How to address mcelog warning?

I recently installed Fedora 30 (was using a different Linux distribution before). Now every time I resume from suspend I get a notification from Problem Reporting with this message:

The kernel log indicates that hardware errors were detected.
System log may have more information.
The last 20 mcelog lines of system log are:
==========================================
-- Logs begin at Mon 2019-05-06 00:29:56 EDT, end at Sun 2019-05-19 20:14:35 EDT. --
May 19 20:14:34 localhost.localdomain mcelog[1219]: CPUID Vendor Intel Family 6 Model 69
May 19 20:14:34 localhost.localdomain mcelog[1219]: Hardware event. This is not a software error.
May 19 20:14:34 localhost.localdomain mcelog[1219]: MCE 1
May 19 20:14:34 localhost.localdomain mcelog[1219]: CPU 0 BANK 6
May 19 20:14:34 localhost.localdomain mcelog[1219]: MISC 38a0000086 ADDR fef81ac0
May 19 20:14:34 localhost.localdomain mcelog[1219]: TIME 1558311273 Sun May 19 20:14:33 2019
May 19 20:14:34 localhost.localdomain mcelog[1219]: MCG status:
May 19 20:14:34 localhost.localdomain mcelog[1219]: MCi status:
May 19 20:14:34 localhost.localdomain mcelog[1219]: Error overflow
May 19 20:14:34 localhost.localdomain mcelog[1219]: Uncorrected error
May 19 20:14:34 localhost.localdomain mcelog[1219]: MCi_MISC register valid
May 19 20:14:34 localhost.localdomain mcelog[1219]: MCi_ADDR register valid
May 19 20:14:34 localhost.localdomain mcelog[1219]: Processor context corrupt
May 19 20:14:34 localhost.localdomain mcelog[1219]: MCA: corrected filtering (some unreported errors in same region)
May 19 20:14:34 localhost.localdomain mcelog[1219]: Generic CACHE Level-2 Generic Error
May 19 20:14:34 localhost.localdomain mcelog[1219]: STATUS ee0000000040110a MCGSTATUS 0
May 19 20:14:34 localhost.localdomain mcelog[1219]: MCGCAP c07 APICID 0 SOCKETID 0
May 19 20:14:34 localhost.localdomain mcelog[1219]: CPUID Vendor Intel Family 6 Model 69
May 19 20:14:34 localhost.localdomain mcelog[1219]: mcelog: warning: 8 bytes ignored in each record
May 19 20:14:34 localhost.localdomain mcelog[1219]: mcelog: consider an update

Is there something I should do to address this problem. I am not sure what the error is that it is warning about. I have not noticed any problems using the system and it has so far always resumed from suspend cleanly. I am not sure if it is signalling a hardware fault or just a hardware incompatibility with the OS.

1 Like

HI @wsha: welcome to the community! Please do have a look at the introductory posts in #start-here if you haven’t had a chance yet.

From the looks of it, these warnings result from having an older version of mcelog installed. I see a few bugs here requesting the maintainer to update the package: https://bugz.fedoraproject.org/mcelog.

I would “cc” myself to the bug, maybe drop a comment to let the maintainer know that it also affects you. (You can login to bugzilla.redhat.com using the same FAS account you’re using to access AskFedora)

Thanks @FranciscoD. I tried downloading and compiling the latest source code from GitHub, but I still see the same error. I read on the mcelog FAQ that the errors reported are generated by the hardware and the mcelog developers do not do hardware support and will ignore questions about hardware support. So I am still not sure what to do – disable mcelog? figure out a way to ignore this one suspend/resume error? just ignore the error notification that pops up every time I resume?

Unfortunately, I don’t know enough to even say if this is a hardware error or an mcelog error. If it is a hardware error, then it should be reported to your hardware vendor. I don’t know if most vendors let us report hardware issues to them, though. If it’s something simple as a firmware upgrade, they may be able to help, or if you are in warranty, they may be able to replace the faulty bits?

If it is an mcelog bug, maybe just report it to https://bugz.fedoraproject.org/meclog and let the maintainer handle it.

If it isn’t affecting your daily usage, you can ignore it for the time being. I don’t see too many other options :slightly_frowning_face:

@FranciscoD, output says in the beginning:

I’d say it’s a pretty safe bet that it’s real HW event mcelog should warn about, so it’s just doing its job.

Though to be fair I don’t know enough either.

1 Like