mbox series

[v4,00/15] thermal/drivers/tsens: specify nvmem cells in DT rather than parsing them manually

Message ID 20221221020520.1326964-1-dmitry.baryshkov@linaro.org
Headers show
Series thermal/drivers/tsens: specify nvmem cells in DT rather than parsing them manually | expand

Message

Dmitry Baryshkov Dec. 21, 2022, 2:05 a.m. UTC
Historically the tsens driver fetches the calibration data as a blob and
then parses the blob on its own. This results in semi-duplicated code
spreading over the platform-specific functions.

This patch series changes tsens calibration code to per-value nvmem
cells rather than parsing the blob in the driver. For backwards
compatibility the old code is left in place for msm8916, msm8974 and
qcs404, the platforms which have in-tree DT files. For all other
affected platforms the old parsing code has been dropped as a part of
this series.

The code was tested on msm8916 and qcs404 only.

Note: the DTs changes depend on driver changes. Tsens driver will not
work if DT patches are merged, but the driver bits are not. As the
thermal sense is critical for device safety, I'd suggest merging binding
and driver during one merge window and then merging DT changes in the
next merge window.

Note2:
For v3 I have reworked the legacy bindings support into the small and
manageable data-driven piece of code.

I still have included patches to drop legacy support for 8939
(msm8939.dtsi is on the list, I have provided Bryan with the patch to
update it to new bindings) and 8976 (msm8976.dtsi and msm8956.dtsi,
which use single-blob bindings, have been accepted for 6.2, dropping old
bindings support depends on mutual consensuns of platform and thermal
code maintainers). The patches are the last ones in the thermal part of
the series, thus if they are declined, the reset of the series still can
be applied without any problems.

Changes since v3:
- Added a patch to fix the tsens compatible string on msm8956 SoC,
- Fixed num-sensors and slope coefficients for the msm8939 SoC,
- Rewrote code supporting old bindings into the simple data-driven
  parser common to all legacy platforms which made dropping support for
  old bindings less demanding.

Changes since v2:
- Made init_8956 static, as pointed out by the testing robot and by
  AngeloGioacchino Del Regno.

Changes since the RFC:
- Sorted out the msm8976/msm8956, custom slopes are used only for msm8956,
- Implemented proper support for msm8974/apq8084,
- Added tsens_calibrate_common() and ops_v0_1 which can be used in
  common cases,
- Removed superfluous identity hw_ids
- Fixed calibration calculation in tsens_calibrate_nvmem() for
  ONE_PT_CALIB case

Dmitry Baryshkov (20):
  dt-bindings: thermal: tsens: add msm8956 compat
  dt-bindings: thermal: tsens: support per-sensor calibration cells
  dt-bindings: thermal: tsens: add per-sensor cells for msm8974
  thermal/drivers/tsens: Drop unnecessary hw_ids
  thermal/drivers/tsens: Drop msm8976-specific defines
  thermal/drivers/tsens: Sort out msm8976 vs msm8956 data
  thermal/drivers/tsens: limit num_sensors to 9
  thermal/drivers/tsens: fix slope values for msm8939
  thermal/drivers/tsens: Support using nvmem cells for calibration data
  thermal/drivers/tsens: Support using nvmem cells for msm8974
    calibration
  thermal/drivers/tsens: Rework legacy calibration data parsers
  thermal/drivers/tsens: Drop single-cell code for mdm9607
  thermal/drivers/tsens: Drop single-cell code for msm8939
  thermal/drivers/tsens: Drop single-cell code for msm8976/msm8956
  arm64: dts: qcom: msm8956: use SoC-specific compat for tsens
  arm64: dts: qcom: msm8916: specify per-sensor calibration cells
  arm64: dts: qcom: msm8976: specify per-sensor calibration cells
  arm64: dts: qcom: qcs404: specify per-sensor calibration cells
  ARM: dts: qcom-msm8974: specify per-sensor calibration cells
  ARM: dts: qcom-apq8084: specify per-sensor calibration cells

 .../bindings/thermal/qcom-tsens.yaml          | 123 +++-
 arch/arm/boot/dts/qcom-apq8084.dtsi           | 313 ++++++++-
 arch/arm/boot/dts/qcom-msm8974.dtsi           | 313 ++++++++-
 arch/arm64/boot/dts/qcom/msm8916.dtsi         |  81 ++-
 arch/arm64/boot/dts/qcom/msm8956.dtsi         |   4 +
 arch/arm64/boot/dts/qcom/msm8976.dtsi         | 153 +++-
 arch/arm64/boot/dts/qcom/qcs404.dtsi          | 145 +++-
 drivers/thermal/qcom/tsens-v0_1.c             | 653 +++++-------------
 drivers/thermal/qcom/tsens-v1.c               | 340 +++------
 drivers/thermal/qcom/tsens.c                  | 166 +++++
 drivers/thermal/qcom/tsens.h                  |  46 +-
 11 files changed, 1572 insertions(+), 765 deletions(-)

Comments

Konrad Dybcio Dec. 21, 2022, 8:14 p.m. UTC | #1
On 21.12.2022 03:05, Dmitry Baryshkov wrote:
> According to the vendor kernels (msm-3.10, 3.14 and 3.18), msm8939
> uses non-standard slope values for calibrating the sensors. Fill them
> accordingly.
> 
> Fixes: 332bc8ebab2c ("thermal: qcom: tsens-v0_1: Add support for MSM8939")
> Cc: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
> Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
> ---
Reviewed-by: Konrad Dybcio <konrad.dybcio@linaro.org>

Konrad
>  drivers/thermal/qcom/tsens-v0_1.c | 16 +++++++++++++++-
>  1 file changed, 15 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/thermal/qcom/tsens-v0_1.c b/drivers/thermal/qcom/tsens-v0_1.c
> index 57ac23f9d9b7..be173220a939 100644
> --- a/drivers/thermal/qcom/tsens-v0_1.c
> +++ b/drivers/thermal/qcom/tsens-v0_1.c
> @@ -534,6 +534,20 @@ static int calibrate_9607(struct tsens_priv *priv)
>  	return 0;
>  }
>  
> +static int __init init_8939(struct tsens_priv *priv) {
> +	priv->sensor[0].slope = 2911;
> +	priv->sensor[1].slope = 2789;
> +	priv->sensor[2].slope = 2906;
> +	priv->sensor[3].slope = 2763;
> +	priv->sensor[4].slope = 2922;
> +	priv->sensor[5].slope = 2867;
> +	priv->sensor[6].slope = 2833;
> +	priv->sensor[7].slope = 2838;
> +	priv->sensor[8].slope = 2840;
> +
> +	return init_common(priv);
> +}
> +
>  /* v0.1: 8916, 8939, 8974, 9607 */
>  
>  static struct tsens_features tsens_v0_1_feat = {
> @@ -599,7 +613,7 @@ struct tsens_plat_data data_8916 = {
>  };
>  
>  static const struct tsens_ops ops_8939 = {
> -	.init		= init_common,
> +	.init		= init_8939,
>  	.calibrate	= calibrate_8939,
>  	.get_temp	= get_temp_common,
>  };