> From thorpej@lestat.nas.nasa.gov Wed May 20 12:29:06 1998
> To: Ross Harvey <ross@teraflop.com>
> Cc: joe@mediaone.net, port-alpha@netbsd.org
> Subject: Re: multia ide drives 
> From: Jason Thorpe <thorpej@nas.nasa.gov>
> Date: Wed, 20 May 1998 12:17:52 -0700
>
> On Wed, 20 May 1998 10:45:57 -0700 (PDT) 
>  Ross Harvey <ross@teraflop.com> wrote:
>
>  > Unfortunately, every platform routes interrupts differently, and
>  > although the multia and axppci33 use the same sio ("Intel 82378ZB
>  > System I/O") as the eb164 systype, the interrupt isn't established
>  > with special-case code on the lca systems like it is on the eb164.
>  > (See chipset tag pc_pciide_compat_intr_establish if you want to
>  > work on it!)
>
> Um, on the multia, it's not PCI IDE, but rather ISA IDE on the super-i/o
> chip, right?
>
> On the EB164, the PCI IDE is either a separate CMD chip, or a differnet
> function on the Cypress PCI-ISA bridge (on the 164SX).
>

Oh, right. "Forget what I said!" 

(So, ignore the pciide stuff, it's new and is not the way that interrupt
is routed.) So, we must just have a conflict on the multia.