Re: [PATCH 2/5] gpiolib / ACPI: convert to gpiod interfaces

From: Mika Westerberg
Date: Tue Oct 08 2013 - 04:39:51 EST


On Mon, Oct 07, 2013 at 09:54:13PM -0700, Alexandre Courbot wrote:
> > +struct gpio_desc *acpi_get_gpiod(char *path, int pin)
> > {
> > struct gpio_chip *chip;
> > acpi_handle handle;
> > @@ -48,18 +49,15 @@ int acpi_get_gpio(char *path, int pin)
> >
> > status = acpi_get_handle(NULL, path, &handle);
> > if (ACPI_FAILURE(status))
> > - return -ENODEV;
> > + return ERR_PTR(-ENODEV);
> >
> > chip = gpiochip_find(handle, acpi_gpiochip_find);
> > if (!chip)
> > - return -ENODEV;
> > + return ERR_PTR(-ENODEV);
> >
> > - if (!gpio_is_valid(chip->base + pin))
> > - return -EINVAL;
> > -
> > - return chip->base + pin;
> > + return gpio_to_desc(chip->base + pin);
>
> I think you want to avoid using gpio_to_desc(). It is just a
> convenience function provided to ease the transition for consumers
> that still need to rely on GPIO numbers for some reason. The same
> applies to desc_to_gpio(), although the usage you are making of it
> (implemented fallbacks for the integer space) is the one that is
> intended.
>
> The last line could be changed to "return &chip->desc[pin];" after
> checking that 0 <= pin <= chip->ngpio.

I tried that but it doesn't compile anymore :-(

drivers/gpio/gpiolib-acpi.c: In function âacpi_get_gpiodâ:
drivers/gpio/gpiolib-acpi.c:61:2: error: invalid use of undefined type âstruct gpio_descâ
drivers/gpio/gpiolib-acpi.c:61:20: error: dereferencing pointer to incomplete type

Since struct gpio_desc is internal to gpiolib.c we can't dereference it
outside.

The DT version also uses gpio_to_desc().

> This minor issue apart, I really like the fact you are moving this
> under the gpiolib APIs (also appreciate the testing of the new API
> this patchset provides!).
>
> I also wonder whether you could also avoid exporting acpi_get_gpiod*
> (which allow GPIO consumers to shortcut gpiolib) by implementing
> acpi_get_gpio* into the C file (maybe even using gpiod_get()). I see a
> "char *path" parameter though, so maybe that would not be possible at
> the moment.

Good point, I'll check if we can do that.

> Once the gpio_to_desc() issue mentioned above is fixed I think I can give my
>
> Reviewed-by: Alexandre Courbot <acourbot@xxxxxxxxxx>

Thanks!

>
> on patches 2-5.
>
> Thanks,
> Alex.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/