Re: [PATCH] Input: gpio_keys: allocate pins

[Daniel Mack <zonque@gmail.com>]

On 12.10.2012 23:37, Linus Walleij wrote:
> On Fri, Oct 12, 2012 at 11:27 PM, Daniel Mack <zonque@gmail.com> wrote:
>> On 12.10.2012 23:26, Linus Walleij wrote:
>>> On Fri, Oct 12, 2012 at 5:55 PM, Daniel Mack <zonque@gmail.com> wrote:
>>>
>>>> This allows DT driven boards to allocate and configure the pinmux once
>>>> the driver is probed.
>>>>
>>>> Signed-off-by: Daniel Mack <zonque@gmail.com>
>>>> Cc: Linus Walleij <linus.walleij@linaro.org>
>>>> Cc: Dmitry Torokhov <dmitry.torokhov@gmail.com>
>>>
>>> (...)
>>>> +       /* request pin mux */
>>>> +       pinctrl = devm_pinctrl_get_select_default(dev);
>>>> +       if (IS_ERR(pinctrl))
>>>> +               dev_warn(dev, "pins are not configured from the driver\n");
>>>
>>> I think dev_warn() is rather nasty to throw in here, dev_info() is OK.
>>
>> Well, dev_warn is used everywhere else for this particular warning, but
>> I can change that of course.
> 
> Yeah true ... just that I think that in this case most platforms will
> rely on the fall through mechanism where the gpio_get()
> and set_direction will fall through to the special-purpose functions
> in the gpio driver calling out to pinctrl. So it's actually not an error.
> 
> This pinctrl handle is only intended for biasing pins etc, right?
> 
> NO muxing!
> 
> Because we wrote in Documentation/pinctrl.txt that if GPIO
> and pin control handle the same lines, they should be
> implemented in the gpio driver by calling out to pinctrl's
> extern int pinctrl_request_gpio(unsigned gpio);
> extern void pinctrl_free_gpio(unsigned gpio);
> extern int pinctrl_gpio_direction_input(unsigned gpio);
> extern int pinctrl_gpio_direction_output(unsigned gpio);

Hmm. So how is a certain pin muxed to its GPIO function then? And how
can pullup/pulldown features be selected?

I admittedly might lack some background here, and if there's better
solution to what I want to do, I'd be happy to hear about it :)


Many thanks,
Daniel