determining flaky SIMM from log data

From: Steve VanDevender <stevev_at_hexadecimal.uoregon.edu>
Date: Mon, 25 Aug 1997 09:42:45 -0700 (PDT)

A while back I could swear I saw a posting go by about how to determine
the location of a SIMM from the address provided in ECC reports like the
one below:

Aug 22 13:25:34 downpour vmunix: Machine Check error corrected by processor
Aug 22 13:25:34 downpour vmunix: Physical address of error ffffff0012fc6e8f Corrected ECC Error in Memory during D-Cache fill
Aug 22 13:25:34 downpour vmunix: Single Bit error in Quadword 1 at bit<20> in a Data bit
Aug 22 13:25:35 downpour vmunix: CIA Error Reg = 0000000000000000
Aug 22 13:25:35 downpour vmunix: CIA Error Status = 0000000000000000
Aug 22 15:51:06 downpour vmunix: CIA Error Reg = 0000000000000000
Aug 22 15:51:06 downpour vmunix: CIA Error Status = 0000000000000000

So far we have a couple of AlphaStation 500s that get a couple of these
reports a day, which is essentially harmless. If they get worse,
though, I'd love to know how to replace just the SIMM that's causing
trouble.
Received on Mon Aug 25 1997 - 19:06:31 NZST

This archive was generated by hypermail 2.4.0 : Wed Nov 08 2023 - 11:53:36 NZDT