Hi again,
On Sat, Mar 22, 2025 at 02:29:37PM +0100, Salvatore Bonaccorso wrote:
> Hi,
>
> On Mon, Feb 24, 2025 at 06:50:59PM +0100, matte.mb2006.9...@gmail.com wrote:
> > On Fri, 21 Feb 2025 08:55:08 +0100 Salvatore Bonaccorso
> > <car...@debian.org> wrote:
> > > Control: tags -1 + moreinfo
> > >
> > > Hi,
> > >
> > > On Sat, Sep 07, 2024 at 07:04:36PM +0200,
> > matte.mb2006.9...@gmail.com wrote:
> > > > Source: src:linux
> > > > Version: 6.1.0-25-amd64
> > > > Severity: minor
> > > >
> > > > Dear Maintainer,
> > > >
> > > > In the kernel logs I found the following log: "thermal
> > thermal_zone8:
> > > > failed to read out thermal zone (-61)", the only problem I find is
> > that
> > > > power-profiles-daemon doesn't set the performance mode because it
> > says
> > > > that the temperature is too high even if it is not true (bug I
> > already
> > > > reported).
> > >
> > > Can you verify this against the most current kernel in the 6.1.y
> > > stable series? (6.1.128-1). Does the problem shows up as well with a
> > > current kernel from the upper suites ideally 6.12.15-1 to test or
> > > (6.13.3-1~exp1 from experimental)? Is the problem exposed there as
> > > well?
> > >
> > > In the case of yes, can you provide the kernel log showing the
> > problem
> > > and could you report it upstream to the linux-pm developers?
> > >
> > > Regards,
> > > Salvatore
> > >
> > >
> >
> > Hi,
> > the problem goes away with kernel 6.13.4-1~exp1 and the latest
> > firmware-iwlwifi (sid).
> > The power-profiles-daemon problem remains though.
>
> Thanks for reporting back. If you have enough spare cycles you might
> try to verify as well 6.13 mainline, if its fixed there then do a
> bisect of the upstream changes between 6.1 and 6.13.
>
> I would suggest to first narrow down more the upstream versions by
> fetching the kernel revisions from the snapshots service. Once you
> have a close enought range of upstream versions then do the bisection.
>
> if you are able to determine the fix and it's missing in 6.1.y we
> might try to prod upstream to have a backport to 6.1.y as well, but we
> first need to get an idea about the fix.
Might be this one: a8a261774466 ("thermal: core: Call
monitor_thermal_zone() if zone temperature is invalid") in which case
the unstable kernel should as well do not expose anymore the log
error.
Regards,
Salvatore
