18 Feb
2026
18 Feb
'26
11:36 a.m.
I've started working on a disassembly of the Digital VT320 video
terminal, so I'm diving into 8051 assembler.
At the moment, I'm using the D52 disassembler and am getting decent
results, although I wished I hadn't used the analyse option to produce
a starting control file; it doesn't cope at all well with code
involving lots of jump tables because the control flow appears to
break, and it decides that much of the code is just data. Ho hum.
Nevertheless, the code looks pretty understandable at the moment, so
the next stage is to annotate by hand, instead of updating the control
file, and to assemble whenever I make a change so that I can be sure
that I can produce the original ROM.
There isn't an "A52" to match D52, so I was wondering if anyone has
recommendations for a FOSS macro assembler that will take D52's output.
I ask for a macro assembler because there are some modifications I'd
like to make to the source file to clear things up. One of these would
be to cope with the fact that DEC decided that all of their jump tables
would have addresses in little-endian format, despite the 8051 being a
big-endian device. So, the normal "dw" directive is not going to work
for me, and I'll need to do something like:
dwl macro %1
db low(%1), high(%1)
endm
(Totally pulling that syntax out of my bum, btw.)
A quick search finds as31 and naken_asm - anyone use these?
I'd also love to hear recommendations for emulator code for the 8051,
preferably simple C because I'm going to need to hook up a 2681
DUART, 5911 EEPROM and DC7081 video chip (all presumably new code) to
get this working.
Paul.
18 Feb
18 Feb
4:30 p.m.
In article <20260218113616.15345ec9(a)chopoc.localdomain> you write:
I've started working on a disassembly of the
Digital VT320 video
terminal, so I'm diving into 8051 assembler.
Nice! I'm looking forward to it. Based on your VT100 annotated
source listing, it should be a good read.
There isn't an "A52" to match D52, so I
was wondering if anyone has
recommendations for a FOSS macro assembler that will take D52's output.
I haven't used either of these, but maybe the links will help:
gsim51: <https://gsim51.sourceforge.net/>
AS8051: <https://github.com/lukbettale/AS8051>
Unfortunately it doesn't look like either one supports a macro
facility... maybe use the C preprocessor for macros or m4?
--
"The Direct3D Graphics Pipeline" free book
<http://tinyurl.com/d3d-pipeline>
The Terminals Wiki <http://terminals-wiki.org>
The Computer Graphics Museum <http://computergraphicsmuseum.org>
Legalize Adulthood! (my blog) <http://legalizeadulthood.wordpress.com>
5:58 p.m.
Richard via cctalk wrote:
In article
<20260218113616.15345ec9(a)chopoc.localdomain> you write:
I think the very best Macro package for Assmebly where "GASP" Gnu
Assembler Preprocessor" ..but this project is dead since ages....
I've asked some AI about it (couldn't remember the name) and it was part
of the gnu binutils package. Here is a manpage:
https://manpages.debian.org/jessie/binutils-m68hc1x/m68hc12-gasp.1.en.html
This is gasp.c:
https://gnu.googlesource.com/binutils-gdb/+/d9c0bc92686a534d529e8eff558281f…
and this is the Texinfo file
https://gitweb.dragonflybsd.org/~profmakx/dragonfly.git/blob/984263bcb83ad8…
..three different locations :-)
Regards,
Holm
--
Technik Service u. Handel Tiffe, www.tsht.de, Holm Tiffe,
Goethestrasse 15, 09569 Oederan, USt-Id: DE253710583
info(a)tsht.de Tel +49 37292 709778 Mobil: 0172 8790 741
I've started working on a disassembly of the
Digital VT320 video
terminal, so I'm diving into 8051 assembler.
Nice! I'm looking forward to it. Based on your VT100 annotated
source listing, it should be a good read.
There isn't an "A52" to match D52,
so I was wondering if anyone has
recommendations for a FOSS macro assembler that will take D52's output.
I haven't used either of these, but maybe the links will help:
gsim51: <https://gsim51.sourceforge.net/>
AS8051: <https://github.com/lukbettale/AS8051>
Unfortunately it doesn't look like either one supports a macro
facility... maybe use the C preprocessor for macros or m4?
--
"The Direct3D Graphics Pipeline" free book
<http://tinyurl.com/d3d-pipeline>
The Terminals Wiki <http://terminals-wiki.org>
The Computer Graphics Museum <http://computergraphicsmuseum.org>
Legalize Adulthood! (my blog) <http://legalizeadulthood.wordpress.com> 
19 Feb
19 Feb
12:52 p.m.
Matti Nummi via cctalk wrote:
Google for "The Macro Assembler".
Good for several processors.
BR Matti
Not sure if you understood what I wrote..gasp is just a Macro
Preprocessor for Assemblers...
I've compiled it sucessfully now..it works...
Regards,
Holm
--
Technik Service u. Handel Tiffe, www.tsht.de, Holm Tiffe,
Goethestrasse 15, 09569 Oederan, USt-Id: DE253710583
info(a)tsht.de Tel +49 37292 709778 Mobil: 0172 8790 741
20 Feb
20 Feb
11:35 a.m.
Paul
Holm is completely justified in recommending gasp as a macro mangler, as he did to me a
year ago, it builds and it works.
I use it as a front end, via a makefile, to a meta-asembler for developing microcode.
The meta-assembler an AmdAsm clone https://github.com/hveit01/amdasm is a c++ / lex / yacc
marvel which can be tweaked to meet specific requirements; but out ofthe box it lacks a
macro mangler.
Orthogonally, rather than use macros why not use RegEx and your favorite programming
language to fix up the source code's endian issues.
HtH; BR Martin
-----Original Message-----
From: Holm Tiffe via cctalk [mailto:cctalk@classiccmp.org]
Sent: 19 February 2026 12:52
To: General Discussion: On-Topic and Off-Topic Posts <cctalk(a)classiccmp.org>
Cc: Holm Tiffe <holm(a)freibergnet.de>
Subject: [cctalk] Re: Recommendations for 8051 software tools?
Matti Nummi via cctalk wrote:
Google for "The Macro Assembler".
Good for several processors.
BR Matti
Not sure if you understood what I wrote..gasp is just a Macro Preprocessor for
Assemblers...
I've compiled it sucessfully now..it works...
Regards,
Holm
--
Technik Service u. Handel Tiffe, www.tsht.de, Holm Tiffe,
Goethestrasse 15, 09569 Oederan, USt-Id: DE253710583
info(a)tsht.de Tel +49 37292 709778 Mobil: 0172 8790 741
12:32 p.m.
Martin Bishop <mjd.bishop(a)emeritus-solutions.com> wrote:
Orthogonally, rather than use macros why not use RegEx
and your
favorite programming language to fix up the source code's endian
issues.
Well, I try to use Perl as often as possible for exactly that reason,
but the advantage of a macro assembler is that I get the symbol lookups
for free, to save me from brain farts.
So far, I just have these two:
dwl macro address
db low(address),high(address)
endm
which lets me construct the jump tables.
I can't fix up the source to make it big endian because I need to
annotate and rebuild the ROM exactly as it was delivered; I just need a
readable version.
Regards,
Paul
2:08 p.m.
Paul
My concept was to run a pre-proccessing pass to generate the Macro Expansions, your chosen
assembler would then do symbolic substitutions.
The pre-processing could be performed by gasp, a bespoke Perl/RegEx unit,or whatever.
ie dwl address -> db low(address),high(address) ; comment ...
The Perl option would of course permit semantic sugar to be sprinkled to taste
Best Regards
Martin
-----Original Message-----
From: Paul Flo Williams via cctalk [mailto:cctalk@classiccmp.org]
Sent: 20 February 2026 12:32
To: Martin Bishop <mjd.bishop(a)emeritus-solutions.com>
Cc: General Discussion: On-Topic and Off-Topic Posts <cctalk(a)classiccmp.org>rg>; Paul
Flo Williams <paul(a)frixxon.co.uk>
Subject: [cctalk] Re: Recommendations for 8051 software tools?
Martin Bishop <mjd.bishop(a)emeritus-solutions.com> wrote:
Orthogonally, rather than use macros why not use RegEx
and your
favorite programming language to fix up the source code's endian
issues.
Well, I try to use Perl as often as possible for exactly that reason, but the advantage of
a macro assembler is that I get the symbol lookups for free, to save me from brain farts.
So far, I just have these two:
dwl macro address
db low(address),high(address)
endm
which lets me construct the jump tables.
I can't fix up the source to make it big endian because I need to annotate and rebuild
the ROM exactly as it was delivered; I just need a readable version.
Regards,
Paul
19 Feb
19 Feb
2:23 a.m.
On 02/18/2026 6:36 AM EST Paul Flo Williams via cctalk
<cctalk(a)classiccmp.org> wrote:
There isn't an "A52" to match D52, so I was wondering if anyone has
recommendations for a FOSS macro assembler that will take D52's output.
Here is a link to a family of assemblers that you should look at. I have no idea if the
syntax will match, but I know the author goes to a good deal of pain to match manufacturer
syntax for most of them. In the limited use I have made of a few of them, they have
served me well. The author is VERY responsive to bug fixes and they are still under
active development.
http://john.ccac.rwth-aachen.de:8000/as/
Will
You just can't beat the person who never gives up.
Babe Ruth
2:58 a.m.
One free multiple target cross assembler I have used on few occasions
in the past is SB-Projects SB-Assembler. SB-Assembler 3 is written in
Python3 and runs on Linux, MAC and Windows machines.
https://www.sbprojects.net/sbasm/
https://www.sbprojects.net/sbasm/crosses.php
https://www.sbprojects.net/sbasm/download.php
https://github.com/sbprojects/sbasm3
The syntax it uses might need some translation from the syntax used by
other tools. Other than that I found it useful for what I needed when
I used it. Also, I don't think there is any linker / librarian
support, just a single output file, which probably rules it out for
large projects.
12:21 p.m.
On Wed, 18 Feb 2026 21:23:41 -0500 (EST)
Will Cooke via cctalk <cctalk(a)classiccmp.org> wrote:
Here is a link to a family of assemblers that you
should look at. I
have no idea if the syntax will match, but I know the author goes to
a good deal of pain to match manufacturer syntax for most of them.
In the limited use I have made of a few of them, they have served me
well. The author is VERY responsive to bug fixes and they are still
under active development. http://john.ccac.rwth-aachen.de:8000/as/
That's two votes for as/asl then, thanks.
Thank you for all the replies so far. I've loved working in assembler
since first touching a Z80 at school in 1982, but moving between
different processors (particularly older ones) has always been a bit of
a jarring experience because of the tools. I think "assemblers, yeah,
they're a solved problem, a mature set of tools" and then I realise
that, in many cases, "mature" means "rotten" and
"unmaintained"!
The first assembler I looked at for this was Ken Stauffer's as31,
despite it not having a macro facility: as Richard noted, macro
processing can easily be a bolt-on if necessary. However, it is written
in something like C90, and gcc threw so many warnings that I feared it
wouldn't produce an executable. When it did, I realised that, even
changing "db" to ".byte" throughout wasn't going to get my code to
assemble. The parser gets itself confused between binary and
hexadecimal constants if you use suffixed numbers. What is "0bch"? as31
can't work it out. Wasn't going to spend time doing wholesale
replacements for that and its quirks of single/double quoting.
Today, I'm working with W. W. Heinz's asem51, which is freeware, not
FOSS, but it has "standard" directives for specifying data and it
rebuilt a correct ROM binary for me with just a single textual change.
That is enough to keep me on track while I wrangle the code for now.
As far as emulator/simulators go, I've started with Jari Komppa's
emu8051, which I will no doubt outgrow, but it has a really simple
ncurses interface that allows me to jump around the code and try out
various routines to check my understanding of 8051 behaviour. Case in
point: there was a table-searching routine which was supposed to return
Z or NZ for a result, but it was looping with DJNZ which, in my Z80
brain, will always flag Z when it runs to completion. Aha! On the 8051,
DJNZ does not affect flags - weird!
I'll take a look at Seemanta Dutta's gSim51 at some point. I braved the
popups of Sourceforge, ignored its "alpha" status and the fact it was
last updated 13 years ago and downloaded it.
Loving the 8051 and this VT320 code so far. Thank goodness for the
availabilty of schematics and ancient datasheets.
Regards,
Paul
12:55 p.m.
Paul Flo Williams via cctalk wrote:
On Wed, 18 Feb 2026 21:23:41 -0500 (EST)
Will Cooke via cctalk <cctalk(a)classiccmp.org> wrote:
asl seems to be a nice thing..agree fully.
But I want just to mention the Frankenstein Corssassembler package,
I've used it in the past:
https://aminet.net/package/dev/cross/Frankenstein
Regards,
Holm
--
Technik Service u. Handel Tiffe, www.tsht.de, Holm Tiffe,
Goethestrasse 15, 09569 Oederan, USt-Id: DE253710583
info(a)tsht.de Tel +49 37292 709778 Mobil: 0172 8790 741
Here is a link to a family of assemblers that you
should look at. I
have no idea if the syntax will match, but I know the author goes to
a good deal of pain to match manufacturer syntax for most of them.
In the limited use I have made of a few of them, they have served me
well. The author is VERY responsive to bug fixes and they are still
under active development. http://john.ccac.rwth-aachen.de:8000/as/
That's two votes for as/asl then, thanks.
Thank you for all the replies so far. I've loved working in assembler
since first touching a Z80 at school in 1982, but moving between
different processors (particularly older ones) has always been a bit of
a jarring experience because of the tools. I think "assemblers, yeah,
they're a solved problem, a mature set of tools" and then I realise
that, in many cases, "mature" means "rotten" and
"unmaintained"!
The first assembler I looked at for this was Ken Stauffer's as31,
despite it not having a macro facility: as Richard noted, macro
processing can easily be a bolt-on if necessary. However, it is written
in something like C90, and gcc threw so many warnings that I feared it
wouldn't produce an executable. When it did, I realised that, even
changing "db" to ".byte" throughout wasn't going to get my code
to
assemble. The parser gets itself confused between binary and
hexadecimal constants if you use suffixed numbers. What is "0bch"? as31
can't work it out. Wasn't going to spend time doing wholesale
replacements for that and its quirks of single/double quoting.
Today, I'm working with W. W. Heinz's asem51, which is freeware, not
FOSS, but it has "standard" directives for specifying data and it
rebuilt a correct ROM binary for me with just a single textual change.
That is enough to keep me on track while I wrangle the code for now.
As far as emulator/simulators go, I've started with Jari Komppa's
emu8051, which I will no doubt outgrow, but it has a really simple
ncurses interface that allows me to jump around the code and try out
various routines to check my understanding of 8051 behaviour. Case in
point: there was a table-searching routine which was supposed to return
Z or NZ for a result, but it was looping with DJNZ which, in my Z80
brain, will always flag Z when it runs to completion. Aha! On the 8051,
DJNZ does not affect flags - weird!
I'll take a look at Seemanta Dutta's gSim51 at some point. I braved the
popups of Sourceforge, ignored its "alpha" status and the fact it was
last updated 13 years ago and downloaded it.
Loving the 8051 and this VT320 code so far. Thank goodness for the
availabilty of schematics and ancient datasheets.
Regards,
Paul 
10:34 p.m.
Here is a link
to a family of assemblers that you should look at. I
have no idea if the syntax will match, but I know the author goes to
a good deal of pain to match manufacturer syntax for most of them.
In the limited use I have made of a few of them, they have served me
well. The author is VERY responsive to bug fixes and they are still
under active development. http://john.ccac.rwth-aachen.de:8000/as/
That's two votes for as/asl then, thanks. 
20 Feb
20 Feb
12:21 a.m.
On 2026-02-19 3:34 p.m., Cameron Kaiser via cctalk wrote:
Three votes. I use asl (a/k/a Macroassembler AS) for anything that isn't 6502
(prefer xa65) or PowerPC. Most recently did some MC6800 and TMS7000 work in it.
I give it a vote as well. I can't use it ( as far as I can tell ),
because I need 36 bit cross assembler.
Did any of the PDP 10 cross assemblers make it past the 70's?
Ben.
Here is a
link to a family of assemblers that you should look at. I
have no idea if the syntax will match, but I know the author goes to
a good deal of pain to match manufacturer syntax for most of them.
In the limited use I have made of a few of them, they have served me
well. The author is VERY responsive to bug fixes and they are still
under active development. http://john.ccac.rwth-aachen.de:8000/as/
That's two votes for as/asl then, thanks. 
5:51 a.m.
Ben wrote:
I give it a vote as well. I can't use it ( as far
as I can tell ),
because I need 36 bit cross assembler.
Why?
Did any of the PDP 10 cross assemblers make it past
the 70's?
Some. CROSS has support for 6502, 8008, 8080, Z80, 8085, 8048, 6800,
6809, 1802. I see no 8051 in there, but isn't 8048 close?
7:23 a.m.
On 2026-02-19 10:51 p.m., Lars Brinkhoff wrote:
Ben.
Ben wrote:
Not sure what the why is FOR?
I have a new 36 bit computer design, complete with *magic smoke* The
current assembler is a rather hacked piece of software, with no macros.
I give it a vote as well. I can't use it ( as
far as I can tell ),
because I need 36 bit cross assembler.
Why?
Did any of the
PDP 10 cross assemblers make it past the 70's?
Some. CROSS has support for 6502, 8008, 8080, Z80, 8085, 8048, 6800,
6809, 1802. I see no 8051 in there, but isn't 8048 close?
9:33 a.m.
Ben wrote:
I have a new 36 bit computer design, complete with
*magic smoke* The
current assembler is a rather hacked piece of software, with no
macros.
Thank you, that answers my question.
>> Did any of the PDP 10 cross assemblers make it
past the 70's?
Then I suppose you were asking for a cross assembler *targeting* the
PDP-10? I have never heard of such an assembler from the 70's.
Supposedly the initial PDP-6 software was cross assembled from a PDP-4,
but I think that's lost. The only modern PDP-10 cross assembler I know
of is written in Erlang, but I'm guessing it doesn't support macros.
19 Feb
19 Feb
8:09 a.m.
On Wed, 18 Feb 2026, Paul Flo Williams wrote:
At the moment, I'm using the D52 disassembler and
am getting decent
results, although I wished I hadn't used the analyse option to produce
a starting control file; it doesn't cope at all well with code
involving lots of jump tables because the control flow appears to
break, and it decides that much of the code is just data. Ho hum.
Uh well, that is the problem with most disassemblers. Since interactive
disassembly is clearly superiour to normal disassemblers, I am using disx
at the moment, and I can easily write new processor modules for it. I'm
quite surprised that, even though it has only a pure ncurses UI, it is
simple and efficient to use.
Sometimes I use Ghidra, but it is a holy complex thing and has some very
nasty habits like C-style hex constants etc.
I had hoped that Binary Ninja would be a powerful and nice tool, although
commercial, but I have bought a personal license to try it. But this is
the biggest crap that I have ever seen! This is not what I call a
disassembler. It is a graphical whatever thing, and it can't even just
show you a disassembly in a normal fashion. Everything is transformed into
pseudo high-level garbage and so on. So for now I can only recommend to
stay away from it for real disassembly.
Christian
11:45 a.m.
On Thu, 19 Feb 2026 09:09:22 +0100 (CET)
Christian Corti via cctalk <cctalk(a)classiccmp.org> wrote:
Sometimes I use Ghidra, but it is a holy complex thing
and has some
very nasty habits like C-style hex constants etc.
I was wondering if Ghidra would get a mention ;-)
I haven't used it and I might be tempted, if I was tackling a project
that had been written in something 'higher' than assembly language.
However, this is clearly hand written and I have a good amount of
information already about the structures they would pick for the
project.
I had hoped that Binary Ninja would be a powerful and
nice tool,
although commercial, but I have bought a personal license to try it.
But this is the biggest crap that I have ever seen! This is not what
I call a disassembler. It is a graphical whatever thing, and it can't
even just show you a disassembly in a normal fashion. Everything is
transformed into pseudo high-level garbage and so on. So for now I
can only recommend to stay away from it for real disassembly.
I did try this:
https://github.com/msrst/interactive-8051-disassembler
which looked interesting from the screenshots but was quickly
frustrating. I don't know whether I just work differently, but I needed
to separate out code from data and then attach names to the data first,
starting with the I/O from the 8051 itself, UART, NVRAM, keyboard, etc.
The possible names of functions were only going to become apparent
after that was done, other than naming lots of tiny helper functions
around the place.
Regards,
Paul
8:12 a.m.
On Wed, 18 Feb 2026, Paul Flo Williams wrote:
I ask for a macro assembler because there are some
modifications I'd
like to make to the source file to clear things up. One of these would
be to cope with the fact that DEC decided that all of their jump tables
would have addresses in little-endian format, despite the 8051 being a
big-endian device. So, the normal "dw" directive is not going to work
for me, and I'll need to do something like:
Btw. the *best* macro assembler is asl from Alfred Arnold. I use it for
everything, and I have written several backends, e.g. IBM PALM, Dietz
MINCAL and Zuse Z43.
Christian
4:41 p.m.
On 2026-02-19 1:12 a.m., Christian Corti via cctalk wrote:
On Wed, 18 Feb 2026, Paul Flo Williams wrote:
Are there any good macro assemblers that could be self hosting?
Ben.
I ask for a macro assembler because there are
some modifications I'd
like to make to the source file to clear things up. One of these would
be to cope with the fact that DEC decided that all of their jump tables
would have addresses in little-endian format, despite the 8051 being a
big-endian device. So, the normal "dw" directive is not going to work
for me, and I'll need to do something like:
Btw. the *best* macro assembler is asl from Alfred Arnold. I use it for
everything, and I have written several backends, e.g. IBM PALM, Dietz
MINCAL and Zuse Z43.
Christian
2
days inactive
4
days old
21 comments
11 participants
participants (11)
-
ben -
Cameron Kaiser -
Christian Corti -
Glen Slick -
Holm Tiffe -
Lars Brinkhoff -
Martin Bishop -
Matti Nummi -
Paul Flo Williams -
Richard -
wrcooke＠wrcooke.net