On Jun 12,  5:38pm, Jonathan Stone wrote:
} 
} The dc driver in 1.2-Alpha (as of June 12) repeatably hangs a 5000/200
} here, and has done since approximately June 5.  (or perhaps earlier,
} June 5 was the first time in a while I'd rebooted that machine,
} and it's been wedged from then until today)
} 
} The machine hangs immediately after printing
}    adding dc0 subslot 0 offset 2000000
} 
} and sits there until the reset button is pressed.
} 
} I'd appreciate hearing a confirmation from anyone who's successfully
} booted a 3100 or a 2100 with a kernel compiled from June 5 or newer
} sources.

  I would suspect that I've messed up the match/attach code someplace.
The dc device appears to be the only other device that configures on the
mainbus or on an IOASIC [or at least appears to for the autoconfig
stuff].  I've looked at the match/attach code, but I don't see any
obvious problems yet - I need to think about it a bit.  I'd suspect
that the match routine is accessing a bad pointer.

Michael