On (24/02/01 11:05), Sakari Ailus wrote:
[..]
> > > > > if (num_pads >= MEDIA_ENTITY_MAX_PADS)
> > > > > return -E2BIG;
> > > > > @@ -210,15 +211,27 @@ int media_entity_pads_init(struct media_entity *entity, u16 num_pads,
> > > > > media_entity_for_each_pad(entity, iter) {
> > > > > iter->entity = entity;
> > > > > iter->index = i++;
> > > > > +
> > > > > + if (hweight32(iter->flags & (MEDIA_PAD_FL_SINK |
> > > > > + MEDIA_PAD_FL_SOURCE)) != 1) {
> > > > > + ret = -EINVAL;
> > > >
> > > > Can we please have some sort of WARN_ON() or pr_err() here?
> > > > This is a pretty big change.
> > >
> > > Doing proper input validation is hardly anything unusual, is it?
> >
> > Well, function requirements change quite significantly, to the point
> > that drivers that worked before won't work after.
> >
> > > I'm fine with a WARN_ON() though, I'll add that to v8.
> >
> > Thanks!
>
> Actually this was a patchset that was merged quite some time ago. I'll
> post separate patch on this.
Ack.
I just debugged a driver that miraculously stopped working, and it
turned out to be because of this media_entity_pads_init() change.
I think I would have benefited from WARN_ON() or pr_err() there.
