drm/amdgpu: protect fence_process from multiple context

fence_process may be called from kthread, user thread and interrupt context.
it is possible to called concurrently, then will wake up fence queue multiple times.

Signed-off-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Jammy Zhou <Jammy.Zhou@amd.com>

1 files changed, 1 insertions, 1 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
index 855e2196657a..1e68a561bbfe 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
@@ -367,7 +367,7 @@ int amdgpu_ring_init(struct amdgpu_device *adev, struct amdgpu_ring *ring,
 	}
 	ring->next_rptr_gpu_addr = adev->wb.gpu_addr + (ring->next_rptr_offs * 4);
 	ring->next_rptr_cpu_addr = &adev->wb.wb[ring->next_rptr_offs];
-
+	spin_lock_init(&ring->fence_lock);
 	r = amdgpu_fence_driver_start_ring(ring, irq_src, irq_type);
 	if (r) {
 		dev_err(adev->dev, "failed initializing fences (%d).\n", r);