clone_private_mnt(): make sure that caller has CAP_SYS_ADMIN in the right userns

commit c28f922c9d upstream.

What we want is to verify there is that clone won't expose something
hidden by a mount we wouldn't be able to undo.  "Wouldn't be able to undo"
may be a result of MNT_LOCKED on a child, but it may also come from
lacking admin rights in the userns of the namespace mount belongs to.

clone_private_mnt() checks the former, but not the latter.

There's a number of rather confusing CAP_SYS_ADMIN checks in various
userns during the mount, especially with the new mount API; they serve
different purposes and in case of clone_private_mnt() they usually,
but not always end up covering the missing check mentioned above.

Reviewed-by: Christian Brauner <brauner@kernel.org>
Reported-by: "Orlando, Noah" <Noah.Orlando@deshaw.com>
Fixes: 427215d85e ("ovl: prevent private clone if bind mount is not allowed")
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
[ merge conflict resolution: clone_private_mount() was reworked in
  db04662e2f ("fs: allow detached mounts in clone_private_mount()").
  Tweak the relevant ns_capable check so that it works on older kernels ]
Signed-off-by: Noah Orlando <Noah.Orlando@deshaw.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

fs/namespace.c

@@ -2018,6 +2018,11 @@ struct vfsmount *clone_private_mount(const struct path *path)
 	if (!check_mnt(old_mnt))
 		goto invalid;
 
+	if (!ns_capable(old_mnt->mnt_ns->user_ns, CAP_SYS_ADMIN)) {
+		up_read(&namespace_sem);
+		return ERR_PTR(-EPERM);
+	}
+
 	if (has_locked_children(old_mnt, path->dentry))
 		goto invalid;