[comp.lang.c] C compiler for 386 or 486

mjnkowcz@unix1.tcd.ie (Maciej Jankowicz) (04/23/91)

	The PC's using 386 or 486 CPU are present for some time, but
I don't know about any C compilers which are able to generate code 
suitable for them to get the best of the machine. For example Turbo
C can toggle only between 86/88 and 186/286 instruction sets.
	
	I would appreciate very much any references of such products.

	Thank you very much in advance.

	Maciej Jankowicz
	mjnkowcz@unix1.tcd.ie

cjwein@watpix.waterloo.edu (Chris J. Wein) (04/24/91)

In article <mjnkowcz.672352737@unix1.tcd.ie> mjnkowcz@unix1.tcd.ie (Maciej Jankowicz) writes:
>
>	The PC's using 386 or 486 CPU are present for some time, but
>I don't know about any C compilers which are able to generate code 
>suitable for them to get the best of the machine. For example Turbo
>C can toggle only between 86/88 and 186/286 instruction sets.
>	
>	I would appreciate very much any references of such products.
>
>	Thank you very much in advance.
>
>	Maciej Jankowicz
>	mjnkowcz@unix1.tcd.ie

Try Watcom C/386 or Metaware C.  They use Pharlap Dos extender.  Watcom C
is lib compatible with Microsoft C.


-- 
==============================================================================
 Chris Wein                           | cjwein@watcgl.waterloo.edu 
 Computer Graphics Lab, CS Dept.      | cjwein@watcgl.uwaterloo.ca
 University of Waterloo               | (519) 888-4548 

c60b-1eq@e260-1g.berkeley.edu (Noam Mendelson) (04/24/91)

In article <mjnkowcz.672352737@unix1.tcd.ie> mjnkowcz@unix1.tcd.ie (Maciej Jankowicz) writes:
>	The PC's using 386 or 486 CPU are present for some time, but
>I don't know about any C compilers which are able to generate code 
>suitable for them to get the best of the machine. For example Turbo
>C can toggle only between 86/88 and 186/286 instruction sets.

How about gcc for DOS?  It requires a 32-bit processor.  I know
you can get it from sandstorm.berkeley.edu under pub/GNU/gcc, but I'm
sure you can find a more up-to-date version elsewhere (comments,
anyone?).

-- 
+==========================================================================+
| Noam Mendelson   ..!ucbvax!web!c60b-1eq       | "I haven't lost my mind, |
| c60b-1eq@web.Berkeley.EDU                     |  it's backed up on tape  |
| University of California at Berkeley          |  somewhere."             |

mackay@frith.egr.msu.edu (James F. MacKay) (04/24/91)

Intel is supposed to have a 32-bit compiler (386/486), but I have
never tried it.  Give them a call.

dj@ctron.com (DJ Delorie) (04/24/91)

In article <mjnkowcz.672352737@unix1.tcd.ie>, mjnkowcz@unix1.tcd.ie (Maciej Jankowicz) writes:
> 	The PC's using 386 or 486 CPU are present for some time, but
> I don't know about any C compilers which are able to generate code 
> suitable for them to get the best of the machine. For example Turbo
> C can toggle only between 86/88 and 186/286 instruction sets.

grape.ecs.clarkson.edu  Anonymous FTP  ~ftp/pub/msdos/djgcc

Stored there is djgpp, my 386/DOS port of the GCC & G++
compiler, along with sources, documentation, and the required
extender.  It is stored two ways:

djgpp.zip  2.2M  Everything
djgppzip.000 through djgppzip.063  32K chunks - upload the merge
  program there to put them together

The "readme" file is stored separately there also.  Upload it
first and read it before uploading the rest.

I am working on making it available from other sites and via other
means, and will announce as they are available.

DJ
dj@ctron.com

erick@sunee.waterloo.edu (Erick Engelke) (04/24/91)

In article <mjnkowcz.672352737@unix1.tcd.ie> mjnkowcz@unix1.tcd.ie (Maciej Jankowicz) writes:
>
>	The PC's using 386 or 486 CPU are present for some time, but
>I don't know about any C compilers which are able to generate code 
>suitable for them to get the best of the machine. For example Turbo
>C can toggle only between 86/88 and 186/286 instruction sets.

The Watcom 386 compiler is very good.  Novell sellected it as their 386 
compiler.  You will still have to buy a DOS extender or their very pricy
version which includes the extender.  They also have a recently released
p-code option to dramtically reduce the size of non-critical components by
generating fast p-code rather than assembly language.

    C 8.0 / 386  $895
    C 8.0 / 386 prof  $1295
    C 8/0/386/prof/with extender $1790 

I have no affiliation with Watcom, and I used to be unimpressed with their
work.  Their new C development systems are not as pretty as Borland, but they
generate very good, very fast code, and I highly recommend them. 
-- 
----------------------------------------------------------------------------
Erick Engelke                                       Watstar Computer Network
Watstar Network Guy                                   University of Waterloo
Erick@Development.Watstar.UWaterloo.ca              (519) 885-1211 Ext. 2965

jgay@digi.lonestar.org (john gay) (04/24/91)

From article <mjnkowcz.672352737@unix1.tcd.ie>, by mjnkowcz@unix1.tcd.ie (Maciej Jankowicz):

> 	The PC's using 386 or 486 CPU are present for some time, but
> I don't know about any C compilers which are able to generate code 
> suitable for them to get the best of the machine. For example Turbo

GNU C
Watcom C/386 8.0
Zortech C++/386
MetaWare High C 386
NDP C & C++/386

Just off the top of my head.  I am sure that I left some out.

Look in any programmer oriented magazine for adds from software places
that sale these.

Bob.Stout@p6.f506.n106.z1.fidonet.org (Bob Stout) (04/25/91)

Zortech C/C++ has code generation switches to generate both 386 and 486 
specific code. 

jimp@cognos.UUCP (Jim Patterson) (04/26/91)

In article <mjnkowcz.672352737@unix1.tcd.ie> mjnkowcz@unix1.tcd.ie (Maciej Jankowicz) writes:
>
>	The PC's using 386 or 486 CPU are present for some time, but
>I don't know about any C compilers which are able to generate code 
>suitable for them to get the best of the machine.
>	
>	I would appreciate very much any references of such products.

There are some fairly expensive compilers out there, which I have not
personally tried. I believe a company called PharLap makes one.

Or, get DJ Delorie's port of the Gnu CC compiler, which is available
under GNU General Public Licence (i.e. basically free). My experience
with it has been good, but I haven't tried anything large-scale and
there are a number of limitations (major ones are its incompatible
with other extended memory products e.g. Windows, and that 386's need
a 387 for floating point).

DJ's Gnu C compiler was posted in February to c.b.i.p but if you
missed it, here's some info on obtaining GCC. It's an FTP address; I
don't know of any uucp sites that might provide it. A warning: it's
big (the four .zoo files are about 1.6 Megabytes total).

---------------------------------------

File location:

    host:      grape.ecs.clarkson.edu
    login:     ftp
    password:  send your mail address
    directory: /pub/msdos/djgcc

Author:

    DJ Delorie
    24 Kirsten Ave
    Rochester, NH  03867-2954
    USA
-- 
Jim Patterson                              Cognos Incorporated
UUCP:uunet!mitel!cunews!cognos!jimp        P.O. BOX 9707    
PHONE:(613)738-1440 x6112                  3755 Riverside Drive
                                           Ottawa, Ont  K1G 3Z4

mwizard@eecs.cs.pdx.edu (Craig Nelson) (04/28/91)

Bob.Stout@p6.f506.n106.z1.fidonet.org (Bob Stout) writes:

>Zortech C/C++ has code generation switches to generate both 386 and 486 
>specific code. 

So does TCC/TCC++
 []====================================================================[]
 || Craig R. Nelson                | CCSofD Software Inc.              ||
 || Programmer                     | Beaverton, OR, 97005              ||
 || mwizard@eecs.ee.pdx.edu        | (unlisted on the net)             ||

gordon@osiris.cso.uiuc.edu (John Gordon) (04/29/91)

mwizard@eecs.cs.pdx.edu (Craig Nelson) writes:

>Bob.Stout@p6.f506.n106.z1.fidonet.org (Bob Stout) writes:

>>Zortech C/C++ has code generation switches to generate both 386 and 486 
>>specific code. 

>So does TCC/TCC++
> []====================================================================[]
> || Craig R. Nelson                | CCSofD Software Inc.              ||
> || Programmer                     | Beaverton, OR, 97005              ||
> || mwizard@eecs.ee.pdx.edu        | (unlisted on the net)             ||

	It does?  I thought that Turbo C only had switches to do 8088 or
186/286 code.

mwizard@eecs.cs.pdx.edu (Craig Nelson) (04/30/91)

gordon@osiris.cso.uiuc.edu (John Gordon) writes:

>mwizard@eecs.cs.pdx.edu (Craig Nelson) writes:

>>Bob.Stout@p6.f506.n106.z1.fidonet.org (Bob Stout) writes:

>>>Zortech C/C++ has code generation switches to generate both 386 and 486 
>>>specific code. 

>>So does TCC/TCC++

>	It does?  I thought that Turbo C only had switches to do 8088 or
>186/286 code.

	Really. When in doubt, you can always ask a Borland rep :) , but
I generate the code by first compiling command line with the -c -S options,
then use the /3 or .386 options with tasm.exe when recompiling the asm code.
Works for me ...

Later!!!

 []====================================================================[]
 || Craig R. Nelson                | CCSofD Software Inc.              ||
 || Programmer                     | Beaverton, OR, 97005              ||
 || mwizard@eecs.ee.pdx.edu        | (unlisted on the net)             ||

schmke@nntp-server.caltech.edu (Kevin Todd Schmidt) (04/30/91)

In article <2518@pdxgate.UUCP> mwizard@eecs.cs.pdx.edu (Craig Nelson) writes:
>gordon@osiris.cso.uiuc.edu (John Gordon) writes:
>
>>mwizard@eecs.cs.pdx.edu (Craig Nelson) writes:
>
>>>Bob.Stout@p6.f506.n106.z1.fidonet.org (Bob Stout) writes:
>
>>>>Zortech C/C++ has code generation switches to generate both 386 and 486 
>>>>specific code. 
>
>>>So does TCC/TCC++
>
>>	It does?  I thought that Turbo C only had switches to do 8088 or
>>186/286 code.
>
>	Really. When in doubt, you can always ask a Borland rep :) , but
>I generate the code by first compiling command line with the -c -S options,
>then use the /3 or .386 options with tasm.exe when recompiling the asm code.
>Works for me ...

But the asm code you are assembling is still only 8088/8086 code and
doesn't contain any 386/486 specific instructions so what are you gaining?
Nothing really!!!

______________________________________________________________________________
        _  ____   _  |                           |
       / // _  \ / / | Kevin T. Schmidt          | schmke@gap.cco.caltech.edu
      / // /_/ // /  | Jet Propulsion Laboratory |
 _   / // ____// /   | 4800 Oak Grove Dr.        |
/ /_/ // /    / /___ | M/S 301-355               |
\____//_/    /_____/ | Pasadena, CA  91109       |

sci240s@monu6.cc.monash.edu.au (mr w.j. ho) (05/01/91)

mwizard@eecs.cs.pdx.edu (Craig Nelson) writes:

>Bob.Stout@p6.f506.n106.z1.fidonet.org (Bob Stout) writes:

>>Zortech C/C++ has code generation switches to generate both 386 and 486 
>>specific code. 

>So does TCC/TCC++
         ^^^^^^^^^ Really????!! How ?
                   Info. on switches or manual page number would be great!
                   THANK YOU.

-- 
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
^   Wey Jing Ho   Tel: 61-3-5732567   E-mail : sci240s@monu6.cc.monash.edu.au  ^
^ Physics Dept., Monash University ( Caulfield Campus ), Melbourne, AUSTRALIA  ^
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^