> the NE2000 driver (the comment in 3c509.c is a dead give-away ;-)
*laugh*
> may be an unresolved issue with respect to supplying i/o addresses and
> the way in which the cards respond to the ID port during the probe.
> This may mainest itself as things not getting detected when you supply
> "io=0xNNN,0xMMM" but working fine when you do "io=0xMMM,0xNNN" so do
Ah, right. I thought the problem was that the order in which cards would
get autoprobed wasn't deterministic, so if you just say "ether=0,0,eth1"
you can't be sure which card will end up as eth0 and which as eth1.
However, I only have one 3c509 so I've never tried it, and I'm probably
remembering wrong.
phil