In message <B665A8C8.6BE8%peter@boku.net>, "Peter A. Eisch" writes:
>And everything is fine until the interface is brought up.   Whether if the
>interface is brought up with an ifconfig command or via dhclient or
>whichever.

Yup.

>de0: abnormal interrupt: transmit process stopped

Yup.

>Which doesn't really add much more info than what we had before.

It actually tells you exactly what's wrong.  Somehow, the abnormal
interrupt bit, and the specific bit for "transmit process stopped", are
getting set.  I just sent some patches out that fix this in 1.5, and which
also fix the corresponding problem in the Tulip driver.  I believe this
is a VPC bug, but it's a fix.

The patches assume you'll add options VPC_CPU in your config file to compile
in all the hacks.

Anyway, so far as I can tell, the problem is that, any time the card finishes
transmitting, they set the flag for "transmit process stopped", which is
not what they mean to say, but sounds good if you don't know it's an abnormal
interrupt.

-s