CPU hardware errors in Ubuntu 17.04

Question

Can anyone explain to me what these error messages I got when I looked in dmesg? I am new to Ubuntu and to the Linux World.

[ 7.802351] CPU4: Core temperature above threshold, cpu clock throttled (total events = 1)
[ 7.802352] CPU0: Core temperature above threshold, cpu clock throttled (total events = 1)
[ 7.802353] CPU5: Package temperature above threshold, cpu clock throttled (total events = 1)
[ 7.802354] CPU0: Package temperature above threshold, cpu clock throttled (total events = 1)
[ 7.802354] CPU4: Package temperature above threshold, cpu clock throttled (total events = 1)
[ 7.802356] CPU1: Package temperature above threshold, cpu clock throttled (total events = 1)
[ 7.802356] mce: [Hardware Error]: Machine check events logged
[ 7.802362] mce: [Hardware Error]: CPU 4: Machine Check: 0 Bank 128: 00000000880a0003
[ 7.802363] mce: [Hardware Error]: TSC 99561677c
[ 7.802385] mce: [Hardware Error]: PROCESSOR 0:506e3 TIME 1501537538 SOCKET 0 APIC 1 microcode ba
[ 7.802387] mce: [Hardware Error]: CPU 0: Machine Check: 0 Bank 128: 00000000880a0003
[ 7.802387] mce: [Hardware Error]: TSC 995616be4
[ 7.802388] mce: [Hardware Error]: PROCESSOR 0:506e3 TIME 1501537538 SOCKET 0 APIC 0 microcode ba
[ 7.802389] CPU2: Package temperature above threshold, cpu clock throttled (total events = 1)
[ 7.802390] CPU6: Package temperature above threshold, cpu clock throttled (total events = 1)
[ 7.802391] CPU3: Package temperature above threshold, cpu clock throttled (total events = 1)
[ 7.802392] CPU7: Package temperature above threshold, cpu clock throttled (total events = 1)
[ 7.826359] CPU4: Core temperature/speed normal
[ 7.826359] CPU0: Core temperature/speed normal
[ 7.826360] CPU2: Package temperature/speed normal
[ 7.826361] CPU6: Package temperature/speed normal
[ 7.826361] CPU0: Package temperature/speed normal
[ 7.826362] CPU4: Package temperature/speed normal
[ 7.826363] mce: [Hardware Error]: Machine check events logged
[ 7.826367] mce: [Hardware Error]: CPU 4: Machine Check: 0 Bank 128: 00000000880b0002
[ 7.826368] mce: [Hardware Error]: TSC 99916f004
[ 7.826369] mce: [Hardware Error]: PROCESSOR 0:506e3 TIME 1501537538 SOCKET 0 APIC 1 microcode ba
[ 7.826369] mce: [Hardware Error]: CPU 0: Machine Check: 0 Bank 128: 00000000880b0002
[ 7.826370] mce: [Hardware Error]: TSC 99916f2ca
[ 7.826370] mce: [Hardware Error]: PROCESSOR 0:506e3 TIME 1501537538 SOCKET 0 APIC 0 microcode ba
[ 7.826400] CPU1: Package temperature/speed normal
[ 7.826401] CPU5: Package temperature/speed normal
[ 7.826402] CPU3: Package temperature/speed normal
[ 7.826402] CPU7: Package temperature/speed normal
[ 467.922330] CPU4: Core temperature above threshold, cpu clock throttled (total events = 73)
[ 467.922331] CPU0: Core temperature above threshold, cpu clock throttled (total events = 73)
[ 467.922332] CPU7: Package temperature above threshold, cpu clock throttled (total events = 86)
[ 467.922333] CPU3: Package temperature above threshold, cpu clock throttled 

I am running Ubuntu 17.04 with 4.10.0-29-generic kernal

Solution

The CPU is overheating and going into MCE (machine check events)... ie: it's crashing. If you don't see other temperature related events in syslog, it's probably because your CPU cooler/fan/thermal pipe/thermal paste isn't doing the job.

• Check syslog with this terminal command...

  grep -i -e temp -e therm /var/log/syslog*
  

• If the machine is very dirty/dusty, that could play a major role in the machine overheating. Clean it out.

• If your machine has Intel processors, make sure that intel-microcode is installed.

  sudo apt-get update
  sudo apt-get install intel-microcode
  reboot
  

• Install thermald to try and control the temperature.

  sudo apt-get update
  sudo apt-get install thermald
  reboot
  

• Check your BIOS version. Enter your BIOS at power on time, and note the version #. Go to the manufacturer's web site with the make/model of your computer. Go to the support/downloads section, and look to see if there's a newer BIOS.

• Lastly, and very likely, if this is an older machine, the thermal compound that sits between the processors and its heat pipe/fan cooler needs to be re-applied. This requires some technical experience.

Comments

[Rolf Schumacher]

or it is a software bug I do have the same in my log of a new Lenovo T480S see bugs.launchpad.net/ubuntu/+source/linux/+bug/1781924

[gene_wood]

What impact does the intel-micorocode have on the temperature?

[Boris Hamanov]

@gene_wood as I mention, it's probably a dust, fan, or thermal compound problem, and checking microcode is just another step in helping to diagnose the problem remotely. CPU's running old microcode can cause various problems. Seconds to check for it.