1 files changed, 130 insertions, 85 deletions

diff --git a/net/mptcp/pm_netlink.c b/net/mptcp/pm_netlink.c
index f523051f5aef..75af1f701e1d 100644
--- a/net/mptcp/pm_netlink.c
+++ b/net/mptcp/pm_netlink.c
@@ -38,9 +38,6 @@ struct mptcp_pm_add_entry {
 	u8			retrans_times;
 };
 
-#define MAX_ADDR_ID		255
-#define BITMAP_SZ DIV_ROUND_UP(MAX_ADDR_ID + 1, BITS_PER_LONG)
-
 struct pm_nl_pernet {
 	/* protects pernet updates */
 	spinlock_t		lock;
@@ -52,14 +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;
 
@@ -168,11 +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 &&
@@ -183,23 +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
@@ -208,12 +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;
@@ -255,12 +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 *
@@ -429,6 +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 };
@@ -436,22 +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++;
 			}
 		}
 	}
@@ -459,6 +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;
@@ -474,6 +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,
@@ -481,47 +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)
@@ -551,7 +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;
@@ -640,7 +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);
@@ -710,6 +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;
@@ -726,18 +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--;
+		}
 	}
 }
 
@@ -758,6 +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);
@@ -803,7 +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;
@@ -823,16 +875,15 @@ 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 || entry->addr.id > MAX_ADDR_ID) &&
-		    pernet->next_id != 1) {
+		if (!entry->addr.id && pernet->next_id != 1) {
 			pernet->next_id = 1;
 			goto find_next;
 		}
 	}
 
-	if (!entry->addr.id || entry->addr.id > MAX_ADDR_ID)
+	if (!entry->addr.id)
 		goto out;
 
 	__set_bit(entry->addr.id, pernet->id_bitmap);
@@ -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)
@@ -1705,22 +1744,28 @@ next:
 
 static int mptcp_nl_cmd_set_flags(struct sk_buff *skb, struct genl_info *info)
 {
+	struct mptcp_pm_addr_entry addr = { .addr = { .family = AF_UNSPEC }, }, *entry;
 	struct nlattr *attr = info->attrs[MPTCP_PM_ATTR_ADDR];
 	struct pm_nl_pernet *pernet = genl_info_pm_nl(info);
-	struct mptcp_pm_addr_entry addr, *entry;
 	struct net *net = sock_net(skb->sk);
-	u8 bkup = 0;
+	u8 bkup = 0, lookup_by_id = 0;
 	int ret;
 
-	ret = mptcp_pm_parse_addr(attr, info, true, &addr);
+	ret = mptcp_pm_parse_addr(attr, info, false, &addr);
 	if (ret < 0)
 		return ret;
 
 	if (addr.flags & MPTCP_PM_ADDR_FLAG_BACKUP)
 		bkup = 1;
+	if (addr.addr.family == AF_UNSPEC) {
+		lookup_by_id = 1;
+		if (!addr.addr.id)
+			return -EOPNOTSUPP;
+	}
 
 	list_for_each_entry(entry, &pernet->local_addr_list, list) {
-		if (addresses_equal(&entry->addr, &addr.addr, true)) {
+		if ((!lookup_by_id && addresses_equal(&entry->addr, &addr.addr, true)) ||
+		    (lookup_by_id && entry->addr.id == addr.addr.id)) {
 			mptcp_nl_addr_backup(net, &entry->addr, bkup);
 
 			if (bkup)