generic: concurrent IO test with mixed IO types
diff mbox

Message ID 1433760101-25540-1-git-send-email-eguan@redhat.com
State New
Headers show

Commit Message

Eryu Guan June 8, 2015, 10:41 a.m. UTC
Test concurrent buffered I/O, DIO, AIO, mmap I/O and splice I/O on the
same files.

Signed-off-by: Eryu Guan <eguan@redhat.com>
---

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

Comments

Lukas Czerner June 8, 2015, 11:02 a.m. UTC | #1
On Mon, 8 Jun 2015, Eryu Guan wrote:

> Date: Mon,  8 Jun 2015 18:41:41 +0800
> From: Eryu Guan <eguan@redhat.com>
> To: fstests@vger.kernel.org
> Cc: lczerner@redhat.com, Eryu Guan <eguan@redhat.com>
> 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@redhat.com>
> ---
> 
> 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.

> +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 ?

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@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Eryu Guan June 8, 2015, 11:59 a.m. UTC | #2
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@redhat.com>
> > To: fstests@vger.kernel.org
> > Cc: lczerner@redhat.com, Eryu Guan <eguan@redhat.com>
> > 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@redhat.com>
> > ---
> > 
> > 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@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Lukas Czerner June 8, 2015, 12:36 p.m. UTC | #3
On Mon, 8 Jun 2015, Eryu Guan wrote:

> Date: Mon, 8 Jun 2015 19:59:47 +0800
> From: Eryu Guan <eguan@redhat.com>
> To: Lukáš Czerner <lczerner@redhat.com>
> Cc: fstests@vger.kernel.org
> Subject: Re: [PATCH] generic: concurrent IO test with mixed IO types
> 
> 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@redhat.com>
> > > To: fstests@vger.kernel.org
> > > Cc: lczerner@redhat.com, Eryu Guan <eguan@redhat.com>
> > > 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@redhat.com>
> > > ---
> > > 
> > > 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

Ah right :) I missed the quotes, sorry.

Thanks!
-Lukas


> 
> > 
> > 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
> > > 
>
Dave Chinner June 9, 2015, 10:27 p.m. UTC | #4
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@redhat.com> To: fstests@vger.kernel.org Cc:
> > lczerner@redhat.com, Eryu Guan <eguan@redhat.com> 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.
....
> > + +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.

Not better at all - this reduces the easy of debugging problems.
e.g.  if a test is hung, I know where it's temporary output files
are just by looking at the process listing and grabbing the pid of
the test process being run. IOWs, we *want* deterministic temporary
filenames here - obscuring the temporary file names that tests use
doesn't gain us anything.

Yes, I see that some btrfs tests have used mktemp  and a couple of
generic tests use $TEST_DIR/$$ - they should be reverted to /tmp/$$
so they are consistent with the majority of tests and the new test
template....

Cheers,

Dave.

Patch
diff mbox

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/$$
+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
+
+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