From: Zorro Lang <zlang@redhat.com>
To: fstests@vger.kernel.org
Cc: linux-xfs@vger.kernel.org
Subject: [PATCH] fstests/xfs: xfs_repair should junk empty attribute leaf blocks
Date: Wed, 12 Apr 2017 18:41:33 +0800 [thread overview]
Message-ID: <1491993693-9667-1-git-send-email-zlang@redhat.com> (raw)
There was a bug during log replay, the attr/attr3 leaf verifier
reported corruption when encountering a leaf attribute with a
count of 0 in the header, as below:
Metadata corruption detected at xfs_attr3_leaf block 0x480988/0x1000
commit f714016 from xfsprogs has fixed this bug. This test case
will emulate this corruption by xfs_db and use xfs_repair to fix
it.
Signed-off-by: Zorro Lang <zlang@redhat.com>
---
Hi,
xfs_db already support 'write -d' from last year, even though the
manpage doesn't record that. But the '-d' option can't be run for
everything.
I just saw Eric and Darrick send two patches to allow write -d to
dqblks and inodes. But they haven't been merged. In this case, I
hope to change a field of attr3 on V5 XFS, so I have to find a way
to make sure if the current xfs_db can do that. I'm not the one
who learn about xfs_db so much, so if someone feel something wrong,
please feel free to tell me.
I know I can limit this case test on V4 XFS, then there's no 'write -d'
problem. But I hope to cover V5 XFS test if I can.
Thanks,
Zorro
tests/xfs/288 | 103 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
tests/xfs/288.out | 2 ++
tests/xfs/group | 1 +
3 files changed, 106 insertions(+)
create mode 100755 tests/xfs/288
create mode 100644 tests/xfs/288.out
diff --git a/tests/xfs/288 b/tests/xfs/288
new file mode 100755
index 0000000..3dc0a75
--- /dev/null
+++ b/tests/xfs/288
@@ -0,0 +1,103 @@
+#! /bin/bash
+# FS QA Test 288
+#
+# When an attribute leaf block count is 0, xfs_repair should junk
+# that leaf directly (as xfsprogs commit f714016).
+#
+#-----------------------------------------------------------------------
+# Copyright (c) 2017 YOUR NAME HERE. 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
+. ./common/attr
+
+# remove previous $seqres.full before test
+rm -f $seqres.full
+
+# V4 XFS attr type is 'attr'
+write_cmd="write"
+
+# Modify as appropriate.
+_supported_fs xfs
+_supported_os Linux
+_require_scratch
+_require_attrs
+_require_xfs_db_command write
+_require_xfs_db_command addr
+
+_scratch_mkfs_xfs | _filter_mkfs 2>$tmp.mkfs >/dev/null
+. $tmp.mkfs
+
+# If test on XFS with CRC enabled, after change attr's hdr.count we need
+# to recalculate crc.
+if [ $_fs_has_crcs -eq 1 ]; then
+ # This's a trick to check if xfs_db support recalculating
+ # CRCs of attr3
+ _scratch_xfs_db -x -c "daddr 0" \
+ -c "type attr3" \
+ -c "write -d" 2>/dev/null | \
+ grep -q "Cannot recalculate CRCs on this type of object"
+ if [ $? -eq 0 ];then
+ _notrun "xfs_db can't recalculate CRCs of attr3"
+ fi
+ # V5 XFS attr type is 'attr3'
+ write_cmd="write -d"
+fi
+
+_scratch_mount
+
+touch $SCRATCH_MNT/$seq.attrfile
+inum=$(stat -c '%i' $SCRATCH_MNT/$seq.attrfile)
+
+# To get a attr block leaf, we need to extend attr format to extent
+# format at least, and the max inode size is half of filesystem
+# block size, so write half of block size attr to make sure attr
+# out of local format.
+fsblocksize=`xfs_info $SCRATCH_MNT|sed 's/=/ /g'|awk '/^data / { print $3 }'`
+maxisize=$((fsblocksize/2))
+setfattr -n "user.testattr${seq}" -v `python -c "print ($maxisize * 'a')"` \
+ $SCRATCH_MNT/$seq.attrfile
+umount $SCRATCH_MNT
+_scratch_xfs_db -x -c "inode $inum" \
+ -c "addr a.bmx[0].startblock" \
+ -c "$write_cmd hdr.count 0" >> $seqres.full
+
+# This repair should junk above leaf attribute to fix this XFS
+_repair_scratch_fs >> $seqres.full
+
+echo "Silence is golden"
+
+# success, all done
+status=0
+exit
diff --git a/tests/xfs/288.out b/tests/xfs/288.out
new file mode 100644
index 0000000..2958a5c
--- /dev/null
+++ b/tests/xfs/288.out
@@ -0,0 +1,2 @@
+QA output created by 288
+Silence is golden
diff --git a/tests/xfs/group b/tests/xfs/group
index 75769f9..a0a03c0 100644
--- a/tests/xfs/group
+++ b/tests/xfs/group
@@ -285,6 +285,7 @@
285 dangerous_fuzzers dangerous_scrub
286 dangerous_fuzzers dangerous_scrub dangerous_online_repair
287 auto dump quota quick
+288 auto quick repair
290 auto rw prealloc quick ioctl zero
291 auto repair
292 auto mkfs quick
--
2.7.4
reply other threads:[~2017-04-12 10:41 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
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=1491993693-9667-1-git-send-email-zlang@redhat.com \
--to=zlang@redhat.com \
--cc=fstests@vger.kernel.org \
--cc=linux-xfs@vger.kernel.org \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.