interrupt problems running two Xylogics 713 controllers?

[Webster Dove]

The 713 is a 753 lookalike for ESDI disks.  If we put two controllers in
the same chassis at the same (default) interrupt level 2, the second
controllers occasionally gets:

Jan  3 10:28:38 peach vmunix: xdc1: controller not responding
Jan  3 10:28:39 peach vmunix: xdc1: system lost interrupt from controller. SOFTWARE RECOVERED

When things are busy.  When we reverse the order of the controllers we
get:

Jan  3 14:26:10 peach vmunix: xdc0: controller not responding
Jan  3 14:26:13 peach vmunix: xdc0: system lost interrupt from controller. SOFTWARE RECOVERED

When we set the latter controller to priority 3 (at the suggestion of
Xylogics), we get:

Jan  2 18:32:04 peach vmunix: xd4g: read fixed (soft ecc corrected) -- blk #212800, abs blk #212800
Jan  2 18:32:04 peach last message repeated 6 times
Jan  2 18:32:04 peach vmunix: panic: dup biodone

That is, first some ecc errors on a brand new disk that don't appear when
the priorities are the same.  Then a biodone panic (whatever that is!).

Here is our bootup config:

Jan  2 18:32:04 peach vmunix: SunOS Release 4.0.3 (SERVER) #2: Tue Jan 2 17:08:36 EST 1990
Jan  2 18:32:04 peach vmunix: Copyright (c) 1989 by Sun Microsystems, Inc.
Jan  2 18:32:04 peach vmunix: mem = 8192K (0x800000)
Jan  2 18:32:04 peach vmunix: avail mem = 7217152
Jan  2 18:32:04 peach vmunix: Ethernet address = 8:0:20:6:28:ee
Jan  2 18:32:04 peach vmunix: xdc0 at vme16d32 0xee80 vec 0x44 
Jan  2 18:32:04 peach vmunix: xd0 at xdc0 slave 0
Jan  2 18:32:04 peach vmunix: xd0: <CDC Wren VI 94196-766 cyl 1630 alt 2 hd 15 sec 52>
Jan  2 18:32:04 peach vmunix: xd1 at xdc0 slave 1
Jan  2 18:32:04 peach vmunix: xd1: <CDC Wren VI 94196-766 cyl 1630 alt 2 hd 15 sec 52>
Jan  2 18:32:04 peach vmunix: xd2 at xdc0 slave 2
Jan  2 18:32:04 peach vmunix: xd2: <HP-97548EF cyl 1455 alt 2 hd 16 sec 56>
Jan  2 18:32:04 peach vmunix: xd3 at xdc0 slave 3
Jan  2 18:32:04 peach vmunix: xd3: <HP-97548EF cyl 1455 alt 2 hd 16 sec 56>
Jan  2 18:32:04 peach vmunix: xdc1 at vme16d32 0xee90 vec 0x45 
Jan  2 18:32:04 peach vmunix: xd4 at xdc1 slave 0
Jan  2 18:32:04 peach vmunix: xd4: <CDC Wren VI 94196-766 cyl 1630 alt 2 hd 15 sec 52>
Jan  2 18:32:04 peach vmunix: xd5 at xdc1 slave 1
Jan  2 18:32:04 peach vmunix: xd5: <CDC Wren VI 94196-766 cyl 1630 alt 2 hd 15 sec 52>
Jan  2 18:32:04 peach vmunix: se0 at vme24d16 0x300000 vec 0x40 
Jan  2 18:32:04 peach vmunix: st0 at se0 slave 32
Jan  2 18:32:04 peach vmunix: st1 at se0 slave 40
Jan  2 18:32:04 peach vmunix: sd0 at se0 slave 0
Jan  2 18:32:04 peach vmunix: sd1 at se0 slave 1
Jan  2 18:32:04 peach vmunix: sd2 at se0 slave 8
Jan  2 18:32:04 peach vmunix: sd3 at se0 slave 9
Jan  2 18:32:04 peach vmunix: zs0 at obio 0x20000 pri 3 
Jan  2 18:32:04 peach vmunix: zs1 at obio 0x0 pri 3 
Jan  2 18:32:04 peach vmunix: ie0 at vme24d16 0x31ff02 vec 0x74 
Jan  2 18:32:04 peach vmunix: root on xd0a fstype 4.2
Jan  2 18:32:04 peach vmunix: swap on xd0b fstype spec size 39000K
Jan  2 18:32:04 peach vmunix: dump on xd0b fstype spec

Any ideas???

		Dr. Webster Dove
		Special Computing Applications
		Advanced Technology Engineering
		Sanders Associates (a Lockheed Company)
		uunet!rocket!dove