Date:	Wed, 26 Apr 2000 07:06:47 -0400 (EDT)
Message-Id:	<200004261106.HAA23678@indy.delorie.com>
From:	Eli Zaretskii <eliz AT delorie DOT com>
To:	Richard Dawe <richdawe AT bigfoot DOT com>
CC:	djgpp AT delorie DOT com
In-reply-to:	<3905EEC5.836F724B@bigfoot.com> (message from Richard Dawe on
	Tue, 25 Apr 2000 20:15:17 +0100)
Subject:	Re: gdb problem or just silly me - please help
References:	<Pine DOT SUN DOT 3 DOT 91 DOT 1000425112446 DOT 23128H-100000 AT is> <3905EEC5 DOT 836F724B AT bigfoot DOT com>
Reply-To:	djgpp AT delorie DOT com
Errors-To:	nobody AT delorie DOT com
X-Mailing-List:	djgpp AT delorie DOT com
X-Unsubscribes-To:	listserv AT delorie DOT com

> From: Richard Dawe <richdawe AT bigfoot DOT com>
> Date: Tue, 25 Apr 2000 20:15:17 +0100
> 
> Eli Zaretskii wrote:
> > The problem with using -gstabs+ by default is that SYMIFY doesn't
> > support stabs, so -gstabs+ renders stack tracebacks unusable, unless
> > you run the program inside GDB or RHIDE.
> > 
> > If and when someone teaches SYMIFY about stabs, DJGPP will probably
> > switch to -gstabs+ as the default.
> 
> Eli, ISTR that you had a BFD version of symify that did work with stabs+
> debugging. Maybe I am confused, but is this the case? I remember you sent
> me bfdsymify when I had some weird debugging problems, which we never got
> to the bottom of.

Yes, I do have a version of SYMIFY linked against the BFD library, and
that version should support stabs.

However, I use neither C++ nor stabs myself, so I don't have an easy
way of testing how well does it work, and debug whatever bugs might be
there.

At the time, I asked on djgpp-workers who would be interested to use
that version of SYMIFY and help me debug it, but never saw any
responses.  Seems like no one is interested in getting this done.

(Testing and debugging a tool such as SYMIFY requires some knowledge
in how debug info is recorded in the executable, and it also requires
heavy use of Binutils programs such as objcopy and objdump.  So this
is not a task for ``normal'' DJGPP users, which is why I didn't post
the request for help on this news group.)

For that matter, there's lots of things to be done in improving debug
support in general.  For example, adding support for core files would
be nice.  Adding code to SYMIFY to interpret additional parts of the
crash message, to make them tell their story in language understood by
humans, is another nice project.

It's too bad nobody is working in that area (AFAIK).

- Raw text -