On Mon, Feb 10, 2020 at 06:25:35AM -0800, Guenter Roeck wrote: > On 2/10/20 2:22 AM, Linus Walleij wrote: > >The WM831x watchdog driver passes a global GPIO number from > >platform data into this driver, this is discouraged so pass > >a GPIO descriptor instead. > > > >More thorough approaches are possible passing descriptors > >associated with the device through machine descriptor tables, > >but no boardfiles in the kernel currently use this driver > >so it is hard to test. > > > >Cc: Richard Fitzgerald <rf@xxxxxxxxxxxxxxxxxxxxx> > >Cc: Charles Keepax <ckeepax@xxxxxxxxxxxxxxxxxxxxx> > >Cc: Mark Brown <broonie@xxxxxxxxxx> > >Signed-off-by: Linus Walleij <linus.walleij@xxxxxxxxxx> > > Interesting, I don't see evidence of the original patch in > watchdog patchwork. > > Anyway, it seems to me it would be better to remove the gpio code > entirely from this driver. It is instantiated from an mfd driver > which doesn't set the gpio pin. It is quite unlikely that it is > ever going to be used, so we might as well remove it (instead of > modifying it without ability to test it). > We don't have any internal users for the WDT GPIOs here, given the age of the part perhaps removing the GPIO is the simplest solution. Thanks, Charles
