From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753370AbbAOKOe (ORCPT ); Thu, 15 Jan 2015 05:14:34 -0500 Received: from mailout2.samsung.com ([203.254.224.25]:19303 "EHLO mailout2.samsung.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752143AbbAOKO3 (ORCPT ); Thu, 15 Jan 2015 05:14:29 -0500 X-AuditID: cbfee68d-f79296d000004278-58-54b793036e6b From: Namjae Jeon To: "'Brian Foster'" , "'Namjae Jeon'" Cc: david@fromorbit.com, tytso@mit.edu, linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org, linux-ext4@vger.kernel.org, xfs@oss.sgi.com, a.sangwan@samsung.com References: <1421165126-3585-1-git-send-email-linkinjeon@gmail.com> <1421165126-3585-10-git-send-email-linkinjeon@gmail.com> <20150114201710.GB49511@bfoster.bfoster> In-reply-to: <20150114201710.GB49511@bfoster.bfoster> Subject: RE: [PATCH v8 9/11] xfstests: generic/043: Test multiple fallocate insert/collapse range calls Date: Thu, 15 Jan 2015 19:14:26 +0900 Message-id: <004301d030ac$0ad62960$20827c20$@samsung.com> MIME-version: 1.0 Content-type: text/plain; charset=us-ascii Content-transfer-encoding: 7bit X-Mailer: Microsoft Outlook 14.0 Thread-index: AQI8HFfblg/gPWIGszLKUjAczsKp1gJE7vJ0AfQXG+Obx1UzIA== Content-language: ko X-Brightmail-Tracker: H4sIAAAAAAAAA+NgFrrFIsWRmVeSWpSXmKPExsWyRsSkWJd58vYQgw87+CyWTrzEbPHuc5XF lmP3GC2u373FbDFz3h02iz17T7JYXN41h82itecnu8WivluMDpwepxZJeOycdZfdo+nMUWaP 1Re2Mnq833eVzaNvyypGj8+b5ALYo7hsUlJzMstSi/TtErgyHrx4zljw3Lri8IKzbA2Mywy6 GDk5JARMJL7fWssOYYtJXLi3ng3EFhJYyigxc5EmTE1r9yKgGi6g+HRGiZ9r5rFCOH8ZJZYc awPKcHCwCWhL/NkiCtIgIuArcX/5R7AaZoHVjBLnp3RDNaxklHh18iorSBWngKnE61N7mEFs YYFMic8XboHZLAKqEmeO/QI7g1fAUqKnZyGULSjxY/I9FhCbWUBLYv3O40wQtrzE5jVvmSFO VZDYcfY1I8QVThJfr+6FqhGR2PfiHSPIERICjRwSc/bsY4JYJiDxbfIhFpAPJARkJTYdgJoj KXFwxQ2WCYwSs5CsnoVk9Swkq2chWbGAkWUVo2hqQXJBcVJ6kaFecWJucWleul5yfu4mRmCU n/73rHcH4+0D1ocYBTgYlXh4Gfy2hwixJpYVV+YeYjQFumgis5Rocj4wleSVxBsamxlZmJqY GhuZW5opifMqSv0MFhJITyxJzU5NLUgtii8qzUktPsTIxMEp1cDolnBhV1XLajXGiwzR95v2 /N90/Mc75/V3p/zds/5C95dJBvYzpB52as2SXmvrmHL0zWpTtQ/OLzi0Lh+Vikmt62b5Lhm+ 7+tv1i/pUYdUuN+7qjtwGjS47zykeeiI/oFtzu9knPc9DvK4c3z/otLkk3e90znlXZP97q+c s4j1kbTi6cLn6753K7EUZyQaajEXFScCANNNN8ztAgAA X-Brightmail-Tracker: H4sIAAAAAAAAA+NgFlrKKsWRmVeSWpSXmKPExsVy+t9jQV3mydtDDGYfZLFYOvESs8W7z1UW W47dY7S4fvcWs8XMeXfYLPbsPclicXnXHDaL1p6f7BaL+m4xOnB6nFok4bFz1l12j6YzR5k9 Vl/Yyujxft9VNo++LasYPT5vkgtgj2pgtMlITUxJLVJIzUvOT8nMS7dV8g6Od443NTMw1DW0 tDBXUshLzE21VXLxCdB1y8wBukxJoSwxpxQoFJBYXKykb4dpQmiIm64FTGOErm9IEFyPkQEa SFjDmPHgxXPGgufWFYcXnGVrYFxm0MXIySEhYCLR2r2IHcIWk7hwbz1bFyMXh5DAdEaJn2vm sUI4fxkllhxrA6ri4GAT0Jb4s0UUpEFEwFfi/vKPYDXMAqsZJc5P6YZqWMko8erkVVaQKk4B U4nXp/Ywg9jCApkSny/cArNZBFQlzhz7xQZi8wpYSvT0LISyBSV+TL7HAmIzC2hJrN95nAnC lpfYvOYtM8SpChI7zr5mhLjCSeLr1b1QNSIS+168Y5zAKDQLyahZSEbNQjJqFpKWBYwsqxhF UwuSC4qT0nMN9YoTc4tL89L1kvNzNzGCk8gzqR2MKxssDjEKcDAq8fAy+G0PEWJNLCuuzD3E KMHBrCTC25kDFOJNSaysSi3Kjy8qzUktPsRoCvTpRGYp0eR8YILLK4k3NDYxM7I0Mje0MDI2 VxLnVbJvCxESSE8sSc1OTS1ILYLpY+LglGpg7K6arCow/byWV5bRnCt79c7f5G2eOIWdd9k5 m+yoJT6/NwqE7LkUaBc5ocO37kPjC/EHxXFSniossde6Kh3uMxY31GRd+jB3VYz/totrst5k Miv9u3X0jHFGjNfCWGajW/V15eof5h/yXNW0fR9TvndAVcYih8Wb58jNNxKo2J2cfXptdEWq EktxRqKhFnNRcSIAVhnlWzgDAAA= DLP-Filter: Pass X-MTR: 20000000000000000@CPGS X-CFilter-Loop: Reflected Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org > > +_require_scratch > > +_require_xfs_io_command "fiemap" > > +_require_xfs_io_command "finsert" > > +_require_xfs_io_command "fcollapse" > > +_do_die_on_error=y > > What is _do_die_on_error for? Seems like that's only relevant for using > _do()? > > > +src=$SCRATCH_MNT/testfile > > +dest=$SCRATCH_MNT/testfile.dest > > +BLOCKS=100 > > +BSIZE=`get_block_size $SCRATCH_MNT` > > + > > rm -f $seqres.full > > ... to clear out the old .full file. > > > +_scratch_mkfs MKFS_OPTIONS >> $seqres.full 2>&1 > > You don't need MKFS_OPTIONS here. In fact, this currently causes a mkfs > failure (missing $) that we don't detect because we aren't checking that > the mkfs actually succeeds. All we need to do here is: > > _scratch_mkfs >> $seqres.full 2>&1 || _fail "mkfs failed" > > If you dig down into _scratch_mkfs(), you'll see that it already > includes $MKFS_OPTIONS and thus formats the fs as specified by the test > config. > > It might also be a good idea to check that the _scratch_mount below > succeeds as well... > > > +_scratch_mount >> $seqres.full 2>&1 > > +length=$(($BLOCKS * $BSIZE)) > > + > > +# Write file > > +$XFS_IO_PROG -f -c "pwrite 0 $length" -c fsync $src > /dev/null > > +cp $src $dest > > + > > It seems quite unlikely for this to not create a single extent given the > smallish file size and freshly created fs, but who knows with various fs > types, test configurations, test device sizes, etc. Another option could > be to check the starting extent count and verify the ending extent count > matches, rather than assume hardcoded values of 1. > > To be honest, even just including the starting extent count in the > golden output (e.g., add an fiemap command here as well) might be good > enough to distinguish that failure path from something going wrong in > the collapse path, should it ever occur. > > > +# Insert alternate blocks > > +for (( j=0; j < $(($BLOCKS/2)); j++ )); do > > + offset=$((($j*$BSIZE)*2)) > > + $XFS_IO_PROG -c "finsert $offset $BSIZE" $dest > /dev/null > > +done > > + > > +# Check if 100 extents are present > > +$XFS_IO_PROG -c "fiemap -v" $dest | grep "^ *[0-9]*:" |wc -l > > + > > +_check_scratch_fs > > +if [ $? -ne 0 ]; then > > + status=1 > > + exit > > +fi > > + > > +# Collapse alternate blocks > > +for (( j=0; j < $(($BLOCKS/2)); j++ )); do > > + offset=$((($j*$BSIZE))) > > + $XFS_IO_PROG -c "fcollapse $offset $BSIZE" $dest > /dev/null > > +done > > + > > +# Check if 1 extents are present > > +$XFS_IO_PROG -c "fiemap -v" $dest | grep "^ *[0-9]*:" |wc -l > > + > > +# compare original file and test file. > > +cmp $src $dest || _fail "file bytes check failed" > > + > > +_check_scratch_fs > > +if [ $? -ne 0 ]; then > > + status=1 > > + exit > > +fi > > + > > +umount $SCRATCH_MNT > > + > > The scratch device is unmounted and checked after each test that uses it > so the above is unnecessary. I checked your each review points. and updated patch. Could you please review below patch ? Thanks a lot! ------------------------------------------------------------------------------ Subject: [PATCH v9 9/11] xfstests: generic/043: Test multiple fallocate insert/collapse range calls This testcase(043) tries to test finsert range a single alternate block mulitiple times and test merge code of collase range. Signed-off-by: Namjae Jeon Signed-off-by: Ashish Sangwan --- tests/generic/043 | 96 +++++++++++++++++++++++++++++++++++++++++++++++++ tests/generic/043.out | 2 + tests/generic/group | 1 + 3 files changed, 99 insertions(+), 0 deletions(-) create mode 100644 tests/generic/043 create mode 100644 tests/generic/043.out diff --git a/tests/generic/043 b/tests/generic/043 new file mode 100644 index 0000000..a6f91ce --- /dev/null +++ b/tests/generic/043 @@ -0,0 +1,96 @@ +#! /bin/bash +# FS QA Test No. generic/043 +# +# Test multiple fallocate insert/collapse range calls on same file. +# Call insert range a single alternate block multiple times until the file +# is left with 100 extents and as much number of extents. And Call collapse +# range about the previously inserted ranges to test merge code of collapse +# range. Also check for data integrity and file system consistency. +#----------------------------------------------------------------------- +# Copyright (c) 2014 Samsung Electronics. 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 "rm -f $tmp.*; exit \$status" 0 1 2 3 15 + +# get standard environment, filters and checks +. ./common/rc +. ./common/filter + +# real QA test starts here +_supported_fs generic +_supported_os Linux + +_require_scratch +_require_xfs_io_command "fiemap" +_require_xfs_io_command "finsert" +_require_xfs_io_command "fcollapse" +_do_die_on_error=always +src=$SCRATCH_MNT/testfile +dest=$SCRATCH_MNT/testfile.dest +BLOCKS=100 +BSIZE=`get_block_size $SCRATCH_MNT` +rm -f $seqres.full + +_scratch_mkfs >> $seqres.full 2>&1 || _fail "mkfs failed" +_scratch_mount || _fail "mount failed" +length=$(($BLOCKS * $BSIZE)) + +# Write file +_do "$XFS_IO_PROG -f -c \"pwrite 0 $length\" -c fsync $src" +cp $src $dest +extent_before=`$XFS_IO_PROG -c "fiemap -v" $dest | grep "^ *[0-9]*:" |wc -l` + +# Insert alternate blocks +for (( j=0; j < $(($BLOCKS/2)); j++ )); do + offset=$((($j*$BSIZE)*2)) + _do "$XFS_IO_PROG -c \"finsert $offset $BSIZE\" $dest" +done + +# Check if 100 extents are present +$XFS_IO_PROG -c "fiemap -v" $dest | grep "^ *[0-9]*:" |wc -l + +_check_scratch_fs +if [ $? -ne 0 ]; then + status=1 + exit +fi + +# Collapse alternate blocks +for (( j=0; j < $(($BLOCKS/2)); j++ )); do + offset=$((($j*$BSIZE))) + _do "$XFS_IO_PROG -c \"fcollapse $offset $BSIZE\" $dest" +done + +extent_after=`$XFS_IO_PROG -c "fiemap -v" $dest | grep "^ *[0-9]*:" |wc -l` +if [ $extent_before -ne $extent_after ]; then + echo "extents mismatched before = $extent_before after = $extent_after" +fi + +# compare original file and test file. +cmp $src $dest || _fail "file bytes check failed" + +# success, all done +status=0 +exit diff --git a/tests/generic/043.out b/tests/generic/043.out new file mode 100644 index 0000000..28427c5 --- /dev/null +++ b/tests/generic/043.out @@ -0,0 +1,2 @@ +QA output created by 043 +100 diff --git a/tests/generic/group b/tests/generic/group index c0944b3..0a10bdd 100644 --- a/tests/generic/group +++ b/tests/generic/group @@ -45,6 +45,7 @@ 040 auto quick prealloc 041 auto quick prealloc 042 auto quick prealloc +043 auto quick prealloc 053 acl repair auto quick 062 attr udf auto quick 068 other auto freeze dangerous stress -- 1.7.7