- Gpc - gnu.de

clreol, wherex, etc how?
by Toby Ewing 15 Sep '06

15 Sep '06

Hi, all I'm running gpc-20041218, based on gcc 3.3.3, on a pc running RH Enterprise. I have several programs that use basic cursor controls -- wherex, wherey, clreol, and so on -- which I used to grab as externals: Procedure clreol; C; Procedure gotoxy (x : integer; y : integer ); C; Function wherey : integer; C; I have been unable to get the externals to "take" on this version of gpc, though I've tried about 20 variants -- external, asm_name, just about anything I could find in the docs and demos. It seemed logical to just "use crt" instead. Unfortunately, the CRT unit has resisted all my attempts at compilation. Specifically: * crt.pas won't compile unless crtc.c has been compiled first * "gcc -c crtc.c" fails because ncurses.h, panel.h, and term.h are missing. (In fact, panel.h and term.h are in the units directory, but presumably gcc wants them somewhere else.) My system claims to have ncurses 5.4-13 installed, but apparently that is not sufficient. I tried Willett's ncurses.gpc 20060207, but configure failed loudly because it couldn't find ncurses.h (WHICH ncurses.h, and where does it want it?). I tried the PDCurses-2.6 on the GPC libraries page, and here again configure failed (couldn't find Intrinsic.h). I am stumped. What can I do to get simple cursor control in my programs? many thanks, Toby

3 2

internal compiler error when trying to use record variable declared using "absolute" keyword
by Igor Marnat 04 Sep '06

04 Sep '06

Hello, List! Trying to build the program below using gpc 20060325 based on gcc 4.0.3 I've got "Internal compiler error". Sorry for a bit complicated test program but it's the simplest one I could produce to reproduce the bug. Error message: [igor@marnat2 internal_compiler_error]$ make gpc --automake test.pas <var_decl 0x401910d8 Eepromstor.155419 type <pointer_type 0x4018ec3c type <pointer_type 0x4018ea8c Eepromptr type <record_type 0x4018e510 Eepromtype> unsigned SI size <integer_cst 0x40139888 constant invariant 32> unit size <integer_cst 0x401398a0 constant invariant 4> align 32 symtab 0 alias set -1 main_variant <pointer_type 0x4018ea8c Eepromptr> pointer_to_this <pointer_type 0x4018ec3c>> unsigned SI size <integer_cst 0x40139888 32> unit size <integer_cst 0x401398a0 4> align 32 symtab 0 alias set -1 main_variant <pointer_type 0x4018ec3c>> used unsigned ignored SI file test.pas line 26 size <integer_cst 0x40139888 32> unit size <integer_cst 0x401398a0 4> align 32 context <function_decl 0x4018ee58 Eep> (reg/f:SI 0 ax [orig:63 Eepromstor.155419 ] [63]) chain <var_decl 0x40191144 D.606>> test.pas: In function `EAdvLineInfo': test.pas:31: internal compiler error: in expand_expr_real_1, at expr.c:6800 Please submit a full bug report, with preprocessed source if appropriate. See <URL:http://www.gnu-pascal.de/todo.html> for instructions. make: *** [all] Error 1 Offending program: program test; type LinesInfoPtr = ^LinesInfoType; LinesInfoType = array [0..99] of word; type EEPROMType = record LinesInfo: word; end; type StoragePtr = ^StorageType; StorageType = array [0..99] of byte; type EEPROMPtr = ^EEPROMType; var EEPROMStor: StoragePtr; EEPROMData: EEPROMPtr absolute EEPromStor; function EEP ( W: word): pointer; begin EEP := pointer (word (EEPromData) + W) end; function EAdvLineInfo (Line: ShortWord): word; begin EAdvLineInfo := LinesInfoPtr (EEP (EEPROMData^.LinesInfo))^ [Line]; end; begin writeln (integer (EAdvLineInfo (1))); end. ~ -- Best regards, Igor mailto:marny@rambler.ru

2 2

GCC front-end for IBM XL Pascal/6000 VS Mode
by Shirley Goldrei 01 Sep '06

01 Sep '06

Is anyone aware of a front-end for gcc that understands IBM XL Pascal/ 6000 VS Mode Pascal dialect? Failing that, has anyone had experience in porting or automatically translating (i.e. "transforming") XL Pascal (VS Mode) to GPC? Many thanks, Shirley Shirley Goldrei PhD Student Computing Department Macquarie University

2 1

GPC, CGI units, and program size
by Prof. A Olowofoyeku (The African Chief) 30 Aug '06

30 Aug '06

Hi I have downloaded Frank's cgiprogs package, for the purpose of building a cgi program for an adsl router (running MontaVista Linux on a mipsfple processor). First, hats off to Frank. The program compiled and runs fine on the router. The problem is the size of the compiled program (320kb when statically linked with libgpc.a, and 250kb when linked to libgpc.so, which itself is 320kb). This is a significant amount of space for a single program to take on a router. My question (and probably only Frank can answer this) is, how can one reduce the size of this program? The program uses the gpc and cgi units only. From the GPC unit it calls "Execute" and "GetTempDirectory", and from the cgi unit it calls: GetVar IsCGI CGIInitVars If it is not possible to reduce the program size, then I guess that would have to be it - I just thought I'd ask anyway :) Perhaps there is a C cgi library? Thanks. Best regards, The Chief --------- Prof. Abimbola Olowofoyeku (The African Chief) Web: http://www.greatchief.plus.com/

3 12

results of cross compilation gpc20060325 for ppc-linux target on i386-linux platform
by Igor Marnat 29 Aug '06

29 Aug '06

Hello List. I've cross built gpc20060325 for ppc-linux target. In order to do that I've made the changes attached. The second and the third changes were needed because otherwise compilation process called xgpc (cross-compiler just built) with native CFLAGS (-O2 -g -march=i386 -mcpu=i686) which it didn't want to accept. The first change allowed me to install the result compiler as ppc-linux-gpc instead of ppc-linuxppc-linux-gpc as it wanted to do first. Perhaps some changes are to be made in sources. diff -urN gpc-20060325/p/Make-lang.in gpc-20060325-ppc/p/Make-lang.in --- gpc-20060325/p/Make-lang.in 2006-03-25 06:33:50.000000000 +0300 +++ gpc-20060325-ppc/p/Make-lang.in 2006-08-29 19:04:08.057019832 +0400 @@ -179,7 +179,8 @@ # Actual names to use when installing a cross-compiler. GPC_CROSS_NAME_=`t='$(program_transform_cross_name)'; echo gpc | sed $$t` -GPC_CROSS_NAME_34=$(target_noncanonical)`t='$(program_transform_name)'; echo gpc | sed $$t` +#GPC_CROSS_NAME_34=$(target_noncanonical)`t='$(program_transform_name)'; echo gpc | sed $$t` +GPC_CROSS_NAME_34=`t='$(program_transform_name)'; echo gpc | sed $$t` GPC_CROSS_NAME=$(GPC_CROSS_NAME_$(GPC_GCC_34_)) # The GPC to use for compiling libgpc.a @@ -228,8 +229,8 @@ RTS_FLAGS_TO_PASS_ALL=\ GCC_VERSION="$(GPC_GCC_VERSION)" \ SHELL="$(SHELL)" \ - CFLAGS="$(CFLAGS)" \ - PFLAGS="$(PFLAGS)" \ + CFLAGS="-O2 -g -mcpu=405" \ + PFLAGS="-O2 -g -mcpu=405" \ AR_FLAGS="$(AR_FLAGS)" \ RTSFLAGS="$(RTSFLAGS)" \ DESTDIR="../.." @@ -778,7 +779,7 @@ $(srcdir)/p/script/mkdir-p p/rts src=`cd $(srcdir) && pwd` && \ cd ./p/rts && \ - $(RTS_COMPILERS) CONFIG_SITE=nonexisting-file \ + $(RTS_COMPILERS) $(RTS_FLAGS_TO_PASS) CONFIG_SITE=nonexisting-file \ $$src/p/rts/configure --silent > /dev/null --srcdir=$$src/p/rts \ `if [ x$(WITH_SHARED) = xyes ]; then echo '--with-shared'; fi` Best regards, Igor Marnat mailto:marny@rambler.ru

1 0

test results for gpc 20060325
by Igor Marnat 28 Aug '06

28 Aug '06

Hello List! Here are the results of building test suite on my machine. Results: rm -f todo/a.out todo/*.exe todo/*.o todo/*.s todo/*.i todo/*.gpi todo/*.gpd todo/core GP= PC="gpc" PFLAGS=" --autobuild -g -O3 -W -Wall -Wno-unused " PFLAGS_NO_PATHS="-g -O3 -W -Wall -Wno-unused " SRCDIR="." TEST_MAKE_FLAG=test-make-flag "./test_run" "*.pas" | tee test_log | "./test_sum" -d Test Run By igor on 2006-08-28 19:22:41 Native configuration is i686-pc-linux-gnu (marnat2) === gpc tests === Running target any Running testsuite ... UNSUPPORTED: aregextest.pas FAIL: fjf322.pas FAIL: fjf403b.pas FAIL: fjf587b.pas FAIL: fjf779a.pas FAIL: fjf779b.pas FAIL: fjf779e.pas FAIL: fjf779f.pas FAIL: fjf779g.pas FAIL: goto8.pas FAIL: nicola4c.pas === gpc Summary === # of tests 5070 # of expected passes 5059 # of unexpected failures 10 # of unsupported tests 1 gpc version 20060325, based on gcc-4.0.3 Machine configuration: RedHat Linux, CPU: Intel 32-bit Kernel: 2.6.7, -- Best regards, Igor Marnat mailto:marny@rambler.ru

1 0

Bug with packed arrays
by Jonas Maebe 27 Aug '06

27 Aug '06

Hello, While working on packed array support for FPC and verifying against GPC, I discovered a bug in GPC's packed array support. The following program prints 0 as second line instead of 1540222835 *** type ta = packed 0..$7fffffff; tb = packed array[0..40] of ta; function test(i: longint): integer; var b: tb; begin b[1] := $5BCDEF73; test := b[i]; end; begin writeln(sizeof(tb)); writeln(test(1)); end. *** Jonas

5 10

GPC @ 64 bit
by Silvio a Beccara 25 Aug '06

25 Aug '06

Good morning, which flags should I use to compile 64-bit applications on a 64-bit machine with GPC? Thanks Silvio a Beccara -- Email.it, the professional e-mail, gratis per te: http://www.email.it/f Sponsor: Vinci 100 giorni di Sheba: In palio tanti set di prodotti Sheba per la gioia del tuo gatto! Clicca qui: http://adv.email.it/cgi-bin/foclick.cgi?mid=5267&d=8-8

2 1

Problem with --enable-keyword and --enable-predefined-identifier
by J. David Bryan 17 Aug '06

17 Aug '06

Running gpc-20060325 on i486-pc-mingw32 (gcc-3.4.6), I wish to use an "asm" statement and the predefined identifiers "Cardinal" and "ParamCount" in an Extended Pascal program. However, the following program fails to compile: {$ extended-pascal } {$ enable-keyword asm } {$ enable-predefined-identifier Cardinal } {$ enable-predefined-identifier ParamCount } program Predefs; var C : Cardinal; I : integer; begin asm ('nop'); I := ParamCount; C := 1 end. With "gpc -c predefs.pas", I get: predefs.pas:1: warning: compiler directives are a UCSD Pascal extension predefs.pas: In main program: predefs.pas:12: error: `asm' is an extension of Borland Pascal, traditional predefs.pas:12: error: Macintosh Pascal predefs.pas:13: error: `ParamCount' is an extension of Borland Pascal Note from the above that "Cardinal" is being enabled by the directives, whereas "ParamCount" and "asm" are not. I get the same results if I specify the compiler switches on the command line (e.g., "--enable-keyword=asm", etc.) instead of within the program. I see the same behavior with gpc-20051116. Is this a compiler bug, or am I doing something wrong? -- Dave

3 14

which is faster: inc(x) 'or' x:=x+1 ?
by wangtj＠Safe-mail.net 17 Aug '06

17 Aug '06

as title thanks.

3 2