Hi Philipp,

> > On the other hand, the fallback is a really nice feature which could
> > remove duplicated code. But if the fallback is not present by default,
> > it makes it cumbersome to use IMO.
> 
> And it's not easy to automatically determine whether RESET_GPIO is
> actually required, because that depends on both device tree and
> individual drivers.

Well, the driver is small enough to be included always, I'd think.
What about adding this to RESET_GPIO:

	default y if GPIOLIB

?

> In short, if both of_parse_phandle_with_args() return -ENOENT, we
> should continue to silently return (optional ? NULL : -ENOENT), even if
> CONFIG_RESET_GPIO=n.

I think I got it.

> I think the message should be pr_err() level if we return an error that
> will cause the consumer driver probe to fail.

Ok, will send v2 tomorrow.

Thanks for the fast review!

   Wolfram