Merge tag 'v5.7-rc6' into rdma.git for-next

Linux 5.7-rc6

Conflict in drivers/net/ethernet/mellanox/mlx5/core/steering/dr_send.c
resolved by deleting dr_cq_event, matching how netdev resolved it.

Required for dependencies in the following patches.

Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>

5 files changed, 256 insertions, 139 deletions

diff --git a/net/mptcp/options.c b/net/mptcp/options.c
index faf57585b892..45497af23906 100644
--- a/net/mptcp/options.c
+++ b/net/mptcp/options.c
@@ -16,10 +16,10 @@ static bool mptcp_cap_flag_sha256(u8 flags)
 	return (flags & MPTCP_CAP_FLAG_MASK) == MPTCP_CAP_HMAC_SHA256;
 }
 
-void mptcp_parse_option(const struct sk_buff *skb, const unsigned char *ptr,
-			int opsize, struct tcp_options_received *opt_rx)
+static void mptcp_parse_option(const struct sk_buff *skb,
+			       const unsigned char *ptr, int opsize,
+			       struct mptcp_options_received *mp_opt)
 {
-	struct mptcp_options_received *mp_opt = &opt_rx->mptcp;
 	u8 subtype = *ptr >> 4;
 	int expected_opsize;
 	u8 version;
@@ -283,12 +283,20 @@ void mptcp_parse_option(const struct sk_buff *skb, const unsigned char *ptr,
 }
 
 void mptcp_get_options(const struct sk_buff *skb,
-		       struct tcp_options_received *opt_rx)
+		       struct mptcp_options_received *mp_opt)
 {
-	const unsigned char *ptr;
 	const struct tcphdr *th = tcp_hdr(skb);
-	int length = (th->doff * 4) - sizeof(struct tcphdr);
+	const unsigned char *ptr;
+	int length;
+
+	/* initialize option status */
+	mp_opt->mp_capable = 0;
+	mp_opt->mp_join = 0;
+	mp_opt->add_addr = 0;
+	mp_opt->rm_addr = 0;
+	mp_opt->dss = 0;
 
+	length = (th->doff * 4) - sizeof(struct tcphdr);
 	ptr = (const unsigned char *)(th + 1);
 
 	while (length > 0) {
@@ -308,7 +316,7 @@ void mptcp_get_options(const struct sk_buff *skb,
 			if (opsize > length)
 				return;	/* don't parse partial options */
 			if (opcode == TCPOPT_MPTCP)
-				mptcp_parse_option(skb, ptr, opsize, opt_rx);
+				mptcp_parse_option(skb, ptr, opsize, mp_opt);
 			ptr += opsize - 2;
 			length -= opsize;
 		}
@@ -344,28 +352,6 @@ bool mptcp_syn_options(struct sock *sk, const struct sk_buff *skb,
 	return false;
 }
 
-void mptcp_rcv_synsent(struct sock *sk)
-{
-	struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(sk);
-	struct tcp_sock *tp = tcp_sk(sk);
-
-	if (subflow->request_mptcp && tp->rx_opt.mptcp.mp_capable) {
-		subflow->mp_capable = 1;
-		subflow->can_ack = 1;
-		subflow->remote_key = tp->rx_opt.mptcp.sndr_key;
-		pr_debug("subflow=%p, remote_key=%llu", subflow,
-			 subflow->remote_key);
-	} else if (subflow->request_join && tp->rx_opt.mptcp.mp_join) {
-		subflow->mp_join = 1;
-		subflow->thmac = tp->rx_opt.mptcp.thmac;
-		subflow->remote_nonce = tp->rx_opt.mptcp.nonce;
-		pr_debug("subflow=%p, thmac=%llu, remote_nonce=%u", subflow,
-			 subflow->thmac, subflow->remote_nonce);
-	} else if (subflow->request_mptcp) {
-		tcp_sk(sk)->is_mptcp = 0;
-	}
-}
-
 /* MP_JOIN client subflow must wait for 4th ack before sending any data:
  * TCP can't schedule delack timer before the subflow is fully established.
  * MPTCP uses the delack timer to do 3rd ack retransmissions
@@ -709,7 +695,7 @@ static bool check_fully_established(struct mptcp_sock *msk, struct sock *sk,
 	if (TCP_SKB_CB(skb)->seq != subflow->ssn_offset + 1)
 		return subflow->mp_capable;
 
-	if (mp_opt->use_ack) {
+	if (mp_opt->dss && mp_opt->use_ack) {
 		/* subflows are fully established as soon as we get any
 		 * additional ack.
 		 */
@@ -717,8 +703,6 @@ static bool check_fully_established(struct mptcp_sock *msk, struct sock *sk,
 		goto fully_established;
 	}
 
-	WARN_ON_ONCE(subflow->can_ack);
-
 	/* If the first established packet does not contain MP_CAPABLE + data
 	 * then fallback to TCP
 	 */
@@ -728,6 +712,8 @@ static bool check_fully_established(struct mptcp_sock *msk, struct sock *sk,
 		return false;
 	}
 
+	if (unlikely(!READ_ONCE(msk->pm.server_side)))
+		pr_warn_once("bogus mpc option on established client sk");
 	subflow->fully_established = 1;
 	subflow->remote_key = mp_opt->sndr_key;
 	subflow->can_ack = 1;
@@ -819,41 +805,41 @@ void mptcp_incoming_options(struct sock *sk, struct sk_buff *skb,
 {
 	struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(sk);
 	struct mptcp_sock *msk = mptcp_sk(subflow->conn);
-	struct mptcp_options_received *mp_opt;
+	struct mptcp_options_received mp_opt;
 	struct mptcp_ext *mpext;
 
-	mp_opt = &opt_rx->mptcp;
-	if (!check_fully_established(msk, sk, subflow, skb, mp_opt))
+	mptcp_get_options(skb, &mp_opt);
+	if (!check_fully_established(msk, sk, subflow, skb, &mp_opt))
 		return;
 
-	if (mp_opt->add_addr && add_addr_hmac_valid(msk, mp_opt)) {
+	if (mp_opt.add_addr && add_addr_hmac_valid(msk, &mp_opt)) {
 		struct mptcp_addr_info addr;
 
-		addr.port = htons(mp_opt->port);
-		addr.id = mp_opt->addr_id;
-		if (mp_opt->family == MPTCP_ADDR_IPVERSION_4) {
+		addr.port = htons(mp_opt.port);
+		addr.id = mp_opt.addr_id;
+		if (mp_opt.family == MPTCP_ADDR_IPVERSION_4) {
 			addr.family = AF_INET;
-			addr.addr = mp_opt->addr;
+			addr.addr = mp_opt.addr;
 		}
 #if IS_ENABLED(CONFIG_MPTCP_IPV6)
-		else if (mp_opt->family == MPTCP_ADDR_IPVERSION_6) {
+		else if (mp_opt.family == MPTCP_ADDR_IPVERSION_6) {
 			addr.family = AF_INET6;
-			addr.addr6 = mp_opt->addr6;
+			addr.addr6 = mp_opt.addr6;
 		}
 #endif
-		if (!mp_opt->echo)
+		if (!mp_opt.echo)
 			mptcp_pm_add_addr_received(msk, &addr);
-		mp_opt->add_addr = 0;
+		mp_opt.add_addr = 0;
 	}
 
-	if (!mp_opt->dss)
+	if (!mp_opt.dss)
 		return;
 
 	/* we can't wait for recvmsg() to update the ack_seq, otherwise
 	 * monodirectional flows will stuck
 	 */
-	if (mp_opt->use_ack)
-		update_una(msk, mp_opt);
+	if (mp_opt.use_ack)
+		update_una(msk, &mp_opt);
 
 	mpext = skb_ext_add(skb, SKB_EXT_MPTCP);
 	if (!mpext)
@@ -861,8 +847,8 @@ void mptcp_incoming_options(struct sock *sk, struct sk_buff *skb,
 
 	memset(mpext, 0, sizeof(*mpext));
 
-	if (mp_opt->use_map) {
-		if (mp_opt->mpc_map) {
+	if (mp_opt.use_map) {
+		if (mp_opt.mpc_map) {
 			/* this is an MP_CAPABLE carrying MPTCP data
 			 * we know this map the first chunk of data
 			 */
@@ -872,16 +858,16 @@ void mptcp_incoming_options(struct sock *sk, struct sk_buff *skb,
 			mpext->subflow_seq = 1;
 			mpext->dsn64 = 1;
 			mpext->mpc_map = 1;
+			mpext->data_fin = 0;
 		} else {
-			mpext->data_seq = mp_opt->data_seq;
-			mpext->subflow_seq = mp_opt->subflow_seq;
-			mpext->dsn64 = mp_opt->dsn64;
+			mpext->data_seq = mp_opt.data_seq;
+			mpext->subflow_seq = mp_opt.subflow_seq;
+			mpext->dsn64 = mp_opt.dsn64;
+			mpext->data_fin = mp_opt.data_fin;
 		}
-		mpext->data_len = mp_opt->data_len;
+		mpext->data_len = mp_opt.data_len;
 		mpext->use_map = 1;
 	}
-
-	mpext->data_fin = mp_opt->data_fin;
 }
 
 void mptcp_write_options(__be32 *ptr, struct mptcp_out_options *opts)
diff --git a/net/mptcp/pm_netlink.c b/net/mptcp/pm_netlink.c
index 86d61ab34c7c..b78edf237ba0 100644
--- a/net/mptcp/pm_netlink.c
+++ b/net/mptcp/pm_netlink.c
@@ -599,12 +599,14 @@ static int mptcp_nl_fill_addr(struct sk_buff *skb,
 	    nla_put_s32(skb, MPTCP_PM_ADDR_ATTR_IF_IDX, entry->ifindex))
 		goto nla_put_failure;
 
-	if (addr->family == AF_INET)
-		nla_put_in_addr(skb, MPTCP_PM_ADDR_ATTR_ADDR4,
-				addr->addr.s_addr);
+	if (addr->family == AF_INET &&
+	    nla_put_in_addr(skb, MPTCP_PM_ADDR_ATTR_ADDR4,
+			    addr->addr.s_addr))
+		goto nla_put_failure;
 #if IS_ENABLED(CONFIG_MPTCP_IPV6)
-	else if (addr->family == AF_INET6)
-		nla_put_in6_addr(skb, MPTCP_PM_ADDR_ATTR_ADDR6, &addr->addr6);
+	else if (addr->family == AF_INET6 &&
+		 nla_put_in6_addr(skb, MPTCP_PM_ADDR_ATTR_ADDR6, &addr->addr6))
+		goto nla_put_failure;
 #endif
 	nla_nest_end(skb, attr);
 	return 0;
diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c
index 939a5045181a..32ea8d35489a 100644
--- a/net/mptcp/protocol.c
+++ b/net/mptcp/protocol.c
@@ -97,12 +97,7 @@ static struct socket *__mptcp_tcp_fallback(struct mptcp_sock *msk)
 	if (likely(!__mptcp_needs_tcp_fallback(msk)))
 		return NULL;
 
-	if (msk->subflow) {
-		release_sock((struct sock *)msk);
-		return msk->subflow;
-	}
-
-	return NULL;
+	return msk->subflow;
 }
 
 static bool __mptcp_can_create_subflow(const struct mptcp_sock *msk)
@@ -734,9 +729,10 @@ static int mptcp_sendmsg(struct sock *sk, struct msghdr *msg, size_t len)
 			goto out;
 	}
 
+fallback:
 	ssock = __mptcp_tcp_fallback(msk);
 	if (unlikely(ssock)) {
-fallback:
+		release_sock(sk);
 		pr_debug("fallback passthrough");
 		ret = sock_sendmsg(ssock, msg);
 		return ret >= 0 ? ret + copied : (copied ? copied : ret);
@@ -769,8 +765,14 @@ fallback:
 		if (ret < 0)
 			break;
 		if (ret == 0 && unlikely(__mptcp_needs_tcp_fallback(msk))) {
+			/* Can happen for passive sockets:
+			 * 3WHS negotiated MPTCP, but first packet after is
+			 * plain TCP (e.g. due to middlebox filtering unknown
+			 * options).
+			 *
+			 * Fall back to TCP.
+			 */
 			release_sock(ssk);
-			ssock = __mptcp_tcp_fallback(msk);
 			goto fallback;
 		}
 
@@ -883,6 +885,7 @@ static int mptcp_recvmsg(struct sock *sk, struct msghdr *msg, size_t len,
 	ssock = __mptcp_tcp_fallback(msk);
 	if (unlikely(ssock)) {
 fallback:
+		release_sock(sk);
 		pr_debug("fallback-read subflow=%p",
 			 mptcp_subflow_ctx(ssock->sk));
 		copied = sock_recvmsg(ssock, msg, flags);
@@ -1313,11 +1316,12 @@ static void mptcp_copy_inaddrs(struct sock *msk, const struct sock *ssk)
 
 static int mptcp_disconnect(struct sock *sk, int flags)
 {
-	lock_sock(sk);
-	__mptcp_clear_xmit(sk);
-	release_sock(sk);
-	mptcp_cancel_work(sk);
-	return tcp_disconnect(sk, flags);
+	/* Should never be called.
+	 * inet_stream_connect() calls ->disconnect, but that
+	 * refers to the subflow socket, not the mptcp one.
+	 */
+	WARN_ON_ONCE(1);
+	return 0;
 }
 
 #if IS_ENABLED(CONFIG_MPTCP_IPV6)
@@ -1329,7 +1333,9 @@ static struct ipv6_pinfo *mptcp_inet6_sk(const struct sock *sk)
 }
 #endif
 
-struct sock *mptcp_sk_clone(const struct sock *sk, struct request_sock *req)
+struct sock *mptcp_sk_clone(const struct sock *sk,
+			    const struct mptcp_options_received *mp_opt,
+			    struct request_sock *req)
 {
 	struct mptcp_subflow_request_sock *subflow_req = mptcp_subflow_rsk(req);
 	struct sock *nsk = sk_clone_lock(sk, GFP_ATOMIC);
@@ -1352,26 +1358,30 @@ struct sock *mptcp_sk_clone(const struct sock *sk, struct request_sock *req)
 	msk->subflow = NULL;
 
 	if (unlikely(mptcp_token_new_accept(subflow_req->token, nsk))) {
+		nsk->sk_state = TCP_CLOSE;
 		bh_unlock_sock(nsk);
 
 		/* we can't call into mptcp_close() here - possible BH context
-		 * free the sock directly
+		 * free the sock directly.
+		 * sk_clone_lock() sets nsk refcnt to two, hence call sk_free()
+		 * too.
 		 */
-		nsk->sk_prot->destroy(nsk);
+		sk_common_release(nsk);
 		sk_free(nsk);
 		return NULL;
 	}
 
 	msk->write_seq = subflow_req->idsn + 1;
 	atomic64_set(&msk->snd_una, msk->write_seq);
-	if (subflow_req->remote_key_valid) {
+	if (mp_opt->mp_capable) {
 		msk->can_ack = true;
-		msk->remote_key = subflow_req->remote_key;
+		msk->remote_key = mp_opt->sndr_key;
 		mptcp_crypto_key_sha(msk->remote_key, NULL, &ack_seq);
 		ack_seq++;
 		msk->ack_seq = 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);
 	bh_unlock_sock(nsk);
@@ -1428,6 +1438,7 @@ static struct sock *mptcp_accept(struct sock *sk, int flags, int *err,
 		newsk = new_mptcp_sock;
 		mptcp_copy_inaddrs(newsk, ssk);
 		list_add(&subflow->node, &msk->conn_list);
+		inet_sk_state_store(newsk, TCP_ESTABLISHED);
 
 		bh_unlock_sock(new_mptcp_sock);
 
@@ -1467,12 +1478,11 @@ static int mptcp_setsockopt(struct sock *sk, int level, int optname,
 	 */
 	lock_sock(sk);
 	ssock = __mptcp_tcp_fallback(msk);
+	release_sock(sk);
 	if (ssock)
 		return tcp_setsockopt(ssock->sk, level, optname, optval,
 				      optlen);
 
-	release_sock(sk);
-
 	return -EOPNOTSUPP;
 }
 
@@ -1492,12 +1502,11 @@ static int mptcp_getsockopt(struct sock *sk, int level, int optname,
 	 */
 	lock_sock(sk);
 	ssock = __mptcp_tcp_fallback(msk);
+	release_sock(sk);
 	if (ssock)
 		return tcp_getsockopt(ssock->sk, level, optname, optval,
 				      option);
 
-	release_sock(sk);
-
 	return -EOPNOTSUPP;
 }
 
@@ -1620,6 +1629,8 @@ bool mptcp_finish_join(struct sock *sk)
 
 	ret = mptcp_pm_allow_new_subflow(msk);
 	if (ret) {
+		subflow->map_seq = msk->ack_seq;
+
 		/* active connections are already on conn_list */
 		spin_lock_bh(&msk->join_list_lock);
 		if (!WARN_ON_ONCE(!list_empty(&subflow->node)))
@@ -1774,6 +1785,8 @@ static int mptcp_listen(struct socket *sock, int backlog)
 		goto unlock;
 	}
 
+	sock_set_flag(sock->sk, SOCK_RCU_FREE);
+
 	err = ssock->ops->listen(ssock, backlog);
 	inet_sk_state_store(sock->sk, inet_sk_state_load(ssock->sk));
 	if (!err)
diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h
index 67448002a2d7..e4ca6320ce76 100644
--- a/net/mptcp/protocol.h
+++ b/net/mptcp/protocol.h
@@ -91,6 +91,45 @@
 #define MPTCP_WORK_RTX		2
 #define MPTCP_WORK_EOF		3
 
+struct mptcp_options_received {
+	u64	sndr_key;
+	u64	rcvr_key;
+	u64	data_ack;
+	u64	data_seq;
+	u32	subflow_seq;
+	u16	data_len;
+	u16	mp_capable : 1,
+		mp_join : 1,
+		dss : 1,
+		add_addr : 1,
+		rm_addr : 1,
+		family : 4,
+		echo : 1,
+		backup : 1;
+	u32	token;
+	u32	nonce;
+	u64	thmac;
+	u8	hmac[20];
+	u8	join_id;
+	u8	use_map:1,
+		dsn64:1,
+		data_fin:1,
+		use_ack:1,
+		ack64:1,
+		mpc_map:1,
+		__unused:2;
+	u8	addr_id;
+	u8	rm_id;
+	union {
+		struct in_addr	addr;
+#if IS_ENABLED(CONFIG_MPTCP_IPV6)
+		struct in6_addr	addr6;
+#endif
+	};
+	u64	ahmac;
+	u16	port;
+};
+
 static inline __be32 mptcp_option(u8 subopt, u8 len, u8 nib, u8 field)
 {
 	return htonl((TCPOPT_MPTCP << 24) | (len << 16) | (subopt << 12) |
@@ -206,12 +245,10 @@ struct mptcp_subflow_request_sock {
 	struct	tcp_request_sock sk;
 	u16	mp_capable : 1,
 		mp_join : 1,
-		backup : 1,
-		remote_key_valid : 1;
+		backup : 1;
 	u8	local_id;
 	u8	remote_id;
 	u64	local_key;
-	u64	remote_key;
 	u64	idsn;
 	u32	token;
 	u32	ssn_offset;
@@ -332,9 +369,11 @@ void mptcp_proto_init(void);
 int mptcp_proto_v6_init(void);
 #endif
 
-struct sock *mptcp_sk_clone(const struct sock *sk, struct request_sock *req);
+struct sock *mptcp_sk_clone(const struct sock *sk,
+			    const struct mptcp_options_received *mp_opt,
+			    struct request_sock *req);
 void mptcp_get_options(const struct sk_buff *skb,
-		       struct tcp_options_received *opt_rx);
+		       struct mptcp_options_received *mp_opt);
 
 void mptcp_finish_connect(struct sock *sk);
 void mptcp_data_ready(struct sock *sk, struct sock *ssk);
diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c
index 50a8bea987c6..4931a29a6f08 100644
--- a/net/mptcp/subflow.c
+++ b/net/mptcp/subflow.c
@@ -124,16 +124,14 @@ static void subflow_init_req(struct request_sock *req,
 {
 	struct mptcp_subflow_context *listener = mptcp_subflow_ctx(sk_listener);
 	struct mptcp_subflow_request_sock *subflow_req = mptcp_subflow_rsk(req);
-	struct tcp_options_received rx_opt;
+	struct mptcp_options_received mp_opt;
 
 	pr_debug("subflow_req=%p, listener=%p", subflow_req, listener);
 
-	memset(&rx_opt.mptcp, 0, sizeof(rx_opt.mptcp));
-	mptcp_get_options(skb, &rx_opt);
+	mptcp_get_options(skb, &mp_opt);
 
 	subflow_req->mp_capable = 0;
 	subflow_req->mp_join = 0;
-	subflow_req->remote_key_valid = 0;
 
 #ifdef CONFIG_TCP_MD5SIG
 	/* no MPTCP if MD5SIG is enabled on this socket or we may run out of
@@ -143,16 +141,16 @@ static void subflow_init_req(struct request_sock *req,
 		return;
 #endif
 
-	if (rx_opt.mptcp.mp_capable) {
+	if (mp_opt.mp_capable) {
 		SUBFLOW_REQ_INC_STATS(req, MPTCP_MIB_MPCAPABLEPASSIVE);
 
-		if (rx_opt.mptcp.mp_join)
+		if (mp_opt.mp_join)
 			return;
-	} else if (rx_opt.mptcp.mp_join) {
+	} else if (mp_opt.mp_join) {
 		SUBFLOW_REQ_INC_STATS(req, MPTCP_MIB_JOINSYNRX);
 	}
 
-	if (rx_opt.mptcp.mp_capable && listener->request_mptcp) {
+	if (mp_opt.mp_capable && listener->request_mptcp) {
 		int err;
 
 		err = mptcp_token_new_request(req);
@@ -160,13 +158,13 @@ static void subflow_init_req(struct request_sock *req,
 			subflow_req->mp_capable = 1;
 
 		subflow_req->ssn_offset = TCP_SKB_CB(skb)->seq;
-	} else if (rx_opt.mptcp.mp_join && listener->request_mptcp) {
+	} else if (mp_opt.mp_join && listener->request_mptcp) {
 		subflow_req->ssn_offset = TCP_SKB_CB(skb)->seq;
 		subflow_req->mp_join = 1;
-		subflow_req->backup = rx_opt.mptcp.backup;
-		subflow_req->remote_id = rx_opt.mptcp.join_id;
-		subflow_req->token = rx_opt.mptcp.token;
-		subflow_req->remote_nonce = rx_opt.mptcp.nonce;
+		subflow_req->backup = mp_opt.backup;
+		subflow_req->remote_id = mp_opt.join_id;
+		subflow_req->token = mp_opt.token;
+		subflow_req->remote_nonce = mp_opt.nonce;
 		pr_debug("token=%u, remote_nonce=%u", subflow_req->token,
 			 subflow_req->remote_nonce);
 		if (!subflow_token_join_request(req, skb)) {
@@ -222,23 +220,47 @@ static bool subflow_thmac_valid(struct mptcp_subflow_context *subflow)
 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 tcp_sock *tp = tcp_sk(sk);
 
 	subflow->icsk_af_ops->sk_rx_dst_set(sk, skb);
 
-	if (inet_sk_state_load(parent) != TCP_ESTABLISHED) {
+	if (inet_sk_state_load(parent) == TCP_SYN_SENT) {
 		inet_sk_state_store(parent, TCP_ESTABLISHED);
 		parent->sk_state_change(parent);
 	}
 
-	if (subflow->conn_finished || !tcp_sk(sk)->is_mptcp)
+	/* be sure no special action on any packet other than syn-ack */
+	if (subflow->conn_finished)
+		return;
+
+	subflow->conn_finished = 1;
+
+	mptcp_get_options(skb, &mp_opt);
+	if (subflow->request_mptcp && mp_opt.mp_capable) {
+		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);
+	} else if (subflow->request_join && mp_opt.mp_join) {
+		subflow->mp_join = 1;
+		subflow->thmac = mp_opt.thmac;
+		subflow->remote_nonce = mp_opt.nonce;
+		pr_debug("subflow=%p, thmac=%llu, remote_nonce=%u", subflow,
+			 subflow->thmac, subflow->remote_nonce);
+	} else if (subflow->request_mptcp) {
+		tp->is_mptcp = 0;
+	}
+
+	if (!tp->is_mptcp)
 		return;
 
 	if (subflow->mp_capable) {
 		pr_debug("subflow=%p, remote_key=%llu", mptcp_subflow_ctx(sk),
 			 subflow->remote_key);
 		mptcp_finish_connect(sk);
-		subflow->conn_finished = 1;
 
 		if (skb) {
 			pr_debug("synack seq=%u", TCP_SKB_CB(skb)->seq);
@@ -265,7 +287,6 @@ static void subflow_finish_connect(struct sock *sk, const struct sk_buff *skb)
 		if (!mptcp_finish_join(sk))
 			goto do_reset;
 
-		subflow->conn_finished = 1;
 		MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_JOINSYNACKRX);
 	} else {
 do_reset:
@@ -323,7 +344,7 @@ drop:
 
 /* validate hmac received in third ACK */
 static bool subflow_hmac_valid(const struct request_sock *req,
-			       const struct tcp_options_received *rx_opt)
+			       const struct mptcp_options_received *mp_opt)
 {
 	const struct mptcp_subflow_request_sock *subflow_req;
 	u8 hmac[MPTCPOPT_HMAC_LEN];
@@ -340,13 +361,53 @@ static bool subflow_hmac_valid(const struct request_sock *req,
 			      subflow_req->local_nonce, hmac);
 
 	ret = true;
-	if (crypto_memneq(hmac, rx_opt->mptcp.hmac, sizeof(hmac)))
+	if (crypto_memneq(hmac, mp_opt->hmac, sizeof(hmac)))
 		ret = false;
 
 	sock_put((struct sock *)msk);
 	return ret;
 }
 
+static void mptcp_sock_destruct(struct sock *sk)
+{
+	/* if new mptcp socket isn't accepted, it is free'd
+	 * from the tcp listener sockets request queue, linked
+	 * from req->sk.  The tcp socket is released.
+	 * This calls the ULP release function which will
+	 * also remove the mptcp socket, via
+	 * sock_put(ctx->conn).
+	 *
+	 * Problem is that the mptcp socket will not be in
+	 * SYN_RECV state and doesn't have SOCK_DEAD flag.
+	 * Both result in warnings from inet_sock_destruct.
+	 */
+
+	if (sk->sk_state == TCP_SYN_RECV) {
+		sk->sk_state = TCP_CLOSE;
+		WARN_ON_ONCE(sk->sk_socket);
+		sock_orphan(sk);
+	}
+
+	inet_sock_destruct(sk);
+}
+
+static void mptcp_force_close(struct sock *sk)
+{
+	inet_sk_state_store(sk, TCP_CLOSE);
+	sk_common_release(sk);
+}
+
+static void subflow_ulp_fallback(struct sock *sk,
+				 struct mptcp_subflow_context *old_ctx)
+{
+	struct inet_connection_sock *icsk = inet_csk(sk);
+
+	mptcp_subflow_tcp_fallback(sk, old_ctx);
+	icsk->icsk_ulp_ops = NULL;
+	rcu_assign_pointer(icsk->icsk_ulp_data, NULL);
+	tcp_sk(sk)->is_mptcp = 0;
+}
+
 static struct sock *subflow_syn_recv_sock(const struct sock *sk,
 					  struct sk_buff *skb,
 					  struct request_sock *req,
@@ -356,13 +417,18 @@ static struct sock *subflow_syn_recv_sock(const struct sock *sk,
 {
 	struct mptcp_subflow_context *listener = mptcp_subflow_ctx(sk);
 	struct mptcp_subflow_request_sock *subflow_req;
-	struct tcp_options_received opt_rx;
+	struct mptcp_options_received mp_opt;
 	bool fallback_is_fatal = false;
 	struct sock *new_msk = NULL;
+	bool fallback = false;
 	struct sock *child;
 
 	pr_debug("listener=%p, req=%p, conn=%p", listener, req, listener->conn);
 
+	/* we need later a valid 'mp_capable' value even when options are not
+	 * parsed
+	 */
+	mp_opt.mp_capable = 0;
 	if (tcp_rsk(req)->is_mptcp == 0)
 		goto create_child;
 
@@ -377,26 +443,21 @@ static struct sock *subflow_syn_recv_sock(const struct sock *sk,
 			goto create_msk;
 		}
 
-		opt_rx.mptcp.mp_capable = 0;
-		mptcp_get_options(skb, &opt_rx);
-		if (opt_rx.mptcp.mp_capable) {
-			subflow_req->remote_key = opt_rx.mptcp.sndr_key;
-			subflow_req->remote_key_valid = 1;
-		} else {
-			subflow_req->mp_capable = 0;
+		mptcp_get_options(skb, &mp_opt);
+		if (!mp_opt.mp_capable) {
+			fallback = true;
 			goto create_child;
 		}
 
 create_msk:
-		new_msk = mptcp_sk_clone(listener->conn, req);
+		new_msk = mptcp_sk_clone(listener->conn, &mp_opt, req);
 		if (!new_msk)
-			subflow_req->mp_capable = 0;
+			fallback = true;
 	} else if (subflow_req->mp_join) {
 		fallback_is_fatal = true;
-		opt_rx.mptcp.mp_join = 0;
-		mptcp_get_options(skb, &opt_rx);
-		if (!opt_rx.mptcp.mp_join ||
-		    !subflow_hmac_valid(req, &opt_rx)) {
+		mptcp_get_options(skb, &mp_opt);
+		if (!mp_opt.mp_join ||
+		    !subflow_hmac_valid(req, &mp_opt)) {
 			SUBFLOW_REQ_INC_STATS(req, MPTCP_MIB_JOINACKMAC);
 			return NULL;
 		}
@@ -409,12 +470,18 @@ create_child:
 	if (child && *own_req) {
 		struct mptcp_subflow_context *ctx = mptcp_subflow_ctx(child);
 
-		/* we have null ctx on TCP fallback, which is fatal on
-		 * MPJ handshake
+		/* we need to fallback on ctx allocation failure and on pre-reqs
+		 * checking above. In the latter scenario we additionally need
+		 * to reset the context to non MPTCP status.
 		 */
-		if (!ctx) {
+		if (!ctx || fallback) {
 			if (fallback_is_fatal)
 				goto close_child;
+
+			if (ctx) {
+				subflow_ulp_fallback(child, ctx);
+				kfree_rcu(ctx, rcu);
+			}
 			goto out;
 		}
 
@@ -422,10 +489,17 @@ create_child:
 			/* new mpc subflow takes ownership of the newly
 			 * created mptcp socket
 			 */
-			inet_sk_state_store(new_msk, TCP_ESTABLISHED);
+			new_msk->sk_destruct = mptcp_sock_destruct;
 			mptcp_pm_new_connection(mptcp_sk(new_msk), 1);
 			ctx->conn = new_msk;
 			new_msk = NULL;
+
+			/* with OoO packets we can reach here without ingress
+			 * mpc option
+			 */
+			ctx->remote_key = mp_opt.sndr_key;
+			ctx->fully_established = mp_opt.mp_capable;
+			ctx->can_ack = mp_opt.mp_capable;
 		} else if (ctx->mp_join) {
 			struct mptcp_sock *owner;
 
@@ -444,7 +518,14 @@ create_child:
 out:
 	/* dispose of the left over mptcp master, if any */
 	if (unlikely(new_msk))
-		sock_put(new_msk);
+		mptcp_force_close(new_msk);
+
+	/* check for expected invariant - should never trigger, just help
+	 * catching eariler subtle bugs
+	 */
+	WARN_ON_ONCE(child && *own_req && tcp_sk(child)->is_mptcp &&
+		     (!mptcp_subflow_ctx(child) ||
+		      !mptcp_subflow_ctx(child)->conn));
 	return child;
 
 close_child:
@@ -931,6 +1012,16 @@ int mptcp_subflow_create_socket(struct sock *sk, struct socket **new_sock)
 	if (err)
 		return err;
 
+	/* the newly created socket really belongs to the owning MPTCP master
+	 * socket, even if for additional subflows the allocation is performed
+	 * by a kernel workqueue. Adjust inode references, so that the
+	 * procfs/diag interaces really show this one belonging to the correct
+	 * user.
+	 */
+	SOCK_INODE(sf)->i_ino = SOCK_INODE(sk->sk_socket)->i_ino;
+	SOCK_INODE(sf)->i_uid = SOCK_INODE(sk->sk_socket)->i_uid;
+	SOCK_INODE(sf)->i_gid = SOCK_INODE(sk->sk_socket)->i_gid;
+
 	subflow = mptcp_subflow_ctx(sf->sk);
 	pr_debug("subflow=%p", subflow);
 
@@ -1047,17 +1138,6 @@ static void subflow_ulp_release(struct sock *sk)
 	kfree_rcu(ctx, rcu);
 }
 
-static void subflow_ulp_fallback(struct sock *sk,
-				 struct mptcp_subflow_context *old_ctx)
-{
-	struct inet_connection_sock *icsk = inet_csk(sk);
-
-	mptcp_subflow_tcp_fallback(sk, old_ctx);
-	icsk->icsk_ulp_ops = NULL;
-	rcu_assign_pointer(icsk->icsk_ulp_data, NULL);
-	tcp_sk(sk)->is_mptcp = 0;
-}
-
 static void subflow_ulp_clone(const struct request_sock *req,
 			      struct sock *newsk,
 			      const gfp_t priority)
@@ -1091,9 +1171,6 @@ static void subflow_ulp_clone(const struct request_sock *req,
 		 * is fully established only after we receive the remote key
 		 */
 		new_ctx->mp_capable = 1;
-		new_ctx->fully_established = subflow_req->remote_key_valid;
-		new_ctx->can_ack = subflow_req->remote_key_valid;
-		new_ctx->remote_key = subflow_req->remote_key;
 		new_ctx->local_key = subflow_req->local_key;
 		new_ctx->token = subflow_req->token;
 		new_ctx->ssn_offset = subflow_req->ssn_offset;