drm/amdgpu/SRIOV:implement guilty job TDR for(V2)

1,TDR will kickout guilty job if it hang exceed the threshold
of the given one from kernel paramter "job_hang_limit", that
way a bad command stream will not infinitly cause GPU hang.

by default this threshold is 1 so a job will be kicked out
after it hang.

2,if a job timeout TDR routine will not reset all sched/ring,
instead if will only reset on the givn one which is indicated
by @job of amdgpu_sriov_gpu_reset, that way we don't need to
reset and recover each sched/ring if we already know which job
cause GPU hang.

3,unblock sriov_gpu_reset for AI family.

V2:
1:put kickout guilty job after sched parked.
2:since parking scheduler prior to kickout already occupies a
while, we can do last check on the in question job before
doing hw_reset.

TODO:
1:when a job is considered as guilty, we should mark some flag
in its fence status flag, and let UMD side aware that this
fence signaling is not due to job complete but job hang.

2:if gpu reset cause all video memory lost, we need introduce
a new policy to implement TDR, like drop all jobs not yet
signaled, and all IOCTL on this device will return ERROR
DEVICE_LOST.
this will be implemented later.

Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

2 files changed, 17 insertions, 1 deletions

diff --git a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c b/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c
index fea96a765cf1..38cea6fb25a8 100644
--- a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c
+++ b/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c
@@ -409,9 +409,18 @@ void amd_sched_hw_job_reset(struct amd_gpu_scheduler *sched)
 					      &s_job->s_fence->cb)) {
 			dma_fence_put(s_job->s_fence->parent);
 			s_job->s_fence->parent = NULL;
+			atomic_dec(&sched->hw_rq_count);
 		}
 	}
-	atomic_set(&sched->hw_rq_count, 0);
+	spin_unlock(&sched->job_list_lock);
+}
+
+void amd_sched_job_kickout(struct amd_sched_job *s_job)
+{
+	struct amd_gpu_scheduler *sched = s_job->sched;
+
+	spin_lock(&sched->job_list_lock);
+	list_del_init(&s_job->node);
 	spin_unlock(&sched->job_list_lock);
 }
 
diff --git a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.h b/drivers/gpu/drm/amd/scheduler/gpu_scheduler.h
index 924d4a5899e1..f9d8f28efd16 100644
--- a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.h
+++ b/drivers/gpu/drm/amd/scheduler/gpu_scheduler.h
@@ -81,6 +81,7 @@ struct amd_sched_job {
 	struct list_head		node;
 	struct delayed_work		work_tdr;
 	uint64_t			id;
+	atomic_t karma;
 };
 
 extern const struct dma_fence_ops amd_sched_fence_ops_scheduled;
@@ -96,6 +97,11 @@ static inline struct amd_sched_fence *to_amd_sched_fence(struct dma_fence *f)
 	return NULL;
 }
 
+static inline bool amd_sched_invalidate_job(struct amd_sched_job *s_job, int threshold)
+{
+	return (s_job && atomic_inc_return(&s_job->karma) > threshold);
+}
+
 /**
  * Define the backend operations called by the scheduler,
  * these functions should be implemented in driver side
@@ -160,4 +166,5 @@ void amd_sched_hw_job_reset(struct amd_gpu_scheduler *sched);
 void amd_sched_job_recovery(struct amd_gpu_scheduler *sched);
 bool amd_sched_dependency_optimized(struct dma_fence* fence,
 				    struct amd_sched_entity *entity);
+void amd_sched_job_kickout(struct amd_sched_job *s_job);
 #endif