Merge branch 'mptcp-msg_fastopen-and-tfo-listener-side-support'

Matthieu Baerts says:

====================
mptcp: MSG_FASTOPEN and TFO listener side support

Before this series, only the initiator of a connection was able to combine
both TCP FastOpen and MPTCP when using TCP_FASTOPEN_CONNECT socket option.

These new patches here add (in theory) the full support of TFO with MPTCP,
which means:

 - MSG_FASTOPEN sendmsg flag support (patch 1/8)
 - TFO support for the listener side (patches 2-5/8)
 - TCP_FASTOPEN socket option (patch 6/8)
 - TCP_FASTOPEN_KEY socket option (patch 7/8)

To support TFO for the server side, a few preparation patches are needed
(patches 2 to 5/8). Some of them were inspired by a previous work from
Benjamin Hesmans.

Note that TFO support with MPTCP has been validated with selftests
(patch 8/8) but also with Packetdrill tests running with a modified
but still very WIP version supporting MPTCP. Both the modified tool
and the tests are available online:

  https://github.com/multipath-tcp/packetdrill/
====================

Link: https://lore.kernel.org/r/20221125222958.958636-1-matthieu.baerts@tessares.netSigned-off-by: Jakub Kicinski <kuba@kernel.org>

net/mptcp/Makefile

@@ -2,7 +2,7 @@
 obj-$(CONFIG_MPTCP) += mptcp.o
 
 mptcp-y := protocol.o subflow.o options.o token.o crypto.o ctrl.o pm.o diag.o \
-	   mib.o pm_netlink.o sockopt.o pm_userspace.o
+	   mib.o pm_netlink.o sockopt.o pm_userspace.o fastopen.o
 
 obj-$(CONFIG_SYN_COOKIES) += syncookies.o
 obj-$(CONFIG_INET_MPTCP_DIAG) += mptcp_diag.o

net/mptcp/fastopen.c

+// SPDX-License-Identifier: GPL-2.0
+/* MPTCP Fast Open Mechanism
+ *
+ * Copyright (c) 2021-2022, Dmytro SHYTYI
+ */
+
+#include "protocol.h"
+
+void mptcp_fastopen_subflow_synack_set_params(struct mptcp_subflow_context *subflow,
+					      struct request_sock *req)
+{
+	struct sock *ssk = subflow->tcp_sock;
+	struct sock *sk = subflow->conn;
+	struct sk_buff *skb;
+	struct tcp_sock *tp;
+
+	tp = tcp_sk(ssk);
+
+	subflow->is_mptfo = 1;
+
+	skb = skb_peek(&ssk->sk_receive_queue);
+	if (WARN_ON_ONCE(!skb))
+		return;
+
+	/* dequeue the skb from sk receive queue */
+	__skb_unlink(skb, &ssk->sk_receive_queue);
+	skb_ext_reset(skb);
+	skb_orphan(skb);
+
+	/* We copy the fastopen data, but that don't belong to the mptcp sequence
+	 * space, need to offset it in the subflow sequence, see mptcp_subflow_get_map_offset()
+	 */
+	tp->copied_seq += skb->len;
+	subflow->ssn_offset += skb->len;
+
+	/* initialize a dummy sequence number, we will update it at MPC
+	 * completion, if needed
+	 */
+	MPTCP_SKB_CB(skb)->map_seq = -skb->len;
+	MPTCP_SKB_CB(skb)->end_seq = 0;
+	MPTCP_SKB_CB(skb)->offset = 0;
+	MPTCP_SKB_CB(skb)->has_rxtstamp = TCP_SKB_CB(skb)->has_rxtstamp;
+
+	mptcp_data_lock(sk);
+
+	mptcp_set_owner_r(skb, sk);
+	__skb_queue_tail(&sk->sk_receive_queue, skb);
+
+	sk->sk_data_ready(sk);
+
+	mptcp_data_unlock(sk);
+}
+
+void mptcp_fastopen_gen_msk_ackseq(struct mptcp_sock *msk, struct mptcp_subflow_context *subflow,
+				   const struct mptcp_options_received *mp_opt)
+{
+	struct sock *sk = (struct sock *)msk;
+	struct sk_buff *skb;
+
+	mptcp_data_lock(sk);
+	skb = skb_peek_tail(&sk->sk_receive_queue);
+	if (skb) {
+		WARN_ON_ONCE(MPTCP_SKB_CB(skb)->end_seq);
+		pr_debug("msk %p moving seq %llx -> %llx end_seq %llx -> %llx", sk,
+			 MPTCP_SKB_CB(skb)->map_seq, MPTCP_SKB_CB(skb)->map_seq + msk->ack_seq,
+			 MPTCP_SKB_CB(skb)->end_seq, MPTCP_SKB_CB(skb)->end_seq + msk->ack_seq);
+		MPTCP_SKB_CB(skb)->map_seq += msk->ack_seq;
+		MPTCP_SKB_CB(skb)->end_seq += msk->ack_seq;
+	}
+
+	pr_debug("msk=%p ack_seq=%llx", msk, msk->ack_seq);
+	mptcp_data_unlock(sk);
+}

net/mptcp/options.c

@@ -26,6 +26,7 @@ static void mptcp_parse_option(const struct sk_buff *skb,
 {
 	u8 subtype = *ptr >> 4;
 	int expected_opsize;
+	u16 subopt;
 	u8 version;
 	u8 flags;
 	u8 i;
@@ -38,11 +39,15 @@ static void mptcp_parse_option(const struct sk_buff *skb,
 				expected_opsize = TCPOLEN_MPTCP_MPC_ACK_DATA;
 			else
 				expected_opsize = TCPOLEN_MPTCP_MPC_ACK;
+			subopt = OPTION_MPTCP_MPC_ACK;
 		} else {
-			if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_ACK)
+			if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_ACK) {
 				expected_opsize = TCPOLEN_MPTCP_MPC_SYNACK;
-			else
+				subopt = OPTION_MPTCP_MPC_SYNACK;
+			} else {
 				expected_opsize = TCPOLEN_MPTCP_MPC_SYN;
+				subopt = OPTION_MPTCP_MPC_SYN;
+			}
 		}
 
 		/* Cfr RFC 8684 Section 3.3.0:
@@ -85,7 +90,7 @@ static void mptcp_parse_option(const struct sk_buff *skb,
 
 		mp_opt->deny_join_id0 = !!(flags & MPTCP_CAP_DENY_JOIN_ID0);
 
-		mp_opt->suboptions |= OPTIONS_MPTCP_MPC;
+		mp_opt->suboptions |= subopt;
 		if (opsize >= TCPOLEN_MPTCP_MPC_SYNACK) {
 			mp_opt->sndr_key = get_unaligned_be64(ptr);
 			ptr += 8;
@@ -934,7 +939,7 @@ static bool check_fully_established(struct mptcp_sock *msk, struct sock *ssk,
 		    subflow->mp_join && (mp_opt->suboptions & OPTIONS_MPTCP_MPJ) &&
 		    !subflow->request_join)
 			tcp_send_ack(ssk);
-		goto fully_established;
+		goto check_notify;
 	}
 
 	/* we must process OoO packets before the first subflow is fully
@@ -945,17 +950,20 @@ static bool check_fully_established(struct mptcp_sock *msk, struct sock *ssk,
 	if (TCP_SKB_CB(skb)->seq != subflow->ssn_offset + 1) {
 		if (subflow->mp_join)
 			goto reset;
+		if (subflow->is_mptfo && mp_opt->suboptions & OPTION_MPTCP_MPC_ACK)
+			goto set_fully_established;
 		return subflow->mp_capable;
 	}
 
-	if (((mp_opt->suboptions & OPTION_MPTCP_DSS) && mp_opt->use_ack) ||
-	    ((mp_opt->suboptions & OPTION_MPTCP_ADD_ADDR) && !mp_opt->echo)) {
+	if (subflow->remote_key_valid &&
+	    (((mp_opt->suboptions & OPTION_MPTCP_DSS) && mp_opt->use_ack) ||
+	     ((mp_opt->suboptions & OPTION_MPTCP_ADD_ADDR) && !mp_opt->echo))) {
 		/* subflows are fully established as soon as we get any
 		 * additional ack, including ADD_ADDR.
 		 */
 		subflow->fully_established = 1;
 		WRITE_ONCE(msk->fully_established, true);
-		goto fully_established;
+		goto check_notify;
 	}
 
 	/* If the first established packet does not contain MP_CAPABLE + data
@@ -974,11 +982,12 @@ static bool check_fully_established(struct mptcp_sock *msk, struct sock *ssk,
 	if (mp_opt->deny_join_id0)
 		WRITE_ONCE(msk->pm.remote_deny_join_id0, true);
 
+set_fully_established:
 	if (unlikely(!READ_ONCE(msk->pm.server_side)))
 		pr_warn_once("bogus mpc option on established client sk");
 	mptcp_subflow_fully_established(subflow, mp_opt);
 
-fully_established:
+check_notify:
 	/* if the subflow is not already linked into the conn_list, we can't
 	 * notify the PM: this subflow is still on the listener queue
 	 * and the PM possibly acquiring the subflow lock could race with

net/mptcp/protocol.c

@@ -36,15 +36,6 @@ struct mptcp6_sock {
 };
 #endif
 
-struct mptcp_skb_cb {
-	u64 map_seq;
-	u64 end_seq;
-	u32 offset;
-	u8  has_rxtstamp:1;
-};
-
-#define MPTCP_SKB_CB(__skb)	((struct mptcp_skb_cb *)&((__skb)->cb[0]))
-
 enum {
 	MPTCP_CMSG_TS = BIT(0),
 	MPTCP_CMSG_INQ = BIT(1),
@@ -200,7 +191,7 @@ static void mptcp_rfree(struct sk_buff *skb)
 	mptcp_rmem_uncharge(sk, len);
 }
 
-static void mptcp_set_owner_r(struct sk_buff *skb, struct sock *sk)
+void mptcp_set_owner_r(struct sk_buff *skb, struct sock *sk)
 {
 	skb_orphan(skb);
 	skb->sk = sk;
@@ -1711,17 +1702,14 @@ static int mptcp_sendmsg(struct sock *sk, struct msghdr *msg, size_t len)
 	int ret = 0;
 	long timeo;
 
-	/* we don't support FASTOPEN yet */
-	if (msg->msg_flags & MSG_FASTOPEN)
-		return -EOPNOTSUPP;
-
 	/* silently ignore everything else */
-	msg->msg_flags &= MSG_MORE | MSG_DONTWAIT | MSG_NOSIGNAL;
+	msg->msg_flags &= MSG_MORE | MSG_DONTWAIT | MSG_NOSIGNAL | MSG_FASTOPEN;
 
 	lock_sock(sk);
 
 	ssock = __mptcp_nmpc_socket(msk);
-	if (unlikely(ssock && inet_sk(ssock->sk)->defer_connect)) {
+	if (unlikely(ssock && (inet_sk(ssock->sk)->defer_connect ||
+			       msg->msg_flags & MSG_FASTOPEN))) {
 		int copied_syn = 0;
 
 		ret = mptcp_sendmsg_fastopen(sk, ssock->sk, msg, len, &copied_syn);
@@ -3048,7 +3036,6 @@ struct sock *mptcp_sk_clone(const struct sock *sk,
 	struct mptcp_subflow_request_sock *subflow_req = mptcp_subflow_rsk(req);
 	struct sock *nsk = sk_clone_lock(sk, GFP_ATOMIC);
 	struct mptcp_sock *msk;
-	u64 ack_seq;
 
 	if (!nsk)
 		return NULL;
@@ -3074,15 +3061,6 @@ struct sock *mptcp_sk_clone(const struct sock *sk,
 	msk->wnd_end = msk->snd_nxt + req->rsk_rcv_wnd;
 	msk->setsockopt_seq = mptcp_sk(sk)->setsockopt_seq;
 
-	if (mp_opt->suboptions & OPTIONS_MPTCP_MPC) {
-		msk->can_ack = true;
-		msk->remote_key = mp_opt->sndr_key;
-		mptcp_crypto_key_sha(msk->remote_key, NULL, &ack_seq);
-		ack_seq++;
-		WRITE_ONCE(msk->ack_seq, ack_seq);
-		atomic64_set(&msk->rcv_wnd_sent, ack_seq);
-	}
-
 	sock_reset_flag(nsk, SOCK_RCU_FREE);
 	/* will be fully established after successful MPC subflow creation */
 	inet_sk_state_store(nsk, TCP_SYN_RECV);
@@ -3355,7 +3333,6 @@ void mptcp_finish_connect(struct sock *ssk)
 	struct mptcp_subflow_context *subflow;
 	struct mptcp_sock *msk;
 	struct sock *sk;
-	u64 ack_seq;
 
 	subflow = mptcp_subflow_ctx(ssk);
 	sk = subflow->conn;
@@ -3363,22 +3340,16 @@ void mptcp_finish_connect(struct sock *ssk)
 
 	pr_debug("msk=%p, token=%u", sk, subflow->token);
 
-	mptcp_crypto_key_sha(subflow->remote_key, NULL, &ack_seq);
-	ack_seq++;
-	subflow->map_seq = ack_seq;
+	subflow->map_seq = subflow->iasn;
 	subflow->map_subflow_seq = 1;
 
 	/* the socket is not connected yet, no msk/subflow ops can access/race
 	 * accessing the field below
 	 */
-	WRITE_ONCE(msk->remote_key, subflow->remote_key);
 	WRITE_ONCE(msk->local_key, subflow->local_key);
 	WRITE_ONCE(msk->write_seq, subflow->idsn + 1);
 	WRITE_ONCE(msk->snd_nxt, msk->write_seq);
-	WRITE_ONCE(msk->ack_seq, ack_seq);
-	WRITE_ONCE(msk->can_ack, 1);
 	WRITE_ONCE(msk->snd_una, msk->write_seq);
-	atomic64_set(&msk->rcv_wnd_sent, ack_seq);
 
 	mptcp_pm_new_connection(msk, ssk, 0);
 

net/mptcp/protocol.h

@@ -126,6 +126,15 @@
 #define MPTCP_CONNECTED		6
 #define MPTCP_RESET_SCHEDULER	7
 
+struct mptcp_skb_cb {
+	u64 map_seq;
+	u64 end_seq;
+	u32 offset;
+	u8  has_rxtstamp:1;
+};
+
+#define MPTCP_SKB_CB(__skb)	((struct mptcp_skb_cb *)&((__skb)->cb[0]))
+
 static inline bool before64(__u64 seq1, __u64 seq2)
 {
 	return (__s64)(seq1 - seq2) < 0;
@@ -467,17 +476,22 @@ struct mptcp_subflow_context {
 		send_fastclose : 1,
 		send_infinite_map : 1,
 		rx_eof : 1,
-		can_ack : 1,        /* only after processing the remote a key */
+		remote_key_valid : 1,        /* received the peer key from */
 		disposable : 1,	    /* ctx can be free at ulp release time */
 		stale : 1,	    /* unable to snd/rcv data, do not use for xmit */
 		local_id_valid : 1, /* local_id is correctly initialized */
-		valid_csum_seen : 1;        /* at least one csum validated */
+		valid_csum_seen : 1,        /* at least one csum validated */
+		is_mptfo : 1,	    /* subflow is doing TFO */
+		__unused : 8;
 	enum mptcp_data_avail data_avail;
 	u32	remote_nonce;
 	u64	thmac;
 	u32	local_nonce;
 	u32	remote_token;
-	u8	hmac[MPTCPOPT_HMAC_LEN];
+	union {
+		u8	hmac[MPTCPOPT_HMAC_LEN]; /* MPJ subflow only */
+		u64	iasn;	    /* initial ack sequence number, MPC subflows only */
+	};
 	u8	local_id;
 	u8	remote_id;
 	u8	reset_seen:1;
@@ -603,7 +617,7 @@ unsigned int mptcp_stale_loss_cnt(const struct net *net);
 int mptcp_get_pm_type(const struct net *net);
 void mptcp_copy_inaddrs(struct sock *msk, const struct sock *ssk);
 void mptcp_subflow_fully_established(struct mptcp_subflow_context *subflow,
-				     struct mptcp_options_received *mp_opt);
+				     const struct mptcp_options_received *mp_opt);
 bool __mptcp_retransmit_pending_data(struct sock *sk);
 void mptcp_check_and_set_pending(struct sock *sk);
 void __mptcp_push_pending(struct sock *sk, unsigned int flags);
@@ -619,6 +633,7 @@ void mptcp_sock_graft(struct sock *sk, struct socket *parent);
 struct socket *__mptcp_nmpc_socket(const struct mptcp_sock *msk);
 bool __mptcp_close(struct sock *sk, long timeout);
 void mptcp_cancel_work(struct sock *sk);
+void mptcp_set_owner_r(struct sk_buff *skb, struct sock *sk);
 
 bool mptcp_addresses_equal(const struct mptcp_addr_info *a,
 			   const struct mptcp_addr_info *b, bool use_port);
@@ -826,6 +841,11 @@ void mptcp_event_addr_announced(const struct sock *ssk, const struct mptcp_addr_
 void mptcp_event_addr_removed(const struct mptcp_sock *msk, u8 id);
 bool mptcp_userspace_pm_active(const struct mptcp_sock *msk);
 
+void mptcp_fastopen_gen_msk_ackseq(struct mptcp_sock *msk, struct mptcp_subflow_context *subflow,
+				   const struct mptcp_options_received *mp_opt);
+void mptcp_fastopen_subflow_synack_set_params(struct mptcp_subflow_context *subflow,
+					      struct request_sock *req);
+
 static inline bool mptcp_pm_should_add_signal(struct mptcp_sock *msk)
 {
 	return READ_ONCE(msk->pm.addr_signal) &

net/mptcp/sockopt.c

@@ -559,7 +559,9 @@ static bool mptcp_supported_sockopt(int level, int optname)
 		case TCP_NOTSENT_LOWAT:
 		case TCP_TX_DELAY:
 		case TCP_INQ:
+		case TCP_FASTOPEN:
 		case TCP_FASTOPEN_CONNECT:
+		case TCP_FASTOPEN_KEY:
 		case TCP_FASTOPEN_NO_COOKIE:
 			return true;
 		}
@@ -569,9 +571,6 @@ static bool mptcp_supported_sockopt(int level, int optname)
 		/* TCP_REPAIR, TCP_REPAIR_QUEUE, TCP_QUEUE_SEQ, TCP_REPAIR_OPTIONS,
 		 * TCP_REPAIR_WINDOW are not supported, better avoid this mess
 		 */
-		/* TCP_FASTOPEN_KEY, TCP_FASTOPEN are not supported because
-		 * fastopen for the listener side is currently unsupported
-		 */
 	}
 	return false;
 }
@@ -801,7 +800,9 @@ static int mptcp_setsockopt_sol_tcp(struct mptcp_sock *msk, int optname,
 		/* See tcp.c: TCP_DEFER_ACCEPT does not fail */
 		mptcp_setsockopt_first_sf_only(msk, SOL_TCP, optname, optval, optlen);
 		return 0;
+	case TCP_FASTOPEN:
 	case TCP_FASTOPEN_CONNECT:
+	case TCP_FASTOPEN_KEY:
 	case TCP_FASTOPEN_NO_COOKIE:
 		return mptcp_setsockopt_first_sf_only(msk, SOL_TCP, optname,
 						      optval, optlen);
@@ -1166,7 +1167,9 @@ static int mptcp_getsockopt_sol_tcp(struct mptcp_sock *msk, int optname,
 	case TCP_INFO:
 	case TCP_CC_INFO:
 	case TCP_DEFER_ACCEPT:
+	case TCP_FASTOPEN:
 	case TCP_FASTOPEN_CONNECT:
+	case TCP_FASTOPEN_KEY:
 	case TCP_FASTOPEN_NO_COOKIE:
 		return mptcp_getsockopt_first_sf_only(msk, SOL_TCP, optname,
 						      optval, optlen);

net/mptcp/subflow.c

@@ -307,7 +307,48 @@ static struct dst_entry *subflow_v4_route_req(const struct sock *sk,
 	return NULL;
 }
 
+static void subflow_prep_synack(const struct sock *sk, struct request_sock *req,
+				struct tcp_fastopen_cookie *foc,
+				enum tcp_synack_type synack_type)
+{
+	struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(sk);
+	struct inet_request_sock *ireq = inet_rsk(req);
+
+	/* clear tstamp_ok, as needed depending on cookie */
+	if (foc && foc->len > -1)
+		ireq->tstamp_ok = 0;
+
+	if (synack_type == TCP_SYNACK_FASTOPEN)
+		mptcp_fastopen_subflow_synack_set_params(subflow, req);
+}
+
+static int subflow_v4_send_synack(const struct sock *sk, struct dst_entry *dst,
+				  struct flowi *fl,
+				  struct request_sock *req,
+				  struct tcp_fastopen_cookie *foc,
+				  enum tcp_synack_type synack_type,
+				  struct sk_buff *syn_skb)
+{
+	subflow_prep_synack(sk, req, foc, synack_type);
+
+	return tcp_request_sock_ipv4_ops.send_synack(sk, dst, fl, req, foc,
+						     synack_type, syn_skb);
+}
+
 #if IS_ENABLED(CONFIG_MPTCP_IPV6)
+static int subflow_v6_send_synack(const struct sock *sk, struct dst_entry *dst,
+				  struct flowi *fl,
+				  struct request_sock *req,
+				  struct tcp_fastopen_cookie *foc,
+				  enum tcp_synack_type synack_type,
+				  struct sk_buff *syn_skb)
+{
+	subflow_prep_synack(sk, req, foc, synack_type);
+
+	return tcp_request_sock_ipv6_ops.send_synack(sk, dst, fl, req, foc,
+						     synack_type, syn_skb);
+}
+
 static struct dst_entry *subflow_v6_route_req(const struct sock *sk,
 					      struct sk_buff *skb,
 					      struct flowi *fl,
@@ -392,11 +433,33 @@ static void mptcp_set_connected(struct sock *sk)
 	mptcp_data_unlock(sk);
 }
 
+static void subflow_set_remote_key(struct mptcp_sock *msk,
+				   struct mptcp_subflow_context *subflow,
+				   const struct mptcp_options_received *mp_opt)
+{
+	/* active MPC subflow will reach here multiple times:
+	 * at subflow_finish_connect() time and at 4th ack time
+	 */
+	if (subflow->remote_key_valid)
+		return;
+
+	subflow->remote_key_valid = 1;
+	subflow->remote_key = mp_opt->sndr_key;
+	mptcp_crypto_key_sha(subflow->remote_key, NULL, &subflow->iasn);
+	subflow->iasn++;
+
+	WRITE_ONCE(msk->remote_key, subflow->remote_key);
+	WRITE_ONCE(msk->ack_seq, subflow->iasn);
+	WRITE_ONCE(msk->can_ack, true);
+	atomic64_set(&msk->rcv_wnd_sent, subflow->iasn);
+}
+
 static void subflow_finish_connect(struct sock *sk, const struct sk_buff *skb)
 {
 	struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(sk);
 	struct mptcp_options_received mp_opt;
 	struct sock *parent = subflow->conn;
+	struct mptcp_sock *msk;
 
 	subflow->icsk_af_ops->sk_rx_dst_set(sk, skb);
 
@@ -404,6 +467,7 @@ static void subflow_finish_connect(struct sock *sk, const struct sk_buff *skb)
 	if (subflow->conn_finished)
 		return;
 
+	msk = mptcp_sk(parent);
 	mptcp_propagate_sndbuf(parent, sk);
 	subflow->rel_write_seq = 1;
 	subflow->conn_finished = 1;
@@ -416,19 +480,16 @@ static void subflow_finish_connect(struct sock *sk, const struct sk_buff *skb)
 			MPTCP_INC_STATS(sock_net(sk),
 					MPTCP_MIB_MPCAPABLEACTIVEFALLBACK);
 			mptcp_do_fallback(sk);
-			pr_fallback(mptcp_sk(subflow->conn));
+			pr_fallback(msk);
 			goto fallback;
 		}
 
 		if (mp_opt.suboptions & OPTION_MPTCP_CSUMREQD)
-			WRITE_ONCE(mptcp_sk(parent)->csum_enabled, true);
+			WRITE_ONCE(msk->csum_enabled, true);
 		if (mp_opt.deny_join_id0)
-			WRITE_ONCE(mptcp_sk(parent)->pm.remote_deny_join_id0, true);
+			WRITE_ONCE(msk->pm.remote_deny_join_id0, true);
 		subflow->mp_capable = 1;
-		subflow->can_ack = 1;
-		subflow->remote_key = mp_opt.sndr_key;
-		pr_debug("subflow=%p, remote_key=%llu", subflow,
-			 subflow->remote_key);
+		subflow_set_remote_key(msk, subflow, &mp_opt);
 		MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_MPCAPABLEACTIVEACK);
 		mptcp_finish_connect(sk);
 		mptcp_set_connected(parent);
@@ -466,7 +527,7 @@ static void subflow_finish_connect(struct sock *sk, const struct sk_buff *skb)
 		subflow->mp_join = 1;
 		MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_JOINSYNACKRX);
 
-		if (subflow_use_different_dport(mptcp_sk(parent), sk)) {
+		if (subflow_use_different_dport(msk, sk)) {
 			pr_debug("synack inet_dport=%d %d",
 				 ntohs(inet_sk(sk)->inet_dport),
 				 ntohs(inet_sk(parent)->inet_dport));
@@ -474,7 +535,7 @@ static void subflow_finish_connect(struct sock *sk, const struct sk_buff *skb)
 		}
 	} else if (mptcp_check_fallback(sk)) {
 fallback:
-		mptcp_rcv_space_init(mptcp_sk(parent), sk);
+		mptcp_rcv_space_init(msk, sk);
 		mptcp_set_connected(parent);
 	}
 	return;
@@ -637,14 +698,16 @@ static void subflow_drop_ctx(struct sock *ssk)
 }
 
 void mptcp_subflow_fully_established(struct mptcp_subflow_context *subflow,
-				     struct mptcp_options_received *mp_opt)
+				     const struct mptcp_options_received *mp_opt)
 {
 	struct mptcp_sock *msk = mptcp_sk(subflow->conn);
 
-	subflow->remote_key = mp_opt->sndr_key;
+	subflow_set_remote_key(msk, subflow, mp_opt);
 	subflow->fully_established = 1;
-	subflow->can_ack = 1;
 	WRITE_ONCE(msk->fully_established, true);
+
+	if (subflow->is_mptfo)
+		mptcp_fastopen_gen_msk_ackseq(msk, subflow, mp_opt);
 }
 
 static struct sock *subflow_syn_recv_sock(const struct sock *sk,
@@ -760,7 +823,7 @@ static struct sock *subflow_syn_recv_sock(const struct sock *sk,
 			/* with OoO packets we can reach here without ingress
 			 * mpc option
 			 */
-			if (mp_opt.suboptions & OPTIONS_MPTCP_MPC)
+			if (mp_opt.suboptions & OPTION_MPTCP_MPC_ACK)
 				mptcp_subflow_fully_established(ctx, &mp_opt);
 		} else if (ctx->mp_join) {
 			struct mptcp_sock *owner;
@@ -1198,16 +1261,8 @@ static bool subflow_check_data_avail(struct sock *ssk)
 		if (WARN_ON_ONCE(!skb))
 			goto no_data;
 
-		/* if msk lacks the remote key, this subflow must provide an
-		 * MP_CAPABLE-based mapping
-		 */
-		if (unlikely(!READ_ONCE(msk->can_ack))) {
-			if (!subflow->mpc_map)
-				goto fallback;
-			WRITE_ONCE(msk->remote_key, subflow->remote_key);
-			WRITE_ONCE(msk->ack_seq, subflow->map_seq);
-			WRITE_ONCE(msk->can_ack, true);
-		}
+		if (unlikely(!READ_ONCE(msk->can_ack)))
+			goto fallback;
 
 		old_ack = READ_ONCE(msk->ack_seq);
 		ack_seq = mptcp_subflow_get_mapped_dsn(subflow);
@@ -1480,6 +1535,7 @@ int __mptcp_subflow_connect(struct sock *sk, const struct mptcp_addr_info *loc,
 
 	mptcp_pm_get_flags_and_ifindex_by_id(msk, local_id,
 					     &flags, &ifindex);
+	subflow->remote_key_valid = 1;
 	subflow->remote_key = msk->remote_key;
 	subflow->local_key = msk->local_key;
 	subflow->token = msk->token;
@@ -1873,6 +1929,7 @@ static void subflow_ulp_clone(const struct request_sock *req,
 		new_ctx->ssn_offset = subflow_req->ssn_offset;
 		new_ctx->mp_join = 1;
 		new_ctx->fully_established = 1;
+		new_ctx->remote_key_valid = 1;
 		new_ctx->backup = subflow_req->backup;
 		new_ctx->remote_id = subflow_req->remote_id;
 		new_ctx->token = subflow_req->token;
@@ -1929,6 +1986,7 @@ void __init mptcp_subflow_init(void)
 
 	subflow_request_sock_ipv4_ops = tcp_request_sock_ipv4_ops;
 	subflow_request_sock_ipv4_ops.route_req = subflow_v4_route_req;
+	subflow_request_sock_ipv4_ops.send_synack = subflow_v4_send_synack;
 
 	subflow_specific = ipv4_specific;
 	subflow_specific.conn_request = subflow_v4_conn_request;
@@ -1942,6 +2000,7 @@ void __init mptcp_subflow_init(void)
 #if IS_ENABLED(CONFIG_MPTCP_IPV6)
 	subflow_request_sock_ipv6_ops = tcp_request_sock_ipv6_ops;
 	subflow_request_sock_ipv6_ops.route_req = subflow_v6_route_req;
+	subflow_request_sock_ipv6_ops.send_synack = subflow_v6_send_synack;
 
 	subflow_v6_specific = ipv6_specific;
 	subflow_v6_specific.conn_request = subflow_v6_conn_request;

tools/testing/selftests/net/mptcp/mptcp_connect.c

@@ -83,6 +83,7 @@ struct cfg_cmsg_types {
 
 struct cfg_sockopt_types {
 	unsigned int transparent:1;
+	unsigned int mptfo:1;
 };
 
 struct tcp_inq_state {
@@ -90,6 +91,13 @@ struct tcp_inq_state {
 	bool expect_eof;
 };
 
+struct wstate {
+	char buf[8192];
+	unsigned int len;
+	unsigned int off;
+	unsigned int total_len;
+};
+
 static struct tcp_inq_state tcp_inq;
 
 static struct cfg_cmsg_types cfg_cmsg_types;
@@ -232,6 +240,14 @@ static void set_transparent(int fd, int pf)
 	}
 }
 
+static void set_mptfo(int fd, int pf)
+{
+	int qlen = 25;
+
+	if (setsockopt(fd, IPPROTO_TCP, TCP_FASTOPEN, &qlen, sizeof(qlen)) == -1)
+		perror("TCP_FASTOPEN");
+}
+
 static int do_ulp_so(int sock, const char *name)
 {
 	return setsockopt(sock, IPPROTO_TCP, TCP_ULP, name, strlen(name));
@@ -300,6 +316,9 @@ static int sock_listen_mptcp(const char * const listenaddr,
 		if (cfg_sockopt_types.transparent)
 			set_transparent(sock, pf);
 
+		if (cfg_sockopt_types.mptfo)
+			set_mptfo(sock, pf);
+
 		if (bind(sock, a->ai_addr, a->ai_addrlen) == 0)
 			break; /* success */
 
@@ -330,13 +349,15 @@ static int sock_listen_mptcp(const char * const listenaddr,
 
 static int sock_connect_mptcp(const char * const remoteaddr,
 			      const char * const port, int proto,
-			      struct addrinfo **peer)
+			      struct addrinfo **peer,
+			      int infd, struct wstate *winfo)
 {
 	struct addrinfo hints = {
 		.ai_protocol = IPPROTO_TCP,
 		.ai_socktype = SOCK_STREAM,
 	};
 	struct addrinfo *a, *addr;
+	int syn_copied = 0;
 	int sock = -1;
 
 	hints.ai_family = pf;
@@ -354,14 +375,34 @@ static int sock_connect_mptcp(const char * const remoteaddr,
 		if (cfg_mark)
 			set_mark(sock, cfg_mark);
 
-		if (connect(sock, a->ai_addr, a->ai_addrlen) == 0) {
-			*peer = a;
-			break; /* success */
+		if (cfg_sockopt_types.mptfo) {
+			if (!winfo->total_len)
+				winfo->total_len = winfo->len = read(infd, winfo->buf,
+								     sizeof(winfo->buf));
+
+			syn_copied = sendto(sock, winfo->buf, winfo->len, MSG_FASTOPEN,
+					    a->ai_addr, a->ai_addrlen);
+			if (syn_copied >= 0) {
+				winfo->off = syn_copied;
+				winfo->len -= syn_copied;
+				*peer = a;
+				break; /* success */
+			}
+		} else {
+			if (connect(sock, a->ai_addr, a->ai_addrlen) == 0) {
+				*peer = a;
+				break; /* success */
+			}
+		}
+		if (cfg_sockopt_types.mptfo) {
+			perror("sendto()");
+			close(sock);
+			sock = -1;
+		} else {
+			perror("connect()");
+			close(sock);
+			sock = -1;
 		}
-
-		perror("connect()");
-		close(sock);
-		sock = -1;
 	}
 
 	freeaddrinfo(addr);
@@ -571,14 +612,14 @@ static void shut_wr(int fd)
 	shutdown(fd, SHUT_WR);
 }
 
-static int copyfd_io_poll(int infd, int peerfd, int outfd, bool *in_closed_after_out)
+static int copyfd_io_poll(int infd, int peerfd, int outfd,
+			  bool *in_closed_after_out, struct wstate *winfo)
 {
 	struct pollfd fds = {
 		.fd = peerfd,
 		.events = POLLIN | POLLOUT,
 	};
-	unsigned int woff = 0, wlen = 0, total_wlen = 0, total_rlen = 0;
-	char wbuf[8192];
+	unsigned int total_wlen = 0, total_rlen = 0;
 
 	set_nonblock(peerfd, true);
 
@@ -638,19 +679,19 @@ static int copyfd_io_poll(int infd, int peerfd, int outfd, bool *in_closed_after
 		}
 
 		if (fds.revents & POLLOUT) {
-			if (wlen == 0) {
-				woff = 0;
-				wlen = read(infd, wbuf, sizeof(wbuf));
+			if (winfo->len == 0) {
+				winfo->off = 0;
+				winfo->len = read(infd, winfo->buf, sizeof(winfo->buf));
 			}
 
-			if (wlen > 0) {
+			if (winfo->len > 0) {
 				ssize_t bw;
 
 				/* limit the total amount of written data to the trunc value */
-				if (cfg_truncate > 0 && wlen + total_wlen > cfg_truncate)
-					wlen = cfg_truncate - total_wlen;
+				if (cfg_truncate > 0 && winfo->len + total_wlen > cfg_truncate)
+					winfo->len = cfg_truncate - total_wlen;
 
-				bw = do_rnd_write(peerfd, wbuf + woff, wlen);
+				bw = do_rnd_write(peerfd, winfo->buf + winfo->off, winfo->len);
 				if (bw < 0) {
 					if (cfg_rcv_trunc)
 						return 0;
@@ -658,10 +699,10 @@ static int copyfd_io_poll(int infd, int peerfd, int outfd, bool *in_closed_after
 					return 111;
 				}
 
-				woff += bw;
-				wlen -= bw;
+				winfo->off += bw;
+				winfo->len -= bw;
 				total_wlen += bw;
-			} else if (wlen == 0) {
+			} else if (winfo->len == 0) {
 				/* We have no more data to send. */
 				fds.events &= ~POLLOUT;
 
@@ -717,10 +758,26 @@ static int do_recvfile(int infd, int outfd)
 	return (int)r;
 }
 
-static int do_mmap(int infd, int outfd, unsigned int size)
+static int spool_buf(int fd, struct wstate *winfo)
+{
+	while (winfo->len) {
+		int ret = write(fd, winfo->buf + winfo->off, winfo->len);
+
+		if (ret < 0) {
+			perror("write");
+			return 4;
+		}
+		winfo->off += ret;
+		winfo->len -= ret;
+	}
+	return 0;
+}
+
+static int do_mmap(int infd, int outfd, unsigned int size,
+		   struct wstate *winfo)
 {
 	char *inbuf = mmap(NULL, size, PROT_READ, MAP_SHARED, infd, 0);
-	ssize_t ret = 0, off = 0;
+	ssize_t ret = 0, off = winfo->total_len;
 	size_t rem;
 
 	if (inbuf == MAP_FAILED) {
@@ -728,7 +785,11 @@ static int do_mmap(int infd, int outfd, unsigned int size)
 		return 1;
 	}
 
-	rem = size;
+	ret = spool_buf(outfd, winfo);
+	if (ret < 0)
+		return ret;
+
+	rem = size - winfo->total_len;
 
 	while (rem > 0) {
 		ret = write(outfd, inbuf + off, rem);
@@ -772,8 +833,16 @@ static int get_infd_size(int fd)
 	return (int)count;
 }
 
-static int do_sendfile(int infd, int outfd, unsigned int count)
+static int do_sendfile(int infd, int outfd, unsigned int count,
+		       struct wstate *winfo)
 {
+	int ret = spool_buf(outfd, winfo);
+
+	if (ret < 0)
+		return ret;
+
+	count -= winfo->total_len;
+
 	while (count > 0) {
 		ssize_t r;
 
@@ -790,7 +859,8 @@ static int do_sendfile(int infd, int outfd, unsigned int count)
 }
 
 static int copyfd_io_mmap(int infd, int peerfd, int outfd,
-			  unsigned int size, bool *in_closed_after_out)
+			  unsigned int size, bool *in_closed_after_out,
+			  struct wstate *winfo)
 {
 	int err;
 
@@ -799,9 +869,9 @@ static int copyfd_io_mmap(int infd, int peerfd, int outfd,
 		if (err)
 			return err;
 
-		err = do_mmap(infd, peerfd, size);
+		err = do_mmap(infd, peerfd, size, winfo);
 	} else {
-		err = do_mmap(infd, peerfd, size);
+		err = do_mmap(infd, peerfd, size, winfo);
 		if (err)
 			return err;
 
@@ -815,7 +885,7 @@ static int copyfd_io_mmap(int infd, int peerfd, int outfd,
 }
 
 static int copyfd_io_sendfile(int infd, int peerfd, int outfd,
-			      unsigned int size, bool *in_closed_after_out)
+			      unsigned int size, bool *in_closed_after_out, struct wstate *winfo)
 {
 	int err;
 
@@ -824,9 +894,9 @@ static int copyfd_io_sendfile(int infd, int peerfd, int outfd,
 		if (err)
 			return err;
 
-		err = do_sendfile(infd, peerfd, size);
+		err = do_sendfile(infd, peerfd, size, winfo);
 	} else {
-		err = do_sendfile(infd, peerfd, size);
+		err = do_sendfile(infd, peerfd, size, winfo);
 		if (err)
 			return err;
 
@@ -839,7 +909,7 @@ static int copyfd_io_sendfile(int infd, int peerfd, int outfd,
 	return err;
 }
 
-static int copyfd_io(int infd, int peerfd, int outfd, bool close_peerfd)
+static int copyfd_io(int infd, int peerfd, int outfd, bool close_peerfd, struct wstate *winfo)
 {
 	bool in_closed_after_out = false;
 	struct timespec start, end;
@@ -851,21 +921,24 @@ static int copyfd_io(int infd, int peerfd, int outfd, bool close_peerfd)
 
 	switch (cfg_mode) {
 	case CFG_MODE_POLL:
-		ret = copyfd_io_poll(infd, peerfd, outfd, &in_closed_after_out);
+		ret = copyfd_io_poll(infd, peerfd, outfd, &in_closed_after_out,
+				     winfo);
 		break;
 
 	case CFG_MODE_MMAP:
 		file_size = get_infd_size(infd);
 		if (file_size < 0)
 			return file_size;
-		ret = copyfd_io_mmap(infd, peerfd, outfd, file_size, &in_closed_after_out);
+		ret = copyfd_io_mmap(infd, peerfd, outfd, file_size,
+				     &in_closed_after_out, winfo);
 		break;
 
 	case CFG_MODE_SENDFILE:
 		file_size = get_infd_size(infd);
 		if (file_size < 0)
 			return file_size;
-		ret = copyfd_io_sendfile(infd, peerfd, outfd, file_size, &in_closed_after_out);
+		ret = copyfd_io_sendfile(infd, peerfd, outfd, file_size,
+					 &in_closed_after_out, winfo);
 		break;
 
 	default:
@@ -999,6 +1072,7 @@ static void maybe_close(int fd)
 int main_loop_s(int listensock)
 {
 	struct sockaddr_storage ss;
+	struct wstate winfo;
 	struct pollfd polls;
 	socklen_t salen;
 	int remotesock;
@@ -1033,7 +1107,8 @@ int main_loop_s(int listensock)
 
 		SOCK_TEST_TCPULP(remotesock, 0);
 
-		copyfd_io(fd, remotesock, 1, true);
+		memset(&winfo, 0, sizeof(winfo));
+		copyfd_io(fd, remotesock, 1, true, &winfo);
 	} else {
 		perror("accept");
 		return 1;
@@ -1130,6 +1205,11 @@ static void parse_setsock_options(const char *name)
 		return;
 	}
 
+	if (strncmp(name, "MPTFO", len) == 0) {
+		cfg_sockopt_types.mptfo = 1;
+		return;
+	}
+
 	fprintf(stderr, "Unrecognized setsockopt option %s\n", name);
 	exit(1);
 }
@@ -1166,11 +1246,18 @@ void xdisconnect(int fd, int addrlen)
 
 int main_loop(void)
 {
-	int fd, ret, fd_in = 0;
+	int fd = 0, ret, fd_in = 0;
 	struct addrinfo *peer;
+	struct wstate winfo;
+
+	if (cfg_input && cfg_sockopt_types.mptfo) {
+		fd_in = open(cfg_input, O_RDONLY);
+		if (fd < 0)
+			xerror("can't open %s:%d", cfg_input, errno);
+	}
 
-	/* listener is ready. */
-	fd = sock_connect_mptcp(cfg_host, cfg_port, cfg_sock_proto, &peer);
+	memset(&winfo, 0, sizeof(winfo));
+	fd = sock_connect_mptcp(cfg_host, cfg_port, cfg_sock_proto, &peer, fd_in, &winfo);
 	if (fd < 0)
 		return 2;
 
@@ -1186,14 +1273,13 @@ int main_loop(void)
 	if (cfg_cmsg_types.cmsg_enabled)
 		apply_cmsg_types(fd, &cfg_cmsg_types);
 
-	if (cfg_input) {
+	if (cfg_input && !cfg_sockopt_types.mptfo) {
 		fd_in = open(cfg_input, O_RDONLY);
 		if (fd < 0)
 			xerror("can't open %s:%d", cfg_input, errno);
 	}
 
-	/* close the client socket open only if we are not going to reconnect */
-	ret = copyfd_io(fd_in, fd, 1, 0);
+	ret = copyfd_io(fd_in, fd, 1, 0, &winfo);
 	if (ret)
 		return ret;
 
@@ -1210,6 +1296,7 @@ int main_loop(void)
 			xerror("can't reconnect: %d", errno);
 		if (cfg_input)
 			close(fd_in);
+		memset(&winfo, 0, sizeof(winfo));
 		goto again;
 	} else {
 		close(fd);

tools/testing/selftests/net/mptcp/mptcp_connect.sh

@@ -762,6 +762,23 @@ run_tests_peekmode()
 	run_tests_lo "$ns1" "$ns1" dead:beef:1::1 1 "-P ${peekmode}"
 }
 
+run_tests_mptfo()
+{
+	echo "INFO: with MPTFO start"
+	ip netns exec "$ns1" sysctl -q net.ipv4.tcp_fastopen=2
+	ip netns exec "$ns2" sysctl -q net.ipv4.tcp_fastopen=1
+
+	run_tests_lo "$ns1" "$ns2" 10.0.1.1 0 "-o MPTFO"
+	run_tests_lo "$ns1" "$ns2" 10.0.1.1 0 "-o MPTFO"
+
+	run_tests_lo "$ns1" "$ns2" dead:beef:1::1 0 "-o MPTFO"
+	run_tests_lo "$ns1" "$ns2" dead:beef:1::1 0 "-o MPTFO"
+
+	ip netns exec "$ns1" sysctl -q net.ipv4.tcp_fastopen=0
+	ip netns exec "$ns2" sysctl -q net.ipv4.tcp_fastopen=0
+	echo "INFO: with MPTFO end"
+}
+
 run_tests_disconnect()
 {
 	local peekmode="$1"
@@ -901,6 +918,10 @@ run_tests_peekmode "saveWithPeek"
 run_tests_peekmode "saveAfterPeek"
 stop_if_error "Tests with peek mode have failed"
 
+# MPTFO (MultiPath TCP Fatopen tests)
+run_tests_mptfo
+stop_if_error "Tests with MPTFO have failed"
+
 # connect to ns4 ip address, ns2 should intercept/proxy
 run_test_transparent 10.0.3.1 "tproxy ipv4"
 run_test_transparent dead:beef:3::1 "tproxy ipv6"