Date: Wed, 1 May 1996 19:49:22 +0800 (GMT+0800)
From: Orlando Andico <orly AT abigail DOT eee DOT upd DOT edu DOT ph>
To: Jan Hansen <jaha AT daimi DOT aau DOT dk>
cc: djgpp AT delorie DOT com
Subject: Re: GDB breakpoint problem
In-Reply-To: <4m2f1t$s2s@gjallar.daimi.aau.dk>
Message-ID: <Pine.LNX.3.91.960501194656.2446A-100000@abigail.eee.upd.edu.ph>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII

Strange problem, <ick!>

We really gotta accept that MS-DOG even with DJGPP ain't UNIX. I would be 
willing to bet that your problem is caused by too small a DPMI stack; use 
stubedit to enlarge it. You can also try using symify to print the call 
traceback. If you see some really weird functions (nothing like your 
functions, something like __djgpp_something) then it's almost certainly 
some shitty stack problem (that is, if the program really isn't at fault, 
which would be your implication if it works perfectly on a UNIX 
workstation). 

On 29 Apr 1996, Jan Hansen wrote:

> I have a strange problem while using GDB.
> Its not just one program causing the trouble, it won't work at all.
> First i set a breakpoint in a line. Then i run the program.
> When the execution reaches the breakpoint, i get this message:
>   exception SIGSEGV, segmentation fault
> And then i can't continue (using n or s)
> 
> i'm using a 120MHz Pentium with 8Mb ram.
> The debugger is running in a Windows 3.11 dos window.
> Both compiler and debugger is version 2.? 
> 
> I have tried GDB on a UNIX workstation, and that works just fine.
> Please help!!!!
> 
> Btw. is it somehow possible to start a program by singlestepping (n), 
> Without running and then stopping the execution with a breakpoint first.
> 
> Thanks. 
> Jan Hansen.
>