tipc: introduce UDP replicast

This patch introduces UDP replicast. A concept where we emulate
multicast by sending multiple unicast messages to configured peers.

The purpose of replicast is mainly to be able to use TIPC in cloud
environments where IP multicast is disabled. Using replicas to unicast
multicast messages is costly as we have to copy each skb and send the
copies individually.

Signed-off-by: Richard Alpe <richard.alpe@ericsson.com>
Reviewed-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

1 files changed, 106 insertions, 12 deletions

diff --git a/net/tipc/udp_media.c b/net/tipc/udp_media.c
index b8ec1a18241e..6b938cc15daf 100644
--- a/net/tipc/udp_media.c
+++ b/net/tipc/udp_media.c
@@ -49,6 +49,7 @@
 #include "core.h"
 #include "bearer.h"
 #include "netlink.h"
+#include "msg.h"
 
 /* IANA assigned UDP port */
 #define UDP_PORT_DEFAULT	6118
@@ -70,6 +71,13 @@ struct udp_media_addr {
 	};
 };
 
+/* struct udp_replicast - container for UDP remote addresses */
+struct udp_replicast {
+	struct udp_media_addr addr;
+	struct rcu_head rcu;
+	struct list_head list;
+};
+
 /**
  * struct udp_bearer - ip/udp bearer data structure
  * @bearer:	associated generic tipc bearer
@@ -82,6 +90,7 @@ struct udp_bearer {
 	struct socket *ubsock;
 	u32 ifindex;
 	struct work_struct work;
+	struct udp_replicast rcast;
 };
 
 static int tipc_udp_is_mcast_addr(struct udp_media_addr *addr)
@@ -203,29 +212,75 @@ static int tipc_udp_send_msg(struct net *net, struct sk_buff *skb,
 {
 	struct udp_media_addr *src = (struct udp_media_addr *)&b->addr.value;
 	struct udp_media_addr *dst = (struct udp_media_addr *)&addr->value;
+	struct udp_replicast *rcast;
 	struct udp_bearer *ub;
 	int err = 0;
 
 	if (skb_headroom(skb) < UDP_MIN_HEADROOM) {
 		err = pskb_expand_head(skb, UDP_MIN_HEADROOM, 0, GFP_ATOMIC);
 		if (err)
-			goto tx_error;
+			goto out;
 	}
 
 	skb_set_inner_protocol(skb, htons(ETH_P_TIPC));
 	ub = rcu_dereference_rtnl(b->media_ptr);
 	if (!ub) {
 		err = -ENODEV;
-		goto tx_error;
+		goto out;
 	}
 
-	return tipc_udp_xmit(net, skb, ub, src, dst);
+	if (!addr->broadcast || list_empty(&ub->rcast.list))
+		return tipc_udp_xmit(net, skb, ub, src, dst);
 
-tx_error:
+	/* Replicast, send an skb to each configured IP address */
+	list_for_each_entry_rcu(rcast, &ub->rcast.list, list) {
+		struct sk_buff *_skb;
+
+		_skb = pskb_copy(skb, GFP_ATOMIC);
+		if (!_skb) {
+			err = -ENOMEM;
+			goto out;
+		}
+
+		err = tipc_udp_xmit(net, _skb, ub, src, &rcast->addr);
+		if (err) {
+			kfree_skb(_skb);
+			goto out;
+		}
+	}
+	err = 0;
+out:
 	kfree_skb(skb);
 	return err;
 }
 
+static int tipc_udp_rcast_add(struct tipc_bearer *b,
+			      struct udp_media_addr *addr)
+{
+	struct udp_replicast *rcast;
+	struct udp_bearer *ub;
+
+	ub = rcu_dereference_rtnl(b->media_ptr);
+	if (!ub)
+		return -ENODEV;
+
+	rcast = kmalloc(sizeof(*rcast), GFP_ATOMIC);
+	if (!rcast)
+		return -ENOMEM;
+
+	memcpy(&rcast->addr, addr, sizeof(struct udp_media_addr));
+
+	if (ntohs(addr->proto) == ETH_P_IP)
+		pr_info("New replicast peer: %pI4\n", &rcast->addr.ipv4);
+#if IS_ENABLED(CONFIG_IPV6)
+	else if (ntohs(addr->proto) == ETH_P_IPV6)
+		pr_info("New replicast peer: %pI6\n", &rcast->addr.ipv6);
+#endif
+
+	list_add_rcu(&rcast->list, &ub->rcast.list);
+	return 0;
+}
+
 /* tipc_udp_recv - read data from bearer socket */
 static int tipc_udp_recv(struct sock *sk, struct sk_buff *skb)
 {
@@ -320,6 +375,32 @@ static int tipc_parse_udp_addr(struct nlattr *nla, struct udp_media_addr *addr,
 	return -EADDRNOTAVAIL;
 }
 
+int tipc_udp_nl_bearer_add(struct tipc_bearer *b, struct nlattr *attr)
+{
+	int err;
+	struct udp_media_addr addr = {0};
+	struct nlattr *opts[TIPC_NLA_UDP_MAX + 1];
+	struct udp_media_addr *dst;
+
+	if (nla_parse_nested(opts, TIPC_NLA_UDP_MAX, attr, tipc_nl_udp_policy))
+		return -EINVAL;
+
+	if (!opts[TIPC_NLA_UDP_REMOTE])
+		return -EINVAL;
+
+	err = tipc_parse_udp_addr(opts[TIPC_NLA_UDP_REMOTE], &addr, NULL);
+	if (err)
+		return err;
+
+	dst = (struct udp_media_addr *)&b->bcast_addr.value;
+	if (tipc_udp_is_mcast_addr(dst)) {
+		pr_err("Can't add remote ip to TIPC UDP multicast bearer\n");
+		return -EINVAL;
+	}
+
+	return tipc_udp_rcast_add(b, &addr);
+}
+
 /**
  * tipc_udp_enable - callback to create a new udp bearer instance
  * @net:	network namespace
@@ -334,7 +415,7 @@ static int tipc_udp_enable(struct net *net, struct tipc_bearer *b,
 {
 	int err = -EINVAL;
 	struct udp_bearer *ub;
-	struct udp_media_addr *remote;
+	struct udp_media_addr remote = {0};
 	struct udp_media_addr local = {0};
 	struct udp_port_cfg udp_conf = {0};
 	struct udp_tunnel_sock_cfg tuncfg = {NULL};
@@ -344,6 +425,8 @@ static int tipc_udp_enable(struct net *net, struct tipc_bearer *b,
 	if (!ub)
 		return -ENOMEM;
 
+	INIT_LIST_HEAD(&ub->rcast.list);
+
 	if (!attrs[TIPC_NLA_BEARER_UDP_OPTS])
 		goto err;
 
@@ -362,9 +445,7 @@ static int tipc_udp_enable(struct net *net, struct tipc_bearer *b,
 	if (err)
 		goto err;
 
-	remote = (struct udp_media_addr *)&b->bcast_addr.value;
-	memset(remote, 0, sizeof(struct udp_media_addr));
-	err = tipc_parse_udp_addr(opts[TIPC_NLA_UDP_REMOTE], remote, NULL);
+	err = tipc_parse_udp_addr(opts[TIPC_NLA_UDP_REMOTE], &remote, NULL);
 	if (err)
 		goto err;
 
@@ -409,10 +490,17 @@ static int tipc_udp_enable(struct net *net, struct tipc_bearer *b,
 	tuncfg.encap_destroy = NULL;
 	setup_udp_tunnel_sock(net, ub->ubsock, &tuncfg);
 
-	if (tipc_udp_is_mcast_addr(remote)) {
-		if (enable_mcast(ub, remote))
-			goto err;
-	}
+	/**
+	 * The bcast media address port is used for all peers and the ip
+	 * is used if it's a multicast address.
+	 */
+	memcpy(&b->bcast_addr.value, &remote, sizeof(remote));
+	if (tipc_udp_is_mcast_addr(&remote))
+		err = enable_mcast(ub, &remote);
+	else
+		err = tipc_udp_rcast_add(b, &remote);
+	if (err)
+		goto err;
 
 	return 0;
 err:
@@ -424,6 +512,12 @@ err:
 static void cleanup_bearer(struct work_struct *work)
 {
 	struct udp_bearer *ub = container_of(work, struct udp_bearer, work);
+	struct udp_replicast *rcast, *tmp;
+
+	list_for_each_entry_safe(rcast, tmp, &ub->rcast.list, list) {
+		list_del_rcu(&rcast->list);
+		kfree_rcu(rcast, rcu);
+	}
 
 	if (ub->ubsock)
 		udp_tunnel_sock_release(ub->ubsock);