Commit f9e69496 authored by Christian König's avatar Christian König Committed by Alex Deucher
Browse files

drm/amdgpu: workaround for TLB seq race 



It can happen that we query the sequence value before the callback
had a chance to run.

Workaround that by grabbing the fence lock and releasing it again.
Should be replaced by hw handling soon.

Signed-off-by: default avatarChristian König <christian.koenig@amd.com>
CC: stable@vger.kernel.org # 5.19+
Fixes: 5255e146 ("drm/amdgpu: rework TLB flushing")
Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2113


Acked-by: default avatarAlex Deucher <alexander.deucher@amd.com>
Acked-by: default avatarPhilip Yang <Philip.Yang@amd.com>
Tested-by: default avatarStefan Springer <stefanspr94@gmail.com>
Signed-off-by: default avatarAlex Deucher <alexander.deucher@amd.com>
parent a0fba3e0

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h

+15 −0
Original line number Diff line number Diff line
@@ -492,6 +492,21 @@ void amdgpu_debugfs_vm_bo_info(struct amdgpu_vm *vm, struct seq_file *m); 
 */

static inline uint64_t amdgpu_vm_tlb_seq(struct amdgpu_vm *vm)

{

	unsigned long flags;

	spinlock_t *lock;



	/*

	 * Workaround to stop racing between the fence signaling and handling

	 * the cb. The lock is static after initially setting it up, just make

	 * sure that the dma_fence structure isn't freed up.

	 */

	rcu_read_lock();

	lock = vm->last_tlb_flush->lock;

	rcu_read_unlock();



	spin_lock_irqsave(lock, flags);

	spin_unlock_irqrestore(lock, flags);



	return atomic64_read(&vm->tlb_seq);

}