May 2000 - Gpc

Re: Is it possible that GPC doesnt recognize "file of char" but "text" ?
by Frank Heckenbach 03 May '00

03 May '00

Tim Kaulmann wrote: > A friend told me that his GPC Compiler did not accept "file of char" at his > variable declaration part. He tried "text" instead and it worked. > > Is this solved in a newer version of GPC or is this an unknown bug ? AFAIK, GPC has always supported `file of Char' (at least, for several years, and I don't have an older version to test -- if he's still using GPC 2.0 or such an old version, he might be better off upgrading, also for other reasons). It is true, however, that GPC doesn't treat `file of Char' and `Text' as equivalent (e.g., Readln/Writeln applies to `Text' files only). That's according to the standard, AFAIK, and the same with most other Pascal compilers. But ISTR there are a few compilers (Sun Pascal?) that treat them as equivalent, so if he tried code written for these compilers, it might be that he used `file of Char' where actually `Text' was correct. If that's not the cause, please get some sample code from him, and state the compiler version used. 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

Is it possible that GPC doesnt recognize "file of char" but "text" ?
by Tim Kaulmann 03 May '00

03 May '00

Hello again. A friend told me that his GPC Compiler did not accept "file of char" at his variable declaration part. He tried "text" instead and it worked. Is this solved in a newer version of GPC or is this an unknown bug ? Greetings, Tim

1 0

BP toys in GPC
by Mariusz Zynel 03 May '00

03 May '00

Hi First of all let me thank Authors for supplying SetTextBuf in GPC (20000401 snapshot). There are still, however, some BP functions missing in GPC: SeekEOLN, SeekEOF, Include and Exclude. Providing them in GPC would make porting from BP easier. Maybe the following code could be utilized: const acTAB = #009; acLF = #010; acCR = #013; acWS = #032; type AnySetElement = 0..255; AnySet = set of AnySetElement; function SeekEoln(var aFile: Text): Boolean; var CurPos: Integer; CurChar: Char; begin if not Eoln(aFile) and not Eof(aFile) then begin repeat CurPos := FilePos(aFile); Read(aFile, CurChar); until ((CurChar <> acTAB) and (CurChar <> acWS)) or Eof(aFile); Seek(aFile, CurPos); end; SeekEoln := Eoln(aFile); end; function SeekEof(var aFile: Text): Boolean; begin while SeekEoln(aFile) and not Eof(aFile) do Readln(aFile); SeekEof := Eof(aFile); end; procedure Include(var aSet: AnySet; aElement: AnySetElement); begin aSet := aSet + [aElement]; end; procedure Exclude(var aSet: AnySet; aElement: AnySetElement); begin aSet := aSet - [aElement]; end; Regards -- . Mariusz Zynel Institute of Mathematics | phone: +48 (85) 745 75 52 University of Bialystok | Akademicka 2 St | mailto:mariusz@math.uwb.edu.pl 15-267 Bialystok | Poland | http://math.uwb.edu.pl/~mariusz/

2 1

Re: BP toys in GPC
by Frank Heckenbach 02 May '00

02 May '00

Mariusz Zynel wrote: > First of all let me thank Authors for supplying SetTextBuf in GPC > (20000401 snapshot). There are still, however, some BP functions > missing in GPC: SeekEOLN, SeekEOF, Include and Exclude. Providing > them in GPC would make porting from BP easier. Thanks for the hint. In fact, Include and Exclude have been added on Feb 19, so if you have any luck with 20000401, you have these routines already... > Maybe the following > code could be utilized: > > const > acTAB = #009; > acLF = #010; > acCR = #013; > acWS = #032; Just a few comments, no offense intended: I think it's better not to assume a particular character encoding (even though it's unlikely to change nowadays). ' ' is probably easier than #32, and for Tab, LF and CR, GPC supports "\t", "\n" and "\r" (in the releases up to now only if `char-escapes' is enabled, starting from the next release always in double-quoted strings). Also, it better to avoid `Seek'ing. It might be problematic on systems that have to do translation on text files (i.e., Dos -- but I don't feel like trying this out now), and is probably a little inefficient on any system. Since the BP routines actually skip the white space found, and Pascal provides look-ahead to the next character in text files with `FileVar^', seeking can easily be avoided here. I've put similar code (BP compatible in that it treats everything < ' ' as white space) into the RTS now. However, in BP the argument is optional (and defaults to Input), so we need compiler magic in GPC to do this as well. Peter, will you do this (I hope it's easy to do, just like with a number of other I/O routines)? 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 on Solaris (fwd)
by Veselic 02 May '00

02 May '00

Dear Dr. Gerwinski, dear GPC-ers, After having installed the new version of gpc all my sources (originally written for Turbo Pascal, hundreds of them) are compiled smoothly, BUT ONE. Here is the story: System description: SunOS mph 5.7 Generic_106541-08 sun4u sparc SUNW,Ultra-2 GPC description: GNU Pascal version 19991030, based on 2.95.2 19991024 (release), compiled Nov 1$ Copyright (C) 1987, 89, 92-97, 1998 Free Software Foundation, Inc. Here I try compiling: gpc --borland-pascal -o sj sj.p sj.p: In function `Gjgt_dec_new': sj.p:466: Internal compiler error in `emit_library_call_value', at calls.c:3198 Please submit a full bug report. See <URL:http://www.gnu.org/software/gcc/faq.html#bugreport> for instructions. Additional information: 1. The corresponding area in the source is: 465 x := sqrt( beta * w ) ; 466 y := 0.5 * alfa / x ; 467 z := 0.5 * gama / x ; 2. another source (named ej.p) which differs from sj.p ONLY in having 'extended' whenever sj.p has 'single', COMPILES AND EXECUTES SMOOTHLY. 3. On my PC-Linux with the gpc version GNU Pascal version 19990118, based on gcc-2.8.1, compiled Jul 6 1999 16:08:58. Copyright (C) 1987, 89, 92-97, 1998 Free Software Foundation, Inc. all sources, including the incriminated sj.p, compile OK. Very grateful in advance for your help, K. Veselic.

1 0

Building 20000401 snapshot on Solaris 8 x86
by Mariusz Zynel 01 May '00

01 May '00

Hi All I've found 20000401 snapshot and tried to build it on my Solaris 8 x86 box. I faced following problems: 1. there are bash extensions like ! [ ... ] used in Makefiles: - Make-lang-2.95.in, - Make-lang-old.in, - Make-lang.in, and - Make-lang - but this one belongs to gcc distribution. In result some precompiled objects do not install. 2. xgpc failed to compile rts modules like heap.pas etc.: ../.././xgpc -B../.././ -c -I. -W -Wall -Wpointer-arith -Wwrite-strings -Wmissing-prototypes -Wmissing-declarations -g -O2 -DRTS_RELEASE_STRING="'__GPC_RTS_VERSION_`cat /export/home/admin/soft/gcc-2.95.2/gcc/p/rts/rts_release`__'" -DBSD_RTS=0 `if [ x"@with_shared@" = xyes ] ; then echo -fPIC; fi` --unit-path=/export/home/admin/soft/gcc-2.95.2/gcc/p/rts --automake heap.pas xgpc: Internal compiler error: program gpc1 got fatal signal 11 gpc1: ../.././xgpc exited with status 1 The workaround was to remove RTS_WARN flag in p/rts/Makefile 3. GNU make and M4 are necessary to build GPC, but this is a minor problem as we all use GNU tools. Now to link a Pascal program I have to either, use --automake/build, or specify objects it depends on. What is the reason for this modification? Is there any switch to force GPC to find necessary units without --automake/build, as it was before? Thanks in advance for any suggestions? -- . Mariusz Zynel Institute of Mathematics | phone: +48 (85) 745 75 52 University of Bialystok | Akademicka 2 St | mailto:mariusz@math.uwb.edu.pl 15-267 Bialystok | Poland | http://math.uwb.edu.pl/~mariusz/

2 2

bug
by Toby Ewing 01 May '00

01 May '00

Hi, all. I've been fighting a bug for a while, and when I stripped it down to its basics, I was surprised. My test program is appended. The issue involves arrays of size > 255. When I use Ymax = Zmax = 250, the program runs fine. When I use Ymax = Zmax = 260, the portion of the program that runs in Main works fine, but the procedure stops at y = 7, z = 260 - the program hangs, with ^C ineffective. I'm running a pentium 400 under Win98. I've had no trouble with this in small arrays, but 256 seems to be a divide of some sort. Any suggestions appreciated. Toby Toby Ewing Soil Scientist, Iowa State University ------------------------------------------------------------------- Program TestLgArray; Const Xmax = 25 ; {physical space - grid units} Ymax = 260; {physical space - grid units} Zmax = 260; Type Compary = array[1..Ymax, 1..Zmax] of integer; var Comp : Compary; x, y, z : integer; {________________________________________________ } Procedure Zero; var Comp : Compary; x, y, z : integer; begin writeln('Procedure:'); readln; for y := 1 to Ymax do begin {zero the comp array} for z := 1 to Zmax do Comp[y,z] := 0; writeln('y=', y:3, ' z=', z:1); end; end; {________________________________________________ } begin {Main Program} writeln('Main:'); readln; for y := 1 to Ymax do begin {zero the comp array} for z := 1 to Zmax do Comp[y,z] := 0; writeln('y=', y:3, ' z=', z:1); end; Zero; write(' done'); end.

4 3

gpc 19991930 and gcc2.95.2 on LinuxPPC
by Thanh To 01 May '00

01 May '00

Hi, I follow the instruction below trying to build gpc 19991030 and gcc2.95.2 on my LinuxPPC tar xvfz gcc-2.95.2.tar.gz tar xvfz gpc-19991030.tar.gz mv gpc/p gcc-2.95.2/gcc/ mkdir gcc-2.95.2/gpc cd .. ./configure However when I do the make, I got the following error message: make[2]: Leaving directory `/export/home/accounts/tto/gcc-2.95.2/gcc/java' gcc -DIN_GCC -DHAIFA -g -O2 -DHAVE_CONFIG_H -o cc1obj toplev.o version.o tree.o print-tree.o s tor-layout.o fold-const.o function.o stmt.o except.o expr.o calls.o expmed.o explow.o optabs.o intl. o varasm.o rtl.o print-rtl.o rtlanal.o emit-rtl.o genrtl.o real.o dbxout.o sdbout.o dwarfout.o dwarf 2out.o xcoffout.o bitmap.o alias.o gcse.o integrate.o jump.o cse.o loop.o unroll.o flow.o stupid.o c ombine.o varray.o regclass.o regmove.o local-alloc.o global.o reload.o reload1.o caller-save.o insn- peep.o reorg.o haifa-sched.o final.o recog.o reg-stack.o insn-opinit.o insn-recog.o insn-extract.o i nsn-output.o insn-emit.o lcm.o profile.o insn-attrtab.o rs6000.o getpwd.o convert.o mbchar.o dyn-st ring.o splay-tree.o graph.o sbitmap.o resource.o hash.o objc-parse.o objc-act.o c-lex.o c-pragma.o c -decl.o c-typeck.o c-convert.o c-aux-info.o c-common.o c-iterate.o obstack.o ../libiberty/li biberty.a cd p; make "AR_FLAGS_FOR_TARGET=rc" "AR_FOR_TARGET=ar" "BISON=bison" "BISONFLAGS=" "CFLAGS=-g -O2" " CLIB=" "GCC_FOR_TARGET=/zubat/export/home/accounts/tto/gcc-2.95.2/gcc/xgcc -B/zubat/export/home/acco unts/tto/gcc-2.95.2/gcc/ -B/usr/local/powerpc-unknown-linux-gnu/bin/ -I/usr/local/powerpc-unknown-li nux-gnu/include" "LDFLAGS=" "LEX=flex" "LEXFLAGS=" "LN=ln" "LN_S=ln -s" "MAKEINFO=/zubat/export/home /accounts/tto/gcc-2.95.2/texinfo/makeinfo/makeinfo " "MAKEINFOFLAGS=" "RANLIB_FOR_TARGET=ranlib" "RA NLIB_TEST_FOR_TARGET=[ -f ranlib ] || ( [ "powerpc-unknown-linux-gnu" = "powerpc-unknown-linux-gnu" ] && [ -f /usr/bin/ranlib -o -f /bin/ranlib ] )" "SHELL=/bin/sh" "STAGE_PREFIX=" "exeext=" "build_ex eext=" "objext=.o" "exec_prefix=/usr/local" "prefix=/usr/local" "local_prefix=/usr/local" "gxx_inclu de_dir=/usr/local/lib/gcc-lib/powerpc-unknown-linux-gnu/2.95.2/../../../..`echo /usr/local | sed -e 's|^/usr/local||' -e 's|/[^/]*|/..|g'`/include/g++-3" "tooldir=/usr/local/powerpc-unknown-linux-gnu" "gcc_tooldir=/usr/local/lib/gcc-lib/powerpc-unknown-linux-gnu/2.95.2/../../../../powerpc-unknown-li nux-gnu" "bindir=/usr/local/bin" "libsubdir=/usr/local/lib/gcc-lib/powerpc-unknown-linux-gnu/2.95.2" "datadir=/usr/local/share" "distdir=../tmp/\$(subdir)" "localedir=/usr/local/share/locale" "CC=gcc" "CROSS=" "X_CFLAGS=" "EXTRA_LIBS=" "MAYBE_TARGET_DEFAULT=" "MAYBE_USE_COLLECT2=" "SCHED_CFLAGS=-DHA IFA" "includedir=`echo '/usr/local/include'|sed -e 's/\\$/\\\\&&/g'`" "libsubdir=`echo '/usr/local/l ib/gcc-lib/powerpc-unknown-linux-gnu/2.95.2'|sed -e 's/\\$/\\\\&&/g'`" "target_alias=powerpc-unknown -linux-gnu" "version=2.95.2" "target_alias=powerpc-unknown-linux-gnu" "program_transform_name=s,x,x, " "program_transform_cross_name=s,^,powerpc-unknown-linux-gnu-," "exeext=" "GCC_FOR_TARGET=/zubat/ex port/home/accounts/tto/gcc-2.95.2/gcc/xgcc -B/zubat/export/home/accounts/tto/gcc-2.95.2/gcc/ -B/usr/ local/powerpc-unknown-linux-gnu/bin/ -I/usr/local/powerpc-unknown-linux-gnu/include" "GPC_FOR_TARGET =./xgpc -B./" "GPCSOLIBSHORTNAME=libgpc.so" "GPCSOLIBDIR=lib" "WITH_SHARED=@with_shared@" ../gpc1 make[2]: Entering directory `/export/home/accounts/tto/gcc-2.95.2/gcc/p' gcc -c -DIN_GCC -DGPC -g -O2 -DHAIFA -I. -I.. -I. -I./.. -I./../config -I./../../include ./.. /stor-layout.c ./../stor-layout.c:64: initializer element is not constant make[2]: *** [stor-layout.o] Error 1 make[2]: Leaving directory `/export/home/accounts/tto/gcc-2.95.2/gcc/p' make[1]: *** [gpc1] Error 2 make[1]: Leaving directory `/export/home/accounts/tto/gcc-2.95.2/gcc' make: *** [all-gcc] Error 2 How can I fix this problem? Thank you. Thanh To

1 0

String types and relational operators
by Tomasz Kowaltowski 01 May '00

01 May '00

We are using for teaching purposes the GPC compiler under Linux (RedHat 6.1) which was installed from the following RPM packages: gpc-19990118-1.i386.rpm gpc-extras-19990118-1.i386.rpm Our Pascal programs are compiled with the "--borland-pascal" option. I noticed a problem with the type "String" (and the derived types like "String[10]") and the relational operators which do not seem to return correct results. The following test program seems illustrate this point: ---------------------------------------------------------------------- program test; procedure Compare(var x,y : String); begin if x=y then Writeln('"',x,'" = "',y,'"') else if x>y then Writeln('"',x,'" > "',y,'"') else if x<y then Writeln('"',x,'" < "',y,'"') else Writeln('"',x,'" ? "',y,'"') end; { Compare } var a,b : String; Empty,Zeros : String; begin a := 'abcd'; b := 'abce'; Empty := ''; Zeros := '0000'; Compare(Zeros,Zeros); Compare(Zeros,a); Compare(a,Zeros); Compare(Empty,Empty); Compare(Empty,a); Compare(a,Empty); Compare(a,a); Compare(a,b) end. ---------------------------------------------------------------------- The following output was produced: "0000" = "0000" "0000" > "abcd" [*] "abcd" ? "0000" [*] "" = "" "" > "abcd" [*] "abcd" ? "" [*] "abcd" = "abcd" "abcd" > "abce" [*] which shows that some results are wrong (marked with [*]). Any suggestions as to what might be wrong? -- Tomasz Kowaltowski

4 3

Help to work with GPC on Windows NT
by Anuradha 01 May '00

01 May '00

Hello, I want to compile Pascal programs using GNU-pascal. I FTPed the zip files from the agnes.dida.physik.uni-essen.de/gnu-pascal

3 2