From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 12010 invoked by alias); 17 Nov 2008 23:35:46 -0000 Received: (qmail 11979 invoked by uid 22791); 17 Nov 2008 23:35:45 -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, 17 Nov 2008 23:34:55 +0000 Received: from nan.false.org (localhost [127.0.0.1]) by nan.false.org (Postfix) with ESMTP id 0470C10DE1; Mon, 17 Nov 2008 23:34:52 +0000 (GMT) Received: from caradoc.them.org (209.195.188.212.nauticom.net [209.195.188.212]) by nan.false.org (Postfix) with ESMTP id AF1DC10D50; Mon, 17 Nov 2008 23:34:52 +0000 (GMT) Received: from drow by caradoc.them.org with local (Exim 4.69) (envelope-from ) id 1L2DcK-0001nl-2T; Mon, 17 Nov 2008 18:34:52 -0500 Date: Tue, 18 Nov 2008 12:59:00 -0000 From: Daniel Jacobowitz To: Doug Evans Cc: Ulrich Weigand , gdb-patches@sourceware.org Subject: Re: [RFA] fix popping of dummy frame if inferior gets signal with unwindonsignal Message-ID: <20081117233452.GA6909@caradoc.them.org> Mail-Followup-To: Doug Evans , Ulrich Weigand , gdb-patches@sourceware.org References: <20081117200140.GA18828@caradoc.them.org> <200811172042.mAHKg6rd008540@d12av02.megacenter.de.ibm.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: 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: 2008-11/txt/msg00454.txt.bz2 On Mon, Nov 17, 2008 at 02:37:10PM -0800, Doug Evans wrote: > On Mon, Nov 17, 2008 at 12:42 PM, Ulrich Weigand wrote: > > Daniel Jacobowitz wrote: > >> On Mon, Nov 17, 2008 at 08:50:13PM +0100, Ulrich Weigand wrote: > >> > I'm not sure what the point is of having two copies of the regcache > >> > preserved across the full extent of the inferior call ... > >> > >> Maybe in case we get to somewhere we can not backtrace from? > > > > OK, good point. The inf_status will always work ... > > I don't understand why two copies are needed. > Can either of you elaborate? Hypothetically, if you call a function which completely trashes the registers but not the previous stack, we might not be able to find the dummy frame to pop. So having it saved somewhere else seems reasonable. I don't know how to structure that in a useful least-surprising way though... -- Daniel Jacobowitz CodeSourcery