1 files changed, 36 insertions, 19 deletions

diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c
index c2a8392254dc..a57f3eab7b6a 100644
--- a/net/mptcp/protocol.c
+++ b/net/mptcp/protocol.c
@@ -363,8 +363,6 @@ static void mptcp_check_data_fin_ack(struct sock *sk)
 
 	/* Look for an acknowledged DATA_FIN */
 	if (mptcp_pending_data_fin_ack(sk)) {
-		mptcp_stop_timer(sk);
-
 		WRITE_ONCE(msk->snd_data_fin_enable, 0);
 
 		switch (sk->sk_state) {
@@ -458,7 +456,18 @@ static bool mptcp_subflow_cleanup_rbuf(struct sock *ssk)
 static void mptcp_cleanup_rbuf(struct mptcp_sock *msk)
 {
 	struct sock *ack_hint = READ_ONCE(msk->ack_hint);
+	int old_space = READ_ONCE(msk->old_wspace);
 	struct mptcp_subflow_context *subflow;
+	struct sock *sk = (struct sock *)msk;
+	bool cleanup;
+
+	/* this is a simple superset of what tcp_cleanup_rbuf() implements
+	 * so that we don't have to acquire the ssk socket lock most of the time
+	 * to do actually nothing
+	 */
+	cleanup = __mptcp_space(sk) - old_space >= max(0, old_space);
+	if (!cleanup)
+		return;
 
 	/* if the hinted ssk is still active, try to use it */
 	if (likely(ack_hint)) {
@@ -1565,6 +1574,9 @@ out:
 		mptcp_set_timeout(sk, ssk);
 		tcp_push(ssk, 0, info.mss_now, tcp_sk(ssk)->nonagle,
 			 info.size_goal);
+		if (!mptcp_timer_pending(sk))
+			mptcp_reset_timer(sk);
+
 		if (msk->snd_data_fin_enable &&
 		    msk->snd_nxt + 1 == msk->write_seq)
 			mptcp_schedule_work(sk);
@@ -1868,7 +1880,7 @@ static void __mptcp_splice_receive_queue(struct sock *sk)
 	skb_queue_splice_tail_init(&sk->sk_receive_queue, &msk->receive_queue);
 }
 
-static bool __mptcp_move_skbs(struct mptcp_sock *msk, unsigned int rcv)
+static bool __mptcp_move_skbs(struct mptcp_sock *msk)
 {
 	struct sock *sk = (struct sock *)msk;
 	unsigned int moved = 0;
@@ -1888,13 +1900,10 @@ static bool __mptcp_move_skbs(struct mptcp_sock *msk, unsigned int rcv)
 
 		slowpath = lock_sock_fast(ssk);
 		mptcp_data_lock(sk);
+		__mptcp_update_rmem(sk);
 		done = __mptcp_move_skbs_from_subflow(msk, ssk, &moved);
 		mptcp_data_unlock(sk);
-		if (moved && rcv) {
-			WRITE_ONCE(msk->rmem_pending, min(rcv, moved));
-			tcp_cleanup_rbuf(ssk, 1);
-			WRITE_ONCE(msk->rmem_pending, 0);
-		}
+		tcp_cleanup_rbuf(ssk, moved);
 		unlock_sock_fast(ssk, slowpath);
 	} while (!done);
 
@@ -1907,6 +1916,7 @@ static bool __mptcp_move_skbs(struct mptcp_sock *msk, unsigned int rcv)
 		ret |= __mptcp_ofo_queue(msk);
 		__mptcp_splice_receive_queue(sk);
 		mptcp_data_unlock(sk);
+		mptcp_cleanup_rbuf(msk);
 	}
 	if (ret)
 		mptcp_check_data_fin((struct sock *)msk);
@@ -1936,7 +1946,7 @@ static int mptcp_recvmsg(struct sock *sk, struct msghdr *msg, size_t len,
 	target = sock_rcvlowat(sk, flags & MSG_WAITALL, len);
 
 	while (copied < len) {
-		int bytes_read, old_space;
+		int bytes_read;
 
 		bytes_read = __mptcp_recvmsg_mskq(msk, msg, len - copied);
 		if (unlikely(bytes_read < 0)) {
@@ -1947,14 +1957,11 @@ static int mptcp_recvmsg(struct sock *sk, struct msghdr *msg, size_t len,
 
 		copied += bytes_read;
 
-		if (skb_queue_empty(&msk->receive_queue) &&
-		    __mptcp_move_skbs(msk, len - copied))
-			continue;
-
 		/* be sure to advertise window change */
-		old_space = READ_ONCE(msk->old_wspace);
-		if ((tcp_space(sk) - old_space) >= old_space)
-			mptcp_cleanup_rbuf(msk);
+		mptcp_cleanup_rbuf(msk);
+
+		if (skb_queue_empty(&msk->receive_queue) && __mptcp_move_skbs(msk))
+			continue;
 
 		/* only the master socket status is relevant here. The exit
 		 * conditions mirror closely tcp_recvmsg()
@@ -1982,7 +1989,7 @@ static int mptcp_recvmsg(struct sock *sk, struct msghdr *msg, size_t len,
 				/* race breaker: the shutdown could be after the
 				 * previous receive queue check
 				 */
-				if (__mptcp_move_skbs(msk, len - copied))
+				if (__mptcp_move_skbs(msk))
 					continue;
 				break;
 			}
@@ -2015,7 +2022,7 @@ static int mptcp_recvmsg(struct sock *sk, struct msghdr *msg, size_t len,
 		/* .. race-breaker: ssk might have gotten new data
 		 * after last __mptcp_move_skbs() returned false.
 		 */
-		if (unlikely(__mptcp_move_skbs(msk, 0)))
+		if (unlikely(__mptcp_move_skbs(msk)))
 			set_bit(MPTCP_DATA_READY, &msk->flags);
 	} else if (unlikely(!test_bit(MPTCP_DATA_READY, &msk->flags))) {
 		/* data to read but mptcp_wait_data() cleared DATA_READY */
@@ -2275,6 +2282,7 @@ static void mptcp_worker(struct work_struct *work)
 	if (!test_and_clear_bit(MPTCP_WORK_RTX, &msk->flags))
 		goto unlock;
 
+	__mptcp_clean_una(sk);
 	dfrag = mptcp_rtx_head(sk);
 	if (!dfrag)
 		goto unlock;
@@ -2943,6 +2951,8 @@ static void mptcp_release_cb(struct sock *sk)
 		mptcp_push_pending(sk, 0);
 		spin_lock_bh(&sk->sk_lock.slock);
 	}
+	if (test_and_clear_bit(MPTCP_ERROR_REPORT, &mptcp_sk(sk)->flags))
+		__mptcp_error_report(sk);
 
 	/* clear any wmem reservation and errors */
 	__mptcp_update_wmem(sk);
@@ -3319,7 +3329,7 @@ static __poll_t mptcp_check_writeable(struct mptcp_sock *msk)
 	struct sock *sk = (struct sock *)msk;
 
 	if (unlikely(sk->sk_shutdown & SEND_SHUTDOWN))
-		return 0;
+		return EPOLLOUT | EPOLLWRNORM;
 
 	if (sk_stream_is_writeable(sk))
 		return EPOLLOUT | EPOLLWRNORM;
@@ -3352,9 +3362,16 @@ static __poll_t mptcp_poll(struct file *file, struct socket *sock,
 		mask |= mptcp_check_readable(msk);
 		mask |= mptcp_check_writeable(msk);
 	}
+	if (sk->sk_shutdown == SHUTDOWN_MASK || state == TCP_CLOSE)
+		mask |= EPOLLHUP;
 	if (sk->sk_shutdown & RCV_SHUTDOWN)
 		mask |= EPOLLIN | EPOLLRDNORM | EPOLLRDHUP;
 
+	/* This barrier is coupled with smp_wmb() in tcp_reset() */
+	smp_rmb();
+	if (sk->sk_err)
+		mask |= EPOLLERR;
+
 	return mask;
 }