Gpc September 1999

gpc@gnu.de

44 participants
93 discussions

Re: make error
by Frank Heckenbach 05 Sep '99

05 Sep '99

"Dmitry S. Luhtionov" wrote: > This message has been displayed when i try build gpc-19990813.tar.gz > > OS: FreeBSD 3.2 > > # make > `echo gcc | sed 's/gcc/gpc/'` -c -I. -Wall -Wpointer-arith -Wmissing-prototypes > -Wmissing-declarations -O2 -DBSD_RTS=0 `if [ x = xyes ] ; then echo -fPIC; fi` > --unit-path= --automake files.pas > files.pas: In function `Sleepmicroseconds': > files.pas:257: argument does not match conformant array formal parameter > files.pas:257: incompatible type for argument 1 of `Ioselect' > files.pas:257: argument does not match conformant array formal parameter > files.pas:257: too few arguments to function `Ioselect' > files.pas: In function `Stringtfdd_rewrite': > files.pas:310: assignment from incompatible pointer type > *** Error code 1 The problem is (I suppose) that an older installed gpc (`echo gcc | sed 's/gcc/gpc/'`) is used to build the RTS rather than the GPC just created. This cannot work since the RTS relies on the latest compiler features. Did you specify any configure or make options (like GCC_FOR_TARGET) or environment variables? If so, try without them. 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: re; new snapshot
by Frank Heckenbach 05 Sep '99

05 Sep '99

"Russ Whitaker" wrote: > ran into two little bugs: > > 1: Since compiling previous snapshot changed motherboards here > so i486 became i586. Ran into the "ar" bug again. While a soft link > fixes the problem, changing line 74 of Makefile.in does it nicely: > > From: > AR = ar > To: > AR = /usr/bin/ar I think /usr/bin/ar might not be correcton every system. You said you ran into this bug again. When was the first time, I don't remember, and what was the cure then? Why does `AR = ar' not work on yor system, if ar is in /usr/bin (and /usr/bin is certainly in your $PATH, isn't it)? Besides, since this is the GCC Makefile.in, we can't really make any changes there, anyway... > 2. The date in gpc_release doesn't match the snapshot date Yep. Shouldn't be a serious problem, however. ;-) 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

header files
by Simon Nath 05 Sep '99

05 Sep '99

Hi, I'm trying to modularize my pascal programs into multiple source files. I was able to do this using Units but now I'm having trouble defining data types and global variables in a single file that can be included by pascal source files. Is there an interface file or definition file concept in GNU Pascal? Thanks -simon ____________________________________________________________________ Get your own FREE, personal Netscape WebMail account today at http://webmail.netscape.com.

4 3

Re: Internal compiler error with arithmetic or arrays?
by Frank Heckenbach 05 Sep '99

05 Sep '99

Matthias Friedrich wrote: > Hello! > > Who can help me with my gpc-compiler error? I think its an arithmetic problem, > maybe the declaration of Single-Arrays... > > > > The following code produces an 'Internal compiler error: program gpc1 got fatal > signal 6' on my Solaris2.6 system: Unfortunately, real arithmetic is buggy under Solaris. I added your program to our test suite (matt1.pas). 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: gpc-19990813+gcc-2.95.1/alphaev56-dec-osf4.0d => failure
by Frank Heckenbach 05 Sep '99

05 Sep '99

Alexandre Oliva wrote: > --- gcc/p/rts/rts.c~ Fri Aug 6 12:49:33 1999 > +++ gcc/p/rts/rts.c Sun Aug 29 15:29:31 1999 > @@ -330,6 +330,10 @@ > > static clock_t initial_clocks = 0; > > +#ifndef CLOCKS_PER_SEC > +#define CLOCKS_PER_SEC 1000 > +#endif > + > int > _p_get_cpu_time (MicroSecond) > int *MicroSecond; Hmpf. Isn't there another way to get the value on this system (perhaps in another header file, or under another name)? The code above would set it to 1000 whenever it's not defined. This might be wrong on a system that uses a value other than 1000 and does not define CLOCKS_PER_SEC in the headers we include. 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: DIV bug found in GPC 19990510
by Frank Heckenbach 05 Sep '99

05 Sep '99

"GPC on Ken Linder's Linux Box" wrote: > Howdy all... > Found an oddity with the DIV operation. The following line correctly > produces a compile error (as it should): > > aninteger := 12.1 div 5.5; > > However, the following compiles & runs, behaving like a REAL division > operation: > > writeln( 12.1 div 5.5 ); > > Very strange indeed! Thanks for the report (kevin14.pas). 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

Compiling GPC
by Dr A Olowofoyeku 05 Sep '99

05 Sep '99

Peter Gerwinski wrote: >The normal procedure is to type first `./configure' and then >`make LANGUAGES=pascal' in the GCC directory that contains the >GPC source in a subdirectory `p'. However I know that mingw32 >needs special care. You are right. For one, there is no such program as "configure". This must be a Linux thing, or am I missing something here? > Better ask the list for help. Ok. Anyone managed to compile GPC for Mingw32 successfully? If so, can you please give me any tips? Thanks. Best regards, The Chief -------- Dr. Abimbola A. Olowofoyeku (The African Chief) Email: laa12(a)keele.ac.uk Homepage: http://ourworld.compuserve.com/homepages/African_Chief/ Author of: Chief's Installer Pro v5.20a for Win32 ftp://ftp.demon.co.uk/pub/ibmpc/win95/apps/chief32/chief52a.zip

1 0

gpc-
by Jochen 05 Sep '99

05 Sep '99

Hi All, first of all: Is there an uptodate archive of the mailing list ? I try to compile a gcc-2.95.1 patched/extended by the gpc-19990813 sources/patch. (Same results with plain gcc-2.95.) This is on alpha-de-linux-gnu and i586-ps-linux-gnu. After applying the gcc-2.95 patch and moving the p directory, I do a configure and the make bootstrap. I always get the following error about object files that cannot be build (right after libiberty.a is build): stage1/xgcc -Bstage1/ -B/usr/alphaev5-dec-linux-gnu/bin/ -DGPC -DIN_GCC -DHAIFA -O2 -O3 -pipe -Wall -DHAVE_CONFIG_H -I. -I../../gcc-2.95.1/gcc -I../../gcc-2.95.1/gcc/config -I../../gcc-2.95.1/gcc/../include \ -DGPC_MAJOR="\"2\"" \ -DGPC_MINOR="\"0\"" \ -DGPC_RELEASE="\"`cat ../../gcc-2.95.1/gcc/p/gpc_release`\"" \ -DSTANDARD_STARTFILE_PREFIX=\"/usr/lib/\" -DSTANDARD_EXEC_PREFIX=\"/usr/lib/gcc-lib/\" -DDEFAULT_TARGET_VERSION=\"2.95.1\" -DDEFAULT_TARGET_MACHINE=\"alphaev5-dec-linux-gnu\" -DTOOLDIR_BASE_PREFIX=\"/usr/\" \ -c `echo ../../gcc-2.95.1/gcc/p/gpc.c | sed 's,^\./,,'` -o p/gpc.o ../../gcc-2.95.1/gcc/p/gpc.c: In function `read_specs': ../../gcc-2.95.1/gcc/p/gpc.c:1770: warning: implicit declaration of function `read' ../../gcc-2.95.1/gcc/p/gpc.c:1774: warning: implicit declaration of function `close' ../../gcc-2.95.1/gcc/p/gpc.c: In function `delete_if_ordinary': ../../gcc-2.95.1/gcc/p/gpc.c:2092: warning: implicit declaration of function `unlink' ../../gcc-2.95.1/gcc/p/gpc.c: In function `putenv_from_prefixes': ../../gcc-2.95.1/gcc/p/gpc.c:2269: warning: implicit declaration of function `putenv' ../../gcc-2.95.1/gcc/p/gpc.c: In function `find_a_file': ../../gcc-2.95.1/gcc/p/gpc.c:2299: warning: implicit declaration of function `access' ../../gcc-2.95.1/gcc/p/gpc.c: In function `init_automake_temp_file_name': ../../gcc-2.95.1/gcc/p/gpc.c:4821: warning: unused variable `is_pascal' ../../gcc-2.95.1/gcc/p/gpc.c: In function `fatal_error': ../../gcc-2.95.1/gcc/p/gpc.c:5069: warning: implicit declaration of function `getpid' ../../gcc-2.95.1/gcc/p/gpc.c: In function `main': ../../gcc-2.95.1/gcc/p/gpc.c:5087: warning: unused variable `buf' make[2]: *** No rule to make target `choose-temp.o', needed by `xgpc'. Stop. make[2]: Leaving directory `/tmp/gcc/gcc' make[1]: *** [bootstrap] Error 2 make[1]: Leaving directory `/tmp/gcc/gcc' make: *** [bootstrap] Error 2 Is this a known problem ? Any ideas or help ? PS: I am currently using the available binaries for gpc build against gcc-2.8.1 for evaluation. This is not a solution, though, since I want to use it in a mixed C++/FORTRAN/PASCAL project where I need the newer gcc-2.95. -- Jochen Heinrich-Heine-Universität Düsseldorf jochen(a)uni-duesseldorf.de Institut für Physikalische Chemie I phone ++49-211-8113681 Universitätsstr. 26.43.02.29 fax ++49-211-8115195 40225 Düsseldorf, Germany www-public.rz.uni-duesseldorf.de/~jochen

4 4

Division by 'Single' not allowed ?
by Matthias Friedrich 05 Sep '99

05 Sep '99

Hello! In my previous mail to this mailing list yesterday I had an 'Internal Compiler Error'. Now I changed the last few lines to: For i:=1 to II DO if h_c[i,j,k]=1 then begin if dx[i]<0 then test:=-(exp(ln(-dx[i])*(-1))) else test:=(exp(ln(dx[i])*(-1))); du:=abs((0.5*(UU[i,j,k]+UU[i+1,j,k]))*test); This works! And it means to me that the division by a 'Single'-Variable is not always allowed in GNUPascal. Is this true (or is there a bug) ?? Hope somebody can helpme Thanks for your answers Matthias _/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/ - Matthias Friedrich - email: mafri(a)ipg.uni-freiburg.de

1 0

Problems building new snapshot under EGCS
by david＠tcs01.demon.co.uk 05 Sep '99

05 Sep '99

In message <38919B7D.19990813155556.FOO-5D74.frank(a)g-n-u.de> Frank Heckenbach writes: > Hello, folks, > > a new GPC snapshot is available: > > ftp://agnes.dida.physik.uni-essen.de/gnu-pascal/alpha/gpc-19990813.tar.gz > Help! I'm having lots of problems building the new snapshot under RedHat Linux 6.0 (kernel 2.2.5-15 on i586) using EGCS (source RPM egcs-1.1.2-12, appears to be version 2.91.66). I unpacked the egcs source into a directory tree, moved the snapshot's p directory into gcc, applied the gcc-egcs-2.91.66 patch (after a bit of editing of directory names within the patch file), did a ./configure, and am now trying make LANGUAGES=pascal. Is this the correct sequence of actions? The INSTALL file in the snapshot doesn't seem to cover building GPC under EGCS. I guess my first question is: should I be using EGCS or would I be better off with a plain gcc source, and if so what version? Various things I've come across so far: The p/gpc_release file contains 19990610, I think it should contain 19990813. I think there is something strange about the way the contents of the gpc_release file is accessed - at the point where I see the make doing a cd p, I then see an error message from cat that it cannot find the gpc_release file. It looks as if there is one too few "../"s on the front of the path to the file. In p/gpc-common.c at line 2893, TREU_OVERFLOW should be TREE_OVERFLOW, I think. (Not a GPC problem): In my gcc/flow.c at line 1493, the code says REG_N_SEDS, it should be REG_N_SETS, I think. Finally, I come to a halt with make[1]: Entering directory `/home/david/egcs/objs/gcc' gcc -DIN_GCC -g -O2 -DHAVE_CONFIG_H -o cc1 c-parse.o c-lang.o c-lex.o c-pragma.o c-decl.o c-typeck.o c-convert.o c-aux-info.o c-common collect2: ld terminated with signal 11 [Segmentation fault], core dumped make[1]: *** [cc1] Error 1 make: *** [all-gcc] Error 2 make[1]: Leaving directory `/home/david/egcs/objs/gcc' I think the GCC line has been truncated in copying it between systems. Has anyone succeeded in compiling this snapshot in a similar environment? -- David James mailto:david@tcs01.demon.co.uk Special Stage Rally results archive <URL:http://www.tcs01.demon.co.uk/>

3 2

← Newer
1
2
3
4
5
6
7
8
9
10
Older →

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 September 1999