On 2015/08/05 10:08, Qu Wenruo wrote:
> The regression is introduced in v4.2-rc1, with the big btrfs qgroup
> change.
> The problem is, qgroup reserved space is never freed, causing even we
> increase the limit, we can still hit the EDQUOT much faster than it
> should.
>
> Reported-by: Tsutomu Itoh <t-i...@jp.fujitsu.com>
> Signed-off-by: Qu Wenruo <quwen...@cn.fujitsu.com>
> ---
> tests/btrfs/089 | 88
> +++++++++++++++++++++++++++++++++++++++++++++++++++++
> tests/btrfs/089.out | 5 +++
> tests/btrfs/group | 1 +
> 3 files changed, 94 insertions(+)
> create mode 100755 tests/btrfs/089
> create mode 100644 tests/btrfs/089.out
>
> diff --git a/tests/btrfs/089 b/tests/btrfs/089
> new file mode 100755
> index 0000000..82db96c
> --- /dev/null
> +++ b/tests/btrfs/089
> @@ -0,0 +1,88 @@
> +#! /bin/bash
> +# FS QA Test 089
> +#
> +# Regression test for btrfs qgroup reserved space leak.
> +#
> +# Due to qgroup reserved space leak, EDQUOT can be trigged even it's not
> +# over limit after previous write.
> +#
> +#-----------------------------------------------------------------------
> +# Copyright (c) 2015 Fujitsu. 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 $tmp.*
> +}
> +
> +# get standard environment, filters and checks
> +. ./common/rc
> +. ./common/filter
> +
> +# real QA test starts here
> +
> +# Modify as appropriate.
> +_supported_fs btrfs
> +_supported_os Linux
> +_require_scratch
> +_need_to_be_root
> +
> +# Use big blocksize to ensure there is still enough space left
> +# for metadata reserve after hitting EDQUOT
> +BLOCKSIZE=$(( 2 * 1024 * 1024 ))
> +FILESIZE=$(( 128 * 1024 * 1024 )) # 128Mbytes
> +
> +# The last block won't be able to finish write, as metadata takes
> +# $NODESIZE space, causing the last block triggering EDQUOT
> +LENGTH=$(( $FILESIZE - $BLOCKSIZE ))
> +
> +_scratch_mkfs >>$seqres.full 2>&1
> +_scratch_mount
> +_require_fs_space $SCRATCH_MNT $(($FILESIZE * 2 / 1024))
> +
> +_run_btrfs_util_prog quota enable $SCRATCH_MNT
> +_run_btrfs_util_prog qgroup limit $FILESIZE 5 $SCRATCH_MNT
> +
> +$XFS_IO_PROG -f -c "pwrite -b $BLOCKSIZE 0 $LENGTH" \
> + $SCRATCH_MNT/foo | _filter_xfs_io
> +
> +# A sync is needed to trigger a commit_transaction.
> +# As the reserved space freeing happens at commit_transaction time,
> +# without a transaction commit, no reserved space needs freeing and
> +# won't trigger the bug.
> +sync
Isn't '$BTRFS_UTIL_PROG filesystem sync' better instead of 'sync'?
Thanks,
Tsutomu
> +
> +# Double the limit to allow further write
> +_run_btrfs_util_prog qgroup limit $(($FILESIZE * 2)) 5 $SCRATCH_MNT
> +
> +# Test whether further write can succeed
> +$XFS_IO_PROG -f -c "pwrite -b $BLOCKSIZE $LENGTH $LENGTH" \
> + $SCRATCH_MNT/foo | _filter_xfs_io
> +
> +# success, all done
> +status=0
> +exit
> diff --git a/tests/btrfs/089.out b/tests/btrfs/089.out
> new file mode 100644
> index 0000000..396888f
> --- /dev/null
> +++ b/tests/btrfs/089.out
> @@ -0,0 +1,5 @@
> +QA output created by 089
> +wrote 132120576/132120576 bytes at offset 0
> +XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
> +wrote 132120576/132120576 bytes at offset 132120576
> +XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
> diff --git a/tests/btrfs/group b/tests/btrfs/group
> index ffe18bf..225b532 100644
> --- a/tests/btrfs/group
> +++ b/tests/btrfs/group
> @@ -91,6 +91,7 @@
> 086 auto quick clone
> 087 auto quick send
> 088 auto quick metadata
> +089 auto quick qgroup
> 090 auto quick metadata
> 091 auto quick qgroup
> 092 auto quick send
>
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
