Re: [PATCH v1 0/3] gpio: Introduce helper to dispose IRQ mappings

[Andy Shevchenko]

On Thu, Oct 22, 2026 at 07:05:55AM -0700, Bartosz Golaszewski wrote:
> On Thu, 22 Oct 2026 15:32:44 +0200, Andy Shevchenko
> <andriy.shevchenko@xxxxxxxxxxxxxxx> said:
> > On Thu, Oct 22, 2026 at 02:11:38AM -0700, Bartosz Golaszewski wrote:
> >> On Wed, 21 Oct 2026 22:44:42 +0200, Andy Shevchenko
> >> <andriy.shevchenko@xxxxxxxxxxxxxxx> said:
> >> > I have noticed that a few drivers use almost the same code to dispose IRQ
> >> > mappings. This series to deduplicate the efforts. I only converted a couple
> >> > of the users (there are potentially more) in order to stay in the same
> >> > subsystem.
> >> The "few" being 2 testing drivers of which one (mockup) should ideally go away.
> >> Do we really want to add more code to GPIO core that will only be used in
> >> testing module(s)? I don't think this is useful, sorry.
> >
> > I meant a few. As I pointed these were chosen explicitly for the sake of not
> > crossing the subsystem boundaries. I looked at Broadcom one, it is so big mess
> > but I think I can get an advantage on it. Also there are other drivers that
> > just need more work to check (and many of them are outside of drivers/gpio),
> > but I'm pretty sure the new API has a good potential.
> 
> Would it be possible to provide at least one more user other than the testing
> drivers?

Sure.

-- 
With Best Regards,
Andy Shevchenko