zebra: set connected route metric based on the devaddr metric

MACVLAN devices are typically used for applications such as VRR/VRRP that
require a second MAC address (virtual). These devices have a corresponding
SVI/VLAN device -
root@TORC11:~# ip addr show vlan1002
39: vlan1002@bridge: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 9152 qdisc noqueue master vrf1 state UP group default
    link/ether 00:02:00:00:00:2e brd ff:ff:ff:ff:ff:ff
    inet6 2001:aa:1::2/64 scope global
       valid_lft forever preferred_lft forever
root@TORC11:~# ip addr show vlan1002-v0
40: vlan1002-v0@vlan1002: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 9152 qdisc noqueue master vrf1 state UP group default
    link/ether 00:00:5e:00:01:01 brd ff:ff:ff:ff:ff:ff
    inet6 2001:aa:1::a/64 metric 1024 scope global
       valid_lft forever preferred_lft forever
root@TORC11:~#

The macvlan device is used primarily for RX (VR-IP/VR-MAC). And TX is via
the SVI. To acheive that functionality the macvlan network's metric
is set to a higher value.

Zebra currently ignores the devaddr metric sent by the kernel and hardcodes
it to 0. This commit eliminates that hardcoding. If the devaddr metric
is available (METRIC_MAX) it is used for setting up the connected route
otherwise we fallback to the dev/interface metric.

Setting the macvlan metric to a higher value ensures that zebra will always
select the connected route on the SVI (and subsequently use it for next hop
resolution etc.) -
root@TORC11:~# vtysh -c "show ip route vrf vrf1 2001:aa:1::/64"
Routing entry for 2001:aa:1::/64
  Known via "connected", distance 0, metric 1024, vrf vrf1
  Last update 11:30:56 ago
  * directly connected, vlan1002-v0

Routing entry for 2001:aa:1::/64
  Known via "connected", distance 0, metric 0, vrf vrf1, best
  Last update 11:30:56 ago
  * directly connected, vlan1002

root@TORC11:~#

Ticket: CM-23511
Signed-off-by: Anuradha Karuppiah <anuradhak@cumulusnetworks.com>

include/linux/if_addr.h

@@ -0,0 +1,69 @@
+/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
+#ifndef __LINUX_IF_ADDR_H
+#define __LINUX_IF_ADDR_H
+
+#include <linux/types.h>
+#include <linux/netlink.h>
+
+struct ifaddrmsg {
+	__u8		ifa_family;
+	__u8		ifa_prefixlen;	/* The prefix length		*/
+	__u8		ifa_flags;	/* Flags			*/
+	__u8		ifa_scope;	/* Address scope		*/
+	__u32		ifa_index;	/* Link index			*/
+};
+
+/*
+ * Important comment:
+ * IFA_ADDRESS is prefix address, rather than local interface address.
+ * It makes no difference for normally configured broadcast interfaces,
+ * but for point-to-point IFA_ADDRESS is DESTINATION address,
+ * local address is supplied in IFA_LOCAL attribute.
+ *
+ * IFA_FLAGS is a u32 attribute that extends the u8 field ifa_flags.
+ * If present, the value from struct ifaddrmsg will be ignored.
+ */
+enum {
+	IFA_UNSPEC,
+	IFA_ADDRESS,
+	IFA_LOCAL,
+	IFA_LABEL,
+	IFA_BROADCAST,
+	IFA_ANYCAST,
+	IFA_CACHEINFO,
+	IFA_MULTICAST,
+	IFA_FLAGS,
+	IFA_RT_PRIORITY,  /* u32, priority/metric for prefix route */
+	__IFA_MAX,
+};
+
+#define IFA_MAX (__IFA_MAX - 1)
+
+/* ifa_flags */
+#define IFA_F_SECONDARY		0x01
+#define IFA_F_TEMPORARY		IFA_F_SECONDARY
+
+#define	IFA_F_NODAD		0x02
+#define IFA_F_OPTIMISTIC	0x04
+#define IFA_F_DADFAILED		0x08
+#define	IFA_F_HOMEADDRESS	0x10
+#define IFA_F_DEPRECATED	0x20
+#define IFA_F_TENTATIVE		0x40
+#define IFA_F_PERMANENT		0x80
+#define IFA_F_MANAGETEMPADDR	0x100
+#define IFA_F_NOPREFIXROUTE	0x200
+#define IFA_F_MCAUTOJOIN	0x400
+#define IFA_F_STABLE_PRIVACY	0x800
+
+struct ifa_cacheinfo {
+	__u32	ifa_prefered;
+	__u32	ifa_valid;
+	__u32	cstamp; /* created timestamp, hundredths of seconds */
+	__u32	tstamp; /* updated timestamp, hundredths of seconds */
+};
+
+/* backwards compatibility for userspace */
+#define IFA_RTA(r)  ((struct rtattr*)(((char*)(r)) + NLMSG_ALIGN(sizeof(struct ifaddrmsg))))
+#define IFA_PAYLOAD(n) NLMSG_PAYLOAD(n,sizeof(struct ifaddrmsg))
+
+#endif

include/subdir.am

@@ -1,4 +1,5 @@
 noinst_HEADERS += \
+	include/linux/if_addr.h \
 	include/linux/if_bridge.h \
 	include/linux/if_link.h \
 	include/linux/lwtunnel.h \

lib/if.h

@@ -341,6 +341,8 @@ DECLARE_QOBJ_TYPE(interface)
 DECLARE_HOOK(if_add, (struct interface * ifp), (ifp))
 DECLARE_KOOH(if_del, (struct interface * ifp), (ifp))
 
+#define METRIC_MAX (~0)
+
 /* Connected address structure. */
 struct connected {
 	/* Attached interface. */
@@ -388,6 +390,13 @@ struct connected {
 
 	/* Label for Linux 2.2.X and upper. */
 	char *label;
+
+	/*
+	 * Used for setting the connected route's cost. If the metric
+	 * here is set to METRIC_MAX the connected route falls back to
+	 * "struct interface"
+	 */
+	uint32_t metric;
 };
 
 /* Nbr Connected address structure. */

zebra/connected.c

@@ -209,6 +209,7 @@ void connected_up(struct interface *ifp, struct connected *ifc)
 		.ifindex = ifp->ifindex,
 		.vrf_id = ifp->vrf_id,
 	};
+	uint32_t metric;
 
 	if (!CHECK_FLAG(ifc->conf, ZEBRA_IFC_REAL))
 		return;
@@ -243,11 +244,13 @@ void connected_up(struct interface *ifp, struct connected *ifc)
 		break;
 	}
 
+	metric = (ifc->metric < (uint32_t)METRIC_MAX) ?
+				ifc->metric : ifp->metric;
 	rib_add(afi, SAFI_UNICAST, ifp->vrf_id, ZEBRA_ROUTE_CONNECT, 0, 0, &p,
-		NULL, &nh, RT_TABLE_MAIN, ifp->metric, 0, 0, 0);
+		NULL, &nh, RT_TABLE_MAIN, metric, 0, 0, 0);
 
 	rib_add(afi, SAFI_MULTICAST, ifp->vrf_id, ZEBRA_ROUTE_CONNECT, 0, 0, &p,
-		NULL, &nh, RT_TABLE_MAIN, ifp->metric, 0, 0, 0);
+		NULL, &nh, RT_TABLE_MAIN, metric, 0, 0, 0);
 
 	if (IS_ZEBRA_DEBUG_RIB_DETAILED) {
 		char buf[PREFIX_STRLEN];
@@ -276,7 +279,7 @@ void connected_up(struct interface *ifp, struct connected *ifc)
 /* Add connected IPv4 route to the interface. */
 void connected_add_ipv4(struct interface *ifp, int flags, struct in_addr *addr,
 			uint16_t prefixlen, struct in_addr *broad,
-			const char *label)
+			const char *label, uint32_t metric)
 {
 	struct prefix_ipv4 *p;
 	struct connected *ifc;
@@ -288,6 +291,7 @@ void connected_add_ipv4(struct interface *ifp, int flags, struct in_addr *addr,
 	ifc = connected_new();
 	ifc->ifp = ifp;
 	ifc->flags = flags;
+	ifc->metric = metric;
 	/* If we get a notification from the kernel,
 	 * we can safely assume the address is known to the kernel */
 	SET_FLAG(ifc->conf, ZEBRA_IFC_QUEUED);
@@ -500,7 +504,7 @@ void connected_delete_ipv4(struct interface *ifp, int flags,
 /* Add connected IPv6 route to the interface. */
 void connected_add_ipv6(struct interface *ifp, int flags, struct in6_addr *addr,
 			struct in6_addr *broad, uint16_t prefixlen,
-			const char *label)
+			const char *label, uint32_t metric)
 {
 	struct prefix_ipv6 *p;
 	struct connected *ifc;
@@ -512,6 +516,7 @@ void connected_add_ipv6(struct interface *ifp, int flags, struct in6_addr *addr,
 	ifc = connected_new();
 	ifc->ifp = ifp;
 	ifc->flags = flags;
+	ifc->metric = metric;
 	/* If we get a notification from the kernel,
 	 * we can safely assume the address is known to the kernel */
 	SET_FLAG(ifc->conf, ZEBRA_IFC_QUEUED);

zebra/connected.h

@@ -36,7 +36,8 @@ extern struct connected *connected_check_ptp(struct interface *ifp,
 
 extern void connected_add_ipv4(struct interface *ifp, int flags,
 			       struct in_addr *addr, uint16_t prefixlen,
-			       struct in_addr *broad, const char *label);
+			       struct in_addr *broad, const char *label,
+			       uint32_t metric);
 
 extern void connected_delete_ipv4(struct interface *ifp, int flags,
 				  struct in_addr *addr, uint16_t prefixlen,
@@ -49,7 +50,8 @@ extern void connected_down(struct interface *ifp, struct connected *ifc);
 
 extern void connected_add_ipv6(struct interface *ifp, int flags,
 			       struct in6_addr *address, struct in6_addr *broad,
-			       uint16_t prefixlen, const char *label);
+			       uint16_t prefixlen, const char *label,
+			       uint32_t metric);
 extern void connected_delete_ipv6(struct interface *ifp,
 				  struct in6_addr *address,
 				  struct in6_addr *broad, uint16_t prefixlen);

zebra/if_ioctl.c

@@ -236,7 +236,8 @@ static int if_getaddrs(void)
 			}
 
 			connected_add_ipv4(ifp, flags, &addr->sin_addr,
-					   prefixlen, dest_pnt, NULL);
+					   prefixlen, dest_pnt, NULL,
+					   METRIC_MAX);
 		}
 		if (ifap->ifa_addr->sa_family == AF_INET6) {
 			struct sockaddr_in6 *addr;
@@ -258,7 +259,7 @@ static int if_getaddrs(void)
 #endif
 
 			connected_add_ipv6(ifp, flags, &addr->sin6_addr, NULL,
-					   prefixlen, NULL);
+					   prefixlen, NULL, METRIC_MAX);
 		}
 	}
 

zebra/if_ioctl_solaris.c

@@ -302,10 +302,11 @@ static int if_get_addr(struct interface *ifp, struct sockaddr *addr,
 	/* Set address to the interface. */
 	if (af == AF_INET)
 		connected_add_ipv4(ifp, flags, &SIN(addr)->sin_addr, prefixlen,
-				   (struct in_addr *)dest_pnt, label);
+				   (struct in_addr *)dest_pnt, label,
+				   METRIC_MAX);
 	else if (af == AF_INET6)
 		connected_add_ipv6(ifp, flags, &SIN6(addr)->sin6_addr, NULL,
-				   prefixlen, label);
+				   prefixlen, label, METRIC_MAX);
 
 	return 0;
 }

zebra/if_netlink.c

@@ -924,6 +924,7 @@ int netlink_interface_addr(struct nlmsghdr *h, ns_id_t ns_id, int startup)
 	uint8_t flags = 0;
 	char *label = NULL;
 	struct zebra_ns *zns;
+	uint32_t metric = METRIC_MAX;
 
 	zns = zebra_ns_lookup(ns_id);
 	ifa = NLMSG_DATA(h);
@@ -1032,6 +1033,9 @@ int netlink_interface_addr(struct nlmsghdr *h, ns_id_t ns_id, int startup)
 	if (label && strcmp(ifp->name, label) == 0)
 		label = NULL;
 
+	if (tb[IFA_RT_PRIORITY])
+		metric = *(uint32_t *)RTA_DATA(tb[IFA_RT_PRIORITY]);
+
 	/* Register interface address to the interface. */
 	if (ifa->ifa_family == AF_INET) {
 		if (ifa->ifa_prefixlen > IPV4_MAX_BITLEN) {
@@ -1044,7 +1048,8 @@ int netlink_interface_addr(struct nlmsghdr *h, ns_id_t ns_id, int startup)
 		if (h->nlmsg_type == RTM_NEWADDR)
 			connected_add_ipv4(ifp, flags, (struct in_addr *)addr,
 					   ifa->ifa_prefixlen,
-					   (struct in_addr *)broad, label);
+					   (struct in_addr *)broad, label,
+					   metric);
 		else
 			connected_delete_ipv4(
 				ifp, flags, (struct in_addr *)addr,
@@ -1070,7 +1075,8 @@ int netlink_interface_addr(struct nlmsghdr *h, ns_id_t ns_id, int startup)
 				connected_add_ipv6(ifp, flags,
 						   (struct in6_addr *)addr,
 						   (struct in6_addr *)broad,
-						   ifa->ifa_prefixlen, label);
+						   ifa->ifa_prefixlen, label,
+						   metric);
 		} else
 			connected_delete_ipv6(ifp, (struct in6_addr *)addr,
 					      (struct in6_addr *)broad,

zebra/kernel_socket.c

@@ -906,7 +906,8 @@ int ifam_read(struct ifa_msghdr *ifam)
 			connected_add_ipv4(ifp, flags, &addr.sin.sin_addr,
 					   ip_masklen(mask.sin.sin_addr),
 					   &brd.sin.sin_addr,
-					   (isalias ? ifname : NULL));
+					   (isalias ? ifname : NULL),
+					   METRIC_MAX);
 		else
 			connected_delete_ipv4(ifp, flags, &addr.sin.sin_addr,
 					      ip_masklen(mask.sin.sin_addr),
@@ -923,7 +924,8 @@ int ifam_read(struct ifa_msghdr *ifam)
 			connected_add_ipv6(ifp, flags, &addr.sin6.sin6_addr,
 					   NULL,
 					   ip6_masklen(mask.sin6.sin6_addr),
-					   (isalias ? ifname : NULL));
+					   (isalias ? ifname : NULL),
+					   METRIC_MAX);
 		else
 			connected_delete_ipv6(ifp, &addr.sin6.sin6_addr, NULL,
 					      ip6_masklen(mask.sin6.sin6_addr));