Gpc March 2000

gpc@gnu.de

28 participants
42 discussions

enumerated types problem
by miller＠otago.ac.nz 08 Mar '00

08 Mar '00

Greetings from New Zealand, It appears to me that in some situations gpc refuses to compile because it loses information about enumerated types defined in units. I'm using gpc version 19990118, based on gcc-2.8.1, under linux. (I gave up on using gpc under OS/2). Below is a minimal example--about 50 lines of source code divided across 2 units and a main program, all in separate files. Please tell me if I've done something wrong or if you know of a work-around. Thanks, Jeff Miller miller(a)otago.ac.nz http://psy.otago.ac.nz/miller Unit gnutstb1; { this unit is in file gnutstb1.pas } Interface Type MyType = (eenie, meenie, minie, mo, eenie2, meenie2, minie2, mo2, eenie3, meenie3, minie3, mo3, eenie4, meenie4, minie4, mo4, eenie5, meenie5, minie5, mo5, eenie6, meenie6, minie6, mo6, eenie7, meenie7, minie7, mo7, eenie8, meenie8, minie8, mo8, eenie9, meenie9, minie9, mo9 ); Basic = Object BoolArr : Array[MyType] of Boolean; Constructor Construct; End; Implementation Constructor Basic.Construct; Begin BoolArr[eenie] := True; BoolArr[meenie] := False; BoolArr[minie] := False; BoolArr[mo] := False; End; End. { end of file gnutstb1.pas } Unit gnutstb2; { this unit is in file gnutstb2.pas } Interface Uses gnutstb1; Type Derived1 = Object(Basic) Unused : Char; Constructor Construct; End; Implementation Constructor Derived1.Construct; Begin Basic.Construct; BoolArr[meenie] := True; End; End. { end of file gnutstb2.pas } Program GnuTstB3; { This main program is in file gnutstb3.pas } Uses gnutstb1, gnutstb2; Type Derived2 = Object(Derived1) Unused2 : Char; Constructor Construct; End; Constructor Derived2.Construct; Begin Derived1.Construct; BoolArr[minie] := True; { gpc says "type mismatch in array index" } End; Var D2 : Derived2; Begin D2.Construct; If D2.BoolArr[minie] { gpc again says "type mismatch in array index" } Then Writeln('True') Else Writeln('False'); End. { end of file gnutstb3.pas } When I issue the command gpc gnutstb3.pas --automake --borland-pascal I get the output gnutstb3.pas: In function `Derived2_Construct': gnutstb3.pas:17: type mismatch in array index gnutstb3.pas: In function `program_Gnutstb3': gnutstb3.pas:24: type mismatch in array index These files were made OK (as far as I can tell): gnutstb1.o gnutstb2.gpi gnutstb1.gpi gnutstb2.o

2 1

GC & GPC question.
by Joe Blask 08 Mar '00

08 Mar '00

One of the things that steers me away from Borland Pascal is its lack of a runtime garbage collector. Does GNU Pascal have a runtime garbage collector? Sincerely, ==Joe Blask==

4 6

RE: Mingw version of GPC based on latest GCC 2.95?
by Mason Ip 08 Mar '00

08 Mar '00

It is very nice of you, Dr A Olowofoyeku, to provide such a binary. But why is it labelled as gpc-19991111-etc? The latest source I could find was gpc-19991030, from ftp://agnes.dida.physik.uni-essen.de/gnu-pascal/alpha/. BTW, could gpc provide a CVS source, so that we can fetch the latest snapshot? > ---------- > From: Dr A Olowofoyeku[SMTP:laa12@law.keele.ac.uk] > Sent: Wednesday, March 08, 2000 4:22 AM > To: Mark Taylor; gpc(a)gnu.de > Subject: Re: Mingw version of GPC based on latest GCC 2.95? > > On 8 Mar 00, at 9:58, Mark Taylor wrote: > > > > > Has anyone built the Cygwin or Mingw version of GPC based on the > > latest GCC 2.95? > > > > Does anyone plan to do this? > > I have built a Cygwin compiler (which is also able to produce Mingw > programs), for gcc 2.95.2. You can download it from: > ftp://agnes.dida.physik.uni-essen.de/home/chief/ > > > > If not, I will have a go myself, but this might be a big effort as I > > have never built GCC or GPC before. If anyone can offer pointers I > > would be most gratefull. > > It is relatively straightforward to build a Cygwin compiler, as long as > you have the full and latest Cygwin distribution. I am due to write > some notes for building GPC on Cygwin, but I don't have the time > yet. If you can understand bash scripts, I can send you the > (commented) bash script that I use to automate all the steps. You > should be able to work from that. > > > If successfull, I would then be in a position > > to offer my services in helping with further development of the > > compiler. Something I would be interested in would be recoding GPC in > > Pascal (I am sure this would make it easier for more people to > > contribute their help) - would this be feasible, or does GPC *have* to > > be coded in C for some reason? > > Yes. AFAIK all GNU compilers *have* to be coded in C. Large parts > of the GPC runtime library are of course coded in Pascal. > > Best regards, The Chief > ----- > Dr Abimbola A Olowofoyeku (The African Chief) > Email: African_Chief(a)bigfoot.com > Author of Chief's Installer Pro v5.22 for Win32 > http://www.bigfoot.com/~African_Chief/chief32.htm >

2 1

Mingw version of GPC based on latest GCC 2.95?
by Mark Taylor 08 Mar '00

08 Mar '00

Has anyone built the Cygwin or Mingw version of GPC based on the latest GCC 2.95? Does anyone plan to do this? If not, I will have a go myself, but this might be a big effort as I have never built GCC or GPC before. If anyone can offer pointers I would be most gratefull. If successfull, I would then be in a position to offer my services in helping with further development of the compiler. Something I would be interested in would be recoding GPC in Pascal (I am sure this would make it easier for more people to contribute their help) - would this be feasible, or does GPC *have* to be coded in C for some reason? Mark. -- Mark Taylor, Department of Corporate Information & Computing Services, Extension 21145. (0114) 222 1145 http://www.shef.ac.uk/misc/personal/ad1mt The opinions expressed in this email are mine and not those of the University of Sheffield.

3 2

GPC and X-Windows
by ROMANIUK ALEŠ 08 Mar '00

08 Mar '00

Hi ! I have been using GPC on Sun/Solaris for a while, and at this point I want to know, If there exist X-Windows programing library, and If so, where can I get it. Regards Alex ============================ Alex Romaniuk, External Computer Systems (Technical Departement) Tel.+386 61 1722543 (Fax. 1722590) http://www.mobitel.si, Slovenia ===========================

3 2

Re: Cannot produce executable
by Russ Whitaker 02 Mar '00

02 Mar '00

Using gpc version 19991030, based on 2.95.2 19991024 (release) output is: HELLO yes Hope this helps Russ <russwhit(a)mind.net> ---------- > From: miller(a)otago.ac.nz > To: gpc(a)gnu.de > Cc: miller(a)psy.otago.ac.nz > Subject: Cannot produce executable > Date: Wednesday, March 01, 2000 2:55 PM > > Greetings, > > I have some statistical programs that I would like to release under GPL. > They were originally written in Borland Pascal and Speed Pascal (the latter > is a borland-like pascal for OS/2), and all the code compiles and runs fine > using those two compilers. Now I am trying to get everything to > compile and run under gpc, so that the code will be useful to more > people. It looks like I'll have a lot of questions along the way, > and I'd appreciate any help that can be provided. > > I'm using a version of gpc that I installed as a binary from the file > gpc-19990118_i386-pc-emx0_9d.zip > > I can compile and run little test programs that use no units, > or that use only type definitions from units, but I get "undefined symbol" > messages if the main program refers to any variables or procedures > defined in the interface part of the unit. > > Here is a small example, adapted from one posted recently. > > The file def.pas has the following unit: > > UNIT modul; > INTERFACE > > TYPE > CARDINAL = 0..65535; > LONGBITSET = SET OF 0..63; > > Var ModS : Integer; > > PROCEDURE HELLO; > > IMPLEMENTATION > > PROCEDURE HELLO; > BEGIN > WRITELN('HELLO'); > END; > > END. > > The file test.pas has this test program. > > program test; > USES modul; > var > i:CARDINAL; > c:LONGBITSET; > begin > ModS := 0; > Hello; > c:=[1]; > i:=1; > if (i in c) then writeln('yes'); > end. > > gpc -c def.pas works fine, but > gpc test.pas > yields the messages > d:\tcpip\tmp\ccc00799: Undefined symbol _Mods referenced from text segment > d:\tcpip\tmp\ccc00799: Undefined symbol _Hello referenced from text segment > > d:\tcpip\tmp is my temp directory, and I noticed that running gpc leaves > a lot lot of small files called cca000*.exe in the temp directory. > > Can anyone tell me what I'm doing wrong? > > Thank you for your time, > > Jeff Miller > Dept of Psychology > University of Otago > Dunedin, New Zealand > miller(a)otago.ac.nz > http://psy.otago.ac.nz/miller >

1 0

Cannot produce executable
by miller＠otago.ac.nz 02 Mar '00

02 Mar '00

Greetings, I have some statistical programs that I would like to release under GPL. They were originally written in Borland Pascal and Speed Pascal (the latter is a borland-like pascal for OS/2), and all the code compiles and runs fine using those two compilers. Now I am trying to get everything to compile and run under gpc, so that the code will be useful to more people. It looks like I'll have a lot of questions along the way, and I'd appreciate any help that can be provided. I'm using a version of gpc that I installed as a binary from the file gpc-19990118_i386-pc-emx0_9d.zip I can compile and run little test programs that use no units, or that use only type definitions from units, but I get "undefined symbol" messages if the main program refers to any variables or procedures defined in the interface part of the unit. Here is a small example, adapted from one posted recently. The file def.pas has the following unit: UNIT modul; INTERFACE TYPE CARDINAL = 0..65535; LONGBITSET = SET OF 0..63; Var ModS : Integer; PROCEDURE HELLO; IMPLEMENTATION PROCEDURE HELLO; BEGIN WRITELN('HELLO'); END; END. The file test.pas has this test program. program test; USES modul; var i:CARDINAL; c:LONGBITSET; begin ModS := 0; Hello; c:=[1]; i:=1; if (i in c) then writeln('yes'); end. gpc -c def.pas works fine, but gpc test.pas yields the messages d:\tcpip\tmp\ccc00799: Undefined symbol _Mods referenced from text segment d:\tcpip\tmp\ccc00799: Undefined symbol _Hello referenced from text segment d:\tcpip\tmp is my temp directory, and I noticed that running gpc leaves a lot lot of small files called cca000*.exe in the temp directory. Can anyone tell me what I'm doing wrong? Thank you for your time, Jeff Miller Dept of Psychology University of Otago Dunedin, New Zealand miller(a)otago.ac.nz http://psy.otago.ac.nz/miller

2 1

Re: install gpc
by Frank Heckenbach 02 Mar '00

02 Mar '00

Ronald Bianchin wrote: > I've a problem with install gpc. > After unpack gpc-19990118 I've move the files > to /opt/gpc, copy /p/* to gcc/p/* and > execute patch -s -p1 < p/diffs/gcc-2.8.1.diff. You have to unpack the sources of GCC in a matching version before. There are several possible versions as shown in the diff file names. 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

Re: Cannot produce executable
by Frank Heckenbach 02 Mar '00

02 Mar '00

miller(a)otago.ac.nz wrote: > I have some statistical programs that I would like to release under GPL. > They were originally written in Borland Pascal and Speed Pascal (the latter > is a borland-like pascal for OS/2), and all the code compiles and runs fine > using those two compilers. Now I am trying to get everything to > compile and run under gpc, so that the code will be useful to more > people. It looks like I'll have a lot of questions along the way, > and I'd appreciate any help that can be provided. > > I'm using a version of gpc that I installed as a binary from the file > gpc-19990118_i386-pc-emx0_9d.zip > > I can compile and run little test programs that use no units, > or that use only type definitions from units, but I get "undefined symbol" > messages if the main program refers to any variables or procedures > defined in the interface part of the unit. Try compiling with `--automake' which also compiles units automatically, so don't have to compile them with `-c' before. If this doesn't work (automake is a bit buggy under Dos), try `--autobuild' instead. 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

Re: Bug in UNIT definitions
by Frank Heckenbach 02 Mar '00

02 Mar '00

Jose Oliver Gil wrote: > Hello again, > > Now, I have updated to gpc version 19990118, based > on egcs-2.91.60 19981201 (egcs-1.1.1 release). > > I have found out another bug with this version, > when I define two local variables, with the same > name, in two different units but in the same file, > e.g. > > UNIT TEST1; > INTERFACE > IMPLEMENTATION > var Repeated:integer; > END. > > UNIT TEST2; > INTERFACE > IMPLEMENTATION > var Repeated:integer; > END. > > the following error is displayed > > /tmp/cca09330.s: Assembler messages: > /tmp/cca09330.s:31: Error: Ignoring attempt to re-define symbol > /tmp/cca09330.s:31: Error: Rest of line ignored. First ignored character > is `,'. > > If I am defining the variable Repeated in the implementation zone, > why does the assembler think that it's a re-definition? > Is it a compiler bug? Yes. Multiple units/modules/programs in one source file are not supported perfectly. It will be fixed sometime, but don't hold your breath... For now, it's best to split your sources into separate files. 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

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 March 2000