spin_trylock/spin_unlock panic in 2.6.9-42.0.3.EL

[Jon Lewis <jlewis@lewis.org>]

I have a system running CentOS 4.4 (kernel 2.6.9-42.0.3.EL) which crashed 
several times in just over an hour a couple nights ago.  Each time, the 
last messages / only indication of trouble was:

net/ipv4/icmp.c:239: spin_trylock(net/core/sock.c:f7c27a60) already locked
by net/ipv4/icmp.c/239
kernel panic - not syncing:
net/ipv4/icmp.c:251:spin_unlock(net/core/sock.c:f7c27a60) not locked

In irc, someone suggested this might be caused by bad hardware, but I have 
time believing that hardware going bad could cause the kernel to lose 
track of spinlocks, especially in exactly the same way repeatedly.

The system is an old Tyan Tiger 200, single PIII-933, 1GB RAM, acting as 
both a mail/web server and a NAT router using the two built in Intel 
825579 FE interfaces.

This system has been in use and stable for several months prior to these 
crashes.

----------------------------------------------------------------------
 Jon Lewis                   |  I route
 Senior Network Engineer     |  therefore you are
 Atlantic Net                |
_________ http://www.lewis.org/~jlewis/pgp for PGP public key_________
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/