From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from cn.fujitsu.com ([59.151.112.132]:51376 "EHLO heian.cn.fujitsu.com" rhost-flags-OK-FAIL-OK-FAIL) by vger.kernel.org with ESMTP id S1752449AbcF2DYk (ORCPT ); Tue, 28 Jun 2016 23:24:40 -0400 Received: from G08CNEXCHPEKD01.g08.fujitsu.local (unknown [10.167.33.80]) by cn.fujitsu.com (Postfix) with ESMTP id D69D541C0BB7 for ; Wed, 29 Jun 2016 11:24:23 +0800 (CST) Subject: Re: [PATCH] btrfs: check truncate can update file size correctly when truncate fails References: <1466667400-21639-1-git-send-email-wangxg.fnst@cn.fujitsu.com> <20160623082718.GF3226@eguan.usersys.redhat.com> From: Wang Xiaoguang Message-ID: <57733F06.9000103@cn.fujitsu.com> Date: Wed, 29 Jun 2016 11:22:46 +0800 MIME-Version: 1.0 In-Reply-To: <20160623082718.GF3226@eguan.usersys.redhat.com> Content-Type: text/plain; charset="windows-1252"; format=flowed Content-Transfer-Encoding: 7bit Sender: fstests-owner@vger.kernel.org To: fstests@vger.kernel.org List-ID: hello, Thanks for your comments. I will update the patch according to your comments and make it to generic tests. Regards, Xiaoguang Wang On 06/23/2016 04:27 PM, Eryu Guan wrote: > On Thu, Jun 23, 2016 at 03:36:40PM +0800, Wang Xiaoguang wrote: >> In btrfs, when truncate operation fails for enospc reason, file may still >> have some disk blocks, but it will fail to update filesize accordingly. >> >> Signed-off-by: Wang Xiaoguang >> --- >> tests/btrfs/124 | 86 +++++++++++++++++++++++++++++++++++++++++++++++++++++ >> tests/btrfs/124.out | 2 ++ >> tests/btrfs/group | 1 + >> 3 files changed, 89 insertions(+) >> create mode 100755 tests/btrfs/124 >> create mode 100644 tests/btrfs/124.out >> >> diff --git a/tests/btrfs/124 b/tests/btrfs/124 >> new file mode 100755 >> index 0000000..1a13ddf >> --- /dev/null >> +++ b/tests/btrfs/124 >> @@ -0,0 +1,86 @@ >> +#! /bin/bash >> +# FS QA Test 124 >> +# >> +# Test whether truncate can update file size correctly when truncate fails >> +# >> +# When truncate operation fails for enospc reason, file will still >> +# have some disk blocks, but it will fail to update filesize accordingly. >> +# >> +#----------------------------------------------------------------------- >> +# Copyright (c) 2016 Fujitsu. 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.* >> +} >> + >> +# get standard environment, filters and checks >> +. ./common/rc >> +. ./common/filter >> +. ./common/reflink >> + >> +# remove previous $seqres.full before test >> +rm -f $seqres.full >> + >> +# Modify as appropriate. >> +_supported_fs generic >> +_supported_os Linux >> +_supported_fs btrfs > Two "_supported_fs" here. > > And I don't see any other btrfs-specific configs except MKFS_OPTIONS, is > it possible to make it a generic test? > >> +_require_scratch_reflink >> + >> +MKFS_OPTIONS="-O ^mixed-bg" > Need comments on why testing with this MKFS_OPTIONS. > >> +_scratch_mkfs_sized $((101 * 1024 * 1024)) >> $seqres.full 2>&1 > 101M is too small to make btrfs on ppc64 host for me. > >> +_scratch_mount >> + >> +blocksize=$((128 * 1024)) >> +file="$SCRATCH_MNT/testfile" >> +filesize=$((12 * 1024 * 1024 * 1024)) > Comments on these numbers picked would be good, i.e. why 128k blocksize, > why 12M filesize. Any special reason or just good enough for the test? > >> +nr=$((filesize / blocksize)) >> + >> +# write the initial block for later reflink >> +_pwrite_byte 0xcdcdcdcd 0 $blocksize $file > /dev/null >> + >> +# use reflink to create the rest of the file, whose all extents are all >> +# pointing to the first extent >> +for ((i = 1; i < $nr; i++)); do >> + _reflink_range $file 0 $file $(($i * $blocksize)) $blocksize > /dev/null >> +done > It's not clear to me in test description that why reflink is needed in > this test. Perhaps test description and comments above can be updated to > reflect the purpose of reflink? > >> + >> +$XFS_IO_PROG -f -c "truncate 0" $file > /dev/null 2>&1 >> +disk_usage=`du $file | $AWK_PROG '{print $1}'` >> +new_filesize=`ls -l $file | $AWK_PROG '{print $5}'` > I think "stat -c %s $file" is much easier :) > >> +if [ $disk_usage -gt 0 ] && [ $new_filesize -eq 0 ]; then >> + echo "after truncate, file size is 0, but file still owns disk blocks" >> + status=1 >> + exit >> +fi >> + >> +echo "Silence is golden" >> +# success, all done >> +status=0 >> +exit >> diff --git a/tests/btrfs/124.out b/tests/btrfs/124.out >> new file mode 100644 >> index 0000000..d7339e6 >> --- /dev/null >> +++ b/tests/btrfs/124.out >> @@ -0,0 +1,2 @@ >> +QA output created by 124 >> +Silence is golden >> diff --git a/tests/btrfs/group b/tests/btrfs/group >> index 5a26ed7..8b5050e 100644 >> --- a/tests/btrfs/group >> +++ b/tests/btrfs/group >> @@ -126,3 +126,4 @@ >> 121 auto quick snapshot qgroup >> 122 auto quick snapshot qgroup >> 123 auto quick qgroup >> +124 auto quick metadata > Add it to 'clone' group as well? > > Thanks, > Eryu > >