- Gpc - gnu.de

Problems building gpc-20030830 on mingw
by Markus Gerwinski 03 Apr '04

03 Apr '04

Hi folks, I'm currently trying to build gpc-20030830 on Windows XP with mingw. After unpacking, copying and `configure --enable-languages=pascal`ing, `make` aborts with the error message: make[1]: Circular libgcc.a <- pascal dependency dropped. In file included from config/i386/mingw32.h:31, from tconfig.h:16, from libgcc2.c:36: config/i386/cygwin.h:31:19: stdio.h: No such file or directory In file included from config/i386/mingw32.h:31, from tconfig.h:16, from libgcc2.c:36: config/i386/cygwin.h:447: parse error before '*' token config/i386/cygwin.h:447: warning: function declaration isn't a prototype config/i386/cygwin.h:449: parse error before '*' token config/i386/cygwin.h:449: warning: function declaration isn't a prototype In file included from libgcc2.c:37: tsystem.h:63:19: stdio.h: No such file or directory tsystem.h:66:23: sys/types.h: No such file or directory tsystem.h:69:19: errno.h: No such file or directory tsystem.h:76:20: string.h: No such file or directory tsystem.h:80:20: stdlib.h: No such file or directory tsystem.h:81:20: unistd.h: No such file or directory In file included from C:/msys/1.0/mysoftware/gcc-3.2.1/gcc/include/limits.h:132, from C:/msys/1.0/mysoftware/gcc-3.2.1/gcc/include/syslimits.h:7, from C:/msys/1.0/mysoftware/gcc-3.2.1/gcc/include/limits.h:11, from tsystem.h:84, from libgcc2.c:37: include/limits.h:132:75: limits.h: No such file or directory In file included from libgcc2.c:37: tsystem.h:87:18: time.h: No such file or directory make[2]: *** [libgcc/./_muldi3.o] Error 1 make[1]: *** [libgcc.a] Error 2 make: *** [all-gcc] Error 2 After in the Makefile.in I added -I$(srcdir)/fixinc/tests/base to the include path, only this error message remained: rm: `include' is a directory rm: `include' is a directory config.h is unchanged hconfig.h is unchanged tm_p.h is unchanged In file included from config/i386/mingw32.h:31, from config.h:18, from cppdefault.c:25: config/i386/cygwin.h:447: parse error before '*' token config/i386/cygwin.h:447: warning: function declaration isn't a prototype config/i386/cygwin.h:449: parse error before '*' token config/i386/cygwin.h:449: warning: function declaration isn't a prototype make[1]: *** [cppdefault.o] Error 1 make: *** [all-gcc] Error 2 These lines in the source refer to a macro(?) named FILE that seemingly isn't defined anywhere. Does anyone have an idea what I'm doing wrong? -- My MinGW installation directory is just C:\MinGW, so all of the important system headers should be found in C:\MinGW\include. I'm compiling from within my msys directory tree, which is situated at C:\msys\1.0. Any ideas will be welcome. Thanks in advance, Markus

3 20

Graph
by Prof. Harley Flanders 03 Apr '04

03 Apr '04

Does anyone know how to get the Graph unit to compile? Thank you, HF =========================================== Prof. Harley Flanders 3533 Windemere Court Ann Arbor, MI 48105-2867 Home: 734 668 1546 harley(a)umich.edu Visiting Professor University of Michigan Dept. of Mathematics, East Hall Ann Arbor, MI 48109 734 764 6451 Secretary: 734 764 6485 FAX: 734 763 0937 ===========================================

4 15

File names and ChDir
by Frank Heckenbach 03 Apr '04

03 Apr '04

Hi, there's a strange case when using relative file names and changing the directory. (Incidentally I was made aware of this by a bugfix in the ncurses library (C code) some while ago ... ;-) Consider this program: program Foo; var i: Integer; f: Text; begin Assign (f, 'test.dat'); Rewrite (f); WriteLn (f, 42); Close (f); ChDir ('..'); Reset (f); ReadLn (f, i); WriteLn (i) end. It fails under both GPC and BP because `Reset' tries to open a file `test.dat' in a different directory than it was created. Of course, this effect isn't limited to `Assign'. The issue applies to any way of storing file names, provided that changing directories is possible at all. A solution would be to store the expanded file name (i.e., turn it into an absolute path whenever possible -- I currently don't know any case where it wouldn't be possible). Of course, this wouldn't be strictly BP compatible, but I wonder if this is a "good" or "bad" incompatibility. (Of course, the difference would be subtler and more confusing if there actually was a file `test.dat' in the new directory.) Also, clearly a new `Assign' statement after the change of directory would refer to the new directory, as would be expected. How do other compilers behave? What would be the expected behaviour of code like above if it actually occurs in practice? Frank -- Frank Heckenbach, frank(a)g-n-u.de, http://fjf.gnu.de/, 7977168E GPC To-Do list, latest features, fixed bugs: http://www.gnu-pascal.de/todo.html GPC download signing key: 51FF C1F0 1A77 C6C2 4482 4DDC 117A 9773 7F88 1707

5 7

Re: internal error in gpc
by Adriaan van Os 02 Apr '04

02 Apr '04

In the Macintosh Pascal mailing list (<http://pascal-central.com/maillist.html>) Scott Lindhurst wrote: > I just started playing around with gpc, working on getting some old > code to work with it, and have found a bug. This is version 3.3.2d1, > just downloaded a few days ago with the accompanying GPCInterfacesB2, > on OS X 10.2.8. > > I get the message "gpc: Internal error: Illegal instruction (program > gpc1)" when compiling the following program: > > program bug; > > uses > GPCMacOSAll, GPCStringsAll; {No error if remove GPCStringsAll} > > var > tix: UInt32; {No error if change to integer or longint} > begin > tix := 1; > > if tix > TickCount then {error while processing this line} > tix := 2; > end. > > My makefile generates this command line (which is going to be wrapped, > sorry): > gpc --automake --unit-path=./:/Developer/Pascal/GPCPInterfaces/ > --object-path=./build/ --unit-destination-path=./build/ > --executable-path=./build/ > -Wl,-framework,Carbon,-framework,ApplicationServices -o ./bug -Wl,-x > -O2 -Wall -Wno-identifier-case --ignore-garbage-after-dot > --longjmp-all-nonlocal-labels ./bug.pas > > but the internal error can be reproduced with just > gpc --unit-path=/Developer/Pascal/GPCPInterfaces/ ./bug.pas > > I have the workaround of using integer or longint instead of UInt32. Peter N. Lewis helped out by noting that "limit stacksize 8096" solves the problem. This is true, but still it is strange that the compiler needs such a huge amount of stack for compilation. So, I wonder if maybe there is indeed a problem in the compiler, solved only "by accident" by a stacklimit of nearly a megabyte. When I set "limit stacksize 4096", the backtrace reads: Date/Time: 2003-11-07 11:36:46 +0100 OS Version: 10.2.8 (Build 6R73) Host: G4.local. Command: gpc1 PID: 437 Exception: EXC_BAD_ACCESS (0x0001) Codes: KERN_INVALID_ADDRESS (0x0001) at 0xbff7fff0 Thread 0 Crashed: #0 0x000a9dc8 in tree_size (tree.c:161) #1 0x000aa024 in make_node (tree.c:237) #2 0x000c6980 in alloc_node (stringpool.c:70) #3 0x0024b4b0 in ht_lookup (hashtable.c:162) #4 0x000c6b1c in get_identifier (stringpool.c:109) #5 0x0001bb04 in get_operator (expressions.c:425) #6 0x0001c7e4 in build_pascal_binary_op (expressions.c:661) #7 0x0001c514 in parser_build_binary_op (expressions.c:583) #8 0x0000dd48 in yyuserAction (parse.c:4580) #9 0x000136c8 in yydoAction (parse.c:8238) #10 0x00013190 in yyglrReduce (parse.c:8290) #11 0x000124d0 in main_yyparse (parse.c:8991) #12 0x000a49c0 in compile_file (toplev.c:2134) #13 0x000a9c38 in do_compile (toplev.c:5402) #14 0x000a9d04 in toplev_main (toplev.c:5432) #15 0x00001fa8 in _start (crt.c:267) #16 0x00001e28 in start PPC Thread State: srr0: 0x000a9dc8 srr1: 0x0200f930 vrsave: 0x00000000 xer: 0x20000000 lr: 0x000a9dc0 ctr: 0x000c6968 mq: 0x00000000 r0: 0x000aa024 r1: 0xbff80040 r2: 0x6c616773 r3: 0xbff80080 r4: 0x0000002a r5: 0x0000002a r6: 0x00000001 r7: 0x00000002 r8: 0x727900e4 r9: 0x002e8a28 r10: 0x02a4f000 r11: 0x00020000 r12: 0x48042447 r13: 0x00000000 r14: 0x00000000 r15: 0x00000000 r16: 0x00000000 r17: 0x00000000 r18: 0x00387e40 r19: 0x002eb814 r20: 0x00000000 r21: 0x00000001 r22: 0xbff80200 r23: 0x0000002a r24: 0x0000b04b r25: 0xe972469b r26: 0x00883890 r27: 0x00000078 r28: 0x00000000 r29: 0x00011a6c r30: 0x00000001 r31: 0x000a9dc0 [G4:~] adriaan% gpc -v Reading specs from /Developer/Pascal/gpc332d1/lib/gcc-lib/powerpc-apple-darwin/3.3.2/specs Configured with: ../gpc-332d1/configure --enable-languages=pascal,c --prefix=/Developer/Pascal/gpc332d1 --enable-threads=posix --target=powerpc-apple-darwin Thread model: posix gpc version 20030830, based on gcc-3.3.2 The same problem also occurs with the previous alpha and gcc-3.3, but the backtrace is different: Date/Time: 2003-11-07 11:34:03 +0100 OS Version: 10.2.8 (Build 6R73) Host: G4.local. Command: gpc1 PID: 404 Exception: EXC_BAD_ACCESS (0x0001) Codes: KERN_INVALID_ADDRESS (0x0001) at 0xbff7ffe0 Thread 0 Crashed: #0 0x000c3b74 in alloc_page (ggc-page.c:660) #1 0x000c2b24 in ggc_alloc (ggc-page.c:1026) #2 0x0009c430 in make_node (tree.c:292) #3 0x000c25ac in alloc_node (stringpool.c:70) #4 0x0023cd30 in ht_lookup (hashtable.c:161) #5 0x000c2748 in get_identifier (stringpool.c:109) #6 0x000097d4 in get_operator (expressions.c:426) #7 0x0000a478 in build_pascal_binary_op (expressions.c:656) #8 0x0000a1ac in parser_build_binary_op (expressions.c:578) #9 0x0002d8d8 in main_yyparse (parse.c:2400) #10 0x00096e88 in compile_file (toplev.c:2134) #11 0x0009c03c in do_compile (toplev.c:5370) #12 0x0009c108 in toplev_main (toplev.c:5400) #13 0x000018a8 in _start (crt.c:267) #14 0x00001728 in start PPC Thread State: srr0: 0x000c3b74 srr1: 0x0000f930 vrsave: 0x00000000 xer: 0x00000000 lr: 0x000c3b6c ctr: 0x0009c3a0 mq: 0x00000000 r0: 0x000c2b24 r1: 0xbff80060 r2: 0x6c616773 r3: 0x00000023 r4: 0xbff80280 r5: 0x00000027 r6: 0x00000000 r7: 0x00000003 r8: 0xbff80100 r9: 0x002db53c r10: 0x00000001 r11: 0x0000002c r12: 0x88048444 r13: 0x00000003 r14: 0x000000c8 r15: 0xbfffec80 r16: 0x00000034 r17: 0x00000000 r18: 0x0036c9a0 r19: 0x002d94e4 r20: 0x00000000 r21: 0x02fa3e70 r22: 0x00000001 r23: 0xbff80280 r24: 0x00302ad0 r25: 0x00302ad0 r26: 0x00824890 r27: 0x0000008c r28: 0x03423e70 r29: 0x0000002c r30: 0x002fc69c r31: 0x000c3b6c The needed units can be found at <www.microbizz.nl/GPCMacOSAll.pas.bz2> and <www.microbizz.nl/GPCStringsAll.pas.bz2>. They were too large to include with this message. Regards, Adriaan van Os

4 6

interface and implementation
by mkhaldi＠gosympatico.ca 02 Apr '04

02 Apr '04

I have a question about implementation and interface. If i want only to get the interface from a module which has an implementation module, how must I do to get also the symbol init_XXXXX?? example : ---------------------------------------------- module a interface; export a =all; type string80 = string(80); procedure p; end. module a implementation; procedure p; ...... ...... end. module b interface; export b =all; import a; end. module b impleementation; .... end. I just want to get the type definition without the implementation of module a. In the b.o file, init_A symbol appears. ------------------------------------------------- Actually, I compile my modules in 2 steps. The first one is with --interface-only to get gpi file. Then, I compile my module with --implementation-only. Sometimes, I only need the interface files. But, importing a module always adds the init_XXX symbol. So what is the solution?

3 2

Open source tax
by Adriaan van Os 01 Apr '04

01 Apr '04

Today, the government in The Netherlands introduced a tax on free and open source software. The purpose of the tax is to "protect the local commercial software industry from unfair compettion from open source and free software projects". For details, see linux news at <http://www.nedlinux.nl/modules/news/index.php?action=showitem&id=1001>. Regards, Adriaan van Os

3 3

multiprocessor use w/ GPC compiled program (OT?)
by Boris Herman 01 Apr '04

01 Apr '04

Hi list, I have compiled a single-threaded (and quite non MP aware) pascal program that does a lot of numerical computations. One calculation takes about 5 days on my 1.25 GHz DDR G4 (single cpu) and I need about 200 computations to do. Well, when I run the program on my G4 it takes up all CPU time (cca 93%-95%). I thought of speeding the whole thing by running it on a dual processor machine so I got someone to run it on a dual 2 GHz G5. So here comes the problem. If only one instance is run on a dual G5 the Activity monitor shows 50% CPU load (which is understandable) but the graph alternates between 100% load on 1st CPU and 100% load on 2nd CPU. I thought that is an anomaly of Activity Monitor. However, when two instances are run, the time required to execute each task doubles even though the Activity monitor shows both cpus at 100% load. My program doesn't really do anything much - it performs gazillions of computations in arrays that all together consume less than 20 Mb of memory. Am I doing anything wrong or shouldn't I expect better performance on a dual cpu unit? Highest regards, Boris

6 5

Re: Open source tax
by Theo.CarrBrion＠mitrefinch.co.uk 01 Apr '04

01 Apr '04

How about a tax on software that has an unfair monopoly. Theo Carr-Brion -----Original Message----- From: Adriaan van Os [mailto:gpc@microbizz.nl] Sent: 01 April 2004 14:56 To: gpc(a)gnu.de Subject: Open source tax Today, the government in The Netherlands introduced a tax on free and open source software. The purpose of the tax is to "protect the local commercial software industry from unfair compettion from open source and free software projects". For details, see linux news at <http://www.nedlinux.nl/modules/news/index.php?action=showitem&id=1001>. Regards, Adriaan van Os

1 0

packed records' address again
by Francesco Bonomi 31 Mar '04

31 Mar '04

I have been following the "cannot take address of packed record" thread and now I have a similar problem. I perfectly understand that doing a strcopy is an implicit conversion, but I think that the program gets carried away during checks :-) I had a VERY complex data structure and it took me quite a while to reduce it to the minimum, please bear with me if it's anyway a bit intricate. let's have the following definitions: type char_arr = array[1..30] of char; smallrec = RECORD arr : char_arr; END; largerec = packed RECORD one_rec : smallrec; many_rec : array[1..5] of smallrec; one_array : char_arr; many_arrays : array[1..5] of char_arr; END; VAR r:largerec; c:char_arr; The largerec record is packed, and contains an assortment of structures. A lot of assignments are allowed, like: r.one_array :=c; r.one_rec.arr :=c; r.many_arrays[1] :=c; This single one is not allowed: r.many_rec[1].arr:=c; BTW, assigning a single char is allowed: r.many_rec[1].arr[1]:=#0; So I guess it must be a check that deals with implicits type conversions. but: 1) I am not converting strings, I am assigning a char_arr to another char_arr 2) if the "r.one_rec.arr :=c" assignment is allowed, I cannot see how the "r.many_rec[1].arr:=c;" can be more "dangerous" in addressing mode. I am not trying to be fussy just for the fun of it, but my structure is (as life goes :-) just the single variant that will not work. I am porting an old source code to gpc, while keeping the data file compatibility, so I will explore the unpack-access-pack solution, (hoping that it will not degrade performances too much) BTW (and here I go into newbie mode) how can I check if there is after all any physical difference between a packed and an un-packed record? Is there something like a length() to measure a record's size in memory?? Also, do I understand correctly that the proposed solution: > {$pack-struct, maximum-field-alignment 8} > foo = record > sig : array [ 0..3 ] OF Char; > end; > {$no-pack-struct, maximum-field-alignment 0} would change the structure of a "file of foo" ? Thanks! Francesco Bonomi Francesco Bonomi Si.lab s.r.l. Via Citille 13 50022 Greve in Chianti FI- Italy Tel +39 - 055 8544805 Fax +39 - 055 8547418 http//:www.silab.it

6 8

gpc reference reader
by Russell Whitaker 29 Mar '04

29 Mar '04

hi The attachment, readhelp.pas.gz, is a simple, fast reader for the reference portion of the gpc manual. Uncompress with gunzip. The first few lines after the start is a "readme". Be sure to read it before compiling. The plan here is to merge this program with an updated version of the simple Pascal Editor I contributated a little over a year ago. When done will contribute the result. Meanwhile, perhaps this little program will encourage someone to work on the gpc manual. russ

1 0