mce Hardware Error Machine check events logged -

botzko · 06-26-2014, 08:21 AM

Hello,

I have acer apspire v3-772G with Slackware 14.1 on it. Today I found this in my dmesg and /var/log/messages

Code:

Jun 26 14:24:51 darkstar kernel: [  300.359668] mce: [Hardware Error]: Machine check events logged

I checked /var/log/messages and found this error start on Jun 18 after upgrade to kernel 3.15.1. This log message appear every time 5 minutes after boot. I execute mcelog and here is the output:

Code:

HARDWARE ERROR. This is *NOT* a software problem!
Please contact your hardware vendor
MCE 0
CPU 0 BANK 8 
MISC 38a0000086 ADDR ffb06d80 
TIME 1403785936 Thu Jun 26 15:32:16 2014
MCG status:
MCi status:
Uncorrected error
MCi_MISC register valid
MCi_ADDR register valid
Processor context corrupt
MCA: corrected filtering (some unreported errors in same region)
Generic CACHE Level-2 Generic Error
STATUS ae0000000040110a MCGSTATUS 0
MCGCAP c09 APICID 0 SOCKETID 0 
CPUID Vendor Intel Family 6 Model 60

Can someone explain me what's failing (and what is the exact problem)? I can not find other clues for hardware problems.

OS: Slackware 14.1 64bit
Kernel: Linux darkstar 3.15.1-btz #2 SMP Thu Jun 19 11:01:39 EEST 2014 x86_64 Intel(R) Core(TM) i7-4702MQ CPU @ 2.20GHz GenuineIntel GNU/Linux
Laptop: Acer Apspire V3-772G

smallpond · 06-26-2014, 04:04 PM

Quote:

Generic CACHE Level-2 Generic Error

means an error was detected in the CPU L2 cache. This means either the cache has bad bits or the BIOS has not set up the CPU correctly. Is this the original CPU for the laptop?

botzko · 06-27-2014, 02:04 AM

Yes it is the original CPU. I'm using microcode (release date 2014-05-23) but the errors are before its installation.

Code:

# dmesg | grep microcode
[   39.001830] microcode: CPU0 sig=0x306c3, pf=0x10, revision=0x16
[   39.001853] microcode: CPU1 sig=0x306c3, pf=0x10, revision=0x16
[   39.001876] microcode: CPU2 sig=0x306c3, pf=0x10, revision=0x16
[   39.001888] microcode: CPU3 sig=0x306c3, pf=0x10, revision=0x16
[   39.001900] microcode: CPU4 sig=0x306c3, pf=0x10, revision=0x16
[   39.001910] microcode: CPU5 sig=0x306c3, pf=0x10, revision=0x16
[   39.001931] microcode: CPU6 sig=0x306c3, pf=0x10, revision=0x16
[   39.001942] microcode: CPU7 sig=0x306c3, pf=0x10, revision=0x16
[   39.002001] microcode: Microcode Update Driver: v2.00 <tigran@aivazian.fsnet.co.uk>, Peter Oruba
[   39.002264] microcode: CPU0 sig=0x306c3, pf=0x10, revision=0x16
[   39.004256] microcode: CPU0 updated to revision 0x1a, date = 2014-05-23
[   39.004421] microcode: CPU1 sig=0x306c3, pf=0x10, revision=0x16
[   39.005588] microcode: CPU1 updated to revision 0x1a, date = 2014-05-23
[   39.005736] microcode: CPU2 sig=0x306c3, pf=0x10, revision=0x16
[   39.006657] microcode: CPU2 updated to revision 0x1a, date = 2014-05-23
[   39.006796] microcode: CPU3 sig=0x306c3, pf=0x10, revision=0x16
[   39.007661] microcode: CPU3 updated to revision 0x1a, date = 2014-05-23
[   39.007790] microcode: CPU4 sig=0x306c3, pf=0x10, revision=0x16
[   39.008656] microcode: CPU4 updated to revision 0x1a, date = 2014-05-23
[   39.008807] microcode: CPU5 sig=0x306c3, pf=0x10, revision=0x16
[   39.009916] microcode: CPU5 updated to revision 0x1a, date = 2014-05-23
[   39.010073] microcode: CPU6 sig=0x306c3, pf=0x10, revision=0x16
[   39.010939] microcode: CPU6 updated to revision 0x1a, date = 2014-05-23
[   39.011089] microcode: CPU7 sig=0x306c3, pf=0x10, revision=0x16
[   39.012082] microcode: CPU7 updated to revision 0x1a, date = 2014-05-23

Just checked for new bios version and there is one (release date 2014/04/28) so I will try to upgrade it (but my microcode is newer).

Do you know why it's reported only one time after each boot. MCE check for errors every 5 minutes (as I know) but the error is reported only once and it's 5 minutes after boot (on the first/second check), no such errors before next boot.
Do you know a way to reproduce the problem any time (it will be good if laptop freez/crash) so I can make a demo to the service guy.

Thank you.

business_kid · 06-27-2014, 12:31 PM

iirc, memtest86 checks the cache. Why not run it and see?

botzko · 06-30-2014, 12:55 AM

I upgrade bios and no logs from mce. I'll try memtest86 thank you.