Commit 222b5f04 authored by Andrey Grodzovsky's avatar Andrey Grodzovsky Committed by Alex Deucher

drm/sched: Refactor ring mirror list handling.

Decauple sched threads stop and start and ring mirror
list handling from the policy of what to do about the
guilty jobs.
When stoppping the sched thread and detaching sched fences
from non signaled HW fenes wait for all signaled HW fences
to complete before rerunning the jobs.

v2: Fix resubmission of guilty job into HW after refactoring.

v4:
Full restart for all the jobs, not only from guilty ring.
Extract karma increase into standalone function.

v5:
Rework waiting for signaled jobs without relying on the job
struct itself as those might already be freed for non 'guilty'
job's schedulers.
Expose karma increase to drivers.

v6:
Use list_for_each_entry_safe_continue and drm_sched_process_job
in case fence already signaled.
Call drm_sched_increase_karma only once for amdgpu and add documentation.

v7:
Wait only for the latest job's fence.
Suggested-by: default avatarChristian Koenig <Christian.Koenig@amd.com>
Signed-off-by: default avatarAndrey Grodzovsky <andrey.grodzovsky@amd.com>
Reviewed-by: default avatarChristian König <christian.koenig@amd.com>
Signed-off-by: default avatarAlex Deucher <alexander.deucher@amd.com>
parent 713b64a5
...@@ -3313,17 +3313,15 @@ static int amdgpu_device_pre_asic_reset(struct amdgpu_device *adev, ...@@ -3313,17 +3313,15 @@ static int amdgpu_device_pre_asic_reset(struct amdgpu_device *adev,
if (!ring || !ring->sched.thread) if (!ring || !ring->sched.thread)
continue; continue;
kthread_park(ring->sched.thread); drm_sched_stop(&ring->sched);
if (job && job->base.sched != &ring->sched)
continue;
drm_sched_hw_job_reset(&ring->sched, job ? &job->base : NULL);
/* after all hw jobs are reset, hw fence is meaningless, so force_completion */ /* after all hw jobs are reset, hw fence is meaningless, so force_completion */
amdgpu_fence_driver_force_completion(ring); amdgpu_fence_driver_force_completion(ring);
} }
if(job)
drm_sched_increase_karma(&job->base);
if (!amdgpu_sriov_vf(adev)) { if (!amdgpu_sriov_vf(adev)) {
...@@ -3469,14 +3467,10 @@ static void amdgpu_device_post_asic_reset(struct amdgpu_device *adev, ...@@ -3469,14 +3467,10 @@ static void amdgpu_device_post_asic_reset(struct amdgpu_device *adev,
if (!ring || !ring->sched.thread) if (!ring || !ring->sched.thread)
continue; continue;
/* only need recovery sched of the given job's ring if (!adev->asic_reset_res)
* or all rings (in the case @job is NULL) drm_sched_resubmit_jobs(&ring->sched);
* after above amdgpu_reset accomplished
*/
if ((!job || job->base.sched == &ring->sched) && !adev->asic_reset_res)
drm_sched_job_recovery(&ring->sched);
kthread_unpark(ring->sched.thread); drm_sched_start(&ring->sched, !adev->asic_reset_res);
} }
if (!amdgpu_device_has_dc_support(adev)) { if (!amdgpu_device_has_dc_support(adev)) {
......
...@@ -109,16 +109,19 @@ static void etnaviv_sched_timedout_job(struct drm_sched_job *sched_job) ...@@ -109,16 +109,19 @@ static void etnaviv_sched_timedout_job(struct drm_sched_job *sched_job)
} }
/* block scheduler */ /* block scheduler */
kthread_park(gpu->sched.thread); drm_sched_stop(&gpu->sched);
drm_sched_hw_job_reset(&gpu->sched, sched_job);
if(sched_job)
drm_sched_increase_karma(sched_job);
/* get the GPU back into the init state */ /* get the GPU back into the init state */
etnaviv_core_dump(gpu); etnaviv_core_dump(gpu);
etnaviv_gpu_recover_hang(gpu); etnaviv_gpu_recover_hang(gpu);
drm_sched_resubmit_jobs(&gpu->sched);
/* restart scheduler after GPU is usable again */ /* restart scheduler after GPU is usable again */
drm_sched_job_recovery(&gpu->sched); drm_sched_start(&gpu->sched, true);
kthread_unpark(gpu->sched.thread);
} }
static void etnaviv_sched_free_job(struct drm_sched_job *sched_job) static void etnaviv_sched_free_job(struct drm_sched_job *sched_job)
......
...@@ -60,8 +60,6 @@ ...@@ -60,8 +60,6 @@
static void drm_sched_process_job(struct dma_fence *f, struct dma_fence_cb *cb); static void drm_sched_process_job(struct dma_fence *f, struct dma_fence_cb *cb);
static void drm_sched_expel_job_unlocked(struct drm_sched_job *s_job);
/** /**
* drm_sched_rq_init - initialize a given run queue struct * drm_sched_rq_init - initialize a given run queue struct
* *
...@@ -335,6 +333,51 @@ static void drm_sched_job_timedout(struct work_struct *work) ...@@ -335,6 +333,51 @@ static void drm_sched_job_timedout(struct work_struct *work)
spin_unlock_irqrestore(&sched->job_list_lock, flags); spin_unlock_irqrestore(&sched->job_list_lock, flags);
} }
/**
* drm_sched_increase_karma - Update sched_entity guilty flag
*
* @bad: The job guilty of time out
*
* Increment on every hang caused by the 'bad' job. If this exceeds the hang
* limit of the scheduler then the respective sched entity is marked guilty and
* jobs from it will not be scheduled further
*/
void drm_sched_increase_karma(struct drm_sched_job *bad)
{
int i;
struct drm_sched_entity *tmp;
struct drm_sched_entity *entity;
struct drm_gpu_scheduler *sched = bad->sched;
/* don't increase @bad's karma if it's from KERNEL RQ,
* because sometimes GPU hang would cause kernel jobs (like VM updating jobs)
* corrupt but keep in mind that kernel jobs always considered good.
*/
if (bad->s_priority != DRM_SCHED_PRIORITY_KERNEL) {
atomic_inc(&bad->karma);
for (i = DRM_SCHED_PRIORITY_MIN; i < DRM_SCHED_PRIORITY_KERNEL;
i++) {
struct drm_sched_rq *rq = &sched->sched_rq[i];
spin_lock(&rq->lock);
list_for_each_entry_safe(entity, tmp, &rq->entities, list) {
if (bad->s_fence->scheduled.context ==
entity->fence_context) {
if (atomic_read(&bad->karma) >
bad->sched->hang_limit)
if (entity->guilty)
atomic_set(entity->guilty, 1);
break;
}
}
spin_unlock(&rq->lock);
if (&entity->list != &rq->entities)
break;
}
}
}
EXPORT_SYMBOL(drm_sched_increase_karma);
/** /**
* drm_sched_hw_job_reset - stop the scheduler if it contains the bad job * drm_sched_hw_job_reset - stop the scheduler if it contains the bad job
* *
...@@ -342,13 +385,20 @@ static void drm_sched_job_timedout(struct work_struct *work) ...@@ -342,13 +385,20 @@ static void drm_sched_job_timedout(struct work_struct *work)
* @bad: bad scheduler job * @bad: bad scheduler job
* *
*/ */
void drm_sched_hw_job_reset(struct drm_gpu_scheduler *sched, struct drm_sched_job *bad) void drm_sched_stop(struct drm_gpu_scheduler *sched)
{ {
struct drm_sched_job *s_job; struct drm_sched_job *s_job;
struct drm_sched_entity *entity, *tmp;
unsigned long flags; unsigned long flags;
int i; struct dma_fence *last_fence = NULL;
kthread_park(sched->thread);
/*
* Verify all the signaled jobs in mirror list are removed from the ring
* by waiting for the latest job to enter the list. This should insure that
* also all the previous jobs that were in flight also already singaled
* and removed from the list.
*/
spin_lock_irqsave(&sched->job_list_lock, flags); spin_lock_irqsave(&sched->job_list_lock, flags);
list_for_each_entry_reverse(s_job, &sched->ring_mirror_list, node) { list_for_each_entry_reverse(s_job, &sched->ring_mirror_list, node) {
if (s_job->s_fence->parent && if (s_job->s_fence->parent &&
...@@ -357,35 +407,20 @@ void drm_sched_hw_job_reset(struct drm_gpu_scheduler *sched, struct drm_sched_jo ...@@ -357,35 +407,20 @@ void drm_sched_hw_job_reset(struct drm_gpu_scheduler *sched, struct drm_sched_jo
dma_fence_put(s_job->s_fence->parent); dma_fence_put(s_job->s_fence->parent);
s_job->s_fence->parent = NULL; s_job->s_fence->parent = NULL;
atomic_dec(&sched->hw_rq_count); atomic_dec(&sched->hw_rq_count);
} else {
last_fence = dma_fence_get(&s_job->s_fence->finished);
break;
} }
} }
spin_unlock_irqrestore(&sched->job_list_lock, flags); spin_unlock_irqrestore(&sched->job_list_lock, flags);
if (bad && bad->s_priority != DRM_SCHED_PRIORITY_KERNEL) { if (last_fence) {
atomic_inc(&bad->karma); dma_fence_wait(last_fence, false);
/* don't increase @bad's karma if it's from KERNEL RQ, dma_fence_put(last_fence);
* becuase sometimes GPU hang would cause kernel jobs (like VM updating jobs)
* corrupt but keep in mind that kernel jobs always considered good.
*/
for (i = DRM_SCHED_PRIORITY_MIN; i < DRM_SCHED_PRIORITY_KERNEL; i++ ) {
struct drm_sched_rq *rq = &sched->sched_rq[i];
spin_lock(&rq->lock);
list_for_each_entry_safe(entity, tmp, &rq->entities, list) {
if (bad->s_fence->scheduled.context == entity->fence_context) {
if (atomic_read(&bad->karma) > bad->sched->hang_limit)
if (entity->guilty)
atomic_set(entity->guilty, 1);
break;
}
}
spin_unlock(&rq->lock);
if (&entity->list != &rq->entities)
break;
}
} }
} }
EXPORT_SYMBOL(drm_sched_hw_job_reset);
EXPORT_SYMBOL(drm_sched_stop);
/** /**
* drm_sched_job_recovery - recover jobs after a reset * drm_sched_job_recovery - recover jobs after a reset
...@@ -393,33 +428,21 @@ EXPORT_SYMBOL(drm_sched_hw_job_reset); ...@@ -393,33 +428,21 @@ EXPORT_SYMBOL(drm_sched_hw_job_reset);
* @sched: scheduler instance * @sched: scheduler instance
* *
*/ */
void drm_sched_job_recovery(struct drm_gpu_scheduler *sched) void drm_sched_start(struct drm_gpu_scheduler *sched, bool full_recovery)
{ {
struct drm_sched_job *s_job, *tmp; struct drm_sched_job *s_job, *tmp;
bool found_guilty = false;
unsigned long flags; unsigned long flags;
int r; int r;
if (!full_recovery)
goto unpark;
spin_lock_irqsave(&sched->job_list_lock, flags); spin_lock_irqsave(&sched->job_list_lock, flags);
list_for_each_entry_safe(s_job, tmp, &sched->ring_mirror_list, node) { list_for_each_entry_safe(s_job, tmp, &sched->ring_mirror_list, node) {
struct drm_sched_fence *s_fence = s_job->s_fence; struct drm_sched_fence *s_fence = s_job->s_fence;
struct dma_fence *fence; struct dma_fence *fence = s_job->s_fence->parent;
uint64_t guilty_context;
if (!found_guilty && atomic_read(&s_job->karma) > sched->hang_limit) {
found_guilty = true;
guilty_context = s_job->s_fence->scheduled.context;
}
if (found_guilty && s_job->s_fence->scheduled.context == guilty_context)
dma_fence_set_error(&s_fence->finished, -ECANCELED);
spin_unlock_irqrestore(&sched->job_list_lock, flags);
fence = sched->ops->run_job(s_job);
atomic_inc(&sched->hw_rq_count);
if (fence) { if (fence) {
s_fence->parent = dma_fence_get(fence);
r = dma_fence_add_callback(fence, &s_fence->cb, r = dma_fence_add_callback(fence, &s_fence->cb,
drm_sched_process_job); drm_sched_process_job);
if (r == -ENOENT) if (r == -ENOENT)
...@@ -427,18 +450,47 @@ void drm_sched_job_recovery(struct drm_gpu_scheduler *sched) ...@@ -427,18 +450,47 @@ void drm_sched_job_recovery(struct drm_gpu_scheduler *sched)
else if (r) else if (r)
DRM_ERROR("fence add callback failed (%d)\n", DRM_ERROR("fence add callback failed (%d)\n",
r); r);
dma_fence_put(fence); } else
} else {
if (s_fence->finished.error < 0)
drm_sched_expel_job_unlocked(s_job);
drm_sched_process_job(NULL, &s_fence->cb); drm_sched_process_job(NULL, &s_fence->cb);
} }
spin_lock_irqsave(&sched->job_list_lock, flags);
}
drm_sched_start_timeout(sched); drm_sched_start_timeout(sched);
spin_unlock_irqrestore(&sched->job_list_lock, flags); spin_unlock_irqrestore(&sched->job_list_lock, flags);
unpark:
kthread_unpark(sched->thread);
} }
EXPORT_SYMBOL(drm_sched_job_recovery); EXPORT_SYMBOL(drm_sched_start);
/**
* drm_sched_resubmit_jobs - helper to relunch job from mirror ring list
*
* @sched: scheduler instance
*
*/
void drm_sched_resubmit_jobs(struct drm_gpu_scheduler *sched)
{
struct drm_sched_job *s_job, *tmp;
uint64_t guilty_context;
bool found_guilty = false;
/*TODO DO we need spinlock here ? */
list_for_each_entry_safe(s_job, tmp, &sched->ring_mirror_list, node) {
struct drm_sched_fence *s_fence = s_job->s_fence;
if (!found_guilty && atomic_read(&s_job->karma) > sched->hang_limit) {
found_guilty = true;
guilty_context = s_job->s_fence->scheduled.context;
}
if (found_guilty && s_job->s_fence->scheduled.context == guilty_context)
dma_fence_set_error(&s_fence->finished, -ECANCELED);
s_job->s_fence->parent = sched->ops->run_job(s_job);
atomic_inc(&sched->hw_rq_count);
}
}
EXPORT_SYMBOL(drm_sched_resubmit_jobs);
/** /**
* drm_sched_job_init - init a scheduler job * drm_sched_job_init - init a scheduler job
...@@ -634,26 +686,14 @@ static int drm_sched_main(void *param) ...@@ -634,26 +686,14 @@ static int drm_sched_main(void *param)
DRM_ERROR("fence add callback failed (%d)\n", DRM_ERROR("fence add callback failed (%d)\n",
r); r);
dma_fence_put(fence); dma_fence_put(fence);
} else { } else
if (s_fence->finished.error < 0)
drm_sched_expel_job_unlocked(sched_job);
drm_sched_process_job(NULL, &s_fence->cb); drm_sched_process_job(NULL, &s_fence->cb);
}
wake_up(&sched->job_scheduled); wake_up(&sched->job_scheduled);
} }
return 0; return 0;
} }
static void drm_sched_expel_job_unlocked(struct drm_sched_job *s_job)
{
struct drm_gpu_scheduler *sched = s_job->sched;
spin_lock(&sched->job_list_lock);
list_del_init(&s_job->node);
spin_unlock(&sched->job_list_lock);
}
/** /**
* drm_sched_init - Init a gpu scheduler instance * drm_sched_init - Init a gpu scheduler instance
* *
......
...@@ -234,18 +234,21 @@ v3d_gpu_reset_for_timeout(struct v3d_dev *v3d, struct drm_sched_job *sched_job) ...@@ -234,18 +234,21 @@ v3d_gpu_reset_for_timeout(struct v3d_dev *v3d, struct drm_sched_job *sched_job)
for (q = 0; q < V3D_MAX_QUEUES; q++) { for (q = 0; q < V3D_MAX_QUEUES; q++) {
struct drm_gpu_scheduler *sched = &v3d->queue[q].sched; struct drm_gpu_scheduler *sched = &v3d->queue[q].sched;
kthread_park(sched->thread); drm_sched_stop(sched);
drm_sched_hw_job_reset(sched, (sched_job->sched == sched ?
sched_job : NULL)); if(sched_job)
drm_sched_increase_karma(sched_job);
} }
/* get the GPU back into the init state */ /* get the GPU back into the init state */
v3d_reset(v3d); v3d_reset(v3d);
for (q = 0; q < V3D_MAX_QUEUES; q++)
drm_sched_resubmit_jobs(sched_job->sched);
/* Unblock schedulers and restart their jobs. */ /* Unblock schedulers and restart their jobs. */
for (q = 0; q < V3D_MAX_QUEUES; q++) { for (q = 0; q < V3D_MAX_QUEUES; q++) {
drm_sched_job_recovery(&v3d->queue[q].sched); drm_sched_start(&v3d->queue[q].sched, true);
kthread_unpark(v3d->queue[q].sched.thread);
} }
mutex_unlock(&v3d->reset_lock); mutex_unlock(&v3d->reset_lock);
......
...@@ -298,9 +298,10 @@ int drm_sched_job_init(struct drm_sched_job *job, ...@@ -298,9 +298,10 @@ int drm_sched_job_init(struct drm_sched_job *job,
void *owner); void *owner);
void drm_sched_job_cleanup(struct drm_sched_job *job); void drm_sched_job_cleanup(struct drm_sched_job *job);
void drm_sched_wakeup(struct drm_gpu_scheduler *sched); void drm_sched_wakeup(struct drm_gpu_scheduler *sched);
void drm_sched_hw_job_reset(struct drm_gpu_scheduler *sched, void drm_sched_stop(struct drm_gpu_scheduler *sched);
struct drm_sched_job *job); void drm_sched_start(struct drm_gpu_scheduler *sched, bool full_recovery);
void drm_sched_job_recovery(struct drm_gpu_scheduler *sched); void drm_sched_resubmit_jobs(struct drm_gpu_scheduler *sched);
void drm_sched_increase_karma(struct drm_sched_job *bad);
bool drm_sched_dependency_optimized(struct dma_fence* fence, bool drm_sched_dependency_optimized(struct dma_fence* fence,
struct drm_sched_entity *entity); struct drm_sched_entity *entity);
void drm_sched_fault(struct drm_gpu_scheduler *sched); void drm_sched_fault(struct drm_gpu_scheduler *sched);
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment