[PATCH 3/3] drm/amdgpu: Refresh rq selection for job after ASIC reset

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



A ring might become unusable after reset, if that the case
drm_sched_entity_select_rq will choose another, working rq
to run the job if there is one.
Also, skip recovery of ring which is not ready.

Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@xxxxxxx>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 16 +++++++++++++++-
 1 file changed, 15 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index d11489e..3124ca1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -3355,10 +3355,24 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev,
 	else
 		r = amdgpu_device_reset(adev);
 
+	/*
+	 * After reboot a ring might fail in which case this will
+	 * move the job to different rq if possible
+	 */
+	if (job) {
+		drm_sched_entity_select_rq(job->base.entity);
+		if (job->base.entity->rq) {
+			job->base.sched = job->base.entity->rq->sched;
+		} else {
+			job->base.sched = NULL;
+			r = -ENOENT;
+		}
+	}
+
 	for (i = 0; i < AMDGPU_MAX_RINGS; ++i) {
 		struct amdgpu_ring *ring = adev->rings[i];
 
-		if (!ring || !ring->sched.thread)
+		if (!ring || !ring->ready || !ring->sched.thread)
 			continue;
 
 		/* only need recovery sched of the given job's ring
-- 
2.7.4

_______________________________________________
amd-gfx mailing list
amd-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/amd-gfx




[Index of Archives]     [Linux USB Devel]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux