Hi Maxime,

On Thu, 22 May 2025 17:43:37 +0200
Maxime Ripard <[email protected]> wrote:

> On Fri, May 16, 2025 at 06:48:38PM +0200, Luca Ceresoli wrote:
> > Some users of DRM bridges may need to execute specific code just before
> > deallocation.
> > 
> > As of now the only known user would be KUnit tests.
> > 
> > Suggested-by: Maxime Ripard <[email protected]>
> > Signed-off-by: Luca Ceresoli <[email protected]>
> > 
> > ---
> > 
> > This patch is new in v8. The .destroy callback had appeared in v5 as well
> > [5], but as part of a larger patch and for different reason that do not
> > apply anymore.
> > 
> > [5] 
> > https://lore.kernel.org/all/[email protected]/#t
> > ---
> >  drivers/gpu/drm/drm_bridge.c |  2 ++
> >  include/drm/drm_bridge.h     | 10 ++++++++++
> >  2 files changed, 12 insertions(+)
> > 
> > diff --git a/drivers/gpu/drm/drm_bridge.c b/drivers/gpu/drm/drm_bridge.c
> > index 
> > b4c89ec01998b849018ce031c7cd84614e65e710..6185cb29fe3162264f0912c09c205fb467975dee
> >  100644
> > --- a/drivers/gpu/drm/drm_bridge.c
> > +++ b/drivers/gpu/drm/drm_bridge.c
> > @@ -203,6 +203,8 @@ static void __drm_bridge_free(struct kref *kref)
> >  {
> >     struct drm_bridge *bridge = container_of(kref, struct drm_bridge, 
> > refcount);
> >  
> > +   if (bridge->funcs->destroy)
> > +           bridge->funcs->destroy(bridge);
> >     kfree(bridge->container);
> >  }
> >  
> > diff --git a/include/drm/drm_bridge.h b/include/drm/drm_bridge.h
> > index 
> > 4e418a29a9ff9d014d6ac0910a5d9bcf7118195e..3ccd493faa580845c2ed1166f398eca27b464261
> >  100644
> > --- a/include/drm/drm_bridge.h
> > +++ b/include/drm/drm_bridge.h
> > @@ -86,6 +86,16 @@ struct drm_bridge_funcs {
> >      */
> >     void (*detach)(struct drm_bridge *bridge);
> >  
> > +   /**
> > +    * @destroy:
> > +    *
> > +    * This callback is invoked when the bridge is about to be
> > +    * deallocated.
> > +    *
> > +    * The @destroy callback is optional.
> > +    */
> > +   void (*destroy)(struct drm_bridge *bridge);
> > +  
> 
> destroy is before detach in alphabetical order, but otherwise it looks
> good to me.

I saw the struct is not alpha-ordered right now, so I did not get it
should be, and it looked like keeping .attach and .detach nearby would
be good.

> Once fixed,
> Acked-by: Maxime Ripard <[email protected]>

OK, will send new iteration with .destroy before .detach, thanks for
the review.

Luca

-- 
Luca Ceresoli, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com

Reply via email to