Re: [PATCH v5 08/14] mfd: lm3533: Convert to use OF bindings

ср, 22 лип. 2026 р. о 04:30 Jonathan Cameron <[email protected]> пише:
>
> On Tue, 14 Jul 2026 16:57:01 +0300
> Svyatoslav Ryhel <[email protected]> wrote:
>
> > пт, 3 лип. 2026 р. о 14:03 Johan Hovold <[email protected]> пише:
> > >
> > > On Wed, Jun 17, 2026 at 11:00:25AM +0300, Svyatoslav Ryhel wrote:
> > > > Since there are no users of this driver via platform data, remove the
> > > > platform data support and switch to using Device Tree bindings.
> > > >
> > > > Signed-off-by: Svyatoslav Ryhel <[email protected]>
> > > > Reviewed-by: Daniel Thompson (RISCstar) <[email protected]> #for backlight
> > > >       indio_dev->name = dev_name(&pdev->dev);
> > > > -     iio_device_set_parent(indio_dev, pdev->dev.parent);
> > >
> > > Why are you reparenting the iio device here?
> > >
> >
> > Because every cell has its own binding now and using phandle to parent
> > when device has its own node is not a good practice.
>
> It is however a good point.  So should at very least the ABI break should
> be called out in the patch description with a brief mention of what the
> path change actually is.
>

Ok, this is a good idea to add chapter into commit description.

> >
> > > That's an ABI break.
> > >
> >
> > This driver does not have any active users in the kernel and no
> > activity for more then 2 years.
>
> ...
>
> > > > +     u32 count = 0, reg, nchilds;
> > >
> > > Don't mix multiple declarations with initialisation like this.
> > >
> >
> > Checkpatch does not complain on style issue, hence this is not prohibited.
>
> Prohibition vs taste.  I fully agree with John on this one!  Those mixes are
> messy and hard to read.
>
> Checkpatch is a helpful tool but there are both rules and preferred style things
> it does not catch.
>

Ok, fine.

> Jonathan
>
> >
> > > >       int ret;
> > > >