Bug 218652

Summary:	acpitz sensors regression on Linux 6.7+ on Framework 16
Product:	ACPI	Reporter:	Quentin Smith (qsmith)
Component:	Power-Thermal	Assignee:	acpi_power-thermal
Status:	RESOLVED DUPLICATE
Severity:	normal	CC:	mario.limonciello, rjw
Priority:	P3
Hardware:	All
OS:	Linux
Kernel Version:		Subsystem:
Regression:	Yes	Bisected commit-id:
Attachments:	possible patch

Description Quentin Smith 2024-03-28 04:47:57 UTC

After upgrading from 6.1.78 to 6.8.1, I discovered that the `acpitz-*` sensors had gone missing. I believe that was caused by this commit:

https://github.com/torvalds/linux/commit/f04256a8f7de2c13619b636cec1109e596804229

because the kernel is now logging at boot:

```
ACPI: thermal: [Firmware Bug]: Invalid critical threshold (-274000)
ACPI: thermal: [Firmware Bug]: No valid trip points!
ACPI: thermal: [Firmware Bug]: Invalid critical threshold (-274000)
ACPI: thermal: [Firmware Bug]: No valid trip points!
ACPI: thermal: [Firmware Bug]: Invalid critical threshold (-274000)
ACPI: thermal: [Firmware Bug]: No valid trip points!
ACPI: thermal: [Firmware Bug]: Invalid critical threshold (-274000)
ACPI: thermal: [Firmware Bug]: No valid trip points!
```

If I’m reading the laptop's DSDT correctly, the `_HOT` threshold is reported as 0x1218 (463.2 degrees) and the `_CRT` threshold is 0x12E0 (483.2 degrees). The old code would always create the thermal zones, but after that commit the kernel ignores any threshold above 448 degrees as implausibly high, and if there were no valid thresholds it ignores the whole thermal zone.

Even if there are no plausible thresholds, the thermal zone should still be created so the sensor values can be read.

I reported this in the Framework forum at https://community.frame.work/t/firmware-bug-acpi-table-error-causes-missing-sensors-on-linux-6-7/47767, but Mario Limonciello from AMD asked me to file an upstream bug since the commit message says "No intentional functional impact."

Comment 1 Artem S. Tashkinov 2024-03-29 09:04:17 UTC

> I believe that was caused by this commit:

Can you confirm reverting this commit fixes the issue for you?

Comment 2 Mario Limonciello (AMD) 2024-03-29 15:17:47 UTC

Created attachment 306058 [details]
possible patch

Can you see if this patch helps?

Comment 3 Mario Limonciello (AMD) 2024-03-30 14:16:08 UTC


*** This bug has been marked as a duplicate of bug 218586 ***