From: Eryu Guan <eguan@redhat.com>
To: Amir Goldstein <amir73il@gmail.com>
Cc: Miklos Szeredi <miklos@szeredi.hu>, Xiong Zhou <xzhou@redhat.com>,
linux-unionfs@vger.kernel.org, fstests@vger.kernel.org
Subject: Re: [PATCH 4/5] overlay/017: verify constant inode number after rename
Date: Fri, 28 Apr 2017 13:47:20 +0800 [thread overview]
Message-ID: <20170428054720.GM26397@eguan.usersys.redhat.com> (raw)
In-Reply-To: <1493305775-12559-5-git-send-email-amir73il@gmail.com>
On Thu, Apr 27, 2017 at 06:09:34PM +0300, Amir Goldstein wrote:
> The test verifies constant inode number after copy up.
> Verify that inode number remains constant also after rename
> and drop caches (when overlayfs needs to find the lower
> inodes in another location).
>
> Signed-off-by: Amir Goldstein <amir73il@gmail.com>
> ---
> tests/overlay/017 | 22 ++++++++++++++++++----
> 1 file changed, 18 insertions(+), 4 deletions(-)
>
> diff --git a/tests/overlay/017 b/tests/overlay/017
> index f3bf454..1cf684d 100755
> --- a/tests/overlay/017
> +++ b/tests/overlay/017
> @@ -8,7 +8,8 @@
> # - modify A to trigger copy up
> # - stat file A shows inode number Y != X
> #
> -# Also test if d_ino of readdir entries changes after copy up.
> +# Also test if d_ino of readdir entries changes after copy up
> +# and if inode numbers persist after rename and drop caches.
> #
> #-----------------------------------------------------------------------
> #
> @@ -94,6 +95,8 @@ _scratch_mount
>
>
> rm -f $tmp.*
> +testdir=$SCRATCH_MNT/test
> +mkdir -p $testdir
>
> # Test stable stat(2) st_ino
>
> @@ -106,18 +109,29 @@ for f in $FILES; do
> done
>
> # Record inode numbers after copy up
> -record_inode_numbers $SCRATCH_MNT $tmp.after
> +record_inode_numbers $SCRATCH_MNT $tmp.after_copyup
inode numbers are saved after copyup and will be tested by diff with
$tmp.before, but find (d_ino) is not tested after a pure copyup...
> +
> +for f in $FILES; do
> + # move to another dir
> + mv $SCRATCH_MNT/$f $testdir/
> +done
> +
> +echo 3 > /proc/sys/vm/drop_caches
> +
> +# Record inode numbers after rename and drop caches
> +record_inode_numbers $testdir $tmp.after_move
>
> # Test stable readdir(3)/getdents(2) d_ino
>
> # find by inode number - expect to find file by inode number
> cat $tmp.before | while read ino f; do
> - find $SCRATCH_MNT/ -inum $ino -maxdepth 1 | grep -q $f || \
> + find $testdir/ -inum $ino -maxdepth 1 | grep -q $f || \
> echo "$f not found by ino $ino"
> done
it's only tested here, after a rename and a drop_caches.
IMO, it's better to test both find and stat after each operation that
would cause inode number change. So how about factoring out a helper
that does this inode number check (find and diff), and calling it after
each operation?
e.g. a new helper called check_inode_number(), then
<create files>
<copyup>
check_inode_number [with necessary args]
<rename>
check_inode_number
<drop_caches>
check_inode_number
<cycle mount>
check_inode_number
Thanks,
Eryu
>
> # Compare before..after - expect silence
> -diff $tmp.before $tmp.after
> +diff -u $tmp.before $tmp.after_copyup
> +diff -u $tmp.after_copyup $tmp.after_move
>
> echo "Silence is golden"
> status=0
> --
> 2.7.4
>
next prev parent reply other threads:[~2017-04-28 5:47 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-04-27 15:09 [PATCH 0/5] fstests: more tests for overlay constant inode numbers Amir Goldstein
2017-04-27 15:09 ` [PATCH 1/5] overlay/017: silence test output Amir Goldstein
2017-04-28 5:36 ` Eryu Guan
2017-04-28 5:45 ` Amir Goldstein
2017-04-28 5:50 ` Eryu Guan
2017-04-28 6:34 ` Amir Goldstein
2017-04-28 6:50 ` Eryu Guan
2017-04-27 15:09 ` [PATCH 2/5] overlay/017: use af_unix to create socket test file Amir Goldstein
2017-04-27 15:09 ` [PATCH 3/5] overlay/017: create a helper to record inode number Amir Goldstein
2017-04-27 15:09 ` [PATCH 4/5] overlay/017: verify constant inode number after rename Amir Goldstein
2017-04-28 5:47 ` Eryu Guan [this message]
2017-04-28 5:50 ` Amir Goldstein
2017-04-27 15:09 ` [PATCH 5/5] overlay/017: test persistent inode numbers after mount cycle Amir Goldstein
2017-04-28 5:30 ` [PATCH 0/5] fstests: more tests for overlay constant inode numbers Eryu Guan
2017-04-28 5:39 ` Amir Goldstein
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20170428054720.GM26397@eguan.usersys.redhat.com \
--to=eguan@redhat.com \
--cc=amir73il@gmail.com \
--cc=fstests@vger.kernel.org \
--cc=linux-unionfs@vger.kernel.org \
--cc=miklos@szeredi.hu \
--cc=xzhou@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox