Tracking down zero UNIBUS interrupt vectors

[v.wales%ucla-locus at sri-unix.UUCP]

From:            Rich Wales <v.wales at ucla-locus>

On our VAX 11/780 running 4.1BSD, every so often some device on the
UNIBUS starts spewing forth zero interrupt vectors.  After about
250,000 of these, of course, UNIX's solution is to do a UNIBUS reset.

Clearly, it would be nice if I could figure out which device is causing
all these zero vectors -- but I can't seem to figure out any way to get
the culprit to 'fess up, since (Catch-22!) the only way I can think of
to identify an interrupting device on the UNIBUS is by its vector.

By adding a few lines to sys/locore.s and dev/uba.c, I was able to tell
that the guilty device is interrupting at IPL 15.  That doesn't really
help me much, though, because just about EVERY device I have interrupts
at IPL 15.

Does anyone out there have any helpful hints?

Our UNIBUS configuration is as follows, by the way:

	    1   SI 9400 disk controller
	    1   ABLE DH/DM
	    6   DEC DZ-11's
	    1   Proteon V2LNI interface
	    1   Interlan Ethernet interface

Everything we have interrupts at IPL 15, except for the LNI interface
and the DM half of the DH/DM (both of which interrupt at IPL 14).

-- Rich <wales at UCLA-LOCUS>