From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from ipmail06.adl2.internode.on.net ([150.101.137.129]:57678 "EHLO ipmail06.adl2.internode.on.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1725267AbfCFFHI (ORCPT ); Wed, 6 Mar 2019 00:07:08 -0500 Date: Wed, 6 Mar 2019 16:07:05 +1100 From: Dave Chinner Subject: Re: [PATCH] Documenting the crash-recovery guarantees of Linux file systems Message-ID: <20190306050705.GB23020@dastard> References: <1551841140-3708-1-git-send-email-jaya@cs.utexas.edu> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1551841140-3708-1-git-send-email-jaya@cs.utexas.edu> Sender: fstests-owner@vger.kernel.org To: Jayashree Cc: fstests@vger.kernel.org, linux-fsdevel@vger.kernel.org, linux-doc@vger.kernel.org, vijay@cs.utexas.edu, amir73il@gmail.com, tytso@mit.edu, fdmanana@gmail.com, chao@kernel.org List-ID: On Tue, Mar 05, 2019 at 08:59:00PM -0600, Jayashree wrote: > In this file, we document the crash-recovery guarantees > provided by four Linux file systems - xfs, ext4, F2FS and btrfs. We also > present Dave Chinner's proposal of Strictly-Ordered Metadata Consistency > (SOMC), which is provided by xfs. It is not clear to us if other file systems > provide SOMC; we would be happy to modify the document if file-system > developers claim that their system provides (or aims to provide) SOMC. I haven't had time to read this yet, but I will point out that fstests assumes that filesystems that run "_require_metadata_journaling" tests present SOMC semantics to users. i.e. we are not testing POSIX semantics in fstests, we are testing for SOMC crash-recovery semantics because POSIX does not define a useful set of semantics. That was the whole point of genericising that test infrastructure - to make sure all the linux journalling filesystems behave the same way in the same circumstances... Cheers, Dave. -- Dave Chinner david@fromorbit.com