Re: [PATCH] gpio: aggregator: constify struct configfs_attribute

[Thomas Weißschuh]

On Thu, Oct 29, 2026 at 04:43:44AM -0400, Bartosz Golaszewski wrote:
> On Thu, 29 Oct 2026 09:41:02 +0100, Bartosz Golaszewski <brgl@xxxxxxxxxx> said:
> > On Thu, 29 Oct 2026 08:57:15 +0100, Geert Uytterhoeven <geert@xxxxxxxxxxxxxx> said:
> >> Thanks for your patch!
> >>
> >> On Wed, 21 Oct 2026 at 14:31, Bartosz Golaszewski
> >> <bartosz.golaszewski@xxxxxxxxxxxxxxxx> wrote:
> >>> In preparation for the constification of configfs attributes, make the
> >>> existing instances of struct configfs_attribute in gpio-aggregator const
> >>> and use the new st_attrs_const field of struct config_item_type.
> >>
> >> This has a hard dependency on commit f7528cd38b87fb59 ("configfs:
> >> Allow the registration of const struct configfs_attribute"), which is
> >> not yet in gpio/for-next, thus  breaking your build.
> >>
> >>> Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@xxxxxxxxxxxxxxxx>
> >>
> >> Since you have to rebase anyway ;-)
> >> Reviewed-by: Geert Uytterhoeven <geert+renesas@xxxxxxxxx>
> >>
> >
> > Ah, dang it. :( I stupidly assumed it must be in mainline when Thomas
> > recommended it. :(
> >
> > Thomas: Unless there's an immutable branch I could pull, I'll need to back all
> > of them out of my branch and have them wait for the next cycle I suppose?
> >
> > Bart
> >
> 
> Ah, seems like commit
> 
> commit d4d9eb226e4c2af8ab390bb470477fa68072fa2a
> Merge: 32543b09681a1 6c1fa55e287a6
> Author: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
> Date:   Tue Oct 20 16:25:43 2026 -1000
> 
>     Merge tag 'configfs-7.4-rc1' of
> git://git.kernel.org/pub/scm/linux/kernel/git/leitao/linux
> 
>     Pull configfs updates from Breno Leitao:
> 
>      - Attribute structures can now be const (Thomas Weißschuh)
> 
> pulled it into Linus' branch so it's just a matter of post-poning them until
> v7.4-rc1.

This is what I meant with "the necessary infrastructure will be part of the
next merge window" when suggesting this change. I could have made that clearer.

Early adopter issues :-)


Thomas