All of lore.kernel.org
 help / color / mirror / Atom feed
From: Zorro Lang <zlang@redhat.com>
To: fstests@vger.kernel.org
Cc: linux-xfs@vger.kernel.org
Subject: [PATCH v2] fstests/xfs: xfs_repair should junk empty attribute leaf blocks
Date: Wed, 12 Apr 2017 23:39:37 +0800	[thread overview]
Message-ID: <1492011577-13436-1-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,

V2 did below changes:
1) remove the requirement for xfs_db's write and addr commands
2) check if xfs_db support write -d option
3) turn to get filesystem block size from _filter_mkfs output
4) add more comments

Thanks,
Zorro

 tests/xfs/288     | 110 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
 tests/xfs/288.out |   2 +
 tests/xfs/group   |   1 +
 3 files changed, 113 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..81e7b50
--- /dev/null
+++ b/tests/xfs/288
@@ -0,0 +1,110 @@
+#! /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 Red Hat Inc.  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 $seqres.full
+
+# V4 XFS attr type is 'attr'
+write_cmd="write"
+
+# Modify as appropriate.
+_supported_fs xfs
+_supported_os Linux
+_require_scratch
+_require_attrs
+
+_scratch_mkfs_xfs | _filter_mkfs 2>$tmp.mkfs >/dev/null
+. $tmp.mkfs
+
+# If test on XFS with CRC enabled, after change attr's hdr.count we need
+# to recalculate crc by use write -d option
+if [ $_fs_has_crcs -eq 1 ]; then
+	# check if xfs_db support write -d option
+	_scratch_xfs_db -x -c "sb" -c "write -d" 2>&1 | \
+		grep -q "invalid option"
+	if [ $? -eq 0 ];then
+		_notrun "xfs_db doesn't support write -d option"
+	fi
+	# 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 an 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.
+maxisize=$((dbsize/2))
+$SETFATTR_PROG -n "user.testattr${seq}" \
+	       -v `python -c "print ($maxisize * 'a')"` \
+	       $SCRATCH_MNT/$seq.attrfile
+
+_scratch_unmount
+# manually corrupt the XFS, by set the header count of attr to 0
+_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 and fix this XFS, then
+# SCRATCH_DEV shouldn't be corrupted after this case done.
+_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..f27a7b6 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 fuzzers
 290 auto rw prealloc quick ioctl zero
 291 auto repair
 292 auto mkfs quick
-- 
2.7.4


             reply	other threads:[~2017-04-12 15:39 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-04-12 15:39 Zorro Lang [this message]
2017-04-12 16:09 ` [PATCH v2] fstests/xfs: xfs_repair should junk empty attribute leaf blocks Eric Sandeen
2017-04-12 16:24   ` Zorro Lang
2017-04-12 16:14 ` Darrick J. Wong
2017-04-12 16:20   ` Eric Sandeen
2017-04-12 16:27   ` Zorro Lang
2017-04-12 16:55     ` Darrick J. Wong
2017-04-12 17:13       ` Zorro Lang

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=1492011577-13436-1-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.