]> git.saurik.com Git - apple/xnu.git/blobdiff - bsd/netinet6/ip6_input.c
xnu-7195.101.1.tar.gz
[apple/xnu.git] / bsd / netinet6 / ip6_input.c
index 64221cf429a7ad064207127c088beb1ab1ea3b94..e1b9091a596a0cbbdd77b7639b30a35e41ad551e 100644 (file)
@@ -1,5 +1,30 @@
-/*     $FreeBSD: src/sys/netinet6/ip6_input.c,v 1.11.2.10 2001/07/24 19:10:18 brooks Exp $     */
-/*     $KAME: ip6_input.c,v 1.194 2001/05/27 13:28:35 itojun Exp $     */
+/*
+ * Copyright (c) 2003-2020 Apple Inc. All rights reserved.
+ *
+ * @APPLE_OSREFERENCE_LICENSE_HEADER_START@
+ *
+ * This file contains Original Code and/or Modifications of Original Code
+ * as defined in and that are subject to the Apple Public Source License
+ * Version 2.0 (the 'License'). You may not use this file except in
+ * compliance with the License. The rights granted to you under the License
+ * may not be used to create, or enable the creation or redistribution of,
+ * unlawful or unlicensed copies of an Apple operating system, or to
+ * circumvent, violate, or enable the circumvention or violation of, any
+ * terms of an Apple operating system software license agreement.
+ *
+ * Please obtain a copy of the License at
+ * http://www.opensource.apple.com/apsl/ and read it before using this file.
+ *
+ * The Original Code and all software distributed under the License are
+ * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
+ * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
+ * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
+ * Please see the License for the specific language governing rights and
+ * limitations under the License.
+ *
+ * @APPLE_OSREFERENCE_LICENSE_HEADER_END@
+ */
 
 /*
  * Copyright (C) 1995, 1996, 1997, and 1998 WIDE Project.
@@ -65,7 +90,6 @@
  *     @(#)ip_input.c  8.2 (Berkeley) 1/4/94
  */
 
-
 #include <sys/param.h>
 #include <sys/systm.h>
 #include <sys/malloc.h>
 #include <sys/time.h>
 #include <sys/kernel.h>
 #include <sys/syslog.h>
+#include <sys/sysctl.h>
 #include <sys/proc.h>
 #include <sys/kauth.h>
+#include <sys/mcache.h>
+
+#include <mach/mach_time.h>
+#include <mach/sdt.h>
+#include <pexpert/pexpert.h>
+#include <dev/random/randomdev.h>
 
 #include <net/if.h>
 #include <net/if_var.h>
 #include <net/if_types.h>
 #include <net/if_dl.h>
 #include <net/route.h>
+#include <net/kpi_protocol.h>
+#include <net/ntstat.h>
+#include <net/init.h>
+#include <net/net_osdep.h>
+#include <net/net_perf.h>
 
 #include <netinet/in.h>
 #include <netinet/in_systm.h>
 #if INET
 #include <netinet/ip.h>
 #include <netinet/ip_icmp.h>
-#endif /*INET*/
+#endif /* INET */
+#include <netinet/kpi_ipfilter_var.h>
 #include <netinet/ip6.h>
+#include <netinet/udp.h>
 #include <netinet6/in6_var.h>
 #include <netinet6/ip6_var.h>
 #include <netinet/in_pcb.h>
 #include <netinet/icmp6.h>
 #include <netinet6/in6_ifattach.h>
 #include <netinet6/nd6.h>
-#include <netinet6/in6_prefix.h>
+#include <netinet6/scope6_var.h>
+#include <netinet6/ip6protosw.h>
 
 #if IPSEC
 #include <netinet6/ipsec.h>
-#if INET6
 #include <netinet6/ipsec6.h>
-#endif
 extern int ipsec_bypass;
-extern lck_mtx_t *sadb_mutex;
-#endif
-
-#include <netinet6/ip6_fw.h>
-
-#include <netinet/kpi_ipfilter_var.h>
+#endif /* IPSEC */
 
-#include <netinet6/ip6protosw.h>
+#if DUMMYNET
+#include <netinet/ip_dummynet.h>
+#endif /* DUMMYNET */
 
 /* we need it for NLOOP. */
 #include "loop.h"
-#include "faith.h"
 
-#include <net/net_osdep.h>
+#if PF
+#include <net/pfvar.h>
+#endif /* PF */
 
-extern struct domain inet6domain;
-extern struct ip6protosw inet6sw[];
+#include <os/log.h>
 
-struct ip6protosw *  ip6_protox[IPPROTO_MAX];
-static int ip6qmaxlen = IFQ_MAXLEN;
-struct in6_ifaddr *in6_ifaddrs;
+struct ip6protosw *ip6_protox[IPPROTO_MAX];
 
-int ip6_forward_srcrt;                 /* XXX */
-int ip6_sourcecheck;                   /* XXX */
-int ip6_sourcecheck_interval;          /* XXX */
-const int int6intrq_present = 1;
+static lck_grp_attr_t   *in6_ifaddr_rwlock_grp_attr;
+static lck_grp_t        *in6_ifaddr_rwlock_grp;
+static lck_attr_t       *in6_ifaddr_rwlock_attr;
+decl_lck_rw_data(, in6_ifaddr_rwlock);
 
-int ip6_ours_check_algorithm;
-int in6_init2done = 0;
+/* Protected by in6_ifaddr_rwlock */
+struct in6_ifaddrhead in6_ifaddrhead;
+struct in6_ifaddrhashhead * in6_ifaddrhashtbl;
+uint32_t in6_ifaddrhmask;
 
+#define IN6ADDR_NHASH    61
+u_int32_t in6addr_nhash = 0;                  /* hash table size */
+u_int32_t in6addr_hashp = 0;                  /* next largest prime */
 
-/* firewall hooks */
-ip6_fw_chk_t *ip6_fw_chk_ptr;
-ip6_fw_ctl_t *ip6_fw_ctl_ptr;
-int ip6_fw_enable = 1;
+
+#define IN6_IFSTAT_REQUIRE_ALIGNED_64(f)        \
+       _CASSERT(!(offsetof(struct in6_ifstat, f) % sizeof (uint64_t)))
+
+#define ICMP6_IFSTAT_REQUIRE_ALIGNED_64(f)      \
+       _CASSERT(!(offsetof(struct icmp6_ifstat, f) % sizeof (uint64_t)))
 
 struct ip6stat ip6stat;
 
-#ifdef __APPLE__
-struct ifqueue ip6intrq;
-lck_mtx_t              *ip6_mutex;
-lck_mtx_t              *dad6_mutex;
-lck_mtx_t              *nd6_mutex;
-lck_mtx_t                      *prefix6_mutex;
-lck_attr_t             *ip6_mutex_attr;
-lck_grp_t                      *ip6_mutex_grp;
-lck_grp_attr_t *ip6_mutex_grp_attr;
-extern lck_mtx_t       *inet6_domain_mutex;
+decl_lck_mtx_data(, proxy6_lock);
+decl_lck_mtx_data(static, dad6_mutex_data);
+decl_lck_mtx_data(static, nd6_mutex_data);
+decl_lck_mtx_data(static, prefix6_mutex_data);
+lck_mtx_t               *dad6_mutex = &dad6_mutex_data;
+lck_mtx_t               *nd6_mutex = &nd6_mutex_data;
+lck_mtx_t               *prefix6_mutex = &prefix6_mutex_data;
+#ifdef ENABLE_ADDRSEL
+decl_lck_mtx_data(static, addrsel_mutex_data);
+lck_mtx_t               *addrsel_mutex = &addrsel_mutex_data;
 #endif
-extern int loopattach_done;
+static lck_attr_t       *ip6_mutex_attr;
+static lck_grp_t        *ip6_mutex_grp;
+static lck_grp_attr_t   *ip6_mutex_grp_attr;
 
-static void ip6_init2(void *);
-static struct mbuf *ip6_setdstifaddr(struct mbuf *, struct in6_ifaddr *);
+extern int loopattach_done;
+extern void addrsel_policy_init(void);
 
+static int sysctl_reset_ip6_input_stats SYSCTL_HANDLER_ARGS;
+static int sysctl_ip6_input_measure_bins SYSCTL_HANDLER_ARGS;
+static int sysctl_ip6_input_getperf SYSCTL_HANDLER_ARGS;
+static void ip6_init_delayed(void);
 static int ip6_hopopts_input(u_int32_t *, u_int32_t *, struct mbuf **, int *);
-#if PULLDOWN_TEST
-static struct mbuf *ip6_pullexthdr(struct mbuf *, size_t, int);
-#endif
 
-#ifdef __APPLE__
-void gifattach(void);
-void faithattach(void);
-void stfattach(void);
-#endif
+static void in6_ifaddrhashtbl_init(void);
+
+#if NSTF
+extern void stfattach(void);
+#endif /* NSTF */
+
+SYSCTL_DECL(_net_inet6_ip6);
+
+static uint32_t ip6_adj_clear_hwcksum = 0;
+SYSCTL_UINT(_net_inet6_ip6, OID_AUTO, adj_clear_hwcksum,
+    CTLFLAG_RW | CTLFLAG_LOCKED, &ip6_adj_clear_hwcksum, 0,
+    "Invalidate hwcksum info when adjusting length");
+
+static uint32_t ip6_adj_partial_sum = 1;
+SYSCTL_UINT(_net_inet6_ip6, OID_AUTO, adj_partial_sum,
+    CTLFLAG_RW | CTLFLAG_LOCKED, &ip6_adj_partial_sum, 0,
+    "Perform partial sum adjustment of trailing bytes at IP layer");
+
+static int ip6_input_measure = 0;
+SYSCTL_PROC(_net_inet6_ip6, OID_AUTO, input_perf,
+    CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_LOCKED,
+    &ip6_input_measure, 0, sysctl_reset_ip6_input_stats, "I", "Do time measurement");
+
+static uint64_t ip6_input_measure_bins = 0;
+SYSCTL_PROC(_net_inet6_ip6, OID_AUTO, input_perf_bins,
+    CTLTYPE_QUAD | CTLFLAG_RW | CTLFLAG_LOCKED, &ip6_input_measure_bins, 0,
+    sysctl_ip6_input_measure_bins, "I",
+    "bins for chaining performance data histogram");
+
+static net_perf_t net_perf;
+SYSCTL_PROC(_net_inet6_ip6, OID_AUTO, input_perf_data,
+    CTLTYPE_STRUCT | CTLFLAG_RD | CTLFLAG_LOCKED,
+    0, 0, sysctl_ip6_input_getperf, "S,net_perf",
+    "IP6 input performance data (struct net_perf, net/net_perf.h)");
+
+/*
+ * ip6_checkinterface controls the receive side of the models for multihoming
+ * that are discussed in RFC 1122.
+ *
+ * sysctl_ip6_checkinterface values are:
+ *  IP6_CHECKINTERFACE_WEAK_ES:
+ *     This corresponds to the Weak End-System model where incoming packets from
+ *     any interface are accepted provided the destination address of the incoming packet
+ *     is assigned to some interface.
+ *
+ *  IP6_CHECKINTERFACE_HYBRID_ES:
+ *     The Hybrid End-System model use the Strong End-System for tunnel interfaces
+ *     (ipsec and utun) and the weak End-System model for other interfaces families.
+ *     This prevents a rogue middle box to probe for signs of TCP connections
+ *     that use the tunnel interface.
+ *
+ *  IP6_CHECKINTERFACE_STRONG_ES:
+ *     The Strong model model requires the packet arrived on an interface that
+ *     is assigned the destination address of the packet.
+ *
+ * Since the routing table and transmit implementation do not implement the Strong ES model,
+ * setting this to a value different from IP6_CHECKINTERFACE_WEAK_ES may lead to unexpected results.
+ *
+ * When forwarding is enabled, the system reverts to the Weak ES model as a router
+ * is expected by design to receive packets from several interfaces to the same address.
+ */
+#define IP6_CHECKINTERFACE_WEAK_ES       0
+#define IP6_CHECKINTERFACE_HYBRID_ES     1
+#define IP6_CHECKINTERFACE_STRONG_ES     2
+
+static int ip6_checkinterface = IP6_CHECKINTERFACE_HYBRID_ES;
+
+static int sysctl_ip6_checkinterface SYSCTL_HANDLER_ARGS;
+SYSCTL_PROC(_net_inet6_ip6, OID_AUTO, check_interface,
+    CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_LOCKED,
+    0, 0, sysctl_ip6_checkinterface, "I", "Verify packet arrives on correct interface");
+
+#if (DEBUG || DEVELOPMENT)
+#define IP6_CHECK_IFDEBUG 1
+#else
+#define IP6_CHECK_IFDEBUG 0
+#endif /* (DEBUG || DEVELOPMENT) */
+static int ip6_checkinterface_debug = IP6_CHECK_IFDEBUG;
+SYSCTL_INT(_net_inet6_ip6, OID_AUTO, checkinterface_debug, CTLFLAG_RW | CTLFLAG_LOCKED,
+    &ip6_checkinterface_debug, IP6_CHECK_IFDEBUG, "");
+
+typedef enum ip6_check_if_result {
+       IP6_CHECK_IF_NONE = 0,
+       IP6_CHECK_IF_OURS = 1,
+       IP6_CHECK_IF_DROP = 2,
+       IP6_CHECK_IF_FORWARD = 3
+} ip6_check_if_result_t;
+
+static ip6_check_if_result_t ip6_input_check_interface(struct mbuf *, struct ip6_hdr *, struct ifnet *, struct route_in6 *rin6, struct ifnet **);
+
+/*
+ * On platforms which require strict alignment (currently for anything but
+ * i386 or x86_64), check if the IP header pointer is 32-bit aligned; if not,
+ * copy the contents of the mbuf chain into a new chain, and free the original
+ * one.  Create some head room in the first mbuf of the new chain, in case
+ * it's needed later on.
+ *
+ * RFC 2460 says that IPv6 headers are 64-bit aligned, but network interfaces
+ * mostly align to 32-bit boundaries.  Care should be taken never to use 64-bit
+ * load/store operations on the fields in IPv6 headers.
+ */
+#if defined(__i386__) || defined(__x86_64__)
+#define IP6_HDR_ALIGNMENT_FIXUP(_m, _ifp, _action) do { } while (0)
+#else /* !__i386__ && !__x86_64__ */
+#define IP6_HDR_ALIGNMENT_FIXUP(_m, _ifp, _action) do {                 \
+       if (!IP6_HDR_ALIGNED_P(mtod(_m, caddr_t))) {                    \
+               struct mbuf *_n;                                        \
+               struct ifnet *__ifp = (_ifp);                           \
+               atomic_add_64(&(__ifp)->if_alignerrs, 1);               \
+               if (((_m)->m_flags & M_PKTHDR) &&                       \
+                   (_m)->m_pkthdr.pkt_hdr != NULL)                     \
+                       (_m)->m_pkthdr.pkt_hdr = NULL;                  \
+               _n = m_defrag_offset(_m, max_linkhdr, M_NOWAIT);        \
+               if (_n == NULL) {                                       \
+                       ip6stat.ip6s_toosmall++;                        \
+                       m_freem(_m);                                    \
+                       (_m) = NULL;                                    \
+                       _action;                                        \
+               } else {                                                \
+                       VERIFY(_n != (_m));                             \
+                       (_m) = _n;                                      \
+               }                                                       \
+       }                                                               \
+} while (0)
+#endif /* !__i386__ && !__x86_64__ */
 
 static void
-ip6_proto_input(
-       protocol_family_t       protocol,
-       mbuf_t                          packet)
+ip6_proto_input(protocol_family_t protocol, mbuf_t packet)
 {
+#pragma unused(protocol)
+#if INET
+       struct timeval start_tv;
+       if (ip6_input_measure) {
+               net_perf_start_time(&net_perf, &start_tv);
+       }
+#endif /* INET */
        ip6_input(packet);
+#if INET
+       if (ip6_input_measure) {
+               net_perf_measure_time(&net_perf, &start_tv, 1);
+               net_perf_histogram(&net_perf, 1);
+       }
+#endif /* INET */
 }
 
 /*
@@ -186,27 +359,63 @@ ip6_proto_input(
  * All protocols not implemented in kernel go to raw IP6 protocol handler.
  */
 void
-ip6_init()
+ip6_init(struct ip6protosw *pp, struct domain *dp)
 {
-       struct ip6protosw *pr;
-       int i;
+       static int ip6_initialized = 0;
+       struct protosw *pr;
        struct timeval tv;
-       extern lck_mtx_t *domain_proto_mtx;
+       int i;
+       domain_unguard_t unguard;
 
-#if DIAGNOSTIC
-       if (sizeof(struct protosw) != sizeof(struct ip6protosw))
-               panic("sizeof(protosw) != sizeof(ip6protosw)");
-#endif
-       pr = (struct ip6protosw *)pffindproto_locked(PF_INET6, IPPROTO_RAW, SOCK_RAW);
-       if (pr == 0)
-               panic("ip6_init");
-       for (i = 0; i < IPPROTO_MAX; i++)
-               ip6_protox[i] = pr;
-       for (pr = (struct ip6protosw*)inet6domain.dom_protosw; pr; pr = pr->pr_next) {
-               if(!((unsigned int)pr->pr_domain)) continue;    /* If uninitialized, skip */
-               if (pr->pr_domain->dom_family == PF_INET6 &&
-                   pr->pr_protocol && pr->pr_protocol != IPPROTO_RAW) {
-                       ip6_protox[pr->pr_protocol] = pr;
+       domain_proto_mtx_lock_assert_held();
+       VERIFY((pp->pr_flags & (PR_INITIALIZED | PR_ATTACHED)) == PR_ATTACHED);
+
+       _CASSERT((sizeof(struct ip6_hdr) +
+           sizeof(struct icmp6_hdr)) <= _MHLEN);
+
+       if (ip6_initialized) {
+               return;
+       }
+       ip6_initialized = 1;
+
+       eventhandler_lists_ctxt_init(&in6_evhdlr_ctxt);
+       (void)EVENTHANDLER_REGISTER(&in6_evhdlr_ctxt, in6_event,
+           in6_eventhdlr_callback, eventhandler_entry_dummy_arg,
+           EVENTHANDLER_PRI_ANY);
+
+       eventhandler_lists_ctxt_init(&in6_clat46_evhdlr_ctxt);
+       (void)EVENTHANDLER_REGISTER(&in6_clat46_evhdlr_ctxt, in6_clat46_event,
+           in6_clat46_eventhdlr_callback, eventhandler_entry_dummy_arg,
+           EVENTHANDLER_PRI_ANY);
+
+       for (i = 0; i < IN6_EVENT_MAX; i++) {
+               VERIFY(in6_event2kev_array[i].in6_event_code == i);
+       }
+
+       pr = pffindproto_locked(PF_INET6, IPPROTO_RAW, SOCK_RAW);
+       if (pr == NULL) {
+               panic("%s: Unable to find [PF_INET6,IPPROTO_RAW,SOCK_RAW]\n",
+                   __func__);
+               /* NOTREACHED */
+       }
+
+       /* Initialize the entire ip6_protox[] array to IPPROTO_RAW. */
+       for (i = 0; i < IPPROTO_MAX; i++) {
+               ip6_protox[i] = (struct ip6protosw *)pr;
+       }
+       /*
+        * Cycle through IP protocols and put them into the appropriate place
+        * in ip6_protox[], skipping protocols IPPROTO_{IP,RAW}.
+        */
+       VERIFY(dp == inet6domain && dp->dom_family == PF_INET6);
+       TAILQ_FOREACH(pr, &dp->dom_protosw, pr_entry) {
+               VERIFY(pr->pr_domain == dp);
+               if (pr->pr_protocol != 0 && pr->pr_protocol != IPPROTO_RAW) {
+                       /* Be careful to only index valid IP protocols. */
+                       if (pr->pr_protocol < IPPROTO_MAX) {
+                               ip6_protox[pr->pr_protocol] =
+                                   (struct ip6protosw *)pr;
+                       }
                }
        }
 
@@ -215,188 +424,498 @@ ip6_init()
        ip6_mutex_grp = lck_grp_alloc_init("ip6", ip6_mutex_grp_attr);
        ip6_mutex_attr = lck_attr_alloc_init();
 
-       if ((ip6_mutex = lck_mtx_alloc_init(ip6_mutex_grp, ip6_mutex_attr)) == NULL) {
-               printf("ip6_init: can't alloc ip6_mutex\n");
-               return;
-       }
-       if ((dad6_mutex = lck_mtx_alloc_init(ip6_mutex_grp, ip6_mutex_attr)) == NULL) {
-               printf("ip6_init: can't alloc dad6_mutex\n");
-               return;
-       }
-       if ((nd6_mutex = lck_mtx_alloc_init(ip6_mutex_grp, ip6_mutex_attr)) == NULL) {
-               printf("ip6_init: can't alloc nd6_mutex\n");
-               return;
-       }
+       lck_mtx_init(dad6_mutex, ip6_mutex_grp, ip6_mutex_attr);
+       lck_mtx_init(nd6_mutex, ip6_mutex_grp, ip6_mutex_attr);
+       lck_mtx_init(prefix6_mutex, ip6_mutex_grp, ip6_mutex_attr);
+       scope6_init(ip6_mutex_grp, ip6_mutex_attr);
 
-       if ((prefix6_mutex = lck_mtx_alloc_init(ip6_mutex_grp, ip6_mutex_attr)) == NULL) {
-               printf("ip6_init: can't alloc prefix6_mutex\n");
-               return;
-       }
+#ifdef ENABLE_ADDRSEL
+       lck_mtx_init(addrsel_mutex, ip6_mutex_grp, ip6_mutex_attr);
+#endif
 
-       inet6domain.dom_flags = DOM_REENTRANT;  
+       lck_mtx_init(&proxy6_lock, ip6_mutex_grp, ip6_mutex_attr);
+
+       in6_ifaddr_rwlock_grp_attr = lck_grp_attr_alloc_init();
+       in6_ifaddr_rwlock_grp = lck_grp_alloc_init("in6_ifaddr_rwlock",
+           in6_ifaddr_rwlock_grp_attr);
+       in6_ifaddr_rwlock_attr = lck_attr_alloc_init();
+       lck_rw_init(&in6_ifaddr_rwlock, in6_ifaddr_rwlock_grp,
+           in6_ifaddr_rwlock_attr);
+
+       TAILQ_INIT(&in6_ifaddrhead);
+       in6_ifaddrhashtbl_init();
+
+       IN6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_receive);
+       IN6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_hdrerr);
+       IN6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_toobig);
+       IN6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_noroute);
+       IN6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_addrerr);
+       IN6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_protounknown);
+       IN6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_truncated);
+       IN6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_discard);
+       IN6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_deliver);
+       IN6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_out_forward);
+       IN6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_out_request);
+       IN6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_out_discard);
+       IN6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_out_fragok);
+       IN6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_out_fragfail);
+       IN6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_out_fragcreat);
+       IN6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_reass_reqd);
+       IN6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_reass_ok);
+       IN6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_reass_fail);
+       IN6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_mcast);
+       IN6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_out_mcast);
+
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_msg);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_error);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_dstunreach);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_adminprohib);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_timeexceed);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_paramprob);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_pkttoobig);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_echo);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_echoreply);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_routersolicit);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_routeradvert);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_neighborsolicit);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_neighboradvert);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_redirect);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_mldquery);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_mldreport);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_in_mlddone);
+
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_out_msg);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_out_error);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_out_dstunreach);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_out_adminprohib);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_out_timeexceed);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_out_paramprob);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_out_pkttoobig);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_out_echo);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_out_echoreply);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_out_routersolicit);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_out_routeradvert);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_out_neighborsolicit);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_out_neighboradvert);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_out_redirect);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_out_mldquery);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_out_mldreport);
+       ICMP6_IFSTAT_REQUIRE_ALIGNED_64(ifs6_out_mlddone);
+
+       getmicrotime(&tv);
+       ip6_desync_factor =
+           (RandomULong() ^ tv.tv_usec) % MAX_TEMP_DESYNC_FACTOR;
+
+       PE_parse_boot_argn("ip6_checkinterface", &i, sizeof(i));
+       switch (i) {
+       case IP6_CHECKINTERFACE_WEAK_ES:
+       case IP6_CHECKINTERFACE_HYBRID_ES:
+       case IP6_CHECKINTERFACE_STRONG_ES:
+               ip6_checkinterface = i;
+               break;
+       default:
+               break;
+       }
 
-       ip6intrq.ifq_maxlen = ip6qmaxlen;
+       in6_ifaddr_init();
+       ip6_moptions_init();
        nd6_init();
        frag6_init();
-       icmp6_init();
+       icmp6_init(NULL, dp);
+       addrsel_policy_init();
+
        /*
-        * in many cases, random() here does NOT return random number
-        * as initialization during bootstrap time occur in fixed order.
+        * P2P interfaces often route the local address to the loopback
+        * interface. At this point, lo0 hasn't been initialized yet, which
+        * means that we need to delay the IPv6 configuration of lo0.
         */
-       microtime(&tv);
-       ip6_flow_seq = random() ^ tv.tv_usec;
-       microtime(&tv);
-       ip6_desync_factor = (random() ^ tv.tv_usec) % MAX_TEMP_DESYNC_FACTOR;
-       timeout(ip6_init2, (caddr_t)0, 1 * hz);
-
-       lck_mtx_unlock(domain_proto_mtx);       
-       proto_register_input(PF_INET6, ip6_proto_input, NULL);
-       lck_mtx_lock(domain_proto_mtx); 
+       net_init_add(ip6_init_delayed);
+
+       unguard = domain_unguard_deploy();
+       i = proto_register_input(PF_INET6, ip6_proto_input, NULL, 0);
+       if (i != 0) {
+               panic("%s: failed to register PF_INET6 protocol: %d\n",
+                   __func__, i);
+               /* NOTREACHED */
+       }
+       domain_unguard_release(unguard);
 }
 
 static void
-ip6_init2(dummy)
-       void *dummy;
+ip6_init_delayed(void)
 {
+       (void) in6_ifattach_prelim(lo_ifp);
+
+       /* timer for regeneranation of temporary addresses randomize ID */
+       timeout(in6_tmpaddrtimer, NULL,
+           (ip6_temp_preferred_lifetime - ip6_desync_factor -
+           ip6_temp_regen_advance) * hz);
+
+#if NSTF
+       stfattach();
+#endif /* NSTF */
+}
+
+static void
+ip6_input_adjust(struct mbuf *m, struct ip6_hdr *ip6, uint32_t plen,
+    struct ifnet *inifp)
+{
+       boolean_t adjust = TRUE;
+       uint32_t tot_len = sizeof(*ip6) + plen;
+
+       ASSERT(m_pktlen(m) > tot_len);
+
        /*
-        * to route local address of p2p link to loopback,
-        * assign loopback address first.
+        * Invalidate hardware checksum info if ip6_adj_clear_hwcksum
+        * is set; useful to handle buggy drivers.  Note that this
+        * should not be enabled by default, as we may get here due
+        * to link-layer padding.
         */
-       if (loopattach_done == 0) {
-               timeout(ip6_init2, (caddr_t)0, 1 * hz);
-               return;
+       if (ip6_adj_clear_hwcksum &&
+           (m->m_pkthdr.csum_flags & CSUM_DATA_VALID) &&
+           !(inifp->if_flags & IFF_LOOPBACK) &&
+           !(m->m_pkthdr.pkt_flags & PKTF_LOOP)) {
+               m->m_pkthdr.csum_flags &= ~CSUM_DATA_VALID;
+               m->m_pkthdr.csum_data = 0;
+               ip6stat.ip6s_adj_hwcsum_clr++;
+       }
+
+       /*
+        * If partial checksum information is available, subtract
+        * out the partial sum of postpended extraneous bytes, and
+        * update the checksum metadata accordingly.  By doing it
+        * here, the upper layer transport only needs to adjust any
+        * prepended extraneous bytes (else it will do both.)
+        */
+       if (ip6_adj_partial_sum &&
+           (m->m_pkthdr.csum_flags & (CSUM_DATA_VALID | CSUM_PARTIAL)) ==
+           (CSUM_DATA_VALID | CSUM_PARTIAL)) {
+               m->m_pkthdr.csum_rx_val = m_adj_sum16(m,
+                   m->m_pkthdr.csum_rx_start, m->m_pkthdr.csum_rx_start,
+                   (tot_len - m->m_pkthdr.csum_rx_start),
+                   m->m_pkthdr.csum_rx_val);
+       } else if ((m->m_pkthdr.csum_flags &
+           (CSUM_DATA_VALID | CSUM_PARTIAL)) ==
+           (CSUM_DATA_VALID | CSUM_PARTIAL)) {
+               /*
+                * If packet has partial checksum info and we decided not
+                * to subtract the partial sum of postpended extraneous
+                * bytes here (not the default case), leave that work to
+                * be handled by the other layers.  For now, only TCP, UDP
+                * layers are capable of dealing with this.  For all other
+                * protocols (including fragments), trim and ditch the
+                * partial sum as those layers might not implement partial
+                * checksumming (or adjustment) at all.
+                */
+               if (ip6->ip6_nxt == IPPROTO_TCP ||
+                   ip6->ip6_nxt == IPPROTO_UDP) {
+                       adjust = FALSE;
+               } else {
+                       m->m_pkthdr.csum_flags &= ~CSUM_DATA_VALID;
+                       m->m_pkthdr.csum_data = 0;
+                       ip6stat.ip6s_adj_hwcsum_clr++;
+               }
        }
-       in6_ifattach(&loif[0], NULL, NULL);
 
-#ifdef __APPLE__
-       /* nd6_timer_init */
-       timeout(nd6_timer, (caddr_t)0, hz);
+       if (adjust) {
+               ip6stat.ip6s_adj++;
+               if (m->m_len == m->m_pkthdr.len) {
+                       m->m_len = tot_len;
+                       m->m_pkthdr.len = tot_len;
+               } else {
+                       m_adj(m, tot_len - m->m_pkthdr.len);
+               }
+       }
+}
+static ip6_check_if_result_t
+ip6_input_check_interface(struct mbuf *m, struct ip6_hdr *ip6, struct ifnet *inifp, struct route_in6 *rin6, struct ifnet **deliverifp)
+{
+       struct in6_ifaddr *ia6 = NULL;
+       struct in6_addr tmp_dst = ip6->ip6_dst; /* copy to avoid unaligned access */
+       struct in6_ifaddr *best_ia6 = NULL;
+       ip6_check_if_result_t result = IP6_CHECK_IF_NONE;
 
-       /* router renumbering prefix list maintenance */
-       timeout(in6_rr_timer, (caddr_t)0, hz);
+       *deliverifp = NULL;
 
-       /* timer for regeneranation of temporary addresses randomize ID */
-       timeout(in6_tmpaddrtimer, (caddr_t)0,
-               (ip6_temp_preferred_lifetime - ip6_desync_factor -
-                      ip6_temp_regen_advance) * hz);
+       /*
+        * Check for exact addresses in the hash bucket.
+        */
+       lck_rw_lock_shared(&in6_ifaddr_rwlock);
+       TAILQ_FOREACH(ia6, IN6ADDR_HASH(&tmp_dst), ia6_hash) {
+               /*
+                * TODO: should we accept loopbacl
+                */
+               if (IN6_ARE_ADDR_EQUAL(&ia6->ia_addr.sin6_addr, &tmp_dst)) {
+                       if ((ia6->ia6_flags & (IN6_IFF_NOTREADY | IN6_IFF_CLAT46))) {
+                               continue;
+                       }
+                       best_ia6 = ia6;
+                       if (ia6->ia_ifp == inifp) {
+                               /*
+                                * TODO: should we also accept locally originated packets
+                                * or from loopback ???
+                                */
+                               break;
+                       }
+                       /*
+                        * Continue the loop in case there's a exact match with another
+                        * interface
+                        */
+               }
+       }
+       if (best_ia6 != NULL) {
+               if (best_ia6->ia_ifp != inifp && ip6_forwarding == 0 &&
+                   ((ip6_checkinterface == IP6_CHECKINTERFACE_HYBRID_ES &&
+                   (best_ia6->ia_ifp->if_family == IFNET_FAMILY_IPSEC ||
+                   best_ia6->ia_ifp->if_family == IFNET_FAMILY_UTUN)) ||
+                   ip6_checkinterface == IP6_CHECKINTERFACE_STRONG_ES)) {
+                       /*
+                        * Drop when interface address check is strict and forwarding
+                        * is disabled
+                        */
+                       result = IP6_CHECK_IF_DROP;
+               } else {
+                       result = IP6_CHECK_IF_OURS;
+                       *deliverifp = best_ia6->ia_ifp;
+                       ip6_setdstifaddr_info(m, 0, best_ia6);
+               }
+       }
+       lck_rw_done(&in6_ifaddr_rwlock);
 
-#if NGIF
-       gifattach();
-#endif
-#if NFAITH
-       faithattach();
-#endif
-#if NSTF
-       stfattach();
-#endif
-#else
-       /* nd6_timer_init */
+       if (result == IP6_CHECK_IF_NONE) {
+               /*
+                * Slow path: route lookup.
+                */
+               struct sockaddr_in6 *dst6;
 
-       callout_init(&nd6_timer_ch);
-       callout_reset(&nd6_timer_ch, hz, nd6_timer, NULL);
+               dst6 = SIN6(&rin6->ro_dst);
+               dst6->sin6_len = sizeof(struct sockaddr_in6);
+               dst6->sin6_family = AF_INET6;
+               dst6->sin6_addr = ip6->ip6_dst;
 
-       /* router renumbering prefix list maintenance */
-       callout_init(&in6_rr_timer_ch);
-       callout_reset(&in6_rr_timer_ch, hz, in6_rr_timer, NULL);
+               rtalloc_scoped_ign((struct route *)rin6,
+                   RTF_PRCLONING, IFSCOPE_NONE);
+               if (rin6->ro_rt != NULL) {
+                       RT_LOCK_SPIN(rin6->ro_rt);
+               }
 
-       /* timer for regeneranation of temporary addresses randomize ID */
-       callout_reset(&in6_tmpaddrtimer_ch,
-                     (ip6_temp_preferred_lifetime - ip6_desync_factor -
-                      ip6_temp_regen_advance) * hz,
-                     in6_tmpaddrtimer, NULL);
+#define rt6_key(r) (SIN6((r)->rt_nodes->rn_key))
+
+               /*
+                * Accept the packet if the forwarding interface to the destination
+                * according to the routing table is the loopback interface,
+                * unless the associated route has a gateway.
+                * Note that this approach causes to accept a packet if there is a
+                * route to the loopback interface for the destination of the packet.
+                * But we think it's even useful in some situations, e.g. when using
+                * a special daemon which wants to intercept the packet.
+                *
+                * XXX: some OSes automatically make a cloned route for the destination
+                * of an outgoing packet.  If the outgoing interface of the packet
+                * is a loopback one, the kernel would consider the packet to be
+                * accepted, even if we have no such address assinged on the interface.
+                * We check the cloned flag of the route entry to reject such cases,
+                * assuming that route entries for our own addresses are not made by
+                * cloning (it should be true because in6_addloop explicitly installs
+                * the host route).  However, we might have to do an explicit check
+                * while it would be less efficient.  Or, should we rather install a
+                * reject route for such a case?
+                */
+               if (rin6->ro_rt != NULL &&
+                   (rin6->ro_rt->rt_flags & (RTF_HOST | RTF_GATEWAY)) == RTF_HOST &&
+#if RTF_WASCLONED
+                   !(rin6->ro_rt->rt_flags & RTF_WASCLONED) &&
 #endif
+                   rin6->ro_rt->rt_ifp->if_type == IFT_LOOP) {
+                       ia6 = (struct in6_ifaddr *)rin6->ro_rt->rt_ifa;
+                       /*
+                        * Packets to a tentative, duplicated, or somehow invalid
+                        * address must not be accepted.
+                        *
+                        * For performance, test without acquiring the address lock;
+                        * a lot of things in the address are set once and never
+                        * changed (e.g. ia_ifp.)
+                        */
+                       if (!(ia6->ia6_flags & IN6_IFF_NOTREADY)) {
+                               /* this address is ready */
+                               result = IP6_CHECK_IF_OURS;
+                               *deliverifp = ia6->ia_ifp;       /* correct? */
+                               /*
+                                * record dst address information into mbuf.
+                                */
+                               (void) ip6_setdstifaddr_info(m, 0, ia6);
+                       }
+               }
 
-       in6_init2done = 1;
-}
+               if (rin6->ro_rt != NULL) {
+                       RT_UNLOCK(rin6->ro_rt);
+               }
+       }
 
-#if __FreeBSD__
-/* cheat */
-/* This must be after route_init(), which is now SI_ORDER_THIRD */
-SYSINIT(netinet6init2, SI_SUB_PROTO_DOMAIN, SI_ORDER_MIDDLE, ip6_init2, NULL);
-#endif
+       if (result == IP6_CHECK_IF_NONE) {
+               if (ip6_forwarding == 0) {
+                       result = IP6_CHECK_IF_DROP;
+               } else {
+                       result = IP6_CHECK_IF_FORWARD;
+                       ip6_setdstifaddr_info(m, inifp->if_index, NULL);
+               }
+       }
+
+       if (result == IP6_CHECK_IF_OURS && *deliverifp != inifp) {
+               ASSERT(*deliverifp != NULL);
+               ip6stat.ip6s_rcv_if_weak_match++;
+
+               /*  Logging is too noisy when forwarding is enabled */
+               if (ip6_checkinterface_debug != IP6_CHECKINTERFACE_WEAK_ES && ip6_forwarding != 0) {
+                       char src_str[MAX_IPv6_STR_LEN];
+                       char dst_str[MAX_IPv6_STR_LEN];
 
-extern struct  route_in6 ip6_forward_rt;
+                       inet_ntop(AF_INET6, &ip6->ip6_src, src_str, sizeof(src_str));
+                       inet_ntop(AF_INET6, &ip6->ip6_dst, dst_str, sizeof(dst_str));
+                       os_log_info(OS_LOG_DEFAULT,
+                           "%s: weak ES interface match to %s for packet from %s to %s proto %u received via %s",
+                           __func__, (*deliverifp)->if_xname, src_str, dst_str, ip6->ip6_nxt, inifp->if_xname);
+               }
+       } else if (result == IP6_CHECK_IF_DROP) {
+               ip6stat.ip6s_rcv_if_no_match++;
+               if (ip6_checkinterface_debug > 0) {
+                       char src_str[MAX_IPv6_STR_LEN];
+                       char dst_str[MAX_IPv6_STR_LEN];
+
+                       inet_ntop(AF_INET6, &ip6->ip6_src, src_str, sizeof(src_str));
+                       inet_ntop(AF_INET6, &ip6->ip6_dst, dst_str, sizeof(dst_str));
+                       os_log(OS_LOG_DEFAULT,
+                           "%s: no interface match for packet from %s to %s proto %u received via %s",
+                           __func__, src_str, dst_str, ip6->ip6_nxt, inifp->if_xname);
+               }
+       }
+
+       return result;
+}
 
 void
-ip6_input(m)
-       struct mbuf *m;
+ip6_input(struct mbuf *m)
 {
        struct ip6_hdr *ip6;
        int off = sizeof(struct ip6_hdr), nest;
        u_int32_t plen;
        u_int32_t rtalert = ~0;
        int nxt = 0, ours = 0;
-       struct ifnet *deliverifp = NULL;
-       ipfilter_t inject_ipfref = 0;
-       int seen;
+       struct ifnet *inifp, *deliverifp = NULL;
+       ipfilter_t inject_ipfref = NULL;
+       int seen = 1;
+#if DUMMYNET
+       struct m_tag *tag;
+       struct ip_fw_args args = {};
+#endif /* DUMMYNET */
+       struct route_in6 rin6 = {};
 
        /*
-        * No need to proccess packet twice if we've 
-        * already seen it
+        * Check if the packet we received is valid after interface filter
+        * processing
         */
-       inject_ipfref = ipf_get_inject_filter(m);
-       if (inject_ipfref != 0) {
-               ip6 = mtod(m, struct ip6_hdr *);
-               nxt = ip6->ip6_nxt;
-               seen = 0;
-               goto injectit;
-       } else
-               seen = 1;
-       
+       MBUF_INPUT_CHECK(m, m->m_pkthdr.rcvif);
+       inifp = m->m_pkthdr.rcvif;
+       VERIFY(inifp != NULL);
+
+       /* Perform IP header alignment fixup, if needed */
+       IP6_HDR_ALIGNMENT_FIXUP(m, inifp, return );
+
+       m->m_pkthdr.pkt_flags &= ~PKTF_FORWARDED;
 #if IPSEC
        /*
         * should the inner packet be considered authentic?
         * see comment in ah4_input().
         */
-       if (m) {
-               m->m_flags &= ~M_AUTHIPHDR;
-               m->m_flags &= ~M_AUTHIPDGM;
-       }
-#endif
+       m->m_flags &= ~M_AUTHIPHDR;
+       m->m_flags &= ~M_AUTHIPDGM;
+#endif /* IPSEC */
 
        /*
         * make sure we don't have onion peering information into m_aux.
         */
        ip6_delaux(m);
 
-       lck_mtx_lock(ip6_mutex);
+#if DUMMYNET
+       if ((tag = m_tag_locate(m, KERNEL_MODULE_TAG_ID,
+           KERNEL_TAG_TYPE_DUMMYNET, NULL)) != NULL) {
+               struct dn_pkt_tag       *dn_tag;
+
+               dn_tag = (struct dn_pkt_tag *)(tag + 1);
+
+               args.fwa_pf_rule = dn_tag->dn_pf_rule;
+
+               m_tag_delete(m, tag);
+       }
+
+       if (args.fwa_pf_rule) {
+               ip6 = mtod(m, struct ip6_hdr *); /* In case PF got disabled */
+
+               goto check_with_pf;
+       }
+#endif /* DUMMYNET */
+
+       /*
+        * No need to proccess packet twice if we've already seen it.
+        */
+       inject_ipfref = ipf_get_inject_filter(m);
+       if (inject_ipfref != NULL) {
+               ip6 = mtod(m, struct ip6_hdr *);
+               nxt = ip6->ip6_nxt;
+               seen = 0;
+               goto injectit;
+       } else {
+               seen = 1;
+       }
+
        /*
         * mbuf statistics
         */
        if (m->m_flags & M_EXT) {
-               if (m->m_next)
+               if (m->m_next != NULL) {
                        ip6stat.ip6s_mext2m++;
-               else
+               } else {
                        ip6stat.ip6s_mext1++;
+               }
        } else {
-#define M2MMAX (sizeof(ip6stat.ip6s_m2m)/sizeof(ip6stat.ip6s_m2m[0]))
-               if (m->m_next) {
-                       if (m->m_flags & M_LOOP) {
-                               ip6stat.ip6s_m2m[loif[0].if_index]++;   /* XXX */
-                       } else if (m->m_pkthdr.rcvif->if_index < M2MMAX)
-                               ip6stat.ip6s_m2m[m->m_pkthdr.rcvif->if_index]++;
-                       else
+#define M2MMAX  (sizeof (ip6stat.ip6s_m2m) / sizeof (ip6stat.ip6s_m2m[0]))
+               if (m->m_next != NULL) {
+                       if (m->m_pkthdr.pkt_flags & PKTF_LOOP) {
+                               /* XXX */
+                               ip6stat.ip6s_m2m[ifnet_index(lo_ifp)]++;
+                       } else if (inifp->if_index < M2MMAX) {
+                               ip6stat.ip6s_m2m[inifp->if_index]++;
+                       } else {
                                ip6stat.ip6s_m2m[0]++;
-               } else
+                       }
+               } else {
                        ip6stat.ip6s_m1++;
+               }
 #undef M2MMAX
        }
 
-       in6_ifstat_inc(m->m_pkthdr.rcvif, ifs6_in_receive);
+       /*
+        * Drop the packet if IPv6 operation is disabled on the interface.
+        */
+       if (inifp->if_eflags & IFEF_IPV6_DISABLED) {
+               goto bad;
+       }
+
+       in6_ifstat_inc_na(inifp, ifs6_in_receive);
        ip6stat.ip6s_total++;
 
-#ifndef PULLDOWN_TEST
        /*
         * L2 bridge code and some other code can return mbuf chain
         * that does not conform to KAME requirement.  too bad.
         * XXX: fails to join if interface MTU > MCLBYTES.  jumbogram?
         */
-       if (m && m->m_next != NULL && m->m_pkthdr.len < MCLBYTES) {
+       if (m->m_next != NULL && m->m_pkthdr.len < MCLBYTES) {
                struct mbuf *n;
 
-               MGETHDR(n, M_DONTWAIT, MT_HEADER);
-               if (n)
+               MGETHDR(n, M_DONTWAIT, MT_HEADER);      /* MAC-OK */
+               if (n) {
                        M_COPY_PKTHDR(n, m);
+               }
                if (n && m->m_pkthdr.len > MHLEN) {
                        MCLGET(n, M_DONTWAIT);
                        if ((n->m_flags & M_EXT) == 0) {
@@ -405,9 +924,7 @@ ip6_input(m)
                        }
                }
                if (n == NULL) {
-                       m_freem(m);
-                       lck_mtx_unlock(ip6_mutex);
-                       return; /*ENOBUFS*/
+                       goto bad;
                }
 
                m_copydata(m, 0, m->m_pkthdr.len, mtod(n, caddr_t));
@@ -415,18 +932,13 @@ ip6_input(m)
                m_freem(m);
                m = n;
        }
-       IP6_EXTHDR_CHECK(m, 0, sizeof(struct ip6_hdr),
-               {lck_mtx_unlock(ip6_mutex); return;}); 
-#endif
+       IP6_EXTHDR_CHECK(m, 0, sizeof(struct ip6_hdr), { goto done; });
 
        if (m->m_len < sizeof(struct ip6_hdr)) {
-               struct ifnet *inifp;
-               inifp = m->m_pkthdr.rcvif;
                if ((m = m_pullup(m, sizeof(struct ip6_hdr))) == 0) {
                        ip6stat.ip6s_toosmall++;
                        in6_ifstat_inc(inifp, ifs6_in_hdrerr);
-                       lck_mtx_unlock(ip6_mutex);
-                       return;
+                       goto done;
                }
        }
 
@@ -434,46 +946,40 @@ ip6_input(m)
 
        if ((ip6->ip6_vfc & IPV6_VERSION_MASK) != IPV6_VERSION) {
                ip6stat.ip6s_badvers++;
-               in6_ifstat_inc(m->m_pkthdr.rcvif, ifs6_in_hdrerr);
+               in6_ifstat_inc(inifp, ifs6_in_hdrerr);
                goto bad;
        }
 
        ip6stat.ip6s_nxthist[ip6->ip6_nxt]++;
 
-       /*
-        * Check with the firewall...
-        */
-       if (ip6_fw_enable && ip6_fw_chk_ptr) {
-               u_short port = 0;
-               /* If ipfw says divert, we have to just drop packet */
-               /* use port as a dummy argument */
-               if ((*ip6_fw_chk_ptr)(&ip6, NULL, &port, &m)) {
-                       m_freem(m);
-                       m = NULL;
-               }
-               if (!m) {
-                       lck_mtx_unlock(ip6_mutex);
-                       return;
-               }
-       }
-
        /*
         * Check against address spoofing/corruption.
         */
+       if (!(m->m_pkthdr.pkt_flags & PKTF_LOOP) &&
+           IN6_IS_ADDR_LOOPBACK(&ip6->ip6_src)) {
+               ip6stat.ip6s_badscope++;
+               in6_ifstat_inc(inifp, ifs6_in_addrerr);
+               goto bad;
+       }
        if (IN6_IS_ADDR_MULTICAST(&ip6->ip6_src) ||
            IN6_IS_ADDR_UNSPECIFIED(&ip6->ip6_dst)) {
                /*
                 * XXX: "badscope" is not very suitable for a multicast source.
                 */
                ip6stat.ip6s_badscope++;
-               in6_ifstat_inc(m->m_pkthdr.rcvif, ifs6_in_addrerr);
+               in6_ifstat_inc(inifp, ifs6_in_addrerr);
                goto bad;
        }
-       if ((IN6_IS_ADDR_LOOPBACK(&ip6->ip6_src) ||
-            IN6_IS_ADDR_LOOPBACK(&ip6->ip6_dst)) &&
-           (m->m_pkthdr.rcvif->if_flags & IFF_LOOPBACK) == 0) {
+       if (IN6_IS_ADDR_MC_INTFACELOCAL(&ip6->ip6_dst) &&
+           !(m->m_pkthdr.pkt_flags & PKTF_LOOP)) {
+               /*
+                * In this case, the packet should come from the loopback
+                * interface.  However, we cannot just check the if_flags,
+                * because ip6_mloopback() passes the "actual" interface
+                * as the outgoing/incoming interface.
+                */
                ip6stat.ip6s_badscope++;
-               in6_ifstat_inc(m->m_pkthdr.rcvif, ifs6_in_addrerr);
+               in6_ifstat_inc(inifp, ifs6_in_addrerr);
                goto bad;
        }
 
@@ -481,18 +987,18 @@ ip6_input(m)
         * The following check is not documented in specs.  A malicious
         * party may be able to use IPv4 mapped addr to confuse tcp/udp stack
         * and bypass security checks (act as if it was from 127.0.0.1 by using
-        * IPv6 src ::ffff:127.0.0.1).  Be cautious.
+        * IPv6 src ::ffff:127.0.0.1).  Be cautious.
         *
         * This check chokes if we are in an SIIT cloud.  As none of BSDs
         * support IPv4-less kernel compilation, we cannot support SIIT
         * environment at all.  So, it makes more sense for us to reject any
         * malicious packets for non-SIIT environment, than try to do a
-        * partical support for SIIT environment.
+        * partial support for SIIT environment.
         */
        if (IN6_IS_ADDR_V4MAPPED(&ip6->ip6_src) ||
            IN6_IS_ADDR_V4MAPPED(&ip6->ip6_dst)) {
                ip6stat.ip6s_badscope++;
-               in6_ifstat_inc(m->m_pkthdr.rcvif, ifs6_in_addrerr);
+               in6_ifstat_inc(inifp, ifs6_in_addrerr);
                goto bad;
        }
 #if 0
@@ -506,211 +1012,144 @@ ip6_input(m)
        if (IN6_IS_ADDR_V4COMPAT(&ip6->ip6_src) ||
            IN6_IS_ADDR_V4COMPAT(&ip6->ip6_dst)) {
                ip6stat.ip6s_badscope++;
-               in6_ifstat_inc(m->m_pkthdr.rcvif, ifs6_in_addrerr);
+               in6_ifstat_inc(inifp, ifs6_in_addrerr);
                goto bad;
        }
 #endif
 
+       /*
+        * Naively assume we can attribute inbound data to the route we would
+        * use to send to this destination. Asymetric routing breaks this
+        * assumption, but it still allows us to account for traffic from
+        * a remote node in the routing table.
+        * this has a very significant performance impact so we bypass
+        * if nstat_collect is disabled. We may also bypass if the
+        * protocol is tcp in the future because tcp will have a route that
+        * we can use to attribute the data to. That does mean we would not
+        * account for forwarded tcp traffic.
+        */
+       if (nstat_collect) {
+               struct rtentry *rte =
+                   ifnet_cached_rtlookup_inet6(inifp, &ip6->ip6_src);
+               if (rte != NULL) {
+                       nstat_route_rx(rte, 1, m->m_pkthdr.len, 0);
+                       rtfree(rte);
+               }
+       }
+
+#if DUMMYNET
+check_with_pf:
+#endif /* DUMMYNET */
+#if PF
+       /* Invoke inbound packet filter */
+       if (PF_IS_ENABLED) {
+               int error;
+#if DUMMYNET
+               error = pf_af_hook(inifp, NULL, &m, AF_INET6, TRUE, &args);
+#else /* !DUMMYNET */
+               error = pf_af_hook(inifp, NULL, &m, AF_INET6, TRUE, NULL);
+#endif /* !DUMMYNET */
+               if (error != 0 || m == NULL) {
+                       if (m != NULL) {
+                               panic("%s: unexpected packet %p\n",
+                                   __func__, m);
+                               /* NOTREACHED */
+                       }
+                       /* Already freed by callee */
+                       goto done;
+               }
+               ip6 = mtod(m, struct ip6_hdr *);
+       }
+#endif /* PF */
+
        /* drop packets if interface ID portion is already filled */
-       if ((m->m_pkthdr.rcvif->if_flags & IFF_LOOPBACK) == 0) {
+       if (!(inifp->if_flags & IFF_LOOPBACK) &&
+           !(m->m_pkthdr.pkt_flags & PKTF_LOOP)) {
                if (IN6_IS_SCOPE_LINKLOCAL(&ip6->ip6_src) &&
                    ip6->ip6_src.s6_addr16[1]) {
                        ip6stat.ip6s_badscope++;
                        goto bad;
                }
-               if (IN6_IS_SCOPE_LINKLOCAL(&ip6->ip6_dst) &&
+               if (IN6_IS_SCOPE_EMBED(&ip6->ip6_dst) &&
                    ip6->ip6_dst.s6_addr16[1]) {
                        ip6stat.ip6s_badscope++;
                        goto bad;
                }
        }
 
-       if (IN6_IS_SCOPE_LINKLOCAL(&ip6->ip6_src))
-               ip6->ip6_src.s6_addr16[1]
-                       = htons(m->m_pkthdr.rcvif->if_index);
-       if (IN6_IS_SCOPE_LINKLOCAL(&ip6->ip6_dst))
-               ip6->ip6_dst.s6_addr16[1]
-                       = htons(m->m_pkthdr.rcvif->if_index);
-
-#if 0 /* this case seems to be unnecessary. (jinmei, 20010401) */
-       /*
-        * We use rt->rt_ifp to determine if the address is ours or not.
-        * If rt_ifp is lo0, the address is ours.
-        * The problem here is, rt->rt_ifp for fe80::%lo0/64 is set to lo0,
-        * so any address under fe80::%lo0/64 will be mistakenly considered
-        * local.  The special case is supplied to handle the case properly
-        * by actually looking at interface addresses
-        * (using in6ifa_ifpwithaddr).
-        */
-       if ((m->m_pkthdr.rcvif->if_flags & IFF_LOOPBACK) != 0 &&
-           IN6_IS_ADDR_LINKLOCAL(&ip6->ip6_dst)) {
-           struct in6_ifaddr *ia6;
-               if (!(ia6 = in6ifa_ifpwithaddr(m->m_pkthdr.rcvif, &ip6->ip6_dst))) {
-                       lck_mtx_unlock(ip6_mutex);
-                       icmp6_error(m, ICMP6_DST_UNREACH,
-                           ICMP6_DST_UNREACH_ADDR, 0);
-                       /* m is already freed */
-                       return;
-               }
-               ifafree(&ia6->ia_ifa);
-
-               ours = 1;
-               deliverifp = m->m_pkthdr.rcvif;
-               goto hbhcheck;
+       if (m->m_pkthdr.pkt_flags & PKTF_IFAINFO) {
+               if (IN6_IS_SCOPE_LINKLOCAL(&ip6->ip6_src)) {
+                       ip6->ip6_src.s6_addr16[1] =
+                           htons(m->m_pkthdr.src_ifindex);
+               }
+               if (IN6_IS_SCOPE_EMBED(&ip6->ip6_dst)) {
+                       ip6->ip6_dst.s6_addr16[1] =
+                           htons(m->m_pkthdr.dst_ifindex);
+               }
+       } else {
+               if (IN6_IS_SCOPE_LINKLOCAL(&ip6->ip6_src)) {
+                       ip6->ip6_src.s6_addr16[1] = htons(inifp->if_index);
+               }
+               if (IN6_IS_SCOPE_EMBED(&ip6->ip6_dst)) {
+                       ip6->ip6_dst.s6_addr16[1] = htons(inifp->if_index);
+               }
        }
-#endif
 
        /*
         * Multicast check
         */
        if (IN6_IS_ADDR_MULTICAST(&ip6->ip6_dst)) {
-               struct  in6_multi *in6m = 0;
+               struct  in6_multi *in6m = NULL;
 
-               in6_ifstat_inc(m->m_pkthdr.rcvif, ifs6_in_mcast);
+               in6_ifstat_inc_na(inifp, ifs6_in_mcast);
                /*
                 * See if we belong to the destination multicast group on the
                 * arrival interface.
                 */
-               IN6_LOOKUP_MULTI(ip6->ip6_dst, m->m_pkthdr.rcvif, in6m);
-               if (in6m)
+               in6_multihead_lock_shared();
+               IN6_LOOKUP_MULTI(&ip6->ip6_dst, inifp, in6m);
+               in6_multihead_lock_done();
+               if (in6m != NULL) {
+                       IN6M_REMREF(in6m);
                        ours = 1;
-               else if (!ip6_mrouter) {
+               } else if (!nd6_prproxy) {
                        ip6stat.ip6s_notmember++;
                        ip6stat.ip6s_cantforward++;
-                       in6_ifstat_inc(m->m_pkthdr.rcvif, ifs6_in_discard);
+                       in6_ifstat_inc(inifp, ifs6_in_discard);
                        goto bad;
                }
-               deliverifp = m->m_pkthdr.rcvif;
-               goto hbhcheck;
-       }
-
-       /*
-        *  Unicast check
-        */
-       switch (ip6_ours_check_algorithm) {
-       default:
+               deliverifp = inifp;
                /*
-                * XXX: I intentionally broke our indentation rule here,
-                *      since this switch-case is just for measurement and
-                *      therefore should soon be removed.
+                * record dst address information into mbuf, if we don't have one yet.
+                * note that we are unable to record it, if the address is not listed
+                * as our interface address (e.g. multicast addresses, etc.)
                 */
-       if (ip6_forward_rt.ro_rt != NULL &&
-           (ip6_forward_rt.ro_rt->rt_flags & RTF_UP) != 0 && 
-           IN6_ARE_ADDR_EQUAL(&ip6->ip6_dst,
-                              &((struct sockaddr_in6 *)(&ip6_forward_rt.ro_dst))->sin6_addr))
-               ip6stat.ip6s_forward_cachehit++;
-       else {
-               struct sockaddr_in6 *dst6;
-
-               if (ip6_forward_rt.ro_rt) {
-                       /* route is down or destination is different */
-                       ip6stat.ip6s_forward_cachemiss++;
-                       rtfree(ip6_forward_rt.ro_rt);
-                       ip6_forward_rt.ro_rt = 0;
+               if (deliverifp != NULL) {
+                       struct in6_ifaddr *ia6 = NULL;
+
+                       ia6 = in6_ifawithifp(deliverifp, &ip6->ip6_dst);
+                       if (ia6 != NULL) {
+                               (void) ip6_setdstifaddr_info(m, 0, ia6);
+                               IFA_REMREF(&ia6->ia_ifa);
+                       } else {
+                               (void) ip6_setdstifaddr_info(m, inifp->if_index, NULL);
+                       }
                }
-
-               bzero(&ip6_forward_rt.ro_dst, sizeof(struct sockaddr_in6));
-               dst6 = (struct sockaddr_in6 *)&ip6_forward_rt.ro_dst;
-               dst6->sin6_len = sizeof(struct sockaddr_in6);
-               dst6->sin6_family = AF_INET6;
-               dst6->sin6_addr = ip6->ip6_dst;
-#if SCOPEDROUTING
-               ip6_forward_rt.ro_dst.sin6_scope_id =
-                       in6_addr2scopeid(m->m_pkthdr.rcvif, &ip6->ip6_dst);
-#endif
-
-               rtalloc_ign((struct route *)&ip6_forward_rt, RTF_PRCLONING);
-       }
-
-#define rt6_key(r) ((struct sockaddr_in6 *)((r)->rt_nodes->rn_key))
-
-       /*
-        * Accept the packet if the forwarding interface to the destination
-        * according to the routing table is the loopback interface,
-        * unless the associated route has a gateway.
-        * Note that this approach causes to accept a packet if there is a
-        * route to the loopback interface for the destination of the packet.
-        * But we think it's even useful in some situations, e.g. when using
-        * a special daemon which wants to intercept the packet.
-        *
-        * XXX: some OSes automatically make a cloned route for the destination
-        * of an outgoing packet.  If the outgoing interface of the packet
-        * is a loopback one, the kernel would consider the packet to be
-        * accepted, even if we have no such address assinged on the interface.
-        * We check the cloned flag of the route entry to reject such cases,
-        * assuming that route entries for our own addresses are not made by
-        * cloning (it should be true because in6_addloop explicitly installs
-        * the host route).  However, we might have to do an explicit check
-        * while it would be less efficient.  Or, should we rather install a
-        * reject route for such a case?
-        */
-       if (ip6_forward_rt.ro_rt &&
-           (ip6_forward_rt.ro_rt->rt_flags &
-            (RTF_HOST|RTF_GATEWAY)) == RTF_HOST &&
-#if RTF_WASCLONED
-           !(ip6_forward_rt.ro_rt->rt_flags & RTF_WASCLONED) &&
-#endif
-#if RTF_CLONED
-           !(ip6_forward_rt.ro_rt->rt_flags & RTF_CLONED) &&
-#endif
-#if 0
-           /*
-            * The check below is redundant since the comparison of
-            * the destination and the key of the rtentry has
-            * already done through looking up the routing table.
-            */
-           IN6_ARE_ADDR_EQUAL(&ip6->ip6_dst,
-                               &rt6_key(ip6_forward_rt.ro_rt)->sin6_addr)
-#endif
-           ip6_forward_rt.ro_rt->rt_ifp->if_type == IFT_LOOP) {
-               struct in6_ifaddr *ia6 =
-                       (struct in6_ifaddr *)ip6_forward_rt.ro_rt->rt_ifa;
-
-               /*
-                * record address information into m_aux.
-                */
-               (void)ip6_setdstifaddr(m, ia6);
-
+               goto hbhcheck;
+       } else {
                /*
-                * packets to a tentative, duplicated, or somehow invalid
-                * address must not be accepted.
+                * Unicast check
                 */
-               if (!(ia6->ia6_flags & IN6_IFF_NOTREADY)) {
-                       /* this address is ready */
+               ip6_check_if_result_t check_if_result = IP6_CHECK_IF_NONE;
+               check_if_result = ip6_input_check_interface(m, ip6, inifp, &rin6, &deliverifp);
+               ASSERT(check_if_result != IP6_CHECK_IF_NONE);
+               if (check_if_result == IP6_CHECK_IF_OURS) {
                        ours = 1;
-                       deliverifp = ia6->ia_ifp;       /* correct? */
-                       /* Count the packet in the ip address stats */
-#ifndef __APPLE__
-
-                       ia6->ia_ifa.if_ipackets++;
-                       ia6->ia_ifa.if_ibytes += m->m_pkthdr.len;
-#endif
                        goto hbhcheck;
-               } else {
-                       /* address is not ready, so discard the packet. */
-                       nd6log((LOG_INFO,
-                           "ip6_input: packet to an unready address %s->%s\n",
-                           ip6_sprintf(&ip6->ip6_src),
-                           ip6_sprintf(&ip6->ip6_dst)));
+               } else if (check_if_result == IP6_CHECK_IF_DROP) {
                        goto bad;
                }
        }
-       } /* XXX indentation (see above) */
-
-       /*
-        * FAITH(Firewall Aided Internet Translator)
-        */
-#if defined(NFAITH) && 0 < NFAITH
-       if (ip6_keepfaith) {
-               if (ip6_forward_rt.ro_rt && ip6_forward_rt.ro_rt->rt_ifp
-                && ip6_forward_rt.ro_rt->rt_ifp->if_type == IFT_FAITH) {
-                       /* XXX do we need more sanity checks? */
-                       ours = 1;
-                       deliverifp = ip6_forward_rt.ro_rt->rt_ifp; /* faith */
-                       goto hbhcheck;
-               }
-       }
-#endif
 
        /*
         * Now there is no reason to process the packet if it's not our own
@@ -718,32 +1157,20 @@ ip6_input(m)
         */
        if (!ip6_forwarding) {
                ip6stat.ip6s_cantforward++;
-               in6_ifstat_inc(m->m_pkthdr.rcvif, ifs6_in_discard);
-               goto bad;
-       }
-
-  hbhcheck:
-       /*
-        * record address information into m_aux, if we don't have one yet.
-        * note that we are unable to record it, if the address is not listed
-        * as our interface address (e.g. multicast addresses, addresses
-        * within FAITH prefixes and such).
-        */
-       if (deliverifp && !ip6_getdstifaddr(m)) {
-               struct in6_ifaddr *ia6;
-
-               ia6 = in6_ifawithifp(deliverifp, &ip6->ip6_dst);
-               if (ia6) {
-                       if (!ip6_setdstifaddr(m, ia6)) {
-                               /*
-                                * XXX maybe we should drop the packet here,
-                                * as we could not provide enough information
-                                * to the upper layers.
-                                */
-                       }
+               in6_ifstat_inc(inifp, ifs6_in_discard);
+               /*
+                * Raise a kernel event if the packet received on cellular
+                * interface is not intended for local host.
+                * For now limit it to ICMPv6 packets.
+                */
+               if (inifp->if_type == IFT_CELLULAR &&
+                   ip6->ip6_nxt == IPPROTO_ICMPV6) {
+                       in6_ifstat_inc(inifp, ifs6_cantfoward_icmp6);
                }
+               goto bad;
        }
 
+hbhcheck:
        /*
         * Process Hop-by-Hop options header if it's contained.
         * m may be modified in ip6_hopopts_input().
@@ -753,12 +1180,17 @@ ip6_input(m)
        if (ip6->ip6_nxt == IPPROTO_HOPOPTS) {
                struct ip6_hbh *hbh;
 
+               /*
+                * Mark the packet to imply that HBH option has been checked.
+                * This can only be true is the packet came in unfragmented
+                * or if the option is in the first fragment
+                */
+               m->m_pkthdr.pkt_flags |= PKTF_HBH_CHKED;
                if (ip6_hopopts_input(&plen, &rtalert, &m, &off)) {
-#if 0  /*touches NULL pointer*/
-                       in6_ifstat_inc(m->m_pkthdr.rcvif, ifs6_in_discard);
+#if 0   /* touches NULL pointer */
+                       in6_ifstat_inc(inifp, ifs6_in_discard);
 #endif
-                       lck_mtx_unlock(ip6_mutex);
-                       return; /* m have already been freed */
+                       goto done;      /* m have already been freed */
                }
 
                /* adjust pointer */
@@ -772,40 +1204,43 @@ ip6_input(m)
                if (ip6->ip6_plen == 0 && plen == 0) {
                        /*
                         * Note that if a valid jumbo payload option is
-                        * contained, ip6_hoptops_input() must set a valid
-                        * (non-zero) payload length to the variable plen. 
+                        * contained, ip6_hopopts_input() must set a valid
+                        * (non-zero) payload length to the variable plen.
                         */
                        ip6stat.ip6s_badoptions++;
-                       in6_ifstat_inc(m->m_pkthdr.rcvif, ifs6_in_discard);
-                       in6_ifstat_inc(m->m_pkthdr.rcvif, ifs6_in_hdrerr);
-                       lck_mtx_unlock(ip6_mutex);
-                       icmp6_error(m, ICMP6_PARAM_PROB,
-                                   ICMP6_PARAMPROB_HEADER,
-                                   (caddr_t)&ip6->ip6_plen - (caddr_t)ip6);
-                       return;
+                       in6_ifstat_inc(inifp, ifs6_in_discard);
+                       in6_ifstat_inc(inifp, ifs6_in_hdrerr);
+                       icmp6_error(m, ICMP6_PARAM_PROB, ICMP6_PARAMPROB_HEADER,
+                           (int)((caddr_t)&ip6->ip6_plen - (caddr_t)ip6));
+                       goto done;
                }
-#ifndef PULLDOWN_TEST
                /* ip6_hopopts_input() ensures that mbuf is contiguous */
                hbh = (struct ip6_hbh *)(ip6 + 1);
-#else
-               IP6_EXTHDR_GET(hbh, struct ip6_hbh *, m, sizeof(struct ip6_hdr),
-                       sizeof(struct ip6_hbh));
-               if (hbh == NULL) {
-                       ip6stat.ip6s_tooshort++;
-                       lck_mtx_unlock(ip6_mutex);
-                       return;
-               }
-#endif
                nxt = hbh->ip6h_nxt;
 
                /*
-                * accept the packet if a router alert option is included
-                * and we act as an IPv6 router.
+                * If we are acting as a router and the packet contains a
+                * router alert option, see if we know the option value.
+                * Currently, we only support the option value for MLD, in which
+                * case we should pass the packet to the multicast routing
+                * daemon.
                 */
-               if (rtalert != ~0 && ip6_forwarding)
-                       ours = 1;
-       } else
+               if (rtalert != ~0 && ip6_forwarding) {
+                       switch (rtalert) {
+                       case IP6OPT_RTALERT_MLD:
+                               ours = 1;
+                               break;
+                       default:
+                               /*
+                                * RFC2711 requires unrecognized values must be
+                                * silently ignored.
+                                */
+                               break;
+                       }
+               }
+       } else {
                nxt = ip6->ip6_nxt;
+       }
 
        /*
         * Check that the amount of data in the buffers
@@ -815,45 +1250,53 @@ ip6_input(m)
         */
        if (m->m_pkthdr.len - sizeof(struct ip6_hdr) < plen) {
                ip6stat.ip6s_tooshort++;
-               in6_ifstat_inc(m->m_pkthdr.rcvif, ifs6_in_truncated);
+               in6_ifstat_inc(inifp, ifs6_in_truncated);
                goto bad;
        }
        if (m->m_pkthdr.len > sizeof(struct ip6_hdr) + plen) {
-               if (m->m_len == m->m_pkthdr.len) {
-                       m->m_len = sizeof(struct ip6_hdr) + plen;
-                       m->m_pkthdr.len = sizeof(struct ip6_hdr) + plen;
-               } else
-                       m_adj(m, sizeof(struct ip6_hdr) + plen - m->m_pkthdr.len);
+               ip6_input_adjust(m, ip6, plen, inifp);
        }
 
        /*
         * Forward if desirable.
         */
        if (IN6_IS_ADDR_MULTICAST(&ip6->ip6_dst)) {
-               /*
-                * If we are acting as a multicast router, all
-                * incoming multicast packets are passed to the
-                * kernel-level multicast forwarding function.
-                * The packet is returned (relatively) intact; if
-                * ip6_mforward() returns a non-zero value, the packet
-                * must be discarded, else it may be accepted below.
-                */
-               if (ip6_mrouter && ip6_mforward(ip6, m->m_pkthdr.rcvif, m)) {
-                       ip6stat.ip6s_cantforward++;
-                       m_freem(m);
-                       lck_mtx_unlock(ip6_mutex);
-                       return;
+               if (!ours && nd6_prproxy) {
+                       /*
+                        * If this isn't for us, this might be a Neighbor
+                        * Solicitation (dst is solicited-node multicast)
+                        * against an address in one of the proxied prefixes;
+                        * if so, claim the packet and let icmp6_input()
+                        * handle the rest.
+                        */
+                       ours = nd6_prproxy_isours(m, ip6, NULL, IFSCOPE_NONE);
+                       VERIFY(!ours ||
+                           (m->m_pkthdr.pkt_flags & PKTF_PROXY_DST));
                }
                if (!ours) {
-                       m_freem(m);
-                       lck_mtx_unlock(ip6_mutex);
-                       return;
+                       goto bad;
                }
        } else if (!ours) {
-               ip6_forward(m, 0, 1);
-               lck_mtx_unlock(ip6_mutex);
-               return;
-       }       
+               /*
+                * The unicast forwarding function might return the packet
+                * if we are proxying prefix(es), and if the packet is an
+                * ICMPv6 packet that has failed the zone checks, but is
+                * targetted towards a proxied address (this is optimized by
+                * way of RTF_PROXY test.)  If so, claim the packet as ours
+                * and let icmp6_input() handle the rest.  The packet's hop
+                * limit value is kept intact (it's not decremented).  This
+                * is for supporting Neighbor Unreachability Detection between
+                * proxied nodes on different links (src is link-local, dst
+                * is target address.)
+                */
+               if ((m = ip6_forward(m, &rin6, 0)) == NULL) {
+                       goto done;
+               }
+               VERIFY(rin6.ro_rt != NULL);
+               VERIFY(m->m_pkthdr.pkt_flags & PKTF_PROXY_DST);
+               deliverifp = rin6.ro_rt->rt_ifp;
+               ours = 1;
+       }
 
        ip6 = mtod(m, struct ip6_hdr *);
 
@@ -869,7 +1312,7 @@ ip6_input(m)
        if (IN6_IS_ADDR_V4MAPPED(&ip6->ip6_src) ||
            IN6_IS_ADDR_V4MAPPED(&ip6->ip6_dst)) {
                ip6stat.ip6s_badscope++;
-               in6_ifstat_inc(m->m_pkthdr.rcvif, ifs6_in_addrerr);
+               in6_ifstat_inc(inifp, ifs6_in_addrerr);
                goto bad;
        }
 
@@ -877,18 +1320,39 @@ ip6_input(m)
         * Tell launch routine the next header
         */
        ip6stat.ip6s_delivered++;
-       in6_ifstat_inc(deliverifp, ifs6_in_deliver);
+       in6_ifstat_inc_na(deliverifp, ifs6_in_deliver);
 
-       lck_mtx_unlock(ip6_mutex);
 injectit:
        nest = 0;
 
+       /*
+        * Perform IP header alignment fixup again, if needed.  Note that
+        * we do it once for the outermost protocol, and we assume each
+        * protocol handler wouldn't mess with the alignment afterwards.
+        */
+       IP6_HDR_ALIGNMENT_FIXUP(m, inifp, return );
+
        while (nxt != IPPROTO_DONE) {
                struct ipfilter *filter;
-               
+               int (*pr_input)(struct mbuf **, int *, int);
+
+               /*
+                * This would imply either IPPROTO_HOPOPTS was not the first
+                * option or it did not come in the first fragment.
+                */
+               if (nxt == IPPROTO_HOPOPTS &&
+                   (m->m_pkthdr.pkt_flags & PKTF_HBH_CHKED) == 0) {
+                       /*
+                        * This implies that HBH option was not contained
+                        * in the first fragment
+                        */
+                       ip6stat.ip6s_badoptions++;
+                       goto bad;
+               }
+
                if (ip6_hdrnestlimit && (++nest > ip6_hdrnestlimit)) {
                        ip6stat.ip6s_toomanyhdr++;
-                       goto badunlocked;
+                       goto bad;
                }
 
                /*
@@ -897,25 +1361,9 @@ injectit:
                 */
                if (m->m_pkthdr.len < off) {
                        ip6stat.ip6s_tooshort++;
-                       in6_ifstat_inc(m->m_pkthdr.rcvif, ifs6_in_truncated);
-                       goto badunlocked;
-               }
-
-#if 0
-               /*
-                * do we need to do it for every header?  yeah, other
-                * functions can play with it (like re-allocate and copy).
-                */
-               mhist = ip6_addaux(m);
-               if (mhist && M_TRAILINGSPACE(mhist) >= sizeof(nxt)) {
-                       hist = mtod(mhist, caddr_t) + mhist->m_len;
-                       bcopy(&nxt, hist, sizeof(nxt));
-                       mhist->m_len += sizeof(nxt);
-               } else {
-                       ip6stat.ip6s_toomanyhdr++;
+                       in6_ifstat_inc(inifp, ifs6_in_truncated);
                        goto bad;
                }
-#endif
 
 #if IPSEC
                /*
@@ -923,88 +1371,157 @@ injectit:
                 * note that we do not visit this with protocols with pcb layer
                 * code - like udp/tcp/raw ip.
                 */
-               if ((ipsec_bypass == 0) && (ip6_protox[nxt]->pr_flags & PR_LASTHDR) != 0) {
-                       lck_mtx_lock(sadb_mutex);
+               if ((ipsec_bypass == 0) &&
+                   (ip6_protox[nxt]->pr_flags & PR_LASTHDR) != 0) {
                        if (ipsec6_in_reject(m, NULL)) {
-                               ipsec6stat.in_polvio++;
-                               lck_mtx_unlock(sadb_mutex);
-                               goto badunlocked;
-                       }
-                       lck_mtx_unlock(sadb_mutex);
+                               IPSEC_STAT_INCREMENT(ipsec6stat.in_polvio);
+                               goto bad;
+                       }
                }
-#endif
+#endif /* IPSEC */
 
                /*
-                * Call IP filter on last header only
+                * Call IP filter
                 */
-               if ((ip6_protox[nxt]->pr_flags & PR_LASTHDR) != 0 && !TAILQ_EMPTY(&ipv6_filters)) {
+               if (!TAILQ_EMPTY(&ipv6_filters) && !IFNET_IS_INTCOPROC(inifp)) {
                        ipf_ref();
                        TAILQ_FOREACH(filter, &ipv6_filters, ipf_link) {
                                if (seen == 0) {
-                                       if ((struct ipfilter *)inject_ipfref == filter)
+                                       if ((struct ipfilter *)inject_ipfref ==
+                                           filter) {
                                                seen = 1;
+                                       }
                                } else if (filter->ipf_filter.ipf_input) {
                                        errno_t result;
-                                       
+
                                        result = filter->ipf_filter.ipf_input(
-                                               filter->ipf_filter.cookie, (mbuf_t*)&m, off, nxt);
+                                               filter->ipf_filter.cookie,
+                                               (mbuf_t *)&m, off, (uint8_t)nxt);
                                        if (result == EJUSTRETURN) {
                                                ipf_unref();
-                                               return;
+                                               goto done;
                                        }
                                        if (result != 0) {
                                                ipf_unref();
-                                               m_freem(m);
-                                               return;
+                                               goto bad;
                                        }
                                }
                        }
                        ipf_unref();
                }
-               if (!(ip6_protox[nxt]->pr_flags & PR_PROTOLOCK)) {
+
+               DTRACE_IP6(receive, struct mbuf *, m, struct inpcb *, NULL,
+                   struct ip6_hdr *, ip6, struct ifnet *, inifp,
+                   struct ip *, NULL, struct ip6_hdr *, ip6);
+
+               if ((pr_input = ip6_protox[nxt]->pr_input) == NULL) {
+                       m_freem(m);
+                       m = NULL;
+                       nxt = IPPROTO_DONE;
+               } else if (!(ip6_protox[nxt]->pr_flags & PR_PROTOLOCK)) {
                        lck_mtx_lock(inet6_domain_mutex);
-                       nxt = (*ip6_protox[nxt]->pr_input)(&m, &off);
+                       nxt = pr_input(&m, &off, nxt);
                        lck_mtx_unlock(inet6_domain_mutex);
+               } else {
+                       nxt = pr_input(&m, &off, nxt);
                }
-               else
-                       nxt = (*ip6_protox[nxt]->pr_input)(&m, &off);
        }
-       return;
- bad:
-       lck_mtx_unlock(ip6_mutex);
- badunlocked:
-       m_freem(m);
-       return;
+done:
+       ROUTE_RELEASE(&rin6);
+       return;
+bad:
+       m_freem(m);
+       goto done;
+}
+
+void
+ip6_setsrcifaddr_info(struct mbuf *m, uint32_t src_idx, struct in6_ifaddr *ia6)
+{
+       VERIFY(m->m_flags & M_PKTHDR);
+
+       /*
+        * If the source ifaddr is specified, pick up the information
+        * from there; otherwise just grab the passed-in ifindex as the
+        * caller may not have the ifaddr available.
+        */
+       if (ia6 != NULL) {
+               m->m_pkthdr.pkt_flags |= PKTF_IFAINFO;
+               m->m_pkthdr.src_ifindex = ia6->ia_ifp->if_index;
+
+               /* See IN6_IFF comments in in6_var.h */
+               m->m_pkthdr.src_iff = (ia6->ia6_flags & 0xffff);
+       } else {
+               m->m_pkthdr.src_iff = 0;
+               m->m_pkthdr.src_ifindex = (uint16_t)src_idx;
+               if (src_idx != 0) {
+                       m->m_pkthdr.pkt_flags |= PKTF_IFAINFO;
+               }
+       }
+}
+
+void
+ip6_setdstifaddr_info(struct mbuf *m, uint32_t dst_idx, struct in6_ifaddr *ia6)
+{
+       VERIFY(m->m_flags & M_PKTHDR);
+
+       /*
+        * If the destination ifaddr is specified, pick up the information
+        * from there; otherwise just grab the passed-in ifindex as the
+        * caller may not have the ifaddr available.
+        */
+       if (ia6 != NULL) {
+               m->m_pkthdr.pkt_flags |= PKTF_IFAINFO;
+               m->m_pkthdr.dst_ifindex = ia6->ia_ifp->if_index;
+
+               /* See IN6_IFF comments in in6_var.h */
+               m->m_pkthdr.dst_iff = (ia6->ia6_flags & 0xffff);
+       } else {
+               m->m_pkthdr.dst_iff = 0;
+               m->m_pkthdr.dst_ifindex = (uint16_t)dst_idx;
+               if (dst_idx != 0) {
+                       m->m_pkthdr.pkt_flags |= PKTF_IFAINFO;
+               }
+       }
+}
+
+int
+ip6_getsrcifaddr_info(struct mbuf *m, uint32_t *src_idx, uint32_t *ia6f)
+{
+       VERIFY(m->m_flags & M_PKTHDR);
+
+       if (!(m->m_pkthdr.pkt_flags & PKTF_IFAINFO)) {
+               return -1;
+       }
+
+       if (src_idx != NULL) {
+               *src_idx = m->m_pkthdr.src_ifindex;
+       }
+
+       if (ia6f != NULL) {
+               *ia6f = m->m_pkthdr.src_iff;
+       }
+
+       return 0;
 }
 
-/*
- * set/grab in6_ifaddr correspond to IPv6 destination address.
- * XXX backward compatibility wrapper
- */
-static struct mbuf *
-ip6_setdstifaddr(m, ia6)
-       struct mbuf *m;
-       struct in6_ifaddr *ia6;
+int
+ip6_getdstifaddr_info(struct mbuf *m, uint32_t *dst_idx, uint32_t *ia6f)
 {
-       struct mbuf *n;
+       VERIFY(m->m_flags & M_PKTHDR);
 
-       n = ip6_addaux(m);
-       if (n)
-               mtod(n, struct ip6aux *)->ip6a_dstia6 = ia6;
-       return n;       /* NULL if failed to set */
-}
+       if (!(m->m_pkthdr.pkt_flags & PKTF_IFAINFO)) {
+               return -1;
+       }
 
-struct in6_ifaddr *
-ip6_getdstifaddr(m)
-       struct mbuf *m;
-{
-       struct mbuf *n;
+       if (dst_idx != NULL) {
+               *dst_idx = m->m_pkthdr.dst_ifindex;
+       }
 
-       n = ip6_findaux(m);
-       if (n)
-               return mtod(n, struct ip6aux *)->ip6a_dstia6;
-       else
-               return NULL;
+       if (ia6f != NULL) {
+               *ia6f = m->m_pkthdr.dst_iff;
+       }
+
+       return 0;
 }
 
 /*
@@ -1012,11 +1529,8 @@ ip6_getdstifaddr(m)
  * included, the real payload length will be stored in plenp.
  */
 static int
-ip6_hopopts_input(plenp, rtalertp, mp, offp)
-       u_int32_t *plenp;
-       u_int32_t *rtalertp;    /* XXX: should be stored more smart way */
-       struct mbuf **mp;
-       int *offp;
+ip6_hopopts_input(uint32_t *plenp, uint32_t *rtalertp, struct mbuf **mp,
+    int *offp)
 {
        struct mbuf *m = *mp;
        int off = *offp, hbhlen;
@@ -1024,39 +1538,24 @@ ip6_hopopts_input(plenp, rtalertp, mp, offp)
        u_int8_t *opt;
 
        /* validation of the length of the header */
-#ifndef PULLDOWN_TEST
-       IP6_EXTHDR_CHECK(m, off, sizeof(*hbh), return -1);
+       IP6_EXTHDR_CHECK(m, off, sizeof(*hbh), return (-1));
        hbh = (struct ip6_hbh *)(mtod(m, caddr_t) + off);
        hbhlen = (hbh->ip6h_len + 1) << 3;
 
-       IP6_EXTHDR_CHECK(m, off, hbhlen, return -1);
+       IP6_EXTHDR_CHECK(m, off, hbhlen, return (-1));
        hbh = (struct ip6_hbh *)(mtod(m, caddr_t) + off);
-#else
-       IP6_EXTHDR_GET(hbh, struct ip6_hbh *, m,
-               sizeof(struct ip6_hdr), sizeof(struct ip6_hbh));
-       if (hbh == NULL) {
-               ip6stat.ip6s_tooshort++;
-               return -1;
-       }
-       hbhlen = (hbh->ip6h_len + 1) << 3;
-       IP6_EXTHDR_GET(hbh, struct ip6_hbh *, m, sizeof(struct ip6_hdr),
-               hbhlen);
-       if (hbh == NULL) {
-               ip6stat.ip6s_tooshort++;
-               return -1;
-       }
-#endif
        off += hbhlen;
        hbhlen -= sizeof(struct ip6_hbh);
        opt = (u_int8_t *)hbh + sizeof(struct ip6_hbh);
 
        if (ip6_process_hopopts(m, (u_int8_t *)hbh + sizeof(struct ip6_hbh),
-                               hbhlen, rtalertp, plenp) < 0)
-               return(-1);
+           hbhlen, rtalertp, plenp) < 0) {
+               return -1;
+       }
 
        *offp = off;
        *mp = m;
-       return(0);
+       return 0;
 }
 
 /*
@@ -1070,12 +1569,8 @@ ip6_hopopts_input(plenp, rtalertp, mp, offp)
  * opthead + hbhlen is located in continuous memory region.
  */
 int
-ip6_process_hopopts(m, opthead, hbhlen, rtalertp, plenp)
-       struct mbuf *m;
-       u_int8_t *opthead;
-       int hbhlen;
-       u_int32_t *rtalertp;
-       u_int32_t *plenp;
+ip6_process_hopopts(struct mbuf *m, u_int8_t *opthead, int hbhlen,
+    u_int32_t *rtalertp, u_int32_t *plenp)
 {
        struct ip6_hdr *ip6;
        int optlen = 0;
@@ -1096,7 +1591,7 @@ ip6_process_hopopts(m, opthead, hbhlen, rtalertp, plenp)
                        }
                        optlen = *(opt + 1) + 2;
                        break;
-               case IP6OPT_RTALERT:
+               case IP6OPT_ROUTER_ALERT:
                        /* XXX may need check for alignment */
                        if (hbhlen < IP6OPT_RTALERT_LEN) {
                                ip6stat.ip6s_toosmall++;
@@ -1104,12 +1599,10 @@ ip6_process_hopopts(m, opthead, hbhlen, rtalertp, plenp)
                        }
                        if (*(opt + 1) != IP6OPT_RTALERT_LEN - 2) {
                                /* XXX stat */
-                               lck_mtx_unlock(ip6_mutex);
                                icmp6_error(m, ICMP6_PARAM_PROB,
-                                           ICMP6_PARAMPROB_HEADER,
-                                           erroff + opt + 1 - opthead);
-                               lck_mtx_lock(ip6_mutex);
-                               return(-1);
+                                   ICMP6_PARAMPROB_HEADER,
+                                   (int)(erroff + opt + 1 - opthead));
+                               return -1;
                        }
                        optlen = IP6OPT_RTALERT_LEN;
                        bcopy((caddr_t)(opt + 2), (caddr_t)&rtalert_val, 2);
@@ -1123,12 +1616,10 @@ ip6_process_hopopts(m, opthead, hbhlen, rtalertp, plenp)
                        }
                        if (*(opt + 1) != IP6OPT_JUMBO_LEN - 2) {
                                /* XXX stat */
-                               lck_mtx_unlock(ip6_mutex);
                                icmp6_error(m, ICMP6_PARAM_PROB,
-                                           ICMP6_PARAMPROB_HEADER,
-                                           erroff + opt + 1 - opthead);
-                               lck_mtx_lock(ip6_mutex);
-                               return(-1);
+                                   ICMP6_PARAMPROB_HEADER,
+                                   (int)(erroff + opt + 1 - opthead));
+                               return -1;
                        }
                        optlen = IP6OPT_JUMBO_LEN;
 
@@ -1139,12 +1630,10 @@ ip6_process_hopopts(m, opthead, hbhlen, rtalertp, plenp)
                        ip6 = mtod(m, struct ip6_hdr *);
                        if (ip6->ip6_plen) {
                                ip6stat.ip6s_badoptions++;
-                               lck_mtx_unlock(ip6_mutex);
                                icmp6_error(m, ICMP6_PARAM_PROB,
-                                           ICMP6_PARAMPROB_HEADER,
-                                           erroff + opt - opthead);
-                               lck_mtx_lock(ip6_mutex);
-                               return(-1);
+                                   ICMP6_PARAMPROB_HEADER,
+                                   (int)(erroff + opt - opthead));
+                               return -1;
                        }
 
                        /*
@@ -1165,12 +1654,10 @@ ip6_process_hopopts(m, opthead, hbhlen, rtalertp, plenp)
                         */
                        if (*plenp != 0) {
                                ip6stat.ip6s_badoptions++;
-                               lck_mtx_unlock(ip6_mutex);
                                icmp6_error(m, ICMP6_PARAM_PROB,
-                                           ICMP6_PARAMPROB_HEADER,
-                                           erroff + opt + 2 - opthead);
-                               lck_mtx_lock(ip6_mutex);
-                               return(-1);
+                                   ICMP6_PARAMPROB_HEADER,
+                                   (int)(erroff + opt + 2 - opthead));
+                               return -1;
                        }
 #endif
 
@@ -1179,37 +1666,34 @@ ip6_process_hopopts(m, opthead, hbhlen, rtalertp, plenp)
                         */
                        if (jumboplen <= IPV6_MAXPACKET) {
                                ip6stat.ip6s_badoptions++;
-                               lck_mtx_unlock(ip6_mutex);
                                icmp6_error(m, ICMP6_PARAM_PROB,
-                                           ICMP6_PARAMPROB_HEADER,
-                                           erroff + opt + 2 - opthead);
-                               lck_mtx_lock(ip6_mutex);
-                               return(-1);
+                                   ICMP6_PARAMPROB_HEADER,
+                                   (int)(erroff + opt + 2 - opthead));
+                               return -1;
                        }
                        *plenp = jumboplen;
 
                        break;
-               default:                /* unknown option */
+               default:                /* unknown option */
                        if (hbhlen < IP6OPT_MINLEN) {
                                ip6stat.ip6s_toosmall++;
                                goto bad;
                        }
                        optlen = ip6_unknown_opt(opt, m,
-                           erroff + opt - opthead, 1);
+                           erroff + opt - opthead);
                        if (optlen == -1) {
-                               /* ip6_unknown opt unlocked ip6_mutex */
-                               return(-1);
+                               return -1;
                        }
                        optlen += 2;
                        break;
                }
        }
 
-       return(0);
+       return 0;
 
-  bad: 
+bad:
        m_freem(m);
-       return(-1);
+       return -1;
 }
 
 /*
@@ -1219,103 +1703,216 @@ ip6_process_hopopts(m, opthead, hbhlen, rtalertp, plenp)
  * is not continuous in order to return an ICMPv6 error.
  */
 int
-ip6_unknown_opt(optp, m, off, locked)
-       u_int8_t *optp;
-       struct mbuf *m;
-       int off;
-       int locked;
+ip6_unknown_opt(uint8_t *optp, struct mbuf *m, size_t off)
 {
        struct ip6_hdr *ip6;
 
        switch (IP6OPT_TYPE(*optp)) {
        case IP6OPT_TYPE_SKIP: /* ignore the option */
-               return((int)*(optp + 1));
-       case IP6OPT_TYPE_DISCARD:       /* silently discard */
+               return (int)*(optp + 1);
+
+       case IP6OPT_TYPE_DISCARD:       /* silently discard */
                m_freem(m);
-               return(-1);
+               return -1;
+
        case IP6OPT_TYPE_FORCEICMP: /* send ICMP even if multicasted */
                ip6stat.ip6s_badoptions++;
-               if (locked)
-                       lck_mtx_unlock(ip6_mutex);
-               icmp6_error(m, ICMP6_PARAM_PROB, ICMP6_PARAMPROB_OPTION, off);
-               if (locked)
-                       lck_mtx_lock(ip6_mutex);
-               return(-1);
+               icmp6_error(m, ICMP6_PARAM_PROB, ICMP6_PARAMPROB_OPTION, (int)off);
+               return -1;
+
        case IP6OPT_TYPE_ICMP: /* send ICMP if not multicasted */
                ip6stat.ip6s_badoptions++;
                ip6 = mtod(m, struct ip6_hdr *);
                if (IN6_IS_ADDR_MULTICAST(&ip6->ip6_dst) ||
-                   (m->m_flags & (M_BCAST|M_MCAST)))
+                   (m->m_flags & (M_BCAST | M_MCAST))) {
                        m_freem(m);
-               else {
-                       if (locked)
-                               lck_mtx_unlock(ip6_mutex);
+               } else {
                        icmp6_error(m, ICMP6_PARAM_PROB,
-                                   ICMP6_PARAMPROB_OPTION, off);
-                       if (locked)
-                               lck_mtx_lock(ip6_mutex);
+                           ICMP6_PARAMPROB_OPTION, (int)off);
                }
-               return(-1);
+               return -1;
        }
 
-       m_freem(m);             /* XXX: NOTREACHED */
-       return(-1);
+       m_freem(m);             /* XXX: NOTREACHED */
+       return -1;
 }
 
 /*
  * Create the "control" list for this pcb.
- * The function will not modify mbuf chain at all.
+ * These functions will not modify mbuf chain at all.
  *
- * with KAME mbuf chain restriction:
+ * With KAME mbuf chain restriction:
  * The routine will be called from upper layer handlers like tcp6_input().
  * Thus the routine assumes that the caller (tcp6_input) have already
  * called IP6_EXTHDR_CHECK() and all the extension headers are located in the
  * very first mbuf on the mbuf chain.
+ *
+ * ip6_savecontrol_v4 will handle those options that are possible to be
+ * set on a v4-mapped socket.
+ * ip6_savecontrol will directly call ip6_savecontrol_v4 to handle those
+ * options and handle the v6-only ones itself.
  */
-void
-ip6_savecontrol(in6p, mp, ip6, m)
-       struct inpcb *in6p;
-       struct mbuf **mp;
-       struct ip6_hdr *ip6;
-       struct mbuf *m;
+struct mbuf **
+ip6_savecontrol_v4(struct inpcb *inp, struct mbuf *m, struct mbuf **mp,
+    int *v4only)
 {
-       int rthdr_exist = 0;
+       struct ip6_hdr *ip6 = mtod(m, struct ip6_hdr *);
 
-#if SO_TIMESTAMP
-       if ((in6p->in6p_socket->so_options & SO_TIMESTAMP) != 0) {
+       if ((inp->inp_socket->so_options & SO_TIMESTAMP) != 0) {
                struct timeval tv;
 
-               microtime(&tv);
-               *mp = sbcreatecontrol((caddr_t) &tv, sizeof(tv),
-                                     SCM_TIMESTAMP, SOL_SOCKET);
-               if (*mp) {
-                       mp = &(*mp)->m_next;
+               getmicrotime(&tv);
+               mp = sbcreatecontrol_mbuf((caddr_t)&tv, sizeof(tv),
+                   SCM_TIMESTAMP, SOL_SOCKET, mp);
+               if (*mp == NULL) {
+                       return NULL;
+               }
+       }
+       if ((inp->inp_socket->so_options & SO_TIMESTAMP_MONOTONIC) != 0) {
+               uint64_t time;
+
+               time = mach_absolute_time();
+               mp = sbcreatecontrol_mbuf((caddr_t)&time, sizeof(time),
+                   SCM_TIMESTAMP_MONOTONIC, SOL_SOCKET, mp);
+               if (*mp == NULL) {
+                       return NULL;
+               }
+       }
+       if ((inp->inp_socket->so_options & SO_TIMESTAMP_CONTINUOUS) != 0) {
+               uint64_t time;
+
+               time = mach_continuous_time();
+               mp = sbcreatecontrol_mbuf((caddr_t)&time, sizeof(time),
+                   SCM_TIMESTAMP_CONTINUOUS, SOL_SOCKET, mp);
+               if (*mp == NULL) {
+                       return NULL;
+               }
+       }
+       if ((inp->inp_socket->so_flags & SOF_RECV_TRAFFIC_CLASS) != 0) {
+               int tc = m_get_traffic_class(m);
+
+               mp = sbcreatecontrol_mbuf((caddr_t)&tc, sizeof(tc),
+                   SO_TRAFFIC_CLASS, SOL_SOCKET, mp);
+               if (*mp == NULL) {
+                       return NULL;
                }
        }
+
+#define IS2292(inp, x, y)       (((inp)->inp_flags & IN6P_RFC2292) ? (x) : (y))
+       if ((ip6->ip6_vfc & IPV6_VERSION_MASK) != IPV6_VERSION) {
+               if (v4only != NULL) {
+                       *v4only = 1;
+               }
+
+               // Send ECN flags for v4-mapped addresses
+               if ((inp->inp_flags & IN6P_TCLASS) != 0) {
+                       struct ip *ip_header = mtod(m, struct ip *);
+
+                       int tclass = (int)(ip_header->ip_tos);
+                       mp = sbcreatecontrol_mbuf((caddr_t)&tclass, sizeof(tclass),
+                           IPV6_TCLASS, IPPROTO_IPV6, mp);
+                       if (*mp == NULL) {
+                               return NULL;
+                       }
+               }
+
+               // Send IN6P_PKTINFO for v4-mapped address
+               if ((inp->inp_flags & IN6P_PKTINFO) != 0
+#if CONTENT_FILTER
+                   /* Content Filter needs to see local address */
+                   || (inp->inp_socket->so_cfil_db != NULL)
 #endif
+                   ) {
+                       struct in6_pktinfo pi6 = {
+                               .ipi6_addr = IN6ADDR_V4MAPPED_INIT,
+                               .ipi6_ifindex = (m && m->m_pkthdr.rcvif) ? m->m_pkthdr.rcvif->if_index : 0,
+                       };
+
+                       struct ip *ip_header = mtod(m, struct ip *);
+                       bcopy(&ip_header->ip_dst, &pi6.ipi6_addr.s6_addr32[3], sizeof(struct in_addr));
+
+                       mp = sbcreatecontrol_mbuf((caddr_t)&pi6,
+                           sizeof(struct in6_pktinfo),
+                           IS2292(inp, IPV6_2292PKTINFO, IPV6_PKTINFO),
+                           IPPROTO_IPV6, mp);
+                       if (*mp == NULL) {
+                               return NULL;
+                       }
+               }
+               return mp;
+       }
 
        /* RFC 2292 sec. 5 */
-       if ((in6p->in6p_flags & IN6P_PKTINFO) != 0) {
+       if ((inp->inp_flags & IN6P_PKTINFO) != 0
+#if CONTENT_FILTER
+           /* Content Filter needs to see local address */
+           || (inp->inp_socket->so_cfil_db != NULL)
+#endif
+           ) {
                struct in6_pktinfo pi6;
+
                bcopy(&ip6->ip6_dst, &pi6.ipi6_addr, sizeof(struct in6_addr));
-               if (IN6_IS_SCOPE_LINKLOCAL(&pi6.ipi6_addr))
-                       pi6.ipi6_addr.s6_addr16[1] = 0;
-               pi6.ipi6_ifindex = (m && m->m_pkthdr.rcvif)
-                                       ? m->m_pkthdr.rcvif->if_index
-                                       : 0;
-               *mp = sbcreatecontrol((caddr_t) &pi6,
-                       sizeof(struct in6_pktinfo), IPV6_PKTINFO,
-                       IPPROTO_IPV6);
-               if (*mp)
-                       mp = &(*mp)->m_next;
-       }
-
-       if ((in6p->in6p_flags & IN6P_HOPLIMIT) != 0) {
+               in6_clearscope(&pi6.ipi6_addr); /* XXX */
+               pi6.ipi6_ifindex =
+                   (m && m->m_pkthdr.rcvif) ? m->m_pkthdr.rcvif->if_index : 0;
+
+               mp = sbcreatecontrol_mbuf((caddr_t)&pi6,
+                   sizeof(struct in6_pktinfo),
+                   IS2292(inp, IPV6_2292PKTINFO, IPV6_PKTINFO),
+                   IPPROTO_IPV6, mp);
+               if (*mp == NULL) {
+                       return NULL;
+               }
+       }
+
+       if ((inp->inp_flags & IN6P_HOPLIMIT) != 0) {
                int hlim = ip6->ip6_hlim & 0xff;
-               *mp = sbcreatecontrol((caddr_t) &hlim,
-                       sizeof(int), IPV6_HOPLIMIT, IPPROTO_IPV6);
-               if (*mp)
-                       mp = &(*mp)->m_next;
+
+               mp = sbcreatecontrol_mbuf((caddr_t)&hlim, sizeof(int),
+                   IS2292(inp, IPV6_2292HOPLIMIT, IPV6_HOPLIMIT),
+                   IPPROTO_IPV6, mp);
+               if (*mp == NULL) {
+                       return NULL;
+               }
+       }
+
+       if (v4only != NULL) {
+               *v4only = 0;
+       }
+       return mp;
+}
+
+int
+ip6_savecontrol(struct inpcb *in6p, struct mbuf *m, struct mbuf **mp)
+{
+       struct mbuf **np;
+       struct ip6_hdr *ip6 = mtod(m, struct ip6_hdr *);
+       int v4only = 0;
+
+       *mp = NULL;
+       np = ip6_savecontrol_v4(in6p, m, mp, &v4only);
+       if (np == NULL) {
+               goto no_mbufs;
+       }
+
+       mp = np;
+       if (v4only) {
+               return 0;
+       }
+
+       if ((in6p->inp_flags & IN6P_TCLASS) != 0) {
+               u_int32_t flowinfo;
+               int tclass;
+
+               flowinfo = (u_int32_t)ntohl(ip6->ip6_flow & IPV6_FLOWINFO_MASK);
+               flowinfo >>= 20;
+
+               tclass = flowinfo & 0xff;
+               mp = sbcreatecontrol_mbuf((caddr_t)&tclass, sizeof(tclass),
+                   IPV6_TCLASS, IPPROTO_IPV6, mp);
+               if (*mp == NULL) {
+                       goto no_mbufs;
+               }
        }
 
        /*
@@ -1323,93 +1920,41 @@ ip6_savecontrol(in6p, mp, ip6, m)
         * privilege for the option (see ip6_ctloutput), but it might be too
         * strict, since there might be some hop-by-hop options which can be
         * returned to normal user.
-        * See RFC 2292 section 6.
+        * See also RFC 2292 section 6 (or RFC 3542 section 8).
         */
-       if ((in6p->in6p_flags & IN6P_HOPOPTS) != 0) {
+       if ((in6p->inp_flags & IN6P_HOPOPTS) != 0) {
                /*
                 * Check if a hop-by-hop options header is contatined in the
                 * received packet, and if so, store the options as ancillary
                 * data. Note that a hop-by-hop options header must be
-                * just after the IPv6 header, which fact is assured through
-                * the IPv6 input processing.
+                * just after the IPv6 header, which is assured through the
+                * IPv6 input processing.
                 */
                ip6 = mtod(m, struct ip6_hdr *);
                if (ip6->ip6_nxt == IPPROTO_HOPOPTS) {
                        struct ip6_hbh *hbh;
                        int hbhlen = 0;
-#if PULLDOWN_TEST
-                       struct mbuf *ext;
-#endif
-
-#ifndef PULLDOWN_TEST
                        hbh = (struct ip6_hbh *)(ip6 + 1);
                        hbhlen = (hbh->ip6h_len + 1) << 3;
-#else
-                       ext = ip6_pullexthdr(m, sizeof(struct ip6_hdr),
-                           ip6->ip6_nxt);
-                       if (ext == NULL) {
-                               ip6stat.ip6s_tooshort++;
-                               return;
-                       }
-                       hbh = mtod(ext, struct ip6_hbh *);
-                       hbhlen = (hbh->ip6h_len + 1) << 3;
-                       if (hbhlen != ext->m_len) {
-                               m_freem(ext);
-                               ip6stat.ip6s_tooshort++;
-                               return;
-                       }
-#endif
 
                        /*
-                        * XXX: We copy whole the header even if a jumbo
-                        * payload option is included, which option is to
-                        * be removed before returning in the RFC 2292.
-                        * Note: this constraint is removed in 2292bis.
+                        * XXX: We copy the whole header even if a
+                        * jumbo payload option is included, the option which
+                        * is to be removed before returning according to
+                        * RFC2292.
+                        * Note: this constraint is removed in RFC3542
                         */
-                       *mp = sbcreatecontrol((caddr_t)hbh, hbhlen,
-                                             IPV6_HOPOPTS, IPPROTO_IPV6);
-                       if (*mp)
-                               mp = &(*mp)->m_next;
-#if PULLDOWN_TEST
-                       m_freem(ext);
-#endif
-               }
-       }
-
-       /* IPV6_DSTOPTS and IPV6_RTHDR socket options */
-       if ((in6p->in6p_flags & (IN6P_DSTOPTS | IN6P_RTHDRDSTOPTS)) != 0) {
-               int proto, off, nxt;
-
-               /*
-                * go through the header chain to see if a routing header is
-                * contained in the packet. We need this information to store
-                * destination options headers (if any) properly.
-                * XXX: performance issue. We should record this info when
-                * processing extension headers in incoming routine.
-                * (todo) use m_aux? 
-                */
-               proto = IPPROTO_IPV6;
-               off = 0;
-               nxt = -1;
-               while (1) {
-                       int newoff;
+                       mp = sbcreatecontrol_mbuf((caddr_t)hbh, hbhlen,
+                           IS2292(in6p, IPV6_2292HOPOPTS, IPV6_HOPOPTS),
+                           IPPROTO_IPV6, mp);
 
-                       newoff = ip6_nexthdr(m, off, proto, &nxt);
-                       if (newoff < 0)
-                               break;
-                       if (newoff < off) /* invalid, check for safety */
-                               break;
-                       if ((proto = nxt) == IPPROTO_ROUTING) {
-                               rthdr_exist = 1;
-                               break;
+                       if (*mp == NULL) {
+                               goto no_mbufs;
                        }
-                       off = newoff;
                }
        }
 
-       if ((in6p->in6p_flags &
-            (IN6P_RTHDR | IN6P_DSTOPTS | IN6P_RTHDRDSTOPTS)) != 0) {
-               ip6 = mtod(m, struct ip6_hdr *);
+       if ((in6p->inp_flags & (IN6P_RTHDR | IN6P_DSTOPTS)) != 0) {
                int nxt = ip6->ip6_nxt, off = sizeof(struct ip6_hdr);
 
                /*
@@ -1419,12 +1964,9 @@ ip6_savecontrol(in6p, mp, ip6, m)
                 * Note that the order of the headers remains in
                 * the chain of ancillary data.
                 */
-               while (1) {     /* is explicit loop prevention necessary? */
+               while (1) {     /* is explicit loop prevention necessary? */
                        struct ip6_ext *ip6e = NULL;
                        int elen;
-#if PULLDOWN_TEST
-                       struct mbuf *ext = NULL;
-#endif
 
                        /*
                         * if it is not an extension header, don't try to
@@ -1440,54 +1982,43 @@ ip6_savecontrol(in6p, mp, ip6, m)
                                goto loopend;
                        }
 
-#ifndef PULLDOWN_TEST
-                       if (off + sizeof(*ip6e) > m->m_len)
+                       if (off + sizeof(*ip6e) > m->m_len) {
                                goto loopend;
+                       }
                        ip6e = (struct ip6_ext *)(mtod(m, caddr_t) + off);
-                       if (nxt == IPPROTO_AH)
+                       if (nxt == IPPROTO_AH) {
                                elen = (ip6e->ip6e_len + 2) << 2;
-                       else
+                       } else {
                                elen = (ip6e->ip6e_len + 1) << 3;
-                       if (off + elen > m->m_len)
-                               goto loopend;
-#else
-                       ext = ip6_pullexthdr(m, off, nxt);
-                       if (ext == NULL) {
-                               ip6stat.ip6s_tooshort++;
-                               return;
                        }
-                       ip6e = mtod(ext, struct ip6_ext *);
-                       if (nxt == IPPROTO_AH)
-                               elen = (ip6e->ip6e_len + 2) << 2;
-                       else
-                               elen = (ip6e->ip6e_len + 1) << 3;
-                       if (elen != ext->m_len) {
-                               m_freem(ext);
-                               ip6stat.ip6s_tooshort++;
-                               return;
+                       if (off + elen > m->m_len) {
+                               goto loopend;
                        }
-#endif
 
                        switch (nxt) {
                        case IPPROTO_DSTOPTS:
-                               if ((in6p->in6p_flags & IN6P_DSTOPTS) == 0)
+                               if (!(in6p->inp_flags & IN6P_DSTOPTS)) {
                                        break;
+                               }
 
-                               *mp = sbcreatecontrol((caddr_t)ip6e, elen,
-                                                     IPV6_DSTOPTS,
-                                                     IPPROTO_IPV6);
-                               if (*mp)
-                                       mp = &(*mp)->m_next;
+                               mp = sbcreatecontrol_mbuf((caddr_t)ip6e, elen,
+                                   IS2292(in6p, IPV6_2292DSTOPTS,
+                                   IPV6_DSTOPTS), IPPROTO_IPV6, mp);
+                               if (*mp == NULL) {
+                                       goto no_mbufs;
+                               }
                                break;
                        case IPPROTO_ROUTING:
-                               if (!in6p->in6p_flags & IN6P_RTHDR)
+                               if (!(in6p->inp_flags & IN6P_RTHDR)) {
                                        break;
+                               }
 
-                               *mp = sbcreatecontrol((caddr_t)ip6e, elen,
-                                                     IPV6_RTHDR,
-                                                     IPPROTO_IPV6);
-                               if (*mp)
-                                       mp = &(*mp)->m_next;
+                               mp = sbcreatecontrol_mbuf((caddr_t)ip6e, elen,
+                                   IS2292(in6p, IPV6_2292RTHDR, IPV6_RTHDR),
+                                   IPPROTO_IPV6, mp);
+                               if (*mp == NULL) {
+                                       goto no_mbufs;
+                               }
                                break;
                        case IPPROTO_HOPOPTS:
                        case IPPROTO_AH: /* is it possible? */
@@ -1495,88 +2026,81 @@ ip6_savecontrol(in6p, mp, ip6, m)
 
                        default:
                                /*
-                                * other cases have been filtered in the above.
+                                * other cases have been filtered in the above.
                                 * none will visit this case.  here we supply
                                 * the code just in case (nxt overwritten or
                                 * other cases).
                                 */
-#if PULLDOWN_TEST
-                               m_freem(ext);
-#endif
                                goto loopend;
-
                        }
 
                        /* proceed with the next header. */
                        off += elen;
                        nxt = ip6e->ip6e_nxt;
                        ip6e = NULL;
-#if PULLDOWN_TEST
-                       m_freem(ext);
-                       ext = NULL;
-#endif
                }
-         loopend:
+loopend:
                ;
        }
-
+       return 0;
+no_mbufs:
+       ip6stat.ip6s_pktdropcntrl++;
+       /* XXX increment a stat to show the failure */
+       return ENOBUFS;
 }
+#undef IS2292
 
-#if PULLDOWN_TEST
-/*
- * pull single extension header from mbuf chain.  returns single mbuf that
- * contains the result, or NULL on error.
- */
-static struct mbuf *
-ip6_pullexthdr(m, off, nxt)
-       struct mbuf *m;
-       size_t off;
-       int nxt;
+void
+ip6_notify_pmtu(struct inpcb *in6p, struct sockaddr_in6 *dst, u_int32_t *mtu)
 {
-       struct ip6_ext ip6e;
-       size_t elen;
-       struct mbuf *n;
+       struct socket *so;
+       struct mbuf *m_mtu;
+       struct ip6_mtuinfo mtuctl;
 
-#if DIAGNOSTIC
-       switch (nxt) {
-       case IPPROTO_DSTOPTS:
-       case IPPROTO_ROUTING:
-       case IPPROTO_HOPOPTS:
-       case IPPROTO_AH: /* is it possible? */
-               break;
-       default:
-               printf("ip6_pullexthdr: invalid nxt=%d\n", nxt);
+       so =  in6p->inp_socket;
+
+       if ((in6p->inp_flags & IN6P_MTU) == 0) {
+               return;
+       }
+
+       if (mtu == NULL) {
+               return;
+       }
+
+#ifdef DIAGNOSTIC
+       if (so == NULL) {               /* I believe this is impossible */
+               panic("ip6_notify_pmtu: socket is NULL");
+               /* NOTREACHED */
        }
 #endif
 
-       m_copydata(m, off, sizeof(ip6e), (caddr_t)&ip6e);
-       if (nxt == IPPROTO_AH)
-               elen = (ip6e.ip6e_len + 2) << 2;
-       else
-               elen = (ip6e.ip6e_len + 1) << 3;
+       if (IN6_IS_ADDR_UNSPECIFIED(&in6p->in6p_faddr) &&
+           (so->so_proto == NULL || so->so_proto->pr_protocol == IPPROTO_TCP)) {
+               return;
+       }
 
-       MGET(n, M_DONTWAIT, MT_DATA);
-       if (n && elen >= MLEN) {
-               MCLGET(n, M_DONTWAIT);
-               if ((n->m_flags & M_EXT) == 0) {
-                       m_free(n);
-                       n = NULL;
-               }
+       if (!IN6_IS_ADDR_UNSPECIFIED(&in6p->in6p_faddr) &&
+           !IN6_ARE_ADDR_EQUAL(&in6p->in6p_faddr, &dst->sin6_addr)) {
+               return;
+       }
+
+       bzero(&mtuctl, sizeof(mtuctl));         /* zero-clear for safety */
+       mtuctl.ip6m_mtu = *mtu;
+       mtuctl.ip6m_addr = *dst;
+       if (sa6_recoverscope(&mtuctl.ip6m_addr, TRUE)) {
+               return;
        }
-       if (!n)
-               return NULL;
 
-       n->m_len = 0;
-       if (elen >= M_TRAILINGSPACE(n)) {
-               m_free(n);
-               return NULL;
+       if ((m_mtu = sbcreatecontrol((caddr_t)&mtuctl, sizeof(mtuctl),
+           IPV6_PATHMTU, IPPROTO_IPV6)) == NULL) {
+               return;
        }
 
-       m_copydata(m, off, elen, mtod(n, caddr_t));
-       n->m_len = elen;
-       return n;
+       if (sbappendaddr(&so->so_rcv, SA(dst), NULL, m_mtu, NULL) == 0) {
+               return;
+       }
+       sorwakeup(so);
 }
-#endif
 
 /*
  * Get pointer to the previous header followed by the header
@@ -1591,15 +2115,13 @@ ip6_pullexthdr(m, off, nxt)
  * we develop `neater' mechanism to process extension headers.
  */
 char *
-ip6_get_prevhdr(m, off)
-       struct mbuf *m;
-       int off;
+ip6_get_prevhdr(struct mbuf *m, int off)
 {
        struct ip6_hdr *ip6 = mtod(m, struct ip6_hdr *);
 
-       if (off == sizeof(struct ip6_hdr))
-               return(&ip6->ip6_nxt);
-       else {
+       if (off == sizeof(struct ip6_hdr)) {
+               return (char *)&ip6->ip6_nxt;
+       else {
                int len, nxt;
                struct ip6_ext *ip6e = NULL;
 
@@ -1621,10 +2143,11 @@ ip6_get_prevhdr(m, off)
                        }
                        nxt = ip6e->ip6e_nxt;
                }
-               if (ip6e)
-                       return(&ip6e->ip6e_nxt);
-               else
+               if (ip6e) {
+                       return (char *)&ip6e->ip6e_nxt;
+               } else {
                        return NULL;
+               }
        }
 }
 
@@ -1632,29 +2155,27 @@ ip6_get_prevhdr(m, off)
  * get next header offset.  m will be retained.
  */
 int
-ip6_nexthdr(m, off, proto, nxtp)
-       struct mbuf *m;
-       int off;
-       int proto;
-       int *nxtp;
+ip6_nexthdr(struct mbuf *m, int off, int proto, int *nxtp)
 {
        struct ip6_hdr ip6;
        struct ip6_ext ip6e;
        struct ip6_frag fh;
 
        /* just in case */
-       if (m == NULL)
-               panic("ip6_nexthdr: m == NULL");
-       if ((m->m_flags & M_PKTHDR) == 0 || m->m_pkthdr.len < off)
+       VERIFY(m != NULL);
+       if ((m->m_flags & M_PKTHDR) == 0 || m->m_pkthdr.len < off) {
                return -1;
+       }
 
        switch (proto) {
        case IPPROTO_IPV6:
-               if (m->m_pkthdr.len < off + sizeof(ip6))
+               if (m->m_pkthdr.len < off + sizeof(ip6)) {
                        return -1;
+               }
                m_copydata(m, off, sizeof(ip6), (caddr_t)&ip6);
-               if (nxtp)
+               if (nxtp) {
                        *nxtp = ip6.ip6_nxt;
+               }
                off += sizeof(ip6);
                return off;
 
@@ -1663,34 +2184,41 @@ ip6_nexthdr(m, off, proto, nxtp)
                 * terminate parsing if it is not the first fragment,
                 * it does not make sense to parse through it.
                 */
-               if (m->m_pkthdr.len < off + sizeof(fh))
+               if (m->m_pkthdr.len < off + sizeof(fh)) {
                        return -1;
+               }
                m_copydata(m, off, sizeof(fh), (caddr_t)&fh);
                /* IP6F_OFF_MASK = 0xfff8(BigEndian), 0xf8ff(LittleEndian) */
-               if (fh.ip6f_offlg & IP6F_OFF_MASK)
+               if (fh.ip6f_offlg & IP6F_OFF_MASK) {
                        return -1;
-               if (nxtp)
+               }
+               if (nxtp) {
                        *nxtp = fh.ip6f_nxt;
+               }
                off += sizeof(struct ip6_frag);
                return off;
 
        case IPPROTO_AH:
-               if (m->m_pkthdr.len < off + sizeof(ip6e))
+               if (m->m_pkthdr.len < off + sizeof(ip6e)) {
                        return -1;
+               }
                m_copydata(m, off, sizeof(ip6e), (caddr_t)&ip6e);
-               if (nxtp)
+               if (nxtp) {
                        *nxtp = ip6e.ip6e_nxt;
+               }
                off += (ip6e.ip6e_len + 2) << 2;
                return off;
 
        case IPPROTO_HOPOPTS:
        case IPPROTO_ROUTING:
        case IPPROTO_DSTOPTS:
-               if (m->m_pkthdr.len < off + sizeof(ip6e))
+               if (m->m_pkthdr.len < off + sizeof(ip6e)) {
                        return -1;
+               }
                m_copydata(m, off, sizeof(ip6e), (caddr_t)&ip6e);
-               if (nxtp)
+               if (nxtp) {
                        *nxtp = ip6e.ip6e_nxt;
+               }
                off += (ip6e.ip6e_len + 1) << 3;
                return off;
 
@@ -1703,19 +2231,13 @@ ip6_nexthdr(m, off, proto, nxtp)
        default:
                return -1;
        }
-
-       return -1;
 }
 
 /*
  * get offset for the last header in the chain.  m will be kept untainted.
  */
 int
-ip6_lasthdr(m, off, proto, nxtp)
-       struct mbuf *m;
-       int off;
-       int proto;
-       int *nxtp;
+ip6_lasthdr(struct mbuf *m, int off, int proto, int *nxtp)
 {
        int newoff;
        int nxt;
@@ -1726,78 +2248,253 @@ ip6_lasthdr(m, off, proto, nxtp)
        }
        while (1) {
                newoff = ip6_nexthdr(m, off, proto, nxtp);
-               if (newoff < 0)
+               if (newoff < 0) {
                        return off;
-               else if (newoff < off)
-                       return -1;      /* invalid */
-               else if (newoff == off)
+               } else if (newoff < off) {
+                       return -1;    /* invalid */
+               } else if (newoff == off) {
                        return newoff;
+               }
 
                off = newoff;
                proto = *nxtp;
        }
 }
 
-struct mbuf *
-ip6_addaux(m)
-       struct mbuf *m;
+boolean_t
+ip6_pkt_has_ulp(struct mbuf *m)
 {
-       struct mbuf *n;
+       int off = 0, nxt = IPPROTO_NONE;
 
-#if DIAGNOSTIC
-       if (sizeof(struct ip6aux) > MHLEN)
-               panic("assumption failed on sizeof(ip6aux)");
-#endif
-       n = m_aux_find(m, AF_INET6, -1);
-       if (n) {
-               if (n->m_len < sizeof(struct ip6aux)) {
-                       printf("conflicting use of ip6aux");
-                       return NULL;
+       off = ip6_lasthdr(m, 0, IPPROTO_IPV6, &nxt);
+       if (off < 0 || m->m_pkthdr.len < off) {
+               return FALSE;
+       }
+
+       switch (nxt) {
+       case IPPROTO_TCP:
+               if (off + sizeof(struct tcphdr) > m->m_pkthdr.len) {
+                       return FALSE;
                }
-       } else {
-               n = m_aux_add(m, AF_INET6, -1);
-               if (n) {
-                       n->m_len = sizeof(struct ip6aux);
-                       bzero(mtod(n, caddr_t), n->m_len);
+               break;
+       case IPPROTO_UDP:
+               if (off + sizeof(struct udphdr) > m->m_pkthdr.len) {
+                       return FALSE;
                }
+               break;
+       case IPPROTO_ICMPV6:
+               if (off + sizeof(uint32_t) > m->m_pkthdr.len) {
+                       return FALSE;
+               }
+               break;
+       case IPPROTO_NONE:
+               return TRUE;
+       case IPPROTO_ESP:
+               return TRUE;
+       case IPPROTO_IPCOMP:
+               return TRUE;
+       default:
+               return FALSE;
        }
-       return n;
+       return TRUE;
 }
 
-struct mbuf *
-ip6_findaux(m)
-       struct mbuf *m;
+struct ip6aux *
+ip6_addaux(struct mbuf *m)
 {
-       struct mbuf *n;
-
-       n = m_aux_find(m, AF_INET6, -1);
-       if (n && n->m_len < sizeof(struct ip6aux)) {
-               printf("conflicting use of ip6aux");
-               n = NULL;
+       struct m_tag            *tag;
+
+       /* Check if one is already allocated */
+       tag = m_tag_locate(m, KERNEL_MODULE_TAG_ID,
+           KERNEL_TAG_TYPE_INET6, NULL);
+       if (tag == NULL) {
+               /* Allocate a tag */
+               tag = m_tag_create(KERNEL_MODULE_TAG_ID, KERNEL_TAG_TYPE_INET6,
+                   sizeof(struct ip6aux), M_DONTWAIT, m);
+
+               /* Attach it to the mbuf */
+               if (tag) {
+                       m_tag_prepend(m, tag);
+               }
        }
-       return n;
+
+       return tag ? (struct ip6aux *)(tag + 1) : NULL;
+}
+
+struct ip6aux *
+ip6_findaux(struct mbuf *m)
+{
+       struct m_tag    *tag;
+
+       tag = m_tag_locate(m, KERNEL_MODULE_TAG_ID,
+           KERNEL_TAG_TYPE_INET6, NULL);
+
+       return tag ? (struct ip6aux *)(tag + 1) : NULL;
 }
 
 void
-ip6_delaux(m)
-       struct mbuf *m;
+ip6_delaux(struct mbuf *m)
 {
-       struct mbuf *n;
+       struct m_tag    *tag;
+
+       tag = m_tag_locate(m, KERNEL_MODULE_TAG_ID,
+           KERNEL_TAG_TYPE_INET6, NULL);
+       if (tag) {
+               m_tag_delete(m, tag);
+       }
+}
 
-       n = m_aux_find(m, AF_INET6, -1);
-       if (n)
-               m_aux_delete(m, n);
+/*
+ * Drain callback
+ */
+void
+ip6_drain(void)
+{
+       frag6_drain();          /* fragments */
+       in6_rtqdrain();         /* protocol cloned routes */
+       nd6_drain(NULL);        /* cloned routes: ND6 */
 }
 
 /*
  * System control for IP6
  */
 
-u_char inet6ctlerrmap[PRC_NCMDS] = {
-       0,              0,              0,              0,
-       0,              EMSGSIZE,       EHOSTDOWN,      EHOSTUNREACH,
-       EHOSTUNREACH,   EHOSTUNREACH,   ECONNREFUSED,   ECONNREFUSED,
-       EMSGSIZE,       EHOSTUNREACH,   0,              0,
-       0,              0,              0,              0,
+u_char  inet6ctlerrmap[PRC_NCMDS] = {
+       0, 0, 0, 0,
+       0, EMSGSIZE, EHOSTDOWN, EHOSTUNREACH,
+       EHOSTUNREACH, EHOSTUNREACH, ECONNREFUSED, ECONNREFUSED,
+       EMSGSIZE, EHOSTUNREACH, 0, 0,
+       0, 0, 0, 0,
        ENOPROTOOPT
 };
+
+static int
+sysctl_reset_ip6_input_stats SYSCTL_HANDLER_ARGS
+{
+#pragma unused(arg1, arg2)
+       int error, i;
+
+       i = ip6_input_measure;
+       error = sysctl_handle_int(oidp, &i, 0, req);
+       if (error || req->newptr == USER_ADDR_NULL) {
+               goto done;
+       }
+       /* impose bounds */
+       if (i < 0 || i > 1) {
+               error = EINVAL;
+               goto done;
+       }
+       if (ip6_input_measure != i && i == 1) {
+               net_perf_initialize(&net_perf, ip6_input_measure_bins);
+       }
+       ip6_input_measure = i;
+done:
+       return error;
+}
+
+static int
+sysctl_ip6_input_measure_bins SYSCTL_HANDLER_ARGS
+{
+#pragma unused(arg1, arg2)
+       int error;
+       uint64_t i;
+
+       i = ip6_input_measure_bins;
+       error = sysctl_handle_quad(oidp, &i, 0, req);
+       if (error || req->newptr == USER_ADDR_NULL) {
+               goto done;
+       }
+       /* validate data */
+       if (!net_perf_validate_bins(i)) {
+               error = EINVAL;
+               goto done;
+       }
+       ip6_input_measure_bins = i;
+done:
+       return error;
+}
+
+static int
+sysctl_ip6_input_getperf SYSCTL_HANDLER_ARGS
+{
+#pragma unused(oidp, arg1, arg2)
+       if (req->oldptr == USER_ADDR_NULL) {
+               req->oldlen = (size_t)sizeof(struct net_perf);
+       }
+
+       return SYSCTL_OUT(req, &net_perf, MIN(sizeof(net_perf), req->oldlen));
+}
+
+
+/*
+ * Initialize IPv6 source address hash table.
+ */
+static void
+in6_ifaddrhashtbl_init(void)
+{
+       int i, k, p;
+
+       if (in6_ifaddrhashtbl != NULL) {
+               return;
+       }
+
+       PE_parse_boot_argn("ina6ddr_nhash", &in6addr_nhash,
+           sizeof(in6addr_nhash));
+       if (in6addr_nhash == 0) {
+               in6addr_nhash = IN6ADDR_NHASH;
+       }
+
+       MALLOC(in6_ifaddrhashtbl, struct in6_ifaddrhashhead *,
+           in6addr_nhash * sizeof(*in6_ifaddrhashtbl),
+           M_IFADDR, M_WAITOK | M_ZERO);
+       if (in6_ifaddrhashtbl == NULL) {
+               panic("in6_ifaddrhashtbl allocation failed");
+       }
+
+       /*
+        * Generate the next largest prime greater than in6addr_nhash.
+        */
+       k = (in6addr_nhash % 2 == 0) ? in6addr_nhash + 1 : in6addr_nhash + 2;
+       for (;;) {
+               p = 1;
+               for (i = 3; i * i <= k; i += 2) {
+                       if (k % i == 0) {
+                               p = 0;
+                       }
+               }
+               if (p == 1) {
+                       break;
+               }
+               k += 2;
+       }
+       in6addr_hashp = k;
+}
+
+static int
+sysctl_ip6_checkinterface SYSCTL_HANDLER_ARGS
+{
+#pragma unused(arg1, arg2)
+       int error, i;
+
+       i = ip6_checkinterface;
+       error = sysctl_handle_int(oidp, &i, 0, req);
+       if (error || req->newptr == USER_ADDR_NULL) {
+               return error;
+       }
+
+       switch (i) {
+       case IP6_CHECKINTERFACE_WEAK_ES:
+       case IP6_CHECKINTERFACE_HYBRID_ES:
+       case IP6_CHECKINTERFACE_STRONG_ES:
+               if (ip6_checkinterface != i) {
+                       ip6_checkinterface = i;
+                       os_log(OS_LOG_DEFAULT, "%s: ip6_checkinterface is now %d\n",
+                           __func__, ip6_checkinterface);
+               }
+               break;
+       default:
+               error = EINVAL;
+               break;
+       }
+       return error;
+}