Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S264344AbUD0UkF (ORCPT ); Tue, 27 Apr 2004 16:40:05 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S264345AbUD0UkF (ORCPT ); Tue, 27 Apr 2004 16:40:05 -0400 Received: from kinesis.swishmail.com ([209.10.110.86]:26631 "EHLO kinesis.swishmail.com") by vger.kernel.org with ESMTP id S264344AbUD0Uj4 (ORCPT ); Tue, 27 Apr 2004 16:39:56 -0400 Message-ID: <408EC60B.50507@techsource.com> Date: Tue, 27 Apr 2004 16:43:55 -0400 From: Timothy Miller MIME-Version: 1.0 To: Ken Moffat CC: linux-kernel@vger.kernel.org Subject: Re: IDE throughput in 2.6 - it's good! References: <408E7E79.9080405@techsource.com> In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1733 Lines: 48 Ken Moffat wrote: > On Tue, 27 Apr 2004, Timothy Miller wrote: > > >> >>Ken Moffat wrote: >> >> >>>So, despite the numbers shown by hdparm looking worse, when only one >>>user is doing anything the performance is actually improved. I've no >>>idea which changes have achieved this, but thanks to whoever were >>>involved. >> >> >>I've done tests using dd to and from the raw block device under 2.4 and >>2.6. Memory size (kernel boot param mem=) doesn't seem to affect >>performance, so I assume that means that dd to and from the raw block >>device is unbuffered. When I compare read and write speeds between 2.4 >>and 2.6, 2.6 is definately slower. The last 2.6 kernel I tried this >>with is 2.6.5. >> > > > Well, my original test used cp, sync, rm, sync. I've no statistics > from running 2.4 on this box to compare against. > Based on my experience, cp and anything else that uses the filesystem gets buffered. I can tell this because, without sync, the throughput varies with memory size. Furthermore, I wanted to know raw throughput, so I used a block device so I could eliminate filesystem overhead. Reading and writing the block device does not seem to be buffered because the run time is not affected by memory (host RAM) size. That is, unless dd does an implicit sync. The numbers I get when using dd to and from one of my drives under a 2.4 kernel with the drive connected to the on-board IDE controller are roughly the same as published benchmarks. - 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/