Bug 219090

Summary:	[Regression] non-stop ACPI ([\_SB.PC00.LPCB.HEC.TSR1]) and thermal zone error
Product:	ACPI	Reporter:	Tom Yan (tom.ty89)
Component:	Power-Thermal	Assignee:	acpi_power-thermal
Status:	RESOLVED PATCH_ALREADY_AVAILABLE
Severity:	high	CC:	regressions
Priority:	P3
Hardware:	Intel
OS:	Linux
Kernel Version:	6.10	Subsystem:
Regression:	Yes	Bisected commit-id:

Description Tom Yan 2024-07-24 05:39:06 UTC

Up til 6.9.10, I would see these errors every boot *twice*:

ACPI BIOS Error (bug): Could not resolve symbol [\_SB.PC00.LPCB.HEC.TSR1], AE_NOT_FOUND (20230628/psargs-330)
ACPI Error: Aborting method \_SB.PC00.LPCB.H_EC.SEN1._TMP due to previous error (AE_NOT_FOUND) (20230628/psparse-529)

(they do get occasionally repeated again)

and these *once*:

thermal thermal_zone2: failed to read out thermal zone (-5)
thermal thermal_zone7: failed to read out thermal zone (-61)

But since the distro I use (Arch) has moved on to 6.10, these are spammed *non-stop*:

ACPI BIOS Error (bug): Could not resolve symbol [\_SB.PC00.LPCB.HEC.TSR1], AE_NOT_FOUND (20240322/psargs-330)
ACPI Error: Aborting method \_SB.PC00.LPCB.H_EC.SEN1._TMP due to previous error (AE_NOT_FOUND) (20240322/psparse-529)
thermal thermal_zone1: failed to read out thermal zone (-5)


in the frequency of *three to four times per second*.

For the record, this is now consecutively repeated 5 times every boot as well:

thermal thermal_zone7: failed to read out thermal zone (-61)

But at least this one would stop.

Comment 1 Artem S. Tashkinov 2024-07-24 17:06:34 UTC

Would be great if you performed git bisect:

https://docs.kernel.org/admin-guide/bug-bisect.html

Comment 2 The Linux kernel's regression tracker (Thorsten Leemhuis) 2024-07-25 05:51:55 UTC

(In reply to Artem S. Tashkinov from comment #1)
> Would be great if you performed git bisect:

No need to, the problem is known[1] and a fix heading towards mainline:
https://lore.kernel.org/linux-pm/2348857.ElGaqSPkdT@rjwysocki.net/

[1] a quick search on lore would have found existing reports and breadcrumbs to the fix; Artem, it IMHO would be good if you could do a quick search before asking people to bisect, as that is a huge effort for many people

Comment 3 Artem S. Tashkinov 2024-07-25 14:46:00 UTC

(In reply to The Linux kernel's regression tracker (Thorsten Leemhuis) from comment #2)
> Artem, it IMHO would be good if you could do a quick search
> before asking people to bisect, as that is a huge effort for many people

Searching LKML is anything but easy due to an insane amount of posts.

And the message you've cited needs to be carefully read and comprehended to realize that it's indeed connected to this bug report.

Would be great if Linux had two mailing lists: one solely for patches (and patches discussion) and the other one for any other discussions (regressions, ideas, talks) but I can only dream.