drm/amdgpu: handle eviction fence race

The eviction process can get into a race condition between the eviction
fence suspend work (which replaces the old fence with new) and kms_close
(which destroys the fence and doesn't expect a new one).

This patch:
- adds a flag to indicate that fd is closing, so fence replacement is
  not required (evf_mgr->fd_closing)
- adds a flush_work() during the ev_fence_destroy routine

V2: Addressed review comments from Christian:
    - Do not use mutex to sync
    - Use flush_work and wait for suspend_work to be done

V3: Fixed state machine for queue->active, which adds into race between
    suspend/resume and queue ops

Cc: Alex Deucher <[email protected]>
Cc: Christian König <[email protected]>
Reviewed-by: Christian König <[email protected]>
Signed-off-by: Shashank Sharma <[email protected]>
Signed-off-by: Arvind Yadav <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>

1 files changed, 2 insertions, 1 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_userqueue.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_userqueue.c
index 57e502d014a1..cba51bdf2e2c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_userqueue.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_userqueue.c
@@ -614,9 +614,10 @@ void amdgpu_userq_mgr_fini(struct amdgpu_userq_mgr *userq_mgr)
 
 	cancel_delayed_work(&userq_mgr->resume_work);
 
+	mutex_lock(&userq_mgr->userq_mutex);
 	idr_for_each_entry(&userq_mgr->userq_idr, queue, queue_id)
 		amdgpu_userqueue_cleanup(userq_mgr, queue, queue_id);
-
 	idr_destroy(&userq_mgr->userq_idr);
+	mutex_unlock(&userq_mgr->userq_mutex);
 	mutex_destroy(&userq_mgr->userq_mutex);
 }