dri-devel.lists.freedesktop.org archive mirror
 help / color / mirror / Atom feed
* [PATCH v2] drm/sched: Fix racy access to drm_sched_entity.dependency
@ 2025-09-01 12:40 Pierre-Eric Pelloux-Prayer
  2025-09-02  7:27 ` Philipp Stanner
  0 siblings, 1 reply; 6+ messages in thread
From: Pierre-Eric Pelloux-Prayer @ 2025-09-01 12:40 UTC (permalink / raw)
  To: Matthew Brost, Danilo Krummrich, Philipp Stanner,
	Christian König, Maarten Lankhorst, Maxime Ripard,
	Thomas Zimmermann, David Airlie, Simona Vetter, Tvrtko Ursulin,
	Pierre-Eric Pelloux-Prayer
  Cc: dri-devel, linux-kernel

The drm_sched_job_unschedulable trace point can access
entity->dependency after it was cleared by the callback
installed in drm_sched_entity_add_dependency_cb, causing:

BUG: kernel NULL pointer dereference, address: 0000000000000020
[...]
Workqueue: comp_1.1.0 drm_sched_run_job_work [gpu_sched]
RIP: 0010:trace_event_raw_event_drm_sched_job_unschedulable+0x70/0xd0 [gpu_sched]

To fix this we either need to keep a reference to the fence before
setting up the callbacks, or move the trace_drm_sched_job_unschedulable
calls into drm_sched_entity_add_dependency_cb where they can be
done earlier.

Fixes: 76d97c870f29 ("drm/sched: Trace dependencies for GPU jobs")

Signed-off-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
---
 drivers/gpu/drm/scheduler/sched_entity.c | 11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/scheduler/sched_entity.c b/drivers/gpu/drm/scheduler/sched_entity.c
index 8867b95ab089..3d06f72531ba 100644
--- a/drivers/gpu/drm/scheduler/sched_entity.c
+++ b/drivers/gpu/drm/scheduler/sched_entity.c
@@ -391,7 +391,8 @@ EXPORT_SYMBOL(drm_sched_entity_set_priority);
  * Add a callback to the current dependency of the entity to wake up the
  * scheduler when the entity becomes available.
  */
-static bool drm_sched_entity_add_dependency_cb(struct drm_sched_entity *entity)
+static bool drm_sched_entity_add_dependency_cb(struct drm_sched_entity *entity,
+					       struct drm_sched_job *sched_job)
 {
 	struct drm_gpu_scheduler *sched = entity->rq->sched;
 	struct dma_fence *fence = entity->dependency;
@@ -421,6 +422,10 @@ static bool drm_sched_entity_add_dependency_cb(struct drm_sched_entity *entity)
 		entity->dependency = fence;
 	}
 
+	if (trace_drm_sched_job_unschedulable_enabled() &&
+	    !test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &entity->dependency->flags))
+		trace_drm_sched_job_unschedulable(sched_job, entity->dependency);
+
 	if (!dma_fence_add_callback(entity->dependency, &entity->cb,
 				    drm_sched_entity_wakeup))
 		return true;
@@ -461,10 +466,8 @@ struct drm_sched_job *drm_sched_entity_pop_job(struct drm_sched_entity *entity)
 
 	while ((entity->dependency =
 			drm_sched_job_dependency(sched_job, entity))) {
-		if (drm_sched_entity_add_dependency_cb(entity)) {
-			trace_drm_sched_job_unschedulable(sched_job, entity->dependency);
+		if (drm_sched_entity_add_dependency_cb(entity, sched_job))
 			return NULL;
-		}
 	}
 
 	/* skip jobs from entity that marked guilty */
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 6+ messages in thread

* Re: [PATCH v2] drm/sched: Fix racy access to drm_sched_entity.dependency
  2025-09-01 12:40 [PATCH v2] drm/sched: Fix racy access to drm_sched_entity.dependency Pierre-Eric Pelloux-Prayer
@ 2025-09-02  7:27 ` Philipp Stanner
  2025-09-02  7:59   ` Tvrtko Ursulin
  0 siblings, 1 reply; 6+ messages in thread
From: Philipp Stanner @ 2025-09-02  7:27 UTC (permalink / raw)
  To: Pierre-Eric Pelloux-Prayer, Matthew Brost, Danilo Krummrich,
	Philipp Stanner, Christian König, Maarten Lankhorst,
	Maxime Ripard, Thomas Zimmermann, David Airlie, Simona Vetter,
	Tvrtko Ursulin
  Cc: dri-devel, linux-kernel

On Mon, 2025-09-01 at 14:40 +0200, Pierre-Eric Pelloux-Prayer wrote:
> The drm_sched_job_unschedulable trace point can access
> entity->dependency after it was cleared by the callback
> installed in drm_sched_entity_add_dependency_cb, causing:
> 
> BUG: kernel NULL pointer dereference, address: 0000000000000020
> [...]
> Workqueue: comp_1.1.0 drm_sched_run_job_work [gpu_sched]
> RIP: 0010:trace_event_raw_event_drm_sched_job_unschedulable+0x70/0xd0 [gpu_sched]
> 
> To fix this we either need to keep a reference to the fence before
> setting up the callbacks, or move the trace_drm_sched_job_unschedulable
> calls into drm_sched_entity_add_dependency_cb where they can be
> done earlier.
> 
> Fixes: 76d97c870f29 ("drm/sched: Trace dependencies for GPU jobs")
> 
> Signed-off-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>

Applied to drm-misc-next

Thx
P.

> ---
>  drivers/gpu/drm/scheduler/sched_entity.c | 11 +++++++----
>  1 file changed, 7 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/gpu/drm/scheduler/sched_entity.c b/drivers/gpu/drm/scheduler/sched_entity.c
> index 8867b95ab089..3d06f72531ba 100644
> --- a/drivers/gpu/drm/scheduler/sched_entity.c
> +++ b/drivers/gpu/drm/scheduler/sched_entity.c
> @@ -391,7 +391,8 @@ EXPORT_SYMBOL(drm_sched_entity_set_priority);
>   * Add a callback to the current dependency of the entity to wake up the
>   * scheduler when the entity becomes available.
>   */
> -static bool drm_sched_entity_add_dependency_cb(struct drm_sched_entity *entity)
> +static bool drm_sched_entity_add_dependency_cb(struct drm_sched_entity *entity,
> +					       struct drm_sched_job *sched_job)
>  {
>  	struct drm_gpu_scheduler *sched = entity->rq->sched;
>  	struct dma_fence *fence = entity->dependency;
> @@ -421,6 +422,10 @@ static bool drm_sched_entity_add_dependency_cb(struct drm_sched_entity *entity)
>  		entity->dependency = fence;
>  	}
>  
> +	if (trace_drm_sched_job_unschedulable_enabled() &&
> +	    !test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &entity->dependency->flags))
> +		trace_drm_sched_job_unschedulable(sched_job, entity->dependency);
> +
>  	if (!dma_fence_add_callback(entity->dependency, &entity->cb,
>  				    drm_sched_entity_wakeup))
>  		return true;
> @@ -461,10 +466,8 @@ struct drm_sched_job *drm_sched_entity_pop_job(struct drm_sched_entity *entity)
>  
>  	while ((entity->dependency =
>  			drm_sched_job_dependency(sched_job, entity))) {
> -		if (drm_sched_entity_add_dependency_cb(entity)) {
> -			trace_drm_sched_job_unschedulable(sched_job, entity->dependency);
> +		if (drm_sched_entity_add_dependency_cb(entity, sched_job))
>  			return NULL;
> -		}
>  	}
>  
>  	/* skip jobs from entity that marked guilty */


^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [PATCH v2] drm/sched: Fix racy access to drm_sched_entity.dependency
  2025-09-02  7:27 ` Philipp Stanner
@ 2025-09-02  7:59   ` Tvrtko Ursulin
  2025-09-02  8:18     ` Philipp Stanner
  0 siblings, 1 reply; 6+ messages in thread
From: Tvrtko Ursulin @ 2025-09-02  7:59 UTC (permalink / raw)
  To: phasta, Pierre-Eric Pelloux-Prayer, Matthew Brost,
	Danilo Krummrich, Christian König, Maarten Lankhorst,
	Maxime Ripard, Thomas Zimmermann, David Airlie, Simona Vetter
  Cc: dri-devel, linux-kernel


On 02/09/2025 08:27, Philipp Stanner wrote:
> On Mon, 2025-09-01 at 14:40 +0200, Pierre-Eric Pelloux-Prayer wrote:
>> The drm_sched_job_unschedulable trace point can access
>> entity->dependency after it was cleared by the callback
>> installed in drm_sched_entity_add_dependency_cb, causing:
>>
>> BUG: kernel NULL pointer dereference, address: 0000000000000020
>> [...]
>> Workqueue: comp_1.1.0 drm_sched_run_job_work [gpu_sched]
>> RIP: 0010:trace_event_raw_event_drm_sched_job_unschedulable+0x70/0xd0 [gpu_sched]
>>
>> To fix this we either need to keep a reference to the fence before
>> setting up the callbacks, or move the trace_drm_sched_job_unschedulable
>> calls into drm_sched_entity_add_dependency_cb where they can be
>> done earlier.
>>
>> Fixes: 76d97c870f29 ("drm/sched: Trace dependencies for GPU jobs")
>>
>> Signed-off-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
>> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
> 
> Applied to drm-misc-next

Shouldn't it have been drm-misc-fixes?

Regards,

Tvrtko

>> ---
>>   drivers/gpu/drm/scheduler/sched_entity.c | 11 +++++++----
>>   1 file changed, 7 insertions(+), 4 deletions(-)
>>
>> diff --git a/drivers/gpu/drm/scheduler/sched_entity.c b/drivers/gpu/drm/scheduler/sched_entity.c
>> index 8867b95ab089..3d06f72531ba 100644
>> --- a/drivers/gpu/drm/scheduler/sched_entity.c
>> +++ b/drivers/gpu/drm/scheduler/sched_entity.c
>> @@ -391,7 +391,8 @@ EXPORT_SYMBOL(drm_sched_entity_set_priority);
>>    * Add a callback to the current dependency of the entity to wake up the
>>    * scheduler when the entity becomes available.
>>    */
>> -static bool drm_sched_entity_add_dependency_cb(struct drm_sched_entity *entity)
>> +static bool drm_sched_entity_add_dependency_cb(struct drm_sched_entity *entity,
>> +					       struct drm_sched_job *sched_job)
>>   {
>>   	struct drm_gpu_scheduler *sched = entity->rq->sched;
>>   	struct dma_fence *fence = entity->dependency;
>> @@ -421,6 +422,10 @@ static bool drm_sched_entity_add_dependency_cb(struct drm_sched_entity *entity)
>>   		entity->dependency = fence;
>>   	}
>>   
>> +	if (trace_drm_sched_job_unschedulable_enabled() &&
>> +	    !test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &entity->dependency->flags))
>> +		trace_drm_sched_job_unschedulable(sched_job, entity->dependency);
>> +
>>   	if (!dma_fence_add_callback(entity->dependency, &entity->cb,
>>   				    drm_sched_entity_wakeup))
>>   		return true;
>> @@ -461,10 +466,8 @@ struct drm_sched_job *drm_sched_entity_pop_job(struct drm_sched_entity *entity)
>>   
>>   	while ((entity->dependency =
>>   			drm_sched_job_dependency(sched_job, entity))) {
>> -		if (drm_sched_entity_add_dependency_cb(entity)) {
>> -			trace_drm_sched_job_unschedulable(sched_job, entity->dependency);
>> +		if (drm_sched_entity_add_dependency_cb(entity, sched_job))
>>   			return NULL;
>> -		}
>>   	}
>>   
>>   	/* skip jobs from entity that marked guilty */
> 


^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [PATCH v2] drm/sched: Fix racy access to drm_sched_entity.dependency
  2025-09-02  7:59   ` Tvrtko Ursulin
@ 2025-09-02  8:18     ` Philipp Stanner
  2025-09-02  8:22       ` Philipp Stanner
  0 siblings, 1 reply; 6+ messages in thread
From: Philipp Stanner @ 2025-09-02  8:18 UTC (permalink / raw)
  To: Tvrtko Ursulin, phasta, Pierre-Eric Pelloux-Prayer, Matthew Brost,
	Danilo Krummrich, Christian König, Maarten Lankhorst,
	Maxime Ripard, Thomas Zimmermann, David Airlie, Simona Vetter
  Cc: dri-devel, linux-kernel

On Tue, 2025-09-02 at 08:59 +0100, Tvrtko Ursulin wrote:
> 
> On 02/09/2025 08:27, Philipp Stanner wrote:
> > On Mon, 2025-09-01 at 14:40 +0200, Pierre-Eric Pelloux-Prayer wrote:
> > > The drm_sched_job_unschedulable trace point can access
> > > entity->dependency after it was cleared by the callback
> > > installed in drm_sched_entity_add_dependency_cb, causing:
> > > 
> > > BUG: kernel NULL pointer dereference, address: 0000000000000020
> > > [...]
> > > Workqueue: comp_1.1.0 drm_sched_run_job_work [gpu_sched]
> > > RIP: 0010:trace_event_raw_event_drm_sched_job_unschedulable+0x70/0xd0 [gpu_sched]
> > > 
> > > To fix this we either need to keep a reference to the fence before
> > > setting up the callbacks, or move the trace_drm_sched_job_unschedulable
> > > calls into drm_sched_entity_add_dependency_cb where they can be
> > > done earlier.
> > > 
> > > Fixes: 76d97c870f29 ("drm/sched: Trace dependencies for GPU jobs")
> > > 
> > > Signed-off-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
> > > Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
> > 
> > Applied to drm-misc-next
> 
> Shouldn't it have been drm-misc-fixes?

I considered that, but thought not: the fixed commit is only in this rc
(v6.17), and in this case the committer guidelines say it should go to
misc-next:

https://drm.pages.freedesktop.org/maintainer-tools/committer/committer-drm-misc.html

Same reason we don't need to +Cc stable.

But correct me if I made a mistake.

P.


> 
> Regards,
> 
> Tvrtko
> 
> > > ---
> > >   drivers/gpu/drm/scheduler/sched_entity.c | 11 +++++++----
> > >   1 file changed, 7 insertions(+), 4 deletions(-)
> > > 
> > > diff --git a/drivers/gpu/drm/scheduler/sched_entity.c b/drivers/gpu/drm/scheduler/sched_entity.c
> > > index 8867b95ab089..3d06f72531ba 100644
> > > --- a/drivers/gpu/drm/scheduler/sched_entity.c
> > > +++ b/drivers/gpu/drm/scheduler/sched_entity.c
> > > @@ -391,7 +391,8 @@ EXPORT_SYMBOL(drm_sched_entity_set_priority);
> > >    * Add a callback to the current dependency of the entity to wake up the
> > >    * scheduler when the entity becomes available.
> > >    */
> > > -static bool drm_sched_entity_add_dependency_cb(struct drm_sched_entity *entity)
> > > +static bool drm_sched_entity_add_dependency_cb(struct drm_sched_entity *entity,
> > > +					       struct drm_sched_job *sched_job)
> > >   {
> > >   	struct drm_gpu_scheduler *sched = entity->rq->sched;
> > >   	struct dma_fence *fence = entity->dependency;
> > > @@ -421,6 +422,10 @@ static bool drm_sched_entity_add_dependency_cb(struct drm_sched_entity *entity)
> > >   		entity->dependency = fence;
> > >   	}
> > >   
> > > +	if (trace_drm_sched_job_unschedulable_enabled() &&
> > > +	    !test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &entity->dependency->flags))
> > > +		trace_drm_sched_job_unschedulable(sched_job, entity->dependency);
> > > +
> > >   	if (!dma_fence_add_callback(entity->dependency, &entity->cb,
> > >   				    drm_sched_entity_wakeup))
> > >   		return true;
> > > @@ -461,10 +466,8 @@ struct drm_sched_job *drm_sched_entity_pop_job(struct drm_sched_entity *entity)
> > >   
> > >   	while ((entity->dependency =
> > >   			drm_sched_job_dependency(sched_job, entity))) {
> > > -		if (drm_sched_entity_add_dependency_cb(entity)) {
> > > -			trace_drm_sched_job_unschedulable(sched_job, entity->dependency);
> > > +		if (drm_sched_entity_add_dependency_cb(entity, sched_job))
> > >   			return NULL;
> > > -		}
> > >   	}
> > >   
> > >   	/* skip jobs from entity that marked guilty */
> > 
> 


^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [PATCH v2] drm/sched: Fix racy access to drm_sched_entity.dependency
  2025-09-02  8:18     ` Philipp Stanner
@ 2025-09-02  8:22       ` Philipp Stanner
  2025-09-02 11:54         ` Philipp Stanner
  0 siblings, 1 reply; 6+ messages in thread
From: Philipp Stanner @ 2025-09-02  8:22 UTC (permalink / raw)
  To: phasta, Tvrtko Ursulin, Pierre-Eric Pelloux-Prayer, Matthew Brost,
	Danilo Krummrich, Christian König, Maarten Lankhorst,
	Maxime Ripard, Thomas Zimmermann, David Airlie, Simona Vetter
  Cc: dri-devel, linux-kernel

On Tue, 2025-09-02 at 10:18 +0200, Philipp Stanner wrote:
> On Tue, 2025-09-02 at 08:59 +0100, Tvrtko Ursulin wrote:
> > 
> > On 02/09/2025 08:27, Philipp Stanner wrote:
> > > On Mon, 2025-09-01 at 14:40 +0200, Pierre-Eric Pelloux-Prayer wrote:
> > > > The drm_sched_job_unschedulable trace point can access
> > > > entity->dependency after it was cleared by the callback
> > > > installed in drm_sched_entity_add_dependency_cb, causing:
> > > > 
> > > > BUG: kernel NULL pointer dereference, address: 0000000000000020
> > > > [...]
> > > > Workqueue: comp_1.1.0 drm_sched_run_job_work [gpu_sched]
> > > > RIP: 0010:trace_event_raw_event_drm_sched_job_unschedulable+0x70/0xd0 [gpu_sched]
> > > > 
> > > > To fix this we either need to keep a reference to the fence before
> > > > setting up the callbacks, or move the trace_drm_sched_job_unschedulable
> > > > calls into drm_sched_entity_add_dependency_cb where they can be
> > > > done earlier.
> > > > 
> > > > Fixes: 76d97c870f29 ("drm/sched: Trace dependencies for GPU jobs")
> > > > 
> > > > Signed-off-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
> > > > Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
> > > 
> > > Applied to drm-misc-next
> > 
> > Shouldn't it have been drm-misc-fixes?
> 
> I considered that, but thought not: the fixed commit is only in this rc
> (v6.17), and in this case the committer guidelines say it should go to
> misc-next:
> 
> https://drm.pages.freedesktop.org/maintainer-tools/committer/committer-drm-misc.html

OK, wait, the opposite is the case. I slipped to the wrong branch in
the diagram.

Oh dear. My coffee machine broke a few days ago, and it shows.

My bad, sorry. Let me ping the drm maintainers to sort that one out.

P.



> 
> Same reason we don't need to +Cc stable.
> 
> But correct me if I made a mistake.
> 
> P.
> 
> 
> > 
> > Regards,
> > 
> > Tvrtko
> > 
> > > > ---
> > > >   drivers/gpu/drm/scheduler/sched_entity.c | 11 +++++++----
> > > >   1 file changed, 7 insertions(+), 4 deletions(-)
> > > > 
> > > > diff --git a/drivers/gpu/drm/scheduler/sched_entity.c b/drivers/gpu/drm/scheduler/sched_entity.c
> > > > index 8867b95ab089..3d06f72531ba 100644
> > > > --- a/drivers/gpu/drm/scheduler/sched_entity.c
> > > > +++ b/drivers/gpu/drm/scheduler/sched_entity.c
> > > > @@ -391,7 +391,8 @@ EXPORT_SYMBOL(drm_sched_entity_set_priority);
> > > >    * Add a callback to the current dependency of the entity to wake up the
> > > >    * scheduler when the entity becomes available.
> > > >    */
> > > > -static bool drm_sched_entity_add_dependency_cb(struct drm_sched_entity *entity)
> > > > +static bool drm_sched_entity_add_dependency_cb(struct drm_sched_entity *entity,
> > > > +					       struct drm_sched_job *sched_job)
> > > >   {
> > > >   	struct drm_gpu_scheduler *sched = entity->rq->sched;
> > > >   	struct dma_fence *fence = entity->dependency;
> > > > @@ -421,6 +422,10 @@ static bool drm_sched_entity_add_dependency_cb(struct drm_sched_entity *entity)
> > > >   		entity->dependency = fence;
> > > >   	}
> > > >   
> > > > +	if (trace_drm_sched_job_unschedulable_enabled() &&
> > > > +	    !test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &entity->dependency->flags))
> > > > +		trace_drm_sched_job_unschedulable(sched_job, entity->dependency);
> > > > +
> > > >   	if (!dma_fence_add_callback(entity->dependency, &entity->cb,
> > > >   				    drm_sched_entity_wakeup))
> > > >   		return true;
> > > > @@ -461,10 +466,8 @@ struct drm_sched_job *drm_sched_entity_pop_job(struct drm_sched_entity *entity)
> > > >   
> > > >   	while ((entity->dependency =
> > > >   			drm_sched_job_dependency(sched_job, entity))) {
> > > > -		if (drm_sched_entity_add_dependency_cb(entity)) {
> > > > -			trace_drm_sched_job_unschedulable(sched_job, entity->dependency);
> > > > +		if (drm_sched_entity_add_dependency_cb(entity, sched_job))
> > > >   			return NULL;
> > > > -		}
> > > >   	}
> > > >   
> > > >   	/* skip jobs from entity that marked guilty */
> > > 
> > 
> 


^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [PATCH v2] drm/sched: Fix racy access to drm_sched_entity.dependency
  2025-09-02  8:22       ` Philipp Stanner
@ 2025-09-02 11:54         ` Philipp Stanner
  0 siblings, 0 replies; 6+ messages in thread
From: Philipp Stanner @ 2025-09-02 11:54 UTC (permalink / raw)
  To: phasta, Tvrtko Ursulin, Pierre-Eric Pelloux-Prayer, Matthew Brost,
	Danilo Krummrich, Christian König, Maarten Lankhorst,
	Maxime Ripard, Thomas Zimmermann, David Airlie, Simona Vetter
  Cc: dri-devel, linux-kernel

On Tue, 2025-09-02 at 10:22 +0200, Philipp Stanner wrote:
> On Tue, 2025-09-02 at 10:18 +0200, Philipp Stanner wrote:
> > On Tue, 2025-09-02 at 08:59 +0100, Tvrtko Ursulin wrote:
> > > 
> > > On 02/09/2025 08:27, Philipp Stanner wrote:
> > > > On Mon, 2025-09-01 at 14:40 +0200, Pierre-Eric Pelloux-Prayer wrote:
> > > > > The drm_sched_job_unschedulable trace point can access
> > > > > entity->dependency after it was cleared by the callback
> > > > > installed in drm_sched_entity_add_dependency_cb, causing:
> > > > > 
> > > > > BUG: kernel NULL pointer dereference, address: 0000000000000020
> > > > > [...]
> > > > > Workqueue: comp_1.1.0 drm_sched_run_job_work [gpu_sched]
> > > > > RIP: 0010:trace_event_raw_event_drm_sched_job_unschedulable+0x70/0xd0 [gpu_sched]
> > > > > 
> > > > > To fix this we either need to keep a reference to the fence before
> > > > > setting up the callbacks, or move the trace_drm_sched_job_unschedulable
> > > > > calls into drm_sched_entity_add_dependency_cb where they can be
> > > > > done earlier.
> > > > > 
> > > > > Fixes: 76d97c870f29 ("drm/sched: Trace dependencies for GPU jobs")
> > > > > 
> > > > > Signed-off-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
> > > > > Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
> > > > 
> > > > Applied to drm-misc-next
> > > 
> > > Shouldn't it have been drm-misc-fixes?
> > 
> > I considered that, but thought not: the fixed commit is only in this rc
> > (v6.17), and in this case the committer guidelines say it should go to
> > misc-next:
> > 
> > https://drm.pages.freedesktop.org/maintainer-tools/committer/committer-drm-misc.html
> 
> OK, wait, the opposite is the case. I slipped to the wrong branch in
> the diagram.
> 
> Oh dear. My coffee machine broke a few days ago, and it shows.
> 
> My bad, sorry. Let me ping the drm maintainers to sort that one out.

Maxime was so kind to fix this for me.
Merci, Maxime, and sorry guys.

Is now on drm-misc-fixes and we're good.

P.

> 
> P.
> 
> 
> 
> > 
> > Same reason we don't need to +Cc stable.
> > 
> > But correct me if I made a mistake.
> > 
> > P.
> > 
> > 
> > > 
> > > Regards,
> > > 
> > > Tvrtko
> > > 
> > > > > ---
> > > > >   drivers/gpu/drm/scheduler/sched_entity.c | 11 +++++++----
> > > > >   1 file changed, 7 insertions(+), 4 deletions(-)
> > > > > 
> > > > > diff --git a/drivers/gpu/drm/scheduler/sched_entity.c b/drivers/gpu/drm/scheduler/sched_entity.c
> > > > > index 8867b95ab089..3d06f72531ba 100644
> > > > > --- a/drivers/gpu/drm/scheduler/sched_entity.c
> > > > > +++ b/drivers/gpu/drm/scheduler/sched_entity.c
> > > > > @@ -391,7 +391,8 @@ EXPORT_SYMBOL(drm_sched_entity_set_priority);
> > > > >    * Add a callback to the current dependency of the entity to wake up the
> > > > >    * scheduler when the entity becomes available.
> > > > >    */
> > > > > -static bool drm_sched_entity_add_dependency_cb(struct drm_sched_entity *entity)
> > > > > +static bool drm_sched_entity_add_dependency_cb(struct drm_sched_entity *entity,
> > > > > +					       struct drm_sched_job *sched_job)
> > > > >   {
> > > > >   	struct drm_gpu_scheduler *sched = entity->rq->sched;
> > > > >   	struct dma_fence *fence = entity->dependency;
> > > > > @@ -421,6 +422,10 @@ static bool drm_sched_entity_add_dependency_cb(struct drm_sched_entity *entity)
> > > > >   		entity->dependency = fence;
> > > > >   	}
> > > > >   
> > > > > +	if (trace_drm_sched_job_unschedulable_enabled() &&
> > > > > +	    !test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &entity->dependency->flags))
> > > > > +		trace_drm_sched_job_unschedulable(sched_job, entity->dependency);
> > > > > +
> > > > >   	if (!dma_fence_add_callback(entity->dependency, &entity->cb,
> > > > >   				    drm_sched_entity_wakeup))
> > > > >   		return true;
> > > > > @@ -461,10 +466,8 @@ struct drm_sched_job *drm_sched_entity_pop_job(struct drm_sched_entity *entity)
> > > > >   
> > > > >   	while ((entity->dependency =
> > > > >   			drm_sched_job_dependency(sched_job, entity))) {
> > > > > -		if (drm_sched_entity_add_dependency_cb(entity)) {
> > > > > -			trace_drm_sched_job_unschedulable(sched_job, entity->dependency);
> > > > > +		if (drm_sched_entity_add_dependency_cb(entity, sched_job))
> > > > >   			return NULL;
> > > > > -		}
> > > > >   	}
> > > > >   
> > > > >   	/* skip jobs from entity that marked guilty */
> > > > 
> > > 
> > 
> 


^ permalink raw reply	[flat|nested] 6+ messages in thread

end of thread, other threads:[~2025-09-02 11:54 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-09-01 12:40 [PATCH v2] drm/sched: Fix racy access to drm_sched_entity.dependency Pierre-Eric Pelloux-Prayer
2025-09-02  7:27 ` Philipp Stanner
2025-09-02  7:59   ` Tvrtko Ursulin
2025-09-02  8:18     ` Philipp Stanner
2025-09-02  8:22       ` Philipp Stanner
2025-09-02 11:54         ` Philipp Stanner

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).