> > > > DEBUG eb164_iointr() vec = 0x900 irq = 0
> > > > scsibus0: waiting 2 seconds for devices to settle...
> > > > 
> > > > unexpected machine check:
> > > 
> > > Did this kernel work once in this machine ?
> > 
> > I think so, modulo different PCI cards, and being
> > connected to a hard drive rather than a CD/DVD burner.
> > 
> > Is the CMD controller known for failing on powerup?
> 
> No. These controllers are used in some popular sparc systems too (Ultra/5).
> 
> > 
> > I'm assuming that a fresh batch of "Vector allocation failed"
> > wouldn't have damaged the part?
> 
> I don't think it can. I suspect a software issue at the SRM level (not
> assigning a proper address to the DMA registers). Maybe the adapter you
> added changed a non-volatile state somewhere. If possible, try to remove
> all the adapters you added, and reset the SRM to factory defaults (I don't know
> how to do that last part, but I guess there is a way to do it - maybe involves
> a jumper setting).

I finally made a little progress:

I tried booting from a couple of CDs.  SRM successfully read data from the
drive and attempted to boot.  The boots failed, but it doesn't look like
it was the CMD controller's fault.  So... the controller is NOT dead!