[PATCH v2] ext4/021: make sure the fdatasync subprocess exits

[PATCH v2] ext4/021: make sure the fdatasync subprocess exits

[zhangyi (F)]

Now we just kill fdatasync_work process and wait nothing after the
test, so a busy unmount failure may appear if the fdatasync syscall
doesn't return in time.

  umount: /tmp/scratch: target is busy.
  mount: /tmp/scratch: /dev/sdb already mounted on /tmp/scratch.
  !!! failed to remount /dev/sdb on /tmp/scratch

This patch wait the xfs_io fdatasync subprocess exit to make sure
_check_scratch_fs success.

Signed-off-by: zhangyi (F) <yi.zhang@huawei.com>
---
 tests/ext4/021 | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/tests/ext4/021 b/tests/ext4/021
index 519737e1..3fc38e89 100755
--- a/tests/ext4/021
+++ b/tests/ext4/021
@@ -80,6 +80,10 @@ _scratch_mount
 
 do_fdatasync_work()
 {
+	# Wait for running subcommand before exitting so that
+	# mountpoint is not busy when we try to unmount it
+	trap "wait; exit" SIGTERM
+
 	while [ 1 ]; do
 		$XFS_IO_PROG -f -c "fdatasync" $SCRATCH_MNT/testfile
 	done
@@ -89,6 +93,7 @@ do_fdatasync_work &
 datasync_work_pid=$!
 sleep 10
 kill $datasync_work_pid >/dev/null 2>&1
+wait
 
 # success, all done
 status=0
-- 
2.17.2

Re: [PATCH v2] ext4/021: make sure the fdatasync subprocess exits

[Theodore Y. Ts'o]

On Wed, Feb 26, 2020 at 11:22:56AM +0800, zhangyi (F) wrote:
> Now we just kill fdatasync_work process and wait nothing after the
> test, so a busy unmount failure may appear if the fdatasync syscall
> doesn't return in time.
> 
>   umount: /tmp/scratch: target is busy.
>   mount: /tmp/scratch: /dev/sdb already mounted on /tmp/scratch.
>   !!! failed to remount /dev/sdb on /tmp/scratch
> 
> This patch wait the xfs_io fdatasync subprocess exit to make sure
> _check_scratch_fs success.
> 
> Signed-off-by: zhangyi (F) <yi.zhang@huawei.com>

Looks good, thanks!

Reviewed-by: Theodore Ts'o <tytso@mit.edu>