Interpreting MCA error output
royce.williams at gmail.com
Sat Oct 1 18:11:45 UTC 2011
On Sat, Oct 1, 2011 at 2:23 AM, Jeremy Chadwick
<freebsd at jdc.parodius.com> wrote:
> Decoding the MCE can be done using Linux's mcelog program -- you'll need
> to download the source and apply the patch by hand *and* put in place a
> heavily modified version of memstream.c -- which requires a lot of
> patching to work on FreeBSD, and can only be used to decode
> ASCII-provided MCEs; DMI support does not work. So, you have to apply
> patches then use "mcelog --no-dmi --ascii" and provide the MCE text via
> stdin (or use --file).
I'm glad to see this thread; I have a different error, for which I
wanted to make sure I was fixing the right problem before randomly
> John Baldwin tends to keep up-to-date patches for mcelog here:
> The last build of mcelog I did on FreeBSD was for mcelog-1.0pre2, which
> John's patch (at the time) did not work with. I made my own patch
> (dated 2011/02/11), but it looks like John has since updated his patch.
> If you need/want mine, I can put it up on the web.
That would be very useful as a crosscheck.
I found one additional intermediate patch from John, posted on
2011-04-26, that appears to have not been merged into ~jhb/mcelog/ :
That patch got me up and going (using 'gmake FREEBSD=yes i386=yes')
My problem is different from the original poster; looks a bit more serious:
royce at heffalump$ ./mcelog --no-dmi --ascii
MCA: Bank 1, Status 0x9400000000000151
MCA: Global Cap 0x0000000000000105, Status 0x0000000000000000
MCA: Vendor "AuthenticAMD", ID 0xfc0, APIC ID 0
MCA: CPU 0 COR ICACHE L1 IRD error
MCA: Address 0xc089d890
HARDWARE ERROR. This is *NOT* a software problem!
Please contact your hardware vendor
CPU 0 1 instruction cache
memory/cache error 'instruction fetch mem transaction, instruction
transaction, level 1'
STATUS 9400000000000151 MCGSTATUS 0
MCGCAP 105 APICID 0 SOCKETID 0
CPUID Vendor AMD Family 15 Model 12
> A few moments ago I tried to download mcelog from the official site, but
> ftp.kernel.org is presently returning NXDOMAIN for me (e.g. A record not
> found). The same goes for git.kernel.org. Great.....
kernel.org is still down from the compromise a few weeks ago. At
least one kernel.org mirror still has mcelog-1.0pre2:
The main mcelog page also has a link to how to get it from GitHub
while kernel.org is down:
Agreed that a port and a CGI would be higher leverage, but these
breadcrumbs should help in the short term.
More information about the freebsd-stable