On Mon, Jun 08, 2015 at 01:02:48PM +0200, Lukáš Czerner wrote:
> On Mon, 8 Jun 2015, Eryu Guan wrote:
>
> > Date: Mon, 8 Jun 2015 18:41:41 +0800
> > From: Eryu Guan <eguan@xxxxxxxxxx>
> > To: fstests@xxxxxxxxxxxxxxx
> > Cc: lczerner@xxxxxxxxxx, Eryu Guan <eguan@xxxxxxxxxx>
> > Subject: [PATCH] generic: concurrent IO test with mixed IO types
> >
> > Test concurrent buffered I/O, DIO, AIO, mmap I/O and splice I/O on the
> > same files.
>
> Hi Eryu,
>
> Great! Thanks for turning it into a test. Couple of comments below.
>
> >
> > Signed-off-by: Eryu Guan <eguan@xxxxxxxxxx>
> > ---
> >
> > This fio job file has been proven to be potent, it triggers WARNINGs on ext4
> > and xfs with 4.1-rc6 kernel.
> >
> > ext4: WARNING: at fs/ext4/inode.c:1328
> > xfs: WARNING: CPU: 7 PID: 3090 at fs/xfs/xfs_file.c:726 xfs_file_dio_aio_write+0x176/0x2a8 [xfs]()
> >
> > The ext4 issue should be fixed by Lukas's patch
> > ext4: fix reservation release on invalidatepage for delalloc fs
> >
> > And it ever paniced kernel in mm code and hung xfs.
> >
> > I reduced the numjobs and iodepth to reduce the test time(~25s on my test host)
> > and scale them by $LOAD_FACTOR. And it still could trigger the warning on ext4
> > and xfs with reduced workload.
> >
> > tests/generic/090 | 121 ++++++++++++++++++++++++++++++++++++++++++++++++++
> > tests/generic/090.out | 2 +
> > tests/generic/group | 1 +
> > 3 files changed, 124 insertions(+)
> > create mode 100755 tests/generic/090
> > create mode 100644 tests/generic/090.out
> >
> > diff --git a/tests/generic/090 b/tests/generic/090
> > new file mode 100755
> > index 0000000..748e935
> > --- /dev/null
> > +++ b/tests/generic/090
> > @@ -0,0 +1,121 @@
> > +#! /bin/bash
> > +# FS QA Test generic/090
> > +#
> > +# Concurrent mixed I/O (buffer I/O, aiodio, mmap, splice) on the same files
> > +#
> > +#-----------------------------------------------------------------------
> > +# Copyright (c) 2015 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/$$
>
> Not a big deal, but using mktemp woud probably be better.
Will do in v2.
>
> > +fio_config=$tmp.fio
> > +status=1 # failure is the default!
> > +trap "_cleanup; exit \$status" 0 1 2 3 15
> > +
> > +_cleanup()
> > +{
> > + cd /
> > + rm -f $tmp.*
> > +}
> > +
> > +# get standard environment, filters and checks
> > +. ./common/rc
> > +. ./common/filter
> > +
> > +# real QA test starts here
> > +_supported_fs generic
> > +_supported_os Linux
> > +_require_scratch
> > +
> > +iodepth=$((16 * LOAD_FACTOR))
> > +iodepth_batch=$((8 * LOAD_FACTOR))
> > +numjobs=$((5 * LOAD_FACTOR))
> > +cat >$fio_config <<EOF
> > +[global]
> > +bs=8k
> > +iodepth=$iodepth
> > +iodepth_batch=$iodepth_batch
> > +randrepeat=1
> > +size=1m
> > +directory=$SCRATCH_MNT
> > +numjobs=$numjobs
> > +[job1]
> > +ioengine=sync
> > +bs=1k
> > +direct=1
> > +rw=randread
> > +filename=file1:file2
> > +[job2]
> > +ioengine=libaio
> > +rw=randwrite
> > +direct=1
> > +filename=file1:file2
> > +[job3]
> > +bs=1k
> > +ioengine=posixaio
> > +rw=randwrite
> > +direct=1
> > +filename=file1:file2
> > +[job4]
> > +ioengine=splice
> > +direct=1
> > +rw=randwrite
> > +filename=file1:file2
> > +[job5]
> > +bs=1k
> > +ioengine=sync
> > +rw=randread
> > +filename=file1:file2
> > +[job6]
> > +ioengine=posixaio
> > +rw=randwrite
> > +filename=file1:file2
> > +[job7]
> > +ioengine=splice
> > +rw=randwrite
> > +filename=file1:file2
> > +[job8]
> > +ioengine=mmap
> > +rw=randwrite
> > +bs=1k
> > +filename=file1:file2
> > +[job9]
> > +ioengine=mmap
> > +rw=randwrite
> > +direct=1
> > +filename=file1:file2
> > +EOF
> > +# with ioengine=mmap and direct=1, fio requires bs to be at least pagesize,
> > +# which is a fio built-in var.
> > +echo 'bs=$pagesize' >> $fio_config
>
> Can't this be part of the the fio script creation above ?
The 'here-document' replaces every var to its value, e.g. $numjobs => 10
(if LOAD_FACTOR is 1), so we don't have $pagesize set, and it'll be
empty in the fio config file, that's not what we want.
"bs=$pagesize" is literally what we want in the config file, $pagesize
is a special var for fio, it will be replaced by real pagesize
internally by fio.
cat >file <<'EOF'
text
EOF
This kind of 'here-document' ('' around EOF) won't do any variable
expansion, but we need $numjobs to be expanded. So they can't be merged
together.
Thanks,
Eryu
>
> Thanks!
> -Lukas
>
> > +
> > +rm -f $seqres.full
> > +
> > +_require_fio $fio_config
> > +
> > +echo "Silence is golden"
> > +$FIO_PROG $fio_config >>$seqres.full 2>&1
> > +
> > +# all done, expect no hang no oops no fs corruption,
> > +# _check_dmesg and _check_filesystems will do the check work for us
> > +status=0
> > +exit
> > diff --git a/tests/generic/090.out b/tests/generic/090.out
> > new file mode 100644
> > index 0000000..2b5100d
> > --- /dev/null
> > +++ b/tests/generic/090.out
> > @@ -0,0 +1,2 @@
> > +QA output created by 090
> > +Silence is golden
> > diff --git a/tests/generic/group b/tests/generic/group
> > index 0c8964c..2e534a5 100644
> > --- a/tests/generic/group
> > +++ b/tests/generic/group
> > @@ -92,6 +92,7 @@
> > 087 perms auto quick
> > 088 perms auto quick
> > 089 metadata auto
> > +090 auto rw stress
> > 091 rw auto quick
> > 092 auto quick prealloc
> > 093 attr cap udf auto
> >
--
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
