net: bridge: add support for IGMP/MLD stats and export them via netlink

This patch adds stats support for the currently used IGMP/MLD types by the
bridge. The stats are per-port (plus one stat per-bridge) and per-direction
(RX/TX). The stats are exported via netlink via the new linkxstats API
(RTM_GETSTATS). In order to minimize the performance impact, a new option
is used to enable/disable the stats - multicast_stats_enabled, similar to
the recent vlan stats. Also in order to avoid multiple IGMP/MLD type
lookups and checks, we make use of the current "igmp" member of the bridge
private skb->cb region to record the type on Rx (both host-generated and
external packets pass by multicast_rcv()). We can do that since the igmp
member was used as a boolean and all the valid IGMP/MLD types are positive
values. The normal bridge fast-path is not affected at all, the only
affected paths are the flooding ones and since we make use of the IGMP/MLD
type, we can quickly determine if the packet should be counted using
cache-hot data (cb's igmp member). We add counters for:
* IGMP Queries
* IGMP Leaves
* IGMP v1/v2/v3 reports

* MLD Queries
* MLD Leaves
* MLD v1/v2 reports

These are invaluable when monitoring or debugging complex multicast setups
with bridges.
Signed-off-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

include/uapi/linux/if_bridge.h

@@ -247,8 +247,34 @@ enum {
 enum {
 	BRIDGE_XSTATS_UNSPEC,
 	BRIDGE_XSTATS_VLAN,
+	BRIDGE_XSTATS_MCAST,
+	BRIDGE_XSTATS_PAD,
 	__BRIDGE_XSTATS_MAX
 };
 #define BRIDGE_XSTATS_MAX (__BRIDGE_XSTATS_MAX - 1)
 
+enum {
+	BR_MCAST_DIR_RX,
+	BR_MCAST_DIR_TX,
+	BR_MCAST_DIR_SIZE
+};
+
+/* IGMP/MLD statistics */
+struct br_mcast_stats {
+	__u64 igmp_queries[BR_MCAST_DIR_SIZE];
+	__u64 igmp_leaves[BR_MCAST_DIR_SIZE];
+	__u64 igmp_v1reports[BR_MCAST_DIR_SIZE];
+	__u64 igmp_v2reports[BR_MCAST_DIR_SIZE];
+	__u64 igmp_v3reports[BR_MCAST_DIR_SIZE];
+	__u64 igmp_parse_errors;
+
+	__u64 mld_queries[BR_MCAST_DIR_SIZE];
+	__u64 mld_leaves[BR_MCAST_DIR_SIZE];
+	__u64 mld_v1reports[BR_MCAST_DIR_SIZE];
+	__u64 mld_v2reports[BR_MCAST_DIR_SIZE];
+	__u64 mld_parse_errors;
+
+	__u64 mcast_bytes[BR_MCAST_DIR_SIZE];
+	__u64 mcast_packets[BR_MCAST_DIR_SIZE];
+};
 #endif /* _UAPI_LINUX_IF_BRIDGE_H */

include/uapi/linux/if_link.h

@@ -273,6 +273,7 @@ enum {
 	IFLA_BR_VLAN_DEFAULT_PVID,
 	IFLA_BR_PAD,
 	IFLA_BR_VLAN_STATS_ENABLED,
+	IFLA_BR_MCAST_STATS_ENABLED,
 	__IFLA_BR_MAX,
 };
 

net/bridge/br_device.c

@@ -104,8 +104,16 @@ static int br_dev_init(struct net_device *dev)
 		return -ENOMEM;
 
 	err = br_vlan_init(br);
-	if (err)
+	if (err) {
 		free_percpu(br->stats);
+		return err;
+	}
+
+	err = br_multicast_init_stats(br);
+	if (err) {
+		free_percpu(br->stats);
+		br_vlan_flush(br);
+	}
 	br_set_lockdep_class(dev);
 
 	return err;

net/bridge/br_forward.c

@@ -198,8 +198,10 @@ static void br_flood(struct net_bridge *br, struct sk_buff *skb,
 					   struct sk_buff *skb),
 		     bool unicast)
 {
-	struct net_bridge_port *p;
+	u8 igmp_type = br_multicast_igmp_type(skb);
+	__be16 proto = skb->protocol;
 	struct net_bridge_port *prev;
+	struct net_bridge_port *p;
 
 	prev = NULL;
 
@@ -218,6 +220,9 @@ static void br_flood(struct net_bridge *br, struct sk_buff *skb,
 		prev = maybe_deliver(prev, p, skb, __packet_hook);
 		if (IS_ERR(prev))
 			goto out;
+		if (prev == p)
+			br_multicast_count(p->br, p, proto, igmp_type,
+					   BR_MCAST_DIR_TX);
 	}
 
 	if (!prev)
@@ -257,9 +262,12 @@ static void br_multicast_flood(struct net_bridge_mdb_entry *mdst,
 					struct sk_buff *skb))
 {
 	struct net_device *dev = BR_INPUT_SKB_CB(skb)->brdev;
+	u8 igmp_type = br_multicast_igmp_type(skb);
 	struct net_bridge *br = netdev_priv(dev);
 	struct net_bridge_port *prev = NULL;
 	struct net_bridge_port_group *p;
+	__be16 proto = skb->protocol;
+
 	struct hlist_node *rp;
 
 	rp = rcu_dereference(hlist_first_rcu(&br->router_list));
@@ -277,6 +285,9 @@ static void br_multicast_flood(struct net_bridge_mdb_entry *mdst,
 		prev = maybe_deliver(prev, port, skb, __packet_hook);
 		if (IS_ERR(prev))
 			goto out;
+		if (prev == port)
+			br_multicast_count(port->br, port, proto, igmp_type,
+					   BR_MCAST_DIR_TX);
 
 		if ((unsigned long)lport >= (unsigned long)port)
 			p = rcu_dereference(p->next);

net/bridge/br_if.c

@@ -345,8 +345,8 @@ static int find_portno(struct net_bridge *br)
 static struct net_bridge_port *new_nbp(struct net_bridge *br,
 				       struct net_device *dev)
 {
-	int index;
 	struct net_bridge_port *p;
+	int index, err;
 
 	index = find_portno(br);
 	if (index < 0)
@@ -366,7 +366,12 @@ static struct net_bridge_port *new_nbp(struct net_bridge *br,
 	br_init_port(p);
 	br_set_state(p, BR_STATE_DISABLED);
 	br_stp_port_timer_init(p);
-	br_multicast_add_port(p);
+	err = br_multicast_add_port(p);
+	if (err) {
+		dev_put(dev);
+		kfree(p);
+		p = ERR_PTR(err);
+	}
 
 	return p;
 }

net/bridge/br_input.c

@@ -60,6 +60,9 @@ static int br_pass_frame_up(struct sk_buff *skb)
 	skb = br_handle_vlan(br, vg, skb);
 	if (!skb)
 		return NET_RX_DROP;
+	/* update the multicast stats if the packet is IGMP/MLD */
+	br_multicast_count(br, NULL, skb->protocol, br_multicast_igmp_type(skb),
+			   BR_MCAST_DIR_TX);
 
 	return NF_HOOK(NFPROTO_BRIDGE, NF_BR_LOCAL_IN,
 		       dev_net(indev), NULL, skb, indev, NULL,

net/bridge/br_netlink.c

@@ -851,6 +851,7 @@ static const struct nla_policy br_policy[IFLA_BR_MAX + 1] = {
 	[IFLA_BR_NF_CALL_ARPTABLES] = { .type = NLA_U8 },
 	[IFLA_BR_VLAN_DEFAULT_PVID] = { .type = NLA_U16 },
 	[IFLA_BR_VLAN_STATS_ENABLED] = { .type = NLA_U8 },
+	[IFLA_BR_MCAST_STATS_ENABLED] = { .type = NLA_U8 },
 };
 
 static int br_changelink(struct net_device *brdev, struct nlattr *tb[],
@@ -1055,6 +1056,13 @@ static int br_changelink(struct net_device *brdev, struct nlattr *tb[],
 
 		br->multicast_startup_query_interval = clock_t_to_jiffies(val);
 	}
+
+	if (data[IFLA_BR_MCAST_STATS_ENABLED]) {
+		__u8 mcast_stats;
+
+		mcast_stats = nla_get_u8(data[IFLA_BR_MCAST_STATS_ENABLED]);
+		br->multicast_stats_enabled = !!mcast_stats;
+	}
 #endif
 #if IS_ENABLED(CONFIG_BRIDGE_NETFILTER)
 	if (data[IFLA_BR_NF_CALL_IPTABLES]) {
@@ -1110,6 +1118,7 @@ static size_t br_get_size(const struct net_device *brdev)
 	       nla_total_size(sizeof(u8)) +     /* IFLA_BR_MCAST_SNOOPING */
 	       nla_total_size(sizeof(u8)) +     /* IFLA_BR_MCAST_QUERY_USE_IFADDR */
 	       nla_total_size(sizeof(u8)) +     /* IFLA_BR_MCAST_QUERIER */
+	       nla_total_size(sizeof(u8)) +     /* IFLA_BR_MCAST_STATS_ENABLED */
 	       nla_total_size(sizeof(u32)) +    /* IFLA_BR_MCAST_HASH_ELASTICITY */
 	       nla_total_size(sizeof(u32)) +    /* IFLA_BR_MCAST_HASH_MAX */
 	       nla_total_size(sizeof(u32)) +    /* IFLA_BR_MCAST_LAST_MEMBER_CNT */
@@ -1187,6 +1196,8 @@ static int br_fill_info(struct sk_buff *skb, const struct net_device *brdev)
 	    nla_put_u8(skb, IFLA_BR_MCAST_QUERY_USE_IFADDR,
 		       br->multicast_query_use_ifaddr) ||
 	    nla_put_u8(skb, IFLA_BR_MCAST_QUERIER, br->multicast_querier) ||
+	    nla_put_u8(skb, IFLA_BR_MCAST_STATS_ENABLED,
+		       br->multicast_stats_enabled) ||
 	    nla_put_u32(skb, IFLA_BR_MCAST_HASH_ELASTICITY,
 			br->hash_elasticity) ||
 	    nla_put_u32(skb, IFLA_BR_MCAST_HASH_MAX, br->hash_max) ||
@@ -1242,21 +1253,21 @@ static size_t bridge_get_linkxstats_size(const struct net_device *dev)
 	int numvls = 0;
 
 	vg = br_vlan_group(br);
-	if (!vg)
-		return 0;
-
-	/* we need to count all, even placeholder entries */
-	list_for_each_entry(v, &vg->vlan_list, vlist)
-		numvls++;
+	if (vg) {
+		/* we need to count all, even placeholder entries */
+		list_for_each_entry(v, &vg->vlan_list, vlist)
+			numvls++;
+	}
 
-	/* account for the vlans and the link xstats type nest attribute */
 	return numvls * nla_total_size(sizeof(struct bridge_vlan_xstats)) +
+	       nla_total_size(sizeof(struct br_mcast_stats)) +
 	       nla_total_size(0);
 }
 
 static size_t brport_get_linkxstats_size(const struct net_device *dev)
 {
-	return nla_total_size(0);
+	return nla_total_size(sizeof(struct br_mcast_stats)) +
+	       nla_total_size(0);
 }
 
 static size_t br_get_linkxstats_size(const struct net_device *dev, int attr)
@@ -1280,37 +1291,50 @@ static int bridge_fill_linkxstats(struct sk_buff *skb,
 				  int *prividx)
 {
 	struct net_bridge *br = netdev_priv(dev);
+	struct nlattr *nla __maybe_unused;
 	struct net_bridge_vlan_group *vg;
 	struct net_bridge_vlan *v;
 	struct nlattr *nest;
 	int vl_idx = 0;
 
-	vg = br_vlan_group(br);
-	if (!vg)
-		goto out;
 	nest = nla_nest_start(skb, LINK_XSTATS_TYPE_BRIDGE);
 	if (!nest)
 		return -EMSGSIZE;
-	list_for_each_entry(v, &vg->vlan_list, vlist) {
-		struct bridge_vlan_xstats vxi;
-		struct br_vlan_stats stats;
 
-		if (++vl_idx < *prividx)
-			continue;
-		memset(&vxi, 0, sizeof(vxi));
-		vxi.vid = v->vid;
-		br_vlan_get_stats(v, &stats);
-		vxi.rx_bytes = stats.rx_bytes;
-		vxi.rx_packets = stats.rx_packets;
-		vxi.tx_bytes = stats.tx_bytes;
-		vxi.tx_packets = stats.tx_packets;
-
-		if (nla_put(skb, BRIDGE_XSTATS_VLAN, sizeof(vxi), &vxi))
+	vg = br_vlan_group(br);
+	if (vg) {
+		list_for_each_entry(v, &vg->vlan_list, vlist) {
+			struct bridge_vlan_xstats vxi;
+			struct br_vlan_stats stats;
+
+			if (++vl_idx < *prividx)
+				continue;
+			memset(&vxi, 0, sizeof(vxi));
+			vxi.vid = v->vid;
+			br_vlan_get_stats(v, &stats);
+			vxi.rx_bytes = stats.rx_bytes;
+			vxi.rx_packets = stats.rx_packets;
+			vxi.tx_bytes = stats.tx_bytes;
+			vxi.tx_packets = stats.tx_packets;
+
+			if (nla_put(skb, BRIDGE_XSTATS_VLAN, sizeof(vxi), &vxi))
+				goto nla_put_failure;
+		}
+	}
+
+#ifdef CONFIG_BRIDGE_IGMP_SNOOPING
+	if (++vl_idx >= *prividx) {
+		nla = nla_reserve_64bit(skb, BRIDGE_XSTATS_MCAST,
+					sizeof(struct br_mcast_stats),
+					BRIDGE_XSTATS_PAD);
+		if (!nla)
 			goto nla_put_failure;
+		br_multicast_get_stats(br, NULL, nla_data(nla));
 	}
+#endif
 	nla_nest_end(skb, nest);
 	*prividx = 0;
-out:
+
 	return 0;
 
 nla_put_failure:
@@ -1324,11 +1348,26 @@ static int brport_fill_linkxstats(struct sk_buff *skb,
 				  const struct net_device *dev,
 				  int *prividx)
 {
+	struct net_bridge_port *p = br_port_get_rtnl(dev);
+	struct nlattr *nla __maybe_unused;
 	struct nlattr *nest;
 
+	if (!p)
+		return 0;
+
 	nest = nla_nest_start(skb, LINK_XSTATS_TYPE_BRIDGE);
 	if (!nest)
 		return -EMSGSIZE;
+#ifdef CONFIG_BRIDGE_IGMP_SNOOPING
+	nla = nla_reserve_64bit(skb, BRIDGE_XSTATS_MCAST,
+				sizeof(struct br_mcast_stats),
+				BRIDGE_XSTATS_PAD);
+	if (!nla) {
+		nla_nest_end(skb, nest);
+		return -EMSGSIZE;
+	}
+	br_multicast_get_stats(p->br, p, nla_data(nla));
+#endif
 	nla_nest_end(skb, nest);
 
 	return 0;

net/bridge/br_private.h

@@ -75,6 +75,12 @@ struct bridge_mcast_querier {
 	struct br_ip addr;
 	struct net_bridge_port __rcu	*port;
 };
+
+/* IGMP/MLD statistics */
+struct bridge_mcast_stats {
+	struct br_mcast_stats mstats;
+	struct u64_stats_sync syncp;
+};
 #endif
 
 struct br_vlan_stats {
@@ -229,6 +235,7 @@ struct net_bridge_port
 	struct bridge_mcast_own_query	ip6_own_query;
 #endif /* IS_ENABLED(CONFIG_IPV6) */
 	unsigned char			multicast_router;
+	struct bridge_mcast_stats	__percpu *mcast_stats;
 	struct timer_list		multicast_router_timer;
 	struct hlist_head		mglist;
 	struct hlist_node		rlist;
@@ -315,6 +322,7 @@ struct net_bridge
 	u8				multicast_querier:1;
 	u8				multicast_query_use_ifaddr:1;
 	u8				has_ipv6_addr:1;
+	u8				multicast_stats_enabled:1;
 
 	u32				hash_elasticity;
 	u32				hash_max;
@@ -337,6 +345,7 @@ struct net_bridge
 	struct bridge_mcast_other_query	ip4_other_query;
 	struct bridge_mcast_own_query	ip4_own_query;
 	struct bridge_mcast_querier	ip4_querier;
+	struct bridge_mcast_stats	__percpu *mcast_stats;
 #if IS_ENABLED(CONFIG_IPV6)
 	struct bridge_mcast_other_query	ip6_other_query;
 	struct bridge_mcast_own_query	ip6_own_query;
@@ -543,7 +552,7 @@ int br_multicast_rcv(struct net_bridge *br, struct net_bridge_port *port,
 		     struct sk_buff *skb, u16 vid);
 struct net_bridge_mdb_entry *br_mdb_get(struct net_bridge *br,
 					struct sk_buff *skb, u16 vid);
-void br_multicast_add_port(struct net_bridge_port *port);
+int br_multicast_add_port(struct net_bridge_port *port);
 void br_multicast_del_port(struct net_bridge_port *port);
 void br_multicast_enable_port(struct net_bridge_port *port);
 void br_multicast_disable_port(struct net_bridge_port *port);
@@ -576,6 +585,12 @@ void br_mdb_notify(struct net_device *dev, struct net_bridge_port *port,
 		   struct br_ip *group, int type, u8 flags);
 void br_rtr_notify(struct net_device *dev, struct net_bridge_port *port,
 		   int type);
+void br_multicast_count(struct net_bridge *br, const struct net_bridge_port *p,
+			__be16 proto, u8 type, u8 dir);
+int br_multicast_init_stats(struct net_bridge *br);
+void br_multicast_get_stats(const struct net_bridge *br,
+			    const struct net_bridge_port *p,
+			    struct br_mcast_stats *dest);
 
 #define mlock_dereference(X, br) \
 	rcu_dereference_protected(X, lockdep_is_held(&br->multicast_lock))
@@ -623,6 +638,11 @@ static inline bool br_multicast_querier_exists(struct net_bridge *br,
 		return false;
 	}
 }
+
+static inline int br_multicast_igmp_type(const struct sk_buff *skb)
+{
+	return BR_INPUT_SKB_CB(skb)->igmp;
+}
 #else
 static inline int br_multicast_rcv(struct net_bridge *br,
 				   struct net_bridge_port *port,
@@ -638,8 +658,9 @@ static inline struct net_bridge_mdb_entry *br_mdb_get(struct net_bridge *br,
 	return NULL;
 }
 
-static inline void br_multicast_add_port(struct net_bridge_port *port)
+static inline int br_multicast_add_port(struct net_bridge_port *port)
 {
+	return 0;
 }
 
 static inline void br_multicast_del_port(struct net_bridge_port *port)
@@ -695,6 +716,22 @@ static inline void br_mdb_init(void)
 static inline void br_mdb_uninit(void)
 {
 }
+
+static inline void br_multicast_count(struct net_bridge *br,
+				      const struct net_bridge_port *p,
+				      __be16 proto, u8 type, u8 dir)
+{
+}
+
+static inline int br_multicast_init_stats(struct net_bridge *br)
+{
+	return 0;
+}
+
+static inline int br_multicast_igmp_type(const struct sk_buff *skb)
+{
+	return 0;
+}
 #endif
 
 /* br_vlan.c */

net/bridge/br_sysfs_br.c

@@ -618,6 +618,30 @@ static ssize_t multicast_startup_query_interval_store(
 	return store_bridge_parm(d, buf, len, set_startup_query_interval);
 }
 static DEVICE_ATTR_RW(multicast_startup_query_interval);
+
+static ssize_t multicast_stats_enabled_show(struct device *d,
+					    struct device_attribute *attr,
+					    char *buf)
+{
+	struct net_bridge *br = to_bridge(d);
+
+	return sprintf(buf, "%u\n", br->multicast_stats_enabled);
+}
+
+static int set_stats_enabled(struct net_bridge *br, unsigned long val)
+{
+	br->multicast_stats_enabled = !!val;
+	return 0;
+}
+
+static ssize_t multicast_stats_enabled_store(struct device *d,
+					     struct device_attribute *attr,
+					     const char *buf,
+					     size_t len)
+{
+	return store_bridge_parm(d, buf, len, set_stats_enabled);
+}
+static DEVICE_ATTR_RW(multicast_stats_enabled);
 #endif
 #if IS_ENABLED(CONFIG_BRIDGE_NETFILTER)
 static ssize_t nf_call_iptables_show(
@@ -784,6 +808,7 @@ static struct attribute *bridge_attrs[] = {
 	&dev_attr_multicast_query_interval.attr,
 	&dev_attr_multicast_query_response_interval.attr,
 	&dev_attr_multicast_startup_query_interval.attr,
+	&dev_attr_multicast_stats_enabled.attr,
 #endif
 #if IS_ENABLED(CONFIG_BRIDGE_NETFILTER)
 	&dev_attr_nf_call_iptables.attr,