On Wed, Jul 23, 2014 at 03:38:17PM -0400, Rob Clark wrote:
...
> @@ -2702,10 +2895,11 @@ static int drm_mode_cursor_universal(struct drm_crtc 
> *crtc,
>        * setplane_internal will take care of deref'ing either the old or new
>        * framebuffer depending on success.
>        */
> -     ret = setplane_internal(crtc->cursor, crtc, fb,
> +     ret = setplane_internal(crtc->cursor,
> +                             crtc->base.id, fb ? fb->base.id : 0,
>                               crtc_x, crtc_y, crtc_w, crtc_h,
>                               0, 0, src_w, src_h);
> -
> +     drm_framebuffer_unreference(fb);

I haven't had time to look through this in depth yet, but can't we wind
up unreferencing NULL here?

Also, the comment above this call should probably be updated/removed
since the locking works a bit differently now?


Matt

-- 
Matt Roper
Graphics Software Engineer
IoTG Platform Enabling & Development
Intel Corporation
(916) 356-2795

Reply via email to