Is The M9312 Boot Module Essential?

[Noel Chiappa]

    >> On Feb 19, 2022, at 10:51 AM, Noel Chiappa wrote:

    >> The -11/34 (not the /34A) has something unusual for grant timeouts,
    >> but I forget the details. I'll look it up.

And here it is...

    > From: Fritz Mueller

    > I think you are thinking of the M9302, Noel: a far-side terminator card
    > with integrated SACK turnaround?

No; the M8264 Sack Timeout module. What's an M8264, you say? Well, there's
next to nothing in print about them, but I think I've managed to assemble
enough distant clues to work out their story.


Start with EK-11034-OP-PRE2 (I have a hard-copy of it); it gives a clue as to
how it all started. In 3.10.2, "End-of-Bus Terminator", it says:

  "As a result of this [SACK turnaround] circuitry [on the M9302], the SACK
  timeout feature found on other processors is not required"

So if i) a device requests a grant, and then drops the request at _just_ the
right time (so a grant gets sent out when there's no device waiting to grab
it), and ii) there's a break in that grant line (maybe a missing grant
continuity card) before it gets to the M9302, which can turn it around as a
SACK , then ... the KD11-E CPU will hang!

The M8264 was apparently the first attempt to deal with this.


Like I said, there's next to nothing in print about them. EK-11034-UG-001, in
Section 1.2, "System Description", does list "M8264 SACK Timeout module
(11/34 only)" in the list of components - but says nothing else _at all_
about it!

There is one page of circuit diagram of it, in:

  http://www.bitsavers.org/pdf/dec/pdp11/1134/MP00082_1134_Vol2_Sep76.pdf

on pg. 149. The rest of the prints for it (e.g. the PCB layout) aren't there,
but I happen to have one - it's a quad board with only a few components on
it. Looking at the circuit diagram, it's mostly just a 9602 retriggerable,
resettable one shot. (There's also a synchronous 4-bit up/down counter, but
that's just there to count events, and display the count in some LEDs -
probably just to make sure it isn't happening too often.)

Since I'm not a real hardware type, I'm not absolutely certain just from
looking at the circuit diagram exactly what it does, or how, but
EK-KD1EA-MM-001 Section 4.7.2.4, "No-SACK Timeout Circuitry", shows a very
similar circuit, and says it "asserts BUS SACK ... [if the device] does not
assert SACK within 22 usec after a grant line has been enabled." Presumably
the M8264 does the same thing.

Interestingly, that circuit appears in the KD11-EA prints on pg. K2-10; the
KD11-E prints have a blank space on that page where this circuit is in the
KD11-EA prints.

Since the M9302 appears in EK-11034-OP-PRE2, with SACK turnaround, I deduce
that the M8264 was produced _after_ that came out, and post-dates the M9302,
to fix the potential CPU hang issue I described - and was later dropped when
the -11/34 switched to the KD11-EA, with that circuit built in.


I'll do a page on the CHWiki about the M8264, and include an image of one.

I figure I might use my M8264 on my -11/04, which also doesn't have SACK
timeout (on the BG lines, for sure; it looks like it might have it on the NPG
line). The M8264 doesn't tie into the CPU, it just looks at UNIBUS lines, so
it can be plugged into any UNIBUS machine (near the start of the bus, since
the grant lines it monitors are wired sequentially).

	Noel