Re: [PATCH 2/3] thermal/drivers/intel: Use generic trip points for processor_thermal_device_pci
From: srinivas pandruvada
Date: Mon Jan 23 2023 - 14:32:00 EST
Hi Daniel,
On Mon, 2023-01-23 at 19:02 +0100, Daniel Lezcano wrote:
>
> Hi Srinivas,
>
>
> On 18/01/2023 20:09, srinivas pandruvada wrote:
> > On Wed, 2023-01-18 at 19:16 +0100, Daniel Lezcano wrote:
> > > The thermal framework gives the possibility to register the trip
> > > points with the thermal zone. When that is done, no get_trip_*
> > > ops
> > > are
> > > needed and they can be removed.
> > >
> > > Convert ops content logic into generic trip points and register
> > > them
> > > with the
> > > thermal zone.
> > >
> > In this scheme is the assumption is that trip point temperature
> > never
> > changes? If firmware updated the trip temperature, what needs to be
> > done?
>
> I'm a bit confused about the situation where the firmware can change
> the
> trip point in the back of the OSPM.
>
> Does the firmware send a notification about the trip change? Or does
> it
> assume the OSPM will be reading the trip point while
> monitoring/polling
> the thermal zone ?
Firmware sends an ACPI notification. For example INT3403.
https://elixir.bootlin.com/linux/latest/C/ident/INT3403_PERF_TRIP_POINT_CHANGED
>
> Is the question for this particular driver?
This PCH driver trips are not changed by firmware hence we don't have
to worry about here.
Thanks,
Srinivas
>
> If the trip point is changed by the userspace (via sysfs),
> thermal_zone_set_trip() is used which in turn changes the thermal
> trip
> temperature directly in the generic structure and then calls the back
> set_trip_temp.
>