From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 6861 invoked by alias); 14 Jan 2004 15:07:06 -0000 Mailing-List: contact gdb-patches-help@sources.redhat.com; run by ezmlm Precedence: bulk List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-patches-owner@sources.redhat.com Received: (qmail 6840 invoked from network); 14 Jan 2004 15:07:05 -0000 Received: from unknown (HELO nevyn.them.org) (66.93.172.17) by sources.redhat.com with SMTP; 14 Jan 2004 15:07:05 -0000 Received: from drow by nevyn.them.org with local (Exim 4.30 #1 (Debian)) id 1Agmbo-0001CD-FW; Wed, 14 Jan 2004 10:07:04 -0500 Date: Wed, 14 Jan 2004 15:07:00 -0000 From: Daniel Jacobowitz To: r6144 Cc: gdb-patches@sources.redhat.com Subject: Re: [PATCH][newbie] Make gdb accept -feliminate-dwarf2-dups stuff Message-ID: <20040114150704.GA4517@nevyn.them.org> Mail-Followup-To: r6144 , gdb-patches@sources.redhat.com References: <16389.22843.352984.894253@localhost.localdomain> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <16389.22843.352984.894253@localhost.localdomain> User-Agent: Mutt/1.5.1i X-SW-Source: 2004-01/txt/msg00370.txt.bz2 On Wed, Jan 14, 2004 at 10:59:07PM +0800, r6144 wrote: Content-Description: message body text > Hello, everyone: > > Current CVS gcc has an option "-feliminate-dwarf2-dups" that > generates dwarf2 debug info containing references across compilation > units, which current gdb seems to be unable to parse. I wrote a > quick-and-dirty patch (against 20040103 snapshot) to make it work. > > The patch isn't really clean (in particular it contains a known memory > leak as indicated by the FIXME), and I don't really expect it to be > included. However, I think this patch *might* be useful to someone, > since "-feliminate-dwarf2-dups" seems to reduce debug info size > significantly for large projects (the effect is more clear when the > DW* symbols for DWARF-2 relocation are deleted), which may be helpful > to distributors making debug-info packages. Maybe some of you may > tell me some way to improve the patch, as well as to make it > accessible to those needing similar functionality. > > Hopefully you'll CC me since I'm not subscribed. Thanks. Unfortunately, we can't fix up and include the patch unless you file a copyright assignment with the FSF. You can do that, or wait - a whole lot of people have needed this feature lately, and if no one with an assignment beats me to it I'll probably do it myself this month. I skimmed your patch and it looks like you're doing it rather differently than I planned to. > +#include Neat, I hadn't seen this before... -- Daniel Jacobowitz MontaVista Software Debian GNU/Linux Developer