From mboxrd@z Thu Jan 1 00:00:00 1970 From: Harvey Harrison Subject: Re: [PATCH] rndis: remove private wrapper of __constant_cpu_to_le32 Date: Wed, 11 Feb 2009 14:16:13 -0800 Message-ID: <1234390573.5566.13.camel@brick> References: <1234389410.5566.5.camel@brick> <200902111412.18233.david-b@pacbell.net> Mime-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 7bit Cc: Greg KH , John Linville , David Miller , linux-netdev To: David Brownell Return-path: Received: from rn-out-0910.google.com ([64.233.170.186]:15953 "EHLO rn-out-0910.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754589AbZBKWRw (ORCPT ); Wed, 11 Feb 2009 17:17:52 -0500 Received: by rn-out-0910.google.com with SMTP id k40so467343rnd.17 for ; Wed, 11 Feb 2009 14:17:51 -0800 (PST) In-Reply-To: <200902111412.18233.david-b@pacbell.net> Sender: netdev-owner@vger.kernel.org List-ID: On Wed, 2009-02-11 at 14:12 -0800, David Brownell wrote: > On Wednesday 11 February 2009, Harvey Harrison wrote: > > Use cpu_to_le32 directly as it handles constant folding now, replace direct > > uses of __constant_cpu_to_{endian} as well. > > > > Signed-off-by: Harvey Harrison > > Acked-by: David Brownell > > ... though I'm trusting you that constant folding really *does* work > now, as it didn't when all this code was first written. Easily enough > seen if it doesn't: build on big-endian HW will break on all the > "case cpu_to_le32(constant):" branches. ;) Yes, and large parts of the kernel are already moved to using the base versions already. I fixed the constant-folding bits during the creation/consolidation of linux/swab.h Harvey