[xubuntu-users] troubleshoot mcelog Machine Check Exception messages

[Rob Ward]

On 24/10/14 05:02, scar wrote:
> hi i have a computer that has been hard rebooting over and over... i
> thought it was a kernel/software problem as the machine froze while
> installing a kernel update.  i took the opportunity to wipe it clean and
> install xubuntu on it.  the installation went well and i was able to
> restore my backup, but now the machine is freezing again during boot-up.
>   this time, however, it is not rebooting automatically and has provided
> a Machine Check Exception output, which i ran through mcelog on my laptop:
>
> $ sudo mcelog --ascii
> CPU 0: Machine Check Exception: 4 Bank 4: b200001000010c0f
> CPU 0: Machine Check Exception: 4 Bank 4: b200001000010c0f
> TSC 1295c993c5
> TSC 1295c993c5
> PROCESSOR 2:40ff2 TIME 1414042838 SOCKET 0 APIC 0 microcode 0
> Hardware event. This is not a software error.
> CPU 0 BANK 4 TSC 1295c993c5
> TIME 1414042838 Wed Oct 22 22:40:38 2014
> STATUS b200001000010c0f MCGSTATUS 4
> CPUID Vendor AMD Family 15 Model 79
> PROCESSOR 2:40ff2 TIME 1414042838 SOCKET 0 APIC 0 microcode 0
> $
>
> i had to type in the 3 lines manually and then mcelog produced the
> output starting with "Hardware event...." (it echoed the first two lines
> after i typed them in).  so i'm trying to nail down what hardware is
> faulty so i can fix it.  all i've been able to find out there is
> 'contact your hardware vendor'... so is that my motherboard
> manufacturer?  or the CPU manufacturer?  or what?
>
> i did run memtest overnight for 12 hours, 13+ passes and no errors, so
> it seems the RAM is OK....
>
>
I had similar problems until I had thrown away a cheap Hard Drive power 
cable splitter for the power supplies to my SATA drives and gave them 
own higher quality cabling from another separate lead off the power 
supply.  (I cut off a standard IDE drive power plug and spliced some 
proper SATA cabling from a wrecked power supply onto it). Booting 
problems disappeared completely.

Cheers, Rob