From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jeff Garzik Subject: Re: siimage.c -serious- error fixed. Performance/stability issues resolved. Date: Tue, 02 Dec 2003 19:06:24 -0500 Sender: linux-ide-owner@vger.kernel.org Message-ID: <3FCD2900.1020002@pobox.com> References: <62929.67.101.15.58.1070362667.squirrel@mail.clanhk.org> <200312022016.13725.bzolnier@elka.pw.edu.pl> <33385.209.163.161.49.1070400815.squirrel@mail.clanhk.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Return-path: Received: from parcelfarce.linux.theplanet.co.uk ([195.92.249.252]:17835 "EHLO www.linux.org.uk") by vger.kernel.org with ESMTP id S264378AbTLCAGm (ORCPT ); Tue, 2 Dec 2003 19:06:42 -0500 In-Reply-To: <33385.209.163.161.49.1070400815.squirrel@mail.clanhk.org> List-Id: linux-ide@vger.kernel.org To: Ryan Earl Cc: Bartlomiej Zolnierkiewicz , linux-ide@vger.kernel.org Ryan Earl wrote: > The 15KB request buffer "fix"--*ahem* workaround--never addressed the real > issue. You want to get rid of all the 15KB hoocus-poocus. All that does > is kill your disk performance, it never solved the problem, merely made > your system more stable with the problem still around. Essentially, you > should never have that buffer set to 15KB and with this fix all that noise > should just disappear from the driver/kernel. Not correct. As has been discussed many times, with certain Seagate drives, sector_count % 15 == 1 && sector_count != 1 causes failure. This is what the 15KB limit addresses. Read libata SII driver for more info. Jeff