Re: [PATCH v2] drm/i915/gt: Do not destroy uninitialized timer

[Jani Nikula <jani.nikula@linux.intel.com>]

On Thu, 01 Oct 2026, Krzysztof Karas <krzysztof.karas@intel.com> wrote:
> If intel_ring_begin() fails in live_emit_pte_full_ring(), then
> cleanup is performed on an uninitialized timer.
>
> Prevent that by jumping to a dedicated label for driver removal
> cleanup.
>
> Assisted-by: Copilot-Claude-Opus-5

Nowadays it's sufficient to state, "Assisted-by: LLM".

> Suggested-by: Jani Nikula <jani.nikula@linux.intel.com>
> Signed-off-by: Krzysztof Karas <krzysztof.karas@intel.com>
> ---
>  drivers/gpu/drm/i915/gt/selftest_migrate.c | 9 +++++----
>  1 file changed, 5 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/gt/selftest_migrate.c b/drivers/gpu/drm/i915/gt/selftest_migrate.c
> index 5b986a7687d9..5e3a9b5a83eb 100644
> --- a/drivers/gpu/drm/i915/gt/selftest_migrate.c
> +++ b/drivers/gpu/drm/i915/gt/selftest_migrate.c
> @@ -651,17 +651,18 @@ static int live_emit_pte_full_ring(void *arg)
>  	len = emit_pte(rq, &it, obj->pat_index, false, 0, CHUNK_SZ);
>  	if (!len) {
>  		err = -EINVAL;
> -		goto out_rq;
> +		goto out_timer;
>  	}
>  	if (len < 0) {
>  		err = len;
> -		goto out_rq;
> +		goto out_timer;
>  	}
>  
> -out_rq:
> -	i915_request_add(rq); /* GEM_BUG_ON(rq->reserved_space > ring->space)? */
> +out_timer:
>  	timer_delete_sync(&st.timer);
>  	timer_destroy_on_stack(&st.timer);
> +out_rq:
> +	i915_request_add(rq); /* GEM_BUG_ON(rq->reserved_space > ring->space)? */

This seems like the proper thing to do.

I can't think of any reason why changing the order would be bad, but
then this is not really my area of expertise. With that,

Reviewed-by: Jani Nikula <jani.nikula@intel.com>


>  out_unpin:
>  	intel_context_unpin(ce);
>  out_put:

-- 
Jani Nikula, Intel