14 Sep
2017
14 Sep
'17
3:39 p.m.
From: Alec Muffett
"threaded code" in the old sense could be
smaller than the equivalent
CISC binary on the same machine
One can think of 'threaded code' as code for a new virtual machine, one
specialized to the task at hand.
For those who really want to delve in some depth, see the chapter "Turning
Cousins into Sisters" (Chapter 15, pg. 365) in "Computer Engineering: A DEC
View of Hardware Systems Design", by Bell, Mudge and McNamara.
Interesting factoid: The PDP-11 initially used a threaded FORTRAN
implementation. In line with the observation above (about a new virtual
machine), DEC actually looked into writing microcode for the -11/60 (which had
a writeable control store) to implement the FORTRAN virtual machine.
Noel
14 Sep
14 Sep
4:24 p.m.
New subject: B Source Code
On Thu, Sep 14, 2017 at 9:39 AM, Noel Chiappa <jnc(a)mercury.lcs.mit.edu> wrote:
Huh. I happened to have that book on my shelf here at work and it's a great
description. That chapter cites a CACM paper from June 1973; citation here:
http://dl.acm.org/citation.cfm?id=362270
The paper is short; only 4 or so pages.
Now I'm confused about the timeline: I thought B was obsolete by this time
and that C was on the rise. Was the term "threaded code" in use earlier? Ah,
I see a note at the bottom of the paper that it was received in June, 1971 and
revised in December 1972, but not published until June 1973 (2 years after
initial submission). Given that the technique is described in the context of a
FORTRAN IV compiler for the PDP-11 that must have been in existence at
the time it was submitted in 1971, it seems reasonable to believe that the
technique was known in advance.
- Dan C.
From: Alec
Muffett
"threaded code" in the old sense could
be smaller than the equivalent
CISC binary on the same machine
One can think of 'threaded code' as code for a new virtual machine, one
specialized to the task at hand.
For those who really want to delve in some depth, see the chapter "Turning
Cousins into Sisters" (Chapter 15, pg. 365) in "Computer Engineering: A DEC
View of Hardware Systems Design", by Bell, Mudge and McNamara. 
9:46 p.m.
New subject: B Source Code
On Thu, Sep 14, 2017 at 10:24:08AM -0400, Dan Cross wrote:
Was the term "threaded code" in use earlier?
Forth.
Which was say 1968 - 1970 for its inception, how did it's early
versions refer to the technique? I certainly recall it being
used as the term when I learnt about Forth (which was much later than that).
DF
15 Sep
15 Sep
5:59 a.m.
New subject: B Source Code
On Thu, 14 Sep 2017, Noel Chiappa wrote:
Interesting factoid: The PDP-11 initially used a
threaded FORTRAN
implementation. In line with the observation above (about a new virtual
machine), DEC actually looked into writing microcode for the -11/60
(which had a writeable control store) to implement the FORTRAN virtual
machine.
Did anyone actually use the WCS? I had visions of implementing CSAV and
CRET on our -60, but never did get around to it. Too late now...
--
Dave Horsfall DTM (VK2KFU) "Those who don't understand security will
suffer."
1:43 p.m.
New subject: B Source Code
I thought Reuters re-microcoded them to be the fastest PDP8 around...
So was the DEC Field Service story...
Bill
-----Original Message-----
From: Dave Horsfall <dave(a)horsfall.org>
To: The Eunuchs Hysterical Society <tuhs(a)tuhs.org>
Sent: Fri, 15 Sep 2017 0:00
Subject: Re: [TUHS] B Source Code
On Thu, 14 Sep 2017, Noel Chiappa wrote:
Interesting factoid: The PDP-11 initially used a
threaded FORTRAN
implementation. In line with the observation above (about a new virtual
machine), DEC actually looked into writing microcode for the -11/60
(which had a writeable control store) to implement the FORTRAN virtual
machine.
Did anyone actually use the WCS? I had visions of implementing CSAV and
CRET on our -60, but never did get around to it. Too late now...
--
Dave Horsfall DTM (VK2KFU) "Those who don't understand security will
suffer."
19 Sep
19 Sep
3:20 a.m.
New subject: B Source Code
On Fri, 15 Sep 2017, William Pechter wrote:
[ WCS on the 11/60 ]
I thought Reuters re-microcoded them to be the fastest
PDP8 around...
I'd heard that some newspaper house in Australia ordered a bunch, and when
the contract fell through DEC were stuck with them, so they practically
gave them away to Universities.
--
Dave Horsfall DTM (VK2KFU) "Those who don't understand security will
suffer."
15 Sep
15 Sep
2:44 p.m.
New subject: B Source Code
CMU/CS implemented CSAV and CRET on the 11/40e (6th edition++). IMO - it
was a mixed bag. I've forgotten the statistics, it was a few ticks fasters
for saving the registers, which was good. The primary thing it did was
save a small amount of address space on the non-I/D 40 - which was helpful
for applications as they got larger and were running out of address
space...... But it was PITA in practice because it meant that binaries
compiled on IUS and SUS (the UNIX two systems in CS), would not run
anywhere else on campus which were mostly 11/34 or later 11/34A's and it a
networked environment (which we were just starting to create) the seams
tended to show a little more than we would have liked.
IIRC I eventually got a emulator working at Mellon Institute and EE systems
so they binaries would not core dump, but we just recompiled we ran into
them.
The real solution to address space issue was when DEC released the 11/44
which was separate I/D (11/70 class) and the thunk work that we did in
2.9BSD in trying to move 4.1/4.2 code to the 11.
FYI: Danny Klein and I should have the compiler somewhere (and the
microcode) is I ever get my CMU archives off tape. It's on my list of
things to look for.
Clem
On Thu, Sep 14, 2017 at 11:59 PM, Dave Horsfall <dave(a)horsfall.org> wrote:
On Thu, 14 Sep 2017, Noel Chiappa wrote:
Interesting factoid: The PDP-11 initially used a threaded FORTRAN
implementation. In line with the observation
above (about a new virtual
machine), DEC actually looked into writing microcode for the -11/60 (which
had a writeable control store) to implement the FORTRAN virtual machine.
Did anyone actually use the WCS? I had visions of implementing CSAV and
CRET on our -60, but never did get around to it. Too late now...
--
Dave Horsfall DTM (VK2KFU) "Those who don't understand security will
suffer."
3:03 p.m.
New subject: B Source Code
Also, in my Tektronix archives... we had an 11/60 with WSC and IIRC Steve
Glaser took a stab at moving the CMU stuff. The microengines between the
11/40 and 11/60 were different, so the code had to be rewritten. A big
issue, again IIRC, was the microcode compiler/tools for the WSC ran on RSX
so it meant UNIX was not running, which was not popular.
But around the same time, unexpected our group had the chance to get an
fairly loaded cast off seperate I/D based 11/70 which had been running
RSTS/Cobol for a business support team in sales. By the time, Steve got
anything working on the 11/60, I had the 11/70 (aka "teklabs") up and
installed; and I don't remember what happened to the WSC stuff; since we
had better solution to small address space problem of adding a '17th
address bit' with hardware.
Clem
Clem
On Fri, Sep 15, 2017 at 8:44 AM, Clem Cole <clemc(a)ccc.com> wrote:
CMU/CS implemented CSAV and CRET on the 11/40e (6th
edition++). IMO - it
was a mixed bag. I've forgotten the statistics, it was a few ticks fasters
for saving the registers, which was good. The primary thing it did was
save a small amount of address space on the non-I/D 40 - which was helpful
for applications as they got larger and were running out of address
space...... But it was PITA in practice because it meant that binaries
compiled on IUS and SUS (the UNIX two systems in CS), would not run
anywhere else on campus which were mostly 11/34 or later 11/34A's and it a
networked environment (which we were just starting to create) the seams
tended to show a little more than we would have liked.
IIRC I eventually got a emulator working at Mellon Institute and EE
systems so they binaries would not core dump, but we just recompiled we ran
into them.
The real solution to address space issue was when DEC released the 11/44
which was separate I/D (11/70 class) and the thunk work that we did in
2.9BSD in trying to move 4.1/4.2 code to the 11.
FYI: Danny Klein and I should have the compiler somewhere (and the
microcode) is I ever get my CMU archives off tape. It's on my list of
things to look for.
Clem
On Thu, Sep 14, 2017 at 11:59 PM, Dave Horsfall <dave(a)horsfall.org> wrote:
On Thu, 14 Sep 2017, Noel Chiappa wrote:
Interesting factoid: The PDP-11 initially used a threaded FORTRAN
implementation. In line with the observation
above (about a new virtual
machine), DEC actually looked into writing microcode for the -11/60 (which
had a writeable control store) to implement the FORTRAN virtual machine.
Did anyone actually use the WCS? I had visions of implementing CSAV and
CRET on our -60, but never did get around to it. Too late now...
--
Dave Horsfall DTM (VK2KFU) "Those who don't understand security will
suffer."
4:16 p.m.
New subject: Rich's B compiler, was Re: B Source Code
I mentioned some of this list's discussions about overlays and threaded
code in the pub last night, and my friend Rich Wareham said that he had
implemented a B compiler as a fun project, using Python and LLVM.
I thought it might be of interest, so here's a link:
https://github.com/rjw57/rbc
The doc/pdf subdirectory includes copies of the reference documents Rich
used as a specification for B.
Tony.
--
f.anthony.n.finch <dot(a)dotat.at> http://dotat.at/ - I xn--zr8h punycode
Shannon, Rockall: Northerly 4 or 5, occasionally 6 at first in east. Moderate
or rough. Showers. Good.
2619
days inactive
2624
days old
8 comments
7 participants
participants (7)
-
Clem Cole -
Dan Cross -
Dave Horsfall -
Derek Fawcus -
jnc＠mercury.lcs.mit.edu -
Tony Finch -
William Pechter