Mesa (master): i965: Track format and aux usage in the render cache
Jason Ekstrand
jekstrand at kemper.freedesktop.org
Wed Jan 17 05:41:38 UTC 2018
Module: Mesa
Branch: master
Commit: d84275b884244a2fd3a6e67ceb2a5277e5edf89a
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=d84275b884244a2fd3a6e67ceb2a5277e5edf89a
Author: Jason Ekstrand <jason.ekstrand at intel.com>
Date: Wed Dec 13 17:25:26 2017 -0800
i965: Track format and aux usage in the render cache
This lets us perform render cache flushes whenever a surface goes from
being used with one aux+format to a different aux+format.
This is the "proper" fix for https://bugs.freedesktop.org/102435.
ee57b15ec764736e2d5360beaef9fb2045ed0f68 which was really just a partial
revert of 3e57e9494c2279580ad6a83ab8c065d01e7e634e was just a hack to
get rid of a hang in a bunch of Valve games. This solves the actual
problem responsible for the hang and lets us enable CCS_E once again.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=102435
Reviewed-by: Iago Toral Quiroga <itoral at igalia.com>
Reviewed-by: Kenneth Graunke <kenneth at whitecape.org>
Cc: "17.3" <mesa-stable at lists.freedesktop.org>
---
src/mesa/drivers/dri/i965/brw_context.h | 2 +-
src/mesa/drivers/dri/i965/brw_draw.c | 20 +++++---
src/mesa/drivers/dri/i965/genX_blorp_exec.c | 14 ++++--
src/mesa/drivers/dri/i965/intel_fbo.c | 75 +++++++++++++++++++++++------
src/mesa/drivers/dri/i965/intel_fbo.h | 8 ++-
5 files changed, 92 insertions(+), 27 deletions(-)
diff --git a/src/mesa/drivers/dri/i965/brw_context.h b/src/mesa/drivers/dri/i965/brw_context.h
index 8d8ab71093..3cbc2e8c13 100644
--- a/src/mesa/drivers/dri/i965/brw_context.h
+++ b/src/mesa/drivers/dri/i965/brw_context.h
@@ -764,7 +764,7 @@ struct brw_context
* and would need flushing before being used from another cache domain that
* isn't coherent with it (i.e. the sampler).
*/
- struct set *render_cache;
+ struct hash_table *render_cache;
/**
* Set of struct brw_bo * that have been used as a depth buffer within this
diff --git a/src/mesa/drivers/dri/i965/brw_draw.c b/src/mesa/drivers/dri/i965/brw_draw.c
index 1f86378f5e..96e014dc1f 100644
--- a/src/mesa/drivers/dri/i965/brw_draw.c
+++ b/src/mesa/drivers/dri/i965/brw_draw.c
@@ -503,13 +503,17 @@ brw_predraw_resolve_framebuffer(struct brw_context *brw)
mesa_format mesa_format =
_mesa_get_render_format(ctx, intel_rb_format(irb));
enum isl_format isl_format = brw_isl_format_for_mesa_format(mesa_format);
+ bool blend_enabled = ctx->Color.BlendEnabled & (1 << i);
+ enum isl_aux_usage aux_usage =
+ intel_miptree_render_aux_usage(brw, irb->mt, isl_format,
+ blend_enabled);
intel_miptree_prepare_render(brw, irb->mt, irb->mt_level,
irb->mt_layer, irb->layer_count,
- isl_format,
- ctx->Color.BlendEnabled & (1 << i));
+ isl_format, blend_enabled);
- brw_cache_flush_for_render(brw, irb->mt->bo);
+ brw_cache_flush_for_render(brw, irb->mt->bo,
+ isl_format, aux_usage);
}
}
@@ -575,12 +579,16 @@ brw_postdraw_set_buffers_need_resolve(struct brw_context *brw)
mesa_format mesa_format =
_mesa_get_render_format(ctx, intel_rb_format(irb));
enum isl_format isl_format = brw_isl_format_for_mesa_format(mesa_format);
+ bool blend_enabled = ctx->Color.BlendEnabled & (1 << i);
+ enum isl_aux_usage aux_usage =
+ intel_miptree_render_aux_usage(brw, irb->mt, isl_format,
+ blend_enabled);
+
+ brw_render_cache_add_bo(brw, irb->mt->bo, isl_format, aux_usage);
- brw_render_cache_add_bo(brw, irb->mt->bo);
intel_miptree_finish_render(brw, irb->mt, irb->mt_level,
irb->mt_layer, irb->layer_count,
- isl_format,
- ctx->Color.BlendEnabled & (1 << i));
+ isl_format, blend_enabled);
}
}
diff --git a/src/mesa/drivers/dri/i965/genX_blorp_exec.c b/src/mesa/drivers/dri/i965/genX_blorp_exec.c
index e8bc52e5d7..062171af60 100644
--- a/src/mesa/drivers/dri/i965/genX_blorp_exec.c
+++ b/src/mesa/drivers/dri/i965/genX_blorp_exec.c
@@ -239,8 +239,11 @@ genX(blorp_exec)(struct blorp_batch *batch,
*/
if (params->src.enabled)
brw_cache_flush_for_read(brw, params->src.addr.buffer);
- if (params->dst.enabled)
- brw_cache_flush_for_render(brw, params->dst.addr.buffer);
+ if (params->dst.enabled) {
+ brw_cache_flush_for_render(brw, params->dst.addr.buffer,
+ params->dst.view.format,
+ params->dst.aux_usage);
+ }
if (params->depth.enabled)
brw_cache_flush_for_depth(brw, params->depth.addr.buffer);
if (params->stencil.enabled)
@@ -310,8 +313,11 @@ retry:
!params->stencil.enabled;
brw->ib.index_size = -1;
- if (params->dst.enabled)
- brw_render_cache_add_bo(brw, params->dst.addr.buffer);
+ if (params->dst.enabled) {
+ brw_render_cache_add_bo(brw, params->dst.addr.buffer,
+ params->dst.view.format,
+ params->dst.aux_usage);
+ }
if (params->depth.enabled)
brw_depth_cache_add_bo(brw, params->depth.addr.buffer);
if (params->stencil.enabled)
diff --git a/src/mesa/drivers/dri/i965/intel_fbo.c b/src/mesa/drivers/dri/i965/intel_fbo.c
index 75c85ecb63..056f494e2f 100644
--- a/src/mesa/drivers/dri/i965/intel_fbo.c
+++ b/src/mesa/drivers/dri/i965/intel_fbo.c
@@ -972,14 +972,13 @@ intel_renderbuffer_move_to_temp(struct brw_context *brw,
void
brw_cache_sets_clear(struct brw_context *brw)
{
- struct set_entry *entry;
+ struct hash_entry *render_entry;
+ hash_table_foreach(brw->render_cache, render_entry)
+ _mesa_hash_table_remove(brw->render_cache, render_entry);
- set_foreach(brw->render_cache, entry) {
- _mesa_set_remove(brw->render_cache, entry);
- }
-
- set_foreach(brw->depth_cache, entry)
- _mesa_set_remove(brw->depth_cache, entry);
+ struct set_entry *depth_entry;
+ set_foreach(brw->depth_cache, depth_entry)
+ _mesa_set_remove(brw->depth_cache, depth_entry);
}
/**
@@ -1018,28 +1017,76 @@ flush_depth_and_render_caches(struct brw_context *brw, struct brw_bo *bo)
void
brw_cache_flush_for_read(struct brw_context *brw, struct brw_bo *bo)
{
- if (_mesa_set_search(brw->render_cache, bo) ||
+ if (_mesa_hash_table_search(brw->render_cache, bo) ||
_mesa_set_search(brw->depth_cache, bo))
flush_depth_and_render_caches(brw, bo);
}
+static void *
+format_aux_tuple(enum isl_format format, enum isl_aux_usage aux_usage)
+{
+ return (void *)(uintptr_t)((uint32_t)format << 8 | aux_usage);
+}
+
void
-brw_cache_flush_for_render(struct brw_context *brw, struct brw_bo *bo)
+brw_cache_flush_for_render(struct brw_context *brw, struct brw_bo *bo,
+ enum isl_format format,
+ enum isl_aux_usage aux_usage)
{
if (_mesa_set_search(brw->depth_cache, bo))
flush_depth_and_render_caches(brw, bo);
+
+ /* Check to see if this bo has been used by a previous rendering operation
+ * but with a different format or aux usage. If it has, flush the render
+ * cache so we ensure that it's only in there with one format or aux usage
+ * at a time.
+ *
+ * Even though it's not obvious, this can easily happen in practice.
+ * Suppose a client is blending on a surface with sRGB encode enabled on
+ * gen9. This implies that you get AUX_USAGE_CCS_D at best. If the client
+ * then disables sRGB decode and continues blending we will flip on
+ * AUX_USAGE_CCS_E without doing any sort of resolve in-between (this is
+ * perfectly valid since CCS_E is a subset of CCS_D). However, this means
+ * that we have fragments in-flight which are rendering with UNORM+CCS_E
+ * and other fragments in-flight with SRGB+CCS_D on the same surface at the
+ * same time and the pixel scoreboard and color blender are trying to sort
+ * it all out. This ends badly (i.e. GPU hangs).
+ *
+ * To date, we have never observed GPU hangs or even corruption to be
+ * associated with switching the format, only the aux usage. However,
+ * there are comments in various docs which indicate that the render cache
+ * isn't 100% resilient to format changes. We may as well be conservative
+ * and flush on format changes too. We can always relax this later if we
+ * find it to be a performance problem.
+ */
+ struct hash_entry *entry = _mesa_hash_table_search(brw->render_cache, bo);
+ if (entry && entry->data != format_aux_tuple(format, aux_usage))
+ flush_depth_and_render_caches(brw, bo);
}
void
-brw_render_cache_add_bo(struct brw_context *brw, struct brw_bo *bo)
+brw_render_cache_add_bo(struct brw_context *brw, struct brw_bo *bo,
+ enum isl_format format,
+ enum isl_aux_usage aux_usage)
{
- _mesa_set_add(brw->render_cache, bo);
+#ifndef NDEBUG
+ struct hash_entry *entry = _mesa_hash_table_search(brw->render_cache, bo);
+ if (entry) {
+ /* Otherwise, someone didn't do a flush_for_render and that would be
+ * very bad indeed.
+ */
+ assert(entry->data == format_aux_tuple(format, aux_usage));
+ }
+#endif
+
+ _mesa_hash_table_insert(brw->render_cache, bo,
+ format_aux_tuple(format, aux_usage));
}
void
brw_cache_flush_for_depth(struct brw_context *brw, struct brw_bo *bo)
{
- if (_mesa_set_search(brw->render_cache, bo))
+ if (_mesa_hash_table_search(brw->render_cache, bo))
flush_depth_and_render_caches(brw, bo);
}
@@ -1066,8 +1113,8 @@ intel_fbo_init(struct brw_context *brw)
dd->EGLImageTargetRenderbufferStorage =
intel_image_target_renderbuffer_storage;
- brw->render_cache = _mesa_set_create(brw, _mesa_hash_pointer,
- _mesa_key_pointer_equal);
+ brw->render_cache = _mesa_hash_table_create(brw, _mesa_hash_pointer,
+ _mesa_key_pointer_equal);
brw->depth_cache = _mesa_set_create(brw, _mesa_hash_pointer,
_mesa_key_pointer_equal);
}
diff --git a/src/mesa/drivers/dri/i965/intel_fbo.h b/src/mesa/drivers/dri/i965/intel_fbo.h
index 10be4bbc7d..88a5b6732b 100644
--- a/src/mesa/drivers/dri/i965/intel_fbo.h
+++ b/src/mesa/drivers/dri/i965/intel_fbo.h
@@ -236,9 +236,13 @@ intel_renderbuffer_upsample(struct brw_context *brw,
void brw_cache_sets_clear(struct brw_context *brw);
void brw_cache_flush_for_read(struct brw_context *brw, struct brw_bo *bo);
-void brw_cache_flush_for_render(struct brw_context *brw, struct brw_bo *bo);
+void brw_cache_flush_for_render(struct brw_context *brw, struct brw_bo *bo,
+ enum isl_format format,
+ enum isl_aux_usage aux_usage);
void brw_cache_flush_for_depth(struct brw_context *brw, struct brw_bo *bo);
-void brw_render_cache_add_bo(struct brw_context *brw, struct brw_bo *bo);
+void brw_render_cache_add_bo(struct brw_context *brw, struct brw_bo *bo,
+ enum isl_format format,
+ enum isl_aux_usage aux_usage);
void brw_depth_cache_add_bo(struct brw_context *brw, struct brw_bo *bo);
unsigned
More information about the mesa-commit
mailing list