Date: Tue, 24 Dec 2002 07:49:41 +0200 (IST)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
X-Sender: eliz AT is
To: Richard Dawe <rich AT phekda DOT freeserve DOT co DOT uk>
cc: djgpp-workers AT delorie DOT com
Subject: Re: RESEND: gdb 5.2.1 segfaults while debugging fileutils 4.1
In-Reply-To: <3E070DD2.6420C1E4@phekda.freeserve.co.uk>
Message-ID: <Pine.SUN.3.91.1021224074416.24775A-100000@is>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Reply-To: djgpp-workers AT delorie DOT com
Errors-To: nobody AT delorie DOT com
X-Mailing-List: djgpp-workers AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk


On Mon, 23 Dec 2002, Richard Dawe wrote:

> I've found the problem or, at least, a cause. I had compiled lib/*.c with a
> different version of gcc, so it had different debugging info to compiled
> src/*.c - stabs ("dbx") vs. DWARF-2. Does gdb support this? (I don't see why
> not, but perhaps it has not been tested much.)

Thanks for digging into this.  It doesn't surprise me that GDB has 
trouble working with mixed stabs/dwarf-2 debug info: I doubt if anybody 
has ever tried to debug such a program.

Please try to come up with a minimal complete program and compilation 
procedure that reproduces this problem.  It would be nice if the way the 
library was compiled had no effect on the outcome; ideally, the library 
should be compiled with no debug info at all.

Armed with such a test case, please post the details to the GDB mailing 
list <gdb AT sources DOT redhat DOT com>.  (I think several bugs which led to 
crashes were fixed in symtab.c and related code lately, but I don't know 
whether this bug is one of them.  However, it does mean that someone is 
actively working on symtab.c, so this might be a good time to report 
such bugs.)