progress report

[David G. Conroy]

I have a lot of other things to do lately, but I continue to make forward
progress on the design of my small pdp10.

I spent a bunch of time writing a couple of new diagnostics for the ITS
specific parts of the microcode; interrupts, traps, address translation,
address translation failure reporting, XCTR, and so on. Found a couple of
bugs, once again related to the fact that I (perhaps unwisely) decided *not*
to include hardware to cover the write-read hazard on one of the operand
busses feeding the alu (the bus where the check is complicated because it
has to take source, destination, and accumulator numbers into account), and
which required that the code be shuffled about a little.

[ And yes, I should write a program which traces through the code and points
out the places where the hazard causes the old value to be used, and there
isn't a pragma indicating that I'm being tricky and want the old value. ]

I also spent some time getting ITS exec ddt and nsalv to work. I can load a
file containing ddt and nsalv into the simulator, and I can use ddt, format
a fake disk in a big file, and successfully salvage a fake disk full of
files (a simh rm03 image, with the 1KW blocks swizzled into natural order,
and the data reformatted the way I want it). Not only is the combination of
the micromachine and the microcode continuing to act like a pdp10, but I
have versions of ddt and nsalv which know how to drive a 16450-style console
and an IDE-style disk, at least in a simplistic way (after I found the AOSL
in the new disk driver which should have been an AOJL).

dgc

[Karl Ramm]

"David G. Conroy" <d...@spies.com> writes:
> have versions of ddt and nsalv which know how to drive a 16450-style console
> and an IDE-style disk, at least in a simplistic way (after I found the AOSL
> in the new disk driver which should have been an AOJL).

Have you given any thought to hooking up an ethernet controller to this
one? The software would appear to be a pile of work, at least in the ITS
case, but it seems like it would be interesting... I suspect, even, that
if the hardware were to be made available somehow, @i(someone) would write
a driver for it.

This causes me to remember another question: How "available" or
"reproducible" do you intend these designs to be? Your web site appears to
have just enough of the rope necessary to reproduce a PDP-8/X, but not the
-4/X. Your incipient -10 is the first of your designs to really make me
wonder about this issue, though. :-)

Karl Ramm <pos...@1ts.org>