foo et al,
i have no idea what card you guys are talking about, but from your
discussion, i'm going to assume that this must have been an ISA card.
(but i do remember saying that the computer shouldn't be posting ;) ).
there are some various technical reasons why you should NOT probe an ISA
card. they all boil down to the same thing: you can hang your machine
badly. as in, the scheduler stoping dead in its tracks.
i don't think the linux kernel autoprobes *ANY* isa cards for precisely this
reason. so what you need to do is load ISA nic cards at boot time, with a
bootup script.
when you modprobe a card, the driver looks for the correct base address for
the card's IO. but this is exactly what the kernel is trying to avoid in
the first place. what you need to do is pass the base address to the
module. ie:
>
> modprobe -a smc-ultra io=0x320
>
and, actually, we could've (prolly should've) passed an IRQ to the card
too. again, there is *NO* safe way to autoprobe isa cards.
i see no reason to do what mike did if this was a pci card...
jessica, the person you're trying to reach is almost certainly not on vox.
pete
On Mon, 9 Oct 2000, Foo Lim wrote:
> On Mon, 9 Oct 2000, Jessica wrote:
>
> > Can the person who was working on the computer with the SMC Elite Ultra
> > card (I believe it was the old obnoxious p133 that spent most of the day
> > defragging) tell me whether or not they got it working? I just stumbled
> > upon some information about this card that may be relevant if it didn't
> > get working.
> >
> > (Mike, I think that this was you)
> >
> > jessica
>
> Hi Jessica,
>
> I don't know what Mike did exactly. I helped out at the end of the day,
> right before Pete said "that computer shouldn't be POSTing" :)
>
> We finally added the line:
>
> modprobe -a smc-ultra io=0x320
>
> at the end of rc.local, and it seemed to work ok. I didn't have time to
> test it. Again, I dunno what he did before that caused the need for
> passing an io address to the card.
>
> FL
>
>
