Re: [PATCH v2 4/7] gpiolib: remove gpio_set_debounce

[Andy Shevchenko <andriy.shevchenko@linux.intel.com>]

On Wed, Feb 01, 2023 at 11:07:35AM -0800, Dmitry Torokhov wrote:
> On Wed, Feb 01, 2023 at 08:32:06PM +0200, Andy Shevchenko wrote:
> > On Tue, Jan 31, 2023 at 09:44:31PM -0800, Dmitry Torokhov wrote:
> > > On Fri, Jan 27, 2023 at 11:11:46AM +0100, Arnd Bergmann wrote:

...

> > > > -	return !gpio_get_value(ts->gpio_pendown);
> > > > +	return !gpiod_get_value(ts->gpio_pendown);
> > > 
> > > No, we can not blindly do that without checking annotations on GPIOs.
> > 
> > But this is easy to fix, i.e. use raw API, no?
> 
> I'd rather not (I hope I can make this driver respect declared polarity
> at some point), so for debounce we could do:
> 
> 	gpiod_set_debounce(gpio_to_gpiod(), ...);
> 
> in ads7846 for now, and get rid of gpio_set_debounce() as a publc API.

This will work and we can keep it for a while (gpio_to_desc(), I believe you
meant this one, is part of the new API to keep this bridge for the cases like
this).

Arnd, are you going to send a v3? It would be really nice to have less
collisions next cycle if your series is applied.

-- 
With Best Regards,
Andy Shevchenko