68K C compiler

[BSvK]

I am looking for Digital Research C compiler for cp/m 68k originally
included in the distribution disks. Or any other C compiler for cp/m 68k.
Any suggestions ?

[James M. Knox]

"BSvK" <bshu...@redshift.com> wrote in
news:tu0n9qg...@corp.supernews.com:

> I am looking for Digital Research C compiler for cp/m 68k originally
> included in the distribution disks. Or any other C compiler for cp/m 68k.
> Any suggestions ?

The Alcyon C compiler was included with CP/M-68K. There were three
versions. The first was pretty lame (no floating point, etc.). The second
was much better and included floating point. The third version, just
before DRI started to self-destruct, was questionable... I personally was
never able to get it to work correctly, although I was told by others that
they did manage. [I guess technically I still am waiting on a response
from DRI to the Software Difficulty Report I turned in. <G>]

I believe that the files you want are on one of the CP/M web sites
(hopefully someone here will chime in with the URL). If not, I can e:mail
you what you need. I have the DRI disk's on line, but not the manuals.

jmk

-----------------------------------------------
James M. Knox
TriSoft ph 512-385-0316
1109-A Shady Lane fax 512-366-4331
Austin, Tx 78721 jk...@trisoft.com
-----------------------------------------------

[Charles Richmond]

"James M. Knox" wrote:
>
> "BSvK" <bshu...@redshift.com> wrote in
> news:tu0n9qg...@corp.supernews.com:
>
> > I am looking for Digital Research C compiler for cp/m 68k originally
> > included in the distribution disks. Or any other C compiler for cp/m 68k.
> > Any suggestions ?
>
> The Alcyon C compiler was included with CP/M-68K. There were three
> versions. The first was pretty lame (no floating point, etc.). The second
> was much better and included floating point. The third version, just
> before DRI started to self-destruct, was questionable... I personally was
> never able to get it to work correctly, although I was told by others that
> they did manage. [I guess technically I still am waiting on a response
> from DRI to the Software Difficulty Report I turned in. <G>]
>
> I believe that the files you want are on one of the CP/M web sites
> (hopefully someone here will chime in with the URL). If not, I can e:mail
> you what you need. I have the DRI disk's on line, but not the manuals.
>

I believe that you can download the Alcyon C compiler at:

<http://www.deltasoft.com/downloads-gemworld.htm>

Search down the page for "GEMKIT"...you have to download the
*whole* developer's kit to get the C compiler...

--
+-------------------------------------------------------------+
| Charles and Francis Richmond <rich...@plano.net> |
+-------------------------------------------------------------+

[coinst...@gmail.com]

On Thursday, November 1, 2001 at 8:28:23 AM UTC-6, James M. Knox wrote:

> > I am looking for Digital Research C compiler for cp/m 68k originally
> > included in the distribution disks. Or any other C compiler for cp/m 68k.
> > Any suggestions ?
>
> The Alcyon C compiler was included with CP/M-68K. There were three
> versions. The first was pretty lame (no floating point, etc.). The second
> was much better and included floating point. The third version, just
> before DRI started to self-destruct, was questionable... I personally was
> never able to get it to work correctly, although I was told by others that
> they did manage. [I guess technically I still am waiting on a response
> from DRI to the Software Difficulty Report I turned in. <G>]
>

I recently ported CP/M-68K version 1.3 to a repurposed SPX-MPU board with a MC68302 CPU.
https://www.retrobrewcomputers.org/forum/index.php?t=msg&th=201&start=0&

I downloaded the CP/M v1.3 binary recently from cpm.z80.de The C compiler appears to be working. I'm able to compile and run the micro EMAC for CP/M-68K as well as gkermit for 68k. So is the C compiler fixed now or is there a particular type of problem I have not yet encountered?

[BobH]

I never got the floating point to work in 1.3. Everything else worked as
I remember.

BobH

[David Schultz]

On 10/23/2017 02:56 PM, BobH wrote:
> I never got the floating point to work in 1.3. Everything else worked as
> I remember.
>
> BobH

It does seem to be broken. This simple test program failed to produce
correct output:

float a, b;

a = 43;
b = 27;
printf("%f %f %f\n", a, b, a * b);

In addition I tried using the 1.2 compiler with the 1.3 libraries and
the 1.3 compiler with the 1.2 libraries. Nothing worked. So there are
problems in the compiler and the libraries.

Too bad as it does appear to have fixed some of the known bugs.


--
David W. Schultz
http://home.earthlink.net/~david.schultz
Well, boys, I reckon this is it - nuclear combat toe to toe with the
Roosskies. - Major Kong
https://www.youtube.com/watch?v=TIoBrob3bjI

[coinst...@gmail.com]

Thank you all for your comments.

I compile David's sample floating program using Motorola fast floating point format (compile with c.sub, link with clinkf.sub) I got:
;.000011 ;.000008 <1.001844

The same program using IEEE floating point format (compile with ce.sub, link with clinke.sub) I got:
11<.001<=1 A<.0023?A C035.0=3;18

Interestingly, compile with ce.sub and link with clinkf.sub gets this:
0.000000 0.000000 0.000000

Obvious the wrong answers, but at least it looks like floating point values.

[David Schultz]

I looked through the data section for cp68, c068, and c168 and noticed
that the version in the 1.3 distribution included text indicating a 4.9
version number. There is no such text in the 1.2 distribution.

I then looked at the source code for the compiler (from
http://www.cpm.z80.de/source.html ) and found a file version.c which
indicates 4.3.