helped wanted with ethernet problem.

[John McBride]

Every once in awhile my work machine crashes fairly hard. It has the 
home directory on nfs and if the network drops I have to reset the box 
to recover.

In /var/log/messages:

Feb 12 16:13:41 ws12 kernel: eth0: Too much work at interrupt,
IntrStatus=0x0040.
Feb 12 16:13:45 ws12 kernel: nfs: server lnx-fs not responding,
still trying
Feb 12 16:13:54 ws12 last message repeated 6 times
Feb 12 16:14:45 ws12 ypbind[773]: broadcast: RPC: Timed out.

dmesg has this to say about the card:

8139too Fast Ethernet driver 0.9.26
PCI: Enabling device 00:09.0 (0004 -> 0007)
PCI: Found IRQ 5 for device 00:09.0
PCI: Sharing IRQ 5 with 00:04.2
PCI: Sharing IRQ 5 with 00:0d.0
divert: allocating divert_blk for eth0
eth0: SMC1211TX EZCard 10/100 (RealTek RTL8139) at 0xd09c7000,
00:e0:xx:xx:xx:xx
, IRQ 5
eth0:  Identified 8139 chip type 'RTL-8139 rev K'
divert: freeing divert_blk for eth0
ip_tables: (C) 2000-2002 Netfilter core team
8139too Fast Ethernet driver 0.9.26
PCI: Found IRQ 5 for device 00:09.0
PCI: Sharing IRQ 5 with 00:04.2
PCI: Sharing IRQ 5 with 00:0d.0
divert: allocating divert_blk for eth0
eth0: SMC1211TX EZCard 10/100 (RealTek RTL8139) at 0xd09d3000,
00:e0:xx:xx:xx:xx
, IRQ 5
eth0:  Identified 8139 chip type 'RTL-8139 rev K'
eth0: link up, 10Mbps, half-duplex, lpa 0x0000
ip_conntrack version 2.1 (2047 buckets, 16376 max) - 292 bytes per conntrack

Does "too much work" mean the system is "exceptionally busy"? I've been 
doing resource limit and performance testing and have been running 90% 
CPU and surfing out into swap pretty heavily in recent weeks, although 
it seems to do this even when the system is fairly idle.

It's happened maybe 8-10 times since fc1 came out.

TIA,
John