Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753137AbZFCAEX (ORCPT ); Tue, 2 Jun 2009 20:04:23 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751713AbZFCAEQ (ORCPT ); Tue, 2 Jun 2009 20:04:16 -0400 Received: from fgwmail5.fujitsu.co.jp ([192.51.44.35]:34508 "EHLO fgwmail5.fujitsu.co.jp" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751343AbZFCAEP (ORCPT ); Tue, 2 Jun 2009 20:04:15 -0400 Date: Wed, 3 Jun 2009 09:02:43 +0900 From: KAMEZAWA Hiroyuki To: Andrew Morton Cc: Dhaval Giani , menage@google.com, lizf@cn.fujitsu.com, balbir@linux.vnet.ibm.com, linux-kernel@vger.kernel.org, bharata@linux.vnet.ibm.com, libcg-devel@lists.sourceforge.net Subject: Re: + cgroups-forbid-noprefix-if-mounting-more-than-just-cpuset-subsystem. patch added to -mm tree Message-Id: <20090603090243.5a2f566c.kamezawa.hiroyu@jp.fujitsu.com> In-Reply-To: <20090602133534.d2e453f6.akpm@linux-foundation.org> References: <200906020602.n5262wcu008560@imap1.linux-foundation.org> <20090602090944.GB16753@linux.vnet.ibm.com> <6599ad830906020908s7cb6afe6pd83ca35b82d148a4@mail.gmail.com> <20090602174042.GD16753@linux.vnet.ibm.com> <20090602133534.d2e453f6.akpm@linux-foundation.org> Organization: FUJITSU Co. LTD. X-Mailer: Sylpheed 2.5.0 (GTK+ 2.10.14; i686-pc-mingw32) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 2182 Lines: 80 On Tue, 2 Jun 2009 13:35:34 -0700 Andrew Morton wrote: > On Tue, 2 Jun 2009 23:10:42 +0530 > Dhaval Giani wrote: > > > On Tue, Jun 02, 2009 at 09:08:04AM -0700, Paul Menage wrote: > > > On Tue, Jun 2, 2009 at 2:09 AM, Dhaval Giani wrote: > > > > > > > > I am not sure if this is a good idea. For libcgroup, we would then be > > > > adding a special case for just cpuset. I would rather that we allow it > > > > either for all the subsystems or none of them. > > > > > > > > > > libcgroup shouldn't be using the noprefix option. Its only intentded > > > use is to allow the legacy "cpuset" filesystem type to be mounted and > > > to see the same fileset as it had before the cgroups transition. > > > > > > > It does not. But if some user is using that option, we need to be in a > > position to handle it. > > > > I am quite happy not supporting the noprefix option in the library if it > > is fine. > > > > fyi, the above discussion transitions akpm into the "confused" state. > I'll keep the patch on hold until akpm transitions back out of that > state. > Traffic control... [What "noprefix" is] - When "noprefix" is used, the name of file under cgroup is.. [Without noprefix] ... (dir)/subsysname.filename [With noprefix] ... (dir)/filename Then, cpuset's files will be [Without noprefix] ... (dir)/cpuset.xxx [With noprefix] ... (dir)/xxx This is for _backward compatibility_. [Problem] cpustat subsys has "stat" file. memory subsys has "stat" file. So, these cannot be mounted at the same mount point with "noprefix". Considering arbitrary subsys can be mounted at the same point, allowing "noprefix" other than cpuset just makes user-land complex, and "noprefix" itself is troublesome, it breaks naming rule for cgroup files. Then, I vote for that this patch should go. Thanks, -Kame > -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/