Also change the device probing scheme to use something a bit more rational.
A current side-effect is that nubus cards are double-mapped. I expect
to fix that shortly.
Also change splclock() to block everything but serial hardware interrupts.
reject them for simple commands. Leo W. reports that this does
not work for some systems. It's better than what we have now,
though. From Paul Goyette <paul@pgoyette.bdt.com>.
MSG_REJECT_MESSAGE instead of MSG_ABORT when we get an extended message.
This allows a DEC drive to work when it attempts to negotiate synch.
transfers on us.
I added two new bitfields to ncr_softc. They flag which devices have
been selected, and which ones have failed to go to message out when
selected with ATN. If a target fails to go to message out the first
time it is selected, it goes into the latter bitfield. If it fails
thereafter, we treat it as we did before.