Ken Hornstein <kenh@cmf.nrl.navy.mil> writes:
> What happens is that there's a way to tell the ISA (not the CardBus, AFAIK)
> PCIC chip to generate an interrupt on a particular IRQ.  Linux uses this
> to test all the valid IRQs and builds a mask of working IRQs.
> 
> The problem for us is that interrupts are off at autoconfiguration time.
> I don't know how to solve/work around this problem.

We could conceivably turn on interrupts (briefly) during autoconfig in 
order to figure this out. I mean, we don't do that, but we *could*, if 
need be. As an alternative, we might be able to figure out a way to
detect interrupts firing without actually enabling interrupts. Some
interrupt controllers let you do this, but I don't remember the PC's
controller allowing that. We might also be able to come up with some
other intriguing or gross hack, depending.

Perry