summaryrefslogtreecommitdiffstats
path: root/drivers/gpu/drm/i915
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/gpu/drm/i915')
-rw-r--r--drivers/gpu/drm/i915/i915_debugfs.c6
-rw-r--r--drivers/gpu/drm/i915/i915_gem.c63
-rw-r--r--drivers/gpu/drm/i915/i915_gem_execbuffer.c17
-rw-r--r--drivers/gpu/drm/i915/i915_gem_request.c2
-rw-r--r--drivers/gpu/drm/i915/i915_gem_request.h2
-rw-r--r--drivers/gpu/drm/i915/i915_guc_submission.c4
-rw-r--r--drivers/gpu/drm/i915/i915_sw_fence.c35
-rw-r--r--drivers/gpu/drm/i915/i915_sw_fence.h2
-rw-r--r--drivers/gpu/drm/i915/i915_vma.c5
-rw-r--r--drivers/gpu/drm/i915/intel_acpi.c14
-rw-r--r--drivers/gpu/drm/i915/intel_display.c30
-rw-r--r--drivers/gpu/drm/i915/intel_dp_aux_backlight.c2
-rw-r--r--drivers/gpu/drm/i915/intel_lrc.c6
-rw-r--r--drivers/gpu/drm/i915/intel_ringbuffer.c41
-rw-r--r--drivers/gpu/drm/i915/intel_ringbuffer.h19
15 files changed, 155 insertions, 93 deletions
diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index d689e511744e..4bd1467c17b1 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -292,6 +292,8 @@ static int per_file_stats(int id, void *ptr, void *data)
struct file_stats *stats = data;
struct i915_vma *vma;
+ lockdep_assert_held(&obj->base.dev->struct_mutex);
+
stats->count++;
stats->total += obj->base.size;
if (!obj->bind_count)
@@ -476,6 +478,8 @@ static int i915_gem_object_info(struct seq_file *m, void *data)
struct drm_i915_gem_request *request;
struct task_struct *task;
+ mutex_lock(&dev->struct_mutex);
+
memset(&stats, 0, sizeof(stats));
stats.file_priv = file->driver_priv;
spin_lock(&file->table_lock);
@@ -487,7 +491,6 @@ static int i915_gem_object_info(struct seq_file *m, void *data)
* still alive (e.g. get_pid(current) => fork() => exit()).
* Therefore, we need to protect this ->comm access using RCU.
*/
- mutex_lock(&dev->struct_mutex);
request = list_first_entry_or_null(&file_priv->mm.request_list,
struct drm_i915_gem_request,
client_link);
@@ -497,6 +500,7 @@ static int i915_gem_object_info(struct seq_file *m, void *data)
PIDTYPE_PID);
print_file_stats(m, task ? task->comm : "<unknown>", stats);
rcu_read_unlock();
+
mutex_unlock(&dev->struct_mutex);
}
mutex_unlock(&dev->filelist_mutex);
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 462031cbd77f..615f0a855222 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -2285,8 +2285,8 @@ i915_gem_object_get_pages_gtt(struct drm_i915_gem_object *obj)
struct page *page;
unsigned long last_pfn = 0; /* suppress gcc warning */
unsigned int max_segment;
+ gfp_t noreclaim;
int ret;
- gfp_t gfp;
/* Assert that the object is not currently in any GPU domain. As it
* wasn't in the GTT, there shouldn't be any way it could have been in
@@ -2315,22 +2315,31 @@ rebuild_st:
* Fail silently without starting the shrinker
*/
mapping = obj->base.filp->f_mapping;
- gfp = mapping_gfp_constraint(mapping, ~(__GFP_IO | __GFP_RECLAIM));
- gfp |= __GFP_NORETRY | __GFP_NOWARN;
+ noreclaim = mapping_gfp_constraint(mapping,
+ ~(__GFP_IO | __GFP_RECLAIM));
+ noreclaim |= __GFP_NORETRY | __GFP_NOWARN;
+
sg = st->sgl;
st->nents = 0;
for (i = 0; i < page_count; i++) {
- page = shmem_read_mapping_page_gfp(mapping, i, gfp);
- if (unlikely(IS_ERR(page))) {
- i915_gem_shrink(dev_priv,
- page_count,
- I915_SHRINK_BOUND |
- I915_SHRINK_UNBOUND |
- I915_SHRINK_PURGEABLE);
+ const unsigned int shrink[] = {
+ I915_SHRINK_BOUND | I915_SHRINK_UNBOUND | I915_SHRINK_PURGEABLE,
+ 0,
+ }, *s = shrink;
+ gfp_t gfp = noreclaim;
+
+ do {
page = shmem_read_mapping_page_gfp(mapping, i, gfp);
- }
- if (unlikely(IS_ERR(page))) {
- gfp_t reclaim;
+ if (likely(!IS_ERR(page)))
+ break;
+
+ if (!*s) {
+ ret = PTR_ERR(page);
+ goto err_sg;
+ }
+
+ i915_gem_shrink(dev_priv, 2 * page_count, *s++);
+ cond_resched();
/* We've tried hard to allocate the memory by reaping
* our own buffer, now let the real VM do its job and
@@ -2340,15 +2349,26 @@ rebuild_st:
* defer the oom here by reporting the ENOMEM back
* to userspace.
*/
- reclaim = mapping_gfp_mask(mapping);
- reclaim |= __GFP_NORETRY; /* reclaim, but no oom */
-
- page = shmem_read_mapping_page_gfp(mapping, i, reclaim);
- if (IS_ERR(page)) {
- ret = PTR_ERR(page);
- goto err_sg;
+ if (!*s) {
+ /* reclaim and warn, but no oom */
+ gfp = mapping_gfp_mask(mapping);
+
+ /* Our bo are always dirty and so we require
+ * kswapd to reclaim our pages (direct reclaim
+ * does not effectively begin pageout of our
+ * buffers on its own). However, direct reclaim
+ * only waits for kswapd when under allocation
+ * congestion. So as a result __GFP_RECLAIM is
+ * unreliable and fails to actually reclaim our
+ * dirty pages -- unless you try over and over
+ * again with !__GFP_NORETRY. However, we still
+ * want to fail this allocation rather than
+ * trigger the out-of-memory killer and for
+ * this we want the future __GFP_MAYFAIL.
+ */
}
- }
+ } while (1);
+
if (!i ||
sg->length >= max_segment ||
page_to_pfn(page) != last_pfn + 1) {
@@ -4222,6 +4242,7 @@ i915_gem_object_create(struct drm_i915_private *dev_priv, u64 size)
mapping = obj->base.filp->f_mapping;
mapping_set_gfp_mask(mapping, mask);
+ GEM_BUG_ON(!(mapping_gfp_mask(mapping) & __GFP_RECLAIM));
i915_gem_object_init(obj, &i915_gem_object_ops);
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index a3e59c8ef27b..9ad13eeed904 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -546,11 +546,12 @@ repeat:
}
static int
-i915_gem_execbuffer_relocate_entry(struct drm_i915_gem_object *obj,
+i915_gem_execbuffer_relocate_entry(struct i915_vma *vma,
struct eb_vmas *eb,
struct drm_i915_gem_relocation_entry *reloc,
struct reloc_cache *cache)
{
+ struct drm_i915_gem_object *obj = vma->obj;
struct drm_i915_private *dev_priv = to_i915(obj->base.dev);
struct drm_gem_object *target_obj;
struct drm_i915_gem_object *target_i915_obj;
@@ -628,6 +629,16 @@ i915_gem_execbuffer_relocate_entry(struct drm_i915_gem_object *obj,
return -EINVAL;
}
+ /*
+ * If we write into the object, we need to force the synchronisation
+ * barrier, either with an asynchronous clflush or if we executed the
+ * patching using the GPU (though that should be serialised by the
+ * timeline). To be completely sure, and since we are required to
+ * do relocations we are already stalling, disable the user's opt
+ * of our synchronisation.
+ */
+ vma->exec_entry->flags &= ~EXEC_OBJECT_ASYNC;
+
ret = relocate_entry(obj, reloc, cache, target_offset);
if (ret)
return ret;
@@ -678,7 +689,7 @@ i915_gem_execbuffer_relocate_vma(struct i915_vma *vma,
do {
u64 offset = r->presumed_offset;
- ret = i915_gem_execbuffer_relocate_entry(vma->obj, eb, r, &cache);
+ ret = i915_gem_execbuffer_relocate_entry(vma, eb, r, &cache);
if (ret)
goto out;
@@ -726,7 +737,7 @@ i915_gem_execbuffer_relocate_vma_slow(struct i915_vma *vma,
reloc_cache_init(&cache, eb->i915);
for (i = 0; i < entry->relocation_count; i++) {
- ret = i915_gem_execbuffer_relocate_entry(vma->obj, eb, &relocs[i], &cache);
+ ret = i915_gem_execbuffer_relocate_entry(vma, eb, &relocs[i], &cache);
if (ret)
break;
}
diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c
index 5ddbc9499775..a74d0ac737cb 100644
--- a/drivers/gpu/drm/i915/i915_gem_request.c
+++ b/drivers/gpu/drm/i915/i915_gem_request.c
@@ -623,7 +623,7 @@ i915_gem_request_alloc(struct intel_engine_cs *engine,
* GPU processing the request, we never over-estimate the
* position of the head.
*/
- req->head = req->ring->tail;
+ req->head = req->ring->emit;
/* Check that we didn't interrupt ourselves with a new request */
GEM_BUG_ON(req->timeline->seqno != req->fence.seqno);
diff --git a/drivers/gpu/drm/i915/i915_gem_request.h b/drivers/gpu/drm/i915/i915_gem_request.h
index 129c58bb4805..a4a920c4c454 100644
--- a/drivers/gpu/drm/i915/i915_gem_request.h
+++ b/drivers/gpu/drm/i915/i915_gem_request.h
@@ -123,7 +123,7 @@ struct drm_i915_gem_request {
* It is used by the driver to then queue the request for execution.
*/
struct i915_sw_fence submit;
- wait_queue_t submitq;
+ wait_queue_entry_t submitq;
wait_queue_head_t execute;
/* A list of everyone we wait upon, and everyone who waits upon us.
diff --git a/drivers/gpu/drm/i915/i915_guc_submission.c b/drivers/gpu/drm/i915/i915_guc_submission.c
index 1642fff9cf13..ab5140ba108d 100644
--- a/drivers/gpu/drm/i915/i915_guc_submission.c
+++ b/drivers/gpu/drm/i915/i915_guc_submission.c
@@ -480,9 +480,7 @@ static void guc_wq_item_append(struct i915_guc_client *client,
GEM_BUG_ON(freespace < wqi_size);
/* The GuC firmware wants the tail index in QWords, not bytes */
- tail = rq->tail;
- assert_ring_tail_valid(rq->ring, rq->tail);
- tail >>= 3;
+ tail = intel_ring_set_tail(rq->ring, rq->tail) >> 3;
GEM_BUG_ON(tail > WQ_RING_TAIL_MAX);
/* For now workqueue item is 4 DWs; workqueue buffer is 2 pages. So we
diff --git a/drivers/gpu/drm/i915/i915_sw_fence.c b/drivers/gpu/drm/i915/i915_sw_fence.c
index a277f8eb7beb..380de4360b8a 100644
--- a/drivers/gpu/drm/i915/i915_sw_fence.c
+++ b/drivers/gpu/drm/i915/i915_sw_fence.c
@@ -152,7 +152,7 @@ static void __i915_sw_fence_wake_up_all(struct i915_sw_fence *fence,
struct list_head *continuation)
{
wait_queue_head_t *x = &fence->wait;
- wait_queue_t *pos, *next;
+ wait_queue_entry_t *pos, *next;
unsigned long flags;
debug_fence_deactivate(fence);
@@ -160,31 +160,30 @@ static void __i915_sw_fence_wake_up_all(struct i915_sw_fence *fence,
/*
* To prevent unbounded recursion as we traverse the graph of
- * i915_sw_fences, we move the task_list from this, the next ready
- * fence, to the tail of the original fence's task_list
+ * i915_sw_fences, we move the entry list from this, the next ready
+ * fence, to the tail of the original fence's entry list
* (and so added to the list to be woken).
*/
spin_lock_irqsave_nested(&x->lock, flags, 1 + !!continuation);
if (continuation) {
- list_for_each_entry_safe(pos, next, &x->task_list, task_list) {
+ list_for_each_entry_safe(pos, next, &x->head, entry) {
if (pos->func == autoremove_wake_function)
pos->func(pos, TASK_NORMAL, 0, continuation);
else
- list_move_tail(&pos->task_list, continuation);
+ list_move_tail(&pos->entry, continuation);
}
} else {
LIST_HEAD(extra);
do {
- list_for_each_entry_safe(pos, next,
- &x->task_list, task_list)
+ list_for_each_entry_safe(pos, next, &x->head, entry)
pos->func(pos, TASK_NORMAL, 0, &extra);
if (list_empty(&extra))
break;
- list_splice_tail_init(&extra, &x->task_list);
+ list_splice_tail_init(&extra, &x->head);
} while (1);
}
spin_unlock_irqrestore(&x->lock, flags);
@@ -254,9 +253,9 @@ void i915_sw_fence_commit(struct i915_sw_fence *fence)
__i915_sw_fence_commit(fence);
}
-static int i915_sw_fence_wake(wait_queue_t *wq, unsigned mode, int flags, void *key)
+static int i915_sw_fence_wake(wait_queue_entry_t *wq, unsigned mode, int flags, void *key)
{
- list_del(&wq->task_list);
+ list_del(&wq->entry);
__i915_sw_fence_complete(wq->private, key);
i915_sw_fence_put(wq->private);
if (wq->flags & I915_SW_FENCE_FLAG_ALLOC)
@@ -267,7 +266,7 @@ static int i915_sw_fence_wake(wait_queue_t *wq, unsigned mode, int flags, void *
static bool __i915_sw_fence_check_if_after(struct i915_sw_fence *fence,
const struct i915_sw_fence * const signaler)
{
- wait_queue_t *wq;
+ wait_queue_entry_t *wq;
if (__test_and_set_bit(I915_SW_FENCE_CHECKED_BIT, &fence->flags))
return false;
@@ -275,7 +274,7 @@ static bool __i915_sw_fence_check_if_after(struct i915_sw_fence *fence,
if (fence == signaler)
return true;
- list_for_each_entry(wq, &fence->wait.task_list, task_list) {
+ list_for_each_entry(wq, &fence->wait.head, entry) {
if (wq->func != i915_sw_fence_wake)
continue;
@@ -288,12 +287,12 @@ static bool __i915_sw_fence_check_if_after(struct i915_sw_fence *fence,
static void __i915_sw_fence_clear_checked_bit(struct i915_sw_fence *fence)
{
- wait_queue_t *wq;
+ wait_queue_entry_t *wq;
if (!__test_and_clear_bit(I915_SW_FENCE_CHECKED_BIT, &fence->flags))
return;
- list_for_each_entry(wq, &fence->wait.task_list, task_list) {
+ list_for_each_entry(wq, &fence->wait.head, entry) {
if (wq->func != i915_sw_fence_wake)
continue;
@@ -320,7 +319,7 @@ static bool i915_sw_fence_check_if_after(struct i915_sw_fence *fence,
static int __i915_sw_fence_await_sw_fence(struct i915_sw_fence *fence,
struct i915_sw_fence *signaler,
- wait_queue_t *wq, gfp_t gfp)
+ wait_queue_entry_t *wq, gfp_t gfp)
{
unsigned long flags;
int pending;
@@ -350,7 +349,7 @@ static int __i915_sw_fence_await_sw_fence(struct i915_sw_fence *fence,
pending |= I915_SW_FENCE_FLAG_ALLOC;
}
- INIT_LIST_HEAD(&wq->task_list);
+ INIT_LIST_HEAD(&wq->entry);
wq->flags = pending;
wq->func = i915_sw_fence_wake;
wq->private = i915_sw_fence_get(fence);
@@ -359,7 +358,7 @@ static int __i915_sw_fence_await_sw_fence(struct i915_sw_fence *fence,
spin_lock_irqsave(&signaler->wait.lock, flags);
if (likely(!i915_sw_fence_done(signaler))) {
- __add_wait_queue_tail(&signaler->wait, wq);
+ __add_wait_queue_entry_tail(&signaler->wait, wq);
pending = 1;
} else {
i915_sw_fence_wake(wq, 0, 0, NULL);
@@ -372,7 +371,7 @@ static int __i915_sw_fence_await_sw_fence(struct i915_sw_fence *fence,
int i915_sw_fence_await_sw_fence(struct i915_sw_fence *fence,
struct i915_sw_fence *signaler,
- wait_queue_t *wq)
+ wait_queue_entry_t *wq)
{
return __i915_sw_fence_await_sw_fence(fence, signaler, wq, 0);
}
diff --git a/drivers/gpu/drm/i915/i915_sw_fence.h b/drivers/gpu/drm/i915/i915_sw_fence.h
index d31cefbbcc04..fd3c3bf6c8b7 100644
--- a/drivers/gpu/drm/i915/i915_sw_fence.h
+++ b/drivers/gpu/drm/i915/i915_sw_fence.h
@@ -66,7 +66,7 @@ void i915_sw_fence_commit(struct i915_sw_fence *fence);
int i915_sw_fence_await_sw_fence(struct i915_sw_fence *fence,
struct i915_sw_fence *after,
- wait_queue_t *wq);
+ wait_queue_entry_t *wq);
int i915_sw_fence_await_sw_fence_gfp(struct i915_sw_fence *fence,
struct i915_sw_fence *after,
gfp_t gfp);
diff --git a/drivers/gpu/drm/i915/i915_vma.c b/drivers/gpu/drm/i915/i915_vma.c
index 1aba47024656..f066e2d785f5 100644
--- a/drivers/gpu/drm/i915/i915_vma.c
+++ b/drivers/gpu/drm/i915/i915_vma.c
@@ -650,6 +650,11 @@ int i915_vma_unbind(struct i915_vma *vma)
break;
}
+ if (!ret) {
+ ret = i915_gem_active_retire(&vma->last_fence,
+ &vma->vm->i915->drm.struct_mutex);
+ }
+
__i915_vma_unpin(vma);
if (ret)
return ret;
diff --git a/drivers/gpu/drm/i915/intel_acpi.c b/drivers/gpu/drm/i915/intel_acpi.c
index eb638a1e69d2..42fb436f6cdc 100644
--- a/drivers/gpu/drm/i915/intel_acpi.c
+++ b/drivers/gpu/drm/i915/intel_acpi.c
@@ -15,13 +15,9 @@ static struct intel_dsm_priv {
acpi_handle dhandle;
} intel_dsm_priv;
-static const u8 intel_dsm_guid[] = {
- 0xd3, 0x73, 0xd8, 0x7e,
- 0xd0, 0xc2,
- 0x4f, 0x4e,
- 0xa8, 0x54,
- 0x0f, 0x13, 0x17, 0xb0, 0x1c, 0x2c
-};
+static const guid_t intel_dsm_guid =
+ GUID_INIT(0x7ed873d3, 0xc2d0, 0x4e4f,
+ 0xa8, 0x54, 0x0f, 0x13, 0x17, 0xb0, 0x1c, 0x2c);
static char *intel_dsm_port_name(u8 id)
{
@@ -80,7 +76,7 @@ static void intel_dsm_platform_mux_info(void)
int i;
union acpi_object *pkg, *connector_count;
- pkg = acpi_evaluate_dsm_typed(intel_dsm_priv.dhandle, intel_dsm_guid,
+ pkg = acpi_evaluate_dsm_typed(intel_dsm_priv.dhandle, &intel_dsm_guid,
INTEL_DSM_REVISION_ID, INTEL_DSM_FN_PLATFORM_MUX_INFO,
NULL, ACPI_TYPE_PACKAGE);
if (!pkg) {
@@ -118,7 +114,7 @@ static bool intel_dsm_pci_probe(struct pci_dev *pdev)
if (!dhandle)
return false;
- if (!acpi_check_dsm(dhandle, intel_dsm_guid, INTEL_DSM_REVISION_ID,
+ if (!acpi_check_dsm(dhandle, &intel_dsm_guid, INTEL_DSM_REVISION_ID,
1 << INTEL_DSM_FN_PLATFORM_MUX_INFO)) {
DRM_DEBUG_KMS("no _DSM method for intel device\n");
return false;
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 96b0b01677e2..9106ea32b048 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -120,7 +120,8 @@ static void intel_crtc_init_scalers(struct intel_crtc *crtc,
static void skylake_pfit_enable(struct intel_crtc *crtc);
static void ironlake_pfit_disable(struct intel_crtc *crtc, bool force);
static void ironlake_pfit_enable(struct intel_crtc *crtc);
-static void intel_modeset_setup_hw_state(struct drm_device *dev);
+static void intel_modeset_setup_hw_state(struct drm_device *dev,
+ struct drm_modeset_acquire_ctx *ctx);
static void intel_pre_disable_primary_noatomic(struct drm_crtc *crtc);
struct intel_limit {
@@ -3449,7 +3450,7 @@ __intel_display_resume(struct drm_device *dev,
struct drm_crtc *crtc;
int i, ret;
- intel_modeset_setup_hw_state(dev);
+ intel_modeset_setup_hw_state(dev, ctx);
i915_redisable_vga(to_i915(dev));
if (!state)
@@ -5825,7 +5826,8 @@ static void i9xx_crtc_disable(struct intel_crtc_state *old_crtc_state,
intel_update_watermarks(intel_crtc);
}
-static void intel_crtc_disable_noatomic(struct drm_crtc *crtc)
+static void intel_crtc_disable_noatomic(struct drm_crtc *crtc,
+ struct drm_modeset_acquire_ctx *ctx)
{
struct intel_encoder *encoder;
struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
@@ -5855,7 +5857,7 @@ static void intel_crtc_disable_noatomic(struct drm_crtc *crtc)
return;
}
- state->acquire_ctx = crtc->dev->mode_config.acquire_ctx;
+ state->acquire_ctx = ctx;
/* Everything's already locked, -EDEADLK can't happen. */
crtc_state = intel_atomic_get_crtc_state(state, intel_crtc);
@@ -15030,7 +15032,7 @@ int intel_modeset_init(struct drm_device *dev)
intel_setup_outputs(dev_priv);
drm_modeset_lock_all(dev);
- intel_modeset_setup_hw_state(dev);
+ intel_modeset_setup_hw_state(dev, dev->mode_config.acquire_ctx);
drm_modeset_unlock_all(dev);
for_each_intel_crtc(dev, crtc) {
@@ -15067,13 +15069,13 @@ int intel_modeset_init(struct drm_device *dev)
return 0;
}
-static void intel_enable_pipe_a(struct drm_device *dev)
+static void intel_enable_pipe_a(struct drm_device *dev,
+ struct drm_modeset_acquire_ctx *ctx)
{
struct intel_connector *connector;
struct drm_connector_list_iter conn_iter;
struct drm_connector *crt = NULL;
struct intel_load_detect_pipe load_detect_temp;
- struct drm_modeset_acquire_ctx *ctx = dev->mode_config.acquire_ctx;
int ret;
/* We can't just switch on the pipe A, we need to set things up with a
@@ -15145,7 +15147,8 @@ static bool has_pch_trancoder(struct drm_i915_private *dev_priv,
(HAS_PCH_LPT_H(dev_priv) && pch_transcoder == TRANSCODER_A);
}
-static void intel_sanitize_crtc(struct intel_crtc *crtc)
+static void intel_sanitize_crtc(struct intel_crtc *crtc,
+ struct drm_modeset_acquire_ctx *ctx)
{
struct drm_device *dev = crtc->base.dev;
struct drm_i915_private *dev_priv = to_i915(dev);
@@ -15191,7 +15194,7 @@ static void intel_sanitize_crtc(struct intel_crtc *crtc)
plane = crtc->plane;
crtc->base.primary->state->visible = true;
crtc->plane = !plane;
- intel_crtc_disable_noatomic(&crtc->base);
+ intel_crtc_disable_noatomic(&crtc->base, ctx);
crtc->plane = plane;
}
@@ -15201,13 +15204,13 @@ static void intel_sanitize_crtc(struct intel_crtc *crtc)
* resume. Force-enable the pipe to fix this, the update_dpms
* call below we restore the pipe to the right state, but leave
* the required bits on. */
- intel_enable_pipe_a(dev);
+ intel_enable_pipe_a(dev, ctx);
}
/* Adjust the state of the output pipe according to whether we
* have active connectors/encoders. */
if (crtc->active && !intel_crtc_has_encoders(crtc))
- intel_crtc_disable_noatomic(&crtc->base);
+ intel_crtc_disable_noatomic(&crtc->base, ctx);
if (crtc->active || HAS_GMCH_DISPLAY(dev_priv)) {
/*
@@ -15505,7 +15508,8 @@ get_encoder_power_domains(struct drm_i915_private *dev_priv)
* and sanitizes it to the current state
*/
static void
-intel_modeset_setup_hw_state(struct drm_device *dev)
+intel_modeset_setup_hw_state(struct drm_device *dev,
+ struct drm_modeset_acquire_ctx *ctx)
{
struct drm_i915_private *dev_priv = to_i915(dev);
enum pipe pipe;
@@ -15525,7 +15529,7 @@ intel_modeset_setup_hw_state(struct drm_device *dev)
for_each_pipe(dev_priv, pipe) {
crtc = intel_get_crtc_for_pipe(dev_priv, pipe);
- intel_sanitize_crtc(crtc);
+ intel_sanitize_crtc(crtc, ctx);
intel_dump_pipe_config(crtc, crtc->config,
"[setup_hw_state]");
}
diff --git a/drivers/gpu/drm/i915/intel_dp_aux_backlight.c b/drivers/gpu/drm/i915/intel_dp_aux_backlight.c
index 6532e226db29..40ba3134545e 100644
--- a/drivers/gpu/drm/i915/intel_dp_aux_backlight.c
+++ b/drivers/gpu/drm/i915/intel_dp_aux_backlight.c
@@ -119,8 +119,6 @@ static int intel_dp_aux_setup_backlight(struct intel_connector *connector,
struct intel_dp *intel_dp = enc_to_intel_dp(&connector->encoder->base);
struct intel_panel *panel = &connector->panel;
- intel_dp_aux_enable_backlight(connector);
-
if (intel_dp->edp_dpcd[2] & DP_EDP_BACKLIGHT_BRIGHTNESS_BYTE_COUNT)
panel->backlight.max = 0xFFFF;
else
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index dac4e003c1f3..62f44d3e7c43 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -326,8 +326,7 @@ static u64 execlists_update_context(struct drm_i915_gem_request *rq)
rq->ctx->ppgtt ?: rq->i915->mm.aliasing_ppgtt;
u32 *reg_state = ce->lrc_reg_state;
- assert_ring_tail_valid(rq->ring, rq->tail);
- reg_state[CTX_RING_TAIL+1] = rq->tail;
+ reg_state[CTX_RING_TAIL+1] = intel_ring_set_tail(rq->ring, rq->tail);
/* True 32b PPGTT with dynamic page allocation: update PDP
* registers and point the unallocated PDPs to scratch page.
@@ -2036,8 +2035,7 @@ void intel_lr_context_resume(struct drm_i915_private *dev_priv)
ce->state->obj->mm.dirty = true;
i915_gem_object_unpin_map(ce->state->obj);
- ce->ring->head = ce->ring->tail = 0;
- intel_ring_update_space(ce->ring);
+ intel_ring_reset(ce->ring, 0);
}
}
}
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 66a2b8b83972..513a0f4b469b 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -49,7 +49,7 @@ static int __intel_ring_space(int head, int tail, int size)
void intel_ring_update_space(struct intel_ring *ring)
{
- ring->space = __intel_ring_space(ring->head, ring->tail, ring->size);
+ ring->space = __intel_ring_space(ring->head, ring->emit, ring->size);
}
static int
@@ -774,8 +774,8 @@ static void i9xx_submit_request(struct drm_i915_gem_request *request)
i915_gem_request_submit(request);
- assert_ring_tail_valid(request->ring, request->tail);
- I915_WRITE_TAIL(request->engine, request->tail);
+ I915_WRITE_TAIL(request->engine,
+ intel_ring_set_tail(request->ring, request->tail));
}
static void i9xx_emit_breadcrumb(struct drm_i915_gem_request *req, u32 *cs)
@@ -1316,11 +1316,23 @@ err:
return PTR_ERR(addr);
}
+void intel_ring_reset(struct intel_ring *ring, u32 tail)
+{
+ GEM_BUG_ON(!list_empty(&ring->request_list));
+ ring->tail = tail;
+ ring->head = tail;
+ ring->emit = tail;
+ intel_ring_update_space(ring);
+}
+
void intel_ring_unpin(struct intel_ring *ring)
{
GEM_BUG_ON(!ring->vma);
GEM_BUG_ON(!ring->vaddr);
+ /* Discard any unused bytes beyond that submitted to hw. */
+ intel_ring_reset(ring, ring->tail);
+
if (i915_vma_is_map_and_fenceable(ring->vma))
i915_vma_unpin_iomap(ring->vma);
else
@@ -1562,8 +1574,9 @@ void intel_legacy_submission_resume(struct drm_i915_private *dev_priv)
struct intel_engine_cs *engine;
enum intel_engine_id id;
+ /* Restart from the beginning of the rings for convenience */
for_each_engine(engine, dev_priv, id)
- engine->buffer->head = engine->buffer->tail;
+ intel_ring_reset(engine->buffer, 0);
}
static int ring_request_alloc(struct drm_i915_gem_request *request)
@@ -1616,7 +1629,7 @@ static int wait_for_space(struct drm_i915_gem_request *req, int bytes)
unsigned space;
/* Would completion of this request free enough space? */
- space = __intel_ring_space(target->postfix, ring->tail,
+ space = __intel_ring_space(target->postfix, ring->emit,
ring->size);
if (space >= bytes)
break;
@@ -1641,8 +1654,8 @@ static int wait_for_space(struct drm_i915_gem_request *req, int bytes)
u32 *intel_ring_begin(struct drm_i915_gem_request *req, int num_dwords)
{
struct intel_ring *ring = req->ring;
- int remain_actual = ring->size - ring->tail;
- int remain_usable = ring->effective_size - ring->tail;
+ int remain_actual = ring->size - ring->emit;
+ int remain_usable = ring->effective_size - ring->emit;
int bytes = num_dwords * sizeof(u32);
int total_bytes, wait_bytes;
bool need_wrap = false;
@@ -1678,17 +1691,17 @@ u32 *intel_ring_begin(struct drm_i915_gem_request *req, int num_dwords)
if (unlikely(need_wrap)) {
GEM_BUG_ON(remain_actual > ring->space);
- GEM_BUG_ON(ring->tail + remain_actual > ring->size);
+ GEM_BUG_ON(ring->emit + remain_actual > ring->size);
/* Fill the tail with MI_NOOP */
- memset(ring->vaddr + ring->tail, 0, remain_actual);
- ring->tail = 0;
+ memset(ring->vaddr + ring->emit, 0, remain_actual);
+ ring->emit = 0;
ring->space -= remain_actual;
}
- GEM_BUG_ON(ring->tail > ring->size - bytes);
- cs = ring->vaddr + ring->tail;
- ring->tail += bytes;
+ GEM_BUG_ON(ring->emit > ring->size - bytes);
+ cs = ring->vaddr + ring->emit;
+ ring->emit += bytes;
ring->space -= bytes;
GEM_BUG_ON(ring->space < 0);
@@ -1699,7 +1712,7 @@ u32 *intel_ring_begin(struct drm_i915_gem_request *req, int num_dwords)
int intel_ring_cacheline_align(struct drm_i915_gem_request *req)
{
int num_dwords =
- (req->ring->tail & (CACHELINE_BYTES - 1)) / sizeof(uint32_t);
+ (req->ring->emit & (CACHELINE_BYTES - 1)) / sizeof(uint32_t);
u32 *cs;
if (num_dwords == 0)
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index a82a0807f64d..f7144fe09613 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -145,6 +145,7 @@ struct intel_ring {
u32 head;
u32 tail;
+ u32 emit;
int space;
int size;
@@ -488,6 +489,8 @@ intel_write_status_page(struct intel_engine_cs *engine, int reg, u32 value)
struct intel_ring *
intel_engine_create_ring(struct intel_engine_cs *engine, int size);
int intel_ring_pin(struct intel_ring *ring, unsigned int offset_bias);
+void intel_ring_reset(struct intel_ring *ring, u32 tail);
+void intel_ring_update_space(struct intel_ring *ring);
void intel_ring_unpin(struct intel_ring *ring);
void intel_ring_free(struct intel_ring *ring);
@@ -511,7 +514,7 @@ intel_ring_advance(struct drm_i915_gem_request *req, u32 *cs)
* reserved for the command packet (i.e. the value passed to
* intel_ring_begin()).
*/
- GEM_BUG_ON((req->ring->vaddr + req->ring->tail) != cs);
+ GEM_BUG_ON((req->ring->vaddr + req->ring->emit) != cs);
}
static inline u32
@@ -540,7 +543,19 @@ assert_ring_tail_valid(const struct intel_ring *ring, unsigned int tail)
GEM_BUG_ON(tail >= ring->size);
}
-void intel_ring_update_space(struct intel_ring *ring);
+static inline unsigned int
+intel_ring_set_tail(struct intel_ring *ring, unsigned int tail)
+{
+ /* Whilst writes to the tail are strictly order, there is no
+ * serialisation between readers and the writers. The tail may be
+ * read by i915_gem_request_retire() just as it is being updated
+ * by execlists, as although the breadcrumb is complete, the context
+ * switch hasn't been seen.
+ */
+ assert_ring_tail_valid(ring, tail);
+ ring->tail = tail;
+ return tail;
+}
void intel_engine_init_global_seqno(struct intel_engine_cs *engine, u32 seqno);
OpenPOWER on IntegriCloud