On Wed, 8 May 2024 12:54:35 -0300
Fernando Yang <hagisf@xxxxxx> wrote:
> Switching to the _scoped() version can make the error
> handling more natural instead of delayed until direct
> mode was released.
>
> Co-developed-by: Eduardo Figueredo <eduardofp@xxxxxx>
> Signed-off-by: Eduardo Figueredo <eduardofp@xxxxxx>
> Signed-off-by: Fernando Yang <hagisf@xxxxxx>
> ---
> drivers/iio/adc/ad7266.c | 15 +++++++++------
> 1 file changed, 9 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/iio/adc/ad7266.c b/drivers/iio/adc/ad7266.c
> index 8b03d4469..3fc34a3a8 100644
> --- a/drivers/iio/adc/ad7266.c
> +++ b/drivers/iio/adc/ad7266.c
> @@ -63,12 +63,14 @@ static int ad7266_powerdown(struct ad7266_state *st)
> static int ad7266_preenable(struct iio_dev *indio_dev)
> {
> struct ad7266_state *st = iio_priv(indio_dev);
> +
> return ad7266_wakeup(st);
> }
>
> static int ad7266_postdisable(struct iio_dev *indio_dev)
> {
> struct ad7266_state *st = iio_priv(indio_dev);
> +
> return ad7266_powerdown(st);
> }
>
> @@ -151,15 +153,16 @@ static int ad7266_read_raw(struct iio_dev *indio_dev,
>
> switch (m) {
> case IIO_CHAN_INFO_RAW:
> - iio_device_claim_direct_scoped(return -EBUSY, indio_dev)
This isn't the existing code. I'm guessing you have another commit ahead
of this one on your branch?
> + iio_device_claim_direct_scoped(return -EBUSY, indio_dev) {
> ret = ad7266_read_single(st, val, chan->address);
Not checking ret here is a bug in the existing code. Please fix that as
a precursor patch before making the _scoped change.
The check will have to be immediately after iio_device_release_direct_mode()
in the existing code.
>
> - *val = (*val >> 2) & 0xfff;
> - if (chan->scan_type.sign == 's')
> - *val = sign_extend32(*val,
> - chan->scan_type.realbits - 1);
> + *val = (*val >> 2) & 0xfff;
> + if (chan->scan_type.sign == 's')
> + *val = sign_extend32(*val,
> + chan->scan_type.realbits - 1);
>
> - return IIO_VAL_INT;
> + return IIO_VAL_INT;
> + }
> unreachable();
> case IIO_CHAN_INFO_SCALE:
> scale_mv = st->vref_mv;
