A major overhaul of the CARP implementation. The ip_carp.c was started

from scratch, copying needed functionality from the old implemenation
on demand, with a thorough review of all code. The main change is that
interface layer has been removed from the CARP. Now redundant addresses
are configured exactly on the interfaces, they run on.

The CARP configuration itself is, as before, configured and read via
SIOCSVH/SIOCGVH ioctls. A new prefix created with SIOCAIFADDR or
SIOCAIFADDR_IN6 may now be configured to a particular virtual host id,
which makes the prefix redundant.

ifconfig(8) semantics has been changed too: now one doesn't need
to clone carpXX interface, he/she should directly configure a vhid
on a Ethernet interface.

To supply vhid data from the kernel to an application the getifaddrs(8)
function had been changed to pass ifam_data with each address. [1]

The new implementation definitely closes all PRs related to carp(4)
being an interface, and may close several others. It also allows
to run a single redundant IP per interface.

Big thanks to Bjoern Zeeb for his help with inet6 part of patch, for
idea on using ifam_data and for several rounds of reviewing!

PR:		kern/117000, kern/126945, kern/126714, kern/120130, kern/117448
Reviewed by:	bz
Submitted by:	bz [1]

6 files changed, 39 insertions, 9 deletions

diff --git a/sys/net/if.c b/sys/net/if.c
index 8c6c24c..4377343 100644
--- a/sys/net/if.c
+++ b/sys/net/if.c
@@ -130,17 +130,19 @@ void	(*lagg_linkstate_p)(struct ifnet *ifp, int state);
 /* These are external hooks for CARP. */
 void	(*carp_linkstate_p)(struct ifnet *ifp);
 #if defined(INET) || defined(INET6)
-struct ifnet *(*carp_forus_p)(struct ifnet *ifp, u_char *dhost);
+int	(*carp_forus_p)(struct ifnet *ifp, u_char *dhost);
 int	(*carp_output_p)(struct ifnet *ifp, struct mbuf *m,
-    struct sockaddr *sa, struct rtentry *rt);
+    struct sockaddr *sa);
+int	(*carp_ioctl_p)(struct ifreq *, u_long, struct thread *);   
+int	(*carp_attach_p)(struct ifaddr *, int);
+void	(*carp_detach_p)(struct ifaddr *);
 #endif
 #ifdef INET
-int (*carp_iamatch_p)(struct ifnet *, struct in_ifaddr *, struct in_addr *,
-    u_int8_t **);
+int	(*carp_iamatch_p)(struct ifaddr *, uint8_t **);
 #endif
 #ifdef INET6
 struct ifaddr *(*carp_iamatch6_p)(struct ifnet *ifp, struct in6_addr *taddr6);
-caddr_t (*carp_macmatch6_p)(struct ifnet *ifp, struct mbuf *m,
+caddr_t	(*carp_macmatch6_p)(struct ifnet *ifp, struct mbuf *m,
     const struct in6_addr *taddr);
 #endif
 
@@ -2506,6 +2508,16 @@ ifioctl(struct socket *so, u_long cmd, caddr_t data, struct thread *td)
 		error = if_getgroupmembers((struct ifgroupreq *)data);
 		CURVNET_RESTORE();
 		return (error);
+#if defined(INET) || defined(INET6)
+	case SIOCSVH:
+	case SIOCGVH:
+		if (carp_ioctl_p == NULL)
+			error = EPROTONOSUPPORT;
+		else
+			error = (*carp_ioctl_p)(ifr, cmd, td);
+		CURVNET_RESTORE();
+		return (error);
+#endif
 	}
 
 	ifp = ifunit_ref(ifr->ifr_name);
diff --git a/sys/net/if.h b/sys/net/if.h
index 4f2dc6f..08420f4 100644
--- a/sys/net/if.h
+++ b/sys/net/if.h
@@ -85,7 +85,7 @@ struct if_data {
 	u_char	ifi_addrlen;		/* media address length */
 	u_char	ifi_hdrlen;		/* media header length */
 	u_char	ifi_link_state;		/* current link state */
-	u_char	ifi_spare_char1;	/* spare byte */
+	u_char	ifi_vhid;		/* carp vhid */
 	u_char	ifi_spare_char2;	/* spare byte */
 	u_char	ifi_datalen;		/* length of this data struct */
 	u_long	ifi_mtu;		/* maximum transmission unit */
@@ -267,6 +267,8 @@ struct ifa_msghdr {
 	int	ifam_flags;	/* value of ifa_flags */
 	u_short	ifam_index;	/* index for associated ifp */
 	int	ifam_metric;	/* value of ifa_metric */
+	struct	if_data ifam_data;/* statistics and other data about if or
+				 * address */
 };
 
 /*
@@ -357,6 +359,7 @@ struct ifaliasreq {
 	struct	sockaddr ifra_addr;
 	struct	sockaddr ifra_broadaddr;
 	struct	sockaddr ifra_mask;
+	int	ifra_vhid;
 };
 
 struct ifmediareq {
diff --git a/sys/net/if_ethersubr.c b/sys/net/if_ethersubr.c
index 9eae683..253371a 100644
--- a/sys/net/if_ethersubr.c
+++ b/sys/net/if_ethersubr.c
@@ -397,7 +397,7 @@ ether_output(struct ifnet *ifp, struct mbuf *m,
 
 #if defined(INET) || defined(INET6)
 	if (ifp->if_carp &&
-	    (error = (*carp_output_p)(ifp, m, dst, NULL)))
+	    (error = (*carp_output_p)(ifp, m, dst)))
 		goto bad;
 #endif
 
diff --git a/sys/net/if_types.h b/sys/net/if_types.h
index c2effac..fe6ab5e 100644
--- a/sys/net/if_types.h
+++ b/sys/net/if_types.h
@@ -250,6 +250,5 @@
 #define	IFT_ENC		0xf4
 #define	IFT_PFLOG	0xf6
 #define	IFT_PFSYNC	0xf7
-#define	IFT_CARP	0xf8	/* Common Address Redundancy Protocol */
 #define IFT_IPXIP	0xf9	/* IPX over IP tunneling; no longer used. */
 #endif /* !_NET_IF_TYPES_H_ */
diff --git a/sys/net/if_var.h b/sys/net/if_var.h
index 1cb8440..ea60944 100644
--- a/sys/net/if_var.h
+++ b/sys/net/if_var.h
@@ -69,6 +69,7 @@ struct	rt_addrinfo;
 struct	socket;
 struct	ether_header;
 struct	carp_if;
+struct	carp_softc;
 struct  ifvlantrunk;
 struct	route;
 struct	vnet;
@@ -729,6 +730,7 @@ struct ifaddr {
 	struct	sockaddr *ifa_netmask;	/* used to determine subnet */
 	struct	if_data if_data;	/* not all members are meaningful */
 	struct	ifnet *ifa_ifp;		/* back-pointer to interface */
+	struct	carp_softc *ifa_carp;	/* pointer to CARP data */
 	TAILQ_ENTRY(ifaddr) ifa_link;	/* queue macro glue */
 	void	(*ifa_rtrequest)	/* check or clean routes (+ or -)'d */
 		(int, struct rtentry *, struct rt_addrinfo *);
diff --git a/sys/net/rtsock.c b/sys/net/rtsock.c
index b2313c0..3520824 100644
--- a/sys/net/rtsock.c
+++ b/sys/net/rtsock.c
@@ -63,6 +63,7 @@
 
 #include <netinet/in.h>
 #include <netinet/if_ether.h>
+#include <netinet/ip_carp.h>
 #ifdef INET6
 #include <netinet6/scope6_var.h>
 #endif
@@ -83,7 +84,7 @@ struct if_data32 {
 	uint8_t	ifi_addrlen;
 	uint8_t	ifi_hdrlen;
 	uint8_t	ifi_link_state;
-	uint8_t	ifi_spare_char1;
+	uint8_t	ifi_vhid;
 	uint8_t	ifi_spare_char2;
 	uint8_t	ifi_datalen;
 	uint32_t ifi_mtu;
@@ -122,6 +123,9 @@ MALLOC_DEFINE(M_RTABLE, "routetbl", "routing tables");
 static struct	sockaddr route_src = { 2, PF_ROUTE, };
 static struct	sockaddr sa_zero   = { sizeof(sa_zero), AF_INET, };
 
+/* These are external hooks for CARP. */
+int	(*carp_get_vhid_p)(struct ifaddr *);
+
 /*
  * Used by rtsock/raw_input callback code to decide whether to filter the update
  * notification to a socket bound to a particular FIB.
@@ -1508,6 +1512,7 @@ copy_ifdata32(struct if_data *src, struct if_data32 *dst)
 	CP(*src, *dst, ifi_addrlen);
 	CP(*src, *dst, ifi_hdrlen);
 	CP(*src, *dst, ifi_link_state);
+	CP(*src, *dst, ifi_vhid);
 	dst->ifi_datalen = sizeof(struct if_data32);
 	CP(*src, *dst, ifi_mtu);
 	CP(*src, *dst, ifi_metric);
@@ -1559,6 +1564,9 @@ sysctl_iflist(int af, struct walkarg *w)
 				ifm32->ifm_flags = ifp->if_flags |
 				    ifp->if_drv_flags;
 				copy_ifdata32(&ifp->if_data, &ifm32->ifm_data);
+				if (carp_get_vhid_p != NULL)
+					ifm32->ifm_data.ifi_vhid =
+					    (*carp_get_vhid_p)(ifa);
 				ifm32->ifm_addrs = info.rti_addrs;
 				error = SYSCTL_OUT(w->w_req, (caddr_t)ifm32,
 				    len);
@@ -1569,6 +1577,9 @@ sysctl_iflist(int af, struct walkarg *w)
 			ifm->ifm_index = ifp->if_index;
 			ifm->ifm_flags = ifp->if_flags | ifp->if_drv_flags;
 			ifm->ifm_data = ifp->if_data;
+			if (carp_get_vhid_p != NULL)
+				ifm->ifm_data.ifi_vhid =
+				    (*carp_get_vhid_p)(ifa);
 			ifm->ifm_addrs = info.rti_addrs;
 			error = SYSCTL_OUT(w->w_req, (caddr_t)ifm, len);
 #ifdef COMPAT_FREEBSD32
@@ -1595,6 +1606,9 @@ sysctl_iflist(int af, struct walkarg *w)
 				ifam->ifam_flags = ifa->ifa_flags;
 				ifam->ifam_metric = ifa->ifa_metric;
 				ifam->ifam_addrs = info.rti_addrs;
+				if (carp_get_vhid_p != NULL)
+					ifam->ifam_data.ifi_vhid =
+					    (*carp_get_vhid_p)(ifa);
 				error = SYSCTL_OUT(w->w_req, w->w_tmem, len);
 				if (error)
 					goto done;