From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 5778 invoked by alias); 29 Sep 2008 18:03:25 -0000 Received: (qmail 5768 invoked by uid 22791); 29 Sep 2008 18:03:24 -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; Mon, 29 Sep 2008 18:02:29 +0000 Received: from nan.false.org (localhost [127.0.0.1]) by nan.false.org (Postfix) with ESMTP id AC36D10D24; Mon, 29 Sep 2008 18:02:26 +0000 (GMT) Received: from caradoc.them.org (22.svnf5.xdsl.nauticom.net [209.195.183.55]) by nan.false.org (Postfix) with ESMTP id 799191019C; Mon, 29 Sep 2008 18:02:26 +0000 (GMT) Received: from drow by caradoc.them.org with local (Exim 4.69) (envelope-from ) id 1KkN4j-0006pl-RM; Mon, 29 Sep 2008 14:02:25 -0400 Date: Mon, 29 Sep 2008 18:03:00 -0000 From: Daniel Jacobowitz To: Joel Brobecker Cc: Pedro Alves , gdb-patches@sourceware.org, Tom Tromey Subject: Re: PR gdb/856 Message-ID: <20080929180225.GA25971@caradoc.them.org> Mail-Followup-To: Joel Brobecker , Pedro Alves , gdb-patches@sourceware.org, Tom Tromey References: <20080929165712.GA3811@adacore.com> <200809291811.39619.pedro@codesourcery.com> <20080929171750.GB3811@adacore.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20080929171750.GB3811@adacore.com> User-Agent: Mutt/1.5.17 (2008-05-11) X-IsSubscribed: yes Mailing-List: contact gdb-patches-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-patches-owner@sourceware.org X-SW-Source: 2008-09/txt/msg00560.txt.bz2 On Mon, Sep 29, 2008 at 10:17:50AM -0700, Joel Brobecker wrote: > > Daniel's answer is here, but I never got back to it: > > http://sourceware.org/ml/gdb-patches/2008-06/msg00036.html > > PC and block: I don't see why passing the block would be required > for inlining support - as I understand it, we can always compute > the actual block from the PC. No? No. There can be more than one block at the PC. Normally this isn't a big deal. Anything from blocks further out is visible further in. There are two exceptions. One of them is already a problem and GDB does not support it helpfully: shadowed variables. { int x; { int x; ... } } The other is more noticeable; inlining. When you say "up" you move to the call site. And the call site might be in a different file, and have different macros defined. -- Daniel Jacobowitz CodeSourcery