| Message ID | 20240809-auxadc_thermal-v11-1-af36cc74f3a3@chromium.org |
|---|---|
| State | Superseded |
| Headers | show |
| Series | [v11] thermal/drivers/mediatek: add another get_temp ops for thermal sensors | expand |
On 09/08/2024 10:44, Hsin-Te Yuan wrote: > From: James Lo <james.lo@mediatek.com> > > 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/ Please elaborate a bit more the description to stick to the changes. What is the difference between MT8183 Kukui devices and older ones, for instance ? > Signed-off-by: Michael Kao <michael.kao@mediatek.com> > Signed-off-by: Hsin-Yi Wang <hsinyi@chromium.org> > Signed-off-by: Ben Tseng <ben.tseng@mediatek.com> > Signed-off-by: James Lo <james.lo@mediatek.com> > Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> > Signed-off-by: Hsin-Te Yuan <yuanhsinte@chromium.org> > --- > Resurrecting this patch specifically for MediaTek MT8183 Kukui devices. > > Changes in V11: > - Rebase on kernel v6.11-rc2 > - Use mtk_thermal_temp_is_valid in mtk_read_sensor_temp just like > mtk_thermal_bank_temperature > - Change the error handling of devm_thermal_of_zone_register return > value > - link to V10: https://lore.kernel.org/lkml/20220519101044.16765-1-james.lo@mediatek.com/ > > Changes in V10: > - Rebase to kernel-v5.18-rc7 > - Resend > > Changes in V9: > - Rebase to kernel-v5.14-rc1 > - Bind raw_to_mcelsius_v1 or raw_to_mcelsius_v2 to compatible > data of struct mtk_thermal_data > - Remove duplicate struct 'mtk_thermal_bank' > - Remove unnecessary if condition check > - Return error if any thermal zone fail to register > > Changes in V8: > - Rebase to kernel-v5.13-rc1 > - Resend > > Changes in v7: > - Fix build error in v6. > > Changes in v6: > - Rebase to kernel-5.11-rc1. > - [1/3] > - add interrupts property. > - [2/3] > - add the Tested-by in the commit message. > - [3/3] > - use the mt->conf->msr[id] instead of conf->msr[id] in the > _get_sensor_temp and mtk_thermal_bank_temperature. > - remove the redundant space in _get_sensor_temp and > mtk_read_sensor_temp. > - change kmalloc to dev_kmalloc in mtk_thermal_probe. > > Changes in v5: > - Rebase to kernel-5.9-rc1. > - Revise the title of cover letter. > - Drop "[v4,7/7] thermal: mediatek: use spinlock to protect PTPCORESEL" > - [2/2] > - Add the judgement to the version of raw_to_mcelsius. > > Changes in v4: > - Rebase to kernel-5.6-rc1. > - [1/7] > - Squash thermal zone settings in the dtsi from [v3,5/8] > arm64: dts: mt8183: Increase polling frequency for CPU thermal zone. > - Remove the property of interrupts and mediatek,hw-reset-temp. > - [2/7] > - Correct commit message. > - [4/7] > - Change the target temperature to the 80C and change the commit message. > - [6/7] > - Adjust newline alignment. > - Fix the judgement on the return value of registering thermal zone. > > Changes in v3: > - Rebase to kernel-5.5-rc1. > - [1/8] > - Update sustainable power of cpu, tzts1~5 and tztsABB. > - [7/8] > - Bypass the failure that non cpu_thermal sensor is not find in thermal-zones > in dts, which is normal for mt8173, so prompt a warning here instead of > failing. > > 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) > > Changes in v2: > - [1/8] > - Add the sustainable-power,trips,cooling-maps to the tzts1~tztsABB. > - [4/8] > - Add the min opp of cpu throttle. > --- > > --- > drivers/thermal/mediatek/auxadc_thermal.c | 71 +++++++++++++++++++++++++++---- > 1 file changed, 63 insertions(+), 8 deletions(-) > > diff --git a/drivers/thermal/mediatek/auxadc_thermal.c b/drivers/thermal/mediatek/auxadc_thermal.c > index 9ee2e7283435..8b50d560bbf9 100644 > --- a/drivers/thermal/mediatek/auxadc_thermal.c > +++ b/drivers/thermal/mediatek/auxadc_thermal.c > @@ -847,7 +847,8 @@ static int mtk_thermal_bank_temperature(struct mtk_thermal_bank *bank) > > static int mtk_read_temp(struct thermal_zone_device *tz, int *temperature) > { > - struct mtk_thermal *mt = thermal_zone_device_priv(tz); > + struct mtk_thermal_bank *bank = thermal_zone_device_priv(tz); > + struct mtk_thermal *mt = bank->mt; Where is used this variable ? > int i; > int tempmax = INT_MIN; > > @@ -866,10 +867,46 @@ static int mtk_read_temp(struct thermal_zone_device *tz, int *temperature) > return 0; > } > > +static int mtk_read_sensor_temp(struct thermal_zone_device *tz, int *temperature) > +{ > + struct mtk_thermal_bank *bank = thermal_zone_device_priv(tz); > + struct mtk_thermal *mt = bank->mt; > + const struct mtk_thermal_data *conf = mt->conf; > + int id = bank->id - 1; > + int temp = INT_MIN; > + u32 raw; > + > + if (id < 0) > + return -EACCES; This test should not be done here: 1. EACCES is permission denied 2. The bank id should be checked at init time > + > + raw = readl(mt->thermal_base + conf->msr[id]); > + > + temp = mt->raw_to_mcelsius(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 (!mtk_thermal_temp_is_valid(temp)) { unlikely(!mtk_thermal_temp_is_valid(temp)) ? > + temp = THERMAL_TEMP_INVALID; What is the point of assigning this value and return just after ? > + return -EAGAIN; > + } > + > + *temperature = temp; > + > + return 0; > +} > + > static const struct thermal_zone_device_ops mtk_thermal_ops = { > .get_temp = mtk_read_temp, > }; > > +static const struct thermal_zone_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) > @@ -1199,6 +1236,7 @@ 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_bank *tz; > > mt = devm_kzalloc(&pdev->dev, sizeof(*mt), GFP_KERNEL); > if (!mt) > @@ -1285,14 +1323,31 @@ static int mtk_thermal_probe(struct platform_device *pdev) > mtk_thermal_init_bank(mt, i, apmixed_phys_base, > auxadc_phys_base, ctrl_id); > > - tzdev = devm_thermal_of_zone_register(&pdev->dev, 0, mt, > - &mtk_thermal_ops); > - if (IS_ERR(tzdev)) > - return PTR_ERR(tzdev); > + for (i = 0; i < mt->conf->num_sensors + 1; i++) { Why "num_sensors + 1" ? > + tz = devm_kmalloc(&pdev->dev, sizeof(*tz), GFP_KERNEL); > + if (!tz) > + return -ENOMEM; > + > + tz->mt = mt; > + tz->id = i; > + > + tzdev = devm_thermal_of_zone_register(&pdev->dev, i, > + tz, (i == 0) ? > + &mtk_thermal_ops : &mtk_thermal_sensor_ops); Don't you want to keep the init routine and depending on the platform add the loop with the extra sensors ? That may help to make the code cleaner instead of doing tests like (i == 0) ? &mtk_thermal_ops : &mtk_thermal_sensor_ops > + if (IS_ERR(tzdev)) { > + if (PTR_ERR(tzdev) == -ENODEV) { > + dev_warn(&pdev->dev, "can't find thermal sensor %d\n", i); > + continue; > + } > + if (PTR_ERR(tzdev) != -EACCES) The id correctness should be checked in this loop, not rely on a EACCES returned by a implicit call to thermal_zone_get_temp() > + return PTR_ERR(tzdev); > + } > > - ret = devm_thermal_add_hwmon_sysfs(&pdev->dev, tzdev); > - if (ret) > - dev_warn(&pdev->dev, "error in thermal_add_hwmon_sysfs"); > + ret = devm_thermal_add_hwmon_sysfs(&pdev->dev, tzdev); > + if (ret) > + dev_warn(&pdev->dev, "error in thermal_add_hwmon_sysfs: %d\n", ret); > + } > > return 0; > } > > --- > base-commit: ee9a43b7cfe2d8a3520335fea7d8ce71b8cabd9d > change-id: 20240809-auxadc_thermal-9be338ec8b1c > > Best regards,
On Mon, Aug 19, 2024 at 6:33 PM Daniel Lezcano <daniel.lezcano@linaro.org> wrote: > > On 09/08/2024 10:44, Hsin-Te Yuan wrote: > > From: James Lo <james.lo@mediatek.com> > > > > 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/ > > Please elaborate a bit more the description to stick to the changes. > > What is the difference between MT8183 Kukui devices and older ones, for > instance ? > > > Signed-off-by: Michael Kao <michael.kao@mediatek.com> > > Signed-off-by: Hsin-Yi Wang <hsinyi@chromium.org> > > Signed-off-by: Ben Tseng <ben.tseng@mediatek.com> > > Signed-off-by: James Lo <james.lo@mediatek.com> > > Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> > > Signed-off-by: Hsin-Te Yuan <yuanhsinte@chromium.org> > > --- > > Resurrecting this patch specifically for MediaTek MT8183 Kukui devices. > > > > Changes in V11: > > - Rebase on kernel v6.11-rc2 > > - Use mtk_thermal_temp_is_valid in mtk_read_sensor_temp just like > > mtk_thermal_bank_temperature > > - Change the error handling of devm_thermal_of_zone_register return > > value > > - link to V10: https://lore.kernel.org/lkml/20220519101044.16765-1-james.lo@mediatek.com/ > > > > Changes in V10: > > - Rebase to kernel-v5.18-rc7 > > - Resend > > > > Changes in V9: > > - Rebase to kernel-v5.14-rc1 > > - Bind raw_to_mcelsius_v1 or raw_to_mcelsius_v2 to compatible > > data of struct mtk_thermal_data > > - Remove duplicate struct 'mtk_thermal_bank' > > - Remove unnecessary if condition check > > - Return error if any thermal zone fail to register > > > > Changes in V8: > > - Rebase to kernel-v5.13-rc1 > > - Resend > > > > Changes in v7: > > - Fix build error in v6. > > > > Changes in v6: > > - Rebase to kernel-5.11-rc1. > > - [1/3] > > - add interrupts property. > > - [2/3] > > - add the Tested-by in the commit message. > > - [3/3] > > - use the mt->conf->msr[id] instead of conf->msr[id] in the > > _get_sensor_temp and mtk_thermal_bank_temperature. > > - remove the redundant space in _get_sensor_temp and > > mtk_read_sensor_temp. > > - change kmalloc to dev_kmalloc in mtk_thermal_probe. > > > > Changes in v5: > > - Rebase to kernel-5.9-rc1. > > - Revise the title of cover letter. > > - Drop "[v4,7/7] thermal: mediatek: use spinlock to protect PTPCORESEL" > > - [2/2] > > - Add the judgement to the version of raw_to_mcelsius. > > > > Changes in v4: > > - Rebase to kernel-5.6-rc1. > > - [1/7] > > - Squash thermal zone settings in the dtsi from [v3,5/8] > > arm64: dts: mt8183: Increase polling frequency for CPU thermal zone. > > - Remove the property of interrupts and mediatek,hw-reset-temp. > > - [2/7] > > - Correct commit message. > > - [4/7] > > - Change the target temperature to the 80C and change the commit message. > > - [6/7] > > - Adjust newline alignment. > > - Fix the judgement on the return value of registering thermal zone. > > > > Changes in v3: > > - Rebase to kernel-5.5-rc1. > > - [1/8] > > - Update sustainable power of cpu, tzts1~5 and tztsABB. > > - [7/8] > > - Bypass the failure that non cpu_thermal sensor is not find in thermal-zones > > in dts, which is normal for mt8173, so prompt a warning here instead of > > failing. > > > > 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) > > > > Changes in v2: > > - [1/8] > > - Add the sustainable-power,trips,cooling-maps to the tzts1~tztsABB. > > - [4/8] > > - Add the min opp of cpu throttle. > > --- > > > > --- > > drivers/thermal/mediatek/auxadc_thermal.c | 71 +++++++++++++++++++++++++++---- > > 1 file changed, 63 insertions(+), 8 deletions(-) > > > > diff --git a/drivers/thermal/mediatek/auxadc_thermal.c b/drivers/thermal/mediatek/auxadc_thermal.c > > index 9ee2e7283435..8b50d560bbf9 100644 > > --- a/drivers/thermal/mediatek/auxadc_thermal.c > > +++ b/drivers/thermal/mediatek/auxadc_thermal.c > > @@ -847,7 +847,8 @@ static int mtk_thermal_bank_temperature(struct mtk_thermal_bank *bank) > > > > static int mtk_read_temp(struct thermal_zone_device *tz, int *temperature) > > { > > - struct mtk_thermal *mt = thermal_zone_device_priv(tz); > > + struct mtk_thermal_bank *bank = thermal_zone_device_priv(tz); > > + struct mtk_thermal *mt = bank->mt; > > Where is used this variable ? > This is used by the for loop in this function. > > int i; > > int tempmax = INT_MIN; > > > > @@ -866,10 +867,46 @@ static int mtk_read_temp(struct thermal_zone_device *tz, int *temperature) > > return 0; > > } > > > > +static int mtk_read_sensor_temp(struct thermal_zone_device *tz, int *temperature) > > +{ > > + struct mtk_thermal_bank *bank = thermal_zone_device_priv(tz); > > + struct mtk_thermal *mt = bank->mt; > > + const struct mtk_thermal_data *conf = mt->conf; > > + int id = bank->id - 1; > > + int temp = INT_MIN; > > + u32 raw; > > + > > + if (id < 0) > > + return -EACCES; > > This test should not be done here: > > 1. EACCES is permission denied > > 2. The bank id should be checked at init time > > > + > > + raw = readl(mt->thermal_base + conf->msr[id]); > > + > > + temp = mt->raw_to_mcelsius(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 (!mtk_thermal_temp_is_valid(temp)) { > > unlikely(!mtk_thermal_temp_is_valid(temp)) ? > > > + temp = THERMAL_TEMP_INVALID; > > What is the point of assigning this value and return just after ? > I'll remove this line. > > + return -EAGAIN; > > + } > > + > > + *temperature = temp; > > + > > + return 0; > > +} > > + > > static const struct thermal_zone_device_ops mtk_thermal_ops = { > > .get_temp = mtk_read_temp, > > }; > > > > +static const struct thermal_zone_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) > > @@ -1199,6 +1236,7 @@ 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_bank *tz; > > > > mt = devm_kzalloc(&pdev->dev, sizeof(*mt), GFP_KERNEL); > > if (!mt) > > @@ -1285,14 +1323,31 @@ static int mtk_thermal_probe(struct platform_device *pdev) > > mtk_thermal_init_bank(mt, i, apmixed_phys_base, > > auxadc_phys_base, ctrl_id); > > > > - tzdev = devm_thermal_of_zone_register(&pdev->dev, 0, mt, > > - &mtk_thermal_ops); > > - if (IS_ERR(tzdev)) > > - return PTR_ERR(tzdev); > > + for (i = 0; i < mt->conf->num_sensors + 1; i++) { > > Why "num_sensors + 1" ? > Because we have the one with id 0 which represents the max of all the temperature and num_sensors real sensors. > > + tz = devm_kmalloc(&pdev->dev, sizeof(*tz), GFP_KERNEL); > > + if (!tz) > > + return -ENOMEM; > > + > > + tz->mt = mt; > > + tz->id = i; > > + > > + tzdev = devm_thermal_of_zone_register(&pdev->dev, i, > > + tz, (i == 0) ? > > + &mtk_thermal_ops : &mtk_thermal_sensor_ops); > > Don't you want to keep the init routine and depending on the platform > add the loop with the extra sensors ? That may help to make the code > cleaner instead of doing tests like (i == 0) ? &mtk_thermal_ops : > &mtk_thermal_sensor_ops > How about moving the registration with i = 0 out of this loop. Although some platforms don't use those extra sensors, they can read the temperature from these extra sensors by simply adding thermal zones in devicetree. > > + if (IS_ERR(tzdev)) { > > + if (PTR_ERR(tzdev) == -ENODEV) { > > + dev_warn(&pdev->dev, "can't find thermal sensor %d\n", i); > > + continue; > > + } > > + if (PTR_ERR(tzdev) != -EACCES) > > The id correctness should be checked in this loop, not rely on a EACCES > returned by a implicit call to thermal_zone_get_temp() > Since we use the for loop here, checking the id correctness looks meaningless. I'll change to return error directly and remove the id correctness checking from mtk_read_sensor_temp. > > + return PTR_ERR(tzdev); > > + } > > > > - ret = devm_thermal_add_hwmon_sysfs(&pdev->dev, tzdev); > > - if (ret) > > - dev_warn(&pdev->dev, "error in thermal_add_hwmon_sysfs"); > > + ret = devm_thermal_add_hwmon_sysfs(&pdev->dev, tzdev); > > + if (ret) > > + dev_warn(&pdev->dev, "error in thermal_add_hwmon_sysfs: %d\n", ret); > > + } > > > > return 0; > > } > > > > --- > > base-commit: ee9a43b7cfe2d8a3520335fea7d8ce71b8cabd9d > > change-id: 20240809-auxadc_thermal-9be338ec8b1c > > > > Best regards, > > > -- > <http://www.linaro.org/> Linaro.org │ Open source software for ARM SoCs > > Follow Linaro: <http://www.facebook.com/pages/Linaro> Facebook | > <http://twitter.com/#!/linaroorg> Twitter | > <http://www.linaro.org/linaro-blog/> Blog
diff --git a/drivers/thermal/mediatek/auxadc_thermal.c b/drivers/thermal/mediatek/auxadc_thermal.c index 9ee2e7283435..8b50d560bbf9 100644 --- a/drivers/thermal/mediatek/auxadc_thermal.c +++ b/drivers/thermal/mediatek/auxadc_thermal.c @@ -847,7 +847,8 @@ static int mtk_thermal_bank_temperature(struct mtk_thermal_bank *bank) static int mtk_read_temp(struct thermal_zone_device *tz, int *temperature) { - struct mtk_thermal *mt = thermal_zone_device_priv(tz); + struct mtk_thermal_bank *bank = thermal_zone_device_priv(tz); + struct mtk_thermal *mt = bank->mt; int i; int tempmax = INT_MIN; @@ -866,10 +867,46 @@ static int mtk_read_temp(struct thermal_zone_device *tz, int *temperature) return 0; } +static int mtk_read_sensor_temp(struct thermal_zone_device *tz, int *temperature) +{ + struct mtk_thermal_bank *bank = thermal_zone_device_priv(tz); + struct mtk_thermal *mt = bank->mt; + const struct mtk_thermal_data *conf = mt->conf; + int id = bank->id - 1; + int temp = INT_MIN; + u32 raw; + + if (id < 0) + return -EACCES; + + raw = readl(mt->thermal_base + conf->msr[id]); + + temp = mt->raw_to_mcelsius(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 (!mtk_thermal_temp_is_valid(temp)) { + temp = THERMAL_TEMP_INVALID; + return -EAGAIN; + } + + *temperature = temp; + + return 0; +} + static const struct thermal_zone_device_ops mtk_thermal_ops = { .get_temp = mtk_read_temp, }; +static const struct thermal_zone_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) @@ -1199,6 +1236,7 @@ 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_bank *tz; mt = devm_kzalloc(&pdev->dev, sizeof(*mt), GFP_KERNEL); if (!mt) @@ -1285,14 +1323,31 @@ static int mtk_thermal_probe(struct platform_device *pdev) mtk_thermal_init_bank(mt, i, apmixed_phys_base, auxadc_phys_base, ctrl_id); - tzdev = devm_thermal_of_zone_register(&pdev->dev, 0, mt, - &mtk_thermal_ops); - if (IS_ERR(tzdev)) - return PTR_ERR(tzdev); + for (i = 0; i < mt->conf->num_sensors + 1; i++) { + tz = devm_kmalloc(&pdev->dev, sizeof(*tz), GFP_KERNEL); + if (!tz) + return -ENOMEM; + + tz->mt = mt; + tz->id = i; + + tzdev = devm_thermal_of_zone_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, "can't find thermal sensor %d\n", i); + continue; + } + if (PTR_ERR(tzdev) != -EACCES) + return PTR_ERR(tzdev); + } - ret = devm_thermal_add_hwmon_sysfs(&pdev->dev, tzdev); - if (ret) - dev_warn(&pdev->dev, "error in thermal_add_hwmon_sysfs"); + ret = devm_thermal_add_hwmon_sysfs(&pdev->dev, tzdev); + if (ret) + dev_warn(&pdev->dev, "error in thermal_add_hwmon_sysfs: %d\n", ret); + } return 0; }