From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S964860AbWBMUgr (ORCPT ); Mon, 13 Feb 2006 15:36:47 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S964864AbWBMUgr (ORCPT ); Mon, 13 Feb 2006 15:36:47 -0500 Received: from smtp.osdl.org ([65.172.181.4]:62384 "EHLO smtp.osdl.org") by vger.kernel.org with ESMTP id S964860AbWBMUgq (ORCPT ); Mon, 13 Feb 2006 15:36:46 -0500 Date: Mon, 13 Feb 2006 12:35:47 -0800 From: Andrew Morton To: Roman Zippel Cc: mingo@elte.hu, tglx@linutronix.de, linux-kernel@vger.kernel.org Subject: Re: [PATCH 02/13] hrtimer: remove useless const Message-Id: <20060213123547.41af78c4.akpm@osdl.org> In-Reply-To: References: <1139830116.2480.464.camel@localhost.localdomain> <20060213114612.GA30500@elte.hu> X-Mailer: Sylpheed version 1.0.4 (GTK+ 1.2.10; i386-redhat-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Roman Zippel wrote: > > Hi, > > On Mon, 13 Feb 2006, Ingo Molnar wrote: > > > your patch makes code larger on gcc3. Please investigate why. > > Ok, I checked with 3.3.6 and 3.4.5 and adding const to ktime_divns/ > posix_cpu_nsleep fixes the problem (actually the kernel becomes even > smaller because posix_cpu_clock_getres is better off without const), both > are complex and noncritical functions. > Andrew, I'd really prefer to keep the patch and I'll send a patch to add > the const where it's actually an improvement. > There's been enough churn here that it'll be counterproductive for me to continue to munge on your patches. I'll drop the lot and will await rev #2. I don't think any of this was critical for 2.6.16. If there's something which you really do think needs to be in 2.6.16 then please carefully describe the reasons for that, and make sure that the most critical patches are staged at the head of the series.