Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932462AbVKZDJX (ORCPT ); Fri, 25 Nov 2005 22:09:23 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S932683AbVKZDJX (ORCPT ); Fri, 25 Nov 2005 22:09:23 -0500 Received: from ns.ustc.edu.cn ([202.38.64.1]:42969 "EHLO mx1.ustc.edu.cn") by vger.kernel.org with ESMTP id S932462AbVKZDJX (ORCPT ); Fri, 25 Nov 2005 22:09:23 -0500 Date: Sat, 26 Nov 2005 11:17:55 +0800 From: Wu Fengguang To: Diego Calleja Cc: linux-kernel@vger.kernel.org, akpm@osdl.org Subject: Re: [PATCH 00/19] Adaptive read-ahead V8 Message-ID: <20051126031755.GA7226@mail.ustc.edu.cn> Mail-Followup-To: Wu Fengguang , Diego Calleja , linux-kernel@vger.kernel.org, akpm@osdl.org References: <20051125151210.993109000@localhost.localdomain> <20051125164317.c42c0639.diegocg@gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20051125164317.c42c0639.diegocg@gmail.com> User-Agent: Mutt/1.5.11 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1377 Lines: 32 On Fri, Nov 25, 2005 at 04:43:17PM +0100, Diego Calleja wrote: > Recently, a openoffice hacker wrote in his blog that the kernel was > culprit of applications not starting as fast as in other systems. > Most of the reasons he gave were wrong, but there was a interesting > one: When you start your system, you've lots of free memory. Since > you have lots of memory, he said it was reasonable to expect that > kernel would readahead *heavily* everything it can to fill that > memory as soon as possible (hoping that what you readahead'ed was > part of the kde/gnome/openoffice libraries etc) and go back to the > normal behaviour when your free memory is used by caches etc. > "Teorically" it looks like a nice heuristic for desktops. Does > adaptative readahead does something like this? It's interesting ;) In fact some distributions do have a read-ahead script to preload files on startup. The readahead system call should be enough for this purpose: NAME readahead - perform file readahead into page cache SYNOPSIS #include ssize_t readahead(int fd, off64_t *offset, size_t count); Thanks, Wu - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/