Merge tag 'net-next-6.17' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next

Pull networking updates from Jakub Kicinski:
 "Core & protocols:

   - Wrap datapath globals into net_aligned_data, to avoid false sharing

   - Preserve MSG_ZEROCOPY in forwarding (e.g. out of a container)

   - Add SO_INQ and SCM_INQ support to AF_UNIX

   - Add SIOCINQ support to AF_VSOCK

   - Add TCP_MAXSEG sockopt to MPTCP

   - Add IPv6 force_forwarding sysctl to enable forwarding per interface

   - Make TCP validation of whether packet fully fits in the receive
     window and the rcv_buf more strict. With increased use of HW
     aggregation a single "packet" can be multiple 100s of kB

   - Add MSG_MORE flag to optimize large TCP transmissions via sockmap,
     improves latency up to 33% for sockmap users

   - Convert TCP send queue handling from tasklet to BH workque

   - Improve BPF iteration over TCP sockets to see each socket exactly
     once

   - Remove obsolete and unused TCP RFC3517/RFC6675 loss recovery code

   - Support enabling kernel threads for NAPI processing on per-NAPI
     instance basis rather than a whole device. Fully stop the kernel
     NAPI thread when threaded NAPI gets disabled. Previously thread
     would stick around until ifdown due to tricky synchronization

   - Allow multicast routing to take effect on locally-generated packets

   - Add output interface argument for End.X in segment routing

   - MCTP: add support for gateway routing, improve bind() handling

   - Don't require rtnl_lock when fetching an IPv6 neighbor over Netlink

   - Add a new neighbor flag ("extern_valid"), which cedes refresh
     responsibilities to userspace. This is needed for EVPN multi-homing
     where a neighbor entry for a multi-homed host needs to be synced
     across all the VTEPs among which the host is multi-homed

   - Support NUD_PERMANENT for proxy neighbor entries

   - Add a new queuing discipline for IETF RFC9332 DualQ Coupled AQM

   - Add sequence numbers to netconsole messages. Unregister
     netconsole's console when all net targets are removed. Code
     refactoring. Add a number of selftests

   - Align IPSec inbound SA lookup to RFC 4301. Only SPI and protocol
     should be used for an inbound SA lookup

   - Support inspecting ref_tracker state via DebugFS

   - Don't force bonding advertisement frames tx to ~333 ms boundaries.
     Add broadcast_neighbor option to send ARP/ND on all bonded links

   - Allow providing upcall pid for the 'execute' command in openvswitch

   - Remove DCCP support from Netfilter's conntrack

   - Disallow multiple packet duplications in the queuing layer

   - Prevent use of deprecated iptables code on PREEMPT_RT

  Driver API:

   - Support RSS and hashing configuration over ethtool Netlink

   - Add dedicated ethtool callbacks for getting and setting hashing
     fields

   - Add support for power budget evaluation strategy in PSE /
     Power-over-Ethernet. Generate Netlink events for overcurrent etc

   - Support DPLL phase offset monitoring across all device inputs.
     Support providing clock reference and SYNC over separate DPLL
     inputs

   - Support traffic classes in devlink rate API for bandwidth
     management

   - Remove rtnl_lock dependency from UDP tunnel port configuration

  Device drivers:

   - Add a new Broadcom driver for 800G Ethernet (bnge)

   - Add a standalone driver for Microchip ZL3073x DPLL

   - Remove IBM's NETIUCV device driver

   - Ethernet high-speed NICs:
      - Broadcom (bnxt):
         - support zero-copy Tx of DMABUF memory
         - take page size into account for page pool recycling rings
      - Intel (100G, ice, idpf):
         - idpf: XDP and AF_XDP support preparations
         - idpf: add flow steering
         - add link_down_events statistic
         - clean up the TSPLL code
         - preparations for live VM migration
      - nVidia/Mellanox:
         - support zero-copy Rx/Tx interfaces (DMABUF and io_uring)
         - optimize context memory usage for matchers
         - expose serial numbers in devlink info
         - support PCIe congestion metrics
      - Meta (fbnic):
         - add 25G, 50G, and 100G link modes to phylink
         - support dumping FW logs
      - Marvell/Cavium:
         - support for CN20K generation of the Octeon chips
      - Amazon:
         - add HW clock (without timestamping, just hypervisor time access)

   - Ethernet virtual:
      - VirtIO net:
         - support segmentation of UDP-tunnel-encapsulated packets
      - Google (gve):
         - support packet timestamping and clock synchronization
      - Microsoft vNIC:
         - add handler for device-originated servicing events
         - allow dynamic MSI-X vector allocation
         - support Tx bandwidth clamping

   - Ethernet NICs consumer, and embedded:
      - AMD:
         - amd-xgbe: hardware timestamping and PTP clock support
      - Broadcom integrated MACs (bcmgenet, bcmasp):
         - use napi_complete_done() return value to support NAPI polling
         - add support for re-starting auto-negotiation
      - Broadcom switches (b53):
         - support BCM5325 switches
         - add bcm63xx EPHY power control
      - Synopsys (stmmac):
         - lots of code refactoring and cleanups
      - TI:
         - icssg-prueth: read firmware-names from device tree
         - icssg: PRP offload support
      - Microchip:
         - lan78xx: convert to PHYLINK for improved PHY and MAC management
         - ksz: add KSZ8463 switch support
      - Intel:
         - support similar queue priority scheme in multi-queue and
           time-sensitive networking (taprio)
         - support packet pre-emption in both
      - RealTek (r8169):
         - enable EEE at 5Gbps on RTL8126
      - Airoha:
         - add PPPoE offload support
         - MDIO bus controller for Airoha AN7583

   - Ethernet PHYs:
      - support for the IPQ5018 internal GE PHY
      - micrel KSZ9477 switch-integrated PHYs:
         - add MDI/MDI-X control support
         - add RX error counters
         - add cable test support
         - add Signal Quality Indicator (SQI) reporting
      - dp83tg720: improve reset handling and reduce link recovery time
      - support bcm54811 (and its MII-Lite interface type)
      - air_en8811h: support resume/suspend
      - support PHY counters for QCA807x and QCA808x
      - support WoL for QCA807x

   - CAN drivers:
      - rcar_canfd: support for Transceiver Delay Compensation
      - kvaser: report FW versions via devlink dev info

   - WiFi:
      - extended regulatory info support (6 GHz)
      - add statistics and beacon monitor for Multi-Link Operation (MLO)
      - support S1G aggregation, improve S1G support
      - add Radio Measurement action fields
      - support per-radio RTS threshold
      - some work around how FIPS affects wifi, which was wrong (RC4 is
        used by TKIP, not only WEP)
      - improvements for unsolicited probe response handling

   - WiFi drivers:
      - RealTek (rtw88):
         - IBSS mode for SDIO devices
      - RealTek (rtw89):
         - BT coexistence for MLO/WiFi7
         - concurrent station + P2P support
         - support for USB devices RTL8851BU/RTL8852BU
      - Intel (iwlwifi):
         - use embedded PNVM in (to be released) FW images to fix
           compatibility issues
         - many cleanups (unused FW APIs, PCIe code, WoWLAN)
         - some FIPS interoperability
      - MediaTek (mt76):
         - firmware recovery improvements
         - more MLO work
      - Qualcomm/Atheros (ath12k):
         - fix scan on multi-radio devices
         - more EHT/Wi-Fi 7 features
         - encapsulation/decapsulation offload
      - Broadcom (brcm80211):
         - support SDIO 43751 device

   - Bluetooth:
      - hci_event: add support for handling LE BIG Sync Lost event
      - ISO: add socket option to report packet seqnum via CMSG
      - ISO: support SCM_TIMESTAMPING for ISO TS

   - Bluetooth drivers:
      - intel_pcie: support Function Level Reset
      - nxpuart: add support for 4M baudrate
      - nxpuart: implement powerup sequence, reset, FW dump, and FW loading"

* tag 'net-next-6.17' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next: (1742 commits)
  dpll: zl3073x: Fix build failure
  selftests: bpf: fix legacy netfilter options
  ipv6: annotate data-races around rt->fib6_nsiblings
  ipv6: fix possible infinite loop in fib6_info_uses_dev()
  ipv6: prevent infinite loop in rt6_nlmsg_size()
  ipv6: add a retry logic in net6_rt_notify()
  vrf: Drop existing dst reference in vrf_ip6_input_dst
  net/sched: taprio: align entry index attr validation with mqprio
  net: fsl_pq_mdio: use dev_err_probe
  selftests: rtnetlink.sh: remove esp4_offload after test
  vsock: remove unnecessary null check in vsock_getname()
  igb: xsk: solve negative overflow of nb_pkts in zerocopy mode
  stmmac: xsk: fix negative overflow of budget in zerocopy mode
  dt-bindings: ieee802154: Convert at86rf230.txt yaml format
  net: dsa: microchip: Disable PTP function of KSZ8463
  net: dsa: microchip: Setup fiber ports for KSZ8463
  net: dsa: microchip: Write switch MAC address differently for KSZ8463
  net: dsa: microchip: Use different registers for KSZ8463
  net: dsa: microchip: Add KSZ8463 switch support to KSZ DSA driver
  dt-bindings: net: dsa: microchip: Add KSZ8463 switch support
  ...

1 files changed, 581 insertions, 217 deletions

diff --git a/net/mctp/test/route-test.c b/net/mctp/test/route-test.c
index 06c1897b685a..fb6b46a952cb 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;
+/* keep clangd happy when compiled outside of the route.c include */
+#include <net/mctp.h>
+#include <net/mctpdevice.h>
 
-	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);
-
-	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 +1121,480 @@ 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);
+
+	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));
+
+	kfree_skb(skb2);
+	__mctp_route_test_fini(test, dev, &dst, &tpq, sock);
+}
+
+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);
+}
+
+struct mctp_bind_lookup_test {
+	/* header of incoming message */
+	struct mctp_hdr hdr;
+	u8 ty;
+	/* mctp network of incoming interface (smctp_network) */
+	unsigned int net;
+
+	/* expected socket, matches .name in lookup_binds, NULL for dropped */
+	const char *expect;
+};
+
+/* Single-packet TO-set message */
+#define LK(src, dst) RX_HDR(1, (src), (dst), FL_S | FL_E | FL_TO)
+
+/* Input message test cases for bind lookup tests.
+ *
+ * 10 and 11 are local EIDs.
+ * 20 and 21 are remote EIDs.
+ */
+static const struct mctp_bind_lookup_test mctp_bind_lookup_tests[] = {
+	/* both local-eid and remote-eid binds, remote eid is preferenced */
+	{ .hdr = LK(20, 10),  .ty = 1, .net = 1, .expect = "remote20" },
+
+	{ .hdr = LK(20, 255), .ty = 1, .net = 1, .expect = "remote20" },
+	{ .hdr = LK(20, 0),   .ty = 1, .net = 1, .expect = "remote20" },
+	{ .hdr = LK(0, 255),  .ty = 1, .net = 1, .expect = "any" },
+	{ .hdr = LK(0, 11),   .ty = 1, .net = 1, .expect = "any" },
+	{ .hdr = LK(0, 0),    .ty = 1, .net = 1, .expect = "any" },
+	{ .hdr = LK(0, 10),   .ty = 1, .net = 1, .expect = "local10" },
+	{ .hdr = LK(21, 10),  .ty = 1, .net = 1, .expect = "local10" },
+	{ .hdr = LK(21, 11),  .ty = 1, .net = 1, .expect = "remote21local11" },
+
+	/* both src and dest set to eid=99. unusual, but accepted
+	 * by MCTP stack currently.
+	 */
+	{ .hdr = LK(99, 99),  .ty = 1, .net = 1, .expect = "any" },
+
+	/* unbound smctp_type */
+	{ .hdr = LK(20, 10),  .ty = 3, .net = 1, .expect = NULL },
+
+	/* smctp_network tests */
+
+	{ .hdr = LK(0, 0),    .ty = 1, .net = 7, .expect = "any" },
+	{ .hdr = LK(21, 10),  .ty = 1, .net = 2, .expect = "any" },
+
+	/* remote EID 20 matches, but MCTP_NET_ANY in "remote20" resolved
+	 * to net=1, so lookup doesn't match "remote20"
+	 */
+	{ .hdr = LK(20, 10),  .ty = 1, .net = 3, .expect = "any" },
+
+	{ .hdr = LK(21, 10),  .ty = 1, .net = 3, .expect = "remote21net3" },
+	{ .hdr = LK(21, 10),  .ty = 1, .net = 4, .expect = "remote21net4" },
+	{ .hdr = LK(21, 10),  .ty = 1, .net = 5, .expect = "remote21net5" },
+
+	{ .hdr = LK(21, 10),  .ty = 1, .net = 5, .expect = "remote21net5" },
+
+	{ .hdr = LK(99, 10),  .ty = 1, .net = 8, .expect = "local10net8" },
+
+	{ .hdr = LK(99, 10),  .ty = 1, .net = 9, .expect = "anynet9" },
+	{ .hdr = LK(0, 0),    .ty = 1, .net = 9, .expect = "anynet9" },
+	{ .hdr = LK(99, 99),  .ty = 1, .net = 9, .expect = "anynet9" },
+	{ .hdr = LK(20, 10),  .ty = 1, .net = 9, .expect = "anynet9" },
+};
+
+/* Binds to create during the lookup tests */
+static const struct mctp_test_bind_setup lookup_binds[] = {
+	/* any address and net, type 1 */
+	{ .name = "any", .bind_addr = MCTP_ADDR_ANY,
+		.bind_net = MCTP_NET_ANY, .bind_type = 1, },
+	/* local eid 10, net 1 (resolved from MCTP_NET_ANY) */
+	{ .name = "local10", .bind_addr = 10,
+		.bind_net = MCTP_NET_ANY, .bind_type = 1, },
+	/* local eid 10, net 8 */
+	{ .name = "local10net8", .bind_addr = 10,
+		.bind_net = 8, .bind_type = 1, },
+	/* any EID, net 9 */
+	{ .name = "anynet9", .bind_addr = MCTP_ADDR_ANY,
+		.bind_net = 9, .bind_type = 1, },
+
+	/* remote eid 20, net 1, any local eid */
+	{ .name = "remote20", .bind_addr = MCTP_ADDR_ANY,
+		.bind_net = MCTP_NET_ANY, .bind_type = 1,
+		.have_peer = true, .peer_addr = 20, .peer_net = MCTP_NET_ANY, },
+
+	/* remote eid 20, net 1, local eid 11 */
+	{ .name = "remote21local11", .bind_addr = 11,
+		.bind_net = MCTP_NET_ANY, .bind_type = 1,
+		.have_peer = true, .peer_addr = 21, .peer_net = MCTP_NET_ANY, },
+
+	/* remote eid 21, specific net=3 for connect() */
+	{ .name = "remote21net3", .bind_addr = MCTP_ADDR_ANY,
+		.bind_net = MCTP_NET_ANY, .bind_type = 1,
+		.have_peer = true, .peer_addr = 21, .peer_net = 3, },
+
+	/* remote eid 21, net 4 for bind, specific net=4 for connect() */
+	{ .name = "remote21net4", .bind_addr = MCTP_ADDR_ANY,
+		.bind_net = 4, .bind_type = 1,
+		.have_peer = true, .peer_addr = 21, .peer_net = 4, },
+
+	/* remote eid 21, net 5 for bind, specific net=5 for connect() */
+	{ .name = "remote21net5", .bind_addr = MCTP_ADDR_ANY,
+		.bind_net = 5, .bind_type = 1,
+		.have_peer = true, .peer_addr = 21, .peer_net = 5, },
+};
+
+static void mctp_bind_lookup_desc(const struct mctp_bind_lookup_test *t,
+				  char *desc)
+{
+	snprintf(desc, KUNIT_PARAM_DESC_SIZE,
+		 "{src %d dst %d ty %d net %d expect %s}",
+		 t->hdr.src, t->hdr.dest, t->ty, t->net, t->expect);
+}
+
+KUNIT_ARRAY_PARAM(mctp_bind_lookup, mctp_bind_lookup_tests,
+		  mctp_bind_lookup_desc);
+
+static void mctp_test_bind_lookup(struct kunit *test)
+{
+	const struct mctp_bind_lookup_test *rx;
+	struct socket *socks[ARRAY_SIZE(lookup_binds)];
+	struct sk_buff *skb_pkt = NULL, *skb_sock = NULL;
+	struct socket *sock_ty0, *sock_expect = NULL;
+	struct mctp_test_pktqueue tpq;
+	struct mctp_test_dev *dev;
+	struct mctp_dst dst;
+	int rc;
+
+	rx = test->param_value;
+
+	__mctp_route_test_init(test, &dev, &dst, &tpq, &sock_ty0, rx->net);
+	/* Create all binds */
+	for (size_t i = 0; i < ARRAY_SIZE(lookup_binds); i++) {
+		mctp_test_bind_run(test, &lookup_binds[i],
+				   &rc, &socks[i]);
+		KUNIT_ASSERT_EQ(test, rc, 0);
+
+		/* Record the expected receive socket */
+		if (rx->expect &&
+		    strcmp(rx->expect, lookup_binds[i].name) == 0) {
+			KUNIT_ASSERT_NULL(test, sock_expect);
+			sock_expect = socks[i];
+		}
+	}
+	KUNIT_ASSERT_EQ(test, !!sock_expect, !!rx->expect);
+
+	/* Create test message */
+	skb_pkt = mctp_test_create_skb_data(&rx->hdr, &rx->ty);
+	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, skb_pkt);
+	mctp_test_skb_set_dev(skb_pkt, dev);
+	mctp_test_pktqueue_init(&tpq);
+
+	rc = mctp_dst_input(&dst, skb_pkt);
+	if (rx->expect) {
+		/* Test the message is received on the expected socket */
+		KUNIT_EXPECT_EQ(test, rc, 0);
+		skb_sock = skb_recv_datagram(sock_expect->sk,
+					     MSG_DONTWAIT, &rc);
+		if (!skb_sock) {
+			/* Find which socket received it instead */
+			for (size_t i = 0; i < ARRAY_SIZE(lookup_binds); i++) {
+				skb_sock = skb_recv_datagram(socks[i]->sk,
+							     MSG_DONTWAIT, &rc);
+				if (skb_sock) {
+					KUNIT_FAIL(test,
+						   "received on incorrect socket '%s', expect '%s'",
+						   lookup_binds[i].name,
+						   rx->expect);
+					goto cleanup;
+				}
+			}
+			KUNIT_FAIL(test, "no message received");
+		}
+	} else {
+		KUNIT_EXPECT_NE(test, rc, 0);
+	}
+
+cleanup:
+	kfree_skb(skb_sock);
+	kfree_skb(skb_pkt);
+
+	/* Drop all binds */
+	for (size_t i = 0; i < ARRAY_SIZE(lookup_binds); i++)
+		sock_release(socks[i]);
+
+	__mctp_route_test_fini(test, dev, &dst, &tpq, sock_ty0);
+}
+
 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 +1611,17 @@ 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),
+	KUNIT_CASE_PARAM(mctp_test_bind_lookup, mctp_bind_lookup_gen_params),
 	{}
 };
 
 static struct kunit_suite mctp_test_suite = {
-	.name = "mctp",
+	.name = "mctp-route",
 	.test_cases = mctp_test_cases,
 };