xfs
[Top] [All Lists]

Re: [PATCH v2] xfs: log record crc mismatch test

To: Eric Sandeen <sandeen@xxxxxxxxxxx>
Subject: Re: [PATCH v2] xfs: log record crc mismatch test
From: Alexander Tsvetkov <alexander.tsvetkov@xxxxxxxxxx>
Date: Wed, 13 May 2015 17:52:31 +0300
Cc: Dave Chinner <david@xxxxxxxxxxxxx>, fstests@xxxxxxxxxxxxxxx, xfs@xxxxxxxxxxx
Delivered-to: xfs@xxxxxxxxxxx
In-reply-to: <553E6723.8090909@xxxxxxxxxxx>
References: <553E5505.4070804@xxxxxxxxxx> <553E6723.8090909@xxxxxxxxxxx>
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:31.0) Gecko/20100101 Thunderbird/31.6.0
Hi Eric,

On 04/27/2015 07:43 PM, Eric Sandeen wrote:
On 4/27/15 10:25 AM, Alexander Tsvetkov wrote:
Added new test case for simulation of V5 filesystem
log crc record corruption using of external log.

Signed-off-by: Alexander Tsvetkov <alexander.tsvetkov@xxxxxxxxxx>
---
  tests/xfs/005     | 28 ++++++++++++++++++++++++++--
  tests/xfs/005.out |  3 +++
  2 files changed, 29 insertions(+), 2 deletions(-)

diff --git a/tests/xfs/005 b/tests/xfs/005
index fade4bb..ba0dcbb 100755
--- a/tests/xfs/005
+++ b/tests/xfs/005
@@ -1,7 +1,7 @@
  #! /bin/bash
  # FS QA Test No. 005
  #
-# Test that a bad crc on a primary V5 superblock will fail the mount
+# Test that a bad crc on a V5 filesystem will fail the mount
bad crc on what?

  #
  # 10e6e65 xfs: be more forgiving of a v4 secondary sb w/ junk in v5 fields
  # inadvertently caused primary SB CRC failures to not error out, this
@@ -55,10 +55,12 @@ _supported_os Linux

  _require_scratch_nocheck
  _require_xfs_mkfs_crc
+_require_logdev
+logsize=2560
No, this means that xfs/05 will no longer work without an external logdev,
right?  So you've stopped testing the internal case, and we've lost that
coverage.

And, we generally don't want to add new testcases to old tests because then
it gets harder to investigate regressions.  It'd probably be best to
create a new test which specifically tests log crc corruption, and make
it handle both internal and external logs automatically.
I've prepared the test with both cases, see at the bottom of this mail


Thanks,
-Eric

  rm -f $seqres.full

-_scratch_mkfs_xfs -m crc=1 >> $seqres.full 2>&1 || _fail "mkfs failed"
+_scratch_mkfs_xfs -m crc=1 -lsize=${logsize}b >> $seqres.full 2>&1 || _fail "mkfs 
failed"

  # Zap the crc.  xfs_db updates the CRC post-write, so poke it directly
  $XFS_IO_PROG -c "pwrite 224 4" -c fsync $SCRATCH_DEV | _filter_xfs_io
@@ -66,6 +68,28 @@ $XFS_IO_PROG -c "pwrite 224 4" -c fsync $SCRATCH_DEV | 
_filter_xfs_io
  # should FAIL, the crc is bad; golden output contains mount failure
  _scratch_mount 2>&1 | filter_mount

+_scratch_unmount &> /dev/null
+_scratch_mkfs_xfs -m crc=1 -lsize=${logsize}b  >> $seqres.full 2>&1 || _fail "mkfs 
failed"
+_scratch_mount >> $seqres.full 2>&1 || _fail "mount failed"
+
+# Run filesytem exercising for some time
+_param="-f link=10 -f creat=10 -f mkdir=10 -f truncate=5 -f symlink=10"
+$FSSTRESS_PROG $_param -s 1 $FSSTRESS_AVOID -n 1000 -d $SCRATCH_MNT &> 
/dev/null
+sleep 3
+
+# Shutdown filesystem and unmount
+src/godown -v $SCRATCH_MNT >> $seqres.full
+killall -q fsstress
+wait
+sleep 3
+_scratch_unmount
+
+# Zap log record crc.
+$XFS_IO_PROG -c "pwrite 5000 5" $SCRATCH_LOGDEV | _filter_xfs_io
+
+# should FAIL, log record crc is bad; golden output contains mount failure
+_scratch_mount 2>&1 | filter_mount
+
  # success, all done
  status=0
  exit
diff --git a/tests/xfs/005.out b/tests/xfs/005.out
index 13d0125..8907301 100644
--- a/tests/xfs/005.out
+++ b/tests/xfs/005.out
@@ -2,3 +2,6 @@ QA output created by 005
  wrote 4/4 bytes at offset 224
  XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
  mount: Structure needs cleaning
+wrote 5/5 bytes at offset 5000
+XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
+mount: Structure needs cleaning
Added new test for simulation of V5 filesystem
log crc record corruption in internal and external logs

Signed-off-by: Alexander Tsvetkov <alexander.tsvetkov@xxxxxxxxxx>
---
tests/xfs/306 | 110 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
 tests/xfs/306.out |   7 ++++
 tests/xfs/group   |   1 +
 3 files changed, 118 insertions(+)
 create mode 100755 tests/xfs/306
 create mode 100644 tests/xfs/306.out

diff --git a/tests/xfs/306 b/tests/xfs/306
new file mode 100755
index 0000000..6e025fe
--- /dev/null
+++ b/tests/xfs/306
@@ -0,0 +1,110 @@
+#! /bin/bash
+# FS QA Test No. 306
+#
+# Test that bad crc on a V5 log record will fail the mount
+#
+#-----------------------------------------------------------------------
+# Copyright (c) 2015 Oracle and/or its affiliates.  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.*
+}
+
+filter_mount()
+{
+    sed -e "s/mount .* failed: //"
+}
+
+filter_io()
+{
+    sed -e "s/at offset .*/at offset XXX/"
+}
+
+# $1 - log start offset
+# $2 - log device
+do_test()
+{
+    off=$1
+    dev=$2
+
+    _scratch_unmount >/dev/null 2>&1
+    _scratch_mount >> $seqres.full 2>&1 || _fail "mount failed"
+
+    # Run filesytem exercising for some time
+    param="-f link=10 -f creat=10 -f mkdir=10 -f truncate=5 -f symlink=10"
+ $FSSTRESS_PROG $param -s 1 $FSSTRESS_AVOID -n 1000 -d $SCRATCH_MNT &> /dev/null
+    sleep 3
+
+    # Shutdown filesystem and unmount
+    src/godown -v $SCRATCH_MNT >> $seqres.full
+    killall -q fsstress
+    wait
+    sleep 3
+    _scratch_unmount
+
+    # Zap log record crc
+    off=$((off + 5000))
+    $XFS_IO_PROG -c "pwrite $off 5" $dev | _filter_xfs_io | filter_io
+
+ # should FAIL, log record crc is bad; golden output contains mount failure
+    _scratch_mount 2>&1 | filter_mount
+}
+
+# get standard environment, filters and checks
+. ./common/rc
+. ./common/filter
+
+_supported_fs xfs
+_supported_os Linux
+
+_require_scratch_nocheck
+_require_xfs_mkfs_crc
+_require_logdev
+
+rm -f $seqres.full
+fssize=$((32 * 1024 * 1024))
+MKFS_OPTIONS="-m crc=1 $MKFS_OPTIONS"
+_scratch_mkfs_sized $fssize 2> /dev/null | _filter_mkfs 2> $tmp.mkfs > /dev/null
+. $tmp.mkfs
+
+# internal log corruption
+unset USE_EXTERNAL
+_scratch_mkfs_sized $fssize >> $seqres.full 2>&1 || _fail "mkfs failed"
+logstart=$($XFS_DB_PROG -c "sb 0" -c "print" $SCRATCH_DEV | grep logstart | cut -f3 -d' ')
+offset=$((logstart * lbsize))
+do_test $offset $SCRATCH_DEV
+
+# external log corruption
+USE_EXTERNAL="yes"
+_scratch_mkfs_sized $fssize >> $seqres.full 2>&1 || _fail "mkfs failed"
+do_test 0 $SCRATCH_LOGDEV
+
+# success, all done
+status=0
+exit
diff --git a/tests/xfs/306.out b/tests/xfs/306.out
new file mode 100644
index 0000000..aef98f2
--- /dev/null
+++ b/tests/xfs/306.out
@@ -0,0 +1,7 @@
+QA output created by 306
+wrote 5/5 bytes at offset XXX
+XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
+mount: Structure needs cleaning
+wrote 5/5 bytes at offset XXX
+XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
+mount: Structure needs cleaning
diff --git a/tests/xfs/group b/tests/xfs/group
index 58144d2..f108377 100644
--- a/tests/xfs/group
+++ b/tests/xfs/group
@@ -191,3 +191,4 @@
 303 auto quick quota
 304 auto quick quota
 305 auto quota
+306 auto quick
--
1.9.3

<Prev in Thread] Current Thread [Next in Thread>
  • Re: [PATCH v2] xfs: log record crc mismatch test, Alexander Tsvetkov <=