* [PATCH] generic: test Btrfs delalloc accounting overflow
@ 2017-06-02 8:23 Omar Sandoval
2017-06-02 9:07 ` Nikolay Borisov
2017-06-03 7:37 ` Christoph Hellwig
0 siblings, 2 replies; 8+ messages in thread
From: Omar Sandoval @ 2017-06-02 8:23 UTC (permalink / raw)
To: fstests, linux-btrfs; +Cc: kernel-team
From: Omar Sandoval <osandov@fb.com>
This is a regression test for "[PATCH] Btrfs: fix delalloc accounting
leak caused by u32 overflow". It creates a bunch of delalloc extents and
merges them together to make sure the accounting is done right.
Signed-off-by: Omar Sandoval <osandov@fb.com>
---
tests/generic/438 | 77 +++++++++++++++++++++++++++++++++++++++++++++++++++
tests/generic/438.out | 2 ++
tests/generic/group | 1 +
3 files changed, 80 insertions(+)
create mode 100755 tests/generic/438
create mode 100644 tests/generic/438.out
diff --git a/tests/generic/438 b/tests/generic/438
new file mode 100755
index 00000000..8edac03d
--- /dev/null
+++ b/tests/generic/438
@@ -0,0 +1,77 @@
+#! /bin/bash
+# FS QA Test 438
+#
+# Test delayed allocation with a large number of extents that are merged.
+# Regression test for patch "Btrfs: fix delalloc accounting leak caused by u32
+# overflow".
+#
+#-----------------------------------------------------------------------
+# Copyright (c) 2017 Facebook. 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
+
+test_file="$TEST_DIR/$seq"
+
+_cleanup()
+{
+ cd /
+ rm -f $tmp.* "$test_file"
+}
+
+. ./common/rc
+. ./common/filter
+
+rm -f $seqres.full
+
+_supported_fs generic
+_supported_os Linux
+_require_test
+
+# Create 32k extents. All of these extents will be accounted as outstanding and
+# reserved.
+for ((i = 0; i < 32 * 1024; i++)); do
+ xfs_io -f -c "pwrite $((2 * 4096 * i)) 4096" "$test_file" >>"$seqres.full"
+done
+
+# Fill in the gaps between the created extents. The outstanding extents will
+# all be merged into 1, but there will still be 32k reserved.
+for ((i = 0; i < 32 * 1024; i++)); do
+ xfs_io -f -c "pwrite $((2 * 4096 * i + 1)) 4096" "$test_file" >>"$seqres.full"
+done
+
+# Flush the delayed allocations.
+sync
+
+# Make sure that we didn't leak any metadata space.
+if [[ $FSTYP = btrfs ]]; then
+ uuid="$(findmnt -n -o UUID "$TEST_DIR")"
+ cd "/sys/fs/btrfs/$uuid/allocation"
+ echo "$(($(cat metadata/bytes_may_use) - $(cat global_rsv_reserved))) bytes leaked" | grep -v '^0 '
+fi
+
+echo "Silence is golden"
+
+status=0
+exit
diff --git a/tests/generic/438.out b/tests/generic/438.out
new file mode 100644
index 00000000..4968f4d7
--- /dev/null
+++ b/tests/generic/438.out
@@ -0,0 +1,2 @@
+QA output created by 438
+Silence is golden
diff --git a/tests/generic/group b/tests/generic/group
index 438c2990..c804b05d 100644
--- a/tests/generic/group
+++ b/tests/generic/group
@@ -440,3 +440,4 @@
435 auto encrypt
436 auto quick rw
437 auto quick
+438 auto
--
2.13.0
^ permalink raw reply related [flat|nested] 8+ messages in thread
* Re: [PATCH] generic: test Btrfs delalloc accounting overflow
2017-06-02 8:23 [PATCH] generic: test Btrfs delalloc accounting overflow Omar Sandoval
@ 2017-06-02 9:07 ` Nikolay Borisov
2017-06-02 12:46 ` David Sterba
2017-06-03 7:37 ` Christoph Hellwig
1 sibling, 1 reply; 8+ messages in thread
From: Nikolay Borisov @ 2017-06-02 9:07 UTC (permalink / raw)
To: Omar Sandoval, fstests, linux-btrfs; +Cc: kernel-team
On 2.06.2017 11:23, Omar Sandoval wrote:
> From: Omar Sandoval <osandov@fb.com>
>
> This is a regression test for "[PATCH] Btrfs: fix delalloc accounting
> leak caused by u32 overflow". It creates a bunch of delalloc extents and
> merges them together to make sure the accounting is done right.
>
> Signed-off-by: Omar Sandoval <osandov@fb.com>
> ---
> tests/generic/438 | 77 +++++++++++++++++++++++++++++++++++++++++++++++++++
> tests/generic/438.out | 2 ++
> tests/generic/group | 1 +
> 3 files changed, 80 insertions(+)
> create mode 100755 tests/generic/438
> create mode 100644 tests/generic/438.out
>
> diff --git a/tests/generic/438 b/tests/generic/438
> new file mode 100755
> index 00000000..8edac03d
> --- /dev/null
> +++ b/tests/generic/438
> @@ -0,0 +1,77 @@
> +#! /bin/bash
> +# FS QA Test 438
> +#
> +# Test delayed allocation with a large number of extents that are merged.
> +# Regression test for patch "Btrfs: fix delalloc accounting leak caused by u32
> +# overflow".
> +#
> +#-----------------------------------------------------------------------
> +# Copyright (c) 2017 Facebook. 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
> +
> +test_file="$TEST_DIR/$seq"
> +
> +_cleanup()
> +{
> + cd /
> + rm -f $tmp.* "$test_file"
> +}
> +
> +. ./common/rc
> +. ./common/filter
> +
> +rm -f $seqres.full
> +
> +_supported_fs generic
> +_supported_os Linux
> +_require_test
> +
> +# Create 32k extents. All of these extents will be accounted as outstanding and
> +# reserved.
> +for ((i = 0; i < 32 * 1024; i++)); do
> + xfs_io -f -c "pwrite $((2 * 4096 * i)) 4096" "$test_file" >>"$seqres.full"
> +done
> +
> +# Fill in the gaps between the created extents. The outstanding extents will
> +# all be merged into 1, but there will still be 32k reserved.
> +for ((i = 0; i < 32 * 1024; i++)); do
> + xfs_io -f -c "pwrite $((2 * 4096 * i + 1)) 4096" "$test_file" >>"$seqres.full"
> +done
> +
> +# Flush the delayed allocations.
> +sync
> +
> +# Make sure that we didn't leak any metadata space.
> +if [[ $FSTYP = btrfs ]]; then
> + uuid="$(findmnt -n -o UUID "$TEST_DIR")"
if we are on btrfs and we don't have findmnt this test will likely fail.
Perhaps include a _require_command findmnt
> + cd "/sys/fs/btrfs/$uuid/allocation"
> + echo "$(($(cat metadata/bytes_may_use) - $(cat global_rsv_reserved))) bytes leaked" | grep -v '^0 '
> +fi
> +
> +echo "Silence is golden"
> +
> +status=0
> +exit
> diff --git a/tests/generic/438.out b/tests/generic/438.out
> new file mode 100644
> index 00000000..4968f4d7
> --- /dev/null
> +++ b/tests/generic/438.out
> @@ -0,0 +1,2 @@
> +QA output created by 438
> +Silence is golden
> diff --git a/tests/generic/group b/tests/generic/group
> index 438c2990..c804b05d 100644
> --- a/tests/generic/group
> +++ b/tests/generic/group
> @@ -440,3 +440,4 @@
> 435 auto encrypt
> 436 auto quick rw
> 437 auto quick
> +438 auto
>
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] generic: test Btrfs delalloc accounting overflow
2017-06-02 9:07 ` Nikolay Borisov
@ 2017-06-02 12:46 ` David Sterba
2017-06-03 7:01 ` Eryu Guan
0 siblings, 1 reply; 8+ messages in thread
From: David Sterba @ 2017-06-02 12:46 UTC (permalink / raw)
To: Nikolay Borisov; +Cc: Omar Sandoval, fstests, linux-btrfs, kernel-team
On Fri, Jun 02, 2017 at 12:07:37PM +0300, Nikolay Borisov wrote:
> > +# Make sure that we didn't leak any metadata space.
> > +if [[ $FSTYP = btrfs ]]; then
> > + uuid="$(findmnt -n -o UUID "$TEST_DIR")"
>
> if we are on btrfs and we don't have findmnt this test will likely fail.
> Perhaps include a _require_command findmnt
I think utilities like findmnt should be checked at the beginning of the
whole testuiste, not in each test that uses them. As findmnt is part of
util-linux, missing it would also mean that eg 'mount' is missing.
Highly unlikely.
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] generic: test Btrfs delalloc accounting overflow
2017-06-02 12:46 ` David Sterba
@ 2017-06-03 7:01 ` Eryu Guan
0 siblings, 0 replies; 8+ messages in thread
From: Eryu Guan @ 2017-06-03 7:01 UTC (permalink / raw)
To: dsterba; +Cc: Nikolay Borisov, Omar Sandoval, fstests, linux-btrfs, kernel-team
On Fri, Jun 02, 2017 at 02:46:52PM +0200, David Sterba wrote:
> On Fri, Jun 02, 2017 at 12:07:37PM +0300, Nikolay Borisov wrote:
> > > +# Make sure that we didn't leak any metadata space.
> > > +if [[ $FSTYP = btrfs ]]; then
> > > + uuid="$(findmnt -n -o UUID "$TEST_DIR")"
> >
> > if we are on btrfs and we don't have findmnt this test will likely fail.
> > Perhaps include a _require_command findmnt
>
> I think utilities like findmnt should be checked at the beginning of the
> whole testuiste, not in each test that uses them. As findmnt is part of
Agreed. I think we can define a FINDMNT_PROG in common/config and refuse
to run any test if it's mising, as what we did to $MOUNT_PROG and other
must-have commands. (There's already a bare call to findmnt in
common/rc, change it to call $FINDMNT_PROG too).
We can do this in a separate patch.
Thanks,
Eryu
> util-linux, missing it would also mean that eg 'mount' is missing.
> Highly unlikely.
> --
> 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
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] generic: test Btrfs delalloc accounting overflow
2017-06-02 8:23 [PATCH] generic: test Btrfs delalloc accounting overflow Omar Sandoval
2017-06-02 9:07 ` Nikolay Borisov
@ 2017-06-03 7:37 ` Christoph Hellwig
2017-06-07 0:03 ` Omar Sandoval
1 sibling, 1 reply; 8+ messages in thread
From: Christoph Hellwig @ 2017-06-03 7:37 UTC (permalink / raw)
To: Omar Sandoval; +Cc: fstests, linux-btrfs, kernel-team
This looks like a btrfs-specific test, and not like a generic one
to me.
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] generic: test Btrfs delalloc accounting overflow
2017-06-03 7:37 ` Christoph Hellwig
@ 2017-06-07 0:03 ` Omar Sandoval
2017-06-07 3:08 ` Eryu Guan
0 siblings, 1 reply; 8+ messages in thread
From: Omar Sandoval @ 2017-06-07 0:03 UTC (permalink / raw)
To: Christoph Hellwig; +Cc: fstests, linux-btrfs, kernel-team
On Sat, Jun 03, 2017 at 12:37:00AM -0700, Christoph Hellwig wrote:
> This looks like a btrfs-specific test, and not like a generic one
> to me.
Nothing about the workload itself is btrfs-specific, we just have the
extra check at the end. But I don't really care, I can make it a btrfs
test unless Eryu has already applied it.
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] generic: test Btrfs delalloc accounting overflow
2017-06-07 0:03 ` Omar Sandoval
@ 2017-06-07 3:08 ` Eryu Guan
2017-06-07 3:12 ` Omar Sandoval
0 siblings, 1 reply; 8+ messages in thread
From: Eryu Guan @ 2017-06-07 3:08 UTC (permalink / raw)
To: Omar Sandoval; +Cc: Christoph Hellwig, fstests, linux-btrfs, kernel-team
On Tue, Jun 06, 2017 at 05:03:05PM -0700, Omar Sandoval wrote:
> On Sat, Jun 03, 2017 at 12:37:00AM -0700, Christoph Hellwig wrote:
> > This looks like a btrfs-specific test, and not like a generic one
> > to me.
>
> Nothing about the workload itself is btrfs-specific, we just have the
> extra check at the end. But I don't really care, I can make it a btrfs
> test unless Eryu has already applied it.
What's not very clear to me is that how do we check the accountings for
non-btrfs filesystems, i.e. what would cause a test failure on non-btrfs
filesystems besides an xfs_io write failure? I'd prefer a btrfs-specific
test if there's no good way to do the check. Or if we want to keep it a
generic test, some comments on the non-btrfs case would be good. I'm
fine with either way.
Thanks,
Eryu
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] generic: test Btrfs delalloc accounting overflow
2017-06-07 3:08 ` Eryu Guan
@ 2017-06-07 3:12 ` Omar Sandoval
0 siblings, 0 replies; 8+ messages in thread
From: Omar Sandoval @ 2017-06-07 3:12 UTC (permalink / raw)
To: Eryu Guan; +Cc: Christoph Hellwig, fstests, linux-btrfs, kernel-team
On Wed, Jun 07, 2017 at 11:08:01AM +0800, Eryu Guan wrote:
> On Tue, Jun 06, 2017 at 05:03:05PM -0700, Omar Sandoval wrote:
> > On Sat, Jun 03, 2017 at 12:37:00AM -0700, Christoph Hellwig wrote:
> > > This looks like a btrfs-specific test, and not like a generic one
> > > to me.
> >
> > Nothing about the workload itself is btrfs-specific, we just have the
> > extra check at the end. But I don't really care, I can make it a btrfs
> > test unless Eryu has already applied it.
>
> What's not very clear to me is that how do we check the accountings for
> non-btrfs filesystems, i.e. what would cause a test failure on non-btrfs
> filesystems besides an xfs_io write failure? I'd prefer a btrfs-specific
> test if there's no good way to do the check. Or if we want to keep it a
> generic test, some comments on the non-btrfs case would be good. I'm
> fine with either way.
>
> Thanks,
> Eryu
The only thing I can imagine other filesystems hitting would be an early
enospc or some sort of kernel warning, so I'll resend it as a btrfs test.
^ permalink raw reply [flat|nested] 8+ messages in thread
end of thread, other threads:[~2017-06-07 3:12 UTC | newest]
Thread overview: 8+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2017-06-02 8:23 [PATCH] generic: test Btrfs delalloc accounting overflow Omar Sandoval
2017-06-02 9:07 ` Nikolay Borisov
2017-06-02 12:46 ` David Sterba
2017-06-03 7:01 ` Eryu Guan
2017-06-03 7:37 ` Christoph Hellwig
2017-06-07 0:03 ` Omar Sandoval
2017-06-07 3:08 ` Eryu Guan
2017-06-07 3:12 ` Omar Sandoval
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.