[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAJZ5v0heL_wa=bimMDijn-x6G0SxsMf=yGhKZAe282P5+h2O3w@mail.gmail.com>
Date: Thu, 14 Nov 2024 12:53:11 +0100
From: "Rafael J. Wysocki" <rafael@...nel.org>
To: "lihuisong (C)" <lihuisong@...wei.com>
Cc: "Rafael J. Wysocki" <rafael@...nel.org>, linux-pm@...r.kernel.org, linux-acpi@...r.kernel.org,
linux-kernel@...r.kernel.org, rui.zhang@...el.com, liuyonglong@...wei.com,
zhanjie9@...ilicon.com, zhenglifeng1@...wei.com
Subject: Re: [PATCH v2] ACPI: thermal: Support for linking devices associated
with the thermal zone
Hi,
On Thu, Nov 14, 2024 at 9:37 AM lihuisong (C) <lihuisong@...wei.com> wrote:
>
> Hi Rafael,
>
> 在 2024/11/13 17:26, Rafael J. Wysocki 写道:
> > On Wed, Nov 13, 2024 at 10:07 AM Huisong Li <lihuisong@...wei.com> wrote:
> >> There are many 'cdevX' files which link cooling devices under
> >> '/sys/class/thermal/thermal_zoneX/'. These devices contain active cooling
> >> devices and passive cooling devices. And user cann't directly know which
> >> devices temperature is represented by the thermal zone.
> >>
> >> However, ACPI spec provides a '_TZD' object which evaluates to a package
> >> of device names. Each name corresponds to a device in the ACPI namespace
> >> that is associated with the thermal zone. The temperature reported by the
> >> thermal zone is roughly correspondent to that of each of the devices.
> >>
> >> User can get all devices a thermal zone measured by the 'measures'
> >> directory under the thermal zone device.
> > Well, that's kind of clear, but what exactly is the use case? Why
> > does the user need to know that?
> IMO, this makes thermal zone information more friendly.
> For instance, user can directly know the temperature of CPUs or other
> devices is roughly represented by which thermal zone.
> This may offer the convenience for further usersapce application.
>
> BTW, the '_TZD' method is similar to the '_PMD' in acpi power meter.
> Since ACPI spec provides them, they should also have a role in their
> existence.
So there is no specific use case, but it is possible that somebody may
want to use this information, IIUC.
Well, let's defer making kernel changes until there is a user
wanting/needing this information. Then we'll decide how to expose it.
For one, I'm not convinced that exposing it under the ACPI
representation of a thermal zone is going to be really useful.
Powered by blists - more mailing lists