diff options
27 files changed, 2881 insertions, 2256 deletions
diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile index b65c06f..8b8bbc7 100644 --- a/drivers/gpu/drm/i915/Makefile +++ b/drivers/gpu/drm/i915/Makefile @@ -11,6 +11,7 @@ i915-y := i915_drv.o i915_dma.o i915_irq.o \ i915_gem_evict.o \ i915_gem_execbuffer.o \ i915_gem_gtt.o \ + i915_gem_stolen.o \ i915_gem_tiling.o \ i915_sysfs.o \ i915_trace_points.o \ diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c index a8db386..950f72a 100644 --- a/drivers/gpu/drm/i915/i915_debugfs.c +++ b/drivers/gpu/drm/i915/i915_debugfs.c @@ -47,7 +47,6 @@ enum { FLUSHING_LIST, INACTIVE_LIST, PINNED_LIST, - DEFERRED_FREE_LIST, }; static const char *yesno(int v) @@ -178,18 +177,10 @@ static int i915_gem_object_list_info(struct seq_file *m, void *data) seq_printf(m, "Inactive:\n"); head = &dev_priv->mm.inactive_list; break; - case PINNED_LIST: - seq_printf(m, "Pinned:\n"); - head = &dev_priv->mm.pinned_list; - break; case FLUSHING_LIST: seq_printf(m, "Flushing:\n"); head = &dev_priv->mm.flushing_list; break; - case DEFERRED_FREE_LIST: - seq_printf(m, "Deferred free:\n"); - head = &dev_priv->mm.deferred_free_list; - break; default: mutex_unlock(&dev->struct_mutex); return -EINVAL; @@ -252,21 +243,11 @@ static int i915_gem_object_info(struct seq_file *m, void* data) count, mappable_count, size, mappable_size); size = count = mappable_size = mappable_count = 0; - count_objects(&dev_priv->mm.pinned_list, mm_list); - seq_printf(m, " %u [%u] pinned objects, %zu [%zu] bytes\n", - count, mappable_count, size, mappable_size); - - size = count = mappable_size = mappable_count = 0; count_objects(&dev_priv->mm.inactive_list, mm_list); seq_printf(m, " %u [%u] inactive objects, %zu [%zu] bytes\n", count, mappable_count, size, mappable_size); size = count = mappable_size = mappable_count = 0; - count_objects(&dev_priv->mm.deferred_free_list, mm_list); - seq_printf(m, " %u [%u] freed objects, %zu [%zu] bytes\n", - count, mappable_count, size, mappable_size); - - size = count = mappable_size = mappable_count = 0; list_for_each_entry(obj, &dev_priv->mm.gtt_list, gtt_list) { if (obj->fault_mappable) { size += obj->gtt_space->size; @@ -294,6 +275,7 @@ static int i915_gem_gtt_info(struct seq_file *m, void* data) { struct drm_info_node *node = (struct drm_info_node *) m->private; struct drm_device *dev = node->minor->dev; + uintptr_t list = (uintptr_t) node->info_ent->data; struct drm_i915_private *dev_priv = dev->dev_private; struct drm_i915_gem_object *obj; size_t total_obj_size, total_gtt_size; @@ -305,6 +287,9 @@ static int i915_gem_gtt_info(struct seq_file *m, void* data) total_obj_size = total_gtt_size = count = 0; list_for_each_entry(obj, &dev_priv->mm.gtt_list, gtt_list) { + if (list == PINNED_LIST && obj->pin_count == 0) + continue; + seq_printf(m, " "); describe_obj(m, obj); seq_printf(m, "\n"); @@ -321,7 +306,6 @@ static int i915_gem_gtt_info(struct seq_file *m, void* data) return 0; } - static int i915_gem_pageflip_info(struct seq_file *m, void *data) { struct drm_info_node *node = (struct drm_info_node *) m->private; @@ -430,10 +414,6 @@ static void i915_ring_seqno_info(struct seq_file *m, if (ring->get_seqno) { seq_printf(m, "Current sequence (%s): %d\n", ring->name, ring->get_seqno(ring)); - seq_printf(m, "Waiter sequence (%s): %d\n", - ring->name, ring->waiting_seqno); - seq_printf(m, "IRQ sequence (%s): %d\n", - ring->name, ring->irq_seqno); } } @@ -602,69 +582,6 @@ static int i915_hws_info(struct seq_file *m, void *data) return 0; } -static int i915_ringbuffer_data(struct seq_file *m, void *data) -{ - struct drm_info_node *node = (struct drm_info_node *) m->private; - struct drm_device *dev = node->minor->dev; - drm_i915_private_t *dev_priv = dev->dev_private; - struct intel_ring_buffer *ring; - int ret; - - ret = mutex_lock_interruptible(&dev->struct_mutex); - if (ret) - return ret; - - ring = &dev_priv->ring[(uintptr_t)node->info_ent->data]; - if (!ring->obj) { - seq_printf(m, "No ringbuffer setup\n"); - } else { - const u8 __iomem *virt = ring->virtual_start; - uint32_t off; - - for (off = 0; off < ring->size; off += 4) { - uint32_t *ptr = (uint32_t *)(virt + off); - seq_printf(m, "%08x : %08x\n", off, *ptr); - } - } - mutex_unlock(&dev->struct_mutex); - - return 0; -} - -static int i915_ringbuffer_info(struct seq_file *m, void *data) -{ - struct drm_info_node *node = (struct drm_info_node *) m->private; - struct drm_device *dev = node->minor->dev; - drm_i915_private_t *dev_priv = dev->dev_private; - struct intel_ring_buffer *ring; - int ret; - - ring = &dev_priv->ring[(uintptr_t)node->info_ent->data]; - if (ring->size == 0) - return 0; - - ret = mutex_lock_interruptible(&dev->struct_mutex); - if (ret) - return ret; - - seq_printf(m, "Ring %s:\n", ring->name); - seq_printf(m, " Head : %08x\n", I915_READ_HEAD(ring) & HEAD_ADDR); - seq_printf(m, " Tail : %08x\n", I915_READ_TAIL(ring) & TAIL_ADDR); - seq_printf(m, " Size : %08x\n", ring->size); - seq_printf(m, " Active : %08x\n", intel_ring_get_active_head(ring)); - seq_printf(m, " NOPID : %08x\n", I915_READ_NOPID(ring)); - if (IS_GEN6(dev) || IS_GEN7(dev)) { - seq_printf(m, " Sync 0 : %08x\n", I915_READ_SYNC_0(ring)); - seq_printf(m, " Sync 1 : %08x\n", I915_READ_SYNC_1(ring)); - } - seq_printf(m, " Control : %08x\n", I915_READ_CTL(ring)); - seq_printf(m, " Start : %08x\n", I915_READ_START(ring)); - - mutex_unlock(&dev->struct_mutex); - - return 0; -} - static const char *ring_str(int ring) { switch (ring) { @@ -766,31 +683,35 @@ static void i915_ring_error_state(struct seq_file *m, error->semaphore_mboxes[ring][1]); } seq_printf(m, " seqno: 0x%08x\n", error->seqno[ring]); + seq_printf(m, " waiting: %s\n", yesno(error->waiting[ring])); seq_printf(m, " ring->head: 0x%08x\n", error->cpu_ring_head[ring]); seq_printf(m, " ring->tail: 0x%08x\n", error->cpu_ring_tail[ring]); } +struct i915_error_state_file_priv { + struct drm_device *dev; + struct drm_i915_error_state *error; +}; + static int i915_error_state(struct seq_file *m, void *unused) { - struct drm_info_node *node = (struct drm_info_node *) m->private; - struct drm_device *dev = node->minor->dev; + struct i915_error_state_file_priv *error_priv = m->private; + struct drm_device *dev = error_priv->dev; drm_i915_private_t *dev_priv = dev->dev_private; - struct drm_i915_error_state *error; - unsigned long flags; + struct drm_i915_error_state *error = error_priv->error; int i, j, page, offset, elt; - spin_lock_irqsave(&dev_priv->error_lock, flags); - if (!dev_priv->first_error) { + if (!error) { seq_printf(m, "no error state collected\n"); - goto out; + return 0; } - error = dev_priv->first_error; seq_printf(m, "Time: %ld s %ld us\n", error->time.tv_sec, error->time.tv_usec); seq_printf(m, "PCI ID: 0x%04x\n", dev->pci_device); seq_printf(m, "EIR: 0x%08x\n", error->eir); + seq_printf(m, "IER: 0x%08x\n", error->ier); seq_printf(m, "PGTBL_ER: 0x%08x\n", error->pgtbl_er); for (i = 0; i < dev_priv->num_fence_regs; i++) @@ -867,12 +788,71 @@ static int i915_error_state(struct seq_file *m, void *unused) if (error->display) intel_display_print_error_state(m, dev, error->display); -out: + return 0; +} + +static ssize_t +i915_error_state_write(struct file *filp, + const char __user *ubuf, + size_t cnt, + loff_t *ppos) +{ + struct seq_file *m = filp->private_data; + struct i915_error_state_file_priv *error_priv = m->private; + struct drm_device *dev = error_priv->dev; + + DRM_DEBUG_DRIVER("Resetting error state\n"); + + mutex_lock(&dev->struct_mutex); + i915_destroy_error_state(dev); + mutex_unlock(&dev->struct_mutex); + + return cnt; +} + +static int i915_error_state_open(struct inode *inode, struct file *file) +{ + struct drm_device *dev = inode->i_private; + drm_i915_private_t *dev_priv = dev->dev_private; + struct i915_error_state_file_priv *error_priv; + unsigned long flags; + + error_priv = kzalloc(sizeof(*error_priv), GFP_KERNEL); + if (!error_priv) + return -ENOMEM; + + error_priv->dev = dev; + + spin_lock_irqsave(&dev_priv->error_lock, flags); + error_priv->error = dev_priv->first_error; + if (error_priv->error) + kref_get(&error_priv->error->ref); spin_unlock_irqrestore(&dev_priv->error_lock, flags); - return 0; + return single_open(file, i915_error_state, error_priv); } +static int i915_error_state_release(struct inode *inode, struct file *file) +{ + struct seq_file *m = file->private_data; + struct i915_error_state_file_priv *error_priv = m->private; + + if (error_priv->error) + kref_put(&error_priv->error->ref, i915_error_state_free); + kfree(error_priv); + + return single_release(inode, file); +} + +static const struct file_operations i915_error_state_fops = { + .owner = THIS_MODULE, + .open = i915_error_state_open, + .read = seq_read, + .write = i915_error_state_write, + .llseek = default_llseek, + .release = i915_error_state_release, +}; + static int i915_rstdby_delays(struct seq_file *m, void *unused) { struct drm_info_node *node = (struct drm_info_node *) m->private; @@ -1356,17 +1336,25 @@ static int i915_opregion(struct seq_file *m, void *unused) struct drm_device *dev = node->minor->dev; drm_i915_private_t *dev_priv = dev->dev_private; struct intel_opregion *opregion = &dev_priv->opregion; + void *data = kmalloc(OPREGION_SIZE, GFP_KERNEL); int ret; + if (data == NULL) + return -ENOMEM; + ret = mutex_lock_interruptible(&dev->struct_mutex); if (ret) - return ret; + goto out; - if (opregion->header) - seq_write(m, opregion->header, OPREGION_SIZE); + if (opregion->header) { + memcpy_fromio(data, opregion->header, OPREGION_SIZE); + seq_write(m, data, OPREGION_SIZE); + } mutex_unlock(&dev->struct_mutex); +out: + kfree(data); return 0; } @@ -1659,6 +1647,65 @@ static const struct file_operations i915_wedged_fops = { }; static ssize_t +i915_ring_stop_read(struct file *filp, + char __user *ubuf, + size_t max, + loff_t *ppos) +{ + struct drm_device *dev = filp->private_data; + drm_i915_private_t *dev_priv = dev->dev_private; + char buf[20]; + int len; + + len = snprintf(buf, sizeof(buf), + "0x%08x\n", dev_priv->stop_rings); + + if (len > sizeof(buf)) + len = sizeof(buf); + + return simple_read_from_buffer(ubuf, max, ppos, buf, len); +} + +static ssize_t +i915_ring_stop_write(struct file *filp, + const char __user *ubuf, + size_t cnt, + loff_t *ppos) +{ + struct drm_device *dev = filp->private_data; + struct drm_i915_private *dev_priv = dev->dev_private; + char buf[20]; + int val = 0; + + if (cnt > 0) { + if (cnt > sizeof(buf) - 1) + return -EINVAL; + + if (copy_from_user(buf, ubuf, cnt)) + return -EFAULT; + buf[cnt] = 0; + + val = simple_strtoul(buf, NULL, 0); + } + + DRM_DEBUG_DRIVER("Stopping rings 0x%08x\n", val); + + mutex_lock(&dev->struct_mutex); + dev_priv->stop_rings = val; + mutex_unlock(&dev->struct_mutex); + + return cnt; +} + +static const struct file_operations i915_ring_stop_fops = { + .owner = THIS_MODULE, + .open = simple_open, + .read = i915_ring_stop_read, + .write = i915_ring_stop_write, + .llseek = default_llseek, +}; + +static ssize_t i915_max_freq_read(struct file *filp, char __user *ubuf, size_t max, @@ -1900,11 +1947,10 @@ static struct drm_info_list i915_debugfs_list[] = { {"i915_capabilities", i915_capabilities, 0}, {"i915_gem_objects", i915_gem_object_info, 0}, {"i915_gem_gtt", i915_gem_gtt_info, 0}, + {"i915_gem_pinned", i915_gem_gtt_info, 0, (void *) PINNED_LIST}, {"i915_gem_active", i915_gem_object_list_info, 0, (void *) ACTIVE_LIST}, {"i915_gem_flushing", i915_gem_object_list_info, 0, (void *) FLUSHING_LIST}, {"i915_gem_inactive", i915_gem_object_list_info, 0, (void *) INACTIVE_LIST}, - {"i915_gem_pinned", i915_gem_object_list_info, 0, (void *) PINNED_LIST}, - {"i915_gem_deferred_free", i915_gem_object_list_info, 0, (void *) DEFERRED_FREE_LIST}, {"i915_gem_pageflip", i915_gem_pageflip_info, 0}, {"i915_gem_request", i915_gem_request_info, 0}, {"i915_gem_seqno", i915_gem_seqno_info, 0}, @@ -1913,13 +1959,6 @@ static struct drm_info_list i915_debugfs_list[] = { {"i915_gem_hws", i915_hws_info, 0, (void *)RCS}, {"i915_gem_hws_blt", i915_hws_info, 0, (void *)BCS}, {"i915_gem_hws_bsd", i915_hws_info, 0, (void *)VCS}, - {"i915_ringbuffer_data", i915_ringbuffer_data, 0, (void *)RCS}, - {"i915_ringbuffer_info", i915_ringbuffer_info, 0, (void *)RCS}, - {"i915_bsd_ringbuffer_data", i915_ringbuffer_data, 0, (void *)VCS}, - {"i915_bsd_ringbuffer_info", i915_ringbuffer_info, 0, (void *)VCS}, - {"i915_blt_ringbuffer_data", i915_ringbuffer_data, 0, (void *)BCS}, - {"i915_blt_ringbuffer_info", i915_ringbuffer_info, 0, (void *)BCS}, - {"i915_error_state", i915_error_state, 0}, {"i915_rstdby_delays", i915_rstdby_delays, 0}, {"i915_cur_delayinfo", i915_cur_delayinfo, 0}, {"i915_delayfreq_table", i915_delayfreq_table, 0}, @@ -1965,6 +2004,17 @@ int i915_debugfs_init(struct drm_minor *minor) &i915_cache_sharing_fops); if (ret) return ret; + ret = i915_debugfs_create(minor->debugfs_root, minor, + "i915_ring_stop", + &i915_ring_stop_fops); + if (ret) + return ret; + + ret = i915_debugfs_create(minor->debugfs_root, minor, + "i915_error_state", + &i915_error_state_fops); + if (ret) + return ret; return drm_debugfs_create_files(i915_debugfs_list, I915_DEBUGFS_ENTRIES, @@ -1983,6 +2033,8 @@ void i915_debugfs_cleanup(struct drm_minor *minor) 1, minor); drm_debugfs_remove_files((struct drm_info_list *) &i915_cache_sharing_fops, 1, minor); + drm_debugfs_remove_files((struct drm_info_list *) &i915_ring_stop_fops, + 1, minor); } #endif /* CONFIG_DEBUG_FS */ diff --git a/drivers/gpu/drm/i915/i915_dma.c b/drivers/gpu/drm/i915/i915_dma.c index 068958c..006ea47 100644 --- a/drivers/gpu/drm/i915/i915_dma.c +++ b/drivers/gpu/drm/i915/i915_dma.c @@ -36,17 +36,63 @@ #include "i915_drm.h" #include "i915_drv.h" #include "i915_trace.h" -#include "../../../platform/x86/intel_ips.h" #include <linux/pci.h> #include <linux/vgaarb.h> #include <linux/acpi.h> #include <linux/pnp.h> #include <linux/vga_switcheroo.h> #include <linux/slab.h> -#include <linux/module.h> #include <acpi/video.h> #include <asm/pat.h> +#define LP_RING(d) (&((struct drm_i915_private *)(d))->ring[RCS]) + +#define BEGIN_LP_RING(n) \ + intel_ring_begin(LP_RING(dev_priv), (n)) + +#define OUT_RING(x) \ + intel_ring_emit(LP_RING(dev_priv), x) + +#define ADVANCE_LP_RING() \ + intel_ring_advance(LP_RING(dev_priv)) + +/** + * Lock test for when it's just for synchronization of ring access. + * + * In that case, we don't need to do it when GEM is initialized as nobody else + * has access to the ring. + */ +#define RING_LOCK_TEST_WITH_RETURN(dev, file) do { \ + if (LP_RING(dev->dev_private)->obj == NULL) \ + LOCK_TEST_WITH_RETURN(dev, file); \ +} while (0) + +static inline u32 +intel_read_legacy_status_page(struct drm_i915_private *dev_priv, int reg) +{ + if (I915_NEED_GFX_HWS(dev_priv->dev)) + return ioread32(dev_priv->dri1.gfx_hws_cpu_addr + reg); + else + return intel_read_status_page(LP_RING(dev_priv), reg); +} + +#define READ_HWSP(dev_priv, reg) intel_read_legacy_status_page(dev_priv, reg) +#define READ_BREADCRUMB(dev_priv) READ_HWSP(dev_priv, I915_BREADCRUMB_INDEX) +#define I915_BREADCRUMB_INDEX 0x21 + +void i915_update_dri1_breadcrumb(struct drm_device *dev) +{ + drm_i915_private_t *dev_priv = dev->dev_private; + struct drm_i915_master_private *master_priv; + + if (dev->primary->master) { + master_priv = dev->primary->master->driver_priv; + if (master_priv->sarea_priv) + master_priv->sarea_priv->last_dispatch = + READ_BREADCRUMB(dev_priv); + } +} + static void i915_write_hws_pga(struct drm_device *dev) { drm_i915_private_t *dev_priv = dev->dev_private; @@ -100,7 +146,7 @@ static void i915_free_hws(struct drm_device *dev) if (ring->status_page.gfx_addr) { ring->status_page.gfx_addr = 0; - drm_core_ioremapfree(&dev_priv->hws_map, dev); + iounmap(dev_priv->dri1.gfx_hws_cpu_addr); } /* Need to rewrite hardware status page */ @@ -198,7 +244,7 @@ static int i915_initialize(struct drm_device * dev, drm_i915_init_t * init) /* Allow hardware batchbuffers unless told otherwise. */ - dev_priv->allow_batchbuffer = 1; + dev_priv->dri1.allow_batchbuffer = 1; return 0; } @@ -210,7 +256,7 @@ static int i915_dma_resume(struct drm_device * dev) DRM_DEBUG_DRIVER("%s\n", __func__); - if (ring->map.handle == NULL) { + if (ring->virtual_start == NULL) { DRM_ERROR("can not ioremap virtual address for" " ring buffer\n"); return -ENOMEM; @@ -239,6 +285,9 @@ static int i915_dma_init(struct drm_device *dev, void *data, drm_i915_init_t *init = data; int retcode = 0; + if (drm_core_check_feature(dev, DRIVER_MODESET)) + return -ENODEV; + switch (init->func) { case I915_INIT_DMA: retcode = i915_initialize(dev, init); @@ -581,6 +630,9 @@ static int i915_flush_ioctl(struct drm_device *dev, void *data, { int ret; + if (drm_core_check_feature(dev, DRIVER_MODESET)) + return -ENODEV; + RING_LOCK_TEST_WITH_RETURN(dev, file_priv); mutex_lock(&dev->struct_mutex); @@ -601,7 +653,10 @@ static int i915_batchbuffer(struct drm_device *dev, void *data, int ret; struct drm_clip_rect *cliprects = NULL; - if (!dev_priv->allow_batchbuffer) { + if (drm_core_check_feature(dev, DRIVER_MODESET)) + return -ENODEV; + + if (!dev_priv->dri1.allow_batchbuffer) { DRM_ERROR("Batchbuffer ioctl disabled\n"); return -EINVAL; } @@ -658,6 +713,9 @@ static int i915_cmdbuffer(struct drm_device *dev, void *data, DRM_DEBUG_DRIVER("i915 cmdbuffer, buf %p sz %d cliprects %d\n", cmdbuf->buf, cmdbuf->sz, cmdbuf->num_cliprects); + if (drm_core_check_feature(dev, DRIVER_MODESET)) + return -ENODEV; + RING_LOCK_TEST_WITH_RETURN(dev, file_priv); if (cmdbuf->num_cliprects < 0) @@ -709,11 +767,166 @@ fail_batch_free: return ret; } +static int i915_emit_irq(struct drm_device * dev) +{ + drm_i915_private_t *dev_priv = dev->dev_private; + struct drm_i915_master_private *master_priv = dev->primary->master->driver_priv; + + i915_kernel_lost_context(dev); + + DRM_DEBUG_DRIVER("\n"); + + dev_priv->counter++; + if (dev_priv->counter > 0x7FFFFFFFUL) + dev_priv->counter = 1; + if (master_priv->sarea_priv) + master_priv->sarea_priv->last_enqueue = dev_priv->counter; + + if (BEGIN_LP_RING(4) == 0) { + OUT_RING(MI_STORE_DWORD_INDEX); + OUT_RING(I915_BREADCRUMB_INDEX << MI_STORE_DWORD_INDEX_SHIFT); + OUT_RING(dev_priv->counter); + OUT_RING(MI_USER_INTERRUPT); + ADVANCE_LP_RING(); + } + + return dev_priv->counter; +} + +static int i915_wait_irq(struct drm_device * dev, int irq_nr) +{ + drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; + struct drm_i915_master_private *master_priv = dev->primary->master->driver_priv; + int ret = 0; + struct intel_ring_buffer *ring = LP_RING(dev_priv); + + DRM_DEBUG_DRIVER("irq_nr=%d breadcrumb=%d\n", irq_nr, + READ_BREADCRUMB(dev_priv)); + + if (READ_BREADCRUMB(dev_priv) >= irq_nr) { + if (master_priv->sarea_priv) + master_priv->sarea_priv->last_dispatch = READ_BREADCRUMB(dev_priv); + return 0; + } + + if (master_priv->sarea_priv) + master_priv->sarea_priv->perf_boxes |= I915_BOX_WAIT; + + if (ring->irq_get(ring)) { + DRM_WAIT_ON(ret, ring->irq_queue, 3 * DRM_HZ, + READ_BREADCRUMB(dev_priv) >= irq_nr); + ring->irq_put(ring); + } else if (wait_for(READ_BREADCRUMB(dev_priv) >= irq_nr, 3000)) + ret = -EBUSY; + + if (ret == -EBUSY) { + DRM_ERROR("EBUSY -- rec: %d emitted: %d\n", + READ_BREADCRUMB(dev_priv), (int)dev_priv->counter); + } + + return ret; +} + +/* Needs the lock as it touches the ring. + */ +static int i915_irq_emit(struct drm_device *dev, void *data, + struct drm_file *file_priv) +{ + drm_i915_private_t *dev_priv = dev->dev_private; + drm_i915_irq_emit_t *emit = data; + int result; + + if (drm_core_check_feature(dev, DRIVER_MODESET)) + return -ENODEV; + + if (!dev_priv || !LP_RING(dev_priv)->virtual_start) { + DRM_ERROR("called with no initialization\n"); + return -EINVAL; + } + + RING_LOCK_TEST_WITH_RETURN(dev, file_priv); + + mutex_lock(&dev->struct_mutex); + result = i915_emit_irq(dev); + mutex_unlock(&dev->struct_mutex); + + if (DRM_COPY_TO_USER(emit->irq_seq, &result, sizeof(int))) { + DRM_ERROR("copy_to_user\n"); + return -EFAULT; + } + + return 0; +} + +/* Doesn't need the hardware lock. + */ +static int i915_irq_wait(struct drm_device *dev, void *data, + struct drm_file *file_priv) +{ + drm_i915_private_t *dev_priv = dev->dev_private; + drm_i915_irq_wait_t *irqwait = data; + + if (drm_core_check_feature(dev, DRIVER_MODESET)) + return -ENODEV; + + if (!dev_priv) { + DRM_ERROR("called with no initialization\n"); + return -EINVAL; + } + + return i915_wait_irq(dev, irqwait->irq_seq); +} + +static int i915_vblank_pipe_get(struct drm_device *dev, void *data, + struct drm_file *file_priv) +{ + drm_i915_private_t *dev_priv = dev->dev_private; + drm_i915_vblank_pipe_t *pipe = data; + + if (drm_core_check_feature(dev, DRIVER_MODESET)) + return -ENODEV; + + if (!dev_priv) { + DRM_ERROR("called with no initialization\n"); + return -EINVAL; + } + + pipe->pipe = DRM_I915_VBLANK_PIPE_A | DRM_I915_VBLANK_PIPE_B; + + return 0; +} + +/** + * Schedule buffer swap at given vertical blank. + */ +static int i915_vblank_swap(struct drm_device *dev, void *data, + struct drm_file *file_priv) +{ + /* The delayed swap mechanism was fundamentally racy, and has been + * removed. The model was that the client requested a delayed flip/swap + * from the kernel, then waited for vblank before continuing to perform + * rendering. The problem was that the kernel might wake the client + * up before it dispatched the vblank swap (since the lock has to be + * held while touching the ringbuffer), in which case the client would + * clear and start the next frame before the swap occurred, and + * flicker would occur in addition to likely missing the vblank. + * + * In the absence of this ioctl, userland falls back to a correct path + * of waiting for a vblank, then dispatching the swap on its own. + * Context switching to userland and back is plenty fast enough for + * meeting the requirements of vblank swapping. + */ + return -EINVAL; +} + static int i915_flip_bufs(struct drm_device *dev, void *data, struct drm_file *file_priv) { int ret; + if (drm_core_check_feature(dev, DRIVER_MODESET)) + return -ENODEV; + DRM_DEBUG_DRIVER("%s\n", __func__); RING_LOCK_TEST_WITH_RETURN(dev, file_priv); @@ -742,7 +955,7 @@ static int i915_getparam(struct drm_device *dev, void *data, value = dev->pdev->irq ? 1 : 0; break; case I915_PARAM_ALLOW_BATCHBUFFER: - value = dev_priv->allow_batchbuffer ? 1 : 0; + value = dev_priv->dri1.allow_batchbuffer ? 1 : 0; break; case I915_PARAM_LAST_DISPATCH: value = READ_BREADCRUMB(dev_priv); @@ -751,7 +964,7 @@ static int i915_getparam(struct drm_device *dev, void *data, value = dev->pci_device; break; case I915_PARAM_HAS_GEM: - value = dev_priv->has_gem; + value = 1; break; case I915_PARAM_NUM_FENCES_AVAIL: value = dev_priv->num_fence_regs - dev_priv->fence_reg_start; @@ -764,7 +977,7 @@ static int i915_getparam(struct drm_device *dev, void *data, break; case I915_PARAM_HAS_EXECBUF2: /* depends on GEM */ - value = dev_priv->has_gem; + value = 1; break; case I915_PARAM_HAS_BSD: value = HAS_BSD(dev); @@ -822,10 +1035,9 @@ static int i915_setparam(struct drm_device *dev, void *data, case I915_SETPARAM_USE_MI_BATCHBUFFER_START: break; case I915_SETPARAM_TEX_LRU_LOG_GRANULARITY: - dev_priv->tex_lru_log_granularity = param->value; break; case I915_SETPARAM_ALLOW_BATCHBUFFER: - dev_priv->allow_batchbuffer = param->value; + dev_priv->dri1.allow_batchbuffer = param->value ? 1 : 0; break; case I915_SETPARAM_NUM_USED_FENCES: if (param->value > dev_priv->num_fence_regs || @@ -850,6 +1062,9 @@ static int i915_set_status_page(struct drm_device *dev, void *data, drm_i915_hws_addr_t *hws = data; struct intel_ring_buffer *ring = LP_RING(dev_priv); + if (drm_core_check_feature(dev, DRIVER_MODESET)) + return -ENODEV; + if (!I915_NEED_GFX_HWS(dev)) return -EINVAL; @@ -867,23 +1082,17 @@ static int i915_set_status_page(struct drm_device *dev, void *data, ring->status_page.gfx_addr = hws->addr & (0x1ffff<<12); - dev_priv->hws_map.offset = dev->agp->base + hws->addr; - dev_priv->hws_map.size = 4*1024; - dev_priv->hws_map.type = 0; - dev_priv->hws_map.flags = 0; - dev_priv->hws_map.mtrr = 0; - - drm_core_ioremap_wc(&dev_priv->hws_map, dev); - if (dev_priv->hws_map.handle == NULL) { + dev_priv->dri1.gfx_hws_cpu_addr = ioremap_wc(dev->agp->base + hws->addr, + 4096); + if (dev_priv->dri1.gfx_hws_cpu_addr == NULL) { i915_dma_cleanup(dev); ring->status_page.gfx_addr = 0; DRM_ERROR("can not ioremap virtual address for" " G33 hw status page\n"); return -ENOMEM; } - ring->status_page.page_addr = - (void __force __iomem *)dev_priv->hws_map.handle; - memset_io(ring->status_page.page_addr, 0, PAGE_SIZE); + + memset_io(dev_priv->dri1.gfx_hws_cpu_addr, 0, PAGE_SIZE); I915_WRITE(HWS_PGA, ring->status_page.gfx_addr); DRM_DEBUG_DRIVER("load hws HWS_PGA with gfx mem 0x%x\n", @@ -1019,133 +1228,6 @@ intel_teardown_mchbar(struct drm_device *dev) release_resource(&dev_priv->mch_res); } -#define PTE_ADDRESS_MASK 0xfffff000 -#define PTE_ADDRESS_MASK_HIGH 0x000000f0 /* i915+ */ -#define PTE_MAPPING_TYPE_UNCACHED (0 << 1) -#define PTE_MAPPING_TYPE_DCACHE (1 << 1) /* i830 only */ -#define PTE_MAPPING_TYPE_CACHED (3 << 1) -#define PTE_MAPPING_TYPE_MASK (3 << 1) -#define PTE_VALID (1 << 0) - -/** - * i915_stolen_to_phys - take an offset into stolen memory and turn it into - * a physical one - * @dev: drm device - * @offset: address to translate - * - * Some chip functions require allocations from stolen space and need the - * physical address of the memory in question. - */ -static unsigned long i915_stolen_to_phys(struct drm_device *dev, u32 offset) -{ - struct drm_i915_private *dev_priv = dev->dev_private; - struct pci_dev *pdev = dev_priv->bridge_dev; - u32 base; - -#if 0 - /* On the machines I have tested the Graphics Base of Stolen Memory - * is unreliable, so compute the base by subtracting the stolen memory - * from the Top of Low Usable DRAM which is where the BIOS places - * the graphics stolen memory. - */ - if (INTEL_INFO(dev)->gen > 3 || IS_G33(dev)) { - /* top 32bits are reserved = 0 */ - pci_read_config_dword(pdev, 0xA4, &base); - } else { - /* XXX presume 8xx is the same as i915 */ - pci_bus_read_config_dword(pdev->bus, 2, 0x5C, &base); - } -#else - if (INTEL_INFO(dev)->gen > 3 || IS_G33(dev)) { - u16 val; - pci_read_config_word(pdev, 0xb0, &val); - base = val >> 4 << 20; - } else { - u8 val; - pci_read_config_byte(pdev, 0x9c, &val); - base = val >> 3 << 27; - } - base -= dev_priv->mm.gtt->stolen_size; -#endif - - return base + offset; -} - -static void i915_warn_stolen(struct drm_device *dev) -{ - DRM_ERROR("not enough stolen space for compressed buffer, disabling\n"); - DRM_ERROR("hint: you may be able to increase stolen memory size in the BIOS to avoid this\n"); -} - -static void i915_setup_compression(struct drm_device *dev, int size) -{ - struct drm_i915_private *dev_priv = dev->dev_private; - struct drm_mm_node *compressed_fb, *uninitialized_var(compressed_llb); - unsigned long cfb_base; - unsigned long ll_base = 0; - - /* Just in case the BIOS is doing something questionable. */ - intel_disable_fbc(dev); - - compressed_fb = drm_mm_search_free(&dev_priv->mm.stolen, size, 4096, 0); - if (compressed_fb) - compressed_fb = drm_mm_get_block(compressed_fb, size, 4096); - if (!compressed_fb) - goto err; - - cfb_base = i915_stolen_to_phys(dev, compressed_fb->start); - if (!cfb_base) - goto err_fb; - - if (!(IS_GM45(dev) || HAS_PCH_SPLIT(dev))) { - compressed_llb = drm_mm_search_free(&dev_priv->mm.stolen, - 4096, 4096, 0); - if (compressed_llb) - compressed_llb = drm_mm_get_block(compressed_llb, - 4096, 4096); - if (!compressed_llb) - goto err_fb; - - ll_base = i915_stolen_to_phys(dev, compressed_llb->start); - if (!ll_base) - goto err_llb; - } - - dev_priv->cfb_size = size; - - dev_priv->compressed_fb = compressed_fb; - if (HAS_PCH_SPLIT(dev)) - I915_WRITE(ILK_DPFC_CB_BASE, compressed_fb->start); - else if (IS_GM45(dev)) { - I915_WRITE(DPFC_CB_BASE, compressed_fb->start); - } else { - I915_WRITE(FBC_CFB_BASE, cfb_base); - I915_WRITE(FBC_LL_BASE, ll_base); - dev_priv->compressed_llb = compressed_llb; - } - - DRM_DEBUG_KMS("FBC base 0x%08lx, ll base 0x%08lx, size %dM\n", - cfb_base, ll_base, size >> 20); - return; - -err_llb: - drm_mm_put_block(compressed_llb); -err_fb: - drm_mm_put_block(compressed_fb); -err: - dev_priv->no_fbc_reason = FBC_STOLEN_TOO_SMALL; - i915_warn_stolen(dev); -} - -static void i915_cleanup_compression(struct drm_device *dev) -{ - struct drm_i915_private *dev_priv = dev->dev_private; - - drm_mm_put_block(dev_priv->compressed_fb); - if (dev_priv->compressed_llb) - drm_mm_put_block(dev_priv->compressed_llb); -} - /* true = enable decode, false = disable decoder */ static unsigned int i915_vga_set_decode(void *cookie, bool state) { @@ -1189,88 +1271,6 @@ static bool i915_switcheroo_can_switch(struct pci_dev *pdev) return can_switch; } -static bool -intel_enable_ppgtt(struct drm_device *dev) -{ - if (i915_enable_ppgtt >= 0) - return i915_enable_ppgtt; - -#ifdef CONFIG_INTEL_IOMMU - /* Disable ppgtt on SNB if VT-d is on. */ - if (INTEL_INFO(dev)->gen == 6 && intel_iommu_gfx_mapped) - return false; -#endif - - return true; -} - -static int i915_load_gem_init(struct drm_device *dev) -{ - struct drm_i915_private *dev_priv = dev->dev_private; - unsigned long prealloc_size, gtt_size, mappable_size; - int ret; - - prealloc_size = dev_priv->mm.gtt->stolen_size; - gtt_size = dev_priv->mm.gtt->gtt_total_entries << PAGE_SHIFT; - mappable_size = dev_priv->mm.gtt->gtt_mappable_entries << PAGE_SHIFT; - - /* Basic memrange allocator for stolen space */ - drm_mm_init(&dev_priv->mm.stolen, 0, prealloc_size); - - mutex_lock(&dev->struct_mutex); - if (intel_enable_ppgtt(dev) && HAS_ALIASING_PPGTT(dev)) { - /* PPGTT pdes are stolen from global gtt ptes, so shrink the - * aperture accordingly when using aliasing ppgtt. */ - gtt_size -= I915_PPGTT_PD_ENTRIES*PAGE_SIZE; - - i915_gem_init_global_gtt(dev, 0, mappable_size, gtt_size); - - ret = i915_gem_init_aliasing_ppgtt(dev); - if (ret) { - mutex_unlock(&dev->struct_mutex); - return ret; - } - } else { - /* Let GEM Manage all of the aperture. - * - * However, leave one page at the end still bound to the scratch - * page. There are a number of places where the hardware - * apparently prefetches past the end of the object, and we've - * seen multiple hangs with the GPU head pointer stuck in a - * batchbuffer bound at the last page of the aperture. One page - * should be enough to keep any prefetching inside of the - * aperture. - */ - i915_gem_init_global_gtt(dev, 0, mappable_size, - gtt_size); - } - - ret = i915_gem_init_hw(dev); - mutex_unlock(&dev->struct_mutex); - if (ret) { - i915_gem_cleanup_aliasing_ppgtt(dev); - return ret; - } - - /* Try to set up FBC with a reasonable compressed buffer size */ - if (I915_HAS_FBC(dev) && i915_powersave) { - int cfb_size; - - /* Leave 1M for line length buffer & misc. */ - - /* Try to get a 32M buffer... */ - if (prealloc_size > (36*1024*1024)) - cfb_size = 32*1024*1024; - else /* fall back to 7/8 of the stolen space */ - cfb_size = prealloc_size * 7 / 8; - i915_setup_compression(dev, cfb_size); - } - - /* Allow hardware batchbuffers unless told otherwise. */ - dev_priv->allow_batchbuffer = 1; - return 0; -} - static int i915_load_modeset_init(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; @@ -1300,15 +1300,18 @@ static int i915_load_modeset_init(struct drm_device *dev) if (ret) goto cleanup_vga_client; - /* IIR "flip pending" bit means done if this bit is set */ - if (IS_GEN3(dev) && (I915_READ(ECOSKPD) & ECO_FLIP_DONE)) - dev_priv->flip_pending_is_done = true; + /* Initialise stolen first so that we may reserve preallocated + * objects for the BIOS to KMS transition. + */ + ret = i915_gem_init_stolen(dev); + if (ret) + goto cleanup_vga_switcheroo; intel_modeset_init(dev); - ret = i915_load_gem_init(dev); + ret = i915_gem_init(dev); if (ret) - goto cleanup_vga_switcheroo; + goto cleanup_gem_stolen; intel_modeset_gem_init(dev); @@ -1338,6 +1341,8 @@ cleanup_gem: i915_gem_cleanup_ringbuffer(dev); mutex_unlock(&dev->struct_mutex); i915_gem_cleanup_aliasing_ppgtt(dev); +cleanup_gem_stolen: + i915_gem_cleanup_stolen(dev); cleanup_vga_switcheroo: vga_switcheroo_unregister_client(dev->pdev); cleanup_vga_client: @@ -1370,575 +1375,6 @@ void i915_master_destroy(struct drm_device *dev, struct drm_master *master) master->driver_priv = NULL; } -static void i915_pineview_get_mem_freq(struct drm_device *dev) -{ - drm_i915_private_t *dev_priv = dev->dev_private; - u32 tmp; - - tmp = I915_READ(CLKCFG); - - switch (tmp & CLKCFG_FSB_MASK) { - case CLKCFG_FSB_533: - dev_priv->fsb_freq = 533; /* 133*4 */ - break; - case CLKCFG_FSB_800: - dev_priv->fsb_freq = 800; /* 200*4 */ - break; - case CLKCFG_FSB_667: - dev_priv->fsb_freq = 667; /* 167*4 */ - break; - case CLKCFG_FSB_400: - dev_priv->fsb_freq = 400; /* 100*4 */ - break; - } - - switch (tmp & CLKCFG_MEM_MASK) { - case CLKCFG_MEM_533: - dev_priv->mem_freq = 533; - break; - case CLKCFG_MEM_667: - dev_priv->mem_freq = 667; - break; - case CLKCFG_MEM_800: - dev_priv->mem_freq = 800; - break; - } - - /* detect pineview DDR3 setting */ - tmp = I915_READ(CSHRDDR3CTL); - dev_priv->is_ddr3 = (tmp & CSHRDDR3CTL_DDR3) ? 1 : 0; -} - -static void i915_ironlake_get_mem_freq(struct drm_device *dev) -{ - drm_i915_private_t *dev_priv = dev->dev_private; - u16 ddrpll, csipll; - - ddrpll = I915_READ16(DDRMPLL1); - csipll = I915_READ16(CSIPLL0); - - switch (ddrpll & 0xff) { - case 0xc: - dev_priv->mem_freq = 800; - break; - case 0x10: - dev_priv->mem_freq = 1066; - break; - case 0x14: - dev_priv->mem_freq = 1333; - break; - case 0x18: - dev_priv->mem_freq = 1600; - break; - default: - DRM_DEBUG_DRIVER("unknown memory frequency 0x%02x\n", - ddrpll & 0xff); - dev_priv->mem_freq = 0; - break; - } - - dev_priv->r_t = dev_priv->mem_freq; - - switch (csipll & 0x3ff) { - case 0x00c: - dev_priv->fsb_freq = 3200; - break; - case 0x00e: - dev_priv->fsb_freq = 3733; - break; - case 0x010: - dev_priv->fsb_freq = 4266; - break; - case 0x012: - dev_priv->fsb_freq = 4800; - break; - case 0x014: - dev_priv->fsb_freq = 5333; - break; - case 0x016: - dev_priv->fsb_freq = 5866; - break; - case 0x018: - dev_priv->fsb_freq = 6400; - break; - default: - DRM_DEBUG_DRIVER("unknown fsb frequency 0x%04x\n", - csipll & 0x3ff); - dev_priv->fsb_freq = 0; - break; - } - - if (dev_priv->fsb_freq == 3200) { - dev_priv->c_m = 0; - } else if (dev_priv->fsb_freq > 3200 && dev_priv->fsb_freq <= 4800) { - dev_priv->c_m = 1; - } else { - dev_priv->c_m = 2; - } -} - -static const struct cparams { - u16 i; - u16 t; - u16 m; - u16 c; -} cparams[] = { - { 1, 1333, 301, 28664 }, - { 1, 1066, 294, 24460 }, - { 1, 800, 294, 25192 }, - { 0, 1333, 276, 27605 }, - { 0, 1066, 276, 27605 }, - { 0, 800, 231, 23784 }, -}; - -unsigned long i915_chipset_val(struct drm_i915_private *dev_priv) -{ - u64 total_count, diff, ret; - u32 count1, count2, count3, m = 0, c = 0; - unsigned long now = jiffies_to_msecs(jiffies), diff1; - int i; - - diff1 = now - dev_priv->last_time1; - - /* Prevent division-by-zero if we are asking too fast. - * Also, we don't get interesting results if we are polling - * faster than once in 10ms, so just return the saved value - * in such cases. - */ - if (diff1 <= 10) - return dev_priv->chipset_power; - - count1 = I915_READ(DMIEC); - count2 = I915_READ(DDREC); - count3 = I915_READ(CSIEC); - - total_count = count1 + count2 + count3; - - /* FIXME: handle per-counter overflow */ - if (total_count < dev_priv->last_count1) { - diff = ~0UL - dev_priv->last_count1; - diff += total_count; - } else { - diff = total_count - dev_priv->last_count1; - } - - for (i = 0; i < ARRAY_SIZE(cparams); i++) { - if (cparams[i].i == dev_priv->c_m && - cparams[i].t == dev_priv->r_t) { - m = cparams[i].m; - c = cparams[i].c; - break; - } - } - - diff = div_u64(diff, diff1); - ret = ((m * diff) + c); - ret = div_u64(ret, 10); - - dev_priv->last_count1 = total_count; - dev_priv->last_time1 = now; - - dev_priv->chipset_power = ret; - - return ret; -} - -unsigned long i915_mch_val(struct drm_i915_private *dev_priv) -{ - unsigned long m, x, b; - u32 tsfs; - - tsfs = I915_READ(TSFS); - - m = ((tsfs & TSFS_SLOPE_MASK) >> TSFS_SLOPE_SHIFT); - x = I915_READ8(TR1); - - b = tsfs & TSFS_INTR_MASK; - - return ((m * x) / 127) - b; -} - -static u16 pvid_to_extvid(struct drm_i915_private *dev_priv, u8 pxvid) -{ - static const struct v_table { - u16 vd; /* in .1 mil */ - u16 vm; /* in .1 mil */ - } v_table[] = { - { 0, 0, }, - { 375, 0, }, - { 500, 0, }, - { 625, 0, }, - { 750, 0, }, - { 875, 0, }, - { 1000, 0, }, - { 1125, 0, }, - { 4125, 3000, }, - { 4125, 3000, }, - { 4125, 3000, }, - { 4125, 3000, }, - { 4125, 3000, }, - { 4125, 3000, }, - { 4125, 3000, }, - { 4125, 3000, }, - { 4125, 3000, }, - { 4125, 3000, }, - { 4125, 3000, }, - { 4125, 3000, }, - { 4125, 3000, }, - { 4125, 3000, }, - { 4125, 3000, }, - { 4125, 3000, }, - { 4125, 3000, }, - { 4125, 3000, }, - { 4125, 3000, }, - { 4125, 3000, }, - { 4125, 3000, }, - { 4125, 3000, }, - { 4125, 3000, }, - { 4125, 3000, }, - { 4250, 3125, }, - { 4375, 3250, }, - { 4500, 3375, }, - { 4625, 3500, }, - { 4750, 3625, }, - { 4875, 3750, }, - { 5000, 3875, }, - { 5125, 4000, }, - { 5250, 4125, }, - { 5375, 4250, }, - { 5500, 4375, }, - { 5625, 4500, }, - { 5750, 4625, }, - { 5875, 4750, }, - { 6000, 4875, }, - { 6125, 5000, }, - { 6250, 5125, }, - { 6375, 5250, }, - { 6500, 5375, }, - { 6625, 5500, }, - { 6750, 5625, }, - { 6875, 5750, }, - { 7000, 5875, }, - { 7125, 6000, }, - { 7250, 6125, }, - { 7375, 6250, }, - { 7500, 6375, }, - { 7625, 6500, }, - { 7750, 6625, }, - { 7875, 6750, }, - { 8000, 6875, }, - { 8125, 7000, }, - { 8250, 7125, }, - { 8375, 7250, }, - { 8500, 7375, }, - { 8625, 7500, }, - { 8750, 7625, }, - { 8875, 7750, }, - { 9000, 7875, }, - { 9125, 8000, }, - { 9250, 8125, }, - { 9375, 8250, }, - { 9500, 8375, }, - { 9625, 8500, }, - { 9750, 8625, }, - { 9875, 8750, }, - { 10000, 8875, }, - { 10125, 9000, }, - { 10250, 9125, }, - { 10375, 9250, }, - { 10500, 9375, }, - { 10625, 9500, }, - { 10750, 9625, }, - { 10875, 9750, }, - { 11000, 9875, }, - { 11125, 10000, }, - { 11250, 10125, }, - { 11375, 10250, }, - { 11500, 10375, }, - { 11625, 10500, }, - { 11750, 10625, }, - { 11875, 10750, }, - { 12000, 10875, }, - { 12125, 11000, }, - { 12250, 11125, }, - { 12375, 11250, }, - { 12500, 11375, }, - { 12625, 11500, }, - { 12750, 11625, }, - { 12875, 11750, }, - { 13000, 11875, }, - { 13125, 12000, }, - { 13250, 12125, }, - { 13375, 12250, }, - { 13500, 12375, }, - { 13625, 12500, }, - { 13750, 12625, }, - { 13875, 12750, }, - { 14000, 12875, }, - { 14125, 13000, }, - { 14250, 13125, }, - { 14375, 13250, }, - { 14500, 13375, }, - { 14625, 13500, }, - { 14750, 13625, }, - { 14875, 13750, }, - { 15000, 13875, }, - { 15125, 14000, }, - { 15250, 14125, }, - { 15375, 14250, }, - { 15500, 14375, }, - { 15625, 14500, }, - { 15750, 14625, }, - { 15875, 14750, }, - { 16000, 14875, }, - { 16125, 15000, }, - }; - if (dev_priv->info->is_mobile) - return v_table[pxvid].vm; - else - return v_table[pxvid].vd; -} - -void i915_update_gfx_val(struct drm_i915_private *dev_priv) -{ - struct timespec now, diff1; - u64 diff; - unsigned long diffms; - u32 count; - - if (dev_priv->info->gen != 5) - return; - - getrawmonotonic(&now); - diff1 = timespec_sub(now, dev_priv->last_time2); - - /* Don't divide by 0 */ - diffms = diff1.tv_sec * 1000 + diff1.tv_nsec / 1000000; - if (!diffms) - return; - - count = I915_READ(GFXEC); - - if (count < dev_priv->last_count2) { - diff = ~0UL - dev_priv->last_count2; - diff += count; - } else { - diff = count - dev_priv->last_count2; - } - - dev_priv->last_count2 = count; - dev_priv->last_time2 = now; - - /* More magic constants... */ - diff = diff * 1181; - diff = div_u64(diff, diffms * 10); - dev_priv->gfx_power = diff; -} - -unsigned long i915_gfx_val(struct drm_i915_private *dev_priv) -{ - unsigned long t, corr, state1, corr2, state2; - u32 pxvid, ext_v; - - pxvid = I915_READ(PXVFREQ_BASE + (dev_priv->cur_delay * 4)); - pxvid = (pxvid >> 24) & 0x7f; - ext_v = pvid_to_extvid(dev_priv, pxvid); - - state1 = ext_v; - - t = i915_mch_val(dev_priv); - - /* Revel in the empirically derived constants */ - - /* Correction factor in 1/100000 units */ - if (t > 80) - corr = ((t * 2349) + 135940); - else if (t >= 50) - corr = ((t * 964) + 29317); - else /* < 50 */ - corr = ((t * 301) + 1004); - - corr = corr * ((150142 * state1) / 10000 - 78642); - corr /= 100000; - corr2 = (corr * dev_priv->corr); - - state2 = (corr2 * state1) / 10000; - state2 /= 100; /* convert to mW */ - - i915_update_gfx_val(dev_priv); - - return dev_priv->gfx_power + state2; -} - -/* Global for IPS driver to get at the current i915 device */ -static struct drm_i915_private *i915_mch_dev; -/* - * Lock protecting IPS related data structures - * - i915_mch_dev - * - dev_priv->max_delay - * - dev_priv->min_delay - * - dev_priv->fmax - * - dev_priv->gpu_busy - */ -static DEFINE_SPINLOCK(mchdev_lock); - -/** - * i915_read_mch_val - return value for IPS use - * - * Calculate and return a value for the IPS driver to use when deciding whether - * we have thermal and power headroom to increase CPU or GPU power budget. - */ -unsigned long i915_read_mch_val(void) -{ - struct drm_i915_private *dev_priv; - unsigned long chipset_val, graphics_val, ret = 0; - - spin_lock(&mchdev_lock); - if (!i915_mch_dev) - goto out_unlock; - dev_priv = i915_mch_dev; - - chipset_val = i915_chipset_val(dev_priv); - graphics_val = i915_gfx_val(dev_priv); - - ret = chipset_val + graphics_val; - -out_unlock: - spin_unlock(&mchdev_lock); - - return ret; -} -EXPORT_SYMBOL_GPL(i915_read_mch_val); - -/** - * i915_gpu_raise - raise GPU frequency limit - * - * Raise the limit; IPS indicates we have thermal headroom. - */ -bool i915_gpu_raise(void) -{ - struct drm_i915_private *dev_priv; - bool ret = true; - - spin_lock(&mchdev_lock); - if (!i915_mch_dev) { - ret = false; - goto out_unlock; - } - dev_priv = i915_mch_dev; - - if (dev_priv->max_delay > dev_priv->fmax) - dev_priv->max_delay--; - -out_unlock: - spin_unlock(&mchdev_lock); - - return ret; -} -EXPORT_SYMBOL_GPL(i915_gpu_raise); - -/** - * i915_gpu_lower - lower GPU frequency limit - * - * IPS indicates we're close to a thermal limit, so throttle back the GPU - * frequency maximum. - */ -bool i915_gpu_lower(void) -{ - struct drm_i915_private *dev_priv; - bool ret = true; - - spin_lock(&mchdev_lock); - if (!i915_mch_dev) { - ret = false; - goto out_unlock; - } - dev_priv = i915_mch_dev; - - if (dev_priv->max_delay < dev_priv->min_delay) - dev_priv->max_delay++; - -out_unlock: - spin_unlock(&mchdev_lock); - - return ret; -} -EXPORT_SYMBOL_GPL(i915_gpu_lower); - -/** - * i915_gpu_busy - indicate GPU business to IPS - * - * Tell the IPS driver whether or not the GPU is busy. - */ -bool i915_gpu_busy(void) -{ - struct drm_i915_private *dev_priv; - bool ret = false; - - spin_lock(&mchdev_lock); - if (!i915_mch_dev) - goto out_unlock; - dev_priv = i915_mch_dev; - - ret = dev_priv->busy; - -out_unlock: - spin_unlock(&mchdev_lock); - - return ret; -} -EXPORT_SYMBOL_GPL(i915_gpu_busy); - -/** - * i915_gpu_turbo_disable - disable graphics turbo - * - * Disable graphics turbo by resetting the max frequency and setting the - * current frequency to the default. - */ -bool i915_gpu_turbo_disable(void) -{ - struct drm_i915_private *dev_priv; - bool ret = true; - - spin_lock(&mchdev_lock); - if (!i915_mch_dev) { - ret = false; - goto out_unlock; - } - dev_priv = i915_mch_dev; - - dev_priv->max_delay = dev_priv->fstart; - - if (!ironlake_set_drps(dev_priv->dev, dev_priv->fstart)) - ret = false; - -out_unlock: - spin_unlock(&mchdev_lock); - - return ret; -} -EXPORT_SYMBOL_GPL(i915_gpu_turbo_disable); - -/** - * Tells the intel_ips driver that the i915 driver is now loaded, if - * IPS got loaded first. - * - * This awkward dance is so that neither module has to depend on the - * other in order for IPS to do the appropriate communication of - * GPU turbo limits to i915. - */ -static void -ips_ping_for_i915_load(void) -{ - void (*link)(void); - - link = symbol_get(ips_link_to_i915_driver); - if (link) { - link(); - symbol_put(ips_link_to_i915_driver); - } -} - static void i915_mtrr_setup(struct drm_i915_private *dev_priv, unsigned long base, unsigned long size) @@ -2072,9 +1508,6 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags) goto out_mtrrfree; } - /* enable GEM by default */ - dev_priv->has_gem = 1; - intel_irq_init(dev); /* Try to make sure MCHBAR is enabled before poking at it */ @@ -2094,11 +1527,6 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags) goto out_gem_unload; } - if (IS_PINEVIEW(dev)) - i915_pineview_get_mem_freq(dev); - else if (IS_GEN5(dev)) - i915_ironlake_get_mem_freq(dev); - /* On the 945G/GM, the chipset reports the MSI capability on the * integrated graphics even though the support isn't actually there * according to the published specs. It doesn't appear to function @@ -2151,14 +1579,8 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags) setup_timer(&dev_priv->hangcheck_timer, i915_hangcheck_elapsed, (unsigned long) dev); - if (IS_GEN5(dev)) { - spin_lock(&mchdev_lock); - i915_mch_dev = dev_priv; - dev_priv->mchdev_lock = &mchdev_lock; - spin_unlock(&mchdev_lock); - - ips_ping_for_i915_load(); - } + if (IS_GEN5(dev)) + intel_gpu_ips_init(dev_priv); return 0; @@ -2193,9 +1615,7 @@ int i915_driver_unload(struct drm_device *dev) struct drm_i915_private *dev_priv = dev->dev_private; int ret; - spin_lock(&mchdev_lock); - i915_mch_dev = NULL; - spin_unlock(&mchdev_lock); + intel_gpu_ips_teardown(); i915_teardown_sysfs(dev); @@ -2203,9 +1623,10 @@ int i915_driver_unload(struct drm_device *dev) unregister_shrinker(&dev_priv->mm.inactive_shrinker); mutex_lock(&dev->struct_mutex); - ret = i915_gpu_idle(dev, true); + ret = i915_gpu_idle(dev); if (ret) DRM_ERROR("failed to idle hardware: %d\n", ret); + i915_gem_retire_requests(dev); mutex_unlock(&dev->struct_mutex); /* Cancel the retire work handler, which should be idle now. */ @@ -2257,8 +1678,7 @@ int i915_driver_unload(struct drm_device *dev) i915_gem_cleanup_ringbuffer(dev); mutex_unlock(&dev->struct_mutex); i915_gem_cleanup_aliasing_ppgtt(dev); - if (I915_HAS_FBC(dev) && i915_powersave) - i915_cleanup_compression(dev); + i915_gem_cleanup_stolen(dev); drm_mm_takedown(&dev_priv->mm.stolen); intel_cleanup_overlay(dev); @@ -2351,7 +1771,7 @@ struct drm_ioctl_desc i915_ioctls[] = { DRM_IOCTL_DEF_DRV(I915_INIT_HEAP, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY), DRM_IOCTL_DEF_DRV(I915_CMDBUFFER, i915_cmdbuffer, DRM_AUTH), DRM_IOCTL_DEF_DRV(I915_DESTROY_HEAP, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY), - DRM_IOCTL_DEF_DRV(I915_SET_VBLANK_PIPE, i915_vblank_pipe_set, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY), + DRM_IOCTL_DEF_DRV(I915_SET_VBLANK_PIPE, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY), DRM_IOCTL_DEF_DRV(I915_GET_VBLANK_PIPE, i915_vblank_pipe_get, DRM_AUTH), DRM_IOCTL_DEF_DRV(I915_VBLANK_SWAP, i915_vblank_swap, DRM_AUTH), DRM_IOCTL_DEF_DRV(I915_HWS_ADDR, i915_set_status_page, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY), diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c index 3effcf7..77b7a50 100644 --- a/drivers/gpu/drm/i915/i915_drv.c +++ b/drivers/gpu/drm/i915/i915_drv.c @@ -377,18 +377,23 @@ void intel_detect_pch(struct drm_device *dev) if (id == INTEL_PCH_IBX_DEVICE_ID_TYPE) { dev_priv->pch_type = PCH_IBX; + dev_priv->num_pch_pll = 2; DRM_DEBUG_KMS("Found Ibex Peak PCH\n"); } else if (id == INTEL_PCH_CPT_DEVICE_ID_TYPE) { dev_priv->pch_type = PCH_CPT; + dev_priv->num_pch_pll = 2; DRM_DEBUG_KMS("Found CougarPoint PCH\n"); } else if (id == INTEL_PCH_PPT_DEVICE_ID_TYPE) { /* PantherPoint is CPT compatible */ dev_priv->pch_type = PCH_CPT; + dev_priv->num_pch_pll = 2; DRM_DEBUG_KMS("Found PatherPoint PCH\n"); } else if (id == INTEL_PCH_LPT_DEVICE_ID_TYPE) { dev_priv->pch_type = PCH_LPT; + dev_priv->num_pch_pll = 0; DRM_DEBUG_KMS("Found LynxPoint PCH\n"); } + BUG_ON(dev_priv->num_pch_pll > I915_NUM_PLLS); } pci_dev_put(pch); } @@ -433,7 +438,7 @@ void __gen6_gt_force_wake_mt_get(struct drm_i915_private *dev_priv) while (count++ < 50 && (I915_READ_NOTRACE(FORCEWAKE_MT_ACK) & 1)) udelay(10); - I915_WRITE_NOTRACE(FORCEWAKE_MT, (1<<16) | 1); + I915_WRITE_NOTRACE(FORCEWAKE_MT, _MASKED_BIT_ENABLE(1)); POSTING_READ(FORCEWAKE_MT); count = 0; @@ -475,7 +480,7 @@ void __gen6_gt_force_wake_put(struct drm_i915_private *dev_priv) void __gen6_gt_force_wake_mt_put(struct drm_i915_private *dev_priv) { - I915_WRITE_NOTRACE(FORCEWAKE_MT, (1<<16) | 0); + I915_WRITE_NOTRACE(FORCEWAKE_MT, _MASKED_BIT_DISABLE(1)); /* The below doubles as a POSTING_READ */ gen6_gt_check_fifodbg(dev_priv); } @@ -668,7 +673,7 @@ int i915_resume(struct drm_device *dev) return 0; } -static int i8xx_do_reset(struct drm_device *dev, u8 flags) +static int i8xx_do_reset(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; @@ -702,11 +707,12 @@ static int i965_reset_complete(struct drm_device *dev) { u8 gdrst; pci_read_config_byte(dev->pdev, I965_GDRST, &gdrst); - return gdrst & 0x1; + return (gdrst & GRDOM_RESET_ENABLE) == 0; } -static int i965_do_reset(struct drm_device *dev, u8 flags) +static int i965_do_reset(struct drm_device *dev) { + int ret; u8 gdrst; /* @@ -715,20 +721,43 @@ static int i965_do_reset(struct drm_device *dev, u8 flags) * triggers the reset; when done, the hardware will clear it. */ pci_read_config_byte(dev->pdev, I965_GDRST, &gdrst); - pci_write_config_byte(dev->pdev, I965_GDRST, gdrst | flags | 0x1); + pci_write_config_byte(dev->pdev, I965_GDRST, + gdrst | GRDOM_RENDER | + GRDOM_RESET_ENABLE); + ret = wait_for(i965_reset_complete(dev), 500); + if (ret) + return ret; + + /* We can't reset render&media without also resetting display ... */ + pci_read_config_byte(dev->pdev, I965_GDRST, &gdrst); + pci_write_config_byte(dev->pdev, I965_GDRST, + gdrst | GRDOM_MEDIA | + GRDOM_RESET_ENABLE); return wait_for(i965_reset_complete(dev), 500); } -static int ironlake_do_reset(struct drm_device *dev, u8 flags) +static int ironlake_do_reset(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; - u32 gdrst = I915_READ(MCHBAR_MIRROR_BASE + ILK_GDSR); - I915_WRITE(MCHBAR_MIRROR_BASE + ILK_GDSR, gdrst | flags | 0x1); + u32 gdrst; + int ret; + + gdrst = I915_READ(MCHBAR_MIRROR_BASE + ILK_GDSR); + I915_WRITE(MCHBAR_MIRROR_BASE + ILK_GDSR, + gdrst | GRDOM_RENDER | GRDOM_RESET_ENABLE); + ret = wait_for(I915_READ(MCHBAR_MIRROR_BASE + ILK_GDSR) & 0x1, 500); + if (ret) + return ret; + + /* We can't reset render&media without also resetting display ... */ + gdrst = I915_READ(MCHBAR_MIRROR_BASE + ILK_GDSR); + I915_WRITE(MCHBAR_MIRROR_BASE + ILK_GDSR, + gdrst | GRDOM_MEDIA | GRDOM_RESET_ENABLE); return wait_for(I915_READ(MCHBAR_MIRROR_BASE + ILK_GDSR) & 0x1, 500); } -static int gen6_do_reset(struct drm_device *dev, u8 flags) +static int gen6_do_reset(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; int ret; @@ -763,10 +792,44 @@ static int gen6_do_reset(struct drm_device *dev, u8 flags) return ret; } +static int intel_gpu_reset(struct drm_device *dev) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + int ret = -ENODEV; + + switch (INTEL_INFO(dev)->gen) { + case 7: + case 6: + ret = gen6_do_reset(dev); + break; + case 5: + ret = ironlake_do_reset(dev); + break; + case 4: + ret = i965_do_reset(dev); + break; + case 2: + ret = i8xx_do_reset(dev); + break; + } + + /* Also reset the gpu hangman. */ + if (dev_priv->stop_rings) { + DRM_DEBUG("Simulated gpu hang, resetting stop_rings\n"); + dev_priv->stop_rings = 0; + if (ret == -ENODEV) { + DRM_ERROR("Reset not implemented, but ignoring " + "error for simulated gpu hangs\n"); + ret = 0; + } + } + + return ret; +} + /** * i915_reset - reset chip after a hang * @dev: drm device to reset - * @flags: reset domains * * Reset the chip. Useful if a hang is detected. Returns zero on successful * reset or otherwise an error code. @@ -779,14 +842,9 @@ static int gen6_do_reset(struct drm_device *dev, u8 flags) * - re-init interrupt state * - re-init display */ -int i915_reset(struct drm_device *dev, u8 flags) +int i915_reset(struct drm_device *dev) { drm_i915_private_t *dev_priv = dev->dev_private; - /* - * We really should only reset the display subsystem if we actually - * need to - */ - bool need_display = true; int ret; if (!i915_try_reset) @@ -795,26 +853,16 @@ int i915_reset(struct drm_device *dev, u8 flags) if (!mutex_trylock(&dev->struct_mutex)) return -EBUSY; + dev_priv->stop_rings = 0; + i915_gem_reset(dev); ret = -ENODEV; - if (get_seconds() - dev_priv->last_gpu_reset < 5) { + if (get_seconds() - dev_priv->last_gpu_reset < 5) DRM_ERROR("GPU hanging too fast, declaring wedged!\n"); - } else switch (INTEL_INFO(dev)->gen) { - case 7: - case 6: - ret = gen6_do_reset(dev, flags); - break; - case 5: - ret = ironlake_do_reset(dev, flags); - break; - case 4: - ret = i965_do_reset(dev, flags); - break; - case 2: - ret = i8xx_do_reset(dev, flags); - break; - } + else + ret = intel_gpu_reset(dev); + dev_priv->last_gpu_reset = get_seconds(); if (ret) { DRM_ERROR("Failed to reset chip.\n"); @@ -856,23 +904,9 @@ int i915_reset(struct drm_device *dev, u8 flags) intel_modeset_init_hw(dev); drm_irq_uninstall(dev); - drm_mode_config_reset(dev); drm_irq_install(dev); - - mutex_lock(&dev->struct_mutex); - } - - mutex_unlock(&dev->struct_mutex); - - /* - * Perform a full modeset as on later generations, e.g. Ironlake, we may - * need to retrain the display link and cannot just restore the register - * values. - */ - if (need_display) { - mutex_lock(&dev->mode_config.mutex); - drm_helper_resume_force_mode(dev); - mutex_unlock(&dev->mode_config.mutex); + } else { + mutex_unlock(&dev->struct_mutex); } return 0; diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 69e1539..e03a4f8 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -39,6 +39,7 @@ #include <drm/intel-gtt.h> #include <linux/backlight.h> #include <linux/intel-iommu.h> +#include <linux/kref.h> /* General customization: */ @@ -78,6 +79,16 @@ enum port { #define for_each_pipe(p) for ((p) = 0; (p) < dev_priv->num_pipe; (p)++) +struct intel_pch_pll { + int refcount; /* count of number of CRTCs sharing this PLL */ + int active; /* count of number of active CRTCs (i.e. DPMS on) */ + bool on; /* is the PLL actually active? Disabled during modeset */ + int pll_reg; + int fp0_reg; + int fp1_reg; +}; +#define I915_NUM_PLLS 2 + /* Interface history: * * 1.1: Original. @@ -122,11 +133,11 @@ struct opregion_asle; struct drm_i915_private; struct intel_opregion { - struct opregion_header *header; - struct opregion_acpi *acpi; - struct opregion_swsci *swsci; - struct opregion_asle *asle; - void *vbt; + struct opregion_header __iomem *header; + struct opregion_acpi __iomem *acpi; + struct opregion_swsci __iomem *swsci; + struct opregion_asle __iomem *asle; + void __iomem *vbt; u32 __iomem *lid_state; }; #define OPREGION_SIZE (8*1024) @@ -161,8 +172,11 @@ struct sdvo_device_mapping { struct intel_display_error_state; struct drm_i915_error_state { + struct kref ref; u32 eir; u32 pgtbl_er; + u32 ier; + bool waiting[I915_NUM_RINGS]; u32 pipestat[I915_MAX_PIPES]; u32 tail[I915_NUM_RINGS]; u32 head[I915_NUM_RINGS]; @@ -228,11 +242,13 @@ struct drm_i915_display_funcs { void (*update_wm)(struct drm_device *dev); void (*update_sprite_wm)(struct drm_device *dev, int pipe, uint32_t sprite_width, int pixel_size); + void (*sanitize_pm)(struct drm_device *dev); int (*crtc_mode_set)(struct drm_crtc *crtc, struct drm_display_mode *mode, struct drm_display_mode *adjusted_mode, int x, int y, struct drm_framebuffer *old_fb); + void (*off)(struct drm_crtc *crtc); void (*write_eld)(struct drm_connector *connector, struct drm_crtc *crtc); void (*fdi_link_train)(struct drm_crtc *crtc); @@ -328,7 +344,6 @@ typedef struct drm_i915_private { const struct intel_device_info *info; - int has_gem; int relative_constants_mode; void __iomem *regs; @@ -357,7 +372,6 @@ typedef struct drm_i915_private { drm_dma_handle_t *status_page_dmah; uint32_t counter; - drm_local_map_t hws_map; struct drm_i915_gem_object *pwrctx; struct drm_i915_gem_object *renderctx; @@ -386,11 +400,9 @@ typedef struct drm_i915_private { u32 hotplug_supported_mask; struct work_struct hotplug_work; - int tex_lru_log_granularity; - int allow_batchbuffer; unsigned int sr01, adpa, ppcr, dvob, dvoc, lvds; - int vblank_pipe; int num_pipe; + int num_pch_pll; /* For hangcheck timer */ #define DRM_I915_HANGCHECK_PERIOD 1500 /* in ms */ @@ -402,6 +414,8 @@ typedef struct drm_i915_private { uint32_t last_instdone; uint32_t last_instdone1; + unsigned int stop_rings; + unsigned long cfb_size; unsigned int cfb_fb; enum plane cfb_plane; @@ -453,6 +467,7 @@ typedef struct drm_i915_private { unsigned int fsb_freq, mem_freq, is_ddr3; spinlock_t error_lock; + /* Protected by dev->error_lock. */ struct drm_i915_error_state *first_error; struct work_struct error_work; struct completion error_completion; @@ -677,24 +692,10 @@ typedef struct drm_i915_private { */ struct list_head inactive_list; - /** - * LRU list of objects which are not in the ringbuffer but - * are still pinned in the GTT. - */ - struct list_head pinned_list; - /** LRU list of objects with fence regs on them. */ struct list_head fence_list; /** - * List of objects currently pending being freed. - * - * These objects are no longer in use, but due to a signal - * we were prevented from freeing them at the appointed time. - */ - struct list_head deferred_free_list; - - /** * We leave the user IRQ off as much as possible, * but this means that requests will finish and never * be retired once the system goes idle. Set a timer to @@ -742,6 +743,16 @@ typedef struct drm_i915_private { size_t object_memory; u32 object_count; } mm; + + /* Old dri1 support infrastructure, beware the dragons ya fools entering + * here! */ + struct { + unsigned allow_batchbuffer : 1; + u32 __iomem *gfx_hws_cpu_addr; + } dri1; + + /* Kernel Modesetting */ + struct sdvo_device_mapping sdvo_mappings[2]; /* indicate whether the LVDS_BORDER should be enabled or not */ unsigned int lvds_border_bits; @@ -751,7 +762,8 @@ typedef struct drm_i915_private { struct drm_crtc *plane_to_crtc_mapping[3]; struct drm_crtc *pipe_to_crtc_mapping[3]; wait_queue_head_t pending_flip_queue; - bool flip_pending_is_done; + + struct intel_pch_pll pch_plls[I915_NUM_PLLS]; /* Reclocking support */ bool render_reclock_avail; @@ -869,7 +881,14 @@ struct drm_i915_gem_object { * Current tiling mode for the object. */ unsigned int tiling_mode:2; - unsigned int tiling_changed:1; + /** + * Whether the tiling parameters for the currently associated fence + * register have changed. Note that for the purposes of tracking + * tiling changes we also treat the unfenced register, the register + * slot that the object occupies whilst it executes a fenced + * command (such as BLT on gen2/3), as a "fence". + */ + unsigned int fence_dirty:1; /** How many users have pinned this object in GTT space. The following * users can each hold at most one reference: pwrite/pread, pin_ioctl @@ -1116,6 +1135,7 @@ extern int i915_master_create(struct drm_device *dev, struct drm_master *master) extern void i915_master_destroy(struct drm_device *dev, struct drm_master *master); /* i915_dma.c */ +void i915_update_dri1_breadcrumb(struct drm_device *dev); extern void i915_kernel_lost_context(struct drm_device * dev); extern int i915_driver_load(struct drm_device *, unsigned long flags); extern int i915_driver_unload(struct drm_device *); @@ -1133,7 +1153,7 @@ extern long i915_compat_ioctl(struct file *filp, unsigned int cmd, extern int i915_emit_box(struct drm_device *dev, struct drm_clip_rect *box, int DR1, int DR4); -extern int i915_reset(struct drm_device *dev, u8 flags); +extern int i915_reset(struct drm_device *dev); extern unsigned long i915_chipset_val(struct drm_i915_private *dev_priv); extern unsigned long i915_mch_val(struct drm_i915_private *dev_priv); extern unsigned long i915_gfx_val(struct drm_i915_private *dev_priv); @@ -1143,19 +1163,10 @@ extern void i915_update_gfx_val(struct drm_i915_private *dev_priv); /* i915_irq.c */ void i915_hangcheck_elapsed(unsigned long data); void i915_handle_error(struct drm_device *dev, bool wedged); -extern int i915_irq_emit(struct drm_device *dev, void *data, - struct drm_file *file_priv); -extern int i915_irq_wait(struct drm_device *dev, void *data, - struct drm_file *file_priv); extern void intel_irq_init(struct drm_device *dev); -extern int i915_vblank_pipe_set(struct drm_device *dev, void *data, - struct drm_file *file_priv); -extern int i915_vblank_pipe_get(struct drm_device *dev, void *data, - struct drm_file *file_priv); -extern int i915_vblank_swap(struct drm_device *dev, void *data, - struct drm_file *file_priv); +void i915_error_state_free(struct kref *error_ref); void i915_enable_pipestat(drm_i915_private_t *dev_priv, int pipe, u32 mask); @@ -1287,18 +1298,18 @@ int __must_check i915_gem_object_set_domain(struct drm_i915_gem_object *obj, uint32_t read_domains, uint32_t write_domain); int __must_check i915_gem_object_finish_gpu(struct drm_i915_gem_object *obj); +int __must_check i915_gem_init(struct drm_device *dev); int __must_check i915_gem_init_hw(struct drm_device *dev); void i915_gem_init_swizzling(struct drm_device *dev); void i915_gem_init_ppgtt(struct drm_device *dev); void i915_gem_cleanup_ringbuffer(struct drm_device *dev); -int __must_check i915_gpu_idle(struct drm_device *dev, bool do_retire); +int __must_check i915_gpu_idle(struct drm_device *dev); int __must_check i915_gem_idle(struct drm_device *dev); int __must_check i915_add_request(struct intel_ring_buffer *ring, struct drm_file *file, struct drm_i915_gem_request *request); int __must_check i915_wait_request(struct intel_ring_buffer *ring, - uint32_t seqno, - bool do_retire); + uint32_t seqno); int i915_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf); int __must_check i915_gem_object_set_to_gtt_domain(struct drm_i915_gem_object *obj, @@ -1349,10 +1360,11 @@ void i915_gem_init_global_gtt(struct drm_device *dev, /* i915_gem_evict.c */ int __must_check i915_gem_evict_something(struct drm_device *dev, int min_size, unsigned alignment, bool mappable); -int __must_check i915_gem_evict_everything(struct drm_device *dev, - bool purgeable_only); -int __must_check i915_gem_evict_inactive(struct drm_device *dev, - bool purgeable_only); +int i915_gem_evict_everything(struct drm_device *dev, bool purgeable_only); + +/* i915_gem_stolen.c */ +int i915_gem_init_stolen(struct drm_device *dev); +void i915_gem_cleanup_stolen(struct drm_device *dev); /* i915_gem_tiling.c */ void i915_gem_detect_bit_6_swizzle(struct drm_device *dev); @@ -1467,28 +1479,6 @@ extern void intel_display_print_error_state(struct seq_file *m, struct intel_display_error_state *error); #endif -#define LP_RING(d) (&((struct drm_i915_private *)(d))->ring[RCS]) - -#define BEGIN_LP_RING(n) \ - intel_ring_begin(LP_RING(dev_priv), (n)) - -#define OUT_RING(x) \ - intel_ring_emit(LP_RING(dev_priv), x) - -#define ADVANCE_LP_RING() \ - intel_ring_advance(LP_RING(dev_priv)) - -/** - * Lock test for when it's just for synchronization of ring access. - * - * In that case, we don't need to do it when GEM is initialized as nobody else - * has access to the ring. - */ -#define RING_LOCK_TEST_WITH_RETURN(dev, file) do { \ - if (LP_RING(dev->dev_private)->obj == NULL) \ - LOCK_TEST_WITH_RETURN(dev, file); \ -} while (0) - /* On SNB platform, before reading ring registers forcewake bit * must be set to prevent GT core from power down and stale values being * returned. diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index dd87937..44a5f24 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -46,7 +46,6 @@ static int i915_gem_phys_pwrite(struct drm_device *dev, struct drm_i915_gem_object *obj, struct drm_i915_gem_pwrite *args, struct drm_file *file); -static void i915_gem_free_object_tail(struct drm_i915_gem_object *obj); static void i915_gem_write_fence(struct drm_device *dev, int reg, struct drm_i915_gem_object *obj); @@ -66,7 +65,7 @@ static inline void i915_gem_object_fence_lost(struct drm_i915_gem_object *obj) /* As we do not have an associated fence register, we will force * a tiling change if we ever need to acquire one. */ - obj->tiling_changed = false; + obj->fence_dirty = false; obj->fence_reg = I915_FENCE_REG_NONE; } @@ -132,7 +131,7 @@ int i915_mutex_lock_interruptible(struct drm_device *dev) static inline bool i915_gem_object_is_inactive(struct drm_i915_gem_object *obj) { - return obj->gtt_space && !obj->active && obj->pin_count == 0; + return !obj->active; } int @@ -141,6 +140,9 @@ i915_gem_init_ioctl(struct drm_device *dev, void *data, { struct drm_i915_gem_init *args = data; + if (drm_core_check_feature(dev, DRIVER_MODESET)) + return -ENODEV; + if (args->gtt_start >= args->gtt_end || (args->gtt_end | args->gtt_start) & (PAGE_SIZE - 1)) return -EINVAL; @@ -166,13 +168,11 @@ i915_gem_get_aperture_ioctl(struct drm_device *dev, void *data, struct drm_i915_gem_object *obj; size_t pinned; - if (!(dev->driver->driver_features & DRIVER_GEM)) - return -ENODEV; - pinned = 0; mutex_lock(&dev->struct_mutex); - list_for_each_entry(obj, &dev_priv->mm.pinned_list, mm_list) - pinned += obj->gtt_space->size; + list_for_each_entry(obj, &dev_priv->mm.gtt_list, gtt_list) + if (obj->pin_count) + pinned += obj->gtt_space->size; mutex_unlock(&dev->struct_mutex); args->aper_size = dev_priv->mm.gtt_total; @@ -243,6 +243,7 @@ i915_gem_create_ioctl(struct drm_device *dev, void *data, struct drm_file *file) { struct drm_i915_gem_create *args = data; + return i915_gem_create(file, dev, args->size, &args->handle); } @@ -282,8 +283,8 @@ __copy_to_user_swizzled(char __user *cpu_vaddr, } static inline int -__copy_from_user_swizzled(char __user *gpu_vaddr, int gpu_offset, - const char *cpu_vaddr, +__copy_from_user_swizzled(char *gpu_vaddr, int gpu_offset, + const char __user *cpu_vaddr, int length) { int ret, cpu_offset = 0; @@ -558,11 +559,14 @@ fast_user_write(struct io_mapping *mapping, char __user *user_data, int length) { - char *vaddr_atomic; + void __iomem *vaddr_atomic; + void *vaddr; unsigned long unwritten; vaddr_atomic = io_mapping_map_atomic_wc(mapping, page_base); - unwritten = __copy_from_user_inatomic_nocache(vaddr_atomic + page_offset, + /* We can use the cpu mem copy function because this is X86. */ + vaddr = (void __force*)vaddr_atomic + page_offset; + unwritten = __copy_from_user_inatomic_nocache(vaddr, user_data, length); io_mapping_unmap_atomic(vaddr_atomic); return unwritten; @@ -925,9 +929,6 @@ i915_gem_set_domain_ioctl(struct drm_device *dev, void *data, uint32_t write_domain = args->write_domain; int ret; - if (!(dev->driver->driver_features & DRIVER_GEM)) - return -ENODEV; - /* Only handle setting domains to types used by the CPU. */ if (write_domain & I915_GEM_GPU_DOMAINS) return -EINVAL; @@ -981,9 +982,6 @@ i915_gem_sw_finish_ioctl(struct drm_device *dev, void *data, struct drm_i915_gem_object *obj; int ret = 0; - if (!(dev->driver->driver_features & DRIVER_GEM)) - return -ENODEV; - ret = i915_mutex_lock_interruptible(dev); if (ret) return ret; @@ -1019,9 +1017,6 @@ i915_gem_mmap_ioctl(struct drm_device *dev, void *data, struct drm_gem_object *obj; unsigned long addr; - if (!(dev->driver->driver_features & DRIVER_GEM)) - return -ENODEV; - obj = drm_gem_object_lookup(dev, file, args->handle); if (obj == NULL) return -ENOENT; @@ -1247,9 +1242,6 @@ i915_gem_mmap_gtt(struct drm_file *file, struct drm_i915_gem_object *obj; int ret; - if (!(dev->driver->driver_features & DRIVER_GEM)) - return -ENODEV; - ret = i915_mutex_lock_interruptible(dev); if (ret) return ret; @@ -1307,9 +1299,6 @@ i915_gem_mmap_gtt_ioctl(struct drm_device *dev, void *data, { struct drm_i915_gem_mmap_gtt *args = data; - if (!(dev->driver->driver_features & DRIVER_GEM)) - return -ENODEV; - return i915_gem_mmap_gtt(file, dev, args->handle, &args->offset); } @@ -1450,10 +1439,7 @@ i915_gem_object_move_to_inactive(struct drm_i915_gem_object *obj) struct drm_device *dev = obj->base.dev; struct drm_i915_private *dev_priv = dev->dev_private; - if (obj->pin_count != 0) - list_move_tail(&obj->mm_list, &dev_priv->mm.pinned_list); - else - list_move_tail(&obj->mm_list, &dev_priv->mm.inactive_list); + list_move_tail(&obj->mm_list, &dev_priv->mm.inactive_list); BUG_ON(!list_empty(&obj->gpu_write_list)); BUG_ON(!obj->active); @@ -1779,20 +1765,6 @@ i915_gem_retire_requests(struct drm_device *dev) drm_i915_private_t *dev_priv = dev->dev_private; int i; - if (!list_empty(&dev_priv->mm.deferred_free_list)) { - struct drm_i915_gem_object *obj, *next; - - /* We must be careful that during unbind() we do not - * accidentally infinitely recurse into retire requests. - * Currently: - * retire -> free -> unbind -> wait -> retire_ring - */ - list_for_each_entry_safe(obj, next, - &dev_priv->mm.deferred_free_list, - mm_list) - i915_gem_free_object_tail(obj); - } - for (i = 0; i < I915_NUM_RINGS; i++) i915_gem_retire_requests_ring(&dev_priv->ring[i]); } @@ -1845,20 +1817,10 @@ i915_gem_retire_work_handler(struct work_struct *work) mutex_unlock(&dev->struct_mutex); } -/** - * Waits for a sequence number to be signaled, and cleans up the - * request and object lists appropriately for that event. - */ -int -i915_wait_request(struct intel_ring_buffer *ring, - uint32_t seqno, - bool do_retire) +static int +i915_gem_check_wedge(struct drm_i915_private *dev_priv) { - drm_i915_private_t *dev_priv = ring->dev->dev_private; - u32 ier; - int ret = 0; - - BUG_ON(seqno == 0); + BUG_ON(!mutex_is_locked(&dev_priv->dev->struct_mutex)); if (atomic_read(&dev_priv->mm.wedged)) { struct completion *x = &dev_priv->error_completion; @@ -1873,6 +1835,20 @@ i915_wait_request(struct intel_ring_buffer *ring, return recovery_complete ? -EIO : -EAGAIN; } + return 0; +} + +/* + * Compare seqno against outstanding lazy request. Emit a request if they are + * equal. + */ +static int +i915_gem_check_olr(struct intel_ring_buffer *ring, u32 seqno) +{ + int ret = 0; + + BUG_ON(!mutex_is_locked(&ring->dev->struct_mutex)); + if (seqno == ring->outstanding_lazy_request) { struct drm_i915_gem_request *request; @@ -1886,56 +1862,67 @@ i915_wait_request(struct intel_ring_buffer *ring, return ret; } - seqno = request->seqno; + BUG_ON(seqno != request->seqno); } - if (!i915_seqno_passed(ring->get_seqno(ring), seqno)) { - if (HAS_PCH_SPLIT(ring->dev)) - ier = I915_READ(DEIER) | I915_READ(GTIER); - else if (IS_VALLEYVIEW(ring->dev)) - ier = I915_READ(GTIER) | I915_READ(VLV_IER); - else - ier = I915_READ(IER); - if (!ier) { - DRM_ERROR("something (likely vbetool) disabled " - "interrupts, re-enabling\n"); - ring->dev->driver->irq_preinstall(ring->dev); - ring->dev->driver->irq_postinstall(ring->dev); - } + return ret; +} + +static int __wait_seqno(struct intel_ring_buffer *ring, u32 seqno, + bool interruptible) +{ + drm_i915_private_t *dev_priv = ring->dev->dev_private; + int ret = 0; - trace_i915_gem_request_wait_begin(ring, seqno); + if (i915_seqno_passed(ring->get_seqno(ring), seqno)) + return 0; - ring->waiting_seqno = seqno; - if (ring->irq_get(ring)) { - if (dev_priv->mm.interruptible) - ret = wait_event_interruptible(ring->irq_queue, - i915_seqno_passed(ring->get_seqno(ring), seqno) - || atomic_read(&dev_priv->mm.wedged)); - else - wait_event(ring->irq_queue, - i915_seqno_passed(ring->get_seqno(ring), seqno) - || atomic_read(&dev_priv->mm.wedged)); + trace_i915_gem_request_wait_begin(ring, seqno); + if (WARN_ON(!ring->irq_get(ring))) + return -ENODEV; - ring->irq_put(ring); - } else if (wait_for_atomic(i915_seqno_passed(ring->get_seqno(ring), - seqno) || - atomic_read(&dev_priv->mm.wedged), 3000)) - ret = -EBUSY; - ring->waiting_seqno = 0; +#define EXIT_COND \ + (i915_seqno_passed(ring->get_seqno(ring), seqno) || \ + atomic_read(&dev_priv->mm.wedged)) - trace_i915_gem_request_wait_end(ring, seqno); - } + if (interruptible) + ret = wait_event_interruptible(ring->irq_queue, + EXIT_COND); + else + wait_event(ring->irq_queue, EXIT_COND); + + ring->irq_put(ring); + trace_i915_gem_request_wait_end(ring, seqno); +#undef EXIT_COND + + return ret; +} + +/** + * Waits for a sequence number to be signaled, and cleans up the + * request and object lists appropriately for that event. + */ +int +i915_wait_request(struct intel_ring_buffer *ring, + uint32_t seqno) +{ + drm_i915_private_t *dev_priv = ring->dev->dev_private; + int ret = 0; + + BUG_ON(seqno == 0); + + ret = i915_gem_check_wedge(dev_priv); + if (ret) + return ret; + + ret = i915_gem_check_olr(ring, seqno); + if (ret) + return ret; + + ret = __wait_seqno(ring, seqno, dev_priv->mm.interruptible); if (atomic_read(&dev_priv->mm.wedged)) ret = -EAGAIN; - /* Directly dispatch request retiring. While we have the work queue - * to handle this, the waiter on a request often wants an associated - * buffer to have made it to the inactive list, and we would need - * a separate wait queue to handle that. - */ - if (ret == 0 && do_retire) - i915_gem_retire_requests_ring(ring); - return ret; } @@ -1957,10 +1944,10 @@ i915_gem_object_wait_rendering(struct drm_i915_gem_object *obj) * it. */ if (obj->active) { - ret = i915_wait_request(obj->ring, obj->last_rendering_seqno, - true); + ret = i915_wait_request(obj->ring, obj->last_rendering_seqno); if (ret) return ret; + i915_gem_retire_requests_ring(obj->ring); } return 0; @@ -1998,22 +1985,9 @@ i915_gem_object_sync(struct drm_i915_gem_object *obj, if (seqno <= from->sync_seqno[idx]) return 0; - if (seqno == from->outstanding_lazy_request) { - struct drm_i915_gem_request *request; - - request = kzalloc(sizeof(*request), GFP_KERNEL); - if (request == NULL) - return -ENOMEM; - - ret = i915_add_request(from, NULL, request); - if (ret) { - kfree(request); - return ret; - } - - seqno = request->seqno; - } - + ret = i915_gem_check_olr(obj->ring, seqno); + if (ret) + return ret; ret = to->sync_to(to, from, seqno); if (!ret) @@ -2064,7 +2038,7 @@ i915_gem_object_unbind(struct drm_i915_gem_object *obj) } ret = i915_gem_object_finish_gpu(obj); - if (ret == -ERESTARTSYS) + if (ret) return ret; /* Continue on if we fail due to EIO, the GPU is hung so we * should be safe and we need to cleanup or else we might @@ -2091,7 +2065,7 @@ i915_gem_object_unbind(struct drm_i915_gem_object *obj) /* release the fence reg _after_ flushing */ ret = i915_gem_object_put_fence(obj); - if (ret == -ERESTARTSYS) + if (ret) return ret; trace_i915_gem_object_unbind(obj); @@ -2143,7 +2117,7 @@ i915_gem_flush_ring(struct intel_ring_buffer *ring, return 0; } -static int i915_ring_idle(struct intel_ring_buffer *ring, bool do_retire) +static int i915_ring_idle(struct intel_ring_buffer *ring) { int ret; @@ -2157,18 +2131,17 @@ static int i915_ring_idle(struct intel_ring_buffer *ring, bool do_retire) return ret; } - return i915_wait_request(ring, i915_gem_next_request_seqno(ring), - do_retire); + return i915_wait_request(ring, i915_gem_next_request_seqno(ring)); } -int i915_gpu_idle(struct drm_device *dev, bool do_retire) +int i915_gpu_idle(struct drm_device *dev) { drm_i915_private_t *dev_priv = dev->dev_private; int ret, i; /* Flush everything onto the inactive list. */ for (i = 0; i < I915_NUM_RINGS; i++) { - ret = i915_ring_idle(&dev_priv->ring[i], do_retire); + ret = i915_ring_idle(&dev_priv->ring[i]); if (ret) return ret; } @@ -2357,9 +2330,7 @@ i915_gem_object_flush_fence(struct drm_i915_gem_object *obj) } if (obj->last_fenced_seqno) { - ret = i915_wait_request(obj->ring, - obj->last_fenced_seqno, - false); + ret = i915_wait_request(obj->ring, obj->last_fenced_seqno); if (ret) return ret; @@ -2454,7 +2425,7 @@ i915_gem_object_get_fence(struct drm_i915_gem_object *obj) /* Have we updated the tiling parameters upon the object and so * will need to serialise the write to the associated fence register? */ - if (obj->tiling_changed) { + if (obj->fence_dirty) { ret = i915_gem_object_flush_fence(obj); if (ret) return ret; @@ -2463,7 +2434,7 @@ i915_gem_object_get_fence(struct drm_i915_gem_object *obj) /* Just update our place in the LRU if our fence is getting reused. */ if (obj->fence_reg != I915_FENCE_REG_NONE) { reg = &dev_priv->fence_regs[obj->fence_reg]; - if (!obj->tiling_changed) { + if (!obj->fence_dirty) { list_move_tail(®->lru_list, &dev_priv->mm.fence_list); return 0; @@ -2486,7 +2457,7 @@ i915_gem_object_get_fence(struct drm_i915_gem_object *obj) return 0; i915_gem_object_update_fence(obj, reg, enable); - obj->tiling_changed = false; + obj->fence_dirty = false; return 0; } @@ -2732,6 +2703,7 @@ i915_gem_object_flush_cpu_write_domain(struct drm_i915_gem_object *obj) int i915_gem_object_set_to_gtt_domain(struct drm_i915_gem_object *obj, bool write) { + drm_i915_private_t *dev_priv = obj->base.dev->dev_private; uint32_t old_write_domain, old_read_domains; int ret; @@ -2772,6 +2744,10 @@ i915_gem_object_set_to_gtt_domain(struct drm_i915_gem_object *obj, bool write) old_read_domains, old_write_domain); + /* And bump the LRU for this access */ + if (i915_gem_object_is_inactive(obj)) + list_move_tail(&obj->mm_list, &dev_priv->mm.inactive_list); + return 0; } @@ -3020,28 +2996,7 @@ i915_gem_ring_throttle(struct drm_device *dev, struct drm_file *file) if (seqno == 0) return 0; - ret = 0; - if (!i915_seqno_passed(ring->get_seqno(ring), seqno)) { - /* And wait for the seqno passing without holding any locks and - * causing extra latency for others. This is safe as the irq - * generation is designed to be run atomically and so is - * lockless. - */ - if (ring->irq_get(ring)) { - ret = wait_event_interruptible(ring->irq_queue, - i915_seqno_passed(ring->get_seqno(ring), seqno) - || atomic_read(&dev_priv->mm.wedged)); - ring->irq_put(ring); - - if (ret == 0 && atomic_read(&dev_priv->mm.wedged)) - ret = -EIO; - } else if (wait_for_atomic(i915_seqno_passed(ring->get_seqno(ring), - seqno) || - atomic_read(&dev_priv->mm.wedged), 3000)) { - ret = -EBUSY; - } - } - + ret = __wait_seqno(ring, seqno, true); if (ret == 0) queue_delayed_work(dev_priv->wq, &dev_priv->mm.retire_work, 0); @@ -3053,12 +3008,9 @@ i915_gem_object_pin(struct drm_i915_gem_object *obj, uint32_t alignment, bool map_and_fenceable) { - struct drm_device *dev = obj->base.dev; - struct drm_i915_private *dev_priv = dev->dev_private; int ret; BUG_ON(obj->pin_count == DRM_I915_GEM_OBJECT_MAX_PIN_COUNT); - WARN_ON(i915_verify_lists(dev)); if (obj->gtt_space != NULL) { if ((alignment && obj->gtt_offset & (alignment - 1)) || @@ -3086,34 +3038,20 @@ i915_gem_object_pin(struct drm_i915_gem_object *obj, if (!obj->has_global_gtt_mapping && map_and_fenceable) i915_gem_gtt_bind_object(obj, obj->cache_level); - if (obj->pin_count++ == 0) { - if (!obj->active) - list_move_tail(&obj->mm_list, - &dev_priv->mm.pinned_list); - } + obj->pin_count++; obj->pin_mappable |= map_and_fenceable; - WARN_ON(i915_verify_lists(dev)); return 0; } void i915_gem_object_unpin(struct drm_i915_gem_object *obj) { - struct drm_device *dev = obj->base.dev; - drm_i915_private_t *dev_priv = dev->dev_private; - - WARN_ON(i915_verify_lists(dev)); BUG_ON(obj->pin_count == 0); BUG_ON(obj->gtt_space == NULL); - if (--obj->pin_count == 0) { - if (!obj->active) - list_move_tail(&obj->mm_list, - &dev_priv->mm.inactive_list); + if (--obj->pin_count == 0) obj->pin_mappable = false; - } - WARN_ON(i915_verify_lists(dev)); } int @@ -3237,20 +3175,9 @@ i915_gem_busy_ioctl(struct drm_device *dev, void *data, if (obj->base.write_domain & I915_GEM_GPU_DOMAINS) { ret = i915_gem_flush_ring(obj->ring, 0, obj->base.write_domain); - } else if (obj->ring->outstanding_lazy_request == - obj->last_rendering_seqno) { - struct drm_i915_gem_request *request; - - /* This ring is not being cleared by active usage, - * so emit a request to do so. - */ - request = kzalloc(sizeof(*request), GFP_KERNEL); - if (request) { - ret = i915_add_request(obj->ring, NULL, request); - if (ret) - kfree(request); - } else - ret = -ENOMEM; + } else { + ret = i915_gem_check_olr(obj->ring, + obj->last_rendering_seqno); } /* Update the active list for the hardware's current position. @@ -3386,21 +3313,29 @@ int i915_gem_init_object(struct drm_gem_object *obj) return 0; } -static void i915_gem_free_object_tail(struct drm_i915_gem_object *obj) +void i915_gem_free_object(struct drm_gem_object *gem_obj) { + struct drm_i915_gem_object *obj = to_intel_bo(gem_obj); struct drm_device *dev = obj->base.dev; drm_i915_private_t *dev_priv = dev->dev_private; - int ret; - - ret = i915_gem_object_unbind(obj); - if (ret == -ERESTARTSYS) { - list_move(&obj->mm_list, - &dev_priv->mm.deferred_free_list); - return; - } trace_i915_gem_object_destroy(obj); + if (obj->phys_obj) + i915_gem_detach_phys_object(dev, obj); + + obj->pin_count = 0; + if (WARN_ON(i915_gem_object_unbind(obj) == -ERESTARTSYS)) { + bool was_interruptible; + + was_interruptible = dev_priv->mm.interruptible; + dev_priv->mm.interruptible = false; + + WARN_ON(i915_gem_object_unbind(obj)); + + dev_priv->mm.interruptible = was_interruptible; + } + if (obj->base.map_list.map) drm_gem_free_mmap_offset(&obj->base); @@ -3411,20 +3346,6 @@ static void i915_gem_free_object_tail(struct drm_i915_gem_object *obj) kfree(obj); } -void i915_gem_free_object(struct drm_gem_object *gem_obj) -{ - struct drm_i915_gem_object *obj = to_intel_bo(gem_obj); - struct drm_device *dev = obj->base.dev; - - while (obj->pin_count > 0) - i915_gem_object_unpin(obj); - - if (obj->phys_obj) - i915_gem_detach_phys_object(dev, obj); - - i915_gem_free_object_tail(obj); -} - int i915_gem_idle(struct drm_device *dev) { @@ -3438,20 +3359,16 @@ i915_gem_idle(struct drm_device *dev) return 0; } - ret = i915_gpu_idle(dev, true); + ret = i915_gpu_idle(dev); if (ret) { mutex_unlock(&dev->struct_mutex); return ret; } + i915_gem_retire_requests(dev); /* Under UMS, be paranoid and evict. */ - if (!drm_core_check_feature(dev, DRIVER_MODESET)) { - ret = i915_gem_evict_inactive(dev, false); - if (ret) { - mutex_unlock(&dev->struct_mutex); - return ret; - } - } + if (!drm_core_check_feature(dev, DRIVER_MODESET)) + i915_gem_evict_everything(dev, false); i915_gem_reset_fences(dev); @@ -3489,9 +3406,9 @@ void i915_gem_init_swizzling(struct drm_device *dev) I915_WRITE(TILECTL, I915_READ(TILECTL) | TILECTL_SWZCTL); if (IS_GEN6(dev)) - I915_WRITE(ARB_MODE, ARB_MODE_ENABLE(ARB_MODE_SWIZZLE_SNB)); + I915_WRITE(ARB_MODE, _MASKED_BIT_ENABLE(ARB_MODE_SWIZZLE_SNB)); else - I915_WRITE(ARB_MODE, ARB_MODE_ENABLE(ARB_MODE_SWIZZLE_IVB)); + I915_WRITE(ARB_MODE, _MASKED_BIT_ENABLE(ARB_MODE_SWIZZLE_IVB)); } void i915_gem_init_ppgtt(struct drm_device *dev) @@ -3540,7 +3457,7 @@ void i915_gem_init_ppgtt(struct drm_device *dev) ecochk = I915_READ(GAM_ECOCHK); I915_WRITE(GAM_ECOCHK, ecochk | ECOCHK_SNB_BIT | ECOCHK_PPGTT_CACHE64B); - I915_WRITE(GFX_MODE, GFX_MODE_ENABLE(GFX_PPGTT_ENABLE)); + I915_WRITE(GFX_MODE, _MASKED_BIT_ENABLE(GFX_PPGTT_ENABLE)); } else if (INTEL_INFO(dev)->gen >= 7) { I915_WRITE(GAM_ECOCHK, ECOCHK_PPGTT_CACHE64B); /* GFX_MODE is per-ring on gen7+ */ @@ -3551,7 +3468,7 @@ void i915_gem_init_ppgtt(struct drm_device *dev) if (INTEL_INFO(dev)->gen >= 7) I915_WRITE(RING_MODE_GEN7(ring), - GFX_MODE_ENABLE(GFX_PPGTT_ENABLE)); + _MASKED_BIT_ENABLE(GFX_PPGTT_ENABLE)); I915_WRITE(RING_PP_DIR_DCLV(ring), PP_DIR_DCLV_2G); I915_WRITE(RING_PP_DIR_BASE(ring), pd_offset); @@ -3595,6 +3512,71 @@ cleanup_render_ring: return ret; } +static bool +intel_enable_ppgtt(struct drm_device *dev) +{ + if (i915_enable_ppgtt >= 0) + return i915_enable_ppgtt; + +#ifdef CONFIG_INTEL_IOMMU + /* Disable ppgtt on SNB if VT-d is on. */ + if (INTEL_INFO(dev)->gen == 6 && intel_iommu_gfx_mapped) + return false; +#endif + + return true; +} + +int i915_gem_init(struct drm_device *dev) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + unsigned long gtt_size, mappable_size; + int ret; + + gtt_size = dev_priv->mm.gtt->gtt_total_entries << PAGE_SHIFT; + mappable_size = dev_priv->mm.gtt->gtt_mappable_entries << PAGE_SHIFT; + + mutex_lock(&dev->struct_mutex); + if (intel_enable_ppgtt(dev) && HAS_ALIASING_PPGTT(dev)) { + /* PPGTT pdes are stolen from global gtt ptes, so shrink the + * aperture accordingly when using aliasing ppgtt. */ + gtt_size -= I915_PPGTT_PD_ENTRIES*PAGE_SIZE; + + i915_gem_init_global_gtt(dev, 0, mappable_size, gtt_size); + + ret = i915_gem_init_aliasing_ppgtt(dev); + if (ret) { + mutex_unlock(&dev->struct_mutex); + return ret; + } + } else { + /* Let GEM Manage all of the aperture. + * + * However, leave one page at the end still bound to the scratch + * page. There are a number of places where the hardware + * apparently prefetches past the end of the object, and we've + * seen multiple hangs with the GPU head pointer stuck in a + * batchbuffer bound at the last page of the aperture. One page + * should be enough to keep any prefetching inside of the + * aperture. + */ + i915_gem_init_global_gtt(dev, 0, mappable_size, + gtt_size); + } + + ret = i915_gem_init_hw(dev); + mutex_unlock(&dev->struct_mutex); + if (ret) { + i915_gem_cleanup_aliasing_ppgtt(dev); + return ret; + } + + /* Allow hardware batchbuffers unless told otherwise, but not for KMS. */ + if (!drm_core_check_feature(dev, DRIVER_MODESET)) + dev_priv->dri1.allow_batchbuffer = 1; + return 0; +} + void i915_gem_cleanup_ringbuffer(struct drm_device *dev) { @@ -3694,9 +3676,7 @@ i915_gem_load(struct drm_device *dev) INIT_LIST_HEAD(&dev_priv->mm.active_list); INIT_LIST_HEAD(&dev_priv->mm.flushing_list); INIT_LIST_HEAD(&dev_priv->mm.inactive_list); - INIT_LIST_HEAD(&dev_priv->mm.pinned_list); INIT_LIST_HEAD(&dev_priv->mm.fence_list); - INIT_LIST_HEAD(&dev_priv->mm.deferred_free_list); INIT_LIST_HEAD(&dev_priv->mm.gtt_list); for (i = 0; i < I915_NUM_RINGS; i++) init_ring_lists(&dev_priv->ring[i]); @@ -3708,12 +3688,8 @@ i915_gem_load(struct drm_device *dev) /* On GEN3 we really need to make sure the ARB C3 LP bit is set */ if (IS_GEN3(dev)) { - u32 tmp = I915_READ(MI_ARB_STATE); - if (!(tmp & MI_ARB_C3_LP_WRITE_ENABLE)) { - /* arb state is a masked write, so set bit + bit in mask */ - tmp = MI_ARB_C3_LP_WRITE_ENABLE | (MI_ARB_C3_LP_WRITE_ENABLE << MI_ARB_MASK_SHIFT); - I915_WRITE(MI_ARB_STATE, tmp); - } + I915_WRITE(MI_ARB_STATE, + _MASKED_BIT_ENABLE(MI_ARB_C3_LP_WRITE_ENABLE)); } dev_priv->relative_constants_mode = I915_EXEC_CONSTANTS_REL_GENERAL; @@ -4016,7 +3992,7 @@ rescan: * This has a dramatic impact to reduce the number of * OOM-killer events whilst running the GPU aggressively. */ - if (i915_gpu_idle(dev, true) == 0) + if (i915_gpu_idle(dev) == 0) goto rescan; } mutex_unlock(&dev->struct_mutex); diff --git a/drivers/gpu/drm/i915/i915_gem_debug.c b/drivers/gpu/drm/i915/i915_gem_debug.c index cc93cac..a4f6aaa 100644 --- a/drivers/gpu/drm/i915/i915_gem_debug.c +++ b/drivers/gpu/drm/i915/i915_gem_debug.c @@ -114,22 +114,6 @@ i915_verify_lists(struct drm_device *dev) } } - list_for_each_entry(obj, &dev_priv->mm.pinned_list, list) { - if (obj->base.dev != dev || - !atomic_read(&obj->base.refcount.refcount)) { - DRM_ERROR("freed pinned %p\n", obj); - err++; - break; - } else if (!obj->pin_count || obj->active || - (obj->base.write_domain & I915_GEM_GPU_DOMAINS)) { - DRM_ERROR("invalid pinned %p (p %d a %d w %x)\n", - obj, - obj->pin_count, obj->active, - obj->base.write_domain); - err++; - } - } - return warned = err; } #endif /* WATCH_INACTIVE */ diff --git a/drivers/gpu/drm/i915/i915_gem_evict.c b/drivers/gpu/drm/i915/i915_gem_evict.c index 21a8271..3bcf045 100644 --- a/drivers/gpu/drm/i915/i915_gem_evict.c +++ b/drivers/gpu/drm/i915/i915_gem_evict.c @@ -35,6 +35,9 @@ static bool mark_free(struct drm_i915_gem_object *obj, struct list_head *unwind) { + if (obj->pin_count) + return false; + list_add(&obj->exec_list, unwind); return drm_mm_scan_add_block(obj->gtt_space); } @@ -90,7 +93,7 @@ i915_gem_evict_something(struct drm_device *dev, int min_size, /* Now merge in the soon-to-be-expired objects... */ list_for_each_entry(obj, &dev_priv->mm.active_list, mm_list) { /* Does the object require an outstanding flush? */ - if (obj->base.write_domain || obj->pin_count) + if (obj->base.write_domain) continue; if (mark_free(obj, &unwind_list)) @@ -99,14 +102,11 @@ i915_gem_evict_something(struct drm_device *dev, int min_size, /* Finally add anything with a pending flush (in order of retirement) */ list_for_each_entry(obj, &dev_priv->mm.flushing_list, mm_list) { - if (obj->pin_count) - continue; - if (mark_free(obj, &unwind_list)) goto found; } list_for_each_entry(obj, &dev_priv->mm.active_list, mm_list) { - if (!obj->base.write_domain || obj->pin_count) + if (!obj->base.write_domain) continue; if (mark_free(obj, &unwind_list)) @@ -166,8 +166,9 @@ int i915_gem_evict_everything(struct drm_device *dev, bool purgeable_only) { drm_i915_private_t *dev_priv = dev->dev_private; - int ret; + struct drm_i915_gem_object *obj, *next; bool lists_empty; + int ret,i; lists_empty = (list_empty(&dev_priv->mm.inactive_list) && list_empty(&dev_priv->mm.flushing_list) && @@ -177,31 +178,30 @@ i915_gem_evict_everything(struct drm_device *dev, bool purgeable_only) trace_i915_gem_evict_everything(dev, purgeable_only); - /* Flush everything (on to the inactive lists) and evict */ - ret = i915_gpu_idle(dev, true); + ret = i915_gpu_idle(dev); if (ret) return ret; - BUG_ON(!list_empty(&dev_priv->mm.flushing_list)); + /* The gpu_idle will flush everything in the write domain to the + * active list. Then we must move everything off the active list + * with retire requests. + */ + for (i = 0; i < I915_NUM_RINGS; i++) + if (WARN_ON(!list_empty(&dev_priv->ring[i].gpu_write_list))) + return -EBUSY; - return i915_gem_evict_inactive(dev, purgeable_only); -} + i915_gem_retire_requests(dev); -/** Unbinds all inactive objects. */ -int -i915_gem_evict_inactive(struct drm_device *dev, bool purgeable_only) -{ - drm_i915_private_t *dev_priv = dev->dev_private; - struct drm_i915_gem_object *obj, *next; + BUG_ON(!list_empty(&dev_priv->mm.flushing_list)); + /* Having flushed everything, unbind() should never raise an error */ list_for_each_entry_safe(obj, next, &dev_priv->mm.inactive_list, mm_list) { if (!purgeable_only || obj->madv != I915_MADV_WILLNEED) { - int ret = i915_gem_object_unbind(obj); - if (ret) - return ret; + if (obj->pin_count == 0) + WARN_ON(i915_gem_object_unbind(obj)); } } - return 0; + return ret; } diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c index c77bfa9a..206b9bb 100644 --- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c +++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c @@ -1116,11 +1116,17 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data, return -EINVAL; } + if (INTEL_INFO(dev)->gen >= 5) { + DRM_DEBUG("clip rectangles are only valid on pre-gen5\n"); + return -EINVAL; + } + if (args->num_cliprects > UINT_MAX / sizeof(*cliprects)) { DRM_DEBUG("execbuf with %u cliprects\n", args->num_cliprects); return -EINVAL; } + cliprects = kmalloc(args->num_cliprects * sizeof(*cliprects), GFP_KERNEL); if (cliprects == NULL) { @@ -1225,9 +1231,10 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data, * so every billion or so execbuffers, we need to stall * the GPU in order to reset the counters. */ - ret = i915_gpu_idle(dev, true); + ret = i915_gpu_idle(dev); if (ret) goto err; + i915_gem_retire_requests(dev); BUG_ON(ring->sync_seqno[i]); } diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c index 25c8bf9..29d573c 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c @@ -317,7 +317,7 @@ static bool do_idling(struct drm_i915_private *dev_priv) if (unlikely(dev_priv->mm.gtt->do_idle_maps)) { dev_priv->mm.interruptible = false; - if (i915_gpu_idle(dev_priv->dev, false)) { + if (i915_gpu_idle(dev_priv->dev)) { DRM_ERROR("Couldn't idle GPU\n"); /* Wait a bit, in hopes it avoids the hang */ udelay(10); diff --git a/drivers/gpu/drm/i915/i915_gem_stolen.c b/drivers/gpu/drm/i915/i915_gem_stolen.c new file mode 100644 index 0000000..ada2e90 --- /dev/null +++ b/drivers/gpu/drm/i915/i915_gem_stolen.c @@ -0,0 +1,202 @@ +/* + * Copyright © 2008-2012 Intel Corporation + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice (including the next + * paragraph) shall be included in all copies or substantial portions of the + * Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + * + * Authors: + * Eric Anholt <eric@anholt.net> + * Chris Wilson <chris@chris-wilson.co.uk> + * + */ + +#include "drmP.h" +#include "drm.h" +#include "i915_drm.h" +#include "i915_drv.h" + +/* + * The BIOS typically reserves some of the system's memory for the exclusive + * use of the integrated graphics. This memory is no longer available for + * use by the OS and so the user finds that his system has less memory + * available than he put in. We refer to this memory as stolen. + * + * The BIOS will allocate its framebuffer from the stolen memory. Our + * goal is try to reuse that object for our own fbcon which must always + * be available for panics. Anything else we can reuse the stolen memory + * for is a boon. + */ + +#define PTE_ADDRESS_MASK 0xfffff000 +#define PTE_ADDRESS_MASK_HIGH 0x000000f0 /* i915+ */ +#define PTE_MAPPING_TYPE_UNCACHED (0 << 1) +#define PTE_MAPPING_TYPE_DCACHE (1 << 1) /* i830 only */ +#define PTE_MAPPING_TYPE_CACHED (3 << 1) +#define PTE_MAPPING_TYPE_MASK (3 << 1) +#define PTE_VALID (1 << 0) + +/** + * i915_stolen_to_phys - take an offset into stolen memory and turn it into + * a physical one + * @dev: drm device + * @offset: address to translate + * + * Some chip functions require allocations from stolen space and need the + * physical address of the memory in question. + */ +static unsigned long i915_stolen_to_phys(struct drm_device *dev, u32 offset) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + struct pci_dev *pdev = dev_priv->bridge_dev; + u32 base; + +#if 0 + /* On the machines I have tested the Graphics Base of Stolen Memory + * is unreliable, so compute the base by subtracting the stolen memory + * from the Top of Low Usable DRAM which is where the BIOS places + * the graphics stolen memory. + */ + if (INTEL_INFO(dev)->gen > 3 || IS_G33(dev)) { + /* top 32bits are reserved = 0 */ + pci_read_config_dword(pdev, 0xA4, &base); + } else { + /* XXX presume 8xx is the same as i915 */ + pci_bus_read_config_dword(pdev->bus, 2, 0x5C, &base); + } +#else + if (INTEL_INFO(dev)->gen > 3 || IS_G33(dev)) { + u16 val; + pci_read_config_word(pdev, 0xb0, &val); + base = val >> 4 << 20; + } else { + u8 val; + pci_read_config_byte(pdev, 0x9c, &val); + base = val >> 3 << 27; + } + base -= dev_priv->mm.gtt->stolen_size; +#endif + + return base + offset; +} + +static void i915_warn_stolen(struct drm_device *dev) +{ + DRM_INFO("not enough stolen space for compressed buffer, disabling\n"); + DRM_INFO("hint: you may be able to increase stolen memory size in the BIOS to avoid this\n"); +} + +static void i915_setup_compression(struct drm_device *dev, int size) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + struct drm_mm_node *compressed_fb, *uninitialized_var(compressed_llb); + unsigned long cfb_base; + unsigned long ll_base = 0; + + /* Just in case the BIOS is doing something questionable. */ + intel_disable_fbc(dev); + + compressed_fb = drm_mm_search_free(&dev_priv->mm.stolen, size, 4096, 0); + if (compressed_fb) + compressed_fb = drm_mm_get_block(compressed_fb, size, 4096); + if (!compressed_fb) + goto err; + + cfb_base = i915_stolen_to_phys(dev, compressed_fb->start); + if (!cfb_base) + goto err_fb; + + if (!(IS_GM45(dev) || HAS_PCH_SPLIT(dev))) { + compressed_llb = drm_mm_search_free(&dev_priv->mm.stolen, + 4096, 4096, 0); + if (compressed_llb) + compressed_llb = drm_mm_get_block(compressed_llb, + 4096, 4096); + if (!compressed_llb) + goto err_fb; + + ll_base = i915_stolen_to_phys(dev, compressed_llb->start); + if (!ll_base) + goto err_llb; + } + + dev_priv->cfb_size = size; + + dev_priv->compressed_fb = compressed_fb; + if (HAS_PCH_SPLIT(dev)) + I915_WRITE(ILK_DPFC_CB_BASE, compressed_fb->start); + else if (IS_GM45(dev)) { + I915_WRITE(DPFC_CB_BASE, compressed_fb->start); + } else { + I915_WRITE(FBC_CFB_BASE, cfb_base); + I915_WRITE(FBC_LL_BASE, ll_base); + dev_priv->compressed_llb = compressed_llb; + } + + DRM_DEBUG_KMS("FBC base 0x%08lx, ll base 0x%08lx, size %dM\n", + cfb_base, ll_base, size >> 20); + return; + +err_llb: + drm_mm_put_block(compressed_llb); +err_fb: + drm_mm_put_block(compressed_fb); +err: + dev_priv->no_fbc_reason = FBC_STOLEN_TOO_SMALL; + i915_warn_stolen(dev); +} + +static void i915_cleanup_compression(struct drm_device *dev) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + + drm_mm_put_block(dev_priv->compressed_fb); + if (dev_priv->compressed_llb) + drm_mm_put_block(dev_priv->compressed_llb); +} + +void i915_gem_cleanup_stolen(struct drm_device *dev) +{ + if (I915_HAS_FBC(dev) && i915_powersave) + i915_cleanup_compression(dev); +} + +int i915_gem_init_stolen(struct drm_device *dev) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + unsigned long prealloc_size = dev_priv->mm.gtt->stolen_size; + + /* Basic memrange allocator for stolen space */ + drm_mm_init(&dev_priv->mm.stolen, 0, prealloc_size); + + /* Try to set up FBC with a reasonable compressed buffer size */ + if (I915_HAS_FBC(dev) && i915_powersave) { + int cfb_size; + + /* Leave 1M for line length buffer & misc. */ + + /* Try to get a 32M buffer... */ + if (prealloc_size > (36*1024*1024)) + cfb_size = 32*1024*1024; + else /* fall back to 7/8 of the stolen space */ + cfb_size = prealloc_size * 7 / 8; + i915_setup_compression(dev, cfb_size); + } + + return 0; +} diff --git a/drivers/gpu/drm/i915/i915_gem_tiling.c b/drivers/gpu/drm/i915/i915_gem_tiling.c index 1a93066..b964df5 100644 --- a/drivers/gpu/drm/i915/i915_gem_tiling.c +++ b/drivers/gpu/drm/i915/i915_gem_tiling.c @@ -354,9 +354,15 @@ i915_gem_set_tiling(struct drm_device *dev, void *data, /* We need to rebind the object if its current allocation * no longer meets the alignment restrictions for its new * tiling mode. Otherwise we can just leave it alone, but - * need to ensure that any fence register is cleared. + * need to ensure that any fence register is updated before + * the next fenced (either through the GTT or by the BLT unit + * on older GPUs) access. + * + * After updating the tiling parameters, we then flag whether + * we need to update an associated fence register. Note this + * has to also include the unfenced register the GPU uses + * whilst executing a fenced command for an untiled object. */ - i915_gem_release_mmap(obj); obj->map_and_fenceable = obj->gtt_space == NULL || @@ -374,9 +380,15 @@ i915_gem_set_tiling(struct drm_device *dev, void *data, } if (ret == 0) { - obj->tiling_changed = true; + obj->fence_dirty = + obj->fenced_gpu_access || + obj->fence_reg != I915_FENCE_REG_NONE; + obj->tiling_mode = args->tiling_mode; obj->stride = args->stride; + + /* Force the fence to be reacquired for GTT access */ + i915_gem_release_mmap(obj); } } /* we have to maintain this existing ABI... */ diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c index ab023ca..b4999b5 100644 --- a/drivers/gpu/drm/i915/i915_irq.c +++ b/drivers/gpu/drm/i915/i915_irq.c @@ -37,35 +37,6 @@ #include "i915_trace.h" #include "intel_drv.h" -#define MAX_NOPID ((u32)~0) - -/** - * Interrupts that are always left unmasked. - * - * Since pipe events are edge-triggered from the PIPESTAT register to IIR, - * we leave them always unmasked in IMR and then control enabling them through - * PIPESTAT alone. - */ -#define I915_INTERRUPT_ENABLE_FIX \ - (I915_ASLE_INTERRUPT | \ - I915_DISPLAY_PIPE_A_EVENT_INTERRUPT | \ - I915_DISPLAY_PIPE_B_EVENT_INTERRUPT | \ - I915_DISPLAY_PLANE_A_FLIP_PENDING_INTERRUPT | \ - I915_DISPLAY_PLANE_B_FLIP_PENDING_INTERRUPT | \ - I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT) - -/** Interrupts that we mask and unmask at runtime. */ -#define I915_INTERRUPT_ENABLE_VAR (I915_USER_INTERRUPT | I915_BSD_USER_INTERRUPT) - -#define I915_PIPE_VBLANK_STATUS (PIPE_START_VBLANK_INTERRUPT_STATUS |\ - PIPE_VBLANK_INTERRUPT_STATUS) - -#define I915_PIPE_VBLANK_ENABLE (PIPE_START_VBLANK_INTERRUPT_ENABLE |\ - PIPE_VBLANK_INTERRUPT_ENABLE) - -#define DRM_I915_VBLANK_PIPE_ALL (DRM_I915_VBLANK_PIPE_A | \ - DRM_I915_VBLANK_PIPE_B) - /* For display hotplug interrupt */ static void ironlake_enable_display_irq(drm_i915_private_t *dev_priv, u32 mask) @@ -360,15 +331,12 @@ static void notify_ring(struct drm_device *dev, struct intel_ring_buffer *ring) { struct drm_i915_private *dev_priv = dev->dev_private; - u32 seqno; if (ring->obj == NULL) return; - seqno = ring->get_seqno(ring); - trace_i915_gem_request_complete(ring, seqno); + trace_i915_gem_request_complete(ring, ring->get_seqno(ring)); - ring->irq_seqno = seqno; wake_up_all(&ring->irq_queue); if (i915_enable_hangcheck) { dev_priv->hangcheck_count = 0; @@ -541,17 +509,13 @@ static irqreturn_t valleyview_irq_handler(DRM_IRQ_ARGS) if (iir & I915_DISPLAY_PIPE_A_VBLANK_INTERRUPT) { drm_handle_vblank(dev, 0); vblank++; - if (!dev_priv->flip_pending_is_done) { - intel_finish_page_flip(dev, 0); - } + intel_finish_page_flip(dev, 0); } if (iir & I915_DISPLAY_PIPE_B_VBLANK_INTERRUPT) { drm_handle_vblank(dev, 1); vblank++; - if (!dev_priv->flip_pending_is_done) { - intel_finish_page_flip(dev, 0); - } + intel_finish_page_flip(dev, 0); } if (pipe_stats[pipe] & PIPE_LEGACY_BLC_EVENT_STATUS) @@ -618,7 +582,6 @@ static irqreturn_t ivybridge_irq_handler(DRM_IRQ_ARGS) drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; int ret = IRQ_NONE; u32 de_iir, gt_iir, de_ier, pch_iir, pm_iir; - struct drm_i915_master_private *master_priv; atomic_inc(&dev_priv->irq_received); @@ -637,13 +600,6 @@ static irqreturn_t ivybridge_irq_handler(DRM_IRQ_ARGS) ret = IRQ_HANDLED; - if (dev->primary->master) { - master_priv = dev->primary->master->driver_priv; - if (master_priv->sarea_priv) - master_priv->sarea_priv->last_dispatch = - READ_BREADCRUMB(dev_priv); - } - snb_gt_irq_handler(dev, dev_priv, gt_iir); if (de_iir & DE_GSE_IVB) @@ -659,12 +615,20 @@ static irqreturn_t ivybridge_irq_handler(DRM_IRQ_ARGS) intel_finish_page_flip_plane(dev, 1); } + if (de_iir & DE_PLANEC_FLIP_DONE_IVB) { + intel_prepare_page_flip(dev, 2); + intel_finish_page_flip_plane(dev, 2); + } + if (de_iir & DE_PIPEA_VBLANK_IVB) drm_handle_vblank(dev, 0); if (de_iir & DE_PIPEB_VBLANK_IVB) drm_handle_vblank(dev, 1); + if (de_iir & DE_PIPEC_VBLANK_IVB) + drm_handle_vblank(dev, 2); + /* check event from PCH */ if (de_iir & DE_PCH_EVENT_IVB) { if (pch_iir & SDE_HOTPLUG_MASK_CPT) @@ -705,7 +669,6 @@ static irqreturn_t ironlake_irq_handler(DRM_IRQ_ARGS) int ret = IRQ_NONE; u32 de_iir, gt_iir, de_ier, pch_iir, pm_iir; u32 hotplug_mask; - struct drm_i915_master_private *master_priv; atomic_inc(&dev_priv->irq_received); @@ -730,13 +693,6 @@ static irqreturn_t ironlake_irq_handler(DRM_IRQ_ARGS) ret = IRQ_HANDLED; - if (dev->primary->master) { - master_priv = dev->primary->master->driver_priv; - if (master_priv->sarea_priv) - master_priv->sarea_priv->last_dispatch = - READ_BREADCRUMB(dev_priv); - } - if (IS_GEN5(dev)) ilk_gt_irq_handler(dev, dev_priv, gt_iir); else @@ -810,7 +766,7 @@ static void i915_error_work_func(struct work_struct *work) if (atomic_read(&dev_priv->mm.wedged)) { DRM_DEBUG_DRIVER("resetting chip\n"); kobject_uevent_env(&dev->primary->kdev.kobj, KOBJ_CHANGE, reset_event); - if (!i915_reset(dev, GRDOM_RENDER)) { + if (!i915_reset(dev)) { atomic_set(&dev_priv->mm.wedged, 0); kobject_uevent_env(&dev->primary->kdev.kobj, KOBJ_CHANGE, reset_done_event); } @@ -902,10 +858,11 @@ i915_error_object_free(struct drm_i915_error_object *obj) kfree(obj); } -static void -i915_error_state_free(struct drm_device *dev, - struct drm_i915_error_state *error) +void +i915_error_state_free(struct kref *error_ref) { + struct drm_i915_error_state *error = container_of(error_ref, + typeof(*error), ref); int i; for (i = 0; i < ARRAY_SIZE(error->ring); i++) { @@ -918,37 +875,56 @@ i915_error_state_free(struct drm_device *dev, kfree(error->overlay); kfree(error); } - -static u32 capture_bo_list(struct drm_i915_error_buffer *err, - int count, - struct list_head *head) +static void capture_bo(struct drm_i915_error_buffer *err, + struct drm_i915_gem_object *obj) +{ + err->size = obj->base.size; + err->name = obj->base.name; + err->seqno = obj->last_rendering_seqno; + err->gtt_offset = obj->gtt_offset; + err->read_domains = obj->base.read_domains; + err->write_domain = obj->base.write_domain; + err->fence_reg = obj->fence_reg; + err->pinned = 0; + if (obj->pin_count > 0) + err->pinned = 1; + if (obj->user_pin_count > 0) + err->pinned = -1; + err->tiling = obj->tiling_mode; + err->dirty = obj->dirty; + err->purgeable = obj->madv != I915_MADV_WILLNEED; + err->ring = obj->ring ? obj->ring->id : -1; + err->cache_level = obj->cache_level; +} + +static u32 capture_active_bo(struct drm_i915_error_buffer *err, + int count, struct list_head *head) { struct drm_i915_gem_object *obj; int i = 0; list_for_each_entry(obj, head, mm_list) { - err->size = obj->base.size; - err->name = obj->base.name; - err->seqno = obj->last_rendering_seqno; - err->gtt_offset = obj->gtt_offset; - err->read_domains = obj->base.read_domains; - err->write_domain = obj->base.write_domain; - err->fence_reg = obj->fence_reg; - err->pinned = 0; - if (obj->pin_count > 0) - err->pinned = 1; - if (obj->user_pin_count > 0) - err->pinned = -1; - err->tiling = obj->tiling_mode; - err->dirty = obj->dirty; - err->purgeable = obj->madv != I915_MADV_WILLNEED; - err->ring = obj->ring ? obj->ring->id : -1; - err->cache_level = obj->cache_level; - + capture_bo(err++, obj); if (++i == count) break; + } + + return i; +} + +static u32 capture_pinned_bo(struct drm_i915_error_buffer *err, + int count, struct list_head *head) +{ + struct drm_i915_gem_object *obj; + int i = 0; + + list_for_each_entry(obj, head, gtt_list) { + if (obj->pin_count == 0) + continue; - err++; + capture_bo(err++, obj); + if (++i == count) + break; } return i; @@ -1045,6 +1021,7 @@ static void i915_record_ring_state(struct drm_device *dev, error->instdone[ring->id] = I915_READ(INSTDONE); } + error->waiting[ring->id] = waitqueue_active(&ring->irq_queue); error->instpm[ring->id] = I915_READ(RING_INSTPM(ring->mmio_base)); error->seqno[ring->id] = ring->get_seqno(ring); error->acthd[ring->id] = intel_ring_get_active_head(ring); @@ -1134,8 +1111,19 @@ static void i915_capture_error_state(struct drm_device *dev) DRM_INFO("capturing error event; look for more information in /debug/dri/%d/i915_error_state\n", dev->primary->index); + kref_init(&error->ref); error->eir = I915_READ(EIR); error->pgtbl_er = I915_READ(PGTBL_ER); + + if (HAS_PCH_SPLIT(dev)) + error->ier = I915_READ(DEIER) | I915_READ(GTIER); + else if (IS_VALLEYVIEW(dev)) + error->ier = I915_READ(GTIER) | I915_READ(VLV_IER); + else if (IS_GEN2(dev)) + error->ier = I915_READ16(IER); + else + error->ier = I915_READ(IER); + for_each_pipe(pipe) error->pipestat[pipe] = I915_READ(PIPESTAT(pipe)); @@ -1155,8 +1143,9 @@ static void i915_capture_error_state(struct drm_device *dev) list_for_each_entry(obj, &dev_priv->mm.active_list, mm_list) i++; error->active_bo_count = i; - list_for_each_entry(obj, &dev_priv->mm.pinned_list, mm_list) - i++; + list_for_each_entry(obj, &dev_priv->mm.gtt_list, gtt_list) + if (obj->pin_count) + i++; error->pinned_bo_count = i - error->active_bo_count; error->active_bo = NULL; @@ -1171,15 +1160,15 @@ static void i915_capture_error_state(struct drm_device *dev) if (error->active_bo) error->active_bo_count = - capture_bo_list(error->active_bo, - error->active_bo_count, - &dev_priv->mm.active_list); + capture_active_bo(error->active_bo, + error->active_bo_count, + &dev_priv->mm.active_list); if (error->pinned_bo) error->pinned_bo_count = - capture_bo_list(error->pinned_bo, - error->pinned_bo_count, - &dev_priv->mm.pinned_list); + capture_pinned_bo(error->pinned_bo, + error->pinned_bo_count, + &dev_priv->mm.gtt_list); do_gettimeofday(&error->time); @@ -1194,7 +1183,7 @@ static void i915_capture_error_state(struct drm_device *dev) spin_unlock_irqrestore(&dev_priv->error_lock, flags); if (error) - i915_error_state_free(dev, error); + i915_error_state_free(&error->ref); } void i915_destroy_error_state(struct drm_device *dev) @@ -1209,7 +1198,7 @@ void i915_destroy_error_state(struct drm_device *dev) spin_unlock_irqrestore(&dev_priv->error_lock, flags); if (error) - i915_error_state_free(dev, error); + kref_put(&error->ref, i915_error_state_free); } #else #define i915_capture_error_state(x) @@ -1385,248 +1374,6 @@ static void i915_pageflip_stall_check(struct drm_device *dev, int pipe) } } -static irqreturn_t i915_driver_irq_handler(DRM_IRQ_ARGS) -{ - struct drm_device *dev = (struct drm_device *) arg; - drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; - struct drm_i915_master_private *master_priv; - u32 iir, new_iir; - u32 pipe_stats[I915_MAX_PIPES]; - u32 vblank_status; - int vblank = 0; - unsigned long irqflags; - int irq_received; - int ret = IRQ_NONE, pipe; - bool blc_event = false; - - atomic_inc(&dev_priv->irq_received); - - iir = I915_READ(IIR); - - if (INTEL_INFO(dev)->gen >= 4) - vblank_status = PIPE_START_VBLANK_INTERRUPT_STATUS; - else - vblank_status = PIPE_VBLANK_INTERRUPT_STATUS; - - for (;;) { - irq_received = iir != 0; - - /* Can't rely on pipestat interrupt bit in iir as it might - * have been cleared after the pipestat interrupt was received. - * It doesn't set the bit in iir again, but it still produces - * interrupts (for non-MSI). - */ - spin_lock_irqsave(&dev_priv->irq_lock, irqflags); - if (iir & I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT) - i915_handle_error(dev, false); - - for_each_pipe(pipe) { - int reg = PIPESTAT(pipe); - pipe_stats[pipe] = I915_READ(reg); - - /* - * Clear the PIPE*STAT regs before the IIR - */ - if (pipe_stats[pipe] & 0x8000ffff) { - if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS) - DRM_DEBUG_DRIVER("pipe %c underrun\n", - pipe_name(pipe)); - I915_WRITE(reg, pipe_stats[pipe]); - irq_received = 1; - } - } - spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags); - - if (!irq_received) - break; - - ret = IRQ_HANDLED; - - /* Consume port. Then clear IIR or we'll miss events */ - if ((I915_HAS_HOTPLUG(dev)) && - (iir & I915_DISPLAY_PORT_INTERRUPT)) { - u32 hotplug_status = I915_READ(PORT_HOTPLUG_STAT); - - DRM_DEBUG_DRIVER("hotplug event received, stat 0x%08x\n", - hotplug_status); - if (hotplug_status & dev_priv->hotplug_supported_mask) - queue_work(dev_priv->wq, - &dev_priv->hotplug_work); - - I915_WRITE(PORT_HOTPLUG_STAT, hotplug_status); - I915_READ(PORT_HOTPLUG_STAT); - } - - I915_WRITE(IIR, iir); - new_iir = I915_READ(IIR); /* Flush posted writes */ - - if (dev->primary->master) { - master_priv = dev->primary->master->driver_priv; - if (master_priv->sarea_priv) - master_priv->sarea_priv->last_dispatch = - READ_BREADCRUMB(dev_priv); - } - - if (iir & I915_USER_INTERRUPT) - notify_ring(dev, &dev_priv->ring[RCS]); - if (iir & I915_BSD_USER_INTERRUPT) - notify_ring(dev, &dev_priv->ring[VCS]); - - if (iir & I915_DISPLAY_PLANE_A_FLIP_PENDING_INTERRUPT) { - intel_prepare_page_flip(dev, 0); - if (dev_priv->flip_pending_is_done) - intel_finish_page_flip_plane(dev, 0); - } - - if (iir & I915_DISPLAY_PLANE_B_FLIP_PENDING_INTERRUPT) { - intel_prepare_page_flip(dev, 1); - if (dev_priv->flip_pending_is_done) - intel_finish_page_flip_plane(dev, 1); - } - - for_each_pipe(pipe) { - if (pipe_stats[pipe] & vblank_status && - drm_handle_vblank(dev, pipe)) { - vblank++; - if (!dev_priv->flip_pending_is_done) { - i915_pageflip_stall_check(dev, pipe); - intel_finish_page_flip(dev, pipe); - } - } - - if (pipe_stats[pipe] & PIPE_LEGACY_BLC_EVENT_STATUS) - blc_event = true; - } - - - if (blc_event || (iir & I915_ASLE_INTERRUPT)) - intel_opregion_asle_intr(dev); - - /* With MSI, interrupts are only generated when iir - * transitions from zero to nonzero. If another bit got - * set while we were handling the existing iir bits, then - * we would never get another interrupt. - * - * This is fine on non-MSI as well, as if we hit this path - * we avoid exiting the interrupt handler only to generate - * another one. - * - * Note that for MSI this could cause a stray interrupt report - * if an interrupt landed in the time between writing IIR and - * the posting read. This should be rare enough to never - * trigger the 99% of 100,000 interrupts test for disabling - * stray interrupts. - */ - iir = new_iir; - } - - return ret; -} - -static int i915_emit_irq(struct drm_device * dev) -{ - drm_i915_private_t *dev_priv = dev->dev_private; - struct drm_i915_master_private *master_priv = dev->primary->master->driver_priv; - - i915_kernel_lost_context(dev); - - DRM_DEBUG_DRIVER("\n"); - - dev_priv->counter++; - if (dev_priv->counter > 0x7FFFFFFFUL) - dev_priv->counter = 1; - if (master_priv->sarea_priv) - master_priv->sarea_priv->last_enqueue = dev_priv->counter; - - if (BEGIN_LP_RING(4) == 0) { - OUT_RING(MI_STORE_DWORD_INDEX); - OUT_RING(I915_BREADCRUMB_INDEX << MI_STORE_DWORD_INDEX_SHIFT); - OUT_RING(dev_priv->counter); - OUT_RING(MI_USER_INTERRUPT); - ADVANCE_LP_RING(); - } - - return dev_priv->counter; -} - -static int i915_wait_irq(struct drm_device * dev, int irq_nr) -{ - drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; - struct drm_i915_master_private *master_priv = dev->primary->master->driver_priv; - int ret = 0; - struct intel_ring_buffer *ring = LP_RING(dev_priv); - - DRM_DEBUG_DRIVER("irq_nr=%d breadcrumb=%d\n", irq_nr, - READ_BREADCRUMB(dev_priv)); - - if (READ_BREADCRUMB(dev_priv) >= irq_nr) { - if (master_priv->sarea_priv) - master_priv->sarea_priv->last_dispatch = READ_BREADCRUMB(dev_priv); - return 0; - } - - if (master_priv->sarea_priv) - master_priv->sarea_priv->perf_boxes |= I915_BOX_WAIT; - - if (ring->irq_get(ring)) { - DRM_WAIT_ON(ret, ring->irq_queue, 3 * DRM_HZ, - READ_BREADCRUMB(dev_priv) >= irq_nr); - ring->irq_put(ring); - } else if (wait_for(READ_BREADCRUMB(dev_priv) >= irq_nr, 3000)) - ret = -EBUSY; - - if (ret == -EBUSY) { - DRM_ERROR("EBUSY -- rec: %d emitted: %d\n", - READ_BREADCRUMB(dev_priv), (int)dev_priv->counter); - } - - return ret; -} - -/* Needs the lock as it touches the ring. - */ -int i915_irq_emit(struct drm_device *dev, void *data, - struct drm_file *file_priv) -{ - drm_i915_private_t *dev_priv = dev->dev_private; - drm_i915_irq_emit_t *emit = data; - int result; - - if (!dev_priv || !LP_RING(dev_priv)->virtual_start) { - DRM_ERROR("called with no initialization\n"); - return -EINVAL; - } - - RING_LOCK_TEST_WITH_RETURN(dev, file_priv); - - mutex_lock(&dev->struct_mutex); - result = i915_emit_irq(dev); - mutex_unlock(&dev->struct_mutex); - - if (DRM_COPY_TO_USER(emit->irq_seq, &result, sizeof(int))) { - DRM_ERROR("copy_to_user\n"); - return -EFAULT; - } - - return 0; -} - -/* Doesn't need the hardware lock. - */ -int i915_irq_wait(struct drm_device *dev, void *data, - struct drm_file *file_priv) -{ - drm_i915_private_t *dev_priv = dev->dev_private; - drm_i915_irq_wait_t *irqwait = data; - - if (!dev_priv) { - DRM_ERROR("called with no initialization\n"); - return -EINVAL; - } - - return i915_wait_irq(dev, irqwait->irq_seq); -} - /* Called from drm generic code, passed 'crtc' which * we use as a pipe index */ @@ -1648,7 +1395,7 @@ static int i915_enable_vblank(struct drm_device *dev, int pipe) /* maintain vblank delivery even in deep C-states */ if (dev_priv->info->gen == 3) - I915_WRITE(INSTPM, INSTPM_AGPBUSY_DIS << 16); + I915_WRITE(INSTPM, _MASKED_BIT_DISABLE(INSTPM_AGPBUSY_DIS)); spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags); return 0; @@ -1679,8 +1426,8 @@ static int ivybridge_enable_vblank(struct drm_device *dev, int pipe) return -EINVAL; spin_lock_irqsave(&dev_priv->irq_lock, irqflags); - ironlake_enable_display_irq(dev_priv, (pipe == 0) ? - DE_PIPEA_VBLANK_IVB : DE_PIPEB_VBLANK_IVB); + ironlake_enable_display_irq(dev_priv, + DE_PIPEA_VBLANK_IVB << (5 * pipe)); spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags); return 0; @@ -1722,8 +1469,7 @@ static void i915_disable_vblank(struct drm_device *dev, int pipe) spin_lock_irqsave(&dev_priv->irq_lock, irqflags); if (dev_priv->info->gen == 3) - I915_WRITE(INSTPM, - INSTPM_AGPBUSY_DIS << 16 | INSTPM_AGPBUSY_DIS); + I915_WRITE(INSTPM, _MASKED_BIT_ENABLE(INSTPM_AGPBUSY_DIS)); i915_disable_pipestat(dev_priv, pipe, PIPE_VBLANK_INTERRUPT_ENABLE | @@ -1748,8 +1494,8 @@ static void ivybridge_disable_vblank(struct drm_device *dev, int pipe) unsigned long irqflags; spin_lock_irqsave(&dev_priv->irq_lock, irqflags); - ironlake_disable_display_irq(dev_priv, (pipe == 0) ? - DE_PIPEA_VBLANK_IVB : DE_PIPEB_VBLANK_IVB); + ironlake_disable_display_irq(dev_priv, + DE_PIPEA_VBLANK_IVB << (pipe * 5)); spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags); } @@ -1774,61 +1520,6 @@ static void valleyview_disable_vblank(struct drm_device *dev, int pipe) spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags); } - -/* Set the vblank monitor pipe - */ -int i915_vblank_pipe_set(struct drm_device *dev, void *data, - struct drm_file *file_priv) -{ - drm_i915_private_t *dev_priv = dev->dev_private; - - if (!dev_priv) { - DRM_ERROR("called with no initialization\n"); - return -EINVAL; - } - - return 0; -} - -int i915_vblank_pipe_get(struct drm_device *dev, void *data, - struct drm_file *file_priv) -{ - drm_i915_private_t *dev_priv = dev->dev_private; - drm_i915_vblank_pipe_t *pipe = data; - - if (!dev_priv) { - DRM_ERROR("called with no initialization\n"); - return -EINVAL; - } - - pipe->pipe = DRM_I915_VBLANK_PIPE_A | DRM_I915_VBLANK_PIPE_B; - - return 0; -} - -/** - * Schedule buffer swap at given vertical blank. - */ -int i915_vblank_swap(struct drm_device *dev, void *data, - struct drm_file *file_priv) -{ - /* The delayed swap mechanism was fundamentally racy, and has been - * removed. The model was that the client requested a delayed flip/swap - * from the kernel, then waited for vblank before continuing to perform - * rendering. The problem was that the kernel might wake the client - * up before it dispatched the vblank swap (since the lock has to be - * held while touching the ringbuffer), in which case the client would - * clear and start the next frame before the swap occurred, and - * flicker would occur in addition to likely missing the vblank. - * - * In the absence of this ioctl, userland falls back to a correct path - * of waiting for a vblank, then dispatching the swap on its own. - * Context switching to userland and back is plenty fast enough for - * meeting the requirements of vblank swapping. - */ - return -EINVAL; -} - static u32 ring_last_seqno(struct intel_ring_buffer *ring) { @@ -1838,14 +1529,17 @@ ring_last_seqno(struct intel_ring_buffer *ring) static bool i915_hangcheck_ring_idle(struct intel_ring_buffer *ring, bool *err) { + /* We don't check whether the ring even exists before calling this + * function. Hence check whether it's initialized. */ + if (ring->obj == NULL) + return true; + if (list_empty(&ring->request_list) || i915_seqno_passed(ring->get_seqno(ring), ring_last_seqno(ring))) { /* Issue a wake-up to catch stuck h/w. */ - if (ring->waiting_seqno && waitqueue_active(&ring->irq_queue)) { - DRM_ERROR("Hangcheck timer elapsed... %s idle [waiting on %d, at %d], missed IRQ?\n", - ring->name, - ring->waiting_seqno, - ring->get_seqno(ring)); + if (waitqueue_active(&ring->irq_queue)) { + DRM_ERROR("Hangcheck timer elapsed... %s idle\n", + ring->name); wake_up_all(&ring->irq_queue); *err = true; } @@ -1973,10 +1667,6 @@ static void ironlake_irq_preinstall(struct drm_device *dev) atomic_set(&dev_priv->irq_received, 0); - INIT_WORK(&dev_priv->hotplug_work, i915_hotplug_work_func); - INIT_WORK(&dev_priv->error_work, i915_error_work_func); - if (IS_GEN6(dev) || IS_IVYBRIDGE(dev)) - INIT_WORK(&dev_priv->rps_work, gen6_pm_rps_work); I915_WRITE(HWSTAM, 0xeffe); @@ -2004,9 +1694,6 @@ static void valleyview_irq_preinstall(struct drm_device *dev) atomic_set(&dev_priv->irq_received, 0); - INIT_WORK(&dev_priv->hotplug_work, i915_hotplug_work_func); - INIT_WORK(&dev_priv->error_work, i915_error_work_func); - /* VLV magic */ I915_WRITE(VLV_IMR, 0); I915_WRITE(RING_IMR(RENDER_RING_BASE), 0); @@ -2061,13 +1748,6 @@ static int ironlake_irq_postinstall(struct drm_device *dev) u32 render_irqs; u32 hotplug_mask; - DRM_INIT_WAITQUEUE(&dev_priv->ring[RCS].irq_queue); - if (HAS_BSD(dev)) - DRM_INIT_WAITQUEUE(&dev_priv->ring[VCS].irq_queue); - if (HAS_BLT(dev)) - DRM_INIT_WAITQUEUE(&dev_priv->ring[BCS].irq_queue); - - dev_priv->vblank_pipe = DRM_I915_VBLANK_PIPE_A | DRM_I915_VBLANK_PIPE_B; dev_priv->irq_mask = ~display_mask; /* should always can generate irq */ @@ -2130,26 +1810,24 @@ static int ivybridge_irq_postinstall(struct drm_device *dev) { drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; /* enable kind of interrupts always enabled */ - u32 display_mask = DE_MASTER_IRQ_CONTROL | DE_GSE_IVB | - DE_PCH_EVENT_IVB | DE_PLANEA_FLIP_DONE_IVB | - DE_PLANEB_FLIP_DONE_IVB; + u32 display_mask = + DE_MASTER_IRQ_CONTROL | DE_GSE_IVB | DE_PCH_EVENT_IVB | + DE_PLANEC_FLIP_DONE_IVB | + DE_PLANEB_FLIP_DONE_IVB | + DE_PLANEA_FLIP_DONE_IVB; u32 render_irqs; u32 hotplug_mask; - DRM_INIT_WAITQUEUE(&dev_priv->ring[RCS].irq_queue); - if (HAS_BSD(dev)) - DRM_INIT_WAITQUEUE(&dev_priv->ring[VCS].irq_queue); - if (HAS_BLT(dev)) - DRM_INIT_WAITQUEUE(&dev_priv->ring[BCS].irq_queue); - - dev_priv->vblank_pipe = DRM_I915_VBLANK_PIPE_A | DRM_I915_VBLANK_PIPE_B; dev_priv->irq_mask = ~display_mask; /* should always can generate irq */ I915_WRITE(DEIIR, I915_READ(DEIIR)); I915_WRITE(DEIMR, dev_priv->irq_mask); - I915_WRITE(DEIER, display_mask | DE_PIPEA_VBLANK_IVB | - DE_PIPEB_VBLANK_IVB); + I915_WRITE(DEIER, + display_mask | + DE_PIPEC_VBLANK_IVB | + DE_PIPEB_VBLANK_IVB | + DE_PIPEA_VBLANK_IVB); POSTING_READ(DEIER); dev_priv->gt_irq_mask = ~0; @@ -2192,16 +1870,9 @@ static int valleyview_irq_postinstall(struct drm_device *dev) dev_priv->irq_mask = ~enable_mask; - - DRM_INIT_WAITQUEUE(&dev_priv->ring[RCS].irq_queue); - DRM_INIT_WAITQUEUE(&dev_priv->ring[VCS].irq_queue); - DRM_INIT_WAITQUEUE(&dev_priv->ring[BCS].irq_queue); - dev_priv->pipestat[0] = 0; dev_priv->pipestat[1] = 0; - dev_priv->vblank_pipe = DRM_I915_VBLANK_PIPE_A | DRM_I915_VBLANK_PIPE_B; - /* Hack for broken MSIs on VLV */ pci_write_config_dword(dev_priv->dev->pdev, 0x94, 0xfee00000); pci_read_config_word(dev->pdev, 0x98, &msid); @@ -2268,15 +1939,413 @@ static int valleyview_irq_postinstall(struct drm_device *dev) return 0; } -static void i915_driver_irq_preinstall(struct drm_device * dev) +static void valleyview_irq_uninstall(struct drm_device *dev) +{ + drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; + int pipe; + + if (!dev_priv) + return; + + for_each_pipe(pipe) + I915_WRITE(PIPESTAT(pipe), 0xffff); + + I915_WRITE(HWSTAM, 0xffffffff); + I915_WRITE(PORT_HOTPLUG_EN, 0); + I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT)); + for_each_pipe(pipe) + I915_WRITE(PIPESTAT(pipe), 0xffff); + I915_WRITE(VLV_IIR, 0xffffffff); + I915_WRITE(VLV_IMR, 0xffffffff); + I915_WRITE(VLV_IER, 0x0); + POSTING_READ(VLV_IER); +} + +static void ironlake_irq_uninstall(struct drm_device *dev) +{ + drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; + + if (!dev_priv) + return; + + I915_WRITE(HWSTAM, 0xffffffff); + + I915_WRITE(DEIMR, 0xffffffff); + I915_WRITE(DEIER, 0x0); + I915_WRITE(DEIIR, I915_READ(DEIIR)); + + I915_WRITE(GTIMR, 0xffffffff); + I915_WRITE(GTIER, 0x0); + I915_WRITE(GTIIR, I915_READ(GTIIR)); + + I915_WRITE(SDEIMR, 0xffffffff); + I915_WRITE(SDEIER, 0x0); + I915_WRITE(SDEIIR, I915_READ(SDEIIR)); +} + +static void i8xx_irq_preinstall(struct drm_device * dev) { drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; int pipe; atomic_set(&dev_priv->irq_received, 0); - INIT_WORK(&dev_priv->hotplug_work, i915_hotplug_work_func); - INIT_WORK(&dev_priv->error_work, i915_error_work_func); + for_each_pipe(pipe) + I915_WRITE(PIPESTAT(pipe), 0); + I915_WRITE16(IMR, 0xffff); + I915_WRITE16(IER, 0x0); + POSTING_READ16(IER); +} + +static int i8xx_irq_postinstall(struct drm_device *dev) +{ + drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; + + dev_priv->pipestat[0] = 0; + dev_priv->pipestat[1] = 0; + + I915_WRITE16(EMR, + ~(I915_ERROR_PAGE_TABLE | I915_ERROR_MEMORY_REFRESH)); + + /* Unmask the interrupts that we always want on. */ + dev_priv->irq_mask = + ~(I915_DISPLAY_PIPE_A_EVENT_INTERRUPT | + I915_DISPLAY_PIPE_B_EVENT_INTERRUPT | + I915_DISPLAY_PLANE_A_FLIP_PENDING_INTERRUPT | + I915_DISPLAY_PLANE_B_FLIP_PENDING_INTERRUPT | + I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT); + I915_WRITE16(IMR, dev_priv->irq_mask); + + I915_WRITE16(IER, + I915_DISPLAY_PIPE_A_EVENT_INTERRUPT | + I915_DISPLAY_PIPE_B_EVENT_INTERRUPT | + I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT | + I915_USER_INTERRUPT); + POSTING_READ16(IER); + + return 0; +} + +static irqreturn_t i8xx_irq_handler(DRM_IRQ_ARGS) +{ + struct drm_device *dev = (struct drm_device *) arg; + drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; + u16 iir, new_iir; + u32 pipe_stats[2]; + unsigned long irqflags; + int irq_received; + int pipe; + u16 flip_mask = + I915_DISPLAY_PLANE_A_FLIP_PENDING_INTERRUPT | + I915_DISPLAY_PLANE_B_FLIP_PENDING_INTERRUPT; + + atomic_inc(&dev_priv->irq_received); + + iir = I915_READ16(IIR); + if (iir == 0) + return IRQ_NONE; + + while (iir & ~flip_mask) { + /* Can't rely on pipestat interrupt bit in iir as it might + * have been cleared after the pipestat interrupt was received. + * It doesn't set the bit in iir again, but it still produces + * interrupts (for non-MSI). + */ + spin_lock_irqsave(&dev_priv->irq_lock, irqflags); + if (iir & I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT) + i915_handle_error(dev, false); + + for_each_pipe(pipe) { + int reg = PIPESTAT(pipe); + pipe_stats[pipe] = I915_READ(reg); + + /* + * Clear the PIPE*STAT regs before the IIR + */ + if (pipe_stats[pipe] & 0x8000ffff) { + if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS) + DRM_DEBUG_DRIVER("pipe %c underrun\n", + pipe_name(pipe)); + I915_WRITE(reg, pipe_stats[pipe]); + irq_received = 1; + } + } + spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags); + + I915_WRITE16(IIR, iir & ~flip_mask); + new_iir = I915_READ16(IIR); /* Flush posted writes */ + + i915_update_dri1_breadcrumb(dev); + + if (iir & I915_USER_INTERRUPT) + notify_ring(dev, &dev_priv->ring[RCS]); + + if (pipe_stats[0] & PIPE_VBLANK_INTERRUPT_STATUS && + drm_handle_vblank(dev, 0)) { + if (iir & I915_DISPLAY_PLANE_A_FLIP_PENDING_INTERRUPT) { + intel_prepare_page_flip(dev, 0); + intel_finish_page_flip(dev, 0); + flip_mask &= ~I915_DISPLAY_PLANE_A_FLIP_PENDING_INTERRUPT; + } + } + + if (pipe_stats[1] & PIPE_VBLANK_INTERRUPT_STATUS && + drm_handle_vblank(dev, 1)) { + if (iir & I915_DISPLAY_PLANE_B_FLIP_PENDING_INTERRUPT) { + intel_prepare_page_flip(dev, 1); + intel_finish_page_flip(dev, 1); + flip_mask &= ~I915_DISPLAY_PLANE_B_FLIP_PENDING_INTERRUPT; + } + } + + iir = new_iir; + } + + return IRQ_HANDLED; +} + +static void i8xx_irq_uninstall(struct drm_device * dev) +{ + drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; + int pipe; + + for_each_pipe(pipe) { + /* Clear enable bits; then clear status bits */ + I915_WRITE(PIPESTAT(pipe), 0); + I915_WRITE(PIPESTAT(pipe), I915_READ(PIPESTAT(pipe))); + } + I915_WRITE16(IMR, 0xffff); + I915_WRITE16(IER, 0x0); + I915_WRITE16(IIR, I915_READ16(IIR)); +} + +static void i915_irq_preinstall(struct drm_device * dev) +{ + drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; + int pipe; + + atomic_set(&dev_priv->irq_received, 0); + + if (I915_HAS_HOTPLUG(dev)) { + I915_WRITE(PORT_HOTPLUG_EN, 0); + I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT)); + } + + I915_WRITE16(HWSTAM, 0xeffe); + for_each_pipe(pipe) + I915_WRITE(PIPESTAT(pipe), 0); + I915_WRITE(IMR, 0xffffffff); + I915_WRITE(IER, 0x0); + POSTING_READ(IER); +} + +static int i915_irq_postinstall(struct drm_device *dev) +{ + drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; + u32 enable_mask; + + dev_priv->pipestat[0] = 0; + dev_priv->pipestat[1] = 0; + + I915_WRITE(EMR, ~(I915_ERROR_PAGE_TABLE | I915_ERROR_MEMORY_REFRESH)); + + /* Unmask the interrupts that we always want on. */ + dev_priv->irq_mask = + ~(I915_ASLE_INTERRUPT | + I915_DISPLAY_PIPE_A_EVENT_INTERRUPT | + I915_DISPLAY_PIPE_B_EVENT_INTERRUPT | + I915_DISPLAY_PLANE_A_FLIP_PENDING_INTERRUPT | + I915_DISPLAY_PLANE_B_FLIP_PENDING_INTERRUPT | + I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT); + + enable_mask = + I915_ASLE_INTERRUPT | + I915_DISPLAY_PIPE_A_EVENT_INTERRUPT | + I915_DISPLAY_PIPE_B_EVENT_INTERRUPT | + I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT | + I915_USER_INTERRUPT; + + if (I915_HAS_HOTPLUG(dev)) { + /* Enable in IER... */ + enable_mask |= I915_DISPLAY_PORT_INTERRUPT; + /* and unmask in IMR */ + dev_priv->irq_mask &= ~I915_DISPLAY_PORT_INTERRUPT; + } + + I915_WRITE(IMR, dev_priv->irq_mask); + I915_WRITE(IER, enable_mask); + POSTING_READ(IER); + + if (I915_HAS_HOTPLUG(dev)) { + u32 hotplug_en = I915_READ(PORT_HOTPLUG_EN); + + if (dev_priv->hotplug_supported_mask & HDMIB_HOTPLUG_INT_STATUS) + hotplug_en |= HDMIB_HOTPLUG_INT_EN; + if (dev_priv->hotplug_supported_mask & HDMIC_HOTPLUG_INT_STATUS) + hotplug_en |= HDMIC_HOTPLUG_INT_EN; + if (dev_priv->hotplug_supported_mask & HDMID_HOTPLUG_INT_STATUS) + hotplug_en |= HDMID_HOTPLUG_INT_EN; + if (dev_priv->hotplug_supported_mask & SDVOC_HOTPLUG_INT_STATUS) + hotplug_en |= SDVOC_HOTPLUG_INT_EN; + if (dev_priv->hotplug_supported_mask & SDVOB_HOTPLUG_INT_STATUS) + hotplug_en |= SDVOB_HOTPLUG_INT_EN; + if (dev_priv->hotplug_supported_mask & CRT_HOTPLUG_INT_STATUS) { + hotplug_en |= CRT_HOTPLUG_INT_EN; + hotplug_en |= CRT_HOTPLUG_VOLTAGE_COMPARE_50; + } + + /* Ignore TV since it's buggy */ + + I915_WRITE(PORT_HOTPLUG_EN, hotplug_en); + } + + intel_opregion_enable_asle(dev); + + return 0; +} + +static irqreturn_t i915_irq_handler(DRM_IRQ_ARGS) +{ + struct drm_device *dev = (struct drm_device *) arg; + drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; + u32 iir, new_iir, pipe_stats[I915_MAX_PIPES]; + unsigned long irqflags; + u32 flip_mask = + I915_DISPLAY_PLANE_A_FLIP_PENDING_INTERRUPT | + I915_DISPLAY_PLANE_B_FLIP_PENDING_INTERRUPT; + u32 flip[2] = { + I915_DISPLAY_PLANE_A_FLIP_PENDING_INTERRUPT, + I915_DISPLAY_PLANE_B_FLIP_PENDING_INTERRUPT + }; + int pipe, ret = IRQ_NONE; + + atomic_inc(&dev_priv->irq_received); + + iir = I915_READ(IIR); + do { + bool irq_received = (iir & ~flip_mask) != 0; + bool blc_event = false; + + /* Can't rely on pipestat interrupt bit in iir as it might + * have been cleared after the pipestat interrupt was received. + * It doesn't set the bit in iir again, but it still produces + * interrupts (for non-MSI). + */ + spin_lock_irqsave(&dev_priv->irq_lock, irqflags); + if (iir & I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT) + i915_handle_error(dev, false); + + for_each_pipe(pipe) { + int reg = PIPESTAT(pipe); + pipe_stats[pipe] = I915_READ(reg); + + /* Clear the PIPE*STAT regs before the IIR */ + if (pipe_stats[pipe] & 0x8000ffff) { + if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS) + DRM_DEBUG_DRIVER("pipe %c underrun\n", + pipe_name(pipe)); + I915_WRITE(reg, pipe_stats[pipe]); + irq_received = true; + } + } + spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags); + + if (!irq_received) + break; + + /* Consume port. Then clear IIR or we'll miss events */ + if ((I915_HAS_HOTPLUG(dev)) && + (iir & I915_DISPLAY_PORT_INTERRUPT)) { + u32 hotplug_status = I915_READ(PORT_HOTPLUG_STAT); + + DRM_DEBUG_DRIVER("hotplug event received, stat 0x%08x\n", + hotplug_status); + if (hotplug_status & dev_priv->hotplug_supported_mask) + queue_work(dev_priv->wq, + &dev_priv->hotplug_work); + + I915_WRITE(PORT_HOTPLUG_STAT, hotplug_status); + POSTING_READ(PORT_HOTPLUG_STAT); + } + + I915_WRITE(IIR, iir & ~flip_mask); + new_iir = I915_READ(IIR); /* Flush posted writes */ + + if (iir & I915_USER_INTERRUPT) + notify_ring(dev, &dev_priv->ring[RCS]); + + for_each_pipe(pipe) { + int plane = pipe; + if (IS_MOBILE(dev)) + plane = !plane; + if (pipe_stats[pipe] & PIPE_VBLANK_INTERRUPT_STATUS && + drm_handle_vblank(dev, pipe)) { + if (iir & flip[plane]) { + intel_prepare_page_flip(dev, plane); + intel_finish_page_flip(dev, pipe); + flip_mask &= ~flip[plane]; + } + } + + if (pipe_stats[pipe] & PIPE_LEGACY_BLC_EVENT_STATUS) + blc_event = true; + } + + if (blc_event || (iir & I915_ASLE_INTERRUPT)) + intel_opregion_asle_intr(dev); + + /* With MSI, interrupts are only generated when iir + * transitions from zero to nonzero. If another bit got + * set while we were handling the existing iir bits, then + * we would never get another interrupt. + * + * This is fine on non-MSI as well, as if we hit this path + * we avoid exiting the interrupt handler only to generate + * another one. + * + * Note that for MSI this could cause a stray interrupt report + * if an interrupt landed in the time between writing IIR and + * the posting read. This should be rare enough to never + * trigger the 99% of 100,000 interrupts test for disabling + * stray interrupts. + */ + ret = IRQ_HANDLED; + iir = new_iir; + } while (iir & ~flip_mask); + + i915_update_dri1_breadcrumb(dev); + + return ret; +} + +static void i915_irq_uninstall(struct drm_device * dev) +{ + drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; + int pipe; + + if (I915_HAS_HOTPLUG(dev)) { + I915_WRITE(PORT_HOTPLUG_EN, 0); + I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT)); + } + + I915_WRITE16(HWSTAM, 0xffff); + for_each_pipe(pipe) { + /* Clear enable bits; then clear status bits */ + I915_WRITE(PIPESTAT(pipe), 0); + I915_WRITE(PIPESTAT(pipe), I915_READ(PIPESTAT(pipe))); + } + I915_WRITE(IMR, 0xffffffff); + I915_WRITE(IER, 0x0); + + I915_WRITE(IIR, I915_READ(IIR)); +} + +static void i965_irq_preinstall(struct drm_device * dev) +{ + drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; + int pipe; + + atomic_set(&dev_priv->irq_received, 0); if (I915_HAS_HOTPLUG(dev)) { I915_WRITE(PORT_HOTPLUG_EN, 0); @@ -2291,20 +2360,25 @@ static void i915_driver_irq_preinstall(struct drm_device * dev) POSTING_READ(IER); } -/* - * Must be called after intel_modeset_init or hotplug interrupts won't be - * enabled correctly. - */ -static int i915_driver_irq_postinstall(struct drm_device *dev) +static int i965_irq_postinstall(struct drm_device *dev) { drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; - u32 enable_mask = I915_INTERRUPT_ENABLE_FIX | I915_INTERRUPT_ENABLE_VAR; + u32 enable_mask; u32 error_mask; - dev_priv->vblank_pipe = DRM_I915_VBLANK_PIPE_A | DRM_I915_VBLANK_PIPE_B; - /* Unmask the interrupts that we always want on. */ - dev_priv->irq_mask = ~I915_INTERRUPT_ENABLE_FIX; + dev_priv->irq_mask = ~(I915_ASLE_INTERRUPT | + I915_DISPLAY_PIPE_A_EVENT_INTERRUPT | + I915_DISPLAY_PIPE_B_EVENT_INTERRUPT | + I915_DISPLAY_PLANE_A_FLIP_PENDING_INTERRUPT | + I915_DISPLAY_PLANE_B_FLIP_PENDING_INTERRUPT | + I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT); + + enable_mask = ~dev_priv->irq_mask; + enable_mask |= I915_USER_INTERRUPT; + + if (IS_G4X(dev)) + enable_mask |= I915_BSD_USER_INTERRUPT; dev_priv->pipestat[0] = 0; dev_priv->pipestat[1] = 0; @@ -2371,55 +2445,124 @@ static int i915_driver_irq_postinstall(struct drm_device *dev) return 0; } -static void valleyview_irq_uninstall(struct drm_device *dev) +static irqreturn_t i965_irq_handler(DRM_IRQ_ARGS) { + struct drm_device *dev = (struct drm_device *) arg; drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; - int pipe; + u32 iir, new_iir; + u32 pipe_stats[I915_MAX_PIPES]; + unsigned long irqflags; + int irq_received; + int ret = IRQ_NONE, pipe; - if (!dev_priv) - return; + atomic_inc(&dev_priv->irq_received); - dev_priv->vblank_pipe = 0; + iir = I915_READ(IIR); - for_each_pipe(pipe) - I915_WRITE(PIPESTAT(pipe), 0xffff); + for (;;) { + bool blc_event = false; - I915_WRITE(HWSTAM, 0xffffffff); - I915_WRITE(PORT_HOTPLUG_EN, 0); - I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT)); - for_each_pipe(pipe) - I915_WRITE(PIPESTAT(pipe), 0xffff); - I915_WRITE(VLV_IIR, 0xffffffff); - I915_WRITE(VLV_IMR, 0xffffffff); - I915_WRITE(VLV_IER, 0x0); - POSTING_READ(VLV_IER); -} + irq_received = iir != 0; -static void ironlake_irq_uninstall(struct drm_device *dev) -{ - drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; + /* Can't rely on pipestat interrupt bit in iir as it might + * have been cleared after the pipestat interrupt was received. + * It doesn't set the bit in iir again, but it still produces + * interrupts (for non-MSI). + */ + spin_lock_irqsave(&dev_priv->irq_lock, irqflags); + if (iir & I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT) + i915_handle_error(dev, false); - if (!dev_priv) - return; + for_each_pipe(pipe) { + int reg = PIPESTAT(pipe); + pipe_stats[pipe] = I915_READ(reg); + + /* + * Clear the PIPE*STAT regs before the IIR + */ + if (pipe_stats[pipe] & 0x8000ffff) { + if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS) + DRM_DEBUG_DRIVER("pipe %c underrun\n", + pipe_name(pipe)); + I915_WRITE(reg, pipe_stats[pipe]); + irq_received = 1; + } + } + spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags); - dev_priv->vblank_pipe = 0; + if (!irq_received) + break; - I915_WRITE(HWSTAM, 0xffffffff); + ret = IRQ_HANDLED; - I915_WRITE(DEIMR, 0xffffffff); - I915_WRITE(DEIER, 0x0); - I915_WRITE(DEIIR, I915_READ(DEIIR)); + /* Consume port. Then clear IIR or we'll miss events */ + if ((I915_HAS_HOTPLUG(dev)) && + (iir & I915_DISPLAY_PORT_INTERRUPT)) { + u32 hotplug_status = I915_READ(PORT_HOTPLUG_STAT); - I915_WRITE(GTIMR, 0xffffffff); - I915_WRITE(GTIER, 0x0); - I915_WRITE(GTIIR, I915_READ(GTIIR)); + DRM_DEBUG_DRIVER("hotplug event received, stat 0x%08x\n", + hotplug_status); + if (hotplug_status & dev_priv->hotplug_supported_mask) + queue_work(dev_priv->wq, + &dev_priv->hotplug_work); - I915_WRITE(SDEIMR, 0xffffffff); - I915_WRITE(SDEIER, 0x0); - I915_WRITE(SDEIIR, I915_READ(SDEIIR)); + I915_WRITE(PORT_HOTPLUG_STAT, hotplug_status); + I915_READ(PORT_HOTPLUG_STAT); + } + + I915_WRITE(IIR, iir); + new_iir = I915_READ(IIR); /* Flush posted writes */ + + if (iir & I915_USER_INTERRUPT) + notify_ring(dev, &dev_priv->ring[RCS]); + if (iir & I915_BSD_USER_INTERRUPT) + notify_ring(dev, &dev_priv->ring[VCS]); + + if (iir & I915_DISPLAY_PLANE_A_FLIP_PENDING_INTERRUPT) + intel_prepare_page_flip(dev, 0); + + if (iir & I915_DISPLAY_PLANE_B_FLIP_PENDING_INTERRUPT) + intel_prepare_page_flip(dev, 1); + + for_each_pipe(pipe) { + if (pipe_stats[pipe] & PIPE_START_VBLANK_INTERRUPT_STATUS && + drm_handle_vblank(dev, pipe)) { + i915_pageflip_stall_check(dev, pipe); + intel_finish_page_flip(dev, pipe); + } + + if (pipe_stats[pipe] & PIPE_LEGACY_BLC_EVENT_STATUS) + blc_event = true; + } + + + if (blc_event || (iir & I915_ASLE_INTERRUPT)) + intel_opregion_asle_intr(dev); + + /* With MSI, interrupts are only generated when iir + * transitions from zero to nonzero. If another bit got + * set while we were handling the existing iir bits, then + * we would never get another interrupt. + * + * This is fine on non-MSI as well, as if we hit this path + * we avoid exiting the interrupt handler only to generate + * another one. + * + * Note that for MSI this could cause a stray interrupt report + * if an interrupt landed in the time between writing IIR and + * the posting read. This should be rare enough to never + * trigger the 99% of 100,000 interrupts test for disabling + * stray interrupts. + */ + iir = new_iir; + } + + i915_update_dri1_breadcrumb(dev); + + return ret; } -static void i915_driver_irq_uninstall(struct drm_device * dev) +static void i965_irq_uninstall(struct drm_device * dev) { drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; int pipe; @@ -2427,8 +2570,6 @@ static void i915_driver_irq_uninstall(struct drm_device * dev) if (!dev_priv) return; - dev_priv->vblank_pipe = 0; - if (I915_HAS_HOTPLUG(dev)) { I915_WRITE(PORT_HOTPLUG_EN, 0); I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT)); @@ -2448,6 +2589,12 @@ static void i915_driver_irq_uninstall(struct drm_device * dev) void intel_irq_init(struct drm_device *dev) { + struct drm_i915_private *dev_priv = dev->dev_private; + + INIT_WORK(&dev_priv->hotplug_work, i915_hotplug_work_func); + INIT_WORK(&dev_priv->error_work, i915_error_work_func); + INIT_WORK(&dev_priv->rps_work, gen6_pm_rps_work); + dev->driver->get_vblank_counter = i915_get_vblank_counter; dev->max_vblank_count = 0xffffff; /* only 24 bits of frame count */ if (IS_G4X(dev) || IS_GEN5(dev) || IS_GEN6(dev) || IS_IVYBRIDGE(dev) || @@ -2485,10 +2632,25 @@ void intel_irq_init(struct drm_device *dev) dev->driver->enable_vblank = ironlake_enable_vblank; dev->driver->disable_vblank = ironlake_disable_vblank; } else { - dev->driver->irq_preinstall = i915_driver_irq_preinstall; - dev->driver->irq_postinstall = i915_driver_irq_postinstall; - dev->driver->irq_uninstall = i915_driver_irq_uninstall; - dev->driver->irq_handler = i915_driver_irq_handler; + if (INTEL_INFO(dev)->gen == 2) { + dev->driver->irq_preinstall = i8xx_irq_preinstall; + dev->driver->irq_postinstall = i8xx_irq_postinstall; + dev->driver->irq_handler = i8xx_irq_handler; + dev->driver->irq_uninstall = i8xx_irq_uninstall; + } else if (INTEL_INFO(dev)->gen == 3) { + /* IIR "flip pending" means done if this bit is set */ + I915_WRITE(ECOSKPD, _MASKED_BIT_DISABLE(ECO_FLIP_DONE)); + + dev->driver->irq_preinstall = i915_irq_preinstall; + dev->driver->irq_postinstall = i915_irq_postinstall; + dev->driver->irq_uninstall = i915_irq_uninstall; + dev->driver->irq_handler = i915_irq_handler; + } else { + dev->driver->irq_preinstall = i965_irq_preinstall; + dev->driver->irq_postinstall = i965_irq_postinstall; + dev->driver->irq_uninstall = i965_irq_uninstall; + dev->driver->irq_handler = i965_irq_handler; + } dev->driver->enable_vblank = i915_enable_vblank; dev->driver->disable_vblank = i915_disable_vblank; } diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index 5ac9837..10e71a9 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -29,6 +29,9 @@ #define _PORT(port, a, b) ((a) + (port)*((b)-(a))) +#define _MASKED_BIT_ENABLE(a) (((a) << 16) | (a)) +#define _MASKED_BIT_DISABLE(a) ((a) << 16) + /* * The Bridge device's PCI config space has information about the * fb aperture size and the amount of pre-reserved memory. @@ -79,6 +82,7 @@ #define GRDOM_FULL (0<<2) #define GRDOM_RENDER (1<<2) #define GRDOM_MEDIA (3<<2) +#define GRDOM_RESET_ENABLE (1<<0) #define GEN6_MBCUNIT_SNPCR 0x900c /* for LLC config */ #define GEN6_MBC_SNPCR_SHIFT 21 @@ -425,8 +429,6 @@ #define ARB_MODE 0x04030 #define ARB_MODE_SWIZZLE_SNB (1<<4) #define ARB_MODE_SWIZZLE_IVB (1<<5) -#define ARB_MODE_ENABLE(x) GFX_MODE_ENABLE(x) -#define ARB_MODE_DISABLE(x) GFX_MODE_DISABLE(x) #define RENDER_HWS_PGA_GEN7 (0x04080) #define RING_FAULT_REG(ring) (0x4094 + 0x100*(ring)->id) #define DONE_REG 0x40b0 @@ -514,9 +516,6 @@ #define GFX_PSMI_GRANULARITY (1<<10) #define GFX_PPGTT_ENABLE (1<<9) -#define GFX_MODE_ENABLE(bit) (((bit) << 16) | (bit)) -#define GFX_MODE_DISABLE(bit) (((bit) << 16) | (0)) - #define SCPD0 0x0209c /* 915+ only */ #define IER 0x020a0 #define IIR 0x020a4 @@ -572,7 +571,6 @@ #define LM_BURST_LENGTH 0x00000700 #define LM_FIFO_WATERMARK 0x0000001F #define MI_ARB_STATE 0x020e4 /* 915+ only */ -#define MI_ARB_MASK_SHIFT 16 /* shift for enable bits */ /* Make render/texture TLB fetches lower priorty than associated data * fetches. This is not turned on by default @@ -637,7 +635,6 @@ #define MI_ARB_DISPLAY_PRIORITY_B_A (1 << 0) /* display B > display A */ #define CACHE_MODE_0 0x02120 /* 915+ only */ -#define CM0_MASK_SHIFT 16 #define CM0_IZ_OPT_DISABLE (1<<6) #define CM0_ZR_OPT_DISABLE (1<<5) #define CM0_STC_EVICT_DISABLE_LRA_SNB (1<<5) @@ -3224,11 +3221,14 @@ #define DE_PCH_EVENT_IVB (1<<28) #define DE_DP_A_HOTPLUG_IVB (1<<27) #define DE_AUX_CHANNEL_A_IVB (1<<26) +#define DE_SPRITEC_FLIP_DONE_IVB (1<<14) +#define DE_PLANEC_FLIP_DONE_IVB (1<<13) +#define DE_PIPEC_VBLANK_IVB (1<<10) #define DE_SPRITEB_FLIP_DONE_IVB (1<<9) -#define DE_SPRITEA_FLIP_DONE_IVB (1<<4) #define DE_PLANEB_FLIP_DONE_IVB (1<<8) -#define DE_PLANEA_FLIP_DONE_IVB (1<<3) #define DE_PIPEB_VBLANK_IVB (1<<5) +#define DE_SPRITEA_FLIP_DONE_IVB (1<<4) +#define DE_PLANEA_FLIP_DONE_IVB (1<<3) #define DE_PIPEA_VBLANK_IVB (1<<0) #define VLV_MASTER_IER 0x4400c /* Gunit master IER */ @@ -3402,15 +3402,15 @@ #define _PCH_DPLL_A 0xc6014 #define _PCH_DPLL_B 0xc6018 -#define PCH_DPLL(pipe) (pipe == 0 ? _PCH_DPLL_A : _PCH_DPLL_B) +#define _PCH_DPLL(pll) (pll == 0 ? _PCH_DPLL_A : _PCH_DPLL_B) #define _PCH_FPA0 0xc6040 #define FP_CB_TUNE (0x3<<22) #define _PCH_FPA1 0xc6044 #define _PCH_FPB0 0xc6048 #define _PCH_FPB1 0xc604c -#define PCH_FP0(pipe) (pipe == 0 ? _PCH_FPA0 : _PCH_FPB0) -#define PCH_FP1(pipe) (pipe == 0 ? _PCH_FPA1 : _PCH_FPB1) +#define _PCH_FP0(pll) (pll == 0 ? _PCH_FPA0 : _PCH_FPB0) +#define _PCH_FP1(pll) (pll == 0 ? _PCH_FPA1 : _PCH_FPB1) #define PCH_DPLL_TEST 0xc606c diff --git a/drivers/gpu/drm/i915/i915_suspend.c b/drivers/gpu/drm/i915/i915_suspend.c index 0c3e3bf..73a5c3c 100644 --- a/drivers/gpu/drm/i915/i915_suspend.c +++ b/drivers/gpu/drm/i915/i915_suspend.c @@ -40,7 +40,7 @@ static bool i915_pipe_enabled(struct drm_device *dev, enum pipe pipe) return false; if (HAS_PCH_SPLIT(dev)) - dpll_reg = PCH_DPLL(pipe); + dpll_reg = _PCH_DPLL(pipe); else dpll_reg = (pipe == PIPE_A) ? _DPLL_A : _DPLL_B; diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 8c239f2..6b41390 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -384,28 +384,6 @@ out_unlock: return val; } -static void intel_dpio_write(struct drm_i915_private *dev_priv, int reg, - u32 val) -{ - unsigned long flags; - - spin_lock_irqsave(&dev_priv->dpio_lock, flags); - if (wait_for_atomic_us((I915_READ(DPIO_PKT) & DPIO_BUSY) == 0, 100)) { - DRM_ERROR("DPIO idle wait timed out\n"); - goto out_unlock; - } - - I915_WRITE(DPIO_DATA, val); - I915_WRITE(DPIO_REG, reg); - I915_WRITE(DPIO_PKT, DPIO_RID | DPIO_OP_WRITE | DPIO_PORTID | - DPIO_BYTE); - if (wait_for_atomic_us((I915_READ(DPIO_PKT) & DPIO_BUSY) == 0, 100)) - DRM_ERROR("DPIO write wait timed out\n"); - -out_unlock: - spin_unlock_irqrestore(&dev_priv->dpio_lock, flags); -} - static void vlv_init_dpio(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; @@ -911,26 +889,28 @@ static void assert_pll(struct drm_i915_private *dev_priv, /* For ILK+ */ static void assert_pch_pll(struct drm_i915_private *dev_priv, - enum pipe pipe, bool state) + struct intel_crtc *intel_crtc, bool state) { int reg; u32 val; bool cur_state; + if (!intel_crtc->pch_pll) { + WARN(1, "asserting PCH PLL enabled with no PLL\n"); + return; + } + if (HAS_PCH_CPT(dev_priv->dev)) { u32 pch_dpll; pch_dpll = I915_READ(PCH_DPLL_SEL); /* Make sure the selected PLL is enabled to the transcoder */ - WARN(!((pch_dpll >> (4 * pipe)) & 8), - "transcoder %d PLL not enabled\n", pipe); - - /* Convert the transcoder pipe number to a pll pipe number */ - pipe = (pch_dpll >> (4 * pipe)) & 1; + WARN(!((pch_dpll >> (4 * intel_crtc->pipe)) & 8), + "transcoder %d PLL not enabled\n", intel_crtc->pipe); } - reg = PCH_DPLL(pipe); + reg = intel_crtc->pch_pll->pll_reg; val = I915_READ(reg); cur_state = !!(val & DPLL_VCO_ENABLE); WARN(cur_state != state, @@ -1306,60 +1286,79 @@ static void intel_disable_pll(struct drm_i915_private *dev_priv, enum pipe pipe) * The PCH PLL needs to be enabled before the PCH transcoder, since it * drives the transcoder clock. */ -static void intel_enable_pch_pll(struct drm_i915_private *dev_priv, - enum pipe pipe) +static void intel_enable_pch_pll(struct intel_crtc *intel_crtc) { + struct drm_i915_private *dev_priv = intel_crtc->base.dev->dev_private; + struct intel_pch_pll *pll = intel_crtc->pch_pll; int reg; u32 val; - if (pipe > 1) - return; - /* PCH only available on ILK+ */ BUG_ON(dev_priv->info->gen < 5); + BUG_ON(pll == NULL); + BUG_ON(pll->refcount == 0); + + DRM_DEBUG_KMS("enable PCH PLL %x (active %d, on? %d)for crtc %d\n", + pll->pll_reg, pll->active, pll->on, + intel_crtc->base.base.id); /* PCH refclock must be enabled first */ assert_pch_refclk_enabled(dev_priv); - reg = PCH_DPLL(pipe); + if (pll->active++ && pll->on) { + assert_pch_pll_enabled(dev_priv, intel_crtc); + return; + } + + DRM_DEBUG_KMS("enabling PCH PLL %x\n", pll->pll_reg); + + reg = pll->pll_reg; val = I915_READ(reg); val |= DPLL_VCO_ENABLE; I915_WRITE(reg, val); POSTING_READ(reg); udelay(200); + + pll->on = true; } -static void intel_disable_pch_pll(struct drm_i915_private *dev_priv, - enum pipe pipe) +static void intel_disable_pch_pll(struct intel_crtc *intel_crtc) { + struct drm_i915_private *dev_priv = intel_crtc->base.dev->dev_private; + struct intel_pch_pll *pll = intel_crtc->pch_pll; int reg; - u32 val, pll_mask = TRANSC_DPLL_ENABLE | TRANSC_DPLLB_SEL, - pll_sel = TRANSC_DPLL_ENABLE; - - if (pipe > 1) - return; + u32 val; /* PCH only available on ILK+ */ BUG_ON(dev_priv->info->gen < 5); + if (pll == NULL) + return; - /* Make sure transcoder isn't still depending on us */ - assert_transcoder_disabled(dev_priv, pipe); + BUG_ON(pll->refcount == 0); - if (pipe == 0) - pll_sel |= TRANSC_DPLLA_SEL; - else if (pipe == 1) - pll_sel |= TRANSC_DPLLB_SEL; + DRM_DEBUG_KMS("disable PCH PLL %x (active %d, on? %d) for crtc %d\n", + pll->pll_reg, pll->active, pll->on, + intel_crtc->base.base.id); - - if ((I915_READ(PCH_DPLL_SEL) & pll_mask) == pll_sel) + BUG_ON(pll->active == 0); + if (--pll->active) { + assert_pch_pll_enabled(dev_priv, intel_crtc); return; + } + + DRM_DEBUG_KMS("disabling PCH PLL %x\n", pll->pll_reg); + + /* Make sure transcoder isn't still depending on us */ + assert_transcoder_disabled(dev_priv, intel_crtc->pipe); - reg = PCH_DPLL(pipe); + reg = pll->pll_reg; val = I915_READ(reg); val &= ~DPLL_VCO_ENABLE; I915_WRITE(reg, val); POSTING_READ(reg); udelay(200); + + pll->on = false; } static void intel_enable_transcoder(struct drm_i915_private *dev_priv, @@ -1373,7 +1372,7 @@ static void intel_enable_transcoder(struct drm_i915_private *dev_priv, BUG_ON(dev_priv->info->gen < 5); /* Make sure PCH DPLL is enabled */ - assert_pch_pll_enabled(dev_priv, pipe); + assert_pch_pll_enabled(dev_priv, to_intel_crtc(crtc)); /* FDI must be feeding us bits for PCH ports */ assert_fdi_tx_enabled(dev_priv, pipe); @@ -2507,26 +2506,6 @@ static void ironlake_fdi_disable(struct drm_crtc *crtc) udelay(100); } -/* - * When we disable a pipe, we need to clear any pending scanline wait events - * to avoid hanging the ring, which we assume we are waiting on. - */ -static void intel_clear_scanline_wait(struct drm_device *dev) -{ - struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_ring_buffer *ring; - u32 tmp; - - if (IS_GEN2(dev)) - /* Can't break the hang on i8xx */ - return; - - ring = LP_RING(dev_priv); - tmp = I915_READ_CTL(ring); - if (tmp & RING_WAIT) - I915_WRITE_CTL(ring, tmp); -} - static void intel_crtc_wait_for_pending_flips(struct drm_crtc *crtc) { struct drm_device *dev = crtc->dev; @@ -2578,29 +2557,36 @@ static void ironlake_pch_enable(struct drm_crtc *crtc) struct drm_i915_private *dev_priv = dev->dev_private; struct intel_crtc *intel_crtc = to_intel_crtc(crtc); int pipe = intel_crtc->pipe; - u32 reg, temp, transc_sel; + u32 reg, temp; /* For PCH output, training FDI link */ dev_priv->display.fdi_link_train(crtc); - intel_enable_pch_pll(dev_priv, pipe); + intel_enable_pch_pll(intel_crtc); if (HAS_PCH_CPT(dev)) { - transc_sel = intel_crtc->use_pll_a ? TRANSC_DPLLA_SEL : - TRANSC_DPLLB_SEL; + u32 sel; - /* Be sure PCH DPLL SEL is set */ temp = I915_READ(PCH_DPLL_SEL); - if (pipe == 0) { - temp &= ~(TRANSA_DPLLB_SEL); - temp |= (TRANSA_DPLL_ENABLE | TRANSA_DPLLA_SEL); - } else if (pipe == 1) { - temp &= ~(TRANSB_DPLLB_SEL); - temp |= (TRANSB_DPLL_ENABLE | TRANSB_DPLLB_SEL); - } else if (pipe == 2) { - temp &= ~(TRANSC_DPLLB_SEL); - temp |= (TRANSC_DPLL_ENABLE | transc_sel); + switch (pipe) { + default: + case 0: + temp |= TRANSA_DPLL_ENABLE; + sel = TRANSA_DPLLB_SEL; + break; + case 1: + temp |= TRANSB_DPLL_ENABLE; + sel = TRANSB_DPLLB_SEL; + break; + case 2: + temp |= TRANSC_DPLL_ENABLE; + sel = TRANSC_DPLLB_SEL; + break; } + if (intel_crtc->pch_pll->pll_reg == _PCH_DPLL_B) + temp |= sel; + else + temp &= ~sel; I915_WRITE(PCH_DPLL_SEL, temp); } @@ -2658,6 +2644,82 @@ static void ironlake_pch_enable(struct drm_crtc *crtc) intel_enable_transcoder(dev_priv, pipe); } +static void intel_put_pch_pll(struct intel_crtc *intel_crtc) +{ + struct intel_pch_pll *pll = intel_crtc->pch_pll; + + if (pll == NULL) + return; + + if (pll->refcount == 0) { + WARN(1, "bad PCH PLL refcount\n"); + return; + } + + --pll->refcount; + intel_crtc->pch_pll = NULL; +} + +static struct intel_pch_pll *intel_get_pch_pll(struct intel_crtc *intel_crtc, u32 dpll, u32 fp) +{ + struct drm_i915_private *dev_priv = intel_crtc->base.dev->dev_private; + struct intel_pch_pll *pll; + int i; + + pll = intel_crtc->pch_pll; + if (pll) { + DRM_DEBUG_KMS("CRTC:%d reusing existing PCH PLL %x\n", + intel_crtc->base.base.id, pll->pll_reg); + goto prepare; + } + + for (i = 0; i < dev_priv->num_pch_pll; i++) { + pll = &dev_priv->pch_plls[i]; + + /* Only want to check enabled timings first */ + if (pll->refcount == 0) + continue; + + if (dpll == (I915_READ(pll->pll_reg) & 0x7fffffff) && + fp == I915_READ(pll->fp0_reg)) { + DRM_DEBUG_KMS("CRTC:%d sharing existing PCH PLL %x (refcount %d, ative %d)\n", + intel_crtc->base.base.id, + pll->pll_reg, pll->refcount, pll->active); + + goto found; + } + } + + /* Ok no matching timings, maybe there's a free one? */ + for (i = 0; i < dev_priv->num_pch_pll; i++) { + pll = &dev_priv->pch_plls[i]; + if (pll->refcount == 0) { + DRM_DEBUG_KMS("CRTC:%d allocated PCH PLL %x\n", + intel_crtc->base.base.id, pll->pll_reg); + goto found; + } + } + + return NULL; + +found: + intel_crtc->pch_pll = pll; + pll->refcount++; + DRM_DEBUG_DRIVER("using pll %d for pipe %d\n", i, intel_crtc->pipe); +prepare: /* separate function? */ + DRM_DEBUG_DRIVER("switching PLL %x off\n", pll->pll_reg); + + /* Wait for the clocks to stabilize before rewriting the regs */ + I915_WRITE(pll->pll_reg, dpll & ~DPLL_VCO_ENABLE); + POSTING_READ(pll->pll_reg); + udelay(150); + + I915_WRITE(pll->fp0_reg, fp); + I915_WRITE(pll->pll_reg, dpll & ~DPLL_VCO_ENABLE); + pll->on = false; + return pll; +} + void intel_cpt_verify_modeset(struct drm_device *dev, int pipe) { struct drm_i915_private *dev_priv = dev->dev_private; @@ -2802,8 +2864,7 @@ static void ironlake_crtc_disable(struct drm_crtc *crtc) } /* disable PCH DPLL */ - if (!intel_crtc->no_pll) - intel_disable_pch_pll(dev_priv, pipe); + intel_disable_pch_pll(intel_crtc); /* Switch from PCDclk to Rawclk */ reg = FDI_RX_CTL(pipe); @@ -2831,7 +2892,6 @@ static void ironlake_crtc_disable(struct drm_crtc *crtc) mutex_lock(&dev->struct_mutex); intel_update_fbc(dev); - intel_clear_scanline_wait(dev); mutex_unlock(&dev->struct_mutex); } @@ -2859,6 +2919,12 @@ static void ironlake_crtc_dpms(struct drm_crtc *crtc, int mode) } } +static void ironlake_crtc_off(struct drm_crtc *crtc) +{ + struct intel_crtc *intel_crtc = to_intel_crtc(crtc); + intel_put_pch_pll(intel_crtc); +} + static void intel_crtc_dpms_overlay(struct intel_crtc *intel_crtc, bool enable) { if (!enable && intel_crtc->overlay) { @@ -2930,7 +2996,6 @@ static void i9xx_crtc_disable(struct drm_crtc *crtc) intel_crtc->active = false; intel_update_fbc(dev); intel_update_watermarks(dev); - intel_clear_scanline_wait(dev); } static void i9xx_crtc_dpms(struct drm_crtc *crtc, int mode) @@ -2950,6 +3015,10 @@ static void i9xx_crtc_dpms(struct drm_crtc *crtc, int mode) } } +static void i9xx_crtc_off(struct drm_crtc *crtc) +{ +} + /** * Sets the power management mode of the pipe and plane. */ @@ -2997,8 +3066,11 @@ static void intel_crtc_disable(struct drm_crtc *crtc) { struct drm_crtc_helper_funcs *crtc_funcs = crtc->helper_private; struct drm_device *dev = crtc->dev; + struct drm_i915_private *dev_priv = dev->dev_private; crtc_funcs->dpms(crtc, DRM_MODE_DPMS_OFF); + dev_priv->display.off(crtc); + assert_plane_disabled(dev->dev_private, to_intel_crtc(crtc)->plane); assert_pipe_disabled(dev->dev_private, to_intel_crtc(crtc)->pipe); @@ -3822,7 +3894,6 @@ static int i9xx_crtc_mode_set(struct drm_crtc *crtc, I915_WRITE(DSPCNTR(plane), dspcntr); POSTING_READ(DSPCNTR(plane)); - intel_enable_plane(dev_priv, plane, pipe); ret = intel_pipe_set_base(crtc, x, y, old_fb); @@ -4241,29 +4312,18 @@ static int ironlake_crtc_mode_set(struct drm_crtc *crtc, DRM_DEBUG_KMS("Mode for pipe %d:\n", pipe); drm_mode_debug_printmodeline(mode); - /* PCH eDP needs FDI, but CPU eDP does not */ - if (!intel_crtc->no_pll) { - if (!is_cpu_edp) { - I915_WRITE(PCH_FP0(pipe), fp); - I915_WRITE(PCH_DPLL(pipe), dpll & ~DPLL_VCO_ENABLE); + /* CPU eDP is the only output that doesn't need a PCH PLL of its own */ + if (!is_cpu_edp) { + struct intel_pch_pll *pll; - POSTING_READ(PCH_DPLL(pipe)); - udelay(150); - } - } else { - if (dpll == (I915_READ(PCH_DPLL(0)) & 0x7fffffff) && - fp == I915_READ(PCH_FP0(0))) { - intel_crtc->use_pll_a = true; - DRM_DEBUG_KMS("using pipe a dpll\n"); - } else if (dpll == (I915_READ(PCH_DPLL(1)) & 0x7fffffff) && - fp == I915_READ(PCH_FP0(1))) { - intel_crtc->use_pll_a = false; - DRM_DEBUG_KMS("using pipe b dpll\n"); - } else { - DRM_DEBUG_KMS("no matching PLL configuration for pipe 2\n"); + pll = intel_get_pch_pll(intel_crtc, dpll, fp); + if (pll == NULL) { + DRM_DEBUG_DRIVER("failed to find PLL for pipe %d\n", + pipe); return -EINVAL; } - } + } else + intel_put_pch_pll(intel_crtc); /* The LVDS pin pair needs to be on before the DPLLs are enabled. * This is an exception to the general rule that mode_set doesn't turn @@ -4320,11 +4380,11 @@ static int ironlake_crtc_mode_set(struct drm_crtc *crtc, I915_WRITE(TRANSDPLINK_N1(pipe), 0); } - if (!intel_crtc->no_pll && (!edp_encoder || is_pch_edp)) { - I915_WRITE(PCH_DPLL(pipe), dpll); + if (intel_crtc->pch_pll) { + I915_WRITE(intel_crtc->pch_pll->pll_reg, dpll); /* Wait for the clocks to stabilize. */ - POSTING_READ(PCH_DPLL(pipe)); + POSTING_READ(intel_crtc->pch_pll->pll_reg); udelay(150); /* The pixel multiplier can only be updated once the @@ -4332,20 +4392,20 @@ static int ironlake_crtc_mode_set(struct drm_crtc *crtc, * * So write it again. */ - I915_WRITE(PCH_DPLL(pipe), dpll); + I915_WRITE(intel_crtc->pch_pll->pll_reg, dpll); } intel_crtc->lowfreq_avail = false; - if (!intel_crtc->no_pll) { + if (intel_crtc->pch_pll) { if (is_lvds && has_reduced_clock && i915_powersave) { - I915_WRITE(PCH_FP1(pipe), fp2); + I915_WRITE(intel_crtc->pch_pll->fp1_reg, fp2); intel_crtc->lowfreq_avail = true; if (HAS_PIPE_CXSR(dev)) { DRM_DEBUG_KMS("enabling CxSR downclocking\n"); pipeconf |= PIPECONF_CXSR_DOWNCLOCK; } } else { - I915_WRITE(PCH_FP1(pipe), fp); + I915_WRITE(intel_crtc->pch_pll->fp1_reg, fp); if (HAS_PIPE_CXSR(dev)) { DRM_DEBUG_KMS("disabling CxSR downclocking\n"); pipeconf &= ~PIPECONF_CXSR_DOWNCLOCK; @@ -4777,9 +4837,6 @@ static void intel_crtc_update_cursor(struct drm_crtc *crtc, else i9xx_update_cursor(crtc, base); } - - if (visible) - intel_mark_busy(dev, to_intel_framebuffer(crtc->fb)->obj); } static int intel_crtc_cursor_set(struct drm_crtc *crtc, @@ -5303,7 +5360,6 @@ struct drm_display_mode *intel_crtc_mode_get(struct drm_device *dev, mode->vsync_end = ((vsync & 0xffff0000) >> 16) + 1; drm_mode_set_name(mode); - drm_mode_set_crtcinfo(mode, 0); return mode; } @@ -5475,9 +5531,10 @@ void intel_mark_busy(struct drm_device *dev, struct drm_i915_gem_object *obj) if (!drm_core_check_feature(dev, DRIVER_MODESET)) return; - if (!dev_priv->busy) + if (!dev_priv->busy) { + intel_sanitize_pm(dev); dev_priv->busy = true; - else + } else mod_timer(&dev_priv->idle_timer, jiffies + msecs_to_jiffies(GPU_IDLE_TIMEOUT)); @@ -5653,16 +5710,17 @@ static int intel_gen2_queue_flip(struct drm_device *dev, struct intel_crtc *intel_crtc = to_intel_crtc(crtc); unsigned long offset; u32 flip_mask; + struct intel_ring_buffer *ring = &dev_priv->ring[RCS]; int ret; - ret = intel_pin_and_fence_fb_obj(dev, obj, LP_RING(dev_priv)); + ret = intel_pin_and_fence_fb_obj(dev, obj, ring); if (ret) goto err; /* Offset into the new buffer for cases of shared fbs between CRTCs */ offset = crtc->y * fb->pitches[0] + crtc->x * fb->bits_per_pixel/8; - ret = BEGIN_LP_RING(6); + ret = intel_ring_begin(ring, 6); if (ret) goto err_unpin; @@ -5673,14 +5731,14 @@ static int intel_gen2_queue_flip(struct drm_device *dev, flip_mask = MI_WAIT_FOR_PLANE_B_FLIP; else flip_mask = MI_WAIT_FOR_PLANE_A_FLIP; - OUT_RING(MI_WAIT_FOR_EVENT | flip_mask); - OUT_RING(MI_NOOP); - OUT_RING(MI_DISPLAY_FLIP | - MI_DISPLAY_FLIP_PLANE(intel_crtc->plane)); - OUT_RING(fb->pitches[0]); - OUT_RING(obj->gtt_offset + offset); - OUT_RING(0); /* aux display base address, unused */ - ADVANCE_LP_RING(); + intel_ring_emit(ring, MI_WAIT_FOR_EVENT | flip_mask); + intel_ring_emit(ring, MI_NOOP); + intel_ring_emit(ring, MI_DISPLAY_FLIP | + MI_DISPLAY_FLIP_PLANE(intel_crtc->plane)); + intel_ring_emit(ring, fb->pitches[0]); + intel_ring_emit(ring, obj->gtt_offset + offset); + intel_ring_emit(ring, 0); /* aux display base address, unused */ + intel_ring_advance(ring); return 0; err_unpin: @@ -5698,16 +5756,17 @@ static int intel_gen3_queue_flip(struct drm_device *dev, struct intel_crtc *intel_crtc = to_intel_crtc(crtc); unsigned long offset; u32 flip_mask; + struct intel_ring_buffer *ring = &dev_priv->ring[RCS]; int ret; - ret = intel_pin_and_fence_fb_obj(dev, obj, LP_RING(dev_priv)); + ret = intel_pin_and_fence_fb_obj(dev, obj, ring); if (ret) goto err; /* Offset into the new buffer for cases of shared fbs between CRTCs */ offset = crtc->y * fb->pitches[0] + crtc->x * fb->bits_per_pixel/8; - ret = BEGIN_LP_RING(6); + ret = intel_ring_begin(ring, 6); if (ret) goto err_unpin; @@ -5715,15 +5774,15 @@ static int intel_gen3_queue_flip(struct drm_device *dev, flip_mask = MI_WAIT_FOR_PLANE_B_FLIP; else flip_mask = MI_WAIT_FOR_PLANE_A_FLIP; - OUT_RING(MI_WAIT_FOR_EVENT | flip_mask); - OUT_RING(MI_NOOP); - OUT_RING(MI_DISPLAY_FLIP_I915 | - MI_DISPLAY_FLIP_PLANE(intel_crtc->plane)); - OUT_RING(fb->pitches[0]); - OUT_RING(obj->gtt_offset + offset); - OUT_RING(MI_NOOP); - - ADVANCE_LP_RING(); + intel_ring_emit(ring, MI_WAIT_FOR_EVENT | flip_mask); + intel_ring_emit(ring, MI_NOOP); + intel_ring_emit(ring, MI_DISPLAY_FLIP_I915 | + MI_DISPLAY_FLIP_PLANE(intel_crtc->plane)); + intel_ring_emit(ring, fb->pitches[0]); + intel_ring_emit(ring, obj->gtt_offset + offset); + intel_ring_emit(ring, MI_NOOP); + + intel_ring_advance(ring); return 0; err_unpin: @@ -5740,13 +5799,14 @@ static int intel_gen4_queue_flip(struct drm_device *dev, struct drm_i915_private *dev_priv = dev->dev_private; struct intel_crtc *intel_crtc = to_intel_crtc(crtc); uint32_t pf, pipesrc; + struct intel_ring_buffer *ring = &dev_priv->ring[RCS]; int ret; - ret = intel_pin_and_fence_fb_obj(dev, obj, LP_RING(dev_priv)); + ret = intel_pin_and_fence_fb_obj(dev, obj, ring); if (ret) goto err; - ret = BEGIN_LP_RING(4); + ret = intel_ring_begin(ring, 4); if (ret) goto err_unpin; @@ -5754,10 +5814,10 @@ static int intel_gen4_queue_flip(struct drm_device *dev, * Display Registers (which do not change across a page-flip) * so we need only reprogram the base address. */ - OUT_RING(MI_DISPLAY_FLIP | - MI_DISPLAY_FLIP_PLANE(intel_crtc->plane)); - OUT_RING(fb->pitches[0]); - OUT_RING(obj->gtt_offset | obj->tiling_mode); + intel_ring_emit(ring, MI_DISPLAY_FLIP | + MI_DISPLAY_FLIP_PLANE(intel_crtc->plane)); + intel_ring_emit(ring, fb->pitches[0]); + intel_ring_emit(ring, obj->gtt_offset | obj->tiling_mode); /* XXX Enabling the panel-fitter across page-flip is so far * untested on non-native modes, so ignore it for now. @@ -5765,8 +5825,8 @@ static int intel_gen4_queue_flip(struct drm_device *dev, */ pf = 0; pipesrc = I915_READ(PIPESRC(intel_crtc->pipe)) & 0x0fff0fff; - OUT_RING(pf | pipesrc); - ADVANCE_LP_RING(); + intel_ring_emit(ring, pf | pipesrc); + intel_ring_advance(ring); return 0; err_unpin: @@ -5782,21 +5842,22 @@ static int intel_gen6_queue_flip(struct drm_device *dev, { struct drm_i915_private *dev_priv = dev->dev_private; struct intel_crtc *intel_crtc = to_intel_crtc(crtc); + struct intel_ring_buffer *ring = &dev_priv->ring[RCS]; uint32_t pf, pipesrc; int ret; - ret = intel_pin_and_fence_fb_obj(dev, obj, LP_RING(dev_priv)); + ret = intel_pin_and_fence_fb_obj(dev, obj, ring); if (ret) goto err; - ret = BEGIN_LP_RING(4); + ret = intel_ring_begin(ring, 4); if (ret) goto err_unpin; - OUT_RING(MI_DISPLAY_FLIP | - MI_DISPLAY_FLIP_PLANE(intel_crtc->plane)); - OUT_RING(fb->pitches[0] | obj->tiling_mode); - OUT_RING(obj->gtt_offset); + intel_ring_emit(ring, MI_DISPLAY_FLIP | + MI_DISPLAY_FLIP_PLANE(intel_crtc->plane)); + intel_ring_emit(ring, fb->pitches[0] | obj->tiling_mode); + intel_ring_emit(ring, obj->gtt_offset); /* Contrary to the suggestions in the documentation, * "Enable Panel Fitter" does not seem to be required when page @@ -5806,8 +5867,8 @@ static int intel_gen6_queue_flip(struct drm_device *dev, */ pf = 0; pipesrc = I915_READ(PIPESRC(intel_crtc->pipe)) & 0x0fff0fff; - OUT_RING(pf | pipesrc); - ADVANCE_LP_RING(); + intel_ring_emit(ring, pf | pipesrc); + intel_ring_advance(ring); return 0; err_unpin: @@ -6027,6 +6088,23 @@ static const struct drm_crtc_funcs intel_crtc_funcs = { .page_flip = intel_crtc_page_flip, }; +static void intel_pch_pll_init(struct drm_device *dev) +{ + drm_i915_private_t *dev_priv = dev->dev_private; + int i; + + if (dev_priv->num_pch_pll == 0) { + DRM_DEBUG_KMS("No PCH PLLs on this hardware, skipping initialisation\n"); + return; + } + + for (i = 0; i < dev_priv->num_pch_pll; i++) { + dev_priv->pch_plls[i].pll_reg = _PCH_DPLL(i); + dev_priv->pch_plls[i].fp0_reg = _PCH_FP0(i); + dev_priv->pch_plls[i].fp1_reg = _PCH_FP1(i); + } +} + static void intel_crtc_init(struct drm_device *dev, int pipe) { drm_i915_private_t *dev_priv = dev->dev_private; @@ -6064,8 +6142,6 @@ static void intel_crtc_init(struct drm_device *dev, int pipe) intel_crtc->bpp = 24; /* default for pre-Ironlake */ if (HAS_PCH_SPLIT(dev)) { - if (pipe == 2 && IS_IVYBRIDGE(dev)) - intel_crtc->no_pll = true; intel_helper_funcs.prepare = ironlake_crtc_prepare; intel_helper_funcs.commit = ironlake_crtc_commit; } else { @@ -6084,15 +6160,12 @@ static void intel_crtc_init(struct drm_device *dev, int pipe) int intel_get_pipe_from_crtc_id(struct drm_device *dev, void *data, struct drm_file *file) { - drm_i915_private_t *dev_priv = dev->dev_private; struct drm_i915_get_pipe_from_crtc_id *pipe_from_crtc_id = data; struct drm_mode_object *drmmode_obj; struct intel_crtc *crtc; - if (!dev_priv) { - DRM_ERROR("called with no initialization\n"); - return -EINVAL; - } + if (!drm_core_check_feature(dev, DRIVER_MODESET)) + return -ENODEV; drmmode_obj = drm_mode_object_find(dev, pipe_from_crtc_id->crtc_id, DRM_MODE_OBJECT_CRTC); @@ -6348,10 +6421,12 @@ static void intel_init_display(struct drm_device *dev) if (HAS_PCH_SPLIT(dev)) { dev_priv->display.dpms = ironlake_crtc_dpms; dev_priv->display.crtc_mode_set = ironlake_crtc_mode_set; + dev_priv->display.off = ironlake_crtc_off; dev_priv->display.update_plane = ironlake_update_plane; } else { dev_priv->display.dpms = i9xx_crtc_dpms; dev_priv->display.crtc_mode_set = i9xx_crtc_mode_set; + dev_priv->display.off = i9xx_crtc_off; dev_priv->display.update_plane = i9xx_update_plane; } @@ -6614,6 +6689,8 @@ void intel_modeset_init(struct drm_device *dev) DRM_DEBUG_KMS("plane %d init failed: %d\n", i, ret); } + intel_pch_pll_init(dev); + /* Just disable it once at startup */ i915_disable_vga(dev); intel_setup_outputs(dev); diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 44cf32c..a1a5ce7 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -688,7 +688,7 @@ intel_dp_mode_fixup(struct drm_encoder *encoder, struct drm_display_mode *mode, int lane_count, clock; int max_lane_count = intel_dp_max_lane_count(intel_dp); int max_clock = intel_dp_max_link_bw(intel_dp) == DP_LINK_BW_2_7 ? 1 : 0; - int bpp; + int bpp, mode_rate; static int bws[2] = { DP_LINK_BW_1_62, DP_LINK_BW_2_7 }; if (is_edp(intel_dp) && intel_dp->panel_fixed_mode) { @@ -702,24 +702,30 @@ intel_dp_mode_fixup(struct drm_encoder *encoder, struct drm_display_mode *mode, mode->clock = intel_dp->panel_fixed_mode->clock; } + DRM_DEBUG_KMS("DP link computation with max lane count %i " + "max bw %02x pixel clock %iKHz\n", + max_lane_count, bws[max_clock], mode->clock); + if (!intel_dp_adjust_dithering(intel_dp, mode, adjusted_mode)) return false; bpp = adjusted_mode->private_flags & INTEL_MODE_DP_FORCE_6BPC ? 18 : 24; + mode_rate = intel_dp_link_required(mode->clock, bpp); for (lane_count = 1; lane_count <= max_lane_count; lane_count <<= 1) { for (clock = 0; clock <= max_clock; clock++) { int link_avail = intel_dp_max_data_rate(intel_dp_link_clock(bws[clock]), lane_count); - if (intel_dp_link_required(mode->clock, bpp) - <= link_avail) { + if (mode_rate <= link_avail) { intel_dp->link_bw = bws[clock]; intel_dp->lane_count = lane_count; adjusted_mode->clock = intel_dp_link_clock(intel_dp->link_bw); - DRM_DEBUG_KMS("Display port link bw %02x lane " - "count %d clock %d\n", + DRM_DEBUG_KMS("DP link bw %02x lane " + "count %d clock %d bpp %d\n", intel_dp->link_bw, intel_dp->lane_count, - adjusted_mode->clock); + adjusted_mode->clock, bpp); + DRM_DEBUG_KMS("DP link bw required %i available %i\n", + mode_rate, link_avail); return true; } } @@ -2439,6 +2445,7 @@ intel_dp_init(struct drm_device *dev, int output_reg) } intel_encoder->crtc_mask = (1 << 0) | (1 << 1) | (1 << 2); + connector->interlace_allowed = true; connector->doublescan_allowed = 0; diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index 7a7cae7..e5ee166 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -183,8 +183,8 @@ struct intel_crtc { bool cursor_visible; unsigned int bpp; - bool no_pll; /* tertiary pipe for IVB */ - bool use_pll_a; + /* We can share PLLs across outputs if the timings match */ + struct intel_pch_pll *pch_pll; }; struct intel_plane { @@ -238,6 +238,8 @@ struct cxsr_latency { #define DIP_TYPE_AVI 0x82 #define DIP_VERSION_AVI 0x2 #define DIP_LEN_AVI 13 +#define DIP_AVI_PR_1 0 +#define DIP_AVI_PR_2 1 #define DIP_TYPE_SPD 0x83 #define DIP_VERSION_SPD 0x1 @@ -271,8 +273,8 @@ struct dip_infoframe { uint8_t ITC_EC_Q_SC; /* PB4 - VIC 6:0 */ uint8_t VIC; - /* PB5 - PR 3:0 */ - uint8_t PR; + /* PB5 - YQ 7:6, CN 5:4, PR 3:0 */ + uint8_t YQ_CN_PR; /* PB6 to PB13 */ uint16_t top_bar_end; uint16_t bottom_bar_start; @@ -346,6 +348,8 @@ extern int intel_plane_init(struct drm_device *dev, enum pipe pipe); extern void intel_flush_display_plane(struct drm_i915_private *dev_priv, enum plane plane); +void intel_sanitize_pm(struct drm_device *dev); + /* intel_panel.c */ extern void intel_fixed_panel_mode(struct drm_display_mode *fixed_mode, struct drm_display_mode *adjusted_mode); @@ -405,10 +409,6 @@ extern void intel_enable_clock_gating(struct drm_device *dev); extern void ironlake_disable_rc6(struct drm_device *dev); extern void ironlake_enable_drps(struct drm_device *dev); extern void ironlake_disable_drps(struct drm_device *dev); -extern void gen6_enable_rps(struct drm_i915_private *dev_priv); -extern void gen6_update_ring_freq(struct drm_i915_private *dev_priv); -extern void gen6_disable_rps(struct drm_device *dev); -extern void intel_init_emon(struct drm_device *dev); extern int intel_pin_and_fence_fb_obj(struct drm_device *dev, struct drm_i915_gem_object *obj, @@ -466,5 +466,13 @@ extern void intel_init_pm(struct drm_device *dev); extern bool intel_fbc_enabled(struct drm_device *dev); extern void intel_enable_fbc(struct drm_crtc *crtc, unsigned long interval); extern void intel_update_fbc(struct drm_device *dev); +/* IPS */ +extern void intel_gpu_ips_init(struct drm_i915_private *dev_priv); +extern void intel_gpu_ips_teardown(void); + +extern void gen6_enable_rps(struct drm_i915_private *dev_priv); +extern void gen6_update_ring_freq(struct drm_i915_private *dev_priv); +extern void gen6_disable_rps(struct drm_device *dev); +extern void intel_init_emon(struct drm_device *dev); #endif /* __INTEL_DRV_H__ */ diff --git a/drivers/gpu/drm/i915/intel_hdmi.c b/drivers/gpu/drm/i915/intel_hdmi.c index 1eef50d..bf21875 100644 --- a/drivers/gpu/drm/i915/intel_hdmi.c +++ b/drivers/gpu/drm/i915/intel_hdmi.c @@ -220,7 +220,8 @@ static void intel_set_infoframe(struct drm_encoder *encoder, intel_hdmi->write_infoframe(encoder, frame); } -static void intel_hdmi_set_avi_infoframe(struct drm_encoder *encoder) +static void intel_hdmi_set_avi_infoframe(struct drm_encoder *encoder, + struct drm_display_mode *adjusted_mode) { struct dip_infoframe avi_if = { .type = DIP_TYPE_AVI, @@ -228,6 +229,9 @@ static void intel_hdmi_set_avi_infoframe(struct drm_encoder *encoder) .len = DIP_LEN_AVI, }; + if (adjusted_mode->flags & DRM_MODE_FLAG_DBLCLK) + avi_if.body.avi.YQ_CN_PR |= DIP_AVI_PR_2; + intel_set_infoframe(encoder, &avi_if); } @@ -290,7 +294,7 @@ static void intel_hdmi_mode_set(struct drm_encoder *encoder, I915_WRITE(intel_hdmi->sdvox_reg, sdvox); POSTING_READ(intel_hdmi->sdvox_reg); - intel_hdmi_set_avi_infoframe(encoder); + intel_hdmi_set_avi_infoframe(encoder, adjusted_mode); intel_hdmi_set_spd_infoframe(encoder); } diff --git a/drivers/gpu/drm/i915/intel_opregion.c b/drivers/gpu/drm/i915/intel_opregion.c index 34929ae..18bd0af 100644 --- a/drivers/gpu/drm/i915/intel_opregion.c +++ b/drivers/gpu/drm/i915/intel_opregion.c @@ -151,7 +151,7 @@ struct opregion_asle { static u32 asle_set_backlight(struct drm_device *dev, u32 bclp) { struct drm_i915_private *dev_priv = dev->dev_private; - struct opregion_asle *asle = dev_priv->opregion.asle; + struct opregion_asle __iomem *asle = dev_priv->opregion.asle; u32 max; if (!(bclp & ASLE_BCLP_VALID)) @@ -163,7 +163,7 @@ static u32 asle_set_backlight(struct drm_device *dev, u32 bclp) max = intel_panel_get_max_backlight(dev); intel_panel_set_backlight(dev, bclp * max / 255); - asle->cblv = (bclp*0x64)/0xff | ASLE_CBLV_VALID; + iowrite32((bclp*0x64)/0xff | ASLE_CBLV_VALID, &asle->cblv); return 0; } @@ -200,14 +200,14 @@ static u32 asle_set_pfit(struct drm_device *dev, u32 pfit) void intel_opregion_asle_intr(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; - struct opregion_asle *asle = dev_priv->opregion.asle; + struct opregion_asle __iomem *asle = dev_priv->opregion.asle; u32 asle_stat = 0; u32 asle_req; if (!asle) return; - asle_req = asle->aslc & ASLE_REQ_MSK; + asle_req = ioread32(&asle->aslc) & ASLE_REQ_MSK; if (!asle_req) { DRM_DEBUG_DRIVER("non asle set request??\n"); @@ -215,31 +215,31 @@ void intel_opregion_asle_intr(struct drm_device *dev) } if (asle_req & ASLE_SET_ALS_ILLUM) - asle_stat |= asle_set_als_illum(dev, asle->alsi); + asle_stat |= asle_set_als_illum(dev, ioread32(&asle->alsi)); if (asle_req & ASLE_SET_BACKLIGHT) - asle_stat |= asle_set_backlight(dev, asle->bclp); + asle_stat |= asle_set_backlight(dev, ioread32(&asle->bclp)); if (asle_req & ASLE_SET_PFIT) - asle_stat |= asle_set_pfit(dev, asle->pfit); + asle_stat |= asle_set_pfit(dev, ioread32(&asle->pfit)); if (asle_req & ASLE_SET_PWM_FREQ) - asle_stat |= asle_set_pwm_freq(dev, asle->pfmb); + asle_stat |= asle_set_pwm_freq(dev, ioread32(&asle->pfmb)); - asle->aslc = asle_stat; + iowrite32(asle_stat, &asle->aslc); } void intel_opregion_gse_intr(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; - struct opregion_asle *asle = dev_priv->opregion.asle; + struct opregion_asle __iomem *asle = dev_priv->opregion.asle; u32 asle_stat = 0; u32 asle_req; if (!asle) return; - asle_req = asle->aslc & ASLE_REQ_MSK; + asle_req = ioread32(&asle->aslc) & ASLE_REQ_MSK; if (!asle_req) { DRM_DEBUG_DRIVER("non asle set request??\n"); @@ -252,7 +252,7 @@ void intel_opregion_gse_intr(struct drm_device *dev) } if (asle_req & ASLE_SET_BACKLIGHT) - asle_stat |= asle_set_backlight(dev, asle->bclp); + asle_stat |= asle_set_backlight(dev, ioread32(&asle->bclp)); if (asle_req & ASLE_SET_PFIT) { DRM_DEBUG_DRIVER("Pfit is not supported\n"); @@ -264,7 +264,7 @@ void intel_opregion_gse_intr(struct drm_device *dev) asle_stat |= ASLE_PWM_FREQ_FAILED; } - asle->aslc = asle_stat; + iowrite32(asle_stat, &asle->aslc); } #define ASLE_ALS_EN (1<<0) #define ASLE_BLC_EN (1<<1) @@ -274,15 +274,16 @@ void intel_opregion_gse_intr(struct drm_device *dev) void intel_opregion_enable_asle(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; - struct opregion_asle *asle = dev_priv->opregion.asle; + struct opregion_asle __iomem *asle = dev_priv->opregion.asle; if (asle) { if (IS_MOBILE(dev)) intel_enable_asle(dev); - asle->tche = ASLE_ALS_EN | ASLE_BLC_EN | ASLE_PFIT_EN | - ASLE_PFMB_EN; - asle->ardy = 1; + iowrite32(ASLE_ALS_EN | ASLE_BLC_EN | ASLE_PFIT_EN | + ASLE_PFMB_EN, + &asle->tche); + iowrite32(1, &asle->ardy); } } @@ -300,7 +301,7 @@ static int intel_opregion_video_event(struct notifier_block *nb, Linux, these are handled by the dock, button and video drivers. */ - struct opregion_acpi *acpi; + struct opregion_acpi __iomem *acpi; struct acpi_bus_event *event = data; int ret = NOTIFY_OK; @@ -312,10 +313,11 @@ static int intel_opregion_video_event(struct notifier_block *nb, acpi = system_opregion->acpi; - if (event->type == 0x80 && !(acpi->cevt & 0x1)) + if (event->type == 0x80 && + (ioread32(&acpi->cevt) & 1) == 0) ret = NOTIFY_BAD; - acpi->csts = 0; + iowrite32(0, &acpi->csts); return ret; } @@ -339,6 +341,7 @@ static void intel_didl_outputs(struct drm_device *dev) struct acpi_device *acpi_dev, *acpi_cdev, *acpi_video_bus = NULL; unsigned long long device_id; acpi_status status; + u32 temp; int i = 0; handle = DEVICE_ACPI_HANDLE(&dev->pdev->dev); @@ -373,7 +376,8 @@ static void intel_didl_outputs(struct drm_device *dev) if (ACPI_SUCCESS(status)) { if (!device_id) goto blind_set; - opregion->acpi->didl[i] = (u32)(device_id & 0x0f0f); + iowrite32((u32)(device_id & 0x0f0f), + &opregion->acpi->didl[i]); i++; } } @@ -381,7 +385,7 @@ static void intel_didl_outputs(struct drm_device *dev) end: /* If fewer than 8 outputs, the list must be null terminated */ if (i < 8) - opregion->acpi->didl[i] = 0; + iowrite32(0, &opregion->acpi->didl[i]); return; blind_set: @@ -415,7 +419,9 @@ blind_set: output_type = ACPI_LVDS_OUTPUT; break; } - opregion->acpi->didl[i] |= (1<<31) | output_type | i; + temp = ioread32(&opregion->acpi->didl[i]); + iowrite32(temp | (1<<31) | output_type | i, + &opregion->acpi->didl[i]); i++; } goto end; @@ -436,8 +442,8 @@ void intel_opregion_init(struct drm_device *dev) /* Notify BIOS we are ready to handle ACPI video ext notifs. * Right now, all the events are handled by the ACPI video module. * We don't actually need to do anything with them. */ - opregion->acpi->csts = 0; - opregion->acpi->drdy = 1; + iowrite32(0, &opregion->acpi->csts); + iowrite32(1, &opregion->acpi->drdy); system_opregion = opregion; register_acpi_notifier(&intel_opregion_notifier); @@ -456,7 +462,7 @@ void intel_opregion_fini(struct drm_device *dev) return; if (opregion->acpi) { - opregion->acpi->drdy = 0; + iowrite32(0, &opregion->acpi->drdy); system_opregion = NULL; unregister_acpi_notifier(&intel_opregion_notifier); @@ -476,8 +482,9 @@ int intel_opregion_setup(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; struct intel_opregion *opregion = &dev_priv->opregion; - void *base; + void __iomem *base; u32 asls, mboxes; + char buf[sizeof(OPREGION_SIGNATURE)]; int err = 0; pci_read_config_dword(dev->pdev, PCI_ASLS, &asls); @@ -491,7 +498,9 @@ int intel_opregion_setup(struct drm_device *dev) if (!base) return -ENOMEM; - if (memcmp(base, OPREGION_SIGNATURE, 16)) { + memcpy_fromio(buf, base, sizeof(buf)); + + if (memcmp(buf, OPREGION_SIGNATURE, 16)) { DRM_DEBUG_DRIVER("opregion signature mismatch\n"); err = -EINVAL; goto err_out; @@ -501,7 +510,7 @@ int intel_opregion_setup(struct drm_device *dev) opregion->lid_state = base + ACPI_CLID; - mboxes = opregion->header->mboxes; + mboxes = ioread32(&opregion->header->mboxes); if (mboxes & MBOX_ACPI) { DRM_DEBUG_DRIVER("Public ACPI methods supported\n"); opregion->acpi = base + OPREGION_ACPI_OFFSET; diff --git a/drivers/gpu/drm/i915/intel_overlay.c b/drivers/gpu/drm/i915/intel_overlay.c index 80b331c..458743d 100644 --- a/drivers/gpu/drm/i915/intel_overlay.c +++ b/drivers/gpu/drm/i915/intel_overlay.c @@ -187,14 +187,14 @@ struct intel_overlay { void (*flip_tail)(struct intel_overlay *); }; -static struct overlay_registers * +static struct overlay_registers __iomem * intel_overlay_map_regs(struct intel_overlay *overlay) { drm_i915_private_t *dev_priv = overlay->dev->dev_private; - struct overlay_registers *regs; + struct overlay_registers __iomem *regs; if (OVERLAY_NEEDS_PHYSICAL(overlay->dev)) - regs = overlay->reg_bo->phys_obj->handle->vaddr; + regs = (struct overlay_registers __iomem *)overlay->reg_bo->phys_obj->handle->vaddr; else regs = io_mapping_map_wc(dev_priv->mm.gtt_mapping, overlay->reg_bo->gtt_offset); @@ -203,7 +203,7 @@ intel_overlay_map_regs(struct intel_overlay *overlay) } static void intel_overlay_unmap_regs(struct intel_overlay *overlay, - struct overlay_registers *regs) + struct overlay_registers __iomem *regs) { if (!OVERLAY_NEEDS_PHYSICAL(overlay->dev)) io_mapping_unmap(regs); @@ -215,20 +215,21 @@ static int intel_overlay_do_wait_request(struct intel_overlay *overlay, { struct drm_device *dev = overlay->dev; drm_i915_private_t *dev_priv = dev->dev_private; + struct intel_ring_buffer *ring = &dev_priv->ring[RCS]; int ret; BUG_ON(overlay->last_flip_req); - ret = i915_add_request(LP_RING(dev_priv), NULL, request); + ret = i915_add_request(ring, NULL, request); if (ret) { kfree(request); return ret; } overlay->last_flip_req = request->seqno; overlay->flip_tail = tail; - ret = i915_wait_request(LP_RING(dev_priv), overlay->last_flip_req, - true); + ret = i915_wait_request(ring, overlay->last_flip_req); if (ret) return ret; + i915_gem_retire_requests(dev); overlay->last_flip_req = 0; return 0; @@ -262,7 +263,7 @@ i830_activate_pipe_a(struct drm_device *dev) DRM_DEBUG_DRIVER("Enabling pipe A in order to enable overlay\n"); mode = drm_mode_duplicate(dev, &vesa_640x480); - drm_mode_set_crtcinfo(mode, 0); + if (!drm_crtc_helper_set_mode(&crtc->base, mode, crtc->base.x, crtc->base.y, crtc->base.fb)) @@ -287,6 +288,7 @@ static int intel_overlay_on(struct intel_overlay *overlay) { struct drm_device *dev = overlay->dev; struct drm_i915_private *dev_priv = dev->dev_private; + struct intel_ring_buffer *ring = &dev_priv->ring[RCS]; struct drm_i915_gem_request *request; int pipe_a_quirk = 0; int ret; @@ -306,17 +308,17 @@ static int intel_overlay_on(struct intel_overlay *overlay) goto out; } - ret = BEGIN_LP_RING(4); + ret = intel_ring_begin(ring, 4); if (ret) { kfree(request); goto out; } - OUT_RING(MI_OVERLAY_FLIP | MI_OVERLAY_ON); - OUT_RING(overlay->flip_addr | OFC_UPDATE); - OUT_RING(MI_WAIT_FOR_EVENT | MI_WAIT_FOR_OVERLAY_FLIP); - OUT_RING(MI_NOOP); - ADVANCE_LP_RING(); + intel_ring_emit(ring, MI_OVERLAY_FLIP | MI_OVERLAY_ON); + intel_ring_emit(ring, overlay->flip_addr | OFC_UPDATE); + intel_ring_emit(ring, MI_WAIT_FOR_EVENT | MI_WAIT_FOR_OVERLAY_FLIP); + intel_ring_emit(ring, MI_NOOP); + intel_ring_advance(ring); ret = intel_overlay_do_wait_request(overlay, request, NULL); out: @@ -332,6 +334,7 @@ static int intel_overlay_continue(struct intel_overlay *overlay, { struct drm_device *dev = overlay->dev; drm_i915_private_t *dev_priv = dev->dev_private; + struct intel_ring_buffer *ring = &dev_priv->ring[RCS]; struct drm_i915_gem_request *request; u32 flip_addr = overlay->flip_addr; u32 tmp; @@ -351,16 +354,16 @@ static int intel_overlay_continue(struct intel_overlay *overlay, if (tmp & (1 << 17)) DRM_DEBUG("overlay underrun, DOVSTA: %x\n", tmp); - ret = BEGIN_LP_RING(2); + ret = intel_ring_begin(ring, 2); if (ret) { kfree(request); return ret; } - OUT_RING(MI_OVERLAY_FLIP | MI_OVERLAY_CONTINUE); - OUT_RING(flip_addr); - ADVANCE_LP_RING(); + intel_ring_emit(ring, MI_OVERLAY_FLIP | MI_OVERLAY_CONTINUE); + intel_ring_emit(ring, flip_addr); + intel_ring_advance(ring); - ret = i915_add_request(LP_RING(dev_priv), NULL, request); + ret = i915_add_request(ring, NULL, request); if (ret) { kfree(request); return ret; @@ -401,6 +404,7 @@ static int intel_overlay_off(struct intel_overlay *overlay) { struct drm_device *dev = overlay->dev; struct drm_i915_private *dev_priv = dev->dev_private; + struct intel_ring_buffer *ring = &dev_priv->ring[RCS]; u32 flip_addr = overlay->flip_addr; struct drm_i915_gem_request *request; int ret; @@ -417,20 +421,20 @@ static int intel_overlay_off(struct intel_overlay *overlay) * of the hw. Do it in both cases */ flip_addr |= OFC_UPDATE; - ret = BEGIN_LP_RING(6); + ret = intel_ring_begin(ring, 6); if (ret) { kfree(request); return ret; } /* wait for overlay to go idle */ - OUT_RING(MI_OVERLAY_FLIP | MI_OVERLAY_CONTINUE); - OUT_RING(flip_addr); - OUT_RING(MI_WAIT_FOR_EVENT | MI_WAIT_FOR_OVERLAY_FLIP); + intel_ring_emit(ring, MI_OVERLAY_FLIP | MI_OVERLAY_CONTINUE); + intel_ring_emit(ring, flip_addr); + intel_ring_emit(ring, MI_WAIT_FOR_EVENT | MI_WAIT_FOR_OVERLAY_FLIP); /* turn overlay off */ - OUT_RING(MI_OVERLAY_FLIP | MI_OVERLAY_OFF); - OUT_RING(flip_addr); - OUT_RING(MI_WAIT_FOR_EVENT | MI_WAIT_FOR_OVERLAY_FLIP); - ADVANCE_LP_RING(); + intel_ring_emit(ring, MI_OVERLAY_FLIP | MI_OVERLAY_OFF); + intel_ring_emit(ring, flip_addr); + intel_ring_emit(ring, MI_WAIT_FOR_EVENT | MI_WAIT_FOR_OVERLAY_FLIP); + intel_ring_advance(ring); return intel_overlay_do_wait_request(overlay, request, intel_overlay_off_tail); @@ -442,15 +446,16 @@ static int intel_overlay_recover_from_interrupt(struct intel_overlay *overlay) { struct drm_device *dev = overlay->dev; drm_i915_private_t *dev_priv = dev->dev_private; + struct intel_ring_buffer *ring = &dev_priv->ring[RCS]; int ret; if (overlay->last_flip_req == 0) return 0; - ret = i915_wait_request(LP_RING(dev_priv), overlay->last_flip_req, - true); + ret = i915_wait_request(ring, overlay->last_flip_req); if (ret) return ret; + i915_gem_retire_requests(dev); if (overlay->flip_tail) overlay->flip_tail(overlay); @@ -467,6 +472,7 @@ static int intel_overlay_release_old_vid(struct intel_overlay *overlay) { struct drm_device *dev = overlay->dev; drm_i915_private_t *dev_priv = dev->dev_private; + struct intel_ring_buffer *ring = &dev_priv->ring[RCS]; int ret; /* Only wait if there is actually an old frame to release to @@ -483,15 +489,15 @@ static int intel_overlay_release_old_vid(struct intel_overlay *overlay) if (request == NULL) return -ENOMEM; - ret = BEGIN_LP_RING(2); + ret = intel_ring_begin(ring, 2); if (ret) { kfree(request); return ret; } - OUT_RING(MI_WAIT_FOR_EVENT | MI_WAIT_FOR_OVERLAY_FLIP); - OUT_RING(MI_NOOP); - ADVANCE_LP_RING(); + intel_ring_emit(ring, MI_WAIT_FOR_EVENT | MI_WAIT_FOR_OVERLAY_FLIP); + intel_ring_emit(ring, MI_NOOP); + intel_ring_advance(ring); ret = intel_overlay_do_wait_request(overlay, request, intel_overlay_release_old_vid_tail); @@ -619,14 +625,15 @@ static const u16 uv_static_hcoeffs[N_HORIZ_UV_TAPS * N_PHASES] = { 0x3000, 0x0800, 0x3000 }; -static void update_polyphase_filter(struct overlay_registers *regs) +static void update_polyphase_filter(struct overlay_registers __iomem *regs) { - memcpy(regs->Y_HCOEFS, y_static_hcoeffs, sizeof(y_static_hcoeffs)); - memcpy(regs->UV_HCOEFS, uv_static_hcoeffs, sizeof(uv_static_hcoeffs)); + memcpy_toio(regs->Y_HCOEFS, y_static_hcoeffs, sizeof(y_static_hcoeffs)); + memcpy_toio(regs->UV_HCOEFS, uv_static_hcoeffs, + sizeof(uv_static_hcoeffs)); } static bool update_scaling_factors(struct intel_overlay *overlay, - struct overlay_registers *regs, + struct overlay_registers __iomem *regs, struct put_image_params *params) { /* fixed point with a 12 bit shift */ @@ -665,16 +672,19 @@ static bool update_scaling_factors(struct intel_overlay *overlay, overlay->old_xscale = xscale; overlay->old_yscale = yscale; - regs->YRGBSCALE = (((yscale & FRACT_MASK) << 20) | - ((xscale >> FP_SHIFT) << 16) | - ((xscale & FRACT_MASK) << 3)); + iowrite32(((yscale & FRACT_MASK) << 20) | + ((xscale >> FP_SHIFT) << 16) | + ((xscale & FRACT_MASK) << 3), + ®s->YRGBSCALE); - regs->UVSCALE = (((yscale_UV & FRACT_MASK) << 20) | - ((xscale_UV >> FP_SHIFT) << 16) | - ((xscale_UV & FRACT_MASK) << 3)); + iowrite32(((yscale_UV & FRACT_MASK) << 20) | + ((xscale_UV >> FP_SHIFT) << 16) | + ((xscale_UV & FRACT_MASK) << 3), + ®s->UVSCALE); - regs->UVSCALEV = ((((yscale >> FP_SHIFT) << 16) | - ((yscale_UV >> FP_SHIFT) << 0))); + iowrite32((((yscale >> FP_SHIFT) << 16) | + ((yscale_UV >> FP_SHIFT) << 0)), + ®s->UVSCALEV); if (scale_changed) update_polyphase_filter(regs); @@ -683,30 +693,32 @@ static bool update_scaling_factors(struct intel_overlay *overlay, } static void update_colorkey(struct intel_overlay *overlay, - struct overlay_registers *regs) + struct overlay_registers __iomem *regs) { u32 key = overlay->color_key; switch (overlay->crtc->base.fb->bits_per_pixel) { case 8: - regs->DCLRKV = 0; - regs->DCLRKM = CLK_RGB8I_MASK | DST_KEY_ENABLE; + iowrite32(0, ®s->DCLRKV); + iowrite32(CLK_RGB8I_MASK | DST_KEY_ENABLE, ®s->DCLRKM); break; case 16: if (overlay->crtc->base.fb->depth == 15) { - regs->DCLRKV = RGB15_TO_COLORKEY(key); - regs->DCLRKM = CLK_RGB15_MASK | DST_KEY_ENABLE; + iowrite32(RGB15_TO_COLORKEY(key), ®s->DCLRKV); + iowrite32(CLK_RGB15_MASK | DST_KEY_ENABLE, + ®s->DCLRKM); } else { - regs->DCLRKV = RGB16_TO_COLORKEY(key); - regs->DCLRKM = CLK_RGB16_MASK | DST_KEY_ENABLE; + iowrite32(RGB16_TO_COLORKEY(key), ®s->DCLRKV); + iowrite32(CLK_RGB16_MASK | DST_KEY_ENABLE, + ®s->DCLRKM); } break; case 24: case 32: - regs->DCLRKV = key; - regs->DCLRKM = CLK_RGB24_MASK | DST_KEY_ENABLE; + iowrite32(key, ®s->DCLRKV); + iowrite32(CLK_RGB24_MASK | DST_KEY_ENABLE, ®s->DCLRKM); break; } } @@ -761,9 +773,10 @@ static int intel_overlay_do_put_image(struct intel_overlay *overlay, struct put_image_params *params) { int ret, tmp_width; - struct overlay_registers *regs; + struct overlay_registers __iomem *regs; bool scale_changed = false; struct drm_device *dev = overlay->dev; + u32 swidth, swidthsw, sheight, ostride; BUG_ON(!mutex_is_locked(&dev->struct_mutex)); BUG_ON(!mutex_is_locked(&dev->mode_config.mutex)); @@ -782,16 +795,18 @@ static int intel_overlay_do_put_image(struct intel_overlay *overlay, goto out_unpin; if (!overlay->active) { + u32 oconfig; regs = intel_overlay_map_regs(overlay); if (!regs) { ret = -ENOMEM; goto out_unpin; } - regs->OCONFIG = OCONF_CC_OUT_8BIT; + oconfig = OCONF_CC_OUT_8BIT; if (IS_GEN4(overlay->dev)) - regs->OCONFIG |= OCONF_CSC_MODE_BT709; - regs->OCONFIG |= overlay->crtc->pipe == 0 ? + oconfig |= OCONF_CSC_MODE_BT709; + oconfig |= overlay->crtc->pipe == 0 ? OCONF_PIPE_A : OCONF_PIPE_B; + iowrite32(oconfig, ®s->OCONFIG); intel_overlay_unmap_regs(overlay, regs); ret = intel_overlay_on(overlay); @@ -805,42 +820,46 @@ static int intel_overlay_do_put_image(struct intel_overlay *overlay, goto out_unpin; } - regs->DWINPOS = (params->dst_y << 16) | params->dst_x; - regs->DWINSZ = (params->dst_h << 16) | params->dst_w; + iowrite32((params->dst_y << 16) | params->dst_x, ®s->DWINPOS); + iowrite32((params->dst_h << 16) | params->dst_w, ®s->DWINSZ); if (params->format & I915_OVERLAY_YUV_PACKED) tmp_width = packed_width_bytes(params->format, params->src_w); else tmp_width = params->src_w; - regs->SWIDTH = params->src_w; - regs->SWIDTHSW = calc_swidthsw(overlay->dev, - params->offset_Y, tmp_width); - regs->SHEIGHT = params->src_h; - regs->OBUF_0Y = new_bo->gtt_offset + params->offset_Y; - regs->OSTRIDE = params->stride_Y; + swidth = params->src_w; + swidthsw = calc_swidthsw(overlay->dev, params->offset_Y, tmp_width); + sheight = params->src_h; + iowrite32(new_bo->gtt_offset + params->offset_Y, ®s->OBUF_0Y); + ostride = params->stride_Y; if (params->format & I915_OVERLAY_YUV_PLANAR) { int uv_hscale = uv_hsubsampling(params->format); int uv_vscale = uv_vsubsampling(params->format); u32 tmp_U, tmp_V; - regs->SWIDTH |= (params->src_w/uv_hscale) << 16; + swidth |= (params->src_w/uv_hscale) << 16; tmp_U = calc_swidthsw(overlay->dev, params->offset_U, params->src_w/uv_hscale); tmp_V = calc_swidthsw(overlay->dev, params->offset_V, params->src_w/uv_hscale); - regs->SWIDTHSW |= max_t(u32, tmp_U, tmp_V) << 16; - regs->SHEIGHT |= (params->src_h/uv_vscale) << 16; - regs->OBUF_0U = new_bo->gtt_offset + params->offset_U; - regs->OBUF_0V = new_bo->gtt_offset + params->offset_V; - regs->OSTRIDE |= params->stride_UV << 16; + swidthsw |= max_t(u32, tmp_U, tmp_V) << 16; + sheight |= (params->src_h/uv_vscale) << 16; + iowrite32(new_bo->gtt_offset + params->offset_U, ®s->OBUF_0U); + iowrite32(new_bo->gtt_offset + params->offset_V, ®s->OBUF_0V); + ostride |= params->stride_UV << 16; } + iowrite32(swidth, ®s->SWIDTH); + iowrite32(swidthsw, ®s->SWIDTHSW); + iowrite32(sheight, ®s->SHEIGHT); + iowrite32(ostride, ®s->OSTRIDE); + scale_changed = update_scaling_factors(overlay, regs, params); update_colorkey(overlay, regs); - regs->OCMD = overlay_cmd_reg(params); + iowrite32(overlay_cmd_reg(params), ®s->OCMD); intel_overlay_unmap_regs(overlay, regs); @@ -860,7 +879,7 @@ out_unpin: int intel_overlay_switch_off(struct intel_overlay *overlay) { - struct overlay_registers *regs; + struct overlay_registers __iomem *regs; struct drm_device *dev = overlay->dev; int ret; @@ -879,7 +898,7 @@ int intel_overlay_switch_off(struct intel_overlay *overlay) return ret; regs = intel_overlay_map_regs(overlay); - regs->OCMD = 0; + iowrite32(0, ®s->OCMD); intel_overlay_unmap_regs(overlay, regs); ret = intel_overlay_off(overlay); @@ -1109,11 +1128,7 @@ int intel_overlay_put_image(struct drm_device *dev, void *data, struct put_image_params *params; int ret; - if (!dev_priv) { - DRM_ERROR("called with no initialization\n"); - return -EINVAL; - } - + /* No need to check for DRIVER_MODESET - we don't set it up then. */ overlay = dev_priv->overlay; if (!overlay) { DRM_DEBUG("userspace bug: no overlay\n"); @@ -1250,10 +1265,11 @@ out_free: } static void update_reg_attrs(struct intel_overlay *overlay, - struct overlay_registers *regs) + struct overlay_registers __iomem *regs) { - regs->OCLRC0 = (overlay->contrast << 18) | (overlay->brightness & 0xff); - regs->OCLRC1 = overlay->saturation; + iowrite32((overlay->contrast << 18) | (overlay->brightness & 0xff), + ®s->OCLRC0); + iowrite32(overlay->saturation, ®s->OCLRC1); } static bool check_gamma_bounds(u32 gamma1, u32 gamma2) @@ -1306,14 +1322,10 @@ int intel_overlay_attrs(struct drm_device *dev, void *data, struct drm_intel_overlay_attrs *attrs = data; drm_i915_private_t *dev_priv = dev->dev_private; struct intel_overlay *overlay; - struct overlay_registers *regs; + struct overlay_registers __iomem *regs; int ret; - if (!dev_priv) { - DRM_ERROR("called with no initialization\n"); - return -EINVAL; - } - + /* No need to check for DRIVER_MODESET - we don't set it up then. */ overlay = dev_priv->overlay; if (!overlay) { DRM_DEBUG("userspace bug: no overlay\n"); @@ -1396,7 +1408,7 @@ void intel_setup_overlay(struct drm_device *dev) drm_i915_private_t *dev_priv = dev->dev_private; struct intel_overlay *overlay; struct drm_i915_gem_object *reg_bo; - struct overlay_registers *regs; + struct overlay_registers __iomem *regs; int ret; if (!HAS_OVERLAY(dev)) @@ -1451,7 +1463,7 @@ void intel_setup_overlay(struct drm_device *dev) if (!regs) goto out_unpin_bo; - memset(regs, 0, sizeof(struct overlay_registers)); + memset_io(regs, 0, sizeof(struct overlay_registers)); update_polyphase_filter(regs); update_reg_attrs(overlay, regs); @@ -1499,14 +1511,17 @@ struct intel_overlay_error_state { u32 isr; }; -static struct overlay_registers * +static struct overlay_registers __iomem * intel_overlay_map_regs_atomic(struct intel_overlay *overlay) { drm_i915_private_t *dev_priv = overlay->dev->dev_private; - struct overlay_registers *regs; + struct overlay_registers __iomem *regs; if (OVERLAY_NEEDS_PHYSICAL(overlay->dev)) - regs = overlay->reg_bo->phys_obj->handle->vaddr; + /* Cast to make sparse happy, but it's wc memory anyway, so + * equivalent to the wc io mapping on X86. */ + regs = (struct overlay_registers __iomem *) + overlay->reg_bo->phys_obj->handle->vaddr; else regs = io_mapping_map_atomic_wc(dev_priv->mm.gtt_mapping, overlay->reg_bo->gtt_offset); @@ -1515,7 +1530,7 @@ intel_overlay_map_regs_atomic(struct intel_overlay *overlay) } static void intel_overlay_unmap_regs_atomic(struct intel_overlay *overlay, - struct overlay_registers *regs) + struct overlay_registers __iomem *regs) { if (!OVERLAY_NEEDS_PHYSICAL(overlay->dev)) io_mapping_unmap_atomic(regs); @@ -1540,9 +1555,9 @@ intel_overlay_capture_error_state(struct drm_device *dev) error->dovsta = I915_READ(DOVSTA); error->isr = I915_READ(ISR); if (OVERLAY_NEEDS_PHYSICAL(overlay->dev)) - error->base = (long) overlay->reg_bo->phys_obj->handle->vaddr; + error->base = (__force long)overlay->reg_bo->phys_obj->handle->vaddr; else - error->base = (long) overlay->reg_bo->gtt_offset; + error->base = overlay->reg_bo->gtt_offset; regs = intel_overlay_map_regs_atomic(overlay); if (!regs) diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 36940a3..4389234 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -28,6 +28,8 @@ #include <linux/cpufreq.h> #include "i915_drv.h" #include "intel_drv.h" +#include "../../../platform/x86/intel_ips.h" +#include <linux/module.h> /* FBC, or Frame Buffer Compression, is a technique employed to compress the * framebuffer contents in-memory, aiming at reducing the required bandwidth @@ -524,6 +526,113 @@ out_disable: } } +static void i915_pineview_get_mem_freq(struct drm_device *dev) +{ + drm_i915_private_t *dev_priv = dev->dev_private; + u32 tmp; + + tmp = I915_READ(CLKCFG); + + switch (tmp & CLKCFG_FSB_MASK) { + case CLKCFG_FSB_533: + dev_priv->fsb_freq = 533; /* 133*4 */ + break; + case CLKCFG_FSB_800: + dev_priv->fsb_freq = 800; /* 200*4 */ + break; + case CLKCFG_FSB_667: + dev_priv->fsb_freq = 667; /* 167*4 */ + break; + case CLKCFG_FSB_400: + dev_priv->fsb_freq = 400; /* 100*4 */ + break; + } + + switch (tmp & CLKCFG_MEM_MASK) { + case CLKCFG_MEM_533: + dev_priv->mem_freq = 533; + break; + case CLKCFG_MEM_667: + dev_priv->mem_freq = 667; + break; + case CLKCFG_MEM_800: + dev_priv->mem_freq = 800; + break; + } + + /* detect pineview DDR3 setting */ + tmp = I915_READ(CSHRDDR3CTL); + dev_priv->is_ddr3 = (tmp & CSHRDDR3CTL_DDR3) ? 1 : 0; +} + +static void i915_ironlake_get_mem_freq(struct drm_device *dev) +{ + drm_i915_private_t *dev_priv = dev->dev_private; + u16 ddrpll, csipll; + + ddrpll = I915_READ16(DDRMPLL1); + csipll = I915_READ16(CSIPLL0); + + switch (ddrpll & 0xff) { + case 0xc: + dev_priv->mem_freq = 800; + break; + case 0x10: + dev_priv->mem_freq = 1066; + break; + case 0x14: + dev_priv->mem_freq = 1333; + break; + case 0x18: + dev_priv->mem_freq = 1600; + break; + default: + DRM_DEBUG_DRIVER("unknown memory frequency 0x%02x\n", + ddrpll & 0xff); + dev_priv->mem_freq = 0; + break; + } + + dev_priv->r_t = dev_priv->mem_freq; + + switch (csipll & 0x3ff) { + case 0x00c: + dev_priv->fsb_freq = 3200; + break; + case 0x00e: + dev_priv->fsb_freq = 3733; + break; + case 0x010: + dev_priv->fsb_freq = 4266; + break; + case 0x012: + dev_priv->fsb_freq = 4800; + break; + case 0x014: + dev_priv->fsb_freq = 5333; + break; + case 0x016: + dev_priv->fsb_freq = 5866; + break; + case 0x018: + dev_priv->fsb_freq = 6400; + break; + default: + DRM_DEBUG_DRIVER("unknown fsb frequency 0x%04x\n", + csipll & 0x3ff); + dev_priv->fsb_freq = 0; + break; + } + + if (dev_priv->fsb_freq == 3200) { + dev_priv->c_m = 0; + } else if (dev_priv->fsb_freq > 3200 && dev_priv->fsb_freq <= 4800) { + dev_priv->c_m = 1; + } else { + dev_priv->c_m = 2; + } +} + static const struct cxsr_latency cxsr_latency_table[] = { {1, 0, 800, 400, 3382, 33382, 3983, 33983}, /* DDR2-400 SC */ {1, 0, 800, 667, 3354, 33354, 3807, 33807}, /* DDR2-667 SC */ @@ -562,7 +671,7 @@ static const struct cxsr_latency cxsr_latency_table[] = { {0, 1, 400, 800, 6042, 36042, 6584, 36584}, /* DDR3-800 SC */ }; -const struct cxsr_latency *intel_get_cxsr_latency(int is_desktop, +static const struct cxsr_latency *intel_get_cxsr_latency(int is_desktop, int is_ddr3, int fsb, int mem) @@ -2436,6 +2545,7 @@ static int ironlake_setup_rc6(struct drm_device *dev) void ironlake_enable_rc6(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; + struct intel_ring_buffer *ring = &dev_priv->ring[RCS]; int ret; /* rc6 disabled by default due to repeated reports of hanging during @@ -2455,31 +2565,31 @@ void ironlake_enable_rc6(struct drm_device *dev) * GPU can automatically power down the render unit if given a page * to save state. */ - ret = BEGIN_LP_RING(6); + ret = intel_ring_begin(ring, 6); if (ret) { ironlake_teardown_rc6(dev); mutex_unlock(&dev->struct_mutex); return; } - OUT_RING(MI_SUSPEND_FLUSH | MI_SUSPEND_FLUSH_EN); - OUT_RING(MI_SET_CONTEXT); - OUT_RING(dev_priv->renderctx->gtt_offset | - MI_MM_SPACE_GTT | - MI_SAVE_EXT_STATE_EN | - MI_RESTORE_EXT_STATE_EN | - MI_RESTORE_INHIBIT); - OUT_RING(MI_SUSPEND_FLUSH); - OUT_RING(MI_NOOP); - OUT_RING(MI_FLUSH); - ADVANCE_LP_RING(); + intel_ring_emit(ring, MI_SUSPEND_FLUSH | MI_SUSPEND_FLUSH_EN); + intel_ring_emit(ring, MI_SET_CONTEXT); + intel_ring_emit(ring, dev_priv->renderctx->gtt_offset | + MI_MM_SPACE_GTT | + MI_SAVE_EXT_STATE_EN | + MI_RESTORE_EXT_STATE_EN | + MI_RESTORE_INHIBIT); + intel_ring_emit(ring, MI_SUSPEND_FLUSH); + intel_ring_emit(ring, MI_NOOP); + intel_ring_emit(ring, MI_FLUSH); + intel_ring_advance(ring); /* * Wait for the command parser to advance past MI_SET_CONTEXT. The HW * does an implicit flush, combined with MI_FLUSH above, it should be * safe to assume that renderctx is valid */ - ret = intel_wait_ring_idle(LP_RING(dev_priv)); + ret = intel_wait_ring_idle(ring); if (ret) { DRM_ERROR("failed to enable ironlake power power savings\n"); ironlake_teardown_rc6(dev); @@ -2507,6 +2617,485 @@ static unsigned long intel_pxfreq(u32 vidfreq) return freq; } +static const struct cparams { + u16 i; + u16 t; + u16 m; + u16 c; +} cparams[] = { + { 1, 1333, 301, 28664 }, + { 1, 1066, 294, 24460 }, + { 1, 800, 294, 25192 }, + { 0, 1333, 276, 27605 }, + { 0, 1066, 276, 27605 }, + { 0, 800, 231, 23784 }, +}; + +unsigned long i915_chipset_val(struct drm_i915_private *dev_priv) +{ + u64 total_count, diff, ret; + u32 count1, count2, count3, m = 0, c = 0; + unsigned long now = jiffies_to_msecs(jiffies), diff1; + int i; + + diff1 = now - dev_priv->last_time1; + + /* Prevent division-by-zero if we are asking too fast. + * Also, we don't get interesting results if we are polling + * faster than once in 10ms, so just return the saved value + * in such cases. + */ + if (diff1 <= 10) + return dev_priv->chipset_power; + + count1 = I915_READ(DMIEC); + count2 = I915_READ(DDREC); + count3 = I915_READ(CSIEC); + + total_count = count1 + count2 + count3; + + /* FIXME: handle per-counter overflow */ + if (total_count < dev_priv->last_count1) { + diff = ~0UL - dev_priv->last_count1; + diff += total_count; + } else { + diff = total_count - dev_priv->last_count1; + } + + for (i = 0; i < ARRAY_SIZE(cparams); i++) { + if (cparams[i].i == dev_priv->c_m && + cparams[i].t == dev_priv->r_t) { + m = cparams[i].m; + c = cparams[i].c; + break; + } + } + + diff = div_u64(diff, diff1); + ret = ((m * diff) + c); + ret = div_u64(ret, 10); + + dev_priv->last_count1 = total_count; + dev_priv->last_time1 = now; + + dev_priv->chipset_power = ret; + + return ret; +} + +unsigned long i915_mch_val(struct drm_i915_private *dev_priv) +{ + unsigned long m, x, b; + u32 tsfs; + + tsfs = I915_READ(TSFS); + + m = ((tsfs & TSFS_SLOPE_MASK) >> TSFS_SLOPE_SHIFT); + x = I915_READ8(TR1); + + b = tsfs & TSFS_INTR_MASK; + + return ((m * x) / 127) - b; +} + +static u16 pvid_to_extvid(struct drm_i915_private *dev_priv, u8 pxvid) +{ + static const struct v_table { + u16 vd; /* in .1 mil */ + u16 vm; /* in .1 mil */ + } v_table[] = { + { 0, 0, }, + { 375, 0, }, + { 500, 0, }, + { 625, 0, }, + { 750, 0, }, + { 875, 0, }, + { 1000, 0, }, + { 1125, 0, }, + { 4125, 3000, }, + { 4125, 3000, }, + { 4125, 3000, }, + { 4125, 3000, }, + { 4125, 3000, }, + { 4125, 3000, }, + { 4125, 3000, }, + { 4125, 3000, }, + { 4125, 3000, }, + { 4125, 3000, }, + { 4125, 3000, }, + { 4125, 3000, }, + { 4125, 3000, }, + { 4125, 3000, }, + { 4125, 3000, }, + { 4125, 3000, }, + { 4125, 3000, }, + { 4125, 3000, }, + { 4125, 3000, }, + { 4125, 3000, }, + { 4125, 3000, }, + { 4125, 3000, }, + { 4125, 3000, }, + { 4125, 3000, }, + { 4250, 3125, }, + { 4375, 3250, }, + { 4500, 3375, }, + { 4625, 3500, }, + { 4750, 3625, }, + { 4875, 3750, }, + { 5000, 3875, }, + { 5125, 4000, }, + { 5250, 4125, }, + { 5375, 4250, }, + { 5500, 4375, }, + { 5625, 4500, }, + { 5750, 4625, }, + { 5875, 4750, }, + { 6000, 4875, }, + { 6125, 5000, }, + { 6250, 5125, }, + { 6375, 5250, }, + { 6500, 5375, }, + { 6625, 5500, }, + { 6750, 5625, }, + { 6875, 5750, }, + { 7000, 5875, }, + { 7125, 6000, }, + { 7250, 6125, }, + { 7375, 6250, }, + { 7500, 6375, }, + { 7625, 6500, }, + { 7750, 6625, }, + { 7875, 6750, }, + { 8000, 6875, }, + { 8125, 7000, }, + { 8250, 7125, }, + { 8375, 7250, }, + { 8500, 7375, }, + { 8625, 7500, }, + { 8750, 7625, }, + { 8875, 7750, }, + { 9000, 7875, }, + { 9125, 8000, }, + { 9250, 8125, }, + { 9375, 8250, }, + { 9500, 8375, }, + { 9625, 8500, }, + { 9750, 8625, }, + { 9875, 8750, }, + { 10000, 8875, }, + { 10125, 9000, }, + { 10250, 9125, }, + { 10375, 9250, }, + { 10500, 9375, }, + { 10625, 9500, }, + { 10750, 9625, }, + { 10875, 9750, }, + { 11000, 9875, }, + { 11125, 10000, }, + { 11250, 10125, }, + { 11375, 10250, }, + { 11500, 10375, }, + { 11625, 10500, }, + { 11750, 10625, }, + { 11875, 10750, }, + { 12000, 10875, }, + { 12125, 11000, }, + { 12250, 11125, }, + { 12375, 11250, }, + { 12500, 11375, }, + { 12625, 11500, }, + { 12750, 11625, }, + { 12875, 11750, }, + { 13000, 11875, }, + { 13125, 12000, }, + { 13250, 12125, }, + { 13375, 12250, }, + { 13500, 12375, }, + { 13625, 12500, }, + { 13750, 12625, }, + { 13875, 12750, }, + { 14000, 12875, }, + { 14125, 13000, }, + { 14250, 13125, }, + { 14375, 13250, }, + { 14500, 13375, }, + { 14625, 13500, }, + { 14750, 13625, }, + { 14875, 13750, }, + { 15000, 13875, }, + { 15125, 14000, }, + { 15250, 14125, }, + { 15375, 14250, }, + { 15500, 14375, }, + { 15625, 14500, }, + { 15750, 14625, }, + { 15875, 14750, }, + { 16000, 14875, }, + { 16125, 15000, }, + }; + if (dev_priv->info->is_mobile) + return v_table[pxvid].vm; + else + return v_table[pxvid].vd; +} + +void i915_update_gfx_val(struct drm_i915_private *dev_priv) +{ + struct timespec now, diff1; + u64 diff; + unsigned long diffms; + u32 count; + + if (dev_priv->info->gen != 5) + return; + + getrawmonotonic(&now); + diff1 = timespec_sub(now, dev_priv->last_time2); + + /* Don't divide by 0 */ + diffms = diff1.tv_sec * 1000 + diff1.tv_nsec / 1000000; + if (!diffms) + return; + + count = I915_READ(GFXEC); + + if (count < dev_priv->last_count2) { + diff = ~0UL - dev_priv->last_count2; + diff += count; + } else { + diff = count - dev_priv->last_count2; + } + + dev_priv->last_count2 = count; + dev_priv->last_time2 = now; + + /* More magic constants... */ + diff = diff * 1181; + diff = div_u64(diff, diffms * 10); + dev_priv->gfx_power = diff; +} + +unsigned long i915_gfx_val(struct drm_i915_private *dev_priv) +{ + unsigned long t, corr, state1, corr2, state2; + u32 pxvid, ext_v; + + pxvid = I915_READ(PXVFREQ_BASE + (dev_priv->cur_delay * 4)); + pxvid = (pxvid >> 24) & 0x7f; + ext_v = pvid_to_extvid(dev_priv, pxvid); + + state1 = ext_v; + + t = i915_mch_val(dev_priv); + + /* Revel in the empirically derived constants */ + + /* Correction factor in 1/100000 units */ + if (t > 80) + corr = ((t * 2349) + 135940); + else if (t >= 50) + corr = ((t * 964) + 29317); + else /* < 50 */ + corr = ((t * 301) + 1004); + + corr = corr * ((150142 * state1) / 10000 - 78642); + corr /= 100000; + corr2 = (corr * dev_priv->corr); + + state2 = (corr2 * state1) / 10000; + state2 /= 100; /* convert to mW */ + + i915_update_gfx_val(dev_priv); + + return dev_priv->gfx_power + state2; +} + +/* Global for IPS driver to get at the current i915 device */ +static struct drm_i915_private *i915_mch_dev; +/* + * Lock protecting IPS related data structures + * - i915_mch_dev + * - dev_priv->max_delay + * - dev_priv->min_delay + * - dev_priv->fmax + * - dev_priv->gpu_busy + */ +static DEFINE_SPINLOCK(mchdev_lock); + +/** + * i915_read_mch_val - return value for IPS use + * + * Calculate and return a value for the IPS driver to use when deciding whether + * we have thermal and power headroom to increase CPU or GPU power budget. + */ +unsigned long i915_read_mch_val(void) +{ + struct drm_i915_private *dev_priv; + unsigned long chipset_val, graphics_val, ret = 0; + + spin_lock(&mchdev_lock); + if (!i915_mch_dev) + goto out_unlock; + dev_priv = i915_mch_dev; + + chipset_val = i915_chipset_val(dev_priv); + graphics_val = i915_gfx_val(dev_priv); + + ret = chipset_val + graphics_val; + +out_unlock: + spin_unlock(&mchdev_lock); + + return ret; +} +EXPORT_SYMBOL_GPL(i915_read_mch_val); + +/** + * i915_gpu_raise - raise GPU frequency limit + * + * Raise the limit; IPS indicates we have thermal headroom. + */ +bool i915_gpu_raise(void) +{ + struct drm_i915_private *dev_priv; + bool ret = true; + + spin_lock(&mchdev_lock); + if (!i915_mch_dev) { + ret = false; + goto out_unlock; + } + dev_priv = i915_mch_dev; + + if (dev_priv->max_delay > dev_priv->fmax) + dev_priv->max_delay--; + +out_unlock: + spin_unlock(&mchdev_lock); + + return ret; +} +EXPORT_SYMBOL_GPL(i915_gpu_raise); + +/** + * i915_gpu_lower - lower GPU frequency limit + * + * IPS indicates we're close to a thermal limit, so throttle back the GPU + * frequency maximum. + */ +bool i915_gpu_lower(void) +{ + struct drm_i915_private *dev_priv; + bool ret = true; + + spin_lock(&mchdev_lock); + if (!i915_mch_dev) { + ret = false; + goto out_unlock; + } + dev_priv = i915_mch_dev; + + if (dev_priv->max_delay < dev_priv->min_delay) + dev_priv->max_delay++; + +out_unlock: + spin_unlock(&mchdev_lock); + + return ret; +} +EXPORT_SYMBOL_GPL(i915_gpu_lower); + +/** + * i915_gpu_busy - indicate GPU business to IPS + * + * Tell the IPS driver whether or not the GPU is busy. + */ +bool i915_gpu_busy(void) +{ + struct drm_i915_private *dev_priv; + bool ret = false; + + spin_lock(&mchdev_lock); + if (!i915_mch_dev) + goto out_unlock; + dev_priv = i915_mch_dev; + + ret = dev_priv->busy; + +out_unlock: + spin_unlock(&mchdev_lock); + + return ret; +} +EXPORT_SYMBOL_GPL(i915_gpu_busy); + +/** + * i915_gpu_turbo_disable - disable graphics turbo + * + * Disable graphics turbo by resetting the max frequency and setting the + * current frequency to the default. + */ +bool i915_gpu_turbo_disable(void) +{ + struct drm_i915_private *dev_priv; + bool ret = true; + + spin_lock(&mchdev_lock); + if (!i915_mch_dev) { + ret = false; + goto out_unlock; + } + dev_priv = i915_mch_dev; + + dev_priv->max_delay = dev_priv->fstart; + + if (!ironlake_set_drps(dev_priv->dev, dev_priv->fstart)) + ret = false; + +out_unlock: + spin_unlock(&mchdev_lock); + + return ret; +} +EXPORT_SYMBOL_GPL(i915_gpu_turbo_disable); + +/** + * Tells the intel_ips driver that the i915 driver is now loaded, if + * IPS got loaded first. + * + * This awkward dance is so that neither module has to depend on the + * other in order for IPS to do the appropriate communication of + * GPU turbo limits to i915. + */ +static void +ips_ping_for_i915_load(void) +{ + void (*link)(void); + + link = symbol_get(ips_link_to_i915_driver); + if (link) { + link(); + symbol_put(ips_link_to_i915_driver); + } +} + +void intel_gpu_ips_init(struct drm_i915_private *dev_priv) +{ + spin_lock(&mchdev_lock); + i915_mch_dev = dev_priv; + dev_priv->mchdev_lock = &mchdev_lock; + spin_unlock(&mchdev_lock); + + ips_ping_for_i915_load(); +} + +void intel_gpu_ips_teardown(void) +{ + spin_lock(&mchdev_lock); + i915_mch_dev = NULL; + spin_unlock(&mchdev_lock); +} + void intel_init_emon(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; @@ -2663,9 +3252,8 @@ static void gen6_init_clock_gating(struct drm_device *dev) I915_WRITE(WM2_LP_ILK, 0); I915_WRITE(WM1_LP_ILK, 0); - /* clear masked bit */ I915_WRITE(CACHE_MODE_0, - CM0_STC_EVICT_DISABLE_LRA_SNB << CM0_MASK_SHIFT); + _MASKED_BIT_DISABLE(CM0_STC_EVICT_DISABLE_LRA_SNB)); I915_WRITE(GEN6_UCGCTL1, I915_READ(GEN6_UCGCTL1) | @@ -2776,6 +3364,10 @@ static void ivybridge_init_clock_gating(struct drm_device *dev) } gen7_setup_fixed_func_scheduler(dev_priv); + + /* WaDisable4x2SubspanOptimization */ + I915_WRITE(CACHE_MODE_1, + _MASKED_BIT_ENABLE(PIXEL_SUBSPAN_COLLECT_OPT_DISABLE)); } static void valleyview_init_clock_gating(struct drm_device *dev) @@ -2821,9 +3413,8 @@ static void valleyview_init_clock_gating(struct drm_device *dev) intel_flush_display_plane(dev_priv, pipe); } - I915_WRITE(CACHE_MODE_1, I915_READ(CACHE_MODE_1) | - (PIXEL_SUBSPAN_COLLECT_OPT_DISABLE << 16) | - PIXEL_SUBSPAN_COLLECT_OPT_DISABLE); + I915_WRITE(CACHE_MODE_1, + _MASKED_BIT_ENABLE(PIXEL_SUBSPAN_COLLECT_OPT_DISABLE)); } static void g4x_init_clock_gating(struct drm_device *dev) @@ -2875,6 +3466,9 @@ static void gen3_init_clock_gating(struct drm_device *dev) dstate |= DSTATE_PLL_D3_OFF | DSTATE_GFX_CLOCK_GATING | DSTATE_DOT_CLOCK_GATING; I915_WRITE(D_STATE, dstate); + + if (IS_PINEVIEW(dev)) + I915_WRITE(ECOSKPD, _MASKED_BIT_ENABLE(ECO_GATING_CX_ONLY)); } static void i85x_init_clock_gating(struct drm_device *dev) @@ -2931,6 +3525,41 @@ void intel_init_clock_gating(struct drm_device *dev) dev_priv->display.init_pch_clock_gating(dev); } +static void gen6_sanitize_pm(struct drm_device *dev) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + u32 limits, delay, old; + + gen6_gt_force_wake_get(dev_priv); + + old = limits = I915_READ(GEN6_RP_INTERRUPT_LIMITS); + /* Make sure we continue to get interrupts + * until we hit the minimum or maximum frequencies. + */ + limits &= ~(0x3f << 16 | 0x3f << 24); + delay = dev_priv->cur_delay; + if (delay < dev_priv->max_delay) + limits |= (dev_priv->max_delay & 0x3f) << 24; + if (delay > dev_priv->min_delay) + limits |= (dev_priv->min_delay & 0x3f) << 16; + + if (old != limits) { + DRM_ERROR("Power management discrepancy: GEN6_RP_INTERRUPT_LIMITS expected %08x, was %08x\n", + limits, old); + I915_WRITE(GEN6_RP_INTERRUPT_LIMITS, limits); + } + + gen6_gt_force_wake_put(dev_priv); +} + +void intel_sanitize_pm(struct drm_device *dev) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + + if (dev_priv->display.sanitize_pm) + dev_priv->display.sanitize_pm(dev); +} + /* Set up chip specific power management-related functions */ void intel_init_pm(struct drm_device *dev) { @@ -2953,6 +3582,12 @@ void intel_init_pm(struct drm_device *dev) /* 855GM needs testing */ } + /* For cxsr */ + if (IS_PINEVIEW(dev)) + i915_pineview_get_mem_freq(dev); + else if (IS_GEN5(dev)) + i915_ironlake_get_mem_freq(dev); + /* For FIFO watermark updates */ if (HAS_PCH_SPLIT(dev)) { dev_priv->display.force_wake_get = __gen6_gt_force_wake_get; @@ -3007,6 +3642,7 @@ void intel_init_pm(struct drm_device *dev) dev_priv->display.update_wm = NULL; } dev_priv->display.init_clock_gating = gen6_init_clock_gating; + dev_priv->display.sanitize_pm = gen6_sanitize_pm; } else if (IS_IVYBRIDGE(dev)) { /* FIXME: detect B0+ stepping and use auto training */ if (SNB_READ_WM0_LATENCY()) { @@ -3018,6 +3654,7 @@ void intel_init_pm(struct drm_device *dev) dev_priv->display.update_wm = NULL; } dev_priv->display.init_clock_gating = ivybridge_init_clock_gating; + dev_priv->display.sanitize_pm = gen6_sanitize_pm; } else dev_priv->display.update_wm = NULL; } else if (IS_VALLEYVIEW(dev)) { diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index b5ef7c1..b59b6d5 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -401,12 +401,11 @@ static int init_render_ring(struct intel_ring_buffer *ring) int ret = init_ring_common(ring); if (INTEL_INFO(dev)->gen > 3) { - int mode = VS_TIMER_DISPATCH << 16 | VS_TIMER_DISPATCH; - I915_WRITE(MI_MODE, mode); + I915_WRITE(MI_MODE, _MASKED_BIT_ENABLE(VS_TIMER_DISPATCH)); if (IS_GEN7(dev)) I915_WRITE(GFX_MODE_GEN7, - GFX_MODE_DISABLE(GFX_TLB_INVALIDATE_ALWAYS) | - GFX_MODE_ENABLE(GFX_REPLAY_MODE)); + _MASKED_BIT_DISABLE(GFX_TLB_INVALIDATE_ALWAYS) | + _MASKED_BIT_ENABLE(GFX_REPLAY_MODE)); } if (INTEL_INFO(dev)->gen >= 5) { @@ -415,19 +414,19 @@ static int init_render_ring(struct intel_ring_buffer *ring) return ret; } - if (INTEL_INFO(dev)->gen >= 6) { - I915_WRITE(INSTPM, - INSTPM_FORCE_ORDERING << 16 | INSTPM_FORCE_ORDERING); - + if (IS_GEN6(dev)) { /* From the Sandybridge PRM, volume 1 part 3, page 24: * "If this bit is set, STCunit will have LRA as replacement * policy. [...] This bit must be reset. LRA replacement * policy is not supported." */ I915_WRITE(CACHE_MODE_0, - CM0_STC_EVICT_DISABLE_LRA_SNB << CM0_MASK_SHIFT); + _MASKED_BIT_DISABLE(CM0_STC_EVICT_DISABLE_LRA_SNB)); } + if (INTEL_INFO(dev)->gen >= 6) + I915_WRITE(INSTPM, _MASKED_BIT_ENABLE(INSTPM_FORCE_ORDERING)); + return ret; } @@ -621,17 +620,18 @@ gen5_ring_get_irq(struct intel_ring_buffer *ring) { struct drm_device *dev = ring->dev; drm_i915_private_t *dev_priv = dev->dev_private; + unsigned long flags; if (!dev->irq_enabled) return false; - spin_lock(&ring->irq_lock); + spin_lock_irqsave(&dev_priv->irq_lock, flags); if (ring->irq_refcount++ == 0) { dev_priv->gt_irq_mask &= ~ring->irq_enable_mask; I915_WRITE(GTIMR, dev_priv->gt_irq_mask); POSTING_READ(GTIMR); } - spin_unlock(&ring->irq_lock); + spin_unlock_irqrestore(&dev_priv->irq_lock, flags); return true; } @@ -641,14 +641,15 @@ gen5_ring_put_irq(struct intel_ring_buffer *ring) { struct drm_device *dev = ring->dev; drm_i915_private_t *dev_priv = dev->dev_private; + unsigned long flags; - spin_lock(&ring->irq_lock); + spin_lock_irqsave(&dev_priv->irq_lock, flags); if (--ring->irq_refcount == 0) { dev_priv->gt_irq_mask |= ring->irq_enable_mask; I915_WRITE(GTIMR, dev_priv->gt_irq_mask); POSTING_READ(GTIMR); } - spin_unlock(&ring->irq_lock); + spin_unlock_irqrestore(&dev_priv->irq_lock, flags); } static bool @@ -656,17 +657,18 @@ i9xx_ring_get_irq(struct intel_ring_buffer *ring) { struct drm_device *dev = ring->dev; drm_i915_private_t *dev_priv = dev->dev_private; + unsigned long flags; if (!dev->irq_enabled) return false; - spin_lock(&ring->irq_lock); + spin_lock_irqsave(&dev_priv->irq_lock, flags); if (ring->irq_refcount++ == 0) { dev_priv->irq_mask &= ~ring->irq_enable_mask; I915_WRITE(IMR, dev_priv->irq_mask); POSTING_READ(IMR); } - spin_unlock(&ring->irq_lock); + spin_unlock_irqrestore(&dev_priv->irq_lock, flags); return true; } @@ -676,14 +678,52 @@ i9xx_ring_put_irq(struct intel_ring_buffer *ring) { struct drm_device *dev = ring->dev; drm_i915_private_t *dev_priv = dev->dev_private; + unsigned long flags; - spin_lock(&ring->irq_lock); + spin_lock_irqsave(&dev_priv->irq_lock, flags); if (--ring->irq_refcount == 0) { dev_priv->irq_mask |= ring->irq_enable_mask; I915_WRITE(IMR, dev_priv->irq_mask); POSTING_READ(IMR); } - spin_unlock(&ring->irq_lock); + spin_unlock_irqrestore(&dev_priv->irq_lock, flags); +} + +static bool +i8xx_ring_get_irq(struct intel_ring_buffer *ring) +{ + struct drm_device *dev = ring->dev; + drm_i915_private_t *dev_priv = dev->dev_private; + unsigned long flags; + + if (!dev->irq_enabled) + return false; + + spin_lock_irqsave(&dev_priv->irq_lock, flags); + if (ring->irq_refcount++ == 0) { + dev_priv->irq_mask &= ~ring->irq_enable_mask; + I915_WRITE16(IMR, dev_priv->irq_mask); + POSTING_READ16(IMR); + } + spin_unlock_irqrestore(&dev_priv->irq_lock, flags); + + return true; +} + +static void +i8xx_ring_put_irq(struct intel_ring_buffer *ring) +{ + struct drm_device *dev = ring->dev; + drm_i915_private_t *dev_priv = dev->dev_private; + unsigned long flags; + + spin_lock_irqsave(&dev_priv->irq_lock, flags); + if (--ring->irq_refcount == 0) { + dev_priv->irq_mask |= ring->irq_enable_mask; + I915_WRITE16(IMR, dev_priv->irq_mask); + POSTING_READ16(IMR); + } + spin_unlock_irqrestore(&dev_priv->irq_lock, flags); } void intel_ring_setup_status_page(struct intel_ring_buffer *ring) @@ -762,6 +802,7 @@ gen6_ring_get_irq(struct intel_ring_buffer *ring) { struct drm_device *dev = ring->dev; drm_i915_private_t *dev_priv = dev->dev_private; + unsigned long flags; if (!dev->irq_enabled) return false; @@ -771,14 +812,14 @@ gen6_ring_get_irq(struct intel_ring_buffer *ring) * blt/bsd rings on ivb. */ gen6_gt_force_wake_get(dev_priv); - spin_lock(&ring->irq_lock); + spin_lock_irqsave(&dev_priv->irq_lock, flags); if (ring->irq_refcount++ == 0) { I915_WRITE_IMR(ring, ~ring->irq_enable_mask); dev_priv->gt_irq_mask &= ~ring->irq_enable_mask; I915_WRITE(GTIMR, dev_priv->gt_irq_mask); POSTING_READ(GTIMR); } - spin_unlock(&ring->irq_lock); + spin_unlock_irqrestore(&dev_priv->irq_lock, flags); return true; } @@ -788,15 +829,16 @@ gen6_ring_put_irq(struct intel_ring_buffer *ring) { struct drm_device *dev = ring->dev; drm_i915_private_t *dev_priv = dev->dev_private; + unsigned long flags; - spin_lock(&ring->irq_lock); + spin_lock_irqsave(&dev_priv->irq_lock, flags); if (--ring->irq_refcount == 0) { I915_WRITE_IMR(ring, ~0); dev_priv->gt_irq_mask |= ring->irq_enable_mask; I915_WRITE(GTIMR, dev_priv->gt_irq_mask); POSTING_READ(GTIMR); } - spin_unlock(&ring->irq_lock); + spin_unlock_irqrestore(&dev_priv->irq_lock, flags); gen6_gt_force_wake_put(dev_priv); } @@ -858,7 +900,6 @@ i915_dispatch_execbuffer(struct intel_ring_buffer *ring, static void cleanup_status_page(struct intel_ring_buffer *ring) { - drm_i915_private_t *dev_priv = ring->dev->dev_private; struct drm_i915_gem_object *obj; obj = ring->status_page.obj; @@ -869,14 +910,11 @@ static void cleanup_status_page(struct intel_ring_buffer *ring) i915_gem_object_unpin(obj); drm_gem_object_unreference(&obj->base); ring->status_page.obj = NULL; - - memset(&dev_priv->hws_map, 0, sizeof(dev_priv->hws_map)); } static int init_status_page(struct intel_ring_buffer *ring) { struct drm_device *dev = ring->dev; - drm_i915_private_t *dev_priv = dev->dev_private; struct drm_i915_gem_object *obj; int ret; @@ -897,7 +935,6 @@ static int init_status_page(struct intel_ring_buffer *ring) ring->status_page.gfx_addr = obj->gtt_offset; ring->status_page.page_addr = kmap(obj->pages[0]); if (ring->status_page.page_addr == NULL) { - memset(&dev_priv->hws_map, 0, sizeof(dev_priv->hws_map)); goto err_unpin; } ring->status_page.obj = obj; @@ -930,7 +967,6 @@ static int intel_init_ring_buffer(struct drm_device *dev, ring->size = 32 * PAGE_SIZE; init_waitqueue_head(&ring->irq_queue); - spin_lock_init(&ring->irq_lock); if (I915_NEED_GFX_HWS(dev)) { ret = init_status_page(ring); @@ -951,20 +987,14 @@ static int intel_init_ring_buffer(struct drm_device *dev, if (ret) goto err_unref; - ring->map.size = ring->size; - ring->map.offset = dev->agp->base + obj->gtt_offset; - ring->map.type = 0; - ring->map.flags = 0; - ring->map.mtrr = 0; - - drm_core_ioremap_wc(&ring->map, dev); - if (ring->map.handle == NULL) { + ring->virtual_start = ioremap_wc(dev->agp->base + obj->gtt_offset, + ring->size); + if (ring->virtual_start == NULL) { DRM_ERROR("Failed to map ringbuffer.\n"); ret = -EINVAL; goto err_unpin; } - ring->virtual_start = ring->map.handle; ret = ring->init(ring); if (ret) goto err_unmap; @@ -980,7 +1010,7 @@ static int intel_init_ring_buffer(struct drm_device *dev, return 0; err_unmap: - drm_core_ioremapfree(&ring->map, dev); + iounmap(ring->virtual_start); err_unpin: i915_gem_object_unpin(obj); err_unref: @@ -1008,7 +1038,7 @@ void intel_cleanup_ring_buffer(struct intel_ring_buffer *ring) I915_WRITE_CTL(ring, 0); - drm_core_ioremapfree(&ring->map, ring->dev); + iounmap(ring->virtual_start); i915_gem_object_unpin(ring->obj); drm_gem_object_unreference(&ring->obj->base); @@ -1022,7 +1052,7 @@ void intel_cleanup_ring_buffer(struct intel_ring_buffer *ring) static int intel_wrap_ring_buffer(struct intel_ring_buffer *ring) { - unsigned int *virt; + uint32_t __iomem *virt; int rem = ring->size - ring->tail; if (ring->space < rem) { @@ -1031,12 +1061,10 @@ static int intel_wrap_ring_buffer(struct intel_ring_buffer *ring) return ret; } - virt = (unsigned int *)(ring->virtual_start + ring->tail); - rem /= 8; - while (rem--) { - *virt++ = MI_NOOP; - *virt++ = MI_NOOP; - } + virt = ring->virtual_start + ring->tail; + rem /= 4; + while (rem--) + iowrite32(MI_NOOP, virt++); ring->tail = 0; ring->space = ring_space(ring); @@ -1057,9 +1085,11 @@ static int intel_ring_wait_seqno(struct intel_ring_buffer *ring, u32 seqno) was_interruptible = dev_priv->mm.interruptible; dev_priv->mm.interruptible = false; - ret = i915_wait_request(ring, seqno, true); + ret = i915_wait_request(ring, seqno); dev_priv->mm.interruptible = was_interruptible; + if (!ret) + i915_gem_retire_requests_ring(ring); return ret; } @@ -1133,15 +1163,12 @@ int intel_wait_ring_buffer(struct intel_ring_buffer *ring, int n) return ret; trace_i915_ring_wait_begin(ring); - if (drm_core_check_feature(dev, DRIVER_GEM)) - /* With GEM the hangcheck timer should kick us out of the loop, - * leaving it early runs the risk of corrupting GEM state (due - * to running on almost untested codepaths). But on resume - * timers don't work yet, so prevent a complete hang in that - * case by choosing an insanely large timeout. */ - end = jiffies + 60 * HZ; - else - end = jiffies + 3 * HZ; + /* With GEM the hangcheck timer should kick us out of the loop, + * leaving it early runs the risk of corrupting GEM state (due + * to running on almost untested codepaths). But on resume + * timers don't work yet, so prevent a complete hang in that + * case by choosing an insanely large timeout. */ + end = jiffies + 60 * HZ; do { ring->head = I915_READ_HEAD(ring); @@ -1193,7 +1220,11 @@ int intel_ring_begin(struct intel_ring_buffer *ring, void intel_ring_advance(struct intel_ring_buffer *ring) { + struct drm_i915_private *dev_priv = ring->dev->dev_private; + ring->tail &= ring->size - 1; + if (dev_priv->stop_rings & intel_ring_flag(ring)) + return; ring->write_tail(ring, ring->tail); } @@ -1318,8 +1349,13 @@ int intel_init_render_ring_buffer(struct drm_device *dev) else ring->flush = gen4_render_ring_flush; ring->get_seqno = ring_get_seqno; - ring->irq_get = i9xx_ring_get_irq; - ring->irq_put = i9xx_ring_put_irq; + if (IS_GEN2(dev)) { + ring->irq_get = i8xx_ring_get_irq; + ring->irq_put = i8xx_ring_put_irq; + } else { + ring->irq_get = i9xx_ring_get_irq; + ring->irq_put = i9xx_ring_put_irq; + } ring->irq_enable_mask = I915_USER_INTERRUPT; } ring->write_tail = ring_write_tail; @@ -1366,8 +1402,13 @@ int intel_render_ring_init_dri(struct drm_device *dev, u64 start, u32 size) else ring->flush = gen4_render_ring_flush; ring->get_seqno = ring_get_seqno; - ring->irq_get = i9xx_ring_get_irq; - ring->irq_put = i9xx_ring_put_irq; + if (IS_GEN2(dev)) { + ring->irq_get = i8xx_ring_get_irq; + ring->irq_put = i8xx_ring_put_irq; + } else { + ring->irq_get = i9xx_ring_get_irq; + ring->irq_put = i9xx_ring_put_irq; + } ring->irq_enable_mask = I915_USER_INTERRUPT; ring->write_tail = ring_write_tail; if (INTEL_INFO(dev)->gen >= 4) @@ -1392,20 +1433,13 @@ int intel_render_ring_init_dri(struct drm_device *dev, u64 start, u32 size) if (IS_I830(ring->dev)) ring->effective_size -= 128; - ring->map.offset = start; - ring->map.size = size; - ring->map.type = 0; - ring->map.flags = 0; - ring->map.mtrr = 0; - - drm_core_ioremap_wc(&ring->map, dev); - if (ring->map.handle == NULL) { + ring->virtual_start = ioremap_wc(start, size); + if (ring->virtual_start == NULL) { DRM_ERROR("can not ioremap virtual address for" " ring buffer\n"); return -ENOMEM; } - ring->virtual_start = (void __force __iomem *)ring->map.handle; return 0; } diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h index 06a66ad..baba757 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.h +++ b/drivers/gpu/drm/i915/intel_ringbuffer.h @@ -2,7 +2,7 @@ #define _INTEL_RINGBUFFER_H_ struct intel_hw_status_page { - u32 __iomem *page_addr; + u32 *page_addr; unsigned int gfx_addr; struct drm_i915_gem_object *obj; }; @@ -56,12 +56,9 @@ struct intel_ring_buffer { */ u32 last_retired_head; - spinlock_t irq_lock; - u32 irq_refcount; + u32 irq_refcount; /* protected by dev_priv->irq_lock */ u32 irq_enable_mask; /* bitmask to enable ring interrupt */ - u32 irq_seqno; /* last seq seem at irq time */ u32 trace_irq_seqno; - u32 waiting_seqno; u32 sync_seqno[I915_NUM_RINGS-1]; bool __must_check (*irq_get)(struct intel_ring_buffer *ring); void (*irq_put)(struct intel_ring_buffer *ring); @@ -118,7 +115,6 @@ struct intel_ring_buffer { u32 outstanding_lazy_request; wait_queue_head_t irq_queue; - drm_local_map_t map; void *private; }; @@ -152,7 +148,9 @@ static inline u32 intel_read_status_page(struct intel_ring_buffer *ring, int reg) { - return ioread32(ring->status_page.page_addr + reg); + /* Ensure that the compiler doesn't optimize away the load. */ + barrier(); + return ring->status_page.page_addr[reg]; } /** @@ -170,10 +168,7 @@ intel_read_status_page(struct intel_ring_buffer *ring, * * The area from dword 0x20 to 0x3ff is available for driver usage. */ -#define READ_HWSP(dev_priv, reg) intel_read_status_page(LP_RING(dev_priv), reg) -#define READ_BREADCRUMB(dev_priv) READ_HWSP(dev_priv, I915_BREADCRUMB_INDEX) #define I915_GEM_HWS_INDEX 0x20 -#define I915_BREADCRUMB_INDEX 0x21 void intel_cleanup_ring_buffer(struct intel_ring_buffer *ring); diff --git a/drivers/gpu/drm/i915/intel_sdvo.c b/drivers/gpu/drm/i915/intel_sdvo.c index 3d9dfa5..7d3f238 100644 --- a/drivers/gpu/drm/i915/intel_sdvo.c +++ b/drivers/gpu/drm/i915/intel_sdvo.c @@ -747,18 +747,18 @@ static void intel_sdvo_get_dtd_from_mode(struct intel_sdvo_dtd *dtd, uint16_t h_sync_offset, v_sync_offset; int mode_clock; - width = mode->crtc_hdisplay; - height = mode->crtc_vdisplay; + width = mode->hdisplay; + height = mode->vdisplay; /* do some mode translations */ - h_blank_len = mode->crtc_hblank_end - mode->crtc_hblank_start; - h_sync_len = mode->crtc_hsync_end - mode->crtc_hsync_start; + h_blank_len = mode->htotal - mode->hdisplay; + h_sync_len = mode->hsync_end - mode->hsync_start; - v_blank_len = mode->crtc_vblank_end - mode->crtc_vblank_start; - v_sync_len = mode->crtc_vsync_end - mode->crtc_vsync_start; + v_blank_len = mode->vtotal - mode->vdisplay; + v_sync_len = mode->vsync_end - mode->vsync_start; - h_sync_offset = mode->crtc_hsync_start - mode->crtc_hblank_start; - v_sync_offset = mode->crtc_vsync_start - mode->crtc_vblank_start; + h_sync_offset = mode->hsync_start - mode->hdisplay; + v_sync_offset = mode->vsync_start - mode->vdisplay; mode_clock = mode->clock; mode_clock /= intel_mode_get_pixel_multiplier(mode) ?: 1; @@ -1578,9 +1578,6 @@ end: intel_sdvo->sdvo_lvds_fixed_mode = drm_mode_duplicate(connector->dev, newmode); - drm_mode_set_crtcinfo(intel_sdvo->sdvo_lvds_fixed_mode, - 0); - intel_sdvo->is_lvds = true; break; } diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c index fbf03b9..2a20fb0 100644 --- a/drivers/gpu/drm/i915/intel_sprite.c +++ b/drivers/gpu/drm/i915/intel_sprite.c @@ -110,14 +110,18 @@ ivb_update_plane(struct drm_plane *plane, struct drm_framebuffer *fb, * when scaling is disabled. */ if (crtc_w != src_w || crtc_h != src_h) { - dev_priv->sprite_scaling_enabled = true; - intel_update_watermarks(dev); - intel_wait_for_vblank(dev, pipe); + if (!dev_priv->sprite_scaling_enabled) { + dev_priv->sprite_scaling_enabled = true; + intel_update_watermarks(dev); + intel_wait_for_vblank(dev, pipe); + } sprscale = SPRITE_SCALE_ENABLE | (src_w << 16) | src_h; } else { - dev_priv->sprite_scaling_enabled = false; - /* potentially re-enable LP watermarks */ - intel_update_watermarks(dev); + if (dev_priv->sprite_scaling_enabled) { + dev_priv->sprite_scaling_enabled = false; + /* potentially re-enable LP watermarks */ + intel_update_watermarks(dev); + } } I915_WRITE(SPRSTRIDE(pipe), fb->pitches[0]); @@ -151,6 +155,9 @@ ivb_disable_plane(struct drm_plane *plane) /* Activate double buffered register update */ I915_MODIFY_DISPBASE(SPRSURF(pipe), 0); POSTING_READ(SPRSURF(pipe)); + + dev_priv->sprite_scaling_enabled = false; + intel_update_watermarks(dev); } static int @@ -551,14 +558,13 @@ int intel_sprite_set_colorkey(struct drm_device *dev, void *data, struct drm_file *file_priv) { struct drm_intel_sprite_colorkey *set = data; - struct drm_i915_private *dev_priv = dev->dev_private; struct drm_mode_object *obj; struct drm_plane *plane; struct intel_plane *intel_plane; int ret = 0; - if (!dev_priv) - return -EINVAL; + if (!drm_core_check_feature(dev, DRIVER_MODESET)) + return -ENODEV; /* Make sure we don't try to enable both src & dest simultaneously */ if ((set->flags & (I915_SET_COLORKEY_DESTINATION | I915_SET_COLORKEY_SOURCE)) == (I915_SET_COLORKEY_DESTINATION | I915_SET_COLORKEY_SOURCE)) @@ -585,14 +591,13 @@ int intel_sprite_get_colorkey(struct drm_device *dev, void *data, struct drm_file *file_priv) { struct drm_intel_sprite_colorkey *get = data; - struct drm_i915_private *dev_priv = dev->dev_private; struct drm_mode_object *obj; struct drm_plane *plane; struct intel_plane *intel_plane; int ret = 0; - if (!dev_priv) - return -EINVAL; + if (!drm_core_check_feature(dev, DRIVER_MODESET)) + return -ENODEV; mutex_lock(&dev->mode_config.mutex); diff --git a/drivers/gpu/drm/i915/intel_tv.c b/drivers/gpu/drm/i915/intel_tv.c index 67f444d..3346612 100644 --- a/drivers/gpu/drm/i915/intel_tv.c +++ b/drivers/gpu/drm/i915/intel_tv.c @@ -1249,11 +1249,8 @@ intel_tv_detect(struct drm_connector *connector, bool force) int type; mode = reported_modes[0]; - drm_mode_set_crtcinfo(&mode, 0); - if (intel_tv->base.base.crtc && intel_tv->base.base.crtc->enabled) { - type = intel_tv_detect_type(intel_tv, connector); - } else if (force) { + if (force) { struct intel_load_detect_pipe tmp; if (intel_get_load_detect_pipe(&intel_tv->base, connector, |