From: Jeff Mahoney <jeffm@suse.com>
To: linux-btrfs <linux-btrfs@vger.kernel.org>, fstests@vger.kernel.org
Cc: Qu Wenruo <quwenruo@cn.fujitsu.com>
Subject: [PATCH 1/2] btrfs/131: test for umount of read-only fs when quota rescan is paused
Date: Tue, 16 Aug 2016 14:30:21 -0400 [thread overview]
Message-ID: <2f102cc7-a6b2-e23d-9c56-de6c6bc5110d@suse.com> (raw)
Ensure that we can unmount a read-only file system when quota rescan
is paused from a previous read-write mount.
If the kernel has a separate bug where we are returning early while
waiting for the rescan worker, we can use that to un-hang the test,
and report both errors.
This issue is resolved by the following patch for the Linux kernel:
"btrfs: properly track when rescan worker is running"
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
---
tests/btrfs/131 | 100 ++++++++++++++++++++++++++++++++++++++++++++++++++++
tests/btrfs/131.out | 3 ++
tests/btrfs/group | 1 +
3 files changed, 104 insertions(+)
create mode 100755 tests/btrfs/131
create mode 100644 tests/btrfs/131.out
diff --git a/tests/btrfs/131 b/tests/btrfs/131
new file mode 100755
index 0000000..56c38a2
--- /dev/null
+++ b/tests/btrfs/131
@@ -0,0 +1,100 @@
+#! /bin/bash
+# FS QA Test 131
+#
+# Test for bug where read-only mounts will hang on umount when
+# a qgroup rescan was paused. This also tests whether that hung
+# umount can be unhung by trying to make use of a separate bug that
+# means we can interrupt the wait for the rescan worker. If that
+# happens, we report both errors.
+#
+#-----------------------------------------------------------------------
+# Copyright (c) 2016 SUSE. All Rights Reserved.
+#
+# This program is free software; you can redistribute it and/or
+# modify it under the terms of the GNU General Public License as
+# published by the Free Software Foundation.
+#
+# This program is distributed in the hope that it would be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write the Free Software Foundation,
+# Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
+#-----------------------------------------------------------------------
+#
+
+seq=`basename $0`
+seqres=$RESULT_DIR/$seq
+echo "QA output created by $seq"
+
+here=`pwd`
+tmp=/tmp/$$
+status=1 # failure is the default!
+trap "_cleanup; exit \$status" 0 1 2 3 15
+
+_cleanup()
+{
+ cd /
+ rm -f $tmp.*
+}
+
+# get standard environment, filters and checks
+. ./common/rc
+. ./common/filter
+
+# remove previous $seqres.full before test
+rm -f $seqres.full
+
+# real QA test starts here
+
+# Modify as appropriate.
+_supported_fs btrfs
+_supported_os Linux
+
+# We'll exit with a quota rescan paused
+_require_scratch_nocheck
+
+_require_btrfs
+BTRFS_DEBUG_TREE_PROG="`set_prog_path btrfs-debug-tree`"
+_require_command "$BTRFS_DEBUG_TREE_PROG" btrfs-debug-tree
+TIMEOUT_PROG="`set_prog_path timeout`"
+_require_command "$TIMEOUT_PROG" timeout
+
+rm -f $seqres.full
+_scratch_mkfs >>$seqres.full 2>&1
+
+_scratch_mount
+cp -aR /lib/modules/$(uname -r) $SCRATCH_MNT
+
+# A qgroup rescan on an empty or small file system completes nearly
+# immediately. We need to ensure that it runs long enough that it will
+# be paused on umount. Snapshots slow down the rescan so we should see
+# the race without a lot of data. This is an arbitrary number that
+# works on a ramdisk so it should be sufficient for any storage.
+for n in $(seq 1 100); do
+ _run_btrfs_util_prog subvolume snapshot $SCRATCH_MNT $SCRATCH_MNT/$n
+done
+_run_btrfs_util_prog quota enable $SCRATCH_MNT
+_scratch_unmount
+
+echo "read-write umount completed"
+
+# Confirm that the rescan is paused
+if ! $BTRFS_DEBUG_TREE_PROG $SCRATCH_DEV | \
+ egrep -q 'flags ON|SCANNING|INCONSISTENT scan'; then
+ echo "qgroup rescan not paused."
+fi
+_scratch_mount -r
+
+# If the bug exists, this will hang. If we can kill it, that's another bug.
+$TIMEOUT_PROG 10 umount $SCRATCH_MNT
+if test $? -eq 124 ; then
+ echo "umount hung but was killed"
+fi
+echo "read-only umount completed"
+
+# success, all done
+status=0
+exit
diff --git a/tests/btrfs/131.out b/tests/btrfs/131.out
new file mode 100644
index 0000000..845a501
--- /dev/null
+++ b/tests/btrfs/131.out
@@ -0,0 +1,3 @@
+QA output created by 131
+read-write umount completed
+read-only umount completed
diff --git a/tests/btrfs/group b/tests/btrfs/group
index 6b29c05..929fa21 100644
--- a/tests/btrfs/group
+++ b/tests/btrfs/group
@@ -133,3 +133,4 @@
128 auto quick send
129 auto quick send
130 auto clone send
+131 auto quick qgroup
--
1.8.5.6
--
Jeff Mahoney
SUSE Labs
next reply other threads:[~2016-08-16 18:43 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-08-16 18:30 Jeff Mahoney [this message]
2016-08-17 8:45 ` [PATCH 1/2] btrfs/131: test for umount of read-only fs when quota rescan is paused Eryu Guan
2016-08-17 8:59 ` Qu Wenruo
2016-08-19 19:20 ` Jeff Mahoney
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=2f102cc7-a6b2-e23d-9c56-de6c6bc5110d@suse.com \
--to=jeffm@suse.com \
--cc=fstests@vger.kernel.org \
--cc=linux-btrfs@vger.kernel.org \
--cc=quwenruo@cn.fujitsu.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;
as well as URLs for NNTP newsgroup(s).