| Message ID | 20160808084814.GC27776@eguan.usersys.redhat.com (mailing list archive) |
|---|---|
| State | Not Applicable, archived |
| Headers | show |
On Mon, Aug 08, 2016 at 04:48:14PM +0800, Eryu Guan wrote: > On Thu, Jul 21, 2016 at 04:47:45PM -0700, Darrick J. Wong wrote: > > More tests for the reverse mapping functionality. > > > > Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com> > > --- > > tests/xfs/876 | 76 +++++++++++++++++++++++++++++++++++++++++++++++ > > tests/xfs/876.out | 4 ++ > > tests/xfs/877 | 85 +++++++++++++++++++++++++++++++++++++++++++++++++++++ > > tests/xfs/877.out | 10 ++++++ > > tests/xfs/group | 4 ++ > > 5 files changed, 178 insertions(+), 1 deletion(-) > > create mode 100755 tests/xfs/876 > > create mode 100644 tests/xfs/876.out > > create mode 100755 tests/xfs/877 > > create mode 100644 tests/xfs/877.out > > > > > > diff --git a/tests/xfs/876 b/tests/xfs/876 > > new file mode 100755 > > index 0000000..cf73d40 > > --- /dev/null > > +++ b/tests/xfs/876 > > @@ -0,0 +1,76 @@ > > +#! /bin/bash > > +# FS QA Test No. 876 > > +# > > +# Create a big enough rmapbt that we tickle a fdblocks accounting bug. > > This test failed post-test fs check as well > > "sb_fdblocks 3912360, counted 3912522" > > Seems like it's a known bug? (Yes, it should be fixed in the -wtf branch, as noted in a previous reply.) > > +# > > +#----------------------------------------------------------------------- > > +# Copyright (c) 2016, 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 "$tmp".* > > +} > > + > > +# get standard environment, filters and checks > > +. ./common/rc > > +. ./common/filter > > +. ./common/attr > > Seems this is not needed. > > > +. ./common/reflink > > + > > +# real QA test starts here > > +_supported_os Linux > > +_require_scratch > > +_require_xfs_scratch_rmapbt > > +_require_scratch_reflink > > +_require_test_program "punch-alternating" > > + > > +rm -f "$seqres.full" > > + > > +echo "+ create scratch fs" > > +_scratch_mkfs > "$seqres.full" 2>&1 > > + > > +echo "+ mount fs image" > > +_scratch_mount > > +blksz="$(stat -f $SCRATCH_MNT -c '%S')" > > +isize=$(xfs_info $SCRATCH_MNT | grep isize | sed -e 's/^.*isize=\([0-9]*\).*$/\1/g') > > We usually get isize from the output of _filter_mkfs, e.g. > > _scratch_mkfs | _filter_mkfs 2>$tmp.mkfs >/dev/null > . $tmp.mkfs > > Then all mkfs info is set accordingly, including $isize > > But in this test, isize is got but not used. Remove the line? Ok. > > + > > +bt_ptrs=$(( (blksz - 56) / 44 )) > > +bt_recs=$(( (blksz - 56) / 24 )) > > + > > +blocks=$((bt_ptrs * bt_recs + 1)) > > +_require_fs_space $SCRATCH_MNT $(( (2 * blocks * blksz) * 5 / 4096 )) > > +len=$((blocks * blksz)) > > + > > +echo "+ make some files" > > +$XFS_IO_PROG -f -c "falloc 0 $len" -c "pwrite -S 0x68 -b 1048576 0 $len" $SCRATCH_MNT/f1 >> $seqres.full > > +$XFS_IO_PROG -f -c "falloc 0 $len" -c "pwrite -S 0x68 -b 1048576 0 $len" $SCRATCH_MNT/f2 >> $seqres.full > > +./src/punch-alternating $SCRATCH_MNT/f1 >> "$seqres.full" > > +./src/punch-alternating $SCRATCH_MNT/f2 >> "$seqres.full" > > + > > +# success, all done > > +status=0 > > +exit > > diff --git a/tests/xfs/876.out b/tests/xfs/876.out > > new file mode 100644 > > index 0000000..96c1970 > > --- /dev/null > > +++ b/tests/xfs/876.out > > @@ -0,0 +1,4 @@ > > +QA output created by 876 > > ++ create scratch fs > > ++ mount fs image > > ++ make some files > > diff --git a/tests/xfs/877 b/tests/xfs/877 > > new file mode 100755 > > index 0000000..9700263 > > --- /dev/null > > +++ b/tests/xfs/877 > > @@ -0,0 +1,85 @@ > > +#! /bin/bash > > +# FS QA Test No. 877 > > +# > > +# Make sure query_range returns -EINVAL if lowkey > highkey. > > This test failed too, is this expected (again, tested with your > djwong-experimental branck kernel and xfsprogs)? > > [root@dhcp-66-86-11 xfstests]# diff -u tests/xfs/877.out /root/workspace/xfstests/results//xfs_reflink/xfs/877.out.bad > --- tests/xfs/877.out 2016-08-08 16:22:35.151000000 +0800 > +++ /root/workspace/xfstests/results//xfs_reflink/xfs/877.out.bad 2016-08-08 16:23:12.405000000 +0800 > @@ -7,4 +7,3 @@ > Unmount > Try a regular fsmap > Try a bad fsmap > -Error 22 while querying fsmap btree. During the code review of the XFS function that underlies GETFSMAP (query_range), someone suggested that it should return -EINVAL if someone queries start > end rather than simply returning no results, so this is the xfstest to ensure that happens. (It's in the -wtf branch.) > > +# > > +#----------------------------------------------------------------------- > > +# Copyright (c) 2016, 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".* $metadump_file > > +} > > + > > +# get standard environment, filters and checks > > +. ./common/rc > > +. ./common/filter > > +. ./common/attr > > Seems this is not needed. > > > + > > +# real QA test starts here > > +_supported_os Linux > > _supported_fs xfs > > > +_require_xfs_scratch_rmapbt > > _require_xfs_io_command commands used in this test? e.g. falloc, fpunch, fcollapse and finsert Ok. > And '_require_command "$XFS_DB_PROG" "xfs_db"' I'm assuming this is a new requirement to declare explicitly that we're going to use xfs_db? All of the existing tests/xfs/ tests assume that XFS_DB_PROG is always set. > > > + > > +rm -f "$seqres.full" > > + > > +echo "Format and mount" > > +_scratch_mkfs > "$seqres.full" 2>&1 > > +_scratch_mount > > +blksz=65536 > > +blocks=16 > > +len=$((blocks * blksz)) > > + > > +echo "Create some files" > > +$XFS_IO_PROG -f -c "falloc 0 $len" -c "pwrite -S 0x68 -b 1048576 0 $len" $SCRATCH_MNT/f1 >> $seqres.full > > + > > +echo "Manipulate file" > > +$XFS_IO_PROG -c "fpunch $blksz $blksz" \ > > + -c "fzero $((3 * blksz)) $blksz" \ > > + -c "pwrite -S 0x69 $((5 * blksz)) $blksz" \ > > + -c "fpunch $((7 * blksz)) $blksz" \ > > + -c "fsync" \ > > + -c "pwrite -S 0x70 $((7 * blksz)) $blksz" \ > > + -c "fcollapse $((9 * blksz)) $blksz" \ > > + -c "finsert $((10 * blksz)) $blksz" $SCRATCH_MNT/f1 >> $seqres.full > > + > > +echo "Check file" > > +md5sum $SCRATCH_MNT/f1 | _filter_scratch > > +od -tx1 -Ad -c $SCRATCH_MNT/f1 >> $seqres.full > > + > > +echo "Unmount" > > +_scratch_unmount > > + > > +echo "Try a regular fsmap" > > +$XFS_DB_PROG -c 'fsmap' $SCRATCH_DEV >> $seqres.full > > +$XFS_DB_PROG -c 'fsmap 0 5' $SCRATCH_DEV >> $seqres.full > > + > > +echo "Try a bad fsmap" > > +$XFS_DB_PROG -c 'fsmap 5 4' $SCRATCH_DEV > > + > > +# success, all done > > +status=0 > > +exit > > diff --git a/tests/xfs/877.out b/tests/xfs/877.out > > new file mode 100644 > > index 0000000..4a2242f > > --- /dev/null > > +++ b/tests/xfs/877.out > > @@ -0,0 +1,10 @@ > > +QA output created by 877 > > +Format and mount > > +Create some files > > +Manipulate file > > +Check file > > +e45c5707fcf6817e914ffb6ce37a0ac7 SCRATCH_MNT/f1 > > +Unmount > > +Try a regular fsmap > > +Try a bad fsmap > > +Error 22 while querying fsmap btree. > > diff --git a/tests/xfs/group b/tests/xfs/group > > index 16f7947..f81f899 100644 > > --- a/tests/xfs/group > > +++ b/tests/xfs/group > > @@ -3,8 +3,8 @@ > > 003 db auto quick > > 004 db auto quick > > 005 auto quick > > -007 auto quota quick > > 006 auto quick mount > > +007 auto quota quick > > This is already fixed in upstream, not needed. Ok. It'll probably fix itself when I rebase. --D > > Thanks, > Eryu > > > 008 rw ioctl auto quick > > 009 rw ioctl auto prealloc quick > > 010 auto quick repair > > @@ -309,3 +309,5 @@ > > 873 auto quick clone fsr > > 874 auto quick clone fsr > > 875 auto quick clone fsr quota > > +876 auto quick rmap clone > > +877 auto quick rmap clone > >
--- tests/xfs/877.out 2016-08-08 16:22:35.151000000 +0800 +++ /root/workspace/xfstests/results//xfs_reflink/xfs/877.out.bad 2016-08-08 16:23:12.405000000 +0800 @@ -7,4 +7,3 @@ Unmount Try a regular fsmap Try a bad fsmap -Error 22 while querying fsmap btree.