On Sep 23 2004, 14:22, Paul Koning wrote:
> >>>>> "Pete" == Pete Turnbull <pete_at_dunnington.u-net.com> writes:
> The PDP11 architecture handbook says CIS was an option on the
> Micro/PDP-11, the 11/23-plus, the 11/24, and the 11/44. (Hm, I had
> forgotten about the 11/44...) Micro/PDP-11 seems like another name
> for 11/23. But elsewhere (in the intro) it also shows the LSI11/2 as
> a machine with CIS option. That's the 11/03, I believe.
Yes, the MicroPDP-11 was a specifc range of models of 11/23, 11/73, et
seq in BA23 and BA123 cabinets. The most common 11/03 was an LSI-11/2.
The term LSI-11 refers to either the original set of boards including
the M7264 which could be used to build a system, or more specifically
to the M7264 KD11-F processor card. A PDP-11/03 is, strictly speaking,
a DEC-packaged system, the early ones being an BA11-M box, or an 11V03
system in a BA11-N and with RX01 drives. LSI-11/2 is the later
processor card, used in BA11-N boxes to make some 11V03 and 11T03 (with
RL01 drives) systems.
The reason I asked my question is just that I've never seen or heard of
a CIS option for the 11/03, and I can't think where you'd put it.
There are no spare sockets on the board and not nearly enough of the
internal bus is brought out. Well, there's a spare socket on a basic
M7264 (the original quad board) and on a basic M7270 (the later and
more common LSI-11/2 dual-height board), but they're for the KEV11
(EIS/FIS) option. Most of the 11/03 processors I've seen have the
KEV11 already fitted.
It's also possible to add a Writable Control Store (KUV11) to an 11/03
(M7264) and that can hold user-defined microcode or a version of the
EIS/FIS. AFAIK that only works with an LSI-11 (M7264), not an
LSI-11/2, and anyway it's a quad-height board, not a MICROM chip (it
has a cable that plugs into the socket usually occupied by the KEV11).
I have both the DIGITAL Microcomputer Handbook (1976) that was released
with the LSI-11 series, and the later Microcomputers and Memories
Handbook (1982) which describe the 11/03, and neither mentions CIS,
except for the 11/23-plus (in the 1982 handbook). The 1976 Handbook
describes all the instructions, including the EIS and FIS, but not CIS.
It also describes or lists various languages but not COBOL or DIBOL
(which would have been its raison d'etre).
I've also been told (by someone on the list) that the 11/03 processors
don't have enough microcode address space for CIS, and that fits with
what I know: CIS is BIG -- bigger than the all rest of the instruction
set put together, including the floating point additions, on an 11/23.
So, whilst willing to believe that CIS exists for an 11/03 if someone
has it, I think something else is much more likely :-) Of course, if
it does exist, I'd be very interested to know more about it!
--
Pete Peter Turnbull
Network Manager
University of York
Received on Thu Sep 23 2004 - 17:44:10 BST