Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S933534AbcK2NG3 (ORCPT ); Tue, 29 Nov 2016 08:06:29 -0500 Received: from mail-wm0-f43.google.com ([74.125.82.43]:34972 "EHLO mail-wm0-f43.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S933366AbcK2NGU (ORCPT ); Tue, 29 Nov 2016 08:06:20 -0500 MIME-Version: 1.0 In-Reply-To: References: <20161125212934.GB2622@veci.piliscsaba.szeredi.hu> From: Amir Goldstein Date: Tue, 29 Nov 2016 14:03:06 +0200 Message-ID: Subject: Re: [POC/RFC PATCH] overlayfs: constant inode numbers To: Miklos Szeredi Cc: "linux-unionfs@vger.kernel.org" , linux-fsdevel , linux-kernel Content-Type: text/plain; charset=UTF-8 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 2449 Lines: 54 On Tue, Nov 29, 2016 at 1:34 PM, Amir Goldstein wrote: > On Tue, Nov 29, 2016 at 12:16 PM, Miklos Szeredi wrote: >> On Mon, Nov 28, 2016 at 12:56 PM, Amir Goldstein wrote: >> >> [...] >> >> >>> Well, not sure if you noticed my redirect_fh (rediect by file handle) work. >>> If differs from redirect by path in 2 major ways: >>> 1. Like OVL_XATTR_INO, redirect is set on copy up (but only for dirs) >>> 2. Lookup is much simpler (and most likely faster) then full path lookup >>> >>> It would be trivial to set oe->ino of merged dir from lower most entry in >>> ovl_lookup(). >> >> Okay. In fact always using the handle looks like a better option >> overall. File handle should be unique for the lifetime of the >> filesystem, while inode numbers may be reused. >> >> Biggest drawback of the file handle based redirects is exactly that: >> makes backing up the overlay basically impossible, since file handles >> won't work after a backup + restore. But as an optimization, in >> addition to path based redirects it would work fine and provide a good >> way to get the stable ino. >> > > Not sure that I understand what you are suggesting, but I would be happy > to make the needed adjustments to redirect_fh per your request if you clarify > what you mean. From what I understand: > > 1. If redirect_dir=fh (and supported by layers), store lower handle > on dir copy up in new xattr OVL_XATTR_FH > 2. In ovl_rename(), set OVL_XATTR_REDIRECT regardless of OVL_XATTR_FH > 3. In ovl_lookup_single(), carry both d.redirct and d.redirect_fh to next layer > 4. In ovl_lookup_layer(), lookup by handle first then by path > > Mind you that unlike OVL_XATTR_INO, OVL_XATTR_FH points to *next* inode > rather then *lowest* inode, so it does not really improve anything wrt > getting the > stable inode. Stable inode of merged dir is available after lookup in > oe->lowerstack[oe->numlower - 1].dentry->d_inode->i_ino > regardless of whether lookup was by handle or by path or no redirect at all, so > not sure what you meant by "... and provide a good way to get the stable ino." > Either I managed to confuse you, or I am missing something? > Perhaps you meant for non-dir: 5. If redirect_dir=fh, *propagate* lowest-handle on non-dir copy up 6. In ovl_lookup() of non-dir, decode lowest-handle to set oe->ino This is a good way to avoid stale OVL_XATTR_INO after backup+restore. Amir.