Re: [PATCH v2] iio: adc: ad7887: invert/rework external ref logic

From: Jonathan Cameron
Date: Sat Oct 10 2020 - 19:07:11 EST


On Fri, 2 Oct 2020 11:27:23 +0300
Alexandru Ardelean <alexandru.ardelean@xxxxxxxxxx> wrote:

> This change inverts/reworks the logic to use an external reference via a
> provided regulator.
>
> Now the driver tries to obtain a regulator. If one is found, then it is
> used. The rest of the driver logic already checks if there is a non-NULL
> reference to a regulator, so it should be fine.
>
> Signed-off-by: Alexandru Ardelean <alexandru.ardelean@xxxxxxxxxx>
Applied to the togreg branch of iio.git and pushed out as testing for
all the normal reasons.

Interestingly it seems something odd happened to my email and I was missing
random threads/part threads. I've pulled them off lore.kernel.org but if
I seem to have lost anything let me know.

If anyone has a bounce message from me, please send it over as I'm curious
as to what went wrong!

Thanks,

Jonathan

> ---
>
> Changelog v1 -> v2:
> * remove omitted '!pdata->use_onchip_ref' check; the field was removed from
> the platform data, but was still used
>
> drivers/iio/adc/ad7887.c | 12 ++++++++----
> include/linux/platform_data/ad7887.h | 4 ----
> 2 files changed, 8 insertions(+), 8 deletions(-)
>
> diff --git a/drivers/iio/adc/ad7887.c b/drivers/iio/adc/ad7887.c
> index 037bcb47693c..99a480ad3985 100644
> --- a/drivers/iio/adc/ad7887.c
> +++ b/drivers/iio/adc/ad7887.c
> @@ -246,11 +246,15 @@ static int ad7887_probe(struct spi_device *spi)
>
> st = iio_priv(indio_dev);
>
> - if (!pdata || !pdata->use_onchip_ref) {
> - st->reg = devm_regulator_get(&spi->dev, "vref");
> - if (IS_ERR(st->reg))
> + st->reg = devm_regulator_get_optional(&spi->dev, "vref");
> + if (IS_ERR(st->reg)) {
> + if (PTR_ERR(st->reg) != -ENODEV)
> return PTR_ERR(st->reg);
>
> + st->reg = NULL;
> + }
> +
> + if (st->reg) {
> ret = regulator_enable(st->reg);
> if (ret)
> return ret;
> @@ -269,7 +273,7 @@ static int ad7887_probe(struct spi_device *spi)
> /* Setup default message */
>
> mode = AD7887_PM_MODE4;
> - if (!pdata || !pdata->use_onchip_ref)
> + if (!st->reg)
> mode |= AD7887_REF_DIS;
> if (pdata && pdata->en_dual)
> mode |= AD7887_DUAL;
> diff --git a/include/linux/platform_data/ad7887.h b/include/linux/platform_data/ad7887.h
> index 732af46b2d16..9b4dca6ae70b 100644
> --- a/include/linux/platform_data/ad7887.h
> +++ b/include/linux/platform_data/ad7887.h
> @@ -13,13 +13,9 @@
> * second input channel, and Vref is internally connected to Vdd. If set to
> * false the device is used in single channel mode and AIN1/Vref is used as
> * VREF input.
> - * @use_onchip_ref: Whether to use the onchip reference. If set to true the
> - * internal 2.5V reference is used. If set to false a external reference is
> - * used.
> */
> struct ad7887_platform_data {
> bool en_dual;
> - bool use_onchip_ref;
> };
>
> #endif /* IIO_ADC_AD7887_H_ */