X-Authentication-Warning: acp3bf.physik.rwth-aachen.de: broeker owned process doing -bs
Date: Tue, 20 Jul 1999 15:47:15 +0200 (MET DST)
From: Hans-Bernhard Broeker <broeker AT physik DOT rwth-aachen DOT de>
X-Sender: broeker AT acp3bf
To: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
cc: djgpp AT delorie DOT com
Subject: Re: Bizarre debugging format problem
In-Reply-To: <Pine.SUN.3.91.990720162123.20658E-100000@is>
Message-ID: <Pine.LNX.3.93.990720153851.212G-100000@acp3bf>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Reply-To: djgpp AT delorie DOT com
X-Mailing-List: djgpp AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk


On Tue, 20 Jul 1999, Eli Zaretskii wrote:
> On Tue, 20 Jul 1999, Hans-Bernhard Broeker wrote:

> > There is a definite problem with resolution of source line information
> > in COFF format which I observed in binutils-2.8.1, as available from the
> > net, and it's sitting in the BFD library, which is used by both gdb
> > and binutils.
> 
> I'm not sure I understand.  Are you telling that Binutils 2.8.1 emit
> incorrect line info in the COFF debugging info?  If so, are you talking
> about stock Binutils 2.8.1 from GNU FTP sites or the version that is in
> DJGPP archives on SimTel? 

If memory serves, both of them. I compiled the DJGPP sources, but AFAIK,
the relevant part is not changed from the official FSF ones.

To put this in context: no, I haven't ever seen this problem with *gdb*,
only with binutils themselves (objdump -dl, gprof). That's why I asked the
original poster to make sure he's using stock DJGPP gdb, and not something
else. 

> And if there are real problems, could you describe them some more?  I 
> don't think anybody has ever said anything about such problems here.

I'm referring to the problems I observed in gprof. 'gprof -l', which also
relies on fetching line-number information, fails utterly if used on a
'gcc -g' compiled program, but works as soon as I either use 'gcc -gstabs'
or change the BFD library. The problem is that the code addresses in the
line number tables are stored as offsets, but with different bases for
COFF and for stabs debug infos.

I'm currently in the process of packaging up the changes I need for full
applicability of 'gprof'. They'll be relative to bnu281s.zip, first, and I
plan on sending them to the djgpp-workers for testing, before offering
them to the FSF. 

Hans-Bernhard Broeker (broeker AT physik DOT rwth-aachen DOT de)
Even if all the snow were burnt, ashes would remain.