SUN 3 Bus Arbitration Problem

[JARMOLOWSKI at esdsdf.dnet.ge.com]

| schmitz at fas.ri.cmu.edu (Donald Schmitz) writes:
| 
| This suggested a bus arbitration problem: one of the SUN, disk controller,
| or Ironic's add-on CPU aren't playing by the rules, ie. try to use the bus
| when someone else still thinks they own it.  I looked at the bus request,
| bus busy, bus grant (at various places) and address strobe lines using a
| logic analyzer, and never saw an invalid arbitration cycle (although I had
| no way of triggering exactly when the SUN crashed). Unfortunately, the
| site has no spares for anything and there often aren't spares in all of
| Korea.  Before I tell them to buy a spare everything, wait 1-6 months for
| it to show, and then start trying combinations of boards, I'd like to know
| if there are any known problems with the SUN hardware in this regard, in
| particular I seem to remember rumors that the SUN 3-260 is not completely
| VME compliant.

We built a similar setup, although we made heavy use of VME bus.  We too
were never able to get a 3/260 to work.  One problem we found is that the
CG2 color board and the mouse live on the VME bus.  The SMD disk does too.
We had to give up and use the machine as a software development platform.