Hi Andy,
On Wed, Nov 29, 2023 at 3:23 PM Andy Shevchenko
<andriy.shevchenko@xxxxxxxxxxxxxxx> wrote:
> On Wed, Nov 29, 2023 at 03:41:55PM +0200, Andy Shevchenko wrote:
> > On Wed, Nov 29, 2023 at 12:21:45PM +0100, Geert Uytterhoeven wrote:
> > > On Tue, Nov 28, 2023 at 9:04 PM Andy Shevchenko
> > > <andriy.shevchenko@xxxxxxxxxxxxxxx> wrote:
> > > > It's unclear why it's not a const from day 1. Make the pins member
> > > > const in struct group_desc. Update necessary APIs.
>
> ...
>
> > > > int pinctrl_generic_add_group(struct pinctrl_dev *pctldev, const char *name,
> > > > - int *gpins, int ngpins, void *data);
> > > > + const int *pins, int num_pins, void *data);
> > > >
> > > > int pinctrl_generic_remove_group(struct pinctrl_dev *pctldev,
> > > > unsigned int group_selector);
> > >
> > > Probably this is also the right moment to change all of these to arrays
> > > of unsigned ints? Else you will have mixed int/unsigned int after
> > > "[PATCH v3 13/22] pinctrl: core: Embed struct pingroup into struct
> > > group_desc", and purely unsigned int after "[PATCH v3 22/22] pinctrl:
> > > core: Remove unused members from struct group_desc".
> >
> > Hmm... Can it be done later?
> >
> > I can, of course try to change the parameter here to be unsigned, but it most
> > likely fail the build for those drivers means need more patches, more delay to
> > this series.
> >
> > Linus?
>
> On the first glance updating API here does not fail the build.
That's what I had expected, as drivers already pass int or unsigned int
arrays anyway.
> Lemme incorporate this into v4.
Thanks!
> Meanwhile the drivers I left untouched, it might be separate changes
> to convert from int to const unsigned int.
Sounds fine to me.
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@xxxxxxxxxxxxxx
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
