From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mx1.redhat.com ([209.132.183.28]:45132 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751542AbdKXFFo (ORCPT ); Fri, 24 Nov 2017 00:05:44 -0500 Date: Fri, 24 Nov 2017 13:05:41 +0800 From: Eryu Guan Subject: Re: [PATCH] overlay: add a test for multiple redirects to the same lower dir Message-ID: <20171124050541.GY2749@eguan.usersys.redhat.com> References: <1511447224-2301-1-git-send-email-amir73il@gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1511447224-2301-1-git-send-email-amir73il@gmail.com> Sender: fstests-owner@vger.kernel.org To: Amir Goldstein Cc: Miklos Szeredi , zhangyi , Chandan Rajendra , linux-unionfs@vger.kernel.org, fstests@vger.kernel.org List-ID: On Thu, Nov 23, 2017 at 04:27:04PM +0200, Amir Goldstein wrote: > Multiple redirects to the same lower dir will falsely return the > same st_ino/st_dev for two different upper dirs and will cause 'diff' > to falsely report that content of directories is the same when it is not. > > This is a test for a regression introduced in kernel v4.12 by > commit 72b608f08528 ("ovl: constant st_ino/st_dev across copy up"), > but also the first xfstest to require the redirect_dir feature that > was introduced as an opt-in feature in kernel v4.10. > > Signed-off-by: Amir Goldstein > --- > > Eryu, > > This test is for a "bug" that has not been acknoledges by Miklos > as a kernel bug yet. It may well fall within the jurisdiction of > fsck.overlayfs. > > IMO, cp -a of upper files and dirs qualifies to the statement in > Documentation/filesystems/overlayfs.txt: > "Offline changes, when the overlay is not mounted, are allowed to either > the upper or the lower trees." > > So unless Miklos objects to ever fixing this "bug"? > I suggest that we include the failing test until kernel is fixed. Thanks for the info! Then I'll wait until it's been acked as a bug. Thanks, Eryu