On Wed, Jun 20, 2018 at 09:30:01AM -0700, Darrick J. Wong wrote:
> On Wed, Jun 20, 2018 at 04:41:14PM +0800, Zorro Lang wrote:
> > Run several duperemove processes with fsstress on same directory at
> > same time. Make sure the race won't break the fs or kernel.
> >
> > Signed-off-by: Zorro Lang <zlang@xxxxxxxxxx>
> > ---
> >
> > V2 did below changes:
> > 1) do sleep 1 after kill processes
> > 2) change SLEEP_TIME to sleep_time
> > 3) add the case to stress group
> >
> > Thanks,
> > Zorro
> >
> > tests/shared/010 | 111 +++++++++++++++++++++++++++++++++++++++++++++++++++
> > tests/shared/010.out | 2 +
> > tests/shared/group | 1 +
> > 3 files changed, 114 insertions(+)
> > create mode 100755 tests/shared/010
> > create mode 100644 tests/shared/010.out
> >
> > diff --git a/tests/shared/010 b/tests/shared/010
> > new file mode 100755
> > index 00000000..c449c247
> > --- /dev/null
> > +++ b/tests/shared/010
> > @@ -0,0 +1,111 @@
> > +#! /bin/bash
> > +# FS QA Test 010
> > +#
> > +# Dedup & random I/O race test, do multi-threads fsstress and dedupe on
> > +# same directory/files
> > +#
> > +#-----------------------------------------------------------------------
> > +# Copyright (c) 2018 Red Hat Inc. 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.*
> > + kill_all_stress
> > +}
> > +
> > +# get standard environment, filters and checks
> > +. ./common/rc
> > +. ./common/filter
> > +. ./common/reflink
> > +
> > +# remove previous $seqres.full before test
> > +rm -f $seqres.full
> > +
> > +# real QA test starts here
> > +
> > +# duperemove only supports btrfs and xfs (with reflink feature).
> > +# Add other filesystems if it supports more later.
> > +_supported_fs xfs btrfs
> > +_supported_os Linux
> > +_require_scratch_dedupe
> > +_require_command "$DUPEREMOVE_PROG" duperemove
> > +_require_command "$KILLALL_PROG" killall
> > +
> > +_scratch_mkfs > $seqres.full 2>&1
> > +_scratch_mount >> $seqres.full 2>&1
> > +
> > +function kill_all_stress()
> > +{
> > + local f=1
> > + local d=1
> > +
> > + # kill the bash process which loop run duperemove
> > + if [ -n "$loop_dedup_pid" ]; then
> > + kill $loop_dedup_pid > /dev/null 2>&1
> > + wait $loop_dedup_pid > /dev/null 2>&1
> > + loop_dedup_pid=""
> > + fi
> > +
> > + # Make sure all fsstress and duperemove processes get killed
> > + while [ $((f + d)) -ne 0 ]; do
> > + $KILLALL_PROG -q $FSSTRESS_PROG > /dev/null 2>&1
> > + $KILLALL_PROG -q $DUPEREMOVE_PROG > /dev/null 2>&1
> > + sleep 1
> > + f=`ps -eLf | grep $FSSTRESS_PROG | grep -v "grep" | wc -l`
> > + d=`ps -eLf | grep $DUPEREMOVE_PROG | grep -v "grep" | wc -l`
> > + done
> > +}
> > +
> > +sleep_time=$((50 * TIME_FACTOR))
> > +
> > +# Start fsstress
> > +fsstress_opts="-r -n 1000 -p $((5 * LOAD_FACTOR))"
> > +$FSSTRESS_PROG $fsstress_opts -d $SCRATCH_MNT -l 0 >> $seqres.full 2>&1 &
> > +loop_dedup_pid=""
> > +# Start several dedupe processes on same directory
> > +for ((i = 0; i < $((2 * LOAD_FACTOR)); i++)); do
> > + while true; do
> > + $DUPEREMOVE_PROG -dr --dedupe-options=same $SCRATCH_MNT/ \
> > + >>$seqres.full 2>&1
>
> /me wonders why not just touch $TEST_DIR/run, have this loop do:
>
> while test -e $TEST_DIR/run; do
> duperemove...
> done
>
> and then rm -f $TEST_DIR/run to end the loop? Then you don't need the
> complex machinery to shut down the bash loop and kill the fsstress and
> duperemove processes.
Yeah, this looks cleaner to me.
Zorro, could you please take a look and update the test as Darrick
suggested? Also, you could update the fstests repo and use new 'new'
script to generate new test template which has correct SPDX tag :)
>
> Otherwise looks decent,
Thanks for reviewing!
Eryu
>
> --D
>
> > + done &
> > + loop_dedup_pid="$! $loop_dedup_pid"
> > +done
> > +
> > +# End the test after $sleep_time seconds
> > +sleep $sleep_time
> > +kill_all_stress
> > +
> > +# umount and mount again, verify pagecache contents don't mutate and a fresh
> > +# read from the disk also doesn't show mutations.
> > +find $testdir -type f -exec md5sum {} \; > $TEST_DIR/${seq}md5.sum
> > +_scratch_cycle_mount
> > +md5sum -c --quiet $TEST_DIR/${seq}md5.sum
> > +
> > +echo "Silence is golden"
> > +status=0
> > +exit
> > diff --git a/tests/shared/010.out b/tests/shared/010.out
> > new file mode 100644
> > index 00000000..1d83a8d6
> > --- /dev/null
> > +++ b/tests/shared/010.out
> > @@ -0,0 +1,2 @@
> > +QA output created by 010
> > +Silence is golden
> > diff --git a/tests/shared/group b/tests/shared/group
> > index 9c484794..094da27d 100644
> > --- a/tests/shared/group
> > +++ b/tests/shared/group
> > @@ -12,6 +12,7 @@
> > 007 dangerous_fuzzers
> > 008 auto stress dedupe
> > 009 auto stress dedupe
> > +010 auto stress dedupe
> > 032 mkfs auto quick
> > 272 auto enospc rw
> > 289 auto quick
> > --
> > 2.14.4
> >
> > --
> > To unsubscribe from this list: send the line "unsubscribe fstests" in
> > the body of a message to majordomo@xxxxxxxxxxxxxxx
> > More majordomo info at http://vger.kernel.org/majordomo-info.html
> --
> To unsubscribe from this list: send the line "unsubscribe fstests" in
> the body of a message to majordomo@xxxxxxxxxxxxxxx
> More majordomo info at http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
