bybrooklyn/linux
1
0
Fork 0
mirror of https://github.com/torvalds/linux.git synced 2026-04-27 11:02:31 -04:00
Code Issues Packages Projects Releases Wiki Activity

drm/amdgpu: Follow up change to previous drm scheduler change.

Browse Source 
Align refcount behaviour for amdgpu_job embedded HW fence with
classic pointer style HW fences by increasing refcount each
time emit is called so amdgpu code doesn't need to make workarounds
using amdgpu_job.job_run_counter to keep the HW fence refcount balanced.

Also since in the previous patch we resumed setting s_fence->parent to NULL
in drm_sched_stop switch to directly checking if job->hw_fence is
signaled to short circuit reset if already signed.

Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Tested-by: Yiqing Yao <yiqing.yao@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This commit is contained in:
Andrey Grodzovsky
2022-06-20 17:33:55 -04:00
committed by Alex Deucher
parent 45ecaea738
commit 9ae55f030d
4 changed files with 29 additions and 11 deletions
Download Patch File Download Diff File Expand all files Collapse all files

27
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
View File

@@ -5010,16 +5010,32 @@ static void amdgpu_device_recheck_guilty_jobs(
/* clear job's guilty and depend the folowing step to decide the real one */
drm_sched_reset_karma(s_job);
/* for the real bad job, it will be resubmitted twice, adding a dma_fence_get
* to make sure fence is balanced */
dma_fence_get(s_job->s_fence->parent);
drm_sched_resubmit_jobs_ext(&ring->sched, 1);
if (!s_job->s_fence->parent) {
DRM_WARN("Failed to get a HW fence for job!");
continue;
}
ret = dma_fence_wait_timeout(s_job->s_fence->parent, false, ring->sched.timeout);
if (ret == 0) { /* timeout */
DRM_ERROR("Found the real bad job! ring:%s, job_id:%llx\n",
ring->sched.name, s_job->id);
amdgpu_fence_driver_isr_toggle(adev, true);
/* Clear this failed job from fence array */
amdgpu_fence_driver_clear_job_fences(ring);
amdgpu_fence_driver_isr_toggle(adev, false);
/* Since the job won't signal and we go for
* another resubmit drop this parent pointer
*/
dma_fence_put(s_job->s_fence->parent);
s_job->s_fence->parent = NULL;
/* set guilty */
drm_sched_increase_karma(s_job);
retry:
@@ -5048,7 +5064,6 @@ retry:
/* got the hw fence, signal finished fence */
atomic_dec(ring->sched.score);
dma_fence_put(s_job->s_fence->parent);
dma_fence_get(&s_job->s_fence->finished);
dma_fence_signal(&s_job->s_fence->finished);
dma_fence_put(&s_job->s_fence->finished);
@@ -5221,8 +5236,8 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev,
*
* job->base holds a reference to parent fence
*/
if (job && job->base.s_fence->parent &&
dma_fence_is_signaled(job->base.s_fence->parent)) {
if (job && (job->hw_fence.ops != NULL) &&
dma_fence_is_signaled(&job->hw_fence)) {
job_signaled = true;
dev_info(adev->dev, "Guilty job already signaled, skipping HW reset");
goto skip_hw_reset;