ocfs2: drop acl cache for directories too

commit 9c0f0a0 upstream.

ocfs2_data_convert_worker() is currently dropping any cached acl info
for FILE before down-converting meta lock.  It should also drop for
DIRECTORY.  Otherwise the second acl lookup returns the cached one (from
VFS layer) which could be already stale.

The problem we are seeing is that the acl changes on one node doesn't
get refreshed on other nodes in the following case:

  Node 1                    Node 2
  --------------            ----------------
  getfacl dir1

                            getfacl dir1    <-- this is OK

  setfacl -m u:user1:rwX dir1
  getfacl dir1   <-- see the change for user1

                            getfacl dir1    <-- can't see change for user1

Link: https://lkml.kernel.org/r/20210903012631.6099-1-wen.gang.wang@oracle.com
Signed-off-by: Wengang Wang <wen.gang.wang@oracle.com>
Reviewed-by: Joseph Qi <joseph.qi@linux.alibaba.com>
Cc: Mark Fasheh <mark@fasheh.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Junxiao Bi <junxiao.bi@oracle.com>
Cc: Changwei Ge <gechangwei@live.cn>
Cc: Gang He <ghe@suse.com>
Cc: Jun Piao <piaojun@huawei.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

fs/ocfs2/dlmglue.c

@@ -3939,7 +3939,7 @@ static int ocfs2_data_convert_worker(struct ocfs2_lock_res *lockres,
 		oi = OCFS2_I(inode);
 		oi->ip_dir_lock_gen++;
 		mlog(0, "generation: %u\n", oi->ip_dir_lock_gen);
-		goto out;
+		goto out_forget;
 	}
 
 	if (!S_ISREG(inode->i_mode))
@@ -3970,6 +3970,7 @@ static int ocfs2_data_convert_worker(struct ocfs2_lock_res *lockres,
 		filemap_fdatawait(mapping);
 	}
 
+out_forget:
 	forget_all_cached_acls(inode);
 
 out: