From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S262762AbVF3Arr (ORCPT ); Wed, 29 Jun 2005 20:47:47 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S262761AbVF3Arr (ORCPT ); Wed, 29 Jun 2005 20:47:47 -0400 Received: from mailout.stusta.mhn.de ([141.84.69.5]:43026 "HELO mailout.stusta.mhn.de") by vger.kernel.org with SMTP id S262762AbVF3Ark (ORCPT ); Wed, 29 Jun 2005 20:47:40 -0400 Date: Thu, 30 Jun 2005 02:47:38 +0200 From: Adrian Bunk To: ocfs2-devel@oss.oracle.com, linux-kernel@vger.kernel.org Subject: [-mm patch] CONFIGFS_FS: "If unsure, say N." Message-ID: <20050630004738.GA27478@stusta.de> References: <20050624080315.GC26545@stusta.de> <20050629213038.GA23823@ca-server1.us.oracle.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20050629213038.GA23823@ca-server1.us.oracle.com> User-Agent: Mutt/1.5.9i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Wed, Jun 29, 2005 at 02:30:38PM -0700, Joel Becker wrote: > On Fri, Jun 24, 2005 at 10:03:15AM +0200, Adrian Bunk wrote: > > I haven't found any reason why CONFIGFS_FS is user-visible. > > Other parts of the kernel using configfs should simply select it. > > Doesn't work for external modules that might want to use it. > Imagine that configfs gets merged before OCFS2, which depends on it. I was surprised if configfs was merged with zero users in the kernel. But I get your point, what about the patch below? > Joel cu Adrian <-- snip --> Make it clear that users usually shouldn't manually enable CONFIGFS_FS. Signed-off-by: Adrian Bunk --- linux-2.6.12-mm2-full/fs/Kconfig.old 2005-06-30 01:51:51.000000000 +0200 +++ linux-2.6.12-mm2-full/fs/Kconfig 2005-06-30 01:54:01.000000000 +0200 @@ -934,13 +934,11 @@ tristate "Userspace-driven configuration filesystem (EXPERIMENTAL)" depends on EXPERIMENTAL help - configfs is a ram-based filesystem that provides the converse - of sysfs's functionality. Where sysfs is a filesystem-based - view of kernel objects, configfs is a filesystem-based manager - of kernel objects, or config_items. + This option is provided for the case where no in-kernel-tree + modules require configfs, but a module built outside the kernel + tree does. Such modules require Y or M here. - Both sysfs and configfs can and should exist together on the - same system. One is not a replacement for the other. + If unsure, say N. config RELAYFS_FS tristate "Relayfs file system support"