Re: [PATCH] 64 bit scsi read/write

From: "Ken Hirsch" <kenhirsch@myself.com>
To: "Chris Wedgwood" <cw@f00f.org>, "John Alvord" <jalvo@mbay.net>
Cc: "Daniel Phillips" <phillips@bonn-fries.net>,
	"Alan Cox" <alan@lxorguk.ukuu.org.uk>,
	"Andrew Morton" <andrewm@uow.edu.au>,
	"Andreas Dilger" <adilger@turbolinux.com>,
	"Albert D. Cahalan" <acahalan@cs.uml.edu>,
	"Ben LaHaise" <bcrl@redhat.com>,
	"Ragnar Kjxrstad" <kernel@ragnark.vestdata.no>,
	<linux-fsdevel@vger.kernel.org>, <linux-kernel@vger.kernel.org>,
	<mike@bigstorage.com>, <kevin@bigstorage.com>,
	<linux-lvm@sistina.com>
Subject: Re: [PATCH] 64 bit scsi read/write
Date: Sun, 15 Jul 2001 09:16:09 -0400	[thread overview]
Message-ID: <005501c10d30$54e0e260$7c853dd0@hppav> (raw)
In-Reply-To: <Pine.LNX.4.20.0107142304010.17541-100000@otter.mbay.net> <20010715180752.B7993@weta.f00f.org>

Chris Wedgwood <cw@f00f.org> wrote:
> On Sat, Jul 14, 2001 at 11:05:36PM -0700, John Alvord wrote:
>
>     In the IBM solution to this (1977-78, VM/CMS) the critical data was
>     written at the begining and the end of the block. If the two data
items
>     didn't match then the block was rejected.
>
> Neat.
>
>
> Simple and effective.  Presumably you can also checksum the block, and
> check that.

The first technique is not sufficient with modern disk controllers, which
may reorder sector writes within a block.  A checksum, especially a robust
CRC32, is sufficient, but rather expensive.

Mohan has a clever technique that is computationally trivial and only uses
one bit per sector: http://www.almaden.ibm.com/u/mohan/ICDE95.pdf

Unfortunately, it's also patented:
http://www.delphion.com/details?pn=US05418940__

Perhaps IBM will clarify their position with respect to free software and
patents in the upcoming conference.