Dec 19 00:13:08 yoda kernel: eth0: Tx timed out, lost interrupt? TSR=0x3,
ISR=0x3, t=20.
Dec 19 00:13:09 yoda kernel: eth0: Tx timed out, lost interrupt? TSR=0x3,
ISR=0x43, t=52.
Dec 19 00:13:09 yoda kernel: eth0: Tx timed out, lost interrupt? TSR=0x3,
ISR=0x43, t=52.
Dec 19 00:13:10 yoda kernel: eth0: Tx timed out, lost interrupt? TSR=0x3,
ISR=0x43, t=25.
When this happens, eth0 is "dead" until it's been ifconfig eth0 down;
ifconfig eth0 up. This exact same thing happened back in October 21, when
I was running 2.0.31-pre9-SMP.
Oct 21 18:45:37 yoda kernel: eth0: Tx timed out, lost interrupt? TSR=0x3,
ISR=0x3, t=40.
Oct 21 18:45:38 yoda kernel: eth0: Tx timed out, lost interrupt? TSR=0x3,
ISR=0x43, t=38.
Oct 21 18:45:38 yoda kernel: eth0: Tx timed out, lost interrupt? TSR=0x3,
ISR=0x43, t=46.
Oct 21 18:45:39 yoda kernel: eth0: Tx timed out, lost interrupt? TSR=0x3,
ISR=0x43, t=61.
My guess is this is some sort of SMP issue with the SMC Ultra driver,
since this system ran several months without these etherfreezes with
non-smp kernels as SMP didn't seem stable enough back in April when the
system was built.
I just looked at patch-2.0.33, and saw no obvious changes (nothing at all
with the SMC Ultra driver) that might help with this.
Is anyone else seeing this problem with SMP? Should I yank the card and
put in a tulip?
The system is a dual P150, 128mb RAM, Buslogic 948, 3 IBM SCSI disks,
Trident 9440, SMC 8216T Ultra.
------------------------------------------------------------------
Jon Lewis <jlewis@fdt.net> | Unsolicited commercial e-mail will
Network Administrator | be proof-read for $199/message.
Florida Digital Turnpike |
______http://inorganic5.fdt.net/~jlewis/pgp for PGP public key____