From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 15191 invoked by alias); 28 Mar 2008 02:51:32 -0000 Received: (qmail 15182 invoked by uid 22791); 28 Mar 2008 02:51:32 -0000 X-Spam-Check-By: sourceware.org Received: from NaN.false.org (HELO nan.false.org) (208.75.86.248) by sourceware.org (qpsmtpd/0.31) with ESMTP; Fri, 28 Mar 2008 02:51:12 +0000 Received: from nan.false.org (localhost [127.0.0.1]) by nan.false.org (Postfix) with ESMTP id 82461983AA; Fri, 28 Mar 2008 02:51:10 +0000 (GMT) Received: from caradoc.them.org (22.svnf5.xdsl.nauticom.net [209.195.183.55]) by nan.false.org (Postfix) with ESMTP id 635DC983A1; Fri, 28 Mar 2008 02:51:10 +0000 (GMT) Received: from drow by caradoc.them.org with local (Exim 4.69) (envelope-from ) id 1Jf4gP-0000vH-2l; Thu, 27 Mar 2008 22:51:09 -0400 Date: Sun, 30 Mar 2008 21:26:00 -0000 From: Daniel Jacobowitz To: Doug Evans Cc: gdb Subject: Re: constructor breakpoints Message-ID: <20080328025108.GA3529@caradoc.them.org> Mail-Followup-To: Doug Evans , gdb References: MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.17 (2007-12-11) X-IsSubscribed: yes Mailing-List: contact gdb-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-owner@sourceware.org X-SW-Source: 2008-03/txt/msg00271.txt.bz2 On Thu, Mar 27, 2008 at 05:04:57PM -0700, Doug Evans wrote: > Having recently tripped over this (not being able to set breakpoints > on constructors), and having read as much of the sordid history of > this as I can find, I wonder what the current thinking is on how to > fix this. Has there been any progress? Not that I know of. Hey, at least we fixed the _other_ breakpoints in constructors bug. > The ?HERE's are because symbol lookup finds the class, not the > constructor. Being able to specifically ask for the constructor seems > like a reasonable way to go. This could be done by encoding the > request in the name (looking up c::c), or specifying the class of > object to lookup (e.g. LOC_BLOCK), or rearranging the symbol table > such that one can lookup the class, and then from that lookup (or > iterate) over members that match certain parameters. [I'm just > throwing out a range of possibilities.] I like that last one - I always have - but I'm a little worried about how much work it would be in practice; neither stabs nor dwarf symbol info is really laid out quite right for it. -- Daniel Jacobowitz CodeSourcery