Got a machine that reports itself at boot time as

	Digital AlphaPC 164LX 533 MHz, 531MHz

with an NCR SCSI interface.  The NCR card is narrow-only, and we had
occasion to want to put a wide device on the box.  So we tried to plug
in a Qlogic card.  With this in place, dmesg describes the interfaces as

isp0 at pci0 dev 6 function 0
isp0: interrupting at eb164 irq 0
isp0: Board Revision 1020, loaded F/W Revision 7.55
isp0: Last F/W revision was 5.57
isp0: driver initiated bus reset
scsibus0 at isp0: 16 targets, 8 luns per target

ncr0 at pci0 dev 9 function 0: ncr 53c810 fast10 scsi
ncr0: interrupting at eb164 irq 3
ncr0: minsync=25, maxsync=206, maxoffs=8, 16 dwords burst, normal dma fifo
ncr0: single-ended, open drain IRQ driver
ncr0: restart (scsi reset).
scsibus1 at ncr0: 8 targets, 8 luns per target

The boot device is

sd0 at scsibus1 targ 0 lun 0: <SEAGATE, ST39140N, 1498> SCSI2 0/direct fixed
sd0(ncr0:0:0): 10.0 MB/s (100 ns, offset 8)
sd0: 8683MB, 9006 cyl, 8 head, 246 sec, 512 bytes/sect x 17783240 sectors

and seems to work fine...until I touch the tape drive that's the reason
we wanted a wide SCSI chain.  Then something goes wonky, and the next
attempt to touch sd0 produces errors and a hung system.  Unfortunately
the system hangs hard enough that I don't have the hang messages;
tomorrow I hope to set up another machine to capture them.

What I find baffling is that attempting to use one SCSI chain would
completely wedge a disk on the other SCSI chain.  I can't see what
could be doing that; surely the SCSI infrastructure isn't getting
confused between the two chains?

The tape drive does seem to work okay on other (non-NetBSD) machines,
but even if it's somehow incompatible, I could see it wedging scsibus0,
but not scsibus1.

This machine is running 1.4 (1999-05-13); it's a production machine, so
I don't update it casually.  Are there known problems in 1.4 that could
explain this?

					der Mouse

			       mouse@rodents.montreal.qc.ca
		     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B