Merge branch 'mptcp-next'

Mat Martineau says:

====================
mptcp: New features and cleanup

These patches have been tested in the MPTCP tree for a longer than usual
time (thanks to holiday schedules), and are ready for the net-next
branch. Changes include feature updates, small fixes, refactoring, and
some selftest changes.

Patch 1 fixes an OUTQ ioctl issue with TCP fallback sockets.

Patches 2, 3, and 6 add support of the MPTCP fastclose option (quick
shutdown of the full MPTCP connection, similar to TCP RST in regular
TCP), and a related self test.

Patch 4 cleans up some accept and poll code that is no longer needed
after the fastclose changes.

Patch 5 add userspace disconnect using AF_UNSPEC, which is used when
testing fastclose and makes the MPTCP socket's handling of AF_UNSPEC in
connect() more TCP-like.

Patches 7-11 refactor subflow creation to make better use of multiple
local endpoints and to better handle individual connection failures when
creating multiple subflows. Includes self test updates.

Patch 12 cleans up the way subflows are added to the MPTCP connection
list, eliminating the need for calls throughout the MPTCP code that had
to check the intermediate "join list" for entries to shift over to the
main "connection list".

Patch 13 refactors the MPTCP release_cb flags to use separate storage
for values only accessed with the socket lock held (no atomic ops
needed), and for values that need atomic operations.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

net/mptcp/options.c

@@ -768,6 +768,28 @@ static noinline bool mptcp_established_options_rst(struct sock *sk, struct sk_bu
 	return true;
 }
 
+static bool mptcp_established_options_fastclose(struct sock *sk,
+						unsigned int *size,
+						unsigned int remaining,
+						struct mptcp_out_options *opts)
+{
+	struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(sk);
+	struct mptcp_sock *msk = mptcp_sk(subflow->conn);
+
+	if (likely(!subflow->send_fastclose))
+		return false;
+
+	if (remaining < TCPOLEN_MPTCP_FASTCLOSE)
+		return false;
+
+	*size = TCPOLEN_MPTCP_FASTCLOSE;
+	opts->suboptions |= OPTION_MPTCP_FASTCLOSE;
+	opts->rcvr_key = msk->remote_key;
+
+	pr_debug("FASTCLOSE key=%llu", opts->rcvr_key);
+	return true;
+}
+
 static bool mptcp_established_options_mp_fail(struct sock *sk,
 					      unsigned int *size,
 					      unsigned int remaining,
@@ -806,10 +828,12 @@ bool mptcp_established_options(struct sock *sk, struct sk_buff *skb,
 		return false;
 
 	if (unlikely(skb && TCP_SKB_CB(skb)->tcp_flags & TCPHDR_RST)) {
-		if (mptcp_established_options_mp_fail(sk, &opt_size, remaining, opts)) {
+		if (mptcp_established_options_fastclose(sk, &opt_size, remaining, opts) ||
+		    mptcp_established_options_mp_fail(sk, &opt_size, remaining, opts)) {
 			*size += opt_size;
 			remaining -= opt_size;
 		}
+		/* MP_RST can be used with MP_FASTCLOSE and MP_FAIL if there is room */
 		if (mptcp_established_options_rst(sk, skb, &opt_size, remaining, opts)) {
 			*size += opt_size;
 			remaining -= opt_size;
@@ -1251,17 +1275,8 @@ void mptcp_write_options(__be32 *ptr, const struct tcp_sock *tp,
 		ptr += 2;
 	}
 
-	/* RST is mutually exclusive with everything else */
-	if (unlikely(OPTION_MPTCP_RST & opts->suboptions)) {
-		*ptr++ = mptcp_option(MPTCPOPT_RST,
-				      TCPOLEN_MPTCP_RST,
-				      opts->reset_transient,
-				      opts->reset_reason);
-		return;
-	}
-
-	/* DSS, MPC, MPJ and ADD_ADDR are mutually exclusive, see
-	 * mptcp_established_options*()
+	/* DSS, MPC, MPJ, ADD_ADDR, FASTCLOSE and RST are mutually exclusive,
+	 * see mptcp_established_options*()
 	 */
 	if (likely(OPTION_MPTCP_DSS & opts->suboptions)) {
 		struct mptcp_ext *mpext = &opts->ext_copy;
@@ -1370,27 +1385,29 @@ void mptcp_write_options(__be32 *ptr, const struct tcp_sock *tp,
 
 		/* MPC is additionally mutually exclusive with MP_PRIO */
 		goto mp_capable_done;
-	} else if (OPTION_MPTCP_MPJ_SYN & opts->suboptions) {
-		*ptr++ = mptcp_option(MPTCPOPT_MP_JOIN,
-				      TCPOLEN_MPTCP_MPJ_SYN,
-				      opts->backup, opts->join_id);
-		put_unaligned_be32(opts->token, ptr);
-		ptr += 1;
-		put_unaligned_be32(opts->nonce, ptr);
-		ptr += 1;
-	} else if (OPTION_MPTCP_MPJ_SYNACK & opts->suboptions) {
-		*ptr++ = mptcp_option(MPTCPOPT_MP_JOIN,
-				      TCPOLEN_MPTCP_MPJ_SYNACK,
-				      opts->backup, opts->join_id);
-		put_unaligned_be64(opts->thmac, ptr);
-		ptr += 2;
-		put_unaligned_be32(opts->nonce, ptr);
-		ptr += 1;
-	} else if (OPTION_MPTCP_MPJ_ACK & opts->suboptions) {
-		*ptr++ = mptcp_option(MPTCPOPT_MP_JOIN,
-				      TCPOLEN_MPTCP_MPJ_ACK, 0, 0);
-		memcpy(ptr, opts->hmac, MPTCPOPT_HMAC_LEN);
-		ptr += 5;
+	} else if (OPTIONS_MPTCP_MPJ & opts->suboptions) {
+		if (OPTION_MPTCP_MPJ_SYN & opts->suboptions) {
+			*ptr++ = mptcp_option(MPTCPOPT_MP_JOIN,
+					      TCPOLEN_MPTCP_MPJ_SYN,
+					      opts->backup, opts->join_id);
+			put_unaligned_be32(opts->token, ptr);
+			ptr += 1;
+			put_unaligned_be32(opts->nonce, ptr);
+			ptr += 1;
+		} else if (OPTION_MPTCP_MPJ_SYNACK & opts->suboptions) {
+			*ptr++ = mptcp_option(MPTCPOPT_MP_JOIN,
+					      TCPOLEN_MPTCP_MPJ_SYNACK,
+					      opts->backup, opts->join_id);
+			put_unaligned_be64(opts->thmac, ptr);
+			ptr += 2;
+			put_unaligned_be32(opts->nonce, ptr);
+			ptr += 1;
+		} else {
+			*ptr++ = mptcp_option(MPTCPOPT_MP_JOIN,
+					      TCPOLEN_MPTCP_MPJ_ACK, 0, 0);
+			memcpy(ptr, opts->hmac, MPTCPOPT_HMAC_LEN);
+			ptr += 5;
+		}
 	} else if (OPTION_MPTCP_ADD_ADDR & opts->suboptions) {
 		u8 len = TCPOLEN_MPTCP_ADD_ADDR_BASE;
 		u8 echo = MPTCP_ADDR_ECHO;
@@ -1447,6 +1464,24 @@ void mptcp_write_options(__be32 *ptr, const struct tcp_sock *tp,
 				ptr += 1;
 			}
 		}
+	} else if (unlikely(OPTION_MPTCP_FASTCLOSE & opts->suboptions)) {
+		/* FASTCLOSE is mutually exclusive with others except RST */
+		*ptr++ = mptcp_option(MPTCPOPT_MP_FASTCLOSE,
+				      TCPOLEN_MPTCP_FASTCLOSE,
+				      0, 0);
+		put_unaligned_be64(opts->rcvr_key, ptr);
+		ptr += 2;
+
+		if (OPTION_MPTCP_RST & opts->suboptions)
+			goto mp_rst;
+		return;
+	} else if (unlikely(OPTION_MPTCP_RST & opts->suboptions)) {
+mp_rst:
+		*ptr++ = mptcp_option(MPTCPOPT_RST,
+				      TCPOLEN_MPTCP_RST,
+				      opts->reset_transient,
+				      opts->reset_reason);
+		return;
 	}
 
 	if (OPTION_MPTCP_PRIO & opts->suboptions) {

net/mptcp/pm.c

@@ -172,9 +172,28 @@ void mptcp_pm_subflow_established(struct mptcp_sock *msk)
 	spin_unlock_bh(&pm->lock);
 }
 
-void mptcp_pm_subflow_closed(struct mptcp_sock *msk, u8 id)
+void mptcp_pm_subflow_check_next(struct mptcp_sock *msk, const struct sock *ssk,
+				 const struct mptcp_subflow_context *subflow)
 {
-	pr_debug("msk=%p", msk);
+	struct mptcp_pm_data *pm = &msk->pm;
+	bool update_subflows;
+
+	update_subflows = (ssk->sk_state == TCP_CLOSE) &&
+			  (subflow->request_join || subflow->mp_join);
+	if (!READ_ONCE(pm->work_pending) && !update_subflows)
+		return;
+
+	spin_lock_bh(&pm->lock);
+	if (update_subflows)
+		pm->subflows--;
+
+	/* Even if this subflow is not really established, tell the PM to try
+	 * to pick the next ones, if possible.
+	 */
+	if (mptcp_pm_nl_check_work_pending(msk))
+		mptcp_pm_schedule_work(msk, MPTCP_PM_SUBFLOW_ESTABLISHED);
+
+	spin_unlock_bh(&pm->lock);
 }
 
 void mptcp_pm_add_addr_received(struct mptcp_sock *msk,
@@ -356,7 +375,7 @@ void mptcp_pm_subflow_chk_stale(const struct mptcp_sock *msk, struct sock *ssk)
 	}
 }
 
-void mptcp_pm_data_init(struct mptcp_sock *msk)
+void mptcp_pm_data_reset(struct mptcp_sock *msk)
 {
 	msk->pm.add_addr_signaled = 0;
 	msk->pm.add_addr_accepted = 0;
@@ -370,11 +389,16 @@ void mptcp_pm_data_init(struct mptcp_sock *msk)
 	WRITE_ONCE(msk->pm.accept_subflow, false);
 	WRITE_ONCE(msk->pm.remote_deny_join_id0, false);
 	msk->pm.status = 0;
+	bitmap_fill(msk->pm.id_avail_bitmap, MPTCP_PM_MAX_ADDR_ID + 1);
+
+	mptcp_pm_nl_data_init(msk);
+}
 
+void mptcp_pm_data_init(struct mptcp_sock *msk)
+{
 	spin_lock_init(&msk->pm.lock);
 	INIT_LIST_HEAD(&msk->pm.anno_list);
-
-	mptcp_pm_nl_data_init(msk);
+	mptcp_pm_data_reset(msk);
 }
 
 void __init mptcp_pm_init(void)

net/mptcp/pm_netlink.c

@@ -38,10 +38,6 @@ struct mptcp_pm_add_entry {
 	u8			retrans_times;
 };
 
-/* max value of mptcp_addr_info.id */
-#define MAX_ADDR_ID		U8_MAX
-#define BITMAP_SZ DIV_ROUND_UP(MAX_ADDR_ID + 1, BITS_PER_LONG)
-
 struct pm_nl_pernet {
 	/* protects pernet updates */
 	spinlock_t		lock;
@@ -53,14 +49,14 @@ struct pm_nl_pernet {
 	unsigned int		local_addr_max;
 	unsigned int		subflows_max;
 	unsigned int		next_id;
-	unsigned long		id_bitmap[BITMAP_SZ];
+	DECLARE_BITMAP(id_bitmap, MPTCP_PM_MAX_ADDR_ID + 1);
 };
 
 #define MPTCP_PM_ADDR_MAX	8
 #define ADD_ADDR_RETRANS_MAX	3
 
 static bool addresses_equal(const struct mptcp_addr_info *a,
-			    struct mptcp_addr_info *b, bool use_port)
+			    const struct mptcp_addr_info *b, bool use_port)
 {
 	bool addr_equals = false;
 
@@ -169,11 +165,13 @@ select_local_address(const struct pm_nl_pernet *pernet,
 	msk_owned_by_me(msk);
 
 	rcu_read_lock();
-	__mptcp_flush_join_list(msk);
 	list_for_each_entry_rcu(entry, &pernet->local_addr_list, list) {
 		if (!(entry->flags & MPTCP_PM_ADDR_FLAG_SUBFLOW))
 			continue;
 
+		if (!test_bit(entry->addr.id, msk->pm.id_avail_bitmap))
+			continue;
+
 		if (entry->addr.family != sk->sk_family) {
 #if IS_ENABLED(CONFIG_MPTCP_IPV6)
 			if ((entry->addr.family == AF_INET &&
@@ -184,23 +182,17 @@ select_local_address(const struct pm_nl_pernet *pernet,
 				continue;
 		}
 
-		/* avoid any address already in use by subflows and
-		 * pending join
-		 */
-		if (!lookup_subflow_by_saddr(&msk->conn_list, &entry->addr)) {
-			ret = entry;
-			break;
-		}
+		ret = entry;
+		break;
 	}
 	rcu_read_unlock();
 	return ret;
 }
 
 static struct mptcp_pm_addr_entry *
-select_signal_address(struct pm_nl_pernet *pernet, unsigned int pos)
+select_signal_address(struct pm_nl_pernet *pernet, struct mptcp_sock *msk)
 {
 	struct mptcp_pm_addr_entry *entry, *ret = NULL;
-	int i = 0;
 
 	rcu_read_lock();
 	/* do not keep any additional per socket state, just signal
@@ -209,12 +201,14 @@ select_signal_address(struct pm_nl_pernet *pernet, unsigned int pos)
 	 * can lead to additional addresses not being announced.
 	 */
 	list_for_each_entry_rcu(entry, &pernet->local_addr_list, list) {
+		if (!test_bit(entry->addr.id, msk->pm.id_avail_bitmap))
+			continue;
+
 		if (!(entry->flags & MPTCP_PM_ADDR_FLAG_SIGNAL))
 			continue;
-		if (i++ == pos) {
-			ret = entry;
-			break;
-		}
+
+		ret = entry;
+		break;
 	}
 	rcu_read_unlock();
 	return ret;
@@ -256,12 +250,17 @@ unsigned int mptcp_pm_get_local_addr_max(struct mptcp_sock *msk)
 }
 EXPORT_SYMBOL_GPL(mptcp_pm_get_local_addr_max);
 
-static void check_work_pending(struct mptcp_sock *msk)
+bool mptcp_pm_nl_check_work_pending(struct mptcp_sock *msk)
 {
-	if (msk->pm.add_addr_signaled == mptcp_pm_get_add_addr_signal_max(msk) &&
-	    (msk->pm.local_addr_used == mptcp_pm_get_local_addr_max(msk) ||
-	     msk->pm.subflows == mptcp_pm_get_subflows_max(msk)))
+	struct pm_nl_pernet *pernet = net_generic(sock_net((struct sock *)msk), pm_nl_pernet_id);
+
+	if (msk->pm.subflows == mptcp_pm_get_subflows_max(msk) ||
+	    (find_next_and_bit(pernet->id_bitmap, msk->pm.id_avail_bitmap,
+			       MPTCP_PM_MAX_ADDR_ID + 1, 0) == MPTCP_PM_MAX_ADDR_ID + 1)) {
 		WRITE_ONCE(msk->pm.work_pending, false);
+		return false;
+	}
+	return true;
 }
 
 struct mptcp_pm_add_entry *
@@ -430,6 +429,7 @@ static bool lookup_address_in_vec(struct mptcp_addr_info *addrs, unsigned int nr
 static unsigned int fill_remote_addresses_vec(struct mptcp_sock *msk, bool fullmesh,
 					      struct mptcp_addr_info *addrs)
 {
+	bool deny_id0 = READ_ONCE(msk->pm.remote_deny_join_id0);
 	struct sock *sk = (struct sock *)msk, *ssk;
 	struct mptcp_subflow_context *subflow;
 	struct mptcp_addr_info remote = { 0 };
@@ -437,22 +437,28 @@ static unsigned int fill_remote_addresses_vec(struct mptcp_sock *msk, bool fullm
 	int i = 0;
 
 	subflows_max = mptcp_pm_get_subflows_max(msk);
+	remote_address((struct sock_common *)sk, &remote);
 
 	/* Non-fullmesh endpoint, fill in the single entry
 	 * corresponding to the primary MPC subflow remote address
 	 */
 	if (!fullmesh) {
-		remote_address((struct sock_common *)sk, &remote);
+		if (deny_id0)
+			return 0;
+
 		msk->pm.subflows++;
 		addrs[i++] = remote;
 	} else {
 		mptcp_for_each_subflow(msk, subflow) {
 			ssk = mptcp_subflow_tcp_sock(subflow);
-			remote_address((struct sock_common *)ssk, &remote);
-			if (!lookup_address_in_vec(addrs, i, &remote) &&
+			remote_address((struct sock_common *)ssk, &addrs[i]);
+			if (deny_id0 && addresses_equal(&addrs[i], &remote, false))
+				continue;
+
+			if (!lookup_address_in_vec(addrs, i, &addrs[i]) &&
 			    msk->pm.subflows < subflows_max) {
 				msk->pm.subflows++;
-				addrs[i++] = remote;
+				i++;
 			}
 		}
 	}
@@ -460,6 +466,35 @@ static unsigned int fill_remote_addresses_vec(struct mptcp_sock *msk, bool fullm
 	return i;
 }
 
+static struct mptcp_pm_addr_entry *
+__lookup_addr_by_id(struct pm_nl_pernet *pernet, unsigned int id)
+{
+	struct mptcp_pm_addr_entry *entry;
+
+	list_for_each_entry(entry, &pernet->local_addr_list, list) {
+		if (entry->addr.id == id)
+			return entry;
+	}
+	return NULL;
+}
+
+static int
+lookup_id_by_addr(struct pm_nl_pernet *pernet, const struct mptcp_addr_info *addr)
+{
+	struct mptcp_pm_addr_entry *entry;
+	int ret = -1;
+
+	rcu_read_lock();
+	list_for_each_entry(entry, &pernet->local_addr_list, list) {
+		if (addresses_equal(&entry->addr, addr, entry->addr.port)) {
+			ret = entry->addr.id;
+			break;
+		}
+	}
+	rcu_read_unlock();
+	return ret;
+}
+
 static void mptcp_pm_create_subflow_or_signal_addr(struct mptcp_sock *msk)
 {
 	struct sock *sk = (struct sock *)msk;
@@ -475,6 +510,19 @@ static void mptcp_pm_create_subflow_or_signal_addr(struct mptcp_sock *msk)
 	local_addr_max = mptcp_pm_get_local_addr_max(msk);
 	subflows_max = mptcp_pm_get_subflows_max(msk);
 
+	/* do lazy endpoint usage accounting for the MPC subflows */
+	if (unlikely(!(msk->pm.status & BIT(MPTCP_PM_MPC_ENDPOINT_ACCOUNTED))) && msk->first) {
+		struct mptcp_addr_info mpc_addr;
+		int mpc_id;
+
+		local_address((struct sock_common *)msk->first, &mpc_addr);
+		mpc_id = lookup_id_by_addr(pernet, &mpc_addr);
+		if (mpc_id >= 0)
+			__clear_bit(mpc_id, msk->pm.id_avail_bitmap);
+
+		msk->pm.status |= BIT(MPTCP_PM_MPC_ENDPOINT_ACCOUNTED);
+	}
+
 	pr_debug("local %d:%d signal %d:%d subflows %d:%d\n",
 		 msk->pm.local_addr_used, local_addr_max,
 		 msk->pm.add_addr_signaled, add_addr_signal_max,
@@ -482,47 +530,41 @@ static void mptcp_pm_create_subflow_or_signal_addr(struct mptcp_sock *msk)
 
 	/* check first for announce */
 	if (msk->pm.add_addr_signaled < add_addr_signal_max) {
-		local = select_signal_address(pernet,
-					      msk->pm.add_addr_signaled);
+		local = select_signal_address(pernet, msk);
 
 		if (local) {
 			if (mptcp_pm_alloc_anno_list(msk, local)) {
+				__clear_bit(local->addr.id, msk->pm.id_avail_bitmap);
 				msk->pm.add_addr_signaled++;
 				mptcp_pm_announce_addr(msk, &local->addr, false);
 				mptcp_pm_nl_addr_send_ack(msk);
 			}
-		} else {
-			/* pick failed, avoid fourther attempts later */
-			msk->pm.local_addr_used = add_addr_signal_max;
 		}
-
-		check_work_pending(msk);
 	}
 
 	/* check if should create a new subflow */
-	if (msk->pm.local_addr_used < local_addr_max &&
-	    msk->pm.subflows < subflows_max &&
-	    !READ_ONCE(msk->pm.remote_deny_join_id0)) {
+	while (msk->pm.local_addr_used < local_addr_max &&
+	       msk->pm.subflows < subflows_max) {
+		struct mptcp_addr_info addrs[MPTCP_PM_ADDR_MAX];
+		bool fullmesh;
+		int i, nr;
+
 		local = select_local_address(pernet, msk);
-		if (local) {
-			bool fullmesh = !!(local->flags & MPTCP_PM_ADDR_FLAG_FULLMESH);
-			struct mptcp_addr_info addrs[MPTCP_PM_ADDR_MAX];
-			int i, nr;
+		if (!local)
+			break;
 
-			msk->pm.local_addr_used++;
-			check_work_pending(msk);
-			nr = fill_remote_addresses_vec(msk, fullmesh, addrs);
-			spin_unlock_bh(&msk->pm.lock);
-			for (i = 0; i < nr; i++)
-				__mptcp_subflow_connect(sk, &local->addr, &addrs[i]);
-			spin_lock_bh(&msk->pm.lock);
-			return;
-		}
+		fullmesh = !!(local->flags & MPTCP_PM_ADDR_FLAG_FULLMESH);
 
-		/* lookup failed, avoid fourther attempts later */
-		msk->pm.local_addr_used = local_addr_max;
-		check_work_pending(msk);
+		msk->pm.local_addr_used++;
+		nr = fill_remote_addresses_vec(msk, fullmesh, addrs);
+		if (nr)
+			__clear_bit(local->addr.id, msk->pm.id_avail_bitmap);
+		spin_unlock_bh(&msk->pm.lock);
+		for (i = 0; i < nr; i++)
+			__mptcp_subflow_connect(sk, &local->addr, &addrs[i]);
+		spin_lock_bh(&msk->pm.lock);
 	}
+	mptcp_pm_nl_check_work_pending(msk);
 }
 
 static void mptcp_pm_nl_fully_established(struct mptcp_sock *msk)
@@ -552,7 +594,6 @@ static unsigned int fill_local_addresses_vec(struct mptcp_sock *msk,
 	subflows_max = mptcp_pm_get_subflows_max(msk);
 
 	rcu_read_lock();
-	__mptcp_flush_join_list(msk);
 	list_for_each_entry_rcu(entry, &pernet->local_addr_list, list) {
 		if (!(entry->flags & MPTCP_PM_ADDR_FLAG_FULLMESH))
 			continue;
@@ -641,7 +682,6 @@ void mptcp_pm_nl_addr_send_ack(struct mptcp_sock *msk)
 	    !mptcp_pm_should_rm_signal(msk))
 		return;
 
-	__mptcp_flush_join_list(msk);
 	subflow = list_first_entry_or_null(&msk->conn_list, typeof(*subflow), node);
 	if (subflow) {
 		struct sock *ssk = mptcp_subflow_tcp_sock(subflow);
@@ -711,6 +751,8 @@ static void mptcp_pm_nl_rm_addr_or_subflow(struct mptcp_sock *msk,
 		return;
 
 	for (i = 0; i < rm_list->nr; i++) {
+		bool removed = false;
+
 		list_for_each_entry_safe(subflow, tmp, &msk->conn_list, node) {
 			struct sock *ssk = mptcp_subflow_tcp_sock(subflow);
 			int how = RCV_SHUTDOWN | SEND_SHUTDOWN;
@@ -727,18 +769,24 @@ static void mptcp_pm_nl_rm_addr_or_subflow(struct mptcp_sock *msk,
 				 i, rm_list->ids[i], subflow->local_id, subflow->remote_id);
 			spin_unlock_bh(&msk->pm.lock);
 			mptcp_subflow_shutdown(sk, ssk, how);
+
+			/* the following takes care of updating the subflows counter */
 			mptcp_close_ssk(sk, ssk, subflow);
 			spin_lock_bh(&msk->pm.lock);
 
-			if (rm_type == MPTCP_MIB_RMADDR) {
-				msk->pm.add_addr_accepted--;
-				WRITE_ONCE(msk->pm.accept_addr, true);
-			} else if (rm_type == MPTCP_MIB_RMSUBFLOW) {
-				msk->pm.local_addr_used--;
-			}
-			msk->pm.subflows--;
+			removed = true;
 			__MPTCP_INC_STATS(sock_net(sk), rm_type);
 		}
+		__set_bit(rm_list->ids[1], msk->pm.id_avail_bitmap);
+		if (!removed)
+			continue;
+
+		if (rm_type == MPTCP_MIB_RMADDR) {
+			msk->pm.add_addr_accepted--;
+			WRITE_ONCE(msk->pm.accept_addr, true);
+		} else if (rm_type == MPTCP_MIB_RMSUBFLOW) {
+			msk->pm.local_addr_used--;
+		}
 	}
 }
 
@@ -759,6 +807,9 @@ void mptcp_pm_nl_work(struct mptcp_sock *msk)
 
 	msk_owned_by_me(msk);
 
+	if (!(pm->status & MPTCP_PM_WORK_MASK))
+		return;
+
 	spin_lock_bh(&msk->pm.lock);
 
 	pr_debug("msk=%p status=%x", msk, pm->status);
@@ -804,7 +855,7 @@ static int mptcp_pm_nl_append_new_local_addr(struct pm_nl_pernet *pernet,
 	/* to keep the code simple, don't do IDR-like allocation for address ID,
 	 * just bail when we exceed limits
 	 */
-	if (pernet->next_id == MAX_ADDR_ID)
+	if (pernet->next_id == MPTCP_PM_MAX_ADDR_ID)
 		pernet->next_id = 1;
 	if (pernet->addrs >= MPTCP_PM_ADDR_MAX)
 		goto out;
@@ -824,7 +875,7 @@ static int mptcp_pm_nl_append_new_local_addr(struct pm_nl_pernet *pernet,
 	if (!entry->addr.id) {
 find_next:
 		entry->addr.id = find_next_zero_bit(pernet->id_bitmap,
-						    MAX_ADDR_ID + 1,
+						    MPTCP_PM_MAX_ADDR_ID + 1,
 						    pernet->next_id);
 		if (!entry->addr.id && pernet->next_id != 1) {
 			pernet->next_id = 1;
@@ -1191,18 +1242,6 @@ static int mptcp_nl_cmd_add_addr(struct sk_buff *skb, struct genl_info *info)
 	return 0;
 }
 
-static struct mptcp_pm_addr_entry *
-__lookup_addr_by_id(struct pm_nl_pernet *pernet, unsigned int id)
-{
-	struct mptcp_pm_addr_entry *entry;
-
-	list_for_each_entry(entry, &pernet->local_addr_list, list) {
-		if (entry->addr.id == id)
-			return entry;
-	}
-	return NULL;
-}
-
 int mptcp_pm_get_flags_and_ifindex_by_id(struct net *net, unsigned int id,
 					 u8 *flags, int *ifindex)
 {
@@ -1461,7 +1500,7 @@ static int mptcp_nl_cmd_flush_addrs(struct sk_buff *skb, struct genl_info *info)
 	list_splice_init(&pernet->local_addr_list, &free_list);
 	__reset_counters(pernet);
 	pernet->next_id = 1;
-	bitmap_zero(pernet->id_bitmap, MAX_ADDR_ID + 1);
+	bitmap_zero(pernet->id_bitmap, MPTCP_PM_MAX_ADDR_ID + 1);
 	spin_unlock_bh(&pernet->lock);
 	mptcp_nl_remove_addrs_list(sock_net(skb->sk), &free_list);
 	synchronize_rcu();
@@ -1571,7 +1610,7 @@ static int mptcp_nl_cmd_dump_addrs(struct sk_buff *msg,
 	pernet = net_generic(net, pm_nl_pernet_id);
 
 	spin_lock_bh(&pernet->lock);
-	for (i = id; i < MAX_ADDR_ID + 1; i++) {
+	for (i = id; i < MPTCP_PM_MAX_ADDR_ID + 1; i++) {
 		if (test_bit(i, pernet->id_bitmap)) {
 			entry = __lookup_addr_by_id(pernet, i);
 			if (!entry)

net/mptcp/protocol.c

@@ -763,7 +763,7 @@ static bool move_skbs_to_msk(struct mptcp_sock *msk, struct sock *ssk)
 		if (!sock_owned_by_user(sk))
 			__mptcp_error_report(sk);
 		else
-			set_bit(MPTCP_ERROR_REPORT,  &msk->flags);
+			__set_bit(MPTCP_ERROR_REPORT,  &msk->cb_flags);
 	}
 
 	/* If the moves have caught up with the DATA_FIN sequence number
@@ -808,47 +808,38 @@ void mptcp_data_ready(struct sock *sk, struct sock *ssk)
 	mptcp_data_unlock(sk);
 }
 
-static bool mptcp_do_flush_join_list(struct mptcp_sock *msk)
+static bool __mptcp_finish_join(struct mptcp_sock *msk, struct sock *ssk)
 {
-	struct mptcp_subflow_context *subflow;
-	bool ret = false;
+	struct sock *sk = (struct sock *)msk;
 
-	if (likely(list_empty(&msk->join_list)))
+	if (sk->sk_state != TCP_ESTABLISHED)
 		return false;
 
-	spin_lock_bh(&msk->join_list_lock);
-	list_for_each_entry(subflow, &msk->join_list, node) {
-		u32 sseq = READ_ONCE(subflow->setsockopt_seq);
-
-		mptcp_propagate_sndbuf((struct sock *)msk, mptcp_subflow_tcp_sock(subflow));
-		if (READ_ONCE(msk->setsockopt_seq) != sseq)
-			ret = true;
-	}
-	list_splice_tail_init(&msk->join_list, &msk->conn_list);
-	spin_unlock_bh(&msk->join_list_lock);
-
-	return ret;
-}
-
-void __mptcp_flush_join_list(struct mptcp_sock *msk)
-{
-	if (likely(!mptcp_do_flush_join_list(msk)))
-		return;
+	/* attach to msk socket only after we are sure we will deal with it
+	 * at close time
+	 */
+	if (sk->sk_socket && !ssk->sk_socket)
+		mptcp_sock_graft(ssk, sk->sk_socket);
 
-	if (!test_and_set_bit(MPTCP_WORK_SYNC_SETSOCKOPT, &msk->flags))
-		mptcp_schedule_work((struct sock *)msk);
+	mptcp_propagate_sndbuf((struct sock *)msk, ssk);
+	mptcp_sockopt_sync_locked(msk, ssk);
+	return true;
 }
 
-static void mptcp_flush_join_list(struct mptcp_sock *msk)
+static void __mptcp_flush_join_list(struct sock *sk)
 {
-	bool sync_needed = test_and_clear_bit(MPTCP_WORK_SYNC_SETSOCKOPT, &msk->flags);
-
-	might_sleep();
+	struct mptcp_subflow_context *tmp, *subflow;
+	struct mptcp_sock *msk = mptcp_sk(sk);
 
-	if (!mptcp_do_flush_join_list(msk) && !sync_needed)
-		return;
+	list_for_each_entry_safe(subflow, tmp, &msk->join_list, node) {
+		struct sock *ssk = mptcp_subflow_tcp_sock(subflow);
+		bool slow = lock_sock_fast(ssk);
 
-	mptcp_sockopt_sync_all(msk);
+		list_move_tail(&subflow->node, &msk->conn_list);
+		if (!__mptcp_finish_join(msk, ssk))
+			mptcp_subflow_reset(ssk);
+		unlock_sock_fast(ssk, slow);
+	}
 }
 
 static bool mptcp_timer_pending(struct sock *sk)
@@ -1526,9 +1517,8 @@ static void mptcp_update_post_push(struct mptcp_sock *msk,
 
 void mptcp_check_and_set_pending(struct sock *sk)
 {
-	if (mptcp_send_head(sk) &&
-	    !test_bit(MPTCP_PUSH_PENDING, &mptcp_sk(sk)->flags))
-		set_bit(MPTCP_PUSH_PENDING, &mptcp_sk(sk)->flags);
+	if (mptcp_send_head(sk))
+		mptcp_sk(sk)->push_pending |= BIT(MPTCP_PUSH_PENDING);
 }
 
 void __mptcp_push_pending(struct sock *sk, unsigned int flags)
@@ -1549,7 +1539,6 @@ void __mptcp_push_pending(struct sock *sk, unsigned int flags)
 			int ret = 0;
 
 			prev_ssk = ssk;
-			__mptcp_flush_join_list(msk);
 			ssk = mptcp_subflow_get_send(msk);
 
 			/* First check. If the ssk has changed since
@@ -1954,7 +1943,6 @@ static bool __mptcp_move_skbs(struct mptcp_sock *msk)
 	unsigned int moved = 0;
 	bool ret, done;
 
-	mptcp_flush_join_list(msk);
 	do {
 		struct sock *ssk = mptcp_subflow_recv_lookup(msk);
 		bool slowpath;
@@ -2145,7 +2133,7 @@ static void mptcp_retransmit_timer(struct timer_list *t)
 			mptcp_schedule_work(sk);
 	} else {
 		/* delegate our work to tcp_release_cb() */
-		set_bit(MPTCP_RETRANSMIT, &msk->flags);
+		__set_bit(MPTCP_RETRANSMIT, &msk->cb_flags);
 	}
 	bh_unlock_sock(sk);
 	sock_put(sk);
@@ -2253,6 +2241,10 @@ bool __mptcp_retransmit_pending_data(struct sock *sk)
 	return true;
 }
 
+/* flags for __mptcp_close_ssk() */
+#define MPTCP_CF_PUSH		BIT(1)
+#define MPTCP_CF_FASTCLOSE	BIT(2)
+
 /* subflow sockets can be either outgoing (connect) or incoming
  * (accept).
  *
@@ -2262,22 +2254,37 @@ bool __mptcp_retransmit_pending_data(struct sock *sk)
  * parent socket.
  */
 static void __mptcp_close_ssk(struct sock *sk, struct sock *ssk,
-			      struct mptcp_subflow_context *subflow)
+			      struct mptcp_subflow_context *subflow,
+			      unsigned int flags)
 {
 	struct mptcp_sock *msk = mptcp_sk(sk);
-	bool need_push;
+	bool need_push, dispose_it;
 
-	list_del(&subflow->node);
+	dispose_it = !msk->subflow || ssk != msk->subflow->sk;
+	if (dispose_it)
+		list_del(&subflow->node);
 
 	lock_sock_nested(ssk, SINGLE_DEPTH_NESTING);
 
+	if (flags & MPTCP_CF_FASTCLOSE)
+		subflow->send_fastclose = 1;
+
+	need_push = (flags & MPTCP_CF_PUSH) && __mptcp_retransmit_pending_data(sk);
+	if (!dispose_it) {
+		tcp_disconnect(ssk, 0);
+		msk->subflow->state = SS_UNCONNECTED;
+		mptcp_subflow_ctx_reset(subflow);
+		release_sock(ssk);
+
+		goto out;
+	}
+
 	/* if we are invoked by the msk cleanup code, the subflow is
 	 * already orphaned
 	 */
 	if (ssk->sk_socket)
 		sock_orphan(ssk);
 
-	need_push = __mptcp_retransmit_pending_data(sk);
 	subflow->disposable = 1;
 
 	/* if ssk hit tcp_done(), tcp_cleanup_ulp() cleared the related ops
@@ -2297,14 +2304,12 @@ static void __mptcp_close_ssk(struct sock *sk, struct sock *ssk,
 
 	sock_put(ssk);
 
-	if (ssk == msk->last_snd)
-		msk->last_snd = NULL;
-
 	if (ssk == msk->first)
 		msk->first = NULL;
 
-	if (msk->subflow && ssk == msk->subflow->sk)
-		mptcp_dispose_initial_subflow(msk);
+out:
+	if (ssk == msk->last_snd)
+		msk->last_snd = NULL;
 
 	if (need_push)
 		__mptcp_push_pending(sk, 0);
@@ -2315,7 +2320,13 @@ void mptcp_close_ssk(struct sock *sk, struct sock *ssk,
 {
 	if (sk->sk_state == TCP_ESTABLISHED)
 		mptcp_event(MPTCP_EVENT_SUB_CLOSED, mptcp_sk(sk), ssk, GFP_KERNEL);
-	__mptcp_close_ssk(sk, ssk, subflow);
+
+	/* subflow aborted before reaching the fully_established status
+	 * attempt the creation of the next subflow
+	 */
+	mptcp_pm_subflow_check_next(mptcp_sk(sk), ssk, subflow);
+
+	__mptcp_close_ssk(sk, ssk, subflow, MPTCP_CF_PUSH);
 }
 
 static unsigned int mptcp_sync_mss(struct sock *sk, u32 pmtu)
@@ -2467,12 +2478,10 @@ static void mptcp_worker(struct work_struct *work)
 		goto unlock;
 
 	mptcp_check_data_fin_ack(sk);
-	mptcp_flush_join_list(msk);
 
 	mptcp_check_fastclose(msk);
 
-	if (msk->pm.status)
-		mptcp_pm_nl_work(msk);
+	mptcp_pm_nl_work(msk);
 
 	if (test_and_clear_bit(MPTCP_WORK_EOF, &msk->flags))
 		mptcp_check_for_eof(msk);
@@ -2506,8 +2515,6 @@ static int __mptcp_init_sock(struct sock *sk)
 {
 	struct mptcp_sock *msk = mptcp_sk(sk);
 
-	spin_lock_init(&msk->join_list_lock);
-
 	INIT_LIST_HEAD(&msk->conn_list);
 	INIT_LIST_HEAD(&msk->join_list);
 	INIT_LIST_HEAD(&msk->rtx_queue);
@@ -2533,9 +2540,20 @@ static int __mptcp_init_sock(struct sock *sk)
 	return 0;
 }
 
-static int mptcp_init_sock(struct sock *sk)
+static void mptcp_ca_reset(struct sock *sk)
 {
 	struct inet_connection_sock *icsk = inet_csk(sk);
+
+	tcp_assign_congestion_control(sk);
+	strcpy(mptcp_sk(sk)->ca_name, icsk->icsk_ca_ops->name);
+
+	/* no need to keep a reference to the ops, the name will suffice */
+	tcp_cleanup_congestion_control(sk);
+	icsk->icsk_ca_ops = NULL;
+}
+
+static int mptcp_init_sock(struct sock *sk)
+{
 	struct net *net = sock_net(sk);
 	int ret;
 
@@ -2556,12 +2574,7 @@ static int mptcp_init_sock(struct sock *sk)
 	/* fetch the ca name; do it outside __mptcp_init_sock(), so that clone will
 	 * propagate the correct value
 	 */
-	tcp_assign_congestion_control(sk);
-	strcpy(mptcp_sk(sk)->ca_name, icsk->icsk_ca_ops->name);
-
-	/* no need to keep a reference to the ops, the name will suffice */
-	tcp_cleanup_congestion_control(sk);
-	icsk->icsk_ca_ops = NULL;
+	mptcp_ca_reset(sk);
 
 	sk_sockets_allocated_inc(sk);
 	sk->sk_rcvbuf = sock_net(sk)->ipv4.sysctl_tcp_rmem[1];
@@ -2666,6 +2679,7 @@ static void __mptcp_check_send_data_fin(struct sock *sk)
 	 * state now
 	 */
 	if (__mptcp_check_fallback(msk)) {
+		WRITE_ONCE(msk->snd_una, msk->write_seq);
 		if ((1 << sk->sk_state) & (TCPF_CLOSING | TCPF_LAST_ACK)) {
 			inet_sk_state_store(sk, TCP_CLOSE);
 			mptcp_close_wake_up(sk);
@@ -2674,7 +2688,6 @@ static void __mptcp_check_send_data_fin(struct sock *sk)
 		}
 	}
 
-	mptcp_flush_join_list(msk);
 	mptcp_for_each_subflow(msk, subflow) {
 		struct sock *tcp_sk = mptcp_subflow_tcp_sock(subflow);
 
@@ -2707,21 +2720,20 @@ static void __mptcp_destroy_sock(struct sock *sk)
 
 	might_sleep();
 
-	/* be sure to always acquire the join list lock, to sync vs
-	 * mptcp_finish_join().
-	 */
-	spin_lock_bh(&msk->join_list_lock);
-	list_splice_tail_init(&msk->join_list, &msk->conn_list);
-	spin_unlock_bh(&msk->join_list_lock);
+	/* 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);
 	sk_stop_timer(sk, &sk->sk_timer);
 	msk->pm.status = 0;
 
+	/* clears msk->subflow, allowing the following loop to close
+	 * even the initial subflow
+	 */
+	mptcp_dispose_initial_subflow(msk);
 	list_for_each_entry_safe(subflow, tmp, &conn_list, node) {
 		struct sock *ssk = mptcp_subflow_tcp_sock(subflow);
-		__mptcp_close_ssk(sk, ssk, subflow);
+		__mptcp_close_ssk(sk, ssk, subflow, 0);
 	}
 
 	sk->sk_prot->destroy(sk);
@@ -2732,7 +2744,6 @@ static void __mptcp_destroy_sock(struct sock *sk)
 	xfrm_sk_free_policy(sk);
 
 	sk_refcnt_debug_release(sk);
-	mptcp_dispose_initial_subflow(msk);
 	sock_put(sk);
 }
 
@@ -2768,6 +2779,9 @@ static void mptcp_close(struct sock *sk, long timeout)
 
 	sock_hold(sk);
 	pr_debug("msk=%p state=%d", sk, sk->sk_state);
+	if (mptcp_sk(sk)->token)
+		mptcp_event(MPTCP_EVENT_CLOSED, mptcp_sk(sk), NULL, GFP_KERNEL);
+
 	if (sk->sk_state == TCP_CLOSE) {
 		__mptcp_destroy_sock(sk);
 		do_cancel_work = true;
@@ -2778,9 +2792,6 @@ static void mptcp_close(struct sock *sk, long timeout)
 	if (do_cancel_work)
 		mptcp_cancel_work(sk);
 
-	if (mptcp_sk(sk)->token)
-		mptcp_event(MPTCP_EVENT_CLOSED, mptcp_sk(sk), NULL, GFP_KERNEL);
-
 	sock_put(sk);
 }
 
@@ -2812,15 +2823,38 @@ static int mptcp_disconnect(struct sock *sk, int flags)
 	struct mptcp_subflow_context *subflow;
 	struct mptcp_sock *msk = mptcp_sk(sk);
 
-	mptcp_do_flush_join_list(msk);
+	inet_sk_state_store(sk, TCP_CLOSE);
 
 	mptcp_for_each_subflow(msk, subflow) {
 		struct sock *ssk = mptcp_subflow_tcp_sock(subflow);
 
-		lock_sock(ssk);
-		tcp_disconnect(ssk, flags);
-		release_sock(ssk);
+		__mptcp_close_ssk(sk, ssk, subflow, MPTCP_CF_FASTCLOSE);
 	}
+
+	sk_stop_timer(sk, &msk->sk.icsk_retransmit_timer);
+	sk_stop_timer(sk, &sk->sk_timer);
+
+	if (mptcp_sk(sk)->token)
+		mptcp_event(MPTCP_EVENT_CLOSED, mptcp_sk(sk), NULL, GFP_KERNEL);
+
+	mptcp_destroy_common(msk);
+	msk->last_snd = NULL;
+	WRITE_ONCE(msk->flags, 0);
+	msk->cb_flags = 0;
+	msk->push_pending = 0;
+	msk->recovery = false;
+	msk->can_ack = false;
+	msk->fully_established = false;
+	msk->rcv_data_fin = false;
+	msk->snd_data_fin_enable = false;
+	msk->rcv_fastclose = false;
+	msk->use_64bit_ack = false;
+	WRITE_ONCE(msk->csum_enabled, mptcp_is_checksum_enabled(sock_net(sk)));
+	mptcp_pm_data_reset(msk);
+	mptcp_ca_reset(sk);
+
+	sk->sk_shutdown = 0;
+	sk_error_report(sk);
 	return 0;
 }
 
@@ -2960,9 +2994,11 @@ void mptcp_destroy_common(struct mptcp_sock *msk)
 	__mptcp_clear_xmit(sk);
 
 	/* move to sk_receive_queue, sk_stream_kill_queues will purge it */
+	mptcp_data_lock(sk);
 	skb_queue_splice_tail_init(&msk->receive_queue, &sk->sk_receive_queue);
 	__skb_queue_purge(&sk->sk_receive_queue);
 	skb_rbtree_purge(&msk->out_of_order_queue);
+	mptcp_data_unlock(sk);
 
 	/* move all the rx fwd alloc into the sk_mem_reclaim_final in
 	 * inet_sock_destruct() will dispose it
@@ -2986,7 +3022,7 @@ void __mptcp_data_acked(struct sock *sk)
 	if (!sock_owned_by_user(sk))
 		__mptcp_clean_una(sk);
 	else
-		set_bit(MPTCP_CLEAN_UNA, &mptcp_sk(sk)->flags);
+		__set_bit(MPTCP_CLEAN_UNA, &mptcp_sk(sk)->cb_flags);
 
 	if (mptcp_pending_data_fin_ack(sk))
 		mptcp_schedule_work(sk);
@@ -3005,20 +3041,23 @@ void __mptcp_check_push(struct sock *sk, struct sock *ssk)
 		else if (xmit_ssk)
 			mptcp_subflow_delegate(mptcp_subflow_ctx(xmit_ssk), MPTCP_DELEGATE_SEND);
 	} else {
-		set_bit(MPTCP_PUSH_PENDING, &mptcp_sk(sk)->flags);
+		__set_bit(MPTCP_PUSH_PENDING, &mptcp_sk(sk)->cb_flags);
 	}
 }
 
+#define MPTCP_FLAGS_PROCESS_CTX_NEED (BIT(MPTCP_PUSH_PENDING) | \
+				      BIT(MPTCP_RETRANSMIT) | \
+				      BIT(MPTCP_FLUSH_JOIN_LIST))
+
 /* processes deferred events and flush wmem */
 static void mptcp_release_cb(struct sock *sk)
+	__must_hold(&sk->sk_lock.slock)
 {
-	for (;;) {
-		unsigned long flags = 0;
+	struct mptcp_sock *msk = mptcp_sk(sk);
 
-		if (test_and_clear_bit(MPTCP_PUSH_PENDING, &mptcp_sk(sk)->flags))
-			flags |= BIT(MPTCP_PUSH_PENDING);
-		if (test_and_clear_bit(MPTCP_RETRANSMIT, &mptcp_sk(sk)->flags))
-			flags |= BIT(MPTCP_RETRANSMIT);
+	for (;;) {
+		unsigned long flags = (msk->cb_flags & MPTCP_FLAGS_PROCESS_CTX_NEED) |
+				      msk->push_pending;
 		if (!flags)
 			break;
 
@@ -3029,8 +3068,11 @@ static void mptcp_release_cb(struct sock *sk)
 		 *    datapath acquires the msk socket spinlock while helding
 		 *    the subflow socket lock
 		 */
-
+		msk->push_pending = 0;
+		msk->cb_flags &= ~flags;
 		spin_unlock_bh(&sk->sk_lock.slock);
+		if (flags & BIT(MPTCP_FLUSH_JOIN_LIST))
+			__mptcp_flush_join_list(sk);
 		if (flags & BIT(MPTCP_PUSH_PENDING))
 			__mptcp_push_pending(sk, 0);
 		if (flags & BIT(MPTCP_RETRANSMIT))
@@ -3043,11 +3085,11 @@ static void mptcp_release_cb(struct sock *sk)
 	/* be sure to set the current sk state before tacking actions
 	 * depending on sk_state
 	 */
-	if (test_and_clear_bit(MPTCP_CONNECTED, &mptcp_sk(sk)->flags))
+	if (__test_and_clear_bit(MPTCP_CONNECTED, &msk->cb_flags))
 		__mptcp_set_connected(sk);
-	if (test_and_clear_bit(MPTCP_CLEAN_UNA, &mptcp_sk(sk)->flags))
+	if (__test_and_clear_bit(MPTCP_CLEAN_UNA, &msk->cb_flags))
 		__mptcp_clean_una_wakeup(sk);
-	if (test_and_clear_bit(MPTCP_ERROR_REPORT, &mptcp_sk(sk)->flags))
+	if (__test_and_clear_bit(MPTCP_ERROR_REPORT, &msk->cb_flags))
 		__mptcp_error_report(sk);
 
 	__mptcp_update_rmem(sk);
@@ -3089,7 +3131,7 @@ void mptcp_subflow_process_delegated(struct sock *ssk)
 		if (!sock_owned_by_user(sk))
 			__mptcp_subflow_push_pending(sk, ssk);
 		else
-			set_bit(MPTCP_PUSH_PENDING, &mptcp_sk(sk)->flags);
+			__set_bit(MPTCP_PUSH_PENDING, &mptcp_sk(sk)->cb_flags);
 		mptcp_data_unlock(sk);
 		mptcp_subflow_delegated_done(subflow, MPTCP_DELEGATE_SEND);
 	}
@@ -3175,8 +3217,7 @@ bool mptcp_finish_join(struct sock *ssk)
 	struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(ssk);
 	struct mptcp_sock *msk = mptcp_sk(subflow->conn);
 	struct sock *parent = (void *)msk;
-	struct socket *parent_sock;
-	bool ret;
+	bool ret = true;
 
 	pr_debug("msk=%p, subflow=%p", msk, subflow);
 
@@ -3189,35 +3230,38 @@ bool mptcp_finish_join(struct sock *ssk)
 	if (!msk->pm.server_side)
 		goto out;
 
-	if (!mptcp_pm_allow_new_subflow(msk)) {
-		subflow->reset_reason = MPTCP_RST_EPROHIBIT;
-		return false;
-	}
+	if (!mptcp_pm_allow_new_subflow(msk))
+		goto err_prohibited;
+
+	if (WARN_ON_ONCE(!list_empty(&subflow->node)))
+		goto err_prohibited;
 
-	/* active connections are already on conn_list, and we can't acquire
-	 * msk lock here.
-	 * use the join list lock as synchronization point and double-check
-	 * msk status to avoid racing with __mptcp_destroy_sock()
+	/* active connections are already on conn_list.
+	 * If we can't acquire msk socket lock here, let the release callback
+	 * handle it
 	 */
-	spin_lock_bh(&msk->join_list_lock);
-	ret = inet_sk_state_load(parent) == TCP_ESTABLISHED;
-	if (ret && !WARN_ON_ONCE(!list_empty(&subflow->node))) {
-		list_add_tail(&subflow->node, &msk->join_list);
+	mptcp_data_lock(parent);
+	if (!sock_owned_by_user(parent)) {
+		ret = __mptcp_finish_join(msk, ssk);
+		if (ret) {
+			sock_hold(ssk);
+			list_add_tail(&subflow->node, &msk->conn_list);
+		}
+	} else {
 		sock_hold(ssk);
+		list_add_tail(&subflow->node, &msk->join_list);
+		__set_bit(MPTCP_FLUSH_JOIN_LIST, &msk->cb_flags);
 	}
-	spin_unlock_bh(&msk->join_list_lock);
+	mptcp_data_unlock(parent);
+
 	if (!ret) {
+err_prohibited:
 		subflow->reset_reason = MPTCP_RST_EPROHIBIT;
 		return false;
 	}
 
-	/* attach to msk socket only after we are sure he will deal with us
-	 * at close time
-	 */
-	parent_sock = READ_ONCE(parent->sk_socket);
-	if (parent_sock && !ssk->sk_socket)
-		mptcp_sock_graft(ssk, parent_sock);
 	subflow->map_seq = READ_ONCE(msk->ack_seq);
+
 out:
 	mptcp_event(MPTCP_EVENT_SUB_ESTABLISHED, msk, ssk, GFP_ATOMIC);
 	return true;
@@ -3352,9 +3396,20 @@ static int mptcp_stream_connect(struct socket *sock, struct sockaddr *uaddr,
 	struct mptcp_sock *msk = mptcp_sk(sock->sk);
 	struct mptcp_subflow_context *subflow;
 	struct socket *ssock;
-	int err;
+	int err = -EINVAL;
 
 	lock_sock(sock->sk);
+	if (uaddr) {
+		if (addr_len < sizeof(uaddr->sa_family))
+			goto unlock;
+
+		if (uaddr->sa_family == AF_UNSPEC) {
+			err = mptcp_disconnect(sock->sk, flags);
+			sock->state = err ? SS_DISCONNECTING : SS_UNCONNECTED;
+			goto unlock;
+		}
+	}
+
 	if (sock->state != SS_UNCONNECTED && msk->subflow) {
 		/* pending connection or invalid state, let existing subflow
 		 * cope with that
@@ -3364,10 +3419,8 @@ static int mptcp_stream_connect(struct socket *sock, struct sockaddr *uaddr,
 	}
 
 	ssock = __mptcp_nmpc_socket(msk);
-	if (!ssock) {
-		err = -EINVAL;
+	if (!ssock)
 		goto unlock;
-	}
 
 	mptcp_token_destroy(msk);
 	inet_sk_state_store(sock->sk, TCP_SYN_SENT);
@@ -3441,17 +3494,9 @@ static int mptcp_stream_accept(struct socket *sock, struct socket *newsock,
 
 	pr_debug("msk=%p", msk);
 
-	lock_sock(sock->sk);
-	if (sock->sk->sk_state != TCP_LISTEN)
-		goto unlock_fail;
-
 	ssock = __mptcp_nmpc_socket(msk);
 	if (!ssock)
-		goto unlock_fail;
-
-	clear_bit(MPTCP_DATA_READY, &msk->flags);
-	sock_hold(ssock->sk);
-	release_sock(sock->sk);
+		return -EINVAL;
 
 	err = ssock->ops->accept(sock, newsock, flags, kern);
 	if (err == 0 && !mptcp_is_tcpsk(newsock->sk)) {
@@ -3481,7 +3526,6 @@ static int mptcp_stream_accept(struct socket *sock, struct socket *newsock,
 		/* set ssk->sk_socket of accept()ed flows to mptcp socket.
 		 * This is needed so NOSPACE flag can be set from tcp stack.
 		 */
-		mptcp_flush_join_list(msk);
 		mptcp_for_each_subflow(msk, subflow) {
 			struct sock *ssk = mptcp_subflow_tcp_sock(subflow);
 
@@ -3491,14 +3535,7 @@ static int mptcp_stream_accept(struct socket *sock, struct socket *newsock,
 		release_sock(newsk);
 	}
 
-	if (inet_csk_listen_poll(ssock->sk))
-		set_bit(MPTCP_DATA_READY, &msk->flags);
-	sock_put(ssock->sk);
 	return err;
-
-unlock_fail:
-	release_sock(sock->sk);
-	return -EINVAL;
 }
 
 static __poll_t mptcp_check_readable(struct mptcp_sock *msk)
@@ -3544,8 +3581,12 @@ static __poll_t mptcp_poll(struct file *file, struct socket *sock,
 
 	state = inet_sk_state_load(sk);
 	pr_debug("msk=%p state=%d flags=%lx", msk, state, msk->flags);
-	if (state == TCP_LISTEN)
-		return test_bit(MPTCP_DATA_READY, &msk->flags) ? EPOLLIN | EPOLLRDNORM : 0;
+	if (state == TCP_LISTEN) {
+		if (WARN_ON_ONCE(!msk->subflow || !msk->subflow->sk))
+			return 0;
+
+		return inet_csk_listen_poll(msk->subflow->sk);
+	}
 
 	if (state != TCP_SYN_SENT && state != TCP_SYN_RECV) {
 		mask |= mptcp_check_readable(msk);

net/mptcp/protocol.h

@@ -110,19 +110,20 @@
 /* MPTCP TCPRST flags */
 #define MPTCP_RST_TRANSIENT	BIT(0)
 
-/* MPTCP socket flags */
-#define MPTCP_DATA_READY	0
+/* MPTCP socket atomic flags */
 #define MPTCP_NOSPACE		1
 #define MPTCP_WORK_RTX		2
 #define MPTCP_WORK_EOF		3
 #define MPTCP_FALLBACK_DONE	4
 #define MPTCP_WORK_CLOSE_SUBFLOW 5
-#define MPTCP_PUSH_PENDING	6
-#define MPTCP_CLEAN_UNA		7
-#define MPTCP_ERROR_REPORT	8
-#define MPTCP_RETRANSMIT	9
-#define MPTCP_WORK_SYNC_SETSOCKOPT 10
-#define MPTCP_CONNECTED		11
+
+/* MPTCP socket release cb flags */
+#define MPTCP_PUSH_PENDING	1
+#define MPTCP_CLEAN_UNA		2
+#define MPTCP_ERROR_REPORT	3
+#define MPTCP_RETRANSMIT	4
+#define MPTCP_FLUSH_JOIN_LIST	5
+#define MPTCP_CONNECTED		6
 
 static inline bool before64(__u64 seq1, __u64 seq2)
 {
@@ -174,16 +175,25 @@ enum mptcp_pm_status {
 	MPTCP_PM_ADD_ADDR_SEND_ACK,
 	MPTCP_PM_RM_ADDR_RECEIVED,
 	MPTCP_PM_ESTABLISHED,
-	MPTCP_PM_ALREADY_ESTABLISHED,	/* persistent status, set after ESTABLISHED event */
 	MPTCP_PM_SUBFLOW_ESTABLISHED,
+	MPTCP_PM_ALREADY_ESTABLISHED,	/* persistent status, set after ESTABLISHED event */
+	MPTCP_PM_MPC_ENDPOINT_ACCOUNTED /* persistent status, set after MPC local address is
+					 * accounted int id_avail_bitmap
+					 */
 };
 
+/* Status bits below MPTCP_PM_ALREADY_ESTABLISHED need pm worker actions */
+#define MPTCP_PM_WORK_MASK ((1 << MPTCP_PM_ALREADY_ESTABLISHED) - 1)
+
 enum mptcp_addr_signal_status {
 	MPTCP_ADD_ADDR_SIGNAL,
 	MPTCP_ADD_ADDR_ECHO,
 	MPTCP_RM_ADDR_SIGNAL,
 };
 
+/* max value of mptcp_addr_info.id */
+#define MPTCP_PM_MAX_ADDR_ID		U8_MAX
+
 struct mptcp_pm_data {
 	struct mptcp_addr_info local;
 	struct mptcp_addr_info remote;
@@ -202,6 +212,7 @@ struct mptcp_pm_data {
 	u8		local_addr_used;
 	u8		subflows;
 	u8		status;
+	DECLARE_BITMAP(id_avail_bitmap, MPTCP_PM_MAX_ADDR_ID + 1);
 	struct mptcp_rm_list rm_list_tx;
 	struct mptcp_rm_list rm_list_rx;
 };
@@ -241,6 +252,8 @@ struct mptcp_sock {
 	u32		token;
 	int		rmem_released;
 	unsigned long	flags;
+	unsigned long	cb_flags;
+	unsigned long	push_pending;
 	bool		recovery;		/* closing subflow write queue reinjected */
 	bool		can_ack;
 	bool		fully_established;
@@ -252,7 +265,6 @@ struct mptcp_sock {
 	u8		recvmsg_inq:1,
 			cork:1,
 			nodelay:1;
-	spinlock_t	join_list_lock;
 	struct work_struct work;
 	struct sk_buff  *ooo_last_skb;
 	struct rb_root  out_of_order_queue;
@@ -395,6 +407,9 @@ DECLARE_PER_CPU(struct mptcp_delegated_action, mptcp_delegated_actions);
 /* MPTCP subflow context */
 struct mptcp_subflow_context {
 	struct	list_head node;/* conn_list of subflows */
+
+	char	reset_start[0];
+
 	unsigned long avg_pacing_rate; /* protected by msk socket lock */
 	u64	local_key;
 	u64	remote_key;
@@ -423,6 +438,7 @@ struct mptcp_subflow_context {
 		backup : 1,
 		send_mp_prio : 1,
 		send_mp_fail : 1,
+		send_fastclose : 1,
 		rx_eof : 1,
 		can_ack : 1,        /* only after processing the remote a key */
 		disposable : 1,	    /* ctx can be free at ulp release time */
@@ -441,6 +457,9 @@ struct mptcp_subflow_context {
 	u8	stale_count;
 
 	long	delegated_status;
+
+	char	reset_end[0];
+
 	struct	list_head delegated_node;   /* link into delegated_action, protected by local BH */
 
 	u32	setsockopt_seq;
@@ -472,6 +491,13 @@ mptcp_subflow_tcp_sock(const struct mptcp_subflow_context *subflow)
 	return subflow->tcp_sock;
 }
 
+static inline void
+mptcp_subflow_ctx_reset(struct mptcp_subflow_context *subflow)
+{
+	memset(subflow->reset_start, 0, subflow->reset_end - subflow->reset_start);
+	subflow->request_mptcp = 1;
+}
+
 static inline u64
 mptcp_subflow_get_map_offset(const struct mptcp_subflow_context *subflow)
 {
@@ -486,15 +512,6 @@ mptcp_subflow_get_mapped_dsn(const struct mptcp_subflow_context *subflow)
 	return subflow->map_seq + mptcp_subflow_get_map_offset(subflow);
 }
 
-static inline void mptcp_add_pending_subflow(struct mptcp_sock *msk,
-					     struct mptcp_subflow_context *subflow)
-{
-	sock_hold(mptcp_subflow_tcp_sock(subflow));
-	spin_lock_bh(&msk->join_list_lock);
-	list_add_tail(&subflow->node, &msk->join_list);
-	spin_unlock_bh(&msk->join_list_lock);
-}
-
 void mptcp_subflow_process_delegated(struct sock *ssk);
 
 static inline void mptcp_subflow_delegate(struct mptcp_subflow_context *subflow, int action)
@@ -659,7 +676,6 @@ void __mptcp_data_acked(struct sock *sk);
 void __mptcp_error_report(struct sock *sk);
 void mptcp_subflow_eof(struct sock *sk);
 bool mptcp_update_rcv_data_fin(struct mptcp_sock *msk, u64 data_fin_seq, bool use_64bit);
-void __mptcp_flush_join_list(struct mptcp_sock *msk);
 static inline bool mptcp_data_fin_enabled(const struct mptcp_sock *msk)
 {
 	return READ_ONCE(msk->snd_data_fin_enable) &&
@@ -712,6 +728,7 @@ void mptcp_crypto_hmac_sha(u64 key1, u64 key2, u8 *msg, int len, void *hmac);
 
 void __init mptcp_pm_init(void);
 void mptcp_pm_data_init(struct mptcp_sock *msk);
+void mptcp_pm_data_reset(struct mptcp_sock *msk);
 void mptcp_pm_subflow_chk_stale(const struct mptcp_sock *msk, struct sock *ssk);
 void mptcp_pm_nl_subflow_chk_stale(const struct mptcp_sock *msk, struct sock *ssk);
 void mptcp_pm_new_connection(struct mptcp_sock *msk, const struct sock *ssk, int server_side);
@@ -719,7 +736,9 @@ void mptcp_pm_fully_established(struct mptcp_sock *msk, const struct sock *ssk,
 bool mptcp_pm_allow_new_subflow(struct mptcp_sock *msk);
 void mptcp_pm_connection_closed(struct mptcp_sock *msk);
 void mptcp_pm_subflow_established(struct mptcp_sock *msk);
-void mptcp_pm_subflow_closed(struct mptcp_sock *msk, u8 id);
+bool mptcp_pm_nl_check_work_pending(struct mptcp_sock *msk);
+void mptcp_pm_subflow_check_next(struct mptcp_sock *msk, const struct sock *ssk,
+				 const struct mptcp_subflow_context *subflow);
 void mptcp_pm_add_addr_received(struct mptcp_sock *msk,
 				const struct mptcp_addr_info *addr);
 void mptcp_pm_add_addr_echoed(struct mptcp_sock *msk,
@@ -816,7 +835,7 @@ unsigned int mptcp_pm_get_subflows_max(struct mptcp_sock *msk);
 unsigned int mptcp_pm_get_local_addr_max(struct mptcp_sock *msk);
 
 void mptcp_sockopt_sync(struct mptcp_sock *msk, struct sock *ssk);
-void mptcp_sockopt_sync_all(struct mptcp_sock *msk);
+void mptcp_sockopt_sync_locked(struct mptcp_sock *msk, struct sock *ssk);
 
 static inline struct mptcp_ext *mptcp_get_ext(const struct sk_buff *skb)
 {

net/mptcp/sockopt.c

@@ -1285,27 +1285,15 @@ void mptcp_sockopt_sync(struct mptcp_sock *msk, struct sock *ssk)
 	}
 }
 
-void mptcp_sockopt_sync_all(struct mptcp_sock *msk)
+void mptcp_sockopt_sync_locked(struct mptcp_sock *msk, struct sock *ssk)
 {
-	struct mptcp_subflow_context *subflow;
-	struct sock *sk = (struct sock *)msk;
-	u32 seq;
-
-	seq = sockopt_seq_reset(sk);
+	struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(ssk);
 
-	mptcp_for_each_subflow(msk, subflow) {
-		struct sock *ssk = mptcp_subflow_tcp_sock(subflow);
-		u32 sseq = READ_ONCE(subflow->setsockopt_seq);
+	msk_owned_by_me(msk);
 
-		if (sseq != msk->setsockopt_seq) {
-			__mptcp_sockopt_sync(msk, ssk);
-			WRITE_ONCE(subflow->setsockopt_seq, seq);
-		} else if (sseq != seq) {
-			WRITE_ONCE(subflow->setsockopt_seq, seq);
-		}
+	if (READ_ONCE(subflow->setsockopt_seq) != msk->setsockopt_seq) {
+		sync_socket_options(msk, ssk);
 
-		cond_resched();
+		subflow->setsockopt_seq = msk->setsockopt_seq;
 	}
-
-	msk->setsockopt_seq = seq;
 }

net/mptcp/subflow.c

@@ -388,7 +388,7 @@ static void mptcp_set_connected(struct sock *sk)
 	if (!sock_owned_by_user(sk))
 		__mptcp_set_connected(sk);
 	else
-		set_bit(MPTCP_CONNECTED, &mptcp_sk(sk)->flags);
+		__set_bit(MPTCP_CONNECTED, &mptcp_sk(sk)->cb_flags);
 	mptcp_data_unlock(sk);
 }
 
@@ -1274,7 +1274,7 @@ static void subflow_error_report(struct sock *ssk)
 	if (!sock_owned_by_user(sk))
 		__mptcp_error_report(sk);
 	else
-		set_bit(MPTCP_ERROR_REPORT,  &mptcp_sk(sk)->flags);
+		__set_bit(MPTCP_ERROR_REPORT,  &mptcp_sk(sk)->cb_flags);
 	mptcp_data_unlock(sk);
 }
 
@@ -1293,7 +1293,6 @@ static void subflow_data_ready(struct sock *sk)
 		if (reqsk_queue_empty(&inet_csk(sk)->icsk_accept_queue))
 			return;
 
-		set_bit(MPTCP_DATA_READY, &msk->flags);
 		parent->sk_data_ready(parent);
 		return;
 	}
@@ -1442,7 +1441,8 @@ int __mptcp_subflow_connect(struct sock *sk, const struct mptcp_addr_info *loc,
 	subflow->request_bkup = !!(flags & MPTCP_PM_ADDR_FLAG_BACKUP);
 	mptcp_info2sockaddr(remote, &addr, ssk->sk_family);
 
-	mptcp_add_pending_subflow(msk, subflow);
+	sock_hold(ssk);
+	list_add_tail(&subflow->node, &msk->conn_list);
 	err = kernel_connect(sf, (struct sockaddr *)&addr, addrlen, O_NONBLOCK);
 	if (err && err != -EINPROGRESS)
 		goto failed_unlink;
@@ -1453,9 +1453,7 @@ int __mptcp_subflow_connect(struct sock *sk, const struct mptcp_addr_info *loc,
 	return err;
 
 failed_unlink:
-	spin_lock_bh(&msk->join_list_lock);
 	list_del(&subflow->node);
-	spin_unlock_bh(&msk->join_list_lock);
 	sock_put(mptcp_subflow_tcp_sock(subflow));
 
 failed:

net/mptcp/token.c

@@ -384,6 +384,7 @@ void mptcp_token_destroy(struct mptcp_sock *msk)
 		bucket->chain_len--;
 	}
 	spin_unlock_bh(&bucket->lock);
+	WRITE_ONCE(msk->token, 0);
 }
 
 void __init mptcp_token_init(void)

tools/testing/selftests/net/mptcp/config

@@ -17,4 +17,5 @@ CONFIG_NFT_TPROXY=m
 CONFIG_NFT_SOCKET=m
 CONFIG_IP_ADVANCED_ROUTER=y
 CONFIG_IP_MULTIPLE_TABLES=y
+CONFIG_IP_NF_TARGET_REJECT=m
 CONFIG_IPV6_MULTIPLE_TABLES=y

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

@@ -16,6 +16,7 @@
 #include <unistd.h>
 #include <time.h>
 
+#include <sys/ioctl.h>
 #include <sys/poll.h>
 #include <sys/sendfile.h>
 #include <sys/stat.h>
@@ -28,6 +29,7 @@
 
 #include <linux/tcp.h>
 #include <linux/time_types.h>
+#include <linux/sockios.h>
 
 extern int optind;
 
@@ -68,6 +70,8 @@ static unsigned int cfg_time;
 static unsigned int cfg_do_w;
 static int cfg_wait;
 static uint32_t cfg_mark;
+static char *cfg_input;
+static int cfg_repeat = 1;
 
 struct cfg_cmsg_types {
 	unsigned int cmsg_enabled:1;
@@ -91,22 +95,31 @@ static struct cfg_sockopt_types cfg_sockopt_types;
 
 static void die_usage(void)
 {
-	fprintf(stderr, "Usage: mptcp_connect [-6] [-u] [-s MPTCP|TCP] [-p port] [-m mode]"
-		"[-l] [-w sec] [-t num] [-T num] connect_address\n");
+	fprintf(stderr, "Usage: mptcp_connect [-6] [-c cmsg] [-i file] [-I num] [-j] [-l] "
+		"[-m mode] [-M mark] [-o option] [-p port] [-P mode] [-j] [-l] [-r num] "
+		"[-s MPTCP|TCP] [-S num] [-r num] [-t num] [-T num] [-u] [-w sec] connect_address\n");
 	fprintf(stderr, "\t-6 use ipv6\n");
-	fprintf(stderr, "\t-t num -- set poll timeout to num\n");
-	fprintf(stderr, "\t-T num -- set expected runtime to num ms\n");
-	fprintf(stderr, "\t-S num -- set SO_SNDBUF to num\n");
-	fprintf(stderr, "\t-R num -- set SO_RCVBUF to num\n");
-	fprintf(stderr, "\t-p num -- use port num\n");
-	fprintf(stderr, "\t-s [MPTCP|TCP] -- use mptcp(default) or tcp sockets\n");
+	fprintf(stderr, "\t-c cmsg -- test cmsg type <cmsg>\n");
+	fprintf(stderr, "\t-i file -- read the data to send from the given file instead of stdin");
+	fprintf(stderr, "\t-I num -- repeat the transfer 'num' times. In listen mode accepts num "
+		"incoming connections, in client mode, disconnect and reconnect to the server\n");
+	fprintf(stderr, "\t-j     -- add additional sleep at connection start and tear down "
+		"-- for MPJ tests\n");
+	fprintf(stderr, "\t-l     -- listens mode, accepts incoming connection\n");
 	fprintf(stderr, "\t-m [poll|mmap|sendfile] -- use poll(default)/mmap+write/sendfile\n");
 	fprintf(stderr, "\t-M mark -- set socket packet mark\n");
-	fprintf(stderr, "\t-w num -- wait num sec before closing the socket\n");
-	fprintf(stderr, "\t-c cmsg -- test cmsg type <cmsg>\n");
 	fprintf(stderr, "\t-o option -- test sockopt <option>\n");
+	fprintf(stderr, "\t-p num -- use port num\n");
 	fprintf(stderr,
 		"\t-P [saveWithPeek|saveAfterPeek] -- save data with/after MSG_PEEK form tcp socket\n");
+	fprintf(stderr, "\t-t num -- set poll timeout to num\n");
+	fprintf(stderr, "\t-T num -- set expected runtime to num ms\n");
+	fprintf(stderr, "\t-r num -- enable slow mode, limiting each write to num bytes "
+		"-- for remove addr tests\n");
+	fprintf(stderr, "\t-R num -- set SO_RCVBUF to num\n");
+	fprintf(stderr, "\t-s [MPTCP|TCP] -- use mptcp(default) or tcp sockets\n");
+	fprintf(stderr, "\t-S num -- set SO_SNDBUF to num\n");
+	fprintf(stderr, "\t-w num -- wait num sec before closing the socket\n");
 	exit(1);
 }
 
@@ -310,7 +323,8 @@ static int sock_listen_mptcp(const char * const listenaddr,
 }
 
 static int sock_connect_mptcp(const char * const remoteaddr,
-			      const char * const port, int proto)
+			      const char * const port, int proto,
+			      struct addrinfo **peer)
 {
 	struct addrinfo hints = {
 		.ai_protocol = IPPROTO_TCP,
@@ -334,8 +348,10 @@ 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)
+		if (connect(sock, a->ai_addr, a->ai_addrlen) == 0) {
+			*peer = a;
 			break; /* success */
+		}
 
 		perror("connect()");
 		close(sock);
@@ -524,14 +540,17 @@ static ssize_t do_rnd_read(const int fd, char *buf, const size_t len)
 	return ret;
 }
 
-static void set_nonblock(int fd)
+static void set_nonblock(int fd, bool nonblock)
 {
 	int flags = fcntl(fd, F_GETFL);
 
 	if (flags == -1)
 		return;
 
-	fcntl(fd, F_SETFL, flags | O_NONBLOCK);
+	if (nonblock)
+		fcntl(fd, F_SETFL, flags | O_NONBLOCK);
+	else
+		fcntl(fd, F_SETFL, flags & ~O_NONBLOCK);
 }
 
 static int copyfd_io_poll(int infd, int peerfd, int outfd, bool *in_closed_after_out)
@@ -543,7 +562,7 @@ static int copyfd_io_poll(int infd, int peerfd, int outfd, bool *in_closed_after
 	unsigned int woff = 0, wlen = 0;
 	char wbuf[8192];
 
-	set_nonblock(peerfd);
+	set_nonblock(peerfd, true);
 
 	for (;;) {
 		char rbuf[8192];
@@ -638,7 +657,6 @@ static int copyfd_io_poll(int infd, int peerfd, int outfd, bool *in_closed_after
 	if (cfg_remove)
 		usleep(cfg_wait);
 
-	close(peerfd);
 	return 0;
 }
 
@@ -780,7 +798,7 @@ static int copyfd_io_sendfile(int infd, int peerfd, int outfd,
 	return err;
 }
 
-static int copyfd_io(int infd, int peerfd, int outfd)
+static int copyfd_io(int infd, int peerfd, int outfd, bool close_peerfd)
 {
 	bool in_closed_after_out = false;
 	struct timespec start, end;
@@ -819,6 +837,9 @@ static int copyfd_io(int infd, int peerfd, int outfd)
 	if (ret)
 		return ret;
 
+	if (close_peerfd)
+		close(peerfd);
+
 	if (cfg_time) {
 		unsigned int delta_ms;
 
@@ -930,7 +951,7 @@ static void maybe_close(int fd)
 {
 	unsigned int r = rand();
 
-	if (!(cfg_join || cfg_remove) && (r & 1))
+	if (!(cfg_join || cfg_remove || cfg_repeat > 1) && (r & 1))
 		close(fd);
 }
 
@@ -940,7 +961,9 @@ int main_loop_s(int listensock)
 	struct pollfd polls;
 	socklen_t salen;
 	int remotesock;
+	int fd = 0;
 
+again:
 	polls.fd = listensock;
 	polls.events = POLLIN;
 
@@ -961,14 +984,27 @@ int main_loop_s(int listensock)
 		check_sockaddr(pf, &ss, salen);
 		check_getpeername(remotesock, &ss, salen);
 
+		if (cfg_input) {
+			fd = open(cfg_input, O_RDONLY);
+			if (fd < 0)
+				xerror("can't open %s: %d", cfg_input, errno);
+		}
+
 		SOCK_TEST_TCPULP(remotesock, 0);
 
-		return copyfd_io(0, remotesock, 1);
+		copyfd_io(fd, remotesock, 1, true);
+	} else {
+		perror("accept");
+		return 1;
 	}
 
-	perror("accept");
+	if (--cfg_repeat > 0) {
+		if (cfg_input)
+			close(fd);
+		goto again;
+	}
 
-	return 1;
+	return 0;
 }
 
 static void init_rng(void)
@@ -1057,15 +1093,47 @@ static void parse_setsock_options(const char *name)
 	exit(1);
 }
 
+void xdisconnect(int fd, int addrlen)
+{
+	struct sockaddr_storage empty;
+	int msec_sleep = 10;
+	int queued = 1;
+	int i;
+
+	shutdown(fd, SHUT_WR);
+
+	/* while until the pending data is completely flushed, the later
+	 * disconnect will bypass/ignore/drop any pending data.
+	 */
+	for (i = 0; ; i += msec_sleep) {
+		if (ioctl(fd, SIOCOUTQ, &queued) < 0)
+			xerror("can't query out socket queue: %d", errno);
+
+		if (!queued)
+			break;
+
+		if (i > poll_timeout)
+			xerror("timeout while waiting for spool to complete");
+		usleep(msec_sleep * 1000);
+	}
+
+	memset(&empty, 0, sizeof(empty));
+	empty.ss_family = AF_UNSPEC;
+	if (connect(fd, (struct sockaddr *)&empty, addrlen) < 0)
+		xerror("can't disconnect: %d", errno);
+}
+
 int main_loop(void)
 {
-	int fd;
+	int fd, ret, fd_in = 0;
+	struct addrinfo *peer;
 
 	/* listener is ready. */
-	fd = sock_connect_mptcp(cfg_host, cfg_port, cfg_sock_proto);
+	fd = sock_connect_mptcp(cfg_host, cfg_port, cfg_sock_proto, &peer);
 	if (fd < 0)
 		return 2;
 
+again:
 	check_getpeername_connect(fd);
 
 	SOCK_TEST_TCPULP(fd, cfg_sock_proto);
@@ -1077,7 +1145,31 @@ int main_loop(void)
 	if (cfg_cmsg_types.cmsg_enabled)
 		apply_cmsg_types(fd, &cfg_cmsg_types);
 
-	return copyfd_io(0, fd, 1);
+	if (cfg_input) {
+		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, cfg_repeat == 1);
+	if (ret)
+		return ret;
+
+	if (--cfg_repeat > 0) {
+		xdisconnect(fd, peer->ai_addrlen);
+
+		/* the socket could be unblocking at this point, we need the
+		 * connect to be blocking
+		 */
+		set_nonblock(fd, false);
+		if (connect(fd, peer->ai_addr, peer->ai_addrlen))
+			xerror("can't reconnect: %d", errno);
+		if (cfg_input)
+			close(fd_in);
+		goto again;
+	}
+	return 0;
 }
 
 int parse_proto(const char *proto)
@@ -1162,7 +1254,7 @@ static void parse_opts(int argc, char **argv)
 {
 	int c;
 
-	while ((c = getopt(argc, argv, "6jr:lp:s:ht:T:m:S:R:w:M:P:c:o:")) != -1) {
+	while ((c = getopt(argc, argv, "6c:hi:I:jlm:M:o:p:P:r:R:s:S:t:T:w:")) != -1) {
 		switch (c) {
 		case 'j':
 			cfg_join = true;
@@ -1176,6 +1268,12 @@ static void parse_opts(int argc, char **argv)
 			if (cfg_do_w <= 0)
 				cfg_do_w = 50;
 			break;
+		case 'i':
+			cfg_input = optarg;
+			break;
+		case 'I':
+			cfg_repeat = atoi(optarg);
+			break;
 		case 'l':
 			listen_mode = true;
 			break;

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

@@ -7,6 +7,7 @@ optstring="S:R:d:e:l:r:h4cm:f:tC"
 ret=0
 sin=""
 sout=""
+cin_disconnect=""
 cin=""
 cout=""
 ksft_skip=4
@@ -24,6 +25,7 @@ options_log=true
 do_tcp=0
 checksum=false
 filesize=0
+connect_per_transfer=1
 
 if [ $tc_loss -eq 100 ];then
 	tc_loss=1%
@@ -127,6 +129,7 @@ TEST_COUNT=0
 
 cleanup()
 {
+	rm -f "$cin_disconnect" "$cout_disconnect"
 	rm -f "$cin" "$cout"
 	rm -f "$sin" "$sout"
 	rm -f "$capout"
@@ -149,6 +152,8 @@ sout=$(mktemp)
 cin=$(mktemp)
 cout=$(mktemp)
 capout=$(mktemp)
+cin_disconnect="$cin".disconnect
+cout_disconnect="$cout".disconnect
 trap cleanup EXIT
 
 for i in "$ns1" "$ns2" "$ns3" "$ns4";do
@@ -500,8 +505,8 @@ do_transfer()
 	cookies=${cookies##*=}
 
 	if [ ${cl_proto} = "MPTCP" ] && [ ${srv_proto} = "MPTCP" ]; then
-		expect_synrx=$((stat_synrx_last_l+1))
-		expect_ackrx=$((stat_ackrx_last_l+1))
+		expect_synrx=$((stat_synrx_last_l+$connect_per_transfer))
+		expect_ackrx=$((stat_ackrx_last_l+$connect_per_transfer))
 	fi
 
 	if [ ${stat_synrx_now_l} -lt ${expect_synrx} ]; then
@@ -738,6 +743,33 @@ run_tests_peekmode()
 	run_tests_lo "$ns1" "$ns1" dead:beef:1::1 1 "-P ${peekmode}"
 }
 
+run_tests_disconnect()
+{
+	local peekmode="$1"
+	local old_cin=$cin
+	local old_sin=$sin
+
+	cat $cin $cin $cin > "$cin".disconnect
+
+	# force do_transfer to cope with the multiple tranmissions
+	sin="$cin.disconnect"
+	sin_disconnect=$old_sin
+	cin="$cin.disconnect"
+	cin_disconnect="$old_cin"
+	connect_per_transfer=3
+
+	echo "INFO: disconnect"
+	run_tests_lo "$ns1" "$ns1" 10.0.1.1 1 "-I 3 -i $old_cin"
+	run_tests_lo "$ns1" "$ns1" dead:beef:1::1 1 "-I 3 -i $old_cin"
+
+	# restore previous status
+	cout=$old_cout
+	cout_disconnect="$cout".disconnect
+	cin=$old_cin
+	cin_disconnect="$cin".disconnect
+	connect_per_transfer=1
+}
+
 display_time()
 {
 	time_end=$(date +%s)
@@ -853,6 +885,9 @@ stop_if_error "Tests with peek mode 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"
+stop_if_error "Tests with tproxy have failed"
+
+run_tests_disconnect
 
 display_time
 exit $ret

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

@@ -937,6 +937,22 @@ chk_link_usage()
 	fi
 }
 
+wait_for_tw()
+{
+	local timeout_ms=$((timeout_poll * 1000))
+	local time=0
+	local ns=$1
+
+	while [ $time -lt $timeout_ms ]; do
+		local cnt=$(ip netns exec $ns ss -t state time-wait |wc -l)
+
+		[ "$cnt" = 1 ] && return 1
+		time=$((time + 100))
+		sleep 0.1
+	done
+	return 1
+}
+
 subflows_tests()
 {
 	reset
@@ -994,6 +1010,61 @@ subflows_tests()
 	chk_join_nr "single subflow, dev" 1 1 1
 }
 
+subflows_error_tests()
+{
+	# If a single subflow is configured, and matches the MPC src
+	# address, no additional subflow should be created
+	reset
+	ip netns exec $ns1 ./pm_nl_ctl limits 0 1
+	ip netns exec $ns2 ./pm_nl_ctl limits 0 1
+	ip netns exec $ns2 ./pm_nl_ctl add 10.0.1.2 flags subflow
+	run_tests $ns1 $ns2 10.0.1.1 0 0 0 slow
+	chk_join_nr "no MPC reuse with single endpoint" 0 0 0
+
+	# multiple subflows, with subflow creation error
+	reset
+	ip netns exec $ns1 ./pm_nl_ctl limits 0 2
+	ip netns exec $ns2 ./pm_nl_ctl limits 0 2
+	ip netns exec $ns2 ./pm_nl_ctl add 10.0.3.2 flags subflow
+	ip netns exec $ns2 ./pm_nl_ctl add 10.0.2.2 flags subflow
+	ip netns exec $ns1 iptables -A INPUT -s 10.0.3.2 -p tcp -j REJECT
+	run_tests $ns1 $ns2 10.0.1.1 0 0 0 slow
+	chk_join_nr "multi subflows, with failing subflow" 1 1 1
+
+	# multiple subflows, with subflow timeout on MPJ
+	reset
+	ip netns exec $ns1 ./pm_nl_ctl limits 0 2
+	ip netns exec $ns2 ./pm_nl_ctl limits 0 2
+	ip netns exec $ns2 ./pm_nl_ctl add 10.0.3.2 flags subflow
+	ip netns exec $ns2 ./pm_nl_ctl add 10.0.2.2 flags subflow
+	ip netns exec $ns1 iptables -A INPUT -s 10.0.3.2 -p tcp -j DROP
+	run_tests $ns1 $ns2 10.0.1.1 0 0 0 slow
+	chk_join_nr "multi subflows, with subflow timeout" 1 1 1
+
+	# multiple subflows, check that the endpoint corresponding to
+	# closed subflow (due to reset) is not reused if additional
+	# subflows are added later
+	reset
+	ip netns exec $ns1 ./pm_nl_ctl limits 0 1
+	ip netns exec $ns2 ./pm_nl_ctl limits 0 1
+	ip netns exec $ns2 ./pm_nl_ctl add 10.0.3.2 flags subflow
+	ip netns exec $ns1 iptables -A INPUT -s 10.0.3.2 -p tcp -j REJECT
+	run_tests $ns1 $ns2 10.0.1.1 0 0 0 slow &
+
+	# updates in the child shell do not have any effect here, we
+	# need to bump the test counter for the above case
+	TEST_COUNT=$((TEST_COUNT+1))
+
+	# mpj subflow will be in TW after the reset
+	wait_for_tw $ns2
+	ip netns exec $ns2 ./pm_nl_ctl add 10.0.2.2 flags subflow
+	wait
+
+	# additional subflow could be created only if the PM select
+	# the later endpoint, skipping the already used one
+	chk_join_nr "multi subflows, fair usage on close" 1 1 1
+}
+
 signal_address_tests()
 {
 	# add_address, unused
@@ -1071,7 +1142,10 @@ signal_address_tests()
 	ip netns exec $ns2 ./pm_nl_ctl add 10.0.3.2 flags signal
 	ip netns exec $ns2 ./pm_nl_ctl add 10.0.4.2 flags signal
 	run_tests $ns1 $ns2 10.0.1.1
-	chk_add_nr 4 4
+
+	# the server will not signal the address terminating
+	# the MPC subflow
+	chk_add_nr 3 3
 }
 
 link_failure_tests()
@@ -1802,6 +1876,7 @@ fullmesh_tests()
 all_tests()
 {
 	subflows_tests
+	subflows_error_tests
 	signal_address_tests
 	link_failure_tests
 	add_addr_timeout_tests
@@ -1821,6 +1896,7 @@ usage()
 {
 	echo "mptcp_join usage:"
 	echo "  -f subflows_tests"
+	echo "  -e subflows_error_tests"
 	echo "  -s signal_address_tests"
 	echo "  -l link_failure_tests"
 	echo "  -t add_addr_timeout_tests"
@@ -1869,11 +1945,14 @@ if [ $do_all_tests -eq 1 ]; then
 	exit $ret
 fi
 
-while getopts 'fsltra64bpkdmchCS' opt; do
+while getopts 'fesltra64bpkdmchCS' opt; do
 	case $opt in
 		f)
 			subflows_tests
 			;;
+		e)
+			subflows_error_tests
+			;;
 		s)
 			signal_address_tests
 			;;