On Sun, May 23, 1999 at 11:26:59PM -0500, Brian C. Grayson wrote:
> > I have one, runnning with a GENERIC 1.4 kernel in a k6/II 350 machine,
> > 64Mb 100Mhz simms, asus p5a motherboard.
> > It's connected to a 100Mb HUB. Usage is: NFS (with a 10Mb server) for my
> > $HOME, cvs update from cvs.netbsd.org to local disk then
> > "tar cvf - src | rsh otherbox 'gzip -9 > /zip/src.tgz'" (this one uses the
> > full 100MBs). Also some package builds from a read-only NFS mounted
> > pkgsrc.

I just put another Dlink board in a k6II/350 machine with a no-name AT
motherboard (it has an Acer Labs chipset too), connected at 100Mbs full-duplex
to a 3com switch. I get regular:
May 25 13:50:28 antifer /netbsd: vr0: transmit underrun
May 25 13:50:28 antifer /netbsd: vr0: transmit aborted

During NFS activity with the 10Mb NFS server. NFS for this mount point is then
dead, it seems that a pattern of packet always triggers the error, and as
the host can't transmit it it thinks the NFS server is down.
Switching to 10Mbs half-duplex helps a little.
tried playing with the BIOS settings without luck. This machine also have an
AHC2940 scsi board, no IDE.
This is really strange because the driver is supposed to put the board in
store-and-forward mode, so transmit underrun should not happen.

Then I compiled a 1.4C kernel and got 2 strange things:
- a panic in the filesystem code from the sync syscall
- a reset. No messages, no debugger, the box just rebooted.
I don't know if this is related to the rhine ethernet. Changing some
BIOS settings to lower values allowed me to boot multi-user and start X.
Still investigating ...

--
Manuel Bouyer, LIP6, Universite Paris VI.           Manuel.Bouyer@lip6.fr
--