From mboxrd@z Thu Jan 1 00:00:00 1970 From: Mark Lord Subject: Re: ata_std_qc_defer not good enough for FIS-based switching ? Date: Tue, 29 Apr 2008 13:17:06 -0400 Message-ID: <48175812.7000106@rtr.ca> References: <48163C5D.9050605@rtr.ca> <48164AE8.4070106@rtr.ca> <481659B5.7090703@gmail.com> <48175664.5030702@rtr.ca> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit Return-path: Received: from rtr.ca ([76.10.145.34]:3798 "EHLO mail.rtr.ca" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750989AbYD2RRH (ORCPT ); Tue, 29 Apr 2008 13:17:07 -0400 In-Reply-To: <48175664.5030702@rtr.ca> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Tejun Heo Cc: Jeff Garzik , IDE/ATA development list , Alan Cox Mark Lord wrote: .. > For the first time ever, I can now *heavily* mix PIO commands > with NCQ/non-NCQ DMA commands, to drives on an attached PM > (well, two attached PMs even), and the thing doesn't croak. > Woo-hoo! .. Here is the simple torture test I'm using for this. It might be fun for others to try something similar on the other LLDs which have PMP support: 1. QuadCore CPU, with 4GB of RAM (lots of simultaneous threads). 2. In one window, do this (lots of overlapping PIO commands): while ( true ) ; do for d in /dev/sd? ; do bash -c "hdparm -C $d &" ; done ; sleep 0.1 ; done 3. In a second window, do this (lots of overlapping DMA/NCQ reads): while (true) ; do for d in /dev/sd? ; do bash -c "hdparm -t $d &" ; done ; sleep 6 ; done Or something like that. Cheers