[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-Id: <1535394566-18837-1-git-send-email-amir73il@gmail.com>
Date: Mon, 27 Aug 2018 21:29:24 +0300
From: Amir Goldstein <amir73il@...il.com>
To: Eryu Guan <guaneryu@...il.com>
Cc: Theodore Ts'o <tytso@....edu>, fstests@...r.kernel.org,
linux-ext4@...r.kernel.org, linux-unionfs@...r.kernel.org
Subject: [PATCH 0/2] fstests: fix overlay test failure over ext4
Eryu,
As I reported earlier, test generic/009 was failing ./check -overlay
run due to a test bug. The test belongs to a class of tests that have
explicit check for a certain $FSTYP, which does not apply when FSTYP is
overlay.
I ran all the tests that check for FSTYP = ext4 with overlay over ext4
and all the tests that check for FSTYP = xfs with overlay over xfs.
generic/009 was the only test that failed because of the exlicit check.
The rest of the tests either passes or did not run for a good reason.
Strangely, tests which _require_seek_data_hole(), that has the exact same
check as generic/009 didn't fail with overlay over ext4. In fact, those
tests also did not with with bare ext4 even when I removed the call
to _ext4_disable_extent_zeroout() from _require_seek_data_hole()???
I tried to come up with a more generic solution, but didn't like any
ideas I had, so I ended up fixing generic/009 in place in the hope that
you either accept it as is or propose a better solution.
The fix depends on the user either setting FSTYP=ext4 in a config file
or setting the variable OVL_BASE_FSTYP before running the test.
I will soon send a patch for kvm-xfstests to Ted that does the latter.
Thanks,
Amir.
Amir Goldstein (2):
overlay: do not set OVL_BASE_FSTYP value to "overlay"
generic/009: fix test failure with overlay over ext4
common/config | 2 +-
tests/generic/009 | 3 +++
2 files changed, 4 insertions(+), 1 deletion(-)
--
2.7.4
Powered by blists - more mailing lists