From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:34348) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1W3mFK-0002Kz-Lc for qemu-devel@nongnu.org; Thu, 16 Jan 2014 07:40:36 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1W3mFE-0007qZ-K0 for qemu-devel@nongnu.org; Thu, 16 Jan 2014 07:40:30 -0500 Received: from mx1.redhat.com ([209.132.183.28]:51517) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1W3mFE-0007qP-CO for qemu-devel@nongnu.org; Thu, 16 Jan 2014 07:40:24 -0500 Date: Thu, 16 Jan 2014 20:40:04 +0800 From: Fam Zheng Message-ID: <20140116124004.GB10829@T430.nay.redhat.com> References: <1389775713-996-1-git-send-email-famz@redhat.com> <1389775713-996-3-git-send-email-famz@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: Subject: Re: [Qemu-devel] [PATCH v16 2/9] rules.mak: allow per object cflags and libs List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Peter Maydell Cc: Kevin Wolf , Stefan Hajnoczi , Michael Tokarev , QEMU Developers , Alex Bligh , Paolo Bonzini , Miroslav Rezanina , =?iso-8859-1?Q?Llu=EDs?= Vilanova , Richard Henderson On Thu, 01/16 11:04, Peter Maydell wrote: > On 15 January 2014 08:48, Fam Zheng wrote:.objs. > > Similarly, > > > > foo.o-cflags := $(FOO_CFLAGS) > > > > is also supported. > > I noticed that we already support per-object cflags via: > > $(obj)/adlib.o $(obj)/fmopl.o: QEMU_CFLAGS += -DBUILD_Y8950=0 > > (this example from audio/Makefile.objs). Is your new > method better in some way? Are we going to convert > users of the old-style method? > I added this new method as it's cleaner to use, especially when defining multiple-object module (implemented in this series): foo.mo-objs := bar.o biz.o qux.o , compared to $(obj)/foo.mo-objs := $(obj)/bar.o $(obj)/biz.o $(obj)/qux.o or slightly better $(obj)/foo.mo-objs := $(addprefix $(obj), bar.o, biz.o, qux.o) I've already converted block/curl.o's cflags to this, later in this series. So yes I think this is a better syntax. Thanks, Fam