strange network crash [Linux taco 2.4.20 #7 SMP Thu Jul 3 21:48:15 CDT 2003 i686 unknown]

I had a very strange network crash and was wondering if anyone has any
insight.

/var/adm/syslog shows it starting:
Feb 2 07:24:21 taco kernel: eth0: Host error, FIFO diagnostic
register 0000.
Feb 2 07:24:21 taco kernel: eth0: PCI bus error, bus status 800000a0

this continues many times a second (~30 times a second) for about 45
minutes. An occassional :
eth0: Too much work in interrupt, status e003.
is in there too.

At 08:12:10, syslog gets garbled with what looks like uuencoded text
for a few screens and then about 20 lines appear from the ftp log that
has dates back in October and that doesn't even exist!

The system was still active, but with no networking from the time it
all started.

We had a motherboard burn out a few days earlier, and it was a
replacement. (at least it wasn't another mobo prob (maybe))

Any thoughts? DOS attack? post-crash fs hiccup? bad mobo?

thanks -
Alan