Re: [PATCH] gpio: pxa: make controllers without pinctrl work again

[Robert Jarzmik <robert.jarzmik@xxxxxxx>]

Daniel Mack <daniel@xxxxxxxxxx> writes:

> Hi,
>
> I have to get back to this one. The fall-out that made me implement the patch
> below in the first place was fixed with the following patch:
>
>   e41f76830d (ARM: pxa: dts: add gpio-ranges to gpio controller)
>
> ... but now another issue came up.
>
> To recap, the pxa3xx driver uses the pinctrl-single driver since a while which
> does not implement a .gpio_set_direction() callback. The pinmux core will simply
> return 0 in this case, and the pxa3xx gpio driver hence thinks the pinctrl
> driver did its job and returns as well.
>
> This effectively makes pxa_gpio_direction_{input,output} no-ops. Interestingly,
> I only noticed that when I tried to use the w1-gpio driver which makes use of
> gpiolib's open-drain emulation. All other use-cases work just fine.
>
> So the question is how this is supposed to work. I see two options:
>
> 1) Apply the patch below, so we don't use use the pinctrl driver on PXA3xx.
>
> 2) Change pinmux_gpio_direction() and make it return -ENOSYS or something in
> case there is no .gpio_set_direction() implementation. I'm not sure whether
> that's a good idea though, and what the side-effects would be.

Linus, this question is rather for you : is it normal that whether an
pincontroller is available or not, pinctrl_gpio_direction_input() returns 0 (as
stated in the function pinmux_gpio_direction()).

If that is the case, how a gpio driver can know if the gpio direction will be
handled in the pin controller and not take action itself ?

As for Daniel, there is another solution, ie remove this from
pxa_gpio_direction_input() and pxa_gpio_direction_output():
	if (!ret)
		return 0;

Cheers.

-- 
Robert
--
To unsubscribe from this list: send the line "unsubscribe linux-gpio" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html