From: Nick Piggin Subject: Re: [patch 2/2] fs: introduce mnt_clone_write Date: Tue, 10 Mar 2009 16:08:41 +0100 Message-ID: <20090310150841.GB8579@wotan.suse.de> References: <20090310143718.GB15977@wotan.suse.de> <20090310143801.GC15977@wotan.suse.de> <20090310145558.GY25995@parisc-linux.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Linux Kernel Mailing List , linux-fsdevel@vger.kernel.org, Andrew Morton , Dave Hansen To: Matthew Wilcox Return-path: Received: from ns2.suse.de ([195.135.220.15]:52958 "EHLO mx2.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754410AbZCJPIo (ORCPT ); Tue, 10 Mar 2009 11:08:44 -0400 Content-Disposition: inline In-Reply-To: <20090310145558.GY25995@parisc-linux.org> Sender: linux-fsdevel-owner@vger.kernel.org List-ID: On Tue, Mar 10, 2009 at 08:55:58AM -0600, Matthew Wilcox wrote: > On Tue, Mar 10, 2009 at 03:38:01PM +0100, Nick Piggin wrote: > > +EXPORT_SYMBOL_GPL(mnt_clone_write); > > +EXPORT_SYMBOL_GPL(mnt_want_write_file); > > You don't add any users of these functions in this patch ... is there a > good reason to export them? Only to encourage modular code to use them (as a replacement for mnt_want_write -- which itself is exported). If that's not the right way to go, I don't mind waiting for a user to turn up. [I see many ioctls in btrfs (coming first in the alphabet) could probably easily use mnt_want_write_file (otoh probably none are so performance critical).]