On Mon, Feb 13, 2017 at 07:55:43PM -0500, Theodore Ts'o wrote:
> From: Zorro Lang <zlang@xxxxxxxxxx>
>
> This case will do function test for mount bind operation, it will
> verify below semantics:
>
> ---------------------------------------------------------------------------
> | BIND MOUNT OPERATION |
> |**************************************************************************
> |source(A)->| shared | private | slave | unbindable |
> | dest(B) | | | | |
> | | | | | | |
> | v | | | | |
> |**************************************************************************
> | shared | shared | shared | shared & slave | invalid |
> | | | | | |
> |non-shared| shared | private | slave | invalid |
> ***************************************************************************
>
> This case use fsstress to produce a little rand load, to sure basic
> operations on the bind mountpoint won't cause hang or panic etc.
>
> Signed-off-by: Zorro Lang <zlang@xxxxxxxxxx>
> Signed-off-by: Theodore Ts'o <tytso@xxxxxxx>
> ---
> tests/generic/409 | 182 +++++++++++++++++++++++++
> tests/generic/409.out | 365 ++++++++++++++++++++++++++++++++++++++++++++++++++
> tests/generic/group | 1 +
> 3 files changed, 548 insertions(+)
> create mode 100644 tests/generic/409
File mode for new case is better to be 755, so after running tests git
diff won't complain about the mode change.
> create mode 100644 tests/generic/409.out
>
> diff --git a/tests/generic/409 b/tests/generic/409
> new file mode 100644
> index 00000000..88bda6c5
> --- /dev/null
> +++ b/tests/generic/409
> @@ -0,0 +1,182 @@
> +#! /bin/bash
> +# FS QA Test 409
> +#
> +# Test mount shared subtrees, verify the bind semantics:
> +#
> +# ---------------------------------------------------------------------------
> +# | BIND MOUNT OPERATION |
> +# |**************************************************************************
> +# |source(A)->| shared | private | slave | unbindable |
> +# | dest(B) | | | | |
> +# | | | | | | |
> +# | v | | | | |
> +# |**************************************************************************
> +# | shared | shared | shared | shared & slave | invalid |
> +# | | | | | |
> +# |non-shared| shared | private | slave | invalid |
> +# ***************************************************************************
> +#
> +#-----------------------------------------------------------------------
> +# Copyright (c) 2016 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/$$
> +status=1 # failure is the default!
> +trap "_cleanup; exit \$status" 0 1 2 3 15
> +
> +_cleanup()
> +{
> + cd /
> + rm -f $tmp.*
> + _clear_mount_stack
> +}
> +
> +# get standard environment, filters and checks
> +. ./common/rc
> +. ./common/filter
> +
> +# remove previous $seqres.full before test
> +rm -f $seqres.full
> +
> +# real QA test starts here
> +_supported_fs generic
> +_supported_os Linux
> +_require_scratch
> +_require_test
Need a "_require_block_device $SCRATCH_DEV", otherwise test on
NFS/overlay etc. will fail, because we assume $SCRATCH_DEV is a block
device.
> +
> +fs_stress()
> +{
> + local target=$1
> +
> + $FSSTRESS_PROG -z -n 50 -p 3 \
> + -f creat=5 \
> + -f mkdir=5 \
> + -f link=2 \
> + -f rename=1 \
> + -f rmdir=2 \
> + -f unlink=1 \
> + -f symlink=1 \
> + -f write=1 \
> + -f read=1 \
> + -f chown=1 \
> + -f getdents=1 \
> + -f fiemap=1 \
> + -d $target >/dev/null
> + sync
> +}
> +
> +# prepare some mountpoint dir
> +MNTHEAD=$TEST_DIR/$seq
> +mkdir $MNTHEAD 2>>$seqres.full
> +mpA=$MNTHEAD/"$$"_mpA
> +mpB=$MNTHEAD/"$$"_mpB
> +mpC=$MNTHEAD/"$$"_mpC
> +mpD=$MNTHEAD/"$$"_mpD
> +
> +find_mnt()
> +{
> + echo "------"
> + findmnt -n -o TARGET,SOURCE $SCRATCH_DEV | \
> + sed -e "s;$mpA;mpA;g" \
> + -e "s;$mpB;mpB;g" \
> + -e "s;$mpC;mpC;g" \
> + -e "s;$mpD;mpD;g" | \
> + _filter_spaces | _filter_scratch | \
> + _filter_test_dir | sort
> + echo "======"
> +}
> +
> +start_test()
> +{
> + local type=$1
> +
> + _get_mount $SCRATCH_DEV $MNTHEAD
> + mount --make-"${type}" $MNTHEAD
> + rm -rf $mpA $mpB $mpC $mpD 2>/dev/null
> + mkdir $mpA $mpB $mpC $mpD
> + _scratch_mkfs >$seqres.full 2>&1
We should _scratch_mkfs before mounting $SCRATCH_DEV on $MNTHEAD, and
since the filesystem is newly created, the "rm -rf" line seems not
necessary.
> +}
> +
> +end_test()
> +{
> + rm -rf $mpA $mpB $mpC $mpD 2>/dev/null
> + _clear_mount_stack
Unmount all mount points then remove them, otherwise rm always fails
because of "Device is busy".
> +}
> +
> +bind_run()
> +{
> + local source=$1
> + local dest=$2
> +
> + start_test $dest
> +
> + echo "bind $source on $dest"
> + _get_mount $SCRATCH_DEV $mpA
> + mkdir -p $mpA/dir 2>/dev/null
> + mount --make-shared $mpA
> + _get_mount --bind $mpA $mpB
> + mount --make-"$source" $mpB
> + # maybe unbindable at here
> + _get_mount --bind $mpB $mpC 2>/dev/null
> + if [ $? -ne 0 ];then
> + find_mnt
> + end_test
> + return 0
> + fi
> + _get_mount --bind $mpC $mpD
> + for m in $mpA $mpB $mpC $mpD
> + do
> + _get_mount $SCRATCH_DEV $m/dir
> + fs_stress $m/dir
> + find_mnt
> + _put_mount
> + done
> +
> + end_test
> +}
> +
> +bind_test()
> +{
> + local dest=$1
dest is not used, can be removed.
The other two tests have similar issues, I'll fix them all.
Thanks,
Eryu
--
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
