github-mirror/linux
1
0
Fork 0
mirror of https://github.com/torvalds/linux.git synced 2026-07-28 10:09:10 +02:00
Code Issues Actions Packages Projects Releases Wiki Activity Graph

drm/amdgpu/userq: fix indefinite fence wait during GPU reset

Browse Source 
pre_reset only force-completes fences of MAPPED queues. A queue in any
other state (e.g. mid-eviction) keeps its last_fence pending; after a
GPU reset that fence never signals, so the eviction/suspend worker and
process teardown (amdgpu_evf_mgr_flush_suspend) wait on it forever and
wedge the machine:

  INFO: task kworker/6:28 blocked for more than 120 seconds.
  Workqueue: events amdgpu_eviction_fence_suspend_worker [amdgpu]
  Call Trace:
   dma_fence_wait_timeout+0x7e/0x130
   amdgpu_userq_evict+0x67/0x140 [amdgpu]
   amdgpu_eviction_fence_suspend_worker+0xd8/0x160 [amdgpu]
   process_scheduled_works+0xa6/0x420

Force-complete every queue's fence regardless of state. The unmap and
mark-hung step stays gated on MAPPED, since unmapping a queue that is
not mapped is invalid.

Fixes: 290f46cf57 ("drm/amdgpu: Implement user queue reset functionality")
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Jesse Zhang <Jesse.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 9102b39fa924dcc3dc75a3137bfa9633c40b88c0)
Cc: stable@vger.kernel.org
This commit is contained in:
Jesse Zhang 2026-07-10 16:45:42 +08:00 committed by Alex Deucher
parent f9ad396aa8
commit 5d75ec2e5f
1 changed files with 12 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
drivers/gpu/drm/amd/amdgpu/amdgpu_userq.c
View File

@ -1376,16 +1376,19 @@ void amdgpu_userq_pre_reset(struct amdgpu_device *adev)
/* TODO: We probably need a new lock for the queue state */
xa_for_each(&adev->userq_doorbell_xa, queue_id, queue) {
if (queue->state != AMDGPU_USERQ_STATE_MAPPED)
continue;
userq_funcs = adev->userq_funcs[queue->queue_type];
userq_funcs->unmap(queue);
/* just mark all queues as hung at this point.
* if unmap succeeds, we could map again
* in amdgpu_userq_post_reset() if vram is not lost
if (queue->state == AMDGPU_USERQ_STATE_MAPPED) {
userq_funcs = adev->userq_funcs[queue->queue_type];
userq_funcs->unmap(queue);
/* just mark all queues as hung at this point.
* if unmap succeeds, we could map again
* in amdgpu_userq_post_reset() if vram is not lost
*/
queue->state = AMDGPU_USERQ_STATE_HUNG;
}
/* Force-complete any pending fence regardless of queue state so
* that eviction/suspend and queue teardown waiters don't block
* forever on a fence that will never signal after the reset.
*/
queue->state = AMDGPU_USERQ_STATE_HUNG;
amdgpu_userq_fence_driver_force_completion(queue);
}
}