> That just doesn't make sense.. x86 changes I could imagine blaming on the
> irq.c changes (and even that I'd have a hard time doing), but that file
> isn't even used on m68k..
> The only common theme would be the generic networking code.
Interesting bit of information:
We had a Cisco ATM -> 100BaseT switch that just died. I had a dual P5/200
using the tulip driver on the same ethernet segment. With kernel 2.1.122 it
was averaging about 2-3 hours uptime before it would die nasty. ( lock:
completely unresponsice, or net layer would die: no pings, but could still
logon console )
I couldn't figure it out.
Then the Cisco box died. We replaced it with a HP 4000 stack. Now the box
has been running ever since the switch was replaced ( 2 days ). I was
curious about that, so I check to see what was different:
I'm not getting any of these any more:
Sep 8 17:51:40 angband kernel: eth0: Oversized Ethernet frame spanned
multiple buffers, status 7fffceff!
Sep 8 17:51:40 angband kernel: eth0: Oversized Ethernet frame spanned
multiple buffers, status 063185a2!
Food for thought.
--Jack
Wishing without work is like fishing without bait.
-- Frank Tyger
====-=-==== Jack (John) Cummings == cummings@nyquist.ee.ualberta.ca ====-=-====
========== for pgp public key --- finger the above address, and pray ==========
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/