On Tue, Oct 31, 2023 at 10:45:57AM +0200, Jouni Högander wrote:
>       intel_atomic_commit_tail(state);
>  }
>  
> -static int
> -intel_atomic_commit_ready(struct i915_sw_fence *fence,
> -                       enum i915_sw_fence_notify notify)
> -{
> -     struct intel_atomic_state *state =
> -             container_of(fence, struct intel_atomic_state, commit_ready);
> -
> -     switch (notify) {
> -     case FENCE_COMPLETE:
> -             /* we do blocking waits in the worker, nothing to do here */
> -             break;
> -     case FENCE_FREE:
> -             {
> -                     struct drm_i915_private *i915 = 
> to_i915(state->base.dev);
> -                     struct intel_atomic_helper *helper =
> -                             &i915->display.atomic_helper;
> -
> -                     if (llist_add(&state->freed, &helper->free_list))
> -                             queue_work(i915->unordered_wq, 
> &helper->free_work);

BTW looks like we're left with a bunch of dead code
around this free_list stuff...

-- 
Ville Syrjälä
Intel

Reply via email to