fstests: generic: splitted large dio write could trigger assertion on btrfs
diff mbox

Message ID 1484259726-3561-1-git-send-email-bo.li.liu@oracle.com
State Not Applicable
Headers show

Commit Message

Liu Bo Jan. 12, 2017, 10:22 p.m. UTC
On btrfs, if a large dio write (>=128MB) got splitted, the outstanding_extents
assertion would complain.  Note that CONFIG_BTRFS_ASSERT is required.

Regression test for
  Btrfs: adjust outstanding_extents counter properly when dio write is split

Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
---
I didn't figure out how to check if CONFIG_BTRFS_ASSERT is enabled, but since
there is no btrfs specific stuff in the test case, it might be better to not
have such a _require check b/c it doesn't make sense to other FS.

 tests/generic/392     | 75 +++++++++++++++++++++++++++++++++++++++++++++++++++
 tests/generic/392.out |  2 ++
 tests/generic/group   |  1 +
 3 files changed, 78 insertions(+)
 create mode 100755 tests/generic/392
 create mode 100644 tests/generic/392.out

Comments

Eryu Guan Jan. 16, 2017, 4:30 a.m. UTC | #1
On Thu, Jan 12, 2017 at 02:22:06PM -0800, Liu Bo wrote:
> On btrfs, if a large dio write (>=128MB) got splitted, the outstanding_extents
> assertion would complain.  Note that CONFIG_BTRFS_ASSERT is required.
> 
> Regression test for
>   Btrfs: adjust outstanding_extents counter properly when dio write is split
> 
> Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
> ---
> I didn't figure out how to check if CONFIG_BTRFS_ASSERT is enabled, but since
> there is no btrfs specific stuff in the test case, it might be better to not
> have such a _require check b/c it doesn't make sense to other FS.
> 
>  tests/generic/392     | 75 +++++++++++++++++++++++++++++++++++++++++++++++++++
>  tests/generic/392.out |  2 ++
>  tests/generic/group   |  1 +
>  3 files changed, 78 insertions(+)
>  create mode 100755 tests/generic/392
>  create mode 100644 tests/generic/392.out

There're over 400 generic tests now, seems your repo hasn't been updated
for a long time :)

> 
> diff --git a/tests/generic/392 b/tests/generic/392
> new file mode 100755
> index 0000000..4d88c44
> --- /dev/null
> +++ b/tests/generic/392
> @@ -0,0 +1,75 @@
> +#! /bin/bash
> +# FS QA Test generic/392
> +#
> +# If a larger dio write (size >= 128M) got splitted, the assertion in endio
> +# would complain (CONFIG_BTRFS_ASSERT is required).
> +#
> +# Regression test for
> +#   Btrfs: adjust outstanding_extents counter properly when dio write is split
> +#
> +#-----------------------------------------------------------------------
> +# Copyright (c) 2017 Liu Bo.  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.*
> +}
> +
> +# get standard environment, filters and checks
> +. ./common/rc
> +. ./common/filter
> +
> +# remove previous $seqres.full before test
> +rm -f $seqres.full
> +
> +# real QA test starts here
> +
> +# Modify as appropriate.
> +_supported_fs generic
> +_supported_os Linux
> +_require_scratch
> +_require_odirect
> +
> +# 2G / 1K
> +fsblock=$((1 << 21))
> +fssize=$((1 << 31))
> +_require_fs_space $SCRATCH_MNT $fsblock

You should mkfs & mount $SCRATCH_DEV first before _require_fs_space,
otherwise it's testing against the wrong fs.

> +
> +_scratch_mkfs_sized $fssize >> $seqres.full 2>&1

_scratch_mkfs_sized also make sure SCRATCH_DEV is big enough to make the
filesystem, in this test _require_fs_space and _scratch_mkfs_sized both
could do the work, only one is needed.

But not all filesystems have _scratch_mkfs_sized support, I'd prefer
using _require_fs_space.

Thanks,
Eryu

> +_scratch_mount >> $seqres.full 2>&1
> +
> +echo "Silence is golden"
> +
> +blocksize=$(( (128 + 1) * 2 * 1024 * 1024))
> +$XFS_IO_PROG -f -d -c "pwrite -b ${blocksize} 0 ${blocksize}" $SCRATCH_MNT/testfile.$seq >> $seqres.full 2>&1
> +
> +_scratch_unmount
> +
> +# success, all done
> +status=0
> +exit
> diff --git a/tests/generic/392.out b/tests/generic/392.out
> new file mode 100644
> index 0000000..665233c
> --- /dev/null
> +++ b/tests/generic/392.out
> @@ -0,0 +1,2 @@
> +QA output created by 392
> +Silence is golden
> diff --git a/tests/generic/group b/tests/generic/group
> index 2c16bd1..1631933 100644
> --- a/tests/generic/group
> +++ b/tests/generic/group
> @@ -394,3 +394,4 @@
>  389 auto quick acl
>  390 auto freeze stress dangerous
>  391 auto quick rw
> +392 auto quick dangerous
> -- 
> 2.5.0
> 
> --
> 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
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Liu Bo Jan. 25, 2017, 10:16 p.m. UTC | #2
On Mon, Jan 16, 2017 at 12:30:59PM +0800, Eryu Guan wrote:
> On Thu, Jan 12, 2017 at 02:22:06PM -0800, Liu Bo wrote:
> > On btrfs, if a large dio write (>=128MB) got splitted, the outstanding_extents
> > assertion would complain.  Note that CONFIG_BTRFS_ASSERT is required.
> > 
> > Regression test for
> >   Btrfs: adjust outstanding_extents counter properly when dio write is split
> > 
> > Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
> > ---
> > I didn't figure out how to check if CONFIG_BTRFS_ASSERT is enabled, but since
> > there is no btrfs specific stuff in the test case, it might be better to not
> > have such a _require check b/c it doesn't make sense to other FS.
> > 
> >  tests/generic/392     | 75 +++++++++++++++++++++++++++++++++++++++++++++++++++
> >  tests/generic/392.out |  2 ++
> >  tests/generic/group   |  1 +
> >  3 files changed, 78 insertions(+)
> >  create mode 100755 tests/generic/392
> >  create mode 100644 tests/generic/392.out
> 
> There're over 400 generic tests now, seems your repo hasn't been updated
> for a long time :)
> 

(sorry for the late reply.)

OK :p

> > 
> > diff --git a/tests/generic/392 b/tests/generic/392
> > new file mode 100755
> > index 0000000..4d88c44
> > --- /dev/null
> > +++ b/tests/generic/392
> > @@ -0,0 +1,75 @@
> > +#! /bin/bash
> > +# FS QA Test generic/392
> > +#
> > +# If a larger dio write (size >= 128M) got splitted, the assertion in endio
> > +# would complain (CONFIG_BTRFS_ASSERT is required).
> > +#
> > +# Regression test for
> > +#   Btrfs: adjust outstanding_extents counter properly when dio write is split
> > +#
> > +#-----------------------------------------------------------------------
> > +# Copyright (c) 2017 Liu Bo.  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.*
> > +}
> > +
> > +# get standard environment, filters and checks
> > +. ./common/rc
> > +. ./common/filter
> > +
> > +# remove previous $seqres.full before test
> > +rm -f $seqres.full
> > +
> > +# real QA test starts here
> > +
> > +# Modify as appropriate.
> > +_supported_fs generic
> > +_supported_os Linux
> > +_require_scratch
> > +_require_odirect
> > +
> > +# 2G / 1K
> > +fsblock=$((1 << 21))
> > +fssize=$((1 << 31))
> > +_require_fs_space $SCRATCH_MNT $fsblock
> 
> You should mkfs & mount $SCRATCH_DEV first before _require_fs_space,
> otherwise it's testing against the wrong fs.
> 
> > +
> > +_scratch_mkfs_sized $fssize >> $seqres.full 2>&1
> 
> _scratch_mkfs_sized also make sure SCRATCH_DEV is big enough to make the
> filesystem, in this test _require_fs_space and _scratch_mkfs_sized both
> could do the work, only one is needed.
> 
> But not all filesystems have _scratch_mkfs_sized support, I'd prefer
> using _require_fs_space.
> 

Make sense, thanks for the comments.

Thanks,

-liubo

> Thanks,
> Eryu
> 
> > +_scratch_mount >> $seqres.full 2>&1
> > +
> > +echo "Silence is golden"
> > +
> > +blocksize=$(( (128 + 1) * 2 * 1024 * 1024))
> > +$XFS_IO_PROG -f -d -c "pwrite -b ${blocksize} 0 ${blocksize}" $SCRATCH_MNT/testfile.$seq >> $seqres.full 2>&1
> > +
> > +_scratch_unmount
> > +
> > +# success, all done
> > +status=0
> > +exit
> > diff --git a/tests/generic/392.out b/tests/generic/392.out
> > new file mode 100644
> > index 0000000..665233c
> > --- /dev/null
> > +++ b/tests/generic/392.out
> > @@ -0,0 +1,2 @@
> > +QA output created by 392
> > +Silence is golden
> > diff --git a/tests/generic/group b/tests/generic/group
> > index 2c16bd1..1631933 100644
> > --- a/tests/generic/group
> > +++ b/tests/generic/group
> > @@ -394,3 +394,4 @@
> >  389 auto quick acl
> >  390 auto freeze stress dangerous
> >  391 auto quick rw
> > +392 auto quick dangerous
> > -- 
> > 2.5.0
> > 
> > --
> > 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
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Patch
diff mbox

diff --git a/tests/generic/392 b/tests/generic/392
new file mode 100755
index 0000000..4d88c44
--- /dev/null
+++ b/tests/generic/392
@@ -0,0 +1,75 @@ 
+#! /bin/bash
+# FS QA Test generic/392
+#
+# If a larger dio write (size >= 128M) got splitted, the assertion in endio
+# would complain (CONFIG_BTRFS_ASSERT is required).
+#
+# Regression test for
+#   Btrfs: adjust outstanding_extents counter properly when dio write is split
+#
+#-----------------------------------------------------------------------
+# Copyright (c) 2017 Liu Bo.  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.*
+}
+
+# get standard environment, filters and checks
+. ./common/rc
+. ./common/filter
+
+# remove previous $seqres.full before test
+rm -f $seqres.full
+
+# real QA test starts here
+
+# Modify as appropriate.
+_supported_fs generic
+_supported_os Linux
+_require_scratch
+_require_odirect
+
+# 2G / 1K
+fsblock=$((1 << 21))
+fssize=$((1 << 31))
+_require_fs_space $SCRATCH_MNT $fsblock
+
+_scratch_mkfs_sized $fssize >> $seqres.full 2>&1
+_scratch_mount >> $seqres.full 2>&1
+
+echo "Silence is golden"
+
+blocksize=$(( (128 + 1) * 2 * 1024 * 1024))
+$XFS_IO_PROG -f -d -c "pwrite -b ${blocksize} 0 ${blocksize}" $SCRATCH_MNT/testfile.$seq >> $seqres.full 2>&1
+
+_scratch_unmount
+
+# success, all done
+status=0
+exit
diff --git a/tests/generic/392.out b/tests/generic/392.out
new file mode 100644
index 0000000..665233c
--- /dev/null
+++ b/tests/generic/392.out
@@ -0,0 +1,2 @@ 
+QA output created by 392
+Silence is golden
diff --git a/tests/generic/group b/tests/generic/group
index 2c16bd1..1631933 100644
--- a/tests/generic/group
+++ b/tests/generic/group
@@ -394,3 +394,4 @@ 
 389 auto quick acl
 390 auto freeze stress dangerous
 391 auto quick rw
+392 auto quick dangerous