From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756118AbXK3T4w (ORCPT ); Fri, 30 Nov 2007 14:56:52 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1753609AbXK3T4a (ORCPT ); Fri, 30 Nov 2007 14:56:30 -0500 Received: from pentafluge.infradead.org ([213.146.154.40]:60239 "EHLO pentafluge.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752910AbXK3T43 (ORCPT ); Fri, 30 Nov 2007 14:56:29 -0500 Date: Fri, 30 Nov 2007 11:51:50 -0800 From: Greg KH To: Cornelia Huck , Kay Sievers , Alan Stern Cc: Kernel development list , Jonathan Corbet , Randy Dunlap Subject: [RFC] kobject_init changes Message-ID: <20071130195150.GA4659@kroah.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.5.16 (2007-06-09) Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org New thread time... Here are two patches, one that adds a two new functions, kobject_init_ng() and kobject_init_and_add() and then a patch that shows how a number of different users are converted over to these new functions. I choose the name "kobject_init_ng()" for now, so that I can do a patch series that converts the differnet usages in the kernel tree, and then do a global rename all at once, making it easier to find any potential problems that might happen. Same goes for the _and_add() function, which is really just kobject_register() but without the uevent call, and it has proper memory cleanup. Any objections to these changes? My plans with this are: - clean up the rest of the kernel tree usages of kobject_init() - rename kobject_init_ng() to kobject_init(). - remove all users of kobject_register() and use the new functions and a uevent call. - fix up the new kobject_create_and_register() function to be create_and_add() instead. The last one will require some patch reworking in my tree, but I have four days of meetings next week, and two plane rides, so I'll need something real to do :) I suppose if I really get bored, I can fix up the cdev stuff, but that's going to take a lot more work... thanks, greg k-h