Commit fcf00f8d authored by Xiaogang Chen's avatar Xiaogang Chen Committed by Alex Deucher
Browse files

drm/amdkfd: Remove skiping userptr buffer mapping when mmu notifier marks it as invalid 



mmu notifier does not always hold mm->sem during call back. That causes
a race condition between kfd userprt buffer mapping and mmu notifier
which leds to gpu shadder or SDMA access userptr buffer before it has been
mapped to gpu VM. Always map userptr buffer to avoid that though it may make
some userprt buffers mapped two times.

Suggested-by: default avatarFelix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: default avatarXiaogang Chen <xiaogang.chen@amd.com>
Reviewed-by: default avatarFelix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: default avatarAlex Deucher <alexander.deucher@amd.com>
parent 79b72db6

drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c

+0 −10
Original line number Diff line number Diff line
@@ -1907,16 +1907,6 @@ int amdgpu_amdkfd_gpuvm_map_memory_to_gpu( 
	 */

	mutex_lock(&mem->process_info->lock);



	/* Lock mmap-sem. If we find an invalid userptr BO, we can be

	 * sure that the MMU notifier is no longer running

	 * concurrently and the queues are actually stopped

	 */

	if (amdgpu_ttm_tt_get_usermm(bo->tbo.ttm)) {

		mmap_write_lock(current->mm);

		is_invalid_userptr = atomic_read(&mem->invalid);

		mmap_write_unlock(current->mm);

	}



	mutex_lock(&mem->lock);



	domain = mem->domain;