Merge branch 'net-mctp-add-support-for-gateway-routing'

Jeremy Kerr says:

====================
net: mctp: Add support for gateway routing

This series adds a gateway route type for the MCTP core, allowing
non-local EIDs as the match for a route.

Example setup using the mctp tools:

    mctp route add 9 via mctpi2c0
    mctp neigh add 9 dev mctpi2c0 lladdr 0x1d
    mctp route add 10 gw 9

- will route packets to eid 10 through mctpi2c0, using a dest lladdr
of 0x1d (ie, that of the directly-attached eid 9).

The core change to support this is the introduction of a struct
mctp_dst, which represents the result of a route lookup. Since this
involves a bit of surgery through the routing code, we add a few tests
along the way.

We're introducing an ABI change in the new RTM_{NEW,GET,DEL}ROUTE
netlink formats, with the support for a RTA_GATEWAY attribute. Because
we need a network ID specified to fully-qualify a gateway EID, the
RTA_GATEWAY attribute carries the (net, eid) tuple in full:

    struct mctp_fq_addr {
        unsigned int net;
        mctp_eid_t eid;
    }

Of course, any questions, comments etc are most welcome.

Signed-off-by: Jeremy Kerr <jk@codeconstruct.com.au>
====================

Link: https://patch.msgid.link/20250702-dev-forwarding-v5-0-1468191da8a4@codeconstruct.com.au
Signed-off-by: Paolo Abeni <pabeni@redhat.com>

8 files changed, 1324 insertions, 457 deletions

diff --git a/include/net/mctp.h b/include/net/mctp.h
index 07d458990113..ac4f4ecdfc24 100644
--- a/include/net/mctp.h
+++ b/include/net/mctp.h
@@ -183,8 +183,8 @@ struct mctp_sk_key {
 struct mctp_skb_cb {
 	unsigned int	magic;
 	unsigned int	net;
-	int		ifindex; /* extended/direct addressing if set */
-	mctp_eid_t	src;
+	/* fields below provide extended addressing for ingress to recvmsg() */
+	int		ifindex;
 	unsigned char	halen;
 	unsigned char	haddr[MAX_ADDR_LEN];
 };
@@ -222,6 +222,8 @@ struct mctp_flow {
 	struct mctp_sk_key *key;
 };
 
+struct mctp_dst;
+
 /* Route definition.
  *
  * These are held in the pernet->mctp.routes list, with RCU protection for
@@ -229,16 +231,25 @@ struct mctp_flow {
  * dropped on NETDEV_UNREGISTER events.
  *
  * Updates to the route table are performed under rtnl; all reads under RCU,
- * so routes cannot be referenced over a RCU grace period. Specifically: A
- * caller cannot block between mctp_route_lookup and mctp_route_release()
+ * so routes cannot be referenced over a RCU grace period.
  */
 struct mctp_route {
 	mctp_eid_t		min, max;
 
 	unsigned char		type;
+
 	unsigned int		mtu;
-	struct mctp_dev		*dev;
-	int			(*output)(struct mctp_route *route,
+
+	enum {
+		MCTP_ROUTE_DIRECT,
+		MCTP_ROUTE_GATEWAY,
+	} dst_type;
+	union {
+		struct mctp_dev	*dev;
+		struct mctp_fq_addr gateway;
+	};
+
+	int			(*output)(struct mctp_dst *dst,
 					  struct sk_buff *skb);
 
 	struct list_head	list;
@@ -246,12 +257,35 @@ struct mctp_route {
 	struct rcu_head		rcu;
 };
 
+/* Route lookup result: dst. Represents the results of a routing decision,
+ * but is only held over the individual routing operation.
+ *
+ * Will typically be stored on the caller stack, and must be released after
+ * usage.
+ */
+struct mctp_dst {
+	struct mctp_dev *dev;
+	unsigned int mtu;
+	mctp_eid_t nexthop;
+
+	/* set for direct addressing */
+	unsigned char halen;
+	unsigned char haddr[MAX_ADDR_LEN];
+
+	int (*output)(struct mctp_dst *dst, struct sk_buff *skb);
+};
+
+int mctp_dst_from_extaddr(struct mctp_dst *dst, struct net *net, int ifindex,
+			  unsigned char halen, const unsigned char *haddr);
+
 /* route interfaces */
-struct mctp_route *mctp_route_lookup(struct net *net, unsigned int dnet,
-				     mctp_eid_t daddr);
+int mctp_route_lookup(struct net *net, unsigned int dnet,
+		      mctp_eid_t daddr, struct mctp_dst *dst);
+
+void mctp_dst_release(struct mctp_dst *dst);
 
 /* always takes ownership of skb */
-int mctp_local_output(struct sock *sk, struct mctp_route *rt,
+int mctp_local_output(struct sock *sk, struct mctp_dst *dst,
 		      struct sk_buff *skb, mctp_eid_t daddr, u8 req_tag);
 
 void mctp_key_unref(struct mctp_sk_key *key);
diff --git a/include/uapi/linux/mctp.h b/include/uapi/linux/mctp.h
index e1db65df9359..19ad12a0cd4b 100644
--- a/include/uapi/linux/mctp.h
+++ b/include/uapi/linux/mctp.h
@@ -37,6 +37,14 @@ struct sockaddr_mctp_ext {
 	__u8			smctp_haddr[MAX_ADDR_LEN];
 };
 
+/* A "fully qualified" MCTP address, which includes the system-local network ID,
+ * required to uniquely resolve a routable EID.
+ */
+struct mctp_fq_addr {
+	unsigned int	net;
+	mctp_eid_t	eid;
+};
+
 #define MCTP_NET_ANY		0x0
 
 #define MCTP_ADDR_NULL		0x00
diff --git a/net/mctp/af_mctp.c b/net/mctp/af_mctp.c
index 9b12ca97f412..aef74308c18e 100644
--- a/net/mctp/af_mctp.c
+++ b/net/mctp/af_mctp.c
@@ -97,8 +97,8 @@ static int mctp_sendmsg(struct socket *sock, struct msghdr *msg, size_t len)
 	struct sock *sk = sock->sk;
 	struct mctp_sock *msk = container_of(sk, struct mctp_sock, sk);
 	struct mctp_skb_cb *cb;
-	struct mctp_route *rt;
 	struct sk_buff *skb = NULL;
+	struct mctp_dst dst;
 	int hlen;
 
 	if (addr) {
@@ -133,34 +133,30 @@ static int mctp_sendmsg(struct socket *sock, struct msghdr *msg, size_t len)
 	if (msk->addr_ext && addrlen >= sizeof(struct sockaddr_mctp_ext)) {
 		DECLARE_SOCKADDR(struct sockaddr_mctp_ext *,
 				 extaddr, msg->msg_name);
-		struct net_device *dev;
-
-		rc = -EINVAL;
-		rcu_read_lock();
-		dev = dev_get_by_index_rcu(sock_net(sk), extaddr->smctp_ifindex);
-		/* check for correct halen */
-		if (dev && extaddr->smctp_halen == dev->addr_len) {
-			hlen = LL_RESERVED_SPACE(dev) + sizeof(struct mctp_hdr);
-			rc = 0;
-		}
-		rcu_read_unlock();
+
+		if (!mctp_sockaddr_ext_is_ok(extaddr))
+			return -EINVAL;
+
+		rc = mctp_dst_from_extaddr(&dst, sock_net(sk),
+					   extaddr->smctp_ifindex,
+					   extaddr->smctp_halen,
+					   extaddr->smctp_haddr);
 		if (rc)
-			goto err_free;
-		rt = NULL;
+			return rc;
+
 	} else {
-		rt = mctp_route_lookup(sock_net(sk), addr->smctp_network,
-				       addr->smctp_addr.s_addr);
-		if (!rt) {
-			rc = -EHOSTUNREACH;
-			goto err_free;
-		}
-		hlen = LL_RESERVED_SPACE(rt->dev->dev) + sizeof(struct mctp_hdr);
+		rc = mctp_route_lookup(sock_net(sk), addr->smctp_network,
+				       addr->smctp_addr.s_addr, &dst);
+		if (rc)
+			return rc;
 	}
 
+	hlen = LL_RESERVED_SPACE(dst.dev->dev) + sizeof(struct mctp_hdr);
+
 	skb = sock_alloc_send_skb(sk, hlen + 1 + len,
 				  msg->msg_flags & MSG_DONTWAIT, &rc);
 	if (!skb)
-		return rc;
+		goto err_release_dst;
 
 	skb_reserve(skb, hlen);
 
@@ -175,30 +171,16 @@ static int mctp_sendmsg(struct socket *sock, struct msghdr *msg, size_t len)
 	cb = __mctp_cb(skb);
 	cb->net = addr->smctp_network;
 
-	if (!rt) {
-		/* fill extended address in cb */
-		DECLARE_SOCKADDR(struct sockaddr_mctp_ext *,
-				 extaddr, msg->msg_name);
-
-		if (!mctp_sockaddr_ext_is_ok(extaddr) ||
-		    extaddr->smctp_halen > sizeof(cb->haddr)) {
-			rc = -EINVAL;
-			goto err_free;
-		}
-
-		cb->ifindex = extaddr->smctp_ifindex;
-		/* smctp_halen is checked above */
-		cb->halen = extaddr->smctp_halen;
-		memcpy(cb->haddr, extaddr->smctp_haddr, cb->halen);
-	}
-
-	rc = mctp_local_output(sk, rt, skb, addr->smctp_addr.s_addr,
+	rc = mctp_local_output(sk, &dst, skb, addr->smctp_addr.s_addr,
 			       addr->smctp_tag);
 
+	mctp_dst_release(&dst);
 	return rc ? : len;
 
 err_free:
 	kfree_skb(skb);
+err_release_dst:
+	mctp_dst_release(&dst);
 	return rc;
 }
 
@@ -793,3 +775,7 @@ MODULE_DESCRIPTION("MCTP core");
 MODULE_AUTHOR("Jeremy Kerr <jk@codeconstruct.com.au>");
 
 MODULE_ALIAS_NETPROTO(PF_MCTP);
+
+#if IS_ENABLED(CONFIG_MCTP_TEST)
+#include "test/sock-test.c"
+#endif
diff --git a/net/mctp/route.c b/net/mctp/route.c
index d9c8e5a5f9ce..a20d6b11d418 100644
--- a/net/mctp/route.c
+++ b/net/mctp/route.c
@@ -17,6 +17,8 @@
 #include <linux/rtnetlink.h>
 #include <linux/skbuff.h>
 
+#include <kunit/static_stub.h>
+
 #include <uapi/linux/if_arp.h>
 
 #include <net/mctp.h>
@@ -32,7 +34,7 @@ static const unsigned long mctp_key_lifetime = 6 * CONFIG_HZ;
 static void mctp_flow_prepare_output(struct sk_buff *skb, struct mctp_dev *dev);
 
 /* route output callbacks */
-static int mctp_route_discard(struct mctp_route *route, struct sk_buff *skb)
+static int mctp_dst_discard(struct mctp_dst *dst, struct sk_buff *skb)
 {
 	kfree_skb(skb);
 	return 0;
@@ -368,7 +370,7 @@ static int mctp_frag_queue(struct mctp_sk_key *key, struct sk_buff *skb)
 	return 0;
 }
 
-static int mctp_route_input(struct mctp_route *route, struct sk_buff *skb)
+static int mctp_dst_input(struct mctp_dst *dst, struct sk_buff *skb)
 {
 	struct mctp_sk_key *key, *any_key = NULL;
 	struct net *net = dev_net(skb->dev);
@@ -392,6 +394,9 @@ static int mctp_route_input(struct mctp_route *route, struct sk_buff *skb)
 	 */
 	skb_orphan(skb);
 
+	if (skb->pkt_type == PACKET_OUTGOING)
+		skb->pkt_type = PACKET_LOOPBACK;
+
 	/* ensure we have enough data for a header and a type */
 	if (skb->len < sizeof(struct mctp_hdr) + 1)
 		goto out;
@@ -556,39 +561,31 @@ out:
 	return rc;
 }
 
-static unsigned int mctp_route_mtu(struct mctp_route *rt)
+static int mctp_dst_output(struct mctp_dst *dst, struct sk_buff *skb)
 {
-	return rt->mtu ?: READ_ONCE(rt->dev->dev->mtu);
-}
-
-static int mctp_route_output(struct mctp_route *route, struct sk_buff *skb)
-{
-	struct mctp_skb_cb *cb = mctp_cb(skb);
-	struct mctp_hdr *hdr = mctp_hdr(skb);
 	char daddr_buf[MAX_ADDR_LEN];
 	char *daddr = NULL;
-	unsigned int mtu;
 	int rc;
 
 	skb->protocol = htons(ETH_P_MCTP);
+	skb->pkt_type = PACKET_OUTGOING;
 
-	mtu = READ_ONCE(skb->dev->mtu);
-	if (skb->len > mtu) {
+	if (skb->len > dst->mtu) {
 		kfree_skb(skb);
 		return -EMSGSIZE;
 	}
 
-	if (cb->ifindex) {
-		/* direct route; use the hwaddr we stashed in sendmsg */
-		if (cb->halen != skb->dev->addr_len) {
+	/* direct route; use the hwaddr we stashed in sendmsg */
+	if (dst->halen) {
+		if (dst->halen != skb->dev->addr_len) {
 			/* sanity check, sendmsg should have already caught this */
 			kfree_skb(skb);
 			return -EMSGSIZE;
 		}
-		daddr = cb->haddr;
+		daddr = dst->haddr;
 	} else {
 		/* If lookup fails let the device handle daddr==NULL */
-		if (mctp_neigh_lookup(route->dev, hdr->dest, daddr_buf) == 0)
+		if (mctp_neigh_lookup(dst->dev, dst->nexthop, daddr_buf) == 0)
 			daddr = daddr_buf;
 	}
 
@@ -599,7 +596,7 @@ static int mctp_route_output(struct mctp_route *route, struct sk_buff *skb)
 		return -EHOSTUNREACH;
 	}
 
-	mctp_flow_prepare_output(skb, route->dev);
+	mctp_flow_prepare_output(skb, dst->dev);
 
 	rc = dev_queue_xmit(skb);
 	if (rc)
@@ -612,7 +609,8 @@ static int mctp_route_output(struct mctp_route *route, struct sk_buff *skb)
 static void mctp_route_release(struct mctp_route *rt)
 {
 	if (refcount_dec_and_test(&rt->refs)) {
-		mctp_dev_put(rt->dev);
+		if (rt->dst_type == MCTP_ROUTE_DIRECT)
+			mctp_dev_put(rt->dev);
 		kfree_rcu(rt, rcu);
 	}
 }
@@ -628,7 +626,7 @@ static struct mctp_route *mctp_route_alloc(void)
 
 	INIT_LIST_HEAD(&rt->list);
 	refcount_set(&rt->refs, 1);
-	rt->output = mctp_route_discard;
+	rt->output = mctp_dst_discard;
 
 	return rt;
 }
@@ -801,10 +799,16 @@ static struct mctp_sk_key *mctp_lookup_prealloc_tag(struct mctp_sock *msk,
 }
 
 /* routing lookups */
+static unsigned int mctp_route_netid(struct mctp_route *rt)
+{
+	return rt->dst_type == MCTP_ROUTE_DIRECT ?
+		READ_ONCE(rt->dev->net) : rt->gateway.net;
+}
+
 static bool mctp_rt_match_eid(struct mctp_route *rt,
 			      unsigned int net, mctp_eid_t eid)
 {
-	return READ_ONCE(rt->dev->net) == net &&
+	return mctp_route_netid(rt) == net &&
 		rt->min <= eid && rt->max >= eid;
 }
 
@@ -813,54 +817,150 @@ static bool mctp_rt_compare_exact(struct mctp_route *rt1,
 				  struct mctp_route *rt2)
 {
 	ASSERT_RTNL();
-	return rt1->dev->net == rt2->dev->net &&
+	return mctp_route_netid(rt1) == mctp_route_netid(rt2) &&
 		rt1->min == rt2->min &&
 		rt1->max == rt2->max;
 }
 
-struct mctp_route *mctp_route_lookup(struct net *net, unsigned int dnet,
-				     mctp_eid_t daddr)
+/* must only be called on a direct route, as the final output hop */
+static void mctp_dst_from_route(struct mctp_dst *dst, mctp_eid_t eid,
+				unsigned int mtu, struct mctp_route *route)
+{
+	mctp_dev_hold(route->dev);
+	dst->nexthop = eid;
+	dst->dev = route->dev;
+	dst->mtu = READ_ONCE(dst->dev->dev->mtu);
+	if (mtu)
+		dst->mtu = min(dst->mtu, mtu);
+	dst->halen = 0;
+	dst->output = route->output;
+}
+
+int mctp_dst_from_extaddr(struct mctp_dst *dst, struct net *net, int ifindex,
+			  unsigned char halen, const unsigned char *haddr)
 {
-	struct mctp_route *tmp, *rt = NULL;
+	struct net_device *netdev;
+	struct mctp_dev *dev;
+	int rc = -ENOENT;
+
+	if (halen > sizeof(dst->haddr))
+		return -EINVAL;
 
 	rcu_read_lock();
 
-	list_for_each_entry_rcu(tmp, &net->mctp.routes, list) {
-		/* TODO: add metrics */
-		if (mctp_rt_match_eid(tmp, dnet, daddr)) {
-			if (refcount_inc_not_zero(&tmp->refs)) {
-				rt = tmp;
-				break;
-			}
-		}
+	netdev = dev_get_by_index_rcu(net, ifindex);
+	if (!netdev)
+		goto out_unlock;
+
+	if (netdev->addr_len != halen) {
+		rc = -EINVAL;
+		goto out_unlock;
 	}
 
+	dev = __mctp_dev_get(netdev);
+	if (!dev)
+		goto out_unlock;
+
+	dst->dev = dev;
+	dst->mtu = READ_ONCE(netdev->mtu);
+	dst->halen = halen;
+	dst->output = mctp_dst_output;
+	dst->nexthop = 0;
+	memcpy(dst->haddr, haddr, halen);
+
+	rc = 0;
+
+out_unlock:
 	rcu_read_unlock();
+	return rc;
+}
 
-	return rt;
+void mctp_dst_release(struct mctp_dst *dst)
+{
+	mctp_dev_put(dst->dev);
+}
+
+static struct mctp_route *mctp_route_lookup_single(struct net *net,
+						   unsigned int dnet,
+						   mctp_eid_t daddr)
+{
+	struct mctp_route *rt;
+
+	list_for_each_entry_rcu(rt, &net->mctp.routes, list) {
+		if (mctp_rt_match_eid(rt, dnet, daddr))
+			return rt;
+	}
+
+	return NULL;
 }
 
-static struct mctp_route *mctp_route_lookup_null(struct net *net,
-						 struct net_device *dev)
+/* populates *dst on successful lookup, if set */
+int mctp_route_lookup(struct net *net, unsigned int dnet,
+		      mctp_eid_t daddr, struct mctp_dst *dst)
 {
-	struct mctp_route *tmp, *rt = NULL;
+	const unsigned int max_depth = 32;
+	unsigned int depth, mtu = 0;
+	int rc = -EHOSTUNREACH;
 
 	rcu_read_lock();
 
-	list_for_each_entry_rcu(tmp, &net->mctp.routes, list) {
-		if (tmp->dev->dev == dev && tmp->type == RTN_LOCAL &&
-		    refcount_inc_not_zero(&tmp->refs)) {
-			rt = tmp;
+	for (depth = 0; depth < max_depth; depth++) {
+		struct mctp_route *rt;
+
+		rt = mctp_route_lookup_single(net, dnet, daddr);
+		if (!rt)
 			break;
+
+		/* clamp mtu to the smallest in the path, allowing 0
+		 * to specify no restrictions
+		 */
+		if (mtu && rt->mtu)
+			mtu = min(mtu, rt->mtu);
+		else
+			mtu = mtu ?: rt->mtu;
+
+		if (rt->dst_type == MCTP_ROUTE_DIRECT) {
+			if (dst)
+				mctp_dst_from_route(dst, daddr, mtu, rt);
+			rc = 0;
+			break;
+
+		} else if (rt->dst_type == MCTP_ROUTE_GATEWAY) {
+			daddr = rt->gateway.eid;
 		}
 	}
 
 	rcu_read_unlock();
 
-	return rt;
+	return rc;
 }
 
-static int mctp_do_fragment_route(struct mctp_route *rt, struct sk_buff *skb,
+static int mctp_route_lookup_null(struct net *net, struct net_device *dev,
+				  struct mctp_dst *dst)
+{
+	int rc = -EHOSTUNREACH;
+	struct mctp_route *rt;
+
+	rcu_read_lock();
+
+	list_for_each_entry_rcu(rt, &net->mctp.routes, list) {
+		if (rt->dst_type != MCTP_ROUTE_DIRECT || rt->type != RTN_LOCAL)
+			continue;
+
+		if (rt->dev->dev != dev)
+			continue;
+
+		mctp_dst_from_route(dst, 0, 0, rt);
+		rc = 0;
+		break;
+	}
+
+	rcu_read_unlock();
+
+	return rc;
+}
+
+static int mctp_do_fragment_route(struct mctp_dst *dst, struct sk_buff *skb,
 				  unsigned int mtu, u8 tag)
 {
 	const unsigned int hlen = sizeof(struct mctp_hdr);
@@ -933,7 +1033,7 @@ static int mctp_do_fragment_route(struct mctp_route *rt, struct sk_buff *skb,
 		skb_ext_copy(skb2, skb);
 
 		/* do route */
-		rc = rt->output(rt, skb2);
+		rc = dst->output(dst, skb2);
 		if (rc)
 			break;
 
@@ -945,68 +1045,34 @@ static int mctp_do_fragment_route(struct mctp_route *rt, struct sk_buff *skb,
 	return rc;
 }
 
-int mctp_local_output(struct sock *sk, struct mctp_route *rt,
+int mctp_local_output(struct sock *sk, struct mctp_dst *dst,
 		      struct sk_buff *skb, mctp_eid_t daddr, u8 req_tag)
 {
 	struct mctp_sock *msk = container_of(sk, struct mctp_sock, sk);
-	struct mctp_skb_cb *cb = mctp_cb(skb);
-	struct mctp_route tmp_rt = {0};
 	struct mctp_sk_key *key;
 	struct mctp_hdr *hdr;
 	unsigned long flags;
 	unsigned int netid;
 	unsigned int mtu;
 	mctp_eid_t saddr;
-	bool ext_rt;
 	int rc;
 	u8 tag;
 
-	rc = -ENODEV;
-
-	if (rt) {
-		ext_rt = false;
-		if (WARN_ON(!rt->dev))
-			goto out_release;
-
-	} else if (cb->ifindex) {
-		struct net_device *dev;
-
-		ext_rt = true;
-		rt = &tmp_rt;
-
-		rcu_read_lock();
-		dev = dev_get_by_index_rcu(sock_net(sk), cb->ifindex);
-		if (!dev) {
-			rcu_read_unlock();
-			goto out_free;
-		}
-		rt->dev = __mctp_dev_get(dev);
-		rcu_read_unlock();
-
-		if (!rt->dev)
-			goto out_release;
-
-		/* establish temporary route - we set up enough to keep
-		 * mctp_route_output happy
-		 */
-		rt->output = mctp_route_output;
-		rt->mtu = 0;
+	KUNIT_STATIC_STUB_REDIRECT(mctp_local_output, sk, dst, skb, daddr,
+				   req_tag);
 
-	} else {
-		rc = -EINVAL;
-		goto out_free;
-	}
+	rc = -ENODEV;
 
-	spin_lock_irqsave(&rt->dev->addrs_lock, flags);
-	if (rt->dev->num_addrs == 0) {
+	spin_lock_irqsave(&dst->dev->addrs_lock, flags);
+	if (dst->dev->num_addrs == 0) {
 		rc = -EHOSTUNREACH;
 	} else {
 		/* use the outbound interface's first address as our source */
-		saddr = rt->dev->addrs[0];
+		saddr = dst->dev->addrs[0];
 		rc = 0;
 	}
-	spin_unlock_irqrestore(&rt->dev->addrs_lock, flags);
-	netid = READ_ONCE(rt->dev->net);
+	spin_unlock_irqrestore(&dst->dev->addrs_lock, flags);
+	netid = READ_ONCE(dst->dev->net);
 
 	if (rc)
 		goto out_release;
@@ -1032,15 +1098,13 @@ int mctp_local_output(struct sock *sk, struct mctp_route *rt,
 		tag = req_tag & MCTP_TAG_MASK;
 	}
 
+	skb->pkt_type = PACKET_OUTGOING;
 	skb->protocol = htons(ETH_P_MCTP);
 	skb->priority = 0;
 	skb_reset_transport_header(skb);
 	skb_push(skb, sizeof(struct mctp_hdr));
 	skb_reset_network_header(skb);
-	skb->dev = rt->dev->dev;
-
-	/* cb->net will have been set on initial ingress */
-	cb->src = saddr;
+	skb->dev = dst->dev->dev;
 
 	/* set up common header fields */
 	hdr = mctp_hdr(skb);
@@ -1048,73 +1112,64 @@ int mctp_local_output(struct sock *sk, struct mctp_route *rt,
 	hdr->dest = daddr;
 	hdr->src = saddr;
 
-	mtu = mctp_route_mtu(rt);
+	mtu = dst->mtu;
 
 	if (skb->len + sizeof(struct mctp_hdr) <= mtu) {
 		hdr->flags_seq_tag = MCTP_HDR_FLAG_SOM |
 			MCTP_HDR_FLAG_EOM | tag;
-		rc = rt->output(rt, skb);
+		rc = dst->output(dst, skb);
 	} else {
-		rc = mctp_do_fragment_route(rt, skb, mtu, tag);
+		rc = mctp_do_fragment_route(dst, skb, mtu, tag);
 	}
 
 	/* route output functions consume the skb, even on error */
 	skb = NULL;
 
 out_release:
-	if (!ext_rt)
-		mctp_route_release(rt);
-
-	mctp_dev_put(tmp_rt.dev);
-
-out_free:
 	kfree_skb(skb);
 	return rc;
 }
 
 /* route management */
-static int mctp_route_add(struct mctp_dev *mdev, mctp_eid_t daddr_start,
-			  unsigned int daddr_extent, unsigned int mtu,
-			  unsigned char type)
+
+/* mctp_route_add(): Add the provided route, previously allocated via
+ * mctp_route_alloc(). On success, takes ownership of @rt, which includes a
+ * hold on rt->dev for usage in the route table. On failure a caller will want
+ * to mctp_route_release().
+ *
+ * We expect that the caller has set rt->type, rt->dst_type, rt->min, rt->max,
+ * rt->mtu and either rt->dev (with a reference held appropriately) or
+ * rt->gateway. Other fields will be populated.
+ */
+static int mctp_route_add(struct net *net, struct mctp_route *rt)
 {
-	int (*rtfn)(struct mctp_route *rt, struct sk_buff *skb);
-	struct net *net = dev_net(mdev->dev);
-	struct mctp_route *rt, *ert;
+	struct mctp_route *ert;
 
-	if (!mctp_address_unicast(daddr_start))
+	if (!mctp_address_unicast(rt->min) || !mctp_address_unicast(rt->max))
 		return -EINVAL;
 
-	if (daddr_extent > 0xff || daddr_start + daddr_extent >= 255)
+	if (rt->dst_type == MCTP_ROUTE_DIRECT && !rt->dev)
 		return -EINVAL;
 
-	switch (type) {
+	if (rt->dst_type == MCTP_ROUTE_GATEWAY && !rt->gateway.eid)
+		return -EINVAL;
+
+	switch (rt->type) {
 	case RTN_LOCAL:
-		rtfn = mctp_route_input;
+		rt->output = mctp_dst_input;
 		break;
 	case RTN_UNICAST:
-		rtfn = mctp_route_output;
+		rt->output = mctp_dst_output;
 		break;
 	default:
 		return -EINVAL;
 	}
 
-	rt = mctp_route_alloc();
-	if (!rt)
-		return -ENOMEM;
-
-	rt->min = daddr_start;
-	rt->max = daddr_start + daddr_extent;
-	rt->mtu = mtu;
-	rt->dev = mdev;
-	mctp_dev_hold(rt->dev);
-	rt->type = type;
-	rt->output = rtfn;
-
 	ASSERT_RTNL();
+
 	/* Prevent duplicate identical routes. */
 	list_for_each_entry(ert, &net->mctp.routes, list) {
 		if (mctp_rt_compare_exact(rt, ert)) {
-			mctp_route_release(rt);
 			return -EEXIST;
 		}
 	}
@@ -1124,10 +1179,10 @@ static int mctp_route_add(struct mctp_dev *mdev, mctp_eid_t daddr_start,
 	return 0;
 }
 
-static int mctp_route_remove(struct mctp_dev *mdev, mctp_eid_t daddr_start,
-			     unsigned int daddr_extent, unsigned char type)
+static int mctp_route_remove(struct net *net, unsigned int netid,
+			     mctp_eid_t daddr_start, unsigned int daddr_extent,
+			     unsigned char type)
 {
-	struct net *net = dev_net(mdev->dev);
 	struct mctp_route *rt, *tmp;
 	mctp_eid_t daddr_end;
 	bool dropped;
@@ -1141,7 +1196,7 @@ static int mctp_route_remove(struct mctp_dev *mdev, mctp_eid_t daddr_start,
 	ASSERT_RTNL();
 
 	list_for_each_entry_safe(rt, tmp, &net->mctp.routes, list) {
-		if (rt->dev == mdev &&
+		if (mctp_route_netid(rt) == netid &&
 		    rt->min == daddr_start && rt->max == daddr_end &&
 		    rt->type == type) {
 			list_del_rcu(&rt->list);
@@ -1156,12 +1211,32 @@ static int mctp_route_remove(struct mctp_dev *mdev, mctp_eid_t daddr_start,
 
 int mctp_route_add_local(struct mctp_dev *mdev, mctp_eid_t addr)
 {
-	return mctp_route_add(mdev, addr, 0, 0, RTN_LOCAL);
+	struct mctp_route *rt;
+	int rc;
+
+	rt = mctp_route_alloc();
+	if (!rt)
+		return -ENOMEM;
+
+	rt->min = addr;
+	rt->max = addr;
+	rt->dst_type = MCTP_ROUTE_DIRECT;
+	rt->dev = mdev;
+	rt->type = RTN_LOCAL;
+
+	mctp_dev_hold(rt->dev);
+
+	rc = mctp_route_add(dev_net(mdev->dev), rt);
+	if (rc)
+		mctp_route_release(rt);
+
+	return rc;
 }
 
 int mctp_route_remove_local(struct mctp_dev *mdev, mctp_eid_t addr)
 {
-	return mctp_route_remove(mdev, addr, 0, RTN_LOCAL);
+	return mctp_route_remove(dev_net(mdev->dev), mdev->net,
+				 addr, 0, RTN_LOCAL);
 }
 
 /* removes all entries for a given device */
@@ -1172,7 +1247,7 @@ void mctp_route_remove_dev(struct mctp_dev *mdev)
 
 	ASSERT_RTNL();
 	list_for_each_entry_safe(rt, tmp, &net->mctp.routes, list) {
-		if (rt->dev == mdev) {
+		if (rt->dst_type == MCTP_ROUTE_DIRECT && rt->dev == mdev) {
 			list_del_rcu(&rt->list);
 			/* TODO: immediate RTM_DELROUTE */
 			mctp_route_release(rt);
@@ -1189,8 +1264,9 @@ static int mctp_pkttype_receive(struct sk_buff *skb, struct net_device *dev,
 	struct net *net = dev_net(dev);
 	struct mctp_dev *mdev;
 	struct mctp_skb_cb *cb;
-	struct mctp_route *rt;
+	struct mctp_dst dst;
 	struct mctp_hdr *mh;
+	int rc;
 
 	rcu_read_lock();
 	mdev = __mctp_dev_get(dev);
@@ -1232,17 +1308,17 @@ static int mctp_pkttype_receive(struct sk_buff *skb, struct net_device *dev,
 	cb->net = READ_ONCE(mdev->net);
 	cb->ifindex = dev->ifindex;
 
-	rt = mctp_route_lookup(net, cb->net, mh->dest);
+	rc = mctp_route_lookup(net, cb->net, mh->dest, &dst);
 
 	/* NULL EID, but addressed to our physical address */
-	if (!rt && mh->dest == MCTP_ADDR_NULL && skb->pkt_type == PACKET_HOST)
-		rt = mctp_route_lookup_null(net, dev);
+	if (rc && mh->dest == MCTP_ADDR_NULL && skb->pkt_type == PACKET_HOST)
+		rc = mctp_route_lookup_null(net, dev, &dst);
 
-	if (!rt)
+	if (rc)
 		goto err_drop;
 
-	rt->output(rt, skb);
-	mctp_route_release(rt);
+	dst.output(&dst, skb);
+	mctp_dst_release(&dst);
 	mctp_dev_put(mdev);
 
 	return NET_RX_SUCCESS;
@@ -1264,19 +1340,28 @@ static const struct nla_policy rta_mctp_policy[RTA_MAX + 1] = {
 	[RTA_DST]		= { .type = NLA_U8 },
 	[RTA_METRICS]		= { .type = NLA_NESTED },
 	[RTA_OIF]		= { .type = NLA_U32 },
+	[RTA_GATEWAY]		= NLA_POLICY_EXACT_LEN(sizeof(struct mctp_fq_addr)),
+};
+
+static const struct nla_policy rta_metrics_policy[RTAX_MAX + 1] = {
+	[RTAX_MTU]		= { .type = NLA_U32 },
 };
 
-/* Common part for RTM_NEWROUTE and RTM_DELROUTE parsing.
- * tb must hold RTA_MAX+1 elements.
+/* base parsing; common to both _lookup and _populate variants.
+ *
+ * For gateway routes (which have a RTA_GATEWAY, and no RTA_OIF), we populate
+ * *gatweayp. for direct routes (RTA_OIF, no RTA_GATEWAY), we populate *mdev.
  */
-static int mctp_route_nlparse(struct sk_buff *skb, struct nlmsghdr *nlh,
-			      struct netlink_ext_ack *extack,
-			      struct nlattr **tb, struct rtmsg **rtm,
-			      struct mctp_dev **mdev, mctp_eid_t *daddr_start)
+static int mctp_route_nlparse_common(struct net *net, struct nlmsghdr *nlh,
+				     struct netlink_ext_ack *extack,
+				     struct nlattr **tb, struct rtmsg **rtm,
+				     struct mctp_dev **mdev,
+				     struct mctp_fq_addr *gatewayp,
+				     mctp_eid_t *daddr_start)
 {
-	struct net *net = sock_net(skb->sk);
+	struct mctp_fq_addr *gateway = NULL;
+	unsigned int ifindex = 0;
 	struct net_device *dev;
-	unsigned int ifindex;
 	int rc;
 
 	rc = nlmsg_parse(nlh, sizeof(struct rtmsg), tb, RTA_MAX,
@@ -1292,11 +1377,44 @@ static int mctp_route_nlparse(struct sk_buff *skb, struct nlmsghdr *nlh,
 	}
 	*daddr_start = nla_get_u8(tb[RTA_DST]);
 
-	if (!tb[RTA_OIF]) {
-		NL_SET_ERR_MSG(extack, "ifindex missing");
+	if (tb[RTA_OIF])
+		ifindex = nla_get_u32(tb[RTA_OIF]);
+
+	if (tb[RTA_GATEWAY])
+		gateway = nla_data(tb[RTA_GATEWAY]);
+
+	if (ifindex && gateway) {
+		NL_SET_ERR_MSG(extack,
+			       "cannot specify both ifindex and gateway");
+		return -EINVAL;
+
+	} else if (ifindex) {
+		dev = __dev_get_by_index(net, ifindex);
+		if (!dev) {
+			NL_SET_ERR_MSG(extack, "bad ifindex");
+			return -ENODEV;
+		}
+		*mdev = mctp_dev_get_rtnl(dev);
+		if (!*mdev)
+			return -ENODEV;
+		gatewayp->eid = 0;
+
+	} else if (gateway) {
+		if (!mctp_address_unicast(gateway->eid)) {
+			NL_SET_ERR_MSG(extack, "bad gateway");
+			return -EINVAL;
+		}
+
+		gatewayp->eid = gateway->eid;
+		gatewayp->net = gateway->net != MCTP_NET_ANY ?
+			gateway->net :
+			READ_ONCE(net->mctp.default_net);
+		*mdev = NULL;
+
+	} else {
+		NL_SET_ERR_MSG(extack, "no route output provided");
 		return -EINVAL;
 	}
-	ifindex = nla_get_u32(tb[RTA_OIF]);
 
 	*rtm = nlmsg_data(nlh);
 	if ((*rtm)->rtm_family != AF_MCTP) {
@@ -1304,82 +1422,157 @@ static int mctp_route_nlparse(struct sk_buff *skb, struct nlmsghdr *nlh,
 		return -EINVAL;
 	}
 
-	dev = __dev_get_by_index(net, ifindex);
-	if (!dev) {
-		NL_SET_ERR_MSG(extack, "bad ifindex");
-		return -ENODEV;
-	}
-	*mdev = mctp_dev_get_rtnl(dev);
-	if (!*mdev)
-		return -ENODEV;
-
-	if (dev->flags & IFF_LOOPBACK) {
-		NL_SET_ERR_MSG(extack, "no routes to loopback");
+	if ((*rtm)->rtm_type != RTN_UNICAST) {
+		NL_SET_ERR_MSG(extack, "rtm_type must be RTN_UNICAST");
 		return -EINVAL;
 	}
 
 	return 0;
 }
 
-static const struct nla_policy rta_metrics_policy[RTAX_MAX + 1] = {
-	[RTAX_MTU]		= { .type = NLA_U32 },
-};
-
-static int mctp_newroute(struct sk_buff *skb, struct nlmsghdr *nlh,
-			 struct netlink_ext_ack *extack)
+/* Route parsing for lookup operations; we only need the "route target"
+ * components (ie., network and dest-EID range).
+ */
+static int mctp_route_nlparse_lookup(struct net *net, struct nlmsghdr *nlh,
+				     struct netlink_ext_ack *extack,
+				     unsigned char *type, unsigned int *netid,
+				     mctp_eid_t *daddr_start,
+				     unsigned int *daddr_extent)
 {
 	struct nlattr *tb[RTA_MAX + 1];
+	struct mctp_fq_addr gw;
+	struct mctp_dev *mdev;
+	struct rtmsg *rtm;
+	int rc;
+
+	rc = mctp_route_nlparse_common(net, nlh, extack, tb, &rtm,
+				       &mdev, &gw, daddr_start);
+	if (rc)
+		return rc;
+
+	if (mdev) {
+		*netid = mdev->net;
+	} else if (gw.eid) {
+		*netid = gw.net;
+	} else {
+		/* bug: _nlparse_common should not allow this */
+		return -1;
+	}
+
+	*type = rtm->rtm_type;
+	*daddr_extent = rtm->rtm_dst_len;
+
+	return 0;
+}
+
+/* Full route parse for RTM_NEWROUTE: populate @rt. On success,
+ * MCTP_ROUTE_DIRECT routes (ie, those with a direct dev) will hold a reference
+ * to that dev.
+ */
+static int mctp_route_nlparse_populate(struct net *net, struct nlmsghdr *nlh,
+				       struct netlink_ext_ack *extack,
+				       struct mctp_route *rt)
+{
 	struct nlattr *tbx[RTAX_MAX + 1];
+	struct nlattr *tb[RTA_MAX + 1];
+	unsigned int daddr_extent;
+	struct mctp_fq_addr gw;
 	mctp_eid_t daddr_start;
-	struct mctp_dev *mdev;
+	struct mctp_dev *dev;
 	struct rtmsg *rtm;
-	unsigned int mtu;
+	u32 mtu = 0;
 	int rc;
 
-	rc = mctp_route_nlparse(skb, nlh, extack, tb,
-				&rtm, &mdev, &daddr_start);
-	if (rc < 0)
+	rc = mctp_route_nlparse_common(net, nlh, extack, tb, &rtm,
+				       &dev, &gw, &daddr_start);
+	if (rc)
 		return rc;
 
-	if (rtm->rtm_type != RTN_UNICAST) {
-		NL_SET_ERR_MSG(extack, "rtm_type must be RTN_UNICAST");
+	daddr_extent = rtm->rtm_dst_len;
+
+	if (daddr_extent > 0xff || daddr_extent + daddr_start >= 255) {
+		NL_SET_ERR_MSG(extack, "invalid eid range");
 		return -EINVAL;
 	}
 
-	mtu = 0;
 	if (tb[RTA_METRICS]) {
 		rc = nla_parse_nested(tbx, RTAX_MAX, tb[RTA_METRICS],
 				      rta_metrics_policy, NULL);
-		if (rc < 0)
+		if (rc < 0) {
+			NL_SET_ERR_MSG(extack, "incorrect RTA_METRICS format");
 			return rc;
+		}
 		if (tbx[RTAX_MTU])
 			mtu = nla_get_u32(tbx[RTAX_MTU]);
 	}
 
-	rc = mctp_route_add(mdev, daddr_start, rtm->rtm_dst_len, mtu,
-			    rtm->rtm_type);
+	rt->type = rtm->rtm_type;
+	rt->min = daddr_start;
+	rt->max = daddr_start + daddr_extent;
+	rt->mtu = mtu;
+	if (gw.eid) {
+		rt->dst_type = MCTP_ROUTE_GATEWAY;
+		rt->gateway.eid = gw.eid;
+		rt->gateway.net = gw.net;
+	} else {
+		rt->dst_type = MCTP_ROUTE_DIRECT;
+		rt->dev = dev;
+		mctp_dev_hold(rt->dev);
+	}
+
+	return 0;
+}
+
+static int mctp_newroute(struct sk_buff *skb, struct nlmsghdr *nlh,
+			 struct netlink_ext_ack *extack)
+{
+	struct net *net = sock_net(skb->sk);
+	struct mctp_route *rt;
+	int rc;
+
+	rt = mctp_route_alloc();
+	if (!rt)
+		return -ENOMEM;
+
+	rc = mctp_route_nlparse_populate(net, nlh, extack, rt);
+	if (rc < 0)
+		goto err_free;
+
+	if (rt->dst_type == MCTP_ROUTE_DIRECT &&
+	    rt->dev->dev->flags & IFF_LOOPBACK) {
+		NL_SET_ERR_MSG(extack, "no routes to loopback");
+		rc = -EINVAL;
+		goto err_free;
+	}
+
+	rc = mctp_route_add(net, rt);
+	if (!rc)
+		return 0;
+
+err_free:
+	mctp_route_release(rt);
 	return rc;
 }
 
 static int mctp_delroute(struct sk_buff *skb, struct nlmsghdr *nlh,
 			 struct netlink_ext_ack *extack)
 {
-	struct nlattr *tb[RTA_MAX + 1];
+	struct net *net = sock_net(skb->sk);
+	unsigned int netid, daddr_extent;
+	unsigned char type = RTN_UNSPEC;
 	mctp_eid_t daddr_start;
-	struct mctp_dev *mdev;
-	struct rtmsg *rtm;
 	int rc;
 
-	rc = mctp_route_nlparse(skb, nlh, extack, tb,
-				&rtm, &mdev, &daddr_start);
+	rc = mctp_route_nlparse_lookup(net, nlh, extack, &type, &netid,
+				       &daddr_start, &daddr_extent);
 	if (rc < 0)
 		return rc;
 
 	/* we only have unicast routes */
-	if (rtm->rtm_type != RTN_UNICAST)
+	if (type != RTN_UNICAST)
 		return -EINVAL;
 
-	rc = mctp_route_remove(mdev, daddr_start, rtm->rtm_dst_len, RTN_UNICAST);
+	rc = mctp_route_remove(net, netid, daddr_start, daddr_extent, type);
 	return rc;
 }
 
@@ -1405,7 +1598,6 @@ static int mctp_fill_rtinfo(struct sk_buff *skb, struct mctp_route *rt,
 	hdr->rtm_tos = 0;
 	hdr->rtm_table = RT_TABLE_DEFAULT;
 	hdr->rtm_protocol = RTPROT_STATIC; /* everything is user-defined */
-	hdr->rtm_scope = RT_SCOPE_LINK; /* TODO: scope in mctp_route? */
 	hdr->rtm_type = rt->type;
 
 	if (nla_put_u8(skb, RTA_DST, rt->min))
@@ -1422,13 +1614,17 @@ static int mctp_fill_rtinfo(struct sk_buff *skb, struct mctp_route *rt,
 
 	nla_nest_end(skb, metrics);
 
-	if (rt->dev) {
+	if (rt->dst_type == MCTP_ROUTE_DIRECT) {
+		hdr->rtm_scope = RT_SCOPE_LINK;
 		if (nla_put_u32(skb, RTA_OIF, rt->dev->dev->ifindex))
 			goto cancel;
+	} else if (rt->dst_type == MCTP_ROUTE_GATEWAY) {
+		hdr->rtm_scope = RT_SCOPE_UNIVERSE;
+		if (nla_put(skb, RTA_GATEWAY,
+			    sizeof(rt->gateway), &rt->gateway))
+			goto cancel;
 	}
 
-	/* TODO: conditional neighbour physaddr? */
-
 	nlmsg_end(skb, nlh);
 
 	return 0;
diff --git a/net/mctp/test/route-test.c b/net/mctp/test/route-test.c
index 06c1897b685a..7a398f41b621 100644
--- a/net/mctp/test/route-test.c
+++ b/net/mctp/test/route-test.c
@@ -2,132 +2,11 @@
 
 #include <kunit/test.h>
 
-#include "utils.h"
-
-struct mctp_test_route {
-	struct mctp_route	rt;
-	struct sk_buff_head	pkts;
-};
-
-static int mctp_test_route_output(struct mctp_route *rt, struct sk_buff *skb)
-{
-	struct mctp_test_route *test_rt = container_of(rt, struct mctp_test_route, rt);
-
-	skb_queue_tail(&test_rt->pkts, skb);
-
-	return 0;
-}
-
-/* local version of mctp_route_alloc() */
-static struct mctp_test_route *mctp_route_test_alloc(void)
-{
-	struct mctp_test_route *rt;
-
-	rt = kzalloc(sizeof(*rt), GFP_KERNEL);
-	if (!rt)
-		return NULL;
-
-	INIT_LIST_HEAD(&rt->rt.list);
-	refcount_set(&rt->rt.refs, 1);
-	rt->rt.output = mctp_test_route_output;
-
-	skb_queue_head_init(&rt->pkts);
-
-	return rt;
-}
-
-static struct mctp_test_route *mctp_test_create_route(struct net *net,
-						      struct mctp_dev *dev,
-						      mctp_eid_t eid,
-						      unsigned int mtu)
-{
-	struct mctp_test_route *rt;
-
-	rt = mctp_route_test_alloc();
-	if (!rt)
-		return NULL;
-
-	rt->rt.min = eid;
-	rt->rt.max = eid;
-	rt->rt.mtu = mtu;
-	rt->rt.type = RTN_UNSPEC;
-	if (dev)
-		mctp_dev_hold(dev);
-	rt->rt.dev = dev;
-
-	list_add_rcu(&rt->rt.list, &net->mctp.routes);
-
-	return rt;
-}
-
-static void mctp_test_route_destroy(struct kunit *test,
-				    struct mctp_test_route *rt)
-{
-	unsigned int refs;
-
-	rtnl_lock();
-	list_del_rcu(&rt->rt.list);
-	rtnl_unlock();
-
-	skb_queue_purge(&rt->pkts);
-	if (rt->rt.dev)
-		mctp_dev_put(rt->rt.dev);
-
-	refs = refcount_read(&rt->rt.refs);
-	KUNIT_ASSERT_EQ_MSG(test, refs, 1, "route ref imbalance");
-
-	kfree_rcu(&rt->rt, rcu);
-}
-
-static void mctp_test_skb_set_dev(struct sk_buff *skb,
-				  struct mctp_test_dev *dev)
-{
-	struct mctp_skb_cb *cb;
-
-	cb = mctp_cb(skb);
-	cb->net = READ_ONCE(dev->mdev->net);
-	skb->dev = dev->ndev;
-}
-
-static struct sk_buff *mctp_test_create_skb(const struct mctp_hdr *hdr,
-					    unsigned int data_len)
-{
-	size_t hdr_len = sizeof(*hdr);
-	struct sk_buff *skb;
-	unsigned int i;
-	u8 *buf;
-
-	skb = alloc_skb(hdr_len + data_len, GFP_KERNEL);
-	if (!skb)
-		return NULL;
-
-	__mctp_cb(skb);
-	memcpy(skb_put(skb, hdr_len), hdr, hdr_len);
-
-	buf = skb_put(skb, data_len);
-	for (i = 0; i < data_len; i++)
-		buf[i] = i & 0xff;
-
-	return skb;
-}
-
-static struct sk_buff *__mctp_test_create_skb_data(const struct mctp_hdr *hdr,
-						   const void *data,
-						   size_t data_len)
-{
-	size_t hdr_len = sizeof(*hdr);
-	struct sk_buff *skb;
-
-	skb = alloc_skb(hdr_len + data_len, GFP_KERNEL);
-	if (!skb)
-		return NULL;
-
-	__mctp_cb(skb);
-	memcpy(skb_put(skb, hdr_len), hdr, hdr_len);
-	memcpy(skb_put(skb, data_len), data, data_len);
+/* keep clangd happy when compiled outside of the route.c include */
+#include <net/mctp.h>
+#include <net/mctpdevice.h>
 
-	return skb;
-}
+#include "utils.h"
 
 #define mctp_test_create_skb_data(h, d) \
 	__mctp_test_create_skb_data(h, d, sizeof(*d))
@@ -141,8 +20,10 @@ struct mctp_frag_test {
 static void mctp_test_fragment(struct kunit *test)
 {
 	const struct mctp_frag_test *params;
+	struct mctp_test_pktqueue tpq;
 	int rc, i, n, mtu, msgsize;
-	struct mctp_test_route *rt;
+	struct mctp_test_dev *dev;
+	struct mctp_dst dst;
 	struct sk_buff *skb;
 	struct mctp_hdr hdr;
 	u8 seq;
@@ -159,13 +40,15 @@ static void mctp_test_fragment(struct kunit *test)
 	skb = mctp_test_create_skb(&hdr, msgsize);
 	KUNIT_ASSERT_TRUE(test, skb);
 
-	rt = mctp_test_create_route(&init_net, NULL, 10, mtu);
-	KUNIT_ASSERT_TRUE(test, rt);
+	dev = mctp_test_create_dev();
+	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, dev);
+
+	mctp_test_dst_setup(test, &dst, dev, &tpq, mtu);
 
-	rc = mctp_do_fragment_route(&rt->rt, skb, mtu, MCTP_TAG_OWNER);
+	rc = mctp_do_fragment_route(&dst, skb, mtu, MCTP_TAG_OWNER);
 	KUNIT_EXPECT_FALSE(test, rc);
 
-	n = rt->pkts.qlen;
+	n = tpq.pkts.qlen;
 
 	KUNIT_EXPECT_EQ(test, n, params->n_frags);
 
@@ -178,7 +61,7 @@ static void mctp_test_fragment(struct kunit *test)
 		first = i == 0;
 		last = i == (n - 1);
 
-		skb2 = skb_dequeue(&rt->pkts);
+		skb2 = skb_dequeue(&tpq.pkts);
 
 		if (!skb2)
 			break;
@@ -216,7 +99,8 @@ static void mctp_test_fragment(struct kunit *test)
 		kfree_skb(skb2);
 	}
 
-	mctp_test_route_destroy(test, rt);
+	mctp_test_dst_release(&dst, &tpq);
+	mctp_test_destroy_dev(dev);
 }
 
 static const struct mctp_frag_test mctp_frag_tests[] = {
@@ -246,25 +130,30 @@ struct mctp_rx_input_test {
 static void mctp_test_rx_input(struct kunit *test)
 {
 	const struct mctp_rx_input_test *params;
+	struct mctp_test_pktqueue tpq;
 	struct mctp_test_route *rt;
 	struct mctp_test_dev *dev;
 	struct sk_buff *skb;
 
 	params = test->param_value;
+	test->priv = &tpq;
 
 	dev = mctp_test_create_dev();
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, dev);
 
-	rt = mctp_test_create_route(&init_net, dev->mdev, 8, 68);
+	rt = mctp_test_create_route_direct(&init_net, dev->mdev, 8, 68);
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, rt);
 
 	skb = mctp_test_create_skb(&params->hdr, 1);
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, skb);
 
+	mctp_test_pktqueue_init(&tpq);
+
 	mctp_pkttype_receive(skb, dev->ndev, &mctp_packet_type, NULL);
 
-	KUNIT_EXPECT_EQ(test, !!rt->pkts.qlen, params->input);
+	KUNIT_EXPECT_EQ(test, !!tpq.pkts.qlen, params->input);
 
+	skb_queue_purge(&tpq.pkts);
 	mctp_test_route_destroy(test, rt);
 	mctp_test_destroy_dev(dev);
 }
@@ -292,12 +181,12 @@ KUNIT_ARRAY_PARAM(mctp_rx_input, mctp_rx_input_tests,
 /* set up a local dev, route on EID 8, and a socket listening on type 0 */
 static void __mctp_route_test_init(struct kunit *test,
 				   struct mctp_test_dev **devp,
-				   struct mctp_test_route **rtp,
+				   struct mctp_dst *dst,
+				   struct mctp_test_pktqueue *tpq,
 				   struct socket **sockp,
 				   unsigned int netid)
 {
 	struct sockaddr_mctp addr = {0};
-	struct mctp_test_route *rt;
 	struct mctp_test_dev *dev;
 	struct socket *sock;
 	int rc;
@@ -307,8 +196,7 @@ static void __mctp_route_test_init(struct kunit *test,
 	if (netid != MCTP_NET_ANY)
 		WRITE_ONCE(dev->mdev->net, netid);
 
-	rt = mctp_test_create_route(&init_net, dev->mdev, 8, 68);
-	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, rt);
+	mctp_test_dst_setup(test, dst, dev, tpq, 68);
 
 	rc = sock_create_kern(&init_net, AF_MCTP, SOCK_DGRAM, 0, &sock);
 	KUNIT_ASSERT_EQ(test, rc, 0);
@@ -320,18 +208,18 @@ static void __mctp_route_test_init(struct kunit *test,
 	rc = kernel_bind(sock, (struct sockaddr *)&addr, sizeof(addr));
 	KUNIT_ASSERT_EQ(test, rc, 0);
 
-	*rtp = rt;
 	*devp = dev;
 	*sockp = sock;
 }
 
 static void __mctp_route_test_fini(struct kunit *test,
 				   struct mctp_test_dev *dev,
-				   struct mctp_test_route *rt,
+				   struct mctp_dst *dst,
+				   struct mctp_test_pktqueue *tpq,
 				   struct socket *sock)
 {
 	sock_release(sock);
-	mctp_test_route_destroy(test, rt);
+	mctp_test_dst_release(dst, tpq);
 	mctp_test_destroy_dev(dev);
 }
 
@@ -344,22 +232,24 @@ struct mctp_route_input_sk_test {
 static void mctp_test_route_input_sk(struct kunit *test)
 {
 	const struct mctp_route_input_sk_test *params;
+	struct mctp_test_pktqueue tpq;
 	struct sk_buff *skb, *skb2;
-	struct mctp_test_route *rt;
 	struct mctp_test_dev *dev;
+	struct mctp_dst dst;
 	struct socket *sock;
 	int rc;
 
 	params = test->param_value;
 
-	__mctp_route_test_init(test, &dev, &rt, &sock, MCTP_NET_ANY);
+	__mctp_route_test_init(test, &dev, &dst, &tpq, &sock, MCTP_NET_ANY);
 
 	skb = mctp_test_create_skb_data(&params->hdr, &params->type);
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, skb);
 
 	mctp_test_skb_set_dev(skb, dev);
+	mctp_test_pktqueue_init(&tpq);
 
-	rc = mctp_route_input(&rt->rt, skb);
+	rc = mctp_dst_input(&dst, skb);
 
 	if (params->deliver) {
 		KUNIT_EXPECT_EQ(test, rc, 0);
@@ -376,7 +266,7 @@ static void mctp_test_route_input_sk(struct kunit *test)
 		KUNIT_EXPECT_NULL(test, skb2);
 	}
 
-	__mctp_route_test_fini(test, dev, rt, sock);
+	__mctp_route_test_fini(test, dev, &dst, &tpq, sock);
 }
 
 #define FL_S	(MCTP_HDR_FLAG_SOM)
@@ -413,16 +303,17 @@ struct mctp_route_input_sk_reasm_test {
 static void mctp_test_route_input_sk_reasm(struct kunit *test)
 {
 	const struct mctp_route_input_sk_reasm_test *params;
+	struct mctp_test_pktqueue tpq;
 	struct sk_buff *skb, *skb2;
-	struct mctp_test_route *rt;
 	struct mctp_test_dev *dev;
+	struct mctp_dst dst;
 	struct socket *sock;
 	int i, rc;
 	u8 c;
 
 	params = test->param_value;
 
-	__mctp_route_test_init(test, &dev, &rt, &sock, MCTP_NET_ANY);
+	__mctp_route_test_init(test, &dev, &dst, &tpq, &sock, MCTP_NET_ANY);
 
 	for (i = 0; i < params->n_hdrs; i++) {
 		c = i;
@@ -431,7 +322,7 @@ static void mctp_test_route_input_sk_reasm(struct kunit *test)
 
 		mctp_test_skb_set_dev(skb, dev);
 
-		rc = mctp_route_input(&rt->rt, skb);
+		rc = mctp_dst_input(&dst, skb);
 	}
 
 	skb2 = skb_recv_datagram(sock->sk, MSG_DONTWAIT, &rc);
@@ -445,7 +336,7 @@ static void mctp_test_route_input_sk_reasm(struct kunit *test)
 		KUNIT_EXPECT_NULL(test, skb2);
 	}
 
-	__mctp_route_test_fini(test, dev, rt, sock);
+	__mctp_route_test_fini(test, dev, &dst, &tpq, sock);
 }
 
 #define RX_FRAG(f, s) RX_HDR(1, 10, 8, FL_TO | (f) | ((s) << MCTP_HDR_SEQ_SHIFT))
@@ -547,7 +438,7 @@ struct mctp_route_input_sk_keys_test {
 static void mctp_test_route_input_sk_keys(struct kunit *test)
 {
 	const struct mctp_route_input_sk_keys_test *params;
-	struct mctp_test_route *rt;
+	struct mctp_test_pktqueue tpq;
 	struct sk_buff *skb, *skb2;
 	struct mctp_test_dev *dev;
 	struct mctp_sk_key *key;
@@ -555,6 +446,7 @@ static void mctp_test_route_input_sk_keys(struct kunit *test)
 	struct mctp_sock *msk;
 	struct socket *sock;
 	unsigned long flags;
+	struct mctp_dst dst;
 	unsigned int net;
 	int rc;
 	u8 c;
@@ -565,8 +457,7 @@ static void mctp_test_route_input_sk_keys(struct kunit *test)
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, dev);
 	net = READ_ONCE(dev->mdev->net);
 
-	rt = mctp_test_create_route(&init_net, dev->mdev, 8, 68);
-	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, rt);
+	mctp_test_dst_setup(test, &dst, dev, &tpq, 68);
 
 	rc = sock_create_kern(&init_net, AF_MCTP, SOCK_DGRAM, 0, &sock);
 	KUNIT_ASSERT_EQ(test, rc, 0);
@@ -592,7 +483,7 @@ static void mctp_test_route_input_sk_keys(struct kunit *test)
 
 	mctp_test_skb_set_dev(skb, dev);
 
-	rc = mctp_route_input(&rt->rt, skb);
+	rc = mctp_dst_input(&dst, skb);
 
 	/* (potentially) receive message */
 	skb2 = skb_recv_datagram(sock->sk, MSG_DONTWAIT, &rc);
@@ -606,7 +497,7 @@ static void mctp_test_route_input_sk_keys(struct kunit *test)
 		skb_free_datagram(sock->sk, skb2);
 
 	mctp_key_unref(key);
-	__mctp_route_test_fini(test, dev, rt, sock);
+	__mctp_route_test_fini(test, dev, &dst, &tpq, sock);
 }
 
 static const struct mctp_route_input_sk_keys_test mctp_route_input_sk_keys_tests[] = {
@@ -681,7 +572,8 @@ KUNIT_ARRAY_PARAM(mctp_route_input_sk_keys, mctp_route_input_sk_keys_tests,
 struct test_net {
 	unsigned int netid;
 	struct mctp_test_dev *dev;
-	struct mctp_test_route *rt;
+	struct mctp_test_pktqueue tpq;
+	struct mctp_dst dst;
 	struct socket *sock;
 	struct sk_buff *skb;
 	struct mctp_sk_key *key;
@@ -699,18 +591,20 @@ mctp_test_route_input_multiple_nets_bind_init(struct kunit *test,
 
 	t->msg.data = t->netid;
 
-	__mctp_route_test_init(test, &t->dev, &t->rt, &t->sock, t->netid);
+	__mctp_route_test_init(test, &t->dev, &t->dst, &t->tpq, &t->sock,
+			       t->netid);
 
 	t->skb = mctp_test_create_skb_data(&hdr, &t->msg);
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, t->skb);
 	mctp_test_skb_set_dev(t->skb, t->dev);
+	mctp_test_pktqueue_init(&t->tpq);
 }
 
 static void
 mctp_test_route_input_multiple_nets_bind_fini(struct kunit *test,
 					      struct test_net *t)
 {
-	__mctp_route_test_fini(test, t->dev, t->rt, t->sock);
+	__mctp_route_test_fini(test, t->dev, &t->dst, &t->tpq, t->sock);
 }
 
 /* Test that skbs from different nets (otherwise identical) get routed to their
@@ -731,9 +625,9 @@ static void mctp_test_route_input_multiple_nets_bind(struct kunit *test)
 	mctp_test_route_input_multiple_nets_bind_init(test, &t1);
 	mctp_test_route_input_multiple_nets_bind_init(test, &t2);
 
-	rc = mctp_route_input(&t1.rt->rt, t1.skb);
+	rc = mctp_dst_input(&t1.dst, t1.skb);
 	KUNIT_ASSERT_EQ(test, rc, 0);
-	rc = mctp_route_input(&t2.rt->rt, t2.skb);
+	rc = mctp_dst_input(&t2.dst, t2.skb);
 	KUNIT_ASSERT_EQ(test, rc, 0);
 
 	rx_skb1 = skb_recv_datagram(t1.sock->sk, MSG_DONTWAIT, &rc);
@@ -767,7 +661,8 @@ mctp_test_route_input_multiple_nets_key_init(struct kunit *test,
 
 	t->msg.data = t->netid;
 
-	__mctp_route_test_init(test, &t->dev, &t->rt, &t->sock, t->netid);
+	__mctp_route_test_init(test, &t->dev, &t->dst, &t->tpq, &t->sock,
+			       t->netid);
 
 	msk = container_of(t->sock->sk, struct mctp_sock, sk);
 
@@ -790,7 +685,7 @@ mctp_test_route_input_multiple_nets_key_fini(struct kunit *test,
 					     struct test_net *t)
 {
 	mctp_key_unref(t->key);
-	__mctp_route_test_fini(test, t->dev, t->rt, t->sock);
+	__mctp_route_test_fini(test, t->dev, &t->dst, &t->tpq, t->sock);
 }
 
 /* test that skbs from different nets (otherwise identical) get routed to their
@@ -812,9 +707,9 @@ static void mctp_test_route_input_multiple_nets_key(struct kunit *test)
 	mctp_test_route_input_multiple_nets_key_init(test, &t1);
 	mctp_test_route_input_multiple_nets_key_init(test, &t2);
 
-	rc = mctp_route_input(&t1.rt->rt, t1.skb);
+	rc = mctp_dst_input(&t1.dst, t1.skb);
 	KUNIT_ASSERT_EQ(test, rc, 0);
-	rc = mctp_route_input(&t2.rt->rt, t2.skb);
+	rc = mctp_dst_input(&t2.dst, t2.skb);
 	KUNIT_ASSERT_EQ(test, rc, 0);
 
 	rx_skb1 = skb_recv_datagram(t1.sock->sk, MSG_DONTWAIT, &rc);
@@ -843,13 +738,14 @@ static void mctp_test_route_input_multiple_nets_key(struct kunit *test)
 static void mctp_test_route_input_sk_fail_single(struct kunit *test)
 {
 	const struct mctp_hdr hdr = RX_HDR(1, 10, 8, FL_S | FL_E | FL_TO);
-	struct mctp_test_route *rt;
+	struct mctp_test_pktqueue tpq;
 	struct mctp_test_dev *dev;
+	struct mctp_dst dst;
 	struct socket *sock;
 	struct sk_buff *skb;
 	int rc;
 
-	__mctp_route_test_init(test, &dev, &rt, &sock, MCTP_NET_ANY);
+	__mctp_route_test_init(test, &dev, &dst, &tpq, &sock, MCTP_NET_ANY);
 
 	/* No rcvbuf space, so delivery should fail. __sock_set_rcvbuf will
 	 * clamp the minimum to SOCK_MIN_RCVBUF, so we open-code this.
@@ -865,14 +761,14 @@ static void mctp_test_route_input_sk_fail_single(struct kunit *test)
 	mctp_test_skb_set_dev(skb, dev);
 
 	/* do route input, which should fail */
-	rc = mctp_route_input(&rt->rt, skb);
+	rc = mctp_dst_input(&dst, skb);
 	KUNIT_EXPECT_NE(test, rc, 0);
 
 	/* we should hold the only reference to skb */
 	KUNIT_EXPECT_EQ(test, refcount_read(&skb->users), 1);
 	kfree_skb(skb);
 
-	__mctp_route_test_fini(test, dev, rt, sock);
+	__mctp_route_test_fini(test, dev, &dst, &tpq, sock);
 }
 
 /* Input route to socket, using a fragmented message, where sock delivery fails.
@@ -880,14 +776,15 @@ static void mctp_test_route_input_sk_fail_single(struct kunit *test)
 static void mctp_test_route_input_sk_fail_frag(struct kunit *test)
 {
 	const struct mctp_hdr hdrs[2] = { RX_FRAG(FL_S, 0), RX_FRAG(FL_E, 1) };
-	struct mctp_test_route *rt;
+	struct mctp_test_pktqueue tpq;
 	struct mctp_test_dev *dev;
 	struct sk_buff *skbs[2];
+	struct mctp_dst dst;
 	struct socket *sock;
 	unsigned int i;
 	int rc;
 
-	__mctp_route_test_init(test, &dev, &rt, &sock, MCTP_NET_ANY);
+	__mctp_route_test_init(test, &dev, &dst, &tpq, &sock, MCTP_NET_ANY);
 
 	lock_sock(sock->sk);
 	WRITE_ONCE(sock->sk->sk_rcvbuf, 0);
@@ -904,11 +801,11 @@ static void mctp_test_route_input_sk_fail_frag(struct kunit *test)
 	/* first route input should succeed, we're only queueing to the
 	 * frag list
 	 */
-	rc = mctp_route_input(&rt->rt, skbs[0]);
+	rc = mctp_dst_input(&dst, skbs[0]);
 	KUNIT_EXPECT_EQ(test, rc, 0);
 
 	/* final route input should fail to deliver to the socket */
-	rc = mctp_route_input(&rt->rt, skbs[1]);
+	rc = mctp_dst_input(&dst, skbs[1]);
 	KUNIT_EXPECT_NE(test, rc, 0);
 
 	/* we should hold the only reference to both skbs */
@@ -918,7 +815,7 @@ static void mctp_test_route_input_sk_fail_frag(struct kunit *test)
 	KUNIT_EXPECT_EQ(test, refcount_read(&skbs[1]->users), 1);
 	kfree_skb(skbs[1]);
 
-	__mctp_route_test_fini(test, dev, rt, sock);
+	__mctp_route_test_fini(test, dev, &dst, &tpq, sock);
 }
 
 /* Input route to socket, using a fragmented message created from clones.
@@ -933,23 +830,22 @@ static void mctp_test_route_input_cloned_frag(struct kunit *test)
 		RX_FRAG(FL_S, 0),
 		RX_FRAG(FL_E, 1),
 	};
-	struct mctp_test_route *rt;
+	const size_t data_len = 3; /* arbitrary */
+	u8 compare[3 * ARRAY_SIZE(hdrs)];
+	u8 flat[3 * ARRAY_SIZE(hdrs)];
+	struct mctp_test_pktqueue tpq;
 	struct mctp_test_dev *dev;
 	struct sk_buff *skb[5];
 	struct sk_buff *rx_skb;
+	struct mctp_dst dst;
 	struct socket *sock;
-	size_t data_len;
-	u8 compare[100];
-	u8 flat[100];
 	size_t total;
 	void *p;
 	int rc;
 
-	/* Arbitrary length */
-	data_len = 3;
 	total = data_len + sizeof(struct mctp_hdr);
 
-	__mctp_route_test_init(test, &dev, &rt, &sock, MCTP_NET_ANY);
+	__mctp_route_test_init(test, &dev, &dst, &tpq, &sock, MCTP_NET_ANY);
 
 	/* Create a single skb initially with concatenated packets */
 	skb[0] = mctp_test_create_skb(&hdrs[0], 5 * total);
@@ -988,7 +884,7 @@ static void mctp_test_route_input_cloned_frag(struct kunit *test)
 
 	/* Feed the fragments into MCTP core */
 	for (int i = 0; i < 5; i++) {
-		rc = mctp_route_input(&rt->rt, skb[i]);
+		rc = mctp_dst_input(&dst, skb[i]);
 		KUNIT_EXPECT_EQ(test, rc, 0);
 	}
 
@@ -1026,29 +922,29 @@ static void mctp_test_route_input_cloned_frag(struct kunit *test)
 		kfree_skb(skb[i]);
 	}
 
-	__mctp_route_test_fini(test, dev, rt, sock);
+	__mctp_route_test_fini(test, dev, &dst, &tpq, sock);
 }
 
 #if IS_ENABLED(CONFIG_MCTP_FLOWS)
 
 static void mctp_test_flow_init(struct kunit *test,
 				struct mctp_test_dev **devp,
-				struct mctp_test_route **rtp,
+				struct mctp_dst *dst,
+				struct mctp_test_pktqueue *tpq,
 				struct socket **sock,
 				struct sk_buff **skbp,
 				unsigned int len)
 {
-	struct mctp_test_route *rt;
 	struct mctp_test_dev *dev;
 	struct sk_buff *skb;
 
 	/* we have a slightly odd routing setup here; the test route
 	 * is for EID 8, which is our local EID. We don't do a routing
 	 * lookup, so that's fine - all we require is a path through
-	 * mctp_local_output, which will call rt->output on whatever
+	 * mctp_local_output, which will call dst->output on whatever
 	 * route we provide
 	 */
-	__mctp_route_test_init(test, &dev, &rt, sock, MCTP_NET_ANY);
+	__mctp_route_test_init(test, &dev, dst, tpq, sock, MCTP_NET_ANY);
 
 	/* Assign a single EID. ->addrs is freed on mctp netdev release */
 	dev->mdev->addrs = kmalloc(sizeof(u8), GFP_KERNEL);
@@ -1061,42 +957,41 @@ static void mctp_test_flow_init(struct kunit *test,
 	skb_reserve(skb, sizeof(struct mctp_hdr) + 1);
 	memset(skb_put(skb, len), 0, len);
 
-	/* take a ref for the route, we'll decrement in local output */
-	refcount_inc(&rt->rt.refs);
 
 	*devp = dev;
-	*rtp = rt;
 	*skbp = skb;
 }
 
 static void mctp_test_flow_fini(struct kunit *test,
 				struct mctp_test_dev *dev,
-				struct mctp_test_route *rt,
+				struct mctp_dst *dst,
+				struct mctp_test_pktqueue *tpq,
 				struct socket *sock)
 {
-	__mctp_route_test_fini(test, dev, rt, sock);
+	__mctp_route_test_fini(test, dev, dst, tpq, sock);
 }
 
 /* test that an outgoing skb has the correct MCTP extension data set */
 static void mctp_test_packet_flow(struct kunit *test)
 {
+	struct mctp_test_pktqueue tpq;
 	struct sk_buff *skb, *skb2;
-	struct mctp_test_route *rt;
 	struct mctp_test_dev *dev;
+	struct mctp_dst dst;
 	struct mctp_flow *flow;
 	struct socket *sock;
-	u8 dst = 8;
+	u8 dst_eid = 8;
 	int n, rc;
 
-	mctp_test_flow_init(test, &dev, &rt, &sock, &skb, 30);
+	mctp_test_flow_init(test, &dev, &dst, &tpq, &sock, &skb, 30);
 
-	rc = mctp_local_output(sock->sk, &rt->rt, skb, dst, MCTP_TAG_OWNER);
+	rc = mctp_local_output(sock->sk, &dst, skb, dst_eid, MCTP_TAG_OWNER);
 	KUNIT_ASSERT_EQ(test, rc, 0);
 
-	n = rt->pkts.qlen;
+	n = tpq.pkts.qlen;
 	KUNIT_ASSERT_EQ(test, n, 1);
 
-	skb2 = skb_dequeue(&rt->pkts);
+	skb2 = skb_dequeue(&tpq.pkts);
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, skb2);
 
 	flow = skb_ext_find(skb2, SKB_EXT_MCTP);
@@ -1105,7 +1000,7 @@ static void mctp_test_packet_flow(struct kunit *test)
 	KUNIT_ASSERT_PTR_EQ(test, flow->key->sk, sock->sk);
 
 	kfree_skb(skb2);
-	mctp_test_flow_fini(test, dev, rt, sock);
+	mctp_test_flow_fini(test, dev, &dst, &tpq, sock);
 }
 
 /* test that outgoing skbs, after fragmentation, all have the correct MCTP
@@ -1113,26 +1008,27 @@ static void mctp_test_packet_flow(struct kunit *test)
  */
 static void mctp_test_fragment_flow(struct kunit *test)
 {
+	struct mctp_test_pktqueue tpq;
 	struct mctp_flow *flows[2];
 	struct sk_buff *tx_skbs[2];
-	struct mctp_test_route *rt;
 	struct mctp_test_dev *dev;
+	struct mctp_dst dst;
 	struct sk_buff *skb;
 	struct socket *sock;
-	u8 dst = 8;
+	u8 dst_eid = 8;
 	int n, rc;
 
-	mctp_test_flow_init(test, &dev, &rt, &sock, &skb, 100);
+	mctp_test_flow_init(test, &dev, &dst, &tpq, &sock, &skb, 100);
 
-	rc = mctp_local_output(sock->sk, &rt->rt, skb, dst, MCTP_TAG_OWNER);
+	rc = mctp_local_output(sock->sk, &dst, skb, dst_eid, MCTP_TAG_OWNER);
 	KUNIT_ASSERT_EQ(test, rc, 0);
 
-	n = rt->pkts.qlen;
+	n = tpq.pkts.qlen;
 	KUNIT_ASSERT_EQ(test, n, 2);
 
 	/* both resulting packets should have the same flow data */
-	tx_skbs[0] = skb_dequeue(&rt->pkts);
-	tx_skbs[1] = skb_dequeue(&rt->pkts);
+	tx_skbs[0] = skb_dequeue(&tpq.pkts);
+	tx_skbs[1] = skb_dequeue(&tpq.pkts);
 
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, tx_skbs[0]);
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, tx_skbs[1]);
@@ -1148,7 +1044,7 @@ static void mctp_test_fragment_flow(struct kunit *test)
 
 	kfree_skb(tx_skbs[0]);
 	kfree_skb(tx_skbs[1]);
-	mctp_test_flow_fini(test, dev, rt, sock);
+	mctp_test_flow_fini(test, dev, &dst, &tpq, sock);
 }
 
 #else
@@ -1166,15 +1062,16 @@ static void mctp_test_fragment_flow(struct kunit *test)
 /* Test that outgoing skbs cause a suitable tag to be created */
 static void mctp_test_route_output_key_create(struct kunit *test)
 {
+	const u8 dst_eid = 26, src_eid = 15;
+	struct mctp_test_pktqueue tpq;
 	const unsigned int netid = 50;
-	const u8 dst = 26, src = 15;
-	struct mctp_test_route *rt;
 	struct mctp_test_dev *dev;
 	struct mctp_sk_key *key;
 	struct netns_mctp *mns;
 	unsigned long flags;
 	struct socket *sock;
 	struct sk_buff *skb;
+	struct mctp_dst dst;
 	bool empty, single;
 	const int len = 2;
 	int rc;
@@ -1183,15 +1080,14 @@ static void mctp_test_route_output_key_create(struct kunit *test)
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, dev);
 	WRITE_ONCE(dev->mdev->net, netid);
 
-	rt = mctp_test_create_route(&init_net, dev->mdev, dst, 68);
-	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, rt);
+	mctp_test_dst_setup(test, &dst, dev, &tpq, 68);
 
 	rc = sock_create_kern(&init_net, AF_MCTP, SOCK_DGRAM, 0, &sock);
 	KUNIT_ASSERT_EQ(test, rc, 0);
 
 	dev->mdev->addrs = kmalloc(sizeof(u8), GFP_KERNEL);
 	dev->mdev->num_addrs = 1;
-	dev->mdev->addrs[0] = src;
+	dev->mdev->addrs[0] = src_eid;
 
 	skb = alloc_skb(sizeof(struct mctp_hdr) + 1 + len, GFP_KERNEL);
 	KUNIT_ASSERT_TRUE(test, skb);
@@ -1199,8 +1095,6 @@ static void mctp_test_route_output_key_create(struct kunit *test)
 	skb_reserve(skb, sizeof(struct mctp_hdr) + 1 + len);
 	memset(skb_put(skb, len), 0, len);
 
-	refcount_inc(&rt->rt.refs);
-
 	mns = &sock_net(sock->sk)->mctp;
 
 	/* We assume we're starting from an empty keys list, which requires
@@ -1211,7 +1105,7 @@ static void mctp_test_route_output_key_create(struct kunit *test)
 	spin_unlock_irqrestore(&mns->keys_lock, flags);
 	KUNIT_ASSERT_TRUE(test, empty);
 
-	rc = mctp_local_output(sock->sk, &rt->rt, skb, dst, MCTP_TAG_OWNER);
+	rc = mctp_local_output(sock->sk, &dst, skb, dst_eid, MCTP_TAG_OWNER);
 	KUNIT_ASSERT_EQ(test, rc, 0);
 
 	key = NULL;
@@ -1227,16 +1121,295 @@ static void mctp_test_route_output_key_create(struct kunit *test)
 	KUNIT_ASSERT_TRUE(test, single);
 
 	KUNIT_EXPECT_EQ(test, key->net, netid);
-	KUNIT_EXPECT_EQ(test, key->local_addr, src);
-	KUNIT_EXPECT_EQ(test, key->peer_addr, dst);
+	KUNIT_EXPECT_EQ(test, key->local_addr, src_eid);
+	KUNIT_EXPECT_EQ(test, key->peer_addr, dst_eid);
 	/* key has incoming tag, so inverse of what we sent */
 	KUNIT_EXPECT_FALSE(test, key->tag & MCTP_TAG_OWNER);
 
 	sock_release(sock);
-	mctp_test_route_destroy(test, rt);
+	mctp_test_dst_release(&dst, &tpq);
 	mctp_test_destroy_dev(dev);
 }
 
+static void mctp_test_route_extaddr_input(struct kunit *test)
+{
+	static const unsigned char haddr[] = { 0xaa, 0x55 };
+	struct mctp_test_pktqueue tpq;
+	struct mctp_skb_cb *cb, *cb2;
+	const unsigned int len = 40;
+	struct mctp_test_dev *dev;
+	struct sk_buff *skb, *skb2;
+	struct mctp_dst dst;
+	struct mctp_hdr hdr;
+	struct socket *sock;
+	int rc;
+
+	hdr.ver = 1;
+	hdr.src = 10;
+	hdr.dest = 8;
+	hdr.flags_seq_tag = FL_S | FL_E | FL_TO;
+
+	__mctp_route_test_init(test, &dev, &dst, &tpq, &sock, MCTP_NET_ANY);
+
+	skb = mctp_test_create_skb(&hdr, len);
+	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, skb);
+
+	/* set our hardware addressing data */
+	cb = mctp_cb(skb);
+	memcpy(cb->haddr, haddr, sizeof(haddr));
+	cb->halen = sizeof(haddr);
+
+	mctp_test_skb_set_dev(skb, dev);
+
+	rc = mctp_dst_input(&dst, skb);
+	KUNIT_ASSERT_EQ(test, rc, 0);
+
+	mctp_test_dst_release(&dst, &tpq);
+
+	skb2 = skb_recv_datagram(sock->sk, MSG_DONTWAIT, &rc);
+	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, skb2);
+	KUNIT_ASSERT_EQ(test, skb2->len, len);
+
+	cb2 = mctp_cb(skb2);
+
+	/* Received SKB should have the hardware addressing as set above.
+	 * We're likely to have the same actual cb here (ie., cb == cb2),
+	 * but it's the comparison that we care about
+	 */
+	KUNIT_EXPECT_EQ(test, cb2->halen, sizeof(haddr));
+	KUNIT_EXPECT_MEMEQ(test, cb2->haddr, haddr, sizeof(haddr));
+
+	skb_free_datagram(sock->sk, skb2);
+	mctp_test_destroy_dev(dev);
+}
+
+static void mctp_test_route_gw_lookup(struct kunit *test)
+{
+	struct mctp_test_route *rt1, *rt2;
+	struct mctp_dst dst = { 0 };
+	struct mctp_test_dev *dev;
+	int rc;
+
+	dev = mctp_test_create_dev();
+	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, dev);
+
+	/* 8 (local) -> 10 (gateway) via 9 (direct) */
+	rt1 = mctp_test_create_route_direct(&init_net, dev->mdev, 9, 0);
+	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, rt1);
+	rt2 = mctp_test_create_route_gw(&init_net, dev->mdev->net, 10, 9, 0);
+	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, rt2);
+
+	rc = mctp_route_lookup(&init_net, dev->mdev->net, 10, &dst);
+	KUNIT_EXPECT_EQ(test, rc, 0);
+	KUNIT_EXPECT_PTR_EQ(test, dst.dev, dev->mdev);
+	KUNIT_EXPECT_EQ(test, dst.mtu, dev->ndev->mtu);
+	KUNIT_EXPECT_EQ(test, dst.nexthop, 9);
+	KUNIT_EXPECT_EQ(test, dst.halen, 0);
+
+	mctp_dst_release(&dst);
+
+	mctp_test_route_destroy(test, rt2);
+	mctp_test_route_destroy(test, rt1);
+	mctp_test_destroy_dev(dev);
+}
+
+static void mctp_test_route_gw_loop(struct kunit *test)
+{
+	struct mctp_test_route *rt1, *rt2;
+	struct mctp_dst dst = { 0 };
+	struct mctp_test_dev *dev;
+	int rc;
+
+	dev = mctp_test_create_dev();
+	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, dev);
+
+	/* two routes using each other as the gw */
+	rt1 = mctp_test_create_route_gw(&init_net, dev->mdev->net, 9, 10, 0);
+	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, rt1);
+	rt2 = mctp_test_create_route_gw(&init_net, dev->mdev->net, 10, 9, 0);
+	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, rt2);
+
+	/* this should fail, rather than infinite-loop */
+	rc = mctp_route_lookup(&init_net, dev->mdev->net, 10, &dst);
+	KUNIT_EXPECT_NE(test, rc, 0);
+
+	mctp_test_route_destroy(test, rt2);
+	mctp_test_route_destroy(test, rt1);
+	mctp_test_destroy_dev(dev);
+}
+
+struct mctp_route_gw_mtu_test {
+	/* working away from the local stack */
+	unsigned int dev, neigh, gw, dst;
+	unsigned int exp;
+};
+
+static void mctp_route_gw_mtu_to_desc(const struct mctp_route_gw_mtu_test *t,
+				      char *desc)
+{
+	sprintf(desc, "dev %d, neigh %d, gw %d, dst %d -> %d",
+		t->dev, t->neigh, t->gw, t->dst, t->exp);
+}
+
+static const struct mctp_route_gw_mtu_test mctp_route_gw_mtu_tests[] = {
+	/* no route-specific MTUs */
+	{ 68, 0, 0, 0, 68 },
+	{ 100, 0, 0, 0, 100 },
+	/* one route MTU (smaller than dev mtu), others unrestricted */
+	{ 100, 68, 0, 0, 68 },
+	{ 100, 0, 68, 0, 68 },
+	{ 100, 0, 0, 68, 68 },
+	/* smallest applied, regardless of order */
+	{ 100, 99, 98, 68, 68 },
+	{ 99, 100, 98, 68, 68 },
+	{ 98, 99, 100, 68, 68 },
+	{ 68, 98, 99, 100, 68 },
+};
+
+KUNIT_ARRAY_PARAM(mctp_route_gw_mtu, mctp_route_gw_mtu_tests,
+		  mctp_route_gw_mtu_to_desc);
+
+static void mctp_test_route_gw_mtu(struct kunit *test)
+{
+	const struct mctp_route_gw_mtu_test *mtus = test->param_value;
+	struct mctp_test_route *rt1, *rt2, *rt3;
+	struct mctp_dst dst = { 0 };
+	struct mctp_test_dev *dev;
+	struct mctp_dev *mdev;
+	unsigned int netid;
+	int rc;
+
+	dev = mctp_test_create_dev();
+	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, dev);
+	dev->ndev->mtu = mtus->dev;
+	mdev = dev->mdev;
+	netid = mdev->net;
+
+	/* 8 (local) -> 11 (dst) via 10 (gw) via 9 (neigh) */
+	rt1 = mctp_test_create_route_direct(&init_net, mdev, 9, mtus->neigh);
+	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, rt1);
+
+	rt2 = mctp_test_create_route_gw(&init_net, netid, 10, 9, mtus->gw);
+	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, rt2);
+
+	rt3 = mctp_test_create_route_gw(&init_net, netid, 11, 10, mtus->dst);
+	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, rt3);
+
+	rc = mctp_route_lookup(&init_net, dev->mdev->net, 11, &dst);
+	KUNIT_EXPECT_EQ(test, rc, 0);
+	KUNIT_EXPECT_EQ(test, dst.mtu, mtus->exp);
+
+	mctp_dst_release(&dst);
+
+	mctp_test_route_destroy(test, rt3);
+	mctp_test_route_destroy(test, rt2);
+	mctp_test_route_destroy(test, rt1);
+	mctp_test_destroy_dev(dev);
+}
+
+#define MCTP_TEST_LLADDR_LEN 2
+struct mctp_test_llhdr {
+	unsigned int magic;
+	unsigned char src[MCTP_TEST_LLADDR_LEN];
+	unsigned char dst[MCTP_TEST_LLADDR_LEN];
+};
+
+static const unsigned int mctp_test_llhdr_magic = 0x5c78339c;
+
+static int test_dev_header_create(struct sk_buff *skb, struct net_device *dev,
+				  unsigned short type, const void *daddr,
+				  const void *saddr, unsigned int len)
+{
+	struct kunit *test = current->kunit_test;
+	struct mctp_test_llhdr *hdr;
+
+	hdr = skb_push(skb, sizeof(*hdr));
+	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, hdr);
+	skb_reset_mac_header(skb);
+
+	hdr->magic = mctp_test_llhdr_magic;
+	memcpy(&hdr->src, saddr, sizeof(hdr->src));
+	memcpy(&hdr->dst, daddr, sizeof(hdr->dst));
+
+	return 0;
+}
+
+/* Test the dst_output path for a gateway-routed skb: we should have it
+ * lookup the nexthop EID in the neighbour table, and call into
+ * header_ops->create to resolve that to a lladdr. Our mock header_ops->create
+ * will just set a synthetic link-layer header, which we check after transmit.
+ */
+static void mctp_test_route_gw_output(struct kunit *test)
+{
+	const unsigned char haddr_self[MCTP_TEST_LLADDR_LEN] = { 0xaa, 0x03 };
+	const unsigned char haddr_peer[MCTP_TEST_LLADDR_LEN] = { 0xaa, 0x02 };
+	const struct header_ops ops = {
+		.create = test_dev_header_create,
+	};
+	struct mctp_neigh neigh = { 0 };
+	struct mctp_test_llhdr *ll_hdr;
+	struct mctp_dst dst = { 0 };
+	struct mctp_hdr hdr = { 0 };
+	struct mctp_test_dev *dev;
+	struct sk_buff *skb;
+	unsigned char *buf;
+	int i, rc;
+
+	dev = mctp_test_create_dev_lladdr(sizeof(haddr_self), haddr_self);
+	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, dev);
+	dev->ndev->header_ops = &ops;
+
+	dst.dev = dev->mdev;
+	__mctp_dev_get(dst.dev->dev);
+	dst.mtu = 68;
+	dst.nexthop = 9;
+
+	/* simple mctp_neigh_add for the gateway (not dest!) endpoint */
+	INIT_LIST_HEAD(&neigh.list);
+	neigh.dev = dev->mdev;
+	mctp_dev_hold(dev->mdev);
+	neigh.eid = 9;
+	neigh.source = MCTP_NEIGH_STATIC;
+	memcpy(neigh.ha, haddr_peer, sizeof(haddr_peer));
+	list_add_rcu(&neigh.list, &init_net.mctp.neighbours);
+
+	hdr.ver = 1;
+	hdr.src = 8;
+	hdr.dest = 10;
+	hdr.flags_seq_tag = FL_S | FL_E | FL_TO;
+
+	/* construct enough for a future link-layer header, the provided
+	 * mctp header, and 4 bytes of data
+	 */
+	skb = alloc_skb(sizeof(*ll_hdr) + sizeof(hdr) + 4, GFP_KERNEL);
+	skb->dev = dev->ndev;
+	__mctp_cb(skb);
+
+	skb_reserve(skb, sizeof(*ll_hdr));
+
+	memcpy(skb_put(skb, sizeof(hdr)), &hdr, sizeof(hdr));
+	buf = skb_put(skb, 4);
+	for (i = 0; i < 4; i++)
+		buf[i] = i;
+
+	/* extra ref over the dev_xmit */
+	skb_get(skb);
+
+	rc = mctp_dst_output(&dst, skb);
+	KUNIT_EXPECT_EQ(test, rc, 0);
+
+	mctp_dst_release(&dst);
+	list_del_rcu(&neigh.list);
+	mctp_dev_put(dev->mdev);
+
+	/* check that we have our header created with the correct neighbour */
+	ll_hdr = (void *)skb_mac_header(skb);
+	KUNIT_EXPECT_EQ(test, ll_hdr->magic, mctp_test_llhdr_magic);
+	KUNIT_EXPECT_MEMEQ(test, ll_hdr->src, haddr_self, sizeof(haddr_self));
+	KUNIT_EXPECT_MEMEQ(test, ll_hdr->dst, haddr_peer, sizeof(haddr_peer));
+	kfree_skb(skb);
+}
+
 static struct kunit_case mctp_test_cases[] = {
 	KUNIT_CASE_PARAM(mctp_test_fragment, mctp_frag_gen_params),
 	KUNIT_CASE_PARAM(mctp_test_rx_input, mctp_rx_input_gen_params),
@@ -1253,11 +1426,16 @@ static struct kunit_case mctp_test_cases[] = {
 	KUNIT_CASE(mctp_test_fragment_flow),
 	KUNIT_CASE(mctp_test_route_output_key_create),
 	KUNIT_CASE(mctp_test_route_input_cloned_frag),
+	KUNIT_CASE(mctp_test_route_extaddr_input),
+	KUNIT_CASE(mctp_test_route_gw_lookup),
+	KUNIT_CASE(mctp_test_route_gw_loop),
+	KUNIT_CASE_PARAM(mctp_test_route_gw_mtu, mctp_route_gw_mtu_gen_params),
+	KUNIT_CASE(mctp_test_route_gw_output),
 	{}
 };
 
 static struct kunit_suite mctp_test_suite = {
-	.name = "mctp",
+	.name = "mctp-route",
 	.test_cases = mctp_test_cases,
 };
 
diff --git a/net/mctp/test/sock-test.c b/net/mctp/test/sock-test.c
new file mode 100644
index 000000000000..4eb3a724dca3
--- /dev/null
+++ b/net/mctp/test/sock-test.c
@@ -0,0 +1,229 @@
+// SPDX-License-Identifier: GPL-2.0
+
+#include <kunit/static_stub.h>
+#include <kunit/test.h>
+
+#include <linux/socket.h>
+#include <linux/spinlock.h>
+
+#include "utils.h"
+
+static const u8 dev_default_lladdr[] = { 0x01, 0x02 };
+
+/* helper for simple sock setup: single device, with dev_default_lladdr as its
+ * hardware address, assigned with a local EID 8, and a route to EID 9
+ */
+static void __mctp_sock_test_init(struct kunit *test,
+				  struct mctp_test_dev **devp,
+				  struct mctp_test_route **rtp,
+				  struct socket **sockp)
+{
+	struct mctp_test_route *rt;
+	struct mctp_test_dev *dev;
+	struct socket *sock;
+	unsigned long flags;
+	u8 *addrs;
+	int rc;
+
+	dev = mctp_test_create_dev_lladdr(sizeof(dev_default_lladdr),
+					  dev_default_lladdr);
+	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, dev);
+
+	addrs = kmalloc(1, GFP_KERNEL);
+	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, addrs);
+	addrs[0] = 8;
+
+	spin_lock_irqsave(&dev->mdev->addrs_lock, flags);
+	dev->mdev->num_addrs = 1;
+	swap(addrs, dev->mdev->addrs);
+	spin_unlock_irqrestore(&dev->mdev->addrs_lock, flags);
+
+	kfree(addrs);
+
+	rt = mctp_test_create_route_direct(dev_net(dev->ndev), dev->mdev, 9, 0);
+	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, rt);
+
+	rc = sock_create_kern(&init_net, AF_MCTP, SOCK_DGRAM, 0, &sock);
+	KUNIT_ASSERT_EQ(test, rc, 0);
+
+	*devp = dev;
+	*rtp = rt;
+	*sockp = sock;
+}
+
+static void __mctp_sock_test_fini(struct kunit *test,
+				  struct mctp_test_dev *dev,
+				  struct mctp_test_route *rt,
+				  struct socket *sock)
+{
+	sock_release(sock);
+	mctp_test_route_destroy(test, rt);
+	mctp_test_destroy_dev(dev);
+}
+
+struct mctp_test_sock_local_output_config {
+	struct mctp_test_dev *dev;
+	size_t halen;
+	u8 haddr[MAX_ADDR_LEN];
+	bool invoked;
+	int rc;
+};
+
+static int mctp_test_sock_local_output(struct sock *sk,
+				       struct mctp_dst *dst,
+				       struct sk_buff *skb,
+				       mctp_eid_t daddr, u8 req_tag)
+{
+	struct kunit *test = kunit_get_current_test();
+	struct mctp_test_sock_local_output_config *cfg = test->priv;
+
+	KUNIT_EXPECT_PTR_EQ(test, dst->dev, cfg->dev->mdev);
+	KUNIT_EXPECT_EQ(test, dst->halen, cfg->halen);
+	KUNIT_EXPECT_MEMEQ(test, dst->haddr, cfg->haddr, dst->halen);
+
+	cfg->invoked = true;
+
+	kfree_skb(skb);
+
+	return cfg->rc;
+}
+
+static void mctp_test_sock_sendmsg_extaddr(struct kunit *test)
+{
+	struct sockaddr_mctp_ext addr = {
+		.smctp_base = {
+			.smctp_family = AF_MCTP,
+			.smctp_tag = MCTP_TAG_OWNER,
+			.smctp_network = MCTP_NET_ANY,
+		},
+	};
+	struct mctp_test_sock_local_output_config cfg = { 0 };
+	u8 haddr[] = { 0xaa, 0x01 };
+	u8 buf[4] = { 0, 1, 2, 3 };
+	struct mctp_test_route *rt;
+	struct msghdr msg = { 0 };
+	struct mctp_test_dev *dev;
+	struct mctp_sock *msk;
+	struct socket *sock;
+	ssize_t send_len;
+	struct kvec vec = {
+		.iov_base = buf,
+		.iov_len = sizeof(buf),
+	};
+
+	__mctp_sock_test_init(test, &dev, &rt, &sock);
+
+	/* Expect to see the dst configured up with the addressing data we
+	 * provide in the struct sockaddr_mctp_ext
+	 */
+	cfg.dev = dev;
+	cfg.halen = sizeof(haddr);
+	memcpy(cfg.haddr, haddr, sizeof(haddr));
+
+	test->priv = &cfg;
+
+	kunit_activate_static_stub(test, mctp_local_output,
+				   mctp_test_sock_local_output);
+
+	/* enable and configure direct addressing */
+	msk = container_of(sock->sk, struct mctp_sock, sk);
+	msk->addr_ext = true;
+
+	addr.smctp_ifindex = dev->ndev->ifindex;
+	addr.smctp_halen = sizeof(haddr);
+	memcpy(addr.smctp_haddr, haddr, sizeof(haddr));
+
+	msg.msg_name = &addr;
+	msg.msg_namelen = sizeof(addr);
+
+	iov_iter_kvec(&msg.msg_iter, ITER_SOURCE, &vec, 1, sizeof(buf));
+	send_len = mctp_sendmsg(sock, &msg, sizeof(buf));
+	KUNIT_EXPECT_EQ(test, send_len, sizeof(buf));
+	KUNIT_EXPECT_TRUE(test, cfg.invoked);
+
+	__mctp_sock_test_fini(test, dev, rt, sock);
+}
+
+static void mctp_test_sock_recvmsg_extaddr(struct kunit *test)
+{
+	struct sockaddr_mctp_ext recv_addr = { 0 };
+	u8 rcv_buf[1], rcv_data[] = { 0, 1 };
+	u8 haddr[] = { 0xaa, 0x02 };
+	struct mctp_test_route *rt;
+	struct mctp_test_dev *dev;
+	struct mctp_skb_cb *cb;
+	struct mctp_sock *msk;
+	struct sk_buff *skb;
+	struct mctp_hdr hdr;
+	struct socket *sock;
+	struct msghdr msg;
+	ssize_t recv_len;
+	int rc;
+	struct kvec vec = {
+		.iov_base = rcv_buf,
+		.iov_len = sizeof(rcv_buf),
+	};
+
+	__mctp_sock_test_init(test, &dev, &rt, &sock);
+
+	/* enable extended addressing on recv */
+	msk = container_of(sock->sk, struct mctp_sock, sk);
+	msk->addr_ext = true;
+
+	/* base incoming header, using a nul-EID dest */
+	hdr.ver = 1;
+	hdr.dest = 0;
+	hdr.src = 9;
+	hdr.flags_seq_tag = MCTP_HDR_FLAG_SOM | MCTP_HDR_FLAG_EOM |
+			    MCTP_HDR_FLAG_TO;
+
+	skb = mctp_test_create_skb_data(&hdr, &rcv_data);
+	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, skb);
+
+	mctp_test_skb_set_dev(skb, dev);
+
+	/* set incoming extended address data */
+	cb = mctp_cb(skb);
+	cb->halen = sizeof(haddr);
+	cb->ifindex = dev->ndev->ifindex;
+	memcpy(cb->haddr, haddr, sizeof(haddr));
+
+	/* Deliver to socket. The route input path pulls the network header,
+	 * leaving skb data at type byte onwards. recvmsg will consume the
+	 * type for addr.smctp_type
+	 */
+	skb_pull(skb, sizeof(hdr));
+	rc = sock_queue_rcv_skb(sock->sk, skb);
+	KUNIT_ASSERT_EQ(test, rc, 0);
+
+	msg.msg_name = &recv_addr;
+	msg.msg_namelen = sizeof(recv_addr);
+	iov_iter_kvec(&msg.msg_iter, ITER_DEST, &vec, 1, sizeof(rcv_buf));
+
+	recv_len = mctp_recvmsg(sock, &msg, sizeof(rcv_buf),
+				MSG_DONTWAIT | MSG_TRUNC);
+
+	KUNIT_EXPECT_EQ(test, recv_len, sizeof(rcv_buf));
+
+	/* expect our extended address to be populated from hdr and cb */
+	KUNIT_EXPECT_EQ(test, msg.msg_namelen, sizeof(recv_addr));
+	KUNIT_EXPECT_EQ(test, recv_addr.smctp_base.smctp_family, AF_MCTP);
+	KUNIT_EXPECT_EQ(test, recv_addr.smctp_ifindex, dev->ndev->ifindex);
+	KUNIT_EXPECT_EQ(test, recv_addr.smctp_halen, sizeof(haddr));
+	KUNIT_EXPECT_MEMEQ(test, recv_addr.smctp_haddr, haddr, sizeof(haddr));
+
+	__mctp_sock_test_fini(test, dev, rt, sock);
+}
+
+static struct kunit_case mctp_test_cases[] = {
+	KUNIT_CASE(mctp_test_sock_sendmsg_extaddr),
+	KUNIT_CASE(mctp_test_sock_recvmsg_extaddr),
+	{}
+};
+
+static struct kunit_suite mctp_test_suite = {
+	.name = "mctp-sock",
+	.test_cases = mctp_test_cases,
+};
+
+kunit_test_suite(mctp_test_suite);
diff --git a/net/mctp/test/utils.c b/net/mctp/test/utils.c
index 565763eb0211..01f5af416b81 100644
--- a/net/mctp/test/utils.c
+++ b/net/mctp/test/utils.c
@@ -26,19 +26,22 @@ static void mctp_test_dev_setup(struct net_device *ndev)
 	ndev->type = ARPHRD_MCTP;
 	ndev->mtu = MCTP_DEV_TEST_MTU;
 	ndev->hard_header_len = 0;
-	ndev->addr_len = 0;
 	ndev->tx_queue_len = DEFAULT_TX_QUEUE_LEN;
 	ndev->flags = IFF_NOARP;
 	ndev->netdev_ops = &mctp_test_netdev_ops;
 	ndev->needs_free_netdev = true;
 }
 
-struct mctp_test_dev *mctp_test_create_dev(void)
+static struct mctp_test_dev *__mctp_test_create_dev(unsigned short lladdr_len,
+						    const unsigned char *lladdr)
 {
 	struct mctp_test_dev *dev;
 	struct net_device *ndev;
 	int rc;
 
+	if (WARN_ON(lladdr_len > MAX_ADDR_LEN))
+		return NULL;
+
 	ndev = alloc_netdev(sizeof(*dev), "mctptest%d", NET_NAME_ENUM,
 			    mctp_test_dev_setup);
 	if (!ndev)
@@ -46,6 +49,8 @@ struct mctp_test_dev *mctp_test_create_dev(void)
 
 	dev = netdev_priv(ndev);
 	dev->ndev = ndev;
+	ndev->addr_len = lladdr_len;
+	dev_addr_set(ndev, lladdr);
 
 	rc = register_netdev(ndev);
 	if (rc) {
@@ -61,8 +66,195 @@ struct mctp_test_dev *mctp_test_create_dev(void)
 	return dev;
 }
 
+struct mctp_test_dev *mctp_test_create_dev(void)
+{
+	return __mctp_test_create_dev(0, NULL);
+}
+
+struct mctp_test_dev *mctp_test_create_dev_lladdr(unsigned short lladdr_len,
+						  const unsigned char *lladdr)
+{
+	return __mctp_test_create_dev(lladdr_len, lladdr);
+}
+
 void mctp_test_destroy_dev(struct mctp_test_dev *dev)
 {
 	mctp_dev_put(dev->mdev);
 	unregister_netdev(dev->ndev);
 }
+
+static const unsigned int test_pktqueue_magic = 0x5f713aef;
+
+void mctp_test_pktqueue_init(struct mctp_test_pktqueue *tpq)
+{
+	tpq->magic = test_pktqueue_magic;
+	skb_queue_head_init(&tpq->pkts);
+}
+
+static int mctp_test_dst_output(struct mctp_dst *dst, struct sk_buff *skb)
+{
+	struct kunit *test = current->kunit_test;
+	struct mctp_test_pktqueue *tpq = test->priv;
+
+	KUNIT_ASSERT_EQ(test, tpq->magic, test_pktqueue_magic);
+
+	skb_queue_tail(&tpq->pkts, skb);
+
+	return 0;
+}
+
+/* local version of mctp_route_alloc() */
+static struct mctp_test_route *mctp_route_test_alloc(void)
+{
+	struct mctp_test_route *rt;
+
+	rt = kzalloc(sizeof(*rt), GFP_KERNEL);
+	if (!rt)
+		return NULL;
+
+	INIT_LIST_HEAD(&rt->rt.list);
+	refcount_set(&rt->rt.refs, 1);
+	rt->rt.output = mctp_test_dst_output;
+
+	return rt;
+}
+
+struct mctp_test_route *mctp_test_create_route_direct(struct net *net,
+						      struct mctp_dev *dev,
+						      mctp_eid_t eid,
+						      unsigned int mtu)
+{
+	struct mctp_test_route *rt;
+
+	rt = mctp_route_test_alloc();
+	if (!rt)
+		return NULL;
+
+	rt->rt.min = eid;
+	rt->rt.max = eid;
+	rt->rt.mtu = mtu;
+	rt->rt.type = RTN_UNSPEC;
+	rt->rt.dst_type = MCTP_ROUTE_DIRECT;
+	if (dev)
+		mctp_dev_hold(dev);
+	rt->rt.dev = dev;
+
+	list_add_rcu(&rt->rt.list, &net->mctp.routes);
+
+	return rt;
+}
+
+struct mctp_test_route *mctp_test_create_route_gw(struct net *net,
+						  unsigned int netid,
+						  mctp_eid_t eid,
+						  mctp_eid_t gw,
+						  unsigned int mtu)
+{
+	struct mctp_test_route *rt;
+
+	rt = mctp_route_test_alloc();
+	if (!rt)
+		return NULL;
+
+	rt->rt.min = eid;
+	rt->rt.max = eid;
+	rt->rt.mtu = mtu;
+	rt->rt.type = RTN_UNSPEC;
+	rt->rt.dst_type = MCTP_ROUTE_GATEWAY;
+	rt->rt.gateway.eid = gw;
+	rt->rt.gateway.net = netid;
+
+	list_add_rcu(&rt->rt.list, &net->mctp.routes);
+
+	return rt;
+}
+
+/* Convenience function for our test dst; release with mctp_test_dst_release()
+ */
+void mctp_test_dst_setup(struct kunit *test, struct mctp_dst *dst,
+			 struct mctp_test_dev *dev,
+			 struct mctp_test_pktqueue *tpq, unsigned int mtu)
+{
+	KUNIT_EXPECT_NOT_ERR_OR_NULL(test, dev);
+
+	memset(dst, 0, sizeof(*dst));
+
+	dst->dev = dev->mdev;
+	__mctp_dev_get(dst->dev->dev);
+	dst->mtu = mtu;
+	dst->output = mctp_test_dst_output;
+	mctp_test_pktqueue_init(tpq);
+	test->priv = tpq;
+}
+
+void mctp_test_dst_release(struct mctp_dst *dst,
+			   struct mctp_test_pktqueue *tpq)
+{
+	mctp_dst_release(dst);
+	skb_queue_purge(&tpq->pkts);
+}
+
+void mctp_test_route_destroy(struct kunit *test, struct mctp_test_route *rt)
+{
+	unsigned int refs;
+
+	rtnl_lock();
+	list_del_rcu(&rt->rt.list);
+	rtnl_unlock();
+
+	if (rt->rt.dst_type == MCTP_ROUTE_DIRECT && rt->rt.dev)
+		mctp_dev_put(rt->rt.dev);
+
+	refs = refcount_read(&rt->rt.refs);
+	KUNIT_ASSERT_EQ_MSG(test, refs, 1, "route ref imbalance");
+
+	kfree_rcu(&rt->rt, rcu);
+}
+
+void mctp_test_skb_set_dev(struct sk_buff *skb, struct mctp_test_dev *dev)
+{
+	struct mctp_skb_cb *cb;
+
+	cb = mctp_cb(skb);
+	cb->net = READ_ONCE(dev->mdev->net);
+	skb->dev = dev->ndev;
+}
+
+struct sk_buff *mctp_test_create_skb(const struct mctp_hdr *hdr,
+				     unsigned int data_len)
+{
+	size_t hdr_len = sizeof(*hdr);
+	struct sk_buff *skb;
+	unsigned int i;
+	u8 *buf;
+
+	skb = alloc_skb(hdr_len + data_len, GFP_KERNEL);
+	if (!skb)
+		return NULL;
+
+	__mctp_cb(skb);
+	memcpy(skb_put(skb, hdr_len), hdr, hdr_len);
+
+	buf = skb_put(skb, data_len);
+	for (i = 0; i < data_len; i++)
+		buf[i] = i & 0xff;
+
+	return skb;
+}
+
+struct sk_buff *__mctp_test_create_skb_data(const struct mctp_hdr *hdr,
+					    const void *data, size_t data_len)
+{
+	size_t hdr_len = sizeof(*hdr);
+	struct sk_buff *skb;
+
+	skb = alloc_skb(hdr_len + data_len, GFP_KERNEL);
+	if (!skb)
+		return NULL;
+
+	__mctp_cb(skb);
+	memcpy(skb_put(skb, hdr_len), hdr, hdr_len);
+	memcpy(skb_put(skb, data_len), data, data_len);
+
+	return skb;
+}
diff --git a/net/mctp/test/utils.h b/net/mctp/test/utils.h
index df6aa1c03440..f10d1d9066cc 100644
--- a/net/mctp/test/utils.h
+++ b/net/mctp/test/utils.h
@@ -3,6 +3,11 @@
 #ifndef __NET_MCTP_TEST_UTILS_H
 #define __NET_MCTP_TEST_UTILS_H
 
+#include <uapi/linux/netdevice.h>
+
+#include <net/mctp.h>
+#include <net/mctpdevice.h>
+
 #include <kunit/test.h>
 
 #define MCTP_DEV_TEST_MTU	68
@@ -10,11 +15,50 @@
 struct mctp_test_dev {
 	struct net_device *ndev;
 	struct mctp_dev *mdev;
+
+	unsigned short lladdr_len;
+	unsigned char lladdr[MAX_ADDR_LEN];
 };
 
 struct mctp_test_dev;
 
+struct mctp_test_route {
+	struct mctp_route	rt;
+};
+
+struct mctp_test_pktqueue {
+	unsigned int magic;
+	struct sk_buff_head pkts;
+};
+
 struct mctp_test_dev *mctp_test_create_dev(void);
+struct mctp_test_dev *mctp_test_create_dev_lladdr(unsigned short lladdr_len,
+						  const unsigned char *lladdr);
 void mctp_test_destroy_dev(struct mctp_test_dev *dev);
 
+struct mctp_test_route *mctp_test_create_route_direct(struct net *net,
+						      struct mctp_dev *dev,
+						      mctp_eid_t eid,
+						      unsigned int mtu);
+struct mctp_test_route *mctp_test_create_route_gw(struct net *net,
+						  unsigned int netid,
+						  mctp_eid_t eid,
+						  mctp_eid_t gw,
+						  unsigned int mtu);
+void mctp_test_dst_setup(struct kunit *test, struct mctp_dst *dst,
+			 struct mctp_test_dev *dev,
+			 struct mctp_test_pktqueue *tpq, unsigned int mtu);
+void mctp_test_dst_release(struct mctp_dst *dst,
+			   struct mctp_test_pktqueue *tpq);
+void mctp_test_pktqueue_init(struct mctp_test_pktqueue *tpq);
+void mctp_test_route_destroy(struct kunit *test, struct mctp_test_route *rt);
+void mctp_test_skb_set_dev(struct sk_buff *skb, struct mctp_test_dev *dev);
+struct sk_buff *mctp_test_create_skb(const struct mctp_hdr *hdr,
+				     unsigned int data_len);
+struct sk_buff *__mctp_test_create_skb_data(const struct mctp_hdr *hdr,
+					    const void *data, size_t data_len);
+
+#define mctp_test_create_skb_data(h, d) \
+	__mctp_test_create_skb_data(h, d, sizeof(*d))
+
 #endif /* __NET_MCTP_TEST_UTILS_H */