[PATCH 4.14 31/77] ceph: flush dirty inodes before proceeding with remount

From: Greg Kroah-Hartman
Date: Thu May 23 2019 - 15:14:37 EST


From: Jeff Layton <jlayton@xxxxxxxxxx>

commit 00abf69dd24f4444d185982379c5cc3bb7b6d1fc upstream.

xfstest generic/452 was triggering a "Busy inodes after umount" warning.
ceph was allowing the mount to go read-only without first flushing out
dirty inodes in the cache. Ensure we sync out the filesystem before
allowing a remount to proceed.

Cc: stable@xxxxxxxxxxxxxxx
Link: http://tracker.ceph.com/issues/39571
Signed-off-by: Jeff Layton <jlayton@xxxxxxxxxx>
Reviewed-by: "Yan, Zheng" <zyan@xxxxxxxxxx>
Signed-off-by: Ilya Dryomov <idryomov@xxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>

---
fs/ceph/super.c | 7 +++++++
1 file changed, 7 insertions(+)

--- a/fs/ceph/super.c
+++ b/fs/ceph/super.c
@@ -768,6 +768,12 @@ static void ceph_umount_begin(struct sup
return;
}

+static int ceph_remount(struct super_block *sb, int *flags, char *data)
+{
+ sync_filesystem(sb);
+ return 0;
+}
+
static const struct super_operations ceph_super_ops = {
.alloc_inode = ceph_alloc_inode,
.destroy_inode = ceph_destroy_inode,
@@ -775,6 +781,7 @@ static const struct super_operations cep
.drop_inode = ceph_drop_inode,
.sync_fs = ceph_sync_fs,
.put_super = ceph_put_super,
+ .remount_fs = ceph_remount,
.show_options = ceph_show_options,
.statfs = ceph_statfs,
.umount_begin = ceph_umount_begin,