From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Miller Subject: Re: [PATCH 00/68] ide2libata Date: Sun, 31 Jan 2010 23:47:26 -0800 (PST) Message-ID: <20100131.234726.124893364.davem@davemloft.net> References: <20100129160308.21495.14120.sendpatchset@localhost> <4B6355DF.9050001@garzik.org> <20100129232540.41f6c4f6@lxorguk.ukuu.org.uk> Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Return-path: Received: from 74-93-104-97-Washington.hfc.comcastbusiness.net ([74.93.104.97]:32805 "EHLO sunset.davemloft.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751042Ab0BAHrN (ORCPT ); Mon, 1 Feb 2010 02:47:13 -0500 In-Reply-To: <20100129232540.41f6c4f6@lxorguk.ukuu.org.uk> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: alan@lxorguk.ukuu.org.uk Cc: jeff@garzik.org, bzolnier@gmail.com, linux-ide@vger.kernel.org, linux-kernel@vger.kernel.org From: Alan Cox Date: Fri, 29 Jan 2010 23:25:40 +0000 > The old IDE "maintenance mode" seems to be drifting - the rate of change > is rather high for that claim. I think there's another angle to this. By making the IDE layer build from the same code as the ATA driver, the legacy IDE layer gets an indirect tester base. I like that. However what I don't like is how this is implemented. We shouldn't pretend the data structures are the same by using macros in some header file, we should truly abstract out the data types properly such that these drivers in fact use the same datastructures. For the price of a few series of data structure morphs, we eliminate the tester-base issue of legacy IDE. There's one driver for both ATA and legacy IDE, the stuff in front is just a presentation and probing layer, nothing more. The cost and risk is testing the morphing changes, but I think the scales tip towards making this truly worth it. Bart, thanks for working on this and publishing what you came up with.