diff mbox series

[v2] thermal: imx: implement runtime PM support

Message ID 20211019130809.21281-1-o.rempel@pengutronix.de (mailing list archive)
State New, archived
Headers show
Series [v2] thermal: imx: implement runtime PM support | expand

Commit Message

Oleksij Rempel Oct. 19, 2021, 1:08 p.m. UTC
Starting with commit d92ed2c9d3ff ("thermal: imx: Use driver's local
data to decide whether to run a measurement") this driver stared using
irq_enabled flag to make decision to power on/off the thermal core. This
triggered a regression, where after reaching critical temperature, alarm
IRQ handler set irq_enabled to false,  disabled thermal core and was not
able read temperature and disable cooling sequence.

In case the cooling device is "CPU/GPU freq", the system will run with
reduce performance until next reboot.

To solve this issue, we need to move all parts implementing hand made
runtime power management and let it handle actual runtime PM framework.

Fixes: d92ed2c9d3ff ("thermal: imx: Use driver's local data to decide whether to run a measurement")
Signed-off-by: Oleksij Rempel <o.rempel@pengutronix.de>
---
 drivers/thermal/imx_thermal.c | 143 +++++++++++++++++++++-------------
 1 file changed, 89 insertions(+), 54 deletions(-)

Comments

Oleksij Rempel Oct. 20, 2021, 5:04 a.m. UTC | #1
Hi Petr and Michal,

I forgot to add you for v2 in CC. Please test/review this version.

On Tue, Oct 19, 2021 at 03:08:09PM +0200, Oleksij Rempel wrote:
> Starting with commit d92ed2c9d3ff ("thermal: imx: Use driver's local
> data to decide whether to run a measurement") this driver stared using
> irq_enabled flag to make decision to power on/off the thermal core. This
> triggered a regression, where after reaching critical temperature, alarm
> IRQ handler set irq_enabled to false,  disabled thermal core and was not
> able read temperature and disable cooling sequence.
> 
> In case the cooling device is "CPU/GPU freq", the system will run with
> reduce performance until next reboot.
> 
> To solve this issue, we need to move all parts implementing hand made
> runtime power management and let it handle actual runtime PM framework.
> 
> Fixes: d92ed2c9d3ff ("thermal: imx: Use driver's local data to decide whether to run a measurement")
> Signed-off-by: Oleksij Rempel <o.rempel@pengutronix.de>
> ---
>  drivers/thermal/imx_thermal.c | 143 +++++++++++++++++++++-------------
>  1 file changed, 89 insertions(+), 54 deletions(-)
> 
> diff --git a/drivers/thermal/imx_thermal.c b/drivers/thermal/imx_thermal.c
> index 2c7473d86a59..cb5a4354fc75 100644
> --- a/drivers/thermal/imx_thermal.c
> +++ b/drivers/thermal/imx_thermal.c
> @@ -15,6 +15,7 @@
>  #include <linux/regmap.h>
>  #include <linux/thermal.h>
>  #include <linux/nvmem-consumer.h>
> +#include <linux/pm_runtime.h>
>  
>  #define REG_SET		0x4
>  #define REG_CLR		0x8
> @@ -194,6 +195,7 @@ static struct thermal_soc_data thermal_imx7d_data = {
>  };
>  
>  struct imx_thermal_data {
> +	struct device *dev;
>  	struct cpufreq_policy *policy;
>  	struct thermal_zone_device *tz;
>  	struct thermal_cooling_device *cdev;
> @@ -252,44 +254,15 @@ static int imx_get_temp(struct thermal_zone_device *tz, int *temp)
>  	const struct thermal_soc_data *soc_data = data->socdata;
>  	struct regmap *map = data->tempmon;
>  	unsigned int n_meas;
> -	bool wait, run_measurement;
>  	u32 val;
> +	int ret;
>  
> -	run_measurement = !data->irq_enabled;
> -	if (!run_measurement) {
> -		/* Check if a measurement is currently in progress */
> -		regmap_read(map, soc_data->temp_data, &val);
> -		wait = !(val & soc_data->temp_valid_mask);
> -	} else {
> -		/*
> -		 * Every time we measure the temperature, we will power on the
> -		 * temperature sensor, enable measurements, take a reading,
> -		 * disable measurements, power off the temperature sensor.
> -		 */
> -		regmap_write(map, soc_data->sensor_ctrl + REG_CLR,
> -			    soc_data->power_down_mask);
> -		regmap_write(map, soc_data->sensor_ctrl + REG_SET,
> -			    soc_data->measure_temp_mask);
> -
> -		wait = true;
> -	}
> -
> -	/*
> -	 * According to the temp sensor designers, it may require up to ~17us
> -	 * to complete a measurement.
> -	 */
> -	if (wait)
> -		usleep_range(20, 50);
> +	ret = pm_runtime_resume_and_get(data->dev);
> +	if (ret < 0)
> +		return ret;
>  
>  	regmap_read(map, soc_data->temp_data, &val);
>  
> -	if (run_measurement) {
> -		regmap_write(map, soc_data->sensor_ctrl + REG_CLR,
> -			     soc_data->measure_temp_mask);
> -		regmap_write(map, soc_data->sensor_ctrl + REG_SET,
> -			     soc_data->power_down_mask);
> -	}
> -
>  	if ((val & soc_data->temp_valid_mask) == 0) {
>  		dev_dbg(&tz->device, "temp measurement never finished\n");
>  		return -EAGAIN;
> @@ -328,6 +301,8 @@ static int imx_get_temp(struct thermal_zone_device *tz, int *temp)
>  		enable_irq(data->irq);
>  	}
>  
> +	pm_runtime_put(data->dev);
> +
>  	return 0;
>  }
>  
> @@ -335,24 +310,16 @@ static int imx_change_mode(struct thermal_zone_device *tz,
>  			   enum thermal_device_mode mode)
>  {
>  	struct imx_thermal_data *data = tz->devdata;
> -	struct regmap *map = data->tempmon;
> -	const struct thermal_soc_data *soc_data = data->socdata;
>  
>  	if (mode == THERMAL_DEVICE_ENABLED) {
> -		regmap_write(map, soc_data->sensor_ctrl + REG_CLR,
> -			     soc_data->power_down_mask);
> -		regmap_write(map, soc_data->sensor_ctrl + REG_SET,
> -			     soc_data->measure_temp_mask);
> +		pm_runtime_get(data->dev);
>  
>  		if (!data->irq_enabled) {
>  			data->irq_enabled = true;
>  			enable_irq(data->irq);
>  		}
>  	} else {
> -		regmap_write(map, soc_data->sensor_ctrl + REG_CLR,
> -			     soc_data->measure_temp_mask);
> -		regmap_write(map, soc_data->sensor_ctrl + REG_SET,
> -			     soc_data->power_down_mask);
> +		pm_runtime_put(data->dev);
>  
>  		if (data->irq_enabled) {
>  			disable_irq(data->irq);
> @@ -393,6 +360,11 @@ static int imx_set_trip_temp(struct thermal_zone_device *tz, int trip,
>  			     int temp)
>  {
>  	struct imx_thermal_data *data = tz->devdata;
> +	int ret;
> +
> +	ret = pm_runtime_resume_and_get(data->dev);
> +	if (ret < 0)
> +		return ret;
>  
>  	/* do not allow changing critical threshold */
>  	if (trip == IMX_TRIP_CRITICAL)
> @@ -406,6 +378,8 @@ static int imx_set_trip_temp(struct thermal_zone_device *tz, int trip,
>  
>  	imx_set_alarm_temp(data, temp);
>  
> +	pm_runtime_put(data->dev);
> +
>  	return 0;
>  }
>  
> @@ -681,6 +655,8 @@ static int imx_thermal_probe(struct platform_device *pdev)
>  	if (!data)
>  		return -ENOMEM;
>  
> +	data->dev = &pdev->dev;
> +
>  	map = syscon_regmap_lookup_by_phandle(pdev->dev.of_node, "fsl,tempmon");
>  	if (IS_ERR(map)) {
>  		ret = PTR_ERR(map);
> @@ -801,6 +777,14 @@ static int imx_thermal_probe(struct platform_device *pdev)
>  	regmap_write(map, data->socdata->sensor_ctrl + REG_SET,
>  		     data->socdata->measure_temp_mask);
>  
> +	/* the core was configured and enabled just before */
> +	pm_runtime_set_active(&pdev->dev);
> +	pm_runtime_enable(data->dev);
> +
> +	ret = pm_runtime_resume_and_get(data->dev);
> +	if (ret < 0)
> +		goto disable_runtime_pm;
> +
>  	data->irq_enabled = true;
>  	ret = thermal_zone_device_enable(data->tz);
>  	if (ret)
> @@ -814,10 +798,15 @@ static int imx_thermal_probe(struct platform_device *pdev)
>  		goto thermal_zone_unregister;
>  	}
>  
> +	pm_runtime_put(data->dev);
> +
>  	return 0;
>  
>  thermal_zone_unregister:
>  	thermal_zone_device_unregister(data->tz);
> +disable_runtime_pm:
> +	pm_runtime_put_noidle(data->dev);
> +	pm_runtime_disable(data->dev);
>  clk_disable:
>  	clk_disable_unprepare(data->thermal_clk);
>  legacy_cleanup:
> @@ -829,13 +818,9 @@ static int imx_thermal_probe(struct platform_device *pdev)
>  static int imx_thermal_remove(struct platform_device *pdev)
>  {
>  	struct imx_thermal_data *data = platform_get_drvdata(pdev);
> -	struct regmap *map = data->tempmon;
>  
> -	/* Disable measurements */
> -	regmap_write(map, data->socdata->sensor_ctrl + REG_SET,
> -		     data->socdata->power_down_mask);
> -	if (!IS_ERR(data->thermal_clk))
> -		clk_disable_unprepare(data->thermal_clk);
> +	pm_runtime_put_noidle(data->dev);
> +	pm_runtime_disable(data->dev);
>  
>  	thermal_zone_device_unregister(data->tz);
>  	imx_thermal_unregister_legacy_cooling(data);
> @@ -858,29 +843,79 @@ static int __maybe_unused imx_thermal_suspend(struct device *dev)
>  	ret = thermal_zone_device_disable(data->tz);
>  	if (ret)
>  		return ret;
> +
> +	return pm_runtime_force_suspend(data->dev);
> +}
> +
> +static int __maybe_unused imx_thermal_resume(struct device *dev)
> +{
> +	struct imx_thermal_data *data = dev_get_drvdata(dev);
> +	int ret;
> +
> +	ret = pm_runtime_force_resume(data->dev);
> +	if (ret)
> +		return ret;
> +	/* Enabled thermal sensor after resume */
> +	return thermal_zone_device_enable(data->tz);
> +}
> +
> +static int __maybe_unused imx_thermal_runtime_suspend(struct device *dev)
> +{
> +	struct imx_thermal_data *data = dev_get_drvdata(dev);
> +	const struct thermal_soc_data *socdata = data->socdata;
> +	struct regmap *map = data->tempmon;
> +	int ret;
> +
> +	ret = regmap_write(map, socdata->sensor_ctrl + REG_CLR,
> +			   socdata->measure_temp_mask);
> +	if (ret)
> +		return ret;
> +
> +	ret = regmap_write(map, socdata->sensor_ctrl + REG_SET,
> +			   socdata->power_down_mask);
> +	if (ret)
> +		return ret;
> +
>  	clk_disable_unprepare(data->thermal_clk);
>  
>  	return 0;
>  }
>  
> -static int __maybe_unused imx_thermal_resume(struct device *dev)
> +static int __maybe_unused imx_thermal_runtime_resume(struct device *dev)
>  {
>  	struct imx_thermal_data *data = dev_get_drvdata(dev);
> +	const struct thermal_soc_data *socdata = data->socdata;
> +	struct regmap *map = data->tempmon;
>  	int ret;
>  
>  	ret = clk_prepare_enable(data->thermal_clk);
>  	if (ret)
>  		return ret;
> -	/* Enabled thermal sensor after resume */
> -	ret = thermal_zone_device_enable(data->tz);
> +
> +	ret = regmap_write(map, socdata->sensor_ctrl + REG_CLR,
> +			   socdata->power_down_mask);
> +	if (ret)
> +		return ret;
> +
> +	ret = regmap_write(map, socdata->sensor_ctrl + REG_SET,
> +			   socdata->measure_temp_mask);
>  	if (ret)
>  		return ret;
>  
> +	/*
> +	 * According to the temp sensor designers, it may require up to ~17us
> +	 * to complete a measurement.
> +	 */
> +	usleep_range(20, 50);
> +
>  	return 0;
>  }
>  
> -static SIMPLE_DEV_PM_OPS(imx_thermal_pm_ops,
> -			 imx_thermal_suspend, imx_thermal_resume);
> +static const struct dev_pm_ops imx_thermal_pm_ops = {
> +	SET_SYSTEM_SLEEP_PM_OPS(imx_thermal_suspend, imx_thermal_resume)
> +	SET_RUNTIME_PM_OPS(imx_thermal_runtime_suspend,
> +			   imx_thermal_runtime_resume, NULL)
> +};
>  
>  static struct platform_driver imx_thermal = {
>  	.driver = {
> -- 
> 2.30.2
> 
>
Petr Benes Oct. 20, 2021, 3:53 p.m. UTC | #2
On Wed, 20 Oct 2021 at 07:05, Oleksij Rempel <o.rempel@pengutronix.de> wrote:
>
> Hi Petr and Michal,
>
> I forgot to add you for v2 in CC. Please test/review this version.

Hi Oleksij,

It works good. with PM as well as without PM. The only minor issue I found is,
that the first temperature reading (when the driver probes) fails. That is
(val & soc_data->temp_valid_mask) == 0) holds true. How does
pm_runtime_resume_and_get() behave in imx_thermal_probe()?
Does it go through imx_thermal_runtime_resume() with usleep_range()?

>
> On Tue, Oct 19, 2021 at 03:08:09PM +0200, Oleksij Rempel wrote:
> > Starting with commit d92ed2c9d3ff ("thermal: imx: Use driver's local
> > data to decide whether to run a measurement") this driver stared using
> > irq_enabled flag to make decision to power on/off the thermal core. This
> > triggered a regression, where after reaching critical temperature, alarm
> > IRQ handler set irq_enabled to false,  disabled thermal core and was not
> > able read temperature and disable cooling sequence.
> >
> > In case the cooling device is "CPU/GPU freq", the system will run with
> > reduce performance until next reboot.
> >
> > To solve this issue, we need to move all parts implementing hand made
> > runtime power management and let it handle actual runtime PM framework.
> >
> > Fixes: d92ed2c9d3ff ("thermal: imx: Use driver's local data to decide whether to run a measurement")
> > Signed-off-by: Oleksij Rempel <o.rempel@pengutronix.de>
> > ---
> >  drivers/thermal/imx_thermal.c | 143 +++++++++++++++++++++-------------
> >  1 file changed, 89 insertions(+), 54 deletions(-)
> >
> > diff --git a/drivers/thermal/imx_thermal.c b/drivers/thermal/imx_thermal.c
> > index 2c7473d86a59..cb5a4354fc75 100644
> > --- a/drivers/thermal/imx_thermal.c
> > +++ b/drivers/thermal/imx_thermal.c
> > @@ -15,6 +15,7 @@
> >  #include <linux/regmap.h>
> >  #include <linux/thermal.h>
> >  #include <linux/nvmem-consumer.h>
> > +#include <linux/pm_runtime.h>
> >
> >  #define REG_SET              0x4
> >  #define REG_CLR              0x8
> > @@ -194,6 +195,7 @@ static struct thermal_soc_data thermal_imx7d_data = {
> >  };
> >
> >  struct imx_thermal_data {
> > +     struct device *dev;
> >       struct cpufreq_policy *policy;
> >       struct thermal_zone_device *tz;
> >       struct thermal_cooling_device *cdev;
> > @@ -252,44 +254,15 @@ static int imx_get_temp(struct thermal_zone_device *tz, int *temp)
> >       const struct thermal_soc_data *soc_data = data->socdata;
> >       struct regmap *map = data->tempmon;
> >       unsigned int n_meas;
> > -     bool wait, run_measurement;
> >       u32 val;
> > +     int ret;
> >
> > -     run_measurement = !data->irq_enabled;
> > -     if (!run_measurement) {
> > -             /* Check if a measurement is currently in progress */
> > -             regmap_read(map, soc_data->temp_data, &val);
> > -             wait = !(val & soc_data->temp_valid_mask);
> > -     } else {
> > -             /*
> > -              * Every time we measure the temperature, we will power on the
> > -              * temperature sensor, enable measurements, take a reading,
> > -              * disable measurements, power off the temperature sensor.
> > -              */
> > -             regmap_write(map, soc_data->sensor_ctrl + REG_CLR,
> > -                         soc_data->power_down_mask);
> > -             regmap_write(map, soc_data->sensor_ctrl + REG_SET,
> > -                         soc_data->measure_temp_mask);
> > -
> > -             wait = true;
> > -     }
> > -
> > -     /*
> > -      * According to the temp sensor designers, it may require up to ~17us
> > -      * to complete a measurement.
> > -      */
> > -     if (wait)
> > -             usleep_range(20, 50);
> > +     ret = pm_runtime_resume_and_get(data->dev);
> > +     if (ret < 0)
> > +             return ret;
> >
> >       regmap_read(map, soc_data->temp_data, &val);
> >
> > -     if (run_measurement) {
> > -             regmap_write(map, soc_data->sensor_ctrl + REG_CLR,
> > -                          soc_data->measure_temp_mask);
> > -             regmap_write(map, soc_data->sensor_ctrl + REG_SET,
> > -                          soc_data->power_down_mask);
> > -     }
> > -
> >       if ((val & soc_data->temp_valid_mask) == 0) {
> >               dev_dbg(&tz->device, "temp measurement never finished\n");
> >               return -EAGAIN;
> > @@ -328,6 +301,8 @@ static int imx_get_temp(struct thermal_zone_device *tz, int *temp)
> >               enable_irq(data->irq);
> >       }
> >
> > +     pm_runtime_put(data->dev);
> > +
> >       return 0;
> >  }
> >
> > @@ -335,24 +310,16 @@ static int imx_change_mode(struct thermal_zone_device *tz,
> >                          enum thermal_device_mode mode)
> >  {
> >       struct imx_thermal_data *data = tz->devdata;
> > -     struct regmap *map = data->tempmon;
> > -     const struct thermal_soc_data *soc_data = data->socdata;
> >
> >       if (mode == THERMAL_DEVICE_ENABLED) {
> > -             regmap_write(map, soc_data->sensor_ctrl + REG_CLR,
> > -                          soc_data->power_down_mask);
> > -             regmap_write(map, soc_data->sensor_ctrl + REG_SET,
> > -                          soc_data->measure_temp_mask);
> > +             pm_runtime_get(data->dev);
> >
> >               if (!data->irq_enabled) {
> >                       data->irq_enabled = true;
> >                       enable_irq(data->irq);
> >               }
> >       } else {
> > -             regmap_write(map, soc_data->sensor_ctrl + REG_CLR,
> > -                          soc_data->measure_temp_mask);
> > -             regmap_write(map, soc_data->sensor_ctrl + REG_SET,
> > -                          soc_data->power_down_mask);
> > +             pm_runtime_put(data->dev);
> >
> >               if (data->irq_enabled) {
> >                       disable_irq(data->irq);
> > @@ -393,6 +360,11 @@ static int imx_set_trip_temp(struct thermal_zone_device *tz, int trip,
> >                            int temp)
> >  {
> >       struct imx_thermal_data *data = tz->devdata;
> > +     int ret;
> > +
> > +     ret = pm_runtime_resume_and_get(data->dev);
> > +     if (ret < 0)
> > +             return ret;
> >
> >       /* do not allow changing critical threshold */
> >       if (trip == IMX_TRIP_CRITICAL)
> > @@ -406,6 +378,8 @@ static int imx_set_trip_temp(struct thermal_zone_device *tz, int trip,
> >
> >       imx_set_alarm_temp(data, temp);
> >
> > +     pm_runtime_put(data->dev);
> > +
> >       return 0;
> >  }
> >
> > @@ -681,6 +655,8 @@ static int imx_thermal_probe(struct platform_device *pdev)
> >       if (!data)
> >               return -ENOMEM;
> >
> > +     data->dev = &pdev->dev;
> > +
> >       map = syscon_regmap_lookup_by_phandle(pdev->dev.of_node, "fsl,tempmon");
> >       if (IS_ERR(map)) {
> >               ret = PTR_ERR(map);
> > @@ -801,6 +777,14 @@ static int imx_thermal_probe(struct platform_device *pdev)
> >       regmap_write(map, data->socdata->sensor_ctrl + REG_SET,
> >                    data->socdata->measure_temp_mask);
> >
> > +     /* the core was configured and enabled just before */
> > +     pm_runtime_set_active(&pdev->dev);
> > +     pm_runtime_enable(data->dev);
> > +
> > +     ret = pm_runtime_resume_and_get(data->dev);
> > +     if (ret < 0)
> > +             goto disable_runtime_pm;
> > +
> >       data->irq_enabled = true;
> >       ret = thermal_zone_device_enable(data->tz);
> >       if (ret)
> > @@ -814,10 +798,15 @@ static int imx_thermal_probe(struct platform_device *pdev)
> >               goto thermal_zone_unregister;
> >       }
> >
> > +     pm_runtime_put(data->dev);
> > +
> >       return 0;
> >
> >  thermal_zone_unregister:
> >       thermal_zone_device_unregister(data->tz);
> > +disable_runtime_pm:
> > +     pm_runtime_put_noidle(data->dev);
> > +     pm_runtime_disable(data->dev);
> >  clk_disable:
> >       clk_disable_unprepare(data->thermal_clk);
> >  legacy_cleanup:
> > @@ -829,13 +818,9 @@ static int imx_thermal_probe(struct platform_device *pdev)
> >  static int imx_thermal_remove(struct platform_device *pdev)
> >  {
> >       struct imx_thermal_data *data = platform_get_drvdata(pdev);
> > -     struct regmap *map = data->tempmon;
> >
> > -     /* Disable measurements */
> > -     regmap_write(map, data->socdata->sensor_ctrl + REG_SET,
> > -                  data->socdata->power_down_mask);
> > -     if (!IS_ERR(data->thermal_clk))
> > -             clk_disable_unprepare(data->thermal_clk);
> > +     pm_runtime_put_noidle(data->dev);
> > +     pm_runtime_disable(data->dev);
> >
> >       thermal_zone_device_unregister(data->tz);
> >       imx_thermal_unregister_legacy_cooling(data);
> > @@ -858,29 +843,79 @@ static int __maybe_unused imx_thermal_suspend(struct device *dev)
> >       ret = thermal_zone_device_disable(data->tz);
> >       if (ret)
> >               return ret;
> > +
> > +     return pm_runtime_force_suspend(data->dev);
> > +}
> > +
> > +static int __maybe_unused imx_thermal_resume(struct device *dev)
> > +{
> > +     struct imx_thermal_data *data = dev_get_drvdata(dev);
> > +     int ret;
> > +
> > +     ret = pm_runtime_force_resume(data->dev);
> > +     if (ret)
> > +             return ret;
> > +     /* Enabled thermal sensor after resume */
> > +     return thermal_zone_device_enable(data->tz);
> > +}
> > +
> > +static int __maybe_unused imx_thermal_runtime_suspend(struct device *dev)
> > +{
> > +     struct imx_thermal_data *data = dev_get_drvdata(dev);
> > +     const struct thermal_soc_data *socdata = data->socdata;
> > +     struct regmap *map = data->tempmon;
> > +     int ret;
> > +
> > +     ret = regmap_write(map, socdata->sensor_ctrl + REG_CLR,
> > +                        socdata->measure_temp_mask);
> > +     if (ret)
> > +             return ret;
> > +
> > +     ret = regmap_write(map, socdata->sensor_ctrl + REG_SET,
> > +                        socdata->power_down_mask);
> > +     if (ret)
> > +             return ret;
> > +
> >       clk_disable_unprepare(data->thermal_clk);
> >
> >       return 0;
> >  }
> >
> > -static int __maybe_unused imx_thermal_resume(struct device *dev)
> > +static int __maybe_unused imx_thermal_runtime_resume(struct device *dev)
> >  {
> >       struct imx_thermal_data *data = dev_get_drvdata(dev);
> > +     const struct thermal_soc_data *socdata = data->socdata;
> > +     struct regmap *map = data->tempmon;
> >       int ret;
> >
> >       ret = clk_prepare_enable(data->thermal_clk);
> >       if (ret)
> >               return ret;
> > -     /* Enabled thermal sensor after resume */
> > -     ret = thermal_zone_device_enable(data->tz);
> > +
> > +     ret = regmap_write(map, socdata->sensor_ctrl + REG_CLR,
> > +                        socdata->power_down_mask);
> > +     if (ret)
> > +             return ret;
> > +
> > +     ret = regmap_write(map, socdata->sensor_ctrl + REG_SET,
> > +                        socdata->measure_temp_mask);
> >       if (ret)
> >               return ret;
> >
> > +     /*
> > +      * According to the temp sensor designers, it may require up to ~17us
> > +      * to complete a measurement.
> > +      */
> > +     usleep_range(20, 50);
> > +
> >       return 0;
> >  }
> >
> > -static SIMPLE_DEV_PM_OPS(imx_thermal_pm_ops,
> > -                      imx_thermal_suspend, imx_thermal_resume);
> > +static const struct dev_pm_ops imx_thermal_pm_ops = {
> > +     SET_SYSTEM_SLEEP_PM_OPS(imx_thermal_suspend, imx_thermal_resume)
> > +     SET_RUNTIME_PM_OPS(imx_thermal_runtime_suspend,
> > +                        imx_thermal_runtime_resume, NULL)
> > +};
> >
> >  static struct platform_driver imx_thermal = {
> >       .driver = {
> > --
> > 2.30.2
> >
> >
>
> --
> Pengutronix e.K.                           |                             |
> Steuerwalder Str. 21                       | http://www.pengutronix.de/  |
> 31137 Hildesheim, Germany                  | Phone: +49-5121-206917-0    |
> Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |
Oleksij Rempel Oct. 21, 2021, 7:20 a.m. UTC | #3
Hi Petr,

On Wed, Oct 20, 2021 at 05:53:03PM +0200, Petr Benes wrote:
> On Wed, 20 Oct 2021 at 07:05, Oleksij Rempel <o.rempel@pengutronix.de> wrote:
> >
> > Hi Petr and Michal,
> >
> > I forgot to add you for v2 in CC. Please test/review this version.
> 
> Hi Oleksij,
> 
> It works good. with PM as well as without PM. The only minor issue I found is,
> that the first temperature reading (when the driver probes) fails. That is
> (val & soc_data->temp_valid_mask) == 0) holds true. How does
> pm_runtime_resume_and_get() behave in imx_thermal_probe()?
> Does it go through imx_thermal_runtime_resume() with usleep_range()?

On the first temperature reading, the PM and part of HW is not
initialized. Current probe sequence is racy and has at least following
issues:
- thermal_zone_device_register is executed before HW init was completed.
  It kind of worked before my patch, becaus part of reinit was done by
  temperature init. It  worked, since the irq_enabled flag was not set,
  but potentially would run enable_irq() two times if device is
  overheated on probe.
- the imx_thermal core is potentially disable after first race
  condition:
  CPU0					CPU1
  thermal_zone_device_register()
					imx_get_temp()
  					irq_enabled == false
						power_up
						read_temp
  power_up
  						power_down
  irq_enabled = true;

  ... at this point imx_thermal is powered down for some amount of time,
  over temperature IRQ will not be triggered for some amount of time.

- if some part after thermal_zone_device_register() would fail or
  deferred, the worker polling temperature will run in to NULL pointer.
  This issue already happened...

After migrating to runtime PM, one of issues started to be visible even
on normal conditions.
I'll send one more patch with reworking probe sequence.

> >
> > On Tue, Oct 19, 2021 at 03:08:09PM +0200, Oleksij Rempel wrote:
> > > Starting with commit d92ed2c9d3ff ("thermal: imx: Use driver's local
> > > data to decide whether to run a measurement") this driver stared using
> > > irq_enabled flag to make decision to power on/off the thermal core. This
> > > triggered a regression, where after reaching critical temperature, alarm
> > > IRQ handler set irq_enabled to false,  disabled thermal core and was not
> > > able read temperature and disable cooling sequence.
> > >
> > > In case the cooling device is "CPU/GPU freq", the system will run with
> > > reduce performance until next reboot.
> > >
> > > To solve this issue, we need to move all parts implementing hand made
> > > runtime power management and let it handle actual runtime PM framework.
> > >
> > > Fixes: d92ed2c9d3ff ("thermal: imx: Use driver's local data to decide whether to run a measurement")
> > > Signed-off-by: Oleksij Rempel <o.rempel@pengutronix.de>
> > > ---
> > >  drivers/thermal/imx_thermal.c | 143 +++++++++++++++++++++-------------
> > >  1 file changed, 89 insertions(+), 54 deletions(-)
> > >
> > > diff --git a/drivers/thermal/imx_thermal.c b/drivers/thermal/imx_thermal.c
> > > index 2c7473d86a59..cb5a4354fc75 100644
> > > --- a/drivers/thermal/imx_thermal.c
> > > +++ b/drivers/thermal/imx_thermal.c
> > > @@ -15,6 +15,7 @@
> > >  #include <linux/regmap.h>
> > >  #include <linux/thermal.h>
> > >  #include <linux/nvmem-consumer.h>
> > > +#include <linux/pm_runtime.h>
> > >
> > >  #define REG_SET              0x4
> > >  #define REG_CLR              0x8
> > > @@ -194,6 +195,7 @@ static struct thermal_soc_data thermal_imx7d_data = {
> > >  };
> > >
> > >  struct imx_thermal_data {
> > > +     struct device *dev;
> > >       struct cpufreq_policy *policy;
> > >       struct thermal_zone_device *tz;
> > >       struct thermal_cooling_device *cdev;
> > > @@ -252,44 +254,15 @@ static int imx_get_temp(struct thermal_zone_device *tz, int *temp)
> > >       const struct thermal_soc_data *soc_data = data->socdata;
> > >       struct regmap *map = data->tempmon;
> > >       unsigned int n_meas;
> > > -     bool wait, run_measurement;
> > >       u32 val;
> > > +     int ret;
> > >
> > > -     run_measurement = !data->irq_enabled;
> > > -     if (!run_measurement) {
> > > -             /* Check if a measurement is currently in progress */
> > > -             regmap_read(map, soc_data->temp_data, &val);
> > > -             wait = !(val & soc_data->temp_valid_mask);
> > > -     } else {
> > > -             /*
> > > -              * Every time we measure the temperature, we will power on the
> > > -              * temperature sensor, enable measurements, take a reading,
> > > -              * disable measurements, power off the temperature sensor.
> > > -              */
> > > -             regmap_write(map, soc_data->sensor_ctrl + REG_CLR,
> > > -                         soc_data->power_down_mask);
> > > -             regmap_write(map, soc_data->sensor_ctrl + REG_SET,
> > > -                         soc_data->measure_temp_mask);
> > > -
> > > -             wait = true;
> > > -     }
> > > -
> > > -     /*
> > > -      * According to the temp sensor designers, it may require up to ~17us
> > > -      * to complete a measurement.
> > > -      */
> > > -     if (wait)
> > > -             usleep_range(20, 50);
> > > +     ret = pm_runtime_resume_and_get(data->dev);
> > > +     if (ret < 0)
> > > +             return ret;
> > >
> > >       regmap_read(map, soc_data->temp_data, &val);
> > >
> > > -     if (run_measurement) {
> > > -             regmap_write(map, soc_data->sensor_ctrl + REG_CLR,
> > > -                          soc_data->measure_temp_mask);
> > > -             regmap_write(map, soc_data->sensor_ctrl + REG_SET,
> > > -                          soc_data->power_down_mask);
> > > -     }
> > > -
> > >       if ((val & soc_data->temp_valid_mask) == 0) {
> > >               dev_dbg(&tz->device, "temp measurement never finished\n");
> > >               return -EAGAIN;
> > > @@ -328,6 +301,8 @@ static int imx_get_temp(struct thermal_zone_device *tz, int *temp)
> > >               enable_irq(data->irq);
> > >       }
> > >
> > > +     pm_runtime_put(data->dev);
> > > +
> > >       return 0;
> > >  }
> > >
> > > @@ -335,24 +310,16 @@ static int imx_change_mode(struct thermal_zone_device *tz,
> > >                          enum thermal_device_mode mode)
> > >  {
> > >       struct imx_thermal_data *data = tz->devdata;
> > > -     struct regmap *map = data->tempmon;
> > > -     const struct thermal_soc_data *soc_data = data->socdata;
> > >
> > >       if (mode == THERMAL_DEVICE_ENABLED) {
> > > -             regmap_write(map, soc_data->sensor_ctrl + REG_CLR,
> > > -                          soc_data->power_down_mask);
> > > -             regmap_write(map, soc_data->sensor_ctrl + REG_SET,
> > > -                          soc_data->measure_temp_mask);
> > > +             pm_runtime_get(data->dev);
> > >
> > >               if (!data->irq_enabled) {
> > >                       data->irq_enabled = true;
> > >                       enable_irq(data->irq);
> > >               }
> > >       } else {
> > > -             regmap_write(map, soc_data->sensor_ctrl + REG_CLR,
> > > -                          soc_data->measure_temp_mask);
> > > -             regmap_write(map, soc_data->sensor_ctrl + REG_SET,
> > > -                          soc_data->power_down_mask);
> > > +             pm_runtime_put(data->dev);
> > >
> > >               if (data->irq_enabled) {
> > >                       disable_irq(data->irq);
> > > @@ -393,6 +360,11 @@ static int imx_set_trip_temp(struct thermal_zone_device *tz, int trip,
> > >                            int temp)
> > >  {
> > >       struct imx_thermal_data *data = tz->devdata;
> > > +     int ret;
> > > +
> > > +     ret = pm_runtime_resume_and_get(data->dev);
> > > +     if (ret < 0)
> > > +             return ret;
> > >
> > >       /* do not allow changing critical threshold */
> > >       if (trip == IMX_TRIP_CRITICAL)
> > > @@ -406,6 +378,8 @@ static int imx_set_trip_temp(struct thermal_zone_device *tz, int trip,
> > >
> > >       imx_set_alarm_temp(data, temp);
> > >
> > > +     pm_runtime_put(data->dev);
> > > +
> > >       return 0;
> > >  }
> > >
> > > @@ -681,6 +655,8 @@ static int imx_thermal_probe(struct platform_device *pdev)
> > >       if (!data)
> > >               return -ENOMEM;
> > >
> > > +     data->dev = &pdev->dev;
> > > +
> > >       map = syscon_regmap_lookup_by_phandle(pdev->dev.of_node, "fsl,tempmon");
> > >       if (IS_ERR(map)) {
> > >               ret = PTR_ERR(map);
> > > @@ -801,6 +777,14 @@ static int imx_thermal_probe(struct platform_device *pdev)
> > >       regmap_write(map, data->socdata->sensor_ctrl + REG_SET,
> > >                    data->socdata->measure_temp_mask);
> > >
> > > +     /* the core was configured and enabled just before */
> > > +     pm_runtime_set_active(&pdev->dev);
> > > +     pm_runtime_enable(data->dev);
> > > +
> > > +     ret = pm_runtime_resume_and_get(data->dev);
> > > +     if (ret < 0)
> > > +             goto disable_runtime_pm;
> > > +
> > >       data->irq_enabled = true;
> > >       ret = thermal_zone_device_enable(data->tz);
> > >       if (ret)
> > > @@ -814,10 +798,15 @@ static int imx_thermal_probe(struct platform_device *pdev)
> > >               goto thermal_zone_unregister;
> > >       }
> > >
> > > +     pm_runtime_put(data->dev);
> > > +
> > >       return 0;
> > >
> > >  thermal_zone_unregister:
> > >       thermal_zone_device_unregister(data->tz);
> > > +disable_runtime_pm:
> > > +     pm_runtime_put_noidle(data->dev);
> > > +     pm_runtime_disable(data->dev);
> > >  clk_disable:
> > >       clk_disable_unprepare(data->thermal_clk);
> > >  legacy_cleanup:
> > > @@ -829,13 +818,9 @@ static int imx_thermal_probe(struct platform_device *pdev)
> > >  static int imx_thermal_remove(struct platform_device *pdev)
> > >  {
> > >       struct imx_thermal_data *data = platform_get_drvdata(pdev);
> > > -     struct regmap *map = data->tempmon;
> > >
> > > -     /* Disable measurements */
> > > -     regmap_write(map, data->socdata->sensor_ctrl + REG_SET,
> > > -                  data->socdata->power_down_mask);
> > > -     if (!IS_ERR(data->thermal_clk))
> > > -             clk_disable_unprepare(data->thermal_clk);
> > > +     pm_runtime_put_noidle(data->dev);
> > > +     pm_runtime_disable(data->dev);
> > >
> > >       thermal_zone_device_unregister(data->tz);
> > >       imx_thermal_unregister_legacy_cooling(data);
> > > @@ -858,29 +843,79 @@ static int __maybe_unused imx_thermal_suspend(struct device *dev)
> > >       ret = thermal_zone_device_disable(data->tz);
> > >       if (ret)
> > >               return ret;
> > > +
> > > +     return pm_runtime_force_suspend(data->dev);
> > > +}
> > > +
> > > +static int __maybe_unused imx_thermal_resume(struct device *dev)
> > > +{
> > > +     struct imx_thermal_data *data = dev_get_drvdata(dev);
> > > +     int ret;
> > > +
> > > +     ret = pm_runtime_force_resume(data->dev);
> > > +     if (ret)
> > > +             return ret;
> > > +     /* Enabled thermal sensor after resume */
> > > +     return thermal_zone_device_enable(data->tz);
> > > +}
> > > +
> > > +static int __maybe_unused imx_thermal_runtime_suspend(struct device *dev)
> > > +{
> > > +     struct imx_thermal_data *data = dev_get_drvdata(dev);
> > > +     const struct thermal_soc_data *socdata = data->socdata;
> > > +     struct regmap *map = data->tempmon;
> > > +     int ret;
> > > +
> > > +     ret = regmap_write(map, socdata->sensor_ctrl + REG_CLR,
> > > +                        socdata->measure_temp_mask);
> > > +     if (ret)
> > > +             return ret;
> > > +
> > > +     ret = regmap_write(map, socdata->sensor_ctrl + REG_SET,
> > > +                        socdata->power_down_mask);
> > > +     if (ret)
> > > +             return ret;
> > > +
> > >       clk_disable_unprepare(data->thermal_clk);
> > >
> > >       return 0;
> > >  }
> > >
> > > -static int __maybe_unused imx_thermal_resume(struct device *dev)
> > > +static int __maybe_unused imx_thermal_runtime_resume(struct device *dev)
> > >  {
> > >       struct imx_thermal_data *data = dev_get_drvdata(dev);
> > > +     const struct thermal_soc_data *socdata = data->socdata;
> > > +     struct regmap *map = data->tempmon;
> > >       int ret;
> > >
> > >       ret = clk_prepare_enable(data->thermal_clk);
> > >       if (ret)
> > >               return ret;
> > > -     /* Enabled thermal sensor after resume */
> > > -     ret = thermal_zone_device_enable(data->tz);
> > > +
> > > +     ret = regmap_write(map, socdata->sensor_ctrl + REG_CLR,
> > > +                        socdata->power_down_mask);
> > > +     if (ret)
> > > +             return ret;
> > > +
> > > +     ret = regmap_write(map, socdata->sensor_ctrl + REG_SET,
> > > +                        socdata->measure_temp_mask);
> > >       if (ret)
> > >               return ret;
> > >
> > > +     /*
> > > +      * According to the temp sensor designers, it may require up to ~17us
> > > +      * to complete a measurement.
> > > +      */
> > > +     usleep_range(20, 50);
> > > +
> > >       return 0;
> > >  }
> > >
> > > -static SIMPLE_DEV_PM_OPS(imx_thermal_pm_ops,
> > > -                      imx_thermal_suspend, imx_thermal_resume);
> > > +static const struct dev_pm_ops imx_thermal_pm_ops = {
> > > +     SET_SYSTEM_SLEEP_PM_OPS(imx_thermal_suspend, imx_thermal_resume)
> > > +     SET_RUNTIME_PM_OPS(imx_thermal_runtime_suspend,
> > > +                        imx_thermal_runtime_resume, NULL)
> > > +};
> > >
> > >  static struct platform_driver imx_thermal = {
> > >       .driver = {
> > > --
> > > 2.30.2
> > >
> > >
> >
> > --
> > Pengutronix e.K.                           |                             |
> > Steuerwalder Str. 21                       | http://www.pengutronix.de/  |
> > 31137 Hildesheim, Germany                  | Phone: +49-5121-206917-0    |
> > Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |
>
Daniel Lezcano Oct. 21, 2021, 7:41 a.m. UTC | #4
On 21/10/2021 09:20, Oleksij Rempel wrote:
> Hi Petr,
> 
> On Wed, Oct 20, 2021 at 05:53:03PM +0200, Petr Benes wrote:
>> On Wed, 20 Oct 2021 at 07:05, Oleksij Rempel <o.rempel@pengutronix.de> wrote:
>>>
>>> Hi Petr and Michal,
>>>
>>> I forgot to add you for v2 in CC. Please test/review this version.
>>
>> Hi Oleksij,
>>
>> It works good. with PM as well as without PM. The only minor issue I found is,
>> that the first temperature reading (when the driver probes) fails. That is
>> (val & soc_data->temp_valid_mask) == 0) holds true. How does
>> pm_runtime_resume_and_get() behave in imx_thermal_probe()?
>> Does it go through imx_thermal_runtime_resume() with usleep_range()?
> 
> On the first temperature reading, the PM and part of HW is not
> initialized. Current probe sequence is racy and has at least following
> issues:
> - thermal_zone_device_register is executed before HW init was completed.
>   It kind of worked before my patch, becaus part of reinit was done by
>   temperature init. It  worked, since the irq_enabled flag was not set,
>   but potentially would run enable_irq() two times if device is
>   overheated on probe.
> - the imx_thermal core is potentially disable after first race
>   condition:
>   CPU0					CPU1
>   thermal_zone_device_register()
> 					imx_get_temp()
>   					irq_enabled == false
> 						power_up
> 						read_temp
>   power_up
>   						power_down
>   irq_enabled = true;
> 
>   ... at this point imx_thermal is powered down for some amount of time,
>   over temperature IRQ will not be triggered for some amount of time.
> 
> - if some part after thermal_zone_device_register() would fail or
>   deferred, the worker polling temperature will run in to NULL pointer.
>   This issue already happened...
> 
> After migrating to runtime PM, one of issues started to be visible even
> on normal conditions.
> I'll send one more patch with reworking probe sequence.

Are you planning to send a v3 with this patch? Or a separate patch?

[ ... ]
Oleksij Rempel Oct. 21, 2021, 7:44 a.m. UTC | #5
On Thu, Oct 21, 2021 at 09:41:35AM +0200, Daniel Lezcano wrote:
> On 21/10/2021 09:20, Oleksij Rempel wrote:
> > Hi Petr,
> > 
> > On Wed, Oct 20, 2021 at 05:53:03PM +0200, Petr Benes wrote:
> >> On Wed, 20 Oct 2021 at 07:05, Oleksij Rempel <o.rempel@pengutronix.de> wrote:
> >>>
> >>> Hi Petr and Michal,
> >>>
> >>> I forgot to add you for v2 in CC. Please test/review this version.
> >>
> >> Hi Oleksij,
> >>
> >> It works good. with PM as well as without PM. The only minor issue I found is,
> >> that the first temperature reading (when the driver probes) fails. That is
> >> (val & soc_data->temp_valid_mask) == 0) holds true. How does
> >> pm_runtime_resume_and_get() behave in imx_thermal_probe()?
> >> Does it go through imx_thermal_runtime_resume() with usleep_range()?
> > 
> > On the first temperature reading, the PM and part of HW is not
> > initialized. Current probe sequence is racy and has at least following
> > issues:
> > - thermal_zone_device_register is executed before HW init was completed.
> >   It kind of worked before my patch, becaus part of reinit was done by
> >   temperature init. It  worked, since the irq_enabled flag was not set,
> >   but potentially would run enable_irq() two times if device is
> >   overheated on probe.
> > - the imx_thermal core is potentially disable after first race
> >   condition:
> >   CPU0					CPU1
> >   thermal_zone_device_register()
> > 					imx_get_temp()
> >   					irq_enabled == false
> > 						power_up
> > 						read_temp
> >   power_up
> >   						power_down
> >   irq_enabled = true;
> > 
> >   ... at this point imx_thermal is powered down for some amount of time,
> >   over temperature IRQ will not be triggered for some amount of time.
> > 
> > - if some part after thermal_zone_device_register() would fail or
> >   deferred, the worker polling temperature will run in to NULL pointer.
> >   This issue already happened...
> > 
> > After migrating to runtime PM, one of issues started to be visible even
> > on normal conditions.
> > I'll send one more patch with reworking probe sequence.
> 
> Are you planning to send a v3 with this patch? Or a separate patch?

I'm OK with both variants. What do you prefer?

I'll do i on top of PM patch to reduce refactoring overhead, if you OK
about it.

Regards,
Oleksij
Daniel Lezcano Oct. 21, 2021, 7:56 a.m. UTC | #6
On 21/10/2021 09:44, Oleksij Rempel wrote:
> On Thu, Oct 21, 2021 at 09:41:35AM +0200, Daniel Lezcano wrote:
>> On 21/10/2021 09:20, Oleksij Rempel wrote:
>>> Hi Petr,
>>>
>>> On Wed, Oct 20, 2021 at 05:53:03PM +0200, Petr Benes wrote:
>>>> On Wed, 20 Oct 2021 at 07:05, Oleksij Rempel <o.rempel@pengutronix.de> wrote:
>>>>>
>>>>> Hi Petr and Michal,
>>>>>
>>>>> I forgot to add you for v2 in CC. Please test/review this version.
>>>>
>>>> Hi Oleksij,
>>>>
>>>> It works good. with PM as well as without PM. The only minor issue I found is,
>>>> that the first temperature reading (when the driver probes) fails. That is
>>>> (val & soc_data->temp_valid_mask) == 0) holds true. How does
>>>> pm_runtime_resume_and_get() behave in imx_thermal_probe()?
>>>> Does it go through imx_thermal_runtime_resume() with usleep_range()?
>>>
>>> On the first temperature reading, the PM and part of HW is not
>>> initialized. Current probe sequence is racy and has at least following
>>> issues:
>>> - thermal_zone_device_register is executed before HW init was completed.
>>>   It kind of worked before my patch, becaus part of reinit was done by
>>>   temperature init. It  worked, since the irq_enabled flag was not set,
>>>   but potentially would run enable_irq() two times if device is
>>>   overheated on probe.
>>> - the imx_thermal core is potentially disable after first race
>>>   condition:
>>>   CPU0					CPU1
>>>   thermal_zone_device_register()
>>> 					imx_get_temp()
>>>   					irq_enabled == false
>>> 						power_up
>>> 						read_temp
>>>   power_up
>>>   						power_down
>>>   irq_enabled = true;
>>>
>>>   ... at this point imx_thermal is powered down for some amount of time,
>>>   over temperature IRQ will not be triggered for some amount of time.
>>>
>>> - if some part after thermal_zone_device_register() would fail or
>>>   deferred, the worker polling temperature will run in to NULL pointer.
>>>   This issue already happened...
>>>
>>> After migrating to runtime PM, one of issues started to be visible even
>>> on normal conditions.
>>> I'll send one more patch with reworking probe sequence.
>>
>> Are you planning to send a v3 with this patch? Or a separate patch?
> 
> I'm OK with both variants. What do you prefer?
> 
> I'll do i on top of PM patch to reduce refactoring overhead, if you OK
> about it.

I prefer you resend a couple of patches but change the subject of this
patch to something like "thermal/drivers/imx: Fix disabled sensor after
handling trip temperature" in order to reflect the problem, not the
solution.

btw: nice fix
Oleksij Rempel Oct. 21, 2021, 5:20 p.m. UTC | #7
Hi Petr,

On Wed, Oct 20, 2021 at 05:53:03PM +0200, Petr Benes wrote:
> On Wed, 20 Oct 2021 at 07:05, Oleksij Rempel <o.rempel@pengutronix.de> wrote:
> >
> > Hi Petr and Michal,
> >
> > I forgot to add you for v2 in CC. Please test/review this version.
> 
> Hi Oleksij,
> 
> It works good. with PM as well as without PM. The only minor issue I found is,
> that the first temperature reading (when the driver probes) fails. That is
> (val & soc_data->temp_valid_mask) == 0) holds true. How does
> pm_runtime_resume_and_get() behave in imx_thermal_probe()?
> Does it go through imx_thermal_runtime_resume() with usleep_range()?

How exactly did you reproduce it? Even more or less understanding how
this can potentially happen, i never had this issue on my HW. Is it something
HW specific?

Even after executing this command:
echo disabled > /sys/class/thermal/thermal_zone0/mode
cat /sys/class/thermal/thermal_zone0/temp

In this case, IRQ is disabled and on each manual temp read, driver starts
actively using runtime PM. I still never get (val & soc_data->temp_valid_mask)
== 0)..

> >
> > On Tue, Oct 19, 2021 at 03:08:09PM +0200, Oleksij Rempel wrote:
> > > Starting with commit d92ed2c9d3ff ("thermal: imx: Use driver's local
> > > data to decide whether to run a measurement") this driver stared using
> > > irq_enabled flag to make decision to power on/off the thermal core. This
> > > triggered a regression, where after reaching critical temperature, alarm
> > > IRQ handler set irq_enabled to false,  disabled thermal core and was not
> > > able read temperature and disable cooling sequence.
> > >
> > > In case the cooling device is "CPU/GPU freq", the system will run with
> > > reduce performance until next reboot.
> > >
> > > To solve this issue, we need to move all parts implementing hand made
> > > runtime power management and let it handle actual runtime PM framework.
> > >
> > > Fixes: d92ed2c9d3ff ("thermal: imx: Use driver's local data to decide whether to run a measurement")
> > > Signed-off-by: Oleksij Rempel <o.rempel@pengutronix.de>
> > > ---
> > >  drivers/thermal/imx_thermal.c | 143 +++++++++++++++++++++-------------
> > >  1 file changed, 89 insertions(+), 54 deletions(-)
> > >
> > > diff --git a/drivers/thermal/imx_thermal.c b/drivers/thermal/imx_thermal.c
> > > index 2c7473d86a59..cb5a4354fc75 100644
> > > --- a/drivers/thermal/imx_thermal.c
> > > +++ b/drivers/thermal/imx_thermal.c
> > > @@ -15,6 +15,7 @@
> > >  #include <linux/regmap.h>
> > >  #include <linux/thermal.h>
> > >  #include <linux/nvmem-consumer.h>
> > > +#include <linux/pm_runtime.h>
> > >
> > >  #define REG_SET              0x4
> > >  #define REG_CLR              0x8
> > > @@ -194,6 +195,7 @@ static struct thermal_soc_data thermal_imx7d_data = {
> > >  };
> > >
> > >  struct imx_thermal_data {
> > > +     struct device *dev;
> > >       struct cpufreq_policy *policy;
> > >       struct thermal_zone_device *tz;
> > >       struct thermal_cooling_device *cdev;
> > > @@ -252,44 +254,15 @@ static int imx_get_temp(struct thermal_zone_device *tz, int *temp)
> > >       const struct thermal_soc_data *soc_data = data->socdata;
> > >       struct regmap *map = data->tempmon;
> > >       unsigned int n_meas;
> > > -     bool wait, run_measurement;
> > >       u32 val;
> > > +     int ret;
> > >
> > > -     run_measurement = !data->irq_enabled;
> > > -     if (!run_measurement) {
> > > -             /* Check if a measurement is currently in progress */
> > > -             regmap_read(map, soc_data->temp_data, &val);
> > > -             wait = !(val & soc_data->temp_valid_mask);
> > > -     } else {
> > > -             /*
> > > -              * Every time we measure the temperature, we will power on the
> > > -              * temperature sensor, enable measurements, take a reading,
> > > -              * disable measurements, power off the temperature sensor.
> > > -              */
> > > -             regmap_write(map, soc_data->sensor_ctrl + REG_CLR,
> > > -                         soc_data->power_down_mask);
> > > -             regmap_write(map, soc_data->sensor_ctrl + REG_SET,
> > > -                         soc_data->measure_temp_mask);
> > > -
> > > -             wait = true;
> > > -     }
> > > -
> > > -     /*
> > > -      * According to the temp sensor designers, it may require up to ~17us
> > > -      * to complete a measurement.
> > > -      */
> > > -     if (wait)
> > > -             usleep_range(20, 50);
> > > +     ret = pm_runtime_resume_and_get(data->dev);
> > > +     if (ret < 0)
> > > +             return ret;
> > >
> > >       regmap_read(map, soc_data->temp_data, &val);
> > >
> > > -     if (run_measurement) {
> > > -             regmap_write(map, soc_data->sensor_ctrl + REG_CLR,
> > > -                          soc_data->measure_temp_mask);
> > > -             regmap_write(map, soc_data->sensor_ctrl + REG_SET,
> > > -                          soc_data->power_down_mask);
> > > -     }
> > > -
> > >       if ((val & soc_data->temp_valid_mask) == 0) {
> > >               dev_dbg(&tz->device, "temp measurement never finished\n");
> > >               return -EAGAIN;
> > > @@ -328,6 +301,8 @@ static int imx_get_temp(struct thermal_zone_device *tz, int *temp)
> > >               enable_irq(data->irq);
> > >       }
> > >
> > > +     pm_runtime_put(data->dev);
> > > +
> > >       return 0;
> > >  }
> > >
> > > @@ -335,24 +310,16 @@ static int imx_change_mode(struct thermal_zone_device *tz,
> > >                          enum thermal_device_mode mode)
> > >  {
> > >       struct imx_thermal_data *data = tz->devdata;
> > > -     struct regmap *map = data->tempmon;
> > > -     const struct thermal_soc_data *soc_data = data->socdata;
> > >
> > >       if (mode == THERMAL_DEVICE_ENABLED) {
> > > -             regmap_write(map, soc_data->sensor_ctrl + REG_CLR,
> > > -                          soc_data->power_down_mask);
> > > -             regmap_write(map, soc_data->sensor_ctrl + REG_SET,
> > > -                          soc_data->measure_temp_mask);
> > > +             pm_runtime_get(data->dev);
> > >
> > >               if (!data->irq_enabled) {
> > >                       data->irq_enabled = true;
> > >                       enable_irq(data->irq);
> > >               }
> > >       } else {
> > > -             regmap_write(map, soc_data->sensor_ctrl + REG_CLR,
> > > -                          soc_data->measure_temp_mask);
> > > -             regmap_write(map, soc_data->sensor_ctrl + REG_SET,
> > > -                          soc_data->power_down_mask);
> > > +             pm_runtime_put(data->dev);
> > >
> > >               if (data->irq_enabled) {
> > >                       disable_irq(data->irq);
> > > @@ -393,6 +360,11 @@ static int imx_set_trip_temp(struct thermal_zone_device *tz, int trip,
> > >                            int temp)
> > >  {
> > >       struct imx_thermal_data *data = tz->devdata;
> > > +     int ret;
> > > +
> > > +     ret = pm_runtime_resume_and_get(data->dev);
> > > +     if (ret < 0)
> > > +             return ret;
> > >
> > >       /* do not allow changing critical threshold */
> > >       if (trip == IMX_TRIP_CRITICAL)
> > > @@ -406,6 +378,8 @@ static int imx_set_trip_temp(struct thermal_zone_device *tz, int trip,
> > >
> > >       imx_set_alarm_temp(data, temp);
> > >
> > > +     pm_runtime_put(data->dev);
> > > +
> > >       return 0;
> > >  }
> > >
> > > @@ -681,6 +655,8 @@ static int imx_thermal_probe(struct platform_device *pdev)
> > >       if (!data)
> > >               return -ENOMEM;
> > >
> > > +     data->dev = &pdev->dev;
> > > +
> > >       map = syscon_regmap_lookup_by_phandle(pdev->dev.of_node, "fsl,tempmon");
> > >       if (IS_ERR(map)) {
> > >               ret = PTR_ERR(map);
> > > @@ -801,6 +777,14 @@ static int imx_thermal_probe(struct platform_device *pdev)
> > >       regmap_write(map, data->socdata->sensor_ctrl + REG_SET,
> > >                    data->socdata->measure_temp_mask);
> > >
> > > +     /* the core was configured and enabled just before */
> > > +     pm_runtime_set_active(&pdev->dev);
> > > +     pm_runtime_enable(data->dev);
> > > +
> > > +     ret = pm_runtime_resume_and_get(data->dev);
> > > +     if (ret < 0)
> > > +             goto disable_runtime_pm;
> > > +
> > >       data->irq_enabled = true;
> > >       ret = thermal_zone_device_enable(data->tz);
> > >       if (ret)
> > > @@ -814,10 +798,15 @@ static int imx_thermal_probe(struct platform_device *pdev)
> > >               goto thermal_zone_unregister;
> > >       }
> > >
> > > +     pm_runtime_put(data->dev);
> > > +
> > >       return 0;
> > >
> > >  thermal_zone_unregister:
> > >       thermal_zone_device_unregister(data->tz);
> > > +disable_runtime_pm:
> > > +     pm_runtime_put_noidle(data->dev);
> > > +     pm_runtime_disable(data->dev);
> > >  clk_disable:
> > >       clk_disable_unprepare(data->thermal_clk);
> > >  legacy_cleanup:
> > > @@ -829,13 +818,9 @@ static int imx_thermal_probe(struct platform_device *pdev)
> > >  static int imx_thermal_remove(struct platform_device *pdev)
> > >  {
> > >       struct imx_thermal_data *data = platform_get_drvdata(pdev);
> > > -     struct regmap *map = data->tempmon;
> > >
> > > -     /* Disable measurements */
> > > -     regmap_write(map, data->socdata->sensor_ctrl + REG_SET,
> > > -                  data->socdata->power_down_mask);
> > > -     if (!IS_ERR(data->thermal_clk))
> > > -             clk_disable_unprepare(data->thermal_clk);
> > > +     pm_runtime_put_noidle(data->dev);
> > > +     pm_runtime_disable(data->dev);
> > >
> > >       thermal_zone_device_unregister(data->tz);
> > >       imx_thermal_unregister_legacy_cooling(data);
> > > @@ -858,29 +843,79 @@ static int __maybe_unused imx_thermal_suspend(struct device *dev)
> > >       ret = thermal_zone_device_disable(data->tz);
> > >       if (ret)
> > >               return ret;
> > > +
> > > +     return pm_runtime_force_suspend(data->dev);
> > > +}
> > > +
> > > +static int __maybe_unused imx_thermal_resume(struct device *dev)
> > > +{
> > > +     struct imx_thermal_data *data = dev_get_drvdata(dev);
> > > +     int ret;
> > > +
> > > +     ret = pm_runtime_force_resume(data->dev);
> > > +     if (ret)
> > > +             return ret;
> > > +     /* Enabled thermal sensor after resume */
> > > +     return thermal_zone_device_enable(data->tz);
> > > +}
> > > +
> > > +static int __maybe_unused imx_thermal_runtime_suspend(struct device *dev)
> > > +{
> > > +     struct imx_thermal_data *data = dev_get_drvdata(dev);
> > > +     const struct thermal_soc_data *socdata = data->socdata;
> > > +     struct regmap *map = data->tempmon;
> > > +     int ret;
> > > +
> > > +     ret = regmap_write(map, socdata->sensor_ctrl + REG_CLR,
> > > +                        socdata->measure_temp_mask);
> > > +     if (ret)
> > > +             return ret;
> > > +
> > > +     ret = regmap_write(map, socdata->sensor_ctrl + REG_SET,
> > > +                        socdata->power_down_mask);
> > > +     if (ret)
> > > +             return ret;
> > > +
> > >       clk_disable_unprepare(data->thermal_clk);
> > >
> > >       return 0;
> > >  }
> > >
> > > -static int __maybe_unused imx_thermal_resume(struct device *dev)
> > > +static int __maybe_unused imx_thermal_runtime_resume(struct device *dev)
> > >  {
> > >       struct imx_thermal_data *data = dev_get_drvdata(dev);
> > > +     const struct thermal_soc_data *socdata = data->socdata;
> > > +     struct regmap *map = data->tempmon;
> > >       int ret;
> > >
> > >       ret = clk_prepare_enable(data->thermal_clk);
> > >       if (ret)
> > >               return ret;
> > > -     /* Enabled thermal sensor after resume */
> > > -     ret = thermal_zone_device_enable(data->tz);
> > > +
> > > +     ret = regmap_write(map, socdata->sensor_ctrl + REG_CLR,
> > > +                        socdata->power_down_mask);
> > > +     if (ret)
> > > +             return ret;
> > > +
> > > +     ret = regmap_write(map, socdata->sensor_ctrl + REG_SET,
> > > +                        socdata->measure_temp_mask);
> > >       if (ret)
> > >               return ret;
> > >
> > > +     /*
> > > +      * According to the temp sensor designers, it may require up to ~17us
> > > +      * to complete a measurement.
> > > +      */
> > > +     usleep_range(20, 50);
> > > +
> > >       return 0;
> > >  }
> > >
> > > -static SIMPLE_DEV_PM_OPS(imx_thermal_pm_ops,
> > > -                      imx_thermal_suspend, imx_thermal_resume);
> > > +static const struct dev_pm_ops imx_thermal_pm_ops = {
> > > +     SET_SYSTEM_SLEEP_PM_OPS(imx_thermal_suspend, imx_thermal_resume)
> > > +     SET_RUNTIME_PM_OPS(imx_thermal_runtime_suspend,
> > > +                        imx_thermal_runtime_resume, NULL)
> > > +};
> > >
> > >  static struct platform_driver imx_thermal = {
> > >       .driver = {
> > > --
> > > 2.30.2
> > >
> > >
> >
> > --
> > Pengutronix e.K.                           |                             |
> > Steuerwalder Str. 21                       | http://www.pengutronix.de/  |
> > 31137 Hildesheim, Germany                  | Phone: +49-5121-206917-0    |
> > Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |
> 
>
Petr Benes Oct. 25, 2021, 11:06 a.m. UTC | #8
Hi Oleksij,

On Thu, 21 Oct 2021 at 19:21, Oleksij Rempel <o.rempel@pengutronix.de> wrote:
>
> Hi Petr,
>
> On Wed, Oct 20, 2021 at 05:53:03PM +0200, Petr Benes wrote:
> > On Wed, 20 Oct 2021 at 07:05, Oleksij Rempel <o.rempel@pengutronix.de> wrote:
> > >
> > > Hi Petr and Michal,
> > >
> > > I forgot to add you for v2 in CC. Please test/review this version.
> >
> > Hi Oleksij,
> >
> > It works good. with PM as well as without PM. The only minor issue I found is,
> > that the first temperature reading (when the driver probes) fails. That is
> > (val & soc_data->temp_valid_mask) == 0) holds true. How does
> > pm_runtime_resume_and_get() behave in imx_thermal_probe()?
> > Does it go through imx_thermal_runtime_resume() with usleep_range()?
>
> How exactly did you reproduce it? Even more or less understanding how

I just placed my debug print into get_temp()

    if ((val & soc_data->temp_valid_mask) == 0) {
        dev_dbg(&tz->device, "temp measurement never finished\n");
        printk("Wrong temperature reading!!!!!!\n");
        return -EAGAIN;
    }

> this can potentially happen, i never had this issue on my HW. Is it something
> HW specific?

IMHO it is just product of the following sequence:

pm_runtime_set_active(&pdev->dev);
pm_runtime_enable(data->dev);
pm_runtime_resume_and_get(data->dev);
thermal_zone_device_enable(data->tz);

With assumption imx_thermal_runtime_resume() didn't run,
hence the sensor didn't get enough time to come up.

I didn't have time to spend it on and you have better knowledge of the
area. If it is not that straightforward I can try to diagnose it better.


>
> Even after executing this command:
> echo disabled > /sys/class/thermal/thermal_zone0/mode
> cat /sys/class/thermal/thermal_zone0/temp
>
> In this case, IRQ is disabled and on each manual temp read, driver starts
> actively using runtime PM. I still never get (val & soc_data->temp_valid_mask)
> == 0)..
>
> > >
> > > On Tue, Oct 19, 2021 at 03:08:09PM +0200, Oleksij Rempel wrote:
> > > > Starting with commit d92ed2c9d3ff ("thermal: imx: Use driver's local
> > > > data to decide whether to run a measurement") this driver stared using
> > > > irq_enabled flag to make decision to power on/off the thermal core. This
> > > > triggered a regression, where after reaching critical temperature, alarm
> > > > IRQ handler set irq_enabled to false,  disabled thermal core and was not
> > > > able read temperature and disable cooling sequence.
> > > >
> > > > In case the cooling device is "CPU/GPU freq", the system will run with
> > > > reduce performance until next reboot.
> > > >
> > > > To solve this issue, we need to move all parts implementing hand made
> > > > runtime power management and let it handle actual runtime PM framework.
> > > >
> > > > Fixes: d92ed2c9d3ff ("thermal: imx: Use driver's local data to decide whether to run a measurement")
> > > > Signed-off-by: Oleksij Rempel <o.rempel@pengutronix.de>
> > > > ---
> > > >  drivers/thermal/imx_thermal.c | 143 +++++++++++++++++++++-------------
> > > >  1 file changed, 89 insertions(+), 54 deletions(-)
> > > >
> > > > diff --git a/drivers/thermal/imx_thermal.c b/drivers/thermal/imx_thermal.c
> > > > index 2c7473d86a59..cb5a4354fc75 100644
> > > > --- a/drivers/thermal/imx_thermal.c
> > > > +++ b/drivers/thermal/imx_thermal.c
> > > > @@ -15,6 +15,7 @@
> > > >  #include <linux/regmap.h>
> > > >  #include <linux/thermal.h>
> > > >  #include <linux/nvmem-consumer.h>
> > > > +#include <linux/pm_runtime.h>
> > > >
> > > >  #define REG_SET              0x4
> > > >  #define REG_CLR              0x8
> > > > @@ -194,6 +195,7 @@ static struct thermal_soc_data thermal_imx7d_data = {
> > > >  };
> > > >
> > > >  struct imx_thermal_data {
> > > > +     struct device *dev;
> > > >       struct cpufreq_policy *policy;
> > > >       struct thermal_zone_device *tz;
> > > >       struct thermal_cooling_device *cdev;
> > > > @@ -252,44 +254,15 @@ static int imx_get_temp(struct thermal_zone_device *tz, int *temp)
> > > >       const struct thermal_soc_data *soc_data = data->socdata;
> > > >       struct regmap *map = data->tempmon;
> > > >       unsigned int n_meas;
> > > > -     bool wait, run_measurement;
> > > >       u32 val;
> > > > +     int ret;
> > > >
> > > > -     run_measurement = !data->irq_enabled;
> > > > -     if (!run_measurement) {
> > > > -             /* Check if a measurement is currently in progress */
> > > > -             regmap_read(map, soc_data->temp_data, &val);
> > > > -             wait = !(val & soc_data->temp_valid_mask);
> > > > -     } else {
> > > > -             /*
> > > > -              * Every time we measure the temperature, we will power on the
> > > > -              * temperature sensor, enable measurements, take a reading,
> > > > -              * disable measurements, power off the temperature sensor.
> > > > -              */
> > > > -             regmap_write(map, soc_data->sensor_ctrl + REG_CLR,
> > > > -                         soc_data->power_down_mask);
> > > > -             regmap_write(map, soc_data->sensor_ctrl + REG_SET,
> > > > -                         soc_data->measure_temp_mask);
> > > > -
> > > > -             wait = true;
> > > > -     }
> > > > -
> > > > -     /*
> > > > -      * According to the temp sensor designers, it may require up to ~17us
> > > > -      * to complete a measurement.
> > > > -      */
> > > > -     if (wait)
> > > > -             usleep_range(20, 50);
> > > > +     ret = pm_runtime_resume_and_get(data->dev);
> > > > +     if (ret < 0)
> > > > +             return ret;
> > > >
> > > >       regmap_read(map, soc_data->temp_data, &val);
> > > >
> > > > -     if (run_measurement) {
> > > > -             regmap_write(map, soc_data->sensor_ctrl + REG_CLR,
> > > > -                          soc_data->measure_temp_mask);
> > > > -             regmap_write(map, soc_data->sensor_ctrl + REG_SET,
> > > > -                          soc_data->power_down_mask);
> > > > -     }
> > > > -
> > > >       if ((val & soc_data->temp_valid_mask) == 0) {
> > > >               dev_dbg(&tz->device, "temp measurement never finished\n");
> > > >               return -EAGAIN;
> > > > @@ -328,6 +301,8 @@ static int imx_get_temp(struct thermal_zone_device *tz, int *temp)
> > > >               enable_irq(data->irq);
> > > >       }
> > > >
> > > > +     pm_runtime_put(data->dev);
> > > > +
> > > >       return 0;
> > > >  }
> > > >
> > > > @@ -335,24 +310,16 @@ static int imx_change_mode(struct thermal_zone_device *tz,
> > > >                          enum thermal_device_mode mode)
> > > >  {
> > > >       struct imx_thermal_data *data = tz->devdata;
> > > > -     struct regmap *map = data->tempmon;
> > > > -     const struct thermal_soc_data *soc_data = data->socdata;
> > > >
> > > >       if (mode == THERMAL_DEVICE_ENABLED) {
> > > > -             regmap_write(map, soc_data->sensor_ctrl + REG_CLR,
> > > > -                          soc_data->power_down_mask);
> > > > -             regmap_write(map, soc_data->sensor_ctrl + REG_SET,
> > > > -                          soc_data->measure_temp_mask);
> > > > +             pm_runtime_get(data->dev);
> > > >
> > > >               if (!data->irq_enabled) {
> > > >                       data->irq_enabled = true;
> > > >                       enable_irq(data->irq);
> > > >               }
> > > >       } else {
> > > > -             regmap_write(map, soc_data->sensor_ctrl + REG_CLR,
> > > > -                          soc_data->measure_temp_mask);
> > > > -             regmap_write(map, soc_data->sensor_ctrl + REG_SET,
> > > > -                          soc_data->power_down_mask);
> > > > +             pm_runtime_put(data->dev);
> > > >
> > > >               if (data->irq_enabled) {
> > > >                       disable_irq(data->irq);
> > > > @@ -393,6 +360,11 @@ static int imx_set_trip_temp(struct thermal_zone_device *tz, int trip,
> > > >                            int temp)
> > > >  {
> > > >       struct imx_thermal_data *data = tz->devdata;
> > > > +     int ret;
> > > > +
> > > > +     ret = pm_runtime_resume_and_get(data->dev);
> > > > +     if (ret < 0)
> > > > +             return ret;
> > > >
> > > >       /* do not allow changing critical threshold */
> > > >       if (trip == IMX_TRIP_CRITICAL)
> > > > @@ -406,6 +378,8 @@ static int imx_set_trip_temp(struct thermal_zone_device *tz, int trip,
> > > >
> > > >       imx_set_alarm_temp(data, temp);
> > > >
> > > > +     pm_runtime_put(data->dev);
> > > > +
> > > >       return 0;
> > > >  }
> > > >
> > > > @@ -681,6 +655,8 @@ static int imx_thermal_probe(struct platform_device *pdev)
> > > >       if (!data)
> > > >               return -ENOMEM;
> > > >
> > > > +     data->dev = &pdev->dev;
> > > > +
> > > >       map = syscon_regmap_lookup_by_phandle(pdev->dev.of_node, "fsl,tempmon");
> > > >       if (IS_ERR(map)) {
> > > >               ret = PTR_ERR(map);
> > > > @@ -801,6 +777,14 @@ static int imx_thermal_probe(struct platform_device *pdev)
> > > >       regmap_write(map, data->socdata->sensor_ctrl + REG_SET,
> > > >                    data->socdata->measure_temp_mask);
> > > >
> > > > +     /* the core was configured and enabled just before */
> > > > +     pm_runtime_set_active(&pdev->dev);
> > > > +     pm_runtime_enable(data->dev);
> > > > +
> > > > +     ret = pm_runtime_resume_and_get(data->dev);
> > > > +     if (ret < 0)
> > > > +             goto disable_runtime_pm;
> > > > +
> > > >       data->irq_enabled = true;
> > > >       ret = thermal_zone_device_enable(data->tz);
> > > >       if (ret)
> > > > @@ -814,10 +798,15 @@ static int imx_thermal_probe(struct platform_device *pdev)
> > > >               goto thermal_zone_unregister;
> > > >       }
> > > >
> > > > +     pm_runtime_put(data->dev);
> > > > +
> > > >       return 0;
> > > >
> > > >  thermal_zone_unregister:
> > > >       thermal_zone_device_unregister(data->tz);
> > > > +disable_runtime_pm:
> > > > +     pm_runtime_put_noidle(data->dev);
> > > > +     pm_runtime_disable(data->dev);
> > > >  clk_disable:
> > > >       clk_disable_unprepare(data->thermal_clk);
> > > >  legacy_cleanup:
> > > > @@ -829,13 +818,9 @@ static int imx_thermal_probe(struct platform_device *pdev)
> > > >  static int imx_thermal_remove(struct platform_device *pdev)
> > > >  {
> > > >       struct imx_thermal_data *data = platform_get_drvdata(pdev);
> > > > -     struct regmap *map = data->tempmon;
> > > >
> > > > -     /* Disable measurements */
> > > > -     regmap_write(map, data->socdata->sensor_ctrl + REG_SET,
> > > > -                  data->socdata->power_down_mask);
> > > > -     if (!IS_ERR(data->thermal_clk))
> > > > -             clk_disable_unprepare(data->thermal_clk);
> > > > +     pm_runtime_put_noidle(data->dev);
> > > > +     pm_runtime_disable(data->dev);
> > > >
> > > >       thermal_zone_device_unregister(data->tz);
> > > >       imx_thermal_unregister_legacy_cooling(data);
> > > > @@ -858,29 +843,79 @@ static int __maybe_unused imx_thermal_suspend(struct device *dev)
> > > >       ret = thermal_zone_device_disable(data->tz);
> > > >       if (ret)
> > > >               return ret;
> > > > +
> > > > +     return pm_runtime_force_suspend(data->dev);
> > > > +}
> > > > +
> > > > +static int __maybe_unused imx_thermal_resume(struct device *dev)
> > > > +{
> > > > +     struct imx_thermal_data *data = dev_get_drvdata(dev);
> > > > +     int ret;
> > > > +
> > > > +     ret = pm_runtime_force_resume(data->dev);
> > > > +     if (ret)
> > > > +             return ret;
> > > > +     /* Enabled thermal sensor after resume */
> > > > +     return thermal_zone_device_enable(data->tz);
> > > > +}
> > > > +
> > > > +static int __maybe_unused imx_thermal_runtime_suspend(struct device *dev)
> > > > +{
> > > > +     struct imx_thermal_data *data = dev_get_drvdata(dev);
> > > > +     const struct thermal_soc_data *socdata = data->socdata;
> > > > +     struct regmap *map = data->tempmon;
> > > > +     int ret;
> > > > +
> > > > +     ret = regmap_write(map, socdata->sensor_ctrl + REG_CLR,
> > > > +                        socdata->measure_temp_mask);
> > > > +     if (ret)
> > > > +             return ret;
> > > > +
> > > > +     ret = regmap_write(map, socdata->sensor_ctrl + REG_SET,
> > > > +                        socdata->power_down_mask);
> > > > +     if (ret)
> > > > +             return ret;
> > > > +
> > > >       clk_disable_unprepare(data->thermal_clk);
> > > >
> > > >       return 0;
> > > >  }
> > > >
> > > > -static int __maybe_unused imx_thermal_resume(struct device *dev)
> > > > +static int __maybe_unused imx_thermal_runtime_resume(struct device *dev)
> > > >  {
> > > >       struct imx_thermal_data *data = dev_get_drvdata(dev);
> > > > +     const struct thermal_soc_data *socdata = data->socdata;
> > > > +     struct regmap *map = data->tempmon;
> > > >       int ret;
> > > >
> > > >       ret = clk_prepare_enable(data->thermal_clk);
> > > >       if (ret)
> > > >               return ret;
> > > > -     /* Enabled thermal sensor after resume */
> > > > -     ret = thermal_zone_device_enable(data->tz);
> > > > +
> > > > +     ret = regmap_write(map, socdata->sensor_ctrl + REG_CLR,
> > > > +                        socdata->power_down_mask);
> > > > +     if (ret)
> > > > +             return ret;
> > > > +
> > > > +     ret = regmap_write(map, socdata->sensor_ctrl + REG_SET,
> > > > +                        socdata->measure_temp_mask);
> > > >       if (ret)
> > > >               return ret;
> > > >
> > > > +     /*
> > > > +      * According to the temp sensor designers, it may require up to ~17us
> > > > +      * to complete a measurement.
> > > > +      */
> > > > +     usleep_range(20, 50);
> > > > +
> > > >       return 0;
> > > >  }
> > > >
> > > > -static SIMPLE_DEV_PM_OPS(imx_thermal_pm_ops,
> > > > -                      imx_thermal_suspend, imx_thermal_resume);
> > > > +static const struct dev_pm_ops imx_thermal_pm_ops = {
> > > > +     SET_SYSTEM_SLEEP_PM_OPS(imx_thermal_suspend, imx_thermal_resume)
> > > > +     SET_RUNTIME_PM_OPS(imx_thermal_runtime_suspend,
> > > > +                        imx_thermal_runtime_resume, NULL)
> > > > +};
> > > >
> > > >  static struct platform_driver imx_thermal = {
> > > >       .driver = {
> > > > --
> > > > 2.30.2
> > > >
> > > >
> > >
> > > --
> > > Pengutronix e.K.                           |                             |
> > > Steuerwalder Str. 21                       | http://www.pengutronix.de/  |
> > > 31137 Hildesheim, Germany                  | Phone: +49-5121-206917-0    |
> > > Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |
> >
> >
>
> --
> Pengutronix e.K.                           |                             |
> Steuerwalder Str. 21                       | http://www.pengutronix.de/  |
> 31137 Hildesheim, Germany                  | Phone: +49-5121-206917-0    |
> Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |
Michal Vokáč Nov. 10, 2021, 10:07 a.m. UTC | #9
On 25. 10. 21 13:06, Petr Benes wrote:
> Hi Oleksij,
> 
> On Thu, 21 Oct 2021 at 19:21, Oleksij Rempel <o.rempel@pengutronix.de> wrote:
>>
>> Hi Petr,
>>
>> On Wed, Oct 20, 2021 at 05:53:03PM +0200, Petr Benes wrote:
>>> On Wed, 20 Oct 2021 at 07:05, Oleksij Rempel <o.rempel@pengutronix.de> wrote:
>>>>
>>>> Hi Petr and Michal,
>>>>
>>>> I forgot to add you for v2 in CC. Please test/review this version.
>>>
>>> Hi Oleksij,
>>>
>>> It works good. with PM as well as without PM. The only minor issue I found is,
>>> that the first temperature reading (when the driver probes) fails. That is
>>> (val & soc_data->temp_valid_mask) == 0) holds true. How does
>>> pm_runtime_resume_and_get() behave in imx_thermal_probe()?
>>> Does it go through imx_thermal_runtime_resume() with usleep_range()?
>>
>> How exactly did you reproduce it? Even more or less understanding how
> 
> I just placed my debug print into get_temp()
> 
>      if ((val & soc_data->temp_valid_mask) == 0) {
>          dev_dbg(&tz->device, "temp measurement never finished\n");
>          printk("Wrong temperature reading!!!!!!\n");
>          return -EAGAIN;
>      }
> 
>> this can potentially happen, i never had this issue on my HW. Is it something
>> HW specific?
> 
> IMHO it is just product of the following sequence:
> 
> pm_runtime_set_active(&pdev->dev);
> pm_runtime_enable(data->dev);
> pm_runtime_resume_and_get(data->dev);
> thermal_zone_device_enable(data->tz);
> 
> With assumption imx_thermal_runtime_resume() didn't run,
> hence the sensor didn't get enough time to come up.
> 
> I didn't have time to spend it on and you have better knowledge of the
> area. If it is not that straightforward I can try to diagnose it better.
>
Hi Oleksij,
Did you manage to further debug and reproduce this problem?
Do you plan to send the v3?

Regarding your question about the HW - this problem occured once we
upgraded the SoC on our SBC from i.MX6DL to i.MX6Q/QP. With the DualLite
we never had this problem but the Quad is getting hot quite fast.
We have pretty limited cooling options so the core is operated at its
upper temperature limits when fully loaded.

Best regards,
Michal
Oleksij Rempel Nov. 11, 2021, 9:16 a.m. UTC | #10
On Wed, Nov 10, 2021 at 11:07:31AM +0100, Michal Vokáč wrote:
> On 25. 10. 21 13:06, Petr Benes wrote:
> > Hi Oleksij,
> > 
> > On Thu, 21 Oct 2021 at 19:21, Oleksij Rempel <o.rempel@pengutronix.de> wrote:
> > > 
> > > Hi Petr,
> > > 
> > > On Wed, Oct 20, 2021 at 05:53:03PM +0200, Petr Benes wrote:
> > > > On Wed, 20 Oct 2021 at 07:05, Oleksij Rempel <o.rempel@pengutronix.de> wrote:
> > > > > 
> > > > > Hi Petr and Michal,
> > > > > 
> > > > > I forgot to add you for v2 in CC. Please test/review this version.
> > > > 
> > > > Hi Oleksij,
> > > > 
> > > > It works good. with PM as well as without PM. The only minor issue I found is,
> > > > that the first temperature reading (when the driver probes) fails. That is
> > > > (val & soc_data->temp_valid_mask) == 0) holds true. How does
> > > > pm_runtime_resume_and_get() behave in imx_thermal_probe()?
> > > > Does it go through imx_thermal_runtime_resume() with usleep_range()?
> > > 
> > > How exactly did you reproduce it? Even more or less understanding how
> > 
> > I just placed my debug print into get_temp()
> > 
> >      if ((val & soc_data->temp_valid_mask) == 0) {
> >          dev_dbg(&tz->device, "temp measurement never finished\n");
> >          printk("Wrong temperature reading!!!!!!\n");
> >          return -EAGAIN;
> >      }
> > 
> > > this can potentially happen, i never had this issue on my HW. Is it something
> > > HW specific?
> > 
> > IMHO it is just product of the following sequence:
> > 
> > pm_runtime_set_active(&pdev->dev);
> > pm_runtime_enable(data->dev);
> > pm_runtime_resume_and_get(data->dev);
> > thermal_zone_device_enable(data->tz);
> > 
> > With assumption imx_thermal_runtime_resume() didn't run,
> > hence the sensor didn't get enough time to come up.
> > 
> > I didn't have time to spend it on and you have better knowledge of the
> > area. If it is not that straightforward I can try to diagnose it better.
> > 
> Hi Oleksij,
> Did you manage to further debug and reproduce this problem?
> Do you plan to send the v3?
> 
> Regarding your question about the HW - this problem occured once we
> upgraded the SoC on our SBC from i.MX6DL to i.MX6Q/QP. With the DualLite
> we never had this problem but the Quad is getting hot quite fast.
> We have pretty limited cooling options so the core is operated at its
> upper temperature limits when fully loaded.

Hi Michal,

Sorry, I was busy and lost this topic from my radar. I was not able to
reproduce it on my i.MX6Q and i.MX6QP died after other thermal voltage
experiments. Please, if you able to reproduce it, try to investigate
what is wrong, for example increasing wakeup time and/or and tracing
sleap/wake/get sequences.

Regards,
Oleksij
Petr Benes Nov. 15, 2021, 3:02 p.m. UTC | #11
Hi Oleksij,

On Thu, 11 Nov 2021 at 10:16, Oleksij Rempel <o.rempel@pengutronix.de> wrote:
>
> On Wed, Nov 10, 2021 at 11:07:31AM +0100, Michal Vokáč wrote:
> > On 25. 10. 21 13:06, Petr Benes wrote:
> > > Hi Oleksij,
> > >
> > > On Thu, 21 Oct 2021 at 19:21, Oleksij Rempel <o.rempel@pengutronix.de> wrote:
> > > >
> > > > Hi Petr,
> > > >
> > > > On Wed, Oct 20, 2021 at 05:53:03PM +0200, Petr Benes wrote:
> > > > > On Wed, 20 Oct 2021 at 07:05, Oleksij Rempel <o.rempel@pengutronix.de> wrote:
> > > > > >
> > > > > > Hi Petr and Michal,
> > > > > >
> > > > > > I forgot to add you for v2 in CC. Please test/review this version.
> > > > >
> > > > > Hi Oleksij,
> > > > >
> > > > > It works good. with PM as well as without PM. The only minor issue I found is,
> > > > > that the first temperature reading (when the driver probes) fails. That is
> > > > > (val & soc_data->temp_valid_mask) == 0) holds true. How does
> > > > > pm_runtime_resume_and_get() behave in imx_thermal_probe()?
> > > > > Does it go through imx_thermal_runtime_resume() with usleep_range()?
> > > >
> > > > How exactly did you reproduce it? Even more or less understanding how
> > >
> > > I just placed my debug print into get_temp()
> > >
> > >      if ((val & soc_data->temp_valid_mask) == 0) {
> > >          dev_dbg(&tz->device, "temp measurement never finished\n");
> > >          printk("Wrong temperature reading!!!!!!\n");
> > >          return -EAGAIN;
> > >      }
> > >
> > > > this can potentially happen, i never had this issue on my HW. Is it something
> > > > HW specific?
> > >
> > > IMHO it is just product of the following sequence:
> > >
> > > pm_runtime_set_active(&pdev->dev);
> > > pm_runtime_enable(data->dev);
> > > pm_runtime_resume_and_get(data->dev);
> > > thermal_zone_device_enable(data->tz);
> > >
> > > With assumption imx_thermal_runtime_resume() didn't run,
> > > hence the sensor didn't get enough time to come up.
> > >
> > > I didn't have time to spend it on and you have better knowledge of the
> > > area. If it is not that straightforward I can try to diagnose it better.
> > >
> > Hi Oleksij,
> > Did you manage to further debug and reproduce this problem?
> > Do you plan to send the v3?
> >
> > Regarding your question about the HW - this problem occured once we
> > upgraded the SoC on our SBC from i.MX6DL to i.MX6Q/QP. With the DualLite
> > we never had this problem but the Quad is getting hot quite fast.
> > We have pretty limited cooling options so the core is operated at its
> > upper temperature limits when fully loaded.
>
> Hi Michal,
>
> Sorry, I was busy and lost this topic from my radar. I was not able to
> reproduce it on my i.MX6Q and i.MX6QP died after other thermal voltage
> experiments. Please, if you able to reproduce it, try to investigate
> what is wrong, for example increasing wakeup time and/or and tracing
> sleap/wake/get sequences.

Seems it is just as easy as calling usleep_range(20, 50) when you switch on
the sensor and enable temperature measurement in imx_thermal_probe().
So, we are sure the sensor is configured and _ready_.

You call pm_runtime_set_active(), pm_runtime_enable(), and
pm_runtime_resume_and_get(). The last one doesn't call the resume
callback (which correctly handles waiting for the sensor) as the device
is already active.

On some SoCs the timing leads to a failure of the temperature readout
in thermal_zone_device_enable() which follows. I've seen it on i.MX6DL,
but cannot reproduce it on i.MX6QP for example.

Regards,
Petr

>
> Regards,
> Oleksij
> --
> Pengutronix e.K.                           |                             |
> Steuerwalder Str. 21                       | http://www.pengutronix.de/  |
> 31137 Hildesheim, Germany                  | Phone: +49-5121-206917-0    |
> Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |
Oleksij Rempel Nov. 17, 2021, 10:24 a.m. UTC | #12
On Mon, Nov 15, 2021 at 04:02:07PM +0100, Petr Benes wrote:
> Hi Oleksij,
> 
> On Thu, 11 Nov 2021 at 10:16, Oleksij Rempel <o.rempel@pengutronix.de> wrote:
> >
> > On Wed, Nov 10, 2021 at 11:07:31AM +0100, Michal Vokáč wrote:
> > > On 25. 10. 21 13:06, Petr Benes wrote:
> > > > Hi Oleksij,
> > > >
> > > > On Thu, 21 Oct 2021 at 19:21, Oleksij Rempel <o.rempel@pengutronix.de> wrote:
> > > > >
> > > > > Hi Petr,
> > > > >
> > > > > On Wed, Oct 20, 2021 at 05:53:03PM +0200, Petr Benes wrote:
> > > > > > On Wed, 20 Oct 2021 at 07:05, Oleksij Rempel <o.rempel@pengutronix.de> wrote:
> > > > > > >
> > > > > > > Hi Petr and Michal,
> > > > > > >
> > > > > > > I forgot to add you for v2 in CC. Please test/review this version.
> > > > > >
> > > > > > Hi Oleksij,
> > > > > >
> > > > > > It works good. with PM as well as without PM. The only minor issue I found is,
> > > > > > that the first temperature reading (when the driver probes) fails. That is
> > > > > > (val & soc_data->temp_valid_mask) == 0) holds true. How does
> > > > > > pm_runtime_resume_and_get() behave in imx_thermal_probe()?
> > > > > > Does it go through imx_thermal_runtime_resume() with usleep_range()?
> > > > >
> > > > > How exactly did you reproduce it? Even more or less understanding how
> > > >
> > > > I just placed my debug print into get_temp()
> > > >
> > > >      if ((val & soc_data->temp_valid_mask) == 0) {
> > > >          dev_dbg(&tz->device, "temp measurement never finished\n");
> > > >          printk("Wrong temperature reading!!!!!!\n");
> > > >          return -EAGAIN;
> > > >      }
> > > >
> > > > > this can potentially happen, i never had this issue on my HW. Is it something
> > > > > HW specific?
> > > >
> > > > IMHO it is just product of the following sequence:
> > > >
> > > > pm_runtime_set_active(&pdev->dev);
> > > > pm_runtime_enable(data->dev);
> > > > pm_runtime_resume_and_get(data->dev);
> > > > thermal_zone_device_enable(data->tz);
> > > >
> > > > With assumption imx_thermal_runtime_resume() didn't run,
> > > > hence the sensor didn't get enough time to come up.
> > > >
> > > > I didn't have time to spend it on and you have better knowledge of the
> > > > area. If it is not that straightforward I can try to diagnose it better.
> > > >
> > > Hi Oleksij,
> > > Did you manage to further debug and reproduce this problem?
> > > Do you plan to send the v3?
> > >
> > > Regarding your question about the HW - this problem occured once we
> > > upgraded the SoC on our SBC from i.MX6DL to i.MX6Q/QP. With the DualLite
> > > we never had this problem but the Quad is getting hot quite fast.
> > > We have pretty limited cooling options so the core is operated at its
> > > upper temperature limits when fully loaded.
> >
> > Hi Michal,
> >
> > Sorry, I was busy and lost this topic from my radar. I was not able to
> > reproduce it on my i.MX6Q and i.MX6QP died after other thermal voltage
> > experiments. Please, if you able to reproduce it, try to investigate
> > what is wrong, for example increasing wakeup time and/or and tracing
> > sleap/wake/get sequences.
> 
> Seems it is just as easy as calling usleep_range(20, 50) when you switch on
> the sensor and enable temperature measurement in imx_thermal_probe().
> So, we are sure the sensor is configured and _ready_.
> 
> You call pm_runtime_set_active(), pm_runtime_enable(), and
> pm_runtime_resume_and_get(). The last one doesn't call the resume
> callback (which correctly handles waiting for the sensor) as the device
> is already active.

Ok, thx! It makes sense. I'll send new version with the fix.

Regards,
Oleksij
diff mbox series

Patch

diff --git a/drivers/thermal/imx_thermal.c b/drivers/thermal/imx_thermal.c
index 2c7473d86a59..cb5a4354fc75 100644
--- a/drivers/thermal/imx_thermal.c
+++ b/drivers/thermal/imx_thermal.c
@@ -15,6 +15,7 @@ 
 #include <linux/regmap.h>
 #include <linux/thermal.h>
 #include <linux/nvmem-consumer.h>
+#include <linux/pm_runtime.h>
 
 #define REG_SET		0x4
 #define REG_CLR		0x8
@@ -194,6 +195,7 @@  static struct thermal_soc_data thermal_imx7d_data = {
 };
 
 struct imx_thermal_data {
+	struct device *dev;
 	struct cpufreq_policy *policy;
 	struct thermal_zone_device *tz;
 	struct thermal_cooling_device *cdev;
@@ -252,44 +254,15 @@  static int imx_get_temp(struct thermal_zone_device *tz, int *temp)
 	const struct thermal_soc_data *soc_data = data->socdata;
 	struct regmap *map = data->tempmon;
 	unsigned int n_meas;
-	bool wait, run_measurement;
 	u32 val;
+	int ret;
 
-	run_measurement = !data->irq_enabled;
-	if (!run_measurement) {
-		/* Check if a measurement is currently in progress */
-		regmap_read(map, soc_data->temp_data, &val);
-		wait = !(val & soc_data->temp_valid_mask);
-	} else {
-		/*
-		 * Every time we measure the temperature, we will power on the
-		 * temperature sensor, enable measurements, take a reading,
-		 * disable measurements, power off the temperature sensor.
-		 */
-		regmap_write(map, soc_data->sensor_ctrl + REG_CLR,
-			    soc_data->power_down_mask);
-		regmap_write(map, soc_data->sensor_ctrl + REG_SET,
-			    soc_data->measure_temp_mask);
-
-		wait = true;
-	}
-
-	/*
-	 * According to the temp sensor designers, it may require up to ~17us
-	 * to complete a measurement.
-	 */
-	if (wait)
-		usleep_range(20, 50);
+	ret = pm_runtime_resume_and_get(data->dev);
+	if (ret < 0)
+		return ret;
 
 	regmap_read(map, soc_data->temp_data, &val);
 
-	if (run_measurement) {
-		regmap_write(map, soc_data->sensor_ctrl + REG_CLR,
-			     soc_data->measure_temp_mask);
-		regmap_write(map, soc_data->sensor_ctrl + REG_SET,
-			     soc_data->power_down_mask);
-	}
-
 	if ((val & soc_data->temp_valid_mask) == 0) {
 		dev_dbg(&tz->device, "temp measurement never finished\n");
 		return -EAGAIN;
@@ -328,6 +301,8 @@  static int imx_get_temp(struct thermal_zone_device *tz, int *temp)
 		enable_irq(data->irq);
 	}
 
+	pm_runtime_put(data->dev);
+
 	return 0;
 }
 
@@ -335,24 +310,16 @@  static int imx_change_mode(struct thermal_zone_device *tz,
 			   enum thermal_device_mode mode)
 {
 	struct imx_thermal_data *data = tz->devdata;
-	struct regmap *map = data->tempmon;
-	const struct thermal_soc_data *soc_data = data->socdata;
 
 	if (mode == THERMAL_DEVICE_ENABLED) {
-		regmap_write(map, soc_data->sensor_ctrl + REG_CLR,
-			     soc_data->power_down_mask);
-		regmap_write(map, soc_data->sensor_ctrl + REG_SET,
-			     soc_data->measure_temp_mask);
+		pm_runtime_get(data->dev);
 
 		if (!data->irq_enabled) {
 			data->irq_enabled = true;
 			enable_irq(data->irq);
 		}
 	} else {
-		regmap_write(map, soc_data->sensor_ctrl + REG_CLR,
-			     soc_data->measure_temp_mask);
-		regmap_write(map, soc_data->sensor_ctrl + REG_SET,
-			     soc_data->power_down_mask);
+		pm_runtime_put(data->dev);
 
 		if (data->irq_enabled) {
 			disable_irq(data->irq);
@@ -393,6 +360,11 @@  static int imx_set_trip_temp(struct thermal_zone_device *tz, int trip,
 			     int temp)
 {
 	struct imx_thermal_data *data = tz->devdata;
+	int ret;
+
+	ret = pm_runtime_resume_and_get(data->dev);
+	if (ret < 0)
+		return ret;
 
 	/* do not allow changing critical threshold */
 	if (trip == IMX_TRIP_CRITICAL)
@@ -406,6 +378,8 @@  static int imx_set_trip_temp(struct thermal_zone_device *tz, int trip,
 
 	imx_set_alarm_temp(data, temp);
 
+	pm_runtime_put(data->dev);
+
 	return 0;
 }
 
@@ -681,6 +655,8 @@  static int imx_thermal_probe(struct platform_device *pdev)
 	if (!data)
 		return -ENOMEM;
 
+	data->dev = &pdev->dev;
+
 	map = syscon_regmap_lookup_by_phandle(pdev->dev.of_node, "fsl,tempmon");
 	if (IS_ERR(map)) {
 		ret = PTR_ERR(map);
@@ -801,6 +777,14 @@  static int imx_thermal_probe(struct platform_device *pdev)
 	regmap_write(map, data->socdata->sensor_ctrl + REG_SET,
 		     data->socdata->measure_temp_mask);
 
+	/* the core was configured and enabled just before */
+	pm_runtime_set_active(&pdev->dev);
+	pm_runtime_enable(data->dev);
+
+	ret = pm_runtime_resume_and_get(data->dev);
+	if (ret < 0)
+		goto disable_runtime_pm;
+
 	data->irq_enabled = true;
 	ret = thermal_zone_device_enable(data->tz);
 	if (ret)
@@ -814,10 +798,15 @@  static int imx_thermal_probe(struct platform_device *pdev)
 		goto thermal_zone_unregister;
 	}
 
+	pm_runtime_put(data->dev);
+
 	return 0;
 
 thermal_zone_unregister:
 	thermal_zone_device_unregister(data->tz);
+disable_runtime_pm:
+	pm_runtime_put_noidle(data->dev);
+	pm_runtime_disable(data->dev);
 clk_disable:
 	clk_disable_unprepare(data->thermal_clk);
 legacy_cleanup:
@@ -829,13 +818,9 @@  static int imx_thermal_probe(struct platform_device *pdev)
 static int imx_thermal_remove(struct platform_device *pdev)
 {
 	struct imx_thermal_data *data = platform_get_drvdata(pdev);
-	struct regmap *map = data->tempmon;
 
-	/* Disable measurements */
-	regmap_write(map, data->socdata->sensor_ctrl + REG_SET,
-		     data->socdata->power_down_mask);
-	if (!IS_ERR(data->thermal_clk))
-		clk_disable_unprepare(data->thermal_clk);
+	pm_runtime_put_noidle(data->dev);
+	pm_runtime_disable(data->dev);
 
 	thermal_zone_device_unregister(data->tz);
 	imx_thermal_unregister_legacy_cooling(data);
@@ -858,29 +843,79 @@  static int __maybe_unused imx_thermal_suspend(struct device *dev)
 	ret = thermal_zone_device_disable(data->tz);
 	if (ret)
 		return ret;
+
+	return pm_runtime_force_suspend(data->dev);
+}
+
+static int __maybe_unused imx_thermal_resume(struct device *dev)
+{
+	struct imx_thermal_data *data = dev_get_drvdata(dev);
+	int ret;
+
+	ret = pm_runtime_force_resume(data->dev);
+	if (ret)
+		return ret;
+	/* Enabled thermal sensor after resume */
+	return thermal_zone_device_enable(data->tz);
+}
+
+static int __maybe_unused imx_thermal_runtime_suspend(struct device *dev)
+{
+	struct imx_thermal_data *data = dev_get_drvdata(dev);
+	const struct thermal_soc_data *socdata = data->socdata;
+	struct regmap *map = data->tempmon;
+	int ret;
+
+	ret = regmap_write(map, socdata->sensor_ctrl + REG_CLR,
+			   socdata->measure_temp_mask);
+	if (ret)
+		return ret;
+
+	ret = regmap_write(map, socdata->sensor_ctrl + REG_SET,
+			   socdata->power_down_mask);
+	if (ret)
+		return ret;
+
 	clk_disable_unprepare(data->thermal_clk);
 
 	return 0;
 }
 
-static int __maybe_unused imx_thermal_resume(struct device *dev)
+static int __maybe_unused imx_thermal_runtime_resume(struct device *dev)
 {
 	struct imx_thermal_data *data = dev_get_drvdata(dev);
+	const struct thermal_soc_data *socdata = data->socdata;
+	struct regmap *map = data->tempmon;
 	int ret;
 
 	ret = clk_prepare_enable(data->thermal_clk);
 	if (ret)
 		return ret;
-	/* Enabled thermal sensor after resume */
-	ret = thermal_zone_device_enable(data->tz);
+
+	ret = regmap_write(map, socdata->sensor_ctrl + REG_CLR,
+			   socdata->power_down_mask);
+	if (ret)
+		return ret;
+
+	ret = regmap_write(map, socdata->sensor_ctrl + REG_SET,
+			   socdata->measure_temp_mask);
 	if (ret)
 		return ret;
 
+	/*
+	 * According to the temp sensor designers, it may require up to ~17us
+	 * to complete a measurement.
+	 */
+	usleep_range(20, 50);
+
 	return 0;
 }
 
-static SIMPLE_DEV_PM_OPS(imx_thermal_pm_ops,
-			 imx_thermal_suspend, imx_thermal_resume);
+static const struct dev_pm_ops imx_thermal_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(imx_thermal_suspend, imx_thermal_resume)
+	SET_RUNTIME_PM_OPS(imx_thermal_runtime_suspend,
+			   imx_thermal_runtime_resume, NULL)
+};
 
 static struct platform_driver imx_thermal = {
 	.driver = {