Gpc May 2001

gpc@gnu.de

26 participants
40 discussions

Module/unit could not be imported
by Casper ter Kuile 29 May '01

29 May '01

Hi, I encounter the following problem regarding GPC: I have a program with the statement: uses {Crt,Dos,Larrayd,} sysutils, initiadel, goodiesdel, pathsdel, ozsimu3zdel, simuldel, decondel, shiftdel, efuvriv3del; And I try to compile and link this and receive this error: shic3_011del.p:36: module/unit interface `Sysutils' could not be imported Anyone an idea what is going wrong and how to fix it? Thanks! Casper. Casper ter Kuile, Lab. for Radiation Research (LSO), National Institute for Public Healthg and the Environment (RIVM) PB 1, 3720 BA, Bilthoven Building A6, roomr 104, pb 33 Phone: 030-2742647 (030-2742515) E-mail: casper.ter.kuile(a)rivm.nl

3 2

BUG report for gpc-20010516.tar.gz
by Mirsad Todorovac 27 May '01

27 May '01

Dear Sir or Madam, I have to report a bug in your code which prevents your PASCAL compiler to be built unless -DEGCS -DEGCS92 and -DEGCS95 are all defined (using 2.95.3 back-end). This has not been set by ``configure'', so manual intervention was needed. I can't tell if defining -DEGCS97 would be better than not, but I was satisfied (as for now) with having a PASCAL running on the system!!! (alpha-dec-osf4.0b box, gcc-2.95.3 compiler suite, gmake, gsed) Fix: define "-DEGCS -DEGCS92 -DEGCS95" in CFLAGS, meaning run: % gmake CFLAGS="-O2 -DEGCS -DEGCS92 -DEGCS95" RTSFLAGS=-Wall The compiling breaks, for example, on conflicting definitions of the return type of xmalloc(), one in gbe.h having char * as return value and the other, in libiberty.h having PTR as return value. Other problems are some conflicting prototypes having const defined in global includes, but ommited in /p/ source tree ... Also, never in the installation instruction it isn't mentioned that GPC needs a smarter sed - this could be addressed by ``configure'' script too. Also, very helpful would be if the building of PASCAL compiler wouldn't include "fixing" of C include files again, which takes anormous amount of time on Alpha/Digital-Unix box. Thank you very much for your time. Mirsad Todorovac

2 1

Thanks! (was RE: [Fwd: Re: Newbie question ...])
by da Silva, Joe 25 May '01

25 May '01

Many thanks for that clarification, Frank! As for the EXE compression by WDOSX, the documentation certainly suggests this is genuine compression, not simply stripping the symbol table ... :-) Best regards, Joe. > -----Original Message----- > From: Frank Heckenbach [SMTP:frank@g-n-u.de] > Sent: Friday, 25 May 2001 1:43 > To: gpc(a)gnu.de > Subject: Re: [Fwd: Re: Newbie question ...] > > da Silva, Joe wrote: > > > I have now tried WDOSX 0.96 and it works with the GPC > > "Hello world" test program! Not only did it convert the > > executable to a stand-alone application (built-in DPMI > > services), it also compressed the executable to nearly > > half it's original size! :-) > > Nice, but maybe it only stripped the symbol table (like `strip' > does). > > > OK, one final newbie question : The online documentation > > (http://agnes.dida.physik.uni-essen.de/~gnu-pascal/gnu.html) > > says that GPC can be used to compile "commercial, > > non-free" applications. Is this a modification of the standard > > GNU GPL, or is this an interpretation of the GPL? > > It's a special exception. All files of the Run Time System contain a > license statement like the following: > > : This file is part of GNU Pascal. > : > : GNU Pascal is free software; you can redistribute it and/or modify > : it under the terms of the GNU General Public License as published by > : the Free Software Foundation; either version 2, or (at your option) > : any later version. > : > : GNU Pascal is distributed in the hope that it will be useful, > : but WITHOUT ANY WARRANTY; without even the implied warranty of > : MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the > : GNU General Public License for more details. > : > : You should have received a copy of the GNU General Public License > : along with GNU Pascal; see the file COPYING. If not, write to the > : Free Software Foundation, 59 Temple Place - Suite 330, Boston, MA > : 02111-1307, USA. > : > : As a special exception, if you link this file with files compiled > : with a GNU compiler to produce an executable, this does not cause > : the resulting executable to be covered by the GNU General Public > : License. This exception does not however invalidate any other > : reasons why the executable file might be covered by the GNU General > : Public License. > > The compiler itself does not contain the exception, but that's not > necessary, because the output of a GPL's program does not > automatically fall under the GPL (unless it does for other reasons). > So, only if you copy/link some actual code from the compiler's > sources in your program, it would have to be GPL'd. > > Please note that some units distributed with GPC, or contributed on > other places have different licences, so check the license of each > unit you use (but that applies to any compiler, of course). > > > The > > reason I ask, is that a number of people, including (I think) > > DJ Delorie (of DJGPP fame), have a different interpretation. > > I apologise if this question has been asked before, but > > no matter how many times I read the GPL or look at the > > information of the FSF and GNU, I just can't figure it out. > > AFAIK, the DJGPP library is released under the LGPL which does allow > linking to non-free programs, though under some conditions (e.g., > provide relinkable object files, but see the LGPL for the exact > details). Of course, as above, this does not apply for using the > code of GCC or any other GPL'd program ported to DJGPP. > > Frank > > -- > Frank Heckenbach, frank(a)g-n-u.de, http://fjf.gnu.de/ > GPC To-Do list, latest features, fixed bugs: > http://agnes.dida.physik.uni-essen.de/~gnu-pascal/todo.html

1 0

Old Bug: Global variable not initialised properly
by Ian Thurlbeck 24 May '01

24 May '01

Dear All After 2 years of using gpc I thought I'd better update to a newer release, so I have installed gcc-2.95.3 and gpc-20010512. It appears than an old bug has resurfaced. A global variable declared in a module is not initialised properly. The following code does NOT print out the string 'OK': -------main.h------------ var s: string(40); external; procedure print_globals; external; ------------------------- -------main.p------------ program main(input,output); #include "main.h" begin s:='OK'; print_globals; end. ------------------------- -------lib.p------------- module lib(input,output); #include "main.h" var s:string(40); procedure print_globals; begin writeln(s); end; end. ------------------------- Compile with 'gpc -o main main.p lib.p' If you move the declaration of string 's' from lib.p to main.p it works. This was fixed about 2 years ago (by Peter I think) but it seems to have gotten un-fixed. The initialisation code for the Main program, when the variable 's' is defined in the module in lib.p, is not correct. If you do: # gpc -c lib.p # gpc -S main.p # vi main.s and add "movl $40,S" before .L6: in init_pascal_main_program, then # as -V -Qy -o main.o main.s GNU assembler version 2.10.91 (i386-redhat-linux) using BFD version 2.10.91.0.2 # gpc -o main main.o lib.o # ./main OK you get the expected result. Hope there is enough information to fix the problem. Many thanks Ian -- Ian Thurlbeck http://www.stams.strath.ac.uk/ Statistics and Modelling Science, University of Strathclyde Livingstone Tower, 26 Richmond Street, Glasgow, UK, G1 1XH Tel: +44 (0)141 548 3667 Fax: +44 (0)141 552 2079 Â°Ã¯S

3 7

RE: [Fwd: Re: Newbie question ...]
by da Silva, Joe 24 May '01

24 May '01

Good news! I have now tried WDOSX 0.96 and it works with the GPC "Hello world" test program! Not only did it convert the executable to a stand-alone application (built-in DPMI services), it also compressed the executable to nearly half it's original size! :-) OK, one final newbie question : The online documentation (http://agnes.dida.physik.uni-essen.de/~gnu-pascal/gnu.html) says that GPC can be used to compile "commercial, non-free" applications. Is this a modification of the standard GNU GPL, or is this an interpretation of the GPL? The reason I ask, is that a number of people, including (I think) DJ Delorie (of DJGPP fame), have a different interpretation. I apologise if this question has been asked before, but no matter how many times I read the GPL or look at the information of the FSF and GNU, I just can't figure it out. Regards, Joe. > -----Original Message----- > From: da Silva, Joe [SMTP:Joe.daSilva@emailmetering.com] > Sent: Wednesday, 23 May 2001 18:49 > To: gpc > Subject: RE: [Fwd: Re: Newbie question ...] > > Thanks for the info., Maurice. > > WDOSX is a DOS-extender for NASM, TASM, MASM, Borland C, > Borland Delphi, MSC, Watcom C, TMT Pascal and DJGPP ... > home page : http://www.wuschel.demon.co.uk/ > > I have now downloaded the pmode13b archive, but when I follow > the links for COFF, I reach a "dead end" (can't find domain) at : > http://www.abwillms.demon.co.uk/prog/kernel2.zip :-( > > Anyway, it looks like a new version of WDOSX was released > just last month, so I will try again with this release ... > > Joe. > > -----Original Message----- > > From: Maurice Lombardi [SMTP:Maurice.Lombardi@ujf-grenoble.fr] > > Sent: Wednesday, 23 May 2001 1:45 > > To: gpc > > Subject: [Fwd: Re: Newbie question ...] > > > > > > da Silva, Joe wrote: > > > > > Greetings. > > > > > > I hope I haven't missed the answer to this question somewhere (I have > > > looked at the online information, but couldn't find it ...) : > > > > > > Is there a stub/extender available, for the DJGPP build of the > > compiler, > > > that provides DPMI services for an application, instead of loading > > > CWSDPMI? I tried stubbing the "Hello world" example using WDOSX, > > > but this simply crashed the computer ... > > > > I do not know what is WDOSX > > DJGPP gives pmode13b/s in the v2misc directory (free of course). > > This stub can be bound to the COFF image to give a standalone > executable. > > See the doc included in the distribution. > > Maurice > > > > -- > > Maurice Lombardi > > Laboratoire de Spectrometrie Physique, > > Universite Joseph Fourier de Grenoble, BP87 > > 38402 Saint Martin d'Heres Cedex FRANCE > > Tel: 33 (0)4 76 51 47 51 > > Fax: 33 (0)4 76 51 45 44 > > mailto:Maurice.Lombardi@ujf-grenoble.fr

2 1

RE: [Fwd: Re: Newbie question ...]
by da Silva, Joe 23 May '01

23 May '01

Thanks for the info., Maurice. WDOSX is a DOS-extender for NASM, TASM, MASM, Borland C, Borland Delphi, MSC, Watcom C, TMT Pascal and DJGPP ... home page : http://www.wuschel.demon.co.uk/ I have now downloaded the pmode13b archive, but when I follow the links for COFF, I reach a "dead end" (can't find domain) at : http://www.abwillms.demon.co.uk/prog/kernel2.zip :-( Anyway, it looks like a new version of WDOSX was released just last month, so I will try again with this release ... Joe. > -----Original Message----- > From: Maurice Lombardi [SMTP:Maurice.Lombardi@ujf-grenoble.fr] > Sent: Wednesday, 23 May 2001 1:45 > To: gpc > Subject: [Fwd: Re: Newbie question ...] > > > da Silva, Joe wrote: > > > Greetings. > > > > I hope I haven't missed the answer to this question somewhere (I have > > looked at the online information, but couldn't find it ...) : > > > > Is there a stub/extender available, for the DJGPP build of the > compiler, > > that provides DPMI services for an application, instead of loading > > CWSDPMI? I tried stubbing the "Hello world" example using WDOSX, > > but this simply crashed the computer ... > > I do not know what is WDOSX > DJGPP gives pmode13b/s in the v2misc directory (free of course). > This stub can be bound to the COFF image to give a standalone executable. > See the doc included in the distribution. > Maurice > > -- > Maurice Lombardi > Laboratoire de Spectrometrie Physique, > Universite Joseph Fourier de Grenoble, BP87 > 38402 Saint Martin d'Heres Cedex FRANCE > Tel: 33 (0)4 76 51 47 51 > Fax: 33 (0)4 76 51 45 44 > mailto:Maurice.Lombardi@ujf-grenoble.fr

1 0

[Fwd: Re: Newbie question ...]
by Maurice Lombardi 22 May '01

22 May '01

da Silva, Joe wrote: > Greetings. > > I hope I haven't missed the answer to this question somewhere (I have > looked at the online information, but couldn't find it ...) : > > Is there a stub/extender available, for the DJGPP build of the compiler, > that provides DPMI services for an application, instead of loading > CWSDPMI? I tried stubbing the "Hello world" example using WDOSX, > but this simply crashed the computer ... I do not know what is WDOSX DJGPP gives pmode13b/s in the v2misc directory (free of course). This stub can be bound to the COFF image to give a standalone executable. See the doc included in the distribution. Maurice -- Maurice Lombardi Laboratoire de Spectrometrie Physique, Universite Joseph Fourier de Grenoble, BP87 38402 Saint Martin d'Heres Cedex FRANCE Tel: 33 (0)4 76 51 47 51 Fax: 33 (0)4 76 51 45 44 mailto:Maurice.Lombardi@ujf-grenoble.fr

1 0

Newbie question ...
by da Silva, Joe 22 May '01

22 May '01

Greetings. I hope I haven't missed the answer to this question somewhere (I have looked at the online information, but couldn't find it ...) : Is there a stub/extender available, for the DJGPP build of the compiler, that provides DPMI services for an application, instead of loading CWSDPMI? I tried stubbing the "Hello world" example using WDOSX, but this simply crashed the computer ... Regards, Joe.

1 0

reset() regression
by Ian Thurlbeck 21 May '01

21 May '01

Hello all I've just noticed a small difference with the gpc-20010512 compiler from the one I was using (gpc-19991030). If you do: program restest(input,output); var s:string(40); t:text; begin s:=""; {$i-} reset(t,s); if (IOResult = 0) then writeln('File exists') else writeln('File does not exist'); end. the 1999 compiled code produces "File does not exist" whilst the 2001 code produces "File exists". Not sure if you consider this worth fixing given I can test for "" before I do the reset() easily enough. ( A "" filename is produced by some legacy code I maintain, when someone presses RETURN to a filename: prompt. Not very sensible to go passing it to reset(), but there you go. ) Yours Ian -- Ian Thurlbeck http://www.stams.strath.ac.uk/ Statistics and Modelling Science, University of Strathclyde Livingstone Tower, 26 Richmond Street, Glasgow, UK, G1 1XH Tel: +44 (0)141 548 3667 Fax: +44 (0)141 552 2079

2 1

Re: Compiling latest gpc on latest cygwin (fwd)
by Matthias Braun 16 May '01

16 May '01

Forgot the attachements :) I don't attach gpcpp.c (cause it's 360k) I'll send it to anyone asking. On Wed, 16 May 2001, Prof. A Olowofoyeku (The African Chief) wrote: > On 15 May 2001, at 16:39, Matthias Braun wrote: > > > Hello, > > I've got problems with compiling gpc snapshot from 12.5 and latest cygwin > > with gcc 2.95.3-4 I did like described in the faq, and I had some problems > > I solved: > > -There was no patch for gcc 2.95.3-4 so I took the one for 2.95.3 > > -GPLUSPLUS_INCLUDE_DIR and TOOL_INCDLUDE_DIR wasn't declared in > > gpcpp.c, I found an #undeclare in cygwin.h and deletet it. > > I am not sure that you should do that. I think you should trust that > whoever undeclared it knew what they were doing. I normally can compile > GPC for both Cygwin and Mingw without touching anything other than to > apply the GPC patch. So I looked into the cccp.c file which seems to be the c counterpart to gpcpp.c, here they did a #ifdef TOOL_INCLUDE_DIR around the critical parts, so I did the same for gpcpp.c and restored original cygwin.h. But this gives the same effect as before. > > After that compiling worked, but if I use the compile now, the following > > appears: (gpc -v test.pas) > > Do you have all the latest GNU tools and headers and libraries? You > shouldn't have had to change cygwin.h to be able to build GPC > successfully. Yes, I used the setup utility from cygwin.com and installed experimental releases (gcc package isn't marked experimental anyway). I installed all the tools there. > [...] > > GNU Pascal version 2.95.3-4 (cygwin special) (i586-pc-cygwin32) compiled > > by GNU C version 2.95.3-3 (cygwin special). GNU Pascal version is actually > > 20010512, based on gcc-2.95.3-4 (cygwin special) > > as -o /PlatteC/WINDOWS/TEMP/ccc85291 /PlatteC/WINDOWS/TEMP/ccb85291 > > gpc.exe: Internal compiler error in `do_spec_1', at p/gpc.c:4500 > > I have never seen this problem before. The GPC that you built is > obviously broken. Why? I have no idea. Could it be a problem with your > "specs" file? Perhaps Peter or Frank might have an idea what this error > means. I reinstalled everything, I did some tests, compiling works if I use the following sequence: gpc -c -o test.o test.pas gcc test.o -o test.exe -lgpc So It seems only an error reading the specs file, but I don't understand the spec file syntax :/. So I've attached it along with the changed gpcpp.c to this mail. Matze

3 2

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

1996

1995

Gpc May 2001