From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S262370AbVFUXzp (ORCPT ); Tue, 21 Jun 2005 19:55:45 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S262435AbVFUXz3 (ORCPT ); Tue, 21 Jun 2005 19:55:29 -0400 Received: from mail.dvmed.net ([216.237.124.58]:2474 "EHLO mail.dvmed.net") by vger.kernel.org with ESMTP id S262370AbVFUXjP (ORCPT ); Tue, 21 Jun 2005 19:39:15 -0400 Message-ID: <42B8A51A.8020208@pobox.com> Date: Tue, 21 Jun 2005 19:39:06 -0400 From: Jeff Garzik User-Agent: Mozilla Thunderbird 1.0.2-6 (X11/20050513) X-Accept-Language: en-us, en MIME-Version: 1.0 To: Christoph Lameter CC: Robert Love , Andrew Morton , linux-kernel@vger.kernel.org Subject: Re: -mm -> 2.6.13 merge status References: <20050620235458.5b437274.akpm@osdl.org> <42B831B4.9020603@pobox.com> <1119368364.3949.236.camel@betsy> <42B8987F.9000607@pobox.com> In-Reply-To: Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Score: 0.0 (/) Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Christoph Lameter wrote: > On Tue, 21 Jun 2005, Jeff Garzik wrote: > > >>Non-blocking file I/O is an open issue. >> >>AIO is probably a better path. > > > AIO is requiring you to poll and check if I/O is complete. select() does Incorrect. The entire point of AIO is that its an async callback system, when the I/O is complete... just like the kernel's internal I/O request queue system. Jeff