[PATCH] iio: common: cros_ec_sensors: use sensor_scale union member for CALIBSCALE

[PATCH] iio: common: cros_ec_sensors: use sensor_scale union member for CALIBSCALE

[Arnav Kapoor]

The IIO_CHAN_INFO_CALIBSCALE read and write paths send
MOTIONSENSE_CMD_SENSOR_SCALE, but set the flags through the
sensor_offset member of the request union, while the scale values in
the write path are filled through sensor_scale.

This only works because the sensor_offset and sensor_scale structs of
struct ec_params_motion_sense happen to place flags at the same offset.
Use the sensor_scale member that matches the command, so all fields of
the request are set through the same struct.

No functional change.

Assisted-by: LLM
Signed-off-by: Arnav Kapoor <kapoorarnav43@gmail.com>
---
Compile-tested only (W=1); I don't have ChromeOS EC hardware. Both
union members place flags at the same offset, so the generated host
command is unchanged.

 drivers/iio/common/cros_ec_sensors/cros_ec_sensors.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/drivers/iio/common/cros_ec_sensors/cros_ec_sensors.c b/drivers/iio/common/cros_ec_sensors/cros_ec_sensors.c
index 551b86236abc..2839714f5bc8 100644
--- a/drivers/iio/common/cros_ec_sensors/cros_ec_sensors.c
+++ b/drivers/iio/common/cros_ec_sensors/cros_ec_sensors.c
@@ -70,7 +70,7 @@ static int cros_ec_sensors_read(struct iio_dev *indio_dev,
 		break;
 	case IIO_CHAN_INFO_CALIBSCALE:
 		st->core.param.cmd = MOTIONSENSE_CMD_SENSOR_SCALE;
-		st->core.param.sensor_offset.flags = 0;
+		st->core.param.sensor_scale.flags = 0;
 
 		ret = cros_ec_motion_send_host_cmd(&st->core, 0);
 		if (ret == -EPROTO || ret == -EOPNOTSUPP) {
@@ -178,8 +178,7 @@ static int cros_ec_sensors_write(struct iio_dev *indio_dev,
 		/* Send to EC for each axis, even if not complete */
 
 		st->core.param.cmd = MOTIONSENSE_CMD_SENSOR_SCALE;
-		st->core.param.sensor_offset.flags =
-			MOTION_SENSE_SET_OFFSET;
+		st->core.param.sensor_scale.flags = MOTION_SENSE_SET_OFFSET;
 		for (i = CROS_EC_SENSOR_X; i < CROS_EC_SENSOR_MAX_AXIS; i++)
 			st->core.param.sensor_scale.scale[i] =
 				st->core.calib[i].scale;
-- 
2.53.0

Re: [PATCH] iio: common: cros_ec_sensors: use sensor_scale union member for CALIBSCALE

[Joshua Crofts]

On Sun, 11 Oct 2026 10:24:08 +0530
Arnav Kapoor <kapoorarnav43@gmail.com> wrote:

> The IIO_CHAN_INFO_CALIBSCALE read and write paths send
> MOTIONSENSE_CMD_SENSOR_SCALE, but set the flags through the
> sensor_offset member of the request union, while the scale values in
> the write path are filled through sensor_scale.
> 
> This only works because the sensor_offset and sensor_scale structs of
> struct ec_params_motion_sense happen to place flags at the same offset.
> Use the sensor_scale member that matches the command, so all fields of
> the request are set through the same struct.
> 
> No functional change.
> 
> Assisted-by: LLM
> Signed-off-by: Arnav Kapoor <kapoorarnav43@gmail.com>
> ---

Please add a Fixes tag. Otherwise looks good to me!

Reviewed-by: Joshua Crofts <joshua.crofts1@gmail.com>

-- 
Kind regards,
Joshua Crofts

Re: [PATCH] iio: common: cros_ec_sensors: use sensor_scale union member for CALIBSCALE

[Arnav Kapoor]

On Sun, 11 Oct 2026 19:16:41 +0200, Joshua Crofts wrote:
> Please add a Fixes tag. Otherwise looks good to me!
>
> Reviewed-by: Joshua Crofts <joshua.crofts1@gmail.com>

Thanks for the review. The mismatch was introduced when CALIBSCALE
support was added:

Fixes: ed1f2e85da79 ("iio: cros_ec: Add calibscale for 3d MEMS")

Jonathan, could you pick this up when applying, or would you prefer a v2?

Thanks,
Arnav

Re: [PATCH] iio: common: cros_ec_sensors: use sensor_scale union member for CALIBSCALE

[Andy Shevchenko]

On Tue, Oct 13, 2026 at 02:47:30AM +0530, Arnav Kapoor wrote:
> On Sun, 11 Oct 2026 19:16:41 +0200, Joshua Crofts wrote:
> > Please add a Fixes tag. Otherwise looks good to me!
> >
> > Reviewed-by: Joshua Crofts <joshua.crofts1@gmail.com>
> 
> Thanks for the review. The mismatch was introduced when CALIBSCALE
> support was added:
> 
> Fixes: ed1f2e85da79 ("iio: cros_ec: Add calibscale for 3d MEMS")
> 
> Jonathan, could you pick this up when applying, or would you prefer a v2?

No need to resend, `b4` will scan the whole email thread for the tags.

-- 
With Best Regards,
Andy Shevchenko

Re: [PATCH] iio: common: cros_ec_sensors: use sensor_scale union member for CALIBSCALE

[Jonathan Cameron]

On Tue, 13 Oct 2026 22:55:27 +0300
Andy Shevchenko <andriy.shevchenko@intel.com> wrote:

> On Tue, Oct 13, 2026 at 02:47:30AM +0530, Arnav Kapoor wrote:
> > On Sun, 11 Oct 2026 19:16:41 +0200, Joshua Crofts wrote:  
> > > Please add a Fixes tag. Otherwise looks good to me!
> > >
> > > Reviewed-by: Joshua Crofts <joshua.crofts1@gmail.com>  
> > 
> > Thanks for the review. The mismatch was introduced when CALIBSCALE
> > support was added:
> > 
> > Fixes: ed1f2e85da79 ("iio: cros_ec: Add calibscale for 3d MEMS")
> > 
> > Jonathan, could you pick this up when applying, or would you prefer a v2?  
> 
> No need to resend, `b4` will scan the whole email thread for the tags.
If it could put them in the right order for 'late' fixes tags it would
be perfect (easy fix though so don't resend!)
> 

This driver tends to be well maintained so whilst I'm fine
with the change I'll wait a bit for a tag from one of the maintainers.

Jonathan

Re: [PATCH] iio: common: cros_ec_sensors: use sensor_scale union member for CALIBSCALE

[Tzung-Bi Shih]

On Sun, Oct 18, 2026 at 12:02:06AM +0100, Jonathan Cameron wrote:
> On Tue, 13 Oct 2026 22:55:27 +0300
> Andy Shevchenko <andriy.shevchenko@intel.com> wrote:
> 
> > On Tue, Oct 13, 2026 at 02:47:30AM +0530, Arnav Kapoor wrote:
> > > On Sun, 11 Oct 2026 19:16:41 +0200, Joshua Crofts wrote:  
> > > > Please add a Fixes tag. Otherwise looks good to me!
> > > >
> > > > Reviewed-by: Joshua Crofts <joshua.crofts1@gmail.com>  
> > > 
> > > Thanks for the review. The mismatch was introduced when CALIBSCALE
> > > support was added:
> > > 
> > > Fixes: ed1f2e85da79 ("iio: cros_ec: Add calibscale for 3d MEMS")
> > > 
> > > Jonathan, could you pick this up when applying, or would you prefer a v2?  
> > 
> > No need to resend, `b4` will scan the whole email thread for the tags.
> If it could put them in the right order for 'late' fixes tags it would
> be perfect (easy fix though so don't resend!)
> > 
> 
> This driver tends to be well maintained so whilst I'm fine
> with the change I'll wait a bit for a tag from one of the maintainers.

Sorry for the delay, I somehow missed this in my inbox.

Reviewed-by: Tzung-Bi Shih <tzungbi@kernel.org>