mptcp: really cope with fastopen race

Fastopen and PM-trigger subflow shutdown can race, as reported by
syzkaller.

In my first attempt to close such race, I missed the fact that
the subflow status can change again before the subflow_state_change
callback is invoked.

Address the issue additionally copying with all the states directly
reachable from TCP_FIN_WAIT1.

Fixes: 1e777f3 ("mptcp: add MSG_FASTOPEN sendmsg flag support")
Fixes: 4fd19a3 ("mptcp: fix inconsistent state on fastopen race")
Reported-by: syzbot+c53d4d3ddb327e80bc51@syzkaller.appspotmail.com
Closes: #458
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Mat Martineau <martineau@kernel.org>

net/mptcp/protocol.h

@@ -1129,7 +1129,8 @@ static inline bool subflow_simultaneous_connect(struct sock *sk)
 {
 	struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(sk);
 
-	return (1 << sk->sk_state) & (TCPF_ESTABLISHED | TCPF_FIN_WAIT1) &&
+	return (1 << sk->sk_state) &
+	       (TCPF_ESTABLISHED | TCPF_FIN_WAIT1 | TCPF_FIN_WAIT2 | TCPF_CLOSING) &&
 	       is_active_ssk(subflow) &&
 	       !subflow->conn_finished;
 }