drm/radeon: Allow write-combined CPU mappings of BOs in GTT (v2)

v2: fix rebase onto drm-fixes

Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

21 files changed, 70 insertions, 47 deletions

diff --git a/drivers/gpu/drm/radeon/cik.c b/drivers/gpu/drm/radeon/cik.c
index 64100fc..e6576b4 100644
--- a/drivers/gpu/drm/radeon/cik.c
+++ b/drivers/gpu/drm/radeon/cik.c
@@ -4676,7 +4676,7 @@ static int cik_mec_init(struct radeon_device *rdev)
 		r = radeon_bo_create(rdev,
 				     rdev->mec.num_mec *rdev->mec.num_pipe * MEC_HPD_SIZE * 2,
 				     PAGE_SIZE, true,
-				     RADEON_GEM_DOMAIN_GTT, NULL,
+				     RADEON_GEM_DOMAIN_GTT, 0, NULL,
 				     &rdev->mec.hpd_eop_obj);
 		if (r) {
 			dev_warn(rdev->dev, "(%d) create HDP EOP bo failed\n", r);
@@ -4846,7 +4846,7 @@ static int cik_cp_compute_resume(struct radeon_device *rdev)
 			r = radeon_bo_create(rdev,
 					     sizeof(struct bonaire_mqd),
 					     PAGE_SIZE, true,
-					     RADEON_GEM_DOMAIN_GTT, NULL,
+					     RADEON_GEM_DOMAIN_GTT, 0, NULL,
 					     &rdev->ring[idx].mqd_obj);
 			if (r) {
 				dev_warn(rdev->dev, "(%d) create MQD bo failed\n", r);
diff --git a/drivers/gpu/drm/radeon/cik_sdma.c b/drivers/gpu/drm/radeon/cik_sdma.c
index 60aa1cb..4b2ab76 100644
--- a/drivers/gpu/drm/radeon/cik_sdma.c
+++ b/drivers/gpu/drm/radeon/cik_sdma.c
@@ -771,7 +771,8 @@ void cik_sdma_vm_set_page(struct radeon_device *rdev,
 
 	trace_radeon_vm_set_page(pe, addr, count, incr, flags);
 
-	if (flags == R600_PTE_GART) {
+	/* XXX: How to distinguish between GART and other system memory pages? */
+	if (flags & R600_PTE_SYSTEM) {
 		uint64_t src = rdev->gart.table_addr + (addr >> 12) * 8;
 		while (count) {
 			unsigned bytes = count * 8;
diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c
index ae7923c..27ebaaa 100644
--- a/drivers/gpu/drm/radeon/evergreen.c
+++ b/drivers/gpu/drm/radeon/evergreen.c
@@ -4022,7 +4022,8 @@ int sumo_rlc_init(struct radeon_device *rdev)
 		/* save restore block */
 		if (rdev->rlc.save_restore_obj == NULL) {
 			r = radeon_bo_create(rdev, dws * 4, PAGE_SIZE, true,
-					     RADEON_GEM_DOMAIN_VRAM, NULL, &rdev->rlc.save_restore_obj);
+					     RADEON_GEM_DOMAIN_VRAM, 0, NULL,
+					     &rdev->rlc.save_restore_obj);
 			if (r) {
 				dev_warn(rdev->dev, "(%d) create RLC sr bo failed\n", r);
 				return r;
@@ -4100,7 +4101,8 @@ int sumo_rlc_init(struct radeon_device *rdev)
 
 		if (rdev->rlc.clear_state_obj == NULL) {
 			r = radeon_bo_create(rdev, dws * 4, PAGE_SIZE, true,
-					     RADEON_GEM_DOMAIN_VRAM, NULL, &rdev->rlc.clear_state_obj);
+					     RADEON_GEM_DOMAIN_VRAM, 0, NULL,
+					     &rdev->rlc.clear_state_obj);
 			if (r) {
 				dev_warn(rdev->dev, "(%d) create RLC c bo failed\n", r);
 				sumo_rlc_fini(rdev);
@@ -4174,8 +4176,10 @@ int sumo_rlc_init(struct radeon_device *rdev)
 
 	if (rdev->rlc.cp_table_size) {
 		if (rdev->rlc.cp_table_obj == NULL) {
-			r = radeon_bo_create(rdev, rdev->rlc.cp_table_size, PAGE_SIZE, true,
-					     RADEON_GEM_DOMAIN_VRAM, NULL, &rdev->rlc.cp_table_obj);
+			r = radeon_bo_create(rdev, rdev->rlc.cp_table_size,
+					     PAGE_SIZE, true,
+					     RADEON_GEM_DOMAIN_VRAM, 0, NULL,
+					     &rdev->rlc.cp_table_obj);
 			if (r) {
 				dev_warn(rdev->dev, "(%d) create RLC cp table bo failed\n", r);
 				sumo_rlc_fini(rdev);
diff --git a/drivers/gpu/drm/radeon/r600.c b/drivers/gpu/drm/radeon/r600.c
index e6a2243..c17ff5d 100644
--- a/drivers/gpu/drm/radeon/r600.c
+++ b/drivers/gpu/drm/radeon/r600.c
@@ -1338,7 +1338,7 @@ int r600_vram_scratch_init(struct radeon_device *rdev)
 	if (rdev->vram_scratch.robj == NULL) {
 		r = radeon_bo_create(rdev, RADEON_GPU_PAGE_SIZE,
 				     PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM,
-				     NULL, &rdev->vram_scratch.robj);
+				     0, NULL, &rdev->vram_scratch.robj);
 		if (r) {
 			return r;
 		}
@@ -3226,7 +3226,7 @@ int r600_ih_ring_alloc(struct radeon_device *rdev)
 	if (rdev->ih.ring_obj == NULL) {
 		r = radeon_bo_create(rdev, rdev->ih.ring_size,
 				     PAGE_SIZE, true,
-				     RADEON_GEM_DOMAIN_GTT,
+				     RADEON_GEM_DOMAIN_GTT, 0,
 				     NULL, &rdev->ih.ring_obj);
 		if (r) {
 			DRM_ERROR("radeon: failed to create ih ring buffer (%d).\n", r);
diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h
index b1eea04..67e9931 100644
--- a/drivers/gpu/drm/radeon/radeon.h
+++ b/drivers/gpu/drm/radeon/radeon.h
@@ -468,6 +468,7 @@ struct radeon_bo {
 	struct ttm_placement		placement;
 	struct ttm_buffer_object	tbo;
 	struct ttm_bo_kmap_obj		kmap;
+	u32				flags;
 	unsigned			pin_count;
 	void				*kptr;
 	u32				tiling_flags;
@@ -548,7 +549,7 @@ int radeon_gem_init(struct radeon_device *rdev);
 void radeon_gem_fini(struct radeon_device *rdev);
 int radeon_gem_object_create(struct radeon_device *rdev, int size,
 				int alignment, int initial_domain,
-				bool discardable, bool kernel,
+				u32 flags, bool discardable, bool kernel,
 				struct drm_gem_object **obj);
 
 int radeon_mode_dumb_create(struct drm_file *file_priv,
diff --git a/drivers/gpu/drm/radeon/radeon_benchmark.c b/drivers/gpu/drm/radeon/radeon_benchmark.c
index 6e05a2e..69f5695 100644
--- a/drivers/gpu/drm/radeon/radeon_benchmark.c
+++ b/drivers/gpu/drm/radeon/radeon_benchmark.c
@@ -97,7 +97,7 @@ static void radeon_benchmark_move(struct radeon_device *rdev, unsigned size,
 	int time;
 
 	n = RADEON_BENCHMARK_ITERATIONS;
-	r = radeon_bo_create(rdev, size, PAGE_SIZE, true, sdomain, NULL, &sobj);
+	r = radeon_bo_create(rdev, size, PAGE_SIZE, true, sdomain, 0, NULL, &sobj);
 	if (r) {
 		goto out_cleanup;
 	}
@@ -109,7 +109,7 @@ static void radeon_benchmark_move(struct radeon_device *rdev, unsigned size,
 	if (r) {
 		goto out_cleanup;
 	}
-	r = radeon_bo_create(rdev, size, PAGE_SIZE, true, ddomain, NULL, &dobj);
+	r = radeon_bo_create(rdev, size, PAGE_SIZE, true, ddomain, 0, NULL, &dobj);
 	if (r) {
 		goto out_cleanup;
 	}
diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c
index 697add2..61990bd 100644
--- a/drivers/gpu/drm/radeon/radeon_device.c
+++ b/drivers/gpu/drm/radeon/radeon_device.c
@@ -385,7 +385,8 @@ int radeon_wb_init(struct radeon_device *rdev)
 
 	if (rdev->wb.wb_obj == NULL) {
 		r = radeon_bo_create(rdev, RADEON_GPU_PAGE_SIZE, PAGE_SIZE, true,
-				     RADEON_GEM_DOMAIN_GTT, NULL, &rdev->wb.wb_obj);
+				     RADEON_GEM_DOMAIN_GTT, 0, NULL,
+				     &rdev->wb.wb_obj);
 		if (r) {
 			dev_warn(rdev->dev, "(%d) create WB bo failed\n", r);
 			return r;
diff --git a/drivers/gpu/drm/radeon/radeon_fb.c b/drivers/gpu/drm/radeon/radeon_fb.c
index db598d7..477ea0d 100644
--- a/drivers/gpu/drm/radeon/radeon_fb.c
+++ b/drivers/gpu/drm/radeon/radeon_fb.c
@@ -127,7 +127,7 @@ static int radeonfb_create_pinned_object(struct radeon_fbdev *rfbdev,
 	aligned_size = ALIGN(size, PAGE_SIZE);
 	ret = radeon_gem_object_create(rdev, aligned_size, 0,
 				       RADEON_GEM_DOMAIN_VRAM,
-				       false, true,
+				       0, false, true,
 				       &gobj);
 	if (ret) {
 		printk(KERN_ERR "failed to allocate framebuffer (%d)\n",
diff --git a/drivers/gpu/drm/radeon/radeon_gart.c b/drivers/gpu/drm/radeon/radeon_gart.c
index d684642..a053a07 100644
--- a/drivers/gpu/drm/radeon/radeon_gart.c
+++ b/drivers/gpu/drm/radeon/radeon_gart.c
@@ -128,7 +128,7 @@ int radeon_gart_table_vram_alloc(struct radeon_device *rdev)
 	if (rdev->gart.robj == NULL) {
 		r = radeon_bo_create(rdev, rdev->gart.table_size,
 				     PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM,
-				     NULL, &rdev->gart.robj);
+				     0, NULL, &rdev->gart.robj);
 		if (r) {
 			return r;
 		}
diff --git a/drivers/gpu/drm/radeon/radeon_gem.c b/drivers/gpu/drm/radeon/radeon_gem.c
index fdd189b..ce1d61e 100644
--- a/drivers/gpu/drm/radeon/radeon_gem.c
+++ b/drivers/gpu/drm/radeon/radeon_gem.c
@@ -42,7 +42,7 @@ void radeon_gem_object_free(struct drm_gem_object *gobj)
 
 int radeon_gem_object_create(struct radeon_device *rdev, int size,
 				int alignment, int initial_domain,
-				bool discardable, bool kernel,
+				u32 flags, bool discardable, bool kernel,
 				struct drm_gem_object **obj)
 {
 	struct radeon_bo *robj;
@@ -64,7 +64,8 @@ int radeon_gem_object_create(struct radeon_device *rdev, int size,
 	}
 
 retry:
-	r = radeon_bo_create(rdev, size, alignment, kernel, initial_domain, NULL, &robj);
+	r = radeon_bo_create(rdev, size, alignment, kernel, initial_domain,
+			     flags, NULL, &robj);
 	if (r) {
 		if (r != -ERESTARTSYS) {
 			if (initial_domain == RADEON_GEM_DOMAIN_VRAM) {
@@ -252,8 +253,8 @@ int radeon_gem_create_ioctl(struct drm_device *dev, void *data,
 	/* create a gem object to contain this object in */
 	args->size = roundup(args->size, PAGE_SIZE);
 	r = radeon_gem_object_create(rdev, args->size, args->alignment,
-					args->initial_domain, false,
-					false, &gobj);
+				     args->initial_domain, args->flags,
+				     false, false, &gobj);
 	if (r) {
 		up_read(&rdev->exclusive_lock);
 		r = radeon_gem_handle_lockup(rdev, r);
@@ -461,11 +462,6 @@ int radeon_gem_va_ioctl(struct drm_device *dev, void *data,
 		args->operation = RADEON_VA_RESULT_ERROR;
 		return -EINVAL;
 	}
-	if (!(args->flags & RADEON_VM_PAGE_SNOOPED)) {
-		dev_err(&dev->pdev->dev, "only supported snooped mapping for now\n");
-		args->operation = RADEON_VA_RESULT_ERROR;
-		return -EINVAL;
-	}
 
 	switch (args->operation) {
 	case RADEON_VA_MAP:
@@ -572,7 +568,7 @@ int radeon_mode_dumb_create(struct drm_file *file_priv,
 	args->size = ALIGN(args->size, PAGE_SIZE);
 
 	r = radeon_gem_object_create(rdev, args->size, 0,
-				     RADEON_GEM_DOMAIN_VRAM,
+				     RADEON_GEM_DOMAIN_VRAM, 0,
 				     false, ttm_bo_type_device,
 				     &gobj);
 	if (r)
diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c
index 6c717b2..1b05a4d 100644
--- a/drivers/gpu/drm/radeon/radeon_object.c
+++ b/drivers/gpu/drm/radeon/radeon_object.c
@@ -114,15 +114,23 @@ void radeon_ttm_placement_from_domain(struct radeon_bo *rbo, u32 domain)
 		rbo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED |
 					TTM_PL_FLAG_VRAM;
 	if (domain & RADEON_GEM_DOMAIN_GTT) {
-		if (rbo->rdev->flags & RADEON_IS_AGP) {
-			rbo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_TT;
+		if (rbo->flags & RADEON_GEM_GTT_UC) {
+			rbo->placements[c++] = TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_TT;
+		} else if ((rbo->flags & RADEON_GEM_GTT_WC) ||
+			   (rbo->rdev->flags & RADEON_IS_AGP)) {
+			rbo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED |
+				TTM_PL_FLAG_TT;
 		} else {
 			rbo->placements[c++] = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_TT;
 		}
 	}
 	if (domain & RADEON_GEM_DOMAIN_CPU) {
-		if (rbo->rdev->flags & RADEON_IS_AGP) {
-			rbo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_SYSTEM;
+		if (rbo->flags & RADEON_GEM_GTT_UC) {
+			rbo->placements[c++] = TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_SYSTEM;
+		} else if ((rbo->flags & RADEON_GEM_GTT_WC) ||
+		    rbo->rdev->flags & RADEON_IS_AGP) {
+			rbo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED |
+				TTM_PL_FLAG_SYSTEM;
 		} else {
 			rbo->placements[c++] = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_SYSTEM;
 		}
@@ -146,7 +154,7 @@ void radeon_ttm_placement_from_domain(struct radeon_bo *rbo, u32 domain)
 
 int radeon_bo_create(struct radeon_device *rdev,
 		     unsigned long size, int byte_align, bool kernel, u32 domain,
-		     struct sg_table *sg, struct radeon_bo **bo_ptr)
+		     u32 flags, struct sg_table *sg, struct radeon_bo **bo_ptr)
 {
 	struct radeon_bo *bo;
 	enum ttm_bo_type type;
@@ -183,6 +191,12 @@ int radeon_bo_create(struct radeon_device *rdev,
 	bo->initial_domain = domain & (RADEON_GEM_DOMAIN_VRAM |
 	                               RADEON_GEM_DOMAIN_GTT |
 	                               RADEON_GEM_DOMAIN_CPU);
+
+	bo->flags = flags;
+	/* PCI GART is always snooped */
+	if (!(rdev->flags & RADEON_IS_PCIE))
+		bo->flags &= ~(RADEON_GEM_GTT_WC | RADEON_GEM_GTT_UC);
+
 	radeon_ttm_placement_from_domain(bo, domain);
 	/* Kernel allocation are uninterruptible */
 	down_read(&rdev->pm.mclk_lock);
diff --git a/drivers/gpu/drm/radeon/radeon_object.h b/drivers/gpu/drm/radeon/radeon_object.h
index 5a873f3..972ec0f 100644
--- a/drivers/gpu/drm/radeon/radeon_object.h
+++ b/drivers/gpu/drm/radeon/radeon_object.h
@@ -124,7 +124,7 @@ extern int radeon_bo_wait(struct radeon_bo *bo, u32 *mem_type,
 
 extern int radeon_bo_create(struct radeon_device *rdev,
 			    unsigned long size, int byte_align,
-			    bool kernel, u32 domain,
+			    bool kernel, u32 domain, u32 flags,
 			    struct sg_table *sg,
 			    struct radeon_bo **bo_ptr);
 extern int radeon_bo_kmap(struct radeon_bo *bo, void **ptr);
@@ -170,7 +170,8 @@ static inline void * radeon_sa_bo_cpu_addr(struct radeon_sa_bo *sa_bo)
 
 extern int radeon_sa_bo_manager_init(struct radeon_device *rdev,
 				     struct radeon_sa_manager *sa_manager,
-				     unsigned size, u32 align, u32 domain);
+				     unsigned size, u32 align, u32 domain,
+				     u32 flags);
 extern void radeon_sa_bo_manager_fini(struct radeon_device *rdev,
 				      struct radeon_sa_manager *sa_manager);
 extern int radeon_sa_bo_manager_start(struct radeon_device *rdev,
diff --git a/drivers/gpu/drm/radeon/radeon_prime.c b/drivers/gpu/drm/radeon/radeon_prime.c
index 2007456..f7e48d3 100644
--- a/drivers/gpu/drm/radeon/radeon_prime.c
+++ b/drivers/gpu/drm/radeon/radeon_prime.c
@@ -65,7 +65,7 @@ struct drm_gem_object *radeon_gem_prime_import_sg_table(struct drm_device *dev,
 	int ret;
 
 	ret = radeon_bo_create(rdev, size, PAGE_SIZE, false,
-			       RADEON_GEM_DOMAIN_GTT, sg, &bo);
+			       RADEON_GEM_DOMAIN_GTT, 0, sg, &bo);
 	if (ret)
 		return ERR_PTR(ret);
 
diff --git a/drivers/gpu/drm/radeon/radeon_ring.c b/drivers/gpu/drm/radeon/radeon_ring.c
index f8050f5..71439f0 100644
--- a/drivers/gpu/drm/radeon/radeon_ring.c
+++ b/drivers/gpu/drm/radeon/radeon_ring.c
@@ -204,7 +204,7 @@ int radeon_ib_pool_init(struct radeon_device *rdev)
 	r = radeon_sa_bo_manager_init(rdev, &rdev->ring_tmp_bo,
 				      RADEON_IB_POOL_SIZE*64*1024,
 				      RADEON_GPU_PAGE_SIZE,
-				      RADEON_GEM_DOMAIN_GTT);
+				      RADEON_GEM_DOMAIN_GTT, 0);
 	if (r) {
 		return r;
 	}
@@ -640,7 +640,7 @@ int radeon_ring_init(struct radeon_device *rdev, struct radeon_ring *ring, unsig
 	/* Allocate ring buffer */
 	if (ring->ring_obj == NULL) {
 		r = radeon_bo_create(rdev, ring->ring_size, PAGE_SIZE, true,
-				     RADEON_GEM_DOMAIN_GTT,
+				     RADEON_GEM_DOMAIN_GTT, 0,
 				     NULL, &ring->ring_obj);
 		if (r) {
 			dev_err(rdev->dev, "(%d) ring create failed\n", r);
diff --git a/drivers/gpu/drm/radeon/radeon_sa.c b/drivers/gpu/drm/radeon/radeon_sa.c
index adcf3e2..b84f97c 100644
--- a/drivers/gpu/drm/radeon/radeon_sa.c
+++ b/drivers/gpu/drm/radeon/radeon_sa.c
@@ -49,7 +49,7 @@ static void radeon_sa_bo_try_free(struct radeon_sa_manager *sa_manager);
 
 int radeon_sa_bo_manager_init(struct radeon_device *rdev,
 			      struct radeon_sa_manager *sa_manager,
-			      unsigned size, u32 align, u32 domain)
+			      unsigned size, u32 align, u32 domain, u32 flags)
 {
 	int i, r;
 
@@ -65,7 +65,7 @@ int radeon_sa_bo_manager_init(struct radeon_device *rdev,
 	}
 
 	r = radeon_bo_create(rdev, size, align, true,
-			     domain, NULL, &sa_manager->bo);
+			     domain, flags, NULL, &sa_manager->bo);
 	if (r) {
 		dev_err(rdev->dev, "(%d) failed to allocate bo for manager\n", r);
 		return r;
diff --git a/drivers/gpu/drm/radeon/radeon_test.c b/drivers/gpu/drm/radeon/radeon_test.c
index 3a13e0d..9c5b66c 100644
--- a/drivers/gpu/drm/radeon/radeon_test.c
+++ b/drivers/gpu/drm/radeon/radeon_test.c
@@ -73,7 +73,7 @@ static void radeon_do_test_moves(struct radeon_device *rdev, int flag)
 	}
 
 	r = radeon_bo_create(rdev, size, PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM,
-			     NULL, &vram_obj);
+			     0, NULL, &vram_obj);
 	if (r) {
 		DRM_ERROR("Failed to create VRAM object\n");
 		goto out_cleanup;
@@ -93,7 +93,7 @@ static void radeon_do_test_moves(struct radeon_device *rdev, int flag)
 		struct radeon_fence *fence = NULL;
 
 		r = radeon_bo_create(rdev, size, PAGE_SIZE, true,
-				     RADEON_GEM_DOMAIN_GTT, NULL, gtt_obj + i);
+				     RADEON_GEM_DOMAIN_GTT, 0, NULL, gtt_obj + i);
 		if (r) {
 			DRM_ERROR("Failed to create GTT object %d\n", i);
 			goto out_lclean;
diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c
index 7fb7c1c..72afe82 100644
--- a/drivers/gpu/drm/radeon/radeon_ttm.c
+++ b/drivers/gpu/drm/radeon/radeon_ttm.c
@@ -730,7 +730,7 @@ int radeon_ttm_init(struct radeon_device *rdev)
 	radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
 
 	r = radeon_bo_create(rdev, 256 * 1024, PAGE_SIZE, true,
-			     RADEON_GEM_DOMAIN_VRAM,
+			     RADEON_GEM_DOMAIN_VRAM, 0,
 			     NULL, &rdev->stollen_vga_memory);
 	if (r) {
 		return r;
diff --git a/drivers/gpu/drm/radeon/radeon_uvd.c b/drivers/gpu/drm/radeon/radeon_uvd.c
index a4ad270..6bf55ec 100644
--- a/drivers/gpu/drm/radeon/radeon_uvd.c
+++ b/drivers/gpu/drm/radeon/radeon_uvd.c
@@ -117,7 +117,7 @@ int radeon_uvd_init(struct radeon_device *rdev)
 	bo_size = RADEON_GPU_PAGE_ALIGN(rdev->uvd_fw->size + 8) +
 		  RADEON_UVD_STACK_SIZE + RADEON_UVD_HEAP_SIZE;
 	r = radeon_bo_create(rdev, bo_size, PAGE_SIZE, true,
-			     RADEON_GEM_DOMAIN_VRAM, NULL, &rdev->uvd.vcpu_bo);
+			     RADEON_GEM_DOMAIN_VRAM, 0, NULL, &rdev->uvd.vcpu_bo);
 	if (r) {
 		dev_err(rdev->dev, "(%d) failed to allocate UVD bo\n", r);
 		return r;
@@ -674,7 +674,7 @@ int radeon_uvd_get_create_msg(struct radeon_device *rdev, int ring,
 	int r, i;
 
 	r = radeon_bo_create(rdev, 1024, PAGE_SIZE, true,
-			     RADEON_GEM_DOMAIN_VRAM, NULL, &bo);
+			     RADEON_GEM_DOMAIN_VRAM, 0, NULL, &bo);
 	if (r)
 		return r;
 
@@ -720,7 +720,7 @@ int radeon_uvd_get_destroy_msg(struct radeon_device *rdev, int ring,
 	int r, i;
 
 	r = radeon_bo_create(rdev, 1024, PAGE_SIZE, true,
-			     RADEON_GEM_DOMAIN_VRAM, NULL, &bo);
+			     RADEON_GEM_DOMAIN_VRAM, 0, NULL, &bo);
 	if (r)
 		return r;
 
diff --git a/drivers/gpu/drm/radeon/radeon_vce.c b/drivers/gpu/drm/radeon/radeon_vce.c
index aa21c31..f9b70a4 100644
--- a/drivers/gpu/drm/radeon/radeon_vce.c
+++ b/drivers/gpu/drm/radeon/radeon_vce.c
@@ -126,7 +126,7 @@ int radeon_vce_init(struct radeon_device *rdev)
 	size = RADEON_GPU_PAGE_ALIGN(rdev->vce_fw->size) +
 	       RADEON_VCE_STACK_SIZE + RADEON_VCE_HEAP_SIZE;
 	r = radeon_bo_create(rdev, size, PAGE_SIZE, true,
-			     RADEON_GEM_DOMAIN_VRAM, NULL, &rdev->vce.vcpu_bo);
+			     RADEON_GEM_DOMAIN_VRAM, 0, NULL, &rdev->vce.vcpu_bo);
 	if (r) {
 		dev_err(rdev->dev, "(%d) failed to allocate VCE bo\n", r);
 		return r;
diff --git a/drivers/gpu/drm/radeon/radeon_vm.c b/drivers/gpu/drm/radeon/radeon_vm.c
index 725d366..f0a8cd7 100644
--- a/drivers/gpu/drm/radeon/radeon_vm.c
+++ b/drivers/gpu/drm/radeon/radeon_vm.c
@@ -510,7 +510,7 @@ int radeon_vm_bo_set_addr(struct radeon_device *rdev,
 
 		r = radeon_bo_create(rdev, RADEON_VM_PTE_COUNT * 8,
 				     RADEON_GPU_PAGE_SIZE, true,
-				     RADEON_GEM_DOMAIN_VRAM, NULL, &pt);
+				     RADEON_GEM_DOMAIN_VRAM, 0, NULL, &pt);
 		if (r)
 			return r;
 
@@ -858,6 +858,7 @@ int radeon_vm_bo_update(struct radeon_device *rdev,
 
 	bo_va->flags &= ~RADEON_VM_PAGE_VALID;
 	bo_va->flags &= ~RADEON_VM_PAGE_SYSTEM;
+	bo_va->flags &= ~RADEON_VM_PAGE_SNOOPED;
 	if (mem) {
 		addr = mem->start << PAGE_SHIFT;
 		if (mem->mem_type != TTM_PL_SYSTEM) {
@@ -866,6 +867,9 @@ int radeon_vm_bo_update(struct radeon_device *rdev,
 		}
 		if (mem->mem_type == TTM_PL_TT) {
 			bo_va->flags |= RADEON_VM_PAGE_SYSTEM;
+			if (!(bo_va->bo->flags & (RADEON_GEM_GTT_WC | RADEON_GEM_GTT_UC)))
+				bo_va->flags |= RADEON_VM_PAGE_SNOOPED;
+
 		} else {
 			addr += rdev->vm_manager.vram_base_offset;
 		}
@@ -1031,7 +1035,7 @@ int radeon_vm_init(struct radeon_device *rdev, struct radeon_vm *vm)
 	}
 
 	r = radeon_bo_create(rdev, pd_size, align, true,
-			     RADEON_GEM_DOMAIN_VRAM, NULL,
+			     RADEON_GEM_DOMAIN_VRAM, 0, NULL,
 			     &vm->page_directory);
 	if (r)
 		return r;
diff --git a/drivers/gpu/drm/radeon/si_dma.c b/drivers/gpu/drm/radeon/si_dma.c
index e24c94b..c9da341 100644
--- a/drivers/gpu/drm/radeon/si_dma.c
+++ b/drivers/gpu/drm/radeon/si_dma.c
@@ -79,7 +79,8 @@ void si_dma_vm_set_page(struct radeon_device *rdev,
 
 	trace_radeon_vm_set_page(pe, addr, count, incr, flags);
 
-	if (flags == R600_PTE_GART) {
+	/* XXX: How to distinguish between GART and other system memory pages? */
+	if (flags & R600_PTE_SYSTEM) {
 		uint64_t src = rdev->gart.table_addr + (addr >> 12) * 8;
 		while (count) {
 			unsigned bytes = count * 8;