jnc at mercury.lcs.mit.edu
Tue Nov 18 14:22:51 CST 2014
> From: John Wallace
> Perhaps not so obviously,11/23 boards were available in variants with
> or without floating point, and (iirc) with or without memory management
> unit (MMU) (correction welcome).
No, you're right on target. (Although one couldn't have FP without the MMU;
the FP registers were actually in the MMU chip, see MicroComputers+Memories
'82 pp. 350-351.
> is it vaguely conceivable that the diagnostic in question needs a piece
> of hardware that isn't on the board in question
Hmm. Possible, but from what I know, unlikely. That test claims to test:
Part 1 - the CPU checks out the basic PDP11 instructions in every addressing
mode with various data patterns.
Part 2 - tests all trap instructions, trap overflow conditions R6,
interrupts, the reset and wait instruction.
Part 3 - tests the EXTENDED INSTRUCTION SET, the ASH, ASHC, MUL and DIV
all of which are, AFAIK, part of the basic chip. (Unlike the 11/03 and
11/40, where MUL et al are part of the optional EIS.)
Although I see 'interrupts' listed there - I wonder what they are using
to generate the interrupt? The console serial line? Must be, I can't think
of any other device in the config he was testing.
Ben, you didn't have a discontinuity in the grant jumpering, did you?
What kind of backplane do you have, and in what order/slots were the
boards plugged in?
> or is seriously broken on the board in question?
That is the other possibiliy.
Ben, do you get the same error with both CPU cards? Do you have spare
memory/serial line cards you can swap in?
More information about the cctalk