Izumi Tsutsui asked again;

>> why should we stay with
>> the old-and-archaic leftover of softintr()/netintr() scheme?  I want
>> to see a cool solution in MI area.
>
> I know we should switch more modern way like Solaris as Jolitz says,
> but for now I'd just like to know if current implementation will work
> as expected or not...

softintr()/netintr() code path runs in true interrupt context which MIPS
interrupt machinary provides.  Not all NetBSD capable processors can
do that, then it would be necessary to craft pseudo interrupt threads to go.
The sole issue is whether softintr()/netintr() code path is preempted by
themselves.

Toru Nishimura/ALKYL Technology