Merge branch 'mptcp-fixes'

Mat Martineau says:

====================
mptcp: More v5.13 fixes

Here's another batch of MPTCP fixes for v5.13.

Patch 1 cleans up memory accounting between the MPTCP-level socket and
the subflows to more reliably transfer forward allocated memory under
pressure.

Patch 2 wakes up socket readers more reliably.

Patch 3 changes a WARN_ONCE to a pr_debug.

Patch 4 changes the selftests to only use syncookies in test cases where
they do not cause spurious failures.

Patch 5 modifies socket error reporting to avoid a possible soft lockup.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

4 files changed, 88 insertions, 84 deletions

diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c
index 5edc686faff1..632350018fb6 100644
--- a/net/mptcp/protocol.c
+++ b/net/mptcp/protocol.c
@@ -280,11 +280,13 @@ static bool __mptcp_move_skb(struct mptcp_sock *msk, struct sock *ssk,
 
 	/* try to fetch required memory from subflow */
 	if (!sk_rmem_schedule(sk, skb, skb->truesize)) {
-		if (ssk->sk_forward_alloc < skb->truesize)
-			goto drop;
-		__sk_mem_reclaim(ssk, skb->truesize);
-		if (!sk_rmem_schedule(sk, skb, skb->truesize))
+		int amount = sk_mem_pages(skb->truesize) << SK_MEM_QUANTUM_SHIFT;
+
+		if (ssk->sk_forward_alloc < amount)
 			goto drop;
+
+		ssk->sk_forward_alloc -= amount;
+		sk->sk_forward_alloc += amount;
 	}
 
 	/* the skb map_seq accounts for the skb offset:
@@ -668,18 +670,22 @@ static bool __mptcp_ofo_queue(struct mptcp_sock *msk)
 /* In most cases we will be able to lock the mptcp socket.  If its already
  * owned, we need to defer to the work queue to avoid ABBA deadlock.
  */
-static void move_skbs_to_msk(struct mptcp_sock *msk, struct sock *ssk)
+static bool move_skbs_to_msk(struct mptcp_sock *msk, struct sock *ssk)
 {
 	struct sock *sk = (struct sock *)msk;
 	unsigned int moved = 0;
 
 	if (inet_sk_state_load(sk) == TCP_CLOSE)
-		return;
-
-	mptcp_data_lock(sk);
+		return false;
 
 	__mptcp_move_skbs_from_subflow(msk, ssk, &moved);
 	__mptcp_ofo_queue(msk);
+	if (unlikely(ssk->sk_err)) {
+		if (!sock_owned_by_user(sk))
+			__mptcp_error_report(sk);
+		else
+			set_bit(MPTCP_ERROR_REPORT,  &msk->flags);
+	}
 
 	/* If the moves have caught up with the DATA_FIN sequence number
 	 * it's time to ack the DATA_FIN and change socket state, but
@@ -688,7 +694,7 @@ static void move_skbs_to_msk(struct mptcp_sock *msk, struct sock *ssk)
 	 */
 	if (mptcp_pending_data_fin(sk, NULL))
 		mptcp_schedule_work(sk);
-	mptcp_data_unlock(sk);
+	return moved > 0;
 }
 
 void mptcp_data_ready(struct sock *sk, struct sock *ssk)
@@ -696,7 +702,6 @@ void mptcp_data_ready(struct sock *sk, struct sock *ssk)
 	struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(ssk);
 	struct mptcp_sock *msk = mptcp_sk(sk);
 	int sk_rbuf, ssk_rbuf;
-	bool wake;
 
 	/* The peer can send data while we are shutting down this
 	 * subflow at msk destruction time, but we must avoid enqueuing
@@ -705,28 +710,22 @@ void mptcp_data_ready(struct sock *sk, struct sock *ssk)
 	if (unlikely(subflow->disposable))
 		return;
 
-	/* move_skbs_to_msk below can legitly clear the data_avail flag,
-	 * but we will need later to properly woke the reader, cache its
-	 * value
-	 */
-	wake = subflow->data_avail == MPTCP_SUBFLOW_DATA_AVAIL;
-	if (wake)
-		set_bit(MPTCP_DATA_READY, &msk->flags);
-
 	ssk_rbuf = READ_ONCE(ssk->sk_rcvbuf);
 	sk_rbuf = READ_ONCE(sk->sk_rcvbuf);
 	if (unlikely(ssk_rbuf > sk_rbuf))
 		sk_rbuf = ssk_rbuf;
 
-	/* over limit? can't append more skbs to msk */
+	/* over limit? can't append more skbs to msk, Also, no need to wake-up*/
 	if (atomic_read(&sk->sk_rmem_alloc) > sk_rbuf)
-		goto wake;
-
-	move_skbs_to_msk(msk, ssk);
+		return;
 
-wake:
-	if (wake)
+	/* Wake-up the reader only for in-sequence data */
+	mptcp_data_lock(sk);
+	if (move_skbs_to_msk(msk, ssk)) {
+		set_bit(MPTCP_DATA_READY, &msk->flags);
 		sk->sk_data_ready(sk);
+	}
+	mptcp_data_unlock(sk);
 }
 
 static bool mptcp_do_flush_join_list(struct mptcp_sock *msk)
@@ -858,7 +857,7 @@ static struct sock *mptcp_subflow_recv_lookup(const struct mptcp_sock *msk)
 	sock_owned_by_me(sk);
 
 	mptcp_for_each_subflow(msk, subflow) {
-		if (subflow->data_avail)
+		if (READ_ONCE(subflow->data_avail))
 			return mptcp_subflow_tcp_sock(subflow);
 	}
 
@@ -1955,6 +1954,9 @@ static bool __mptcp_move_skbs(struct mptcp_sock *msk)
 		done = __mptcp_move_skbs_from_subflow(msk, ssk, &moved);
 		mptcp_data_unlock(sk);
 		tcp_cleanup_rbuf(ssk, moved);
+
+		if (unlikely(ssk->sk_err))
+			__mptcp_error_report(sk);
 		unlock_sock_fast(ssk, slowpath);
 	} while (!done);
 
diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h
index 0c6f99c67345..385796f0ef19 100644
--- a/net/mptcp/protocol.h
+++ b/net/mptcp/protocol.h
@@ -362,7 +362,6 @@ mptcp_subflow_rsk(const struct request_sock *rsk)
 enum mptcp_data_avail {
 	MPTCP_SUBFLOW_NODATA,
 	MPTCP_SUBFLOW_DATA_AVAIL,
-	MPTCP_SUBFLOW_OOO_DATA
 };
 
 struct mptcp_delegated_action {
diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c
index ef3d037f984a..be1de4084196 100644
--- a/net/mptcp/subflow.c
+++ b/net/mptcp/subflow.c
@@ -784,10 +784,10 @@ static u64 expand_seq(u64 old_seq, u16 old_data_len, u64 seq)
 	return seq | ((old_seq + old_data_len + 1) & GENMASK_ULL(63, 32));
 }
 
-static void warn_bad_map(struct mptcp_subflow_context *subflow, u32 ssn)
+static void dbg_bad_map(struct mptcp_subflow_context *subflow, u32 ssn)
 {
-	WARN_ONCE(1, "Bad mapping: ssn=%d map_seq=%d map_data_len=%d",
-		  ssn, subflow->map_subflow_seq, subflow->map_data_len);
+	pr_debug("Bad mapping: ssn=%d map_seq=%d map_data_len=%d",
+		 ssn, subflow->map_subflow_seq, subflow->map_data_len);
 }
 
 static bool skb_is_fully_mapped(struct sock *ssk, struct sk_buff *skb)
@@ -812,13 +812,13 @@ static bool validate_mapping(struct sock *ssk, struct sk_buff *skb)
 		/* Mapping covers data later in the subflow stream,
 		 * currently unsupported.
 		 */
-		warn_bad_map(subflow, ssn);
+		dbg_bad_map(subflow, ssn);
 		return false;
 	}
 	if (unlikely(!before(ssn, subflow->map_subflow_seq +
 				  subflow->map_data_len))) {
 		/* Mapping does covers past subflow data, invalid */
-		warn_bad_map(subflow, ssn + skb->len);
+		dbg_bad_map(subflow, ssn);
 		return false;
 	}
 	return true;
@@ -1000,7 +1000,7 @@ static bool subflow_check_data_avail(struct sock *ssk)
 	struct sk_buff *skb;
 
 	if (!skb_peek(&ssk->sk_receive_queue))
-		subflow->data_avail = 0;
+		WRITE_ONCE(subflow->data_avail, 0);
 	if (subflow->data_avail)
 		return true;
 
@@ -1039,18 +1039,13 @@ static bool subflow_check_data_avail(struct sock *ssk)
 		ack_seq = mptcp_subflow_get_mapped_dsn(subflow);
 		pr_debug("msk ack_seq=%llx subflow ack_seq=%llx", old_ack,
 			 ack_seq);
-		if (ack_seq == old_ack) {
-			subflow->data_avail = MPTCP_SUBFLOW_DATA_AVAIL;
-			break;
-		} else if (after64(ack_seq, old_ack)) {
-			subflow->data_avail = MPTCP_SUBFLOW_OOO_DATA;
-			break;
+		if (unlikely(before64(ack_seq, old_ack))) {
+			mptcp_subflow_discard_data(ssk, skb, old_ack - ack_seq);
+			continue;
 		}
 
-		/* only accept in-sequence mapping. Old values are spurious
-		 * retransmission
-		 */
-		mptcp_subflow_discard_data(ssk, skb, old_ack - ack_seq);
+		WRITE_ONCE(subflow->data_avail, MPTCP_SUBFLOW_DATA_AVAIL);
+		break;
 	}
 	return true;
 
@@ -1065,12 +1060,11 @@ fallback:
 		 * subflow_error_report() will introduce the appropriate barriers
 		 */
 		ssk->sk_err = EBADMSG;
-		ssk->sk_error_report(ssk);
 		tcp_set_state(ssk, TCP_CLOSE);
 		subflow->reset_transient = 0;
 		subflow->reset_reason = MPTCP_RST_EMPTCP;
 		tcp_send_active_reset(ssk, GFP_ATOMIC);
-		subflow->data_avail = 0;
+		WRITE_ONCE(subflow->data_avail, 0);
 		return false;
 	}
 
@@ -1080,7 +1074,7 @@ fallback:
 	subflow->map_seq = READ_ONCE(msk->ack_seq);
 	subflow->map_data_len = skb->len;
 	subflow->map_subflow_seq = tcp_sk(ssk)->copied_seq - subflow->ssn_offset;
-	subflow->data_avail = MPTCP_SUBFLOW_DATA_AVAIL;
+	WRITE_ONCE(subflow->data_avail, MPTCP_SUBFLOW_DATA_AVAIL);
 	return true;
 }
 
@@ -1092,7 +1086,7 @@ bool mptcp_subflow_data_available(struct sock *sk)
 	if (subflow->map_valid &&
 	    mptcp_subflow_get_map_offset(subflow) >= subflow->map_data_len) {
 		subflow->map_valid = 0;
-		subflow->data_avail = 0;
+		WRITE_ONCE(subflow->data_avail, 0);
 
 		pr_debug("Done with mapping: seq=%u data_len=%u",
 			 subflow->map_subflow_seq,
@@ -1120,41 +1114,6 @@ void mptcp_space(const struct sock *ssk, int *space, int *full_space)
 	*full_space = tcp_full_space(sk);
 }
 
-static void subflow_data_ready(struct sock *sk)
-{
-	struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(sk);
-	u16 state = 1 << inet_sk_state_load(sk);
-	struct sock *parent = subflow->conn;
-	struct mptcp_sock *msk;
-
-	msk = mptcp_sk(parent);
-	if (state & TCPF_LISTEN) {
-		/* MPJ subflow are removed from accept queue before reaching here,
-		 * avoid stray wakeups
-		 */
-		if (reqsk_queue_empty(&inet_csk(sk)->icsk_accept_queue))
-			return;
-
-		set_bit(MPTCP_DATA_READY, &msk->flags);
-		parent->sk_data_ready(parent);
-		return;
-	}
-
-	WARN_ON_ONCE(!__mptcp_check_fallback(msk) && !subflow->mp_capable &&
-		     !subflow->mp_join && !(state & TCPF_CLOSE));
-
-	if (mptcp_subflow_data_available(sk))
-		mptcp_data_ready(parent, sk);
-}
-
-static void subflow_write_space(struct sock *ssk)
-{
-	struct sock *sk = mptcp_subflow_ctx(ssk)->conn;
-
-	mptcp_propagate_sndbuf(sk, ssk);
-	mptcp_write_space(sk);
-}
-
 void __mptcp_error_report(struct sock *sk)
 {
 	struct mptcp_subflow_context *subflow;
@@ -1195,6 +1154,43 @@ static void subflow_error_report(struct sock *ssk)
 	mptcp_data_unlock(sk);
 }
 
+static void subflow_data_ready(struct sock *sk)
+{
+	struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(sk);
+	u16 state = 1 << inet_sk_state_load(sk);
+	struct sock *parent = subflow->conn;
+	struct mptcp_sock *msk;
+
+	msk = mptcp_sk(parent);
+	if (state & TCPF_LISTEN) {
+		/* MPJ subflow are removed from accept queue before reaching here,
+		 * avoid stray wakeups
+		 */
+		if (reqsk_queue_empty(&inet_csk(sk)->icsk_accept_queue))
+			return;
+
+		set_bit(MPTCP_DATA_READY, &msk->flags);
+		parent->sk_data_ready(parent);
+		return;
+	}
+
+	WARN_ON_ONCE(!__mptcp_check_fallback(msk) && !subflow->mp_capable &&
+		     !subflow->mp_join && !(state & TCPF_CLOSE));
+
+	if (mptcp_subflow_data_available(sk))
+		mptcp_data_ready(parent, sk);
+	else if (unlikely(sk->sk_err))
+		subflow_error_report(sk);
+}
+
+static void subflow_write_space(struct sock *ssk)
+{
+	struct sock *sk = mptcp_subflow_ctx(ssk)->conn;
+
+	mptcp_propagate_sndbuf(sk, ssk);
+	mptcp_write_space(sk);
+}
+
 static struct inet_connection_sock_af_ops *
 subflow_default_af_ops(struct sock *sk)
 {
@@ -1505,6 +1501,8 @@ static void subflow_state_change(struct sock *sk)
 	 */
 	if (mptcp_subflow_data_available(sk))
 		mptcp_data_ready(parent, sk);
+	else if (unlikely(sk->sk_err))
+		subflow_error_report(sk);
 
 	subflow_sched_work_if_closed(mptcp_sk(parent), sk);
 
diff --git a/tools/testing/selftests/net/mptcp/mptcp_connect.sh b/tools/testing/selftests/net/mptcp/mptcp_connect.sh
index 9ca5f1ba461e..2b495dc8d78e 100755
--- a/tools/testing/selftests/net/mptcp/mptcp_connect.sh
+++ b/tools/testing/selftests/net/mptcp/mptcp_connect.sh
@@ -197,9 +197,6 @@ ip -net "$ns4" link set ns4eth3 up
 ip -net "$ns4" route add default via 10.0.3.2
 ip -net "$ns4" route add default via dead:beef:3::2
 
-# use TCP syn cookies, even if no flooding was detected.
-ip netns exec "$ns2" sysctl -q net.ipv4.tcp_syncookies=2
-
 set_ethtool_flags() {
 	local ns="$1"
 	local dev="$2"
@@ -737,6 +734,14 @@ for sender in $ns1 $ns2 $ns3 $ns4;do
 		exit $ret
 	fi
 
+	# ns1<->ns2 is not subject to reordering/tc delays. Use it to test
+	# mptcp syncookie support.
+	if [ $sender = $ns1 ]; then
+		ip netns exec "$ns2" sysctl -q net.ipv4.tcp_syncookies=2
+	else
+		ip netns exec "$ns2" sysctl -q net.ipv4.tcp_syncookies=1
+	fi
+
 	run_tests "$ns2" $sender 10.0.1.2
 	run_tests "$ns2" $sender dead:beef:1::2
 	run_tests "$ns2" $sender 10.0.2.1