Re: [PATCH 3/6] pinctrl: armada-37xx: Add gpio support

[Gregory CLEMENT <gregory.clement@xxxxxxxxxxxxxxxxxx>]

Hi Linus,
 
 On ven., déc. 30 2016, Linus Walleij <linus.walleij@xxxxxxxxxx> wrote:

> On Thu, Dec 22, 2016 at 6:24 PM, Gregory CLEMENT
> <gregory.clement@xxxxxxxxxxxxxxxxxx> wrote:
>
>> GPIO management is pretty simple and is part of the same IP than the pin
>> controller for the Armada 37xx SoCs.  This patch adds the GPIO support to
>> the pinctrl-armada-37xx.c file, it also allows sharing common functions
>> between the gpiolib and the pinctrl drivers.
>>
>> Signed-off-by: Gregory CLEMENT <gregory.clement@xxxxxxxxxxxxxxxxxx>
>
> Some remarks:
>
>> +static int armada_37xx_gpio_get_direction(struct gpio_chip *chip,
>> +                                         unsigned int offset)
>> +{
>> +       struct armada_37xx_pinctrl *info = gpiochip_get_data(chip);
>> +       unsigned int reg = OUTPUT_EN;
>> +       unsigned int val, mask;
>> +
>> +       if (offset >= GPIO_PER_REG) {
>> +               offset -= GPIO_PER_REG;
>> +               reg += sizeof(u32);
>> +       }
>
> Add a comment saying we never have more than two registers?
> If there would be three registers this would fail, right?

I added the comment

>
>> +       mask = BIT(offset);
>> +
>> +       regmap_read(info->regmap, reg, &val);
>>
>> +       return (val & mask) == 0;
>
> Use this:
>
> return !(val & mask);

done but I could tou explain the advantage of doing it?

>
>> +static int armada_37xx_gpio_get(struct gpio_chip *chip, unsigned int offset)
>> +{
>> +       struct armada_37xx_pinctrl *info = gpiochip_get_data(chip);
>> +       unsigned int reg = INPUT_VAL;
>> +       unsigned int val, mask;
>> +
>> +       if (offset >= GPIO_PER_REG) {
>> +               offset -= GPIO_PER_REG;
>> +               reg += sizeof(u32);
>> +       }
>> +       mask = BIT(offset);
>
> This code is repeating. Break out a static (inline?) helper to do
> this.

done

>
>> +static int armada_37xx_gpiolib_register(struct platform_device *pdev,
>> +                                       struct armada_37xx_pinctrl *info)
>
> Nit: gpiochip_register or so is more to the point.
>
>> +       ret = gpiochip_add_pin_range(&info->gpio_chip, dev_name(dev), 0,
>> +                                    pinbase, info->data->nr_pins);
>> +       if (ret)
>> +               return ret;
>
> Why do you do this?
>
> Why not just put the ranges into the device tree? We already support
> this in the gpiolib core and it is helpful.
>
> See Documentation/devicetree/bindings/gpio/gpio.txt
> and other DTS files for gpio-ranges.

Following your review, I tried to use it but it didn't work for
me. When the second pin controller was probed then there was collision
for the gpio number. I tried several combination without any luck.

So for now I left it aside.

I can show you the errors message I get and the binding I used if you
are interested.


Thanks,

Gregory

-- 
Gregory Clement, Free Electrons
Kernel, drivers, real-time and embedded Linux
development, consulting, training and support.
http://free-electrons.com
--
To unsubscribe from this list: send the line "unsubscribe linux-gpio" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html