From owner-freebsd-arch@FreeBSD.ORG  Mon Sep 20 05:56:00 2010
Return-Path: <owner-freebsd-arch@FreeBSD.ORG>
Delivered-To: freebsd-arch@freebsd.org
Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34])
	by hub.freebsd.org (Postfix) with ESMTP id 73F8F1065672;
	Mon, 20 Sep 2010 05:56:00 +0000 (UTC)
	(envelope-from davidxu@freebsd.org)
Received: from freefall.freebsd.org (freefall.freebsd.org
	[IPv6:2001:4f8:fff6::28])
	by mx1.freebsd.org (Postfix) with ESMTP id 62AC48FC16;
	Mon, 20 Sep 2010 05:56:00 +0000 (UTC)
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by freefall.freebsd.org (8.14.4/8.14.4) with ESMTP id o8K5twM4058772;
	Mon, 20 Sep 2010 05:55:59 GMT (envelope-from davidxu@freebsd.org)
Message-ID: <4C9767EE.3000802@freebsd.org>
Date: Mon, 20 Sep 2010 13:55:58 +0000
From: David Xu <davidxu@freebsd.org>
User-Agent: Thunderbird 2.0.0.24 (X11/20100630)
MIME-Version: 1.0
To: David Xu <davidxu@freebsd.org>, freebsd-arch@freebsd.org,
	Marcel Moolenaar <xcllnt@mac.com>, Bruce Evans <brde@optusnet.com.au>
References: <4C972E83.7090603@freebsd.org> <20100920024542.GA25717@zim.MIT.EDU>
In-Reply-To: <20100920024542.GA25717@zim.MIT.EDU>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Cc: 
Subject: Re: Per-source CFLAGS again
X-BeenThere: freebsd-arch@freebsd.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Discussion related to FreeBSD architecture <freebsd-arch.freebsd.org>
List-Unsubscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-arch>,
	<mailto:freebsd-arch-request@freebsd.org?subject=unsubscribe>
List-Archive: <http://lists.freebsd.org/pipermail/freebsd-arch>
List-Post: <mailto:freebsd-arch@freebsd.org>
List-Help: <mailto:freebsd-arch-request@freebsd.org?subject=help>
List-Subscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-arch>,
	<mailto:freebsd-arch-request@freebsd.org?subject=subscribe>
X-List-Received-Date: Mon, 20 Sep 2010 05:56:00 -0000

David Schultz wrote:
> The reason I proposed the feature was so that we could compile the
> parts of libc that come from a third party (namely, gdtoa) with
> different warning flags than the rest of libc.  It was not
> committed because someone else proposed a simpler way to get make
> to use different flags for gdtoa.
> 
> I suppose it wouldn't hurt to commit it it's needed for something
> else, but I think Marcel also had some philosophical objections to
> the proposal.

This time, the compiler's function is required to generate some
stack unwinding info, it is not compiler's warning flag.
without modification to sys.mk:
I have to hack the libc's Makefile, but obviously, CWARNFLAGS is not
the right place.

Index: Makefile
===================================================================
--- Makefile	
+++ Makefile	
@@ -145,3 +145,5 @@
  SSP_CFLAGS:=	${SSP_CFLAGS:S/^-fstack-protector-all$/-fstack-protector/}
  # Disable stack protection for SSP symbols.
  SSP_CFLAGS:=	${.IMPSRC:N*/stack_protector.c:C/^.+$/${SSP_CFLAGS}/}
+# Generate stack unwinding table for cancellation points
+CWARNFLAGS:= 
${.IMPSRC:Ncancelpoints_*:C/^.+$/${CWARNFLAGS}/:C/^$/-fexceptions 
${CWARNFLAGS}/}