From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751090AbVI2MJ2 (ORCPT ); Thu, 29 Sep 2005 08:09:28 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751112AbVI2MJ2 (ORCPT ); Thu, 29 Sep 2005 08:09:28 -0400 Received: from 85-50-2-236.bcn1.adsl.uni2.es ([85.50.2.236]:40812 "EHLO puil.ghetto") by vger.kernel.org with ESMTP id S1751090AbVI2MJ1 (ORCPT ); Thu, 29 Sep 2005 08:09:27 -0400 Date: Thu, 29 Sep 2005 14:09:13 +0200 To: linux-kernel@vger.kernel.org Subject: Fastest way to do socket IO on Linux Message-ID: <20050929120912.GA12054@larroy.com> Mail-Followup-To: linux-kernel@vger.kernel.org Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline Content-Transfer-Encoding: 8bit User-Agent: Mutt/1.5.9i From: piotr@larroy.com (Pedro Larroy) Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Hello Which is the fastest way to do socket IO from an scalable perspective? Afaik AIO doesn't work on sockets yet, and select has too much overhead when you have to test if many fds are set, epoll could be it? Is expected to have AIO working on sockets some day? Regards. -- Pedro Larroy Tovar, pedro at larroy dot com | http://pedro.larroy.com/ * Las patentes de programación son nocivas para la innovación * http://proinnova.hispalinux.es/