6.6.3 Hardware Error

[fazzeren]

Oct 30 10:30:34 Tower kernel: mce: [Hardware Error]: Machine check events logged
Oct 30 10:30:34 Tower kernel: mce: [Hardware Error]: CPU 0: Machine Check: 0 Bank 17: ee2000000004017a
Oct 30 10:30:34 Tower kernel: mce: [Hardware Error]: TSC 0 ADDR 5f000000 MISC 4f00031e0000086 
Oct 30 10:30:34 Tower kernel: mce: [Hardware Error]: PROCESSOR 0:306f2 TIME 1540909820 SOCKET 0 APIC 0 microcode 3d
Oct 30 10:30:34 Tower kernel: mce: [Hardware Error]: Machine check events logged
Oct 30 10:30:34 Tower kernel: mce: [Hardware Error]: CPU 0: Machine Check: 0 Bank 18: ee2000000004017a
Oct 30 10:30:34 Tower kernel: mce: [Hardware Error]: TSC 0 ADDR 5f100000 MISC 44f00031e0000086 
Oct 30 10:30:34 Tower kernel: mce: [Hardware Error]: PROCESSOR 0:306f2 TIME 1540909820 SOCKET 0 APIC 0 microcode 3d
Oct 30 10:30:34 Tower kernel: mce: [Hardware Error]: CPU 0: Machine Check: 0 Bank 19: ee2000000004017a
Oct 30 10:30:34 Tower kernel: mce: [Hardware Error]: TSC 0 ADDR 5f100080 MISC 84f00031e0000086 
Oct 30 10:30:34 Tower kernel: mce: [Hardware Error]: PROCESSOR 0:306f2 TIME 1540909820 SOCKET 0 APIC 0 microcode 3d

tower-diagnostics-20181101-1615.zip

[Squid]

It's happening super early in the boot process, so unless the event log in the BIOS (if it has one) shows anything, then you probably won't be able to determine what the cause is or if its a bios bug etc.

 

IE: if everything is stable, and the mce's don't happen after array start, then don't worry excessively about it.  Alternatively, you can email Intel support and ask them what it means.

[John_M]

It appears to be happening during the CPU initialisation. That's the second one I've seen today! Looking for a BIOS update won't hurt.

[jonp]

Machine checks can occur for lots of different reasons.  On my system, I started seeing them at one point when my heatsink loosened contact with the CPU and I would run games in a VM.  This would obviously cause the CPU temp to rise and it would also fire off these machine check errors occasionally.  Simply tightening that down fixed the issue for me.  MCEs definitely point to a hardware-specific problem though (most of the time).