From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 10995 invoked by alias); 20 Mar 2003 22:19:33 -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 10977 invoked from network); 20 Mar 2003 22:19:32 -0000 Received: from unknown (HELO crack.them.org) (65.125.64.184) by sources.redhat.com with SMTP; 20 Mar 2003 22:19:32 -0000 Received: from nevyn.them.org ([66.93.61.169] ident=mail) by crack.them.org with asmtp (Exim 3.12 #1 (Debian)) id 18wAHB-0001VO-00; Thu, 20 Mar 2003 18:20:49 -0600 Received: from drow by nevyn.them.org with local (Exim 3.36 #1 (Debian)) id 18w8Nh-0001NV-00; Thu, 20 Mar 2003 17:19:25 -0500 Date: Thu, 20 Mar 2003 22:19:00 -0000 From: Daniel Jacobowitz To: Andrew Cagney Cc: Adam Fedor , GDB Patches Subject: Re: [RFA] Compile objc-lang.c, objc-exp.tab.c [1/5] Message-ID: <20030320221924.GA25955@nevyn.them.org> Mail-Followup-To: Andrew Cagney , Adam Fedor , GDB Patches References: <3E16093D.3070907@doc.com> <3E7A3226.9040904@redhat.com> <20030320213945.GA17345@nevyn.them.org> <3E7A3D14.1020100@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <3E7A3D14.1020100@redhat.com> User-Agent: Mutt/1.5.1i X-SW-Source: 2003-03/txt/msg00442.txt.bz2 On Thu, Mar 20, 2003 at 05:13:40PM -0500, Andrew Cagney wrote: > >Not without some substantial thought. We use cplus_demangle in a lot > >of places where we don't even know what the language is supposed to be > >- for minsyms, during lookups, et cetera. > > > >Certainly it needs to be thought about. At least objc's mangling is > >probably not entirely ambiguous with C++/Java's? I don't know. > > > >On Thu, Mar 20, 2003 at 04:27:02PM -0500, Andrew Cagney wrote: > > > >>Daniel, > >> > >>Would it be possible to make the cplus_demangle() method part of the > >>language vector? That way code like the patch below could be reduced to: > >> > >> /* Return demangled language symbol, or NULL. */ > >> language_demangle (current_language, arg); > >> > >>This would in turn allow Adam to just add an equivalent objc_demangle() > >>method to the objc language vector, and hence eliminate the need to > >>always link in objc-lang.c. > > As well as then? The places where objc is adding calls to the demangler > the language is known. My gut reaction is that it's just clutter until we decide how to solve the problem of not knowing demanglings. But it'll do for now. I'd like a comment along the lines of: /* FIXME: sometimes the demangler is invoked when we don't know the language, so we can't use this everywhere. */ -- Daniel Jacobowitz MontaVista Software Debian GNU/Linux Developer