Re: [PATCH v2 11/16] iio: adc: max1027: Prevent single channel accesses during buffer reads

From: Jonathan Cameron
Date: Sun Sep 05 2021 - 11:34:43 EST


On Thu, 2 Sep 2021 23:14:32 +0200
Miquel Raynal <miquel.raynal@xxxxxxxxxxx> wrote:

> When hardware buffers are enabled (the cnvst pin being the trigger), one
> should not mess with the device state by requesting a single channel
> read.
>
> There is already a iio_buffer_enabled() check in *_read_single_value()
> to merely prevent this situation but the check is inconsistent since
> buffers can be enabled after the if clause anyway. Instead, use the core
> mutex by calling iio_device_claim/release_direct_mode().
>
> Signed-off-by: Miquel Raynal <miquel.raynal@xxxxxxxxxxx>
> ---
> drivers/iio/adc/max1027.c | 9 +++++----
> 1 file changed, 5 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/iio/adc/max1027.c b/drivers/iio/adc/max1027.c
> index f4cb5c75604b..57f62ea2d7aa 100644
> --- a/drivers/iio/adc/max1027.c
> +++ b/drivers/iio/adc/max1027.c
> @@ -296,10 +296,9 @@ static int max1027_read_single_value(struct iio_dev *indio_dev,
> int ret;
> struct max1027_state *st = iio_priv(indio_dev);
>
> - if (iio_buffer_enabled(indio_dev)) {
> - dev_warn(&indio_dev->dev, "trigger mode already enabled");
> - return -EBUSY;
> - }
> + ret = iio_device_claim_direct_mode(indio_dev);
> + if (ret)
> + return ret;

>
> /* Configure conversion register with the requested chan */
> st->reg = MAX1027_CONV_REG | MAX1027_CHAN(chan->channel) |
> @@ -325,6 +324,8 @@ static int max1027_read_single_value(struct iio_dev *indio_dev,
> if (ret < 0)
> return ret;

Must also be released in error paths. Treat it like a lock (there is one
internally) so to avoid permanent deadlock we must release it even if things
are going wrong

>
> + iio_device_release_direct_mode(indio_dev);
> +
> *val = be16_to_cpu(st->buffer[0]);
>
> return IIO_VAL_INT;