
+nv50_wndw_flush_set(struct nv50_wndw *wndw, u32 *interlock,
 		    struct nv50_wndw_atom *asyw)
 {
 	if (interlock) {
@@ -139,7 +141,9 @@ nv50_wndw_flush_set(struct nv50_wndw *wndw, u32 interlock,
 		wndw->immd->update(wndw, interlock);
 	}
 
-	return wndw->func->update ? wndw->func->update(wndw, interlock) : 0;
+	interlock[wndw->interlock.type] |= wndw->interlock.data;
+	if (wndw->func->update)
+		wndw->func->update(wndw, interlock);
 }
 
 void
@@ -445,7 +449,9 @@ nv50_wndw_init(struct nv50_wndw *wndw)
 int
 nv50_wndw_new_(const struct nv50_wndw_func *func, struct drm_device *dev,
 	       enum drm_plane_type type, const char *name, int index,
-	       const u32 *format, u32 heads, struct nv50_wndw **pwndw)
+	       const u32 *format, u32 heads,
+	       enum nv50_disp_interlock_type interlock_type, u32 interlock_data,
+	       struct nv50_wndw **pwndw)
 {
 	struct nv50_wndw *wndw;
 	int nformat;
@@ -455,6 +461,9 @@ nv50_wndw_new_(const struct nv50_wndw_func *func, struct drm_device *dev,
 		return -ENOMEM;
 	wndw->func = func;
 	wndw->id = index;
+	wndw->interlock.type = interlock_type;
+	wndw->interlock.data = interlock_data;
+	wndw->ctxdma.parent = &wndw->wndw.base.user;
 
 	wndw->ctxdma.parent = &wndw->wndw.base.user;
 	INIT_LIST_HEAD(&wndw->ctxdma.list);
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.h b/drivers/gpu/drm/nouveau/dispnv50/wndw.h
index 8672c28..c26796c 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/wndw.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.h
@@ -15,6 +15,7 @@ struct nv50_wndw {
 	const struct nv50_wndw_func *func;
 	const struct nv50_wimm_func *immd;
 	int id;
+	struct nv50_disp_interlock interlock;
 
 	struct {
 		struct nvif_object *parent;
@@ -34,13 +35,14 @@ struct nv50_wndw {
 
 int nv50_wndw_new_(const struct nv50_wndw_func *, struct drm_device *,
 		   enum drm_plane_type, const char *name, int index,
-		   const u32 *format, u32 heads, struct nv50_wndw **);
+		   const u32 *format, enum nv50_disp_interlock_type,
+		   u32 interlock_data, u32 heads, struct nv50_wndw **);
 void nv50_wndw_init(struct nv50_wndw *);
 void nv50_wndw_fini(struct nv50_wndw *);
-u32 nv50_wndw_flush_set(struct nv50_wndw *, u32 interlock,
-			struct nv50_wndw_atom *);
-u32 nv50_wndw_flush_clr(struct nv50_wndw *, u32 interlock, bool flush,
-			struct nv50_wndw_atom *);
+void nv50_wndw_flush_set(struct nv50_wndw *, u32 *interlock,
+			 struct nv50_wndw_atom *);
+void nv50_wndw_flush_clr(struct nv50_wndw *, u32 *interlock, bool flush,
+			 struct nv50_wndw_atom *);
 void nv50_wndw_ntfy_enable(struct nv50_wndw *, struct nv50_wndw_atom *);
 int nv50_wndw_wait_armed(struct nv50_wndw *, struct nv50_wndw_atom *);
 
@@ -63,7 +65,7 @@ struct nv50_wndw_func {
 	void (*image_clr)(struct nv50_wndw *);
 	void (*lut)(struct nv50_wndw *, struct nv50_wndw_atom *);
 
-	u32 (*update)(struct nv50_wndw *, u32 interlock);
+	void (*update)(struct nv50_wndw *, u32 *interlock);
 };
 
 extern const struct drm_plane_funcs nv50_wndw;
@@ -71,6 +73,6 @@ extern const struct drm_plane_funcs nv50_wndw;
 struct nv50_wimm_func {
 	void (*point)(struct nv50_wndw *, struct nv50_wndw_atom *);
 
-	u32 (*update)(struct nv50_wndw *, u32 interlock);
+	void (*update)(struct nv50_wndw *, u32 *interlock);
 };
 #endif
-- 
cgit v1.1


From 04fc14be7726edbb34404f69297e74061a8a9563 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50-: decouple window state changes, and update
 method submisssion

This will be required to support Volta.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/disp.c | 16 ++++++++++++++++
 drivers/gpu/drm/nouveau/dispnv50/wndw.c |  4 ----
 2 files changed, 16 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c
index e80d11c..0f20200 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/disp.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c
@@ -1690,6 +1690,14 @@ nv50_disp_atomic_commit_tail(struct drm_atomic_state *state)
 	/* Flush disable. */
 	if (interlock[NV50_DISP_INTERLOCK_CORE]) {
 		if (atom->flush_disable) {
+			for_each_new_plane_in_state(state, plane, new_plane_state, i) {
+				struct nv50_wndw *wndw = nv50_wndw(plane);
+				if (interlock[wndw->interlock.type] & wndw->interlock.data) {
+					if (wndw->func->update)
+						wndw->func->update(wndw, interlock);
+				}
+			}
+
 			nv50_disp_atomic_commit_core(drm, interlock);
 			memset(interlock, 0x00, sizeof(interlock));
 		}
@@ -1751,6 +1759,14 @@ nv50_disp_atomic_commit_tail(struct drm_atomic_state *state)
 	}
 
 	/* Flush update. */
+	for_each_new_plane_in_state(state, plane, new_plane_state, i) {
+		struct nv50_wndw *wndw = nv50_wndw(plane);
+		if (interlock[wndw->interlock.type] & wndw->interlock.data) {
+			if (wndw->func->update)
+				wndw->func->update(wndw, interlock);
+		}
+	}
+
 	if (interlock[NV50_DISP_INTERLOCK_CORE]) {
 		if (interlock[NV50_DISP_INTERLOCK_BASE] ||
 		    !atom->state.legacy_cursor_update)
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.c b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
index 8f62c2a..0fba4e0 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/wndw.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
@@ -119,8 +119,6 @@ nv50_wndw_flush_clr(struct nv50_wndw *wndw, u32 *interlock, bool flush,
 	if (clr.image) wndw->func->image_clr(wndw);
 
 	interlock[wndw->interlock.type] |= wndw->interlock.data;
-	if (flush)
-		wndw->func->update(wndw, interlock);
 }
 
 void
@@ -142,8 +140,6 @@ nv50_wndw_flush_set(struct nv50_wndw *wndw, u32 *interlock,
 	}
 
 	interlock[wndw->interlock.type] |= wndw->interlock.data;
-	if (wndw->func->update)
-		wndw->func->update(wndw, interlock);
 }
 
 void
-- 
cgit v1.1


From 45a2945a3759479c08a4aceaee181639c92f9d48 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50-: simplify swap interval handling

This is just cleaning up some left-overs from when we needed a custom
legacy page flip implementation.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/atom.h |  1 -
 drivers/gpu/drm/nouveau/dispnv50/wndw.c | 11 +++--------
 2 files changed, 3 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/dispnv50/atom.h b/drivers/gpu/drm/nouveau/dispnv50/atom.h
index 53638ee..b5b8a12 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/atom.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/atom.h
@@ -135,7 +135,6 @@ nv50_head_atom_get(struct drm_atomic_state *state, struct drm_crtc *crtc)
 
 struct nv50_wndw_atom {
 	struct drm_plane_state state;
-	u8 interval;
 
 	struct {
 		u32  handle;
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.c b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
index 0fba4e0..06d1696 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/wndw.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
@@ -208,11 +208,6 @@ nv50_wndw_atomic_check_acquire(struct nv50_wndw *wndw,
 	if (ret)
 		return ret;
 
-	if (asyh->state.pageflip_flags & DRM_MODE_PAGE_FLIP_ASYNC)
-		asyw->interval = 0;
-	else
-		asyw->interval = 1;
-
 	if (asyw->image.kind) {
 		asyw->image.layout = 0;
 		if (drm->client.device.info.chipset >= 0xc0)
@@ -231,10 +226,11 @@ nv50_wndw_atomic_check_acquire(struct nv50_wndw *wndw,
 		return ret;
 
 	if (asyw->set.image) {
-		if (!(asyw->image.mode = asyw->interval ? 0 : 1))
-			asyw->image.interval = asyw->interval;
+		if (!(asyh->state.pageflip_flags & DRM_MODE_PAGE_FLIP_ASYNC))
+			asyw->image.interval = 1;
 		else
 			asyw->image.interval = 0;
+		asyw->image.mode = asyw->image.interval ? 0 : 1;
 	}
 
 	return 0;
@@ -371,7 +367,6 @@ nv50_wndw_atomic_duplicate_state(struct drm_plane *plane)
 	if (!(asyw = kmalloc(sizeof(*asyw), GFP_KERNEL)))
 		return NULL;
 	__drm_atomic_helper_plane_duplicate_state(plane, &asyw->state);
-	asyw->interval = 1;
 	asyw->sema = armw->sema;
 	asyw->ntfy = armw->ntfy;
 	asyw->image = armw->image;
-- 
cgit v1.1


From 859b456b6b19a19761883cf52993dec645a36152 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50-: store window visibility in state

Window visibility is going to become a little more complicated with the
upcoming LUT changes, so store the calculated value to avoid needing to
recalculate the armed state again.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/atom.h |  2 ++
 drivers/gpu/drm/nouveau/dispnv50/wndw.c | 27 +++++++++++++++++++--------
 2 files changed, 21 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/dispnv50/atom.h b/drivers/gpu/drm/nouveau/dispnv50/atom.h
index b5b8a12..fefb9ca 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/atom.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/atom.h
@@ -136,6 +136,8 @@ nv50_head_atom_get(struct drm_atomic_state *state, struct drm_crtc *crtc)
 struct nv50_wndw_atom {
 	struct drm_plane_state state;
 
+	bool visible;
+
 	struct {
 		u32  handle;
 		u16  offset:12;
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.c b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
index 06d1696..4a685d7 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/wndw.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
@@ -244,26 +244,33 @@ nv50_wndw_atomic_check(struct drm_plane *plane, struct drm_plane_state *state)
 	struct nv50_wndw_atom *armw = nv50_wndw_atom(wndw->plane.state);
 	struct nv50_wndw_atom *asyw = nv50_wndw_atom(state);
 	struct nv50_head_atom *harm = NULL, *asyh = NULL;
-	bool varm = false, asyv = false, asym = false;
+	bool modeset = false;
 	int ret;
 
 	NV_ATOMIC(drm, "%s atomic_check\n", plane->name);
+
+	/* Fetch the assembly state for the head the window will belong to,
+	 * and determine whether the window will be visible.
+	 */
 	if (asyw->state.crtc) {
 		asyh = nv50_head_atom_get(asyw->state.state, asyw->state.crtc);
 		if (IS_ERR(asyh))
 			return PTR_ERR(asyh);
-		asym = drm_atomic_crtc_needs_modeset(&asyh->state);
-		asyv = asyh->state.active;
+		modeset = drm_atomic_crtc_needs_modeset(&asyh->state);
+		asyw->visible = asyh->state.active;
+	} else {
+		asyw->visible = false;
 	}
 
+	/* Fetch assembly state for the head the window used to belong to. */
 	if (armw->state.crtc) {
 		harm = nv50_head_atom_get(asyw->state.state, armw->state.crtc);
 		if (IS_ERR(harm))
 			return PTR_ERR(harm);
-		varm = harm->state.crtc->state->active;
 	}
 
-	if (asyv) {
+	/* Calculate new window state. */
+	if (asyw->visible) {
 		asyw->point.x = asyw->state.crtc_x;
 		asyw->point.y = asyw->state.crtc_y;
 		if (memcmp(&armw->point, &asyw->point, sizeof(asyw->point)))
@@ -273,18 +280,22 @@ nv50_wndw_atomic_check(struct drm_plane *plane, struct drm_plane_state *state)
 		if (ret)
 			return ret;
 	} else
-	if (varm) {
+	if (armw->visible) {
 		nv50_wndw_atomic_check_release(wndw, asyw, harm);
 	} else {
 		return 0;
 	}
 
-	if (!asyv || asym) {
+	/* Aside from the obvious case where the window is actively being
+	 * disabled, we might also need to temporarily disable the window
+	 * when performing certain modeset operations.
+	 */
+	if (!asyw->visible || modeset) {
 		asyw->clr.ntfy = armw->ntfy.handle != 0;
 		asyw->clr.sema = armw->sema.handle != 0;
 		if (wndw->func->image_clr)
 			asyw->clr.image = armw->image.handle[0] != 0;
-		asyw->set.lut = wndw->func->lut && asyv;
+		asyw->set.lut = wndw->func->lut && asyw->visible;
 	}
 
 	return 0;
-- 
cgit v1.1


From e349a05dc8faad6b27700383945a1783612cbae6 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50-: plane updates don't always require image_set()

When only the position of a window changes, there's no need to submit
an image update as well.

Will be required to support the overlays, and Volta windows.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/base507c.c |  1 -
 drivers/gpu/drm/nouveau/dispnv50/wndw.c     | 63 +++++++++++++++--------------
 2 files changed, 32 insertions(+), 32 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/dispnv50/base507c.c b/drivers/gpu/drm/nouveau/dispnv50/base507c.c
index 819403f..d8d3516 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/base507c.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/base507c.c
@@ -185,7 +185,6 @@ base507c_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
 	asyh->base.h = asyw->state.fb->height;
 
 	asyw->lut.enable = 1;
-	asyw->set.image = true;
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.c b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
index 4a685d7..0f6de60 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/wndw.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
@@ -190,7 +190,8 @@ nv50_wndw_atomic_check_acquire_rgb(struct nv50_wndw_atom *asyw)
 }
 
 static int
-nv50_wndw_atomic_check_acquire(struct nv50_wndw *wndw,
+nv50_wndw_atomic_check_acquire(struct nv50_wndw *wndw, bool modeset,
+			       struct nv50_wndw_atom *armw,
 			       struct nv50_wndw_atom *asyw,
 			       struct nv50_head_atom *asyh)
 {
@@ -200,40 +201,44 @@ nv50_wndw_atomic_check_acquire(struct nv50_wndw *wndw,
 
 	NV_ATOMIC(drm, "%s acquire\n", wndw->plane.name);
 
-	asyw->image.w = fb->base.width;
-	asyw->image.h = fb->base.height;
-	asyw->image.kind = fb->nvbo->kind;
+	if (asyw->state.fb != armw->state.fb || !armw->visible || modeset) {
+		asyw->image.w = fb->base.width;
+		asyw->image.h = fb->base.height;
+		asyw->image.kind = fb->nvbo->kind;
 
-	ret = nv50_wndw_atomic_check_acquire_rgb(asyw);
-	if (ret)
-		return ret;
-
-	if (asyw->image.kind) {
-		asyw->image.layout = 0;
-		if (drm->client.device.info.chipset >= 0xc0)
-			asyw->image.block = fb->nvbo->mode >> 4;
-		else
-			asyw->image.block = fb->nvbo->mode;
-		asyw->image.pitch[0] = (fb->base.pitches[0] / 4) << 4;
-	} else {
-		asyw->image.layout = 1;
-		asyw->image.block  = 0;
-		asyw->image.pitch[0] = fb->base.pitches[0];
-	}
+		ret = nv50_wndw_atomic_check_acquire_rgb(asyw);
+		if (ret)
+			return ret;
 
-	ret = wndw->func->acquire(wndw, asyw, asyh);
-	if (ret)
-		return ret;
+		if (asyw->image.kind) {
+			asyw->image.layout = 0;
+			if (drm->client.device.info.chipset >= 0xc0)
+				asyw->image.block = fb->nvbo->mode >> 4;
+			else
+				asyw->image.block = fb->nvbo->mode;
+			asyw->image.pitch[0] = (fb->base.pitches[0] / 4) << 4;
+		} else {
+			asyw->image.layout = 1;
+			asyw->image.block  = 0;
+			asyw->image.pitch[0] = fb->base.pitches[0];
+		}
 
-	if (asyw->set.image) {
 		if (!(asyh->state.pageflip_flags & DRM_MODE_PAGE_FLIP_ASYNC))
 			asyw->image.interval = 1;
 		else
 			asyw->image.interval = 0;
 		asyw->image.mode = asyw->image.interval ? 0 : 1;
+		asyw->set.image = wndw->func->image_set != NULL;
 	}
 
-	return 0;
+	if (wndw->immd) {
+		asyw->point.x = asyw->state.crtc_x;
+		asyw->point.y = asyw->state.crtc_y;
+		if (memcmp(&armw->point, &asyw->point, sizeof(asyw->point)))
+			asyw->set.point = true;
+	}
+
+	return wndw->func->acquire(wndw, asyw, asyh);
 }
 
 int
@@ -271,12 +276,8 @@ nv50_wndw_atomic_check(struct drm_plane *plane, struct drm_plane_state *state)
 
 	/* Calculate new window state. */
 	if (asyw->visible) {
-		asyw->point.x = asyw->state.crtc_x;
-		asyw->point.y = asyw->state.crtc_y;
-		if (memcmp(&armw->point, &asyw->point, sizeof(asyw->point)))
-			asyw->set.point = true;
-
-		ret = nv50_wndw_atomic_check_acquire(wndw, asyw, asyh);
+		ret = nv50_wndw_atomic_check_acquire(wndw, modeset,
+						     armw, asyw, asyh);
 		if (ret)
 			return ret;
 	} else
-- 
cgit v1.1


From 119608a7f3f1ef899f1f98d05306340b92834836 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50-: handle degamma LUT from window channels

Required to eventually support DRM colour management APIs, and to
support Volta.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/Kbuild     |   1 +
 drivers/gpu/drm/nouveau/dispnv50/atom.h     |  25 +++--
 drivers/gpu/drm/nouveau/dispnv50/base.h     |   3 +-
 drivers/gpu/drm/nouveau/dispnv50/base507c.c |  44 +++++---
 drivers/gpu/drm/nouveau/dispnv50/base827c.c |   4 +-
 drivers/gpu/drm/nouveau/dispnv50/base907c.c |  43 +++++++-
 drivers/gpu/drm/nouveau/dispnv50/disp.c     |  11 ++
 drivers/gpu/drm/nouveau/dispnv50/head.c     | 160 +++++++++++-----------------
 drivers/gpu/drm/nouveau/dispnv50/head.h     |  18 ++--
 drivers/gpu/drm/nouveau/dispnv50/head507d.c |  29 +++--
 drivers/gpu/drm/nouveau/dispnv50/head827d.c |  17 +--
 drivers/gpu/drm/nouveau/dispnv50/head907d.c |  33 +++---
 drivers/gpu/drm/nouveau/dispnv50/head917d.c |  34 +++++-
 drivers/gpu/drm/nouveau/dispnv50/lut.c      |  95 +++++++++++++++++
 drivers/gpu/drm/nouveau/dispnv50/lut.h      |  15 +++
 drivers/gpu/drm/nouveau/dispnv50/wndw.c     |  93 +++++++++++++++-
 drivers/gpu/drm/nouveau/dispnv50/wndw.h     |   8 +-
 17 files changed, 462 insertions(+), 171 deletions(-)
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/lut.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/lut.h

diff --git a/drivers/gpu/drm/nouveau/dispnv50/Kbuild b/drivers/gpu/drm/nouveau/dispnv50/Kbuild
index 3e53484..d074bb8 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/Kbuild
+++ b/drivers/gpu/drm/nouveau/dispnv50/Kbuild
@@ -1,4 +1,5 @@
 nouveau-y += dispnv50/disp.o
+nouveau-y += dispnv50/lut.o
 
 nouveau-y += dispnv50/core.o
 nouveau-y += dispnv50/core507d.o
diff --git a/drivers/gpu/drm/nouveau/dispnv50/atom.h b/drivers/gpu/drm/nouveau/dispnv50/atom.h
index fefb9ca..3e9e883 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/atom.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/atom.h
@@ -17,6 +17,11 @@ struct nv50_head_atom {
 	struct drm_crtc_state state;
 
 	struct {
+		u32 mask;
+		u32 olut;
+	} wndw;
+
+	struct {
 		u16 iW;
 		u16 iH;
 		u16 oW;
@@ -47,8 +52,9 @@ struct nv50_head_atom {
 		bool visible;
 		u32 handle;
 		u64 offset:40;
-		u8  mode:4;
-	} ilut;
+		u8 buffer:1;
+		u8 mode:4;
+	} olut;
 
 	struct {
 		bool visible;
@@ -107,7 +113,7 @@ struct nv50_head_atom {
 
 	union nv50_head_atom_mask {
 		struct {
-			bool ilut:1;
+			bool olut:1;
 			bool core:1;
 			bool curs:1;
 			bool view:1;
@@ -136,6 +142,7 @@ nv50_head_atom_get(struct drm_atomic_state *state, struct drm_crtc *crtc)
 struct nv50_wndw_atom {
 	struct drm_plane_state state;
 
+	struct drm_property_blob *ilut;
 	bool visible;
 
 	struct {
@@ -152,8 +159,14 @@ struct nv50_wndw_atom {
 	} sema;
 
 	struct {
-		u8 enable:2;
-	} lut;
+		u32 handle;
+		struct {
+			u64 offset:40;
+			u8  buffer:1;
+			u8  enable:2;
+			u8  mode:4;
+		} i;
+	} xlut;
 
 	struct {
 		u8  mode:2;
@@ -180,8 +193,8 @@ struct nv50_wndw_atom {
 		struct {
 			bool ntfy:1;
 			bool sema:1;
+			bool xlut:1;
 			bool image:1;
-			bool lut:1;
 			bool point:1;
 		};
 		u8 mask;
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base.h b/drivers/gpu/drm/nouveau/dispnv50/base.h
index 71fc103..87ec839 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/base.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/base.h
@@ -17,8 +17,9 @@ void base507c_ntfy_reset(struct nouveau_bo *, u32);
 void base507c_ntfy_set(struct nv50_wndw *, struct nv50_wndw_atom *);
 void base507c_ntfy_clr(struct nv50_wndw *);
 int base507c_ntfy_wait_begun(struct nouveau_bo *, u32, struct nvif_device *);
+void base507c_xlut_set(struct nv50_wndw *, struct nv50_wndw_atom *);
+void base507c_xlut_clr(struct nv50_wndw *);
 void base507c_image_clr(struct nv50_wndw *);
-void base507c_lut(struct nv50_wndw *, struct nv50_wndw_atom *);
 void base507c_update(struct nv50_wndw *, u32 *);
 
 int base827c_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base507c.c b/drivers/gpu/drm/nouveau/dispnv50/base507c.c
index d8d3516..5d664d7 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/base507c.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/base507c.c
@@ -40,17 +40,6 @@ base507c_update(struct nv50_wndw *wndw, u32 *interlock)
 }
 
 void
-base507c_lut(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
-{
-	u32 *push;
-	if ((push = evo_wait(&wndw->wndw, 2))) {
-		evo_mthd(push, 0x00e0, 1);
-		evo_data(push, asyw->lut.enable << 30);
-		evo_kick(push, &wndw->wndw);
-	}
-}
-
-void
 base507c_image_clr(struct nv50_wndw *wndw)
 {
 	u32 *push;
@@ -86,6 +75,28 @@ base507c_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 	}
 }
 
+void
+base507c_xlut_clr(struct nv50_wndw *wndw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 2))) {
+		evo_mthd(push, 0x00e0, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+void
+base507c_xlut_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 2))) {
+		evo_mthd(push, 0x00e0, 1);
+		evo_data(push, 0x40000000);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
 int
 base507c_ntfy_wait_begun(struct nouveau_bo *bo, u32 offset,
 			 struct nvif_device *device)
@@ -177,14 +188,17 @@ base507c_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
 	if (ret)
 		return ret;
 
+	if (!wndw->func->ilut) {
+		if ((asyh->base.cpp != 1) ^ (fb->format->cpp[0] != 1))
+			asyh->state.color_mgmt_changed = true;
+	}
+
 	asyh->base.depth = fb->format->depth;
 	asyh->base.cpp = fb->format->cpp[0];
 	asyh->base.x = asyw->state.src.x1 >> 16;
 	asyh->base.y = asyw->state.src.y1 >> 16;
 	asyh->base.w = asyw->state.fb->width;
 	asyh->base.h = asyw->state.fb->height;
-
-	asyw->lut.enable = 1;
 	return 0;
 }
 
@@ -213,9 +227,11 @@ base507c = {
 	.ntfy_set = base507c_ntfy_set,
 	.ntfy_clr = base507c_ntfy_clr,
 	.ntfy_wait_begun = base507c_ntfy_wait_begun,
+	.olut_core = 1,
+	.xlut_set = base507c_xlut_set,
+	.xlut_clr = base507c_xlut_clr,
 	.image_set = base507c_image_set,
 	.image_clr = base507c_image_clr,
-	.lut = base507c_lut,
 	.update = base507c_update,
 };
 
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base827c.c b/drivers/gpu/drm/nouveau/dispnv50/base827c.c
index 240a640..d886858 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/base827c.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/base827c.c
@@ -53,9 +53,11 @@ base827c = {
 	.ntfy_set = base507c_ntfy_set,
 	.ntfy_clr = base507c_ntfy_clr,
 	.ntfy_wait_begun = base507c_ntfy_wait_begun,
+	.olut_core = 1,
+	.xlut_set = base507c_xlut_set,
+	.xlut_clr = base507c_xlut_clr,
 	.image_set = base827c_image_set,
 	.image_clr = base507c_image_clr,
-	.lut = base507c_lut,
 	.update = base507c_update,
 };
 
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base907c.c b/drivers/gpu/drm/nouveau/dispnv50/base907c.c
index 6c32a4e..2643592 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/base907c.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/base907c.c
@@ -43,6 +43,44 @@ base907c_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 	}
 }
 
+static void
+base907c_xlut_clr(struct nv50_wndw *wndw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 6))) {
+		evo_mthd(push, 0x00e0, 1);
+		evo_data(push, 0x00000000);
+		evo_mthd(push, 0x00e8, 1);
+		evo_data(push, 0x00000000);
+		evo_mthd(push, 0x00fc, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+base907c_xlut_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 6))) {
+		evo_mthd(push, 0x00e0, 3);
+		evo_data(push, asyw->xlut.i.enable << 30 |
+			       asyw->xlut.i.mode << 24);
+		evo_data(push, asyw->xlut.i.offset >> 8);
+		evo_data(push, 0x40000000);
+		evo_mthd(push, 0x00fc, 1);
+		evo_data(push, asyw->xlut.handle);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+base907c_ilut(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	asyw->xlut.i.mode = 7;
+	asyw->xlut.i.enable = 2;
+}
+
 static const struct nv50_wndw_func
 base907c = {
 	.acquire = base507c_acquire,
@@ -53,9 +91,12 @@ base907c = {
 	.ntfy_set = base507c_ntfy_set,
 	.ntfy_clr = base507c_ntfy_clr,
 	.ntfy_wait_begun = base507c_ntfy_wait_begun,
+	.ilut = base907c_ilut,
+	.olut_core = true,
+	.xlut_set = base907c_xlut_set,
+	.xlut_clr = base907c_xlut_clr,
 	.image_set = base907c_image_set,
 	.image_clr = base507c_image_clr,
-	.lut = base507c_lut,
 	.update = base507c_update,
 };
 
diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c
index 0f20200..6c860e8 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/disp.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c
@@ -1971,8 +1971,19 @@ nv50_disp_atomic_check(struct drm_device *dev, struct drm_atomic_state *state)
 	struct nv50_atom *atom = nv50_atom(state);
 	struct drm_connector_state *old_connector_state, *new_connector_state;
 	struct drm_connector *connector;
+	struct drm_crtc_state *new_crtc_state;
+	struct drm_crtc *crtc;
 	int ret, i;
 
+	/* We need to handle colour management on a per-plane basis. */
+	for_each_new_crtc_in_state(state, crtc, new_crtc_state, i) {
+		if (new_crtc_state->color_mgmt_changed) {
+			ret = drm_atomic_add_affected_planes(state, crtc);
+			if (ret)
+				return ret;
+		}
+	}
+
 	ret = drm_atomic_helper_check(dev, state);
 	if (ret)
 		return ret;
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head.c b/drivers/gpu/drm/nouveau/dispnv50/head.c
index 2eb7fdb..ca83006 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/head.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/head.c
@@ -30,56 +30,6 @@
 #include <drm/drm_atomic_helper.h>
 #include <drm/drm_crtc_helper.h>
 #include "nouveau_connector.h"
-#include "nouveau_bo.h"
-
-static void
-nv50_head_lut_load(struct drm_property_blob *blob, int mode,
-		   struct nouveau_bo *nvbo)
-{
-	struct drm_color_lut *in = (struct drm_color_lut *)blob->data;
-	void __iomem *lut = (u8 *)nvbo_kmap_obj_iovirtual(nvbo);
-	const int size = blob->length / sizeof(*in);
-	int bits, shift, i;
-	u16 zero, r, g, b;
-
-	/* This can't happen.. But it shuts the compiler up. */
-	if (WARN_ON(size != 256))
-		return;
-
-	switch (mode) {
-	case 0: /* LORES. */
-	case 1: /* HIRES. */
-		bits = 11;
-		shift = 3;
-		zero = 0x0000;
-		break;
-	case 7: /* INTERPOLATE_257_UNITY_RANGE. */
-		bits = 14;
-		shift = 0;
-		zero = 0x6000;
-		break;
-	default:
-		WARN_ON(1);
-		return;
-	}
-
-	for (i = 0; i < size; i++) {
-		r = (drm_color_lut_extract(in[i].  red, bits) + zero) << shift;
-		g = (drm_color_lut_extract(in[i].green, bits) + zero) << shift;
-		b = (drm_color_lut_extract(in[i]. blue, bits) + zero) << shift;
-		writew(r, lut + (i * 0x08) + 0);
-		writew(g, lut + (i * 0x08) + 2);
-		writew(b, lut + (i * 0x08) + 4);
-	}
-
-	/* INTERPOLATE modes require a "next" entry to interpolate with,
-	 * so we replicate the last entry to deal with this for now.
-	 */
-	writew(r, lut + (i * 0x08) + 0);
-	writew(g, lut + (i * 0x08) + 2);
-	writew(b, lut + (i * 0x08) + 4);
-}
-
 void
 nv50_head_flush_clr(struct nv50_head *head,
 		    struct nv50_head_atom *asyh, bool flush)
@@ -87,7 +37,7 @@ nv50_head_flush_clr(struct nv50_head *head,
 	union nv50_head_atom_mask clr = {
 		.mask = asyh->clr.mask & ~(flush ? 0 : asyh->set.mask),
 	};
-	if (clr.ilut) head->func->ilut_clr(head);
+	if (clr.olut) head->func->olut_clr(head);
 	if (clr.core) head->func->core_clr(head);
 	if (clr.curs) head->func->curs_clr(head);
 }
@@ -97,16 +47,14 @@ nv50_head_flush_set(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
 	if (asyh->set.view   ) head->func->view    (head, asyh);
 	if (asyh->set.mode   ) head->func->mode    (head, asyh);
-	if (asyh->set.ilut   ) {
-		struct nouveau_bo *nvbo = head->ilut.nvbo[head->ilut.next];
-		struct drm_property_blob *blob = asyh->state.gamma_lut;
-		if (blob)
-			nv50_head_lut_load(blob, asyh->ilut.mode, nvbo);
-		asyh->ilut.offset = nvbo->bo.offset;
-		head->ilut.next ^= 1;
-		head->func->ilut_set(head, asyh);
-	}
 	if (asyh->set.core   ) head->func->core_set(head, asyh);
+	if (asyh->set.olut   ) {
+		asyh->olut.offset = nv50_lut_load(&head->olut,
+						  asyh->olut.mode <= 1,
+						  asyh->olut.buffer,
+						  asyh->state.gamma_lut);
+		head->func->olut_set(head, asyh);
+	}
 	if (asyh->set.curs   ) head->func->curs_set(head, asyh);
 	if (asyh->set.base   ) head->func->base    (head, asyh);
 	if (asyh->set.ovly   ) head->func->ovly    (head, asyh);
@@ -240,35 +188,37 @@ nv50_head_atomic_check_view(struct nv50_head_atom *armh,
 	asyh->set.view = true;
 }
 
-static void
+static int
 nv50_head_atomic_check_lut(struct nv50_head *head,
-			   struct nv50_head_atom *armh,
 			   struct nv50_head_atom *asyh)
 {
 	struct nv50_disp *disp = nv50_disp(head->base.base.dev);
-
-	/* An I8 surface without an input LUT makes no sense, and
-	 * EVO will throw an error if you try.
-	 *
-	 * Legacy clients actually cause this due to the order in
-	 * which they call ioctls, so we will enable the LUT with
-	 * whatever contents the buffer already contains to avoid
-	 * triggering the error check.
-	 */
-	if (!asyh->state.gamma_lut && asyh->base.cpp != 1) {
-		asyh->ilut.handle = 0;
-		asyh->clr.ilut = armh->ilut.visible;
-		return;
+	struct drm_property_blob *olut = asyh->state.gamma_lut;
+
+	/* Determine whether core output LUT should be enabled. */
+	if (olut) {
+		/* Check if any window(s) have stolen the core output LUT
+		 * to as an input LUT for legacy gamma + I8 colour format.
+		 */
+		if (asyh->wndw.olut) {
+			/* If any window has stolen the core output LUT,
+			 * all of them must.
+			 */
+			if (asyh->wndw.olut != asyh->wndw.mask)
+				return -EINVAL;
+			olut = NULL;
+		}
 	}
 
-	if (disp->disp->object.oclass < GF110_DISP) {
-		asyh->ilut.mode = (asyh->base.cpp == 1) ? 0 : 1;
-		asyh->set.ilut = true;
-	} else {
-		asyh->ilut.mode = 7;
-		asyh->set.ilut = asyh->state.color_mgmt_changed;
+	if (!olut) {
+		asyh->olut.handle = 0;
+		return 0;
 	}
-	asyh->ilut.handle = disp->core->chan.vram.handle;
+
+	asyh->olut.handle = disp->core->chan.vram.handle;
+	asyh->olut.buffer = !asyh->olut.buffer;
+	head->func->olut(head, asyh);
+	return 0;
 }
 
 static void
@@ -360,9 +310,13 @@ nv50_head_atomic_check(struct drm_crtc *crtc, struct drm_crtc_state *state)
 			nv50_head_atomic_check_mode(head, asyh);
 
 		if (asyh->state.color_mgmt_changed ||
-		    asyh->base.cpp != armh->base.cpp)
-			nv50_head_atomic_check_lut(head, armh, asyh);
-		asyh->ilut.visible = asyh->ilut.handle != 0;
+		    memcmp(&armh->wndw, &asyh->wndw, sizeof(asyh->wndw))) {
+			int ret = nv50_head_atomic_check_lut(head, asyh);
+			if (ret)
+				return ret;
+
+			asyh->olut.visible = asyh->olut.handle != 0;
+		}
 
 		if (asyc) {
 			if (asyc->set.scaler)
@@ -373,13 +327,16 @@ nv50_head_atomic_check(struct drm_crtc *crtc, struct drm_crtc_state *state)
 				nv50_head_atomic_check_procamp(armh, asyh, asyc);
 		}
 
-		if (head->func->core_calc)
+		if (head->func->core_calc) {
 			head->func->core_calc(head, asyh);
+			if (!asyh->core.visible)
+				asyh->olut.visible = false;
+		}
 
 		asyh->set.base = armh->base.cpp != asyh->base.cpp;
 		asyh->set.ovly = armh->ovly.cpp != asyh->ovly.cpp;
 	} else {
-		asyh->ilut.visible = false;
+		asyh->olut.visible = false;
 		asyh->core.visible = false;
 		asyh->curs.visible = false;
 		asyh->base.cpp = 0;
@@ -402,11 +359,19 @@ nv50_head_atomic_check(struct drm_crtc *crtc, struct drm_crtc_state *state)
 		if (armh->curs.visible) {
 			asyh->clr.curs = true;
 		}
+
+		if (asyh->olut.visible) {
+			if (memcmp(&armh->olut, &asyh->olut, sizeof(asyh->olut)))
+				asyh->set.olut = true;
+		} else
+		if (armh->olut.visible) {
+			asyh->clr.olut = true;
+		}
 	} else {
-		asyh->clr.ilut = armh->ilut.visible;
+		asyh->clr.olut = armh->olut.visible;
 		asyh->clr.core = armh->core.visible;
 		asyh->clr.curs = armh->curs.visible;
-		asyh->set.ilut = asyh->ilut.visible;
+		asyh->set.olut = asyh->olut.visible;
 		asyh->set.core = asyh->core.visible;
 		asyh->set.curs = asyh->curs.visible;
 	}
@@ -438,9 +403,10 @@ nv50_head_atomic_duplicate_state(struct drm_crtc *crtc)
 	if (!(asyh = kmalloc(sizeof(*asyh), GFP_KERNEL)))
 		return NULL;
 	__drm_atomic_helper_crtc_duplicate_state(crtc, &asyh->state);
+	asyh->wndw = armh->wndw;
 	asyh->view = armh->view;
 	asyh->mode = armh->mode;
-	asyh->ilut = armh->ilut;
+	asyh->olut = armh->olut;
 	asyh->core = armh->core;
 	asyh->curs = armh->curs;
 	asyh->base = armh->base;
@@ -477,11 +443,7 @@ static void
 nv50_head_destroy(struct drm_crtc *crtc)
 {
 	struct nv50_head *head = nv50_head(crtc);
-	int i;
-
-	for (i = 0; i < ARRAY_SIZE(head->ilut.nvbo); i++)
-		nouveau_bo_unmap_unpin_unref(&head->ilut.nvbo[i]);
-
+	nv50_lut_fini(&head->olut);
 	drm_crtc_cleanup(crtc);
 	kfree(head);
 }
@@ -505,7 +467,7 @@ nv50_head_create(struct drm_device *dev, int index)
 	struct nv50_head *head;
 	struct nv50_wndw *curs, *wndw;
 	struct drm_crtc *crtc;
-	int ret, i;
+	int ret;
 
 	head = kzalloc(sizeof(*head), GFP_KERNEL);
 	if (!head)
@@ -527,10 +489,8 @@ nv50_head_create(struct drm_device *dev, int index)
 	drm_crtc_helper_add(crtc, &nv50_head_help);
 	drm_mode_crtc_set_gamma_size(crtc, 256);
 
-	for (i = 0; i < ARRAY_SIZE(head->ilut.nvbo); i++) {
-		ret = nouveau_bo_new_pin_map(&drm->client, 1025 * 8, 0x100,
-					     TTM_PL_FLAG_VRAM,
-					     &head->ilut.nvbo[i]);
+	if (head->func->olut_set) {
+		ret = nv50_lut_init(disp, &drm->client.mmu, &head->olut);
 		if (ret)
 			goto out;
 	}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head.h b/drivers/gpu/drm/nouveau/dispnv50/head.h
index d00cebd..0802271 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/head.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/head.h
@@ -3,16 +3,14 @@
 #define nv50_head(c) container_of((c), struct nv50_head, base.base)
 #include "disp.h"
 #include "atom.h"
+#include "lut.h"
 
 #include "nouveau_crtc.h"
 
 struct nv50_head {
 	const struct nv50_head_func *func;
 	struct nouveau_crtc base;
-	struct {
-		struct nouveau_bo *nvbo[2];
-		int next;
-	} ilut;
+	struct nv50_lut olut;
 };
 
 int nv50_head_create(struct drm_device *, int index);
@@ -22,8 +20,9 @@ void nv50_head_flush_clr(struct nv50_head *, struct nv50_head_atom *, bool y);
 struct nv50_head_func {
 	void (*view)(struct nv50_head *, struct nv50_head_atom *);
 	void (*mode)(struct nv50_head *, struct nv50_head_atom *);
-	void (*ilut_set)(struct nv50_head *, struct nv50_head_atom *);
-	void (*ilut_clr)(struct nv50_head *);
+	void (*olut)(struct nv50_head *, struct nv50_head_atom *);
+	void (*olut_set)(struct nv50_head *, struct nv50_head_atom *);
+	void (*olut_clr)(struct nv50_head *);
 	void (*core_calc)(struct nv50_head *, struct nv50_head_atom *);
 	void (*core_set)(struct nv50_head *, struct nv50_head_atom *);
 	void (*core_clr)(struct nv50_head *);
@@ -39,6 +38,7 @@ struct nv50_head_func {
 extern const struct nv50_head_func head507d;
 void head507d_view(struct nv50_head *, struct nv50_head_atom *);
 void head507d_mode(struct nv50_head *, struct nv50_head_atom *);
+void head507d_olut(struct nv50_head *, struct nv50_head_atom *);
 void head507d_core_calc(struct nv50_head *, struct nv50_head_atom *);
 void head507d_core_clr(struct nv50_head *);
 void head507d_base(struct nv50_head *, struct nv50_head_atom *);
@@ -51,13 +51,13 @@ extern const struct nv50_head_func head827d;
 extern const struct nv50_head_func head907d;
 void head907d_view(struct nv50_head *, struct nv50_head_atom *);
 void head907d_mode(struct nv50_head *, struct nv50_head_atom *);
-void head907d_ilut_set(struct nv50_head *, struct nv50_head_atom *);
-void head907d_ilut_clr(struct nv50_head *);
+void head907d_olut(struct nv50_head *, struct nv50_head_atom *);
+void head907d_olut_set(struct nv50_head *, struct nv50_head_atom *);
+void head907d_olut_clr(struct nv50_head *);
 void head907d_core_set(struct nv50_head *, struct nv50_head_atom *);
 void head907d_core_clr(struct nv50_head *);
 void head907d_curs_set(struct nv50_head *, struct nv50_head_atom *);
 void head907d_curs_clr(struct nv50_head *);
-void head907d_base(struct nv50_head *, struct nv50_head_atom *);
 void head907d_ovly(struct nv50_head *, struct nv50_head_atom *);
 void head907d_procamp(struct nv50_head *, struct nv50_head_atom *);
 void head907d_or(struct nv50_head *, struct nv50_head_atom *);
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head507d.c b/drivers/gpu/drm/nouveau/dispnv50/head507d.c
index 5f06fa1..75575c33 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/head507d.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/head507d.c
@@ -165,6 +165,7 @@ head507d_core_set(struct nv50_head *head, struct nv50_head_atom *asyh)
 		 * without also updating HeadSetOffsetCursor.
 		 */
 		asyh->set.curs = asyh->curs.visible;
+		asyh->set.olut = asyh->olut.handle != 0;
 	}
 }
 
@@ -178,8 +179,8 @@ head507d_core_calc(struct nv50_head *head, struct nv50_head_atom *asyh)
 		asyh->core.w = asyh->base.w;
 		asyh->core.h = asyh->base.h;
 	} else
-	if ((asyh->core.visible = asyh->curs.visible) ||
-	    (asyh->core.visible = asyh->ilut.visible)) {
+	if ((asyh->core.visible = (asyh->ovly.cpp != 0)) ||
+	    (asyh->core.visible = asyh->curs.visible)) {
 		/*XXX: We need to either find some way of having the
 		 *     primary base layer appear black, while still
 		 *     being able to display the other layers, or we
@@ -200,31 +201,40 @@ head507d_core_calc(struct nv50_head *head, struct nv50_head_atom *asyh)
 }
 
 static void
-head507d_ilut_clr(struct nv50_head *head)
+head507d_olut_clr(struct nv50_head *head)
 {
 	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
 	if ((push = evo_wait(core, 2))) {
 		evo_mthd(push, 0x0840 + (head->base.index * 0x400), 1);
-		evo_data(push, 0x40000000);
+		evo_data(push, 0x00000000);
 		evo_kick(push, core);
 	}
 }
 
 static void
-head507d_ilut_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+head507d_olut_set(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
 	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
 	if ((push = evo_wait(core, 3))) {
 		evo_mthd(push, 0x0840 + (head->base.index * 0x400), 2);
-		evo_data(push, 0x80000000 | asyh->ilut.mode << 30);
-		evo_data(push, asyh->ilut.offset >> 8);
+		evo_data(push, 0x80000000 | asyh->olut.mode << 30);
+		evo_data(push, asyh->olut.offset >> 8);
 		evo_kick(push, core);
 	}
 }
 
 void
+head507d_olut(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	if (asyh->base.cpp == 1)
+		asyh->olut.mode = 0;
+	else
+		asyh->olut.mode = 1;
+}
+
+void
 head507d_mode(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
 	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
@@ -269,8 +279,9 @@ const struct nv50_head_func
 head507d = {
 	.view = head507d_view,
 	.mode = head507d_mode,
-	.ilut_set = head507d_ilut_set,
-	.ilut_clr = head507d_ilut_clr,
+	.olut = head507d_olut,
+	.olut_set = head507d_olut_set,
+	.olut_clr = head507d_olut_clr,
 	.core_calc = head507d_core_calc,
 	.core_set = head507d_core_set,
 	.core_clr = head507d_core_clr,
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head827d.c b/drivers/gpu/drm/nouveau/dispnv50/head827d.c
index 84ce595..ddc143b 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/head827d.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/head827d.c
@@ -74,13 +74,13 @@ head827d_core_set(struct nv50_head *head, struct nv50_head_atom *asyh)
 }
 
 static void
-head827d_ilut_clr(struct nv50_head *head)
+head827d_olut_clr(struct nv50_head *head)
 {
 	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
 	if ((push = evo_wait(core, 4))) {
 		evo_mthd(push, 0x0840 + (head->base.index * 0x400), 1);
-		evo_data(push, 0x40000000);
+		evo_data(push, 0x00000000);
 		evo_mthd(push, 0x085c + (head->base.index * 0x400), 1);
 		evo_data(push, 0x00000000);
 		evo_kick(push, core);
@@ -88,16 +88,16 @@ head827d_ilut_clr(struct nv50_head *head)
 }
 
 static void
-head827d_ilut_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+head827d_olut_set(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
 	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
 	if ((push = evo_wait(core, 5))) {
 		evo_mthd(push, 0x0840 + (head->base.index * 0x400), 2);
-		evo_data(push, 0x80000000 | asyh->ilut.mode << 30);
-		evo_data(push, asyh->ilut.offset >> 8);
+		evo_data(push, 0x80000000 | asyh->olut.mode << 30);
+		evo_data(push, asyh->olut.offset >> 8);
 		evo_mthd(push, 0x085c + (head->base.index * 0x400), 1);
-		evo_data(push, asyh->ilut.handle);
+		evo_data(push, asyh->olut.handle);
 		evo_kick(push, core);
 	}
 }
@@ -106,8 +106,9 @@ const struct nv50_head_func
 head827d = {
 	.view = head507d_view,
 	.mode = head507d_mode,
-	.ilut_set = head827d_ilut_set,
-	.ilut_clr = head827d_ilut_clr,
+	.olut = head507d_olut,
+	.olut_set = head827d_olut_set,
+	.olut_clr = head827d_olut_clr,
 	.core_calc = head507d_core_calc,
 	.core_set = head827d_core_set,
 	.core_clr = head507d_core_clr,
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head907d.c b/drivers/gpu/drm/nouveau/dispnv50/head907d.c
index 0035ecc..0fa0159 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/head907d.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/head907d.c
@@ -91,7 +91,7 @@ head907d_ovly(struct nv50_head *head, struct nv50_head_atom *asyh)
 	}
 }
 
-void
+static void
 head907d_base(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
 	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
@@ -182,13 +182,13 @@ head907d_core_set(struct nv50_head *head, struct nv50_head_atom *asyh)
 }
 
 void
-head907d_ilut_clr(struct nv50_head *head)
+head907d_olut_clr(struct nv50_head *head)
 {
 	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
 	if ((push = evo_wait(core, 4))) {
-		evo_mthd(push, 0x0440 + (head->base.index * 0x300), 1);
-		evo_data(push, 0x03000000);
+		evo_mthd(push, 0x0448 + (head->base.index * 0x300), 1);
+		evo_data(push, 0x00000000);
 		evo_mthd(push, 0x045c + (head->base.index * 0x300), 1);
 		evo_data(push, 0x00000000);
 		evo_kick(push, core);
@@ -196,23 +196,27 @@ head907d_ilut_clr(struct nv50_head *head)
 }
 
 void
-head907d_ilut_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+head907d_olut_set(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
 	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
 	u32 *push;
-	if ((push = evo_wait(core, 7))) {
-		evo_mthd(push, 0x0440 + (head->base.index * 0x300), 4);
-		evo_data(push, 0x80000000 | asyh->ilut.mode << 24);
-		evo_data(push, asyh->ilut.offset >> 8);
-		evo_data(push, 0x00000000);
-		evo_data(push, 0x00000000);
+	if ((push = evo_wait(core, 5))) {
+		evo_mthd(push, 0x0448 + (head->base.index * 0x300), 2);
+		evo_data(push, 0x80000000 | asyh->olut.mode << 24);
+		evo_data(push, asyh->olut.offset >> 8);
 		evo_mthd(push, 0x045c + (head->base.index * 0x300), 1);
-		evo_data(push, asyh->ilut.handle);
+		evo_data(push, asyh->olut.handle);
 		evo_kick(push, core);
 	}
 }
 
 void
+head907d_olut(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	asyh->olut.mode = 7;
+}
+
+void
 head907d_mode(struct nv50_head *head, struct nv50_head_atom *asyh)
 {
 	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
@@ -259,8 +263,9 @@ const struct nv50_head_func
 head907d = {
 	.view = head907d_view,
 	.mode = head907d_mode,
-	.ilut_set = head907d_ilut_set,
-	.ilut_clr = head907d_ilut_clr,
+	.olut = head907d_olut,
+	.olut_set = head907d_olut_set,
+	.olut_clr = head907d_olut_clr,
 	.core_calc = head507d_core_calc,
 	.core_set = head907d_core_set,
 	.core_clr = head907d_core_clr,
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head917d.c b/drivers/gpu/drm/nouveau/dispnv50/head917d.c
index 5341ea3..5f65451 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/head917d.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/head917d.c
@@ -36,18 +36,46 @@ head917d_dither(struct nv50_head *head, struct nv50_head_atom *asyh)
 	}
 }
 
+static void
+head917d_base(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 bounds = 0;
+	u32 *push;
+
+	if (asyh->base.cpp) {
+		switch (asyh->base.cpp) {
+		case 8: bounds |= 0x00000500; break;
+		case 4: bounds |= 0x00000300; break;
+		case 2: bounds |= 0x00000100; break;
+		case 1: bounds |= 0x00000000; break;
+		default:
+			WARN_ON(1);
+			break;
+		}
+		bounds |= 0x00020001;
+	}
+
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x04d0 + head->base.index * 0x300, 1);
+		evo_data(push, bounds);
+		evo_kick(push, core);
+	}
+}
+
 const struct nv50_head_func
 head917d = {
 	.view = head907d_view,
 	.mode = head907d_mode,
-	.ilut_set = head907d_ilut_set,
-	.ilut_clr = head907d_ilut_clr,
+	.olut = head907d_olut,
+	.olut_set = head907d_olut_set,
+	.olut_clr = head907d_olut_clr,
 	.core_calc = head507d_core_calc,
 	.core_set = head907d_core_set,
 	.core_clr = head907d_core_clr,
 	.curs_set = head907d_curs_set,
 	.curs_clr = head907d_curs_clr,
-	.base = head907d_base,
+	.base = head917d_base,
 	.ovly = head907d_ovly,
 	.dither = head917d_dither,
 	.procamp = head907d_procamp,
diff --git a/drivers/gpu/drm/nouveau/dispnv50/lut.c b/drivers/gpu/drm/nouveau/dispnv50/lut.c
new file mode 100644
index 0000000..a6b96ae2
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/lut.c
@@ -0,0 +1,95 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "lut.h"
+#include "disp.h"
+
+#include <drm/drm_color_mgmt.h>
+#include <drm/drm_mode.h>
+#include <drm/drm_property.h>
+
+#include <nvif/class.h>
+
+u32
+nv50_lut_load(struct nv50_lut *lut, bool legacy, int buffer,
+	      struct drm_property_blob *blob)
+{
+	struct drm_color_lut *in = (struct drm_color_lut *)blob->data;
+	void __iomem *mem = lut->mem[buffer].object.map.ptr;
+	const int size = blob->length / sizeof(*in);
+	int bits, shift, i;
+	u16 zero, r, g, b;
+	u32 addr = lut->mem[buffer].addr;
+
+	/* This can't happen.. But it shuts the compiler up. */
+	if (WARN_ON(size != 256))
+		return 0;
+
+	if (legacy) {
+		bits = 11;
+		shift = 3;
+		zero = 0x0000;
+	} else {
+		bits = 14;
+		shift = 0;
+		zero = 0x6000;
+	}
+
+	for (i = 0; i < size; i++) {
+		r = (drm_color_lut_extract(in[i].  red, bits) + zero) << shift;
+		g = (drm_color_lut_extract(in[i].green, bits) + zero) << shift;
+		b = (drm_color_lut_extract(in[i]. blue, bits) + zero) << shift;
+		writew(r, mem + (i * 0x08) + 0);
+		writew(g, mem + (i * 0x08) + 2);
+		writew(b, mem + (i * 0x08) + 4);
+	}
+
+	/* INTERPOLATE modes require a "next" entry to interpolate with,
+	 * so we replicate the last entry to deal with this for now.
+	 */
+	writew(r, mem + (i * 0x08) + 0);
+	writew(g, mem + (i * 0x08) + 2);
+	writew(b, mem + (i * 0x08) + 4);
+	return addr;
+}
+
+void
+nv50_lut_fini(struct nv50_lut *lut)
+{
+	int i;
+	for (i = 0; i < ARRAY_SIZE(lut->mem); i++)
+		nvif_mem_fini(&lut->mem[i]);
+}
+
+int
+nv50_lut_init(struct nv50_disp *disp, struct nvif_mmu *mmu,
+	      struct nv50_lut *lut)
+{
+	const u32 size = disp->disp->object.oclass < GF110_DISP ? 257 : 1025;
+	int i;
+	for (i = 0; i < ARRAY_SIZE(lut->mem); i++) {
+		int ret = nvif_mem_init_map(mmu, NVIF_MEM_VRAM, size * 8,
+					    &lut->mem[i]);
+		if (ret)
+			return ret;
+	}
+	return 0;
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/lut.h b/drivers/gpu/drm/nouveau/dispnv50/lut.h
new file mode 100644
index 0000000..6d7b835
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/lut.h
@@ -0,0 +1,15 @@
+#ifndef __NV50_KMS_LUT_H__
+#define __NV50_KMS_LUT_H__
+#include <nvif/mem.h>
+struct drm_property_blob;
+struct nv50_disp;
+
+struct nv50_lut {
+	struct nvif_mem mem[2];
+};
+
+int nv50_lut_init(struct nv50_disp *, struct nvif_mmu *, struct nv50_lut *);
+void nv50_lut_fini(struct nv50_lut *);
+u32 nv50_lut_load(struct nv50_lut *, bool legacy, int buffer,
+		  struct drm_property_blob *);
+#endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.c b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
index 0f6de60..fbaf8b7 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/wndw.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
@@ -116,6 +116,7 @@ nv50_wndw_flush_clr(struct nv50_wndw *wndw, u32 *interlock, bool flush,
 	};
 	if (clr.sema ) wndw->func-> sema_clr(wndw);
 	if (clr.ntfy ) wndw->func-> ntfy_clr(wndw);
+	if (clr.xlut ) wndw->func-> xlut_clr(wndw);
 	if (clr.image) wndw->func->image_clr(wndw);
 
 	interlock[wndw->interlock.type] |= wndw->interlock.data;
@@ -133,7 +134,18 @@ nv50_wndw_flush_set(struct nv50_wndw *wndw, u32 *interlock,
 	if (asyw->set.sema ) wndw->func->sema_set (wndw, asyw);
 	if (asyw->set.ntfy ) wndw->func->ntfy_set (wndw, asyw);
 	if (asyw->set.image) wndw->func->image_set(wndw, asyw);
-	if (asyw->set.lut  ) wndw->func->lut      (wndw, asyw);
+
+	if (asyw->set.xlut ) {
+		if (asyw->ilut) {
+			asyw->xlut.i.offset =
+				nv50_lut_load(&wndw->ilut,
+					      asyw->xlut.i.mode <= 1,
+					      asyw->xlut.i.buffer,
+					      asyw->ilut);
+		}
+		wndw->func->xlut_set(wndw, asyw);
+	}
+
 	if (asyw->set.point) {
 		wndw->immd->point(wndw, asyw);
 		wndw->immd->update(wndw, interlock);
@@ -241,7 +253,56 @@ nv50_wndw_atomic_check_acquire(struct nv50_wndw *wndw, bool modeset,
 	return wndw->func->acquire(wndw, asyw, asyh);
 }
 
-int
+static void
+nv50_wndw_atomic_check_lut(struct nv50_wndw *wndw,
+			   struct nv50_wndw_atom *armw,
+			   struct nv50_wndw_atom *asyw,
+			   struct nv50_head_atom *asyh)
+{
+	struct drm_property_blob *ilut = asyh->state.degamma_lut;
+
+	/* I8 format without an input LUT makes no sense, and the
+	 * HW error-checks for this.
+	 *
+	 * In order to handle legacy gamma, when there's no input
+	 * LUT we need to steal the output LUT and use it instead.
+	 */
+	if (!ilut && asyw->state.fb->format->format == DRM_FORMAT_C8) {
+		/* This should be an error, but there's legacy clients
+		 * that do a modeset before providing a gamma table.
+		 *
+		 * We keep the window disabled to avoid angering HW.
+		 */
+		if (!(ilut = asyh->state.gamma_lut)) {
+			asyw->visible = false;
+			return;
+		}
+
+		if (wndw->func->ilut)
+			asyh->wndw.olut |= BIT(wndw->id);
+	} else {
+		asyh->wndw.olut &= ~BIT(wndw->id);
+	}
+
+	/* Recalculate LUT state. */
+	memset(&asyw->xlut, 0x00, sizeof(asyw->xlut));
+	if ((asyw->ilut = wndw->func->ilut ? ilut : NULL)) {
+		wndw->func->ilut(wndw, asyw);
+		asyw->xlut.handle = wndw->wndw.vram.handle;
+		asyw->xlut.i.buffer = !asyw->xlut.i.buffer;
+		asyw->set.xlut = true;
+	}
+
+	/* Handle setting base SET_OUTPUT_LUT_LO_ENABLE_USE_CORE_LUT. */
+	if (wndw->func->olut_core &&
+	    (!armw->visible || (armw->xlut.handle && !asyw->xlut.handle)))
+		asyw->set.xlut = true;
+
+	/* Can't do an immediate flip while changing the LUT. */
+	asyh->state.pageflip_flags &= ~DRM_MODE_PAGE_FLIP_ASYNC;
+}
+
+static int
 nv50_wndw_atomic_check(struct drm_plane *plane, struct drm_plane_state *state)
 {
 	struct nouveau_drm *drm = nouveau_drm(plane->dev);
@@ -274,15 +335,26 @@ nv50_wndw_atomic_check(struct drm_plane *plane, struct drm_plane_state *state)
 			return PTR_ERR(harm);
 	}
 
+	/* LUT configuration can potentially cause the window to be disabled. */
+	if (asyw->visible && wndw->func->xlut_set &&
+	    (!armw->visible ||
+	     asyh->state.color_mgmt_changed ||
+	     asyw->state.fb->format->format !=
+	     armw->state.fb->format->format))
+		nv50_wndw_atomic_check_lut(wndw, armw, asyw, asyh);
+
 	/* Calculate new window state. */
 	if (asyw->visible) {
 		ret = nv50_wndw_atomic_check_acquire(wndw, modeset,
 						     armw, asyw, asyh);
 		if (ret)
 			return ret;
+
+		asyh->wndw.mask |= BIT(wndw->id);
 	} else
 	if (armw->visible) {
 		nv50_wndw_atomic_check_release(wndw, asyw, harm);
+		harm->wndw.mask &= ~BIT(wndw->id);
 	} else {
 		return 0;
 	}
@@ -294,9 +366,9 @@ nv50_wndw_atomic_check(struct drm_plane *plane, struct drm_plane_state *state)
 	if (!asyw->visible || modeset) {
 		asyw->clr.ntfy = armw->ntfy.handle != 0;
 		asyw->clr.sema = armw->sema.handle != 0;
+		asyw->clr.xlut = armw->xlut.handle != 0;
 		if (wndw->func->image_clr)
 			asyw->clr.image = armw->image.handle[0] != 0;
-		asyw->set.lut = wndw->func->lut && asyw->visible;
 	}
 
 	return 0;
@@ -381,9 +453,10 @@ nv50_wndw_atomic_duplicate_state(struct drm_plane *plane)
 	__drm_atomic_helper_plane_duplicate_state(plane, &asyw->state);
 	asyw->sema = armw->sema;
 	asyw->ntfy = armw->ntfy;
+	asyw->ilut = NULL;
+	asyw->xlut = armw->xlut;
 	asyw->image = armw->image;
 	asyw->point = armw->point;
-	asyw->lut = armw->lut;
 	asyw->clr.mask = 0;
 	asyw->set.mask = 0;
 	return &asyw->state;
@@ -417,6 +490,9 @@ nv50_wndw_destroy(struct drm_plane *plane)
 	nvif_notify_fini(&wndw->notify);
 	nv50_dmac_destroy(&wndw->wimm);
 	nv50_dmac_destroy(&wndw->wndw);
+
+	nv50_lut_fini(&wndw->ilut);
+
 	drm_plane_cleanup(&wndw->plane);
 	kfree(wndw);
 }
@@ -456,6 +532,9 @@ nv50_wndw_new_(const struct nv50_wndw_func *func, struct drm_device *dev,
 	       enum nv50_disp_interlock_type interlock_type, u32 interlock_data,
 	       struct nv50_wndw **pwndw)
 {
+	struct nouveau_drm *drm = nouveau_drm(dev);
+	struct nvif_mmu *mmu = &drm->client.mmu;
+	struct nv50_disp *disp = nv50_disp(dev);
 	struct nv50_wndw *wndw;
 	int nformat;
 	int ret;
@@ -484,6 +563,12 @@ nv50_wndw_new_(const struct nv50_wndw_func *func, struct drm_device *dev,
 
 	drm_plane_helper_add(&wndw->plane, &nv50_wndw_helper);
 
+	if (wndw->func->ilut) {
+		ret = nv50_lut_init(disp, mmu, &wndw->ilut);
+		if (ret)
+			return ret;
+	}
+
 	wndw->notify.func = nv50_wndw_notify;
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.h b/drivers/gpu/drm/nouveau/dispnv50/wndw.h
index c26796c..223cf3f 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/wndw.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.h
@@ -3,6 +3,7 @@
 #define nv50_wndw(p) container_of((p), struct nv50_wndw, plane)
 #include "disp.h"
 #include "atom.h"
+#include "lut.h"
 
 #include <nvif/notify.h>
 
@@ -24,6 +25,8 @@ struct nv50_wndw {
 
 	struct drm_plane plane;
 
+	struct nv50_lut ilut;
+
 	struct nv50_dmac wndw;
 	struct nv50_dmac wimm;
 
@@ -61,9 +64,12 @@ struct nv50_wndw_func {
 	void (*ntfy_clr)(struct nv50_wndw *);
 	int (*ntfy_wait_begun)(struct nouveau_bo *, u32 offset,
 			       struct nvif_device *);
+	void (*ilut)(struct nv50_wndw *, struct nv50_wndw_atom *);
+	bool olut_core;
+	void (*xlut_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
+	void (*xlut_clr)(struct nv50_wndw *);
 	void (*image_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
 	void (*image_clr)(struct nv50_wndw *);
-	void (*lut)(struct nv50_wndw *, struct nv50_wndw_atom *);
 
 	void (*update)(struct nv50_wndw *, u32 *interlock);
 };
-- 
cgit v1.1


From b05d873808c77fedd25130b0355acc0da1c11e19 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50-: separate blocklinear vs linear pitch

Will be required to support Volta.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/atom.h     |  6 ++++--
 drivers/gpu/drm/nouveau/dispnv50/base507c.c |  5 +++--
 drivers/gpu/drm/nouveau/dispnv50/base827c.c |  5 +++--
 drivers/gpu/drm/nouveau/dispnv50/base907c.c |  5 +++--
 drivers/gpu/drm/nouveau/dispnv50/head507d.c |  8 +++++---
 drivers/gpu/drm/nouveau/dispnv50/head827d.c |  3 ++-
 drivers/gpu/drm/nouveau/dispnv50/head907d.c |  3 ++-
 drivers/gpu/drm/nouveau/dispnv50/wndw.c     | 10 ++++++----
 8 files changed, 28 insertions(+), 17 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/dispnv50/atom.h b/drivers/gpu/drm/nouveau/dispnv50/atom.h
index 3e9e883..0409947 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/atom.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/atom.h
@@ -63,7 +63,8 @@ struct nv50_head_atom {
 		u8  format;
 		u8  kind:7;
 		u8  layout:1;
-		u8  block:4;
+		u8  blockh:4;
+		u16 blocks:12;
 		u32 pitch:20;
 		u16 x;
 		u16 y;
@@ -175,7 +176,8 @@ struct nv50_wndw_atom {
 		u8  format;
 		u8  kind:7;
 		u8  layout:1;
-		u8  block:4;
+		u8  blockh:4;
+		u16 blocks[3];
 		u32 pitch[3];
 		u16 w;
 		u16 h;
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base507c.c b/drivers/gpu/drm/nouveau/dispnv50/base507c.c
index 5d664d7..d5e295c 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/base507c.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/base507c.c
@@ -67,8 +67,9 @@ base507c_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 		evo_data(push, 0x00000000);
 		evo_data(push, asyw->image.h << 16 | asyw->image.w);
 		evo_data(push, asyw->image.layout << 20 |
-			       asyw->image.pitch[0] |
-			       asyw->image.block);
+			       (asyw->image.pitch[0] >> 8) << 8 |
+			       asyw->image.blocks[0] << 8 |
+			       asyw->image.blockh);
 		evo_data(push, asyw->image.kind << 16 |
 			       asyw->image.format << 8);
 		evo_kick(push, &wndw->wndw);
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base827c.c b/drivers/gpu/drm/nouveau/dispnv50/base827c.c
index d886858..7364681 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/base827c.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/base827c.c
@@ -36,8 +36,9 @@ base827c_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 		evo_data(push, 0x00000000);
 		evo_data(push, asyw->image.h << 16 | asyw->image.w);
 		evo_data(push, asyw->image.layout << 20 |
-			       asyw->image.pitch[0] |
-			       asyw->image.block);
+			       (asyw->image.pitch[0] >> 8) << 8 |
+			       asyw->image.blocks[0] << 8 |
+			       asyw->image.blockh);
 		evo_data(push, asyw->image.format << 8);
 		evo_kick(push, &wndw->wndw);
 	}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base907c.c b/drivers/gpu/drm/nouveau/dispnv50/base907c.c
index 2643592..8edc059 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/base907c.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/base907c.c
@@ -36,8 +36,9 @@ base907c_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 		evo_data(push, 0x00000000);
 		evo_data(push, asyw->image.h << 16 | asyw->image.w);
 		evo_data(push, asyw->image.layout << 24 |
-			       asyw->image.pitch[0] |
-			       asyw->image.block);
+			       (asyw->image.pitch[0] >> 8) << 8 |
+			       asyw->image.blocks[0] << 8 |
+			       asyw->image.blockh);
 		evo_data(push, asyw->image.format << 8);
 		evo_kick(push, &wndw->wndw);
 	}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head507d.c b/drivers/gpu/drm/nouveau/dispnv50/head507d.c
index 75575c33..8a8aa9b 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/head507d.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/head507d.c
@@ -151,8 +151,9 @@ head507d_core_set(struct nv50_head *head, struct nv50_head_atom *asyh)
 		evo_mthd(push, 0x0868 + head->base.index * 0x400, 4);
 		evo_data(push, asyh->core.h << 16 | asyh->core.w);
 		evo_data(push, asyh->core.layout << 20 |
-			       asyh->core.pitch >> 8 << 8 |
-			       asyh->core.block);
+			       (asyh->core.pitch >> 8) << 8 |
+			       asyh->core.blocks << 8 |
+			       asyh->core.blockh);
 		evo_data(push, asyh->core.kind << 16 |
 			       asyh->core.format << 8);
 		evo_data(push, asyh->core.handle);
@@ -196,7 +197,8 @@ head507d_core_calc(struct nv50_head *head, struct nv50_head_atom *asyh)
 	asyh->core.format = 0xcf;
 	asyh->core.kind = 0;
 	asyh->core.layout = 1;
-	asyh->core.block = 0;
+	asyh->core.blockh = 0;
+	asyh->core.blocks = 0;
 	asyh->core.pitch = ALIGN(asyh->core.w, 64) * 4;
 }
 
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head827d.c b/drivers/gpu/drm/nouveau/dispnv50/head827d.c
index ddc143b..ae33e21 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/head827d.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/head827d.c
@@ -64,7 +64,8 @@ head827d_core_set(struct nv50_head *head, struct nv50_head_atom *asyh)
 		evo_data(push, asyh->core.h << 16 | asyh->core.w);
 		evo_data(push, asyh->core.layout << 20 |
 			       (asyh->core.pitch >> 8) << 8 |
-			       asyh->core.block);
+			       asyh->core.blocks << 8 |
+			       asyh->core.blockh);
 		evo_data(push, asyh->core.format << 8);
 		evo_data(push, asyh->core.handle);
 		evo_mthd(push, 0x08c0 + head->base.index * 0x400, 1);
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head907d.c b/drivers/gpu/drm/nouveau/dispnv50/head907d.c
index 0fa0159..a05dfcc 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/head907d.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/head907d.c
@@ -172,7 +172,8 @@ head907d_core_set(struct nv50_head *head, struct nv50_head_atom *asyh)
 		evo_data(push, asyh->core.h << 16 | asyh->core.w);
 		evo_data(push, asyh->core.layout << 24 |
 			       (asyh->core.pitch >> 8) << 8 |
-			       asyh->core.block);
+			       asyh->core.blocks << 8 |
+			       asyh->core.blockh);
 		evo_data(push, asyh->core.format << 8);
 		evo_data(push, asyh->core.handle);
 		evo_mthd(push, 0x04b0 + head->base.index * 0x300, 1);
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.c b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
index fbaf8b7..b96dc3d 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/wndw.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
@@ -225,13 +225,15 @@ nv50_wndw_atomic_check_acquire(struct nv50_wndw *wndw, bool modeset,
 		if (asyw->image.kind) {
 			asyw->image.layout = 0;
 			if (drm->client.device.info.chipset >= 0xc0)
-				asyw->image.block = fb->nvbo->mode >> 4;
+				asyw->image.blockh = fb->nvbo->mode >> 4;
 			else
-				asyw->image.block = fb->nvbo->mode;
-			asyw->image.pitch[0] = (fb->base.pitches[0] / 4) << 4;
+				asyw->image.blockh = fb->nvbo->mode;
+			asyw->image.blocks[0] = fb->base.pitches[0] / 64;
+			asyw->image.pitch[0] = 0;
 		} else {
 			asyw->image.layout = 1;
-			asyw->image.block  = 0;
+			asyw->image.blockh = 0;
+			asyw->image.blocks[0] = 0;
 			asyw->image.pitch[0] = fb->base.pitches[0];
 		}
 
-- 
cgit v1.1


From 01d380ab4f702fffa6da60c4b006547b8dd66de8 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/gk104-: support additional cursor sizes

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/atom.h     |  2 +-
 drivers/gpu/drm/nouveau/dispnv50/curs507a.c | 22 +++++++---------------
 drivers/gpu/drm/nouveau/dispnv50/head.h     |  8 ++++++++
 drivers/gpu/drm/nouveau/dispnv50/head507d.c | 28 ++++++++++++++++++++++++++++
 drivers/gpu/drm/nouveau/dispnv50/head827d.c |  2 ++
 drivers/gpu/drm/nouveau/dispnv50/head907d.c |  2 ++
 drivers/gpu/drm/nouveau/dispnv50/head917d.c | 17 +++++++++++++++++
 7 files changed, 65 insertions(+), 16 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/dispnv50/atom.h b/drivers/gpu/drm/nouveau/dispnv50/atom.h
index 0409947..3d059df 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/atom.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/atom.h
@@ -76,7 +76,7 @@ struct nv50_head_atom {
 		bool visible;
 		u32 handle;
 		u64 offset:40;
-		u8  layout:1;
+		u8  layout:2;
 		u8  format:1;
 	} curs;
 
diff --git a/drivers/gpu/drm/nouveau/dispnv50/curs507a.c b/drivers/gpu/drm/nouveau/dispnv50/curs507a.c
index ba05bcb1..291c081 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/curs507a.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/curs507a.c
@@ -21,6 +21,7 @@
  */
 #include "curs.h"
 #include "core.h"
+#include "head.h"
 
 #include <nvif/cl507a.h>
 
@@ -70,6 +71,7 @@ static int
 curs507a_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
 		 struct nv50_head_atom *asyh)
 {
+	struct nv50_head *head = nv50_head(asyw->state.crtc);
 	int ret;
 
 	ret = drm_atomic_helper_check_plane_state(&asyw->state, &asyh->state,
@@ -80,24 +82,14 @@ curs507a_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
 	if (ret || !asyh->curs.visible)
 		return ret;
 
-	switch (asyw->state.fb->width) {
-	case 32: asyh->curs.layout = 0; break;
-	case 64: asyh->curs.layout = 1; break;
-	default:
-		return -EINVAL;
-	}
-
-	if (asyw->state.fb->width != asyw->state.fb->height)
+	if (asyw->image.w != asyw->image.h)
 		return -EINVAL;
 
-	switch (asyw->image.format) {
-	case 0xcf: asyh->curs.format = 1; break;
-	default:
-		WARN_ON(1);
-		return -EINVAL;
-	}
+	ret = head->func->curs_layout(head, asyw, asyh);
+	if (ret)
+		return ret;
 
-	return 0;
+	return head->func->curs_format(head, asyw, asyh);
 }
 
 static const u32
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head.h b/drivers/gpu/drm/nouveau/dispnv50/head.h
index 0802271..8f2c3ff 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/head.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/head.h
@@ -26,6 +26,10 @@ struct nv50_head_func {
 	void (*core_calc)(struct nv50_head *, struct nv50_head_atom *);
 	void (*core_set)(struct nv50_head *, struct nv50_head_atom *);
 	void (*core_clr)(struct nv50_head *);
+	int (*curs_layout)(struct nv50_head *, struct nv50_wndw_atom *,
+			   struct nv50_head_atom *);
+	int (*curs_format)(struct nv50_head *, struct nv50_wndw_atom *,
+			   struct nv50_head_atom *);
 	void (*curs_set)(struct nv50_head *, struct nv50_head_atom *);
 	void (*curs_clr)(struct nv50_head *);
 	void (*base)(struct nv50_head *, struct nv50_head_atom *);
@@ -41,6 +45,10 @@ void head507d_mode(struct nv50_head *, struct nv50_head_atom *);
 void head507d_olut(struct nv50_head *, struct nv50_head_atom *);
 void head507d_core_calc(struct nv50_head *, struct nv50_head_atom *);
 void head507d_core_clr(struct nv50_head *);
+int head507d_curs_layout(struct nv50_head *, struct nv50_wndw_atom *,
+			 struct nv50_head_atom *);
+int head507d_curs_format(struct nv50_head *, struct nv50_wndw_atom *,
+			 struct nv50_head_atom *);
 void head507d_base(struct nv50_head *, struct nv50_head_atom *);
 void head507d_ovly(struct nv50_head *, struct nv50_head_atom *);
 void head507d_dither(struct nv50_head *, struct nv50_head_atom *);
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head507d.c b/drivers/gpu/drm/nouveau/dispnv50/head507d.c
index 8a8aa9b..5b6a280 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/head507d.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/head507d.c
@@ -128,6 +128,32 @@ head507d_curs_set(struct nv50_head *head, struct nv50_head_atom *asyh)
 	}
 }
 
+int
+head507d_curs_format(struct nv50_head *head, struct nv50_wndw_atom *asyw,
+		     struct nv50_head_atom *asyh)
+{
+	switch (asyw->image.format) {
+	case 0xcf: asyh->curs.format = 1; break;
+	default:
+		WARN_ON(1);
+		return -EINVAL;
+	}
+	return 0;
+}
+
+int
+head507d_curs_layout(struct nv50_head *head, struct nv50_wndw_atom *asyw,
+		     struct nv50_head_atom *asyh)
+{
+	switch (asyw->image.w) {
+	case 32: asyh->curs.layout = 0; break;
+	case 64: asyh->curs.layout = 1; break;
+	default:
+		return -EINVAL;
+	}
+	return 0;
+}
+
 void
 head507d_core_clr(struct nv50_head *head)
 {
@@ -287,6 +313,8 @@ head507d = {
 	.core_calc = head507d_core_calc,
 	.core_set = head507d_core_set,
 	.core_clr = head507d_core_clr,
+	.curs_layout = head507d_curs_layout,
+	.curs_format = head507d_curs_format,
 	.curs_set = head507d_curs_set,
 	.curs_clr = head507d_curs_clr,
 	.base = head507d_base,
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head827d.c b/drivers/gpu/drm/nouveau/dispnv50/head827d.c
index ae33e21..af5e7bd 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/head827d.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/head827d.c
@@ -113,6 +113,8 @@ head827d = {
 	.core_calc = head507d_core_calc,
 	.core_set = head827d_core_set,
 	.core_clr = head507d_core_clr,
+	.curs_layout = head507d_curs_layout,
+	.curs_format = head507d_curs_format,
 	.curs_set = head827d_curs_set,
 	.curs_clr = head827d_curs_clr,
 	.base = head507d_base,
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head907d.c b/drivers/gpu/drm/nouveau/dispnv50/head907d.c
index a05dfcc..c09620f 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/head907d.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/head907d.c
@@ -270,6 +270,8 @@ head907d = {
 	.core_calc = head507d_core_calc,
 	.core_set = head907d_core_set,
 	.core_clr = head907d_core_clr,
+	.curs_layout = head507d_curs_layout,
+	.curs_format = head507d_curs_format,
 	.curs_set = head907d_curs_set,
 	.curs_clr = head907d_curs_clr,
 	.base = head907d_base,
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head917d.c b/drivers/gpu/drm/nouveau/dispnv50/head917d.c
index 5f65451..4c019a4 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/head917d.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/head917d.c
@@ -63,6 +63,21 @@ head917d_base(struct nv50_head *head, struct nv50_head_atom *asyh)
 	}
 }
 
+static int
+head917d_curs_layout(struct nv50_head *head, struct nv50_wndw_atom *asyw,
+		     struct nv50_head_atom *asyh)
+{
+	switch (asyw->state.fb->width) {
+	case  32: asyh->curs.layout = 0; break;
+	case  64: asyh->curs.layout = 1; break;
+	case 128: asyh->curs.layout = 2; break;
+	case 256: asyh->curs.layout = 3; break;
+	default:
+		return -EINVAL;
+	}
+	return 0;
+}
+
 const struct nv50_head_func
 head917d = {
 	.view = head907d_view,
@@ -73,6 +88,8 @@ head917d = {
 	.core_calc = head507d_core_calc,
 	.core_set = head907d_core_set,
 	.core_clr = head907d_core_clr,
+	.curs_layout = head917d_curs_layout,
+	.curs_format = head507d_curs_format,
 	.curs_set = head907d_curs_set,
 	.curs_clr = head907d_curs_clr,
 	.base = head917d_base,
-- 
cgit v1.1


From 88b600d421a5550cd56e13f2eda34cbefe417c28 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/gk104-: add support for [XA]2R10G10B10 formats

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/Kbuild     |  1 +
 drivers/gpu/drm/nouveau/dispnv50/base.c     |  4 +--
 drivers/gpu/drm/nouveau/dispnv50/base.h     |  3 ++
 drivers/gpu/drm/nouveau/dispnv50/base907c.c |  2 +-
 drivers/gpu/drm/nouveau/dispnv50/base917c.c | 48 +++++++++++++++++++++++++++++
 drivers/gpu/drm/nouveau/dispnv50/wndw.c     |  2 ++
 6 files changed, 57 insertions(+), 3 deletions(-)
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/base917c.c

diff --git a/drivers/gpu/drm/nouveau/dispnv50/Kbuild b/drivers/gpu/drm/nouveau/dispnv50/Kbuild
index d074bb8..7c337fd 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/Kbuild
+++ b/drivers/gpu/drm/nouveau/dispnv50/Kbuild
@@ -27,6 +27,7 @@ nouveau-y += dispnv50/base.o
 nouveau-y += dispnv50/base507c.o
 nouveau-y += dispnv50/base827c.o
 nouveau-y += dispnv50/base907c.o
+nouveau-y += dispnv50/base917c.o
 
 nouveau-y += dispnv50/curs.o
 nouveau-y += dispnv50/curs507a.o
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base.c b/drivers/gpu/drm/nouveau/dispnv50/base.c
index 5f184ab..7c752ac 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/base.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/base.c
@@ -31,8 +31,8 @@ nv50_base_new(struct nouveau_drm *drm, int head, struct nv50_wndw **pwndw)
 		int version;
 		int (*new)(struct nouveau_drm *, int, s32, struct nv50_wndw **);
 	} bases[] = {
-		{ GK110_DISP_BASE_CHANNEL_DMA, 0, base907c_new },
-		{ GK104_DISP_BASE_CHANNEL_DMA, 0, base907c_new },
+		{ GK110_DISP_BASE_CHANNEL_DMA, 0, base917c_new },
+		{ GK104_DISP_BASE_CHANNEL_DMA, 0, base917c_new },
 		{ GF110_DISP_BASE_CHANNEL_DMA, 0, base907c_new },
 		{ GT214_DISP_BASE_CHANNEL_DMA, 0, base827c_new },
 		{ GT200_DISP_BASE_CHANNEL_DMA, 0, base827c_new },
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base.h b/drivers/gpu/drm/nouveau/dispnv50/base.h
index 87ec839..7afd9e2 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/base.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/base.h
@@ -25,6 +25,9 @@ void base507c_update(struct nv50_wndw *, u32 *);
 int base827c_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
 
 int base907c_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+extern const struct nv50_wndw_func base907c;
+
+int base917c_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
 
 int nv50_base_new(struct nouveau_drm *, int head, struct nv50_wndw **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base907c.c b/drivers/gpu/drm/nouveau/dispnv50/base907c.c
index 8edc059..a562fc9 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/base907c.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/base907c.c
@@ -82,7 +82,7 @@ base907c_ilut(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
 	asyw->xlut.i.enable = 2;
 }
 
-static const struct nv50_wndw_func
+const struct nv50_wndw_func
 base907c = {
 	.acquire = base507c_acquire,
 	.release = base507c_release,
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base917c.c b/drivers/gpu/drm/nouveau/dispnv50/base917c.c
new file mode 100644
index 0000000..54d705b
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/base917c.c
@@ -0,0 +1,48 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "base.h"
+#include "atom.h"
+
+const u32
+base917c_format[] = {
+	DRM_FORMAT_C8,
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_RGB565,
+	DRM_FORMAT_XRGB1555,
+	DRM_FORMAT_ARGB1555,
+	DRM_FORMAT_XBGR2101010,
+	DRM_FORMAT_ABGR2101010,
+	DRM_FORMAT_XBGR8888,
+	DRM_FORMAT_ABGR8888,
+	DRM_FORMAT_XRGB2101010,
+	DRM_FORMAT_ARGB2101010,
+	0
+};
+
+int
+base917c_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return base507c_new_(&base907c, base917c_format, drm, head, oclass,
+			     0x00000002 << (head * 4), pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.c b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
index b96dc3d..861fb0e 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/wndw.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
@@ -194,6 +194,8 @@ nv50_wndw_atomic_check_acquire_rgb(struct nv50_wndw_atom *asyw)
 	case DRM_FORMAT_ABGR2101010: asyw->image.format = 0xd1; break;
 	case DRM_FORMAT_XBGR8888   :
 	case DRM_FORMAT_ABGR8888   : asyw->image.format = 0xd5; break;
+	case DRM_FORMAT_XRGB2101010:
+	case DRM_FORMAT_ARGB2101010: asyw->image.format = 0xdf; break;
 	default:
 		WARN_ON(1);
 		return -EINVAL;
-- 
cgit v1.1


From 2ce7f38629891eeaf3e5d406add102a3fa6f6632 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/kms/nv50-: initial overlay support

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/Kbuild     |   1 +
 drivers/gpu/drm/nouveau/dispnv50/atom.h     |  11 +++
 drivers/gpu/drm/nouveau/dispnv50/base.h     |   2 -
 drivers/gpu/drm/nouveau/dispnv50/curs.h     |   1 -
 drivers/gpu/drm/nouveau/dispnv50/disp.h     |   3 +
 drivers/gpu/drm/nouveau/dispnv50/head507d.c |   3 +-
 drivers/gpu/drm/nouveau/dispnv50/head907d.c |   2 +
 drivers/gpu/drm/nouveau/dispnv50/oimm507b.c |   6 +-
 drivers/gpu/drm/nouveau/dispnv50/ovly.c     |   2 +-
 drivers/gpu/drm/nouveau/dispnv50/ovly.h     |  14 +++
 drivers/gpu/drm/nouveau/dispnv50/ovly507e.c | 144 ++++++++++++++++++++++++++++
 drivers/gpu/drm/nouveau/dispnv50/ovly827e.c |  66 ++++++++++++-
 drivers/gpu/drm/nouveau/dispnv50/ovly907e.c |  38 +++++++-
 drivers/gpu/drm/nouveau/dispnv50/ovly917e.c |  45 +++++++++
 drivers/gpu/drm/nouveau/dispnv50/wndw.c     |  35 ++++++-
 drivers/gpu/drm/nouveau/dispnv50/wndw.h     |   6 ++
 16 files changed, 364 insertions(+), 15 deletions(-)
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/ovly917e.c

diff --git a/drivers/gpu/drm/nouveau/dispnv50/Kbuild b/drivers/gpu/drm/nouveau/dispnv50/Kbuild
index 7c337fd..ebd18cb 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/Kbuild
+++ b/drivers/gpu/drm/nouveau/dispnv50/Kbuild
@@ -40,3 +40,4 @@ nouveau-y += dispnv50/ovly.o
 nouveau-y += dispnv50/ovly507e.o
 nouveau-y += dispnv50/ovly827e.o
 nouveau-y += dispnv50/ovly907e.o
+nouveau-y += dispnv50/ovly917e.o
diff --git a/drivers/gpu/drm/nouveau/dispnv50/atom.h b/drivers/gpu/drm/nouveau/dispnv50/atom.h
index 3d059df..d8337e7 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/atom.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/atom.h
@@ -173,6 +173,7 @@ struct nv50_wndw_atom {
 		u8  mode:2;
 		u8  interval:4;
 
+		u8  colorspace:2;
 		u8  format;
 		u8  kind:7;
 		u8  layout:1;
@@ -187,6 +188,15 @@ struct nv50_wndw_atom {
 	} image;
 
 	struct {
+		u16 sx;
+		u16 sy;
+		u16 sw;
+		u16 sh;
+		u16 dw;
+		u16 dh;
+	} scale;
+
+	struct {
 		u16 x;
 		u16 y;
 	} point;
@@ -197,6 +207,7 @@ struct nv50_wndw_atom {
 			bool sema:1;
 			bool xlut:1;
 			bool image:1;
+			bool scale:1;
 			bool point:1;
 		};
 		u8 mask;
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base.h b/drivers/gpu/drm/nouveau/dispnv50/base.h
index 7afd9e2..e7f14f2 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/base.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/base.h
@@ -13,10 +13,8 @@ void base507c_release(struct nv50_wndw *, struct nv50_wndw_atom *,
 		      struct nv50_head_atom *);
 void base507c_sema_set(struct nv50_wndw *, struct nv50_wndw_atom *);
 void base507c_sema_clr(struct nv50_wndw *);
-void base507c_ntfy_reset(struct nouveau_bo *, u32);
 void base507c_ntfy_set(struct nv50_wndw *, struct nv50_wndw_atom *);
 void base507c_ntfy_clr(struct nv50_wndw *);
-int base507c_ntfy_wait_begun(struct nouveau_bo *, u32, struct nvif_device *);
 void base507c_xlut_set(struct nv50_wndw *, struct nv50_wndw_atom *);
 void base507c_xlut_clr(struct nv50_wndw *);
 void base507c_image_clr(struct nv50_wndw *);
diff --git a/drivers/gpu/drm/nouveau/dispnv50/curs.h b/drivers/gpu/drm/nouveau/dispnv50/curs.h
index 2285247..8edac45 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/curs.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/curs.h
@@ -6,7 +6,6 @@ int curs507a_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
 int curs507a_new_(const struct nv50_wimm_func *, struct nouveau_drm *,
 		  int head, s32 oclass, u32 interlock_data,
 		  struct nv50_wndw **);
-extern const struct nv50_wimm_func curs507a;
 
 int curs907a_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
 
diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.h b/drivers/gpu/drm/nouveau/dispnv50/disp.h
index f3a963b..a89b83f 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/disp.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.h
@@ -16,6 +16,9 @@ struct nv50_disp {
 #define NV50_DISP_BASE_SEM0(c)                    NV50_DISP_WNDW_SEM0(0 + (c))
 #define NV50_DISP_BASE_SEM1(c)                    NV50_DISP_WNDW_SEM1(0 + (c))
 #define NV50_DISP_BASE_NTFY(c)                    NV50_DISP_WNDW_NTFY(0 + (c))
+#define NV50_DISP_OVLY_SEM0(c)                    NV50_DISP_WNDW_SEM0(4 + (c))
+#define NV50_DISP_OVLY_SEM1(c)                    NV50_DISP_WNDW_SEM1(4 + (c))
+#define NV50_DISP_OVLY_NTFY(c)                    NV50_DISP_WNDW_NTFY(4 + (c))
 	struct nouveau_bo *sync;
 
 	struct mutex mutex;
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head507d.c b/drivers/gpu/drm/nouveau/dispnv50/head507d.c
index 5b6a280..51bc599 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/head507d.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/head507d.c
@@ -58,7 +58,6 @@ head507d_ovly(struct nv50_head *head, struct nv50_head_atom *asyh)
 
 	if (asyh->ovly.cpp) {
 		switch (asyh->ovly.cpp) {
-		case 8: bounds |= 0x00000500; break;
 		case 4: bounds |= 0x00000300; break;
 		case 2: bounds |= 0x00000100; break;
 		default:
@@ -66,6 +65,8 @@ head507d_ovly(struct nv50_head *head, struct nv50_head_atom *asyh)
 			break;
 		}
 		bounds |= 0x00000001;
+	} else {
+		bounds |= 0x00000100;
 	}
 
 	if ((push = evo_wait(core, 2))) {
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head907d.c b/drivers/gpu/drm/nouveau/dispnv50/head907d.c
index c09620f..6339071 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/head907d.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/head907d.c
@@ -82,6 +82,8 @@ head907d_ovly(struct nv50_head *head, struct nv50_head_atom *asyh)
 			break;
 		}
 		bounds |= 0x00000001;
+	} else {
+		bounds |= 0x00000100;
 	}
 
 	if ((push = evo_wait(core, 2))) {
diff --git a/drivers/gpu/drm/nouveau/dispnv50/oimm507b.c b/drivers/gpu/drm/nouveau/dispnv50/oimm507b.c
index c4baca8..2ee404b 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/oimm507b.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/oimm507b.c
@@ -23,10 +23,6 @@
 
 #include <nvif/cl507b.h>
 
-static const struct nv50_wimm_func
-oimm507b = {
-};
-
 static int
 oimm507b_init_(const struct nv50_wimm_func *func, struct nouveau_drm *drm,
 	       s32 oclass, struct nv50_wndw *wndw)
@@ -52,5 +48,5 @@ oimm507b_init_(const struct nv50_wimm_func *func, struct nouveau_drm *drm,
 int
 oimm507b_init(struct nouveau_drm *drm, s32 oclass, struct nv50_wndw *wndw)
 {
-	return oimm507b_init_(&oimm507b, drm, oclass, wndw);
+	return oimm507b_init_(&curs507a, drm, oclass, wndw);
 }
diff --git a/drivers/gpu/drm/nouveau/dispnv50/ovly.c b/drivers/gpu/drm/nouveau/dispnv50/ovly.c
index be0f16f..90c246d 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/ovly.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/ovly.c
@@ -32,7 +32,7 @@ nv50_ovly_new(struct nouveau_drm *drm, int head, struct nv50_wndw **pwndw)
 		int version;
 		int (*new)(struct nouveau_drm *, int, s32, struct nv50_wndw **);
 	} ovlys[] = {
-		{ GK104_DISP_OVERLAY_CONTROL_DMA, 0, ovly907e_new },
+		{ GK104_DISP_OVERLAY_CONTROL_DMA, 0, ovly917e_new },
 		{ GF110_DISP_OVERLAY_CONTROL_DMA, 0, ovly907e_new },
 		{ GT214_DISP_OVERLAY_CHANNEL_DMA, 0, ovly827e_new },
 		{ GT200_DISP_OVERLAY_CHANNEL_DMA, 0, ovly827e_new },
diff --git a/drivers/gpu/drm/nouveau/dispnv50/ovly.h b/drivers/gpu/drm/nouveau/dispnv50/ovly.h
index d149ef6..4869d52 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/ovly.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/ovly.h
@@ -6,11 +6,25 @@ int ovly507e_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
 int ovly507e_new_(const struct nv50_wndw_func *, const u32 *format,
 		  struct nouveau_drm *, int head, s32 oclass,
 		  u32 interlock_data, struct nv50_wndw **);
+int ovly507e_acquire(struct nv50_wndw *, struct nv50_wndw_atom *,
+		     struct nv50_head_atom *);
+void ovly507e_release(struct nv50_wndw *, struct nv50_wndw_atom *,
+		      struct nv50_head_atom *);
+void ovly507e_ntfy_set(struct nv50_wndw *, struct nv50_wndw_atom *);
+void ovly507e_ntfy_clr(struct nv50_wndw *);
+void ovly507e_image_clr(struct nv50_wndw *);
+void ovly507e_scale_set(struct nv50_wndw *, struct nv50_wndw_atom *);
+void ovly507e_update(struct nv50_wndw *, u32 *);
 
 extern const u32 ovly827e_format[];
+void ovly827e_ntfy_reset(struct nouveau_bo *, u32);
+int ovly827e_ntfy_wait_begun(struct nouveau_bo *, u32, struct nvif_device *);
+
+extern const struct nv50_wndw_func ovly907e;
 
 int ovly827e_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
 int ovly907e_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+int ovly917e_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
 
 int nv50_ovly_new(struct nouveau_drm *, int head, struct nv50_wndw **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/ovly507e.c b/drivers/gpu/drm/nouveau/dispnv50/ovly507e.c
index 732eea3..cc41766 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/ovly507e.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/ovly507e.c
@@ -20,17 +20,149 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  */
 #include "ovly.h"
+#include "atom.h"
+
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_plane_helper.h>
 
 #include <nvif/cl507e.h>
+#include <nvif/event.h>
+
+void
+ovly507e_update(struct nv50_wndw *wndw, u32 *interlock)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 2))) {
+		evo_mthd(push, 0x0080, 1);
+		evo_data(push, interlock[NV50_DISP_INTERLOCK_CORE]);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+void
+ovly507e_scale_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 4))) {
+		evo_mthd(push, 0x00e0, 3);
+		evo_data(push, asyw->scale.sy << 16 | asyw->scale.sx);
+		evo_data(push, asyw->scale.sh << 16 | asyw->scale.sw);
+		evo_data(push, asyw->scale.dw);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+void
+ovly507e_image_clr(struct nv50_wndw *wndw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 4))) {
+		evo_mthd(push, 0x0084, 1);
+		evo_data(push, 0x00000000);
+		evo_mthd(push, 0x00c0, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+ovly507e_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 12))) {
+		evo_mthd(push, 0x0084, 1);
+		evo_data(push, asyw->image.interval << 4);
+		evo_mthd(push, 0x00c0, 1);
+		evo_data(push, asyw->image.handle[0]);
+		evo_mthd(push, 0x0100, 1);
+		evo_data(push, 0x00000002);
+		evo_mthd(push, 0x0800, 1);
+		evo_data(push, asyw->image.offset[0] >> 8);
+		evo_mthd(push, 0x0808, 3);
+		evo_data(push, asyw->image.h << 16 | asyw->image.w);
+		evo_data(push, asyw->image.layout << 20 |
+			       (asyw->image.pitch[0] >> 8) << 8 |
+			       asyw->image.blocks[0] << 8 |
+			       asyw->image.blockh);
+		evo_data(push, asyw->image.kind << 16 |
+			       asyw->image.format << 8 |
+			       asyw->image.colorspace);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+void
+ovly507e_ntfy_clr(struct nv50_wndw *wndw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 2))) {
+		evo_mthd(push, 0x00a4, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+void
+ovly507e_ntfy_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 3))) {
+		evo_mthd(push, 0x00a0, 2);
+		evo_data(push, asyw->ntfy.awaken << 30 | asyw->ntfy.offset);
+		evo_data(push, asyw->ntfy.handle);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+void
+ovly507e_release(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
+		 struct nv50_head_atom *asyh)
+{
+	asyh->ovly.cpp = 0;
+}
+
+int
+ovly507e_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
+		 struct nv50_head_atom *asyh)
+{
+	const struct drm_framebuffer *fb = asyw->state.fb;
+	int ret;
+
+	ret = drm_atomic_helper_check_plane_state(&asyw->state, &asyh->state,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  true, true);
+	if (ret)
+		return ret;
+
+	asyh->ovly.cpp = fb->format->cpp[0];
+	return 0;
+}
 
 #include "nouveau_bo.h"
 
 static const struct nv50_wndw_func
 ovly507e = {
+	.acquire = ovly507e_acquire,
+	.release = ovly507e_release,
+	.ntfy_set = ovly507e_ntfy_set,
+	.ntfy_clr = ovly507e_ntfy_clr,
+	.ntfy_reset = base507c_ntfy_reset,
+	.ntfy_wait_begun = base507c_ntfy_wait_begun,
+	.image_set = ovly507e_image_set,
+	.image_clr = ovly507e_image_clr,
+	.scale_set = ovly507e_scale_set,
+	.update = ovly507e_update,
 };
 
 static const u32
 ovly507e_format[] = {
+	DRM_FORMAT_YUYV,
+	DRM_FORMAT_UYVY,
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_XRGB1555,
+	DRM_FORMAT_ARGB1555,
 	0
 };
 
@@ -61,6 +193,18 @@ ovly507e_new_(const struct nv50_wndw_func *func, const u32 *format,
 		return ret;
 	}
 
+	ret = nvif_notify_init(&wndw->wndw.base.user, wndw->notify.func, false,
+			       NV50_DISP_OVERLAY_CHANNEL_DMA_V0_NTFY_UEVENT,
+			       &(struct nvif_notify_uevent_req) {},
+			       sizeof(struct nvif_notify_uevent_req),
+			       sizeof(struct nvif_notify_uevent_rep),
+			       &wndw->notify);
+	if (ret)
+		return ret;
+
+	wndw->ntfy = NV50_DISP_OVLY_NTFY(wndw->id);
+	wndw->sema = NV50_DISP_OVLY_SEM0(wndw->id);
+	wndw->data = 0x00000000;
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/nouveau/dispnv50/ovly827e.c b/drivers/gpu/drm/nouveau/dispnv50/ovly827e.c
index a8115f1..aaa9fe5 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/ovly827e.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/ovly827e.c
@@ -20,17 +20,81 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  */
 #include "ovly.h"
+#include "atom.h"
 
 #include <nouveau_bo.h>
 
-#include <nvif/cl507e.h>
+static void
+ovly827e_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 12))) {
+		evo_mthd(push, 0x0084, 1);
+		evo_data(push, asyw->image.interval << 4);
+		evo_mthd(push, 0x00c0, 1);
+		evo_data(push, asyw->image.handle[0]);
+		evo_mthd(push, 0x0100, 1);
+		evo_data(push, 0x00000002);
+		evo_mthd(push, 0x0800, 1);
+		evo_data(push, asyw->image.offset[0] >> 8);
+		evo_mthd(push, 0x0808, 3);
+		evo_data(push, asyw->image.h << 16 | asyw->image.w);
+		evo_data(push, asyw->image.layout << 20 |
+			       (asyw->image.pitch[0] >> 8) << 8 |
+			       asyw->image.blocks[0] << 8 |
+			       asyw->image.blockh);
+		evo_data(push, asyw->image.format << 8 |
+			       asyw->image.colorspace);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+int
+ovly827e_ntfy_wait_begun(struct nouveau_bo *bo, u32 offset,
+			 struct nvif_device *device)
+{
+	s64 time = nvif_msec(device, 2000ULL,
+		u32 data = nouveau_bo_rd32(bo, offset / 4 + 3);
+		if ((data & 0xffff0000) == 0xffff0000)
+			break;
+		usleep_range(1, 2);
+	);
+	return time < 0 ? time : 0;
+}
+
+void
+ovly827e_ntfy_reset(struct nouveau_bo *bo, u32 offset)
+{
+	nouveau_bo_wr32(bo, offset / 4 + 0, 0x00000000);
+	nouveau_bo_wr32(bo, offset / 4 + 1, 0x00000000);
+	nouveau_bo_wr32(bo, offset / 4 + 2, 0x00000000);
+	nouveau_bo_wr32(bo, offset / 4 + 3, 0x80000000);
+}
 
 static const struct nv50_wndw_func
 ovly827e = {
+	.acquire = ovly507e_acquire,
+	.release = ovly507e_release,
+	.ntfy_set = ovly507e_ntfy_set,
+	.ntfy_clr = ovly507e_ntfy_clr,
+	.ntfy_reset = ovly827e_ntfy_reset,
+	.ntfy_wait_begun = ovly827e_ntfy_wait_begun,
+	.image_set = ovly827e_image_set,
+	.image_clr = ovly507e_image_clr,
+	.scale_set = ovly507e_scale_set,
+	.update = ovly507e_update,
 };
 
 const u32
 ovly827e_format[] = {
+	DRM_FORMAT_YUYV,
+	DRM_FORMAT_UYVY,
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_XRGB1555,
+	DRM_FORMAT_ARGB1555,
+	DRM_FORMAT_XBGR2101010,
+	DRM_FORMAT_ABGR2101010,
 	0
 };
 
diff --git a/drivers/gpu/drm/nouveau/dispnv50/ovly907e.c b/drivers/gpu/drm/nouveau/dispnv50/ovly907e.c
index f50da646..a3ce530 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/ovly907e.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/ovly907e.c
@@ -20,9 +20,45 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  */
 #include "ovly.h"
+#include "atom.h"
 
-static const struct nv50_wndw_func
+static void
+ovly907e_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 12))) {
+		evo_mthd(push, 0x0084, 1);
+		evo_data(push, asyw->image.interval << 4);
+		evo_mthd(push, 0x00c0, 1);
+		evo_data(push, asyw->image.handle[0]);
+		evo_mthd(push, 0x0100, 1);
+		evo_data(push, 0x00000002);
+		evo_mthd(push, 0x0400, 1);
+		evo_data(push, asyw->image.offset[0] >> 8);
+		evo_mthd(push, 0x0408, 3);
+		evo_data(push, asyw->image.h << 16 | asyw->image.w);
+		evo_data(push, asyw->image.layout << 24 |
+			       (asyw->image.pitch[0] >> 8) << 8 |
+			       asyw->image.blocks[0] << 8 |
+			       asyw->image.blockh);
+		evo_data(push, asyw->image.format << 8 |
+			       asyw->image.colorspace);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+const struct nv50_wndw_func
 ovly907e = {
+	.acquire = ovly507e_acquire,
+	.release = ovly507e_release,
+	.ntfy_set = ovly507e_ntfy_set,
+	.ntfy_clr = ovly507e_ntfy_clr,
+	.ntfy_reset = ovly827e_ntfy_reset,
+	.ntfy_wait_begun = ovly827e_ntfy_wait_begun,
+	.image_set = ovly907e_image_set,
+	.image_clr = ovly507e_image_clr,
+	.scale_set = ovly507e_scale_set,
+	.update = ovly507e_update,
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/dispnv50/ovly917e.c b/drivers/gpu/drm/nouveau/dispnv50/ovly917e.c
new file mode 100644
index 0000000..505fa7e
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/ovly917e.c
@@ -0,0 +1,45 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "ovly.h"
+
+static const u32
+ovly917e_format[] = {
+	DRM_FORMAT_YUYV,
+	DRM_FORMAT_UYVY,
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_XRGB1555,
+	DRM_FORMAT_ARGB1555,
+	DRM_FORMAT_XBGR2101010,
+	DRM_FORMAT_ABGR2101010,
+	DRM_FORMAT_XRGB2101010,
+	DRM_FORMAT_ARGB2101010,
+	0
+};
+
+int
+ovly917e_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return ovly507e_new_(&ovly907e, ovly917e_format, drm, head, oclass,
+			     0x00000004 << (head * 4), pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.c b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
index 861fb0e..c7c08fa 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/wndw.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
@@ -146,6 +146,7 @@ nv50_wndw_flush_set(struct nv50_wndw *wndw, u32 *interlock,
 		wndw->func->xlut_set(wndw, asyw);
 	}
 
+	if (asyw->set.scale) wndw->func->scale_set(wndw, asyw);
 	if (asyw->set.point) {
 		wndw->immd->point(wndw, asyw);
 		wndw->immd->update(wndw, interlock);
@@ -181,6 +182,20 @@ nv50_wndw_atomic_check_release(struct nv50_wndw *wndw,
 }
 
 static int
+nv50_wndw_atomic_check_acquire_yuv(struct nv50_wndw_atom *asyw)
+{
+	switch (asyw->state.fb->format->format) {
+	case DRM_FORMAT_YUYV: asyw->image.format = 0x28; break;
+	case DRM_FORMAT_UYVY: asyw->image.format = 0x29; break;
+	default:
+		WARN_ON(1);
+		return -EINVAL;
+	}
+	asyw->image.colorspace = 1;
+	return 0;
+}
+
+static int
 nv50_wndw_atomic_check_acquire_rgb(struct nv50_wndw_atom *asyw)
 {
 	switch (asyw->state.fb->format->format) {
@@ -197,9 +212,9 @@ nv50_wndw_atomic_check_acquire_rgb(struct nv50_wndw_atom *asyw)
 	case DRM_FORMAT_XRGB2101010:
 	case DRM_FORMAT_ARGB2101010: asyw->image.format = 0xdf; break;
 	default:
-		WARN_ON(1);
 		return -EINVAL;
 	}
+	asyw->image.colorspace = 0;
 	return 0;
 }
 
@@ -221,8 +236,11 @@ nv50_wndw_atomic_check_acquire(struct nv50_wndw *wndw, bool modeset,
 		asyw->image.kind = fb->nvbo->kind;
 
 		ret = nv50_wndw_atomic_check_acquire_rgb(asyw);
-		if (ret)
-			return ret;
+		if (ret) {
+			ret = nv50_wndw_atomic_check_acquire_yuv(asyw);
+			if (ret)
+				return ret;
+		}
 
 		if (asyw->image.kind) {
 			asyw->image.layout = 0;
@@ -247,6 +265,17 @@ nv50_wndw_atomic_check_acquire(struct nv50_wndw *wndw, bool modeset,
 		asyw->set.image = wndw->func->image_set != NULL;
 	}
 
+	if (wndw->func->scale_set) {
+		asyw->scale.sx = asyw->state.src_x >> 16;
+		asyw->scale.sy = asyw->state.src_y >> 16;
+		asyw->scale.sw = asyw->state.src_w >> 16;
+		asyw->scale.sh = asyw->state.src_h >> 16;
+		asyw->scale.dw = asyw->state.crtc_w;
+		asyw->scale.dh = asyw->state.crtc_h;
+		if (memcmp(&armw->scale, &asyw->scale, sizeof(asyw->scale)))
+			asyw->set.scale = true;
+	}
+
 	if (wndw->immd) {
 		asyw->point.x = asyw->state.crtc_x;
 		asyw->point.y = asyw->state.crtc_y;
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.h b/drivers/gpu/drm/nouveau/dispnv50/wndw.h
index 223cf3f..745304d 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/wndw.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.h
@@ -70,15 +70,21 @@ struct nv50_wndw_func {
 	void (*xlut_clr)(struct nv50_wndw *);
 	void (*image_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
 	void (*image_clr)(struct nv50_wndw *);
+	void (*scale_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
 
 	void (*update)(struct nv50_wndw *, u32 *interlock);
 };
 
 extern const struct drm_plane_funcs nv50_wndw;
 
+void base507c_ntfy_reset(struct nouveau_bo *, u32);
+int base507c_ntfy_wait_begun(struct nouveau_bo *, u32, struct nvif_device *);
+
 struct nv50_wimm_func {
 	void (*point)(struct nv50_wndw *, struct nv50_wndw_atom *);
 
 	void (*update)(struct nv50_wndw *, u32 *interlock);
 };
+
+extern const struct nv50_wimm_func curs507a;
 #endif
-- 
cgit v1.1


From 890c85f3ee106e2f13cb510d3ee73f0214e1c620 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/core: increase maximum number of copy engines to 9

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/include/nvkm/core/device.h | 9 ++++++---
 drivers/gpu/drm/nouveau/nvkm/core/subdev.c         | 3 +++
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c  | 6 ++++++
 3 files changed, 15 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/include/nvkm/core/device.h b/drivers/gpu/drm/nouveau/include/nvkm/core/device.h
index f2f9b9e..08c52e3 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/core/device.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/core/device.h
@@ -38,7 +38,10 @@ enum nvkm_devidx {
 	NVKM_ENGINE_CE3,
 	NVKM_ENGINE_CE4,
 	NVKM_ENGINE_CE5,
-	NVKM_ENGINE_CE_LAST = NVKM_ENGINE_CE5,
+	NVKM_ENGINE_CE6,
+	NVKM_ENGINE_CE7,
+	NVKM_ENGINE_CE8,
+	NVKM_ENGINE_CE_LAST = NVKM_ENGINE_CE8,
 
 	NVKM_ENGINE_CIPHER,
 	NVKM_ENGINE_DISP,
@@ -145,7 +148,7 @@ struct nvkm_device {
 	struct nvkm_volt *volt;
 
 	struct nvkm_engine *bsp;
-	struct nvkm_engine *ce[6];
+	struct nvkm_engine *ce[9];
 	struct nvkm_engine *cipher;
 	struct nvkm_disp *disp;
 	struct nvkm_dma *dma;
@@ -217,7 +220,7 @@ struct nvkm_device_chip {
 	int (*volt    )(struct nvkm_device *, int idx, struct nvkm_volt **);
 
 	int (*bsp     )(struct nvkm_device *, int idx, struct nvkm_engine **);
-	int (*ce[6]   )(struct nvkm_device *, int idx, struct nvkm_engine **);
+	int (*ce[9]   )(struct nvkm_device *, int idx, struct nvkm_engine **);
 	int (*cipher  )(struct nvkm_device *, int idx, struct nvkm_engine **);
 	int (*disp    )(struct nvkm_device *, int idx, struct nvkm_disp **);
 	int (*dma     )(struct nvkm_device *, int idx, struct nvkm_dma **);
diff --git a/drivers/gpu/drm/nouveau/nvkm/core/subdev.c b/drivers/gpu/drm/nouveau/nvkm/core/subdev.c
index b96f9e2..03f676c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/core/subdev.c
+++ b/drivers/gpu/drm/nouveau/nvkm/core/subdev.c
@@ -61,6 +61,9 @@ nvkm_subdev_name[NVKM_SUBDEV_NR] = {
 	[NVKM_ENGINE_CE3     ] = "ce3",
 	[NVKM_ENGINE_CE4     ] = "ce4",
 	[NVKM_ENGINE_CE5     ] = "ce5",
+	[NVKM_ENGINE_CE6     ] = "ce6",
+	[NVKM_ENGINE_CE7     ] = "ce7",
+	[NVKM_ENGINE_CE8     ] = "ce8",
 	[NVKM_ENGINE_CIPHER  ] = "cipher",
 	[NVKM_ENGINE_DISP    ] = "disp",
 	[NVKM_ENGINE_DMAOBJ  ] = "dma",
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 5c79c79..b9b6bef 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2471,6 +2471,9 @@ nvkm_device_engine(struct nvkm_device *device, int index)
 	_(CE3    , device->ce[3]   ,  device->ce[3]);
 	_(CE4    , device->ce[4]   ,  device->ce[4]);
 	_(CE5    , device->ce[5]   ,  device->ce[5]);
+	_(CE6    , device->ce[6]   ,  device->ce[6]);
+	_(CE7    , device->ce[7]   ,  device->ce[7]);
+	_(CE8    , device->ce[8]   ,  device->ce[8]);
 	_(CIPHER , device->cipher  ,  device->cipher);
 	_(DISP   , device->disp    , &device->disp->engine);
 	_(DMAOBJ , device->dma     , &device->dma->engine);
@@ -2925,6 +2928,9 @@ nvkm_device_ctor(const struct nvkm_device_func *func,
 		_(NVKM_ENGINE_CE3     ,    ce[3]);
 		_(NVKM_ENGINE_CE4     ,    ce[4]);
 		_(NVKM_ENGINE_CE5     ,    ce[5]);
+		_(NVKM_ENGINE_CE6     ,    ce[6]);
+		_(NVKM_ENGINE_CE7     ,    ce[7]);
+		_(NVKM_ENGINE_CE8     ,    ce[8]);
 		_(NVKM_ENGINE_CIPHER  ,   cipher);
 		_(NVKM_ENGINE_DISP    ,     disp);
 		_(NVKM_ENGINE_DMAOBJ  ,      dma);
-- 
cgit v1.1


From c1f856bb99499f82420d74886884d193e9d63db7 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/core: recognise gv100

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/include/nvif/cl0080.h      | 1 +
 drivers/gpu/drm/nouveau/include/nvkm/core/device.h | 1 +
 drivers/gpu/drm/nouveau/nouveau_abi16.c            | 1 +
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c  | 7 +++++++
 drivers/gpu/drm/nouveau/nvkm/engine/device/user.c  | 1 +
 5 files changed, 11 insertions(+)

diff --git a/drivers/gpu/drm/nouveau/include/nvif/cl0080.h b/drivers/gpu/drm/nouveau/include/nvif/cl0080.h
index 49c1c90..4f52331 100644
--- a/drivers/gpu/drm/nouveau/include/nvif/cl0080.h
+++ b/drivers/gpu/drm/nouveau/include/nvif/cl0080.h
@@ -31,6 +31,7 @@ struct nv_device_info_v0 {
 #define NV_DEVICE_INFO_V0_KEPLER                                           0x08
 #define NV_DEVICE_INFO_V0_MAXWELL                                          0x09
 #define NV_DEVICE_INFO_V0_PASCAL                                           0x0a
+#define NV_DEVICE_INFO_V0_VOLTA                                            0x0b
 	__u8  family;
 	__u8  pad06[2];
 	__u64 ram_size;
diff --git a/drivers/gpu/drm/nouveau/include/nvkm/core/device.h b/drivers/gpu/drm/nouveau/include/nvkm/core/device.h
index 08c52e3..d83d834 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/core/device.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/core/device.h
@@ -113,6 +113,7 @@ struct nvkm_device {
 		NV_E0    = 0xe0,
 		GM100    = 0x110,
 		GP100    = 0x130,
+		GV100    = 0x140,
 	} card_type;
 	u32 chipset;
 	u8  chiprev;
diff --git a/drivers/gpu/drm/nouveau/nouveau_abi16.c b/drivers/gpu/drm/nouveau/nouveau_abi16.c
index ea24727..e2211bb 100644
--- a/drivers/gpu/drm/nouveau/nouveau_abi16.c
+++ b/drivers/gpu/drm/nouveau/nouveau_abi16.c
@@ -103,6 +103,7 @@ nouveau_abi16_swclass(struct nouveau_drm *drm)
 	case NV_DEVICE_INFO_V0_KEPLER:
 	case NV_DEVICE_INFO_V0_MAXWELL:
 	case NV_DEVICE_INFO_V0_PASCAL:
+	case NV_DEVICE_INFO_V0_VOLTA:
 		return NVIF_CLASS_SW_GF100;
 	}
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index b9b6bef..e0e4d28 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2394,6 +2394,11 @@ nv13b_chipset = {
 	.sw = gf100_sw_new,
 };
 
+static const struct nvkm_device_chip
+nv140_chipset = {
+	.name = "GV100",
+};
+
 static int
 nvkm_device_event_ctor(struct nvkm_object *object, void *data, u32 size,
 		       struct nvkm_notify *notify)
@@ -2750,6 +2755,7 @@ nvkm_device_ctor(const struct nvkm_device_func *func,
 			case 0x110:
 			case 0x120: device->card_type = GM100; break;
 			case 0x130: device->card_type = GP100; break;
+			case 0x140: device->card_type = GV100; break;
 			default:
 				break;
 			}
@@ -2841,6 +2847,7 @@ nvkm_device_ctor(const struct nvkm_device_func *func,
 		case 0x137: device->chip = &nv137_chipset; break;
 		case 0x138: device->chip = &nv138_chipset; break;
 		case 0x13b: device->chip = &nv13b_chipset; break;
+		case 0x140: device->chip = &nv140_chipset; break;
 		default:
 			nvdev_error(device, "unknown chipset (%08x)\n", boot0);
 			goto done;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c
index 600bdb8..dde6bba 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c
@@ -174,6 +174,7 @@ nvkm_udevice_info(struct nvkm_udevice *udev, void *data, u32 size)
 	case NV_E0: args->v0.family = NV_DEVICE_INFO_V0_KEPLER; break;
 	case GM100: args->v0.family = NV_DEVICE_INFO_V0_MAXWELL; break;
 	case GP100: args->v0.family = NV_DEVICE_INFO_V0_PASCAL; break;
+	case GV100: args->v0.family = NV_DEVICE_INFO_V0_VOLTA; break;
 	default:
 		args->v0.family = 0;
 		break;
-- 
cgit v1.1


From 893855d8215f5ec6b0e1fac399960405c8237c53 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/pci/gv100: initial support

Appears to be compatible with GP100.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index e0e4d28..3674db6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2397,6 +2397,7 @@ nv13b_chipset = {
 static const struct nvkm_device_chip
 nv140_chipset = {
 	.name = "GV100",
+	.pci = gp100_pci_new,
 };
 
 static int
-- 
cgit v1.1


From 75e482efd38a40497e06c217e6ae9f92940b218e Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/bios/gv100: initial support

No real surprises here so far.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c      | 1 +
 drivers/gpu/drm/nouveau/nvkm/subdev/bios/shadowramin.c | 5 ++++-
 2 files changed, 5 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 3674db6..9394522 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2397,6 +2397,7 @@ nv13b_chipset = {
 static const struct nvkm_device_chip
 nv140_chipset = {
 	.name = "GV100",
+	.bios = nvkm_bios_new,
 	.pci = gp100_pci_new,
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/shadowramin.c b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/shadowramin.c
index 0f537c2..3634cd0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/shadowramin.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/shadowramin.c
@@ -78,7 +78,10 @@ pramin_init(struct nvkm_bios *bios, const char *name)
 	 * important as we don't want to be touching vram on an
 	 * uninitialised board
 	 */
-	addr = nvkm_rd32(device, 0x619f04);
+	if (device->card_type >= GV100)
+		addr = nvkm_rd32(device, 0x625f04);
+	else
+		addr = nvkm_rd32(device, 0x619f04);
 	if (!(addr & 0x00000008)) {
 		nvkm_debug(subdev, "... not enabled\n");
 		return ERR_PTR(-ENODEV);
-- 
cgit v1.1


From 6827c9a8683d8102479ba7add0cc7ab181143c82 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/bios/pll: limits table 5.0

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/subdev/bios/pll.c | 19 ++++++++++++++++++-
 1 file changed, 18 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/pll.c b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/pll.c
index 2ca23a9..e6e804c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/pll.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/pll.c
@@ -193,7 +193,10 @@ pll_map_type(struct nvkm_bios *bios, u8 type, u32 *reg, u8 *ver, u8 *len)
 		data += hdr;
 		while (cnt--) {
 			if (nvbios_rd08(bios, data + 0) == type) {
-				*reg = nvbios_rd32(bios, data + 3);
+				if (*ver < 0x50)
+					*reg = nvbios_rd32(bios, data + 3);
+				else
+					*reg = 0;
 				return data;
 			}
 			data += *len;
@@ -361,6 +364,20 @@ nvbios_pll_parse(struct nvkm_bios *bios, u32 type, struct nvbios_pll *info)
 		info->min_p = nvbios_rd08(bios, data + 12);
 		info->max_p = nvbios_rd08(bios, data + 13);
 		break;
+	case 0x50:
+		info->refclk = nvbios_rd16(bios, data + 1) * 1000;
+		/* info->refclk_alt = nvbios_rd16(bios, data + 3) * 1000; */
+		info->vco1.min_freq = nvbios_rd16(bios, data + 5) * 1000;
+		info->vco1.max_freq = nvbios_rd16(bios, data + 7) * 1000;
+		info->vco1.min_inputfreq = nvbios_rd16(bios, data + 9) * 1000;
+		info->vco1.max_inputfreq = nvbios_rd16(bios, data + 11) * 1000;
+		info->vco1.min_m = nvbios_rd08(bios, data + 13);
+		info->vco1.max_m = nvbios_rd08(bios, data + 14);
+		info->vco1.min_n = nvbios_rd08(bios, data + 15);
+		info->vco1.max_n = nvbios_rd08(bios, data + 16);
+		info->min_p = nvbios_rd08(bios, data + 17);
+		info->max_p = nvbios_rd08(bios, data + 18);
+		break;
 	default:
 		nvkm_error(subdev, "unknown pll limits version 0x%02x\n", ver);
 		return -EINVAL;
-- 
cgit v1.1


From 8769dc989c53e5ed38460b7585d55f381f51e9d8 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/devinit/gv100: initial support

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 .../gpu/drm/nouveau/include/nvkm/subdev/devinit.h  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/subdev/devinit/Kbuild |  1 +
 .../gpu/drm/nouveau/nvkm/subdev/devinit/gm200.c    |  2 +-
 .../gpu/drm/nouveau/nvkm/subdev/devinit/gv100.c    | 79 ++++++++++++++++++++++
 drivers/gpu/drm/nouveau/nvkm/subdev/devinit/nv50.h |  2 +
 6 files changed, 85 insertions(+), 1 deletion(-)
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/subdev/devinit/gv100.c

diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/devinit.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/devinit.h
index 4055806..486e763 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/devinit.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/devinit.h
@@ -30,4 +30,5 @@ int mcp89_devinit_new(struct nvkm_device *, int, struct nvkm_devinit **);
 int gf100_devinit_new(struct nvkm_device *, int, struct nvkm_devinit **);
 int gm107_devinit_new(struct nvkm_device *, int, struct nvkm_devinit **);
 int gm200_devinit_new(struct nvkm_device *, int, struct nvkm_devinit **);
+int gv100_devinit_new(struct nvkm_device *, int, struct nvkm_devinit **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 9394522..4a2d413 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2398,6 +2398,7 @@ static const struct nvkm_device_chip
 nv140_chipset = {
 	.name = "GV100",
 	.bios = nvkm_bios_new,
+	.devinit = gv100_devinit_new,
 	.pci = gp100_pci_new,
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/Kbuild
index eac88e3..50a4369 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/Kbuild
@@ -12,3 +12,4 @@ nvkm-y += nvkm/subdev/devinit/mcp89.o
 nvkm-y += nvkm/subdev/devinit/gf100.o
 nvkm-y += nvkm/subdev/devinit/gm107.o
 nvkm-y += nvkm/subdev/devinit/gm200.o
+nvkm-y += nvkm/subdev/devinit/gv100.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/gm200.c b/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/gm200.c
index 1730371..b80618e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/gm200.c
@@ -107,7 +107,7 @@ pmu_load(struct nv50_devinit *init, u8 type, bool post,
 	return pmu_exec(init, pmu.init_addr_pmu), 0;
 }
 
-static int
+int
 gm200_devinit_post(struct nvkm_devinit *base, bool post)
 {
 	struct nv50_devinit *init = nv50_devinit(base);
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/gv100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/gv100.c
new file mode 100644
index 0000000..fbde682
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/gv100.c
@@ -0,0 +1,79 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "nv50.h"
+
+#include <subdev/bios.h>
+#include <subdev/bios/pll.h>
+#include <subdev/clk/pll.h>
+
+static int
+gv100_devinit_pll_set(struct nvkm_devinit *init, u32 type, u32 freq)
+{
+	struct nvkm_subdev *subdev = &init->subdev;
+	struct nvkm_device *device = subdev->device;
+	struct nvbios_pll info;
+	int head = type - PLL_VPLL0;
+	int N, fN, M, P;
+	int ret;
+
+	ret = nvbios_pll_parse(device->bios, type, &info);
+	if (ret)
+		return ret;
+
+	ret = gt215_pll_calc(subdev, &info, freq, &N, &fN, &M, &P);
+	if (ret < 0)
+		return ret;
+
+	switch (info.type) {
+	case PLL_VPLL0:
+	case PLL_VPLL1:
+	case PLL_VPLL2:
+	case PLL_VPLL3:
+		nvkm_wr32(device, 0x00ef10 + (head * 0x40), fN << 16);
+		nvkm_wr32(device, 0x00ef04 + (head * 0x40), (P << 16) |
+							    (N <<  8) |
+							    (M <<  0));
+		break;
+	default:
+		nvkm_warn(subdev, "%08x/%dKhz unimplemented\n", type, freq);
+		ret = -EINVAL;
+		break;
+	}
+
+	return ret;
+}
+
+static const struct nvkm_devinit_func
+gv100_devinit = {
+	.preinit = gf100_devinit_preinit,
+	.init = nv50_devinit_init,
+	.post = gm200_devinit_post,
+	.pll_set = gv100_devinit_pll_set,
+	.disable = gm107_devinit_disable,
+};
+
+int
+gv100_devinit_new(struct nvkm_device *device, int index,
+		struct nvkm_devinit **pinit)
+{
+	return nv50_devinit_new_(&gv100_devinit, device, index, pinit);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/nv50.h b/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/nv50.h
index 315ebaf..9b9f0dc 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/nv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/nv50.h
@@ -24,4 +24,6 @@ int  gf100_devinit_pll_set(struct nvkm_devinit *, u32, u32);
 void gf100_devinit_preinit(struct nvkm_devinit *);
 
 u64  gm107_devinit_disable(struct nvkm_devinit *);
+
+int gm200_devinit_post(struct nvkm_devinit *, bool);
 #endif
-- 
cgit v1.1


From a1c771a5cb86e2a45bb4516b40f6127112aaa464 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/top/gv100: initial support

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c | 1 +
 drivers/gpu/drm/nouveau/nvkm/subdev/top/gk104.c   | 3 ++-
 2 files changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 4a2d413..777ba68 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2400,6 +2400,7 @@ nv140_chipset = {
 	.bios = nvkm_bios_new,
 	.devinit = gv100_devinit_new,
 	.pci = gp100_pci_new,
+	.top = gk104_top_new,
 };
 
 static int
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/top/gk104.c b/drivers/gpu/drm/nouveau/nvkm/subdev/top/gk104.c
index fea4957..4f1f3e8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/top/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/top/gk104.c
@@ -48,7 +48,8 @@ gk104_top_oneinit(struct nvkm_top *top)
 		case 0x00000001: /* DATA */
 			inst        = (data & 0x3c000000) >> 26;
 			info->addr  = (data & 0x00fff000);
-			info->fault = (data & 0x000000f8) >> 3;
+			if (data & 0x00000004)
+				info->fault = (data & 0x000003f8) >> 3;
 			break;
 		case 0x00000002: /* ENUM */
 			if (data & 0x00000020)
-- 
cgit v1.1


From 46fe1a813adf9abcf71d0a6641ef2de4ce443485 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/ibus/gv100: initial support

Appears to be compatible with GM200.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 777ba68..def4acd 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2399,6 +2399,7 @@ nv140_chipset = {
 	.name = "GV100",
 	.bios = nvkm_bios_new,
 	.devinit = gv100_devinit_new,
+	.ibus = gm200_ibus_new,
 	.pci = gp100_pci_new,
 	.top = gk104_top_new,
 };
-- 
cgit v1.1


From 8afbcca54976abd47d3cd77ffd99aa8103944483 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/gpio/gv100: initial support

Appears to be compatible with GK104.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index def4acd..5fbc086 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2399,6 +2399,7 @@ nv140_chipset = {
 	.name = "GV100",
 	.bios = nvkm_bios_new,
 	.devinit = gv100_devinit_new,
+	.gpio = gk104_gpio_new,
 	.ibus = gm200_ibus_new,
 	.pci = gp100_pci_new,
 	.top = gk104_top_new,
-- 
cgit v1.1


From d2e3b57d81992442221f00938d137568ee895953 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/i2c/gv100: initial support

Appears to be compatible with GM200.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 5fbc086..92290de 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2400,6 +2400,7 @@ nv140_chipset = {
 	.bios = nvkm_bios_new,
 	.devinit = gv100_devinit_new,
 	.gpio = gk104_gpio_new,
+	.i2c = gm200_i2c_new,
 	.ibus = gm200_ibus_new,
 	.pci = gp100_pci_new,
 	.top = gk104_top_new,
-- 
cgit v1.1


From 292550499af0117c7137071d8a0aee93fab81f44 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/fuse/gv100: initial support

Appears to be compatible with GM107.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 92290de..a9a087a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2399,6 +2399,7 @@ nv140_chipset = {
 	.name = "GV100",
 	.bios = nvkm_bios_new,
 	.devinit = gv100_devinit_new,
+	.fuse = gm107_fuse_new,
 	.gpio = gk104_gpio_new,
 	.i2c = gm200_i2c_new,
 	.ibus = gm200_ibus_new,
-- 
cgit v1.1


From 41af75bd3569eeb105635455cc7a99a930dc35ff Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/mc/gv100: initial support

Appears to be compatible with GP100.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index a9a087a..d330675 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2403,6 +2403,7 @@ nv140_chipset = {
 	.gpio = gk104_gpio_new,
 	.i2c = gm200_i2c_new,
 	.ibus = gm200_ibus_new,
+	.mc = gp100_mc_new,
 	.pci = gp100_pci_new,
 	.top = gk104_top_new,
 };
-- 
cgit v1.1


From 9506bd24072af64dae998ac5b62c16b3492fc8e8 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/bus/gv100: initial support

Appears to be compatible with GF100.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index d330675..9b2b3b8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2398,6 +2398,7 @@ static const struct nvkm_device_chip
 nv140_chipset = {
 	.name = "GV100",
 	.bios = nvkm_bios_new,
+	.bus = gf100_bus_new,
 	.devinit = gv100_devinit_new,
 	.fuse = gm107_fuse_new,
 	.gpio = gk104_gpio_new,
-- 
cgit v1.1


From 936240c9bbb2f10fe11b8fbd4aca79f8f5b1b1da Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/tmr/gv100: initial support

Appears to be compatible with GK20A.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 9b2b3b8..87d50bd 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2406,6 +2406,7 @@ nv140_chipset = {
 	.ibus = gm200_ibus_new,
 	.mc = gp100_mc_new,
 	.pci = gp100_pci_new,
+	.timer = gk20a_timer_new,
 	.top = gk104_top_new,
 };
 
-- 
cgit v1.1


From a4a0cfb6420152de2725e1e407fd86301aceccd1 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/imem/gv100: initial support

Can't imagine this will be any different.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 87d50bd..8528e83 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2404,6 +2404,7 @@ nv140_chipset = {
 	.gpio = gk104_gpio_new,
 	.i2c = gm200_i2c_new,
 	.ibus = gm200_ibus_new,
+	.imem = nv50_instmem_new,
 	.mc = gp100_mc_new,
 	.pci = gp100_pci_new,
 	.timer = gk20a_timer_new,
-- 
cgit v1.1


From 3582942c2820a4dfcd64585140bc6e1ad72c1130 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/fb/gv100: initial support

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/include/nvkm/subdev/fb.h  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c |  1 +
 drivers/gpu/drm/nouveau/nvkm/subdev/fb/Kbuild     |  1 +
 drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp100.c    |  2 +-
 drivers/gpu/drm/nouveau/nvkm/subdev/fb/gv100.c    | 46 +++++++++++++++++++++++
 drivers/gpu/drm/nouveau/nvkm/subdev/fb/priv.h     |  2 +
 6 files changed, 52 insertions(+), 1 deletion(-)
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/subdev/fb/gv100.c

diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/fb.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/fb.h
index 92be0e5..96ccc62 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/fb.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/fb.h
@@ -83,6 +83,7 @@ int gm20b_fb_new(struct nvkm_device *, int, struct nvkm_fb **);
 int gp100_fb_new(struct nvkm_device *, int, struct nvkm_fb **);
 int gp102_fb_new(struct nvkm_device *, int, struct nvkm_fb **);
 int gp10b_fb_new(struct nvkm_device *, int, struct nvkm_fb **);
+int gv100_fb_new(struct nvkm_device *, int, struct nvkm_fb **);
 
 #include <subdev/bios.h>
 #include <subdev/bios/ramcfg.h>
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 8528e83..170a7c5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2400,6 +2400,7 @@ nv140_chipset = {
 	.bios = nvkm_bios_new,
 	.bus = gf100_bus_new,
 	.devinit = gv100_devinit_new,
+	.fb = gv100_fb_new,
 	.fuse = gm107_fuse_new,
 	.gpio = gk104_gpio_new,
 	.i2c = gm200_i2c_new,
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/Kbuild
index b4f22cc..9696109 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/Kbuild
@@ -30,6 +30,7 @@ nvkm-y += nvkm/subdev/fb/gm20b.o
 nvkm-y += nvkm/subdev/fb/gp100.o
 nvkm-y += nvkm/subdev/fb/gp102.o
 nvkm-y += nvkm/subdev/fb/gp10b.o
+nvkm-y += nvkm/subdev/fb/gv100.o
 
 nvkm-y += nvkm/subdev/fb/ram.o
 nvkm-y += nvkm/subdev/fb/ramnv04.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp100.c
index d0a47b9..dffe1f5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp100.c
@@ -26,7 +26,7 @@
 
 #include <core/memory.h>
 
-static void
+void
 gp100_fb_init_unkn(struct nvkm_fb *base)
 {
 	struct nvkm_device *device = gf100_fb(base)->base.subdev.device;
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gv100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gv100.c
new file mode 100644
index 0000000..3c5e02e
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gv100.c
@@ -0,0 +1,46 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "gf100.h"
+#include "ram.h"
+
+static int
+gv100_fb_init_page(struct nvkm_fb *fb)
+{
+	return (fb->page == 16) ? 0 : -EINVAL;
+}
+
+static const struct nvkm_fb_func
+gv100_fb = {
+	.dtor = gf100_fb_dtor,
+	.oneinit = gf100_fb_oneinit,
+	.init = gp100_fb_init,
+	.init_page = gv100_fb_init_page,
+	.init_unkn = gp100_fb_init_unkn,
+	.ram_new = gp100_ram_new,
+	.default_bigpage = 16,
+};
+
+int
+gv100_fb_new(struct nvkm_device *device, int index, struct nvkm_fb **pfb)
+{
+	return gf100_fb_new_(&gv100_fb, device, index, pfb);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/priv.h b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/priv.h
index 414a423..2857f31 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/priv.h
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/priv.h
@@ -68,4 +68,6 @@ int gf100_fb_oneinit(struct nvkm_fb *);
 int gf100_fb_init_page(struct nvkm_fb *);
 
 int gm200_fb_init_page(struct nvkm_fb *);
+
+void gp100_fb_init_unkn(struct nvkm_fb *);
 #endif
-- 
cgit v1.1


From 1bce57250ad226e410f9a1a55c0722f075b01652 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/ltc/gv100: initial support

Appears to be compatible with GP102.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 170a7c5..8da7b44 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2406,6 +2406,7 @@ nv140_chipset = {
 	.i2c = gm200_i2c_new,
 	.ibus = gm200_ibus_new,
 	.imem = nv50_instmem_new,
+	.ltc = gp102_ltc_new,
 	.mc = gp100_mc_new,
 	.pci = gp100_pci_new,
 	.timer = gk20a_timer_new,
-- 
cgit v1.1


From edf50395c7c5e8563843eb586aae57c7ac1214ed Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/mmu/gv100: initial support

VEID support hacked in here, as it's the most convenient place for now.

Will be refined once it's better understood.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/include/nvkm/subdev/mmu.h  |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c  |  1 +
 drivers/gpu/drm/nouveau/nvkm/subdev/mmu/Kbuild     |  2 +
 drivers/gpu/drm/nouveau/nvkm/subdev/mmu/gv100.c    | 43 +++++++++++
 drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmm.h      |  3 +
 drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmmgv100.c | 87 ++++++++++++++++++++++
 6 files changed, 137 insertions(+)
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/subdev/mmu/gv100.c
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmmgv100.c

diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/mmu.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/mmu.h
index baab933..6885955 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/mmu.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/mmu.h
@@ -129,4 +129,5 @@ int gm200_mmu_new(struct nvkm_device *, int, struct nvkm_mmu **);
 int gm20b_mmu_new(struct nvkm_device *, int, struct nvkm_mmu **);
 int gp100_mmu_new(struct nvkm_device *, int, struct nvkm_mmu **);
 int gp10b_mmu_new(struct nvkm_device *, int, struct nvkm_mmu **);
+int gv100_mmu_new(struct nvkm_device *, int, struct nvkm_mmu **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 8da7b44..46dff27 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2408,6 +2408,7 @@ nv140_chipset = {
 	.imem = nv50_instmem_new,
 	.ltc = gp102_ltc_new,
 	.mc = gp100_mc_new,
+	.mmu = gv100_mmu_new,
 	.pci = gp100_pci_new,
 	.timer = gk20a_timer_new,
 	.top = gk104_top_new,
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/Kbuild
index 67ee983..58a24e3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/Kbuild
@@ -12,6 +12,7 @@ nvkm-y += nvkm/subdev/mmu/gm200.o
 nvkm-y += nvkm/subdev/mmu/gm20b.o
 nvkm-y += nvkm/subdev/mmu/gp100.o
 nvkm-y += nvkm/subdev/mmu/gp10b.o
+nvkm-y += nvkm/subdev/mmu/gv100.o
 
 nvkm-y += nvkm/subdev/mmu/mem.o
 nvkm-y += nvkm/subdev/mmu/memnv04.o
@@ -31,6 +32,7 @@ nvkm-y += nvkm/subdev/mmu/vmmgm200.o
 nvkm-y += nvkm/subdev/mmu/vmmgm20b.o
 nvkm-y += nvkm/subdev/mmu/vmmgp100.o
 nvkm-y += nvkm/subdev/mmu/vmmgp10b.o
+nvkm-y += nvkm/subdev/mmu/vmmgv100.o
 
 nvkm-y += nvkm/subdev/mmu/umem.o
 nvkm-y += nvkm/subdev/mmu/ummu.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/gv100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/gv100.c
new file mode 100644
index 0000000..f666cb5
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/gv100.c
@@ -0,0 +1,43 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "mem.h"
+#include "vmm.h"
+
+#include <core/option.h>
+
+#include <nvif/class.h>
+
+static const struct nvkm_mmu_func
+gv100_mmu = {
+	.dma_bits = 47,
+	.mmu = {{ -1, -1, NVIF_CLASS_MMU_GF100}},
+	.mem = {{ -1,  0, NVIF_CLASS_MEM_GF100}, gf100_mem_new, gf100_mem_map },
+	.vmm = {{ -1, -1, NVIF_CLASS_VMM_GP100}, gv100_vmm_new },
+	.kind = gm200_mmu_kind,
+	.kind_sys = true,
+};
+
+int
+gv100_mmu_new(struct nvkm_device *device, int index, struct nvkm_mmu **pmmu)
+{
+	return nvkm_mmu_new_(&gv100_mmu, device, index, pmmu);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmm.h b/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmm.h
index da06e64..1a3b0a3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmm.h
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmm.h
@@ -236,6 +236,9 @@ int gp100_vmm_new(struct nvkm_mmu *, u64, u64, void *, u32,
 int gp10b_vmm_new(struct nvkm_mmu *, u64, u64, void *, u32,
 		  struct lock_class_key *, const char *,
 		  struct nvkm_vmm **);
+int gv100_vmm_new(struct nvkm_mmu *, u64, u64, void *, u32,
+		  struct lock_class_key *, const char *,
+		  struct nvkm_vmm **);
 
 #define VMM_PRINT(l,v,p,f,a...) do {                                           \
 	struct nvkm_vmm *_vmm = (v);                                           \
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmmgv100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmmgv100.c
new file mode 100644
index 0000000..2fa40c1
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmmgv100.c
@@ -0,0 +1,87 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "vmm.h"
+
+#include <subdev/fb.h>
+#include <subdev/ltc.h>
+
+#include <nvif/ifc00d.h>
+#include <nvif/unpack.h>
+
+int
+gv100_vmm_join(struct nvkm_vmm *vmm, struct nvkm_memory *inst)
+{
+	u64 data[2], mask;
+	int ret = gp100_vmm_join(vmm, inst), i;
+	if (ret)
+		return ret;
+
+	nvkm_kmap(inst);
+	data[0] = nvkm_ro32(inst, 0x200);
+	data[1] = nvkm_ro32(inst, 0x204);
+	mask = BIT_ULL(0);
+
+	nvkm_wo32(inst, 0x21c, 0x00000000);
+
+	for (i = 0; i < 64; i++) {
+		if (mask & BIT_ULL(i)) {
+			nvkm_wo32(inst, 0x2a4 + (i * 0x10), data[1]);
+			nvkm_wo32(inst, 0x2a0 + (i * 0x10), data[0]);
+		} else {
+			nvkm_wo32(inst, 0x2a4 + (i * 0x10), 0x00000001);
+			nvkm_wo32(inst, 0x2a0 + (i * 0x10), 0x00000001);
+		}
+		nvkm_wo32(inst, 0x2a8 + (i * 0x10), 0x00000000);
+	}
+
+	nvkm_wo32(inst, 0x298, lower_32_bits(mask));
+	nvkm_wo32(inst, 0x29c, upper_32_bits(mask));
+	nvkm_done(inst);
+	return 0;
+}
+
+static const struct nvkm_vmm_func
+gv100_vmm = {
+	.join = gv100_vmm_join,
+	.part = gf100_vmm_part,
+	.aper = gf100_vmm_aper,
+	.valid = gp100_vmm_valid,
+	.flush = gp100_vmm_flush,
+	.page = {
+		{ 47, &gp100_vmm_desc_16[4], NVKM_VMM_PAGE_Sxxx },
+		{ 38, &gp100_vmm_desc_16[3], NVKM_VMM_PAGE_Sxxx },
+		{ 29, &gp100_vmm_desc_16[2], NVKM_VMM_PAGE_Sxxx },
+		{ 21, &gp100_vmm_desc_16[1], NVKM_VMM_PAGE_SVxC },
+		{ 16, &gp100_vmm_desc_16[0], NVKM_VMM_PAGE_SVxC },
+		{ 12, &gp100_vmm_desc_12[0], NVKM_VMM_PAGE_SVHx },
+		{}
+	}
+};
+
+int
+gv100_vmm_new(struct nvkm_mmu *mmu, u64 addr, u64 size, void *argv, u32 argc,
+	      struct lock_class_key *key, const char *name,
+	      struct nvkm_vmm **pvmm)
+{
+	return nv04_vmm_new_(&gv100_vmm, mmu, 0, addr, size,
+			     argv, argc, key, name, pvmm);
+}
-- 
cgit v1.1


From 013b7b37739ca883b2dd5ef979e0e250ac3dafc1 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/bar/gv100: initial support

Appears to be compatible with GM107.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 46dff27..6aaa3d9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2397,6 +2397,7 @@ nv13b_chipset = {
 static const struct nvkm_device_chip
 nv140_chipset = {
 	.name = "GV100",
+	.bar = gm107_bar_new,
 	.bios = nvkm_bios_new,
 	.bus = gf100_bus_new,
 	.devinit = gv100_devinit_new,
-- 
cgit v1.1


From 8b811951c604e417b4511e3d17a75bb8c84b8f08 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/fault/gv100: initial support

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 .../gpu/drm/nouveau/include/nvkm/subdev/fault.h    |   6 +-
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c  |   1 +
 drivers/gpu/drm/nouveau/nvkm/subdev/fault/Kbuild   |   1 +
 drivers/gpu/drm/nouveau/nvkm/subdev/fault/gv100.c  | 206 +++++++++++++++++++++
 4 files changed, 213 insertions(+), 1 deletion(-)
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/subdev/fault/gv100.c

diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/fault.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/fault.h
index 8e9bc30..5a77498 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/fault.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/fault.h
@@ -1,15 +1,18 @@
 #ifndef __NVKM_FAULT_H__
 #define __NVKM_FAULT_H__
 #include <core/subdev.h>
+#include <core/notify.h>
 
 struct nvkm_fault {
 	const struct nvkm_fault_func *func;
 	struct nvkm_subdev subdev;
 
-	struct nvkm_fault_buffer *buffer[1];
+	struct nvkm_fault_buffer *buffer[2];
 	int buffer_nr;
 
 	struct nvkm_event event;
+
+	struct nvkm_notify nrpfb;
 };
 
 struct nvkm_fault_data {
@@ -26,4 +29,5 @@ struct nvkm_fault_data {
 };
 
 int gp100_fault_new(struct nvkm_device *, int, struct nvkm_fault **);
+int gv100_fault_new(struct nvkm_device *, int, struct nvkm_fault **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 6aaa3d9..7f0385d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2401,6 +2401,7 @@ nv140_chipset = {
 	.bios = nvkm_bios_new,
 	.bus = gf100_bus_new,
 	.devinit = gv100_devinit_new,
+	.fault = gv100_fault_new,
 	.fb = gv100_fb_new,
 	.fuse = gm107_fuse_new,
 	.gpio = gk104_gpio_new,
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fault/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/Kbuild
index 807ea40..45bb46f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/fault/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/Kbuild
@@ -1,2 +1,3 @@
 nvkm-y += nvkm/subdev/fault/base.o
 nvkm-y += nvkm/subdev/fault/gp100.o
+nvkm-y += nvkm/subdev/fault/gv100.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fault/gv100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/gv100.c
new file mode 100644
index 0000000..73c7728
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/gv100.c
@@ -0,0 +1,206 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "priv.h"
+
+#include <core/memory.h>
+#include <subdev/mmu.h>
+#include <engine/fifo.h>
+
+static void
+gv100_fault_buffer_process(struct nvkm_fault_buffer *buffer)
+{
+	struct nvkm_device *device = buffer->fault->subdev.device;
+	struct nvkm_memory *mem = buffer->mem;
+	const u32 foff = buffer->id * 0x14;
+	u32 get = nvkm_rd32(device, 0x100e2c + foff);
+	u32 put = nvkm_rd32(device, 0x100e30 + foff);
+	if (put == get)
+		return;
+
+	nvkm_kmap(mem);
+	while (get != put) {
+		const u32   base = get * buffer->fault->func->buffer.entry_size;
+		const u32 instlo = nvkm_ro32(mem, base + 0x00);
+		const u32 insthi = nvkm_ro32(mem, base + 0x04);
+		const u32 addrlo = nvkm_ro32(mem, base + 0x08);
+		const u32 addrhi = nvkm_ro32(mem, base + 0x0c);
+		const u32 timelo = nvkm_ro32(mem, base + 0x10);
+		const u32 timehi = nvkm_ro32(mem, base + 0x14);
+		const u32  info0 = nvkm_ro32(mem, base + 0x18);
+		const u32  info1 = nvkm_ro32(mem, base + 0x1c);
+		struct nvkm_fault_data info;
+
+		if (++get == buffer->entries)
+			get = 0;
+		nvkm_wr32(device, 0x100e2c + foff, get);
+
+		info.addr   = ((u64)addrhi << 32) | addrlo;
+		info.inst   = ((u64)insthi << 32) | instlo;
+		info.time   = ((u64)timehi << 32) | timelo;
+		info.engine = (info0 & 0x000000ff);
+		info.valid  = (info1 & 0x80000000) >> 31;
+		info.gpc    = (info1 & 0x1f000000) >> 24;
+		info.hub    = (info1 & 0x00100000) >> 20;
+		info.access = (info1 & 0x000f0000) >> 16;
+		info.client = (info1 & 0x00007f00) >> 8;
+		info.reason = (info1 & 0x0000001f);
+
+		nvkm_fifo_fault(device->fifo, &info);
+	}
+	nvkm_done(mem);
+}
+
+static void
+gv100_fault_buffer_fini(struct nvkm_fault_buffer *buffer)
+{
+	struct nvkm_device *device = buffer->fault->subdev.device;
+	const u32 intr = buffer->id ? 0x08000000 : 0x20000000;
+	const u32 foff = buffer->id * 0x14;
+
+	nvkm_mask(device, 0x100a34, intr, intr);
+	nvkm_mask(device, 0x100e34 + foff, 0x80000000, 0x00000000);
+}
+
+static void
+gv100_fault_buffer_init(struct nvkm_fault_buffer *buffer)
+{
+	struct nvkm_device *device = buffer->fault->subdev.device;
+	const u32 intr = buffer->id ? 0x08000000 : 0x20000000;
+	const u32 foff = buffer->id * 0x14;
+
+	nvkm_mask(device, 0x100e34 + foff, 0xc0000000, 0x40000000);
+	nvkm_wr32(device, 0x100e28 + foff, upper_32_bits(buffer->vma->addr));
+	nvkm_wr32(device, 0x100e24 + foff, lower_32_bits(buffer->vma->addr));
+	nvkm_mask(device, 0x100e34 + foff, 0x80000000, 0x80000000);
+	nvkm_mask(device, 0x100a2c, intr, intr);
+}
+
+static u32
+gv100_fault_buffer_entries(struct nvkm_fault_buffer *buffer)
+{
+	struct nvkm_device *device = buffer->fault->subdev.device;
+	const u32 foff = buffer->id * 0x14;
+	nvkm_mask(device, 0x100e34 + foff, 0x40000000, 0x40000000);
+	return nvkm_rd32(device, 0x100e34 + foff) & 0x000fffff;
+}
+
+static int
+gv100_fault_ntfy_nrpfb(struct nvkm_notify *notify)
+{
+	struct nvkm_fault *fault = container_of(notify, typeof(*fault), nrpfb);
+	gv100_fault_buffer_process(fault->buffer[0]);
+	return NVKM_NOTIFY_KEEP;
+}
+
+static void
+gv100_fault_intr_fault(struct nvkm_fault *fault)
+{
+	struct nvkm_subdev *subdev = &fault->subdev;
+	struct nvkm_device *device = subdev->device;
+	struct nvkm_fault_data info;
+	const u32 addrlo = nvkm_rd32(device, 0x100e4c);
+	const u32 addrhi = nvkm_rd32(device, 0x100e50);
+	const u32  info0 = nvkm_rd32(device, 0x100e54);
+	const u32 insthi = nvkm_rd32(device, 0x100e58);
+	const u32  info1 = nvkm_rd32(device, 0x100e5c);
+
+	info.addr = ((u64)addrhi << 32) | addrlo;
+	info.inst = ((u64)insthi << 32) | (info0 & 0xfffff000);
+	info.time = 0;
+	info.engine = (info0 & 0x000000ff);
+	info.valid  = (info1 & 0x80000000) >> 31;
+	info.gpc    = (info1 & 0x1f000000) >> 24;
+	info.hub    = (info1 & 0x00100000) >> 20;
+	info.access = (info1 & 0x000f0000) >> 16;
+	info.client = (info1 & 0x00007f00) >> 8;
+	info.reason = (info1 & 0x0000001f);
+
+	nvkm_fifo_fault(device->fifo, &info);
+}
+
+static void
+gv100_fault_intr(struct nvkm_fault *fault)
+{
+	struct nvkm_subdev *subdev = &fault->subdev;
+	struct nvkm_device *device = subdev->device;
+	u32 stat = nvkm_rd32(device, 0x100a20);
+
+	if (stat & 0x80000000) {
+		gv100_fault_intr_fault(fault);
+		nvkm_wr32(device, 0x100e60, 0x80000000);
+		stat &= ~0x80000000;
+	}
+
+	if (stat & 0x20000000) {
+		if (fault->buffer[0]) {
+			nvkm_event_send(&fault->event, 1, 0, NULL, 0);
+			stat &= ~0x20000000;
+		}
+	}
+
+	if (stat) {
+		nvkm_debug(subdev, "intr %08x\n", stat);
+	}
+}
+
+static void
+gv100_fault_fini(struct nvkm_fault *fault)
+{
+	nvkm_notify_put(&fault->nrpfb);
+	nvkm_mask(fault->subdev.device, 0x100a34, 0x80000000, 0x80000000);
+}
+
+static void
+gv100_fault_init(struct nvkm_fault *fault)
+{
+	nvkm_mask(fault->subdev.device, 0x100a2c, 0x80000000, 0x80000000);
+	nvkm_notify_get(&fault->nrpfb);
+}
+
+static const struct nvkm_fault_func
+gv100_fault = {
+	.init = gv100_fault_init,
+	.fini = gv100_fault_fini,
+	.intr = gv100_fault_intr,
+	.buffer.nr = 2,
+	.buffer.entry_size = 32,
+	.buffer.entries = gv100_fault_buffer_entries,
+	.buffer.init = gv100_fault_buffer_init,
+	.buffer.fini = gv100_fault_buffer_fini,
+};
+
+int
+gv100_fault_new(struct nvkm_device *device, int index,
+		struct nvkm_fault **pfault)
+{
+	struct nvkm_fault *fault;
+	int ret;
+
+	ret = nvkm_fault_new_(&gv100_fault, device, index, &fault);
+	*pfault = fault;
+	if (ret)
+		return ret;
+
+	return nvkm_notify_init(&fault->buffer[0]->object, &fault->event,
+				gv100_fault_ntfy_nrpfb, false, NULL, 0, 0,
+				&fault->nrpfb);
+}
-- 
cgit v1.1


From ada0c562814ca466386d9bf1e61fd5ee46f2a72c Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/pmu/gv100: initial support

Appears to be compatible with GP102.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 7f0385d..12e24eb 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2412,6 +2412,7 @@ nv140_chipset = {
 	.mc = gp100_mc_new,
 	.mmu = gv100_mmu_new,
 	.pci = gp100_pci_new,
+	.pmu = gp102_pmu_new,
 	.timer = gk20a_timer_new,
 	.top = gk104_top_new,
 };
-- 
cgit v1.1


From 24a7513c1026fb1b2d42df1c31ea2da56c1604e2 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/therm/gv100: initial support

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 12e24eb..965c433 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2413,6 +2413,7 @@ nv140_chipset = {
 	.mmu = gv100_mmu_new,
 	.pci = gp100_pci_new,
 	.pmu = gp102_pmu_new,
+	.therm = gp100_therm_new,
 	.timer = gk20a_timer_new,
 	.top = gk104_top_new,
 };
-- 
cgit v1.1


From 6fb566b913728fe2dadc8271a568583b8854af93 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/dma/gv100: initial support

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/include/nvkm/engine/dma.h  |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c  |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/dma/Kbuild     |   2 +
 drivers/gpu/drm/nouveau/nvkm/engine/dma/gv100.c    |  34 ++++++
 drivers/gpu/drm/nouveau/nvkm/engine/dma/user.h     |   2 +
 .../gpu/drm/nouveau/nvkm/engine/dma/usergv100.c    | 119 +++++++++++++++++++++
 6 files changed, 159 insertions(+)
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/dma/gv100.c
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/dma/usergv100.c

diff --git a/drivers/gpu/drm/nouveau/include/nvkm/engine/dma.h b/drivers/gpu/drm/nouveau/include/nvkm/engine/dma.h
index 0f9c1c7..f0c1b2c 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/engine/dma.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/engine/dma.h
@@ -27,4 +27,5 @@ int nv04_dma_new(struct nvkm_device *, int, struct nvkm_dma **);
 int nv50_dma_new(struct nvkm_device *, int, struct nvkm_dma **);
 int gf100_dma_new(struct nvkm_device *, int, struct nvkm_dma **);
 int gf119_dma_new(struct nvkm_device *, int, struct nvkm_dma **);
+int gv100_dma_new(struct nvkm_device *, int, struct nvkm_dma **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 965c433..f2c3fe1 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2416,6 +2416,7 @@ nv140_chipset = {
 	.therm = gp100_therm_new,
 	.timer = gk20a_timer_new,
 	.top = gk104_top_new,
+	.dma = gv100_dma_new,
 };
 
 static int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/dma/Kbuild b/drivers/gpu/drm/nouveau/nvkm/engine/dma/Kbuild
index c4a2ce9..e96d1f5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/dma/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/dma/Kbuild
@@ -3,9 +3,11 @@ nvkm-y += nvkm/engine/dma/nv04.o
 nvkm-y += nvkm/engine/dma/nv50.o
 nvkm-y += nvkm/engine/dma/gf100.o
 nvkm-y += nvkm/engine/dma/gf119.o
+nvkm-y += nvkm/engine/dma/gv100.o
 
 nvkm-y += nvkm/engine/dma/user.o
 nvkm-y += nvkm/engine/dma/usernv04.o
 nvkm-y += nvkm/engine/dma/usernv50.o
 nvkm-y += nvkm/engine/dma/usergf100.o
 nvkm-y += nvkm/engine/dma/usergf119.o
+nvkm-y += nvkm/engine/dma/usergv100.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/dma/gv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/dma/gv100.c
new file mode 100644
index 0000000..c65a4c2
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/dma/gv100.c
@@ -0,0 +1,34 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "priv.h"
+#include "user.h"
+
+static const struct nvkm_dma_func
+gv100_dma = {
+	.class_new = gv100_dmaobj_new,
+};
+
+int
+gv100_dma_new(struct nvkm_device *device, int index, struct nvkm_dma **pdma)
+{
+	return nvkm_dma_new_(&gv100_dma, device, index, pdma);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/dma/user.h b/drivers/gpu/drm/nouveau/nvkm/engine/dma/user.h
index 4bbac8a..9fe01fd 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/dma/user.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/dma/user.h
@@ -16,4 +16,6 @@ int gf100_dmaobj_new(struct nvkm_dma *, const struct nvkm_oclass *, void *, u32,
 		     struct nvkm_dmaobj **);
 int gf119_dmaobj_new(struct nvkm_dma *, const struct nvkm_oclass *, void *, u32,
 		     struct nvkm_dmaobj **);
+int gv100_dmaobj_new(struct nvkm_dma *, const struct nvkm_oclass *, void *, u32,
+		     struct nvkm_dmaobj **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/dma/usergv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/dma/usergv100.c
new file mode 100644
index 0000000..39eba9f
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/dma/usergv100.c
@@ -0,0 +1,119 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#define gv100_dmaobj(p) container_of((p), struct gv100_dmaobj, base)
+#include "user.h"
+
+#include <core/client.h>
+#include <core/gpuobj.h>
+#include <subdev/fb.h>
+
+#include <nvif/cl0002.h>
+#include <nvif/unpack.h>
+
+struct gv100_dmaobj {
+	struct nvkm_dmaobj base;
+	u32 flags0;
+};
+
+static int
+gv100_dmaobj_bind(struct nvkm_dmaobj *base, struct nvkm_gpuobj *parent,
+		  int align, struct nvkm_gpuobj **pgpuobj)
+{
+	struct gv100_dmaobj *dmaobj = gv100_dmaobj(base);
+	struct nvkm_device *device = dmaobj->base.dma->engine.subdev.device;
+	u64 start = dmaobj->base.start >> 8;
+	u64 limit = dmaobj->base.limit >> 8;
+	int ret;
+
+	ret = nvkm_gpuobj_new(device, 24, align, false, parent, pgpuobj);
+	if (ret == 0) {
+		nvkm_kmap(*pgpuobj);
+		nvkm_wo32(*pgpuobj, 0x00, dmaobj->flags0);
+		nvkm_wo32(*pgpuobj, 0x04, lower_32_bits(start));
+		nvkm_wo32(*pgpuobj, 0x08, upper_32_bits(start));
+		nvkm_wo32(*pgpuobj, 0x0c, lower_32_bits(limit));
+		nvkm_wo32(*pgpuobj, 0x10, upper_32_bits(limit));
+		nvkm_done(*pgpuobj);
+	}
+
+	return ret;
+}
+
+static const struct nvkm_dmaobj_func
+gv100_dmaobj_func = {
+	.bind = gv100_dmaobj_bind,
+};
+
+int
+gv100_dmaobj_new(struct nvkm_dma *dma, const struct nvkm_oclass *oclass,
+		 void *data, u32 size, struct nvkm_dmaobj **pdmaobj)
+{
+	union {
+		struct gf119_dma_v0 v0;
+	} *args;
+	struct nvkm_object *parent = oclass->parent;
+	struct gv100_dmaobj *dmaobj;
+	u32 kind, page;
+	int ret;
+
+	if (!(dmaobj = kzalloc(sizeof(*dmaobj), GFP_KERNEL)))
+		return -ENOMEM;
+	*pdmaobj = &dmaobj->base;
+
+	ret = nvkm_dmaobj_ctor(&gv100_dmaobj_func, dma, oclass,
+			       &data, &size, &dmaobj->base);
+	if (ret)
+		return ret;
+
+	ret  = -ENOSYS;
+	args = data;
+
+	nvif_ioctl(parent, "create gv100 dma size %d\n", size);
+	if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false))) {
+		nvif_ioctl(parent,
+			   "create gv100 dma vers %d page %d kind %02x\n",
+			   args->v0.version, args->v0.page, args->v0.kind);
+		kind = args->v0.kind != 0;
+		page = args->v0.page != 0;
+	} else
+	if (size == 0) {
+		kind = 0;
+		page = GF119_DMA_V0_PAGE_SP;
+	} else
+		return ret;
+
+	if (kind)
+		dmaobj->flags0 |= 0x00100000;
+	if (page)
+		dmaobj->flags0 |= 0x00000040;
+	dmaobj->flags0 |= 0x00000004; /* rw */
+
+	switch (dmaobj->base.target) {
+	case NV_MEM_TARGET_VRAM       : dmaobj->flags0 |= 0x00000001; break;
+	case NV_MEM_TARGET_PCI        : dmaobj->flags0 |= 0x00000002; break;
+	case NV_MEM_TARGET_PCI_NOSNOOP: dmaobj->flags0 |= 0x00000003; break;
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
-- 
cgit v1.1


From 290ffeafcc1a953aa287c8a7bf7f6d9af25b7e77 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:47 +1000
Subject: drm/nouveau/disp/gv100: initial support

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/include/nvif/class.h       |   7 +
 drivers/gpu/drm/nouveau/include/nvif/clc37b.h      |  11 +
 drivers/gpu/drm/nouveau/include/nvif/clc37e.h      |  13 +
 drivers/gpu/drm/nouveau/include/nvkm/engine/disp.h |   1 +
 drivers/gpu/drm/nouveau/nvif/disp.c                |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c  |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild    |  13 +
 .../gpu/drm/nouveau/nvkm/engine/disp/changv100.c   |  34 ++
 .../gpu/drm/nouveau/nvkm/engine/disp/channv50.h    |  14 +
 .../gpu/drm/nouveau/nvkm/engine/disp/coregv100.c   | 204 ++++++++++
 .../gpu/drm/nouveau/nvkm/engine/disp/cursgv100.c   |  81 ++++
 .../gpu/drm/nouveau/nvkm/engine/disp/dmacgv100.c   |  77 ++++
 drivers/gpu/drm/nouveau/nvkm/engine/disp/gv100.c   | 427 +++++++++++++++++++++
 .../gpu/drm/nouveau/nvkm/engine/disp/hdmigv100.c   |  85 ++++
 drivers/gpu/drm/nouveau/nvkm/engine/disp/head.h    |   4 +
 .../gpu/drm/nouveau/nvkm/engine/disp/headgf119.c   |   2 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/headgv100.c   | 105 +++++
 drivers/gpu/drm/nouveau/nvkm/engine/disp/ior.h     |  10 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c    |   9 +-
 drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h    |   9 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/rootgv100.c   |  52 +++
 .../gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h    |   1 +
 .../gpu/drm/nouveau/nvkm/engine/disp/sorgm200.c    |   6 +-
 .../gpu/drm/nouveau/nvkm/engine/disp/sorgv100.c    | 120 ++++++
 .../gpu/drm/nouveau/nvkm/engine/disp/wimmgv100.c   |  82 ++++
 .../gpu/drm/nouveau/nvkm/engine/disp/wndwgv100.c   | 184 +++++++++
 26 files changed, 1544 insertions(+), 9 deletions(-)
 create mode 100644 drivers/gpu/drm/nouveau/include/nvif/clc37b.h
 create mode 100644 drivers/gpu/drm/nouveau/include/nvif/clc37e.h
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/changv100.c
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/coregv100.c
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgv100.c
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgv100.c
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/gv100.c
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/hdmigv100.c
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/headgv100.c
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgv100.c
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgv100.c
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/wimmgv100.c
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/disp/wndwgv100.c

diff --git a/drivers/gpu/drm/nouveau/include/nvif/class.h b/drivers/gpu/drm/nouveau/include/nvif/class.h
index a7c5bf5..8c9aa55 100644
--- a/drivers/gpu/drm/nouveau/include/nvif/class.h
+++ b/drivers/gpu/drm/nouveau/include/nvif/class.h
@@ -79,6 +79,7 @@
 #define GM200_DISP                                    /* cl5070.h */ 0x00009570
 #define GP100_DISP                                    /* cl5070.h */ 0x00009770
 #define GP102_DISP                                    /* cl5070.h */ 0x00009870
+#define GV100_DISP                                    /* cl5070.h */ 0x0000c370
 
 #define NV31_MPEG                                                    0x00003174
 #define G82_MPEG                                                     0x00008274
@@ -90,6 +91,7 @@
 #define GT214_DISP_CURSOR                             /* cl507a.h */ 0x0000857a
 #define GF110_DISP_CURSOR                             /* cl507a.h */ 0x0000907a
 #define GK104_DISP_CURSOR                             /* cl507a.h */ 0x0000917a
+#define GV100_DISP_CURSOR                             /* cl507a.h */ 0x0000c37a
 
 #define NV50_DISP_OVERLAY                             /* cl507b.h */ 0x0000507b
 #define G82_DISP_OVERLAY                              /* cl507b.h */ 0x0000827b
@@ -97,6 +99,8 @@
 #define GF110_DISP_OVERLAY                            /* cl507b.h */ 0x0000907b
 #define GK104_DISP_OVERLAY                            /* cl507b.h */ 0x0000917b
 
+#define GV100_DISP_WINDOW_IMM_CHANNEL_DMA             /* clc37b.h */ 0x0000c37b
+
 #define NV50_DISP_BASE_CHANNEL_DMA                    /* cl507c.h */ 0x0000507c
 #define G82_DISP_BASE_CHANNEL_DMA                     /* cl507c.h */ 0x0000827c
 #define GT200_DISP_BASE_CHANNEL_DMA                   /* cl507c.h */ 0x0000837c
@@ -117,6 +121,7 @@
 #define GM200_DISP_CORE_CHANNEL_DMA                   /* cl507d.h */ 0x0000957d
 #define GP100_DISP_CORE_CHANNEL_DMA                   /* cl507d.h */ 0x0000977d
 #define GP102_DISP_CORE_CHANNEL_DMA                   /* cl507d.h */ 0x0000987d
+#define GV100_DISP_CORE_CHANNEL_DMA                   /* cl507d.h */ 0x0000c37d
 
 #define NV50_DISP_OVERLAY_CHANNEL_DMA                 /* cl507e.h */ 0x0000507e
 #define G82_DISP_OVERLAY_CHANNEL_DMA                  /* cl507e.h */ 0x0000827e
@@ -125,6 +130,8 @@
 #define GF110_DISP_OVERLAY_CONTROL_DMA                /* cl507e.h */ 0x0000907e
 #define GK104_DISP_OVERLAY_CONTROL_DMA                /* cl507e.h */ 0x0000917e
 
+#define GV100_DISP_WINDOW_CHANNEL_DMA                 /* clc37e.h */ 0x0000c37e
+
 #define NV50_TESLA                                                   0x00005097
 #define G82_TESLA                                                    0x00008297
 #define GT200_TESLA                                                  0x00008397
diff --git a/drivers/gpu/drm/nouveau/include/nvif/clc37b.h b/drivers/gpu/drm/nouveau/include/nvif/clc37b.h
new file mode 100644
index 0000000..89b1818
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/include/nvif/clc37b.h
@@ -0,0 +1,11 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __NVIF_CLC37B_H__
+#define __NVIF_CLC37B_H__
+
+struct nvc37b_window_imm_channel_dma_v0 {
+	__u8  version;
+	__u8  index;
+	__u8  pad02[6];
+	__u64 pushbuf;
+};
+#endif
diff --git a/drivers/gpu/drm/nouveau/include/nvif/clc37e.h b/drivers/gpu/drm/nouveau/include/nvif/clc37e.h
new file mode 100644
index 0000000..899db9e
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/include/nvif/clc37e.h
@@ -0,0 +1,13 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __NVIF_CLC37E_H__
+#define __NVIF_CLC37E_H__
+
+struct nvc37e_window_channel_dma_v0 {
+	__u8  version;
+	__u8  index;
+	__u8  pad02[6];
+	__u64 pushbuf;
+};
+
+#define NVC37E_WINDOW_CHANNEL_DMA_V0_NTFY_UEVENT                           0x00
+#endif
diff --git a/drivers/gpu/drm/nouveau/include/nvkm/engine/disp.h b/drivers/gpu/drm/nouveau/include/nvkm/engine/disp.h
index e83193d..ef7dc08 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/engine/disp.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/engine/disp.h
@@ -35,4 +35,5 @@ int gm107_disp_new(struct nvkm_device *, int, struct nvkm_disp **);
 int gm200_disp_new(struct nvkm_device *, int, struct nvkm_disp **);
 int gp100_disp_new(struct nvkm_device *, int, struct nvkm_disp **);
 int gp102_disp_new(struct nvkm_device *, int, struct nvkm_disp **);
+int gv100_disp_new(struct nvkm_device *, int, struct nvkm_disp **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvif/disp.c b/drivers/gpu/drm/nouveau/nvif/disp.c
index 7006482..18c7d06 100644
--- a/drivers/gpu/drm/nouveau/nvif/disp.c
+++ b/drivers/gpu/drm/nouveau/nvif/disp.c
@@ -34,6 +34,7 @@ int
 nvif_disp_ctor(struct nvif_device *device, s32 oclass, struct nvif_disp *disp)
 {
 	static const struct nvif_mclass disps[] = {
+		{ GV100_DISP, -1 },
 		{ GP102_DISP, -1 },
 		{ GP100_DISP, -1 },
 		{ GM200_DISP, -1 },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index f2c3fe1..43f6b7a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2416,6 +2416,7 @@ nv140_chipset = {
 	.therm = gp100_therm_new,
 	.timer = gk20a_timer_new,
 	.top = gk104_top_new,
+	.disp = gv100_disp_new,
 	.dma = gv100_dma_new,
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild b/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild
index b580581..3d485db 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild
@@ -14,12 +14,14 @@ nvkm-y += nvkm/engine/disp/gm107.o
 nvkm-y += nvkm/engine/disp/gm200.o
 nvkm-y += nvkm/engine/disp/gp100.o
 nvkm-y += nvkm/engine/disp/gp102.o
+nvkm-y += nvkm/engine/disp/gv100.o
 nvkm-y += nvkm/engine/disp/vga.o
 
 nvkm-y += nvkm/engine/disp/head.o
 nvkm-y += nvkm/engine/disp/headnv04.o
 nvkm-y += nvkm/engine/disp/headnv50.o
 nvkm-y += nvkm/engine/disp/headgf119.o
+nvkm-y += nvkm/engine/disp/headgv100.o
 
 nvkm-y += nvkm/engine/disp/ior.o
 nvkm-y += nvkm/engine/disp/dacnv50.o
@@ -35,6 +37,7 @@ nvkm-y += nvkm/engine/disp/sorgf119.o
 nvkm-y += nvkm/engine/disp/sorgk104.o
 nvkm-y += nvkm/engine/disp/sorgm107.o
 nvkm-y += nvkm/engine/disp/sorgm200.o
+nvkm-y += nvkm/engine/disp/sorgv100.o
 
 nvkm-y += nvkm/engine/disp/outp.o
 nvkm-y += nvkm/engine/disp/dp.o
@@ -47,6 +50,7 @@ nvkm-y += nvkm/engine/disp/hdmig84.o
 nvkm-y += nvkm/engine/disp/hdmigt215.o
 nvkm-y += nvkm/engine/disp/hdmigf119.o
 nvkm-y += nvkm/engine/disp/hdmigk104.o
+nvkm-y += nvkm/engine/disp/hdmigv100.o
 
 nvkm-y += nvkm/engine/disp/conn.o
 
@@ -63,13 +67,16 @@ nvkm-y += nvkm/engine/disp/rootgm107.o
 nvkm-y += nvkm/engine/disp/rootgm200.o
 nvkm-y += nvkm/engine/disp/rootgp100.o
 nvkm-y += nvkm/engine/disp/rootgp102.o
+nvkm-y += nvkm/engine/disp/rootgv100.o
 
 nvkm-y += nvkm/engine/disp/channv50.o
 nvkm-y += nvkm/engine/disp/changf119.o
+nvkm-y += nvkm/engine/disp/changv100.o
 
 nvkm-y += nvkm/engine/disp/dmacnv50.o
 nvkm-y += nvkm/engine/disp/dmacgf119.o
 nvkm-y += nvkm/engine/disp/dmacgp102.o
+nvkm-y += nvkm/engine/disp/dmacgv100.o
 
 nvkm-y += nvkm/engine/disp/basenv50.o
 nvkm-y += nvkm/engine/disp/baseg84.o
@@ -82,6 +89,7 @@ nvkm-y += nvkm/engine/disp/coreg94.o
 nvkm-y += nvkm/engine/disp/coregf119.o
 nvkm-y += nvkm/engine/disp/coregk104.o
 nvkm-y += nvkm/engine/disp/coregp102.o
+nvkm-y += nvkm/engine/disp/coregv100.o
 
 nvkm-y += nvkm/engine/disp/ovlynv50.o
 nvkm-y += nvkm/engine/disp/ovlyg84.o
@@ -90,12 +98,17 @@ nvkm-y += nvkm/engine/disp/ovlygf119.o
 nvkm-y += nvkm/engine/disp/ovlygk104.o
 nvkm-y += nvkm/engine/disp/ovlygp102.o
 
+nvkm-y += nvkm/engine/disp/wimmgv100.o
+
+nvkm-y += nvkm/engine/disp/wndwgv100.o
+
 nvkm-y += nvkm/engine/disp/piocnv50.o
 nvkm-y += nvkm/engine/disp/piocgf119.o
 
 nvkm-y += nvkm/engine/disp/cursnv50.o
 nvkm-y += nvkm/engine/disp/cursgf119.o
 nvkm-y += nvkm/engine/disp/cursgp102.o
+nvkm-y += nvkm/engine/disp/cursgv100.o
 
 nvkm-y += nvkm/engine/disp/oimmnv50.o
 nvkm-y += nvkm/engine/disp/oimmgf119.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/changv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/changv100.c
new file mode 100644
index 0000000..75247c9
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/changv100.c
@@ -0,0 +1,34 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "channv50.h"
+
+const struct nvkm_event_func
+gv100_disp_chan_uevent = {
+	.ctor = nv50_disp_chan_uevent_ctor,
+};
+
+u64
+gv100_disp_chan_user(struct nv50_disp_chan *chan, u64 *psize)
+{
+	*psize = 0x1000;
+	return 0x690000 + ((chan->chid.user - 1) * 0x1000);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
index 391b007..adc9d76 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
@@ -57,6 +57,11 @@ void gf119_disp_core_fini(struct nv50_disp_chan *);
 
 extern const struct nv50_disp_chan_func gp102_disp_dmac_func;
 
+u64 gv100_disp_chan_user(struct nv50_disp_chan *, u64 *);
+int gv100_disp_dmac_init(struct nv50_disp_chan *);
+void gv100_disp_dmac_fini(struct nv50_disp_chan *);
+int gv100_disp_dmac_bind(struct nv50_disp_chan *, struct nvkm_object *, u32);
+
 int nv50_disp_curs_new_(const struct nv50_disp_chan_func *,
 			struct nv50_disp *, int ctrl, int user,
 			const struct nvkm_oclass *, void *argv, u32 argc,
@@ -132,6 +137,15 @@ int gp102_disp_core_new(const struct nvkm_oclass *, void *, u32,
 int gp102_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
 			struct nv50_disp *, struct nvkm_object **);
 
+int gv100_disp_curs_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
+int gv100_disp_wimm_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
+int gv100_disp_core_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
+int gv100_disp_wndw_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
+
 struct nv50_disp_mthd_list {
 	u32 mthd;
 	u32 addr;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregv100.c
new file mode 100644
index 0000000..4592d0e
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregv100.c
@@ -0,0 +1,204 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "channv50.h"
+
+#include <subdev/timer.h>
+
+const struct nv50_disp_mthd_list
+gv100_disp_core_mthd_base = {
+	.mthd = 0x0000,
+	.addr = 0x000000,
+	.data = {
+		{ 0x0200, 0x680200 },
+		{ 0x0208, 0x680208 },
+		{ 0x020c, 0x68020c },
+		{ 0x0210, 0x680210 },
+		{ 0x0214, 0x680214 },
+		{ 0x0218, 0x680218 },
+		{ 0x021c, 0x68021c },
+		{}
+	}
+};
+
+const struct nv50_disp_mthd_list
+gv100_disp_core_mthd_sor = {
+	.mthd = 0x0020,
+	.addr = 0x000020,
+	.data = {
+		{ 0x0300, 0x680300 },
+		{ 0x0304, 0x680304 },
+		{ 0x0308, 0x680308 },
+		{ 0x030c, 0x68030c },
+		{}
+	}
+};
+
+static const struct nv50_disp_mthd_list
+gv100_disp_core_mthd_wndw = {
+	.mthd = 0x0080,
+	.addr = 0x000080,
+	.data = {
+		{ 0x1000, 0x681000 },
+		{ 0x1004, 0x681004 },
+		{ 0x1008, 0x681008 },
+		{ 0x100c, 0x68100c },
+		{ 0x1010, 0x681010 },
+		{}
+	}
+};
+
+static const struct nv50_disp_mthd_list
+gv100_disp_core_mthd_head = {
+	.mthd = 0x0400,
+	.addr = 0x000400,
+	.data = {
+		{ 0x2000, 0x682000 },
+		{ 0x2004, 0x682004 },
+		{ 0x2008, 0x682008 },
+		{ 0x200c, 0x68200c },
+		{ 0x2014, 0x682014 },
+		{ 0x2018, 0x682018 },
+		{ 0x201c, 0x68201c },
+		{ 0x2020, 0x682020 },
+		{ 0x2028, 0x682028 },
+		{ 0x202c, 0x68202c },
+		{ 0x2030, 0x682030 },
+		{ 0x2038, 0x682038 },
+		{ 0x203c, 0x68203c },
+		{ 0x2048, 0x682048 },
+		{ 0x204c, 0x68204c },
+		{ 0x2050, 0x682050 },
+		{ 0x2054, 0x682054 },
+		{ 0x2058, 0x682058 },
+		{ 0x205c, 0x68205c },
+		{ 0x2060, 0x682060 },
+		{ 0x2064, 0x682064 },
+		{ 0x2068, 0x682068 },
+		{ 0x206c, 0x68206c },
+		{ 0x2070, 0x682070 },
+		{ 0x2074, 0x682074 },
+		{ 0x2078, 0x682078 },
+		{ 0x207c, 0x68207c },
+		{ 0x2080, 0x682080 },
+		{ 0x2088, 0x682088 },
+		{ 0x2090, 0x682090 },
+		{ 0x209c, 0x68209c },
+		{ 0x20a0, 0x6820a0 },
+		{ 0x20a4, 0x6820a4 },
+		{ 0x20a8, 0x6820a8 },
+		{ 0x20ac, 0x6820ac },
+		{ 0x218c, 0x68218c },
+		{ 0x2194, 0x682194 },
+		{ 0x2198, 0x682198 },
+		{ 0x219c, 0x68219c },
+		{ 0x21a0, 0x6821a0 },
+		{ 0x21a4, 0x6821a4 },
+		{ 0x2214, 0x682214 },
+		{ 0x2218, 0x682218 },
+		{}
+	}
+};
+
+static const struct nv50_disp_chan_mthd
+gv100_disp_core_mthd = {
+	.name = "Core",
+	.addr = 0x000000,
+	.prev = 0x008000,
+	.data = {
+		{ "Global", 1, &gv100_disp_core_mthd_base },
+		{    "SOR", 4, &gv100_disp_core_mthd_sor  },
+		{ "WINDOW", 8, &gv100_disp_core_mthd_wndw },
+		{   "HEAD", 4, &gv100_disp_core_mthd_head },
+		{}
+	}
+};
+
+static int
+gv100_disp_core_idle(struct nv50_disp_chan *chan)
+{
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	nvkm_msec(device, 2000,
+		u32 stat = nvkm_rd32(device, 0x610630);
+		if ((stat & 0x001f0000) == 0x000b0000)
+			return 0;
+	);
+	return -EBUSY;
+}
+
+static u64
+gv100_disp_core_user(struct nv50_disp_chan *chan, u64 *psize)
+{
+	*psize = 0x10000;
+	return 0x680000;
+}
+
+static void
+gv100_disp_core_intr(struct nv50_disp_chan *chan, bool en)
+{
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	const u32 mask = 0x00000001;
+	const u32 data = en ? mask : 0;
+	nvkm_mask(device, 0x611dac, mask, data);
+}
+
+static void
+gv100_disp_core_fini(struct nv50_disp_chan *chan)
+{
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	nvkm_mask(device, 0x6104e0, 0x00000010, 0x00000000);
+	gv100_disp_core_idle(chan);
+	nvkm_mask(device, 0x6104e0, 0x00000002, 0x00000000);
+}
+
+static int
+gv100_disp_core_init(struct nv50_disp_chan *chan)
+{
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+
+	nvkm_wr32(device, 0x610b24, lower_32_bits(chan->push));
+	nvkm_wr32(device, 0x610b20, upper_32_bits(chan->push));
+	nvkm_wr32(device, 0x610b28, 0x00000001);
+	nvkm_wr32(device, 0x610b2c, 0x00000040);
+
+	nvkm_mask(device, 0x6104e0, 0x00000010, 0x00000010);
+	nvkm_wr32(device, 0x680000, 0x00000000);
+	nvkm_wr32(device, 0x6104e0, 0x00000013);
+	return gv100_disp_core_idle(chan);
+}
+
+static const struct nv50_disp_chan_func
+gv100_disp_core = {
+	.init = gv100_disp_core_init,
+	.fini = gv100_disp_core_fini,
+	.intr = gv100_disp_core_intr,
+	.user = gv100_disp_core_user,
+	.bind = gv100_disp_dmac_bind,
+};
+
+int
+gv100_disp_core_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_core_new_(&gv100_disp_core, &gv100_disp_core_mthd,
+				   disp, 0, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgv100.c
new file mode 100644
index 0000000..a3e4f69
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgv100.c
@@ -0,0 +1,81 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "channv50.h"
+
+#include <subdev/timer.h>
+
+static int
+gv100_disp_curs_idle(struct nv50_disp_chan *chan)
+{
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	const u32 soff = (chan->chid.ctrl - 1) * 0x04;
+	nvkm_msec(device, 2000,
+		u32 stat = nvkm_rd32(device, 0x610664 + soff);
+		if ((stat & 0x00070000) == 0x00040000)
+			return 0;
+	);
+	return -EBUSY;
+}
+
+static void
+gv100_disp_curs_intr(struct nv50_disp_chan *chan, bool en)
+{
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	const u32 mask = 0x00010000 << chan->head;
+	const u32 data = en ? mask : 0;
+	nvkm_mask(device, 0x611dac, mask, data);
+}
+
+static void
+gv100_disp_curs_fini(struct nv50_disp_chan *chan)
+{
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	const u32 hoff = chan->chid.ctrl * 4;
+	nvkm_mask(device, 0x6104e0 + hoff, 0x00000010, 0x00000010);
+	gv100_disp_curs_idle(chan);
+	nvkm_mask(device, 0x6104e0 + hoff, 0x00000001, 0x00000000);
+}
+
+static int
+gv100_disp_curs_init(struct nv50_disp_chan *chan)
+{
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	nvkm_wr32(device, 0x6104e0 + chan->chid.ctrl * 4, 0x00000001);
+	return gv100_disp_curs_idle(chan);
+}
+
+static const struct nv50_disp_chan_func
+gv100_disp_curs = {
+	.init = gv100_disp_curs_init,
+	.fini = gv100_disp_curs_fini,
+	.intr = gv100_disp_curs_intr,
+	.user = gv100_disp_chan_user,
+};
+
+int
+gv100_disp_curs_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_curs_new_(&gv100_disp_curs, disp, 73, 73,
+				   oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgv100.c
new file mode 100644
index 0000000..eac0e42
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgv100.c
@@ -0,0 +1,77 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "channv50.h"
+
+#include <core/ramht.h>
+#include <subdev/timer.h>
+
+static int
+gv100_disp_dmac_idle(struct nv50_disp_chan *chan)
+{
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	const u32 soff = (chan->chid.ctrl - 1) * 0x04;
+	nvkm_msec(device, 2000,
+		u32 stat = nvkm_rd32(device, 0x610664 + soff);
+		if ((stat & 0x000f0000) == 0x00040000)
+			return 0;
+	);
+	return -EBUSY;
+}
+
+int
+gv100_disp_dmac_bind(struct nv50_disp_chan *chan,
+		     struct nvkm_object *object, u32 handle)
+{
+	return nvkm_ramht_insert(chan->disp->ramht, object,
+				 chan->chid.user, -9, handle,
+				 chan->chid.user << 25 | 0x00000040);
+}
+
+void
+gv100_disp_dmac_fini(struct nv50_disp_chan *chan)
+{
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	const u32 coff = chan->chid.ctrl * 0x04;
+	nvkm_mask(device, 0x6104e0 + coff, 0x00000010, 0x00000000);
+	gv100_disp_dmac_idle(chan);
+	nvkm_mask(device, 0x6104e0 + coff, 0x00000002, 0x00000000);
+}
+
+int
+gv100_disp_dmac_init(struct nv50_disp_chan *chan)
+{
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	const u32 uoff = (chan->chid.ctrl - 1) * 0x1000;
+	const u32 poff = chan->chid.ctrl * 0x10;
+	const u32 coff = chan->chid.ctrl * 0x04;
+
+	nvkm_wr32(device, 0x610b24 + poff, lower_32_bits(chan->push));
+	nvkm_wr32(device, 0x610b20 + poff, upper_32_bits(chan->push));
+	nvkm_wr32(device, 0x610b28 + poff, 0x00000001);
+	nvkm_wr32(device, 0x610b2c + poff, 0x00000040);
+
+	nvkm_mask(device, 0x6104e0 + coff, 0x00000010, 0x00000010);
+	nvkm_wr32(device, 0x690000 + uoff, 0x00000000);
+	nvkm_wr32(device, 0x6104e0 + coff, 0x00000013);
+	return gv100_disp_dmac_idle(chan);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gv100.c
new file mode 100644
index 0000000..d0a7e34
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gv100.c
@@ -0,0 +1,427 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "nv50.h"
+#include "head.h"
+#include "ior.h"
+#include "channv50.h"
+#include "rootnv50.h"
+
+#include <core/gpuobj.h>
+#include <subdev/timer.h>
+
+static int
+gv100_disp_wndw_cnt(struct nvkm_disp *disp, unsigned long *pmask)
+{
+	struct nvkm_device *device = disp->engine.subdev.device;
+	*pmask = nvkm_rd32(device, 0x610064);
+	return (nvkm_rd32(device, 0x610074) & 0x03f00000) >> 20;
+}
+
+static void
+gv100_disp_super(struct work_struct *work)
+{
+	struct nv50_disp *disp =
+		container_of(work, struct nv50_disp, supervisor);
+	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	struct nvkm_head *head;
+	u32 stat = nvkm_rd32(device, 0x6107a8);
+	u32 mask[4];
+
+	nvkm_debug(subdev, "supervisor %d: %08x\n", ffs(disp->super), stat);
+	list_for_each_entry(head, &disp->base.head, head) {
+		mask[head->id] = nvkm_rd32(device, 0x6107ac + (head->id * 4));
+		HEAD_DBG(head, "%08x", mask[head->id]);
+	}
+
+	if (disp->super & 0x00000001) {
+		nv50_disp_chan_mthd(disp->chan[0], NV_DBG_DEBUG);
+		nv50_disp_super_1(disp);
+		list_for_each_entry(head, &disp->base.head, head) {
+			if (!(mask[head->id] & 0x00001000))
+				continue;
+			nv50_disp_super_1_0(disp, head);
+		}
+	} else
+	if (disp->super & 0x00000002) {
+		list_for_each_entry(head, &disp->base.head, head) {
+			if (!(mask[head->id] & 0x00001000))
+				continue;
+			nv50_disp_super_2_0(disp, head);
+		}
+		nvkm_outp_route(&disp->base);
+		list_for_each_entry(head, &disp->base.head, head) {
+			if (!(mask[head->id] & 0x00010000))
+				continue;
+			nv50_disp_super_2_1(disp, head);
+		}
+		list_for_each_entry(head, &disp->base.head, head) {
+			if (!(mask[head->id] & 0x00001000))
+				continue;
+			nv50_disp_super_2_2(disp, head);
+		}
+	} else
+	if (disp->super & 0x00000004) {
+		list_for_each_entry(head, &disp->base.head, head) {
+			if (!(mask[head->id] & 0x00001000))
+				continue;
+			nv50_disp_super_3_0(disp, head);
+		}
+	}
+
+	list_for_each_entry(head, &disp->base.head, head)
+		nvkm_wr32(device, 0x6107ac + (head->id * 4), 0x00000000);
+	nvkm_wr32(device, 0x6107a8, 0x80000000);
+}
+
+static void
+gv100_disp_exception(struct nv50_disp *disp, int chid)
+{
+	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	u32 stat = nvkm_rd32(device, 0x611020 + (chid * 12));
+	u32 type = (stat & 0x00007000) >> 12;
+	u32 mthd = (stat & 0x00000fff) << 2;
+	u32 data = nvkm_rd32(device, 0x611024 + (chid * 12));
+	u32 code = nvkm_rd32(device, 0x611028 + (chid * 12));
+
+	nvkm_error(subdev, "chid %d %08x [type %d mthd %04x] "
+			   "data %08x code %08x\n",
+		   chid, stat, type, mthd, data, code);
+
+	if (chid < ARRAY_SIZE(disp->chan) && disp->chan[chid]) {
+		switch (mthd) {
+		case 0x0200:
+			nv50_disp_chan_mthd(disp->chan[chid], NV_DBG_ERROR);
+			break;
+		default:
+			break;
+		}
+	}
+
+	nvkm_wr32(device, 0x611020 + (chid * 12), 0x90000000);
+}
+
+static void
+gv100_disp_intr_ctrl_disp(struct nv50_disp *disp)
+{
+	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	u32 stat = nvkm_rd32(device, 0x611c30);
+
+	if (stat & 0x00000007) {
+		disp->super = (stat & 0x00000007);
+		queue_work(disp->wq, &disp->supervisor);
+		nvkm_wr32(device, 0x611860, disp->super);
+		stat &= ~0x00000007;
+	}
+
+	/*TODO: I would guess this is VBIOS_RELEASE, however, NFI how to
+	 *      ACK it, nor does RM appear to bother.
+	 */
+	if (stat & 0x00000008)
+		stat &= ~0x00000008;
+
+	if (stat & 0x00000100) {
+		unsigned long wndws = nvkm_rd32(device, 0x611858);
+		unsigned long other = nvkm_rd32(device, 0x61185c);
+		int wndw;
+
+		nvkm_wr32(device, 0x611858, wndws);
+		nvkm_wr32(device, 0x61185c, other);
+
+		/* AWAKEN_OTHER_CORE. */
+		if (other & 0x00000001)
+			nv50_disp_chan_uevent_send(disp, 0);
+
+		/* AWAKEN_WIN_CH(n). */
+		for_each_set_bit(wndw, &wndws, disp->wndw.nr) {
+			nv50_disp_chan_uevent_send(disp, 1 + wndw);
+		}
+	}
+
+	if (stat)
+		nvkm_warn(subdev, "ctrl %08x\n", stat);
+}
+
+static void
+gv100_disp_intr_exc_other(struct nv50_disp *disp)
+{
+	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	u32 stat = nvkm_rd32(device, 0x611854);
+	unsigned long mask;
+	int head;
+
+	if (stat & 0x00000001) {
+		nvkm_wr32(device, 0x611854, 0x00000001);
+		gv100_disp_exception(disp, 0);
+		stat &= ~0x00000001;
+	}
+
+	if ((mask = (stat & 0x00ff0000) >> 16)) {
+		for_each_set_bit(head, &mask, disp->wndw.nr) {
+			nvkm_wr32(device, 0x611854, 0x00010000 << head);
+			gv100_disp_exception(disp, 73 + head);
+			stat &= ~(0x00010000 << head);
+		}
+	}
+
+	if (stat) {
+		nvkm_warn(subdev, "exception %08x\n", stat);
+		nvkm_wr32(device, 0x611854, stat);
+	}
+}
+
+static void
+gv100_disp_intr_exc_winim(struct nv50_disp *disp)
+{
+	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	unsigned long stat = nvkm_rd32(device, 0x611850);
+	int wndw;
+
+	for_each_set_bit(wndw, &stat, disp->wndw.nr) {
+		nvkm_wr32(device, 0x611850, BIT(wndw));
+		gv100_disp_exception(disp, 33 + wndw);
+		stat &= ~BIT(wndw);
+	}
+
+	if (stat) {
+		nvkm_warn(subdev, "wimm %08x\n", (u32)stat);
+		nvkm_wr32(device, 0x611850, stat);
+	}
+}
+
+static void
+gv100_disp_intr_exc_win(struct nv50_disp *disp)
+{
+	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	unsigned long stat = nvkm_rd32(device, 0x61184c);
+	int wndw;
+
+	for_each_set_bit(wndw, &stat, disp->wndw.nr) {
+		nvkm_wr32(device, 0x61184c, BIT(wndw));
+		gv100_disp_exception(disp, 1 + wndw);
+		stat &= ~BIT(wndw);
+	}
+
+	if (stat) {
+		nvkm_warn(subdev, "wndw %08x\n", (u32)stat);
+		nvkm_wr32(device, 0x61184c, stat);
+	}
+}
+
+static void
+gv100_disp_intr_head_timing(struct nv50_disp *disp, int head)
+{
+	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	u32 stat = nvkm_rd32(device, 0x611800 + (head * 0x04));
+
+	/* LAST_DATA, LOADV. */
+	if (stat & 0x00000003) {
+		nvkm_wr32(device, 0x611800 + (head * 0x04), stat & 0x00000003);
+		stat &= ~0x00000003;
+	}
+
+	if (stat & 0x00000004) {
+		nvkm_disp_vblank(&disp->base, head);
+		nvkm_wr32(device, 0x611800 + (head * 0x04), 0x00000004);
+		stat &= ~0x00000004;
+	}
+
+	if (stat) {
+		nvkm_warn(subdev, "head %08x\n", stat);
+		nvkm_wr32(device, 0x611800 + (head * 0x04), stat);
+	}
+}
+
+static void
+gv100_disp_intr(struct nv50_disp *disp)
+{
+	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	u32 stat = nvkm_rd32(device, 0x611ec0);
+	unsigned long mask;
+	int head;
+
+	if ((mask = (stat & 0x000000ff))) {
+		for_each_set_bit(head, &mask, 8) {
+			gv100_disp_intr_head_timing(disp, head);
+			stat &= ~BIT(head);
+		}
+	}
+
+	if (stat & 0x00000200) {
+		gv100_disp_intr_exc_win(disp);
+		stat &= ~0x00000200;
+	}
+
+	if (stat & 0x00000400) {
+		gv100_disp_intr_exc_winim(disp);
+		stat &= ~0x00000400;
+	}
+
+	if (stat & 0x00000800) {
+		gv100_disp_intr_exc_other(disp);
+		stat &= ~0x00000800;
+	}
+
+	if (stat & 0x00001000) {
+		gv100_disp_intr_ctrl_disp(disp);
+		stat &= ~0x00001000;
+	}
+
+	if (stat)
+		nvkm_warn(subdev, "intr %08x\n", stat);
+}
+
+static void
+gv100_disp_fini(struct nv50_disp *disp)
+{
+	struct nvkm_device *device = disp->base.engine.subdev.device;
+	nvkm_wr32(device, 0x611db0, 0x00000000);
+}
+
+static int
+gv100_disp_init(struct nv50_disp *disp)
+{
+	struct nvkm_device *device = disp->base.engine.subdev.device;
+	struct nvkm_head *head;
+	int i, j;
+	u32 tmp;
+
+	/* Claim ownership of display. */
+	if (nvkm_rd32(device, 0x6254e8) & 0x00000002) {
+		nvkm_mask(device, 0x6254e8, 0x00000001, 0x00000000);
+		if (nvkm_msec(device, 2000,
+			if (!(nvkm_rd32(device, 0x6254e8) & 0x00000002))
+				break;
+		) < 0)
+			return -EBUSY;
+	}
+
+	/* Lock pin capabilities. */
+	tmp = nvkm_rd32(device, 0x610068);
+	nvkm_wr32(device, 0x640008, tmp);
+
+	/* SOR capabilities. */
+	for (i = 0; i < disp->sor.nr; i++) {
+		tmp = nvkm_rd32(device, 0x61c000 + (i * 0x800));
+		nvkm_mask(device, 0x640000, 0x00000100 << i, 0x00000100 << i);
+		nvkm_wr32(device, 0x640144 + (i * 0x08), tmp);
+	}
+
+	/* Head capabilities. */
+	list_for_each_entry(head, &disp->base.head, head) {
+		const int id = head->id;
+
+		/* RG. */
+		tmp = nvkm_rd32(device, 0x616300 + (id * 0x800));
+		nvkm_wr32(device, 0x640048 + (id * 0x020), tmp);
+
+		/* POSTCOMP. */
+		for (j = 0; j < 6 * 4; j += 4) {
+			tmp = nvkm_rd32(device, 0x616100 + (id * 0x800) + j);
+			nvkm_wr32(device, 0x640030 + (id * 0x20) + j, tmp);
+		}
+	}
+
+	/* Window capabilities. */
+	for (i = 0; i < disp->wndw.nr; i++) {
+		nvkm_mask(device, 0x640004, 1 << i, 1 << i);
+		for (j = 0; j < 6 * 4; j += 4) {
+			tmp = nvkm_rd32(device, 0x630050 + (i * 0x800) + j);
+			nvkm_wr32(device, 0x6401e4 + (i * 0x20) + j, tmp);
+		}
+	}
+
+	/* IHUB capabilities. */
+	for (i = 0; i < 4; i++) {
+		tmp = nvkm_rd32(device, 0x62e000 + (i * 0x04));
+		nvkm_wr32(device, 0x640010 + (i * 0x04), tmp);
+	}
+
+	nvkm_mask(device, 0x610078, 0x00000001, 0x00000001);
+
+	/* Setup instance memory. */
+	switch (nvkm_memory_target(disp->inst->memory)) {
+	case NVKM_MEM_TARGET_VRAM: tmp = 0x00000001; break;
+	case NVKM_MEM_TARGET_NCOH: tmp = 0x00000002; break;
+	case NVKM_MEM_TARGET_HOST: tmp = 0x00000003; break;
+	default:
+		break;
+	}
+	nvkm_wr32(device, 0x610010, 0x00000008 | tmp);
+	nvkm_wr32(device, 0x610014, disp->inst->addr >> 16);
+
+	/* CTRL_DISP: AWAKEN, ERROR, SUPERVISOR[1-3]. */
+	nvkm_wr32(device, 0x611cf0, 0x00000187); /* MSK. */
+	nvkm_wr32(device, 0x611db0, 0x00000187); /* EN. */
+
+	/* EXC_OTHER: CURSn, CORE. */
+	nvkm_wr32(device, 0x611cec, disp->head.mask << 16 |
+				    0x00000001); /* MSK. */
+	nvkm_wr32(device, 0x611dac, 0x00000000); /* EN. */
+
+	/* EXC_WINIM. */
+	nvkm_wr32(device, 0x611ce8, disp->wndw.mask); /* MSK. */
+	nvkm_wr32(device, 0x611da8, 0x00000000); /* EN. */
+
+	/* EXC_WIN. */
+	nvkm_wr32(device, 0x611ce4, disp->wndw.mask); /* MSK. */
+	nvkm_wr32(device, 0x611da4, 0x00000000); /* EN. */
+
+	/* HEAD_TIMING(n): VBLANK. */
+	list_for_each_entry(head, &disp->base.head, head) {
+		const u32 hoff = head->id * 4;
+		nvkm_wr32(device, 0x611cc0 + hoff, 0x00000004); /* MSK. */
+		nvkm_wr32(device, 0x611d80 + hoff, 0x00000000); /* EN. */
+	}
+
+	/* OR. */
+	nvkm_wr32(device, 0x611cf4, 0x00000000); /* MSK. */
+	nvkm_wr32(device, 0x611db4, 0x00000000); /* EN. */
+	return 0;
+}
+
+static const struct nv50_disp_func
+gv100_disp = {
+	.init = gv100_disp_init,
+	.fini = gv100_disp_fini,
+	.intr = gv100_disp_intr,
+	.uevent = &gv100_disp_chan_uevent,
+	.super = gv100_disp_super,
+	.root = &gv100_disp_root_oclass,
+	.wndw = { .cnt = gv100_disp_wndw_cnt },
+	.head = { .cnt = gv100_head_cnt, .new = gv100_head_new },
+	.sor = { .cnt = gv100_sor_cnt, .new = gv100_sor_new },
+	.ramht_size = 0x2000,
+};
+
+int
+gv100_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
+{
+	return nv50_disp_new_(&gv100_disp, device, index, pdisp);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/hdmigv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/hdmigv100.c
new file mode 100644
index 0000000..6e3c450
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/hdmigv100.c
@@ -0,0 +1,85 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "hdmi.h"
+
+void
+gv100_hdmi_ctrl(struct nvkm_ior *ior, int head, bool enable, u8 max_ac_packet,
+		u8 rekey, u8 *avi, u8 avi_size, u8 *vendor, u8 vendor_size)
+{
+	struct nvkm_device *device = ior->disp->engine.subdev.device;
+	const u32 ctrl = 0x40000000 * enable |
+			 max_ac_packet << 16 |
+			 rekey;
+	const u32 hoff = head * 0x800;
+	const u32 hdmi = head * 0x400;
+	struct packed_hdmi_infoframe avi_infoframe;
+	struct packed_hdmi_infoframe vendor_infoframe;
+
+	pack_hdmi_infoframe(&avi_infoframe, avi, avi_size);
+	pack_hdmi_infoframe(&vendor_infoframe, vendor, vendor_size);
+
+	if (!(ctrl & 0x40000000)) {
+		nvkm_mask(device, 0x6165c0 + hoff, 0x40000000, 0x00000000);
+		nvkm_mask(device, 0x6f0100 + hdmi, 0x00000001, 0x00000000);
+		nvkm_mask(device, 0x6f00c0 + hdmi, 0x00000001, 0x00000000);
+		nvkm_mask(device, 0x6f0000 + hdmi, 0x00000001, 0x00000000);
+		return;
+	}
+
+	/* AVI InfoFrame (AVI). */
+	nvkm_mask(device, 0x6f0000 + hdmi, 0x00000001, 0x00000000);
+	if (avi_size) {
+		nvkm_wr32(device, 0x6f0008 + hdmi, avi_infoframe.header);
+		nvkm_wr32(device, 0x6f000c + hdmi, avi_infoframe.subpack0_low);
+		nvkm_wr32(device, 0x6f0010 + hdmi, avi_infoframe.subpack0_high);
+		nvkm_wr32(device, 0x6f0014 + hdmi, avi_infoframe.subpack1_low);
+		nvkm_wr32(device, 0x6f0018 + hdmi, avi_infoframe.subpack1_high);
+		nvkm_mask(device, 0x6f0000 + hdmi, 0x00000001, 0x00000001);
+	}
+
+	/* Vendor-specific InfoFrame (VSI). */
+	nvkm_mask(device, 0x6f0100 + hdmi, 0x00010001, 0x00000000);
+	if (vendor_size) {
+		nvkm_wr32(device, 0x6f0108 + hdmi, vendor_infoframe.header);
+		nvkm_wr32(device, 0x6f010c + hdmi, vendor_infoframe.subpack0_low);
+		nvkm_wr32(device, 0x6f0110 + hdmi, vendor_infoframe.subpack0_high);
+		nvkm_wr32(device, 0x6f0110 + hdmi, 0x00000000);
+		nvkm_wr32(device, 0x6f0114 + hdmi, 0x00000000);
+		nvkm_wr32(device, 0x6f0118 + hdmi, 0x00000000);
+		nvkm_wr32(device, 0x6f011c + hdmi, 0x00000000);
+		nvkm_wr32(device, 0x6f0120 + hdmi, 0x00000000);
+		nvkm_wr32(device, 0x6f0124 + hdmi, 0x00000000);
+		nvkm_mask(device, 0x6f0100 + hdmi, 0x00000001, 0x00000001);
+	}
+
+
+	/* General Control (GCP). */
+	nvkm_mask(device, 0x6f00c0 + hdmi, 0x00000001, 0x00000000);
+	nvkm_wr32(device, 0x6f00cc + hdmi, 0x00000010);
+	nvkm_mask(device, 0x6f00c0 + hdmi, 0x00000001, 0x00000001);
+
+	/* Audio Clock Regeneration (ACR). */
+	nvkm_wr32(device, 0x6f0080 + hdmi, 0x82000000);
+
+	/* NV_PDISP_SF_HDMI_CTRL. */
+	nvkm_mask(device, 0x6165c0 + hoff, 0x401f007f, ctrl);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/head.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/head.h
index 4a5d789..7d55faf 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/head.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/head.h
@@ -58,4 +58,8 @@ int nv50_head_new(struct nvkm_disp *, int id);
 
 int gf119_head_cnt(struct nvkm_disp *, unsigned long *);
 int gf119_head_new(struct nvkm_disp *, int id);
+void gf119_head_rgclk(struct nvkm_head *, int);
+
+int gv100_head_cnt(struct nvkm_disp *, unsigned long *);
+int gv100_head_new(struct nvkm_disp *, int id);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/headgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/headgf119.c
index bcbdaaf..e86298b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/headgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/headgf119.c
@@ -39,7 +39,7 @@ gf119_head_vblank_get(struct nvkm_head *head)
 	nvkm_mask(device, 0x6100c0 + hoff, 0x00000001, 0x00000001);
 }
 
-static void
+void
 gf119_head_rgclk(struct nvkm_head *head, int div)
 {
 	struct nvkm_device *device = head->disp->engine.subdev.device;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/headgv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/headgv100.c
new file mode 100644
index 0000000..1a061b4
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/headgv100.c
@@ -0,0 +1,105 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "head.h"
+
+static void
+gv100_head_vblank_put(struct nvkm_head *head)
+{
+	struct nvkm_device *device = head->disp->engine.subdev.device;
+	nvkm_mask(device, 0x611d80 + (head->id * 4), 0x00000004, 0x00000000);
+}
+
+static void
+gv100_head_vblank_get(struct nvkm_head *head)
+{
+	struct nvkm_device *device = head->disp->engine.subdev.device;
+	nvkm_mask(device, 0x611d80 + (head->id * 4), 0x00000004, 0x00000004);
+}
+
+static void
+gv100_head_rgpos(struct nvkm_head *head, u16 *hline, u16 *vline)
+{
+	struct nvkm_device *device = head->disp->engine.subdev.device;
+	const u32 hoff = head->id * 0x800;
+	/* vline read locks hline. */
+	*vline = nvkm_rd32(device, 0x616330 + hoff) & 0x0000ffff;
+	*hline = nvkm_rd32(device, 0x616334 + hoff) & 0x0000ffff;
+}
+
+static void
+gv100_head_state(struct nvkm_head *head, struct nvkm_head_state *state)
+{
+	struct nvkm_device *device = head->disp->engine.subdev.device;
+	const u32 hoff = (state == &head->arm) * 0x8000 + head->id * 0x400;
+	u32 data;
+
+	data = nvkm_rd32(device, 0x682064 + hoff);
+	state->vtotal = (data & 0xffff0000) >> 16;
+	state->htotal = (data & 0x0000ffff);
+	data = nvkm_rd32(device, 0x682068 + hoff);
+	state->vsynce = (data & 0xffff0000) >> 16;
+	state->hsynce = (data & 0x0000ffff);
+	data = nvkm_rd32(device, 0x68206c + hoff);
+	state->vblanke = (data & 0xffff0000) >> 16;
+	state->hblanke = (data & 0x0000ffff);
+	data = nvkm_rd32(device, 0x682070 + hoff);
+	state->vblanks = (data & 0xffff0000) >> 16;
+	state->hblanks = (data & 0x0000ffff);
+	state->hz = nvkm_rd32(device, 0x68200c + hoff);
+
+	data = nvkm_rd32(device, 0x682004 + hoff);
+	switch ((data & 0x000000f0) >> 4) {
+	case 5: state->or.depth = 30; break;
+	case 4: state->or.depth = 24; break;
+	case 1: state->or.depth = 18; break;
+	default:
+		state->or.depth = 18;
+		WARN_ON(1);
+		break;
+	}
+}
+
+static const struct nvkm_head_func
+gv100_head = {
+	.state = gv100_head_state,
+	.rgpos = gv100_head_rgpos,
+	.rgclk = gf119_head_rgclk,
+	.vblank_get = gv100_head_vblank_get,
+	.vblank_put = gv100_head_vblank_put,
+};
+
+int
+gv100_head_new(struct nvkm_disp *disp, int id)
+{
+	struct nvkm_device *device = disp->engine.subdev.device;
+	if (!(nvkm_rd32(device, 0x610060) & (0x00000001 << id)))
+		return 0;
+	return nvkm_head_new_(&gv100_head, disp, id);
+}
+
+int
+gv100_head_cnt(struct nvkm_disp *disp, unsigned long *pmask)
+{
+	struct nvkm_device *device = disp->engine.subdev.device;
+	*pmask = nvkm_rd32(device, 0x610060) & 0x000000ff;
+	return nvkm_rd32(device, 0x610074) & 0x0000000f;
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ior.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ior.h
index 9d43ab2..e0b4e0c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ior.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ior.h
@@ -30,7 +30,7 @@ struct nvkm_ior {
 			UNKNOWN
 		} proto:3;
 		unsigned link:2;
-		unsigned head:4;
+		unsigned head:8;
 	} arm, asy;
 
 	/* Armed DP state. */
@@ -133,10 +133,15 @@ void gf119_sor_dp_watermark(struct nvkm_ior *, int, u8);
 
 void gm107_sor_dp_pattern(struct nvkm_ior *, int);
 
+void gm200_sor_route_set(struct nvkm_outp *, struct nvkm_ior *);
+int gm200_sor_route_get(struct nvkm_outp *, int *);
+void gm200_sor_dp_drive(struct nvkm_ior *, int, int, int, int, int);
+
 void g84_hdmi_ctrl(struct nvkm_ior *, int, bool, u8, u8, u8 *, u8 , u8 *, u8);
 void gt215_hdmi_ctrl(struct nvkm_ior *, int, bool, u8, u8, u8 *, u8 , u8 *, u8);
 void gf119_hdmi_ctrl(struct nvkm_ior *, int, bool, u8, u8, u8 *, u8 , u8 *, u8);
 void gk104_hdmi_ctrl(struct nvkm_ior *, int, bool, u8, u8, u8 *, u8 , u8 *, u8);
+void gv100_hdmi_ctrl(struct nvkm_ior *, int, bool, u8, u8, u8 *, u8 , u8 *, u8);
 
 void gt215_hda_hpd(struct nvkm_ior *, int, bool);
 void gt215_hda_eld(struct nvkm_ior *, u8 *, u8);
@@ -178,4 +183,7 @@ int gf119_sor_new(struct nvkm_disp *, int);
 int gk104_sor_new(struct nvkm_disp *, int);
 int gm107_sor_new(struct nvkm_disp *, int);
 int gm200_sor_new(struct nvkm_disp *, int);
+
+int gv100_sor_cnt(struct nvkm_disp *, unsigned long *);
+int gv100_sor_new(struct nvkm_disp *, int);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
index 4a37c44..f89c7b9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
@@ -88,6 +88,12 @@ nv50_disp_oneinit_(struct nvkm_disp *base)
 	struct nvkm_device *device = subdev->device;
 	int ret, i;
 
+	if (func->wndw.cnt) {
+		disp->wndw.nr = func->wndw.cnt(&disp->base, &disp->wndw.mask);
+		nvkm_debug(subdev, "Window(s): %d (%08lx)\n",
+			   disp->wndw.nr, disp->wndw.mask);
+	}
+
 	disp->head.nr = func->head.cnt(&disp->base, &disp->head.mask);
 	nvkm_debug(subdev, "  Head(s): %d (%02lx)\n",
 		   disp->head.nr, disp->head.mask);
@@ -133,7 +139,8 @@ nv50_disp_oneinit_(struct nvkm_disp *base)
 	if (ret)
 		return ret;
 
-	return nvkm_ramht_new(device, 0x1000, 0, disp->inst, &disp->ramht);
+	return nvkm_ramht_new(device, func->ramht_size ? func->ramht_size :
+			      0x1000, 0, disp->inst, &disp->ramht);
 }
 
 static const struct nvkm_disp_func
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
index 77aa2c8..8580382 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
@@ -18,7 +18,7 @@ struct nv50_disp {
 	struct {
 		unsigned long mask;
 		int nr;
-	} head, dac;
+	} wndw, head, dac;
 
 	struct {
 		unsigned long mask;
@@ -35,7 +35,7 @@ struct nv50_disp {
 	struct nvkm_gpuobj *inst;
 	struct nvkm_ramht *ramht;
 
-	struct nv50_disp_chan *chan[21];
+	struct nv50_disp_chan *chan[81];
 };
 
 void nv50_disp_super_1(struct nv50_disp *);
@@ -62,7 +62,9 @@ struct nv50_disp_func {
 	struct {
 		int (*cnt)(struct nvkm_disp *, unsigned long *mask);
 		int (*new)(struct nvkm_disp *, int id);
-	} head, dac, sor, pior;
+	} wndw, head, dac, sor, pior;
+
+	u16 ramht_size;
 };
 
 int nv50_disp_init(struct nv50_disp *);
@@ -86,4 +88,5 @@ int  nv50_disp_chan_uevent_ctor(struct nvkm_object *, void *, u32,
 void nv50_disp_chan_uevent_send(struct nv50_disp *, int);
 
 extern const struct nvkm_event_func gf119_disp_chan_uevent;
+extern const struct nvkm_event_func gv100_disp_chan_uevent;
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgv100.c
new file mode 100644
index 0000000..9c658d6
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgv100.c
@@ -0,0 +1,52 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "rootnv50.h"
+#include "channv50.h"
+
+#include <nvif/class.h>
+
+static const struct nv50_disp_root_func
+gv100_disp_root = {
+	.user = {
+		{{0,0,GV100_DISP_CURSOR                }, gv100_disp_curs_new },
+		{{0,0,GV100_DISP_WINDOW_IMM_CHANNEL_DMA}, gv100_disp_wimm_new },
+		{{0,0,GV100_DISP_CORE_CHANNEL_DMA      }, gv100_disp_core_new },
+		{{0,0,GV100_DISP_WINDOW_CHANNEL_DMA    }, gv100_disp_wndw_new },
+		{}
+	},
+};
+
+static int
+gv100_disp_root_new(struct nvkm_disp *disp, const struct nvkm_oclass *oclass,
+		    void *data, u32 size, struct nvkm_object **pobject)
+{
+	return nv50_disp_root_new_(&gv100_disp_root, disp, oclass,
+				   data, size, pobject);
+}
+
+const struct nvkm_disp_oclass
+gv100_disp_root_oclass = {
+	.base.oclass = GV100_DISP,
+	.base.minver = -1,
+	.base.maxver = -1,
+	.ctor = gv100_disp_root_new,
+};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
index 9983a42..6ca4f91 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
@@ -36,4 +36,5 @@ extern const struct nvkm_disp_oclass gm107_disp_root_oclass;
 extern const struct nvkm_disp_oclass gm200_disp_root_oclass;
 extern const struct nvkm_disp_oclass gp100_disp_root_oclass;
 extern const struct nvkm_disp_oclass gp102_disp_root_oclass;
+extern const struct nvkm_disp_oclass gv100_disp_root_oclass;
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm200.c
index 8bc019b..d892bdf 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm200.c
@@ -23,7 +23,7 @@
  */
 #include "ior.h"
 
-static void
+void
 gm200_sor_dp_drive(struct nvkm_ior *sor, int ln, int pc, int dc, int pe, int pu)
 {
 	struct nvkm_device *device = sor->disp->engine.subdev.device;
@@ -45,7 +45,7 @@ gm200_sor_dp_drive(struct nvkm_ior *sor, int ln, int pc, int dc, int pe, int pu)
 	nvkm_wr32(device, 0x61c13c + loff, data[3] | (pc << shift));
 }
 
-static void
+void
 gm200_sor_route_set(struct nvkm_outp *outp, struct nvkm_ior *ior)
 {
 	struct nvkm_device *device = outp->disp->engine.subdev.device;
@@ -62,7 +62,7 @@ gm200_sor_route_set(struct nvkm_outp *outp, struct nvkm_ior *ior)
 		nvkm_mask(device, 0x612388 + moff, 0x0000001f, link << 4 | sor);
 }
 
-static int
+int
 gm200_sor_route_get(struct nvkm_outp *outp, int *link)
 {
 	struct nvkm_device *device = outp->disp->engine.subdev.device;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgv100.c
new file mode 100644
index 0000000..040db8a
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgv100.c
@@ -0,0 +1,120 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "ior.h"
+
+#include <subdev/timer.h>
+
+static void
+gv100_sor_dp_watermark(struct nvkm_ior *sor, int head, u8 watermark)
+{
+	struct nvkm_device *device = sor->disp->engine.subdev.device;
+	const u32 hoff = head * 0x800;
+	nvkm_mask(device, 0x616550 + hoff, 0x0c00003f, 0x08000000 | watermark);
+}
+
+static void
+gv100_sor_dp_audio_sym(struct nvkm_ior *sor, int head, u16 h, u32 v)
+{
+	struct nvkm_device *device = sor->disp->engine.subdev.device;
+	const u32 hoff = head * 0x800;
+	nvkm_mask(device, 0x616568 + hoff, 0x0000ffff, h);
+	nvkm_mask(device, 0x61656c + hoff, 0x00ffffff, v);
+}
+
+static void
+gv100_sor_dp_audio(struct nvkm_ior *sor, int head, bool enable)
+{
+	struct nvkm_device *device = sor->disp->engine.subdev.device;
+	const u32 hoff = 0x800 * head;
+	const u32 data = 0x80000000 | (0x00000001 * enable);
+	const u32 mask = 0x8000000d;
+	nvkm_mask(device, 0x616560 + hoff, mask, data);
+	nvkm_msec(device, 2000,
+		if (!(nvkm_rd32(device, 0x616560 + hoff) & 0x80000000))
+			break;
+	);
+}
+
+static void
+gv100_sor_state(struct nvkm_ior *sor, struct nvkm_ior_state *state)
+{
+	struct nvkm_device *device = sor->disp->engine.subdev.device;
+	const u32 coff = (state == &sor->arm) * 0x8000 + sor->id * 0x20;
+	u32 ctrl = nvkm_rd32(device, 0x680300 + coff);
+
+	state->proto_evo = (ctrl & 0x00000f00) >> 8;
+	switch (state->proto_evo) {
+	case 0: state->proto = LVDS; state->link = 1; break;
+	case 1: state->proto = TMDS; state->link = 1; break;
+	case 2: state->proto = TMDS; state->link = 2; break;
+	case 5: state->proto = TMDS; state->link = 3; break;
+	case 8: state->proto =   DP; state->link = 1; break;
+	case 9: state->proto =   DP; state->link = 2; break;
+	default:
+		state->proto = UNKNOWN;
+		break;
+	}
+
+	state->head = ctrl & 0x000000ff;
+}
+
+static const struct nvkm_ior_func
+gv100_sor = {
+	.route = {
+		.get = gm200_sor_route_get,
+		.set = gm200_sor_route_set,
+	},
+	.state = gv100_sor_state,
+	.power = nv50_sor_power,
+	.clock = gf119_sor_clock,
+	.hdmi = {
+		.ctrl = gv100_hdmi_ctrl,
+	},
+	.dp = {
+		.lanes = { 0, 1, 2, 3 },
+		.links = gf119_sor_dp_links,
+		.power = g94_sor_dp_power,
+		.pattern = gm107_sor_dp_pattern,
+		.drive = gm200_sor_dp_drive,
+		.audio = gv100_sor_dp_audio,
+		.audio_sym = gv100_sor_dp_audio_sym,
+		.watermark = gv100_sor_dp_watermark,
+	},
+	.hda = {
+		.hpd = gf119_hda_hpd,
+		.eld = gf119_hda_eld,
+	},
+};
+
+int
+gv100_sor_new(struct nvkm_disp *disp, int id)
+{
+	return nvkm_ior_new_(&gv100_sor, disp, SOR, id);
+}
+
+int
+gv100_sor_cnt(struct nvkm_disp *disp, unsigned long *pmask)
+{
+	struct nvkm_device *device = disp->engine.subdev.device;
+	*pmask = (nvkm_rd32(device, 0x610060) & 0x0000ff00) >> 8;
+	return (nvkm_rd32(device, 0x610074) & 0x00000f00) >> 8;
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/wimmgv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/wimmgv100.c
new file mode 100644
index 0000000..89d7833
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/wimmgv100.c
@@ -0,0 +1,82 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "channv50.h"
+
+#include <core/client.h>
+
+#include <nvif/clc37b.h>
+#include <nvif/unpack.h>
+
+static void
+gv100_disp_wimm_intr(struct nv50_disp_chan *chan, bool en)
+{
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	const u32 mask = 0x00000001 << chan->head;
+	const u32 data = en ? mask : 0;
+	nvkm_mask(device, 0x611da8, mask, data);
+}
+
+const struct nv50_disp_chan_func
+gv100_disp_wimm = {
+	.init = gv100_disp_dmac_init,
+	.fini = gv100_disp_dmac_fini,
+	.intr = gv100_disp_wimm_intr,
+	.user = gv100_disp_chan_user,
+};
+
+static int
+gv100_disp_wimm_new_(const struct nv50_disp_chan_func *func,
+		     const struct nv50_disp_chan_mthd *mthd,
+		     struct nv50_disp *disp, int chid,
+		     const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		     struct nvkm_object **pobject)
+{
+	union {
+		struct nvc37b_window_imm_channel_dma_v0 v0;
+	} *args = argv;
+	struct nvkm_object *parent = oclass->parent;
+	int wndw, ret = -ENOSYS;
+	u64 push;
+
+	nvif_ioctl(parent, "create window imm channel dma size %d\n", argc);
+	if (!(ret = nvif_unpack(ret, &argv, &argc, args->v0, 0, 0, false))) {
+		nvif_ioctl(parent, "create window imm channel dma vers %d "
+				   "pushbuf %016llx index %d\n",
+			   args->v0.version, args->v0.pushbuf, args->v0.index);
+		if (!(disp->wndw.mask & BIT(args->v0.index)))
+			return -EINVAL;
+		push = args->v0.pushbuf;
+		wndw = args->v0.index;
+	} else
+		return ret;
+
+	return nv50_disp_dmac_new_(func, mthd, disp, chid + wndw,
+				   wndw, push, oclass, pobject);
+}
+
+int
+gv100_disp_wimm_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return gv100_disp_wimm_new_(&gv100_disp_wimm, NULL, disp, 33,
+				    oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/wndwgv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/wndwgv100.c
new file mode 100644
index 0000000..9891180
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/wndwgv100.c
@@ -0,0 +1,184 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "channv50.h"
+
+#include <core/client.h>
+
+#include <nvif/clc37e.h>
+#include <nvif/unpack.h>
+
+static const struct nv50_disp_mthd_list
+gv100_disp_wndw_mthd_base = {
+	.mthd = 0x0000,
+	.addr = 0x000000,
+	.data = {
+		{ 0x0200, 0x690200 },
+		{ 0x020c, 0x69020c },
+		{ 0x0210, 0x690210 },
+		{ 0x0214, 0x690214 },
+		{ 0x0218, 0x690218 },
+		{ 0x021c, 0x69021c },
+		{ 0x0220, 0x690220 },
+		{ 0x0224, 0x690224 },
+		{ 0x0228, 0x690228 },
+		{ 0x022c, 0x69022c },
+		{ 0x0230, 0x690230 },
+		{ 0x0234, 0x690234 },
+		{ 0x0238, 0x690238 },
+		{ 0x0240, 0x690240 },
+		{ 0x0244, 0x690244 },
+		{ 0x0248, 0x690248 },
+		{ 0x024c, 0x69024c },
+		{ 0x0250, 0x690250 },
+		{ 0x0254, 0x690254 },
+		{ 0x0260, 0x690260 },
+		{ 0x0264, 0x690264 },
+		{ 0x0268, 0x690268 },
+		{ 0x026c, 0x69026c },
+		{ 0x0270, 0x690270 },
+		{ 0x0274, 0x690274 },
+		{ 0x0280, 0x690280 },
+		{ 0x0284, 0x690284 },
+		{ 0x0288, 0x690288 },
+		{ 0x028c, 0x69028c },
+		{ 0x0290, 0x690290 },
+		{ 0x0298, 0x690298 },
+		{ 0x029c, 0x69029c },
+		{ 0x02a0, 0x6902a0 },
+		{ 0x02a4, 0x6902a4 },
+		{ 0x02a8, 0x6902a8 },
+		{ 0x02ac, 0x6902ac },
+		{ 0x02b0, 0x6902b0 },
+		{ 0x02b4, 0x6902b4 },
+		{ 0x02b8, 0x6902b8 },
+		{ 0x02bc, 0x6902bc },
+		{ 0x02c0, 0x6902c0 },
+		{ 0x02c4, 0x6902c4 },
+		{ 0x02c8, 0x6902c8 },
+		{ 0x02cc, 0x6902cc },
+		{ 0x02d0, 0x6902d0 },
+		{ 0x02d4, 0x6902d4 },
+		{ 0x02d8, 0x6902d8 },
+		{ 0x02dc, 0x6902dc },
+		{ 0x02e0, 0x6902e0 },
+		{ 0x02e4, 0x6902e4 },
+		{ 0x02e8, 0x6902e8 },
+		{ 0x02ec, 0x6902ec },
+		{ 0x02f0, 0x6902f0 },
+		{ 0x02f4, 0x6902f4 },
+		{ 0x02f8, 0x6902f8 },
+		{ 0x02fc, 0x6902fc },
+		{ 0x0300, 0x690300 },
+		{ 0x0304, 0x690304 },
+		{ 0x0308, 0x690308 },
+		{ 0x0310, 0x690310 },
+		{ 0x0314, 0x690314 },
+		{ 0x0318, 0x690318 },
+		{ 0x031c, 0x69031c },
+		{ 0x0320, 0x690320 },
+		{ 0x0324, 0x690324 },
+		{ 0x0328, 0x690328 },
+		{ 0x032c, 0x69032c },
+		{ 0x033c, 0x69033c },
+		{ 0x0340, 0x690340 },
+		{ 0x0344, 0x690344 },
+		{ 0x0348, 0x690348 },
+		{ 0x034c, 0x69034c },
+		{ 0x0350, 0x690350 },
+		{ 0x0354, 0x690354 },
+		{ 0x0358, 0x690358 },
+		{ 0x0364, 0x690364 },
+		{ 0x0368, 0x690368 },
+		{ 0x036c, 0x69036c },
+		{ 0x0370, 0x690370 },
+		{ 0x0374, 0x690374 },
+		{ 0x0380, 0x690380 },
+		{}
+	}
+};
+
+const struct nv50_disp_chan_mthd
+gv100_disp_wndw_mthd = {
+	.name = "Base",
+	.addr = 0x001000,
+	.prev = 0x000800,
+	.data = {
+		{ "Global", 1, &gv100_disp_wndw_mthd_base },
+		{}
+	}
+};
+
+static void
+gv100_disp_wndw_intr(struct nv50_disp_chan *chan, bool en)
+{
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	const u32 mask = 0x00000001 << chan->head;
+	const u32 data = en ? mask : 0;
+	nvkm_mask(device, 0x611da4, mask, data);
+}
+
+const struct nv50_disp_chan_func
+gv100_disp_wndw = {
+	.init = gv100_disp_dmac_init,
+	.fini = gv100_disp_dmac_fini,
+	.intr = gv100_disp_wndw_intr,
+	.user = gv100_disp_chan_user,
+	.bind = gv100_disp_dmac_bind,
+};
+
+static int
+gv100_disp_wndw_new_(const struct nv50_disp_chan_func *func,
+		     const struct nv50_disp_chan_mthd *mthd,
+		     struct nv50_disp *disp, int chid,
+		     const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		     struct nvkm_object **pobject)
+{
+	union {
+		struct nvc37e_window_channel_dma_v0 v0;
+	} *args = argv;
+	struct nvkm_object *parent = oclass->parent;
+	int wndw, ret = -ENOSYS;
+	u64 push;
+
+	nvif_ioctl(parent, "create window channel dma size %d\n", argc);
+	if (!(ret = nvif_unpack(ret, &argv, &argc, args->v0, 0, 0, false))) {
+		nvif_ioctl(parent, "create window channel dma vers %d "
+				   "pushbuf %016llx index %d\n",
+			   args->v0.version, args->v0.pushbuf, args->v0.index);
+		if (!(disp->wndw.mask & BIT(args->v0.index)))
+			return -EINVAL;
+		push = args->v0.pushbuf;
+		wndw = args->v0.index;
+	} else
+		return ret;
+
+	return nv50_disp_dmac_new_(func, mthd, disp, chid + wndw,
+				   wndw, push, oclass, pobject);
+}
+
+int
+gv100_disp_wndw_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return gv100_disp_wndw_new_(&gv100_disp_wndw, &gv100_disp_wndw_mthd,
+				    disp, 1, oclass, argv, argc, pobject);
+}
-- 
cgit v1.1


From facaed62b4cba3a6334fc1798fa8f51ea6a1962d Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:48 +1000
Subject: drm/nouveau/kms/gv100: initial support

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/dispnv50/Kbuild     |   8 +
 drivers/gpu/drm/nouveau/dispnv50/atom.h     |   8 +-
 drivers/gpu/drm/nouveau/dispnv50/core.c     |   1 +
 drivers/gpu/drm/nouveau/dispnv50/core.h     |   3 +
 drivers/gpu/drm/nouveau/dispnv50/corec37d.c | 110 +++++++++++
 drivers/gpu/drm/nouveau/dispnv50/curs.c     |   1 +
 drivers/gpu/drm/nouveau/dispnv50/curs.h     |   1 +
 drivers/gpu/drm/nouveau/dispnv50/cursc37a.c |  50 +++++
 drivers/gpu/drm/nouveau/dispnv50/disp.c     |   6 +
 drivers/gpu/drm/nouveau/dispnv50/disp.h     |   4 +
 drivers/gpu/drm/nouveau/dispnv50/head.c     |  13 +-
 drivers/gpu/drm/nouveau/dispnv50/head.h     |   4 +
 drivers/gpu/drm/nouveau/dispnv50/head917d.c |   2 +-
 drivers/gpu/drm/nouveau/dispnv50/headc37d.c | 212 +++++++++++++++++++++
 drivers/gpu/drm/nouveau/dispnv50/sorc37d.c  |  39 ++++
 drivers/gpu/drm/nouveau/dispnv50/wimm.c     |  47 +++++
 drivers/gpu/drm/nouveau/dispnv50/wimm.h     |   8 +
 drivers/gpu/drm/nouveau/dispnv50/wimmc37b.c |  86 +++++++++
 drivers/gpu/drm/nouveau/dispnv50/wndw.c     |  38 +++-
 drivers/gpu/drm/nouveau/dispnv50/wndw.h     |   6 +
 drivers/gpu/drm/nouveau/dispnv50/wndwc37e.c | 278 ++++++++++++++++++++++++++++
 21 files changed, 918 insertions(+), 7 deletions(-)
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/corec37d.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/cursc37a.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/headc37d.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/sorc37d.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/wimm.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/wimm.h
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/wimmc37b.c
 create mode 100644 drivers/gpu/drm/nouveau/dispnv50/wndwc37e.c

diff --git a/drivers/gpu/drm/nouveau/dispnv50/Kbuild b/drivers/gpu/drm/nouveau/dispnv50/Kbuild
index ebd18cb..849b0f4 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/Kbuild
+++ b/drivers/gpu/drm/nouveau/dispnv50/Kbuild
@@ -6,6 +6,7 @@ nouveau-y += dispnv50/core507d.o
 nouveau-y += dispnv50/core827d.o
 nouveau-y += dispnv50/core907d.o
 nouveau-y += dispnv50/core917d.o
+nouveau-y += dispnv50/corec37d.o
 
 nouveau-y += dispnv50/dac507d.o
 nouveau-y += dispnv50/dac907d.o
@@ -14,14 +15,20 @@ nouveau-y += dispnv50/pior507d.o
 
 nouveau-y += dispnv50/sor507d.o
 nouveau-y += dispnv50/sor907d.o
+nouveau-y += dispnv50/sorc37d.o
 
 nouveau-y += dispnv50/head.o
 nouveau-y += dispnv50/head507d.o
 nouveau-y += dispnv50/head827d.o
 nouveau-y += dispnv50/head907d.o
 nouveau-y += dispnv50/head917d.o
+nouveau-y += dispnv50/headc37d.o
+
+nouveau-y += dispnv50/wimm.o
+nouveau-y += dispnv50/wimmc37b.o
 
 nouveau-y += dispnv50/wndw.o
+nouveau-y += dispnv50/wndwc37e.o
 
 nouveau-y += dispnv50/base.o
 nouveau-y += dispnv50/base507c.o
@@ -32,6 +39,7 @@ nouveau-y += dispnv50/base917c.o
 nouveau-y += dispnv50/curs.o
 nouveau-y += dispnv50/curs507a.o
 nouveau-y += dispnv50/curs907a.o
+nouveau-y += dispnv50/cursc37a.o
 
 nouveau-y += dispnv50/oimm.o
 nouveau-y += dispnv50/oimm507b.o
diff --git a/drivers/gpu/drm/nouveau/dispnv50/atom.h b/drivers/gpu/drm/nouveau/dispnv50/atom.h
index d8337e7..908feb1 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/atom.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/atom.h
@@ -54,6 +54,9 @@ struct nv50_head_atom {
 		u64 offset:40;
 		u8 buffer:1;
 		u8 mode:4;
+		u8 size:2;
+		u8 range:2;
+		u8 output_mode:2;
 	} olut;
 
 	struct {
@@ -77,7 +80,7 @@ struct nv50_head_atom {
 		u32 handle;
 		u64 offset:40;
 		u8  layout:2;
-		u8  format:1;
+		u8  format:8;
 	} curs;
 
 	struct {
@@ -166,6 +169,9 @@ struct nv50_wndw_atom {
 			u8  buffer:1;
 			u8  enable:2;
 			u8  mode:4;
+			u8  size:2;
+			u8  range:2;
+			u8  output_mode:2;
 		} i;
 	} xlut;
 
diff --git a/drivers/gpu/drm/nouveau/dispnv50/core.c b/drivers/gpu/drm/nouveau/dispnv50/core.c
index f87cbaa..f3c49ad 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/core.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/core.c
@@ -42,6 +42,7 @@ nv50_core_new(struct nouveau_drm *drm, struct nv50_core **pcore)
 		int version;
 		int (*new)(struct nouveau_drm *, s32, struct nv50_core **);
 	} cores[] = {
+		{ GV100_DISP_CORE_CHANNEL_DMA, 0, corec37d_new },
 		{ GP102_DISP_CORE_CHANNEL_DMA, 0, core917d_new },
 		{ GP100_DISP_CORE_CHANNEL_DMA, 0, core917d_new },
 		{ GM200_DISP_CORE_CHANNEL_DMA, 0, core917d_new },
diff --git a/drivers/gpu/drm/nouveau/dispnv50/core.h b/drivers/gpu/drm/nouveau/dispnv50/core.h
index c490d7d..8470df9 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/core.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/core.h
@@ -44,4 +44,7 @@ extern const struct nv50_outp_func dac907d;
 extern const struct nv50_outp_func sor907d;
 
 int core917d_new(struct nouveau_drm *, s32, struct nv50_core **);
+
+int corec37d_new(struct nouveau_drm *, s32, struct nv50_core **);
+extern const struct nv50_outp_func sorc37d;
 #endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/corec37d.c b/drivers/gpu/drm/nouveau/dispnv50/corec37d.c
new file mode 100644
index 0000000..b5c17c9
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/corec37d.c
@@ -0,0 +1,110 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "core.h"
+#include "head.h"
+
+#include <nouveau_bo.h>
+
+static void
+corec37d_update(struct nv50_core *core, u32 *interlock, bool ntfy)
+{
+	u32 *push;
+	if ((push = evo_wait(&core->chan, 9))) {
+		if (ntfy) {
+			evo_mthd(push, 0x020c, 1);
+			evo_data(push, 0x00001000 | NV50_DISP_CORE_NTFY);
+		}
+
+		evo_mthd(push, 0x0218, 2);
+		evo_data(push, interlock[NV50_DISP_INTERLOCK_CURS]);
+		evo_data(push, interlock[NV50_DISP_INTERLOCK_WNDW]);
+		evo_mthd(push, 0x0200, 1);
+		evo_data(push, 0x00000001);
+
+		if (ntfy) {
+			evo_mthd(push, 0x020c, 1);
+			evo_data(push, 0x00000000);
+		}
+		evo_kick(push, &core->chan);
+	}
+}
+
+int
+corec37d_ntfy_wait_done(struct nouveau_bo *bo, u32 offset,
+			struct nvif_device *device)
+{
+	u32 data;
+	s64 time = nvif_msec(device, 2000ULL,
+		data = nouveau_bo_rd32(bo, offset / 4 + 0);
+		if ((data & 0xc0000000) == 0x80000000)
+			break;
+		usleep_range(1, 2);
+	);
+	return time < 0 ? time : 0;
+}
+
+void
+corec37d_ntfy_init(struct nouveau_bo *bo, u32 offset)
+{
+	nouveau_bo_wr32(bo, offset / 4 + 0, 0x00000000);
+	nouveau_bo_wr32(bo, offset / 4 + 1, 0x00000000);
+	nouveau_bo_wr32(bo, offset / 4 + 2, 0x00000000);
+	nouveau_bo_wr32(bo, offset / 4 + 3, 0x00000000);
+}
+
+void
+corec37d_init(struct nv50_core *core)
+{
+	const u32 windows = 8; /*XXX*/
+	u32 *push, i;
+	if ((push = evo_wait(&core->chan, 2 + 6 * windows + 2))) {
+		evo_mthd(push, 0x0208, 1);
+		evo_data(push, core->chan.sync.handle);
+		for (i = 0; i < windows; i++) {
+			evo_mthd(push, 0x1000 + (i * 0x080), 3);
+			evo_data(push, i >> 1);
+			evo_data(push, 0x00000017);
+			evo_data(push, 0x00000000);
+			evo_mthd(push, 0x1010 + (i * 0x080), 1);
+			evo_data(push, 0x00127fff);
+		}
+		evo_mthd(push, 0x0200, 1);
+		evo_data(push, 0x00000001);
+		evo_kick(push, &core->chan);
+	}
+}
+
+static const struct nv50_core_func
+corec37d = {
+	.init = corec37d_init,
+	.ntfy_init = corec37d_ntfy_init,
+	.ntfy_wait_done = corec37d_ntfy_wait_done,
+	.update = corec37d_update,
+	.head = &headc37d,
+	.sor = &sorc37d,
+};
+
+int
+corec37d_new(struct nouveau_drm *drm, s32 oclass, struct nv50_core **pcore)
+{
+	return core507d_new_(&corec37d, drm, oclass, pcore);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/curs.c b/drivers/gpu/drm/nouveau/dispnv50/curs.c
index fb842ed..f592087 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/curs.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/curs.c
@@ -31,6 +31,7 @@ nv50_curs_new(struct nouveau_drm *drm, int head, struct nv50_wndw **pwndw)
 		int version;
 		int (*new)(struct nouveau_drm *, int, s32, struct nv50_wndw **);
 	} curses[] = {
+		{ GV100_DISP_CURSOR, 0, cursc37a_new },
 		{ GK104_DISP_CURSOR, 0, curs907a_new },
 		{ GF110_DISP_CURSOR, 0, curs907a_new },
 		{ GT214_DISP_CURSOR, 0, curs507a_new },
diff --git a/drivers/gpu/drm/nouveau/dispnv50/curs.h b/drivers/gpu/drm/nouveau/dispnv50/curs.h
index 8edac45..23aff5f 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/curs.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/curs.h
@@ -8,6 +8,7 @@ int curs507a_new_(const struct nv50_wimm_func *, struct nouveau_drm *,
 		  struct nv50_wndw **);
 
 int curs907a_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+int cursc37a_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
 
 int nv50_curs_new(struct nouveau_drm *, int head, struct nv50_wndw **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/cursc37a.c b/drivers/gpu/drm/nouveau/dispnv50/cursc37a.c
new file mode 100644
index 0000000..23fb29d
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/cursc37a.c
@@ -0,0 +1,50 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "curs.h"
+#include "atom.h"
+
+static void
+cursc37a_update(struct nv50_wndw *wndw, u32 *interlock)
+{
+	nvif_wr32(&wndw->wimm.base.user, 0x0200, 0x00000001);
+}
+
+static void
+cursc37a_point(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	nvif_wr32(&wndw->wimm.base.user, 0x0208, asyw->point.y << 16 |
+						 asyw->point.x);
+}
+
+static const struct nv50_wimm_func
+cursc37a = {
+	.point = cursc37a_point,
+	.update = cursc37a_update,
+};
+
+int
+cursc37a_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return curs507a_new_(&cursc37a, drm, head, oclass,
+			     0x00000001 << head, pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c
index 6c860e8..b83465a 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/disp.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c
@@ -154,6 +154,9 @@ nv50_dmac_create(struct nvif_device *device, struct nvif_object *disp,
 	if (ret)
 		return ret;
 
+	if (!syncbuf)
+		return 0;
+
 	ret = nvif_object_init(&dmac->base.user, 0xf0000000, NV_DMA_IN_MEMORY,
 			       &(struct nv_dma_v0) {
 					.target = NV_DMA_V0_TARGET_VRAM,
@@ -2170,6 +2173,9 @@ nv50_display_create(struct drm_device *dev)
 		goto out;
 
 	/* create crtc objects to represent the hw heads */
+	if (disp->disp->object.oclass >= GV100_DISP)
+		crtcs = nvif_rd32(&device->object, 0x610060) & 0xff;
+	else
 	if (disp->disp->object.oclass >= GF110_DISP)
 		crtcs = nvif_rd32(&device->object, 0x612004) & 0xf;
 	else
diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.h b/drivers/gpu/drm/nouveau/dispnv50/disp.h
index a89b83f..e48c5eb 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/disp.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.h
@@ -36,11 +36,15 @@ struct nv50_disp_interlock {
 		NV50_DISP_INTERLOCK_CURS,
 		NV50_DISP_INTERLOCK_BASE,
 		NV50_DISP_INTERLOCK_OVLY,
+		NV50_DISP_INTERLOCK_WNDW,
+		NV50_DISP_INTERLOCK_WIMM,
 		NV50_DISP_INTERLOCK__SIZE
 	} type;
 	u32 data;
 };
 
+void corec37d_ntfy_init(struct nouveau_bo *, u32);
+
 struct nv50_chan {
 	struct nvif_object user;
 	struct nvif_device *device;
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head.c b/drivers/gpu/drm/nouveau/dispnv50/head.c
index ca83006..4f57e53 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/head.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/head.c
@@ -475,7 +475,16 @@ nv50_head_create(struct drm_device *dev, int index)
 
 	head->func = disp->core->func->head;
 	head->base.index = index;
-	ret = nv50_base_new(drm, head->base.index, &wndw);
+
+	if (disp->disp->object.oclass < GV100_DISP) {
+		ret = nv50_ovly_new(drm, head->base.index, &wndw);
+		ret = nv50_base_new(drm, head->base.index, &wndw);
+	} else {
+		ret = nv50_wndw_new(drm, DRM_PLANE_TYPE_OVERLAY,
+				    head->base.index * 2 + 1, &wndw);
+		ret = nv50_wndw_new(drm, DRM_PLANE_TYPE_PRIMARY,
+				    head->base.index * 2 + 0, &wndw);
+	}
 	if (ret == 0)
 		ret = nv50_curs_new(drm, head->base.index, &curs);
 	if (ret) {
@@ -495,8 +504,6 @@ nv50_head_create(struct drm_device *dev, int index)
 			goto out;
 	}
 
-	/* allocate overlay resources */
-	ret = nv50_ovly_new(drm, head->base.index, &wndw);
 out:
 	if (ret)
 		nv50_head_destroy(crtc);
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head.h b/drivers/gpu/drm/nouveau/dispnv50/head.h
index 8f2c3ff..37b3248 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/head.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/head.h
@@ -71,4 +71,8 @@ void head907d_procamp(struct nv50_head *, struct nv50_head_atom *);
 void head907d_or(struct nv50_head *, struct nv50_head_atom *);
 
 extern const struct nv50_head_func head917d;
+int head917d_curs_layout(struct nv50_head *, struct nv50_wndw_atom *,
+			 struct nv50_head_atom *);
+
+extern const struct nv50_head_func headc37d;
 #endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head917d.c b/drivers/gpu/drm/nouveau/dispnv50/head917d.c
index 4c019a4..303df84 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/head917d.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/head917d.c
@@ -63,7 +63,7 @@ head917d_base(struct nv50_head *head, struct nv50_head_atom *asyh)
 	}
 }
 
-static int
+int
 head917d_curs_layout(struct nv50_head *head, struct nv50_wndw_atom *asyw,
 		     struct nv50_head_atom *asyh)
 {
diff --git a/drivers/gpu/drm/nouveau/dispnv50/headc37d.c b/drivers/gpu/drm/nouveau/dispnv50/headc37d.c
new file mode 100644
index 0000000..989c140
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/headc37d.c
@@ -0,0 +1,212 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "head.h"
+#include "atom.h"
+#include "core.h"
+
+static void
+headc37d_or(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		/*XXX: This is a dirty hack until OR depth handling is
+		 *     improved later for deep colour etc.
+		 */
+		switch (asyh->or.depth) {
+		case 6: asyh->or.depth = 5; break;
+		case 5: asyh->or.depth = 4; break;
+		case 2: asyh->or.depth = 1; break;
+		case 0:	asyh->or.depth = 4; break;
+		default:
+			WARN_ON(1);
+			break;
+		}
+
+		evo_mthd(push, 0x2004 + (head->base.index * 0x400), 1);
+		evo_data(push, 0x00000001 |
+			       asyh->or.depth << 4 |
+			       asyh->or.nvsync << 3 |
+			       asyh->or.nhsync << 2);
+		evo_kick(push, core);
+	}
+}
+
+static void
+headc37d_procamp(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x2000 + (head->base.index * 0x400), 1);
+		evo_data(push, 0x80000000 |
+			       asyh->procamp.sat.sin << 16 |
+			       asyh->procamp.sat.cos << 4);
+		evo_kick(push, core);
+	}
+}
+
+static void
+headc37d_dither(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x2018 + (head->base.index * 0x0400), 1);
+		evo_data(push, asyh->dither.mode << 8 |
+			       asyh->dither.bits << 4 |
+			       asyh->dither.enable);
+		evo_kick(push, core);
+	}
+}
+
+static void
+headc37d_curs_clr(struct nv50_head *head)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 4))) {
+		evo_mthd(push, 0x209c + head->base.index * 0x400, 1);
+		evo_data(push, 0x000000cf);
+		evo_mthd(push, 0x2088 + head->base.index * 0x400, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, core);
+	}
+}
+
+static void
+headc37d_curs_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 7))) {
+		evo_mthd(push, 0x209c + head->base.index * 0x400, 2);
+		evo_data(push, 0x80000000 |
+			       asyh->curs.layout << 8 |
+			       asyh->curs.format << 0);
+		evo_data(push, 0x000072ff);
+		evo_mthd(push, 0x2088 + head->base.index * 0x400, 1);
+		evo_data(push, asyh->curs.handle);
+		evo_mthd(push, 0x2090 + head->base.index * 0x400, 1);
+		evo_data(push, asyh->curs.offset >> 8);
+		evo_kick(push, core);
+	}
+}
+
+static int
+headc37d_curs_format(struct nv50_head *head, struct nv50_wndw_atom *asyw,
+		     struct nv50_head_atom *asyh)
+{
+	asyh->curs.format = asyw->image.format;
+	return 0;
+}
+
+static void
+headc37d_olut_clr(struct nv50_head *head)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x20ac + (head->base.index * 0x400), 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, core);
+	}
+}
+
+static void
+headc37d_olut_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 4))) {
+		evo_mthd(push, 0x20a4 + (head->base.index * 0x400), 3);
+		evo_data(push, asyh->olut.output_mode << 8 |
+			       asyh->olut.range << 4 |
+			       asyh->olut.size);
+		evo_data(push, asyh->olut.offset >> 8);
+		evo_data(push, asyh->olut.handle);
+		evo_kick(push, core);
+	}
+}
+
+static void
+headc37d_olut(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	asyh->olut.mode = 2;
+	asyh->olut.size = 0;
+	asyh->olut.range = 0;
+	asyh->olut.output_mode = 1;
+}
+
+static void
+headc37d_mode(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	struct nv50_head_mode *m = &asyh->mode;
+	u32 *push;
+	if ((push = evo_wait(core, 12))) {
+		evo_mthd(push, 0x2064 + (head->base.index * 0x400), 5);
+		evo_data(push, (m->v.active  << 16) | m->h.active );
+		evo_data(push, (m->v.synce   << 16) | m->h.synce  );
+		evo_data(push, (m->v.blanke  << 16) | m->h.blanke );
+		evo_data(push, (m->v.blanks  << 16) | m->h.blanks );
+		evo_data(push, (m->v.blank2e << 16) | m->v.blank2s);
+		evo_mthd(push, 0x200c + (head->base.index * 0x400), 1);
+		evo_data(push, m->clock * 1000);
+		evo_mthd(push, 0x2028 + (head->base.index * 0x400), 1);
+		evo_data(push, m->clock * 1000);
+		/*XXX: HEAD_USAGE_BOUNDS, doesn't belong here. */
+		evo_mthd(push, 0x2030 + (head->base.index * 0x400), 1);
+		evo_data(push, 0x00000124);
+		evo_kick(push, core);
+	}
+}
+
+static void
+headc37d_view(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 4))) {
+		evo_mthd(push, 0x204c + (head->base.index * 0x400), 1);
+		evo_data(push, (asyh->view.iH << 16) | asyh->view.iW);
+		evo_mthd(push, 0x2058 + (head->base.index * 0x400), 1);
+		evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
+		evo_kick(push, core);
+	}
+}
+
+const struct nv50_head_func
+headc37d = {
+	.view = headc37d_view,
+	.mode = headc37d_mode,
+	.olut = headc37d_olut,
+	.olut_set = headc37d_olut_set,
+	.olut_clr = headc37d_olut_clr,
+	.curs_layout = head917d_curs_layout,
+	.curs_format = headc37d_curs_format,
+	.curs_set = headc37d_curs_set,
+	.curs_clr = headc37d_curs_clr,
+	.dither = headc37d_dither,
+	.procamp = headc37d_procamp,
+	.or = headc37d_or,
+};
diff --git a/drivers/gpu/drm/nouveau/dispnv50/sorc37d.c b/drivers/gpu/drm/nouveau/dispnv50/sorc37d.c
new file mode 100644
index 0000000..dff0592
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/sorc37d.c
@@ -0,0 +1,39 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "core.h"
+
+static void
+sorc37d_ctrl(struct nv50_core *core, int or, u32 ctrl,
+	     struct nv50_head_atom *asyh)
+{
+	u32 *push;
+	if ((push = evo_wait(&core->chan, 2))) {
+		evo_mthd(push, 0x0300 + (or * 0x20), 1);
+		evo_data(push, ctrl);
+		evo_kick(push, &core->chan);
+	}
+}
+
+const struct nv50_outp_func
+sorc37d = {
+	.ctrl = sorc37d_ctrl,
+};
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wimm.c b/drivers/gpu/drm/nouveau/dispnv50/wimm.c
new file mode 100644
index 0000000..fc36e06
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/wimm.c
@@ -0,0 +1,47 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "wimm.h"
+
+#include <nvif/class.h>
+
+int
+nv50_wimm_init(struct nouveau_drm *drm, struct nv50_wndw *wndw)
+{
+	struct {
+		s32 oclass;
+		int version;
+		int (*init)(struct nouveau_drm *, s32, struct nv50_wndw *);
+	} wimms[] = {
+		{ GV100_DISP_WINDOW_IMM_CHANNEL_DMA, 0, wimmc37b_init },
+		{}
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	int cid;
+
+	cid = nvif_mclass(&disp->disp->object, wimms);
+	if (cid < 0) {
+		NV_ERROR(drm, "No supported window immediate class\n");
+		return cid;
+	}
+
+	return wimms[cid].init(drm, wimms[cid].oclass, wndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wimm.h b/drivers/gpu/drm/nouveau/dispnv50/wimm.h
new file mode 100644
index 0000000..3630523
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/wimm.h
@@ -0,0 +1,8 @@
+#ifndef __NV50_KMS_WIMM_H__
+#define __NV50_KMS_WIMM_H__
+#include "wndw.h"
+
+int nv50_wimm_init(struct nouveau_drm *drm, struct nv50_wndw *);
+
+int wimmc37b_init(struct nouveau_drm *, s32, struct nv50_wndw *);
+#endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wimmc37b.c b/drivers/gpu/drm/nouveau/dispnv50/wimmc37b.c
new file mode 100644
index 0000000..9103b84
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/wimmc37b.c
@@ -0,0 +1,86 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "wimm.h"
+#include "atom.h"
+#include "wndw.h"
+
+#include <nvif/clc37b.h>
+
+static void
+wimmc37b_update(struct nv50_wndw *wndw, u32 *interlock)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wimm, 2))) {
+		evo_mthd(push, 0x0200, 1);
+		if (interlock[NV50_DISP_INTERLOCK_WNDW] & wndw->interlock.data)
+			evo_data(push, 0x00000003);
+		else
+			evo_data(push, 0x00000001);
+		evo_kick(push, &wndw->wimm);
+	}
+}
+
+static void
+wimmc37b_point(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wimm, 2))) {
+		evo_mthd(push, 0x0208, 1);
+		evo_data(push, asyw->point.y << 16 | asyw->point.x);
+		evo_kick(push, &wndw->wimm);
+	}
+}
+
+static const struct nv50_wimm_func
+wimmc37b = {
+	.point = wimmc37b_point,
+	.update = wimmc37b_update,
+};
+
+static int
+wimmc37b_init_(const struct nv50_wimm_func *func, struct nouveau_drm *drm,
+	       s32 oclass, struct nv50_wndw *wndw)
+{
+	struct nvc37b_window_imm_channel_dma_v0 args = {
+		.pushbuf = 0xb0007b00 | wndw->id,
+		.index = wndw->id,
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	int ret;
+
+	ret = nv50_dmac_create(&drm->client.device, &disp->disp->object,
+			       &oclass, 0, &args, sizeof(args), 0,
+			       &wndw->wimm);
+	if (ret) {
+		NV_ERROR(drm, "wimm%04x allocation failed: %d\n", oclass, ret);
+		return ret;
+	}
+
+	wndw->immd = func;
+	return 0;
+}
+
+int
+wimmc37b_init(struct nouveau_drm *drm, s32 oclass, struct nv50_wndw *wndw)
+{
+	return wimmc37b_init_(&wimmc37b, drm, oclass, wndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.c b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
index c7c08fa..224963b5 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/wndw.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
@@ -20,6 +20,7 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  */
 #include "wndw.h"
+#include "wimm.h"
 
 #include <nvif/class.h>
 #include <nvif/cl0002.h>
@@ -148,11 +149,15 @@ nv50_wndw_flush_set(struct nv50_wndw *wndw, u32 *interlock,
 
 	if (asyw->set.scale) wndw->func->scale_set(wndw, asyw);
 	if (asyw->set.point) {
+		if (asyw->set.point = false, asyw->set.mask)
+			interlock[wndw->interlock.type] |= wndw->interlock.data;
+		interlock[NV50_DISP_INTERLOCK_WIMM] |= wndw->interlock.data;
+
 		wndw->immd->point(wndw, asyw);
 		wndw->immd->update(wndw, interlock);
+	} else {
+		interlock[wndw->interlock.type] |= wndw->interlock.data;
 	}
-
-	interlock[wndw->interlock.type] |= wndw->interlock.data;
 }
 
 void
@@ -605,3 +610,32 @@ nv50_wndw_new_(const struct nv50_wndw_func *func, struct drm_device *dev,
 	wndw->notify.func = nv50_wndw_notify;
 	return 0;
 }
+
+int
+nv50_wndw_new(struct nouveau_drm *drm, enum drm_plane_type type, int index,
+	      struct nv50_wndw **pwndw)
+{
+	struct {
+		s32 oclass;
+		int version;
+		int (*new)(struct nouveau_drm *, enum drm_plane_type,
+			   int, s32, struct nv50_wndw **);
+	} wndws[] = {
+		{ GV100_DISP_WINDOW_CHANNEL_DMA, 0, wndwc37e_new },
+		{}
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	int cid, ret;
+
+	cid = nvif_mclass(&disp->disp->object, wndws);
+	if (cid < 0) {
+		NV_ERROR(drm, "No supported window class\n");
+		return cid;
+	}
+
+	ret = wndws[cid].new(drm, type, index, wndws[cid].oclass, pwndw);
+	if (ret)
+		return ret;
+
+	return nv50_wimm_init(drm, *pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.h b/drivers/gpu/drm/nouveau/dispnv50/wndw.h
index 745304d..b0b6428 100644
--- a/drivers/gpu/drm/nouveau/dispnv50/wndw.h
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.h
@@ -87,4 +87,10 @@ struct nv50_wimm_func {
 };
 
 extern const struct nv50_wimm_func curs507a;
+
+int wndwc37e_new(struct nouveau_drm *, enum drm_plane_type, int, s32,
+		 struct nv50_wndw **);
+
+int nv50_wndw_new(struct nouveau_drm *, enum drm_plane_type, int index,
+		  struct nv50_wndw **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndwc37e.c b/drivers/gpu/drm/nouveau/dispnv50/wndwc37e.c
new file mode 100644
index 0000000..44afb0f
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndwc37e.c
@@ -0,0 +1,278 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "wndw.h"
+#include "atom.h"
+
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_plane_helper.h>
+#include <nouveau_bo.h>
+
+#include <nvif/clc37e.h>
+
+static void
+wndwc37e_ilut_clr(struct nv50_wndw *wndw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 2))) {
+		evo_mthd(push, 0x02b8, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+wndwc37e_ilut_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 4))) {
+		evo_mthd(push, 0x02b0, 3);
+		evo_data(push, asyw->xlut.i.output_mode << 8 |
+			       asyw->xlut.i.range << 4 |
+			       asyw->xlut.i.size);
+		evo_data(push, asyw->xlut.i.offset >> 8);
+		evo_data(push, asyw->xlut.handle);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+wndwc37e_ilut(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	asyw->xlut.i.mode = 2;
+	asyw->xlut.i.size = 0;
+	asyw->xlut.i.range = 0;
+	asyw->xlut.i.output_mode = 1;
+}
+
+static void
+wndwc37e_image_clr(struct nv50_wndw *wndw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 4))) {
+		evo_mthd(push, 0x0308, 1);
+		evo_data(push, 0x00000000);
+		evo_mthd(push, 0x0240, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+wndwc37e_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+
+	if (!(push = evo_wait(&wndw->wndw, 25)))
+		return;
+
+	evo_mthd(push, 0x0308, 1);
+	evo_data(push, asyw->image.mode << 4 | asyw->image.interval);
+	evo_mthd(push, 0x0224, 4);
+	evo_data(push, asyw->image.h << 16 | asyw->image.w);
+	evo_data(push, asyw->image.layout << 4 | asyw->image.blockh);
+	evo_data(push, asyw->image.colorspace << 8 | asyw->image.format);
+	evo_data(push, asyw->image.blocks[0] | (asyw->image.pitch[0] >> 6));
+	evo_mthd(push, 0x0240, 1);
+	evo_data(push, asyw->image.handle[0]);
+	evo_mthd(push, 0x0260, 1);
+	evo_data(push, asyw->image.offset[0] >> 8);
+	evo_mthd(push, 0x0290, 1);
+	evo_data(push, (asyw->state.src_y >> 16) << 16 |
+		       (asyw->state.src_x >> 16));
+	evo_mthd(push, 0x0298, 1);
+	evo_data(push, (asyw->state.src_h >> 16) << 16 |
+		       (asyw->state.src_w >> 16));
+	evo_mthd(push, 0x02a4, 1);
+	evo_data(push, asyw->state.crtc_h << 16 |
+		       asyw->state.crtc_w);
+
+	/*XXX: Composition-related stuff.  Need to implement properly. */
+	evo_mthd(push, 0x02ec, 1);
+	evo_data(push, (2 - (wndw->id & 1)) << 4);
+	evo_mthd(push, 0x02f4, 5);
+	evo_data(push, 0x00000011);
+	evo_data(push, 0xffff0000);
+	evo_data(push, 0xffff0000);
+	evo_data(push, 0xffff0000);
+	evo_data(push, 0xffff0000);
+	evo_kick(push, &wndw->wndw);
+}
+
+static void
+wndwc37e_ntfy_clr(struct nv50_wndw *wndw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 2))) {
+		evo_mthd(push, 0x021c, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+wndwc37e_ntfy_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 3))) {
+		evo_mthd(push, 0x021c, 2);
+		evo_data(push, asyw->ntfy.handle);
+		evo_data(push, asyw->ntfy.offset | asyw->ntfy.awaken);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+wndwc37e_sema_clr(struct nv50_wndw *wndw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 2))) {
+		evo_mthd(push, 0x0218, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+wndwc37e_sema_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 5))) {
+		evo_mthd(push, 0x020c, 4);
+		evo_data(push, asyw->sema.offset);
+		evo_data(push, asyw->sema.acquire);
+		evo_data(push, asyw->sema.release);
+		evo_data(push, asyw->sema.handle);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+wndwc37e_update(struct nv50_wndw *wndw, u32 *interlock)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 5))) {
+		evo_mthd(push, 0x0370, 2);
+		evo_data(push, interlock[NV50_DISP_INTERLOCK_CURS] << 1 |
+			       interlock[NV50_DISP_INTERLOCK_CORE]);
+		evo_data(push, interlock[NV50_DISP_INTERLOCK_WNDW]);
+		evo_mthd(push, 0x0200, 1);
+		if (interlock[NV50_DISP_INTERLOCK_WIMM] & wndw->interlock.data)
+			evo_data(push, 0x00001001);
+		else
+			evo_data(push, 0x00000001);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+wndwc37e_release(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
+		 struct nv50_head_atom *asyh)
+{
+}
+
+static int
+wndwc37e_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
+		 struct nv50_head_atom *asyh)
+{
+	return drm_atomic_helper_check_plane_state(&asyw->state, &asyh->state,
+						   DRM_PLANE_HELPER_NO_SCALING,
+						   DRM_PLANE_HELPER_NO_SCALING,
+						   true, true);
+}
+
+static const u32
+wndwc37e_format[] = {
+	DRM_FORMAT_C8,
+	DRM_FORMAT_YUYV,
+	DRM_FORMAT_UYVY,
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_RGB565,
+	DRM_FORMAT_XRGB1555,
+	DRM_FORMAT_ARGB1555,
+	DRM_FORMAT_XBGR2101010,
+	DRM_FORMAT_ABGR2101010,
+	DRM_FORMAT_XBGR8888,
+	DRM_FORMAT_ABGR8888,
+	DRM_FORMAT_XRGB2101010,
+	DRM_FORMAT_ARGB2101010,
+	0
+};
+
+static const struct nv50_wndw_func
+wndwc37e = {
+	.acquire = wndwc37e_acquire,
+	.release = wndwc37e_release,
+	.sema_set = wndwc37e_sema_set,
+	.sema_clr = wndwc37e_sema_clr,
+	.ntfy_set = wndwc37e_ntfy_set,
+	.ntfy_clr = wndwc37e_ntfy_clr,
+	.ntfy_reset = corec37d_ntfy_init,
+	.ntfy_wait_begun = base507c_ntfy_wait_begun,
+	.ilut = wndwc37e_ilut,
+	.xlut_set = wndwc37e_ilut_set,
+	.xlut_clr = wndwc37e_ilut_clr,
+	.image_set = wndwc37e_image_set,
+	.image_clr = wndwc37e_image_clr,
+	.update = wndwc37e_update,
+};
+
+static int
+wndwc37e_new_(const struct nv50_wndw_func *func, struct nouveau_drm *drm,
+	      enum drm_plane_type type, int index, s32 oclass, u32 heads,
+	      struct nv50_wndw **pwndw)
+{
+	struct nvc37e_window_channel_dma_v0 args = {
+		.pushbuf = 0xb0007e00 | index,
+		.index = index,
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	struct nv50_wndw *wndw;
+	int ret;
+
+	ret = nv50_wndw_new_(func, drm->dev, type, "wndw", index,
+			     wndwc37e_format, heads, NV50_DISP_INTERLOCK_WNDW,
+			     BIT(index), &wndw);
+	if (*pwndw = wndw, ret)
+		return ret;
+
+	ret = nv50_dmac_create(&drm->client.device, &disp->disp->object,
+			       &oclass, 0, &args, sizeof(args),
+			       disp->sync->bo.offset, &wndw->wndw);
+	if (ret) {
+		NV_ERROR(drm, "qndw%04x allocation failed: %d\n", oclass, ret);
+		return ret;
+	}
+
+	wndw->ntfy = NV50_DISP_WNDW_NTFY(wndw->id);
+	wndw->sema = NV50_DISP_WNDW_SEM0(wndw->id);
+	wndw->data = 0x00000000;
+	return 0;
+}
+
+int
+wndwc37e_new(struct nouveau_drm *drm, enum drm_plane_type type, int index,
+	     s32 oclass, struct nv50_wndw **pwndw)
+{
+	return wndwc37e_new_(&wndwc37e, drm, type, index, oclass,
+			     BIT(index >> 1), pwndw);
+}
-- 
cgit v1.1


From 37e1c45a58b5c1f699d583483f612462418dd2ee Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:48 +1000
Subject: drm/nouveau/fifo/gv100: initial support

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/include/nvif/class.h       |   3 +
 drivers/gpu/drm/nouveau/include/nvif/device.h      |   3 +
 drivers/gpu/drm/nouveau/include/nvif/user.h        |  19 ++
 drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h |   1 +
 drivers/gpu/drm/nouveau/nouveau_chan.c             |   3 +-
 drivers/gpu/drm/nouveau/nouveau_dma.c              |   5 +
 drivers/gpu/drm/nouveau/nouveau_drm.c              |   8 +
 drivers/gpu/drm/nouveau/nvif/Kbuild                |   4 +
 drivers/gpu/drm/nouveau/nvif/device.c              |   2 +
 drivers/gpu/drm/nouveau/nvif/user.c                |  64 +++++
 drivers/gpu/drm/nouveau/nvif/userc361.c            |  33 +++
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c  |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/Kbuild    |   4 +
 .../gpu/drm/nouveau/nvkm/engine/fifo/changk104.h   |  14 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c   |  10 +
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h   |   6 +
 .../gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c |  28 +-
 .../gpu/drm/nouveau/nvkm/engine/fifo/gpfifogv100.c | 225 +++++++++++++++
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gv100.c   | 306 +++++++++++++++++++++
 drivers/gpu/drm/nouveau/nvkm/engine/fifo/user.h    |   6 +
 .../gpu/drm/nouveau/nvkm/engine/fifo/usergv100.c   |  45 +++
 21 files changed, 779 insertions(+), 11 deletions(-)
 create mode 100644 drivers/gpu/drm/nouveau/include/nvif/user.h
 create mode 100644 drivers/gpu/drm/nouveau/nvif/user.c
 create mode 100644 drivers/gpu/drm/nouveau/nvif/userc361.c
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogv100.c
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/fifo/gv100.c
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/fifo/user.h
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/fifo/usergv100.c

diff --git a/drivers/gpu/drm/nouveau/include/nvif/class.h b/drivers/gpu/drm/nouveau/include/nvif/class.h
index 8c9aa55..a2fdbc6 100644
--- a/drivers/gpu/drm/nouveau/include/nvif/class.h
+++ b/drivers/gpu/drm/nouveau/include/nvif/class.h
@@ -52,6 +52,8 @@
 
 #define NV04_DISP                                     /* cl0046.h */ 0x00000046
 
+#define VOLTA_USERMODE_A                                             0x0000c361
+
 #define NV03_CHANNEL_DMA                              /* cl506b.h */ 0x0000006b
 #define NV10_CHANNEL_DMA                              /* cl506b.h */ 0x0000006e
 #define NV17_CHANNEL_DMA                              /* cl506b.h */ 0x0000176e
@@ -66,6 +68,7 @@
 #define KEPLER_CHANNEL_GPFIFO_B                       /* cla06f.h */ 0x0000a16f
 #define MAXWELL_CHANNEL_GPFIFO_A                      /* cla06f.h */ 0x0000b06f
 #define PASCAL_CHANNEL_GPFIFO_A                       /* cla06f.h */ 0x0000c06f
+#define VOLTA_CHANNEL_GPFIFO_A                        /* cla06f.h */ 0x0000c36f
 
 #define NV50_DISP                                     /* cl5070.h */ 0x00005070
 #define G82_DISP                                      /* cl5070.h */ 0x00008270
diff --git a/drivers/gpu/drm/nouveau/include/nvif/device.h b/drivers/gpu/drm/nouveau/include/nvif/device.h
index 76fe21e..ef839bd 100644
--- a/drivers/gpu/drm/nouveau/include/nvif/device.h
+++ b/drivers/gpu/drm/nouveau/include/nvif/device.h
@@ -4,6 +4,7 @@
 
 #include <nvif/object.h>
 #include <nvif/cl0080.h>
+#include <nvif/user.h>
 
 struct nvif_device {
 	struct nvif_object object;
@@ -13,6 +14,8 @@ struct nvif_device {
 		u64 engines;
 	} *runlist;
 	int runlists;
+
+	struct nvif_user user;
 };
 
 int  nvif_device_init(struct nvif_object *, u32 handle, s32 oclass, void *, u32,
diff --git a/drivers/gpu/drm/nouveau/include/nvif/user.h b/drivers/gpu/drm/nouveau/include/nvif/user.h
new file mode 100644
index 0000000..03c1182
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/include/nvif/user.h
@@ -0,0 +1,19 @@
+#ifndef __NVIF_USER_H__
+#define __NVIF_USER_H__
+#include <nvif/object.h>
+struct nvif_device;
+
+struct nvif_user {
+	const struct nvif_user_func *func;
+	struct nvif_object object;
+};
+
+struct nvif_user_func {
+	void (*doorbell)(struct nvif_user *, u32 token);
+};
+
+int nvif_user_init(struct nvif_device *);
+void nvif_user_fini(struct nvif_device *);
+
+extern const struct nvif_user_func nvif_userc361;
+#endif
diff --git a/drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h b/drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h
index 0d96ede..7e39fbe 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h
@@ -73,4 +73,5 @@ int gm200_fifo_new(struct nvkm_device *, int, struct nvkm_fifo **);
 int gm20b_fifo_new(struct nvkm_device *, int, struct nvkm_fifo **);
 int gp100_fifo_new(struct nvkm_device *, int, struct nvkm_fifo **);
 int gp10b_fifo_new(struct nvkm_device *, int, struct nvkm_fifo **);
+int gv100_fifo_new(struct nvkm_device *, int, struct nvkm_fifo **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nouveau_chan.c b/drivers/gpu/drm/nouveau/nouveau_chan.c
index 97900e9..92d3115 100644
--- a/drivers/gpu/drm/nouveau/nouveau_chan.c
+++ b/drivers/gpu/drm/nouveau/nouveau_chan.c
@@ -220,7 +220,8 @@ nouveau_channel_ind(struct nouveau_drm *drm, struct nvif_device *device,
 		    u64 runlist, struct nouveau_channel **pchan)
 {
 	struct nouveau_cli *cli = (void *)device->object.client;
-	static const u16 oclasses[] = { PASCAL_CHANNEL_GPFIFO_A,
+	static const u16 oclasses[] = { VOLTA_CHANNEL_GPFIFO_A,
+					PASCAL_CHANNEL_GPFIFO_A,
 					MAXWELL_CHANNEL_GPFIFO_A,
 					KEPLER_CHANNEL_GPFIFO_B,
 					KEPLER_CHANNEL_GPFIFO_A,
diff --git a/drivers/gpu/drm/nouveau/nouveau_dma.c b/drivers/gpu/drm/nouveau/nouveau_dma.c
index e0664d2..945afd3 100644
--- a/drivers/gpu/drm/nouveau/nouveau_dma.c
+++ b/drivers/gpu/drm/nouveau/nouveau_dma.c
@@ -28,6 +28,8 @@
 #include "nouveau_dma.h"
 #include "nouveau_vmm.h"
 
+#include <nvif/user.h>
+
 void
 OUT_RINGp(struct nouveau_channel *chan, const void *data, unsigned nr_dwords)
 {
@@ -82,6 +84,7 @@ READ_GET(struct nouveau_channel *chan, uint64_t *prev_get, int *timeout)
 void
 nv50_dma_push(struct nouveau_channel *chan, u64 offset, int length)
 {
+	struct nvif_user *user = &chan->drm->client.device.user;
 	struct nouveau_bo *pb = chan->push.buffer;
 	int ip = (chan->dma.ib_put * 2) + chan->dma.ib_base;
 
@@ -97,6 +100,8 @@ nv50_dma_push(struct nouveau_channel *chan, u64 offset, int length)
 	nouveau_bo_rd32(pb, 0);
 
 	nvif_wr32(&chan->user, 0x8c, chan->dma.ib_put);
+	if (user->func && user->func->doorbell)
+		user->func->doorbell(user, chan->chid);
 	chan->dma.ib_free--;
 }
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c
index 8e506c5..775443c 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_drm.c
@@ -39,6 +39,7 @@
 
 #include <nvif/driver.h>
 #include <nvif/fifo.h>
+#include <nvif/user.h>
 
 #include <nvif/class.h>
 #include <nvif/cl0002.h>
@@ -310,6 +311,12 @@ nouveau_accel_init(struct nouveau_drm *drm)
 	if (ret)
 		return;
 
+	if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_VOLTA) {
+		ret = nvif_user_init(device);
+		if (ret)
+			return;
+	}
+
 	/* initialise synchronisation routines */
 	/*XXX: this is crap, but the fence/channel stuff is a little
 	 *     backwards in some places.  this will be fixed.
@@ -341,6 +348,7 @@ nouveau_accel_init(struct nouveau_drm *drm)
 		case KEPLER_CHANNEL_GPFIFO_B:
 		case MAXWELL_CHANNEL_GPFIFO_A:
 		case PASCAL_CHANNEL_GPFIFO_A:
+		case VOLTA_CHANNEL_GPFIFO_A:
 			ret = nvc0_fence_create(drm);
 			break;
 		default:
diff --git a/drivers/gpu/drm/nouveau/nvif/Kbuild b/drivers/gpu/drm/nouveau/nvif/Kbuild
index 3db1250..42e8c85 100644
--- a/drivers/gpu/drm/nouveau/nvif/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvif/Kbuild
@@ -8,3 +8,7 @@ nvif-y += nvif/mem.o
 nvif-y += nvif/mmu.o
 nvif-y += nvif/notify.o
 nvif-y += nvif/vmm.o
+
+# Usermode classes
+nvif-y += nvif/user.o
+nvif-y += nvif/userc361.o
diff --git a/drivers/gpu/drm/nouveau/nvif/device.c b/drivers/gpu/drm/nouveau/nvif/device.c
index ca5eb3d..1ec101b 100644
--- a/drivers/gpu/drm/nouveau/nvif/device.c
+++ b/drivers/gpu/drm/nouveau/nvif/device.c
@@ -37,6 +37,7 @@ nvif_device_time(struct nvif_device *device)
 void
 nvif_device_fini(struct nvif_device *device)
 {
+	nvif_user_fini(device);
 	kfree(device->runlist);
 	device->runlist = NULL;
 	nvif_object_fini(&device->object);
@@ -49,6 +50,7 @@ nvif_device_init(struct nvif_object *parent, u32 handle, s32 oclass,
 	int ret = nvif_object_init(parent, handle, oclass, data, size,
 				   &device->object);
 	device->runlist = NULL;
+	device->user.func = NULL;
 	if (ret == 0) {
 		device->info.version = 0;
 		ret = nvif_object_mthd(&device->object, NV_DEVICE_V0_INFO,
diff --git a/drivers/gpu/drm/nouveau/nvif/user.c b/drivers/gpu/drm/nouveau/nvif/user.c
new file mode 100644
index 0000000..10da3cd
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvif/user.c
@@ -0,0 +1,64 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include <nvif/user.h>
+#include <nvif/device.h>
+
+#include <nvif/class.h>
+
+void
+nvif_user_fini(struct nvif_device *device)
+{
+	if (device->user.func) {
+		nvif_object_fini(&device->user.object);
+		device->user.func = NULL;
+	}
+}
+
+int
+nvif_user_init(struct nvif_device *device)
+{
+	struct {
+		s32 oclass;
+		int version;
+		const struct nvif_user_func *func;
+	} users[] = {
+		{ VOLTA_USERMODE_A, -1, &nvif_userc361 },
+		{}
+	};
+	int cid, ret;
+
+	if (device->user.func)
+		return 0;
+
+	cid = nvif_mclass(&device->object, users);
+	if (cid < 0)
+		return cid;
+
+	ret = nvif_object_init(&device->object, 0, users[cid].oclass, NULL, 0,
+			       &device->user.object);
+	if (ret)
+		return ret;
+
+	nvif_object_map(&device->user.object, NULL, 0);
+	device->user.func = users[cid].func;
+	return 0;
+}
diff --git a/drivers/gpu/drm/nouveau/nvif/userc361.c b/drivers/gpu/drm/nouveau/nvif/userc361.c
new file mode 100644
index 0000000..19f9958
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvif/userc361.c
@@ -0,0 +1,33 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include <nvif/user.h>
+
+static void
+nvif_userc361_doorbell(struct nvif_user *user, u32 token)
+{
+	nvif_wr32(&user->object, 0x90, token);
+}
+
+const struct nvif_user_func
+nvif_userc361 = {
+	.doorbell = nvif_userc361_doorbell,
+};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 43f6b7a..95a56d4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2418,6 +2418,7 @@ nv140_chipset = {
 	.top = gk104_top_new,
 	.disp = gv100_disp_new,
 	.dma = gv100_dma_new,
+	.fifo = gv100_fifo_new,
 };
 
 static int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/Kbuild b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/Kbuild
index b888ea6..f004085 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/Kbuild
@@ -15,6 +15,7 @@ nvkm-y += nvkm/engine/fifo/gm200.o
 nvkm-y += nvkm/engine/fifo/gm20b.o
 nvkm-y += nvkm/engine/fifo/gp100.o
 nvkm-y += nvkm/engine/fifo/gp10b.o
+nvkm-y += nvkm/engine/fifo/gv100.o
 
 nvkm-y += nvkm/engine/fifo/chan.o
 nvkm-y += nvkm/engine/fifo/channv50.o
@@ -31,3 +32,6 @@ nvkm-y += nvkm/engine/fifo/gpfifonv50.o
 nvkm-y += nvkm/engine/fifo/gpfifog84.o
 nvkm-y += nvkm/engine/fifo/gpfifogf100.o
 nvkm-y += nvkm/engine/fifo/gpfifogk104.o
+nvkm-y += nvkm/engine/fifo/gpfifogv100.o
+
+nvkm-y += nvkm/engine/fifo/usergv100.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/changk104.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/changk104.h
index 391e864..8e28ba6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/changk104.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/changk104.h
@@ -20,6 +20,20 @@ struct gk104_fifo_chan {
 	} engn[NVKM_SUBDEV_NR];
 };
 
+extern const struct nvkm_fifo_chan_func gk104_fifo_gpfifo_func;
+
 int gk104_fifo_gpfifo_new(struct gk104_fifo *, const struct nvkm_oclass *,
 			  void *data, u32 size, struct nvkm_object **);
+void *gk104_fifo_gpfifo_dtor(struct nvkm_fifo_chan *);
+void gk104_fifo_gpfifo_init(struct nvkm_fifo_chan *);
+void gk104_fifo_gpfifo_fini(struct nvkm_fifo_chan *);
+int gk104_fifo_gpfifo_engine_ctor(struct nvkm_fifo_chan *, struct nvkm_engine *,
+				  struct nvkm_object *);
+void gk104_fifo_gpfifo_engine_dtor(struct nvkm_fifo_chan *,
+				   struct nvkm_engine *);
+int gk104_fifo_gpfifo_kick(struct gk104_fifo_chan *);
+int gk104_fifo_gpfifo_kick_locked(struct gk104_fifo_chan *);
+
+int gv100_fifo_gpfifo_new(struct gk104_fifo *, const struct nvkm_oclass *,
+			  void *data, u32 size, struct nvkm_object **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
index 767e0ab..a990464 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
@@ -103,6 +103,10 @@ gk104_fifo_class_new(struct nvkm_fifo *base, const struct nvkm_oclass *oclass,
 	if (oclass->engn == &fifo->func->chan) {
 		const struct gk104_fifo_chan_user *user = oclass->engn;
 		return user->ctor(fifo, oclass, argv, argc, pobject);
+	} else
+	if (oclass->engn == &fifo->func->user) {
+		const struct gk104_fifo_user_user *user = oclass->engn;
+		return user->ctor(oclass, argv, argc, pobject);
 	}
 	WARN_ON(1);
 	return -EINVAL;
@@ -115,6 +119,12 @@ gk104_fifo_class_get(struct nvkm_fifo *base, int index,
 	struct gk104_fifo *fifo = gk104_fifo(base);
 	int c = 0;
 
+	if (fifo->func->user.ctor && c++ == index) {
+		oclass->base =  fifo->func->user.user;
+		oclass->engn = &fifo->func->user;
+		return 0;
+	}
+
 	if (fifo->func->chan.ctor && c++ == index) {
 		oclass->base =  fifo->func->chan.user;
 		oclass->engn = &fifo->func->chan;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
index 1d182d8..d295b81 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
@@ -63,6 +63,12 @@ struct gk104_fifo_func {
 			     struct nvkm_memory *, u32 offset);
 	} *runlist;
 
+	struct gk104_fifo_user_user {
+		struct nvkm_sclass user;
+		int (*ctor)(const struct nvkm_oclass *, void *, u32,
+			    struct nvkm_object **);
+	} user;
+
 	struct gk104_fifo_chan_user {
 		struct nvkm_sclass user;
 		int (*ctor)(struct gk104_fifo *, const struct nvkm_oclass *,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c
index 60e7d72..118b37a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c
@@ -34,8 +34,8 @@
 #include <nvif/cla06f.h>
 #include <nvif/unpack.h>
 
-static int
-gk104_fifo_gpfifo_kick(struct gk104_fifo_chan *chan)
+int
+gk104_fifo_gpfifo_kick_locked(struct gk104_fifo_chan *chan)
 {
 	struct gk104_fifo *fifo = chan->fifo;
 	struct nvkm_subdev *subdev = &fifo->base.engine.subdev;
@@ -44,7 +44,6 @@ gk104_fifo_gpfifo_kick(struct gk104_fifo_chan *chan)
 	struct nvkm_fifo_cgrp *cgrp = chan->cgrp;
 	int ret = 0;
 
-	mutex_lock(&subdev->mutex);
 	if (cgrp)
 		nvkm_wr32(device, 0x002634, cgrp->id | 0x01000000);
 	else
@@ -59,7 +58,16 @@ gk104_fifo_gpfifo_kick(struct gk104_fifo_chan *chan)
 		nvkm_fifo_recover_chan(&fifo->base, chan->base.chid);
 		ret = -ETIMEDOUT;
 	}
-	mutex_unlock(&subdev->mutex);
+	return ret;
+}
+
+int
+gk104_fifo_gpfifo_kick(struct gk104_fifo_chan *chan)
+{
+	int ret;
+	mutex_lock(&chan->base.fifo->engine.subdev.mutex);
+	ret = gk104_fifo_gpfifo_kick_locked(chan);
+	mutex_unlock(&chan->base.fifo->engine.subdev.mutex);
 	return ret;
 }
 
@@ -138,7 +146,7 @@ gk104_fifo_gpfifo_engine_init(struct nvkm_fifo_chan *base,
 	return 0;
 }
 
-static void
+void
 gk104_fifo_gpfifo_engine_dtor(struct nvkm_fifo_chan *base,
 			      struct nvkm_engine *engine)
 {
@@ -147,7 +155,7 @@ gk104_fifo_gpfifo_engine_dtor(struct nvkm_fifo_chan *base,
 	nvkm_gpuobj_del(&chan->engn[engine->subdev.index].inst);
 }
 
-static int
+int
 gk104_fifo_gpfifo_engine_ctor(struct nvkm_fifo_chan *base,
 			      struct nvkm_engine *engine,
 			      struct nvkm_object *object)
@@ -172,7 +180,7 @@ gk104_fifo_gpfifo_engine_ctor(struct nvkm_fifo_chan *base,
 			       chan->engn[engn].vma, NULL, 0);
 }
 
-static void
+void
 gk104_fifo_gpfifo_fini(struct nvkm_fifo_chan *base)
 {
 	struct gk104_fifo_chan *chan = gk104_fifo_chan(base);
@@ -190,7 +198,7 @@ gk104_fifo_gpfifo_fini(struct nvkm_fifo_chan *base)
 	nvkm_wr32(device, 0x800000 + coff, 0x00000000);
 }
 
-static void
+void
 gk104_fifo_gpfifo_init(struct nvkm_fifo_chan *base)
 {
 	struct gk104_fifo_chan *chan = gk104_fifo_chan(base);
@@ -210,7 +218,7 @@ gk104_fifo_gpfifo_init(struct nvkm_fifo_chan *base)
 	}
 }
 
-static void *
+void *
 gk104_fifo_gpfifo_dtor(struct nvkm_fifo_chan *base)
 {
 	struct gk104_fifo_chan *chan = gk104_fifo_chan(base);
@@ -218,7 +226,7 @@ gk104_fifo_gpfifo_dtor(struct nvkm_fifo_chan *base)
 	return chan;
 }
 
-static const struct nvkm_fifo_chan_func
+const struct nvkm_fifo_chan_func
 gk104_fifo_gpfifo_func = {
 	.dtor = gk104_fifo_gpfifo_dtor,
 	.init = gk104_fifo_gpfifo_init,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogv100.c
new file mode 100644
index 0000000..9598853
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogv100.c
@@ -0,0 +1,225 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "changk104.h"
+#include "cgrp.h"
+
+#include <core/client.h>
+#include <core/gpuobj.h>
+
+#include <nvif/cla06f.h>
+#include <nvif/unpack.h>
+
+static int
+gv100_fifo_gpfifo_engine_valid(struct gk104_fifo_chan *chan, bool ce, bool valid)
+{
+	struct nvkm_subdev *subdev = &chan->base.fifo->engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	const u32 mask = ce ? 0x00020000 : 0x00010000;
+	const u32 data = valid ? mask : 0x00000000;
+	int ret;
+
+	/* Block runlist to prevent the channel from being rescheduled. */
+	mutex_lock(&subdev->mutex);
+	nvkm_mask(device, 0x002630, BIT(chan->runl), BIT(chan->runl));
+
+	/* Preempt the channel. */
+	ret = gk104_fifo_gpfifo_kick_locked(chan);
+	if (ret == 0) {
+		/* Update engine context validity. */
+		nvkm_kmap(chan->base.inst);
+		nvkm_mo32(chan->base.inst, 0x0ac, mask, data);
+		nvkm_done(chan->base.inst);
+	}
+
+	/* Resume runlist. */
+	nvkm_mask(device, 0x002630, BIT(chan->runl), 0);
+	mutex_unlock(&subdev->mutex);
+	return ret;
+}
+
+static int
+gv100_fifo_gpfifo_engine_fini(struct nvkm_fifo_chan *base,
+			      struct nvkm_engine *engine, bool suspend)
+{
+	struct gk104_fifo_chan *chan = gk104_fifo_chan(base);
+	struct nvkm_gpuobj *inst = chan->base.inst;
+	int ret;
+
+	if (engine->subdev.index >= NVKM_ENGINE_CE0 &&
+	    engine->subdev.index <= NVKM_ENGINE_CE_LAST)
+		return gk104_fifo_gpfifo_kick(chan);
+
+	ret = gv100_fifo_gpfifo_engine_valid(chan, false, false);
+	if (ret && suspend)
+		return ret;
+
+	nvkm_kmap(inst);
+	nvkm_wo32(inst, 0x0210, 0x00000000);
+	nvkm_wo32(inst, 0x0214, 0x00000000);
+	nvkm_done(inst);
+	return ret;
+}
+
+static int
+gv100_fifo_gpfifo_engine_init(struct nvkm_fifo_chan *base,
+			      struct nvkm_engine *engine)
+{
+	struct gk104_fifo_chan *chan = gk104_fifo_chan(base);
+	struct nvkm_gpuobj *inst = chan->base.inst;
+	u64 addr;
+
+	if (engine->subdev.index >= NVKM_ENGINE_CE0 &&
+	    engine->subdev.index <= NVKM_ENGINE_CE_LAST)
+		return 0;
+
+	addr = chan->engn[engine->subdev.index].vma->addr;
+	nvkm_kmap(inst);
+	nvkm_wo32(inst, 0x210, lower_32_bits(addr) | 0x00000004);
+	nvkm_wo32(inst, 0x214, upper_32_bits(addr));
+	nvkm_done(inst);
+
+	return gv100_fifo_gpfifo_engine_valid(chan, false, true);
+}
+
+const struct nvkm_fifo_chan_func
+gv100_fifo_gpfifo_func = {
+	.dtor = gk104_fifo_gpfifo_dtor,
+	.init = gk104_fifo_gpfifo_init,
+	.fini = gk104_fifo_gpfifo_fini,
+	.ntfy = gf100_fifo_chan_ntfy,
+	.engine_ctor = gk104_fifo_gpfifo_engine_ctor,
+	.engine_dtor = gk104_fifo_gpfifo_engine_dtor,
+	.engine_init = gv100_fifo_gpfifo_engine_init,
+	.engine_fini = gv100_fifo_gpfifo_engine_fini,
+};
+
+static int
+gv100_fifo_gpfifo_new_(struct gk104_fifo *fifo, u64 *runlists, u16 *chid,
+		       u64 vmm, u64 ioffset, u64 ilength,
+		       const struct nvkm_oclass *oclass,
+		       struct nvkm_object **pobject)
+{
+	struct gk104_fifo_chan *chan;
+	int runlist = ffs(*runlists) -1, ret, i;
+	unsigned long engm;
+	u64 subdevs = 0;
+	u64 usermem;
+
+	if (!vmm || runlist < 0 || runlist >= fifo->runlist_nr)
+		return -EINVAL;
+	*runlists = BIT_ULL(runlist);
+
+	engm = fifo->runlist[runlist].engm;
+	for_each_set_bit(i, &engm, fifo->engine_nr) {
+		if (fifo->engine[i].engine)
+			subdevs |= BIT_ULL(fifo->engine[i].engine->subdev.index);
+	}
+
+	/* Allocate the channel. */
+	if (!(chan = kzalloc(sizeof(*chan), GFP_KERNEL)))
+		return -ENOMEM;
+	*pobject = &chan->base.object;
+	chan->fifo = fifo;
+	chan->runl = runlist;
+	INIT_LIST_HEAD(&chan->head);
+
+	ret = nvkm_fifo_chan_ctor(&gv100_fifo_gpfifo_func, &fifo->base,
+				  0x1000, 0x1000, true, vmm, 0, subdevs,
+				  1, fifo->user.bar->addr, 0x200,
+				  oclass, &chan->base);
+	if (ret)
+		return ret;
+
+	*chid = chan->base.chid;
+
+	/* Hack to support GPUs where even individual channels should be
+	 * part of a channel group.
+	 */
+	if (fifo->func->cgrp_force) {
+		if (!(chan->cgrp = kmalloc(sizeof(*chan->cgrp), GFP_KERNEL)))
+			return -ENOMEM;
+		chan->cgrp->id = chan->base.chid;
+		INIT_LIST_HEAD(&chan->cgrp->head);
+		INIT_LIST_HEAD(&chan->cgrp->chan);
+		chan->cgrp->chan_nr = 0;
+	}
+
+	/* Clear channel control registers. */
+	usermem = chan->base.chid * 0x200;
+	ilength = order_base_2(ilength / 8);
+
+	nvkm_kmap(fifo->user.mem);
+	for (i = 0; i < 0x200; i += 4)
+		nvkm_wo32(fifo->user.mem, usermem + i, 0x00000000);
+	nvkm_done(fifo->user.mem);
+	usermem = nvkm_memory_addr(fifo->user.mem) + usermem;
+
+	/* RAMFC */
+	nvkm_kmap(chan->base.inst);
+	nvkm_wo32(chan->base.inst, 0x008, lower_32_bits(usermem));
+	nvkm_wo32(chan->base.inst, 0x00c, upper_32_bits(usermem));
+	nvkm_wo32(chan->base.inst, 0x010, 0x0000face);
+	nvkm_wo32(chan->base.inst, 0x030, 0x7ffff902);
+	nvkm_wo32(chan->base.inst, 0x048, lower_32_bits(ioffset));
+	nvkm_wo32(chan->base.inst, 0x04c, upper_32_bits(ioffset) |
+					  (ilength << 16));
+	nvkm_wo32(chan->base.inst, 0x084, 0x20400000);
+	nvkm_wo32(chan->base.inst, 0x094, 0x30000001);
+	nvkm_wo32(chan->base.inst, 0x0e4, 0x00000020);
+	nvkm_wo32(chan->base.inst, 0x0e8, chan->base.chid);
+	nvkm_wo32(chan->base.inst, 0x0f4, 0x00001100);
+	nvkm_wo32(chan->base.inst, 0x0f8, 0x10003080);
+	nvkm_mo32(chan->base.inst, 0x218, 0x00000000, 0x00000000);
+	nvkm_wo32(chan->base.inst, 0x220, 0x020a1000);
+	nvkm_wo32(chan->base.inst, 0x224, 0x00000000);
+	nvkm_done(chan->base.inst);
+	return gv100_fifo_gpfifo_engine_valid(chan, true, true);
+}
+
+int
+gv100_fifo_gpfifo_new(struct gk104_fifo *fifo, const struct nvkm_oclass *oclass,
+		      void *data, u32 size, struct nvkm_object **pobject)
+{
+	struct nvkm_object *parent = oclass->parent;
+	union {
+		struct kepler_channel_gpfifo_a_v0 v0;
+	} *args = data;
+	int ret = -ENOSYS;
+
+	nvif_ioctl(parent, "create channel gpfifo size %d\n", size);
+	if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false))) {
+		nvif_ioctl(parent, "create channel gpfifo vers %d vmm %llx "
+				   "ioffset %016llx ilength %08x "
+				   "runlist %016llx\n",
+			   args->v0.version, args->v0.vmm, args->v0.ioffset,
+			   args->v0.ilength, args->v0.runlist);
+		return gv100_fifo_gpfifo_new_(fifo,
+					      &args->v0.runlist,
+					      &args->v0.chid,
+					       args->v0.vmm,
+					       args->v0.ioffset,
+					       args->v0.ilength,
+					      oclass, pobject);
+	}
+
+	return ret;
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gv100.c
new file mode 100644
index 0000000..4e1d159
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gv100.c
@@ -0,0 +1,306 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "gk104.h"
+#include "cgrp.h"
+#include "changk104.h"
+#include "user.h"
+
+#include <core/gpuobj.h>
+
+#include <nvif/class.h>
+
+static void
+gv100_fifo_runlist_chan(struct gk104_fifo_chan *chan,
+			struct nvkm_memory *memory, u32 offset)
+{
+	struct nvkm_memory *usermem = chan->fifo->user.mem;
+	const u64 user = nvkm_memory_addr(usermem) + (chan->base.chid * 0x200);
+	const u64 inst = chan->base.inst->addr;
+
+	nvkm_wo32(memory, offset + 0x0, lower_32_bits(user));
+	nvkm_wo32(memory, offset + 0x4, upper_32_bits(user));
+	nvkm_wo32(memory, offset + 0x8, lower_32_bits(inst) | chan->base.chid);
+	nvkm_wo32(memory, offset + 0xc, upper_32_bits(inst));
+}
+
+static void
+gv100_fifo_runlist_cgrp(struct nvkm_fifo_cgrp *cgrp,
+			struct nvkm_memory *memory, u32 offset)
+{
+	nvkm_wo32(memory, offset + 0x0, (128 << 24) | (3 << 16) | 0x00000001);
+	nvkm_wo32(memory, offset + 0x4, cgrp->chan_nr);
+	nvkm_wo32(memory, offset + 0x8, cgrp->id);
+	nvkm_wo32(memory, offset + 0xc, 0x00000000);
+}
+
+const struct gk104_fifo_runlist_func
+gv100_fifo_runlist = {
+	.size = 16,
+	.cgrp = gv100_fifo_runlist_cgrp,
+	.chan = gv100_fifo_runlist_chan,
+};
+
+static const struct nvkm_enum
+gv100_fifo_fault_gpcclient[] = {
+	{ 0x00, "T1_0" },
+	{ 0x01, "T1_1" },
+	{ 0x02, "T1_2" },
+	{ 0x03, "T1_3" },
+	{ 0x04, "T1_4" },
+	{ 0x05, "T1_5" },
+	{ 0x06, "T1_6" },
+	{ 0x07, "T1_7" },
+	{ 0x08, "PE_0" },
+	{ 0x09, "PE_1" },
+	{ 0x0a, "PE_2" },
+	{ 0x0b, "PE_3" },
+	{ 0x0c, "PE_4" },
+	{ 0x0d, "PE_5" },
+	{ 0x0e, "PE_6" },
+	{ 0x0f, "PE_7" },
+	{ 0x10, "RAST" },
+	{ 0x11, "GCC" },
+	{ 0x12, "GPCCS" },
+	{ 0x13, "PROP_0" },
+	{ 0x14, "PROP_1" },
+	{ 0x15, "PROP_2" },
+	{ 0x16, "PROP_3" },
+	{ 0x17, "GPM" },
+	{ 0x18, "LTP_UTLB_0" },
+	{ 0x19, "LTP_UTLB_1" },
+	{ 0x1a, "LTP_UTLB_2" },
+	{ 0x1b, "LTP_UTLB_3" },
+	{ 0x1c, "LTP_UTLB_4" },
+	{ 0x1d, "LTP_UTLB_5" },
+	{ 0x1e, "LTP_UTLB_6" },
+	{ 0x1f, "LTP_UTLB_7" },
+	{ 0x20, "RGG_UTLB" },
+	{ 0x21, "T1_8" },
+	{ 0x22, "T1_9" },
+	{ 0x23, "T1_10" },
+	{ 0x24, "T1_11" },
+	{ 0x25, "T1_12" },
+	{ 0x26, "T1_13" },
+	{ 0x27, "T1_14" },
+	{ 0x28, "T1_15" },
+	{ 0x29, "TPCCS_0" },
+	{ 0x2a, "TPCCS_1" },
+	{ 0x2b, "TPCCS_2" },
+	{ 0x2c, "TPCCS_3" },
+	{ 0x2d, "TPCCS_4" },
+	{ 0x2e, "TPCCS_5" },
+	{ 0x2f, "TPCCS_6" },
+	{ 0x30, "TPCCS_7" },
+	{ 0x31, "PE_8" },
+	{ 0x32, "PE_9" },
+	{ 0x33, "TPCCS_8" },
+	{ 0x34, "TPCCS_9" },
+	{ 0x35, "T1_16" },
+	{ 0x36, "T1_17" },
+	{ 0x37, "T1_18" },
+	{ 0x38, "T1_19" },
+	{ 0x39, "PE_10" },
+	{ 0x3a, "PE_11" },
+	{ 0x3b, "TPCCS_10" },
+	{ 0x3c, "TPCCS_11" },
+	{ 0x3d, "T1_20" },
+	{ 0x3e, "T1_21" },
+	{ 0x3f, "T1_22" },
+	{ 0x40, "T1_23" },
+	{ 0x41, "PE_12" },
+	{ 0x42, "PE_13" },
+	{ 0x43, "TPCCS_12" },
+	{ 0x44, "TPCCS_13" },
+	{ 0x45, "T1_24" },
+	{ 0x46, "T1_25" },
+	{ 0x47, "T1_26" },
+	{ 0x48, "T1_27" },
+	{ 0x49, "PE_14" },
+	{ 0x4a, "PE_15" },
+	{ 0x4b, "TPCCS_14" },
+	{ 0x4c, "TPCCS_15" },
+	{ 0x4d, "T1_28" },
+	{ 0x4e, "T1_29" },
+	{ 0x4f, "T1_30" },
+	{ 0x50, "T1_31" },
+	{ 0x51, "PE_16" },
+	{ 0x52, "PE_17" },
+	{ 0x53, "TPCCS_16" },
+	{ 0x54, "TPCCS_17" },
+	{ 0x55, "T1_32" },
+	{ 0x56, "T1_33" },
+	{ 0x57, "T1_34" },
+	{ 0x58, "T1_35" },
+	{ 0x59, "PE_18" },
+	{ 0x5a, "PE_19" },
+	{ 0x5b, "TPCCS_18" },
+	{ 0x5c, "TPCCS_19" },
+	{ 0x5d, "T1_36" },
+	{ 0x5e, "T1_37" },
+	{ 0x5f, "T1_38" },
+	{ 0x60, "T1_39" },
+	{}
+};
+
+static const struct nvkm_enum
+gv100_fifo_fault_hubclient[] = {
+	{ 0x00, "VIP" },
+	{ 0x01, "CE0" },
+	{ 0x02, "CE1" },
+	{ 0x03, "DNISO" },
+	{ 0x04, "FE" },
+	{ 0x05, "FECS" },
+	{ 0x06, "HOST" },
+	{ 0x07, "HOST_CPU" },
+	{ 0x08, "HOST_CPU_NB" },
+	{ 0x09, "ISO" },
+	{ 0x0a, "MMU" },
+	{ 0x0b, "NVDEC" },
+	{ 0x0d, "NVENC1" },
+	{ 0x0e, "NISO" },
+	{ 0x0f, "P2P" },
+	{ 0x10, "PD" },
+	{ 0x11, "PERF" },
+	{ 0x12, "PMU" },
+	{ 0x13, "RASTERTWOD" },
+	{ 0x14, "SCC" },
+	{ 0x15, "SCC_NB" },
+	{ 0x16, "SEC" },
+	{ 0x17, "SSYNC" },
+	{ 0x18, "CE2" },
+	{ 0x19, "XV" },
+	{ 0x1a, "MMU_NB" },
+	{ 0x1b, "NVENC0" },
+	{ 0x1c, "DFALCON" },
+	{ 0x1d, "SKED" },
+	{ 0x1e, "AFALCON" },
+	{ 0x1f, "DONT_CARE" },
+	{ 0x20, "HSCE0" },
+	{ 0x21, "HSCE1" },
+	{ 0x22, "HSCE2" },
+	{ 0x23, "HSCE3" },
+	{ 0x24, "HSCE4" },
+	{ 0x25, "HSCE5" },
+	{ 0x26, "HSCE6" },
+	{ 0x27, "HSCE7" },
+	{ 0x28, "HSCE8" },
+	{ 0x29, "HSCE9" },
+	{ 0x2a, "HSHUB" },
+	{ 0x2b, "PTP_X0" },
+	{ 0x2c, "PTP_X1" },
+	{ 0x2d, "PTP_X2" },
+	{ 0x2e, "PTP_X3" },
+	{ 0x2f, "PTP_X4" },
+	{ 0x30, "PTP_X5" },
+	{ 0x31, "PTP_X6" },
+	{ 0x32, "PTP_X7" },
+	{ 0x33, "NVENC2" },
+	{ 0x34, "VPR_SCRUBBER0" },
+	{ 0x35, "VPR_SCRUBBER1" },
+	{ 0x36, "DWBIF" },
+	{ 0x37, "FBFALCON" },
+	{ 0x38, "CE_SHIM" },
+	{ 0x39, "GSP" },
+	{}
+};
+
+static const struct nvkm_enum
+gv100_fifo_fault_reason[] = {
+	{ 0x00, "PDE" },
+	{ 0x01, "PDE_SIZE" },
+	{ 0x02, "PTE" },
+	{ 0x03, "VA_LIMIT_VIOLATION" },
+	{ 0x04, "UNBOUND_INST_BLOCK" },
+	{ 0x05, "PRIV_VIOLATION" },
+	{ 0x06, "RO_VIOLATION" },
+	{ 0x07, "WO_VIOLATION" },
+	{ 0x08, "PITCH_MASK_VIOLATION" },
+	{ 0x09, "WORK_CREATION" },
+	{ 0x0a, "UNSUPPORTED_APERTURE" },
+	{ 0x0b, "COMPRESSION_FAILURE" },
+	{ 0x0c, "UNSUPPORTED_KIND" },
+	{ 0x0d, "REGION_VIOLATION" },
+	{ 0x0e, "POISONED" },
+	{ 0x0f, "ATOMIC_VIOLATION" },
+	{}
+};
+
+static const struct nvkm_enum
+gv100_fifo_fault_engine[] = {
+	{ 0x01, "DISPLAY" },
+	{ 0x03, "PTP" },
+	{ 0x04, "BAR1", NULL, NVKM_SUBDEV_BAR },
+	{ 0x05, "BAR2", NULL, NVKM_SUBDEV_INSTMEM },
+	{ 0x06, "PWR_PMU" },
+	{ 0x08, "IFB", NULL, NVKM_ENGINE_IFB },
+	{ 0x09, "PERF" },
+	{ 0x1f, "PHYSICAL" },
+	{ 0x20, "HOST0" },
+	{ 0x21, "HOST1" },
+	{ 0x22, "HOST2" },
+	{ 0x23, "HOST3" },
+	{ 0x24, "HOST4" },
+	{ 0x25, "HOST5" },
+	{ 0x26, "HOST6" },
+	{ 0x27, "HOST7" },
+	{ 0x28, "HOST8" },
+	{ 0x29, "HOST9" },
+	{ 0x2a, "HOST10" },
+	{ 0x2b, "HOST11" },
+	{ 0x2c, "HOST12" },
+	{ 0x2d, "HOST13" },
+	{}
+};
+
+static const struct nvkm_enum
+gv100_fifo_fault_access[] = {
+	{ 0x0, "VIRT_READ" },
+	{ 0x1, "VIRT_WRITE" },
+	{ 0x2, "VIRT_ATOMIC" },
+	{ 0x3, "VIRT_PREFETCH" },
+	{ 0x4, "VIRT_ATOMIC_WEAK" },
+	{ 0x8, "PHYS_READ" },
+	{ 0x9, "PHYS_WRITE" },
+	{ 0xa, "PHYS_ATOMIC" },
+	{ 0xb, "PHYS_PREFETCH" },
+	{}
+};
+
+static const struct gk104_fifo_func
+gv100_fifo = {
+	.init_pbdma_timeout = gk208_fifo_init_pbdma_timeout,
+	.fault.access = gv100_fifo_fault_access,
+	.fault.engine = gv100_fifo_fault_engine,
+	.fault.reason = gv100_fifo_fault_reason,
+	.fault.hubclient = gv100_fifo_fault_hubclient,
+	.fault.gpcclient = gv100_fifo_fault_gpcclient,
+	.runlist = &gv100_fifo_runlist,
+	.user = {{-1,-1,VOLTA_USERMODE_A      }, gv100_fifo_user_new   },
+	.chan = {{ 0, 0,VOLTA_CHANNEL_GPFIFO_A}, gv100_fifo_gpfifo_new },
+	.cgrp_force = true,
+};
+
+int
+gv100_fifo_new(struct nvkm_device *device, int index, struct nvkm_fifo **pfifo)
+{
+	return gk104_fifo_new_(&gv100_fifo, device, index, 4096, pfifo);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/user.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/user.h
new file mode 100644
index 0000000..ed84092
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/user.h
@@ -0,0 +1,6 @@
+#ifndef __NVKM_FIFO_USER_H__
+#define __NVKM_FIFO_USER_H__
+#include "priv.h"
+int gv100_fifo_user_new(const struct nvkm_oclass *, void *, u32,
+			struct nvkm_object **);
+#endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/usergv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/usergv100.c
new file mode 100644
index 0000000..3dc3b8b
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/usergv100.c
@@ -0,0 +1,45 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "user.h"
+
+static int
+gv100_fifo_user_map(struct nvkm_object *object, void *argv, u32 argc,
+		    enum nvkm_object_map *type, u64 *addr, u64 *size)
+{
+	struct nvkm_device *device = object->engine->subdev.device;
+	*addr = 0x810000 + device->func->resource_addr(device, 0);
+	*size = 0x010000;
+	*type = NVKM_OBJECT_MAP_IO;
+	return 0;
+}
+
+static const struct nvkm_object_func
+gv100_fifo_user = {
+	.map = gv100_fifo_user_map,
+};
+
+int
+gv100_fifo_user_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nvkm_object **pobject)
+{
+	return nvkm_object_new_(&gv100_fifo_user, oclass, argv, argc, pobject);
+}
-- 
cgit v1.1


From 6e1f34e33c17f633ebbd383cab429c820ec0c7b0 Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:48 +1000
Subject: drm/nouveau/ce/gv100: initial support

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/include/nvif/class.h      |  1 +
 drivers/gpu/drm/nouveau/include/nvkm/engine/ce.h  |  1 +
 drivers/gpu/drm/nouveau/nouveau_bo.c              |  2 ++
 drivers/gpu/drm/nouveau/nvkm/engine/ce/Kbuild     |  1 +
 drivers/gpu/drm/nouveau/nvkm/engine/ce/gv100.c    | 40 +++++++++++++++++++++++
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c |  9 +++++
 6 files changed, 54 insertions(+)
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/ce/gv100.c

diff --git a/drivers/gpu/drm/nouveau/include/nvif/class.h b/drivers/gpu/drm/nouveau/include/nvif/class.h
index a2fdbc6..8688342 100644
--- a/drivers/gpu/drm/nouveau/include/nvif/class.h
+++ b/drivers/gpu/drm/nouveau/include/nvif/class.h
@@ -180,6 +180,7 @@
 #define MAXWELL_DMA_COPY_A                                           0x0000b0b5
 #define PASCAL_DMA_COPY_A                                            0x0000c0b5
 #define PASCAL_DMA_COPY_B                                            0x0000c1b5
+#define VOLTA_DMA_COPY_A                                             0x0000c3b5
 
 #define FERMI_DECOMPRESS                                             0x000090b8
 
diff --git a/drivers/gpu/drm/nouveau/include/nvkm/engine/ce.h b/drivers/gpu/drm/nouveau/include/nvkm/engine/ce.h
index 5532459..fc295e1 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/engine/ce.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/engine/ce.h
@@ -10,4 +10,5 @@ int gm107_ce_new(struct nvkm_device *, int, struct nvkm_engine **);
 int gm200_ce_new(struct nvkm_device *, int, struct nvkm_engine **);
 int gp100_ce_new(struct nvkm_device *, int, struct nvkm_engine **);
 int gp102_ce_new(struct nvkm_device *, int, struct nvkm_engine **);
+int gv100_ce_new(struct nvkm_device *, int, struct nvkm_engine **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.c b/drivers/gpu/drm/nouveau/nouveau_bo.c
index ab61c03..7214022 100644
--- a/drivers/gpu/drm/nouveau/nouveau_bo.c
+++ b/drivers/gpu/drm/nouveau/nouveau_bo.c
@@ -1141,6 +1141,8 @@ nouveau_bo_move_init(struct nouveau_drm *drm)
 			    struct ttm_mem_reg *, struct ttm_mem_reg *);
 		int (*init)(struct nouveau_channel *, u32 handle);
 	} _methods[] = {
+		{  "COPY", 4, 0xc3b5, nve0_bo_move_copy, nve0_bo_move_init },
+		{  "GRCE", 0, 0xc3b5, nve0_bo_move_copy, nvc0_bo_move_init },
 		{  "COPY", 4, 0xc1b5, nve0_bo_move_copy, nve0_bo_move_init },
 		{  "GRCE", 0, 0xc1b5, nve0_bo_move_copy, nvc0_bo_move_init },
 		{  "COPY", 4, 0xc0b5, nve0_bo_move_copy, nve0_bo_move_init },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/ce/Kbuild b/drivers/gpu/drm/nouveau/nvkm/engine/ce/Kbuild
index 255d81c..80d7844 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/ce/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/ce/Kbuild
@@ -5,3 +5,4 @@ nvkm-y += nvkm/engine/ce/gm107.o
 nvkm-y += nvkm/engine/ce/gm200.o
 nvkm-y += nvkm/engine/ce/gp100.o
 nvkm-y += nvkm/engine/ce/gp102.o
+nvkm-y += nvkm/engine/ce/gv100.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/ce/gv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/ce/gv100.c
new file mode 100644
index 0000000..fcda3de
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/ce/gv100.c
@@ -0,0 +1,40 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "priv.h"
+
+#include <nvif/class.h>
+
+static const struct nvkm_engine_func
+gv100_ce = {
+	.intr = gp100_ce_intr,
+	.sclass = {
+		{ -1, -1, VOLTA_DMA_COPY_A },
+		{}
+	}
+};
+
+int
+gv100_ce_new(struct nvkm_device *device, int index,
+	     struct nvkm_engine **pengine)
+{
+	return nvkm_engine_new_(&gv100_ce, device, index, true, pengine);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 95a56d4..5fa3061 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2417,6 +2417,15 @@ nv140_chipset = {
 	.timer = gk20a_timer_new,
 	.top = gk104_top_new,
 	.disp = gv100_disp_new,
+	.ce[0] = gv100_ce_new,
+	.ce[1] = gv100_ce_new,
+	.ce[2] = gv100_ce_new,
+	.ce[3] = gv100_ce_new,
+	.ce[4] = gv100_ce_new,
+	.ce[5] = gv100_ce_new,
+	.ce[6] = gv100_ce_new,
+	.ce[7] = gv100_ce_new,
+	.ce[8] = gv100_ce_new,
 	.dma = gv100_dma_new,
 	.fifo = gv100_fifo_new,
 };
-- 
cgit v1.1


From d521097f58bdfdc9966b8d10754074c8524133dd Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Tue, 8 May 2018 20:39:48 +1000
Subject: drm/nouveau/gr/gv100: initial support

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/include/nvif/class.h       |   3 +
 drivers/gpu/drm/nouveau/include/nvkm/engine/gr.h   |   1 +
 drivers/gpu/drm/nouveau/nvkm/engine/device/base.c  |   4 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/Kbuild      |   2 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c  |  10 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h  |   5 +
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgv100.c  | 215 +++++++++++++++++++++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c     |  10 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h     |   6 +-
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gv100.c     | 120 ++++++++++++
 drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue.c     |   1 +
 .../gpu/drm/nouveau/nvkm/subdev/secboot/gp108.c    |  21 ++
 12 files changed, 395 insertions(+), 3 deletions(-)
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgv100.c
 create mode 100644 drivers/gpu/drm/nouveau/nvkm/engine/gr/gv100.c

diff --git a/drivers/gpu/drm/nouveau/include/nvif/class.h b/drivers/gpu/drm/nouveau/include/nvif/class.h
index 8688342..6db56bd 100644
--- a/drivers/gpu/drm/nouveau/include/nvif/class.h
+++ b/drivers/gpu/drm/nouveau/include/nvif/class.h
@@ -155,6 +155,8 @@
 #define PASCAL_A                                      /* cl9097.h */ 0x0000c097
 #define PASCAL_B                                      /* cl9097.h */ 0x0000c197
 
+#define VOLTA_A                                       /* cl9097.h */ 0x0000c397
+
 #define NV74_BSP                                                     0x000074b0
 
 #define GT212_MSVLD                                                  0x000085b1
@@ -194,6 +196,7 @@
 #define MAXWELL_COMPUTE_B                                            0x0000b1c0
 #define PASCAL_COMPUTE_A                                             0x0000c0c0
 #define PASCAL_COMPUTE_B                                             0x0000c1c0
+#define VOLTA_COMPUTE_A                                              0x0000c3c0
 
 #define NV74_CIPHER                                                  0x000074c1
 #endif
diff --git a/drivers/gpu/drm/nouveau/include/nvkm/engine/gr.h b/drivers/gpu/drm/nouveau/include/nvkm/engine/gr.h
index 33b2f2e..ba1518f 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/engine/gr.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/engine/gr.h
@@ -48,4 +48,5 @@ int gp102_gr_new(struct nvkm_device *, int, struct nvkm_gr **);
 int gp104_gr_new(struct nvkm_device *, int, struct nvkm_gr **);
 int gp107_gr_new(struct nvkm_device *, int, struct nvkm_gr **);
 int gp10b_gr_new(struct nvkm_device *, int, struct nvkm_gr **);
+int gv100_gr_new(struct nvkm_device *, int, struct nvkm_gr **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 5fa3061..e294013 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2413,6 +2413,7 @@ nv140_chipset = {
 	.mmu = gv100_mmu_new,
 	.pci = gp100_pci_new,
 	.pmu = gp102_pmu_new,
+	.secboot = gp108_secboot_new,
 	.therm = gp100_therm_new,
 	.timer = gk20a_timer_new,
 	.top = gk104_top_new,
@@ -2428,6 +2429,9 @@ nv140_chipset = {
 	.ce[8] = gv100_ce_new,
 	.dma = gv100_dma_new,
 	.fifo = gv100_fifo_new,
+	.gr = gv100_gr_new,
+	.nvdec = gp102_nvdec_new,
+	.sec2 = gp102_sec2_new,
 };
 
 static int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/Kbuild b/drivers/gpu/drm/nouveau/nvkm/engine/gr/Kbuild
index 42342b4..93e3733 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/Kbuild
@@ -36,6 +36,7 @@ nvkm-y += nvkm/engine/gr/gp102.o
 nvkm-y += nvkm/engine/gr/gp104.o
 nvkm-y += nvkm/engine/gr/gp107.o
 nvkm-y += nvkm/engine/gr/gp10b.o
+nvkm-y += nvkm/engine/gr/gv100.o
 
 nvkm-y += nvkm/engine/gr/ctxnv40.o
 nvkm-y += nvkm/engine/gr/ctxnv50.o
@@ -57,3 +58,4 @@ nvkm-y += nvkm/engine/gr/ctxgp100.o
 nvkm-y += nvkm/engine/gr/ctxgp102.o
 nvkm-y += nvkm/engine/gr/ctxgp104.o
 nvkm-y += nvkm/engine/gr/ctxgp107.o
+nvkm-y += nvkm/engine/gr/ctxgv100.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index 1ed63ed..f0f5a51 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1396,10 +1396,14 @@ gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	gf100_grctx_generate_floorsweep(gr);
 
+	if (grctx->r400088) grctx->r400088(gr, false);
 	if (gr->fuc_bundle)
 		gf100_gr_icmd(gr, gr->fuc_bundle);
 	else
 		gf100_gr_icmd(gr, grctx->icmd);
+	if (grctx->sw_veid_bundle_init)
+		gf100_gr_icmd(gr, grctx->sw_veid_bundle_init);
+	if (grctx->r400088) grctx->r400088(gr, true);
 
 	nvkm_wr32(device, 0x404154, idle_timeout);
 
@@ -1448,6 +1452,9 @@ gf100_grctx_generate(struct gf100_gr *gr)
 			break;
 	);
 
+	if (grctx->unkn88c)
+		grctx->unkn88c(gr, true);
+
 	/* Reset FECS. */
 	nvkm_wr32(device, 0x409614, 0x00000070);
 	nvkm_usec(device, 10, NVKM_DELAY);
@@ -1455,6 +1462,9 @@ gf100_grctx_generate(struct gf100_gr *gr)
 	nvkm_usec(device, 10, NVKM_DELAY);
 	nvkm_rd32(device, 0x409614);
 
+	if (grctx->unkn88c)
+		grctx->unkn88c(gr, false);
+
 	/* NV_PGRAPH_FE_PWR_MODE_AUTO. */
 	nvkm_wr32(device, 0x404170, 0x00000010);
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
index 9ce3d00..33e932b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
@@ -21,6 +21,7 @@ void gf100_grctx_mmio_item(struct gf100_grctx *, u32 addr, u32 data, int s, int)
 #define mmio_wr32(a,b,c) mmio_refn((a), (b), (c),  0, -1)
 
 struct gf100_grctx_func {
+	void (*unkn88c)(struct gf100_gr *, bool on);
 	/* main context generation function */
 	void  (*main)(struct gf100_gr *, struct gf100_grctx *);
 	/* context-specific modify-on-first-load list generation function */
@@ -35,6 +36,7 @@ struct gf100_grctx_func {
 	/* indirect context data, generated with icmds/mthds */
 	const struct gf100_gr_pack *icmd;
 	const struct gf100_gr_pack *mthd;
+	const struct gf100_gr_pack *sw_veid_bundle_init;
 	/* bundle circular buffer */
 	void (*bundle)(struct gf100_grctx *);
 	u32 bundle_size;
@@ -66,6 +68,7 @@ struct gf100_grctx_func {
 	void (*tpc_mask)(struct gf100_gr *);
 	void (*smid_config)(struct gf100_gr *);
 	/* misc other things */
+	void (*r400088)(struct gf100_gr *, bool);
 	void (*r419cb8)(struct gf100_gr *);
 	void (*r418800)(struct gf100_gr *);
 	void (*r419eb0)(struct gf100_gr *);
@@ -148,6 +151,8 @@ extern const struct gf100_grctx_func gp104_grctx;
 
 extern const struct gf100_grctx_func gp107_grctx;
 
+extern const struct gf100_grctx_func gv100_grctx;
+
 /* context init value lists */
 
 extern const struct gf100_gr_pack gf100_grctx_pack_icmd[];
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgv100.c
new file mode 100644
index 0000000..0990765
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgv100.c
@@ -0,0 +1,215 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "ctxgf100.h"
+
+/*******************************************************************************
+ * PGRAPH context implementation
+ ******************************************************************************/
+
+static const struct gf100_gr_init
+gv100_grctx_init_sw_veid_bundle_init_0[] = {
+	{ 0x00001000, 64, 0x00100000, 0x00000008 },
+	{ 0x00000941, 64, 0x00100000, 0x00000000 },
+	{ 0x0000097e, 64, 0x00100000, 0x00000000 },
+	{ 0x0000097f, 64, 0x00100000, 0x00000100 },
+	{ 0x0000035c, 64, 0x00100000, 0x00000000 },
+	{ 0x0000035d, 64, 0x00100000, 0x00000000 },
+	{ 0x00000a08, 64, 0x00100000, 0x00000000 },
+	{ 0x00000a09, 64, 0x00100000, 0x00000000 },
+	{ 0x00000a0a, 64, 0x00100000, 0x00000000 },
+	{ 0x00000352, 64, 0x00100000, 0x00000000 },
+	{ 0x00000353, 64, 0x00100000, 0x00000000 },
+	{ 0x00000358, 64, 0x00100000, 0x00000000 },
+	{ 0x00000359, 64, 0x00100000, 0x00000000 },
+	{ 0x00000370, 64, 0x00100000, 0x00000000 },
+	{ 0x00000371, 64, 0x00100000, 0x00000000 },
+	{ 0x00000372, 64, 0x00100000, 0x000fffff },
+	{ 0x00000366, 64, 0x00100000, 0x00000000 },
+	{ 0x00000367, 64, 0x00100000, 0x00000000 },
+	{ 0x00000368, 64, 0x00100000, 0x00000fff },
+	{ 0x00000623, 64, 0x00100000, 0x00000000 },
+	{ 0x00000624, 64, 0x00100000, 0x00000000 },
+	{ 0x0001e100,  1, 0x00000001, 0x02000001 },
+	{}
+};
+
+static const struct gf100_gr_pack
+gv100_grctx_pack_sw_veid_bundle_init[] = {
+	{ gv100_grctx_init_sw_veid_bundle_init_0 },
+	{}
+};
+
+static void
+gv100_grctx_generate_attrib(struct gf100_grctx *info)
+{
+	struct gf100_gr *gr = info->gr;
+	const struct gf100_grctx_func *grctx = gr->func->grctx;
+	const u32  alpha = grctx->alpha_nr;
+	const u32 attrib = grctx->attrib_nr;
+	const u32   gfxp = grctx->gfxp_nr;
+	const int s = 12;
+	const int max_batches = 0xffff;
+	u32 size = grctx->alpha_nr_max * gr->tpc_total;
+	u32 ao = 0;
+	u32 bo = ao + size;
+	int gpc, ppc, b, n = 0;
+
+	size += grctx->gfxp_nr * gr->tpc_total;
+	size = ((size * 0x20) + 128) & ~127;
+	b = mmio_vram(info, size, (1 << s), false);
+
+	mmio_refn(info, 0x418810, 0x80000000, s, b);
+	mmio_refn(info, 0x419848, 0x10000000, s, b);
+	mmio_refn(info, 0x419c2c, 0x10000000, s, b);
+	mmio_refn(info, 0x419e00, 0x00000000, s, b);
+	mmio_wr32(info, 0x419e04, 0x80000000 | size >> 7);
+	mmio_wr32(info, 0x405830, attrib);
+	mmio_wr32(info, 0x40585c, alpha);
+	mmio_wr32(info, 0x4064c4, ((alpha / 4) << 16) | max_batches);
+
+	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
+		for (ppc = 0; ppc < gr->ppc_nr[gpc]; ppc++, n++) {
+			const u32 as =  alpha * gr->ppc_tpc_nr[gpc][ppc];
+			const u32 bs = attrib * gr->ppc_tpc_nr[gpc][ppc];
+			const u32 gs =   gfxp * gr->ppc_tpc_nr[gpc][ppc];
+			const u32 u = 0x418ea0 + (n * 0x04);
+			const u32 o = PPC_UNIT(gpc, ppc, 0);
+			if (!(gr->ppc_mask[gpc] & (1 << ppc)))
+				continue;
+			mmio_wr32(info, o + 0xc0, gs);
+			mmio_wr32(info, o + 0xf4, bo);
+			mmio_wr32(info, o + 0xf0, bs);
+			bo += gs;
+			mmio_wr32(info, o + 0xe4, as);
+			mmio_wr32(info, o + 0xf8, ao);
+			ao += grctx->alpha_nr_max * gr->ppc_tpc_nr[gpc][ppc];
+			mmio_wr32(info, u, bs);
+		}
+	}
+
+	mmio_wr32(info, 0x4181e4, 0x00000100);
+	mmio_wr32(info, 0x41befc, 0x00000100);
+}
+
+static void
+gv100_grctx_generate_rop_mapping(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	u32 data;
+	int i, j;
+
+	/* Pack tile map into register format. */
+	nvkm_wr32(device, 0x418bb8, (gr->tpc_total << 8) |
+				     gr->screen_tile_row_offset);
+	for (i = 0; i < 11; i++) {
+		for (data = 0, j = 0; j < 6; j++)
+			data |= (gr->tile[i * 6 + j] & 0x1f) << (j * 5);
+		nvkm_wr32(device, 0x418b08 + (i * 4), data);
+		nvkm_wr32(device, 0x41bf00 + (i * 4), data);
+		nvkm_wr32(device, 0x40780c + (i * 4), data);
+	}
+
+	/* GPC_BROADCAST.TP_BROADCAST */
+	nvkm_wr32(device, 0x41bfd0, (gr->tpc_total << 8) |
+				     gr->screen_tile_row_offset);
+	for (i = 0, j = 1; i < 5; i++, j += 4) {
+		u8 v19 = (1 << (j + 0)) % gr->tpc_total;
+		u8 v20 = (1 << (j + 1)) % gr->tpc_total;
+		u8 v21 = (1 << (j + 2)) % gr->tpc_total;
+		u8 v22 = (1 << (j + 3)) % gr->tpc_total;
+		nvkm_wr32(device, 0x41bfb0 + (i * 4), (v22 << 24) |
+						      (v21 << 16) |
+						      (v20 <<  8) |
+						       v19);
+	}
+
+	/* UNK78xx */
+	nvkm_wr32(device, 0x4078bc, (gr->tpc_total << 8) |
+				     gr->screen_tile_row_offset);
+}
+
+static void
+gv100_grctx_generate_r400088(struct gf100_gr *gr, bool on)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x400088, 0x00060000, on ? 0x00060000 : 0x00000000);
+}
+
+static void
+gv100_grctx_generate_sm_id(struct gf100_gr *gr, int gpc, int tpc, int sm)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x608), sm);
+	nvkm_wr32(device, GPC_UNIT(gpc, 0x0c10 + tpc * 4), sm);
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x088), sm);
+}
+
+static void
+gv100_grctx_generate_unkn(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x41980c, 0x00000010, 0x00000010);
+	nvkm_mask(device, 0x41be08, 0x00000004, 0x00000004);
+	nvkm_mask(device, 0x4064c0, 0x80000000, 0x80000000);
+	nvkm_mask(device, 0x405800, 0x08000000, 0x08000000);
+	nvkm_mask(device, 0x419c00, 0x00000008, 0x00000008);
+}
+
+static void
+gv100_grctx_unkn88c(struct gf100_gr *gr, bool on)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	const u32 mask = 0x00000010, data = on ? mask : 0x00000000;
+	nvkm_mask(device, 0x40988c, mask, data);
+	nvkm_rd32(device, 0x40988c);
+	nvkm_mask(device, 0x41a88c, mask, data);
+	nvkm_rd32(device, 0x41a88c);
+	nvkm_mask(device, 0x408a14, mask, data);
+	nvkm_rd32(device, 0x408a14);
+}
+
+const struct gf100_grctx_func
+gv100_grctx = {
+	.unkn88c = gv100_grctx_unkn88c,
+	.main = gf100_grctx_generate_main,
+	.unkn = gv100_grctx_generate_unkn,
+	.sw_veid_bundle_init = gv100_grctx_pack_sw_veid_bundle_init,
+	.bundle = gm107_grctx_generate_bundle,
+	.bundle_size = 0x3000,
+	.bundle_min_gpm_fifo_depth = 0x180,
+	.bundle_token_limit = 0x1680,
+	.pagepool = gp100_grctx_generate_pagepool,
+	.pagepool_size = 0x20000,
+	.attrib = gv100_grctx_generate_attrib,
+	.attrib_nr_max = 0x6c0,
+	.attrib_nr = 0x480,
+	.alpha_nr_max = 0xc00,
+	.alpha_nr = 0x800,
+	.gfxp_nr = 0xd10,
+	.sm_id = gv100_grctx_generate_sm_id,
+	.rop_mapping = gv100_grctx_generate_rop_mapping,
+	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
+	.r406500 = gm200_grctx_generate_r406500,
+	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
+	.smid_config = gp100_grctx_generate_smid_config,
+	.r400088 = gv100_grctx_generate_r400088,
+};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 86ae5c7..8dd4bd71 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -987,7 +987,7 @@ gf100_gr_trap_gpc_rop(struct gf100_gr *gr, int gpc)
 	nvkm_wr32(device, GPC_UNIT(gpc, 0x0420), 0xc0000000);
 }
 
-static const struct nvkm_enum gf100_mp_warp_error[] = {
+const struct nvkm_enum gf100_mp_warp_error[] = {
 	{ 0x01, "STACK_ERROR" },
 	{ 0x02, "API_STACK_ERROR" },
 	{ 0x03, "RET_EMPTY_STACK_ERROR" },
@@ -1012,7 +1012,7 @@ static const struct nvkm_enum gf100_mp_warp_error[] = {
 	{}
 };
 
-static const struct nvkm_bitfield gf100_mp_global_error[] = {
+const struct nvkm_bitfield gf100_mp_global_error[] = {
 	{ 0x00000001, "SM_TO_SM_FAULT" },
 	{ 0x00000002, "L1_ERROR" },
 	{ 0x00000004, "MULTIPLE_WARP_ERRORS" },
@@ -2113,6 +2113,9 @@ gf100_gr_init(struct gf100_gr *gr)
 	struct nvkm_device *device = gr->base.engine.subdev.device;
 	int gpc, tpc, rop;
 
+	if (gr->func->init_419bd8)
+		gr->func->init_419bd8(gr);
+
 	gr->func->init_gpc_mmu(gr);
 
 	if (gr->fuc_sw_nonctx)
@@ -2213,6 +2216,9 @@ gf100_gr_init(struct gf100_gr *gr)
 
 	gf100_gr_zbc_init(gr);
 
+	if (gr->func->init_4188a4)
+		gr->func->init_4188a4(gr);
+
 	return gf100_gr_init_ctxctl(gr);
 }
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index edf6eda..dc46cf01 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -149,6 +149,7 @@ struct gf100_gr_func {
 	void (*oneinit_tiles)(struct gf100_gr *);
 	void (*oneinit_sm_id)(struct gf100_gr *);
 	int (*init)(struct gf100_gr *);
+	void (*init_419bd8)(struct gf100_gr *);
 	void (*init_gpc_mmu)(struct gf100_gr *);
 	void (*init_r405a14)(struct gf100_gr *);
 	void (*init_bios)(struct gf100_gr *);
@@ -170,6 +171,7 @@ struct gf100_gr_func {
 	void (*init_504430)(struct gf100_gr *, int gpc, int tpc);
 	void (*init_shader_exceptions)(struct gf100_gr *, int gpc, int tpc);
 	void (*init_400054)(struct gf100_gr *);
+	void (*init_4188a4)(struct gf100_gr *);
 	void (*trap_mp)(struct gf100_gr *, int gpc, int tpc);
 	void (*set_hww_esr_report_mask)(struct gf100_gr *);
 	const struct gf100_gr_pack *mmio;
@@ -266,7 +268,7 @@ extern const struct nvkm_object_func gf100_fermi;
 struct gf100_gr_init {
 	u32 addr;
 	u8  count;
-	u8  pitch;
+	u32 pitch;
 	u32 data;
 };
 
@@ -337,6 +339,8 @@ extern const struct gf100_gr_init gf100_gr_init_fe_1[];
 extern const struct gf100_gr_init gf100_gr_init_pe_1[];
 void gf100_gr_init_gpc_mmu(struct gf100_gr *);
 void gf100_gr_trap_mp(struct gf100_gr *, int, int);
+extern const struct nvkm_bitfield gf100_mp_global_error[];
+extern const struct nvkm_enum gf100_mp_warp_error[];
 
 extern const struct gf100_gr_init gf104_gr_init_ds_0[];
 extern const struct gf100_gr_init gf104_gr_init_tex_0[];
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gv100.c
new file mode 100644
index 0000000..19173ea
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gv100.c
@@ -0,0 +1,120 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "gf100.h"
+#include "ctxgf100.h"
+
+#include <nvif/class.h>
+
+static void
+gv100_gr_trap_mp(struct gf100_gr *gr, int gpc, int tpc)
+{
+	struct nvkm_subdev *subdev = &gr->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	u32 werr = nvkm_rd32(device, TPC_UNIT(gpc, tpc, 0x730));
+	u32 gerr = nvkm_rd32(device, TPC_UNIT(gpc, tpc, 0x734));
+	const struct nvkm_enum *warp;
+	char glob[128];
+
+	nvkm_snprintbf(glob, sizeof(glob), gf100_mp_global_error, gerr);
+	warp = nvkm_enum_find(gf100_mp_warp_error, werr & 0xffff);
+
+	nvkm_error(subdev, "GPC%i/TPC%i/MP trap: "
+			   "global %08x [%s] warp %04x [%s]\n",
+		   gpc, tpc, gerr, glob, werr, warp ? warp->name : "");
+
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x730), 0x00000000);
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x734), gerr);
+}
+
+static void
+gv100_gr_init_4188a4(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x4188a4, 0x03000000, 0x03000000);
+}
+
+static void
+gv100_gr_init_shader_exceptions(struct gf100_gr *gr, int gpc, int tpc)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	int sm;
+	for (sm = 0; sm < 0x100; sm += 0x80) {
+		nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x728 + sm), 0x0085eb64);
+		nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x610), 0x00000001);
+		nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x72c + sm), 0x00000004);
+	}
+}
+
+static void
+gv100_gr_init_504430(struct gf100_gr *gr, int gpc, int tpc)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x430), 0x403f0000);
+}
+
+static void
+gv100_gr_init_419bd8(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x419bd8, 0x00000700, 0x00000000);
+}
+
+static const struct gf100_gr_func
+gv100_gr = {
+	.oneinit_tiles = gm200_gr_oneinit_tiles,
+	.oneinit_sm_id = gm200_gr_oneinit_sm_id,
+	.init = gf100_gr_init,
+	.init_419bd8 = gv100_gr_init_419bd8,
+	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gm200_gr_init_num_active_ltcs,
+	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
+	.init_swdx_pes_mask = gp102_gr_init_swdx_pes_mask,
+	.init_fecs_exceptions = gp100_gr_init_fecs_exceptions,
+	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
+	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
+	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
+	.init_504430 = gv100_gr_init_504430,
+	.init_shader_exceptions = gv100_gr_init_shader_exceptions,
+	.init_4188a4 = gv100_gr_init_4188a4,
+	.trap_mp = gv100_gr_trap_mp,
+	.rops = gm200_gr_rops,
+	.gpc_nr = 6,
+	.tpc_nr = 5,
+	.ppc_nr = 3,
+	.grctx = &gv100_grctx,
+	.zbc = &gp102_gr_zbc,
+	.sclass = {
+		{ -1, -1, FERMI_TWOD_A },
+		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
+		{ -1, -1, VOLTA_A, &gf100_fermi },
+		{ -1, -1, VOLTA_COMPUTE_A },
+		{}
+	}
+};
+
+int
+gv100_gr_new(struct nvkm_device *device, int index, struct nvkm_gr **pgr)
+{
+	return gm200_gr_new_(&gv100_gr, device, index, pgr);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue.c b/drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue.c
index 58a59b7..771e16a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue.c
+++ b/drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue.c
@@ -506,6 +506,7 @@ nvkm_msgqueue_new(u32 version, struct nvkm_falcon *falcon,
 		break;
 	case 0x0148cdec:
 	case 0x015ccf3e:
+	case 0x0167d263:
 		ret = msgqueue_0148cdec_new(falcon, sb, queue);
 		break;
 	default:
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gp108.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gp108.c
index e8c27ec..737a8d5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gp108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gp108.c
@@ -65,3 +65,24 @@ MODULE_FIRMWARE("nvidia/gp108/nvdec/scrubber.bin");
 MODULE_FIRMWARE("nvidia/gp108/sec2/desc.bin");
 MODULE_FIRMWARE("nvidia/gp108/sec2/image.bin");
 MODULE_FIRMWARE("nvidia/gp108/sec2/sig.bin");
+
+MODULE_FIRMWARE("nvidia/gv100/acr/bl.bin");
+MODULE_FIRMWARE("nvidia/gv100/acr/unload_bl.bin");
+MODULE_FIRMWARE("nvidia/gv100/acr/ucode_load.bin");
+MODULE_FIRMWARE("nvidia/gv100/acr/ucode_unload.bin");
+MODULE_FIRMWARE("nvidia/gv100/gr/fecs_bl.bin");
+MODULE_FIRMWARE("nvidia/gv100/gr/fecs_inst.bin");
+MODULE_FIRMWARE("nvidia/gv100/gr/fecs_data.bin");
+MODULE_FIRMWARE("nvidia/gv100/gr/fecs_sig.bin");
+MODULE_FIRMWARE("nvidia/gv100/gr/gpccs_bl.bin");
+MODULE_FIRMWARE("nvidia/gv100/gr/gpccs_inst.bin");
+MODULE_FIRMWARE("nvidia/gv100/gr/gpccs_data.bin");
+MODULE_FIRMWARE("nvidia/gv100/gr/gpccs_sig.bin");
+MODULE_FIRMWARE("nvidia/gv100/gr/sw_ctx.bin");
+MODULE_FIRMWARE("nvidia/gv100/gr/sw_nonctx.bin");
+MODULE_FIRMWARE("nvidia/gv100/gr/sw_bundle_init.bin");
+MODULE_FIRMWARE("nvidia/gv100/gr/sw_method_init.bin");
+MODULE_FIRMWARE("nvidia/gv100/nvdec/scrubber.bin");
+MODULE_FIRMWARE("nvidia/gv100/sec2/desc.bin");
+MODULE_FIRMWARE("nvidia/gv100/sec2/image.bin");
+MODULE_FIRMWARE("nvidia/gv100/sec2/sig.bin");
-- 
cgit v1.1


From 54b202f1d83074074562f645e9d1ba4e7b6f1cca Mon Sep 17 00:00:00 2001
From: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Date: Tue, 24 Apr 2018 15:15:10 +0200
Subject: drm/nouveau: fix mode_valid's return type

The method struct drm_connector_helper_funcs::mode_valid is defined
as returning an 'enum drm_mode_status' but the driver implementation
for this method uses an 'int' for it.

Fix this by using 'enum drm_mode_status' in the driver too.

Signed-off-by: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nouveau_connector.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/nouveau/nouveau_connector.c b/drivers/gpu/drm/nouveau/nouveau_connector.c
index 18e3239..7b557c3 100644
--- a/drivers/gpu/drm/nouveau/nouveau_connector.c
+++ b/drivers/gpu/drm/nouveau/nouveau_connector.c
@@ -1005,7 +1005,7 @@ get_tmds_link_bandwidth(struct drm_connector *connector, bool hdmi)
 		return 112000;
 }
 
-static int
+static enum drm_mode_status
 nouveau_connector_mode_valid(struct drm_connector *connector,
 			     struct drm_display_mode *mode)
 {
-- 
cgit v1.1


From f43cda5c76922777f4fe5026ee5984364ae5a918 Mon Sep 17 00:00:00 2001
From: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Date: Tue, 24 Apr 2018 15:15:38 +0200
Subject: drm/nouveau: fix nouveau_dsm_get_client_id()'s return type

The method struct vga_switcheroo_handler::get_client_id() is defined
as returning an 'enum vga_switcheroo_client_id' but the implementation
in this driver, nouveau_dsm_get_client_id(), returns an 'int'.

Fix this by returning 'enum vga_switcheroo_client_id' in this driver too.

Signed-off-by: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nouveau_acpi.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/nouveau/nouveau_acpi.c b/drivers/gpu/drm/nouveau/nouveau_acpi.c
index 5ffcb66..ffb1958 100644
--- a/drivers/gpu/drm/nouveau/nouveau_acpi.c
+++ b/drivers/gpu/drm/nouveau/nouveau_acpi.c
@@ -193,7 +193,7 @@ static int nouveau_dsm_power_state(enum vga_switcheroo_client_id id,
 	return nouveau_dsm_set_discrete_state(nouveau_dsm_priv.dhandle, state);
 }
 
-static int nouveau_dsm_get_client_id(struct pci_dev *pdev)
+static enum vga_switcheroo_client_id nouveau_dsm_get_client_id(struct pci_dev *pdev)
 {
 	/* easy option one - intel vendor ID means Integrated */
 	if (pdev->vendor == PCI_VENDOR_ID_INTEL)
-- 
cgit v1.1


From 7a22c737faef99d6f75d7049c1a2f6f0fdefb1ec Mon Sep 17 00:00:00 2001
From: Ilia Mirkin <imirkin@alum.mit.edu>
Date: Sun, 22 Apr 2018 17:47:12 -0400
Subject: drm/nouveau: fix temp/pwm visibility, skip hwmon when no sensors
 exist

A NV34 GPU was seeing temp and pwm entries in hwmon, which would error
out when read. These should not have been visible, but also the whole
hwmon object should just not have been registered in the first place.

Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nouveau_hwmon.c | 16 +++++++++++++---
 1 file changed, 13 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nouveau_hwmon.c b/drivers/gpu/drm/nouveau/nouveau_hwmon.c
index 7c96564..44178b4 100644
--- a/drivers/gpu/drm/nouveau/nouveau_hwmon.c
+++ b/drivers/gpu/drm/nouveau/nouveau_hwmon.c
@@ -327,7 +327,7 @@ nouveau_temp_is_visible(const void *data, u32 attr, int channel)
 	struct nouveau_drm *drm = nouveau_drm((struct drm_device *)data);
 	struct nvkm_therm *therm = nvxx_therm(&drm->client.device);
 
-	if (therm && therm->attr_get && nvkm_therm_temp_get(therm) < 0)
+	if (!therm || !therm->attr_get || nvkm_therm_temp_get(therm) < 0)
 		return 0;
 
 	switch (attr) {
@@ -351,8 +351,8 @@ nouveau_pwm_is_visible(const void *data, u32 attr, int channel)
 	struct nouveau_drm *drm = nouveau_drm((struct drm_device *)data);
 	struct nvkm_therm *therm = nvxx_therm(&drm->client.device);
 
-	if (therm && therm->attr_get && therm->fan_get &&
-				therm->fan_get(therm) < 0)
+	if (!therm || !therm->attr_get || !therm->fan_get ||
+	    therm->fan_get(therm) < 0)
 		return 0;
 
 	switch (attr) {
@@ -707,13 +707,20 @@ nouveau_hwmon_init(struct drm_device *dev)
 {
 #if defined(CONFIG_HWMON) || (defined(MODULE) && defined(CONFIG_HWMON_MODULE))
 	struct nouveau_drm *drm = nouveau_drm(dev);
+	struct nvkm_iccsense *iccsense = nvxx_iccsense(&drm->client.device);
 	struct nvkm_therm *therm = nvxx_therm(&drm->client.device);
+	struct nvkm_volt *volt = nvxx_volt(&drm->client.device);
 	const struct attribute_group *special_groups[N_ATTR_GROUPS];
 	struct nouveau_hwmon *hwmon;
 	struct device *hwmon_dev;
 	int ret = 0;
 	int i = 0;
 
+	if (!iccsense && !therm && !volt) {
+		NV_DEBUG(drm, "Skipping hwmon registration\n");
+		return 0;
+	}
+
 	hwmon = drm->hwmon = kzalloc(sizeof(*hwmon), GFP_KERNEL);
 	if (!hwmon)
 		return -ENOMEM;
@@ -749,6 +756,9 @@ nouveau_hwmon_fini(struct drm_device *dev)
 #if defined(CONFIG_HWMON) || (defined(MODULE) && defined(CONFIG_HWMON_MODULE))
 	struct nouveau_hwmon *hwmon = nouveau_hwmon(dev);
 
+	if (!hwmon)
+		return;
+
 	if (hwmon->hwmon)
 		hwmon_device_unregister(hwmon->hwmon);
 
-- 
cgit v1.1


From dd3b89be3eafd1c9977e350e81c5556230319101 Mon Sep 17 00:00:00 2001
From: Arushi Singhal <arushisinghal19971997@gmail.com>
Date: Tue, 8 May 2018 23:13:09 +1000
Subject: drm/nouveau/clk: Use list_for_each_entry_from_reverse

It's better to use "list_for_each_entry_from_reverse" for iterating list
than "for loop" as it makes the code more clear to read.
This patch replace "for loop" with "list_for_each_entry_from_reverse"
and "start" variable with "cstate" which helps in refactoring
the code and also "cstate" variable is more commonly used in the other
functions.

changes in v2:
"start" variable is removed, before "cstate" variable was removed
but "cstate" is more common so preferred "cstate" over "start".

Signed-off-by: Arushi Singhal <arushisinghal19971997@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/subdev/clk/base.c | 10 ++++------
 1 file changed, 4 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/clk/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/clk/base.c
index 81c3567..ba6a868 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/clk/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/clk/base.c
@@ -109,18 +109,17 @@ nvkm_cstate_valid(struct nvkm_clk *clk, struct nvkm_cstate *cstate,
 
 static struct nvkm_cstate *
 nvkm_cstate_find_best(struct nvkm_clk *clk, struct nvkm_pstate *pstate,
-		      struct nvkm_cstate *start)
+		      struct nvkm_cstate *cstate)
 {
 	struct nvkm_device *device = clk->subdev.device;
 	struct nvkm_volt *volt = device->volt;
-	struct nvkm_cstate *cstate;
 	int max_volt;
 
-	if (!pstate || !start)
+	if (!pstate || !cstate)
 		return NULL;
 
 	if (!volt)
-		return start;
+		return cstate;
 
 	max_volt = volt->max_uv;
 	if (volt->max0_id != 0xff)
@@ -133,8 +132,7 @@ nvkm_cstate_find_best(struct nvkm_clk *clk, struct nvkm_pstate *pstate,
 		max_volt = min(max_volt,
 			       nvkm_volt_map(volt, volt->max2_id, clk->temp));
 
-	for (cstate = start; &cstate->head != &pstate->list;
-	     cstate = list_prev_entry(cstate, head)) {
+	list_for_each_entry_from_reverse(cstate, &pstate->list, head) {
 		if (nvkm_cstate_valid(clk, cstate, max_volt, clk->temp))
 			break;
 	}
-- 
cgit v1.1


From 6c46d01f25bcf74608d09645c27c35c3f3940ebe Mon Sep 17 00:00:00 2001
From: Ben Skeggs <bskeggs@redhat.com>
Date: Wed, 16 May 2018 12:07:32 +1000
Subject: drm/nouveau/gr/gf100-: insert some WFIs during gr init

Inserted wait-for-gr-idle in the places it seems that RM does it, seems
to prevent some random mmio timeouts on Quadro GV100.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
---
 drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c | 4 ++++
 drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c    | 2 ++
 2 files changed, 6 insertions(+)

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index f0f5a51..e813a3f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -1385,6 +1385,8 @@ gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 		gf100_gr_mmio(gr, gr->fuc_sw_ctx);
 	}
 
+	gf100_gr_wait_idle(gr);
+
 	idle_timeout = nvkm_mask(device, 0x404154, 0xffffffff, 0x00000000);
 
 	grctx->pagepool(info);
@@ -1396,6 +1398,8 @@ gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	gf100_grctx_generate_floorsweep(gr);
 
+	gf100_gr_wait_idle(gr);
+
 	if (grctx->r400088) grctx->r400088(gr, false);
 	if (gr->fuc_bundle)
 		gf100_gr_icmd(gr, gr->fuc_bundle);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 8dd4bd71..70d3d41 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -2123,6 +2123,8 @@ gf100_gr_init(struct gf100_gr *gr)
 	else
 		gf100_gr_mmio(gr, gr->func->mmio);
 
+	gf100_gr_wait_idle(gr);
+
 	if (gr->func->init_r405a14)
 		gr->func->init_r405a14(gr);
 
-- 
cgit v1.1


From 2f8a6da866eff746a9f8c7745790f3765baeb589 Mon Sep 17 00:00:00 2001
From: Emil Goode <emil.fsw@goode.io>
Date: Wed, 16 May 2018 12:22:04 +0200
Subject: gpu: host1x: Fix compiler errors by converting to dma_addr_t
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The compiler is complaining with the following errors:

drivers/gpu/host1x/cdma.c:94:48: error:
	passing argument 3 of ‘dma_alloc_wc’ from incompatible pointer type
	[-Werror=incompatible-pointer-types]

drivers/gpu/host1x/cdma.c:113:48: error:
	passing argument 3 of ‘dma_alloc_wc’ from incompatible pointer type
	[-Werror=incompatible-pointer-types]

The expected pointer type of the third argument to dma_alloc_wc() is
dma_addr_t but phys_addr_t is passed.

Change the phys member of struct push_buffer to be dma_addr_t so that we
pass the correct type to dma_alloc_wc().
Also check pb->mapped for non-NULL in the destroy function as that is the
right way of checking if dma_alloc_wc() was successful.

Signed-off-by: Emil Goode <emil.fsw@goode.io>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/host1x/cdma.c | 2 +-
 drivers/gpu/host1x/cdma.h | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/host1x/cdma.c b/drivers/gpu/host1x/cdma.c
index cf6caa9..69bb773 100644
--- a/drivers/gpu/host1x/cdma.c
+++ b/drivers/gpu/host1x/cdma.c
@@ -51,7 +51,7 @@ static void host1x_pushbuffer_destroy(struct push_buffer *pb)
 	struct host1x_cdma *cdma = pb_to_cdma(pb);
 	struct host1x *host1x = cdma_to_host1x(cdma);
 
-	if (!pb->phys)
+	if (!pb->mapped)
 		return;
 
 	if (host1x->domain) {
diff --git a/drivers/gpu/host1x/cdma.h b/drivers/gpu/host1x/cdma.h
index 286d493..446ee1a 100644
--- a/drivers/gpu/host1x/cdma.h
+++ b/drivers/gpu/host1x/cdma.h
@@ -44,7 +44,7 @@ struct host1x_job;
 struct push_buffer {
 	void *mapped;			/* mapped pushbuffer memory */
 	dma_addr_t dma;			/* device address of pushbuffer */
-	phys_addr_t phys;		/* physical address of pushbuffer */
+	dma_addr_t phys;		/* physical address of pushbuffer */
 	u32 fence;			/* index we've written */
 	u32 pos;			/* index to write to */
 	u32 size;
-- 
cgit v1.1


From d066b246d482f69553e58d52f746377ce3966b66 Mon Sep 17 00:00:00 2001
From: Russell King <rmk+kernel@armlinux.org.uk>
Date: Tue, 20 Feb 2018 10:22:22 +0100
Subject: drm/etnaviv: correct timeout calculation

The old way did clamp the jiffy conversion and thus caused the timeouts
to become negative after some time. Also it didn't work with userspace
which actually fills the upper 32bits of the 64bit timestamp value.

clock_gettime() is 32-bit on 32-bit architectures. Using 64-bit timespec
math, like we do in this commit, means that when a wrap occurs, the
specified timeout goes into the past and we can't request a timeout in
the future. As the Linux implementation of CLOCK_MONOTONIC is reasonable
and starts at 0, the first such timer wrap will occur after approx. 68
years of system uptime.

Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
---
 drivers/gpu/drm/etnaviv/etnaviv_drv.h | 25 +++++++++++++++++--------
 1 file changed, 17 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/etnaviv/etnaviv_drv.h b/drivers/gpu/drm/etnaviv/etnaviv_drv.h
index ddb17ee5..17a43da 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_drv.h
+++ b/drivers/gpu/drm/etnaviv/etnaviv_drv.h
@@ -26,6 +26,7 @@
 #include <linux/pm_runtime.h>
 #include <linux/slab.h>
 #include <linux/list.h>
+#include <linux/time64.h>
 #include <linux/types.h>
 #include <linux/sizes.h>
 
@@ -132,19 +133,27 @@ static inline bool fence_after_eq(u32 a, u32 b)
 	return (s32)(a - b) >= 0;
 }
 
+/*
+ * Etnaviv timeouts are specified wrt CLOCK_MONOTONIC, not jiffies.
+ * We need to calculate the timeout in terms of number of jiffies
+ * between the specified timeout and the current CLOCK_MONOTONIC time.
+ */
 static inline unsigned long etnaviv_timeout_to_jiffies(
 	const struct timespec *timeout)
 {
-	unsigned long timeout_jiffies = timespec_to_jiffies(timeout);
-	unsigned long start_jiffies = jiffies;
-	unsigned long remaining_jiffies;
+	struct timespec64 ts, to;
+
+	to = timespec_to_timespec64(*timeout);
+
+	ktime_get_ts64(&ts);
+
+	/* timeouts before "now" have already expired */
+	if (timespec64_compare(&to, &ts) <= 0)
+		return 0;
 
-	if (time_after(start_jiffies, timeout_jiffies))
-		remaining_jiffies = 0;
-	else
-		remaining_jiffies = timeout_jiffies - start_jiffies;
+	ts = timespec64_sub(to, ts);
 
-	return remaining_jiffies;
+	return timespec64_to_jiffies(&ts);
 }
 
 #endif /* __ETNAVIV_DRV_H__ */
-- 
cgit v1.1


From ccae45928fc43d78d6ba7d0c6965b142c922a446 Mon Sep 17 00:00:00 2001
From: Lucas Stach <l.stach@pengutronix.de>
Date: Fri, 9 Mar 2018 12:53:34 +0100
Subject: drm/etnaviv: remove cycling through MMU address space

This was useful on MMUv1 GPUs, which don't generate proper faults,
when the GPU write caches weren't fully understood and not properly
handled by the kernel driver. As this has been fixed for quite some
time, the cycling though the MMU address space needlessly spreads
out the MMU mappings.

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
---
 drivers/gpu/drm/etnaviv/etnaviv_mmu.c | 16 +---------------
 drivers/gpu/drm/etnaviv/etnaviv_mmu.h |  1 -
 2 files changed, 1 insertion(+), 16 deletions(-)

diff --git a/drivers/gpu/drm/etnaviv/etnaviv_mmu.c b/drivers/gpu/drm/etnaviv/etnaviv_mmu.c
index 49e0497..e8e8c4f 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_mmu.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_mmu.c
@@ -162,22 +162,10 @@ static int etnaviv_iommu_find_iova(struct etnaviv_iommu *mmu,
 		bool found;
 
 		ret = drm_mm_insert_node_in_range(&mmu->mm, node,
-						  size, 0, 0,
-						  mmu->last_iova, U64_MAX,
-						  mode);
+						  size, 0, 0, 0, U64_MAX, mode);
 		if (ret != -ENOSPC)
 			break;
 
-		/*
-		 * If we did not search from the start of the MMU region,
-		 * try again in case there are free slots.
-		 */
-		if (mmu->last_iova) {
-			mmu->last_iova = 0;
-			mmu->need_flush = true;
-			continue;
-		}
-
 		/* Try to retire some entries */
 		drm_mm_scan_init(&scan, &mmu->mm, size, 0, 0, mode);
 
@@ -274,7 +262,6 @@ int etnaviv_iommu_map_gem(struct etnaviv_iommu *mmu,
 	if (ret < 0)
 		goto unlock;
 
-	mmu->last_iova = node->start + etnaviv_obj->base.size;
 	mapping->iova = node->start;
 	ret = etnaviv_iommu_map(mmu, node->start, sgt, etnaviv_obj->base.size,
 				ETNAVIV_PROT_READ | ETNAVIV_PROT_WRITE);
@@ -381,7 +368,6 @@ int etnaviv_iommu_get_suballoc_va(struct etnaviv_gpu *gpu, dma_addr_t paddr,
 			mutex_unlock(&mmu->lock);
 			return ret;
 		}
-		mmu->last_iova = vram_node->start + size;
 		gpu->mmu->need_flush = true;
 		mutex_unlock(&mmu->lock);
 
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_mmu.h b/drivers/gpu/drm/etnaviv/etnaviv_mmu.h
index ab603f5..a339ec5 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_mmu.h
+++ b/drivers/gpu/drm/etnaviv/etnaviv_mmu.h
@@ -59,7 +59,6 @@ struct etnaviv_iommu {
 	struct mutex lock;
 	struct list_head mappings;
 	struct drm_mm mm;
-	u32 last_iova;
 	bool need_flush;
 };
 
-- 
cgit v1.1


From a98b1e7808a8a9faf7aa3a6318a1f3400f0ee628 Mon Sep 17 00:00:00 2001
From: Lucas Stach <l.stach@pengutronix.de>
Date: Thu, 19 Apr 2018 15:55:40 +0200
Subject: drm/etnaviv: remove register logging

I'm not aware of any case where tracing GPU register manipulation at the
kernel level would have been useful. It only adds more indirections and
adds to the code size.

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
---
 drivers/gpu/drm/etnaviv/Kconfig       |  8 ------
 drivers/gpu/drm/etnaviv/etnaviv_drv.c | 51 -----------------------------------
 drivers/gpu/drm/etnaviv/etnaviv_drv.h |  5 ----
 drivers/gpu/drm/etnaviv/etnaviv_gpu.c |  4 ++-
 drivers/gpu/drm/etnaviv/etnaviv_gpu.h |  4 +--
 5 files changed, 5 insertions(+), 67 deletions(-)

diff --git a/drivers/gpu/drm/etnaviv/Kconfig b/drivers/gpu/drm/etnaviv/Kconfig
index e5bfeca..041a77e 100644
--- a/drivers/gpu/drm/etnaviv/Kconfig
+++ b/drivers/gpu/drm/etnaviv/Kconfig
@@ -22,11 +22,3 @@ config DRM_ETNAVIV_THERMAL
 	help
 	  Compile in support for thermal throttling.
 	  Say Y unless you want to risk burning your SoC.
-
-config DRM_ETNAVIV_REGISTER_LOGGING
-	bool "enable ETNAVIV register logging"
-	depends on DRM_ETNAVIV
-	help
-	  Compile in support for logging register reads/writes in a format
-	  that can be parsed by envytools demsm tool.  If enabled, register
-	  logging can be switched on via etnaviv.reglog=y module param.
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_drv.c b/drivers/gpu/drm/etnaviv/etnaviv_drv.c
index ab50090..0aa543d 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_drv.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_drv.c
@@ -25,57 +25,6 @@
 #include "etnaviv_mmu.h"
 #include "etnaviv_perfmon.h"
 
-#ifdef CONFIG_DRM_ETNAVIV_REGISTER_LOGGING
-static bool reglog;
-MODULE_PARM_DESC(reglog, "Enable register read/write logging");
-module_param(reglog, bool, 0600);
-#else
-#define reglog 0
-#endif
-
-void __iomem *etnaviv_ioremap(struct platform_device *pdev, const char *name,
-		const char *dbgname)
-{
-	struct resource *res;
-	void __iomem *ptr;
-
-	if (name)
-		res = platform_get_resource_byname(pdev, IORESOURCE_MEM, name);
-	else
-		res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-
-	ptr = devm_ioremap_resource(&pdev->dev, res);
-	if (IS_ERR(ptr)) {
-		dev_err(&pdev->dev, "failed to ioremap %s: %ld\n", name,
-			PTR_ERR(ptr));
-		return ptr;
-	}
-
-	if (reglog)
-		dev_printk(KERN_DEBUG, &pdev->dev, "IO:region %s 0x%p %08zx\n",
-			   dbgname, ptr, (size_t)resource_size(res));
-
-	return ptr;
-}
-
-void etnaviv_writel(u32 data, void __iomem *addr)
-{
-	if (reglog)
-		printk(KERN_DEBUG "IO:W %p %08x\n", addr, data);
-
-	writel(data, addr);
-}
-
-u32 etnaviv_readl(const void __iomem *addr)
-{
-	u32 val = readl(addr);
-
-	if (reglog)
-		printk(KERN_DEBUG "IO:R %p %08x\n", addr, val);
-
-	return val;
-}
-
 /*
  * DRM operations:
  */
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_drv.h b/drivers/gpu/drm/etnaviv/etnaviv_drv.h
index 17a43da..763cf5b 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_drv.h
+++ b/drivers/gpu/drm/etnaviv/etnaviv_drv.h
@@ -102,11 +102,6 @@ void etnaviv_gem_describe_objects(struct etnaviv_drm_private *priv,
 	struct seq_file *m);
 #endif
 
-void __iomem *etnaviv_ioremap(struct platform_device *pdev, const char *name,
-		const char *dbgname);
-void etnaviv_writel(u32 data, void __iomem *addr);
-u32 etnaviv_readl(const void __iomem *addr);
-
 #define DBG(fmt, ...) DRM_DEBUG(fmt"\n", ##__VA_ARGS__)
 #define VERB(fmt, ...) if (0) DRM_DEBUG(fmt"\n", ##__VA_ARGS__)
 
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gpu.c b/drivers/gpu/drm/etnaviv/etnaviv_gpu.c
index 8a88799..08c5875 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_gpu.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_gpu.c
@@ -1735,6 +1735,7 @@ static int etnaviv_gpu_platform_probe(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;
 	struct etnaviv_gpu *gpu;
+	struct resource *res;
 	int err;
 
 	gpu = devm_kzalloc(dev, sizeof(*gpu), GFP_KERNEL);
@@ -1746,7 +1747,8 @@ static int etnaviv_gpu_platform_probe(struct platform_device *pdev)
 	mutex_init(&gpu->fence_idr_lock);
 
 	/* Map registers: */
-	gpu->mmio = etnaviv_ioremap(pdev, NULL, dev_name(gpu->dev));
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	gpu->mmio = devm_ioremap_resource(&pdev->dev, res);
 	if (IS_ERR(gpu->mmio))
 		return PTR_ERR(gpu->mmio);
 
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gpu.h b/drivers/gpu/drm/etnaviv/etnaviv_gpu.h
index 3c30055..6052093 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_gpu.h
+++ b/drivers/gpu/drm/etnaviv/etnaviv_gpu.h
@@ -161,12 +161,12 @@ struct etnaviv_gpu {
 
 static inline void gpu_write(struct etnaviv_gpu *gpu, u32 reg, u32 data)
 {
-	etnaviv_writel(data, gpu->mmio + reg);
+	writel(data, gpu->mmio + reg);
 }
 
 static inline u32 gpu_read(struct etnaviv_gpu *gpu, u32 reg)
 {
-	return etnaviv_readl(gpu->mmio + reg);
+	return readl(gpu->mmio + reg);
 }
 
 static inline bool fence_completed(struct etnaviv_gpu *gpu, u32 fence)
-- 
cgit v1.1


From 1af998b27c6c63d43c491783144ad0310d13a747 Mon Sep 17 00:00:00 2001
From: Lucas Stach <l.stach@pengutronix.de>
Date: Tue, 17 Apr 2018 12:00:46 +0200
Subject: drm/etnaviv: switch MMU page tables to writecombine memory

We are likely to write multiple page entries at once and already ensure
proper write buffer flushing before GPU submit, so this improves CPU
time usage in the submit path without any downsides.

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Philipp Zabel <p.zabel@pengutronix.de>
---
 drivers/gpu/drm/etnaviv/etnaviv_iommu.c    | 34 +++++++-------
 drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c | 74 +++++++++++++-----------------
 2 files changed, 49 insertions(+), 59 deletions(-)

diff --git a/drivers/gpu/drm/etnaviv/etnaviv_iommu.c b/drivers/gpu/drm/etnaviv/etnaviv_iommu.c
index 4b9b11c..4ada190 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_iommu.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_iommu.c
@@ -47,11 +47,10 @@ static int __etnaviv_iommu_init(struct etnaviv_iommuv1_domain *etnaviv_domain)
 	u32 *p;
 	int i;
 
-	etnaviv_domain->base.bad_page_cpu = dma_alloc_coherent(
-						etnaviv_domain->base.dev,
-						SZ_4K,
-						&etnaviv_domain->base.bad_page_dma,
-						GFP_KERNEL);
+	etnaviv_domain->base.bad_page_cpu =
+			dma_alloc_wc(etnaviv_domain->base.dev, SZ_4K,
+				     &etnaviv_domain->base.bad_page_dma,
+				     GFP_KERNEL);
 	if (!etnaviv_domain->base.bad_page_cpu)
 		return -ENOMEM;
 
@@ -59,14 +58,14 @@ static int __etnaviv_iommu_init(struct etnaviv_iommuv1_domain *etnaviv_domain)
 	for (i = 0; i < SZ_4K / 4; i++)
 		*p++ = 0xdead55aa;
 
-	etnaviv_domain->pgtable_cpu =
-			dma_alloc_coherent(etnaviv_domain->base.dev, PT_SIZE,
-					   &etnaviv_domain->pgtable_dma,
-					   GFP_KERNEL);
+	etnaviv_domain->pgtable_cpu = dma_alloc_wc(etnaviv_domain->base.dev,
+						   PT_SIZE,
+						   &etnaviv_domain->pgtable_dma,
+						   GFP_KERNEL);
 	if (!etnaviv_domain->pgtable_cpu) {
-		dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
-				  etnaviv_domain->base.bad_page_cpu,
-				  etnaviv_domain->base.bad_page_dma);
+		dma_free_wc(etnaviv_domain->base.dev, SZ_4K,
+			    etnaviv_domain->base.bad_page_cpu,
+			    etnaviv_domain->base.bad_page_dma);
 		return -ENOMEM;
 	}
 
@@ -81,13 +80,12 @@ static void etnaviv_iommuv1_domain_free(struct etnaviv_iommu_domain *domain)
 	struct etnaviv_iommuv1_domain *etnaviv_domain =
 			to_etnaviv_domain(domain);
 
-	dma_free_coherent(etnaviv_domain->base.dev, PT_SIZE,
-			  etnaviv_domain->pgtable_cpu,
-			  etnaviv_domain->pgtable_dma);
+	dma_free_wc(etnaviv_domain->base.dev, PT_SIZE,
+		    etnaviv_domain->pgtable_cpu, etnaviv_domain->pgtable_dma);
 
-	dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
-			  etnaviv_domain->base.bad_page_cpu,
-			  etnaviv_domain->base.bad_page_dma);
+	dma_free_wc(etnaviv_domain->base.dev, SZ_4K,
+		    etnaviv_domain->base.bad_page_cpu,
+		    etnaviv_domain->base.bad_page_dma);
 
 	kfree(etnaviv_domain);
 }
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c b/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c
index 9752dbd..47785d61 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c
@@ -104,11 +104,10 @@ static int etnaviv_iommuv2_init(struct etnaviv_iommuv2_domain *etnaviv_domain)
 	int ret, i, j;
 
 	/* allocate scratch page */
-	etnaviv_domain->base.bad_page_cpu = dma_alloc_coherent(
-						etnaviv_domain->base.dev,
-						SZ_4K,
-						&etnaviv_domain->base.bad_page_dma,
-						GFP_KERNEL);
+	etnaviv_domain->base.bad_page_cpu =
+			dma_alloc_wc(etnaviv_domain->base.dev, SZ_4K,
+				     &etnaviv_domain->base.bad_page_dma,
+				     GFP_KERNEL);
 	if (!etnaviv_domain->base.bad_page_cpu) {
 		ret = -ENOMEM;
 		goto fail_mem;
@@ -117,19 +116,17 @@ static int etnaviv_iommuv2_init(struct etnaviv_iommuv2_domain *etnaviv_domain)
 	for (i = 0; i < SZ_4K / 4; i++)
 		*p++ = 0xdead55aa;
 
-	etnaviv_domain->pta_cpu = dma_alloc_coherent(etnaviv_domain->base.dev,
-						     SZ_4K,
-						     &etnaviv_domain->pta_dma,
-						     GFP_KERNEL);
+	etnaviv_domain->pta_cpu = dma_alloc_wc(etnaviv_domain->base.dev,
+					       SZ_4K, &etnaviv_domain->pta_dma,
+					       GFP_KERNEL);
 	if (!etnaviv_domain->pta_cpu) {
 		ret = -ENOMEM;
 		goto fail_mem;
 	}
 
-	etnaviv_domain->mtlb_cpu = dma_alloc_coherent(etnaviv_domain->base.dev,
-						  SZ_4K,
-						  &etnaviv_domain->mtlb_dma,
-						  GFP_KERNEL);
+	etnaviv_domain->mtlb_cpu = dma_alloc_wc(etnaviv_domain->base.dev,
+						SZ_4K, &etnaviv_domain->mtlb_dma,
+						GFP_KERNEL);
 	if (!etnaviv_domain->mtlb_cpu) {
 		ret = -ENOMEM;
 		goto fail_mem;
@@ -138,10 +135,9 @@ static int etnaviv_iommuv2_init(struct etnaviv_iommuv2_domain *etnaviv_domain)
 	/* pre-populate STLB pages (may want to switch to on-demand later) */
 	for (i = 0; i < MMUv2_MAX_STLB_ENTRIES; i++) {
 		etnaviv_domain->stlb_cpu[i] =
-				dma_alloc_coherent(etnaviv_domain->base.dev,
-						   SZ_4K,
-						   &etnaviv_domain->stlb_dma[i],
-						   GFP_KERNEL);
+				dma_alloc_wc(etnaviv_domain->base.dev, SZ_4K,
+					     &etnaviv_domain->stlb_dma[i],
+					     GFP_KERNEL);
 		if (!etnaviv_domain->stlb_cpu[i]) {
 			ret = -ENOMEM;
 			goto fail_mem;
@@ -158,25 +154,23 @@ static int etnaviv_iommuv2_init(struct etnaviv_iommuv2_domain *etnaviv_domain)
 
 fail_mem:
 	if (etnaviv_domain->base.bad_page_cpu)
-		dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
-				  etnaviv_domain->base.bad_page_cpu,
-				  etnaviv_domain->base.bad_page_dma);
+		dma_free_wc(etnaviv_domain->base.dev, SZ_4K,
+			    etnaviv_domain->base.bad_page_cpu,
+			    etnaviv_domain->base.bad_page_dma);
 
 	if (etnaviv_domain->pta_cpu)
-		dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
-				  etnaviv_domain->pta_cpu,
-				  etnaviv_domain->pta_dma);
+		dma_free_wc(etnaviv_domain->base.dev, SZ_4K,
+			    etnaviv_domain->pta_cpu, etnaviv_domain->pta_dma);
 
 	if (etnaviv_domain->mtlb_cpu)
-		dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
-				  etnaviv_domain->mtlb_cpu,
-				  etnaviv_domain->mtlb_dma);
+		dma_free_wc(etnaviv_domain->base.dev, SZ_4K,
+			    etnaviv_domain->mtlb_cpu, etnaviv_domain->mtlb_dma);
 
 	for (i = 0; i < MMUv2_MAX_STLB_ENTRIES; i++) {
 		if (etnaviv_domain->stlb_cpu[i])
-			dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
-					  etnaviv_domain->stlb_cpu[i],
-					  etnaviv_domain->stlb_dma[i]);
+			dma_free_wc(etnaviv_domain->base.dev, SZ_4K,
+				    etnaviv_domain->stlb_cpu[i],
+				    etnaviv_domain->stlb_dma[i]);
 	}
 
 	return ret;
@@ -188,23 +182,21 @@ static void etnaviv_iommuv2_domain_free(struct etnaviv_iommu_domain *domain)
 			to_etnaviv_domain(domain);
 	int i;
 
-	dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
-			  etnaviv_domain->base.bad_page_cpu,
-			  etnaviv_domain->base.bad_page_dma);
+	dma_free_wc(etnaviv_domain->base.dev, SZ_4K,
+		    etnaviv_domain->base.bad_page_cpu,
+		    etnaviv_domain->base.bad_page_dma);
 
-	dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
-			  etnaviv_domain->pta_cpu,
-			  etnaviv_domain->pta_dma);
+	dma_free_wc(etnaviv_domain->base.dev, SZ_4K,
+		    etnaviv_domain->pta_cpu, etnaviv_domain->pta_dma);
 
-	dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
-			  etnaviv_domain->mtlb_cpu,
-			  etnaviv_domain->mtlb_dma);
+	dma_free_wc(etnaviv_domain->base.dev, SZ_4K,
+		    etnaviv_domain->mtlb_cpu, etnaviv_domain->mtlb_dma);
 
 	for (i = 0; i < MMUv2_MAX_STLB_ENTRIES; i++) {
 		if (etnaviv_domain->stlb_cpu[i])
-			dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
-					  etnaviv_domain->stlb_cpu[i],
-					  etnaviv_domain->stlb_dma[i]);
+			dma_free_wc(etnaviv_domain->base.dev, SZ_4K,
+				    etnaviv_domain->stlb_cpu[i],
+				    etnaviv_domain->stlb_dma[i]);
 	}
 
 	vfree(etnaviv_domain);
-- 
cgit v1.1


From a1fb6f204f956cc8385c31600354e2039978ebb4 Mon Sep 17 00:00:00 2001
From: Lucas Stach <l.stach@pengutronix.de>
Date: Tue, 17 Apr 2018 12:15:13 +0200
Subject: drm/etnaviv: mmuv2: allocate 2nd level page tables on demand

With etnaviv not being tied into the IOMMU framework anymore, the MMU
functions will only be called under sleeping locks. Thus we are able
to allocate the memory for the 2nd level page tables on demand without
having to deal with memory allocation in atomic context.

This speeds up driver intitialization on MMUv2 GPU cores, as we don't
need to preallocate all the page table memory and also reduces memory
consumption for most workloads, as most of them won't use the full
GPU virtual address space.

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Philipp Zabel <p.zabel@pengutronix.de>
---
 drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c | 61 ++++++++++++++++--------------
 1 file changed, 33 insertions(+), 28 deletions(-)

diff --git a/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c b/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c
index 47785d61..6336fdc 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c
@@ -47,8 +47,8 @@ struct etnaviv_iommuv2_domain {
 	u32 *mtlb_cpu;
 	dma_addr_t mtlb_dma;
 	/* S(lave) TLB aka second level pagetable */
-	u32 *stlb_cpu[1024];
-	dma_addr_t stlb_dma[1024];
+	u32 *stlb_cpu[MMUv2_MAX_STLB_ENTRIES];
+	dma_addr_t stlb_dma[MMUv2_MAX_STLB_ENTRIES];
 };
 
 static struct etnaviv_iommuv2_domain *
@@ -57,13 +57,36 @@ to_etnaviv_domain(struct etnaviv_iommu_domain *domain)
 	return container_of(domain, struct etnaviv_iommuv2_domain, base);
 }
 
+static int
+etnaviv_iommuv2_ensure_stlb(struct etnaviv_iommuv2_domain *etnaviv_domain,
+			    int stlb)
+{
+	if (etnaviv_domain->stlb_cpu[stlb])
+		return 0;
+
+	etnaviv_domain->stlb_cpu[stlb] =
+			dma_alloc_wc(etnaviv_domain->base.dev, SZ_4K,
+				     &etnaviv_domain->stlb_dma[stlb],
+				     GFP_KERNEL);
+
+	if (!etnaviv_domain->stlb_cpu[stlb])
+		return -ENOMEM;
+
+	memset32(etnaviv_domain->stlb_cpu[stlb], MMUv2_PTE_EXCEPTION,
+		 SZ_4K / sizeof(u32));
+
+	etnaviv_domain->mtlb_cpu[stlb] = etnaviv_domain->stlb_dma[stlb] |
+						      MMUv2_PTE_PRESENT;
+	return 0;
+}
+
 static int etnaviv_iommuv2_map(struct etnaviv_iommu_domain *domain,
 			       unsigned long iova, phys_addr_t paddr,
 			       size_t size, int prot)
 {
 	struct etnaviv_iommuv2_domain *etnaviv_domain =
 			to_etnaviv_domain(domain);
-	int mtlb_entry, stlb_entry;
+	int mtlb_entry, stlb_entry, ret;
 	u32 entry = (u32)paddr | MMUv2_PTE_PRESENT;
 
 	if (size != SZ_4K)
@@ -75,6 +98,10 @@ static int etnaviv_iommuv2_map(struct etnaviv_iommu_domain *domain,
 	mtlb_entry = (iova & MMUv2_MTLB_MASK) >> MMUv2_MTLB_SHIFT;
 	stlb_entry = (iova & MMUv2_STLB_MASK) >> MMUv2_STLB_SHIFT;
 
+	ret = etnaviv_iommuv2_ensure_stlb(etnaviv_domain, mtlb_entry);
+	if (ret)
+		return ret;
+
 	etnaviv_domain->stlb_cpu[mtlb_entry][stlb_entry] = entry;
 
 	return 0;
@@ -101,7 +128,7 @@ static size_t etnaviv_iommuv2_unmap(struct etnaviv_iommu_domain *domain,
 static int etnaviv_iommuv2_init(struct etnaviv_iommuv2_domain *etnaviv_domain)
 {
 	u32 *p;
-	int ret, i, j;
+	int ret, i;
 
 	/* allocate scratch page */
 	etnaviv_domain->base.bad_page_cpu =
@@ -132,23 +159,8 @@ static int etnaviv_iommuv2_init(struct etnaviv_iommuv2_domain *etnaviv_domain)
 		goto fail_mem;
 	}
 
-	/* pre-populate STLB pages (may want to switch to on-demand later) */
-	for (i = 0; i < MMUv2_MAX_STLB_ENTRIES; i++) {
-		etnaviv_domain->stlb_cpu[i] =
-				dma_alloc_wc(etnaviv_domain->base.dev, SZ_4K,
-					     &etnaviv_domain->stlb_dma[i],
-					     GFP_KERNEL);
-		if (!etnaviv_domain->stlb_cpu[i]) {
-			ret = -ENOMEM;
-			goto fail_mem;
-		}
-		p = etnaviv_domain->stlb_cpu[i];
-		for (j = 0; j < SZ_4K / 4; j++)
-			*p++ = MMUv2_PTE_EXCEPTION;
-
-		etnaviv_domain->mtlb_cpu[i] = etnaviv_domain->stlb_dma[i] |
-					      MMUv2_PTE_PRESENT;
-	}
+	memset32(etnaviv_domain->mtlb_cpu, MMUv2_PTE_EXCEPTION,
+		 MMUv2_MAX_STLB_ENTRIES);
 
 	return 0;
 
@@ -166,13 +178,6 @@ fail_mem:
 		dma_free_wc(etnaviv_domain->base.dev, SZ_4K,
 			    etnaviv_domain->mtlb_cpu, etnaviv_domain->mtlb_dma);
 
-	for (i = 0; i < MMUv2_MAX_STLB_ENTRIES; i++) {
-		if (etnaviv_domain->stlb_cpu[i])
-			dma_free_wc(etnaviv_domain->base.dev, SZ_4K,
-				    etnaviv_domain->stlb_cpu[i],
-				    etnaviv_domain->stlb_dma[i]);
-	}
-
 	return ret;
 }
 
-- 
cgit v1.1


From 931e97f3afd80bd9671d92f6934306a56012cae8 Mon Sep 17 00:00:00 2001
From: Lucas Stach <l.stach@pengutronix.de>
Date: Fri, 4 May 2018 11:58:45 +0200
Subject: drm/etnaviv: mmuv2: support 40 bit phys address

MMUv2 supports up to 40 bits of physical address by folding the upper
8 bits into bits [4:11] of the PTE.

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
---
 drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c b/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c
index 6336fdc..72bd010 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c
@@ -87,11 +87,14 @@ static int etnaviv_iommuv2_map(struct etnaviv_iommu_domain *domain,
 	struct etnaviv_iommuv2_domain *etnaviv_domain =
 			to_etnaviv_domain(domain);
 	int mtlb_entry, stlb_entry, ret;
-	u32 entry = (u32)paddr | MMUv2_PTE_PRESENT;
+	u32 entry = lower_32_bits(paddr) | MMUv2_PTE_PRESENT;
 
 	if (size != SZ_4K)
 		return -EINVAL;
 
+	if (IS_ENABLED(CONFIG_PHYS_ADDR_T_64BIT))
+		entry |= (upper_32_bits(paddr) & 0xff) << 4;
+
 	if (prot & ETNAVIV_PROT_WRITE)
 		entry |= MMUv2_PTE_WRITEABLE;
 
-- 
cgit v1.1


From f6ffbd4fc1a1caafe2ab840993b917fba5324598 Mon Sep 17 00:00:00 2001
From: Lucas Stach <l.stach@pengutronix.de>
Date: Tue, 8 May 2018 16:20:54 +0200
Subject: drm/etnaviv: replace license text with SPDX tags

This replaces the repetitive GPL-2.0 license text in code and header files
with the SPDX tags. Generated hardware headers aren't changed, as any changes
there need to be done in the upstream rnndb repository.

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
---
 drivers/gpu/drm/etnaviv/etnaviv_buffer.c     | 16 ++--------------
 drivers/gpu/drm/etnaviv/etnaviv_cmd_parser.c | 15 ++-------------
 drivers/gpu/drm/etnaviv/etnaviv_cmdbuf.c     | 15 ++-------------
 drivers/gpu/drm/etnaviv/etnaviv_cmdbuf.h     | 13 +------------
 drivers/gpu/drm/etnaviv/etnaviv_drv.c        | 15 ++-------------
 drivers/gpu/drm/etnaviv/etnaviv_drv.h        | 15 ++-------------
 drivers/gpu/drm/etnaviv/etnaviv_dump.c       | 15 ++-------------
 drivers/gpu/drm/etnaviv/etnaviv_dump.h       | 16 ++--------------
 drivers/gpu/drm/etnaviv/etnaviv_gem.c        | 15 ++-------------
 drivers/gpu/drm/etnaviv/etnaviv_gem.h        | 15 ++-------------
 drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c  | 16 ++--------------
 drivers/gpu/drm/etnaviv/etnaviv_gpu.c        | 15 ++-------------
 drivers/gpu/drm/etnaviv/etnaviv_gpu.h        | 15 ++-------------
 drivers/gpu/drm/etnaviv/etnaviv_hwdb.c       | 13 +------------
 drivers/gpu/drm/etnaviv/etnaviv_iommu.c      | 15 ++-------------
 drivers/gpu/drm/etnaviv/etnaviv_iommu.h      | 15 ++-------------
 drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c   | 15 ++-------------
 drivers/gpu/drm/etnaviv/etnaviv_mmu.c        | 15 ++-------------
 drivers/gpu/drm/etnaviv/etnaviv_mmu.h        | 15 ++-------------
 drivers/gpu/drm/etnaviv/etnaviv_perfmon.c    | 13 +------------
 drivers/gpu/drm/etnaviv/etnaviv_perfmon.h    | 13 +------------
 drivers/gpu/drm/etnaviv/etnaviv_sched.c      | 13 +------------
 drivers/gpu/drm/etnaviv/etnaviv_sched.h      | 13 +------------
 23 files changed, 40 insertions(+), 296 deletions(-)

diff --git a/drivers/gpu/drm/etnaviv/etnaviv_buffer.c b/drivers/gpu/drm/etnaviv/etnaviv_buffer.c
index bfc6d4a..7fea748 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_buffer.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_buffer.c
@@ -1,18 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
- * Copyright (C) 2014 Etnaviv Project
- * Author: Christian Gmeiner <christian.gmeiner@gmail.com>
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2014-2018 Etnaviv Project
  */
 
 #include "etnaviv_cmdbuf.h"
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_cmd_parser.c b/drivers/gpu/drm/etnaviv/etnaviv_cmd_parser.c
index 68e6d37..b106e8b 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_cmd_parser.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_cmd_parser.c
@@ -1,17 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
- * Copyright (C) 2015 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2015-2018 Etnaviv Project
  */
 
 #include <linux/kernel.h>
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_cmdbuf.c b/drivers/gpu/drm/etnaviv/etnaviv_cmdbuf.c
index 3746827..a3c44f1 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_cmdbuf.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_cmdbuf.c
@@ -1,17 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
- * Copyright (C) 2017 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2017-2018 Etnaviv Project
  */
 
 #include <drm/drm_mm.h>
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_cmdbuf.h b/drivers/gpu/drm/etnaviv/etnaviv_cmdbuf.h
index ddc3f7e..acb68c6 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_cmdbuf.h
+++ b/drivers/gpu/drm/etnaviv/etnaviv_cmdbuf.h
@@ -1,17 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 */
 /*
  * Copyright (C) 2017 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
 #ifndef __ETNAVIV_CMDBUF_H__
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_drv.c b/drivers/gpu/drm/etnaviv/etnaviv_drv.c
index 0aa543d..144fd8b 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_drv.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_drv.c
@@ -1,17 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
- * Copyright (C) 2015 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2015-2018 Etnaviv Project
  */
 
 #include <linux/component.h>
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_drv.h b/drivers/gpu/drm/etnaviv/etnaviv_drv.h
index 763cf5b..d36c7bb 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_drv.h
+++ b/drivers/gpu/drm/etnaviv/etnaviv_drv.h
@@ -1,17 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 */
 /*
- * Copyright (C) 2015 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2015-2018 Etnaviv Project
  */
 
 #ifndef __ETNAVIV_DRV_H__
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_dump.c b/drivers/gpu/drm/etnaviv/etnaviv_dump.c
index 48aef6c..9146e30 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_dump.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_dump.c
@@ -1,17 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
- * Copyright (C) 2015 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2015-2018 Etnaviv Project
  */
 
 #include <linux/devcoredump.h>
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_dump.h b/drivers/gpu/drm/etnaviv/etnaviv_dump.h
index 97f2f8d..2d916c2 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_dump.h
+++ b/drivers/gpu/drm/etnaviv/etnaviv_dump.h
@@ -1,20 +1,8 @@
+/* SPDX-License-Identifier: GPL-2.0 */
 /*
  * Copyright (C) 2015 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
- *
- * Etnaviv devcoredump file definitions
  */
+
 #ifndef ETNAVIV_DUMP_H
 #define ETNAVIV_DUMP_H
 
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gem.c b/drivers/gpu/drm/etnaviv/etnaviv_gem.c
index fcc969f..209ef12 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_gem.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_gem.c
@@ -1,17 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
- * Copyright (C) 2015 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2015-2018 Etnaviv Project
  */
 
 #include <linux/spinlock.h>
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gem.h b/drivers/gpu/drm/etnaviv/etnaviv_gem.h
index 93e696f..76079c22 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_gem.h
+++ b/drivers/gpu/drm/etnaviv/etnaviv_gem.h
@@ -1,17 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 */
 /*
- * Copyright (C) 2015 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2015-2018 Etnaviv Project
  */
 
 #ifndef __ETNAVIV_GEM_H__
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c b/drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c
index 5704305..0566171 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c
@@ -1,18 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
- * Copyright (C) 2013 Red Hat
- * Author: Rob Clark <robdclark@gmail.com>
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2014-2018 Etnaviv Project
  */
 
 #include <linux/dma-buf.h>
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gpu.c b/drivers/gpu/drm/etnaviv/etnaviv_gpu.c
index 08c5875..686f655 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_gpu.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_gpu.c
@@ -1,17 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
- * Copyright (C) 2015 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2015-2018 Etnaviv Project
  */
 
 #include <linux/component.h>
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gpu.h b/drivers/gpu/drm/etnaviv/etnaviv_gpu.h
index 6052093..dd430f0 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_gpu.h
+++ b/drivers/gpu/drm/etnaviv/etnaviv_gpu.h
@@ -1,17 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 */
 /*
- * Copyright (C) 2015 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2015-2018 Etnaviv Project
  */
 
 #ifndef __ETNAVIV_GPU_H__
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_hwdb.c b/drivers/gpu/drm/etnaviv/etnaviv_hwdb.c
index ea08bb3..39b463d 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_hwdb.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_hwdb.c
@@ -1,17 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * Copyright (C) 2018 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
 #include "etnaviv_gpu.h"
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_iommu.c b/drivers/gpu/drm/etnaviv/etnaviv_iommu.c
index 4ada190..b163bdb 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_iommu.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_iommu.c
@@ -1,17 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
- * Copyright (C) 2014 Christian Gmeiner <christian.gmeiner@gmail.com>
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2014-2018 Etnaviv Project
  */
 
 #include <linux/platform_device.h>
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_iommu.h b/drivers/gpu/drm/etnaviv/etnaviv_iommu.h
index 01d59bf..b279404 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_iommu.h
+++ b/drivers/gpu/drm/etnaviv/etnaviv_iommu.h
@@ -1,17 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 */
 /*
- * Copyright (C) 2014 Christian Gmeiner <christian.gmeiner@gmail.com>
-  *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2014-2018 Etnaviv Project
  */
 
 #ifndef __ETNAVIV_IOMMU_H__
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c b/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c
index 72bd010..71fbc1f 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c
@@ -1,17 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
- * Copyright (C) 2016 Etnaviv Project
-  *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2016-2018 Etnaviv Project
  */
 
 #include <linux/platform_device.h>
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_mmu.c b/drivers/gpu/drm/etnaviv/etnaviv_mmu.c
index e8e8c4f..8069f9f 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_mmu.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_mmu.c
@@ -1,17 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
- * Copyright (C) 2015 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2015-2018 Etnaviv Project
  */
 
 #include "common.xml.h"
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_mmu.h b/drivers/gpu/drm/etnaviv/etnaviv_mmu.h
index a339ec5..a0db17f 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_mmu.h
+++ b/drivers/gpu/drm/etnaviv/etnaviv_mmu.h
@@ -1,17 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 */
 /*
- * Copyright (C) 2015 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2015-2018 Etnaviv Project
  */
 
 #ifndef __ETNAVIV_MMU_H__
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_perfmon.c b/drivers/gpu/drm/etnaviv/etnaviv_perfmon.c
index 26dddfc..9980d81 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_perfmon.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_perfmon.c
@@ -1,18 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * Copyright (C) 2017 Etnaviv Project
  * Copyright (C) 2017 Zodiac Inflight Innovations
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
 #include "etnaviv_gpu.h"
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_perfmon.h b/drivers/gpu/drm/etnaviv/etnaviv_perfmon.h
index c1653c6..4a9d508 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_perfmon.h
+++ b/drivers/gpu/drm/etnaviv/etnaviv_perfmon.h
@@ -1,18 +1,7 @@
+/* SPDX-License-Identifier: GPL-2.0 */
 /*
  * Copyright (C) 2017 Etnaviv Project
  * Copyright (C) 2017 Zodiac Inflight Innovations
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
 #ifndef __ETNAVIV_PERFMON_H__
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_sched.c b/drivers/gpu/drm/etnaviv/etnaviv_sched.c
index 6cf0775..a74eb57 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_sched.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_sched.c
@@ -1,17 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * Copyright (C) 2017 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
 #include <linux/kthread.h>
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_sched.h b/drivers/gpu/drm/etnaviv/etnaviv_sched.h
index 097635f..c0a6796 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_sched.h
+++ b/drivers/gpu/drm/etnaviv/etnaviv_sched.h
@@ -1,17 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 */
 /*
  * Copyright (C) 2017 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
 #ifndef __ETNAVIV_SCHED_H__
-- 
cgit v1.1


From fcdfa432a5b0569e8c5399effa950c71940b5889 Mon Sep 17 00:00:00 2001
From: Oded Gabbay <oded.gabbay@gmail.com>
Date: Fri, 18 May 2018 22:18:16 +0300
Subject: drm/amdgpu: conditionally compile amdgpu's amdkfd files

In case CONFIG_HSA_AMD is not chosen, there is no need to compile amdkfd
files that reside inside amdgpu dirver. In addition, because amdkfd
depends on x86_64 architecture and amdgpu is not, compiling amdkfd files
under i386 architecture can cause compiler errors and warnings.

This patch modifies amdgpu's makefile to build amdkfd files only if
CONFIG_HSA_AMD is chosen. The only file to be compiled unconditionally
is amdgpu_amdkfd.c

There are stub functions that are compiled only if amdkfd is not
compiled. In that case, calls from amdgpu driver proper will go to those
functions instead of the real functions.

v2: instead of using function pointers, use stub functions

v3: initialize kgd2kfd to NULL in case amdkfd is not compiled

Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdgpu/Makefile        | 13 +++++++--
 drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c | 47 ++++++++++++++++++++++++++++++
 drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h | 12 ++++----
 3 files changed, 63 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/Makefile b/drivers/gpu/drm/amd/amdgpu/Makefile
index a51c5a9..bfd332c 100644
--- a/drivers/gpu/drm/amd/amdgpu/Makefile
+++ b/drivers/gpu/drm/amd/amdgpu/Makefile
@@ -56,8 +56,7 @@ amdgpu-y += amdgpu_device.o amdgpu_kms.o \
 
 # add asic specific block
 amdgpu-$(CONFIG_DRM_AMDGPU_CIK)+= cik.o cik_ih.o kv_smc.o kv_dpm.o \
-	ci_smc.o ci_dpm.o dce_v8_0.o gfx_v7_0.o cik_sdma.o uvd_v4_2.o vce_v2_0.o \
-	amdgpu_amdkfd_gfx_v7.o
+	ci_smc.o ci_dpm.o dce_v8_0.o gfx_v7_0.o cik_sdma.o uvd_v4_2.o vce_v2_0.o
 
 amdgpu-$(CONFIG_DRM_AMDGPU_SI)+= si.o gmc_v6_0.o gfx_v6_0.o si_ih.o si_dma.o dce_v6_0.o si_dpm.o si_smc.o
 
@@ -132,13 +131,21 @@ amdgpu-y += \
 	vcn_v1_0.o
 
 # add amdkfd interfaces
+amdgpu-y += amdgpu_amdkfd.o
+
+ifneq ($(CONFIG_HSA_AMD),)
 amdgpu-y += \
-	 amdgpu_amdkfd.o \
 	 amdgpu_amdkfd_fence.o \
 	 amdgpu_amdkfd_gpuvm.o \
 	 amdgpu_amdkfd_gfx_v8.o \
 	 amdgpu_amdkfd_gfx_v9.o
 
+ifneq ($(CONFIG_DRM_AMDGPU_CIK),)
+amdgpu-y += amdgpu_amdkfd_gfx_v7.o
+endif
+
+endif
+
 # add cgs
 amdgpu-y += amdgpu_cgs.o
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
index bd36ee9..95fcbd8 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
@@ -50,15 +50,21 @@ int amdgpu_amdkfd_init(void)
 		kgd2kfd = NULL;
 	}
 
+
 #elif defined(CONFIG_HSA_AMD)
+
 	ret = kgd2kfd_init(KFD_INTERFACE_VERSION, &kgd2kfd);
 	if (ret)
 		kgd2kfd = NULL;
 
 #else
+	kgd2kfd = NULL;
 	ret = -ENOENT;
 #endif
+
+#if defined(CONFIG_HSA_AMD_MODULE) || defined(CONFIG_HSA_AMD)
 	amdgpu_amdkfd_gpuvm_init_mem_limits();
+#endif
 
 	return ret;
 }
@@ -464,3 +470,44 @@ bool amdgpu_amdkfd_is_kfd_vmid(struct amdgpu_device *adev, u32 vmid)
 
 	return false;
 }
+
+#if !defined(CONFIG_HSA_AMD_MODULE) && !defined(CONFIG_HSA_AMD)
+bool amdkfd_fence_check_mm(struct dma_fence *f, struct mm_struct *mm)
+{
+	return false;
+}
+
+void amdgpu_amdkfd_unreserve_system_memory_limit(struct amdgpu_bo *bo)
+{
+}
+
+void amdgpu_amdkfd_gpuvm_destroy_cb(struct amdgpu_device *adev,
+					struct amdgpu_vm *vm)
+{
+}
+
+struct amdgpu_amdkfd_fence *to_amdgpu_amdkfd_fence(struct dma_fence *f)
+{
+	return NULL;
+}
+
+int amdgpu_amdkfd_evict_userptr(struct kgd_mem *mem, struct mm_struct *mm)
+{
+	return 0;
+}
+
+struct kfd2kgd_calls *amdgpu_amdkfd_gfx_7_get_functions(void)
+{
+	return NULL;
+}
+
+struct kfd2kgd_calls *amdgpu_amdkfd_gfx_8_0_get_functions(void)
+{
+	return NULL;
+}
+
+struct kfd2kgd_calls *amdgpu_amdkfd_gfx_9_0_get_functions(void)
+{
+	return NULL;
+}
+#endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
index 12367a9..a8418a3 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
@@ -156,14 +156,14 @@ uint64_t amdgpu_amdkfd_get_vram_usage(struct kgd_dev *kgd);
 
 /* GPUVM API */
 int amdgpu_amdkfd_gpuvm_create_process_vm(struct kgd_dev *kgd, void **vm,
-					  void **process_info,
-					  struct dma_fence **ef);
+					void **process_info,
+					struct dma_fence **ef);
 int amdgpu_amdkfd_gpuvm_acquire_process_vm(struct kgd_dev *kgd,
-					   struct file *filp,
-					   void **vm, void **process_info,
-					   struct dma_fence **ef);
+					struct file *filp,
+					void **vm, void **process_info,
+					struct dma_fence **ef);
 void amdgpu_amdkfd_gpuvm_destroy_cb(struct amdgpu_device *adev,
-				    struct amdgpu_vm *vm);
+				struct amdgpu_vm *vm);
 void amdgpu_amdkfd_gpuvm_destroy_process_vm(struct kgd_dev *kgd, void *vm);
 uint32_t amdgpu_amdkfd_gpuvm_get_process_page_dir(void *vm);
 int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu(
-- 
cgit v1.1


From 24c94e166dfe89839129b8e0fae208b6af60d6f1 Mon Sep 17 00:00:00 2001
From: Thierry Reding <treding@nvidia.com>
Date: Sat, 5 May 2018 08:45:47 +0200
Subject: gpu: host1x: Remove wait check support

The job submission userspace ABI doesn't support this and there are no
plans to implement it, so all of this code is dead and can be removed.

Reviewed-by: Dmitry Osipenko <digetx@gmail.com>
Tested-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/drm.c        |  62 +------------------
 drivers/gpu/host1x/dev.h           |   8 ---
 drivers/gpu/host1x/hw/channel_hw.c |   3 +-
 drivers/gpu/host1x/hw/syncpt_hw.c  |  11 ----
 drivers/gpu/host1x/job.c           | 124 +------------------------------------
 drivers/gpu/host1x/syncpt.c        |   6 --
 drivers/gpu/host1x/syncpt.h        |   3 -
 include/linux/host1x.h             |  15 +----
 include/trace/events/host1x.h      |  16 +++--
 9 files changed, 14 insertions(+), 234 deletions(-)

diff --git a/drivers/gpu/drm/tegra/drm.c b/drivers/gpu/drm/tegra/drm.c
index 3cdef65..204b10e 100644
--- a/drivers/gpu/drm/tegra/drm.c
+++ b/drivers/gpu/drm/tegra/drm.c
@@ -321,46 +321,14 @@ static int host1x_reloc_copy_from_user(struct host1x_reloc *dest,
 	return 0;
 }
 
-static int host1x_waitchk_copy_from_user(struct host1x_waitchk *dest,
-					 struct drm_tegra_waitchk __user *src,
-					 struct drm_file *file)
-{
-	u32 cmdbuf;
-	int err;
-
-	err = get_user(cmdbuf, &src->handle);
-	if (err < 0)
-		return err;
-
-	err = get_user(dest->offset, &src->offset);
-	if (err < 0)
-		return err;
-
-	err = get_user(dest->syncpt_id, &src->syncpt);
-	if (err < 0)
-		return err;
-
-	err = get_user(dest->thresh, &src->thresh);
-	if (err < 0)
-		return err;
-
-	dest->bo = host1x_bo_lookup(file, cmdbuf);
-	if (!dest->bo)
-		return -ENOENT;
-
-	return 0;
-}
-
 int tegra_drm_submit(struct tegra_drm_context *context,
 		     struct drm_tegra_submit *args, struct drm_device *drm,
 		     struct drm_file *file)
 {
 	unsigned int num_cmdbufs = args->num_cmdbufs;
 	unsigned int num_relocs = args->num_relocs;
-	unsigned int num_waitchks = args->num_waitchks;
 	struct drm_tegra_cmdbuf __user *user_cmdbufs;
 	struct drm_tegra_reloc __user *user_relocs;
-	struct drm_tegra_waitchk __user *user_waitchks;
 	struct drm_tegra_syncpt __user *user_syncpt;
 	struct drm_tegra_syncpt syncpt;
 	struct host1x *host1x = dev_get_drvdata(drm->dev->parent);
@@ -372,7 +340,6 @@ int tegra_drm_submit(struct tegra_drm_context *context,
 
 	user_cmdbufs = u64_to_user_ptr(args->cmdbufs);
 	user_relocs = u64_to_user_ptr(args->relocs);
-	user_waitchks = u64_to_user_ptr(args->waitchks);
 	user_syncpt = u64_to_user_ptr(args->syncpts);
 
 	/* We don't yet support other than one syncpt_incr struct per submit */
@@ -384,12 +351,11 @@ int tegra_drm_submit(struct tegra_drm_context *context,
 		return -EINVAL;
 
 	job = host1x_job_alloc(context->channel, args->num_cmdbufs,
-			       args->num_relocs, args->num_waitchks);
+			       args->num_relocs);
 	if (!job)
 		return -ENOMEM;
 
 	job->num_relocs = args->num_relocs;
-	job->num_waitchk = args->num_waitchks;
 	job->client = (u32)args->context;
 	job->class = context->client->base.class;
 	job->serialize = true;
@@ -398,7 +364,7 @@ int tegra_drm_submit(struct tegra_drm_context *context,
 	 * Track referenced BOs so that they can be unreferenced after the
 	 * submission is complete.
 	 */
-	num_refs = num_cmdbufs + num_relocs * 2 + num_waitchks;
+	num_refs = num_cmdbufs + num_relocs * 2;
 
 	refs = kmalloc_array(num_refs, sizeof(*refs), GFP_KERNEL);
 	if (!refs) {
@@ -489,30 +455,6 @@ int tegra_drm_submit(struct tegra_drm_context *context,
 		}
 	}
 
-	/* copy and resolve waitchks from submit */
-	while (num_waitchks--) {
-		struct host1x_waitchk *wait = &job->waitchk[num_waitchks];
-		struct tegra_bo *obj;
-
-		err = host1x_waitchk_copy_from_user(
-			wait, &user_waitchks[num_waitchks], file);
-		if (err < 0)
-			goto fail;
-
-		obj = host1x_to_tegra_bo(wait->bo);
-		refs[num_refs++] = &obj->gem;
-
-		/*
-		 * The unaligned offset will cause an unaligned write during
-		 * of the waitchks patching, corrupting the commands stream.
-		 */
-		if (wait->offset & 3 ||
-		    wait->offset >= obj->gem.size) {
-			err = -EINVAL;
-			goto fail;
-		}
-	}
-
 	if (copy_from_user(&syncpt, user_syncpt, sizeof(syncpt))) {
 		err = -EFAULT;
 		goto fail;
diff --git a/drivers/gpu/host1x/dev.h b/drivers/gpu/host1x/dev.h
index 43e9fab..36f44ff 100644
--- a/drivers/gpu/host1x/dev.h
+++ b/drivers/gpu/host1x/dev.h
@@ -78,7 +78,6 @@ struct host1x_syncpt_ops {
 	void (*load_wait_base)(struct host1x_syncpt *syncpt);
 	u32 (*load)(struct host1x_syncpt *syncpt);
 	int (*cpu_incr)(struct host1x_syncpt *syncpt);
-	int (*patch_wait)(struct host1x_syncpt *syncpt, void *patch_addr);
 	void (*assign_to_channel)(struct host1x_syncpt *syncpt,
 	                          struct host1x_channel *channel);
 	void (*enable_protection)(struct host1x *host);
@@ -183,13 +182,6 @@ static inline int host1x_hw_syncpt_cpu_incr(struct host1x *host,
 	return host->syncpt_op->cpu_incr(sp);
 }
 
-static inline int host1x_hw_syncpt_patch_wait(struct host1x *host,
-					      struct host1x_syncpt *sp,
-					      void *patch_addr)
-{
-	return host->syncpt_op->patch_wait(sp, patch_addr);
-}
-
 static inline void host1x_hw_syncpt_assign_to_channel(
 	struct host1x *host, struct host1x_syncpt *sp,
 	struct host1x_channel *ch)
diff --git a/drivers/gpu/host1x/hw/channel_hw.c b/drivers/gpu/host1x/hw/channel_hw.c
index 9af7587..4c95550 100644
--- a/drivers/gpu/host1x/hw/channel_hw.c
+++ b/drivers/gpu/host1x/hw/channel_hw.c
@@ -104,8 +104,7 @@ static int channel_submit(struct host1x_job *job)
 	sp = host->syncpt + job->syncpt_id;
 	trace_host1x_channel_submit(dev_name(ch->dev),
 				    job->num_gathers, job->num_relocs,
-				    job->num_waitchk, job->syncpt_id,
-				    job->syncpt_incrs);
+				    job->syncpt_id, job->syncpt_incrs);
 
 	/* before error checks, return current max */
 	prev_max = job->syncpt_end = host1x_syncpt_read_max(sp);
diff --git a/drivers/gpu/host1x/hw/syncpt_hw.c b/drivers/gpu/host1x/hw/syncpt_hw.c
index 7dfd47d..a23bb33 100644
--- a/drivers/gpu/host1x/hw/syncpt_hw.c
+++ b/drivers/gpu/host1x/hw/syncpt_hw.c
@@ -96,16 +96,6 @@ static int syncpt_cpu_incr(struct host1x_syncpt *sp)
 	return 0;
 }
 
-/* remove a wait pointed to by patch_addr */
-static int syncpt_patch_wait(struct host1x_syncpt *sp, void *patch_addr)
-{
-	u32 override = host1x_class_host_wait_syncpt(HOST1X_SYNCPT_RESERVED, 0);
-
-	*((u32 *)patch_addr) = override;
-
-	return 0;
-}
-
 /**
  * syncpt_assign_to_channel() - Assign syncpoint to channel
  * @sp: syncpoint
@@ -156,7 +146,6 @@ static const struct host1x_syncpt_ops host1x_syncpt_ops = {
 	.load_wait_base = syncpt_read_wait_base,
 	.load = syncpt_load,
 	.cpu_incr = syncpt_cpu_incr,
-	.patch_wait = syncpt_patch_wait,
 	.assign_to_channel = syncpt_assign_to_channel,
 	.enable_protection = syncpt_enable_protection,
 };
diff --git a/drivers/gpu/host1x/job.c b/drivers/gpu/host1x/job.c
index db509ab..3cbfc6e 100644
--- a/drivers/gpu/host1x/job.c
+++ b/drivers/gpu/host1x/job.c
@@ -34,8 +34,7 @@
 #define HOST1X_WAIT_SYNCPT_OFFSET 0x8
 
 struct host1x_job *host1x_job_alloc(struct host1x_channel *ch,
-				    u32 num_cmdbufs, u32 num_relocs,
-				    u32 num_waitchks)
+				    u32 num_cmdbufs, u32 num_relocs)
 {
 	struct host1x_job *job = NULL;
 	unsigned int num_unpins = num_cmdbufs + num_relocs;
@@ -46,7 +45,6 @@ struct host1x_job *host1x_job_alloc(struct host1x_channel *ch,
 	total = sizeof(struct host1x_job) +
 		(u64)num_relocs * sizeof(struct host1x_reloc) +
 		(u64)num_unpins * sizeof(struct host1x_job_unpin_data) +
-		(u64)num_waitchks * sizeof(struct host1x_waitchk) +
 		(u64)num_cmdbufs * sizeof(struct host1x_job_gather) +
 		(u64)num_unpins * sizeof(dma_addr_t) +
 		(u64)num_unpins * sizeof(u32 *);
@@ -66,8 +64,6 @@ struct host1x_job *host1x_job_alloc(struct host1x_channel *ch,
 	mem += num_relocs * sizeof(struct host1x_reloc);
 	job->unpins = num_unpins ? mem : NULL;
 	mem += num_unpins * sizeof(struct host1x_job_unpin_data);
-	job->waitchk = num_waitchks ? mem : NULL;
-	mem += num_waitchks * sizeof(struct host1x_waitchk);
 	job->gathers = num_cmdbufs ? mem : NULL;
 	mem += num_cmdbufs * sizeof(struct host1x_job_gather);
 	job->addr_phys = num_unpins ? mem : NULL;
@@ -111,73 +107,6 @@ void host1x_job_add_gather(struct host1x_job *job, struct host1x_bo *bo,
 }
 EXPORT_SYMBOL(host1x_job_add_gather);
 
-/*
- * NULL an already satisfied WAIT_SYNCPT host method, by patching its
- * args in the command stream. The method data is changed to reference
- * a reserved (never given out or incr) HOST1X_SYNCPT_RESERVED syncpt
- * with a matching threshold value of 0, so is guaranteed to be popped
- * by the host HW.
- */
-static void host1x_syncpt_patch_offset(struct host1x_syncpt *sp,
-				       struct host1x_bo *h, u32 offset)
-{
-	void *patch_addr = NULL;
-
-	/* patch the wait */
-	patch_addr = host1x_bo_kmap(h, offset >> PAGE_SHIFT);
-	if (patch_addr) {
-		host1x_syncpt_patch_wait(sp,
-					 patch_addr + (offset & ~PAGE_MASK));
-		host1x_bo_kunmap(h, offset >> PAGE_SHIFT, patch_addr);
-	} else
-		pr_err("Could not map cmdbuf for wait check\n");
-}
-
-/*
- * Check driver supplied waitchk structs for syncpt thresholds
- * that have already been satisfied and NULL the comparison (to
- * avoid a wrap condition in the HW).
- */
-static int do_waitchks(struct host1x_job *job, struct host1x *host,
-		       struct host1x_job_gather *g)
-{
-	struct host1x_bo *patch = g->bo;
-	int i;
-
-	/* compare syncpt vs wait threshold */
-	for (i = 0; i < job->num_waitchk; i++) {
-		struct host1x_waitchk *wait = &job->waitchk[i];
-		struct host1x_syncpt *sp =
-			host1x_syncpt_get(host, wait->syncpt_id);
-
-		/* validate syncpt id */
-		if (wait->syncpt_id > host1x_syncpt_nb_pts(host))
-			continue;
-
-		/* skip all other gathers */
-		if (patch != wait->bo)
-			continue;
-
-		trace_host1x_syncpt_wait_check(wait->bo, wait->offset,
-					       wait->syncpt_id, wait->thresh,
-					       host1x_syncpt_read_min(sp));
-
-		if (host1x_syncpt_is_expired(sp, wait->thresh)) {
-			dev_dbg(host->dev,
-				"drop WAIT id %u (%s) thresh 0x%x, min 0x%x\n",
-				wait->syncpt_id, sp->name, wait->thresh,
-				host1x_syncpt_read_min(sp));
-
-			host1x_syncpt_patch_offset(sp, patch,
-						   g->offset + wait->offset);
-		}
-
-		wait->bo = NULL;
-	}
-
-	return 0;
-}
-
 static unsigned int pin_job(struct host1x *host, struct host1x_job *job)
 {
 	unsigned int i;
@@ -331,17 +260,6 @@ static bool check_reloc(struct host1x_reloc *reloc, struct host1x_bo *cmdbuf,
 	return true;
 }
 
-static bool check_wait(struct host1x_waitchk *wait, struct host1x_bo *cmdbuf,
-		       unsigned int offset)
-{
-	offset *= sizeof(u32);
-
-	if (wait->bo != cmdbuf || wait->offset != offset)
-		return false;
-
-	return true;
-}
-
 struct host1x_firewall {
 	struct host1x_job *job;
 	struct device *dev;
@@ -349,9 +267,6 @@ struct host1x_firewall {
 	unsigned int num_relocs;
 	struct host1x_reloc *reloc;
 
-	unsigned int num_waitchks;
-	struct host1x_waitchk *waitchk;
-
 	struct host1x_bo *cmdbuf;
 	unsigned int offset;
 
@@ -378,20 +293,6 @@ static int check_register(struct host1x_firewall *fw, unsigned long offset)
 		fw->reloc++;
 	}
 
-	if (offset == HOST1X_WAIT_SYNCPT_OFFSET) {
-		if (fw->class != HOST1X_CLASS_HOST1X)
-			return -EINVAL;
-
-		if (!fw->num_waitchks)
-			return -EINVAL;
-
-		if (!check_wait(fw->waitchk, fw->cmdbuf, fw->offset))
-			return -EINVAL;
-
-		fw->num_waitchks--;
-		fw->waitchk++;
-	}
-
 	return 0;
 }
 
@@ -556,8 +457,6 @@ static inline int copy_gathers(struct host1x_job *job, struct device *dev)
 	fw.dev = dev;
 	fw.reloc = job->relocarray;
 	fw.num_relocs = job->num_relocs;
-	fw.waitchk = job->waitchk;
-	fw.num_waitchks = job->num_waitchk;
 	fw.class = job->class;
 
 	for (i = 0; i < job->num_gathers; i++) {
@@ -604,8 +503,8 @@ static inline int copy_gathers(struct host1x_job *job, struct device *dev)
 		offset += g->words * sizeof(u32);
 	}
 
-	/* No relocs and waitchks should remain at this point */
-	if (fw.num_relocs || fw.num_waitchks)
+	/* No relocs should remain at this point */
+	if (fw.num_relocs)
 		return -EINVAL;
 
 	return 0;
@@ -616,19 +515,6 @@ int host1x_job_pin(struct host1x_job *job, struct device *dev)
 	int err;
 	unsigned int i, j;
 	struct host1x *host = dev_get_drvdata(dev->parent);
-	DECLARE_BITMAP(waitchk_mask, host1x_syncpt_nb_pts(host));
-
-	bitmap_zero(waitchk_mask, host1x_syncpt_nb_pts(host));
-	for (i = 0; i < job->num_waitchk; i++) {
-		u32 syncpt_id = job->waitchk[i].syncpt_id;
-
-		if (syncpt_id < host1x_syncpt_nb_pts(host))
-			set_bit(syncpt_id, waitchk_mask);
-	}
-
-	/* get current syncpt values for waitchk */
-	for_each_set_bit(i, waitchk_mask, host1x_syncpt_nb_pts(host))
-		host1x_syncpt_load(host->syncpt + i);
 
 	/* pin memory */
 	err = pin_job(host, job);
@@ -663,10 +549,6 @@ int host1x_job_pin(struct host1x_job *job, struct device *dev)
 		err = do_relocs(job, g);
 		if (err)
 			break;
-
-		err = do_waitchks(job, host, g);
-		if (err)
-			break;
 	}
 
 out:
diff --git a/drivers/gpu/host1x/syncpt.c b/drivers/gpu/host1x/syncpt.c
index a2a952a..a108669 100644
--- a/drivers/gpu/host1x/syncpt.c
+++ b/drivers/gpu/host1x/syncpt.c
@@ -373,12 +373,6 @@ bool host1x_syncpt_is_expired(struct host1x_syncpt *sp, u32 thresh)
 		return (s32)(current_val - thresh) >= 0;
 }
 
-/* remove a wait pointed to by patch_addr */
-int host1x_syncpt_patch_wait(struct host1x_syncpt *sp, void *patch_addr)
-{
-	return host1x_hw_syncpt_patch_wait(sp->host, sp, patch_addr);
-}
-
 int host1x_syncpt_init(struct host1x *host)
 {
 	struct host1x_syncpt_base *bases;
diff --git a/drivers/gpu/host1x/syncpt.h b/drivers/gpu/host1x/syncpt.h
index 9d88d37..d98e223 100644
--- a/drivers/gpu/host1x/syncpt.h
+++ b/drivers/gpu/host1x/syncpt.h
@@ -124,7 +124,4 @@ static inline int host1x_syncpt_is_valid(struct host1x_syncpt *sp)
 	return sp->id < host1x_syncpt_nb_pts(sp->host);
 }
 
-/* Patch a wait by replacing it with a wait for syncpt 0 value 0 */
-int host1x_syncpt_patch_wait(struct host1x_syncpt *sp, void *patch_addr);
-
 #endif
diff --git a/include/linux/host1x.h b/include/linux/host1x.h
index ddf7f9c..f66bece 100644
--- a/include/linux/host1x.h
+++ b/include/linux/host1x.h
@@ -192,13 +192,6 @@ struct host1x_reloc {
 	unsigned long shift;
 };
 
-struct host1x_waitchk {
-	struct host1x_bo *bo;
-	u32 offset;
-	u32 syncpt_id;
-	u32 thresh;
-};
-
 struct host1x_job {
 	/* When refcount goes to zero, job can be freed */
 	struct kref ref;
@@ -215,11 +208,6 @@ struct host1x_job {
 	struct host1x_job_gather *gathers;
 	unsigned int num_gathers;
 
-	/* Wait checks to be processed at submit time */
-	struct host1x_waitchk *waitchk;
-	unsigned int num_waitchk;
-	u32 waitchk_mask;
-
 	/* Array of handles to be pinned & unpinned */
 	struct host1x_reloc *relocarray;
 	unsigned int num_relocs;
@@ -261,8 +249,7 @@ struct host1x_job {
 };
 
 struct host1x_job *host1x_job_alloc(struct host1x_channel *ch,
-				    u32 num_cmdbufs, u32 num_relocs,
-				    u32 num_waitchks);
+				    u32 num_cmdbufs, u32 num_relocs);
 void host1x_job_add_gather(struct host1x_job *job, struct host1x_bo *mem_id,
 			   u32 words, u32 offset);
 struct host1x_job *host1x_job_get(struct host1x_job *job);
diff --git a/include/trace/events/host1x.h b/include/trace/events/host1x.h
index 6311636..a37ef73 100644
--- a/include/trace/events/host1x.h
+++ b/include/trace/events/host1x.h
@@ -115,16 +115,15 @@ TRACE_EVENT(host1x_cdma_push_gather,
 );
 
 TRACE_EVENT(host1x_channel_submit,
-	TP_PROTO(const char *name, u32 cmdbufs, u32 relocs, u32 waitchks,
-			u32 syncpt_id, u32 syncpt_incrs),
+	TP_PROTO(const char *name, u32 cmdbufs, u32 relocs, u32 syncpt_id,
+		 u32 syncpt_incrs),
 
-	TP_ARGS(name, cmdbufs, relocs, waitchks, syncpt_id, syncpt_incrs),
+	TP_ARGS(name, cmdbufs, relocs, syncpt_id, syncpt_incrs),
 
 	TP_STRUCT__entry(
 		__field(const char *, name)
 		__field(u32, cmdbufs)
 		__field(u32, relocs)
-		__field(u32, waitchks)
 		__field(u32, syncpt_id)
 		__field(u32, syncpt_incrs)
 	),
@@ -133,15 +132,14 @@ TRACE_EVENT(host1x_channel_submit,
 		__entry->name = name;
 		__entry->cmdbufs = cmdbufs;
 		__entry->relocs = relocs;
-		__entry->waitchks = waitchks;
 		__entry->syncpt_id = syncpt_id;
 		__entry->syncpt_incrs = syncpt_incrs;
 	),
 
-	TP_printk("name=%s, cmdbufs=%u, relocs=%u, waitchks=%d,"
-		"syncpt_id=%u, syncpt_incrs=%u",
-	  __entry->name, __entry->cmdbufs, __entry->relocs, __entry->waitchks,
-	  __entry->syncpt_id, __entry->syncpt_incrs)
+	TP_printk("name=%s, cmdbufs=%u, relocs=%u, syncpt_id=%u, "
+		  "syncpt_incrs=%u",
+		  __entry->name, __entry->cmdbufs, __entry->relocs,
+		  __entry->syncpt_id, __entry->syncpt_incrs)
 );
 
 TRACE_EVENT(host1x_channel_submitted,
-- 
cgit v1.1


From bf3d41ccabb53c57e19fcfc8b81d790043ac2bed Mon Sep 17 00:00:00 2001
From: Thierry Reding <treding@nvidia.com>
Date: Wed, 16 May 2018 14:12:33 +0200
Subject: gpu: host1x: Store pointer to client in jobs

Rather than storing some identifier derived from the application
context that can't be used concretely anywhere, store a pointer to the
client directly so that accesses can be made directly through that
client object.

Reviewed-by: Dmitry Osipenko <digetx@gmail.com>
Tested-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/drm.c | 5 +++--
 drivers/gpu/host1x/cdma.c   | 2 +-
 drivers/gpu/host1x/cdma.h   | 2 +-
 include/linux/host1x.h      | 3 ++-
 4 files changed, 7 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/tegra/drm.c b/drivers/gpu/drm/tegra/drm.c
index 204b10e..8f29323 100644
--- a/drivers/gpu/drm/tegra/drm.c
+++ b/drivers/gpu/drm/tegra/drm.c
@@ -325,6 +325,7 @@ int tegra_drm_submit(struct tegra_drm_context *context,
 		     struct drm_tegra_submit *args, struct drm_device *drm,
 		     struct drm_file *file)
 {
+	struct host1x_client *client = &context->client->base;
 	unsigned int num_cmdbufs = args->num_cmdbufs;
 	unsigned int num_relocs = args->num_relocs;
 	struct drm_tegra_cmdbuf __user *user_cmdbufs;
@@ -356,8 +357,8 @@ int tegra_drm_submit(struct tegra_drm_context *context,
 		return -ENOMEM;
 
 	job->num_relocs = args->num_relocs;
-	job->client = (u32)args->context;
-	job->class = context->client->base.class;
+	job->client = client;
+	job->class = client->class;
 	job->serialize = true;
 
 	/*
diff --git a/drivers/gpu/host1x/cdma.c b/drivers/gpu/host1x/cdma.c
index 69bb773..91df51e 100644
--- a/drivers/gpu/host1x/cdma.c
+++ b/drivers/gpu/host1x/cdma.c
@@ -247,7 +247,7 @@ static void cdma_start_timer_locked(struct host1x_cdma *cdma,
 static void stop_cdma_timer_locked(struct host1x_cdma *cdma)
 {
 	cancel_delayed_work(&cdma->timeout.wq);
-	cdma->timeout.client = 0;
+	cdma->timeout.client = NULL;
 }
 
 /*
diff --git a/drivers/gpu/host1x/cdma.h b/drivers/gpu/host1x/cdma.h
index 446ee1a..e97e17b 100644
--- a/drivers/gpu/host1x/cdma.h
+++ b/drivers/gpu/host1x/cdma.h
@@ -58,7 +58,7 @@ struct buffer_timeout {
 	u32 syncpt_val;			/* syncpt value when completed */
 	ktime_t start_ktime;		/* starting time */
 	/* context timeout information */
-	int client;
+	struct host1x_client *client;
 };
 
 enum cdma_event {
diff --git a/include/linux/host1x.h b/include/linux/host1x.h
index f66bece..0632010 100644
--- a/include/linux/host1x.h
+++ b/include/linux/host1x.h
@@ -202,7 +202,8 @@ struct host1x_job {
 	/* Channel where job is submitted to */
 	struct host1x_channel *channel;
 
-	u32 client;
+	/* client where the job originated */
+	struct host1x_client *client;
 
 	/* Gathers and their memory */
 	struct host1x_job_gather *gathers;
-- 
cgit v1.1


From d4ad3ad9b81b73f568227563988b67708291900b Mon Sep 17 00:00:00 2001
From: Thierry Reding <treding@nvidia.com>
Date: Fri, 23 Mar 2018 13:31:24 +0100
Subject: gpu: host1x: Cleanup loop variable usage

Use unsigned int where possible and don't unnecessarily initialize the
loop variable.

Reviewed-by: Dmitry Osipenko <digetx@gmail.com>
Tested-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/host1x/debug.c  | 2 +-
 drivers/gpu/host1x/intr.c   | 2 +-
 drivers/gpu/host1x/job.c    | 4 ++--
 drivers/gpu/host1x/syncpt.c | 2 +-
 4 files changed, 5 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/host1x/debug.c b/drivers/gpu/host1x/debug.c
index dc77ec4..329e4a3 100644
--- a/drivers/gpu/host1x/debug.c
+++ b/drivers/gpu/host1x/debug.c
@@ -103,7 +103,7 @@ static void show_syncpts(struct host1x *m, struct output *o)
 
 static void show_all(struct host1x *m, struct output *o, bool show_fifo)
 {
-	int i;
+	unsigned int i;
 
 	host1x_hw_show_mlocks(m, o);
 	show_syncpts(m, o);
diff --git a/drivers/gpu/host1x/intr.c b/drivers/gpu/host1x/intr.c
index 8b4fad0..6028cf7 100644
--- a/drivers/gpu/host1x/intr.c
+++ b/drivers/gpu/host1x/intr.c
@@ -144,7 +144,7 @@ static const action_handler action_handlers[HOST1X_INTR_ACTION_COUNT] = {
 static void run_handlers(struct list_head completed[HOST1X_INTR_ACTION_COUNT])
 {
 	struct list_head *head = completed;
-	int i;
+	unsigned int i;
 
 	for (i = 0; i < HOST1X_INTR_ACTION_COUNT; ++i, ++head) {
 		action_handler handler = action_handlers[i];
diff --git a/drivers/gpu/host1x/job.c b/drivers/gpu/host1x/job.c
index 3cbfc6e..2be0bca 100644
--- a/drivers/gpu/host1x/job.c
+++ b/drivers/gpu/host1x/job.c
@@ -196,10 +196,10 @@ unpin:
 
 static int do_relocs(struct host1x_job *job, struct host1x_job_gather *g)
 {
-	int i = 0;
 	u32 last_page = ~0;
 	void *cmdbuf_page_addr = NULL;
 	struct host1x_bo *cmdbuf = g->bo;
+	unsigned int i;
 
 	/* pin & patch the relocs for one gather */
 	for (i = 0; i < job->num_relocs; i++) {
@@ -451,7 +451,7 @@ static inline int copy_gathers(struct host1x_job *job, struct device *dev)
 	struct host1x_firewall fw;
 	size_t size = 0;
 	size_t offset = 0;
-	int i;
+	unsigned int i;
 
 	fw.job = job;
 	fw.dev = dev;
diff --git a/drivers/gpu/host1x/syncpt.c b/drivers/gpu/host1x/syncpt.c
index a108669..088c05d 100644
--- a/drivers/gpu/host1x/syncpt.c
+++ b/drivers/gpu/host1x/syncpt.c
@@ -57,8 +57,8 @@ static struct host1x_syncpt *host1x_syncpt_alloc(struct host1x *host,
 						 struct host1x_client *client,
 						 unsigned long flags)
 {
-	int i;
 	struct host1x_syncpt *sp = host->syncpt;
+	unsigned int i;
 	char *name;
 
 	mutex_lock(&host->syncpt_mutex);
-- 
cgit v1.1


From ac330f45c7ca5b92e78b369c7034160947f03b8d Mon Sep 17 00:00:00 2001
From: Thierry Reding <treding@nvidia.com>
Date: Wed, 16 May 2018 14:29:33 +0200
Subject: gpu: host1x: Drop unnecessary host1x argument

Functions taking a pointer to a host1x syncpoint as an argument don't
need to specify a pointer to a host1x instance because it can be
obtained from the syncpoint.

Reviewed-by: Dmitry Osipenko <digetx@gmail.com>
Tested-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/host1x/hw/channel_hw.c |  2 +-
 drivers/gpu/host1x/intr.c          | 14 ++++++--------
 drivers/gpu/host1x/intr.h          |  8 +++++---
 drivers/gpu/host1x/syncpt.c        |  2 +-
 4 files changed, 13 insertions(+), 13 deletions(-)

diff --git a/drivers/gpu/host1x/hw/channel_hw.c b/drivers/gpu/host1x/hw/channel_hw.c
index 4c95550..d188f90 100644
--- a/drivers/gpu/host1x/hw/channel_hw.c
+++ b/drivers/gpu/host1x/hw/channel_hw.c
@@ -164,7 +164,7 @@ static int channel_submit(struct host1x_job *job)
 	trace_host1x_channel_submitted(dev_name(ch->dev), prev_max, syncval);
 
 	/* schedule a submit complete interrupt */
-	err = host1x_intr_add_action(host, job->syncpt_id, syncval,
+	err = host1x_intr_add_action(host, sp, syncval,
 				     HOST1X_INTR_ACTION_SUBMIT_COMPLETE, ch,
 				     completed_waiter, NULL);
 	completed_waiter = NULL;
diff --git a/drivers/gpu/host1x/intr.c b/drivers/gpu/host1x/intr.c
index 6028cf7..9629c00 100644
--- a/drivers/gpu/host1x/intr.c
+++ b/drivers/gpu/host1x/intr.c
@@ -211,11 +211,11 @@ static void syncpt_thresh_work(struct work_struct *work)
 				host1x_syncpt_load(host->syncpt + id));
 }
 
-int host1x_intr_add_action(struct host1x *host, unsigned int id, u32 thresh,
-			   enum host1x_intr_action action, void *data,
-			   struct host1x_waitlist *waiter, void **ref)
+int host1x_intr_add_action(struct host1x *host, struct host1x_syncpt *syncpt,
+			   u32 thresh, enum host1x_intr_action action,
+			   void *data, struct host1x_waitlist *waiter,
+			   void **ref)
 {
-	struct host1x_syncpt *syncpt;
 	int queue_was_empty;
 
 	if (waiter == NULL) {
@@ -234,19 +234,17 @@ int host1x_intr_add_action(struct host1x *host, unsigned int id, u32 thresh,
 	waiter->data = data;
 	waiter->count = 1;
 
-	syncpt = host->syncpt + id;
-
 	spin_lock(&syncpt->intr.lock);
 
 	queue_was_empty = list_empty(&syncpt->intr.wait_head);
 
 	if (add_waiter_to_queue(waiter, &syncpt->intr.wait_head)) {
 		/* added at head of list - new threshold value */
-		host1x_hw_intr_set_syncpt_threshold(host, id, thresh);
+		host1x_hw_intr_set_syncpt_threshold(host, syncpt->id, thresh);
 
 		/* added as first waiter - enable interrupt */
 		if (queue_was_empty)
-			host1x_hw_intr_enable_syncpt_intr(host, id);
+			host1x_hw_intr_enable_syncpt_intr(host, syncpt->id);
 	}
 
 	spin_unlock(&syncpt->intr.lock);
diff --git a/drivers/gpu/host1x/intr.h b/drivers/gpu/host1x/intr.h
index 1370c2b..6db96af 100644
--- a/drivers/gpu/host1x/intr.h
+++ b/drivers/gpu/host1x/intr.h
@@ -22,6 +22,7 @@
 #include <linux/interrupt.h>
 #include <linux/workqueue.h>
 
+struct host1x_syncpt;
 struct host1x;
 
 enum host1x_intr_action {
@@ -75,9 +76,10 @@ struct host1x_waitlist {
  *
  * This is a non-blocking api.
  */
-int host1x_intr_add_action(struct host1x *host, unsigned int id, u32 thresh,
-	enum host1x_intr_action action, void *data,
-	struct host1x_waitlist *waiter, void **ref);
+int host1x_intr_add_action(struct host1x *host, struct host1x_syncpt *syncpt,
+			   u32 thresh, enum host1x_intr_action action,
+			   void *data, struct host1x_waitlist *waiter,
+			   void **ref);
 
 /*
  * Unreference an action submitted to host1x_intr_add_action().
diff --git a/drivers/gpu/host1x/syncpt.c b/drivers/gpu/host1x/syncpt.c
index 088c05d..a5dbf1b 100644
--- a/drivers/gpu/host1x/syncpt.c
+++ b/drivers/gpu/host1x/syncpt.c
@@ -255,7 +255,7 @@ int host1x_syncpt_wait(struct host1x_syncpt *sp, u32 thresh, long timeout,
 	}
 
 	/* schedule a wakeup when the syncpoint value is reached */
-	err = host1x_intr_add_action(sp->host, sp->id, thresh,
+	err = host1x_intr_add_action(sp->host, sp, thresh,
 				     HOST1X_INTR_ACTION_WAKEUP_INTERRUPTIBLE,
 				     &wq, waiter, &ref);
 	if (err)
-- 
cgit v1.1


From 06490bb99e1840ab2b6814af7356e8b4ab0e3ee6 Mon Sep 17 00:00:00 2001
From: Thierry Reding <treding@nvidia.com>
Date: Wed, 16 May 2018 16:58:44 +0200
Subject: gpu: host1x: Rename relocarray -> relocs for consistency

All other array variables use a plural, and this is the only one using
the *array suffix. This is confusing, so rename it for consistency.

Reviewed-by: Dmitry Osipenko <digetx@gmail.com>
Tested-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/drm.c | 4 ++--
 drivers/gpu/host1x/job.c    | 8 ++++----
 include/linux/host1x.h      | 2 +-
 3 files changed, 7 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/tegra/drm.c b/drivers/gpu/drm/tegra/drm.c
index 8f29323..bfbd3a8 100644
--- a/drivers/gpu/drm/tegra/drm.c
+++ b/drivers/gpu/drm/tegra/drm.c
@@ -426,13 +426,13 @@ int tegra_drm_submit(struct tegra_drm_context *context,
 		struct host1x_reloc *reloc;
 		struct tegra_bo *obj;
 
-		err = host1x_reloc_copy_from_user(&job->relocarray[num_relocs],
+		err = host1x_reloc_copy_from_user(&job->relocs[num_relocs],
 						  &user_relocs[num_relocs], drm,
 						  file);
 		if (err < 0)
 			goto fail;
 
-		reloc = &job->relocarray[num_relocs];
+		reloc = &job->relocs[num_relocs];
 		obj = host1x_to_tegra_bo(reloc->cmdbuf.bo);
 		refs[num_refs++] = &obj->gem;
 
diff --git a/drivers/gpu/host1x/job.c b/drivers/gpu/host1x/job.c
index 2be0bca..9d6d3e1 100644
--- a/drivers/gpu/host1x/job.c
+++ b/drivers/gpu/host1x/job.c
@@ -60,7 +60,7 @@ struct host1x_job *host1x_job_alloc(struct host1x_channel *ch,
 
 	/* Redistribute memory to the structs  */
 	mem += sizeof(struct host1x_job);
-	job->relocarray = num_relocs ? mem : NULL;
+	job->relocs = num_relocs ? mem : NULL;
 	mem += num_relocs * sizeof(struct host1x_reloc);
 	job->unpins = num_unpins ? mem : NULL;
 	mem += num_unpins * sizeof(struct host1x_job_unpin_data);
@@ -115,7 +115,7 @@ static unsigned int pin_job(struct host1x *host, struct host1x_job *job)
 	job->num_unpins = 0;
 
 	for (i = 0; i < job->num_relocs; i++) {
-		struct host1x_reloc *reloc = &job->relocarray[i];
+		struct host1x_reloc *reloc = &job->relocs[i];
 		struct sg_table *sgt;
 		dma_addr_t phys_addr;
 
@@ -203,7 +203,7 @@ static int do_relocs(struct host1x_job *job, struct host1x_job_gather *g)
 
 	/* pin & patch the relocs for one gather */
 	for (i = 0; i < job->num_relocs; i++) {
-		struct host1x_reloc *reloc = &job->relocarray[i];
+		struct host1x_reloc *reloc = &job->relocs[i];
 		u32 reloc_addr = (job->reloc_addr_phys[i] +
 				  reloc->target.offset) >> reloc->shift;
 		u32 *target;
@@ -455,7 +455,7 @@ static inline int copy_gathers(struct host1x_job *job, struct device *dev)
 
 	fw.job = job;
 	fw.dev = dev;
-	fw.reloc = job->relocarray;
+	fw.reloc = job->relocs;
 	fw.num_relocs = job->num_relocs;
 	fw.class = job->class;
 
diff --git a/include/linux/host1x.h b/include/linux/host1x.h
index 0632010..dcb6140 100644
--- a/include/linux/host1x.h
+++ b/include/linux/host1x.h
@@ -210,7 +210,7 @@ struct host1x_job {
 	unsigned int num_gathers;
 
 	/* Array of handles to be pinned & unpinned */
-	struct host1x_reloc *relocarray;
+	struct host1x_reloc *relocs;
 	unsigned int num_relocs;
 	struct host1x_job_unpin_data *unpins;
 	unsigned int num_unpins;
-- 
cgit v1.1


From 326bbd79fd61716841585a52d5b68f48f4e6644e Mon Sep 17 00:00:00 2001
From: Thierry Reding <treding@nvidia.com>
Date: Wed, 16 May 2018 17:01:43 +0200
Subject: gpu: host1x: Use not explicitly sized types

The number of words and the offset in a gather don't need to be
explicitly sized, so make them unsigned int instead.

Reviewed-by: Dmitry Osipenko <digetx@gmail.com>
Tested-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/host1x/job.c | 11 ++++++-----
 drivers/gpu/host1x/job.h |  4 ++--
 include/linux/host1x.h   |  4 ++--
 3 files changed, 10 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/host1x/job.c b/drivers/gpu/host1x/job.c
index 9d6d3e1..e2f4a4d 100644
--- a/drivers/gpu/host1x/job.c
+++ b/drivers/gpu/host1x/job.c
@@ -96,13 +96,14 @@ void host1x_job_put(struct host1x_job *job)
 EXPORT_SYMBOL(host1x_job_put);
 
 void host1x_job_add_gather(struct host1x_job *job, struct host1x_bo *bo,
-			   u32 words, u32 offset)
+			   unsigned int words, unsigned int offset)
 {
-	struct host1x_job_gather *cur_gather = &job->gathers[job->num_gathers];
+	struct host1x_job_gather *gather = &job->gathers[job->num_gathers];
+
+	gather->words = words;
+	gather->bo = bo;
+	gather->offset = offset;
 
-	cur_gather->words = words;
-	cur_gather->bo = bo;
-	cur_gather->offset = offset;
 	job->num_gathers++;
 }
 EXPORT_SYMBOL(host1x_job_add_gather);
diff --git a/drivers/gpu/host1x/job.h b/drivers/gpu/host1x/job.h
index 4bda51d..188400e 100644
--- a/drivers/gpu/host1x/job.h
+++ b/drivers/gpu/host1x/job.h
@@ -20,10 +20,10 @@
 #define __HOST1X_JOB_H
 
 struct host1x_job_gather {
-	u32 words;
+	unsigned int words;
 	dma_addr_t base;
 	struct host1x_bo *bo;
-	u32 offset;
+	unsigned int offset;
 	bool handled;
 };
 
diff --git a/include/linux/host1x.h b/include/linux/host1x.h
index dcb6140..89110d8 100644
--- a/include/linux/host1x.h
+++ b/include/linux/host1x.h
@@ -251,8 +251,8 @@ struct host1x_job {
 
 struct host1x_job *host1x_job_alloc(struct host1x_channel *ch,
 				    u32 num_cmdbufs, u32 num_relocs);
-void host1x_job_add_gather(struct host1x_job *job, struct host1x_bo *mem_id,
-			   u32 words, u32 offset);
+void host1x_job_add_gather(struct host1x_job *job, struct host1x_bo *bo,
+			   unsigned int words, unsigned int offset);
 struct host1x_job *host1x_job_get(struct host1x_job *job);
 void host1x_job_put(struct host1x_job *job);
 int host1x_job_pin(struct host1x_job *job, struct device *dev);
-- 
cgit v1.1


From c850ece71f71c2a68a9921c52fb5fd8d3ec2b8d7 Mon Sep 17 00:00:00 2001
From: Thierry Reding <treding@nvidia.com>
Date: Sat, 5 May 2018 08:12:53 +0200
Subject: drm/tegra: Use proper arguments for DRM_TEGRA_CLOSE_CHANNEL IOCTL

A separate data structure exists for the DRM_TEGRA_CLOSE_CHANNEL IOCTL,
but it is currently unused. The IOCTL was using the data structure for
the DRM_TEGRA_OPEN_CHANNEL IOCTL.

Reviewed-by: Dmitry Osipenko <digetx@gmail.com>
Tested-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 include/uapi/drm/tegra_drm.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/include/uapi/drm/tegra_drm.h b/include/uapi/drm/tegra_drm.h
index d954f8c..99e15d8 100644
--- a/include/uapi/drm/tegra_drm.h
+++ b/include/uapi/drm/tegra_drm.h
@@ -193,7 +193,7 @@ struct drm_tegra_gem_get_flags {
 #define DRM_IOCTL_TEGRA_SYNCPT_INCR DRM_IOWR(DRM_COMMAND_BASE + DRM_TEGRA_SYNCPT_INCR, struct drm_tegra_syncpt_incr)
 #define DRM_IOCTL_TEGRA_SYNCPT_WAIT DRM_IOWR(DRM_COMMAND_BASE + DRM_TEGRA_SYNCPT_WAIT, struct drm_tegra_syncpt_wait)
 #define DRM_IOCTL_TEGRA_OPEN_CHANNEL DRM_IOWR(DRM_COMMAND_BASE + DRM_TEGRA_OPEN_CHANNEL, struct drm_tegra_open_channel)
-#define DRM_IOCTL_TEGRA_CLOSE_CHANNEL DRM_IOWR(DRM_COMMAND_BASE + DRM_TEGRA_CLOSE_CHANNEL, struct drm_tegra_open_channel)
+#define DRM_IOCTL_TEGRA_CLOSE_CHANNEL DRM_IOWR(DRM_COMMAND_BASE + DRM_TEGRA_CLOSE_CHANNEL, struct drm_tegra_close_channel)
 #define DRM_IOCTL_TEGRA_GET_SYNCPT DRM_IOWR(DRM_COMMAND_BASE + DRM_TEGRA_GET_SYNCPT, struct drm_tegra_get_syncpt)
 #define DRM_IOCTL_TEGRA_SUBMIT DRM_IOWR(DRM_COMMAND_BASE + DRM_TEGRA_SUBMIT, struct drm_tegra_submit)
 #define DRM_IOCTL_TEGRA_GET_SYNCPT_BASE DRM_IOWR(DRM_COMMAND_BASE + DRM_TEGRA_GET_SYNCPT_BASE, struct drm_tegra_get_syncpt_base)
-- 
cgit v1.1


From 4bd91a5b5dbb8b536208396c3d032cba8e3c3913 Mon Sep 17 00:00:00 2001
From: Thierry Reding <treding@nvidia.com>
Date: Wed, 16 May 2018 18:49:04 +0200
Subject: drm/tegra: gem: Fill in missing export info

Set the owner and name of the exported DMA-BUF in addition to the
already filled-in fields.

Reviewed-by: Dmitry Osipenko <digetx@gmail.com>
Tested-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/gem.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/gpu/drm/tegra/gem.c b/drivers/gpu/drm/tegra/gem.c
index 1c40117..00a5c9f 100644
--- a/drivers/gpu/drm/tegra/gem.c
+++ b/drivers/gpu/drm/tegra/gem.c
@@ -649,6 +649,8 @@ struct dma_buf *tegra_gem_prime_export(struct drm_device *drm,
 {
 	DEFINE_DMA_BUF_EXPORT_INFO(exp_info);
 
+	exp_info.exp_name = KBUILD_MODNAME;
+	exp_info.owner = drm->driver->fops->owner;
 	exp_info.ops = &tegra_gem_prime_dmabuf_ops;
 	exp_info.size = gem->size;
 	exp_info.flags = flags;
-- 
cgit v1.1


From 995c5a509fb032ddd83eff4f3772c7fc8ff0b7ec Mon Sep 17 00:00:00 2001
From: Thierry Reding <treding@nvidia.com>
Date: Mon, 19 Mar 2018 17:20:46 +0100
Subject: drm/tegra: dc: Support rotation property

Currently only the DRM_MODE_REFLECT_Y rotation is supported. The driver
already supports reflection on the Y axis via a custom flag which is not
very useful because it requires custom userspace. Add the standard
rotation property that supports 0 degree rotation and Y axis reflection
for primary and overlay planes to provide a better interface than the
custom flag.

v2: keep custom flag for ABI compatibility (Dmitry)

Reviewed-by: Dmitry Osipenko <digetx@gmail.com>
Tested-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/dc.c    | 26 +++++++++++++++++++++++++-
 drivers/gpu/drm/tegra/plane.c |  1 +
 drivers/gpu/drm/tegra/plane.h |  2 ++
 3 files changed, 28 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/tegra/dc.c b/drivers/gpu/drm/tegra/dc.c
index 31e12a9..c3afe7b2 100644
--- a/drivers/gpu/drm/tegra/dc.c
+++ b/drivers/gpu/drm/tegra/dc.c
@@ -596,6 +596,7 @@ static int tegra_plane_atomic_check(struct drm_plane *plane,
 				    struct drm_plane_state *state)
 {
 	struct tegra_plane_state *plane_state = to_tegra_plane_state(state);
+	unsigned int rotation = DRM_MODE_ROTATE_0 | DRM_MODE_REFLECT_Y;
 	struct tegra_bo_tiling *tiling = &plane_state->tiling;
 	struct tegra_plane *tegra = to_tegra_plane(plane);
 	struct tegra_dc *dc = to_tegra_dc(state->crtc);
@@ -633,6 +634,13 @@ static int tegra_plane_atomic_check(struct drm_plane *plane,
 		return -EINVAL;
 	}
 
+	rotation = drm_rotation_simplify(state->rotation, rotation);
+
+	if (rotation & DRM_MODE_REFLECT_Y)
+		plane_state->bottom_up = true;
+	else
+		plane_state->bottom_up = false;
+
 	/*
 	 * Tegra doesn't support different strides for U and V planes so we
 	 * error out if the user tries to display a framebuffer with such a
@@ -693,7 +701,7 @@ static void tegra_plane_atomic_update(struct drm_plane *plane,
 	window.dst.w = drm_rect_width(&plane->state->dst);
 	window.dst.h = drm_rect_height(&plane->state->dst);
 	window.bits_per_pixel = fb->format->cpp[0] * 8;
-	window.bottom_up = tegra_fb_is_bottom_up(fb);
+	window.bottom_up = tegra_fb_is_bottom_up(fb) || state->bottom_up;
 
 	/* copy from state */
 	window.zpos = plane->state->normalized_zpos;
@@ -776,6 +784,14 @@ static struct drm_plane *tegra_primary_plane_create(struct drm_device *drm,
 	drm_plane_helper_add(&plane->base, &tegra_plane_helper_funcs);
 	drm_plane_create_zpos_property(&plane->base, plane->index, 0, 255);
 
+	err = drm_plane_create_rotation_property(&plane->base,
+						 DRM_MODE_ROTATE_0,
+						 DRM_MODE_ROTATE_0 |
+						 DRM_MODE_REFLECT_Y);
+	if (err < 0)
+		dev_err(dc->dev, "failed to create rotation property: %d\n",
+			err);
+
 	return &plane->base;
 }
 
@@ -1053,6 +1069,14 @@ static struct drm_plane *tegra_dc_overlay_plane_create(struct drm_device *drm,
 	drm_plane_helper_add(&plane->base, &tegra_plane_helper_funcs);
 	drm_plane_create_zpos_property(&plane->base, plane->index, 0, 255);
 
+	err = drm_plane_create_rotation_property(&plane->base,
+						 DRM_MODE_ROTATE_0,
+						 DRM_MODE_ROTATE_0 |
+						 DRM_MODE_REFLECT_Y);
+	if (err < 0)
+		dev_err(dc->dev, "failed to create rotation property: %d\n",
+			err);
+
 	return &plane->base;
 }
 
diff --git a/drivers/gpu/drm/tegra/plane.c b/drivers/gpu/drm/tegra/plane.c
index 0406c2e..d068e8a 100644
--- a/drivers/gpu/drm/tegra/plane.c
+++ b/drivers/gpu/drm/tegra/plane.c
@@ -56,6 +56,7 @@ tegra_plane_atomic_duplicate_state(struct drm_plane *plane)
 	copy->tiling = state->tiling;
 	copy->format = state->format;
 	copy->swap = state->swap;
+	copy->bottom_up = state->bottom_up;
 	copy->opaque = state->opaque;
 
 	for (i = 0; i < 2; i++)
diff --git a/drivers/gpu/drm/tegra/plane.h b/drivers/gpu/drm/tegra/plane.h
index 7360ddf..e79e6b4 100644
--- a/drivers/gpu/drm/tegra/plane.h
+++ b/drivers/gpu/drm/tegra/plane.h
@@ -46,6 +46,8 @@ struct tegra_plane_state {
 	u32 format;
 	u32 swap;
 
+	bool bottom_up;
+
 	/* used for legacy blending support only */
 	struct tegra_plane_legacy_blending_state blending[2];
 	bool opaque;
-- 
cgit v1.1


From f3b3cfcc3f09490ffb8e1e997e8a8695a6a55b1b Mon Sep 17 00:00:00 2001
From: Thierry Reding <treding@nvidia.com>
Date: Wed, 16 May 2018 17:05:04 +0200
Subject: drm/tegra: Track client version

Userspace needs to know the version of the interface implemented by a
client so it can create the proper command streams. Allow individual
drivers to store this version along with the client so that it can be
returned to userspace upon opening a channel.

Acked-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/drm.h | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/tegra/drm.h b/drivers/gpu/drm/tegra/drm.h
index f47a605..92d2487 100644
--- a/drivers/gpu/drm/tegra/drm.h
+++ b/drivers/gpu/drm/tegra/drm.h
@@ -91,6 +91,7 @@ struct tegra_drm_client {
 	struct host1x_client base;
 	struct list_head list;
 
+	unsigned int version;
 	const struct tegra_drm_client_ops *ops;
 };
 
-- 
cgit v1.1


From 840fd213fca23b185f71b45a5b563e4e9b6d1759 Mon Sep 17 00:00:00 2001
From: Thierry Reding <treding@nvidia.com>
Date: Wed, 16 May 2018 17:06:36 +0200
Subject: drm/tegra: gr2d: Track interface version

Set the interface version implemented by the gr2d module. This allows
userspace to pass the correct command stream when programming the gr2d
module.

Reviewed-by: Dmitry Osipenko <digetx@gmail.com>
Tested-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/gr2d.c | 22 ++++++++++++++++++++--
 1 file changed, 20 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/tegra/gr2d.c b/drivers/gpu/drm/tegra/gr2d.c
index 2cd0f66..673059f 100644
--- a/drivers/gpu/drm/tegra/gr2d.c
+++ b/drivers/gpu/drm/tegra/gr2d.c
@@ -8,17 +8,24 @@
 
 #include <linux/clk.h>
 #include <linux/iommu.h>
+#include <linux/of_device.h>
 
 #include "drm.h"
 #include "gem.h"
 #include "gr2d.h"
 
+struct gr2d_soc {
+	unsigned int version;
+};
+
 struct gr2d {
 	struct iommu_group *group;
 	struct tegra_drm_client client;
 	struct host1x_channel *channel;
 	struct clk *clk;
 
+	const struct gr2d_soc *soc;
+
 	DECLARE_BITMAP(addr_regs, GR2D_NUM_REGS);
 };
 
@@ -150,9 +157,17 @@ static const struct tegra_drm_client_ops gr2d_ops = {
 	.submit = tegra_drm_submit,
 };
 
+static const struct gr2d_soc tegra20_gr2d_soc = {
+	.version = 0x20,
+};
+
+static const struct gr2d_soc tegra30_gr2d_soc = {
+	.version = 0x30,
+};
+
 static const struct of_device_id gr2d_match[] = {
-	{ .compatible = "nvidia,tegra30-gr2d" },
-	{ .compatible = "nvidia,tegra20-gr2d" },
+	{ .compatible = "nvidia,tegra30-gr2d", .data = &tegra20_gr2d_soc },
+	{ .compatible = "nvidia,tegra20-gr2d", .data = &tegra30_gr2d_soc },
 	{ },
 };
 MODULE_DEVICE_TABLE(of, gr2d_match);
@@ -185,6 +200,8 @@ static int gr2d_probe(struct platform_device *pdev)
 	if (!gr2d)
 		return -ENOMEM;
 
+	gr2d->soc = of_device_get_match_data(dev);
+
 	syncpts = devm_kzalloc(dev, sizeof(*syncpts), GFP_KERNEL);
 	if (!syncpts)
 		return -ENOMEM;
@@ -209,6 +226,7 @@ static int gr2d_probe(struct platform_device *pdev)
 	gr2d->client.base.num_syncpts = 1;
 
 	INIT_LIST_HEAD(&gr2d->client.list);
+	gr2d->client.version = gr2d->soc->version;
 	gr2d->client.ops = &gr2d_ops;
 
 	err = host1x_client_register(&gr2d->client.base);
-- 
cgit v1.1


From 33f150ea82ff029ec7e00345c6fbf00e44a8fd60 Mon Sep 17 00:00:00 2001
From: Thierry Reding <treding@nvidia.com>
Date: Wed, 16 May 2018 17:07:38 +0200
Subject: drm/tegra: gr3d: Track interface version

Set the interface version implemented by the gr3d module. This allows
userspace to pass the correct command stream when programming the gr3d
module.

Reviewed-by: Dmitry Osipenko <digetx@gmail.com>
Tested-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/gr3d.c | 28 +++++++++++++++++++++++++---
 1 file changed, 25 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/tegra/gr3d.c b/drivers/gpu/drm/tegra/gr3d.c
index b00002f..4778ae9 100644
--- a/drivers/gpu/drm/tegra/gr3d.c
+++ b/drivers/gpu/drm/tegra/gr3d.c
@@ -11,6 +11,7 @@
 #include <linux/host1x.h>
 #include <linux/iommu.h>
 #include <linux/module.h>
+#include <linux/of_device.h>
 #include <linux/platform_device.h>
 #include <linux/reset.h>
 
@@ -20,6 +21,10 @@
 #include "gem.h"
 #include "gr3d.h"
 
+struct gr3d_soc {
+	unsigned int version;
+};
+
 struct gr3d {
 	struct iommu_group *group;
 	struct tegra_drm_client client;
@@ -29,6 +34,8 @@ struct gr3d {
 	struct reset_control *rst_secondary;
 	struct reset_control *rst;
 
+	const struct gr3d_soc *soc;
+
 	DECLARE_BITMAP(addr_regs, GR3D_NUM_REGS);
 };
 
@@ -151,10 +158,22 @@ static const struct tegra_drm_client_ops gr3d_ops = {
 	.submit = tegra_drm_submit,
 };
 
+static const struct gr3d_soc tegra20_gr3d_soc = {
+	.version = 0x20,
+};
+
+static const struct gr3d_soc tegra30_gr3d_soc = {
+	.version = 0x30,
+};
+
+static const struct gr3d_soc tegra114_gr3d_soc = {
+	.version = 0x35,
+};
+
 static const struct of_device_id tegra_gr3d_match[] = {
-	{ .compatible = "nvidia,tegra114-gr3d" },
-	{ .compatible = "nvidia,tegra30-gr3d" },
-	{ .compatible = "nvidia,tegra20-gr3d" },
+	{ .compatible = "nvidia,tegra114-gr3d", .data = &tegra114_gr3d_soc },
+	{ .compatible = "nvidia,tegra30-gr3d", .data = &tegra30_gr3d_soc },
+	{ .compatible = "nvidia,tegra20-gr3d", .data = &tegra20_gr3d_soc },
 	{ }
 };
 MODULE_DEVICE_TABLE(of, tegra_gr3d_match);
@@ -276,6 +295,8 @@ static int gr3d_probe(struct platform_device *pdev)
 	if (!gr3d)
 		return -ENOMEM;
 
+	gr3d->soc = of_device_get_match_data(&pdev->dev);
+
 	syncpts = devm_kzalloc(&pdev->dev, sizeof(*syncpts), GFP_KERNEL);
 	if (!syncpts)
 		return -ENOMEM;
@@ -333,6 +354,7 @@ static int gr3d_probe(struct platform_device *pdev)
 	gr3d->client.base.num_syncpts = 1;
 
 	INIT_LIST_HEAD(&gr3d->client.list);
+	gr3d->client.version = gr3d->soc->version;
 	gr3d->client.ops = &gr3d_ops;
 
 	err = host1x_client_register(&gr3d->client.base);
-- 
cgit v1.1


From acae8a9d054daa75a01e34b18f3627e6df330622 Mon Sep 17 00:00:00 2001
From: Thierry Reding <treding@nvidia.com>
Date: Wed, 16 May 2018 17:08:04 +0200
Subject: drm/tegra: vic: Track interface version

Set the interface version implemented by the VIC module. This allows
userspace to pass the correct command stream when programming the VIC
module.

Reviewed-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/gpu/drm/tegra/vic.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/gpu/drm/tegra/vic.c b/drivers/gpu/drm/tegra/vic.c
index f5794dd..9f657a6 100644
--- a/drivers/gpu/drm/tegra/vic.c
+++ b/drivers/gpu/drm/tegra/vic.c
@@ -25,6 +25,7 @@
 
 struct vic_config {
 	const char *firmware;
+	unsigned int version;
 };
 
 struct vic {
@@ -264,18 +265,21 @@ static const struct tegra_drm_client_ops vic_ops = {
 
 static const struct vic_config vic_t124_config = {
 	.firmware = NVIDIA_TEGRA_124_VIC_FIRMWARE,
+	.version = 0x40,
 };
 
 #define NVIDIA_TEGRA_210_VIC_FIRMWARE "nvidia/tegra210/vic04_ucode.bin"
 
 static const struct vic_config vic_t210_config = {
 	.firmware = NVIDIA_TEGRA_210_VIC_FIRMWARE,
+	.version = 0x21,
 };
 
 #define NVIDIA_TEGRA_186_VIC_FIRMWARE "nvidia/tegra186/vic04_ucode.bin"
 
 static const struct vic_config vic_t186_config = {
 	.firmware = NVIDIA_TEGRA_186_VIC_FIRMWARE,
+	.version = 0x18,
 };
 
 static const struct of_device_id vic_match[] = {
@@ -342,6 +346,7 @@ static int vic_probe(struct platform_device *pdev)
 	vic->dev = dev;
 
 	INIT_LIST_HEAD(&vic->client.list);
+	vic->client.version = vic->config->version;
 	vic->client.ops = &vic_ops;
 
 	err = host1x_client_register(&vic->client.base);
-- 
cgit v1.1


From 3fdbab5f5689a656fa719df752ca7608bcf66c99 Mon Sep 17 00:00:00 2001
From: Evan Quan <evan.quan@amd.com>
Date: Mon, 26 Mar 2018 11:43:04 +0800
Subject: drm/amd/powerplay: update vega20 cg flags (v2)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

v2: remove duplicate flag.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/soc15.c | 18 +++++++++++++++++-
 1 file changed, 17 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index 4e065c6..63135cf 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -665,7 +665,23 @@ static int soc15_common_early_init(void *handle)
 		adev->external_rev_id = adev->rev_id + 0x14;
 		break;
 	case CHIP_VEGA20:
-		adev->cg_flags = 0;
+		adev->cg_flags = AMD_CG_SUPPORT_GFX_MGCG |
+			AMD_CG_SUPPORT_GFX_MGLS |
+			AMD_CG_SUPPORT_GFX_CGCG |
+			AMD_CG_SUPPORT_GFX_CGLS |
+			AMD_CG_SUPPORT_GFX_3D_CGCG |
+			AMD_CG_SUPPORT_GFX_3D_CGLS |
+			AMD_CG_SUPPORT_GFX_CP_LS |
+			AMD_CG_SUPPORT_MC_LS |
+			AMD_CG_SUPPORT_MC_MGCG |
+			AMD_CG_SUPPORT_SDMA_MGCG |
+			AMD_CG_SUPPORT_SDMA_LS |
+			AMD_CG_SUPPORT_BIF_MGCG |
+			AMD_CG_SUPPORT_BIF_LS |
+			AMD_CG_SUPPORT_HDP_MGCG |
+			AMD_CG_SUPPORT_ROM_MGCG |
+			AMD_CG_SUPPORT_VCE_MGCG |
+			AMD_CG_SUPPORT_UVD_MGCG;
 		adev->pg_flags = 0;
 		adev->external_rev_id = adev->rev_id + 0x28;
 		break;
-- 
cgit v1.1


From 602ed6c69b128b77050e178aca9e945d969f3aa8 Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Tue, 24 Apr 2018 11:20:16 +0800
Subject: drm/amdgpu: Disable ip modules that are not ready yet
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Please enable above ips on soc15.c when they're available.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/soc15.c | 14 +++++++++-----
 1 file changed, 9 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index 63135cf..295bc9c 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -514,9 +514,11 @@ int soc15_set_ip_blocks(struct amdgpu_device *adev)
 		amdgpu_device_ip_block_add(adev, &vega10_common_ip_block);
 		amdgpu_device_ip_block_add(adev, &gmc_v9_0_ip_block);
 		amdgpu_device_ip_block_add(adev, &vega10_ih_ip_block);
-		amdgpu_device_ip_block_add(adev, &psp_v3_1_ip_block);
-		if (!amdgpu_sriov_vf(adev))
-			amdgpu_device_ip_block_add(adev, &pp_smu_ip_block);
+		if (adev->asic_type != CHIP_VEGA20) {
+			amdgpu_device_ip_block_add(adev, &psp_v3_1_ip_block);
+			if (!amdgpu_sriov_vf(adev))
+				amdgpu_device_ip_block_add(adev, &pp_smu_ip_block);
+		}
 		if (adev->enable_virtual_display || amdgpu_sriov_vf(adev))
 			amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block);
 #if defined(CONFIG_DRM_AMD_DC)
@@ -527,8 +529,10 @@ int soc15_set_ip_blocks(struct amdgpu_device *adev)
 #endif
 		amdgpu_device_ip_block_add(adev, &gfx_v9_0_ip_block);
 		amdgpu_device_ip_block_add(adev, &sdma_v4_0_ip_block);
-		amdgpu_device_ip_block_add(adev, &uvd_v7_0_ip_block);
-		amdgpu_device_ip_block_add(adev, &vce_v4_0_ip_block);
+		if (adev->asic_type != CHIP_VEGA20) {
+			amdgpu_device_ip_block_add(adev, &uvd_v7_0_ip_block);
+			amdgpu_device_ip_block_add(adev, &vce_v4_0_ip_block);
+		}
 		break;
 	case CHIP_RAVEN:
 		amdgpu_device_ip_block_add(adev, &vega10_common_ip_block);
-- 
cgit v1.1


From 2bb795f5ba9cd676536858a978b9df06f473af88 Mon Sep 17 00:00:00 2001
From: James Zhu <James.Zhu@amd.com>
Date: Tue, 15 May 2018 14:25:46 -0500
Subject: drm/amdgpu/vg20:Restruct uvd to support multiple uvds

Vega20 has dual-UVD. Need Restruct amdgpu_device::uvd to support
multiple uvds. There are no any logical changes here.

Signed-off-by: James Zhu <James.Zhu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c     |   6 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c       |   4 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_queue_mgr.c |   4 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c       | 102 +++++++++----------
 drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h       |  19 ++--
 drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c         |  27 +++---
 drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c         |  25 ++---
 drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c         |  77 +++++++--------
 drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c         | 135 +++++++++++++-------------
 9 files changed, 205 insertions(+), 194 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
index d09fcab..1070f40 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
@@ -376,14 +376,14 @@ int amdgpu_fence_driver_start_ring(struct amdgpu_ring *ring,
 	struct amdgpu_device *adev = ring->adev;
 	uint64_t index;
 
-	if (ring != &adev->uvd.ring) {
+	if (ring != &adev->uvd.inst->ring) {
 		ring->fence_drv.cpu_addr = &adev->wb.wb[ring->fence_offs];
 		ring->fence_drv.gpu_addr = adev->wb.gpu_addr + (ring->fence_offs * 4);
 	} else {
 		/* put fence directly behind firmware */
 		index = ALIGN(adev->uvd.fw->size, 8);
-		ring->fence_drv.cpu_addr = adev->uvd.cpu_addr + index;
-		ring->fence_drv.gpu_addr = adev->uvd.gpu_addr + index;
+		ring->fence_drv.cpu_addr = adev->uvd.inst->cpu_addr + index;
+		ring->fence_drv.gpu_addr = adev->uvd.inst->gpu_addr + index;
 	}
 	amdgpu_fence_write(ring, atomic_read(&ring->fence_drv.last_seq));
 	amdgpu_irq_get(adev, irq_src, irq_type);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
index eb4785e..5620ed2 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
@@ -348,7 +348,7 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 			break;
 		case AMDGPU_HW_IP_UVD:
 			type = AMD_IP_BLOCK_TYPE_UVD;
-			ring_mask = adev->uvd.ring.ready ? 1 : 0;
+			ring_mask = adev->uvd.inst->ring.ready ? 1 : 0;
 			ib_start_alignment = AMDGPU_GPU_PAGE_SIZE;
 			ib_size_alignment = 16;
 			break;
@@ -362,7 +362,7 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 		case AMDGPU_HW_IP_UVD_ENC:
 			type = AMD_IP_BLOCK_TYPE_UVD;
 			for (i = 0; i < adev->uvd.num_enc_rings; i++)
-				ring_mask |= ((adev->uvd.ring_enc[i].ready ? 1 : 0) << i);
+				ring_mask |= ((adev->uvd.inst->ring_enc[i].ready ? 1 : 0) << i);
 			ib_start_alignment = AMDGPU_GPU_PAGE_SIZE;
 			ib_size_alignment = 1;
 			break;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_queue_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_queue_mgr.c
index 262c126..2458d38 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_queue_mgr.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_queue_mgr.c
@@ -77,13 +77,13 @@ static int amdgpu_identity_map(struct amdgpu_device *adev,
 		*out_ring = &adev->sdma.instance[ring].ring;
 		break;
 	case AMDGPU_HW_IP_UVD:
-		*out_ring = &adev->uvd.ring;
+		*out_ring = &adev->uvd.inst->ring;
 		break;
 	case AMDGPU_HW_IP_VCE:
 		*out_ring = &adev->vce.ring[ring];
 		break;
 	case AMDGPU_HW_IP_UVD_ENC:
-		*out_ring = &adev->uvd.ring_enc[ring];
+		*out_ring = &adev->uvd.inst->ring_enc[ring];
 		break;
 	case AMDGPU_HW_IP_VCN_DEC:
 		*out_ring = &adev->vcn.ring_dec;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
index fd1e9cd..02683a0 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
@@ -129,7 +129,7 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
 	unsigned version_major, version_minor, family_id;
 	int i, r;
 
-	INIT_DELAYED_WORK(&adev->uvd.idle_work, amdgpu_uvd_idle_work_handler);
+	INIT_DELAYED_WORK(&adev->uvd.inst->idle_work, amdgpu_uvd_idle_work_handler);
 
 	switch (adev->asic_type) {
 #ifdef CONFIG_DRM_AMDGPU_CIK
@@ -237,16 +237,16 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
 		bo_size += AMDGPU_GPU_PAGE_ALIGN(le32_to_cpu(hdr->ucode_size_bytes) + 8);
 
 	r = amdgpu_bo_create_kernel(adev, bo_size, PAGE_SIZE,
-				    AMDGPU_GEM_DOMAIN_VRAM, &adev->uvd.vcpu_bo,
-				    &adev->uvd.gpu_addr, &adev->uvd.cpu_addr);
+				    AMDGPU_GEM_DOMAIN_VRAM, &adev->uvd.inst->vcpu_bo,
+				    &adev->uvd.inst->gpu_addr, &adev->uvd.inst->cpu_addr);
 	if (r) {
 		dev_err(adev->dev, "(%d) failed to allocate UVD bo\n", r);
 		return r;
 	}
 
-	ring = &adev->uvd.ring;
+	ring = &adev->uvd.inst->ring;
 	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
-	r = drm_sched_entity_init(&ring->sched, &adev->uvd.entity,
+	r = drm_sched_entity_init(&ring->sched, &adev->uvd.inst->entity,
 				  rq, NULL);
 	if (r != 0) {
 		DRM_ERROR("Failed setting up UVD run queue.\n");
@@ -254,8 +254,8 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
 	}
 
 	for (i = 0; i < adev->uvd.max_handles; ++i) {
-		atomic_set(&adev->uvd.handles[i], 0);
-		adev->uvd.filp[i] = NULL;
+		atomic_set(&adev->uvd.inst->handles[i], 0);
+		adev->uvd.inst->filp[i] = NULL;
 	}
 
 	/* from uvd v5.0 HW addressing capacity increased to 64 bits */
@@ -285,18 +285,18 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
 int amdgpu_uvd_sw_fini(struct amdgpu_device *adev)
 {
 	int i;
-	kfree(adev->uvd.saved_bo);
+	kfree(adev->uvd.inst->saved_bo);
 
-	drm_sched_entity_fini(&adev->uvd.ring.sched, &adev->uvd.entity);
+	drm_sched_entity_fini(&adev->uvd.inst->ring.sched, &adev->uvd.inst->entity);
 
-	amdgpu_bo_free_kernel(&adev->uvd.vcpu_bo,
-			      &adev->uvd.gpu_addr,
-			      (void **)&adev->uvd.cpu_addr);
+	amdgpu_bo_free_kernel(&adev->uvd.inst->vcpu_bo,
+			      &adev->uvd.inst->gpu_addr,
+			      (void **)&adev->uvd.inst->cpu_addr);
 
-	amdgpu_ring_fini(&adev->uvd.ring);
+	amdgpu_ring_fini(&adev->uvd.inst->ring);
 
 	for (i = 0; i < AMDGPU_MAX_UVD_ENC_RINGS; ++i)
-		amdgpu_ring_fini(&adev->uvd.ring_enc[i]);
+		amdgpu_ring_fini(&adev->uvd.inst->ring_enc[i]);
 
 	release_firmware(adev->uvd.fw);
 
@@ -309,29 +309,29 @@ int amdgpu_uvd_suspend(struct amdgpu_device *adev)
 	void *ptr;
 	int i;
 
-	if (adev->uvd.vcpu_bo == NULL)
+	if (adev->uvd.inst->vcpu_bo == NULL)
 		return 0;
 
-	cancel_delayed_work_sync(&adev->uvd.idle_work);
+	cancel_delayed_work_sync(&adev->uvd.inst->idle_work);
 
 	/* only valid for physical mode */
 	if (adev->asic_type < CHIP_POLARIS10) {
 		for (i = 0; i < adev->uvd.max_handles; ++i)
-			if (atomic_read(&adev->uvd.handles[i]))
+			if (atomic_read(&adev->uvd.inst->handles[i]))
 				break;
 
 		if (i == adev->uvd.max_handles)
 			return 0;
 	}
 
-	size = amdgpu_bo_size(adev->uvd.vcpu_bo);
-	ptr = adev->uvd.cpu_addr;
+	size = amdgpu_bo_size(adev->uvd.inst->vcpu_bo);
+	ptr = adev->uvd.inst->cpu_addr;
 
-	adev->uvd.saved_bo = kmalloc(size, GFP_KERNEL);
-	if (!adev->uvd.saved_bo)
+	adev->uvd.inst->saved_bo = kmalloc(size, GFP_KERNEL);
+	if (!adev->uvd.inst->saved_bo)
 		return -ENOMEM;
 
-	memcpy_fromio(adev->uvd.saved_bo, ptr, size);
+	memcpy_fromio(adev->uvd.inst->saved_bo, ptr, size);
 
 	return 0;
 }
@@ -341,16 +341,16 @@ int amdgpu_uvd_resume(struct amdgpu_device *adev)
 	unsigned size;
 	void *ptr;
 
-	if (adev->uvd.vcpu_bo == NULL)
+	if (adev->uvd.inst->vcpu_bo == NULL)
 		return -EINVAL;
 
-	size = amdgpu_bo_size(adev->uvd.vcpu_bo);
-	ptr = adev->uvd.cpu_addr;
+	size = amdgpu_bo_size(adev->uvd.inst->vcpu_bo);
+	ptr = adev->uvd.inst->cpu_addr;
 
-	if (adev->uvd.saved_bo != NULL) {
-		memcpy_toio(ptr, adev->uvd.saved_bo, size);
-		kfree(adev->uvd.saved_bo);
-		adev->uvd.saved_bo = NULL;
+	if (adev->uvd.inst->saved_bo != NULL) {
+		memcpy_toio(ptr, adev->uvd.inst->saved_bo, size);
+		kfree(adev->uvd.inst->saved_bo);
+		adev->uvd.inst->saved_bo = NULL;
 	} else {
 		const struct common_firmware_header *hdr;
 		unsigned offset;
@@ -358,14 +358,14 @@ int amdgpu_uvd_resume(struct amdgpu_device *adev)
 		hdr = (const struct common_firmware_header *)adev->uvd.fw->data;
 		if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) {
 			offset = le32_to_cpu(hdr->ucode_array_offset_bytes);
-			memcpy_toio(adev->uvd.cpu_addr, adev->uvd.fw->data + offset,
+			memcpy_toio(adev->uvd.inst->cpu_addr, adev->uvd.fw->data + offset,
 				    le32_to_cpu(hdr->ucode_size_bytes));
 			size -= le32_to_cpu(hdr->ucode_size_bytes);
 			ptr += le32_to_cpu(hdr->ucode_size_bytes);
 		}
 		memset_io(ptr, 0, size);
 		/* to restore uvd fence seq */
-		amdgpu_fence_driver_force_completion(&adev->uvd.ring);
+		amdgpu_fence_driver_force_completion(&adev->uvd.inst->ring);
 	}
 
 	return 0;
@@ -373,12 +373,12 @@ int amdgpu_uvd_resume(struct amdgpu_device *adev)
 
 void amdgpu_uvd_free_handles(struct amdgpu_device *adev, struct drm_file *filp)
 {
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
 	int i, r;
 
 	for (i = 0; i < adev->uvd.max_handles; ++i) {
-		uint32_t handle = atomic_read(&adev->uvd.handles[i]);
-		if (handle != 0 && adev->uvd.filp[i] == filp) {
+		uint32_t handle = atomic_read(&adev->uvd.inst->handles[i]);
+		if (handle != 0 && adev->uvd.inst->filp[i] == filp) {
 			struct dma_fence *fence;
 
 			r = amdgpu_uvd_get_destroy_msg(ring, handle,
@@ -391,8 +391,8 @@ void amdgpu_uvd_free_handles(struct amdgpu_device *adev, struct drm_file *filp)
 			dma_fence_wait(fence, false);
 			dma_fence_put(fence);
 
-			adev->uvd.filp[i] = NULL;
-			atomic_set(&adev->uvd.handles[i], 0);
+			adev->uvd.inst->filp[i] = NULL;
+			atomic_set(&adev->uvd.inst->handles[i], 0);
 		}
 	}
 }
@@ -696,13 +696,13 @@ static int amdgpu_uvd_cs_msg(struct amdgpu_uvd_cs_ctx *ctx,
 
 		/* try to alloc a new handle */
 		for (i = 0; i < adev->uvd.max_handles; ++i) {
-			if (atomic_read(&adev->uvd.handles[i]) == handle) {
+			if (atomic_read(&adev->uvd.inst->handles[i]) == handle) {
 				DRM_ERROR("Handle 0x%x already in use!\n", handle);
 				return -EINVAL;
 			}
 
-			if (!atomic_cmpxchg(&adev->uvd.handles[i], 0, handle)) {
-				adev->uvd.filp[i] = ctx->parser->filp;
+			if (!atomic_cmpxchg(&adev->uvd.inst->handles[i], 0, handle)) {
+				adev->uvd.inst->filp[i] = ctx->parser->filp;
 				return 0;
 			}
 		}
@@ -719,8 +719,8 @@ static int amdgpu_uvd_cs_msg(struct amdgpu_uvd_cs_ctx *ctx,
 
 		/* validate the handle */
 		for (i = 0; i < adev->uvd.max_handles; ++i) {
-			if (atomic_read(&adev->uvd.handles[i]) == handle) {
-				if (adev->uvd.filp[i] != ctx->parser->filp) {
+			if (atomic_read(&adev->uvd.inst->handles[i]) == handle) {
+				if (adev->uvd.inst->filp[i] != ctx->parser->filp) {
 					DRM_ERROR("UVD handle collision detected!\n");
 					return -EINVAL;
 				}
@@ -734,7 +734,7 @@ static int amdgpu_uvd_cs_msg(struct amdgpu_uvd_cs_ctx *ctx,
 	case 2:
 		/* it's a destroy msg, free the handle */
 		for (i = 0; i < adev->uvd.max_handles; ++i)
-			atomic_cmpxchg(&adev->uvd.handles[i], handle, 0);
+			atomic_cmpxchg(&adev->uvd.inst->handles[i], handle, 0);
 		amdgpu_bo_kunmap(bo);
 		return 0;
 
@@ -810,7 +810,7 @@ static int amdgpu_uvd_cs_pass2(struct amdgpu_uvd_cs_ctx *ctx)
 		}
 
 		if ((cmd == 0 || cmd == 0x3) &&
-		    (start >> 28) != (ctx->parser->adev->uvd.gpu_addr >> 28)) {
+		    (start >> 28) != (ctx->parser->adev->uvd.inst->gpu_addr >> 28)) {
 			DRM_ERROR("msg/fb buffer %LX-%LX out of 256MB segment!\n",
 				  start, end);
 			return -EINVAL;
@@ -1043,7 +1043,7 @@ static int amdgpu_uvd_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo,
 		if (r)
 			goto err_free;
 
-		r = amdgpu_job_submit(job, ring, &adev->uvd.entity,
+		r = amdgpu_job_submit(job, ring, &adev->uvd.inst->entity,
 				      AMDGPU_FENCE_OWNER_UNDEFINED, &f);
 		if (r)
 			goto err_free;
@@ -1131,8 +1131,8 @@ int amdgpu_uvd_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle,
 static void amdgpu_uvd_idle_work_handler(struct work_struct *work)
 {
 	struct amdgpu_device *adev =
-		container_of(work, struct amdgpu_device, uvd.idle_work.work);
-	unsigned fences = amdgpu_fence_count_emitted(&adev->uvd.ring);
+		container_of(work, struct amdgpu_device, uvd.inst->idle_work.work);
+	unsigned fences = amdgpu_fence_count_emitted(&adev->uvd.inst->ring);
 
 	if (fences == 0) {
 		if (adev->pm.dpm_enabled) {
@@ -1146,7 +1146,7 @@ static void amdgpu_uvd_idle_work_handler(struct work_struct *work)
 							       AMD_CG_STATE_GATE);
 		}
 	} else {
-		schedule_delayed_work(&adev->uvd.idle_work, UVD_IDLE_TIMEOUT);
+		schedule_delayed_work(&adev->uvd.inst->idle_work, UVD_IDLE_TIMEOUT);
 	}
 }
 
@@ -1158,7 +1158,7 @@ void amdgpu_uvd_ring_begin_use(struct amdgpu_ring *ring)
 	if (amdgpu_sriov_vf(adev))
 		return;
 
-	set_clocks = !cancel_delayed_work_sync(&adev->uvd.idle_work);
+	set_clocks = !cancel_delayed_work_sync(&adev->uvd.inst->idle_work);
 	if (set_clocks) {
 		if (adev->pm.dpm_enabled) {
 			amdgpu_dpm_enable_uvd(adev, true);
@@ -1175,7 +1175,7 @@ void amdgpu_uvd_ring_begin_use(struct amdgpu_ring *ring)
 void amdgpu_uvd_ring_end_use(struct amdgpu_ring *ring)
 {
 	if (!amdgpu_sriov_vf(ring->adev))
-		schedule_delayed_work(&ring->adev->uvd.idle_work, UVD_IDLE_TIMEOUT);
+		schedule_delayed_work(&ring->adev->uvd.inst->idle_work, UVD_IDLE_TIMEOUT);
 }
 
 /**
@@ -1209,7 +1209,7 @@ int amdgpu_uvd_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 	} else if (r < 0) {
 		DRM_ERROR("amdgpu: fence wait failed (%ld).\n", r);
 	} else {
-		DRM_DEBUG("ib test on ring %d succeeded\n",  ring->idx);
+		DRM_DEBUG("ib test on ring %d succeeded\n", ring->idx);
 		r = 0;
 	}
 
@@ -1237,7 +1237,7 @@ uint32_t amdgpu_uvd_used_handles(struct amdgpu_device *adev)
 		 * necessarily linear. So we need to count
 		 * all non-zero handles.
 		 */
-		if (atomic_read(&adev->uvd.handles[i]))
+		if (atomic_read(&adev->uvd.inst->handles[i]))
 			used_handles++;
 	}
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h
index 32ea20b..b1579fb 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h
@@ -31,30 +31,37 @@
 #define AMDGPU_UVD_SESSION_SIZE		(50*1024)
 #define AMDGPU_UVD_FIRMWARE_OFFSET	256
 
+#define AMDGPU_MAX_UVD_INSTANCES			2
+
 #define AMDGPU_UVD_FIRMWARE_SIZE(adev)    \
 	(AMDGPU_GPU_PAGE_ALIGN(le32_to_cpu(((const struct common_firmware_header *)(adev)->uvd.fw->data)->ucode_size_bytes) + \
 			       8) - AMDGPU_UVD_FIRMWARE_OFFSET)
 
-struct amdgpu_uvd {
+struct amdgpu_uvd_inst {
 	struct amdgpu_bo	*vcpu_bo;
 	void			*cpu_addr;
 	uint64_t		gpu_addr;
-	unsigned		fw_version;
 	void			*saved_bo;
-	unsigned		max_handles;
 	atomic_t		handles[AMDGPU_MAX_UVD_HANDLES];
 	struct drm_file		*filp[AMDGPU_MAX_UVD_HANDLES];
 	struct delayed_work	idle_work;
-	const struct firmware	*fw;	/* UVD firmware */
 	struct amdgpu_ring	ring;
 	struct amdgpu_ring	ring_enc[AMDGPU_MAX_UVD_ENC_RINGS];
 	struct amdgpu_irq_src	irq;
-	bool			address_64_bit;
-	bool			use_ctx_buf;
 	struct drm_sched_entity entity;
 	struct drm_sched_entity entity_enc;
 	uint32_t                srbm_soft_reset;
+};
+
+struct amdgpu_uvd {
+	const struct firmware	*fw;	/* UVD firmware */
+	unsigned		fw_version;
+	unsigned		max_handles;
 	unsigned		num_enc_rings;
+	uint8_t		num_uvd_inst;
+	bool			address_64_bit;
+	bool			use_ctx_buf;
+	struct amdgpu_uvd_inst		inst[AMDGPU_MAX_UVD_INSTANCES];
 };
 
 int amdgpu_uvd_sw_init(struct amdgpu_device *adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
index 87cbb14..5f22135d 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
@@ -93,6 +93,7 @@ static void uvd_v4_2_ring_set_wptr(struct amdgpu_ring *ring)
 static int uvd_v4_2_early_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+	adev->uvd.num_uvd_inst = 1;
 
 	uvd_v4_2_set_ring_funcs(adev);
 	uvd_v4_2_set_irq_funcs(adev);
@@ -107,7 +108,7 @@ static int uvd_v4_2_sw_init(void *handle)
 	int r;
 
 	/* UVD TRAP */
-	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 124, &adev->uvd.irq);
+	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 124, &adev->uvd.inst->irq);
 	if (r)
 		return r;
 
@@ -119,9 +120,9 @@ static int uvd_v4_2_sw_init(void *handle)
 	if (r)
 		return r;
 
-	ring = &adev->uvd.ring;
+	ring = &adev->uvd.inst->ring;
 	sprintf(ring->name, "uvd");
-	r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.irq, 0);
+	r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.inst->irq, 0);
 
 	return r;
 }
@@ -150,7 +151,7 @@ static void uvd_v4_2_enable_mgcg(struct amdgpu_device *adev,
 static int uvd_v4_2_hw_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
 	uint32_t tmp;
 	int r;
 
@@ -208,7 +209,7 @@ done:
 static int uvd_v4_2_hw_fini(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
 
 	if (RREG32(mmUVD_STATUS) != 0)
 		uvd_v4_2_stop(adev);
@@ -251,7 +252,7 @@ static int uvd_v4_2_resume(void *handle)
  */
 static int uvd_v4_2_start(struct amdgpu_device *adev)
 {
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
 	uint32_t rb_bufsz;
 	int i, j, r;
 	u32 tmp;
@@ -536,7 +537,7 @@ static void uvd_v4_2_mc_resume(struct amdgpu_device *adev)
 	uint32_t size;
 
 	/* programm the VCPU memory controller bits 0-27 */
-	addr = (adev->uvd.gpu_addr + AMDGPU_UVD_FIRMWARE_OFFSET) >> 3;
+	addr = (adev->uvd.inst->gpu_addr + AMDGPU_UVD_FIRMWARE_OFFSET) >> 3;
 	size = AMDGPU_UVD_FIRMWARE_SIZE(adev) >> 3;
 	WREG32(mmUVD_VCPU_CACHE_OFFSET0, addr);
 	WREG32(mmUVD_VCPU_CACHE_SIZE0, size);
@@ -553,11 +554,11 @@ static void uvd_v4_2_mc_resume(struct amdgpu_device *adev)
 	WREG32(mmUVD_VCPU_CACHE_SIZE2, size);
 
 	/* bits 28-31 */
-	addr = (adev->uvd.gpu_addr >> 28) & 0xF;
+	addr = (adev->uvd.inst->gpu_addr >> 28) & 0xF;
 	WREG32(mmUVD_LMI_ADDR_EXT, (addr << 12) | (addr << 0));
 
 	/* bits 32-39 */
-	addr = (adev->uvd.gpu_addr >> 32) & 0xFF;
+	addr = (adev->uvd.inst->gpu_addr >> 32) & 0xFF;
 	WREG32(mmUVD_LMI_EXT40_ADDR, addr | (0x9 << 16) | (0x1 << 31));
 
 	WREG32(mmUVD_UDEC_ADDR_CONFIG, adev->gfx.config.gb_addr_config);
@@ -664,7 +665,7 @@ static int uvd_v4_2_process_interrupt(struct amdgpu_device *adev,
 				      struct amdgpu_iv_entry *entry)
 {
 	DRM_DEBUG("IH: UVD TRAP\n");
-	amdgpu_fence_process(&adev->uvd.ring);
+	amdgpu_fence_process(&adev->uvd.inst->ring);
 	return 0;
 }
 
@@ -753,7 +754,7 @@ static const struct amdgpu_ring_funcs uvd_v4_2_ring_funcs = {
 
 static void uvd_v4_2_set_ring_funcs(struct amdgpu_device *adev)
 {
-	adev->uvd.ring.funcs = &uvd_v4_2_ring_funcs;
+	adev->uvd.inst->ring.funcs = &uvd_v4_2_ring_funcs;
 }
 
 static const struct amdgpu_irq_src_funcs uvd_v4_2_irq_funcs = {
@@ -763,8 +764,8 @@ static const struct amdgpu_irq_src_funcs uvd_v4_2_irq_funcs = {
 
 static void uvd_v4_2_set_irq_funcs(struct amdgpu_device *adev)
 {
-	adev->uvd.irq.num_types = 1;
-	adev->uvd.irq.funcs = &uvd_v4_2_irq_funcs;
+	adev->uvd.inst->irq.num_types = 1;
+	adev->uvd.inst->irq.funcs = &uvd_v4_2_irq_funcs;
 }
 
 const struct amdgpu_ip_block_version uvd_v4_2_ip_block =
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
index 6445d55..f5d074a 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
@@ -89,6 +89,7 @@ static void uvd_v5_0_ring_set_wptr(struct amdgpu_ring *ring)
 static int uvd_v5_0_early_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+	adev->uvd.num_uvd_inst = 1;
 
 	uvd_v5_0_set_ring_funcs(adev);
 	uvd_v5_0_set_irq_funcs(adev);
@@ -103,7 +104,7 @@ static int uvd_v5_0_sw_init(void *handle)
 	int r;
 
 	/* UVD TRAP */
-	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 124, &adev->uvd.irq);
+	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 124, &adev->uvd.inst->irq);
 	if (r)
 		return r;
 
@@ -115,9 +116,9 @@ static int uvd_v5_0_sw_init(void *handle)
 	if (r)
 		return r;
 
-	ring = &adev->uvd.ring;
+	ring = &adev->uvd.inst->ring;
 	sprintf(ring->name, "uvd");
-	r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.irq, 0);
+	r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.inst->irq, 0);
 
 	return r;
 }
@@ -144,7 +145,7 @@ static int uvd_v5_0_sw_fini(void *handle)
 static int uvd_v5_0_hw_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
 	uint32_t tmp;
 	int r;
 
@@ -204,7 +205,7 @@ done:
 static int uvd_v5_0_hw_fini(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
 
 	if (RREG32(mmUVD_STATUS) != 0)
 		uvd_v5_0_stop(adev);
@@ -253,9 +254,9 @@ static void uvd_v5_0_mc_resume(struct amdgpu_device *adev)
 
 	/* programm memory controller bits 0-27 */
 	WREG32(mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW,
-			lower_32_bits(adev->uvd.gpu_addr));
+			lower_32_bits(adev->uvd.inst->gpu_addr));
 	WREG32(mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH,
-			upper_32_bits(adev->uvd.gpu_addr));
+			upper_32_bits(adev->uvd.inst->gpu_addr));
 
 	offset = AMDGPU_UVD_FIRMWARE_OFFSET;
 	size = AMDGPU_UVD_FIRMWARE_SIZE(adev);
@@ -287,7 +288,7 @@ static void uvd_v5_0_mc_resume(struct amdgpu_device *adev)
  */
 static int uvd_v5_0_start(struct amdgpu_device *adev)
 {
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
 	uint32_t rb_bufsz, tmp;
 	uint32_t lmi_swap_cntl;
 	uint32_t mp_swap_cntl;
@@ -586,7 +587,7 @@ static int uvd_v5_0_process_interrupt(struct amdgpu_device *adev,
 				      struct amdgpu_iv_entry *entry)
 {
 	DRM_DEBUG("IH: UVD TRAP\n");
-	amdgpu_fence_process(&adev->uvd.ring);
+	amdgpu_fence_process(&adev->uvd.inst->ring);
 	return 0;
 }
 
@@ -861,7 +862,7 @@ static const struct amdgpu_ring_funcs uvd_v5_0_ring_funcs = {
 
 static void uvd_v5_0_set_ring_funcs(struct amdgpu_device *adev)
 {
-	adev->uvd.ring.funcs = &uvd_v5_0_ring_funcs;
+	adev->uvd.inst->ring.funcs = &uvd_v5_0_ring_funcs;
 }
 
 static const struct amdgpu_irq_src_funcs uvd_v5_0_irq_funcs = {
@@ -871,8 +872,8 @@ static const struct amdgpu_irq_src_funcs uvd_v5_0_irq_funcs = {
 
 static void uvd_v5_0_set_irq_funcs(struct amdgpu_device *adev)
 {
-	adev->uvd.irq.num_types = 1;
-	adev->uvd.irq.funcs = &uvd_v5_0_irq_funcs;
+	adev->uvd.inst->irq.num_types = 1;
+	adev->uvd.inst->irq.funcs = &uvd_v5_0_irq_funcs;
 }
 
 const struct amdgpu_ip_block_version uvd_v5_0_ip_block =
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
index ca6ab56..dc39169 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
@@ -91,7 +91,7 @@ static uint64_t uvd_v6_0_enc_ring_get_rptr(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
 
-	if (ring == &adev->uvd.ring_enc[0])
+	if (ring == &adev->uvd.inst->ring_enc[0])
 		return RREG32(mmUVD_RB_RPTR);
 	else
 		return RREG32(mmUVD_RB_RPTR2);
@@ -121,7 +121,7 @@ static uint64_t uvd_v6_0_enc_ring_get_wptr(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
 
-	if (ring == &adev->uvd.ring_enc[0])
+	if (ring == &adev->uvd.inst->ring_enc[0])
 		return RREG32(mmUVD_RB_WPTR);
 	else
 		return RREG32(mmUVD_RB_WPTR2);
@@ -152,7 +152,7 @@ static void uvd_v6_0_enc_ring_set_wptr(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
 
-	if (ring == &adev->uvd.ring_enc[0])
+	if (ring == &adev->uvd.inst->ring_enc[0])
 		WREG32(mmUVD_RB_WPTR,
 			lower_32_bits(ring->wptr));
 	else
@@ -375,6 +375,7 @@ error:
 static int uvd_v6_0_early_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+	adev->uvd.num_uvd_inst = 1;
 
 	if (!(adev->flags & AMD_IS_APU) &&
 	    (RREG32_SMC(ixCC_HARVEST_FUSES) & CC_HARVEST_FUSES__UVD_DISABLE_MASK))
@@ -399,14 +400,14 @@ static int uvd_v6_0_sw_init(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
 	/* UVD TRAP */
-	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 124, &adev->uvd.irq);
+	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 124, &adev->uvd.inst->irq);
 	if (r)
 		return r;
 
 	/* UVD ENC TRAP */
 	if (uvd_v6_0_enc_support(adev)) {
 		for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
-			r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, i + 119, &adev->uvd.irq);
+			r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, i + 119, &adev->uvd.inst->irq);
 			if (r)
 				return r;
 		}
@@ -418,17 +419,17 @@ static int uvd_v6_0_sw_init(void *handle)
 
 	if (!uvd_v6_0_enc_support(adev)) {
 		for (i = 0; i < adev->uvd.num_enc_rings; ++i)
-			adev->uvd.ring_enc[i].funcs = NULL;
+			adev->uvd.inst->ring_enc[i].funcs = NULL;
 
-		adev->uvd.irq.num_types = 1;
+		adev->uvd.inst->irq.num_types = 1;
 		adev->uvd.num_enc_rings = 0;
 
 		DRM_INFO("UVD ENC is disabled\n");
 	} else {
 		struct drm_sched_rq *rq;
-		ring = &adev->uvd.ring_enc[0];
+		ring = &adev->uvd.inst->ring_enc[0];
 		rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
-		r = drm_sched_entity_init(&ring->sched, &adev->uvd.entity_enc,
+		r = drm_sched_entity_init(&ring->sched, &adev->uvd.inst->entity_enc,
 					  rq, NULL);
 		if (r) {
 			DRM_ERROR("Failed setting up UVD ENC run queue.\n");
@@ -440,17 +441,17 @@ static int uvd_v6_0_sw_init(void *handle)
 	if (r)
 		return r;
 
-	ring = &adev->uvd.ring;
+	ring = &adev->uvd.inst->ring;
 	sprintf(ring->name, "uvd");
-	r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.irq, 0);
+	r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.inst->irq, 0);
 	if (r)
 		return r;
 
 	if (uvd_v6_0_enc_support(adev)) {
 		for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
-			ring = &adev->uvd.ring_enc[i];
+			ring = &adev->uvd.inst->ring_enc[i];
 			sprintf(ring->name, "uvd_enc%d", i);
-			r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.irq, 0);
+			r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.inst->irq, 0);
 			if (r)
 				return r;
 		}
@@ -469,10 +470,10 @@ static int uvd_v6_0_sw_fini(void *handle)
 		return r;
 
 	if (uvd_v6_0_enc_support(adev)) {
-		drm_sched_entity_fini(&adev->uvd.ring_enc[0].sched, &adev->uvd.entity_enc);
+		drm_sched_entity_fini(&adev->uvd.inst->ring_enc[0].sched, &adev->uvd.inst->entity_enc);
 
 		for (i = 0; i < adev->uvd.num_enc_rings; ++i)
-			amdgpu_ring_fini(&adev->uvd.ring_enc[i]);
+			amdgpu_ring_fini(&adev->uvd.inst->ring_enc[i]);
 	}
 
 	return amdgpu_uvd_sw_fini(adev);
@@ -488,7 +489,7 @@ static int uvd_v6_0_sw_fini(void *handle)
 static int uvd_v6_0_hw_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
 	uint32_t tmp;
 	int i, r;
 
@@ -532,7 +533,7 @@ static int uvd_v6_0_hw_init(void *handle)
 
 	if (uvd_v6_0_enc_support(adev)) {
 		for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
-			ring = &adev->uvd.ring_enc[i];
+			ring = &adev->uvd.inst->ring_enc[i];
 			ring->ready = true;
 			r = amdgpu_ring_test_ring(ring);
 			if (r) {
@@ -563,7 +564,7 @@ done:
 static int uvd_v6_0_hw_fini(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
 
 	if (RREG32(mmUVD_STATUS) != 0)
 		uvd_v6_0_stop(adev);
@@ -611,9 +612,9 @@ static void uvd_v6_0_mc_resume(struct amdgpu_device *adev)
 
 	/* programm memory controller bits 0-27 */
 	WREG32(mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW,
-			lower_32_bits(adev->uvd.gpu_addr));
+			lower_32_bits(adev->uvd.inst->gpu_addr));
 	WREG32(mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH,
-			upper_32_bits(adev->uvd.gpu_addr));
+			upper_32_bits(adev->uvd.inst->gpu_addr));
 
 	offset = AMDGPU_UVD_FIRMWARE_OFFSET;
 	size = AMDGPU_UVD_FIRMWARE_SIZE(adev);
@@ -726,7 +727,7 @@ static void cz_set_uvd_clock_gating_branches(struct amdgpu_device *adev,
  */
 static int uvd_v6_0_start(struct amdgpu_device *adev)
 {
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
 	uint32_t rb_bufsz, tmp;
 	uint32_t lmi_swap_cntl;
 	uint32_t mp_swap_cntl;
@@ -866,14 +867,14 @@ static int uvd_v6_0_start(struct amdgpu_device *adev)
 	WREG32_FIELD(UVD_RBC_RB_CNTL, RB_NO_FETCH, 0);
 
 	if (uvd_v6_0_enc_support(adev)) {
-		ring = &adev->uvd.ring_enc[0];
+		ring = &adev->uvd.inst->ring_enc[0];
 		WREG32(mmUVD_RB_RPTR, lower_32_bits(ring->wptr));
 		WREG32(mmUVD_RB_WPTR, lower_32_bits(ring->wptr));
 		WREG32(mmUVD_RB_BASE_LO, ring->gpu_addr);
 		WREG32(mmUVD_RB_BASE_HI, upper_32_bits(ring->gpu_addr));
 		WREG32(mmUVD_RB_SIZE, ring->ring_size / 4);
 
-		ring = &adev->uvd.ring_enc[1];
+		ring = &adev->uvd.inst->ring_enc[1];
 		WREG32(mmUVD_RB_RPTR2, lower_32_bits(ring->wptr));
 		WREG32(mmUVD_RB_WPTR2, lower_32_bits(ring->wptr));
 		WREG32(mmUVD_RB_BASE_LO2, ring->gpu_addr);
@@ -1158,10 +1159,10 @@ static bool uvd_v6_0_check_soft_reset(void *handle)
 		srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset, SRBM_SOFT_RESET, SOFT_RESET_UVD, 1);
 
 	if (srbm_soft_reset) {
-		adev->uvd.srbm_soft_reset = srbm_soft_reset;
+		adev->uvd.inst->srbm_soft_reset = srbm_soft_reset;
 		return true;
 	} else {
-		adev->uvd.srbm_soft_reset = 0;
+		adev->uvd.inst->srbm_soft_reset = 0;
 		return false;
 	}
 }
@@ -1170,7 +1171,7 @@ static int uvd_v6_0_pre_soft_reset(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	if (!adev->uvd.srbm_soft_reset)
+	if (!adev->uvd.inst->srbm_soft_reset)
 		return 0;
 
 	uvd_v6_0_stop(adev);
@@ -1182,9 +1183,9 @@ static int uvd_v6_0_soft_reset(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 	u32 srbm_soft_reset;
 
-	if (!adev->uvd.srbm_soft_reset)
+	if (!adev->uvd.inst->srbm_soft_reset)
 		return 0;
-	srbm_soft_reset = adev->uvd.srbm_soft_reset;
+	srbm_soft_reset = adev->uvd.inst->srbm_soft_reset;
 
 	if (srbm_soft_reset) {
 		u32 tmp;
@@ -1212,7 +1213,7 @@ static int uvd_v6_0_post_soft_reset(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	if (!adev->uvd.srbm_soft_reset)
+	if (!adev->uvd.inst->srbm_soft_reset)
 		return 0;
 
 	mdelay(5);
@@ -1238,17 +1239,17 @@ static int uvd_v6_0_process_interrupt(struct amdgpu_device *adev,
 
 	switch (entry->src_id) {
 	case 124:
-		amdgpu_fence_process(&adev->uvd.ring);
+		amdgpu_fence_process(&adev->uvd.inst->ring);
 		break;
 	case 119:
 		if (likely(uvd_v6_0_enc_support(adev)))
-			amdgpu_fence_process(&adev->uvd.ring_enc[0]);
+			amdgpu_fence_process(&adev->uvd.inst->ring_enc[0]);
 		else
 			int_handled = false;
 		break;
 	case 120:
 		if (likely(uvd_v6_0_enc_support(adev)))
-			amdgpu_fence_process(&adev->uvd.ring_enc[1]);
+			amdgpu_fence_process(&adev->uvd.inst->ring_enc[1]);
 		else
 			int_handled = false;
 		break;
@@ -1612,10 +1613,10 @@ static const struct amdgpu_ring_funcs uvd_v6_0_enc_ring_vm_funcs = {
 static void uvd_v6_0_set_ring_funcs(struct amdgpu_device *adev)
 {
 	if (adev->asic_type >= CHIP_POLARIS10) {
-		adev->uvd.ring.funcs = &uvd_v6_0_ring_vm_funcs;
+		adev->uvd.inst->ring.funcs = &uvd_v6_0_ring_vm_funcs;
 		DRM_INFO("UVD is enabled in VM mode\n");
 	} else {
-		adev->uvd.ring.funcs = &uvd_v6_0_ring_phys_funcs;
+		adev->uvd.inst->ring.funcs = &uvd_v6_0_ring_phys_funcs;
 		DRM_INFO("UVD is enabled in physical mode\n");
 	}
 }
@@ -1625,7 +1626,7 @@ static void uvd_v6_0_set_enc_ring_funcs(struct amdgpu_device *adev)
 	int i;
 
 	for (i = 0; i < adev->uvd.num_enc_rings; ++i)
-		adev->uvd.ring_enc[i].funcs = &uvd_v6_0_enc_ring_vm_funcs;
+		adev->uvd.inst->ring_enc[i].funcs = &uvd_v6_0_enc_ring_vm_funcs;
 
 	DRM_INFO("UVD ENC is enabled in VM mode\n");
 }
@@ -1638,11 +1639,11 @@ static const struct amdgpu_irq_src_funcs uvd_v6_0_irq_funcs = {
 static void uvd_v6_0_set_irq_funcs(struct amdgpu_device *adev)
 {
 	if (uvd_v6_0_enc_support(adev))
-		adev->uvd.irq.num_types = adev->uvd.num_enc_rings + 1;
+		adev->uvd.inst->irq.num_types = adev->uvd.num_enc_rings + 1;
 	else
-		adev->uvd.irq.num_types = 1;
+		adev->uvd.inst->irq.num_types = 1;
 
-	adev->uvd.irq.funcs = &uvd_v6_0_irq_funcs;
+	adev->uvd.inst->irq.funcs = &uvd_v6_0_irq_funcs;
 }
 
 const struct amdgpu_ip_block_version uvd_v6_0_ip_block =
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
index 0ca63d5..66d4bea 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
@@ -72,7 +72,7 @@ static uint64_t uvd_v7_0_enc_ring_get_rptr(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
 
-	if (ring == &adev->uvd.ring_enc[0])
+	if (ring == &adev->uvd.inst->ring_enc[0])
 		return RREG32_SOC15(UVD, 0, mmUVD_RB_RPTR);
 	else
 		return RREG32_SOC15(UVD, 0, mmUVD_RB_RPTR2);
@@ -106,7 +106,7 @@ static uint64_t uvd_v7_0_enc_ring_get_wptr(struct amdgpu_ring *ring)
 	if (ring->use_doorbell)
 		return adev->wb.wb[ring->wptr_offs];
 
-	if (ring == &adev->uvd.ring_enc[0])
+	if (ring == &adev->uvd.inst->ring_enc[0])
 		return RREG32_SOC15(UVD, 0, mmUVD_RB_WPTR);
 	else
 		return RREG32_SOC15(UVD, 0, mmUVD_RB_WPTR2);
@@ -144,7 +144,7 @@ static void uvd_v7_0_enc_ring_set_wptr(struct amdgpu_ring *ring)
 		return;
 	}
 
-	if (ring == &adev->uvd.ring_enc[0])
+	if (ring == &adev->uvd.inst->ring_enc[0])
 		WREG32_SOC15(UVD, 0, mmUVD_RB_WPTR,
 			lower_32_bits(ring->wptr));
 	else
@@ -170,8 +170,8 @@ static int uvd_v7_0_enc_ring_test_ring(struct amdgpu_ring *ring)
 
 	r = amdgpu_ring_alloc(ring, 16);
 	if (r) {
-		DRM_ERROR("amdgpu: uvd enc failed to lock ring %d (%d).\n",
-			  ring->idx, r);
+		DRM_ERROR("amdgpu: uvd enc failed to lock (%d)ring %d (%d).\n",
+			  ring->me, ring->idx, r);
 		return r;
 	}
 	amdgpu_ring_write(ring, HEVC_ENC_CMD_END);
@@ -184,11 +184,11 @@ static int uvd_v7_0_enc_ring_test_ring(struct amdgpu_ring *ring)
 	}
 
 	if (i < adev->usec_timeout) {
-		DRM_DEBUG("ring test on %d succeeded in %d usecs\n",
-			 ring->idx, i);
+		DRM_DEBUG("(%d)ring test on %d succeeded in %d usecs\n",
+			 ring->me, ring->idx, i);
 	} else {
-		DRM_ERROR("amdgpu: ring %d test failed\n",
-			  ring->idx);
+		DRM_ERROR("amdgpu: (%d)ring %d test failed\n",
+			  ring->me, ring->idx);
 		r = -ETIMEDOUT;
 	}
 
@@ -342,24 +342,24 @@ static int uvd_v7_0_enc_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 
 	r = uvd_v7_0_enc_get_create_msg(ring, 1, NULL);
 	if (r) {
-		DRM_ERROR("amdgpu: failed to get create msg (%ld).\n", r);
+		DRM_ERROR("amdgpu: (%d)failed to get create msg (%ld).\n", ring->me, r);
 		goto error;
 	}
 
 	r = uvd_v7_0_enc_get_destroy_msg(ring, 1, true, &fence);
 	if (r) {
-		DRM_ERROR("amdgpu: failed to get destroy ib (%ld).\n", r);
+		DRM_ERROR("amdgpu: (%d)failed to get destroy ib (%ld).\n", ring->me, r);
 		goto error;
 	}
 
 	r = dma_fence_wait_timeout(fence, false, timeout);
 	if (r == 0) {
-		DRM_ERROR("amdgpu: IB test timed out.\n");
+		DRM_ERROR("amdgpu: (%d)IB test timed out.\n", ring->me);
 		r = -ETIMEDOUT;
 	} else if (r < 0) {
-		DRM_ERROR("amdgpu: fence wait failed (%ld).\n", r);
+		DRM_ERROR("amdgpu: (%d)fence wait failed (%ld).\n", ring->me, r);
 	} else {
-		DRM_DEBUG("ib test on ring %d succeeded\n", ring->idx);
+		DRM_DEBUG("ib test on (%d)ring %d succeeded\n", ring->me, ring->idx);
 		r = 0;
 	}
 error:
@@ -370,6 +370,7 @@ error:
 static int uvd_v7_0_early_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+	adev->uvd.num_uvd_inst = 1;
 
 	if (amdgpu_sriov_vf(adev))
 		adev->uvd.num_enc_rings = 1;
@@ -390,13 +391,13 @@ static int uvd_v7_0_sw_init(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
 	/* UVD TRAP */
-	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_UVD, 124, &adev->uvd.irq);
+	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_UVD, 124, &adev->uvd.inst->irq);
 	if (r)
 		return r;
 
 	/* UVD ENC TRAP */
 	for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
-		r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_UVD, i + 119, &adev->uvd.irq);
+		r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_UVD, i + 119, &adev->uvd.inst->irq);
 		if (r)
 			return r;
 	}
@@ -415,9 +416,9 @@ static int uvd_v7_0_sw_init(void *handle)
 		DRM_INFO("PSP loading UVD firmware\n");
 	}
 
-	ring = &adev->uvd.ring_enc[0];
+	ring = &adev->uvd.inst->ring_enc[0];
 	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
-	r = drm_sched_entity_init(&ring->sched, &adev->uvd.entity_enc,
+	r = drm_sched_entity_init(&ring->sched, &adev->uvd.inst->entity_enc,
 				  rq, NULL);
 	if (r) {
 		DRM_ERROR("Failed setting up UVD ENC run queue.\n");
@@ -428,15 +429,15 @@ static int uvd_v7_0_sw_init(void *handle)
 	if (r)
 		return r;
 	if (!amdgpu_sriov_vf(adev)) {
-		ring = &adev->uvd.ring;
+		ring = &adev->uvd.inst->ring;
 		sprintf(ring->name, "uvd");
-		r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.irq, 0);
+		r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.inst->irq, 0);
 		if (r)
 			return r;
 	}
 
 	for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
-		ring = &adev->uvd.ring_enc[i];
+		ring = &adev->uvd.inst->ring_enc[i];
 		sprintf(ring->name, "uvd_enc%d", i);
 		if (amdgpu_sriov_vf(adev)) {
 			ring->use_doorbell = true;
@@ -449,7 +450,7 @@ static int uvd_v7_0_sw_init(void *handle)
 			else
 				ring->doorbell_index = AMDGPU_DOORBELL64_UVD_RING2_3 * 2 + 1;
 		}
-		r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.irq, 0);
+		r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.inst->irq, 0);
 		if (r)
 			return r;
 	}
@@ -472,10 +473,10 @@ static int uvd_v7_0_sw_fini(void *handle)
 	if (r)
 		return r;
 
-	drm_sched_entity_fini(&adev->uvd.ring_enc[0].sched, &adev->uvd.entity_enc);
+	drm_sched_entity_fini(&adev->uvd.inst->ring_enc[0].sched, &adev->uvd.inst->entity_enc);
 
 	for (i = 0; i < adev->uvd.num_enc_rings; ++i)
-		amdgpu_ring_fini(&adev->uvd.ring_enc[i]);
+		amdgpu_ring_fini(&adev->uvd.inst->ring_enc[i]);
 
 	return amdgpu_uvd_sw_fini(adev);
 }
@@ -490,7 +491,7 @@ static int uvd_v7_0_sw_fini(void *handle)
 static int uvd_v7_0_hw_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
 	uint32_t tmp;
 	int i, r;
 
@@ -543,7 +544,7 @@ static int uvd_v7_0_hw_init(void *handle)
 	}
 
 	for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
-		ring = &adev->uvd.ring_enc[i];
+		ring = &adev->uvd.inst->ring_enc[i];
 		ring->ready = true;
 		r = amdgpu_ring_test_ring(ring);
 		if (r) {
@@ -569,7 +570,7 @@ done:
 static int uvd_v7_0_hw_fini(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
 
 	if (!amdgpu_sriov_vf(adev))
 		uvd_v7_0_stop(adev);
@@ -627,9 +628,9 @@ static void uvd_v7_0_mc_resume(struct amdgpu_device *adev)
 		offset = 0;
 	} else {
 		WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW,
-			lower_32_bits(adev->uvd.gpu_addr));
+			lower_32_bits(adev->uvd.inst->gpu_addr));
 		WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH,
-			upper_32_bits(adev->uvd.gpu_addr));
+			upper_32_bits(adev->uvd.inst->gpu_addr));
 		offset = size;
 	}
 
@@ -638,16 +639,16 @@ static void uvd_v7_0_mc_resume(struct amdgpu_device *adev)
 	WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_SIZE0, size);
 
 	WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW,
-			lower_32_bits(adev->uvd.gpu_addr + offset));
+			lower_32_bits(adev->uvd.inst->gpu_addr + offset));
 	WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH,
-			upper_32_bits(adev->uvd.gpu_addr + offset));
+			upper_32_bits(adev->uvd.inst->gpu_addr + offset));
 	WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_OFFSET1, (1 << 21));
 	WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_SIZE1, AMDGPU_UVD_HEAP_SIZE);
 
 	WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_LOW,
-			lower_32_bits(adev->uvd.gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE));
+			lower_32_bits(adev->uvd.inst->gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE));
 	WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_HIGH,
-			upper_32_bits(adev->uvd.gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE));
+			upper_32_bits(adev->uvd.inst->gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE));
 	WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_OFFSET2, (2 << 21));
 	WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_SIZE2,
 			AMDGPU_UVD_STACK_SIZE + (AMDGPU_UVD_SESSION_SIZE * 40));
@@ -688,10 +689,10 @@ static int uvd_v7_0_mmsch_start(struct amdgpu_device *adev,
 	/* 4, set resp to zero */
 	WREG32_SOC15(VCE, 0, mmVCE_MMSCH_VF_MAILBOX_RESP, 0);
 
-	WDOORBELL32(adev->uvd.ring_enc[0].doorbell_index, 0);
-	adev->wb.wb[adev->uvd.ring_enc[0].wptr_offs] = 0;
-	adev->uvd.ring_enc[0].wptr = 0;
-	adev->uvd.ring_enc[0].wptr_old = 0;
+	WDOORBELL32(adev->uvd.inst->ring_enc[0].doorbell_index, 0);
+	adev->wb.wb[adev->uvd.inst->ring_enc[0].wptr_offs] = 0;
+	adev->uvd.inst->ring_enc[0].wptr = 0;
+	adev->uvd.inst->ring_enc[0].wptr_old = 0;
 
 	/* 5, kick off the initialization and wait until VCE_MMSCH_VF_MAILBOX_RESP becomes non-zero */
 	WREG32_SOC15(VCE, 0, mmVCE_MMSCH_VF_MAILBOX_HOST, 0x10000001);
@@ -742,7 +743,7 @@ static int uvd_v7_0_sriov_start(struct amdgpu_device *adev)
 
 		init_table += header->uvd_table_offset;
 
-		ring = &adev->uvd.ring;
+		ring = &adev->uvd.inst->ring;
 		ring->wptr = 0;
 		size = AMDGPU_GPU_PAGE_ALIGN(adev->uvd.fw->size + 4);
 
@@ -757,9 +758,9 @@ static int uvd_v7_0_sriov_start(struct amdgpu_device *adev)
 			offset = 0;
 		} else {
 			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW),
-						    lower_32_bits(adev->uvd.gpu_addr));
+						    lower_32_bits(adev->uvd.inst->gpu_addr));
 			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH),
-						    upper_32_bits(adev->uvd.gpu_addr));
+						    upper_32_bits(adev->uvd.inst->gpu_addr));
 			offset = size;
 		}
 
@@ -768,16 +769,16 @@ static int uvd_v7_0_sriov_start(struct amdgpu_device *adev)
 		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_SIZE0), size);
 
 		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW),
-					    lower_32_bits(adev->uvd.gpu_addr + offset));
+					    lower_32_bits(adev->uvd.inst->gpu_addr + offset));
 		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH),
-					    upper_32_bits(adev->uvd.gpu_addr + offset));
+					    upper_32_bits(adev->uvd.inst->gpu_addr + offset));
 		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_OFFSET1), (1 << 21));
 		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_SIZE1), AMDGPU_UVD_HEAP_SIZE);
 
 		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_LOW),
-					    lower_32_bits(adev->uvd.gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE));
+					    lower_32_bits(adev->uvd.inst->gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE));
 		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_HIGH),
-					    upper_32_bits(adev->uvd.gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE));
+					    upper_32_bits(adev->uvd.inst->gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE));
 		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_OFFSET2), (2 << 21));
 		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_SIZE2),
 					    AMDGPU_UVD_STACK_SIZE + (AMDGPU_UVD_SESSION_SIZE * 40));
@@ -841,7 +842,7 @@ static int uvd_v7_0_sriov_start(struct amdgpu_device *adev)
 		tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_FETCH, 1);
 		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RBC_RB_CNTL), tmp);
 
-		ring = &adev->uvd.ring_enc[0];
+		ring = &adev->uvd.inst->ring_enc[0];
 		ring->wptr = 0;
 		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_BASE_LO), ring->gpu_addr);
 		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_BASE_HI), upper_32_bits(ring->gpu_addr));
@@ -874,7 +875,7 @@ static int uvd_v7_0_sriov_start(struct amdgpu_device *adev)
  */
 static int uvd_v7_0_start(struct amdgpu_device *adev)
 {
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
 	uint32_t rb_bufsz, tmp;
 	uint32_t lmi_swap_cntl;
 	uint32_t mp_swap_cntl;
@@ -1027,14 +1028,14 @@ static int uvd_v7_0_start(struct amdgpu_device *adev)
 	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_RBC_RB_CNTL), 0,
 			~UVD_RBC_RB_CNTL__RB_NO_FETCH_MASK);
 
-	ring = &adev->uvd.ring_enc[0];
+	ring = &adev->uvd.inst->ring_enc[0];
 	WREG32_SOC15(UVD, 0, mmUVD_RB_RPTR, lower_32_bits(ring->wptr));
 	WREG32_SOC15(UVD, 0, mmUVD_RB_WPTR, lower_32_bits(ring->wptr));
 	WREG32_SOC15(UVD, 0, mmUVD_RB_BASE_LO, ring->gpu_addr);
 	WREG32_SOC15(UVD, 0, mmUVD_RB_BASE_HI, upper_32_bits(ring->gpu_addr));
 	WREG32_SOC15(UVD, 0, mmUVD_RB_SIZE, ring->ring_size / 4);
 
-	ring = &adev->uvd.ring_enc[1];
+	ring = &adev->uvd.inst->ring_enc[1];
 	WREG32_SOC15(UVD, 0, mmUVD_RB_RPTR2, lower_32_bits(ring->wptr));
 	WREG32_SOC15(UVD, 0, mmUVD_RB_WPTR2, lower_32_bits(ring->wptr));
 	WREG32_SOC15(UVD, 0, mmUVD_RB_BASE_LO2, ring->gpu_addr);
@@ -1162,8 +1163,8 @@ static int uvd_v7_0_ring_test_ring(struct amdgpu_ring *ring)
 	WREG32_SOC15(UVD, 0, mmUVD_CONTEXT_ID, 0xCAFEDEAD);
 	r = amdgpu_ring_alloc(ring, 3);
 	if (r) {
-		DRM_ERROR("amdgpu: cp failed to lock ring %d (%d).\n",
-			  ring->idx, r);
+		DRM_ERROR("amdgpu: (%d)cp failed to lock ring %d (%d).\n",
+			  ring->me, ring->idx, r);
 		return r;
 	}
 	amdgpu_ring_write(ring,
@@ -1178,11 +1179,11 @@ static int uvd_v7_0_ring_test_ring(struct amdgpu_ring *ring)
 	}
 
 	if (i < adev->usec_timeout) {
-		DRM_DEBUG("ring test on %d succeeded in %d usecs\n",
-			 ring->idx, i);
+		DRM_DEBUG("(%d)ring test on %d succeeded in %d usecs\n",
+			 ring->me, ring->idx, i);
 	} else {
-		DRM_ERROR("amdgpu: ring %d test failed (0x%08X)\n",
-			  ring->idx, tmp);
+		DRM_ERROR("(%d)amdgpu: ring %d test failed (0x%08X)\n",
+			  ring->me, ring->idx, tmp);
 		r = -EINVAL;
 	}
 	return r;
@@ -1365,10 +1366,10 @@ static bool uvd_v7_0_check_soft_reset(void *handle)
 				SRBM_SOFT_RESET, SOFT_RESET_UVD, 1);
 
 	if (srbm_soft_reset) {
-		adev->uvd.srbm_soft_reset = srbm_soft_reset;
+		adev->uvd.inst->srbm_soft_reset = srbm_soft_reset;
 		return true;
 	} else {
-		adev->uvd.srbm_soft_reset = 0;
+		adev->uvd.inst->srbm_soft_reset = 0;
 		return false;
 	}
 }
@@ -1377,7 +1378,7 @@ static int uvd_v7_0_pre_soft_reset(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	if (!adev->uvd.srbm_soft_reset)
+	if (!adev->uvd.inst->srbm_soft_reset)
 		return 0;
 
 	uvd_v7_0_stop(adev);
@@ -1389,9 +1390,9 @@ static int uvd_v7_0_soft_reset(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 	u32 srbm_soft_reset;
 
-	if (!adev->uvd.srbm_soft_reset)
+	if (!adev->uvd.inst->srbm_soft_reset)
 		return 0;
-	srbm_soft_reset = adev->uvd.srbm_soft_reset;
+	srbm_soft_reset = adev->uvd.inst->srbm_soft_reset;
 
 	if (srbm_soft_reset) {
 		u32 tmp;
@@ -1419,7 +1420,7 @@ static int uvd_v7_0_post_soft_reset(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	if (!adev->uvd.srbm_soft_reset)
+	if (!adev->uvd.inst->srbm_soft_reset)
 		return 0;
 
 	mdelay(5);
@@ -1444,14 +1445,14 @@ static int uvd_v7_0_process_interrupt(struct amdgpu_device *adev,
 	DRM_DEBUG("IH: UVD TRAP\n");
 	switch (entry->src_id) {
 	case 124:
-		amdgpu_fence_process(&adev->uvd.ring);
+		amdgpu_fence_process(&adev->uvd.inst->ring);
 		break;
 	case 119:
-		amdgpu_fence_process(&adev->uvd.ring_enc[0]);
+		amdgpu_fence_process(&adev->uvd.inst->ring_enc[0]);
 		break;
 	case 120:
 		if (!amdgpu_sriov_vf(adev))
-			amdgpu_fence_process(&adev->uvd.ring_enc[1]);
+			amdgpu_fence_process(&adev->uvd.inst->ring_enc[1]);
 		break;
 	default:
 		DRM_ERROR("Unhandled interrupt: %d %d\n",
@@ -1719,7 +1720,7 @@ static const struct amdgpu_ring_funcs uvd_v7_0_enc_ring_vm_funcs = {
 
 static void uvd_v7_0_set_ring_funcs(struct amdgpu_device *adev)
 {
-	adev->uvd.ring.funcs = &uvd_v7_0_ring_vm_funcs;
+	adev->uvd.inst->ring.funcs = &uvd_v7_0_ring_vm_funcs;
 	DRM_INFO("UVD is enabled in VM mode\n");
 }
 
@@ -1728,7 +1729,7 @@ static void uvd_v7_0_set_enc_ring_funcs(struct amdgpu_device *adev)
 	int i;
 
 	for (i = 0; i < adev->uvd.num_enc_rings; ++i)
-		adev->uvd.ring_enc[i].funcs = &uvd_v7_0_enc_ring_vm_funcs;
+		adev->uvd.inst->ring_enc[i].funcs = &uvd_v7_0_enc_ring_vm_funcs;
 
 	DRM_INFO("UVD ENC is enabled in VM mode\n");
 }
@@ -1740,8 +1741,8 @@ static const struct amdgpu_irq_src_funcs uvd_v7_0_irq_funcs = {
 
 static void uvd_v7_0_set_irq_funcs(struct amdgpu_device *adev)
 {
-	adev->uvd.irq.num_types = adev->uvd.num_enc_rings + 1;
-	adev->uvd.irq.funcs = &uvd_v7_0_irq_funcs;
+	adev->uvd.inst->irq.num_types = adev->uvd.num_enc_rings + 1;
+	adev->uvd.inst->irq.funcs = &uvd_v7_0_irq_funcs;
 }
 
 const struct amdgpu_ip_block_version uvd_v7_0_ip_block =
-- 
cgit v1.1


From 10dd74eac4dba963bfa97f5092040aa75ff742d6 Mon Sep 17 00:00:00 2001
From: James Zhu <James.Zhu@amd.com>
Date: Tue, 15 May 2018 14:31:24 -0500
Subject: drm/amdgpu/vg20:Restruct uvd.inst to support multiple instances

Vega20 has dual-UVD. Need add multiple instances support for uvd.
Restruct uvd.inst, using uvd.inst[0] to replace uvd.inst->.
Repurpose amdgpu_ring::me for instance index, and initialize to 0.
There are no any logical changes here.

Signed-off-by: James Zhu <James.Zhu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c |    6 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c   |   12 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c  |    1 +
 drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c   |  229 +++----
 drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c     | 1002 +++++++++++++++--------------
 5 files changed, 660 insertions(+), 590 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
index 1070f40..39ec6b8 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
@@ -376,14 +376,14 @@ int amdgpu_fence_driver_start_ring(struct amdgpu_ring *ring,
 	struct amdgpu_device *adev = ring->adev;
 	uint64_t index;
 
-	if (ring != &adev->uvd.inst->ring) {
+	if (ring != &adev->uvd.inst[ring->me].ring) {
 		ring->fence_drv.cpu_addr = &adev->wb.wb[ring->fence_offs];
 		ring->fence_drv.gpu_addr = adev->wb.gpu_addr + (ring->fence_offs * 4);
 	} else {
 		/* put fence directly behind firmware */
 		index = ALIGN(adev->uvd.fw->size, 8);
-		ring->fence_drv.cpu_addr = adev->uvd.inst->cpu_addr + index;
-		ring->fence_drv.gpu_addr = adev->uvd.inst->gpu_addr + index;
+		ring->fence_drv.cpu_addr = adev->uvd.inst[ring->me].cpu_addr + index;
+		ring->fence_drv.gpu_addr = adev->uvd.inst[ring->me].gpu_addr + index;
 	}
 	amdgpu_fence_write(ring, atomic_read(&ring->fence_drv.last_seq));
 	amdgpu_irq_get(adev, irq_src, irq_type);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
index 5620ed2..91517b1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
@@ -286,7 +286,7 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 	struct drm_crtc *crtc;
 	uint32_t ui32 = 0;
 	uint64_t ui64 = 0;
-	int i, found;
+	int i, j, found;
 	int ui32_size = sizeof(ui32);
 
 	if (!info->return_size || !info->return_pointer)
@@ -348,7 +348,8 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 			break;
 		case AMDGPU_HW_IP_UVD:
 			type = AMD_IP_BLOCK_TYPE_UVD;
-			ring_mask = adev->uvd.inst->ring.ready ? 1 : 0;
+			for (i = 0; i < adev->uvd.num_uvd_inst; i++)
+				ring_mask |= ((adev->uvd.inst[i].ring.ready ? 1 : 0) << i);
 			ib_start_alignment = AMDGPU_GPU_PAGE_SIZE;
 			ib_size_alignment = 16;
 			break;
@@ -361,8 +362,11 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 			break;
 		case AMDGPU_HW_IP_UVD_ENC:
 			type = AMD_IP_BLOCK_TYPE_UVD;
-			for (i = 0; i < adev->uvd.num_enc_rings; i++)
-				ring_mask |= ((adev->uvd.inst->ring_enc[i].ready ? 1 : 0) << i);
+			for (i = 0; i < adev->uvd.num_uvd_inst; i++)
+				for (j = 0; j < adev->uvd.num_enc_rings; j++)
+					ring_mask |=
+					((adev->uvd.inst[i].ring_enc[j].ready ? 1 : 0) <<
+					(j + i * adev->uvd.num_enc_rings));
 			ib_start_alignment = AMDGPU_GPU_PAGE_SIZE;
 			ib_size_alignment = 1;
 			break;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
index 49cad08..c6850b6 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
@@ -362,6 +362,7 @@ void amdgpu_ring_fini(struct amdgpu_ring *ring)
 
 	dma_fence_put(ring->vmid_wait);
 	ring->vmid_wait = NULL;
+	ring->me = 0;
 
 	ring->adev->rings[ring->idx] = NULL;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
index 02683a0..e961492 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
@@ -127,7 +127,7 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
 	const char *fw_name;
 	const struct common_firmware_header *hdr;
 	unsigned version_major, version_minor, family_id;
-	int i, r;
+	int i, j, r;
 
 	INIT_DELAYED_WORK(&adev->uvd.inst->idle_work, amdgpu_uvd_idle_work_handler);
 
@@ -236,28 +236,30 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
 	if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP)
 		bo_size += AMDGPU_GPU_PAGE_ALIGN(le32_to_cpu(hdr->ucode_size_bytes) + 8);
 
-	r = amdgpu_bo_create_kernel(adev, bo_size, PAGE_SIZE,
-				    AMDGPU_GEM_DOMAIN_VRAM, &adev->uvd.inst->vcpu_bo,
-				    &adev->uvd.inst->gpu_addr, &adev->uvd.inst->cpu_addr);
-	if (r) {
-		dev_err(adev->dev, "(%d) failed to allocate UVD bo\n", r);
-		return r;
-	}
+	for (j = 0; j < adev->uvd.num_uvd_inst; j++) {
 
-	ring = &adev->uvd.inst->ring;
-	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
-	r = drm_sched_entity_init(&ring->sched, &adev->uvd.inst->entity,
-				  rq, NULL);
-	if (r != 0) {
-		DRM_ERROR("Failed setting up UVD run queue.\n");
-		return r;
-	}
+		r = amdgpu_bo_create_kernel(adev, bo_size, PAGE_SIZE,
+					    AMDGPU_GEM_DOMAIN_VRAM, &adev->uvd.inst[j].vcpu_bo,
+					    &adev->uvd.inst[j].gpu_addr, &adev->uvd.inst[j].cpu_addr);
+		if (r) {
+			dev_err(adev->dev, "(%d) failed to allocate UVD bo\n", r);
+			return r;
+		}
 
-	for (i = 0; i < adev->uvd.max_handles; ++i) {
-		atomic_set(&adev->uvd.inst->handles[i], 0);
-		adev->uvd.inst->filp[i] = NULL;
-	}
+		ring = &adev->uvd.inst[j].ring;
+		rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
+		r = drm_sched_entity_init(&ring->sched, &adev->uvd.inst[j].entity,
+					  rq, NULL);
+		if (r != 0) {
+			DRM_ERROR("Failed setting up UVD(%d) run queue.\n", j);
+			return r;
+		}
 
+		for (i = 0; i < adev->uvd.max_handles; ++i) {
+			atomic_set(&adev->uvd.inst[j].handles[i], 0);
+			adev->uvd.inst[j].filp[i] = NULL;
+		}
+	}
 	/* from uvd v5.0 HW addressing capacity increased to 64 bits */
 	if (!amdgpu_device_ip_block_version_cmp(adev, AMD_IP_BLOCK_TYPE_UVD, 5, 0))
 		adev->uvd.address_64_bit = true;
@@ -284,20 +286,22 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
 
 int amdgpu_uvd_sw_fini(struct amdgpu_device *adev)
 {
-	int i;
-	kfree(adev->uvd.inst->saved_bo);
+	int i, j;
 
-	drm_sched_entity_fini(&adev->uvd.inst->ring.sched, &adev->uvd.inst->entity);
+	for (j = 0; j < adev->uvd.num_uvd_inst; ++j) {
+		kfree(adev->uvd.inst[j].saved_bo);
 
-	amdgpu_bo_free_kernel(&adev->uvd.inst->vcpu_bo,
-			      &adev->uvd.inst->gpu_addr,
-			      (void **)&adev->uvd.inst->cpu_addr);
+		drm_sched_entity_fini(&adev->uvd.inst[j].ring.sched, &adev->uvd.inst[j].entity);
 
-	amdgpu_ring_fini(&adev->uvd.inst->ring);
+		amdgpu_bo_free_kernel(&adev->uvd.inst[j].vcpu_bo,
+				      &adev->uvd.inst[j].gpu_addr,
+				      (void **)&adev->uvd.inst[j].cpu_addr);
 
-	for (i = 0; i < AMDGPU_MAX_UVD_ENC_RINGS; ++i)
-		amdgpu_ring_fini(&adev->uvd.inst->ring_enc[i]);
+		amdgpu_ring_fini(&adev->uvd.inst[j].ring);
 
+		for (i = 0; i < AMDGPU_MAX_UVD_ENC_RINGS; ++i)
+			amdgpu_ring_fini(&adev->uvd.inst[j].ring_enc[i]);
+	}
 	release_firmware(adev->uvd.fw);
 
 	return 0;
@@ -307,32 +311,33 @@ int amdgpu_uvd_suspend(struct amdgpu_device *adev)
 {
 	unsigned size;
 	void *ptr;
-	int i;
+	int i, j;
 
-	if (adev->uvd.inst->vcpu_bo == NULL)
-		return 0;
+	for (j = 0; j < adev->uvd.num_uvd_inst; ++j) {
+		if (adev->uvd.inst[j].vcpu_bo == NULL)
+			continue;
 
-	cancel_delayed_work_sync(&adev->uvd.inst->idle_work);
+		cancel_delayed_work_sync(&adev->uvd.inst[j].idle_work);
 
-	/* only valid for physical mode */
-	if (adev->asic_type < CHIP_POLARIS10) {
-		for (i = 0; i < adev->uvd.max_handles; ++i)
-			if (atomic_read(&adev->uvd.inst->handles[i]))
-				break;
+		/* only valid for physical mode */
+		if (adev->asic_type < CHIP_POLARIS10) {
+			for (i = 0; i < adev->uvd.max_handles; ++i)
+				if (atomic_read(&adev->uvd.inst[j].handles[i]))
+					break;
 
-		if (i == adev->uvd.max_handles)
-			return 0;
-	}
-
-	size = amdgpu_bo_size(adev->uvd.inst->vcpu_bo);
-	ptr = adev->uvd.inst->cpu_addr;
+			if (i == adev->uvd.max_handles)
+				continue;
+		}
 
-	adev->uvd.inst->saved_bo = kmalloc(size, GFP_KERNEL);
-	if (!adev->uvd.inst->saved_bo)
-		return -ENOMEM;
+		size = amdgpu_bo_size(adev->uvd.inst[j].vcpu_bo);
+		ptr = adev->uvd.inst[j].cpu_addr;
 
-	memcpy_fromio(adev->uvd.inst->saved_bo, ptr, size);
+		adev->uvd.inst[j].saved_bo = kmalloc(size, GFP_KERNEL);
+		if (!adev->uvd.inst[j].saved_bo)
+			return -ENOMEM;
 
+		memcpy_fromio(adev->uvd.inst[j].saved_bo, ptr, size);
+	}
 	return 0;
 }
 
@@ -340,59 +345,65 @@ int amdgpu_uvd_resume(struct amdgpu_device *adev)
 {
 	unsigned size;
 	void *ptr;
+	int i;
 
-	if (adev->uvd.inst->vcpu_bo == NULL)
-		return -EINVAL;
+	for (i = 0; i < adev->uvd.num_uvd_inst; i++) {
+		if (adev->uvd.inst[i].vcpu_bo == NULL)
+			return -EINVAL;
 
-	size = amdgpu_bo_size(adev->uvd.inst->vcpu_bo);
-	ptr = adev->uvd.inst->cpu_addr;
+		size = amdgpu_bo_size(adev->uvd.inst[i].vcpu_bo);
+		ptr = adev->uvd.inst[i].cpu_addr;
 
-	if (adev->uvd.inst->saved_bo != NULL) {
-		memcpy_toio(ptr, adev->uvd.inst->saved_bo, size);
-		kfree(adev->uvd.inst->saved_bo);
-		adev->uvd.inst->saved_bo = NULL;
-	} else {
-		const struct common_firmware_header *hdr;
-		unsigned offset;
-
-		hdr = (const struct common_firmware_header *)adev->uvd.fw->data;
-		if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) {
-			offset = le32_to_cpu(hdr->ucode_array_offset_bytes);
-			memcpy_toio(adev->uvd.inst->cpu_addr, adev->uvd.fw->data + offset,
-				    le32_to_cpu(hdr->ucode_size_bytes));
-			size -= le32_to_cpu(hdr->ucode_size_bytes);
-			ptr += le32_to_cpu(hdr->ucode_size_bytes);
+		if (adev->uvd.inst[i].saved_bo != NULL) {
+			memcpy_toio(ptr, adev->uvd.inst[i].saved_bo, size);
+			kfree(adev->uvd.inst[i].saved_bo);
+			adev->uvd.inst[i].saved_bo = NULL;
+		} else {
+			const struct common_firmware_header *hdr;
+			unsigned offset;
+
+			hdr = (const struct common_firmware_header *)adev->uvd.fw->data;
+			if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) {
+				offset = le32_to_cpu(hdr->ucode_array_offset_bytes);
+				memcpy_toio(adev->uvd.inst[i].cpu_addr, adev->uvd.fw->data + offset,
+					    le32_to_cpu(hdr->ucode_size_bytes));
+				size -= le32_to_cpu(hdr->ucode_size_bytes);
+				ptr += le32_to_cpu(hdr->ucode_size_bytes);
+			}
+			memset_io(ptr, 0, size);
+			/* to restore uvd fence seq */
+			amdgpu_fence_driver_force_completion(&adev->uvd.inst[i].ring);
 		}
-		memset_io(ptr, 0, size);
-		/* to restore uvd fence seq */
-		amdgpu_fence_driver_force_completion(&adev->uvd.inst->ring);
 	}
-
 	return 0;
 }
 
 void amdgpu_uvd_free_handles(struct amdgpu_device *adev, struct drm_file *filp)
 {
-	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
-	int i, r;
+	struct amdgpu_ring *ring;
+	int i, j, r;
 
-	for (i = 0; i < adev->uvd.max_handles; ++i) {
-		uint32_t handle = atomic_read(&adev->uvd.inst->handles[i]);
-		if (handle != 0 && adev->uvd.inst->filp[i] == filp) {
-			struct dma_fence *fence;
-
-			r = amdgpu_uvd_get_destroy_msg(ring, handle,
-						       false, &fence);
-			if (r) {
-				DRM_ERROR("Error destroying UVD (%d)!\n", r);
-				continue;
-			}
+	for (j = 0; j < adev->uvd.num_uvd_inst; j++) {
+		ring = &adev->uvd.inst[j].ring;
 
-			dma_fence_wait(fence, false);
-			dma_fence_put(fence);
+		for (i = 0; i < adev->uvd.max_handles; ++i) {
+			uint32_t handle = atomic_read(&adev->uvd.inst[j].handles[i]);
+			if (handle != 0 && adev->uvd.inst[j].filp[i] == filp) {
+				struct dma_fence *fence;
+
+				r = amdgpu_uvd_get_destroy_msg(ring, handle,
+							       false, &fence);
+				if (r) {
+					DRM_ERROR("Error destroying UVD(%d) %d!\n", j, r);
+					continue;
+				}
 
-			adev->uvd.inst->filp[i] = NULL;
-			atomic_set(&adev->uvd.inst->handles[i], 0);
+				dma_fence_wait(fence, false);
+				dma_fence_put(fence);
+
+				adev->uvd.inst[j].filp[i] = NULL;
+				atomic_set(&adev->uvd.inst[j].handles[i], 0);
+			}
 		}
 	}
 }
@@ -667,15 +678,16 @@ static int amdgpu_uvd_cs_msg(struct amdgpu_uvd_cs_ctx *ctx,
 	void *ptr;
 	long r;
 	int i;
+	uint32_t ip_instance = ctx->parser->job->ring->me;
 
 	if (offset & 0x3F) {
-		DRM_ERROR("UVD messages must be 64 byte aligned!\n");
+		DRM_ERROR("UVD(%d) messages must be 64 byte aligned!\n", ip_instance);
 		return -EINVAL;
 	}
 
 	r = amdgpu_bo_kmap(bo, &ptr);
 	if (r) {
-		DRM_ERROR("Failed mapping the UVD message (%ld)!\n", r);
+		DRM_ERROR("Failed mapping the UVD(%d) message (%ld)!\n", ip_instance, r);
 		return r;
 	}
 
@@ -685,7 +697,7 @@ static int amdgpu_uvd_cs_msg(struct amdgpu_uvd_cs_ctx *ctx,
 	handle = msg[2];
 
 	if (handle == 0) {
-		DRM_ERROR("Invalid UVD handle!\n");
+		DRM_ERROR("Invalid UVD(%d) handle!\n", ip_instance);
 		return -EINVAL;
 	}
 
@@ -696,18 +708,18 @@ static int amdgpu_uvd_cs_msg(struct amdgpu_uvd_cs_ctx *ctx,
 
 		/* try to alloc a new handle */
 		for (i = 0; i < adev->uvd.max_handles; ++i) {
-			if (atomic_read(&adev->uvd.inst->handles[i]) == handle) {
-				DRM_ERROR("Handle 0x%x already in use!\n", handle);
+			if (atomic_read(&adev->uvd.inst[ip_instance].handles[i]) == handle) {
+				DRM_ERROR("(%d)Handle 0x%x already in use!\n", ip_instance, handle);
 				return -EINVAL;
 			}
 
-			if (!atomic_cmpxchg(&adev->uvd.inst->handles[i], 0, handle)) {
-				adev->uvd.inst->filp[i] = ctx->parser->filp;
+			if (!atomic_cmpxchg(&adev->uvd.inst[ip_instance].handles[i], 0, handle)) {
+				adev->uvd.inst[ip_instance].filp[i] = ctx->parser->filp;
 				return 0;
 			}
 		}
 
-		DRM_ERROR("No more free UVD handles!\n");
+		DRM_ERROR("No more free UVD(%d) handles!\n", ip_instance);
 		return -ENOSPC;
 
 	case 1:
@@ -719,27 +731,27 @@ static int amdgpu_uvd_cs_msg(struct amdgpu_uvd_cs_ctx *ctx,
 
 		/* validate the handle */
 		for (i = 0; i < adev->uvd.max_handles; ++i) {
-			if (atomic_read(&adev->uvd.inst->handles[i]) == handle) {
-				if (adev->uvd.inst->filp[i] != ctx->parser->filp) {
-					DRM_ERROR("UVD handle collision detected!\n");
+			if (atomic_read(&adev->uvd.inst[ip_instance].handles[i]) == handle) {
+				if (adev->uvd.inst[ip_instance].filp[i] != ctx->parser->filp) {
+					DRM_ERROR("UVD(%d) handle collision detected!\n", ip_instance);
 					return -EINVAL;
 				}
 				return 0;
 			}
 		}
 
-		DRM_ERROR("Invalid UVD handle 0x%x!\n", handle);
+		DRM_ERROR("Invalid UVD(%d) handle 0x%x!\n", ip_instance, handle);
 		return -ENOENT;
 
 	case 2:
 		/* it's a destroy msg, free the handle */
 		for (i = 0; i < adev->uvd.max_handles; ++i)
-			atomic_cmpxchg(&adev->uvd.inst->handles[i], handle, 0);
+			atomic_cmpxchg(&adev->uvd.inst[ip_instance].handles[i], handle, 0);
 		amdgpu_bo_kunmap(bo);
 		return 0;
 
 	default:
-		DRM_ERROR("Illegal UVD message type (%d)!\n", msg_type);
+		DRM_ERROR("Illegal UVD(%d) message type (%d)!\n", ip_instance, msg_type);
 		return -EINVAL;
 	}
 	BUG();
@@ -1043,7 +1055,7 @@ static int amdgpu_uvd_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo,
 		if (r)
 			goto err_free;
 
-		r = amdgpu_job_submit(job, ring, &adev->uvd.inst->entity,
+		r = amdgpu_job_submit(job, ring, &adev->uvd.inst[ring->me].entity,
 				      AMDGPU_FENCE_OWNER_UNDEFINED, &f);
 		if (r)
 			goto err_free;
@@ -1189,27 +1201,28 @@ int amdgpu_uvd_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 {
 	struct dma_fence *fence;
 	long r;
+	uint32_t ip_instance = ring->me;
 
 	r = amdgpu_uvd_get_create_msg(ring, 1, NULL);
 	if (r) {
-		DRM_ERROR("amdgpu: failed to get create msg (%ld).\n", r);
+		DRM_ERROR("amdgpu: (%d)failed to get create msg (%ld).\n", ip_instance, r);
 		goto error;
 	}
 
 	r = amdgpu_uvd_get_destroy_msg(ring, 1, true, &fence);
 	if (r) {
-		DRM_ERROR("amdgpu: failed to get destroy ib (%ld).\n", r);
+		DRM_ERROR("amdgpu: (%d)failed to get destroy ib (%ld).\n", ip_instance, r);
 		goto error;
 	}
 
 	r = dma_fence_wait_timeout(fence, false, timeout);
 	if (r == 0) {
-		DRM_ERROR("amdgpu: IB test timed out.\n");
+		DRM_ERROR("amdgpu: (%d)IB test timed out.\n", ip_instance);
 		r = -ETIMEDOUT;
 	} else if (r < 0) {
-		DRM_ERROR("amdgpu: fence wait failed (%ld).\n", r);
+		DRM_ERROR("amdgpu: (%d)fence wait failed (%ld).\n", ip_instance, r);
 	} else {
-		DRM_DEBUG("ib test on ring %d succeeded\n", ring->idx);
+		DRM_DEBUG("ib test on (%d)ring %d succeeded\n", ip_instance, ring->idx);
 		r = 0;
 	}
 
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
index 66d4bea..08f3b6c 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
@@ -58,7 +58,7 @@ static uint64_t uvd_v7_0_ring_get_rptr(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
 
-	return RREG32_SOC15(UVD, 0, mmUVD_RBC_RB_RPTR);
+	return RREG32_SOC15(UVD, ring->me, mmUVD_RBC_RB_RPTR);
 }
 
 /**
@@ -72,10 +72,10 @@ static uint64_t uvd_v7_0_enc_ring_get_rptr(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
 
-	if (ring == &adev->uvd.inst->ring_enc[0])
-		return RREG32_SOC15(UVD, 0, mmUVD_RB_RPTR);
+	if (ring == &adev->uvd.inst[ring->me].ring_enc[0])
+		return RREG32_SOC15(UVD, ring->me, mmUVD_RB_RPTR);
 	else
-		return RREG32_SOC15(UVD, 0, mmUVD_RB_RPTR2);
+		return RREG32_SOC15(UVD, ring->me, mmUVD_RB_RPTR2);
 }
 
 /**
@@ -89,7 +89,7 @@ static uint64_t uvd_v7_0_ring_get_wptr(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
 
-	return RREG32_SOC15(UVD, 0, mmUVD_RBC_RB_WPTR);
+	return RREG32_SOC15(UVD, ring->me, mmUVD_RBC_RB_WPTR);
 }
 
 /**
@@ -106,10 +106,10 @@ static uint64_t uvd_v7_0_enc_ring_get_wptr(struct amdgpu_ring *ring)
 	if (ring->use_doorbell)
 		return adev->wb.wb[ring->wptr_offs];
 
-	if (ring == &adev->uvd.inst->ring_enc[0])
-		return RREG32_SOC15(UVD, 0, mmUVD_RB_WPTR);
+	if (ring == &adev->uvd.inst[ring->me].ring_enc[0])
+		return RREG32_SOC15(UVD, ring->me, mmUVD_RB_WPTR);
 	else
-		return RREG32_SOC15(UVD, 0, mmUVD_RB_WPTR2);
+		return RREG32_SOC15(UVD, ring->me, mmUVD_RB_WPTR2);
 }
 
 /**
@@ -123,7 +123,7 @@ static void uvd_v7_0_ring_set_wptr(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
 
-	WREG32_SOC15(UVD, 0, mmUVD_RBC_RB_WPTR, lower_32_bits(ring->wptr));
+	WREG32_SOC15(UVD, ring->me, mmUVD_RBC_RB_WPTR, lower_32_bits(ring->wptr));
 }
 
 /**
@@ -144,11 +144,11 @@ static void uvd_v7_0_enc_ring_set_wptr(struct amdgpu_ring *ring)
 		return;
 	}
 
-	if (ring == &adev->uvd.inst->ring_enc[0])
-		WREG32_SOC15(UVD, 0, mmUVD_RB_WPTR,
+	if (ring == &adev->uvd.inst[ring->me].ring_enc[0])
+		WREG32_SOC15(UVD, ring->me, mmUVD_RB_WPTR,
 			lower_32_bits(ring->wptr));
 	else
-		WREG32_SOC15(UVD, 0, mmUVD_RB_WPTR2,
+		WREG32_SOC15(UVD, ring->me, mmUVD_RB_WPTR2,
 			lower_32_bits(ring->wptr));
 }
 
@@ -387,19 +387,21 @@ static int uvd_v7_0_sw_init(void *handle)
 {
 	struct amdgpu_ring *ring;
 	struct drm_sched_rq *rq;
-	int i, r;
+	int i, j, r;
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	/* UVD TRAP */
-	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_UVD, 124, &adev->uvd.inst->irq);
-	if (r)
-		return r;
-
-	/* UVD ENC TRAP */
-	for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
-		r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_UVD, i + 119, &adev->uvd.inst->irq);
+	for (j = 0; j < adev->uvd.num_uvd_inst; j++) {
+		/* UVD TRAP */
+		r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_UVD, 124, &adev->uvd.inst[j].irq);
 		if (r)
 			return r;
+
+		/* UVD ENC TRAP */
+		for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
+			r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_UVD, i + 119, &adev->uvd.inst[j].irq);
+			if (r)
+				return r;
+		}
 	}
 
 	r = amdgpu_uvd_sw_init(adev);
@@ -416,43 +418,48 @@ static int uvd_v7_0_sw_init(void *handle)
 		DRM_INFO("PSP loading UVD firmware\n");
 	}
 
-	ring = &adev->uvd.inst->ring_enc[0];
-	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
-	r = drm_sched_entity_init(&ring->sched, &adev->uvd.inst->entity_enc,
-				  rq, NULL);
-	if (r) {
-		DRM_ERROR("Failed setting up UVD ENC run queue.\n");
-		return r;
+	for (j = 0; j < adev->uvd.num_uvd_inst; j++) {
+		ring = &adev->uvd.inst[j].ring_enc[0];
+		rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
+		r = drm_sched_entity_init(&ring->sched, &adev->uvd.inst[j].entity_enc,
+					  rq, NULL);
+		if (r) {
+			DRM_ERROR("(%d)Failed setting up UVD ENC run queue.\n", j);
+			return r;
+		}
 	}
 
 	r = amdgpu_uvd_resume(adev);
 	if (r)
 		return r;
-	if (!amdgpu_sriov_vf(adev)) {
-		ring = &adev->uvd.inst->ring;
-		sprintf(ring->name, "uvd");
-		r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.inst->irq, 0);
-		if (r)
-			return r;
-	}
 
-	for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
-		ring = &adev->uvd.inst->ring_enc[i];
-		sprintf(ring->name, "uvd_enc%d", i);
-		if (amdgpu_sriov_vf(adev)) {
-			ring->use_doorbell = true;
-
-			/* currently only use the first enconding ring for
-			 * sriov, so set unused location for other unused rings.
-			 */
-			if (i == 0)
-				ring->doorbell_index = AMDGPU_DOORBELL64_UVD_RING0_1 * 2;
-			else
-				ring->doorbell_index = AMDGPU_DOORBELL64_UVD_RING2_3 * 2 + 1;
+	for (j = 0; j < adev->uvd.num_uvd_inst; j++) {
+		if (!amdgpu_sriov_vf(adev)) {
+			ring = &adev->uvd.inst[j].ring;
+			sprintf(ring->name, "uvd<%d>", j);
+			r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.inst[j].irq, 0);
+			if (r)
+				return r;
+		}
+
+		for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
+			ring = &adev->uvd.inst[j].ring_enc[i];
+			sprintf(ring->name, "uvd_enc%d<%d>", i, j);
+			if (amdgpu_sriov_vf(adev)) {
+				ring->use_doorbell = true;
+
+				/* currently only use the first enconding ring for
+				 * sriov, so set unused location for other unused rings.
+				 */
+				if (i == 0)
+					ring->doorbell_index = AMDGPU_DOORBELL64_UVD_RING0_1 * 2;
+				else
+					ring->doorbell_index = AMDGPU_DOORBELL64_UVD_RING2_3 * 2 + 1;
+			}
+			r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.inst[j].irq, 0);
+			if (r)
+				return r;
 		}
-		r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.inst->irq, 0);
-		if (r)
-			return r;
 	}
 
 	r = amdgpu_virt_alloc_mm_table(adev);
@@ -464,7 +471,7 @@ static int uvd_v7_0_sw_init(void *handle)
 
 static int uvd_v7_0_sw_fini(void *handle)
 {
-	int i, r;
+	int i, j, r;
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
 	amdgpu_virt_free_mm_table(adev);
@@ -473,11 +480,12 @@ static int uvd_v7_0_sw_fini(void *handle)
 	if (r)
 		return r;
 
-	drm_sched_entity_fini(&adev->uvd.inst->ring_enc[0].sched, &adev->uvd.inst->entity_enc);
-
-	for (i = 0; i < adev->uvd.num_enc_rings; ++i)
-		amdgpu_ring_fini(&adev->uvd.inst->ring_enc[i]);
+	for (j = 0; j < adev->uvd.num_uvd_inst; ++j) {
+		drm_sched_entity_fini(&adev->uvd.inst[j].ring_enc[0].sched, &adev->uvd.inst[j].entity_enc);
 
+		for (i = 0; i < adev->uvd.num_enc_rings; ++i)
+			amdgpu_ring_fini(&adev->uvd.inst[j].ring_enc[i]);
+	}
 	return amdgpu_uvd_sw_fini(adev);
 }
 
@@ -491,9 +499,9 @@ static int uvd_v7_0_sw_fini(void *handle)
 static int uvd_v7_0_hw_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
+	struct amdgpu_ring *ring;
 	uint32_t tmp;
-	int i, r;
+	int i, j, r;
 
 	if (amdgpu_sriov_vf(adev))
 		r = uvd_v7_0_sriov_start(adev);
@@ -502,57 +510,60 @@ static int uvd_v7_0_hw_init(void *handle)
 	if (r)
 		goto done;
 
-	if (!amdgpu_sriov_vf(adev)) {
-		ring->ready = true;
-		r = amdgpu_ring_test_ring(ring);
-		if (r) {
-			ring->ready = false;
-			goto done;
+	for (j = 0; j < adev->uvd.num_uvd_inst; ++j) {
+		ring = &adev->uvd.inst[j].ring;
+
+		if (!amdgpu_sriov_vf(adev)) {
+			ring->ready = true;
+			r = amdgpu_ring_test_ring(ring);
+			if (r) {
+				ring->ready = false;
+				goto done;
+			}
+
+			r = amdgpu_ring_alloc(ring, 10);
+			if (r) {
+				DRM_ERROR("amdgpu: (%d)ring failed to lock UVD ring (%d).\n", j, r);
+				goto done;
+			}
+
+			tmp = PACKET0(SOC15_REG_OFFSET(UVD, j,
+				mmUVD_SEMA_WAIT_FAULT_TIMEOUT_CNTL), 0);
+			amdgpu_ring_write(ring, tmp);
+			amdgpu_ring_write(ring, 0xFFFFF);
+
+			tmp = PACKET0(SOC15_REG_OFFSET(UVD, j,
+				mmUVD_SEMA_WAIT_INCOMPLETE_TIMEOUT_CNTL), 0);
+			amdgpu_ring_write(ring, tmp);
+			amdgpu_ring_write(ring, 0xFFFFF);
+
+			tmp = PACKET0(SOC15_REG_OFFSET(UVD, j,
+				mmUVD_SEMA_SIGNAL_INCOMPLETE_TIMEOUT_CNTL), 0);
+			amdgpu_ring_write(ring, tmp);
+			amdgpu_ring_write(ring, 0xFFFFF);
+
+			/* Clear timeout status bits */
+			amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, j,
+				mmUVD_SEMA_TIMEOUT_STATUS), 0));
+			amdgpu_ring_write(ring, 0x8);
+
+			amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, j,
+				mmUVD_SEMA_CNTL), 0));
+			amdgpu_ring_write(ring, 3);
+
+			amdgpu_ring_commit(ring);
 		}
 
-		r = amdgpu_ring_alloc(ring, 10);
-		if (r) {
-			DRM_ERROR("amdgpu: ring failed to lock UVD ring (%d).\n", r);
-			goto done;
+		for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
+			ring = &adev->uvd.inst[j].ring_enc[i];
+			ring->ready = true;
+			r = amdgpu_ring_test_ring(ring);
+			if (r) {
+				ring->ready = false;
+				goto done;
+			}
 		}
-
-		tmp = PACKET0(SOC15_REG_OFFSET(UVD, 0,
-			mmUVD_SEMA_WAIT_FAULT_TIMEOUT_CNTL), 0);
-		amdgpu_ring_write(ring, tmp);
-		amdgpu_ring_write(ring, 0xFFFFF);
-
-		tmp = PACKET0(SOC15_REG_OFFSET(UVD, 0,
-			mmUVD_SEMA_WAIT_INCOMPLETE_TIMEOUT_CNTL), 0);
-		amdgpu_ring_write(ring, tmp);
-		amdgpu_ring_write(ring, 0xFFFFF);
-
-		tmp = PACKET0(SOC15_REG_OFFSET(UVD, 0,
-			mmUVD_SEMA_SIGNAL_INCOMPLETE_TIMEOUT_CNTL), 0);
-		amdgpu_ring_write(ring, tmp);
-		amdgpu_ring_write(ring, 0xFFFFF);
-
-		/* Clear timeout status bits */
-		amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0,
-			mmUVD_SEMA_TIMEOUT_STATUS), 0));
-		amdgpu_ring_write(ring, 0x8);
-
-		amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0,
-			mmUVD_SEMA_CNTL), 0));
-		amdgpu_ring_write(ring, 3);
-
-		amdgpu_ring_commit(ring);
 	}
-
-	for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
-		ring = &adev->uvd.inst->ring_enc[i];
-		ring->ready = true;
-		r = amdgpu_ring_test_ring(ring);
-		if (r) {
-			ring->ready = false;
-			goto done;
-		}
-	}
-
 done:
 	if (!r)
 		DRM_INFO("UVD and UVD ENC initialized successfully.\n");
@@ -570,7 +581,7 @@ done:
 static int uvd_v7_0_hw_fini(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
+	int i;
 
 	if (!amdgpu_sriov_vf(adev))
 		uvd_v7_0_stop(adev);
@@ -579,7 +590,8 @@ static int uvd_v7_0_hw_fini(void *handle)
 		DRM_DEBUG("For SRIOV client, shouldn't do anything.\n");
 	}
 
-	ring->ready = false;
+	for (i = 0; i < adev->uvd.num_uvd_inst; ++i)
+		adev->uvd.inst[i].ring.ready = false;
 
 	return 0;
 }
@@ -619,48 +631,51 @@ static void uvd_v7_0_mc_resume(struct amdgpu_device *adev)
 {
 	uint32_t size = AMDGPU_UVD_FIRMWARE_SIZE(adev);
 	uint32_t offset;
+	int i;
 
-	if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) {
-		WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW,
-			lower_32_bits(adev->firmware.ucode[AMDGPU_UCODE_ID_UVD].mc_addr));
-		WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH,
-			upper_32_bits(adev->firmware.ucode[AMDGPU_UCODE_ID_UVD].mc_addr));
-		offset = 0;
-	} else {
-		WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW,
-			lower_32_bits(adev->uvd.inst->gpu_addr));
-		WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH,
-			upper_32_bits(adev->uvd.inst->gpu_addr));
-		offset = size;
-	}
+	for (i = 0; i < adev->uvd.num_uvd_inst; ++i) {
+		if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) {
+			WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW,
+				lower_32_bits(adev->firmware.ucode[AMDGPU_UCODE_ID_UVD].mc_addr));
+			WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH,
+				upper_32_bits(adev->firmware.ucode[AMDGPU_UCODE_ID_UVD].mc_addr));
+			offset = 0;
+		} else {
+			WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW,
+				lower_32_bits(adev->uvd.inst[i].gpu_addr));
+			WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH,
+				upper_32_bits(adev->uvd.inst[i].gpu_addr));
+			offset = size;
+		}
 
-	WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_OFFSET0,
-				AMDGPU_UVD_FIRMWARE_OFFSET >> 3);
-	WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_SIZE0, size);
-
-	WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW,
-			lower_32_bits(adev->uvd.inst->gpu_addr + offset));
-	WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH,
-			upper_32_bits(adev->uvd.inst->gpu_addr + offset));
-	WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_OFFSET1, (1 << 21));
-	WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_SIZE1, AMDGPU_UVD_HEAP_SIZE);
-
-	WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_LOW,
-			lower_32_bits(adev->uvd.inst->gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE));
-	WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_HIGH,
-			upper_32_bits(adev->uvd.inst->gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE));
-	WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_OFFSET2, (2 << 21));
-	WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_SIZE2,
-			AMDGPU_UVD_STACK_SIZE + (AMDGPU_UVD_SESSION_SIZE * 40));
-
-	WREG32_SOC15(UVD, 0, mmUVD_UDEC_ADDR_CONFIG,
-			adev->gfx.config.gb_addr_config);
-	WREG32_SOC15(UVD, 0, mmUVD_UDEC_DB_ADDR_CONFIG,
-			adev->gfx.config.gb_addr_config);
-	WREG32_SOC15(UVD, 0, mmUVD_UDEC_DBW_ADDR_CONFIG,
-			adev->gfx.config.gb_addr_config);
-
-	WREG32_SOC15(UVD, 0, mmUVD_GP_SCRATCH4, adev->uvd.max_handles);
+		WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_OFFSET0,
+					AMDGPU_UVD_FIRMWARE_OFFSET >> 3);
+		WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_SIZE0, size);
+
+		WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW,
+				lower_32_bits(adev->uvd.inst[i].gpu_addr + offset));
+		WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH,
+				upper_32_bits(adev->uvd.inst[i].gpu_addr + offset));
+		WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_OFFSET1, (1 << 21));
+		WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_SIZE1, AMDGPU_UVD_HEAP_SIZE);
+
+		WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_LOW,
+				lower_32_bits(adev->uvd.inst[i].gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE));
+		WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_HIGH,
+				upper_32_bits(adev->uvd.inst[i].gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE));
+		WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_OFFSET2, (2 << 21));
+		WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_SIZE2,
+				AMDGPU_UVD_STACK_SIZE + (AMDGPU_UVD_SESSION_SIZE * 40));
+
+		WREG32_SOC15(UVD, i, mmUVD_UDEC_ADDR_CONFIG,
+				adev->gfx.config.gb_addr_config);
+		WREG32_SOC15(UVD, i, mmUVD_UDEC_DB_ADDR_CONFIG,
+				adev->gfx.config.gb_addr_config);
+		WREG32_SOC15(UVD, i, mmUVD_UDEC_DBW_ADDR_CONFIG,
+				adev->gfx.config.gb_addr_config);
+
+		WREG32_SOC15(UVD, i, mmUVD_GP_SCRATCH4, adev->uvd.max_handles);
+	}
 }
 
 static int uvd_v7_0_mmsch_start(struct amdgpu_device *adev,
@@ -670,6 +685,7 @@ static int uvd_v7_0_mmsch_start(struct amdgpu_device *adev,
 	uint64_t addr = table->gpu_addr;
 	struct mmsch_v1_0_init_header *header = (struct mmsch_v1_0_init_header *)table->cpu_addr;
 	uint32_t size;
+	int i;
 
 	size = header->header_size + header->vce_table_size + header->uvd_table_size;
 
@@ -689,11 +705,12 @@ static int uvd_v7_0_mmsch_start(struct amdgpu_device *adev,
 	/* 4, set resp to zero */
 	WREG32_SOC15(VCE, 0, mmVCE_MMSCH_VF_MAILBOX_RESP, 0);
 
-	WDOORBELL32(adev->uvd.inst->ring_enc[0].doorbell_index, 0);
-	adev->wb.wb[adev->uvd.inst->ring_enc[0].wptr_offs] = 0;
-	adev->uvd.inst->ring_enc[0].wptr = 0;
-	adev->uvd.inst->ring_enc[0].wptr_old = 0;
-
+	for (i = 0; i < adev->uvd.num_uvd_inst; ++i) {
+		WDOORBELL32(adev->uvd.inst[i].ring_enc[0].doorbell_index, 0);
+		adev->wb.wb[adev->uvd.inst[i].ring_enc[0].wptr_offs] = 0;
+		adev->uvd.inst[i].ring_enc[0].wptr = 0;
+		adev->uvd.inst[i].ring_enc[0].wptr_old = 0;
+	}
 	/* 5, kick off the initialization and wait until VCE_MMSCH_VF_MAILBOX_RESP becomes non-zero */
 	WREG32_SOC15(VCE, 0, mmVCE_MMSCH_VF_MAILBOX_HOST, 0x10000001);
 
@@ -726,6 +743,7 @@ static int uvd_v7_0_sriov_start(struct amdgpu_device *adev)
 	struct mmsch_v1_0_cmd_end end = { {0} };
 	uint32_t *init_table = adev->virt.mm_table.cpu_addr;
 	struct mmsch_v1_0_init_header *header = (struct mmsch_v1_0_init_header *)init_table;
+	uint8_t i = 0;
 
 	direct_wt.cmd_header.command_type = MMSCH_COMMAND__DIRECT_REG_WRITE;
 	direct_rd_mod_wt.cmd_header.command_type = MMSCH_COMMAND__DIRECT_REG_READ_MODIFY_WRITE;
@@ -743,120 +761,121 @@ static int uvd_v7_0_sriov_start(struct amdgpu_device *adev)
 
 		init_table += header->uvd_table_offset;
 
-		ring = &adev->uvd.inst->ring;
-		ring->wptr = 0;
-		size = AMDGPU_GPU_PAGE_ALIGN(adev->uvd.fw->size + 4);
-
-		MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_STATUS),
-						   0xFFFFFFFF, 0x00000004);
-		/* mc resume*/
-		if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) {
-			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW),
-						    lower_32_bits(adev->firmware.ucode[AMDGPU_UCODE_ID_UVD].mc_addr));
-			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH),
-						    upper_32_bits(adev->firmware.ucode[AMDGPU_UCODE_ID_UVD].mc_addr));
-			offset = 0;
-		} else {
-			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW),
-						    lower_32_bits(adev->uvd.inst->gpu_addr));
-			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH),
-						    upper_32_bits(adev->uvd.inst->gpu_addr));
-			offset = size;
+		for (i = 0; i < adev->uvd.num_uvd_inst; ++i) {
+			ring = &adev->uvd.inst[i].ring;
+			ring->wptr = 0;
+			size = AMDGPU_GPU_PAGE_ALIGN(adev->uvd.fw->size + 4);
+
+			MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_STATUS),
+							   0xFFFFFFFF, 0x00000004);
+			/* mc resume*/
+			if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) {
+				MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW),
+							    lower_32_bits(adev->firmware.ucode[AMDGPU_UCODE_ID_UVD].mc_addr));
+				MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH),
+							    upper_32_bits(adev->firmware.ucode[AMDGPU_UCODE_ID_UVD].mc_addr));
+				offset = 0;
+			} else {
+				MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW),
+							    lower_32_bits(adev->uvd.inst[i].gpu_addr));
+				MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH),
+							    upper_32_bits(adev->uvd.inst[i].gpu_addr));
+				offset = size;
+			}
+
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_OFFSET0),
+						    AMDGPU_UVD_FIRMWARE_OFFSET >> 3);
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_SIZE0), size);
+
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW),
+						    lower_32_bits(adev->uvd.inst[i].gpu_addr + offset));
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH),
+						    upper_32_bits(adev->uvd.inst[i].gpu_addr + offset));
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_OFFSET1), (1 << 21));
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_SIZE1), AMDGPU_UVD_HEAP_SIZE);
+
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_LOW),
+						    lower_32_bits(adev->uvd.inst[i].gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE));
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_HIGH),
+						    upper_32_bits(adev->uvd.inst[i].gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE));
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_OFFSET2), (2 << 21));
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_SIZE2),
+						    AMDGPU_UVD_STACK_SIZE + (AMDGPU_UVD_SESSION_SIZE * 40));
+
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_GP_SCRATCH4), adev->uvd.max_handles);
+			/* mc resume end*/
+
+			/* disable clock gating */
+			MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_CGC_CTRL),
+							   ~UVD_CGC_CTRL__DYN_CLOCK_MODE_MASK, 0);
+
+			/* disable interupt */
+			MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_MASTINT_EN),
+							   ~UVD_MASTINT_EN__VCPU_EN_MASK, 0);
+
+			/* stall UMC and register bus before resetting VCPU */
+			MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_CTRL2),
+							   ~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK,
+							   UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
+
+			/* put LMI, VCPU, RBC etc... into reset */
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_SOFT_RESET),
+						    (uint32_t)(UVD_SOFT_RESET__LMI_SOFT_RESET_MASK |
+							       UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK |
+							       UVD_SOFT_RESET__LBSI_SOFT_RESET_MASK |
+							       UVD_SOFT_RESET__RBC_SOFT_RESET_MASK |
+							       UVD_SOFT_RESET__CSM_SOFT_RESET_MASK |
+							       UVD_SOFT_RESET__CXW_SOFT_RESET_MASK |
+							       UVD_SOFT_RESET__TAP_SOFT_RESET_MASK |
+							       UVD_SOFT_RESET__LMI_UMC_SOFT_RESET_MASK));
+
+			/* initialize UVD memory controller */
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_CTRL),
+						    (uint32_t)((0x40 << UVD_LMI_CTRL__WRITE_CLEAN_TIMER__SHIFT) |
+							       UVD_LMI_CTRL__WRITE_CLEAN_TIMER_EN_MASK |
+							       UVD_LMI_CTRL__DATA_COHERENCY_EN_MASK |
+							       UVD_LMI_CTRL__VCPU_DATA_COHERENCY_EN_MASK |
+							       UVD_LMI_CTRL__REQ_MODE_MASK |
+							       0x00100000L));
+
+			/* take all subblocks out of reset, except VCPU */
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_SOFT_RESET),
+						    UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
+
+			/* enable VCPU clock */
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CNTL),
+						    UVD_VCPU_CNTL__CLK_EN_MASK);
+
+			/* enable master interrupt */
+			MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_MASTINT_EN),
+							   ~(UVD_MASTINT_EN__VCPU_EN_MASK|UVD_MASTINT_EN__SYS_EN_MASK),
+							   (UVD_MASTINT_EN__VCPU_EN_MASK|UVD_MASTINT_EN__SYS_EN_MASK));
+
+			/* clear the bit 4 of UVD_STATUS */
+			MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_STATUS),
+							   ~(2 << UVD_STATUS__VCPU_REPORT__SHIFT), 0);
+
+			/* force RBC into idle state */
+			size = order_base_2(ring->ring_size);
+			tmp = REG_SET_FIELD(0, UVD_RBC_RB_CNTL, RB_BUFSZ, size);
+			tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_FETCH, 1);
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_RBC_RB_CNTL), tmp);
+
+			ring = &adev->uvd.inst[i].ring_enc[0];
+			ring->wptr = 0;
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_RB_BASE_LO), ring->gpu_addr);
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_RB_BASE_HI), upper_32_bits(ring->gpu_addr));
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_RB_SIZE), ring->ring_size / 4);
+
+			/* boot up the VCPU */
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_SOFT_RESET), 0);
+
+			/* enable UMC */
+			MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_CTRL2),
+											   ~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK, 0);
+
+			MMSCH_V1_0_INSERT_DIRECT_POLL(SOC15_REG_OFFSET(UVD, i, mmUVD_STATUS), 0x02, 0x02);
 		}
-
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_OFFSET0),
-					    AMDGPU_UVD_FIRMWARE_OFFSET >> 3);
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_SIZE0), size);
-
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW),
-					    lower_32_bits(adev->uvd.inst->gpu_addr + offset));
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH),
-					    upper_32_bits(adev->uvd.inst->gpu_addr + offset));
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_OFFSET1), (1 << 21));
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_SIZE1), AMDGPU_UVD_HEAP_SIZE);
-
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_LOW),
-					    lower_32_bits(adev->uvd.inst->gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE));
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_HIGH),
-					    upper_32_bits(adev->uvd.inst->gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE));
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_OFFSET2), (2 << 21));
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_SIZE2),
-					    AMDGPU_UVD_STACK_SIZE + (AMDGPU_UVD_SESSION_SIZE * 40));
-
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_GP_SCRATCH4), adev->uvd.max_handles);
-		/* mc resume end*/
-
-		/* disable clock gating */
-		MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_CGC_CTRL),
-						   ~UVD_CGC_CTRL__DYN_CLOCK_MODE_MASK, 0);
-
-		/* disable interupt */
-		MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_MASTINT_EN),
-						   ~UVD_MASTINT_EN__VCPU_EN_MASK, 0);
-
-		/* stall UMC and register bus before resetting VCPU */
-		MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_CTRL2),
-						   ~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK,
-						   UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
-
-		/* put LMI, VCPU, RBC etc... into reset */
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_SOFT_RESET),
-					    (uint32_t)(UVD_SOFT_RESET__LMI_SOFT_RESET_MASK |
-						       UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK |
-						       UVD_SOFT_RESET__LBSI_SOFT_RESET_MASK |
-						       UVD_SOFT_RESET__RBC_SOFT_RESET_MASK |
-						       UVD_SOFT_RESET__CSM_SOFT_RESET_MASK |
-						       UVD_SOFT_RESET__CXW_SOFT_RESET_MASK |
-						       UVD_SOFT_RESET__TAP_SOFT_RESET_MASK |
-						       UVD_SOFT_RESET__LMI_UMC_SOFT_RESET_MASK));
-
-		/* initialize UVD memory controller */
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_CTRL),
-					    (uint32_t)((0x40 << UVD_LMI_CTRL__WRITE_CLEAN_TIMER__SHIFT) |
-						       UVD_LMI_CTRL__WRITE_CLEAN_TIMER_EN_MASK |
-						       UVD_LMI_CTRL__DATA_COHERENCY_EN_MASK |
-						       UVD_LMI_CTRL__VCPU_DATA_COHERENCY_EN_MASK |
-						       UVD_LMI_CTRL__REQ_MODE_MASK |
-						       0x00100000L));
-
-		/* take all subblocks out of reset, except VCPU */
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_SOFT_RESET),
-					    UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
-
-		/* enable VCPU clock */
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CNTL),
-					    UVD_VCPU_CNTL__CLK_EN_MASK);
-
-		/* enable master interrupt */
-		MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_MASTINT_EN),
-						   ~(UVD_MASTINT_EN__VCPU_EN_MASK|UVD_MASTINT_EN__SYS_EN_MASK),
-						   (UVD_MASTINT_EN__VCPU_EN_MASK|UVD_MASTINT_EN__SYS_EN_MASK));
-
-		/* clear the bit 4 of UVD_STATUS */
-		MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_STATUS),
-						   ~(2 << UVD_STATUS__VCPU_REPORT__SHIFT), 0);
-
-		/* force RBC into idle state */
-		size = order_base_2(ring->ring_size);
-		tmp = REG_SET_FIELD(0, UVD_RBC_RB_CNTL, RB_BUFSZ, size);
-		tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_FETCH, 1);
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RBC_RB_CNTL), tmp);
-
-		ring = &adev->uvd.inst->ring_enc[0];
-		ring->wptr = 0;
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_BASE_LO), ring->gpu_addr);
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_BASE_HI), upper_32_bits(ring->gpu_addr));
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_SIZE), ring->ring_size / 4);
-
-		/* boot up the VCPU */
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_SOFT_RESET), 0);
-
-		/* enable UMC */
-		MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_CTRL2),
-										   ~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK, 0);
-
-		MMSCH_V1_0_INSERT_DIRECT_POLL(SOC15_REG_OFFSET(UVD, 0, mmUVD_STATUS), 0x02, 0x02);
-
 		/* add end packet */
 		memcpy((void *)init_table, &end, sizeof(struct mmsch_v1_0_cmd_end));
 		table_size += sizeof(struct mmsch_v1_0_cmd_end) / 4;
@@ -875,15 +894,17 @@ static int uvd_v7_0_sriov_start(struct amdgpu_device *adev)
  */
 static int uvd_v7_0_start(struct amdgpu_device *adev)
 {
-	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
+	struct amdgpu_ring *ring;
 	uint32_t rb_bufsz, tmp;
 	uint32_t lmi_swap_cntl;
 	uint32_t mp_swap_cntl;
-	int i, j, r;
+	int i, j, k, r;
 
-	/* disable DPG */
-	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_POWER_STATUS), 0,
-			~UVD_POWER_STATUS__UVD_PG_MODE_MASK);
+	for (k = 0; k < adev->uvd.num_uvd_inst; ++k) {
+		/* disable DPG */
+		WREG32_P(SOC15_REG_OFFSET(UVD, k, mmUVD_POWER_STATUS), 0,
+				~UVD_POWER_STATUS__UVD_PG_MODE_MASK);
+	}
 
 	/* disable byte swapping */
 	lmi_swap_cntl = 0;
@@ -891,157 +912,159 @@ static int uvd_v7_0_start(struct amdgpu_device *adev)
 
 	uvd_v7_0_mc_resume(adev);
 
-	/* disable clock gating */
-	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_CGC_CTRL), 0,
-			~UVD_CGC_CTRL__DYN_CLOCK_MODE_MASK);
-
-	/* disable interupt */
-	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_MASTINT_EN), 0,
-			~UVD_MASTINT_EN__VCPU_EN_MASK);
-
-	/* stall UMC and register bus before resetting VCPU */
-	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_CTRL2),
-			UVD_LMI_CTRL2__STALL_ARB_UMC_MASK,
-			~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
-	mdelay(1);
-
-	/* put LMI, VCPU, RBC etc... into reset */
-	WREG32_SOC15(UVD, 0, mmUVD_SOFT_RESET,
-		UVD_SOFT_RESET__LMI_SOFT_RESET_MASK |
-		UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK |
-		UVD_SOFT_RESET__LBSI_SOFT_RESET_MASK |
-		UVD_SOFT_RESET__RBC_SOFT_RESET_MASK |
-		UVD_SOFT_RESET__CSM_SOFT_RESET_MASK |
-		UVD_SOFT_RESET__CXW_SOFT_RESET_MASK |
-		UVD_SOFT_RESET__TAP_SOFT_RESET_MASK |
-		UVD_SOFT_RESET__LMI_UMC_SOFT_RESET_MASK);
-	mdelay(5);
+	for (k = 0; k < adev->uvd.num_uvd_inst; ++k) {
+		ring = &adev->uvd.inst[k].ring;
+		/* disable clock gating */
+		WREG32_P(SOC15_REG_OFFSET(UVD, k, mmUVD_CGC_CTRL), 0,
+				~UVD_CGC_CTRL__DYN_CLOCK_MODE_MASK);
 
-	/* initialize UVD memory controller */
-	WREG32_SOC15(UVD, 0, mmUVD_LMI_CTRL,
-		(0x40 << UVD_LMI_CTRL__WRITE_CLEAN_TIMER__SHIFT) |
-		UVD_LMI_CTRL__WRITE_CLEAN_TIMER_EN_MASK |
-		UVD_LMI_CTRL__DATA_COHERENCY_EN_MASK |
-		UVD_LMI_CTRL__VCPU_DATA_COHERENCY_EN_MASK |
-		UVD_LMI_CTRL__REQ_MODE_MASK |
-		0x00100000L);
+		/* disable interupt */
+		WREG32_P(SOC15_REG_OFFSET(UVD, k, mmUVD_MASTINT_EN), 0,
+				~UVD_MASTINT_EN__VCPU_EN_MASK);
+
+		/* stall UMC and register bus before resetting VCPU */
+		WREG32_P(SOC15_REG_OFFSET(UVD, k, mmUVD_LMI_CTRL2),
+				UVD_LMI_CTRL2__STALL_ARB_UMC_MASK,
+				~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
+		mdelay(1);
+
+		/* put LMI, VCPU, RBC etc... into reset */
+		WREG32_SOC15(UVD, k, mmUVD_SOFT_RESET,
+			UVD_SOFT_RESET__LMI_SOFT_RESET_MASK |
+			UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK |
+			UVD_SOFT_RESET__LBSI_SOFT_RESET_MASK |
+			UVD_SOFT_RESET__RBC_SOFT_RESET_MASK |
+			UVD_SOFT_RESET__CSM_SOFT_RESET_MASK |
+			UVD_SOFT_RESET__CXW_SOFT_RESET_MASK |
+			UVD_SOFT_RESET__TAP_SOFT_RESET_MASK |
+			UVD_SOFT_RESET__LMI_UMC_SOFT_RESET_MASK);
+		mdelay(5);
+
+		/* initialize UVD memory controller */
+		WREG32_SOC15(UVD, k, mmUVD_LMI_CTRL,
+			(0x40 << UVD_LMI_CTRL__WRITE_CLEAN_TIMER__SHIFT) |
+			UVD_LMI_CTRL__WRITE_CLEAN_TIMER_EN_MASK |
+			UVD_LMI_CTRL__DATA_COHERENCY_EN_MASK |
+			UVD_LMI_CTRL__VCPU_DATA_COHERENCY_EN_MASK |
+			UVD_LMI_CTRL__REQ_MODE_MASK |
+			0x00100000L);
 
 #ifdef __BIG_ENDIAN
-	/* swap (8 in 32) RB and IB */
-	lmi_swap_cntl = 0xa;
-	mp_swap_cntl = 0;
+		/* swap (8 in 32) RB and IB */
+		lmi_swap_cntl = 0xa;
+		mp_swap_cntl = 0;
 #endif
-	WREG32_SOC15(UVD, 0, mmUVD_LMI_SWAP_CNTL, lmi_swap_cntl);
-	WREG32_SOC15(UVD, 0, mmUVD_MP_SWAP_CNTL, mp_swap_cntl);
-
-	WREG32_SOC15(UVD, 0, mmUVD_MPC_SET_MUXA0, 0x40c2040);
-	WREG32_SOC15(UVD, 0, mmUVD_MPC_SET_MUXA1, 0x0);
-	WREG32_SOC15(UVD, 0, mmUVD_MPC_SET_MUXB0, 0x40c2040);
-	WREG32_SOC15(UVD, 0, mmUVD_MPC_SET_MUXB1, 0x0);
-	WREG32_SOC15(UVD, 0, mmUVD_MPC_SET_ALU, 0);
-	WREG32_SOC15(UVD, 0, mmUVD_MPC_SET_MUX, 0x88);
-
-	/* take all subblocks out of reset, except VCPU */
-	WREG32_SOC15(UVD, 0, mmUVD_SOFT_RESET,
-			UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
-	mdelay(5);
+		WREG32_SOC15(UVD, k, mmUVD_LMI_SWAP_CNTL, lmi_swap_cntl);
+		WREG32_SOC15(UVD, k, mmUVD_MP_SWAP_CNTL, mp_swap_cntl);
 
-	/* enable VCPU clock */
-	WREG32_SOC15(UVD, 0, mmUVD_VCPU_CNTL,
-			UVD_VCPU_CNTL__CLK_EN_MASK);
+		WREG32_SOC15(UVD, k, mmUVD_MPC_SET_MUXA0, 0x40c2040);
+		WREG32_SOC15(UVD, k, mmUVD_MPC_SET_MUXA1, 0x0);
+		WREG32_SOC15(UVD, k, mmUVD_MPC_SET_MUXB0, 0x40c2040);
+		WREG32_SOC15(UVD, k, mmUVD_MPC_SET_MUXB1, 0x0);
+		WREG32_SOC15(UVD, k, mmUVD_MPC_SET_ALU, 0);
+		WREG32_SOC15(UVD, k, mmUVD_MPC_SET_MUX, 0x88);
 
-	/* enable UMC */
-	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_CTRL2), 0,
-			~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
+		/* take all subblocks out of reset, except VCPU */
+		WREG32_SOC15(UVD, k, mmUVD_SOFT_RESET,
+				UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
+		mdelay(5);
 
-	/* boot up the VCPU */
-	WREG32_SOC15(UVD, 0, mmUVD_SOFT_RESET, 0);
-	mdelay(10);
+		/* enable VCPU clock */
+		WREG32_SOC15(UVD, k, mmUVD_VCPU_CNTL,
+				UVD_VCPU_CNTL__CLK_EN_MASK);
 
-	for (i = 0; i < 10; ++i) {
-		uint32_t status;
+		/* enable UMC */
+		WREG32_P(SOC15_REG_OFFSET(UVD, k, mmUVD_LMI_CTRL2), 0,
+				~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
 
-		for (j = 0; j < 100; ++j) {
-			status = RREG32_SOC15(UVD, 0, mmUVD_STATUS);
+		/* boot up the VCPU */
+		WREG32_SOC15(UVD, k, mmUVD_SOFT_RESET, 0);
+		mdelay(10);
+
+		for (i = 0; i < 10; ++i) {
+			uint32_t status;
+
+			for (j = 0; j < 100; ++j) {
+				status = RREG32_SOC15(UVD, k, mmUVD_STATUS);
+				if (status & 2)
+					break;
+				mdelay(10);
+			}
+			r = 0;
 			if (status & 2)
 				break;
+
+			DRM_ERROR("UVD(%d) not responding, trying to reset the VCPU!!!\n", k);
+			WREG32_P(SOC15_REG_OFFSET(UVD, k, mmUVD_SOFT_RESET),
+					UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK,
+					~UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
+			mdelay(10);
+			WREG32_P(SOC15_REG_OFFSET(UVD, k, mmUVD_SOFT_RESET), 0,
+					~UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
 			mdelay(10);
+			r = -1;
 		}
-		r = 0;
-		if (status & 2)
-			break;
 
-		DRM_ERROR("UVD not responding, trying to reset the VCPU!!!\n");
-		WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_SOFT_RESET),
-				UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK,
-				~UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
-		mdelay(10);
-		WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_SOFT_RESET), 0,
-				~UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
-		mdelay(10);
-		r = -1;
-	}
-
-	if (r) {
-		DRM_ERROR("UVD not responding, giving up!!!\n");
-		return r;
-	}
-	/* enable master interrupt */
-	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_MASTINT_EN),
-		(UVD_MASTINT_EN__VCPU_EN_MASK|UVD_MASTINT_EN__SYS_EN_MASK),
-		~(UVD_MASTINT_EN__VCPU_EN_MASK|UVD_MASTINT_EN__SYS_EN_MASK));
-
-	/* clear the bit 4 of UVD_STATUS */
-	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_STATUS), 0,
-			~(2 << UVD_STATUS__VCPU_REPORT__SHIFT));
-
-	/* force RBC into idle state */
-	rb_bufsz = order_base_2(ring->ring_size);
-	tmp = REG_SET_FIELD(0, UVD_RBC_RB_CNTL, RB_BUFSZ, rb_bufsz);
-	tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_BLKSZ, 1);
-	tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_FETCH, 1);
-	tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_WPTR_POLL_EN, 0);
-	tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_UPDATE, 1);
-	tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_RPTR_WR_EN, 1);
-	WREG32_SOC15(UVD, 0, mmUVD_RBC_RB_CNTL, tmp);
-
-	/* set the write pointer delay */
-	WREG32_SOC15(UVD, 0, mmUVD_RBC_RB_WPTR_CNTL, 0);
-
-	/* set the wb address */
-	WREG32_SOC15(UVD, 0, mmUVD_RBC_RB_RPTR_ADDR,
-			(upper_32_bits(ring->gpu_addr) >> 2));
-
-	/* programm the RB_BASE for ring buffer */
-	WREG32_SOC15(UVD, 0, mmUVD_LMI_RBC_RB_64BIT_BAR_LOW,
-			lower_32_bits(ring->gpu_addr));
-	WREG32_SOC15(UVD, 0, mmUVD_LMI_RBC_RB_64BIT_BAR_HIGH,
-			upper_32_bits(ring->gpu_addr));
-
-	/* Initialize the ring buffer's read and write pointers */
-	WREG32_SOC15(UVD, 0, mmUVD_RBC_RB_RPTR, 0);
-
-	ring->wptr = RREG32_SOC15(UVD, 0, mmUVD_RBC_RB_RPTR);
-	WREG32_SOC15(UVD, 0, mmUVD_RBC_RB_WPTR,
-			lower_32_bits(ring->wptr));
-
-	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_RBC_RB_CNTL), 0,
-			~UVD_RBC_RB_CNTL__RB_NO_FETCH_MASK);
-
-	ring = &adev->uvd.inst->ring_enc[0];
-	WREG32_SOC15(UVD, 0, mmUVD_RB_RPTR, lower_32_bits(ring->wptr));
-	WREG32_SOC15(UVD, 0, mmUVD_RB_WPTR, lower_32_bits(ring->wptr));
-	WREG32_SOC15(UVD, 0, mmUVD_RB_BASE_LO, ring->gpu_addr);
-	WREG32_SOC15(UVD, 0, mmUVD_RB_BASE_HI, upper_32_bits(ring->gpu_addr));
-	WREG32_SOC15(UVD, 0, mmUVD_RB_SIZE, ring->ring_size / 4);
+		if (r) {
+			DRM_ERROR("UVD(%d) not responding, giving up!!!\n", k);
+			return r;
+		}
+		/* enable master interrupt */
+		WREG32_P(SOC15_REG_OFFSET(UVD, k, mmUVD_MASTINT_EN),
+			(UVD_MASTINT_EN__VCPU_EN_MASK|UVD_MASTINT_EN__SYS_EN_MASK),
+			~(UVD_MASTINT_EN__VCPU_EN_MASK|UVD_MASTINT_EN__SYS_EN_MASK));
 
-	ring = &adev->uvd.inst->ring_enc[1];
-	WREG32_SOC15(UVD, 0, mmUVD_RB_RPTR2, lower_32_bits(ring->wptr));
-	WREG32_SOC15(UVD, 0, mmUVD_RB_WPTR2, lower_32_bits(ring->wptr));
-	WREG32_SOC15(UVD, 0, mmUVD_RB_BASE_LO2, ring->gpu_addr);
-	WREG32_SOC15(UVD, 0, mmUVD_RB_BASE_HI2, upper_32_bits(ring->gpu_addr));
-	WREG32_SOC15(UVD, 0, mmUVD_RB_SIZE2, ring->ring_size / 4);
+		/* clear the bit 4 of UVD_STATUS */
+		WREG32_P(SOC15_REG_OFFSET(UVD, k, mmUVD_STATUS), 0,
+				~(2 << UVD_STATUS__VCPU_REPORT__SHIFT));
 
+		/* force RBC into idle state */
+		rb_bufsz = order_base_2(ring->ring_size);
+		tmp = REG_SET_FIELD(0, UVD_RBC_RB_CNTL, RB_BUFSZ, rb_bufsz);
+		tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_BLKSZ, 1);
+		tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_FETCH, 1);
+		tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_WPTR_POLL_EN, 0);
+		tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_UPDATE, 1);
+		tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_RPTR_WR_EN, 1);
+		WREG32_SOC15(UVD, k, mmUVD_RBC_RB_CNTL, tmp);
+
+		/* set the write pointer delay */
+		WREG32_SOC15(UVD, k, mmUVD_RBC_RB_WPTR_CNTL, 0);
+
+		/* set the wb address */
+		WREG32_SOC15(UVD, k, mmUVD_RBC_RB_RPTR_ADDR,
+				(upper_32_bits(ring->gpu_addr) >> 2));
+
+		/* programm the RB_BASE for ring buffer */
+		WREG32_SOC15(UVD, k, mmUVD_LMI_RBC_RB_64BIT_BAR_LOW,
+				lower_32_bits(ring->gpu_addr));
+		WREG32_SOC15(UVD, k, mmUVD_LMI_RBC_RB_64BIT_BAR_HIGH,
+				upper_32_bits(ring->gpu_addr));
+
+		/* Initialize the ring buffer's read and write pointers */
+		WREG32_SOC15(UVD, k, mmUVD_RBC_RB_RPTR, 0);
+
+		ring->wptr = RREG32_SOC15(UVD, k, mmUVD_RBC_RB_RPTR);
+		WREG32_SOC15(UVD, k, mmUVD_RBC_RB_WPTR,
+				lower_32_bits(ring->wptr));
+
+		WREG32_P(SOC15_REG_OFFSET(UVD, k, mmUVD_RBC_RB_CNTL), 0,
+				~UVD_RBC_RB_CNTL__RB_NO_FETCH_MASK);
+
+		ring = &adev->uvd.inst[k].ring_enc[0];
+		WREG32_SOC15(UVD, k, mmUVD_RB_RPTR, lower_32_bits(ring->wptr));
+		WREG32_SOC15(UVD, k, mmUVD_RB_WPTR, lower_32_bits(ring->wptr));
+		WREG32_SOC15(UVD, k, mmUVD_RB_BASE_LO, ring->gpu_addr);
+		WREG32_SOC15(UVD, k, mmUVD_RB_BASE_HI, upper_32_bits(ring->gpu_addr));
+		WREG32_SOC15(UVD, k, mmUVD_RB_SIZE, ring->ring_size / 4);
+
+		ring = &adev->uvd.inst[k].ring_enc[1];
+		WREG32_SOC15(UVD, k, mmUVD_RB_RPTR2, lower_32_bits(ring->wptr));
+		WREG32_SOC15(UVD, k, mmUVD_RB_WPTR2, lower_32_bits(ring->wptr));
+		WREG32_SOC15(UVD, k, mmUVD_RB_BASE_LO2, ring->gpu_addr);
+		WREG32_SOC15(UVD, k, mmUVD_RB_BASE_HI2, upper_32_bits(ring->gpu_addr));
+		WREG32_SOC15(UVD, k, mmUVD_RB_SIZE2, ring->ring_size / 4);
+	}
 	return 0;
 }
 
@@ -1054,26 +1077,30 @@ static int uvd_v7_0_start(struct amdgpu_device *adev)
  */
 static void uvd_v7_0_stop(struct amdgpu_device *adev)
 {
-	/* force RBC into idle state */
-	WREG32_SOC15(UVD, 0, mmUVD_RBC_RB_CNTL, 0x11010101);
-
-	/* Stall UMC and register bus before resetting VCPU */
-	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_CTRL2),
-			UVD_LMI_CTRL2__STALL_ARB_UMC_MASK,
-			~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
-	mdelay(1);
-
-	/* put VCPU into reset */
-	WREG32_SOC15(UVD, 0, mmUVD_SOFT_RESET,
-			UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
-	mdelay(5);
+	uint8_t i = 0;
+
+	for (i = 0; i < adev->uvd.num_uvd_inst; ++i) {
+		/* force RBC into idle state */
+		WREG32_SOC15(UVD, i, mmUVD_RBC_RB_CNTL, 0x11010101);
 
-	/* disable VCPU clock */
-	WREG32_SOC15(UVD, 0, mmUVD_VCPU_CNTL, 0x0);
+		/* Stall UMC and register bus before resetting VCPU */
+		WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_CTRL2),
+				UVD_LMI_CTRL2__STALL_ARB_UMC_MASK,
+				~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
+		mdelay(1);
 
-	/* Unstall UMC and register bus */
-	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_CTRL2), 0,
-			~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
+		/* put VCPU into reset */
+		WREG32_SOC15(UVD, i, mmUVD_SOFT_RESET,
+				UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
+		mdelay(5);
+
+		/* disable VCPU clock */
+		WREG32_SOC15(UVD, i, mmUVD_VCPU_CNTL, 0x0);
+
+		/* Unstall UMC and register bus */
+		WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_CTRL2), 0,
+				~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
+	}
 }
 
 /**
@@ -1092,26 +1119,26 @@ static void uvd_v7_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq
 	WARN_ON(flags & AMDGPU_FENCE_FLAG_64BIT);
 
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_CONTEXT_ID), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_CONTEXT_ID), 0));
 	amdgpu_ring_write(ring, seq);
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA0), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GPCOM_VCPU_DATA0), 0));
 	amdgpu_ring_write(ring, addr & 0xffffffff);
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA1), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GPCOM_VCPU_DATA1), 0));
 	amdgpu_ring_write(ring, upper_32_bits(addr) & 0xff);
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_CMD), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GPCOM_VCPU_CMD), 0));
 	amdgpu_ring_write(ring, 0);
 
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA0), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GPCOM_VCPU_DATA0), 0));
 	amdgpu_ring_write(ring, 0);
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA1), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GPCOM_VCPU_DATA1), 0));
 	amdgpu_ring_write(ring, 0);
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_CMD), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GPCOM_VCPU_CMD), 0));
 	amdgpu_ring_write(ring, 2);
 }
 
@@ -1160,7 +1187,7 @@ static int uvd_v7_0_ring_test_ring(struct amdgpu_ring *ring)
 	unsigned i;
 	int r;
 
-	WREG32_SOC15(UVD, 0, mmUVD_CONTEXT_ID, 0xCAFEDEAD);
+	WREG32_SOC15(UVD, ring->me, mmUVD_CONTEXT_ID, 0xCAFEDEAD);
 	r = amdgpu_ring_alloc(ring, 3);
 	if (r) {
 		DRM_ERROR("amdgpu: (%d)cp failed to lock ring %d (%d).\n",
@@ -1168,11 +1195,11 @@ static int uvd_v7_0_ring_test_ring(struct amdgpu_ring *ring)
 		return r;
 	}
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_CONTEXT_ID), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_CONTEXT_ID), 0));
 	amdgpu_ring_write(ring, 0xDEADBEEF);
 	amdgpu_ring_commit(ring);
 	for (i = 0; i < adev->usec_timeout; i++) {
-		tmp = RREG32_SOC15(UVD, 0, mmUVD_CONTEXT_ID);
+		tmp = RREG32_SOC15(UVD, ring->me, mmUVD_CONTEXT_ID);
 		if (tmp == 0xDEADBEEF)
 			break;
 		DRM_UDELAY(1);
@@ -1204,17 +1231,17 @@ static void uvd_v7_0_ring_emit_ib(struct amdgpu_ring *ring,
 	struct amdgpu_device *adev = ring->adev;
 
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_RBC_IB_VMID), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_LMI_RBC_IB_VMID), 0));
 	amdgpu_ring_write(ring, vmid);
 
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_RBC_IB_64BIT_BAR_LOW), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_LMI_RBC_IB_64BIT_BAR_LOW), 0));
 	amdgpu_ring_write(ring, lower_32_bits(ib->gpu_addr));
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_RBC_IB_64BIT_BAR_HIGH), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_LMI_RBC_IB_64BIT_BAR_HIGH), 0));
 	amdgpu_ring_write(ring, upper_32_bits(ib->gpu_addr));
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_RBC_IB_SIZE), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_RBC_IB_SIZE), 0));
 	amdgpu_ring_write(ring, ib->length_dw);
 }
 
@@ -1242,13 +1269,13 @@ static void uvd_v7_0_ring_emit_wreg(struct amdgpu_ring *ring,
 	struct amdgpu_device *adev = ring->adev;
 
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA0), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GPCOM_VCPU_DATA0), 0));
 	amdgpu_ring_write(ring, reg << 2);
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA1), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GPCOM_VCPU_DATA1), 0));
 	amdgpu_ring_write(ring, val);
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_CMD), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GPCOM_VCPU_CMD), 0));
 	amdgpu_ring_write(ring, 8);
 }
 
@@ -1258,16 +1285,16 @@ static void uvd_v7_0_ring_emit_reg_wait(struct amdgpu_ring *ring, uint32_t reg,
 	struct amdgpu_device *adev = ring->adev;
 
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA0), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GPCOM_VCPU_DATA0), 0));
 	amdgpu_ring_write(ring, reg << 2);
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA1), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GPCOM_VCPU_DATA1), 0));
 	amdgpu_ring_write(ring, val);
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GP_SCRATCH8), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GP_SCRATCH8), 0));
 	amdgpu_ring_write(ring, mask);
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_CMD), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GPCOM_VCPU_CMD), 0));
 	amdgpu_ring_write(ring, 12);
 }
 
@@ -1292,7 +1319,7 @@ static void uvd_v7_0_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count)
 	struct amdgpu_device *adev = ring->adev;
 
 	for (i = 0; i < count; i++)
-		amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_NO_OP), 0));
+		amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_NO_OP), 0));
 
 }
 
@@ -1360,16 +1387,16 @@ static bool uvd_v7_0_check_soft_reset(void *handle)
 
 	if (REG_GET_FIELD(tmp, SRBM_STATUS, UVD_RQ_PENDING) ||
 	    REG_GET_FIELD(tmp, SRBM_STATUS, UVD_BUSY) ||
-	    (RREG32_SOC15(UVD, 0, mmUVD_STATUS) &
+	    (RREG32_SOC15(UVD, ring->me, mmUVD_STATUS) &
 		    AMDGPU_UVD_STATUS_BUSY_MASK))
 		srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset,
 				SRBM_SOFT_RESET, SOFT_RESET_UVD, 1);
 
 	if (srbm_soft_reset) {
-		adev->uvd.inst->srbm_soft_reset = srbm_soft_reset;
+		adev->uvd.inst[ring->me].srbm_soft_reset = srbm_soft_reset;
 		return true;
 	} else {
-		adev->uvd.inst->srbm_soft_reset = 0;
+		adev->uvd.inst[ring->me].srbm_soft_reset = 0;
 		return false;
 	}
 }
@@ -1378,7 +1405,7 @@ static int uvd_v7_0_pre_soft_reset(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	if (!adev->uvd.inst->srbm_soft_reset)
+	if (!adev->uvd.inst[ring->me].srbm_soft_reset)
 		return 0;
 
 	uvd_v7_0_stop(adev);
@@ -1390,9 +1417,9 @@ static int uvd_v7_0_soft_reset(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 	u32 srbm_soft_reset;
 
-	if (!adev->uvd.inst->srbm_soft_reset)
+	if (!adev->uvd.inst[ring->me].srbm_soft_reset)
 		return 0;
-	srbm_soft_reset = adev->uvd.inst->srbm_soft_reset;
+	srbm_soft_reset = adev->uvd.inst[ring->me].srbm_soft_reset;
 
 	if (srbm_soft_reset) {
 		u32 tmp;
@@ -1420,7 +1447,7 @@ static int uvd_v7_0_post_soft_reset(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	if (!adev->uvd.inst->srbm_soft_reset)
+	if (!adev->uvd.inst[ring->me].srbm_soft_reset)
 		return 0;
 
 	mdelay(5);
@@ -1442,17 +1469,29 @@ static int uvd_v7_0_process_interrupt(struct amdgpu_device *adev,
 				      struct amdgpu_irq_src *source,
 				      struct amdgpu_iv_entry *entry)
 {
+	uint32_t ip_instance;
+
+	switch (entry->client_id) {
+	case SOC15_IH_CLIENTID_UVD:
+		ip_instance = 0;
+		break;
+	default:
+		DRM_ERROR("Unhandled client id: %d\n", entry->client_id);
+		return 0;
+	}
+
 	DRM_DEBUG("IH: UVD TRAP\n");
+
 	switch (entry->src_id) {
 	case 124:
-		amdgpu_fence_process(&adev->uvd.inst->ring);
+		amdgpu_fence_process(&adev->uvd.inst[ip_instance].ring);
 		break;
 	case 119:
-		amdgpu_fence_process(&adev->uvd.inst->ring_enc[0]);
+		amdgpu_fence_process(&adev->uvd.inst[ip_instance].ring_enc[0]);
 		break;
 	case 120:
 		if (!amdgpu_sriov_vf(adev))
-			amdgpu_fence_process(&adev->uvd.inst->ring_enc[1]);
+			amdgpu_fence_process(&adev->uvd.inst[ip_instance].ring_enc[1]);
 		break;
 	default:
 		DRM_ERROR("Unhandled interrupt: %d %d\n",
@@ -1468,9 +1507,9 @@ static void uvd_v7_0_set_sw_clock_gating(struct amdgpu_device *adev)
 {
 	uint32_t data, data1, data2, suvd_flags;
 
-	data = RREG32_SOC15(UVD, 0, mmUVD_CGC_CTRL);
-	data1 = RREG32_SOC15(UVD, 0, mmUVD_SUVD_CGC_GATE);
-	data2 = RREG32_SOC15(UVD, 0, mmUVD_SUVD_CGC_CTRL);
+	data = RREG32_SOC15(UVD, ring->me, mmUVD_CGC_CTRL);
+	data1 = RREG32_SOC15(UVD, ring->me, mmUVD_SUVD_CGC_GATE);
+	data2 = RREG32_SOC15(UVD, ring->me, mmUVD_SUVD_CGC_CTRL);
 
 	data &= ~(UVD_CGC_CTRL__CLK_OFF_DELAY_MASK |
 		  UVD_CGC_CTRL__CLK_GATE_DLY_TIMER_MASK);
@@ -1514,18 +1553,18 @@ static void uvd_v7_0_set_sw_clock_gating(struct amdgpu_device *adev)
 			UVD_SUVD_CGC_CTRL__SDB_MODE_MASK);
 	data1 |= suvd_flags;
 
-	WREG32_SOC15(UVD, 0, mmUVD_CGC_CTRL, data);
-	WREG32_SOC15(UVD, 0, mmUVD_CGC_GATE, 0);
-	WREG32_SOC15(UVD, 0, mmUVD_SUVD_CGC_GATE, data1);
-	WREG32_SOC15(UVD, 0, mmUVD_SUVD_CGC_CTRL, data2);
+	WREG32_SOC15(UVD, ring->me, mmUVD_CGC_CTRL, data);
+	WREG32_SOC15(UVD, ring->me, mmUVD_CGC_GATE, 0);
+	WREG32_SOC15(UVD, ring->me, mmUVD_SUVD_CGC_GATE, data1);
+	WREG32_SOC15(UVD, ring->me, mmUVD_SUVD_CGC_CTRL, data2);
 }
 
 static void uvd_v7_0_set_hw_clock_gating(struct amdgpu_device *adev)
 {
 	uint32_t data, data1, cgc_flags, suvd_flags;
 
-	data = RREG32_SOC15(UVD, 0, mmUVD_CGC_GATE);
-	data1 = RREG32_SOC15(UVD, 0, mmUVD_SUVD_CGC_GATE);
+	data = RREG32_SOC15(UVD, ring->me, mmUVD_CGC_GATE);
+	data1 = RREG32_SOC15(UVD, ring->me, mmUVD_SUVD_CGC_GATE);
 
 	cgc_flags = UVD_CGC_GATE__SYS_MASK |
 		UVD_CGC_GATE__UDEC_MASK |
@@ -1557,8 +1596,8 @@ static void uvd_v7_0_set_hw_clock_gating(struct amdgpu_device *adev)
 	data |= cgc_flags;
 	data1 |= suvd_flags;
 
-	WREG32_SOC15(UVD, 0, mmUVD_CGC_GATE, data);
-	WREG32_SOC15(UVD, 0, mmUVD_SUVD_CGC_GATE, data1);
+	WREG32_SOC15(UVD, ring->me, mmUVD_CGC_GATE, data);
+	WREG32_SOC15(UVD, ring->me, mmUVD_SUVD_CGC_GATE, data1);
 }
 
 static void uvd_v7_0_set_bypass_mode(struct amdgpu_device *adev, bool enable)
@@ -1617,7 +1656,7 @@ static int uvd_v7_0_set_powergating_state(void *handle,
 	if (!(adev->pg_flags & AMD_PG_SUPPORT_UVD))
 		return 0;
 
-	WREG32_SOC15(UVD, 0, mmUVD_POWER_STATUS, UVD_POWER_STATUS__UVD_PG_EN_MASK);
+	WREG32_SOC15(UVD, ring->me, mmUVD_POWER_STATUS, UVD_POWER_STATUS__UVD_PG_EN_MASK);
 
 	if (state == AMD_PG_STATE_GATE) {
 		uvd_v7_0_stop(adev);
@@ -1720,18 +1759,27 @@ static const struct amdgpu_ring_funcs uvd_v7_0_enc_ring_vm_funcs = {
 
 static void uvd_v7_0_set_ring_funcs(struct amdgpu_device *adev)
 {
-	adev->uvd.inst->ring.funcs = &uvd_v7_0_ring_vm_funcs;
-	DRM_INFO("UVD is enabled in VM mode\n");
+	int i;
+
+	for (i = 0; i < adev->uvd.num_uvd_inst; i++) {
+		adev->uvd.inst[i].ring.funcs = &uvd_v7_0_ring_vm_funcs;
+		adev->uvd.inst[i].ring.me = i;
+		DRM_INFO("UVD(%d) is enabled in VM mode\n", i);
+	}
 }
 
 static void uvd_v7_0_set_enc_ring_funcs(struct amdgpu_device *adev)
 {
-	int i;
+	int i, j;
 
-	for (i = 0; i < adev->uvd.num_enc_rings; ++i)
-		adev->uvd.inst->ring_enc[i].funcs = &uvd_v7_0_enc_ring_vm_funcs;
+	for (j = 0; j < adev->uvd.num_uvd_inst; j++) {
+		for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
+			adev->uvd.inst[j].ring_enc[i].funcs = &uvd_v7_0_enc_ring_vm_funcs;
+			adev->uvd.inst[j].ring_enc[i].me = j;
+		}
 
-	DRM_INFO("UVD ENC is enabled in VM mode\n");
+		DRM_INFO("UVD(%d) ENC is enabled in VM mode\n", j);
+	}
 }
 
 static const struct amdgpu_irq_src_funcs uvd_v7_0_irq_funcs = {
@@ -1741,8 +1789,12 @@ static const struct amdgpu_irq_src_funcs uvd_v7_0_irq_funcs = {
 
 static void uvd_v7_0_set_irq_funcs(struct amdgpu_device *adev)
 {
-	adev->uvd.inst->irq.num_types = adev->uvd.num_enc_rings + 1;
-	adev->uvd.inst->irq.funcs = &uvd_v7_0_irq_funcs;
+	int i;
+
+	for (i = 0; i < adev->uvd.num_uvd_inst; i++) {
+		adev->uvd.inst[i].irq.num_types = adev->uvd.num_enc_rings + 1;
+		adev->uvd.inst[i].irq.funcs = &uvd_v7_0_irq_funcs;
+	}
 }
 
 const struct amdgpu_ip_block_version uvd_v7_0_ip_block =
-- 
cgit v1.1


From 3b17c622856299a0b0eef02a409edec366a719a7 Mon Sep 17 00:00:00 2001
From: James Zhu <James.Zhu@amd.com>
Date: Mon, 23 Apr 2018 19:11:46 -0400
Subject: drm/amdgpu/vg20:increase 3 rings for AMDGPU_MAX_RINGS

For Vega20, there are two UVD Hardware. One more UVD hardware
adds one decode ring and two encode rings. So AMDGPU_MAX_RINGS
need increase by 3.

Signed-off-by: James Zhu <James.Zhu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h
index 4f8dac2..1513124c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h
@@ -29,7 +29,7 @@
 #include <drm/drm_print.h>
 
 /* max number of rings */
-#define AMDGPU_MAX_RINGS		18
+#define AMDGPU_MAX_RINGS		21
 #define AMDGPU_MAX_GFX_RINGS		1
 #define AMDGPU_MAX_COMPUTE_RINGS	8
 #define AMDGPU_MAX_VCE_RINGS		3
-- 
cgit v1.1


From 9181dba670cf0a0e8e3bda9fa66fecfe7c28b535 Mon Sep 17 00:00:00 2001
From: James Zhu <James.Zhu@amd.com>
Date: Fri, 11 May 2018 13:56:44 -0500
Subject: drm/amdgpu/vg20:Enable the 2nd instance for uvd

For Vega20, set num of uvd instance to 2, to enble 2nd instance.
The IB test build-in registers need update for vega20 2nd instance.

Signed-off-by: James Zhu <James.Zhu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c | 30 ++++++++++++++++--------------
 drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c   |  7 ++++++-
 2 files changed, 22 insertions(+), 15 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
index e961492..0772680 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
@@ -72,11 +72,12 @@
 #define FIRMWARE_VEGA12		"amdgpu/vega12_uvd.bin"
 #define FIRMWARE_VEGA20		"amdgpu/vega20_uvd.bin"
 
-#define mmUVD_GPCOM_VCPU_DATA0_VEGA10 (0x03c4 + 0x7e00)
-#define mmUVD_GPCOM_VCPU_DATA1_VEGA10 (0x03c5 + 0x7e00)
-#define mmUVD_GPCOM_VCPU_CMD_VEGA10 (0x03c3 + 0x7e00)
-#define mmUVD_NO_OP_VEGA10 (0x03ff + 0x7e00)
-#define mmUVD_ENGINE_CNTL_VEGA10 (0x03c6 + 0x7e00)
+/* These are common relative offsets for all asics, from uvd_7_0_offset.h,  */
+#define UVD_GPCOM_VCPU_CMD		0x03c3
+#define UVD_GPCOM_VCPU_DATA0	0x03c4
+#define UVD_GPCOM_VCPU_DATA1	0x03c5
+#define UVD_NO_OP				0x03ff
+#define UVD_BASE_SI				0x3800
 
 /**
  * amdgpu_uvd_cs_ctx - Command submission parser context
@@ -990,6 +991,8 @@ static int amdgpu_uvd_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo,
 	uint64_t addr;
 	long r;
 	int i;
+	unsigned offset_idx = 0;
+	unsigned offset[3] = { UVD_BASE_SI, 0, 0 };
 
 	amdgpu_bo_kunmap(bo);
 	amdgpu_bo_unpin(bo);
@@ -1009,17 +1012,16 @@ static int amdgpu_uvd_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo,
 		goto err;
 
 	if (adev->asic_type >= CHIP_VEGA10) {
-		data[0] = PACKET0(mmUVD_GPCOM_VCPU_DATA0_VEGA10, 0);
-		data[1] = PACKET0(mmUVD_GPCOM_VCPU_DATA1_VEGA10, 0);
-		data[2] = PACKET0(mmUVD_GPCOM_VCPU_CMD_VEGA10, 0);
-		data[3] = PACKET0(mmUVD_NO_OP_VEGA10, 0);
-	} else {
-		data[0] = PACKET0(mmUVD_GPCOM_VCPU_DATA0, 0);
-		data[1] = PACKET0(mmUVD_GPCOM_VCPU_DATA1, 0);
-		data[2] = PACKET0(mmUVD_GPCOM_VCPU_CMD, 0);
-		data[3] = PACKET0(mmUVD_NO_OP, 0);
+		offset_idx = 1 + ring->me;
+		offset[1] = adev->reg_offset[UVD_HWIP][0][1];
+		offset[2] = adev->reg_offset[UVD_HWIP][1][1];
 	}
 
+	data[0] = PACKET0(offset[offset_idx] + UVD_GPCOM_VCPU_DATA0, 0);
+	data[1] = PACKET0(offset[offset_idx] + UVD_GPCOM_VCPU_DATA1, 0);
+	data[2] = PACKET0(offset[offset_idx] + UVD_GPCOM_VCPU_CMD, 0);
+	data[3] = PACKET0(offset[offset_idx] + UVD_NO_OP, 0);
+
 	ib = &job->ibs[0];
 	addr = amdgpu_bo_gpu_offset(bo);
 	ib->ptr[0] = data[0];
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
index 08f3b6c..6b719e1 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
@@ -40,6 +40,8 @@
 #include "mmhub/mmhub_1_0_offset.h"
 #include "mmhub/mmhub_1_0_sh_mask.h"
 
+#define UVD7_MAX_HW_INSTANCES_VEGA20			2
+
 static void uvd_v7_0_set_ring_funcs(struct amdgpu_device *adev);
 static void uvd_v7_0_set_enc_ring_funcs(struct amdgpu_device *adev);
 static void uvd_v7_0_set_irq_funcs(struct amdgpu_device *adev);
@@ -370,7 +372,10 @@ error:
 static int uvd_v7_0_early_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	adev->uvd.num_uvd_inst = 1;
+	if (adev->asic_type == CHIP_VEGA20)
+		adev->uvd.num_uvd_inst = UVD7_MAX_HW_INSTANCES_VEGA20;
+	else
+		adev->uvd.num_uvd_inst = 1;
 
 	if (amdgpu_sriov_vf(adev))
 		adev->uvd.num_enc_rings = 1;
-- 
cgit v1.1


From 915893fd2b7bdb0e1e0a16ca402345ebc60e391b Mon Sep 17 00:00:00 2001
From: James Zhu <James.Zhu@amd.com>
Date: Mon, 23 Apr 2018 20:49:28 -0400
Subject: drm/amdgpu/vg20:Add IH client ID for the 2nd UVD

For Vega20, there are two UVD hardware. Need add
the 2nd IH client ID for the 2nd UVD Hardware.

Signed-off-by: James Zhu <James.Zhu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/include/soc15_ih_clientid.h | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/include/soc15_ih_clientid.h b/drivers/gpu/drm/amd/include/soc15_ih_clientid.h
index a12d4f2..12e196c 100644
--- a/drivers/gpu/drm/amd/include/soc15_ih_clientid.h
+++ b/drivers/gpu/drm/amd/include/soc15_ih_clientid.h
@@ -43,6 +43,7 @@ enum soc15_ih_clientid {
 	SOC15_IH_CLIENTID_SE2SH		= 0x0c,
 	SOC15_IH_CLIENTID_SE3SH		= 0x0d,
 	SOC15_IH_CLIENTID_SYSHUB	= 0x0e,
+	SOC15_IH_CLIENTID_UVD1		= 0x0e,
 	SOC15_IH_CLIENTID_THM		= 0x0f,
 	SOC15_IH_CLIENTID_UVD		= 0x10,
 	SOC15_IH_CLIENTID_VCE0		= 0x11,
-- 
cgit v1.1


From b53a6ebcc55971169e56982fd9131d1a6969a053 Mon Sep 17 00:00:00 2001
From: James Zhu <James.Zhu@amd.com>
Date: Mon, 23 Apr 2018 20:56:01 -0400
Subject: drm/amdgpu/vg20:Enable the 2nd instance IRQ for uvd 7.2

For Vega20, the 2nd instance uvd IRQ using different client id.
Enable the 2nd instance IRQ for uvd 7.2

Signed-off-by: James Zhu <James.Zhu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c | 12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
index 6b719e1..f9a54821 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
@@ -49,6 +49,11 @@ static int uvd_v7_0_start(struct amdgpu_device *adev);
 static void uvd_v7_0_stop(struct amdgpu_device *adev);
 static int uvd_v7_0_sriov_start(struct amdgpu_device *adev);
 
+static int amdgpu_ih_clientid_uvds[] = {
+	SOC15_IH_CLIENTID_UVD,
+	SOC15_IH_CLIENTID_UVD1
+};
+
 /**
  * uvd_v7_0_ring_get_rptr - get read pointer
  *
@@ -397,13 +402,13 @@ static int uvd_v7_0_sw_init(void *handle)
 
 	for (j = 0; j < adev->uvd.num_uvd_inst; j++) {
 		/* UVD TRAP */
-		r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_UVD, 124, &adev->uvd.inst[j].irq);
+		r = amdgpu_irq_add_id(adev, amdgpu_ih_clientid_uvds[j], 124, &adev->uvd.inst[j].irq);
 		if (r)
 			return r;
 
 		/* UVD ENC TRAP */
 		for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
-			r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_UVD, i + 119, &adev->uvd.inst[j].irq);
+			r = amdgpu_irq_add_id(adev, amdgpu_ih_clientid_uvds[j], i + 119, &adev->uvd.inst[j].irq);
 			if (r)
 				return r;
 		}
@@ -1480,6 +1485,9 @@ static int uvd_v7_0_process_interrupt(struct amdgpu_device *adev,
 	case SOC15_IH_CLIENTID_UVD:
 		ip_instance = 0;
 		break;
+	case SOC15_IH_CLIENTID_UVD1:
+		ip_instance = 1;
+		break;
 	default:
 		DRM_ERROR("Unhandled client id: %d\n", entry->client_id);
 		return 0;
-- 
cgit v1.1


From 04305acb9f7fc9978ed7a14bf965802c45ea9682 Mon Sep 17 00:00:00 2001
From: James Zhu <James.Zhu@amd.com>
Date: Mon, 23 Apr 2018 21:00:58 -0400
Subject: drm/amdgpu/vg20:Enable 2nd instance queue maping for uvd 7.2

Enable 2nd instance uvd queue maping for uvd 7.2. For user, only one UVD
instance presents. there is two rings for uvd decode, and
4 rings for uvd encode.

Signed-off-by: James Zhu <James.Zhu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_queue_mgr.c | 14 ++++++++++----
 1 file changed, 10 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_queue_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_queue_mgr.c
index 2458d38..8af16e8 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_queue_mgr.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_queue_mgr.c
@@ -66,6 +66,8 @@ static int amdgpu_identity_map(struct amdgpu_device *adev,
 			       u32 ring,
 			       struct amdgpu_ring **out_ring)
 {
+	u32 instance;
+
 	switch (mapper->hw_ip) {
 	case AMDGPU_HW_IP_GFX:
 		*out_ring = &adev->gfx.gfx_ring[ring];
@@ -77,13 +79,16 @@ static int amdgpu_identity_map(struct amdgpu_device *adev,
 		*out_ring = &adev->sdma.instance[ring].ring;
 		break;
 	case AMDGPU_HW_IP_UVD:
-		*out_ring = &adev->uvd.inst->ring;
+		instance = ring;
+		*out_ring = &adev->uvd.inst[instance].ring;
 		break;
 	case AMDGPU_HW_IP_VCE:
 		*out_ring = &adev->vce.ring[ring];
 		break;
 	case AMDGPU_HW_IP_UVD_ENC:
-		*out_ring = &adev->uvd.inst->ring_enc[ring];
+		instance = ring / adev->uvd.num_enc_rings;
+		*out_ring =
+		&adev->uvd.inst[instance].ring_enc[ring%adev->uvd.num_enc_rings];
 		break;
 	case AMDGPU_HW_IP_VCN_DEC:
 		*out_ring = &adev->vcn.ring_dec;
@@ -240,13 +245,14 @@ int amdgpu_queue_mgr_map(struct amdgpu_device *adev,
 		ip_num_rings = adev->sdma.num_instances;
 		break;
 	case AMDGPU_HW_IP_UVD:
-		ip_num_rings = 1;
+		ip_num_rings = adev->uvd.num_uvd_inst;
 		break;
 	case AMDGPU_HW_IP_VCE:
 		ip_num_rings = adev->vce.num_rings;
 		break;
 	case AMDGPU_HW_IP_UVD_ENC:
-		ip_num_rings = adev->uvd.num_enc_rings;
+		ip_num_rings =
+			adev->uvd.num_enc_rings * adev->uvd.num_uvd_inst;
 		break;
 	case AMDGPU_HW_IP_VCN_DEC:
 		ip_num_rings = 1;
-- 
cgit v1.1


From 705e98d77bc61d234ef5a1867acb38f6d0d40e4f Mon Sep 17 00:00:00 2001
From: James Zhu <James.Zhu@amd.com>
Date: Mon, 30 Apr 2018 08:43:12 -0400
Subject: drm/amdgpu/vg20:Enable UVD/VCE for Vega20

Vega20 ucode load type is set to AMDGPU_FW_LOAD_DIRECT for default.
So UVD/VCE needn't PSP IP block up. UVD/VCE for Vega20 can be enabled
at this moment.

Signed-off-by: James Zhu <James.Zhu@amd.com>
Reviewed-by: Leo Liu <leo.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/soc15.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index 295bc9c..987271b 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -529,10 +529,8 @@ int soc15_set_ip_blocks(struct amdgpu_device *adev)
 #endif
 		amdgpu_device_ip_block_add(adev, &gfx_v9_0_ip_block);
 		amdgpu_device_ip_block_add(adev, &sdma_v4_0_ip_block);
-		if (adev->asic_type != CHIP_VEGA20) {
-			amdgpu_device_ip_block_add(adev, &uvd_v7_0_ip_block);
-			amdgpu_device_ip_block_add(adev, &vce_v4_0_ip_block);
-		}
+		amdgpu_device_ip_block_add(adev, &uvd_v7_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &vce_v4_0_ip_block);
 		break;
 	case CHIP_RAVEN:
 		amdgpu_device_ip_block_add(adev, &vega10_common_ip_block);
-- 
cgit v1.1


From 9883e9d751dad05e8c3ad3c6b769dafc60762c38 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Mon, 14 May 2018 11:50:46 -0500
Subject: drm/amdgpu: add df 3.6 headers
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Needed for vega20.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../drm/amd/include/asic_reg/df/df_3_6_default.h   | 26 ++++++++++++
 .../drm/amd/include/asic_reg/df/df_3_6_offset.h    | 33 +++++++++++++++
 .../drm/amd/include/asic_reg/df/df_3_6_sh_mask.h   | 48 ++++++++++++++++++++++
 3 files changed, 107 insertions(+)
 create mode 100644 drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_default.h
 create mode 100644 drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_offset.h
 create mode 100644 drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_sh_mask.h

diff --git a/drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_default.h b/drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_default.h
new file mode 100644
index 0000000..e58c207
--- /dev/null
+++ b/drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_default.h
@@ -0,0 +1,26 @@
+/*
+ * Copyright (C) 2018  Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included
+ * in all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
+ * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN
+ * AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ */
+#ifndef _df_3_6_DEFAULT_HEADER
+#define _df_3_6_DEFAULT_HEADER
+
+#define mmFabricConfigAccessControl_DEFAULT						0x00000000
+
+#endif
diff --git a/drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_offset.h b/drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_offset.h
new file mode 100644
index 0000000..a9575db
--- /dev/null
+++ b/drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_offset.h
@@ -0,0 +1,33 @@
+/*
+ * Copyright (C) 2018  Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included
+ * in all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
+ * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN
+ * AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ */
+#ifndef _df_3_6_OFFSET_HEADER
+#define _df_3_6_OFFSET_HEADER
+
+#define mmFabricConfigAccessControl									0x0410
+#define mmFabricConfigAccessControl_BASE_IDX								0
+
+#define mmDF_PIE_AON0_DfGlobalClkGater									0x00fc
+#define mmDF_PIE_AON0_DfGlobalClkGater_BASE_IDX								0
+
+#define mmDF_CS_UMC_AON0_DramBaseAddress0								0x0044
+#define mmDF_CS_UMC_AON0_DramBaseAddress0_BASE_IDX							0
+
+#endif
diff --git a/drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_sh_mask.h
new file mode 100644
index 0000000..88f7c69
--- /dev/null
+++ b/drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_sh_mask.h
@@ -0,0 +1,48 @@
+/*
+ * Copyright (C) 2018  Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included
+ * in all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
+ * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN
+ * AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ */
+#ifndef _df_3_6_SH_MASK_HEADER
+#define _df_3_6_SH_MASK_HEADER
+
+/* FabricConfigAccessControl */
+#define FabricConfigAccessControl__CfgRegInstAccEn__SHIFT						0x0
+#define FabricConfigAccessControl__CfgRegInstAccRegLock__SHIFT						0x1
+#define FabricConfigAccessControl__CfgRegInstID__SHIFT							0x10
+#define FabricConfigAccessControl__CfgRegInstAccEn_MASK							0x00000001L
+#define FabricConfigAccessControl__CfgRegInstAccRegLock_MASK						0x00000002L
+#define FabricConfigAccessControl__CfgRegInstID_MASK							0x00FF0000L
+
+/* DF_PIE_AON0_DfGlobalClkGater */
+#define DF_PIE_AON0_DfGlobalClkGater__MGCGMode__SHIFT							0x0
+#define DF_PIE_AON0_DfGlobalClkGater__MGCGMode_MASK							0x0000000FL
+
+/* DF_CS_AON0_DramBaseAddress0 */
+#define DF_CS_UMC_AON0_DramBaseAddress0__AddrRngVal__SHIFT						0x0
+#define DF_CS_UMC_AON0_DramBaseAddress0__LgcyMmioHoleEn__SHIFT						0x1
+#define DF_CS_UMC_AON0_DramBaseAddress0__IntLvNumChan__SHIFT						0x4
+#define DF_CS_UMC_AON0_DramBaseAddress0__IntLvAddrSel__SHIFT						0x8
+#define DF_CS_UMC_AON0_DramBaseAddress0__DramBaseAddr__SHIFT						0xc
+#define DF_CS_UMC_AON0_DramBaseAddress0__AddrRngVal_MASK						0x00000001L
+#define DF_CS_UMC_AON0_DramBaseAddress0__LgcyMmioHoleEn_MASK						0x00000002L
+#define DF_CS_UMC_AON0_DramBaseAddress0__IntLvNumChan_MASK						0x000000F0L
+#define DF_CS_UMC_AON0_DramBaseAddress0__IntLvAddrSel_MASK						0x00000700L
+#define DF_CS_UMC_AON0_DramBaseAddress0__DramBaseAddr_MASK						0xFFFFF000L
+
+#endif
-- 
cgit v1.1


From 13b581502d5101adadfb7ea269ff4c8074ba76cb Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Wed, 4 Apr 2018 14:30:28 +0800
Subject: drm/amdgpu/df: implement df v3_6 callback functions (v2)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

New df helpers for 3.6.

v2: switch to using df 3.6 headers (Alex)

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/Makefile  |   3 +-
 drivers/gpu/drm/amd/amdgpu/df_v3_6.c | 116 +++++++++++++++++++++++++++++++++++
 drivers/gpu/drm/amd/amdgpu/df_v3_6.h |  40 ++++++++++++
 3 files changed, 158 insertions(+), 1 deletion(-)
 create mode 100644 drivers/gpu/drm/amd/amdgpu/df_v3_6.c
 create mode 100644 drivers/gpu/drm/amd/amdgpu/df_v3_6.h

diff --git a/drivers/gpu/drm/amd/amdgpu/Makefile b/drivers/gpu/drm/amd/amdgpu/Makefile
index 012ea37..a51c5a9 100644
--- a/drivers/gpu/drm/amd/amdgpu/Makefile
+++ b/drivers/gpu/drm/amd/amdgpu/Makefile
@@ -67,7 +67,8 @@ amdgpu-y += \
 
 # add DF block
 amdgpu-y += \
-	df_v1_7.o
+	df_v1_7.o \
+	df_v3_6.o
 
 # add GMC block
 amdgpu-y += \
diff --git a/drivers/gpu/drm/amd/amdgpu/df_v3_6.c b/drivers/gpu/drm/amd/amdgpu/df_v3_6.c
new file mode 100644
index 0000000..60608b3
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/df_v3_6.c
@@ -0,0 +1,116 @@
+/*
+ * Copyright 2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#include "amdgpu.h"
+#include "df_v3_6.h"
+
+#include "df/df_3_6_default.h"
+#include "df/df_3_6_offset.h"
+#include "df/df_3_6_sh_mask.h"
+
+static u32 df_v3_6_channel_number[] = {1, 2, 0, 4, 0, 8, 0,
+				       16, 32, 0, 0, 0, 2, 4, 8};
+
+static void df_v3_6_init(struct amdgpu_device *adev)
+{
+}
+
+static void df_v3_6_enable_broadcast_mode(struct amdgpu_device *adev,
+					  bool enable)
+{
+	u32 tmp;
+
+	if (enable) {
+		tmp = RREG32_SOC15(DF, 0, mmFabricConfigAccessControl);
+		tmp &= ~FabricConfigAccessControl__CfgRegInstAccEn_MASK;
+		WREG32_SOC15(DF, 0, mmFabricConfigAccessControl, tmp);
+	} else
+		WREG32_SOC15(DF, 0, mmFabricConfigAccessControl,
+			     mmFabricConfigAccessControl_DEFAULT);
+}
+
+static u32 df_v3_6_get_fb_channel_number(struct amdgpu_device *adev)
+{
+	u32 tmp;
+
+	tmp = RREG32_SOC15(DF, 0, mmDF_CS_UMC_AON0_DramBaseAddress0);
+	tmp &= DF_CS_UMC_AON0_DramBaseAddress0__IntLvNumChan_MASK;
+	tmp >>= DF_CS_UMC_AON0_DramBaseAddress0__IntLvNumChan__SHIFT;
+
+	return tmp;
+}
+
+static u32 df_v3_6_get_hbm_channel_number(struct amdgpu_device *adev)
+{
+	int fb_channel_number;
+
+	fb_channel_number = adev->df_funcs->get_fb_channel_number(adev);
+	if (fb_channel_number > ARRAY_SIZE(df_v3_6_channel_number))
+		fb_channel_number = 0;
+
+	return df_v3_6_channel_number[fb_channel_number];
+}
+
+static void df_v3_6_update_medium_grain_clock_gating(struct amdgpu_device *adev,
+						     bool enable)
+{
+	u32 tmp;
+
+	/* Put DF on broadcast mode */
+	adev->df_funcs->enable_broadcast_mode(adev, true);
+
+	if (enable && (adev->cg_flags & AMD_CG_SUPPORT_DF_MGCG)) {
+		tmp = RREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater);
+		tmp &= ~DF_PIE_AON0_DfGlobalClkGater__MGCGMode_MASK;
+		tmp |= DF_V3_6_MGCG_ENABLE_15_CYCLE_DELAY;
+		WREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater, tmp);
+	} else {
+		tmp = RREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater);
+		tmp &= ~DF_PIE_AON0_DfGlobalClkGater__MGCGMode_MASK;
+		tmp |= DF_V3_6_MGCG_DISABLE;
+		WREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater, tmp);
+	}
+
+	/* Exit broadcast mode */
+	adev->df_funcs->enable_broadcast_mode(adev, false);
+}
+
+static void df_v3_6_get_clockgating_state(struct amdgpu_device *adev,
+					  u32 *flags)
+{
+	u32 tmp;
+
+	/* AMD_CG_SUPPORT_DF_MGCG */
+	tmp = RREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater);
+	if (tmp & DF_V3_6_MGCG_ENABLE_15_CYCLE_DELAY)
+		*flags |= AMD_CG_SUPPORT_DF_MGCG;
+}
+
+const struct amdgpu_df_funcs df_v3_6_funcs = {
+	.init = df_v3_6_init,
+	.enable_broadcast_mode = df_v3_6_enable_broadcast_mode,
+	.get_fb_channel_number = df_v3_6_get_fb_channel_number,
+	.get_hbm_channel_number = df_v3_6_get_hbm_channel_number,
+	.update_medium_grain_clock_gating =
+			df_v3_6_update_medium_grain_clock_gating,
+	.get_clockgating_state = df_v3_6_get_clockgating_state,
+};
diff --git a/drivers/gpu/drm/amd/amdgpu/df_v3_6.h b/drivers/gpu/drm/amd/amdgpu/df_v3_6.h
new file mode 100644
index 0000000..e79c58e
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/df_v3_6.h
@@ -0,0 +1,40 @@
+/*
+ * Copyright 2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef __DF_V3_6_H__
+#define __DF_V3_6_H__
+
+#include "soc15_common.h"
+
+enum DF_V3_6_MGCG {
+	DF_V3_6_MGCG_DISABLE = 0,
+	DF_V3_6_MGCG_ENABLE_00_CYCLE_DELAY = 1,
+	DF_V3_6_MGCG_ENABLE_01_CYCLE_DELAY = 2,
+	DF_V3_6_MGCG_ENABLE_15_CYCLE_DELAY = 13,
+	DF_V3_6_MGCG_ENABLE_31_CYCLE_DELAY = 14,
+	DF_V3_6_MGCG_ENABLE_63_CYCLE_DELAY = 15
+};
+
+extern const struct amdgpu_df_funcs df_v3_6_funcs;
+
+#endif
-- 
cgit v1.1


From 698758bbb3e3e344073f86f2d011cc536d94da49 Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Wed, 4 Apr 2018 14:32:10 +0800
Subject: drm/amdgpu: Switch to use df_v3_6_funcs for vega20 (v2)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

v2: fix whitespace (Alex)

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/soc15.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index 987271b..0e4f67e 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -51,6 +51,7 @@
 #include "gfxhub_v1_0.h"
 #include "mmhub_v1_0.h"
 #include "df_v1_7.h"
+#include "df_v3_6.h"
 #include "vega10_ih.h"
 #include "sdma_v4_0.h"
 #include "uvd_v7_0.h"
@@ -501,7 +502,10 @@ int soc15_set_ip_blocks(struct amdgpu_device *adev)
 	else
 		adev->nbio_funcs = &nbio_v6_1_funcs;
 
-	adev->df_funcs = &df_v1_7_funcs;
+	if (adev->asic_type == CHIP_VEGA20)
+		adev->df_funcs = &df_v3_6_funcs;
+	else
+		adev->df_funcs = &df_v1_7_funcs;
 	adev->nbio_funcs->detect_hw_virt(adev);
 
 	if (amdgpu_sriov_vf(adev))
-- 
cgit v1.1


From 1204a26e03a2b46917f7164e665dfc3b67a0ae1e Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Mon, 22 Jan 2018 19:08:33 +0800
Subject: drm/amdgpu: Add vega20 pci ids
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
index 739e7e0..e33e53c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
@@ -560,6 +560,13 @@ static const struct pci_device_id pciidlist[] = {
 	{0x1002, 0x69A2, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA12},
 	{0x1002, 0x69A3, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA12},
 	{0x1002, 0x69AF, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA12},
+	/* Vega 20 */
+	{0x1002, 0x66A0, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20},
+	{0x1002, 0x66A1, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20},
+	{0x1002, 0x66A2, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20},
+	{0x1002, 0x66A3, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20},
+	{0x1002, 0x66A7, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20},
+	{0x1002, 0x66AF, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20},
 	/* Raven */
 	{0x1002, 0x15dd, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RAVEN|AMD_IS_APU},
 
-- 
cgit v1.1


From 950f23ebdcfc7ca53d32d76631ba6c4e61d0f88e Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Mon, 14 May 2018 11:28:04 -0500
Subject: drm/amdgpu: flag Vega20 as experimental
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Must set amdgpu.exp_hw_support=1 on the kernel command line in
grub to enable support.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
index e33e53c..b0bf2f2 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
@@ -561,12 +561,12 @@ static const struct pci_device_id pciidlist[] = {
 	{0x1002, 0x69A3, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA12},
 	{0x1002, 0x69AF, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA12},
 	/* Vega 20 */
-	{0x1002, 0x66A0, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20},
-	{0x1002, 0x66A1, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20},
-	{0x1002, 0x66A2, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20},
-	{0x1002, 0x66A3, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20},
-	{0x1002, 0x66A7, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20},
-	{0x1002, 0x66AF, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20},
+	{0x1002, 0x66A0, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20|AMD_EXP_HW_SUPPORT},
+	{0x1002, 0x66A1, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20|AMD_EXP_HW_SUPPORT},
+	{0x1002, 0x66A2, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20|AMD_EXP_HW_SUPPORT},
+	{0x1002, 0x66A3, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20|AMD_EXP_HW_SUPPORT},
+	{0x1002, 0x66A7, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20|AMD_EXP_HW_SUPPORT},
+	{0x1002, 0x66AF, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20|AMD_EXP_HW_SUPPORT},
 	/* Raven */
 	{0x1002, 0x15dd, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RAVEN|AMD_IS_APU},
 
-- 
cgit v1.1


From 20b6b7885df58b86d9b2768852bb2c81081e2c93 Mon Sep 17 00:00:00 2001
From: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Date: Tue, 15 May 2018 14:12:21 -0400
Subject: drm/amdgpu: Skip drm_sched_entity related ops for KIQ ring.

Following change 75fbed2 we never initialize or use the GPU
scheduler for KIQ and hence we need to skip KIQ ring when iterating
amdgpu_ctx's scheduler entites.

Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c | 21 ++++++++++++++++++---
 1 file changed, 18 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
index a8e531d..c5bb362 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
@@ -173,9 +173,14 @@ static void amdgpu_ctx_do_release(struct kref *ref)
 
 	ctx = container_of(ref, struct amdgpu_ctx, refcount);
 
-	for (i = 0; i < ctx->adev->num_rings; i++)
+	for (i = 0; i < ctx->adev->num_rings; i++) {
+
+		if (ctx->adev->rings[i] == &ctx->adev->gfx.kiq.ring)
+			continue;
+
 		drm_sched_entity_fini(&ctx->adev->rings[i]->sched,
 			&ctx->rings[i].entity);
+	}
 
 	amdgpu_ctx_fini(ref);
 }
@@ -452,12 +457,17 @@ void amdgpu_ctx_mgr_entity_fini(struct amdgpu_ctx_mgr *mgr)
 		if (!ctx->adev)
 			return;
 
-		for (i = 0; i < ctx->adev->num_rings; i++)
+		for (i = 0; i < ctx->adev->num_rings; i++) {
+
+			if (ctx->adev->rings[i] == &ctx->adev->gfx.kiq.ring)
+				continue;
+
 			if (kref_read(&ctx->refcount) == 1)
 				drm_sched_entity_do_release(&ctx->adev->rings[i]->sched,
 						  &ctx->rings[i].entity);
 			else
 				DRM_ERROR("ctx %p is still alive\n", ctx);
+		}
 	}
 }
 
@@ -474,12 +484,17 @@ void amdgpu_ctx_mgr_entity_cleanup(struct amdgpu_ctx_mgr *mgr)
 		if (!ctx->adev)
 			return;
 
-		for (i = 0; i < ctx->adev->num_rings; i++)
+		for (i = 0; i < ctx->adev->num_rings; i++) {
+
+			if (ctx->adev->rings[i] == &ctx->adev->gfx.kiq.ring)
+				continue;
+
 			if (kref_read(&ctx->refcount) == 1)
 				drm_sched_entity_cleanup(&ctx->adev->rings[i]->sched,
 					&ctx->rings[i].entity);
 			else
 				DRM_ERROR("ctx %p is still alive\n", ctx);
+		}
 	}
 }
 
-- 
cgit v1.1


From 01233b8073455e5d489b95758c3afeb78ff94530 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Mon, 14 May 2018 16:03:01 +0800
Subject: drm/amd/pp: Workaround flickering issue on RV

Screen flickering observed while running 1080p video using
MPV_VAAPI/VDPAU with 4x4K@60 monitors

Need to set higher mclk in this configuration.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
index 2f69bfa..017ef2d 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
@@ -600,7 +600,10 @@ static int smu10_dpm_force_dpm_level(struct pp_hwmgr *hwmgr,
 						data->gfx_min_freq_limit/100);
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetHardMinFclkByFreq,
+						hwmgr->display_config->num_display > 3 ?
+						SMU10_UMD_PSTATE_PEAK_FCLK :
 						SMU10_UMD_PSTATE_MIN_FCLK);
+
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetHardMinSocclkByFreq,
 						SMU10_UMD_PSTATE_MIN_SOCCLK);
-- 
cgit v1.1


From b9245b949885f24e84ae16d99d3898a5f1e0ba24 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Christian=20K=C3=B6nig?= <christian.koenig@amd.com>
Date: Thu, 19 Apr 2018 09:57:21 +0200
Subject: drm/amdgpu: remove unused member
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This lock isn't used any more.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
index 4cf6786..d682708 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
@@ -187,9 +187,6 @@ struct amdgpu_vm {
 	struct amdgpu_vm_pt     root;
 	struct dma_fence	*last_update;
 
-	/* protecting freed */
-	spinlock_t		freed_lock;
-
 	/* Scheduler entity for page table updates */
 	struct drm_sched_entity	entity;
 
-- 
cgit v1.1


From 563e1e664d27292a3b55ca08366dc8c32db52450 Mon Sep 17 00:00:00 2001
From: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Date: Tue, 15 May 2018 14:42:20 -0400
Subject: drm/scheduler: Remove obsolete spinlock.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This spinlock is superfluous, any call to drm_sched_entity_push_job
should already be under a lock together with matching drm_sched_job_init
to match the order of insertion into queue with job's fence seqence
number.

v2:
Improve patch description.
Add functions documentation describing the locking considerations

Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Acked-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/scheduler/gpu_scheduler.c | 15 ++++++++++-----
 include/drm/gpu_scheduler.h               |  1 -
 2 files changed, 10 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/scheduler/gpu_scheduler.c b/drivers/gpu/drm/scheduler/gpu_scheduler.c
index a364fc0..df1578d 100644
--- a/drivers/gpu/drm/scheduler/gpu_scheduler.c
+++ b/drivers/gpu/drm/scheduler/gpu_scheduler.c
@@ -139,7 +139,6 @@ int drm_sched_entity_init(struct drm_gpu_scheduler *sched,
 	entity->last_scheduled = NULL;
 
 	spin_lock_init(&entity->rq_lock);
-	spin_lock_init(&entity->queue_lock);
 	spsc_queue_init(&entity->job_queue);
 
 	atomic_set(&entity->fence_seq, 0);
@@ -413,6 +412,10 @@ drm_sched_entity_pop_job(struct drm_sched_entity *entity)
  *
  * @sched_job		The pointer to job required to submit
  *
+ * Note: To guarantee that the order of insertion to queue matches
+ * the job's fence sequence number this function should be
+ * called with drm_sched_job_init under common lock.
+ *
  * Returns 0 for success, negative error code otherwise.
  */
 void drm_sched_entity_push_job(struct drm_sched_job *sched_job,
@@ -423,11 +426,8 @@ void drm_sched_entity_push_job(struct drm_sched_job *sched_job,
 
 	trace_drm_sched_job(sched_job, entity);
 
-	spin_lock(&entity->queue_lock);
 	first = spsc_queue_push(&entity->job_queue, &sched_job->queue_node);
 
-	spin_unlock(&entity->queue_lock);
-
 	/* first job wakes up scheduler */
 	if (first) {
 		/* Add the entity to the run queue */
@@ -593,7 +593,12 @@ void drm_sched_job_recovery(struct drm_gpu_scheduler *sched)
 }
 EXPORT_SYMBOL(drm_sched_job_recovery);
 
-/* init a sched_job with basic field */
+/**
+ * Init a sched_job with basic field
+ *
+ * Note: Refer to drm_sched_entity_push_job documentation
+ * for locking considerations.
+ */
 int drm_sched_job_init(struct drm_sched_job *job,
 		       struct drm_gpu_scheduler *sched,
 		       struct drm_sched_entity *entity,
diff --git a/include/drm/gpu_scheduler.h b/include/drm/gpu_scheduler.h
index 5238006..dec6558 100644
--- a/include/drm/gpu_scheduler.h
+++ b/include/drm/gpu_scheduler.h
@@ -56,7 +56,6 @@ struct drm_sched_entity {
 	spinlock_t			rq_lock;
 	struct drm_gpu_scheduler	*sched;
 
-	spinlock_t			queue_lock;
 	struct spsc_queue		job_queue;
 
 	atomic_t			fence_seq;
-- 
cgit v1.1


From 2b6dc93a3d439136c3fe11291a506e581b84a327 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Wed, 16 May 2018 08:39:58 -0500
Subject: drm/amdgpu/display: remove VEGAM config option
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Leftover from bringup.  No need to keep it around for
upstream.

Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/Kconfig                         | 7 -------
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c           | 4 ----
 drivers/gpu/drm/amd/display/dc/bios/command_table_helper.c  | 2 --
 drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.c | 2 --
 drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c            | 4 ----
 drivers/gpu/drm/amd/display/dc/core/dc_resource.c           | 4 ----
 drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c       | 4 ----
 drivers/gpu/drm/amd/display/dc/gpio/hw_factory.c            | 2 --
 drivers/gpu/drm/amd/display/dc/gpio/hw_translate.c          | 2 --
 drivers/gpu/drm/amd/display/dc/i2caux/i2caux.c              | 2 --
 drivers/gpu/drm/amd/display/dc/inc/dce_calcs.h              | 2 --
 drivers/gpu/drm/amd/display/include/dal_asic_id.h           | 6 +-----
 drivers/gpu/drm/amd/display/include/dal_types.h             | 2 --
 13 files changed, 1 insertion(+), 42 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/Kconfig b/drivers/gpu/drm/amd/display/Kconfig
index 6dcec9c..a0eef59 100644
--- a/drivers/gpu/drm/amd/display/Kconfig
+++ b/drivers/gpu/drm/amd/display/Kconfig
@@ -34,13 +34,6 @@ config DEBUG_KERNEL_DC
 	  if you want to hit
 	  kdgb_break in assert.
 
-config DRM_AMD_DC_VEGAM
-        bool "VEGAM support"
-        depends on DRM_AMD_DC
-        help
-         Choose this option if you want to have
-         VEGAM support for display engine
-
 config DRM_AMD_DC_VG20
 	bool "Vega20 support"
 	depends on DRM_AMD_DC
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 6f5cb26..6d0dc1f 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -1514,9 +1514,7 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS10:
 	case CHIP_POLARIS12:
-#if defined(CONFIG_DRM_AMD_DC_VEGAM)
 	case CHIP_VEGAM:
-#endif
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
 	case CHIP_VEGA20:
@@ -1710,9 +1708,7 @@ static int dm_early_init(void *handle)
 		adev->mode_info.plane_type = dm_plane_type_default;
 		break;
 	case CHIP_POLARIS10:
-#if defined(CONFIG_DRM_AMD_DC_VEGAM)
 	case CHIP_VEGAM:
-#endif
 		adev->mode_info.num_crtc = 6;
 		adev->mode_info.num_hpd = 6;
 		adev->mode_info.num_dig = 6;
diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper.c b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper.c
index be066c4..253bbb1 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper.c
@@ -51,9 +51,7 @@ bool dal_bios_parser_init_cmd_tbl_helper(
 		return true;
 
 	case DCE_VERSION_11_2:
-#if defined(CONFIG_DRM_AMD_DC_VEGAM)
 	case DCE_VERSION_11_22:
-#endif
 		*h = dal_cmd_tbl_helper_dce112_get_table();
 		return true;
 
diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.c b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.c
index 9b9e069..bbbcef5 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.c
@@ -52,9 +52,7 @@ bool dal_bios_parser_init_cmd_tbl_helper2(
 		return true;
 
 	case DCE_VERSION_11_2:
-#if defined(CONFIG_DRM_AMD_DC_VEGAM)
 	case DCE_VERSION_11_22:
-#endif
 		*h = dal_cmd_tbl_helper_dce112_get_table2();
 		return true;
 #if defined(CONFIG_DRM_AMD_DC_DCN1_0)
diff --git a/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c b/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c
index 4ee3c26..2c4e8f0 100644
--- a/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c
+++ b/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c
@@ -59,10 +59,8 @@ static enum bw_calcs_version bw_calcs_version_from_asic_id(struct hw_asic_id asi
 			return BW_CALCS_VERSION_POLARIS10;
 		if (ASIC_REV_IS_POLARIS11_M(asic_id.hw_internal_rev))
 			return BW_CALCS_VERSION_POLARIS11;
-#if defined(CONFIG_DRM_AMD_DC_VEGAM)
 		if (ASIC_REV_IS_VEGAM(asic_id.hw_internal_rev))
 			return BW_CALCS_VERSION_VEGAM;
-#endif
 		return BW_CALCS_VERSION_INVALID;
 
 	case FAMILY_AI:
@@ -2151,11 +2149,9 @@ void bw_calcs_init(struct bw_calcs_dceip *bw_dceip,
 		dceip.mcifwr_all_surfaces_burst_time = bw_int_to_fixed(0); /* todo: this is a bug*/
 		break;
 	case BW_CALCS_VERSION_POLARIS10:
-#if defined(CONFIG_DRM_AMD_DC_VEGAM)
 		/* TODO: Treat VEGAM the same as P10 for now
 		 * Need to tune the para for VEGAM if needed */
 	case BW_CALCS_VERSION_VEGAM:
-#endif
 		vbios.memory_type = bw_def_gddr5;
 		vbios.dram_channel_width_in_bits = 32;
 		vbios.number_of_dram_channels = asic_id.vram_width / vbios.dram_channel_width_in_bits;
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
index 9eb731f..345835f 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
@@ -79,10 +79,8 @@ enum dce_version resource_parse_asic_id(struct hw_asic_id asic_id)
 				ASIC_REV_IS_POLARIS12_V(asic_id.hw_internal_rev)) {
 			dc_version = DCE_VERSION_11_2;
 		}
-#if defined(CONFIG_DRM_AMD_DC_VEGAM)
 		if (ASIC_REV_IS_VEGAM(asic_id.hw_internal_rev))
 			dc_version = DCE_VERSION_11_22;
-#endif
 		break;
 	case FAMILY_AI:
 		dc_version = DCE_VERSION_12_0;
@@ -129,9 +127,7 @@ struct resource_pool *dc_create_resource_pool(
 			num_virtual_links, dc, asic_id);
 		break;
 	case DCE_VERSION_11_2:
-#if defined(CONFIG_DRM_AMD_DC_VEGAM)
 	case DCE_VERSION_11_22:
-#endif
 		res_pool = dce112_create_resource_pool(
 			num_virtual_links, dc);
 		break;
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c
index 223db98..0570e7e 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c
@@ -590,9 +590,7 @@ static uint32_t dce110_get_pix_clk_dividers(
 			pll_settings, pix_clk_params);
 		break;
 	case DCE_VERSION_11_2:
-#if defined(CONFIG_DRM_AMD_DC_VEGAM)
 	case DCE_VERSION_11_22:
-#endif
 	case DCE_VERSION_12_0:
 #if defined(CONFIG_DRM_AMD_DC_DCN1_0)
 	case DCN_VERSION_1_0:
@@ -982,9 +980,7 @@ static bool dce110_program_pix_clk(
 
 		break;
 	case DCE_VERSION_11_2:
-#if defined(CONFIG_DRM_AMD_DC_VEGAM)
 	case DCE_VERSION_11_22:
-#endif
 	case DCE_VERSION_12_0:
 #if defined(CONFIG_DRM_AMD_DC_DCN1_0)
 	case DCN_VERSION_1_0:
diff --git a/drivers/gpu/drm/amd/display/dc/gpio/hw_factory.c b/drivers/gpu/drm/amd/display/dc/gpio/hw_factory.c
index 61fe484..0caee35 100644
--- a/drivers/gpu/drm/amd/display/dc/gpio/hw_factory.c
+++ b/drivers/gpu/drm/amd/display/dc/gpio/hw_factory.c
@@ -75,9 +75,7 @@ bool dal_hw_factory_init(
 		return true;
 	case DCE_VERSION_11_0:
 	case DCE_VERSION_11_2:
-#if defined(CONFIG_DRM_AMD_DC_VEGAM)
 	case DCE_VERSION_11_22:
-#endif
 		dal_hw_factory_dce110_init(factory);
 		return true;
 	case DCE_VERSION_12_0:
diff --git a/drivers/gpu/drm/amd/display/dc/gpio/hw_translate.c b/drivers/gpu/drm/amd/display/dc/gpio/hw_translate.c
index 910ae2b7..55c7074 100644
--- a/drivers/gpu/drm/amd/display/dc/gpio/hw_translate.c
+++ b/drivers/gpu/drm/amd/display/dc/gpio/hw_translate.c
@@ -72,9 +72,7 @@ bool dal_hw_translate_init(
 	case DCE_VERSION_10_0:
 	case DCE_VERSION_11_0:
 	case DCE_VERSION_11_2:
-#if defined(CONFIG_DRM_AMD_DC_VEGAM)
 	case DCE_VERSION_11_22:
-#endif
 		dal_hw_translate_dce110_init(translate);
 		return true;
 	case DCE_VERSION_12_0:
diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/i2caux.c b/drivers/gpu/drm/amd/display/dc/i2caux/i2caux.c
index c3d7c32..14dc8c9 100644
--- a/drivers/gpu/drm/amd/display/dc/i2caux/i2caux.c
+++ b/drivers/gpu/drm/amd/display/dc/i2caux/i2caux.c
@@ -83,9 +83,7 @@ struct i2caux *dal_i2caux_create(
 	case DCE_VERSION_8_3:
 		return dal_i2caux_dce80_create(ctx);
 	case DCE_VERSION_11_2:
-#if defined(CONFIG_DRM_AMD_DC_VEGAM)
 	case DCE_VERSION_11_22:
-#endif
 		return dal_i2caux_dce112_create(ctx);
 	case DCE_VERSION_11_0:
 		return dal_i2caux_dce110_create(ctx);
diff --git a/drivers/gpu/drm/amd/display/dc/inc/dce_calcs.h b/drivers/gpu/drm/amd/display/dc/inc/dce_calcs.h
index 933ea7a..eece165 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/dce_calcs.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/dce_calcs.h
@@ -43,9 +43,7 @@ enum bw_calcs_version {
 	BW_CALCS_VERSION_POLARIS10,
 	BW_CALCS_VERSION_POLARIS11,
 	BW_CALCS_VERSION_POLARIS12,
-#if defined(CONFIG_DRM_AMD_DC_VEGAM)
 	BW_CALCS_VERSION_VEGAM,
-#endif
 	BW_CALCS_VERSION_STONEY,
 	BW_CALCS_VERSION_VEGA10
 };
diff --git a/drivers/gpu/drm/amd/display/include/dal_asic_id.h b/drivers/gpu/drm/amd/display/include/dal_asic_id.h
index 77d2856..6aeb5a2 100644
--- a/drivers/gpu/drm/amd/display/include/dal_asic_id.h
+++ b/drivers/gpu/drm/amd/display/include/dal_asic_id.h
@@ -86,6 +86,7 @@
 #define VI_POLARIS10_P_A0 80
 #define VI_POLARIS11_M_A0 90
 #define VI_POLARIS12_V_A0 100
+#define VI_VEGAM_A0 110
 
 #define VI_UNKNOWN 0xFF
 
@@ -98,14 +99,9 @@
 		(eChipRev < VI_POLARIS11_M_A0))
 #define ASIC_REV_IS_POLARIS11_M(eChipRev) ((eChipRev >= VI_POLARIS11_M_A0) &&  \
 		(eChipRev < VI_POLARIS12_V_A0))
-#if defined(CONFIG_DRM_AMD_DC_VEGAM)
-#define VI_VEGAM_A0 110
 #define ASIC_REV_IS_POLARIS12_V(eChipRev) ((eChipRev >= VI_POLARIS12_V_A0) && \
 		(eChipRev < VI_VEGAM_A0))
 #define ASIC_REV_IS_VEGAM(eChipRev) (eChipRev >= VI_VEGAM_A0)
-#else
-#define ASIC_REV_IS_POLARIS12_V(eChipRev) (eChipRev >= VI_POLARIS12_V_A0)
-#endif
 
 /* DCE11 */
 #define CZ_CARRIZO_A0 0x01
diff --git a/drivers/gpu/drm/amd/display/include/dal_types.h b/drivers/gpu/drm/amd/display/include/dal_types.h
index 5b1f8ce..840142b 100644
--- a/drivers/gpu/drm/amd/display/include/dal_types.h
+++ b/drivers/gpu/drm/amd/display/include/dal_types.h
@@ -40,9 +40,7 @@ enum dce_version {
 	DCE_VERSION_10_0,
 	DCE_VERSION_11_0,
 	DCE_VERSION_11_2,
-#if defined(CONFIG_DRM_AMD_DC_VEGAM)
 	DCE_VERSION_11_22,
-#endif
 	DCE_VERSION_12_0,
 	DCE_VERSION_MAX,
 	DCN_VERSION_1_0,
-- 
cgit v1.1


From b4b9f944e4ee3d1a268d96d7de2d519b491e8ea5 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Wed, 16 May 2018 15:28:59 -0500
Subject: drm/amdgpu/display: remove VEGA20 config option

Leftover from bringup.  No need to keep it around for
upstream.

Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/Kconfig                |   8 -
 drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c |   2 -
 drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c    |   6 -
 .../drm/amd/display/dc/dce120/dce120_resource.c    | 177 ---------------------
 drivers/gpu/drm/amd/display/include/dal_asic_id.h  |   2 -
 5 files changed, 195 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/Kconfig b/drivers/gpu/drm/amd/display/Kconfig
index a0eef59..d5d4586 100644
--- a/drivers/gpu/drm/amd/display/Kconfig
+++ b/drivers/gpu/drm/amd/display/Kconfig
@@ -34,12 +34,4 @@ config DEBUG_KERNEL_DC
 	  if you want to hit
 	  kdgb_break in assert.
 
-config DRM_AMD_DC_VG20
-	bool "Vega20 support"
-	depends on DRM_AMD_DC
-	help
-		Choose this option if you want to have
-		Vega20 support for display engine
-
-
 endmenu
diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
index 4561673..b8cef7a 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
@@ -1331,9 +1331,7 @@ static enum bp_result bios_parser_get_firmware_info(
 				result = get_firmware_info_v3_2(bp, info);
 				break;
 			case 3:
-#ifdef CONFIG_DRM_AMD_DC_VG20
 				result = get_firmware_info_v3_2(bp, info);
-#endif
 				break;
 			default:
 				break;
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c b/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c
index aa4cf30..f043e5e 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c
@@ -413,18 +413,12 @@ static int dce112_set_clock(
 	/*VBIOS will determine DPREFCLK frequency, so we don't set it*/
 	dce_clk_params.target_clock_frequency = 0;
 	dce_clk_params.clock_type = DCECLOCK_TYPE_DPREFCLK;
-#ifndef CONFIG_DRM_AMD_DC_VG20
-	dce_clk_params.flags.USE_GENLOCK_AS_SOURCE_FOR_DPREFCLK =
-			(dce_clk_params.pll_id ==
-					CLOCK_SOURCE_COMBO_DISPLAY_PLL0);
-#else
 	if (!ASICREV_IS_VEGA20_P(clk->ctx->asic_id.hw_internal_rev))
 		dce_clk_params.flags.USE_GENLOCK_AS_SOURCE_FOR_DPREFCLK =
 			(dce_clk_params.pll_id ==
 					CLOCK_SOURCE_COMBO_DISPLAY_PLL0);
 	else
 		dce_clk_params.flags.USE_GENLOCK_AS_SOURCE_FOR_DPREFCLK = false;
-#endif
 
 	bp->funcs->set_dce_clock(bp, &dce_clk_params);
 
diff --git a/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c b/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c
index 545f35f..2d58dac 100644
--- a/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c
@@ -814,7 +814,6 @@ static void bw_calcs_data_update_from_pplib(struct dc *dc)
 	dm_pp_notify_wm_clock_changes(dc->ctx, &clk_ranges);
 }
 
-#ifdef CONFIG_DRM_AMD_DC_VG20
 static uint32_t read_pipe_fuses(struct dc_context *ctx)
 {
 	uint32_t value = dm_read_reg_soc15(ctx, mmCC_DC_PIPE_DIS, 0);
@@ -1020,182 +1019,6 @@ res_create_fail:
 
 	return false;
 }
-#else
-static bool construct(
-	uint8_t num_virtual_links,
-	struct dc *dc,
-	struct dce110_resource_pool *pool)
-{
-	unsigned int i;
-	struct dc_context *ctx = dc->ctx;
-	struct irq_service_init_data irq_init_data;
-
-	ctx->dc_bios->regs = &bios_regs;
-
-	pool->base.res_cap = &res_cap;
-	pool->base.funcs = &dce120_res_pool_funcs;
-
-	/* TODO: Fill more data from GreenlandAsicCapability.cpp */
-	pool->base.pipe_count = res_cap.num_timing_generator;
-	pool->base.timing_generator_count = pool->base.res_cap->num_timing_generator;
-	pool->base.underlay_pipe_index = NO_UNDERLAY_PIPE;
-
-	dc->caps.max_downscale_ratio = 200;
-	dc->caps.i2c_speed_in_khz = 100;
-	dc->caps.max_cursor_size = 128;
-	dc->caps.dual_link_dvi = true;
-
-	dc->debug = debug_defaults;
-
-	/*************************************************
-	 *  Create resources                             *
-	 *************************************************/
-
-	pool->base.clock_sources[DCE120_CLK_SRC_PLL0] =
-			dce120_clock_source_create(ctx, ctx->dc_bios,
-				CLOCK_SOURCE_COMBO_PHY_PLL0,
-				&clk_src_regs[0], false);
-	pool->base.clock_sources[DCE120_CLK_SRC_PLL1] =
-			dce120_clock_source_create(ctx, ctx->dc_bios,
-				CLOCK_SOURCE_COMBO_PHY_PLL1,
-				&clk_src_regs[1], false);
-	pool->base.clock_sources[DCE120_CLK_SRC_PLL2] =
-			dce120_clock_source_create(ctx, ctx->dc_bios,
-				CLOCK_SOURCE_COMBO_PHY_PLL2,
-				&clk_src_regs[2], false);
-	pool->base.clock_sources[DCE120_CLK_SRC_PLL3] =
-			dce120_clock_source_create(ctx, ctx->dc_bios,
-				CLOCK_SOURCE_COMBO_PHY_PLL3,
-				&clk_src_regs[3], false);
-	pool->base.clock_sources[DCE120_CLK_SRC_PLL4] =
-			dce120_clock_source_create(ctx, ctx->dc_bios,
-				CLOCK_SOURCE_COMBO_PHY_PLL4,
-				&clk_src_regs[4], false);
-	pool->base.clock_sources[DCE120_CLK_SRC_PLL5] =
-			dce120_clock_source_create(ctx, ctx->dc_bios,
-				CLOCK_SOURCE_COMBO_PHY_PLL5,
-				&clk_src_regs[5], false);
-	pool->base.clk_src_count = DCE120_CLK_SRC_TOTAL;
-
-	pool->base.dp_clock_source =
-			dce120_clock_source_create(ctx, ctx->dc_bios,
-				CLOCK_SOURCE_ID_DP_DTO,
-				&clk_src_regs[0], true);
-
-	for (i = 0; i < pool->base.clk_src_count; i++) {
-		if (pool->base.clock_sources[i] == NULL) {
-			dm_error("DC: failed to create clock sources!\n");
-			BREAK_TO_DEBUGGER();
-			goto clk_src_create_fail;
-		}
-	}
-
-	pool->base.display_clock = dce120_disp_clk_create(ctx);
-	if (pool->base.display_clock == NULL) {
-		dm_error("DC: failed to create display clock!\n");
-		BREAK_TO_DEBUGGER();
-		goto disp_clk_create_fail;
-	}
-
-	pool->base.dmcu = dce_dmcu_create(ctx,
-			&dmcu_regs,
-			&dmcu_shift,
-			&dmcu_mask);
-	if (pool->base.dmcu == NULL) {
-		dm_error("DC: failed to create dmcu!\n");
-		BREAK_TO_DEBUGGER();
-		goto res_create_fail;
-	}
-
-	pool->base.abm = dce_abm_create(ctx,
-			&abm_regs,
-			&abm_shift,
-			&abm_mask);
-	if (pool->base.abm == NULL) {
-		dm_error("DC: failed to create abm!\n");
-		BREAK_TO_DEBUGGER();
-		goto res_create_fail;
-	}
-
-	irq_init_data.ctx = dc->ctx;
-	pool->base.irqs = dal_irq_service_dce120_create(&irq_init_data);
-	if (!pool->base.irqs)
-		goto irqs_create_fail;
-
-	for (i = 0; i < pool->base.pipe_count; i++) {
-		pool->base.timing_generators[i] =
-				dce120_timing_generator_create(
-					ctx,
-					i,
-					&dce120_tg_offsets[i]);
-		if (pool->base.timing_generators[i] == NULL) {
-			BREAK_TO_DEBUGGER();
-			dm_error("DC: failed to create tg!\n");
-			goto controller_create_fail;
-		}
-
-		pool->base.mis[i] = dce120_mem_input_create(ctx, i);
-
-		if (pool->base.mis[i] == NULL) {
-			BREAK_TO_DEBUGGER();
-			dm_error(
-				"DC: failed to create memory input!\n");
-			goto controller_create_fail;
-		}
-
-		pool->base.ipps[i] = dce120_ipp_create(ctx, i);
-		if (pool->base.ipps[i] == NULL) {
-			BREAK_TO_DEBUGGER();
-			dm_error(
-				"DC: failed to create input pixel processor!\n");
-			goto controller_create_fail;
-		}
-
-		pool->base.transforms[i] = dce120_transform_create(ctx, i);
-		if (pool->base.transforms[i] == NULL) {
-			BREAK_TO_DEBUGGER();
-			dm_error(
-				"DC: failed to create transform!\n");
-			goto res_create_fail;
-		}
-
-		pool->base.opps[i] = dce120_opp_create(
-			ctx,
-			i);
-		if (pool->base.opps[i] == NULL) {
-			BREAK_TO_DEBUGGER();
-			dm_error(
-				"DC: failed to create output pixel processor!\n");
-		}
-	}
-
-	if (!resource_construct(num_virtual_links, dc, &pool->base,
-			 &res_create_funcs))
-		goto res_create_fail;
-
-	/* Create hardware sequencer */
-	if (!dce120_hw_sequencer_create(dc))
-		goto controller_create_fail;
-
-	dc->caps.max_planes =  pool->base.pipe_count;
-
-	bw_calcs_init(dc->bw_dceip, dc->bw_vbios, dc->ctx->asic_id);
-
-	bw_calcs_data_update_from_pplib(dc);
-
-	return true;
-
-irqs_create_fail:
-controller_create_fail:
-disp_clk_create_fail:
-clk_src_create_fail:
-res_create_fail:
-
-	destruct(pool);
-
-	return false;
-}
-#endif
 
 struct resource_pool *dce120_create_resource_pool(
 	uint8_t num_virtual_links,
diff --git a/drivers/gpu/drm/amd/display/include/dal_asic_id.h b/drivers/gpu/drm/amd/display/include/dal_asic_id.h
index 6aeb5a2..cac069d 100644
--- a/drivers/gpu/drm/amd/display/include/dal_asic_id.h
+++ b/drivers/gpu/drm/amd/display/include/dal_asic_id.h
@@ -115,10 +115,8 @@
 /* DCE12 */
 #define AI_UNKNOWN 0xFF
 
-#ifdef CONFIG_DRM_AMD_DC_VG20
 #define AI_VEGA20_P_A0 40
 #define ASICREV_IS_VEGA20_P(eChipRev) ((eChipRev >= AI_VEGA20_P_A0) && (eChipRev < AI_UNKNOWN))
-#endif
 
 #define AI_GREENLAND_P_A0 1
 #define AI_GREENLAND_P_A1 2
-- 
cgit v1.1


From a1a0c40664fbd0bd1a9fa53e14ccab539005e2ca Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Wed, 16 May 2018 15:34:19 -0500
Subject: drm/amdgpu/display: fix vega12/20 handling in dal_asic_id.h

- Remove unused ASICREV_IS_VEGA12_p() macro
- Fix ASICREV_IS_VEGA12_P() macro to properly check against vega20

Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/include/dal_asic_id.h | 8 +++-----
 1 file changed, 3 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/include/dal_asic_id.h b/drivers/gpu/drm/amd/display/include/dal_asic_id.h
index cac069d..25029ed 100644
--- a/drivers/gpu/drm/amd/display/include/dal_asic_id.h
+++ b/drivers/gpu/drm/amd/display/include/dal_asic_id.h
@@ -115,19 +115,17 @@
 /* DCE12 */
 #define AI_UNKNOWN 0xFF
 
-#define AI_VEGA20_P_A0 40
-#define ASICREV_IS_VEGA20_P(eChipRev) ((eChipRev >= AI_VEGA20_P_A0) && (eChipRev < AI_UNKNOWN))
-
 #define AI_GREENLAND_P_A0 1
 #define AI_GREENLAND_P_A1 2
 #define AI_UNKNOWN 0xFF
 
 #define AI_VEGA12_P_A0 20
+#define AI_VEGA20_P_A0 40
 #define ASICREV_IS_GREENLAND_M(eChipRev)  (eChipRev < AI_VEGA12_P_A0)
 #define ASICREV_IS_GREENLAND_P(eChipRev)  (eChipRev < AI_VEGA12_P_A0)
 
-#define ASICREV_IS_VEGA12_P(eChipRev) ((eChipRev >= AI_VEGA12_P_A0) && (eChipRev < AI_UNKNOWN))
-#define ASICREV_IS_VEGA12_p(eChipRev) ((eChipRev >= AI_VEGA12_P_A0) && (eChipRev < AI_UNKNOWN))
+#define ASICREV_IS_VEGA12_P(eChipRev) ((eChipRev >= AI_VEGA12_P_A0) && (eChipRev < AI_VEGA20_P_A0))
+#define ASICREV_IS_VEGA20_P(eChipRev) ((eChipRev >= AI_VEGA20_P_A0) && (eChipRev < AI_UNKNOWN))
 
 /* DCN1_0 */
 #define INTERNAL_REV_RAVEN_A0             0x00    /* First spin of Raven */
-- 
cgit v1.1


From fa19a6e9d0e7b46bedaa526ba71ff1bf376dd93f Mon Sep 17 00:00:00 2001
From: Dan Carpenter <dan.carpenter@oracle.com>
Date: Thu, 17 May 2018 15:56:05 +0300
Subject: drm/amd/pp: missing curly braces in smu7_enable_sclk_mclk_dpm()

We added some more lines of code to this if statement but forgot to add
curly braces.

Fixes: 0c24e7ef233b ("drm/amd/powerplay: add specific changes for VEGAM in smu7_hwmgr.c")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index 8eb3f51..646c9e9 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -1018,7 +1018,7 @@ static int smu7_enable_sclk_mclk_dpm(struct pp_hwmgr *hwmgr)
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
 
 	/* enable SCLK dpm */
-	if (!data->sclk_dpm_key_disabled)
+	if (!data->sclk_dpm_key_disabled) {
 		if (hwmgr->chip_id == CHIP_VEGAM)
 			smu7_disable_sclk_vce_handshake(hwmgr);
 
@@ -1026,6 +1026,7 @@ static int smu7_enable_sclk_mclk_dpm(struct pp_hwmgr *hwmgr)
 		(0 == smum_send_msg_to_smc(hwmgr, PPSMC_MSG_DPM_Enable)),
 		"Failed to enable SCLK DPM during DPM Start Function!",
 		return -EINVAL);
+	}
 
 	/* enable MCLK dpm */
 	if (0 == data->mclk_dpm_key_disabled) {
-- 
cgit v1.1


From 50da51744f005f4afd44b69c03e6f2068abfaed8 Mon Sep 17 00:00:00 2001
From: Tom St Denis <tom.stdenis@amd.com>
Date: Wed, 9 May 2018 14:22:29 -0400
Subject: drm/amd/amdgpu:  Code comments for the amdgpu_ttm.c driver. (v2)

NFC just comments.

(v2):  Updated based on feedback from Alex Deucher.

Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 348 +++++++++++++++++++++++++++++++-
 1 file changed, 341 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index 69a2b25..e93a0a2 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -63,16 +63,44 @@ static void amdgpu_ttm_debugfs_fini(struct amdgpu_device *adev);
 /*
  * Global memory.
  */
+
+/**
+ * amdgpu_ttm_mem_global_init - Initialize and acquire reference to
+ * memory object
+ *
+ * @ref: Object for initialization.
+ *
+ * This is called by drm_global_item_ref() when an object is being
+ * initialized.
+ */
 static int amdgpu_ttm_mem_global_init(struct drm_global_reference *ref)
 {
 	return ttm_mem_global_init(ref->object);
 }
 
+/**
+ * amdgpu_ttm_mem_global_release - Drop reference to a memory object
+ *
+ * @ref: Object being removed
+ *
+ * This is called by drm_global_item_unref() when an object is being
+ * released.
+ */
 static void amdgpu_ttm_mem_global_release(struct drm_global_reference *ref)
 {
 	ttm_mem_global_release(ref->object);
 }
 
+/**
+ * amdgpu_ttm_global_init - Initialize global TTM memory reference
+ * 							structures.
+ *
+ * @adev:  	AMDGPU device for which the global structures need to be
+ *			registered.
+ *
+ * This is called as part of the AMDGPU ttm init from amdgpu_ttm_init()
+ * during bring up.
+ */
 static int amdgpu_ttm_global_init(struct amdgpu_device *adev)
 {
 	struct drm_global_reference *global_ref;
@@ -80,7 +108,9 @@ static int amdgpu_ttm_global_init(struct amdgpu_device *adev)
 	struct drm_sched_rq *rq;
 	int r;
 
+	/* ensure reference is false in case init fails */
 	adev->mman.mem_global_referenced = false;
+
 	global_ref = &adev->mman.mem_global_ref;
 	global_ref->global_type = DRM_GLOBAL_TTM_MEM;
 	global_ref->size = sizeof(struct ttm_mem_global);
@@ -146,6 +176,18 @@ static int amdgpu_invalidate_caches(struct ttm_bo_device *bdev, uint32_t flags)
 	return 0;
 }
 
+/**
+ * amdgpu_init_mem_type - 	Initialize a memory manager for a specific
+ * 							type of memory request.
+ *
+ * @bdev:	The TTM BO device object (contains a reference to
+ * 			amdgpu_device)
+ * @type:	The type of memory requested
+ * @man:
+ *
+ * This is called by ttm_bo_init_mm() when a buffer object is being
+ * initialized.
+ */
 static int amdgpu_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
 				struct ttm_mem_type_manager *man)
 {
@@ -161,6 +203,7 @@ static int amdgpu_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
 		man->default_caching = TTM_PL_FLAG_CACHED;
 		break;
 	case TTM_PL_TT:
+		/* GTT memory  */
 		man->func = &amdgpu_gtt_mgr_func;
 		man->gpu_offset = adev->gmc.gart_start;
 		man->available_caching = TTM_PL_MASK_CACHING;
@@ -193,6 +236,14 @@ static int amdgpu_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
 	return 0;
 }
 
+/**
+ * amdgpu_evict_flags - Compute placement flags
+ *
+ * @bo: The buffer object to evict
+ * @placement: Possible destination(s) for evicted BO
+ *
+ * Fill in placement data when ttm_bo_evict() is called
+ */
 static void amdgpu_evict_flags(struct ttm_buffer_object *bo,
 				struct ttm_placement *placement)
 {
@@ -204,12 +255,14 @@ static void amdgpu_evict_flags(struct ttm_buffer_object *bo,
 		.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM
 	};
 
+	/* Don't handle scatter gather BOs */
 	if (bo->type == ttm_bo_type_sg) {
 		placement->num_placement = 0;
 		placement->num_busy_placement = 0;
 		return;
 	}
 
+	/* Object isn't an AMDGPU object so ignore */
 	if (!amdgpu_ttm_bo_is_amdgpu_bo(bo)) {
 		placement->placement = &placements;
 		placement->busy_placement = &placements;
@@ -217,10 +270,12 @@ static void amdgpu_evict_flags(struct ttm_buffer_object *bo,
 		placement->num_busy_placement = 1;
 		return;
 	}
+
 	abo = ttm_to_amdgpu_bo(bo);
 	switch (bo->mem.mem_type) {
 	case TTM_PL_VRAM:
 		if (!adev->mman.buffer_funcs_enabled) {
+			/* Move to system memory */
 			amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_CPU);
 		} else if (adev->gmc.visible_vram_size < adev->gmc.real_vram_size &&
 			   !(abo->flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED) &&
@@ -238,6 +293,7 @@ static void amdgpu_evict_flags(struct ttm_buffer_object *bo,
 			abo->placement.busy_placement = &abo->placements[1];
 			abo->placement.num_busy_placement = 1;
 		} else {
+			/* Move to GTT memory */
 			amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_GTT);
 		}
 		break;
@@ -248,6 +304,15 @@ static void amdgpu_evict_flags(struct ttm_buffer_object *bo,
 	*placement = abo->placement;
 }
 
+/**
+ * amdgpu_verify_access - Verify access for a mmap call
+ *
+ * @bo:		The buffer object to map
+ * @filp:	The file pointer from the process performing the mmap
+ *
+ * This is called by ttm_bo_mmap() to verify whether a process
+ * has the right to mmap a BO to their process space.
+ */
 static int amdgpu_verify_access(struct ttm_buffer_object *bo, struct file *filp)
 {
 	struct amdgpu_bo *abo = ttm_to_amdgpu_bo(bo);
@@ -265,6 +330,15 @@ static int amdgpu_verify_access(struct ttm_buffer_object *bo, struct file *filp)
 					  filp->private_data);
 }
 
+/**
+ * amdgpu_move_null - Register memory for a buffer object
+ *
+ * @bo:			The bo to assign the memory to
+ * @new_mem:	The memory to be assigned.
+ *
+ * Assign the memory from new_mem to the memory of the buffer object
+ * bo.
+ */
 static void amdgpu_move_null(struct ttm_buffer_object *bo,
 			     struct ttm_mem_reg *new_mem)
 {
@@ -275,6 +349,10 @@ static void amdgpu_move_null(struct ttm_buffer_object *bo,
 	new_mem->mm_node = NULL;
 }
 
+/**
+ * amdgpu_mm_node_addr -	Compute the GPU relative offset of a GTT
+ * 							buffer.
+ */
 static uint64_t amdgpu_mm_node_addr(struct ttm_buffer_object *bo,
 				    struct drm_mm_node *mm_node,
 				    struct ttm_mem_reg *mem)
@@ -289,9 +367,10 @@ static uint64_t amdgpu_mm_node_addr(struct ttm_buffer_object *bo,
 }
 
 /**
- * amdgpu_find_mm_node - Helper function finds the drm_mm_node
- *  corresponding to @offset. It also modifies the offset to be
- *  within the drm_mm_node returned
+ * amdgpu_find_mm_node -	Helper function finds the drm_mm_node
+ *  						corresponding to @offset. It also modifies
+ * 							the offset to be within the drm_mm_node
+ * 							returned
  */
 static struct drm_mm_node *amdgpu_find_mm_node(struct ttm_mem_reg *mem,
 					       unsigned long *offset)
@@ -430,7 +509,12 @@ error:
 	return r;
 }
 
-
+/**
+ * amdgpu_move_blit - Copy an entire buffer to another buffer
+ *
+ * This is a helper called by amdgpu_bo_move() and
+ * amdgpu_move_vram_ram() to help move buffers to and from VRAM.
+ */
 static int amdgpu_move_blit(struct ttm_buffer_object *bo,
 			    bool evict, bool no_wait_gpu,
 			    struct ttm_mem_reg *new_mem,
@@ -465,6 +549,11 @@ error:
 	return r;
 }
 
+/**
+ * amdgpu_move_vram_ram - Copy VRAM buffer to RAM buffer
+ *
+ * Called by amdgpu_bo_move().
+ */
 static int amdgpu_move_vram_ram(struct ttm_buffer_object *bo, bool evict,
 				struct ttm_operation_ctx *ctx,
 				struct ttm_mem_reg *new_mem)
@@ -477,6 +566,8 @@ static int amdgpu_move_vram_ram(struct ttm_buffer_object *bo, bool evict,
 	int r;
 
 	adev = amdgpu_ttm_adev(bo->bdev);
+
+	/* create space/pages for new_mem in GTT space */
 	tmp_mem = *new_mem;
 	tmp_mem.mm_node = NULL;
 	placement.num_placement = 1;
@@ -491,25 +582,36 @@ static int amdgpu_move_vram_ram(struct ttm_buffer_object *bo, bool evict,
 		return r;
 	}
 
+	/* set caching flags */
 	r = ttm_tt_set_placement_caching(bo->ttm, tmp_mem.placement);
 	if (unlikely(r)) {
 		goto out_cleanup;
 	}
 
+	/* Bind the memory to the GTT space */
 	r = ttm_tt_bind(bo->ttm, &tmp_mem, ctx);
 	if (unlikely(r)) {
 		goto out_cleanup;
 	}
+
+	/* blit VRAM to GTT */
 	r = amdgpu_move_blit(bo, true, ctx->no_wait_gpu, &tmp_mem, old_mem);
 	if (unlikely(r)) {
 		goto out_cleanup;
 	}
+
+	/* move BO (in tmp_mem) to new_mem */
 	r = ttm_bo_move_ttm(bo, ctx, new_mem);
 out_cleanup:
 	ttm_bo_mem_put(bo, &tmp_mem);
 	return r;
 }
 
+/**
+ * amdgpu_move_ram_vram - Copy buffer from RAM to VRAM
+ *
+ * Called by amdgpu_bo_move().
+ */
 static int amdgpu_move_ram_vram(struct ttm_buffer_object *bo, bool evict,
 				struct ttm_operation_ctx *ctx,
 				struct ttm_mem_reg *new_mem)
@@ -522,6 +624,8 @@ static int amdgpu_move_ram_vram(struct ttm_buffer_object *bo, bool evict,
 	int r;
 
 	adev = amdgpu_ttm_adev(bo->bdev);
+
+	/* make space in GTT for old_mem buffer */
 	tmp_mem = *new_mem;
 	tmp_mem.mm_node = NULL;
 	placement.num_placement = 1;
@@ -535,10 +639,14 @@ static int amdgpu_move_ram_vram(struct ttm_buffer_object *bo, bool evict,
 	if (unlikely(r)) {
 		return r;
 	}
+
+	/* move/bind old memory to GTT space */
 	r = ttm_bo_move_ttm(bo, ctx, &tmp_mem);
 	if (unlikely(r)) {
 		goto out_cleanup;
 	}
+
+	/* copy to VRAM */
 	r = amdgpu_move_blit(bo, true, ctx->no_wait_gpu, new_mem, old_mem);
 	if (unlikely(r)) {
 		goto out_cleanup;
@@ -548,6 +656,11 @@ out_cleanup:
 	return r;
 }
 
+/**
+ * amdgpu_bo_move - Move a buffer object to a new memory location
+ *
+ * Called by ttm_bo_handle_move_mem()
+ */
 static int amdgpu_bo_move(struct ttm_buffer_object *bo, bool evict,
 			  struct ttm_operation_ctx *ctx,
 			  struct ttm_mem_reg *new_mem)
@@ -613,6 +726,11 @@ memcpy:
 	return 0;
 }
 
+/**
+ * amdgpu_ttm_io_mem_reserve - Reserve a block of memory during a fault
+ *
+ * Called by ttm_mem_io_reserve() ultimately via ttm_bo_vm_fault()
+ */
 static int amdgpu_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem)
 {
 	struct ttm_mem_type_manager *man = &bdev->man[mem->mem_type];
@@ -690,6 +808,14 @@ struct amdgpu_ttm_tt {
 	uint32_t		last_set_pages;
 };
 
+/**
+ * amdgpu_ttm_tt_get_user_pages - 	Pin pages of memory pointed to
+ * 									by a USERPTR pointer to memory
+ *
+ * Called by amdgpu_gem_userptr_ioctl() and amdgpu_cs_parser_bos().
+ * This provides a wrapper around the get_user_pages() call to provide
+ * device accessible pages that back user memory.
+ */
 int amdgpu_ttm_tt_get_user_pages(struct ttm_tt *ttm, struct page **pages)
 {
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
@@ -719,6 +845,7 @@ int amdgpu_ttm_tt_get_user_pages(struct ttm_tt *ttm, struct page **pages)
 		}
 	}
 
+	/* loop enough times using contiguous pages of memory */
 	do {
 		unsigned num_pages = ttm->num_pages - pinned;
 		uint64_t userptr = gtt->userptr + pinned * PAGE_SIZE;
@@ -757,6 +884,14 @@ release_pages:
 	return r;
 }
 
+/**
+ * amdgpu_ttm_tt_set_user_pages - 	Copy pages in, putting old pages
+ * 									as necessary.
+ *
+ * Called by amdgpu_cs_list_validate().  This creates the page list
+ * that backs user memory and will ultimately be mapped into the device
+ * address space.
+ */
 void amdgpu_ttm_tt_set_user_pages(struct ttm_tt *ttm, struct page **pages)
 {
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
@@ -771,6 +906,11 @@ void amdgpu_ttm_tt_set_user_pages(struct ttm_tt *ttm, struct page **pages)
 	}
 }
 
+/**
+ * amdgpu_ttm_tt_mark_user_page - Mark pages as dirty
+ *
+ * Called while unpinning userptr pages
+ */
 void amdgpu_ttm_tt_mark_user_pages(struct ttm_tt *ttm)
 {
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
@@ -789,7 +929,12 @@ void amdgpu_ttm_tt_mark_user_pages(struct ttm_tt *ttm)
 	}
 }
 
-/* prepare the sg table with the user pages */
+/**
+ * amdgpu_ttm_tt_pin_userptr - 	prepare the sg table with the
+ * 								user pages
+ *
+ * Called by amdgpu_ttm_backend_bind()
+ **/
 static int amdgpu_ttm_tt_pin_userptr(struct ttm_tt *ttm)
 {
 	struct amdgpu_device *adev = amdgpu_ttm_adev(ttm->bdev);
@@ -801,17 +946,20 @@ static int amdgpu_ttm_tt_pin_userptr(struct ttm_tt *ttm)
 	enum dma_data_direction direction = write ?
 		DMA_BIDIRECTIONAL : DMA_TO_DEVICE;
 
+	/* Allocate an SG array and squash pages into it */
 	r = sg_alloc_table_from_pages(ttm->sg, ttm->pages, ttm->num_pages, 0,
 				      ttm->num_pages << PAGE_SHIFT,
 				      GFP_KERNEL);
 	if (r)
 		goto release_sg;
 
+	/* Map SG to device */
 	r = -ENOMEM;
 	nents = dma_map_sg(adev->dev, ttm->sg->sgl, ttm->sg->nents, direction);
 	if (nents != ttm->sg->nents)
 		goto release_sg;
 
+	/* convert SG to linear array of pages and dma addresses */
 	drm_prime_sg_to_page_addr_arrays(ttm->sg, ttm->pages,
 					 gtt->ttm.dma_address, ttm->num_pages);
 
@@ -822,6 +970,9 @@ release_sg:
 	return r;
 }
 
+/**
+ * amdgpu_ttm_tt_unpin_userptr - Unpin and unmap userptr pages
+ */
 static void amdgpu_ttm_tt_unpin_userptr(struct ttm_tt *ttm)
 {
 	struct amdgpu_device *adev = amdgpu_ttm_adev(ttm->bdev);
@@ -835,9 +986,10 @@ static void amdgpu_ttm_tt_unpin_userptr(struct ttm_tt *ttm)
 	if (!ttm->sg->sgl)
 		return;
 
-	/* free the sg table and pages again */
+	/* unmap the pages mapped to the device */
 	dma_unmap_sg(adev->dev, ttm->sg->sgl, ttm->sg->nents, direction);
 
+	/* mark the pages as dirty */
 	amdgpu_ttm_tt_mark_user_pages(ttm);
 
 	sg_free_table(ttm->sg);
@@ -882,6 +1034,12 @@ gart_bind_fail:
 	return r;
 }
 
+/**
+ * amdgpu_ttm_backend_bind - Bind GTT memory
+ *
+ * Called by ttm_tt_bind() on behalf of ttm_bo_handle_move_mem().
+ * This handles binding GTT memory to the device address space.
+ */
 static int amdgpu_ttm_backend_bind(struct ttm_tt *ttm,
 				   struct ttm_mem_reg *bo_mem)
 {
@@ -912,7 +1070,10 @@ static int amdgpu_ttm_backend_bind(struct ttm_tt *ttm,
 		return 0;
 	}
 
+	/* compute PTE flags relevant to this BO memory */
 	flags = amdgpu_ttm_tt_pte_flags(adev, ttm, bo_mem);
+
+	/* bind pages into GART page tables */
 	gtt->offset = (u64)bo_mem->start << PAGE_SHIFT;
 	r = amdgpu_gart_bind(adev, gtt->offset, ttm->num_pages,
 		ttm->pages, gtt->ttm.dma_address, flags);
@@ -923,6 +1084,9 @@ static int amdgpu_ttm_backend_bind(struct ttm_tt *ttm,
 	return r;
 }
 
+/**
+ * amdgpu_ttm_alloc_gart - Allocate GART memory for buffer object
+ */
 int amdgpu_ttm_alloc_gart(struct ttm_buffer_object *bo)
 {
 	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->bdev);
@@ -938,6 +1102,7 @@ int amdgpu_ttm_alloc_gart(struct ttm_buffer_object *bo)
 	    amdgpu_gtt_mgr_has_gart_addr(&bo->mem))
 		return 0;
 
+	/* allocate GTT space */
 	tmp = bo->mem;
 	tmp.mm_node = NULL;
 	placement.num_placement = 1;
@@ -953,7 +1118,10 @@ int amdgpu_ttm_alloc_gart(struct ttm_buffer_object *bo)
 	if (unlikely(r))
 		return r;
 
+	/* compute PTE flags for this buffer object */
 	flags = amdgpu_ttm_tt_pte_flags(adev, bo->ttm, &tmp);
+
+	/* Bind pages */
 	gtt->offset = (u64)tmp.start << PAGE_SHIFT;
 	r = amdgpu_ttm_gart_bind(adev, bo, flags);
 	if (unlikely(r)) {
@@ -969,6 +1137,12 @@ int amdgpu_ttm_alloc_gart(struct ttm_buffer_object *bo)
 	return 0;
 }
 
+/**
+ * amdgpu_ttm_recover_gart - Rebind GTT pages
+ *
+ * Called by amdgpu_gtt_mgr_recover() from amdgpu_device_reset() to
+ * rebind GTT pages during a GPU reset.
+ */
 int amdgpu_ttm_recover_gart(struct ttm_buffer_object *tbo)
 {
 	struct amdgpu_device *adev = amdgpu_ttm_adev(tbo->bdev);
@@ -984,12 +1158,19 @@ int amdgpu_ttm_recover_gart(struct ttm_buffer_object *tbo)
 	return r;
 }
 
+/**
+ * amdgpu_ttm_backend_unbind - Unbind GTT mapped pages
+ *
+ * Called by ttm_tt_unbind() on behalf of ttm_bo_move_ttm() and
+ * ttm_tt_destroy().
+ */
 static int amdgpu_ttm_backend_unbind(struct ttm_tt *ttm)
 {
 	struct amdgpu_device *adev = amdgpu_ttm_adev(ttm->bdev);
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
 	int r;
 
+	/* if the pages have userptr pinning then clear that first */
 	if (gtt->userptr)
 		amdgpu_ttm_tt_unpin_userptr(ttm);
 
@@ -1021,6 +1202,13 @@ static struct ttm_backend_func amdgpu_backend_func = {
 	.destroy = &amdgpu_ttm_backend_destroy,
 };
 
+/**
+ * amdgpu_ttm_tt_create - Create a ttm_tt object for a given BO
+ *
+ * @bo: The buffer object to create a GTT ttm_tt object around
+ *
+ * Called by ttm_tt_create().
+ */
 static struct ttm_tt *amdgpu_ttm_tt_create(struct ttm_buffer_object *bo,
 					   uint32_t page_flags)
 {
@@ -1034,6 +1222,8 @@ static struct ttm_tt *amdgpu_ttm_tt_create(struct ttm_buffer_object *bo,
 		return NULL;
 	}
 	gtt->ttm.ttm.func = &amdgpu_backend_func;
+
+	/* allocate space for the uninitialized page entries */
 	if (ttm_sg_tt_init(&gtt->ttm, bo, page_flags)) {
 		kfree(gtt);
 		return NULL;
@@ -1041,6 +1231,12 @@ static struct ttm_tt *amdgpu_ttm_tt_create(struct ttm_buffer_object *bo,
 	return &gtt->ttm.ttm;
 }
 
+/**
+ * amdgpu_ttm_tt_populate - Map GTT pages visible to the device
+ *
+ * Map the pages of a ttm_tt object to an address space visible
+ * to the underlying device.
+ */
 static int amdgpu_ttm_tt_populate(struct ttm_tt *ttm,
 			struct ttm_operation_ctx *ctx)
 {
@@ -1048,6 +1244,7 @@ static int amdgpu_ttm_tt_populate(struct ttm_tt *ttm,
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
 	bool slave = !!(ttm->page_flags & TTM_PAGE_FLAG_SG);
 
+	/* user pages are bound by amdgpu_ttm_tt_pin_userptr() */
 	if (gtt && gtt->userptr) {
 		ttm->sg = kzalloc(sizeof(struct sg_table), GFP_KERNEL);
 		if (!ttm->sg)
@@ -1072,9 +1269,17 @@ static int amdgpu_ttm_tt_populate(struct ttm_tt *ttm,
 	}
 #endif
 
+	/* fall back to generic helper to populate the page array
+	 * and map them to the device */
 	return ttm_populate_and_map_pages(adev->dev, &gtt->ttm, ctx);
 }
 
+/**
+ * amdgpu_ttm_tt_unpopulate - unmap GTT pages and unpopulate page arrays
+ *
+ * Unmaps pages of a ttm_tt object from the device address space and
+ * unpopulates the page array backing it.
+ */
 static void amdgpu_ttm_tt_unpopulate(struct ttm_tt *ttm)
 {
 	struct amdgpu_device *adev;
@@ -1100,9 +1305,21 @@ static void amdgpu_ttm_tt_unpopulate(struct ttm_tt *ttm)
 	}
 #endif
 
+	/* fall back to generic helper to unmap and unpopulate array */
 	ttm_unmap_and_unpopulate_pages(adev->dev, &gtt->ttm);
 }
 
+/**
+ * amdgpu_ttm_tt_set_userptr -	Initialize userptr GTT ttm_tt
+ * 								for the current task
+ *
+ * @ttm: The ttm_tt object to bind this userptr object to
+ * @addr:  The address in the current tasks VM space to use
+ * @flags: Requirements of userptr object.
+ *
+ * Called by amdgpu_gem_userptr_ioctl() to bind userptr pages
+ * to current task
+ */
 int amdgpu_ttm_tt_set_userptr(struct ttm_tt *ttm, uint64_t addr,
 			      uint32_t flags)
 {
@@ -1127,6 +1344,9 @@ int amdgpu_ttm_tt_set_userptr(struct ttm_tt *ttm, uint64_t addr,
 	return 0;
 }
 
+/**
+ * amdgpu_ttm_tt_get_usermm - Return memory manager for ttm_tt object
+ */
 struct mm_struct *amdgpu_ttm_tt_get_usermm(struct ttm_tt *ttm)
 {
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
@@ -1140,6 +1360,12 @@ struct mm_struct *amdgpu_ttm_tt_get_usermm(struct ttm_tt *ttm)
 	return gtt->usertask->mm;
 }
 
+/**
+ * amdgpu_ttm_tt_affect_userptr -	Determine if a ttm_tt object lays
+ * 									inside an address range for the
+ * 									current task.
+ *
+ */
 bool amdgpu_ttm_tt_affect_userptr(struct ttm_tt *ttm, unsigned long start,
 				  unsigned long end)
 {
@@ -1150,10 +1376,16 @@ bool amdgpu_ttm_tt_affect_userptr(struct ttm_tt *ttm, unsigned long start,
 	if (gtt == NULL || !gtt->userptr)
 		return false;
 
+	/* Return false if no part of the ttm_tt object lies within
+	 * the range
+	 */
 	size = (unsigned long)gtt->ttm.ttm.num_pages * PAGE_SIZE;
 	if (gtt->userptr > end || gtt->userptr + size <= start)
 		return false;
 
+	/* Search the lists of tasks that hold this mapping and see
+	 * if current is one of them.  If it is return false.
+	 */
 	spin_lock(&gtt->guptasklock);
 	list_for_each_entry(entry, &gtt->guptasks, list) {
 		if (entry->task == current) {
@@ -1168,6 +1400,10 @@ bool amdgpu_ttm_tt_affect_userptr(struct ttm_tt *ttm, unsigned long start,
 	return true;
 }
 
+/**
+ * amdgpu_ttm_tt_userptr_invalidated -	Has the ttm_tt object been
+ * 										invalidated?
+ */
 bool amdgpu_ttm_tt_userptr_invalidated(struct ttm_tt *ttm,
 				       int *last_invalidated)
 {
@@ -1178,6 +1414,12 @@ bool amdgpu_ttm_tt_userptr_invalidated(struct ttm_tt *ttm,
 	return prev_invalidated != *last_invalidated;
 }
 
+/**
+ * amdgpu_ttm_tt_userptr_needs_pages -	Have the pages backing this
+ * 										ttm_tt object been invalidated
+ * 										since the last time they've
+ * 										been set?
+ */
 bool amdgpu_ttm_tt_userptr_needs_pages(struct ttm_tt *ttm)
 {
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
@@ -1188,6 +1430,9 @@ bool amdgpu_ttm_tt_userptr_needs_pages(struct ttm_tt *ttm)
 	return atomic_read(&gtt->mmu_invalidations) != gtt->last_set_pages;
 }
 
+/**
+ * amdgpu_ttm_tt_is_readonly - Is the ttm_tt object read only?
+ */
 bool amdgpu_ttm_tt_is_readonly(struct ttm_tt *ttm)
 {
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
@@ -1198,6 +1443,12 @@ bool amdgpu_ttm_tt_is_readonly(struct ttm_tt *ttm)
 	return !!(gtt->userflags & AMDGPU_GEM_USERPTR_READONLY);
 }
 
+/**
+ * amdgpu_ttm_tt_pte_flags - Compute PTE flags for ttm_tt object
+ *
+ * @ttm: The ttm_tt object to compute the flags for
+ * @mem: The memory registry backing this ttm_tt object
+ */
 uint64_t amdgpu_ttm_tt_pte_flags(struct amdgpu_device *adev, struct ttm_tt *ttm,
 				 struct ttm_mem_reg *mem)
 {
@@ -1222,6 +1473,16 @@ uint64_t amdgpu_ttm_tt_pte_flags(struct amdgpu_device *adev, struct ttm_tt *ttm,
 	return flags;
 }
 
+/**
+ * amdgpu_ttm_bo_eviction_valuable -	Check to see if we can evict
+ * 										a buffer object.
+ *
+ * Return true if eviction is sensible.  Called by
+ * ttm_mem_evict_first() on behalf of ttm_bo_mem_force_space()
+ * which tries to evict buffer objects until it can find space
+ * for a new object and by ttm_bo_force_list_clean() which is
+ * used to clean out a memory space.
+ */
 static bool amdgpu_ttm_bo_eviction_valuable(struct ttm_buffer_object *bo,
 					    const struct ttm_place *place)
 {
@@ -1268,6 +1529,19 @@ static bool amdgpu_ttm_bo_eviction_valuable(struct ttm_buffer_object *bo,
 	return ttm_bo_eviction_valuable(bo, place);
 }
 
+/**
+ * amdgpu_ttm_access_memory -	Read or Write memory that backs a
+ * 								buffer object.
+ *
+ * @bo:  The buffer object to read/write
+ * @offset:  Offset into buffer object
+ * @buf:  Secondary buffer to write/read from
+ * @len: Length in bytes of access
+ * @write:  true if writing
+ *
+ * This is used to access VRAM that backs a buffer object via MMIO
+ * access for debugging purposes.
+ */
 static int amdgpu_ttm_access_memory(struct ttm_buffer_object *bo,
 				    unsigned long offset,
 				    void *buf, int len, int write)
@@ -1444,13 +1718,22 @@ error_create:
 	adev->fw_vram_usage.reserved_bo = NULL;
 	return r;
 }
-
+/**
+ * amdgpu_ttm_init -	Init the memory management (ttm) as well as
+ * 						various gtt/vram related fields.
+ *
+ * This initializes all of the memory space pools that the TTM layer
+ * will need such as the GTT space (system memory mapped to the device),
+ * VRAM (on-board memory), and on-chip memories (GDS, GWS, OA) which
+ * can be mapped per VMID.
+ */
 int amdgpu_ttm_init(struct amdgpu_device *adev)
 {
 	uint64_t gtt_size;
 	int r;
 	u64 vis_vram_limit;
 
+	/* initialize global references for vram/gtt */
 	r = amdgpu_ttm_global_init(adev);
 	if (r) {
 		return r;
@@ -1471,6 +1754,7 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 	/* We opt to avoid OOM on system pages allocations */
 	adev->mman.bdev.no_retry = true;
 
+	/* Initialize VRAM pool with all of VRAM divided into pages */
 	r = ttm_bo_init_mm(&adev->mman.bdev, TTM_PL_VRAM,
 				adev->gmc.real_vram_size >> PAGE_SHIFT);
 	if (r) {
@@ -1500,6 +1784,10 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 		return r;
 	}
 
+	/* allocate memory as required for VGA
+	 * This is used for VGA emulation and pre-OS scanout buffers to
+	 * avoid display artifacts while transitioning between pre-OS
+	 * and driver.  */
 	if (adev->gmc.stolen_size) {
 		r = amdgpu_bo_create_kernel(adev, adev->gmc.stolen_size, PAGE_SIZE,
 					    AMDGPU_GEM_DOMAIN_VRAM,
@@ -1511,6 +1799,8 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 	DRM_INFO("amdgpu: %uM of VRAM memory ready\n",
 		 (unsigned) (adev->gmc.real_vram_size / (1024 * 1024)));
 
+	/* Compute GTT size, either bsaed on 3/4th the size of RAM size
+	 * or whatever the user passed on module init */
 	if (amdgpu_gtt_size == -1) {
 		struct sysinfo si;
 
@@ -1521,6 +1811,8 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 	}
 	else
 		gtt_size = (uint64_t)amdgpu_gtt_size << 20;
+
+	/* Initialize GTT memory pool */
 	r = ttm_bo_init_mm(&adev->mman.bdev, TTM_PL_TT, gtt_size >> PAGE_SHIFT);
 	if (r) {
 		DRM_ERROR("Failed initializing GTT heap.\n");
@@ -1529,6 +1821,7 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 	DRM_INFO("amdgpu: %uM of GTT memory ready.\n",
 		 (unsigned)(gtt_size / (1024 * 1024)));
 
+	/* Initialize various on-chip memory pools */
 	adev->gds.mem.total_size = adev->gds.mem.total_size << AMDGPU_GDS_SHIFT;
 	adev->gds.mem.gfx_partition_size = adev->gds.mem.gfx_partition_size << AMDGPU_GDS_SHIFT;
 	adev->gds.mem.cs_partition_size = adev->gds.mem.cs_partition_size << AMDGPU_GDS_SHIFT;
@@ -1568,6 +1861,7 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 		}
 	}
 
+	/* Register debugfs entries for amdgpu_ttm */
 	r = amdgpu_ttm_debugfs_init(adev);
 	if (r) {
 		DRM_ERROR("Failed to init debugfs\n");
@@ -1576,11 +1870,19 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 	return 0;
 }
 
+/**
+ * amdgpu_ttm_late_init -	Handle any late initialization for
+ * 							amdgpu_ttm
+ */
 void amdgpu_ttm_late_init(struct amdgpu_device *adev)
 {
+	/* return the VGA stolen memory (if any) back to VRAM */
 	amdgpu_bo_free_kernel(&adev->stolen_vga_memory, NULL, NULL);
 }
 
+/**
+ * amdgpu_ttm_fini - De-initialize the TTM memory pools
+ */
 void amdgpu_ttm_fini(struct amdgpu_device *adev)
 {
 	if (!adev->mman.initialized)
@@ -1908,6 +2210,11 @@ static const struct drm_info_list amdgpu_ttm_debugfs_list[] = {
 #endif
 };
 
+/**
+ * amdgpu_ttm_vram_read - Linear read access to VRAM
+ *
+ * Accesses VRAM via MMIO for debugging purposes.
+ */
 static ssize_t amdgpu_ttm_vram_read(struct file *f, char __user *buf,
 				    size_t size, loff_t *pos)
 {
@@ -1947,6 +2254,11 @@ static ssize_t amdgpu_ttm_vram_read(struct file *f, char __user *buf,
 	return result;
 }
 
+/**
+ * amdgpu_ttm_vram_write - Linear write access to VRAM
+ *
+ * Accesses VRAM via MMIO for debugging purposes.
+ */
 static ssize_t amdgpu_ttm_vram_write(struct file *f, const char __user *buf,
 				    size_t size, loff_t *pos)
 {
@@ -1995,6 +2307,9 @@ static const struct file_operations amdgpu_ttm_vram_fops = {
 
 #ifdef CONFIG_DRM_AMDGPU_GART_DEBUGFS
 
+/**
+ * amdgpu_ttm_gtt_read - Linear read access to GTT memory
+ */
 static ssize_t amdgpu_ttm_gtt_read(struct file *f, char __user *buf,
 				   size_t size, loff_t *pos)
 {
@@ -2042,6 +2357,13 @@ static const struct file_operations amdgpu_ttm_gtt_fops = {
 
 #endif
 
+/**
+ * amdgpu_iomem_read - Virtual read access to GPU mapped memory
+ *
+ * This function is used to read memory that has been mapped to the
+ * GPU and the known addresses are not physical addresses but instead
+ * bus addresses (e.g., what you'd put in an IB or ring buffer).
+ */
 static ssize_t amdgpu_iomem_read(struct file *f, char __user *buf,
 				 size_t size, loff_t *pos)
 {
@@ -2050,6 +2372,7 @@ static ssize_t amdgpu_iomem_read(struct file *f, char __user *buf,
 	ssize_t result = 0;
 	int r;
 
+	/* retrieve the IOMMU domain if any for this device */
 	dom = iommu_get_domain_for_dev(adev->dev);
 
 	while (size) {
@@ -2062,6 +2385,10 @@ static ssize_t amdgpu_iomem_read(struct file *f, char __user *buf,
 
 		bytes = bytes < size ? bytes : size;
 
+		/* Translate the bus address to a physical address.  If
+		 * the domain is NULL it means there is no IOMMU active
+		 * and the address translation is the identity
+		 */
 		addr = dom ? iommu_iova_to_phys(dom, addr) : addr;
 
 		pfn = addr >> PAGE_SHIFT;
@@ -2086,6 +2413,13 @@ static ssize_t amdgpu_iomem_read(struct file *f, char __user *buf,
 	return result;
 }
 
+/**
+ * amdgpu_iomem_write - Virtual write access to GPU mapped memory
+ *
+ * This function is used to write memory that has been mapped to the
+ * GPU and the known addresses are not physical addresses but instead
+ * bus addresses (e.g., what you'd put in an IB or ring buffer).
+ */
 static ssize_t amdgpu_iomem_write(struct file *f, const char __user *buf,
 				 size_t size, loff_t *pos)
 {
-- 
cgit v1.1


From 652470ac55543fbbdcbce25492a7e370d23a38a0 Mon Sep 17 00:00:00 2001
From: Nayan Deshmukh <nayan26deshmukh@gmail.com>
Date: Wed, 16 May 2018 18:54:18 +0530
Subject: drm/scheduler: fix function name prefix in comments
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

That got missed while moving the files outside of amdgpu.

Signed-off-by: Nayan Deshmukh <nayan26deshmukh@gmail.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/scheduler/sched_fence.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/scheduler/sched_fence.c b/drivers/gpu/drm/scheduler/sched_fence.c
index 69aab086..786b47f 100644
--- a/drivers/gpu/drm/scheduler/sched_fence.c
+++ b/drivers/gpu/drm/scheduler/sched_fence.c
@@ -87,7 +87,7 @@ static bool drm_sched_fence_enable_signaling(struct dma_fence *f)
 }
 
 /**
- * amd_sched_fence_free - free up the fence memory
+ * drm_sched_fence_free - free up the fence memory
  *
  * @rcu: RCU callback head
  *
@@ -103,7 +103,7 @@ static void drm_sched_fence_free(struct rcu_head *rcu)
 }
 
 /**
- * amd_sched_fence_release_scheduled - callback that fence can be freed
+ * drm_sched_fence_release_scheduled - callback that fence can be freed
  *
  * @fence: fence
  *
@@ -118,7 +118,7 @@ static void drm_sched_fence_release_scheduled(struct dma_fence *f)
 }
 
 /**
- * amd_sched_fence_release_finished - drop extra reference
+ * drm_sched_fence_release_finished - drop extra reference
  *
  * @f: fence
  *
-- 
cgit v1.1


From 6f0fd919471cf2477e86e2be9b53ecae37b0e815 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Thu, 17 May 2018 12:33:34 -0500
Subject: drm/amdgpu: count fences from all uvd instances in idle handler
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Current multi-UVD hardware uses a single clock and power source
so handle all instances in the idle handler.

Reviewed-by: James Zhu <James.Zhu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
index 0772680..be2917c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
@@ -1146,7 +1146,11 @@ static void amdgpu_uvd_idle_work_handler(struct work_struct *work)
 {
 	struct amdgpu_device *adev =
 		container_of(work, struct amdgpu_device, uvd.inst->idle_work.work);
-	unsigned fences = amdgpu_fence_count_emitted(&adev->uvd.inst->ring);
+	unsigned fences = 0, i;
+
+	for (i = 0; i < adev->uvd.num_uvd_inst; ++i) {
+		fences += amdgpu_fence_count_emitted(&adev->uvd.inst[i].ring);
+	}
 
 	if (fences == 0) {
 		if (adev->pm.dpm_enabled) {
-- 
cgit v1.1


From 4bd2c5dd763866b827dd7e95b9ea71c47fa06126 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Thu, 17 May 2018 12:45:52 -0500
Subject: drm/amdgpu: Take uvd encode rings into account in idle work (v2)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Take the encode rings into account in the idle work handler.

v2: fix typo: s/num_uvd_inst/num_enc_rings/

Reviewed-by: James Zhu <James.Zhu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
index be2917c..bcf68f8 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
@@ -1146,10 +1146,13 @@ static void amdgpu_uvd_idle_work_handler(struct work_struct *work)
 {
 	struct amdgpu_device *adev =
 		container_of(work, struct amdgpu_device, uvd.inst->idle_work.work);
-	unsigned fences = 0, i;
+	unsigned fences = 0, i, j;
 
 	for (i = 0; i < adev->uvd.num_uvd_inst; ++i) {
 		fences += amdgpu_fence_count_emitted(&adev->uvd.inst[i].ring);
+		for (j = 0; j < adev->uvd.num_enc_rings; ++j) {
+			fences += amdgpu_fence_count_emitted(&adev->uvd.inst[i].ring_enc[j]);
+		}
 	}
 
 	if (fences == 0) {
-- 
cgit v1.1


From 646e906d1d64fdc6bb1a27dac45144dfd8996071 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Thu, 17 May 2018 13:03:05 -0500
Subject: drm/amdgpu: Take vcn encode rings into account in idle work
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Take the encode rings into account in the idle work handler.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
index e5d234c..6046838 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
@@ -205,6 +205,11 @@ static void amdgpu_vcn_idle_work_handler(struct work_struct *work)
 	struct amdgpu_device *adev =
 		container_of(work, struct amdgpu_device, vcn.idle_work.work);
 	unsigned fences = amdgpu_fence_count_emitted(&adev->vcn.ring_dec);
+	unsigned i;
+
+	for (i = 0; i < adev->vcn.num_enc_rings; ++i) {
+		fences += amdgpu_fence_count_emitted(&adev->vcn.ring_enc[i]);
+	}
 
 	if (fences == 0) {
 		if (adev->pm.dpm_enabled) {
-- 
cgit v1.1


From b79655c37b209315d3b533f6d63a3d6f5fcb6f84 Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Thu, 12 Apr 2018 22:40:02 -0400
Subject: drm/amd/display: Cleanup unused SetPlaneConfig

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../amd/display/dc/dce110/dce110_hw_sequencer.c    | 69 ----------------------
 .../drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c  | 10 ----
 drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h  |  5 --
 3 files changed, 84 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
index 2288d0a..ae50042 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
@@ -2269,74 +2269,6 @@ static void program_gamut_remap(struct pipe_ctx *pipe_ctx)
 
 	pipe_ctx->plane_res.xfm->funcs->transform_set_gamut_remap(pipe_ctx->plane_res.xfm, &adjust);
 }
-
-/**
- * TODO REMOVE, USE UPDATE INSTEAD
- */
-static void set_plane_config(
-	const struct dc *dc,
-	struct pipe_ctx *pipe_ctx,
-	struct resource_context *res_ctx)
-{
-	struct mem_input *mi = pipe_ctx->plane_res.mi;
-	struct dc_plane_state *plane_state = pipe_ctx->plane_state;
-	struct xfm_grph_csc_adjustment adjust;
-	struct out_csc_color_matrix tbl_entry;
-	unsigned int i;
-
-	memset(&adjust, 0, sizeof(adjust));
-	memset(&tbl_entry, 0, sizeof(tbl_entry));
-	adjust.gamut_adjust_type = GRAPHICS_GAMUT_ADJUST_TYPE_BYPASS;
-
-	dce_enable_fe_clock(dc->hwseq, mi->inst, true);
-
-	set_default_colors(pipe_ctx);
-	if (pipe_ctx->stream->csc_color_matrix.enable_adjustment == true) {
-		tbl_entry.color_space =
-			pipe_ctx->stream->output_color_space;
-
-		for (i = 0; i < 12; i++)
-			tbl_entry.regval[i] =
-			pipe_ctx->stream->csc_color_matrix.matrix[i];
-
-		pipe_ctx->plane_res.xfm->funcs->opp_set_csc_adjustment
-				(pipe_ctx->plane_res.xfm, &tbl_entry);
-	}
-
-	if (pipe_ctx->stream->gamut_remap_matrix.enable_remap == true) {
-		adjust.gamut_adjust_type = GRAPHICS_GAMUT_ADJUST_TYPE_SW;
-
-		for (i = 0; i < CSC_TEMPERATURE_MATRIX_SIZE; i++)
-			adjust.temperature_matrix[i] =
-				pipe_ctx->stream->gamut_remap_matrix.matrix[i];
-	}
-
-	pipe_ctx->plane_res.xfm->funcs->transform_set_gamut_remap(pipe_ctx->plane_res.xfm, &adjust);
-
-	pipe_ctx->plane_res.scl_data.lb_params.alpha_en = pipe_ctx->bottom_pipe != 0;
-	program_scaler(dc, pipe_ctx);
-
-	program_surface_visibility(dc, pipe_ctx);
-
-	mi->funcs->mem_input_program_surface_config(
-			mi,
-			plane_state->format,
-			&plane_state->tiling_info,
-			&plane_state->plane_size,
-			plane_state->rotation,
-			NULL,
-			false);
-	if (mi->funcs->set_blank)
-		mi->funcs->set_blank(mi, pipe_ctx->plane_state->visible);
-
-	if (dc->config.gpu_vm_support)
-		mi->funcs->mem_input_program_pte_vm(
-				pipe_ctx->plane_res.mi,
-				plane_state->format,
-				&plane_state->tiling_info,
-				plane_state->rotation);
-}
-
 static void update_plane_addr(const struct dc *dc,
 		struct pipe_ctx *pipe_ctx)
 {
@@ -3023,7 +2955,6 @@ static const struct hw_sequencer_funcs dce110_funcs = {
 	.init_hw = init_hw,
 	.apply_ctx_to_hw = dce110_apply_ctx_to_hw,
 	.apply_ctx_for_surface = dce110_apply_ctx_for_surface,
-	.set_plane_config = set_plane_config,
 	.update_plane_addr = update_plane_addr,
 	.update_pending_status = dce110_update_pending_status,
 	.set_input_transfer_func = dce110_set_input_transfer_func,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
index 572fa60..8adb8dc 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -2487,15 +2487,6 @@ static void set_static_screen_control(struct pipe_ctx **pipe_ctx,
 			set_static_screen_control(pipe_ctx[i]->stream_res.tg, value);
 }
 
-static void set_plane_config(
-	const struct dc *dc,
-	struct pipe_ctx *pipe_ctx,
-	struct resource_context *res_ctx)
-{
-	/* TODO */
-	program_gamut_remap(pipe_ctx);
-}
-
 static void dcn10_config_stereo_parameters(
 		struct dc_stream_state *stream, struct crtc_stereo_flags *flags)
 {
@@ -2673,7 +2664,6 @@ static const struct hw_sequencer_funcs dcn10_funcs = {
 	.init_hw = dcn10_init_hw,
 	.apply_ctx_to_hw = dce110_apply_ctx_to_hw,
 	.apply_ctx_for_surface = dcn10_apply_ctx_for_surface,
-	.set_plane_config = set_plane_config,
 	.update_plane_addr = dcn10_update_plane_addr,
 	.update_dchub = dcn10_update_dchub,
 	.update_pending_status = dcn10_update_pending_status,
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
index 29abf3e..63fc6c4 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
@@ -83,11 +83,6 @@ struct hw_sequencer_funcs {
 			int num_planes,
 			struct dc_state *context);
 
-	void (*set_plane_config)(
-			const struct dc *dc,
-			struct pipe_ctx *pipe_ctx,
-			struct resource_context *res_ctx);
-
 	void (*program_gamut_remap)(
 			struct pipe_ctx *pipe_ctx);
 
-- 
cgit v1.1


From eb0e515464e4a1be730c7ac7a01c3ba04c98ea97 Mon Sep 17 00:00:00 2001
From: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Date: Wed, 18 Apr 2018 11:37:53 -0400
Subject: drm/amd/display: get rid of 32.32 unsigned fixed point

32.32 is redundant, 31.32 does everything we use 32.32 for

Signed-off-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../drm/amd/display/amdgpu_dm/amdgpu_dm_color.c    |  14 +-
 drivers/gpu/drm/amd/display/dc/basics/Makefile     |   2 +-
 drivers/gpu/drm/amd/display/dc/basics/conversion.c |  28 +-
 drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c | 176 ++++----
 drivers/gpu/drm/amd/display/dc/basics/fixpt32_32.c | 161 --------
 .../gpu/drm/amd/display/dc/calcs/custom_float.c    |  46 +--
 drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c   |   8 +-
 drivers/gpu/drm/amd/display/dc/core/dc_link.c      |  12 +-
 drivers/gpu/drm/amd/display/dc/core/dc_resource.c  |  68 ++--
 drivers/gpu/drm/amd/display/dc/dc_dp_types.h       |   2 +
 drivers/gpu/drm/amd/display/dc/dc_types.h          |   2 +-
 drivers/gpu/drm/amd/display/dc/dce/dce_abm.c       |   2 +-
 .../gpu/drm/amd/display/dc/dce/dce_clock_source.c  |  60 +--
 drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c    |  26 +-
 drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c      |   2 +-
 drivers/gpu/drm/amd/display/dc/dce/dce_ipp.c       |   6 +-
 .../gpu/drm/amd/display/dc/dce/dce_scl_filters.c   |  48 +--
 .../drm/amd/display/dc/dce/dce_stream_encoder.c    |   8 +-
 drivers/gpu/drm/amd/display/dc/dce/dce_transform.c |  26 +-
 .../amd/display/dc/dce110/dce110_hw_sequencer.c    |  36 +-
 .../drm/amd/display/dc/dce110/dce110_transform_v.c |   8 +-
 .../gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c |  86 ++--
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c   |   2 +-
 .../gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c    |   6 +-
 .../gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c  |  38 +-
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c  |   4 +-
 .../drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c  |  28 +-
 .../amd/display/dc/dcn10/dcn10_stream_encoder.c    |   8 +-
 drivers/gpu/drm/amd/display/dc/irq_types.h         |   2 +
 drivers/gpu/drm/amd/display/include/fixed31_32.h   | 118 +++---
 drivers/gpu/drm/amd/display/include/fixed32_32.h   | 129 ------
 .../drm/amd/display/modules/color/color_gamma.c    | 446 ++++++++++-----------
 32 files changed, 661 insertions(+), 947 deletions(-)
 delete mode 100644 drivers/gpu/drm/amd/display/dc/basics/fixpt32_32.c
 delete mode 100644 drivers/gpu/drm/amd/display/include/fixed32_32.h

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
index e3d90e9..b329393 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
@@ -88,9 +88,9 @@ static void __drm_lut_to_dc_gamma(struct drm_color_lut *lut,
 			g = drm_color_lut_extract(lut[i].green, 16);
 			b = drm_color_lut_extract(lut[i].blue, 16);
 
-			gamma->entries.red[i] = dal_fixed31_32_from_int(r);
-			gamma->entries.green[i] = dal_fixed31_32_from_int(g);
-			gamma->entries.blue[i] = dal_fixed31_32_from_int(b);
+			gamma->entries.red[i] = dc_fixpt_from_int(r);
+			gamma->entries.green[i] = dc_fixpt_from_int(g);
+			gamma->entries.blue[i] = dc_fixpt_from_int(b);
 		}
 		return;
 	}
@@ -101,9 +101,9 @@ static void __drm_lut_to_dc_gamma(struct drm_color_lut *lut,
 		g = drm_color_lut_extract(lut[i].green, 16);
 		b = drm_color_lut_extract(lut[i].blue, 16);
 
-		gamma->entries.red[i] = dal_fixed31_32_from_fraction(r, MAX_DRM_LUT_VALUE);
-		gamma->entries.green[i] = dal_fixed31_32_from_fraction(g, MAX_DRM_LUT_VALUE);
-		gamma->entries.blue[i] = dal_fixed31_32_from_fraction(b, MAX_DRM_LUT_VALUE);
+		gamma->entries.red[i] = dc_fixpt_from_fraction(r, MAX_DRM_LUT_VALUE);
+		gamma->entries.green[i] = dc_fixpt_from_fraction(g, MAX_DRM_LUT_VALUE);
+		gamma->entries.blue[i] = dc_fixpt_from_fraction(b, MAX_DRM_LUT_VALUE);
 	}
 }
 
@@ -208,7 +208,7 @@ void amdgpu_dm_set_ctm(struct dm_crtc_state *crtc)
 	for (i = 0; i < 12; i++) {
 		/* Skip 4th element */
 		if (i % 4 == 3) {
-			stream->gamut_remap_matrix.matrix[i] = dal_fixed31_32_zero;
+			stream->gamut_remap_matrix.matrix[i] = dc_fixpt_zero;
 			continue;
 		}
 
diff --git a/drivers/gpu/drm/amd/display/dc/basics/Makefile b/drivers/gpu/drm/amd/display/dc/basics/Makefile
index bca33bd..b49ea96 100644
--- a/drivers/gpu/drm/amd/display/dc/basics/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/basics/Makefile
@@ -24,7 +24,7 @@
 # It provides the general basic services required by other DAL
 # subcomponents.
 
-BASICS = conversion.o fixpt31_32.o fixpt32_32.o \
+BASICS = conversion.o fixpt31_32.o \
 	logger.o log_helpers.o vector.o
 
 AMD_DAL_BASICS = $(addprefix $(AMDDALPATH)/dc/basics/,$(BASICS))
diff --git a/drivers/gpu/drm/amd/display/dc/basics/conversion.c b/drivers/gpu/drm/amd/display/dc/basics/conversion.c
index 3109649..50b47f1 100644
--- a/drivers/gpu/drm/amd/display/dc/basics/conversion.c
+++ b/drivers/gpu/drm/amd/display/dc/basics/conversion.c
@@ -41,22 +41,22 @@ uint16_t fixed_point_to_int_frac(
 
 	uint16_t result;
 
-	uint16_t d = (uint16_t)dal_fixed31_32_floor(
-		dal_fixed31_32_abs(
+	uint16_t d = (uint16_t)dc_fixpt_floor(
+		dc_fixpt_abs(
 			arg));
 
 	if (d <= (uint16_t)(1 << integer_bits) - (1 / (uint16_t)divisor))
-		numerator = (uint16_t)dal_fixed31_32_round(
-			dal_fixed31_32_mul_int(
+		numerator = (uint16_t)dc_fixpt_round(
+			dc_fixpt_mul_int(
 				arg,
 				divisor));
 	else {
-		numerator = dal_fixed31_32_floor(
-			dal_fixed31_32_sub(
-				dal_fixed31_32_from_int(
+		numerator = dc_fixpt_floor(
+			dc_fixpt_sub(
+				dc_fixpt_from_int(
 					1LL << integer_bits),
-				dal_fixed31_32_recip(
-					dal_fixed31_32_from_int(
+				dc_fixpt_recip(
+					dc_fixpt_from_int(
 						divisor))));
 	}
 
@@ -66,8 +66,8 @@ uint16_t fixed_point_to_int_frac(
 		result = (uint16_t)(
 		(1 << (integer_bits + fractional_bits + 1)) + numerator);
 
-	if ((result != 0) && dal_fixed31_32_lt(
-		arg, dal_fixed31_32_zero))
+	if ((result != 0) && dc_fixpt_lt(
+		arg, dc_fixpt_zero))
 		result |= 1 << (integer_bits + fractional_bits);
 
 	return result;
@@ -84,15 +84,15 @@ void convert_float_matrix(
 	uint32_t buffer_size)
 {
 	const struct fixed31_32 min_2_13 =
-		dal_fixed31_32_from_fraction(S2D13_MIN, DIVIDER);
+		dc_fixpt_from_fraction(S2D13_MIN, DIVIDER);
 	const struct fixed31_32 max_2_13 =
-		dal_fixed31_32_from_fraction(S2D13_MAX, DIVIDER);
+		dc_fixpt_from_fraction(S2D13_MAX, DIVIDER);
 	uint32_t i;
 
 	for (i = 0; i < buffer_size; ++i) {
 		uint32_t reg_value =
 				fixed_point_to_int_frac(
-					dal_fixed31_32_clamp(
+					dc_fixpt_clamp(
 						flt[i],
 						min_2_13,
 						max_2_13),
diff --git a/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c b/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c
index 7191c32..e398ecd 100644
--- a/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c
+++ b/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c
@@ -64,7 +64,7 @@ static inline unsigned long long complete_integer_division_u64(
 #define GET_FRACTIONAL_PART(x) \
 	(FRACTIONAL_PART_MASK & (x))
 
-struct fixed31_32 dal_fixed31_32_from_fraction(
+struct fixed31_32 dc_fixpt_from_fraction(
 	long long numerator,
 	long long denominator)
 {
@@ -118,7 +118,7 @@ struct fixed31_32 dal_fixed31_32_from_fraction(
 	return res;
 }
 
-struct fixed31_32 dal_fixed31_32_from_int_nonconst(
+struct fixed31_32 dc_fixpt_from_int_nonconst(
 	long long arg)
 {
 	struct fixed31_32 res;
@@ -130,7 +130,7 @@ struct fixed31_32 dal_fixed31_32_from_int_nonconst(
 	return res;
 }
 
-struct fixed31_32 dal_fixed31_32_shl(
+struct fixed31_32 dc_fixpt_shl(
 	struct fixed31_32 arg,
 	unsigned char shift)
 {
@@ -144,7 +144,7 @@ struct fixed31_32 dal_fixed31_32_shl(
 	return res;
 }
 
-struct fixed31_32 dal_fixed31_32_add(
+struct fixed31_32 dc_fixpt_add(
 	struct fixed31_32 arg1,
 	struct fixed31_32 arg2)
 {
@@ -158,7 +158,7 @@ struct fixed31_32 dal_fixed31_32_add(
 	return res;
 }
 
-struct fixed31_32 dal_fixed31_32_sub(
+struct fixed31_32 dc_fixpt_sub(
 	struct fixed31_32 arg1,
 	struct fixed31_32 arg2)
 {
@@ -172,7 +172,7 @@ struct fixed31_32 dal_fixed31_32_sub(
 	return res;
 }
 
-struct fixed31_32 dal_fixed31_32_mul(
+struct fixed31_32 dc_fixpt_mul(
 	struct fixed31_32 arg1,
 	struct fixed31_32 arg2)
 {
@@ -213,7 +213,7 @@ struct fixed31_32 dal_fixed31_32_mul(
 	tmp = arg1_fra * arg2_fra;
 
 	tmp = (tmp >> FIXED31_32_BITS_PER_FRACTIONAL_PART) +
-		(tmp >= (unsigned long long)dal_fixed31_32_half.value);
+		(tmp >= (unsigned long long)dc_fixpt_half.value);
 
 	ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value));
 
@@ -225,7 +225,7 @@ struct fixed31_32 dal_fixed31_32_mul(
 	return res;
 }
 
-struct fixed31_32 dal_fixed31_32_sqr(
+struct fixed31_32 dc_fixpt_sqr(
 	struct fixed31_32 arg)
 {
 	struct fixed31_32 res;
@@ -257,7 +257,7 @@ struct fixed31_32 dal_fixed31_32_sqr(
 	tmp = arg_fra * arg_fra;
 
 	tmp = (tmp >> FIXED31_32_BITS_PER_FRACTIONAL_PART) +
-		(tmp >= (unsigned long long)dal_fixed31_32_half.value);
+		(tmp >= (unsigned long long)dc_fixpt_half.value);
 
 	ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value));
 
@@ -266,7 +266,7 @@ struct fixed31_32 dal_fixed31_32_sqr(
 	return res;
 }
 
-struct fixed31_32 dal_fixed31_32_recip(
+struct fixed31_32 dc_fixpt_recip(
 	struct fixed31_32 arg)
 {
 	/*
@@ -276,41 +276,41 @@ struct fixed31_32 dal_fixed31_32_recip(
 
 	ASSERT(arg.value);
 
-	return dal_fixed31_32_from_fraction(
-		dal_fixed31_32_one.value,
+	return dc_fixpt_from_fraction(
+		dc_fixpt_one.value,
 		arg.value);
 }
 
-struct fixed31_32 dal_fixed31_32_sinc(
+struct fixed31_32 dc_fixpt_sinc(
 	struct fixed31_32 arg)
 {
 	struct fixed31_32 square;
 
-	struct fixed31_32 res = dal_fixed31_32_one;
+	struct fixed31_32 res = dc_fixpt_one;
 
 	int n = 27;
 
 	struct fixed31_32 arg_norm = arg;
 
-	if (dal_fixed31_32_le(
-		dal_fixed31_32_two_pi,
-		dal_fixed31_32_abs(arg))) {
-		arg_norm = dal_fixed31_32_sub(
+	if (dc_fixpt_le(
+		dc_fixpt_two_pi,
+		dc_fixpt_abs(arg))) {
+		arg_norm = dc_fixpt_sub(
 			arg_norm,
-			dal_fixed31_32_mul_int(
-				dal_fixed31_32_two_pi,
+			dc_fixpt_mul_int(
+				dc_fixpt_two_pi,
 				(int)div64_s64(
 					arg_norm.value,
-					dal_fixed31_32_two_pi.value)));
+					dc_fixpt_two_pi.value)));
 	}
 
-	square = dal_fixed31_32_sqr(arg_norm);
+	square = dc_fixpt_sqr(arg_norm);
 
 	do {
-		res = dal_fixed31_32_sub(
-			dal_fixed31_32_one,
-			dal_fixed31_32_div_int(
-				dal_fixed31_32_mul(
+		res = dc_fixpt_sub(
+			dc_fixpt_one,
+			dc_fixpt_div_int(
+				dc_fixpt_mul(
 					square,
 					res),
 				n * (n - 1)));
@@ -319,37 +319,37 @@ struct fixed31_32 dal_fixed31_32_sinc(
 	} while (n > 2);
 
 	if (arg.value != arg_norm.value)
-		res = dal_fixed31_32_div(
-			dal_fixed31_32_mul(res, arg_norm),
+		res = dc_fixpt_div(
+			dc_fixpt_mul(res, arg_norm),
 			arg);
 
 	return res;
 }
 
-struct fixed31_32 dal_fixed31_32_sin(
+struct fixed31_32 dc_fixpt_sin(
 	struct fixed31_32 arg)
 {
-	return dal_fixed31_32_mul(
+	return dc_fixpt_mul(
 		arg,
-		dal_fixed31_32_sinc(arg));
+		dc_fixpt_sinc(arg));
 }
 
-struct fixed31_32 dal_fixed31_32_cos(
+struct fixed31_32 dc_fixpt_cos(
 	struct fixed31_32 arg)
 {
 	/* TODO implement argument normalization */
 
-	const struct fixed31_32 square = dal_fixed31_32_sqr(arg);
+	const struct fixed31_32 square = dc_fixpt_sqr(arg);
 
-	struct fixed31_32 res = dal_fixed31_32_one;
+	struct fixed31_32 res = dc_fixpt_one;
 
 	int n = 26;
 
 	do {
-		res = dal_fixed31_32_sub(
-			dal_fixed31_32_one,
-			dal_fixed31_32_div_int(
-				dal_fixed31_32_mul(
+		res = dc_fixpt_sub(
+			dc_fixpt_one,
+			dc_fixpt_div_int(
+				dc_fixpt_mul(
 					square,
 					res),
 				n * (n - 1)));
@@ -372,31 +372,31 @@ static struct fixed31_32 fixed31_32_exp_from_taylor_series(
 {
 	unsigned int n = 9;
 
-	struct fixed31_32 res = dal_fixed31_32_from_fraction(
+	struct fixed31_32 res = dc_fixpt_from_fraction(
 		n + 2,
 		n + 1);
 	/* TODO find correct res */
 
-	ASSERT(dal_fixed31_32_lt(arg, dal_fixed31_32_one));
+	ASSERT(dc_fixpt_lt(arg, dc_fixpt_one));
 
 	do
-		res = dal_fixed31_32_add(
-			dal_fixed31_32_one,
-			dal_fixed31_32_div_int(
-				dal_fixed31_32_mul(
+		res = dc_fixpt_add(
+			dc_fixpt_one,
+			dc_fixpt_div_int(
+				dc_fixpt_mul(
 					arg,
 					res),
 				n));
 	while (--n != 1);
 
-	return dal_fixed31_32_add(
-		dal_fixed31_32_one,
-		dal_fixed31_32_mul(
+	return dc_fixpt_add(
+		dc_fixpt_one,
+		dc_fixpt_mul(
 			arg,
 			res));
 }
 
-struct fixed31_32 dal_fixed31_32_exp(
+struct fixed31_32 dc_fixpt_exp(
 	struct fixed31_32 arg)
 {
 	/*
@@ -406,44 +406,44 @@ struct fixed31_32 dal_fixed31_32_exp(
 	 * where m = round(x / ln(2)), r = x - m * ln(2)
 	 */
 
-	if (dal_fixed31_32_le(
-		dal_fixed31_32_ln2_div_2,
-		dal_fixed31_32_abs(arg))) {
-		int m = dal_fixed31_32_round(
-			dal_fixed31_32_div(
+	if (dc_fixpt_le(
+		dc_fixpt_ln2_div_2,
+		dc_fixpt_abs(arg))) {
+		int m = dc_fixpt_round(
+			dc_fixpt_div(
 				arg,
-				dal_fixed31_32_ln2));
+				dc_fixpt_ln2));
 
-		struct fixed31_32 r = dal_fixed31_32_sub(
+		struct fixed31_32 r = dc_fixpt_sub(
 			arg,
-			dal_fixed31_32_mul_int(
-				dal_fixed31_32_ln2,
+			dc_fixpt_mul_int(
+				dc_fixpt_ln2,
 				m));
 
 		ASSERT(m != 0);
 
-		ASSERT(dal_fixed31_32_lt(
-			dal_fixed31_32_abs(r),
-			dal_fixed31_32_one));
+		ASSERT(dc_fixpt_lt(
+			dc_fixpt_abs(r),
+			dc_fixpt_one));
 
 		if (m > 0)
-			return dal_fixed31_32_shl(
+			return dc_fixpt_shl(
 				fixed31_32_exp_from_taylor_series(r),
 				(unsigned char)m);
 		else
-			return dal_fixed31_32_div_int(
+			return dc_fixpt_div_int(
 				fixed31_32_exp_from_taylor_series(r),
 				1LL << -m);
 	} else if (arg.value != 0)
 		return fixed31_32_exp_from_taylor_series(arg);
 	else
-		return dal_fixed31_32_one;
+		return dc_fixpt_one;
 }
 
-struct fixed31_32 dal_fixed31_32_log(
+struct fixed31_32 dc_fixpt_log(
 	struct fixed31_32 arg)
 {
-	struct fixed31_32 res = dal_fixed31_32_neg(dal_fixed31_32_one);
+	struct fixed31_32 res = dc_fixpt_neg(dc_fixpt_one);
 	/* TODO improve 1st estimation */
 
 	struct fixed31_32 error;
@@ -453,15 +453,15 @@ struct fixed31_32 dal_fixed31_32_log(
 	/* TODO if arg is zero, return -INF */
 
 	do {
-		struct fixed31_32 res1 = dal_fixed31_32_add(
-			dal_fixed31_32_sub(
+		struct fixed31_32 res1 = dc_fixpt_add(
+			dc_fixpt_sub(
 				res,
-				dal_fixed31_32_one),
-			dal_fixed31_32_div(
+				dc_fixpt_one),
+			dc_fixpt_div(
 				arg,
-				dal_fixed31_32_exp(res)));
+				dc_fixpt_exp(res)));
 
-		error = dal_fixed31_32_sub(
+		error = dc_fixpt_sub(
 			res,
 			res1);
 
@@ -472,17 +472,17 @@ struct fixed31_32 dal_fixed31_32_log(
 	return res;
 }
 
-struct fixed31_32 dal_fixed31_32_pow(
+struct fixed31_32 dc_fixpt_pow(
 	struct fixed31_32 arg1,
 	struct fixed31_32 arg2)
 {
-	return dal_fixed31_32_exp(
-		dal_fixed31_32_mul(
-			dal_fixed31_32_log(arg1),
+	return dc_fixpt_exp(
+		dc_fixpt_mul(
+			dc_fixpt_log(arg1),
 			arg2));
 }
 
-int dal_fixed31_32_floor(
+int dc_fixpt_floor(
 	struct fixed31_32 arg)
 {
 	unsigned long long arg_value = abs_i64(arg.value);
@@ -493,12 +493,12 @@ int dal_fixed31_32_floor(
 		return -(int)GET_INTEGER_PART(arg_value);
 }
 
-int dal_fixed31_32_round(
+int dc_fixpt_round(
 	struct fixed31_32 arg)
 {
 	unsigned long long arg_value = abs_i64(arg.value);
 
-	const long long summand = dal_fixed31_32_half.value;
+	const long long summand = dc_fixpt_half.value;
 
 	ASSERT(LLONG_MAX - (long long)arg_value >= summand);
 
@@ -510,13 +510,13 @@ int dal_fixed31_32_round(
 		return -(int)GET_INTEGER_PART(arg_value);
 }
 
-int dal_fixed31_32_ceil(
+int dc_fixpt_ceil(
 	struct fixed31_32 arg)
 {
 	unsigned long long arg_value = abs_i64(arg.value);
 
-	const long long summand = dal_fixed31_32_one.value -
-		dal_fixed31_32_epsilon.value;
+	const long long summand = dc_fixpt_one.value -
+		dc_fixpt_epsilon.value;
 
 	ASSERT(LLONG_MAX - (long long)arg_value >= summand);
 
@@ -531,7 +531,7 @@ int dal_fixed31_32_ceil(
 /* this function is a generic helper to translate fixed point value to
  * specified integer format that will consist of integer_bits integer part and
  * fractional_bits fractional part. For example it is used in
- * dal_fixed31_32_u2d19 to receive 2 bits integer part and 19 bits fractional
+ * dc_fixpt_u2d19 to receive 2 bits integer part and 19 bits fractional
  * part in 32 bits. It is used in hw programming (scaler)
  */
 
@@ -570,35 +570,35 @@ static inline unsigned int clamp_ux_dy(
 		return min_clamp;
 }
 
-unsigned int dal_fixed31_32_u2d19(
+unsigned int dc_fixpt_u2d19(
 	struct fixed31_32 arg)
 {
 	return ux_dy(arg.value, 2, 19);
 }
 
-unsigned int dal_fixed31_32_u0d19(
+unsigned int dc_fixpt_u0d19(
 	struct fixed31_32 arg)
 {
 	return ux_dy(arg.value, 0, 19);
 }
 
-unsigned int dal_fixed31_32_clamp_u0d14(
+unsigned int dc_fixpt_clamp_u0d14(
 	struct fixed31_32 arg)
 {
 	return clamp_ux_dy(arg.value, 0, 14, 1);
 }
 
-unsigned int dal_fixed31_32_clamp_u0d10(
+unsigned int dc_fixpt_clamp_u0d10(
 	struct fixed31_32 arg)
 {
 	return clamp_ux_dy(arg.value, 0, 10, 1);
 }
 
-int dal_fixed31_32_s4d19(
+int dc_fixpt_s4d19(
 	struct fixed31_32 arg)
 {
 	if (arg.value < 0)
-		return -(int)ux_dy(dal_fixed31_32_abs(arg).value, 4, 19);
+		return -(int)ux_dy(dc_fixpt_abs(arg).value, 4, 19);
 	else
 		return ux_dy(arg.value, 4, 19);
 }
diff --git a/drivers/gpu/drm/amd/display/dc/basics/fixpt32_32.c b/drivers/gpu/drm/amd/display/dc/basics/fixpt32_32.c
deleted file mode 100644
index 4d3aaa8..0000000
--- a/drivers/gpu/drm/amd/display/dc/basics/fixpt32_32.c
+++ /dev/null
@@ -1,161 +0,0 @@
-/*
- * Copyright 2012-15 Advanced Micro Devices, Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: AMD
- *
- */
-
-#include "dm_services.h"
-#include "include/fixed32_32.h"
-
-static uint64_t u64_div(uint64_t n, uint64_t d)
-{
-	uint32_t i = 0;
-	uint64_t r;
-	uint64_t q = div64_u64_rem(n, d, &r);
-
-	for (i = 0; i < 32; ++i) {
-		uint64_t sbit = q & (1ULL<<63);
-
-		r <<= 1;
-		r |= sbit ? 1 : 0;
-		q <<= 1;
-		if (r >= d) {
-			r -= d;
-			q |= 1;
-		}
-	}
-
-	if (2*r >= d)
-		q += 1;
-	return q;
-}
-
-struct fixed32_32 dal_fixed32_32_from_fraction(uint32_t n, uint32_t d)
-{
-	struct fixed32_32 fx;
-
-	fx.value = u64_div((uint64_t)n << 32, (uint64_t)d << 32);
-	return fx;
-}
-
-struct fixed32_32 dal_fixed32_32_add(
-	struct fixed32_32 lhs,
-	struct fixed32_32 rhs)
-{
-	struct fixed32_32 fx = {lhs.value + rhs.value};
-
-	ASSERT(fx.value >= rhs.value);
-	return fx;
-}
-
-struct fixed32_32 dal_fixed32_32_add_int(struct fixed32_32 lhs, uint32_t rhs)
-{
-	struct fixed32_32 fx = {lhs.value + ((uint64_t)rhs << 32)};
-
-	ASSERT(fx.value >= (uint64_t)rhs << 32);
-	return fx;
-
-}
-struct fixed32_32 dal_fixed32_32_sub(
-	struct fixed32_32 lhs,
-	struct fixed32_32 rhs)
-{
-	struct fixed32_32 fx;
-
-	ASSERT(lhs.value >= rhs.value);
-	fx.value = lhs.value - rhs.value;
-	return fx;
-}
-
-struct fixed32_32 dal_fixed32_32_sub_int(struct fixed32_32 lhs, uint32_t rhs)
-{
-	struct fixed32_32 fx;
-
-	ASSERT(lhs.value >= ((uint64_t)rhs<<32));
-	fx.value = lhs.value - ((uint64_t)rhs<<32);
-	return fx;
-}
-
-struct fixed32_32 dal_fixed32_32_mul(
-	struct fixed32_32 lhs,
-	struct fixed32_32 rhs)
-{
-	struct fixed32_32 fx;
-	uint64_t lhs_int = lhs.value>>32;
-	uint64_t lhs_frac = (uint32_t)lhs.value;
-	uint64_t rhs_int = rhs.value>>32;
-	uint64_t rhs_frac = (uint32_t)rhs.value;
-	uint64_t ahbh = lhs_int * rhs_int;
-	uint64_t ahbl = lhs_int * rhs_frac;
-	uint64_t albh = lhs_frac * rhs_int;
-	uint64_t albl = lhs_frac * rhs_frac;
-
-	ASSERT((ahbh>>32) == 0);
-
-	fx.value = (ahbh<<32) + ahbl + albh + (albl>>32);
-	return fx;
-
-}
-
-struct fixed32_32 dal_fixed32_32_mul_int(struct fixed32_32 lhs, uint32_t rhs)
-{
-	struct fixed32_32 fx;
-	uint64_t lhsi = (lhs.value>>32) * (uint64_t)rhs;
-	uint64_t lhsf;
-
-	ASSERT((lhsi>>32) == 0);
-	lhsf = ((uint32_t)lhs.value) * (uint64_t)rhs;
-	ASSERT((lhsi<<32) + lhsf >= lhsf);
-	fx.value = (lhsi<<32) + lhsf;
-	return fx;
-}
-
-struct fixed32_32 dal_fixed32_32_div(
-	struct fixed32_32 lhs,
-	struct fixed32_32 rhs)
-{
-	struct fixed32_32 fx;
-
-	fx.value = u64_div(lhs.value, rhs.value);
-	return fx;
-}
-
-struct fixed32_32 dal_fixed32_32_div_int(struct fixed32_32 lhs, uint32_t rhs)
-{
-	struct fixed32_32 fx;
-
-	fx.value = u64_div(lhs.value, (uint64_t)rhs << 32);
-	return fx;
-}
-
-uint32_t dal_fixed32_32_ceil(struct fixed32_32 v)
-{
-	ASSERT((uint32_t)v.value ? (v.value >> 32) + 1 >= 1 : true);
-	return (v.value>>32) + ((uint32_t)v.value ? 1 : 0);
-}
-
-uint32_t dal_fixed32_32_round(struct fixed32_32 v)
-{
-	ASSERT(v.value + (1ULL<<31) >= (1ULL<<31));
-	return (v.value + (1ULL<<31))>>32;
-}
-
diff --git a/drivers/gpu/drm/amd/display/dc/calcs/custom_float.c b/drivers/gpu/drm/amd/display/dc/calcs/custom_float.c
index 7243c37..31d167b 100644
--- a/drivers/gpu/drm/amd/display/dc/calcs/custom_float.c
+++ b/drivers/gpu/drm/amd/display/dc/calcs/custom_float.c
@@ -36,41 +36,41 @@ static bool build_custom_float(
 	uint32_t exp_offset = (1 << (format->exponenta_bits - 1)) - 1;
 
 	const struct fixed31_32 mantissa_constant_plus_max_fraction =
-		dal_fixed31_32_from_fraction(
+		dc_fixpt_from_fraction(
 			(1LL << (format->mantissa_bits + 1)) - 1,
 			1LL << format->mantissa_bits);
 
 	struct fixed31_32 mantiss;
 
-	if (dal_fixed31_32_eq(
+	if (dc_fixpt_eq(
 		value,
-		dal_fixed31_32_zero)) {
+		dc_fixpt_zero)) {
 		*negative = false;
 		*mantissa = 0;
 		*exponenta = 0;
 		return true;
 	}
 
-	if (dal_fixed31_32_lt(
+	if (dc_fixpt_lt(
 		value,
-		dal_fixed31_32_zero)) {
+		dc_fixpt_zero)) {
 		*negative = format->sign;
-		value = dal_fixed31_32_neg(value);
+		value = dc_fixpt_neg(value);
 	} else {
 		*negative = false;
 	}
 
-	if (dal_fixed31_32_lt(
+	if (dc_fixpt_lt(
 		value,
-		dal_fixed31_32_one)) {
+		dc_fixpt_one)) {
 		uint32_t i = 1;
 
 		do {
-			value = dal_fixed31_32_shl(value, 1);
+			value = dc_fixpt_shl(value, 1);
 			++i;
-		} while (dal_fixed31_32_lt(
+		} while (dc_fixpt_lt(
 			value,
-			dal_fixed31_32_one));
+			dc_fixpt_one));
 
 		--i;
 
@@ -81,15 +81,15 @@ static bool build_custom_float(
 		}
 
 		*exponenta = exp_offset - i;
-	} else if (dal_fixed31_32_le(
+	} else if (dc_fixpt_le(
 		mantissa_constant_plus_max_fraction,
 		value)) {
 		uint32_t i = 1;
 
 		do {
-			value = dal_fixed31_32_shr(value, 1);
+			value = dc_fixpt_shr(value, 1);
 			++i;
-		} while (dal_fixed31_32_lt(
+		} while (dc_fixpt_lt(
 			mantissa_constant_plus_max_fraction,
 			value));
 
@@ -98,23 +98,23 @@ static bool build_custom_float(
 		*exponenta = exp_offset;
 	}
 
-	mantiss = dal_fixed31_32_sub(
+	mantiss = dc_fixpt_sub(
 		value,
-		dal_fixed31_32_one);
+		dc_fixpt_one);
 
-	if (dal_fixed31_32_lt(
+	if (dc_fixpt_lt(
 			mantiss,
-			dal_fixed31_32_zero) ||
-		dal_fixed31_32_lt(
-			dal_fixed31_32_one,
+			dc_fixpt_zero) ||
+		dc_fixpt_lt(
+			dc_fixpt_one,
 			mantiss))
-		mantiss = dal_fixed31_32_zero;
+		mantiss = dc_fixpt_zero;
 	else
-		mantiss = dal_fixed31_32_shl(
+		mantiss = dc_fixpt_shl(
 			mantiss,
 			format->mantissa_bits);
 
-	*mantissa = dal_fixed31_32_floor(mantiss);
+	*mantissa = dc_fixpt_floor(mantiss);
 
 	return true;
 }
diff --git a/drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c b/drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c
index a102c19..49a4ea4 100644
--- a/drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c
+++ b/drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c
@@ -873,14 +873,14 @@ bool dcn_validate_bandwidth(
 			}
 
 			if (pipe->plane_state->rotation % 2 == 0) {
-				ASSERT(pipe->plane_res.scl_data.ratios.horz.value != dal_fixed31_32_one.value
+				ASSERT(pipe->plane_res.scl_data.ratios.horz.value != dc_fixpt_one.value
 					|| v->scaler_rec_out_width[input_idx] == v->viewport_width[input_idx]);
-				ASSERT(pipe->plane_res.scl_data.ratios.vert.value != dal_fixed31_32_one.value
+				ASSERT(pipe->plane_res.scl_data.ratios.vert.value != dc_fixpt_one.value
 					|| v->scaler_recout_height[input_idx] == v->viewport_height[input_idx]);
 			} else {
-				ASSERT(pipe->plane_res.scl_data.ratios.horz.value != dal_fixed31_32_one.value
+				ASSERT(pipe->plane_res.scl_data.ratios.horz.value != dc_fixpt_one.value
 					|| v->scaler_recout_height[input_idx] == v->viewport_width[input_idx]);
-				ASSERT(pipe->plane_res.scl_data.ratios.vert.value != dal_fixed31_32_one.value
+				ASSERT(pipe->plane_res.scl_data.ratios.vert.value != dc_fixpt_one.value
 					|| v->scaler_rec_out_width[input_idx] == v->viewport_height[input_idx]);
 			}
 			v->dcc_enable[input_idx] = pipe->plane_state->dcc.enable ? dcn_bw_yes : dcn_bw_no;
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
index b44cf52..ea5d5ff 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
@@ -631,7 +631,7 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 				/* Need to setup mst link_cap struct here
 				 * otherwise dc_link_detect() will leave mst link_cap
 				 * empty which leads to allocate_mst_payload() has "0"
-				 * pbn_per_slot value leading to exception on dal_fixed31_32_div()
+				 * pbn_per_slot value leading to exception on dc_fixpt_div()
 				 */
 				link->verified_link_cap = link->reported_link_cap;
 				return false;
@@ -2059,10 +2059,10 @@ static struct fixed31_32 get_pbn_per_slot(struct dc_stream_state *stream)
 			&stream->sink->link->cur_link_settings;
 	uint32_t link_rate_in_mbps =
 			link_settings->link_rate * LINK_RATE_REF_FREQ_IN_MHZ;
-	struct fixed31_32 mbps = dal_fixed31_32_from_int(
+	struct fixed31_32 mbps = dc_fixpt_from_int(
 			link_rate_in_mbps * link_settings->lane_count);
 
-	return dal_fixed31_32_div_int(mbps, 54);
+	return dc_fixpt_div_int(mbps, 54);
 }
 
 static int get_color_depth(enum dc_color_depth color_depth)
@@ -2103,7 +2103,7 @@ static struct fixed31_32 get_pbn_from_timing(struct pipe_ctx *pipe_ctx)
 	numerator = 64 * PEAK_FACTOR_X1000;
 	denominator = 54 * 8 * 1000 * 1000;
 	kbps *= numerator;
-	peak_kbps = dal_fixed31_32_from_fraction(kbps, denominator);
+	peak_kbps = dc_fixpt_from_fraction(kbps, denominator);
 
 	return peak_kbps;
 }
@@ -2230,7 +2230,7 @@ static enum dc_status allocate_mst_payload(struct pipe_ctx *pipe_ctx)
 	/* slot X.Y for only current stream */
 	pbn_per_slot = get_pbn_per_slot(stream);
 	pbn = get_pbn_from_timing(pipe_ctx);
-	avg_time_slots_per_mtp = dal_fixed31_32_div(pbn, pbn_per_slot);
+	avg_time_slots_per_mtp = dc_fixpt_div(pbn, pbn_per_slot);
 
 	stream_encoder->funcs->set_mst_bandwidth(
 		stream_encoder,
@@ -2247,7 +2247,7 @@ static enum dc_status deallocate_mst_payload(struct pipe_ctx *pipe_ctx)
 	struct link_encoder *link_encoder = link->link_enc;
 	struct stream_encoder *stream_encoder = pipe_ctx->stream_res.stream_enc;
 	struct dp_mst_stream_allocation_table proposed_table = {0};
-	struct fixed31_32 avg_time_slots_per_mtp = dal_fixed31_32_from_int(0);
+	struct fixed31_32 avg_time_slots_per_mtp = dc_fixpt_from_int(0);
 	uint8_t i;
 	bool mst_mode = (link->type == dc_connection_mst_branch);
 	DC_LOGGER_INIT(link->ctx->logger);
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
index 345835f..082458f 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
@@ -496,9 +496,9 @@ static void calculate_viewport(struct pipe_ctx *pipe_ctx)
 	data->viewport_c.x = data->viewport.x / vpc_div;
 	data->viewport_c.y = data->viewport.y / vpc_div;
 	data->inits.h_c = (data->viewport.x % vpc_div) != 0 ?
-			dal_fixed31_32_half : dal_fixed31_32_zero;
+			dc_fixpt_half : dc_fixpt_zero;
 	data->inits.v_c = (data->viewport.y % vpc_div) != 0 ?
-			dal_fixed31_32_half : dal_fixed31_32_zero;
+			dc_fixpt_half : dc_fixpt_zero;
 	/* Round up, assume original video size always even dimensions */
 	data->viewport_c.width = (data->viewport.width + vpc_div - 1) / vpc_div;
 	data->viewport_c.height = (data->viewport.height + vpc_div - 1) / vpc_div;
@@ -627,10 +627,10 @@ static void calculate_scaling_ratios(struct pipe_ctx *pipe_ctx)
 			pipe_ctx->plane_state->rotation == ROTATION_ANGLE_270)
 		rect_swap_helper(&surf_src);
 
-	pipe_ctx->plane_res.scl_data.ratios.horz = dal_fixed31_32_from_fraction(
+	pipe_ctx->plane_res.scl_data.ratios.horz = dc_fixpt_from_fraction(
 					surf_src.width,
 					plane_state->dst_rect.width);
-	pipe_ctx->plane_res.scl_data.ratios.vert = dal_fixed31_32_from_fraction(
+	pipe_ctx->plane_res.scl_data.ratios.vert = dc_fixpt_from_fraction(
 					surf_src.height,
 					plane_state->dst_rect.height);
 
@@ -688,32 +688,32 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *r
 	 * 	init_bot = init + scaling_ratio
 	 * 	init_c = init + truncated_vp_c_offset(from calculate viewport)
 	 */
-	data->inits.h = dal_fixed31_32_div_int(
-			dal_fixed31_32_add_int(data->ratios.horz, data->taps.h_taps + 1), 2);
+	data->inits.h = dc_fixpt_div_int(
+			dc_fixpt_add_int(data->ratios.horz, data->taps.h_taps + 1), 2);
 
-	data->inits.h_c = dal_fixed31_32_add(data->inits.h_c, dal_fixed31_32_div_int(
-			dal_fixed31_32_add_int(data->ratios.horz_c, data->taps.h_taps_c + 1), 2));
+	data->inits.h_c = dc_fixpt_add(data->inits.h_c, dc_fixpt_div_int(
+			dc_fixpt_add_int(data->ratios.horz_c, data->taps.h_taps_c + 1), 2));
 
-	data->inits.v = dal_fixed31_32_div_int(
-			dal_fixed31_32_add_int(data->ratios.vert, data->taps.v_taps + 1), 2);
+	data->inits.v = dc_fixpt_div_int(
+			dc_fixpt_add_int(data->ratios.vert, data->taps.v_taps + 1), 2);
 
-	data->inits.v_c = dal_fixed31_32_add(data->inits.v_c, dal_fixed31_32_div_int(
-			dal_fixed31_32_add_int(data->ratios.vert_c, data->taps.v_taps_c + 1), 2));
+	data->inits.v_c = dc_fixpt_add(data->inits.v_c, dc_fixpt_div_int(
+			dc_fixpt_add_int(data->ratios.vert_c, data->taps.v_taps_c + 1), 2));
 
 
 	/* Adjust for viewport end clip-off */
 	if ((data->viewport.x + data->viewport.width) < (src.x + src.width) && !flip_horz_scan_dir) {
 		int vp_clip = src.x + src.width - data->viewport.width - data->viewport.x;
-		int int_part = dal_fixed31_32_floor(
-				dal_fixed31_32_sub(data->inits.h, data->ratios.horz));
+		int int_part = dc_fixpt_floor(
+				dc_fixpt_sub(data->inits.h, data->ratios.horz));
 
 		int_part = int_part > 0 ? int_part : 0;
 		data->viewport.width += int_part < vp_clip ? int_part : vp_clip;
 	}
 	if ((data->viewport.y + data->viewport.height) < (src.y + src.height) && !flip_vert_scan_dir) {
 		int vp_clip = src.y + src.height - data->viewport.height - data->viewport.y;
-		int int_part = dal_fixed31_32_floor(
-				dal_fixed31_32_sub(data->inits.v, data->ratios.vert));
+		int int_part = dc_fixpt_floor(
+				dc_fixpt_sub(data->inits.v, data->ratios.vert));
 
 		int_part = int_part > 0 ? int_part : 0;
 		data->viewport.height += int_part < vp_clip ? int_part : vp_clip;
@@ -721,8 +721,8 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *r
 	if ((data->viewport_c.x + data->viewport_c.width) < (src.x + src.width) / vpc_div && !flip_horz_scan_dir) {
 		int vp_clip = (src.x + src.width) / vpc_div -
 				data->viewport_c.width - data->viewport_c.x;
-		int int_part = dal_fixed31_32_floor(
-				dal_fixed31_32_sub(data->inits.h_c, data->ratios.horz_c));
+		int int_part = dc_fixpt_floor(
+				dc_fixpt_sub(data->inits.h_c, data->ratios.horz_c));
 
 		int_part = int_part > 0 ? int_part : 0;
 		data->viewport_c.width += int_part < vp_clip ? int_part : vp_clip;
@@ -730,8 +730,8 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *r
 	if ((data->viewport_c.y + data->viewport_c.height) < (src.y + src.height) / vpc_div && !flip_vert_scan_dir) {
 		int vp_clip = (src.y + src.height) / vpc_div -
 				data->viewport_c.height - data->viewport_c.y;
-		int int_part = dal_fixed31_32_floor(
-				dal_fixed31_32_sub(data->inits.v_c, data->ratios.vert_c));
+		int int_part = dc_fixpt_floor(
+				dc_fixpt_sub(data->inits.v_c, data->ratios.vert_c));
 
 		int_part = int_part > 0 ? int_part : 0;
 		data->viewport_c.height += int_part < vp_clip ? int_part : vp_clip;
@@ -741,9 +741,9 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *r
 	if (data->viewport.x && !flip_horz_scan_dir) {
 		int int_part;
 
-		data->inits.h = dal_fixed31_32_add(data->inits.h, dal_fixed31_32_mul_int(
+		data->inits.h = dc_fixpt_add(data->inits.h, dc_fixpt_mul_int(
 				data->ratios.horz, recout_skip->width));
-		int_part = dal_fixed31_32_floor(data->inits.h) - data->viewport.x;
+		int_part = dc_fixpt_floor(data->inits.h) - data->viewport.x;
 		if (int_part < data->taps.h_taps) {
 			int int_adj = data->viewport.x >= (data->taps.h_taps - int_part) ?
 						(data->taps.h_taps - int_part) : data->viewport.x;
@@ -756,15 +756,15 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *r
 			int_part = data->taps.h_taps;
 		}
 		data->inits.h.value &= 0xffffffff;
-		data->inits.h = dal_fixed31_32_add_int(data->inits.h, int_part);
+		data->inits.h = dc_fixpt_add_int(data->inits.h, int_part);
 	}
 
 	if (data->viewport_c.x && !flip_horz_scan_dir) {
 		int int_part;
 
-		data->inits.h_c = dal_fixed31_32_add(data->inits.h_c, dal_fixed31_32_mul_int(
+		data->inits.h_c = dc_fixpt_add(data->inits.h_c, dc_fixpt_mul_int(
 				data->ratios.horz_c, recout_skip->width));
-		int_part = dal_fixed31_32_floor(data->inits.h_c) - data->viewport_c.x;
+		int_part = dc_fixpt_floor(data->inits.h_c) - data->viewport_c.x;
 		if (int_part < data->taps.h_taps_c) {
 			int int_adj = data->viewport_c.x >= (data->taps.h_taps_c - int_part) ?
 					(data->taps.h_taps_c - int_part) : data->viewport_c.x;
@@ -777,15 +777,15 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *r
 			int_part = data->taps.h_taps_c;
 		}
 		data->inits.h_c.value &= 0xffffffff;
-		data->inits.h_c = dal_fixed31_32_add_int(data->inits.h_c, int_part);
+		data->inits.h_c = dc_fixpt_add_int(data->inits.h_c, int_part);
 	}
 
 	if (data->viewport.y && !flip_vert_scan_dir) {
 		int int_part;
 
-		data->inits.v = dal_fixed31_32_add(data->inits.v, dal_fixed31_32_mul_int(
+		data->inits.v = dc_fixpt_add(data->inits.v, dc_fixpt_mul_int(
 				data->ratios.vert, recout_skip->height));
-		int_part = dal_fixed31_32_floor(data->inits.v) - data->viewport.y;
+		int_part = dc_fixpt_floor(data->inits.v) - data->viewport.y;
 		if (int_part < data->taps.v_taps) {
 			int int_adj = data->viewport.y >= (data->taps.v_taps - int_part) ?
 						(data->taps.v_taps - int_part) : data->viewport.y;
@@ -798,15 +798,15 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *r
 			int_part = data->taps.v_taps;
 		}
 		data->inits.v.value &= 0xffffffff;
-		data->inits.v = dal_fixed31_32_add_int(data->inits.v, int_part);
+		data->inits.v = dc_fixpt_add_int(data->inits.v, int_part);
 	}
 
 	if (data->viewport_c.y && !flip_vert_scan_dir) {
 		int int_part;
 
-		data->inits.v_c = dal_fixed31_32_add(data->inits.v_c, dal_fixed31_32_mul_int(
+		data->inits.v_c = dc_fixpt_add(data->inits.v_c, dc_fixpt_mul_int(
 				data->ratios.vert_c, recout_skip->height));
-		int_part = dal_fixed31_32_floor(data->inits.v_c) - data->viewport_c.y;
+		int_part = dc_fixpt_floor(data->inits.v_c) - data->viewport_c.y;
 		if (int_part < data->taps.v_taps_c) {
 			int int_adj = data->viewport_c.y >= (data->taps.v_taps_c - int_part) ?
 					(data->taps.v_taps_c - int_part) : data->viewport_c.y;
@@ -819,12 +819,12 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *r
 			int_part = data->taps.v_taps_c;
 		}
 		data->inits.v_c.value &= 0xffffffff;
-		data->inits.v_c = dal_fixed31_32_add_int(data->inits.v_c, int_part);
+		data->inits.v_c = dc_fixpt_add_int(data->inits.v_c, int_part);
 	}
 
 	/* Interlaced inits based on final vert inits */
-	data->inits.v_bot = dal_fixed31_32_add(data->inits.v, data->ratios.vert);
-	data->inits.v_c_bot = dal_fixed31_32_add(data->inits.v_c, data->ratios.vert_c);
+	data->inits.v_bot = dc_fixpt_add(data->inits.v, data->ratios.vert);
+	data->inits.v_c_bot = dc_fixpt_add(data->inits.v_c, data->ratios.vert_c);
 
 	if (pipe_ctx->plane_state->rotation == ROTATION_ANGLE_90 ||
 			pipe_ctx->plane_state->rotation == ROTATION_ANGLE_270) {
diff --git a/drivers/gpu/drm/amd/display/dc/dc_dp_types.h b/drivers/gpu/drm/amd/display/dc/dc_dp_types.h
index 2726b02..90bccd5 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_dp_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_dp_types.h
@@ -26,6 +26,8 @@
 #ifndef DC_DP_TYPES_H
 #define DC_DP_TYPES_H
 
+#include "os_types.h"
+
 enum dc_lane_count {
 	LANE_COUNT_UNKNOWN = 0,
 	LANE_COUNT_ONE = 1,
diff --git a/drivers/gpu/drm/amd/display/dc/dc_types.h b/drivers/gpu/drm/amd/display/dc/dc_types.h
index 9defe3b..76df253 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_types.h
@@ -25,7 +25,7 @@
 #ifndef DC_TYPES_H_
 #define DC_TYPES_H_
 
-#include "fixed32_32.h"
+#include "os_types.h"
 #include "fixed31_32.h"
 #include "irq_types.h"
 #include "dc_dp_types.h"
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_abm.c b/drivers/gpu/drm/amd/display/dc/dce/dce_abm.c
index fe92a12..29294db 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_abm.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_abm.c
@@ -26,7 +26,7 @@
 #include "dce_abm.h"
 #include "dm_services.h"
 #include "reg_helper.h"
-#include "fixed32_32.h"
+#include "fixed31_32.h"
 #include "dc.h"
 
 #include "atom.h"
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c
index 0570e7e..599c7ab 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c
@@ -657,12 +657,12 @@ static uint32_t dce110_get_d_to_pixel_rate_in_hz(
 			return 0;
 		}
 
-		pix_rate = dal_fixed31_32_from_int(clk_src->ref_freq_khz);
-		pix_rate = dal_fixed31_32_mul_int(pix_rate, 1000);
-		pix_rate = dal_fixed31_32_mul_int(pix_rate, phase);
-		pix_rate = dal_fixed31_32_div_int(pix_rate, modulo);
+		pix_rate = dc_fixpt_from_int(clk_src->ref_freq_khz);
+		pix_rate = dc_fixpt_mul_int(pix_rate, 1000);
+		pix_rate = dc_fixpt_mul_int(pix_rate, phase);
+		pix_rate = dc_fixpt_div_int(pix_rate, modulo);
 
-		return dal_fixed31_32_round(pix_rate);
+		return dc_fixpt_round(pix_rate);
 	} else {
 		return dce110_get_dp_pixel_rate_from_combo_phy_pll(cs, pix_clk_params, pll_settings);
 	}
@@ -711,12 +711,12 @@ static bool calculate_ss(
 		const struct spread_spectrum_data *ss_data,
 		struct delta_sigma_data *ds_data)
 {
-	struct fixed32_32 fb_div;
-	struct fixed32_32 ss_amount;
-	struct fixed32_32 ss_nslip_amount;
-	struct fixed32_32 ss_ds_frac_amount;
-	struct fixed32_32 ss_step_size;
-	struct fixed32_32 modulation_time;
+	struct fixed31_32 fb_div;
+	struct fixed31_32 ss_amount;
+	struct fixed31_32 ss_nslip_amount;
+	struct fixed31_32 ss_ds_frac_amount;
+	struct fixed31_32 ss_step_size;
+	struct fixed31_32 modulation_time;
 
 	if (ds_data == NULL)
 		return false;
@@ -731,42 +731,42 @@ static bool calculate_ss(
 
 	/* compute SS_AMOUNT_FBDIV & SS_AMOUNT_NFRAC_SLIP & SS_AMOUNT_DSFRAC*/
 	/* 6 decimal point support in fractional feedback divider */
-	fb_div  = dal_fixed32_32_from_fraction(
+	fb_div  = dc_fixpt_from_fraction(
 		pll_settings->fract_feedback_divider, 1000000);
-	fb_div = dal_fixed32_32_add_int(fb_div, pll_settings->feedback_divider);
+	fb_div = dc_fixpt_add_int(fb_div, pll_settings->feedback_divider);
 
 	ds_data->ds_frac_amount = 0;
 	/*spreadSpectrumPercentage is in the unit of .01%,
 	 * so have to divided by 100 * 100*/
-	ss_amount = dal_fixed32_32_mul(
-		fb_div, dal_fixed32_32_from_fraction(ss_data->percentage,
+	ss_amount = dc_fixpt_mul(
+		fb_div, dc_fixpt_from_fraction(ss_data->percentage,
 					100 * ss_data->percentage_divider));
-	ds_data->feedback_amount = dal_fixed32_32_floor(ss_amount);
+	ds_data->feedback_amount = dc_fixpt_floor(ss_amount);
 
-	ss_nslip_amount = dal_fixed32_32_sub(ss_amount,
-		dal_fixed32_32_from_int(ds_data->feedback_amount));
-	ss_nslip_amount = dal_fixed32_32_mul_int(ss_nslip_amount, 10);
-	ds_data->nfrac_amount = dal_fixed32_32_floor(ss_nslip_amount);
+	ss_nslip_amount = dc_fixpt_sub(ss_amount,
+		dc_fixpt_from_int(ds_data->feedback_amount));
+	ss_nslip_amount = dc_fixpt_mul_int(ss_nslip_amount, 10);
+	ds_data->nfrac_amount = dc_fixpt_floor(ss_nslip_amount);
 
-	ss_ds_frac_amount = dal_fixed32_32_sub(ss_nslip_amount,
-		dal_fixed32_32_from_int(ds_data->nfrac_amount));
-	ss_ds_frac_amount = dal_fixed32_32_mul_int(ss_ds_frac_amount, 65536);
-	ds_data->ds_frac_amount = dal_fixed32_32_floor(ss_ds_frac_amount);
+	ss_ds_frac_amount = dc_fixpt_sub(ss_nslip_amount,
+		dc_fixpt_from_int(ds_data->nfrac_amount));
+	ss_ds_frac_amount = dc_fixpt_mul_int(ss_ds_frac_amount, 65536);
+	ds_data->ds_frac_amount = dc_fixpt_floor(ss_ds_frac_amount);
 
 	/* compute SS_STEP_SIZE_DSFRAC */
-	modulation_time = dal_fixed32_32_from_fraction(
+	modulation_time = dc_fixpt_from_fraction(
 		pll_settings->reference_freq * 1000,
 		pll_settings->reference_divider * ss_data->modulation_freq_hz);
 
 	if (ss_data->flags.CENTER_SPREAD)
-		modulation_time = dal_fixed32_32_div_int(modulation_time, 4);
+		modulation_time = dc_fixpt_div_int(modulation_time, 4);
 	else
-		modulation_time = dal_fixed32_32_div_int(modulation_time, 2);
+		modulation_time = dc_fixpt_div_int(modulation_time, 2);
 
-	ss_step_size = dal_fixed32_32_div(ss_amount, modulation_time);
+	ss_step_size = dc_fixpt_div(ss_amount, modulation_time);
 	/* SS_STEP_SIZE_DSFRAC_DEC = Int(SS_STEP_SIZE * 2 ^ 16 * 10)*/
-	ss_step_size = dal_fixed32_32_mul_int(ss_step_size, 65536 * 10);
-	ds_data->ds_frac_size =  dal_fixed32_32_floor(ss_step_size);
+	ss_step_size = dc_fixpt_mul_int(ss_step_size, 65536 * 10);
+	ds_data->ds_frac_size =  dc_fixpt_floor(ss_step_size);
 
 	return true;
 }
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c b/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c
index f043e5e..8a581c6 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c
@@ -26,7 +26,7 @@
 #include "dce_clocks.h"
 #include "dm_services.h"
 #include "reg_helper.h"
-#include "fixed32_32.h"
+#include "fixed31_32.h"
 #include "bios_parser_interface.h"
 #include "dc.h"
 #include "dmcu.h"
@@ -228,19 +228,19 @@ static int dce_clocks_get_dp_ref_freq(struct display_clock *clk)
 	 generated according to average value (case as with previous ASICs)
 	  */
 	if (clk_dce->ss_on_dprefclk && clk_dce->dprefclk_ss_divider != 0) {
-		struct fixed32_32 ss_percentage = dal_fixed32_32_div_int(
-				dal_fixed32_32_from_fraction(
+		struct fixed31_32 ss_percentage = dc_fixpt_div_int(
+				dc_fixpt_from_fraction(
 						clk_dce->dprefclk_ss_percentage,
 						clk_dce->dprefclk_ss_divider), 200);
-		struct fixed32_32 adj_dp_ref_clk_khz;
+		struct fixed31_32 adj_dp_ref_clk_khz;
 
-		ss_percentage = dal_fixed32_32_sub(dal_fixed32_32_one,
+		ss_percentage = dc_fixpt_sub(dc_fixpt_one,
 								ss_percentage);
 		adj_dp_ref_clk_khz =
-			dal_fixed32_32_mul_int(
+			dc_fixpt_mul_int(
 				ss_percentage,
 				dp_ref_clk_khz);
-		dp_ref_clk_khz = dal_fixed32_32_floor(adj_dp_ref_clk_khz);
+		dp_ref_clk_khz = dc_fixpt_floor(adj_dp_ref_clk_khz);
 	}
 
 	return dp_ref_clk_khz;
@@ -256,19 +256,19 @@ static int dce_clocks_get_dp_ref_freq_wrkaround(struct display_clock *clk)
 	int dp_ref_clk_khz = 600000;
 
 	if (clk_dce->ss_on_dprefclk && clk_dce->dprefclk_ss_divider != 0) {
-		struct fixed32_32 ss_percentage = dal_fixed32_32_div_int(
-				dal_fixed32_32_from_fraction(
+		struct fixed31_32 ss_percentage = dc_fixpt_div_int(
+				dc_fixpt_from_fraction(
 						clk_dce->dprefclk_ss_percentage,
 						clk_dce->dprefclk_ss_divider), 200);
-		struct fixed32_32 adj_dp_ref_clk_khz;
+		struct fixed31_32 adj_dp_ref_clk_khz;
 
-		ss_percentage = dal_fixed32_32_sub(dal_fixed32_32_one,
+		ss_percentage = dc_fixpt_sub(dc_fixpt_one,
 								ss_percentage);
 		adj_dp_ref_clk_khz =
-			dal_fixed32_32_mul_int(
+			dc_fixpt_mul_int(
 				ss_percentage,
 				dp_ref_clk_khz);
-		dp_ref_clk_khz = dal_fixed32_32_floor(adj_dp_ref_clk_khz);
+		dp_ref_clk_khz = dc_fixpt_floor(adj_dp_ref_clk_khz);
 	}
 
 	return dp_ref_clk_khz;
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c b/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c
index 2ee3d9b..a576b8b 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c
@@ -28,7 +28,7 @@
 #include "dce_dmcu.h"
 #include "dm_services.h"
 #include "reg_helper.h"
-#include "fixed32_32.h"
+#include "fixed31_32.h"
 #include "dc.h"
 
 #define TO_DCE_DMCU(dmcu)\
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_ipp.c b/drivers/gpu/drm/amd/display/dc/dce/dce_ipp.c
index d737e91..5d9506b 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_ipp.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_ipp.c
@@ -195,13 +195,13 @@ static void dce_ipp_program_input_lut(
 
 	for (i = 0; i < gamma->num_entries; i++) {
 		REG_SET(DC_LUT_SEQ_COLOR, 0, DC_LUT_SEQ_COLOR,
-				dal_fixed31_32_round(
+				dc_fixpt_round(
 					gamma->entries.red[i]));
 		REG_SET(DC_LUT_SEQ_COLOR, 0, DC_LUT_SEQ_COLOR,
-				dal_fixed31_32_round(
+				dc_fixpt_round(
 					gamma->entries.green[i]));
 		REG_SET(DC_LUT_SEQ_COLOR, 0, DC_LUT_SEQ_COLOR,
-				dal_fixed31_32_round(
+				dc_fixpt_round(
 					gamma->entries.blue[i]));
 	}
 
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_scl_filters.c b/drivers/gpu/drm/amd/display/dc/dce/dce_scl_filters.c
index 6243450..48862be 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_scl_filters.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_scl_filters.c
@@ -1014,11 +1014,11 @@ static const uint16_t filter_8tap_64p_183[264] = {
 
 const uint16_t *get_filter_3tap_16p(struct fixed31_32 ratio)
 {
-	if (ratio.value < dal_fixed31_32_one.value)
+	if (ratio.value < dc_fixpt_one.value)
 		return filter_3tap_16p_upscale;
-	else if (ratio.value < dal_fixed31_32_from_fraction(4, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(4, 3).value)
 		return filter_3tap_16p_117;
-	else if (ratio.value < dal_fixed31_32_from_fraction(5, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(5, 3).value)
 		return filter_3tap_16p_150;
 	else
 		return filter_3tap_16p_183;
@@ -1026,11 +1026,11 @@ const uint16_t *get_filter_3tap_16p(struct fixed31_32 ratio)
 
 const uint16_t *get_filter_3tap_64p(struct fixed31_32 ratio)
 {
-	if (ratio.value < dal_fixed31_32_one.value)
+	if (ratio.value < dc_fixpt_one.value)
 		return filter_3tap_64p_upscale;
-	else if (ratio.value < dal_fixed31_32_from_fraction(4, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(4, 3).value)
 		return filter_3tap_64p_117;
-	else if (ratio.value < dal_fixed31_32_from_fraction(5, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(5, 3).value)
 		return filter_3tap_64p_150;
 	else
 		return filter_3tap_64p_183;
@@ -1038,11 +1038,11 @@ const uint16_t *get_filter_3tap_64p(struct fixed31_32 ratio)
 
 const uint16_t *get_filter_4tap_16p(struct fixed31_32 ratio)
 {
-	if (ratio.value < dal_fixed31_32_one.value)
+	if (ratio.value < dc_fixpt_one.value)
 		return filter_4tap_16p_upscale;
-	else if (ratio.value < dal_fixed31_32_from_fraction(4, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(4, 3).value)
 		return filter_4tap_16p_117;
-	else if (ratio.value < dal_fixed31_32_from_fraction(5, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(5, 3).value)
 		return filter_4tap_16p_150;
 	else
 		return filter_4tap_16p_183;
@@ -1050,11 +1050,11 @@ const uint16_t *get_filter_4tap_16p(struct fixed31_32 ratio)
 
 const uint16_t *get_filter_4tap_64p(struct fixed31_32 ratio)
 {
-	if (ratio.value < dal_fixed31_32_one.value)
+	if (ratio.value < dc_fixpt_one.value)
 		return filter_4tap_64p_upscale;
-	else if (ratio.value < dal_fixed31_32_from_fraction(4, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(4, 3).value)
 		return filter_4tap_64p_117;
-	else if (ratio.value < dal_fixed31_32_from_fraction(5, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(5, 3).value)
 		return filter_4tap_64p_150;
 	else
 		return filter_4tap_64p_183;
@@ -1062,11 +1062,11 @@ const uint16_t *get_filter_4tap_64p(struct fixed31_32 ratio)
 
 const uint16_t *get_filter_5tap_64p(struct fixed31_32 ratio)
 {
-	if (ratio.value < dal_fixed31_32_one.value)
+	if (ratio.value < dc_fixpt_one.value)
 		return filter_5tap_64p_upscale;
-	else if (ratio.value < dal_fixed31_32_from_fraction(4, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(4, 3).value)
 		return filter_5tap_64p_117;
-	else if (ratio.value < dal_fixed31_32_from_fraction(5, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(5, 3).value)
 		return filter_5tap_64p_150;
 	else
 		return filter_5tap_64p_183;
@@ -1074,11 +1074,11 @@ const uint16_t *get_filter_5tap_64p(struct fixed31_32 ratio)
 
 const uint16_t *get_filter_6tap_64p(struct fixed31_32 ratio)
 {
-	if (ratio.value < dal_fixed31_32_one.value)
+	if (ratio.value < dc_fixpt_one.value)
 		return filter_6tap_64p_upscale;
-	else if (ratio.value < dal_fixed31_32_from_fraction(4, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(4, 3).value)
 		return filter_6tap_64p_117;
-	else if (ratio.value < dal_fixed31_32_from_fraction(5, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(5, 3).value)
 		return filter_6tap_64p_150;
 	else
 		return filter_6tap_64p_183;
@@ -1086,11 +1086,11 @@ const uint16_t *get_filter_6tap_64p(struct fixed31_32 ratio)
 
 const uint16_t *get_filter_7tap_64p(struct fixed31_32 ratio)
 {
-	if (ratio.value < dal_fixed31_32_one.value)
+	if (ratio.value < dc_fixpt_one.value)
 		return filter_7tap_64p_upscale;
-	else if (ratio.value < dal_fixed31_32_from_fraction(4, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(4, 3).value)
 		return filter_7tap_64p_117;
-	else if (ratio.value < dal_fixed31_32_from_fraction(5, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(5, 3).value)
 		return filter_7tap_64p_150;
 	else
 		return filter_7tap_64p_183;
@@ -1098,11 +1098,11 @@ const uint16_t *get_filter_7tap_64p(struct fixed31_32 ratio)
 
 const uint16_t *get_filter_8tap_64p(struct fixed31_32 ratio)
 {
-	if (ratio.value < dal_fixed31_32_one.value)
+	if (ratio.value < dc_fixpt_one.value)
 		return filter_8tap_64p_upscale;
-	else if (ratio.value < dal_fixed31_32_from_fraction(4, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(4, 3).value)
 		return filter_8tap_64p_117;
-	else if (ratio.value < dal_fixed31_32_from_fraction(5, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(5, 3).value)
 		return filter_8tap_64p_150;
 	else
 		return filter_8tap_64p_183;
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
index e265a0a..0a6d483 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
@@ -683,11 +683,11 @@ static void dce110_stream_encoder_set_mst_bandwidth(
 	struct fixed31_32 avg_time_slots_per_mtp)
 {
 	struct dce110_stream_encoder *enc110 = DCE110STRENC_FROM_STRENC(enc);
-	uint32_t x = dal_fixed31_32_floor(
+	uint32_t x = dc_fixpt_floor(
 		avg_time_slots_per_mtp);
-	uint32_t y = dal_fixed31_32_ceil(
-		dal_fixed31_32_shl(
-			dal_fixed31_32_sub_int(
+	uint32_t y = dc_fixpt_ceil(
+		dc_fixpt_shl(
+			dc_fixpt_sub_int(
 				avg_time_slots_per_mtp,
 				x),
 			26));
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_transform.c b/drivers/gpu/drm/amd/display/dc/dce/dce_transform.c
index 832c5da..a02e719 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_transform.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_transform.c
@@ -41,7 +41,7 @@
 #define DC_LOGGER \
 	xfm_dce->base.ctx->logger
 
-#define IDENTITY_RATIO(ratio) (dal_fixed31_32_u2d19(ratio) == (1 << 19))
+#define IDENTITY_RATIO(ratio) (dc_fixpt_u2d19(ratio) == (1 << 19))
 #define GAMUT_MATRIX_SIZE 12
 #define SCL_PHASES 16
 
@@ -256,27 +256,27 @@ static void calculate_inits(
 	struct fixed31_32 v_init;
 
 	inits->h_int_scale_ratio =
-		dal_fixed31_32_u2d19(data->ratios.horz) << 5;
+		dc_fixpt_u2d19(data->ratios.horz) << 5;
 	inits->v_int_scale_ratio =
-		dal_fixed31_32_u2d19(data->ratios.vert) << 5;
+		dc_fixpt_u2d19(data->ratios.vert) << 5;
 
 	h_init =
-		dal_fixed31_32_div_int(
-			dal_fixed31_32_add(
+		dc_fixpt_div_int(
+			dc_fixpt_add(
 				data->ratios.horz,
-				dal_fixed31_32_from_int(data->taps.h_taps + 1)),
+				dc_fixpt_from_int(data->taps.h_taps + 1)),
 				2);
-	inits->h_init.integer = dal_fixed31_32_floor(h_init);
-	inits->h_init.fraction = dal_fixed31_32_u0d19(h_init) << 5;
+	inits->h_init.integer = dc_fixpt_floor(h_init);
+	inits->h_init.fraction = dc_fixpt_u0d19(h_init) << 5;
 
 	v_init =
-		dal_fixed31_32_div_int(
-			dal_fixed31_32_add(
+		dc_fixpt_div_int(
+			dc_fixpt_add(
 				data->ratios.vert,
-				dal_fixed31_32_from_int(data->taps.v_taps + 1)),
+				dc_fixpt_from_int(data->taps.v_taps + 1)),
 				2);
-	inits->v_init.integer = dal_fixed31_32_floor(v_init);
-	inits->v_init.fraction = dal_fixed31_32_u0d19(v_init) << 5;
+	inits->v_init.integer = dc_fixpt_floor(v_init);
+	inits->v_init.fraction = dc_fixpt_u0d19(v_init) << 5;
 }
 
 static void program_scl_ratios_inits(
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
index ae50042..a92fb0a 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
@@ -509,19 +509,19 @@ dce110_translate_regamma_to_hw_format(const struct dc_transfer_func *output_tf,
 	rgb_resulted[hw_points - 1].green = output_tf->tf_pts.green[start_index];
 	rgb_resulted[hw_points - 1].blue = output_tf->tf_pts.blue[start_index];
 
-	arr_points[0].x = dal_fixed31_32_pow(dal_fixed31_32_from_int(2),
-					     dal_fixed31_32_from_int(region_start));
-	arr_points[1].x = dal_fixed31_32_pow(dal_fixed31_32_from_int(2),
-					     dal_fixed31_32_from_int(region_end));
+	arr_points[0].x = dc_fixpt_pow(dc_fixpt_from_int(2),
+					     dc_fixpt_from_int(region_start));
+	arr_points[1].x = dc_fixpt_pow(dc_fixpt_from_int(2),
+					     dc_fixpt_from_int(region_end));
 
 	y_r = rgb_resulted[0].red;
 	y_g = rgb_resulted[0].green;
 	y_b = rgb_resulted[0].blue;
 
-	y1_min = dal_fixed31_32_min(y_r, dal_fixed31_32_min(y_g, y_b));
+	y1_min = dc_fixpt_min(y_r, dc_fixpt_min(y_g, y_b));
 
 	arr_points[0].y = y1_min;
-	arr_points[0].slope = dal_fixed31_32_div(arr_points[0].y,
+	arr_points[0].slope = dc_fixpt_div(arr_points[0].y,
 						 arr_points[0].x);
 
 	y_r = rgb_resulted[hw_points - 1].red;
@@ -531,21 +531,21 @@ dce110_translate_regamma_to_hw_format(const struct dc_transfer_func *output_tf,
 	/* see comment above, m_arrPoints[1].y should be the Y value for the
 	 * region end (m_numOfHwPoints), not last HW point(m_numOfHwPoints - 1)
 	 */
-	y3_max = dal_fixed31_32_max(y_r, dal_fixed31_32_max(y_g, y_b));
+	y3_max = dc_fixpt_max(y_r, dc_fixpt_max(y_g, y_b));
 
 	arr_points[1].y = y3_max;
 
-	arr_points[1].slope = dal_fixed31_32_zero;
+	arr_points[1].slope = dc_fixpt_zero;
 
 	if (output_tf->tf == TRANSFER_FUNCTION_PQ) {
 		/* for PQ, we want to have a straight line from last HW X point,
 		 * and the slope to be such that we hit 1.0 at 10000 nits.
 		 */
-		const struct fixed31_32 end_value = dal_fixed31_32_from_int(125);
+		const struct fixed31_32 end_value = dc_fixpt_from_int(125);
 
-		arr_points[1].slope = dal_fixed31_32_div(
-				dal_fixed31_32_sub(dal_fixed31_32_one, arr_points[1].y),
-				dal_fixed31_32_sub(end_value, arr_points[1].x));
+		arr_points[1].slope = dc_fixpt_div(
+				dc_fixpt_sub(dc_fixpt_one, arr_points[1].y),
+				dc_fixpt_sub(end_value, arr_points[1].x));
 	}
 
 	regamma_params->hw_points_num = hw_points;
@@ -569,16 +569,16 @@ dce110_translate_regamma_to_hw_format(const struct dc_transfer_func *output_tf,
 	i = 1;
 
 	while (i != hw_points + 1) {
-		if (dal_fixed31_32_lt(rgb_plus_1->red, rgb->red))
+		if (dc_fixpt_lt(rgb_plus_1->red, rgb->red))
 			rgb_plus_1->red = rgb->red;
-		if (dal_fixed31_32_lt(rgb_plus_1->green, rgb->green))
+		if (dc_fixpt_lt(rgb_plus_1->green, rgb->green))
 			rgb_plus_1->green = rgb->green;
-		if (dal_fixed31_32_lt(rgb_plus_1->blue, rgb->blue))
+		if (dc_fixpt_lt(rgb_plus_1->blue, rgb->blue))
 			rgb_plus_1->blue = rgb->blue;
 
-		rgb->delta_red = dal_fixed31_32_sub(rgb_plus_1->red, rgb->red);
-		rgb->delta_green = dal_fixed31_32_sub(rgb_plus_1->green, rgb->green);
-		rgb->delta_blue = dal_fixed31_32_sub(rgb_plus_1->blue, rgb->blue);
+		rgb->delta_red = dc_fixpt_sub(rgb_plus_1->red, rgb->red);
+		rgb->delta_green = dc_fixpt_sub(rgb_plus_1->green, rgb->green);
+		rgb->delta_blue = dc_fixpt_sub(rgb_plus_1->blue, rgb->blue);
 
 		++rgb_plus_1;
 		++rgb;
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_transform_v.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_transform_v.c
index 8ba3c12..a7dce06 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_transform_v.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_transform_v.c
@@ -373,13 +373,13 @@ static void calculate_inits(
 	struct rect *chroma_viewport)
 {
 	inits->h_int_scale_ratio_luma =
-		dal_fixed31_32_u2d19(data->ratios.horz) << 5;
+		dc_fixpt_u2d19(data->ratios.horz) << 5;
 	inits->v_int_scale_ratio_luma =
-		dal_fixed31_32_u2d19(data->ratios.vert) << 5;
+		dc_fixpt_u2d19(data->ratios.vert) << 5;
 	inits->h_int_scale_ratio_chroma =
-		dal_fixed31_32_u2d19(data->ratios.horz_c) << 5;
+		dc_fixpt_u2d19(data->ratios.horz_c) << 5;
 	inits->v_int_scale_ratio_chroma =
-		dal_fixed31_32_u2d19(data->ratios.vert_c) << 5;
+		dc_fixpt_u2d19(data->ratios.vert_c) << 5;
 
 	inits->h_init_luma.integer = 1;
 	inits->v_init_luma.integer = 1;
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c
index 96d5878..5d95a99 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c
@@ -169,7 +169,7 @@ bool cm_helper_convert_to_custom_float(
 	}
 
 	if (fixpoint == true)
-		arr_points[1].custom_float_y = dal_fixed31_32_clamp_u0d14(arr_points[1].y);
+		arr_points[1].custom_float_y = dc_fixpt_clamp_u0d14(arr_points[1].y);
 	else if (!convert_to_custom_float_format(arr_points[1].y, &fmt,
 		&arr_points[1].custom_float_y)) {
 		BREAK_TO_DEBUGGER();
@@ -327,19 +327,19 @@ bool cm_helper_translate_curve_to_hw_format(
 	rgb_resulted[hw_points - 1].green = output_tf->tf_pts.green[start_index];
 	rgb_resulted[hw_points - 1].blue = output_tf->tf_pts.blue[start_index];
 
-	arr_points[0].x = dal_fixed31_32_pow(dal_fixed31_32_from_int(2),
-					     dal_fixed31_32_from_int(region_start));
-	arr_points[1].x = dal_fixed31_32_pow(dal_fixed31_32_from_int(2),
-					     dal_fixed31_32_from_int(region_end));
+	arr_points[0].x = dc_fixpt_pow(dc_fixpt_from_int(2),
+					     dc_fixpt_from_int(region_start));
+	arr_points[1].x = dc_fixpt_pow(dc_fixpt_from_int(2),
+					     dc_fixpt_from_int(region_end));
 
 	y_r = rgb_resulted[0].red;
 	y_g = rgb_resulted[0].green;
 	y_b = rgb_resulted[0].blue;
 
-	y1_min = dal_fixed31_32_min(y_r, dal_fixed31_32_min(y_g, y_b));
+	y1_min = dc_fixpt_min(y_r, dc_fixpt_min(y_g, y_b));
 
 	arr_points[0].y = y1_min;
-	arr_points[0].slope = dal_fixed31_32_div(arr_points[0].y, arr_points[0].x);
+	arr_points[0].slope = dc_fixpt_div(arr_points[0].y, arr_points[0].x);
 	y_r = rgb_resulted[hw_points - 1].red;
 	y_g = rgb_resulted[hw_points - 1].green;
 	y_b = rgb_resulted[hw_points - 1].blue;
@@ -347,22 +347,22 @@ bool cm_helper_translate_curve_to_hw_format(
 	/* see comment above, m_arrPoints[1].y should be the Y value for the
 	 * region end (m_numOfHwPoints), not last HW point(m_numOfHwPoints - 1)
 	 */
-	y3_max = dal_fixed31_32_max(y_r, dal_fixed31_32_max(y_g, y_b));
+	y3_max = dc_fixpt_max(y_r, dc_fixpt_max(y_g, y_b));
 
 	arr_points[1].y = y3_max;
 
-	arr_points[1].slope = dal_fixed31_32_zero;
+	arr_points[1].slope = dc_fixpt_zero;
 
 	if (output_tf->tf == TRANSFER_FUNCTION_PQ) {
 		/* for PQ, we want to have a straight line from last HW X point,
 		 * and the slope to be such that we hit 1.0 at 10000 nits.
 		 */
 		const struct fixed31_32 end_value =
-				dal_fixed31_32_from_int(125);
+				dc_fixpt_from_int(125);
 
-		arr_points[1].slope = dal_fixed31_32_div(
-			dal_fixed31_32_sub(dal_fixed31_32_one, arr_points[1].y),
-			dal_fixed31_32_sub(end_value, arr_points[1].x));
+		arr_points[1].slope = dc_fixpt_div(
+			dc_fixpt_sub(dc_fixpt_one, arr_points[1].y),
+			dc_fixpt_sub(end_value, arr_points[1].x));
 	}
 
 	lut_params->hw_points_num = hw_points;
@@ -386,24 +386,24 @@ bool cm_helper_translate_curve_to_hw_format(
 
 	i = 1;
 	while (i != hw_points + 1) {
-		if (dal_fixed31_32_lt(rgb_plus_1->red, rgb->red))
+		if (dc_fixpt_lt(rgb_plus_1->red, rgb->red))
 			rgb_plus_1->red = rgb->red;
-		if (dal_fixed31_32_lt(rgb_plus_1->green, rgb->green))
+		if (dc_fixpt_lt(rgb_plus_1->green, rgb->green))
 			rgb_plus_1->green = rgb->green;
-		if (dal_fixed31_32_lt(rgb_plus_1->blue, rgb->blue))
+		if (dc_fixpt_lt(rgb_plus_1->blue, rgb->blue))
 			rgb_plus_1->blue = rgb->blue;
 
-		rgb->delta_red   = dal_fixed31_32_sub(rgb_plus_1->red,   rgb->red);
-		rgb->delta_green = dal_fixed31_32_sub(rgb_plus_1->green, rgb->green);
-		rgb->delta_blue  = dal_fixed31_32_sub(rgb_plus_1->blue,  rgb->blue);
+		rgb->delta_red   = dc_fixpt_sub(rgb_plus_1->red,   rgb->red);
+		rgb->delta_green = dc_fixpt_sub(rgb_plus_1->green, rgb->green);
+		rgb->delta_blue  = dc_fixpt_sub(rgb_plus_1->blue,  rgb->blue);
 
 		if (fixpoint == true) {
-			rgb->delta_red_reg   = dal_fixed31_32_clamp_u0d10(rgb->delta_red);
-			rgb->delta_green_reg = dal_fixed31_32_clamp_u0d10(rgb->delta_green);
-			rgb->delta_blue_reg  = dal_fixed31_32_clamp_u0d10(rgb->delta_blue);
-			rgb->red_reg         = dal_fixed31_32_clamp_u0d14(rgb->red);
-			rgb->green_reg       = dal_fixed31_32_clamp_u0d14(rgb->green);
-			rgb->blue_reg        = dal_fixed31_32_clamp_u0d14(rgb->blue);
+			rgb->delta_red_reg   = dc_fixpt_clamp_u0d10(rgb->delta_red);
+			rgb->delta_green_reg = dc_fixpt_clamp_u0d10(rgb->delta_green);
+			rgb->delta_blue_reg  = dc_fixpt_clamp_u0d10(rgb->delta_blue);
+			rgb->red_reg         = dc_fixpt_clamp_u0d14(rgb->red);
+			rgb->green_reg       = dc_fixpt_clamp_u0d14(rgb->green);
+			rgb->blue_reg        = dc_fixpt_clamp_u0d14(rgb->blue);
 		}
 
 		++rgb_plus_1;
@@ -489,19 +489,19 @@ bool cm_helper_translate_curve_to_degamma_hw_format(
 	rgb_resulted[hw_points - 1].green = output_tf->tf_pts.green[start_index];
 	rgb_resulted[hw_points - 1].blue = output_tf->tf_pts.blue[start_index];
 
-	arr_points[0].x = dal_fixed31_32_pow(dal_fixed31_32_from_int(2),
-					     dal_fixed31_32_from_int(region_start));
-	arr_points[1].x = dal_fixed31_32_pow(dal_fixed31_32_from_int(2),
-					     dal_fixed31_32_from_int(region_end));
+	arr_points[0].x = dc_fixpt_pow(dc_fixpt_from_int(2),
+					     dc_fixpt_from_int(region_start));
+	arr_points[1].x = dc_fixpt_pow(dc_fixpt_from_int(2),
+					     dc_fixpt_from_int(region_end));
 
 	y_r = rgb_resulted[0].red;
 	y_g = rgb_resulted[0].green;
 	y_b = rgb_resulted[0].blue;
 
-	y1_min = dal_fixed31_32_min(y_r, dal_fixed31_32_min(y_g, y_b));
+	y1_min = dc_fixpt_min(y_r, dc_fixpt_min(y_g, y_b));
 
 	arr_points[0].y = y1_min;
-	arr_points[0].slope = dal_fixed31_32_div(arr_points[0].y, arr_points[0].x);
+	arr_points[0].slope = dc_fixpt_div(arr_points[0].y, arr_points[0].x);
 	y_r = rgb_resulted[hw_points - 1].red;
 	y_g = rgb_resulted[hw_points - 1].green;
 	y_b = rgb_resulted[hw_points - 1].blue;
@@ -509,22 +509,22 @@ bool cm_helper_translate_curve_to_degamma_hw_format(
 	/* see comment above, m_arrPoints[1].y should be the Y value for the
 	 * region end (m_numOfHwPoints), not last HW point(m_numOfHwPoints - 1)
 	 */
-	y3_max = dal_fixed31_32_max(y_r, dal_fixed31_32_max(y_g, y_b));
+	y3_max = dc_fixpt_max(y_r, dc_fixpt_max(y_g, y_b));
 
 	arr_points[1].y = y3_max;
 
-	arr_points[1].slope = dal_fixed31_32_zero;
+	arr_points[1].slope = dc_fixpt_zero;
 
 	if (output_tf->tf == TRANSFER_FUNCTION_PQ) {
 		/* for PQ, we want to have a straight line from last HW X point,
 		 * and the slope to be such that we hit 1.0 at 10000 nits.
 		 */
 		const struct fixed31_32 end_value =
-				dal_fixed31_32_from_int(125);
+				dc_fixpt_from_int(125);
 
-		arr_points[1].slope = dal_fixed31_32_div(
-			dal_fixed31_32_sub(dal_fixed31_32_one, arr_points[1].y),
-			dal_fixed31_32_sub(end_value, arr_points[1].x));
+		arr_points[1].slope = dc_fixpt_div(
+			dc_fixpt_sub(dc_fixpt_one, arr_points[1].y),
+			dc_fixpt_sub(end_value, arr_points[1].x));
 	}
 
 	lut_params->hw_points_num = hw_points;
@@ -548,16 +548,16 @@ bool cm_helper_translate_curve_to_degamma_hw_format(
 
 	i = 1;
 	while (i != hw_points + 1) {
-		if (dal_fixed31_32_lt(rgb_plus_1->red, rgb->red))
+		if (dc_fixpt_lt(rgb_plus_1->red, rgb->red))
 			rgb_plus_1->red = rgb->red;
-		if (dal_fixed31_32_lt(rgb_plus_1->green, rgb->green))
+		if (dc_fixpt_lt(rgb_plus_1->green, rgb->green))
 			rgb_plus_1->green = rgb->green;
-		if (dal_fixed31_32_lt(rgb_plus_1->blue, rgb->blue))
+		if (dc_fixpt_lt(rgb_plus_1->blue, rgb->blue))
 			rgb_plus_1->blue = rgb->blue;
 
-		rgb->delta_red   = dal_fixed31_32_sub(rgb_plus_1->red,   rgb->red);
-		rgb->delta_green = dal_fixed31_32_sub(rgb_plus_1->green, rgb->green);
-		rgb->delta_blue  = dal_fixed31_32_sub(rgb_plus_1->blue,  rgb->blue);
+		rgb->delta_red   = dc_fixpt_sub(rgb_plus_1->red,   rgb->red);
+		rgb->delta_green = dc_fixpt_sub(rgb_plus_1->green, rgb->green);
+		rgb->delta_blue  = dc_fixpt_sub(rgb_plus_1->blue,  rgb->blue);
 
 		++rgb_plus_1;
 		++rgb;
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
index 8c4d9e5..20796da 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
@@ -130,7 +130,7 @@ void dpp_set_gamut_remap_bypass(struct dcn10_dpp *dpp)
 	/* Gamut remap in bypass */
 }
 
-#define IDENTITY_RATIO(ratio) (dal_fixed31_32_u2d19(ratio) == (1 << 19))
+#define IDENTITY_RATIO(ratio) (dc_fixpt_u2d19(ratio) == (1 << 19))
 
 
 bool dpp_get_optimal_number_of_taps(
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c
index 4f373c9..116977e 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c
@@ -811,13 +811,13 @@ void dpp1_program_input_lut(
 	REG_UPDATE(CM_IGAM_LUT_RW_INDEX, CM_IGAM_LUT_RW_INDEX, 0);
 	for (i = 0; i < gamma->num_entries; i++) {
 		REG_SET(CM_IGAM_LUT_SEQ_COLOR, 0, CM_IGAM_LUT_SEQ_COLOR,
-				dal_fixed31_32_round(
+				dc_fixpt_round(
 					gamma->entries.red[i]));
 		REG_SET(CM_IGAM_LUT_SEQ_COLOR, 0, CM_IGAM_LUT_SEQ_COLOR,
-				dal_fixed31_32_round(
+				dc_fixpt_round(
 					gamma->entries.green[i]));
 		REG_SET(CM_IGAM_LUT_SEQ_COLOR, 0, CM_IGAM_LUT_SEQ_COLOR,
-				dal_fixed31_32_round(
+				dc_fixpt_round(
 					gamma->entries.blue[i]));
 	}
 	// Power off LUT memory
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c
index 3eb824d..4ddd627 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c
@@ -169,7 +169,7 @@ static enum dscl_mode_sel dpp1_dscl_get_dscl_mode(
 		const struct scaler_data *data,
 		bool dbg_always_scale)
 {
-	const long long one = dal_fixed31_32_one.value;
+	const long long one = dc_fixpt_one.value;
 
 	if (dpp_base->caps->dscl_data_proc_format == DSCL_DATA_PRCESSING_FIXED_FORMAT) {
 		/* DSCL is processing data in fixed format */
@@ -464,8 +464,8 @@ static enum lb_memory_config dpp1_dscl_find_lb_memory_config(struct dcn10_dpp *d
 	int num_part_y, num_part_c;
 	int vtaps = scl_data->taps.v_taps;
 	int vtaps_c = scl_data->taps.v_taps_c;
-	int ceil_vratio = dal_fixed31_32_ceil(scl_data->ratios.vert);
-	int ceil_vratio_c = dal_fixed31_32_ceil(scl_data->ratios.vert_c);
+	int ceil_vratio = dc_fixpt_ceil(scl_data->ratios.vert);
+	int ceil_vratio_c = dc_fixpt_ceil(scl_data->ratios.vert_c);
 	enum lb_memory_config mem_cfg = LB_MEMORY_CONFIG_0;
 
 	if (dpp->base.ctx->dc->debug.use_max_lb)
@@ -565,52 +565,52 @@ static void dpp1_dscl_set_manual_ratio_init(
 	uint32_t init_int = 0;
 
 	REG_SET(SCL_HORZ_FILTER_SCALE_RATIO, 0,
-			SCL_H_SCALE_RATIO, dal_fixed31_32_u2d19(data->ratios.horz) << 5);
+			SCL_H_SCALE_RATIO, dc_fixpt_u2d19(data->ratios.horz) << 5);
 
 	REG_SET(SCL_VERT_FILTER_SCALE_RATIO, 0,
-			SCL_V_SCALE_RATIO, dal_fixed31_32_u2d19(data->ratios.vert) << 5);
+			SCL_V_SCALE_RATIO, dc_fixpt_u2d19(data->ratios.vert) << 5);
 
 	REG_SET(SCL_HORZ_FILTER_SCALE_RATIO_C, 0,
-			SCL_H_SCALE_RATIO_C, dal_fixed31_32_u2d19(data->ratios.horz_c) << 5);
+			SCL_H_SCALE_RATIO_C, dc_fixpt_u2d19(data->ratios.horz_c) << 5);
 
 	REG_SET(SCL_VERT_FILTER_SCALE_RATIO_C, 0,
-			SCL_V_SCALE_RATIO_C, dal_fixed31_32_u2d19(data->ratios.vert_c) << 5);
+			SCL_V_SCALE_RATIO_C, dc_fixpt_u2d19(data->ratios.vert_c) << 5);
 
 	/*
 	 * 0.24 format for fraction, first five bits zeroed
 	 */
-	init_frac = dal_fixed31_32_u0d19(data->inits.h) << 5;
-	init_int = dal_fixed31_32_floor(data->inits.h);
+	init_frac = dc_fixpt_u0d19(data->inits.h) << 5;
+	init_int = dc_fixpt_floor(data->inits.h);
 	REG_SET_2(SCL_HORZ_FILTER_INIT, 0,
 		SCL_H_INIT_FRAC, init_frac,
 		SCL_H_INIT_INT, init_int);
 
-	init_frac = dal_fixed31_32_u0d19(data->inits.h_c) << 5;
-	init_int = dal_fixed31_32_floor(data->inits.h_c);
+	init_frac = dc_fixpt_u0d19(data->inits.h_c) << 5;
+	init_int = dc_fixpt_floor(data->inits.h_c);
 	REG_SET_2(SCL_HORZ_FILTER_INIT_C, 0,
 		SCL_H_INIT_FRAC_C, init_frac,
 		SCL_H_INIT_INT_C, init_int);
 
-	init_frac = dal_fixed31_32_u0d19(data->inits.v) << 5;
-	init_int = dal_fixed31_32_floor(data->inits.v);
+	init_frac = dc_fixpt_u0d19(data->inits.v) << 5;
+	init_int = dc_fixpt_floor(data->inits.v);
 	REG_SET_2(SCL_VERT_FILTER_INIT, 0,
 		SCL_V_INIT_FRAC, init_frac,
 		SCL_V_INIT_INT, init_int);
 
-	init_frac = dal_fixed31_32_u0d19(data->inits.v_bot) << 5;
-	init_int = dal_fixed31_32_floor(data->inits.v_bot);
+	init_frac = dc_fixpt_u0d19(data->inits.v_bot) << 5;
+	init_int = dc_fixpt_floor(data->inits.v_bot);
 	REG_SET_2(SCL_VERT_FILTER_INIT_BOT, 0,
 		SCL_V_INIT_FRAC_BOT, init_frac,
 		SCL_V_INIT_INT_BOT, init_int);
 
-	init_frac = dal_fixed31_32_u0d19(data->inits.v_c) << 5;
-	init_int = dal_fixed31_32_floor(data->inits.v_c);
+	init_frac = dc_fixpt_u0d19(data->inits.v_c) << 5;
+	init_int = dc_fixpt_floor(data->inits.v_c);
 	REG_SET_2(SCL_VERT_FILTER_INIT_C, 0,
 		SCL_V_INIT_FRAC_C, init_frac,
 		SCL_V_INIT_INT_C, init_int);
 
-	init_frac = dal_fixed31_32_u0d19(data->inits.v_c_bot) << 5;
-	init_int = dal_fixed31_32_floor(data->inits.v_c_bot);
+	init_frac = dc_fixpt_u0d19(data->inits.v_c_bot) << 5;
+	init_int = dc_fixpt_floor(data->inits.v_c_bot);
 	REG_SET_2(SCL_VERT_FILTER_INIT_BOT_C, 0,
 		SCL_V_INIT_FRAC_BOT_C, init_frac,
 		SCL_V_INIT_INT_BOT_C, init_int);
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
index 0cbc83e..185f93b 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
@@ -1054,8 +1054,8 @@ void hubp1_cursor_set_position(
 	ASSERT(param->h_scale_ratio.value);
 
 	if (param->h_scale_ratio.value)
-		dst_x_offset = dal_fixed31_32_floor(dal_fixed31_32_div(
-				dal_fixed31_32_from_int(dst_x_offset),
+		dst_x_offset = dc_fixpt_floor(dc_fixpt_div(
+				dc_fixpt_from_int(dst_x_offset),
 				param->h_scale_ratio));
 
 	if (src_x_offset >= (int)param->viewport_width)
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
index 8adb8dc..50bd754 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -1685,22 +1685,22 @@ static uint16_t fixed_point_to_int_frac(
 
 	uint16_t result;
 
-	uint16_t d = (uint16_t)dal_fixed31_32_floor(
-		dal_fixed31_32_abs(
+	uint16_t d = (uint16_t)dc_fixpt_floor(
+		dc_fixpt_abs(
 			arg));
 
 	if (d <= (uint16_t)(1 << integer_bits) - (1 / (uint16_t)divisor))
-		numerator = (uint16_t)dal_fixed31_32_floor(
-			dal_fixed31_32_mul_int(
+		numerator = (uint16_t)dc_fixpt_floor(
+			dc_fixpt_mul_int(
 				arg,
 				divisor));
 	else {
-		numerator = dal_fixed31_32_floor(
-			dal_fixed31_32_sub(
-				dal_fixed31_32_from_int(
+		numerator = dc_fixpt_floor(
+			dc_fixpt_sub(
+				dc_fixpt_from_int(
 					1LL << integer_bits),
-				dal_fixed31_32_recip(
-					dal_fixed31_32_from_int(
+				dc_fixpt_recip(
+					dc_fixpt_from_int(
 						divisor))));
 	}
 
@@ -1710,8 +1710,8 @@ static uint16_t fixed_point_to_int_frac(
 		result = (uint16_t)(
 		(1 << (integer_bits + fractional_bits + 1)) + numerator);
 
-	if ((result != 0) && dal_fixed31_32_lt(
-		arg, dal_fixed31_32_zero))
+	if ((result != 0) && dc_fixpt_lt(
+		arg, dc_fixpt_zero))
 		result |= 1 << (integer_bits + fractional_bits);
 
 	return result;
@@ -1725,8 +1725,8 @@ void build_prescale_params(struct  dc_bias_and_scale *bias_and_scale,
 			&& plane_state->input_csc_color_matrix.enable_adjustment
 			&& plane_state->coeff_reduction_factor.value != 0) {
 		bias_and_scale->scale_blue = fixed_point_to_int_frac(
-			dal_fixed31_32_mul(plane_state->coeff_reduction_factor,
-					dal_fixed31_32_from_fraction(256, 255)),
+			dc_fixpt_mul(plane_state->coeff_reduction_factor,
+					dc_fixpt_from_fraction(256, 255)),
 				2,
 				13);
 		bias_and_scale->scale_red = bias_and_scale->scale_blue;
@@ -1995,7 +1995,7 @@ static void dcn10_blank_pixel_data(
 
 static void set_hdr_multiplier(struct pipe_ctx *pipe_ctx)
 {
-	struct fixed31_32 multiplier = dal_fixed31_32_from_fraction(
+	struct fixed31_32 multiplier = dc_fixpt_from_fraction(
 			pipe_ctx->plane_state->sdr_white_level, 80);
 	uint32_t hw_mult = 0x1f000; // 1.0 default multiplier
 	struct custom_float_format fmt;
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c
index befd863..653b7b2 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c
@@ -603,11 +603,11 @@ void enc1_stream_encoder_set_mst_bandwidth(
 	struct fixed31_32 avg_time_slots_per_mtp)
 {
 	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
-	uint32_t x = dal_fixed31_32_floor(
+	uint32_t x = dc_fixpt_floor(
 		avg_time_slots_per_mtp);
-	uint32_t y = dal_fixed31_32_ceil(
-		dal_fixed31_32_shl(
-			dal_fixed31_32_sub_int(
+	uint32_t y = dc_fixpt_ceil(
+		dc_fixpt_shl(
+			dc_fixpt_sub_int(
 				avg_time_slots_per_mtp,
 				x),
 			26));
diff --git a/drivers/gpu/drm/amd/display/dc/irq_types.h b/drivers/gpu/drm/amd/display/dc/irq_types.h
index cc3b1bc..0b5f3a2 100644
--- a/drivers/gpu/drm/amd/display/dc/irq_types.h
+++ b/drivers/gpu/drm/amd/display/dc/irq_types.h
@@ -26,6 +26,8 @@
 #ifndef __DAL_IRQ_TYPES_H__
 #define __DAL_IRQ_TYPES_H__
 
+#include "os_types.h"
+
 struct dc_context;
 
 typedef void (*interrupt_handler)(void *);
diff --git a/drivers/gpu/drm/amd/display/include/fixed31_32.h b/drivers/gpu/drm/amd/display/include/fixed31_32.h
index 16cbdb4..b5b8d7d 100644
--- a/drivers/gpu/drm/amd/display/include/fixed31_32.h
+++ b/drivers/gpu/drm/amd/display/include/fixed31_32.h
@@ -50,16 +50,16 @@ struct fixed31_32 {
  * Useful constants
  */
 
-static const struct fixed31_32 dal_fixed31_32_zero = { 0 };
-static const struct fixed31_32 dal_fixed31_32_epsilon = { 1LL };
-static const struct fixed31_32 dal_fixed31_32_half = { 0x80000000LL };
-static const struct fixed31_32 dal_fixed31_32_one = { 0x100000000LL };
+static const struct fixed31_32 dc_fixpt_zero = { 0 };
+static const struct fixed31_32 dc_fixpt_epsilon = { 1LL };
+static const struct fixed31_32 dc_fixpt_half = { 0x80000000LL };
+static const struct fixed31_32 dc_fixpt_one = { 0x100000000LL };
 
-static const struct fixed31_32 dal_fixed31_32_pi = { 13493037705LL };
-static const struct fixed31_32 dal_fixed31_32_two_pi = { 26986075409LL };
-static const struct fixed31_32 dal_fixed31_32_e = { 11674931555LL };
-static const struct fixed31_32 dal_fixed31_32_ln2 = { 2977044471LL };
-static const struct fixed31_32 dal_fixed31_32_ln2_div_2 = { 1488522236LL };
+static const struct fixed31_32 dc_fixpt_pi = { 13493037705LL };
+static const struct fixed31_32 dc_fixpt_two_pi = { 26986075409LL };
+static const struct fixed31_32 dc_fixpt_e = { 11674931555LL };
+static const struct fixed31_32 dc_fixpt_ln2 = { 2977044471LL };
+static const struct fixed31_32 dc_fixpt_ln2_div_2 = { 1488522236LL };
 
 /*
  * @brief
@@ -70,7 +70,7 @@ static const struct fixed31_32 dal_fixed31_32_ln2_div_2 = { 1488522236LL };
  * @brief
  * result = numerator / denominator
  */
-struct fixed31_32 dal_fixed31_32_from_fraction(
+struct fixed31_32 dc_fixpt_from_fraction(
 	long long numerator,
 	long long denominator);
 
@@ -78,8 +78,8 @@ struct fixed31_32 dal_fixed31_32_from_fraction(
  * @brief
  * result = arg
  */
-struct fixed31_32 dal_fixed31_32_from_int_nonconst(long long arg);
-static inline struct fixed31_32 dal_fixed31_32_from_int(long long arg)
+struct fixed31_32 dc_fixpt_from_int_nonconst(long long arg);
+static inline struct fixed31_32 dc_fixpt_from_int(long long arg)
 {
 	if (__builtin_constant_p(arg)) {
 		struct fixed31_32 res;
@@ -87,7 +87,7 @@ static inline struct fixed31_32 dal_fixed31_32_from_int(long long arg)
 		res.value = arg << FIXED31_32_BITS_PER_FRACTIONAL_PART;
 		return res;
 	} else
-		return dal_fixed31_32_from_int_nonconst(arg);
+		return dc_fixpt_from_int_nonconst(arg);
 }
 
 /*
@@ -99,7 +99,7 @@ static inline struct fixed31_32 dal_fixed31_32_from_int(long long arg)
  * @brief
  * result = -arg
  */
-static inline struct fixed31_32 dal_fixed31_32_neg(struct fixed31_32 arg)
+static inline struct fixed31_32 dc_fixpt_neg(struct fixed31_32 arg)
 {
 	struct fixed31_32 res;
 
@@ -112,10 +112,10 @@ static inline struct fixed31_32 dal_fixed31_32_neg(struct fixed31_32 arg)
  * @brief
  * result = abs(arg) := (arg >= 0) ? arg : -arg
  */
-static inline struct fixed31_32 dal_fixed31_32_abs(struct fixed31_32 arg)
+static inline struct fixed31_32 dc_fixpt_abs(struct fixed31_32 arg)
 {
 	if (arg.value < 0)
-		return dal_fixed31_32_neg(arg);
+		return dc_fixpt_neg(arg);
 	else
 		return arg;
 }
@@ -129,7 +129,7 @@ static inline struct fixed31_32 dal_fixed31_32_abs(struct fixed31_32 arg)
  * @brief
  * result = arg1 < arg2
  */
-static inline bool dal_fixed31_32_lt(struct fixed31_32 arg1,
+static inline bool dc_fixpt_lt(struct fixed31_32 arg1,
 				     struct fixed31_32 arg2)
 {
 	return arg1.value < arg2.value;
@@ -139,7 +139,7 @@ static inline bool dal_fixed31_32_lt(struct fixed31_32 arg1,
  * @brief
  * result = arg1 <= arg2
  */
-static inline bool dal_fixed31_32_le(struct fixed31_32 arg1,
+static inline bool dc_fixpt_le(struct fixed31_32 arg1,
 				     struct fixed31_32 arg2)
 {
 	return arg1.value <= arg2.value;
@@ -149,7 +149,7 @@ static inline bool dal_fixed31_32_le(struct fixed31_32 arg1,
  * @brief
  * result = arg1 == arg2
  */
-static inline bool dal_fixed31_32_eq(struct fixed31_32 arg1,
+static inline bool dc_fixpt_eq(struct fixed31_32 arg1,
 				     struct fixed31_32 arg2)
 {
 	return arg1.value == arg2.value;
@@ -159,7 +159,7 @@ static inline bool dal_fixed31_32_eq(struct fixed31_32 arg1,
  * @brief
  * result = min(arg1, arg2) := (arg1 <= arg2) ? arg1 : arg2
  */
-static inline struct fixed31_32 dal_fixed31_32_min(struct fixed31_32 arg1,
+static inline struct fixed31_32 dc_fixpt_min(struct fixed31_32 arg1,
 						   struct fixed31_32 arg2)
 {
 	if (arg1.value <= arg2.value)
@@ -172,7 +172,7 @@ static inline struct fixed31_32 dal_fixed31_32_min(struct fixed31_32 arg1,
  * @brief
  * result = max(arg1, arg2) := (arg1 <= arg2) ? arg2 : arg1
  */
-static inline struct fixed31_32 dal_fixed31_32_max(struct fixed31_32 arg1,
+static inline struct fixed31_32 dc_fixpt_max(struct fixed31_32 arg1,
 						   struct fixed31_32 arg2)
 {
 	if (arg1.value <= arg2.value)
@@ -187,14 +187,14 @@ static inline struct fixed31_32 dal_fixed31_32_max(struct fixed31_32 arg1,
  * result = | arg, when min_value < arg < max_value
  *          | max_value, when arg >= max_value
  */
-static inline struct fixed31_32 dal_fixed31_32_clamp(
+static inline struct fixed31_32 dc_fixpt_clamp(
 	struct fixed31_32 arg,
 	struct fixed31_32 min_value,
 	struct fixed31_32 max_value)
 {
-	if (dal_fixed31_32_le(arg, min_value))
+	if (dc_fixpt_le(arg, min_value))
 		return min_value;
-	else if (dal_fixed31_32_le(max_value, arg))
+	else if (dc_fixpt_le(max_value, arg))
 		return max_value;
 	else
 		return arg;
@@ -209,7 +209,7 @@ static inline struct fixed31_32 dal_fixed31_32_clamp(
  * @brief
  * result = arg << shift
  */
-struct fixed31_32 dal_fixed31_32_shl(
+struct fixed31_32 dc_fixpt_shl(
 	struct fixed31_32 arg,
 	unsigned char shift);
 
@@ -217,7 +217,7 @@ struct fixed31_32 dal_fixed31_32_shl(
  * @brief
  * result = arg >> shift
  */
-static inline struct fixed31_32 dal_fixed31_32_shr(
+static inline struct fixed31_32 dc_fixpt_shr(
 	struct fixed31_32 arg,
 	unsigned char shift)
 {
@@ -235,7 +235,7 @@ static inline struct fixed31_32 dal_fixed31_32_shr(
  * @brief
  * result = arg1 + arg2
  */
-struct fixed31_32 dal_fixed31_32_add(
+struct fixed31_32 dc_fixpt_add(
 	struct fixed31_32 arg1,
 	struct fixed31_32 arg2);
 
@@ -243,18 +243,18 @@ struct fixed31_32 dal_fixed31_32_add(
  * @brief
  * result = arg1 + arg2
  */
-static inline struct fixed31_32 dal_fixed31_32_add_int(struct fixed31_32 arg1,
+static inline struct fixed31_32 dc_fixpt_add_int(struct fixed31_32 arg1,
 						       int arg2)
 {
-	return dal_fixed31_32_add(arg1,
-				  dal_fixed31_32_from_int(arg2));
+	return dc_fixpt_add(arg1,
+				  dc_fixpt_from_int(arg2));
 }
 
 /*
  * @brief
  * result = arg1 - arg2
  */
-struct fixed31_32 dal_fixed31_32_sub(
+struct fixed31_32 dc_fixpt_sub(
 	struct fixed31_32 arg1,
 	struct fixed31_32 arg2);
 
@@ -262,11 +262,11 @@ struct fixed31_32 dal_fixed31_32_sub(
  * @brief
  * result = arg1 - arg2
  */
-static inline struct fixed31_32 dal_fixed31_32_sub_int(struct fixed31_32 arg1,
+static inline struct fixed31_32 dc_fixpt_sub_int(struct fixed31_32 arg1,
 						       int arg2)
 {
-	return dal_fixed31_32_sub(arg1,
-				  dal_fixed31_32_from_int(arg2));
+	return dc_fixpt_sub(arg1,
+				  dc_fixpt_from_int(arg2));
 }
 
 
@@ -279,7 +279,7 @@ static inline struct fixed31_32 dal_fixed31_32_sub_int(struct fixed31_32 arg1,
  * @brief
  * result = arg1 * arg2
  */
-struct fixed31_32 dal_fixed31_32_mul(
+struct fixed31_32 dc_fixpt_mul(
 	struct fixed31_32 arg1,
 	struct fixed31_32 arg2);
 
@@ -288,39 +288,39 @@ struct fixed31_32 dal_fixed31_32_mul(
  * @brief
  * result = arg1 * arg2
  */
-static inline struct fixed31_32 dal_fixed31_32_mul_int(struct fixed31_32 arg1,
+static inline struct fixed31_32 dc_fixpt_mul_int(struct fixed31_32 arg1,
 						       int arg2)
 {
-	return dal_fixed31_32_mul(arg1,
-				  dal_fixed31_32_from_int(arg2));
+	return dc_fixpt_mul(arg1,
+				  dc_fixpt_from_int(arg2));
 }
 
 /*
  * @brief
  * result = square(arg) := arg * arg
  */
-struct fixed31_32 dal_fixed31_32_sqr(
+struct fixed31_32 dc_fixpt_sqr(
 	struct fixed31_32 arg);
 
 /*
  * @brief
  * result = arg1 / arg2
  */
-static inline struct fixed31_32 dal_fixed31_32_div_int(struct fixed31_32 arg1,
+static inline struct fixed31_32 dc_fixpt_div_int(struct fixed31_32 arg1,
 						       long long arg2)
 {
-	return dal_fixed31_32_from_fraction(arg1.value,
-					    dal_fixed31_32_from_int(arg2).value);
+	return dc_fixpt_from_fraction(arg1.value,
+					    dc_fixpt_from_int(arg2).value);
 }
 
 /*
  * @brief
  * result = arg1 / arg2
  */
-static inline struct fixed31_32 dal_fixed31_32_div(struct fixed31_32 arg1,
+static inline struct fixed31_32 dc_fixpt_div(struct fixed31_32 arg1,
 						   struct fixed31_32 arg2)
 {
-	return dal_fixed31_32_from_fraction(arg1.value,
+	return dc_fixpt_from_fraction(arg1.value,
 					    arg2.value);
 }
 
@@ -336,7 +336,7 @@ static inline struct fixed31_32 dal_fixed31_32_div(struct fixed31_32 arg1,
  * @note
  * No special actions taken in case argument is zero.
  */
-struct fixed31_32 dal_fixed31_32_recip(
+struct fixed31_32 dc_fixpt_recip(
 	struct fixed31_32 arg);
 
 /*
@@ -352,7 +352,7 @@ struct fixed31_32 dal_fixed31_32_recip(
  * Argument specified in radians,
  * internally it's normalized to [-2pi...2pi] range.
  */
-struct fixed31_32 dal_fixed31_32_sinc(
+struct fixed31_32 dc_fixpt_sinc(
 	struct fixed31_32 arg);
 
 /*
@@ -363,7 +363,7 @@ struct fixed31_32 dal_fixed31_32_sinc(
  * Argument specified in radians,
  * internally it's normalized to [-2pi...2pi] range.
  */
-struct fixed31_32 dal_fixed31_32_sin(
+struct fixed31_32 dc_fixpt_sin(
 	struct fixed31_32 arg);
 
 /*
@@ -376,7 +376,7 @@ struct fixed31_32 dal_fixed31_32_sin(
  * passing arguments outside that range
  * will cause incorrect result!
  */
-struct fixed31_32 dal_fixed31_32_cos(
+struct fixed31_32 dc_fixpt_cos(
 	struct fixed31_32 arg);
 
 /*
@@ -391,7 +391,7 @@ struct fixed31_32 dal_fixed31_32_cos(
  * @note
  * Currently, function is verified for abs(arg) <= 1.
  */
-struct fixed31_32 dal_fixed31_32_exp(
+struct fixed31_32 dc_fixpt_exp(
 	struct fixed31_32 arg);
 
 /*
@@ -404,7 +404,7 @@ struct fixed31_32 dal_fixed31_32_exp(
  * Currently, no special actions taken
  * in case of invalid argument(s). Take care!
  */
-struct fixed31_32 dal_fixed31_32_log(
+struct fixed31_32 dc_fixpt_log(
 	struct fixed31_32 arg);
 
 /*
@@ -419,7 +419,7 @@ struct fixed31_32 dal_fixed31_32_log(
  * @note
  * Currently, abs(arg1) should be less than 1. Take care!
  */
-struct fixed31_32 dal_fixed31_32_pow(
+struct fixed31_32 dc_fixpt_pow(
 	struct fixed31_32 arg1,
 	struct fixed31_32 arg2);
 
@@ -432,21 +432,21 @@ struct fixed31_32 dal_fixed31_32_pow(
  * @brief
  * result = floor(arg) := greatest integer lower than or equal to arg
  */
-int dal_fixed31_32_floor(
+int dc_fixpt_floor(
 	struct fixed31_32 arg);
 
 /*
  * @brief
  * result = round(arg) := integer nearest to arg
  */
-int dal_fixed31_32_round(
+int dc_fixpt_round(
 	struct fixed31_32 arg);
 
 /*
  * @brief
  * result = ceil(arg) := lowest integer greater than or equal to arg
  */
-int dal_fixed31_32_ceil(
+int dc_fixpt_ceil(
 	struct fixed31_32 arg);
 
 /* the following two function are used in scaler hw programming to convert fixed
@@ -455,20 +455,20 @@ int dal_fixed31_32_ceil(
  * fractional
  */
 
-unsigned int dal_fixed31_32_u2d19(
+unsigned int dc_fixpt_u2d19(
 	struct fixed31_32 arg);
 
-unsigned int dal_fixed31_32_u0d19(
+unsigned int dc_fixpt_u0d19(
 	struct fixed31_32 arg);
 
 
-unsigned int dal_fixed31_32_clamp_u0d14(
+unsigned int dc_fixpt_clamp_u0d14(
 	struct fixed31_32 arg);
 
-unsigned int dal_fixed31_32_clamp_u0d10(
+unsigned int dc_fixpt_clamp_u0d10(
 	struct fixed31_32 arg);
 
-int dal_fixed31_32_s4d19(
+int dc_fixpt_s4d19(
 	struct fixed31_32 arg);
 
 #endif
diff --git a/drivers/gpu/drm/amd/display/include/fixed32_32.h b/drivers/gpu/drm/amd/display/include/fixed32_32.h
deleted file mode 100644
index 9c70341..0000000
--- a/drivers/gpu/drm/amd/display/include/fixed32_32.h
+++ /dev/null
@@ -1,129 +0,0 @@
-/*
- * Copyright 2012-15 Advanced Micro Devices, Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: AMD
- *
- */
-
-
-#ifndef __DAL_FIXED32_32_H__
-#define __DAL_FIXED32_32_H__
-
-#include "os_types.h"
-
-struct fixed32_32 {
-	uint64_t value;
-};
-
-static const struct fixed32_32 dal_fixed32_32_zero = { 0 };
-static const struct fixed32_32 dal_fixed32_32_one = { 0x100000000LL };
-static const struct fixed32_32 dal_fixed32_32_half = { 0x80000000LL };
-
-struct fixed32_32 dal_fixed32_32_from_fraction(uint32_t n, uint32_t d);
-static inline struct fixed32_32 dal_fixed32_32_from_int(uint32_t value)
-{
-	struct fixed32_32 fx;
-
-	fx.value = (uint64_t)value<<32;
-	return fx;
-}
-
-struct fixed32_32 dal_fixed32_32_add(
-	struct fixed32_32 lhs,
-	struct fixed32_32 rhs);
-struct fixed32_32 dal_fixed32_32_add_int(
-	struct fixed32_32 lhs,
-	uint32_t rhs);
-struct fixed32_32 dal_fixed32_32_sub(
-	struct fixed32_32 lhs,
-	struct fixed32_32 rhs);
-struct fixed32_32 dal_fixed32_32_sub_int(
-	struct fixed32_32 lhs,
-	uint32_t rhs);
-struct fixed32_32 dal_fixed32_32_mul(
-	struct fixed32_32 lhs,
-	struct fixed32_32 rhs);
-struct fixed32_32 dal_fixed32_32_mul_int(
-	struct fixed32_32 lhs,
-	uint32_t rhs);
-struct fixed32_32 dal_fixed32_32_div(
-	struct fixed32_32 lhs,
-	struct fixed32_32 rhs);
-struct fixed32_32 dal_fixed32_32_div_int(
-	struct fixed32_32 lhs,
-	uint32_t rhs);
-
-static inline struct fixed32_32 dal_fixed32_32_min(struct fixed32_32 lhs,
-						   struct fixed32_32 rhs)
-{
-	return (lhs.value < rhs.value) ? lhs : rhs;
-}
-
-static inline struct fixed32_32 dal_fixed32_32_max(struct fixed32_32 lhs,
-						   struct fixed32_32 rhs)
-{
-	return (lhs.value > rhs.value) ? lhs : rhs;
-}
-
-static inline bool dal_fixed32_32_gt(struct fixed32_32 lhs, struct fixed32_32 rhs)
-{
-	return lhs.value > rhs.value;
-}
-
-static inline bool dal_fixed32_32_gt_int(struct fixed32_32 lhs, uint32_t rhs)
-{
-	return lhs.value > ((uint64_t)rhs<<32);
-}
-
-static inline bool dal_fixed32_32_lt(struct fixed32_32 lhs, struct fixed32_32 rhs)
-{
-	return lhs.value < rhs.value;
-}
-
-static inline bool dal_fixed32_32_lt_int(struct fixed32_32 lhs, uint32_t rhs)
-{
-	return lhs.value < ((uint64_t)rhs<<32);
-}
-
-static inline bool dal_fixed32_32_le(struct fixed32_32 lhs, struct fixed32_32 rhs)
-{
-	return lhs.value <= rhs.value;
-}
-
-static inline bool dal_fixed32_32_le_int(struct fixed32_32 lhs, uint32_t rhs)
-{
-	return lhs.value <= ((uint64_t)rhs<<32);
-}
-
-static inline bool dal_fixed32_32_eq(struct fixed32_32 lhs, struct fixed32_32 rhs)
-{
-	return lhs.value == rhs.value;
-}
-
-uint32_t dal_fixed32_32_ceil(struct fixed32_32 value);
-static inline uint32_t dal_fixed32_32_floor(struct fixed32_32 value)
-{
-	return value.value>>32;
-}
-
-uint32_t dal_fixed32_32_round(struct fixed32_32 value);
-
-#endif
diff --git a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c
index 15e5b72..29d2ec8 100644
--- a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c
+++ b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c
@@ -43,7 +43,7 @@ static bool de_pq_initialized; /* = false; */
 /* one-time setup of X points */
 void setup_x_points_distribution(void)
 {
-	struct fixed31_32 region_size = dal_fixed31_32_from_int(128);
+	struct fixed31_32 region_size = dc_fixpt_from_int(128);
 	int32_t segment;
 	uint32_t seg_offset;
 	uint32_t index;
@@ -53,8 +53,8 @@ void setup_x_points_distribution(void)
 	coordinates_x[MAX_HW_POINTS + 1].x = region_size;
 
 	for (segment = 6; segment > (6 - NUM_REGIONS); segment--) {
-		region_size = dal_fixed31_32_div_int(region_size, 2);
-		increment = dal_fixed31_32_div_int(region_size,
+		region_size = dc_fixpt_div_int(region_size, 2);
+		increment = dc_fixpt_div_int(region_size,
 						NUM_PTS_IN_REGION);
 		seg_offset = (segment + (NUM_REGIONS - 7)) * NUM_PTS_IN_REGION;
 		coordinates_x[seg_offset].x = region_size;
@@ -62,7 +62,7 @@ void setup_x_points_distribution(void)
 		for (index = seg_offset + 1;
 				index < seg_offset + NUM_PTS_IN_REGION;
 				index++) {
-			coordinates_x[index].x = dal_fixed31_32_add
+			coordinates_x[index].x = dc_fixpt_add
 					(coordinates_x[index-1].x, increment);
 		}
 	}
@@ -72,63 +72,63 @@ static void compute_pq(struct fixed31_32 in_x, struct fixed31_32 *out_y)
 {
 	/* consts for PQ gamma formula. */
 	const struct fixed31_32 m1 =
-		dal_fixed31_32_from_fraction(159301758, 1000000000);
+		dc_fixpt_from_fraction(159301758, 1000000000);
 	const struct fixed31_32 m2 =
-		dal_fixed31_32_from_fraction(7884375, 100000);
+		dc_fixpt_from_fraction(7884375, 100000);
 	const struct fixed31_32 c1 =
-		dal_fixed31_32_from_fraction(8359375, 10000000);
+		dc_fixpt_from_fraction(8359375, 10000000);
 	const struct fixed31_32 c2 =
-		dal_fixed31_32_from_fraction(188515625, 10000000);
+		dc_fixpt_from_fraction(188515625, 10000000);
 	const struct fixed31_32 c3 =
-		dal_fixed31_32_from_fraction(186875, 10000);
+		dc_fixpt_from_fraction(186875, 10000);
 
 	struct fixed31_32 l_pow_m1;
 	struct fixed31_32 base;
 
-	if (dal_fixed31_32_lt(in_x, dal_fixed31_32_zero))
-		in_x = dal_fixed31_32_zero;
+	if (dc_fixpt_lt(in_x, dc_fixpt_zero))
+		in_x = dc_fixpt_zero;
 
-	l_pow_m1 = dal_fixed31_32_pow(in_x, m1);
-	base = dal_fixed31_32_div(
-			dal_fixed31_32_add(c1,
-					(dal_fixed31_32_mul(c2, l_pow_m1))),
-			dal_fixed31_32_add(dal_fixed31_32_one,
-					(dal_fixed31_32_mul(c3, l_pow_m1))));
-	*out_y = dal_fixed31_32_pow(base, m2);
+	l_pow_m1 = dc_fixpt_pow(in_x, m1);
+	base = dc_fixpt_div(
+			dc_fixpt_add(c1,
+					(dc_fixpt_mul(c2, l_pow_m1))),
+			dc_fixpt_add(dc_fixpt_one,
+					(dc_fixpt_mul(c3, l_pow_m1))));
+	*out_y = dc_fixpt_pow(base, m2);
 }
 
 static void compute_de_pq(struct fixed31_32 in_x, struct fixed31_32 *out_y)
 {
 	/* consts for dePQ gamma formula. */
 	const struct fixed31_32 m1 =
-		dal_fixed31_32_from_fraction(159301758, 1000000000);
+		dc_fixpt_from_fraction(159301758, 1000000000);
 	const struct fixed31_32 m2 =
-		dal_fixed31_32_from_fraction(7884375, 100000);
+		dc_fixpt_from_fraction(7884375, 100000);
 	const struct fixed31_32 c1 =
-		dal_fixed31_32_from_fraction(8359375, 10000000);
+		dc_fixpt_from_fraction(8359375, 10000000);
 	const struct fixed31_32 c2 =
-		dal_fixed31_32_from_fraction(188515625, 10000000);
+		dc_fixpt_from_fraction(188515625, 10000000);
 	const struct fixed31_32 c3 =
-		dal_fixed31_32_from_fraction(186875, 10000);
+		dc_fixpt_from_fraction(186875, 10000);
 
 	struct fixed31_32 l_pow_m1;
 	struct fixed31_32 base, div;
 
 
-	if (dal_fixed31_32_lt(in_x, dal_fixed31_32_zero))
-		in_x = dal_fixed31_32_zero;
+	if (dc_fixpt_lt(in_x, dc_fixpt_zero))
+		in_x = dc_fixpt_zero;
 
-	l_pow_m1 = dal_fixed31_32_pow(in_x,
-			dal_fixed31_32_div(dal_fixed31_32_one, m2));
-	base = dal_fixed31_32_sub(l_pow_m1, c1);
+	l_pow_m1 = dc_fixpt_pow(in_x,
+			dc_fixpt_div(dc_fixpt_one, m2));
+	base = dc_fixpt_sub(l_pow_m1, c1);
 
-	if (dal_fixed31_32_lt(base, dal_fixed31_32_zero))
-		base = dal_fixed31_32_zero;
+	if (dc_fixpt_lt(base, dc_fixpt_zero))
+		base = dc_fixpt_zero;
 
-	div = dal_fixed31_32_sub(c2, dal_fixed31_32_mul(c3, l_pow_m1));
+	div = dc_fixpt_sub(c2, dc_fixpt_mul(c3, l_pow_m1));
 
-	*out_y = dal_fixed31_32_pow(dal_fixed31_32_div(base, div),
-			dal_fixed31_32_div(dal_fixed31_32_one, m1));
+	*out_y = dc_fixpt_pow(dc_fixpt_div(base, div),
+			dc_fixpt_div(dc_fixpt_one, m1));
 
 }
 /* one-time pre-compute PQ values - only for sdr_white_level 80 */
@@ -138,14 +138,14 @@ void precompute_pq(void)
 	struct fixed31_32 x;
 	const struct hw_x_point *coord_x = coordinates_x + 32;
 	struct fixed31_32 scaling_factor =
-			dal_fixed31_32_from_fraction(80, 10000);
+			dc_fixpt_from_fraction(80, 10000);
 
 	/* pow function has problems with arguments too small */
 	for (i = 0; i < 32; i++)
-		pq_table[i] = dal_fixed31_32_zero;
+		pq_table[i] = dc_fixpt_zero;
 
 	for (i = 32; i <= MAX_HW_POINTS; i++) {
-		x = dal_fixed31_32_mul(coord_x->x, scaling_factor);
+		x = dc_fixpt_mul(coord_x->x, scaling_factor);
 		compute_pq(x, &pq_table[i]);
 		++coord_x;
 	}
@@ -158,7 +158,7 @@ void precompute_de_pq(void)
 	struct fixed31_32  y;
 	uint32_t begin_index, end_index;
 
-	struct fixed31_32 scaling_factor = dal_fixed31_32_from_int(125);
+	struct fixed31_32 scaling_factor = dc_fixpt_from_int(125);
 
 	/* X points is 2^-25 to 2^7
 	 * De-gamma X is 2^-12 to 2^0 – we are skipping first -12-(-25) = 13 regions
@@ -167,11 +167,11 @@ void precompute_de_pq(void)
 	end_index = begin_index + 12 * NUM_PTS_IN_REGION;
 
 	for (i = 0; i <= begin_index; i++)
-		de_pq_table[i] = dal_fixed31_32_zero;
+		de_pq_table[i] = dc_fixpt_zero;
 
 	for (; i <= end_index; i++) {
 		compute_de_pq(coordinates_x[i].x, &y);
-		de_pq_table[i] = dal_fixed31_32_mul(y, scaling_factor);
+		de_pq_table[i] = dc_fixpt_mul(y, scaling_factor);
 	}
 
 	for (; i <= MAX_HW_POINTS; i++)
@@ -195,15 +195,15 @@ static void build_coefficients(struct gamma_coefficients *coefficients, bool is_
 	uint32_t index = is_2_4 == true ? 0:1;
 
 	do {
-		coefficients->a0[i] = dal_fixed31_32_from_fraction(
+		coefficients->a0[i] = dc_fixpt_from_fraction(
 			numerator01[index], 10000000);
-		coefficients->a1[i] = dal_fixed31_32_from_fraction(
+		coefficients->a1[i] = dc_fixpt_from_fraction(
 			numerator02[index], 1000);
-		coefficients->a2[i] = dal_fixed31_32_from_fraction(
+		coefficients->a2[i] = dc_fixpt_from_fraction(
 			numerator03[index], 1000);
-		coefficients->a3[i] = dal_fixed31_32_from_fraction(
+		coefficients->a3[i] = dc_fixpt_from_fraction(
 			numerator04[index], 1000);
-		coefficients->user_gamma[i] = dal_fixed31_32_from_fraction(
+		coefficients->user_gamma[i] = dc_fixpt_from_fraction(
 			numerator05[index], 1000);
 
 		++i;
@@ -218,33 +218,33 @@ static struct fixed31_32 translate_from_linear_space(
 	struct fixed31_32 a3,
 	struct fixed31_32 gamma)
 {
-	const struct fixed31_32 one = dal_fixed31_32_from_int(1);
+	const struct fixed31_32 one = dc_fixpt_from_int(1);
 
-	if (dal_fixed31_32_lt(one, arg))
+	if (dc_fixpt_lt(one, arg))
 		return one;
 
-	if (dal_fixed31_32_le(arg, dal_fixed31_32_neg(a0)))
-		return dal_fixed31_32_sub(
+	if (dc_fixpt_le(arg, dc_fixpt_neg(a0)))
+		return dc_fixpt_sub(
 			a2,
-			dal_fixed31_32_mul(
-				dal_fixed31_32_add(
+			dc_fixpt_mul(
+				dc_fixpt_add(
 					one,
 					a3),
-				dal_fixed31_32_pow(
-					dal_fixed31_32_neg(arg),
-					dal_fixed31_32_recip(gamma))));
-	else if (dal_fixed31_32_le(a0, arg))
-		return dal_fixed31_32_sub(
-			dal_fixed31_32_mul(
-				dal_fixed31_32_add(
+				dc_fixpt_pow(
+					dc_fixpt_neg(arg),
+					dc_fixpt_recip(gamma))));
+	else if (dc_fixpt_le(a0, arg))
+		return dc_fixpt_sub(
+			dc_fixpt_mul(
+				dc_fixpt_add(
 					one,
 					a3),
-				dal_fixed31_32_pow(
+				dc_fixpt_pow(
 					arg,
-					dal_fixed31_32_recip(gamma))),
+					dc_fixpt_recip(gamma))),
 			a2);
 	else
-		return dal_fixed31_32_mul(
+		return dc_fixpt_mul(
 			arg,
 			a1);
 }
@@ -259,25 +259,25 @@ static struct fixed31_32 translate_to_linear_space(
 {
 	struct fixed31_32 linear;
 
-	a0 = dal_fixed31_32_mul(a0, a1);
-	if (dal_fixed31_32_le(arg, dal_fixed31_32_neg(a0)))
+	a0 = dc_fixpt_mul(a0, a1);
+	if (dc_fixpt_le(arg, dc_fixpt_neg(a0)))
 
-		linear = dal_fixed31_32_neg(
-				 dal_fixed31_32_pow(
-				 dal_fixed31_32_div(
-				 dal_fixed31_32_sub(a2, arg),
-				 dal_fixed31_32_add(
-				 dal_fixed31_32_one, a3)), gamma));
+		linear = dc_fixpt_neg(
+				 dc_fixpt_pow(
+				 dc_fixpt_div(
+				 dc_fixpt_sub(a2, arg),
+				 dc_fixpt_add(
+				 dc_fixpt_one, a3)), gamma));
 
-	else if (dal_fixed31_32_le(dal_fixed31_32_neg(a0), arg) &&
-			 dal_fixed31_32_le(arg, a0))
-		linear = dal_fixed31_32_div(arg, a1);
+	else if (dc_fixpt_le(dc_fixpt_neg(a0), arg) &&
+			 dc_fixpt_le(arg, a0))
+		linear = dc_fixpt_div(arg, a1);
 	else
-		linear =  dal_fixed31_32_pow(
-					dal_fixed31_32_div(
-					dal_fixed31_32_add(a2, arg),
-					dal_fixed31_32_add(
-					dal_fixed31_32_one, a3)), gamma);
+		linear =  dc_fixpt_pow(
+					dc_fixpt_div(
+					dc_fixpt_add(a2, arg),
+					dc_fixpt_add(
+					dc_fixpt_one, a3)), gamma);
 
 	return linear;
 }
@@ -352,8 +352,8 @@ static bool find_software_points(
 				right = axis_x[max_number - 1].b;
 		}
 
-		if (dal_fixed31_32_le(left, hw_point) &&
-			dal_fixed31_32_le(hw_point, right)) {
+		if (dc_fixpt_le(left, hw_point) &&
+			dc_fixpt_le(hw_point, right)) {
 			*index_to_start = i;
 			*index_left = i;
 
@@ -366,7 +366,7 @@ static bool find_software_points(
 
 			return true;
 		} else if ((i == *index_to_start) &&
-			dal_fixed31_32_le(hw_point, left)) {
+			dc_fixpt_le(hw_point, left)) {
 			*index_to_start = i;
 			*index_left = i;
 			*index_right = i;
@@ -375,7 +375,7 @@ static bool find_software_points(
 
 			return true;
 		} else if ((i == max_number - 1) &&
-			dal_fixed31_32_le(right, hw_point)) {
+			dc_fixpt_le(right, hw_point)) {
 			*index_to_start = i;
 			*index_left = i;
 			*index_right = i;
@@ -457,17 +457,17 @@ static bool build_custom_gamma_mapping_coefficients_worker(
 		}
 
 		if (hw_pos == HW_POINT_POSITION_MIDDLE)
-			point->coeff = dal_fixed31_32_div(
-				dal_fixed31_32_sub(
+			point->coeff = dc_fixpt_div(
+				dc_fixpt_sub(
 					coord_x,
 					left_pos),
-				dal_fixed31_32_sub(
+				dc_fixpt_sub(
 					right_pos,
 					left_pos));
 		else if (hw_pos == HW_POINT_POSITION_LEFT)
-			point->coeff = dal_fixed31_32_zero;
+			point->coeff = dc_fixpt_zero;
 		else if (hw_pos == HW_POINT_POSITION_RIGHT)
-			point->coeff = dal_fixed31_32_from_int(2);
+			point->coeff = dc_fixpt_from_int(2);
 		else {
 			BREAK_TO_DEBUGGER();
 			return false;
@@ -502,45 +502,45 @@ static struct fixed31_32 calculate_mapped_value(
 
 	if ((point->left_index < 0) || (point->left_index > max_index)) {
 		BREAK_TO_DEBUGGER();
-		return dal_fixed31_32_zero;
+		return dc_fixpt_zero;
 	}
 
 	if ((point->right_index < 0) || (point->right_index > max_index)) {
 		BREAK_TO_DEBUGGER();
-		return dal_fixed31_32_zero;
+		return dc_fixpt_zero;
 	}
 
 	if (point->pos == HW_POINT_POSITION_MIDDLE)
 		if (channel == CHANNEL_NAME_RED)
-			result = dal_fixed31_32_add(
-				dal_fixed31_32_mul(
+			result = dc_fixpt_add(
+				dc_fixpt_mul(
 					point->coeff,
-					dal_fixed31_32_sub(
+					dc_fixpt_sub(
 						rgb[point->right_index].r,
 						rgb[point->left_index].r)),
 				rgb[point->left_index].r);
 		else if (channel == CHANNEL_NAME_GREEN)
-			result = dal_fixed31_32_add(
-				dal_fixed31_32_mul(
+			result = dc_fixpt_add(
+				dc_fixpt_mul(
 					point->coeff,
-					dal_fixed31_32_sub(
+					dc_fixpt_sub(
 						rgb[point->right_index].g,
 						rgb[point->left_index].g)),
 				rgb[point->left_index].g);
 		else
-			result = dal_fixed31_32_add(
-				dal_fixed31_32_mul(
+			result = dc_fixpt_add(
+				dc_fixpt_mul(
 					point->coeff,
-					dal_fixed31_32_sub(
+					dc_fixpt_sub(
 						rgb[point->right_index].b,
 						rgb[point->left_index].b)),
 				rgb[point->left_index].b);
 	else if (point->pos == HW_POINT_POSITION_LEFT) {
 		BREAK_TO_DEBUGGER();
-		result = dal_fixed31_32_zero;
+		result = dc_fixpt_zero;
 	} else {
 		BREAK_TO_DEBUGGER();
-		result = dal_fixed31_32_one;
+		result = dc_fixpt_one;
 	}
 
 	return result;
@@ -558,7 +558,7 @@ static void build_pq(struct pwl_float_data_ex *rgb_regamma,
 	struct fixed31_32 x;
 	struct fixed31_32 output;
 	struct fixed31_32 scaling_factor =
-			dal_fixed31_32_from_fraction(sdr_white_level, 10000);
+			dc_fixpt_from_fraction(sdr_white_level, 10000);
 
 	if (!pq_initialized && sdr_white_level == 80) {
 		precompute_pq();
@@ -579,15 +579,15 @@ static void build_pq(struct pwl_float_data_ex *rgb_regamma,
 		if (sdr_white_level == 80) {
 			output = pq_table[i];
 		} else {
-			x = dal_fixed31_32_mul(coord_x->x, scaling_factor);
+			x = dc_fixpt_mul(coord_x->x, scaling_factor);
 			compute_pq(x, &output);
 		}
 
 		/* should really not happen? */
-		if (dal_fixed31_32_lt(output, dal_fixed31_32_zero))
-			output = dal_fixed31_32_zero;
-		else if (dal_fixed31_32_lt(dal_fixed31_32_one, output))
-			output = dal_fixed31_32_one;
+		if (dc_fixpt_lt(output, dc_fixpt_zero))
+			output = dc_fixpt_zero;
+		else if (dc_fixpt_lt(dc_fixpt_one, output))
+			output = dc_fixpt_one;
 
 		rgb->r = output;
 		rgb->g = output;
@@ -605,7 +605,7 @@ static void build_de_pq(struct pwl_float_data_ex *de_pq,
 	uint32_t i;
 	struct fixed31_32 output;
 
-	struct fixed31_32 scaling_factor = dal_fixed31_32_from_int(125);
+	struct fixed31_32 scaling_factor = dc_fixpt_from_int(125);
 
 	if (!de_pq_initialized) {
 		precompute_de_pq();
@@ -616,9 +616,9 @@ static void build_de_pq(struct pwl_float_data_ex *de_pq,
 	for (i = 0; i <= hw_points_num; i++) {
 		output = de_pq_table[i];
 		/* should really not happen? */
-		if (dal_fixed31_32_lt(output, dal_fixed31_32_zero))
-			output = dal_fixed31_32_zero;
-		else if (dal_fixed31_32_lt(scaling_factor, output))
+		if (dc_fixpt_lt(output, dc_fixpt_zero))
+			output = dc_fixpt_zero;
+		else if (dc_fixpt_lt(scaling_factor, output))
 			output = scaling_factor;
 		de_pq[i].r = output;
 		de_pq[i].g = output;
@@ -670,9 +670,9 @@ static void build_degamma(struct pwl_float_data_ex *curve,
 	end_index = begin_index + 12 * NUM_PTS_IN_REGION;
 
 	while (i != begin_index) {
-		curve[i].r = dal_fixed31_32_zero;
-		curve[i].g = dal_fixed31_32_zero;
-		curve[i].b = dal_fixed31_32_zero;
+		curve[i].r = dc_fixpt_zero;
+		curve[i].g = dc_fixpt_zero;
+		curve[i].b = dc_fixpt_zero;
 		i++;
 	}
 
@@ -684,9 +684,9 @@ static void build_degamma(struct pwl_float_data_ex *curve,
 		i++;
 	}
 	while (i != hw_points_num + 1) {
-		curve[i].r = dal_fixed31_32_one;
-		curve[i].g = dal_fixed31_32_one;
-		curve[i].b = dal_fixed31_32_one;
+		curve[i].r = dc_fixpt_one;
+		curve[i].g = dc_fixpt_one;
+		curve[i].b = dc_fixpt_one;
 		i++;
 	}
 }
@@ -695,8 +695,8 @@ static void scale_gamma(struct pwl_float_data *pwl_rgb,
 		const struct dc_gamma *ramp,
 		struct dividers dividers)
 {
-	const struct fixed31_32 max_driver = dal_fixed31_32_from_int(0xFFFF);
-	const struct fixed31_32 max_os = dal_fixed31_32_from_int(0xFF00);
+	const struct fixed31_32 max_driver = dc_fixpt_from_int(0xFFFF);
+	const struct fixed31_32 max_os = dc_fixpt_from_int(0xFF00);
 	struct fixed31_32 scaler = max_os;
 	uint32_t i;
 	struct pwl_float_data *rgb = pwl_rgb;
@@ -705,9 +705,9 @@ static void scale_gamma(struct pwl_float_data *pwl_rgb,
 	i = 0;
 
 	do {
-		if (dal_fixed31_32_lt(max_os, ramp->entries.red[i]) ||
-			dal_fixed31_32_lt(max_os, ramp->entries.green[i]) ||
-			dal_fixed31_32_lt(max_os, ramp->entries.blue[i])) {
+		if (dc_fixpt_lt(max_os, ramp->entries.red[i]) ||
+			dc_fixpt_lt(max_os, ramp->entries.green[i]) ||
+			dc_fixpt_lt(max_os, ramp->entries.blue[i])) {
 			scaler = max_driver;
 			break;
 		}
@@ -717,40 +717,40 @@ static void scale_gamma(struct pwl_float_data *pwl_rgb,
 	i = 0;
 
 	do {
-		rgb->r = dal_fixed31_32_div(
+		rgb->r = dc_fixpt_div(
 			ramp->entries.red[i], scaler);
-		rgb->g = dal_fixed31_32_div(
+		rgb->g = dc_fixpt_div(
 			ramp->entries.green[i], scaler);
-		rgb->b = dal_fixed31_32_div(
+		rgb->b = dc_fixpt_div(
 			ramp->entries.blue[i], scaler);
 
 		++rgb;
 		++i;
 	} while (i != ramp->num_entries);
 
-	rgb->r = dal_fixed31_32_mul(rgb_last->r,
+	rgb->r = dc_fixpt_mul(rgb_last->r,
 			dividers.divider1);
-	rgb->g = dal_fixed31_32_mul(rgb_last->g,
+	rgb->g = dc_fixpt_mul(rgb_last->g,
 			dividers.divider1);
-	rgb->b = dal_fixed31_32_mul(rgb_last->b,
+	rgb->b = dc_fixpt_mul(rgb_last->b,
 			dividers.divider1);
 
 	++rgb;
 
-	rgb->r = dal_fixed31_32_mul(rgb_last->r,
+	rgb->r = dc_fixpt_mul(rgb_last->r,
 			dividers.divider2);
-	rgb->g = dal_fixed31_32_mul(rgb_last->g,
+	rgb->g = dc_fixpt_mul(rgb_last->g,
 			dividers.divider2);
-	rgb->b = dal_fixed31_32_mul(rgb_last->b,
+	rgb->b = dc_fixpt_mul(rgb_last->b,
 			dividers.divider2);
 
 	++rgb;
 
-	rgb->r = dal_fixed31_32_mul(rgb_last->r,
+	rgb->r = dc_fixpt_mul(rgb_last->r,
 			dividers.divider3);
-	rgb->g = dal_fixed31_32_mul(rgb_last->g,
+	rgb->g = dc_fixpt_mul(rgb_last->g,
 			dividers.divider3);
-	rgb->b = dal_fixed31_32_mul(rgb_last->b,
+	rgb->b = dc_fixpt_mul(rgb_last->b,
 			dividers.divider3);
 }
 
@@ -759,62 +759,62 @@ static void scale_gamma_dx(struct pwl_float_data *pwl_rgb,
 		struct dividers dividers)
 {
 	uint32_t i;
-	struct fixed31_32 min = dal_fixed31_32_zero;
-	struct fixed31_32 max = dal_fixed31_32_one;
+	struct fixed31_32 min = dc_fixpt_zero;
+	struct fixed31_32 max = dc_fixpt_one;
 
-	struct fixed31_32 delta = dal_fixed31_32_zero;
-	struct fixed31_32 offset = dal_fixed31_32_zero;
+	struct fixed31_32 delta = dc_fixpt_zero;
+	struct fixed31_32 offset = dc_fixpt_zero;
 
 	for (i = 0 ; i < ramp->num_entries; i++) {
-		if (dal_fixed31_32_lt(ramp->entries.red[i], min))
+		if (dc_fixpt_lt(ramp->entries.red[i], min))
 			min = ramp->entries.red[i];
 
-		if (dal_fixed31_32_lt(ramp->entries.green[i], min))
+		if (dc_fixpt_lt(ramp->entries.green[i], min))
 			min = ramp->entries.green[i];
 
-		if (dal_fixed31_32_lt(ramp->entries.blue[i], min))
+		if (dc_fixpt_lt(ramp->entries.blue[i], min))
 			min = ramp->entries.blue[i];
 
-		if (dal_fixed31_32_lt(max, ramp->entries.red[i]))
+		if (dc_fixpt_lt(max, ramp->entries.red[i]))
 			max = ramp->entries.red[i];
 
-		if (dal_fixed31_32_lt(max, ramp->entries.green[i]))
+		if (dc_fixpt_lt(max, ramp->entries.green[i]))
 			max = ramp->entries.green[i];
 
-		if (dal_fixed31_32_lt(max, ramp->entries.blue[i]))
+		if (dc_fixpt_lt(max, ramp->entries.blue[i]))
 			max = ramp->entries.blue[i];
 	}
 
-	if (dal_fixed31_32_lt(min, dal_fixed31_32_zero))
-		delta = dal_fixed31_32_neg(min);
+	if (dc_fixpt_lt(min, dc_fixpt_zero))
+		delta = dc_fixpt_neg(min);
 
-	offset = dal_fixed31_32_add(min, max);
+	offset = dc_fixpt_add(min, max);
 
 	for (i = 0 ; i < ramp->num_entries; i++) {
-		pwl_rgb[i].r = dal_fixed31_32_div(
-			dal_fixed31_32_add(
+		pwl_rgb[i].r = dc_fixpt_div(
+			dc_fixpt_add(
 				ramp->entries.red[i], delta), offset);
-		pwl_rgb[i].g = dal_fixed31_32_div(
-			dal_fixed31_32_add(
+		pwl_rgb[i].g = dc_fixpt_div(
+			dc_fixpt_add(
 				ramp->entries.green[i], delta), offset);
-		pwl_rgb[i].b = dal_fixed31_32_div(
-			dal_fixed31_32_add(
+		pwl_rgb[i].b = dc_fixpt_div(
+			dc_fixpt_add(
 				ramp->entries.blue[i], delta), offset);
 
 	}
 
-	pwl_rgb[i].r =  dal_fixed31_32_sub(dal_fixed31_32_mul_int(
+	pwl_rgb[i].r =  dc_fixpt_sub(dc_fixpt_mul_int(
 				pwl_rgb[i-1].r, 2), pwl_rgb[i-2].r);
-	pwl_rgb[i].g =  dal_fixed31_32_sub(dal_fixed31_32_mul_int(
+	pwl_rgb[i].g =  dc_fixpt_sub(dc_fixpt_mul_int(
 				pwl_rgb[i-1].g, 2), pwl_rgb[i-2].g);
-	pwl_rgb[i].b =  dal_fixed31_32_sub(dal_fixed31_32_mul_int(
+	pwl_rgb[i].b =  dc_fixpt_sub(dc_fixpt_mul_int(
 				pwl_rgb[i-1].b, 2), pwl_rgb[i-2].b);
 	++i;
-	pwl_rgb[i].r =  dal_fixed31_32_sub(dal_fixed31_32_mul_int(
+	pwl_rgb[i].r =  dc_fixpt_sub(dc_fixpt_mul_int(
 				pwl_rgb[i-1].r, 2), pwl_rgb[i-2].r);
-	pwl_rgb[i].g =  dal_fixed31_32_sub(dal_fixed31_32_mul_int(
+	pwl_rgb[i].g =  dc_fixpt_sub(dc_fixpt_mul_int(
 				pwl_rgb[i-1].g, 2), pwl_rgb[i-2].g);
-	pwl_rgb[i].b =  dal_fixed31_32_sub(dal_fixed31_32_mul_int(
+	pwl_rgb[i].b =  dc_fixpt_sub(dc_fixpt_mul_int(
 				pwl_rgb[i-1].b, 2), pwl_rgb[i-2].b);
 }
 
@@ -846,40 +846,40 @@ static void scale_user_regamma_ramp(struct pwl_float_data *pwl_rgb,
 
 	i = 0;
 	do {
-		rgb->r = dal_fixed31_32_from_fraction(
+		rgb->r = dc_fixpt_from_fraction(
 				ramp->gamma[i], scaler);
-		rgb->g = dal_fixed31_32_from_fraction(
+		rgb->g = dc_fixpt_from_fraction(
 				ramp->gamma[i + 256], scaler);
-		rgb->b = dal_fixed31_32_from_fraction(
+		rgb->b = dc_fixpt_from_fraction(
 				ramp->gamma[i + 512], scaler);
 
 		++rgb;
 		++i;
 	} while (i != GAMMA_RGB_256_ENTRIES);
 
-	rgb->r = dal_fixed31_32_mul(rgb_last->r,
+	rgb->r = dc_fixpt_mul(rgb_last->r,
 			dividers.divider1);
-	rgb->g = dal_fixed31_32_mul(rgb_last->g,
+	rgb->g = dc_fixpt_mul(rgb_last->g,
 			dividers.divider1);
-	rgb->b = dal_fixed31_32_mul(rgb_last->b,
+	rgb->b = dc_fixpt_mul(rgb_last->b,
 			dividers.divider1);
 
 	++rgb;
 
-	rgb->r = dal_fixed31_32_mul(rgb_last->r,
+	rgb->r = dc_fixpt_mul(rgb_last->r,
 			dividers.divider2);
-	rgb->g = dal_fixed31_32_mul(rgb_last->g,
+	rgb->g = dc_fixpt_mul(rgb_last->g,
 			dividers.divider2);
-	rgb->b = dal_fixed31_32_mul(rgb_last->b,
+	rgb->b = dc_fixpt_mul(rgb_last->b,
 			dividers.divider2);
 
 	++rgb;
 
-	rgb->r = dal_fixed31_32_mul(rgb_last->r,
+	rgb->r = dc_fixpt_mul(rgb_last->r,
 			dividers.divider3);
-	rgb->g = dal_fixed31_32_mul(rgb_last->g,
+	rgb->g = dc_fixpt_mul(rgb_last->g,
 			dividers.divider3);
-	rgb->b = dal_fixed31_32_mul(rgb_last->b,
+	rgb->b = dc_fixpt_mul(rgb_last->b,
 			dividers.divider3);
 }
 
@@ -913,7 +913,7 @@ static void apply_lut_1d(
 	struct fixed31_32 lut2;
 	const int max_lut_index = 4095;
 	const struct fixed31_32 max_lut_index_f =
-			dal_fixed31_32_from_int_nonconst(max_lut_index);
+			dc_fixpt_from_int_nonconst(max_lut_index);
 	int32_t index = 0, index_next = 0;
 	struct fixed31_32 index_f;
 	struct fixed31_32 delta_lut;
@@ -931,10 +931,10 @@ static void apply_lut_1d(
 			else
 				regamma_y = &tf_pts->blue[i];
 
-			norm_y = dal_fixed31_32_mul(max_lut_index_f,
+			norm_y = dc_fixpt_mul(max_lut_index_f,
 						   *regamma_y);
-			index = dal_fixed31_32_floor(norm_y);
-			index_f = dal_fixed31_32_from_int_nonconst(index);
+			index = dc_fixpt_floor(norm_y);
+			index_f = dc_fixpt_from_int_nonconst(index);
 
 			if (index < 0 || index > max_lut_index)
 				continue;
@@ -953,11 +953,11 @@ static void apply_lut_1d(
 			}
 
 			// we have everything now, so interpolate
-			delta_lut = dal_fixed31_32_sub(lut2, lut1);
-			delta_index = dal_fixed31_32_sub(norm_y, index_f);
+			delta_lut = dc_fixpt_sub(lut2, lut1);
+			delta_index = dc_fixpt_sub(norm_y, index_f);
 
-			*regamma_y = dal_fixed31_32_add(lut1,
-				dal_fixed31_32_mul(delta_index, delta_lut));
+			*regamma_y = dc_fixpt_add(lut1,
+				dc_fixpt_mul(delta_index, delta_lut));
 		}
 	}
 }
@@ -973,7 +973,7 @@ static void build_evenly_distributed_points(
 	uint32_t i = 0;
 
 	do {
-		struct fixed31_32 value = dal_fixed31_32_from_fraction(i,
+		struct fixed31_32 value = dc_fixpt_from_fraction(i,
 			numberof_points - 1);
 
 		p->r = value;
@@ -984,21 +984,21 @@ static void build_evenly_distributed_points(
 		++i;
 	} while (i != numberof_points);
 
-	p->r = dal_fixed31_32_div(p_last->r, dividers.divider1);
-	p->g = dal_fixed31_32_div(p_last->g, dividers.divider1);
-	p->b = dal_fixed31_32_div(p_last->b, dividers.divider1);
+	p->r = dc_fixpt_div(p_last->r, dividers.divider1);
+	p->g = dc_fixpt_div(p_last->g, dividers.divider1);
+	p->b = dc_fixpt_div(p_last->b, dividers.divider1);
 
 	++p;
 
-	p->r = dal_fixed31_32_div(p_last->r, dividers.divider2);
-	p->g = dal_fixed31_32_div(p_last->g, dividers.divider2);
-	p->b = dal_fixed31_32_div(p_last->b, dividers.divider2);
+	p->r = dc_fixpt_div(p_last->r, dividers.divider2);
+	p->g = dc_fixpt_div(p_last->g, dividers.divider2);
+	p->b = dc_fixpt_div(p_last->b, dividers.divider2);
 
 	++p;
 
-	p->r = dal_fixed31_32_div(p_last->r, dividers.divider3);
-	p->g = dal_fixed31_32_div(p_last->g, dividers.divider3);
-	p->b = dal_fixed31_32_div(p_last->b, dividers.divider3);
+	p->r = dc_fixpt_div(p_last->r, dividers.divider3);
+	p->g = dc_fixpt_div(p_last->g, dividers.divider3);
+	p->b = dc_fixpt_div(p_last->b, dividers.divider3);
 }
 
 static inline void copy_rgb_regamma_to_coordinates_x(
@@ -1094,7 +1094,7 @@ static void interpolate_user_regamma(uint32_t hw_points_num,
 	struct fixed31_32 *tf_point;
 	struct fixed31_32 hw_x;
 	struct fixed31_32 norm_factor =
-			dal_fixed31_32_from_int_nonconst(255);
+			dc_fixpt_from_int_nonconst(255);
 	struct fixed31_32 norm_x;
 	struct fixed31_32 index_f;
 	struct fixed31_32 lut1;
@@ -1105,9 +1105,9 @@ static void interpolate_user_regamma(uint32_t hw_points_num,
 	i = 0;
 	/* fixed_pt library has problems handling too small values */
 	while (i != 32) {
-		tf_pts->red[i] = dal_fixed31_32_zero;
-		tf_pts->green[i] = dal_fixed31_32_zero;
-		tf_pts->blue[i] = dal_fixed31_32_zero;
+		tf_pts->red[i] = dc_fixpt_zero;
+		tf_pts->green[i] = dc_fixpt_zero;
+		tf_pts->blue[i] = dc_fixpt_zero;
 		++i;
 	}
 	while (i <= hw_points_num + 1) {
@@ -1129,12 +1129,12 @@ static void interpolate_user_regamma(uint32_t hw_points_num,
 			} else
 				hw_x = coordinates_x[i].x;
 
-			norm_x = dal_fixed31_32_mul(norm_factor, hw_x);
-			index = dal_fixed31_32_floor(norm_x);
+			norm_x = dc_fixpt_mul(norm_factor, hw_x);
+			index = dc_fixpt_floor(norm_x);
 			if (index < 0 || index > 255)
 				continue;
 
-			index_f = dal_fixed31_32_from_int_nonconst(index);
+			index_f = dc_fixpt_from_int_nonconst(index);
 			index_next = (index == 255) ? index : index + 1;
 
 			if (color == 0) {
@@ -1149,11 +1149,11 @@ static void interpolate_user_regamma(uint32_t hw_points_num,
 			}
 
 			// we have everything now, so interpolate
-			delta_lut = dal_fixed31_32_sub(lut2, lut1);
-			delta_index = dal_fixed31_32_sub(norm_x, index_f);
+			delta_lut = dc_fixpt_sub(lut2, lut1);
+			delta_index = dc_fixpt_sub(norm_x, index_f);
 
-			*tf_point = dal_fixed31_32_add(lut1,
-				dal_fixed31_32_mul(delta_index, delta_lut));
+			*tf_point = dc_fixpt_add(lut1,
+				dc_fixpt_mul(delta_index, delta_lut));
 		}
 		++i;
 	}
@@ -1168,15 +1168,15 @@ static void build_new_custom_resulted_curve(
 	i = 0;
 
 	while (i != hw_points_num + 1) {
-		tf_pts->red[i] = dal_fixed31_32_clamp(
-			tf_pts->red[i], dal_fixed31_32_zero,
-			dal_fixed31_32_one);
-		tf_pts->green[i] = dal_fixed31_32_clamp(
-			tf_pts->green[i], dal_fixed31_32_zero,
-			dal_fixed31_32_one);
-		tf_pts->blue[i] = dal_fixed31_32_clamp(
-			tf_pts->blue[i], dal_fixed31_32_zero,
-			dal_fixed31_32_one);
+		tf_pts->red[i] = dc_fixpt_clamp(
+			tf_pts->red[i], dc_fixpt_zero,
+			dc_fixpt_one);
+		tf_pts->green[i] = dc_fixpt_clamp(
+			tf_pts->green[i], dc_fixpt_zero,
+			dc_fixpt_one);
+		tf_pts->blue[i] = dc_fixpt_clamp(
+			tf_pts->blue[i], dc_fixpt_zero,
+			dc_fixpt_one);
 
 		++i;
 	}
@@ -1290,9 +1290,9 @@ bool mod_color_calculate_regamma_params(struct dc_transfer_func *output_tf,
 	if (!coeff)
 		goto coeff_alloc_fail;
 
-	dividers.divider1 = dal_fixed31_32_from_fraction(3, 2);
-	dividers.divider2 = dal_fixed31_32_from_int(2);
-	dividers.divider3 = dal_fixed31_32_from_fraction(5, 2);
+	dividers.divider1 = dc_fixpt_from_fraction(3, 2);
+	dividers.divider2 = dc_fixpt_from_int(2);
+	dividers.divider3 = dc_fixpt_from_fraction(5, 2);
 
 	tf = output_tf->tf;
 
@@ -1357,15 +1357,15 @@ bool calculate_user_regamma_coeff(struct dc_transfer_func *output_tf,
 	uint32_t i = 0;
 
 	do {
-		coeff.a0[i] = dal_fixed31_32_from_fraction(
+		coeff.a0[i] = dc_fixpt_from_fraction(
 				regamma->coeff.A0[i], 10000000);
-		coeff.a1[i] = dal_fixed31_32_from_fraction(
+		coeff.a1[i] = dc_fixpt_from_fraction(
 				regamma->coeff.A1[i], 1000);
-		coeff.a2[i] = dal_fixed31_32_from_fraction(
+		coeff.a2[i] = dc_fixpt_from_fraction(
 				regamma->coeff.A2[i], 1000);
-		coeff.a3[i] = dal_fixed31_32_from_fraction(
+		coeff.a3[i] = dc_fixpt_from_fraction(
 				regamma->coeff.A3[i], 1000);
-		coeff.user_gamma[i] = dal_fixed31_32_from_fraction(
+		coeff.user_gamma[i] = dc_fixpt_from_fraction(
 				regamma->coeff.gamma[i], 1000);
 
 		++i;
@@ -1374,9 +1374,9 @@ bool calculate_user_regamma_coeff(struct dc_transfer_func *output_tf,
 	i = 0;
 	/* fixed_pt library has problems handling too small values */
 	while (i != 32) {
-		output_tf->tf_pts.red[i] = dal_fixed31_32_zero;
-		output_tf->tf_pts.green[i] = dal_fixed31_32_zero;
-		output_tf->tf_pts.blue[i] = dal_fixed31_32_zero;
+		output_tf->tf_pts.red[i] = dc_fixpt_zero;
+		output_tf->tf_pts.green[i] = dc_fixpt_zero;
+		output_tf->tf_pts.blue[i] = dc_fixpt_zero;
 		++coord_x;
 		++i;
 	}
@@ -1423,9 +1423,9 @@ bool calculate_user_regamma_ramp(struct dc_transfer_func *output_tf,
 	if (!rgb_regamma)
 		goto rgb_regamma_alloc_fail;
 
-	dividers.divider1 = dal_fixed31_32_from_fraction(3, 2);
-	dividers.divider2 = dal_fixed31_32_from_int(2);
-	dividers.divider3 = dal_fixed31_32_from_fraction(5, 2);
+	dividers.divider1 = dc_fixpt_from_fraction(3, 2);
+	dividers.divider2 = dc_fixpt_from_int(2);
+	dividers.divider3 = dc_fixpt_from_fraction(5, 2);
 
 	scale_user_regamma_ramp(rgb_user, &regamma->ramp, dividers);
 
@@ -1496,9 +1496,9 @@ bool mod_color_calculate_degamma_params(struct dc_transfer_func *input_tf,
 	if (!coeff)
 		goto coeff_alloc_fail;
 
-	dividers.divider1 = dal_fixed31_32_from_fraction(3, 2);
-	dividers.divider2 = dal_fixed31_32_from_int(2);
-	dividers.divider3 = dal_fixed31_32_from_fraction(5, 2);
+	dividers.divider1 = dc_fixpt_from_fraction(3, 2);
+	dividers.divider2 = dc_fixpt_from_int(2);
+	dividers.divider3 = dc_fixpt_from_fraction(5, 2);
 
 	tf = input_tf->tf;
 
-- 
cgit v1.1


From f3ba7a2fd1ebffe7fc6a9c524754db05dcd0c0e4 Mon Sep 17 00:00:00 2001
From: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Date: Wed, 18 Apr 2018 13:54:24 -0400
Subject: drm/amd/display: inline more of fixed point code

Signed-off-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c | 156 ++--------------
 drivers/gpu/drm/amd/display/include/fixed31_32.h   | 207 ++++++++++++---------
 .../drm/amd/display/modules/color/color_gamma.c    |   8 +-
 3 files changed, 135 insertions(+), 236 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c b/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c
index e398ecd..e61dd97d 100644
--- a/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c
+++ b/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c
@@ -64,9 +64,7 @@ static inline unsigned long long complete_integer_division_u64(
 #define GET_FRACTIONAL_PART(x) \
 	(FRACTIONAL_PART_MASK & (x))
 
-struct fixed31_32 dc_fixpt_from_fraction(
-	long long numerator,
-	long long denominator)
+struct fixed31_32 dc_fixpt_from_fraction(long long numerator, long long denominator)
 {
 	struct fixed31_32 res;
 
@@ -118,63 +116,7 @@ struct fixed31_32 dc_fixpt_from_fraction(
 	return res;
 }
 
-struct fixed31_32 dc_fixpt_from_int_nonconst(
-	long long arg)
-{
-	struct fixed31_32 res;
-
-	ASSERT((LONG_MIN <= arg) && (arg <= LONG_MAX));
-
-	res.value = arg << FIXED31_32_BITS_PER_FRACTIONAL_PART;
-
-	return res;
-}
-
-struct fixed31_32 dc_fixpt_shl(
-	struct fixed31_32 arg,
-	unsigned char shift)
-{
-	struct fixed31_32 res;
-
-	ASSERT(((arg.value >= 0) && (arg.value <= LLONG_MAX >> shift)) ||
-		((arg.value < 0) && (arg.value >= LLONG_MIN >> shift)));
-
-	res.value = arg.value << shift;
-
-	return res;
-}
-
-struct fixed31_32 dc_fixpt_add(
-	struct fixed31_32 arg1,
-	struct fixed31_32 arg2)
-{
-	struct fixed31_32 res;
-
-	ASSERT(((arg1.value >= 0) && (LLONG_MAX - arg1.value >= arg2.value)) ||
-		((arg1.value < 0) && (LLONG_MIN - arg1.value <= arg2.value)));
-
-	res.value = arg1.value + arg2.value;
-
-	return res;
-}
-
-struct fixed31_32 dc_fixpt_sub(
-	struct fixed31_32 arg1,
-	struct fixed31_32 arg2)
-{
-	struct fixed31_32 res;
-
-	ASSERT(((arg2.value >= 0) && (LLONG_MIN + arg2.value <= arg1.value)) ||
-		((arg2.value < 0) && (LLONG_MAX + arg2.value >= arg1.value)));
-
-	res.value = arg1.value - arg2.value;
-
-	return res;
-}
-
-struct fixed31_32 dc_fixpt_mul(
-	struct fixed31_32 arg1,
-	struct fixed31_32 arg2)
+struct fixed31_32 dc_fixpt_mul(struct fixed31_32 arg1, struct fixed31_32 arg2)
 {
 	struct fixed31_32 res;
 
@@ -225,8 +167,7 @@ struct fixed31_32 dc_fixpt_mul(
 	return res;
 }
 
-struct fixed31_32 dc_fixpt_sqr(
-	struct fixed31_32 arg)
+struct fixed31_32 dc_fixpt_sqr(struct fixed31_32 arg)
 {
 	struct fixed31_32 res;
 
@@ -266,8 +207,7 @@ struct fixed31_32 dc_fixpt_sqr(
 	return res;
 }
 
-struct fixed31_32 dc_fixpt_recip(
-	struct fixed31_32 arg)
+struct fixed31_32 dc_fixpt_recip(struct fixed31_32 arg)
 {
 	/*
 	 * @note
@@ -281,8 +221,7 @@ struct fixed31_32 dc_fixpt_recip(
 		arg.value);
 }
 
-struct fixed31_32 dc_fixpt_sinc(
-	struct fixed31_32 arg)
+struct fixed31_32 dc_fixpt_sinc(struct fixed31_32 arg)
 {
 	struct fixed31_32 square;
 
@@ -326,16 +265,14 @@ struct fixed31_32 dc_fixpt_sinc(
 	return res;
 }
 
-struct fixed31_32 dc_fixpt_sin(
-	struct fixed31_32 arg)
+struct fixed31_32 dc_fixpt_sin(struct fixed31_32 arg)
 {
 	return dc_fixpt_mul(
 		arg,
 		dc_fixpt_sinc(arg));
 }
 
-struct fixed31_32 dc_fixpt_cos(
-	struct fixed31_32 arg)
+struct fixed31_32 dc_fixpt_cos(struct fixed31_32 arg)
 {
 	/* TODO implement argument normalization */
 
@@ -367,8 +304,7 @@ struct fixed31_32 dc_fixpt_cos(
  *
  * Calculated as Taylor series.
  */
-static struct fixed31_32 fixed31_32_exp_from_taylor_series(
-	struct fixed31_32 arg)
+static struct fixed31_32 fixed31_32_exp_from_taylor_series(struct fixed31_32 arg)
 {
 	unsigned int n = 9;
 
@@ -396,8 +332,7 @@ static struct fixed31_32 fixed31_32_exp_from_taylor_series(
 			res));
 }
 
-struct fixed31_32 dc_fixpt_exp(
-	struct fixed31_32 arg)
+struct fixed31_32 dc_fixpt_exp(struct fixed31_32 arg)
 {
 	/*
 	 * @brief
@@ -440,8 +375,7 @@ struct fixed31_32 dc_fixpt_exp(
 		return dc_fixpt_one;
 }
 
-struct fixed31_32 dc_fixpt_log(
-	struct fixed31_32 arg)
+struct fixed31_32 dc_fixpt_log(struct fixed31_32 arg)
 {
 	struct fixed31_32 res = dc_fixpt_neg(dc_fixpt_one);
 	/* TODO improve 1st estimation */
@@ -472,61 +406,6 @@ struct fixed31_32 dc_fixpt_log(
 	return res;
 }
 
-struct fixed31_32 dc_fixpt_pow(
-	struct fixed31_32 arg1,
-	struct fixed31_32 arg2)
-{
-	return dc_fixpt_exp(
-		dc_fixpt_mul(
-			dc_fixpt_log(arg1),
-			arg2));
-}
-
-int dc_fixpt_floor(
-	struct fixed31_32 arg)
-{
-	unsigned long long arg_value = abs_i64(arg.value);
-
-	if (arg.value >= 0)
-		return (int)GET_INTEGER_PART(arg_value);
-	else
-		return -(int)GET_INTEGER_PART(arg_value);
-}
-
-int dc_fixpt_round(
-	struct fixed31_32 arg)
-{
-	unsigned long long arg_value = abs_i64(arg.value);
-
-	const long long summand = dc_fixpt_half.value;
-
-	ASSERT(LLONG_MAX - (long long)arg_value >= summand);
-
-	arg_value += summand;
-
-	if (arg.value >= 0)
-		return (int)GET_INTEGER_PART(arg_value);
-	else
-		return -(int)GET_INTEGER_PART(arg_value);
-}
-
-int dc_fixpt_ceil(
-	struct fixed31_32 arg)
-{
-	unsigned long long arg_value = abs_i64(arg.value);
-
-	const long long summand = dc_fixpt_one.value -
-		dc_fixpt_epsilon.value;
-
-	ASSERT(LLONG_MAX - (long long)arg_value >= summand);
-
-	arg_value += summand;
-
-	if (arg.value >= 0)
-		return (int)GET_INTEGER_PART(arg_value);
-	else
-		return -(int)GET_INTEGER_PART(arg_value);
-}
 
 /* this function is a generic helper to translate fixed point value to
  * specified integer format that will consist of integer_bits integer part and
@@ -570,32 +449,27 @@ static inline unsigned int clamp_ux_dy(
 		return min_clamp;
 }
 
-unsigned int dc_fixpt_u2d19(
-	struct fixed31_32 arg)
+unsigned int dc_fixpt_u2d19(struct fixed31_32 arg)
 {
 	return ux_dy(arg.value, 2, 19);
 }
 
-unsigned int dc_fixpt_u0d19(
-	struct fixed31_32 arg)
+unsigned int dc_fixpt_u0d19(struct fixed31_32 arg)
 {
 	return ux_dy(arg.value, 0, 19);
 }
 
-unsigned int dc_fixpt_clamp_u0d14(
-	struct fixed31_32 arg)
+unsigned int dc_fixpt_clamp_u0d14(struct fixed31_32 arg)
 {
 	return clamp_ux_dy(arg.value, 0, 14, 1);
 }
 
-unsigned int dc_fixpt_clamp_u0d10(
-	struct fixed31_32 arg)
+unsigned int dc_fixpt_clamp_u0d10(struct fixed31_32 arg)
 {
 	return clamp_ux_dy(arg.value, 0, 10, 1);
 }
 
-int dc_fixpt_s4d19(
-	struct fixed31_32 arg)
+int dc_fixpt_s4d19(struct fixed31_32 arg)
 {
 	if (arg.value < 0)
 		return -(int)ux_dy(dc_fixpt_abs(arg).value, 4, 19);
diff --git a/drivers/gpu/drm/amd/display/include/fixed31_32.h b/drivers/gpu/drm/amd/display/include/fixed31_32.h
index b5b8d7d..ebfd33e 100644
--- a/drivers/gpu/drm/amd/display/include/fixed31_32.h
+++ b/drivers/gpu/drm/amd/display/include/fixed31_32.h
@@ -70,24 +70,19 @@ static const struct fixed31_32 dc_fixpt_ln2_div_2 = { 1488522236LL };
  * @brief
  * result = numerator / denominator
  */
-struct fixed31_32 dc_fixpt_from_fraction(
-	long long numerator,
-	long long denominator);
+struct fixed31_32 dc_fixpt_from_fraction(long long numerator, long long denominator);
 
 /*
  * @brief
  * result = arg
  */
-struct fixed31_32 dc_fixpt_from_int_nonconst(long long arg);
-static inline struct fixed31_32 dc_fixpt_from_int(long long arg)
+static inline struct fixed31_32 dc_fixpt_from_int(int arg)
 {
-	if (__builtin_constant_p(arg)) {
-		struct fixed31_32 res;
-		BUILD_BUG_ON((LONG_MIN > arg) || (arg > LONG_MAX));
-		res.value = arg << FIXED31_32_BITS_PER_FRACTIONAL_PART;
-		return res;
-	} else
-		return dc_fixpt_from_int_nonconst(arg);
+	struct fixed31_32 res;
+
+	res.value = (long long) arg << FIXED31_32_BITS_PER_FRACTIONAL_PART;
+
+	return res;
 }
 
 /*
@@ -129,8 +124,7 @@ static inline struct fixed31_32 dc_fixpt_abs(struct fixed31_32 arg)
  * @brief
  * result = arg1 < arg2
  */
-static inline bool dc_fixpt_lt(struct fixed31_32 arg1,
-				     struct fixed31_32 arg2)
+static inline bool dc_fixpt_lt(struct fixed31_32 arg1, struct fixed31_32 arg2)
 {
 	return arg1.value < arg2.value;
 }
@@ -139,8 +133,7 @@ static inline bool dc_fixpt_lt(struct fixed31_32 arg1,
  * @brief
  * result = arg1 <= arg2
  */
-static inline bool dc_fixpt_le(struct fixed31_32 arg1,
-				     struct fixed31_32 arg2)
+static inline bool dc_fixpt_le(struct fixed31_32 arg1, struct fixed31_32 arg2)
 {
 	return arg1.value <= arg2.value;
 }
@@ -149,8 +142,7 @@ static inline bool dc_fixpt_le(struct fixed31_32 arg1,
  * @brief
  * result = arg1 == arg2
  */
-static inline bool dc_fixpt_eq(struct fixed31_32 arg1,
-				     struct fixed31_32 arg2)
+static inline bool dc_fixpt_eq(struct fixed31_32 arg1, struct fixed31_32 arg2)
 {
 	return arg1.value == arg2.value;
 }
@@ -159,8 +151,7 @@ static inline bool dc_fixpt_eq(struct fixed31_32 arg1,
  * @brief
  * result = min(arg1, arg2) := (arg1 <= arg2) ? arg1 : arg2
  */
-static inline struct fixed31_32 dc_fixpt_min(struct fixed31_32 arg1,
-						   struct fixed31_32 arg2)
+static inline struct fixed31_32 dc_fixpt_min(struct fixed31_32 arg1, struct fixed31_32 arg2)
 {
 	if (arg1.value <= arg2.value)
 		return arg1;
@@ -172,8 +163,7 @@ static inline struct fixed31_32 dc_fixpt_min(struct fixed31_32 arg1,
  * @brief
  * result = max(arg1, arg2) := (arg1 <= arg2) ? arg2 : arg1
  */
-static inline struct fixed31_32 dc_fixpt_max(struct fixed31_32 arg1,
-						   struct fixed31_32 arg2)
+static inline struct fixed31_32 dc_fixpt_max(struct fixed31_32 arg1, struct fixed31_32 arg2)
 {
 	if (arg1.value <= arg2.value)
 		return arg2;
@@ -209,17 +199,23 @@ static inline struct fixed31_32 dc_fixpt_clamp(
  * @brief
  * result = arg << shift
  */
-struct fixed31_32 dc_fixpt_shl(
-	struct fixed31_32 arg,
-	unsigned char shift);
+static inline struct fixed31_32 dc_fixpt_shl(struct fixed31_32 arg, unsigned char shift)
+{
+	struct fixed31_32 res;
+
+	ASSERT(((arg.value >= 0) && (arg.value <= LLONG_MAX >> shift)) ||
+		((arg.value < 0) && (arg.value >= LLONG_MIN >> shift)));
+
+	res.value = arg.value << shift;
+
+	return res;
+}
 
 /*
  * @brief
  * result = arg >> shift
  */
-static inline struct fixed31_32 dc_fixpt_shr(
-	struct fixed31_32 arg,
-	unsigned char shift)
+static inline struct fixed31_32 dc_fixpt_shr(struct fixed31_32 arg, unsigned char shift)
 {
 	struct fixed31_32 res;
 	res.value = arg.value >> shift;
@@ -235,38 +231,50 @@ static inline struct fixed31_32 dc_fixpt_shr(
  * @brief
  * result = arg1 + arg2
  */
-struct fixed31_32 dc_fixpt_add(
-	struct fixed31_32 arg1,
-	struct fixed31_32 arg2);
+static inline struct fixed31_32 dc_fixpt_add(struct fixed31_32 arg1, struct fixed31_32 arg2)
+{
+	struct fixed31_32 res;
+
+	ASSERT(((arg1.value >= 0) && (LLONG_MAX - arg1.value >= arg2.value)) ||
+		((arg1.value < 0) && (LLONG_MIN - arg1.value <= arg2.value)));
+
+	res.value = arg1.value + arg2.value;
+
+	return res;
+}
 
 /*
  * @brief
  * result = arg1 + arg2
  */
-static inline struct fixed31_32 dc_fixpt_add_int(struct fixed31_32 arg1,
-						       int arg2)
+static inline struct fixed31_32 dc_fixpt_add_int(struct fixed31_32 arg1, int arg2)
 {
-	return dc_fixpt_add(arg1,
-				  dc_fixpt_from_int(arg2));
+	return dc_fixpt_add(arg1, dc_fixpt_from_int(arg2));
 }
 
 /*
  * @brief
  * result = arg1 - arg2
  */
-struct fixed31_32 dc_fixpt_sub(
-	struct fixed31_32 arg1,
-	struct fixed31_32 arg2);
+static inline struct fixed31_32 dc_fixpt_sub(struct fixed31_32 arg1, struct fixed31_32 arg2)
+{
+	struct fixed31_32 res;
+
+	ASSERT(((arg2.value >= 0) && (LLONG_MIN + arg2.value <= arg1.value)) ||
+		((arg2.value < 0) && (LLONG_MAX + arg2.value >= arg1.value)));
+
+	res.value = arg1.value - arg2.value;
+
+	return res;
+}
 
 /*
  * @brief
  * result = arg1 - arg2
  */
-static inline struct fixed31_32 dc_fixpt_sub_int(struct fixed31_32 arg1,
-						       int arg2)
+static inline struct fixed31_32 dc_fixpt_sub_int(struct fixed31_32 arg1, int arg2)
 {
-	return dc_fixpt_sub(arg1,
-				  dc_fixpt_from_int(arg2));
+	return dc_fixpt_sub(arg1, dc_fixpt_from_int(arg2));
 }
 
 
@@ -279,49 +287,40 @@ static inline struct fixed31_32 dc_fixpt_sub_int(struct fixed31_32 arg1,
  * @brief
  * result = arg1 * arg2
  */
-struct fixed31_32 dc_fixpt_mul(
-	struct fixed31_32 arg1,
-	struct fixed31_32 arg2);
+struct fixed31_32 dc_fixpt_mul(struct fixed31_32 arg1, struct fixed31_32 arg2);
 
 
 /*
  * @brief
  * result = arg1 * arg2
  */
-static inline struct fixed31_32 dc_fixpt_mul_int(struct fixed31_32 arg1,
-						       int arg2)
+static inline struct fixed31_32 dc_fixpt_mul_int(struct fixed31_32 arg1, int arg2)
 {
-	return dc_fixpt_mul(arg1,
-				  dc_fixpt_from_int(arg2));
+	return dc_fixpt_mul(arg1, dc_fixpt_from_int(arg2));
 }
 
 /*
  * @brief
  * result = square(arg) := arg * arg
  */
-struct fixed31_32 dc_fixpt_sqr(
-	struct fixed31_32 arg);
+struct fixed31_32 dc_fixpt_sqr(struct fixed31_32 arg);
 
 /*
  * @brief
  * result = arg1 / arg2
  */
-static inline struct fixed31_32 dc_fixpt_div_int(struct fixed31_32 arg1,
-						       long long arg2)
+static inline struct fixed31_32 dc_fixpt_div_int(struct fixed31_32 arg1, long long arg2)
 {
-	return dc_fixpt_from_fraction(arg1.value,
-					    dc_fixpt_from_int(arg2).value);
+	return dc_fixpt_from_fraction(arg1.value, dc_fixpt_from_int(arg2).value);
 }
 
 /*
  * @brief
  * result = arg1 / arg2
  */
-static inline struct fixed31_32 dc_fixpt_div(struct fixed31_32 arg1,
-						   struct fixed31_32 arg2)
+static inline struct fixed31_32 dc_fixpt_div(struct fixed31_32 arg1, struct fixed31_32 arg2)
 {
-	return dc_fixpt_from_fraction(arg1.value,
-					    arg2.value);
+	return dc_fixpt_from_fraction(arg1.value, arg2.value);
 }
 
 /*
@@ -336,8 +335,7 @@ static inline struct fixed31_32 dc_fixpt_div(struct fixed31_32 arg1,
  * @note
  * No special actions taken in case argument is zero.
  */
-struct fixed31_32 dc_fixpt_recip(
-	struct fixed31_32 arg);
+struct fixed31_32 dc_fixpt_recip(struct fixed31_32 arg);
 
 /*
  * @brief
@@ -352,8 +350,7 @@ struct fixed31_32 dc_fixpt_recip(
  * Argument specified in radians,
  * internally it's normalized to [-2pi...2pi] range.
  */
-struct fixed31_32 dc_fixpt_sinc(
-	struct fixed31_32 arg);
+struct fixed31_32 dc_fixpt_sinc(struct fixed31_32 arg);
 
 /*
  * @brief
@@ -363,8 +360,7 @@ struct fixed31_32 dc_fixpt_sinc(
  * Argument specified in radians,
  * internally it's normalized to [-2pi...2pi] range.
  */
-struct fixed31_32 dc_fixpt_sin(
-	struct fixed31_32 arg);
+struct fixed31_32 dc_fixpt_sin(struct fixed31_32 arg);
 
 /*
  * @brief
@@ -376,8 +372,7 @@ struct fixed31_32 dc_fixpt_sin(
  * passing arguments outside that range
  * will cause incorrect result!
  */
-struct fixed31_32 dc_fixpt_cos(
-	struct fixed31_32 arg);
+struct fixed31_32 dc_fixpt_cos(struct fixed31_32 arg);
 
 /*
  * @brief
@@ -391,8 +386,7 @@ struct fixed31_32 dc_fixpt_cos(
  * @note
  * Currently, function is verified for abs(arg) <= 1.
  */
-struct fixed31_32 dc_fixpt_exp(
-	struct fixed31_32 arg);
+struct fixed31_32 dc_fixpt_exp(struct fixed31_32 arg);
 
 /*
  * @brief
@@ -404,8 +398,7 @@ struct fixed31_32 dc_fixpt_exp(
  * Currently, no special actions taken
  * in case of invalid argument(s). Take care!
  */
-struct fixed31_32 dc_fixpt_log(
-	struct fixed31_32 arg);
+struct fixed31_32 dc_fixpt_log(struct fixed31_32 arg);
 
 /*
  * @brief
@@ -419,9 +412,13 @@ struct fixed31_32 dc_fixpt_log(
  * @note
  * Currently, abs(arg1) should be less than 1. Take care!
  */
-struct fixed31_32 dc_fixpt_pow(
-	struct fixed31_32 arg1,
-	struct fixed31_32 arg2);
+static inline struct fixed31_32 dc_fixpt_pow(struct fixed31_32 arg1, struct fixed31_32 arg2)
+{
+	return dc_fixpt_exp(
+		dc_fixpt_mul(
+			dc_fixpt_log(arg1),
+			arg2));
+}
 
 /*
  * @brief
@@ -432,22 +429,56 @@ struct fixed31_32 dc_fixpt_pow(
  * @brief
  * result = floor(arg) := greatest integer lower than or equal to arg
  */
-int dc_fixpt_floor(
-	struct fixed31_32 arg);
+static inline int dc_fixpt_floor(struct fixed31_32 arg)
+{
+	unsigned long long arg_value = arg.value > 0 ? arg.value : -arg.value;
+
+	if (arg.value >= 0)
+		return (int)(arg_value >> FIXED31_32_BITS_PER_FRACTIONAL_PART);
+	else
+		return -(int)(arg_value >> FIXED31_32_BITS_PER_FRACTIONAL_PART);
+}
 
 /*
  * @brief
  * result = round(arg) := integer nearest to arg
  */
-int dc_fixpt_round(
-	struct fixed31_32 arg);
+static inline int dc_fixpt_round(struct fixed31_32 arg)
+{
+	unsigned long long arg_value = arg.value > 0 ? arg.value : -arg.value;
+
+	const long long summand = dc_fixpt_half.value;
+
+	ASSERT(LLONG_MAX - (long long)arg_value >= summand);
+
+	arg_value += summand;
+
+	if (arg.value >= 0)
+		return (int)(arg_value >> FIXED31_32_BITS_PER_FRACTIONAL_PART);
+	else
+		return -(int)(arg_value >> FIXED31_32_BITS_PER_FRACTIONAL_PART);
+}
 
 /*
  * @brief
  * result = ceil(arg) := lowest integer greater than or equal to arg
  */
-int dc_fixpt_ceil(
-	struct fixed31_32 arg);
+static inline int dc_fixpt_ceil(struct fixed31_32 arg)
+{
+	unsigned long long arg_value = arg.value > 0 ? arg.value : -arg.value;
+
+	const long long summand = dc_fixpt_one.value -
+		dc_fixpt_epsilon.value;
+
+	ASSERT(LLONG_MAX - (long long)arg_value >= summand);
+
+	arg_value += summand;
+
+	if (arg.value >= 0)
+		return (int)(arg_value >> FIXED31_32_BITS_PER_FRACTIONAL_PART);
+	else
+		return -(int)(arg_value >> FIXED31_32_BITS_PER_FRACTIONAL_PART);
+}
 
 /* the following two function are used in scaler hw programming to convert fixed
  * point value to format 2 bits from integer part and 19 bits from fractional
@@ -455,20 +486,14 @@ int dc_fixpt_ceil(
  * fractional
  */
 
-unsigned int dc_fixpt_u2d19(
-	struct fixed31_32 arg);
-
-unsigned int dc_fixpt_u0d19(
-	struct fixed31_32 arg);
+unsigned int dc_fixpt_u2d19(struct fixed31_32 arg);
 
+unsigned int dc_fixpt_u0d19(struct fixed31_32 arg);
 
-unsigned int dc_fixpt_clamp_u0d14(
-	struct fixed31_32 arg);
+unsigned int dc_fixpt_clamp_u0d14(struct fixed31_32 arg);
 
-unsigned int dc_fixpt_clamp_u0d10(
-	struct fixed31_32 arg);
+unsigned int dc_fixpt_clamp_u0d10(struct fixed31_32 arg);
 
-int dc_fixpt_s4d19(
-	struct fixed31_32 arg);
+int dc_fixpt_s4d19(struct fixed31_32 arg);
 
 #endif
diff --git a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c
index 29d2ec8..e803b37 100644
--- a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c
+++ b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c
@@ -913,7 +913,7 @@ static void apply_lut_1d(
 	struct fixed31_32 lut2;
 	const int max_lut_index = 4095;
 	const struct fixed31_32 max_lut_index_f =
-			dc_fixpt_from_int_nonconst(max_lut_index);
+			dc_fixpt_from_int(max_lut_index);
 	int32_t index = 0, index_next = 0;
 	struct fixed31_32 index_f;
 	struct fixed31_32 delta_lut;
@@ -934,7 +934,7 @@ static void apply_lut_1d(
 			norm_y = dc_fixpt_mul(max_lut_index_f,
 						   *regamma_y);
 			index = dc_fixpt_floor(norm_y);
-			index_f = dc_fixpt_from_int_nonconst(index);
+			index_f = dc_fixpt_from_int(index);
 
 			if (index < 0 || index > max_lut_index)
 				continue;
@@ -1094,7 +1094,7 @@ static void interpolate_user_regamma(uint32_t hw_points_num,
 	struct fixed31_32 *tf_point;
 	struct fixed31_32 hw_x;
 	struct fixed31_32 norm_factor =
-			dc_fixpt_from_int_nonconst(255);
+			dc_fixpt_from_int(255);
 	struct fixed31_32 norm_x;
 	struct fixed31_32 index_f;
 	struct fixed31_32 lut1;
@@ -1134,7 +1134,7 @@ static void interpolate_user_regamma(uint32_t hw_points_num,
 			if (index < 0 || index > 255)
 				continue;
 
-			index_f = dc_fixpt_from_int_nonconst(index);
+			index_f = dc_fixpt_from_int(index);
 			index_next = (index == 255) ? index : index + 1;
 
 			if (color == 0) {
-- 
cgit v1.1


From e8838df1cb987fe690dfd069824ff08107327607 Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Thu, 19 Apr 2018 10:05:22 -0400
Subject: drm/amd/display: Make DisplayStats work with just DC DisplayStats
 minor

Remove dependency on the old FREESYNC_SW_STATS log mask used by DAL2
Also rename from profiling to displaystats

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/basics/logger.c     |  2 +-
 drivers/gpu/drm/amd/display/include/logger_types.h |  2 +-
 drivers/gpu/drm/amd/display/modules/stats/stats.c  | 81 ++++++++++++----------
 3 files changed, 46 insertions(+), 39 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/basics/logger.c b/drivers/gpu/drm/amd/display/dc/basics/logger.c
index 31bee05..0001a3c 100644
--- a/drivers/gpu/drm/amd/display/dc/basics/logger.c
+++ b/drivers/gpu/drm/amd/display/dc/basics/logger.c
@@ -61,7 +61,7 @@ static const struct dc_log_type_info log_type_info_tbl[] = {
 		{LOG_EVENT_UNDERFLOW,       "Underflow"},
 		{LOG_IF_TRACE,              "InterfaceTrace"},
 		{LOG_DTN,                   "DTN"},
-		{LOG_PROFILING,             "Profiling"}
+		{LOG_DISPLAYSTATS,          "DisplayStats"}
 };
 
 
diff --git a/drivers/gpu/drm/amd/display/include/logger_types.h b/drivers/gpu/drm/amd/display/include/logger_types.h
index b608a08..0a540b9 100644
--- a/drivers/gpu/drm/amd/display/include/logger_types.h
+++ b/drivers/gpu/drm/amd/display/include/logger_types.h
@@ -98,7 +98,7 @@ enum dc_log_type {
 	LOG_EVENT_UNDERFLOW,
 	LOG_IF_TRACE,
 	LOG_PERF_TRACE,
-	LOG_PROFILING,
+	LOG_DISPLAYSTATS,
 
 	LOG_SECTION_TOTAL_COUNT
 };
diff --git a/drivers/gpu/drm/amd/display/modules/stats/stats.c b/drivers/gpu/drm/amd/display/modules/stats/stats.c
index 48e0219..d16aac7 100644
--- a/drivers/gpu/drm/amd/display/modules/stats/stats.c
+++ b/drivers/gpu/drm/amd/display/modules/stats/stats.c
@@ -177,44 +177,51 @@ void mod_stats_dump(struct mod_stats *mod_stats)
 	logger = dc->ctx->logger;
 	time = core_stats->time;
 
-	//LogEntry* pLog = GetLog()->Open(LogMajor_ISR, LogMinor_ISR_FreeSyncSW);
-
-	//if (!pLog->IsDummyEntry())
-	{
-		dm_logger_write(logger, LOG_PROFILING, "==Display Caps==\n");
-		dm_logger_write(logger, LOG_PROFILING, "\n");
-		dm_logger_write(logger, LOG_PROFILING, "\n");
-
-		dm_logger_write(logger, LOG_PROFILING, "==Stats==\n");
-		dm_logger_write(logger, LOG_PROFILING,
-			"render avgRender minWindow midPoint maxWindow vsyncToFlip flipToVsync #vsyncBetweenFlip #frame insertDuration vTotalMin vTotalMax eventTrigs vSyncTime1 vSyncTime2 vSyncTime3 vSyncTime4 vSyncTime5 flags\n");
-
-		for (int i = 0; i < core_stats->index && i < core_stats->entries; i++) {
-			dm_logger_write(logger, LOG_PROFILING,
-					"%u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u",
-					time[i].render_time_in_us,
-					time[i].avg_render_time_in_us_last_ten,
-					time[i].min_window,
-					time[i].lfc_mid_point_in_us,
-					time[i].max_window,
-					time[i].vsync_to_flip_time_in_us,
-					time[i].flip_to_vsync_time_in_us,
-					time[i].num_vsync_between_flips,
-					time[i].num_frames_inserted,
-					time[i].inserted_duration_in_us,
-					time[i].v_total_min,
-					time[i].v_total_max,
-					time[i].event_triggers,
-					time[i].v_sync_time_in_us[0],
-					time[i].v_sync_time_in_us[1],
-					time[i].v_sync_time_in_us[2],
-					time[i].v_sync_time_in_us[3],
-					time[i].v_sync_time_in_us[4],
-					time[i].flags);
-		}
+	dm_logger_write(logger, LOG_DISPLAYSTATS, "==Display Caps==");
+	dm_logger_write(logger, LOG_DISPLAYSTATS, " ");
+
+	dm_logger_write(logger, LOG_DISPLAYSTATS, "==Display Stats==");
+	dm_logger_write(logger, LOG_DISPLAYSTATS, " ");
+
+	dm_logger_write(logger, LOG_DISPLAYSTATS,
+		"%10s %10s %10s %10s %10s"
+			" %11s %11s %17s %10s %14s"
+			" %10s %10s %10s %10s %10s"
+			" %10s %10s %10s %10s",
+		"render", "avgRender",
+		"minWindow", "midPoint", "maxWindow",
+		"vsyncToFlip", "flipToVsync", "vsyncsBetweenFlip",
+		"numFrame", "insertDuration",
+		"vTotalMin", "vTotalMax", "eventTrigs",
+		"vSyncTime1", "vSyncTime2", "vSyncTime3",
+		"vSyncTime4", "vSyncTime5", "flags");
+
+	for (int i = 0; i < core_stats->index && i < core_stats->entries; i++) {
+		dm_logger_write(logger, LOG_DISPLAYSTATS,
+			"%10u %10u %10u %10u %10u"
+				" %11u %11u %17u %10u %14u"
+				" %10u %10u %10u %10u %10u"
+				" %10u %10u %10u %10u",
+			time[i].render_time_in_us,
+			time[i].avg_render_time_in_us_last_ten,
+			time[i].min_window,
+			time[i].lfc_mid_point_in_us,
+			time[i].max_window,
+			time[i].vsync_to_flip_time_in_us,
+			time[i].flip_to_vsync_time_in_us,
+			time[i].num_vsync_between_flips,
+			time[i].num_frames_inserted,
+			time[i].inserted_duration_in_us,
+			time[i].v_total_min,
+			time[i].v_total_max,
+			time[i].event_triggers,
+			time[i].v_sync_time_in_us[0],
+			time[i].v_sync_time_in_us[1],
+			time[i].v_sync_time_in_us[2],
+			time[i].v_sync_time_in_us[3],
+			time[i].v_sync_time_in_us[4],
+			time[i].flags);
 	}
-	//GetLog()->Close(pLog);
-	//GetLog()->UnSetLogMask(LogMajor_ISR, LogMinor_ISR_FreeSyncSW);
 }
 
 void mod_stats_reset_data(struct mod_stats *mod_stats)
-- 
cgit v1.1


From ab9c2062d960df84d41c03efc49cb01071b398c6 Mon Sep 17 00:00:00 2001
From: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Date: Wed, 18 Apr 2018 14:11:43 -0400
Subject: drm/amd/display: add fixed point fractional bit truncation function

Signed-off-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/include/fixed31_32.h | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/include/fixed31_32.h b/drivers/gpu/drm/amd/display/include/fixed31_32.h
index ebfd33e..61f11e2 100644
--- a/drivers/gpu/drm/amd/display/include/fixed31_32.h
+++ b/drivers/gpu/drm/amd/display/include/fixed31_32.h
@@ -496,4 +496,21 @@ unsigned int dc_fixpt_clamp_u0d10(struct fixed31_32 arg);
 
 int dc_fixpt_s4d19(struct fixed31_32 arg);
 
+static inline struct fixed31_32 dc_fixpt_truncate(struct fixed31_32 arg, unsigned int frac_bits)
+{
+	bool negative = arg.value < 0;
+
+	if (frac_bits >= FIXED31_32_BITS_PER_FRACTIONAL_PART) {
+		ASSERT(frac_bits == FIXED31_32_BITS_PER_FRACTIONAL_PART);
+		return arg;
+	}
+
+	if (negative)
+		arg.value = -arg.value;
+	arg.value &= (~0LL) << (FIXED31_32_BITS_PER_FRACTIONAL_PART - frac_bits);
+	if (negative)
+		arg.value = -arg.value;
+	return arg;
+}
+
 #endif
-- 
cgit v1.1


From 0002d3ac8aadcb2850475557de32234b447ba502 Mon Sep 17 00:00:00 2001
From: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Date: Wed, 18 Apr 2018 14:19:23 -0400
Subject: drm/amd/display: truncate scaling ratios and inits to 19 bit
 precision

Signed-off-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_resource.c | 25 +++++++++++++++--------
 1 file changed, 17 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
index 082458f..751f3ac 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
@@ -652,6 +652,14 @@ static void calculate_scaling_ratios(struct pipe_ctx *pipe_ctx)
 		pipe_ctx->plane_res.scl_data.ratios.horz_c.value /= 2;
 		pipe_ctx->plane_res.scl_data.ratios.vert_c.value /= 2;
 	}
+	pipe_ctx->plane_res.scl_data.ratios.horz = dc_fixpt_truncate(
+			pipe_ctx->plane_res.scl_data.ratios.horz, 19);
+	pipe_ctx->plane_res.scl_data.ratios.vert = dc_fixpt_truncate(
+			pipe_ctx->plane_res.scl_data.ratios.vert, 19);
+	pipe_ctx->plane_res.scl_data.ratios.horz_c = dc_fixpt_truncate(
+			pipe_ctx->plane_res.scl_data.ratios.horz_c, 19);
+	pipe_ctx->plane_res.scl_data.ratios.vert_c = dc_fixpt_truncate(
+			pipe_ctx->plane_res.scl_data.ratios.vert_c, 19);
 }
 
 static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *recout_skip)
@@ -688,17 +696,18 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *r
 	 * 	init_bot = init + scaling_ratio
 	 * 	init_c = init + truncated_vp_c_offset(from calculate viewport)
 	 */
-	data->inits.h = dc_fixpt_div_int(
-			dc_fixpt_add_int(data->ratios.horz, data->taps.h_taps + 1), 2);
+	data->inits.h = dc_fixpt_truncate(dc_fixpt_div_int(
+			dc_fixpt_add_int(data->ratios.horz, data->taps.h_taps + 1), 2), 19);
 
-	data->inits.h_c = dc_fixpt_add(data->inits.h_c, dc_fixpt_div_int(
-			dc_fixpt_add_int(data->ratios.horz_c, data->taps.h_taps_c + 1), 2));
+	data->inits.h_c = dc_fixpt_truncate(dc_fixpt_add(data->inits.h_c, dc_fixpt_div_int(
+			dc_fixpt_add_int(data->ratios.horz_c, data->taps.h_taps_c + 1), 2)), 19);
 
-	data->inits.v = dc_fixpt_div_int(
-			dc_fixpt_add_int(data->ratios.vert, data->taps.v_taps + 1), 2);
+	data->inits.v = dc_fixpt_truncate(dc_fixpt_div_int(
+			dc_fixpt_add_int(data->ratios.vert, data->taps.v_taps + 1), 2), 19);
+
+	data->inits.v_c = dc_fixpt_truncate(dc_fixpt_add(data->inits.v_c, dc_fixpt_div_int(
+			dc_fixpt_add_int(data->ratios.vert_c, data->taps.v_taps_c + 1), 2)), 19);
 
-	data->inits.v_c = dc_fixpt_add(data->inits.v_c, dc_fixpt_div_int(
-			dc_fixpt_add_int(data->ratios.vert_c, data->taps.v_taps_c + 1), 2));
 
 
 	/* Adjust for viewport end clip-off */
-- 
cgit v1.1


From 3ba43a59927fbde07414393dfc2b6753cb233e00 Mon Sep 17 00:00:00 2001
From: Charlene Liu <charlene.liu@amd.com>
Date: Wed, 18 Apr 2018 14:31:41 -0400
Subject: drm/amd/display: underflow/blankscreen recovery

[Description]
for any reason, if driver detects HUBP underflow,
if a debug option enabled to enable recovery.
it will kick in a sequence of recovery.

Signed-off-by: Charlene Liu <charlene.liu@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dc.h                |  2 +
 .../gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c    |  8 ++
 .../gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h    |  7 +-
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c  | 24 ++++++
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h  |  3 +
 .../drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c  | 90 +++++++++++++++++++++-
 .../gpu/drm/amd/display/dc/dcn10/dcn10_resource.c  |  1 +
 drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h       |  2 +
 8 files changed, 135 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index cd4f434..1c39c99 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -239,6 +239,8 @@ struct dc_debug {
 	bool az_endpoint_mute_only;
 	bool always_use_regamma;
 	bool p010_mpo_support;
+	bool recovery_enabled;
+
 };
 struct dc_state;
 struct resource_pool;
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c
index b9fb14a..943143e 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c
@@ -476,6 +476,14 @@ void hubbub1_toggle_watermark_change_req(struct hubbub *hubbub)
 			DCHUBBUB_ARB_WATERMARK_CHANGE_REQUEST, watermark_change_req);
 }
 
+void hubbub1_soft_reset(struct hubbub *hubbub, bool reset)
+{
+	uint32_t reset_en = reset ? 1 : 0;
+
+	REG_UPDATE(DCHUBBUB_SOFT_RESET,
+			DCHUBBUB_GLOBAL_SOFT_RESET, reset_en);
+}
+
 static bool hubbub1_dcc_support_swizzle(
 		enum swizzle_mode_values swizzle,
 		unsigned int bytes_per_element,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h
index f479f54..6315a0e 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h
@@ -48,7 +48,8 @@
 	SR(DCHUBBUB_ARB_DF_REQ_OUTSTAND),\
 	SR(DCHUBBUB_GLOBAL_TIMER_CNTL), \
 	SR(DCHUBBUB_TEST_DEBUG_INDEX), \
-	SR(DCHUBBUB_TEST_DEBUG_DATA)
+	SR(DCHUBBUB_TEST_DEBUG_DATA),\
+	SR(DCHUBBUB_SOFT_RESET)
 
 #define HUBBUB_SR_WATERMARK_REG_LIST()\
 	SR(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_A),\
@@ -105,6 +106,7 @@ struct dcn_hubbub_registers {
 	uint32_t DCHUBBUB_SDPIF_AGP_BOT;
 	uint32_t DCHUBBUB_SDPIF_AGP_TOP;
 	uint32_t DCHUBBUB_CRC_CTRL;
+	uint32_t DCHUBBUB_SOFT_RESET;
 };
 
 /* set field name */
@@ -114,6 +116,7 @@ struct dcn_hubbub_registers {
 
 #define HUBBUB_MASK_SH_LIST_DCN(mask_sh)\
 		HUBBUB_SF(DCHUBBUB_GLOBAL_TIMER_CNTL, DCHUBBUB_GLOBAL_TIMER_ENABLE, mask_sh), \
+		HUBBUB_SF(DCHUBBUB_SOFT_RESET, DCHUBBUB_GLOBAL_SOFT_RESET, mask_sh), \
 		HUBBUB_SF(DCHUBBUB_ARB_WATERMARK_CHANGE_CNTL, DCHUBBUB_ARB_WATERMARK_CHANGE_REQUEST, mask_sh), \
 		HUBBUB_SF(DCHUBBUB_ARB_WATERMARK_CHANGE_CNTL, DCHUBBUB_ARB_WATERMARK_CHANGE_DONE_INTERRUPT_DISABLE, mask_sh), \
 		HUBBUB_SF(DCHUBBUB_ARB_DRAM_STATE_CNTL, DCHUBBUB_ARB_ALLOW_SELF_REFRESH_FORCE_VALUE, mask_sh), \
@@ -143,6 +146,7 @@ struct dcn_hubbub_registers {
 		type DCHUBBUB_ARB_SAT_LEVEL;\
 		type DCHUBBUB_ARB_MIN_REQ_OUTSTAND;\
 		type DCHUBBUB_GLOBAL_TIMER_REFDIV;\
+		type DCHUBBUB_GLOBAL_SOFT_RESET; \
 		type SDPIF_FB_TOP;\
 		type SDPIF_FB_BASE;\
 		type SDPIF_FB_OFFSET;\
@@ -201,6 +205,7 @@ void hubbub1_toggle_watermark_change_req(
 void hubbub1_wm_read_state(struct hubbub *hubbub,
 		struct dcn_hubbub_wm *wm);
 
+void hubbub1_soft_reset(struct hubbub *hubbub, bool reset);
 void hubbub1_construct(struct hubbub *hubbub,
 	struct dc_context *ctx,
 	const struct dcn_hubbub_registers *hubbub_regs,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
index 185f93b..d2ab78b 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
@@ -78,6 +78,27 @@ static void hubp1_disconnect(struct hubp *hubp)
 			CURSOR_ENABLE, 0);
 }
 
+static void hubp1_disable_control(struct hubp *hubp, bool disable_hubp)
+{
+	struct dcn10_hubp *hubp1 = TO_DCN10_HUBP(hubp);
+	uint32_t disable = disable_hubp ? 1 : 0;
+
+	REG_UPDATE(DCHUBP_CNTL,
+			HUBP_DISABLE, disable);
+}
+
+static unsigned int hubp1_get_underflow_status(struct hubp *hubp)
+{
+	uint32_t hubp_underflow = 0;
+	struct dcn10_hubp *hubp1 = TO_DCN10_HUBP(hubp);
+
+	REG_GET(DCHUBP_CNTL,
+		HUBP_UNDERFLOW_STATUS,
+		&hubp_underflow);
+
+	return hubp_underflow;
+}
+
 static void hubp1_set_hubp_blank_en(struct hubp *hubp, bool blank)
 {
 	struct dcn10_hubp *hubp1 = TO_DCN10_HUBP(hubp);
@@ -1117,6 +1138,9 @@ static struct hubp_funcs dcn10_hubp_funcs = {
 	.hubp_clk_cntl = hubp1_clk_cntl,
 	.hubp_vtg_sel = hubp1_vtg_sel,
 	.hubp_read_state = hubp1_read_state,
+	.hubp_disable_control =  hubp1_disable_control,
+	.hubp_get_underflow_status = hubp1_get_underflow_status,
+
 };
 
 /*****************************************/
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
index fe9b8c4..af38403 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
@@ -253,6 +253,7 @@
 	HUBP_SF(HUBP0_DCHUBP_CNTL, HUBP_UNDERFLOW_STATUS, mask_sh),\
 	HUBP_SF(HUBP0_DCHUBP_CNTL, HUBP_NO_OUTSTANDING_REQ, mask_sh),\
 	HUBP_SF(HUBP0_DCHUBP_CNTL, HUBP_VTG_SEL, mask_sh),\
+	HUBP_SF(HUBP0_DCHUBP_CNTL, HUBP_DISABLE, mask_sh),\
 	HUBP_SF(HUBP0_DCSURF_ADDR_CONFIG, NUM_PIPES, mask_sh),\
 	HUBP_SF(HUBP0_DCSURF_ADDR_CONFIG, NUM_BANKS, mask_sh),\
 	HUBP_SF(HUBP0_DCSURF_ADDR_CONFIG, PIPE_INTERLEAVE, mask_sh),\
@@ -421,6 +422,7 @@
 
 #define DCN_HUBP_REG_FIELD_LIST(type) \
 	type HUBP_BLANK_EN;\
+	type HUBP_DISABLE;\
 	type HUBP_TTU_DISABLE;\
 	type HUBP_NO_OUTSTANDING_REQ;\
 	type HUBP_VTG_SEL;\
@@ -723,4 +725,5 @@ void hubp1_read_state(struct hubp *hubp);
 
 enum cursor_pitch hubp1_get_cursor_pitch(unsigned int pitch);
 
+
 #endif
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
index 50bd754..be8820d 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -747,6 +747,90 @@ static void reset_back_end_for_pipe(
 					pipe_ctx->pipe_idx, pipe_ctx->stream_res.tg->inst);
 }
 
+static bool dcn10_hw_wa_force_recovery(struct dc *dc)
+{
+	struct hubp *hubp ;
+	unsigned int i;
+	bool need_recover = true;
+
+	if (!dc->debug.recovery_enabled)
+		return false;
+
+	for (i = 0; i < dc->res_pool->pipe_count; i++) {
+		struct pipe_ctx *pipe_ctx =
+			&dc->current_state->res_ctx.pipe_ctx[i];
+		if (pipe_ctx != NULL) {
+			hubp = pipe_ctx->plane_res.hubp;
+			if (hubp != NULL) {
+				if (hubp->funcs->hubp_get_underflow_status(hubp) != 0) {
+					/* one pipe underflow, we will reset all the pipes*/
+					need_recover = true;
+				}
+			}
+		}
+	}
+	if (!need_recover)
+		return false;
+	/*
+	DCHUBP_CNTL:HUBP_BLANK_EN=1
+	DCHUBBUB_SOFT_RESET:DCHUBBUB_GLOBAL_SOFT_RESET=1
+	DCHUBP_CNTL:HUBP_DISABLE=1
+	DCHUBP_CNTL:HUBP_DISABLE=0
+	DCHUBBUB_SOFT_RESET:DCHUBBUB_GLOBAL_SOFT_RESET=0
+	DCSURF_PRIMARY_SURFACE_ADDRESS
+	DCHUBP_CNTL:HUBP_BLANK_EN=0
+	*/
+
+	for (i = 0; i < dc->res_pool->pipe_count; i++) {
+		struct pipe_ctx *pipe_ctx =
+			&dc->current_state->res_ctx.pipe_ctx[i];
+		if (pipe_ctx != NULL) {
+			hubp = pipe_ctx->plane_res.hubp;
+			/*DCHUBP_CNTL:HUBP_BLANK_EN=1*/
+			if (hubp != NULL)
+				hubp->funcs->set_hubp_blank_en(hubp, true);
+		}
+	}
+	/*DCHUBBUB_SOFT_RESET:DCHUBBUB_GLOBAL_SOFT_RESET=1*/
+	hubbub1_soft_reset(dc->res_pool->hubbub, true);
+
+	for (i = 0; i < dc->res_pool->pipe_count; i++) {
+		struct pipe_ctx *pipe_ctx =
+			&dc->current_state->res_ctx.pipe_ctx[i];
+		if (pipe_ctx != NULL) {
+			hubp = pipe_ctx->plane_res.hubp;
+			/*DCHUBP_CNTL:HUBP_DISABLE=1*/
+			if (hubp != NULL)
+				hubp->funcs->hubp_disable_control(hubp, true);
+		}
+	}
+	for (i = 0; i < dc->res_pool->pipe_count; i++) {
+		struct pipe_ctx *pipe_ctx =
+			&dc->current_state->res_ctx.pipe_ctx[i];
+		if (pipe_ctx != NULL) {
+			hubp = pipe_ctx->plane_res.hubp;
+			/*DCHUBP_CNTL:HUBP_DISABLE=0*/
+			if (hubp != NULL)
+				hubp->funcs->hubp_disable_control(hubp, true);
+		}
+	}
+	/*DCHUBBUB_SOFT_RESET:DCHUBBUB_GLOBAL_SOFT_RESET=0*/
+	hubbub1_soft_reset(dc->res_pool->hubbub, false);
+	for (i = 0; i < dc->res_pool->pipe_count; i++) {
+		struct pipe_ctx *pipe_ctx =
+			&dc->current_state->res_ctx.pipe_ctx[i];
+		if (pipe_ctx != NULL) {
+			hubp = pipe_ctx->plane_res.hubp;
+			/*DCHUBP_CNTL:HUBP_BLANK_EN=0*/
+			if (hubp != NULL)
+				hubp->funcs->set_hubp_blank_en(hubp, true);
+		}
+	}
+	return true;
+
+}
+
+
 static void dcn10_verify_allow_pstate_change_high(struct dc *dc)
 {
 	static bool should_log_hw_state; /* prevent hw state log by default */
@@ -755,8 +839,12 @@ static void dcn10_verify_allow_pstate_change_high(struct dc *dc)
 		if (should_log_hw_state) {
 			dcn10_log_hw_state(dc);
 		}
-
 		BREAK_TO_DEBUGGER();
+		if (dcn10_hw_wa_force_recovery(dc)) {
+		/*check again*/
+			if (!hubbub1_verify_allow_pstate_change_high(dc->res_pool->hubbub))
+				BREAK_TO_DEBUGGER();
+		}
 	}
 }
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
index 2c0a315..16c84e9 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
@@ -446,6 +446,7 @@ static const struct dc_debug debug_defaults_drv = {
 		.vsr_support = true,
 		.performance_trace = false,
 		.az_endpoint_mute_only = true,
+		.recovery_enabled = false, /*enable this by default after testing.*/
 };
 
 static const struct dc_debug debug_defaults_diags = {
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h b/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h
index 331f8ff..97df82c 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h
@@ -121,6 +121,8 @@ struct hubp_funcs {
 	void (*hubp_clk_cntl)(struct hubp *hubp, bool enable);
 	void (*hubp_vtg_sel)(struct hubp *hubp, uint32_t otg_inst);
 	void (*hubp_read_state)(struct hubp *hubp);
+	void (*hubp_disable_control)(struct hubp *hubp, bool disable_hubp);
+	unsigned int (*hubp_get_underflow_status)(struct hubp *hubp);
 
 };
 
-- 
cgit v1.1


From 6b8e1eb7c6e059d8bb52f24b13081205242fded9 Mon Sep 17 00:00:00 2001
From: Eric Bernstein <eric.bernstein@amd.com>
Date: Tue, 17 Apr 2018 16:50:28 -0400
Subject: drm/amd/display: Update HW sequencer initialization

Signed-off-by: Eric Bernstein <eric.bernstein@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c |  6 +++---
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h |  2 ++
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c         | 10 +++++-----
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h         | 10 ++++++++++
 4 files changed, 20 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
index be8820d..24bcc5e5 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -849,7 +849,7 @@ static void dcn10_verify_allow_pstate_change_high(struct dc *dc)
 }
 
 /* trigger HW to start disconnect plane from stream on the next vsync */
-static void plane_atomic_disconnect(struct dc *dc, struct pipe_ctx *pipe_ctx)
+void hwss1_plane_atomic_disconnect(struct dc *dc, struct pipe_ctx *pipe_ctx)
 {
 	struct hubp *hubp = pipe_ctx->plane_res.hubp;
 	int dpp_id = pipe_ctx->plane_res.dpp->inst;
@@ -1032,7 +1032,7 @@ static void dcn10_init_hw(struct dc *dc)
 		dc->res_pool->opps[i]->mpcc_disconnect_pending[pipe_ctx->plane_res.mpcc_inst] = true;
 		pipe_ctx->stream_res.opp = dc->res_pool->opps[i];
 
-		plane_atomic_disconnect(dc, pipe_ctx);
+		hwss1_plane_atomic_disconnect(dc, pipe_ctx);
 	}
 
 	for (i = 0; i < dc->res_pool->pipe_count; i++) {
@@ -2267,7 +2267,7 @@ static void dcn10_apply_ctx_for_surface(
 			old_pipe_ctx->plane_state &&
 			old_pipe_ctx->stream_res.tg == tg) {
 
-			plane_atomic_disconnect(dc, old_pipe_ctx);
+			hwss1_plane_atomic_disconnect(dc, old_pipe_ctx);
 			removed_pipe[i] = true;
 
 			DC_LOG_DC(
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h
index 6c526b5..44f734b 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h
@@ -37,4 +37,6 @@ extern void fill_display_configs(
 
 bool is_rgb_cspace(enum dc_color_space output_color_space);
 
+void hwss1_plane_atomic_disconnect(struct dc *dc, struct pipe_ctx *pipe_ctx);
+
 #endif /* __DC_HWSS_DCN10_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c
index c734b7f..f2fbce0 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c
@@ -360,7 +360,7 @@ void optc1_program_timing(
 
 }
 
-static void optc1_set_blank_data_double_buffer(struct timing_generator *optc, bool enable)
+void optc1_set_blank_data_double_buffer(struct timing_generator *optc, bool enable)
 {
 	struct optc *optc1 = DCN10TG_FROM_TG(optc);
 
@@ -1257,20 +1257,20 @@ void optc1_read_otg_state(struct optc *optc1,
 			OPTC_UNDERFLOW_OCCURRED_STATUS, &s->underflow_occurred_status);
 }
 
-static void optc1_clear_optc_underflow(struct timing_generator *optc)
+void optc1_clear_optc_underflow(struct timing_generator *optc)
 {
 	struct optc *optc1 = DCN10TG_FROM_TG(optc);
 
 	REG_UPDATE(OPTC_INPUT_GLOBAL_CONTROL, OPTC_UNDERFLOW_CLEAR, 1);
 }
 
-static void optc1_tg_init(struct timing_generator *optc)
+void optc1_tg_init(struct timing_generator *optc)
 {
 	optc1_set_blank_data_double_buffer(optc, true);
 	optc1_clear_optc_underflow(optc);
 }
 
-static bool optc1_is_tg_enabled(struct timing_generator *optc)
+bool optc1_is_tg_enabled(struct timing_generator *optc)
 {
 	struct optc *optc1 = DCN10TG_FROM_TG(optc);
 	uint32_t otg_enabled = 0;
@@ -1281,7 +1281,7 @@ static bool optc1_is_tg_enabled(struct timing_generator *optc)
 
 }
 
-static bool optc1_is_optc_underflow_occurred(struct timing_generator *optc)
+bool optc1_is_optc_underflow_occurred(struct timing_generator *optc)
 {
 	struct optc *optc1 = DCN10TG_FROM_TG(optc);
 	uint32_t underflow_occurred = 0;
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h
index 89e09e5..c62052f 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h
@@ -497,4 +497,14 @@ void optc1_program_stereo(struct timing_generator *optc,
 
 bool optc1_is_stereo_left_eye(struct timing_generator *optc);
 
+void optc1_clear_optc_underflow(struct timing_generator *optc);
+
+void optc1_tg_init(struct timing_generator *optc);
+
+bool optc1_is_tg_enabled(struct timing_generator *optc);
+
+bool optc1_is_optc_underflow_occurred(struct timing_generator *optc);
+
+void optc1_set_blank_data_double_buffer(struct timing_generator *optc, bool enable);
+
 #endif /* __DC_TIMING_GENERATOR_DCN10_H__ */
-- 
cgit v1.1


From a21ddec61c5ed30b58eea3268ad3e0c69452ebfe Mon Sep 17 00:00:00 2001
From: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Date: Mon, 23 Apr 2018 12:41:34 -0400
Subject: drm/amd/display: fix 31_32_fixpt shift functions

Signed-off-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Reviewed-by: Eric Yang <eric.yang2@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/include/fixed31_32.h | 26 ++++++++++++++++--------
 1 file changed, 18 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/include/fixed31_32.h b/drivers/gpu/drm/amd/display/include/fixed31_32.h
index 61f11e2..bd8a304 100644
--- a/drivers/gpu/drm/amd/display/include/fixed31_32.h
+++ b/drivers/gpu/drm/amd/display/include/fixed31_32.h
@@ -27,6 +27,12 @@
 #define __DAL_FIXED31_32_H__
 
 #define FIXED31_32_BITS_PER_FRACTIONAL_PART 32
+#ifndef LLONG_MIN
+#define LLONG_MIN (1LL<<63)
+#endif
+#ifndef LLONG_MAX
+#define LLONG_MAX (-1LL>>1)
+#endif
 
 /*
  * @brief
@@ -45,6 +51,7 @@ struct fixed31_32 {
 	long long value;
 };
 
+
 /*
  * @brief
  * Useful constants
@@ -201,14 +208,12 @@ static inline struct fixed31_32 dc_fixpt_clamp(
  */
 static inline struct fixed31_32 dc_fixpt_shl(struct fixed31_32 arg, unsigned char shift)
 {
-	struct fixed31_32 res;
-
 	ASSERT(((arg.value >= 0) && (arg.value <= LLONG_MAX >> shift)) ||
-		((arg.value < 0) && (arg.value >= LLONG_MIN >> shift)));
+		((arg.value < 0) && (arg.value >= (LLONG_MIN / (1 << shift)))));
 
-	res.value = arg.value << shift;
+	arg.value = arg.value << shift;
 
-	return res;
+	return arg;
 }
 
 /*
@@ -217,9 +222,14 @@ static inline struct fixed31_32 dc_fixpt_shl(struct fixed31_32 arg, unsigned cha
  */
 static inline struct fixed31_32 dc_fixpt_shr(struct fixed31_32 arg, unsigned char shift)
 {
-	struct fixed31_32 res;
-	res.value = arg.value >> shift;
-	return res;
+	bool negative = arg.value < 0;
+
+	if (negative)
+		arg.value = -arg.value;
+	arg.value = arg.value >> shift;
+	if (negative)
+		arg.value = -arg.value;
+	return arg;
 }
 
 /*
-- 
cgit v1.1


From 7ea034ce8188eaf61ce2b7d4e747e1f6e3bb8aa3 Mon Sep 17 00:00:00 2001
From: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Date: Mon, 23 Apr 2018 14:39:23 -0400
Subject: drm/amd/display: fix a 32 bit shift meant to be 64 warning

Signed-off-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/include/fixed31_32.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/include/fixed31_32.h b/drivers/gpu/drm/amd/display/include/fixed31_32.h
index bd8a304..76f64e9 100644
--- a/drivers/gpu/drm/amd/display/include/fixed31_32.h
+++ b/drivers/gpu/drm/amd/display/include/fixed31_32.h
@@ -209,7 +209,7 @@ static inline struct fixed31_32 dc_fixpt_clamp(
 static inline struct fixed31_32 dc_fixpt_shl(struct fixed31_32 arg, unsigned char shift)
 {
 	ASSERT(((arg.value >= 0) && (arg.value <= LLONG_MAX >> shift)) ||
-		((arg.value < 0) && (arg.value >= (LLONG_MIN / (1 << shift)))));
+		((arg.value < 0) && (arg.value >= (LLONG_MIN / (1LL << shift)))));
 
 	arg.value = arg.value << shift;
 
-- 
cgit v1.1


From 3f460907be1b53441526e644019bcf150c433f59 Mon Sep 17 00:00:00 2001
From: Xingyue Tao <xingyue.tao@amd.com>
Date: Thu, 19 Apr 2018 16:23:12 -0400
Subject: drm/amd/display: Add dc cap to restrict VSR downscaling src size

- Adds int max_downscale_src_width in dc struct
- Checks and does not support if downscale size is more than 4k (width > 3840)

Signed-off-by: Xingyue Tao <xingyue.tao@amd.com>
Reviewed-by: Charlene Liu <Charlene.Liu@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dc.h                   |  1 +
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c      | 16 +++++++++++-----
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c |  1 +
 3 files changed, 13 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index 1c39c99..08b29a7 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -203,6 +203,7 @@ struct dc_debug {
 	bool clock_trace;
 	bool validation_trace;
 	bool bandwidth_calcs_trace;
+	int max_downscale_src_width;
 
 	/* stutter efficiency related */
 	bool disable_stutter;
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
index 20796da..2da1389 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
@@ -145,12 +145,18 @@ bool dpp_get_optimal_number_of_taps(
 	else
 		pixel_width = scl_data->viewport.width;
 
-	/* Some ASICs does not support  FP16 scaling, so we reject modes require this*/
 	if (scl_data->viewport.width  != scl_data->h_active &&
-		scl_data->viewport.height != scl_data->v_active &&
-		dpp->caps->dscl_data_proc_format == DSCL_DATA_PRCESSING_FIXED_FORMAT &&
-		scl_data->format == PIXEL_FORMAT_FP16)
-		return false;
+		scl_data->viewport.height != scl_data->v_active) {
+
+		/* Some ASICs does not support  FP16 scaling, so we reject modes require this*/
+		if (dpp->caps->dscl_data_proc_format == DSCL_DATA_PRCESSING_FIXED_FORMAT &&
+			scl_data->format == PIXEL_FORMAT_FP16)
+			return false;
+
+		if (dpp->ctx->dc->debug.max_downscale_src_width != 0 &&
+			scl_data->viewport.width > dpp->ctx->dc->debug.max_downscale_src_width)
+			return false;
+	}
 
 	/* TODO: add lb check */
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
index 16c84e9..f69f3a5 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
@@ -447,6 +447,7 @@ static const struct dc_debug debug_defaults_drv = {
 		.performance_trace = false,
 		.az_endpoint_mute_only = true,
 		.recovery_enabled = false, /*enable this by default after testing.*/
+		.max_downscale_src_width = 3840,
 };
 
 static const struct dc_debug debug_defaults_diags = {
-- 
cgit v1.1


From 07049507fd1b5813f667bb34e6903369487f9e34 Mon Sep 17 00:00:00 2001
From: Yue Hin Lau <Yuehin.Lau@amd.com>
Date: Wed, 18 Apr 2018 16:07:04 -0400
Subject: drm/amd/display: disable mpo if brightness adjusted

Signed-off-by: Yue Hin Lau <Yuehin.Lau@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dc.h                   | 1 +
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c | 1 +
 2 files changed, 2 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index 08b29a7..7a9f600 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -75,6 +75,7 @@ struct dc_caps {
 	bool dynamic_audio;
 	bool is_apu;
 	bool dual_link_dvi;
+	bool post_blend_color_processing;
 };
 
 struct dc_dcc_surface_param {
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
index f69f3a5..ace2e03 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
@@ -1023,6 +1023,7 @@ static bool construct(
 	dc->caps.max_cursor_size = 256;
 	dc->caps.max_slave_planes = 1;
 	dc->caps.is_apu = true;
+	dc->caps.post_blend_color_processing = false;
 
 	if (dc->ctx->dce_environment == DCE_ENV_PRODUCTION_DRV)
 		dc->debug = debug_defaults_drv;
-- 
cgit v1.1


From a3cb1c1c8e5e494f7630349fbebb79b1787128a1 Mon Sep 17 00:00:00 2001
From: Nikola Cornij <nikola.cornij@amd.com>
Date: Mon, 23 Apr 2018 15:55:36 -0400
Subject: drm/amd/display: Log DTN only after the atomic commit in Diag

Also print HUBP info only if pipe enabled. This fixes having different
DTN logs for different test sequences.

Signed-off-by: Nikola Cornij <nikola.cornij@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c  | 119 +++++++++++----------
 1 file changed, 62 insertions(+), 57 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
index 24bcc5e5..c452972 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -127,24 +127,26 @@ static void dcn10_log_hubp_states(struct dc *dc)
 
 		hubp->funcs->hubp_read_state(hubp);
 
-		DTN_INFO("[%2d]:  %5xh  %6xh  %5d  %6d  %2xh  %2xh  %6xh"
-				"  %6d  %8d  %7d  %8xh",
-				hubp->inst,
-				s->pixel_format,
-				s->inuse_addr_hi,
-				s->viewport_width,
-				s->viewport_height,
-				s->rotation_angle,
-				s->h_mirror_en,
-				s->sw_mode,
-				s->dcc_en,
-				s->blank_en,
-				s->ttu_disable,
-				s->underflow_status);
-		DTN_INFO_MICRO_SEC(s->min_ttu_vblank);
-		DTN_INFO_MICRO_SEC(s->qos_level_low_wm);
-		DTN_INFO_MICRO_SEC(s->qos_level_high_wm);
-		DTN_INFO("\n");
+		if (!s->blank_en) {
+			DTN_INFO("[%2d]:  %5xh  %6xh  %5d  %6d  %2xh  %2xh  %6xh"
+					"  %6d  %8d  %7d  %8xh",
+					hubp->inst,
+					s->pixel_format,
+					s->inuse_addr_hi,
+					s->viewport_width,
+					s->viewport_height,
+					s->rotation_angle,
+					s->h_mirror_en,
+					s->sw_mode,
+					s->dcc_en,
+					s->blank_en,
+					s->ttu_disable,
+					s->underflow_status);
+			DTN_INFO_MICRO_SEC(s->min_ttu_vblank);
+			DTN_INFO_MICRO_SEC(s->qos_level_low_wm);
+			DTN_INFO_MICRO_SEC(s->qos_level_high_wm);
+			DTN_INFO("\n");
+		}
 	}
 
 	DTN_INFO("\n=========RQ========\n");
@@ -155,16 +157,17 @@ static void dcn10_log_hubp_states(struct dc *dc)
 		struct dcn_hubp_state *s = &(TO_DCN10_HUBP(pool->hubps[i])->state);
 		struct _vcs_dpi_display_rq_regs_st *rq_regs = &s->rq_regs;
 
-		DTN_INFO("[%2d]:  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh\n",
-			i, rq_regs->drq_expansion_mode, rq_regs->prq_expansion_mode, rq_regs->mrq_expansion_mode,
-			rq_regs->crq_expansion_mode, rq_regs->plane1_base_address, rq_regs->rq_regs_l.chunk_size,
-			rq_regs->rq_regs_l.min_chunk_size, rq_regs->rq_regs_l.meta_chunk_size,
-			rq_regs->rq_regs_l.min_meta_chunk_size, rq_regs->rq_regs_l.dpte_group_size,
-			rq_regs->rq_regs_l.mpte_group_size, rq_regs->rq_regs_l.swath_height,
-			rq_regs->rq_regs_l.pte_row_height_linear, rq_regs->rq_regs_c.chunk_size, rq_regs->rq_regs_c.min_chunk_size,
-			rq_regs->rq_regs_c.meta_chunk_size, rq_regs->rq_regs_c.min_meta_chunk_size,
-			rq_regs->rq_regs_c.dpte_group_size, rq_regs->rq_regs_c.mpte_group_size,
-			rq_regs->rq_regs_c.swath_height, rq_regs->rq_regs_c.pte_row_height_linear);
+		if (!s->blank_en)
+			DTN_INFO("[%2d]:  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh\n",
+				pool->hubps[i]->inst, rq_regs->drq_expansion_mode, rq_regs->prq_expansion_mode, rq_regs->mrq_expansion_mode,
+				rq_regs->crq_expansion_mode, rq_regs->plane1_base_address, rq_regs->rq_regs_l.chunk_size,
+				rq_regs->rq_regs_l.min_chunk_size, rq_regs->rq_regs_l.meta_chunk_size,
+				rq_regs->rq_regs_l.min_meta_chunk_size, rq_regs->rq_regs_l.dpte_group_size,
+				rq_regs->rq_regs_l.mpte_group_size, rq_regs->rq_regs_l.swath_height,
+				rq_regs->rq_regs_l.pte_row_height_linear, rq_regs->rq_regs_c.chunk_size, rq_regs->rq_regs_c.min_chunk_size,
+				rq_regs->rq_regs_c.meta_chunk_size, rq_regs->rq_regs_c.min_meta_chunk_size,
+				rq_regs->rq_regs_c.dpte_group_size, rq_regs->rq_regs_c.mpte_group_size,
+				rq_regs->rq_regs_c.swath_height, rq_regs->rq_regs_c.pte_row_height_linear);
 	}
 
 	DTN_INFO("========DLG========\n");
@@ -179,27 +182,28 @@ static void dcn10_log_hubp_states(struct dc *dc)
 		struct dcn_hubp_state *s = &(TO_DCN10_HUBP(pool->hubps[i])->state);
 		struct _vcs_dpi_display_dlg_regs_st *dlg_regs = &s->dlg_attr;
 
-		DTN_INFO("[%2d]:  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh"
-			"%  8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh"
-			"  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh\n",
-			i, dlg_regs->refcyc_h_blank_end, dlg_regs->dlg_vblank_end, dlg_regs->min_dst_y_next_start,
-			dlg_regs->refcyc_per_htotal, dlg_regs->refcyc_x_after_scaler, dlg_regs->dst_y_after_scaler,
-			dlg_regs->dst_y_prefetch, dlg_regs->dst_y_per_vm_vblank, dlg_regs->dst_y_per_row_vblank,
-			dlg_regs->dst_y_per_vm_flip, dlg_regs->dst_y_per_row_flip, dlg_regs->ref_freq_to_pix_freq,
-			dlg_regs->vratio_prefetch, dlg_regs->vratio_prefetch_c, dlg_regs->refcyc_per_pte_group_vblank_l,
-			dlg_regs->refcyc_per_pte_group_vblank_c, dlg_regs->refcyc_per_meta_chunk_vblank_l,
-			dlg_regs->refcyc_per_meta_chunk_vblank_c, dlg_regs->refcyc_per_pte_group_flip_l,
-			dlg_regs->refcyc_per_pte_group_flip_c, dlg_regs->refcyc_per_meta_chunk_flip_l,
-			dlg_regs->refcyc_per_meta_chunk_flip_c, dlg_regs->dst_y_per_pte_row_nom_l,
-			dlg_regs->dst_y_per_pte_row_nom_c, dlg_regs->refcyc_per_pte_group_nom_l,
-			dlg_regs->refcyc_per_pte_group_nom_c, dlg_regs->dst_y_per_meta_row_nom_l,
-			dlg_regs->dst_y_per_meta_row_nom_c, dlg_regs->refcyc_per_meta_chunk_nom_l,
-			dlg_regs->refcyc_per_meta_chunk_nom_c, dlg_regs->refcyc_per_line_delivery_pre_l,
-			dlg_regs->refcyc_per_line_delivery_pre_c, dlg_regs->refcyc_per_line_delivery_l,
-			dlg_regs->refcyc_per_line_delivery_c, dlg_regs->chunk_hdl_adjust_cur0, dlg_regs->dst_y_offset_cur1,
-			dlg_regs->chunk_hdl_adjust_cur1, dlg_regs->vready_after_vcount0, dlg_regs->dst_y_delta_drq_limit,
-			dlg_regs->xfc_reg_transfer_delay, dlg_regs->xfc_reg_precharge_delay,
-			dlg_regs->xfc_reg_remote_surface_flip_latency);
+		if (!s->blank_en)
+			DTN_INFO("[%2d]:  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh"
+				"%  8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh"
+				"  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh\n",
+				pool->hubps[i]->inst, dlg_regs->refcyc_h_blank_end, dlg_regs->dlg_vblank_end, dlg_regs->min_dst_y_next_start,
+				dlg_regs->refcyc_per_htotal, dlg_regs->refcyc_x_after_scaler, dlg_regs->dst_y_after_scaler,
+				dlg_regs->dst_y_prefetch, dlg_regs->dst_y_per_vm_vblank, dlg_regs->dst_y_per_row_vblank,
+				dlg_regs->dst_y_per_vm_flip, dlg_regs->dst_y_per_row_flip, dlg_regs->ref_freq_to_pix_freq,
+				dlg_regs->vratio_prefetch, dlg_regs->vratio_prefetch_c, dlg_regs->refcyc_per_pte_group_vblank_l,
+				dlg_regs->refcyc_per_pte_group_vblank_c, dlg_regs->refcyc_per_meta_chunk_vblank_l,
+				dlg_regs->refcyc_per_meta_chunk_vblank_c, dlg_regs->refcyc_per_pte_group_flip_l,
+				dlg_regs->refcyc_per_pte_group_flip_c, dlg_regs->refcyc_per_meta_chunk_flip_l,
+				dlg_regs->refcyc_per_meta_chunk_flip_c, dlg_regs->dst_y_per_pte_row_nom_l,
+				dlg_regs->dst_y_per_pte_row_nom_c, dlg_regs->refcyc_per_pte_group_nom_l,
+				dlg_regs->refcyc_per_pte_group_nom_c, dlg_regs->dst_y_per_meta_row_nom_l,
+				dlg_regs->dst_y_per_meta_row_nom_c, dlg_regs->refcyc_per_meta_chunk_nom_l,
+				dlg_regs->refcyc_per_meta_chunk_nom_c, dlg_regs->refcyc_per_line_delivery_pre_l,
+				dlg_regs->refcyc_per_line_delivery_pre_c, dlg_regs->refcyc_per_line_delivery_l,
+				dlg_regs->refcyc_per_line_delivery_c, dlg_regs->chunk_hdl_adjust_cur0, dlg_regs->dst_y_offset_cur1,
+				dlg_regs->chunk_hdl_adjust_cur1, dlg_regs->vready_after_vcount0, dlg_regs->dst_y_delta_drq_limit,
+				dlg_regs->xfc_reg_transfer_delay, dlg_regs->xfc_reg_precharge_delay,
+				dlg_regs->xfc_reg_remote_surface_flip_latency);
 	}
 
 	DTN_INFO("========TTU========\n");
@@ -210,14 +214,15 @@ static void dcn10_log_hubp_states(struct dc *dc)
 		struct dcn_hubp_state *s = &(TO_DCN10_HUBP(pool->hubps[i])->state);
 		struct _vcs_dpi_display_ttu_regs_st *ttu_regs = &s->ttu_attr;
 
-		DTN_INFO("[%2d]:  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh\n",
-			i, ttu_regs->qos_level_low_wm, ttu_regs->qos_level_high_wm, ttu_regs->min_ttu_vblank,
-			ttu_regs->qos_level_flip, ttu_regs->refcyc_per_req_delivery_pre_l, ttu_regs->refcyc_per_req_delivery_l,
-			ttu_regs->refcyc_per_req_delivery_pre_c, ttu_regs->refcyc_per_req_delivery_c, ttu_regs->refcyc_per_req_delivery_cur0,
-			ttu_regs->refcyc_per_req_delivery_pre_cur0, ttu_regs->refcyc_per_req_delivery_cur1,
-			ttu_regs->refcyc_per_req_delivery_pre_cur1, ttu_regs->qos_level_fixed_l, ttu_regs->qos_ramp_disable_l,
-			ttu_regs->qos_level_fixed_c, ttu_regs->qos_ramp_disable_c, ttu_regs->qos_level_fixed_cur0,
-			ttu_regs->qos_ramp_disable_cur0, ttu_regs->qos_level_fixed_cur1, ttu_regs->qos_ramp_disable_cur1);
+		if (!s->blank_en)
+			DTN_INFO("[%2d]:  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh\n",
+				pool->hubps[i]->inst, ttu_regs->qos_level_low_wm, ttu_regs->qos_level_high_wm, ttu_regs->min_ttu_vblank,
+				ttu_regs->qos_level_flip, ttu_regs->refcyc_per_req_delivery_pre_l, ttu_regs->refcyc_per_req_delivery_l,
+				ttu_regs->refcyc_per_req_delivery_pre_c, ttu_regs->refcyc_per_req_delivery_c, ttu_regs->refcyc_per_req_delivery_cur0,
+				ttu_regs->refcyc_per_req_delivery_pre_cur0, ttu_regs->refcyc_per_req_delivery_cur1,
+				ttu_regs->refcyc_per_req_delivery_pre_cur1, ttu_regs->qos_level_fixed_l, ttu_regs->qos_ramp_disable_l,
+				ttu_regs->qos_level_fixed_c, ttu_regs->qos_ramp_disable_c, ttu_regs->qos_level_fixed_cur0,
+				ttu_regs->qos_ramp_disable_cur0, ttu_regs->qos_level_fixed_cur1, ttu_regs->qos_ramp_disable_cur1);
 	}
 	DTN_INFO("\n");
 }
-- 
cgit v1.1


From cba5e8708ee6123af14ab1f1196353dcda3eb533 Mon Sep 17 00:00:00 2001
From: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Date: Tue, 20 Mar 2018 08:25:16 -0400
Subject: drm/amd/display: update dml to allow sync with DV

Signed-off-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../drm/amd/display/dc/dml/display_mode_enums.h    |  13 +
 .../drm/amd/display/dc/dml/display_mode_structs.h  | 962 +++++++++++----------
 .../gpu/drm/amd/display/dc/dml/dml_inline_defs.h   |  10 +
 3 files changed, 515 insertions(+), 470 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_enums.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_enums.h
index b1ad355..47c19f8 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_enums.h
+++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_enums.h
@@ -108,4 +108,17 @@ enum output_standard {
 	dm_std_uninitialized = 0, dm_std_cvtr2, dm_std_cvt
 };
 
+enum mpc_combine_affinity {
+	dm_mpc_always_when_possible,
+	dm_mpc_reduce_voltage,
+	dm_mpc_reduce_voltage_and_clocks
+};
+
+enum self_refresh_affinity {
+	dm_try_to_allow_self_refresh_and_mclk_switch,
+	dm_allow_self_refresh_and_mclk_switch,
+	dm_allow_self_refresh,
+	dm_neither_self_refresh_nor_mclk_switch
+};
+
 #endif
diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h
index ce750ed..7fa0375 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h
+++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h
@@ -25,39 +25,39 @@
 #ifndef __DISPLAY_MODE_STRUCTS_H__
 #define __DISPLAY_MODE_STRUCTS_H__
 
-typedef struct _vcs_dpi_voltage_scaling_st	voltage_scaling_st;
-typedef struct _vcs_dpi_soc_bounding_box_st	soc_bounding_box_st;
-typedef struct _vcs_dpi_ip_params_st	ip_params_st;
-typedef struct _vcs_dpi_display_pipe_source_params_st	display_pipe_source_params_st;
-typedef struct _vcs_dpi_display_output_params_st	display_output_params_st;
-typedef struct _vcs_dpi_display_bandwidth_st	display_bandwidth_st;
-typedef struct _vcs_dpi_scaler_ratio_depth_st	scaler_ratio_depth_st;
-typedef struct _vcs_dpi_scaler_taps_st	scaler_taps_st;
-typedef struct _vcs_dpi_display_pipe_dest_params_st	display_pipe_dest_params_st;
-typedef struct _vcs_dpi_display_pipe_params_st	display_pipe_params_st;
-typedef struct _vcs_dpi_display_clocks_and_cfg_st	display_clocks_and_cfg_st;
-typedef struct _vcs_dpi_display_e2e_pipe_params_st	display_e2e_pipe_params_st;
-typedef struct _vcs_dpi_dchub_buffer_sizing_st	dchub_buffer_sizing_st;
-typedef struct _vcs_dpi_watermarks_perf_st	watermarks_perf_st;
-typedef struct _vcs_dpi_cstate_pstate_watermarks_st	cstate_pstate_watermarks_st;
-typedef struct _vcs_dpi_wm_calc_pipe_params_st	wm_calc_pipe_params_st;
-typedef struct _vcs_dpi_vratio_pre_st	vratio_pre_st;
-typedef struct _vcs_dpi_display_data_rq_misc_params_st	display_data_rq_misc_params_st;
-typedef struct _vcs_dpi_display_data_rq_sizing_params_st	display_data_rq_sizing_params_st;
-typedef struct _vcs_dpi_display_data_rq_dlg_params_st	display_data_rq_dlg_params_st;
-typedef struct _vcs_dpi_display_cur_rq_dlg_params_st	display_cur_rq_dlg_params_st;
-typedef struct _vcs_dpi_display_rq_dlg_params_st	display_rq_dlg_params_st;
-typedef struct _vcs_dpi_display_rq_sizing_params_st	display_rq_sizing_params_st;
-typedef struct _vcs_dpi_display_rq_misc_params_st	display_rq_misc_params_st;
-typedef struct _vcs_dpi_display_rq_params_st	display_rq_params_st;
-typedef struct _vcs_dpi_display_dlg_regs_st	display_dlg_regs_st;
-typedef struct _vcs_dpi_display_ttu_regs_st	display_ttu_regs_st;
-typedef struct _vcs_dpi_display_data_rq_regs_st	display_data_rq_regs_st;
-typedef struct _vcs_dpi_display_rq_regs_st	display_rq_regs_st;
-typedef struct _vcs_dpi_display_dlg_sys_params_st	display_dlg_sys_params_st;
-typedef struct _vcs_dpi_display_dlg_prefetch_param_st	display_dlg_prefetch_param_st;
-typedef struct _vcs_dpi_display_pipe_clock_st	display_pipe_clock_st;
-typedef struct _vcs_dpi_display_arb_params_st	display_arb_params_st;
+typedef struct _vcs_dpi_voltage_scaling_st voltage_scaling_st;
+typedef struct _vcs_dpi_soc_bounding_box_st soc_bounding_box_st;
+typedef struct _vcs_dpi_ip_params_st ip_params_st;
+typedef struct _vcs_dpi_display_pipe_source_params_st display_pipe_source_params_st;
+typedef struct _vcs_dpi_display_output_params_st display_output_params_st;
+typedef struct _vcs_dpi_display_bandwidth_st display_bandwidth_st;
+typedef struct _vcs_dpi_scaler_ratio_depth_st scaler_ratio_depth_st;
+typedef struct _vcs_dpi_scaler_taps_st scaler_taps_st;
+typedef struct _vcs_dpi_display_pipe_dest_params_st display_pipe_dest_params_st;
+typedef struct _vcs_dpi_display_pipe_params_st display_pipe_params_st;
+typedef struct _vcs_dpi_display_clocks_and_cfg_st display_clocks_and_cfg_st;
+typedef struct _vcs_dpi_display_e2e_pipe_params_st display_e2e_pipe_params_st;
+typedef struct _vcs_dpi_dchub_buffer_sizing_st dchub_buffer_sizing_st;
+typedef struct _vcs_dpi_watermarks_perf_st watermarks_perf_st;
+typedef struct _vcs_dpi_cstate_pstate_watermarks_st cstate_pstate_watermarks_st;
+typedef struct _vcs_dpi_wm_calc_pipe_params_st wm_calc_pipe_params_st;
+typedef struct _vcs_dpi_vratio_pre_st vratio_pre_st;
+typedef struct _vcs_dpi_display_data_rq_misc_params_st display_data_rq_misc_params_st;
+typedef struct _vcs_dpi_display_data_rq_sizing_params_st display_data_rq_sizing_params_st;
+typedef struct _vcs_dpi_display_data_rq_dlg_params_st display_data_rq_dlg_params_st;
+typedef struct _vcs_dpi_display_cur_rq_dlg_params_st display_cur_rq_dlg_params_st;
+typedef struct _vcs_dpi_display_rq_dlg_params_st display_rq_dlg_params_st;
+typedef struct _vcs_dpi_display_rq_sizing_params_st display_rq_sizing_params_st;
+typedef struct _vcs_dpi_display_rq_misc_params_st display_rq_misc_params_st;
+typedef struct _vcs_dpi_display_rq_params_st display_rq_params_st;
+typedef struct _vcs_dpi_display_dlg_regs_st display_dlg_regs_st;
+typedef struct _vcs_dpi_display_ttu_regs_st display_ttu_regs_st;
+typedef struct _vcs_dpi_display_data_rq_regs_st display_data_rq_regs_st;
+typedef struct _vcs_dpi_display_rq_regs_st display_rq_regs_st;
+typedef struct _vcs_dpi_display_dlg_sys_params_st display_dlg_sys_params_st;
+typedef struct _vcs_dpi_display_dlg_prefetch_param_st display_dlg_prefetch_param_st;
+typedef struct _vcs_dpi_display_pipe_clock_st display_pipe_clock_st;
+typedef struct _vcs_dpi_display_arb_params_st display_arb_params_st;
 
 struct _vcs_dpi_voltage_scaling_st {
 	int state;
@@ -72,89 +72,107 @@ struct _vcs_dpi_voltage_scaling_st {
 	double dppclk_mhz;
 };
 
-struct	_vcs_dpi_soc_bounding_box_st	{
-	double	sr_exit_time_us;
-	double	sr_enter_plus_exit_time_us;
-	double	urgent_latency_us;
-	double	writeback_latency_us;
-	double	ideal_dram_bw_after_urgent_percent;
-	unsigned int	max_request_size_bytes;
-	double	downspread_percent;
-	double	dram_page_open_time_ns;
-	double	dram_rw_turnaround_time_ns;
-	double	dram_return_buffer_per_channel_bytes;
-	double	dram_channel_width_bytes;
+struct _vcs_dpi_soc_bounding_box_st {
+	double sr_exit_time_us;
+	double sr_enter_plus_exit_time_us;
+	double urgent_latency_us;
+	double urgent_latency_pixel_data_only_us;
+	double urgent_latency_pixel_mixed_with_vm_data_us;
+	double urgent_latency_vm_data_only_us;
+	double writeback_latency_us;
+	double ideal_dram_bw_after_urgent_percent;
+	double pct_ideal_dram_sdp_bw_after_urgent_pixel_only; // PercentOfIdealDRAMFabricAndSDPPortBWReceivedAfterUrgLatencyPixelDataOnly
+	double pct_ideal_dram_sdp_bw_after_urgent_pixel_and_vm;
+	double pct_ideal_dram_sdp_bw_after_urgent_vm_only;
+	double max_avg_sdp_bw_use_normal_percent;
+	double max_avg_dram_bw_use_normal_percent;
+	unsigned int max_request_size_bytes;
+	double downspread_percent;
+	double dram_page_open_time_ns;
+	double dram_rw_turnaround_time_ns;
+	double dram_return_buffer_per_channel_bytes;
+	double dram_channel_width_bytes;
 	double fabric_datapath_to_dcn_data_return_bytes;
 	double dcn_downspread_percent;
 	double dispclk_dppclk_vco_speed_mhz;
 	double dfs_vco_period_ps;
-	unsigned int	round_trip_ping_latency_dcfclk_cycles;
-	unsigned int	urgent_out_of_order_return_per_channel_bytes;
-	unsigned int	channel_interleave_bytes;
-	unsigned int	num_banks;
-	unsigned int	num_chans;
-	unsigned int	vmm_page_size_bytes;
-	double	dram_clock_change_latency_us;
-	double	writeback_dram_clock_change_latency_us;
-	unsigned int	return_bus_width_bytes;
-	unsigned int	voltage_override;
-	double	xfc_bus_transport_time_us;
-	double	xfc_xbuf_latency_tolerance_us;
+	unsigned int urgent_out_of_order_return_per_channel_pixel_only_bytes;
+	unsigned int urgent_out_of_order_return_per_channel_pixel_and_vm_bytes;
+	unsigned int urgent_out_of_order_return_per_channel_vm_only_bytes;
+	unsigned int round_trip_ping_latency_dcfclk_cycles;
+	unsigned int urgent_out_of_order_return_per_channel_bytes;
+	unsigned int channel_interleave_bytes;
+	unsigned int num_banks;
+	unsigned int num_chans;
+	unsigned int vmm_page_size_bytes;
+	double dram_clock_change_latency_us;
+	double writeback_dram_clock_change_latency_us;
+	unsigned int return_bus_width_bytes;
+	unsigned int voltage_override;
+	double xfc_bus_transport_time_us;
+	double xfc_xbuf_latency_tolerance_us;
+	int use_urgent_burst_bw;
 	struct _vcs_dpi_voltage_scaling_st clock_limits[7];
 };
 
-struct	_vcs_dpi_ip_params_st	{
-	unsigned int	max_inter_dcn_tile_repeaters;
-	unsigned int	num_dsc;
-	unsigned int	odm_capable;
-	unsigned int	rob_buffer_size_kbytes;
-	unsigned int	det_buffer_size_kbytes;
-	unsigned int	dpte_buffer_size_in_pte_reqs;
-	unsigned int	pde_proc_buffer_size_64k_reqs;
-	unsigned int	dpp_output_buffer_pixels;
-	unsigned int	opp_output_buffer_lines;
-	unsigned int	pixel_chunk_size_kbytes;
-	unsigned char	pte_enable;
-	unsigned int	pte_chunk_size_kbytes;
-	unsigned int	meta_chunk_size_kbytes;
-	unsigned int	writeback_chunk_size_kbytes;
-	unsigned int	line_buffer_size_bits;
-	unsigned int	max_line_buffer_lines;
-	unsigned int	writeback_luma_buffer_size_kbytes;
-	unsigned int	writeback_chroma_buffer_size_kbytes;
-	unsigned int	writeback_chroma_line_buffer_width_pixels;
-	unsigned int	max_page_table_levels;
-	unsigned int	max_num_dpp;
-	unsigned int	max_num_otg;
-	unsigned int	cursor_chunk_size;
-	unsigned int	cursor_buffer_size;
-	unsigned int	max_num_wb;
-	unsigned int	max_dchub_pscl_bw_pix_per_clk;
-	unsigned int	max_pscl_lb_bw_pix_per_clk;
-	unsigned int	max_lb_vscl_bw_pix_per_clk;
-	unsigned int	max_vscl_hscl_bw_pix_per_clk;
-	double	max_hscl_ratio;
-	double	max_vscl_ratio;
-	unsigned int	hscl_mults;
-	unsigned int	vscl_mults;
-	unsigned int	max_hscl_taps;
-	unsigned int	max_vscl_taps;
-	unsigned int	xfc_supported;
-	unsigned int	xfc_fill_constant_bytes;
-	double	dispclk_ramp_margin_percent;
-	double	xfc_fill_bw_overhead_percent;
-	double	underscan_factor;
-	unsigned int	min_vblank_lines;
-	unsigned int	dppclk_delay_subtotal;
-	unsigned int	dispclk_delay_subtotal;
-	unsigned int	dcfclk_cstate_latency;
-	unsigned int	dppclk_delay_scl;
-	unsigned int	dppclk_delay_scl_lb_only;
-	unsigned int	dppclk_delay_cnvc_formatter;
-	unsigned int	dppclk_delay_cnvc_cursor;
-	unsigned int	is_line_buffer_bpp_fixed;
-	unsigned int	line_buffer_fixed_bpp;
-	unsigned int	dcc_supported;
+struct _vcs_dpi_ip_params_st {
+	bool gpuvm_enable;
+	bool hostvm_enable;
+	unsigned int gpuvm_max_page_table_levels;
+	unsigned int hostvm_max_page_table_levels;
+	unsigned int hostvm_cached_page_table_levels;
+	unsigned int pte_group_size_bytes;
+	unsigned int max_inter_dcn_tile_repeaters;
+	unsigned int num_dsc;
+	unsigned int odm_capable;
+	unsigned int rob_buffer_size_kbytes;
+	unsigned int det_buffer_size_kbytes;
+	unsigned int dpte_buffer_size_in_pte_reqs;
+	unsigned int pde_proc_buffer_size_64k_reqs;
+	unsigned int dpp_output_buffer_pixels;
+	unsigned int opp_output_buffer_lines;
+	unsigned int pixel_chunk_size_kbytes;
+	unsigned char pte_enable;
+	unsigned int pte_chunk_size_kbytes;
+	unsigned int meta_chunk_size_kbytes;
+	unsigned int writeback_chunk_size_kbytes;
+	unsigned int line_buffer_size_bits;
+	unsigned int max_line_buffer_lines;
+	unsigned int writeback_luma_buffer_size_kbytes;
+	unsigned int writeback_chroma_buffer_size_kbytes;
+	unsigned int writeback_chroma_line_buffer_width_pixels;
+	unsigned int max_page_table_levels;
+	unsigned int max_num_dpp;
+	unsigned int max_num_otg;
+	unsigned int cursor_chunk_size;
+	unsigned int cursor_buffer_size;
+	unsigned int max_num_wb;
+	unsigned int max_dchub_pscl_bw_pix_per_clk;
+	unsigned int max_pscl_lb_bw_pix_per_clk;
+	unsigned int max_lb_vscl_bw_pix_per_clk;
+	unsigned int max_vscl_hscl_bw_pix_per_clk;
+	double max_hscl_ratio;
+	double max_vscl_ratio;
+	unsigned int hscl_mults;
+	unsigned int vscl_mults;
+	unsigned int max_hscl_taps;
+	unsigned int max_vscl_taps;
+	unsigned int xfc_supported;
+	unsigned int xfc_fill_constant_bytes;
+	double dispclk_ramp_margin_percent;
+	double xfc_fill_bw_overhead_percent;
+	double underscan_factor;
+	unsigned int min_vblank_lines;
+	unsigned int dppclk_delay_subtotal;
+	unsigned int dispclk_delay_subtotal;
+	unsigned int dcfclk_cstate_latency;
+	unsigned int dppclk_delay_scl;
+	unsigned int dppclk_delay_scl_lb_only;
+	unsigned int dppclk_delay_cnvc_formatter;
+	unsigned int dppclk_delay_cnvc_cursor;
+	unsigned int is_line_buffer_bpp_fixed;
+	unsigned int line_buffer_fixed_bpp;
+	unsigned int dcc_supported;
 
 	unsigned int IsLineBufferBppFixed;
 	unsigned int LineBufferFixedBpp;
@@ -169,41 +187,45 @@ struct _vcs_dpi_display_xfc_params_st {
 	int xfc_slv_chunk_size_bytes;
 };
 
-struct	_vcs_dpi_display_pipe_source_params_st	{
-	int	source_format;
-	unsigned char	dcc;
-	unsigned int	dcc_override;
-	unsigned int	dcc_rate;
-	unsigned char	dcc_use_global;
-	unsigned char	vm;
-	unsigned char	vm_levels_force_en;
-	unsigned int	vm_levels_force;
-	int	source_scan;
-	int	sw_mode;
-	int	macro_tile_size;
-	unsigned char	is_display_sw;
-	unsigned int	viewport_width;
-	unsigned int	viewport_height;
-	unsigned int	viewport_y_y;
-	unsigned int	viewport_y_c;
-	unsigned int	viewport_width_c;
-	unsigned int	viewport_height_c;
-	unsigned int	data_pitch;
-	unsigned int	data_pitch_c;
-	unsigned int	meta_pitch;
-	unsigned int	meta_pitch_c;
-	unsigned int	cur0_src_width;
-	int	cur0_bpp;
-	unsigned int	cur1_src_width;
-	int	cur1_bpp;
-	int	num_cursors;
-	unsigned char	is_hsplit;
-	unsigned char	dynamic_metadata_enable;
-	unsigned int	dynamic_metadata_lines_before_active;
-	unsigned int	dynamic_metadata_xmit_bytes;
-	unsigned int	hsplit_grp;
-	unsigned char	xfc_enable;
-	unsigned char	xfc_slave;
+struct _vcs_dpi_display_pipe_source_params_st {
+	int source_format;
+	unsigned char dcc;
+	unsigned int dcc_override;
+	unsigned int dcc_rate;
+	unsigned char dcc_use_global;
+	unsigned char vm;
+	bool gpuvm;    // gpuvm enabled
+	bool hostvm;    // hostvm enabled
+	bool gpuvm_levels_force_en;
+	unsigned int gpuvm_levels_force;
+	bool hostvm_levels_force_en;
+	unsigned int hostvm_levels_force;
+	int source_scan;
+	int sw_mode;
+	int macro_tile_size;
+	unsigned char is_display_sw;
+	unsigned int viewport_width;
+	unsigned int viewport_height;
+	unsigned int viewport_y_y;
+	unsigned int viewport_y_c;
+	unsigned int viewport_width_c;
+	unsigned int viewport_height_c;
+	unsigned int data_pitch;
+	unsigned int data_pitch_c;
+	unsigned int meta_pitch;
+	unsigned int meta_pitch_c;
+	unsigned int cur0_src_width;
+	int cur0_bpp;
+	unsigned int cur1_src_width;
+	int cur1_bpp;
+	int num_cursors;
+	unsigned char is_hsplit;
+	unsigned char dynamic_metadata_enable;
+	unsigned int dynamic_metadata_lines_before_active;
+	unsigned int dynamic_metadata_xmit_bytes;
+	unsigned int hsplit_grp;
+	unsigned char xfc_enable;
+	unsigned char xfc_slave;
 	struct _vcs_dpi_display_xfc_params_st xfc_params;
 };
 struct writeback_st {
@@ -219,335 +241,335 @@ struct writeback_st {
 	double wb_vratio;
 };
 
-struct	_vcs_dpi_display_output_params_st	{
-	int	dp_lanes;
-	int	output_bpp;
-	int	dsc_enable;
-	int	wb_enable;
-	int	num_active_wb;
-	int	opp_input_bpc;
-	int	output_type;
-	int	output_format;
-	int	output_standard;
-	int	dsc_slices;
+struct _vcs_dpi_display_output_params_st {
+	int dp_lanes;
+	int output_bpp;
+	int dsc_enable;
+	int wb_enable;
+	int num_active_wb;
+	int output_bpc;
+	int output_type;
+	int output_format;
+	int output_standard;
+	int dsc_slices;
 	struct writeback_st wb;
 };
 
-struct	_vcs_dpi_display_bandwidth_st	{
-	double	total_bw_consumed_gbps;
-	double	guaranteed_urgent_return_bw_gbps;
-};
-
-struct	_vcs_dpi_scaler_ratio_depth_st	{
-	double	hscl_ratio;
-	double	vscl_ratio;
-	double	hscl_ratio_c;
-	double	vscl_ratio_c;
-	double	vinit;
-	double	vinit_c;
-	double	vinit_bot;
-	double	vinit_bot_c;
-	int	lb_depth;
-	int	scl_enable;
-};
-
-struct	_vcs_dpi_scaler_taps_st	{
-	unsigned int	htaps;
-	unsigned int	vtaps;
-	unsigned int	htaps_c;
-	unsigned int	vtaps_c;
-};
-
-struct	_vcs_dpi_display_pipe_dest_params_st	{
-	unsigned int	recout_width;
-	unsigned int	recout_height;
-	unsigned int	full_recout_width;
-	unsigned int	full_recout_height;
-	unsigned int	hblank_start;
-	unsigned int	hblank_end;
-	unsigned int	vblank_start;
-	unsigned int	vblank_end;
-	unsigned int	htotal;
-	unsigned int	vtotal;
-	unsigned int	vactive;
-	unsigned int	hactive;
-	unsigned int	vstartup_start;
-	unsigned int	vupdate_offset;
-	unsigned int	vupdate_width;
-	unsigned int	vready_offset;
-	unsigned char	interlaced;
-	unsigned char	underscan;
-	double	pixel_rate_mhz;
-	unsigned char	synchronized_vblank_all_planes;
-	unsigned char	otg_inst;
-	unsigned char	odm_split_cnt;
-	unsigned char	odm_combine;
-};
-
-struct	_vcs_dpi_display_pipe_params_st	{
-	display_pipe_source_params_st	src;
-	display_pipe_dest_params_st	dest;
-	scaler_ratio_depth_st	scale_ratio_depth;
-	scaler_taps_st	scale_taps;
-};
-
-struct	_vcs_dpi_display_clocks_and_cfg_st	{
-	int	voltage;
-	double	dppclk_mhz;
-	double	refclk_mhz;
-	double	dispclk_mhz;
-	double	dcfclk_mhz;
-	double	socclk_mhz;
-};
-
-struct	_vcs_dpi_display_e2e_pipe_params_st	{
-	display_pipe_params_st	pipe;
-	display_output_params_st	dout;
-	display_clocks_and_cfg_st	clks_cfg;
-};
-
-struct	_vcs_dpi_dchub_buffer_sizing_st	{
-	unsigned int	swath_width_y;
-	unsigned int	swath_height_y;
-	unsigned int	swath_height_c;
-	unsigned int	detail_buffer_size_y;
-};
-
-struct	_vcs_dpi_watermarks_perf_st	{
-	double	stutter_eff_in_active_region_percent;
-	double	urgent_latency_supported_us;
-	double	non_urgent_latency_supported_us;
-	double	dram_clock_change_margin_us;
-	double	dram_access_eff_percent;
-};
-
-struct	_vcs_dpi_cstate_pstate_watermarks_st	{
-	double	cstate_exit_us;
-	double	cstate_enter_plus_exit_us;
-	double	pstate_change_us;
-};
-
-struct	_vcs_dpi_wm_calc_pipe_params_st	{
-	unsigned int	num_dpp;
-	int	voltage;
-	int	output_type;
-	double	dcfclk_mhz;
-	double	socclk_mhz;
-	double	dppclk_mhz;
-	double	pixclk_mhz;
-	unsigned char	interlace_en;
-	unsigned char	pte_enable;
-	unsigned char	dcc_enable;
-	double	dcc_rate;
-	double	bytes_per_pixel_c;
-	double	bytes_per_pixel_y;
-	unsigned int	swath_width_y;
-	unsigned int	swath_height_y;
-	unsigned int	swath_height_c;
-	unsigned int	det_buffer_size_y;
-	double	h_ratio;
-	double	v_ratio;
-	unsigned int	h_taps;
-	unsigned int	h_total;
-	unsigned int	v_total;
-	unsigned int	v_active;
-	unsigned int	e2e_index;
-	double	display_pipe_line_delivery_time;
-	double	read_bw;
-	unsigned int	lines_in_det_y;
-	unsigned int	lines_in_det_y_rounded_down_to_swath;
-	double	full_det_buffering_time;
-	double	dcfclk_deepsleep_mhz_per_plane;
-};
-
-struct	_vcs_dpi_vratio_pre_st	{
-	double	vratio_pre_l;
-	double	vratio_pre_c;
-};
-
-struct	_vcs_dpi_display_data_rq_misc_params_st	{
-	unsigned int	full_swath_bytes;
-	unsigned int	stored_swath_bytes;
-	unsigned int	blk256_height;
-	unsigned int	blk256_width;
-	unsigned int	req_height;
-	unsigned int	req_width;
-};
-
-struct	_vcs_dpi_display_data_rq_sizing_params_st	{
-	unsigned int	chunk_bytes;
-	unsigned int	min_chunk_bytes;
-	unsigned int	meta_chunk_bytes;
-	unsigned int	min_meta_chunk_bytes;
-	unsigned int	mpte_group_bytes;
-	unsigned int	dpte_group_bytes;
-};
-
-struct	_vcs_dpi_display_data_rq_dlg_params_st	{
-	unsigned int	swath_width_ub;
-	unsigned int	swath_height;
-	unsigned int	req_per_swath_ub;
-	unsigned int	meta_pte_bytes_per_frame_ub;
-	unsigned int	dpte_req_per_row_ub;
-	unsigned int	dpte_groups_per_row_ub;
-	unsigned int	dpte_row_height;
-	unsigned int	dpte_bytes_per_row_ub;
-	unsigned int	meta_chunks_per_row_ub;
-	unsigned int	meta_req_per_row_ub;
-	unsigned int	meta_row_height;
-	unsigned int	meta_bytes_per_row_ub;
-};
-
-struct	_vcs_dpi_display_cur_rq_dlg_params_st	{
-	unsigned char	enable;
-	unsigned int	swath_height;
-	unsigned int	req_per_line;
-};
-
-struct	_vcs_dpi_display_rq_dlg_params_st	{
-	display_data_rq_dlg_params_st	rq_l;
-	display_data_rq_dlg_params_st	rq_c;
-	display_cur_rq_dlg_params_st	rq_cur0;
-};
-
-struct	_vcs_dpi_display_rq_sizing_params_st	{
-	display_data_rq_sizing_params_st	rq_l;
-	display_data_rq_sizing_params_st	rq_c;
-};
-
-struct	_vcs_dpi_display_rq_misc_params_st	{
-	display_data_rq_misc_params_st	rq_l;
-	display_data_rq_misc_params_st	rq_c;
-};
-
-struct	_vcs_dpi_display_rq_params_st	{
-	unsigned char	yuv420;
-	unsigned char	yuv420_10bpc;
-	display_rq_misc_params_st	misc;
-	display_rq_sizing_params_st	sizing;
-	display_rq_dlg_params_st	dlg;
-};
-
-struct	_vcs_dpi_display_dlg_regs_st	{
-	unsigned int	refcyc_h_blank_end;
-	unsigned int	dlg_vblank_end;
-	unsigned int	min_dst_y_next_start;
-	unsigned int	refcyc_per_htotal;
-	unsigned int	refcyc_x_after_scaler;
-	unsigned int	dst_y_after_scaler;
-	unsigned int	dst_y_prefetch;
-	unsigned int	dst_y_per_vm_vblank;
-	unsigned int	dst_y_per_row_vblank;
-	unsigned int	dst_y_per_vm_flip;
-	unsigned int	dst_y_per_row_flip;
-	unsigned int	ref_freq_to_pix_freq;
-	unsigned int	vratio_prefetch;
-	unsigned int	vratio_prefetch_c;
-	unsigned int	refcyc_per_pte_group_vblank_l;
-	unsigned int	refcyc_per_pte_group_vblank_c;
-	unsigned int	refcyc_per_meta_chunk_vblank_l;
-	unsigned int	refcyc_per_meta_chunk_vblank_c;
-	unsigned int	refcyc_per_pte_group_flip_l;
-	unsigned int	refcyc_per_pte_group_flip_c;
-	unsigned int	refcyc_per_meta_chunk_flip_l;
-	unsigned int	refcyc_per_meta_chunk_flip_c;
-	unsigned int	dst_y_per_pte_row_nom_l;
-	unsigned int	dst_y_per_pte_row_nom_c;
-	unsigned int	refcyc_per_pte_group_nom_l;
-	unsigned int	refcyc_per_pte_group_nom_c;
-	unsigned int	dst_y_per_meta_row_nom_l;
-	unsigned int	dst_y_per_meta_row_nom_c;
-	unsigned int	refcyc_per_meta_chunk_nom_l;
-	unsigned int	refcyc_per_meta_chunk_nom_c;
-	unsigned int	refcyc_per_line_delivery_pre_l;
-	unsigned int	refcyc_per_line_delivery_pre_c;
-	unsigned int	refcyc_per_line_delivery_l;
-	unsigned int	refcyc_per_line_delivery_c;
-	unsigned int	chunk_hdl_adjust_cur0;
-	unsigned int	chunk_hdl_adjust_cur1;
-	unsigned int	vready_after_vcount0;
-	unsigned int	dst_y_offset_cur0;
-	unsigned int	dst_y_offset_cur1;
-	unsigned int	xfc_reg_transfer_delay;
-	unsigned int	xfc_reg_precharge_delay;
-	unsigned int	xfc_reg_remote_surface_flip_latency;
-	unsigned int	xfc_reg_prefetch_margin;
-	unsigned int	dst_y_delta_drq_limit;
-};
-
-struct	_vcs_dpi_display_ttu_regs_st	{
-	unsigned int	qos_level_low_wm;
-	unsigned int	qos_level_high_wm;
-	unsigned int	min_ttu_vblank;
-	unsigned int	qos_level_flip;
-	unsigned int	refcyc_per_req_delivery_l;
-	unsigned int	refcyc_per_req_delivery_c;
-	unsigned int	refcyc_per_req_delivery_cur0;
-	unsigned int	refcyc_per_req_delivery_cur1;
-	unsigned int	refcyc_per_req_delivery_pre_l;
-	unsigned int	refcyc_per_req_delivery_pre_c;
-	unsigned int	refcyc_per_req_delivery_pre_cur0;
-	unsigned int	refcyc_per_req_delivery_pre_cur1;
-	unsigned int	qos_level_fixed_l;
-	unsigned int	qos_level_fixed_c;
-	unsigned int	qos_level_fixed_cur0;
-	unsigned int	qos_level_fixed_cur1;
-	unsigned int	qos_ramp_disable_l;
-	unsigned int	qos_ramp_disable_c;
-	unsigned int	qos_ramp_disable_cur0;
-	unsigned int	qos_ramp_disable_cur1;
-};
-
-struct	_vcs_dpi_display_data_rq_regs_st	{
-	unsigned int	chunk_size;
-	unsigned int	min_chunk_size;
-	unsigned int	meta_chunk_size;
-	unsigned int	min_meta_chunk_size;
-	unsigned int	dpte_group_size;
-	unsigned int	mpte_group_size;
-	unsigned int	swath_height;
-	unsigned int	pte_row_height_linear;
-};
-
-struct	_vcs_dpi_display_rq_regs_st	{
-	display_data_rq_regs_st	rq_regs_l;
-	display_data_rq_regs_st	rq_regs_c;
-	unsigned int	drq_expansion_mode;
-	unsigned int	prq_expansion_mode;
-	unsigned int	mrq_expansion_mode;
-	unsigned int	crq_expansion_mode;
-	unsigned int	plane1_base_address;
-};
-
-struct	_vcs_dpi_display_dlg_sys_params_st	{
-	double	t_mclk_wm_us;
-	double	t_urg_wm_us;
-	double	t_sr_wm_us;
-	double	t_extra_us;
-	double	mem_trip_us;
-	double	t_srx_delay_us;
-	double	deepsleep_dcfclk_mhz;
-	double	total_flip_bw;
-	unsigned int	total_flip_bytes;
-};
-
-struct	_vcs_dpi_display_dlg_prefetch_param_st	{
-	double	prefetch_bw;
-	unsigned int	flip_bytes;
-};
-
-struct	_vcs_dpi_display_pipe_clock_st	{
-	double	dcfclk_mhz;
-	double	dispclk_mhz;
-	double	socclk_mhz;
-	double	dscclk_mhz[6];
-	double	dppclk_mhz[6];
-};
-
-struct	_vcs_dpi_display_arb_params_st	{
-	int	max_req_outstanding;
-	int	min_req_outstanding;
-	int	sat_level_us;
+struct _vcs_dpi_display_bandwidth_st {
+	double total_bw_consumed_gbps;
+	double guaranteed_urgent_return_bw_gbps;
+};
+
+struct _vcs_dpi_scaler_ratio_depth_st {
+	double hscl_ratio;
+	double vscl_ratio;
+	double hscl_ratio_c;
+	double vscl_ratio_c;
+	double vinit;
+	double vinit_c;
+	double vinit_bot;
+	double vinit_bot_c;
+	int lb_depth;
+	int scl_enable;
+};
+
+struct _vcs_dpi_scaler_taps_st {
+	unsigned int htaps;
+	unsigned int vtaps;
+	unsigned int htaps_c;
+	unsigned int vtaps_c;
+};
+
+struct _vcs_dpi_display_pipe_dest_params_st {
+	unsigned int recout_width;
+	unsigned int recout_height;
+	unsigned int full_recout_width;
+	unsigned int full_recout_height;
+	unsigned int hblank_start;
+	unsigned int hblank_end;
+	unsigned int vblank_start;
+	unsigned int vblank_end;
+	unsigned int htotal;
+	unsigned int vtotal;
+	unsigned int vactive;
+	unsigned int hactive;
+	unsigned int vstartup_start;
+	unsigned int vupdate_offset;
+	unsigned int vupdate_width;
+	unsigned int vready_offset;
+	unsigned char interlaced;
+	unsigned char underscan;
+	double pixel_rate_mhz;
+	unsigned char synchronized_vblank_all_planes;
+	unsigned char otg_inst;
+	unsigned char odm_split_cnt;
+	unsigned char odm_combine;
+};
+
+struct _vcs_dpi_display_pipe_params_st {
+	display_pipe_source_params_st src;
+	display_pipe_dest_params_st dest;
+	scaler_ratio_depth_st scale_ratio_depth;
+	scaler_taps_st scale_taps;
+};
+
+struct _vcs_dpi_display_clocks_and_cfg_st {
+	int voltage;
+	double dppclk_mhz;
+	double refclk_mhz;
+	double dispclk_mhz;
+	double dcfclk_mhz;
+	double socclk_mhz;
+};
+
+struct _vcs_dpi_display_e2e_pipe_params_st {
+	display_pipe_params_st pipe;
+	display_output_params_st dout;
+	display_clocks_and_cfg_st clks_cfg;
+};
+
+struct _vcs_dpi_dchub_buffer_sizing_st {
+	unsigned int swath_width_y;
+	unsigned int swath_height_y;
+	unsigned int swath_height_c;
+	unsigned int detail_buffer_size_y;
+};
+
+struct _vcs_dpi_watermarks_perf_st {
+	double stutter_eff_in_active_region_percent;
+	double urgent_latency_supported_us;
+	double non_urgent_latency_supported_us;
+	double dram_clock_change_margin_us;
+	double dram_access_eff_percent;
+};
+
+struct _vcs_dpi_cstate_pstate_watermarks_st {
+	double cstate_exit_us;
+	double cstate_enter_plus_exit_us;
+	double pstate_change_us;
+};
+
+struct _vcs_dpi_wm_calc_pipe_params_st {
+	unsigned int num_dpp;
+	int voltage;
+	int output_type;
+	double dcfclk_mhz;
+	double socclk_mhz;
+	double dppclk_mhz;
+	double pixclk_mhz;
+	unsigned char interlace_en;
+	unsigned char pte_enable;
+	unsigned char dcc_enable;
+	double dcc_rate;
+	double bytes_per_pixel_c;
+	double bytes_per_pixel_y;
+	unsigned int swath_width_y;
+	unsigned int swath_height_y;
+	unsigned int swath_height_c;
+	unsigned int det_buffer_size_y;
+	double h_ratio;
+	double v_ratio;
+	unsigned int h_taps;
+	unsigned int h_total;
+	unsigned int v_total;
+	unsigned int v_active;
+	unsigned int e2e_index;
+	double display_pipe_line_delivery_time;
+	double read_bw;
+	unsigned int lines_in_det_y;
+	unsigned int lines_in_det_y_rounded_down_to_swath;
+	double full_det_buffering_time;
+	double dcfclk_deepsleep_mhz_per_plane;
+};
+
+struct _vcs_dpi_vratio_pre_st {
+	double vratio_pre_l;
+	double vratio_pre_c;
+};
+
+struct _vcs_dpi_display_data_rq_misc_params_st {
+	unsigned int full_swath_bytes;
+	unsigned int stored_swath_bytes;
+	unsigned int blk256_height;
+	unsigned int blk256_width;
+	unsigned int req_height;
+	unsigned int req_width;
+};
+
+struct _vcs_dpi_display_data_rq_sizing_params_st {
+	unsigned int chunk_bytes;
+	unsigned int min_chunk_bytes;
+	unsigned int meta_chunk_bytes;
+	unsigned int min_meta_chunk_bytes;
+	unsigned int mpte_group_bytes;
+	unsigned int dpte_group_bytes;
+};
+
+struct _vcs_dpi_display_data_rq_dlg_params_st {
+	unsigned int swath_width_ub;
+	unsigned int swath_height;
+	unsigned int req_per_swath_ub;
+	unsigned int meta_pte_bytes_per_frame_ub;
+	unsigned int dpte_req_per_row_ub;
+	unsigned int dpte_groups_per_row_ub;
+	unsigned int dpte_row_height;
+	unsigned int dpte_bytes_per_row_ub;
+	unsigned int meta_chunks_per_row_ub;
+	unsigned int meta_req_per_row_ub;
+	unsigned int meta_row_height;
+	unsigned int meta_bytes_per_row_ub;
+};
+
+struct _vcs_dpi_display_cur_rq_dlg_params_st {
+	unsigned char enable;
+	unsigned int swath_height;
+	unsigned int req_per_line;
+};
+
+struct _vcs_dpi_display_rq_dlg_params_st {
+	display_data_rq_dlg_params_st rq_l;
+	display_data_rq_dlg_params_st rq_c;
+	display_cur_rq_dlg_params_st rq_cur0;
+};
+
+struct _vcs_dpi_display_rq_sizing_params_st {
+	display_data_rq_sizing_params_st rq_l;
+	display_data_rq_sizing_params_st rq_c;
+};
+
+struct _vcs_dpi_display_rq_misc_params_st {
+	display_data_rq_misc_params_st rq_l;
+	display_data_rq_misc_params_st rq_c;
+};
+
+struct _vcs_dpi_display_rq_params_st {
+	unsigned char yuv420;
+	unsigned char yuv420_10bpc;
+	display_rq_misc_params_st misc;
+	display_rq_sizing_params_st sizing;
+	display_rq_dlg_params_st dlg;
+};
+
+struct _vcs_dpi_display_dlg_regs_st {
+	unsigned int refcyc_h_blank_end;
+	unsigned int dlg_vblank_end;
+	unsigned int min_dst_y_next_start;
+	unsigned int refcyc_per_htotal;
+	unsigned int refcyc_x_after_scaler;
+	unsigned int dst_y_after_scaler;
+	unsigned int dst_y_prefetch;
+	unsigned int dst_y_per_vm_vblank;
+	unsigned int dst_y_per_row_vblank;
+	unsigned int dst_y_per_vm_flip;
+	unsigned int dst_y_per_row_flip;
+	unsigned int ref_freq_to_pix_freq;
+	unsigned int vratio_prefetch;
+	unsigned int vratio_prefetch_c;
+	unsigned int refcyc_per_pte_group_vblank_l;
+	unsigned int refcyc_per_pte_group_vblank_c;
+	unsigned int refcyc_per_meta_chunk_vblank_l;
+	unsigned int refcyc_per_meta_chunk_vblank_c;
+	unsigned int refcyc_per_pte_group_flip_l;
+	unsigned int refcyc_per_pte_group_flip_c;
+	unsigned int refcyc_per_meta_chunk_flip_l;
+	unsigned int refcyc_per_meta_chunk_flip_c;
+	unsigned int dst_y_per_pte_row_nom_l;
+	unsigned int dst_y_per_pte_row_nom_c;
+	unsigned int refcyc_per_pte_group_nom_l;
+	unsigned int refcyc_per_pte_group_nom_c;
+	unsigned int dst_y_per_meta_row_nom_l;
+	unsigned int dst_y_per_meta_row_nom_c;
+	unsigned int refcyc_per_meta_chunk_nom_l;
+	unsigned int refcyc_per_meta_chunk_nom_c;
+	unsigned int refcyc_per_line_delivery_pre_l;
+	unsigned int refcyc_per_line_delivery_pre_c;
+	unsigned int refcyc_per_line_delivery_l;
+	unsigned int refcyc_per_line_delivery_c;
+	unsigned int chunk_hdl_adjust_cur0;
+	unsigned int chunk_hdl_adjust_cur1;
+	unsigned int vready_after_vcount0;
+	unsigned int dst_y_offset_cur0;
+	unsigned int dst_y_offset_cur1;
+	unsigned int xfc_reg_transfer_delay;
+	unsigned int xfc_reg_precharge_delay;
+	unsigned int xfc_reg_remote_surface_flip_latency;
+	unsigned int xfc_reg_prefetch_margin;
+	unsigned int dst_y_delta_drq_limit;
+};
+
+struct _vcs_dpi_display_ttu_regs_st {
+	unsigned int qos_level_low_wm;
+	unsigned int qos_level_high_wm;
+	unsigned int min_ttu_vblank;
+	unsigned int qos_level_flip;
+	unsigned int refcyc_per_req_delivery_l;
+	unsigned int refcyc_per_req_delivery_c;
+	unsigned int refcyc_per_req_delivery_cur0;
+	unsigned int refcyc_per_req_delivery_cur1;
+	unsigned int refcyc_per_req_delivery_pre_l;
+	unsigned int refcyc_per_req_delivery_pre_c;
+	unsigned int refcyc_per_req_delivery_pre_cur0;
+	unsigned int refcyc_per_req_delivery_pre_cur1;
+	unsigned int qos_level_fixed_l;
+	unsigned int qos_level_fixed_c;
+	unsigned int qos_level_fixed_cur0;
+	unsigned int qos_level_fixed_cur1;
+	unsigned int qos_ramp_disable_l;
+	unsigned int qos_ramp_disable_c;
+	unsigned int qos_ramp_disable_cur0;
+	unsigned int qos_ramp_disable_cur1;
+};
+
+struct _vcs_dpi_display_data_rq_regs_st {
+	unsigned int chunk_size;
+	unsigned int min_chunk_size;
+	unsigned int meta_chunk_size;
+	unsigned int min_meta_chunk_size;
+	unsigned int dpte_group_size;
+	unsigned int mpte_group_size;
+	unsigned int swath_height;
+	unsigned int pte_row_height_linear;
+};
+
+struct _vcs_dpi_display_rq_regs_st {
+	display_data_rq_regs_st rq_regs_l;
+	display_data_rq_regs_st rq_regs_c;
+	unsigned int drq_expansion_mode;
+	unsigned int prq_expansion_mode;
+	unsigned int mrq_expansion_mode;
+	unsigned int crq_expansion_mode;
+	unsigned int plane1_base_address;
+};
+
+struct _vcs_dpi_display_dlg_sys_params_st {
+	double t_mclk_wm_us;
+	double t_urg_wm_us;
+	double t_sr_wm_us;
+	double t_extra_us;
+	double mem_trip_us;
+	double t_srx_delay_us;
+	double deepsleep_dcfclk_mhz;
+	double total_flip_bw;
+	unsigned int total_flip_bytes;
+};
+
+struct _vcs_dpi_display_dlg_prefetch_param_st {
+	double prefetch_bw;
+	unsigned int flip_bytes;
+};
+
+struct _vcs_dpi_display_pipe_clock_st {
+	double dcfclk_mhz;
+	double dispclk_mhz;
+	double socclk_mhz;
+	double dscclk_mhz[6];
+	double dppclk_mhz[6];
+};
+
+struct _vcs_dpi_display_arb_params_st {
+	int max_req_outstanding;
+	int min_req_outstanding;
+	int sat_level_us;
 };
 
 #endif /*__DISPLAY_MODE_STRUCTS_H__*/
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dml_inline_defs.h b/drivers/gpu/drm/amd/display/dc/dml/dml_inline_defs.h
index f9cf083..e8ce085 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dml_inline_defs.h
+++ b/drivers/gpu/drm/amd/display/dc/dml/dml_inline_defs.h
@@ -35,6 +35,16 @@ static inline double dml_min(double a, double b)
 	return (double) dcn_bw_min2(a, b);
 }
 
+static inline double dml_min3(double a, double b, double c)
+{
+	return dml_min(dml_min(a, b), c);
+}
+
+static inline double dml_min4(double a, double b, double c, double d)
+{
+	return dml_min(dml_min(a, b), dml_min(c, d));
+}
+
 static inline double dml_max(double a, double b)
 {
 	return (double) dcn_bw_max2(a, b);
-- 
cgit v1.1


From 66dec27a987bfcd2572bfc7520826b11340d264f Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Tue, 24 Apr 2018 15:21:33 -0400
Subject: drm/amd/display: Fix up dm logging functionality

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c  |  5 -----
 .../gpu/drm/amd/display/dc/basics/log_helpers.c    |  1 -
 drivers/gpu/drm/amd/display/dc/basics/logger.c     |  1 +
 drivers/gpu/drm/amd/display/dc/dm_services.h       |  4 ----
 drivers/gpu/drm/amd/display/modules/stats/stats.c  | 24 ++++++++++++++--------
 5 files changed, 17 insertions(+), 18 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
index ca0b08b..bd44935 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
@@ -330,11 +330,6 @@ bool dm_helpers_dp_mst_send_payload_allocation(
 	return true;
 }
 
-bool dm_helpers_dc_conn_log(struct dc_context *ctx, struct log_entry *entry, enum dc_log_type event)
-{
-	return true;
-}
-
 void dm_dtn_log_begin(struct dc_context *ctx)
 {}
 
diff --git a/drivers/gpu/drm/amd/display/dc/basics/log_helpers.c b/drivers/gpu/drm/amd/display/dc/basics/log_helpers.c
index 854678a..0214515 100644
--- a/drivers/gpu/drm/amd/display/dc/basics/log_helpers.c
+++ b/drivers/gpu/drm/amd/display/dc/basics/log_helpers.c
@@ -94,7 +94,6 @@ void dc_conn_log(struct dc_context *ctx,
 			dm_logger_append(&entry, "%2.2X ", hex_data[i]);
 
 	dm_logger_append(&entry, "^\n");
-	dm_helpers_dc_conn_log(ctx, &entry, event);
 
 fail:
 	dm_logger_close(&entry);
diff --git a/drivers/gpu/drm/amd/display/dc/basics/logger.c b/drivers/gpu/drm/amd/display/dc/basics/logger.c
index 0001a3c..738a818 100644
--- a/drivers/gpu/drm/amd/display/dc/basics/logger.c
+++ b/drivers/gpu/drm/amd/display/dc/basics/logger.c
@@ -402,3 +402,4 @@ cleanup:
 		entry->max_buf_bytes = 0;
 	}
 }
+
diff --git a/drivers/gpu/drm/amd/display/dc/dm_services.h b/drivers/gpu/drm/amd/display/dc/dm_services.h
index 8eafe1a..4ff9b2b 100644
--- a/drivers/gpu/drm/amd/display/dc/dm_services.h
+++ b/drivers/gpu/drm/amd/display/dc/dm_services.h
@@ -355,10 +355,6 @@ void dm_perf_trace_timestamp(const char *func_name, unsigned int line);
 /*
  * Debug and verification hooks
  */
-bool dm_helpers_dc_conn_log(
-		struct dc_context *ctx,
-		struct log_entry *entry,
-		enum dc_log_type event);
 
 void dm_dtn_log_begin(struct dc_context *ctx);
 void dm_dtn_log_append_v(struct dc_context *ctx, const char *msg, ...);
diff --git a/drivers/gpu/drm/amd/display/modules/stats/stats.c b/drivers/gpu/drm/amd/display/modules/stats/stats.c
index d16aac7..ae2d92b 100644
--- a/drivers/gpu/drm/amd/display/modules/stats/stats.c
+++ b/drivers/gpu/drm/amd/display/modules/stats/stats.c
@@ -168,6 +168,7 @@ void mod_stats_dump(struct mod_stats *mod_stats)
 	struct core_stats *core_stats = NULL;
 	struct stats_time_cache *time = NULL;
 	unsigned int index = 0;
+	struct log_entry log_entry;
 
 	if (mod_stats == NULL)
 		return;
@@ -177,17 +178,22 @@ void mod_stats_dump(struct mod_stats *mod_stats)
 	logger = dc->ctx->logger;
 	time = core_stats->time;
 
-	dm_logger_write(logger, LOG_DISPLAYSTATS, "==Display Caps==");
-	dm_logger_write(logger, LOG_DISPLAYSTATS, " ");
+	dm_logger_open(
+		dc->ctx->logger,
+		&log_entry,
+		LOG_DISPLAYSTATS);
 
-	dm_logger_write(logger, LOG_DISPLAYSTATS, "==Display Stats==");
-	dm_logger_write(logger, LOG_DISPLAYSTATS, " ");
+	dm_logger_append(&log_entry, "==Display Caps==\n");
+	dm_logger_append(&log_entry, "\n");
 
-	dm_logger_write(logger, LOG_DISPLAYSTATS,
+	dm_logger_append(&log_entry, "==Display Stats==\n");
+	dm_logger_append(&log_entry, "\n");
+
+	dm_logger_append(&log_entry,
 		"%10s %10s %10s %10s %10s"
 			" %11s %11s %17s %10s %14s"
 			" %10s %10s %10s %10s %10s"
-			" %10s %10s %10s %10s",
+			" %10s %10s %10s %10s\n",
 		"render", "avgRender",
 		"minWindow", "midPoint", "maxWindow",
 		"vsyncToFlip", "flipToVsync", "vsyncsBetweenFlip",
@@ -197,11 +203,11 @@ void mod_stats_dump(struct mod_stats *mod_stats)
 		"vSyncTime4", "vSyncTime5", "flags");
 
 	for (int i = 0; i < core_stats->index && i < core_stats->entries; i++) {
-		dm_logger_write(logger, LOG_DISPLAYSTATS,
+		dm_logger_append(&log_entry,
 			"%10u %10u %10u %10u %10u"
 				" %11u %11u %17u %10u %14u"
 				" %10u %10u %10u %10u %10u"
-				" %10u %10u %10u %10u",
+				" %10u %10u %10u %10u\n",
 			time[i].render_time_in_us,
 			time[i].avg_render_time_in_us_last_ten,
 			time[i].min_window,
@@ -222,6 +228,8 @@ void mod_stats_dump(struct mod_stats *mod_stats)
 			time[i].v_sync_time_in_us[4],
 			time[i].flags);
 	}
+
+	dm_logger_close(&log_entry);
 }
 
 void mod_stats_reset_data(struct mod_stats *mod_stats)
-- 
cgit v1.1


From 5103c5688518ea16c7f2f864b784c1266cd13c89 Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Tue, 24 Apr 2018 15:36:27 -0400
Subject: drm/amd/display: use macro for logs

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../gpu/drm/amd/display/include/logger_interface.h    |  9 +++++++++
 drivers/gpu/drm/amd/display/modules/stats/stats.c     | 19 ++++++-------------
 2 files changed, 15 insertions(+), 13 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/include/logger_interface.h b/drivers/gpu/drm/amd/display/include/logger_interface.h
index 28dee96..dc98d6d 100644
--- a/drivers/gpu/drm/amd/display/include/logger_interface.h
+++ b/drivers/gpu/drm/amd/display/include/logger_interface.h
@@ -190,4 +190,13 @@ void context_clock_trace(
 	} \
 } while (0)
 
+#define DISPLAY_STATS_BEGIN(entry) \
+	dm_logger_open(dc->ctx->logger, &entry, LOG_DISPLAYSTATS)
+
+#define DISPLAY_STATS(msg, ...) \
+	dm_logger_append(&log_entry, msg, ##__VA_ARGS__)
+
+#define DISPLAY_STATS_END(entry) \
+	dm_logger_close(&entry)
+
 #endif /* __DAL_LOGGER_INTERFACE_H__ */
diff --git a/drivers/gpu/drm/amd/display/modules/stats/stats.c b/drivers/gpu/drm/amd/display/modules/stats/stats.c
index ae2d92b..45acdbc 100644
--- a/drivers/gpu/drm/amd/display/modules/stats/stats.c
+++ b/drivers/gpu/drm/amd/display/modules/stats/stats.c
@@ -178,19 +178,13 @@ void mod_stats_dump(struct mod_stats *mod_stats)
 	logger = dc->ctx->logger;
 	time = core_stats->time;
 
-	dm_logger_open(
-		dc->ctx->logger,
-		&log_entry,
-		LOG_DISPLAYSTATS);
+	DISPLAY_STATS_BEGIN(log_entry);
 
-	dm_logger_append(&log_entry, "==Display Caps==\n");
-	dm_logger_append(&log_entry, "\n");
+	DISPLAY_STATS("==Display Caps==\n");
 
-	dm_logger_append(&log_entry, "==Display Stats==\n");
-	dm_logger_append(&log_entry, "\n");
+	DISPLAY_STATS("==Display Stats==\n");
 
-	dm_logger_append(&log_entry,
-		"%10s %10s %10s %10s %10s"
+	DISPLAY_STATS("%10s %10s %10s %10s %10s"
 			" %11s %11s %17s %10s %14s"
 			" %10s %10s %10s %10s %10s"
 			" %10s %10s %10s %10s\n",
@@ -203,8 +197,7 @@ void mod_stats_dump(struct mod_stats *mod_stats)
 		"vSyncTime4", "vSyncTime5", "flags");
 
 	for (int i = 0; i < core_stats->index && i < core_stats->entries; i++) {
-		dm_logger_append(&log_entry,
-			"%10u %10u %10u %10u %10u"
+		DISPLAY_STATS("%10u %10u %10u %10u %10u"
 				" %11u %11u %17u %10u %14u"
 				" %10u %10u %10u %10u %10u"
 				" %10u %10u %10u %10u\n",
@@ -229,7 +222,7 @@ void mod_stats_dump(struct mod_stats *mod_stats)
 			time[i].flags);
 	}
 
-	dm_logger_close(&log_entry);
+	DISPLAY_STATS_END(log_entry);
 }
 
 void mod_stats_reset_data(struct mod_stats *mod_stats)
-- 
cgit v1.1


From eb815442e840e436108ae4112fa80fc2e7ff47f3 Mon Sep 17 00:00:00 2001
From: Samson Tam <Samson.Tam@amd.com>
Date: Fri, 13 Apr 2018 18:38:56 -0400
Subject: drm/amd/display: don't create new dc_sink if nothing changed at
 detection

Signed-off-by: Samson Tam <Samson.Tam@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_link.c | 95 ++++++++++++++++++++++-----
 1 file changed, 77 insertions(+), 18 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
index ea5d5ff..2fa5218 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
@@ -469,6 +469,13 @@ static void link_disconnect_sink(struct dc_link *link)
 	link->dpcd_sink_count = 0;
 }
 
+static void link_disconnect_remap(struct dc_sink *prev_sink, struct dc_link *link)
+{
+	dc_sink_release(link->local_sink);
+	link->local_sink = prev_sink;
+}
+
+
 static bool detect_dp(
 	struct dc_link *link,
 	struct display_sink_capability *sink_caps,
@@ -551,6 +558,17 @@ static bool detect_dp(
 	return true;
 }
 
+static bool is_same_edid(struct dc_edid *old_edid, struct dc_edid *new_edid)
+{
+	if (old_edid->length != new_edid->length)
+		return false;
+
+	if (new_edid->length == 0)
+		return false;
+
+	return (memcmp(old_edid->raw_edid, new_edid->raw_edid, new_edid->length) == 0);
+}
+
 bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 {
 	struct dc_sink_init_data sink_init_data = { 0 };
@@ -558,9 +576,13 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 	uint8_t i;
 	bool converter_disable_audio = false;
 	struct audio_support *aud_support = &link->dc->res_pool->audio_support;
+	bool same_edid = false;
 	enum dc_edid_status edid_status;
 	struct dc_context *dc_ctx = link->ctx;
 	struct dc_sink *sink = NULL;
+	struct dc_sink *prev_sink = NULL;
+	struct dpcd_caps prev_dpcd_caps;
+	bool same_dpcd = true;
 	enum dc_connection_type new_connection_type = dc_connection_none;
 	DC_LOGGER_INIT(link->ctx->logger);
 	if (link->connector_signal == SIGNAL_TYPE_VIRTUAL)
@@ -575,6 +597,11 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 			link->local_sink)
 		return true;
 
+	prev_sink = link->local_sink;
+	if (prev_sink != NULL) {
+		dc_sink_retain(prev_sink);
+		memcpy(&prev_dpcd_caps, &link->dpcd_caps, sizeof(struct dpcd_caps));
+	}
 	link_disconnect_sink(link);
 
 	if (new_connection_type != dc_connection_none) {
@@ -616,14 +643,25 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 				link,
 				&sink_caps,
 				&converter_disable_audio,
-				aud_support, reason))
+				aud_support, reason)) {
+				if (prev_sink != NULL)
+					dc_sink_release(prev_sink);
 				return false;
+			}
 
+			// Check if dpcp block is the same
+			if (prev_sink != NULL) {
+				if (memcmp(&link->dpcd_caps, &prev_dpcd_caps, sizeof(struct dpcd_caps)))
+					same_dpcd = false;
+			}
 			/* Active dongle downstream unplug */
 			if (link->type == dc_connection_active_dongle
 					&& link->dpcd_caps.sink_count.
-					bits.SINK_COUNT == 0)
+					bits.SINK_COUNT == 0) {
+				if (prev_sink != NULL)
+					dc_sink_release(prev_sink);
 				return true;
+			}
 
 			if (link->type == dc_connection_mst_branch) {
 				LINK_INFO("link=%d, mst branch is now Connected\n",
@@ -634,6 +672,8 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 				 * pbn_per_slot value leading to exception on dc_fixpt_div()
 				 */
 				link->verified_link_cap = link->reported_link_cap;
+				if (prev_sink != NULL)
+					dc_sink_release(prev_sink);
 				return false;
 			}
 
@@ -643,6 +683,8 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 		default:
 			DC_ERROR("Invalid connector type! signal:%d\n",
 				link->connector_signal);
+			if (prev_sink != NULL)
+				dc_sink_release(prev_sink);
 			return false;
 		} /* switch() */
 
@@ -665,6 +707,8 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 		sink = dc_sink_create(&sink_init_data);
 		if (!sink) {
 			DC_ERROR("Failed to create sink!\n");
+			if (prev_sink != NULL)
+				dc_sink_release(prev_sink);
 			return false;
 		}
 
@@ -688,22 +732,33 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 			break;
 		}
 
-		if (link->connector_signal == SIGNAL_TYPE_DISPLAY_PORT &&
-			sink_caps.transaction_type ==
-			DDC_TRANSACTION_TYPE_I2C_OVER_AUX) {
-			/*
-			 * TODO debug why Dell 2413 doesn't like
-			 *  two link trainings
-			 */
+		// Check if edid is the same
+		if ((prev_sink != NULL) && ((edid_status == EDID_THE_SAME) || (edid_status == EDID_OK)))
+			same_edid = is_same_edid(&prev_sink->dc_edid, &sink->dc_edid);
 
-			/* deal with non-mst cases */
-			dp_hbr_verify_link_cap(link, &link->reported_link_cap);
-		}
+		// If both edid and dpcd are the same, then discard new sink and revert back to original sink
+		if ((same_edid) && (same_dpcd)) {
+			link_disconnect_remap(prev_sink, link);
+			sink = prev_sink;
+			prev_sink = NULL;
+		} else {
+			if (link->connector_signal == SIGNAL_TYPE_DISPLAY_PORT &&
+					sink_caps.transaction_type ==
+						DDC_TRANSACTION_TYPE_I2C_OVER_AUX) {
+				/*
+				 * TODO debug why Dell 2413 doesn't like
+				 *  two link trainings
+				 */
+
+				/* deal with non-mst cases */
+				dp_hbr_verify_link_cap(link, &link->reported_link_cap);
+			}
 
-		/* HDMI-DVI Dongle */
-		if (sink->sink_signal == SIGNAL_TYPE_HDMI_TYPE_A &&
-				!sink->edid_caps.edid_hdmi)
-			sink->sink_signal = SIGNAL_TYPE_DVI_SINGLE_LINK;
+			/* HDMI-DVI Dongle */
+			if (sink->sink_signal == SIGNAL_TYPE_HDMI_TYPE_A &&
+					!sink->edid_caps.edid_hdmi)
+				sink->sink_signal = SIGNAL_TYPE_DVI_SINGLE_LINK;
+		}
 
 		/* Connectivity log: detection */
 		for (i = 0; i < sink->dc_edid.length / EDID_BLOCK_SIZE; i++) {
@@ -762,10 +817,14 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 		sink_caps.signal = SIGNAL_TYPE_NONE;
 	}
 
-	LINK_INFO("link=%d, dc_sink_in=%p is now %s\n",
+	LINK_INFO("link=%d, dc_sink_in=%p is now %s prev_sink=%p dpcd same=%d edid same=%d\n",
 		link->link_index, sink,
 		(sink_caps.signal == SIGNAL_TYPE_NONE ?
-			"Disconnected":"Connected"));
+			"Disconnected":"Connected"), prev_sink,
+			same_dpcd, same_edid);
+
+	if (prev_sink != NULL)
+		dc_sink_release(prev_sink);
 
 	return true;
 }
-- 
cgit v1.1


From 8fc06ebc2bb719cddb041bcb14b5ca87adbcd57f Mon Sep 17 00:00:00 2001
From: Xingyue Tao <xingyue.tao@amd.com>
Date: Thu, 19 Apr 2018 16:23:12 -0400
Subject: drm/amd/display: Only limit VSR downscaling when actually downscaling

Signed-off-by: Xingyue Tao <xingyue.tao@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c | 21 ++++++++++-----------
 1 file changed, 10 insertions(+), 11 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
index 2da1389..46a35c7 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
@@ -145,18 +145,17 @@ bool dpp_get_optimal_number_of_taps(
 	else
 		pixel_width = scl_data->viewport.width;
 
+	/* Some ASICs does not support  FP16 scaling, so we reject modes require this*/
 	if (scl_data->viewport.width  != scl_data->h_active &&
-		scl_data->viewport.height != scl_data->v_active) {
-
-		/* Some ASICs does not support  FP16 scaling, so we reject modes require this*/
-		if (dpp->caps->dscl_data_proc_format == DSCL_DATA_PRCESSING_FIXED_FORMAT &&
-			scl_data->format == PIXEL_FORMAT_FP16)
-			return false;
-
-		if (dpp->ctx->dc->debug.max_downscale_src_width != 0 &&
-			scl_data->viewport.width > dpp->ctx->dc->debug.max_downscale_src_width)
-			return false;
-	}
+		scl_data->viewport.height != scl_data->v_active &&
+		dpp->caps->dscl_data_proc_format == DSCL_DATA_PRCESSING_FIXED_FORMAT &&
+		scl_data->format == PIXEL_FORMAT_FP16)
+		return false;
+
+	if (scl_data->viewport.width > scl_data->h_active &&
+		dpp->ctx->dc->debug.max_downscale_src_width != 0 &&
+		scl_data->viewport.width > dpp->ctx->dc->debug.max_downscale_src_width)
+		return false;
 
 	/* TODO: add lb check */
 
-- 
cgit v1.1


From 109ece8d43cdb491a968b3690e947e27225f886e Mon Sep 17 00:00:00 2001
From: Jun Lei <Jun.Lei@amd.com>
Date: Thu, 26 Apr 2018 10:24:25 -0400
Subject: drm/amd/display: constify a few dc_surface_update fields

Signed-off-by: Jun Lei <Jun.Lei@amd.com>
Reviewed-by: Jun Lei <Jun.Lei@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dc.h | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index 7a9f600..9cfde0c 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -503,18 +503,18 @@ struct dc_surface_update {
 	struct dc_plane_state *surface;
 
 	/* isr safe update parameters.  null means no updates */
-	struct dc_flip_addrs *flip_addr;
-	struct dc_plane_info *plane_info;
-	struct dc_scaling_info *scaling_info;
+	const struct dc_flip_addrs *flip_addr;
+	const struct dc_plane_info *plane_info;
+	const struct dc_scaling_info *scaling_info;
 
 	/* following updates require alloc/sleep/spin that is not isr safe,
 	 * null means no updates
 	 */
-	struct dc_gamma *gamma;
-	struct dc_transfer_func *in_transfer_func;
+	const struct dc_gamma *gamma;
+	const struct dc_transfer_func *in_transfer_func;
 
-	struct dc_csc_transform *input_csc_color_matrix;
-	struct fixed31_32 *coeff_reduction_factor;
+	const struct dc_csc_transform *input_csc_color_matrix;
+	const struct fixed31_32 *coeff_reduction_factor;
 };
 
 /*
-- 
cgit v1.1


From 6474b2824d71ac6cd1005aff8841dd8bcfa0901d Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Thu, 26 Apr 2018 10:03:44 -0400
Subject: drm/amd/display: Add fullscreen transitions to log

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../gpu/drm/amd/display/modules/inc/mod_stats.h    |   4 +
 drivers/gpu/drm/amd/display/modules/stats/stats.c  | 137 +++++++++++++++++----
 2 files changed, 114 insertions(+), 27 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/modules/inc/mod_stats.h b/drivers/gpu/drm/amd/display/modules/inc/mod_stats.h
index 3230e2a..3812094 100644
--- a/drivers/gpu/drm/amd/display/modules/inc/mod_stats.h
+++ b/drivers/gpu/drm/amd/display/modules/inc/mod_stats.h
@@ -46,6 +46,10 @@ void mod_stats_dump(struct mod_stats *mod_stats);
 
 void mod_stats_reset_data(struct mod_stats *mod_stats);
 
+void mod_stats_update_event(struct mod_stats *mod_stats,
+		char *event_string,
+		unsigned int length);
+
 void mod_stats_update_flip(struct mod_stats *mod_stats,
 		unsigned long timestamp_in_ns);
 
diff --git a/drivers/gpu/drm/amd/display/modules/stats/stats.c b/drivers/gpu/drm/amd/display/modules/stats/stats.c
index 45acdbc..4b00bae 100644
--- a/drivers/gpu/drm/amd/display/modules/stats/stats.c
+++ b/drivers/gpu/drm/amd/display/modules/stats/stats.c
@@ -36,9 +36,14 @@
 #define DAL_STATS_ENTRIES_REGKEY_DEFAULT	0x00350000
 #define DAL_STATS_ENTRIES_REGKEY_MAX		0x01000000
 
+#define DAL_STATS_EVENT_ENTRIES_DEFAULT		0x00000100
+
 #define MOD_STATS_NUM_VSYNCS			5
+#define MOD_STATS_EVENT_STRING_MAX		512
 
 struct stats_time_cache {
+	unsigned int entry_id;
+
 	unsigned long flip_timestamp_in_ns;
 	unsigned long vupdate_timestamp_in_ns;
 
@@ -63,15 +68,26 @@ struct stats_time_cache {
 	unsigned int flags;
 };
 
+struct stats_event_cache {
+	unsigned int entry_id;
+	char event_string[MOD_STATS_EVENT_STRING_MAX];
+};
+
 struct core_stats {
 	struct mod_stats public;
 	struct dc *dc;
 
+	bool enabled;
+	unsigned int entries;
+	unsigned int event_entries;
+	unsigned int entry_id;
+
 	struct stats_time_cache *time;
 	unsigned int index;
 
-	bool enabled;
-	unsigned int entries;
+	struct stats_event_cache *events;
+	unsigned int event_index;
+
 };
 
 #define MOD_STATS_TO_CORE(mod_stats)\
@@ -125,9 +141,18 @@ struct mod_stats *mod_stats_create(struct dc *dc)
 			else
 				core_stats->entries = reg_data;
 		}
+		core_stats->time = kzalloc(
+			sizeof(struct stats_time_cache) *
+				core_stats->entries,
+						GFP_KERNEL);
 
-		core_stats->time = kzalloc(sizeof(struct stats_time_cache) * core_stats->entries,
+
+		core_stats->event_entries = DAL_STATS_EVENT_ENTRIES_DEFAULT;
+		core_stats->events = kzalloc(
+			sizeof(struct stats_event_cache) *
+				core_stats->event_entries,
 						GFP_KERNEL);
+
 	} else {
 		core_stats->entries = 0;
 	}
@@ -139,6 +164,10 @@ struct mod_stats *mod_stats_create(struct dc *dc)
 	 * handle calculation cases that depend on previous flip data.
 	 */
 	core_stats->index = 1;
+	core_stats->event_index = 0;
+
+	// Keeps track of ordering within the different stats structures
+	core_stats->entry_id = 0;
 
 	return &core_stats->public;
 
@@ -167,6 +196,9 @@ void mod_stats_dump(struct mod_stats *mod_stats)
 	struct dal_logger *logger = NULL;
 	struct core_stats *core_stats = NULL;
 	struct stats_time_cache *time = NULL;
+	struct stats_event_cache *events = NULL;
+	unsigned int time_index = 1;
+	unsigned int event_index = 0;
 	unsigned int index = 0;
 	struct log_entry log_entry;
 
@@ -177,6 +209,7 @@ void mod_stats_dump(struct mod_stats *mod_stats)
 	dc = core_stats->dc;
 	logger = dc->ctx->logger;
 	time = core_stats->time;
+	events = core_stats->events;
 
 	DISPLAY_STATS_BEGIN(log_entry);
 
@@ -196,30 +229,39 @@ void mod_stats_dump(struct mod_stats *mod_stats)
 		"vSyncTime1", "vSyncTime2", "vSyncTime3",
 		"vSyncTime4", "vSyncTime5", "flags");
 
-	for (int i = 0; i < core_stats->index && i < core_stats->entries; i++) {
-		DISPLAY_STATS("%10u %10u %10u %10u %10u"
-				" %11u %11u %17u %10u %14u"
-				" %10u %10u %10u %10u %10u"
-				" %10u %10u %10u %10u\n",
-			time[i].render_time_in_us,
-			time[i].avg_render_time_in_us_last_ten,
-			time[i].min_window,
-			time[i].lfc_mid_point_in_us,
-			time[i].max_window,
-			time[i].vsync_to_flip_time_in_us,
-			time[i].flip_to_vsync_time_in_us,
-			time[i].num_vsync_between_flips,
-			time[i].num_frames_inserted,
-			time[i].inserted_duration_in_us,
-			time[i].v_total_min,
-			time[i].v_total_max,
-			time[i].event_triggers,
-			time[i].v_sync_time_in_us[0],
-			time[i].v_sync_time_in_us[1],
-			time[i].v_sync_time_in_us[2],
-			time[i].v_sync_time_in_us[3],
-			time[i].v_sync_time_in_us[4],
-			time[i].flags);
+	for (int i = 0; i < core_stats->entry_id; i++) {
+		if (event_index < core_stats->event_index &&
+				i == events[event_index].entry_id) {
+			DISPLAY_STATS("%s\n", events[event_index].event_string);
+			event_index++;
+		} else if (time_index < core_stats->index &&
+				i == time[time_index].entry_id) {
+			DISPLAY_STATS("%10u %10u %10u %10u %10u"
+					" %11u %11u %17u %10u %14u"
+					" %10u %10u %10u %10u %10u"
+					" %10u %10u %10u %10u\n",
+				time[time_index].render_time_in_us,
+				time[time_index].avg_render_time_in_us_last_ten,
+				time[time_index].min_window,
+				time[time_index].lfc_mid_point_in_us,
+				time[time_index].max_window,
+				time[time_index].vsync_to_flip_time_in_us,
+				time[time_index].flip_to_vsync_time_in_us,
+				time[time_index].num_vsync_between_flips,
+				time[time_index].num_frames_inserted,
+				time[time_index].inserted_duration_in_us,
+				time[time_index].v_total_min,
+				time[time_index].v_total_max,
+				time[time_index].event_triggers,
+				time[time_index].v_sync_time_in_us[0],
+				time[time_index].v_sync_time_in_us[1],
+				time[time_index].v_sync_time_in_us[2],
+				time[time_index].v_sync_time_in_us[3],
+				time[time_index].v_sync_time_in_us[4],
+				time[time_index].flags);
+
+			time_index++;
+		}
 	}
 
 	DISPLAY_STATS_END(log_entry);
@@ -239,7 +281,46 @@ void mod_stats_reset_data(struct mod_stats *mod_stats)
 	memset(core_stats->time, 0,
 		sizeof(struct stats_time_cache) * core_stats->entries);
 
+	memset(core_stats->events, 0,
+		sizeof(struct stats_event_cache) * core_stats->event_entries);
+
 	core_stats->index = 1;
+	core_stats->event_index = 0;
+
+	// Keeps track of ordering within the different stats structures
+	core_stats->entry_id = 0;
+}
+
+void mod_stats_update_event(struct mod_stats *mod_stats,
+		char *event_string,
+		unsigned int length)
+{
+	struct core_stats *core_stats = NULL;
+	struct stats_event_cache *events = NULL;
+	unsigned int index = 0;
+	unsigned int copy_length = 0;
+
+	if (mod_stats == NULL)
+		return;
+
+	core_stats = MOD_STATS_TO_CORE(mod_stats);
+
+	if (core_stats->index >= core_stats->entries)
+		return;
+
+	events = core_stats->events;
+	index = core_stats->event_index;
+
+	copy_length = length;
+	if (length > MOD_STATS_EVENT_STRING_MAX)
+		copy_length = MOD_STATS_EVENT_STRING_MAX;
+
+	memcpy(&events[index].event_string, event_string, copy_length);
+	events[index].event_string[copy_length - 1] = '\0';
+
+	events[index].entry_id = core_stats->entry_id;
+	core_stats->event_index++;
+	core_stats->entry_id++;
 }
 
 void mod_stats_update_flip(struct mod_stats *mod_stats,
@@ -280,7 +361,9 @@ void mod_stats_update_flip(struct mod_stats *mod_stats,
 			(timestamp_in_ns -
 				time[index - 1].vupdate_timestamp_in_ns) / 1000;
 
+	time[index].entry_id = core_stats->entry_id;
 	core_stats->index++;
+	core_stats->entry_id++;
 }
 
 void mod_stats_update_vupdate(struct mod_stats *mod_stats,
-- 
cgit v1.1


From dab911d535ae24a39b4e383c0dffaa3e5583883d Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Fri, 27 Apr 2018 15:23:23 -0400
Subject: drm/amd/display: fix bug with index check

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/modules/stats/stats.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/modules/stats/stats.c b/drivers/gpu/drm/amd/display/modules/stats/stats.c
index 4b00bae..fe9e4b3 100644
--- a/drivers/gpu/drm/amd/display/modules/stats/stats.c
+++ b/drivers/gpu/drm/amd/display/modules/stats/stats.c
@@ -305,7 +305,7 @@ void mod_stats_update_event(struct mod_stats *mod_stats,
 
 	core_stats = MOD_STATS_TO_CORE(mod_stats);
 
-	if (core_stats->index >= core_stats->entries)
+	if (core_stats->event_index >= core_stats->event_entries)
 		return;
 
 	events = core_stats->events;
-- 
cgit v1.1


From a944744ba517256fcc9311e12c083563cbbe7c88 Mon Sep 17 00:00:00 2001
From: Nikola Cornij <nikola.cornij@amd.com>
Date: Fri, 27 Apr 2018 17:26:25 -0400
Subject: drm/amd/display: Clear underflow status for debug purposes

We want to keep underflow sticky bit on for the longevity tests outside
of test environment. For debug purposes it is, however, useful to clear
underflow status after the test that caused it so that the following
tests are not affected. This change fullfils both requirements by clearing
the underflow only from within Windows or Diags test environment.

Signed-off-by: Nikola Cornij <nikola.cornij@amd.com>
Reviewed-by: Nikola Cornij <Nikola.Cornij@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
index c452972..f8e0576 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -326,6 +326,12 @@ void dcn10_log_hw_state(struct dc *dc)
 				s.h_total,
 				s.v_total,
 				s.underflow_occurred_status);
+
+		// Clear underflow for debug purposes
+		// We want to keep underflow sticky bit on for the longevity tests outside of test environment.
+		// This function is called only from Windows or Diags test environment, hence it's safe to clear
+		// it from here without affecting the original intent.
+		tg->funcs->clear_optc_underflow(tg);
 	}
 	DTN_INFO("\n");
 
-- 
cgit v1.1


From f0cd0a346dfd1df4b691fe38dafb51911392fbce Mon Sep 17 00:00:00 2001
From: Eric Bernstein <eric.bernstein@amd.com>
Date: Thu, 26 Apr 2018 14:06:00 -0400
Subject: drm/amd/display: DCN1 link encoder

Create DCN1 link encoder files and update AUX and HPD register access.

Signed-off-by: Eric Bernstein <eric.bernstein@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_link_hwss.c |    2 -
 drivers/gpu/drm/amd/display/dc/dcn10/Makefile      |    2 +-
 .../drm/amd/display/dc/dcn10/dcn10_link_encoder.c  | 1362 ++++++++++++++++++++
 .../drm/amd/display/dc/dcn10/dcn10_link_encoder.h  |  330 +++++
 .../gpu/drm/amd/display/dc/dcn10/dcn10_resource.c  |   43 +-
 5 files changed, 1716 insertions(+), 23 deletions(-)
 create mode 100644 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.c
 create mode 100644 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.h

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_hwss.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_hwss.c
index 7c866a7..82cd1d6 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link_hwss.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_hwss.c
@@ -11,8 +11,6 @@
 #include "dc_link_dp.h"
 #include "dc_link_ddc.h"
 #include "dm_helpers.h"
-#include "dce/dce_link_encoder.h"
-#include "dce/dce_stream_encoder.h"
 #include "dpcd_defs.h"
 
 enum dc_status core_link_read_dpcd(
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/Makefile b/drivers/gpu/drm/amd/display/dc/dcn10/Makefile
index 5c69743..84f52c6 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/Makefile
@@ -26,7 +26,7 @@ DCN10 = dcn10_resource.o dcn10_ipp.o dcn10_hw_sequencer.o \
 		dcn10_dpp.o dcn10_opp.o dcn10_optc.o \
 		dcn10_hubp.o dcn10_mpc.o \
 		dcn10_dpp_dscl.o dcn10_dpp_cm.o dcn10_cm_common.o \
-		dcn10_hubbub.o dcn10_stream_encoder.o
+		dcn10_hubbub.o dcn10_stream_encoder.o dcn10_link_encoder.o
 
 AMD_DAL_DCN10 = $(addprefix $(AMDDALPATH)/dc/dcn10/,$(DCN10))
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.c
new file mode 100644
index 0000000..21fa40a
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.c
@@ -0,0 +1,1362 @@
+/*
+ * Copyright 2012-15 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#include "reg_helper.h"
+
+#include "core_types.h"
+#include "link_encoder.h"
+#include "dcn10_link_encoder.h"
+#include "stream_encoder.h"
+#include "i2caux_interface.h"
+#include "dc_bios_types.h"
+
+#include "gpio_service_interface.h"
+
+#define CTX \
+	enc10->base.ctx
+#define DC_LOGGER \
+	enc10->base.ctx->logger
+
+#define REG(reg)\
+	(enc10->link_regs->reg)
+
+#undef FN
+#define FN(reg_name, field_name) \
+	enc10->link_shift->field_name, enc10->link_mask->field_name
+
+
+/*
+ * @brief
+ * Trigger Source Select
+ * ASIC-dependent, actual values for register programming
+ */
+#define DCN10_DIG_FE_SOURCE_SELECT_INVALID 0x0
+#define DCN10_DIG_FE_SOURCE_SELECT_DIGA 0x1
+#define DCN10_DIG_FE_SOURCE_SELECT_DIGB 0x2
+#define DCN10_DIG_FE_SOURCE_SELECT_DIGC 0x4
+#define DCN10_DIG_FE_SOURCE_SELECT_DIGD 0x08
+#define DCN10_DIG_FE_SOURCE_SELECT_DIGE 0x10
+#define DCN10_DIG_FE_SOURCE_SELECT_DIGF 0x20
+#define DCN10_DIG_FE_SOURCE_SELECT_DIGG 0x40
+
+enum {
+	DP_MST_UPDATE_MAX_RETRY = 50
+};
+
+
+
+static void aux_initialize(struct dcn10_link_encoder *enc10);
+
+
+static const struct link_encoder_funcs dcn10_lnk_enc_funcs = {
+	.validate_output_with_stream =
+		dcn10_link_encoder_validate_output_with_stream,
+	.hw_init = dcn10_link_encoder_hw_init,
+	.setup = dcn10_link_encoder_setup,
+	.enable_tmds_output = dcn10_link_encoder_enable_tmds_output,
+	.enable_dp_output = dcn10_link_encoder_enable_dp_output,
+	.enable_dp_mst_output = dcn10_link_encoder_enable_dp_mst_output,
+	.disable_output = dcn10_link_encoder_disable_output,
+	.dp_set_lane_settings = dcn10_link_encoder_dp_set_lane_settings,
+	.dp_set_phy_pattern = dcn10_link_encoder_dp_set_phy_pattern,
+	.update_mst_stream_allocation_table =
+		dcn10_link_encoder_update_mst_stream_allocation_table,
+	.psr_program_dp_dphy_fast_training =
+			dcn10_psr_program_dp_dphy_fast_training,
+	.psr_program_secondary_packet = dcn10_psr_program_secondary_packet,
+	.connect_dig_be_to_fe = dcn10_link_encoder_connect_dig_be_to_fe,
+	.enable_hpd = dcn10_link_encoder_enable_hpd,
+	.disable_hpd = dcn10_link_encoder_disable_hpd,
+	.is_dig_enabled = dcn10_is_dig_enabled,
+	.destroy = dcn10_link_encoder_destroy
+};
+
+static enum bp_result link_transmitter_control(
+	struct dcn10_link_encoder *enc10,
+	struct bp_transmitter_control *cntl)
+{
+	enum bp_result result;
+	struct dc_bios *bp = enc10->base.ctx->dc_bios;
+
+	result = bp->funcs->transmitter_control(bp, cntl);
+
+	return result;
+}
+
+static void enable_phy_bypass_mode(
+	struct dcn10_link_encoder *enc10,
+	bool enable)
+{
+	/* This register resides in DP back end block;
+	 * transmitter is used for the offset
+	 */
+	REG_UPDATE(DP_DPHY_CNTL, DPHY_BYPASS, enable);
+
+}
+
+static void disable_prbs_symbols(
+	struct dcn10_link_encoder *enc10,
+	bool disable)
+{
+	/* This register resides in DP back end block;
+	 * transmitter is used for the offset
+	 */
+	REG_UPDATE_4(DP_DPHY_CNTL,
+			DPHY_ATEST_SEL_LANE0, disable,
+			DPHY_ATEST_SEL_LANE1, disable,
+			DPHY_ATEST_SEL_LANE2, disable,
+			DPHY_ATEST_SEL_LANE3, disable);
+}
+
+static void disable_prbs_mode(
+	struct dcn10_link_encoder *enc10)
+{
+	REG_UPDATE(DP_DPHY_PRBS_CNTL, DPHY_PRBS_EN, 0);
+}
+
+static void program_pattern_symbols(
+	struct dcn10_link_encoder *enc10,
+	uint16_t pattern_symbols[8])
+{
+	/* This register resides in DP back end block;
+	 * transmitter is used for the offset
+	 */
+	REG_SET_3(DP_DPHY_SYM0, 0,
+			DPHY_SYM1, pattern_symbols[0],
+			DPHY_SYM2, pattern_symbols[1],
+			DPHY_SYM3, pattern_symbols[2]);
+
+	/* This register resides in DP back end block;
+	 * transmitter is used for the offset
+	 */
+	REG_SET_3(DP_DPHY_SYM1, 0,
+			DPHY_SYM4, pattern_symbols[3],
+			DPHY_SYM5, pattern_symbols[4],
+			DPHY_SYM6, pattern_symbols[5]);
+
+	/* This register resides in DP back end block;
+	 * transmitter is used for the offset
+	 */
+	REG_SET_2(DP_DPHY_SYM2, 0,
+			DPHY_SYM7, pattern_symbols[6],
+			DPHY_SYM8, pattern_symbols[7]);
+}
+
+static void set_dp_phy_pattern_d102(
+	struct dcn10_link_encoder *enc10)
+{
+	/* Disable PHY Bypass mode to setup the test pattern */
+	enable_phy_bypass_mode(enc10, false);
+
+	/* For 10-bit PRBS or debug symbols
+	 * please use the following sequence:
+	 *
+	 * Enable debug symbols on the lanes
+	 */
+	disable_prbs_symbols(enc10, true);
+
+	/* Disable PRBS mode */
+	disable_prbs_mode(enc10);
+
+	/* Program debug symbols to be output */
+	{
+		uint16_t pattern_symbols[8] = {
+			0x2AA, 0x2AA, 0x2AA, 0x2AA,
+			0x2AA, 0x2AA, 0x2AA, 0x2AA
+		};
+
+		program_pattern_symbols(enc10, pattern_symbols);
+	}
+
+	/* Enable phy bypass mode to enable the test pattern */
+
+	enable_phy_bypass_mode(enc10, true);
+}
+
+static void set_link_training_complete(
+	struct dcn10_link_encoder *enc10,
+	bool complete)
+{
+	/* This register resides in DP back end block;
+	 * transmitter is used for the offset
+	 */
+	REG_UPDATE(DP_LINK_CNTL, DP_LINK_TRAINING_COMPLETE, complete);
+
+}
+
+void dcn10_link_encoder_set_dp_phy_pattern_training_pattern(
+	struct link_encoder *enc,
+	uint32_t index)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	/* Write Training Pattern */
+
+	REG_WRITE(DP_DPHY_TRAINING_PATTERN_SEL, index);
+
+	/* Set HW Register Training Complete to false */
+
+	set_link_training_complete(enc10, false);
+
+	/* Disable PHY Bypass mode to output Training Pattern */
+
+	enable_phy_bypass_mode(enc10, false);
+
+	/* Disable PRBS mode */
+	disable_prbs_mode(enc10);
+}
+
+static void setup_panel_mode(
+	struct dcn10_link_encoder *enc10,
+	enum dp_panel_mode panel_mode)
+{
+	uint32_t value;
+
+	ASSERT(REG(DP_DPHY_INTERNAL_CTRL));
+	value = REG_READ(DP_DPHY_INTERNAL_CTRL);
+
+	switch (panel_mode) {
+	case DP_PANEL_MODE_EDP:
+		value = 0x1;
+		break;
+	case DP_PANEL_MODE_SPECIAL:
+		value = 0x11;
+		break;
+	default:
+		value = 0x0;
+		break;
+	}
+
+	REG_WRITE(DP_DPHY_INTERNAL_CTRL, value);
+}
+
+static void set_dp_phy_pattern_symbol_error(
+	struct dcn10_link_encoder *enc10)
+{
+	/* Disable PHY Bypass mode to setup the test pattern */
+	enable_phy_bypass_mode(enc10, false);
+
+	/* program correct panel mode*/
+	setup_panel_mode(enc10, DP_PANEL_MODE_DEFAULT);
+
+	/* A PRBS23 pattern is used for most DP electrical measurements. */
+
+	/* Enable PRBS symbols on the lanes */
+	disable_prbs_symbols(enc10, false);
+
+	/* For PRBS23 Set bit DPHY_PRBS_SEL=1 and Set bit DPHY_PRBS_EN=1 */
+	REG_UPDATE_2(DP_DPHY_PRBS_CNTL,
+			DPHY_PRBS_SEL, 1,
+			DPHY_PRBS_EN, 1);
+
+	/* Enable phy bypass mode to enable the test pattern */
+	enable_phy_bypass_mode(enc10, true);
+}
+
+static void set_dp_phy_pattern_prbs7(
+	struct dcn10_link_encoder *enc10)
+{
+	/* Disable PHY Bypass mode to setup the test pattern */
+	enable_phy_bypass_mode(enc10, false);
+
+	/* A PRBS7 pattern is used for most DP electrical measurements. */
+
+	/* Enable PRBS symbols on the lanes */
+	disable_prbs_symbols(enc10, false);
+
+	/* For PRBS7 Set bit DPHY_PRBS_SEL=0 and Set bit DPHY_PRBS_EN=1 */
+	REG_UPDATE_2(DP_DPHY_PRBS_CNTL,
+			DPHY_PRBS_SEL, 0,
+			DPHY_PRBS_EN, 1);
+
+	/* Enable phy bypass mode to enable the test pattern */
+	enable_phy_bypass_mode(enc10, true);
+}
+
+static void set_dp_phy_pattern_80bit_custom(
+	struct dcn10_link_encoder *enc10,
+	const uint8_t *pattern)
+{
+	/* Disable PHY Bypass mode to setup the test pattern */
+	enable_phy_bypass_mode(enc10, false);
+
+	/* Enable debug symbols on the lanes */
+
+	disable_prbs_symbols(enc10, true);
+
+	/* Enable PHY bypass mode to enable the test pattern */
+	/* TODO is it really needed ? */
+
+	enable_phy_bypass_mode(enc10, true);
+
+	/* Program 80 bit custom pattern */
+	{
+		uint16_t pattern_symbols[8];
+
+		pattern_symbols[0] =
+			((pattern[1] & 0x03) << 8) | pattern[0];
+		pattern_symbols[1] =
+			((pattern[2] & 0x0f) << 6) | ((pattern[1] >> 2) & 0x3f);
+		pattern_symbols[2] =
+			((pattern[3] & 0x3f) << 4) | ((pattern[2] >> 4) & 0x0f);
+		pattern_symbols[3] =
+			(pattern[4] << 2) | ((pattern[3] >> 6) & 0x03);
+		pattern_symbols[4] =
+			((pattern[6] & 0x03) << 8) | pattern[5];
+		pattern_symbols[5] =
+			((pattern[7] & 0x0f) << 6) | ((pattern[6] >> 2) & 0x3f);
+		pattern_symbols[6] =
+			((pattern[8] & 0x3f) << 4) | ((pattern[7] >> 4) & 0x0f);
+		pattern_symbols[7] =
+			(pattern[9] << 2) | ((pattern[8] >> 6) & 0x03);
+
+		program_pattern_symbols(enc10, pattern_symbols);
+	}
+
+	/* Enable phy bypass mode to enable the test pattern */
+
+	enable_phy_bypass_mode(enc10, true);
+}
+
+static void set_dp_phy_pattern_hbr2_compliance_cp2520_2(
+	struct dcn10_link_encoder *enc10,
+	unsigned int cp2520_pattern)
+{
+
+	/* previously there is a register DP_HBR2_EYE_PATTERN
+	 * that is enabled to get the pattern.
+	 * But it does not work with the latest spec change,
+	 * so we are programming the following registers manually.
+	 *
+	 * The following settings have been confirmed
+	 * by Nick Chorney and Sandra Liu
+	 */
+
+	/* Disable PHY Bypass mode to setup the test pattern */
+
+	enable_phy_bypass_mode(enc10, false);
+
+	/* Setup DIG encoder in DP SST mode */
+	enc10->base.funcs->setup(&enc10->base, SIGNAL_TYPE_DISPLAY_PORT);
+
+	/* ensure normal panel mode. */
+	setup_panel_mode(enc10, DP_PANEL_MODE_DEFAULT);
+
+	/* no vbid after BS (SR)
+	 * DP_LINK_FRAMING_CNTL changed history Sandra Liu
+	 * 11000260 / 11000104 / 110000FC
+	 */
+	REG_UPDATE_3(DP_LINK_FRAMING_CNTL,
+			DP_IDLE_BS_INTERVAL, 0xFC,
+			DP_VBID_DISABLE, 1,
+			DP_VID_ENHANCED_FRAME_MODE, 1);
+
+	/* swap every BS with SR */
+	REG_UPDATE(DP_DPHY_SCRAM_CNTL, DPHY_SCRAMBLER_BS_COUNT, 0);
+
+	/* select cp2520 patterns */
+	if (REG(DP_DPHY_HBR2_PATTERN_CONTROL))
+		REG_UPDATE(DP_DPHY_HBR2_PATTERN_CONTROL,
+				DP_DPHY_HBR2_PATTERN_CONTROL, cp2520_pattern);
+	else
+		/* pre-DCE11 can only generate CP2520 pattern 2 */
+		ASSERT(cp2520_pattern == 2);
+
+	/* set link training complete */
+	set_link_training_complete(enc10, true);
+
+	/* disable video stream */
+	REG_UPDATE(DP_VID_STREAM_CNTL, DP_VID_STREAM_ENABLE, 0);
+
+	/* Disable PHY Bypass mode to setup the test pattern */
+	enable_phy_bypass_mode(enc10, false);
+}
+
+static void set_dp_phy_pattern_passthrough_mode(
+	struct dcn10_link_encoder *enc10,
+	enum dp_panel_mode panel_mode)
+{
+	/* program correct panel mode */
+	setup_panel_mode(enc10, panel_mode);
+
+	/* restore LINK_FRAMING_CNTL and DPHY_SCRAMBLER_BS_COUNT
+	 * in case we were doing HBR2 compliance pattern before
+	 */
+	REG_UPDATE_3(DP_LINK_FRAMING_CNTL,
+			DP_IDLE_BS_INTERVAL, 0x2000,
+			DP_VBID_DISABLE, 0,
+			DP_VID_ENHANCED_FRAME_MODE, 1);
+
+	REG_UPDATE(DP_DPHY_SCRAM_CNTL, DPHY_SCRAMBLER_BS_COUNT, 0x1FF);
+
+	/* set link training complete */
+	set_link_training_complete(enc10, true);
+
+	/* Disable PHY Bypass mode to setup the test pattern */
+	enable_phy_bypass_mode(enc10, false);
+
+	/* Disable PRBS mode */
+	disable_prbs_mode(enc10);
+}
+
+/* return value is bit-vector */
+static uint8_t get_frontend_source(
+	enum engine_id engine)
+{
+	switch (engine) {
+	case ENGINE_ID_DIGA:
+		return DCN10_DIG_FE_SOURCE_SELECT_DIGA;
+	case ENGINE_ID_DIGB:
+		return DCN10_DIG_FE_SOURCE_SELECT_DIGB;
+	case ENGINE_ID_DIGC:
+		return DCN10_DIG_FE_SOURCE_SELECT_DIGC;
+	case ENGINE_ID_DIGD:
+		return DCN10_DIG_FE_SOURCE_SELECT_DIGD;
+	case ENGINE_ID_DIGE:
+		return DCN10_DIG_FE_SOURCE_SELECT_DIGE;
+	case ENGINE_ID_DIGF:
+		return DCN10_DIG_FE_SOURCE_SELECT_DIGF;
+	case ENGINE_ID_DIGG:
+		return DCN10_DIG_FE_SOURCE_SELECT_DIGG;
+	default:
+		ASSERT_CRITICAL(false);
+		return DCN10_DIG_FE_SOURCE_SELECT_INVALID;
+	}
+}
+
+static void configure_encoder(
+	struct dcn10_link_encoder *enc10,
+	const struct dc_link_settings *link_settings)
+{
+	/* set number of lanes */
+
+	REG_SET(DP_CONFIG, 0,
+			DP_UDI_LANES, link_settings->lane_count - LANE_COUNT_ONE);
+
+	/* setup scrambler */
+	REG_UPDATE(DP_DPHY_SCRAM_CNTL, DPHY_SCRAMBLER_ADVANCE, 1);
+}
+
+void dcn10_psr_program_dp_dphy_fast_training(struct link_encoder *enc,
+			bool exit_link_training_required)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+
+	if (exit_link_training_required)
+		REG_UPDATE(DP_DPHY_FAST_TRAINING,
+				DPHY_RX_FAST_TRAINING_CAPABLE, 1);
+	else {
+		REG_UPDATE(DP_DPHY_FAST_TRAINING,
+				DPHY_RX_FAST_TRAINING_CAPABLE, 0);
+		/*In DCE 11, we are able to pre-program a Force SR register
+		 * to be able to trigger SR symbol after 5 idle patterns
+		 * transmitted. Upon PSR Exit, DMCU can trigger
+		 * DPHY_LOAD_BS_COUNT_START = 1. Upon writing 1 to
+		 * DPHY_LOAD_BS_COUNT_START and the internal counter
+		 * reaches DPHY_LOAD_BS_COUNT, the next BS symbol will be
+		 * replaced by SR symbol once.
+		 */
+
+		REG_UPDATE(DP_DPHY_BS_SR_SWAP_CNTL, DPHY_LOAD_BS_COUNT, 0x5);
+	}
+}
+
+void dcn10_psr_program_secondary_packet(struct link_encoder *enc,
+			unsigned int sdp_transmit_line_num_deadline)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+
+	REG_UPDATE_2(DP_SEC_CNTL1,
+		DP_SEC_GSP0_LINE_NUM, sdp_transmit_line_num_deadline,
+		DP_SEC_GSP0_PRIORITY, 1);
+}
+
+bool dcn10_is_dig_enabled(struct link_encoder *enc)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	uint32_t value;
+
+	REG_GET(DIG_BE_EN_CNTL, DIG_ENABLE, &value);
+	return value;
+}
+
+static void link_encoder_disable(struct dcn10_link_encoder *enc10)
+{
+	/* reset training pattern */
+	REG_SET(DP_DPHY_TRAINING_PATTERN_SEL, 0,
+			DPHY_TRAINING_PATTERN_SEL, 0);
+
+	/* reset training complete */
+	REG_UPDATE(DP_LINK_CNTL, DP_LINK_TRAINING_COMPLETE, 0);
+
+	/* reset panel mode */
+	setup_panel_mode(enc10, DP_PANEL_MODE_DEFAULT);
+}
+
+static void hpd_initialize(
+	struct dcn10_link_encoder *enc10)
+{
+	/* Associate HPD with DIG_BE */
+	enum hpd_source_id hpd_source = enc10->base.hpd_source;
+
+	REG_UPDATE(DIG_BE_CNTL, DIG_HPD_SELECT, hpd_source);
+}
+
+bool dcn10_link_encoder_validate_dvi_output(
+	const struct dcn10_link_encoder *enc10,
+	enum signal_type connector_signal,
+	enum signal_type signal,
+	const struct dc_crtc_timing *crtc_timing)
+{
+	uint32_t max_pixel_clock = TMDS_MAX_PIXEL_CLOCK;
+
+	if (signal == SIGNAL_TYPE_DVI_DUAL_LINK)
+		max_pixel_clock *= 2;
+
+	/* This handles the case of HDMI downgrade to DVI we don't want to
+	 * we don't want to cap the pixel clock if the DDI is not DVI.
+	 */
+	if (connector_signal != SIGNAL_TYPE_DVI_DUAL_LINK &&
+			connector_signal != SIGNAL_TYPE_DVI_SINGLE_LINK)
+		max_pixel_clock = enc10->base.features.max_hdmi_pixel_clock;
+
+	/* DVI only support RGB pixel encoding */
+	if (crtc_timing->pixel_encoding != PIXEL_ENCODING_RGB)
+		return false;
+
+	/*connect DVI via adpater's HDMI connector*/
+	if ((connector_signal == SIGNAL_TYPE_DVI_SINGLE_LINK ||
+		connector_signal == SIGNAL_TYPE_HDMI_TYPE_A) &&
+		signal != SIGNAL_TYPE_HDMI_TYPE_A &&
+		crtc_timing->pix_clk_khz > TMDS_MAX_PIXEL_CLOCK)
+		return false;
+	if (crtc_timing->pix_clk_khz < TMDS_MIN_PIXEL_CLOCK)
+		return false;
+
+	if (crtc_timing->pix_clk_khz > max_pixel_clock)
+		return false;
+
+	/* DVI supports 6/8bpp single-link and 10/16bpp dual-link */
+	switch (crtc_timing->display_color_depth) {
+	case COLOR_DEPTH_666:
+	case COLOR_DEPTH_888:
+	break;
+	case COLOR_DEPTH_101010:
+	case COLOR_DEPTH_161616:
+		if (signal != SIGNAL_TYPE_DVI_DUAL_LINK)
+			return false;
+	break;
+	default:
+		return false;
+	}
+
+	return true;
+}
+
+static bool dcn10_link_encoder_validate_hdmi_output(
+	const struct dcn10_link_encoder *enc10,
+	const struct dc_crtc_timing *crtc_timing,
+	int adjusted_pix_clk_khz)
+{
+	enum dc_color_depth max_deep_color =
+			enc10->base.features.max_hdmi_deep_color;
+
+	if (max_deep_color < crtc_timing->display_color_depth)
+		return false;
+
+	if (crtc_timing->display_color_depth < COLOR_DEPTH_888)
+		return false;
+	if (adjusted_pix_clk_khz < TMDS_MIN_PIXEL_CLOCK)
+		return false;
+
+	if ((adjusted_pix_clk_khz == 0) ||
+		(adjusted_pix_clk_khz > enc10->base.features.max_hdmi_pixel_clock))
+		return false;
+
+	/* DCE11 HW does not support 420 */
+	if (!enc10->base.features.ycbcr420_supported &&
+			crtc_timing->pixel_encoding == PIXEL_ENCODING_YCBCR420)
+		return false;
+
+	if (!enc10->base.features.flags.bits.HDMI_6GB_EN &&
+		adjusted_pix_clk_khz >= 300000)
+		return false;
+	return true;
+}
+
+bool dcn10_link_encoder_validate_dp_output(
+	const struct dcn10_link_encoder *enc10,
+	const struct dc_crtc_timing *crtc_timing)
+{
+	/* default RGB only */
+	if (crtc_timing->pixel_encoding == PIXEL_ENCODING_RGB)
+		return true;
+
+	if (enc10->base.features.flags.bits.IS_YCBCR_CAPABLE)
+		return true;
+
+	/* for DCE 8.x or later DP Y-only feature,
+	 * we need ASIC cap + FeatureSupportDPYonly, not support 666
+	 */
+	if (crtc_timing->flags.Y_ONLY &&
+		enc10->base.features.flags.bits.IS_YCBCR_CAPABLE &&
+		crtc_timing->display_color_depth != COLOR_DEPTH_666)
+		return true;
+
+	return false;
+}
+
+void dcn10_link_encoder_construct(
+	struct dcn10_link_encoder *enc10,
+	const struct encoder_init_data *init_data,
+	const struct encoder_feature_support *enc_features,
+	const struct dcn10_link_enc_registers *link_regs,
+	const struct dcn10_link_enc_aux_registers *aux_regs,
+	const struct dcn10_link_enc_hpd_registers *hpd_regs,
+	const struct dcn10_link_enc_shift *link_shift,
+	const struct dcn10_link_enc_mask *link_mask)
+{
+	struct bp_encoder_cap_info bp_cap_info = {0};
+	const struct dc_vbios_funcs *bp_funcs = init_data->ctx->dc_bios->funcs;
+	enum bp_result result = BP_RESULT_OK;
+
+	enc10->base.funcs = &dcn10_lnk_enc_funcs;
+	enc10->base.ctx = init_data->ctx;
+	enc10->base.id = init_data->encoder;
+
+	enc10->base.hpd_source = init_data->hpd_source;
+	enc10->base.connector = init_data->connector;
+
+	enc10->base.preferred_engine = ENGINE_ID_UNKNOWN;
+
+	enc10->base.features = *enc_features;
+
+	enc10->base.transmitter = init_data->transmitter;
+
+	/* set the flag to indicate whether driver poll the I2C data pin
+	 * while doing the DP sink detect
+	 */
+
+/*	if (dal_adapter_service_is_feature_supported(as,
+		FEATURE_DP_SINK_DETECT_POLL_DATA_PIN))
+		enc10->base.features.flags.bits.
+			DP_SINK_DETECT_POLL_DATA_PIN = true;*/
+
+	enc10->base.output_signals =
+		SIGNAL_TYPE_DVI_SINGLE_LINK |
+		SIGNAL_TYPE_DVI_DUAL_LINK |
+		SIGNAL_TYPE_LVDS |
+		SIGNAL_TYPE_DISPLAY_PORT |
+		SIGNAL_TYPE_DISPLAY_PORT_MST |
+		SIGNAL_TYPE_EDP |
+		SIGNAL_TYPE_HDMI_TYPE_A;
+
+	/* For DCE 8.0 and 8.1, by design, UNIPHY is hardwired to DIG_BE.
+	 * SW always assign DIG_FE 1:1 mapped to DIG_FE for non-MST UNIPHY.
+	 * SW assign DIG_FE to non-MST UNIPHY first and MST last. So prefer
+	 * DIG is per UNIPHY and used by SST DP, eDP, HDMI, DVI and LVDS.
+	 * Prefer DIG assignment is decided by board design.
+	 * For DCE 8.0, there are only max 6 UNIPHYs, we assume board design
+	 * and VBIOS will filter out 7 UNIPHY for DCE 8.0.
+	 * By this, adding DIGG should not hurt DCE 8.0.
+	 * This will let DCE 8.1 share DCE 8.0 as much as possible
+	 */
+
+	enc10->link_regs = link_regs;
+	enc10->aux_regs = aux_regs;
+	enc10->hpd_regs = hpd_regs;
+	enc10->link_shift = link_shift;
+	enc10->link_mask = link_mask;
+
+	switch (enc10->base.transmitter) {
+	case TRANSMITTER_UNIPHY_A:
+		enc10->base.preferred_engine = ENGINE_ID_DIGA;
+	break;
+	case TRANSMITTER_UNIPHY_B:
+		enc10->base.preferred_engine = ENGINE_ID_DIGB;
+	break;
+	case TRANSMITTER_UNIPHY_C:
+		enc10->base.preferred_engine = ENGINE_ID_DIGC;
+	break;
+	case TRANSMITTER_UNIPHY_D:
+		enc10->base.preferred_engine = ENGINE_ID_DIGD;
+	break;
+	case TRANSMITTER_UNIPHY_E:
+		enc10->base.preferred_engine = ENGINE_ID_DIGE;
+	break;
+	case TRANSMITTER_UNIPHY_F:
+		enc10->base.preferred_engine = ENGINE_ID_DIGF;
+	break;
+	case TRANSMITTER_UNIPHY_G:
+		enc10->base.preferred_engine = ENGINE_ID_DIGG;
+	break;
+	default:
+		ASSERT_CRITICAL(false);
+		enc10->base.preferred_engine = ENGINE_ID_UNKNOWN;
+	}
+
+	/* default to one to mirror Windows behavior */
+	enc10->base.features.flags.bits.HDMI_6GB_EN = 1;
+
+	result = bp_funcs->get_encoder_cap_info(enc10->base.ctx->dc_bios,
+						enc10->base.id, &bp_cap_info);
+
+	/* Override features with DCE-specific values */
+	if (result == BP_RESULT_OK) {
+		enc10->base.features.flags.bits.IS_HBR2_CAPABLE =
+				bp_cap_info.DP_HBR2_EN;
+		enc10->base.features.flags.bits.IS_HBR3_CAPABLE =
+				bp_cap_info.DP_HBR3_EN;
+		enc10->base.features.flags.bits.HDMI_6GB_EN = bp_cap_info.HDMI_6GB_EN;
+	} else {
+		DC_LOG_WARNING("%s: Failed to get encoder_cap_info from VBIOS with error code %d!\n",
+				__func__,
+				result);
+	}
+}
+
+bool dcn10_link_encoder_validate_output_with_stream(
+	struct link_encoder *enc,
+	const struct dc_stream_state *stream)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	bool is_valid;
+
+	switch (stream->signal) {
+	case SIGNAL_TYPE_DVI_SINGLE_LINK:
+	case SIGNAL_TYPE_DVI_DUAL_LINK:
+		is_valid = dcn10_link_encoder_validate_dvi_output(
+			enc10,
+			stream->sink->link->connector_signal,
+			stream->signal,
+			&stream->timing);
+	break;
+	case SIGNAL_TYPE_HDMI_TYPE_A:
+		is_valid = dcn10_link_encoder_validate_hdmi_output(
+				enc10,
+				&stream->timing,
+				stream->phy_pix_clk);
+	break;
+	case SIGNAL_TYPE_DISPLAY_PORT:
+	case SIGNAL_TYPE_DISPLAY_PORT_MST:
+		is_valid = dcn10_link_encoder_validate_dp_output(
+					enc10, &stream->timing);
+	break;
+	case SIGNAL_TYPE_EDP:
+		is_valid = (stream->timing.pixel_encoding == PIXEL_ENCODING_RGB) ? true : false;
+	break;
+	case SIGNAL_TYPE_VIRTUAL:
+		is_valid = true;
+		break;
+	default:
+		is_valid = false;
+	break;
+	}
+
+	return is_valid;
+}
+
+void dcn10_link_encoder_hw_init(
+	struct link_encoder *enc)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	struct bp_transmitter_control cntl = { 0 };
+	enum bp_result result;
+
+	cntl.action = TRANSMITTER_CONTROL_INIT;
+	cntl.engine_id = ENGINE_ID_UNKNOWN;
+	cntl.transmitter = enc10->base.transmitter;
+	cntl.connector_obj_id = enc10->base.connector;
+	cntl.lanes_number = LANE_COUNT_FOUR;
+	cntl.coherent = false;
+	cntl.hpd_sel = enc10->base.hpd_source;
+
+	if (enc10->base.connector.id == CONNECTOR_ID_EDP)
+		cntl.signal = SIGNAL_TYPE_EDP;
+
+	result = link_transmitter_control(enc10, &cntl);
+
+	if (result != BP_RESULT_OK) {
+		DC_LOG_ERROR("%s: Failed to execute VBIOS command table!\n",
+			__func__);
+		BREAK_TO_DEBUGGER();
+		return;
+	}
+
+	if (enc10->base.connector.id == CONNECTOR_ID_LVDS) {
+		cntl.action = TRANSMITTER_CONTROL_BACKLIGHT_BRIGHTNESS;
+
+		result = link_transmitter_control(enc10, &cntl);
+
+		ASSERT(result == BP_RESULT_OK);
+
+	}
+	aux_initialize(enc10);
+
+	/* reinitialize HPD.
+	 * hpd_initialize() will pass DIG_FE id to HW context.
+	 * All other routine within HW context will use fe_engine_offset
+	 * as DIG_FE id even caller pass DIG_FE id.
+	 * So this routine must be called first.
+	 */
+	hpd_initialize(enc10);
+}
+
+void dcn10_link_encoder_destroy(struct link_encoder **enc)
+{
+	kfree(TO_DCN10_LINK_ENC(*enc));
+	*enc = NULL;
+}
+
+void dcn10_link_encoder_setup(
+	struct link_encoder *enc,
+	enum signal_type signal)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+
+	switch (signal) {
+	case SIGNAL_TYPE_EDP:
+	case SIGNAL_TYPE_DISPLAY_PORT:
+		/* DP SST */
+		REG_UPDATE(DIG_BE_CNTL, DIG_MODE, 0);
+		break;
+	case SIGNAL_TYPE_LVDS:
+		/* LVDS */
+		REG_UPDATE(DIG_BE_CNTL, DIG_MODE, 1);
+		break;
+	case SIGNAL_TYPE_DVI_SINGLE_LINK:
+	case SIGNAL_TYPE_DVI_DUAL_LINK:
+		/* TMDS-DVI */
+		REG_UPDATE(DIG_BE_CNTL, DIG_MODE, 2);
+		break;
+	case SIGNAL_TYPE_HDMI_TYPE_A:
+		/* TMDS-HDMI */
+		REG_UPDATE(DIG_BE_CNTL, DIG_MODE, 3);
+		break;
+	case SIGNAL_TYPE_DISPLAY_PORT_MST:
+		/* DP MST */
+		REG_UPDATE(DIG_BE_CNTL, DIG_MODE, 5);
+		break;
+	default:
+		ASSERT_CRITICAL(false);
+		/* invalid mode ! */
+		break;
+	}
+
+}
+
+/* TODO: still need depth or just pass in adjusted pixel clock? */
+void dcn10_link_encoder_enable_tmds_output(
+	struct link_encoder *enc,
+	enum clock_source_id clock_source,
+	enum dc_color_depth color_depth,
+	enum signal_type signal,
+	uint32_t pixel_clock)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	struct bp_transmitter_control cntl = { 0 };
+	enum bp_result result;
+
+	/* Enable the PHY */
+
+	cntl.action = TRANSMITTER_CONTROL_ENABLE;
+	cntl.engine_id = enc->preferred_engine;
+	cntl.transmitter = enc10->base.transmitter;
+	cntl.pll_id = clock_source;
+	cntl.signal = signal;
+	if (cntl.signal == SIGNAL_TYPE_DVI_DUAL_LINK)
+		cntl.lanes_number = 8;
+	else
+		cntl.lanes_number = 4;
+
+	cntl.hpd_sel = enc10->base.hpd_source;
+
+	cntl.pixel_clock = pixel_clock;
+	cntl.color_depth = color_depth;
+
+	result = link_transmitter_control(enc10, &cntl);
+
+	if (result != BP_RESULT_OK) {
+		DC_LOG_ERROR("%s: Failed to execute VBIOS command table!\n",
+			__func__);
+		BREAK_TO_DEBUGGER();
+	}
+}
+
+/* enables DP PHY output */
+void dcn10_link_encoder_enable_dp_output(
+	struct link_encoder *enc,
+	const struct dc_link_settings *link_settings,
+	enum clock_source_id clock_source)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	struct bp_transmitter_control cntl = { 0 };
+	enum bp_result result;
+
+	/* Enable the PHY */
+
+	/* number_of_lanes is used for pixel clock adjust,
+	 * but it's not passed to asic_control.
+	 * We need to set number of lanes manually.
+	 */
+	configure_encoder(enc10, link_settings);
+
+	cntl.action = TRANSMITTER_CONTROL_ENABLE;
+	cntl.engine_id = enc->preferred_engine;
+	cntl.transmitter = enc10->base.transmitter;
+	cntl.pll_id = clock_source;
+	cntl.signal = SIGNAL_TYPE_DISPLAY_PORT;
+	cntl.lanes_number = link_settings->lane_count;
+	cntl.hpd_sel = enc10->base.hpd_source;
+	cntl.pixel_clock = link_settings->link_rate
+						* LINK_RATE_REF_FREQ_IN_KHZ;
+	/* TODO: check if undefined works */
+	cntl.color_depth = COLOR_DEPTH_UNDEFINED;
+
+	result = link_transmitter_control(enc10, &cntl);
+
+	if (result != BP_RESULT_OK) {
+		DC_LOG_ERROR("%s: Failed to execute VBIOS command table!\n",
+			__func__);
+		BREAK_TO_DEBUGGER();
+	}
+}
+
+/* enables DP PHY output in MST mode */
+void dcn10_link_encoder_enable_dp_mst_output(
+	struct link_encoder *enc,
+	const struct dc_link_settings *link_settings,
+	enum clock_source_id clock_source)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	struct bp_transmitter_control cntl = { 0 };
+	enum bp_result result;
+
+	/* Enable the PHY */
+
+	/* number_of_lanes is used for pixel clock adjust,
+	 * but it's not passed to asic_control.
+	 * We need to set number of lanes manually.
+	 */
+	configure_encoder(enc10, link_settings);
+
+	cntl.action = TRANSMITTER_CONTROL_ENABLE;
+	cntl.engine_id = ENGINE_ID_UNKNOWN;
+	cntl.transmitter = enc10->base.transmitter;
+	cntl.pll_id = clock_source;
+	cntl.signal = SIGNAL_TYPE_DISPLAY_PORT_MST;
+	cntl.lanes_number = link_settings->lane_count;
+	cntl.hpd_sel = enc10->base.hpd_source;
+	cntl.pixel_clock = link_settings->link_rate
+						* LINK_RATE_REF_FREQ_IN_KHZ;
+	/* TODO: check if undefined works */
+	cntl.color_depth = COLOR_DEPTH_UNDEFINED;
+
+	result = link_transmitter_control(enc10, &cntl);
+
+	if (result != BP_RESULT_OK) {
+		DC_LOG_ERROR("%s: Failed to execute VBIOS command table!\n",
+			__func__);
+		BREAK_TO_DEBUGGER();
+	}
+}
+/*
+ * @brief
+ * Disable transmitter and its encoder
+ */
+void dcn10_link_encoder_disable_output(
+	struct link_encoder *enc,
+	enum signal_type signal)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	struct bp_transmitter_control cntl = { 0 };
+	enum bp_result result;
+
+	if (!dcn10_is_dig_enabled(enc)) {
+		/* OF_SKIP_POWER_DOWN_INACTIVE_ENCODER */
+		return;
+	}
+	/* Power-down RX and disable GPU PHY should be paired.
+	 * Disabling PHY without powering down RX may cause
+	 * symbol lock loss, on which we will get DP Sink interrupt.
+	 */
+
+	/* There is a case for the DP active dongles
+	 * where we want to disable the PHY but keep RX powered,
+	 * for those we need to ignore DP Sink interrupt
+	 * by checking lane count that has been set
+	 * on the last do_enable_output().
+	 */
+
+	/* disable transmitter */
+	cntl.action = TRANSMITTER_CONTROL_DISABLE;
+	cntl.transmitter = enc10->base.transmitter;
+	cntl.hpd_sel = enc10->base.hpd_source;
+	cntl.signal = signal;
+	cntl.connector_obj_id = enc10->base.connector;
+
+	result = link_transmitter_control(enc10, &cntl);
+
+	if (result != BP_RESULT_OK) {
+		DC_LOG_ERROR("%s: Failed to execute VBIOS command table!\n",
+			__func__);
+		BREAK_TO_DEBUGGER();
+		return;
+	}
+
+	/* disable encoder */
+	if (dc_is_dp_signal(signal))
+		link_encoder_disable(enc10);
+}
+
+void dcn10_link_encoder_dp_set_lane_settings(
+	struct link_encoder *enc,
+	const struct link_training_settings *link_settings)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	union dpcd_training_lane_set training_lane_set = { { 0 } };
+	int32_t lane = 0;
+	struct bp_transmitter_control cntl = { 0 };
+
+	if (!link_settings) {
+		BREAK_TO_DEBUGGER();
+		return;
+	}
+
+	cntl.action = TRANSMITTER_CONTROL_SET_VOLTAGE_AND_PREEMPASIS;
+	cntl.transmitter = enc10->base.transmitter;
+	cntl.connector_obj_id = enc10->base.connector;
+	cntl.lanes_number = link_settings->link_settings.lane_count;
+	cntl.hpd_sel = enc10->base.hpd_source;
+	cntl.pixel_clock = link_settings->link_settings.link_rate *
+						LINK_RATE_REF_FREQ_IN_KHZ;
+
+	for (lane = 0; lane < link_settings->link_settings.lane_count; lane++) {
+		/* translate lane settings */
+
+		training_lane_set.bits.VOLTAGE_SWING_SET =
+			link_settings->lane_settings[lane].VOLTAGE_SWING;
+		training_lane_set.bits.PRE_EMPHASIS_SET =
+			link_settings->lane_settings[lane].PRE_EMPHASIS;
+
+		/* post cursor 2 setting only applies to HBR2 link rate */
+		if (link_settings->link_settings.link_rate == LINK_RATE_HIGH2) {
+			/* this is passed to VBIOS
+			 * to program post cursor 2 level
+			 */
+			training_lane_set.bits.POST_CURSOR2_SET =
+				link_settings->lane_settings[lane].POST_CURSOR2;
+		}
+
+		cntl.lane_select = lane;
+		cntl.lane_settings = training_lane_set.raw;
+
+		/* call VBIOS table to set voltage swing and pre-emphasis */
+		link_transmitter_control(enc10, &cntl);
+	}
+}
+
+/* set DP PHY test and training patterns */
+void dcn10_link_encoder_dp_set_phy_pattern(
+	struct link_encoder *enc,
+	const struct encoder_set_dp_phy_pattern_param *param)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+
+	switch (param->dp_phy_pattern) {
+	case DP_TEST_PATTERN_TRAINING_PATTERN1:
+		dcn10_link_encoder_set_dp_phy_pattern_training_pattern(enc, 0);
+		break;
+	case DP_TEST_PATTERN_TRAINING_PATTERN2:
+		dcn10_link_encoder_set_dp_phy_pattern_training_pattern(enc, 1);
+		break;
+	case DP_TEST_PATTERN_TRAINING_PATTERN3:
+		dcn10_link_encoder_set_dp_phy_pattern_training_pattern(enc, 2);
+		break;
+	case DP_TEST_PATTERN_TRAINING_PATTERN4:
+		dcn10_link_encoder_set_dp_phy_pattern_training_pattern(enc, 3);
+		break;
+	case DP_TEST_PATTERN_D102:
+		set_dp_phy_pattern_d102(enc10);
+		break;
+	case DP_TEST_PATTERN_SYMBOL_ERROR:
+		set_dp_phy_pattern_symbol_error(enc10);
+		break;
+	case DP_TEST_PATTERN_PRBS7:
+		set_dp_phy_pattern_prbs7(enc10);
+		break;
+	case DP_TEST_PATTERN_80BIT_CUSTOM:
+		set_dp_phy_pattern_80bit_custom(
+			enc10, param->custom_pattern);
+		break;
+	case DP_TEST_PATTERN_CP2520_1:
+		set_dp_phy_pattern_hbr2_compliance_cp2520_2(enc10, 1);
+		break;
+	case DP_TEST_PATTERN_CP2520_2:
+		set_dp_phy_pattern_hbr2_compliance_cp2520_2(enc10, 2);
+		break;
+	case DP_TEST_PATTERN_CP2520_3:
+		set_dp_phy_pattern_hbr2_compliance_cp2520_2(enc10, 3);
+		break;
+	case DP_TEST_PATTERN_VIDEO_MODE: {
+		set_dp_phy_pattern_passthrough_mode(
+			enc10, param->dp_panel_mode);
+		break;
+	}
+
+	default:
+		/* invalid phy pattern */
+		ASSERT_CRITICAL(false);
+		break;
+	}
+}
+
+static void fill_stream_allocation_row_info(
+	const struct link_mst_stream_allocation *stream_allocation,
+	uint32_t *src,
+	uint32_t *slots)
+{
+	const struct stream_encoder *stream_enc = stream_allocation->stream_enc;
+
+	if (stream_enc) {
+		*src = stream_enc->id;
+		*slots = stream_allocation->slot_count;
+	} else {
+		*src = 0;
+		*slots = 0;
+	}
+}
+
+/* programs DP MST VC payload allocation */
+void dcn10_link_encoder_update_mst_stream_allocation_table(
+	struct link_encoder *enc,
+	const struct link_mst_stream_allocation_table *table)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	uint32_t value0 = 0;
+	uint32_t value1 = 0;
+	uint32_t value2 = 0;
+	uint32_t slots = 0;
+	uint32_t src = 0;
+	uint32_t retries = 0;
+
+	/* For CZ, there are only 3 pipes. So Virtual channel is up 3.*/
+
+	/* --- Set MSE Stream Attribute -
+	 * Setup VC Payload Table on Tx Side,
+	 * Issue allocation change trigger
+	 * to commit payload on both tx and rx side
+	 */
+
+	/* we should clean-up table each time */
+
+	if (table->stream_count >= 1) {
+		fill_stream_allocation_row_info(
+			&table->stream_allocations[0],
+			&src,
+			&slots);
+	} else {
+		src = 0;
+		slots = 0;
+	}
+
+	REG_UPDATE_2(DP_MSE_SAT0,
+			DP_MSE_SAT_SRC0, src,
+			DP_MSE_SAT_SLOT_COUNT0, slots);
+
+	if (table->stream_count >= 2) {
+		fill_stream_allocation_row_info(
+			&table->stream_allocations[1],
+			&src,
+			&slots);
+	} else {
+		src = 0;
+		slots = 0;
+	}
+
+	REG_UPDATE_2(DP_MSE_SAT0,
+			DP_MSE_SAT_SRC1, src,
+			DP_MSE_SAT_SLOT_COUNT1, slots);
+
+	if (table->stream_count >= 3) {
+		fill_stream_allocation_row_info(
+			&table->stream_allocations[2],
+			&src,
+			&slots);
+	} else {
+		src = 0;
+		slots = 0;
+	}
+
+	REG_UPDATE_2(DP_MSE_SAT1,
+			DP_MSE_SAT_SRC2, src,
+			DP_MSE_SAT_SLOT_COUNT2, slots);
+
+	if (table->stream_count >= 4) {
+		fill_stream_allocation_row_info(
+			&table->stream_allocations[3],
+			&src,
+			&slots);
+	} else {
+		src = 0;
+		slots = 0;
+	}
+
+	REG_UPDATE_2(DP_MSE_SAT1,
+			DP_MSE_SAT_SRC3, src,
+			DP_MSE_SAT_SLOT_COUNT3, slots);
+
+	/* --- wait for transaction finish */
+
+	/* send allocation change trigger (ACT) ?
+	 * this step first sends the ACT,
+	 * then double buffers the SAT into the hardware
+	 * making the new allocation active on the DP MST mode link
+	 */
+
+	/* DP_MSE_SAT_UPDATE:
+	 * 0 - No Action
+	 * 1 - Update SAT with trigger
+	 * 2 - Update SAT without trigger
+	 */
+	REG_UPDATE(DP_MSE_SAT_UPDATE,
+			DP_MSE_SAT_UPDATE, 1);
+
+	/* wait for update to complete
+	 * (i.e. DP_MSE_SAT_UPDATE field is reset to 0)
+	 * then wait for the transmission
+	 * of at least 16 MTP headers on immediate local link.
+	 * i.e. DP_MSE_16_MTP_KEEPOUT field (read only) is reset to 0
+	 * a value of 1 indicates that DP MST mode
+	 * is in the 16 MTP keepout region after a VC has been added.
+	 * MST stream bandwidth (VC rate) can be configured
+	 * after this bit is cleared
+	 */
+	do {
+		udelay(10);
+
+		value0 = REG_READ(DP_MSE_SAT_UPDATE);
+
+		REG_GET(DP_MSE_SAT_UPDATE,
+				DP_MSE_SAT_UPDATE, &value1);
+
+		REG_GET(DP_MSE_SAT_UPDATE,
+				DP_MSE_16_MTP_KEEPOUT, &value2);
+
+		/* bit field DP_MSE_SAT_UPDATE is set to 1 already */
+		if (!value1 && !value2)
+			break;
+		++retries;
+	} while (retries < DP_MST_UPDATE_MAX_RETRY);
+}
+
+void dcn10_link_encoder_connect_dig_be_to_fe(
+	struct link_encoder *enc,
+	enum engine_id engine,
+	bool connect)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	uint32_t field;
+
+	if (engine != ENGINE_ID_UNKNOWN) {
+
+		REG_GET(DIG_BE_CNTL, DIG_FE_SOURCE_SELECT, &field);
+
+		if (connect)
+			field |= get_frontend_source(engine);
+		else
+			field &= ~get_frontend_source(engine);
+
+		REG_UPDATE(DIG_BE_CNTL, DIG_FE_SOURCE_SELECT, field);
+	}
+}
+
+
+#define HPD_REG(reg)\
+	(enc10->hpd_regs->reg)
+
+#define HPD_REG_READ(reg_name) \
+		dm_read_reg(CTX, HPD_REG(reg_name))
+
+#define HPD_REG_UPDATE_N(reg_name, n, ...)	\
+		generic_reg_update_ex(CTX, \
+				HPD_REG(reg_name), \
+				HPD_REG_READ(reg_name), \
+				n, __VA_ARGS__)
+
+#define HPD_REG_UPDATE(reg_name, field, val)	\
+		HPD_REG_UPDATE_N(reg_name, 1, \
+				FN(reg_name, field), val)
+
+void dcn10_link_encoder_enable_hpd(struct link_encoder *enc)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+
+	HPD_REG_UPDATE(DC_HPD_CONTROL,
+			DC_HPD_EN, 1);
+}
+
+void dcn10_link_encoder_disable_hpd(struct link_encoder *enc)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+
+	HPD_REG_UPDATE(DC_HPD_CONTROL,
+			DC_HPD_EN, 0);
+}
+
+
+#define AUX_REG(reg)\
+	(enc10->aux_regs->reg)
+
+#define AUX_REG_READ(reg_name) \
+		dm_read_reg(CTX, AUX_REG(reg_name))
+
+#define AUX_REG_UPDATE_N(reg_name, n, ...)	\
+		generic_reg_update_ex(CTX, \
+				AUX_REG(reg_name), \
+				AUX_REG_READ(reg_name), \
+				n, __VA_ARGS__)
+
+#define AUX_REG_UPDATE(reg_name, field, val)	\
+		AUX_REG_UPDATE_N(reg_name, 1, \
+				FN(reg_name, field), val)
+
+#define AUX_REG_UPDATE_2(reg, f1, v1, f2, v2)	\
+		AUX_REG_UPDATE_N(reg, 2,\
+				FN(reg, f1), v1,\
+				FN(reg, f2), v2)
+
+static void aux_initialize(
+	struct dcn10_link_encoder *enc10)
+{
+	enum hpd_source_id hpd_source = enc10->base.hpd_source;
+
+	AUX_REG_UPDATE_2(AUX_CONTROL,
+			AUX_HPD_SEL, hpd_source,
+			AUX_LS_READ_EN, 0);
+
+	/* 1/4 window (the maximum allowed) */
+	AUX_REG_UPDATE(AUX_DPHY_RX_CONTROL0,
+			AUX_RX_RECEIVE_WINDOW, 1);
+}
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.h
new file mode 100644
index 0000000..2a97cdb
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.h
@@ -0,0 +1,330 @@
+/*
+ * Copyright 2012-15 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ *  and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef __DC_LINK_ENCODER__DCN10_H__
+#define __DC_LINK_ENCODER__DCN10_H__
+
+#include "link_encoder.h"
+
+#define TO_DCN10_LINK_ENC(link_encoder)\
+	container_of(link_encoder, struct dcn10_link_encoder, base)
+
+
+#define AUX_REG_LIST(id)\
+	SRI(AUX_CONTROL, DP_AUX, id), \
+	SRI(AUX_DPHY_RX_CONTROL0, DP_AUX, id)
+
+#define HPD_REG_LIST(id)\
+	SRI(DC_HPD_CONTROL, HPD, id)
+
+#define LE_DCN_COMMON_REG_LIST(id) \
+	SRI(DIG_BE_CNTL, DIG, id), \
+	SRI(DIG_BE_EN_CNTL, DIG, id), \
+	SRI(DP_CONFIG, DP, id), \
+	SRI(DP_DPHY_CNTL, DP, id), \
+	SRI(DP_DPHY_PRBS_CNTL, DP, id), \
+	SRI(DP_DPHY_SCRAM_CNTL, DP, id),\
+	SRI(DP_DPHY_SYM0, DP, id), \
+	SRI(DP_DPHY_SYM1, DP, id), \
+	SRI(DP_DPHY_SYM2, DP, id), \
+	SRI(DP_DPHY_TRAINING_PATTERN_SEL, DP, id), \
+	SRI(DP_LINK_CNTL, DP, id), \
+	SRI(DP_LINK_FRAMING_CNTL, DP, id), \
+	SRI(DP_MSE_SAT0, DP, id), \
+	SRI(DP_MSE_SAT1, DP, id), \
+	SRI(DP_MSE_SAT2, DP, id), \
+	SRI(DP_MSE_SAT_UPDATE, DP, id), \
+	SRI(DP_SEC_CNTL, DP, id), \
+	SRI(DP_VID_STREAM_CNTL, DP, id), \
+	SRI(DP_DPHY_FAST_TRAINING, DP, id), \
+	SRI(DP_SEC_CNTL1, DP, id), \
+	SRI(DP_DPHY_BS_SR_SWAP_CNTL, DP, id), \
+	SRI(DP_DPHY_INTERNAL_CTRL, DP, id), \
+	SRI(DP_DPHY_HBR2_PATTERN_CONTROL, DP, id)
+
+#define LE_DCN10_REG_LIST(id)\
+	LE_DCN_COMMON_REG_LIST(id)
+
+struct dcn10_link_enc_aux_registers {
+	uint32_t AUX_CONTROL;
+	uint32_t AUX_DPHY_RX_CONTROL0;
+};
+
+struct dcn10_link_enc_hpd_registers {
+	uint32_t DC_HPD_CONTROL;
+};
+
+struct dcn10_link_enc_registers {
+	uint32_t DIG_BE_CNTL;
+	uint32_t DIG_BE_EN_CNTL;
+	uint32_t DP_CONFIG;
+	uint32_t DP_DPHY_CNTL;
+	uint32_t DP_DPHY_INTERNAL_CTRL;
+	uint32_t DP_DPHY_PRBS_CNTL;
+	uint32_t DP_DPHY_SCRAM_CNTL;
+	uint32_t DP_DPHY_SYM0;
+	uint32_t DP_DPHY_SYM1;
+	uint32_t DP_DPHY_SYM2;
+	uint32_t DP_DPHY_TRAINING_PATTERN_SEL;
+	uint32_t DP_LINK_CNTL;
+	uint32_t DP_LINK_FRAMING_CNTL;
+	uint32_t DP_MSE_SAT0;
+	uint32_t DP_MSE_SAT1;
+	uint32_t DP_MSE_SAT2;
+	uint32_t DP_MSE_SAT_UPDATE;
+	uint32_t DP_SEC_CNTL;
+	uint32_t DP_VID_STREAM_CNTL;
+	uint32_t DP_DPHY_FAST_TRAINING;
+	uint32_t DP_DPHY_BS_SR_SWAP_CNTL;
+	uint32_t DP_DPHY_HBR2_PATTERN_CONTROL;
+	uint32_t DP_SEC_CNTL1;
+};
+
+#define LE_SF(reg_name, field_name, post_fix)\
+	.field_name = reg_name ## __ ## field_name ## post_fix
+
+#define LINK_ENCODER_MASK_SH_LIST_DCN10(mask_sh)\
+	LE_SF(DIG0_DIG_BE_EN_CNTL, DIG_ENABLE, mask_sh),\
+	LE_SF(DIG0_DIG_BE_CNTL, DIG_HPD_SELECT, mask_sh),\
+	LE_SF(DIG0_DIG_BE_CNTL, DIG_MODE, mask_sh),\
+	LE_SF(DIG0_DIG_BE_CNTL, DIG_FE_SOURCE_SELECT, mask_sh),\
+	LE_SF(DP0_DP_DPHY_CNTL, DPHY_BYPASS, mask_sh),\
+	LE_SF(DP0_DP_DPHY_CNTL, DPHY_ATEST_SEL_LANE0, mask_sh),\
+	LE_SF(DP0_DP_DPHY_CNTL, DPHY_ATEST_SEL_LANE1, mask_sh),\
+	LE_SF(DP0_DP_DPHY_CNTL, DPHY_ATEST_SEL_LANE2, mask_sh),\
+	LE_SF(DP0_DP_DPHY_CNTL, DPHY_ATEST_SEL_LANE3, mask_sh),\
+	LE_SF(DP0_DP_DPHY_PRBS_CNTL, DPHY_PRBS_EN, mask_sh),\
+	LE_SF(DP0_DP_DPHY_PRBS_CNTL, DPHY_PRBS_SEL, mask_sh),\
+	LE_SF(DP0_DP_DPHY_SYM0, DPHY_SYM1, mask_sh),\
+	LE_SF(DP0_DP_DPHY_SYM0, DPHY_SYM2, mask_sh),\
+	LE_SF(DP0_DP_DPHY_SYM0, DPHY_SYM3, mask_sh),\
+	LE_SF(DP0_DP_DPHY_SYM1, DPHY_SYM4, mask_sh),\
+	LE_SF(DP0_DP_DPHY_SYM1, DPHY_SYM5, mask_sh),\
+	LE_SF(DP0_DP_DPHY_SYM1, DPHY_SYM6, mask_sh),\
+	LE_SF(DP0_DP_DPHY_SYM2, DPHY_SYM7, mask_sh),\
+	LE_SF(DP0_DP_DPHY_SYM2, DPHY_SYM8, mask_sh),\
+	LE_SF(DP0_DP_DPHY_SCRAM_CNTL, DPHY_SCRAMBLER_BS_COUNT, mask_sh),\
+	LE_SF(DP0_DP_DPHY_SCRAM_CNTL, DPHY_SCRAMBLER_ADVANCE, mask_sh),\
+	LE_SF(DP0_DP_DPHY_FAST_TRAINING, DPHY_RX_FAST_TRAINING_CAPABLE, mask_sh),\
+	LE_SF(DP0_DP_DPHY_BS_SR_SWAP_CNTL, DPHY_LOAD_BS_COUNT, mask_sh),\
+	LE_SF(DP0_DP_DPHY_TRAINING_PATTERN_SEL, DPHY_TRAINING_PATTERN_SEL, mask_sh),\
+	LE_SF(DP0_DP_DPHY_HBR2_PATTERN_CONTROL, DP_DPHY_HBR2_PATTERN_CONTROL, mask_sh),\
+	LE_SF(DP0_DP_LINK_CNTL, DP_LINK_TRAINING_COMPLETE, mask_sh),\
+	LE_SF(DP0_DP_LINK_FRAMING_CNTL, DP_IDLE_BS_INTERVAL, mask_sh),\
+	LE_SF(DP0_DP_LINK_FRAMING_CNTL, DP_VBID_DISABLE, mask_sh),\
+	LE_SF(DP0_DP_LINK_FRAMING_CNTL, DP_VID_ENHANCED_FRAME_MODE, mask_sh),\
+	LE_SF(DP0_DP_VID_STREAM_CNTL, DP_VID_STREAM_ENABLE, mask_sh),\
+	LE_SF(DP0_DP_CONFIG, DP_UDI_LANES, mask_sh),\
+	LE_SF(DP0_DP_SEC_CNTL1, DP_SEC_GSP0_LINE_NUM, mask_sh),\
+	LE_SF(DP0_DP_SEC_CNTL1, DP_SEC_GSP0_PRIORITY, mask_sh),\
+	LE_SF(DP0_DP_MSE_SAT0, DP_MSE_SAT_SRC0, mask_sh),\
+	LE_SF(DP0_DP_MSE_SAT0, DP_MSE_SAT_SRC1, mask_sh),\
+	LE_SF(DP0_DP_MSE_SAT0, DP_MSE_SAT_SLOT_COUNT0, mask_sh),\
+	LE_SF(DP0_DP_MSE_SAT0, DP_MSE_SAT_SLOT_COUNT1, mask_sh),\
+	LE_SF(DP0_DP_MSE_SAT1, DP_MSE_SAT_SRC2, mask_sh),\
+	LE_SF(DP0_DP_MSE_SAT1, DP_MSE_SAT_SRC3, mask_sh),\
+	LE_SF(DP0_DP_MSE_SAT1, DP_MSE_SAT_SLOT_COUNT2, mask_sh),\
+	LE_SF(DP0_DP_MSE_SAT1, DP_MSE_SAT_SLOT_COUNT3, mask_sh),\
+	LE_SF(DP0_DP_MSE_SAT_UPDATE, DP_MSE_SAT_UPDATE, mask_sh),\
+	LE_SF(DP0_DP_MSE_SAT_UPDATE, DP_MSE_16_MTP_KEEPOUT, mask_sh),\
+	LE_SF(DP_AUX0_AUX_CONTROL, AUX_HPD_SEL, mask_sh),\
+	LE_SF(DP_AUX0_AUX_CONTROL, AUX_LS_READ_EN, mask_sh),\
+	LE_SF(DP_AUX0_AUX_DPHY_RX_CONTROL0, AUX_RX_RECEIVE_WINDOW, mask_sh),\
+	LE_SF(HPD0_DC_HPD_CONTROL, DC_HPD_EN, mask_sh)
+
+#define DCN_LINK_ENCODER_REG_FIELD_LIST(type) \
+	type DIG_ENABLE;\
+	type DIG_HPD_SELECT;\
+	type DIG_MODE;\
+	type DIG_FE_SOURCE_SELECT;\
+	type DPHY_BYPASS;\
+	type DPHY_ATEST_SEL_LANE0;\
+	type DPHY_ATEST_SEL_LANE1;\
+	type DPHY_ATEST_SEL_LANE2;\
+	type DPHY_ATEST_SEL_LANE3;\
+	type DPHY_PRBS_EN;\
+	type DPHY_PRBS_SEL;\
+	type DPHY_SYM1;\
+	type DPHY_SYM2;\
+	type DPHY_SYM3;\
+	type DPHY_SYM4;\
+	type DPHY_SYM5;\
+	type DPHY_SYM6;\
+	type DPHY_SYM7;\
+	type DPHY_SYM8;\
+	type DPHY_SCRAMBLER_BS_COUNT;\
+	type DPHY_SCRAMBLER_ADVANCE;\
+	type DPHY_RX_FAST_TRAINING_CAPABLE;\
+	type DPHY_LOAD_BS_COUNT;\
+	type DPHY_TRAINING_PATTERN_SEL;\
+	type DP_DPHY_HBR2_PATTERN_CONTROL;\
+	type DP_LINK_TRAINING_COMPLETE;\
+	type DP_IDLE_BS_INTERVAL;\
+	type DP_VBID_DISABLE;\
+	type DP_VID_ENHANCED_FRAME_MODE;\
+	type DP_VID_STREAM_ENABLE;\
+	type DP_UDI_LANES;\
+	type DP_SEC_GSP0_LINE_NUM;\
+	type DP_SEC_GSP0_PRIORITY;\
+	type DP_MSE_SAT_SRC0;\
+	type DP_MSE_SAT_SRC1;\
+	type DP_MSE_SAT_SRC2;\
+	type DP_MSE_SAT_SRC3;\
+	type DP_MSE_SAT_SLOT_COUNT0;\
+	type DP_MSE_SAT_SLOT_COUNT1;\
+	type DP_MSE_SAT_SLOT_COUNT2;\
+	type DP_MSE_SAT_SLOT_COUNT3;\
+	type DP_MSE_SAT_UPDATE;\
+	type DP_MSE_16_MTP_KEEPOUT;\
+	type AUX_HPD_SEL;\
+	type AUX_LS_READ_EN;\
+	type AUX_RX_RECEIVE_WINDOW;\
+	type DC_HPD_EN
+
+struct dcn10_link_enc_shift {
+	DCN_LINK_ENCODER_REG_FIELD_LIST(uint8_t);
+};
+
+struct dcn10_link_enc_mask {
+	DCN_LINK_ENCODER_REG_FIELD_LIST(uint32_t);
+};
+
+struct dcn10_link_encoder {
+	struct link_encoder base;
+	const struct dcn10_link_enc_registers *link_regs;
+	const struct dcn10_link_enc_aux_registers *aux_regs;
+	const struct dcn10_link_enc_hpd_registers *hpd_regs;
+	const struct dcn10_link_enc_shift *link_shift;
+	const struct dcn10_link_enc_mask *link_mask;
+};
+
+
+void dcn10_link_encoder_construct(
+	struct dcn10_link_encoder *enc10,
+	const struct encoder_init_data *init_data,
+	const struct encoder_feature_support *enc_features,
+	const struct dcn10_link_enc_registers *link_regs,
+	const struct dcn10_link_enc_aux_registers *aux_regs,
+	const struct dcn10_link_enc_hpd_registers *hpd_regs,
+	const struct dcn10_link_enc_shift *link_shift,
+	const struct dcn10_link_enc_mask *link_mask);
+
+bool dcn10_link_encoder_validate_dvi_output(
+	const struct dcn10_link_encoder *enc10,
+	enum signal_type connector_signal,
+	enum signal_type signal,
+	const struct dc_crtc_timing *crtc_timing);
+
+bool dcn10_link_encoder_validate_rgb_output(
+	const struct dcn10_link_encoder *enc10,
+	const struct dc_crtc_timing *crtc_timing);
+
+bool dcn10_link_encoder_validate_dp_output(
+	const struct dcn10_link_encoder *enc10,
+	const struct dc_crtc_timing *crtc_timing);
+
+bool dcn10_link_encoder_validate_wireless_output(
+	const struct dcn10_link_encoder *enc10,
+	const struct dc_crtc_timing *crtc_timing);
+
+bool dcn10_link_encoder_validate_output_with_stream(
+	struct link_encoder *enc,
+	const struct dc_stream_state *stream);
+
+/****************** HW programming ************************/
+
+/* initialize HW */  /* why do we initialze aux in here? */
+void dcn10_link_encoder_hw_init(struct link_encoder *enc);
+
+void dcn10_link_encoder_destroy(struct link_encoder **enc);
+
+/* program DIG_MODE in DIG_BE */
+/* TODO can this be combined with enable_output? */
+void dcn10_link_encoder_setup(
+	struct link_encoder *enc,
+	enum signal_type signal);
+
+/* enables TMDS PHY output */
+/* TODO: still need depth or just pass in adjusted pixel clock? */
+void dcn10_link_encoder_enable_tmds_output(
+	struct link_encoder *enc,
+	enum clock_source_id clock_source,
+	enum dc_color_depth color_depth,
+	enum signal_type signal,
+	uint32_t pixel_clock);
+
+/* enables DP PHY output */
+void dcn10_link_encoder_enable_dp_output(
+	struct link_encoder *enc,
+	const struct dc_link_settings *link_settings,
+	enum clock_source_id clock_source);
+
+/* enables DP PHY output in MST mode */
+void dcn10_link_encoder_enable_dp_mst_output(
+	struct link_encoder *enc,
+	const struct dc_link_settings *link_settings,
+	enum clock_source_id clock_source);
+
+/* disable PHY output */
+void dcn10_link_encoder_disable_output(
+	struct link_encoder *enc,
+	enum signal_type signal);
+
+/* set DP lane settings */
+void dcn10_link_encoder_dp_set_lane_settings(
+	struct link_encoder *enc,
+	const struct link_training_settings *link_settings);
+
+void dcn10_link_encoder_dp_set_phy_pattern(
+	struct link_encoder *enc,
+	const struct encoder_set_dp_phy_pattern_param *param);
+
+/* programs DP MST VC payload allocation */
+void dcn10_link_encoder_update_mst_stream_allocation_table(
+	struct link_encoder *enc,
+	const struct link_mst_stream_allocation_table *table);
+
+void dcn10_link_encoder_connect_dig_be_to_fe(
+	struct link_encoder *enc,
+	enum engine_id engine,
+	bool connect);
+
+void dcn10_link_encoder_set_dp_phy_pattern_training_pattern(
+	struct link_encoder *enc,
+	uint32_t index);
+
+void dcn10_link_encoder_enable_hpd(struct link_encoder *enc);
+
+void dcn10_link_encoder_disable_hpd(struct link_encoder *enc);
+
+void dcn10_psr_program_dp_dphy_fast_training(struct link_encoder *enc,
+			bool exit_link_training_required);
+
+void dcn10_psr_program_secondary_packet(struct link_encoder *enc,
+			unsigned int sdp_transmit_line_num_deadline);
+
+bool dcn10_is_dig_enabled(struct link_encoder *enc);
+
+#endif /* __DC_LINK_ENCODER__DCN10_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
index ace2e03..df5cb2d 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
@@ -38,7 +38,7 @@
 #include "dcn10/dcn10_hw_sequencer.h"
 #include "dce110/dce110_hw_sequencer.h"
 #include "dcn10/dcn10_opp.h"
-#include "dce/dce_link_encoder.h"
+#include "dcn10/dcn10_link_encoder.h"
 #include "dcn10/dcn10_stream_encoder.h"
 #include "dce/dce_clocks.h"
 #include "dce/dce_clock_source.h"
@@ -214,13 +214,11 @@ static const struct dce_aduio_mask audio_mask = {
 	AUX_REG_LIST(id)\
 }
 
-static const struct dce110_link_enc_aux_registers link_enc_aux_regs[] = {
+static const struct dcn10_link_enc_aux_registers link_enc_aux_regs[] = {
 		aux_regs(0),
 		aux_regs(1),
 		aux_regs(2),
-		aux_regs(3),
-		aux_regs(4),
-		aux_regs(5)
+		aux_regs(3)
 };
 
 #define hpd_regs(id)\
@@ -228,13 +226,11 @@ static const struct dce110_link_enc_aux_registers link_enc_aux_regs[] = {
 	HPD_REG_LIST(id)\
 }
 
-static const struct dce110_link_enc_hpd_registers link_enc_hpd_regs[] = {
+static const struct dcn10_link_enc_hpd_registers link_enc_hpd_regs[] = {
 		hpd_regs(0),
 		hpd_regs(1),
 		hpd_regs(2),
-		hpd_regs(3),
-		hpd_regs(4),
-		hpd_regs(5)
+		hpd_regs(3)
 };
 
 #define link_regs(id)\
@@ -243,14 +239,19 @@ static const struct dce110_link_enc_hpd_registers link_enc_hpd_regs[] = {
 	SRI(DP_DPHY_INTERNAL_CTRL, DP, id) \
 }
 
-static const struct dce110_link_enc_registers link_enc_regs[] = {
+static const struct dcn10_link_enc_registers link_enc_regs[] = {
 	link_regs(0),
 	link_regs(1),
 	link_regs(2),
-	link_regs(3),
-	link_regs(4),
-	link_regs(5),
-	link_regs(6),
+	link_regs(3)
+};
+
+static const struct dcn10_link_enc_shift le_shift = {
+		LINK_ENCODER_MASK_SH_LIST_DCN10(__SHIFT)
+};
+
+static const struct dcn10_link_enc_mask le_mask = {
+		LINK_ENCODER_MASK_SH_LIST_DCN10(_MASK)
 };
 
 #define ipp_regs(id)\
@@ -583,20 +584,22 @@ static const struct encoder_feature_support link_enc_feature = {
 struct link_encoder *dcn10_link_encoder_create(
 	const struct encoder_init_data *enc_init_data)
 {
-	struct dce110_link_encoder *enc110 =
-		kzalloc(sizeof(struct dce110_link_encoder), GFP_KERNEL);
+	struct dcn10_link_encoder *enc10 =
+		kzalloc(sizeof(struct dcn10_link_encoder), GFP_KERNEL);
 
-	if (!enc110)
+	if (!enc10)
 		return NULL;
 
-	dce110_link_encoder_construct(enc110,
+	dcn10_link_encoder_construct(enc10,
 				      enc_init_data,
 				      &link_enc_feature,
 				      &link_enc_regs[enc_init_data->transmitter],
 				      &link_enc_aux_regs[enc_init_data->channel - 1],
-				      &link_enc_hpd_regs[enc_init_data->hpd_source]);
+				      &link_enc_hpd_regs[enc_init_data->hpd_source],
+				      &le_shift,
+				      &le_mask);
 
-	return &enc110->base;
+	return &enc10->base;
 }
 
 struct clock_source *dcn10_clock_source_create(
-- 
cgit v1.1


From 9fcab85c580b31f6eb56dd3a00edd5f5270ad55c Mon Sep 17 00:00:00 2001
From: Anthony Koo <Anthony.Koo@amd.com>
Date: Fri, 27 Apr 2018 20:50:07 -0400
Subject: drm/amd/display: fix memory leaks

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/modules/stats/stats.c | 24 +++++++++++++++--------
 1 file changed, 16 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/modules/stats/stats.c b/drivers/gpu/drm/amd/display/modules/stats/stats.c
index fe9e4b3..3f7d47f 100644
--- a/drivers/gpu/drm/amd/display/modules/stats/stats.c
+++ b/drivers/gpu/drm/amd/display/modules/stats/stats.c
@@ -115,12 +115,12 @@ struct mod_stats *mod_stats_create(struct dc *dc)
 	unsigned int reg_data;
 	int i = 0;
 
+	if (dc == NULL)
+		goto fail_construct;
+
 	core_stats = kzalloc(sizeof(struct core_stats), GFP_KERNEL);
 
 	if (core_stats == NULL)
-		goto fail_alloc_context;
-
-	if (dc == NULL)
 		goto fail_construct;
 
 	core_stats->dc = dc;
@@ -146,6 +146,8 @@ struct mod_stats *mod_stats_create(struct dc *dc)
 				core_stats->entries,
 						GFP_KERNEL);
 
+		if (core_stats->time == NULL)
+			goto fail_construct_time;
 
 		core_stats->event_entries = DAL_STATS_EVENT_ENTRIES_DEFAULT;
 		core_stats->events = kzalloc(
@@ -153,13 +155,13 @@ struct mod_stats *mod_stats_create(struct dc *dc)
 				core_stats->event_entries,
 						GFP_KERNEL);
 
+		if (core_stats->events == NULL)
+			goto fail_construct_events;
+
 	} else {
 		core_stats->entries = 0;
 	}
 
-	if (core_stats->time == NULL)
-		goto fail_construct;
-
 	/* Purposely leave index 0 unused so we don't need special logic to
 	 * handle calculation cases that depend on previous flip data.
 	 */
@@ -171,10 +173,13 @@ struct mod_stats *mod_stats_create(struct dc *dc)
 
 	return &core_stats->public;
 
-fail_construct:
+fail_construct_events:
+	kfree(core_stats->time);
+
+fail_construct_time:
 	kfree(core_stats);
 
-fail_alloc_context:
+fail_construct:
 	return NULL;
 }
 
@@ -186,6 +191,9 @@ void mod_stats_destroy(struct mod_stats *mod_stats)
 		if (core_stats->time != NULL)
 			kfree(core_stats->time);
 
+		if (core_stats->events != NULL)
+			kfree(core_stats->events);
+
 		kfree(core_stats);
 	}
 }
-- 
cgit v1.1


From 5326c4525d1b2d5f1519268dd305e19c9bd4ef56 Mon Sep 17 00:00:00 2001
From: Mikita Lipski <mikita.lipski@amd.com>
Date: Fri, 27 Apr 2018 09:09:52 -0400
Subject: drm/amd/display: Clear connector's edid pointer

Clear connector's edid pointer on coonnector update, when unplugging
the display.

Fix poison EDID when hotplugging on previously used connector.

Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Cc: stable@vger.kernel.org
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 6d0dc1f..1ce10bc 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -911,6 +911,7 @@ amdgpu_dm_update_connector_after_detect(struct amdgpu_dm_connector *aconnector)
 		drm_mode_connector_update_edid_property(connector, NULL);
 		aconnector->num_modes = 0;
 		aconnector->dc_sink = NULL;
+		aconnector->edid = NULL;
 	}
 
 	mutex_unlock(&dev->mode_config.mutex);
-- 
cgit v1.1


From cd3cb7c08754cd5dd1cbccfc2296d6b7dde511f2 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Thu, 17 May 2018 20:21:42 +0800
Subject: drm/amd/pp: Fix build warning in vegam

warning: missing braces around initializer [-Wmissing-braces]

Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c    | 2 +-
 drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c | 4 +++-
 2 files changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c
index cf99c5e..ec38c9f 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c
@@ -320,7 +320,7 @@ int atomctrl_get_memory_pll_dividers_ai(struct pp_hwmgr *hwmgr,
 					pp_atomctrl_memory_clock_param_ai *mpll_param)
 {
 	struct amdgpu_device *adev = hwmgr->adev;
-	COMPUTE_MEMORY_CLOCK_PARAM_PARAMETERS_V2_3 mpll_parameters = {0};
+	COMPUTE_MEMORY_CLOCK_PARAM_PARAMETERS_V2_3 mpll_parameters = {{0}, 0, 0};
 	int result;
 
 	mpll_parameters.ulClock.ulClock = cpu_to_le32(clock_value);
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c
index c9a5633..a40f714 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c
@@ -1366,10 +1366,12 @@ static int vegam_program_memory_timing_parameters(struct pp_hwmgr *hwmgr)
 {
 	struct smu7_hwmgr *hw_data = (struct smu7_hwmgr *)(hwmgr->backend);
 	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
-	struct SMU75_Discrete_MCArbDramTimingTable arb_regs = {0};
+	struct SMU75_Discrete_MCArbDramTimingTable arb_regs;
 	uint32_t i, j;
 	int result = 0;
 
+	memset(&arb_regs, 0, sizeof(SMU75_Discrete_MCArbDramTimingTable));
+
 	for (i = 0; i < hw_data->dpm_table.sclk_table.count; i++) {
 		for (j = 0; j < hw_data->dpm_table.mclk_table.count; j++) {
 			result = vegam_populate_memory_timing_parameters(hwmgr,
-- 
cgit v1.1


From 6ee21dbfe9a79edf6f09d5f3ab1f3c4f0699dbf2 Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Thu, 17 May 2018 13:31:49 -0400
Subject: drm/amdgpu: fix insert nop for VCN decode ring
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

NO_OP register should be writen to 0

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
index 0501746b..7fbbdb1 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
@@ -1048,14 +1048,17 @@ static int vcn_v1_0_process_interrupt(struct amdgpu_device *adev,
 	return 0;
 }
 
-static void vcn_v1_0_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count)
+static void vcn_v1_0_dec_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count)
 {
-	int i;
 	struct amdgpu_device *adev = ring->adev;
+	int i;
 
-	for (i = 0; i < count; i++)
-		amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_NO_OP), 0));
+	WARN_ON(ring->wptr % 2 || count % 2);
 
+	for (i = 0; i < count / 2; i++) {
+		amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_NO_OP), 0));
+		amdgpu_ring_write(ring, 0);
+	}
 }
 
 
@@ -1082,7 +1085,6 @@ static const struct amd_ip_funcs vcn_v1_0_ip_funcs = {
 static const struct amdgpu_ring_funcs vcn_v1_0_dec_ring_vm_funcs = {
 	.type = AMDGPU_RING_TYPE_VCN_DEC,
 	.align_mask = 0xf,
-	.nop = PACKET0(0x81ff, 0),
 	.support_64bit_ptrs = false,
 	.vmhub = AMDGPU_MMHUB,
 	.get_rptr = vcn_v1_0_dec_ring_get_rptr,
@@ -1101,7 +1103,7 @@ static const struct amdgpu_ring_funcs vcn_v1_0_dec_ring_vm_funcs = {
 	.emit_vm_flush = vcn_v1_0_dec_ring_emit_vm_flush,
 	.test_ring = amdgpu_vcn_dec_ring_test_ring,
 	.test_ib = amdgpu_vcn_dec_ring_test_ib,
-	.insert_nop = vcn_v1_0_ring_insert_nop,
+	.insert_nop = vcn_v1_0_dec_ring_insert_nop,
 	.insert_start = vcn_v1_0_dec_ring_insert_start,
 	.insert_end = vcn_v1_0_dec_ring_insert_end,
 	.pad_ib = amdgpu_ring_generic_pad_ib,
-- 
cgit v1.1


From cbb7a239117d45d512fae1806cc7722f68c7b82f Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Thu, 17 May 2018 13:37:50 -0400
Subject: drm/amdgpu: fix insert nop for UVD7 ring
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

NO_OP register should be writen to 0

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
index f9a54821..57d32f2 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
@@ -1325,12 +1325,15 @@ static void uvd_v7_0_ring_emit_vm_flush(struct amdgpu_ring *ring,
 
 static void uvd_v7_0_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count)
 {
-	int i;
 	struct amdgpu_device *adev = ring->adev;
+	int i;
 
-	for (i = 0; i < count; i++)
-		amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_NO_OP), 0));
+	WARN_ON(ring->wptr % 2 || count % 2);
 
+	for (i = 0; i < count / 2; i++) {
+		amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_NO_OP), 0));
+		amdgpu_ring_write(ring, 0);
+	}
 }
 
 static void uvd_v7_0_enc_ring_insert_end(struct amdgpu_ring *ring)
@@ -1710,7 +1713,6 @@ const struct amd_ip_funcs uvd_v7_0_ip_funcs = {
 static const struct amdgpu_ring_funcs uvd_v7_0_ring_vm_funcs = {
 	.type = AMDGPU_RING_TYPE_UVD,
 	.align_mask = 0xf,
-	.nop = PACKET0(0x81ff, 0),
 	.support_64bit_ptrs = false,
 	.vmhub = AMDGPU_MMHUB,
 	.get_rptr = uvd_v7_0_ring_get_rptr,
-- 
cgit v1.1


From 1aac3c918036d6bb0075281d431da3844a058d00 Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Thu, 17 May 2018 13:44:28 -0400
Subject: drm/amdgpu: fix insert nop for UVD6 ring
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

NO_OP register should be writen to 0

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c | 15 +++++++++++++--
 1 file changed, 13 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
index dc39169..bfddf97 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
@@ -1100,6 +1100,18 @@ static void uvd_v6_0_ring_emit_pipeline_sync(struct amdgpu_ring *ring)
 	amdgpu_ring_write(ring, 0xE);
 }
 
+static void uvd_v6_0_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count)
+{
+	int i;
+
+	WARN_ON(ring->wptr % 2 || count % 2);
+
+	for (i = 0; i < count / 2; i++) {
+		amdgpu_ring_write(ring, PACKET0(mmUVD_NO_OP, 0));
+		amdgpu_ring_write(ring, 0);
+	}
+}
+
 static void uvd_v6_0_enc_ring_emit_pipeline_sync(struct amdgpu_ring *ring)
 {
 	uint32_t seq = ring->fence_drv.sync_seq;
@@ -1532,7 +1544,6 @@ static const struct amd_ip_funcs uvd_v6_0_ip_funcs = {
 static const struct amdgpu_ring_funcs uvd_v6_0_ring_phys_funcs = {
 	.type = AMDGPU_RING_TYPE_UVD,
 	.align_mask = 0xf,
-	.nop = PACKET0(mmUVD_NO_OP, 0),
 	.support_64bit_ptrs = false,
 	.get_rptr = uvd_v6_0_ring_get_rptr,
 	.get_wptr = uvd_v6_0_ring_get_wptr,
@@ -1548,7 +1559,7 @@ static const struct amdgpu_ring_funcs uvd_v6_0_ring_phys_funcs = {
 	.emit_hdp_flush = uvd_v6_0_ring_emit_hdp_flush,
 	.test_ring = uvd_v6_0_ring_test_ring,
 	.test_ib = amdgpu_uvd_ring_test_ib,
-	.insert_nop = amdgpu_ring_insert_nop,
+	.insert_nop = uvd_v6_0_ring_insert_nop,
 	.pad_ib = amdgpu_ring_generic_pad_ib,
 	.begin_use = amdgpu_uvd_ring_begin_use,
 	.end_use = amdgpu_uvd_ring_end_use,
-- 
cgit v1.1


From 0232e30623f3761ce9350328d4d96cea8372b114 Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Thu, 17 May 2018 13:52:00 -0400
Subject: drm/amdgpu: fix insert nop for UVD5 ring
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

NO_OP register should be writen to 0

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c | 15 +++++++++++++--
 1 file changed, 13 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
index f5d074a..341ee6d 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
@@ -541,6 +541,18 @@ static void uvd_v5_0_ring_emit_ib(struct amdgpu_ring *ring,
 	amdgpu_ring_write(ring, ib->length_dw);
 }
 
+static void uvd_v5_0_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count)
+{
+	int i;
+
+	WARN_ON(ring->wptr % 2 || count % 2);
+
+	for (i = 0; i < count / 2; i++) {
+		amdgpu_ring_write(ring, PACKET0(mmUVD_NO_OP, 0));
+		amdgpu_ring_write(ring, 0);
+	}
+}
+
 static bool uvd_v5_0_is_idle(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
@@ -841,7 +853,6 @@ static const struct amd_ip_funcs uvd_v5_0_ip_funcs = {
 static const struct amdgpu_ring_funcs uvd_v5_0_ring_funcs = {
 	.type = AMDGPU_RING_TYPE_UVD,
 	.align_mask = 0xf,
-	.nop = PACKET0(mmUVD_NO_OP, 0),
 	.support_64bit_ptrs = false,
 	.get_rptr = uvd_v5_0_ring_get_rptr,
 	.get_wptr = uvd_v5_0_ring_get_wptr,
@@ -854,7 +865,7 @@ static const struct amdgpu_ring_funcs uvd_v5_0_ring_funcs = {
 	.emit_fence = uvd_v5_0_ring_emit_fence,
 	.test_ring = uvd_v5_0_ring_test_ring,
 	.test_ib = amdgpu_uvd_ring_test_ib,
-	.insert_nop = amdgpu_ring_insert_nop,
+	.insert_nop = uvd_v5_0_ring_insert_nop,
 	.pad_ib = amdgpu_ring_generic_pad_ib,
 	.begin_use = amdgpu_uvd_ring_begin_use,
 	.end_use = amdgpu_uvd_ring_end_use,
-- 
cgit v1.1


From def139037bbf9195467fa83c0a299d666e6ed0bb Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Thu, 17 May 2018 13:54:21 -0400
Subject: drm/amdgpu: fix insert nop for UVD4.2 ring
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

NO_OP register should be writen to 0

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c | 15 +++++++++++++--
 1 file changed, 13 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
index 5f22135d..6fed3d7 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
@@ -524,6 +524,18 @@ static void uvd_v4_2_ring_emit_ib(struct amdgpu_ring *ring,
 	amdgpu_ring_write(ring, ib->length_dw);
 }
 
+static void uvd_v4_2_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count)
+{
+	int i;
+
+	WARN_ON(ring->wptr % 2 || count % 2);
+
+	for (i = 0; i < count / 2; i++) {
+		amdgpu_ring_write(ring, PACKET0(mmUVD_NO_OP, 0));
+		amdgpu_ring_write(ring, 0);
+	}
+}
+
 /**
  * uvd_v4_2_mc_resume - memory controller programming
  *
@@ -733,7 +745,6 @@ static const struct amd_ip_funcs uvd_v4_2_ip_funcs = {
 static const struct amdgpu_ring_funcs uvd_v4_2_ring_funcs = {
 	.type = AMDGPU_RING_TYPE_UVD,
 	.align_mask = 0xf,
-	.nop = PACKET0(mmUVD_NO_OP, 0),
 	.support_64bit_ptrs = false,
 	.get_rptr = uvd_v4_2_ring_get_rptr,
 	.get_wptr = uvd_v4_2_ring_get_wptr,
@@ -746,7 +757,7 @@ static const struct amdgpu_ring_funcs uvd_v4_2_ring_funcs = {
 	.emit_fence = uvd_v4_2_ring_emit_fence,
 	.test_ring = uvd_v4_2_ring_test_ring,
 	.test_ib = amdgpu_uvd_ring_test_ib,
-	.insert_nop = amdgpu_ring_insert_nop,
+	.insert_nop = uvd_v4_2_ring_insert_nop,
 	.pad_ib = amdgpu_ring_generic_pad_ib,
 	.begin_use = amdgpu_uvd_ring_begin_use,
 	.end_use = amdgpu_uvd_ring_end_use,
-- 
cgit v1.1


From 6134534ca24f42043cacdd7108026803577f6c59 Mon Sep 17 00:00:00 2001
From: Thierry Reding <treding@nvidia.com>
Date: Wed, 16 May 2018 16:43:11 +0200
Subject: drm/tegra: Add kerneldoc for UAPI

Document the userspace ABI with kerneldoc to provide some information on
how to use it.

v3:
- reword description of arrays and array lengths

v2:
- keep GEM object creation flags for ABI compatibility
- fix typo in struct drm_tegra_syncpt_incr kerneldoc
- fix typos in struct drm_tegra_submit kerneldoc
- reworded some descriptions as suggested

Reviewed-by: Dmitry Osipenko <digetx@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 include/uapi/drm/tegra_drm.h | 490 ++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 481 insertions(+), 9 deletions(-)

diff --git a/include/uapi/drm/tegra_drm.h b/include/uapi/drm/tegra_drm.h
index 99e15d8..c4df3c3 100644
--- a/include/uapi/drm/tegra_drm.h
+++ b/include/uapi/drm/tegra_drm.h
@@ -32,143 +32,615 @@ extern "C" {
 #define DRM_TEGRA_GEM_CREATE_TILED     (1 << 0)
 #define DRM_TEGRA_GEM_CREATE_BOTTOM_UP (1 << 1)
 
+/**
+ * struct drm_tegra_gem_create - parameters for the GEM object creation IOCTL
+ */
 struct drm_tegra_gem_create {
+	/**
+	 * @size:
+	 *
+	 * The size, in bytes, of the buffer object to be created.
+	 */
 	__u64 size;
+
+	/**
+	 * @flags:
+	 *
+	 * A bitmask of flags that influence the creation of GEM objects:
+	 *
+	 * DRM_TEGRA_GEM_CREATE_TILED
+	 *   Use the 16x16 tiling format for this buffer.
+	 *
+	 * DRM_TEGRA_GEM_CREATE_BOTTOM_UP
+	 *   The buffer has a bottom-up layout.
+	 */
 	__u32 flags;
+
+	/**
+	 * @handle:
+	 *
+	 * The handle of the created GEM object. Set by the kernel upon
+	 * successful completion of the IOCTL.
+	 */
 	__u32 handle;
 };
 
+/**
+ * struct drm_tegra_gem_mmap - parameters for the GEM mmap IOCTL
+ */
 struct drm_tegra_gem_mmap {
+	/**
+	 * @handle:
+	 *
+	 * Handle of the GEM object to obtain an mmap offset for.
+	 */
 	__u32 handle;
+
+	/**
+	 * @pad:
+	 *
+	 * Structure padding that may be used in the future. Must be 0.
+	 */
 	__u32 pad;
+
+	/**
+	 * @offset:
+	 *
+	 * The mmap offset for the given GEM object. Set by the kernel upon
+	 * successful completion of the IOCTL.
+	 */
 	__u64 offset;
 };
 
+/**
+ * struct drm_tegra_syncpt_read - parameters for the read syncpoint IOCTL
+ */
 struct drm_tegra_syncpt_read {
+	/**
+	 * @id:
+	 *
+	 * ID of the syncpoint to read the current value from.
+	 */
 	__u32 id;
+
+	/**
+	 * @value:
+	 *
+	 * The current syncpoint value. Set by the kernel upon successful
+	 * completion of the IOCTL.
+	 */
 	__u32 value;
 };
 
+/**
+ * struct drm_tegra_syncpt_incr - parameters for the increment syncpoint IOCTL
+ */
 struct drm_tegra_syncpt_incr {
+	/**
+	 * @id:
+	 *
+	 * ID of the syncpoint to increment.
+	 */
 	__u32 id;
+
+	/**
+	 * @pad:
+	 *
+	 * Structure padding that may be used in the future. Must be 0.
+	 */
 	__u32 pad;
 };
 
+/**
+ * struct drm_tegra_syncpt_wait - parameters for the wait syncpoint IOCTL
+ */
 struct drm_tegra_syncpt_wait {
+	/**
+	 * @id:
+	 *
+	 * ID of the syncpoint to wait on.
+	 */
 	__u32 id;
+
+	/**
+	 * @thresh:
+	 *
+	 * Threshold value for which to wait.
+	 */
 	__u32 thresh;
+
+	/**
+	 * @timeout:
+	 *
+	 * Timeout, in milliseconds, to wait.
+	 */
 	__u32 timeout;
+
+	/**
+	 * @value:
+	 *
+	 * The new syncpoint value after the wait. Set by the kernel upon
+	 * successful completion of the IOCTL.
+	 */
 	__u32 value;
 };
 
 #define DRM_TEGRA_NO_TIMEOUT	(0xffffffff)
 
+/**
+ * struct drm_tegra_open_channel - parameters for the open channel IOCTL
+ */
 struct drm_tegra_open_channel {
+	/**
+	 * @client:
+	 *
+	 * The client ID for this channel.
+	 */
 	__u32 client;
+
+	/**
+	 * @pad:
+	 *
+	 * Structure padding that may be used in the future. Must be 0.
+	 */
 	__u32 pad;
+
+	/**
+	 * @context:
+	 *
+	 * The application context of this channel. Set by the kernel upon
+	 * successful completion of the IOCTL. This context needs to be passed
+	 * to the DRM_TEGRA_CHANNEL_CLOSE or the DRM_TEGRA_SUBMIT IOCTLs.
+	 */
 	__u64 context;
 };
 
+/**
+ * struct drm_tegra_close_channel - parameters for the close channel IOCTL
+ */
 struct drm_tegra_close_channel {
+	/**
+	 * @context:
+	 *
+	 * The application context of this channel. This is obtained from the
+	 * DRM_TEGRA_OPEN_CHANNEL IOCTL.
+	 */
 	__u64 context;
 };
 
+/**
+ * struct drm_tegra_get_syncpt - parameters for the get syncpoint IOCTL
+ */
 struct drm_tegra_get_syncpt {
+	/**
+	 * @context:
+	 *
+	 * The application context identifying the channel for which to obtain
+	 * the syncpoint ID.
+	 */
 	__u64 context;
+
+	/**
+	 * @index:
+	 *
+	 * Index of the client syncpoint for which to obtain the ID.
+	 */
 	__u32 index;
+
+	/**
+	 * @id:
+	 *
+	 * The ID of the given syncpoint. Set by the kernel upon successful
+	 * completion of the IOCTL.
+	 */
 	__u32 id;
 };
 
+/**
+ * struct drm_tegra_get_syncpt_base - parameters for the get wait base IOCTL
+ */
 struct drm_tegra_get_syncpt_base {
+	/**
+	 * @context:
+	 *
+	 * The application context identifying for which channel to obtain the
+	 * wait base.
+	 */
 	__u64 context;
+
+	/**
+	 * @syncpt:
+	 *
+	 * ID of the syncpoint for which to obtain the wait base.
+	 */
 	__u32 syncpt;
+
+	/**
+	 * @id:
+	 *
+	 * The ID of the wait base corresponding to the client syncpoint. Set
+	 * by the kernel upon successful completion of the IOCTL.
+	 */
 	__u32 id;
 };
 
+/**
+ * struct drm_tegra_syncpt - syncpoint increment operation
+ */
 struct drm_tegra_syncpt {
+	/**
+	 * @id:
+	 *
+	 * ID of the syncpoint to operate on.
+	 */
 	__u32 id;
+
+	/**
+	 * @incrs:
+	 *
+	 * Number of increments to perform for the syncpoint.
+	 */
 	__u32 incrs;
 };
 
+/**
+ * struct drm_tegra_cmdbuf - structure describing a command buffer
+ */
 struct drm_tegra_cmdbuf {
+	/**
+	 * @handle:
+	 *
+	 * Handle to a GEM object containing the command buffer.
+	 */
 	__u32 handle;
+
+	/**
+	 * @offset:
+	 *
+	 * Offset, in bytes, into the GEM object identified by @handle at
+	 * which the command buffer starts.
+	 */
 	__u32 offset;
+
+	/**
+	 * @words:
+	 *
+	 * Number of 32-bit words in this command buffer.
+	 */
 	__u32 words;
+
+	/**
+	 * @pad:
+	 *
+	 * Structure padding that may be used in the future. Must be 0.
+	 */
 	__u32 pad;
 };
 
+/**
+ * struct drm_tegra_reloc - GEM object relocation structure
+ */
 struct drm_tegra_reloc {
 	struct {
+		/**
+		 * @cmdbuf.handle:
+		 *
+		 * Handle to the GEM object containing the command buffer for
+		 * which to perform this GEM object relocation.
+		 */
 		__u32 handle;
+
+		/**
+		 * @cmdbuf.offset:
+		 *
+		 * Offset, in bytes, into the command buffer at which to
+		 * insert the relocated address.
+		 */
 		__u32 offset;
 	} cmdbuf;
 	struct {
+		/**
+		 * @target.handle:
+		 *
+		 * Handle to the GEM object to be relocated.
+		 */
 		__u32 handle;
+
+		/**
+		 * @target.offset:
+		 *
+		 * Offset, in bytes, into the target GEM object at which the
+		 * relocated data starts.
+		 */
 		__u32 offset;
 	} target;
+
+	/**
+	 * @shift:
+	 *
+	 * The number of bits by which to shift relocated addresses.
+	 */
 	__u32 shift;
+
+	/**
+	 * @pad:
+	 *
+	 * Structure padding that may be used in the future. Must be 0.
+	 */
 	__u32 pad;
 };
 
+/**
+ * struct drm_tegra_waitchk - wait check structure
+ */
 struct drm_tegra_waitchk {
+	/**
+	 * @handle:
+	 *
+	 * Handle to the GEM object containing a command stream on which to
+	 * perform the wait check.
+	 */
 	__u32 handle;
+
+	/**
+	 * @offset:
+	 *
+	 * Offset, in bytes, of the location in the command stream to perform
+	 * the wait check on.
+	 */
 	__u32 offset;
+
+	/**
+	 * @syncpt:
+	 *
+	 * ID of the syncpoint to wait check.
+	 */
 	__u32 syncpt;
+
+	/**
+	 * @thresh:
+	 *
+	 * Threshold value for which to check.
+	 */
 	__u32 thresh;
 };
 
+/**
+ * struct drm_tegra_submit - job submission structure
+ */
 struct drm_tegra_submit {
+	/**
+	 * @context:
+	 *
+	 * The application context identifying the channel to use for the
+	 * execution of this job.
+	 */
 	__u64 context;
+
+	/**
+	 * @num_syncpts:
+	 *
+	 * The number of syncpoints operated on by this job. This defines the
+	 * length of the array pointed to by @syncpts.
+	 */
 	__u32 num_syncpts;
+
+	/**
+	 * @num_cmdbufs:
+	 *
+	 * The number of command buffers to execute as part of this job. This
+	 * defines the length of the array pointed to by @cmdbufs.
+	 */
 	__u32 num_cmdbufs;
+
+	/**
+	 * @num_relocs:
+	 *
+	 * The number of relocations to perform before executing this job.
+	 * This defines the length of the array pointed to by @relocs.
+	 */
 	__u32 num_relocs;
+
+	/**
+	 * @num_waitchks:
+	 *
+	 * The number of wait checks to perform as part of this job. This
+	 * defines the length of the array pointed to by @waitchks.
+	 */
 	__u32 num_waitchks;
+
+	/**
+	 * @waitchk_mask:
+	 *
+	 * Bitmask of valid wait checks.
+	 */
 	__u32 waitchk_mask;
+
+	/**
+	 * @timeout:
+	 *
+	 * Timeout, in milliseconds, before this job is cancelled.
+	 */
 	__u32 timeout;
+
+	/**
+	 * @syncpts:
+	 *
+	 * A pointer to an array of &struct drm_tegra_syncpt structures that
+	 * specify the syncpoint operations performed as part of this job.
+	 * The number of elements in the array must be equal to the value
+	 * given by @num_syncpts.
+	 */
 	__u64 syncpts;
+
+	/**
+	 * @cmdbufs:
+	 *
+	 * A pointer to an array of &struct drm_tegra_cmdbuf structures that
+	 * define the command buffers to execute as part of this job. The
+	 * number of elements in the array must be equal to the value given
+	 * by @num_syncpts.
+	 */
 	__u64 cmdbufs;
+
+	/**
+	 * @relocs:
+	 *
+	 * A pointer to an array of &struct drm_tegra_reloc structures that
+	 * specify the relocations that need to be performed before executing
+	 * this job. The number of elements in the array must be equal to the
+	 * value given by @num_relocs.
+	 */
 	__u64 relocs;
+
+	/**
+	 * @waitchks:
+	 *
+	 * A pointer to an array of &struct drm_tegra_waitchk structures that
+	 * specify the wait checks to be performed while executing this job.
+	 * The number of elements in the array must be equal to the value
+	 * given by @num_waitchks.
+	 */
 	__u64 waitchks;
-	__u32 fence;		/* Return value */
 
-	__u32 reserved[5];	/* future expansion */
+	/**
+	 * @fence:
+	 *
+	 * The threshold of the syncpoint associated with this job after it
+	 * has been completed. Set by the kernel upon successful completion of
+	 * the IOCTL. This can be used with the DRM_TEGRA_SYNCPT_WAIT IOCTL to
+	 * wait for this job to be finished.
+	 */
+	__u32 fence;
+
+	/**
+	 * @reserved:
+	 *
+	 * This field is reserved for future use. Must be 0.
+	 */
+	__u32 reserved[5];
 };
 
 #define DRM_TEGRA_GEM_TILING_MODE_PITCH 0
 #define DRM_TEGRA_GEM_TILING_MODE_TILED 1
 #define DRM_TEGRA_GEM_TILING_MODE_BLOCK 2
 
+/**
+ * struct drm_tegra_gem_set_tiling - parameters for the set tiling IOCTL
+ */
 struct drm_tegra_gem_set_tiling {
-	/* input */
+	/**
+	 * @handle:
+	 *
+	 * Handle to the GEM object for which to set the tiling parameters.
+	 */
 	__u32 handle;
+
+	/**
+	 * @mode:
+	 *
+	 * The tiling mode to set. Must be one of:
+	 *
+	 * DRM_TEGRA_GEM_TILING_MODE_PITCH
+	 *   pitch linear format
+	 *
+	 * DRM_TEGRA_GEM_TILING_MODE_TILED
+	 *   16x16 tiling format
+	 *
+	 * DRM_TEGRA_GEM_TILING_MODE_BLOCK
+	 *   16Bx2 tiling format
+	 */
 	__u32 mode;
+
+	/**
+	 * @value:
+	 *
+	 * The value to set for the tiling mode parameter.
+	 */
 	__u32 value;
+
+	/**
+	 * @pad:
+	 *
+	 * Structure padding that may be used in the future. Must be 0.
+	 */
 	__u32 pad;
 };
 
+/**
+ * struct drm_tegra_gem_get_tiling - parameters for the get tiling IOCTL
+ */
 struct drm_tegra_gem_get_tiling {
-	/* input */
+	/**
+	 * @handle:
+	 *
+	 * Handle to the GEM object for which to query the tiling parameters.
+	 */
 	__u32 handle;
-	/* output */
+
+	/**
+	 * @mode:
+	 *
+	 * The tiling mode currently associated with the GEM object. Set by
+	 * the kernel upon successful completion of the IOCTL.
+	 */
 	__u32 mode;
+
+	/**
+	 * @value:
+	 *
+	 * The tiling mode parameter currently associated with the GEM object.
+	 * Set by the kernel upon successful completion of the IOCTL.
+	 */
 	__u32 value;
+
+	/**
+	 * @pad:
+	 *
+	 * Structure padding that may be used in the future. Must be 0.
+	 */
 	__u32 pad;
 };
 
 #define DRM_TEGRA_GEM_BOTTOM_UP		(1 << 0)
 #define DRM_TEGRA_GEM_FLAGS		(DRM_TEGRA_GEM_BOTTOM_UP)
 
+/**
+ * struct drm_tegra_gem_set_flags - parameters for the set flags IOCTL
+ */
 struct drm_tegra_gem_set_flags {
-	/* input */
+	/**
+	 * @handle:
+	 *
+	 * Handle to the GEM object for which to set the flags.
+	 */
 	__u32 handle;
-	/* output */
+
+	/**
+	 * @flags:
+	 *
+	 * The flags to set for the GEM object.
+	 */
 	__u32 flags;
 };
 
+/**
+ * struct drm_tegra_gem_get_flags - parameters for the get flags IOCTL
+ */
 struct drm_tegra_gem_get_flags {
-	/* input */
+	/**
+	 * @handle:
+	 *
+	 * Handle to the GEM object for which to query the flags.
+	 */
 	__u32 handle;
-	/* output */
+
+	/**
+	 * @flags:
+	 *
+	 * The flags currently associated with the GEM object. Set by the
+	 * kernel upon successful completion of the IOCTL.
+	 */
 	__u32 flags;
 };
 
-- 
cgit v1.1


From 63e482f172f4680a4ff00305dd140589ae270306 Mon Sep 17 00:00:00 2001
From: Fabio Estevam <fabio.estevam@nxp.com>
Date: Sun, 20 May 2018 12:03:54 -0300
Subject: MAINTAINERS: drm: fsl-dcu: Update to Alison's NXP email address

The freescale.com email domain is not valid anymore, so use
the nxp.com domain instead.

Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
Signed-off-by: Stefan Agner <stefan@agner.ch>
---
 MAINTAINERS | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/MAINTAINERS b/MAINTAINERS
index 0a1410d..268bcfd 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -4665,7 +4665,7 @@ F:	Documentation/devicetree/bindings/display/exynos/
 
 DRM DRIVERS FOR FREESCALE DCU
 M:	Stefan Agner <stefan@agner.ch>
-M:	Alison Wang <alison.wang@freescale.com>
+M:	Alison Wang <alison.wang@nxp.com>
 L:	dri-devel@lists.freedesktop.org
 S:	Supported
 F:	drivers/gpu/drm/fsl-dcu/
-- 
cgit v1.1


From e05f3ea260d6db1fb9aede6af29d7cf8f491f0ed Mon Sep 17 00:00:00 2001
From: Marek Szyprowski <m.szyprowski@samsung.com>
Date: Wed, 23 May 2018 13:00:00 +0200
Subject: dt-bindings: exynos5433-decon: add more required clocks

Proper support for Decon's hardware window no 4 and 5 require enabling
a few more clocks ("aclk_smmu_decon1x", "aclk_xiu_decon1x",
"pclk_smmu_decon1x"). Add those clocks to required clocks list in Decon's
dt bindings.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Acked-by: Krzysztof Kozlowski <krzk@kernel.org>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
---
 .../devicetree/bindings/display/exynos/exynos5433-decon.txt      | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/Documentation/devicetree/bindings/display/exynos/exynos5433-decon.txt b/Documentation/devicetree/bindings/display/exynos/exynos5433-decon.txt
index fc25882..775193e 100644
--- a/Documentation/devicetree/bindings/display/exynos/exynos5433-decon.txt
+++ b/Documentation/devicetree/bindings/display/exynos/exynos5433-decon.txt
@@ -19,7 +19,8 @@ Required properties:
 	  clock-names property.
 - clock-names: list of clock names sorted in the same order as the clocks
 	       property. Must contain "pclk", "aclk_decon", "aclk_smmu_decon0x",
-	       "aclk_xiu_decon0x", "pclk_smmu_decon0x", clk_decon_vclk",
+	       "aclk_xiu_decon0x", "pclk_smmu_decon0x", "aclk_smmu_decon1x",
+	       "aclk_xiu_decon1x", "pclk_smmu_decon1x", clk_decon_vclk",
 	       "sclk_decon_eclk"
 - ports: contains a port which is connected to mic node. address-cells and
 	 size-cells must 1 and 0, respectively.
@@ -34,10 +35,14 @@ decon: decon@13800000 {
 	clocks = <&cmu_disp CLK_ACLK_DECON>, <&cmu_disp CLK_ACLK_SMMU_DECON0X>,
 		<&cmu_disp CLK_ACLK_XIU_DECON0X>,
 		<&cmu_disp CLK_PCLK_SMMU_DECON0X>,
+		<&cmu_disp CLK_ACLK_SMMU_DECON1X>,
+		<&cmu_disp CLK_ACLK_XIU_DECON1X>,
+		<&cmu_disp CLK_PCLK_SMMU_DECON1X>,
 		<&cmu_disp CLK_SCLK_DECON_VCLK>,
 		<&cmu_disp CLK_SCLK_DECON_ECLK>;
 	clock-names = "aclk_decon", "aclk_smmu_decon0x", "aclk_xiu_decon0x",
-		"pclk_smmu_decon0x", "sclk_decon_vclk", "sclk_decon_eclk";
+		"pclk_smmu_decon0x", "aclk_smmu_decon1x", "aclk_xiu_decon1x",
+		"pclk_smmu_decon1x", "sclk_decon_vclk", "sclk_decon_eclk";
 	interrupt-names = "vsync", "lcd_sys";
 	interrupts = <0 202 0>, <0 203 0>;
 
-- 
cgit v1.1


From cb5fba715babc599e82f40d53b6b956efcc8fc25 Mon Sep 17 00:00:00 2001
From: Marek Szyprowski <m.szyprowski@samsung.com>
Date: Wed, 23 May 2018 13:00:02 +0200
Subject: drm/exynos: decon: Add support for hardware windows no 4 and 5

Enable support for 2 more hardware windows. This require enabling a few
more clocks and set proper plane type for all windows. In the new
configuration primary plane uses hardware window no 3 and cursor uses
window no 5. The remaining hardware windows are used for overlays. This
gives us an overlay plane both below and above primary plane for both
Decon and DeconTV (which uses hardware window nr 0 for background).

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
---
 drivers/gpu/drm/exynos/exynos5433_drm_decon.c | 21 ++++++++++++---------
 1 file changed, 12 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/exynos/exynos5433_drm_decon.c b/drivers/gpu/drm/exynos/exynos5433_drm_decon.c
index 1c330f2..82c95c3 100644
--- a/drivers/gpu/drm/exynos/exynos5433_drm_decon.c
+++ b/drivers/gpu/drm/exynos/exynos5433_drm_decon.c
@@ -31,7 +31,10 @@
 #define DSD_CFG_MUX 0x1004
 #define DSD_CFG_MUX_TE_UNMASK_GLOBAL BIT(13)
 
-#define WINDOWS_NR	3
+#define WINDOWS_NR	5
+#define PRIMARY_WIN	2
+#define CURSON_WIN	4
+
 #define MIN_FB_WIDTH_FOR_16WORD_BURST	128
 
 #define I80_HW_TRG	(1 << 0)
@@ -43,6 +46,9 @@ static const char * const decon_clks_name[] = {
 	"aclk_smmu_decon0x",
 	"aclk_xiu_decon0x",
 	"pclk_smmu_decon0x",
+	"aclk_smmu_decon1x",
+	"aclk_xiu_decon1x",
+	"pclk_smmu_decon1x",
 	"sclk_decon_vclk",
 	"sclk_decon_eclk",
 };
@@ -74,9 +80,8 @@ static const uint32_t decon_formats[] = {
 };
 
 static const enum drm_plane_type decon_win_types[WINDOWS_NR] = {
-	DRM_PLANE_TYPE_PRIMARY,
-	DRM_PLANE_TYPE_OVERLAY,
-	DRM_PLANE_TYPE_CURSOR,
+	[PRIMARY_WIN] = DRM_PLANE_TYPE_PRIMARY,
+	[CURSON_WIN] = DRM_PLANE_TYPE_CURSOR,
 };
 
 static inline void decon_set_bits(struct decon_context *ctx, u32 reg, u32 mask,
@@ -552,12 +557,10 @@ static int decon_bind(struct device *dev, struct device *master, void *data)
 	drm_dev->max_vblank_count = 0xffffffff;
 
 	for (win = ctx->first_win; win < WINDOWS_NR; win++) {
-		int tmp = (win == ctx->first_win) ? 0 : win;
-
 		ctx->configs[win].pixel_formats = decon_formats;
 		ctx->configs[win].num_pixel_formats = ARRAY_SIZE(decon_formats);
-		ctx->configs[win].zpos = win;
-		ctx->configs[win].type = decon_win_types[tmp];
+		ctx->configs[win].zpos = win - ctx->first_win;
+		ctx->configs[win].type = decon_win_types[win];
 
 		ret = exynos_plane_init(drm_dev, &ctx->planes[win], win,
 					&ctx->configs[win]);
@@ -565,7 +568,7 @@ static int decon_bind(struct device *dev, struct device *master, void *data)
 			return ret;
 	}
 
-	exynos_plane = &ctx->planes[ctx->first_win];
+	exynos_plane = &ctx->planes[PRIMARY_WIN];
 	out_type = (ctx->out_type & IFTYPE_HDMI) ? EXYNOS_DISPLAY_TYPE_HDMI
 						  : EXYNOS_DISPLAY_TYPE_LCD;
 	ctx->crtc = exynos_drm_crtc_create(drm_dev, &exynos_plane->base,
-- 
cgit v1.1


From bf83060408fea52eccdcf695f3b4b16c71207691 Mon Sep 17 00:00:00 2001
From: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Date: Thu, 17 May 2018 11:18:34 -0400
Subject: Remove calls to suspend/resume atomic helpers from
 amdgpu_device_gpu_recover. (v2)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

First of all it's already being called from the display code from amd_ip_funcs.suspend/resume hooks.
Second of all, the place in amdgpu_device_gpu_recover it's being called is wrong for GPU stalls since
it is called BEFORE we cancel and force completion of all in flight jobs which were not yet processed.
So, as Bas pointed in the ticket we will try to wait for fence  in amdgpu_pm_compute_clocks but the pipe
is hanged so we end up in deadlock.

v2: remove unused variable

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=106500
Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 10 +---------
 1 file changed, 1 insertion(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 0e3f69d..adeb48e 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -3176,7 +3176,6 @@ error:
 int amdgpu_device_gpu_recover(struct amdgpu_device *adev,
 			      struct amdgpu_job *job, bool force)
 {
-	struct drm_atomic_state *state = NULL;
 	int i, r, resched;
 
 	if (!force && !amdgpu_device_ip_check_soft_reset(adev)) {
@@ -3199,10 +3198,6 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev,
 	/* block TTM */
 	resched = ttm_bo_lock_delayed_workqueue(&adev->mman.bdev);
 
-	/* store modesetting */
-	if (amdgpu_device_has_dc_support(adev))
-		state = drm_atomic_helper_suspend(adev->ddev);
-
 	/* block all schedulers and reset given job's ring */
 	for (i = 0; i < AMDGPU_MAX_RINGS; ++i) {
 		struct amdgpu_ring *ring = adev->rings[i];
@@ -3242,10 +3237,7 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev,
 		kthread_unpark(ring->sched.thread);
 	}
 
-	if (amdgpu_device_has_dc_support(adev)) {
-		if (drm_atomic_helper_resume(adev->ddev, state))
-			dev_info(adev->dev, "drm resume failed:%d\n", r);
-	} else {
+	if (!amdgpu_device_has_dc_support(adev)) {
 		drm_helper_resume_force_mode(adev->ddev);
 	}
 
-- 
cgit v1.1


From 99631045862e2994b47285a8cc96bc939ae5b42f Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Thu, 10 May 2018 14:45:12 -0500
Subject: drm/amdgpu: add new DF 1.7 register defs

Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_offset.h  | 4 ++++
 drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_sh_mask.h | 4 ++++
 2 files changed, 8 insertions(+)

diff --git a/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_offset.h b/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_offset.h
index 2b305dd..e6044e2 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_offset.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_offset.h
@@ -30,4 +30,8 @@
 #define mmDF_CS_AON0_DramBaseAddress0									0x0044
 #define mmDF_CS_AON0_DramBaseAddress0_BASE_IDX								0
 
+#define mmDF_CS_AON0_CoherentSlaveModeCtrlA0								0x0214
+#define mmDF_CS_AON0_CoherentSlaveModeCtrlA0_BASE_IDX							0
+
+
 #endif
diff --git a/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_sh_mask.h
index 2ba8497..a78c994 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_sh_mask.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_sh_mask.h
@@ -45,4 +45,8 @@
 #define DF_CS_AON0_DramBaseAddress0__IntLvAddrSel_MASK							0x00000700L
 #define DF_CS_AON0_DramBaseAddress0__DramBaseAddr_MASK							0xFFFFF000L
 
+//DF_CS_AON0_CoherentSlaveModeCtrlA0
+#define DF_CS_AON0_CoherentSlaveModeCtrlA0__ForceParWrRMW__SHIFT					0x3
+#define DF_CS_AON0_CoherentSlaveModeCtrlA0__ForceParWrRMW_MASK						0x00000008L
+
 #endif
-- 
cgit v1.1


From 8f9b2e506129e6eb0d21d163f361dd68a050b974 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Thu, 10 May 2018 14:59:31 -0500
Subject: drm/amdgpu: add new DF callback for ECC setup

The ForceParWrRMW setting needs to be enabled for ECC, but disabled
when ECC is not enabled.

Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu.h | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index 03a2c0be..a59c075 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -1401,6 +1401,8 @@ struct amdgpu_df_funcs {
 						 bool enable);
 	void (*get_clockgating_state)(struct amdgpu_device *adev,
 				      u32 *flags);
+	void (*enable_ecc_force_par_wr_rmw)(struct amdgpu_device *adev,
+					    bool enable);
 };
 /* Define the HW IP blocks will be used in driver , add more if necessary */
 enum amd_hw_ip_block_type {
-- 
cgit v1.1


From 1ca2393b7373d5b0e5a356124fb10fc97e143e88 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Thu, 10 May 2018 15:06:55 -0500
Subject: drm/amdgpu: add a df 1.7 implementation of
 enable_ecc_force_par_wr_rmw

Needed for proper memory setup depending on whether ECC is
enabled on a particular board.

Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/df_v1_7.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/df_v1_7.c b/drivers/gpu/drm/amd/amdgpu/df_v1_7.c
index 4ffda99..9935371 100644
--- a/drivers/gpu/drm/amd/amdgpu/df_v1_7.c
+++ b/drivers/gpu/drm/amd/amdgpu/df_v1_7.c
@@ -102,6 +102,13 @@ static void df_v1_7_get_clockgating_state(struct amdgpu_device *adev,
 		*flags |= AMD_CG_SUPPORT_DF_MGCG;
 }
 
+static void df_v1_7_enable_ecc_force_par_wr_rmw(struct amdgpu_device *adev,
+						bool enable)
+{
+	WREG32_FIELD15(DF, 0, DF_CS_AON0_CoherentSlaveModeCtrlA0,
+		       ForceParWrRMW, enable);
+}
+
 const struct amdgpu_df_funcs df_v1_7_funcs = {
 	.init = df_v1_7_init,
 	.enable_broadcast_mode = df_v1_7_enable_broadcast_mode,
@@ -109,4 +116,5 @@ const struct amdgpu_df_funcs df_v1_7_funcs = {
 	.get_hbm_channel_number = df_v1_7_get_hbm_channel_number,
 	.update_medium_grain_clock_gating = df_v1_7_update_medium_grain_clock_gating,
 	.get_clockgating_state = df_v1_7_get_clockgating_state,
+	.enable_ecc_force_par_wr_rmw = df_v1_7_enable_ecc_force_par_wr_rmw,
 };
-- 
cgit v1.1


From e1d1a7729a62d7b79fb2ab4ac3bc6fc0ebfb6db9 Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Thu, 10 May 2018 15:15:12 -0500
Subject: drm/amdgpu/gmc9: disable partial wr rmw if ECC is not enabled

The vbios mistakenly sets this bit on some boards without ECC.
This can lead to reduced performance in some workloads.  Disable
the bit if the board does not have ECC.

Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
index b60ed28..3c0a85d 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
@@ -675,6 +675,7 @@ static int gmc_v9_0_late_init(void *handle)
 			DRM_INFO("ECC is active.\n");
 		} else if (r == 0) {
 			DRM_INFO("ECC is not present.\n");
+			adev->df_funcs->enable_ecc_force_par_wr_rmw(adev, false);
 		} else {
 			DRM_ERROR("gmc_v9_0_ecc_available() failed. r: %d\n", r);
 			return r;
-- 
cgit v1.1


From 63e138abf0761c7ea3dcb29060bfd48a34e58ecf Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Fri, 18 May 2018 14:24:44 +0800
Subject: drm/amd/pp: Fix static checker warning

error: uninitialized symbol 'xxxx'

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c   | 24 +++++++++-------------
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c   |  3 ++-
 drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c |  6 ++----
 3 files changed, 14 insertions(+), 19 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c
index ec38c9f..7047e29 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c
@@ -1104,10 +1104,8 @@ int atomctrl_get_voltage_evv_on_sclk(
 			GetIndexIntoMasterTable(COMMAND, GetVoltageInfo),
 			(uint32_t *)&get_voltage_info_param_space);
 
-	if (0 != result)
-		return result;
-
-	*voltage = le16_to_cpu(((GET_EVV_VOLTAGE_INFO_OUTPUT_PARAMETER_V1_2 *)
+	*voltage = result ? 0 :
+			le16_to_cpu(((GET_EVV_VOLTAGE_INFO_OUTPUT_PARAMETER_V1_2 *)
 				(&get_voltage_info_param_space))->usVoltageLevel);
 
 	return result;
@@ -1312,8 +1310,7 @@ int atomctrl_read_efuse(struct pp_hwmgr *hwmgr, uint16_t start_index,
 	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, ReadEfuseValue),
 			(uint32_t *)&efuse_param);
-	if (!result)
-		*efuse = le32_to_cpu(efuse_param.ulEfuseValue) & mask;
+	*efuse = result ? 0 : le32_to_cpu(efuse_param.ulEfuseValue) & mask;
 
 	return result;
 }
@@ -1354,11 +1351,8 @@ int atomctrl_get_voltage_evv_on_sclk_ai(struct pp_hwmgr *hwmgr, uint8_t voltage_
 			GetIndexIntoMasterTable(COMMAND, GetVoltageInfo),
 			(uint32_t *)&get_voltage_info_param_space);
 
-	if (0 != result)
-		return result;
-
-	*voltage = le32_to_cpu(((GET_EVV_VOLTAGE_INFO_OUTPUT_PARAMETER_V1_3 *)
-				(&get_voltage_info_param_space))->ulVoltageLevel);
+	*voltage = result ? 0 :
+		le32_to_cpu(((GET_EVV_VOLTAGE_INFO_OUTPUT_PARAMETER_V1_3 *)(&get_voltage_info_param_space))->ulVoltageLevel);
 
 	return result;
 }
@@ -1552,15 +1546,17 @@ void atomctrl_get_voltage_range(struct pp_hwmgr *hwmgr, uint32_t *max_vddc,
 		case CHIP_FIJI:
 			*max_vddc = le32_to_cpu(((ATOM_ASIC_PROFILING_INFO_V3_3 *)profile)->ulMaxVddc/4);
 			*min_vddc = le32_to_cpu(((ATOM_ASIC_PROFILING_INFO_V3_3 *)profile)->ulMinVddc/4);
-			break;
+			return;
 		case CHIP_POLARIS11:
 		case CHIP_POLARIS10:
 		case CHIP_POLARIS12:
 			*max_vddc = le32_to_cpu(((ATOM_ASIC_PROFILING_INFO_V3_6 *)profile)->ulMaxVddc/100);
 			*min_vddc = le32_to_cpu(((ATOM_ASIC_PROFILING_INFO_V3_6 *)profile)->ulMinVddc/100);
-			break;
-		default:
 			return;
+		default:
+			break;
 		}
 	}
+	*max_vddc = 0;
+	*min_vddc = 0;
 }
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index 646c9e9..45e9b8c 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -860,7 +860,8 @@ static void smu7_setup_voltage_range_from_vbios(struct pp_hwmgr *hwmgr)
 	struct phm_ppt_v1_clock_voltage_dependency_table *dep_sclk_table;
 	struct phm_ppt_v1_information *table_info =
 			(struct phm_ppt_v1_information *)(hwmgr->pptable);
-	uint32_t min_vddc, max_vddc;
+	uint32_t min_vddc = 0;
+	uint32_t max_vddc = 0;
 
 	if (!table_info)
 		return;
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
index 64d33b7..d644a9b 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
@@ -283,11 +283,9 @@ int smu7_read_smc_sram_dword(struct pp_hwmgr *hwmgr, uint32_t smc_addr, uint32_t
 
 	result = smu7_set_smc_sram_address(hwmgr, smc_addr, limit);
 
-	if (result)
-		return result;
+	*value = result ? 0 : cgs_read_register(hwmgr->device, mmSMC_IND_DATA_11);
 
-	*value = cgs_read_register(hwmgr->device, mmSMC_IND_DATA_11);
-	return 0;
+	return result;
 }
 
 int smu7_write_smc_sram_dword(struct pp_hwmgr *hwmgr, uint32_t smc_addr, uint32_t value, uint32_t limit)
-- 
cgit v1.1


From e6ee925b795311679dd6e0ebeae6f1dbe983c059 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Fri, 18 May 2018 14:59:46 +0800
Subject: drm/amd/pp: fix a couple locking issues

We should return unlock on the error path

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c   | 31 +++++++++++++---------
 1 file changed, 19 insertions(+), 12 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c
index 99b29ff..c952845 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c
@@ -936,45 +936,49 @@ int smu7_enable_didt_config(struct pp_hwmgr *hwmgr)
 
 			if (hwmgr->chip_id == CHIP_POLARIS10) {
 				result = smu7_program_pt_config_registers(hwmgr, GCCACConfig_Polaris10);
-				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", return result);
+				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", goto error);
 				result = smu7_program_pt_config_registers(hwmgr, DIDTConfig_Polaris10);
-				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", return result);
+				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", goto error);
 			} else if (hwmgr->chip_id == CHIP_POLARIS11) {
 				result = smu7_program_pt_config_registers(hwmgr, GCCACConfig_Polaris11);
-				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", return result);
+				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", goto error);
 				if (hwmgr->is_kicker)
 					result = smu7_program_pt_config_registers(hwmgr, DIDTConfig_Polaris11_Kicker);
 				else
 					result = smu7_program_pt_config_registers(hwmgr, DIDTConfig_Polaris11);
-				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", return result);
+				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", goto error);
 			} else if (hwmgr->chip_id == CHIP_POLARIS12) {
 				result = smu7_program_pt_config_registers(hwmgr, GCCACConfig_Polaris11);
-				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", return result);
+				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", goto error);
 				result = smu7_program_pt_config_registers(hwmgr, DIDTConfig_Polaris12);
-				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", return result);
+				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", goto error);
 			} else if (hwmgr->chip_id == CHIP_VEGAM) {
 				result = smu7_program_pt_config_registers(hwmgr, GCCACConfig_VegaM);
-				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", return result);
+				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", goto error);
 				result = smu7_program_pt_config_registers(hwmgr, DIDTConfig_VegaM);
-				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", return result);
+				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", goto error);
 			}
 		}
 		cgs_write_register(hwmgr->device, mmGRBM_GFX_INDEX, value2);
 
 		result = smu7_enable_didt(hwmgr, true);
-		PP_ASSERT_WITH_CODE((result == 0), "EnableDiDt failed.", return result);
+		PP_ASSERT_WITH_CODE((result == 0), "EnableDiDt failed.", goto error);
 
 		if (hwmgr->chip_id == CHIP_POLARIS11) {
 			result = smum_send_msg_to_smc(hwmgr,
 						(uint16_t)(PPSMC_MSG_EnableDpmDidt));
 			PP_ASSERT_WITH_CODE((0 == result),
-					"Failed to enable DPM DIDT.", return result);
+					"Failed to enable DPM DIDT.", goto error);
 		}
 		mutex_unlock(&adev->grbm_idx_mutex);
 		adev->gfx.rlc.funcs->exit_safe_mode(adev);
 	}
 
 	return 0;
+error:
+	mutex_unlock(&adev->grbm_idx_mutex);
+	adev->gfx.rlc.funcs->exit_safe_mode(adev);
+	return result;
 }
 
 int smu7_disable_didt_config(struct pp_hwmgr *hwmgr)
@@ -992,17 +996,20 @@ int smu7_disable_didt_config(struct pp_hwmgr *hwmgr)
 		result = smu7_enable_didt(hwmgr, false);
 		PP_ASSERT_WITH_CODE((result == 0),
 				"Post DIDT enable clock gating failed.",
-				return result);
+				goto error);
 		if (hwmgr->chip_id == CHIP_POLARIS11) {
 			result = smum_send_msg_to_smc(hwmgr,
 						(uint16_t)(PPSMC_MSG_DisableDpmDidt));
 			PP_ASSERT_WITH_CODE((0 == result),
-					"Failed to disable DPM DIDT.", return result);
+					"Failed to disable DPM DIDT.", goto error);
 		}
 		adev->gfx.rlc.funcs->exit_safe_mode(adev);
 	}
 
 	return 0;
+error:
+	adev->gfx.rlc.funcs->exit_safe_mode(adev);
+	return result;
 }
 
 int smu7_enable_smc_cac(struct pp_hwmgr *hwmgr)
-- 
cgit v1.1


From 34319b329f73eabd7e3baefecf9f71eb8b86db6f Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Wed, 16 May 2018 20:06:53 +0800
Subject: drm/amdgpu: skip CG for VCN when late_init/fini

VCN clockgating is handled manually like VCE and UVD.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index adeb48e..290e279 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -1718,6 +1718,7 @@ static int amdgpu_device_ip_late_set_cg_state(struct amdgpu_device *adev)
 		/* skip CG for VCE/UVD, it's handled specially */
 		if (adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_UVD &&
 		    adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_VCE &&
+		    adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_VCN &&
 		    adev->ip_blocks[i].version->funcs->set_clockgating_state) {
 			/* enable clockgating to save power */
 			r = adev->ip_blocks[i].version->funcs->set_clockgating_state((void *)adev,
@@ -1817,6 +1818,7 @@ static int amdgpu_device_ip_fini(struct amdgpu_device *adev)
 
 		if (adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_UVD &&
 			adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_VCE &&
+			adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_VCN &&
 			adev->ip_blocks[i].version->funcs->set_clockgating_state) {
 			/* ungate blocks before hw fini so that we can shutdown the blocks safely */
 			r = adev->ip_blocks[i].version->funcs->set_clockgating_state((void *)adev,
-- 
cgit v1.1


From ca0b9494633f65ee6779d4c7ca19e799b8308e69 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Wed, 16 May 2018 20:09:09 +0800
Subject: drm/amd/pp: Add smu support for VCN powergating on RV

Add the powerplay callback for powergating VCN (same
as UVD and VCE).

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c | 19 ++++++++++++++++++-
 1 file changed, 18 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
index 017ef2d..85f84f4 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
@@ -1128,6 +1128,23 @@ static int smu10_set_mmhub_powergating_by_smu(struct pp_hwmgr *hwmgr)
 	return smum_send_msg_to_smc(hwmgr, PPSMC_MSG_PowerGateMmHub);
 }
 
+static void smu10_powergate_vcn(struct pp_hwmgr *hwmgr, bool bgate)
+{
+	if (bgate) {
+		amdgpu_device_ip_set_powergating_state(hwmgr->adev,
+						AMD_IP_BLOCK_TYPE_VCN,
+						AMD_PG_STATE_GATE);
+		smum_send_msg_to_smc_with_parameter(hwmgr,
+					PPSMC_MSG_PowerDownVcn, 0);
+	} else {
+		smum_send_msg_to_smc_with_parameter(hwmgr,
+						PPSMC_MSG_PowerUpVcn, 0);
+		amdgpu_device_ip_set_powergating_state(hwmgr->adev,
+						AMD_IP_BLOCK_TYPE_VCN,
+						AMD_PG_STATE_UNGATE);
+	}
+}
+
 static const struct pp_hwmgr_func smu10_hwmgr_funcs = {
 	.backend_init = smu10_hwmgr_backend_init,
 	.backend_fini = smu10_hwmgr_backend_fini,
@@ -1136,7 +1153,7 @@ static const struct pp_hwmgr_func smu10_hwmgr_funcs = {
 	.force_dpm_level = smu10_dpm_force_dpm_level,
 	.get_power_state_size = smu10_get_power_state_size,
 	.powerdown_uvd = NULL,
-	.powergate_uvd = NULL,
+	.powergate_uvd = smu10_powergate_vcn,
 	.powergate_vce = NULL,
 	.get_mclk = smu10_dpm_get_mclk,
 	.get_sclk = smu10_dpm_get_sclk,
-- 
cgit v1.1


From 8dbb8cdf52dbc264d531d1e51f5e311bd1558c21 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Wed, 16 May 2018 20:10:25 +0800
Subject: drm/amdgpu: Add CG/PG flags for VCN

Define new clock and powergating flags for VCN block.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/include/amd_shared.h | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/include/amd_shared.h b/drivers/gpu/drm/amd/include/amd_shared.h
index 33de330..b178176 100644
--- a/drivers/gpu/drm/amd/include/amd_shared.h
+++ b/drivers/gpu/drm/amd/include/amd_shared.h
@@ -92,7 +92,7 @@ enum amd_powergating_state {
 #define AMD_CG_SUPPORT_GFX_3D_CGLS		(1 << 21)
 #define AMD_CG_SUPPORT_DRM_MGCG			(1 << 22)
 #define AMD_CG_SUPPORT_DF_MGCG			(1 << 23)
-
+#define AMD_CG_SUPPORT_VCN_MGCG			(1 << 24)
 /* PG flags */
 #define AMD_PG_SUPPORT_GFX_PG			(1 << 0)
 #define AMD_PG_SUPPORT_GFX_SMG			(1 << 1)
@@ -108,6 +108,7 @@ enum amd_powergating_state {
 #define AMD_PG_SUPPORT_GFX_QUICK_MG		(1 << 11)
 #define AMD_PG_SUPPORT_GFX_PIPELINE		(1 << 12)
 #define AMD_PG_SUPPORT_MMHUB			(1 << 13)
+#define AMD_PG_SUPPORT_VCN			(1 << 14)
 
 enum PP_FEATURE_MASK {
 	PP_SCLK_DPM_MASK = 0x1,
-- 
cgit v1.1


From ac06b4cfd78b79ec6c8306062801a4276a3e0c79 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Thu, 17 May 2018 15:58:53 +0800
Subject: drm/amdgpu: Add SOC15_WAIT_ON_RREG macro define

Add new macro to wait on a register field to be a specific
value.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/soc15_common.h | 15 +++++++++++++++
 1 file changed, 15 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/soc15_common.h b/drivers/gpu/drm/amd/amdgpu/soc15_common.h
index def8650..0942f49 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15_common.h
+++ b/drivers/gpu/drm/amd/amdgpu/soc15_common.h
@@ -47,6 +47,21 @@
 #define WREG32_SOC15_OFFSET(ip, inst, reg, offset, value) \
 	WREG32((adev->reg_offset[ip##_HWIP][inst][reg##_BASE_IDX] + reg) + offset, value)
 
+#define SOC15_WAIT_ON_RREG(ip, inst, reg, expected_value, mask, ret) \
+	do {							\
+		uint32_t tmp_ = RREG32(adev->reg_offset[ip##_HWIP][inst][reg##_BASE_IDX] + reg); \
+		uint32_t loop = adev->usec_timeout;		\
+		while ((tmp_ & (mask)) != (expected_value)) {	\
+			udelay(2);				\
+			tmp_ = RREG32(adev->reg_offset[ip##_HWIP][inst][reg##_BASE_IDX] + reg); \
+			loop--;					\
+			if (!loop) {				\
+				ret = -ETIMEDOUT;		\
+				break;				\
+			}					\
+		}						\
+	} while (0)
+
 #endif
 
 
-- 
cgit v1.1


From ecf81ed98c8df8c6d397f4e044af175481b5b831 Mon Sep 17 00:00:00 2001
From: Andrzej Hajda <a.hajda@samsung.com>
Date: Tue, 8 May 2018 11:36:58 +0200
Subject: drm/exynos/dsi: mask frame-done interrupt

DSI driver is not really interested in this interrupt. It causes only
unnecessary code execution of interrupt handler and could possibly
cause FIFO overflow - as it triggers DSI interrupt handler to process
next DSI transfer. With this patch we will get rid of about 30 IRQ
handler calls per second.

Fixes: e6f988a45857 ("drm/exynos: dsi: add support for Exynos5433")
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
---
 drivers/gpu/drm/exynos/exynos_drm_dsi.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/exynos/exynos_drm_dsi.c b/drivers/gpu/drm/exynos/exynos_drm_dsi.c
index eae44fd..7c3030b 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_dsi.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_dsi.c
@@ -1264,15 +1264,15 @@ static irqreturn_t exynos_dsi_irq(int irq, void *dev_id)
 
 	if (status & DSIM_INT_SW_RST_RELEASE) {
 		u32 mask = ~(DSIM_INT_RX_DONE | DSIM_INT_SFR_FIFO_EMPTY |
-			DSIM_INT_SFR_HDR_FIFO_EMPTY | DSIM_INT_FRAME_DONE |
-			DSIM_INT_RX_ECC_ERR | DSIM_INT_SW_RST_RELEASE);
+			DSIM_INT_SFR_HDR_FIFO_EMPTY | DSIM_INT_RX_ECC_ERR |
+			DSIM_INT_SW_RST_RELEASE);
 		exynos_dsi_write(dsi, DSIM_INTMSK_REG, mask);
 		complete(&dsi->completed);
 		return IRQ_HANDLED;
 	}
 
 	if (!(status & (DSIM_INT_RX_DONE | DSIM_INT_SFR_FIFO_EMPTY |
-			DSIM_INT_FRAME_DONE | DSIM_INT_PLL_STABLE)))
+			DSIM_INT_PLL_STABLE)))
 		return IRQ_HANDLED;
 
 	if (exynos_dsi_transfer_finish(dsi))
-- 
cgit v1.1


From e9497dc2f3e0ead4004231b8d282cb4ecdd36463 Mon Sep 17 00:00:00 2001
From: Marek Szyprowski <m.szyprowski@samsung.com>
Date: Tue, 12 Dec 2017 13:01:15 +0100
Subject: drm/exynos: Fix error value in exynos_drm_crtc_get_by_type()

EPERM is not the correct error value when the driver is not able to get
its resources. Change it to ENODEV.

Reported-by: Russell King - ARM Linux <linux@armlinux.org.uk>
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
---
 drivers/gpu/drm/exynos/exynos_drm_crtc.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/exynos/exynos_drm_crtc.c b/drivers/gpu/drm/exynos/exynos_drm_crtc.c
index dc01342..eea9025 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_crtc.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_crtc.c
@@ -228,7 +228,7 @@ struct exynos_drm_crtc *exynos_drm_crtc_get_by_type(struct drm_device *drm_dev,
 		if (to_exynos_crtc(crtc)->type == out_type)
 			return to_exynos_crtc(crtc);
 
-	return ERR_PTR(-EPERM);
+	return ERR_PTR(-ENODEV);
 }
 
 int exynos_drm_set_possible_crtcs(struct drm_encoder *encoder,
-- 
cgit v1.1


From e9dfe83d8985fcd2e0fac570215a039c24fb6b12 Mon Sep 17 00:00:00 2001
From: Marek Szyprowski <m.szyprowski@samsung.com>
Date: Wed, 23 May 2018 12:15:50 +0200
Subject: drm/exynos: Fix default value for zpos plane property

The default zpos property for all planes in Exynos DRM was fixed as zero.
Fix this by providing proper value provided by hardware drivers, which
typically matches hardware window number.

Reported-by: Seung-Woo Kim <sw0312.kim@samsung.com>
Fixes: e47726a11e11 ("drm/exynos: use generic code for managing zpos plane property")
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
---
 drivers/gpu/drm/exynos/exynos_drm_plane.c | 9 ++++-----
 1 file changed, 4 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/exynos/exynos_drm_plane.c b/drivers/gpu/drm/exynos/exynos_drm_plane.c
index d2a90da..38a2a7f 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_plane.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_plane.c
@@ -289,13 +289,12 @@ static const struct drm_plane_helper_funcs plane_helper_funcs = {
 };
 
 static void exynos_plane_attach_zpos_property(struct drm_plane *plane,
-					      bool immutable)
+					      int zpos, bool immutable)
 {
-	/* FIXME */
 	if (immutable)
-		drm_plane_create_zpos_immutable_property(plane, 0);
+		drm_plane_create_zpos_immutable_property(plane, zpos);
 	else
-		drm_plane_create_zpos_property(plane, 0, 0, MAX_PLANE - 1);
+		drm_plane_create_zpos_property(plane, zpos, 0, MAX_PLANE - 1);
 }
 
 int exynos_plane_init(struct drm_device *dev,
@@ -320,7 +319,7 @@ int exynos_plane_init(struct drm_device *dev,
 	exynos_plane->index = index;
 	exynos_plane->config = config;
 
-	exynos_plane_attach_zpos_property(&exynos_plane->base,
+	exynos_plane_attach_zpos_property(&exynos_plane->base, config->zpos,
 			   !(config->capabilities & EXYNOS_DRM_PLANE_CAP_ZPOS));
 
 	return 0;
-- 
cgit v1.1


From c9dc5abb661b02239eef6cd991700707dc381110 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Thu, 17 May 2018 11:11:22 +0800
Subject: drm/amdgpu: Add static CG control for VCN on RV

Implement proper static clockgating support for VCN.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c | 50 +++++++++++++++++++++++++++--------
 1 file changed, 39 insertions(+), 11 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
index 7fbbdb1..7a36641 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
@@ -288,14 +288,14 @@ static void vcn_v1_0_mc_resume(struct amdgpu_device *adev)
  *
  * Disable clock gating for VCN block
  */
-static void vcn_v1_0_disable_clock_gating(struct amdgpu_device *adev, bool sw)
+static void vcn_v1_0_disable_clock_gating(struct amdgpu_device *adev)
 {
 	uint32_t data;
 
 	/* JPEG disable CGC */
 	data = RREG32_SOC15(VCN, 0, mmJPEG_CGC_CTRL);
 
-	if (sw)
+	if (adev->cg_flags & AMD_CG_SUPPORT_VCN_MGCG)
 		data |= 1 << JPEG_CGC_CTRL__DYN_CLOCK_MODE__SHIFT;
 	else
 		data &= ~JPEG_CGC_CTRL__DYN_CLOCK_MODE_MASK;
@@ -310,7 +310,7 @@ static void vcn_v1_0_disable_clock_gating(struct amdgpu_device *adev, bool sw)
 
 	/* UVD disable CGC */
 	data = RREG32_SOC15(VCN, 0, mmUVD_CGC_CTRL);
-	if (sw)
+	if (adev->cg_flags & AMD_CG_SUPPORT_VCN_MGCG)
 		data |= 1 << UVD_CGC_CTRL__DYN_CLOCK_MODE__SHIFT;
 	else
 		data &= ~ UVD_CGC_CTRL__DYN_CLOCK_MODE_MASK;
@@ -415,13 +415,13 @@ static void vcn_v1_0_disable_clock_gating(struct amdgpu_device *adev, bool sw)
  *
  * Enable clock gating for VCN block
  */
-static void vcn_v1_0_enable_clock_gating(struct amdgpu_device *adev, bool sw)
+static void vcn_v1_0_enable_clock_gating(struct amdgpu_device *adev)
 {
 	uint32_t data = 0;
 
 	/* enable JPEG CGC */
 	data = RREG32_SOC15(VCN, 0, mmJPEG_CGC_CTRL);
-	if (sw)
+	if (adev->cg_flags & AMD_CG_SUPPORT_VCN_MGCG)
 		data |= 1 << JPEG_CGC_CTRL__DYN_CLOCK_MODE__SHIFT;
 	else
 		data |= 0 << JPEG_CGC_CTRL__DYN_CLOCK_MODE__SHIFT;
@@ -435,7 +435,7 @@ static void vcn_v1_0_enable_clock_gating(struct amdgpu_device *adev, bool sw)
 
 	/* enable UVD CGC */
 	data = RREG32_SOC15(VCN, 0, mmUVD_CGC_CTRL);
-	if (sw)
+	if (adev->cg_flags & AMD_CG_SUPPORT_VCN_MGCG)
 		data |= 1 << UVD_CGC_CTRL__DYN_CLOCK_MODE__SHIFT;
 	else
 		data |= 0 << UVD_CGC_CTRL__DYN_CLOCK_MODE__SHIFT;
@@ -500,7 +500,7 @@ static int vcn_v1_0_start(struct amdgpu_device *adev)
 	vcn_v1_0_mc_resume(adev);
 
 	/* disable clock gating */
-	vcn_v1_0_disable_clock_gating(adev, true);
+	vcn_v1_0_disable_clock_gating(adev);
 
 	/* disable interupt */
 	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_MASTINT_EN), 0,
@@ -681,15 +681,43 @@ static int vcn_v1_0_stop(struct amdgpu_device *adev)
 			~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
 
 	/* enable clock gating */
-	vcn_v1_0_enable_clock_gating(adev, true);
+	vcn_v1_0_enable_clock_gating(adev);
 
 	return 0;
 }
 
+bool vcn_v1_0_is_idle(void *handle)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+	return (RREG32_SOC15(VCN, 0, mmUVD_STATUS) == 0x2);
+}
+
+int vcn_v1_0_wait_for_idle(void *handle)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+	int ret = 0;
+
+	SOC15_WAIT_ON_RREG(VCN, 0, mmUVD_STATUS, 0x2, 0x2, ret);
+
+	return ret;
+}
+
 static int vcn_v1_0_set_clockgating_state(void *handle,
 					  enum amd_clockgating_state state)
 {
-	/* needed for driver unload*/
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+	bool enable = (state == AMD_CG_STATE_GATE) ? true : false;
+
+	if (enable) {
+		/* wait for STATUS to clear */
+		if (vcn_v1_0_is_idle(handle))
+			return -EBUSY;
+		vcn_v1_0_enable_clock_gating(adev);
+	} else {
+		/* disable HW gating and enable Sw gating */
+		vcn_v1_0_disable_clock_gating(adev);
+	}
 	return 0;
 }
 
@@ -1072,8 +1100,8 @@ static const struct amd_ip_funcs vcn_v1_0_ip_funcs = {
 	.hw_fini = vcn_v1_0_hw_fini,
 	.suspend = vcn_v1_0_suspend,
 	.resume = vcn_v1_0_resume,
-	.is_idle = NULL /* vcn_v1_0_is_idle */,
-	.wait_for_idle = NULL /* vcn_v1_0_wait_for_idle */,
+	.is_idle = vcn_v1_0_is_idle,
+	.wait_for_idle = vcn_v1_0_wait_for_idle,
 	.check_soft_reset = NULL /* vcn_v1_0_check_soft_reset */,
 	.pre_soft_reset = NULL /* vcn_v1_0_pre_soft_reset */,
 	.soft_reset = NULL /* vcn_v1_0_soft_reset */,
-- 
cgit v1.1


From 79953a60e4476be90fa1767fbf49a76b6a8b01ef Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Thu, 17 May 2018 11:13:51 +0800
Subject: drm/amdgpu: Enable VCN CG by default on RV

Enable VCN clockgating by default on Raven.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/soc15.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index 0e4f67e..2cf9a18 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -710,7 +710,8 @@ static int soc15_common_early_init(void *handle)
 			AMD_CG_SUPPORT_MC_MGCG |
 			AMD_CG_SUPPORT_MC_LS |
 			AMD_CG_SUPPORT_SDMA_MGCG |
-			AMD_CG_SUPPORT_SDMA_LS;
+			AMD_CG_SUPPORT_SDMA_LS |
+			AMD_CG_SUPPORT_VCN_MGCG;
 		adev->pg_flags = AMD_PG_SUPPORT_SDMA;
 
 		if (adev->powerplay.pp_feature & PP_GFXOFF_MASK)
-- 
cgit v1.1


From d58c5d9a42050c93f17ba82aaff0f34a30761ac7 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Thu, 17 May 2018 16:07:02 +0800
Subject: drm/amdgpu: Add VCN static PG support on RV

Implement static powergating suport on VCN.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h | 11 ++++
 drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c   | 92 ++++++++++++++++++++++++++++++++-
 2 files changed, 102 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h
index 2fd7db8..181e6af 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h
@@ -45,6 +45,17 @@
 #define VCN_ENC_CMD_REG_WRITE		0x0000000b
 #define VCN_ENC_CMD_REG_WAIT		0x0000000c
 
+enum engine_status_constants {
+	UVD_PGFSM_STATUS__UVDM_UVDU_PWR_ON = 0x2AAAA0,
+	UVD_PGFSM_CONFIG__UVDM_UVDU_PWR_ON = 0x00000002,
+	UVD_STATUS__UVD_BUSY = 0x00000004,
+	GB_ADDR_CONFIG_DEFAULT = 0x26010011,
+	UVD_STATUS__IDLE = 0x2,
+	UVD_STATUS__BUSY = 0x5,
+	UVD_POWER_STATUS__UVD_POWER_STATUS_TILES_OFF = 0x1,
+	UVD_STATUS__RBC_BUSY = 0x1,
+};
+
 struct amdgpu_vcn {
 	struct amdgpu_bo	*vcpu_bo;
 	void			*cpu_addr;
diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
index 7a36641..dcb60ee 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
@@ -480,6 +480,94 @@ static void vcn_v1_0_enable_clock_gating(struct amdgpu_device *adev)
 	WREG32_SOC15(VCN, 0, mmUVD_SUVD_CGC_CTRL, data);
 }
 
+static void vcn_1_0_disable_static_power_gating(struct amdgpu_device *adev)
+{
+	uint32_t data = 0;
+	int ret;
+
+	if (adev->pg_flags & AMD_PG_SUPPORT_VCN) {
+		data = (1 << UVD_PGFSM_CONFIG__UVDM_PWR_CONFIG__SHIFT
+			| 1 << UVD_PGFSM_CONFIG__UVDU_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDF_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDC_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDB_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDIL_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDIR_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDTD_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDTE_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDE_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDW_PWR_CONFIG__SHIFT);
+
+		WREG32_SOC15(VCN, 0, mmUVD_PGFSM_CONFIG, data);
+		SOC15_WAIT_ON_RREG(VCN, 0, mmUVD_PGFSM_STATUS, UVD_PGFSM_STATUS__UVDM_UVDU_PWR_ON, 0xFFFFFF, ret);
+	} else {
+		data = (1 << UVD_PGFSM_CONFIG__UVDM_PWR_CONFIG__SHIFT
+			| 1 << UVD_PGFSM_CONFIG__UVDU_PWR_CONFIG__SHIFT
+			| 1 << UVD_PGFSM_CONFIG__UVDF_PWR_CONFIG__SHIFT
+			| 1 << UVD_PGFSM_CONFIG__UVDC_PWR_CONFIG__SHIFT
+			| 1 << UVD_PGFSM_CONFIG__UVDB_PWR_CONFIG__SHIFT
+			| 1 << UVD_PGFSM_CONFIG__UVDIL_PWR_CONFIG__SHIFT
+			| 1 << UVD_PGFSM_CONFIG__UVDIR_PWR_CONFIG__SHIFT
+			| 1 << UVD_PGFSM_CONFIG__UVDTD_PWR_CONFIG__SHIFT
+			| 1 << UVD_PGFSM_CONFIG__UVDTE_PWR_CONFIG__SHIFT
+			| 1 << UVD_PGFSM_CONFIG__UVDE_PWR_CONFIG__SHIFT
+			| 1 << UVD_PGFSM_CONFIG__UVDW_PWR_CONFIG__SHIFT);
+		WREG32_SOC15(VCN, 0, mmUVD_PGFSM_CONFIG, data);
+		SOC15_WAIT_ON_RREG(VCN, 0, mmUVD_PGFSM_STATUS, 0,  0xFFFFFFFF, ret);
+	}
+
+	/* polling UVD_PGFSM_STATUS to confirm UVDM_PWR_STATUS , UVDU_PWR_STATUS are 0 (power on) */
+
+	data = RREG32_SOC15(VCN, 0, mmUVD_POWER_STATUS);
+	data &= ~0x103;
+	if (adev->pg_flags & AMD_PG_SUPPORT_VCN)
+		data |= UVD_PGFSM_CONFIG__UVDM_UVDU_PWR_ON | UVD_POWER_STATUS__UVD_PG_EN_MASK;
+
+	WREG32_SOC15(VCN, 0, mmUVD_POWER_STATUS, data);
+}
+
+static void vcn_1_0_enable_static_power_gating(struct amdgpu_device *adev)
+{
+	uint32_t data = 0;
+	int ret;
+
+	if (adev->pg_flags & AMD_PG_SUPPORT_VCN) {
+		/* Before power off, this indicator has to be turned on */
+		data = RREG32_SOC15(VCN, 0, mmUVD_POWER_STATUS);
+		data &= ~UVD_POWER_STATUS__UVD_POWER_STATUS_MASK;
+		data |= UVD_POWER_STATUS__UVD_POWER_STATUS_TILES_OFF;
+		WREG32_SOC15(VCN, 0, mmUVD_POWER_STATUS, data);
+
+
+		data = (2 << UVD_PGFSM_CONFIG__UVDM_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDU_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDF_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDC_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDB_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDIL_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDIR_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDTD_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDTE_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDE_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDW_PWR_CONFIG__SHIFT);
+
+		WREG32_SOC15(VCN, 0, mmUVD_PGFSM_CONFIG, data);
+
+		data = (2 << UVD_PGFSM_STATUS__UVDM_PWR_STATUS__SHIFT
+			| 2 << UVD_PGFSM_STATUS__UVDU_PWR_STATUS__SHIFT
+			| 2 << UVD_PGFSM_STATUS__UVDF_PWR_STATUS__SHIFT
+			| 2 << UVD_PGFSM_STATUS__UVDC_PWR_STATUS__SHIFT
+			| 2 << UVD_PGFSM_STATUS__UVDB_PWR_STATUS__SHIFT
+			| 2 << UVD_PGFSM_STATUS__UVDIL_PWR_STATUS__SHIFT
+			| 2 << UVD_PGFSM_STATUS__UVDIR_PWR_STATUS__SHIFT
+			| 2 << UVD_PGFSM_STATUS__UVDTD_PWR_STATUS__SHIFT
+			| 2 << UVD_PGFSM_STATUS__UVDTE_PWR_STATUS__SHIFT
+			| 2 << UVD_PGFSM_STATUS__UVDE_PWR_STATUS__SHIFT
+			| 2 << UVD_PGFSM_STATUS__UVDW_PWR_STATUS__SHIFT);
+		SOC15_WAIT_ON_RREG(VCN, 0, mmUVD_PGFSM_STATUS, data, 0xFFFFFFFF, ret);
+	}
+}
+
 /**
  * vcn_v1_0_start - start VCN block
  *
@@ -499,6 +587,7 @@ static int vcn_v1_0_start(struct amdgpu_device *adev)
 
 	vcn_v1_0_mc_resume(adev);
 
+	vcn_1_0_disable_static_power_gating(adev);
 	/* disable clock gating */
 	vcn_v1_0_disable_clock_gating(adev);
 
@@ -681,8 +770,9 @@ static int vcn_v1_0_stop(struct amdgpu_device *adev)
 			~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
 
 	/* enable clock gating */
-	vcn_v1_0_enable_clock_gating(adev);
 
+	vcn_v1_0_enable_clock_gating(adev);
+	vcn_1_0_enable_static_power_gating(adev);
 	return 0;
 }
 
-- 
cgit v1.1


From 61c8e90d965ba944d8b56c29c2c7bb9ec34f45d5 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Thu, 17 May 2018 16:03:47 +0800
Subject: drm/amdgpu: Enable VCN static PG by default on RV

Enable static VCN powergating by default on Raven.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/soc15.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index 2cf9a18..68b4a22 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -712,7 +712,8 @@ static int soc15_common_early_init(void *handle)
 			AMD_CG_SUPPORT_SDMA_MGCG |
 			AMD_CG_SUPPORT_SDMA_LS |
 			AMD_CG_SUPPORT_VCN_MGCG;
-		adev->pg_flags = AMD_PG_SUPPORT_SDMA;
+
+		adev->pg_flags = AMD_PG_SUPPORT_SDMA | AMD_PG_SUPPORT_VCN;
 
 		if (adev->powerplay.pp_feature & PP_GFXOFF_MASK)
 			adev->pg_flags |= AMD_PG_SUPPORT_GFX_PG |
-- 
cgit v1.1


From 22cc6c5e1958e5a08b4c44203d1810ab07ce5a16 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Wed, 16 May 2018 20:18:22 +0800
Subject: drm/amdgpu: Add runtime VCN PG support

Enable support for dynamically powering up/down VCN on demand.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c | 16 +++++++++-------
 drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c   | 32 +++++++++++++++++++++-----------
 2 files changed, 30 insertions(+), 18 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
index 6046838..8851bcd 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
@@ -212,11 +212,11 @@ static void amdgpu_vcn_idle_work_handler(struct work_struct *work)
 	}
 
 	if (fences == 0) {
-		if (adev->pm.dpm_enabled) {
-			/* might be used when with pg/cg
+		if (adev->pm.dpm_enabled)
 			amdgpu_dpm_enable_uvd(adev, false);
-			*/
-		}
+		else
+			amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_VCN,
+							       AMD_PG_STATE_GATE);
 	} else {
 		schedule_delayed_work(&adev->vcn.idle_work, VCN_IDLE_TIMEOUT);
 	}
@@ -228,9 +228,11 @@ void amdgpu_vcn_ring_begin_use(struct amdgpu_ring *ring)
 	bool set_clocks = !cancel_delayed_work_sync(&adev->vcn.idle_work);
 
 	if (set_clocks && adev->pm.dpm_enabled) {
-		/* might be used when with pg/cg
-		amdgpu_dpm_enable_uvd(adev, true);
-		*/
+		if (adev->pm.dpm_enabled)
+			amdgpu_dpm_enable_uvd(adev, true);
+		else
+			amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_VCN,
+							       AMD_PG_STATE_UNGATE);
 	}
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
index dcb60ee..110b294 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
@@ -35,7 +35,6 @@
 #include "mmhub/mmhub_9_1_offset.h"
 #include "mmhub/mmhub_9_1_sh_mask.h"
 
-static int vcn_v1_0_start(struct amdgpu_device *adev);
 static int vcn_v1_0_stop(struct amdgpu_device *adev);
 static void vcn_v1_0_set_dec_ring_funcs(struct amdgpu_device *adev);
 static void vcn_v1_0_set_enc_ring_funcs(struct amdgpu_device *adev);
@@ -146,10 +145,6 @@ static int vcn_v1_0_hw_init(void *handle)
 	struct amdgpu_ring *ring = &adev->vcn.ring_dec;
 	int i, r;
 
-	r = vcn_v1_0_start(adev);
-	if (r)
-		goto done;
-
 	ring->ready = true;
 	r = amdgpu_ring_test_ring(ring);
 	if (r) {
@@ -185,11 +180,9 @@ static int vcn_v1_0_hw_fini(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 	struct amdgpu_ring *ring = &adev->vcn.ring_dec;
-	int r;
 
-	r = vcn_v1_0_stop(adev);
-	if (r)
-		return r;
+	if (RREG32_SOC15(VCN, 0, mmUVD_STATUS))
+		vcn_v1_0_stop(adev);
 
 	ring->ready = false;
 
@@ -769,7 +762,7 @@ static int vcn_v1_0_stop(struct amdgpu_device *adev)
 	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_CTRL2), 0,
 			~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
 
-	/* enable clock gating */
+	WREG32_SOC15(VCN, 0, mmUVD_STATUS, 0);
 
 	vcn_v1_0_enable_clock_gating(adev);
 	vcn_1_0_enable_static_power_gating(adev);
@@ -1179,6 +1172,23 @@ static void vcn_v1_0_dec_ring_insert_nop(struct amdgpu_ring *ring, uint32_t coun
 	}
 }
 
+static int vcn_v1_0_set_powergating_state(void *handle,
+					  enum amd_powergating_state state)
+{
+	/* This doesn't actually powergate the VCN block.
+	 * That's done in the dpm code via the SMC.  This
+	 * just re-inits the block as necessary.  The actual
+	 * gating still happens in the dpm code.  We should
+	 * revisit this when there is a cleaner line between
+	 * the smc and the hw blocks
+	 */
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+	if (state == AMD_PG_STATE_GATE)
+		return vcn_v1_0_stop(adev);
+	else
+		return vcn_v1_0_start(adev);
+}
 
 static const struct amd_ip_funcs vcn_v1_0_ip_funcs = {
 	.name = "vcn_v1_0",
@@ -1197,7 +1207,7 @@ static const struct amd_ip_funcs vcn_v1_0_ip_funcs = {
 	.soft_reset = NULL /* vcn_v1_0_soft_reset */,
 	.post_soft_reset = NULL /* vcn_v1_0_post_soft_reset */,
 	.set_clockgating_state = vcn_v1_0_set_clockgating_state,
-	.set_powergating_state = NULL /* vcn_v1_0_set_powergating_state */,
+	.set_powergating_state = vcn_v1_0_set_powergating_state,
 };
 
 static const struct amdgpu_ring_funcs vcn_v1_0_dec_ring_vm_funcs = {
-- 
cgit v1.1


From af4c0f650b563c7b30c1d8cd2bb926247ceb19cc Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Christian=20K=C3=B6nig?= <christian.koenig@amd.com>
Date: Thu, 19 Apr 2018 10:56:02 +0200
Subject: drm/amdgpu: rework VM state machine lock handling v2
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Only the moved state needs a separate spin lock protection. All other
states are protected by reserving the VM anyway.

v2: fix some more incorrect cases

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 66 +++++++++++-----------------------
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h |  4 +--
 2 files changed, 21 insertions(+), 49 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index 1a8f4e0..f0deedc 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -119,9 +119,7 @@ static void amdgpu_vm_bo_base_init(struct amdgpu_vm_bo_base *base,
 	 * is currently evicted. add the bo to the evicted list to make sure it
 	 * is validated on next vm use to avoid fault.
 	 * */
-	spin_lock(&vm->status_lock);
 	list_move_tail(&base->vm_status, &vm->evicted);
-	spin_unlock(&vm->status_lock);
 }
 
 /**
@@ -228,7 +226,6 @@ int amdgpu_vm_validate_pt_bos(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 	struct ttm_bo_global *glob = adev->mman.bdev.glob;
 	int r;
 
-	spin_lock(&vm->status_lock);
 	while (!list_empty(&vm->evicted)) {
 		struct amdgpu_vm_bo_base *bo_base;
 		struct amdgpu_bo *bo;
@@ -236,10 +233,8 @@ int amdgpu_vm_validate_pt_bos(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 		bo_base = list_first_entry(&vm->evicted,
 					   struct amdgpu_vm_bo_base,
 					   vm_status);
-		spin_unlock(&vm->status_lock);
 
 		bo = bo_base->bo;
-		BUG_ON(!bo);
 		if (bo->parent) {
 			r = validate(param, bo);
 			if (r)
@@ -259,13 +254,14 @@ int amdgpu_vm_validate_pt_bos(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 				return r;
 		}
 
-		spin_lock(&vm->status_lock);
-		if (bo->tbo.type != ttm_bo_type_kernel)
+		if (bo->tbo.type != ttm_bo_type_kernel) {
+			spin_lock(&vm->moved_lock);
 			list_move(&bo_base->vm_status, &vm->moved);
-		else
+			spin_unlock(&vm->moved_lock);
+		} else {
 			list_move(&bo_base->vm_status, &vm->relocated);
+		}
 	}
-	spin_unlock(&vm->status_lock);
 
 	return 0;
 }
@@ -279,13 +275,7 @@ int amdgpu_vm_validate_pt_bos(struct amdgpu_device *adev, struct amdgpu_vm *vm,
  */
 bool amdgpu_vm_ready(struct amdgpu_vm *vm)
 {
-	bool ready;
-
-	spin_lock(&vm->status_lock);
-	ready = list_empty(&vm->evicted);
-	spin_unlock(&vm->status_lock);
-
-	return ready;
+	return list_empty(&vm->evicted);
 }
 
 /**
@@ -477,9 +467,7 @@ static int amdgpu_vm_alloc_levels(struct amdgpu_device *adev,
 			pt->parent = amdgpu_bo_ref(parent->base.bo);
 
 			amdgpu_vm_bo_base_init(&entry->base, vm, pt);
-			spin_lock(&vm->status_lock);
 			list_move(&entry->base.vm_status, &vm->relocated);
-			spin_unlock(&vm->status_lock);
 		}
 
 		if (level < AMDGPU_VM_PTB) {
@@ -926,10 +914,8 @@ static void amdgpu_vm_invalidate_level(struct amdgpu_device *adev,
 		if (!entry->base.bo)
 			continue;
 
-		spin_lock(&vm->status_lock);
 		if (list_empty(&entry->base.vm_status))
 			list_add(&entry->base.vm_status, &vm->relocated);
-		spin_unlock(&vm->status_lock);
 		amdgpu_vm_invalidate_level(adev, vm, entry, level + 1);
 	}
 }
@@ -974,7 +960,6 @@ restart:
 		params.func = amdgpu_vm_do_set_ptes;
 	}
 
-	spin_lock(&vm->status_lock);
 	while (!list_empty(&vm->relocated)) {
 		struct amdgpu_vm_bo_base *bo_base, *parent;
 		struct amdgpu_vm_pt *pt, *entry;
@@ -984,13 +969,10 @@ restart:
 					   struct amdgpu_vm_bo_base,
 					   vm_status);
 		list_del_init(&bo_base->vm_status);
-		spin_unlock(&vm->status_lock);
 
 		bo = bo_base->bo->parent;
-		if (!bo) {
-			spin_lock(&vm->status_lock);
+		if (!bo)
 			continue;
-		}
 
 		parent = list_first_entry(&bo->va, struct amdgpu_vm_bo_base,
 					  bo_list);
@@ -999,12 +981,10 @@ restart:
 
 		amdgpu_vm_update_pde(&params, vm, pt, entry);
 
-		spin_lock(&vm->status_lock);
 		if (!vm->use_cpu_for_update &&
 		    (ndw - params.ib->length_dw) < 32)
 			break;
 	}
-	spin_unlock(&vm->status_lock);
 
 	if (vm->use_cpu_for_update) {
 		/* Flush HDP */
@@ -1107,9 +1087,7 @@ static void amdgpu_vm_handle_huge_pages(struct amdgpu_pte_update_params *p,
 		if (entry->huge) {
 			/* Add the entry to the relocated list to update it. */
 			entry->huge = false;
-			spin_lock(&p->vm->status_lock);
 			list_move(&entry->base.vm_status, &p->vm->relocated);
-			spin_unlock(&p->vm->status_lock);
 		}
 		return;
 	}
@@ -1588,8 +1566,9 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev,
 		amdgpu_asic_flush_hdp(adev, NULL);
 	}
 
-	spin_lock(&vm->status_lock);
+	spin_lock(&vm->moved_lock);
 	list_del_init(&bo_va->base.vm_status);
+	spin_unlock(&vm->moved_lock);
 
 	/* If the BO is not in its preferred location add it back to
 	 * the evicted list so that it gets validated again on the
@@ -1599,7 +1578,6 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev,
 	    !(bo->preferred_domains &
 	    amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type)))
 		list_add_tail(&bo_va->base.vm_status, &vm->evicted);
-	spin_unlock(&vm->status_lock);
 
 	list_splice_init(&bo_va->invalids, &bo_va->valids);
 	bo_va->cleared = clear;
@@ -1811,14 +1789,14 @@ int amdgpu_vm_handle_moved(struct amdgpu_device *adev,
 	bool clear;
 	int r = 0;
 
-	spin_lock(&vm->status_lock);
+	spin_lock(&vm->moved_lock);
 	while (!list_empty(&vm->moved)) {
 		struct amdgpu_bo_va *bo_va;
 		struct reservation_object *resv;
 
 		bo_va = list_first_entry(&vm->moved,
 			struct amdgpu_bo_va, base.vm_status);
-		spin_unlock(&vm->status_lock);
+		spin_unlock(&vm->moved_lock);
 
 		resv = bo_va->base.bo->tbo.resv;
 
@@ -1839,9 +1817,9 @@ int amdgpu_vm_handle_moved(struct amdgpu_device *adev,
 		if (!clear && resv != vm->root.base.bo->tbo.resv)
 			reservation_object_unlock(resv);
 
-		spin_lock(&vm->status_lock);
+		spin_lock(&vm->moved_lock);
 	}
-	spin_unlock(&vm->status_lock);
+	spin_unlock(&vm->moved_lock);
 
 	return r;
 }
@@ -1903,10 +1881,10 @@ static void amdgpu_vm_bo_insert_map(struct amdgpu_device *adev,
 		amdgpu_vm_prt_get(adev);
 
 	if (bo && bo->tbo.resv == vm->root.base.bo->tbo.resv) {
-		spin_lock(&vm->status_lock);
+		spin_lock(&vm->moved_lock);
 		if (list_empty(&bo_va->base.vm_status))
 			list_add(&bo_va->base.vm_status, &vm->moved);
-		spin_unlock(&vm->status_lock);
+		spin_unlock(&vm->moved_lock);
 	}
 	trace_amdgpu_vm_bo_map(bo_va, mapping);
 }
@@ -2216,9 +2194,9 @@ void amdgpu_vm_bo_rmv(struct amdgpu_device *adev,
 
 	list_del(&bo_va->base.bo_list);
 
-	spin_lock(&vm->status_lock);
+	spin_lock(&vm->moved_lock);
 	list_del(&bo_va->base.vm_status);
-	spin_unlock(&vm->status_lock);
+	spin_unlock(&vm->moved_lock);
 
 	list_for_each_entry_safe(mapping, next, &bo_va->valids, list) {
 		list_del(&mapping->list);
@@ -2261,28 +2239,24 @@ void amdgpu_vm_bo_invalidate(struct amdgpu_device *adev,
 
 		bo_base->moved = true;
 		if (evicted && bo->tbo.resv == vm->root.base.bo->tbo.resv) {
-			spin_lock(&bo_base->vm->status_lock);
 			if (bo->tbo.type == ttm_bo_type_kernel)
 				list_move(&bo_base->vm_status, &vm->evicted);
 			else
 				list_move_tail(&bo_base->vm_status,
 					       &vm->evicted);
-			spin_unlock(&bo_base->vm->status_lock);
 			continue;
 		}
 
 		if (bo->tbo.type == ttm_bo_type_kernel) {
-			spin_lock(&bo_base->vm->status_lock);
 			if (list_empty(&bo_base->vm_status))
 				list_add(&bo_base->vm_status, &vm->relocated);
-			spin_unlock(&bo_base->vm->status_lock);
 			continue;
 		}
 
-		spin_lock(&bo_base->vm->status_lock);
+		spin_lock(&bo_base->vm->moved_lock);
 		if (list_empty(&bo_base->vm_status))
 			list_add(&bo_base->vm_status, &vm->moved);
-		spin_unlock(&bo_base->vm->status_lock);
+		spin_unlock(&bo_base->vm->moved_lock);
 	}
 }
 
@@ -2391,9 +2365,9 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 	vm->va = RB_ROOT_CACHED;
 	for (i = 0; i < AMDGPU_MAX_VMHUBS; i++)
 		vm->reserved_vmid[i] = NULL;
-	spin_lock_init(&vm->status_lock);
 	INIT_LIST_HEAD(&vm->evicted);
 	INIT_LIST_HEAD(&vm->relocated);
+	spin_lock_init(&vm->moved_lock);
 	INIT_LIST_HEAD(&vm->moved);
 	INIT_LIST_HEAD(&vm->freed);
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
index d682708..0196b9a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
@@ -168,9 +168,6 @@ struct amdgpu_vm {
 	/* tree of virtual addresses mapped */
 	struct rb_root_cached	va;
 
-	/* protecting invalidated */
-	spinlock_t		status_lock;
-
 	/* BOs who needs a validation */
 	struct list_head	evicted;
 
@@ -179,6 +176,7 @@ struct amdgpu_vm {
 
 	/* BOs moved, but not yet updated in the PT */
 	struct list_head	moved;
+	spinlock_t		moved_lock;
 
 	/* BO mappings freed, but not yet updated in the PT */
 	struct list_head	freed;
-- 
cgit v1.1


From 91ccdd24a1955dbec97a6d61322be214b7de1974 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Christian=20K=C3=B6nig?= <christian.koenig@amd.com>
Date: Thu, 19 Apr 2018 11:02:54 +0200
Subject: drm/amdgpu: cleanup amdgpu_vm_validate_pt_bos v2
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Use list_for_each_entry_safe here.

v2: Drop the optimization, it doesn't work as expected.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 19 +++++++------------
 1 file changed, 7 insertions(+), 12 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index f0deedc..3be4d5f 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -224,21 +224,16 @@ int amdgpu_vm_validate_pt_bos(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 			      void *param)
 {
 	struct ttm_bo_global *glob = adev->mman.bdev.glob;
-	int r;
-
-	while (!list_empty(&vm->evicted)) {
-		struct amdgpu_vm_bo_base *bo_base;
-		struct amdgpu_bo *bo;
+	struct amdgpu_vm_bo_base *bo_base, *tmp;
+	int r = 0;
 
-		bo_base = list_first_entry(&vm->evicted,
-					   struct amdgpu_vm_bo_base,
-					   vm_status);
+	list_for_each_entry_safe(bo_base, tmp, &vm->evicted, vm_status) {
+		struct amdgpu_bo *bo = bo_base->bo;
 
-		bo = bo_base->bo;
 		if (bo->parent) {
 			r = validate(param, bo);
 			if (r)
-				return r;
+				break;
 
 			spin_lock(&glob->lru_lock);
 			ttm_bo_move_to_lru_tail(&bo->tbo);
@@ -251,7 +246,7 @@ int amdgpu_vm_validate_pt_bos(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 		    vm->use_cpu_for_update) {
 			r = amdgpu_bo_kmap(bo, NULL);
 			if (r)
-				return r;
+				break;
 		}
 
 		if (bo->tbo.type != ttm_bo_type_kernel) {
@@ -263,7 +258,7 @@ int amdgpu_vm_validate_pt_bos(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 		}
 	}
 
-	return 0;
+	return r;
 }
 
 /**
-- 
cgit v1.1


From 789f3317ed33e34fa97c8918c075c68a62e51a4d Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Christian=20K=C3=B6nig?= <christian.koenig@amd.com>
Date: Thu, 19 Apr 2018 11:08:24 +0200
Subject: drm/amdgpu: further optimize amdgpu_vm_handle_moved
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Splice the moved list to a local one to avoid taking the lock over and
over again.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 27 ++++++++++++++-------------
 1 file changed, 14 insertions(+), 13 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index 3be4d5f..4d88b06 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -1781,19 +1781,18 @@ int amdgpu_vm_clear_freed(struct amdgpu_device *adev,
 int amdgpu_vm_handle_moved(struct amdgpu_device *adev,
 			   struct amdgpu_vm *vm)
 {
+	struct amdgpu_bo_va *bo_va, *tmp;
+	struct list_head moved;
 	bool clear;
-	int r = 0;
+	int r;
 
+	INIT_LIST_HEAD(&moved);
 	spin_lock(&vm->moved_lock);
-	while (!list_empty(&vm->moved)) {
-		struct amdgpu_bo_va *bo_va;
-		struct reservation_object *resv;
-
-		bo_va = list_first_entry(&vm->moved,
-			struct amdgpu_bo_va, base.vm_status);
-		spin_unlock(&vm->moved_lock);
+	list_splice_init(&vm->moved, &moved);
+	spin_unlock(&vm->moved_lock);
 
-		resv = bo_va->base.bo->tbo.resv;
+	list_for_each_entry_safe(bo_va, tmp, &moved, base.vm_status) {
+		struct reservation_object *resv = bo_va->base.bo->tbo.resv;
 
 		/* Per VM BOs never need to bo cleared in the page tables */
 		if (resv == vm->root.base.bo->tbo.resv)
@@ -1806,17 +1805,19 @@ int amdgpu_vm_handle_moved(struct amdgpu_device *adev,
 			clear = true;
 
 		r = amdgpu_vm_bo_update(adev, bo_va, clear);
-		if (r)
+		if (r) {
+			spin_lock(&vm->moved_lock);
+			list_splice(&moved, &vm->moved);
+			spin_unlock(&vm->moved_lock);
 			return r;
+		}
 
 		if (!clear && resv != vm->root.base.bo->tbo.resv)
 			reservation_object_unlock(resv);
 
-		spin_lock(&vm->moved_lock);
 	}
-	spin_unlock(&vm->moved_lock);
 
-	return r;
+	return 0;
 }
 
 /**
-- 
cgit v1.1


From a7f91061c60ad9cac2e6a03b642be6a4f88b3662 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Christian=20K=C3=B6nig?= <christian.koenig@amd.com>
Date: Thu, 19 Apr 2018 13:58:42 +0200
Subject: drm/amdgpu: kmap PDs/PTs in amdgpu_vm_update_directories
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

In theory it is possible that PDs/PTs can move without eviction.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 15 ++++++++-------
 1 file changed, 8 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index 4d88b06..a31afac 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -242,13 +242,6 @@ int amdgpu_vm_validate_pt_bos(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 			spin_unlock(&glob->lru_lock);
 		}
 
-		if (bo->tbo.type == ttm_bo_type_kernel &&
-		    vm->use_cpu_for_update) {
-			r = amdgpu_bo_kmap(bo, NULL);
-			if (r)
-				break;
-		}
-
 		if (bo->tbo.type != ttm_bo_type_kernel) {
 			spin_lock(&vm->moved_lock);
 			list_move(&bo_base->vm_status, &vm->moved);
@@ -940,6 +933,14 @@ restart:
 	params.adev = adev;
 
 	if (vm->use_cpu_for_update) {
+		struct amdgpu_vm_bo_base *bo_base;
+
+		list_for_each_entry(bo_base, &vm->relocated, vm_status) {
+			r = amdgpu_bo_kmap(bo_base->bo, NULL);
+			if (unlikely(r))
+				return r;
+		}
+
 		r = amdgpu_vm_wait_pd(adev, vm, AMDGPU_FENCE_OWNER_VM);
 		if (unlikely(r))
 			return r;
-- 
cgit v1.1


From 862b8c5762e4e2324d18c881ce86062af72b2063 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Christian=20K=C3=B6nig?= <christian.koenig@amd.com>
Date: Thu, 19 Apr 2018 14:22:56 +0200
Subject: drm/amdgpu: consistenly use VM moved flag
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Instead of sometimes checking if the vm_status is empty use the moved
flag and also reset it when the BO leaves the state machine.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 28 +++++++++++++++-------------
 1 file changed, 15 insertions(+), 13 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index a31afac..f5dee4c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -902,8 +902,8 @@ static void amdgpu_vm_invalidate_level(struct amdgpu_device *adev,
 		if (!entry->base.bo)
 			continue;
 
-		if (list_empty(&entry->base.vm_status))
-			list_add(&entry->base.vm_status, &vm->relocated);
+		if (!entry->base.moved)
+			list_move(&entry->base.vm_status, &vm->relocated);
 		amdgpu_vm_invalidate_level(adev, vm, entry, level + 1);
 	}
 }
@@ -964,6 +964,7 @@ restart:
 		bo_base = list_first_entry(&vm->relocated,
 					   struct amdgpu_vm_bo_base,
 					   vm_status);
+		bo_base->moved = false;
 		list_del_init(&bo_base->vm_status);
 
 		bo = bo_base->bo->parent;
@@ -1877,10 +1878,10 @@ static void amdgpu_vm_bo_insert_map(struct amdgpu_device *adev,
 	if (mapping->flags & AMDGPU_PTE_PRT)
 		amdgpu_vm_prt_get(adev);
 
-	if (bo && bo->tbo.resv == vm->root.base.bo->tbo.resv) {
+	if (bo && bo->tbo.resv == vm->root.base.bo->tbo.resv &&
+	    !bo_va->base.moved) {
 		spin_lock(&vm->moved_lock);
-		if (list_empty(&bo_va->base.vm_status))
-			list_add(&bo_va->base.vm_status, &vm->moved);
+		list_move(&bo_va->base.vm_status, &vm->moved);
 		spin_unlock(&vm->moved_lock);
 	}
 	trace_amdgpu_vm_bo_map(bo_va, mapping);
@@ -2233,6 +2234,7 @@ void amdgpu_vm_bo_invalidate(struct amdgpu_device *adev,
 
 	list_for_each_entry(bo_base, &bo->va, bo_list) {
 		struct amdgpu_vm *vm = bo_base->vm;
+		bool was_moved = bo_base->moved;
 
 		bo_base->moved = true;
 		if (evicted && bo->tbo.resv == vm->root.base.bo->tbo.resv) {
@@ -2244,16 +2246,16 @@ void amdgpu_vm_bo_invalidate(struct amdgpu_device *adev,
 			continue;
 		}
 
-		if (bo->tbo.type == ttm_bo_type_kernel) {
-			if (list_empty(&bo_base->vm_status))
-				list_add(&bo_base->vm_status, &vm->relocated);
+		if (was_moved)
 			continue;
-		}
 
-		spin_lock(&bo_base->vm->moved_lock);
-		if (list_empty(&bo_base->vm_status))
-			list_add(&bo_base->vm_status, &vm->moved);
-		spin_unlock(&bo_base->vm->moved_lock);
+		if (bo->tbo.type == ttm_bo_type_kernel) {
+			list_move(&bo_base->vm_status, &vm->relocated);
+		} else {
+			spin_lock(&bo_base->vm->moved_lock);
+			list_move(&bo_base->vm_status, &vm->moved);
+			spin_unlock(&bo_base->vm->moved_lock);
+		}
 	}
 }
 
-- 
cgit v1.1


From 806f043f0253a76248c554ce9f7303bc25e43314 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Christian=20K=C3=B6nig?= <christian.koenig@amd.com>
Date: Thu, 19 Apr 2018 15:01:12 +0200
Subject: drm/amdgpu: move VM BOs on LRU again
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Move all BOs belonging to a VM on the LRU with every submission.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 28 +++++++++++++++++++++++-----
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h |  3 +++
 2 files changed, 26 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index f5dee4c..ccba88c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -251,6 +251,19 @@ int amdgpu_vm_validate_pt_bos(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 		}
 	}
 
+	spin_lock(&glob->lru_lock);
+	list_for_each_entry(bo_base, &vm->idle, vm_status) {
+		struct amdgpu_bo *bo = bo_base->bo;
+
+		if (!bo->parent)
+			continue;
+
+		ttm_bo_move_to_lru_tail(&bo->tbo);
+		if (bo->shadow)
+			ttm_bo_move_to_lru_tail(&bo->shadow->tbo);
+	}
+	spin_unlock(&glob->lru_lock);
+
 	return r;
 }
 
@@ -965,7 +978,7 @@ restart:
 					   struct amdgpu_vm_bo_base,
 					   vm_status);
 		bo_base->moved = false;
-		list_del_init(&bo_base->vm_status);
+		list_move(&bo_base->vm_status, &vm->idle);
 
 		bo = bo_base->bo->parent;
 		if (!bo)
@@ -1571,10 +1584,14 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev,
 	 * the evicted list so that it gets validated again on the
 	 * next command submission.
 	 */
-	if (bo && bo->tbo.resv == vm->root.base.bo->tbo.resv &&
-	    !(bo->preferred_domains &
-	    amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type)))
-		list_add_tail(&bo_va->base.vm_status, &vm->evicted);
+	if (bo && bo->tbo.resv == vm->root.base.bo->tbo.resv) {
+		uint32_t mem_type = bo->tbo.mem.mem_type;
+
+		if (!(bo->preferred_domains & amdgpu_mem_type_to_domain(mem_type)))
+			list_add_tail(&bo_va->base.vm_status, &vm->evicted);
+		else
+			list_add(&bo_va->base.vm_status, &vm->idle);
+	}
 
 	list_splice_init(&bo_va->invalids, &bo_va->valids);
 	bo_va->cleared = clear;
@@ -2368,6 +2385,7 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 	INIT_LIST_HEAD(&vm->relocated);
 	spin_lock_init(&vm->moved_lock);
 	INIT_LIST_HEAD(&vm->moved);
+	INIT_LIST_HEAD(&vm->idle);
 	INIT_LIST_HEAD(&vm->freed);
 
 	/* create scheduler entity for page table updates */
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
index 0196b9a..061b99a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
@@ -178,6 +178,9 @@ struct amdgpu_vm {
 	struct list_head	moved;
 	spinlock_t		moved_lock;
 
+	/* All BOs of this VM not currently in the state machine */
+	struct list_head	idle;
+
 	/* BO mappings freed, but not yet updated in the PT */
 	struct list_head	freed;
 
-- 
cgit v1.1


From 52bf20f414bc34899203307034227ccc95d6adb3 Mon Sep 17 00:00:00 2001
From: Emily Deng <Emily.Deng@amd.com>
Date: Wed, 23 May 2018 15:53:03 +0800
Subject: drm/sched: add rcu_barrier after entity fini
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

To free the fence from the amdgpu_fence_slab, need twice call_rcu, to avoid
the amdgpu_fence_slab_fini call kmem_cache_destroy(amdgpu_fence_slab) before
kmem_cache_free(amdgpu_fence_slab, fence), add rcu_barrier after drm_sched_entity_fini.

The kmem_cache_free(amdgpu_fence_slab, fence)'s call trace as below:
1.drm_sched_entity_fini ->
drm_sched_entity_cleanup ->
dma_fence_put(entity->last_scheduled) ->
drm_sched_fence_release_finished ->
drm_sched_fence_release_scheduled ->
call_rcu(&fence->finished.rcu, drm_sched_fence_free)

2.drm_sched_fence_free ->
dma_fence_put(fence->parent) ->
amdgpu_fence_release ->
call_rcu(&f->rcu, amdgpu_fence_free) ->
kmem_cache_free(amdgpu_fence_slab, fence);

v2:put the barrier before the kmem_cache_destroy
v3:put the dma_fence_put(fence->parent) before call_rcu in
drm_sched_fence_release_scheduled

Signed-off-by: Emily Deng <Emily.Deng@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/scheduler/sched_fence.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/scheduler/sched_fence.c b/drivers/gpu/drm/scheduler/sched_fence.c
index 786b47f..df44616 100644
--- a/drivers/gpu/drm/scheduler/sched_fence.c
+++ b/drivers/gpu/drm/scheduler/sched_fence.c
@@ -98,7 +98,6 @@ static void drm_sched_fence_free(struct rcu_head *rcu)
 	struct dma_fence *f = container_of(rcu, struct dma_fence, rcu);
 	struct drm_sched_fence *fence = to_drm_sched_fence(f);
 
-	dma_fence_put(fence->parent);
 	kmem_cache_free(sched_fence_slab, fence);
 }
 
@@ -114,6 +113,7 @@ static void drm_sched_fence_release_scheduled(struct dma_fence *f)
 {
 	struct drm_sched_fence *fence = to_drm_sched_fence(f);
 
+	dma_fence_put(fence->parent);
 	call_rcu(&fence->finished.rcu, drm_sched_fence_free);
 }
 
-- 
cgit v1.1


From bf20f0ab544d8982af375c87e3d870d45237eecc Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Mon, 21 May 2018 10:16:28 -0500
Subject: drm/amdgpu/pp: remove duplicate assignment
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

is_dpm_running callback was assigned to the same value
twice.  Drop the duplicate.

Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c
index a40f714..2de4895 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c
@@ -2379,6 +2379,5 @@ const struct pp_smumgr_func vegam_smu_funcs = {
 	.update_sclk_threshold = vegam_update_sclk_threshold,
 	.is_hw_avfs_present = vegam_is_hw_avfs_present,
 	.thermal_avfs_enable = vegam_thermal_avfs_enable,
-	.is_dpm_running = vegam_is_dpm_running,
 	.thermal_setup_fan_table = vegam_thermal_setup_fan_table,
 };
-- 
cgit v1.1


From f9fb22a21b380b14f70048fe719875e3523ac7d8 Mon Sep 17 00:00:00 2001
From: Shaoyun Liu <Shaoyun.Liu@amd.com>
Date: Tue, 22 May 2018 11:45:41 -0400
Subject: drm/amdgpu: Update GFX info structure to match what vega20 used

Update to the latest version from the vbios team.

Signed-off-by: Shaoyun Liu <Shaoyun.Liu@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c | 8 +++-----
 drivers/gpu/drm/amd/include/atomfirmware.h       | 3 ++-
 2 files changed, 5 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c
index 7014d58..2369158 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c
@@ -354,11 +354,9 @@ int amdgpu_atomfirmware_get_gfx_info(struct amdgpu_device *adev)
 				le16_to_cpu(gfx_info->v24.gc_gsprim_buff_depth);
 			adev->gfx.config.double_offchip_lds_buf =
 				gfx_info->v24.gc_double_offchip_lds_buffer;
-			adev->gfx.cu_info.wave_front_size = gfx_info->v24.gc_wave_size;
-			adev->gfx.cu_info.max_waves_per_simd =
-				le16_to_cpu(gfx_info->v24.gc_max_waves_per_simd);
-			adev->gfx.cu_info.max_scratch_slots_per_cu =
-				gfx_info->v24.gc_max_scratch_slots_per_cu;
+			adev->gfx.cu_info.wave_front_size = le16_to_cpu(gfx_info->v24.gc_wave_size);
+			adev->gfx.cu_info.max_waves_per_simd = le16_to_cpu(gfx_info->v24.gc_max_waves_per_simd);
+			adev->gfx.cu_info.max_scratch_slots_per_cu = gfx_info->v24.gc_max_scratch_slots_per_cu;
 			adev->gfx.cu_info.lds_size = le16_to_cpu(gfx_info->v24.gc_lds_size);
 			return 0;
 		default:
diff --git a/drivers/gpu/drm/amd/include/atomfirmware.h b/drivers/gpu/drm/amd/include/atomfirmware.h
index fd5e80c..c6c1666 100644
--- a/drivers/gpu/drm/amd/include/atomfirmware.h
+++ b/drivers/gpu/drm/amd/include/atomfirmware.h
@@ -1240,7 +1240,6 @@ struct  atom_gfx_info_v2_4 {
   uint8_t active_cu_per_sh;
   uint8_t active_rb_per_se;
   uint16_t gcgoldenoffset;
-  uint32_t rm21_sram_vmin_value;
   uint16_t gc_num_gprs;
   uint16_t gc_gsprim_buff_depth;
   uint16_t gc_parameter_cache_depth;
@@ -1251,6 +1250,8 @@ struct  atom_gfx_info_v2_4 {
   uint8_t gc_gs_table_depth;
   uint8_t gc_double_offchip_lds_buffer;
   uint8_t gc_max_scratch_slots_per_cu;
+  uint32_t sram_rm_fuses_val;
+  uint32_t sram_custom_rm_fuses_val;
 };
 
 /* 
-- 
cgit v1.1


From b8f3439fa5358ac84d29fa2f4afa115500dec74c Mon Sep 17 00:00:00 2001
From: David Francis <David.Francis@amd.com>
Date: Thu, 24 May 2018 10:40:12 -0400
Subject: drm/amd/display: Remove use of division operator for long longs

In fixed31_32.h, in dc_fixpt_shl,'/' was used for division of one long
long int by another long long int.  As there is no inbuilt long long
int division function in c, gcc inserted its own.  However, gcc does not
link the library that contains this function.  To avoid this, use
bitwise operators instead of /

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: David Francis <David.Francis@amd.com>
Reviewed-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/include/fixed31_32.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/include/fixed31_32.h b/drivers/gpu/drm/amd/display/include/fixed31_32.h
index 76f64e9..bb0d4eb 100644
--- a/drivers/gpu/drm/amd/display/include/fixed31_32.h
+++ b/drivers/gpu/drm/amd/display/include/fixed31_32.h
@@ -209,7 +209,7 @@ static inline struct fixed31_32 dc_fixpt_clamp(
 static inline struct fixed31_32 dc_fixpt_shl(struct fixed31_32 arg, unsigned char shift)
 {
 	ASSERT(((arg.value >= 0) && (arg.value <= LLONG_MAX >> shift)) ||
-		((arg.value < 0) && (arg.value >= (LLONG_MIN / (1LL << shift)))));
+		((arg.value < 0) && (arg.value >= ~(LLONG_MAX >> shift))));
 
 	arg.value = arg.value << shift;
 
-- 
cgit v1.1


From c3032fd9673468783bb20326ead823a2f321a522 Mon Sep 17 00:00:00 2001
From: Tom Stellard <tstellar@redhat.com>
Date: Thu, 24 May 2018 14:07:14 -0700
Subject: drm/amdgpu: Use dev_info() to report amdkfd is not supported for this
 ASIC

This is an important message, so it should be visible to users without
having to enable extra debugging.

Signed-off-by: Tom Stellard <tstellar@redhat.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
index 95fcbd8..8f6f455 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
@@ -103,7 +103,7 @@ void amdgpu_amdkfd_device_probe(struct amdgpu_device *adev)
 		kfd2kgd = amdgpu_amdkfd_gfx_9_0_get_functions();
 		break;
 	default:
-		dev_dbg(adev->dev, "kfd not supported on this ASIC\n");
+		dev_info(adev->dev, "kfd not supported on this ASIC\n");
 		return;
 	}
 
-- 
cgit v1.1


From 12678199c77d196ee23ca2cde4318b6a0e343687 Mon Sep 17 00:00:00 2001
From: Inki Dae <inki.dae@samsung.com>
Date: Wed, 23 May 2018 09:48:15 +0900
Subject: drm/exynos: scaler: fix static checker warning

drivers/gpu/drm/exynos/exynos_drm_scaler.c:402 scaler_task_done()
warn: signedness bug returning '(-22)'

Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
---
 drivers/gpu/drm/exynos/exynos_drm_scaler.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/exynos/exynos_drm_scaler.c b/drivers/gpu/drm/exynos/exynos_drm_scaler.c
index 63b05b7..91d4382 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_scaler.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_scaler.c
@@ -397,7 +397,7 @@ static inline u32 scaler_get_int_status(struct scaler_context *scaler)
 	return scaler_read(SCALER_INT_STATUS);
 }
 
-static inline bool scaler_task_done(u32 val)
+static inline int scaler_task_done(u32 val)
 {
 	return val & SCALER_INT_STATUS_FRAME_END ? 0 : -EINVAL;
 }
-- 
cgit v1.1


From 19832055e2bf5e67f506bac62e4e07326fb545b7 Mon Sep 17 00:00:00 2001
From: Dan Carpenter <dan.carpenter@oracle.com>
Date: Fri, 18 May 2018 01:04:45 -0700
Subject: drm/exynos: fimc: signedness bug in fimc_setup_clocks()

"id" needs to be signed for the error handling to work.

Fixes: 7a2d5c77c558 ("drm/exynos: fimc: Convert driver to IPP v2 core API")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
---
 drivers/gpu/drm/exynos/exynos_drm_fimc.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/exynos/exynos_drm_fimc.c b/drivers/gpu/drm/exynos/exynos_drm_fimc.c
index 4dfbfc7..5ce8402 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_fimc.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_fimc.c
@@ -1200,7 +1200,7 @@ e_clk_free:
 
 int exynos_drm_check_fimc_device(struct device *dev)
 {
-	unsigned int id = of_alias_get_id(dev->of_node, "fimc");
+	int id = of_alias_get_id(dev->of_node, "fimc");
 
 	if (id >= 0 && (BIT(id) & fimc_mask))
 		return 0;
-- 
cgit v1.1


From ebe1d22b57b86b6739f2739b5a0f52435596d84d Mon Sep 17 00:00:00 2001
From: Arnd Bergmann <arnd@arndb.de>
Date: Fri, 25 May 2018 17:50:09 +0200
Subject: drm/amdgpu: fix 32-bit build warning

Casting a pointer to a 64-bit type causes a warning on 32-bit targets:

drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c:473:24: error: cast from pointer to integer of different size [-Werror=pointer-to-int-cast]
          lower_32_bits((uint64_t)wptr));
                        ^
drivers/gpu/drm/amd/amdgpu/amdgpu.h:1701:53: note: in definition of macro 'WREG32'
 #define WREG32(reg, v) amdgpu_mm_wreg(adev, (reg), (v), 0)
                                                     ^
drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c:473:10: note: in expansion of macro 'lower_32_bits'
          lower_32_bits((uint64_t)wptr));
          ^~~~~~~~~~~~~

The correct method is to cast to 'uintptr_t'.

Fixes: d5a114a6c5f7 ("drm/amdgpu: Add GFXv9 kfd2kgd interface functions")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c
index 8f37991..f0c0d39 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c
@@ -470,9 +470,9 @@ static int kgd_hqd_load(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id,
 		WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_HI),
 		       upper_32_bits(guessed_wptr));
 		WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_POLL_ADDR),
-		       lower_32_bits((uint64_t)wptr));
+		       lower_32_bits((uintptr_t)wptr));
 		WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_POLL_ADDR_HI),
-		       upper_32_bits((uint64_t)wptr));
+		       upper_32_bits((uintptr_t)wptr));
 		WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_PQ_WPTR_POLL_CNTL1),
 		       get_queue_mask(adev, pipe_id, queue_id));
 	}
-- 
cgit v1.1


From bfdec234047889f4f6af1ec45c7c502a4405b3fb Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Fri, 18 May 2018 17:07:06 -0400
Subject: drm/amd/display: Implement
 dm_pp_get_clock_levels_by_type_with_latency

This is required so we use the correct minimum clocks for Vega. Without
this pplib will never be able to enter the lowest clock states.

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 .../drm/amd/display/amdgpu_dm/amdgpu_dm_services.c | 46 +++++++++++++++++++++-
 1 file changed, 44 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c
index 0229c7ed..ead3d21 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c
@@ -234,6 +234,34 @@ static void pp_to_dc_clock_levels(
 	}
 }
 
+static void pp_to_dc_clock_levels_with_latency(
+		const struct pp_clock_levels_with_latency *pp_clks,
+		struct dm_pp_clock_levels_with_latency *clk_level_info,
+		enum dm_pp_clock_type dc_clk_type)
+{
+	uint32_t i;
+
+	if (pp_clks->num_levels > DM_PP_MAX_CLOCK_LEVELS) {
+		DRM_INFO("DM_PPLIB: Warning: %s clock: number of levels %d exceeds maximum of %d!\n",
+				DC_DECODE_PP_CLOCK_TYPE(dc_clk_type),
+				pp_clks->num_levels,
+				DM_PP_MAX_CLOCK_LEVELS);
+
+		clk_level_info->num_levels = DM_PP_MAX_CLOCK_LEVELS;
+	} else
+		clk_level_info->num_levels = pp_clks->num_levels;
+
+	DRM_DEBUG("DM_PPLIB: values for %s clock\n",
+			DC_DECODE_PP_CLOCK_TYPE(dc_clk_type));
+
+	for (i = 0; i < clk_level_info->num_levels; i++) {
+		DRM_DEBUG("DM_PPLIB:\t %d\n", pp_clks->data[i].clocks_in_khz);
+		/* translate 10kHz to kHz */
+		clk_level_info->data[i].clocks_in_khz = pp_clks->data[i].clocks_in_khz;
+		clk_level_info->data[i].latency_in_us = pp_clks->data[i].clocks_in_khz;
+	}
+}
+
 bool dm_pp_get_clock_levels_by_type(
 		const struct dc_context *ctx,
 		enum dm_pp_clock_type clk_type,
@@ -311,8 +339,22 @@ bool dm_pp_get_clock_levels_by_type_with_latency(
 	enum dm_pp_clock_type clk_type,
 	struct dm_pp_clock_levels_with_latency *clk_level_info)
 {
-	/* TODO: to be implemented */
-	return false;
+	struct amdgpu_device *adev = ctx->driver_context;
+	void *pp_handle = adev->powerplay.pp_handle;
+	struct pp_clock_levels_with_latency pp_clks = { 0 };
+	const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs;
+
+	if (!pp_funcs->get_clock_by_type_with_latency)
+		return false;
+
+	if (pp_funcs->get_clock_by_type_with_latency(pp_handle,
+						     dc_to_pp_clock_type(clk_type),
+						     &pp_clks))
+		return false;
+
+	pp_to_dc_clock_levels_with_latency(&pp_clks, clk_level_info, clk_type);
+
+	return true;
 }
 
 bool dm_pp_get_clock_levels_by_type_with_voltage(
-- 
cgit v1.1


From adea72c5046f7faffff969ece04c3f31e669edf4 Mon Sep 17 00:00:00 2001
From: kbuild test robot <fengguang.wu@intel.com>
Date: Fri, 25 May 2018 02:54:45 +0800
Subject: drm/amdgpu: vcn_v1_0_is_idle() can be static

Fixes: 9b4c412a654c ("drm/amdgpu: Add static CG control for VCN on RV")
Signed-off-by: kbuild test robot <fengguang.wu@intel.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
index 110b294..29684c3 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
@@ -769,14 +769,14 @@ static int vcn_v1_0_stop(struct amdgpu_device *adev)
 	return 0;
 }
 
-bool vcn_v1_0_is_idle(void *handle)
+static bool vcn_v1_0_is_idle(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
 	return (RREG32_SOC15(VCN, 0, mmUVD_STATUS) == 0x2);
 }
 
-int vcn_v1_0_wait_for_idle(void *handle)
+static int vcn_v1_0_wait_for_idle(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 	int ret = 0;
-- 
cgit v1.1


From 6201e033d77fae5ed6798d3d122643c2fe3c24dd Mon Sep 17 00:00:00 2001
From: Nayan Deshmukh <nayan26deshmukh@gmail.com>
Date: Fri, 25 May 2018 10:15:46 +0530
Subject: drm/scheduler: fix a corner case in dependency optimization
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

When checking for a dependency fence for belonging to the same entity
compare it with scheduled as well finished fence. Earlier we were only
comparing it with the scheduled fence.

Signed-off-by: Nayan Deshmukh <nayan26deshmukh@gmail.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/scheduler/gpu_scheduler.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/scheduler/gpu_scheduler.c b/drivers/gpu/drm/scheduler/gpu_scheduler.c
index df1578d..44d4807 100644
--- a/drivers/gpu/drm/scheduler/gpu_scheduler.c
+++ b/drivers/gpu/drm/scheduler/gpu_scheduler.c
@@ -349,8 +349,13 @@ static bool drm_sched_entity_add_dependency_cb(struct drm_sched_entity *entity)
 	struct dma_fence * fence = entity->dependency;
 	struct drm_sched_fence *s_fence;
 
-	if (fence->context == entity->fence_context) {
-		/* We can ignore fences from ourself */
+	if (fence->context == entity->fence_context ||
+            fence->context == entity->fence_context + 1) {
+                /*
+                 * Fence is a scheduled/finished fence from a job
+                 * which belongs to the same entity, we can ignore
+                 * fences from ourself
+                 */
 		dma_fence_put(entity->dependency);
 		return false;
 	}
-- 
cgit v1.1


From 10dd2b865393bb45526ca342fe69207341f89fd5 Mon Sep 17 00:00:00 2001
From: Harry Wentland <harry.wentland@amd.com>
Date: Tue, 29 May 2018 09:59:13 -0400
Subject: drm/amd/display: Fix wrong latency assignment for VEGA clock levels

Also drop wrong 10kHz comment

Fixes: drm/amd/display: Implement dm_pp_get_clock_levels_by_type_with_latency
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c
index ead3d21..d5e6b45 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c
@@ -256,9 +256,8 @@ static void pp_to_dc_clock_levels_with_latency(
 
 	for (i = 0; i < clk_level_info->num_levels; i++) {
 		DRM_DEBUG("DM_PPLIB:\t %d\n", pp_clks->data[i].clocks_in_khz);
-		/* translate 10kHz to kHz */
 		clk_level_info->data[i].clocks_in_khz = pp_clks->data[i].clocks_in_khz;
-		clk_level_info->data[i].latency_in_us = pp_clks->data[i].clocks_in_khz;
+		clk_level_info->data[i].latency_in_us = pp_clks->data[i].latency_in_us;
 	}
 }
 
-- 
cgit v1.1


From 7bee0572e31f1f5963ecc19d55f64f384d0b155b Mon Sep 17 00:00:00 2001
From: Mikita Lipski <mikita.lipski@amd.com>
Date: Mon, 28 May 2018 11:22:17 -0400
Subject: drm/amd/pp: Add cases for getting phys and disp clks for SMU10

Add case options to retrieve either physical or display clocks with
voltage from SMU controller that are needed by display driver.

Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
index 85f84f4..e160b05 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
@@ -1000,6 +1000,12 @@ static int smu10_get_clock_by_type_with_voltage(struct pp_hwmgr *hwmgr,
 	case amd_pp_soc_clock:
 		pclk_vol_table = pinfo->vdd_dep_on_socclk;
 		break;
+	case amd_pp_disp_clock:
+		pclk_vol_table = pinfo->vdd_dep_on_dispclk;
+		break;
+	case amd_pp_phy_clock:
+		pclk_vol_table = pinfo->vdd_dep_on_phyclk;
+		break;
 	default:
 		return -EINVAL;
 	}
-- 
cgit v1.1


From bda31a24dc5c03fd76832c4d672fba8355e3aa44 Mon Sep 17 00:00:00 2001
From: Deepak Sharma <Deepak.Sharma@amd.com>
Date: Tue, 22 May 2018 15:31:23 -0700
Subject: drm/amdgpu: Use GTT for dumb buffer if sg display enabled (v2)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

When vram size <= THRESHOLD(256M) lets use GTT for dumb buffer
allocation. As SG will be enabled with vram size <= 256M
scan out will not be an issue.

v2: Use amdgpu_display_supported_domains to get supported domain.

Signed-off-by: Deepak Sharma <Deepak.Sharma@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
index 2c8e273..63758db 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
@@ -30,6 +30,7 @@
 #include <drm/drmP.h>
 #include <drm/amdgpu_drm.h>
 #include "amdgpu.h"
+#include "amdgpu_display.h"
 
 void amdgpu_gem_object_free(struct drm_gem_object *gobj)
 {
@@ -749,15 +750,20 @@ int amdgpu_mode_dumb_create(struct drm_file *file_priv,
 	struct amdgpu_device *adev = dev->dev_private;
 	struct drm_gem_object *gobj;
 	uint32_t handle;
+	u32 domain = amdgpu_display_supported_domains(adev);
 	int r;
 
 	args->pitch = amdgpu_align_pitch(adev, args->width,
 					 DIV_ROUND_UP(args->bpp, 8), 0);
 	args->size = (u64)args->pitch * args->height;
 	args->size = ALIGN(args->size, PAGE_SIZE);
+	if (domain == (AMDGPU_GEM_DOMAIN_VRAM | AMDGPU_GEM_DOMAIN_GTT)) {
+		domain = AMDGPU_GEM_DOMAIN_VRAM;
+		if (adev->gmc.real_vram_size <= AMDGPU_SG_THRESHOLD)
+			domain = AMDGPU_GEM_DOMAIN_GTT;
+	}
 
-	r = amdgpu_gem_object_create(adev, args->size, 0,
-				     AMDGPU_GEM_DOMAIN_VRAM,
+	r = amdgpu_gem_object_create(adev, args->size, 0, domain,
 				     AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED,
 				     false, NULL, &gobj);
 	if (r)
-- 
cgit v1.1


From 84b74608442d00fbdcd233e3230b3068b0ab9b18 Mon Sep 17 00:00:00 2001
From: Deepak Sharma <Deepak.Sharma@amd.com>
Date: Fri, 25 May 2018 17:12:29 -0700
Subject: drm/amdgpu: Add helper function to get buffer domain
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Move logic of getting supported domain to a helper
function

Signed-off-by: Deepak Sharma <Deepak.Sharma@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c    | 10 +++-------
 drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 17 ++++++++++++-----
 drivers/gpu/drm/amd/amdgpu/amdgpu_object.h |  3 ++-
 3 files changed, 17 insertions(+), 13 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
index 63758db..556406a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
@@ -750,19 +750,15 @@ int amdgpu_mode_dumb_create(struct drm_file *file_priv,
 	struct amdgpu_device *adev = dev->dev_private;
 	struct drm_gem_object *gobj;
 	uint32_t handle;
-	u32 domain = amdgpu_display_supported_domains(adev);
+	u32 domain;
 	int r;
 
 	args->pitch = amdgpu_align_pitch(adev, args->width,
 					 DIV_ROUND_UP(args->bpp, 8), 0);
 	args->size = (u64)args->pitch * args->height;
 	args->size = ALIGN(args->size, PAGE_SIZE);
-	if (domain == (AMDGPU_GEM_DOMAIN_VRAM | AMDGPU_GEM_DOMAIN_GTT)) {
-		domain = AMDGPU_GEM_DOMAIN_VRAM;
-		if (adev->gmc.real_vram_size <= AMDGPU_SG_THRESHOLD)
-			domain = AMDGPU_GEM_DOMAIN_GTT;
-	}
-
+	domain = amdgpu_bo_get_preferred_pin_domain(adev,
+				amdgpu_display_supported_domains(adev));
 	r = amdgpu_gem_object_create(adev, args->size, 0, domain,
 				     AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED,
 				     false, NULL, &gobj);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
index 6a9e46a..5e4e1bd 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
@@ -703,11 +703,7 @@ int amdgpu_bo_pin_restricted(struct amdgpu_bo *bo, u32 domain,
 	/* This assumes only APU display buffers are pinned with (VRAM|GTT).
 	 * See function amdgpu_display_supported_domains()
 	 */
-	if (domain == (AMDGPU_GEM_DOMAIN_VRAM | AMDGPU_GEM_DOMAIN_GTT)) {
-		domain = AMDGPU_GEM_DOMAIN_VRAM;
-		if (adev->gmc.real_vram_size <= AMDGPU_SG_THRESHOLD)
-			domain = AMDGPU_GEM_DOMAIN_GTT;
-	}
+	domain = amdgpu_bo_get_preferred_pin_domain(adev, domain);
 
 	if (bo->pin_count) {
 		uint32_t mem_type = bo->tbo.mem.mem_type;
@@ -1066,3 +1062,14 @@ u64 amdgpu_bo_gpu_offset(struct amdgpu_bo *bo)
 
 	return bo->tbo.offset;
 }
+
+uint32_t amdgpu_bo_get_preferred_pin_domain(struct amdgpu_device *adev,
+					    uint32_t domain)
+{
+	if (domain == (AMDGPU_GEM_DOMAIN_VRAM | AMDGPU_GEM_DOMAIN_GTT)) {
+		domain = AMDGPU_GEM_DOMAIN_VRAM;
+		if (adev->gmc.real_vram_size <= AMDGPU_SG_THRESHOLD)
+			domain = AMDGPU_GEM_DOMAIN_GTT;
+	}
+	return domain;
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
index 540e03f..7317480 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
@@ -289,7 +289,8 @@ int amdgpu_bo_restore_from_shadow(struct amdgpu_device *adev,
 				  struct reservation_object *resv,
 				  struct dma_fence **fence,
 				  bool direct);
-
+uint32_t amdgpu_bo_get_preferred_pin_domain(struct amdgpu_device *adev,
+					    uint32_t domain);
 
 /*
  * sub allocation
-- 
cgit v1.1


From 4c6530fd66399182d0332c5ed821ea473bdcd7c3 Mon Sep 17 00:00:00 2001
From: Leo Liu <leo.liu@amd.com>
Date: Fri, 25 May 2018 10:53:39 -0400
Subject: drm/amdgpu: remove unnecessary scheduler entity for VCN
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

It should be stateless, and no need for scheduler to take care specially.

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c | 51 +++++++--------------------------
 drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h |  2 --
 2 files changed, 10 insertions(+), 43 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
index 8851bcd..6fd606f 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
@@ -49,8 +49,6 @@ static void amdgpu_vcn_idle_work_handler(struct work_struct *work);
 
 int amdgpu_vcn_sw_init(struct amdgpu_device *adev)
 {
-	struct amdgpu_ring *ring;
-	struct drm_sched_rq *rq;
 	unsigned long bo_size;
 	const char *fw_name;
 	const struct common_firmware_header *hdr;
@@ -102,24 +100,6 @@ int amdgpu_vcn_sw_init(struct amdgpu_device *adev)
 		return r;
 	}
 
-	ring = &adev->vcn.ring_dec;
-	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
-	r = drm_sched_entity_init(&ring->sched, &adev->vcn.entity_dec,
-				  rq, NULL);
-	if (r != 0) {
-		DRM_ERROR("Failed setting up VCN dec run queue.\n");
-		return r;
-	}
-
-	ring = &adev->vcn.ring_enc[0];
-	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
-	r = drm_sched_entity_init(&ring->sched, &adev->vcn.entity_enc,
-				  rq, NULL);
-	if (r != 0) {
-		DRM_ERROR("Failed setting up VCN enc run queue.\n");
-		return r;
-	}
-
 	return 0;
 }
 
@@ -129,10 +109,6 @@ int amdgpu_vcn_sw_fini(struct amdgpu_device *adev)
 
 	kfree(adev->vcn.saved_bo);
 
-	drm_sched_entity_fini(&adev->vcn.ring_dec.sched, &adev->vcn.entity_dec);
-
-	drm_sched_entity_fini(&adev->vcn.ring_enc[0].sched, &adev->vcn.entity_enc);
-
 	amdgpu_bo_free_kernel(&adev->vcn.vcpu_bo,
 			      &adev->vcn.gpu_addr,
 			      (void **)&adev->vcn.cpu_addr);
@@ -278,7 +254,7 @@ int amdgpu_vcn_dec_ring_test_ring(struct amdgpu_ring *ring)
 }
 
 static int amdgpu_vcn_dec_send_msg(struct amdgpu_ring *ring,
-				   struct amdgpu_bo *bo, bool direct,
+				   struct amdgpu_bo *bo,
 				   struct dma_fence **fence)
 {
 	struct amdgpu_device *adev = ring->adev;
@@ -306,19 +282,12 @@ static int amdgpu_vcn_dec_send_msg(struct amdgpu_ring *ring,
 	}
 	ib->length_dw = 16;
 
-	if (direct) {
-		r = amdgpu_ib_schedule(ring, 1, ib, NULL, &f);
-		job->fence = dma_fence_get(f);
-		if (r)
-			goto err_free;
+	r = amdgpu_ib_schedule(ring, 1, ib, NULL, &f);
+	job->fence = dma_fence_get(f);
+	if (r)
+		goto err_free;
 
-		amdgpu_job_free(job);
-	} else {
-		r = amdgpu_job_submit(job, ring, &adev->vcn.entity_dec,
-				      AMDGPU_FENCE_OWNER_UNDEFINED, &f);
-		if (r)
-			goto err_free;
-	}
+	amdgpu_job_free(job);
 
 	amdgpu_bo_fence(bo, f, false);
 	amdgpu_bo_unreserve(bo);
@@ -370,11 +339,11 @@ static int amdgpu_vcn_dec_get_create_msg(struct amdgpu_ring *ring, uint32_t hand
 	for (i = 14; i < 1024; ++i)
 		msg[i] = cpu_to_le32(0x0);
 
-	return amdgpu_vcn_dec_send_msg(ring, bo, true, fence);
+	return amdgpu_vcn_dec_send_msg(ring, bo, fence);
 }
 
 static int amdgpu_vcn_dec_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle,
-			       bool direct, struct dma_fence **fence)
+			       struct dma_fence **fence)
 {
 	struct amdgpu_device *adev = ring->adev;
 	struct amdgpu_bo *bo = NULL;
@@ -396,7 +365,7 @@ static int amdgpu_vcn_dec_get_destroy_msg(struct amdgpu_ring *ring, uint32_t han
 	for (i = 6; i < 1024; ++i)
 		msg[i] = cpu_to_le32(0x0);
 
-	return amdgpu_vcn_dec_send_msg(ring, bo, direct, fence);
+	return amdgpu_vcn_dec_send_msg(ring, bo, fence);
 }
 
 int amdgpu_vcn_dec_ring_test_ib(struct amdgpu_ring *ring, long timeout)
@@ -410,7 +379,7 @@ int amdgpu_vcn_dec_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 		goto error;
 	}
 
-	r = amdgpu_vcn_dec_get_destroy_msg(ring, 1, true, &fence);
+	r = amdgpu_vcn_dec_get_destroy_msg(ring, 1, &fence);
 	if (r) {
 		DRM_ERROR("amdgpu: failed to get destroy ib (%ld).\n", r);
 		goto error;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h
index 181e6af..773010b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h
@@ -67,8 +67,6 @@ struct amdgpu_vcn {
 	struct amdgpu_ring	ring_dec;
 	struct amdgpu_ring	ring_enc[AMDGPU_VCN_MAX_ENC_RINGS];
 	struct amdgpu_irq_src	irq;
-	struct drm_sched_entity entity_dec;
-	struct drm_sched_entity entity_enc;
 	unsigned		num_enc_rings;
 };
 
-- 
cgit v1.1


From e03fd3f300f6184c1264186a4c815e93bf658abb Mon Sep 17 00:00:00 2001
From: Mikita Lipski <mikita.lipski@amd.com>
Date: Wed, 16 May 2018 16:46:18 -0400
Subject: drm/amd/display: Do not limit color depth to 8bpc

Delete if statement that would force any display's color depth higher
than 8 bpc to 8

Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 6 ------
 1 file changed, 6 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 1ce10bc..52e57b5 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -2095,12 +2095,6 @@ convert_color_depth_from_display_info(const struct drm_connector *connector)
 {
 	uint32_t bpc = connector->display_info.bpc;
 
-	/* Limited color depth to 8bit
-	 * TODO: Still need to handle deep color
-	 */
-	if (bpc > 8)
-		bpc = 8;
-
 	switch (bpc) {
 	case 0:
 		/* Temporary Work around, DRM don't parse color depth for
-- 
cgit v1.1


From 2b6199a1d1b70fccd62aed961ba4c2b979ae499c Mon Sep 17 00:00:00 2001
From: Roman Li <Roman.Li@amd.com>
Date: Thu, 3 May 2018 13:29:42 -0400
Subject: drm/amd/display: replace msleep with udelay in fbc path

FBC enabling and disabling path has msleep which leads to
BUG hit when called in atomic context, hence this patch
replaces msleeps with udelays appropriately.

Signed-off-by: Shirish S <shirish.s@amd.com>
Signed-off-by: Roman Li <Roman.Li@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c
index 9150d26..e2994d3 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c
@@ -121,10 +121,10 @@ static void reset_lb_on_vblank(struct dc_context *ctx)
 		frame_count = dm_read_reg(ctx, mmCRTC_STATUS_FRAME_COUNT);
 
 
-		for (retry = 100; retry > 0; retry--) {
+		for (retry = 10000; retry > 0; retry--) {
 			if (frame_count != dm_read_reg(ctx, mmCRTC_STATUS_FRAME_COUNT))
 				break;
-			msleep(1);
+			udelay(10);
 		}
 		if (!retry)
 			dm_error("Frame count did not increase for 100ms.\n");
@@ -147,14 +147,14 @@ static void wait_for_fbc_state_changed(
 	uint32_t addr = mmFBC_STATUS;
 	uint32_t value;
 
-	while (counter < 10) {
+	while (counter < 1000) {
 		value = dm_read_reg(cp110->base.ctx, addr);
 		if (get_reg_field_value(
 			value,
 			FBC_STATUS,
 			FBC_ENABLE_STATUS) == enabled)
 			break;
-		msleep(10);
+		udelay(100);
 		counter++;
 	}
 
-- 
cgit v1.1


From aed15309b9b2009da9dfd209deaf9af8660f50a1 Mon Sep 17 00:00:00 2001
From: Mikita Lipski <mikita.lipski@amd.com>
Date: Tue, 1 May 2018 11:33:25 -0400
Subject: drm/amd/display: Release fake sink

If connector doesn't have a sink, fake sink is created, but
never released as it assumed that its destroyed with the
stream it is used for. But now sink is released before the
stream maintaing refcount consistency.

This way we also avoid assigning anything to connector keeping
all the operation local.

Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 31 ++++++++++++-----------
 1 file changed, 16 insertions(+), 15 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 52e57b5..e2bf4fe 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -2310,27 +2310,22 @@ decide_crtc_timing_for_drm_display_mode(struct drm_display_mode *drm_mode,
 	}
 }
 
-static int create_fake_sink(struct amdgpu_dm_connector *aconnector)
+static struct dc_sink *
+create_fake_sink(struct amdgpu_dm_connector *aconnector)
 {
-	struct dc_sink *sink = NULL;
 	struct dc_sink_init_data sink_init_data = { 0 };
-
+	struct dc_sink *sink = NULL;
 	sink_init_data.link = aconnector->dc_link;
 	sink_init_data.sink_signal = aconnector->dc_link->connector_signal;
 
 	sink = dc_sink_create(&sink_init_data);
 	if (!sink) {
 		DRM_ERROR("Failed to create sink!\n");
-		return -ENOMEM;
+		return NULL;
 	}
-
 	sink->sink_signal = SIGNAL_TYPE_VIRTUAL;
-	aconnector->fake_enable = true;
 
-	aconnector->dc_sink = sink;
-	aconnector->dc_link->local_sink = sink;
-
-	return 0;
+	return sink;
 }
 
 static void set_multisync_trigger_params(
@@ -2393,7 +2388,7 @@ create_stream_for_sink(struct amdgpu_dm_connector *aconnector,
 	struct dc_stream_state *stream = NULL;
 	struct drm_display_mode mode = *drm_mode;
 	bool native_mode_found = false;
-
+	struct dc_sink *sink = NULL;
 	if (aconnector == NULL) {
 		DRM_ERROR("aconnector is NULL!\n");
 		return stream;
@@ -2411,15 +2406,18 @@ create_stream_for_sink(struct amdgpu_dm_connector *aconnector,
 			return stream;
 		}
 
-		if (create_fake_sink(aconnector))
+		sink = create_fake_sink(aconnector);
+		if (!sink)
 			return stream;
+	} else {
+		sink = aconnector->dc_sink;
 	}
 
-	stream = dc_create_stream_for_sink(aconnector->dc_sink);
+	stream = dc_create_stream_for_sink(sink);
 
 	if (stream == NULL) {
 		DRM_ERROR("Failed to create stream for sink!\n");
-		return stream;
+		goto finish;
 	}
 
 	list_for_each_entry(preferred_mode, &aconnector->base.modes, head) {
@@ -2458,12 +2456,15 @@ create_stream_for_sink(struct amdgpu_dm_connector *aconnector,
 	fill_audio_info(
 		&stream->audio_info,
 		drm_connector,
-		aconnector->dc_sink);
+		sink);
 
 	update_stream_signal(stream);
 
 	if (dm_state && dm_state->freesync_capable)
 		stream->ignore_msa_timing_param = true;
+finish:
+	if (sink && sink->sink_signal == SIGNAL_TYPE_VIRTUAL)
+		dc_sink_release(sink);
 
 	return stream;
 }
-- 
cgit v1.1


From fb5fb63aa91aa7b353e3f7f0031299760f65ecf2 Mon Sep 17 00:00:00 2001
From: Charlene Liu <charlene.liu@amd.com>
Date: Tue, 1 May 2018 19:49:03 -0400
Subject: drm/amd/display: add register offset != 0 check.

Signed-off-by: Charlene Liu <charlene.liu@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c     | 3 ++-
 drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c | 2 +-
 2 files changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
index 0a6d483..c063175 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
@@ -72,7 +72,8 @@ static void dce110_update_generic_info_packet(
 	uint32_t max_retries = 50;
 
 	/*we need turn on clock before programming AFMT block*/
-	REG_UPDATE(AFMT_CNTL, AFMT_AUDIO_CLOCK_EN, 1);
+	if (REG(AFMT_CNTL))
+		REG_UPDATE(AFMT_CNTL, AFMT_AUDIO_CLOCK_EN, 1);
 
 	if (REG(AFMT_VBI_PACKET_CONTROL1)) {
 		if (packet_index >= 8)
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
index a92fb0a..c29052b 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
@@ -1004,9 +1004,9 @@ void dce110_disable_stream(struct pipe_ctx *pipe_ctx, int option)
 		/*don't free audio if it is from retrain or internal disable stream*/
 		if (option == FREE_ACQUIRED_RESOURCE && dc->caps.dynamic_audio == true) {
 			/*we have to dynamic arbitrate the audio endpoints*/
-			pipe_ctx->stream_res.audio = NULL;
 			/*we free the resource, need reset is_audio_acquired*/
 			update_audio_usage(&dc->current_state->res_ctx, dc->res_pool, pipe_ctx->stream_res.audio, false);
+			pipe_ctx->stream_res.audio = NULL;
 		}
 
 		/* TODO: notify audio driver for if audio modes list changed
-- 
cgit v1.1


From 9356badb2636b0afe2b34a8133ab246547cdf9ca Mon Sep 17 00:00:00 2001
From: Roman Li <Roman.Li@amd.com>
Date: Thu, 17 May 2018 18:08:54 -0400
Subject: drm/amd/display: check if audio clk enable is applicable

Fixing warning on dce10 with HDMI display.

Signed-off-by: Roman Li <Roman.Li@amd.com>
Reviewed-by: Charlene Liu <Charlene.Liu@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
index c063175..c0e813c 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
@@ -720,7 +720,8 @@ static void dce110_stream_encoder_update_hdmi_info_packets(
 			const uint32_t *content =
 				(const uint32_t *) &info_frame->avi.sb[0];
 			/*we need turn on clock before programming AFMT block*/
-			REG_UPDATE(AFMT_CNTL, AFMT_AUDIO_CLOCK_EN, 1);
+			if (REG(AFMT_CNTL))
+				REG_UPDATE(AFMT_CNTL, AFMT_AUDIO_CLOCK_EN, 1);
 
 			REG_WRITE(AFMT_AVI_INFO0, content[0]);
 
-- 
cgit v1.1


From 01884c02c46ad1f54cfc7eb43633fad199ab8007 Mon Sep 17 00:00:00 2001
From: Eric Bernstein <eric.bernstein@amd.com>
Date: Thu, 10 May 2018 15:12:09 -0400
Subject: drm/amd/display: DP component depth 16 bpc

Add register programming to support 16bpc component
depth for DP.

Signed-off-by: Eric Bernstein <eric.bernstein@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c
index 653b7b2..c928ee4 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c
@@ -319,6 +319,10 @@ void enc1_stream_encoder_dp_set_stream_attribute(
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH,
 				DP_COMPONENT_PIXEL_DEPTH_12BPC);
 		break;
+	case COLOR_DEPTH_161616:
+		REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH,
+				DP_COMPONENT_PIXEL_DEPTH_16BPC);
+		break;
 	default:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH,
 				DP_COMPONENT_PIXEL_DEPTH_6BPC);
-- 
cgit v1.1


From aa6d4a59d686d3172012c19874add120e02c113f Mon Sep 17 00:00:00 2001
From: Eric Bernstein <eric.bernstein@amd.com>
Date: Mon, 14 May 2018 16:55:07 -0400
Subject: drm/amd/display: Set TMZ and DCC for secondary surface

Add register programming to support TMZ and DCC on
secondary surfaces.

Signed-off-by: Eric Bernstein <eric.bernstein@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c | 14 ++++++++++----
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h |  8 ++++++++
 2 files changed, 18 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
index d2ab78b..c28085b 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
@@ -396,11 +396,15 @@ bool hubp1_program_surface_flip_and_addr(
 		if (address->grph_stereo.right_addr.quad_part == 0)
 			break;
 
-		REG_UPDATE_4(DCSURF_SURFACE_CONTROL,
+		REG_UPDATE_8(DCSURF_SURFACE_CONTROL,
 				PRIMARY_SURFACE_TMZ, address->tmz_surface,
 				PRIMARY_SURFACE_TMZ_C, address->tmz_surface,
 				PRIMARY_META_SURFACE_TMZ, address->tmz_surface,
-				PRIMARY_META_SURFACE_TMZ_C, address->tmz_surface);
+				PRIMARY_META_SURFACE_TMZ_C, address->tmz_surface,
+				SECONDARY_SURFACE_TMZ, address->tmz_surface,
+				SECONDARY_SURFACE_TMZ_C, address->tmz_surface,
+				SECONDARY_META_SURFACE_TMZ, address->tmz_surface,
+				SECONDARY_META_SURFACE_TMZ_C, address->tmz_surface);
 
 		if (address->grph_stereo.right_meta_addr.quad_part != 0) {
 
@@ -459,9 +463,11 @@ void hubp1_dcc_control(struct hubp *hubp, bool enable,
 	uint32_t dcc_ind_64b_blk = independent_64b_blks ? 1 : 0;
 	struct dcn10_hubp *hubp1 = TO_DCN10_HUBP(hubp);
 
-	REG_UPDATE_2(DCSURF_SURFACE_CONTROL,
+	REG_UPDATE_4(DCSURF_SURFACE_CONTROL,
 			PRIMARY_SURFACE_DCC_EN, dcc_en,
-			PRIMARY_SURFACE_DCC_IND_64B_BLK, dcc_ind_64b_blk);
+			PRIMARY_SURFACE_DCC_IND_64B_BLK, dcc_ind_64b_blk,
+			SECONDARY_SURFACE_DCC_EN, dcc_en,
+			SECONDARY_SURFACE_DCC_IND_64B_BLK, dcc_ind_64b_blk);
 }
 
 void hubp1_program_surface_config(
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
index af38403..d901d50 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
@@ -312,6 +312,12 @@
 	HUBP_SF(HUBPREQ0_DCSURF_SURFACE_CONTROL, PRIMARY_META_SURFACE_TMZ_C, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCSURF_SURFACE_CONTROL, PRIMARY_SURFACE_DCC_EN, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCSURF_SURFACE_CONTROL, PRIMARY_SURFACE_DCC_IND_64B_BLK, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCSURF_SURFACE_CONTROL, SECONDARY_SURFACE_TMZ, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCSURF_SURFACE_CONTROL, SECONDARY_SURFACE_TMZ_C, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCSURF_SURFACE_CONTROL, SECONDARY_META_SURFACE_TMZ, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCSURF_SURFACE_CONTROL, SECONDARY_META_SURFACE_TMZ_C, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCSURF_SURFACE_CONTROL, SECONDARY_SURFACE_DCC_EN, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCSURF_SURFACE_CONTROL, SECONDARY_SURFACE_DCC_IND_64B_BLK, mask_sh),\
 	HUBP_SF(HUBPRET0_HUBPRET_CONTROL, DET_BUF_PLANE1_BASE_ADDRESS, mask_sh),\
 	HUBP_SF(HUBPRET0_HUBPRET_CONTROL, CROSSBAR_SRC_CB_B, mask_sh),\
 	HUBP_SF(HUBPRET0_HUBPRET_CONTROL, CROSSBAR_SRC_CR_R, mask_sh),\
@@ -489,6 +495,8 @@
 	type SECONDARY_META_SURFACE_TMZ_C;\
 	type PRIMARY_SURFACE_DCC_EN;\
 	type PRIMARY_SURFACE_DCC_IND_64B_BLK;\
+	type SECONDARY_SURFACE_DCC_EN;\
+	type SECONDARY_SURFACE_DCC_IND_64B_BLK;\
 	type DET_BUF_PLANE1_BASE_ADDRESS;\
 	type CROSSBAR_SRC_CB_B;\
 	type CROSSBAR_SRC_CR_R;\
-- 
cgit v1.1


From df099b9b60c6c378b1cd2024d1a21a87459b6614 Mon Sep 17 00:00:00 2001
From: "Leo (Sunpeng) Li" <sunpeng.li@amd.com>
Date: Wed, 16 May 2018 10:31:30 -0400
Subject: drm/amd/display: Destroy connector state on reset

When a DRM mode reset is called on resume, the connector state's
destructor is not called. This leaves a dangling reference on the CRTC
commit object, which was obtained by the connector state during commit
setup.

Signed-off-by: Leo (Sunpeng) Li <sunpeng.li@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index e2bf4fe..d913c0a 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -2709,6 +2709,9 @@ void amdgpu_dm_connector_funcs_reset(struct drm_connector *connector)
 	struct dm_connector_state *state =
 		to_dm_connector_state(connector->state);
 
+	if (connector->state)
+		__drm_atomic_helper_connector_destroy_state(connector->state);
+
 	kfree(state);
 
 	state = kzalloc(sizeof(*state), GFP_KERNEL);
@@ -2719,8 +2722,7 @@ void amdgpu_dm_connector_funcs_reset(struct drm_connector *connector)
 		state->underscan_hborder = 0;
 		state->underscan_vborder = 0;
 
-		connector->state = &state->base;
-		connector->state->connector = connector;
+		__drm_atomic_helper_connector_reset(connector, &state->base);
 	}
 }
 
-- 
cgit v1.1


From 0b19fdc45feffd7569c081fe32a258df3c8ebb9b Mon Sep 17 00:00:00 2001
From: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Date: Thu, 17 May 2018 10:08:10 -0400
Subject: drm/amd/display: fix dscl_manual_ratio_init

This change will fix wb and display scaling when ratios of
4 or more are involved

Signed-off-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c    | 5 +++++
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c      | 3 +--
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h      | 6 +-----
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c | 8 ++++----
 drivers/gpu/drm/amd/display/include/fixed31_32.h      | 2 ++
 5 files changed, 13 insertions(+), 11 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c b/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c
index e61dd97d..f289898 100644
--- a/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c
+++ b/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c
@@ -449,6 +449,11 @@ static inline unsigned int clamp_ux_dy(
 		return min_clamp;
 }
 
+unsigned int dc_fixpt_u3d19(struct fixed31_32 arg)
+{
+	return ux_dy(arg.value, 3, 19);
+}
+
 unsigned int dc_fixpt_u2d19(struct fixed31_32 arg)
 {
 	return ux_dy(arg.value, 2, 19);
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
index 46a35c7..c69fa4b 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
@@ -132,8 +132,7 @@ void dpp_set_gamut_remap_bypass(struct dcn10_dpp *dpp)
 
 #define IDENTITY_RATIO(ratio) (dc_fixpt_u2d19(ratio) == (1 << 19))
 
-
-bool dpp_get_optimal_number_of_taps(
+static bool dpp_get_optimal_number_of_taps(
 		struct dpp *dpp,
 		struct scaler_data *scl_data,
 		const struct scaling_taps *in_taps)
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h
index 5944a3b..e862caf 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h
@@ -1424,12 +1424,8 @@ void dpp1_set_degamma(
 		enum ipp_degamma_mode mode);
 
 void dpp1_set_degamma_pwl(struct dpp *dpp_base,
-								 const struct pwl_params *params);
+		const struct pwl_params *params);
 
-bool dpp_get_optimal_number_of_taps(
-		struct dpp *dpp,
-		struct scaler_data *scl_data,
-		const struct scaling_taps *in_taps);
 
 void dpp_read_state(struct dpp *dpp_base,
 		struct dcn_dpp_state *s);
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c
index 4ddd627..f862fd1 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c
@@ -565,16 +565,16 @@ static void dpp1_dscl_set_manual_ratio_init(
 	uint32_t init_int = 0;
 
 	REG_SET(SCL_HORZ_FILTER_SCALE_RATIO, 0,
-			SCL_H_SCALE_RATIO, dc_fixpt_u2d19(data->ratios.horz) << 5);
+			SCL_H_SCALE_RATIO, dc_fixpt_u3d19(data->ratios.horz) << 5);
 
 	REG_SET(SCL_VERT_FILTER_SCALE_RATIO, 0,
-			SCL_V_SCALE_RATIO, dc_fixpt_u2d19(data->ratios.vert) << 5);
+			SCL_V_SCALE_RATIO, dc_fixpt_u3d19(data->ratios.vert) << 5);
 
 	REG_SET(SCL_HORZ_FILTER_SCALE_RATIO_C, 0,
-			SCL_H_SCALE_RATIO_C, dc_fixpt_u2d19(data->ratios.horz_c) << 5);
+			SCL_H_SCALE_RATIO_C, dc_fixpt_u3d19(data->ratios.horz_c) << 5);
 
 	REG_SET(SCL_VERT_FILTER_SCALE_RATIO_C, 0,
-			SCL_V_SCALE_RATIO_C, dc_fixpt_u2d19(data->ratios.vert_c) << 5);
+			SCL_V_SCALE_RATIO_C, dc_fixpt_u3d19(data->ratios.vert_c) << 5);
 
 	/*
 	 * 0.24 format for fraction, first five bits zeroed
diff --git a/drivers/gpu/drm/amd/display/include/fixed31_32.h b/drivers/gpu/drm/amd/display/include/fixed31_32.h
index bb0d4eb..a981b3e 100644
--- a/drivers/gpu/drm/amd/display/include/fixed31_32.h
+++ b/drivers/gpu/drm/amd/display/include/fixed31_32.h
@@ -496,6 +496,8 @@ static inline int dc_fixpt_ceil(struct fixed31_32 arg)
  * fractional
  */
 
+unsigned int dc_fixpt_u3d19(struct fixed31_32 arg);
+
 unsigned int dc_fixpt_u2d19(struct fixed31_32 arg);
 
 unsigned int dc_fixpt_u0d19(struct fixed31_32 arg);
-- 
cgit v1.1


From 4ea7fc09539bd2399c1fa7acea14529406120d9e Mon Sep 17 00:00:00 2001
From: Mikita Lipski <mikita.lipski@amd.com>
Date: Thu, 3 May 2018 17:08:51 -0400
Subject: drm/amd/display: Do not program interrupt status on disabled crtc

Prevent interrupt programming of a crtc on which the stream is disabled and
it doesn't have an OTG to reference.

Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c
index 4be21bf..a910f01 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c
@@ -555,6 +555,9 @@ static inline int dm_irq_state(struct amdgpu_device *adev,
 		return 0;
 	}
 
+	if (acrtc->otg_inst == -1)
+		return 0;
+
 	irq_source = dal_irq_type + acrtc->otg_inst;
 
 	st = (state == AMDGPU_IRQ_STATE_ENABLE);
-- 
cgit v1.1


From 794550c6eaf791bfd2e8d70e11aa56fdd6361725 Mon Sep 17 00:00:00 2001
From: Nikola Cornij <nikola.cornij@amd.com>
Date: Mon, 7 May 2018 15:35:15 -0400
Subject: drm/amd/display: Read DPCD link caps up to and including
 DP_ADAPTER_CAP

DP 1.4 compliance requires 16 bytes to be read when reading link caps,
i.e. it requires DP_ADAPTER_CAP to be included. Included it for all DP
versions because reading more than required won't fail.

Signed-off-by: Nikola Cornij <nikola.cornij@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
index 7d609c7..3fcb67c 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
@@ -2278,7 +2278,7 @@ static void dp_wa_power_up_0010FA(struct dc_link *link, uint8_t *dpcd_data,
 
 static bool retrieve_link_cap(struct dc_link *link)
 {
-	uint8_t dpcd_data[DP_TRAINING_AUX_RD_INTERVAL - DP_DPCD_REV + 1];
+	uint8_t dpcd_data[DP_ADAPTER_CAP - DP_DPCD_REV + 1];
 
 	union down_stream_port_count down_strm_port_count;
 	union edp_configuration_cap edp_config_cap;
-- 
cgit v1.1


From c733e40c74457ad6aa56cc8b3318e829b8274bef Mon Sep 17 00:00:00 2001
From: Nikola Cornij <nikola.cornij@amd.com>
Date: Wed, 9 May 2018 13:11:35 -0400
Subject: drm/amd/display: Read DP_SINK_COUNT_ESI range on HPD for DP 1.4

DP 1.4 compliance now requires that registers at DP_SINK_COUNT_ESI range
(0x2002-0x2003, 0x200c-0x200f) are read instead of DP_SINK_COUNT range
(0x200-0x2005.

Signed-off-by: Nikola Cornij <nikola.cornij@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c | 35 ++++++++++++++++++++----
 1 file changed, 30 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
index 3fcb67c..7857cb4 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
@@ -1630,17 +1630,42 @@ static enum dc_status read_hpd_rx_irq_data(
 	struct dc_link *link,
 	union hpd_irq_data *irq_data)
 {
+	static enum dc_status retval;
+
 	/* The HW reads 16 bytes from 200h on HPD,
 	 * but if we get an AUX_DEFER, the HW cannot retry
 	 * and this causes the CTS tests 4.3.2.1 - 3.2.4 to
 	 * fail, so we now explicitly read 6 bytes which is
 	 * the req from the above mentioned test cases.
+	 *
+	 * For DP 1.4 we need to read those from 2002h range.
 	 */
-	return core_link_read_dpcd(
-	link,
-	DP_SINK_COUNT,
-	irq_data->raw,
-	sizeof(union hpd_irq_data));
+	if (link->dpcd_caps.dpcd_rev.raw < DPCD_REV_14)
+		retval = core_link_read_dpcd(
+			link,
+			DP_SINK_COUNT,
+			irq_data->raw,
+			sizeof(union hpd_irq_data));
+	else {
+		/* Read 2 bytes at this location,... */
+		retval = core_link_read_dpcd(
+			link,
+			DP_SINK_COUNT_ESI,
+			irq_data->raw,
+			2);
+
+		if (retval != DC_OK)
+			return retval;
+
+		/* ... then read remaining 4 at the other location */
+		retval = core_link_read_dpcd(
+			link,
+			DP_LANE0_1_STATUS_ESI,
+			&irq_data->raw[2],
+			4);
+	}
+
+	return retval;
 }
 
 static bool allow_hpd_rx_irq(const struct dc_link *link)
-- 
cgit v1.1


From 20d4ac659c76034586a3ab79489b0940631a65de Mon Sep 17 00:00:00 2001
From: "Leo (Sunpeng) Li" <sunpeng.li@amd.com>
Date: Tue, 29 May 2018 09:51:51 -0400
Subject: drm/amd/display: Fix BUG_ON during CRTC atomic check update
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

For cases where the CRTC is inactive (DPMS off), where a modeset is not
required, yet the CRTC is still in the atomic state, we should not
attempt to update anything on it.

Previously, we were relying on the modereset_required() helper to check
the above condition. However, the function returns false immediately if
a modeset is not required, ignoring the CRTC's enable/active state
flags. The correct way to filter is by looking at these flags instead.

Fixes: e277adc5a06c "drm/amd/display: Hookup color management functions"
Bugzilla: https://bugs.freedesktop.org/106194

Signed-off-by: Leo (Sunpeng) Li <sunpeng.li@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Tested-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 13 +++++++------
 1 file changed, 7 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index d913c0a..0a06941 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -4765,15 +4765,16 @@ next_crtc:
 		 * We want to do dc stream updates that do not require a
 		 * full modeset below.
 		 */
-		if (!enable || !aconnector || modereset_required(new_crtc_state))
+		if (!(enable && aconnector && new_crtc_state->enable &&
+		      new_crtc_state->active))
 			continue;
 		/*
 		 * Given above conditions, the dc state cannot be NULL because:
-		 * 1. We're attempting to enable a CRTC. Which has a...
-		 * 2. Valid connector attached, and
-		 * 3. User does not want to reset it (disable or mark inactive,
-		 *    which can happen on a CRTC that's already disabled).
-		 * => It currently exists.
+		 * 1. We're in the process of enabling CRTCs (just been added
+		 *    to the dc context, or already is on the context)
+		 * 2. Has a valid connector attached, and
+		 * 3. Is currently active and enabled.
+		 * => The dc stream state currently exists.
 		 */
 		BUG_ON(dm_new_crtc_state->stream == NULL);
 
-- 
cgit v1.1


From 01d98506eca24f1a6b67374ffb65c2e62a56e8e7 Mon Sep 17 00:00:00 2001
From: Emily Deng <Emily.Deng@amd.com>
Date: Wed, 30 May 2018 10:04:25 +0800
Subject: drm/amdgpu: To get gds, gws and oa from adev->gds (v2)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

As now enabled per vm bo feature, the user mode driver won't supply the
bo_list generally, for this case, the gdb_base, gds_size, gws_base, gws_size and
oa_base, oa_size won't be set.

v2: fix warning (Chunming)

Signed-off-by: Emily Deng <Emily.Deng@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 38 ++++++++++++++++++++--------------
 1 file changed, 23 insertions(+), 15 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
index 9c1d491..82312a7 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
@@ -522,6 +522,9 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p,
 	struct amdgpu_bo_list_entry *e;
 	struct list_head duplicates;
 	unsigned i, tries = 10;
+	struct amdgpu_bo *gds;
+	struct amdgpu_bo *gws;
+	struct amdgpu_bo *oa;
 	int r;
 
 	INIT_LIST_HEAD(&p->validated);
@@ -652,31 +655,36 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p,
 
 	amdgpu_cs_report_moved_bytes(p->adev, p->bytes_moved,
 				     p->bytes_moved_vis);
+
 	if (p->bo_list) {
-		struct amdgpu_bo *gds = p->bo_list->gds_obj;
-		struct amdgpu_bo *gws = p->bo_list->gws_obj;
-		struct amdgpu_bo *oa = p->bo_list->oa_obj;
 		struct amdgpu_vm *vm = &fpriv->vm;
 		unsigned i;
 
+		gds = p->bo_list->gds_obj;
+		gws = p->bo_list->gws_obj;
+		oa = p->bo_list->oa_obj;
 		for (i = 0; i < p->bo_list->num_entries; i++) {
 			struct amdgpu_bo *bo = p->bo_list->array[i].robj;
 
 			p->bo_list->array[i].bo_va = amdgpu_vm_bo_find(vm, bo);
 		}
+	} else {
+		gds = p->adev->gds.gds_gfx_bo;
+		gws = p->adev->gds.gws_gfx_bo;
+		oa = p->adev->gds.oa_gfx_bo;
+	}
 
-		if (gds) {
-			p->job->gds_base = amdgpu_bo_gpu_offset(gds);
-			p->job->gds_size = amdgpu_bo_size(gds);
-		}
-		if (gws) {
-			p->job->gws_base = amdgpu_bo_gpu_offset(gws);
-			p->job->gws_size = amdgpu_bo_size(gws);
-		}
-		if (oa) {
-			p->job->oa_base = amdgpu_bo_gpu_offset(oa);
-			p->job->oa_size = amdgpu_bo_size(oa);
-		}
+	if (gds) {
+		p->job->gds_base = amdgpu_bo_gpu_offset(gds);
+		p->job->gds_size = amdgpu_bo_size(gds);
+	}
+	if (gws) {
+		p->job->gws_base = amdgpu_bo_gpu_offset(gws);
+		p->job->gws_size = amdgpu_bo_size(gws);
+	}
+	if (oa) {
+		p->job->oa_base = amdgpu_bo_gpu_offset(oa);
+		p->job->oa_size = amdgpu_bo_size(oa);
 	}
 
 	if (!r && p->uf_entry.robj) {
-- 
cgit v1.1


From ee5309d5f3eba16d7901d29179d03d4336319fc0 Mon Sep 17 00:00:00 2001
From: Chunming Zhou <david1.zhou@amd.com>
Date: Wed, 30 May 2018 11:12:08 +0800
Subject: drm/amdgpu: gds bo must not be per-vm-bo
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

In per-vm-bo case, there could be no bo list.
But gds bo created from user space  must be passed to bo list.
So adding a check to prevent to creat gds bo as per-vm-bo.

Signed-off-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
index 556406a..5fb156a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
@@ -236,6 +236,13 @@ int amdgpu_gem_create_ioctl(struct drm_device *dev, void *data,
 	/* create a gem object to contain this object in */
 	if (args->in.domains & (AMDGPU_GEM_DOMAIN_GDS |
 	    AMDGPU_GEM_DOMAIN_GWS | AMDGPU_GEM_DOMAIN_OA)) {
+		if (flags & AMDGPU_GEM_CREATE_VM_ALWAYS_VALID) {
+			/* if gds bo is created from user space, it must be
+			 * passed to bo list
+			 */
+			DRM_ERROR("GDS bo cannot be per-vm-bo\n");
+			return -EINVAL;
+		}
 		flags |= AMDGPU_GEM_CREATE_NO_CPU_ACCESS;
 		if (args->in.domains == AMDGPU_GEM_DOMAIN_GDS)
 			size = size << AMDGPU_GDS_SHIFT;
-- 
cgit v1.1


From ee7a99c79aa3b15e9b6157f8949a1ad8c170f17f Mon Sep 17 00:00:00 2001
From: Evan Quan <evan.quan@amd.com>
Date: Mon, 28 May 2018 08:53:03 +0800
Subject: drm/amdgpu: correct SMU11 SYSPLL0 clock id values

The SMU11 SYSPLL0 clock ids were assigned wrong values.

Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/include/atomfirmware.h | 12 +++++-------
 1 file changed, 5 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/amd/include/atomfirmware.h b/drivers/gpu/drm/amd/include/atomfirmware.h
index c6c1666..092d800 100644
--- a/drivers/gpu/drm/amd/include/atomfirmware.h
+++ b/drivers/gpu/drm/amd/include/atomfirmware.h
@@ -2026,17 +2026,15 @@ enum atom_smu11_syspll_id {
   SMU11_SYSPLL3_1_ID          = 6,
 };
 
-
 enum atom_smu11_syspll0_clock_id {
-  SMU11_SYSPLL0_SOCCLK_ID   = 0,       //	SOCCLK
-  SMU11_SYSPLL0_MP0CLK_ID   = 1,       //	MP0CLK
-  SMU11_SYSPLL0_DCLK_ID     = 2,       //	DCLK
-  SMU11_SYSPLL0_VCLK_ID     = 3,       //	VCLK
-  SMU11_SYSPLL0_ECLK_ID     = 4,       //	ECLK
+  SMU11_SYSPLL0_ECLK_ID     = 0,       //	ECLK
+  SMU11_SYSPLL0_SOCCLK_ID   = 1,       //	SOCCLK
+  SMU11_SYSPLL0_MP0CLK_ID   = 2,       //	MP0CLK
+  SMU11_SYSPLL0_DCLK_ID     = 3,       //	DCLK
+  SMU11_SYSPLL0_VCLK_ID     = 4,       //	VCLK
   SMU11_SYSPLL0_DCEFCLK_ID  = 5,       //	DCEFCLK
 };
 
-
 enum atom_smu11_syspll1_0_clock_id {
   SMU11_SYSPLL1_0_UCLKA_ID   = 0,       // UCLK_a
 };
-- 
cgit v1.1


From 2f6a18ebd08e3f41a8e01176821b7203232298c4 Mon Sep 17 00:00:00 2001
From: Evan Quan <evan.quan@amd.com>
Date: Mon, 28 May 2018 08:55:09 +0800
Subject: drm/amd/powerplay: bug fixs for getsmuclockinfo

The .syspll_id and .dfsdid are not initialzed correctly. And
le32_to_cpu transfer is needed on the output.

Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c
index c97b0e5..5325661 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c
@@ -496,7 +496,9 @@ int pp_atomfwctrl_get_clk_information_by_clkid(struct pp_hwmgr *hwmgr, BIOS_CLKI
 	uint32_t ix;
 
 	parameters.clk_id = id;
+	parameters.syspll_id = 0;
 	parameters.command = GET_SMU_CLOCK_INFO_V3_1_GET_CLOCK_FREQ;
+	parameters.dfsdid = 0;
 
 	ix = GetIndexIntoMasterCmdTable(getsmuclockinfo);
 
@@ -505,7 +507,7 @@ int pp_atomfwctrl_get_clk_information_by_clkid(struct pp_hwmgr *hwmgr, BIOS_CLKI
 		return -EINVAL;
 
 	output = (struct atom_get_smu_clock_info_output_parameters_v3_1 *)&parameters;
-	*frequency = output->atom_smu_outputclkfreq.smu_clock_freq_hz / 10000;
+	*frequency = le32_to_cpu(output->atom_smu_outputclkfreq.smu_clock_freq_hz) / 10000;
 
 	return 0;
 }
-- 
cgit v1.1


From 102e494001c70901a1de212469d3c8d48dbb301a Mon Sep 17 00:00:00 2001
From: Evan Quan <evan.quan@amd.com>
Date: Mon, 28 May 2018 09:22:09 +0800
Subject: drm/amdgpu: typo fix for vega20 cg flags

The AMD_CG_SUPPORT_HDP_LS was wrongly written as
AMD_CG_SUPPORT_BIF_LS.

Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/soc15.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index 68b4a22..83f2717 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -685,6 +685,7 @@ static int soc15_common_early_init(void *handle)
 			AMD_CG_SUPPORT_BIF_MGCG |
 			AMD_CG_SUPPORT_BIF_LS |
 			AMD_CG_SUPPORT_HDP_MGCG |
+			AMD_CG_SUPPORT_HDP_LS |
 			AMD_CG_SUPPORT_ROM_MGCG |
 			AMD_CG_SUPPORT_VCE_MGCG |
 			AMD_CG_SUPPORT_UVD_MGCG;
-- 
cgit v1.1


From 762f52e9e4d237d8d378b5bb495f64073a9ba481 Mon Sep 17 00:00:00 2001
From: Mikita Lipski <mikita.lipski@amd.com>
Date: Tue, 29 May 2018 17:44:36 -0400
Subject: drm/amd/pp: Connect display_clock_voltage_request to a function
 pointer

Get rid of an empty dublicate of smu10_display_clock_voltage_request

Add display_clock_voltage_request to smu10 functions struct so it
can be called from outside the class and connect the pointer to
the function.

That way Display driver can finally apply clock voltage requests
when needed.

Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c | 64 +++++++++++------------
 1 file changed, 31 insertions(+), 33 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
index e160b05..6a63671 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
@@ -53,8 +53,37 @@ static const unsigned long SMU10_Magic = (unsigned long) PHM_Rv_Magic;
 
 
 static int smu10_display_clock_voltage_request(struct pp_hwmgr *hwmgr,
-		struct pp_display_clock_request *clock_req);
+		struct pp_display_clock_request *clock_req)
+{
+	struct smu10_hwmgr *smu10_data = (struct smu10_hwmgr *)(hwmgr->backend);
+	enum amd_pp_clock_type clk_type = clock_req->clock_type;
+	uint32_t clk_freq = clock_req->clock_freq_in_khz / 1000;
+	PPSMC_Msg        msg;
 
+	switch (clk_type) {
+	case amd_pp_dcf_clock:
+		if (clk_freq == smu10_data->dcf_actual_hard_min_freq)
+			return 0;
+		msg =  PPSMC_MSG_SetHardMinDcefclkByFreq;
+		smu10_data->dcf_actual_hard_min_freq = clk_freq;
+		break;
+	case amd_pp_soc_clock:
+		 msg = PPSMC_MSG_SetHardMinSocclkByFreq;
+		break;
+	case amd_pp_f_clock:
+		if (clk_freq == smu10_data->f_actual_hard_min_freq)
+			return 0;
+		smu10_data->f_actual_hard_min_freq = clk_freq;
+		msg = PPSMC_MSG_SetHardMinFclkByFreq;
+		break;
+	default:
+		pr_info("[DisplayClockVoltageRequest]Invalid Clock Type!");
+		return -EINVAL;
+	}
+	smum_send_msg_to_smc_with_parameter(hwmgr, msg, clk_freq);
+
+	return 0;
+}
 
 static struct smu10_power_state *cast_smu10_ps(struct pp_hw_power_state *hw_ps)
 {
@@ -1023,39 +1052,7 @@ static int smu10_get_clock_by_type_with_voltage(struct pp_hwmgr *hwmgr,
 	return 0;
 }
 
-static int smu10_display_clock_voltage_request(struct pp_hwmgr *hwmgr,
-		struct pp_display_clock_request *clock_req)
-{
-	struct smu10_hwmgr *smu10_data = (struct smu10_hwmgr *)(hwmgr->backend);
-	enum amd_pp_clock_type clk_type = clock_req->clock_type;
-	uint32_t clk_freq = clock_req->clock_freq_in_khz / 1000;
-	PPSMC_Msg        msg;
-
-	switch (clk_type) {
-	case amd_pp_dcf_clock:
-		if (clk_freq == smu10_data->dcf_actual_hard_min_freq)
-			return 0;
-		msg =  PPSMC_MSG_SetHardMinDcefclkByFreq;
-		smu10_data->dcf_actual_hard_min_freq = clk_freq;
-		break;
-	case amd_pp_soc_clock:
-		 msg = PPSMC_MSG_SetHardMinSocclkByFreq;
-		break;
-	case amd_pp_f_clock:
-		if (clk_freq == smu10_data->f_actual_hard_min_freq)
-			return 0;
-		smu10_data->f_actual_hard_min_freq = clk_freq;
-		msg = PPSMC_MSG_SetHardMinFclkByFreq;
-		break;
-	default:
-		pr_info("[DisplayClockVoltageRequest]Invalid Clock Type!");
-		return -EINVAL;
-	}
 
-	smum_send_msg_to_smc_with_parameter(hwmgr, msg, clk_freq);
-
-	return 0;
-}
 
 static int smu10_get_max_high_clocks(struct pp_hwmgr *hwmgr, struct amd_pp_simple_clock_info *clocks)
 {
@@ -1188,6 +1185,7 @@ static const struct pp_hwmgr_func smu10_hwmgr_funcs = {
 	.set_mmhub_powergating_by_smu = smu10_set_mmhub_powergating_by_smu,
 	.smus_notify_pwe = smu10_smus_notify_pwe,
 	.gfx_off_control = smu10_gfx_off_control,
+	.display_clock_voltage_request = smu10_display_clock_voltage_request,
 };
 
 int smu10_init_function_pointers(struct pp_hwmgr *hwmgr)
-- 
cgit v1.1


From 9dac0c3fb41056ae48b93e679c2a796c4dcfa8ed Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Wed, 30 May 2018 09:34:23 -0500
Subject: drm/amdgpu/display: check if ppfuncs exists before using it

Fixes a crash on asics without powerplay yet (e.g., vega20).

Reviewed-by: Rex Zhu<rezhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c
index d5e6b45..5a33461 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c
@@ -343,7 +343,7 @@ bool dm_pp_get_clock_levels_by_type_with_latency(
 	struct pp_clock_levels_with_latency pp_clks = { 0 };
 	const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs;
 
-	if (!pp_funcs->get_clock_by_type_with_latency)
+	if (!pp_funcs || !pp_funcs->get_clock_by_type_with_latency)
 		return false;
 
 	if (pp_funcs->get_clock_by_type_with_latency(pp_handle,
-- 
cgit v1.1


From f619794d716c0615241b73a0a5e78dbc8442de7b Mon Sep 17 00:00:00 2001
From: Eric Anholt <eric@anholt.net>
Date: Mon, 9 Apr 2018 13:58:13 -0700
Subject: drm/vc4: Fix leak of the file_priv that stored the perfmon.

Signed-off-by: Eric Anholt <eric@anholt.net>
Fixes: 65101d8c9108 ("drm/vc4: Expose performance counters to userspace")
Link: https://patchwork.freedesktop.org/patch/msgid/20180409205813.7077-1-eric@anholt.net
Reviewed-by: Boris Brezillon <boris.brezillon@bootlin.com>
---
 drivers/gpu/drm/vc4/vc4_drv.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/vc4/vc4_drv.c b/drivers/gpu/drm/vc4/vc4_drv.c
index d9b8b70..466d0a2 100644
--- a/drivers/gpu/drm/vc4/vc4_drv.c
+++ b/drivers/gpu/drm/vc4/vc4_drv.c
@@ -130,6 +130,7 @@ static void vc4_close(struct drm_device *dev, struct drm_file *file)
 	struct vc4_file *vc4file = file->driver_priv;
 
 	vc4_perfmon_close_file(vc4file);
+	kfree(vc4file);
 }
 
 static const struct vm_operations_struct vc4_vm_ops = {
-- 
cgit v1.1


From fbecef131676c1d18e8e6b42c04e10dc49725e96 Mon Sep 17 00:00:00 2001
From: Arnd Bergmann <arnd@arndb.de>
Date: Fri, 25 May 2018 17:50:12 +0200
Subject: drm/v3d: add CONFIG_MMU dependency

Without CONFIG_MMU, we get a link error:

drivers/gpu/drm/v3d/v3d_bo.o: In function `v3d_gem_fault':
v3d_bo.c:(.text+0x3ca): undefined reference to `vm_insert_mixed'

The other drivers with this problem already depend on CONFIG_MMU,
so let's do the same thing here.

Fixes: 57692c94dcbe ("drm/v3d: Introduce a new DRM driver for Broadcom V3D V3.x+")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Eric Anholt <eric@anholt.net>
Link: https://patchwork.freedesktop.org/patch/msgid/20180525155030.3667352-5-arnd@arndb.de
---
 drivers/gpu/drm/v3d/Kconfig | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/v3d/Kconfig b/drivers/gpu/drm/v3d/Kconfig
index a0c0259..1552bf5 100644
--- a/drivers/gpu/drm/v3d/Kconfig
+++ b/drivers/gpu/drm/v3d/Kconfig
@@ -3,6 +3,7 @@ config DRM_V3D
 	depends on ARCH_BCM || ARCH_BCMSTB || COMPILE_TEST
 	depends on DRM
 	depends on COMMON_CLK
+	depends on MMU
 	select DRM_SCHED
 	help
 	  Choose this option if you have a system that has a Broadcom
-- 
cgit v1.1


From dd166fbd8d468eae7a6e8f7adab4dd1e22b1a780 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Thu, 17 May 2018 16:28:24 +0100
Subject: drm/i915: Nul-terminate legacy debug string
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Make sure that when we don't have any scheduler attributes for the
request, the string is terminated.

Fixes: 247870ac8ea7 ("drm/i915: Build request info on stack before printk")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180517152824.11619-1-chris@chris-wilson.co.uk
(cherry picked from commit 96d4f03c20d04c80026b1ec3643c090cf4f0eb20)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
---
 drivers/gpu/drm/i915/intel_engine_cs.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
index 6bfd7e3..1590375 100644
--- a/drivers/gpu/drm/i915/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/intel_engine_cs.c
@@ -1114,7 +1114,7 @@ static void print_request(struct drm_printer *m,
 			  const char *prefix)
 {
 	const char *name = rq->fence.ops->get_timeline_name(&rq->fence);
-	char buf[80];
+	char buf[80] = "";
 	int x = 0;
 
 	x = print_sched_attr(rq->i915, &rq->sched.attr, buf, x, sizeof(buf));
-- 
cgit v1.1


From 368b554d63948133aca05e63ff8f5f4fbc2804d4 Mon Sep 17 00:00:00 2001
From: Jani Nikula <jani.nikula@intel.com>
Date: Wed, 16 May 2018 11:01:10 +0300
Subject: Revert "drm/i915/edp: Allow alternate fixed mode for eDP if
 available."

This reverts commit dc911f5bd8aacfcf8aabd5c26c88e04c837a938e.

Per the report, no matter what display mode you select with xrandr, the
i915 driver will always select the alternate fixed mode. For the
reporter this means that the display will always run at 40Hz which is
quite annoying. This may be due to the mode comparison.

But there are some other potential issues. The choice of alt_fixed_mode
seems dubious. It's the first non-preferred mode, but there are no
guarantees that the only difference would be refresh rate. Similarly,
there may be more than one preferred mode in the probed modes list, and
the commit changes the preferred mode selection to choose the last one
on the list instead of the first.

(Note that the probed modes list is the raw, unfiltered, unsorted list
of modes from drm_add_edid_modes(), not the pretty result after a
drm_helper_probe_single_connector_modes() call.)

Finally, we already have eerily similar code in place to find the
downclock mode for DRRS that seems like could be reused here.

Back to the drawing board.

Note: This is a hand-crafted revert due to conflicts. If it fails to
backport, please just try reverting the original commit directly.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105469
Reported-by: Rune Petersen <rune@megahurts.dk>
Reported-by: Mark Spencer <n7u4722r35@ynzlx.anonbox.net>
Fixes: dc911f5bd8aa ("drm/i915/edp: Allow alternate fixed mode for eDP if available.")
Cc: Clint Taylor <clinton.a.taylor@intel.com>
Cc: David Weinehall <david.weinehall@linux.intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Paulo Zanoni <paulo.r.zanoni@intel.com>
Cc: Jani Nikula <jani.nikula@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Jim Bride <jim.bride@linux.intel.com>
Cc: Jani Nikula <jani.nikula@linux.intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: intel-gfx@lists.freedesktop.org
Cc: <stable@vger.kernel.org> # v4.14+
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180516080110.22770-1-jani.nikula@intel.com
(cherry picked from commit d93fa1b47b8fcd149b5091f18385304f402a8e15)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
---
 drivers/gpu/drm/i915/intel_dp.c    | 38 +++++---------------------------------
 drivers/gpu/drm/i915/intel_drv.h   |  2 --
 drivers/gpu/drm/i915/intel_dsi.c   |  2 +-
 drivers/gpu/drm/i915/intel_dvo.c   |  2 +-
 drivers/gpu/drm/i915/intel_lvds.c  |  3 +--
 drivers/gpu/drm/i915/intel_panel.c |  6 ------
 6 files changed, 8 insertions(+), 45 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index dde92e4..8320f0e 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -1679,23 +1679,6 @@ static int intel_dp_compute_bpp(struct intel_dp *intel_dp,
 	return bpp;
 }
 
-static bool intel_edp_compare_alt_mode(struct drm_display_mode *m1,
-				       struct drm_display_mode *m2)
-{
-	bool bres = false;
-
-	if (m1 && m2)
-		bres = (m1->hdisplay == m2->hdisplay &&
-			m1->hsync_start == m2->hsync_start &&
-			m1->hsync_end == m2->hsync_end &&
-			m1->htotal == m2->htotal &&
-			m1->vdisplay == m2->vdisplay &&
-			m1->vsync_start == m2->vsync_start &&
-			m1->vsync_end == m2->vsync_end &&
-			m1->vtotal == m2->vtotal);
-	return bres;
-}
-
 /* Adjust link config limits based on compliance test requests. */
 static void
 intel_dp_adjust_compliance_config(struct intel_dp *intel_dp,
@@ -1860,16 +1843,8 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 		pipe_config->has_audio = intel_conn_state->force_audio == HDMI_AUDIO_ON;
 
 	if (intel_dp_is_edp(intel_dp) && intel_connector->panel.fixed_mode) {
-		struct drm_display_mode *panel_mode =
-			intel_connector->panel.alt_fixed_mode;
-		struct drm_display_mode *req_mode = &pipe_config->base.mode;
-
-		if (!intel_edp_compare_alt_mode(req_mode, panel_mode))
-			panel_mode = intel_connector->panel.fixed_mode;
-
-		drm_mode_debug_printmodeline(panel_mode);
-
-		intel_fixed_panel_mode(panel_mode, adjusted_mode);
+		intel_fixed_panel_mode(intel_connector->panel.fixed_mode,
+				       adjusted_mode);
 
 		if (INTEL_GEN(dev_priv) >= 9) {
 			int ret;
@@ -6159,7 +6134,6 @@ static bool intel_edp_init_connector(struct intel_dp *intel_dp,
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct drm_connector *connector = &intel_connector->base;
 	struct drm_display_mode *fixed_mode = NULL;
-	struct drm_display_mode *alt_fixed_mode = NULL;
 	struct drm_display_mode *downclock_mode = NULL;
 	bool has_dpcd;
 	struct drm_display_mode *scan;
@@ -6214,14 +6188,13 @@ static bool intel_edp_init_connector(struct intel_dp *intel_dp,
 	}
 	intel_connector->edid = edid;
 
-	/* prefer fixed mode from EDID if available, save an alt mode also */
+	/* prefer fixed mode from EDID if available */
 	list_for_each_entry(scan, &connector->probed_modes, head) {
 		if ((scan->type & DRM_MODE_TYPE_PREFERRED)) {
 			fixed_mode = drm_mode_duplicate(dev, scan);
 			downclock_mode = intel_dp_drrs_init(
 						intel_connector, fixed_mode);
-		} else if (!alt_fixed_mode) {
-			alt_fixed_mode = drm_mode_duplicate(dev, scan);
+			break;
 		}
 	}
 
@@ -6258,8 +6231,7 @@ static bool intel_edp_init_connector(struct intel_dp *intel_dp,
 			      pipe_name(pipe));
 	}
 
-	intel_panel_init(&intel_connector->panel, fixed_mode, alt_fixed_mode,
-			 downclock_mode);
+	intel_panel_init(&intel_connector->panel, fixed_mode, downclock_mode);
 	intel_connector->panel.backlight.power = intel_edp_backlight_power;
 	intel_panel_setup_backlight(connector, pipe);
 
diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
index d7dbca1..0361130 100644
--- a/drivers/gpu/drm/i915/intel_drv.h
+++ b/drivers/gpu/drm/i915/intel_drv.h
@@ -277,7 +277,6 @@ struct intel_encoder {
 
 struct intel_panel {
 	struct drm_display_mode *fixed_mode;
-	struct drm_display_mode *alt_fixed_mode;
 	struct drm_display_mode *downclock_mode;
 
 	/* backlight */
@@ -1850,7 +1849,6 @@ void intel_overlay_reset(struct drm_i915_private *dev_priv);
 /* intel_panel.c */
 int intel_panel_init(struct intel_panel *panel,
 		     struct drm_display_mode *fixed_mode,
-		     struct drm_display_mode *alt_fixed_mode,
 		     struct drm_display_mode *downclock_mode);
 void intel_panel_fini(struct intel_panel *panel);
 void intel_fixed_panel_mode(const struct drm_display_mode *fixed_mode,
diff --git a/drivers/gpu/drm/i915/intel_dsi.c b/drivers/gpu/drm/i915/intel_dsi.c
index 51a1d68..cf39ca9 100644
--- a/drivers/gpu/drm/i915/intel_dsi.c
+++ b/drivers/gpu/drm/i915/intel_dsi.c
@@ -1846,7 +1846,7 @@ void intel_dsi_init(struct drm_i915_private *dev_priv)
 	connector->display_info.width_mm = fixed_mode->width_mm;
 	connector->display_info.height_mm = fixed_mode->height_mm;
 
-	intel_panel_init(&intel_connector->panel, fixed_mode, NULL, NULL);
+	intel_panel_init(&intel_connector->panel, fixed_mode, NULL);
 	intel_panel_setup_backlight(connector, INVALID_PIPE);
 
 	intel_dsi_add_properties(intel_connector);
diff --git a/drivers/gpu/drm/i915/intel_dvo.c b/drivers/gpu/drm/i915/intel_dvo.c
index eb0c559b2..a70d767 100644
--- a/drivers/gpu/drm/i915/intel_dvo.c
+++ b/drivers/gpu/drm/i915/intel_dvo.c
@@ -536,7 +536,7 @@ void intel_dvo_init(struct drm_i915_private *dev_priv)
 			 */
 			intel_panel_init(&intel_connector->panel,
 					 intel_dvo_get_current_mode(intel_encoder),
-					 NULL, NULL);
+					 NULL);
 			intel_dvo->panel_wants_dither = true;
 		}
 
diff --git a/drivers/gpu/drm/i915/intel_lvds.c b/drivers/gpu/drm/i915/intel_lvds.c
index 8691c86..d8ece90 100644
--- a/drivers/gpu/drm/i915/intel_lvds.c
+++ b/drivers/gpu/drm/i915/intel_lvds.c
@@ -1140,8 +1140,7 @@ void intel_lvds_init(struct drm_i915_private *dev_priv)
 out:
 	mutex_unlock(&dev->mode_config.mutex);
 
-	intel_panel_init(&intel_connector->panel, fixed_mode, NULL,
-			 downclock_mode);
+	intel_panel_init(&intel_connector->panel, fixed_mode, downclock_mode);
 	intel_panel_setup_backlight(connector, INVALID_PIPE);
 
 	lvds_encoder->is_dual_link = compute_is_dual_link_lvds(lvds_encoder);
diff --git a/drivers/gpu/drm/i915/intel_panel.c b/drivers/gpu/drm/i915/intel_panel.c
index 41d00b1..b443278 100644
--- a/drivers/gpu/drm/i915/intel_panel.c
+++ b/drivers/gpu/drm/i915/intel_panel.c
@@ -1928,13 +1928,11 @@ intel_panel_init_backlight_funcs(struct intel_panel *panel)
 
 int intel_panel_init(struct intel_panel *panel,
 		     struct drm_display_mode *fixed_mode,
-		     struct drm_display_mode *alt_fixed_mode,
 		     struct drm_display_mode *downclock_mode)
 {
 	intel_panel_init_backlight_funcs(panel);
 
 	panel->fixed_mode = fixed_mode;
-	panel->alt_fixed_mode = alt_fixed_mode;
 	panel->downclock_mode = downclock_mode;
 
 	return 0;
@@ -1948,10 +1946,6 @@ void intel_panel_fini(struct intel_panel *panel)
 	if (panel->fixed_mode)
 		drm_mode_destroy(intel_connector->base.dev, panel->fixed_mode);
 
-	if (panel->alt_fixed_mode)
-		drm_mode_destroy(intel_connector->base.dev,
-				panel->alt_fixed_mode);
-
 	if (panel->downclock_mode)
 		drm_mode_destroy(intel_connector->base.dev,
 				panel->downclock_mode);
-- 
cgit v1.1


From bc61ec46e83f84b7f42bd775e18567471f1aaaa6 Mon Sep 17 00:00:00 2001
From: Chris Wilson <chris@chris-wilson.co.uk>
Date: Tue, 29 May 2018 14:29:18 +0100
Subject: drm/i915: Remove stale asserts from i915_gem_find_active_request()

Since we use i915_gem_find_active_request() from inside
intel_engine_dump() and may call that at any time, we do not guarantee
that the engine is paused nor that the signal kthreads and irq handler
are suspended, so we cannot assert that the breadcrumb doesn't advance
and that the irq hasn't happened on another CPU signaling the request we
believe to be idle.

The second assert removed (that request->engine == engine) remains
valid, but is now more rigorously checked during retirement.

Fixes: f636edb214a5 ("drm/i915: Make i915_engine_info pretty printer to standalone")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180529132922.6831-1-chris@chris-wilson.co.uk
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
(cherry picked from commit cc7cc5343584d90e74b7c929ff2c9a2ec8b49cfe)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
---
 drivers/gpu/drm/i915/i915_gem.c | 17 ++++++++---------
 1 file changed, 8 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 0a20701..3704f4c 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -2972,23 +2972,22 @@ i915_gem_find_active_request(struct intel_engine_cs *engine)
 	struct i915_request *request, *active = NULL;
 	unsigned long flags;
 
-	/* We are called by the error capture and reset at a random
-	 * point in time. In particular, note that neither is crucially
-	 * ordered with an interrupt. After a hang, the GPU is dead and we
-	 * assume that no more writes can happen (we waited long enough for
-	 * all writes that were in transaction to be flushed) - adding an
+	/*
+	 * We are called by the error capture, reset and to dump engine
+	 * state at random points in time. In particular, note that neither is
+	 * crucially ordered with an interrupt. After a hang, the GPU is dead
+	 * and we assume that no more writes can happen (we waited long enough
+	 * for all writes that were in transaction to be flushed) - adding an
 	 * extra delay for a recent interrupt is pointless. Hence, we do
 	 * not need an engine->irq_seqno_barrier() before the seqno reads.
+	 * At all other times, we must assume the GPU is still running, but
+	 * we only care about the snapshot of this moment.
 	 */
 	spin_lock_irqsave(&engine->timeline.lock, flags);
 	list_for_each_entry(request, &engine->timeline.requests, link) {
 		if (__i915_request_completed(request, request->global_seqno))
 			continue;
 
-		GEM_BUG_ON(request->engine != engine);
-		GEM_BUG_ON(test_bit(DMA_FENCE_FLAG_SIGNALED_BIT,
-				    &request->fence.flags));
-
 		active = request;
 		break;
 	}
-- 
cgit v1.1


From 46defdd6fff70edf6bd21848ee75d927c36e4153 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Wed, 30 May 2018 16:52:22 +0800
Subject: drm/amd/pp: Allow underclocking when od table is empty in vbios

if max od engine clock limit and memory clock limit are not set
in vbios. driver will allow underclocking instand of disable od feature
completely.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.c  | 6 ------
 drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c        | 6 ------
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c             | 5 ++++-
 drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c           | 6 ++++++
 drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c | 6 ------
 5 files changed, 10 insertions(+), 19 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.c b/drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.c
index f0d48b1..35bd987 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.c
@@ -870,12 +870,6 @@ static int init_over_drive_limits(
 	hwmgr->platform_descriptor.maxOverdriveVDDC = 0;
 	hwmgr->platform_descriptor.overdriveVDDCStep = 0;
 
-	if (hwmgr->platform_descriptor.overdriveLimit.engineClock == 0 \
-		|| hwmgr->platform_descriptor.overdriveLimit.memoryClock == 0) {
-		hwmgr->od_enabled = false;
-		pr_debug("OverDrive feature not support by VBIOS\n");
-	}
-
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c b/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c
index ce64dfa..925e171 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c
@@ -1074,12 +1074,6 @@ static int init_overdrive_limits(struct pp_hwmgr *hwmgr,
 				powerplay_table,
 				(const ATOM_FIRMWARE_INFO_V2_1 *)fw_info);
 
-	if (hwmgr->platform_descriptor.overdriveLimit.engineClock == 0
-		&& hwmgr->platform_descriptor.overdriveLimit.memoryClock == 0) {
-		hwmgr->od_enabled = false;
-		pr_debug("OverDrive feature not support by VBIOS\n");
-	}
-
 	return result;
 }
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index 45e9b8c..b763c54 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -791,7 +791,8 @@ static int smu7_setup_dpm_tables_v1(struct pp_hwmgr *hwmgr)
 			data->dpm_table.sclk_table.count++;
 		}
 	}
-
+	if (hwmgr->platform_descriptor.overdriveLimit.engineClock == 0)
+		hwmgr->platform_descriptor.overdriveLimit.engineClock = dep_sclk_table->entries[i-1].clk;
 	/* Initialize Mclk DPM table based on allow Mclk values */
 	data->dpm_table.mclk_table.count = 0;
 	for (i = 0; i < dep_mclk_table->count; i++) {
@@ -806,6 +807,8 @@ static int smu7_setup_dpm_tables_v1(struct pp_hwmgr *hwmgr)
 		}
 	}
 
+	if (hwmgr->platform_descriptor.overdriveLimit.memoryClock == 0)
+		hwmgr->platform_descriptor.overdriveLimit.memoryClock = dep_mclk_table->entries[i-1].clk;
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
index d156b7b..f70dbc8 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
@@ -1311,6 +1311,9 @@ static int vega10_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
 	vega10_setup_default_single_dpm_table(hwmgr,
 			dpm_table,
 			dep_gfx_table);
+	if (hwmgr->platform_descriptor.overdriveLimit.engineClock == 0)
+		hwmgr->platform_descriptor.overdriveLimit.engineClock =
+					dpm_table->dpm_levels[dpm_table->count-1].value;
 	vega10_init_dpm_state(&(dpm_table->dpm_state));
 
 	/* Initialize Mclk DPM table based on allow Mclk values */
@@ -1319,6 +1322,9 @@ static int vega10_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
 	vega10_setup_default_single_dpm_table(hwmgr,
 			dpm_table,
 			dep_mclk_table);
+	if (hwmgr->platform_descriptor.overdriveLimit.memoryClock == 0)
+		hwmgr->platform_descriptor.overdriveLimit.memoryClock =
+					dpm_table->dpm_levels[dpm_table->count-1].value;
 	vega10_init_dpm_state(&(dpm_table->dpm_state));
 
 	data->dpm_table.eclk_table.count = 0;
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c
index 0768d25..16b1a9c 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c
@@ -267,12 +267,6 @@ static int init_over_drive_limits(
 	hwmgr->platform_descriptor.maxOverdriveVDDC = 0;
 	hwmgr->platform_descriptor.overdriveVDDCStep = 0;
 
-	if (hwmgr->platform_descriptor.overdriveLimit.engineClock == 0 ||
-		hwmgr->platform_descriptor.overdriveLimit.memoryClock == 0) {
-		hwmgr->od_enabled = false;
-		pr_debug("OverDrive feature not support by VBIOS\n");
-	}
-
 	return 0;
 }
 
-- 
cgit v1.1


From ac26b0f3fc41b942f03ec9fd0392e3aa1886800a Mon Sep 17 00:00:00 2001
From: Feifei Xu <Feifei.Xu@amd.com>
Date: Thu, 24 May 2018 15:36:57 +0800
Subject: drm/gfx9: Update gc goldensetting for vega20.

Update mmCB_DCC_CONFIG register goldensetting.

Signed-off-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index d7530fd..4f7a72d 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -111,6 +111,7 @@ static const struct soc15_reg_golden golden_settings_gc_9_0_vg10[] =
 
 static const struct soc15_reg_golden golden_settings_gc_9_0_vg20[] =
 {
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_DCC_CONFIG, 0x0f000080, 0x04000080),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_HW_CONTROL_2, 0x0f000000, 0x0a000000),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_HW_CONTROL_3, 0x30000000, 0x10000000),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmGB_ADDR_CONFIG, 0xf3e777ff, 0x22014042),
-- 
cgit v1.1


From 7ba01f9e12bb3f088f617cf69b589ea37bd5d6ed Mon Sep 17 00:00:00 2001
From: Shaoyun Liu <Shaoyun.Liu@amd.com>
Date: Wed, 14 Mar 2018 14:44:58 -0400
Subject: drm/amdgpu: Fix NULL pointer when load kfd driver with PP block is
 disabled

When PP block is disabled, return a fix value(100M) for mclk and sclk on
bare-metal mode. This will cover the emulation mode as well.

Signed-off-by: Shaoyun Liu <Shaoyun.Liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c | 18 +++++++-----------
 1 file changed, 7 insertions(+), 11 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
index bd36ee9..60fc441 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
@@ -336,15 +336,12 @@ void get_local_mem_info(struct kgd_dev *kgd,
 			mem_info->local_mem_size_public,
 			mem_info->local_mem_size_private);
 
-	if (amdgpu_emu_mode == 1) {
-		mem_info->mem_clk_max = 100;
-		return;
-	}
-
 	if (amdgpu_sriov_vf(adev))
 		mem_info->mem_clk_max = adev->clock.default_mclk / 100;
-	else
+	else if (adev->powerplay.pp_funcs)
 		mem_info->mem_clk_max = amdgpu_dpm_get_mclk(adev, false) / 100;
+	else
+		mem_info->mem_clk_max = 100;
 }
 
 uint64_t get_gpu_clock_counter(struct kgd_dev *kgd)
@@ -361,13 +358,12 @@ uint32_t get_max_engine_clock_in_mhz(struct kgd_dev *kgd)
 	struct amdgpu_device *adev = (struct amdgpu_device *)kgd;
 
 	/* the sclk is in quantas of 10kHz */
-	if (amdgpu_emu_mode == 1)
-		return 100;
-
 	if (amdgpu_sriov_vf(adev))
 		return adev->clock.default_sclk / 100;
-
-	return amdgpu_dpm_get_sclk(adev, false) / 100;
+	else if (adev->powerplay.pp_funcs)
+		return amdgpu_dpm_get_sclk(adev, false) / 100;
+	else
+		return 100;
 }
 
 void get_cu_info(struct kgd_dev *kgd, struct kfd_cu_info *cu_info)
-- 
cgit v1.1


From 036286e28e8ece7637cc62934f7dbc89e6656940 Mon Sep 17 00:00:00 2001
From: Colin Ian King <colin.king@canonical.com>
Date: Wed, 30 May 2018 17:41:44 +0100
Subject: drm/amdgpu/df: fix potential array out-of-bounds read

The comparison with the number of elements in array df_v3_7_channel_number
is off-by-one and can produce an array out-of-bounds read if
fb_channel_number is equal to the number of elements of the array. Fix
this by changing the comparison to >= instead of >.

Detected by CoverityScan, CID#1469489 ("Out-of-bounds read")

Fixes: 13b581502d51 ("drm/amdgpu/df: implement df v3_6 callback functions (v2)")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/df_v3_6.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/df_v3_6.c b/drivers/gpu/drm/amd/amdgpu/df_v3_6.c
index 60608b3..d5ebe56 100644
--- a/drivers/gpu/drm/amd/amdgpu/df_v3_6.c
+++ b/drivers/gpu/drm/amd/amdgpu/df_v3_6.c
@@ -64,7 +64,7 @@ static u32 df_v3_6_get_hbm_channel_number(struct amdgpu_device *adev)
 	int fb_channel_number;
 
 	fb_channel_number = adev->df_funcs->get_fb_channel_number(adev);
-	if (fb_channel_number > ARRAY_SIZE(df_v3_6_channel_number))
+	if (fb_channel_number >= ARRAY_SIZE(df_v3_6_channel_number))
 		fb_channel_number = 0;
 
 	return df_v3_6_channel_number[fb_channel_number];
-- 
cgit v1.1


From 161c68eb502e90a5cf25913c577e23217590e43f Mon Sep 17 00:00:00 2001
From: Alex Deucher <alexander.deucher@amd.com>
Date: Thu, 24 May 2018 14:37:39 -0500
Subject: Revert "drm/amdgpu: Add an ATPX quirk for hybrid laptop"

This reverts commit 13b40935cf64f59b93cf1c716a2033488e5a228c.

This was a workaround for a bug in the HDA driver that prevented
the HDA audio chip from going into runtime pm which prevented
the GPU from going into runtime pm.

Bug: https://bugs.freedesktop.org/show_bug.cgi?id=106597
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c
index 1bcb2b2..daa06e7 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c
@@ -569,7 +569,6 @@ static const struct amdgpu_px_quirk amdgpu_px_quirk_list[] = {
 	{ 0x1002, 0x6900, 0x1002, 0x0124, AMDGPU_PX_QUIRK_FORCE_ATPX },
 	{ 0x1002, 0x6900, 0x1028, 0x0812, AMDGPU_PX_QUIRK_FORCE_ATPX },
 	{ 0x1002, 0x6900, 0x1028, 0x0813, AMDGPU_PX_QUIRK_FORCE_ATPX },
-	{ 0x1002, 0x67DF, 0x1028, 0x0774, AMDGPU_PX_QUIRK_FORCE_ATPX },
 	{ 0, 0, 0, 0, 0 },
 };
 
-- 
cgit v1.1


From 79ebc86c41959113009e3d9b17dc2d4547512ac9 Mon Sep 17 00:00:00 2001
From: Abhinav Kumar <abhinavk@codeaurora.org>
Date: Mon, 16 Apr 2018 19:05:30 -0700
Subject: drm/msm/dsi: check return value for video done waits

Check for the return value of wait for video
done waits and print appropriate error message.

Signed-off-by: Abhinav Kumar <abhinavk@codeaurora.org>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
---
 drivers/gpu/drm/msm/dsi/dsi_host.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/msm/dsi/dsi_host.c b/drivers/gpu/drm/msm/dsi/dsi_host.c
index 7a03a94..93a3cdd 100644
--- a/drivers/gpu/drm/msm/dsi/dsi_host.c
+++ b/drivers/gpu/drm/msm/dsi/dsi_host.c
@@ -986,13 +986,19 @@ static void dsi_set_tx_power_mode(int mode, struct msm_dsi_host *msm_host)
 
 static void dsi_wait4video_done(struct msm_dsi_host *msm_host)
 {
+	u32 ret = 0;
+	struct device *dev = &msm_host->pdev->dev;
+
 	dsi_intr_ctrl(msm_host, DSI_IRQ_MASK_VIDEO_DONE, 1);
 
 	reinit_completion(&msm_host->video_comp);
 
-	wait_for_completion_timeout(&msm_host->video_comp,
+	ret = wait_for_completion_timeout(&msm_host->video_comp,
 			msecs_to_jiffies(70));
 
+	if (ret <= 0)
+		dev_err(dev, "wait for video done timed out\n");
+
 	dsi_intr_ctrl(msm_host, DSI_IRQ_MASK_VIDEO_DONE, 0);
 }
 
-- 
cgit v1.1


From 9c5638d78df23e3fc5a6f15972f6fc20939ffe8b Mon Sep 17 00:00:00 2001
From: Abhinav Kumar <abhinavk@codeaurora.org>
Date: Tue, 17 Apr 2018 13:50:18 -0700
Subject: drm/msm/dsi: check video mode engine status before waiting

Make sure the video mode engine is on before waiting
for the video done interrupt.

Changes in v4:
- Move setting enabled to false earlier

Changes in v3:
- Move the return value check to another
  patch

Changes in v2:
- Replace pr_err with dev_err
- Changed error message

Signed-off-by: Abhinav Kumar <abhinavk@codeaurora.org>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
---
 drivers/gpu/drm/msm/dsi/dsi_host.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/msm/dsi/dsi_host.c b/drivers/gpu/drm/msm/dsi/dsi_host.c
index 93a3cdd..2f77038 100644
--- a/drivers/gpu/drm/msm/dsi/dsi_host.c
+++ b/drivers/gpu/drm/msm/dsi/dsi_host.c
@@ -173,6 +173,7 @@ struct msm_dsi_host {
 
 	bool registered;
 	bool power_on;
+	bool enabled;
 	int irq;
 };
 
@@ -1007,7 +1008,7 @@ static void dsi_wait4video_eng_busy(struct msm_dsi_host *msm_host)
 	if (!(msm_host->mode_flags & MIPI_DSI_MODE_VIDEO))
 		return;
 
-	if (msm_host->power_on) {
+	if (msm_host->power_on && msm_host->enabled) {
 		dsi_wait4video_done(msm_host);
 		/* delay 4 ms to skip BLLP */
 		usleep_range(2000, 4000);
@@ -2209,7 +2210,7 @@ int msm_dsi_host_enable(struct mipi_dsi_host *host)
 	 *	pm_runtime_put_autosuspend(&msm_host->pdev->dev);
 	 * }
 	 */
-
+	msm_host->enabled = true;
 	return 0;
 }
 
@@ -2217,6 +2218,7 @@ int msm_dsi_host_disable(struct mipi_dsi_host *host)
 {
 	struct msm_dsi_host *msm_host = to_msm_dsi_host(host);
 
+	msm_host->enabled = false;
 	dsi_op_mode_config(msm_host,
 		!!(msm_host->mode_flags & MIPI_DSI_MODE_VIDEO), false);
 
-- 
cgit v1.1


From f7b6bf20e26384c615a6f5d39eee148c36ed4003 Mon Sep 17 00:00:00 2001
From: Abhinav Kumar <abhinavk@codeaurora.org>
Date: Tue, 17 Apr 2018 13:50:19 -0700
Subject: drm/msm/dsi: implement auto PHY timing calculator for 10nm PHY

Currently the DSI PHY timings are hard-coded for a specific panel
for the 10nm PHY.

Replace this with the auto PHY timing calculator which can calculate
the PHY timings for any panel.

Changes in v4:
- None

Changes in v3:
- None

Changes in v2:
- None

Reviewed-by: Sean Paul <seanpaul@chromium.org>
Reviewed-by: Archit Taneja <architt@codeaurora.org>
Signed-off-by: Abhinav Kumar <abhinavk@codeaurora.org>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
---
 drivers/gpu/drm/msm/dsi/phy/dsi_phy.c      | 109 +++++++++++++++++++++++++++++
 drivers/gpu/drm/msm/dsi/phy/dsi_phy.h      |   2 +
 drivers/gpu/drm/msm/dsi/phy/dsi_phy_10nm.c |  28 --------
 3 files changed, 111 insertions(+), 28 deletions(-)

diff --git a/drivers/gpu/drm/msm/dsi/phy/dsi_phy.c b/drivers/gpu/drm/msm/dsi/phy/dsi_phy.c
index 8e9d5c2..9a9fa0c 100644
--- a/drivers/gpu/drm/msm/dsi/phy/dsi_phy.c
+++ b/drivers/gpu/drm/msm/dsi/phy/dsi_phy.c
@@ -265,6 +265,115 @@ int msm_dsi_dphy_timing_calc_v2(struct msm_dsi_dphy_timing *timing,
 	return 0;
 }
 
+int msm_dsi_dphy_timing_calc_v3(struct msm_dsi_dphy_timing *timing,
+	struct msm_dsi_phy_clk_request *clk_req)
+{
+	const unsigned long bit_rate = clk_req->bitclk_rate;
+	const unsigned long esc_rate = clk_req->escclk_rate;
+	s32 ui, ui_x8, lpx;
+	s32 tmax, tmin;
+	s32 pcnt0 = 50;
+	s32 pcnt1 = 50;
+	s32 pcnt2 = 10;
+	s32 pcnt3 = 30;
+	s32 pcnt4 = 10;
+	s32 pcnt5 = 2;
+	s32 coeff = 1000; /* Precision, should avoid overflow */
+	s32 hb_en, hb_en_ckln;
+	s32 temp;
+
+	if (!bit_rate || !esc_rate)
+		return -EINVAL;
+
+	timing->hs_halfbyte_en = 0;
+	hb_en = 0;
+	timing->hs_halfbyte_en_ckln = 0;
+	hb_en_ckln = 0;
+
+	ui = mult_frac(NSEC_PER_MSEC, coeff, bit_rate / 1000);
+	ui_x8 = ui << 3;
+	lpx = mult_frac(NSEC_PER_MSEC, coeff, esc_rate / 1000);
+
+	temp = S_DIV_ROUND_UP(38 * coeff, ui_x8);
+	tmin = max_t(s32, temp, 0);
+	temp = (95 * coeff) / ui_x8;
+	tmax = max_t(s32, temp, 0);
+	timing->clk_prepare = linear_inter(tmax, tmin, pcnt0, 0, false);
+
+	temp = 300 * coeff - (timing->clk_prepare << 3) * ui;
+	tmin = S_DIV_ROUND_UP(temp, ui_x8) - 1;
+	tmax = (tmin > 255) ? 511 : 255;
+	timing->clk_zero = linear_inter(tmax, tmin, pcnt5, 0, false);
+
+	tmin = DIV_ROUND_UP(60 * coeff + 3 * ui, ui_x8);
+	temp = 105 * coeff + 12 * ui - 20 * coeff;
+	tmax = (temp + 3 * ui) / ui_x8;
+	timing->clk_trail = linear_inter(tmax, tmin, pcnt3, 0, false);
+
+	temp = S_DIV_ROUND_UP(40 * coeff + 4 * ui, ui_x8);
+	tmin = max_t(s32, temp, 0);
+	temp = (85 * coeff + 6 * ui) / ui_x8;
+	tmax = max_t(s32, temp, 0);
+	timing->hs_prepare = linear_inter(tmax, tmin, pcnt1, 0, false);
+
+	temp = 145 * coeff + 10 * ui - (timing->hs_prepare << 3) * ui;
+	tmin = S_DIV_ROUND_UP(temp, ui_x8) - 1;
+	tmax = 255;
+	timing->hs_zero = linear_inter(tmax, tmin, pcnt4, 0, false);
+
+	tmin = DIV_ROUND_UP(60 * coeff + 4 * ui, ui_x8) - 1;
+	temp = 105 * coeff + 12 * ui - 20 * coeff;
+	tmax = (temp / ui_x8) - 1;
+	timing->hs_trail = linear_inter(tmax, tmin, pcnt3, 0, false);
+
+	temp = 50 * coeff + ((hb_en << 2) - 8) * ui;
+	timing->hs_rqst = S_DIV_ROUND_UP(temp, ui_x8);
+
+	tmin = DIV_ROUND_UP(100 * coeff, ui_x8) - 1;
+	tmax = 255;
+	timing->hs_exit = linear_inter(tmax, tmin, pcnt2, 0, false);
+
+	temp = 50 * coeff + ((hb_en_ckln << 2) - 8) * ui;
+	timing->hs_rqst_ckln = S_DIV_ROUND_UP(temp, ui_x8);
+
+	temp = 60 * coeff + 52 * ui - 43 * ui;
+	tmin = DIV_ROUND_UP(temp, ui_x8) - 1;
+	tmax = 63;
+	timing->shared_timings.clk_post =
+		linear_inter(tmax, tmin, pcnt2, 0, false);
+
+	temp = 8 * ui + (timing->clk_prepare << 3) * ui;
+	temp += (((timing->clk_zero + 3) << 3) + 11) * ui;
+	temp += hb_en_ckln ? (((timing->hs_rqst_ckln << 3) + 4) * ui) :
+		(((timing->hs_rqst_ckln << 3) + 8) * ui);
+	tmin = S_DIV_ROUND_UP(temp, ui_x8) - 1;
+	tmax = 63;
+	if (tmin > tmax) {
+		temp = linear_inter(tmax << 1, tmin, pcnt2, 0, false);
+		timing->shared_timings.clk_pre = temp >> 1;
+		timing->shared_timings.clk_pre_inc_by_2 = 1;
+	} else {
+		timing->shared_timings.clk_pre =
+			linear_inter(tmax, tmin, pcnt2, 0, false);
+			timing->shared_timings.clk_pre_inc_by_2 = 0;
+	}
+
+	timing->ta_go = 3;
+	timing->ta_sure = 0;
+	timing->ta_get = 4;
+
+	DBG("%d, %d, %d, %d, %d, %d, %d, %d, %d, %d, %d, %d, %d, %d, %d, %d",
+		timing->shared_timings.clk_pre, timing->shared_timings.clk_post,
+		timing->shared_timings.clk_pre_inc_by_2, timing->clk_zero,
+		timing->clk_trail, timing->clk_prepare, timing->hs_exit,
+		timing->hs_zero, timing->hs_prepare, timing->hs_trail,
+		timing->hs_rqst, timing->hs_rqst_ckln, timing->hs_halfbyte_en,
+		timing->hs_halfbyte_en_ckln, timing->hs_prep_dly,
+		timing->hs_prep_dly_ckln);
+
+	return 0;
+}
+
 void msm_dsi_phy_set_src_pll(struct msm_dsi_phy *phy, int pll_id, u32 reg,
 				u32 bit_mask)
 {
diff --git a/drivers/gpu/drm/msm/dsi/phy/dsi_phy.h b/drivers/gpu/drm/msm/dsi/phy/dsi_phy.h
index c56268c..a24ab80 100644
--- a/drivers/gpu/drm/msm/dsi/phy/dsi_phy.h
+++ b/drivers/gpu/drm/msm/dsi/phy/dsi_phy.h
@@ -101,6 +101,8 @@ int msm_dsi_dphy_timing_calc(struct msm_dsi_dphy_timing *timing,
 			     struct msm_dsi_phy_clk_request *clk_req);
 int msm_dsi_dphy_timing_calc_v2(struct msm_dsi_dphy_timing *timing,
 				struct msm_dsi_phy_clk_request *clk_req);
+int msm_dsi_dphy_timing_calc_v3(struct msm_dsi_dphy_timing *timing,
+				struct msm_dsi_phy_clk_request *clk_req);
 void msm_dsi_phy_set_src_pll(struct msm_dsi_phy *phy, int pll_id, u32 reg,
 				u32 bit_mask);
 int msm_dsi_phy_init_common(struct msm_dsi_phy *phy);
diff --git a/drivers/gpu/drm/msm/dsi/phy/dsi_phy_10nm.c b/drivers/gpu/drm/msm/dsi/phy/dsi_phy_10nm.c
index 0af951a..b3fffc8 100644
--- a/drivers/gpu/drm/msm/dsi/phy/dsi_phy_10nm.c
+++ b/drivers/gpu/drm/msm/dsi/phy/dsi_phy_10nm.c
@@ -79,34 +79,6 @@ static void dsi_phy_hw_v3_0_lane_settings(struct msm_dsi_phy *phy)
 	dsi_phy_write(lane_base + REG_DSI_10nm_PHY_LN_TX_DCTRL(3), 0x04);
 }
 
-static int msm_dsi_dphy_timing_calc_v3(struct msm_dsi_dphy_timing *timing,
-				       struct msm_dsi_phy_clk_request *clk_req)
-{
-	/*
-	 * TODO: These params need to be computed, they're currently hardcoded
-	 * for a 1440x2560@60Hz panel with a byteclk of 100.618 Mhz, and a
-	 * default escape clock of 19.2 Mhz.
-	 */
-
-	timing->hs_halfbyte_en = 0;
-	timing->clk_zero = 0x1c;
-	timing->clk_prepare = 0x07;
-	timing->clk_trail = 0x07;
-	timing->hs_exit = 0x23;
-	timing->hs_zero = 0x21;
-	timing->hs_prepare = 0x07;
-	timing->hs_trail = 0x07;
-	timing->hs_rqst = 0x05;
-	timing->ta_sure = 0x00;
-	timing->ta_go = 0x03;
-	timing->ta_get = 0x04;
-
-	timing->shared_timings.clk_pre = 0x2d;
-	timing->shared_timings.clk_post = 0x0d;
-
-	return 0;
-}
-
 static int dsi_10nm_phy_enable(struct msm_dsi_phy *phy, int src_pll_id,
 			       struct msm_dsi_phy_clk_request *clk_req)
 {
-- 
cgit v1.1


From 20387275142422103c3734238685cbee6ac0dffa Mon Sep 17 00:00:00 2001
From: Sean Paul <seanpaul@chromium.org>
Date: Wed, 28 Feb 2018 14:19:00 -0500
Subject: drm/msm: Mark the crtc->state->event consumed

Don't leave the event != NULL once it's consumed, this is used a signal
to the atomic helpers that the event will be handled by the driver.

Changes in v2:
- None
Changes in v3:
- Rebased on Archit's private_obj set
Changes in v4:
- None

Cc: Jeykumar Sankaran <jsanka@codeaurora.org>
Reviewed-by: Archit Taneja <architt@codeaurora.org>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
---
 drivers/gpu/drm/msm/disp/mdp4/mdp4_crtc.c | 1 +
 drivers/gpu/drm/msm/disp/mdp5/mdp5_crtc.c | 1 +
 2 files changed, 2 insertions(+)

diff --git a/drivers/gpu/drm/msm/disp/mdp4/mdp4_crtc.c b/drivers/gpu/drm/msm/disp/mdp4/mdp4_crtc.c
index 6e5e1aa..b001699 100644
--- a/drivers/gpu/drm/msm/disp/mdp4/mdp4_crtc.c
+++ b/drivers/gpu/drm/msm/disp/mdp4/mdp4_crtc.c
@@ -351,6 +351,7 @@ static void mdp4_crtc_atomic_flush(struct drm_crtc *crtc,
 
 	spin_lock_irqsave(&dev->event_lock, flags);
 	mdp4_crtc->event = crtc->state->event;
+	crtc->state->event = NULL;
 	spin_unlock_irqrestore(&dev->event_lock, flags);
 
 	blend_setup(crtc);
diff --git a/drivers/gpu/drm/msm/disp/mdp5/mdp5_crtc.c b/drivers/gpu/drm/msm/disp/mdp5/mdp5_crtc.c
index 9893e43..76b9608 100644
--- a/drivers/gpu/drm/msm/disp/mdp5/mdp5_crtc.c
+++ b/drivers/gpu/drm/msm/disp/mdp5/mdp5_crtc.c
@@ -708,6 +708,7 @@ static void mdp5_crtc_atomic_flush(struct drm_crtc *crtc,
 
 	spin_lock_irqsave(&dev->event_lock, flags);
 	mdp5_crtc->event = crtc->state->event;
+	crtc->state->event = NULL;
 	spin_unlock_irqrestore(&dev->event_lock, flags);
 
 	/*
-- 
cgit v1.1


From 4e4902324a9b46a9111d5e514301e154f938238a Mon Sep 17 00:00:00 2001
From: Jeykumar Sankaran <jsanka@codeaurora.org>
Date: Tue, 13 Feb 2018 12:42:44 -0500
Subject: drm/msm: Add modifier to mdp_get_format arguments

This change plumbs the new fb modifier through the various mdp/disp
get_format hooks.

Signed-off-by: Jeykumar Sankaran <jsanka@codeaurora.org>
[seanpaul pimped out commit message a bit]
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
---
 drivers/gpu/drm/msm/disp/mdp_format.c | 3 ++-
 drivers/gpu/drm/msm/disp/mdp_kms.h    | 2 +-
 drivers/gpu/drm/msm/msm_fb.c          | 3 ++-
 drivers/gpu/drm/msm/msm_kms.h         | 5 ++++-
 4 files changed, 9 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/msm/disp/mdp_format.c b/drivers/gpu/drm/msm/disp/mdp_format.c
index b4a8aa4..005760b 100644
--- a/drivers/gpu/drm/msm/disp/mdp_format.c
+++ b/drivers/gpu/drm/msm/disp/mdp_format.c
@@ -171,7 +171,8 @@ uint32_t mdp_get_formats(uint32_t *pixel_formats, uint32_t max_formats,
 	return i;
 }
 
-const struct msm_format *mdp_get_format(struct msm_kms *kms, uint32_t format)
+const struct msm_format *mdp_get_format(struct msm_kms *kms, uint32_t format,
+		uint64_t modifier)
 {
 	int i;
 	for (i = 0; i < ARRAY_SIZE(formats); i++) {
diff --git a/drivers/gpu/drm/msm/disp/mdp_kms.h b/drivers/gpu/drm/msm/disp/mdp_kms.h
index 1185487..4fa8dbe 100644
--- a/drivers/gpu/drm/msm/disp/mdp_kms.h
+++ b/drivers/gpu/drm/msm/disp/mdp_kms.h
@@ -98,7 +98,7 @@ struct mdp_format {
 #define MDP_FORMAT_IS_YUV(mdp_format) ((mdp_format)->is_yuv)
 
 uint32_t mdp_get_formats(uint32_t *formats, uint32_t max_formats, bool rgb_only);
-const struct msm_format *mdp_get_format(struct msm_kms *kms, uint32_t format);
+const struct msm_format *mdp_get_format(struct msm_kms *kms, uint32_t format, uint64_t modifier);
 
 /* MDP capabilities */
 #define MDP_CAP_SMP		BIT(0)	/* Shared Memory Pool                 */
diff --git a/drivers/gpu/drm/msm/msm_fb.c b/drivers/gpu/drm/msm/msm_fb.c
index 0e0c872..7a16242 100644
--- a/drivers/gpu/drm/msm/msm_fb.c
+++ b/drivers/gpu/drm/msm/msm_fb.c
@@ -183,7 +183,8 @@ static struct drm_framebuffer *msm_framebuffer_init(struct drm_device *dev,
 	hsub = drm_format_horz_chroma_subsampling(mode_cmd->pixel_format);
 	vsub = drm_format_vert_chroma_subsampling(mode_cmd->pixel_format);
 
-	format = kms->funcs->get_format(kms, mode_cmd->pixel_format);
+	format = kms->funcs->get_format(kms, mode_cmd->pixel_format,
+			mode_cmd->modifier[0]);
 	if (!format) {
 		dev_err(dev->dev, "unsupported pixel format: %4.4s\n",
 				(char *)&mode_cmd->pixel_format);
diff --git a/drivers/gpu/drm/msm/msm_kms.h b/drivers/gpu/drm/msm/msm_kms.h
index 17d5824..aaa329d 100644
--- a/drivers/gpu/drm/msm/msm_kms.h
+++ b/drivers/gpu/drm/msm/msm_kms.h
@@ -48,8 +48,11 @@ struct msm_kms_funcs {
 	/* functions to wait for atomic commit completed on each CRTC */
 	void (*wait_for_crtc_commit_done)(struct msm_kms *kms,
 					struct drm_crtc *crtc);
+	/* get msm_format w/ optional format modifiers from drm_mode_fb_cmd2 */
+	const struct msm_format *(*get_format)(struct msm_kms *kms,
+					const uint32_t format,
+					const uint64_t modifiers);
 	/* misc: */
-	const struct msm_format *(*get_format)(struct msm_kms *kms, uint32_t format);
 	long (*round_pixclk)(struct msm_kms *kms, unsigned long rate,
 			struct drm_encoder *encoder);
 	int (*set_split_display)(struct msm_kms *kms,
-- 
cgit v1.1


From 03c94d60261c28689465a16086eb290ab3012aa3 Mon Sep 17 00:00:00 2001
From: Ben Hutchings <ben.hutchings@codethink.co.uk>
Date: Tue, 3 Apr 2018 23:38:45 +0100
Subject: drm/msm: Fix possible null dereference on failure of get_pages()

Commit 62e3a3e342af changed get_pages() to initialise
msm_gem_object::pages before trying to initialise msm_gem_object::sgt,
so that put_pages() would properly clean up pages in the failure
case.

However, this means that put_pages() now needs to check that
msm_gem_object::sgt is not null before trying to clean it up, and
this check was only applied to part of the cleanup code.  Move
it all into the conditional block.  (Strictly speaking we don't
need to make the kfree() conditional, but since we can't avoid
checking for null ourselves we may as well do so.)

Fixes: 62e3a3e342af ("drm/msm: fix leak in failed get_pages")
Signed-off-by: Ben Hutchings <ben.hutchings@codethink.co.uk>
Reviewed-by: Jordan Crouse <jcrouse@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
---
 drivers/gpu/drm/msm/msm_gem.c | 20 +++++++++++---------
 1 file changed, 11 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/msm/msm_gem.c b/drivers/gpu/drm/msm/msm_gem.c
index 9519647..f583bb4 100644
--- a/drivers/gpu/drm/msm/msm_gem.c
+++ b/drivers/gpu/drm/msm/msm_gem.c
@@ -132,17 +132,19 @@ static void put_pages(struct drm_gem_object *obj)
 	struct msm_gem_object *msm_obj = to_msm_bo(obj);
 
 	if (msm_obj->pages) {
-		/* For non-cached buffers, ensure the new pages are clean
-		 * because display controller, GPU, etc. are not coherent:
-		 */
-		if (msm_obj->flags & (MSM_BO_WC|MSM_BO_UNCACHED))
-			dma_unmap_sg(obj->dev->dev, msm_obj->sgt->sgl,
-					msm_obj->sgt->nents, DMA_BIDIRECTIONAL);
+		if (msm_obj->sgt) {
+			/* For non-cached buffers, ensure the new
+			 * pages are clean because display controller,
+			 * GPU, etc. are not coherent:
+			 */
+			if (msm_obj->flags & (MSM_BO_WC|MSM_BO_UNCACHED))
+				dma_unmap_sg(obj->dev->dev, msm_obj->sgt->sgl,
+					     msm_obj->sgt->nents,
+					     DMA_BIDIRECTIONAL);
 
-		if (msm_obj->sgt)
 			sg_free_table(msm_obj->sgt);
-
-		kfree(msm_obj->sgt);
+			kfree(msm_obj->sgt);
+		}
 
 		if (use_pages(obj))
 			drm_gem_put_pages(obj, msm_obj->pages, true, false);
-- 
cgit v1.1


From cf606fe3300cbd3db44785174934e61e5a2ffc37 Mon Sep 17 00:00:00 2001
From: Stefan Agner <stefan@agner.ch>
Date: Mon, 19 Mar 2018 22:26:32 +0100
Subject: drm/msm/dsi: use correct enum in dsi_get_cmd_fmt

The function dsi_get_cmd_fmt returns enum dsi_cmd_dst_format,
use the correct enum value also for MIPI_DSI_FMT_RGB666/_PACKED.

This has been discovered using clang:
  drivers/gpu/drm/msm/dsi/dsi_host.c:743:35: warning: implicit conversion
        from enumeration type 'enum dsi_vid_dst_format' to different
        enumeration type 'enum dsi_cmd_dst_format' [-Wenum-conversion]
          case MIPI_DSI_FMT_RGB666:       return VID_DST_FORMAT_RGB666;
                                          ~~~~~~ ^~~~~~~~~~~~~~~~~~~~~

Signed-off-by: Stefan Agner <stefan@agner.ch>
Reviewed-by: Archit Taneja <architt@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
---
 drivers/gpu/drm/msm/dsi/dsi_host.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/msm/dsi/dsi_host.c b/drivers/gpu/drm/msm/dsi/dsi_host.c
index 2f77038..8baba30 100644
--- a/drivers/gpu/drm/msm/dsi/dsi_host.c
+++ b/drivers/gpu/drm/msm/dsi/dsi_host.c
@@ -776,7 +776,7 @@ static inline enum dsi_cmd_dst_format dsi_get_cmd_fmt(
 	switch (mipi_fmt) {
 	case MIPI_DSI_FMT_RGB888:	return CMD_DST_FORMAT_RGB888;
 	case MIPI_DSI_FMT_RGB666_PACKED:
-	case MIPI_DSI_FMT_RGB666:	return VID_DST_FORMAT_RGB666;
+	case MIPI_DSI_FMT_RGB666:	return CMD_DST_FORMAT_RGB666;
 	case MIPI_DSI_FMT_RGB565:	return CMD_DST_FORMAT_RGB565;
 	default:			return CMD_DST_FORMAT_RGB888;
 	}
-- 
cgit v1.1


From 641be142bfc241b3e1ce7ecf3f04ef03facfe81e Mon Sep 17 00:00:00 2001
From: Emil Velikov <emil.velikov@collabora.com>
Date: Wed, 28 Mar 2018 17:22:16 +0100
Subject: drm/msm: don't deref error pointer in the msm_fbdev_create error path

Currently the error pointer returned by msm_alloc_stolen_fb gets passed
to drm_framebuffer_remove. The latter handles only NULL pointers, thus
a nasty crash will occur.

Drop the unnecessary fail label and the associated checks - both err and
fb will be set at this stage.

Cc: Rob Clark <robdclark@gmail.com>
Cc: linux-arm-msm@vger.kernel.org
Cc: dri-devel@lists.freedesktop.org
Cc: freedreno@lists.freedesktop.org
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Signed-off-by: Rob Clark <robdclark@gmail.com>
---
 drivers/gpu/drm/msm/msm_fbdev.c | 11 ++---------
 1 file changed, 2 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/msm/msm_fbdev.c b/drivers/gpu/drm/msm/msm_fbdev.c
index c178563..456622b 100644
--- a/drivers/gpu/drm/msm/msm_fbdev.c
+++ b/drivers/gpu/drm/msm/msm_fbdev.c
@@ -92,8 +92,7 @@ static int msm_fbdev_create(struct drm_fb_helper *helper,
 
 	if (IS_ERR(fb)) {
 		dev_err(dev->dev, "failed to allocate fb\n");
-		ret = PTR_ERR(fb);
-		goto fail;
+		return PTR_ERR(fb);
 	}
 
 	bo = msm_framebuffer_bo(fb, 0);
@@ -151,13 +150,7 @@ static int msm_fbdev_create(struct drm_fb_helper *helper,
 
 fail_unlock:
 	mutex_unlock(&dev->struct_mutex);
-fail:
-
-	if (ret) {
-		if (fb)
-			drm_framebuffer_remove(fb);
-	}
-
+	drm_framebuffer_remove(fb);
 	return ret;
 }
 
-- 
cgit v1.1


From 3c9620cdff9e94bc1b8dd688612f0ff46da78e0f Mon Sep 17 00:00:00 2001
From: Daniel Mack <daniel@zonque.org>
Date: Thu, 3 May 2018 14:00:55 +0200
Subject: drm/msm: remove unbalanced mutex unlock

This regression stems from 0e08270a1f01 ("drm/msm: Separate locking of
buffer resources from struct_mutex").

Signed-off-by: Daniel Mack <daniel@zonque.org>
Cc: Sushmita Susheelendra <ssusheel@codeaurora.org>
Cc: Rob Clark <rclark@redhat.com>
Fixes: 0e08270a1f01 ("drm/msm: Separate locking of buffer resources from struct_mutex")
Signed-off-by: Rob Clark <robdclark@gmail.com>
---
 drivers/gpu/drm/msm/dsi/dsi_host.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/drivers/gpu/drm/msm/dsi/dsi_host.c b/drivers/gpu/drm/msm/dsi/dsi_host.c
index 8baba30..c4d23d4 100644
--- a/drivers/gpu/drm/msm/dsi/dsi_host.c
+++ b/drivers/gpu/drm/msm/dsi/dsi_host.c
@@ -1036,7 +1036,6 @@ static int dsi_tx_buf_alloc(struct msm_dsi_host *msm_host, int size)
 
 		ret = msm_gem_get_iova(msm_host->tx_gem_obj,
 				priv->kms->aspace, &iova);
-		mutex_unlock(&dev->struct_mutex);
 		if (ret) {
 			pr_err("%s: failed to get iova, %d\n", __func__, ret);
 			return ret;
-- 
cgit v1.1


From acb1acdb69c85375ee80777336dff4df2682da11 Mon Sep 17 00:00:00 2001
From: Daniel Mack <daniel@zonque.org>
Date: Thu, 3 May 2018 14:00:56 +0200
Subject: drm/msm: use correct aspace pointer in msm_gem_put_iova()

Even though msm_gem_put_iova() is currently a NOP function, the caller
should pass in the address space pointer it used to obtain the object.

Other call sites were changed in 8bdcd949bbe7e ("drm/msm: pass
address-space to _get_iova() and friends"), but this one seems to have
been forgotten.

Signed-off-by: Daniel Mack <daniel@zonque.org>
Cc: Rob Clark <robdclark@gmail.com>
Signed-off-by: Rob Clark <robdclark@gmail.com>
---
 drivers/gpu/drm/msm/dsi/dsi_host.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/msm/dsi/dsi_host.c b/drivers/gpu/drm/msm/dsi/dsi_host.c
index c4d23d4..b916f46 100644
--- a/drivers/gpu/drm/msm/dsi/dsi_host.c
+++ b/drivers/gpu/drm/msm/dsi/dsi_host.c
@@ -1066,9 +1066,10 @@ static int dsi_tx_buf_alloc(struct msm_dsi_host *msm_host, int size)
 static void dsi_tx_buf_free(struct msm_dsi_host *msm_host)
 {
 	struct drm_device *dev = msm_host->dev;
+	struct msm_drm_private *priv = dev->dev_private;
 
 	if (msm_host->tx_gem_obj) {
-		msm_gem_put_iova(msm_host->tx_gem_obj, 0);
+		msm_gem_put_iova(msm_host->tx_gem_obj, priv->kms->aspace);
 		drm_gem_object_put_unlocked(msm_host->tx_gem_obj);
 		msm_host->tx_gem_obj = NULL;
 	}
-- 
cgit v1.1


From 8d58ef346f30cbbeb9213a7eb90c832abf903fa0 Mon Sep 17 00:00:00 2001
From: Archit Taneja <architt@codeaurora.org>
Date: Wed, 21 Feb 2018 09:37:22 -0500
Subject: drm/msm/mdp5: Add global state as a private atomic object

Global shared resources (hwpipes, hwmixers and SMP) for MDP5 are
implemented as a part of atomic state by subclassing drm_atomic_state.

The preferred approach is to use the drm_private_obj infrastructure
available in the atomic core.

mdp5_global_state is introduced as a drm atomic private object. The two
funcs mdp5_get_global_state() and mdp5_get_existing_global_state() are
the two variants that will be used to access mdp5_global_state.

This will replace the existing mdp5_state struct (which subclasses
drm_atomic_state) and the funcs around it. These will be removed later
once we mdp5_global_state is put to use everywhere.

Changes in v3:
- Added glob_state_lock instead of pushing it into the core
- Added to the msm atomic helper patch set
Changes in v4:
- None

Signed-off-by: Archit Taneja <architt@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
---
 drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.c | 87 ++++++++++++++++++++++++++++++++
 drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.h | 25 +++++++++
 2 files changed, 112 insertions(+)

diff --git a/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.c b/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.c
index 6d8e3a9..fcbdef3 100644
--- a/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.c
+++ b/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.c
@@ -106,6 +106,86 @@ static void mdp5_swap_state(struct msm_kms *kms, struct drm_atomic_state *state)
 	swap(to_kms_state(state)->state, mdp5_kms->state);
 }
 
+/* Global/shared object state funcs */
+
+/*
+ * This is a helper that returns the private state currently in operation.
+ * Note that this would return the "old_state" if called in the atomic check
+ * path, and the "new_state" after the atomic swap has been done.
+ */
+struct mdp5_global_state *
+mdp5_get_existing_global_state(struct mdp5_kms *mdp5_kms)
+{
+	return to_mdp5_global_state(mdp5_kms->glob_state.state);
+}
+
+/*
+ * This acquires the modeset lock set aside for global state, creates
+ * a new duplicated private object state.
+ */
+struct mdp5_global_state *mdp5_get_global_state(struct drm_atomic_state *s)
+{
+	struct msm_drm_private *priv = s->dev->dev_private;
+	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(priv->kms));
+	struct drm_private_state *priv_state;
+	int ret;
+
+	ret = drm_modeset_lock(&mdp5_kms->glob_state_lock, s->acquire_ctx);
+	if (ret)
+		return ERR_PTR(ret);
+
+	priv_state = drm_atomic_get_private_obj_state(s, &mdp5_kms->glob_state);
+	if (IS_ERR(priv_state))
+		return ERR_CAST(priv_state);
+
+	return to_mdp5_global_state(priv_state);
+}
+
+static struct drm_private_state *
+mdp5_global_duplicate_state(struct drm_private_obj *obj)
+{
+	struct mdp5_global_state *state;
+
+	state = kmemdup(obj->state, sizeof(*state), GFP_KERNEL);
+	if (!state)
+		return NULL;
+
+	__drm_atomic_helper_private_obj_duplicate_state(obj, &state->base);
+
+	return &state->base;
+}
+
+static void mdp5_global_destroy_state(struct drm_private_obj *obj,
+				      struct drm_private_state *state)
+{
+	struct mdp5_global_state *mdp5_state = to_mdp5_global_state(state);
+
+	kfree(mdp5_state);
+}
+
+static const struct drm_private_state_funcs mdp5_global_state_funcs = {
+	.atomic_duplicate_state = mdp5_global_duplicate_state,
+	.atomic_destroy_state = mdp5_global_destroy_state,
+};
+
+static int mdp5_global_obj_init(struct mdp5_kms *mdp5_kms)
+{
+	struct mdp5_global_state *state;
+
+	drm_modeset_lock_init(&mdp5_kms->glob_state_lock);
+
+	state = kzalloc(sizeof(*state), GFP_KERNEL);
+	if (!state)
+		return -ENOMEM;
+
+	state->mdp5_kms = mdp5_kms;
+
+	drm_atomic_private_obj_init(&mdp5_kms->glob_state,
+				    &state->base,
+				    &mdp5_global_state_funcs);
+	return 0;
+}
+
 static void mdp5_prepare_commit(struct msm_kms *kms, struct drm_atomic_state *state)
 {
 	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(kms));
@@ -727,6 +807,9 @@ static void mdp5_destroy(struct platform_device *pdev)
 	if (mdp5_kms->rpm_enabled)
 		pm_runtime_disable(&pdev->dev);
 
+	drm_atomic_private_obj_fini(&mdp5_kms->glob_state);
+	drm_modeset_lock_fini(&mdp5_kms->glob_state_lock);
+
 	kfree(mdp5_kms->state);
 }
 
@@ -887,6 +970,10 @@ static int mdp5_init(struct platform_device *pdev, struct drm_device *dev)
 		goto fail;
 	}
 
+	ret = mdp5_global_obj_init(mdp5_kms);
+	if (ret)
+		goto fail;
+
 	mdp5_kms->mmio = msm_ioremap(pdev, "mdp_phys", "MDP5");
 	if (IS_ERR(mdp5_kms->mmio)) {
 		ret = PTR_ERR(mdp5_kms->mmio);
diff --git a/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.h b/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.h
index 425a03d..76f0ddf 100644
--- a/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.h
+++ b/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.h
@@ -55,6 +55,13 @@ struct mdp5_kms {
 	struct mdp5_state *state;
 	struct drm_modeset_lock state_lock;
 
+	/*
+	 * Global private object state, Do not access directly, use
+	 * mdp5_global_get_state()
+	 */
+	struct drm_modeset_lock glob_state_lock;
+	struct drm_private_obj glob_state;
+
 	struct mdp5_smp *smp;
 	struct mdp5_ctl_manager *ctlm;
 
@@ -95,6 +102,24 @@ struct mdp5_state {
 struct mdp5_state *__must_check
 mdp5_get_state(struct drm_atomic_state *s);
 
+/* Global private object state for tracking resources that are shared across
+ * multiple kms objects (planes/crtcs/etc).
+ */
+#define to_mdp5_global_state(x) container_of(x, struct mdp5_global_state, base)
+struct mdp5_global_state {
+	struct drm_private_state base;
+
+	struct drm_atomic_state *state;
+	struct mdp5_kms *mdp5_kms;
+
+	struct mdp5_hw_pipe_state hwpipe;
+	struct mdp5_hw_mixer_state hwmixer;
+	struct mdp5_smp_state smp;
+};
+
+struct mdp5_global_state * mdp5_get_existing_global_state(struct mdp5_kms *mdp5_kms);
+struct mdp5_global_state *__must_check mdp5_get_global_state(struct drm_atomic_state *s);
+
 /* Atomic plane state.  Subclasses the base drm_plane_state in order to
  * track assigned hwpipe and hw specific state.
  */
-- 
cgit v1.1


From 7907a0d77cb461f58045763c205a5830be72e97c Mon Sep 17 00:00:00 2001
From: Archit Taneja <architt@codeaurora.org>
Date: Wed, 21 Feb 2018 09:37:23 -0500
Subject: drm/msm/mdp5: Use the new private_obj state

This replaces the usage of the subclassed atomic state (mdp5_state)
with a private_obj state embedded within drm_atomic_state. The latter
method is the preferred approach, since it's simpler to implement
and less prone to errors.

The new API replaces the older and equivalent mdp5_state usage in the
following pattern:
- References to "mdp5_kms->state" (i.e, the old/existing state) is
  replaced with mdp5_get_existing_global_state(). In the atomic_check
  path, this should be called with the glob_state_lock drm_modeset_lock
  alredy taken.
- References to "mdp5_get_state()" are replaced with
  mdp5_get_global_state(). This acquires glob_state_lock and uses
  drm_atomic_get_private_obj_state() to create a new duplicated state.

Changes in v3:
- Acquire glob_state_lock in mdp5_smp.c
- Added to the msm atomic helper patch set
Changes in v4:
- None

Signed-off-by: Archit Taneja <architt@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
---
 drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.c   | 10 ++++++++--
 drivers/gpu/drm/msm/disp/mdp5/mdp5_mixer.c | 12 ++++++------
 drivers/gpu/drm/msm/disp/mdp5/mdp5_pipe.c  | 20 +++++++++++---------
 drivers/gpu/drm/msm/disp/mdp5/mdp5_smp.c   | 17 ++++++++++++-----
 4 files changed, 37 insertions(+), 22 deletions(-)

diff --git a/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.c b/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.c
index fcbdef3..6ada098 100644
--- a/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.c
+++ b/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.c
@@ -190,20 +190,26 @@ static void mdp5_prepare_commit(struct msm_kms *kms, struct drm_atomic_state *st
 {
 	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(kms));
 	struct device *dev = &mdp5_kms->pdev->dev;
+	struct mdp5_global_state *global_state;
+
+	global_state = mdp5_get_existing_global_state(mdp5_kms);
 
 	pm_runtime_get_sync(dev);
 
 	if (mdp5_kms->smp)
-		mdp5_smp_prepare_commit(mdp5_kms->smp, &mdp5_kms->state->smp);
+		mdp5_smp_prepare_commit(mdp5_kms->smp, &global_state->smp);
 }
 
 static void mdp5_complete_commit(struct msm_kms *kms, struct drm_atomic_state *state)
 {
 	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(kms));
 	struct device *dev = &mdp5_kms->pdev->dev;
+	struct mdp5_global_state *global_state;
+
+	global_state = mdp5_get_existing_global_state(mdp5_kms);
 
 	if (mdp5_kms->smp)
-		mdp5_smp_complete_commit(mdp5_kms->smp, &mdp5_kms->state->smp);
+		mdp5_smp_complete_commit(mdp5_kms->smp, &global_state->smp);
 
 	pm_runtime_put_sync(dev);
 }
diff --git a/drivers/gpu/drm/msm/disp/mdp5/mdp5_mixer.c b/drivers/gpu/drm/msm/disp/mdp5/mdp5_mixer.c
index 8a00991..113e6b5 100644
--- a/drivers/gpu/drm/msm/disp/mdp5/mdp5_mixer.c
+++ b/drivers/gpu/drm/msm/disp/mdp5/mdp5_mixer.c
@@ -52,14 +52,14 @@ int mdp5_mixer_assign(struct drm_atomic_state *s, struct drm_crtc *crtc,
 {
 	struct msm_drm_private *priv = s->dev->dev_private;
 	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(priv->kms));
-	struct mdp5_state *state = mdp5_get_state(s);
+	struct mdp5_global_state *global_state = mdp5_get_global_state(s);
 	struct mdp5_hw_mixer_state *new_state;
 	int i;
 
-	if (IS_ERR(state))
-		return PTR_ERR(state);
+	if (IS_ERR(global_state))
+		return PTR_ERR(global_state);
 
-	new_state = &state->hwmixer;
+	new_state = &global_state->hwmixer;
 
 	for (i = 0; i < mdp5_kms->num_hwmixers; i++) {
 		struct mdp5_hw_mixer *cur = mdp5_kms->hwmixers[i];
@@ -129,8 +129,8 @@ int mdp5_mixer_assign(struct drm_atomic_state *s, struct drm_crtc *crtc,
 
 void mdp5_mixer_release(struct drm_atomic_state *s, struct mdp5_hw_mixer *mixer)
 {
-	struct mdp5_state *state = mdp5_get_state(s);
-	struct mdp5_hw_mixer_state *new_state = &state->hwmixer;
+	struct mdp5_global_state *global_state = mdp5_get_global_state(s);
+	struct mdp5_hw_mixer_state *new_state = &global_state->hwmixer;
 
 	if (!mixer)
 		return;
diff --git a/drivers/gpu/drm/msm/disp/mdp5/mdp5_pipe.c b/drivers/gpu/drm/msm/disp/mdp5/mdp5_pipe.c
index ff52c49..1ef26bc6 100644
--- a/drivers/gpu/drm/msm/disp/mdp5/mdp5_pipe.c
+++ b/drivers/gpu/drm/msm/disp/mdp5/mdp5_pipe.c
@@ -24,17 +24,19 @@ int mdp5_pipe_assign(struct drm_atomic_state *s, struct drm_plane *plane,
 {
 	struct msm_drm_private *priv = s->dev->dev_private;
 	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(priv->kms));
-	struct mdp5_state *state;
+	struct mdp5_global_state *new_global_state, *old_global_state;
 	struct mdp5_hw_pipe_state *old_state, *new_state;
 	int i, j;
 
-	state = mdp5_get_state(s);
-	if (IS_ERR(state))
-		return PTR_ERR(state);
+	new_global_state = mdp5_get_global_state(s);
+	if (IS_ERR(new_global_state))
+		return PTR_ERR(new_global_state);
 
-	/* grab old_state after mdp5_get_state(), since now we hold lock: */
-	old_state = &mdp5_kms->state->hwpipe;
-	new_state = &state->hwpipe;
+	/* grab old_state after mdp5_get_global_state(), since now we hold lock: */
+	old_global_state = mdp5_get_existing_global_state(mdp5_kms);
+
+	old_state = &old_global_state->hwpipe;
+	new_state = &new_global_state->hwpipe;
 
 	for (i = 0; i < mdp5_kms->num_hwpipes; i++) {
 		struct mdp5_hw_pipe *cur = mdp5_kms->hwpipes[i];
@@ -107,7 +109,7 @@ int mdp5_pipe_assign(struct drm_atomic_state *s, struct drm_plane *plane,
 		WARN_ON(r_hwpipe);
 
 		DBG("%s: alloc SMP blocks", (*hwpipe)->name);
-		ret = mdp5_smp_assign(mdp5_kms->smp, &state->smp,
+		ret = mdp5_smp_assign(mdp5_kms->smp, &new_global_state->smp,
 				(*hwpipe)->pipe, blkcfg);
 		if (ret)
 			return -ENOMEM;
@@ -132,7 +134,7 @@ void mdp5_pipe_release(struct drm_atomic_state *s, struct mdp5_hw_pipe *hwpipe)
 {
 	struct msm_drm_private *priv = s->dev->dev_private;
 	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(priv->kms));
-	struct mdp5_state *state = mdp5_get_state(s);
+	struct mdp5_global_state *state = mdp5_get_global_state(s);
 	struct mdp5_hw_pipe_state *new_state = &state->hwpipe;
 
 	if (!hwpipe)
diff --git a/drivers/gpu/drm/msm/disp/mdp5/mdp5_smp.c b/drivers/gpu/drm/msm/disp/mdp5/mdp5_smp.c
index ae4983d..96c2b82 100644
--- a/drivers/gpu/drm/msm/disp/mdp5/mdp5_smp.c
+++ b/drivers/gpu/drm/msm/disp/mdp5/mdp5_smp.c
@@ -340,17 +340,20 @@ void mdp5_smp_dump(struct mdp5_smp *smp, struct drm_printer *p)
 	struct mdp5_kms *mdp5_kms = get_kms(smp);
 	struct mdp5_hw_pipe_state *hwpstate;
 	struct mdp5_smp_state *state;
+	struct mdp5_global_state *global_state;
 	int total = 0, i, j;
 
 	drm_printf(p, "name\tinuse\tplane\n");
 	drm_printf(p, "----\t-----\t-----\n");
 
 	if (drm_can_sleep())
-		drm_modeset_lock(&mdp5_kms->state_lock, NULL);
+		drm_modeset_lock(&mdp5_kms->glob_state_lock, NULL);
+
+	global_state = mdp5_get_existing_global_state(mdp5_kms);
 
 	/* grab these *after* we hold the state_lock */
-	hwpstate = &mdp5_kms->state->hwpipe;
-	state = &mdp5_kms->state->smp;
+	hwpstate = &global_state->hwpipe;
+	state = &global_state->smp;
 
 	for (i = 0; i < mdp5_kms->num_hwpipes; i++) {
 		struct mdp5_hw_pipe *hwpipe = mdp5_kms->hwpipes[i];
@@ -374,7 +377,7 @@ void mdp5_smp_dump(struct mdp5_smp *smp, struct drm_printer *p)
 			bitmap_weight(state->state, smp->blk_cnt));
 
 	if (drm_can_sleep())
-		drm_modeset_unlock(&mdp5_kms->state_lock);
+		drm_modeset_unlock(&mdp5_kms->glob_state_lock);
 }
 
 void mdp5_smp_destroy(struct mdp5_smp *smp)
@@ -384,7 +387,8 @@ void mdp5_smp_destroy(struct mdp5_smp *smp)
 
 struct mdp5_smp *mdp5_smp_init(struct mdp5_kms *mdp5_kms, const struct mdp5_smp_block *cfg)
 {
-	struct mdp5_smp_state *state = &mdp5_kms->state->smp;
+	struct mdp5_smp_state *state;
+	struct mdp5_global_state *global_state;
 	struct mdp5_smp *smp = NULL;
 	int ret;
 
@@ -398,6 +402,9 @@ struct mdp5_smp *mdp5_smp_init(struct mdp5_kms *mdp5_kms, const struct mdp5_smp_
 	smp->blk_cnt = cfg->mmb_count;
 	smp->blk_size = cfg->mmb_size;
 
+	global_state = mdp5_get_existing_global_state(mdp5_kms);
+	state = &global_state->smp;
+
 	/* statically tied MMBs cannot be re-allocated: */
 	bitmap_copy(state->state, cfg->reserved_state, smp->blk_cnt);
 	memcpy(smp->reserved, cfg->reserved, sizeof(smp->reserved));
-- 
cgit v1.1


From c21c731d93e8148d926a63797d33075128e60cdd Mon Sep 17 00:00:00 2001
From: Archit Taneja <architt@codeaurora.org>
Date: Wed, 21 Feb 2018 09:37:24 -0500
Subject: drm/msm: Don't subclass drm_atomic_state anymore

With the addition of "private_objs" in drm_atomic_state, we no longer
need to subclass drm_atomic_state to store state of share resources
that don't perfectly fit within planes/crtc/connector state information.
We can now save this state within drm_atomic_state itself using
the private objects.

Remove the infrastructure that allowed subclassing of drm_atomic_state
in the driver.

Changes in v3:
- Added to the msm atomic helper patch set
Changes in v4:
- None

Signed-off-by: Archit Taneja <architt@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
---
 drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.c | 46 --------------------------------
 drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.h | 22 ---------------
 drivers/gpu/drm/msm/msm_atomic.c         | 31 ---------------------
 drivers/gpu/drm/msm/msm_drv.c            |  3 ---
 drivers/gpu/drm/msm/msm_kms.h            | 14 ----------
 5 files changed, 116 deletions(-)

diff --git a/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.c b/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.c
index 6ada098..6e12e27 100644
--- a/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.c
+++ b/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.c
@@ -70,42 +70,6 @@ static int mdp5_hw_init(struct msm_kms *kms)
 	return 0;
 }
 
-struct mdp5_state *mdp5_get_state(struct drm_atomic_state *s)
-{
-	struct msm_drm_private *priv = s->dev->dev_private;
-	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(priv->kms));
-	struct msm_kms_state *state = to_kms_state(s);
-	struct mdp5_state *new_state;
-	int ret;
-
-	if (state->state)
-		return state->state;
-
-	ret = drm_modeset_lock(&mdp5_kms->state_lock, s->acquire_ctx);
-	if (ret)
-		return ERR_PTR(ret);
-
-	new_state = kmalloc(sizeof(*mdp5_kms->state), GFP_KERNEL);
-	if (!new_state)
-		return ERR_PTR(-ENOMEM);
-
-	/* Copy state: */
-	new_state->hwpipe = mdp5_kms->state->hwpipe;
-	new_state->hwmixer = mdp5_kms->state->hwmixer;
-	if (mdp5_kms->smp)
-		new_state->smp = mdp5_kms->state->smp;
-
-	state->state = new_state;
-
-	return new_state;
-}
-
-static void mdp5_swap_state(struct msm_kms *kms, struct drm_atomic_state *state)
-{
-	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(kms));
-	swap(to_kms_state(state)->state, mdp5_kms->state);
-}
-
 /* Global/shared object state funcs */
 
 /*
@@ -315,7 +279,6 @@ static const struct mdp_kms_funcs kms_funcs = {
 		.irq             = mdp5_irq,
 		.enable_vblank   = mdp5_enable_vblank,
 		.disable_vblank  = mdp5_disable_vblank,
-		.swap_state      = mdp5_swap_state,
 		.prepare_commit  = mdp5_prepare_commit,
 		.complete_commit = mdp5_complete_commit,
 		.wait_for_crtc_commit_done = mdp5_wait_for_crtc_commit_done,
@@ -815,8 +778,6 @@ static void mdp5_destroy(struct platform_device *pdev)
 
 	drm_atomic_private_obj_fini(&mdp5_kms->glob_state);
 	drm_modeset_lock_fini(&mdp5_kms->glob_state_lock);
-
-	kfree(mdp5_kms->state);
 }
 
 static int construct_pipes(struct mdp5_kms *mdp5_kms, int cnt,
@@ -969,13 +930,6 @@ static int mdp5_init(struct platform_device *pdev, struct drm_device *dev)
 	mdp5_kms->dev = dev;
 	mdp5_kms->pdev = pdev;
 
-	drm_modeset_lock_init(&mdp5_kms->state_lock);
-	mdp5_kms->state = kzalloc(sizeof(*mdp5_kms->state), GFP_KERNEL);
-	if (!mdp5_kms->state) {
-		ret = -ENOMEM;
-		goto fail;
-	}
-
 	ret = mdp5_global_obj_init(mdp5_kms);
 	if (ret)
 		goto fail;
diff --git a/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.h b/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.h
index 76f0ddf..854dfd3 100644
--- a/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.h
+++ b/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.h
@@ -28,8 +28,6 @@
 #include "mdp5_ctl.h"
 #include "mdp5_smp.h"
 
-struct mdp5_state;
-
 struct mdp5_kms {
 	struct mdp_kms base;
 
@@ -49,12 +47,6 @@ struct mdp5_kms {
 	struct mdp5_cfg_handler *cfg;
 	uint32_t caps;	/* MDP capabilities (MDP_CAP_XXX bits) */
 
-	/**
-	 * Global atomic state.  Do not access directly, use mdp5_get_state()
-	 */
-	struct mdp5_state *state;
-	struct drm_modeset_lock state_lock;
-
 	/*
 	 * Global private object state, Do not access directly, use
 	 * mdp5_global_get_state()
@@ -88,20 +80,6 @@ struct mdp5_kms {
 };
 #define to_mdp5_kms(x) container_of(x, struct mdp5_kms, base)
 
-/* Global atomic state for tracking resources that are shared across
- * multiple kms objects (planes/crtcs/etc).
- *
- * For atomic updates which require modifying global state,
- */
-struct mdp5_state {
-	struct mdp5_hw_pipe_state hwpipe;
-	struct mdp5_hw_mixer_state hwmixer;
-	struct mdp5_smp_state smp;
-};
-
-struct mdp5_state *__must_check
-mdp5_get_state(struct drm_atomic_state *s);
-
 /* Global private object state for tracking resources that are shared across
  * multiple kms objects (planes/crtcs/etc).
  */
diff --git a/drivers/gpu/drm/msm/msm_atomic.c b/drivers/gpu/drm/msm/msm_atomic.c
index bf5f8c3..9d0a0ca 100644
--- a/drivers/gpu/drm/msm/msm_atomic.c
+++ b/drivers/gpu/drm/msm/msm_atomic.c
@@ -224,11 +224,7 @@ int msm_atomic_commit(struct drm_device *dev,
 	 * This is the point of no return - everything below never fails except
 	 * when the hw goes bonghits. Which means we can commit the new state on
 	 * the software side now.
-	 *
-	 * swap driver private state while still holding state_lock
 	 */
-	if (to_kms_state(state)->state)
-		priv->kms->funcs->swap_state(priv->kms, state);
 
 	/*
 	 * Everything below can be run asynchronously without the need to grab
@@ -262,30 +258,3 @@ error:
 	drm_atomic_helper_cleanup_planes(dev, state);
 	return ret;
 }
-
-struct drm_atomic_state *msm_atomic_state_alloc(struct drm_device *dev)
-{
-	struct msm_kms_state *state = kzalloc(sizeof(*state), GFP_KERNEL);
-
-	if (!state || drm_atomic_state_init(dev, &state->base) < 0) {
-		kfree(state);
-		return NULL;
-	}
-
-	return &state->base;
-}
-
-void msm_atomic_state_clear(struct drm_atomic_state *s)
-{
-	struct msm_kms_state *state = to_kms_state(s);
-	drm_atomic_state_default_clear(&state->base);
-	kfree(state->state);
-	state->state = NULL;
-}
-
-void msm_atomic_state_free(struct drm_atomic_state *state)
-{
-	kfree(to_kms_state(state)->state);
-	drm_atomic_state_default_release(state);
-	kfree(state);
-}
diff --git a/drivers/gpu/drm/msm/msm_drv.c b/drivers/gpu/drm/msm/msm_drv.c
index 30cd514..1c89195 100644
--- a/drivers/gpu/drm/msm/msm_drv.c
+++ b/drivers/gpu/drm/msm/msm_drv.c
@@ -42,9 +42,6 @@ static const struct drm_mode_config_funcs mode_config_funcs = {
 	.output_poll_changed = drm_fb_helper_output_poll_changed,
 	.atomic_check = drm_atomic_helper_check,
 	.atomic_commit = msm_atomic_commit,
-	.atomic_state_alloc = msm_atomic_state_alloc,
-	.atomic_state_clear = msm_atomic_state_clear,
-	.atomic_state_free = msm_atomic_state_free,
 };
 
 #ifdef CONFIG_DRM_MSM_REGISTER_LOGGING
diff --git a/drivers/gpu/drm/msm/msm_kms.h b/drivers/gpu/drm/msm/msm_kms.h
index aaa329d..dfd9294 100644
--- a/drivers/gpu/drm/msm/msm_kms.h
+++ b/drivers/gpu/drm/msm/msm_kms.h
@@ -40,8 +40,6 @@ struct msm_kms_funcs {
 	irqreturn_t (*irq)(struct msm_kms *kms);
 	int (*enable_vblank)(struct msm_kms *kms, struct drm_crtc *crtc);
 	void (*disable_vblank)(struct msm_kms *kms, struct drm_crtc *crtc);
-	/* swap global atomic state: */
-	void (*swap_state)(struct msm_kms *kms, struct drm_atomic_state *state);
 	/* modeset, bracketing atomic_commit(): */
 	void (*prepare_commit)(struct msm_kms *kms, struct drm_atomic_state *state);
 	void (*complete_commit)(struct msm_kms *kms, struct drm_atomic_state *state);
@@ -80,18 +78,6 @@ struct msm_kms {
 	struct msm_gem_address_space *aspace;
 };
 
-/**
- * Subclass of drm_atomic_state, to allow kms backend to have driver
- * private global state.  The kms backend can do whatever it wants
- * with the ->state ptr.  On ->atomic_state_clear() the ->state ptr
- * is kfree'd and set back to NULL.
- */
-struct msm_kms_state {
-	struct drm_atomic_state base;
-	void *state;
-};
-#define to_kms_state(x) container_of(x, struct msm_kms_state, base)
-
 static inline void msm_kms_init(struct msm_kms *kms,
 		const struct msm_kms_funcs *funcs)
 {
-- 
cgit v1.1


From 347b90b406ff6b6f3c9f666a527eb46e0fcd7aaf Mon Sep 17 00:00:00 2001
From: Sean Paul <seanpaul@chromium.org>
Date: Wed, 28 Feb 2018 14:18:58 -0500
Subject: drm/msm: Refactor complete_commit() to look more the helpers

Factor out the commit_tail() portions of complete_commit() into a
separate function to facilitate moving to the atomic helpers in future
patches.

Changes in v2:
- None
Changes in v3:
- Rebased on Archit's private_obj set
Changes in v4:
- None

Cc: Jeykumar Sankaran <jsanka@codeaurora.org>
Reviewed-by: Archit Taneja <architt@codeaurora.org>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
---
 drivers/gpu/drm/msm/msm_atomic.c | 25 ++++++++++++++++---------
 1 file changed, 16 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/msm/msm_atomic.c b/drivers/gpu/drm/msm/msm_atomic.c
index 9d0a0ca..c18f0be 100644
--- a/drivers/gpu/drm/msm/msm_atomic.c
+++ b/drivers/gpu/drm/msm/msm_atomic.c
@@ -97,18 +97,12 @@ static void msm_atomic_wait_for_commit_done(struct drm_device *dev,
 	}
 }
 
-/* The (potentially) asynchronous part of the commit.  At this point
- * nothing can fail short of armageddon.
- */
-static void complete_commit(struct msm_commit *c, bool async)
+static void msm_atomic_commit_tail(struct drm_atomic_state *state)
 {
-	struct drm_atomic_state *state = c->state;
 	struct drm_device *dev = state->dev;
 	struct msm_drm_private *priv = dev->dev_private;
 	struct msm_kms *kms = priv->kms;
 
-	drm_atomic_helper_wait_for_fences(dev, state, false);
-
 	kms->funcs->prepare_commit(kms, state);
 
 	drm_atomic_helper_commit_modeset_disables(dev, state);
@@ -135,6 +129,19 @@ static void complete_commit(struct msm_commit *c, bool async)
 	drm_atomic_helper_cleanup_planes(dev, state);
 
 	kms->funcs->complete_commit(kms, state);
+}
+
+/* The (potentially) asynchronous part of the commit.  At this point
+ * nothing can fail short of armageddon.
+ */
+static void complete_commit(struct msm_commit *c)
+{
+	struct drm_atomic_state *state = c->state;
+	struct drm_device *dev = state->dev;
+
+	drm_atomic_helper_wait_for_fences(dev, state, false);
+
+	msm_atomic_commit_tail(state);
 
 	drm_atomic_state_put(state);
 
@@ -143,7 +150,7 @@ static void complete_commit(struct msm_commit *c, bool async)
 
 static void commit_worker(struct work_struct *work)
 {
-	complete_commit(container_of(work, struct msm_commit, work), true);
+	complete_commit(container_of(work, struct msm_commit, work));
 }
 
 /**
@@ -248,7 +255,7 @@ int msm_atomic_commit(struct drm_device *dev,
 		return 0;
 	}
 
-	complete_commit(c, false);
+	complete_commit(c);
 
 	return 0;
 
-- 
cgit v1.1


From db8f4d5d32334b061d0d9c53ec86480377daeaf0 Mon Sep 17 00:00:00 2001
From: Sean Paul <seanpaul@chromium.org>
Date: Tue, 3 Apr 2018 10:42:23 -0400
Subject: drm/msm: Move implicit sync handling to prepare_fb

In preparation for moving to atomic helpers, move the implicit sync
fence handling out of atomic commit and into the plane->prepare_fb()
hook. While we're at it, de-duplicate the mdp*_prepare_fb functions.

Changes in v4:
- Added

Reported-by: Rob Clark <robdclark@gmail.com>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
---
 drivers/gpu/drm/msm/disp/mdp4/mdp4_plane.c | 17 +----------------
 drivers/gpu/drm/msm/disp/mdp5/mdp5_plane.c | 16 +---------------
 drivers/gpu/drm/msm/msm_atomic.c           | 22 ++++++++++++++++++++++
 drivers/gpu/drm/msm/msm_drv.h              |  2 ++
 4 files changed, 26 insertions(+), 31 deletions(-)

diff --git a/drivers/gpu/drm/msm/disp/mdp4/mdp4_plane.c b/drivers/gpu/drm/msm/disp/mdp4/mdp4_plane.c
index 7a1ad3a..20e956e 100644
--- a/drivers/gpu/drm/msm/disp/mdp4/mdp4_plane.c
+++ b/drivers/gpu/drm/msm/disp/mdp4/mdp4_plane.c
@@ -98,21 +98,6 @@ static const struct drm_plane_funcs mdp4_plane_funcs = {
 		.atomic_destroy_state = drm_atomic_helper_plane_destroy_state,
 };
 
-static int mdp4_plane_prepare_fb(struct drm_plane *plane,
-				 struct drm_plane_state *new_state)
-{
-	struct mdp4_plane *mdp4_plane = to_mdp4_plane(plane);
-	struct mdp4_kms *mdp4_kms = get_kms(plane);
-	struct msm_kms *kms = &mdp4_kms->base.base;
-	struct drm_framebuffer *fb = new_state->fb;
-
-	if (!fb)
-		return 0;
-
-	DBG("%s: prepare: FB[%u]", mdp4_plane->name, fb->base.id);
-	return msm_framebuffer_prepare(fb, kms->aspace);
-}
-
 static void mdp4_plane_cleanup_fb(struct drm_plane *plane,
 				  struct drm_plane_state *old_state)
 {
@@ -152,7 +137,7 @@ static void mdp4_plane_atomic_update(struct drm_plane *plane,
 }
 
 static const struct drm_plane_helper_funcs mdp4_plane_helper_funcs = {
-		.prepare_fb = mdp4_plane_prepare_fb,
+		.prepare_fb = msm_atomic_prepare_fb,
 		.cleanup_fb = mdp4_plane_cleanup_fb,
 		.atomic_check = mdp4_plane_atomic_check,
 		.atomic_update = mdp4_plane_atomic_update,
diff --git a/drivers/gpu/drm/msm/disp/mdp5/mdp5_plane.c b/drivers/gpu/drm/msm/disp/mdp5/mdp5_plane.c
index a9f31da..e09bc53 100644
--- a/drivers/gpu/drm/msm/disp/mdp5/mdp5_plane.c
+++ b/drivers/gpu/drm/msm/disp/mdp5/mdp5_plane.c
@@ -245,20 +245,6 @@ static const struct drm_plane_funcs mdp5_plane_funcs = {
 		.atomic_print_state = mdp5_plane_atomic_print_state,
 };
 
-static int mdp5_plane_prepare_fb(struct drm_plane *plane,
-				 struct drm_plane_state *new_state)
-{
-	struct mdp5_kms *mdp5_kms = get_kms(plane);
-	struct msm_kms *kms = &mdp5_kms->base.base;
-	struct drm_framebuffer *fb = new_state->fb;
-
-	if (!new_state->fb)
-		return 0;
-
-	DBG("%s: prepare: FB[%u]", plane->name, fb->base.id);
-	return msm_framebuffer_prepare(fb, kms->aspace);
-}
-
 static void mdp5_plane_cleanup_fb(struct drm_plane *plane,
 				  struct drm_plane_state *old_state)
 {
@@ -543,7 +529,7 @@ static void mdp5_plane_atomic_async_update(struct drm_plane *plane,
 }
 
 static const struct drm_plane_helper_funcs mdp5_plane_helper_funcs = {
-		.prepare_fb = mdp5_plane_prepare_fb,
+		.prepare_fb = msm_atomic_prepare_fb,
 		.cleanup_fb = mdp5_plane_cleanup_fb,
 		.atomic_check = mdp5_plane_atomic_check,
 		.atomic_update = mdp5_plane_atomic_update,
diff --git a/drivers/gpu/drm/msm/msm_atomic.c b/drivers/gpu/drm/msm/msm_atomic.c
index c18f0be..94f9c3e 100644
--- a/drivers/gpu/drm/msm/msm_atomic.c
+++ b/drivers/gpu/drm/msm/msm_atomic.c
@@ -16,6 +16,7 @@
  */
 
 #include "msm_drv.h"
+#include "msm_gem.h"
 #include "msm_kms.h"
 #include "msm_gem.h"
 #include "msm_fence.h"
@@ -97,6 +98,27 @@ static void msm_atomic_wait_for_commit_done(struct drm_device *dev,
 	}
 }
 
+int msm_atomic_prepare_fb(struct drm_plane *plane,
+			  struct drm_plane_state *new_state)
+{
+	struct msm_drm_private *priv = plane->dev->dev_private;
+	struct msm_kms *kms = priv->kms;
+	struct drm_gem_object *obj;
+	struct msm_gem_object *msm_obj;
+	struct dma_fence *fence;
+
+	if (!new_state->fb)
+		return 0;
+
+	obj = msm_framebuffer_bo(new_state->fb, 0);
+	msm_obj = to_msm_bo(obj);
+	fence = reservation_object_get_excl_rcu(msm_obj->resv);
+
+	drm_atomic_set_fence_for_plane(new_state, fence);
+
+	return msm_framebuffer_prepare(new_state->fb, kms->aspace);
+}
+
 static void msm_atomic_commit_tail(struct drm_atomic_state *state)
 {
 	struct drm_device *dev = state->dev;
diff --git a/drivers/gpu/drm/msm/msm_drv.h b/drivers/gpu/drm/msm/msm_drv.h
index 48ed5b9..98e8223 100644
--- a/drivers/gpu/drm/msm/msm_drv.h
+++ b/drivers/gpu/drm/msm/msm_drv.h
@@ -160,6 +160,8 @@ struct msm_format {
 	uint32_t pixel_format;
 };
 
+int msm_atomic_prepare_fb(struct drm_plane *plane,
+			  struct drm_plane_state *new_state);
 int msm_atomic_commit(struct drm_device *dev,
 		struct drm_atomic_state *state, bool nonblock);
 struct drm_atomic_state *msm_atomic_state_alloc(struct drm_device *dev);
-- 
cgit v1.1


From e765ea77b0fd51152e07aa4e6850b81552b76da3 Mon Sep 17 00:00:00 2001
From: Sean Paul <seanpaul@chromium.org>
Date: Wed, 28 Mar 2018 14:41:23 -0400
Subject: drm/msm: Issue queued events when disabling crtc

Ensure that any queued events are issued when disabling the crtc. This
avoids timeouts when we come back and wait for dependencies (like the
previous frame's flip_done).

Changes in v2:
- None
Changes in v3:
- Rebased on Archit's private_obj set
Changes in v4:
- None

Reviewed-by: Archit Taneja <architt@codeaurora.org>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
---
 drivers/gpu/drm/msm/disp/mdp5/mdp5_crtc.c | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/drivers/gpu/drm/msm/disp/mdp5/mdp5_crtc.c b/drivers/gpu/drm/msm/disp/mdp5/mdp5_crtc.c
index 76b9608..1027135 100644
--- a/drivers/gpu/drm/msm/disp/mdp5/mdp5_crtc.c
+++ b/drivers/gpu/drm/msm/disp/mdp5/mdp5_crtc.c
@@ -430,6 +430,7 @@ static void mdp5_crtc_atomic_disable(struct drm_crtc *crtc,
 	struct mdp5_crtc_state *mdp5_cstate = to_mdp5_crtc_state(crtc->state);
 	struct mdp5_kms *mdp5_kms = get_kms(crtc);
 	struct device *dev = &mdp5_kms->pdev->dev;
+	unsigned long flags;
 
 	DBG("%s", crtc->name);
 
@@ -445,6 +446,14 @@ static void mdp5_crtc_atomic_disable(struct drm_crtc *crtc,
 	mdp_irq_unregister(&mdp5_kms->base, &mdp5_crtc->err);
 	pm_runtime_put_sync(dev);
 
+	if (crtc->state->event && !crtc->state->active) {
+		WARN_ON(mdp5_crtc->event);
+		spin_lock_irqsave(&mdp5_kms->dev->event_lock, flags);
+		drm_crtc_send_vblank_event(crtc, crtc->state->event);
+		crtc->state->event = NULL;
+		spin_unlock_irqrestore(&mdp5_kms->dev->event_lock, flags);
+	}
+
 	mdp5_crtc->enabled = false;
 }
 
-- 
cgit v1.1


From 70db18dca4e0130acb0600ad51c33176b6162ccc Mon Sep 17 00:00:00 2001
From: Sean Paul <seanpaul@chromium.org>
Date: Wed, 28 Feb 2018 14:19:01 -0500
Subject: drm/msm: Remove msm_commit/worker, use atomic helper commit

Moving further towards switching fully to the the atomic helpers, this
patch removes the hand-rolled worker nonblock commit code and uses the
atomic helpers commit_work model.

Changes in v2:
- Remove commit_destroy()
- Shuffle order of commit_tail calls to further serialize commits
- Use stall in swap_state to avoid abandoned events on disable
Changes in v3:
- Rebased on Archit's private_obj set
Changes in v4:
- None

Signed-off-by: Sean Paul <seanpaul@chromium.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
---
 drivers/gpu/drm/msm/msm_atomic.c | 153 +++++++++++----------------------------
 drivers/gpu/drm/msm/msm_drv.c    |   1 -
 drivers/gpu/drm/msm/msm_drv.h    |   4 -
 3 files changed, 42 insertions(+), 116 deletions(-)

diff --git a/drivers/gpu/drm/msm/msm_atomic.c b/drivers/gpu/drm/msm/msm_atomic.c
index 94f9c3e..95c7868 100644
--- a/drivers/gpu/drm/msm/msm_atomic.c
+++ b/drivers/gpu/drm/msm/msm_atomic.c
@@ -21,66 +21,6 @@
 #include "msm_gem.h"
 #include "msm_fence.h"
 
-struct msm_commit {
-	struct drm_device *dev;
-	struct drm_atomic_state *state;
-	struct work_struct work;
-	uint32_t crtc_mask;
-};
-
-static void commit_worker(struct work_struct *work);
-
-/* block until specified crtcs are no longer pending update, and
- * atomically mark them as pending update
- */
-static int start_atomic(struct msm_drm_private *priv, uint32_t crtc_mask)
-{
-	int ret;
-
-	spin_lock(&priv->pending_crtcs_event.lock);
-	ret = wait_event_interruptible_locked(priv->pending_crtcs_event,
-			!(priv->pending_crtcs & crtc_mask));
-	if (ret == 0) {
-		DBG("start: %08x", crtc_mask);
-		priv->pending_crtcs |= crtc_mask;
-	}
-	spin_unlock(&priv->pending_crtcs_event.lock);
-
-	return ret;
-}
-
-/* clear specified crtcs (no longer pending update)
- */
-static void end_atomic(struct msm_drm_private *priv, uint32_t crtc_mask)
-{
-	spin_lock(&priv->pending_crtcs_event.lock);
-	DBG("end: %08x", crtc_mask);
-	priv->pending_crtcs &= ~crtc_mask;
-	wake_up_all_locked(&priv->pending_crtcs_event);
-	spin_unlock(&priv->pending_crtcs_event.lock);
-}
-
-static struct msm_commit *commit_init(struct drm_atomic_state *state)
-{
-	struct msm_commit *c = kzalloc(sizeof(*c), GFP_KERNEL);
-
-	if (!c)
-		return NULL;
-
-	c->dev = state->dev;
-	c->state = state;
-
-	INIT_WORK(&c->work, commit_worker);
-
-	return c;
-}
-
-static void commit_destroy(struct msm_commit *c)
-{
-	end_atomic(c->dev->dev_private, c->crtc_mask);
-	kfree(c);
-}
-
 static void msm_atomic_wait_for_commit_done(struct drm_device *dev,
 		struct drm_atomic_state *old_state)
 {
@@ -148,31 +88,37 @@ static void msm_atomic_commit_tail(struct drm_atomic_state *state)
 
 	msm_atomic_wait_for_commit_done(dev, state);
 
-	drm_atomic_helper_cleanup_planes(dev, state);
-
 	kms->funcs->complete_commit(kms, state);
+
+	drm_atomic_helper_wait_for_vblanks(dev, state);
+
+	drm_atomic_helper_commit_hw_done(state);
+
+	drm_atomic_helper_cleanup_planes(dev, state);
 }
 
 /* The (potentially) asynchronous part of the commit.  At this point
  * nothing can fail short of armageddon.
  */
-static void complete_commit(struct msm_commit *c)
+static void commit_tail(struct drm_atomic_state *state)
 {
-	struct drm_atomic_state *state = c->state;
-	struct drm_device *dev = state->dev;
+	drm_atomic_helper_wait_for_fences(state->dev, state, false);
 
-	drm_atomic_helper_wait_for_fences(dev, state, false);
+	drm_atomic_helper_wait_for_dependencies(state);
 
 	msm_atomic_commit_tail(state);
 
-	drm_atomic_state_put(state);
+	drm_atomic_helper_commit_cleanup_done(state);
 
-	commit_destroy(c);
+	drm_atomic_state_put(state);
 }
 
-static void commit_worker(struct work_struct *work)
+static void commit_work(struct work_struct *work)
 {
-	complete_commit(container_of(work, struct msm_commit, work));
+	struct drm_atomic_state *state = container_of(work,
+						      struct drm_atomic_state,
+						      commit_work);
+	commit_tail(state);
 }
 
 /**
@@ -191,17 +137,12 @@ int msm_atomic_commit(struct drm_device *dev,
 		struct drm_atomic_state *state, bool nonblock)
 {
 	struct msm_drm_private *priv = dev->dev_private;
-	struct msm_commit *c;
 	struct drm_crtc *crtc;
 	struct drm_crtc_state *crtc_state;
 	struct drm_plane *plane;
 	struct drm_plane_state *old_plane_state, *new_plane_state;
 	int i, ret;
 
-	ret = drm_atomic_helper_prepare_planes(dev, state);
-	if (ret)
-		return ret;
-
 	/*
 	 * Note that plane->atomic_async_check() should fail if we need
 	 * to re-assign hwpipe or anything that touches global atomic
@@ -209,45 +150,39 @@ int msm_atomic_commit(struct drm_device *dev,
 	 * cases.
 	 */
 	if (state->async_update) {
+		ret = drm_atomic_helper_prepare_planes(dev, state);
+		if (ret)
+			return ret;
+
 		drm_atomic_helper_async_commit(dev, state);
 		drm_atomic_helper_cleanup_planes(dev, state);
 		return 0;
 	}
 
-	c = commit_init(state);
-	if (!c) {
-		ret = -ENOMEM;
-		goto error;
-	}
+	ret = drm_atomic_helper_setup_commit(state, nonblock);
+	if (ret)
+		return ret;
 
-	/*
-	 * Figure out what crtcs we have:
-	 */
-	for_each_new_crtc_in_state(state, crtc, crtc_state, i)
-		c->crtc_mask |= drm_crtc_mask(crtc);
+	INIT_WORK(&state->commit_work, commit_work);
 
-	/*
-	 * Figure out what fence to wait for:
-	 */
-	for_each_oldnew_plane_in_state(state, plane, old_plane_state, new_plane_state, i) {
-		if ((new_plane_state->fb != old_plane_state->fb) && new_plane_state->fb) {
-			struct drm_gem_object *obj = msm_framebuffer_bo(new_plane_state->fb, 0);
-			struct msm_gem_object *msm_obj = to_msm_bo(obj);
-			struct dma_fence *fence = reservation_object_get_excl_rcu(msm_obj->resv);
-
-			drm_atomic_set_fence_for_plane(new_plane_state, fence);
-		}
+	ret = drm_atomic_helper_prepare_planes(dev, state);
+	if (ret)
+		return ret;
+
+	if (!nonblock) {
+		ret = drm_atomic_helper_wait_for_fences(dev, state, true);
+		if (ret)
+			goto error;
 	}
 
 	/*
-	 * Wait for pending updates on any of the same crtc's and then
-	 * mark our set of crtc's as busy:
+	 * This is the point of no return - everything below never fails except
+	 * when the hw goes bonghits. Which means we can commit the new state on
+	 * the software side now.
+	 *
+	 * swap driver private state while still holding state_lock
 	 */
-	ret = start_atomic(dev->dev_private, c->crtc_mask);
-	if (ret)
-		goto err_free;
-
-	BUG_ON(drm_atomic_helper_swap_state(state, false) < 0);
+	BUG_ON(drm_atomic_helper_swap_state(state, true) < 0);
 
 	/*
 	 * This is the point of no return - everything below never fails except
@@ -272,17 +207,13 @@ int msm_atomic_commit(struct drm_device *dev,
 	 */
 
 	drm_atomic_state_get(state);
-	if (nonblock) {
-		queue_work(priv->atomic_wq, &c->work);
-		return 0;
-	}
-
-	complete_commit(c);
+	if (nonblock)
+		queue_work(system_unbound_wq, &state->commit_work);
+	else
+		commit_tail(state);
 
 	return 0;
 
-err_free:
-	kfree(c);
 error:
 	drm_atomic_helper_cleanup_planes(dev, state);
 	return ret;
diff --git a/drivers/gpu/drm/msm/msm_drv.c b/drivers/gpu/drm/msm/msm_drv.c
index 1c89195..9cec74c 100644
--- a/drivers/gpu/drm/msm/msm_drv.c
+++ b/drivers/gpu/drm/msm/msm_drv.c
@@ -381,7 +381,6 @@ static int msm_drm_init(struct device *dev, struct drm_driver *drv)
 
 	priv->wq = alloc_ordered_workqueue("msm", 0);
 	priv->atomic_wq = alloc_ordered_workqueue("msm:atomic", 0);
-	init_waitqueue_head(&priv->pending_crtcs_event);
 
 	INIT_LIST_HEAD(&priv->inactive_list);
 	INIT_LIST_HEAD(&priv->vblank_ctrl.event_list);
diff --git a/drivers/gpu/drm/msm/msm_drv.h b/drivers/gpu/drm/msm/msm_drv.h
index 98e8223..2b26888 100644
--- a/drivers/gpu/drm/msm/msm_drv.h
+++ b/drivers/gpu/drm/msm/msm_drv.h
@@ -117,10 +117,6 @@ struct msm_drm_private {
 	struct workqueue_struct *wq;
 	struct workqueue_struct *atomic_wq;
 
-	/* crtcs pending async atomic updates: */
-	uint32_t pending_crtcs;
-	wait_queue_head_t pending_crtcs_event;
-
 	unsigned int num_planes;
 	struct drm_plane *planes[16];
 
-- 
cgit v1.1


From d14659f5de7d2822764d6944ce7d8d7570ebfd9b Mon Sep 17 00:00:00 2001
From: Sean Paul <seanpaul@chromium.org>
Date: Wed, 28 Feb 2018 14:19:05 -0500
Subject: drm/msm: Switch to atomic_helper_commit()

Now that all of the msm-specific goo is tucked safely away we can switch
over to using the atomic helper commit directly. \o/

Changes in v2:
- None
Changes in v3:
- Rebased on Archit's private_obj set
Changes in v4:
- None

Cc: Abhinav Kumar <abhinavk@codeaurora.org>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
---
 drivers/gpu/drm/msm/msm_atomic.c | 139 +--------------------------------------
 drivers/gpu/drm/msm/msm_drv.c    |   7 +-
 drivers/gpu/drm/msm/msm_drv.h    |   3 +-
 3 files changed, 8 insertions(+), 141 deletions(-)

diff --git a/drivers/gpu/drm/msm/msm_atomic.c b/drivers/gpu/drm/msm/msm_atomic.c
index 95c7868..f0635c3 100644
--- a/drivers/gpu/drm/msm/msm_atomic.c
+++ b/drivers/gpu/drm/msm/msm_atomic.c
@@ -18,8 +18,6 @@
 #include "msm_drv.h"
 #include "msm_gem.h"
 #include "msm_kms.h"
-#include "msm_gem.h"
-#include "msm_fence.h"
 
 static void msm_atomic_wait_for_commit_done(struct drm_device *dev,
 		struct drm_atomic_state *old_state)
@@ -59,7 +57,7 @@ int msm_atomic_prepare_fb(struct drm_plane *plane,
 	return msm_framebuffer_prepare(new_state->fb, kms->aspace);
 }
 
-static void msm_atomic_commit_tail(struct drm_atomic_state *state)
+void msm_atomic_commit_tail(struct drm_atomic_state *state)
 {
 	struct drm_device *dev = state->dev;
 	struct msm_drm_private *priv = dev->dev_private;
@@ -73,19 +71,6 @@ static void msm_atomic_commit_tail(struct drm_atomic_state *state)
 
 	drm_atomic_helper_commit_modeset_enables(dev, state);
 
-	/* NOTE: _wait_for_vblanks() only waits for vblank on
-	 * enabled CRTCs.  So we end up faulting when disabling
-	 * due to (potentially) unref'ing the outgoing fb's
-	 * before the vblank when the disable has latched.
-	 *
-	 * But if it did wait on disabled (or newly disabled)
-	 * CRTCs, that would be racy (ie. we could have missed
-	 * the irq.  We need some way to poll for pipe shut
-	 * down.  Or just live with occasionally hitting the
-	 * timeout in the CRTC disable path (which really should
-	 * not be critical path)
-	 */
-
 	msm_atomic_wait_for_commit_done(dev, state);
 
 	kms->funcs->complete_commit(kms, state);
@@ -96,125 +81,3 @@ static void msm_atomic_commit_tail(struct drm_atomic_state *state)
 
 	drm_atomic_helper_cleanup_planes(dev, state);
 }
-
-/* The (potentially) asynchronous part of the commit.  At this point
- * nothing can fail short of armageddon.
- */
-static void commit_tail(struct drm_atomic_state *state)
-{
-	drm_atomic_helper_wait_for_fences(state->dev, state, false);
-
-	drm_atomic_helper_wait_for_dependencies(state);
-
-	msm_atomic_commit_tail(state);
-
-	drm_atomic_helper_commit_cleanup_done(state);
-
-	drm_atomic_state_put(state);
-}
-
-static void commit_work(struct work_struct *work)
-{
-	struct drm_atomic_state *state = container_of(work,
-						      struct drm_atomic_state,
-						      commit_work);
-	commit_tail(state);
-}
-
-/**
- * drm_atomic_helper_commit - commit validated state object
- * @dev: DRM device
- * @state: the driver state object
- * @nonblock: nonblocking commit
- *
- * This function commits a with drm_atomic_helper_check() pre-validated state
- * object. This can still fail when e.g. the framebuffer reservation fails.
- *
- * RETURNS
- * Zero for success or -errno.
- */
-int msm_atomic_commit(struct drm_device *dev,
-		struct drm_atomic_state *state, bool nonblock)
-{
-	struct msm_drm_private *priv = dev->dev_private;
-	struct drm_crtc *crtc;
-	struct drm_crtc_state *crtc_state;
-	struct drm_plane *plane;
-	struct drm_plane_state *old_plane_state, *new_plane_state;
-	int i, ret;
-
-	/*
-	 * Note that plane->atomic_async_check() should fail if we need
-	 * to re-assign hwpipe or anything that touches global atomic
-	 * state, so we'll never go down the async update path in those
-	 * cases.
-	 */
-	if (state->async_update) {
-		ret = drm_atomic_helper_prepare_planes(dev, state);
-		if (ret)
-			return ret;
-
-		drm_atomic_helper_async_commit(dev, state);
-		drm_atomic_helper_cleanup_planes(dev, state);
-		return 0;
-	}
-
-	ret = drm_atomic_helper_setup_commit(state, nonblock);
-	if (ret)
-		return ret;
-
-	INIT_WORK(&state->commit_work, commit_work);
-
-	ret = drm_atomic_helper_prepare_planes(dev, state);
-	if (ret)
-		return ret;
-
-	if (!nonblock) {
-		ret = drm_atomic_helper_wait_for_fences(dev, state, true);
-		if (ret)
-			goto error;
-	}
-
-	/*
-	 * This is the point of no return - everything below never fails except
-	 * when the hw goes bonghits. Which means we can commit the new state on
-	 * the software side now.
-	 *
-	 * swap driver private state while still holding state_lock
-	 */
-	BUG_ON(drm_atomic_helper_swap_state(state, true) < 0);
-
-	/*
-	 * This is the point of no return - everything below never fails except
-	 * when the hw goes bonghits. Which means we can commit the new state on
-	 * the software side now.
-	 */
-
-	/*
-	 * Everything below can be run asynchronously without the need to grab
-	 * any modeset locks at all under one conditions: It must be guaranteed
-	 * that the asynchronous work has either been cancelled (if the driver
-	 * supports it, which at least requires that the framebuffers get
-	 * cleaned up with drm_atomic_helper_cleanup_planes()) or completed
-	 * before the new state gets committed on the software side with
-	 * drm_atomic_helper_swap_state().
-	 *
-	 * This scheme allows new atomic state updates to be prepared and
-	 * checked in parallel to the asynchronous completion of the previous
-	 * update. Which is important since compositors need to figure out the
-	 * composition of the next frame right after having submitted the
-	 * current layout.
-	 */
-
-	drm_atomic_state_get(state);
-	if (nonblock)
-		queue_work(system_unbound_wq, &state->commit_work);
-	else
-		commit_tail(state);
-
-	return 0;
-
-error:
-	drm_atomic_helper_cleanup_planes(dev, state);
-	return ret;
-}
diff --git a/drivers/gpu/drm/msm/msm_drv.c b/drivers/gpu/drm/msm/msm_drv.c
index 9cec74c..021a0b6 100644
--- a/drivers/gpu/drm/msm/msm_drv.c
+++ b/drivers/gpu/drm/msm/msm_drv.c
@@ -41,7 +41,11 @@ static const struct drm_mode_config_funcs mode_config_funcs = {
 	.fb_create = msm_framebuffer_create,
 	.output_poll_changed = drm_fb_helper_output_poll_changed,
 	.atomic_check = drm_atomic_helper_check,
-	.atomic_commit = msm_atomic_commit,
+	.atomic_commit = drm_atomic_helper_commit,
+};
+
+static const struct drm_mode_config_helper_funcs mode_config_helper_funcs = {
+	.atomic_commit_tail = msm_atomic_commit_tail,
 };
 
 #ifdef CONFIG_DRM_MSM_REGISTER_LOGGING
@@ -438,6 +442,7 @@ static int msm_drm_init(struct device *dev, struct drm_driver *drv)
 	}
 
 	ddev->mode_config.funcs = &mode_config_funcs;
+	ddev->mode_config.helper_private = &mode_config_helper_funcs;
 
 	ret = drm_vblank_init(ddev, priv->num_crtcs);
 	if (ret < 0) {
diff --git a/drivers/gpu/drm/msm/msm_drv.h b/drivers/gpu/drm/msm/msm_drv.h
index 2b26888..b2da1fb 100644
--- a/drivers/gpu/drm/msm/msm_drv.h
+++ b/drivers/gpu/drm/msm/msm_drv.h
@@ -158,8 +158,7 @@ struct msm_format {
 
 int msm_atomic_prepare_fb(struct drm_plane *plane,
 			  struct drm_plane_state *new_state);
-int msm_atomic_commit(struct drm_device *dev,
-		struct drm_atomic_state *state, bool nonblock);
+void msm_atomic_commit_tail(struct drm_atomic_state *state);
 struct drm_atomic_state *msm_atomic_state_alloc(struct drm_device *dev);
 void msm_atomic_state_clear(struct drm_atomic_state *state);
 void msm_atomic_state_free(struct drm_atomic_state *state);
-- 
cgit v1.1


From 74d3a3a70775de356b96b5461c3a204a51496fb3 Mon Sep 17 00:00:00 2001
From: Sean Paul <seanpaul@chromium.org>
Date: Thu, 31 May 2018 14:48:58 -0400
Subject: drm/msm: Fix NULL deref on bind/probe deferral

This patch avoids dereferencing msm_host->dev when it is NULL.

If we find ourselves tearing down dsi before calling
(mdp4|mdp5|dpu)_kms_init(), we'll end up in a state where the dev
pointer is NULL and trying to extract priv from it will fail.

This was introduced in a seemingly innocuous commit to ensure the
arguments to msm_gem_put_iova() are correct (even though that
function has been a stub for ~5 years). Correctness FTW! \o/

Fixes: b01884a286b0 drm/msm: use correct aspace pointer in msm_gem_put_iova()
Cc: Daniel Mack <daniel@zonque.org>
Cc: Rob Clark <robdclark@gmail.com>
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
---
 drivers/gpu/drm/msm/dsi/dsi_host.c | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/msm/dsi/dsi_host.c b/drivers/gpu/drm/msm/dsi/dsi_host.c
index b916f46..2f1a278 100644
--- a/drivers/gpu/drm/msm/dsi/dsi_host.c
+++ b/drivers/gpu/drm/msm/dsi/dsi_host.c
@@ -1066,8 +1066,18 @@ static int dsi_tx_buf_alloc(struct msm_dsi_host *msm_host, int size)
 static void dsi_tx_buf_free(struct msm_dsi_host *msm_host)
 {
 	struct drm_device *dev = msm_host->dev;
-	struct msm_drm_private *priv = dev->dev_private;
+	struct msm_drm_private *priv;
+
+	/*
+	 * This is possible if we're tearing down before we've had a chance to
+	 * fully initialize. A very real possibility if our probe is deferred,
+	 * in which case we'll hit msm_dsi_host_destroy() without having run
+	 * through the dsi_tx_buf_alloc().
+	 */
+	if (!dev)
+		return;
 
+	priv = dev->dev_private;
 	if (msm_host->tx_gem_obj) {
 		msm_gem_put_iova(msm_host->tx_gem_obj, priv->kms->aspace);
 		drm_gem_object_put_unlocked(msm_host->tx_gem_obj);
-- 
cgit v1.1


From 47541443a62a278f0a0f9a03c5403cb15bd62ded Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= <ville.syrjala@linux.intel.com>
Date: Mon, 21 May 2018 21:56:12 +0300
Subject: drm/i915: Remove bogus NV12 PLANE_COLOR_CTL setup
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

We already handle the color encoding mode properly. Remove the broken
NV12 special case.

Cc: Vidya Srinivas <vidya.srinivas@intel.com>
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Fixes: 8ed30ab6aced ("drm/i915: Enable YUV to RGB for Gen10 in Plane Ctrl Reg")
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180521185613.5097-1-ville.syrjala@linux.intel.com
Reviewed-By: Vidya Srinivas <vidya.srinivas@intel.com>
(cherry picked from commit 012d79e6a33f095c293fe2a02b2b3b26d8c6402c)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
---
 drivers/gpu/drm/i915/intel_display.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index ad588d5..dee3a8e 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -3690,11 +3690,6 @@ u32 glk_plane_color_ctl(const struct intel_crtc_state *crtc_state,
 	plane_color_ctl |= glk_plane_color_ctl_alpha(fb->format->format);
 
 	if (intel_format_is_yuv(fb->format->format)) {
-		if (fb->format->format == DRM_FORMAT_NV12) {
-			plane_color_ctl |=
-				PLANE_COLOR_CSC_MODE_YUV709_TO_RGB709;
-			goto out;
-		}
 		if (plane_state->base.color_encoding == DRM_COLOR_YCBCR_BT709)
 			plane_color_ctl |= PLANE_COLOR_CSC_MODE_YUV709_TO_RGB709;
 		else
@@ -3703,7 +3698,7 @@ u32 glk_plane_color_ctl(const struct intel_crtc_state *crtc_state,
 		if (plane_state->base.color_range == DRM_COLOR_YCBCR_FULL_RANGE)
 			plane_color_ctl |= PLANE_COLOR_YUV_RANGE_CORRECTION_DISABLE;
 	}
-out:
+
 	return plane_color_ctl;
 }
 
-- 
cgit v1.1


From 2f08b23d70324bb0f592c71bd36cdc1d2f12face Mon Sep 17 00:00:00 2001
From: Mahesh Kumar <mahesh1.kumar@intel.com>
Date: Fri, 25 May 2018 08:52:38 -0700
Subject: drm/i915/icl: fix icl_unmap/map_plls_to_ports

All connectors may not have best_encoder attached, so don't dereference
encoder pointer for each connector.

Fixes: c27e917e2bda ("drm/i915/icl: add basic support for the ICL clocks")
Reviewed-by: Lucas De Marchi <lucas.demarchi@intel.com>
Signed-off-by: Mahesh Kumar <mahesh1.kumar@intel.com>
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180525155238.7054-1-lucas.demarchi@intel.com
(cherry picked from commit c46ef57d2008e5195f086a538550488b49644db7)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
---
 drivers/gpu/drm/i915/intel_ddi.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c
index b98ac054..f4a8598 100644
--- a/drivers/gpu/drm/i915/intel_ddi.c
+++ b/drivers/gpu/drm/i915/intel_ddi.c
@@ -2453,12 +2453,13 @@ void icl_map_plls_to_ports(struct drm_crtc *crtc,
 	for_each_new_connector_in_state(old_state, conn, conn_state, i) {
 		struct intel_encoder *encoder =
 			to_intel_encoder(conn_state->best_encoder);
-		enum port port = encoder->port;
+		enum port port;
 		uint32_t val;
 
 		if (conn_state->crtc != crtc)
 			continue;
 
+		port = encoder->port;
 		mutex_lock(&dev_priv->dpll_lock);
 
 		val = I915_READ(DPCLKA_CFGCR0_ICL);
@@ -2490,11 +2491,12 @@ void icl_unmap_plls_to_ports(struct drm_crtc *crtc,
 	for_each_old_connector_in_state(old_state, conn, old_conn_state, i) {
 		struct intel_encoder *encoder =
 			to_intel_encoder(old_conn_state->best_encoder);
-		enum port port = encoder->port;
+		enum port port;
 
 		if (old_conn_state->crtc != crtc)
 			continue;
 
+		port = encoder->port;
 		mutex_lock(&dev_priv->dpll_lock);
 		I915_WRITE(DPCLKA_CFGCR0_ICL,
 			   I915_READ(DPCLKA_CFGCR0_ICL) |
-- 
cgit v1.1


From 197af5f2131101f9a6118b238901cb1988f5d7f9 Mon Sep 17 00:00:00 2001
From: Mahesh Kumar <mahesh1.kumar@intel.com>
Date: Thu, 17 May 2018 18:56:26 +0530
Subject: drm/i915/icl: Don't update enabled dbuf slices struct until updated
 in hw

Do not update number of enabled dbuf slices in dev_priv struct until we
actually enable/disable dbuf slice in hw. This is leading to never
updating dbuf slices and resulting in DBuf slice mismatch warning.

Fixes: aa9664ffe863 ("drm/i915/icl: Enable 2nd DBuf slice only when needed")
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Mahesh Kumar <mahesh1.kumar@intel.com>
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180517132626.5885-1-mahesh1.kumar@intel.com
(cherry picked from commit 6ceb7277173597eeed8d635e08db51e35235ab21)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
---
 drivers/gpu/drm/i915/intel_pm.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index b85229e..53aaaa3 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -5150,7 +5150,6 @@ skl_copy_ddb_for_pipe(struct skl_ddb_values *dst,
 	       sizeof(dst->ddb.uv_plane[pipe]));
 	memcpy(dst->ddb.plane[pipe], src->ddb.plane[pipe],
 	       sizeof(dst->ddb.plane[pipe]));
-	dst->ddb.enabled_slices = src->ddb.enabled_slices;
 }
 
 static void
-- 
cgit v1.1


From a0b2ac29415bb44d1c212184c1385a1abe68db5c Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Wed, 23 May 2018 11:18:43 +0800
Subject: drm/amdgpu: fix the missed vcn fw version report
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

It missed vcn.fw_version setting when init vcn microcode, and it will be used to
report vcn ucode version via amdgpu_firmware_info sysfs interface.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
index 6fd606f..127e87b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
@@ -82,6 +82,7 @@ int amdgpu_vcn_sw_init(struct amdgpu_device *adev)
 	}
 
 	hdr = (const struct common_firmware_header *)adev->vcn.fw->data;
+	adev->vcn.fw_version = le32_to_cpu(hdr->ucode_version);
 	family_id = le32_to_cpu(hdr->ucode_version) & 0xff;
 	version_major = (le32_to_cpu(hdr->ucode_version) >> 24) & 0xff;
 	version_minor = (le32_to_cpu(hdr->ucode_version) >> 8) & 0xff;
-- 
cgit v1.1


From 235293901c11705f94744c08582e1ff339ee29b2 Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Sat, 12 May 2018 12:31:12 +0800
Subject: drm/amdgpu: add checking for sos version

The sos ucode version will be changed to align with the value of
mmMP0_SMN_C2PMSG_58. Then we add a checking for this. Meanwhile, we have to be
compatibility backwards. So it adds serveral recent legacy versions as the white
list for the version checking.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/psp_v3_1.c | 27 ++++++++++++++++++++++++++-
 1 file changed, 26 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c
index 0c768e3..727071f 100644
--- a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c
+++ b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c
@@ -47,6 +47,8 @@ MODULE_FIRMWARE("amdgpu/vega20_asd.bin");
 
 #define smnMP1_FIRMWARE_FLAGS 0x3010028
 
+static uint32_t sos_old_versions[] = {1517616, 1510592, 1448594, 1446554};
+
 static int
 psp_v3_1_get_fw_type(struct amdgpu_firmware_info *ucode, enum psp_gfx_fw_type *type)
 {
@@ -210,12 +212,31 @@ static int psp_v3_1_bootloader_load_sysdrv(struct psp_context *psp)
 	return ret;
 }
 
+static bool psp_v3_1_match_version(struct amdgpu_device *adev, uint32_t ver)
+{
+	int i;
+
+	if (ver == adev->psp.sos_fw_version)
+		return true;
+
+	/*
+	 * Double check if the latest four legacy versions.
+	 * If yes, it is still the right version.
+	 */
+	for (i = 0; i < sizeof(sos_old_versions) / sizeof(uint32_t); i++) {
+		if (sos_old_versions[i] == adev->psp.sos_fw_version)
+			return true;
+	}
+
+	return false;
+}
+
 static int psp_v3_1_bootloader_load_sos(struct psp_context *psp)
 {
 	int ret;
 	unsigned int psp_gfxdrv_command_reg = 0;
 	struct amdgpu_device *adev = psp->adev;
-	uint32_t sol_reg;
+	uint32_t sol_reg, ver;
 
 	/* Check sOS sign of life register to confirm sys driver and sOS
 	 * are already been loaded.
@@ -248,6 +269,10 @@ static int psp_v3_1_bootloader_load_sos(struct psp_context *psp)
 			   RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_81),
 			   0, true);
 
+	ver = RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_58);
+	if (!psp_v3_1_match_version(adev, ver))
+		DRM_WARN("SOS version doesn't match\n");
+
 	return ret;
 }
 
-- 
cgit v1.1


From 387f49e5467244b7bcb4cad0946a5d0fcade5f92 Mon Sep 17 00:00:00 2001
From: Junwei Zhang <Jerry.Zhang@amd.com>
Date: Tue, 5 Jun 2018 17:31:51 +0800
Subject: drm/amdgpu: fix clear_all and replace handling in the VM (v2)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

v2: assign bo_va as well

We need to put the lose ends on the invalid list because it is possible
that we need to split up huge pages for them.

Cc: stable@vger.kernel.org
Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Junwei Zhang <Jerry.Zhang@amd.com> (v2)
Reviewed-by: David Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index ccba88c..b0eb2f5 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -2123,7 +2123,8 @@ int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev,
 			before->last = saddr - 1;
 			before->offset = tmp->offset;
 			before->flags = tmp->flags;
-			list_add(&before->list, &tmp->list);
+			before->bo_va = tmp->bo_va;
+			list_add(&before->list, &tmp->bo_va->invalids);
 		}
 
 		/* Remember mapping split at the end */
@@ -2133,7 +2134,8 @@ int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev,
 			after->offset = tmp->offset;
 			after->offset += after->start - tmp->start;
 			after->flags = tmp->flags;
-			list_add(&after->list, &tmp->list);
+			after->bo_va = tmp->bo_va;
+			list_add(&after->list, &tmp->bo_va->invalids);
 		}
 
 		list_del(&tmp->list);
-- 
cgit v1.1


From 06b18f61ee78f8c69417c3a5e4f21ed678662315 Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Fri, 1 Jun 2018 14:41:04 +0800
Subject: drm/amdgpu: fix CG enabling hang with gfxoff enabled
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

After defer the execution of clockgating enabling, at that time, gfx already
enter into "off" state. Howerver, clockgating enabling will use MMIO to access
the gfx registers, then get the gfx hung.

So here we should move the gfx powergating and gfxoff enabling behavior at the
end of initialization behind clockgating.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Cc: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c        | 12 ++++++++++++
 drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c             |  5 -----
 drivers/gpu/drm/amd/powerplay/amd_powerplay.c     |  2 +-
 drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c |  4 ++--
 4 files changed, 15 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 290e279..3317d15 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -1730,6 +1730,18 @@ static int amdgpu_device_ip_late_set_cg_state(struct amdgpu_device *adev)
 			}
 		}
 	}
+
+	if (adev->powerplay.pp_feature & PP_GFXOFF_MASK) {
+		/* enable gfx powergating */
+		amdgpu_device_ip_set_powergating_state(adev,
+						       AMD_IP_BLOCK_TYPE_GFX,
+						       AMD_PG_STATE_GATE);
+		/* enable gfxoff */
+		amdgpu_device_ip_set_powergating_state(adev,
+						       AMD_IP_BLOCK_TYPE_SMC,
+						       AMD_PG_STATE_GATE);
+	}
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index 4f7a72d..95f2773 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -3405,11 +3405,6 @@ static int gfx_v9_0_late_init(void *handle)
 	if (r)
 		return r;
 
-	r = amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_GFX,
-						   AMD_PG_STATE_GATE);
-	if (r)
-		return r;
-
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
index b493369..d0e6e2d 100644
--- a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
+++ b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
@@ -245,7 +245,7 @@ static int pp_set_powergating_state(void *handle,
 	}
 
 	if (hwmgr->hwmgr_func->enable_per_cu_power_gating == NULL) {
-		pr_info("%s was not implemented.\n", __func__);
+		pr_debug("%s was not implemented.\n", __func__);
 		return 0;
 	}
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
index 6a63671..d4bc83e 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
@@ -313,7 +313,7 @@ static int smu10_disable_gfx_off(struct pp_hwmgr *hwmgr)
 
 static int smu10_disable_dpm_tasks(struct pp_hwmgr *hwmgr)
 {
-	return smu10_disable_gfx_off(hwmgr);
+	return 0;
 }
 
 static int smu10_enable_gfx_off(struct pp_hwmgr *hwmgr)
@@ -328,7 +328,7 @@ static int smu10_enable_gfx_off(struct pp_hwmgr *hwmgr)
 
 static int smu10_enable_dpm_tasks(struct pp_hwmgr *hwmgr)
 {
-	return smu10_enable_gfx_off(hwmgr);
+	return 0;
 }
 
 static int smu10_gfx_off_control(struct pp_hwmgr *hwmgr, bool enable)
-- 
cgit v1.1


From 7584498c1c67bba4d1f5a0f7bf48c9a50c3ff5c8 Mon Sep 17 00:00:00 2001
From: Huang Rui <ray.huang@amd.com>
Date: Fri, 18 May 2018 10:39:16 +0800
Subject: drm/amd/powerplay: fix missed hwmgr check warning before call
 gfx_off_control handler

Patch 9667849bbb8d: "drm/amd/powerplay: add control gfxoff enabling in late
init" from Mar 13, 2018, leads to the following static checker warning:

	drivers/gpu/drm/amd/amdgpu/../powerplay/amd_powerplay.c:194
pp_late_init()
	error: we previously assumed 'hwmgr' could be null (see line 185)

drivers/gpu/drm/amd/amdgpu/../powerplay/amd_powerplay.c

This patch fixes the warning to add hwmgr checking.

Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/amd_powerplay.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
index d0e6e2d..0969b65 100644
--- a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
+++ b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
@@ -191,7 +191,8 @@ static int pp_late_init(void *handle)
 	if (adev->pm.smu_prv_buffer_size != 0)
 		pp_reserve_vram_for_smu(adev);
 
-	if (hwmgr->hwmgr_func->gfx_off_control &&
+	if (hwmgr && hwmgr->hwmgr_func &&
+	    hwmgr->hwmgr_func->gfx_off_control &&
 	    (hwmgr->feature_mask & PP_GFXOFF_MASK)) {
 		ret = hwmgr->hwmgr_func->gfx_off_control(hwmgr, true);
 		if (ret)
-- 
cgit v1.1


From 97028037a38ae40c0e06789b71038d3a6045a413 Mon Sep 17 00:00:00 2001
From: Lyude Paul <lyude@redhat.com>
Date: Mon, 4 Jun 2018 15:35:03 -0400
Subject: drm/amdgpu: Grab/put runtime PM references in atomic_commit_tail()
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

So, unfortunately I recently made the discovery that in the upstream
kernel, the only reason that amdgpu is not currently suffering from
issues with runtime PM putting the GPU into suspend while it's driving
displays is due to the fact that on most prime systems, we have sound
devices associated with the GPU that hold their own runtime PM ref for
the GPU.

What this means however, is that in the event that there isn't any kind
of sound device active (which can easily be reproduced by building a
kernel with sound drivers disabled), the GPU will fall asleep even when
there's displays active. This appears to be in part due to the fact that
amdgpu has not actually ever relied on it's rpm_idle() function to be
the only thing keeping it running, and normally grabs it's own power
references whenever there are displays active (as can be seen with the
original pre-DC codepath in amdgpu_display_crtc_set_config() in
amdgpu_display.c). This means it's very likely that this bug was
introduced during the switch over the DC.

So to fix this, we start grabbing runtime PM references every time we
enable a previously disabled CRTC in atomic_commit_tail(). This appears
to be the correct solution, as it matches up with what i915 does in
i915/intel_runtime_pm.c.

The one sideaffect of this is that we ignore the variable that the
pre-DC code used to use for tracking when it needed runtime PM refs,
adev->have_disp_power_ref. This is mainly because there's no way for a
driver to tell whether or not all of it's CRTCs are enabled or disabled
when we've begun committing an atomic state, as there may be CRTC
commits happening in parallel that aren't contained within the atomic
state being committed. So, it's safer to just get/put a reference for
each CRTC being enabled or disabled in the new atomic state.

Signed-off-by: Lyude Paul <lyude@redhat.com>
Acked-by: Christian König <christian.koenig@amd.com>.
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 0a06941..5e9e677 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -46,6 +46,7 @@
 #include <linux/moduleparam.h>
 #include <linux/version.h>
 #include <linux/types.h>
+#include <linux/pm_runtime.h>
 
 #include <drm/drmP.h>
 #include <drm/drm_atomic.h>
@@ -4278,6 +4279,8 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state)
 			if (dm_old_crtc_state->stream)
 				remove_stream(adev, acrtc, dm_old_crtc_state->stream);
 
+			pm_runtime_get_noresume(dev->dev);
+
 			acrtc->enabled = true;
 			acrtc->hw_mode = new_crtc_state->mode;
 			crtc->hwmode = new_crtc_state->mode;
@@ -4466,6 +4469,16 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state)
 		drm_atomic_helper_wait_for_flip_done(dev, state);
 
 	drm_atomic_helper_cleanup_planes(dev, state);
+
+	/* Finally, drop a runtime PM reference for each newly disabled CRTC,
+	 * so we can put the GPU into runtime suspend if we're not driving any
+	 * displays anymore
+	 */
+	pm_runtime_mark_last_busy(dev->dev);
+	for_each_oldnew_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) {
+		if (old_crtc_state->active && !new_crtc_state->active)
+			pm_runtime_put_autosuspend(dev->dev);
+	}
 }
 
 
-- 
cgit v1.1


From c3dade5ef72d9141e28fb86ebd46a9b3f3f4e030 Mon Sep 17 00:00:00 2001
From: Evan Quan <evan.quan@amd.com>
Date: Wed, 6 Jun 2018 11:54:45 +0800
Subject: drm/amd/powerplay: fix wrong clock adjust sequence

The clocks should be adjusted after display configuration changed.
Otherwise, the socclk and memclk may be forced on an unnecessary higher
level.

Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/pp_psm.c | 13 ++++++-------
 1 file changed, 6 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/pp_psm.c b/drivers/gpu/drm/amd/powerplay/hwmgr/pp_psm.c
index 0af13c1..323990b 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/pp_psm.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/pp_psm.c
@@ -265,19 +265,18 @@ int psm_adjust_power_state_dynamic(struct pp_hwmgr *hwmgr, bool skip,
 	if (skip)
 		return 0;
 
-	if (!hwmgr->ps)
-		/*
-		 * for vega12/vega20 which does not support power state manager
-		 * DAL clock limits should also be honoured
-		 */
-		phm_apply_clock_adjust_rules(hwmgr);
-
 	phm_pre_display_configuration_changed(hwmgr);
 
 	phm_display_configuration_changed(hwmgr);
 
 	if (hwmgr->ps)
 		power_state_management(hwmgr, new_ps);
+	else
+		/*
+		 * for vega12/vega20 which does not support power state manager
+		 * DAL clock limits should also be honoured
+		 */
+		phm_apply_clock_adjust_rules(hwmgr);
 
 	phm_notify_smc_display_config_after_ps_adjustment(hwmgr);
 
-- 
cgit v1.1


From c4ff91dd40e2253ab6dd028011469c2c694e1e19 Mon Sep 17 00:00:00 2001
From: Colin Ian King <colin.king@canonical.com>
Date: Wed, 6 Jun 2018 13:18:31 +0100
Subject: drm/amd/pp: initialize result to before or'ing in data

The current use of result is or'ing in values and checking for
a non-zero result, however, result is not initialized to zero
so it potentially contains garbage to start with. Fix this by
initializing it to the first return from the call to
vega10_program_didt_config_registers.

Detected by cppcheck:
"(error) Uninitialized variable: result"

Fixes: 9b7b8154cdb8 ("drm/amd/powerplay: added didt support for vega10")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Acked-by: Huang Rui <ray.huang@amd.com>
[Fix the subject as Colin's comment]
Signed-off-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
---
 drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c
index a9efd855..dbe4b1f 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c
@@ -1104,7 +1104,7 @@ static int vega10_enable_psm_gc_edc_config(struct pp_hwmgr *hwmgr)
 	for (count = 0; count < num_se; count++) {
 		data = GRBM_GFX_INDEX__INSTANCE_BROADCAST_WRITES_MASK | GRBM_GFX_INDEX__SH_BROADCAST_WRITES_MASK | ( count << GRBM_GFX_INDEX__SE_INDEX__SHIFT);
 		WREG32_SOC15(GC, 0, mmGRBM_GFX_INDEX, data);
-		result |= vega10_program_didt_config_registers(hwmgr, PSMSEEDCStallPatternConfig_Vega10, VEGA10_CONFIGREG_DIDT);
+		result = vega10_program_didt_config_registers(hwmgr, PSMSEEDCStallPatternConfig_Vega10, VEGA10_CONFIGREG_DIDT);
 		result |= vega10_program_didt_config_registers(hwmgr, PSMSEEDCStallDelayConfig_Vega10, VEGA10_CONFIGREG_DIDT);
 		result |= vega10_program_didt_config_registers(hwmgr, PSMSEEDCCtrlResetConfig_Vega10, VEGA10_CONFIGREG_DIDT);
 		result |= vega10_program_didt_config_registers(hwmgr, PSMSEEDCCtrlConfig_Vega10, VEGA10_CONFIGREG_DIDT);
-- 
cgit v1.1


From 4b3c641b5f3a1d67c7bec60fb033233504e636ee Mon Sep 17 00:00:00 2001
From: Pratik Vishwakarma <Pratik.Vishwakarma@amd.com>
Date: Thu, 7 Jun 2018 11:48:40 +0530
Subject: drm/amd/display: Fix stale buffer object (bo) use

Fixes stale buffer object (bo) usage for cursor plane

Cursor plane's bo operations are handled in DC code.
Currently, atomic_commit() does not handle bo operations
for cursor plane, as a result the bo assigned for cursor
plane in dm_plane_helper_prepare_fb() is not coherent
with the updates to the same made in dc code.This mismatch
leads to "bo" corruption and hence crashes during S3 entry.

This patch cleans up the code which was added as a hack
for 4.9 version only.

Reviewed-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Signed-off-by: Pratik Vishwakarma <Pratik.Vishwakarma@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 11 -----------
 1 file changed, 11 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 5e9e677..d7d1245 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -3081,17 +3081,6 @@ static int dm_plane_helper_prepare_fb(struct drm_plane *plane,
 		}
 	}
 
-	/* It's a hack for s3 since in 4.9 kernel filter out cursor buffer
-	 * prepare and cleanup in drm_atomic_helper_prepare_planes
-	 * and drm_atomic_helper_cleanup_planes because fb doens't in s3.
-	 * IN 4.10 kernel this code should be removed and amdgpu_device_suspend
-	 * code touching fram buffers should be avoided for DC.
-	 */
-	if (plane->type == DRM_PLANE_TYPE_CURSOR) {
-		struct amdgpu_crtc *acrtc = to_amdgpu_crtc(new_state->crtc);
-
-		acrtc->cursor_bo = obj;
-	}
 	return 0;
 }
 
-- 
cgit v1.1


From f8a5de447f1511917487b43dce96639c29b41219 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@amd.com>
Date: Tue, 12 Jun 2018 14:26:00 +0800
Subject: drm/amd/pp: Fix OD feature enable failed on Vega10 workstation cards

As hw required, soc clock must large than mclk, So we set max soc
clock to OD Max Memory clk.
But on workstation, vbios do not support OD feature, the OD max memory
clock is equal to 0. In this case, driver can support underclocking.
and set od max memory clock to the value in highest memory dpm level.
So the od max memory clock should be less than highest soc clock.
and driver should not change the soc clock.

caused by commit ca57b9b0a156
("drm/amd/pp: Allow underclocking when od table is empty in vbios")

Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
index f70dbc8..05e680d 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
@@ -321,8 +321,12 @@ static int vega10_odn_initial_default_setting(struct pp_hwmgr *hwmgr)
 		odn_table->min_vddc = dep_table[0]->entries[0].vddc;
 
 	i = od_table[2]->count - 1;
-	od_table[2]->entries[i].clk = hwmgr->platform_descriptor.overdriveLimit.memoryClock;
-	od_table[2]->entries[i].vddc = odn_table->max_vddc;
+	od_table[2]->entries[i].clk = hwmgr->platform_descriptor.overdriveLimit.memoryClock > od_table[2]->entries[i].clk ?
+					hwmgr->platform_descriptor.overdriveLimit.memoryClock :
+					od_table[2]->entries[i].clk;
+	od_table[2]->entries[i].vddc = odn_table->max_vddc > od_table[2]->entries[i].vddc ?
+					odn_table->max_vddc :
+					od_table[2]->entries[i].vddc;
 
 	return 0;
 }
@@ -1325,6 +1329,7 @@ static int vega10_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
 	if (hwmgr->platform_descriptor.overdriveLimit.memoryClock == 0)
 		hwmgr->platform_descriptor.overdriveLimit.memoryClock =
 					dpm_table->dpm_levels[dpm_table->count-1].value;
+
 	vega10_init_dpm_state(&(dpm_table->dpm_state));
 
 	data->dpm_table.eclk_table.count = 0;
-- 
cgit v1.1


From b0f6b8090e05a24263207a399b6c48a94034f1e8 Mon Sep 17 00:00:00 2001
From: Shaoyun Liu <Shaoyun.Liu@amd.com>
Date: Tue, 12 Jun 2018 13:35:44 -0400
Subject: drm/amd/include: Update df 3.6 mask and shift definition

The register field hsas been changed in df 3.6, update to correct setting

Signed-off-by: Shaoyun Liu <Shaoyun.Liu@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_sh_mask.h | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_sh_mask.h
index 88f7c69..06fac50 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_sh_mask.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_sh_mask.h
@@ -36,13 +36,13 @@
 /* DF_CS_AON0_DramBaseAddress0 */
 #define DF_CS_UMC_AON0_DramBaseAddress0__AddrRngVal__SHIFT						0x0
 #define DF_CS_UMC_AON0_DramBaseAddress0__LgcyMmioHoleEn__SHIFT						0x1
-#define DF_CS_UMC_AON0_DramBaseAddress0__IntLvNumChan__SHIFT						0x4
-#define DF_CS_UMC_AON0_DramBaseAddress0__IntLvAddrSel__SHIFT						0x8
+#define DF_CS_UMC_AON0_DramBaseAddress0__IntLvNumChan__SHIFT						0x2
+#define DF_CS_UMC_AON0_DramBaseAddress0__IntLvAddrSel__SHIFT						0x9
 #define DF_CS_UMC_AON0_DramBaseAddress0__DramBaseAddr__SHIFT						0xc
 #define DF_CS_UMC_AON0_DramBaseAddress0__AddrRngVal_MASK						0x00000001L
 #define DF_CS_UMC_AON0_DramBaseAddress0__LgcyMmioHoleEn_MASK						0x00000002L
-#define DF_CS_UMC_AON0_DramBaseAddress0__IntLvNumChan_MASK						0x000000F0L
-#define DF_CS_UMC_AON0_DramBaseAddress0__IntLvAddrSel_MASK						0x00000700L
+#define DF_CS_UMC_AON0_DramBaseAddress0__IntLvNumChan_MASK						0x0000003CL
+#define DF_CS_UMC_AON0_DramBaseAddress0__IntLvAddrSel_MASK						0x00000E00L
 #define DF_CS_UMC_AON0_DramBaseAddress0__DramBaseAddr_MASK						0xFFFFF000L
 
 #endif
-- 
cgit v1.1


From cb5ed37f1f9976a5f9d5f677ac9423642e30d10f Mon Sep 17 00:00:00 2001
From: Evan Quan <evan.quan@amd.com>
Date: Tue, 29 May 2018 16:31:05 +0800
Subject: drm/amdgpu: fix parsing indirect register list v2

WARN_ON possible buffer overflow and avoid unnecessary dereference.

v2: change BUG_ON to WARN_ON

Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 15 +++++++++------
 1 file changed, 9 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index 95f2773..a691534 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -1838,13 +1838,15 @@ static void gfx_v9_1_parse_ind_reg_list(int *register_list_format,
 				int indirect_offset,
 				int list_size,
 				int *unique_indirect_regs,
-				int *unique_indirect_reg_count,
+				int unique_indirect_reg_count,
 				int *indirect_start_offsets,
-				int *indirect_start_offsets_count)
+				int *indirect_start_offsets_count,
+				int max_start_offsets_count)
 {
 	int idx;
 
 	for (; indirect_offset < list_size; indirect_offset++) {
+		WARN_ON(*indirect_start_offsets_count >= max_start_offsets_count);
 		indirect_start_offsets[*indirect_start_offsets_count] = indirect_offset;
 		*indirect_start_offsets_count = *indirect_start_offsets_count + 1;
 
@@ -1852,14 +1854,14 @@ static void gfx_v9_1_parse_ind_reg_list(int *register_list_format,
 			indirect_offset += 2;
 
 			/* look for the matching indice */
-			for (idx = 0; idx < *unique_indirect_reg_count; idx++) {
+			for (idx = 0; idx < unique_indirect_reg_count; idx++) {
 				if (unique_indirect_regs[idx] ==
 					register_list_format[indirect_offset] ||
 					!unique_indirect_regs[idx])
 					break;
 			}
 
-			BUG_ON(idx >= *unique_indirect_reg_count);
+			BUG_ON(idx >= unique_indirect_reg_count);
 
 			if (!unique_indirect_regs[idx])
 				unique_indirect_regs[idx] = register_list_format[indirect_offset];
@@ -1894,9 +1896,10 @@ static int gfx_v9_1_init_rlc_save_restore_list(struct amdgpu_device *adev)
 				    adev->gfx.rlc.reg_list_format_direct_reg_list_length,
 				    adev->gfx.rlc.reg_list_format_size_bytes >> 2,
 				    unique_indirect_regs,
-				    &unique_indirect_reg_count,
+				    unique_indirect_reg_count,
 				    indirect_start_offsets,
-				    &indirect_start_offsets_count);
+				    &indirect_start_offsets_count,
+				    ARRAY_SIZE(indirect_start_offsets));
 
 	/* enable auto inc in case it is disabled */
 	tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_SRM_CNTL));
-- 
cgit v1.1


From 333c8d3ef238f21516659b5221532060bae8a128 Mon Sep 17 00:00:00 2001
From: Evan Quan <evan.quan@amd.com>
Date: Tue, 12 Jun 2018 17:01:23 +0800
Subject: drm/amd/powerplay: remove uncessary extra gfxoff control call

Gfxoff is already enabled in amdgpu_device_ip_set_powergating_state.
So, no need to enable it again in pp_late_init.

Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/amd_powerplay.c | 9 ---------
 1 file changed, 9 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
index 0969b65..d567be4 100644
--- a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
+++ b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
@@ -180,7 +180,6 @@ static int pp_late_init(void *handle)
 {
 	struct amdgpu_device *adev = handle;
 	struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle;
-	int ret;
 
 	if (hwmgr && hwmgr->pm_en) {
 		mutex_lock(&hwmgr->smu_lock);
@@ -191,14 +190,6 @@ static int pp_late_init(void *handle)
 	if (adev->pm.smu_prv_buffer_size != 0)
 		pp_reserve_vram_for_smu(adev);
 
-	if (hwmgr && hwmgr->hwmgr_func &&
-	    hwmgr->hwmgr_func->gfx_off_control &&
-	    (hwmgr->feature_mask & PP_GFXOFF_MASK)) {
-		ret = hwmgr->hwmgr_func->gfx_off_control(hwmgr, true);
-		if (ret)
-			pr_err("gfx off enabling failed!\n");
-	}
-
 	return 0;
 }
 
-- 
cgit v1.1


From 5c16f36f6f003b4415237acca59384a074cd8030 Mon Sep 17 00:00:00 2001
From: Kenneth Feng <kenneth.feng@amd.com>
Date: Tue, 12 Jun 2018 15:07:37 +0800
Subject: drm/amd/powerplay: Set higher SCLK&MCLK frequency than dpm7 in OD
 (v2)

Fix the issue that SCLK&MCLK can't be set higher than dpm7 when
OD is enabled in SMU7.

v2: fix warning (Alex)

Signed-off-by: Kenneth Feng <kenneth.feng@amd.com>
Acked-by: Rex Zhu<rezhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index b763c54..f8e866c 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -3755,14 +3755,17 @@ static int smu7_trim_dpm_states(struct pp_hwmgr *hwmgr,
 static int smu7_generate_dpm_level_enable_mask(
 		struct pp_hwmgr *hwmgr, const void *input)
 {
-	int result;
+	int result = 0;
 	const struct phm_set_power_state_input *states =
 			(const struct phm_set_power_state_input *)input;
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
 	const struct smu7_power_state *smu7_ps =
 			cast_const_phw_smu7_power_state(states->pnew_state);
 
-	result = smu7_trim_dpm_states(hwmgr, smu7_ps);
+	/*skip the trim if od is enabled*/
+	if (!hwmgr->od_enabled)
+		result = smu7_trim_dpm_states(hwmgr, smu7_ps);
+
 	if (result)
 		return result;
 
-- 
cgit v1.1