Merge branch 'Support-for-fdb-ECMP-nexthop-groups'

Roopa Prabhu says:

====================
Support for fdb ECMP nexthop groups

This series introduces ecmp nexthops and nexthop groups
for mac fdb entries. In subsequent patches this is used
by the vxlan driver fdb entries. The use case is
E-VPN multihoming [1,2,3] which requires bridged vxlan traffic
to be load balanced to remote switches (vteps) belonging to
the same multi-homed ethernet segment (This is analogous to
a multi-homed LAG but over vxlan).

Changes include new nexthop flag NHA_FDB for nexthops
referenced by fdb entries. These nexthops only have ip.
The patches make sure that routes dont reference such nexthops.

example:
$ip nexthop add id 12 via 172.16.1.2 fdb
$ip nexthop add id 13 via 172.16.1.3 fdb
$ip nexthop add id 102 group 12/13 fdb

$bridge fdb add 02:02:00:00:00:13 dev vxlan1000 nhid 101 self

[1] E-VPN https://tools.ietf.org/html/rfc7432
[2] E-VPN VxLAN: https://tools.ietf.org/html/rfc8365
[3] LPC talk with mention of nexthop groups for L2 ecmp
http://vger.kernel.org/lpc_net2018_talks/scaling_bridge_fdb_database_slidesV3.pdf

v4 -
    - fix error path free_skb in vxlan_xmit_nh
    - fix atomic notifier initialization issue
      (Reported-by: kernel test robot <rong.a.chen@intel.com>)
      The reported error was easy to locate and fix, but i was not
      able to re-test with the robot reproducer script due to some
      other issues with running the script on my test system.

v3 - fix wording in selftest print as pointed out by davidA

v2 -
	- dropped nikolays fixes for nexthop multipath null pointer deref
	  (he will send those separately)
	- added negative tests for route add with fdb nexthop + a few more
	- Fixes for a few  fdb replace conditions found during more testing
	- Moved to rcu_dereference_rtnl in vxlan_fdb_info and consolidate rcu
	  dereferences
	- Fixes to build failures Reported-by: kbuild test robot <lkp@intel.com>
	- DavidA, I am going to send a separate patch for the neighbor code validation
	  for NDA_NH_ID if thats ok.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/net/vxlan.c

@@ -26,6 +26,7 @@
 #include <net/netns/generic.h>
 #include <net/tun_proto.h>
 #include <net/vxlan.h>
+#include <net/nexthop.h>
 
 #if IS_ENABLED(CONFIG_IPV6)
 #include <net/ip6_tunnel.h>
@@ -78,6 +79,9 @@ struct vxlan_fdb {
 	u16		  state;	/* see ndm_state */
 	__be32		  vni;
 	u16		  flags;	/* see ndm_flags and below */
+	struct list_head  nh_list;
+	struct nexthop __rcu *nh;
+	struct vxlan_dev  *vdev;
 };
 
 #define NTF_VXLAN_ADDED_BY_USER 0x100
@@ -174,11 +178,15 @@ static inline struct hlist_head *vs_head(struct net *net, __be16 port)
  */
 static inline struct vxlan_rdst *first_remote_rcu(struct vxlan_fdb *fdb)
 {
+	if (rcu_access_pointer(fdb->nh))
+		return NULL;
 	return list_entry_rcu(fdb->remotes.next, struct vxlan_rdst, list);
 }
 
 static inline struct vxlan_rdst *first_remote_rtnl(struct vxlan_fdb *fdb)
 {
+	if (rcu_access_pointer(fdb->nh))
+		return NULL;
 	return list_first_entry(&fdb->remotes, struct vxlan_rdst, list);
 }
 
@@ -251,9 +259,10 @@ static int vxlan_fdb_info(struct sk_buff *skb, struct vxlan_dev *vxlan,
 {
 	unsigned long now = jiffies;
 	struct nda_cacheinfo ci;
+	bool send_ip, send_eth;
 	struct nlmsghdr *nlh;
+	struct nexthop *nh;
 	struct ndmsg *ndm;
-	bool send_ip, send_eth;
 
 	nlh = nlmsg_put(skb, portid, seq, type, sizeof(*ndm), flags);
 	if (nlh == NULL)
@@ -264,16 +273,21 @@ static int vxlan_fdb_info(struct sk_buff *skb, struct vxlan_dev *vxlan,
 
 	send_eth = send_ip = true;
 
+	nh = rcu_dereference_rtnl(fdb->nh);
 	if (type == RTM_GETNEIGH) {
-		send_ip = !vxlan_addr_any(&rdst->remote_ip);
+		if (rdst) {
+			send_ip = !vxlan_addr_any(&rdst->remote_ip);
+			ndm->ndm_family = send_ip ? rdst->remote_ip.sa.sa_family : AF_INET;
+		} else if (nh) {
+			ndm->ndm_family = nexthop_get_family(nh);
+		}
 		send_eth = !is_zero_ether_addr(fdb->eth_addr);
-		ndm->ndm_family = send_ip ? rdst->remote_ip.sa.sa_family : AF_INET;
 	} else
 		ndm->ndm_family	= AF_BRIDGE;
 	ndm->ndm_state = fdb->state;
 	ndm->ndm_ifindex = vxlan->dev->ifindex;
 	ndm->ndm_flags = fdb->flags;
-	if (rdst->offloaded)
+	if (rdst && rdst->offloaded)
 		ndm->ndm_flags |= NTF_OFFLOADED;
 	ndm->ndm_type = RTN_UNICAST;
 
@@ -284,23 +298,30 @@ static int vxlan_fdb_info(struct sk_buff *skb, struct vxlan_dev *vxlan,
 
 	if (send_eth && nla_put(skb, NDA_LLADDR, ETH_ALEN, &fdb->eth_addr))
 		goto nla_put_failure;
+	if (nh) {
+		if (nla_put_u32(skb, NDA_NH_ID, nh->id))
+			goto nla_put_failure;
+	} else if (rdst) {
+		if (send_ip && vxlan_nla_put_addr(skb, NDA_DST,
+						  &rdst->remote_ip))
+			goto nla_put_failure;
+
+		if (rdst->remote_port &&
+		    rdst->remote_port != vxlan->cfg.dst_port &&
+		    nla_put_be16(skb, NDA_PORT, rdst->remote_port))
+			goto nla_put_failure;
+		if (rdst->remote_vni != vxlan->default_dst.remote_vni &&
+		    nla_put_u32(skb, NDA_VNI, be32_to_cpu(rdst->remote_vni)))
+			goto nla_put_failure;
+		if (rdst->remote_ifindex &&
+		    nla_put_u32(skb, NDA_IFINDEX, rdst->remote_ifindex))
+			goto nla_put_failure;
+	}
 
-	if (send_ip && vxlan_nla_put_addr(skb, NDA_DST, &rdst->remote_ip))
-		goto nla_put_failure;
-
-	if (rdst->remote_port && rdst->remote_port != vxlan->cfg.dst_port &&
-	    nla_put_be16(skb, NDA_PORT, rdst->remote_port))
-		goto nla_put_failure;
-	if (rdst->remote_vni != vxlan->default_dst.remote_vni &&
-	    nla_put_u32(skb, NDA_VNI, be32_to_cpu(rdst->remote_vni)))
-		goto nla_put_failure;
 	if ((vxlan->cfg.flags & VXLAN_F_COLLECT_METADATA) && fdb->vni &&
 	    nla_put_u32(skb, NDA_SRC_VNI,
 			be32_to_cpu(fdb->vni)))
 		goto nla_put_failure;
-	if (rdst->remote_ifindex &&
-	    nla_put_u32(skb, NDA_IFINDEX, rdst->remote_ifindex))
-		goto nla_put_failure;
 
 	ci.ndm_used	 = jiffies_to_clock_t(now - fdb->used);
 	ci.ndm_confirmed = 0;
@@ -401,7 +422,7 @@ static int vxlan_fdb_notify(struct vxlan_dev *vxlan, struct vxlan_fdb *fdb,
 {
 	int err;
 
-	if (swdev_notify) {
+	if (swdev_notify && rd) {
 		switch (type) {
 		case RTM_NEWNEIGH:
 			err = vxlan_fdb_switchdev_call_notifiers(vxlan, fdb, rd,
@@ -793,8 +814,9 @@ static int vxlan_gro_complete(struct sock *sk, struct sk_buff *skb, int nhoff)
 	return eth_gro_complete(skb, nhoff + sizeof(struct vxlanhdr));
 }
 
-static struct vxlan_fdb *vxlan_fdb_alloc(const u8 *mac, __u16 state,
-					 __be32 src_vni, __u16 ndm_flags)
+static struct vxlan_fdb *vxlan_fdb_alloc(struct vxlan_dev *vxlan, const u8 *mac,
+					 __u16 state, __be32 src_vni,
+					 __u16 ndm_flags)
 {
 	struct vxlan_fdb *f;
 
@@ -805,6 +827,9 @@ static struct vxlan_fdb *vxlan_fdb_alloc(const u8 *mac, __u16 state,
 	f->flags = ndm_flags;
 	f->updated = f->used = jiffies;
 	f->vni = src_vni;
+	f->nh = NULL;
+	f->vdev = vxlan;
+	INIT_LIST_HEAD(&f->nh_list);
 	INIT_LIST_HEAD(&f->remotes);
 	memcpy(f->eth_addr, mac, ETH_ALEN);
 
@@ -819,11 +844,78 @@ static void vxlan_fdb_insert(struct vxlan_dev *vxlan, const u8 *mac,
 			   vxlan_fdb_head(vxlan, mac, src_vni));
 }
 
+static int vxlan_fdb_nh_update(struct vxlan_dev *vxlan, struct vxlan_fdb *fdb,
+			       u32 nhid, struct netlink_ext_ack *extack)
+{
+	struct nexthop *old_nh = rtnl_dereference(fdb->nh);
+	struct nh_group *nhg;
+	struct nexthop *nh;
+	int err = -EINVAL;
+
+	if (old_nh && old_nh->id == nhid)
+		return 0;
+
+	nh = nexthop_find_by_id(vxlan->net, nhid);
+	if (!nh) {
+		NL_SET_ERR_MSG(extack, "Nexthop id does not exist");
+		goto err_inval;
+	}
+
+	if (nh) {
+		if (!nexthop_get(nh)) {
+			NL_SET_ERR_MSG(extack, "Nexthop has been deleted");
+			nh = NULL;
+			goto err_inval;
+		}
+		if (!nh->is_fdb_nh) {
+			NL_SET_ERR_MSG(extack, "Nexthop is not a fdb nexthop");
+			goto err_inval;
+		}
+
+		if (!nh->is_group || !nh->nh_grp->mpath) {
+			NL_SET_ERR_MSG(extack, "Nexthop is not a multipath group");
+			goto err_inval;
+		}
+
+		/* check nexthop group family */
+		nhg = rtnl_dereference(nh->nh_grp);
+		switch (vxlan->default_dst.remote_ip.sa.sa_family) {
+		case AF_INET:
+			if (!nhg->has_v4) {
+				err = -EAFNOSUPPORT;
+				NL_SET_ERR_MSG(extack, "Nexthop group family not supported");
+				goto err_inval;
+			}
+			break;
+		case AF_INET6:
+			if (nhg->has_v4) {
+				err = -EAFNOSUPPORT;
+				NL_SET_ERR_MSG(extack, "Nexthop group family not supported");
+				goto err_inval;
+			}
+		}
+	}
+
+	if (old_nh) {
+		list_del_rcu(&fdb->nh_list);
+		nexthop_put(old_nh);
+	}
+	rcu_assign_pointer(fdb->nh, nh);
+	list_add_tail_rcu(&fdb->nh_list, &nh->fdb_list);
+	return 1;
+
+err_inval:
+	if (nh)
+		nexthop_put(nh);
+	return err;
+}
+
 static int vxlan_fdb_create(struct vxlan_dev *vxlan,
 			    const u8 *mac, union vxlan_addr *ip,
 			    __u16 state, __be16 port, __be32 src_vni,
 			    __be32 vni, __u32 ifindex, __u16 ndm_flags,
-			    struct vxlan_fdb **fdb)
+			    u32 nhid, struct vxlan_fdb **fdb,
+			    struct netlink_ext_ack *extack)
 {
 	struct vxlan_rdst *rd = NULL;
 	struct vxlan_fdb *f;
@@ -834,24 +926,37 @@ static int vxlan_fdb_create(struct vxlan_dev *vxlan,
 		return -ENOSPC;
 
 	netdev_dbg(vxlan->dev, "add %pM -> %pIS\n", mac, ip);
-	f = vxlan_fdb_alloc(mac, state, src_vni, ndm_flags);
+	f = vxlan_fdb_alloc(vxlan, mac, state, src_vni, ndm_flags);
 	if (!f)
 		return -ENOMEM;
 
-	rc = vxlan_fdb_append(f, ip, port, vni, ifindex, &rd);
-	if (rc < 0) {
-		kfree(f);
-		return rc;
-	}
+	if (nhid)
+		rc = vxlan_fdb_nh_update(vxlan, f, nhid, extack);
+	else
+		rc = vxlan_fdb_append(f, ip, port, vni, ifindex, &rd);
+	if (rc < 0)
+		goto errout;
 
 	*fdb = f;
 
 	return 0;
+
+errout:
+	kfree(f);
+	return rc;
 }
 
 static void __vxlan_fdb_free(struct vxlan_fdb *f)
 {
 	struct vxlan_rdst *rd, *nd;
+	struct nexthop *nh;
+
+	nh = rcu_dereference_raw(f->nh);
+	if (nh) {
+		rcu_assign_pointer(f->nh, NULL);
+		list_del_rcu(&f->nh_list);
+		nexthop_put(nh);
+	}
 
 	list_for_each_entry_safe(rd, nd, &f->remotes, list) {
 		dst_cache_destroy(&rd->dst_cache);
@@ -875,12 +980,18 @@ static void vxlan_fdb_destroy(struct vxlan_dev *vxlan, struct vxlan_fdb *f,
 	netdev_dbg(vxlan->dev, "delete %pM\n", f->eth_addr);
 
 	--vxlan->addrcnt;
-	if (do_notify)
-		list_for_each_entry(rd, &f->remotes, list)
-			vxlan_fdb_notify(vxlan, f, rd, RTM_DELNEIGH,
+	if (do_notify) {
+		if (rcu_access_pointer(f->nh))
+			vxlan_fdb_notify(vxlan, f, NULL, RTM_DELNEIGH,
 					 swdev_notify, NULL);
+		else
+			list_for_each_entry(rd, &f->remotes, list)
+				vxlan_fdb_notify(vxlan, f, rd, RTM_DELNEIGH,
+						 swdev_notify, NULL);
+	}
 
 	hlist_del_rcu(&f->hlist);
+	f->vdev = NULL;
 	call_rcu(&f->rcu, vxlan_fdb_free);
 }
 
@@ -897,7 +1008,7 @@ static int vxlan_fdb_update_existing(struct vxlan_dev *vxlan,
 				     __u16 state, __u16 flags,
 				     __be16 port, __be32 vni,
 				     __u32 ifindex, __u16 ndm_flags,
-				     struct vxlan_fdb *f,
+				     struct vxlan_fdb *f, u32 nhid,
 				     bool swdev_notify,
 				     struct netlink_ext_ack *extack)
 {
@@ -908,6 +1019,18 @@ static int vxlan_fdb_update_existing(struct vxlan_dev *vxlan,
 	int rc = 0;
 	int err;
 
+	if (nhid && !rcu_access_pointer(f->nh)) {
+		NL_SET_ERR_MSG(extack,
+			       "Cannot replace an existing non nexthop fdb with a nexthop");
+		return -EOPNOTSUPP;
+	}
+
+	if (nhid && (flags & NLM_F_APPEND)) {
+		NL_SET_ERR_MSG(extack,
+			       "Cannot append to a nexthop fdb");
+		return -EOPNOTSUPP;
+	}
+
 	/* Do not allow an externally learned entry to take over an entry added
 	 * by the user.
 	 */
@@ -929,10 +1052,17 @@ static int vxlan_fdb_update_existing(struct vxlan_dev *vxlan,
 		/* Only change unicasts */
 		if (!(is_multicast_ether_addr(f->eth_addr) ||
 		      is_zero_ether_addr(f->eth_addr))) {
-			rc = vxlan_fdb_replace(f, ip, port, vni,
-					       ifindex, &oldrd);
+			if (nhid) {
+				rc = vxlan_fdb_nh_update(vxlan, f, nhid, extack);
+				if (rc < 0)
+					return rc;
+			} else {
+				rc = vxlan_fdb_replace(f, ip, port, vni,
+						       ifindex, &oldrd);
+			}
 			notify |= rc;
 		} else {
+			NL_SET_ERR_MSG(extack, "Cannot replace non-unicast fdb entries");
 			return -EOPNOTSUPP;
 		}
 	}
@@ -962,6 +1092,8 @@ static int vxlan_fdb_update_existing(struct vxlan_dev *vxlan,
 	return 0;
 
 err_notify:
+	if (nhid)
+		return err;
 	if ((flags & NLM_F_REPLACE) && rc)
 		*rd = oldrd;
 	else if ((flags & NLM_F_APPEND) && rc) {
@@ -975,7 +1107,7 @@ static int vxlan_fdb_update_create(struct vxlan_dev *vxlan,
 				   const u8 *mac, union vxlan_addr *ip,
 				   __u16 state, __u16 flags,
 				   __be16 port, __be32 src_vni, __be32 vni,
-				   __u32 ifindex, __u16 ndm_flags,
+				   __u32 ifindex, __u16 ndm_flags, u32 nhid,
 				   bool swdev_notify,
 				   struct netlink_ext_ack *extack)
 {
@@ -990,7 +1122,7 @@ static int vxlan_fdb_update_create(struct vxlan_dev *vxlan,
 
 	netdev_dbg(vxlan->dev, "add %pM -> %pIS\n", mac, ip);
 	rc = vxlan_fdb_create(vxlan, mac, ip, state, port, src_vni,
-			      vni, ifindex, fdb_flags, &f);
+			      vni, ifindex, fdb_flags, nhid, &f, extack);
 	if (rc < 0)
 		return rc;
 
@@ -1012,7 +1144,7 @@ static int vxlan_fdb_update(struct vxlan_dev *vxlan,
 			    const u8 *mac, union vxlan_addr *ip,
 			    __u16 state, __u16 flags,
 			    __be16 port, __be32 src_vni, __be32 vni,
-			    __u32 ifindex, __u16 ndm_flags,
+			    __u32 ifindex, __u16 ndm_flags, u32 nhid,
 			    bool swdev_notify,
 			    struct netlink_ext_ack *extack)
 {
@@ -1028,14 +1160,15 @@ static int vxlan_fdb_update(struct vxlan_dev *vxlan,
 
 		return vxlan_fdb_update_existing(vxlan, ip, state, flags, port,
 						 vni, ifindex, ndm_flags, f,
-						 swdev_notify, extack);
+						 nhid, swdev_notify, extack);
 	} else {
 		if (!(flags & NLM_F_CREATE))
 			return -ENOENT;
 
 		return vxlan_fdb_update_create(vxlan, mac, ip, state, flags,
 					       port, src_vni, vni, ifindex,
-					       ndm_flags, swdev_notify, extack);
+					       ndm_flags, nhid, swdev_notify,
+					       extack);
 	}
 }
 
@@ -1049,7 +1182,7 @@ static void vxlan_fdb_dst_destroy(struct vxlan_dev *vxlan, struct vxlan_fdb *f,
 
 static int vxlan_fdb_parse(struct nlattr *tb[], struct vxlan_dev *vxlan,
 			   union vxlan_addr *ip, __be16 *port, __be32 *src_vni,
-			   __be32 *vni, u32 *ifindex)
+			   __be32 *vni, u32 *ifindex, u32 *nhid)
 {
 	struct net *net = dev_net(vxlan->dev);
 	int err;
@@ -1109,6 +1242,11 @@ static int vxlan_fdb_parse(struct nlattr *tb[], struct vxlan_dev *vxlan,
 		*ifindex = 0;
 	}
 
+	if (tb[NDA_NH_ID])
+		*nhid = nla_get_u32(tb[NDA_NH_ID]);
+	else
+		*nhid = 0;
+
 	return 0;
 }
 
@@ -1123,7 +1261,7 @@ static int vxlan_fdb_add(struct ndmsg *ndm, struct nlattr *tb[],
 	union vxlan_addr ip;
 	__be16 port;
 	__be32 src_vni, vni;
-	u32 ifindex;
+	u32 ifindex, nhid;
 	u32 hash_index;
 	int err;
 
@@ -1133,10 +1271,11 @@ static int vxlan_fdb_add(struct ndmsg *ndm, struct nlattr *tb[],
 		return -EINVAL;
 	}
 
-	if (tb[NDA_DST] == NULL)
+	if (!tb || (!tb[NDA_DST] && !tb[NDA_NH_ID]))
 		return -EINVAL;
 
-	err = vxlan_fdb_parse(tb, vxlan, &ip, &port, &src_vni, &vni, &ifindex);
+	err = vxlan_fdb_parse(tb, vxlan, &ip, &port, &src_vni, &vni, &ifindex,
+			      &nhid);
 	if (err)
 		return err;
 
@@ -1148,7 +1287,7 @@ static int vxlan_fdb_add(struct ndmsg *ndm, struct nlattr *tb[],
 	err = vxlan_fdb_update(vxlan, addr, &ip, ndm->ndm_state, flags,
 			       port, src_vni, vni, ifindex,
 			       ndm->ndm_flags | NTF_VXLAN_ADDED_BY_USER,
-			       true, extack);
+			       nhid, true, extack);
 	spin_unlock_bh(&vxlan->hash_lock[hash_index]);
 
 	return err;
@@ -1159,8 +1298,8 @@ static int __vxlan_fdb_delete(struct vxlan_dev *vxlan,
 			      __be16 port, __be32 src_vni, __be32 vni,
 			      u32 ifindex, bool swdev_notify)
 {
-	struct vxlan_fdb *f;
 	struct vxlan_rdst *rd = NULL;
+	struct vxlan_fdb *f;
 	int err = -ENOENT;
 
 	f = vxlan_find_mac(vxlan, addr, src_vni);
@@ -1195,12 +1334,13 @@ static int vxlan_fdb_delete(struct ndmsg *ndm, struct nlattr *tb[],
 	struct vxlan_dev *vxlan = netdev_priv(dev);
 	union vxlan_addr ip;
 	__be32 src_vni, vni;
-	__be16 port;
-	u32 ifindex;
+	u32 ifindex, nhid;
 	u32 hash_index;
+	__be16 port;
 	int err;
 
-	err = vxlan_fdb_parse(tb, vxlan, &ip, &port, &src_vni, &vni, &ifindex);
+	err = vxlan_fdb_parse(tb, vxlan, &ip, &port, &src_vni, &vni, &ifindex,
+			      &nhid);
 	if (err)
 		return err;
 
@@ -1228,6 +1368,17 @@ static int vxlan_fdb_dump(struct sk_buff *skb, struct netlink_callback *cb,
 		hlist_for_each_entry_rcu(f, &vxlan->fdb_head[h], hlist) {
 			struct vxlan_rdst *rd;
 
+			if (rcu_access_pointer(f->nh)) {
+				err = vxlan_fdb_info(skb, vxlan, f,
+						     NETLINK_CB(cb->skb).portid,
+						     cb->nlh->nlmsg_seq,
+						     RTM_NEWNEIGH,
+						     NLM_F_MULTI, NULL);
+				if (err < 0)
+					goto out;
+				continue;
+			}
+
 			list_for_each_entry_rcu(rd, &f->remotes, list) {
 				if (*idx < cb->args[2])
 					goto skip;
@@ -1311,6 +1462,10 @@ static bool vxlan_snoop(struct net_device *dev,
 		if (f->state & (NUD_PERMANENT | NUD_NOARP))
 			return true;
 
+		/* Don't override an fdb with nexthop with a learnt entry */
+		if (rcu_access_pointer(f->nh))
+			return true;
+
 		if (net_ratelimit())
 			netdev_info(dev,
 				    "%pM migrated from %pIS to %pIS\n",
@@ -1333,7 +1488,7 @@ static bool vxlan_snoop(struct net_device *dev,
 					 vxlan->cfg.dst_port,
 					 vni,
 					 vxlan->default_dst.remote_vni,
-					 ifindex, NTF_SELF, true, NULL);
+					 ifindex, NTF_SELF, 0, true, NULL);
 		spin_unlock(&vxlan->hash_lock[hash_index]);
 	}
 
@@ -2616,6 +2771,38 @@ static void vxlan_xmit_one(struct sk_buff *skb, struct net_device *dev,
 	kfree_skb(skb);
 }
 
+static void vxlan_xmit_nh(struct sk_buff *skb, struct net_device *dev,
+			  struct vxlan_fdb *f, __be32 vni, bool did_rsc)
+{
+	struct vxlan_rdst nh_rdst;
+	struct nexthop *nh;
+	bool do_xmit;
+	u32 hash;
+
+	memset(&nh_rdst, 0, sizeof(struct vxlan_rdst));
+	hash = skb_get_hash(skb);
+
+	rcu_read_lock();
+	nh = rcu_dereference(f->nh);
+	if (!nh) {
+		rcu_read_unlock();
+		goto drop;
+	}
+	do_xmit = vxlan_fdb_nh_path_select(nh, hash, &nh_rdst);
+	rcu_read_unlock();
+
+	if (likely(do_xmit))
+		vxlan_xmit_one(skb, dev, vni, &nh_rdst, did_rsc);
+	else
+		goto drop;
+
+	return;
+
+drop:
+	dev->stats.tx_dropped++;
+	dev_kfree_skb(skb);
+}
+
 /* Transmit local packets over Vxlan
  *
  * Outer IP header inherits ECN and DF from inner header.
@@ -2692,22 +2879,27 @@ static netdev_tx_t vxlan_xmit(struct sk_buff *skb, struct net_device *dev)
 		}
 	}
 
-	list_for_each_entry_rcu(rdst, &f->remotes, list) {
-		struct sk_buff *skb1;
+	if (rcu_access_pointer(f->nh)) {
+		vxlan_xmit_nh(skb, dev, f,
+			      (vni ? : vxlan->default_dst.remote_vni), did_rsc);
+	} else {
+		list_for_each_entry_rcu(rdst, &f->remotes, list) {
+			struct sk_buff *skb1;
 
-		if (!fdst) {
-			fdst = rdst;
-			continue;
+			if (!fdst) {
+				fdst = rdst;
+				continue;
+			}
+			skb1 = skb_clone(skb, GFP_ATOMIC);
+			if (skb1)
+				vxlan_xmit_one(skb1, dev, vni, rdst, did_rsc);
 		}
-		skb1 = skb_clone(skb, GFP_ATOMIC);
-		if (skb1)
-			vxlan_xmit_one(skb1, dev, vni, rdst, did_rsc);
+		if (fdst)
+			vxlan_xmit_one(skb, dev, vni, fdst, did_rsc);
+		else
+			kfree_skb(skb);
 	}
 
-	if (fdst)
-		vxlan_xmit_one(skb, dev, vni, fdst, did_rsc);
-	else
-		kfree_skb(skb);
 	return NETDEV_TX_OK;
 }
 
@@ -3615,7 +3807,7 @@ static int __vxlan_dev_create(struct net *net, struct net_device *dev,
 				       dst->remote_vni,
 				       dst->remote_vni,
 				       dst->remote_ifindex,
-				       NTF_SELF, &f);
+				       NTF_SELF, 0, &f, extack);
 		if (err)
 			return err;
 	}
@@ -4013,7 +4205,7 @@ static int vxlan_changelink(struct net_device *dev, struct nlattr *tb[],
 					       vxlan->cfg.dst_port,
 					       conf.vni, conf.vni,
 					       conf.remote_ifindex,
-					       NTF_SELF, true, extack);
+					       NTF_SELF, 0, true, extack);
 			if (err) {
 				spin_unlock_bh(&vxlan->hash_lock[hash_index]);
 				netdev_adjacent_change_abort(dst->remote_dev,
@@ -4335,7 +4527,7 @@ vxlan_fdb_external_learn_add(struct net_device *dev,
 			       fdb_info->remote_vni,
 			       fdb_info->remote_ifindex,
 			       NTF_USE | NTF_SELF | NTF_EXT_LEARNED,
-			       false, extack);
+			       0, false, extack);
 	spin_unlock_bh(&vxlan->hash_lock[hash_index]);
 
 	return err;
@@ -4410,6 +4602,25 @@ static struct notifier_block vxlan_switchdev_notifier_block __read_mostly = {
 	.notifier_call = vxlan_switchdev_event,
 };
 
+static int vxlan_nexthop_event(struct notifier_block *nb,
+			       unsigned long event, void *ptr)
+{
+	struct nexthop *nh = ptr;
+	struct vxlan_fdb *fdb, *tmp;
+
+	if (!nh || event != NEXTHOP_EVENT_DEL)
+		return NOTIFY_DONE;
+
+	list_for_each_entry_safe(fdb, tmp, &nh->fdb_list, nh_list)
+		vxlan_fdb_destroy(fdb->vdev, fdb, false, false);
+
+	return NOTIFY_DONE;
+}
+
+static struct notifier_block vxlan_nexthop_notifier_block __read_mostly = {
+	.notifier_call = vxlan_nexthop_event,
+};
+
 static __net_init int vxlan_init_net(struct net *net)
 {
 	struct vxlan_net *vn = net_generic(net, vxlan_net_id);
@@ -4421,7 +4632,7 @@ static __net_init int vxlan_init_net(struct net *net)
 	for (h = 0; h < PORT_HASH_SIZE; ++h)
 		INIT_HLIST_HEAD(&vn->sock_list[h]);
 
-	return 0;
+	return register_nexthop_notifier(net, &vxlan_nexthop_notifier_block);
 }
 
 static void vxlan_destroy_tunnels(struct net *net, struct list_head *head)
@@ -4453,6 +4664,8 @@ static void __net_exit vxlan_exit_batch_net(struct list_head *net_list)
 	LIST_HEAD(list);
 
 	rtnl_lock();
+	list_for_each_entry(net, net_list, exit_list)
+		unregister_nexthop_notifier(net, &vxlan_nexthop_notifier_block);
 	list_for_each_entry(net, net_list, exit_list)
 		vxlan_destroy_tunnels(net, &list);
 

include/net/ip6_fib.h

@@ -65,6 +65,7 @@ struct fib6_config {
 	struct nl_info	fc_nlinfo;
 	struct nlattr	*fc_encap;
 	u16		fc_encap_type;
+	bool		fc_is_fdb;
 };
 
 struct fib6_node {

include/net/netns/nexthop.h

@@ -14,5 +14,6 @@ struct netns_nexthop {
 
 	unsigned int		seq;		/* protected by rtnl_mutex */
 	u32			last_id_allocated;
+	struct atomic_notifier_head notifier_chain;
 };
 #endif

include/net/nexthop.h

@@ -10,6 +10,7 @@
 #define __LINUX_NEXTHOP_H
 
 #include <linux/netdevice.h>
+#include <linux/notifier.h>
 #include <linux/route.h>
 #include <linux/types.h>
 #include <net/ip_fib.h>
@@ -26,6 +27,7 @@ struct nh_config {
 	u8		nh_family;
 	u8		nh_protocol;
 	u8		nh_blackhole;
+	u8		nh_fdb;
 	u32		nh_flags;
 
 	int		nh_ifindex;
@@ -52,6 +54,7 @@ struct nh_info {
 
 	u8			family;
 	bool			reject_nh;
+	bool			fdb_nh;
 
 	union {
 		struct fib_nh_common	fib_nhc;
@@ -80,6 +83,7 @@ struct nexthop {
 	struct rb_node		rb_node;    /* entry on netns rbtree */
 	struct list_head	fi_list;    /* v4 entries using nh */
 	struct list_head	f6i_list;   /* v6 entries using nh */
+	struct list_head        fdb_list;   /* fdb entries using this nh */
 	struct list_head	grp_list;   /* nh group entries using this nh */
 	struct net		*net;
 
@@ -88,6 +92,7 @@ struct nexthop {
 	u8			protocol;   /* app managing this nh */
 	u8			nh_flags;
 	bool			is_group;
+	bool			is_fdb_nh;
 
 	refcount_t		refcnt;
 	struct rcu_head		rcu;
@@ -98,6 +103,17 @@ struct nexthop {
 	};
 };
 
+enum nexthop_event_type {
+	NEXTHOP_EVENT_ADD,
+	NEXTHOP_EVENT_DEL
+};
+
+int call_nexthop_notifier(struct notifier_block *nb, struct net *net,
+			  enum nexthop_event_type event_type,
+			  struct nexthop *nh);
+int register_nexthop_notifier(struct net *net, struct notifier_block *nb);
+int unregister_nexthop_notifier(struct net *net, struct notifier_block *nb);
+
 /* caller is holding rcu or rtnl; no reference taken to nexthop */
 struct nexthop *nexthop_find_by_id(struct net *net, u32 id);
 void nexthop_free_rcu(struct rcu_head *head);
@@ -304,4 +320,32 @@ static inline void nexthop_path_fib6_result(struct fib6_result *res, int hash)
 int nexthop_for_each_fib6_nh(struct nexthop *nh,
 			     int (*cb)(struct fib6_nh *nh, void *arg),
 			     void *arg);
+
+static inline int nexthop_get_family(struct nexthop *nh)
+{
+	struct nh_info *nhi = rcu_dereference_rtnl(nh->nh_info);
+
+	return nhi->family;
+}
+
+static inline
+struct fib_nh_common *nexthop_fdb_nhc(struct nexthop *nh)
+{
+	struct nh_info *nhi = rcu_dereference_rtnl(nh->nh_info);
+
+	return &nhi->fib_nhc;
+}
+
+static inline struct fib_nh_common *nexthop_path_fdb_result(struct nexthop *nh,
+							    int hash)
+{
+	struct nh_info *nhi;
+	struct nexthop *nhp;
+
+	nhp = nexthop_select_path(nh, hash);
+	if (unlikely(!nhp))
+		return NULL;
+	nhi = rcu_dereference(nhp->nh_info);
+	return &nhi->fib_nhc;
+}
 #endif

include/net/vxlan.h

@@ -7,6 +7,7 @@
 #include <net/dst_metadata.h>
 #include <net/rtnetlink.h>
 #include <net/switchdev.h>
+#include <net/nexthop.h>
 
 #define IANA_VXLAN_UDP_PORT     4789
 
@@ -487,4 +488,28 @@ static inline void vxlan_flag_attr_error(int attrtype,
 #undef VXLAN_FLAG
 }
 
+static inline bool vxlan_fdb_nh_path_select(struct nexthop *nh,
+					    int hash,
+					    struct vxlan_rdst *rdst)
+{
+	struct fib_nh_common *nhc;
+
+	nhc = nexthop_path_fdb_result(nh, hash);
+	if (unlikely(!nhc))
+		return false;
+
+	switch (nhc->nhc_gw_family) {
+	case AF_INET:
+		rdst->remote_ip.sin.sin_addr.s_addr = nhc->nhc_gw.ipv4;
+		rdst->remote_ip.sa.sa_family = AF_INET;
+		break;
+	case AF_INET6:
+		rdst->remote_ip.sin6.sin6_addr = nhc->nhc_gw.ipv6;
+		rdst->remote_ip.sa.sa_family = AF_INET6;
+		break;
+	}
+
+	return true;
+}
+
 #endif

include/uapi/linux/neighbour.h

@@ -29,6 +29,7 @@ enum {
 	NDA_LINK_NETNSID,
 	NDA_SRC_VNI,
 	NDA_PROTOCOL,  /* Originator of entry */
+	NDA_NH_ID,
 	__NDA_MAX
 };
 

include/uapi/linux/nexthop.h

@@ -49,6 +49,9 @@ enum {
 	NHA_GROUPS,	/* flag; only return nexthop groups in dump */
 	NHA_MASTER,	/* u32;  only return nexthops with given master dev */
 
+	NHA_FDB,	/* flag; nexthop belongs to a bridge fdb */
+	/* if NHA_FDB is added, OIF, BLACKHOLE, ENCAP cannot be set */
+
 	__NHA_MAX,
 };
 

net/core/neighbour.c

@@ -1771,6 +1771,7 @@ static struct neigh_table *neigh_find_table(int family)
 }
 
 const struct nla_policy nda_policy[NDA_MAX+1] = {
+	[NDA_UNSPEC]		= { .strict_start_type = NDA_NH_ID },
 	[NDA_DST]		= { .type = NLA_BINARY, .len = MAX_ADDR_LEN },
 	[NDA_LLADDR]		= { .type = NLA_BINARY, .len = MAX_ADDR_LEN },
 	[NDA_CACHEINFO]		= { .len = sizeof(struct nda_cacheinfo) },
@@ -1781,6 +1782,7 @@ const struct nla_policy nda_policy[NDA_MAX+1] = {
 	[NDA_IFINDEX]		= { .type = NLA_U32 },
 	[NDA_MASTER]		= { .type = NLA_U32 },
 	[NDA_PROTOCOL]		= { .type = NLA_U8 },
+	[NDA_NH_ID]		= { .type = NLA_U32 },
 };
 
 static int neigh_delete(struct sk_buff *skb, struct nlmsghdr *nlh,

net/ipv4/nexthop.c

@@ -33,8 +33,20 @@ static const struct nla_policy rtm_nh_policy[NHA_MAX + 1] = {
 	[NHA_ENCAP]		= { .type = NLA_NESTED },
 	[NHA_GROUPS]		= { .type = NLA_FLAG },
 	[NHA_MASTER]		= { .type = NLA_U32 },
+	[NHA_FDB]		= { .type = NLA_FLAG },
 };
 
+static int call_nexthop_notifiers(struct net *net,
+				  enum fib_event_type event_type,
+				  struct nexthop *nh)
+{
+	int err;
+
+	err = atomic_notifier_call_chain(&net->nexthop.notifier_chain,
+					 event_type, nh);
+	return notifier_to_errno(err);
+}
+
 static unsigned int nh_dev_hashfn(unsigned int val)
 {
 	unsigned int mask = NH_DEV_HASHSIZE - 1;
@@ -107,6 +119,7 @@ static struct nexthop *nexthop_alloc(void)
 		INIT_LIST_HEAD(&nh->fi_list);
 		INIT_LIST_HEAD(&nh->f6i_list);
 		INIT_LIST_HEAD(&nh->grp_list);
+		INIT_LIST_HEAD(&nh->fdb_list);
 	}
 	return nh;
 }
@@ -227,6 +240,9 @@ static int nh_fill_node(struct sk_buff *skb, struct nexthop *nh,
 	if (nla_put_u32(skb, NHA_ID, nh->id))
 		goto nla_put_failure;
 
+	if (nh->is_fdb_nh && nla_put_flag(skb, NHA_FDB))
+		goto nla_put_failure;
+
 	if (nh->is_group) {
 		struct nh_group *nhg = rtnl_dereference(nh->nh_grp);
 
@@ -241,7 +257,7 @@ static int nh_fill_node(struct sk_buff *skb, struct nexthop *nh,
 		if (nla_put_flag(skb, NHA_BLACKHOLE))
 			goto nla_put_failure;
 		goto out;
-	} else {
+	} else if (!nh->is_fdb_nh) {
 		const struct net_device *dev;
 
 		dev = nhi->fib_nhc.nhc_dev;
@@ -387,12 +403,35 @@ static bool valid_group_nh(struct nexthop *nh, unsigned int npaths,
 	return true;
 }
 
+static int nh_check_attr_fdb_group(struct nexthop *nh, u8 *nh_family,
+				   struct netlink_ext_ack *extack)
+{
+	struct nh_info *nhi;
+
+	if (!nh->is_fdb_nh) {
+		NL_SET_ERR_MSG(extack, "FDB nexthop group can only have fdb nexthops");
+		return -EINVAL;
+	}
+
+	nhi = rtnl_dereference(nh->nh_info);
+	if (*nh_family == AF_UNSPEC) {
+		*nh_family = nhi->family;
+	} else if (*nh_family != nhi->family) {
+		NL_SET_ERR_MSG(extack, "FDB nexthop group cannot have mixed family nexthops");
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
 static int nh_check_attr_group(struct net *net, struct nlattr *tb[],
 			       struct netlink_ext_ack *extack)
 {
 	unsigned int len = nla_len(tb[NHA_GROUP]);
+	u8 nh_family = AF_UNSPEC;
 	struct nexthop_grp *nhg;
 	unsigned int i, j;
+	u8 nhg_fdb = 0;
 
 	if (len & (sizeof(struct nexthop_grp) - 1)) {
 		NL_SET_ERR_MSG(extack,
@@ -421,6 +460,8 @@ static int nh_check_attr_group(struct net *net, struct nlattr *tb[],
 		}
 	}
 
+	if (tb[NHA_FDB])
+		nhg_fdb = 1;
 	nhg = nla_data(tb[NHA_GROUP]);
 	for (i = 0; i < len; ++i) {
 		struct nexthop *nh;
@@ -432,11 +473,20 @@ static int nh_check_attr_group(struct net *net, struct nlattr *tb[],
 		}
 		if (!valid_group_nh(nh, len, extack))
 			return -EINVAL;
+
+		if (nhg_fdb && nh_check_attr_fdb_group(nh, &nh_family, extack))
+			return -EINVAL;
+
+		if (!nhg_fdb && nh->is_fdb_nh) {
+			NL_SET_ERR_MSG(extack, "Non FDB nexthop group cannot have fdb nexthops");
+			return -EINVAL;
+		}
 	}
 	for (i = NHA_GROUP + 1; i < __NHA_MAX; ++i) {
 		if (!tb[i])
 			continue;
-
+		if (tb[NHA_FDB])
+			continue;
 		NL_SET_ERR_MSG(extack,
 			       "No other attributes can be set in nexthop groups");
 		return -EINVAL;
@@ -495,6 +545,9 @@ struct nexthop *nexthop_select_path(struct nexthop *nh, int hash)
 		if (hash > atomic_read(&nhge->upper_bound))
 			continue;
 
+		if (nhge->nh->is_fdb_nh)
+			return nhge->nh;
+
 		/* nexthops always check if it is good and does
 		 * not rely on a sysctl for this behavior
 		 */
@@ -564,6 +617,11 @@ int fib6_check_nexthop(struct nexthop *nh, struct fib6_config *cfg,
 {
 	struct nh_info *nhi;
 
+	if (nh->is_fdb_nh) {
+		NL_SET_ERR_MSG(extack, "Route cannot point to a fdb nexthop");
+		return -EINVAL;
+	}
+
 	/* fib6_src is unique to a fib6_info and limits the ability to cache
 	 * routes in fib6_nh within a nexthop that is potentially shared
 	 * across multiple fib entries. If the config wants to use source
@@ -640,6 +698,12 @@ int fib_check_nexthop(struct nexthop *nh, u8 scope,
 {
 	int err = 0;
 
+	if (nh->is_fdb_nh) {
+		NL_SET_ERR_MSG(extack, "Route cannot point to a fdb nexthop");
+		err = -EINVAL;
+		goto out;
+	}
+
 	if (nh->is_group) {
 		struct nh_group *nhg;
 
@@ -773,6 +837,8 @@ static void __remove_nexthop_fib(struct net *net, struct nexthop *nh)
 	bool do_flush = false;
 	struct fib_info *fi;
 
+	call_nexthop_notifiers(net, NEXTHOP_EVENT_DEL, nh);
+
 	list_for_each_entry(fi, &nh->fi_list, nh_list) {
 		fi->fib_flags |= RTNH_F_DEAD;
 		do_flush = true;
@@ -1125,6 +1191,9 @@ static struct nexthop *nexthop_create_group(struct net *net,
 		nh_group_rebalance(nhg);
 	}
 
+	if (cfg->nh_fdb)
+		nh->is_fdb_nh = 1;
+
 	rcu_assign_pointer(nh->nh_grp, nhg);
 
 	return nh;
@@ -1152,7 +1221,7 @@ static int nh_create_ipv4(struct net *net, struct nexthop *nh,
 		.fc_encap = cfg->nh_encap,
 		.fc_encap_type = cfg->nh_encap_type,
 	};
-	u32 tb_id = l3mdev_fib_table(cfg->dev);
+	u32 tb_id = (cfg->dev ? l3mdev_fib_table(cfg->dev) : RT_TABLE_MAIN);
 	int err;
 
 	err = fib_nh_init(net, fib_nh, &fib_cfg, 1, extack);
@@ -1161,6 +1230,9 @@ static int nh_create_ipv4(struct net *net, struct nexthop *nh,
 		goto out;
 	}
 
+	if (nh->is_fdb_nh)
+		goto out;
+
 	/* sets nh_dev if successful */
 	err = fib_check_nh(net, fib_nh, tb_id, 0, extack);
 	if (!err) {
@@ -1186,6 +1258,7 @@ static int nh_create_ipv6(struct net *net,  struct nexthop *nh,
 		.fc_flags = cfg->nh_flags,
 		.fc_encap = cfg->nh_encap,
 		.fc_encap_type = cfg->nh_encap_type,
+		.fc_is_fdb = cfg->nh_fdb,
 	};
 	int err;
 
@@ -1227,6 +1300,9 @@ static struct nexthop *nexthop_create(struct net *net, struct nh_config *cfg,
 	nhi->family = cfg->nh_family;
 	nhi->fib_nhc.nhc_scope = RT_SCOPE_LINK;
 
+	if (cfg->nh_fdb)
+		nh->is_fdb_nh = 1;
+
 	if (cfg->nh_blackhole) {
 		nhi->reject_nh = 1;
 		cfg->nh_ifindex = net->loopback_dev->ifindex;
@@ -1248,7 +1324,8 @@ static struct nexthop *nexthop_create(struct net *net, struct nh_config *cfg,
 	}
 
 	/* add the entry to the device based hash */
-	nexthop_devhash_add(net, nhi);
+	if (!nh->is_fdb_nh)
+		nexthop_devhash_add(net, nhi);
 
 	rcu_assign_pointer(nh->nh_info, nhi);
 
@@ -1352,6 +1429,19 @@ static int rtm_to_nh_config(struct net *net, struct sk_buff *skb,
 	if (tb[NHA_ID])
 		cfg->nh_id = nla_get_u32(tb[NHA_ID]);
 
+	if (tb[NHA_FDB]) {
+		if (tb[NHA_OIF] || tb[NHA_BLACKHOLE] ||
+		    tb[NHA_ENCAP]   || tb[NHA_ENCAP_TYPE]) {
+			NL_SET_ERR_MSG(extack, "Fdb attribute can not be used with encap, oif or blackhole");
+			goto out;
+		}
+		if (nhm->nh_flags) {
+			NL_SET_ERR_MSG(extack, "Unsupported nexthop flags in ancillary header");
+			goto out;
+		}
+		cfg->nh_fdb = nla_get_flag(tb[NHA_FDB]);
+	}
+
 	if (tb[NHA_GROUP]) {
 		if (nhm->nh_family != AF_UNSPEC) {
 			NL_SET_ERR_MSG(extack, "Invalid family for group");
@@ -1375,8 +1465,8 @@ static int rtm_to_nh_config(struct net *net, struct sk_buff *skb,
 
 	if (tb[NHA_BLACKHOLE]) {
 		if (tb[NHA_GATEWAY] || tb[NHA_OIF] ||
-		    tb[NHA_ENCAP]   || tb[NHA_ENCAP_TYPE]) {
-			NL_SET_ERR_MSG(extack, "Blackhole attribute can not be used with gateway or oif");
+		    tb[NHA_ENCAP]   || tb[NHA_ENCAP_TYPE] || tb[NHA_FDB]) {
+			NL_SET_ERR_MSG(extack, "Blackhole attribute can not be used with gateway, oif, encap or fdb");
 			goto out;
 		}
 
@@ -1385,26 +1475,28 @@ static int rtm_to_nh_config(struct net *net, struct sk_buff *skb,
 		goto out;
 	}
 
-	if (!tb[NHA_OIF]) {
-		NL_SET_ERR_MSG(extack, "Device attribute required for non-blackhole nexthops");
+	if (!cfg->nh_fdb && !tb[NHA_OIF]) {
+		NL_SET_ERR_MSG(extack, "Device attribute required for non-blackhole and non-fdb nexthops");
 		goto out;
 	}
 
-	cfg->nh_ifindex = nla_get_u32(tb[NHA_OIF]);
-	if (cfg->nh_ifindex)
-		cfg->dev = __dev_get_by_index(net, cfg->nh_ifindex);
+	if (!cfg->nh_fdb && tb[NHA_OIF]) {
+		cfg->nh_ifindex = nla_get_u32(tb[NHA_OIF]);
+		if (cfg->nh_ifindex)
+			cfg->dev = __dev_get_by_index(net, cfg->nh_ifindex);
 
-	if (!cfg->dev) {
-		NL_SET_ERR_MSG(extack, "Invalid device index");
-		goto out;
-	} else if (!(cfg->dev->flags & IFF_UP)) {
-		NL_SET_ERR_MSG(extack, "Nexthop device is not up");
-		err = -ENETDOWN;
-		goto out;
-	} else if (!netif_carrier_ok(cfg->dev)) {
-		NL_SET_ERR_MSG(extack, "Carrier for nexthop device is down");
-		err = -ENETDOWN;
-		goto out;
+		if (!cfg->dev) {
+			NL_SET_ERR_MSG(extack, "Invalid device index");
+			goto out;
+		} else if (!(cfg->dev->flags & IFF_UP)) {
+			NL_SET_ERR_MSG(extack, "Nexthop device is not up");
+			err = -ENETDOWN;
+			goto out;
+		} else if (!netif_carrier_ok(cfg->dev)) {
+			NL_SET_ERR_MSG(extack, "Carrier for nexthop device is down");
+			err = -ENETDOWN;
+			goto out;
+		}
 	}
 
 	err = -EINVAL;
@@ -1633,7 +1725,7 @@ static bool nh_dump_filtered(struct nexthop *nh, int dev_idx, int master_idx,
 
 static int nh_valid_dump_req(const struct nlmsghdr *nlh, int *dev_idx,
 			     int *master_idx, bool *group_filter,
-			     struct netlink_callback *cb)
+			     bool *fdb_filter, struct netlink_callback *cb)
 {
 	struct netlink_ext_ack *extack = cb->extack;
 	struct nlattr *tb[NHA_MAX + 1];
@@ -1670,6 +1762,9 @@ static int nh_valid_dump_req(const struct nlmsghdr *nlh, int *dev_idx,
 		case NHA_GROUPS:
 			*group_filter = true;
 			break;
+		case NHA_FDB:
+			*fdb_filter = true;
+			break;
 		default:
 			NL_SET_ERR_MSG(extack, "Unsupported attribute in dump request");
 			return -EINVAL;
@@ -1688,17 +1783,17 @@ static int nh_valid_dump_req(const struct nlmsghdr *nlh, int *dev_idx,
 /* rtnl */
 static int rtm_dump_nexthop(struct sk_buff *skb, struct netlink_callback *cb)
 {
+	bool group_filter = false, fdb_filter = false;
 	struct nhmsg *nhm = nlmsg_data(cb->nlh);
 	int dev_filter_idx = 0, master_idx = 0;
 	struct net *net = sock_net(skb->sk);
 	struct rb_root *root = &net->nexthop.rb_root;
-	bool group_filter = false;
 	struct rb_node *node;
 	int idx = 0, s_idx;
 	int err;
 
 	err = nh_valid_dump_req(cb->nlh, &dev_filter_idx, &master_idx,
-				&group_filter, cb);
+				&group_filter, &fdb_filter, cb);
 	if (err < 0)
 		return err;
 
@@ -1783,6 +1878,19 @@ static struct notifier_block nh_netdev_notifier = {
 	.notifier_call = nh_netdev_event,
 };
 
+int register_nexthop_notifier(struct net *net, struct notifier_block *nb)
+{
+	return atomic_notifier_chain_register(&net->nexthop.notifier_chain, nb);
+}
+EXPORT_SYMBOL(register_nexthop_notifier);
+
+int unregister_nexthop_notifier(struct net *net, struct notifier_block *nb)
+{
+	return atomic_notifier_chain_unregister(&net->nexthop.notifier_chain,
+						nb);
+}
+EXPORT_SYMBOL(unregister_nexthop_notifier);
+
 static void __net_exit nexthop_net_exit(struct net *net)
 {
 	rtnl_lock();
@@ -1799,6 +1907,7 @@ static int __net_init nexthop_net_init(struct net *net)
 	net->nexthop.devhash = kzalloc(sz, GFP_KERNEL);
 	if (!net->nexthop.devhash)
 		return -ENOMEM;
+	ATOMIC_INIT_NOTIFIER_HEAD(&net->nexthop.notifier_chain);
 
 	return 0;
 }

net/ipv6/route.c

@@ -3421,6 +3421,11 @@ int fib6_nh_init(struct net *net, struct fib6_nh *fib6_nh,
 #ifdef CONFIG_IPV6_ROUTER_PREF
 	fib6_nh->last_probe = jiffies;
 #endif
+	if (cfg->fc_is_fdb) {
+		fib6_nh->fib_nh_gw6 = cfg->fc_gateway;
+		fib6_nh->fib_nh_gw_family = AF_INET6;
+		return 0;
+	}
 
 	err = -ENODEV;
 	if (cfg->fc_ifindex) {

tools/testing/selftests/net/fib_nexthops.sh

@@ -19,8 +19,8 @@ ret=0
 ksft_skip=4
 
 # all tests in this script. Can be overridden with -t option
-IPV4_TESTS="ipv4_fcnal ipv4_grp_fcnal ipv4_withv6_fcnal ipv4_fcnal_runtime ipv4_compat_mode"
-IPV6_TESTS="ipv6_fcnal ipv6_grp_fcnal ipv6_fcnal_runtime ipv6_compat_mode"
+IPV4_TESTS="ipv4_fcnal ipv4_grp_fcnal ipv4_withv6_fcnal ipv4_fcnal_runtime ipv4_compat_mode ipv4_fdb_grp_fcnal"
+IPV6_TESTS="ipv6_fcnal ipv6_grp_fcnal ipv6_fcnal_runtime ipv6_compat_mode ipv6_fdb_grp_fcnal"
 
 ALL_TESTS="basic ${IPV4_TESTS} ${IPV6_TESTS}"
 TESTS="${ALL_TESTS}"
@@ -146,6 +146,7 @@ setup()
 	create_ns remote
 
 	IP="ip -netns me"
+	BRIDGE="bridge -netns me"
 	set -e
 	$IP li add veth1 type veth peer name veth2
 	$IP li set veth1 up
@@ -280,6 +281,161 @@ stop_ip_monitor()
 	return $rc
 }
 
+check_nexthop_fdb_support()
+{
+	$IP nexthop help 2>&1 | grep -q fdb
+	if [ $? -ne 0 ]; then
+		echo "SKIP: iproute2 too old, missing fdb nexthop support"
+		return $ksft_skip
+	fi
+}
+
+ipv6_fdb_grp_fcnal()
+{
+	local rc
+
+	echo
+	echo "IPv6 fdb groups functional"
+	echo "--------------------------"
+
+	check_nexthop_fdb_support
+	if [ $? -eq $ksft_skip ]; then
+		return $ksft_skip
+	fi
+
+	# create group with multiple nexthops
+	run_cmd "$IP nexthop add id 61 via 2001:db8:91::2 fdb"
+	run_cmd "$IP nexthop add id 62 via 2001:db8:91::3 fdb"
+	run_cmd "$IP nexthop add id 102 group 61/62 fdb"
+	check_nexthop "id 102" "id 102 group 61/62 fdb"
+	log_test $? 0 "Fdb Nexthop group with multiple nexthops"
+
+	## get nexthop group
+	run_cmd "$IP nexthop get id 102"
+	check_nexthop "id 102" "id 102 group 61/62 fdb"
+	log_test $? 0 "Get Fdb nexthop group by id"
+
+	# fdb nexthop group can only contain fdb nexthops
+	run_cmd "$IP nexthop add id 63 via 2001:db8:91::4"
+	run_cmd "$IP nexthop add id 64 via 2001:db8:91::5"
+	run_cmd "$IP nexthop add id 103 group 63/64 fdb"
+	log_test $? 2 "Fdb Nexthop group with non-fdb nexthops"
+
+	# Non fdb nexthop group can not contain fdb nexthops
+	run_cmd "$IP nexthop add id 65 via 2001:db8:91::5 fdb"
+	run_cmd "$IP nexthop add id 66 via 2001:db8:91::6 fdb"
+	run_cmd "$IP nexthop add id 104 group 65/66"
+	log_test $? 2 "Non-Fdb Nexthop group with fdb nexthops"
+
+	# fdb nexthop cannot have blackhole
+	run_cmd "$IP nexthop add id 67 blackhole fdb"
+	log_test $? 2 "Fdb Nexthop with blackhole"
+
+	# fdb nexthop with oif
+	run_cmd "$IP nexthop add id 68 via 2001:db8:91::7 dev veth1 fdb"
+	log_test $? 2 "Fdb Nexthop with oif"
+
+	# fdb nexthop with onlink
+	run_cmd "$IP nexthop add id 68 via 2001:db8:91::7 onlink fdb"
+	log_test $? 2 "Fdb Nexthop with onlink"
+
+	# fdb nexthop with encap
+	run_cmd "$IP nexthop add id 69 encap mpls 101 via 2001:db8:91::8 dev veth1 fdb"
+	log_test $? 2 "Fdb Nexthop with encap"
+
+	run_cmd "$IP link add name vx10 type vxlan id 1010 local 2001:db8:91::9 remote 2001:db8:91::10 dstport 4789 nolearning noudpcsum tos inherit ttl 100"
+	run_cmd "$BRIDGE fdb add 02:02:00:00:00:13 dev vx10 nhid 102 self"
+	log_test $? 0 "Fdb mac add with nexthop group"
+
+	## fdb nexthops can only reference nexthop groups and not nexthops
+	run_cmd "$BRIDGE fdb add 02:02:00:00:00:14 dev vx10 nhid 61 self"
+	log_test $? 255 "Fdb mac add with nexthop"
+
+	run_cmd "$IP -6 ro add 2001:db8:101::1/128 nhid 66"
+	log_test $? 2 "Route add with fdb nexthop"
+
+	run_cmd "$IP -6 ro add 2001:db8:101::1/128 nhid 103"
+	log_test $? 2 "Route add with fdb nexthop group"
+
+	run_cmd "$IP nexthop del id 102"
+	log_test $? 0 "Fdb nexthop delete"
+
+	$IP link del dev vx10
+}
+
+ipv4_fdb_grp_fcnal()
+{
+	local rc
+
+	echo
+	echo "IPv4 fdb groups functional"
+	echo "--------------------------"
+
+	check_nexthop_fdb_support
+	if [ $? -eq $ksft_skip ]; then
+		return $ksft_skip
+	fi
+
+	# create group with multiple nexthops
+	run_cmd "$IP nexthop add id 12 via 172.16.1.2 fdb"
+	run_cmd "$IP nexthop add id 13 via 172.16.1.3 fdb"
+	run_cmd "$IP nexthop add id 102 group 12/13 fdb"
+	check_nexthop "id 102" "id 102 group 12/13 fdb"
+	log_test $? 0 "Fdb Nexthop group with multiple nexthops"
+
+	# get nexthop group
+	run_cmd "$IP nexthop get id 102"
+	check_nexthop "id 102" "id 102 group 12/13 fdb"
+	log_test $? 0 "Get Fdb nexthop group by id"
+
+	# fdb nexthop group can only contain fdb nexthops
+	run_cmd "$IP nexthop add id 14 via 172.16.1.2"
+	run_cmd "$IP nexthop add id 15 via 172.16.1.3"
+	run_cmd "$IP nexthop add id 103 group 14/15 fdb"
+	log_test $? 2 "Fdb Nexthop group with non-fdb nexthops"
+
+	# Non fdb nexthop group can not contain fdb nexthops
+	run_cmd "$IP nexthop add id 16 via 172.16.1.2 fdb"
+	run_cmd "$IP nexthop add id 17 via 172.16.1.3 fdb"
+	run_cmd "$IP nexthop add id 104 group 14/15"
+	log_test $? 2 "Non-Fdb Nexthop group with fdb nexthops"
+
+	# fdb nexthop cannot have blackhole
+	run_cmd "$IP nexthop add id 18 blackhole fdb"
+	log_test $? 2 "Fdb Nexthop with blackhole"
+
+	# fdb nexthop with oif
+	run_cmd "$IP nexthop add id 16 via 172.16.1.2 dev veth1 fdb"
+	log_test $? 2 "Fdb Nexthop with oif"
+
+	# fdb nexthop with onlink
+	run_cmd "$IP nexthop add id 16 via 172.16.1.2 onlink fdb"
+	log_test $? 2 "Fdb Nexthop with onlink"
+
+	# fdb nexthop with encap
+	run_cmd "$IP nexthop add id 17 encap mpls 101 via 172.16.1.2 dev veth1 fdb"
+	log_test $? 2 "Fdb Nexthop with encap"
+
+	run_cmd "$IP link add name vx10 type vxlan id 1010 local 10.0.0.1 remote 10.0.0.2 dstport 4789 nolearning noudpcsum tos inherit ttl 100"
+	run_cmd "$BRIDGE fdb add 02:02:00:00:00:13 dev vx10 nhid 102 self"
+	log_test $? 0 "Fdb mac add with nexthop group"
+
+	# fdb nexthops can only reference nexthop groups and not nexthops
+	run_cmd "$BRIDGE fdb add 02:02:00:00:00:14 dev vx10 nhid 12 self"
+	log_test $? 255 "Fdb mac add with nexthop"
+
+	run_cmd "$IP ro add 172.16.0.0/22 nhid 15"
+	log_test $? 2 "Route add with fdb nexthop"
+
+	run_cmd "$IP ro add 172.16.0.0/22 nhid 103"
+	log_test $? 2 "Route add with fdb nexthop group"
+
+	run_cmd "$IP nexthop del id 102"
+	log_test $? 0 "Fdb nexthop delete"
+
+	$IP link del dev vx10
+}
+
 ################################################################################
 # basic operations (add, delete, replace) on nexthops and nexthop groups
 #