On Mon, Nov 4, 2019 at 10:33 PM Uwe Kleine-König <u.kleine-koenig@xxxxxxxxxxxxxx> wrote: > On Mon, Nov 04, 2019 at 04:32:23PM +0100, Bartosz Golaszewski wrote: > > pon., 4 lis 2019 o 16:11 Miquel Raynal <miquel.raynal@xxxxxxxxxxx> napisał(a): > > > Andy Shevchenko <andy.shevchenko@xxxxxxxxx> wrote on Tue, 15 Oct 2019 > > > 17:55:33 +0300: > > > > > > > Or other way around. PWM registers GPIO (which actually I prefer since > > > > we have PCA9685 case where PWM provides GPIO functionality, though via > > > > different means) > While it's not nice to have a driver that provides two different devices > (here: gpio controller and pwm controller) similar things are not > unseen. And for example the splitting of watchdog > (drivers/watchdog/stmp3xxx_rtc_wdt.c) and rtc > (drivers/rtc/rtc-stmp3xxx.c) of the device in the mx28 is more trouble > than worth. > > So I'd vote for putting it in a single file that lives where the > bigger/more complex part fits to. So assuming that's the GPIO part (as > the driver supports several variants and not all of them have a PWM > function if I'm not mistaken) having it in drivers/gpio is fine for me. For me it sounds more likely that PWM is a *pin function* of a pin controller and actually this GPIO driver should be a pin controller with corresponding function(s). -- With Best Regards, Andy Shevchenko