Re: [PATCH] Input: gpio_keys: allocate pins

[Daniel Mack <zonque@gmail.com>]

On 25.10.2012 10:08, Linus Walleij wrote:
> On Thu, Oct 25, 2012 at 9:58 AM, Lee Jones <lee.jones@linaro.org> wrote:
>> On Fri, 12 Oct 2012, Linus Walleij wrote:
> (...)
>>> Is biasing what you need to do?
> (...)
>>> All I really want is that platforms have a clear idea about
>>> how and where the pins will be handled, and that if GPIO
>>> and pinctrl handle the same lines, they need to interact.
>>>
>>> Yours,
>>> Linus Walleij
>>
>> Friendly poke.
> 
> I don't know how to respond to that? I asked a question about
> what the intent of the patch was and the generic thinking
> behind this approach and it remains unanswered.
> 
> I think I have seen other patches doing the proper thing
> for pinctrl-single by implementing the proper
> pinctrl_request_gpio()
> pinctrl_free_gpio()
> pinctrl_gpio_direction_input()
> pinctrl_gpio_direction_output()
> in that very GPIO driver.
> 
> So I suspect that this patch should be dropped, unless you
> have some other compelling usecase to bring to the show?

I think so too. I misunderstood the concept of pinctrl in this area,
thanks for taking the time of explaining this. Shortly after I worked on
it, I was distracted by other topics so I didn't find time to continue.
But once I will, I'll follow up here again.


Thanks,
Daniel