- Gpc - gnu.de

Stack problems?
by Steve Loft 26 Jan '01

26 Jan '01

I'm using gpc 20010115 on a Linux system. I'm having problems with a program I've written which has recursive function calls, which I thought may be down to blowing the stack. If I run the program in gdb, when it crashes I can't get it to display a backtrace: Program received signal SIGSEGV, Segmentation fault. 0x240804e7 in ?? () (gdb) bt #0 0x240804e7 in ?? () Cannot access memory at address 0xbabffff5 So I compiled it with --stack-checking, and now it crashes much earlier, with a segmentation fault on a 'var' statement. It's at about 20 levels of recursion, and the function only uses three of four integers of local variables; uname -s tells me I have 8196k of stack, so I'm puzzled as to why I'm blowing the stack, if indeed I am. Is a SIGSEGV what you would expect to see as a result of turning on stack checking? Are there any known problems in gpc in this area? Any suggestions as to what I can do to investigate further? Steve

4 6

documents on Compiling ASM
by Anuradha 25 Jan '01

25 Jan '01

Hello , I need a good manual to help me convert the Intel syntax into AT&T syntax. I had a look at : ftp://agnes.dida.physik.uni-essen.de/gnu-pascal/contrib/gpcasm.zip Though it is mentioned that i can have more information in http://www.dcs.warwick.ac.uk/~jules/nasm1.html and http://www.tu-chemnitz.de/~rho/rhide.html I am not able to get the right link because there are some missing letters in the link. can you kindly let me know the complete site? Thank you Regards, Anu

2 1

AT&T <-> Intel assembler converter and indentation program
by Thorsten Hindermann 22 Jan '01

22 Jan '01

Hello together! I have a 2 questions: 1. In the GPC ToDo-list is a point for an AT&T <-> Intel assembler converter. Eike Lange and I discusses to implement this feature in pindent (see anouncement earlier this month). But we are both not very good assembler programmers. The questions: is there is anybody (with very good experience in both assembler languages) who want help us to implement this feature in pindent so please feel free to contribute to this little project. We only talk about inline-assembler and pindent doesn't eat .s-files. Contact us in the GPC- or pindent- (pindent(a)mond.dida.physik.uni-essen.de) mailing-list. Thanx 2. Today I've look in the GPC ToDo-list and I see the indentation program request. If there is any reason to have this point in the list now? Because pindent is in its first version availible (http://mond.dida.physik.uni-essen.de/~pindent/) and Eike and I working on the next version. Thorsten

3 3

Re: verbose output of pcamac.pas
by Frank Heckenbach 17 Jan '01

17 Jan '01

Kocherlakota Harikrishna wrote: > Our main program was actually calling many units and we were getting a lot of errors .So we > thought we will concentrate on pcamac.pas alone and make a .o file by compiling it. As I said, use `-c' then. > Now we have written a simple program to use a procedure from pcamac. > Still we get the similar errors.We have attachedÂ the verbose ouput of it. OK, that's already fewer errors... :-) About the undefined reference to `toCall', `Get_loadpath' and `Moved'. These seem to be identifiers from your code, so I can only guess why they are missing: - You declared them `external' and didn't reference the unit containing them. - They were written in C or assembler and you didn't tell the compiler about it (you can use `{$L foo.c}' or `{$L foo.s}' or specify the file on the command line. - They were written in C or assembler or declared with an explicit `asmname' directive, and upper and lower case letters were mixed up. (Note, while Pascal identifiers are case-insensitive, assembler symbols, including `asmname's, and C identifiers, are case-sensitive.) If this doesn't help you find the problem, I can only help you further if you show me the places in the code where they are defined and used. About the multiple definition of `_p_strrscan'. `_p_strrscan' is the assembler symbol for the RTS routine CStringLastChPos. Do you happen to declare a routine of that name in yuor code? Though it's not reserved, that's currently a known bug that GPC will use the same assembler symbol then, resulting in such a problem. Please use another identifier then. If that's not the case, I don't know what's the problem, without seeing the code. You could try temporarily removing parts of pcamac.pas to find out what's causing the problem... 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: verbose output of pcamac.pas
by Frank Heckenbach 16 Jan '01

16 Jan '01

Kocherlakota Harikrishna wrote: > I am sorry for the inconvenience.Â Herewith we are sending the verbose output again. > > Yes.You are right. pcamac.pas is the main unit and pekxac.pas and pguzaa.pas are called > within it. Oh, pcamac.pas is a unit? I thought it was the main program. To compile a unit, you can use the `-c' option. But normally you don't need to do that at all. Just compile the main program with `--automake', and all units will be compiled when necessary. (Of course, you don't get an executable when compiling just a unit, you need a main program for that -- just like in BP and probably every other compiler.) 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

test -e (used in GPC's config-lang.in) isn't portable
by Alexandre Oliva 15 Jan '01

15 Jan '01

This patch fixes it. -- Alexandre Oliva Enjoy Guarana', see http://www.ic.unicamp.br/~oliva/ Red Hat GCC Developer aoliva(a){cygnus.com, redhat.com} CS PhD student at IC-Unicamp oliva(a){lsd.ic.unicamp.br, gnu.org} Free Software Evangelist *Please* write to mailing lists, not to me

2 6

None
by John Ollason 13 Jan '01

13 Jan '01

Greetings, I think I might have discovered a bug in gpc. A test program follows. John Ollason ------------------- { The following program illustrates a bug discovered in gpc. The program functions correctly with the set 0..31, but with the set 0..32 or greater the second call to card returns zero, and the contents of the set disappear after the first call. The following is the output from gpc --verbose Using builtin specs. gpc version 19990118, based on gcc-2.8.1 ld -m elf_i386 -dynamic-linker /lib/ld-linux.so.2 /usr/lib/crt1.o /usr/lib/crti.o /usr/lib/gcc-lib/i386-redhat-linux/2.8.1/crtbegin.o -L/usr/lib/gcc-lib/i386-redhat-linux/2.8.1 -L/usr/i386-redhat-linux/lib -lgpc -lm -lgcc -lc -lgcc /usr/lib/gcc-lib/i386-redhat-linux/2.8.1/crtend.o /usr/lib/crtn.o /usr/lib/crt1.o(.text+0x18): undefined reference to `main' } program settest(input,output); var i:integer; a: array[1..1] of record stuff:set of 0..32; end; begin with a[1] do begin stuff:=[]; stuff:=stuff+[2]; stuff:=stuff+[3]; writeln(card(stuff)); for i in stuff do writeln(i); writeln(card(stuff)); for i in stuff do writeln(i); end end.

2 2

compilation problems
by Kocherlakota Harikrishna 13 Jan '01

13 Jan '01

Hello, We have some problems in compilation of BP sources with GPC on Red Hat Linux. The following peice of code in our source is giving copilation problem procvar = procedure; const nilproc : procvar = nil; BP supports assinging pointer values to procedure types and we find that GPc doesnot support it. Expecting your advice on solving the above. regards hari

2 5

GPC bug
by Nicola Girardi 13 Jan '01

13 Jan '01

Hi GPC'ers! >From time to time I get a coredump, and sometimes I manage to find the reason. : nikke@RubberBone(tty4):/tmp$ cat bug.pas : program Bug; : type : Pa = ^a; : begin : end. : nikke@RubberBone(tty4):/tmp$ gpc bug.pas : gpc: Internal compiler error: program gpc1 got fatal signal 11 GPC should warn about the dependency (`a' isn't defined anywhere) but shouldn't coredump... (Sorry if the bug was known.) -- Nicola Girardi <nikke(a)swlibero.org>

2 1

Re: Gtk
by Eike Lange 10 Jan '01

10 Jan '01

Hi Tom! I'll forward this to the list, because it might be interesting to others as well. On Tue, Jan 09, 2001 at 09:48:41PM +0100, Tom Verhoeff wrote: > I read about your effort to port Gtk to GPC on the mailing list. > The URLs given there don't seem to work, at least not for me. E.g.: > ftp://agnes.dida.physik.uni-essen.de/~lange/gtkunit-20001119.tar.gz > Can you tell me what the current status of your porting effort is? We (Nicola and I) are working on it. I removed gtkunit-20001119.tar.gz because this unit is outdated and we have nothing new as stable as the old unit. I hope, we finish the unit at the end of january 2001 :-)) We will give an announcement to the list, where to download. Eike

1 0