Re: [PATCH] btrfs/131: send test for rename+recycled ino

[Eryu Guan <eguan@redhat.com>]

On Tue, Aug 23, 2016 at 01:51:58PM -0400, Josef Bacik wrote:
> This is a test for a problem I hit with send/receive.  If you mount with -o
> inode_cache it is possible to recycle inode numbers, and btrfs assumes that when
> you have a recycled inode number that it's parent directory won't be needing a
> rename, but this is not the case.  Without the patch to fix this problem this
> test will panic the box if ASSERT() is turned on, otherwise it'll simply fail
> the fssum.
> 
> Signed-off-by: Josef Bacik <jbacik@fb.com>
> ---
>  tests/btrfs/131     | 91 +++++++++++++++++++++++++++++++++++++++++++++++++++++
>  tests/btrfs/131.out |  2 ++
>  tests/btrfs/group   |  1 +
>  3 files changed, 94 insertions(+)
>  create mode 100644 tests/btrfs/131
>  create mode 100644 tests/btrfs/131.out
> 
> diff --git a/tests/btrfs/131 b/tests/btrfs/131
> new file mode 100644
> index 0000000..596adc1
> --- /dev/null
> +++ b/tests/btrfs/131
> @@ -0,0 +1,91 @@
> +#! /bin/bash
> +# FS QA Test 131
> +#
> +# Make sure btrfs handles send/receive of a file that has been destroyed and
> +# recreated with a recycled inode number and a renamed parent directory.  This
> +# is a test for a fix based on the patch
> +# 

A trailing whitespace in above line :)

> +# Btrfs: handle pending renames with recycled inodes properly
> +#
> +#-----------------------------------------------------------------------
> +# Copyright (c) 2016 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=`mktemp -d`

Usually we define "tmp=/tmp/$$", and tests and some common helpers use
$tmp.xxx for tmp files. I think we should keep the consistency across
tests.

$tmp is used as a tmp dir to store the btrfs send data, I think we can
send the data to some dir in $TEST_DIR, e.g.

send_dir=$TEST_DIR/btrfs_send_$seq

> +status=1	# failure is the default!
> +trap "_cleanup; exit \$status" 0 1 2 3 15
> +
> +_cleanup()
> +{
> +	rm -fr $tmp

Then clean the send dir here, as well as $tmp.*

	cd /
	rm -rf $send_dir
	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 btrfs
> +_supported_os Linux
> +_require_scratch
> +_require_fssum
> +
> +_scratch_mkfs > /dev/null 2>&1
> +_scratch_mount "-o inode_cache"

Need some comments to explain why mount with "-o inode_cache" option.

> +
> +_run_btrfs_util_prog subvolume create $SCRATCH_MNT/base
> +mkdir $SCRATCH_MNT/base/b
> +mkdir $SCRATCH_MNT/base/a
> +mv $SCRATCH_MNT/base/b $SCRATCH_MNT/base/a
> +
> +_run_btrfs_util_prog subvolume snapshot -r $SCRATCH_MNT/base \
> +	$SCRATCH_MNT/snap1
> +_run_btrfs_util_prog send $SCRATCH_MNT/snap1 -f $tmp/1.snap
> +
> +mv $SCRATCH_MNT/base/a $SCRATCH_MNT/base/c
> +rmdir $SCRATCH_MNT/base/c/b
> +
> +# We need to sync to make sure the inode number is recycled
> +sync
> +mkdir $SCRATCH_MNT/base/c/b
> +
> +_run_btrfs_util_prog subvolume snapshot -r $SCRATCH_MNT/base $SCRATCH_MNT/snap2
> +run_check $FSSUM_PROG -A -f -w $tmp/fssum $SCRATCH_MNT/snap2
> +
> +_run_btrfs_util_prog send -p $SCRATCH_MNT/snap1 $SCRATCH_MNT/snap2 -f $tmp/2.snap
> +
> +_scratch_unmount
> +_scratch_mkfs > /dev/null 2>&1
> +_scratch_mount
> +
> +_run_btrfs_util_prog receive $SCRATCH_MNT -f $tmp/1.snap
> +_run_btrfs_util_prog receive $SCRATCH_MNT -f $tmp/2.snap
> +run_check $FSSUM_PROG -r $tmp/fssum $SCRATCH_MNT/snap2

I did see assert failure (and kernel crash) on kernel with ASSERT turned
on, but I didn't see fssum reporting failures on kernel with ASSERT
turned off. Did I miss anything?

[root@dhcp-66-86-11 xfstests]# ./check -s btrfs  btrfs/131
SECTION       -- btrfs
RECREATING    -- btrfs on /dev/sda5
FSTYP         -- btrfs
PLATFORM      -- Linux/x86_64 dhcp-66-86-11 4.8.0-rc2
MKFS_OPTIONS  -- /dev/sda6
MOUNT_OPTIONS -- -o context=system_u:object_r:nfs_t:s0 /dev/sda6 /mnt/testarea/scratch

btrfs/131 1s ... 1s
Ran: btrfs/131
Passed all 1 tests

SECTION       -- btrfs
=========================
Ran: btrfs/131
Passed all 1 tests

Thanks,
Eryu

> +
> +echo "Silence is golden"
> +status=0
> +exit
> diff --git a/tests/btrfs/131.out b/tests/btrfs/131.out
> new file mode 100644
> index 0000000..d118ca9
> --- /dev/null
> +++ b/tests/btrfs/131.out
> @@ -0,0 +1,2 @@
> +QA output created by 131
> +Silence is golden
> diff --git a/tests/btrfs/group b/tests/btrfs/group
> index 6b29c05..933fe10 100644
> --- a/tests/btrfs/group
> +++ b/tests/btrfs/group
> @@ -133,3 +133,4 @@
>  128 auto quick send
>  129 auto quick send
>  130 auto clone send
> +131 auto quick send
> -- 
> 1.8.3.1
> 
> --
> 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