A friend of mine told me, that AMD also had some trouble concerning TLB on that architecture (translation lookaside buffer). Unfortunatelly I have no references for that issue.
I would keep a eye on that error, and if your system must be highly-available, i would even change hardware. Regards, -- Ralf On 09/24/13 10:01, Grant wrote: >> I had a deeper look into the kernel sources: >> >> Your error message is exactly thrown by >> static bool k8_mc1_mce(u16 ec, u8 xec) >> >> So probably you have a K8 ;-) >> >> Have a look at: >> http://www.redhat.com/archives/rhelv5-list/2007-October/msg00075.html > I read it, that one sounds like a correctable ECC RAM error. > >> It *might* be an error concerning ECC error correction. Did you recently >> change any hardware? > No hardware changed in a very long time. > >> Could you attach your /proc/cpuinfo? > Sure, I've attached it. I'm changing hosts and machines shortly and > I've only seen this error once so I'm thinking I don't need to take > action. > > - Grant > > >>> I share this opinion. >>> The message says - even if the error was corrected - that there's >>> something dramatically wrong with your - i suppose - CPU. >>> "Corrected error" might imply, that some low-level feature got disabled >>> in order to prevent furher errors. >>> >>> Does this error appear only once at early boot or frequently? >>> >>> Regards, >>> -- >>> Ralf >>> >>> On 09/23/13 22:07, Volker Armin Hemmann wrote: >>>> Am 23.09.2013 20:59, schrieb Paul Hartman: >>>>> On Mon, Sep 23, 2013 at 1:45 PM, Grant <emailgr...@gmail.com> wrote: >>>>>> Can anyone tell me how to decipher this which has appeared in dmesg? >>>>>> Google wasn't very helpful. >>>>>> >>>>>> [Hardware Error]: MC1 Error: Copyback Parity/Victim error. >>>>>> [Hardware Error]: Error Status: Corrected error, no action required. >>>>>> [Hardware Error]: CPU:3 (10:2:3) MC1_STATUS[-|CE|-|-|-]: >>>>>> 0x9000000000000171 >>>>>> [Hardware Error]: cache level: L1, tx: INSN, mem-tx: EV >>>>> Looks like machine check error, it detected an error in the L1 cache >>>>> on your CPU. >>>>> >>>>> Since it says "Corrected error, no action required" I would not worry >>>>> about it. If that makes you feel any better. :) >>>>> >>>>> >>>> since those errors are rare, I would worry about it.
