On Sat, 7 Dec 2024 20:34:30 +0200
Andy Shevchenko <andriy.shevchenko@xxxxxxxxx> wrote:
> On Sun, Dec 01, 2024 at 01:03:56PM +0000, Jonathan Cameron wrote:
> > On Fri, 22 Nov 2024 13:43:08 -0300
> > Fabio Estevam <festevam@xxxxxxxxx> wrote:
> > > Using gpiod_set_value() to control the reset GPIO causes some verbose
> > > warnings during boot when the reset GPIO is controlled by an I2C IO
> > > expander.
> > >
> > > As the caller can sleep, use the gpiod_set_value_cansleep() variant to
> > > fix the issue.
> > >
> > > Tested on a custom i.MX93 board with a ADS124S08 ADC.
>
> > Hmm. Could be considered a feature, but fair enough as the change is
> > trivial and you are setting it on a real board.
> >
> > Applied to the fixes-togreg branch of iio.git.
>
> ...
>
> > > @@ -183,9 +183,9 @@ static int ads124s_reset(struct iio_dev *indio_dev)
> > > struct ads124s_private *priv = iio_priv(indio_dev);
> > >
> > > if (priv->reset_gpio) {
> > > - gpiod_set_value(priv->reset_gpio, 0);
> > > + gpiod_set_value_cansleep(priv->reset_gpio, 0);
> > > udelay(200);
>
> Obviously if you allow that change, you should switch to fsleep() here.
True, that would be a nice follow up tidying up.
Jonathan
>
> > > - gpiod_set_value(priv->reset_gpio, 1);
> > > + gpiod_set_value_cansleep(priv->reset_gpio, 1);
> > > } else {
> > > return ads124s_write_cmd(indio_dev, ADS124S08_CMD_RESET);
> > > }
>
