From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1758489Ab2D0Bxt (ORCPT ); Thu, 26 Apr 2012 21:53:49 -0400 Received: from mail-vb0-f46.google.com ([209.85.212.46]:40033 "EHLO mail-vb0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1757706Ab2D0Bxs (ORCPT ); Thu, 26 Apr 2012 21:53:48 -0400 Message-ID: <4F99FC28.80504@gmail.com> Date: Thu, 26 Apr 2012 21:53:44 -0400 From: Joe Ceklosky User-Agent: Mozilla/5.0 (X11; Linux i686; rv:11.0) Gecko/20120315 Firefox/11.0 SeaMonkey/2.8 MIME-Version: 1.0 To: linux-kernel@vger.kernel.org Subject: Possible 32-bit PAE kernel I/O slow down with 3.3.X Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org I have notice significant slow downs with disk I/O on a 32-bit PAE 3.3.X kernels. The test machine has 16GB memory, Intel i5 CPU, Intel motherboard, and an Intel 160GB SSD. I copied 536M from one partition on an SSD to another. Yes, I should be using a 64bit kernel with 16GB! Test 1: kernel 3.3.2 32-bit PAE with kernel parm mem=3G time cp -ra testdir/ /tmp real 0m8.989s user 0m0.094s sys 0m1.080s Test 2: kernel 3.3.2 32-bit PAE time cp -ra testdir/ /tmp real 1m17.068s user 0m0.173s sys 0m2.149s Also I tested kernel 3.2.16 32-bit PAE with no memory limits 16GB and it was in line with test 1. I can repeat and provide any debugging information if needed. Thanks, Joe Ceklosky