On Thu, Jun 18, 2015 at 4:48 AM, Eryu Guan <eguan@xxxxxxxxxx> wrote:
> On Wed, Jun 17, 2015 at 12:52:47PM +0100, fdmanana@xxxxxxxxxx wrote:
>> From: Filipe Manana <fdmanana@xxxxxxxx>
>>
>> This test is motivated by an issue found in btrfs.
>>
>> It tests that after syncing the filesystem, adding a xattr to a file,
>> syncing the filesystem again, writing to the file and then doing a fsync
>> against that file, the xattr still exists after a power failure. That is,
>> after the fsync log/journal is replayed, the xattr still exists and with
>> the correct value.
>>
>> The btrfs issue is fixed by the patch titled:
>>
>> "Btrfs: fix fsync xattr loss in the fast fsync path"
>
> If I read the above patch correctly, the issue to be tested here was
> introduced by commit 4f764e515361 ("Btrfs: remove deleted xattrs on
> fsync log replay"), which is in mainline kernel since v4.1-rc1, so the
> test should fail on my v4.1-rc5 kernel, but I didn't see it fail.
>
> Is it reproducible everytime? Or did I miss something?
It is reproducible everytime, but you'll need my fix for the other
test ("Btrfs: fix fsync data loss after append write") in order to
make this one fail. In other words, that other fix just exposes a
problem with commit 4f764e515361 ("Btrfs: remove deleted xattrs on
fsync log replay").
Sorry it wasn't well documented. I'll see if I can make this test fail
even without the other fix applied (i.e. current 4.1-rcs).
thanks
>
>>
>> Signed-off-by: Filipe Manana <fdmanana@xxxxxxxx>
>> ---
>> tests/generic/094 | 112 ++++++++++++++++++++++++++++++++++++++++++++++++++
>> tests/generic/094.out | 29 +++++++++++++
>> tests/generic/group | 1 +
>> 3 files changed, 142 insertions(+)
>> create mode 100755 tests/generic/094
>> create mode 100644 tests/generic/094.out
>>
>> diff --git a/tests/generic/094 b/tests/generic/094
>> new file mode 100755
>> index 0000000..1c6d113
>> --- /dev/null
>> +++ b/tests/generic/094
>> @@ -0,0 +1,112 @@
>> +#! /bin/bash
>> +# FS QA Test No. 094
>> +#
>> +# Test that after syncing the filesystem, adding a xattr to a file, syncing
>> +# the filesystem again, writing to the file and then doing a fsync against that
>> +# file, the xattr still exists after a power failure. That is, after the fsync
>> +# log/journal is replayed, the xattr still exists and with the correct value.
>> +#
>> +# This test is motivated by a bug found in btrfs.
>> +#
>> +#-----------------------------------------------------------------------
>> +# Copyright (C) 2015 SUSE Linux Products GmbH. All Rights Reserved.
>> +# Author: Filipe Manana <fdmanana@xxxxxxxx>
>> +#
>> +# 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!
>> +
>> +_cleanup()
>> +{
>> + _cleanup_flakey
>> + rm -f $tmp.*
>> +}
>> +trap "_cleanup; exit \$status" 0 1 2 3 15
>> +
>> +# get standard environment, filters and checks
>> +. ./common/rc
>> +. ./common/filter
>> +. ./common/dmflakey
>> +. ./common/attr
>> +
>> +# real QA test starts here
>> +_supported_fs generic
>> +_supported_os Linux
>> +_need_to_be_root
>> +_require_scratch
>> +_require_dm_flakey
>> +_require_attrs
>> +_require_metadata_journaling $SCRATCH_DEV
>> +
>> +rm -f $seqres.full
>> +
>> +_scratch_mkfs >> $seqres.full 2>&1
>> +_init_flakey
>> +_mount_flakey
>> +
>> +# Create the test file with some initial data and make sure everything is
>> +# durably persisted. We do fsync before calling sync to make sure that if the
>> +# filesystem is btrfs, we get the flag BTRFS_INODE_NEEDS_FULL_SYNC cleared
>> +# from the btrfs inode - a condition necessary to trigger the issue in btrfs.
>> +$XFS_IO_PROG -f -c "pwrite -S 0xaa 0 32k" \
>> + -c "fsync" \
>> + $SCRATCH_MNT/foo | _filter_xfs_io
>> +sync
>> +
>> +# Add a xattr to our file.
>> +$SETFATTR_PROG -n user.attr -v somevalue $SCRATCH_MNT/foo
>> +
>> +# Sync the filesystem to force a commit of the current btrfs transaction, this
>> +# is a necessary condition to trigger the bug on btrfs.
>> +sync
>> +
>> +# Now update our file's data and fsync the file.
>> +# After a successful fsync, if the fsync log/journal is replayed we expect to
>> +# see the xattr named "user.attr" with a value of "somevalue" (and the updated
>> +# file data of course). Btrfs used to remove the xattr when it replayed the
>> +# fsync log/journal.
>> +$XFS_IO_PROG -c "pwrite -S 0xbb 8K 16K" \
>> + -c "fsync" \
>> + $SCRATCH_MNT/foo | _filter_xfs_io
>> +
>> +echo "File content after fsync and before crash:"
>> +od -t x1 $SCRATCH_MNT/foo
>> +
>> +echo "File xattrs after fsync and before crash:"
>> +$GETFATTR_PROG --absolute-names --dump $SCRATCH_MNT/foo | _filter_scratch
>> +
>> +# Simulate a crash/power loss.
>> +_load_flakey_table $FLAKEY_DROP_WRITES
>> +_unmount_flakey
>> +
>> +# Allow writes again and mount. This makes the fs replay its fsync log.
>> +_load_flakey_table $FLAKEY_ALLOW_WRITES
>> +_mount_flakey
>> +
>> +echo "File content after crash and log replay:"
>> +od -t x1 $SCRATCH_MNT/foo
>> +
>> +echo "File xattrs after crash and log replay:"
>> +$GETFATTR_PROG --absolute-names --dump $SCRATCH_MNT/foo | _filter_scratch
>> +
>> +status=0
>> +exit
>> diff --git a/tests/generic/094.out b/tests/generic/094.out
>> new file mode 100644
>> index 0000000..2e5e0fa
>> --- /dev/null
>> +++ b/tests/generic/094.out
>> @@ -0,0 +1,29 @@
>> +QA output created by 094
>> +wrote 32768/32768 bytes at offset 0
>> +XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
>> +wrote 16384/16384 bytes at offset 8192
>> +XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
>> +File content after fsync and before crash:
>> +0000000 aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa
>> +*
>> +0020000 bb bb bb bb bb bb bb bb bb bb bb bb bb bb bb bb
>> +*
>> +0060000 aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa
>> +*
>> +0100000
>> +File xattrs after fsync and before crash:
>> +# file: SCRATCH_MNT/foo
>> +user.attr="somevalue"
>> +
>> +File content after crash and log replay:
>> +0000000 aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa
>> +*
>> +0020000 bb bb bb bb bb bb bb bb bb bb bb bb bb bb bb bb
>> +*
>> +0060000 aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa
>> +*
>> +0100000
>> +File xattrs after crash and log replay:
>> +# file: SCRATCH_MNT/foo
>> +user.attr="somevalue"
>> +
>
> git am complains about the new line at EOF, but I think it's fine here,
> it's generated by getfattr output and needed to match the output.
>
> Thanks,
> Eryu
>
>> diff --git a/tests/generic/group b/tests/generic/group
>> index ae40fed..18f31a7 100644
>> --- a/tests/generic/group
>> +++ b/tests/generic/group
>> @@ -96,6 +96,7 @@
>> 091 rw auto quick
>> 092 auto quick prealloc
>> 093 attr cap udf auto
>> +094 metadata auto quick
>> 097 udf auto
>> 099 udf auto
>> 100 udf auto
>> --
>> 2.1.3
>>
>> --
>> To unsubscribe from this list: send the line "unsubscribe fstests" in
>> the body of a message to majordomo@xxxxxxxxxxxxxxx
>> More majordomo info at http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe fstests" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
