From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S263754AbTDDPL1 (for ); Fri, 4 Apr 2003 10:11:27 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S263753AbTDDPKR (for ); Fri, 4 Apr 2003 10:10:17 -0500 Received: from ns.virtualhost.dk ([195.184.98.160]:7115 "EHLO virtualhost.dk") by vger.kernel.org with ESMTP id S263749AbTDDPCc (for ); Fri, 4 Apr 2003 10:02:32 -0500 Date: Fri, 4 Apr 2003 17:13:56 +0200 From: Jens Axboe To: Andries Brouwer Cc: Linux Kernel , Alan Cox Subject: Re: [PATCH] only use 48-bit lba when necessary Message-ID: <20030404151356.GA15124@suse.de> References: <20030404122936.GB786@suse.de> <20030404144044.GA14371@win.tue.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20030404144044.GA14371@win.tue.nl> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Fri, Apr 04 2003, Andries Brouwer wrote: > On Fri, Apr 04, 2003 at 02:29:36PM +0200, Jens Axboe wrote: > > > 48-bit lba has a non-significant overhead (twice the outb's, 12 instead > > s/non-// ? Of course :) > > of 6 per command), so it makes sense to use 28-bit lba commands whenever > > we can. > > > > > + if (drive->addressing == 1 && block > 0xfffffff) > > + lba48 = 1; > > Hmm. I wonder whether we should be more cautious, and ask for lba48 > as soon as some part of the interval is past this limit. > (say, block+nsectors > 0xfffffff) > > I don't know whether the standard spells out what happens > at the boundary, but for example the LBA low/mid/high, DEV is required > to contain the sector number at the place the error occurred, > and that is possible only if one stays below the 28-byte sector limit. That might not be a bad idea, just to be on the safe side. I'll do that. -- Jens Axboe