From mboxrd@z Thu Jan 1 00:00:00 1970 From: Mark Lord Subject: Re: Single host/multiple hwifs vs. multiple hosts? Date: Mon, 09 Apr 2012 21:24:59 -0400 Message-ID: <4F838BEB.5090802@teksavvy.com> References: Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Return-path: Received: from ironport-out.teksavvy.com ([206.248.143.162]:60126 "EHLO ironport-out.teksavvy.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753691Ab2DJBeh (ORCPT ); Mon, 9 Apr 2012 21:34:37 -0400 In-Reply-To: Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Geert Uytterhoeven Cc: linux-ide@vger.kernel.org, linux-m68k On 12-04-08 05:12 AM, Geert Uytterhoeven wrote: > On Sun, Mar 4, 2012 at 13:15, Geert Uytterhoeven wrote: >> I'm wondering what are the differences between adding a single IDE host with >> multiple hwifs vs. adding a separate IDE host for each hwif? >> Do both ways work? > > Does no one know? When I first wrote that code, the idea was that if a given chip had multiple IDE ports (hardware interfaces, or hwifs for short), then that's how it would be done. One controller, multiple hwifs. This was before IDE had anything to do with the SCSI subsystem, and allowed for things like controller chips which could operate only a single hwif at any point in time (aka. "serialized"). Nowadays, the code's be hacked beyond recognition from its original form and intent, so who knows! :)