Merge branch 'mptcp-MP_FAIL-timeout'

Mat Martineau says:

====================
mptcp: Timeout for MP_FAIL response

When one peer sends an infinite mapping to coordinate fallback from
MPTCP to regular TCP, the other peer is expected to send a packet with
the MPTCP MP_FAIL option to acknowledge the infinite mapping. Rather
than leave the connection in some half-fallback state, this series adds
a timeout after which the infinite mapping sender will reset the
connection.

Patch 1 adds a fallback self test.

Patches 2-5 make use of the MPTCP socket's retransmit timer to reset the
MPTCP connection if no MP_FAIL was received.

Patches 6 and 7 extends the self test to check MP_FAIL-related MIBs.
====================

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

4 files changed, 94 insertions, 3 deletions

diff --git a/net/mptcp/pm.c b/net/mptcp/pm.c
index 5c36870d3420..14f448d82bb2 100644
--- a/net/mptcp/pm.c
+++ b/net/mptcp/pm.c
@@ -287,11 +287,27 @@ void mptcp_pm_mp_fail_received(struct sock *sk, u64 fail_seq)
 {
 	struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(sk);
 	struct mptcp_sock *msk = mptcp_sk(subflow->conn);
+	struct sock *s = (struct sock *)msk;
 
 	pr_debug("fail_seq=%llu", fail_seq);
 
-	if (!mptcp_has_another_subflow(sk) && READ_ONCE(msk->allow_infinite_fallback))
+	if (mptcp_has_another_subflow(sk) || !READ_ONCE(msk->allow_infinite_fallback))
+		return;
+
+	if (!READ_ONCE(subflow->mp_fail_response_expect)) {
+		pr_debug("send MP_FAIL response and infinite map");
+
+		subflow->send_mp_fail = 1;
+		MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_MPFAILTX);
 		subflow->send_infinite_map = 1;
+	} else if (s && inet_sk_state_load(s) != TCP_CLOSE) {
+		pr_debug("MP_FAIL response received");
+
+		mptcp_data_lock(s);
+		if (inet_sk_state_load(s) != TCP_CLOSE)
+			sk_stop_timer(s, &s->sk_timer);
+		mptcp_data_unlock(s);
+	}
 }
 
 /* path manager helpers */
diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c
index 4581c570ef68..a5d466e6b538 100644
--- a/net/mptcp/protocol.c
+++ b/net/mptcp/protocol.c
@@ -1605,8 +1605,10 @@ void __mptcp_push_pending(struct sock *sk, unsigned int flags)
 
 out:
 	/* ensure the rtx timer is running */
+	mptcp_data_lock(sk);
 	if (!mptcp_timer_pending(sk))
 		mptcp_reset_timer(sk);
+	mptcp_data_unlock(sk);
 	if (copied)
 		__mptcp_check_send_data_fin(sk);
 }
@@ -2167,10 +2169,38 @@ static void mptcp_retransmit_timer(struct timer_list *t)
 	sock_put(sk);
 }
 
+static struct mptcp_subflow_context *
+mp_fail_response_expect_subflow(struct mptcp_sock *msk)
+{
+	struct mptcp_subflow_context *subflow, *ret = NULL;
+
+	mptcp_for_each_subflow(msk, subflow) {
+		if (READ_ONCE(subflow->mp_fail_response_expect)) {
+			ret = subflow;
+			break;
+		}
+	}
+
+	return ret;
+}
+
+static void mptcp_check_mp_fail_response(struct mptcp_sock *msk)
+{
+	struct mptcp_subflow_context *subflow;
+	struct sock *sk = (struct sock *)msk;
+
+	bh_lock_sock(sk);
+	subflow = mp_fail_response_expect_subflow(msk);
+	if (subflow)
+		__set_bit(MPTCP_FAIL_NO_RESPONSE, &msk->flags);
+	bh_unlock_sock(sk);
+}
+
 static void mptcp_timeout_timer(struct timer_list *t)
 {
 	struct sock *sk = from_timer(sk, t, sk_timer);
 
+	mptcp_check_mp_fail_response(mptcp_sk(sk));
 	mptcp_schedule_work(sk);
 	sock_put(sk);
 }
@@ -2491,8 +2521,27 @@ static void __mptcp_retrans(struct sock *sk)
 reset_timer:
 	mptcp_check_and_set_pending(sk);
 
+	mptcp_data_lock(sk);
 	if (!mptcp_timer_pending(sk))
 		mptcp_reset_timer(sk);
+	mptcp_data_unlock(sk);
+}
+
+static void mptcp_mp_fail_no_response(struct mptcp_sock *msk)
+{
+	struct mptcp_subflow_context *subflow;
+	struct sock *ssk;
+	bool slow;
+
+	subflow = mp_fail_response_expect_subflow(msk);
+	if (subflow) {
+		pr_debug("MP_FAIL doesn't respond, reset the subflow");
+
+		ssk = mptcp_subflow_tcp_sock(subflow);
+		slow = lock_sock_fast(ssk);
+		mptcp_subflow_reset(ssk);
+		unlock_sock_fast(ssk, slow);
+	}
 }
 
 static void mptcp_worker(struct work_struct *work)
@@ -2535,6 +2584,9 @@ static void mptcp_worker(struct work_struct *work)
 	if (test_and_clear_bit(MPTCP_WORK_RTX, &msk->flags))
 		__mptcp_retrans(sk);
 
+	if (test_and_clear_bit(MPTCP_FAIL_NO_RESPONSE, &msk->flags))
+		mptcp_mp_fail_no_response(msk);
+
 unlock:
 	release_sock(sk);
 	sock_put(sk);
@@ -2651,8 +2703,10 @@ void mptcp_subflow_shutdown(struct sock *sk, struct sock *ssk, int how)
 		} else {
 			pr_debug("Sending DATA_FIN on subflow %p", ssk);
 			tcp_send_ack(ssk);
+			mptcp_data_lock(sk);
 			if (!mptcp_timer_pending(sk))
 				mptcp_reset_timer(sk);
+			mptcp_data_unlock(sk);
 		}
 		break;
 	}
@@ -2753,8 +2807,10 @@ static void __mptcp_destroy_sock(struct sock *sk)
 	/* join list will be eventually flushed (with rst) at sock lock release time*/
 	list_splice_init(&msk->conn_list, &conn_list);
 
-	sk_stop_timer(sk, &msk->sk.icsk_retransmit_timer);
+	mptcp_data_lock(sk);
+	mptcp_stop_timer(sk);
 	sk_stop_timer(sk, &sk->sk_timer);
+	mptcp_data_unlock(sk);
 	msk->pm.status = 0;
 
 	/* clears msk->subflow, allowing the following loop to close
@@ -2816,7 +2872,9 @@ cleanup:
 		__mptcp_destroy_sock(sk);
 		do_cancel_work = true;
 	} else {
+		mptcp_data_lock(sk);
 		sk_reset_timer(sk, &sk->sk_timer, jiffies + TCP_TIMEWAIT_LEN);
+		mptcp_data_unlock(sk);
 	}
 	release_sock(sk);
 	if (do_cancel_work)
@@ -2861,8 +2919,10 @@ static int mptcp_disconnect(struct sock *sk, int flags)
 		__mptcp_close_ssk(sk, ssk, subflow, MPTCP_CF_FASTCLOSE);
 	}
 
-	sk_stop_timer(sk, &msk->sk.icsk_retransmit_timer);
+	mptcp_data_lock(sk);
+	mptcp_stop_timer(sk);
 	sk_stop_timer(sk, &sk->sk_timer);
+	mptcp_data_unlock(sk);
 
 	if (mptcp_sk(sk)->token)
 		mptcp_event(MPTCP_EVENT_CLOSED, mptcp_sk(sk), NULL, GFP_KERNEL);
diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h
index 61d600693ffd..3a8740fef918 100644
--- a/net/mptcp/protocol.h
+++ b/net/mptcp/protocol.h
@@ -116,6 +116,7 @@
 #define MPTCP_WORK_EOF		3
 #define MPTCP_FALLBACK_DONE	4
 #define MPTCP_WORK_CLOSE_SUBFLOW 5
+#define MPTCP_FAIL_NO_RESPONSE	6
 
 /* MPTCP socket release cb flags */
 #define MPTCP_PUSH_PENDING	1
@@ -448,6 +449,7 @@ struct mptcp_subflow_context {
 		stale : 1,	    /* unable to snd/rcv data, do not use for xmit */
 		local_id_valid : 1; /* local_id is correctly initialized */
 	enum mptcp_data_avail data_avail;
+	bool	mp_fail_response_expect;
 	u32	remote_nonce;
 	u64	thmac;
 	u32	local_nonce;
diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c
index 30ffb00661bb..75c824b67ca9 100644
--- a/net/mptcp/subflow.c
+++ b/net/mptcp/subflow.c
@@ -968,6 +968,7 @@ static enum mapping_status get_mapping_status(struct sock *ssk,
 {
 	struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(ssk);
 	bool csum_reqd = READ_ONCE(msk->csum_enabled);
+	struct sock *sk = (struct sock *)msk;
 	struct mptcp_ext *mpext;
 	struct sk_buff *skb;
 	u16 data_len;
@@ -1009,6 +1010,12 @@ static enum mapping_status get_mapping_status(struct sock *ssk,
 		pr_debug("infinite mapping received");
 		MPTCP_INC_STATS(sock_net(ssk), MPTCP_MIB_INFINITEMAPRX);
 		subflow->map_data_len = 0;
+		if (sk && inet_sk_state_load(sk) != TCP_CLOSE) {
+			mptcp_data_lock(sk);
+			if (inet_sk_state_load(sk) != TCP_CLOSE)
+				sk_stop_timer(sk, &sk->sk_timer);
+			mptcp_data_unlock(sk);
+		}
 		return MAPPING_INVALID;
 	}
 
@@ -1217,6 +1224,12 @@ fallback:
 				tcp_send_active_reset(ssk, GFP_ATOMIC);
 				while ((skb = skb_peek(&ssk->sk_receive_queue)))
 					sk_eat_skb(ssk, skb);
+			} else {
+				WRITE_ONCE(subflow->mp_fail_response_expect, true);
+				/* The data lock is acquired in __mptcp_move_skbs() */
+				sk_reset_timer((struct sock *)msk,
+					       &((struct sock *)msk)->sk_timer,
+					       jiffies + TCP_RTO_MAX);
 			}
 			WRITE_ONCE(subflow->data_avail, MPTCP_SUBFLOW_NODATA);
 			return true;