Re: [PATCH v2] ext4: fsmap tests

[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]



On Fri, Jun 02, 2017 at 03:52:20PM -0700, Darrick J. Wong wrote:
> Test the GETFSMAP ioctl against ext4.
> 
> Signed-off-by: Darrick J. Wong <darrick.wong@xxxxxxxxxx>
> ---
>  tests/ext4/700     |   76 ++++++++++++++++++++++++++++++++++++++++++++++++++++
>  tests/ext4/700.out |    7 +++++
>  tests/ext4/701     |   65 ++++++++++++++++++++++++++++++++++++++++++++
>  tests/ext4/701.out |    3 ++
>  tests/ext4/702     |   66 +++++++++++++++++++++++++++++++++++++++++++++
>  tests/ext4/702.out |    4 +++
>  tests/ext4/group   |    3 ++
>  7 files changed, 224 insertions(+)
>  create mode 100755 tests/ext4/700
>  create mode 100644 tests/ext4/700.out
>  create mode 100755 tests/ext4/701
>  create mode 100644 tests/ext4/701.out
>  create mode 100755 tests/ext4/702
>  create mode 100644 tests/ext4/702.out
> 
> diff --git a/tests/ext4/700 b/tests/ext4/700
> new file mode 100755
> index 0000000..a4a1b83
> --- /dev/null
> +++ b/tests/ext4/700
> @@ -0,0 +1,76 @@
> +#! /bin/bash
> +# FS QA Test No. 700
> +#
> +# Check that getfsmap reports the BG metadata we're expecting.
> +#
> +#-----------------------------------------------------------------------
> +# Copyright (c) 2017, 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".* $TEST_DIR/fsmap $TEST_DIR/testout
> +}
> +
> +# get standard environment, filters and checks
> +. ./common/rc
> +. ./common/filter
> +
> +# real QA test starts here
> +_supported_os Linux
> +_supported_fs ext4
> +_require_scratch
> +_require_xfs_io_command "fsmap"
> +
> +rm -f "$seqres.full"
> +
> +echo "Format and mount"
> +_scratch_mkfs > "$seqres.full" 2>&1
> +_scratch_mount
> +
> +echo "Get fsmap" | tee -a $seqres.full
> +$XFS_IO_PROG -c 'fsmap' $SCRATCH_MNT > $TEST_DIR/fsmap
> +cat $TEST_DIR/fsmap >> $seqres.full
> +
> +echo "Check fs metadata" | tee -a $seqres.full
> +x=$(grep -c 'static fs metadata' $TEST_DIR/fsmap)
> +test $x -gt 0 || echo "No fs metadata?"

This test fails if I use external log dev (with 4.12-rc5 kernel and
djwong-devel branch of your xfsprogs tree).

    --- tests/ext4/026.out      2017-06-20 20:44:11.307128655 +0800
    +++ /root/xfstests/results//ext4/026.out.bad        2017-06-20 21:02:31.361452548 +0800
    @@ -2,6 +2,10 @@
     Format and mount
     Get fsmap
     Check fs metadata
    +No fs metadata?
     Check block bitmap
    +No block bitmaps?
     Check inode bitmap
     ...

Seems this test assmues internal journal? Do we need a new
_require_no_logdev rule? (BTW, ext4/701 fails too if external log is
set).

> +
> +echo "Check block bitmap" | tee -a $seqres.full
> +x=$(grep -c 'special 102:1' $TEST_DIR/fsmap)
> +test $x -gt 0 || echo "No block bitmaps?"
> +
> +echo "Check inode bitmap" | tee -a $seqres.full
> +x=$(grep -c 'special 102:2' $TEST_DIR/fsmap)

It's not clear to me why "special 102:1" and "special 102:2" represent
"block bitmaps" and "inode bitmaps", do you mind adding some comments?

Also, the format of fsmap output seems not match the manpage? In
xfs_io.8 we have
extent: major:minor [startblock..endblock]: owner startoffset..endoffset length

and a fsmap output on ext4 filesystem is like
51: 8:33 [12582912..12583039]: special 102:3 128

So I assume 'special' is the 'owner' field, but '102:3' doesn't seem
like offsets?

> +test $x -gt 0 || echo "No inode bitmaps?"
> +
> +echo "Check inodes" | tee -a $seqres.full
> +x=$(grep -c 'inodes' $TEST_DIR/fsmap)
> +test $x -gt 0 || echo "No inodes?"
> +
> +# success, all done
> +status=0
> +exit
> diff --git a/tests/ext4/700.out b/tests/ext4/700.out
> new file mode 100644
> index 0000000..3984319
> --- /dev/null
> +++ b/tests/ext4/700.out
> @@ -0,0 +1,7 @@
> +QA output created by 700
> +Format and mount
> +Get fsmap
> +Check fs metadata
> +Check block bitmap
> +Check inode bitmap
> +Check inodes
> diff --git a/tests/ext4/701 b/tests/ext4/701
> new file mode 100755
> index 0000000..6ae93d7
> --- /dev/null
> +++ b/tests/ext4/701
> @@ -0,0 +1,65 @@
> +#! /bin/bash
> +# FS QA Test No. 701
> +#
> +# Populate filesystem, check that fsmap -n10000 matches fsmap -n1.
> +#
> +#-----------------------------------------------------------------------
> +# Copyright (c) 2017, 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".* $TEST_DIR/a $TEST_DIR/b
> +}
> +
> +# get standard environment, filters and checks
> +. ./common/rc
> +. ./common/filter
> +. ./common/populate
> +
> +# real QA test starts here
> +_supported_os Linux
> +_supported_fs ext4
> +_require_scratch
> +_require_populate_commands
> +_require_xfs_io_command "fsmap"
> +
> +rm -f "$seqres.full"
> +
> +echo "Format and mount"
> +_scratch_populate_cached nofill > $seqres.full 2>&1

This test would take very long time for me to finish, it's been running
for 1 hour and it's still running. Seems like there're too many extents
to report.

[root@ibm-x3550m3-05 xfsprogs-dev]# du -hs /mnt/testarea/test/a
163M    /mnt/testarea/test/a
[root@ibm-x3550m3-05 xfsprogs-dev]# wc -l /mnt/testarea/test/a
3170842 /mnt/testarea/test/a

I digged into it a bit, it was _scratch_populate_cached that restored a
ext4 image with external journal to current test, which has no external
log used (because I ran this test first with external journal set), so
_scratch_mount failed to mount SCRATCH_DEV and fsmap works on rootfs,
which resulted in very long extents list.

I think we should take LOGDEV into account when deciding if cached fs
image should be used, e.g. I did this update to common/populate and test
worked fine for me.

-       meta_descr="FSTYP ${FSTYP} MKFS_OPTIONS ${MKFS_OPTIONS} SIZE $(blockdev --getsz "${SCRATCH_DEV}") ARGS $@"
+       meta_descr="FSTYP ${FSTYP} MKFS_OPTIONS ${MKFS_OPTIONS} SIZE $(blockdev --getsz "${SCRATCH_DEV}") LOGDEV ${SCRATCH_LOGDEV} USE_EXTERNAL ${USE_EXTERNAL} ARGS $@"

Thanks,
Eryu

> +
> +echo "Compare fsmap" | tee -a $seqres.full
> +_scratch_mount
> +$XFS_IO_PROG -c 'fsmap -n 65536' $SCRATCH_MNT | grep -v 'EXT:' > $TEST_DIR/a
> +$XFS_IO_PROG -c 'fsmap -n 1' $SCRATCH_MNT | grep -v 'EXT:' > $TEST_DIR/b
> +cat $TEST_DIR/a $TEST_DIR/b >> $seqres.full
> +
> +diff -uw $TEST_DIR/a $TEST_DIR/b
> +
> +# success, all done
> +status=0
> +exit
> diff --git a/tests/ext4/701.out b/tests/ext4/701.out
> new file mode 100644
> index 0000000..41ef1c6
> --- /dev/null
> +++ b/tests/ext4/701.out
> @@ -0,0 +1,3 @@
> +QA output created by 701
> +Format and mount
> +Compare fsmap
> diff --git a/tests/ext4/702 b/tests/ext4/702
> new file mode 100755
> index 0000000..ce3ae11
> --- /dev/null
> +++ b/tests/ext4/702
> @@ -0,0 +1,66 @@
> +#! /bin/bash
> +# FS QA Test No. 702
> +#
> +# Check that getfsmap reports external log devices
> +#
> +#-----------------------------------------------------------------------
> +# Copyright (c) 2017, 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".* $TEST_DIR/fsmap $TEST_DIR/testout
> +}
> +
> +# get standard environment, filters and checks
> +. ./common/rc
> +. ./common/filter
> +
> +# real QA test starts here
> +_supported_os Linux
> +_supported_fs ext4
> +_require_logdev
> +_require_scratch
> +_require_xfs_io_command "fsmap"
> +
> +rm -f "$seqres.full"
> +
> +echo "Format and mount"
> +_scratch_mkfs > "$seqres.full" 2>&1
> +_scratch_mount
> +
> +echo "Get fsmap" | tee -a $seqres.full
> +$XFS_IO_PROG -c 'fsmap' $SCRATCH_MNT >> $seqres.full
> +$XFS_IO_PROG -c 'fsmap' $SCRATCH_MNT | tr '[]()' '    ' > $TEST_DIR/fsmap
> +
> +echo "Check device field of FS metadata and journalling log"
> +data_dev=$(grep 'static fs metadata' $TEST_DIR/fsmap | head -n 1 | awk '{print $2}')
> +journal_dev=$(grep 'journalling log' $TEST_DIR/fsmap | head -n 1 | awk '{print $2}')
> +test "${data_dev}" != "${journal_dev}" || echo "data ${data_dev} journal ${journal_dev}?"
> +
> +# success, all done
> +status=0
> +exit
> diff --git a/tests/ext4/702.out b/tests/ext4/702.out
> new file mode 100644
> index 0000000..1d3b7ef
> --- /dev/null
> +++ b/tests/ext4/702.out
> @@ -0,0 +1,4 @@
> +QA output created by 702
> +Format and mount
> +Get fsmap
> +Check device field of FS metadata and journalling log
> diff --git a/tests/ext4/group b/tests/ext4/group
> index 664c059..9007adf 100644
> --- a/tests/ext4/group
> +++ b/tests/ext4/group
> @@ -37,3 +37,6 @@
>  306 auto rw resize quick
>  307 auto ioctl rw defrag
>  308 auto ioctl rw prealloc quick defrag
> +700 auto quick fsmap
> +701 auto quick fsmap
> +702 auto quick fsmap
--
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




[Index of Archives]     [Linux Filesystems Development]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux