Ubuntu – How to troubleshoot CPU temperatures in Ubuntu

cputemperatureUbuntu

I have some questions concerning CPU temperature on my Ubuntu 8.10.

  1. My laptop shuts itself down, which I guess is due to high CPU temperature. I want to confirm this by looking at some system log file that records the reason that the system shuts itself down. Is there any such syslog file? Where is it stored?

  2. Also I have installed libsensors, which gives me different temperatures

    $ sensors  
    acpitz-virtual-0  
    Adapter: Virtual device  
    temp1:       +49.0°C  (crit = +97.0°C)                    
    
    k8temp-pci-00c3  
    Adapter: PCI adapter  
    Core0 Temp:  +57.0°C 
    

    What do "acpitz-virtual-0" and "k8temp-pci-00c3" mean? As well as the meaning of "temp1" and "Core0 Temp"? Are the two temperatures both CPU temperatures?

    Is the temperature given by

    acpi -t
    

    another different measure of CPU temperature?

  3. I also wonder what you will do if the CPU temperature is exceeding some limit that you find is dangerous? I have also installed Computer Temperature Monitor (computertemp), which allows me to set up a limit temperature for alarm as well as command to execute when the limit is reached. So what command will you issue or things will you do when the temperature is exceeding the set limit to protect your laptop, instead of letting it shuts itself down?

Best Answer

On Ubuntu you will have /var/log/pm-*.log as well as the usual syslog.

acpitz-virtual-0 is an unhelpful label for an ACPI thermal zone, probably from an ACPI table at runtime.

k8temp* comes partly from /etc/sensors3.conf and partly from where the system found the chip. You probably have a default ABit configuration that looks like:

chip "k8temp-*"

   label temp1 "Core0 Temp"
   label temp2 "Core0 Temp"
   label temp3 "Core1 Temp"
   label temp4 "Core1 Temp"

If it's in fact an ABit system board you should check the BIOS as it may have have better descriptions.

Regarding acpi -t, sensors(1) is checking both acpi and hardware devices it knows about. In a perfect world sensors(1) would report a superset of what acpi is able to report.

It really shouldn't be possible to overheat a laptop unless it is operated in a rather hot environment. It's more likely that the configuration file or BIOS settings are off, or perhaps some filters need cleaning. If the notebook heatsink was installed using thermal grease, that's known to not age well. (However, thermal grease is unlikely to have been used for original production.) You might be able to regrease it or use a modern thermal interface pad. Don't remove the heatsink unless you are prepared to throw away the old thermal interface and install a new one.

Related Question