From owner-cvs-ports@FreeBSD.ORG  Fri Apr  8 03:54:30 2011
Return-Path: <owner-cvs-ports@FreeBSD.ORG>
Delivered-To: cvs-ports@freebsd.org
Received: by hub.freebsd.org (Postfix, from userid 1033)
	id 1C535106566B; Fri,  8 Apr 2011 03:54:30 +0000 (UTC)
Date: Fri, 8 Apr 2011 03:54:30 +0000
From: Alexey Dokuchaev <danfe@FreeBSD.org>
To: bf1783@gmail.com
Message-ID: <20110408035430.GA56549@FreeBSD.org>
References: <201104072239.p37Md9oh031253@repoman.freebsd.org>
	<BANLkTikmUZsHNjw2Jpe8Y7v5CD6WkZGtoQ@mail.gmail.com>
	<20110408022459.GA40411@FreeBSD.org>
	<20110408023021.GA47293@FreeBSD.org>
	<BANLkTik4HrVvVNU_nBtH8Oyt=81uKjn6og@mail.gmail.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=koi8-r
Content-Disposition: inline
In-Reply-To: <BANLkTik4HrVvVNU_nBtH8Oyt=81uKjn6og@mail.gmail.com>
User-Agent: Mutt/1.4.2.1i
Cc: Rene Ladan <rene@freebsd.org>, cvs-ports@freebsd.org, cvs-all@freebsd.org,
	ports-committers@freebsd.org
Subject: Re: cvs commit: ports/www/chromium Makefile
X-BeenThere: cvs-ports@freebsd.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: CVS commit messages for the ports tree <cvs-ports.freebsd.org>
List-Unsubscribe: <http://lists.freebsd.org/mailman/listinfo/cvs-ports>,
	<mailto:cvs-ports-request@freebsd.org?subject=unsubscribe>
List-Archive: <http://lists.freebsd.org/pipermail/cvs-ports>
List-Post: <mailto:cvs-ports@freebsd.org>
List-Help: <mailto:cvs-ports-request@freebsd.org?subject=help>
List-Subscribe: <http://lists.freebsd.org/mailman/listinfo/cvs-ports>,
	<mailto:cvs-ports-request@freebsd.org?subject=subscribe>
X-List-Received-Date: Fri, 08 Apr 2011 03:54:30 -0000

On Fri, Apr 08, 2011 at 03:47:07AM +0000, b. f. wrote:
> On 4/8/11, Alexey Dokuchaev <danfe@freebsd.org> wrote:
> > On Fri, Apr 08, 2011 at 02:24:59AM +0000, Alexey Dokuchaev wrote:
> >> I've been thinking about this as well.  However, instead of bringing back
> >> this OPTION, I'd suggest adding "|| defined(PACKAGE_BUILDING)" to that
> >> .if to ensure that packages are can be safely run on pre-SSE2 CPUs.
> >
> > Diff attached, seems to DTRT.
> 
> Someone may want to build SSE2-enabled packages for i386 in a
> tinderbox or on a package-building machine, and the option allows them
> to do this, but this change does not.  Also, with this change, you're
> telling the configure script to disable SSE2 when PACKAGE_BUILDING is
> defined even on amd64, where SSE2 is always available.  So I don't
> think it DTRT.

One can always pass desired CPUTYPE is they want to build their package
with specific target.  As for amd64, here I agree, checking for ARCH
makes sence.  I don't mind this heuristics, as long as there is no SSE
option.  OPTIONS like this are bad since they are cached and can stay
unnoticed for long periods of time.

On the other hand, Chromium is web browser, which it most likely is going
to be used on users' desktop, where SSE2 is probably feels like home for
a while already.  :-)

./danfe