drm/i915: Make LRC (un)pinning work on context and engine
Subsystems (2)
PATCH 1/3 drm/i915: Make LRC (un)pinning work on context and engine
From: Tvrtko Ursulin <[email protected]>
Previously intel_lr_context_(un)pin were operating on requests which is in conflict with their names.
If we make them take a context and an engine, it makes the names make more sense and it also makes future fixes possible.
--- drivers/gpu/drm/i915/i915_gem.c | 2 +- drivers/gpu/drm/i915/intel_lrc.c | 48 ++++++++++++++++++++-------------------- drivers/gpu/drm/i915/intel_lrc.h | 3 ++- 3 files changed, 27 insertions(+), 26 deletions(-)
drivers/gpu/drm/i915/i915_gem.c +1 -1
@@ -2681,7 +2681,7 @@ void i915_gem_request_free(struct kref *req_ref)if (ctx) {if (i915.enable_execlists) {if (ctx != req->ring->default_context) -intel_lr_context_unpin (req); +intel_lr_context_unpin (ctx, req->ring); }i915_gem_context_unreference (ctx);
drivers/gpu/drm/i915/intel_lrc.c +24 -24
108 lines not on this page · raw message
drivers/gpu/drm/i915/intel_lrc.h +2 -1
@@ -101,7 +101,8 @@ void intel_lr_context_free(struct intel_context *ctx);uint32_t intel_lr_context_size (struct intel_engine_cs *ring);int intel_lr_context_deferred_alloc (struct intel_context *ctx,struct intel_engine_cs *ring); -void intel_lr_context_unpin (struct drm_i915_gem_request *req); +void intel_lr_context_unpin (struct intel_context *ctx, +struct intel_engine_cs *engine);void intel_lr_context_reset (struct drm_device *dev,struct intel_context *ctx); uint64_t intel_lr_context_descriptor(struct intel_context *ctx,
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH 2/3 drm/i915: Make LRC pinning own a reference to the context
From: Tvrtko Ursulin <[email protected]>
Will simplify the following fix and sounds logical.
--- drivers/gpu/drm/i915/intel_lrc.c | 2 ++ 1 file changed, 2 insertions(+)
drivers/gpu/drm/i915/intel_lrc.c +2 -0
@@ -1105,6 +1105,7 @@ static int intel_lr_context_pin(struct intel_context *ctx, ret =intel_lr_context_do_pin (ctx, engine);if (ret) goto reset_pin_count; +i915_gem_context_reference (ctx); }return ret;
@@ -1130,6 +1131,7 @@ void intel_lr_context_unpin(struct intel_context *ctx, ctx->engine[engine->id].lrc_vma =NULL ; ctx->engine[engine->id].lrc_desc =0 ; ctx->engine[engine->id].lrc_reg_state =NULL ; +i915_gem_context_unreference (ctx); } }
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 2/3] drm/i915: Make LRC pinning own a reference to the context
Quoted 24 lines · Tvrtko Ursulin
From: Tvrtko Ursulin <[email protected]>
Will simplify the following fix and sounds logical.
Signed-off-by Tvrtko Ursulin <[email protected]>Cc Chris Wilson <[email protected]>Cc Nick Hoath <[email protected]>--- drivers/gpu/drm/i915/intel_lrc.c | 2 ++ 1 file changed, 2 insertions(+)drivers/gpu/drm/i915/intel_lrc.c +1
-0@@ -1105,6 +1105,7 @@ static int intel_lr_context_pin(struct intel_context *ctx, ret =intel_lr_context_do_pin (ctx, engine);if (ret) goto reset_pin_count; +i915_gem_context_reference (ctx); }return ret;
drivers/gpu/drm/i915/intel_lrc.c void intel_lr_context_unpin(struct intel_context *ctx, · line 1134
Quoted 4 lines
drivers/gpu/drm/i915/intel_lrc.c +1
-0@@ -1130,6 +1131,7 @@ void intel_lr_context_unpin(struct intel_context *ctx, ctx->engine[engine->id].lrc_vma =NULL ; ctx->engine[engine->id].lrc_desc =0 ; ctx->engine[engine->id].lrc_reg_state =NULL ; +i915_gem_context_unreference (ctx);
I would appreciate some whitespace, especially here, since this is important to be aware of -- we may need to consider that this is the last reference and so have it stand out.
-Chris
Signature · 5 lines
Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH 3/3 drm/i915: Fix premature LRC unpin in GuC mode
In reply to Tvrtko UrsulinFrom: Tvrtko Ursulin <[email protected]>
In GuC mode LRC pinning lifetime depends exclusively on the request liftime. Since that is terminated by the seqno update that opens up a race condition between GPU finishing writing out the context image and the driver unpinning the LRC.
To extend the LRC lifetime we will employ a similar approach to what legacy ringbuffer submission does.
We will start tracking the last submitted context per engine and keep it pinned until it is replaced by another one.
Note that the driver unload path is a bit fragile and could benefit greatly from efforts to unify the legacy and exec list submission code paths.
At the moment i915_gem_context_fini has special casing for the two which are potentialy not needed, and also depends on i915_gem_cleanup_ringbuffer running before itself.
Issue: VIZ-4277
Cc: Chris Wilson <[email protected]>
Cc: Nick Hoath <[email protected]>
---
I cannot test this with GuC but it passes BAT with execlists and some real world smoke tests.
--- drivers/gpu/drm/i915/i915_gem_context.c | 4 +++- drivers/gpu/drm/i915/intel_lrc.c | 7 +++++++ 2 files changed, 10 insertions(+), 1 deletion(-)
drivers/gpu/drm/i915/i915_gem_context.c +3 -1
@@ -438,7 +438,9 @@ void i915_gem_context_fini(struct drm_device *dev)for (i =0 ; i <I915_NUM_RINGS ; i++) {struct intel_engine_cs *ring = &dev_priv->ring[i]; -if (ring->last_context) +if (ring->last_context && i915.enable_execlists) +intel_lr_context_unpin (ring->last_context, ring); +else if (ring->last_context)i915_gem_context_unreference (ring->last_context); ring->default_context =NULL ;
drivers/gpu/drm/i915/intel_lrc.c +7 -0
@@ -918,6 +918,7 @@ int intel_execlists_submission(struct i915_execbuffer_params *params,struct intel_engine_cs *ring = params->ring;struct drm_i915_private *dev_priv = dev->dev_private;struct intel_ringbuffer *ringbuf = params->ctx->engine[ring->id].ringbuf; +struct intel_context *ctx = params->request->ctx;u64 exec_start;int instp_mode;u32 instp_mask;
@@ -982,6 +983,12 @@ int intel_execlists_submission(struct i915_execbuffer_params *params,trace_i915_gem_ring_dispatch (params->request, params->dispatch_flags); +if (ring->last_context && ring->last_context != ctx) { +intel_lr_context_unpin (ring->last_context, ring); +intel_lr_context_pin (ctx, ring); + ring->last_context = ctx; + } +i915_gem_execbuffer_move_to_active (vmas, params->request);i915_gem_execbuffer_retire_commands (params);
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 3/3] drm/i915: Fix premature LRC unpin in GuC mode
Quoted 61 lines · Tvrtko Ursulin
From: Tvrtko Ursulin <[email protected]>
In GuC mode LRC pinning lifetime depends exclusively on the request liftime. Since that is terminated by the seqno update that opens up a race condition between GPU finishing writing out the context image and the driver unpinning the LRC.
To extend the LRC lifetime we will employ a similar approach to what legacy ringbuffer submission does.
We will start tracking the last submitted context per engine and keep it pinned until it is replaced by another one.
Note that the driver unload path is a bit fragile and could benefit greatly from efforts to unify the legacy and exec list submission code paths.
At the moment
i915_gem_context_finihas special casing for the two which are potentialy not needed, and also depends oni915_gem_cleanup_ringbufferrunning before itself.Signed-off-by Tvrtko Ursulin <[email protected]>Issue: VIZ-4277
Cc: Chris Wilson <[email protected]>
Cc: Nick Hoath <[email protected]>
---
I cannot test this with GuC but it passes BAT with execlists and some real world smoke tests.--- drivers/gpu/drm/i915/i915_gem_context.c | 4 +++- drivers/gpu/drm/i915/intel_lrc.c | 7 +++++++ 2 files changed, 10 insertions(+), 1 deletion(-)drivers/gpu/drm/i915/i915_gem_context.c +3
-1@@ -438,7 +438,9 @@ void i915_gem_context_fini(struct drm_device *dev)for (i =0 ; i <I915_NUM_RINGS ; i++) {struct intel_engine_cs *ring = &dev_priv->ring[i]; -if (ring->last_context) +if (ring->last_context && i915.enable_execlists) +intel_lr_context_unpin (ring->last_context, ring); +else if (ring->last_context)i915_gem_context_unreference (ring->last_context); ring->default_context =NULL ;drivers/gpu/drm/i915/intel_lrc.c +1
-0@@ -918,6 +918,7 @@ int intel_execlists_submission(struct i915_execbuffer_params *params,struct intel_engine_cs *ring = params->ring;struct drm_i915_private *dev_priv = dev->dev_private;struct intel_ringbuffer *ringbuf = params->ctx->engine[ring->id].ringbuf; +struct intel_context *ctx = params->request->ctx;u64 exec_start;int instp_mode;u32 instp_mask;
drivers/gpu/drm/i915/intel_lrc.c int intel_execlists_submission(struct i915_execbuffer_params *params, · line 990
Quoted 8 lines
drivers/gpu/drm/i915/intel_lrc.c +5
-0@@ -982,6 +983,12 @@ int intel_execlists_submission(struct i915_execbuffer_params *params,trace_i915_gem_ring_dispatch (params->request, params->dispatch_flags); +if (ring->last_context && ring->last_context != ctx) { +intel_lr_context_unpin (ring->last_context, ring); +intel_lr_context_pin (ctx, ring); + ring->last_context = ctx; + }
I think this is the wrong location and should be part of submitting the context inside the engine (because intel_execlists_submission should not as it is entirely duplicating the common GEM batch submision code and the unique part is engine->add_request()).
Note that it should be:
if (engine->last_context != request->ctx) {
if (engine->last_context)
intel_lr_context_unpin(engine->last_context, engine);
engine->last_context = request->ctx;
intel_lr_context_pin(engine->last_context, engine);
}-Chris
Signature · 5 lines
Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 3/3] drm/i915: Fix premature LRC unpin in GuC mode
Quoted 65 lines · Chris Wilson
Quoted 64 lines · Tvrtko Ursulin
From: Tvrtko Ursulin <[email protected]>
In GuC mode LRC pinning lifetime depends exclusively on the request liftime. Since that is terminated by the seqno update that opens up a race condition between GPU finishing writing out the context image and the driver unpinning the LRC.
To extend the LRC lifetime we will employ a similar approach to what legacy ringbuffer submission does.
We will start tracking the last submitted context per engine and keep it pinned until it is replaced by another one.
Note that the driver unload path is a bit fragile and could benefit greatly from efforts to unify the legacy and exec list submission code paths.
At the moment
i915_gem_context_finihas special casing for the two which are potentialy not needed, and also depends oni915_gem_cleanup_ringbufferrunning before itself.Signed-off-by Tvrtko Ursulin <[email protected]>Issue: VIZ-4277
Cc: Chris Wilson <[email protected]>
Cc: Nick Hoath <[email protected]>
---
I cannot test this with GuC but it passes BAT with execlists and some real world smoke tests.--- drivers/gpu/drm/i915/i915_gem_context.c | 4 +++- drivers/gpu/drm/i915/intel_lrc.c | 7 +++++++ 2 files changed, 10 insertions(+), 1 deletion(-)drivers/gpu/drm/i915/i915_gem_context.c +3
-1@@ -438,7 +438,9 @@ void i915_gem_context_fini(struct drm_device *dev)for (i =0 ; i <I915_NUM_RINGS ; i++) {struct intel_engine_cs *ring = &dev_priv->ring[i]; -if (ring->last_context) +if (ring->last_context && i915.enable_execlists) +intel_lr_context_unpin (ring->last_context, ring); +else if (ring->last_context)i915_gem_context_unreference (ring->last_context); ring->default_context =NULL ;drivers/gpu/drm/i915/intel_lrc.c +1
-0@@ -918,6 +918,7 @@ int intel_execlists_submission(struct i915_execbuffer_params *params,struct intel_engine_cs *ring = params->ring;struct drm_i915_private *dev_priv = dev->dev_private;struct intel_ringbuffer *ringbuf = params->ctx->engine[ring->id].ringbuf; +struct intel_context *ctx = params->request->ctx;u64 exec_start;int instp_mode;u32 instp_mask;@@ -982,6 +983,12 @@ int intel_execlists_submission(struct i915_execbuffer_params *params,trace_i915_gem_ring_dispatch (params->request, params->dispatch_flags);
Quoted 10 lines
drivers/gpu/drm/i915/intel_lrc.c int intel_execlists_submission(struct i915_execbuffer_params *params, · line 990
Quoted 5 lines
drivers/gpu/drm/i915/intel_lrc.c +5
-0+if (ring->last_context && ring->last_context != ctx) { +intel_lr_context_unpin (ring->last_context, ring); +intel_lr_context_pin (ctx, ring); + ring->last_context = ctx; + }I think this is the wrong location and should be part of submitting the context inside the engine (because
intel_execlists_submissionshould not as it is entirely duplicating the common GEM batch submision code and the unique part isengine->add_request()).
So into engine->emit_request you are saying? That works just as well AFAICS, just making sure I understood correctly.
Quoted 8 lines
Note that it should be:
if (engine->last_context != request->ctx) { if (engine->last_context) intel_lr_context_unpin(engine->last_context, engine); engine->last_context = request->ctx; intel_lr_context_pin(engine->last_context, engine); }
Ooops!
Regards,
Tvrtko
_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 3/3] drm/i915: Fix premature LRC unpin in GuC mode
Quoted 78 lines · Tvrtko Ursulin
Quoted 65 lines · Chris Wilson
Quoted 64 lines · Tvrtko Ursulin
From: Tvrtko Ursulin <[email protected]>
In GuC mode LRC pinning lifetime depends exclusively on the request liftime. Since that is terminated by the seqno update that opens up a race condition between GPU finishing writing out the context image and the driver unpinning the LRC.
To extend the LRC lifetime we will employ a similar approach to what legacy ringbuffer submission does.
We will start tracking the last submitted context per engine and keep it pinned until it is replaced by another one.
Note that the driver unload path is a bit fragile and could benefit greatly from efforts to unify the legacy and exec list submission code paths.
At the moment
i915_gem_context_finihas special casing for the two which are potentialy not needed, and also depends oni915_gem_cleanup_ringbufferrunning before itself.Signed-off-by Tvrtko Ursulin <[email protected]>Issue: VIZ-4277
Cc: Chris Wilson <[email protected]>
Cc: Nick Hoath <[email protected]>
---
I cannot test this with GuC but it passes BAT with execlists and some real world smoke tests.--- drivers/gpu/drm/i915/i915_gem_context.c | 4 +++- drivers/gpu/drm/i915/intel_lrc.c | 7 +++++++ 2 files changed, 10 insertions(+), 1 deletion(-)drivers/gpu/drm/i915/i915_gem_context.c +3
-1@@ -438,7 +438,9 @@ void i915_gem_context_fini(struct drm_device *dev)for (i =0 ; i <I915_NUM_RINGS ; i++) {struct intel_engine_cs *ring = &dev_priv->ring[i]; -if (ring->last_context) +if (ring->last_context && i915.enable_execlists) +intel_lr_context_unpin (ring->last_context, ring); +else if (ring->last_context)i915_gem_context_unreference (ring->last_context); ring->default_context =NULL ;drivers/gpu/drm/i915/intel_lrc.c +1
-0@@ -918,6 +918,7 @@ int intel_execlists_submission(struct i915_execbuffer_params *params,struct intel_engine_cs *ring = params->ring;struct drm_i915_private *dev_priv = dev->dev_private;struct intel_ringbuffer *ringbuf = params->ctx->engine[ring->id].ringbuf; +struct intel_context *ctx = params->request->ctx;u64 exec_start;int instp_mode;u32 instp_mask;@@ -982,6 +983,12 @@ int intel_execlists_submission(struct i915_execbuffer_params *params,trace_i915_gem_ring_dispatch (params->request, params->dispatch_flags);Quoted 10 lines
drivers/gpu/drm/i915/intel_lrc.c int intel_execlists_submission(struct i915_execbuffer_params *params, · line 990
Quoted 5 lines
drivers/gpu/drm/i915/intel_lrc.c +5
-0+if (ring->last_context && ring->last_context != ctx) { +intel_lr_context_unpin (ring->last_context, ring); +intel_lr_context_pin (ctx, ring); + ring->last_context = ctx; + }I think this is the wrong location and should be part of submitting the context inside the engine (because
intel_execlists_submissionshould not as it is entirely duplicating the common GEM batch submision code and the unique part isengine->add_request()).
Quoted 2 lines
So into
engine->emit_requestyou are saying? That works just as well AFAICS, just making sure I understood correctly.
Oh, yeah you haven't got the unify add_request/emit_request patch...
So gen8_emit_request(), but considering the other patches in flight, intel_logical_ring_advance_and_submit() will be the ultimate location. -Chris
Signature · 5 lines
Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH v2 1/1 drm/i915: Fix premature LRC unpin in GuC mode
From: Tvrtko Ursulin <[email protected]>
In GuC mode LRC pinning lifetime depends exclusively on the request liftime. Since that is terminated by the seqno update that opens up a race condition between GPU finishing writing out the context image and the driver unpinning the LRC.
To extend the LRC lifetime we will employ a similar approach to what legacy ringbuffer submission does.
We will start tracking the last submitted context per engine and keep it pinned until it is replaced by another one.
Note that the driver unload path is a bit fragile and could benefit greatly from efforts to unify the legacy and exec list submission code paths.
At the moment i915_gem_context_fini has special casing for the two which are potentialy not needed, and also depends on i915_gem_cleanup_ringbuffer running before itself.
v2:
* Move pinning into engine->emit_request and actually fix the reference/unreference logic. (Chris Wilson)
* ring->dev can be NULL on driver unload so use a different route towards it.
Issue: VIZ-4277
Cc: Chris Wilson <[email protected]>
Cc: Nick Hoath <[email protected]>
---
I cannot test this with GuC but it passes BAT with execlists and some real world smoke tests.
--- drivers/gpu/drm/i915/i915_gem_context.c | 4 +++- drivers/gpu/drm/i915/intel_lrc.c | 9 ++++++++- 2 files changed, 11 insertions(+), 2 deletions(-)
drivers/gpu/drm/i915/i915_gem_context.c +3 -1
@@ -438,7 +438,9 @@ void i915_gem_context_fini(struct drm_device *dev)for (i =0 ; i <I915_NUM_RINGS ; i++) {struct intel_engine_cs *ring = &dev_priv->ring[i]; -if (ring->last_context) +if (ring->last_context && i915.enable_execlists) +intel_lr_context_unpin (ring->last_context, ring); +else if (ring->last_context)i915_gem_context_unreference (ring->last_context); ring->default_context =NULL ;
drivers/gpu/drm/i915/intel_lrc.c +8 -1
@@ -1120,7 +1120,7 @@ void intel_lr_context_unpin(struct intel_context *ctx, {struct drm_i915_gem_object *ctx_obj = ctx->engine[engine->id].state; -WARN_ON (!mutex_is_locked (&engine->dev->struct_mutex)); +WARN_ON (!mutex_is_locked (&ctx->i915->dev->struct_mutex));if (WARN_ON_ONCE (!ctx_obj))return ;
@@ -1859,6 +1859,13 @@ static int gen8_emit_request(struct drm_i915_gem_request *request)u32 cmd;int ret; +if (ring->last_context != request->ctx) { +if (ring->last_context) +intel_lr_context_unpin (ring->last_context, ring); +intel_lr_context_pin (request->ctx, ring); + ring->last_context = request->ctx; + } +/* * Reserve space for 2 NOOPs at the end of each request to be * used as a workaround for not being allowed to do lite
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH v2] drm/i915: Fix premature LRC unpin in GuC mode
Quoted 45 lines · Tvrtko Ursulin
From: Tvrtko Ursulin <[email protected]>
In GuC mode LRC pinning lifetime depends exclusively on the request liftime. Since that is terminated by the seqno update that opens up a race condition between GPU finishing writing out the context image and the driver unpinning the LRC.
To extend the LRC lifetime we will employ a similar approach to what legacy ringbuffer submission does.
We will start tracking the last submitted context per engine and keep it pinned until it is replaced by another one.
Note that the driver unload path is a bit fragile and could benefit greatly from efforts to unify the legacy and exec list submission code paths.
At the moment
i915_gem_context_finihas special casing for the two which are potentialy not needed, and also depends oni915_gem_cleanup_ringbufferrunning before itself.v2:
* Move pinning intoengine->emit_requestand actually fix the reference/unreference logic. (Chris Wilson)*
ring->devcan be NULL on driver unload so use a different route towards it.Signed-off-by Tvrtko Ursulin <[email protected]>Issue: VIZ-4277
Cc: Chris Wilson <[email protected]>
Cc: Nick Hoath <[email protected]>
---
I cannot test this with GuC but it passes BAT with execlists and some real world smoke tests.--- drivers/gpu/drm/i915/i915_gem_context.c | 4 +++- drivers/gpu/drm/i915/intel_lrc.c | 9 ++++++++- 2 files changed, 11 insertions(+), 2 deletions(-)drivers/gpu/drm/i915/i915_gem_context.c +0
-0
drivers/gpu/drm/i915/i915_gem_context.c void i915_gem_context_fini(struct drm_device *dev) · line 440
Quoted 3 lines
drivers/gpu/drm/i915/i915_gem_context.c +0
-0@@ -438,7 +438,9 @@ void i915_gem_context_fini(struct drm_device *dev)for (i =0 ; i <I915_NUM_RINGS ; i++) {struct intel_engine_cs *ring = &dev_priv->ring[i];
This is the nasty part where the GPU still has access to the backing pages as we release them. A hard to hit module-unload vs active GPU for sure, but it is something that we can prevent.
The context-fini vs engine-fini ordering is also apparently tied to the use of intel_gpu_reset() here (as it clobbers the GPU without dropping the GEM state, causing a foul up when tearing down the engine).
If we had actually called i915_reset (and so i915_gem_reset) instead, we should expect last_context to be NULL... Oops, Tvrtko you need to inspect i915_gem_context_reset() as well for reseting last_context back to NULL after the GPU reset.
As for here, I would just pencil in a plan to replace this chunk entirely with a call to i915_reset(), or a slightly trimmed down version and tidy up the active GEM cleanup in the process. -Chris
Signature · 5 lines
Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 3/3] drm/i915: Fix premature LRC unpin in GuC mode
In reply to Tvrtko UrsulinQuoted 78 lines · Tvrtko Ursulin
Quoted 65 lines · Chris Wilson
Quoted 64 lines · Tvrtko Ursulin
From: Tvrtko Ursulin <[email protected]>
In GuC mode LRC pinning lifetime depends exclusively on the request liftime. Since that is terminated by the seqno update that opens up a race condition between GPU finishing writing out the context image and the driver unpinning the LRC.
To extend the LRC lifetime we will employ a similar approach to what legacy ringbuffer submission does.
We will start tracking the last submitted context per engine and keep it pinned until it is replaced by another one.
Note that the driver unload path is a bit fragile and could benefit greatly from efforts to unify the legacy and exec list submission code paths.
At the moment
i915_gem_context_finihas special casing for the two which are potentialy not needed, and also depends oni915_gem_cleanup_ringbufferrunning before itself.Signed-off-by Tvrtko Ursulin <[email protected]>Issue: VIZ-4277
Cc: Chris Wilson <[email protected]>
Cc: Nick Hoath <[email protected]>
---
I cannot test this with GuC but it passes BAT with execlists and some real world smoke tests.--- drivers/gpu/drm/i915/i915_gem_context.c | 4 +++- drivers/gpu/drm/i915/intel_lrc.c | 7 +++++++ 2 files changed, 10 insertions(+), 1 deletion(-)drivers/gpu/drm/i915/i915_gem_context.c +3
-1@@ -438,7 +438,9 @@ void i915_gem_context_fini(struct drm_device *dev)for (i =0 ; i <I915_NUM_RINGS ; i++) {struct intel_engine_cs *ring = &dev_priv->ring[i]; -if (ring->last_context) +if (ring->last_context && i915.enable_execlists) +intel_lr_context_unpin (ring->last_context, ring); +else if (ring->last_context)i915_gem_context_unreference (ring->last_context); ring->default_context =NULL ;drivers/gpu/drm/i915/intel_lrc.c +1
-0@@ -918,6 +918,7 @@ int intel_execlists_submission(struct i915_execbuffer_params *params,struct intel_engine_cs *ring = params->ring;struct drm_i915_private *dev_priv = dev->dev_private;struct intel_ringbuffer *ringbuf = params->ctx->engine[ring->id].ringbuf; +struct intel_context *ctx = params->request->ctx;u64 exec_start;int instp_mode;u32 instp_mask;@@ -982,6 +983,12 @@ int intel_execlists_submission(struct i915_execbuffer_params *params,trace_i915_gem_ring_dispatch (params->request, params->dispatch_flags);Quoted 10 lines
drivers/gpu/drm/i915/intel_lrc.c int intel_execlists_submission(struct i915_execbuffer_params *params, · line 990
Quoted 5 lines
drivers/gpu/drm/i915/intel_lrc.c +5
-0+if (ring->last_context && ring->last_context != ctx) { +intel_lr_context_unpin (ring->last_context, ring); +intel_lr_context_pin (ctx, ring); + ring->last_context = ctx; + }I think this is the wrong location and should be part of submitting the context inside the engine (because
intel_execlists_submissionshould not as it is entirely duplicating the common GEM batch submision code and the unique part isengine->add_request()).
Quoted 2 lines
So into
engine->emit_requestyou are saying? That works just as well AFAICS, just making sure I understood correctly.
I think it should go in to intel_logical_ring_advance_and_submit. The extra pinning is being put in place to cover GPU usage of the pin. It should probably therefore go in to the last common place between execlists & GUC, as close to hardware submission as possible.
Quoted 10 lines
Quoted 9 lines
Note that it should be:
if (engine->last_context != request->ctx) { if (engine->last_context) intel_lr_context_unpin(engine->last_context, engine); engine->last_context = request->ctx; intel_lr_context_pin(engine->last_context, engine); }
Quoted 6 lines
Ooops!
Regards,
Tvrtko
_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 11 lines · Tvrtko Ursulin
From: Tvrtko Ursulin <[email protected]>
Previously intel_lr_context_(un)pin were operating on requests which is in conflict with their names.
If we make them take a context and an engine, it makes the names make more sense and it also makes future fixes possible.
Signed-off-by Tvrtko Ursulin <[email protected]>Cc Chris Wilson <[email protected]>Cc Nick Hoath <[email protected]>
Reviewed-by Chris Wilson <[email protected]> -Chris
Signature · 5 lines
Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Patchwork
CI✓ Fi.CI.BAT: success for series starting with [1/3] drm/i915: Make LRC (un)pinning work on context and engine
In reply to Tvrtko Ursulin== Summary ==
== Summary ==
Built on e9545b0b60b73d8be3d41048af5b8f2c1e2fc4c1 drm-intel-nightly: 2016y-01m-20d-13h-55m-37s UTC integration manifest
Test kms_pipe_crc_basic:
Subgroup read-crc-pipe-a-frame-sequence:
dmesg-warn -> PASS (skl-i5k-2)bdw-nuci7 total:143 pass:134 dwarn:0 dfail:0 fail:0 skip:9 bdw-ultra total:143 pass:135 dwarn:1 dfail:1 fail:0 skip:6 bsw-nuc-2 total:143 pass:117 dwarn:2 dfail:0 fail:0 skip:24 byt-nuc total:143 pass:125 dwarn:3 dfail:0 fail:0 skip:15 hsw-brixbox total:143 pass:136 dwarn:0 dfail:0 fail:0 skip:7 hsw-gt2 total:143 pass:139 dwarn:0 dfail:0 fail:0 skip:4 ilk-hp8440p total:143 pass:104 dwarn:1 dfail:0 fail:0 skip:38 ivb-t430s total:143 pass:137 dwarn:0 dfail:0 fail:0 skip:6 skl-i5k-2 total:143 pass:134 dwarn:1 dfail:0 fail:0 skip:8 skl-i7k-2 total:143 pass:134 dwarn:1 dfail:0 fail:0 skip:8 snb-dellxps total:143 pass:129 dwarn:0 dfail:0 fail:0 skip:14 snb-x220t total:143 pass:129 dwarn:0 dfail:0 fail:1 skip:13
Results at /archive/results/CI_IGT_test/Patchwork_1232/
_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx