Re: [PATCH V11 8/9] iio: imu: inv_icm42607: Add Gyroscope to icm42607

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [Medium] Missing `_available` info masks in `INV_ICM42607_GYRO_CHAN` makes `read_avail` dead code and hides capabilities from userspace.
- [Medium] Missing `iio_device_claim_direct()` protection when writing sampling frequency.
--

commit 85b10db1767bd36f30119bab0e0a4f4f63966fc3
Author: Chris Morgan <macromorgan@hotmail.com>

iio: imu: inv_icm42607: Add Gyroscope to icm42607

Add gyroscope functions to the icm42607 driver.

> diff --git a/drivers/iio/imu/inv_icm42607/inv_icm42607_gyro.c b/drivers/iio/imu/inv_icm42607/inv_icm42607_gyro.c
> new file mode 100644
> index 0000000000000..8cabebe2e96b1
> --- /dev/null
> +++ b/drivers/iio/imu/inv_icm42607/inv_icm42607_gyro.c
> @@ -0,0 +1,369 @@

[ ... ]

> +#define INV_ICM42607_GYRO_CHAN(_modifier, _index, _ext_info)	\
> +{								\
> +	.type = IIO_ANGL_VEL,					\
> +	.modified = 1,						\
> +	.channel2 = _modifier,					\
> +	.info_mask_separate =					\
> +		BIT(IIO_CHAN_INFO_RAW),				\
> +	.info_mask_shared_by_type =				\
> +		BIT(IIO_CHAN_INFO_SCALE),			\
> +	.info_mask_shared_by_all =				\
> +		BIT(IIO_CHAN_INFO_SAMP_FREQ),			\

[Severity: Medium]
Will userspace be able to discover the available scales and frequencies?

Since .info_mask_shared_by_type_available and 
.info_mask_shared_by_all_available are not included in this macro, it
appears the IIO core will never query inv_icm42607_gyro_read_avail(),
leaving that function as dead code.

> +	.scan_index = _index,					\
> +	.scan_type = {						\

[ ... ]

> +static int inv_icm42607_gyro_write_raw(struct iio_dev *indio_dev,
> +				       struct iio_chan_spec const *chan,
> +				       int val, int val2, long mask)
> +{
> +	int ret;
> +
> +	if (chan->type != IIO_ANGL_VEL)
> +		return -EINVAL;
> +
> +	switch (mask) {
> +	case IIO_CHAN_INFO_SCALE:
> +		if (!iio_device_claim_direct(indio_dev))
> +			return -EBUSY;
> +		ret = inv_icm42607_gyro_write_scale(indio_dev, val, val2);
> +		iio_device_release_direct(indio_dev);
> +		return ret;
> +	case IIO_CHAN_INFO_SAMP_FREQ:
> +		return inv_icm42607_gyro_write_odr(indio_dev, val, val2);

[Severity: Medium]
Is it safe to write the sampling frequency without claiming direct mode here?

It looks like IIO_CHAN_INFO_SAMP_FREQ bypasses iio_device_claim_direct(),
meaning it could potentially be modified concurrently with active IIO data
buffering.

> +	default:
> +		return -EINVAL;
> +	}
> +}

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260610175455.19006-1-macroalpha82@gmail.com?part=8