From: Eryu Guan <eguan@redhat.com>
To: "Darrick J. Wong" <darrick.wong@oracle.com>
Cc: fstests <fstests@vger.kernel.org>
Subject: Re: [PATCH v2] ext4: fsmap tests
Date: Tue, 20 Jun 2017 23:15:00 +0800 [thread overview]
Message-ID: <20170620151500.GA17669@eguan.usersys.redhat.com> (raw)
In-Reply-To: <20170602225220.GA5634@birch.djwong.org>
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@oracle.com>
> ---
> 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
next prev parent reply other threads:[~2017-06-20 15:15 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-06-02 22:52 [PATCH v2] ext4: fsmap tests Darrick J. Wong
2017-06-16 6:34 ` Eryu Guan
2017-06-16 18:07 ` Darrick J. Wong
2017-06-20 15:15 ` Eryu Guan [this message]
2017-06-20 15:58 ` Darrick J. Wong
2017-06-21 7:43 ` Eryu Guan
2017-06-21 18:01 ` Darrick J. Wong
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20170620151500.GA17669@eguan.usersys.redhat.com \
--to=eguan@redhat.com \
--cc=darrick.wong@oracle.com \
--cc=fstests@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox