Gpc

gpc@gnu.de

5400 discussions

crashing while killing list
by Toby Ewing 11 Jul '01

11 Jul '01

Hi, all I just upgraded to a more recent version of GPC, 2001something. I don't know if that's the root cause, but I'm having trouble killing a linked list. I get a page fault or a GPF (I'm running under win98) when trying to delete the last or next to last record. I've marked the line named in the call traceback. I took my Linsert and Lwipe procedures and put them into a test program (see below). It runs perfectly. But when I run them in the program I'm actually working on -- the program I took the procedures from -- it crashes in Lwipe. Nothing else in the big program even refers to the list. Any suggestions? It's frustrating to be stopped on such an elementary problem. Lists and I have been friends for years, and I'm supposed to be stuck on much more interesting problems! Toby _____________________________ Program ListTest; type PRec = ^Rec; Rec = record ri : integer; Next : PRec; end; var Root : PRec; i : integer; {----------} Procedure LWipe(var Root : PRec); var point : PRec; begin while (Root^.next <> Nil) do begin {<<== program crashes on this line} write('deleting ', root^.ri, ' '); point := Root; Root := Root^.next; Dispose(point); end; if Root <> Nil then begin write('deleting ', root^.ri); Dispose(Root); Root := Nil; end; end; {----------} Procedure LInsert(var num : integer; var Root : PRec); {insert a new record at the *front* of the list} var Site : Rec; temp : PRec; begin with Site do begin ri := num; next := Nil; end; if Root = Nil then begin {starting the list} New(Root); Root^ := Site; end else begin {list already exists} New(temp); {get new memory} temp^ := Site; {move memory} temp^.next := Root; {repoint pointers} Root := temp; end; end; {LInsert} {----------} Begin for i := 1 to 100 do LInsert(i, Root); LWipe(Root); End.

2 1

Fwd: crashing while killing list
by Toby Ewing 11 Jul '01

11 Jul '01

*Groan* once again, programing is primarily a lesson in humility. rpewing(a)lbl.gov wrote: > I just upgraded to a more recent version of GPC, 2001something. I don't > know if that's the root cause, but I'm having trouble killing a linked > list. I get a page fault or a GPF (I'm running under win98) when trying > to delete the last or next to last record. I've marked the line named > in the call traceback. .. I forgot to initialize the pointer to Nil. I think BP did this for me, so I got out of the habit. A corrected version is below. Toby > _____________________________ > > Program ListTest; > > type > PRec = ^Rec; > Rec = record > ri : integer; > Next : PRec; > end; > > var > Root : PRec; > i : integer; > {----------} > > Procedure LWipe(var Root : PRec); > var point : PRec; > begin > while (Root <> Nil) do begin > write('deleting ', root^.ri, ' '); > point := Root; > Root := Root^.next; > Dispose(point); > end; > end; > {----------} > > Procedure LInsert(var num : integer; > var Root : PRec); > {insert a new record at the *front* of the list} > var > Site : Rec; > temp : PRec; > begin > with Site do begin > ri := num; > next := Nil; > end; > if Root = Nil then begin {starting the list} > New(Root); > Root^ := Site; > end else begin {list already exists} > New(temp); {get new memory} > temp^ := Site; {move memory} > temp^.next := Root; {repoint pointers} > Root := temp; > end; > end; {LInsert} > {----------} > > Begin Root := Nil; {<< the line I forgot} > for i := 1 to 100 do LInsert(i, Root); > LWipe(Root); > End. >

1 0

a string question
by Adam Naumowicz 11 Jul '01

11 Jul '01

Hello, I have a basic question concerning strings and array memory representation in GPC. To my amusement I noticed that in the following program the string assigned to an array element cannot be then restored from the array itself. So probably an array of strings is not represented in memory just like an array of simple type, and cannot (?) be simply getmem-ed ? program test; type NameStr=String[8]; shortarray=array[1..3] of NameStr; var n:NameStr; a:^shortarray; i:integer; begin getmem(a,3*sizeof(NameStr)); n:='12345678'; for i:=1 to 3 do begin a^[i]:=n; writeln('n = ',n) writeln('a^[i] = ',a^[i]); end; end. As a result the value of a^[i] is an empty string after the assignment; I will be very grateful for any answers or comments (I use gpc version 20010502, based on 2.95.3 20010315). Regards, Adam Naumowicz -------------------------------------- WWW: http://math.uwb.edu.pl/~adamn/ --------------------------------------

7 16

Object pascal mode for Emacs
by Preben Mikael Bohn 11 Jul '01

11 Jul '01

Hello Does anyone know if there is an Object-Pascal mode for Emacs? I am currently using the 'built-in' Pascal-support, however it dosn't handle objects well. I also tried the Delphi mode (I cannot remember where I got this from) but this dosn't seem to handle GPS-objects well... :-) Best regards Preben Bohn ===== The two rules for success are: 1. Never tell them everything you know. ____________________________________________________________ Do You Yahoo!? Get your free @yahoo.co.uk address at http://mail.yahoo.co.uk or your free @yahoo.ie address at http://mail.yahoo.ie

2 1

Re: [Fwd: RE: Error with asmname]
by Jing Gloria 11 Jul '01

11 Jul '01

Maurice, thanks for taking the time look into my problem. Unfortunately, I have to work with certain constraints on this project. I am trying to find the least amount of work to port the current software to the new hardware platform (SPARC/RISC SunOS to i86 Solaris 7). We will continue to support both hardware platforms and I would prefer one set of source files for both systems. Our SPARC/RISC SunOS software uses an old Pascal compiler which does not have a cross-compiler for i86. Worst case scenario would be to generate a new set of source files for the i86 platform and use the techniques that you suggested. Again, thanks for the suggestions and assistance. I will make a new build of GPC using the latest source code. I will do that tomorrow; I am now at home since the airconditioning unit in my house conked out last night. I have to get it fixed today. Can't do without the AC in this hot Texas summer! Regards, Jing Gloria Texas Instruments

1 0

[Fwd: RE: Error with asmname]
by Maurice Lombardi 11 Jul '01

11 Jul '01

I forward this message to the list to enable people to understand the answer which is at the end. -------- Original Message -------- Objet: RE: Error with asmname Date: Tue, 10 Jul 2001 15:51:51 -0500 De: "Gloria, Jing" <j-gloria(a)ti.com> A: "'Maurice Lombardi'" <Maurice.Lombardi(a)ujf-grenoble.fr> Maurice, pls be kind enough to run my test program in your system and see if it compiles and links correctly. I have attached all three source files and makefile needed. Pls advise which version of GPC you are using. Thanks in advance for your assistance. I will be making a build of the latest GPC from source tomorrow. Regards, Jing Gloria Texas Instruments --------------------------------------------------------- The problem is with try_hello.pas which is also a main pascal program. The gpc-main trick is only to circumvent the case in which a C module happens to define a function with name identical to the default gpc main name. But you cannot have two gpc mains because you have conflicts with gpc main initializers and finalizers. The solution is to write try_hello.pas as a unit, like this: xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx unit try_hello; INTERFACE Var Testcode: integer; asmname 'testcode'; Procedure p_hello; IMPLEMENTATION Procedure p_hello; begin Writeln ('PASCAL says: Hello, world.'); Testcode := TestCode *2; Writeln ('Testcode = ', TestCode); end; End. xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx Your makefile works then. But this is still not the right way of doing. You do not need makefiles with gpc Rewrite try_main as xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx program Hello (Output); uses try_hello; {$L c_hello.c} procedure c_hello; external; Begin Writeln ('Starting in main PASCAL module.'); Writeln ('Calling C routine.'); c_hello; Writeln ('Now back from C.'); P_hello; WRITELN ('That is all folks!!') End. xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx then you can compile only the main with --automake gpc try_main.pas -o try_main.exe --automake gpc automatically compiles try_hello.pas and c_hello.c becauses it has all informations needed in try_main by way of the uses clause for pascal units and {$L } for c modules. Maurice BTW I use djgpp/ gpc 20010623 based on gcc 2.95.3 -- 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 {$gpc-main=Dummy} Var Testcode: integer; asmname 'testcode'; Procedure p_hello; begin Writeln ('PASCAL says: Hello, world.'); Testcode := TestCode *2; Writeln ('Testcode = ', TestCode); end; Begin { dummy block or else the GNU Pascal compiler will complain and generate a compile error.} End. #include <stdio.h> extern int testcode; C_hello() { printf ("Hello, World - from C \n\r"); testcode = 12; } # Test Makefile for Pascal and C compiler. ..KEEP_STATE: # try2: try_main.o c_hello.o try_hello.o gpc -o try2 try_main.o try_hello.o c_hello.o c_hello.o: c_hello.c gcc -O -c c_hello.c try_hello.o: try_hello.pas gpc -c try_hello.pas try_main.o: try_main.pas gpc -c try_main.pas program Hello (Output); procedure c_hello; external; Procedure p_hello; external; Begin Writeln ('Starting in main PASCAL module.'); Writeln ('Calling C routine.'); c_hello; Writeln ('Now back from C.'); P_hello; WRITELN ('That is all folks!!') End.

1 0

[Fwd: Re: Error with asmname]
by Maurice Lombardi 11 Jul '01

11 Jul '01

This message was bounced back due to youe email problemq Gloria, Jing wrote: > Maurice, > > thanks for your inputs. > > I followed your suggestion using the old style __asmname__ > and I get the Pascal module to compile without errors. > But now I get the following error during link. > > rigor% make > gpc -c try_hello.pas > try_hello.pas:2: warning: missing program header > gpc -o try2 try_main.o try_hello.o c_hello.o > Undefined first referenced > symbol in file > testcode try_hello.o > ld: fatal: Symbol referencing errors. No output written to try2 > collect2: ld returned 1 exit status > *** Error code 1 > make: Fatal error: Command failed for target `try2' > > try_hello.pas is the module that is exporting Testcode. How ? Is > there any other flag or directive that I am missing? It > looks like try_hello.pas is expecting testcode to be external. > Difficult to say without having the codes at hand ... > I may have to bite the bullet here and build the latest > released GPC code. I am avoiding this since I am a newbie > on Solaris/Unix systems. Probably so. You seem to try not to use the automake features of gpc, and use C-style programming, with independent compilations of the modules, final linking and the like. But if your programs are properly written, you should have only to type gpc try_main.pas -o try2 --automake and gpc manages to compile all C modules and pascal units that are needed, because all needed information is in the source files. But I do not remember what was posible two years ago, and probably very few people can do. Much effort have been made on the compiler for this to work well. So probably install the last gpc alpha and look in the demos directory the programs gpc_c_pas.pas c_gpc.pas and friends which show how to do that properly. 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: Error with asmname]
by Maurice Lombardi 11 Jul '01

11 Jul '01

This message was bounced back due to your email problems Gloria, Jing wrote: > I would appreciate any assistance to the following problem. > > The following 2 lines of code result in error during compilation: > > Var > Testcode: integer; asmname 'testcode'; > > Error: > try_hello.pas:3: parse error before character constant > try_hello.pas:5: comma missing after identifier `Asmname' > try_hello.pas:5: parse error before `;' > > Note: Line 3 is the line with asmname keyword. > > I am using: > GNU Pascal version 2.95 19990728 (release) (i386-pc-solaris2.7) > compiled by GNU C version 2.95.1 19990816 (release). > > If I remove the asmname 'testcode' from line 3, then the > compilation goes through without any errors. the syntax of asmname for variables has changed on 2000/04/12, to be more consistent with the syntax for procedures and functions. Old syntax was Var Testcode: __asmname__ 'testcode' integer; The new is Var Testcode: integer; asmname 'testcode'; external; external is needed if it is imported from elsewhere (usually C), it is not if you want to export from pascal to C; So use the old syntax if you are in a hurry, but rather upgrade to a recent gpc version (the last alpha is the best). 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

email problems
by Peter Gerwinski 11 Jul '01

11 Jul '01

Hello, during the last 36 hours emails to this list were not delivered. The problem is fixed now. In case you have unsuccessfully tried to post to the list, please retry now. In den letzten 36 Stunden wurde E-Mails an diese Liste nicht zugestellt. Das Problem ist jetzt behoben. Wer in diesem Zeitraum vergeblich versucht hat, E-Mail an die Liste zu schicken, kann es jetzt noch einmal probieren. Peter -- http://home.pages.de/~Peter.Gerwinski/ - G-N-U GmbH: http://www.g-n-u.de Maintainer GNU Pascal - http://home.pages.de/~GNU-Pascal/ - gpc-20010623 GnuPG key fingerprint: 9E7C 0FC4 8A62 5536 1730 A932 9834 65DB 2143 9422 keys: http://www.gerwinski.de/pubkeys/ - AntiSpam: http://spam.abuse.net

1 0

Re: Solaris 8 install & post-install problems
by Frank Heckenbach 11 Jul '01

11 Jul '01

> I have been trying, on behalf of one of our researchers here, to > install the gpc-20010623 on a Solaris 8 (8-04) SunBlade. > > I had install problems buth using Sun's "extra" gcc-2.95.2 sources as > the base and using the gcc-core distribution from the GPC ftp site. > > Mainly, these difficulties appear to arise because things like xmalloc > and concat were being multiply defined. > > So, by a good deal of guesswork, I ended up adding: > > -DEGCS -DEGCS92 -DEGCS95 > > to be defined when doing the make. It compiled. This suggests that configure didn't run correctly since it's supposed to set the correct defines automatically. > However when I try to compile a "hello word" I get the following > at the linker stage > > Undefined first referenced > symbol in file > _p_eq /usr/local/packages/gcc-2.95.2/lib/gcc-lib/sparc-sun-solaris2.8/2.95.2/libgpc.a(string.o) > ld: fatal: Symbol referencing errors. No output written to a.out > collect2: ld returned 1 exit status > > > I guess this is a Pascal thing. > > [...] > > GNU Pascal Compiler PreProcessor version 20010605, based on gcc-2.95.2 19991024 (release) (sparc) ^^^^^^^^ That seems strange since you said you were using 20010623 (this problem on the Sparc was fixed (I hope) on 2001-06-19, so it does make a difference). Does ./gpcpp (in the build directory) claim to be 20010623? If so, apparently the Makefile used an older installed compiler. Otherwise, you either did not get 20010623 (and might want to do so now), or something went wrong in our CVS archives (which I can't check right now). You can probably work around the problem by removing `{$if True}' and the following `{$endif}' in p/rts/string.pas, but I'd appreciate if you try to find out whether there's a real problem left or it was only a version conflict. Frank -- Frank Heckenbach, frank(a)g-n-u.de http://fjf.gnu.de/ PGP and GPG keys: http://fjf.gnu.de/plan

1 0

Jump to page:

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