From: Bugs (bb1@humboldt.edu)
Date: Wed Nov 17 2004 - 12:55:54 EST
My question was:
->
->Hello,
->
->ES40, tru64 5.1B PK4
->2 cpu's (The alpha EV6.7 (21264A) processor operates at 667 MHz)
->
->
->I got this today, anyone know what it is?:
->
-> System panic: panic (cpu 0): pg_nreaders going negative
->
I got responses from these fine folks shown below, many thanks!
I think that what MAY have happned, is that my fellow admin installed
a firmware upgrade in our SAN switch, and rebooted it around that same time.
We have 2 alpha's with mounted disk from a controller serviced
from that switch. The only one that panic'd was the one running steamd.
At least I hope thats all it was. If It happens again, and I get a
final resolution, I will let you know.
Thanks, bugs
==========================================================
From: Benjamin Ingwersen <benjamin@frontierus.com>
sounds like you need the current revision CPU's Rev D1, or if you already
have the most current rev, you need to make a service call..
If you need to purchase the Newer rev CPU's I can sell them for $ 750.00
==========================================================
From: "Dr Thomas.Blinn@HP.com" <tpb@doctor.zk3.dec.com>
#define READERS_DECR(PP) \
MACRO_BEGIN \
RWHOLD_CHECK(PP); \
if ((PP)->pg_nreaders == 0) { \
printf("pp=%P\n", PP); \
panic("pg_nreaders going negative"); \
} \
(PP)->pg_nreaders--; \
MACRO_END
Looks to be used in the "ubc" (unified buffer cache) management.
Should never be negative, hence the panic. Either someone zero-d
it when they shouldn't have or decremented it once too often.
Need to know more (stack trace) to know where it happened, but
it is a definite bug.
==========================================================
Bugs Brouillard Unix system administrator
Humboldt State Univ. Information Technology Services
Arcata, Calif.
email bb1@humboldt.edu
This archive was generated by hypermail 2.1.7 : Sat Apr 12 2008 - 10:50:11 EDT