From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 28595 invoked by alias); 11 Feb 2004 15:39:13 -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 28496 invoked from network); 11 Feb 2004 15:39:12 -0000 Received: from unknown (HELO localhost.redhat.com) (66.30.197.194) by sources.redhat.com with SMTP; 11 Feb 2004 15:39:12 -0000 Received: by localhost.redhat.com (Postfix, from userid 469) id 2DBC31A4484; Wed, 11 Feb 2004 10:35:29 -0500 (EST) From: Elena Zannoni MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <16426.19392.847031.448743@localhost.redhat.com> Date: Wed, 11 Feb 2004 15:39:00 -0000 To: David Carlton Cc: Elena Zannoni , gdb-patches , Daniel Jacobowitz , Michael Elizabeth Chastain Subject: Re: [rfa] teach linespec about nested classes In-Reply-To: References: <16424.12248.555001.90026@localhost.redhat.com> X-SW-Source: 2004-02/txt/msg00291.txt.bz2 David Carlton writes: > On Mon, 9 Feb 2004 20:11:52 -0500, Elena Zannoni said: > > > The approach looks ok, but, how does the HP related comment fit in > > with the new code? I know it came in with the HP merge, which is a > > clue as to its accuracy.... I guess MichaelC found no problems, so > > it should be ok. > > As far as I can tell from that and from other comments elsewhere, HP > must have been the first people with a C++ compiler that supported > namespaces and that GDB supported. So some of the namespace-related > comments look more HP-specific than they were. Also, as far as I can > tell, HP was generating fully-qualified type names long before we > were, so that would have led to some differences as well. But now I > think that the names should look quite similar for DWARF 2 code and HP > code, so those comments shouldn't be relevant any more. > definitely HP were first on a lot of the c++ stuff, yes. And they changed a lot of decode_lie_1 (now decode_compound). > I couldn't think of any reason why looking up every intermediate name > as a class would make any more sense for HP than it would for other > cases. And, as you say, MichaelC found no problems, which is good. > probably their compiler has changed too, in 5+ years (this stuff dates back to 1997). > > I have a new version which incorporates the new comments. Does this > > still work for you? > > The functionality is fine, but the comments are a little off (it > includes the out-of-date HP comments instead of my new ones). How > about this version? I updated my comments to use the same example > that you had used. groan, I missed that. Sure, you can commit this.