From: Ernesto =?utf-8?Q?A=2E_Fern=C3=A1ndez?=
        <ernesto.mnd.fernandez@gmail.com>
Subject: Re: [xfstests PATCH] generic: add test of file mode when setfacl
 fails
Date: Fri, 14 Jul 2017 02:40:12 -0300
Message-ID: <20170714054011.GA4973@debian.home>
References: <cover.1499805484.git.ernesto.mnd.fernandez@gmail.com>
 <20170712114719.GA14398@infradead.org>
 <20170713094003.GA4959@debian.home>
 <20170713115507.GA10435@infradead.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 8bit
Cc: Ernesto =?utf-8?Q?A=2E_Fern=C3=A1ndez?=
        <ernesto.mnd.fernandez@gmail.com>, Jan Kara <jack@suse.com>,
        Theodore Ts'o <tytso@mit.edu>,
        Andreas Dilger <adilger.kernel@dilger.ca>,
        Dave Kleikamp <shaggy@kernel.org>, linux-ext4@vger.kernel.org,
        jfs-discussion@lists.sourceforge.net,
        reiserfs-devel@vger.kernel.org, fstests@vger.kernel.org
To: Christoph Hellwig <hch@infradead.org>
Return-path: <fstests-owner@vger.kernel.org>
Content-Disposition: inline
In-Reply-To: <20170713115507.GA10435@infradead.org>
Sender: fstests-owner@vger.kernel.org
List-Id: linux-ext4.vger.kernel.org

On Thu, Jul 13, 2017 at 04:55:07AM -0700, Christoph Hellwig wrote:
> On Thu, Jul 13, 2017 at 06:40:04AM -0300, Ernesto A. Fernández wrote:
> > Check that the group permission bits of a file are not altered when setfacl
> > fails. At the time of this patch the test fails for at least ext2, ext4 and
> > jfs. It is not run against reiserfs, since xfstests claims that "attrs are
> > not supported by this filesystem type".
> 
> That's odd, as common/config explicitly turns on user xattrs:
> 

xfstests decides if a filesystem supports attributes by trying to set them on
the root inode, but reiserfs does not allow this. If this is actually the
intended behaviour of reiserfs then xfstests should use a different check, but
I'm guessing it's a bug, perhaps in mkfs.reiserfs.

> 
> > 
> > The failure to set acls is induced by filling the device, so to save time
> > this should probably be run with a small TEST_DEV.
> 
> If you run it on the scratch fs instead of the test fs you can use
> _scratch_mkfs_sized to hand craft a small fs, which would be nice
> to get a decent runtime.

That's much better, thank you. I will send a revision soon. For some reason
_scratch_mkfs_sized did not have support for jfs, but it was trivial to add.