From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S933390AbXBCAYB (ORCPT ); Fri, 2 Feb 2007 19:24:01 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S933399AbXBCAYB (ORCPT ); Fri, 2 Feb 2007 19:24:01 -0500 Received: from outpost.ds9a.nl ([213.244.168.210]:46483 "EHLO outpost.ds9a.nl" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S933390AbXBCAYA (ORCPT ); Fri, 2 Feb 2007 19:24:00 -0500 Date: Sat, 3 Feb 2007 01:23:59 +0100 From: bert hubert To: Linus Torvalds Cc: Ingo Molnar , Alan , Zach Brown , linux-kernel@vger.kernel.org, linux-aio@kvack.org, Suparna Bhattacharya , Benjamin LaHaise Subject: Re: [PATCH 2 of 4] Introduce i386 fibril scheduling Message-ID: <20070203002358.GA26422@outpost.ds9a.nl> Mail-Followup-To: bert hubert , Linus Torvalds , Ingo Molnar , Alan , Zach Brown , linux-kernel@vger.kernel.org, linux-aio@kvack.org, Suparna Bhattacharya , Benjamin LaHaise References: <20070202104900.GA13941@elte.hu> <20070202195932.15b9b4ed@localhost.localdomain> <20070202213019.202e8db5@localhost.localdomain> <20070202224259.GB1212@elte.hu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.9i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Fri, Feb 02, 2007 at 03:17:57PM -0800, Linus Torvalds wrote: > threads. But you need to look at what it is we parallelize here, and ask > yourself why we're doing what we're doing, and why people aren't *already* > just using a separate thread for it. Partially this is for the bad reason that creating "i/o threads" (or even processes) has a bad stigma to it, and additionally has always felt crummy. On the first reason, the 'pain' of creating threads is actually rather minor, so this feeling may have been wrong. The main thing is that you don't wantonly create a thousand i/o threads, whereas you conceivably might want to have a thousand outstanding i/o requests. At least I know I want to have that ability. Secondly, the actual mechanics of i/o processes isn't trivial, and feels wasteful with lots of additional copying, or in the case of threads, queueing and posting. Bert -- http://www.PowerDNS.com Open source, database driven DNS Software http://netherlabs.nl Open and Closed source services