[PATCH/RFC] gpio: em: Use dynamic allocation of GPIOs

[PATCH/RFC] gpio: em: Use dynamic allocation of GPIOs

[Geert Uytterhoeven]

Use dynamic allocation of GPIOs instead of looking at the gpio%u alias
in DT.
---
  - Is this correct? Not having to care about the alias would simplify the
    to-be-written DT binding documentation.
  - Completely untested.

Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>

 drivers/gpio/gpio-em.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)

diff --git a/drivers/gpio/gpio-em.c b/drivers/gpio/gpio-em.c
index 21d34d4d473dcefe..c3434146f605748b 100644
--- a/drivers/gpio/gpio-em.c
+++ b/drivers/gpio/gpio-em.c
@@ -330,12 +330,7 @@ static int em_gio_probe(struct platform_device *pdev)
 			goto err0;
 		}
 
-		ret = of_alias_get_id(pdev->dev.of_node, "gpio");
-		if (ret < 0) {
-			dev_err(&pdev->dev, "Couldn't get OF id\n");
-			goto err0;
-		}
-		pdata->gpio_base = ret * 32; /* 32 GPIOs per instance */
+		pdata->gpio_base = -1;
 	}
 
 	gpio_chip = &p->gpio_chip;
-- 
1.9.1

Re: [PATCH/RFC] gpio: em: Use dynamic allocation of GPIOs

[Arnd Bergmann]

On Monday 17 November 2014 15:30:32 Geert Uytterhoeven wrote:
> Use dynamic allocation of GPIOs instead of looking at the gpio%u alias
> in DT.
> ---
>   - Is this correct? Not having to care about the alias would simplify the
>     to-be-written DT binding documentation.
>   - Completely untested.
> 
> Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
> 

This will obviously break any user that expects a particular gpio
number, in particular board files, but I guess also user space
that tries to access the gpios through sysfs.

If you don't have any of those, I guess it's correct.

	Arnd

Re: [PATCH/RFC] gpio: em: Use dynamic allocation of GPIOs

[Geert Uytterhoeven]

Hi Arnd,

On Mon, Nov 17, 2014 at 3:34 PM, Arnd Bergmann <arnd@arndb.de> wrote:
> On Monday 17 November 2014 15:30:32 Geert Uytterhoeven wrote:
>> Use dynamic allocation of GPIOs instead of looking at the gpio%u alias
>> in DT.
>> ---
>>   - Is this correct? Not having to care about the alias would simplify the
>>     to-be-written DT binding documentation.
>>   - Completely untested.
>>
>> Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
>>
>
> This will obviously break any user that expects a particular gpio
> number, in particular board files, but I guess also user space
> that tries to access the gpios through sysfs.
>
> If you don't have any of those, I guess it's correct.

Emev2/kzm9d was the first DT-only shmobile SoC/board, so the kernel
part should be OK.

No idea about userspace.

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

Re: [PATCH/RFC] gpio: em: Use dynamic allocation of GPIOs

[Alexandre Courbot]

On Mon, Nov 17, 2014 at 11:30 PM, Geert Uytterhoeven
<geert+renesas@glider.be> wrote:
> Use dynamic allocation of GPIOs instead of looking at the gpio%u alias
> in DT.
> ---
>   - Is this correct? Not having to care about the alias would simplify the
>     to-be-written DT binding documentation.
>   - Completely untested.
>
> Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
>
>  drivers/gpio/gpio-em.c | 7 +------
>  1 file changed, 1 insertion(+), 6 deletions(-)
>
> diff --git a/drivers/gpio/gpio-em.c b/drivers/gpio/gpio-em.c
> index 21d34d4d473dcefe..c3434146f605748b 100644
> --- a/drivers/gpio/gpio-em.c
> +++ b/drivers/gpio/gpio-em.c
> @@ -330,12 +330,7 @@ static int em_gio_probe(struct platform_device *pdev)
>                         goto err0;
>                 }
>
> -               ret = of_alias_get_id(pdev->dev.of_node, "gpio");
> -               if (ret < 0) {
> -                       dev_err(&pdev->dev, "Couldn't get OF id\n");
> -                       goto err0;
> -               }
> -               pdata->gpio_base = ret * 32; /* 32 GPIOs per instance */
> +               pdata->gpio_base = -1;

User-space might break because of GPIO renumbering. Why not setting
gpio_base to -1 when the property is not present (instead of
triggering an error), keeping support for the property so existing
boards remain safe, and marking the property as deprecated in the
bindings documentation?

Re: [PATCH/RFC] gpio: em: Use dynamic allocation of GPIOs

[Linus Walleij]

On Tue, Nov 18, 2014 at 4:11 AM, Alexandre Courbot <gnurou@gmail.com> wrote:
> On Mon, Nov 17, 2014 at 11:30 PM, Geert Uytterhoeven
> <geert+renesas@glider.be> wrote:

>> -               ret = of_alias_get_id(pdev->dev.of_node, "gpio");
>> -               if (ret < 0) {
>> -                       dev_err(&pdev->dev, "Couldn't get OF id\n");
>> -                       goto err0;
>> -               }
>> -               pdata->gpio_base = ret * 32; /* 32 GPIOs per instance */
>> +               pdata->gpio_base = -1;
>
> User-space might break because of GPIO renumbering. Why not setting
> gpio_base to -1 when the property is not present (instead of
> triggering an error), keeping support for the property so existing
> boards remain safe, and marking the property as deprecated in the
> bindings documentation?

I don't like the way aliases are used here.

These bindings are undocumented by the way :( :( :(

I've applied it to wait and see what happens instead.

Yours,
Linus Walleij

Re: [PATCH/RFC] gpio: em: Use dynamic allocation of GPIOs

[Linus Walleij]

On Mon, Nov 17, 2014 at 3:30 PM, Geert Uytterhoeven
<geert+renesas@glider.be> wrote:

> Use dynamic allocation of GPIOs instead of looking at the gpio%u alias
> in DT.
> ---
>   - Is this correct? Not having to care about the alias would simplify the
>     to-be-written DT binding documentation.
>   - Completely untested.
>
> Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>

Patch applied.

Yours,
Linus Walleij