From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932664AbVIHPq7 (ORCPT ); Thu, 8 Sep 2005 11:46:59 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S932704AbVIHPq6 (ORCPT ); Thu, 8 Sep 2005 11:46:58 -0400 Received: from fed1rmmtao10.cox.net ([68.230.241.29]:60889 "EHLO fed1rmmtao10.cox.net") by vger.kernel.org with ESMTP id S932664AbVIHPqq (ORCPT ); Thu, 8 Sep 2005 11:46:46 -0400 Date: Thu, 8 Sep 2005 08:46:45 -0700 From: Tom Rini To: Jan Beulich Cc: linux-kernel@vger.kernel.org Subject: Re: [PATCH] i386 CFI annotations Message-ID: <20050908154645.GN3966@smtp.west.cox.net> References: <432070850200007800024465@emea1-mh.id2.novell.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <432070850200007800024465@emea1-mh.id2.novell.com> User-Agent: Mutt/1.5.9i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Thu, Sep 08, 2005 at 05:10:29PM +0200, Jan Beulich wrote: > As a foundation for reliable stack unwinding, this adds CFI unwind > annotations to many low-level i386 routines, plus a config option > (available to all architectures) to enable them as well as the > compiler > producing similar information for all C sources. > > Signed-off-by: Jan Beulich x86_64 has proper CFI annotations on things, and just depends on DEBUG_INFO. Perhaps that would be a slightly easier path to use for pushing this along? > + CFI_ADJUST_CFA_OFFSET 4;\ > + /*CFI_REL_OFFSET es, 0;*/\ > pushl %ds; \ > + CFI_ADJUST_CFA_OFFSET 4;\ > + /*CFI_REL_OFFSET ds, 0;*/\ Adding new commented out code never wins new friends. :) > diff -Npru 2.6.13/include/asm-i386/dwarf2.h [snip] > +#ifdef CONFIG_UNWIND_INFO [snip] > +#else [snip] > +#define CFI_STARTPROC ignore Why not just empty defines? -- Tom Rini http://gate.crashing.org/~trini/