From: Eryu Guan <eguan@redhat.com>
To: "Darrick J. Wong" <darrick.wong@oracle.com>
Cc: linux-xfs@vger.kernel.org, fstests@vger.kernel.org
Subject: Re: [PATCH 5/8] reflink: test unlinking a huge extent with a lot of refcount adjustments
Date: Thu, 29 Jun 2017 17:36:14 +0800 [thread overview]
Message-ID: <20170629093614.GO23360@eguan.usersys.redhat.com> (raw)
In-Reply-To: <149808226863.8924.5799576767468365376.stgit@birch.djwong.org>
On Wed, Jun 21, 2017 at 02:57:48PM -0700, Darrick J. Wong wrote:
> From: Darrick J. Wong <darrick.wong@oracle.com>
>
> Test a regression in XFS where we blow out a transaction reservation if
> we create a big file, share every other block, and delete the first
> file. There's nothing particularly fs-specific about this stress test,
> so put it in generic.
>
> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
This test took me 3019s to finish with v4.12-rc7 kernel, and another
host "hung" at "Delete file1" (it had been more than 1 hour, and I lost
my patience and hard-reboot the host). Are they expected results?
If the bug is still existed in latest upstream kernel, I tend to merge
it after the fix landing in linus tree. If v4.12-rc7 doesn't suffer from
this bug, the test time should be reduced.
And another minor nit below.
> ---
> tests/generic/931 | 94 +++++++++++++++++++++++++++++++++++++++++++++++++
> tests/generic/931.out | 6 +++
> tests/generic/group | 1 +
> 3 files changed, 101 insertions(+)
> create mode 100755 tests/generic/931
> create mode 100644 tests/generic/931.out
>
>
> diff --git a/tests/generic/931 b/tests/generic/931
> new file mode 100755
> index 0000000..afadf81
> --- /dev/null
> +++ b/tests/generic/931
> @@ -0,0 +1,94 @@
> +#! /bin/bash
> +# FS QA Test No. 931
> +#
> +# See how well we handle deleting a file with a million refcount extents.
> +#
> +#-----------------------------------------------------------------------
> +# 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 -rf "$tmp".* $testdir/file1
'rm -rf' looks a bit scary, and we're only deleting regular files not
directories, 'rm -f' should be sufficient.
Thanks,
Eryu
> +}
> +
> +# get standard environment, filters and checks
> +. ./common/rc
> +. ./common/filter
> +. ./common/attr
> +. ./common/reflink
> +
> +# real QA test starts here
> +_supported_os Linux
> +_require_scratch_reflink
> +_require_cp_reflink
> +_require_test_program "punch-alternating"
> +
> +rm -f "$seqres.full"
> +
> +echo "Format and mount"
> +_scratch_mkfs > "$seqres.full" 2>&1
> +_scratch_mount >> "$seqres.full" 2>&1
> +
> +testdir="$SCRATCH_MNT/test-$seq"
> +mkdir "$testdir"
> +
> +# Setup for one million blocks, but we'll accept stress testing down to
> +# 2^17 blocks... that should be plenty for anyone.
> +fnr=20
> +free_blocks=$(stat -f -c '%a' "$testdir")
> +blksz=$(_get_block_size "$testdir")
> +space_avail=$((free_blocks * blksz))
> +calc_space() {
> + blocks_needed=$(( 2 ** (fnr + 1) ))
> + space_needed=$((blocks_needed * blksz * 5 / 4))
> +}
> +calc_space
> +while test $space_needed -gt $space_avail; do
> + fnr=$((fnr - 1))
> + calc_space
> +done
> +test $fnr -lt 17 && _notrun "Insufficient space for stress test; would only create $blocks_needed extents ($space_needed/$space_avail blocks)."
> +
> +echo "Create a many-block file"
> +echo "creating $blocks_needed blocks..." >> "$seqres.full"
> +$XFS_IO_PROG -f -c "pwrite -S 0x61 -b 4194304 0 $((2 ** (fnr + 1) * blksz))" "$testdir/file1" >> "$seqres.full"
> +
> +echo "Reflinking file"
> +_cp_reflink $testdir/file1 $testdir/file2
> +
> +echo "Punch file2"
> +echo "Punching file2..." >> "$seqres.full"
> +"$here/src/punch-alternating" "$testdir/file2" >> "$seqres.full"
> +echo "...done" >> "$seqres.full"
> +_scratch_cycle_mount
> +
> +echo "Delete file1"
> +rm -rf $testdir/file1
> +
> +# success, all done
> +status=0
> +exit
> diff --git a/tests/generic/931.out b/tests/generic/931.out
> new file mode 100644
> index 0000000..c7b724e
> --- /dev/null
> +++ b/tests/generic/931.out
> @@ -0,0 +1,6 @@
> +QA output created by 931
> +Format and mount
> +Create a many-block file
> +Reflinking file
> +Punch file2
> +Delete file1
> diff --git a/tests/generic/group b/tests/generic/group
> index ab1e9d3..b0d1844 100644
> --- a/tests/generic/group
> +++ b/tests/generic/group
> @@ -443,3 +443,4 @@
> 438 auto
> 439 auto quick punch
> 440 auto quick encrypt
> +931 auto quick clone
>
next prev parent reply other threads:[~2017-06-29 9:36 UTC|newest]
Thread overview: 34+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-06-21 21:57 [PATCH 0/8] miscellaneous tests Darrick J. Wong
2017-06-21 21:57 ` [PATCH 1/8] ext4: fsmap tests Darrick J. Wong
2017-06-23 7:41 ` Eryu Guan
2017-06-23 15:35 ` Darrick J. Wong
2017-06-21 21:57 ` [PATCH 2/8] xfs/274: flip shared bits to reflect xfsprogs usage Darrick J. Wong
2017-06-21 21:57 ` [PATCH 3/8] xfs: don't allow realtime swap files Darrick J. Wong
2017-06-21 21:57 ` [PATCH 4/8] xfs/040: use compare-libxfs in xfsprogs Darrick J. Wong
2017-06-21 21:57 ` [PATCH 5/8] reflink: test unlinking a huge extent with a lot of refcount adjustments Darrick J. Wong
2017-06-29 9:36 ` Eryu Guan [this message]
2017-06-29 16:07 ` Darrick J. Wong
2017-06-29 17:19 ` Eryu Guan
2017-06-21 21:57 ` [PATCH 6/8] xfs: test that we can SEEK_HOLE/DATA data and holes that are in the CoW fork only Darrick J. Wong
2017-06-21 21:58 ` [PATCH 7/8] xfs: test freeze/rmap repair race Darrick J. Wong
2017-06-29 9:47 ` Eryu Guan
2017-06-29 17:17 ` Darrick J. Wong
2017-06-21 21:58 ` [PATCH 8/8] xfs: scrub while appending to a file Darrick J. Wong
2017-06-22 6:26 ` [PATCH 9/8] common/populate: remember multi-device configurations Darrick J. Wong
2017-06-23 7:49 ` Eryu Guan
2017-06-23 15:34 ` Darrick J. Wong
2017-06-30 4:12 ` [PATCH v2 " Darrick J. Wong
2017-06-23 7:59 ` [PATCH 0/8] miscellaneous tests Eryu Guan
2017-06-30 4:12 ` [PATCH 10/8] common/rc: test that the xfs_io scrub/repair commands actually work Darrick J. Wong
2017-06-30 4:13 ` [PATCH 11/8] common/inject: refactor helpers to use new errortag interface Darrick J. Wong
2017-07-05 12:16 ` Brian Foster
2017-07-05 16:29 ` Darrick J. Wong
2017-07-05 16:45 ` Brian Foster
2017-07-05 21:05 ` Darrick J. Wong
2017-07-06 10:07 ` Brian Foster
2017-06-30 4:13 ` [PATCH 12/8] ext4: don't online scrub ever Darrick J. Wong
2017-06-30 4:22 ` Eryu Guan
2017-06-30 4:26 ` Darrick J. Wong
2017-06-30 4:33 ` Eryu Guan
2017-06-30 4:58 ` [PATCH v2 " Darrick J. Wong
2017-06-30 12:32 ` Eryu Guan
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=20170629093614.GO23360@eguan.usersys.redhat.com \
--to=eguan@redhat.com \
--cc=darrick.wong@oracle.com \
--cc=fstests@vger.kernel.org \
--cc=linux-xfs@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;
as well as URLs for NNTP newsgroup(s).