xfs
[Top] [All Lists]

Re: [PATCH 06/17] xfs: run xfs_repair at the end of each test

To: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Subject: Re: [PATCH 06/17] xfs: run xfs_repair at the end of each test
From: "Darrick J. Wong" <darrick.wong@xxxxxxxxxx>
Date: Mon, 1 Aug 2016 10:08:15 -0700
Cc: david@xxxxxxxxxxxxx, eguan@xxxxxxxxxx, fstests@xxxxxxxxxxxxxxx, linux-btrfs@xxxxxxxxxxxxxxx, xfs@xxxxxxxxxxx
Delivered-to: xfs@xxxxxxxxxxx
In-reply-to: <20160801062719.GH596@xxxxxxxxxxxxx>
References: <146914477514.11762.3144320628851923350.stgit@xxxxxxxxxxxxxxxx> <146914481474.11762.2741429828012981240.stgit@xxxxxxxxxxxxxxxx> <20160801062719.GH596@xxxxxxxxxxxxx>
User-agent: Mutt/1.5.24 (2015-08-30)
On Sun, Jul 31, 2016 at 11:27:19PM -0700, Christoph Hellwig wrote:
> On Thu, Jul 21, 2016 at 04:46:54PM -0700, Darrick J. Wong wrote:
> > Run xfs_repair twice at the end of each test -- once to rebuild
> > the btree indices, and again with -n to check the rebuild work.
> 
> This looks fine to me in general, but shouldn't we have specific
> tests that test the rebuilding in a normal auto run?

We do have specific tests that examine the outputs of rebuilding the
indices (all the fuzzer group tests do this too); this patch enables a
test runner to expand that coverage to all tests.  Running a
rebuilding xfs_repair for all the tests shook out some bugs in the
xfs_repair rmap handling code that only triggered under some of the
non-rmap non-reflink stressor tests.

--D

<Prev in Thread] Current Thread [Next in Thread>