[eepro100] KERNEL: assertion (flags&MSG_PEEK) failed at tcp.c(1463)

Wei Keong chooweikeong@pacific.net.sg
Fri Apr 26 20:37:01 2002

Donald, thanks for your prompt reply.

I've another server down with the same rx buffer problem again. The funny
thing is both servers were down in the morning (5-6 am) where processing
is the lowest. Don't really understand why the kernel has not enough
memory for the network card... Have tried the eepro-diag to see the card
'sleep mode' has been turn on and the results is negative.

Will tuning VM solve the problem? What are the possible side effects? I
still have a couple of server (with similar configs) and I'm not sure if I
should tune the VM, as a preventive messure. Please advise.

Wei Keong

On Wed, 24 Apr 2002, Donald Becker wrote:

> On Thu, 25 Apr 2002, Wei Keong wrote:
> > I've been having problem with this Intel 82557 Ethernet Pro 100 card... hope
> > you guys can help me out.
> >
> > Previously, my system (IBM x340 server) was running Linux 2.4.9 and we had a
> > lot of "eth0: card reports no resources" error. It causes the system to hang
> > once in a while...
> >
> > Ever since we change the kernel to 2.4.18 (patched with 2.4.19-pre6
> > eepro.c), the "card reports no recources error" no longer appear. However,
> > we encountered a new problem after a week...
> >     eth0: can't fill rx buffer (force 0)!
> >     eth0: can't fill rx buffer (force 1)!
> This is the same problem, with a more direct error message.
> The driver cannot get memory from the kernel to fill the Rx descriptor
> list.  Presumably the kernel has temporarily run short of memory due to
> bad VM/memory-management parameters.
> Tune the parameters in /proc/sys/vm/bdflush as described a few days ago
> to avoid this performance problem.
> --
> Donald Becker				becker@scyld.com
> Scyld Computing Corporation		http://www.scyld.com
> 410 Severn Ave. Suite 210		Second Generation Beowulf Clusters
> Annapolis MD 21403			410-990-9993