On 05/12/2025 14:46, Christian König wrote:
On 12/5/25 14:40, Tvrtko Ursulin wrote:
Drop reference to syncobj and timeline fence when aborting the ioctl due
output array being too small.
Signed-off-by: Tvrtko Ursulin <[email protected]>
Fixes: a292fdecd728 ("drm/amdgpu: Implement userqueue signal/wait IOCTL")
Cc: Arunpravin Paneer Selvam <[email protected]>
Cc: Christian König <[email protected]>
Cc: Alex Deucher <[email protected]>
Cc: <[email protected]> # v6.16+
I need to double check the code when I have time, but of hand looks legitimate
to me.
Where are patches #3-#12 from this series?
On amd-gfx only, since 3-12 only contains cleanups I thought not pollute
the inboxes too much.
Regards,
Tvrtko
---
drivers/gpu/drm/amd/amdgpu/amdgpu_userq_fence.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_userq_fence.c
b/drivers/gpu/drm/amd/amdgpu/amdgpu_userq_fence.c
index eba9fb359047..13c5d4462be6 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_userq_fence.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_userq_fence.c
@@ -865,6 +865,7 @@ int amdgpu_userq_wait_ioctl(struct drm_device *dev, void
*data,
dma_fence_unwrap_for_each(f, &iter, fence) {
if (WARN_ON_ONCE(num_fences >=
wait_info->num_fences)) {
r = -EINVAL;
+ dma_fence_put(fence);
goto free_fences;
}
@@ -889,6 +890,7 @@ int amdgpu_userq_wait_ioctl(struct drm_device *dev, void *data,
if (WARN_ON_ONCE(num_fences >= wait_info->num_fences)) {
r = -EINVAL;
+ dma_fence_put(fence);
goto free_fences;
}
