Hi Laurent, Jacopo,
On Wed, Feb 23, 2022 at 04:36:48PM +0200, Laurent Pinchart wrote:
> > > > @@ -3589,13 +3630,19 @@ static int ov5640_enum_mbus_code(struct v4l2_subdev *sd,
> > > > struct v4l2_subdev_state *sd_state,
> > > > struct v4l2_subdev_mbus_code_enum *code)
> > > > {
> > > > - if (code->pad != 0)
> > > > - return -EINVAL;
> > >
> > > You've lost this check.
> > >
> >
> > The driver has a single pad, doesn't core verify that the pad argument
> > is valid ?
>
> Only when called from userspace, not when called from another subdev.
This was historically true but changed in 2019, as the checks when calling
the subdev ops from the kernel and userspace were aligned. This was done by
commit a8fa55078a7784a99a2ce389b5d7456a3be9a941 .
--
Regards,
Sakari Ailus
