On Wed, 14 Aug 2019 11:21:44 +0200 Martin Kaiser <martin@xxxxxxxxx> wrote: > Replace iio_device_register with the devres variant and drop the > explicit function call to unregister the iio device. > > Signed-off-by: Martin Kaiser <martin@xxxxxxxxx> This changes the order of removal compared to probe as now we don't remove the userspace interface before storing to non volatile memory. Admittedly it's a pretty unlikely race to hit but I don't want to set a precedent of judging just how unlikely a race is. I'm also not keen to use dev_add_action_or_reset for this one as the action in remove is not balanced to one in probe - I want to see it clearly laid out in code, as it currently is. So a no on this sort of change in this particular driver. Thanks, Jonathan > --- > drivers/iio/potentiometer/max5481.c | 4 +--- > 1 file changed, 1 insertion(+), 3 deletions(-) > > diff --git a/drivers/iio/potentiometer/max5481.c b/drivers/iio/potentiometer/max5481.c > index 732375b6d131..40b7212e68dc 100644 > --- a/drivers/iio/potentiometer/max5481.c > +++ b/drivers/iio/potentiometer/max5481.c > @@ -162,7 +162,7 @@ static int max5481_probe(struct spi_device *spi) > if (ret < 0) > return ret; > > - return iio_device_register(indio_dev); > + return devm_iio_device_register(&spi->dev, indio_dev); > } > > static int max5481_remove(struct spi_device *spi) > @@ -170,8 +170,6 @@ static int max5481_remove(struct spi_device *spi) > struct iio_dev *indio_dev = dev_get_drvdata(&spi->dev); > struct max5481_data *data = iio_priv(indio_dev); > > - iio_device_unregister(indio_dev); > - > /* save wiper reg to NV reg */ > return max5481_write_cmd(data, MAX5481_COPY_AB_TO_NV, 0); > }
