On Thu, Nov 17, 2022 at 05:01:08PM -0700, Nathan Chancellor wrote:
> On Fri, Nov 18, 2022 at 09:06:36AM +1100, Stephen Rothwell wrote:
> > Hi Nathan,
> > 
> > On Thu, 17 Nov 2022 10:29:33 -0700 Nathan Chancellor <nathan@kernel.org> wrote:
> > >
> > > This resolution is not quite right, as pointed out by clang:
> > > 
> > >     drivers/gpu/drm/vc4/vc4_hdmi.c:351:14: error: variable 'vc4_hdmi' is uninitialized when used here [-Werror,-Wuninitialized]
> > >             mutex_lock(&vc4_hdmi->mutex);
> > >                         ^~~~~~~~
> > >     ./include/linux/mutex.h:187:44: note: expanded from macro 'mutex_lock'
> > >     #define mutex_lock(lock) mutex_lock_nested(lock, 0)
> > >                                                ^~~~
> > >     drivers/gpu/drm/vc4/vc4_hdmi.c:322:27: note: initialize the variable 'vc4_hdmi' to silence this warning
> > >             struct vc4_hdmi *vc4_hdmi;
> > >                                      ^
> > >                                       = NULL
> > >     1 error generated.
> > > 
> > > Obviously, the assignment of vc4_hdmi should be before mutex_lock().
> > 
> > Thanks for pointing that out (silly me :-) ).  I have fixed up the
> > resolution for today.
> 
> Great, thank you so much! One less warning to worry about :)

I actually did the same conflict resolution in drm-tip. I've fixed it up
too, thanks for your report :)

Maxime