On 10/9/26 17:46, Benjamin Cheng wrote:
> This was supposed to loop over the individual evicted list, as the
> kernel and always_valid evicted lists were already handled in
> amdgpu_vm_validate.
>
> Without this, BOs that were created and evicted before a userq was made
> would end up never end up in an eviction fence rearm.
>
> Fixes: 4cdbba5a16aa ("drm/amdgpu: restructure VM state machine v4")
> Signed-off-by: Benjamin Cheng <[email protected]>
Totally my fault, thanks for catching that.
Reviewed-by: Christian König <[email protected]>
> ---
> drivers/gpu/drm/amd/amdgpu/amdgpu_userq.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.c
> b/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.c
> index a90db4d3d4e9..5b9f7f7f3862 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.c
> @@ -993,8 +993,8 @@ amdgpu_userq_bo_validate(struct amdgpu_device *adev,
> struct drm_exec *exec,
> int ret;
>
> spin_lock(&vm->individual_lock);
> - while (!list_empty(&vm->always_valid.evicted)) {
> - bo_va = list_first_entry(&vm->always_valid.evicted,
> + while (!list_empty(&vm->individual.evicted)) {
> + bo_va = list_first_entry(&vm->individual.evicted,
> struct amdgpu_bo_va,
> base.vm_status);
> spin_unlock(&vm->individual_lock);
