[Intel-gfx] [PATCH 04/26] drm/rcar-du: Use for_each_*_in_state

Daniel Vetter daniel at ffwll.ch
Mon May 30 14:54:10 UTC 2016


On Mon, May 30, 2016 at 11:58:27AM +0200, Maarten Lankhorst wrote:
> Op 30-05-16 om 11:18 schreef Laurent Pinchart:
> > Hi Daniel,
> >
> > Thank you for the patch.
> >
> > This looks good to me as the resulting code is mostly similar. However, the 
> > for_each_*_in_state macros end with an for_each_if() that tests if the 
> > object's state is NULL, which isn't present in this code. I'm wondering 
> > whether that was an oversight on my side possibly leading to a crash when 
> > dereferencing a NULL state, or an unneeded check in the macros. Can 
> > atomic_state->*_states[i] be NULL if atomic_state->*[i] is not NULL ?
> Not in any normal case.

Yeah, the drm_atomic_get_*_state functions only ever fill in both of
neither. If this gets out of sync it's a bug ;-)
-Daniel
-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch


More information about the dri-devel mailing list