RE: [PATCH] drivers: iio: dac: ad5592r: add gpio_chip names

[Miclaus, Antoniu]

--
Antoniu Miclăuş

> -----Original Message-----
> From: Jonathan Cameron <jic23@xxxxxxxxxx>
> Sent: Wednesday, November 23, 2022 10:49 PM
> To: Miclaus, Antoniu <Antoniu.Miclaus@xxxxxxxxxx>
> Cc: linux-iio@xxxxxxxxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx
> Subject: Re: [PATCH] drivers: iio: dac: ad5592r: add gpio_chip names
> 
> [External]
> 
> On Thu, 17 Nov 2022 11:01:30 +0200
> Antoniu Miclaus <antoniu.miclaus@xxxxxxxxxx> wrote:
> 
> > Add array of explicit gpio names for the `gpiochip` structure of
> > ad5592r, mainly for debug purposes.
> Can you give an example of when this is more useful than the offset?
> (which I'm assuming is also available when debugging?)
> 
> Jonathan
> 
Hello Jonathan,

I guess this patch is not mandatory. It is a gpio_chip structure member that was left unpopulated.
Also, I think it is nice to see explicitly the gpios under cat /sys/kernel/debug/gpio alongside the others.
This is the current output when trying to debug on a rpi (without the patch applied):

analog@analog:~ $ sudo cat /sys/kernel/debug/gpio
gpiochip0: GPIOs 0-57, parent: platform/fe200000.gpio, pinctrl-bcm2711:
 gpio-0   (ID_SDA              )
 gpio-1   (ID_SCL              )
 gpio-2   (SDA1                )
 gpio-3   (SCL1                )
 gpio-4   (GPIO_GCLK           )
 gpio-5   (GPIO5               )
 gpio-6   (GPIO6               )
....
gpiochip2: GPIOs 496-503, parent: spi/spi0.0, spi0.0, can sleep:

gpiochip1: GPIOs 504-511, parent: platform/soc:firmware:gpio, raspberrypi-exp-gpio, can sleep:
 gpio-504 (BT_ON               )
 gpio-505 (WL_ON               )
 gpio-506 (PWR_LED_OFF         |led1                ) out lo ACTIVE LOW
 gpio-507 (GLOBAL_RESET        )
 gpio-508 (VDD_SD_IO_SEL       |vdd-sd-io           ) out hi
 gpio-509 (CAM_GPIO            |power_ctrl          ) out lo
 gpio-510 (SD_PWR_ON           |sd_vcc_reg          ) out hi
 gpio-511 (SD_OC_N             )

The gpiochip2, corresponding to ad5592r has no explicit gpios, due to the fact that the
gpio names were assigned.

Regards,
> >
> > Since the gpios are configurable via the dts, generic names are
> > used.
> >
> > Signed-off-by: Antoniu Miclaus <antoniu.miclaus@xxxxxxxxxx>
> > ---
> >  drivers/iio/dac/ad5592r-base.c | 5 +++++
> >  1 file changed, 5 insertions(+)
> >
> > diff --git a/drivers/iio/dac/ad5592r-base.c b/drivers/iio/dac/ad5592r-base.c
> > index 7a9b5fc1e579..076bc9ecfb49 100644
> > --- a/drivers/iio/dac/ad5592r-base.c
> > +++ b/drivers/iio/dac/ad5592r-base.c
> > @@ -124,6 +124,10 @@ static int ad5592r_gpio_request(struct gpio_chip
> *chip, unsigned offset)
> >  	return 0;
> >  }
> >
> > +static const char * const ad5592r_gpio_names[] = {
> > +	"GPIO0", "GPIO1", "GPIO2", "GPIO3", "GPIO4", "GPIO5", "GPIO6",
> "GPIO7",
> > +};
> > +
> >  static int ad5592r_gpio_init(struct ad5592r_state *st)
> >  {
> >  	if (!st->gpio_map)
> > @@ -140,6 +144,7 @@ static int ad5592r_gpio_init(struct ad5592r_state
> *st)
> >  	st->gpiochip.set = ad5592r_gpio_set;
> >  	st->gpiochip.request = ad5592r_gpio_request;
> >  	st->gpiochip.owner = THIS_MODULE;
> > +	st->gpiochip.names = ad5592r_gpio_names;
> >
> >  	mutex_init(&st->gpio_lock);
> >