Merge branch 'mptcp-next'

Mat Martineau says:

====================
mptcp: Miscellaneous changes for 5.19

Four separate groups of patches here:

Patch 1 optimizes flag checking when releasing mptcp socket locks.

Patches 2 and 3 update the packet scheduler when subflow priorities
change.

Patch 4 adds some pernet helper functions for MPTCP.

Patches 5-8 add diag support for MPTCP listeners, including a selftest.

====================

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

7 files changed, 193 insertions, 37 deletions

diff --git a/net/mptcp/mptcp_diag.c b/net/mptcp/mptcp_diag.c
index f44125dd6697..dbb6d876a203 100644
--- a/net/mptcp/mptcp_diag.c
+++ b/net/mptcp/mptcp_diag.c
@@ -66,20 +66,103 @@ out_nosk:
 	return err;
 }
 
+struct mptcp_diag_ctx {
+	long s_slot;
+	long s_num;
+	unsigned int l_slot;
+	unsigned int l_num;
+};
+
+static void mptcp_diag_dump_listeners(struct sk_buff *skb, struct netlink_callback *cb,
+				      const struct inet_diag_req_v2 *r,
+				      bool net_admin)
+{
+	struct inet_diag_dump_data *cb_data = cb->data;
+	struct mptcp_diag_ctx *diag_ctx = (void *)cb->ctx;
+	struct nlattr *bc = cb_data->inet_diag_nla_bc;
+	struct net *net = sock_net(skb->sk);
+	int i;
+
+	for (i = diag_ctx->l_slot; i < INET_LHTABLE_SIZE; i++) {
+		struct inet_listen_hashbucket *ilb;
+		struct hlist_nulls_node *node;
+		struct sock *sk;
+		int num = 0;
+
+		ilb = &tcp_hashinfo.listening_hash[i];
+
+		rcu_read_lock();
+		spin_lock(&ilb->lock);
+		sk_nulls_for_each(sk, node, &ilb->nulls_head) {
+			const struct mptcp_subflow_context *ctx = mptcp_subflow_ctx(sk);
+			struct inet_sock *inet = inet_sk(sk);
+			int ret;
+
+			if (num < diag_ctx->l_num)
+				goto next_listen;
+
+			if (!ctx || strcmp(inet_csk(sk)->icsk_ulp_ops->name, "mptcp"))
+				goto next_listen;
+
+			sk = ctx->conn;
+			if (!sk || !net_eq(sock_net(sk), net))
+				goto next_listen;
+
+			if (r->sdiag_family != AF_UNSPEC &&
+			    sk->sk_family != r->sdiag_family)
+				goto next_listen;
+
+			if (r->id.idiag_sport != inet->inet_sport &&
+			    r->id.idiag_sport)
+				goto next_listen;
+
+			if (!refcount_inc_not_zero(&sk->sk_refcnt))
+				goto next_listen;
+
+			ret = sk_diag_dump(sk, skb, cb, r, bc, net_admin);
+
+			sock_put(sk);
+
+			if (ret < 0) {
+				spin_unlock(&ilb->lock);
+				rcu_read_unlock();
+				diag_ctx->l_slot = i;
+				diag_ctx->l_num = num;
+				return;
+			}
+			diag_ctx->l_num = num + 1;
+			num = 0;
+next_listen:
+			++num;
+		}
+		spin_unlock(&ilb->lock);
+		rcu_read_unlock();
+
+		cond_resched();
+		diag_ctx->l_num = 0;
+	}
+
+	diag_ctx->l_num = 0;
+	diag_ctx->l_slot = i;
+}
+
 static void mptcp_diag_dump(struct sk_buff *skb, struct netlink_callback *cb,
 			    const struct inet_diag_req_v2 *r)
 {
 	bool net_admin = netlink_net_capable(cb->skb, CAP_NET_ADMIN);
+	struct mptcp_diag_ctx *diag_ctx = (void *)cb->ctx;
 	struct net *net = sock_net(skb->sk);
 	struct inet_diag_dump_data *cb_data;
 	struct mptcp_sock *msk;
 	struct nlattr *bc;
 
+	BUILD_BUG_ON(sizeof(cb->ctx) < sizeof(*diag_ctx));
+
 	cb_data = cb->data;
 	bc = cb_data->inet_diag_nla_bc;
 
-	while ((msk = mptcp_token_iter_next(net, &cb->args[0], &cb->args[1])) !=
-	       NULL) {
+	while ((msk = mptcp_token_iter_next(net, &diag_ctx->s_slot,
+					    &diag_ctx->s_num)) != NULL) {
 		struct inet_sock *inet = (struct inet_sock *)msk;
 		struct sock *sk = (struct sock *)msk;
 		int ret = 0;
@@ -101,11 +184,14 @@ next:
 		sock_put(sk);
 		if (ret < 0) {
 			/* will retry on the same position */
-			cb->args[1]--;
+			diag_ctx->s_num--;
 			break;
 		}
 		cond_resched();
 	}
+
+	if ((r->idiag_states & TCPF_LISTEN) && r->id.idiag_dport == 0)
+		mptcp_diag_dump_listeners(skb, cb, r, net_admin);
 }
 
 static void mptcp_diag_get_info(struct sock *sk, struct inet_diag_msg *r,
@@ -116,6 +202,19 @@ static void mptcp_diag_get_info(struct sock *sk, struct inet_diag_msg *r,
 
 	r->idiag_rqueue = sk_rmem_alloc_get(sk);
 	r->idiag_wqueue = sk_wmem_alloc_get(sk);
+
+	if (inet_sk_state_load(sk) == TCP_LISTEN) {
+		struct sock *lsk = READ_ONCE(msk->first);
+
+		if (lsk) {
+			/* override with settings from tcp listener,
+			 * so Send-Q will show accept queue.
+			 */
+			r->idiag_rqueue = READ_ONCE(lsk->sk_ack_backlog);
+			r->idiag_wqueue = READ_ONCE(lsk->sk_max_ack_backlog);
+		}
+	}
+
 	if (!info)
 		return;
 
diff --git a/net/mptcp/pm.c b/net/mptcp/pm.c
index 01809eef29b4..8aa0cdb7ad46 100644
--- a/net/mptcp/pm.c
+++ b/net/mptcp/pm.c
@@ -262,14 +262,25 @@ void mptcp_pm_rm_addr_received(struct mptcp_sock *msk,
 	spin_unlock_bh(&pm->lock);
 }
 
-void mptcp_pm_mp_prio_received(struct sock *sk, u8 bkup)
+void mptcp_pm_mp_prio_received(struct sock *ssk, u8 bkup)
 {
-	struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(sk);
+	struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(ssk);
+	struct sock *sk = subflow->conn;
+	struct mptcp_sock *msk;
 
 	pr_debug("subflow->backup=%d, bkup=%d\n", subflow->backup, bkup);
-	subflow->backup = bkup;
+	msk = mptcp_sk(sk);
+	if (subflow->backup != bkup) {
+		subflow->backup = bkup;
+		mptcp_data_lock(sk);
+		if (!sock_owned_by_user(sk))
+			msk->last_snd = NULL;
+		else
+			__set_bit(MPTCP_RESET_SCHEDULER,  &msk->cb_flags);
+		mptcp_data_unlock(sk);
+	}
 
-	mptcp_event(MPTCP_EVENT_SUB_PRIORITY, mptcp_sk(subflow->conn), sk, GFP_ATOMIC);
+	mptcp_event(MPTCP_EVENT_SUB_PRIORITY, msk, ssk, GFP_ATOMIC);
 }
 
 void mptcp_pm_mp_fail_received(struct sock *sk, u64 fail_seq)
diff --git a/net/mptcp/pm_netlink.c b/net/mptcp/pm_netlink.c
index b5e8de6f7507..c20261b612e9 100644
--- a/net/mptcp/pm_netlink.c
+++ b/net/mptcp/pm_netlink.c
@@ -55,6 +55,17 @@ struct pm_nl_pernet {
 #define MPTCP_PM_ADDR_MAX	8
 #define ADD_ADDR_RETRANS_MAX	3
 
+static struct pm_nl_pernet *pm_nl_get_pernet(const struct net *net)
+{
+	return net_generic(net, pm_nl_pernet_id);
+}
+
+static struct pm_nl_pernet *
+pm_nl_get_pernet_from_msk(const struct mptcp_sock *msk)
+{
+	return pm_nl_get_pernet(sock_net((struct sock *)msk));
+}
+
 static bool addresses_equal(const struct mptcp_addr_info *a,
 			    const struct mptcp_addr_info *b, bool use_port)
 {
@@ -206,43 +217,39 @@ select_signal_address(struct pm_nl_pernet *pernet, const struct mptcp_sock *msk)
 
 unsigned int mptcp_pm_get_add_addr_signal_max(const struct mptcp_sock *msk)
 {
-	const struct pm_nl_pernet *pernet;
+	const struct pm_nl_pernet *pernet = pm_nl_get_pernet_from_msk(msk);
 
-	pernet = net_generic(sock_net((const struct sock *)msk), pm_nl_pernet_id);
 	return READ_ONCE(pernet->add_addr_signal_max);
 }
 EXPORT_SYMBOL_GPL(mptcp_pm_get_add_addr_signal_max);
 
 unsigned int mptcp_pm_get_add_addr_accept_max(const struct mptcp_sock *msk)
 {
-	struct pm_nl_pernet *pernet;
+	struct pm_nl_pernet *pernet = pm_nl_get_pernet_from_msk(msk);
 
-	pernet = net_generic(sock_net((struct sock *)msk), pm_nl_pernet_id);
 	return READ_ONCE(pernet->add_addr_accept_max);
 }
 EXPORT_SYMBOL_GPL(mptcp_pm_get_add_addr_accept_max);
 
 unsigned int mptcp_pm_get_subflows_max(const struct mptcp_sock *msk)
 {
-	struct pm_nl_pernet *pernet;
+	struct pm_nl_pernet *pernet = pm_nl_get_pernet_from_msk(msk);
 
-	pernet = net_generic(sock_net((struct sock *)msk), pm_nl_pernet_id);
 	return READ_ONCE(pernet->subflows_max);
 }
 EXPORT_SYMBOL_GPL(mptcp_pm_get_subflows_max);
 
 unsigned int mptcp_pm_get_local_addr_max(const struct mptcp_sock *msk)
 {
-	struct pm_nl_pernet *pernet;
+	struct pm_nl_pernet *pernet = pm_nl_get_pernet_from_msk(msk);
 
-	pernet = net_generic(sock_net((struct sock *)msk), pm_nl_pernet_id);
 	return READ_ONCE(pernet->local_addr_max);
 }
 EXPORT_SYMBOL_GPL(mptcp_pm_get_local_addr_max);
 
 bool mptcp_pm_nl_check_work_pending(struct mptcp_sock *msk)
 {
-	struct pm_nl_pernet *pernet = net_generic(sock_net((struct sock *)msk), pm_nl_pernet_id);
+	struct pm_nl_pernet *pernet = pm_nl_get_pernet_from_msk(msk);
 
 	if (msk->pm.subflows == mptcp_pm_get_subflows_max(msk) ||
 	    (find_next_and_bit(pernet->id_bitmap, msk->pm.id_avail_bitmap,
@@ -508,7 +515,7 @@ static void mptcp_pm_create_subflow_or_signal_addr(struct mptcp_sock *msk)
 	struct pm_nl_pernet *pernet;
 	unsigned int subflows_max;
 
-	pernet = net_generic(sock_net(sk), pm_nl_pernet_id);
+	pernet = pm_nl_get_pernet(sock_net(sk));
 
 	add_addr_signal_max = mptcp_pm_get_add_addr_signal_max(msk);
 	local_addr_max = mptcp_pm_get_local_addr_max(msk);
@@ -604,7 +611,7 @@ static unsigned int fill_local_addresses_vec(struct mptcp_sock *msk,
 	unsigned int subflows_max;
 	int i = 0;
 
-	pernet = net_generic(sock_net(sk), pm_nl_pernet_id);
+	pernet = pm_nl_get_pernet_from_msk(msk);
 	subflows_max = mptcp_pm_get_subflows_max(msk);
 
 	rcu_read_lock();
@@ -727,6 +734,8 @@ static int mptcp_pm_nl_mp_prio_send_ack(struct mptcp_sock *msk,
 		if (!addresses_equal(&local, addr, addr->port))
 			continue;
 
+		if (subflow->backup != bkup)
+			msk->last_snd = NULL;
 		subflow->backup = bkup;
 		subflow->send_mp_prio = 1;
 		subflow->request_bkup = bkup;
@@ -1021,7 +1030,7 @@ int mptcp_pm_nl_get_local_id(struct mptcp_sock *msk, struct sock_common *skc)
 	if (addresses_equal(&msk_local, &skc_local, false))
 		return 0;
 
-	pernet = net_generic(sock_net((struct sock *)msk), pm_nl_pernet_id);
+	pernet = pm_nl_get_pernet_from_msk(msk);
 
 	rcu_read_lock();
 	list_for_each_entry_rcu(entry, &pernet->local_addr_list, list) {
@@ -1212,7 +1221,7 @@ skip_family:
 
 static struct pm_nl_pernet *genl_info_pm_nl(struct genl_info *info)
 {
-	return net_generic(genl_info_net(info), pm_nl_pernet_id);
+	return pm_nl_get_pernet(genl_info_net(info));
 }
 
 static int mptcp_nl_add_subflow_or_signal_addr(struct net *net)
@@ -1306,7 +1315,7 @@ int mptcp_pm_get_flags_and_ifindex_by_id(struct net *net, unsigned int id,
 
 	if (id) {
 		rcu_read_lock();
-		entry = __lookup_addr_by_id(net_generic(net, pm_nl_pernet_id), id);
+		entry = __lookup_addr_by_id(pm_nl_get_pernet(net), id);
 		if (entry) {
 			*flags = entry->flags;
 			*ifindex = entry->ifindex;
@@ -1653,7 +1662,7 @@ static int mptcp_nl_cmd_dump_addrs(struct sk_buff *msg,
 	void *hdr;
 	int i;
 
-	pernet = net_generic(net, pm_nl_pernet_id);
+	pernet = pm_nl_get_pernet(net);
 
 	spin_lock_bh(&pernet->lock);
 	for (i = id; i < MPTCP_PM_MAX_ADDR_ID + 1; i++) {
@@ -2165,7 +2174,7 @@ static struct genl_family mptcp_genl_family __ro_after_init = {
 
 static int __net_init pm_nl_init_net(struct net *net)
 {
-	struct pm_nl_pernet *pernet = net_generic(net, pm_nl_pernet_id);
+	struct pm_nl_pernet *pernet = pm_nl_get_pernet(net);
 
 	INIT_LIST_HEAD_RCU(&pernet->local_addr_list);
 
@@ -2187,7 +2196,7 @@ static void __net_exit pm_nl_exit_net(struct list_head *net_list)
 	struct net *net;
 
 	list_for_each_entry(net, net_list, exit_list) {
-		struct pm_nl_pernet *pernet = net_generic(net, pm_nl_pernet_id);
+		struct pm_nl_pernet *pernet = pm_nl_get_pernet(net);
 
 		/* net is removed from namespace list, can't race with
 		 * other modifiers, also netns core already waited for a
diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c
index 0cbea3b6d0a4..8f54293c1d88 100644
--- a/net/mptcp/protocol.c
+++ b/net/mptcp/protocol.c
@@ -3092,15 +3092,19 @@ static void mptcp_release_cb(struct sock *sk)
 		spin_lock_bh(&sk->sk_lock.slock);
 	}
 
-	/* be sure to set the current sk state before tacking actions
-	 * depending on sk_state
-	 */
-	if (__test_and_clear_bit(MPTCP_CONNECTED, &msk->cb_flags))
-		__mptcp_set_connected(sk);
 	if (__test_and_clear_bit(MPTCP_CLEAN_UNA, &msk->cb_flags))
 		__mptcp_clean_una_wakeup(sk);
-	if (__test_and_clear_bit(MPTCP_ERROR_REPORT, &msk->cb_flags))
-		__mptcp_error_report(sk);
+	if (unlikely(&msk->cb_flags)) {
+		/* be sure to set the current sk state before tacking actions
+		 * depending on sk_state, that is processing MPTCP_ERROR_REPORT
+		 */
+		if (__test_and_clear_bit(MPTCP_CONNECTED, &msk->cb_flags))
+			__mptcp_set_connected(sk);
+		if (__test_and_clear_bit(MPTCP_ERROR_REPORT, &msk->cb_flags))
+			__mptcp_error_report(sk);
+		if (__test_and_clear_bit(MPTCP_RESET_SCHEDULER, &msk->cb_flags))
+			msk->last_snd = NULL;
+	}
 
 	__mptcp_update_rmem(sk);
 }
diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h
index 3c1a3036550f..aca1fb56523f 100644
--- a/net/mptcp/protocol.h
+++ b/net/mptcp/protocol.h
@@ -124,6 +124,7 @@
 #define MPTCP_RETRANSMIT	4
 #define MPTCP_FLUSH_JOIN_LIST	5
 #define MPTCP_CONNECTED		6
+#define MPTCP_RESET_SCHEDULER	7
 
 static inline bool before64(__u64 seq1, __u64 seq2)
 {
diff --git a/net/mptcp/sockopt.c b/net/mptcp/sockopt.c
index f949d22f52bd..826b0c1dae98 100644
--- a/net/mptcp/sockopt.c
+++ b/net/mptcp/sockopt.c
@@ -853,15 +853,11 @@ out:
 
 void mptcp_diag_fill_info(struct mptcp_sock *msk, struct mptcp_info *info)
 {
-	struct sock *sk = &msk->sk.icsk_inet.sk;
 	u32 flags = 0;
-	bool slow;
 	u8 val;
 
 	memset(info, 0, sizeof(*info));
 
-	slow = lock_sock_fast(sk);
-
 	info->mptcpi_subflows = READ_ONCE(msk->pm.subflows);
 	info->mptcpi_add_addr_signal = READ_ONCE(msk->pm.add_addr_signaled);
 	info->mptcpi_add_addr_accepted = READ_ONCE(msk->pm.add_addr_accepted);
@@ -882,8 +878,6 @@ void mptcp_diag_fill_info(struct mptcp_sock *msk, struct mptcp_info *info)
 	info->mptcpi_snd_una = READ_ONCE(msk->snd_una);
 	info->mptcpi_rcv_nxt = READ_ONCE(msk->ack_seq);
 	info->mptcpi_csum_enabled = READ_ONCE(msk->csum_enabled);
-
-	unlock_sock_fast(sk, slow);
 }
 EXPORT_SYMBOL_GPL(mptcp_diag_fill_info);
 
diff --git a/tools/testing/selftests/net/mptcp/diag.sh b/tools/testing/selftests/net/mptcp/diag.sh
index ff821025d309..9dd43d7d957b 100755
--- a/tools/testing/selftests/net/mptcp/diag.sh
+++ b/tools/testing/selftests/net/mptcp/diag.sh
@@ -71,6 +71,43 @@ chk_msk_remote_key_nr()
 		__chk_nr "grep -c remote_key" $*
 }
 
+__chk_listen()
+{
+	local filter="$1"
+	local expected=$2
+
+	shift 2
+	msg=$*
+
+	nr=$(ss -N $ns -Ml "$filter" | grep -c LISTEN)
+	printf "%-50s" "$msg"
+
+	if [ $nr != $expected ]; then
+		echo "[ fail ] expected $expected found $nr"
+		ret=$test_cnt
+	else
+		echo "[  ok  ]"
+	fi
+}
+
+chk_msk_listen()
+{
+	lport=$1
+	local msg="check for listen socket"
+
+	# destination port search should always return empty list
+	__chk_listen "dport $lport" 0 "listen match for dport $lport"
+
+	# should return 'our' mptcp listen socket
+	__chk_listen "sport $lport" 1 "listen match for sport $lport"
+
+	__chk_listen "src inet:0.0.0.0:$lport" 1 "listen match for saddr and sport"
+
+	__chk_listen "" 1 "all listen sockets"
+
+	nr=$(ss -Ml $filter | wc -l)
+}
+
 # $1: ns, $2: port
 wait_local_port_listen()
 {
@@ -113,6 +150,7 @@ echo "a" | \
 				0.0.0.0 >/dev/null &
 wait_local_port_listen $ns 10000
 chk_msk_nr 0 "no msk on netns creation"
+chk_msk_listen 10000
 
 echo "b" | \
 	timeout ${timeout_test} \