1 files changed, 119 insertions, 170 deletions

diff --git a/sys/netinet/ip_input.c b/sys/netinet/ip_input.c
index 44500c46b0d8..82d7acdd0710 100644
--- a/sys/netinet/ip_input.c
+++ b/sys/netinet/ip_input.c
@@ -27,18 +27,16 @@
  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  * SUCH DAMAGE.
- *
- *	@(#)ip_input.c	8.2 (Berkeley) 1/4/94
  */
 
 #include <sys/cdefs.h>
-__FBSDID("$FreeBSD$");
-
 #include "opt_bootp.h"
+#include "opt_inet.h"
 #include "opt_ipstealth.h"
 #include "opt_ipsec.h"
 #include "opt_route.h"
 #include "opt_rss.h"
+#include "opt_sctp.h"
 
 #include <sys/param.h>
 #include <sys/systm.h>
@@ -61,6 +59,7 @@ __FBSDID("$FreeBSD$");
 #include <net/if_types.h>
 #include <net/if_var.h>
 #include <net/if_dl.h>
+#include <net/if_private.h>
 #include <net/pfil.h>
 #include <net/route.h>
 #include <net/route/nhop.h>
@@ -76,13 +75,17 @@ __FBSDID("$FreeBSD$");
 #include <netinet/in_fib.h>
 #include <netinet/in_pcb.h>
 #include <netinet/ip_var.h>
+#include <netinet/ip_encap.h>
 #include <netinet/ip_fw.h>
 #include <netinet/ip_icmp.h>
+#include <netinet/igmp_var.h>
 #include <netinet/ip_options.h>
 #include <machine/in_cksum.h>
 #include <netinet/ip_carp.h>
 #include <netinet/in_rss.h>
-#include <netinet/ip_mroute.h>
+#ifdef SCTP
+#include <netinet/sctp_var.h>
+#endif
 
 #include <netipsec/ipsec_support.h>
 
@@ -96,8 +99,7 @@ CTASSERT(sizeof(struct ip) == 20);
 
 /* IP reassembly functions are defined in ip_reass.c. */
 extern void ipreass_init(void);
-extern void ipreass_drain(void);
-extern void ipreass_slowtimo(void);
+extern void ipreass_vnet_init(void);
 #ifdef VIMAGE
 extern void ipreass_destroy(void);
 #endif
@@ -130,7 +132,9 @@ SYSCTL_BOOL(_net_inet_ip, OID_AUTO, source_address_validation,
     CTLFLAG_VNET | CTLFLAG_RW, &VNET_NAME(ip_sav), true,
     "Drop incoming packets with source address that is a local address");
 
-VNET_DEFINE(pfil_head_t, inet_pfil_head);	/* Packet filter hooks */
+/* Packet filter hooks */
+VNET_DEFINE(pfil_head_t, inet_pfil_head);
+VNET_DEFINE(pfil_head_t, inet_local_pfil_head);
 
 static struct netisr_handler ip_nh = {
 	.nh_name = "ip",
@@ -163,9 +167,11 @@ static struct netisr_handler ip_direct_nh = {
 };
 #endif
 
-extern	struct domain inetdomain;
-extern	struct protosw inetsw[];
-u_char	ip_protox[IPPROTO_MAX];
+ipproto_input_t		*ip_protox[IPPROTO_MAX] = {
+			    [0 ... IPPROTO_MAX - 1] = rip_input };
+ipproto_ctlinput_t	*ip_ctlprotox[IPPROTO_MAX] = {
+			    [0 ... IPPROTO_MAX - 1] = rip_ctlinput };
+
 VNET_DEFINE(struct in_ifaddrhead, in_ifaddrhead);  /* first inet address */
 VNET_DEFINE(struct in_ifaddrhashhead *, in_ifaddrhashtbl); /* inet addr hash table  */
 VNET_DEFINE(u_long, in_ifaddrhmask);		/* mask for hash table */
@@ -301,18 +307,16 @@ SYSCTL_PROC(_net_inet_ip, IPCTL_INTRDQDROPS, intr_direct_queue_drops,
  * IP initialization: fill in IP protocol switch table.
  * All protocols not implemented in kernel go to raw IP protocol handler.
  */
-void
-ip_init(void)
+static void
+ip_vnet_init(void *arg __unused)
 {
 	struct pfil_head_args args;
-	struct protosw *pr;
-	int i;
 
 	CK_STAILQ_INIT(&V_in_ifaddrhead);
 	V_in_ifaddrhashtbl = hashinit(INADDR_NHASH, M_IFADDR, &V_in_ifaddrhmask);
 
 	/* Initialize IP reassembly queue. */
-	ipreass_init();
+	ipreass_vnet_init();
 
 	/* Initialize packet filter hooks. */
 	args.pa_version = PFIL_VERSION;
@@ -321,6 +325,10 @@ ip_init(void)
 	args.pa_headname = PFIL_INET_NAME;
 	V_inet_pfil_head = pfil_head_register(&args);
 
+	args.pa_flags = PFIL_OUT;
+	args.pa_headname = PFIL_INET_LOCAL_NAME;
+	V_inet_local_pfil_head = pfil_head_register(&args);
+
 	if (hhook_head_register(HHOOK_TYPE_IPSEC_IN, AF_INET,
 	    &V_ipsec_hhh_in[HHOOK_IPSEC_INET],
 	    HHOOK_WAITOK | HHOOK_HEADISINVNET) != 0)
@@ -332,42 +340,45 @@ ip_init(void)
 		printf("%s: WARNING: unable to register output helper hook\n",
 		    __func__);
 
-	/* Skip initialization of globals for non-default instances. */
 #ifdef VIMAGE
-	if (!IS_DEFAULT_VNET(curvnet)) {
-		netisr_register_vnet(&ip_nh);
+	netisr_register_vnet(&ip_nh);
 #ifdef	RSS
-		netisr_register_vnet(&ip_direct_nh);
+	netisr_register_vnet(&ip_direct_nh);
 #endif
-		return;
-	}
 #endif
+}
+VNET_SYSINIT(ip_vnet_init, SI_SUB_PROTO_DOMAIN, SI_ORDER_FOURTH,
+    ip_vnet_init, NULL);
 
-	pr = pffindproto(PF_INET, IPPROTO_RAW, SOCK_RAW);
-	if (pr == NULL)
-		panic("ip_init: PF_INET not found");
+static void
+ip_init(const void *unused __unused)
+{
+
+	ipreass_init();
 
-	/* Initialize the entire ip_protox[] array to IPPROTO_RAW. */
-	for (i = 0; i < IPPROTO_MAX; i++)
-		ip_protox[i] = pr - inetsw;
 	/*
-	 * Cycle through IP protocols and put them into the appropriate place
-	 * in ip_protox[].
+	 * Register statically compiled protocols, that are unlikely to
+	 * ever become dynamic.
 	 */
-	for (pr = inetdomain.dom_protosw;
-	    pr < inetdomain.dom_protoswNPROTOSW; pr++)
-		if (pr->pr_domain->dom_family == PF_INET &&
-		    pr->pr_protocol && pr->pr_protocol != IPPROTO_RAW) {
-			/* Be careful to only index valid IP protocols. */
-			if (pr->pr_protocol < IPPROTO_MAX)
-				ip_protox[pr->pr_protocol] = pr - inetsw;
-		}
+	IPPROTO_REGISTER(IPPROTO_ICMP, icmp_input, NULL);
+	IPPROTO_REGISTER(IPPROTO_IGMP, igmp_input, NULL);
+	IPPROTO_REGISTER(IPPROTO_RSVP, rsvp_input, NULL);
+	IPPROTO_REGISTER(IPPROTO_IPV4, encap4_input, NULL);
+	IPPROTO_REGISTER(IPPROTO_MOBILE, encap4_input, NULL);
+	IPPROTO_REGISTER(IPPROTO_ETHERIP, encap4_input, NULL);
+	IPPROTO_REGISTER(IPPROTO_GRE, encap4_input, NULL);
+	IPPROTO_REGISTER(IPPROTO_IPV6, encap4_input, NULL);
+	IPPROTO_REGISTER(IPPROTO_PIM, encap4_input, NULL);
+#ifdef SCTP	/* XXX: has a loadable & static version */
+	IPPROTO_REGISTER(IPPROTO_SCTP, sctp_input, sctp_ctlinput);
+#endif
 
 	netisr_register(&ip_nh);
 #ifdef	RSS
 	netisr_register(&ip_direct_nh);
 #endif
 }
+SYSINIT(ip_init, SI_SUB_PROTO_DOMAIN, SI_ORDER_THIRD, ip_init, NULL);
 
 #ifdef VIMAGE
 static void
@@ -433,8 +444,7 @@ ip_direct_input(struct mbuf *m)
 	}
 #endif /* IPSEC */
 	IPSTAT_INC(ips_delivered);
-	(*inetsw[ip_protox[ip->ip_p]].pr_input)(&m, &hlen, ip->ip_p);
-	return;
+	ip_protox[ip->ip_p](&m, &hlen, ip->ip_p);
 }
 #endif
 
@@ -445,7 +455,6 @@ ip_direct_input(struct mbuf *m)
 void
 ip_input(struct mbuf *m)
 {
-	MROUTER_RLOCK_TRACKER;
 	struct ip *ip = NULL;
 	struct in_ifaddr *ia = NULL;
 	struct ifaddr *ifa;
@@ -512,6 +521,11 @@ ip_input(struct mbuf *m)
 			goto bad;
 		}
 	}
+	/* The unspecified address can appear only as a src address - RFC1122 */
+	if (__predict_false(ntohl(ip->ip_dst.s_addr) == INADDR_ANY)) {
+		IPSTAT_INC(ips_badaddr);
+		goto bad;
+	}
 
 	if (m->m_pkthdr.csum_flags & CSUM_IP_CHECKED) {
 		sum = !(m->m_pkthdr.csum_flags & CSUM_IP_VALID);
@@ -527,12 +541,6 @@ ip_input(struct mbuf *m)
 		goto bad;
 	}
 
-#ifdef ALTQ
-	if (altq_input != NULL && (*altq_input)(m, AF_INET) == 0)
-		/* packet is dropped by traffic conditioner */
-		return;
-#endif
-
 	ip_len = ntohs(ip->ip_len);
 	if (__predict_false(ip_len < hlen)) {
 		IPSTAT_INC(ips_badlen);
@@ -560,8 +568,9 @@ tooshort:
 
 	/*
 	 * Try to forward the packet, but if we fail continue.
-	 * ip_tryforward() does not generate redirects, so fall
-	 * through to normal processing if redirects are required.
+	 * ip_tryforward() may generate redirects these days.
+	 * XXX the logic below falling through to normal processing
+	 * if redirects are required should be revisited as well.
 	 * ip_tryforward() does inbound and outbound packet firewall
 	 * processing. If firewall has decided that destination becomes
 	 * our local address, it sets M_FASTFWD_OURS flag. In this
@@ -574,6 +583,10 @@ tooshort:
 	    IPSEC_CAPS(ipv4, m, IPSEC_CAP_OPERABLE) == 0)
 #endif
 	    ) {
+		/*
+		 * ip_dooptions() was run so we can ignore the source route (or
+		 * any IP options case) case for redirects in ip_tryforward().
+		 */
 		if ((m = ip_tryforward(m)) == NULL)
 			return;
 		if (m->m_flags & M_FASTFWD_OURS) {
@@ -605,11 +618,9 @@ tooshort:
 		goto passin;
 
 	odst = ip->ip_dst;
-	if (pfil_run_hooks(V_inet_pfil_head, &m, ifp, PFIL_IN, NULL) !=
+	if (pfil_mbuf_in(V_inet_pfil_head, &m, ifp, NULL) !=
 	    PFIL_PASS)
 		return;
-	if (m == NULL)			/* consumed by filter */
-		return;
 
 	ip = mtod(m, struct ip *);
 	dchg = (odst.s_addr != ip->ip_dst.s_addr);
@@ -736,7 +747,6 @@ passin:
 		ia = NULL;
 	}
 	if (IN_MULTICAST(ntohl(ip->ip_dst.s_addr))) {
-		MROUTER_RLOCK();
 		/*
 		 * RFC 3927 2.7: Do not forward multicast packets from
 		 * IN_LINKLOCAL.
@@ -751,7 +761,6 @@ passin:
 			 * must be discarded, else it may be accepted below.
 			 */
 			if (ip_mforward && ip_mforward(ip, ifp, m, 0) != 0) {
-				MROUTER_RUNLOCK();
 				IPSTAT_INC(ips_cantforward);
 				m_freem(m);
 				return;
@@ -763,12 +772,10 @@ passin:
 			 * host belongs to their destination groups.
 			 */
 			if (ip->ip_p == IPPROTO_IGMP) {
-				MROUTER_RUNLOCK();
 				goto ours;
 			}
 			IPSTAT_INC(ips_forward);
 		}
-		MROUTER_RUNLOCK();
 		/*
 		 * Assume the packet is for us, to avoid prematurely taking
 		 * a lock on the in_multi hash. Protocols must perform
@@ -810,6 +817,18 @@ ours:
 #endif /* IPSTEALTH */
 
 	/*
+	 * We are going to ship the packet to the local protocol stack. Call the
+	 * filter again for this 'output' action, allowing redirect-like rules
+	 * to adjust the source address.
+	 */
+	if (PFIL_HOOKED_OUT(V_inet_local_pfil_head)) {
+		if (pfil_mbuf_out(V_inet_local_pfil_head, &m, V_loif, NULL) !=
+		    PFIL_PASS)
+			return;
+		ip = mtod(m, struct ip *);
+	}
+
+	/*
 	 * Attempt reassembly; if it succeeds, proceed.
 	 * ip_reass() will return a different mbuf.
 	 */
@@ -835,110 +854,44 @@ ours:
 	 */
 	IPSTAT_INC(ips_delivered);
 
-	(*inetsw[ip_protox[ip->ip_p]].pr_input)(&m, &hlen, ip->ip_p);
+	ip_protox[ip->ip_p](&m, &hlen, ip->ip_p);
 	return;
 bad:
 	m_freem(m);
 }
 
-/*
- * IP timer processing;
- * if a timer expires on a reassembly
- * queue, discard it.
- */
-void
-ip_slowtimo(void)
-{
-	VNET_ITERATOR_DECL(vnet_iter);
-
-	VNET_LIST_RLOCK_NOSLEEP();
-	VNET_FOREACH(vnet_iter) {
-		CURVNET_SET(vnet_iter);
-		ipreass_slowtimo();
-		CURVNET_RESTORE();
-	}
-	VNET_LIST_RUNLOCK_NOSLEEP();
-}
-
-void
-ip_drain(void)
-{
-	VNET_ITERATOR_DECL(vnet_iter);
-
-	VNET_LIST_RLOCK_NOSLEEP();
-	VNET_FOREACH(vnet_iter) {
-		CURVNET_SET(vnet_iter);
-		ipreass_drain();
-		CURVNET_RESTORE();
-	}
-	VNET_LIST_RUNLOCK_NOSLEEP();
-}
-
-/*
- * The protocol to be inserted into ip_protox[] must be already registered
- * in inetsw[], either statically or through pf_proto_register().
- */
 int
-ipproto_register(short ipproto)
+ipproto_register(uint8_t proto, ipproto_input_t input, ipproto_ctlinput_t ctl)
 {
-	struct protosw *pr;
 
-	/* Sanity checks. */
-	if (ipproto <= 0 || ipproto >= IPPROTO_MAX)
-		return (EPROTONOSUPPORT);
+	MPASS(proto > 0);
 
 	/*
 	 * The protocol slot must not be occupied by another protocol
-	 * already.  An index pointing to IPPROTO_RAW is unused.
+	 * already.  An index pointing to rip_input() is unused.
 	 */
-	pr = pffindproto(PF_INET, IPPROTO_RAW, SOCK_RAW);
-	if (pr == NULL)
-		return (EPFNOSUPPORT);
-	if (ip_protox[ipproto] != pr - inetsw)	/* IPPROTO_RAW */
+	if (ip_protox[proto] == rip_input) {
+		ip_protox[proto] = input;
+		ip_ctlprotox[proto] = ctl;
+		return (0);
+	} else
 		return (EEXIST);
-
-	/* Find the protocol position in inetsw[] and set the index. */
-	for (pr = inetdomain.dom_protosw;
-	     pr < inetdomain.dom_protoswNPROTOSW; pr++) {
-		if (pr->pr_domain->dom_family == PF_INET &&
-		    pr->pr_protocol && pr->pr_protocol == ipproto) {
-			ip_protox[pr->pr_protocol] = pr - inetsw;
-			return (0);
-		}
-	}
-	return (EPROTONOSUPPORT);
 }
 
 int
-ipproto_unregister(short ipproto)
+ipproto_unregister(uint8_t proto)
 {
-	struct protosw *pr;
 
-	/* Sanity checks. */
-	if (ipproto <= 0 || ipproto >= IPPROTO_MAX)
-		return (EPROTONOSUPPORT);
+	MPASS(proto > 0);
 
-	/* Check if the protocol was indeed registered. */
-	pr = pffindproto(PF_INET, IPPROTO_RAW, SOCK_RAW);
-	if (pr == NULL)
-		return (EPFNOSUPPORT);
-	if (ip_protox[ipproto] == pr - inetsw)  /* IPPROTO_RAW */
+	if (ip_protox[proto] != rip_input) {
+		ip_protox[proto] = rip_input;
+		ip_ctlprotox[proto] = rip_ctlinput;
+		return (0);
+	} else
 		return (ENOENT);
-
-	/* Reset the protocol slot to IPPROTO_RAW. */
-	ip_protox[ipproto] = pr - inetsw;
-	return (0);
 }
 
-u_char inetctlerrmap[PRC_NCMDS] = {
-	0,		0,		0,		0,
-	0,		EMSGSIZE,	EHOSTDOWN,	EHOSTUNREACH,
-	EHOSTUNREACH,	EHOSTUNREACH,	ECONNREFUSED,	ECONNREFUSED,
-	EMSGSIZE,	EHOSTUNREACH,	0,		0,
-	0,		0,		EHOSTUNREACH,	0,
-	ENOPROTOOPT,	ECONNREFUSED
-};
-
 /*
  * Forward a packet.  If some error occurs return the sender
  * an icmp packet.  Note we can't always generate a meaningful
@@ -1165,8 +1118,8 @@ ip_savecontrol(struct inpcb *inp, struct mbuf **mp, struct ip *ip,
 		} else {
 			bintime(&bt);
 		}
-		*mp = sbcreatecontrol((caddr_t)&bt, sizeof(bt),
-		    SCM_BINTIME, SOL_SOCKET);
+		*mp = sbcreatecontrol(&bt, sizeof(bt), SCM_BINTIME,
+		    SOL_SOCKET, M_NOWAIT);
 		if (*mp != NULL) {
 			mp = &(*mp)->m_next;
 			stamped = true;
@@ -1187,8 +1140,8 @@ ip_savecontrol(struct inpcb *inp, struct mbuf **mp, struct ip *ip,
 		} else {
 			microtime(&tv);
 		}
-		*mp = sbcreatecontrol((caddr_t)&tv, sizeof(tv),
-		    SCM_TIMESTAMP, SOL_SOCKET);
+		*mp = sbcreatecontrol((caddr_t)&tv, sizeof(tv), SCM_TIMESTAMP,
+		    SOL_SOCKET, M_NOWAIT);
 		if (*mp != NULL) {
 			mp = &(*mp)->m_next;
 			stamped = true;
@@ -1206,8 +1159,8 @@ ip_savecontrol(struct inpcb *inp, struct mbuf **mp, struct ip *ip,
 		} else {
 			nanotime(&ts);
 		}
-		*mp = sbcreatecontrol((caddr_t)&ts, sizeof(ts),
-		    SCM_REALTIME, SOL_SOCKET);
+		*mp = sbcreatecontrol(&ts, sizeof(ts), SCM_REALTIME,
+		    SOL_SOCKET, M_NOWAIT);
 		if (*mp != NULL) {
 			mp = &(*mp)->m_next;
 			stamped = true;
@@ -1220,8 +1173,8 @@ ip_savecontrol(struct inpcb *inp, struct mbuf **mp, struct ip *ip,
 			mbuf_tstmp2timespec(m, &ts);
 		else
 			nanouptime(&ts);
-		*mp = sbcreatecontrol((caddr_t)&ts, sizeof(ts),
-		    SCM_MONOTONIC, SOL_SOCKET);
+		*mp = sbcreatecontrol(&ts, sizeof(ts), SCM_MONOTONIC,
+		    SOL_SOCKET, M_NOWAIT);
 		if (*mp != NULL) {
 			mp = &(*mp)->m_next;
 			stamped = true;
@@ -1235,20 +1188,20 @@ ip_savecontrol(struct inpcb *inp, struct mbuf **mp, struct ip *ip,
 		sti.st_info_flags = ST_INFO_HW;
 		if ((m->m_flags & M_TSTMP_HPREC) != 0)
 			sti.st_info_flags |= ST_INFO_HW_HPREC;
-		*mp = sbcreatecontrol((caddr_t)&sti, sizeof(sti), SCM_TIME_INFO,
-		    SOL_SOCKET);
+		*mp = sbcreatecontrol(&sti, sizeof(sti), SCM_TIME_INFO,
+		    SOL_SOCKET, M_NOWAIT);
 		if (*mp != NULL)
 			mp = &(*mp)->m_next;
 	}
 	if (inp->inp_flags & INP_RECVDSTADDR) {
-		*mp = sbcreatecontrol((caddr_t)&ip->ip_dst,
-		    sizeof(struct in_addr), IP_RECVDSTADDR, IPPROTO_IP);
+		*mp = sbcreatecontrol(&ip->ip_dst, sizeof(struct in_addr),
+		    IP_RECVDSTADDR, IPPROTO_IP, M_NOWAIT);
 		if (*mp)
 			mp = &(*mp)->m_next;
 	}
 	if (inp->inp_flags & INP_RECVTTL) {
-		*mp = sbcreatecontrol((caddr_t)&ip->ip_ttl,
-		    sizeof(u_char), IP_RECVTTL, IPPROTO_IP);
+		*mp = sbcreatecontrol(&ip->ip_ttl, sizeof(u_char), IP_RECVTTL,
+		    IPPROTO_IP, M_NOWAIT);
 		if (*mp)
 			mp = &(*mp)->m_next;
 	}
@@ -1259,15 +1212,15 @@ ip_savecontrol(struct inpcb *inp, struct mbuf **mp, struct ip *ip,
 	 */
 	/* options were tossed already */
 	if (inp->inp_flags & INP_RECVOPTS) {
-		*mp = sbcreatecontrol((caddr_t)opts_deleted_above,
-		    sizeof(struct in_addr), IP_RECVOPTS, IPPROTO_IP);
+		*mp = sbcreatecontrol(opts_deleted_above,
+		    sizeof(struct in_addr), IP_RECVOPTS, IPPROTO_IP, M_NOWAIT);
 		if (*mp)
 			mp = &(*mp)->m_next;
 	}
 	/* ip_srcroute doesn't do what we want here, need to fix */
 	if (inp->inp_flags & INP_RECVRETOPTS) {
-		*mp = sbcreatecontrol((caddr_t)ip_srcroute(m),
-		    sizeof(struct in_addr), IP_RECVRETOPTS, IPPROTO_IP);
+		*mp = sbcreatecontrol(ip_srcroute(m), sizeof(struct in_addr),
+		    IP_RECVRETOPTS, IPPROTO_IP, M_NOWAIT);
 		if (*mp)
 			mp = &(*mp)->m_next;
 	}
@@ -1299,14 +1252,14 @@ makedummy:
 			sdl2->sdl_index = 0;
 			sdl2->sdl_nlen = sdl2->sdl_alen = sdl2->sdl_slen = 0;
 		}
-		*mp = sbcreatecontrol((caddr_t)sdl2, sdl2->sdl_len,
-		    IP_RECVIF, IPPROTO_IP);
+		*mp = sbcreatecontrol(sdl2, sdl2->sdl_len, IP_RECVIF,
+		    IPPROTO_IP, M_NOWAIT);
 		if (*mp)
 			mp = &(*mp)->m_next;
 	}
 	if (inp->inp_flags & INP_RECVTOS) {
-		*mp = sbcreatecontrol((caddr_t)&ip->ip_tos,
-		    sizeof(u_char), IP_RECVTOS, IPPROTO_IP);
+		*mp = sbcreatecontrol(&ip->ip_tos, sizeof(u_char), IP_RECVTOS,
+		    IPPROTO_IP, M_NOWAIT);
 		if (*mp)
 			mp = &(*mp)->m_next;
 	}
@@ -1321,12 +1274,12 @@ makedummy:
 		 * XXX should handle the failure of one or the
 		 * other - don't populate both?
 		 */
-		*mp = sbcreatecontrol((caddr_t) &flowid,
-		    sizeof(uint32_t), IP_FLOWID, IPPROTO_IP);
+		*mp = sbcreatecontrol(&flowid, sizeof(uint32_t), IP_FLOWID,
+		    IPPROTO_IP, M_NOWAIT);
 		if (*mp)
 			mp = &(*mp)->m_next;
-		*mp = sbcreatecontrol((caddr_t) &flow_type,
-		    sizeof(uint32_t), IP_FLOWTYPE, IPPROTO_IP);
+		*mp = sbcreatecontrol(&flow_type, sizeof(uint32_t),
+		    IP_FLOWTYPE, IPPROTO_IP, M_NOWAIT);
 		if (*mp)
 			mp = &(*mp)->m_next;
 	}
@@ -1340,8 +1293,8 @@ makedummy:
 		flow_type = M_HASHTYPE_GET(m);
 
 		if (rss_hash2bucket(flowid, flow_type, &rss_bucketid) == 0) {
-			*mp = sbcreatecontrol((caddr_t) &rss_bucketid,
-			   sizeof(uint32_t), IP_RSSBUCKETID, IPPROTO_IP);
+			*mp = sbcreatecontrol(&rss_bucketid, sizeof(uint32_t),
+			    IP_RSSBUCKETID, IPPROTO_IP, M_NOWAIT);
 			if (*mp)
 				mp = &(*mp)->m_next;
 		}
@@ -1364,10 +1317,6 @@ int
 ip_rsvp_init(struct socket *so)
 {
 
-	if (so->so_type != SOCK_RAW ||
-	    so->so_proto->pr_protocol != IPPROTO_RSVP)
-		return EOPNOTSUPP;
-
 	if (V_ip_rsvpd != NULL)
 		return EADDRINUSE;