Merge tag 'amd-drm-next-6.5-2023-06-16' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

amd-drm-next-6.5-2023-06-16:

amdgpu:
- Misc display fixes
- W=1 fixes
- Improve scheduler naming
- DCN 3.1.4 fixes
- kdoc fixes
- Enable W=1
- VCN 4.0 fix
- xgmi fixes
- TOPDOWN fix for large BAR systems
- eDP fix
- PSR fixes
- SubVP fixes
- Freesync fix
- DPIA fix
- SMU 13.0.5 fixes
- vblflash fix
- RAS fixes
- SDMA 4 fix
- BO locking fix
- BO backing store fix
- NBIO 7.9 fixes
- GC 9.4.3 fixes
- GPU reset recovery fixes
- HMM fix

amdkfd:
- Fix NULL check
- Trap fixes
- Queue count fix
- Add event age tracking

radeon:
- fbdev client fix

scheduler:
- Avoid an infinite loop

UAPI:
- Add KFD event age tracking:
  Proposed ROCT-Thunk-Interface:
  https://github.com/RadeonOpenCompute/ROCT-Thunk-Interface/commit/efdbf6cfbc026bd68ac3c35d00dacf84370eb81e
  https://github.com/RadeonOpenCompute/ROCT-Thunk-Interface/commit/1820ae0a2db85b6f584611dc0cde1a00e7c22915
  Proposed ROCR-Runtime:
  https://github.com/RadeonOpenCompute/ROCR-Runtime/compare/master...zhums:ROCR-Runtime:new_event_wait_review
  https://github.com/RadeonOpenCompute/ROCR-Runtime/commit/e1f5bdb88eb882ac798aeca2c00ea3fbb2dba459
  https://github.com/RadeonOpenCompute/ROCR-Runtime/commit/7d26afd14107b5c2a754c1a3f415d89f3aabb503

drm:
- DP MST fix

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230616163548.7706-1-alexander.deucher@amd.com

1 files changed, 25 insertions, 0 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
index 876ec35b8f83..c694b41f6461 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
@@ -692,6 +692,30 @@ void amdgpu_fence_driver_clear_job_fences(struct amdgpu_ring *ring)
 }
 
 /**
+ * amdgpu_fence_driver_set_error - set error code on fences
+ * @ring: the ring which contains the fences
+ * @error: the error code to set
+ *
+ * Set an error code to all the fences pending on the ring.
+ */
+void amdgpu_fence_driver_set_error(struct amdgpu_ring *ring, int error)
+{
+	struct amdgpu_fence_driver *drv = &ring->fence_drv;
+	unsigned long flags;
+
+	spin_lock_irqsave(&drv->lock, flags);
+	for (unsigned int i = 0; i <= drv->num_fences_mask; ++i) {
+		struct dma_fence *fence;
+
+		fence = rcu_dereference_protected(drv->fences[i],
+						  lockdep_is_held(&drv->lock));
+		if (fence && !dma_fence_is_signaled_locked(fence))
+			dma_fence_set_error(fence, error);
+	}
+	spin_unlock_irqrestore(&drv->lock, flags);
+}
+
+/**
  * amdgpu_fence_driver_force_completion - force signal latest fence of ring
  *
  * @ring: fence of the ring to signal
@@ -699,6 +723,7 @@ void amdgpu_fence_driver_clear_job_fences(struct amdgpu_ring *ring)
  */
 void amdgpu_fence_driver_force_completion(struct amdgpu_ring *ring)
 {
+	amdgpu_fence_driver_set_error(ring, -ECANCELED);
 	amdgpu_fence_write(ring, ring->fence_drv.sync_seq);
 	amdgpu_fence_process(ring);
 }