[PATCH v2 0/2] iio: accel: fxls8962af: Fix temperature readings

[PATCH v2 0/2] iio: accel: fxls8962af: Fix temperature readings

[Sean Nyekjaer]

Add the correct scale to get temperature in mili degree Celcius.
Add sign component to temperature scan element.

Signed-off-by: Sean Nyekjaer <sean@geanix.com>
---
Changes in v2:
- Correct offset is applied before scaling component 
- Added sign component to temperature scan element
- Link to v1: https://lore.kernel.org/r/20250501-fxls-v1-1-f54061a07099@geanix.com

---
Sean Nyekjaer (2):
      iio: accel: fxls8962af: Fix temperature calculation
      iio: accel: fxls8962af: Fix sign temperature scan element

 drivers/iio/accel/fxls8962af-core.c | 16 ++++++++++++++--
 1 file changed, 14 insertions(+), 2 deletions(-)
---
base-commit: 609bc31eca06c7408e6860d8b46311ebe45c1fef
change-id: 20250501-fxls-307ef3d6d065

Best regards,
-- 
Sean Nyekjaer <sean@geanix.com>

[PATCH v2 1/2] iio: accel: fxls8962af: Fix temperature calculation

[Sean Nyekjaer]

According to spec temperature should be returned in milli degrees Celsius.
Add in_temp_scale to calculate from Celsius to milli Celsius.

Fixes: a3e0b51884ee ("iio: accel: add support for FXLS8962AF/FXLS8964AF accelerometers")
Cc: stable@vger.kernel.org
Signed-off-by: Sean Nyekjaer <sean@geanix.com>
---
 drivers/iio/accel/fxls8962af-core.c | 15 +++++++++++++--
 1 file changed, 13 insertions(+), 2 deletions(-)

diff --git a/drivers/iio/accel/fxls8962af-core.c b/drivers/iio/accel/fxls8962af-core.c
index bf1d3923a181798a1c884ee08b62d86ab5aed26f..f515222e008493687921879a0b0ef44fd4ae5d10 100644
--- a/drivers/iio/accel/fxls8962af-core.c
+++ b/drivers/iio/accel/fxls8962af-core.c
@@ -134,6 +134,8 @@
 
 /* Raw temp channel offset */
 #define FXLS8962AF_TEMP_CENTER_VAL		25
+/* Raw temp channel scale */
+#define FXLS8962AF_TEMP_SCALE			1000
 
 #define FXLS8962AF_AUTO_SUSPEND_DELAY_MS	2000
 
@@ -439,8 +441,16 @@ static int fxls8962af_read_raw(struct iio_dev *indio_dev,
 		*val = FXLS8962AF_TEMP_CENTER_VAL;
 		return IIO_VAL_INT;
 	case IIO_CHAN_INFO_SCALE:
-		*val = 0;
-		return fxls8962af_read_full_scale(data, val2);
+		switch (chan->type) {
+		case IIO_TEMP:
+			*val = FXLS8962AF_TEMP_SCALE;
+			return IIO_VAL_INT;
+		case IIO_ACCEL:
+			*val = 0;
+			return fxls8962af_read_full_scale(data, val2);
+		default:
+			return -EINVAL;
+		}
 	case IIO_CHAN_INFO_SAMP_FREQ:
 		return fxls8962af_read_samp_freq(data, val, val2);
 	default:
@@ -736,6 +746,7 @@ static const struct iio_event_spec fxls8962af_event[] = {
 	.type = IIO_TEMP, \
 	.address = FXLS8962AF_TEMP_OUT, \
 	.info_mask_separate = BIT(IIO_CHAN_INFO_RAW) | \
+			      BIT(IIO_CHAN_INFO_SCALE) | \
 			      BIT(IIO_CHAN_INFO_OFFSET),\
 	.scan_index = -1, \
 	.scan_type = { \

-- 
2.47.1

[PATCH v2 2/2] iio: accel: fxls8962af: Fix sign temperature scan element

[Sean Nyekjaer]

TEMP_OUT register contains the 8-bit, 2's complement temperature value.
Let's mark the temperature scan element signed.

Fixes: a3e0b51884ee ("iio: accel: add support for FXLS8962AF/FXLS8964AF accelerometers")
Suggested-by: Marcelo Schmitt <marcelo.schmitt1@gmail.com>
Cc: stable@vger.kernel.org
Signed-off-by: Sean Nyekjaer <sean@geanix.com>
---
 drivers/iio/accel/fxls8962af-core.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/iio/accel/fxls8962af-core.c b/drivers/iio/accel/fxls8962af-core.c
index f515222e008493687921879a0b0ef44fd4ae5d10..e1b752e202b877db606a55a978d63ef52894c60d 100644
--- a/drivers/iio/accel/fxls8962af-core.c
+++ b/drivers/iio/accel/fxls8962af-core.c
@@ -750,6 +750,7 @@ static const struct iio_event_spec fxls8962af_event[] = {
 			      BIT(IIO_CHAN_INFO_OFFSET),\
 	.scan_index = -1, \
 	.scan_type = { \
+		.sign = 's', \
 		.realbits = 8, \
 		.storagebits = 8, \
 	}, \

-- 
2.47.1

Re: [PATCH v2 1/2] iio: accel: fxls8962af: Fix temperature calculation

[Andy Shevchenko]

On Fri, May 2, 2025 at 9:15 AM Sean Nyekjaer <sean@geanix.com> wrote:
>
> According to spec temperature should be returned in milli degrees Celsius.
> Add in_temp_scale to calculate from Celsius to milli Celsius.

...

> +/* Raw temp channel scale */
> +#define FXLS8962AF_TEMP_SCALE                  1000

Wouldn't constants from units.h be helpful here?

>  #define FXLS8962AF_AUTO_SUSPEND_DELAY_MS       2000

(2 * MSEC_PER_SEC)

This gives immediately that we want 2 seconds of delay.

-- 
With Best Regards,
Andy Shevchenko

Re: [PATCH v2 1/2] iio: accel: fxls8962af: Fix temperature calculation

[Marcelo Schmitt]

On 05/02, Sean Nyekjaer wrote:
> According to spec temperature should be returned in milli degrees Celsius.
> Add in_temp_scale to calculate from Celsius to milli Celsius.
> 
> Fixes: a3e0b51884ee ("iio: accel: add support for FXLS8962AF/FXLS8964AF accelerometers")
> Cc: stable@vger.kernel.org
> Signed-off-by: Sean Nyekjaer <sean@geanix.com>

Reviewed-by: Marcelo Schmitt <marcelo.schmitt1@gmail.com>

Re: [PATCH v2 2/2] iio: accel: fxls8962af: Fix sign temperature scan element

[Marcelo Schmitt]

On 05/02, Sean Nyekjaer wrote:
> TEMP_OUT register contains the 8-bit, 2's complement temperature value.
> Let's mark the temperature scan element signed.
> 
> Fixes: a3e0b51884ee ("iio: accel: add support for FXLS8962AF/FXLS8964AF accelerometers")
> Suggested-by: Marcelo Schmitt <marcelo.schmitt1@gmail.com>
> Cc: stable@vger.kernel.org
> Signed-off-by: Sean Nyekjaer <sean@geanix.com>

Sort of nitpinking but I think the commit description could be more assertive.
The main idea is that we want to make the scan element signed because the
data read from the TEMP_OUT register is in two's complement format and not
having the scan element marked as a signed number may cause it to be mishandled
and miss displayed. Nevertheless, I do think the patch is good so

Reviewed-by: Marcelo Schmitt <marcelo.schmitt1@gmail.com>

Re: [PATCH v2 1/2] iio: accel: fxls8962af: Fix temperature calculation

[Jonathan Cameron]

On Fri, 2 May 2025 17:19:44 +0300
Andy Shevchenko <andy.shevchenko@gmail.com> wrote:

> On Fri, May 2, 2025 at 9:15 AM Sean Nyekjaer <sean@geanix.com> wrote:
> >
> > According to spec temperature should be returned in milli degrees Celsius.
> > Add in_temp_scale to calculate from Celsius to milli Celsius.  
> 
> ...
> 
> > +/* Raw temp channel scale */
> > +#define FXLS8962AF_TEMP_SCALE                  1000  
> 
> Wouldn't constants from units.h be helpful here?

Whilst you are just continuing local style, I'm not sure
these defines really bring much at all given the TEMP_SCALE
one for instance is only used in one place which is
explicitly getting the temperature scale.  It's not a magic
number that needs a define. It's a number that means it's own
value :)

Using MILLI there would be a nice bit of self documentation
though.
> 
> >  #define FXLS8962AF_AUTO_SUSPEND_DELAY_MS       2000  
> 
> (2 * MSEC_PER_SEC)
> 
> This gives immediately that we want 2 seconds of delay.
> 
True but not part of this patch so that would be a nice
little follow up.  Possibly dropping this define in favour
of using that inline.

Jonathan

Re: [PATCH v2 2/2] iio: accel: fxls8962af: Fix sign temperature scan element

[Jonathan Cameron]

On Sat, 3 May 2025 10:57:09 -0300
Marcelo Schmitt <marcelo.schmitt1@gmail.com> wrote:

> On 05/02, Sean Nyekjaer wrote:
> > TEMP_OUT register contains the 8-bit, 2's complement temperature value.
> > Let's mark the temperature scan element signed.
> > 
> > Fixes: a3e0b51884ee ("iio: accel: add support for FXLS8962AF/FXLS8964AF accelerometers")
> > Suggested-by: Marcelo Schmitt <marcelo.schmitt1@gmail.com>
> > Cc: stable@vger.kernel.org
> > Signed-off-by: Sean Nyekjaer <sean@geanix.com>  
> 
> Sort of nitpinking but I think the commit description could be more assertive.

Agreed.  I might have just tweaked it to
"Mark the temperature element signed."  + some of what Marcelo has below.
but given Andy's request on patch 1 means you are probably doing a v3, please
tidy this up as well.

Thanks and good catches on both of them!

Jonathan

> The main idea is that we want to make the scan element signed because the
> data read from the TEMP_OUT register is in two's complement format and not
> having the scan element marked as a signed number may cause it to be mishandled
> and miss displayed. Nevertheless, I do think the patch is good so
> 
> Reviewed-by: Marcelo Schmitt <marcelo.schmitt1@gmail.com>
>