[v2,4/4] thermal: mediatek: add another get_temp ops for thermal sensors

Provide thermal zone to read thermal sensor in the SoC. We can read all the
thermal sensors value in the SoC by the node /sys/class/thermal/

In mtk_thermal_bank_temperature, return -EAGAIN instead of -EACCESS
on the first read of sensor that often are bogus values.
This can avoid following warning on boot:

  thermal thermal_zone6: failed to read out thermal zone (-13)

Signed-off-by: default avatarMichael Kao <michael.kao@mediatek.com>
Signed-off-by: default avatarHsin-Yi Wang <hsinyi@chromium.org>
Signed-off-by: Amjad Ouled-Ameur <aouledameur@baylibre.com>
Tested-by: Amjad Ouled-Ameur <aouledameur@baylibre.com>
---
 drivers/thermal/mtk_thermal.c | 100 ++++++++++++++++++++++++++--------
 1 file changed, 76 insertions(+), 24 deletions(-)

Hi Christophe,

On 7/20/22 20:54, Christophe JAILLET wrote:
> Le 20/07/2022 à 20:18, Amjad Ouled-Ameur a écrit :
>> Provide thermal zone to read thermal sensor in the SoC. We can read 
>> all the
>> thermal sensors value in the SoC by the node /sys/class/thermal/
>>
>> In mtk_thermal_bank_temperature, return -EAGAIN instead of -EACCESS
>> on the first read of sensor that often are bogus values.
>> This can avoid following warning on boot:
>>
>>    thermal thermal_zone6: failed to read out thermal zone (-13)
>>
>> Signed-off-by: default avatarMichael Kao 
>> <michael.kao-NuS5LvNUpcJWk0Htik3J/w@public.gmane.org>
>> Signed-off-by: default avatarHsin-Yi Wang 
>> <hsinyi-F7+t8E8rja9g9hUCZPvPmw@public.gmane.org>
>> Signed-off-by: Amjad Ouled-Ameur 
>> <aouledameur-rdvid1DuHRBWk0Htik3J/w@public.gmane.org>
>> Tested-by: Amjad Ouled-Ameur 
>> <aouledameur-rdvid1DuHRBWk0Htik3J/w@public.gmane.org>
>> ---
>>   drivers/thermal/mtk_thermal.c | 100 ++++++++++++++++++++++++++--------
>>   1 file changed, 76 insertions(+), 24 deletions(-)
>>
>> diff --git a/drivers/thermal/mtk_thermal.c 
>> b/drivers/thermal/mtk_thermal.c
>> index 1dc276f8c4f1..79b14ce1a08d 100644
>> --- a/drivers/thermal/mtk_thermal.c
>> +++ b/drivers/thermal/mtk_thermal.c
>> @@ -259,6 +259,11 @@ enum mtk_thermal_version {
>>     struct mtk_thermal;
>>   +struct mtk_thermal_zone {
>> +    struct mtk_thermal *mt;
>> +    int id;
>> +};
>> +
>>   struct thermal_bank_cfg {
>>       unsigned int num_sensors;
>>       const int *sensors;
>> @@ -709,6 +714,32 @@ static void mtk_thermal_put_bank(struct 
>> mtk_thermal_bank *bank)
>>           mutex_unlock(&mt->lock);
>>   }
>>   +static u32 _get_sensor_temp(struct mtk_thermal *mt, int id)
>> +{
>> +    u32 raw;
>> +    int temp;
>> +
>> +    const struct mtk_thermal_data *conf = mt->conf;
>> +
>> +    raw = readl(mt->thermal_base + conf->msr[id]);
>> +
>> +    if (mt->conf->version == MTK_THERMAL_V1)
>> +        temp = raw_to_mcelsius_v1(mt, id, raw);
>> +    else
>> +        temp = raw_to_mcelsius_v2(mt, id, raw);
>> +
>> +    /*
>> +     * The first read of a sensor often contains very high bogus
>> +     * temperature value. Filter these out so that the system does
>> +     * not immediately shut down.
>> +     */
>> +
>> +    if (temp > 200000)
>> +        return  -EAGAIN;
>
> This function returns a u32. Is it ok to return -EAGAIN?
>
_get_sensor_temp() should normally return int instead u32, will fix it 
in V3.
> There is also 2 spaces here...
>
>> +    else
>> +        return    temp;
>
> ... and a tab here.
>
will fix them in V3.
>> +}
>> +
>>   /**
>>    * mtk_thermal_bank_temperature - get the temperature of a bank
>>    * @bank:    The bank
>> @@ -721,26 +752,9 @@ static int mtk_thermal_bank_temperature(struct 
>> mtk_thermal_bank *bank)
>>       struct mtk_thermal *mt = bank->mt;
>>       const struct mtk_thermal_data *conf = mt->conf;
>>       int i, temp = INT_MIN, max = INT_MIN;
>> -    u32 raw;
>>         for (i = 0; i < conf->bank_data[bank->id].num_sensors; i++) {
>> -        raw = readl(mt->thermal_base + conf->msr[i]);
>> -
>> -        if (mt->conf->version == MTK_THERMAL_V1) {
>> -            temp = raw_to_mcelsius_v1(
>> -                mt, conf->bank_data[bank->id].sensors[i], raw);
>> -        } else {
>> -            temp = raw_to_mcelsius_v2(
>> -                mt, conf->bank_data[bank->id].sensors[i], raw);
>> -        }
>> -
>> -        /*
>> -         * The first read of a sensor often contains very high bogus
>> -         * temperature value. Filter these out so that the system does
>> -         * not immediately shut down.
>> -         */
>> -        if (temp > 200000)
>> -            temp = 0;
>> +        temp = _get_sensor_temp(mt, i);
>
> Is it ok if _get_sensor_temp() returns -EAGAIN?
>
drivers/thermal/thermal_core.c:update_temperature() checks for -EAGAIN 
after

thermal_zone_get_temp() is called, thus, I think it's good to return 
-EAGAIN in

case of failure.

>>             if (temp > max)
>>               max = temp;
>> @@ -751,7 +765,8 @@ static int mtk_thermal_bank_temperature(struct 
>> mtk_thermal_bank *bank)
>>     static int mtk_read_temp(void *data, int *temperature)
>>   {
>> -    struct mtk_thermal *mt = data;
>> +    struct mtk_thermal_zone *tz = data;
>> +    struct mtk_thermal *mt = tz->mt;
>>       int i;
>>       int tempmax = INT_MIN;
>>   @@ -770,10 +785,28 @@ static int mtk_read_temp(void *data, int 
>> *temperature)
>>       return 0;
>>   }
>>   +static int mtk_read_sensor_temp(void *data, int *temperature)
>> +{
>> +    struct mtk_thermal_zone *tz = data;
>> +    struct mtk_thermal *mt = tz->mt;
>> +    int id = tz->id - 1;
>> +
>> +    if (id < 0)
>> +        return  -EACCES;
>
> 2 spaces.
>

will fix it in V3.
>> +
>> +    *temperature = _get_sensor_temp(mt, id);
>
> If _get_sensor_temp() returns -EAGAIN, should this be propagated to 
> the caller?
>
>> +
>> +    return 0;
>> +}
>> +
>>   static const struct thermal_zone_of_device_ops mtk_thermal_ops = {
>>       .get_temp = mtk_read_temp,
>>   };
>>   +static const struct thermal_zone_of_device_ops 
>> mtk_thermal_sensor_ops = {
>> +    .get_temp = mtk_read_sensor_temp,
>> +};
>> +
>>   static void mtk_thermal_init_bank(struct mtk_thermal *mt, int num,
>>                     u32 apmixed_phys_base, u32 auxadc_phys_base,
>>                     int ctrl_id)
>> @@ -1072,6 +1105,7 @@ static int mtk_thermal_probe(struct 
>> platform_device *pdev)
>>       u64 auxadc_phys_base, apmixed_phys_base;
>>       struct thermal_zone_device *tzdev;
>>       void __iomem *apmixed_base, *auxadc_base;
>> +    struct mtk_thermal_zone *tz;
>>         mt = devm_kzalloc(&pdev->dev, sizeof(*mt), GFP_KERNEL);
>>       if (!mt)
>> @@ -1161,11 +1195,29 @@ static int mtk_thermal_probe(struct 
>> platform_device *pdev)
>>         platform_set_drvdata(pdev, mt);
>>   -    tzdev = devm_thermal_zone_of_sensor_register(&pdev->dev, 0, mt,
>> -                             &mtk_thermal_ops);
>> -    if (IS_ERR(tzdev)) {
>> -        ret = PTR_ERR(tzdev);
>> -        goto err_disable_clk_peri_therm;
>> +    for (i = 0; i < mt->conf->num_sensors + 1; i++) {
>> +        tz = kmalloc(sizeof(*tz), GFP_KERNEL);
>
> Should this memory allocation be a devm_kmalloc(), or is this memory 
> freed at some point by the framework?
>
> (I don't know the thermal_zone API and the patch has no kfree())
AFAIK, thermal API does not free private data, therefore devm_kmalloc() 
should be used.
>
> CJ
>
>> +        if (!tz)
>> +            return -ENOMEM;
>> +
>> +        tz->mt = mt;
>> +        tz->id = i;
>> +
>> +        tzdev = devm_thermal_zone_of_sensor_register(&pdev->dev, i, 
>> tz, (i == 0) ?
>> +                                 &mtk_thermal_ops :
>> +                                 &mtk_thermal_sensor_ops);
>> +
>> +        if (IS_ERR(tzdev)) {
>> +            if (PTR_ERR(tzdev) == -ENODEV) {
>> +                dev_warn(&pdev->dev,
>> +                     "sensor %d not registered in thermal zone in 
>> dt\n", i);
>> +                continue;
>> +            }
>> +            if (PTR_ERR(tzdev) == -EACCES) {
>> +                ret = PTR_ERR(tzdev);
>> +                goto err_disable_clk_peri_therm;
>> +            }
>> +        }
>>       }
>>         ret = devm_thermal_add_hwmon_sysfs(tzdev);
>
Thank you Christophe for the review.

Regards,

Amjad