davidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr) (12/04/90)
I have here a version of gcc for MS-DOS. I am contemplating posting
it, but am not sure how useful it would be. Feedback to me, please.
Pro:
C compiler
DOS extender
seems to allow big 32 bit progs under DOS
Free
Con:
Big. Huge. Megabytes.
No docs.
Needs 386 to run.
Preliminary
Unsupported
--
bill davidsen (davidsen@crdos1.crd.GE.COM -or- uunet!crdgw1!crdos1!davidsen)
VMS is a text-only adventure game. If you win you can use unix.grimesg@sj.ate.slb.com (George Grimes) (12/05/90)
Bill, Count this as one vote for posting gcc for MS-DOS. I'd like to see it and the fact that it requres a '386 is a reasonable contraint. Thanks, George
williams@umaxc.weeg.uiowa.edu (Kent Williams) (12/06/90)
I have ftp'ed and download the GCC package; the loader doesn't work on
my machine (386SX Neat Chipset, VIP MB, with DOS 4.01) I have heard
from one person who got it to compile a 'hello world' program.
The loader refuses to write the output file. It creates it OK, but it
bombs out of one of the other operations in the 'write_file' function
-- fstat'ing it to get its permission bits, and chmod'ing it to make
it executable. It would be very simple to fix this, since neither
operation has any relevance to DOS, but since the loader is broken, I
can't use it to load the fixed version ;-(.
I suspect it is a difference in DOS 3.X versus DOS 4.01 file semantics
-- since I'm getting ENOENT on an operation when the file exists,
apparently the directory entry isn't written out until the file is
closed.
If anyone has GCC working (in particular ld!!!) please mail to me.
Supposedly Uchida has a new version coming out in a few weeks, so hold
off posting!
--
Kent Williams --- williams@umaxc.weeg.uiowa.edu
"'Is this heaven?' --- 'No, this is Iowa'" - from the movie "Field of Dreams"
"This isn't heaven, ... this is Cleveland" - Harry Allard, in "The Stupids Die"davidsen@sixhub.UUCP (Wm E. Davidsen Jr) (12/06/90)
In article <3431@ns-mx.uiowa.edu> williams@umaxc.weeg.uiowa.edu.UUCP (Kent Williams) writes: | If anyone has GCC working (in particular ld!!!) please mail to me. | Supposedly Uchida has a new version coming out in a few weeks, so hold | off posting! Enough people besides me had trouble with this version, so I'm going to wait until a more useful version gets out. -- bill davidsen - davidsen@sixhub.uucp (uunet!crdgw1!sixhub!davidsen) sysop *IX BBS and Public Access UNIX moderator of comp.binaries.ibm.pc and 80386 mailing list "Stupidity, like virtue, is its own reward" -me
chris@iesd.auc.dk (Christian J. Callsen) (12/06/90)
Count this as a vote for posting gcc also. gcc stands as one of the best c-compilers that exists. -Chris C=>8^>
garym@cognos.uucp@uunet.uu.net (Gary Murphy) (12/06/90)
>>>>> On 5 Dec 90 00:16:36 GMT, grimesg@sj.ate.slb.com (George Grimes) said: GG> Bill, GG> Count this as one vote for posting gcc for MS-DOS. I'd like to see it and GG> the fact that it requres a '386 is a reasonable contraint. GG> Thanks, GG> George This is the first I've heard of this constraint. Is there no chance of building and/or running gcc on a 286 machine? Are there any other major constraints? -- o| Gary Murphy |o |------------------------------------------------------------------------| o| uunet!mitel!cunews!cognos!garym garym@cognos.uucp@uunet.uu.net |o | Cognos Inc. P.O. Box 9707 Ottawa K1G 3N3 (613) 738-1338 x5537 | o| "There are many things which do not concern the process" - Joan of Arc |o
jimp@cognos.UUCP (Jim Patterson) (12/07/90)
I'm also interested in gcc for MS-DOS. However, considering that this version is still "experimental", I'd like to suggest that any posting be held off until a reasonably stable version is available. I think the 'net could bear one such large posting, but two within a few months of each other might be a bit much. -- Jim Patterson Cognos Incorporated UUCP:uunet!mitel!cunews!cognos!jimp P.O. BOX 9707 PHONE:(613)738-1440 3755 Riverside Drive NOT a Jays fan (not even a fan) Ottawa, Ont K1G 3Z4
d88-rjo@svart.nada.kth.se (Rickard Jonsson) (12/07/90)
Did anybody get this program to work at all? After having serious problems with installing it ( The install program doesn't work ), and discovering it doesn't work with Quarterdeck's QEMM, it finally gave me: 'Error in COMMAND.COM', when I compiled a hello world program! And then it attempted to reboot the system. Has this happened to anybody else? When is the new version coming out? Besides these problems, the compiler looks VERY good and I think it will be a big hit - a must for every 386 user! /Rick ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ "Who said the friends you made in high school were yours to keep forever?" - Sunny of Miami Vice "As soon as I get a new book, I ALWAYS turn to the last page to see how it ends. That way, if I die before the ending, I'll know how it ends!" - Billy Crystal in "When Harry met Sally" ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
poffen@sj.ate.slb.com (Russ Poffenberger) (12/08/90)
In article <GARYM.90Dec6083731@cognos.uucp@uunet.uu.net> garym@cognos.uucp@uunet.uu.net (Gary Murphy) writes: > >>>>>> On 5 Dec 90 00:16:36 GMT, grimesg@sj.ate.slb.com (George Grimes) said: > > >GG> Bill, >GG> Count this as one vote for posting gcc for MS-DOS. I'd like to see it and >GG> the fact that it requres a '386 is a reasonable contraint. > >GG> Thanks, >GG> George > > >This is the first I've heard of this constraint. Is there no chance >of building and/or running gcc on a 286 machine? Are there any other >major constraints? > Gcc for MSDOS is a port of the famous gcc compiler for UNIX machines. This is a true 32 bit compiler, and I believe that the intent is a straight port (as much as possible) to remain a 32 bit compiler running only on 32 bit machines in protected mode, generating programs for same. A port to a 16 bit machine like an 80286 would be a major undertaking, and given that there are plenty of reasonably good MSDOS compilers out there (Borland, Zortech, MicroSoft, etc, etc), it wouldn't be very interesting. Russ Poffenberger DOMAIN: poffen@sj.ate.slb.com Schlumberger Technologies UUCP: {uunet,decwrl,amdahl}!sjsca4!poffen 1601 Technology Drive CIS: 72401,276 San Jose, Ca. 95110 (408)437-5254
davidsen@sixhub.UUCP (Wm E. Davidsen Jr) (12/09/90)
In article <9103@cognos.UUCP> jimp@cognos.UUCP (Jim Patterson) writes: | I'm also interested in gcc for MS-DOS. However, considering that this | version is still "experimental", I'd like to suggest that any posting | be held off until a reasonably stable version is available. I think there is no question that this version is not going to be useful to enough people to justify sending it out for a week. I got about two parts "post quick" notes and one part horror stories. For something only useful to 386 users in the first place that's not acceptable. By the time the backlog dies down I hope to have a better version. -- bill davidsen - davidsen@sixhub.uucp (uunet!crdgw1!sixhub!davidsen) sysop *IX BBS and Public Access UNIX moderator of comp.binaries.ibm.pc and 80386 mailing list "Stupidity, like virtue, is its own reward" -me
davel@booboo.SanDiego.NCR.COM (David Lord) (12/11/90)
In article <1990Dec7.120521.27553@nada.kth.se> d88-rjo@svart.nada.kth.se (Rickard Jonsson) writes: > Did anybody get this program to work at all? > Besides these problems, the compiler looks VERY good and I think it >will be a big hit - a must for every 386 user! Well call me clueless but could you explain that last statement? Aside from the fact that it doesn't work it looks very good? Just what would be so great about this compiler if it did work?
boylanr@silver.ucs.indiana.edu (ross boylan) (12/11/90)
d88-rjo@svart.nada.kth.se (Rickard Jonsson) writes: >After having serious problems with installing it ( The install program >doesn't work ), and discovering it doesn't work with Quarterdeck's QEMM, I'd like to know more about this--for sure the DOS extender is incompatible with Desqview and QEMM? That would be too bad.
kenny@uh.msc.umn.edu (Kenneth R. Goers) (12/11/90)
Mellow people, Just to protect the person porting GCC, I'd like to tell all you bufoons to lay off him. GCC works fine for me. I've ported a couple programs to DOS already that would not run under it previously, even recompiled gcc. And the guy posting about it out did say it was a pre-release, not the real thing. With all the warped clones out there, I'd say it's quite a while before it works on a majority of them. I still can't understand people bitching about something that's free. kenny.
mra@srchtec.UUCP (Michael Almond) (12/11/90)
In article <3105@uc.msc.umn.edu> kenny@uh.msc.umn.edu (Kenneth R. Goers) writes: > >Mellow people, > >Just to protect the person porting GCC, I'd like to tell all you >bufoons to lay off him. GCC works fine for me. I've ported a couple >programs to DOS already that would not run under it previously, even >recompiled gcc. And the guy posting about it out did say it was a >pre-release, not the real thing. With all the warped clones out >there, I'd say it's quite a while before it works on a majority of >them. > >I still can't understand people bitching about something that's free. I agree. The source code to GNU gcc is freely avialable, unlike a lot of software for the IBM-PC. If you find the compiler not functioning on your system, just get a hold of the source and dig in. --- Michael R. Almond (Georgia Tech Alumnus) mra@srchtec.uucp (registered) search technology, inc. mra%srchtec@salestech.com 4725 peachtree corners cir., suite 200 emory!stiatl!srchtec!mra norcross, georgia 30092 (404) 441-1457 (office) [search]: Systems Engineering Approaches to Research and Development
poffen@sj.ate.slb.com (Russ Poffenberger) (12/12/90)
In article <boylanr.660850366@silver> boylanr@silver.ucs.indiana.edu (ross boylan) writes: >d88-rjo@svart.nada.kth.se (Rickard Jonsson) writes: > >>After having serious problems with installing it ( The install program >>doesn't work ), and discovering it doesn't work with Quarterdeck's QEMM, > >I'd like to know more about this--for sure the DOS extender is incompatible >with Desqview and QEMM? That would be too bad. When you try to use it with QEMM loaded, QEMM reports an exception 13 and asks you if you want to unload QEMM and continue (which locked up for me), or abort the program, or reboot (something like that). I haven't tried it with HIMEM/EMM386 yet, but I suspect it will crash also. Too bad it isn't compliant. Probably just needs to conform to XMS, it may try to grap all of EMS directly. Russ Poffenberger DOMAIN: poffen@sj.ate.slb.com Schlumberger Technologies UUCP: {uunet,decwrl,amdahl}!sjsca4!poffen 1601 Technology Drive CIS: 72401,276 San Jose, Ca. 95110 (408)437-5254
williams@umaxc.weeg.uiowa.edu (Kent Williams) (12/12/90)
uchida san just sent me a new ld that does work with DOS 4.01. If anyone
has had trouble with the linker for gcc, mail me and I'll mail a copy.
He is working on the next release, and the comp.binaries.ibm.pc moderator
has said that gcc will be posted when it is stable. It is semi-stable
now, and is available from ocf.berkeley.edu
For those of you that are still having trouble
1. Get rid of everything in your config.sys and your autoexec.bat that uses
protected mode. This includes highmem.sys, 386max.sys, and
qemm386.sys. It may in fact work with one or more of those, but then
again it may not. I don't know if his dos extender is DPMI compliant.
If it isn't, you can get the source. Fix it, and mail diffs back to
uchida san.
2. If you are running DOS 4.01, then you'll need the fixed ld. I have
a copy and can mail it to whomever needs it.
3. Pay close attention to environment variables, and the config
file, gnucc.cfg. The config file is in the top level directory once
you install, you'll need to copy it into whatever work directory you
want to use. I'm not sure, but I think you can override some of these
things with environment variables. Try gcc -? for info. Here is my gnucc.cfg
-I/gnu/usr/include -O
-L/gnu/usr/lib
-msoft-float
4. Try gcc -n flubber.c. The -n option tells you what gcc is trying
to do to you. You can fool around with the result one command at a
time in order to come up with something useful.
--
Kent Williams --- williams@umaxc.weeg.uiowa.edu
"'Is this heaven?' --- 'No, this is Iowa'" - from the movie "Field of Dreams"
"This isn't heaven, ... this is Cleveland" - Harry Allard, in "The Stupids Die"davidsen@sixhub.UUCP (Wm E. Davidsen Jr) (12/12/90)
In article <3105@uc.msc.umn.edu> kenny@uh.msc.umn.edu (Kenneth R. Goers) writes: | I still can't understand people bitching about something that's free. In the first place I don't think most people are bitching, just noting problems. And if it were true that people shouldn't bitch, why do I get so much mail if something I post isn't just what the readers want? -- bill davidsen - davidsen@sixhub.uucp (uunet!crdgw1!sixhub!davidsen) sysop *IX BBS and Public Access UNIX moderator of comp.binaries.ibm.pc and 80386 mailing list "Stupidity, like virtue, is its own reward" -me
zlsiial@uts.mcc.ac.uk (A.V. Le Blanc) (12/12/90)
I got the gcc from ocf. It runs on some of our 386's and not
on others. An irritating aspect is that the file gnucc.cfg
must be in your current directory when you compile, and that
the default location of the include files, \gnu\usr\include,
is not searched without a special option. Also the SETUP.BAT
file sets up the wrong environment variables. I could fix these
problems -- I have already done some of them -- but I cannot
recompile, for example, gcc.c or the extender xm (run386), since
these require tasm and some other C compiler.
1) Has anyone rewritten the .ASM files as .s files so that
gas can assemble them?
2) Has anyone rewritten the gcc.c file so that gcc can compile
itself?
3) In cases in which gcc-compiled code does not require vast
amounts of memory, can gcc produce an ordinary .exe file?
4) What are the changes required to make ld work on other
machines? This would be easier for me than transferring
recompiled versions, especially since UUencoded files
do not make it here uncorrupted.
Yours,
A. V. Le Blanc
ZLSIIAL@UK.AC.MCC.CMSbmk@ux5.lbl.gov (Brian Kincaid) (12/13/90)
I have had similar experiences with the dos extender (run386.exe) that comes with the gcc-dos package. Using the Phar-Lap dos extender (also called run386.exe), however, things work very well under QEMM. I think that's because QEMM supports VCPI and Phar-Lap uses VCPI. You get the Phar-Lap thing when you buy Phar-Lap or Microway C or F77. The package worked fine this way, but when I tried to compile the math library for with the 387 option (not the emulator mode), various compile time bugs showed up, and I decided to wait for the next version. Also, there are a few routine hacks necessary in the area of makefile options, directory names, etc, to make everything work, but hey, it's free. I am eagerly awaiting the next release of this great tool.
John.Ladwig@p0.f341.n282.z1.FIDONET.ORG (John Ladwig) (12/18/90)
In a message to All <16 Dec 90 04:09> Kenneth R. Goers wrote:
KR> From: kenny@uh.msc.umn.edu (Kenneth R. Goers)
KR> Date: 11 Dec 90 04:12:26 GMT
KR> Organization: Minnesota Supercomputer Center, Minneapolis, MN
KR> Message-ID: <3105@uc.msc.umn.edu>
KR> Newsgroups: comp.binaries.ibm.pc.d
KR> I still can't understand people bitching about something that's free.
My biggest problem with source code distributions exactly..
I can't bitch about missing features or bad implementations, I have to
add/fix them. :-) :-)
--
John Ladwig -- Via FidoNet using ufgate at 1:282/341
INTERNET: John.Ladwig@p0.f341.n282.z1.FIDONET.ORG
UUCP: ...{amdahl,hpda}!bungia!viper!terrabit!341.0!John.Ladwig