1 files changed, 58 insertions, 32 deletions

diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c
index 3ad9c46202fc..b998e9df53ce 100644
--- a/net/mptcp/protocol.c
+++ b/net/mptcp/protocol.c
@@ -825,7 +825,6 @@ static bool __mptcp_finish_join(struct mptcp_sock *msk, struct sock *ssk)
 	if (sk->sk_socket && !ssk->sk_socket)
 		mptcp_sock_graft(ssk, sk->sk_socket);
 
-	mptcp_propagate_sndbuf((struct sock *)msk, ssk);
 	mptcp_sockopt_sync_locked(msk, ssk);
 	return true;
 }
@@ -2316,7 +2315,26 @@ static void __mptcp_close_ssk(struct sock *sk, struct sock *ssk,
 			      unsigned int flags)
 {
 	struct mptcp_sock *msk = mptcp_sk(sk);
-	bool need_push, dispose_it;
+	bool dispose_it, need_push = false;
+
+	/* If the first subflow moved to a close state before accept, e.g. due
+	 * to an incoming reset, mptcp either:
+	 * - if either the subflow or the msk are dead, destroy the context
+	 *   (the subflow socket is deleted by inet_child_forget) and the msk
+	 * - otherwise do nothing at the moment and take action at accept and/or
+	 *   listener shutdown - user-space must be able to accept() the closed
+	 *   socket.
+	 */
+	if (msk->in_accept_queue && msk->first == ssk) {
+		if (!sock_flag(sk, SOCK_DEAD) && !sock_flag(ssk, SOCK_DEAD))
+			return;
+
+		/* ensure later check in mptcp_worker() will dispose the msk */
+		sock_set_flag(sk, SOCK_DEAD);
+		lock_sock_nested(ssk, SINGLE_DEPTH_NESTING);
+		mptcp_subflow_drop_ctx(ssk);
+		goto out_release;
+	}
 
 	dispose_it = !msk->subflow || ssk != msk->subflow->sk;
 	if (dispose_it)
@@ -2343,7 +2361,6 @@ static void __mptcp_close_ssk(struct sock *sk, struct sock *ssk,
 		goto out;
 	}
 
-	sock_orphan(ssk);
 	subflow->disposable = 1;
 
 	/* if ssk hit tcp_done(), tcp_cleanup_ulp() cleared the related ops
@@ -2351,6 +2368,7 @@ static void __mptcp_close_ssk(struct sock *sk, struct sock *ssk,
 	 * reference owned by msk;
 	 */
 	if (!inet_csk(ssk)->icsk_ulp_ops) {
+		WARN_ON_ONCE(!sock_flag(ssk, SOCK_DEAD));
 		kfree_rcu(subflow, rcu);
 	} else {
 		/* otherwise tcp will dispose of the ssk and subflow ctx */
@@ -2360,11 +2378,14 @@ static void __mptcp_close_ssk(struct sock *sk, struct sock *ssk,
 			inet_csk_listen_stop(ssk);
 			mptcp_event_pm_listener(ssk, MPTCP_EVENT_LISTENER_CLOSED);
 		}
+
 		__tcp_close(ssk, 0);
 
 		/* close acquired an extra ref */
 		__sock_put(ssk);
 	}
+
+out_release:
 	release_sock(ssk);
 
 	sock_put(ssk);
@@ -2399,9 +2420,10 @@ static unsigned int mptcp_sync_mss(struct sock *sk, u32 pmtu)
 	return 0;
 }
 
-static void __mptcp_close_subflow(struct mptcp_sock *msk)
+static void __mptcp_close_subflow(struct sock *sk)
 {
 	struct mptcp_subflow_context *subflow, *tmp;
+	struct mptcp_sock *msk = mptcp_sk(sk);
 
 	might_sleep();
 
@@ -2415,16 +2437,17 @@ static void __mptcp_close_subflow(struct mptcp_sock *msk)
 		if (!skb_queue_empty_lockless(&ssk->sk_receive_queue))
 			continue;
 
-		mptcp_close_ssk((struct sock *)msk, ssk, subflow);
+		mptcp_close_ssk(sk, ssk, subflow);
 	}
+
 }
 
-static bool mptcp_check_close_timeout(const struct sock *sk)
+static bool mptcp_should_close(const struct sock *sk)
 {
 	s32 delta = tcp_jiffies32 - inet_csk(sk)->icsk_mtup.probe_timestamp;
 	struct mptcp_subflow_context *subflow;
 
-	if (delta >= TCP_TIMEWAIT_LEN)
+	if (delta >= TCP_TIMEWAIT_LEN || mptcp_sk(sk)->in_accept_queue)
 		return true;
 
 	/* if all subflows are in closed status don't bother with additional
@@ -2609,7 +2632,7 @@ static void mptcp_worker(struct work_struct *work)
 
 	lock_sock(sk);
 	state = sk->sk_state;
-	if (unlikely(state == TCP_CLOSE))
+	if (unlikely((1 << state) & (TCPF_CLOSE | TCPF_LISTEN)))
 		goto unlock;
 
 	mptcp_check_data_fin_ack(sk);
@@ -2624,12 +2647,15 @@ static void mptcp_worker(struct work_struct *work)
 	__mptcp_check_send_data_fin(sk);
 	mptcp_check_data_fin(sk);
 
+	if (test_and_clear_bit(MPTCP_WORK_CLOSE_SUBFLOW, &msk->flags))
+		__mptcp_close_subflow(sk);
+
 	/* There is no point in keeping around an orphaned sk timedout or
 	 * closed, but we need the msk around to reply to incoming DATA_FIN,
 	 * even if it is orphaned and in FIN_WAIT2 state
 	 */
 	if (sock_flag(sk, SOCK_DEAD)) {
-		if (mptcp_check_close_timeout(sk)) {
+		if (mptcp_should_close(sk)) {
 			inet_sk_state_store(sk, TCP_CLOSE);
 			mptcp_do_fastclose(sk);
 		}
@@ -2639,9 +2665,6 @@ static void mptcp_worker(struct work_struct *work)
 		}
 	}
 
-	if (test_and_clear_bit(MPTCP_WORK_CLOSE_SUBFLOW, &msk->flags))
-		__mptcp_close_subflow(msk);
-
 	if (test_and_clear_bit(MPTCP_WORK_RTX, &msk->flags))
 		__mptcp_retrans(sk);
 
@@ -2878,6 +2901,14 @@ static void __mptcp_destroy_sock(struct sock *sk)
 	sock_put(sk);
 }
 
+void __mptcp_unaccepted_force_close(struct sock *sk)
+{
+	sock_set_flag(sk, SOCK_DEAD);
+	inet_sk_state_store(sk, TCP_CLOSE);
+	mptcp_do_fastclose(sk);
+	__mptcp_destroy_sock(sk);
+}
+
 static __poll_t mptcp_check_readable(struct mptcp_sock *msk)
 {
 	/* Concurrent splices from sk_receive_queue into receive_queue will
@@ -3079,6 +3110,7 @@ struct sock *mptcp_sk_clone(const struct sock *sk,
 	msk->local_key = subflow_req->local_key;
 	msk->token = subflow_req->token;
 	msk->subflow = NULL;
+	msk->in_accept_queue = 1;
 	WRITE_ONCE(msk->fully_established, false);
 	if (mp_opt->suboptions & OPTION_MPTCP_CSUMREQD)
 		WRITE_ONCE(msk->csum_enabled, true);
@@ -3096,8 +3128,7 @@ struct sock *mptcp_sk_clone(const struct sock *sk,
 	security_inet_csk_clone(nsk, req);
 	bh_unlock_sock(nsk);
 
-	/* keep a single reference */
-	__sock_put(nsk);
+	/* note: the newly allocated socket refcount is 2 now */
 	return nsk;
 }
 
@@ -3153,8 +3184,6 @@ static struct sock *mptcp_accept(struct sock *sk, int flags, int *err,
 			goto out;
 		}
 
-		/* acquire the 2nd reference for the owning socket */
-		sock_hold(new_mptcp_sock);
 		newsk = new_mptcp_sock;
 		MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_MPCAPABLEPASSIVEACK);
 	} else {
@@ -3705,25 +3734,10 @@ static int mptcp_stream_accept(struct socket *sock, struct socket *newsock,
 		struct sock *newsk = newsock->sk;
 
 		set_bit(SOCK_CUSTOM_SOCKOPT, &newsock->flags);
+		msk->in_accept_queue = 0;
 
 		lock_sock(newsk);
 
-		/* PM/worker can now acquire the first subflow socket
-		 * lock without racing with listener queue cleanup,
-		 * we can notify it, if needed.
-		 *
-		 * Even if remote has reset the initial subflow by now
-		 * the refcnt is still at least one.
-		 */
-		subflow = mptcp_subflow_ctx(msk->first);
-		list_add(&subflow->node, &msk->conn_list);
-		sock_hold(msk->first);
-		if (mptcp_is_fully_established(newsk))
-			mptcp_pm_fully_established(msk, msk->first, GFP_KERNEL);
-
-		mptcp_rcv_space_init(msk, msk->first);
-		mptcp_propagate_sndbuf(newsk, msk->first);
-
 		/* set ssk->sk_socket of accept()ed flows to mptcp socket.
 		 * This is needed so NOSPACE flag can be set from tcp stack.
 		 */
@@ -3733,6 +3747,18 @@ static int mptcp_stream_accept(struct socket *sock, struct socket *newsock,
 			if (!ssk->sk_socket)
 				mptcp_sock_graft(ssk, newsock);
 		}
+
+		/* Do late cleanup for the first subflow as necessary. Also
+		 * deal with bad peers not doing a complete shutdown.
+		 */
+		if (msk->first &&
+		    unlikely(inet_sk_state_load(msk->first) == TCP_CLOSE)) {
+			__mptcp_close_ssk(newsk, msk->first,
+					  mptcp_subflow_ctx(msk->first), 0);
+			if (unlikely(list_empty(&msk->conn_list)))
+				inet_sk_state_store(newsk, TCP_CLOSE);
+		}
+
 		release_sock(newsk);
 	}