| Message ID | 20201128175450.12456-1-kai.heng.feng@canonical.com (mailing list archive) |
|---|---|
| State | Superseded |
| Delegated to: | Zhang Rui |
| Headers | show |
| Series | [1/3] thermal: core: Add indication for userspace usage | expand |
[Added Srinivas] On 28/11/2020 18:54, Kai-Heng Feng wrote: > We are seeing thermal shutdown on Intel based mobile workstations, the > shutdown happens during the first trip handle in > thermal_zone_device_register(): > kernel: thermal thermal_zone15: critical temperature reached (101 C), shutting down > > However, we shouldn't do a thermal shutdown here, since > 1) We may want to use a dedicated daemon, Intel's thermald in this case, > to handle thermal shutdown. > > 2) For ACPI based system, _CRT doesn't mean shutdown unless it's inside > ThermalZone. ACPI Spec, 11.4.4 _CRT (Critical Temperature): > "... If this object it present under a device, the device’s driver > evaluates this object to determine the device’s critical cooling > temperature trip point. This value may then be used by the device’s > driver to program an internal device temperature sensor trip point." > > So a "critical trip" here merely means we should take a more aggressive > cooling method. Well, actually it is stated before: "This object, when defined under a thermal zone, returns the critical temperature at which OSPM must shutdown the system". That is what does the thermal subsystem, no ? > So add an indication to let thermal core know it should leave thermal > device to userspace to handle. You may want to check the 'HOT' trip point and then use the notification mechanism to get notified in userspace and take action from there (eg. offline some CPUs). > Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com> > --- > drivers/thermal/thermal_core.c | 3 +++ > include/linux/thermal.h | 2 ++ > 2 files changed, 5 insertions(+) > > diff --git a/drivers/thermal/thermal_core.c b/drivers/thermal/thermal_core.c > index c6d74bc1c90b..6561e3767529 100644 > --- a/drivers/thermal/thermal_core.c > +++ b/drivers/thermal/thermal_core.c > @@ -1477,6 +1477,9 @@ thermal_zone_device_register(const char *type, int trips, int mask, > goto unregister; > } > > + if (tz->tzp && tz->tzp->userspace) > + thermal_zone_device_disable(tz); > + > mutex_lock(&thermal_list_lock); > list_add_tail(&tz->node, &thermal_tz_list); > mutex_unlock(&thermal_list_lock); > diff --git a/include/linux/thermal.h b/include/linux/thermal.h > index d07ea27e72a9..e8e8fac78fc8 100644 > --- a/include/linux/thermal.h > +++ b/include/linux/thermal.h > @@ -247,6 +247,8 @@ struct thermal_zone_params { > */ > bool no_hwmon; > > + bool userspace; > + > int num_tbps; /* Number of tbp entries */ > struct thermal_bind_params *tbp; > >
> On Nov 30, 2020, at 15:57, Daniel Lezcano <daniel.lezcano@linaro.org> wrote: > > > [Added Srinivas] > > On 28/11/2020 18:54, Kai-Heng Feng wrote: >> We are seeing thermal shutdown on Intel based mobile workstations, the >> shutdown happens during the first trip handle in >> thermal_zone_device_register(): >> kernel: thermal thermal_zone15: critical temperature reached (101 C), shutting down >> >> However, we shouldn't do a thermal shutdown here, since >> 1) We may want to use a dedicated daemon, Intel's thermald in this case, >> to handle thermal shutdown. >> >> 2) For ACPI based system, _CRT doesn't mean shutdown unless it's inside >> ThermalZone. ACPI Spec, 11.4.4 _CRT (Critical Temperature): >> "... If this object it present under a device, the device’s driver >> evaluates this object to determine the device’s critical cooling >> temperature trip point. This value may then be used by the device’s >> driver to program an internal device temperature sensor trip point." >> >> So a "critical trip" here merely means we should take a more aggressive >> cooling method. > > Well, actually it is stated before: > > "This object, when defined under a thermal zone, returns the critical > temperature at which OSPM must shutdown the system". This means specifically for the ACPI ThermalZone in AML, e.g.: ThermalZone (TZ0) { .... Method(_CRT) { ... } } // end of TZ0 However the device is not under any ACPI ThermalZone. > > That is what does the thermal subsystem, no ? > >> So add an indication to let thermal core know it should leave thermal >> device to userspace to handle. > > You may want to check the 'HOT' trip point and then use the notification > mechanism to get notified in userspace and take action from there (eg. > offline some CPUs). For this particular issue we are facing, the thermal shutdown happens in thermal_zone_device_register() and userspace isn't up yet. Kai-Heng > >> Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com> >> --- >> drivers/thermal/thermal_core.c | 3 +++ >> include/linux/thermal.h | 2 ++ >> 2 files changed, 5 insertions(+) >> >> diff --git a/drivers/thermal/thermal_core.c b/drivers/thermal/thermal_core.c >> index c6d74bc1c90b..6561e3767529 100644 >> --- a/drivers/thermal/thermal_core.c >> +++ b/drivers/thermal/thermal_core.c >> @@ -1477,6 +1477,9 @@ thermal_zone_device_register(const char *type, int trips, int mask, >> goto unregister; >> } >> >> + if (tz->tzp && tz->tzp->userspace) >> + thermal_zone_device_disable(tz); >> + >> mutex_lock(&thermal_list_lock); >> list_add_tail(&tz->node, &thermal_tz_list); >> mutex_unlock(&thermal_list_lock); >> diff --git a/include/linux/thermal.h b/include/linux/thermal.h >> index d07ea27e72a9..e8e8fac78fc8 100644 >> --- a/include/linux/thermal.h >> +++ b/include/linux/thermal.h >> @@ -247,6 +247,8 @@ struct thermal_zone_params { >> */ >> bool no_hwmon; >> >> + bool userspace; >> + >> int num_tbps; /* Number of tbp entries */ >> struct thermal_bind_params *tbp; >> >> > > > -- > <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
On Mon, 2020-11-30 at 16:23 +0800, Kai-Heng Feng wrote: > > On Nov 30, 2020, at 15:57, Daniel Lezcano < > > daniel.lezcano@linaro.org> wrote: > > > > > > [Added Srinivas] > > > > On 28/11/2020 18:54, Kai-Heng Feng wrote: > > > We are seeing thermal shutdown on Intel based mobile > > > workstations, the > > > shutdown happens during the first trip handle in > > > thermal_zone_device_register(): > > > kernel: thermal thermal_zone15: critical temperature reached (101 > > > C), shutting down > > > > > > However, we shouldn't do a thermal shutdown here, since > > > 1) We may want to use a dedicated daemon, Intel's thermald in > > > this case, > > > to handle thermal shutdown. > > > > > > 2) For ACPI based system, _CRT doesn't mean shutdown unless it's > > > inside > > > ThermalZone. ACPI Spec, 11.4.4 _CRT (Critical Temperature): > > > "... If this object it present under a device, the device’s > > > driver > > > evaluates this object to determine the device’s critical cooling > > > temperature trip point. This value may then be used by the > > > device’s > > > driver to program an internal device temperature sensor trip > > > point." > > > > > > So a "critical trip" here merely means we should take a more > > > aggressive > > > cooling method. > > > > Well, actually it is stated before: > > > > "This object, when defined under a thermal zone, returns the > > critical > > temperature at which OSPM must shutdown the system". > > This means specifically for the ACPI ThermalZone in AML, e.g.: > > ThermalZone (TZ0) { > .... > Method(_CRT) { ... } > } // end of TZ0 > > However the device is not under any ACPI ThermalZone. > > > That is what does the thermal subsystem, no ? > > > > > So add an indication to let thermal core know it should leave > > > thermal > > > device to userspace to handle. > > > > You may want to check the 'HOT' trip point and then use the > > notification > > mechanism to get notified in userspace and take action from there > > (eg. > > offline some CPUs). > > For this particular issue we are facing, the thermal shutdown happens > in thermal_zone_device_register() and userspace isn't up yet. What about creating an new callback enum thermal_trip_status { THERMAL_TRIP_DISABLED = 0, THERMAL_TRIP_ENABLED, }; int get_trip_status(struct thermal_zone_device *, int trip, enum thermal_trip_status *state); Then in static void handle_thermal_trip(struct thermal_zone_device *tz, int trip) { /* before tz->ops->get_trip_temp(tz, trip, &trip_temp); */ if (tz->ops->get_trip_status) { enum thermal_trip_status *status; if (!tz->ops->get_trip_status(tz, trip, &status)) { if (status == THERMAL_TRIP_DISABLED) return; } } ... ... } This callback will help the cases: - Allows drivers to selectively disable certain trips during init state or system resume where there can be spikes or always. int340x drivers can disable always. - Still give options for drivers to handle critical trip even if they are bound to user space governors. User space process may be dead, so still allow kernel to process graceful shutdown Thanks, Srinivas > > Kai-Heng > > > > Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com> > > > --- > > > drivers/thermal/thermal_core.c | 3 +++ > > > include/linux/thermal.h | 2 ++ > > > 2 files changed, 5 insertions(+) > > > > > > diff --git a/drivers/thermal/thermal_core.c > > > b/drivers/thermal/thermal_core.c > > > index c6d74bc1c90b..6561e3767529 100644 > > > --- a/drivers/thermal/thermal_core.c > > > +++ b/drivers/thermal/thermal_core.c > > > @@ -1477,6 +1477,9 @@ thermal_zone_device_register(const char > > > *type, int trips, int mask, > > > goto unregister; > > > } > > > > > > + if (tz->tzp && tz->tzp->userspace) > > > + thermal_zone_device_disable(tz); > > > + > > > mutex_lock(&thermal_list_lock); > > > list_add_tail(&tz->node, &thermal_tz_list); > > > mutex_unlock(&thermal_list_lock); > > > diff --git a/include/linux/thermal.h b/include/linux/thermal.h > > > index d07ea27e72a9..e8e8fac78fc8 100644 > > > --- a/include/linux/thermal.h > > > +++ b/include/linux/thermal.h > > > @@ -247,6 +247,8 @@ struct thermal_zone_params { > > > */ > > > bool no_hwmon; > > > > > > + bool userspace; > > > + > > > int num_tbps; /* Number of tbp entries */ > > > struct thermal_bind_params *tbp; > > > > > > > > > > -- > > <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
> On Dec 1, 2020, at 00:19, Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com> wrote: > > On Mon, 2020-11-30 at 16:23 +0800, Kai-Heng Feng wrote: >>> On Nov 30, 2020, at 15:57, Daniel Lezcano < >>> daniel.lezcano@linaro.org> wrote: >>> >>> >>> [Added Srinivas] >>> >>> On 28/11/2020 18:54, Kai-Heng Feng wrote: >>>> We are seeing thermal shutdown on Intel based mobile >>>> workstations, the >>>> shutdown happens during the first trip handle in >>>> thermal_zone_device_register(): >>>> kernel: thermal thermal_zone15: critical temperature reached (101 >>>> C), shutting down >>>> >>>> However, we shouldn't do a thermal shutdown here, since >>>> 1) We may want to use a dedicated daemon, Intel's thermald in >>>> this case, >>>> to handle thermal shutdown. >>>> >>>> 2) For ACPI based system, _CRT doesn't mean shutdown unless it's >>>> inside >>>> ThermalZone. ACPI Spec, 11.4.4 _CRT (Critical Temperature): >>>> "... If this object it present under a device, the device’s >>>> driver >>>> evaluates this object to determine the device’s critical cooling >>>> temperature trip point. This value may then be used by the >>>> device’s >>>> driver to program an internal device temperature sensor trip >>>> point." >>>> >>>> So a "critical trip" here merely means we should take a more >>>> aggressive >>>> cooling method. >>> >>> Well, actually it is stated before: >>> >>> "This object, when defined under a thermal zone, returns the >>> critical >>> temperature at which OSPM must shutdown the system". >> >> This means specifically for the ACPI ThermalZone in AML, e.g.: >> >> ThermalZone (TZ0) { >> .... >> Method(_CRT) { ... } >> } // end of TZ0 >> >> However the device is not under any ACPI ThermalZone. >> >>> That is what does the thermal subsystem, no ? >>> >>>> So add an indication to let thermal core know it should leave >>>> thermal >>>> device to userspace to handle. >>> >>> You may want to check the 'HOT' trip point and then use the >>> notification >>> mechanism to get notified in userspace and take action from there >>> (eg. >>> offline some CPUs). >> >> For this particular issue we are facing, the thermal shutdown happens >> in thermal_zone_device_register() and userspace isn't up yet. > > What about creating an new callback > > enum thermal_trip_status { > THERMAL_TRIP_DISABLED = 0, > THERMAL_TRIP_ENABLED, > }; > > int get_trip_status(struct thermal_zone_device *, int trip, enum > thermal_trip_status *state); > > Then in > static void handle_thermal_trip(struct thermal_zone_device *tz, int > trip) > { > > /* before tz->ops->get_trip_temp(tz, trip, &trip_temp); */ > if (tz->ops->get_trip_status) { > enum thermal_trip_status *status; > > if (!tz->ops->get_trip_status(tz, trip, &status)) { > if (status == THERMAL_TRIP_DISABLED) > return; > } > } > ... > ... > > } > > > This callback will help the cases: > - Allows drivers to selectively disable certain trips during init state > or system resume where there can be spikes or always. int340x drivers > can disable always. This sounds really great. This is indeed can happen on system resume, before userspace process thaw. > - Still give options for drivers to handle critical trip even if they > are bound to user space governors. User space process may be dead, so > still allow kernel to process graceful shutdown To make the scenario happen, do we need a new sysfs to let usespace enable it with THERMAL_TRIP_ENABLED? Kai-Heng > > Thanks, > Srinivas > >> >> Kai-Heng >> >>>> Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com> >>>> --- >>>> drivers/thermal/thermal_core.c | 3 +++ >>>> include/linux/thermal.h | 2 ++ >>>> 2 files changed, 5 insertions(+) >>>> >>>> diff --git a/drivers/thermal/thermal_core.c >>>> b/drivers/thermal/thermal_core.c >>>> index c6d74bc1c90b..6561e3767529 100644 >>>> --- a/drivers/thermal/thermal_core.c >>>> +++ b/drivers/thermal/thermal_core.c >>>> @@ -1477,6 +1477,9 @@ thermal_zone_device_register(const char >>>> *type, int trips, int mask, >>>> goto unregister; >>>> } >>>> >>>> + if (tz->tzp && tz->tzp->userspace) >>>> + thermal_zone_device_disable(tz); >>>> + >>>> mutex_lock(&thermal_list_lock); >>>> list_add_tail(&tz->node, &thermal_tz_list); >>>> mutex_unlock(&thermal_list_lock); >>>> diff --git a/include/linux/thermal.h b/include/linux/thermal.h >>>> index d07ea27e72a9..e8e8fac78fc8 100644 >>>> --- a/include/linux/thermal.h >>>> +++ b/include/linux/thermal.h >>>> @@ -247,6 +247,8 @@ struct thermal_zone_params { >>>> */ >>>> bool no_hwmon; >>>> >>>> + bool userspace; >>>> + >>>> int num_tbps; /* Number of tbp entries */ >>>> struct thermal_bind_params *tbp; >>>> >>>> >>> >>> -- >>> <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
On Tue, 2020-12-01 at 02:04 +0800, Kai-Heng Feng wrote: > > On Dec 1, 2020, at 00:19, Srinivas Pandruvada < > > srinivas.pandruvada@linux.intel.com> wrote: > > > > On Mon, 2020-11-30 at 16:23 +0800, Kai-Heng Feng wrote: > > > > On Nov 30, 2020, at 15:57, Daniel Lezcano < > > > > daniel.lezcano@linaro.org> wrote: > > > > > > > > > > > > [Added Srinivas] > > > > > > > > On 28/11/2020 18:54, Kai-Heng Feng wrote: > > > > > We are seeing thermal shutdown on Intel based mobile > > > > > workstations, the > > > > > shutdown happens during the first trip handle in > > > > > thermal_zone_device_register(): > > > > > kernel: thermal thermal_zone15: critical temperature reached > > > > > (101 > > > > > C), shutting down > > > > > > > > > > However, we shouldn't do a thermal shutdown here, since > > > > > 1) We may want to use a dedicated daemon, Intel's thermald in > > > > > this case, > > > > > to handle thermal shutdown. > > > > > > > > > > 2) For ACPI based system, _CRT doesn't mean shutdown unless > > > > > it's > > > > > inside > > > > > ThermalZone. ACPI Spec, 11.4.4 _CRT (Critical Temperature): > > > > > "... If this object it present under a device, the device’s > > > > > driver > > > > > evaluates this object to determine the device’s critical > > > > > cooling > > > > > temperature trip point. This value may then be used by the > > > > > device’s > > > > > driver to program an internal device temperature sensor trip > > > > > point." > > > > > > > > > > So a "critical trip" here merely means we should take a more > > > > > aggressive > > > > > cooling method. > > > > > > > > Well, actually it is stated before: > > > > > > > > "This object, when defined under a thermal zone, returns the > > > > critical > > > > temperature at which OSPM must shutdown the system". > > > > > > This means specifically for the ACPI ThermalZone in AML, e.g.: > > > > > > ThermalZone (TZ0) { > > > .... > > > Method(_CRT) { ... } > > > } // end of TZ0 > > > > > > However the device is not under any ACPI ThermalZone. > > > > > > > That is what does the thermal subsystem, no ? > > > > > > > > > So add an indication to let thermal core know it should leave > > > > > thermal > > > > > device to userspace to handle. > > > > > > > > You may want to check the 'HOT' trip point and then use the > > > > notification > > > > mechanism to get notified in userspace and take action from > > > > there > > > > (eg. > > > > offline some CPUs). > > > > > > For this particular issue we are facing, the thermal shutdown > > > happens > > > in thermal_zone_device_register() and userspace isn't up yet. > > > > What about creating an new callback > > > > enum thermal_trip_status { > > THERMAL_TRIP_DISABLED = 0, > > THERMAL_TRIP_ENABLED, > > }; > > > > int get_trip_status(struct thermal_zone_device *, int trip, enum > > thermal_trip_status *state); > > > > Then in > > static void handle_thermal_trip(struct thermal_zone_device *tz, int > > trip) > > { > > > > /* before tz->ops->get_trip_temp(tz, trip, &trip_temp); */ > > if (tz->ops->get_trip_status) { > > enum thermal_trip_status *status; > > > > if (!tz->ops->get_trip_status(tz, trip, &status)) { > > if (status == THERMAL_TRIP_DISABLED) > > return; > > } > > } > > ... > > ... > > > > } > > > > > > This callback will help the cases: > > - Allows drivers to selectively disable certain trips during init > > state > > or system resume where there can be spikes or always. int340x > > drivers > > can disable always. > > This sounds really great. This is indeed can happen on system resume, > before userspace process thaw. > > > - Still give options for drivers to handle critical trip even if > > they > > are bound to user space governors. User space process may be dead, > > so > > still allow kernel to process graceful shutdown > > To make the scenario happen, do we need a new sysfs to let usespace > enable it with THERMAL_TRIP_ENABLED? This should be drivers call not user space. Thanks, Srinivas > > Kai-Heng > > > Thanks, > > Srinivas > > > > > Kai-Heng > > > > > > > > Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com> > > > > > --- > > > > > drivers/thermal/thermal_core.c | 3 +++ > > > > > include/linux/thermal.h | 2 ++ > > > > > 2 files changed, 5 insertions(+) > > > > > > > > > > diff --git a/drivers/thermal/thermal_core.c > > > > > b/drivers/thermal/thermal_core.c > > > > > index c6d74bc1c90b..6561e3767529 100644 > > > > > --- a/drivers/thermal/thermal_core.c > > > > > +++ b/drivers/thermal/thermal_core.c > > > > > @@ -1477,6 +1477,9 @@ thermal_zone_device_register(const char > > > > > *type, int trips, int mask, > > > > > goto unregister; > > > > > } > > > > > > > > > > + if (tz->tzp && tz->tzp->userspace) > > > > > + thermal_zone_device_disable(tz); > > > > > + > > > > > mutex_lock(&thermal_list_lock); > > > > > list_add_tail(&tz->node, &thermal_tz_list); > > > > > mutex_unlock(&thermal_list_lock); > > > > > diff --git a/include/linux/thermal.h > > > > > b/include/linux/thermal.h > > > > > index d07ea27e72a9..e8e8fac78fc8 100644 > > > > > --- a/include/linux/thermal.h > > > > > +++ b/include/linux/thermal.h > > > > > @@ -247,6 +247,8 @@ struct thermal_zone_params { > > > > > */ > > > > > bool no_hwmon; > > > > > > > > > > + bool userspace; > > > > > + > > > > > int num_tbps; /* Number of tbp entries */ > > > > > struct thermal_bind_params *tbp; > > > > > > > > > > > > > > > > > > -- > > > > <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
> On Dec 1, 2020, at 02:13, Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com> wrote: [snipped] >>> What about creating an new callback >>> >>> enum thermal_trip_status { >>> THERMAL_TRIP_DISABLED = 0, >>> THERMAL_TRIP_ENABLED, >>> }; >>> >>> int get_trip_status(struct thermal_zone_device *, int trip, enum >>> thermal_trip_status *state); >>> >>> Then in >>> static void handle_thermal_trip(struct thermal_zone_device *tz, int >>> trip) >>> { >>> >>> /* before tz->ops->get_trip_temp(tz, trip, &trip_temp); */ >>> if (tz->ops->get_trip_status) { >>> enum thermal_trip_status *status; >>> >>> if (!tz->ops->get_trip_status(tz, trip, &status)) { >>> if (status == THERMAL_TRIP_DISABLED) >>> return; >>> } >>> } >>> ... >>> ... >>> >>> } >>> >>> >>> This callback will help the cases: >>> - Allows drivers to selectively disable certain trips during init >>> state >>> or system resume where there can be spikes or always. int340x >>> drivers >>> can disable always. >> >> This sounds really great. This is indeed can happen on system resume, >> before userspace process thaw. >> >>> - Still give options for drivers to handle critical trip even if >>> they >>> are bound to user space governors. User space process may be dead, >>> so >>> still allow kernel to process graceful shutdown >> >> To make the scenario happen, do we need a new sysfs to let usespace >> enable it with THERMAL_TRIP_ENABLED? > This should be drivers call not user space. Understood. So after thermal_zone_device_register(), the driver can decide to what to return on get_trip_temp(). Let me work on a new patch if there's no other concern. Kai-Heng > > Thanks, > Srinivas
On Tue, 2020-12-01 at 02:22 +0800, Kai-Heng Feng wrote: > > On Dec 1, 2020, at 02:13, Srinivas Pandruvada < > > srinivas.pandruvada@linux.intel.com> wrote: > > [snipped] > > > > > What about creating an new callback > > > > > > > > enum thermal_trip_status { > > > > THERMAL_TRIP_DISABLED = 0, > > > > THERMAL_TRIP_ENABLED, > > > > }; > > > > > > > > int get_trip_status(struct thermal_zone_device *, int trip, > > > > enum > > > > thermal_trip_status *state); > > > > > > > > Then in > > > > static void handle_thermal_trip(struct thermal_zone_device *tz, > > > > int > > > > trip) > > > > { > > > > > > > > /* before tz->ops->get_trip_temp(tz, trip, &trip_temp); */ > > > > if (tz->ops->get_trip_status) { > > > > enum thermal_trip_status *status; > > > > > > > > if (!tz->ops->get_trip_status(tz, trip, &status)) { > > > > if (status == THERMAL_TRIP_DISABLED) > > > > return; > > > > } > > > > } > > > > ... > > > > ... > > > > > > > > } > > > > > > > > > > > > This callback will help the cases: > > > > - Allows drivers to selectively disable certain trips during > > > > init > > > > state > > > > or system resume where there can be spikes or always. int340x > > > > drivers > > > > can disable always. > > > > > > This sounds really great. This is indeed can happen on system > > > resume, > > > before userspace process thaw. > > > > > > > - Still give options for drivers to handle critical trip even > > > > if > > > > they > > > > are bound to user space governors. User space process may be > > > > dead, > > > > so > > > > still allow kernel to process graceful shutdown > > > > > > To make the scenario happen, do we need a new sysfs to let > > > usespace > > > enable it with THERMAL_TRIP_ENABLED? > > This should be drivers call not user space. > > Understood. So after thermal_zone_device_register(), the driver can > decide to what to return on get_trip_temp(). get_trip_status() > Let me work on a new patch if there's no other concern. Better to wait for confirmation from Daniel and others. Thanks, Srinivas > > Kai-Heng > > > Thanks, > > Srinivas
> On Dec 1, 2020, at 02:39, Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com> wrote: > > On Tue, 2020-12-01 at 02:22 +0800, Kai-Heng Feng wrote: >>> On Dec 1, 2020, at 02:13, Srinivas Pandruvada < >>> srinivas.pandruvada@linux.intel.com> wrote: >> >> [snipped] >> >>>>> What about creating an new callback >>>>> >>>>> enum thermal_trip_status { >>>>> THERMAL_TRIP_DISABLED = 0, >>>>> THERMAL_TRIP_ENABLED, >>>>> }; >>>>> >>>>> int get_trip_status(struct thermal_zone_device *, int trip, >>>>> enum >>>>> thermal_trip_status *state); >>>>> >>>>> Then in >>>>> static void handle_thermal_trip(struct thermal_zone_device *tz, >>>>> int >>>>> trip) >>>>> { >>>>> >>>>> /* before tz->ops->get_trip_temp(tz, trip, &trip_temp); */ >>>>> if (tz->ops->get_trip_status) { >>>>> enum thermal_trip_status *status; >>>>> >>>>> if (!tz->ops->get_trip_status(tz, trip, &status)) { >>>>> if (status == THERMAL_TRIP_DISABLED) >>>>> return; >>>>> } >>>>> } >>>>> ... >>>>> ... >>>>> >>>>> } >>>>> >>>>> >>>>> This callback will help the cases: >>>>> - Allows drivers to selectively disable certain trips during >>>>> init >>>>> state >>>>> or system resume where there can be spikes or always. int340x >>>>> drivers >>>>> can disable always. >>>> >>>> This sounds really great. This is indeed can happen on system >>>> resume, >>>> before userspace process thaw. >>>> >>>>> - Still give options for drivers to handle critical trip even >>>>> if >>>>> they >>>>> are bound to user space governors. User space process may be >>>>> dead, >>>>> so >>>>> still allow kernel to process graceful shutdown >>>> >>>> To make the scenario happen, do we need a new sysfs to let >>>> usespace >>>> enable it with THERMAL_TRIP_ENABLED? >>> This should be drivers call not user space. >> >> Understood. So after thermal_zone_device_register(), the driver can >> decide to what to return on get_trip_temp(). > get_trip_status() > >> Let me work on a new patch if there's no other concern. > Better to wait for confirmation from Daniel and others. Daniel, Do you like Srinivas' proposed solution? I hope we can find a solution in upstream kernel soon. Kai-Heng > > Thanks, > Srinivas > >> >> Kai-Heng >> >>> Thanks, >>> Srinivas
On 07/12/2020 06:36, Kai-Heng Feng wrote: > > >> On Dec 1, 2020, at 02:39, Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com> wrote: >> >> On Tue, 2020-12-01 at 02:22 +0800, Kai-Heng Feng wrote: >>>> On Dec 1, 2020, at 02:13, Srinivas Pandruvada < >>>> srinivas.pandruvada@linux.intel.com> wrote: >>> >>> [snipped] >>> >>>>>> What about creating an new callback >>>>>> >>>>>> enum thermal_trip_status { >>>>>> THERMAL_TRIP_DISABLED = 0, >>>>>> THERMAL_TRIP_ENABLED, >>>>>> }; >>>>>> >>>>>> int get_trip_status(struct thermal_zone_device *, int trip, >>>>>> enum >>>>>> thermal_trip_status *state); >>>>>> >>>>>> Then in >>>>>> static void handle_thermal_trip(struct thermal_zone_device *tz, >>>>>> int >>>>>> trip) >>>>>> { >>>>>> >>>>>> /* before tz->ops->get_trip_temp(tz, trip, &trip_temp); */ >>>>>> if (tz->ops->get_trip_status) { >>>>>> enum thermal_trip_status *status; >>>>>> >>>>>> if (!tz->ops->get_trip_status(tz, trip, &status)) { >>>>>> if (status == THERMAL_TRIP_DISABLED) >>>>>> return; >>>>>> } >>>>>> } >>>>>> ... >>>>>> ... >>>>>> >>>>>> } >>>>>> >>>>>> >>>>>> This callback will help the cases: >>>>>> - Allows drivers to selectively disable certain trips during >>>>>> init >>>>>> state >>>>>> or system resume where there can be spikes or always. int340x >>>>>> drivers >>>>>> can disable always. >>>>> >>>>> This sounds really great. This is indeed can happen on system >>>>> resume, >>>>> before userspace process thaw. >>>>> >>>>>> - Still give options for drivers to handle critical trip even >>>>>> if >>>>>> they >>>>>> are bound to user space governors. User space process may be >>>>>> dead, >>>>>> so >>>>>> still allow kernel to process graceful shutdown >>>>> >>>>> To make the scenario happen, do we need a new sysfs to let >>>>> usespace >>>>> enable it with THERMAL_TRIP_ENABLED? >>>> This should be drivers call not user space. >>> >>> Understood. So after thermal_zone_device_register(), the driver can >>> decide to what to return on get_trip_temp(). >> get_trip_status() >> >>> Let me work on a new patch if there's no other concern. >> Better to wait for confirmation from Daniel and others. > > Daniel, > > Do you like Srinivas' proposed solution? > > I hope we can find a solution in upstream kernel soon. (just trying to figure out the full context) If the device is enumerated outside of a thermal zone, the sensor should not register in the thermal zone no ?
On Wed, 2020-12-09 at 10:30 +0100, Daniel Lezcano wrote: > On 07/12/2020 06:36, Kai-Heng Feng wrote: > > > > > On Dec 1, 2020, at 02:39, Srinivas Pandruvada < > > > srinivas.pandruvada@linux.intel.com> wrote: > > > > > > On Tue, 2020-12-01 at 02:22 +0800, Kai-Heng Feng wrote: > > > > > On Dec 1, 2020, at 02:13, Srinivas Pandruvada < > > > > > srinivas.pandruvada@linux.intel.com> wrote: > > > > > > > > [snipped] > > > > > > > > > > > What about creating an new callback > > > > > > > > > > > > > > enum thermal_trip_status { > > > > > > > THERMAL_TRIP_DISABLED = 0, > > > > > > > THERMAL_TRIP_ENABLED, > > > > > > > }; > > > > > > > > > > > > > > int get_trip_status(struct thermal_zone_device *, int > > > > > > > trip, > > > > > > > enum > > > > > > > thermal_trip_status *state); > > > > > > > > > > > > > > Then in > > > > > > > static void handle_thermal_trip(struct > > > > > > > thermal_zone_device *tz, > > > > > > > int > > > > > > > trip) > > > > > > > { > > > > > > > > > > > > > > /* before tz->ops->get_trip_temp(tz, trip, &trip_temp); > > > > > > > */ > > > > > > > if (tz->ops->get_trip_status) { > > > > > > > enum thermal_trip_status *status; > > > > > > > > > > > > > > if (!tz->ops->get_trip_status(tz, trip, &status)) { > > > > > > > if (status == THERMAL_TRIP_DISABLED) > > > > > > > return; > > > > > > > } > > > > > > > } > > > > > > > ... > > > > > > > ... > > > > > > > > > > > > > > } > > > > > > > > > > > > > > > > > > > > > This callback will help the cases: > > > > > > > - Allows drivers to selectively disable certain trips > > > > > > > during > > > > > > > init > > > > > > > state > > > > > > > or system resume where there can be spikes or always. > > > > > > > int340x > > > > > > > drivers > > > > > > > can disable always. > > > > > > > > > > > > This sounds really great. This is indeed can happen on > > > > > > system > > > > > > resume, > > > > > > before userspace process thaw. > > > > > > > > > > > > > - Still give options for drivers to handle critical trip > > > > > > > even > > > > > > > if > > > > > > > they > > > > > > > are bound to user space governors. User space process may > > > > > > > be > > > > > > > dead, > > > > > > > so > > > > > > > still allow kernel to process graceful shutdown > > > > > > > > > > > > To make the scenario happen, do we need a new sysfs to let > > > > > > usespace > > > > > > enable it with THERMAL_TRIP_ENABLED? > > > > > This should be drivers call not user space. > > > > > > > > Understood. So after thermal_zone_device_register(), the driver > > > > can > > > > decide to what to return on get_trip_temp(). > > > get_trip_status() > > > > > > > Let me work on a new patch if there's no other concern. > > > Better to wait for confirmation from Daniel and others. > > > > Daniel, > > > > Do you like Srinivas' proposed solution? > > > > I hope we can find a solution in upstream kernel soon. > > (just trying to figure out the full context) > > If the device is enumerated outside of a thermal zone, the sensor > should > not register in the thermal zone no ? Other trips are fine, so sensor registry is still required for passive and active control. We just need to ignore critical trip. These table are tested by OEM on Windows, where critical trip doesn't result in immediate shutdown. Thanks, Srinivas > > >
diff --git a/drivers/thermal/thermal_core.c b/drivers/thermal/thermal_core.c index c6d74bc1c90b..6561e3767529 100644 --- a/drivers/thermal/thermal_core.c +++ b/drivers/thermal/thermal_core.c @@ -1477,6 +1477,9 @@ thermal_zone_device_register(const char *type, int trips, int mask, goto unregister; } + if (tz->tzp && tz->tzp->userspace) + thermal_zone_device_disable(tz); + mutex_lock(&thermal_list_lock); list_add_tail(&tz->node, &thermal_tz_list); mutex_unlock(&thermal_list_lock); diff --git a/include/linux/thermal.h b/include/linux/thermal.h index d07ea27e72a9..e8e8fac78fc8 100644 --- a/include/linux/thermal.h +++ b/include/linux/thermal.h @@ -247,6 +247,8 @@ struct thermal_zone_params { */ bool no_hwmon; + bool userspace; + int num_tbps; /* Number of tbp entries */ struct thermal_bind_params *tbp;
We are seeing thermal shutdown on Intel based mobile workstations, the shutdown happens during the first trip handle in thermal_zone_device_register(): kernel: thermal thermal_zone15: critical temperature reached (101 C), shutting down However, we shouldn't do a thermal shutdown here, since 1) We may want to use a dedicated daemon, Intel's thermald in this case, to handle thermal shutdown. 2) For ACPI based system, _CRT doesn't mean shutdown unless it's inside ThermalZone. ACPI Spec, 11.4.4 _CRT (Critical Temperature): "... If this object it present under a device, the device’s driver evaluates this object to determine the device’s critical cooling temperature trip point. This value may then be used by the device’s driver to program an internal device temperature sensor trip point." So a "critical trip" here merely means we should take a more aggressive cooling method. So add an indication to let thermal core know it should leave thermal device to userspace to handle. Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com> --- drivers/thermal/thermal_core.c | 3 +++ include/linux/thermal.h | 2 ++ 2 files changed, 5 insertions(+)