* [PATCH v2] generic/707: Test moving directory while being grown
@ 2023-01-30 14:56 Jan Kara
2023-01-30 16:04 ` Zorro Lang
2023-01-30 16:11 ` Bill O'Donnell
0 siblings, 2 replies; 4+ messages in thread
From: Jan Kara @ 2023-01-30 14:56 UTC (permalink / raw)
To: fstests; +Cc: Bill O'Donnell, Jan Kara
Test how the filesystem can handle moving a directory to a different
directory (so that parent pointer gets updated) while it is grown. Ext4
and UDF had a bug where if the directory got converted to a different
type due to growth while rename is running, the filesystem got
corrupted.
Reviewed-by: Bill O'Donnell <bodonnel@redhat.com>
Signed-off-by: Jan Kara <jack@suse.cz>
---
tests/generic/707 | 63 +++++++++++++++++++++++++++++++++++++++++++
tests/generic/707.out | 2 ++
2 files changed, 65 insertions(+)
create mode 100755 tests/generic/707
create mode 100644 tests/generic/707.out
Changes since v1:
* Added TIME_FACTOR multiplier
* Handle background process in _cleanup
* Use absolute paths where possible
diff --git a/tests/generic/707 b/tests/generic/707
new file mode 100755
index 000000000000..d1d299563813
--- /dev/null
+++ b/tests/generic/707
@@ -0,0 +1,63 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2023 Jan Kara, SUSE Linux. All Rights Reserved.
+#
+# FS QA Test 707
+#
+# This is a test verifying whether the filesystem can gracefully handle
+# modifying of a directory while it is being moved, in particular the cases
+# where directory format changes
+#
+. ./common/preamble
+_begin_fstest auto
+
+_supported_fs generic
+_require_scratch
+
+_scratch_mkfs >>$seqres.full 2>&1
+_scratch_mount
+
+_cleanup()
+{
+ if [ -n "$BGPID" ]; then
+ # Stop background process
+ kill -9 $BGPID &>/dev/null
+ wait
+ fi
+}
+
+# Loop multiple times trying to hit the race
+loops=$((100*TIME_FACTOR))
+files=500
+moves=500
+
+create_files()
+{
+ # We use slightly longer file name to make directory grow faster and
+ # hopefully convert between various types
+ for (( i = 0; i < $files; i++ )); do
+ touch somewhatlongerfilename$i
+ done
+}
+
+for (( i = 0; i <= $moves; i++ )); do
+ mkdir $SCRATCH_MNT/dir$i
+done
+
+for (( l = 0; l < $loops; l++ )); do
+ mkdir $SCRATCH_MNT/dir0/dir
+ pushd $SCRATCH_MNT/dir0/dir &>/dev/null
+ create_files &
+ BGPID=$!
+ popd &>/dev/null
+ for (( i = 0; i < $moves; i++ )); do
+ mv $SCRATCH_MNT/dir$i/dir $SCRATCH_MNT/dir$((i+1))/dir
+ done
+ wait
+ BGPID=""
+ rm -r $SCRATCH_MNT/dir$moves/dir
+done
+
+echo "Silence is golden"
+status=0
+exit
diff --git a/tests/generic/707.out b/tests/generic/707.out
new file mode 100644
index 000000000000..8e57a1d8c971
--- /dev/null
+++ b/tests/generic/707.out
@@ -0,0 +1,2 @@
+QA output created by 707
+Silence is golden
--
2.35.3
^ permalink raw reply related [flat|nested] 4+ messages in thread
* Re: [PATCH v2] generic/707: Test moving directory while being grown
2023-01-30 14:56 [PATCH v2] generic/707: Test moving directory while being grown Jan Kara
@ 2023-01-30 16:04 ` Zorro Lang
2023-01-31 12:46 ` Jan Kara
2023-01-30 16:11 ` Bill O'Donnell
1 sibling, 1 reply; 4+ messages in thread
From: Zorro Lang @ 2023-01-30 16:04 UTC (permalink / raw)
To: Jan Kara; +Cc: fstests, Bill O'Donnell
On Mon, Jan 30, 2023 at 03:56:39PM +0100, Jan Kara wrote:
> Test how the filesystem can handle moving a directory to a different
> directory (so that parent pointer gets updated) while it is grown. Ext4
> and UDF had a bug where if the directory got converted to a different
> type due to growth while rename is running, the filesystem got
> corrupted.
>
> Reviewed-by: Bill O'Donnell <bodonnel@redhat.com>
> Signed-off-by: Jan Kara <jack@suse.cz>
> ---
> tests/generic/707 | 63 +++++++++++++++++++++++++++++++++++++++++++
> tests/generic/707.out | 2 ++
> 2 files changed, 65 insertions(+)
> create mode 100755 tests/generic/707
> create mode 100644 tests/generic/707.out
>
> Changes since v1:
> * Added TIME_FACTOR multiplier
> * Handle background process in _cleanup
> * Use absolute paths where possible
>
> diff --git a/tests/generic/707 b/tests/generic/707
> new file mode 100755
> index 000000000000..d1d299563813
> --- /dev/null
> +++ b/tests/generic/707
> @@ -0,0 +1,63 @@
> +#! /bin/bash
> +# SPDX-License-Identifier: GPL-2.0
> +# Copyright (c) 2023 Jan Kara, SUSE Linux. All Rights Reserved.
> +#
> +# FS QA Test 707
> +#
> +# This is a test verifying whether the filesystem can gracefully handle
> +# modifying of a directory while it is being moved, in particular the cases
> +# where directory format changes
> +#
^
A trailing whitespace at here :)
> +. ./common/preamble
> +_begin_fstest auto
> +
> +_supported_fs generic
> +_require_scratch
> +
> +_scratch_mkfs >>$seqres.full 2>&1
> +_scratch_mount
> +
> +_cleanup()
> +{
> + if [ -n "$BGPID" ]; then
> + # Stop background process
> + kill -9 $BGPID &>/dev/null
> + wait
> + fi
Neet to keep the code in common/preamble:_cleanup(), if you cover it:
cd /
rm -f $tmp.*
especially you use pushd and popd.
I forgot to ask, is this case written for someone known fix or change? If so,
better to use _fixed_by_kernel_commit or _wants_kernel_commit.
> +}
> +
> +# Loop multiple times trying to hit the race
> +loops=$((100*TIME_FACTOR))
> +files=500
> +moves=500
> +
> +create_files()
> +{
> + # We use slightly longer file name to make directory grow faster and
> + # hopefully convert between various types
> + for (( i = 0; i < $files; i++ )); do
> + touch somewhatlongerfilename$i
> + done
> +}
> +
> +for (( i = 0; i <= $moves; i++ )); do
> + mkdir $SCRATCH_MNT/dir$i
> +done
> +
> +for (( l = 0; l < $loops; l++ )); do
> + mkdir $SCRATCH_MNT/dir0/dir
> + pushd $SCRATCH_MNT/dir0/dir &>/dev/null
> + create_files &
> + BGPID=$!
> + popd &>/dev/null
> + for (( i = 0; i < $moves; i++ )); do
> + mv $SCRATCH_MNT/dir$i/dir $SCRATCH_MNT/dir$((i+1))/dir
> + done
> + wait
> + BGPID=""
unset BGPID might be better, but anyway :)
Others looks good to me.
Thanks,
Zorro
> + rm -r $SCRATCH_MNT/dir$moves/dir
> +done
> +
> +echo "Silence is golden"
> +status=0
> +exit
> diff --git a/tests/generic/707.out b/tests/generic/707.out
> new file mode 100644
> index 000000000000..8e57a1d8c971
> --- /dev/null
> +++ b/tests/generic/707.out
> @@ -0,0 +1,2 @@
> +QA output created by 707
> +Silence is golden
> --
> 2.35.3
>
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH v2] generic/707: Test moving directory while being grown
2023-01-30 14:56 [PATCH v2] generic/707: Test moving directory while being grown Jan Kara
2023-01-30 16:04 ` Zorro Lang
@ 2023-01-30 16:11 ` Bill O'Donnell
1 sibling, 0 replies; 4+ messages in thread
From: Bill O'Donnell @ 2023-01-30 16:11 UTC (permalink / raw)
To: Jan Kara; +Cc: fstests
On Mon, Jan 30, 2023 at 03:56:39PM +0100, Jan Kara wrote:
> Test how the filesystem can handle moving a directory to a different
> directory (so that parent pointer gets updated) while it is grown. Ext4
> and UDF had a bug where if the directory got converted to a different
> type due to growth while rename is running, the filesystem got
> corrupted.
>
> Reviewed-by: Bill O'Donnell <bodonnel@redhat.com>
> Signed-off-by: Jan Kara <jack@suse.cz>
With Zoro's suggested changes, this looks fine.
Reviewed-by: Bill O'Donnell <bodonnel@redhat.com>
> ---
> tests/generic/707 | 63 +++++++++++++++++++++++++++++++++++++++++++
> tests/generic/707.out | 2 ++
> 2 files changed, 65 insertions(+)
> create mode 100755 tests/generic/707
> create mode 100644 tests/generic/707.out
>
> Changes since v1:
> * Added TIME_FACTOR multiplier
> * Handle background process in _cleanup
> * Use absolute paths where possible
>
> diff --git a/tests/generic/707 b/tests/generic/707
> new file mode 100755
> index 000000000000..d1d299563813
> --- /dev/null
> +++ b/tests/generic/707
> @@ -0,0 +1,63 @@
> +#! /bin/bash
> +# SPDX-License-Identifier: GPL-2.0
> +# Copyright (c) 2023 Jan Kara, SUSE Linux. All Rights Reserved.
> +#
> +# FS QA Test 707
> +#
> +# This is a test verifying whether the filesystem can gracefully handle
> +# modifying of a directory while it is being moved, in particular the cases
> +# where directory format changes
> +#
> +. ./common/preamble
> +_begin_fstest auto
> +
> +_supported_fs generic
> +_require_scratch
> +
> +_scratch_mkfs >>$seqres.full 2>&1
> +_scratch_mount
> +
> +_cleanup()
> +{
> + if [ -n "$BGPID" ]; then
> + # Stop background process
> + kill -9 $BGPID &>/dev/null
> + wait
> + fi
> +}
> +
> +# Loop multiple times trying to hit the race
> +loops=$((100*TIME_FACTOR))
> +files=500
> +moves=500
> +
> +create_files()
> +{
> + # We use slightly longer file name to make directory grow faster and
> + # hopefully convert between various types
> + for (( i = 0; i < $files; i++ )); do
> + touch somewhatlongerfilename$i
> + done
> +}
> +
> +for (( i = 0; i <= $moves; i++ )); do
> + mkdir $SCRATCH_MNT/dir$i
> +done
> +
> +for (( l = 0; l < $loops; l++ )); do
> + mkdir $SCRATCH_MNT/dir0/dir
> + pushd $SCRATCH_MNT/dir0/dir &>/dev/null
> + create_files &
> + BGPID=$!
> + popd &>/dev/null
> + for (( i = 0; i < $moves; i++ )); do
> + mv $SCRATCH_MNT/dir$i/dir $SCRATCH_MNT/dir$((i+1))/dir
> + done
> + wait
> + BGPID=""
> + rm -r $SCRATCH_MNT/dir$moves/dir
> +done
> +
> +echo "Silence is golden"
> +status=0
> +exit
> diff --git a/tests/generic/707.out b/tests/generic/707.out
> new file mode 100644
> index 000000000000..8e57a1d8c971
> --- /dev/null
> +++ b/tests/generic/707.out
> @@ -0,0 +1,2 @@
> +QA output created by 707
> +Silence is golden
> --
> 2.35.3
>
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH v2] generic/707: Test moving directory while being grown
2023-01-30 16:04 ` Zorro Lang
@ 2023-01-31 12:46 ` Jan Kara
0 siblings, 0 replies; 4+ messages in thread
From: Jan Kara @ 2023-01-31 12:46 UTC (permalink / raw)
To: Zorro Lang; +Cc: Jan Kara, fstests, Bill O'Donnell
On Tue 31-01-23 00:04:01, Zorro Lang wrote:
> On Mon, Jan 30, 2023 at 03:56:39PM +0100, Jan Kara wrote:
> > +. ./common/preamble
> > +_begin_fstest auto
> > +
> > +_supported_fs generic
> > +_require_scratch
> > +
> > +_scratch_mkfs >>$seqres.full 2>&1
> > +_scratch_mount
> > +
> > +_cleanup()
> > +{
> > + if [ -n "$BGPID" ]; then
> > + # Stop background process
> > + kill -9 $BGPID &>/dev/null
> > + wait
> > + fi
>
> Neet to keep the code in common/preamble:_cleanup(), if you cover it:
>
> cd /
> rm -f $tmp.*
>
> especially you use pushd and popd.
OK, done.
> I forgot to ask, is this case written for someone known fix or change? If so,
> better to use _fixed_by_kernel_commit or _wants_kernel_commit.
Yes, good point. It tests for UDF & ext4 corruption issue. Ext4 commit is
not yet merged but I'll add there something.
> > +}
> > +
> > +# Loop multiple times trying to hit the race
> > +loops=$((100*TIME_FACTOR))
> > +files=500
> > +moves=500
> > +
> > +create_files()
> > +{
> > + # We use slightly longer file name to make directory grow faster and
> > + # hopefully convert between various types
> > + for (( i = 0; i < $files; i++ )); do
> > + touch somewhatlongerfilename$i
> > + done
> > +}
> > +
> > +for (( i = 0; i <= $moves; i++ )); do
> > + mkdir $SCRATCH_MNT/dir$i
> > +done
> > +
> > +for (( l = 0; l < $loops; l++ )); do
> > + mkdir $SCRATCH_MNT/dir0/dir
> > + pushd $SCRATCH_MNT/dir0/dir &>/dev/null
> > + create_files &
> > + BGPID=$!
> > + popd &>/dev/null
> > + for (( i = 0; i < $moves; i++ )); do
> > + mv $SCRATCH_MNT/dir$i/dir $SCRATCH_MNT/dir$((i+1))/dir
> > + done
> > + wait
> > + BGPID=""
>
> unset BGPID might be better, but anyway :)
Yeah, nicer. Done.
Honza
--
Jan Kara <jack@suse.com>
SUSE Labs, CR
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2023-01-31 12:46 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2023-01-30 14:56 [PATCH v2] generic/707: Test moving directory while being grown Jan Kara
2023-01-30 16:04 ` Zorro Lang
2023-01-31 12:46 ` Jan Kara
2023-01-30 16:11 ` Bill O'Donnell
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox