From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932306AbWKDB0j (ORCPT ); Fri, 3 Nov 2006 20:26:39 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S932307AbWKDB0i (ORCPT ); Fri, 3 Nov 2006 20:26:38 -0500 Received: from omx2-ext.sgi.com ([192.48.171.19]:4328 "EHLO omx2.sgi.com") by vger.kernel.org with ESMTP id S932306AbWKDB0i (ORCPT ); Fri, 3 Nov 2006 20:26:38 -0500 Date: Fri, 3 Nov 2006 17:26:05 -0800 From: Paul Jackson To: Andrew Morton Cc: clameter@sgi.com, linux-kernel@vger.kernel.org Subject: Re: Avoid allocating during interleave from almost full nodes Message-Id: <20061103172605.e646352a.pj@sgi.com> In-Reply-To: <20061103143145.85a9c63f.akpm@osdl.org> References: <20061103134633.a815c7b3.akpm@osdl.org> <20061103143145.85a9c63f.akpm@osdl.org> Organization: SGI X-Mailer: Sylpheed version 2.2.4 (GTK+ 2.8.3; i686-pc-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 Andrew wrote: > But in this application which you are proposing, any correlation with > elapsed walltime is very slight. It's just the wrong baseline to use. > What is the *sense* in it? Ah - but time is cheap as dirt, and scales like the common cold virus. That makes it sinfully attractive for secondary affect placement cache hints like this. What else would you suggest? Same question applies, I suppose, to my zonelist caching patch that is sitting in your *-mm patch stack, where you also had doubts about using wall clock time to decay the fullnode hints. -- I won't rest till it's the best ... Programmer, Linux Scalability Paul Jackson 1.925.600.0401