From: "Florian X" <dos DOT fire AT aon DOT at>
Newsgroups: comp.os.msdos.djgpp
References: <Pine DOT SUN DOT 3 DOT 91 DOT 1010121111642 DOT 13920S-100000 AT is>
Subject: Re: silly question: SIGSEGV
Date: Sun, 21 Jan 2001 11:22:50 +0100
X-Newsreader: Microsoft Outlook Express 4.71.1712.3
X-MimeOLE: Produced By Microsoft MimeOLE V4.71.1712.3
Lines: 74
Message-ID: <3a6ab865$0$25126@SSP1NO25.highway.telekom.at>
NNTP-Posting-Host: 212.183.93.153
X-Trace: newsreader01.vienna.highway.telekom.at 980072549 25126 212.183.93.153
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Reply-To: djgpp AT delorie DOT com


>Do you mean to say that this bug only shows up if you use DLX, and that
>using DLX somehow invalidates the EIP values printed in the crash message?

Here is the printed info from symify:

Call frame traceback EIPs:
  0x000a9b9f _free+247
  0x000016c4 _sf_free+24, line 81 of program.c
  0x00003d38 _dispose+52, line 1478 of program.c
  0xffd94c14 0xffd94c14   <------------- funtktion in the DLX program
  0x00003d23 _dispose+31, line 1476 of program.c
  0x00002870 _obj_done+92, line 657 of program.c
  0x000144c8 _view_done+176, line 129 of program.c
  0x00003d23 _dispose+31, line 1476 of program.c
  0xffd8c2fa 0xffd8c2fa <------------- funtktion in the DLX program
  0x000152cc _view_play_process+176, line 387 of program.c
  0x00003b78 _obj_translate_event+44, line 1242 of program.c
[DR-DOS] E:\SEALSRC>symify seal.exe


>Can you run the program under a debugger?  If so, then both GDB and RHIDE
>should be able to tell you a lot about each EIP value printed in the
>crash message.  Perhaps you didn't use the right commands.

I used it, but they couldn't show me more infos, because of the DLXs and
because some *.c files are
included into the main file program.c via  fe. #include"object.c".

So rhgdb and rhide only opens program.c and cannot show me the line, where
the error is :(((

BTW: SEAL is a Allegro program, and I can debug once or twice, but when I
exit the debuger, I have to use
the RESET-Button. (In Dr-DOS (with and without Dr-DOS DPMI), MS-DOS and
Win95).

>You _must_ find the place in the source code where it crashes.  There's
>no other way.
>
>If nothing else helps, you could selectively remove parts of the code in
>the approximate area where it crashes, until you arrive at a single line
>or maybe a small number of lines whose removal causes the crashes to
>appear.

Yes, I tried it. I think it tries to remove a memory block which wasn't
right allocated via malloc.

All free() commands where replaced by a macro like this:

#define _free(x) sf_free(x)

void  sf_free ( void *rec )
{
  if ( rec ) free(rec);
  rec = 0;
};


>
>> Other people have no problem with this program, some has also.....why??
>
>Memory-related bugs are usually like that.

I hate this bug!! Another problem is, that I haven't developed it, now it is
a open source project, and I have to
find the bug in a very very big program and have to learn the source.

Thanks, florian
www.drdos.org
www.seal.de.vu
-- This mail was written by user of The Arachne Browser - http://arachne.cz/