From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932648AbVHTDW1 (ORCPT ); Fri, 19 Aug 2005 23:22:27 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S932650AbVHTDW1 (ORCPT ); Fri, 19 Aug 2005 23:22:27 -0400 Received: from smtp.istop.com ([66.11.167.126]:13460 "EHLO smtp.istop.com") by vger.kernel.org with ESMTP id S932648AbVHTDW0 (ORCPT ); Fri, 19 Aug 2005 23:22:26 -0400 From: Daniel Phillips To: Greg KH Subject: Re: [PATCH] Permissions don't stick on ConfigFS attributes Date: Sat, 20 Aug 2005 13:23:29 +1000 User-Agent: KMail/1.7.2 Cc: Joel Becker , linux-kernel@vger.kernel.org References: <200508201050.51982.phillips@istop.com> <20050820030117.GA775@kroah.com> In-Reply-To: <20050820030117.GA775@kroah.com> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Content-Disposition: inline Message-Id: <200508201323.29355.phillips@istop.com> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Saturday 20 August 2005 13:01, Greg KH wrote: > On Sat, Aug 20, 2005 at 10:50:51AM +1000, Daniel Phillips wrote: > > So: Integrate with sysfs. > > No, don't. Do you think that Joel would not have already worked with > the sysfs people prior to submitting this? No, he did, and we all > agreed that it should be kept separate. Would you care to recap the reasoning, please? > > Terminology skew. It is a very bad idea to call your configfs files > > "attributes". > > That's what sysfs calls its files. They used the same naming scheme > there. This is nothing that a user ever cares about or sees. It's wrrrrronnnggg. The best you can defend this with is "it's entrenched". > > Memory requirements. ConfigFS pins way too much kernel memory for inodes > > and dentries. > > configfs is not going to have that many nodes at all in memory (compared > to sysfs), so I don't think this is a big problem. The current bloat is unconscionable, for the amount of data that is carried. Are you arguing against fixing it? And what makes you think configfs will never have lots of nodes? Regards, Daniel