From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Aneesh Kumar K. V" Subject: Re: [PATCH -V18 04/13] vfs: Allow handle based open on symlinks Date: Tue, 24 Aug 2010 16:04:32 +0530 Message-ID: References: <1282269097-26166-1-git-send-email-aneesh.kumar@linux.vnet.ibm.com> <1282269097-26166-5-git-send-email-aneesh.kumar@linux.vnet.ibm.com> <20100820083057.GA10039@infradead.org> <20100820195303.20b17210@notabene> <20100820115135.GQ31363@ZenIV.linux.org.uk> <20100821100900.4b15fe08@notabene> <20100821083024.GB3448@amd> <20100824072155.GA3948@amd> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Nick Piggin , Neil Brown , Al Viro , Christoph Hellwig , adilger@sun.com, corbet@lwn.net, hooanon05@yahoo.co.jp, bfields@fieldses.org, miklos@szeredi.hu, linux-fsdevel@vger.kernel.org, sfrench@us.ibm.com, philippe.deniel@CEA.FR, linux-kernel@vger.kernel.org To: Nick Piggin Return-path: In-Reply-To: <20100824072155.GA3948@amd> Sender: linux-kernel-owner@vger.kernel.org List-Id: linux-fsdevel.vger.kernel.org On Tue, 24 Aug 2010 17:21:55 +1000, Nick Piggin wrote: > On Sat, Aug 21, 2010 at 03:12:15PM +0530, Aneesh Kumar K. V wrote: > > On Sat, 21 Aug 2010 18:30:24 +1000, Nick Piggin wrote: > > > Thanks, I had both of the same concerns as Christoph with API > > > change and exposing symlink fds last time I looked at the patces, > > > actually. > > > > > > But they can probably be worked around or avoided. I think the more > > > important thing is whether it is worth supporting. This is > > > all restricted to root (or CAP_DAC_READ_SEARCH) only, right, and > > > what exact semantics they want. I would like to see more discussion > > > of what this enables and some results. > > > > > > For the case of avoiding expensive network revalidations in path name > > > lookup, do we even need to open symlinks? Could the security issues be > > > avoided by always having handle attached to an open fd? > > > > > > > For implementing a userspace file server that use handle for > > representing files (like NFS) we would require to have the ability to do > > different file system operations that can operate on symlink to work on > > handle too. > > Right. Is this a really important goal, I'm wondering? Is it realistic > (ie. to be able to remove the nfs server from the kernel)? > The feature is also needed to implement a 9p virtio pass through file system in Qemu http://thread.gmane.org/gmane.comp.emulators.qemu/68992 -aneesh