On Wed, Sep 14, 2016 at 05:24:00PM +0800, Eryu Guan wrote:
> On Wed, Sep 14, 2016 at 04:02:09PM +0800, Zorro Lang wrote:
> > There's a bug, the ustat() sysetm call is not implemented on someone
> > arch, but xfsprogs uses this to detect whether a filesystem is
> > mounted in platform_check_ismounted() function.
> >
> > If ustat() is not implemented, platform_check_ismounted() return
> > "isn't mounted" if a filesystem is mounted, that's a big problem, it
> > will cause corruption.
> >
> > Signed-off-by: Zorro Lang <zlang@xxxxxxxxxx>
> > ---
> > tests/xfs/284 | 113 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
> > tests/xfs/284.out | 8 ++++
> > tests/xfs/group | 1 +
> > 3 files changed, 122 insertions(+)
> > create mode 100755 tests/xfs/284
> > create mode 100644 tests/xfs/284.out
> >
> > diff --git a/tests/xfs/284 b/tests/xfs/284
> > new file mode 100755
> > index 0000000..53fbcb1
> > --- /dev/null
> > +++ b/tests/xfs/284
> > @@ -0,0 +1,113 @@
> > +#! /bin/bash
> > +# FS QA Test 284
> > +#
> > +# Make sure platform_check_ismounted() do its work. A bug cause
> > +# platform_check_ismounted return "isn't mounted" status, even if
> > +# the filesystem is mounted. That maybe cause corruption, so must
> > +# make sure this function always do its work.
> > +#
> > +#-----------------------------------------------------------------------
> > +# Copyright (c) 2016 Red Hat. 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.*
> > + rm -f $METADUMP_FILE 2>/dev/null
> > + rm -f $COPY_FILE 2>/dev/null
> > +}
> > +
> > +# get standard environment, filters and checks
> > +. ./common/rc
> > +. ./common/filter
> > +
> > +# remove previous $seqres.full before test
> > +rm -f $seqres.full
> > +
> > +# real QA test starts here
> > +_supported_fs xfs
> > +_supported_os Linux
> > +_require_test
> > +_require_scratch
> > +
> > +function filter_mounted()
> > +{
> > + grep "mounted" | _filter_scratch | head -1
> > +}
> > +
> > +METADUMP_FILE="${TEST_DIR}/${seq}_metadump"
> > +COPY_FILE="${TEST_DIR}/${seq}_copyfile"
> > +
> > +# Test dump a mounted device
> > +_scratch_mount
> > +_scratch_metadump $METADUMP_FILE 2>&1 | filter_mounted
> > +[ ${PIPESTATUS[0]} -eq 0 ] && echo "Test dump a mounted device [Failed]"
>
> I don't think these status checks are needed, because if these
> metadump/copy/repair/mkfs commands proceed on mounted fs, the golden
> image will be broken.
Originally I want to check the return status, and print the output to
$seq.full. Because the output can prove platform_check_ismounted() finds
the fs is mounted, the return value can prove these operations didn't break
a mounted fs and return pass. (Maybe it do_log() some messages, but
forgot to exit non-zero :-P. OK, maybe I thought too much:)
Then I found some operations still can return non-zero even if
platform_check_ismounted() return a wrong value (e.g: mkfs.xfs). So
I decided to check the output with the return value together.
So if it's OK to check both of output and return, I hope to retain them.
Or I hope to check return value, and print the output to $seq.full.
>
> > +_scratch_unmount
> > +
> > +# Test restore a mounted device
>
> "restore to a mounted device"
Sure.
>
> > +_scratch_metadump $METADUMP_FILE
> > +_scratch_mount
> > +xfs_mdrestore $METADUMP_FILE $SCRATCH_DEV 2>&1 | filter_mounted
> > +[ ${PIPESTATUS[0]} -eq 0 ] && echo "Test restore a mounted device [Failed]"
> > +_scratch_unmount
> > +
> > +# Test xfs_copy to a mounted device
> > +# If source is mounted, xfs_copy will print a warning, but still
> > +# keep on copying. If target is mounted, xfs_copy will failed.
> > +$XFS_COPY_PROG $SCRATCH_DEV $COPY_FILE >/dev/null
> > +_scratch_mount
> > +$XFS_COPY_PROG $COPY_FILE $SCRATCH_DEV 2>&1 | filter_mounted
> > +[ ${PIPESTATUS[0]} -eq 0 ] && echo "Test xfs_copy to a mounted device [Failed]"
> > +_scratch_unmount
> > +
> > +# Test xfs_db a mounted device without readonly (-r) option
> > +_scratch_mount
> > +$XFS_DB_PROG -c sb $SCRATCH_DEV 2>&1 | filter_mounted
> > +[ ${PIPESTATUS[0]} -eq 0 ] &&
> > + echo "Test xfs_db a mounted device(without -r option) [Failed]"
> > +_scratch_unmount
> > +
> > +# Test mkfs.xfs a mounted device
> > +# Generally mkfs will report device is mounted, but if it can't find
> > +# device is mounted, it'll report device busy.
> > +_scratch_mount
> > +_scratch_mkfs 2>&1 | filter_mounted
> > +[ ${PIPESTATUS[0]} -eq 0 ] && echo "Test mkfs.xfs a mounted device [Failed]"
> > +_scratch_unmount
> > +
> > +# Test xfs_repair (with/without modify flag) a mounted device
> > +_scratch_mount
> > +_scratch_xfs_repair -n 2>&1 | filter_mounted
> > +[ ${PIPESTATUS[0]} -eq 0 ] &&
> > + echo "Test xfs_repair -n a mounted devie [Failed]"
> > +_scratch_xfs_repair 2>&1 | filter_mounted
> > +[ ${PIPESTATUS[0]} -eq 0 ] && echo "Test xfs_repair a mounted devie [Failed]"
> > +_scratch_unmount
> > +
> > +# success, all done
> > +status=0
> > +exit
> > diff --git a/tests/xfs/284.out b/tests/xfs/284.out
> > new file mode 100644
> > index 0000000..302d75e
> > --- /dev/null
> > +++ b/tests/xfs/284.out
> > @@ -0,0 +1,8 @@
> > +QA output created by 284
> > +xfs_metadump: SCRATCH_DEV contains a mounted and writable filesystem
> > +xfs_mdrestore: a filesystem is mounted on target device "SCRATCH_DEV", cannot restore to a mounted filesystem.
> > +xfs_copy: a filesystem is mounted on target device "SCRATCH_DEV".
> > +xfs_db: SCRATCH_DEV contains a mounted filesystem
> > +mkfs.xfs: SCRATCH_DEV contains a mounted filesystem
> > +xfs_repair: SCRATCH_DEV contains a mounted and writable filesystem
> > +xfs_repair: SCRATCH_DEV contains a mounted filesystem
> > diff --git a/tests/xfs/group b/tests/xfs/group
> > index baa0ff6..0ac4bd5 100644
> > --- a/tests/xfs/group
> > +++ b/tests/xfs/group
> > @@ -281,6 +281,7 @@
> > 281 dump ioctl auto quick
> > 282 dump ioctl auto quick
> > 283 dump ioctl auto quick
> > +284 auto quick mount
>
> I'm not sure if this belongs to 'mount' group, it does test against
> mounted device, not actually test mount.
>
> But perhaps we can add 'dump' 'copy' 'mkfs' 'db' 'repair' groups?
Hmm... sure, add them is ok for me:)
>
> Thanks,
> Eryu
>
> > 287 auto dump quota quick
> > 290 auto rw prealloc quick ioctl zero
> > 291 auto repair
> > --
> > 2.7.4
> >
> > --
> > 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 linux-xfs" 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
