Re: [gentoo-user] MCE error

[Mick]

On Saturday 28 Mar 2015 22:48:48 Sebas Pedersen wrote:
> On 28-03-2015 07:37 PM, Volker Armin Hemmann wrote:
> > Am 28.03.2015 um 23:00 schrieb Sebas Pedersen:
> >> On 28-03-2015 06:45 PM, Volker Armin Hemmann wrote:
> >>> Am 28.03.2015 um 14:58 schrieb Sebas Pedersen:
> >>>> Hi guys,
> >>>>
> >>>> From a few days ago I am experimenting an MCE error.
> >>>> Sometimes I turn on the computer and at some point while booting the
> >>>> kernel (after the grub menu) just freezes and puts this:
> >>>>
> >>>> CPU 0: Machine Check Exception: 4 Bank 4: b200000000070f0f
> >>>> TSC f5acc9180
> >>>> PROCESSOR 2:20fc2 TIME 1427486735 SOCKET 0 APIC 0 microcode 0
> >>>>
> >>>> the number for TSC may vary, but the b200000000070f0f it's always
> >>>> the
> >>>> same (at least for now). The error message suggest to parse the
> >>>> above
> >>>> error with mcelog. I did that and the result was:
> >>>>
> >>>> Hardware event. This is not a software error.
> >>>> CPU 0 4 northbridge TSC f5acc9180
> >>>> TIME 1427486735 Fri Mar 27 17:05:35 2015
> >>>>
> >>>>   Northbridge Watchdog error
> >>>>
> >>>>        bit57 = processor context corrupt
> >>>>        bit61 = error uncorrected
> >>>>
> >>>>   bus error 'generic participation, request timed out
> >>>>
> >>>>              generic error mem transaction
> >>>>              generic access, level generic'
> >>>>
> >>>> STATUS b200000000070f0f MCGSTATUS 4
> >>>> CPUID Vendor AMD Family 15 Model 44
> >>>> SOCKET 0 APIC 0 microcode 0
> >>>>
> >>>> The error suggest it's a hardware problem. I replace de RAM with no
> >>>> luck. Same error keeps happening.
> >>>>
> >>>> Any suggestion for identifying the problem or how to procede?
> >>>>
> >>>> Many thanks in advance!
> >>>>
> >>>> Sebas
> >>>
> >>> bios update/microcode update. A google search suggests that you have
> >>> run
> >>> into an errata.
> >>
> >> Oh OK, thank you. Must have miss that in the search. So you are saying
> >> that the error comes from a bios errata (and don't know what microdode
> >> is), and the fix is to update bios?
> >
> > no, possibly from a CPU errata and a bios update might bring in the
> > microcode update that works around that.
>
> I see, thanks for clarifying that. So looks like not too many options,
> either try to update the bios and/or replace the CPU.
>
> I really appreciated you replys and time.
>
> Thanks!,
> Sebas

There's 'CONFIG_MICROCODE=y' and friends in the kernel which along with sys-
apps/microcode-ctl will load what ever is the latest Intel/AMD CPU code 
(firmware) to patch any bugs with instructions that the CPU manufacturers have 
discovered.

-- 
Regards,
Mick