From mboxrd@z Thu Jan 1 00:00:00 1970 From: "J. Bruce Fields" Subject: Re: [patch 3/7] fs, notify: Add file handle entry into inotify_inode_mark Date: Wed, 14 Nov 2012 08:26:22 -0500 Message-ID: <20121114132622.GC23604@fieldses.org> References: <20121112101440.665694060@openvz.org> <2105540.yeyMVrW4mH@deuteros> <50A36ED5.4080505@parallels.com> <2092535.8S9EcgmZCZ@deuteros> <50A37692.1010500@parallels.com> <20121114124549.GB23604@fieldses.org> <20121114130347.GJ16685@moon> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Pavel Emelyanov , Tvrtko Ursulin , David Rientjes , Andrew Morton , linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org, Al Viro , Alexey Dobriyan , James Bottomley , Matthew Helsley , aneesh.kumar@linux.vnet.ibm.com To: Cyrill Gorcunov Return-path: Content-Disposition: inline In-Reply-To: <20121114130347.GJ16685@moon> Sender: linux-kernel-owner@vger.kernel.org List-Id: linux-fsdevel.vger.kernel.org On Wed, Nov 14, 2012 at 05:03:47PM +0400, Cyrill Gorcunov wrote: > On Wed, Nov 14, 2012 at 07:45:49AM -0500, J. Bruce Fields wrote: > > > > > > We can try going this route, what do you think? > > > > I still don't understand why you need a dentry to get the filehandle. > > The current api may ask for one, but it shouldn't really be necessary > > (assuming you don't want parent directory information encoded in the > > filehandle, which I hope you don't). > > As far as I know we don't need parent encoded. So Bruce, you think > to modify exportfs instead to work with inode directly? Looks like the filesystem encode_fh method just takes inodes (with the parent inode allowed to be NULL), so all you'd need would be a version of exportfs_encode_fh that took an inode. (Worst case, if that didn't work, you could fake up a dentry with something like d_obtain_alias, but better not to if it's not necessary.) --b.