LinuxQuestions.org - [SOLVED] CPU temperature error ?

- Linux - Hardware (https://www.linuxquestions.org/questions/linux-hardware-18/)

- - CPU temperature error ? (https://www.linuxquestions.org/questions/linux-hardware-18/cpu-temperature-error-4175657703/)

CPU temperature error ?

I have an old pc that i still use. issue is that sometimes i couldn't boot into the OS(windows XP), it just stalls at bios boot screen, when i do boot into the OS, it freezes frequently (BSOD).

so i boot the pc with slackware installed on an USB drive. here's the output of dmesg:

Code:

possible error:  cpu temp 

[23632.457533] CPU0: Core temperature above threshold, cpu clock throttled (total events = 90096)

[23632.457544] CPU1: Core temperature above threshold, cpu clock throttled (total events = 90085)

[23632.458495] CPU1: Core temperature/speed normal

[23632.458498] CPU0: Core temperature/speed normal

[23849.997017] mce: [Hardware Error]: Machine check events logged

[23942.236038] CPU0: Core temperature above threshold, cpu clock throttled (total events = 92230)

[23942.236050] CPU1: Core temperature above threshold, cpu clock throttled (total events = 92220)

[23942.237289] CPU0: Core temperature/speed normal

[23942.237292] CPU1: Core temperature/speed normal

[23999.997019] mce: [Hardware Error]: Machine check events logged

it seems to tell me the CPU is overheated, the motherboard is a P4M900-M4, i powered it up with side door open, both ps fan and cpu fan are running.

what to do to fix the issue?

a cleanup may help (with a hoover?)

Quote:

Originally Posted by xiongnu (Post 6016768)

i powered it up with side door open, both ps fan and cpu fan are running.

Mostly the internal fans work best with the case closed, they're not big enough to move the air in the whole room.
Or, if you really are overheating: punt a BIG table ventilator beside the case and let IT blow into the cabinet (with the side open, of course, in that case).

And as pan64 already said: clean both ventilators from dust with a vacuumcleaner (at the university we always had one at hand for when we opened a cabinet). Do that too for the ventilator in the power supply (if any, of course).

Have you checked the thermal paste between the processor and heatsink? If not, check that - it might be worn out.

Motherboards and power supplies from the period that P4M900-M4 was made were prone to a plague of electrolytic capacitors that would not last. If you have any doubt about how to check for obviously bad caps, visit badcaps.net. Any misbehavior is possible with failed caps, but especially BSODs. An overheat indication could be real, e.g. from overvoltage caused by failed or failing cap(s), or sensor misread from out of spec circuit caused by failing or failed cap(s).

Those VIA chipsets were no jewels of competence either. :p

Quote:

Originally Posted by mrmazda

Those VIA chipsets were no jewels of competence either.

+1 & Amen to that! Don't start me…

Anyhow another effect of bad caps is varying leakage. The things would momentarily nearly short out, making for noise on the power supply line (5V or 3.3V usually. It was often possible to actually hear bad caps. If you left something for a month, and powered it up in cold weather, you could hear them for a while. Then they would get warm, the leakage would diminish, and the noise would gradually vanish. People spent a long time trying to find suitable electrolytes for capacitors after PCBs were banned, or at least, engineers refused to specify them.
/tangent
I find that funny,and sad actually. People doing military contracts to kill loads of people refusing to harm the earth by using PCBs in the capacitors.
/end tangent