Critical Temperature Reached, H3 Overheating

[Achim Scheidl]

Hi,
I have problems with NanoPi NEO on Allwinner H3 due to sudden sporadic Linux kernel overheating shut down. I'm using the Linux kernel 4.14 on Debian Trixie.
The syslog shows:

Quote

kernel: thermal thermal_zone0: critical temperature reached (103 C), shutting down

The hours before, the kernel shows above 80°C by:

Quote

cat /sys/class/thermal/thermal_zone0/temp

But there is not much load on it. When looking to "top" I see less then 10% as load average and more then 90% idle. "iotop" shows all processes <1%. And when measuring the chips, all are cool at below 40°C

It looks like the CPU suddenly get into a hot-state above 80°C, because it stays there for hours independent from the process load. Even after a reboot command, the system comes up again and is still above unrealistic 80°C even with no load.
A power cycle temporarily fixes the problem. The system comes up with about 50°C and slowly cools down to about 30°C for the next 1 to 3 weeks. Then it suddenly becomes hot again until shutting down.

Any experiences with this, what helps? Is there a kernel patch available?

 Thanks

[Werner]

14 hours ago, Achim Scheidl said:

e Linux kernel 4.14 on Debian Trixie

Where did you get an image with that old kernel?

[Achim Scheidl]

Kernel 4.14 is still part of the image delivered by FriendlyElec with NanoPi NEO.
What you probably want to say: First upgrade to a more recent kernel, maybe this fixes the problem.... 

[Werner]

The image you got is probably fake/non genuine. For such we cannot provide support. Try at the place where you got the image from.

https://docs.armbian.com/User-Guide_Getting-Started/#how-to-check-download-authenticity

 

 

[going]

01.10.2024 в 16:47, Achim Scheidl сказал:

A power cycle temporarily fixes the problem. The system comes up with about 50°C and slowly cools down to about 30°C for the next 1 to 3 weeks. Then it suddenly becomes hot again until shutting down.

This is an old and well-known problem for that time.
The SOC has a controller (ARISC architecture).

The code of that time randomly started the controller at the maximum frequency, for example, when the OS was turned off with the sudo poweroff command.
If you did not pull the plug out of the socket, then the processor chip began to warm up more than when the OS was running.

 

You need to update the u-boot and kernel to the current versions.

[Achim Scheidl]

Wow, this is very important information for me, I will try to fix it with the proposals you made - which will take some time....

Thanks very much für this background info