Re: [PATCH v4 2/2] gpiolib: append SFI helpers for GPIO API

[Alex Courbot <acourbot@nvidia.com>]

On 12/09/2013 10:27 PM, Linus Walleij wrote:
> On Mon, Dec 9, 2013 at 2:02 PM, Andy Shevchenko
> <andriy.shevchenko@linux.intel.com> wrote:
>> On Mon, 2013-12-09 at 15:04 +0200, Mika Westerberg wrote:
>>> On Mon, Dec 09, 2013 at 01:50:29PM +0200, Andy Shevchenko wrote:
>
>>>> +struct gpio_desc *sfi_get_gpiod_by_name(const char *name)
>>>> +{
>>>> +   struct sfi_gpio_table_entry *pentry;
>>>> +
>>>> +   pentry = sfi_gpio_get_entry_by_name(name);
>>>> +   if (IS_ERR(pentry))
>>>> +           return pentry;
>>>> +
>>>> +   return gpio_to_desc(pentry->pin_no);
>>>> +}
>>>
>>> Do you really need add a new file just to have the above supported? I mean
>>> that if you place that function to drivers/gpio/gpiolib.c with the
>>> appropriate:
> (...)
>>
>> I won't have linux/sfi.h in the gpiolib.c. If Linus has no objections, I
>> can do that of course.
>
> #ifdef:s are ugly according to Linus (Documentation/CodingStyle)
> and according to this instance of Linus as well. So I actually prefer
> so have a separate file.

Also agree with that. If anything, I'd like to see the DT/ACPI specific 
functions moved into a different file at some point.

> I really like the way this patch set developed to abstract away
> the SFI internals into this gpiolib sidecar, thanks!

Yes - that's a great followup to Mika's work on ACPI and it's nice to 
see this interface adopted (at least on the producer side).