From mboxrd@z Thu Jan 1 00:00:00 1970 From: James Bottomley Subject: Re: [PATCH RFC 1/3] add fc transport events Date: 12 Jun 2004 23:41:10 -0400 Sender: linux-scsi-owner@vger.kernel.org Message-ID: <1087088793.1730.19.camel@mulgrave> References: <40B597F5.2030501@cs.wisc.edu> Mime-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 7bit Return-path: Received: from stat1.steeleye.com ([65.114.3.130]:12947 "EHLO hancock.sc.steeleye.com") by vger.kernel.org with ESMTP id S264984AbUFMDlV (ORCPT ); Sat, 12 Jun 2004 23:41:21 -0400 In-Reply-To: <40B597F5.2030501@cs.wisc.edu> List-Id: linux-scsi@vger.kernel.org To: Mike Christie Cc: SCSI Mailing List On Thu, 2004-05-27 at 03:25, Mike Christie wrote: > 01-add-host-transport-classdev.patch - adds the transport class_device to > the scsi_host structure. I've been thinking about this quite a bit, and I like the general idea, I just have a quibble with the implementation. The main problem is the adding of a kobject to complement your host transport class (and the device transport class). I don't think we need to do this since, like the transport classes fall into either host or device objects, so the events are similarly divided; thus, we should be able to trigger the hotplug events through the host or device kobject (that come with the embedded struct device) rather than adding another kobject for the purpose. What would be the disadvantage of doing what you propose this way instead of using a separate kobject? James