| Message ID | 20220302140548.1150-1-realwakka@gmail.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | [v2] btrfs: add test for enable/disable quota and create/destroy qgroup repeatedly | expand |
On Wed, Mar 02, 2022 at 02:05:48PM +0000, Sidong Yang wrote: > Test enable/disable quota and create/destroy qgroup repeatedly in > parallel and confirm it does not cause kernel hang. It only happens > in kernel staring with kernel 5.17-rc3. This is a regression test > for the problem reported to linux-btrfs list [1]. > > The hang was recreated using the test case and fixed by kernel patch > titled > > btrfs: qgroup: fix deadlock between rescan worker and remove qgroup > > [1] https://lore.kernel.org/linux-btrfs/20220228014340.21309-1-realwakka@gmail.com/ > > Signed-off-by: Sidong Yang <realwakka@gmail.com> > --- > v2 : fix changelog, comments, no wait pids argument > --- > tests/btrfs/262 | 40 ++++++++++++++++++++++++++++++++++++++++ > tests/btrfs/262.out | 2 ++ > 2 files changed, 42 insertions(+) > create mode 100755 tests/btrfs/262 > create mode 100644 tests/btrfs/262.out > > diff --git a/tests/btrfs/262 b/tests/btrfs/262 > new file mode 100755 > index 00000000..8953a28e > --- /dev/null > +++ b/tests/btrfs/262 > @@ -0,0 +1,40 @@ > +#! /bin/bash > +# SPDX-License-Identifier: GPL-2.0 > +# Copyright (c) 2022 Sidong Yang. All Rights Reserved. > +# > +# FS QA Test 262 > +# > +# Test that running qgroup enable, create, destroy, and disable commands in > +# parallel doesn't result in a deadlock, a crash or any filesystem > +# inconsistency. > +# > +. ./common/preamble > +_begin_fstest auto quick qgroup > + > +# Import common functions. > +. ./common/filter > + > +# real QA test starts here > + > +_supported_fs btrfs > + > +_require_scratch > + > +_scratch_mkfs > /dev/null 2>&1 > +_scratch_mount > + > +for ((i = 0; i < 200; i++)); do > + $BTRFS_UTIL_PROG quota enable $SCRATCH_MNT 2>> $seqres.full & > + $BTRFS_UTIL_PROG qgroup create 1/0 $SCRATCH_MNT 2>> $seqres.full & > + $BTRFS_UTIL_PROG qgroup destroy 1/0 $SCRATCH_MNT 2>> $seqres.full & > + $BTRFS_UTIL_PROG quota disable $SCRATCH_MNT 2>> $seqres.full & > +done > + > +wait > + > +_scratch_unmount The fstests framework will also do the unmount, so this can be removed. You don't need to send a new patch version just to remove that, Eryu can probably amend the patch when he picks it up. Anyway, it looks good and it works as expected. Reviewed-by: Filipe Manana <fdmanana@suse.com> Thanks for making the test case. > + > +# success, all done > +echo "Silence is golden" > +status=0 > +exit > diff --git a/tests/btrfs/262.out b/tests/btrfs/262.out > new file mode 100644 > index 00000000..404badc3 > --- /dev/null > +++ b/tests/btrfs/262.out > @@ -0,0 +1,2 @@ > +QA output created by 262 > +Silence is golden > -- > 2.25.1 >
On Mar 02, 2022 / 14:05, Sidong Yang wrote: > Test enable/disable quota and create/destroy qgroup repeatedly in > parallel and confirm it does not cause kernel hang. It only happens > in kernel staring with kernel 5.17-rc3. This is a regression test > for the problem reported to linux-btrfs list [1]. > > The hang was recreated using the test case and fixed by kernel patch > titled > > btrfs: qgroup: fix deadlock between rescan worker and remove qgroup > > [1] https://lore.kernel.org/linux-btrfs/20220228014340.21309-1-realwakka@gmail.com/ > > Signed-off-by: Sidong Yang <realwakka@gmail.com> As Filipe noted, _scratch_unmount can be removed. Other than that, it looks good to me. Thanks again for catching this! Reviewed-by: Shin'ichiro Kawasaki <shinichiro.kawasaki@wdc.com>
diff --git a/tests/btrfs/262 b/tests/btrfs/262 new file mode 100755 index 00000000..8953a28e --- /dev/null +++ b/tests/btrfs/262 @@ -0,0 +1,40 @@ +#! /bin/bash +# SPDX-License-Identifier: GPL-2.0 +# Copyright (c) 2022 Sidong Yang. All Rights Reserved. +# +# FS QA Test 262 +# +# Test that running qgroup enable, create, destroy, and disable commands in +# parallel doesn't result in a deadlock, a crash or any filesystem +# inconsistency. +# +. ./common/preamble +_begin_fstest auto quick qgroup + +# Import common functions. +. ./common/filter + +# real QA test starts here + +_supported_fs btrfs + +_require_scratch + +_scratch_mkfs > /dev/null 2>&1 +_scratch_mount + +for ((i = 0; i < 200; i++)); do + $BTRFS_UTIL_PROG quota enable $SCRATCH_MNT 2>> $seqres.full & + $BTRFS_UTIL_PROG qgroup create 1/0 $SCRATCH_MNT 2>> $seqres.full & + $BTRFS_UTIL_PROG qgroup destroy 1/0 $SCRATCH_MNT 2>> $seqres.full & + $BTRFS_UTIL_PROG quota disable $SCRATCH_MNT 2>> $seqres.full & +done + +wait + +_scratch_unmount + +# success, all done +echo "Silence is golden" +status=0 +exit diff --git a/tests/btrfs/262.out b/tests/btrfs/262.out new file mode 100644 index 00000000..404badc3 --- /dev/null +++ b/tests/btrfs/262.out @@ -0,0 +1,2 @@ +QA output created by 262 +Silence is golden
Test enable/disable quota and create/destroy qgroup repeatedly in parallel and confirm it does not cause kernel hang. It only happens in kernel staring with kernel 5.17-rc3. This is a regression test for the problem reported to linux-btrfs list [1]. The hang was recreated using the test case and fixed by kernel patch titled btrfs: qgroup: fix deadlock between rescan worker and remove qgroup [1] https://lore.kernel.org/linux-btrfs/20220228014340.21309-1-realwakka@gmail.com/ Signed-off-by: Sidong Yang <realwakka@gmail.com> --- v2 : fix changelog, comments, no wait pids argument --- tests/btrfs/262 | 40 ++++++++++++++++++++++++++++++++++++++++ tests/btrfs/262.out | 2 ++ 2 files changed, 42 insertions(+) create mode 100755 tests/btrfs/262 create mode 100644 tests/btrfs/262.out