[PATCH 1/3] drm/sched: Add callback to mark if sched is ready to work.

[Michel Dänzer]

On 2018-10-18 8:44 p.m., Andrey Grodzovsky wrote:
> Problem:
> A particular scheduler may become unsuable (underlying HW) after
> some event (e.g. GPU reset). If it's later chosen by
> the get free sched. policy a command will fail to be
> submitted.
> 
> Fix:
> Add a driver specific callback to report the sced. status so
> rq with bad sched. can be avoided in favor of working one or
> none in which case job init will fail.
> 
> Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky at amd.com>
> 
> [...]
>  
> @@ -450,7 +457,7 @@ int amdgpu_fence_driver_init_ring(struct amdgpu_ring *ring,
>  
>  		r = drm_sched_init(&ring->sched, &amdgpu_sched_ops,
>  				   num_hw_submission, amdgpu_job_hang_limit,
> -				   timeout, ring->name);
> +				   timeout, ring->name, amdgpu_ring_ready);
>  		if (r) {
>  			DRM_ERROR("Failed to create scheduler on ring %s.\n",
>  				  ring->name);
> diff --git a/drivers/gpu/drm/etnaviv/etnaviv_sched.c b/drivers/gpu/drm/etnaviv/etnaviv_sched.c
> index f8c5f1e..5094013 100644
> --- a/drivers/gpu/drm/etnaviv/etnaviv_sched.c
> +++ b/drivers/gpu/drm/etnaviv/etnaviv_sched.c
> @@ -178,7 +178,8 @@ int etnaviv_sched_init(struct etnaviv_gpu *gpu)
>  
>  	ret = drm_sched_init(&gpu->sched, &etnaviv_sched_ops,
>  			     etnaviv_hw_jobs_limit, etnaviv_job_hang_limit,
> -			     msecs_to_jiffies(500), dev_name(gpu->dev));
> +			     msecs_to_jiffies(500), dev_name(gpu->dev),
> +				 NULL);

Indentation looks wrong for the NULL here...


> diff --git a/drivers/gpu/drm/scheduler/sched_entity.c b/drivers/gpu/drm/scheduler/sched_entity.c
> index 3e22a54..320c77a 100644
> --- a/drivers/gpu/drm/scheduler/sched_entity.c
> +++ b/drivers/gpu/drm/scheduler/sched_entity.c
> @@ -130,7 +130,16 @@ drm_sched_entity_get_free_sched(struct drm_sched_entity *entity)
>  	int i;
>  
>  	for (i = 0; i < entity->num_rq_list; ++i) {
> -		num_jobs = atomic_read(&entity->rq_list[i]->sched->num_jobs);
> +		struct drm_gpu_scheduler *sched = entity->rq_list[i]->sched;
> +
> +		if (entity->rq_list[i]->sched->ready &&
> +				!entity->rq_list[i]->sched->ready(sched)) {

... the second line of the if here...


> diff --git a/drivers/gpu/drm/v3d/v3d_sched.c b/drivers/gpu/drm/v3d/v3d_sched.c
> index 80b641f..e06cc0d 100644
> --- a/drivers/gpu/drm/v3d/v3d_sched.c
> +++ b/drivers/gpu/drm/v3d/v3d_sched.c
> @@ -212,7 +212,8 @@ v3d_sched_init(struct v3d_dev *v3d)
>  			     &v3d_sched_ops,
>  			     hw_jobs_limit, job_hang_limit,
>  			     msecs_to_jiffies(hang_limit_ms),
> -			     "v3d_bin");
> +			     "v3d_bin",
> +				 NULL);

... the NULL here...


> diff --git a/include/drm/gpu_scheduler.h b/include/drm/gpu_scheduler.h
> index 0684dcd..a6e18fe 100644
> --- a/include/drm/gpu_scheduler.h
> +++ b/include/drm/gpu_scheduler.h
> @@ -283,12 +283,15 @@ struct drm_gpu_scheduler {
>  	spinlock_t			job_list_lock;
>  	int				hang_limit;
>  	atomic_t                        num_jobs;
> +
> +	bool 			(*ready)(struct drm_gpu_scheduler *sched);
>  };

... and the new line here.

Which editor are you using?


-- 
Earthling Michel Dänzer               |               http://www.amd.com
Libre software enthusiast             |             Mesa and X developer