]> git.saurik.com Git - apple/xnu.git/blobdiff - bsd/netinet/in_pcb.c
xnu-4903.221.2.tar.gz
[apple/xnu.git] / bsd / netinet / in_pcb.c
index 10efe37f0f35a21a4e241e5df73cacf827ec86ba..3d2e8c91df96fe15e3d041cccf3667849a03cdd0 100644 (file)
@@ -1,8 +1,8 @@
 /*
 /*
- * Copyright (c) 2000-2008 Apple Inc. All rights reserved.
+ * Copyright (c) 2000-2018 Apple Inc. All rights reserved.
  *
  * @APPLE_OSREFERENCE_LICENSE_HEADER_START@
  *
  * @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
  * 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
  * 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.
  * 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.
  * 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,
  * 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,
@@ -22,7 +22,7 @@
  * 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.
  * 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@
  */
 /*
  * @APPLE_OSREFERENCE_LICENSE_HEADER_END@
  */
 /*
 #include <sys/socket.h>
 #include <sys/socketvar.h>
 #include <sys/proc.h>
 #include <sys/socket.h>
 #include <sys/socketvar.h>
 #include <sys/proc.h>
-#ifndef __APPLE__
-#include <sys/jail.h>
-#endif
 #include <sys/kernel.h>
 #include <sys/sysctl.h>
 #include <sys/kernel.h>
 #include <sys/sysctl.h>
+#include <sys/mcache.h>
+#include <sys/kauth.h>
+#include <sys/priv.h>
+#include <sys/proc_uuid_policy.h>
+#include <sys/syslog.h>
+#include <sys/priv.h>
+#include <net/dlil.h>
+
 #include <libkern/OSAtomic.h>
 #include <libkern/OSAtomic.h>
+#include <kern/locks.h>
 
 #include <machine/limits.h>
 
 
 #include <machine/limits.h>
 
-#ifdef __APPLE__
 #include <kern/zalloc.h>
 #include <kern/zalloc.h>
-#endif
 
 #include <net/if.h>
 #include <net/if_types.h>
 #include <net/route.h>
 
 #include <net/if.h>
 #include <net/if_types.h>
 #include <net/route.h>
+#include <net/flowhash.h>
+#include <net/flowadv.h>
+#include <net/nat464_utils.h>
+#include <net/ntstat.h>
 
 #include <netinet/in.h>
 #include <netinet/in_pcb.h>
 
 #include <netinet/in.h>
 #include <netinet/in_pcb.h>
 #include <netinet6/ip6_var.h>
 #endif /* INET6 */
 
 #include <netinet6/ip6_var.h>
 #endif /* INET6 */
 
-#include "faith.h"
-
-#if IPSEC
-#include <netinet6/ipsec.h>
-#include <netkey/key.h>
-#endif /* IPSEC */
-
 #include <sys/kdebug.h>
 #include <sys/random.h>
 
 #include <sys/kdebug.h>
 #include <sys/random.h>
 
-#if IPSEC
-extern int ipsec_bypass;
+#include <dev/random/randomdev.h>
+#include <mach/boolean.h>
+
+#include <pexpert/pexpert.h>
+
+#if NECP
+#include <net/necp.h>
 #endif
 
 #endif
 
-#define DBG_FNC_PCB_LOOKUP     NETDBG_CODE(DBG_NETTCP, (6 << 8))
-#define DBG_FNC_PCB_HLOOKUP    NETDBG_CODE(DBG_NETTCP, ((6 << 8) | 1))
+#include <sys/stat.h>
+#include <sys/ubc.h>
+#include <sys/vnode.h>
+
+static lck_grp_t       *inpcb_lock_grp;
+static lck_attr_t      *inpcb_lock_attr;
+static lck_grp_attr_t  *inpcb_lock_grp_attr;
+decl_lck_mtx_data(static, inpcb_lock);         /* global INPCB lock */
+decl_lck_mtx_data(static, inpcb_timeout_lock);
+
+static TAILQ_HEAD(, inpcbinfo) inpcb_head = TAILQ_HEAD_INITIALIZER(inpcb_head);
 
 
-struct in_addr zeroin_addr;
+static u_int16_t inpcb_timeout_run = 0;        /* INPCB timer is scheduled to run */
+static boolean_t inpcb_garbage_collecting = FALSE; /* gc timer is scheduled */
+static boolean_t inpcb_ticking = FALSE;                /* "slow" timer is scheduled */
+static boolean_t inpcb_fast_timer_on = FALSE;
+
+#define        INPCB_GCREQ_THRESHOLD   50000
+
+static thread_call_t inpcb_thread_call, inpcb_fast_thread_call;
+static void inpcb_sched_timeout(void);
+static void inpcb_sched_lazy_timeout(void);
+static void _inpcb_sched_timeout(unsigned int);
+static void inpcb_timeout(void *, void *);
+const int inpcb_timeout_lazy = 10;     /* 10 seconds leeway for lazy timers */
+extern int tvtohz(struct timeval *);
+
+#if CONFIG_PROC_UUID_POLICY
+static void inp_update_cellular_policy(struct inpcb *, boolean_t);
+#if NECP
+static void inp_update_necp_want_app_policy(struct inpcb *, boolean_t);
+#endif /* NECP */
+#endif /* !CONFIG_PROC_UUID_POLICY */
+
+#define        DBG_FNC_PCB_LOOKUP      NETDBG_CODE(DBG_NETTCP, (6 << 8))
+#define        DBG_FNC_PCB_HLOOKUP     NETDBG_CODE(DBG_NETTCP, ((6 << 8) | 1))
 
 /*
  * These configure the range of local port addresses assigned to
 
 /*
  * These configure the range of local port addresses assigned to
@@ -121,17 +159,12 @@ struct    in_addr zeroin_addr;
  */
 int    ipport_lowfirstauto  = IPPORT_RESERVED - 1;     /* 1023 */
 int    ipport_lowlastauto = IPPORT_RESERVEDSTART;      /* 600 */
  */
 int    ipport_lowfirstauto  = IPPORT_RESERVED - 1;     /* 1023 */
 int    ipport_lowlastauto = IPPORT_RESERVEDSTART;      /* 600 */
-#ifndef __APPLE__
-int    ipport_firstauto = IPPORT_RESERVED;             /* 1024 */
-int    ipport_lastauto  = IPPORT_USERRESERVED;         /* 5000 */
-#else
-int    ipport_firstauto = IPPORT_HIFIRSTAUTO;          /* 49152 */
-int    ipport_lastauto  = IPPORT_HILASTAUTO;           /* 65535 */
-#endif
+int    ipport_firstauto = IPPORT_HIFIRSTAUTO;          /* 49152 */
+int    ipport_lastauto  = IPPORT_HILASTAUTO;           /* 65535 */
 int    ipport_hifirstauto = IPPORT_HIFIRSTAUTO;        /* 49152 */
 int    ipport_hilastauto  = IPPORT_HILASTAUTO;         /* 65535 */
 
 int    ipport_hifirstauto = IPPORT_HIFIRSTAUTO;        /* 49152 */
 int    ipport_hilastauto  = IPPORT_HILASTAUTO;         /* 65535 */
 
-#define RANGECHK(var, min, max) \
+#define        RANGECHK(var, min, max) \
        if ((var) < (min)) { (var) = (min); } \
        else if ((var) > (max)) { (var) = (max); }
 
        if ((var) < (min)) { (var) = (min); } \
        else if ((var) > (max)) { (var) = (max); }
 
@@ -139,8 +172,9 @@ static int
 sysctl_net_ipport_check SYSCTL_HANDLER_ARGS
 {
 #pragma unused(arg1, arg2)
 sysctl_net_ipport_check SYSCTL_HANDLER_ARGS
 {
 #pragma unused(arg1, arg2)
-       int error = sysctl_handle_int(oidp,
-               oidp->oid_arg1, oidp->oid_arg2, req);
+       int error;
+
+       error = sysctl_handle_int(oidp, oidp->oid_arg1, oidp->oid_arg2, req);
        if (!error) {
                RANGECHK(ipport_lowfirstauto, 1, IPPORT_RESERVED - 1);
                RANGECHK(ipport_lowlastauto, 1, IPPORT_RESERVED - 1);
        if (!error) {
                RANGECHK(ipport_lowfirstauto, 1, IPPORT_RESERVED - 1);
                RANGECHK(ipport_lowlastauto, 1, IPPORT_RESERVED - 1);
@@ -149,76 +183,384 @@ sysctl_net_ipport_check SYSCTL_HANDLER_ARGS
                RANGECHK(ipport_hifirstauto, IPPORT_RESERVED, USHRT_MAX);
                RANGECHK(ipport_hilastauto, IPPORT_RESERVED, USHRT_MAX);
        }
                RANGECHK(ipport_hifirstauto, IPPORT_RESERVED, USHRT_MAX);
                RANGECHK(ipport_hilastauto, IPPORT_RESERVED, USHRT_MAX);
        }
-       return error;
+       return (error);
 }
 
 #undef RANGECHK
 
 }
 
 #undef RANGECHK
 
-SYSCTL_NODE(_net_inet_ip, IPPROTO_IP, portrange, CTLFLAG_RW|CTLFLAG_LOCKED, 0, "IP Ports");
-
-SYSCTL_PROC(_net_inet_ip_portrange, OID_AUTO, lowfirst, CTLTYPE_INT|CTLFLAG_RW,
-          &ipport_lowfirstauto, 0, &sysctl_net_ipport_check, "I", "");
-SYSCTL_PROC(_net_inet_ip_portrange, OID_AUTO, lowlast, CTLTYPE_INT|CTLFLAG_RW,
-          &ipport_lowlastauto, 0, &sysctl_net_ipport_check, "I", "");
-SYSCTL_PROC(_net_inet_ip_portrange, OID_AUTO, first, CTLTYPE_INT|CTLFLAG_RW,
-          &ipport_firstauto, 0, &sysctl_net_ipport_check, "I", "");
-SYSCTL_PROC(_net_inet_ip_portrange, OID_AUTO, last, CTLTYPE_INT|CTLFLAG_RW,
-          &ipport_lastauto, 0, &sysctl_net_ipport_check, "I", "");
-SYSCTL_PROC(_net_inet_ip_portrange, OID_AUTO, hifirst, CTLTYPE_INT|CTLFLAG_RW,
-          &ipport_hifirstauto, 0, &sysctl_net_ipport_check, "I", "");
-SYSCTL_PROC(_net_inet_ip_portrange, OID_AUTO, hilast, CTLTYPE_INT|CTLFLAG_RW,
-          &ipport_hilastauto, 0, &sysctl_net_ipport_check, "I", "");
+SYSCTL_NODE(_net_inet_ip, IPPROTO_IP, portrange,
+       CTLFLAG_RW|CTLFLAG_LOCKED, 0, "IP Ports");
+
+SYSCTL_PROC(_net_inet_ip_portrange, OID_AUTO, lowfirst,
+       CTLTYPE_INT|CTLFLAG_RW | CTLFLAG_LOCKED,
+       &ipport_lowfirstauto, 0, &sysctl_net_ipport_check, "I", "");
+SYSCTL_PROC(_net_inet_ip_portrange, OID_AUTO, lowlast,
+       CTLTYPE_INT|CTLFLAG_RW | CTLFLAG_LOCKED,
+       &ipport_lowlastauto, 0, &sysctl_net_ipport_check, "I", "");
+SYSCTL_PROC(_net_inet_ip_portrange, OID_AUTO, first,
+       CTLTYPE_INT|CTLFLAG_RW | CTLFLAG_LOCKED,
+       &ipport_firstauto, 0, &sysctl_net_ipport_check, "I", "");
+SYSCTL_PROC(_net_inet_ip_portrange, OID_AUTO, last,
+       CTLTYPE_INT|CTLFLAG_RW | CTLFLAG_LOCKED,
+       &ipport_lastauto, 0, &sysctl_net_ipport_check, "I", "");
+SYSCTL_PROC(_net_inet_ip_portrange, OID_AUTO, hifirst,
+       CTLTYPE_INT|CTLFLAG_RW | CTLFLAG_LOCKED,
+       &ipport_hifirstauto, 0, &sysctl_net_ipport_check, "I", "");
+SYSCTL_PROC(_net_inet_ip_portrange, OID_AUTO, hilast,
+       CTLTYPE_INT|CTLFLAG_RW | CTLFLAG_LOCKED,
+       &ipport_hilastauto, 0, &sysctl_net_ipport_check, "I", "");
+
+static uint32_t apn_fallbk_debug = 0;
+#define apn_fallbk_log(x)       do { if (apn_fallbk_debug >= 1) log x; } while (0)
+
+#if CONFIG_EMBEDDED
+static boolean_t apn_fallbk_enabled = TRUE;
+
+SYSCTL_DECL(_net_inet);
+SYSCTL_NODE(_net_inet, OID_AUTO, apn_fallback, CTLFLAG_RW|CTLFLAG_LOCKED, 0, "APN Fallback");
+SYSCTL_UINT(_net_inet_apn_fallback, OID_AUTO, enable, CTLFLAG_RW | CTLFLAG_LOCKED,
+    &apn_fallbk_enabled, 0, "APN fallback enable");
+SYSCTL_UINT(_net_inet_apn_fallback, OID_AUTO, debug, CTLFLAG_RW | CTLFLAG_LOCKED,
+    &apn_fallbk_debug, 0, "APN fallback debug enable");
+#else
+static boolean_t apn_fallbk_enabled = FALSE;
+#endif
 
 extern int     udp_use_randomport;
 extern int     tcp_use_randomport;
 
 
 extern int     udp_use_randomport;
 extern int     tcp_use_randomport;
 
+/* Structs used for flowhash computation */
+struct inp_flowhash_key_addr {
+       union {
+               struct in_addr  v4;
+               struct in6_addr v6;
+               u_int8_t        addr8[16];
+               u_int16_t       addr16[8];
+               u_int32_t       addr32[4];
+       } infha;
+};
+
+struct inp_flowhash_key {
+       struct inp_flowhash_key_addr    infh_laddr;
+       struct inp_flowhash_key_addr    infh_faddr;
+       u_int32_t                       infh_lport;
+       u_int32_t                       infh_fport;
+       u_int32_t                       infh_af;
+       u_int32_t                       infh_proto;
+       u_int32_t                       infh_rand1;
+       u_int32_t                       infh_rand2;
+};
+
+static u_int32_t inp_hash_seed = 0;
+
+static int infc_cmp(const struct inpcb *, const struct inpcb *);
+
+/* Flags used by inp_fc_getinp */
+#define        INPFC_SOLOCKED  0x1
+#define        INPFC_REMOVE    0x2
+static struct inpcb *inp_fc_getinp(u_int32_t, u_int32_t);
+
+static void inp_fc_feedback(struct inpcb *);
+extern void tcp_remove_from_time_wait(struct inpcb *inp);
+
+decl_lck_mtx_data(static, inp_fc_lck);
+
+RB_HEAD(inp_fc_tree, inpcb) inp_fc_tree;
+RB_PROTOTYPE(inp_fc_tree, inpcb, infc_link, infc_cmp);
+RB_GENERATE(inp_fc_tree, inpcb, infc_link, infc_cmp);
+
+/*
+ * Use this inp as a key to find an inp in the flowhash tree.
+ * Accesses to it are protected by inp_fc_lck.
+ */
+struct inpcb key_inp;
+
 /*
  * in_pcb.c: manage the Protocol Control Blocks.
 /*
  * in_pcb.c: manage the Protocol Control Blocks.
- *
- * NOTE: It is assumed that most of these functions will be called at
- * splnet(). XXX - There are, unfortunately, a few exceptions to this
- * rule that should be fixed.
  */
 
  */
 
+void
+in_pcbinit(void)
+{
+       static int inpcb_initialized = 0;
+
+       VERIFY(!inpcb_initialized);
+       inpcb_initialized = 1;
+
+       inpcb_lock_grp_attr = lck_grp_attr_alloc_init();
+       inpcb_lock_grp = lck_grp_alloc_init("inpcb", inpcb_lock_grp_attr);
+       inpcb_lock_attr = lck_attr_alloc_init();
+       lck_mtx_init(&inpcb_lock, inpcb_lock_grp, inpcb_lock_attr);
+       lck_mtx_init(&inpcb_timeout_lock, inpcb_lock_grp, inpcb_lock_attr);
+       inpcb_thread_call = thread_call_allocate_with_priority(inpcb_timeout,
+           NULL, THREAD_CALL_PRIORITY_KERNEL);
+       inpcb_fast_thread_call = thread_call_allocate_with_priority(
+           inpcb_timeout, NULL, THREAD_CALL_PRIORITY_KERNEL);
+       if (inpcb_thread_call == NULL || inpcb_fast_thread_call == NULL)
+               panic("unable to alloc the inpcb thread call");
+
+       /*
+        * Initialize data structures required to deliver
+        * flow advisories.
+        */
+       lck_mtx_init(&inp_fc_lck, inpcb_lock_grp, inpcb_lock_attr);
+       lck_mtx_lock(&inp_fc_lck);
+       RB_INIT(&inp_fc_tree);
+       bzero(&key_inp, sizeof(key_inp));
+       lck_mtx_unlock(&inp_fc_lck);
+}
+
+#define        INPCB_HAVE_TIMER_REQ(req)       (((req).intimer_lazy > 0) || \
+       ((req).intimer_fast > 0) || ((req).intimer_nodelay > 0))
+static void
+inpcb_timeout(void *arg0, void *arg1)
+{
+#pragma unused(arg0, arg1)
+       struct inpcbinfo *ipi;
+       boolean_t t, gc;
+       struct intimercount gccnt, tmcnt;
+
+       /*
+        * Update coarse-grained networking timestamp (in sec.); the idea
+        * is to piggy-back on the timeout callout to update the counter
+        * returnable via net_uptime().
+        */
+       net_update_uptime();
+
+       bzero(&gccnt, sizeof(gccnt));
+       bzero(&tmcnt, sizeof(tmcnt));
+
+       lck_mtx_lock_spin(&inpcb_timeout_lock);
+       gc = inpcb_garbage_collecting;
+       inpcb_garbage_collecting = FALSE;
+
+       t = inpcb_ticking;
+       inpcb_ticking = FALSE;
+
+       if (gc || t) {
+               lck_mtx_unlock(&inpcb_timeout_lock);
+
+               lck_mtx_lock(&inpcb_lock);
+               TAILQ_FOREACH(ipi, &inpcb_head, ipi_entry) {
+                       if (INPCB_HAVE_TIMER_REQ(ipi->ipi_gc_req)) {
+                               bzero(&ipi->ipi_gc_req,
+                                       sizeof(ipi->ipi_gc_req));
+                               if (gc && ipi->ipi_gc != NULL) {
+                                       ipi->ipi_gc(ipi);
+                                       gccnt.intimer_lazy +=
+                                           ipi->ipi_gc_req.intimer_lazy;
+                                       gccnt.intimer_fast +=
+                                           ipi->ipi_gc_req.intimer_fast;
+                                       gccnt.intimer_nodelay +=
+                                           ipi->ipi_gc_req.intimer_nodelay;
+                               }
+                       }
+                       if (INPCB_HAVE_TIMER_REQ(ipi->ipi_timer_req)) {
+                               bzero(&ipi->ipi_timer_req,
+                                       sizeof(ipi->ipi_timer_req));
+                               if (t && ipi->ipi_timer != NULL) {
+                                       ipi->ipi_timer(ipi);
+                                       tmcnt.intimer_lazy +=
+                                           ipi->ipi_timer_req.intimer_lazy;
+                                       tmcnt.intimer_fast +=
+                                           ipi->ipi_timer_req.intimer_fast;
+                                       tmcnt.intimer_nodelay +=
+                                           ipi->ipi_timer_req.intimer_nodelay;
+                               }
+                       }
+               }
+               lck_mtx_unlock(&inpcb_lock);
+               lck_mtx_lock_spin(&inpcb_timeout_lock);
+       }
+
+       /* lock was dropped above, so check first before overriding */
+       if (!inpcb_garbage_collecting)
+               inpcb_garbage_collecting = INPCB_HAVE_TIMER_REQ(gccnt);
+       if (!inpcb_ticking)
+               inpcb_ticking = INPCB_HAVE_TIMER_REQ(tmcnt);
+
+       /* re-arm the timer if there's work to do */
+       inpcb_timeout_run--;
+       VERIFY(inpcb_timeout_run >= 0 && inpcb_timeout_run < 2);
+
+       if (gccnt.intimer_nodelay > 0 || tmcnt.intimer_nodelay > 0)
+               inpcb_sched_timeout();
+       else if ((gccnt.intimer_fast + tmcnt.intimer_fast) <= 5)
+               /* be lazy when idle with little activity */
+               inpcb_sched_lazy_timeout();
+       else
+               inpcb_sched_timeout();
+
+       lck_mtx_unlock(&inpcb_timeout_lock);
+}
+
+static void
+inpcb_sched_timeout(void)
+{
+       _inpcb_sched_timeout(0);
+}
+
+static void
+inpcb_sched_lazy_timeout(void)
+{
+       _inpcb_sched_timeout(inpcb_timeout_lazy);
+}
+
+static void
+_inpcb_sched_timeout(unsigned int offset)
+{
+       uint64_t deadline, leeway;
+
+       clock_interval_to_deadline(1, NSEC_PER_SEC, &deadline);
+       LCK_MTX_ASSERT(&inpcb_timeout_lock, LCK_MTX_ASSERT_OWNED);
+       if (inpcb_timeout_run == 0 &&
+           (inpcb_garbage_collecting || inpcb_ticking)) {
+               lck_mtx_convert_spin(&inpcb_timeout_lock);
+               inpcb_timeout_run++;
+               if (offset == 0) {
+                       inpcb_fast_timer_on = TRUE;
+                       thread_call_enter_delayed(inpcb_thread_call,
+                           deadline);
+               } else {
+                       inpcb_fast_timer_on = FALSE;
+                       clock_interval_to_absolutetime_interval(offset,
+                           NSEC_PER_SEC, &leeway);
+                       thread_call_enter_delayed_with_leeway(
+                           inpcb_thread_call, NULL, deadline, leeway,
+                           THREAD_CALL_DELAY_LEEWAY);
+               }
+       } else if (inpcb_timeout_run == 1 &&
+           offset == 0 && !inpcb_fast_timer_on) {
+               /*
+                * Since the request was for a fast timer but the
+                * scheduled timer is a lazy timer, try to schedule
+                * another instance of fast timer also.
+                */
+               lck_mtx_convert_spin(&inpcb_timeout_lock);
+               inpcb_timeout_run++;
+               inpcb_fast_timer_on = TRUE;
+               thread_call_enter_delayed(inpcb_fast_thread_call, deadline);
+       }
+}
+
+void
+inpcb_gc_sched(struct inpcbinfo *ipi, u_int32_t type)
+{
+       u_int32_t gccnt;
+
+       lck_mtx_lock_spin(&inpcb_timeout_lock);
+       inpcb_garbage_collecting = TRUE;
+       gccnt = ipi->ipi_gc_req.intimer_nodelay +
+               ipi->ipi_gc_req.intimer_fast;
+
+       if (gccnt > INPCB_GCREQ_THRESHOLD) {
+               type = INPCB_TIMER_FAST;
+       }
+
+       switch (type) {
+       case INPCB_TIMER_NODELAY:
+               atomic_add_32(&ipi->ipi_gc_req.intimer_nodelay, 1);
+               inpcb_sched_timeout();
+               break;
+       case INPCB_TIMER_FAST:
+               atomic_add_32(&ipi->ipi_gc_req.intimer_fast, 1);
+               inpcb_sched_timeout();
+               break;
+       default:
+               atomic_add_32(&ipi->ipi_gc_req.intimer_lazy, 1);
+               inpcb_sched_lazy_timeout();
+               break;
+       }
+       lck_mtx_unlock(&inpcb_timeout_lock);
+}
+
+void
+inpcb_timer_sched(struct inpcbinfo *ipi, u_int32_t type)
+{
+
+       lck_mtx_lock_spin(&inpcb_timeout_lock);
+       inpcb_ticking = TRUE;
+       switch (type) {
+       case INPCB_TIMER_NODELAY:
+               atomic_add_32(&ipi->ipi_timer_req.intimer_nodelay, 1);
+               inpcb_sched_timeout();
+               break;
+       case INPCB_TIMER_FAST:
+               atomic_add_32(&ipi->ipi_timer_req.intimer_fast, 1);
+               inpcb_sched_timeout();
+               break;
+       default:
+               atomic_add_32(&ipi->ipi_timer_req.intimer_lazy, 1);
+               inpcb_sched_lazy_timeout();
+               break;
+       }
+       lck_mtx_unlock(&inpcb_timeout_lock);
+}
+
+void
+in_pcbinfo_attach(struct inpcbinfo *ipi)
+{
+       struct inpcbinfo *ipi0;
+
+       lck_mtx_lock(&inpcb_lock);
+       TAILQ_FOREACH(ipi0, &inpcb_head, ipi_entry) {
+               if (ipi0 == ipi) {
+                       panic("%s: ipi %p already in the list\n",
+                           __func__, ipi);
+                       /* NOTREACHED */
+               }
+       }
+       TAILQ_INSERT_TAIL(&inpcb_head, ipi, ipi_entry);
+       lck_mtx_unlock(&inpcb_lock);
+}
+
+int
+in_pcbinfo_detach(struct inpcbinfo *ipi)
+{
+       struct inpcbinfo *ipi0;
+       int error = 0;
+
+       lck_mtx_lock(&inpcb_lock);
+       TAILQ_FOREACH(ipi0, &inpcb_head, ipi_entry) {
+               if (ipi0 == ipi)
+                       break;
+       }
+       if (ipi0 != NULL)
+               TAILQ_REMOVE(&inpcb_head, ipi0, ipi_entry);
+       else
+               error = ENXIO;
+       lck_mtx_unlock(&inpcb_lock);
+
+       return (error);
+}
+
 /*
  * Allocate a PCB and associate it with the socket.
  *
  * Returns:    0                       Success
  *             ENOBUFS
  *             ENOMEM
 /*
  * Allocate a PCB and associate it with the socket.
  *
  * Returns:    0                       Success
  *             ENOBUFS
  *             ENOMEM
- *     ipsec_init_policy:???           [IPSEC]
  */
 int
  */
 int
-in_pcballoc(struct socket *so, struct inpcbinfo *pcbinfo, __unused struct proc *p)
+in_pcballoc(struct socket *so, struct inpcbinfo *pcbinfo, struct proc *p)
 {
 {
+#pragma unused(p)
        struct inpcb *inp;
        struct inpcb *inp;
-       caddr_t               temp;
-#if IPSEC
-#ifndef __APPLE__
-       int error;
-#endif
-#endif
+       caddr_t temp;
 #if CONFIG_MACF_NET
        int mac_error;
 #if CONFIG_MACF_NET
        int mac_error;
-#endif
+#endif /* CONFIG_MACF_NET */
 
 
-       if (so->cached_in_sock_layer == 0) {
-#if TEMPDEBUG
-           printf("PCBALLOC calling zalloc for socket %x\n", so);
-#endif
-           inp = (struct inpcb *) zalloc(pcbinfo->ipi_zone);
-           if (inp == NULL)
-                return (ENOBUFS);
-           bzero((caddr_t)inp, sizeof(*inp));
-       }
-       else {
-#if TEMPDEBUG
-           printf("PCBALLOC reusing PCB for socket %x\n", so);
-#endif
-           inp = (struct inpcb *) so->so_saved_pcb;
-           temp = inp->inp_saved_ppcb;
-           bzero((caddr_t) inp, sizeof(*inp));
-           inp->inp_saved_ppcb = temp;
+       if ((so->so_flags1 & SOF1_CACHED_IN_SOCK_LAYER) == 0) {
+               inp = (struct inpcb *)zalloc(pcbinfo->ipi_zone);
+               if (inp == NULL)
+                       return (ENOBUFS);
+               bzero((caddr_t)inp, sizeof (*inp));
+       } else {
+               inp = (struct inpcb *)(void *)so->so_saved_pcb;
+               temp = inp->inp_saved_ppcb;
+               bzero((caddr_t)inp, sizeof (*inp));
+               inp->inp_saved_ppcb = temp;
        }
 
        inp->inp_gencnt = ++pcbinfo->ipi_gencnt;
        }
 
        inp->inp_gencnt = ++pcbinfo->ipi_gencnt;
@@ -227,103 +569,125 @@ in_pcballoc(struct socket *so, struct inpcbinfo *pcbinfo, __unused struct proc *
 #if CONFIG_MACF_NET
        mac_error = mac_inpcb_label_init(inp, M_WAITOK);
        if (mac_error != 0) {
 #if CONFIG_MACF_NET
        mac_error = mac_inpcb_label_init(inp, M_WAITOK);
        if (mac_error != 0) {
-               if (so->cached_in_sock_layer == 0)
+               if ((so->so_flags1 & SOF1_CACHED_IN_SOCK_LAYER) == 0)
                        zfree(pcbinfo->ipi_zone, inp);
                return (mac_error);
        }
        mac_inpcb_label_associate(so, inp);
                        zfree(pcbinfo->ipi_zone, inp);
                return (mac_error);
        }
        mac_inpcb_label_associate(so, inp);
-#endif
+#endif /* CONFIG_MACF_NET */
+       /* make sure inp_stat is always 64-bit aligned */
+       inp->inp_stat = (struct inp_stat *)P2ROUNDUP(inp->inp_stat_store,
+           sizeof (u_int64_t));
+       if (((uintptr_t)inp->inp_stat - (uintptr_t)inp->inp_stat_store) +
+           sizeof (*inp->inp_stat) > sizeof (inp->inp_stat_store)) {
+               panic("%s: insufficient space to align inp_stat", __func__);
+               /* NOTREACHED */
+       }
+
+       /* make sure inp_cstat is always 64-bit aligned */
+       inp->inp_cstat = (struct inp_stat *)P2ROUNDUP(inp->inp_cstat_store,
+           sizeof (u_int64_t));
+       if (((uintptr_t)inp->inp_cstat - (uintptr_t)inp->inp_cstat_store) +
+           sizeof (*inp->inp_cstat) > sizeof (inp->inp_cstat_store)) {
+               panic("%s: insufficient space to align inp_cstat", __func__);
+               /* NOTREACHED */
+       }
+
+       /* make sure inp_wstat is always 64-bit aligned */
+       inp->inp_wstat = (struct inp_stat *)P2ROUNDUP(inp->inp_wstat_store,
+           sizeof (u_int64_t));
+       if (((uintptr_t)inp->inp_wstat - (uintptr_t)inp->inp_wstat_store) +
+           sizeof (*inp->inp_wstat) > sizeof (inp->inp_wstat_store)) {
+               panic("%s: insufficient space to align inp_wstat", __func__);
+               /* NOTREACHED */
+       }
+
+       /* make sure inp_Wstat is always 64-bit aligned */
+       inp->inp_Wstat = (struct inp_stat *)P2ROUNDUP(inp->inp_Wstat_store,
+           sizeof (u_int64_t));
+       if (((uintptr_t)inp->inp_Wstat - (uintptr_t)inp->inp_Wstat_store) +
+           sizeof (*inp->inp_Wstat) > sizeof (inp->inp_Wstat_store)) {
+               panic("%s: insufficient space to align inp_Wstat", __func__);
+               /* NOTREACHED */
+       }
+
        so->so_pcb = (caddr_t)inp;
 
        if (so->so_proto->pr_flags & PR_PCBLOCK) {
        so->so_pcb = (caddr_t)inp;
 
        if (so->so_proto->pr_flags & PR_PCBLOCK) {
-               inp->inpcb_mtx = lck_mtx_alloc_init(pcbinfo->mtx_grp, pcbinfo->mtx_attr);
-               if (inp->inpcb_mtx == NULL) {
-                       printf("in_pcballoc: can't alloc mutex! so=%p\n", so);
-                       return(ENOMEM);
-               }
+               lck_mtx_init(&inp->inpcb_mtx, pcbinfo->ipi_lock_grp,
+                   pcbinfo->ipi_lock_attr);
        }
 
        }
 
-#if IPSEC
-#ifndef __APPLE__
-       if (ipsec_bypass == 0) {
-               error = ipsec_init_policy(so, &inp->inp_sp);
-               if (error != 0) {
-                       zfree(pcbinfo->ipi_zone, inp);
-                       return error;
-               }
-       }
-#endif
-#endif /*IPSEC*/
 #if INET6
 #if INET6
-       if (INP_SOCKAF(so) == AF_INET6 && !ip6_mapped_addr_on)
+       if (SOCK_DOM(so) == PF_INET6 && !ip6_mapped_addr_on)
                inp->inp_flags |= IN6P_IPV6_V6ONLY;
                inp->inp_flags |= IN6P_IPV6_V6ONLY;
-#endif
-       
-#if INET6
+
        if (ip6_auto_flowlabel)
                inp->inp_flags |= IN6P_AUTOFLOWLABEL;
        if (ip6_auto_flowlabel)
                inp->inp_flags |= IN6P_AUTOFLOWLABEL;
-#endif
-       lck_rw_lock_exclusive(pcbinfo->mtx);
+#endif /* INET6 */
+       if (intcoproc_unrestricted)
+               inp->inp_flags2 |= INP2_INTCOPROC_ALLOWED;
+
+       (void) inp_update_policy(inp);
+
+       lck_rw_lock_exclusive(pcbinfo->ipi_lock);
        inp->inp_gencnt = ++pcbinfo->ipi_gencnt;
        inp->inp_gencnt = ++pcbinfo->ipi_gencnt;
-       LIST_INSERT_HEAD(pcbinfo->listhead, inp, inp_list);
+       LIST_INSERT_HEAD(pcbinfo->ipi_listhead, inp, inp_list);
        pcbinfo->ipi_count++;
        pcbinfo->ipi_count++;
-       lck_rw_done(pcbinfo->mtx);
+       lck_rw_done(pcbinfo->ipi_lock);
        return (0);
 }
 
        return (0);
 }
 
-
 /*
 /*
-  in_pcblookup_local_and_cleanup does everything
-  in_pcblookup_local does but it checks for a socket
-  that's going away. Since we know that the lock is
-  held read+write when this funciton is called, we
-  can safely dispose of this socket like the slow
-  timer would usually do and return NULL. This is
-  great for bind.
-*/
-struct inpcb*
-in_pcblookup_local_and_cleanup(
-       struct inpcbinfo *pcbinfo,
-       struct in_addr laddr,
-       u_int lport_arg,
-       int wild_okay)
+ * in_pcblookup_local_and_cleanup does everything
+ * in_pcblookup_local does but it checks for a socket
+ * that's going away. Since we know that the lock is
+ * held read+write when this funciton is called, we
+ * can safely dispose of this socket like the slow
+ * timer would usually do and return NULL. This is
+ * great for bind.
+ */
+struct inpcb *
+in_pcblookup_local_and_cleanup(struct inpcbinfo *pcbinfo, struct in_addr laddr,
+    u_int lport_arg, int wild_okay)
 {
        struct inpcb *inp;
 {
        struct inpcb *inp;
-       
+
        /* Perform normal lookup */
        inp = in_pcblookup_local(pcbinfo, laddr, lport_arg, wild_okay);
        /* Perform normal lookup */
        inp = in_pcblookup_local(pcbinfo, laddr, lport_arg, wild_okay);
-       
+
        /* Check if we found a match but it's waiting to be disposed */
        /* Check if we found a match but it's waiting to be disposed */
-       if (inp && inp->inp_wantcnt == WNT_STOPUSING) {
+       if (inp != NULL && inp->inp_wantcnt == WNT_STOPUSING) {
                struct socket *so = inp->inp_socket;
                struct socket *so = inp->inp_socket;
-               
-               lck_mtx_lock(inp->inpcb_mtx);
-               
+
+               socket_lock(so, 0);
+
                if (so->so_usecount == 0) {
                        if (inp->inp_state != INPCB_STATE_DEAD)
                                in_pcbdetach(inp);
                if (so->so_usecount == 0) {
                        if (inp->inp_state != INPCB_STATE_DEAD)
                                in_pcbdetach(inp);
-                       in_pcbdispose(inp);
+                       in_pcbdispose(inp);     /* will unlock & destroy */
                        inp = NULL;
                        inp = NULL;
-               }
-               else {
-                       lck_mtx_unlock(inp->inpcb_mtx);
+               } else {
+                       socket_unlock(so, 0);
                }
        }
                }
        }
-       
-       return inp;
+
+       return (inp);
 }
 
 }
 
-#ifdef __APPLE_API_PRIVATE
 static void
 in_pcb_conflict_post_msg(u_int16_t port)
 {
 static void
 in_pcb_conflict_post_msg(u_int16_t port)
 {
-       /* 
-        * Radar 5523020 send a kernel event notification if a non-participating socket tries to bind
-        *               the port a socket who has set SOF_NOTIFYCONFLICT owns.
+       /*
+        * Radar 5523020 send a kernel event notification if a
+        * non-participating socket tries to bind the port a socket
+        * who has set SOF_NOTIFYCONFLICT owns.
         */
         */
-       struct kev_msg        ev_msg;
+       struct kev_msg ev_msg;
        struct kev_in_portinuse in_portinuse;
 
        struct kev_in_portinuse in_portinuse;
 
+       bzero(&in_portinuse, sizeof (struct kev_in_portinuse));
+       bzero(&ev_msg, sizeof (struct kev_msg));
        in_portinuse.port = ntohs(port);        /* port in host order */
        in_portinuse.req_pid = proc_selfpid();
        ev_msg.vendor_code = KEV_VENDOR_APPLE;
        in_portinuse.port = ntohs(port);        /* port in host order */
        in_portinuse.req_pid = proc_selfpid();
        ev_msg.vendor_code = KEV_VENDOR_APPLE;
@@ -331,12 +695,15 @@ in_pcb_conflict_post_msg(u_int16_t port)
        ev_msg.kev_subclass = KEV_INET_SUBCLASS;
        ev_msg.event_code = KEV_INET_PORTINUSE;
        ev_msg.dv[0].data_ptr = &in_portinuse;
        ev_msg.kev_subclass = KEV_INET_SUBCLASS;
        ev_msg.event_code = KEV_INET_PORTINUSE;
        ev_msg.dv[0].data_ptr = &in_portinuse;
-       ev_msg.dv[0].data_length      = sizeof(struct kev_in_portinuse);
+       ev_msg.dv[0].data_length = sizeof (struct kev_in_portinuse);
        ev_msg.dv[1].data_length = 0;
        ev_msg.dv[1].data_length = 0;
-       kev_post_msg(&ev_msg);
+       dlil_post_complete_msg(NULL, &ev_msg);
 }
 }
-#endif
+
 /*
 /*
+ * Bind an INPCB to an address and/or port.  This routine should not alter
+ * the caller-supplied local address "nam".
+ *
  * Returns:    0                       Success
  *             EADDRNOTAVAIL           Address not available.
  *             EINVAL                  Invalid argument
  * Returns:    0                       Success
  *             EADDRNOTAVAIL           Address not available.
  *             EINVAL                  Invalid argument
@@ -344,47 +711,58 @@ in_pcb_conflict_post_msg(u_int16_t port)
  *             EACCES                  Permission denied
  *             EADDRINUSE              Address in use
  *             EAGAIN                  Resource unavailable, try again
  *             EACCES                  Permission denied
  *             EADDRINUSE              Address in use
  *             EAGAIN                  Resource unavailable, try again
- *             proc_suser:EPERM        Operation not permitted
+ *             priv_check_cred:EPERM   Operation not permitted
  */
 int
 in_pcbbind(struct inpcb *inp, struct sockaddr *nam, struct proc *p)
 {
        struct socket *so = inp->inp_socket;
        unsigned short *lastport;
  */
 int
 in_pcbbind(struct inpcb *inp, struct sockaddr *nam, struct proc *p)
 {
        struct socket *so = inp->inp_socket;
        unsigned short *lastport;
-       struct sockaddr_in *sin;
        struct inpcbinfo *pcbinfo = inp->inp_pcbinfo;
        u_short lport = 0, rand_port = 0;
        int wild = 0, reuseport = (so->so_options & SO_REUSEPORT);
        int error, randomport, conflict = 0;
        struct inpcbinfo *pcbinfo = inp->inp_pcbinfo;
        u_short lport = 0, rand_port = 0;
        int wild = 0, reuseport = (so->so_options & SO_REUSEPORT);
        int error, randomport, conflict = 0;
+       boolean_t anonport = FALSE;
+       kauth_cred_t cred;
+       struct in_addr laddr;
+       struct ifnet *outif = NULL;
 
        if (TAILQ_EMPTY(&in_ifaddrhead)) /* XXX broken! */
                return (EADDRNOTAVAIL);
 
        if (TAILQ_EMPTY(&in_ifaddrhead)) /* XXX broken! */
                return (EADDRNOTAVAIL);
-       if (inp->inp_lport || inp->inp_laddr.s_addr != INADDR_ANY)
-               return (EINVAL);
-       if ((so->so_options & (SO_REUSEADDR|SO_REUSEPORT)) == 0)
+       if (!(so->so_options & (SO_REUSEADDR|SO_REUSEPORT)))
                wild = 1;
                wild = 1;
+
+       bzero(&laddr, sizeof(laddr));
+
        socket_unlock(so, 0); /* keep reference on socket */
        socket_unlock(so, 0); /* keep reference on socket */
-       lck_rw_lock_exclusive(pcbinfo->mtx);
-       if (nam) {
-               sin = (struct sockaddr_in *)nam;
-               if (nam->sa_len != sizeof (*sin)) {
-                       lck_rw_done(pcbinfo->mtx);
+       lck_rw_lock_exclusive(pcbinfo->ipi_lock);
+       if (inp->inp_lport != 0 || inp->inp_laddr.s_addr != INADDR_ANY) {
+               /* another thread completed the bind */
+               lck_rw_done(pcbinfo->ipi_lock);
+               socket_lock(so, 0);
+               return (EINVAL);
+       }
+
+       if (nam != NULL) {
+               if (nam->sa_len != sizeof (struct sockaddr_in)) {
+                       lck_rw_done(pcbinfo->ipi_lock);
                        socket_lock(so, 0);
                        return (EINVAL);
                }
                        socket_lock(so, 0);
                        return (EINVAL);
                }
-#ifdef notdef
+#if 0
                /*
                 * We should check the family, but old programs
                 * incorrectly fail to initialize it.
                 */
                /*
                 * We should check the family, but old programs
                 * incorrectly fail to initialize it.
                 */
-               if (sin->sin_family != AF_INET) {
-                       lck_rw_done(pcbinfo->mtx);
+               if (nam->sa_family != AF_INET) {
+                       lck_rw_done(pcbinfo->ipi_lock);
                        socket_lock(so, 0);
                        return (EAFNOSUPPORT);
                }
                        socket_lock(so, 0);
                        return (EAFNOSUPPORT);
                }
-#endif
-               lport = sin->sin_port;
-               if (IN_MULTICAST(ntohl(sin->sin_addr.s_addr))) {
+#endif /* 0 */
+               lport = SIN(nam)->sin_port;
+
+               if (IN_MULTICAST(ntohl(SIN(nam)->sin_addr.s_addr))) {
                        /*
                         * Treat SO_REUSEADDR as SO_REUSEPORT for multicast;
                         * allow complete duplication of binding if
                        /*
                         * Treat SO_REUSEADDR as SO_REUSEPORT for multicast;
                         * allow complete duplication of binding if
@@ -394,131 +772,147 @@ in_pcbbind(struct inpcb *inp, struct sockaddr *nam, struct proc *p)
                         */
                        if (so->so_options & SO_REUSEADDR)
                                reuseport = SO_REUSEADDR|SO_REUSEPORT;
                         */
                        if (so->so_options & SO_REUSEADDR)
                                reuseport = SO_REUSEADDR|SO_REUSEPORT;
-               } else if (sin->sin_addr.s_addr != INADDR_ANY) {
+               } else if (SIN(nam)->sin_addr.s_addr != INADDR_ANY) {
+                       struct sockaddr_in sin;
                        struct ifaddr *ifa;
                        struct ifaddr *ifa;
-                       sin->sin_port = 0;              /* yech... */
-                       if ((ifa = ifa_ifwithaddr((struct sockaddr *)sin)) == 0) {
-                               lck_rw_done(pcbinfo->mtx);
+
+                       /* Sanitized for interface address searches */
+                       bzero(&sin, sizeof (sin));
+                       sin.sin_family = AF_INET;
+                       sin.sin_len = sizeof (struct sockaddr_in);
+                       sin.sin_addr.s_addr = SIN(nam)->sin_addr.s_addr;
+
+                       ifa = ifa_ifwithaddr(SA(&sin));
+                       if (ifa == NULL) {
+                               lck_rw_done(pcbinfo->ipi_lock);
                                socket_lock(so, 0);
                                return (EADDRNOTAVAIL);
                                socket_lock(so, 0);
                                return (EADDRNOTAVAIL);
-                       }
-                       else {
-                               ifafree(ifa);
+                       } else {
+                               /*
+                                * Opportunistically determine the outbound
+                                * interface that may be used; this may not
+                                * hold true if we end up using a route
+                                * going over a different interface, e.g.
+                                * when sending to a local address.  This
+                                * will get updated again after sending.
+                                */
+                               IFA_LOCK(ifa);
+                               outif = ifa->ifa_ifp;
+                               IFA_UNLOCK(ifa);
+                               IFA_REMREF(ifa);
                        }
                }
                        }
                }
-               if (lport) {
+               if (lport != 0) {
                        struct inpcb *t;
                        struct inpcb *t;
+                       uid_t u;
 
 
-                       /* GROSS */
 #if !CONFIG_EMBEDDED
 #if !CONFIG_EMBEDDED
-                       if (ntohs(lport) < IPPORT_RESERVED && proc_suser(p)) {
-                               lck_rw_done(pcbinfo->mtx);
-                               socket_lock(so, 0);
-                               return (EACCES);
+                       if (ntohs(lport) < IPPORT_RESERVED &&
+                               SIN(nam)->sin_addr.s_addr != 0) {
+                               cred = kauth_cred_proc_ref(p);
+                               error = priv_check_cred(cred,
+                                   PRIV_NETINET_RESERVEDPORT, 0);
+                               kauth_cred_unref(&cred);
+                               if (error != 0) {
+                                       lck_rw_done(pcbinfo->ipi_lock);
+                                       socket_lock(so, 0);
+                                       return (EACCES);
+                               }
                        }
                        }
-#endif
-                       if (so->so_uid &&
-                           !IN_MULTICAST(ntohl(sin->sin_addr.s_addr))) {
-                               t = in_pcblookup_local_and_cleanup(inp->inp_pcbinfo,
-                                   sin->sin_addr, lport, INPLOOKUP_WILDCARD);
-                               if (t &&
-                                   (ntohl(sin->sin_addr.s_addr) != INADDR_ANY ||
-                                    ntohl(t->inp_laddr.s_addr) != INADDR_ANY ||
-                                    (t->inp_socket->so_options &
-                                        SO_REUSEPORT) == 0) &&
-                                    (so->so_uid != t->inp_socket->so_uid) &&
-                                     ((t->inp_socket->so_flags & SOF_REUSESHAREUID) == 0)) {
-#if INET6
-                                       if (ntohl(sin->sin_addr.s_addr) !=
-                                           INADDR_ANY ||
-                                           ntohl(t->inp_laddr.s_addr) !=
-                                           INADDR_ANY ||
-                                           INP_SOCKAF(so) ==
-                                           INP_SOCKAF(t->inp_socket))
-#endif /* INET6 */
-                                       {
-#ifdef __APPLE_API_PRIVATE
-
-                                               if ((t->inp_socket->so_flags & SOF_NOTIFYCONFLICT) && ((so->so_flags & SOF_NOTIFYCONFLICT) == 0)) 
-                                                       conflict = 1;
-
-                                               lck_rw_done(pcbinfo->mtx);
-
-                                               if (conflict)
-                                                       in_pcb_conflict_post_msg(lport);
-#else
-                                               lck_rw_done(pcbinfo->mtx);
-#endif /* __APPLE_API_PRIVATE */
+#endif /* !CONFIG_EMBEDDED */
+                       if (!IN_MULTICAST(ntohl(SIN(nam)->sin_addr.s_addr)) &&
+                           (u = kauth_cred_getuid(so->so_cred)) != 0 &&
+                           (t = in_pcblookup_local_and_cleanup(
+                           inp->inp_pcbinfo, SIN(nam)->sin_addr, lport,
+                           INPLOOKUP_WILDCARD)) != NULL &&
+                           (SIN(nam)->sin_addr.s_addr != INADDR_ANY ||
+                           t->inp_laddr.s_addr != INADDR_ANY ||
+                           !(t->inp_socket->so_options & SO_REUSEPORT)) &&
+                           (u != kauth_cred_getuid(t->inp_socket->so_cred)) &&
+                           !(t->inp_socket->so_flags & SOF_REUSESHAREUID) &&
+                           (SIN(nam)->sin_addr.s_addr != INADDR_ANY ||
+                           t->inp_laddr.s_addr != INADDR_ANY)) {
+                               if ((t->inp_socket->so_flags &
+                                   SOF_NOTIFYCONFLICT) &&
+                                   !(so->so_flags & SOF_NOTIFYCONFLICT))
+                                       conflict = 1;
+
+                               lck_rw_done(pcbinfo->ipi_lock);
+
+                               if (conflict)
+                                       in_pcb_conflict_post_msg(lport);
 
 
-                                               socket_lock(so, 0);
-                                               return (EADDRINUSE);
-                                       }
-                               }
+                               socket_lock(so, 0);
+                               return (EADDRINUSE);
                        }
                        }
-                       t = in_pcblookup_local_and_cleanup(pcbinfo, sin->sin_addr,
-                           lport, wild);
-                       if (t &&
+                       t = in_pcblookup_local_and_cleanup(pcbinfo,
+                           SIN(nam)->sin_addr, lport, wild);
+                       if (t != NULL &&
                            (reuseport & t->inp_socket->so_options) == 0) {
 #if INET6
                            (reuseport & t->inp_socket->so_options) == 0) {
 #if INET6
-                               if (ip6_mapped_addr_on == 0 ||
-                                   ntohl(sin->sin_addr.s_addr) !=
-                                   INADDR_ANY ||
-                                   ntohl(t->inp_laddr.s_addr) !=
-                                   INADDR_ANY ||
-                                   INP_SOCKAF(so) ==
-                                   INP_SOCKAF(t->inp_socket))
+                               if (SIN(nam)->sin_addr.s_addr != INADDR_ANY ||
+                                   t->inp_laddr.s_addr != INADDR_ANY ||
+                                   SOCK_DOM(so) != PF_INET6 ||
+                                   SOCK_DOM(t->inp_socket) != PF_INET6)
 #endif /* INET6 */
                                {
 #endif /* INET6 */
                                {
-#ifdef __APPLE_API_PRIVATE
 
 
-                                       if ((t->inp_socket->so_flags & SOF_NOTIFYCONFLICT) && ((so->so_flags & SOF_NOTIFYCONFLICT) == 0)) 
+                                       if ((t->inp_socket->so_flags &
+                                           SOF_NOTIFYCONFLICT) &&
+                                           !(so->so_flags & SOF_NOTIFYCONFLICT))
                                                conflict = 1;
 
                                                conflict = 1;
 
-                                       lck_rw_done(pcbinfo->mtx);
+                                       lck_rw_done(pcbinfo->ipi_lock);
 
                                        if (conflict)
                                                in_pcb_conflict_post_msg(lport);
 
                                        if (conflict)
                                                in_pcb_conflict_post_msg(lport);
-#else
-                                       lck_rw_done(pcbinfo->mtx);
-#endif /* __APPLE_API_PRIVATE */
                                        socket_lock(so, 0);
                                        return (EADDRINUSE);
                                }
                        }
                }
                                        socket_lock(so, 0);
                                        return (EADDRINUSE);
                                }
                        }
                }
-               inp->inp_laddr = sin->sin_addr;
+               laddr = SIN(nam)->sin_addr;
        }
        if (lport == 0) {
                u_short first, last;
                int count;
        }
        if (lport == 0) {
                u_short first, last;
                int count;
+               bool found;
 
 
-               randomport = (so->so_flags & SOF_BINDRANDOMPORT) || 
-                       (so->so_type == SOCK_STREAM ? tcp_use_randomport : udp_use_randomport);
-
-               inp->inp_flags |= INP_ANONPORT;
+               randomport = (so->so_flags & SOF_BINDRANDOMPORT) ||
+                   (so->so_type == SOCK_STREAM ? tcp_use_randomport :
+                   udp_use_randomport);
 
 
+               /*
+                * Even though this looks similar to the code in
+                * in6_pcbsetport, the v6 vs v4 checks are different.
+                */
+               anonport = TRUE;
                if (inp->inp_flags & INP_HIGHPORT) {
                        first = ipport_hifirstauto;     /* sysctl */
                        last  = ipport_hilastauto;
                if (inp->inp_flags & INP_HIGHPORT) {
                        first = ipport_hifirstauto;     /* sysctl */
                        last  = ipport_hilastauto;
-                       lastport = &pcbinfo->lasthi;
+                       lastport = &pcbinfo->ipi_lasthi;
                } else if (inp->inp_flags & INP_LOWPORT) {
                } else if (inp->inp_flags & INP_LOWPORT) {
-                       if ((error = proc_suser(p)) != 0) {
-                               lck_rw_done(pcbinfo->mtx);
+                       cred = kauth_cred_proc_ref(p);
+                       error = priv_check_cred(cred,
+                           PRIV_NETINET_RESERVEDPORT, 0);
+                       kauth_cred_unref(&cred);
+                       if (error != 0) {
+                               lck_rw_done(pcbinfo->ipi_lock);
                                socket_lock(so, 0);
                                socket_lock(so, 0);
-                               return error;
+                               return (error);
                        }
                        first = ipport_lowfirstauto;    /* 1023 */
                        last  = ipport_lowlastauto;     /* 600 */
                        }
                        first = ipport_lowfirstauto;    /* 1023 */
                        last  = ipport_lowlastauto;     /* 600 */
-                       lastport = &pcbinfo->lastlow;
+                       lastport = &pcbinfo->ipi_lastlow;
                } else {
                        first = ipport_firstauto;       /* sysctl */
                        last  = ipport_lastauto;
                } else {
                        first = ipport_firstauto;       /* sysctl */
                        last  = ipport_lastauto;
-                       lastport = &pcbinfo->lastport;
+                       lastport = &pcbinfo->ipi_lastport;
                }
                /* No point in randomizing if only one port is available */
 
                if (first == last)
                }
                /* No point in randomizing if only one port is available */
 
                if (first == last)
-                       randomport = 0; 
+                       randomport = 0;
                /*
                 * Simple check to ensure all ports are not used up causing
                 * a deadlock here.
                /*
                 * Simple check to ensure all ports are not used up causing
                 * a deadlock here.
@@ -527,76 +921,284 @@ in_pcbbind(struct inpcb *inp, struct sockaddr *nam, struct proc *p)
                 * is not being tested on each round of the loop.
                 */
                if (first > last) {
                 * is not being tested on each round of the loop.
                 */
                if (first > last) {
+                       struct in_addr lookup_addr;
+
                        /*
                         * counting down
                         */
                        if (randomport) {
                        /*
                         * counting down
                         */
                        if (randomport) {
-                               read_random(&rand_port, sizeof(rand_port));
-                               *lastport = first - (rand_port % (first - last));
+                               read_frandom(&rand_port, sizeof (rand_port));
+                               *lastport =
+                                   first - (rand_port % (first - last));
                        }
                        count = first - last;
 
                        }
                        count = first - last;
 
+                       lookup_addr = (laddr.s_addr != INADDR_ANY) ? laddr :
+                           inp->inp_laddr;
+
+                       found = false;
                        do {
                                if (count-- < 0) {      /* completely used? */
                        do {
                                if (count-- < 0) {      /* completely used? */
-                                       lck_rw_done(pcbinfo->mtx);
+                                       lck_rw_done(pcbinfo->ipi_lock);
                                        socket_lock(so, 0);
                                        socket_lock(so, 0);
-                                       inp->inp_laddr.s_addr = INADDR_ANY;
                                        return (EADDRNOTAVAIL);
                                }
                                --*lastport;
                                if (*lastport > first || *lastport < last)
                                        *lastport = first;
                                lport = htons(*lastport);
                                        return (EADDRNOTAVAIL);
                                }
                                --*lastport;
                                if (*lastport > first || *lastport < last)
                                        *lastport = first;
                                lport = htons(*lastport);
-                       } while (in_pcblookup_local_and_cleanup(pcbinfo,
-                                inp->inp_laddr, lport, wild));
+
+                               found = in_pcblookup_local_and_cleanup(pcbinfo,
+                                   lookup_addr, lport, wild) == NULL;
+                       } while (!found);
                } else {
                } else {
+                       struct in_addr lookup_addr;
+
                        /*
                         * counting up
                         */
                        if (randomport) {
                        /*
                         * counting up
                         */
                        if (randomport) {
-                               read_random(&rand_port, sizeof(rand_port));
-                               *lastport = first + (rand_port % (first - last));
+                               read_frandom(&rand_port, sizeof (rand_port));
+                               *lastport =
+                                   first + (rand_port % (first - last));
                        }
                        count = last - first;
 
                        }
                        count = last - first;
 
+                       lookup_addr = (laddr.s_addr != INADDR_ANY) ? laddr :
+                           inp->inp_laddr;
+
+                       found = false;
                        do {
                                if (count-- < 0) {      /* completely used? */
                        do {
                                if (count-- < 0) {      /* completely used? */
-                                       lck_rw_done(pcbinfo->mtx);
+                                       lck_rw_done(pcbinfo->ipi_lock);
                                        socket_lock(so, 0);
                                        socket_lock(so, 0);
-                                       inp->inp_laddr.s_addr = INADDR_ANY;
                                        return (EADDRNOTAVAIL);
                                }
                                ++*lastport;
                                if (*lastport < first || *lastport > last)
                                        *lastport = first;
                                lport = htons(*lastport);
                                        return (EADDRNOTAVAIL);
                                }
                                ++*lastport;
                                if (*lastport < first || *lastport > last)
                                        *lastport = first;
                                lport = htons(*lastport);
-                       } while (in_pcblookup_local_and_cleanup(pcbinfo,
-                                inp->inp_laddr, lport, wild));
+
+                               found = in_pcblookup_local_and_cleanup(pcbinfo,
+                                   lookup_addr, lport, wild) == NULL;
+                       } while (!found);
                }
        }
        socket_lock(so, 0);
                }
        }
        socket_lock(so, 0);
+
+       /*
+        * We unlocked socket's protocol lock for a long time.
+        * The socket might have been dropped/defuncted.
+        * Checking if world has changed since.
+        */
+       if (inp->inp_state == INPCB_STATE_DEAD) {
+               lck_rw_done(pcbinfo->ipi_lock);
+               return (ECONNABORTED);
+       }
+
+       if (inp->inp_lport != 0 || inp->inp_laddr.s_addr != INADDR_ANY) {
+               lck_rw_done(pcbinfo->ipi_lock);
+               return (EINVAL);
+       }
+
+       if (laddr.s_addr != INADDR_ANY) {
+               inp->inp_laddr = laddr;
+               inp->inp_last_outifp = outif;
+       }
        inp->inp_lport = lport;
        inp->inp_lport = lport;
+       if (anonport)
+               inp->inp_flags |= INP_ANONPORT;
+
        if (in_pcbinshash(inp, 1) != 0) {
                inp->inp_laddr.s_addr = INADDR_ANY;
        if (in_pcbinshash(inp, 1) != 0) {
                inp->inp_laddr.s_addr = INADDR_ANY;
+               inp->inp_last_outifp = NULL;
+
                inp->inp_lport = 0;
                inp->inp_lport = 0;
-               lck_rw_done(pcbinfo->mtx);
+               if (anonport)
+                       inp->inp_flags &= ~INP_ANONPORT;
+               lck_rw_done(pcbinfo->ipi_lock);
                return (EAGAIN);
        }
                return (EAGAIN);
        }
-       lck_rw_done(pcbinfo->mtx);
+       lck_rw_done(pcbinfo->ipi_lock);
        sflt_notify(so, sock_evt_bound, NULL);
        return (0);
 }
 
        sflt_notify(so, sock_evt_bound, NULL);
        return (0);
 }
 
+#define        APN_FALLBACK_IP_FILTER(a)       \
+       (IN_LINKLOCAL(ntohl((a)->sin_addr.s_addr)) || \
+        IN_LOOPBACK(ntohl((a)->sin_addr.s_addr)) || \
+        IN_ZERONET(ntohl((a)->sin_addr.s_addr)) || \
+        IN_MULTICAST(ntohl((a)->sin_addr.s_addr)) || \
+        IN_PRIVATE(ntohl((a)->sin_addr.s_addr)))
+
+#define        APN_FALLBACK_NOTIF_INTERVAL     2 /* Magic Number */
+static uint64_t last_apn_fallback = 0;
+
+static boolean_t
+apn_fallback_required (proc_t proc, struct socket *so, struct sockaddr_in *p_dstv4)
+{
+       uint64_t timenow;
+       struct sockaddr_storage lookup_default_addr;
+       struct rtentry *rt = NULL;
+
+       VERIFY(proc != NULL);
+
+       if (apn_fallbk_enabled  == FALSE)
+               return FALSE;
+
+       if (proc == kernproc)
+               return FALSE;
+
+       if (so && (so->so_options & SO_NOAPNFALLBK))
+               return FALSE;
+
+       timenow = net_uptime();
+       if ((timenow - last_apn_fallback) < APN_FALLBACK_NOTIF_INTERVAL) {
+               apn_fallbk_log((LOG_INFO, "APN fallback notification throttled.\n"));
+               return FALSE;
+       }
+
+       if (p_dstv4 && APN_FALLBACK_IP_FILTER(p_dstv4))
+               return FALSE;
+
+       /* Check if we have unscoped IPv6 default route through cellular */
+       bzero(&lookup_default_addr, sizeof(lookup_default_addr));
+       lookup_default_addr.ss_family = AF_INET6;
+       lookup_default_addr.ss_len = sizeof(struct sockaddr_in6);
+
+       rt = rtalloc1((struct sockaddr *)&lookup_default_addr, 0, 0);
+       if (NULL == rt) {
+               apn_fallbk_log((LOG_INFO, "APN fallback notification could not find "
+                   "unscoped default IPv6 route.\n"));
+               return FALSE;
+       }
+
+       if (!IFNET_IS_CELLULAR(rt->rt_ifp)) {
+               rtfree(rt);
+               apn_fallbk_log((LOG_INFO, "APN fallback notification could not find "
+                   "unscoped default IPv6 route through cellular interface.\n"));
+               return FALSE;
+       }
+
+       /*
+        * We have a default IPv6 route, ensure that
+        * we do not have IPv4 default route before triggering
+        * the event
+        */
+       rtfree(rt);
+       rt = NULL;
+
+       bzero(&lookup_default_addr, sizeof(lookup_default_addr));
+       lookup_default_addr.ss_family = AF_INET;
+       lookup_default_addr.ss_len = sizeof(struct sockaddr_in);
+
+       rt = rtalloc1((struct sockaddr *)&lookup_default_addr, 0, 0);
+
+       if (rt) {
+               rtfree(rt);
+               rt = NULL;
+               apn_fallbk_log((LOG_INFO, "APN fallback notification found unscoped "
+                   "IPv4 default route!\n"));
+               return FALSE;
+       }
+
+       {
+               /*
+                * We disable APN fallback if the binary is not a third-party app.
+                * Note that platform daemons use their process name as a
+                * bundle ID so we filter out bundle IDs without dots.
+                */
+               const char *bundle_id = cs_identity_get(proc);
+               if (bundle_id == NULL ||
+                   bundle_id[0] == '\0' ||
+                   strchr(bundle_id, '.') == NULL ||
+                   strncmp(bundle_id, "com.apple.", sizeof("com.apple.") - 1) == 0) {
+                       apn_fallbk_log((LOG_INFO, "Abort: APN fallback notification found first-"
+                           "party bundle ID \"%s\"!\n", (bundle_id ? bundle_id : "NULL")));
+                       return FALSE;
+               }
+       }
+
+       {
+               /*
+                * The Apple App Store IPv6 requirement started on
+                * June 1st, 2016 at 12:00:00 AM PDT.
+                * We disable APN fallback if the binary is more recent than that.
+                * We check both atime and birthtime since birthtime is not always supported.
+                */
+               static const long ipv6_start_date = 1464764400L;
+               vfs_context_t context;
+               struct stat64 sb;
+               int vn_stat_error;
+
+               bzero(&sb, sizeof(struct stat64));
+               context = vfs_context_create(NULL);
+               vn_stat_error = vn_stat(proc->p_textvp, &sb, NULL, 1, context);
+               (void)vfs_context_rele(context);
+
+               if (vn_stat_error != 0 ||
+                   sb.st_atimespec.tv_sec >= ipv6_start_date ||
+                   sb.st_birthtimespec.tv_sec >= ipv6_start_date) {
+                       apn_fallbk_log((LOG_INFO, "Abort: APN fallback notification found binary "
+                           "too recent! (err %d atime %ld mtime %ld ctime %ld birthtime %ld)\n",
+                           vn_stat_error, sb.st_atimespec.tv_sec, sb.st_mtimespec.tv_sec,
+                           sb.st_ctimespec.tv_sec, sb.st_birthtimespec.tv_sec));
+                       return FALSE;
+               }
+       }
+       return TRUE;
+}
+
+static void
+apn_fallback_trigger(proc_t proc, struct socket *so)
+{
+       pid_t pid = 0;
+       struct kev_msg ev_msg;
+       struct kev_netevent_apnfallbk_data apnfallbk_data;
+
+       last_apn_fallback = net_uptime();
+       pid = proc_pid(proc);
+       uuid_t application_uuid;
+       uuid_clear(application_uuid);
+       proc_getexecutableuuid(proc, application_uuid,
+           sizeof(application_uuid));
+
+       bzero(&ev_msg, sizeof (struct kev_msg));
+       ev_msg.vendor_code      = KEV_VENDOR_APPLE;
+       ev_msg.kev_class        = KEV_NETWORK_CLASS;
+       ev_msg.kev_subclass     = KEV_NETEVENT_SUBCLASS;
+       ev_msg.event_code       = KEV_NETEVENT_APNFALLBACK;
+
+       bzero(&apnfallbk_data, sizeof(apnfallbk_data));
+
+       if (so->so_flags & SOF_DELEGATED) {
+               apnfallbk_data.epid = so->e_pid;
+               uuid_copy(apnfallbk_data.euuid, so->e_uuid);
+       } else {
+               apnfallbk_data.epid = so->last_pid;
+               uuid_copy(apnfallbk_data.euuid, so->last_uuid);
+       }
+
+       ev_msg.dv[0].data_ptr   = &apnfallbk_data;
+       ev_msg.dv[0].data_length = sizeof(apnfallbk_data);
+       kev_post_msg(&ev_msg);
+       apn_fallbk_log((LOG_INFO, "APN fallback notification issued.\n"));
+}
+
 /*
 /*
- *   Transform old in_pcbconnect() into an inner subroutine for new
- *   in_pcbconnect(): Do some validity-checking on the remote
- *   address (in mbuf 'nam') and then determine local host address
- *   (i.e., which interface) to use to access that remote host.
+ * Transform old in_pcbconnect() into an inner subroutine for new
+ * in_pcbconnect(); do some validity-checking on the remote address
+ * (in "nam") and then determine local host address (i.e., which
+ * interface) to use to access that remote host.
+ *
+ * This routine may alter the caller-supplied remote address "nam".
  *
  *
- *   This preserves definition of in_pcbconnect(), while supporting a
- *   slightly different version for T/TCP.  (This is more than
- *   a bit of a kludge, but cleaning up the internal interfaces would
- *   have forced minor changes in every protocol).
+ * The caller may override the bound-to-interface setting of the socket
+ * by specifying the ifscope parameter (e.g. from IP_PKTINFO.)
+ *
+ * This routine might return an ifp with a reference held if the caller
+ * provides a non-NULL outif, even in the error case.  The caller is
+ * responsible for releasing its reference.
  *
  * Returns:    0                       Success
  *             EINVAL                  Invalid argument
  *
  * Returns:    0                       Success
  *             EINVAL                  Invalid argument
@@ -604,149 +1206,264 @@ in_pcbbind(struct inpcb *inp, struct sockaddr *nam, struct proc *p)
  *             EADDRNOTAVAIL           Address not available
  */
 int
  *             EADDRNOTAVAIL           Address not available
  */
 int
-in_pcbladdr(struct inpcb *inp, struct sockaddr *nam,
-           struct sockaddr_in **plocal_sin)
+in_pcbladdr(struct inpcb *inp, struct sockaddr *nam, struct in_addr *laddr,
+    unsigned int ifscope, struct ifnet **outif, int raw)
 {
 {
-       struct in_ifaddr *ia;
-       struct sockaddr_in *sin = (struct sockaddr_in *)nam;
-
-       if (nam->sa_len != sizeof (*sin))
+       struct route *ro = &inp->inp_route;
+       struct in_ifaddr *ia = NULL;
+       struct sockaddr_in sin;
+       int error = 0;
+       boolean_t restricted = FALSE;
+
+       if (outif != NULL)
+               *outif = NULL;
+       if (nam->sa_len != sizeof (struct sockaddr_in))
                return (EINVAL);
                return (EINVAL);
-       if (sin->sin_family != AF_INET)
+       if (SIN(nam)->sin_family != AF_INET)
                return (EAFNOSUPPORT);
                return (EAFNOSUPPORT);
-       if (sin->sin_port == 0)
+       if (raw == 0 && SIN(nam)->sin_port == 0)
                return (EADDRNOTAVAIL);
 
                return (EADDRNOTAVAIL);
 
-       lck_rw_lock_shared(in_ifaddr_rwlock);
-       if (!TAILQ_EMPTY(&in_ifaddrhead)) {
-               /*
-                * If the destination address is INADDR_ANY,
-                * use the primary local address.
-                * If the supplied address is INADDR_BROADCAST,
-                * and the primary interface supports broadcast,
-                * choose the broadcast address for that interface.
-                */
-#define        satosin(sa)     ((struct sockaddr_in *)(sa))
-#define sintosa(sin)   ((struct sockaddr *)(sin))
-#define ifatoia(ifa)   ((struct in_ifaddr *)(ifa))
-               if (sin->sin_addr.s_addr == INADDR_ANY)
-                   sin->sin_addr = IA_SIN(TAILQ_FIRST(&in_ifaddrhead))->sin_addr;
-               else if (sin->sin_addr.s_addr == (u_int32_t)INADDR_BROADCAST &&
-                 (TAILQ_FIRST(&in_ifaddrhead)->ia_ifp->if_flags & IFF_BROADCAST))
-                   sin->sin_addr = satosin(&TAILQ_FIRST(&in_ifaddrhead)->ia_broadaddr)->sin_addr;
+       /*
+        * If the destination address is INADDR_ANY,
+        * use the primary local address.
+        * If the supplied address is INADDR_BROADCAST,
+        * and the primary interface supports broadcast,
+        * choose the broadcast address for that interface.
+        */
+       if (raw == 0 && (SIN(nam)->sin_addr.s_addr == INADDR_ANY ||
+           SIN(nam)->sin_addr.s_addr == (u_int32_t)INADDR_BROADCAST)) {
+               lck_rw_lock_shared(in_ifaddr_rwlock);
+               if (!TAILQ_EMPTY(&in_ifaddrhead)) {
+                       ia = TAILQ_FIRST(&in_ifaddrhead);
+                       IFA_LOCK_SPIN(&ia->ia_ifa);
+                       if (SIN(nam)->sin_addr.s_addr == INADDR_ANY) {
+                               SIN(nam)->sin_addr = IA_SIN(ia)->sin_addr;
+                       } else if (ia->ia_ifp->if_flags & IFF_BROADCAST) {
+                               SIN(nam)->sin_addr =
+                                   SIN(&ia->ia_broadaddr)->sin_addr;
+                       }
+                       IFA_UNLOCK(&ia->ia_ifa);
+                       ia = NULL;
+               }
+               lck_rw_done(in_ifaddr_rwlock);
+       }
+       /*
+        * Otherwise, if the socket has already bound the source, just use it.
+        */
+       if (inp->inp_laddr.s_addr != INADDR_ANY) {
+               VERIFY(ia == NULL);
+               *laddr = inp->inp_laddr;
+               return (0);
        }
        }
-       lck_rw_done(in_ifaddr_rwlock);
 
 
-       if (inp->inp_laddr.s_addr == INADDR_ANY) {
-               struct route *ro;
-               unsigned int ifscope;
+       /*
+        * If the ifscope is specified by the caller (e.g. IP_PKTINFO)
+        * then it overrides the sticky ifscope set for the socket.
+        */
+       if (ifscope == IFSCOPE_NONE && (inp->inp_flags & INP_BOUND_IF))
+               ifscope = inp->inp_boundifp->if_index;
 
 
-               ia = (struct in_ifaddr *)0;
-               ifscope = (inp->inp_flags & INP_BOUND_IF) ?
-                   inp->inp_boundif : IFSCOPE_NONE;
-               /*
-                * If route is known or can be allocated now,
-                * our src addr is taken from the i/f, else punt.
-                * Note that we should check the address family of the cached
-                * destination, in case of sharing the cache with IPv6.
-                */
-               ro = &inp->inp_route;
+       /*
+        * If route is known or can be allocated now,
+        * our src addr is taken from the i/f, else punt.
+        * Note that we should check the address family of the cached
+        * destination, in case of sharing the cache with IPv6.
+        */
+       if (ro->ro_rt != NULL)
+               RT_LOCK_SPIN(ro->ro_rt);
+       if (ROUTE_UNUSABLE(ro) || ro->ro_dst.sa_family != AF_INET ||
+           SIN(&ro->ro_dst)->sin_addr.s_addr != SIN(nam)->sin_addr.s_addr ||
+           (inp->inp_socket->so_options & SO_DONTROUTE)) {
                if (ro->ro_rt != NULL)
                if (ro->ro_rt != NULL)
-                       RT_LOCK_SPIN(ro->ro_rt);
-               if (ro->ro_rt && (ro->ro_dst.sa_family != AF_INET ||
-                   satosin(&ro->ro_dst)->sin_addr.s_addr !=
-                   sin->sin_addr.s_addr ||
-                   inp->inp_socket->so_options & SO_DONTROUTE ||
-                   ro->ro_rt->generation_id != route_generation)) {
                        RT_UNLOCK(ro->ro_rt);
                        RT_UNLOCK(ro->ro_rt);
-                       rtfree(ro->ro_rt);
-                       ro->ro_rt = NULL;
-               }
-               if ((inp->inp_socket->so_options & SO_DONTROUTE) == 0 && /*XXX*/
-                   (ro->ro_rt == NULL || ro->ro_rt->rt_ifp == NULL)) {
-                       if (ro->ro_rt != NULL)
-                               RT_UNLOCK(ro->ro_rt);
-                       /* No route yet, so try to acquire one */
-                       bzero(&ro->ro_dst, sizeof(struct sockaddr_in));
-                       ro->ro_dst.sa_family = AF_INET;
-                       ro->ro_dst.sa_len = sizeof(struct sockaddr_in);
-                       ((struct sockaddr_in *) &ro->ro_dst)->sin_addr =
-                               sin->sin_addr;
-                       rtalloc_scoped_ign(ro, 0, ifscope);
-                       if (ro->ro_rt != NULL)
-                               RT_LOCK_SPIN(ro->ro_rt);
-               }
+               ROUTE_RELEASE(ro);
+       }
+       if (!(inp->inp_socket->so_options & SO_DONTROUTE) &&
+           (ro->ro_rt == NULL || ro->ro_rt->rt_ifp == NULL)) {
+               if (ro->ro_rt != NULL)
+                       RT_UNLOCK(ro->ro_rt);
+               ROUTE_RELEASE(ro);
+               /* No route yet, so try to acquire one */
+               bzero(&ro->ro_dst, sizeof (struct sockaddr_in));
+               ro->ro_dst.sa_family = AF_INET;
+               ro->ro_dst.sa_len = sizeof (struct sockaddr_in);
+               SIN(&ro->ro_dst)->sin_addr = SIN(nam)->sin_addr;
+               rtalloc_scoped(ro, ifscope);
+               if (ro->ro_rt != NULL)
+                       RT_LOCK_SPIN(ro->ro_rt);
+       }
+       /* Sanitized local copy for interface address searches */
+       bzero(&sin, sizeof (sin));
+       sin.sin_family = AF_INET;
+       sin.sin_len = sizeof (struct sockaddr_in);
+       sin.sin_addr.s_addr = SIN(nam)->sin_addr.s_addr;
+       /*
+        * If we did not find (or use) a route, assume dest is reachable
+        * on a directly connected network and try to find a corresponding
+        * interface to take the source address from.
+        */
+       if (ro->ro_rt == NULL) {
+               proc_t proc = current_proc();
+
+               VERIFY(ia == NULL);
+               ia = ifatoia(ifa_ifwithdstaddr(SA(&sin)));
+               if (ia == NULL)
+                       ia = ifatoia(ifa_ifwithnet_scoped(SA(&sin), ifscope));
+               error = ((ia == NULL) ? ENETUNREACH : 0);
+
+               if (apn_fallback_required(proc, inp->inp_socket,
+                   (void *)nam))
+                       apn_fallback_trigger(proc, inp->inp_socket);
+
+               goto done;
+       }
+       RT_LOCK_ASSERT_HELD(ro->ro_rt);
+       /*
+        * If the outgoing interface on the route found is not
+        * a loopback interface, use the address from that interface.
+        */
+       if (!(ro->ro_rt->rt_ifp->if_flags & IFF_LOOPBACK)) {
+               VERIFY(ia == NULL);
                /*
                /*
-                * If we found a route, use the address
-                * corresponding to the outgoing interface
-                * unless it is the loopback (in case a route
-                * to our address on another net goes to loopback).
+                * If the route points to a cellular interface and the
+                * caller forbids our using interfaces of such type,
+                * pretend that there is no route.
+                * Apply the same logic for expensive interfaces.
                 */
                 */
-               if (ro->ro_rt != NULL) {
-                       RT_LOCK_ASSERT_HELD(ro->ro_rt);
-                       if (!(ro->ro_rt->rt_ifp->if_flags & IFF_LOOPBACK)) {
-                               ia = ifatoia(ro->ro_rt->rt_ifa);
-                               if (ia)
-                                       ifaref(&ia->ia_ifa);
-                       }
+               if (inp_restricted_send(inp, ro->ro_rt->rt_ifp)) {
+                       RT_UNLOCK(ro->ro_rt);
+                       ROUTE_RELEASE(ro);
+                       error = EHOSTUNREACH;
+                       restricted = TRUE;
+               } else {
+                       /* Become a regular mutex */
+                       RT_CONVERT_LOCK(ro->ro_rt);
+                       ia = ifatoia(ro->ro_rt->rt_ifa);
+                       IFA_ADDREF(&ia->ia_ifa);
+
+                       /*
+                        * Mark the control block for notification of
+                        * a possible flow that might undergo clat46
+                        * translation.
+                        *
+                        * We defer the decision to a later point when
+                        * inpcb is being disposed off.
+                        * The reason is that we only want to send notification
+                        * if the flow was ever used to send data.
+                        */
+                       if (IS_INTF_CLAT46(ro->ro_rt->rt_ifp))
+                               inp->inp_flags2 |= INP2_CLAT46_FLOW;
+
                        RT_UNLOCK(ro->ro_rt);
                        RT_UNLOCK(ro->ro_rt);
+                       error = 0;
                }
                }
-               if (ia == 0) {
-                       u_short fport = sin->sin_port;
-
-                       sin->sin_port = 0;
-                       ia = ifatoia(ifa_ifwithdstaddr(sintosa(sin)));
-                       if (ia == 0) {
-                               ia = ifatoia(ifa_ifwithnet_scoped(sintosa(sin),
-                                   ifscope));
-                       }
-                       sin->sin_port = fport;
-                       if (ia == 0) {
-                               lck_rw_lock_shared(in_ifaddr_rwlock);
-                               ia = TAILQ_FIRST(&in_ifaddrhead);
-                               if (ia)
-                                       ifaref(&ia->ia_ifa);
-                               lck_rw_done(in_ifaddr_rwlock);
+               goto done;
+       }
+       VERIFY(ro->ro_rt->rt_ifp->if_flags & IFF_LOOPBACK);
+       RT_UNLOCK(ro->ro_rt);
+       /*
+        * The outgoing interface is marked with 'loopback net', so a route
+        * to ourselves is here.
+        * Try to find the interface of the destination address and then
+        * take the address from there. That interface is not necessarily
+        * a loopback interface.
+        */
+       VERIFY(ia == NULL);
+       ia = ifatoia(ifa_ifwithdstaddr(SA(&sin)));
+       if (ia == NULL)
+               ia = ifatoia(ifa_ifwithaddr_scoped(SA(&sin), ifscope));
+       if (ia == NULL)
+               ia = ifatoia(ifa_ifwithnet_scoped(SA(&sin), ifscope));
+       if (ia == NULL) {
+               RT_LOCK(ro->ro_rt);
+               ia = ifatoia(ro->ro_rt->rt_ifa);
+               if (ia != NULL)
+                       IFA_ADDREF(&ia->ia_ifa);
+               RT_UNLOCK(ro->ro_rt);
+       }
+       error = ((ia == NULL) ? ENETUNREACH : 0);
+
+done:
+       /*
+        * If the destination address is multicast and an outgoing
+        * interface has been set as a multicast option, use the
+        * address of that interface as our source address.
+        */
+       if (IN_MULTICAST(ntohl(SIN(nam)->sin_addr.s_addr)) &&
+           inp->inp_moptions != NULL) {
+               struct ip_moptions *imo;
+               struct ifnet *ifp;
+
+               imo = inp->inp_moptions;
+               IMO_LOCK(imo);
+               if (imo->imo_multicast_ifp != NULL && (ia == NULL ||
+                   ia->ia_ifp != imo->imo_multicast_ifp)) {
+                       ifp = imo->imo_multicast_ifp;
+                       if (ia != NULL)
+                               IFA_REMREF(&ia->ia_ifa);
+                       lck_rw_lock_shared(in_ifaddr_rwlock);
+                       TAILQ_FOREACH(ia, &in_ifaddrhead, ia_link) {
+                               if (ia->ia_ifp == ifp)
+                                       break;
                        }
                        }
-                       if (ia == 0)
-                               return (EADDRNOTAVAIL);
+                       if (ia != NULL)
+                               IFA_ADDREF(&ia->ia_ifa);
+                       lck_rw_done(in_ifaddr_rwlock);
+                       if (ia == NULL)
+                               error = EADDRNOTAVAIL;
+                       else
+                               error = 0;
                }
                }
+               IMO_UNLOCK(imo);
+       }
+       /*
+        * Don't do pcblookup call here; return interface in laddr
+        * and exit to caller, that will do the lookup.
+        */
+       if (ia != NULL) {
                /*
                /*
-                * If the destination address is multicast and an outgoing
-                * interface has been set as a multicast option, use the
-                * address of that interface as our source address.
+                * If the source address belongs to a cellular interface
+                * and the socket forbids our using interfaces of such
+                * type, pretend that there is no source address.
+                * Apply the same logic for expensive interfaces.
                 */
                 */
-               if (IN_MULTICAST(ntohl(sin->sin_addr.s_addr)) &&
-                   inp->inp_moptions != NULL) {
-                       struct ip_moptions *imo;
-                       struct ifnet *ifp;
-
-                       imo = inp->inp_moptions;
-                       if (imo->imo_multicast_ifp != NULL && (ia == NULL ||
-                               ia->ia_ifp != imo->imo_multicast_ifp)) {
-                               ifp = imo->imo_multicast_ifp;
-                               if (ia)
-                                       ifafree(&ia->ia_ifa);
-                               lck_rw_lock_shared(in_ifaddr_rwlock);
-                               TAILQ_FOREACH(ia, &in_ifaddrhead, ia_link) {
-                                       if (ia->ia_ifp == ifp)
-                                               break;
-                               }
-                               if (ia)
-                                       ifaref(&ia->ia_ifa);
-                               lck_rw_done(in_ifaddr_rwlock);
-                               if (ia == 0)
-                                       return (EADDRNOTAVAIL);
+               IFA_LOCK_SPIN(&ia->ia_ifa);
+               if (inp_restricted_send(inp, ia->ia_ifa.ifa_ifp)) {
+                       IFA_UNLOCK(&ia->ia_ifa);
+                       error = EHOSTUNREACH;
+                       restricted = TRUE;
+               } else if (error == 0) {
+                       *laddr = ia->ia_addr.sin_addr;
+                       if (outif != NULL) {
+                               struct ifnet *ifp;
+
+                               if (ro->ro_rt != NULL)
+                                       ifp = ro->ro_rt->rt_ifp;
+                               else
+                                       ifp = ia->ia_ifp;
+
+                               VERIFY(ifp != NULL);
+                               IFA_CONVERT_LOCK(&ia->ia_ifa);
+                               ifnet_reference(ifp);   /* for caller */
+                               if (*outif != NULL)
+                                       ifnet_release(*outif);
+                               *outif = ifp;
                        }
                        }
+                       IFA_UNLOCK(&ia->ia_ifa);
+               } else {
+                       IFA_UNLOCK(&ia->ia_ifa);
                }
                }
-               /*
-                * Don't do pcblookup call here; return interface in plocal_sin
-                * and exit to caller, that will do the lookup.
-                */
-               *plocal_sin = &ia->ia_addr;
-               ifafree(&ia->ia_ifa);
+               IFA_REMREF(&ia->ia_ifa);
+               ia = NULL;
+       }
+
+       if (restricted && error == EHOSTUNREACH) {
+               soevent(inp->inp_socket, (SO_FILT_HINT_LOCKED |
+                   SO_FILT_HINT_IFDENIED));
        }
        }
-       return(0);
+
+       return (error);
 }
 
 /*
 }
 
 /*
@@ -755,78 +1472,129 @@ in_pcbladdr(struct inpcb *inp, struct sockaddr *nam,
  * Both address and port must be specified in argument sin.
  * If don't have a local address for this socket yet,
  * then pick one.
  * Both address and port must be specified in argument sin.
  * If don't have a local address for this socket yet,
  * then pick one.
+ *
+ * The caller may override the bound-to-interface setting of the socket
+ * by specifying the ifscope parameter (e.g. from IP_PKTINFO.)
  */
 int
  */
 int
-in_pcbconnect(struct inpcb *inp, struct sockaddr *nam, struct proc *p)
+in_pcbconnect(struct inpcb *inp, struct sockaddr *nam, struct proc *p,
+    unsigned int ifscope, struct ifnet **outif)
 {
 {
-       struct sockaddr_in *ifaddr;
-       struct sockaddr_in *sin = (struct sockaddr_in *)nam;
+       struct in_addr laddr;
+       struct sockaddr_in *sin = (struct sockaddr_in *)(void *)nam;
        struct inpcb *pcb;
        int error;
        struct inpcb *pcb;
        int error;
+       struct socket *so = inp->inp_socket;
+
+#if CONTENT_FILTER
+       if (so)
+               so->so_state_change_cnt++;
+#endif
 
        /*
         *   Call inner routine, to assign local interface address.
         */
 
        /*
         *   Call inner routine, to assign local interface address.
         */
-       if ((error = in_pcbladdr(inp, nam, &ifaddr)) != 0)
-               return(error);
+       if ((error = in_pcbladdr(inp, nam, &laddr, ifscope, outif, 0)) != 0)
+               return (error);
 
 
-       socket_unlock(inp->inp_socket, 0);
+       socket_unlock(so, 0);
        pcb = in_pcblookup_hash(inp->inp_pcbinfo, sin->sin_addr, sin->sin_port,
        pcb = in_pcblookup_hash(inp->inp_pcbinfo, sin->sin_addr, sin->sin_port,
-           inp->inp_laddr.s_addr ? inp->inp_laddr : ifaddr->sin_addr,
+           inp->inp_laddr.s_addr ? inp->inp_laddr : laddr,
            inp->inp_lport, 0, NULL);
            inp->inp_lport, 0, NULL);
-       socket_lock(inp->inp_socket, 0);
+       socket_lock(so, 0);
+
+       /*
+        * Check if the socket is still in a valid state. When we unlock this
+        * embryonic socket, it can get aborted if another thread is closing
+        * the listener (radar 7947600).
+        */
+       if ((so->so_flags & SOF_ABORTED) != 0)
+               return (ECONNREFUSED);
+
        if (pcb != NULL) {
        if (pcb != NULL) {
-               in_pcb_checkstate(pcb, WNT_RELEASE, 0);
+               in_pcb_checkstate(pcb, WNT_RELEASE, pcb == inp ? 1 : 0);
                return (EADDRINUSE);
        }
        if (inp->inp_laddr.s_addr == INADDR_ANY) {
                if (inp->inp_lport == 0) {
                return (EADDRINUSE);
        }
        if (inp->inp_laddr.s_addr == INADDR_ANY) {
                if (inp->inp_lport == 0) {
-                       error = in_pcbbind(inp, (struct sockaddr *)0, p);
+                       error = in_pcbbind(inp, NULL, p);
                        if (error)
                        if (error)
-                           return (error);
+                               return (error);
                }
                }
-               if (!lck_rw_try_lock_exclusive(inp->inp_pcbinfo->mtx)) {
-                       /*lock inversion issue, mostly with udp multicast packets */
-                       socket_unlock(inp->inp_socket, 0);
-                       lck_rw_lock_exclusive(inp->inp_pcbinfo->mtx);
-                       socket_lock(inp->inp_socket, 0);
+               if (!lck_rw_try_lock_exclusive(inp->inp_pcbinfo->ipi_lock)) {
+                       /*
+                        * Lock inversion issue, mostly with udp
+                        * multicast packets.
+                        */
+                       socket_unlock(so, 0);
+                       lck_rw_lock_exclusive(inp->inp_pcbinfo->ipi_lock);
+                       socket_lock(so, 0);
                }
                }
-               inp->inp_laddr = ifaddr->sin_addr;
+               inp->inp_laddr = laddr;
+               /* no reference needed */
+               inp->inp_last_outifp = (outif != NULL) ? *outif : NULL;
                inp->inp_flags |= INP_INADDR_ANY;
                inp->inp_flags |= INP_INADDR_ANY;
-       }
-        else {
-               if (!lck_rw_try_lock_exclusive(inp->inp_pcbinfo->mtx)) {
-                       /*lock inversion issue, mostly with udp multicast packets */
-                       socket_unlock(inp->inp_socket, 0);
-                       lck_rw_lock_exclusive(inp->inp_pcbinfo->mtx);
-                       socket_lock(inp->inp_socket, 0);
+       } else {
+               /*
+                * Usage of IP_PKTINFO, without local port already
+                * speficified will cause kernel to panic,
+                * see rdar://problem/18508185.
+                * For now returning error to avoid a kernel panic
+                * This routines can be refactored and handle this better
+                * in future.
+                */
+               if (inp->inp_lport == 0)
+                       return (EINVAL);
+               if (!lck_rw_try_lock_exclusive(inp->inp_pcbinfo->ipi_lock)) {
+                       /*
+                        * Lock inversion issue, mostly with udp
+                        * multicast packets.
+                        */
+                       socket_unlock(so, 0);
+                       lck_rw_lock_exclusive(inp->inp_pcbinfo->ipi_lock);
+                       socket_lock(so, 0);
                }
        }
        inp->inp_faddr = sin->sin_addr;
        inp->inp_fport = sin->sin_port;
                }
        }
        inp->inp_faddr = sin->sin_addr;
        inp->inp_fport = sin->sin_port;
+       if (nstat_collect && SOCK_PROTO(so) == IPPROTO_UDP)
+               nstat_pcb_invalidate_cache(inp);
        in_pcbrehash(inp);
        in_pcbrehash(inp);
-       lck_rw_done(inp->inp_pcbinfo->mtx);
+       lck_rw_done(inp->inp_pcbinfo->ipi_lock);
        return (0);
 }
 
 void
 in_pcbdisconnect(struct inpcb *inp)
 {
        return (0);
 }
 
 void
 in_pcbdisconnect(struct inpcb *inp)
 {
+       struct socket *so = inp->inp_socket;
+
+       if (nstat_collect && SOCK_PROTO(so) == IPPROTO_UDP)
+               nstat_pcb_cache(inp);
 
        inp->inp_faddr.s_addr = INADDR_ANY;
        inp->inp_fport = 0;
 
 
        inp->inp_faddr.s_addr = INADDR_ANY;
        inp->inp_fport = 0;
 
-       if (!lck_rw_try_lock_exclusive(inp->inp_pcbinfo->mtx)) {
-               /*lock inversion issue, mostly with udp multicast packets */
-               socket_unlock(inp->inp_socket, 0);
-               lck_rw_lock_exclusive(inp->inp_pcbinfo->mtx);
-               socket_lock(inp->inp_socket, 0);
+#if CONTENT_FILTER
+       if (so)
+               so->so_state_change_cnt++;
+#endif
+
+       if (!lck_rw_try_lock_exclusive(inp->inp_pcbinfo->ipi_lock)) {
+               /* lock inversion issue, mostly with udp multicast packets */
+               socket_unlock(so, 0);
+               lck_rw_lock_exclusive(inp->inp_pcbinfo->ipi_lock);
+               socket_lock(so, 0);
        }
 
        in_pcbrehash(inp);
        }
 
        in_pcbrehash(inp);
-       lck_rw_done(inp->inp_pcbinfo->mtx);
-
-       if (inp->inp_socket->so_state & SS_NOFDREF) 
+       lck_rw_done(inp->inp_pcbinfo->ipi_lock);
+       /*
+        * A multipath subflow socket would have its SS_NOFDREF set by default,
+        * so check for SOF_MP_SUBFLOW socket flag before detaching the PCB;
+        * when the socket is closed for real, SOF_MP_SUBFLOW would be cleared.
+        */
+       if (!(so->so_flags & SOF_MP_SUBFLOW) && (so->so_state & SS_NOFDREF))
                in_pcbdetach(inp);
 }
 
                in_pcbdetach(inp);
 }
 
@@ -835,126 +1603,194 @@ in_pcbdetach(struct inpcb *inp)
 {
        struct socket *so = inp->inp_socket;
 
 {
        struct socket *so = inp->inp_socket;
 
-       if (so->so_pcb == 0) { /* we've been called twice */
-               panic("in_pcbdetach: inp=%p so=%p proto=%d so_pcb is null!\n",
-                       inp, so, so->so_proto->pr_protocol);
+       if (so->so_pcb == NULL) {
+               /* PCB has been disposed */
+               panic("%s: inp=%p so=%p proto=%d so_pcb is null!\n", __func__,
+                   inp, so, SOCK_PROTO(so));
+               /* NOTREACHED */
        }
 
 #if IPSEC
        }
 
 #if IPSEC
-       if (ipsec_bypass == 0) {
-               ipsec4_delete_pcbpolicy(inp);
+       if (inp->inp_sp != NULL) {
+               (void) ipsec4_delete_pcbpolicy(inp);
+       }
+#endif /* IPSEC */
+
+       if (inp->inp_stat != NULL && SOCK_PROTO(so) == IPPROTO_UDP) {
+               if (inp->inp_stat->rxpackets == 0 && inp->inp_stat->txpackets == 0) {
+                       INC_ATOMIC_INT64_LIM(net_api_stats.nas_socket_inet_dgram_no_data);
+               }
+       }
+
+       /*
+        * Let NetworkStatistics know this PCB is going away
+        * before we detach it.
+        */
+       if (nstat_collect &&
+           (SOCK_PROTO(so) == IPPROTO_TCP || SOCK_PROTO(so) == IPPROTO_UDP))
+               nstat_pcb_detach(inp);
+
+       /* Free memory buffer held for generating keep alives */
+       if (inp->inp_keepalive_data != NULL) {
+               FREE(inp->inp_keepalive_data, M_TEMP);
+               inp->inp_keepalive_data = NULL;
        }
        }
-#endif /*IPSEC*/
 
        /* mark socket state as dead */
 
        /* mark socket state as dead */
-       if (in_pcb_checkstate(inp, WNT_STOPUSING, 1) != WNT_STOPUSING)
-               panic("in_pcbdetach so=%p prot=%x couldn't set to STOPUSING\n", so, so->so_proto->pr_protocol);
+       if (in_pcb_checkstate(inp, WNT_STOPUSING, 1) != WNT_STOPUSING) {
+               panic("%s: so=%p proto=%d couldn't set to STOPUSING\n",
+                   __func__, so, SOCK_PROTO(so));
+               /* NOTREACHED */
+       }
 
 
-#if TEMPDEBUG
-       if (so->cached_in_sock_layer)
-           printf("in_pcbdetach for cached socket %x flags=%x\n", so, so->so_flags);
-       else
-           printf("in_pcbdetach for allocated socket %x flags=%x\n", so, so->so_flags);
-#endif
-       if ((so->so_flags & SOF_PCBCLEARING) == 0) {
-               struct rtentry *rt;
+       if (!(so->so_flags & SOF_PCBCLEARING)) {
+               struct ip_moptions *imo;
 
                inp->inp_vflag = 0;
 
                inp->inp_vflag = 0;
-               if (inp->inp_options) 
-                       (void)m_free(inp->inp_options);
-               if ((rt = inp->inp_route.ro_rt) != NULL) {
-                       inp->inp_route.ro_rt = NULL;
-                       rtfree(rt);
+               if (inp->inp_options != NULL) {
+                       (void) m_free(inp->inp_options);
+                       inp->inp_options = NULL;
                }
                }
-               ip_freemoptions(inp->inp_moptions);
+               ROUTE_RELEASE(&inp->inp_route);
+               imo = inp->inp_moptions;
                inp->inp_moptions = NULL;
                sofreelastref(so, 0);
                inp->inp_state = INPCB_STATE_DEAD;
                inp->inp_moptions = NULL;
                sofreelastref(so, 0);
                inp->inp_state = INPCB_STATE_DEAD;
-               so->so_flags |= SOF_PCBCLEARING; /* makes sure we're not called twice from so_close */
+
+               /*
+                * Enqueue an event to send kernel event notification
+                * if the flow has to CLAT46 for data packets
+                */
+               if (inp->inp_flags2 & INP2_CLAT46_FLOW) {
+                       /*
+                        * If there has been any exchange of data bytes
+                        * over this flow.
+                        * Schedule a notification to report that flow is
+                        * using client side translation.
+                        */
+                       if (inp->inp_stat != NULL &&
+                           (inp->inp_stat->txbytes != 0 ||
+                            inp->inp_stat->rxbytes !=0)) {
+                               if (so->so_flags & SOF_DELEGATED) {
+                                       in6_clat46_event_enqueue_nwk_wq_entry(
+                                           IN6_CLAT46_EVENT_V4_FLOW,
+                                           so->e_pid,
+                                           so->e_uuid);
+                               } else {
+                                       in6_clat46_event_enqueue_nwk_wq_entry(
+                                            IN6_CLAT46_EVENT_V4_FLOW,
+                                            so->last_pid,
+                                            so->last_uuid);
+                               }
+                       }
+               }
+
+               /* makes sure we're not called twice from so_close */
+               so->so_flags |= SOF_PCBCLEARING;
+
+               inpcb_gc_sched(inp->inp_pcbinfo, INPCB_TIMER_FAST);
+
+               /*
+                * See inp_join_group() for why we need to unlock
+                */
+               if (imo != NULL) {
+                       socket_unlock(so, 0);
+                       IMO_REMREF(imo);
+                       socket_lock(so, 0);
+               }
        }
 }
 
 
        }
 }
 
 
-void 
-in_pcbdispose(struct inpcb *inp) 
+void
+in_pcbdispose(struct inpcb *inp)
 {
        struct socket *so = inp->inp_socket;
        struct inpcbinfo *ipi = inp->inp_pcbinfo;
 
 {
        struct socket *so = inp->inp_socket;
        struct inpcbinfo *ipi = inp->inp_pcbinfo;
 
-#if TEMPDEBUG
-       if (inp->inp_state != INPCB_STATE_DEAD) {
-               printf("in_pcbdispose: not dead yet? so=%p\n", so);
+       if (so != NULL && so->so_usecount != 0) {
+               panic("%s: so %p [%d,%d] usecount %d lockhistory %s\n",
+                   __func__, so, SOCK_DOM(so), SOCK_TYPE(so), so->so_usecount,
+                   solockhistory_nr(so));
+               /* NOTREACHED */
+       } else if (inp->inp_wantcnt != WNT_STOPUSING) {
+               if (so != NULL) {
+                       panic_plain("%s: inp %p invalid wantcnt %d, so %p "
+                           "[%d,%d] usecount %d retaincnt %d state 0x%x "
+                           "flags 0x%x lockhistory %s\n", __func__, inp,
+                           inp->inp_wantcnt, so, SOCK_DOM(so), SOCK_TYPE(so),
+                           so->so_usecount, so->so_retaincnt, so->so_state,
+                           so->so_flags, solockhistory_nr(so));
+                       /* NOTREACHED */
+               } else {
+                       panic("%s: inp %p invalid wantcnt %d no socket\n",
+                           __func__, inp, inp->inp_wantcnt);
+                       /* NOTREACHED */
+               }
        }
        }
-#endif
-
-       if (so && so->so_usecount != 0)
-               panic("in_pcbdispose: use count=%x so=%p\n", so->so_usecount, so);
 
 
-       lck_rw_assert(ipi->mtx, LCK_RW_ASSERT_EXCLUSIVE);
+       LCK_RW_ASSERT(ipi->ipi_lock, LCK_RW_ASSERT_EXCLUSIVE);
 
        inp->inp_gencnt = ++ipi->ipi_gencnt;
 
        inp->inp_gencnt = ++ipi->ipi_gencnt;
-       /*### access ipi in in_pcbremlists */
+       /* access ipi in in_pcbremlists */
        in_pcbremlists(inp);
        in_pcbremlists(inp);
-       
-       if (so) {
+
+       if (so != NULL) {
                if (so->so_proto->pr_flags & PR_PCBLOCK) {
                        sofreelastref(so, 0);
                if (so->so_proto->pr_flags & PR_PCBLOCK) {
                        sofreelastref(so, 0);
-                       if (so->so_rcv.sb_cc || so->so_snd.sb_cc) {
-#if TEMPDEBUG
-                               printf("in_pcbdispose sb not cleaned up so=%p rc_cci=%x snd_cc=%x\n",
-                                       so, so->so_rcv.sb_cc, so->so_snd.sb_cc);        
-#endif
+                       if (so->so_rcv.sb_cc > 0 || so->so_snd.sb_cc > 0) {
+                               /*
+                                * selthreadclear() already called
+                                * during sofreelastref() above.
+                                */
                                sbrelease(&so->so_rcv);
                                sbrelease(&so->so_snd);
                        }
                                sbrelease(&so->so_rcv);
                                sbrelease(&so->so_snd);
                        }
-                       if (so->so_head != NULL)
-                               panic("in_pcbdispose, so=%p head still exist\n", so);
-                       lck_mtx_unlock(inp->inpcb_mtx); 
-                       lck_mtx_free(inp->inpcb_mtx, ipi->mtx_grp);     
+                       if (so->so_head != NULL) {
+                               panic("%s: so=%p head still exist\n",
+                                   __func__, so);
+                               /* NOTREACHED */
+                       }
+                       lck_mtx_unlock(&inp->inpcb_mtx);
+
+#if NECP
+                       necp_inpcb_remove_cb(inp);
+#endif /* NECP */
+
+                       lck_mtx_destroy(&inp->inpcb_mtx, ipi->ipi_lock_grp);
                }
                }
-               so->so_flags |= SOF_PCBCLEARING; /* makes sure we're not called twice from so_close */
-               so->so_saved_pcb = (caddr_t) inp;
-               so->so_pcb = 0; 
-               inp->inp_socket = 0;
+               /* makes sure we're not called twice from so_close */
+               so->so_flags |= SOF_PCBCLEARING;
+               so->so_saved_pcb = (caddr_t)inp;
+               so->so_pcb = NULL;
+               inp->inp_socket = NULL;
 #if CONFIG_MACF_NET
                mac_inpcb_label_destroy(inp);
 #if CONFIG_MACF_NET
                mac_inpcb_label_destroy(inp);
-#endif
+#endif /* CONFIG_MACF_NET */
+#if NECP
+               necp_inpcb_dispose(inp);
+#endif /* NECP */
                /*
                 * In case there a route cached after a detach (possible
                 * in the tcp case), make sure that it is freed before
                 * we deallocate the structure.
                 */
                /*
                 * In case there a route cached after a detach (possible
                 * in the tcp case), make sure that it is freed before
                 * we deallocate the structure.
                 */
-               if (inp->inp_route.ro_rt != NULL) {
-                       rtfree(inp->inp_route.ro_rt);
-                       inp->inp_route.ro_rt = NULL;
-               }
-               if (so->cached_in_sock_layer == 0) {
+               ROUTE_RELEASE(&inp->inp_route);
+               if ((so->so_flags1 & SOF1_CACHED_IN_SOCK_LAYER) == 0) {
                        zfree(ipi->ipi_zone, inp);
                }
                sodealloc(so);
        }
                        zfree(ipi->ipi_zone, inp);
                }
                sodealloc(so);
        }
-#if TEMPDEBUG
-       else
-               printf("in_pcbdispose: no socket for inp=%p\n", inp);
-#endif
 }
 
 /*
 }
 
 /*
- * The calling convention of in_setsockaddr() and in_setpeeraddr() was
+ * The calling convention of in_getsockaddr() and in_getpeeraddr() was
  * modified to match the pru_sockaddr() and pru_peeraddr() entry points
  * in struct pr_usrreqs, so that protocols can just reference then directly
  * modified to match the pru_sockaddr() and pru_peeraddr() entry points
  * in struct pr_usrreqs, so that protocols can just reference then directly
- * without the need for a wrapper function.  The socket must have a valid
- * (i.e., non-nil) PCB, but it should be impossible to get an invalid one
- * except through a kernel programming error, so it is acceptable to panic
- * (or in this case trap) if the PCB is invalid.  (Actually, we don't trap
- * because there actually /is/ a programming error somewhere... XXX)
- *
- * Returns:    0                       Success
- *             ENOBUFS                 No buffer space available
- *             ECONNRESET              Connection reset
+ * without the need for a wrapper function.
  */
 int
  */
 int
-in_setsockaddr(struct socket *so, struct sockaddr **nam)
+in_getsockaddr(struct socket *so, struct sockaddr **nam)
 {
        struct inpcb *inp;
        struct sockaddr_in *sin;
 {
        struct inpcb *inp;
        struct sockaddr_in *sin;
@@ -962,27 +1798,46 @@ in_setsockaddr(struct socket *so, struct sockaddr **nam)
        /*
         * Do the malloc first in case it blocks.
         */
        /*
         * Do the malloc first in case it blocks.
         */
-       MALLOC(sin, struct sockaddr_in *, sizeof *sin, M_SONAME, M_WAITOK);
+       MALLOC(sin, struct sockaddr_in *, sizeof (*sin), M_SONAME, M_WAITOK);
        if (sin == NULL)
        if (sin == NULL)
-               return ENOBUFS;
-       bzero(sin, sizeof *sin);
+               return (ENOBUFS);
+       bzero(sin, sizeof (*sin));
        sin->sin_family = AF_INET;
        sin->sin_family = AF_INET;
-       sin->sin_len = sizeof(*sin);
+       sin->sin_len = sizeof (*sin);
 
 
-       inp = sotoinpcb(so);
-       if (!inp) {
+       if ((inp = sotoinpcb(so)) == NULL) {
                FREE(sin, M_SONAME);
                FREE(sin, M_SONAME);
-               return ECONNRESET;
+               return (EINVAL);
        }
        sin->sin_port = inp->inp_lport;
        sin->sin_addr = inp->inp_laddr;
 
        *nam = (struct sockaddr *)sin;
        }
        sin->sin_port = inp->inp_lport;
        sin->sin_addr = inp->inp_laddr;
 
        *nam = (struct sockaddr *)sin;
-       return 0;
+       return (0);
 }
 
 int
 }
 
 int
-in_setpeeraddr(struct socket *so, struct sockaddr **nam)
+in_getsockaddr_s(struct socket *so, struct sockaddr_in *ss)
+{
+       struct sockaddr_in *sin = ss;
+       struct inpcb *inp;
+
+       VERIFY(ss != NULL);
+       bzero(ss, sizeof (*ss));
+
+       sin->sin_family = AF_INET;
+       sin->sin_len = sizeof (*sin);
+
+       if ((inp = sotoinpcb(so)) == NULL)
+               return (EINVAL);
+
+       sin->sin_port = inp->inp_lport;
+       sin->sin_addr = inp->inp_laddr;
+       return (0);
+}
+
+int
+in_getpeeraddr(struct socket *so, struct sockaddr **nam)
 {
        struct inpcb *inp;
        struct sockaddr_in *sin;
 {
        struct inpcb *inp;
        struct sockaddr_in *sin;
@@ -990,49 +1845,48 @@ in_setpeeraddr(struct socket *so, struct sockaddr **nam)
        /*
         * Do the malloc first in case it blocks.
         */
        /*
         * Do the malloc first in case it blocks.
         */
-       MALLOC(sin, struct sockaddr_in *, sizeof *sin, M_SONAME, M_WAITOK);
+       MALLOC(sin, struct sockaddr_in *, sizeof (*sin), M_SONAME, M_WAITOK);
        if (sin == NULL)
        if (sin == NULL)
-               return ENOBUFS;
+               return (ENOBUFS);
        bzero((caddr_t)sin, sizeof (*sin));
        sin->sin_family = AF_INET;
        bzero((caddr_t)sin, sizeof (*sin));
        sin->sin_family = AF_INET;
-       sin->sin_len = sizeof(*sin);
+       sin->sin_len = sizeof (*sin);
 
 
-       inp = sotoinpcb(so);
-       if (!inp) {
+       if ((inp = sotoinpcb(so)) == NULL) {
                FREE(sin, M_SONAME);
                FREE(sin, M_SONAME);
-               return ECONNRESET;
+               return (EINVAL);
        }
        sin->sin_port = inp->inp_fport;
        sin->sin_addr = inp->inp_faddr;
 
        *nam = (struct sockaddr *)sin;
        }
        sin->sin_port = inp->inp_fport;
        sin->sin_addr = inp->inp_faddr;
 
        *nam = (struct sockaddr *)sin;
-       return 0;
+       return (0);
 }
 
 void
 in_pcbnotifyall(struct inpcbinfo *pcbinfo, struct in_addr faddr,
 }
 
 void
 in_pcbnotifyall(struct inpcbinfo *pcbinfo, struct in_addr faddr,
-               int errno, void (*notify)(struct inpcb *, int))
+    int errno, void (*notify)(struct inpcb *, int))
 {
        struct inpcb *inp;
 
 {
        struct inpcb *inp;
 
-       lck_rw_lock_shared(pcbinfo->mtx);
+       lck_rw_lock_shared(pcbinfo->ipi_lock);
 
 
-       LIST_FOREACH(inp, pcbinfo->listhead, inp_list) {
+       LIST_FOREACH(inp, pcbinfo->ipi_listhead, inp_list) {
 #if INET6
 #if INET6
-               if ((inp->inp_vflag & INP_IPV4) == 0)
+               if (!(inp->inp_vflag & INP_IPV4))
                        continue;
                        continue;
-#endif
+#endif /* INET6 */
                if (inp->inp_faddr.s_addr != faddr.s_addr ||
                    inp->inp_socket == NULL)
                if (inp->inp_faddr.s_addr != faddr.s_addr ||
                    inp->inp_socket == NULL)
-                               continue;
-               if (in_pcb_checkstate(inp, WNT_ACQUIRE, 0) == WNT_STOPUSING) 
+                       continue;
+               if (in_pcb_checkstate(inp, WNT_ACQUIRE, 0) == WNT_STOPUSING)
                        continue;
                socket_lock(inp->inp_socket, 1);
                (*notify)(inp, errno);
                        continue;
                socket_lock(inp->inp_socket, 1);
                (*notify)(inp, errno);
-               (void)in_pcb_checkstate(inp, WNT_RELEASE, 1);
+               (void) in_pcb_checkstate(inp, WNT_RELEASE, 1);
                socket_unlock(inp->inp_socket, 1);
        }
                socket_unlock(inp->inp_socket, 1);
        }
-       lck_rw_done(pcbinfo->mtx);
+       lck_rw_done(pcbinfo->ipi_lock);
 }
 
 /*
 }
 
 /*
@@ -1044,19 +1898,13 @@ in_pcbnotifyall(struct inpcbinfo *pcbinfo, struct in_addr faddr,
 void
 in_losing(struct inpcb *inp)
 {
 void
 in_losing(struct inpcb *inp)
 {
+       boolean_t release = FALSE;
        struct rtentry *rt;
        struct rtentry *rt;
-       struct rt_addrinfo info;
 
        if ((rt = inp->inp_route.ro_rt) != NULL) {
 
        if ((rt = inp->inp_route.ro_rt) != NULL) {
-               struct in_ifaddr *ia;
+               struct in_ifaddr *ia = NULL;
 
 
-               bzero((caddr_t)&info, sizeof(info));
                RT_LOCK(rt);
                RT_LOCK(rt);
-               info.rti_info[RTAX_DST] =
-                       (struct sockaddr *)&inp->inp_route.ro_dst;
-               info.rti_info[RTAX_GATEWAY] = rt->rt_gateway;
-               info.rti_info[RTAX_NETMASK] = rt_mask(rt);
-               rt_missmsg(RTM_LOSING, &info, rt->rt_flags, 0);
                if (rt->rt_flags & RTF_DYNAMIC) {
                        /*
                         * Prevent another thread from modifying rt_key,
                if (rt->rt_flags & RTF_DYNAMIC) {
                        /*
                         * Prevent another thread from modifying rt_key,
@@ -1066,22 +1914,24 @@ in_losing(struct inpcb *inp)
                        rt->rt_flags |= RTF_CONDEMNED;
                        RT_UNLOCK(rt);
                        (void) rtrequest(RTM_DELETE, rt_key(rt),
                        rt->rt_flags |= RTF_CONDEMNED;
                        RT_UNLOCK(rt);
                        (void) rtrequest(RTM_DELETE, rt_key(rt),
-                               rt->rt_gateway, rt_mask(rt), rt->rt_flags,
-                               (struct rtentry **)0);
+                           rt->rt_gateway, rt_mask(rt), rt->rt_flags, NULL);
                } else {
                        RT_UNLOCK(rt);
                }
                /* if the address is gone keep the old route in the pcb */
                } else {
                        RT_UNLOCK(rt);
                }
                /* if the address is gone keep the old route in the pcb */
-               if ((ia = ifa_foraddr(inp->inp_laddr.s_addr)) != NULL) {
-                       inp->inp_route.ro_rt = NULL;
-                       rtfree(rt);
-                       ifafree(&ia->ia_ifa);
+               if (inp->inp_laddr.s_addr != INADDR_ANY &&
+                   (ia = ifa_foraddr(inp->inp_laddr.s_addr)) != NULL) {
+                       /*
+                        * Address is around; ditch the route.  A new route
+                        * can be allocated the next time output is attempted.
+                        */
+                       release = TRUE;
                }
                }
-               /*
-                * A new route can be allocated
-                * the next time output is attempted.
-                */
+               if (ia != NULL)
+                       IFA_REMREF(&ia->ia_ifa);
        }
        }
+       if (rt == NULL || release)
+               ROUTE_RELEASE(&inp->inp_route);
 }
 
 /*
 }
 
 /*
@@ -1089,24 +1939,29 @@ in_losing(struct inpcb *inp)
  * and allocate a (hopefully) better one.
  */
 void
  * and allocate a (hopefully) better one.
  */
 void
-in_rtchange(struct inpcb *inp, __unused int errno)
+in_rtchange(struct inpcb *inp, int errno)
 {
 {
+#pragma unused(errno)
+       boolean_t release = FALSE;
        struct rtentry *rt;
 
        if ((rt = inp->inp_route.ro_rt) != NULL) {
        struct rtentry *rt;
 
        if ((rt = inp->inp_route.ro_rt) != NULL) {
-               struct in_ifaddr *ia;
+               struct in_ifaddr *ia = NULL;
 
 
-               if ((ia = ifa_foraddr(inp->inp_laddr.s_addr)) == NULL) {
-                       return; /* we can't remove the route now. not sure if still ok to use src */
+               /* if address is gone, keep the old route */
+               if (inp->inp_laddr.s_addr != INADDR_ANY &&
+                   (ia = ifa_foraddr(inp->inp_laddr.s_addr)) != NULL) {
+                       /*
+                        * Address is around; ditch the route.  A new route
+                        * can be allocated the next time output is attempted.
+                        */
+                       release = TRUE;
                }
                }
-               ifafree(&ia->ia_ifa);
-               rtfree(rt);
-               inp->inp_route.ro_rt = NULL;
-               /*
-                * A new route can be allocated the next time
-                * output is attempted.
-                */
+               if (ia != NULL)
+                       IFA_REMREF(&ia->ia_ifa);
        }
        }
+       if (rt == NULL || release)
+               ROUTE_RELEASE(&inp->inp_route);
 }
 
 /*
 }
 
 /*
@@ -1114,13 +1969,13 @@ in_rtchange(struct inpcb *inp, __unused int errno)
  */
 struct inpcb *
 in_pcblookup_local(struct inpcbinfo *pcbinfo, struct in_addr laddr,
  */
 struct inpcb *
 in_pcblookup_local(struct inpcbinfo *pcbinfo, struct in_addr laddr,
-                  unsigned int lport_arg, int wild_okay)
+    unsigned int lport_arg, int wild_okay)
 {
        struct inpcb *inp;
        int matchwild = 3, wildcard;
        u_short lport = lport_arg;
 
 {
        struct inpcb *inp;
        int matchwild = 3, wildcard;
        u_short lport = lport_arg;
 
-       KERNEL_DEBUG(DBG_FNC_PCB_LOOKUP | DBG_FUNC_START, 0,0,0,0,0);
+       KERNEL_DEBUG(DBG_FNC_PCB_LOOKUP | DBG_FUNC_START, 0, 0, 0, 0, 0);
 
        if (!wild_okay) {
                struct inpcbhead *head;
 
        if (!wild_okay) {
                struct inpcbhead *head;
@@ -1128,12 +1983,13 @@ in_pcblookup_local(struct inpcbinfo *pcbinfo, struct in_addr laddr,
                 * Look for an unconnected (wildcard foreign addr) PCB that
                 * matches the local address and port we're looking for.
                 */
                 * Look for an unconnected (wildcard foreign addr) PCB that
                 * matches the local address and port we're looking for.
                 */
-               head = &pcbinfo->hashbase[INP_PCBHASH(INADDR_ANY, lport, 0, pcbinfo->hashmask)];
+               head = &pcbinfo->ipi_hashbase[INP_PCBHASH(INADDR_ANY, lport, 0,
+                   pcbinfo->ipi_hashmask)];
                LIST_FOREACH(inp, head, inp_hash) {
 #if INET6
                LIST_FOREACH(inp, head, inp_hash) {
 #if INET6
-                       if ((inp->inp_vflag & INP_IPV4) == 0)
+                       if (!(inp->inp_vflag & INP_IPV4))
                                continue;
                                continue;
-#endif
+#endif /* INET6 */
                        if (inp->inp_faddr.s_addr == INADDR_ANY &&
                            inp->inp_laddr.s_addr == laddr.s_addr &&
                            inp->inp_lport == lport) {
                        if (inp->inp_faddr.s_addr == INADDR_ANY &&
                            inp->inp_laddr.s_addr == laddr.s_addr &&
                            inp->inp_lport == lport) {
@@ -1146,7 +2002,7 @@ in_pcblookup_local(struct inpcbinfo *pcbinfo, struct in_addr laddr,
                /*
                 * Not found.
                 */
                /*
                 * Not found.
                 */
-               KERNEL_DEBUG(DBG_FNC_PCB_LOOKUP | DBG_FUNC_END, 0,0,0,0,0);
+               KERNEL_DEBUG(DBG_FNC_PCB_LOOKUP | DBG_FUNC_END, 0, 0, 0, 0, 0);
                return (NULL);
        } else {
                struct inpcbporthead *porthash;
                return (NULL);
        } else {
                struct inpcbporthead *porthash;
@@ -1158,8 +2014,8 @@ in_pcblookup_local(struct inpcbinfo *pcbinfo, struct in_addr laddr,
                 * First see if this local port is in use by looking on the
                 * port hash list.
                 */
                 * First see if this local port is in use by looking on the
                 * port hash list.
                 */
-               porthash = &pcbinfo->porthashbase[INP_PCBPORTHASH(lport,
-                   pcbinfo->porthashmask)];
+               porthash = &pcbinfo->ipi_porthashbase[INP_PCBPORTHASH(lport,
+                   pcbinfo->ipi_porthashmask)];
                LIST_FOREACH(phd, porthash, phd_hash) {
                        if (phd->phd_port == lport)
                                break;
                LIST_FOREACH(phd, porthash, phd_hash) {
                        if (phd->phd_port == lport)
                                break;
@@ -1172,15 +2028,16 @@ in_pcblookup_local(struct inpcbinfo *pcbinfo, struct in_addr laddr,
                        LIST_FOREACH(inp, &phd->phd_pcblist, inp_portlist) {
                                wildcard = 0;
 #if INET6
                        LIST_FOREACH(inp, &phd->phd_pcblist, inp_portlist) {
                                wildcard = 0;
 #if INET6
-                               if ((inp->inp_vflag & INP_IPV4) == 0)
+                               if (!(inp->inp_vflag & INP_IPV4))
                                        continue;
                                        continue;
-#endif
+#endif /* INET6 */
                                if (inp->inp_faddr.s_addr != INADDR_ANY)
                                        wildcard++;
                                if (inp->inp_laddr.s_addr != INADDR_ANY) {
                                        if (laddr.s_addr == INADDR_ANY)
                                                wildcard++;
                                if (inp->inp_faddr.s_addr != INADDR_ANY)
                                        wildcard++;
                                if (inp->inp_laddr.s_addr != INADDR_ANY) {
                                        if (laddr.s_addr == INADDR_ANY)
                                                wildcard++;
-                                       else if (inp->inp_laddr.s_addr != laddr.s_addr)
+                                       else if (inp->inp_laddr.s_addr !=
+                                           laddr.s_addr)
                                                continue;
                                } else {
                                        if (laddr.s_addr != INADDR_ANY)
                                                continue;
                                } else {
                                        if (laddr.s_addr != INADDR_ANY)
@@ -1195,194 +2052,356 @@ in_pcblookup_local(struct inpcbinfo *pcbinfo, struct in_addr laddr,
                                }
                        }
                }
                                }
                        }
                }
-               KERNEL_DEBUG(DBG_FNC_PCB_LOOKUP | DBG_FUNC_END, match,0,0,0,0);
+               KERNEL_DEBUG(DBG_FNC_PCB_LOOKUP | DBG_FUNC_END, match,
+                   0, 0, 0, 0);
                return (match);
        }
 }
 
 /*
                return (match);
        }
 }
 
 /*
- * Lookup PCB in hash list.
+ * Check if PCB exists in hash list.
  */
  */
-struct inpcb *
-in_pcblookup_hash(
-       struct inpcbinfo *pcbinfo,
-       struct in_addr faddr,
-       u_int fport_arg,
-       struct in_addr laddr,
-       u_int lport_arg,
-       int wildcard,
-       __unused struct ifnet *ifp)
+int
+in_pcblookup_hash_exists(struct inpcbinfo *pcbinfo, struct in_addr faddr,
+    u_int fport_arg, struct in_addr laddr, u_int lport_arg, int wildcard,
+    uid_t *uid, gid_t *gid, struct ifnet *ifp)
 {
        struct inpcbhead *head;
        struct inpcb *inp;
        u_short fport = fport_arg, lport = lport_arg;
 {
        struct inpcbhead *head;
        struct inpcb *inp;
        u_short fport = fport_arg, lport = lport_arg;
+       int found = 0;
+       struct inpcb *local_wild = NULL;
+#if INET6
+       struct inpcb *local_wild_mapped = NULL;
+#endif /* INET6 */
+
+       *uid = UID_MAX;
+       *gid = GID_MAX;
 
        /*
         * We may have found the pcb in the last lookup - check this first.
         */
 
 
        /*
         * We may have found the pcb in the last lookup - check this first.
         */
 
-       lck_rw_lock_shared(pcbinfo->mtx);
+       lck_rw_lock_shared(pcbinfo->ipi_lock);
 
        /*
         * First look for an exact match.
         */
 
        /*
         * First look for an exact match.
         */
-       head = &pcbinfo->hashbase[INP_PCBHASH(faddr.s_addr, lport, fport, pcbinfo->hashmask)];
+       head = &pcbinfo->ipi_hashbase[INP_PCBHASH(faddr.s_addr, lport, fport,
+           pcbinfo->ipi_hashmask)];
        LIST_FOREACH(inp, head, inp_hash) {
 #if INET6
        LIST_FOREACH(inp, head, inp_hash) {
 #if INET6
-               if ((inp->inp_vflag & INP_IPV4) == 0)
+               if (!(inp->inp_vflag & INP_IPV4))
                        continue;
                        continue;
-#endif
+#endif /* INET6 */
+               if (inp_restricted_recv(inp, ifp))
+                       continue;
+
                if (inp->inp_faddr.s_addr == faddr.s_addr &&
                    inp->inp_laddr.s_addr == laddr.s_addr &&
                    inp->inp_fport == fport &&
                    inp->inp_lport == lport) {
                if (inp->inp_faddr.s_addr == faddr.s_addr &&
                    inp->inp_laddr.s_addr == laddr.s_addr &&
                    inp->inp_fport == fport &&
                    inp->inp_lport == lport) {
-                       /*
-                        * Found.
-                        */
-                       if (in_pcb_checkstate(inp, WNT_ACQUIRE, 0) != WNT_STOPUSING) {
-                               lck_rw_done(pcbinfo->mtx);
-                               return (inp);
-                       }
-                       else {  /* it's there but dead, say it isn't found */
-                               lck_rw_done(pcbinfo->mtx);      
-                               return(NULL);
+                       if ((found = (inp->inp_socket != NULL))) {
+                               /*
+                                * Found.
+                                */
+                               *uid = kauth_cred_getuid(
+                                   inp->inp_socket->so_cred);
+                               *gid = kauth_cred_getgid(
+                                   inp->inp_socket->so_cred);
                        }
                        }
+                       lck_rw_done(pcbinfo->ipi_lock);
+                       return (found);
                }
        }
                }
        }
-       if (wildcard) {
-               struct inpcb *local_wild = NULL;
-#if INET6
-               struct inpcb *local_wild_mapped = NULL;
-#endif
 
 
-               head = &pcbinfo->hashbase[INP_PCBHASH(INADDR_ANY, lport, 0, pcbinfo->hashmask)];
-               LIST_FOREACH(inp, head, inp_hash) {
+       if (!wildcard) {
+               /*
+                * Not found.
+                */
+               lck_rw_done(pcbinfo->ipi_lock);
+               return (0);
+       }
+
+       head = &pcbinfo->ipi_hashbase[INP_PCBHASH(INADDR_ANY, lport, 0,
+           pcbinfo->ipi_hashmask)];
+       LIST_FOREACH(inp, head, inp_hash) {
 #if INET6
 #if INET6
-                       if ((inp->inp_vflag & INP_IPV4) == 0)
-                               continue;
-#endif
-                       if (inp->inp_faddr.s_addr == INADDR_ANY &&
-                           inp->inp_lport == lport) {
-#if defined(NFAITH) && NFAITH > 0
-                               if (ifp && ifp->if_type == IFT_FAITH &&
-                                   (inp->inp_flags & INP_FAITH) == 0)
-                                       continue;
-#endif
-                               if (inp->inp_laddr.s_addr == laddr.s_addr) {
-                                       if (in_pcb_checkstate(inp, WNT_ACQUIRE, 0) != WNT_STOPUSING) {
-                                               lck_rw_done(pcbinfo->mtx);
-                                               return (inp);
-                                       }
-                                       else {  /* it's there but dead, say it isn't found */
-                                               lck_rw_done(pcbinfo->mtx);      
-                                               return(NULL);
-                                       }
+               if (!(inp->inp_vflag & INP_IPV4))
+                       continue;
+#endif /* INET6 */
+               if (inp_restricted_recv(inp, ifp))
+                       continue;
+
+               if (inp->inp_faddr.s_addr == INADDR_ANY &&
+                   inp->inp_lport == lport) {
+                       if (inp->inp_laddr.s_addr == laddr.s_addr) {
+                               if ((found = (inp->inp_socket != NULL))) {
+                                       *uid = kauth_cred_getuid(
+                                           inp->inp_socket->so_cred);
+                                       *gid = kauth_cred_getgid(
+                                           inp->inp_socket->so_cred);
                                }
                                }
-                               else if (inp->inp_laddr.s_addr == INADDR_ANY) {
+                               lck_rw_done(pcbinfo->ipi_lock);
+                               return (found);
+                       } else if (inp->inp_laddr.s_addr == INADDR_ANY) {
 #if INET6
 #if INET6
-                                       if (INP_CHECK_SOCKAF(inp->inp_socket,
-                                                            AF_INET6))
-                                               local_wild_mapped = inp;
-                                       else
+                               if (inp->inp_socket &&
+                                   SOCK_CHECK_DOM(inp->inp_socket, PF_INET6))
+                                       local_wild_mapped = inp;
+                               else
 #endif /* INET6 */
                                        local_wild = inp;
 #endif /* INET6 */
                                        local_wild = inp;
-                               }
                        }
                }
                        }
                }
-               if (local_wild == NULL) {
+       }
+       if (local_wild == NULL) {
 #if INET6
 #if INET6
-                       if (local_wild_mapped != NULL) {
-                               if (in_pcb_checkstate(local_wild_mapped, WNT_ACQUIRE, 0) != WNT_STOPUSING) {
-                                       lck_rw_done(pcbinfo->mtx);
-                                       return (local_wild_mapped);
-                               }
-                               else {  /* it's there but dead, say it isn't found */
-                                       lck_rw_done(pcbinfo->mtx);      
-                                       return(NULL);
-                               }
+               if (local_wild_mapped != NULL) {
+                       if ((found = (local_wild_mapped->inp_socket != NULL))) {
+                               *uid = kauth_cred_getuid(
+                                   local_wild_mapped->inp_socket->so_cred);
+                               *gid = kauth_cred_getgid(
+                                   local_wild_mapped->inp_socket->so_cred);
                        }
                        }
-#endif /* INET6 */
-                       lck_rw_done(pcbinfo->mtx);
-                       return (NULL);
-               }
-               if (in_pcb_checkstate(local_wild, WNT_ACQUIRE, 0) != WNT_STOPUSING) {
-                       lck_rw_done(pcbinfo->mtx);
-                       return (local_wild);
-               }
-               else {  /* it's there but dead, say it isn't found */
-                       lck_rw_done(pcbinfo->mtx);      
-                       return(NULL);
+                       lck_rw_done(pcbinfo->ipi_lock);
+                       return (found);
                }
                }
+#endif /* INET6 */
+               lck_rw_done(pcbinfo->ipi_lock);
+               return (0);
        }
        }
-
-       /*
-        * Not found.
-        */
-       lck_rw_done(pcbinfo->mtx);
-       return (NULL);
+       if ((found = (local_wild->inp_socket != NULL))) {
+               *uid = kauth_cred_getuid(
+                   local_wild->inp_socket->so_cred);
+               *gid = kauth_cred_getgid(
+                   local_wild->inp_socket->so_cred);
+       }
+       lck_rw_done(pcbinfo->ipi_lock);
+       return (found);
 }
 
 /*
 }
 
 /*
- * Insert PCB onto various hash lists.
+ * Lookup PCB in hash list.
  */
  */
-int
-in_pcbinshash(struct inpcb *inp, int locked)
+struct inpcb *
+in_pcblookup_hash(struct inpcbinfo *pcbinfo, struct in_addr faddr,
+    u_int fport_arg, struct in_addr laddr, u_int lport_arg, int wildcard,
+    struct ifnet *ifp)
 {
 {
-       struct inpcbhead *pcbhash;
-       struct inpcbporthead *pcbporthash;
-       struct inpcbinfo *pcbinfo = inp->inp_pcbinfo;
-       struct inpcbport *phd;
-       u_int32_t hashkey_faddr;
-
-        if (!locked) {
-                if (!lck_rw_try_lock_exclusive(pcbinfo->mtx)) {
-                /*lock inversion issue, mostly with udp multicast packets */
-                        socket_unlock(inp->inp_socket, 0);
-                        lck_rw_lock_exclusive(pcbinfo->mtx);
-                        socket_lock(inp->inp_socket, 0);
-                }
-        }
-
+       struct inpcbhead *head;
+       struct inpcb *inp;
+       u_short fport = fport_arg, lport = lport_arg;
+       struct inpcb *local_wild = NULL;
 #if INET6
 #if INET6
-       if (inp->inp_vflag & INP_IPV6)
-               hashkey_faddr = inp->in6p_faddr.s6_addr32[3] /* XXX */;
-       else
+       struct inpcb *local_wild_mapped = NULL;
 #endif /* INET6 */
 #endif /* INET6 */
-       hashkey_faddr = inp->inp_faddr.s_addr;
-
-       inp->hash_element = INP_PCBHASH(hashkey_faddr, inp->inp_lport, inp->inp_fport, pcbinfo->hashmask);
-
-       pcbhash = &pcbinfo->hashbase[inp->hash_element];
-
-       pcbporthash = &pcbinfo->porthashbase[INP_PCBPORTHASH(inp->inp_lport,
-           pcbinfo->porthashmask)];
 
        /*
 
        /*
-        * Go through port list and look for a head for this lport.
+        * We may have found the pcb in the last lookup - check this first.
         */
         */
-       LIST_FOREACH(phd, pcbporthash, phd_hash) {
-               if (phd->phd_port == inp->inp_lport)
-                       break;
-       }
+
+       lck_rw_lock_shared(pcbinfo->ipi_lock);
+
        /*
        /*
-        * If none exists, malloc one and tack it on.
+        * First look for an exact match.
         */
         */
-       if (phd == NULL) {
-               MALLOC(phd, struct inpcbport *, sizeof(struct inpcbport), M_PCB, M_WAITOK);
-               if (phd == NULL) {
-                       if (!locked)
-                               lck_rw_done(pcbinfo->mtx);
-                       return (ENOBUFS); /* XXX */
-               }
-               phd->phd_port = inp->inp_lport;
-               LIST_INIT(&phd->phd_pcblist);
-               LIST_INSERT_HEAD(pcbporthash, phd, phd_hash);
-       }
-       inp->inp_phd = phd;
-       LIST_INSERT_HEAD(&phd->phd_pcblist, inp, inp_portlist);
+       head = &pcbinfo->ipi_hashbase[INP_PCBHASH(faddr.s_addr, lport, fport,
+           pcbinfo->ipi_hashmask)];
+       LIST_FOREACH(inp, head, inp_hash) {
+#if INET6
+               if (!(inp->inp_vflag & INP_IPV4))
+                       continue;
+#endif /* INET6 */
+               if (inp_restricted_recv(inp, ifp))
+                       continue;
+
+               if (inp->inp_faddr.s_addr == faddr.s_addr &&
+                   inp->inp_laddr.s_addr == laddr.s_addr &&
+                   inp->inp_fport == fport &&
+                   inp->inp_lport == lport) {
+                       /*
+                        * Found.
+                        */
+                       if (in_pcb_checkstate(inp, WNT_ACQUIRE, 0) !=
+                           WNT_STOPUSING) {
+                               lck_rw_done(pcbinfo->ipi_lock);
+                               return (inp);
+                       } else {
+                               /* it's there but dead, say it isn't found */
+                               lck_rw_done(pcbinfo->ipi_lock);
+                               return (NULL);
+                       }
+               }
+       }
+
+       if (!wildcard) {
+               /*
+                * Not found.
+                */
+               lck_rw_done(pcbinfo->ipi_lock);
+               return (NULL);
+       }
+
+       head = &pcbinfo->ipi_hashbase[INP_PCBHASH(INADDR_ANY, lport, 0,
+           pcbinfo->ipi_hashmask)];
+       LIST_FOREACH(inp, head, inp_hash) {
+#if INET6
+               if (!(inp->inp_vflag & INP_IPV4))
+                       continue;
+#endif /* INET6 */
+               if (inp_restricted_recv(inp, ifp))
+                       continue;
+
+               if (inp->inp_faddr.s_addr == INADDR_ANY &&
+                   inp->inp_lport == lport) {
+                       if (inp->inp_laddr.s_addr == laddr.s_addr) {
+                               if (in_pcb_checkstate(inp, WNT_ACQUIRE, 0) !=
+                                   WNT_STOPUSING) {
+                                       lck_rw_done(pcbinfo->ipi_lock);
+                                       return (inp);
+                               } else {
+                                       /* it's dead; say it isn't found */
+                                       lck_rw_done(pcbinfo->ipi_lock);
+                                       return (NULL);
+                               }
+                       } else if (inp->inp_laddr.s_addr == INADDR_ANY) {
+#if INET6
+                               if (SOCK_CHECK_DOM(inp->inp_socket, PF_INET6))
+                                       local_wild_mapped = inp;
+                               else
+#endif /* INET6 */
+                                       local_wild = inp;
+                       }
+               }
+       }
+       if (local_wild == NULL) {
+#if INET6
+               if (local_wild_mapped != NULL) {
+                       if (in_pcb_checkstate(local_wild_mapped,
+                           WNT_ACQUIRE, 0) != WNT_STOPUSING) {
+                               lck_rw_done(pcbinfo->ipi_lock);
+                               return (local_wild_mapped);
+                       } else {
+                               /* it's dead; say it isn't found */
+                               lck_rw_done(pcbinfo->ipi_lock);
+                               return (NULL);
+                       }
+               }
+#endif /* INET6 */
+               lck_rw_done(pcbinfo->ipi_lock);
+               return (NULL);
+       }
+       if (in_pcb_checkstate(local_wild, WNT_ACQUIRE, 0) != WNT_STOPUSING) {
+               lck_rw_done(pcbinfo->ipi_lock);
+               return (local_wild);
+       }
+       /*
+        * It's either not found or is already dead.
+        */
+       lck_rw_done(pcbinfo->ipi_lock);
+       return (NULL);
+}
+
+/*
+ * @brief      Insert PCB onto various hash lists.
+ *
+ * @param      inp Pointer to internet protocol control block
+ * @param      locked  Implies if ipi_lock (protecting pcb list)
+ *             is already locked or not.
+ *
+ * @return     int error on failure and 0 on success
+ */
+int
+in_pcbinshash(struct inpcb *inp, int locked)
+{
+       struct inpcbhead *pcbhash;
+       struct inpcbporthead *pcbporthash;
+       struct inpcbinfo *pcbinfo = inp->inp_pcbinfo;
+       struct inpcbport *phd;
+       u_int32_t hashkey_faddr;
+
+       if (!locked) {
+               if (!lck_rw_try_lock_exclusive(pcbinfo->ipi_lock)) {
+                       /*
+                        * Lock inversion issue, mostly with udp
+                        * multicast packets
+                        */
+                       socket_unlock(inp->inp_socket, 0);
+                       lck_rw_lock_exclusive(pcbinfo->ipi_lock);
+                       socket_lock(inp->inp_socket, 0);
+               }
+       }
+
+       /*
+        * This routine or its caller may have given up
+        * socket's protocol lock briefly.
+        * During that time the socket may have been dropped.
+        * Safe-guarding against that.
+        */
+       if (inp->inp_state == INPCB_STATE_DEAD) {
+               if (!locked) {
+                       lck_rw_done(pcbinfo->ipi_lock);
+               }
+               return (ECONNABORTED);
+       }
+
+
+#if INET6
+       if (inp->inp_vflag & INP_IPV6)
+               hashkey_faddr = inp->in6p_faddr.s6_addr32[3] /* XXX */;
+       else
+#endif /* INET6 */
+               hashkey_faddr = inp->inp_faddr.s_addr;
+
+       inp->inp_hash_element = INP_PCBHASH(hashkey_faddr, inp->inp_lport,
+           inp->inp_fport, pcbinfo->ipi_hashmask);
+
+       pcbhash = &pcbinfo->ipi_hashbase[inp->inp_hash_element];
+
+       pcbporthash = &pcbinfo->ipi_porthashbase[INP_PCBPORTHASH(inp->inp_lport,
+           pcbinfo->ipi_porthashmask)];
+
+       /*
+        * Go through port list and look for a head for this lport.
+        */
+       LIST_FOREACH(phd, pcbporthash, phd_hash) {
+               if (phd->phd_port == inp->inp_lport)
+                       break;
+       }
+
+       /*
+        * If none exists, malloc one and tack it on.
+        */
+       if (phd == NULL) {
+               MALLOC(phd, struct inpcbport *, sizeof (struct inpcbport),
+                   M_PCB, M_WAITOK);
+               if (phd == NULL) {
+                       if (!locked)
+                               lck_rw_done(pcbinfo->ipi_lock);
+                       return (ENOBUFS); /* XXX */
+               }
+               phd->phd_port = inp->inp_lport;
+               LIST_INIT(&phd->phd_pcblist);
+               LIST_INSERT_HEAD(pcbporthash, phd, phd_hash);
+       }
+
+       VERIFY(!(inp->inp_flags2 & INP2_INHASHLIST));
+
+
+       inp->inp_phd = phd;
+       LIST_INSERT_HEAD(&phd->phd_pcblist, inp, inp_portlist);
        LIST_INSERT_HEAD(pcbhash, inp, inp_hash);
        LIST_INSERT_HEAD(pcbhash, inp, inp_hash);
+       inp->inp_flags2 |= INP2_INHASHLIST;
+
        if (!locked)
        if (!locked)
-               lck_rw_done(pcbinfo->mtx);
+               lck_rw_done(pcbinfo->ipi_lock);
+
+#if NECP
+       // This call catches the original setting of the local address
+       inp_update_necp_policy(inp, NULL, NULL, 0);
+#endif /* NECP */
+
        return (0);
 }
 
        return (0);
 }
 
@@ -1403,129 +2422,185 @@ in_pcbrehash(struct inpcb *inp)
                hashkey_faddr = inp->in6p_faddr.s6_addr32[3] /* XXX */;
        else
 #endif /* INET6 */
                hashkey_faddr = inp->in6p_faddr.s6_addr32[3] /* XXX */;
        else
 #endif /* INET6 */
-       hashkey_faddr = inp->inp_faddr.s_addr;
-       inp->hash_element = INP_PCBHASH(hashkey_faddr, inp->inp_lport, 
-                               inp->inp_fport, inp->inp_pcbinfo->hashmask);
-       head = &inp->inp_pcbinfo->hashbase[inp->hash_element];
+               hashkey_faddr = inp->inp_faddr.s_addr;
+
+       inp->inp_hash_element = INP_PCBHASH(hashkey_faddr, inp->inp_lport,
+           inp->inp_fport, inp->inp_pcbinfo->ipi_hashmask);
+       head = &inp->inp_pcbinfo->ipi_hashbase[inp->inp_hash_element];
+
+       if (inp->inp_flags2 & INP2_INHASHLIST) {
+               LIST_REMOVE(inp, inp_hash);
+               inp->inp_flags2 &= ~INP2_INHASHLIST;
+       }
 
 
-       LIST_REMOVE(inp, inp_hash);
+       VERIFY(!(inp->inp_flags2 & INP2_INHASHLIST));
        LIST_INSERT_HEAD(head, inp, inp_hash);
        LIST_INSERT_HEAD(head, inp, inp_hash);
+       inp->inp_flags2 |= INP2_INHASHLIST;
+
+#if NECP
+       // This call catches updates to the remote addresses
+       inp_update_necp_policy(inp, NULL, NULL, 0);
+#endif /* NECP */
 }
 
 /*
  * Remove PCB from various lists.
 }
 
 /*
  * Remove PCB from various lists.
+ * Must be called pcbinfo lock is held in exclusive mode.
  */
  */
-//###LOCK must be called with list lock held
 void
 in_pcbremlists(struct inpcb *inp)
 {
        inp->inp_gencnt = ++inp->inp_pcbinfo->ipi_gencnt;
 
 void
 in_pcbremlists(struct inpcb *inp)
 {
        inp->inp_gencnt = ++inp->inp_pcbinfo->ipi_gencnt;
 
-       if (inp->inp_lport) {
+       /*
+        * Check if it's in hashlist -- an inp is placed in hashlist when
+        * it's local port gets assigned. So it should also be present
+        * in the port list.
+        */
+       if (inp->inp_flags2 & INP2_INHASHLIST) {
                struct inpcbport *phd = inp->inp_phd;
 
                struct inpcbport *phd = inp->inp_phd;
 
+               VERIFY(phd != NULL && inp->inp_lport > 0);
+
                LIST_REMOVE(inp, inp_hash);
                LIST_REMOVE(inp, inp_hash);
+               inp->inp_hash.le_next = NULL;
+               inp->inp_hash.le_prev = NULL;
+
                LIST_REMOVE(inp, inp_portlist);
                LIST_REMOVE(inp, inp_portlist);
-               if (phd != NULL && (LIST_FIRST(&phd->phd_pcblist) == NULL)) {
+               inp->inp_portlist.le_next = NULL;
+               inp->inp_portlist.le_prev = NULL;
+               if (LIST_EMPTY(&phd->phd_pcblist)) {
                        LIST_REMOVE(phd, phd_hash);
                        FREE(phd, M_PCB);
                }
                        LIST_REMOVE(phd, phd_hash);
                        FREE(phd, M_PCB);
                }
+               inp->inp_phd = NULL;
+               inp->inp_flags2 &= ~INP2_INHASHLIST;
+       }
+       VERIFY(!(inp->inp_flags2 & INP2_INHASHLIST));
+
+       if (inp->inp_flags2 & INP2_TIMEWAIT) {
+               /* Remove from time-wait queue */
+               tcp_remove_from_time_wait(inp);
+               inp->inp_flags2 &= ~INP2_TIMEWAIT;
+               VERIFY(inp->inp_pcbinfo->ipi_twcount != 0);
+               inp->inp_pcbinfo->ipi_twcount--;
+       } else {
+               /* Remove from global inp list if it is not time-wait */
+               LIST_REMOVE(inp, inp_list);
        }
        }
-       LIST_REMOVE(inp, inp_list);
+
+       if (inp->inp_flags2 & INP2_IN_FCTREE) {
+               inp_fc_getinp(inp->inp_flowhash, (INPFC_SOLOCKED|INPFC_REMOVE));
+               VERIFY(!(inp->inp_flags2 & INP2_IN_FCTREE));
+       }
+
        inp->inp_pcbinfo->ipi_count--;
 }
 
        inp->inp_pcbinfo->ipi_count--;
 }
 
-/* Mechanism used to defer the memory release of PCBs
- * The pcb list will contain the pcb until the ripper can clean it up if
- * the following conditions are met: 1) state "DEAD", 2) wantcnt is STOPUSING
- * 3) usecount is null
+/*
+ * Mechanism used to defer the memory release of PCBs
+ * The pcb list will contain the pcb until the reaper can clean it up if
+ * the following conditions are met:
+ *     1) state "DEAD",
+ *     2) wantcnt is STOPUSING
+ *     3) usecount is 0
  * This function will be called to either mark the pcb as
  * This function will be called to either mark the pcb as
-*/
+ */
 int
 in_pcb_checkstate(struct inpcb *pcb, int mode, int locked)
 {
 int
 in_pcb_checkstate(struct inpcb *pcb, int mode, int locked)
 {
-
-       volatile UInt32 *wantcnt        = (volatile UInt32 *)&pcb->inp_wantcnt;
+       volatile UInt32 *wantcnt = (volatile UInt32 *)&pcb->inp_wantcnt;
        UInt32 origwant;
        UInt32 newwant;
 
        switch (mode) {
        UInt32 origwant;
        UInt32 newwant;
 
        switch (mode) {
+       case WNT_STOPUSING:
+               /*
+                * Try to mark the pcb as ready for recycling.  CAS with
+                * STOPUSING, if success we're good, if it's in use, will
+                * be marked later
+                */
+               if (locked == 0)
+                       socket_lock(pcb->inp_socket, 1);
+               pcb->inp_state = INPCB_STATE_DEAD;
 
 
-               case WNT_STOPUSING:     /* try to mark the pcb as ready for recycling */
+stopusing:
+               if (pcb->inp_socket->so_usecount < 0) {
+                       panic("%s: pcb=%p so=%p usecount is negative\n",
+                           __func__, pcb, pcb->inp_socket);
+                       /* NOTREACHED */
+               }
+               if (locked == 0)
+                       socket_unlock(pcb->inp_socket, 1);
 
 
-                       /* compareswap with STOPUSING, if success we're good, if it's in use, will be marked later */
+               inpcb_gc_sched(pcb->inp_pcbinfo, INPCB_TIMER_FAST);
 
 
-                       if (locked == 0)
-                               socket_lock(pcb->inp_socket, 1);
-                       pcb->inp_state = INPCB_STATE_DEAD;
-stopusing:
-                       if (pcb->inp_socket->so_usecount < 0)
-                               panic("in_pcb_checkstate STOP pcb=%p so=%p usecount is negative\n", pcb, pcb->inp_socket);
-                       if (locked == 0)
-                               socket_unlock(pcb->inp_socket, 1);
+               origwant = *wantcnt;
+               if ((UInt16) origwant == 0xffff) /* should stop using */
+                       return (WNT_STOPUSING);
+               newwant = 0xffff;
+               if ((UInt16) origwant == 0) {
+                       /* try to mark it as unsuable now */
+                       OSCompareAndSwap(origwant, newwant, wantcnt);
+               }
+               return (WNT_STOPUSING);
 
 
+       case WNT_ACQUIRE:
+               /*
+                * Try to increase reference to pcb.  If WNT_STOPUSING
+                * should bail out.  If socket state DEAD, try to set count
+                * to STOPUSING, return failed otherwise increase cnt.
+                */
+               do {
                        origwant = *wantcnt;
                        origwant = *wantcnt;
-                       if ((UInt16) origwant == 0xffff ) /* should stop using */
+                       if ((UInt16) origwant == 0xffff) {
+                               /* should stop using */
                                return (WNT_STOPUSING);
                                return (WNT_STOPUSING);
-                       newwant = 0xffff;                       
-                       if ((UInt16) origwant == 0) {/* try to mark it as unsuable now */
-                               OSCompareAndSwap(origwant, newwant, wantcnt) ;
                        }
                        }
-                       return (WNT_STOPUSING);
-                       break;
-
-               case WNT_ACQUIRE:       /* try to increase reference to pcb */
-                                       /* if WNT_STOPUSING should bail out */
-                       /*
-                        * if socket state DEAD, try to set count to STOPUSING, return failed
-                        * otherwise increase cnt
-                        */
-                       do {
-                               origwant = *wantcnt;
-                               if ((UInt16) origwant == 0xffff ) {/* should stop using */
-//                                     printf("in_pcb_checkstate: ACQ PCB was STOPUSING while release. odd pcb=%p\n", pcb);
-                                       return (WNT_STOPUSING);
-                               }
-                               newwant = origwant + 1;         
-                       } while (!OSCompareAndSwap(origwant, newwant, wantcnt));
-                       return (WNT_ACQUIRE);
-                       break;
-
-               case WNT_RELEASE:       /* release reference. if result is null and pcb state is DEAD,
-                                          set wanted bit to STOPUSING
-                                        */
+                       newwant = origwant + 1;
+               } while (!OSCompareAndSwap(origwant, newwant, wantcnt));
+               return (WNT_ACQUIRE);
 
 
-                       if (locked == 0)
-                               socket_lock(pcb->inp_socket, 1);
+       case WNT_RELEASE:
+               /*
+                * Release reference.  If result is null and pcb state
+                * is DEAD, set wanted bit to STOPUSING
+                */
+               if (locked == 0)
+                       socket_lock(pcb->inp_socket, 1);
 
 
-                       do {
-                               origwant = *wantcnt;
-                               if ((UInt16) origwant == 0x0 ) 
-                                       panic("in_pcb_checkstate pcb=%p release with zero count", pcb);
-                               if ((UInt16) origwant == 0xffff ) {/* should stop using */
-#if TEMPDEBUG
-                                       printf("in_pcb_checkstate: REL PCB was STOPUSING while release. odd pcb=%p\n", pcb);
-#endif
-                                       if (locked == 0)
-                                               socket_unlock(pcb->inp_socket, 1);
-                                       return (WNT_STOPUSING);
-                               }
-                               newwant = origwant - 1;         
-                       } while (!OSCompareAndSwap(origwant, newwant, wantcnt));
-
-                       if (pcb->inp_state == INPCB_STATE_DEAD) 
-                               goto stopusing;
-                       if (pcb->inp_socket->so_usecount < 0)
-                               panic("in_pcb_checkstate RELEASE pcb=%p so=%p usecount is negative\n", pcb, pcb->inp_socket);
-                               
-                       if (locked == 0)
-                               socket_unlock(pcb->inp_socket, 1);
-                       return (WNT_RELEASE);
-                       break;
+               do {
+                       origwant = *wantcnt;
+                       if ((UInt16) origwant == 0x0) {
+                               panic("%s: pcb=%p release with zero count",
+                                   __func__, pcb);
+                               /* NOTREACHED */
+                       }
+                       if ((UInt16) origwant == 0xffff) {
+                               /* should stop using */
+                               if (locked == 0)
+                                       socket_unlock(pcb->inp_socket, 1);
+                               return (WNT_STOPUSING);
+                       }
+                       newwant = origwant - 1;
+               } while (!OSCompareAndSwap(origwant, newwant, wantcnt));
+
+               if (pcb->inp_state == INPCB_STATE_DEAD)
+                       goto stopusing;
+               if (pcb->inp_socket->so_usecount < 0) {
+                       panic("%s: RELEASE pcb=%p so=%p usecount is negative\n",
+                           __func__, pcb, pcb->inp_socket);
+                       /* NOTREACHED */
+               }
 
 
-               default:
+               if (locked == 0)
+                       socket_unlock(pcb->inp_socket, 1);
+               return (WNT_RELEASE);
 
 
-                       panic("in_pcb_checkstate: so=%p not a valid state =%x\n", pcb->inp_socket, mode);
+       default:
+               panic("%s: so=%p not a valid state =%x\n", __func__,
+                   pcb->inp_socket, mode);
+               /* NOTREACHED */
        }
 
        /* NOTREACHED */
        }
 
        /* NOTREACHED */
@@ -1538,54 +2613,50 @@ stopusing:
  * not change. We intentionally avoid copying pointers.
  */
 void
  * not change. We intentionally avoid copying pointers.
  */
 void
-inpcb_to_compat(
-       struct inpcb *inp,
-       struct inpcb_compat *inp_compat)
+inpcb_to_compat(struct inpcb *inp, struct inpcb_compat *inp_compat)
 {
 {
-       bzero(inp_compat, sizeof(*inp_compat));
+       bzero(inp_compat, sizeof (*inp_compat));
        inp_compat->inp_fport = inp->inp_fport;
        inp_compat->inp_lport = inp->inp_lport;
        inp_compat->inp_fport = inp->inp_fport;
        inp_compat->inp_lport = inp->inp_lport;
-       inp_compat->nat_owner = inp->nat_owner;
-       inp_compat->nat_cookie = inp->nat_cookie;
+       inp_compat->nat_owner = 0;
+       inp_compat->nat_cookie = 0;
        inp_compat->inp_gencnt = inp->inp_gencnt;
        inp_compat->inp_flags = inp->inp_flags;
        inp_compat->inp_flow = inp->inp_flow;
        inp_compat->inp_vflag = inp->inp_vflag;
        inp_compat->inp_ip_ttl = inp->inp_ip_ttl;
        inp_compat->inp_ip_p = inp->inp_ip_p;
        inp_compat->inp_gencnt = inp->inp_gencnt;
        inp_compat->inp_flags = inp->inp_flags;
        inp_compat->inp_flow = inp->inp_flow;
        inp_compat->inp_vflag = inp->inp_vflag;
        inp_compat->inp_ip_ttl = inp->inp_ip_ttl;
        inp_compat->inp_ip_p = inp->inp_ip_p;
-       inp_compat->inp_dependfaddr.inp6_foreign = inp->inp_dependfaddr.inp6_foreign;
-       inp_compat->inp_dependladdr.inp6_local = inp->inp_dependladdr.inp6_local;
+       inp_compat->inp_dependfaddr.inp6_foreign =
+           inp->inp_dependfaddr.inp6_foreign;
+       inp_compat->inp_dependladdr.inp6_local =
+           inp->inp_dependladdr.inp6_local;
        inp_compat->inp_depend4.inp4_ip_tos = inp->inp_depend4.inp4_ip_tos;
        inp_compat->inp_depend4.inp4_ip_tos = inp->inp_depend4.inp4_ip_tos;
-       inp_compat->inp_depend6.inp6_hlim = inp->inp_depend6.inp6_hlim;
+       inp_compat->inp_depend6.inp6_hlim = 0;
        inp_compat->inp_depend6.inp6_cksum = inp->inp_depend6.inp6_cksum;
        inp_compat->inp_depend6.inp6_cksum = inp->inp_depend6.inp6_cksum;
-       inp_compat->inp_depend6.inp6_ifindex = inp->inp_depend6.inp6_ifindex;
+       inp_compat->inp_depend6.inp6_ifindex = 0;
        inp_compat->inp_depend6.inp6_hops = inp->inp_depend6.inp6_hops;
 }
 
 #if !CONFIG_EMBEDDED
        inp_compat->inp_depend6.inp6_hops = inp->inp_depend6.inp6_hops;
 }
 
 #if !CONFIG_EMBEDDED
-
 void
 void
-inpcb_to_xinpcb64(
-        struct inpcb *inp,
-        struct xinpcb64 *xinp)
-{
-        xinp->inp_fport = inp->inp_fport;
-        xinp->inp_lport = inp->inp_lport;
-        xinp->inp_gencnt = inp->inp_gencnt;
-        xinp->inp_flags = inp->inp_flags;
-        xinp->inp_flow = inp->inp_flow;
-        xinp->inp_vflag = inp->inp_vflag;
-        xinp->inp_ip_ttl = inp->inp_ip_ttl;
-        xinp->inp_ip_p = inp->inp_ip_p;
-        xinp->inp_dependfaddr.inp6_foreign = inp->inp_dependfaddr.inp6_foreign;
-        xinp->inp_dependladdr.inp6_local = inp->inp_dependladdr.inp6_local;
-        xinp->inp_depend4.inp4_ip_tos = inp->inp_depend4.inp4_ip_tos;
-        xinp->inp_depend6.inp6_hlim = inp->inp_depend6.inp6_hlim;
-        xinp->inp_depend6.inp6_cksum = inp->inp_depend6.inp6_cksum;
-       xinp->inp_depend6.inp6_ifindex = inp->inp_depend6.inp6_ifindex;
-        xinp->inp_depend6.inp6_hops = inp->inp_depend6.inp6_hops;
+inpcb_to_xinpcb64(struct inpcb *inp, struct xinpcb64 *xinp)
+{
+       xinp->inp_fport = inp->inp_fport;
+       xinp->inp_lport = inp->inp_lport;
+       xinp->inp_gencnt = inp->inp_gencnt;
+       xinp->inp_flags = inp->inp_flags;
+       xinp->inp_flow = inp->inp_flow;
+       xinp->inp_vflag = inp->inp_vflag;
+       xinp->inp_ip_ttl = inp->inp_ip_ttl;
+       xinp->inp_ip_p = inp->inp_ip_p;
+       xinp->inp_dependfaddr.inp6_foreign = inp->inp_dependfaddr.inp6_foreign;
+       xinp->inp_dependladdr.inp6_local = inp->inp_dependladdr.inp6_local;
+       xinp->inp_depend4.inp4_ip_tos = inp->inp_depend4.inp4_ip_tos;
+       xinp->inp_depend6.inp6_hlim = 0;
+       xinp->inp_depend6.inp6_cksum = inp->inp_depend6.inp6_cksum;
+       xinp->inp_depend6.inp6_ifindex = 0;
+       xinp->inp_depend6.inp6_hops = inp->inp_depend6.inp6_hops;
 }
 }
-
 #endif /* !CONFIG_EMBEDDED */
 
 /*
 #endif /* !CONFIG_EMBEDDED */
 
 /*
@@ -1619,18 +2690,16 @@ inp_route_copyout(struct inpcb *inp, struct route *dst)
 {
        struct route *src = &inp->inp_route;
 
 {
        struct route *src = &inp->inp_route;
 
-       lck_mtx_assert(inp->inpcb_mtx, LCK_MTX_ASSERT_OWNED);
-
-       /* Minor sanity check */
-       if (src->ro_rt != NULL && rt_key(src->ro_rt)->sa_family != AF_INET)
-               panic("%s: wrong or corrupted route: %p", __func__, src);
+       socket_lock_assert_owned(inp->inp_socket);
 
 
-       /* Copy everything (rt, dst, flags) from PCB */
-       bcopy(src, dst, sizeof (*dst));
+       /*
+        * If the route in the PCB is stale or not for IPv4, blow it away;
+        * this is possible in the case of IPv4-mapped address case.
+        */
+       if (ROUTE_UNUSABLE(src) || rt_key(src->ro_rt)->sa_family != AF_INET)
+               ROUTE_RELEASE(src);
 
 
-       /* Hold one reference for the local copy of struct route */
-       if (dst->ro_rt != NULL)
-               RT_ADDREF(dst->ro_rt);
+       route_copyout(dst, src, sizeof (*dst));
 }
 
 void
 }
 
 void
@@ -1638,33 +2707,841 @@ inp_route_copyin(struct inpcb *inp, struct route *src)
 {
        struct route *dst = &inp->inp_route;
 
 {
        struct route *dst = &inp->inp_route;
 
-       lck_mtx_assert(inp->inpcb_mtx, LCK_MTX_ASSERT_OWNED);
+       socket_lock_assert_owned(inp->inp_socket);
 
        /* Minor sanity check */
        if (src->ro_rt != NULL && rt_key(src->ro_rt)->sa_family != AF_INET)
                panic("%s: wrong or corrupted route: %p", __func__, src);
 
 
        /* Minor sanity check */
        if (src->ro_rt != NULL && rt_key(src->ro_rt)->sa_family != AF_INET)
                panic("%s: wrong or corrupted route: %p", __func__, src);
 
-       /* No cached route in the PCB? */
-       if (dst->ro_rt == NULL) {
-               /*
-                * Copy everything (rt, dst, flags) from ip_output();
-                * the reference to the route was held at the time
-                * it was allocated and is kept intact.
-                */
-               bcopy(src, dst, sizeof (*dst));
-       } else if (src->ro_rt != NULL) {
+       route_copyin(src, dst, sizeof (*src));
+}
+
+/*
+ * Handler for setting IP_BOUND_IF/IPV6_BOUND_IF socket option.
+ */
+int
+inp_bindif(struct inpcb *inp, unsigned int ifscope, struct ifnet **pifp)
+{
+       struct ifnet *ifp = NULL;
+
+       ifnet_head_lock_shared();
+       if ((ifscope > (unsigned)if_index) || (ifscope != IFSCOPE_NONE &&
+           (ifp = ifindex2ifnet[ifscope]) == NULL)) {
+               ifnet_head_done();
+               return (ENXIO);
+       }
+       ifnet_head_done();
+
+       VERIFY(ifp != NULL || ifscope == IFSCOPE_NONE);
+
+       /*
+        * A zero interface scope value indicates an "unbind".
+        * Otherwise, take in whatever value the app desires;
+        * the app may already know the scope (or force itself
+        * to such a scope) ahead of time before the interface
+        * gets attached.  It doesn't matter either way; any
+        * route lookup from this point on will require an
+        * exact match for the embedded interface scope.
+        */
+       inp->inp_boundifp = ifp;
+       if (inp->inp_boundifp == NULL)
+               inp->inp_flags &= ~INP_BOUND_IF;
+       else
+               inp->inp_flags |= INP_BOUND_IF;
+
+       /* Blow away any cached route in the PCB */
+       ROUTE_RELEASE(&inp->inp_route);
+
+       if (pifp != NULL)
+               *pifp = ifp;
+
+       return (0);
+}
+
+/*
+ * Handler for setting IP_NO_IFT_CELLULAR/IPV6_NO_IFT_CELLULAR socket option,
+ * as well as for setting PROC_UUID_NO_CELLULAR policy.
+ */
+void
+inp_set_nocellular(struct inpcb *inp)
+{
+       inp->inp_flags |= INP_NO_IFT_CELLULAR;
+
+       /* Blow away any cached route in the PCB */
+       ROUTE_RELEASE(&inp->inp_route);
+}
+
+/*
+ * Handler for clearing IP_NO_IFT_CELLULAR/IPV6_NO_IFT_CELLULAR socket option,
+ * as well as for clearing PROC_UUID_NO_CELLULAR policy.
+ */
+void
+inp_clear_nocellular(struct inpcb *inp)
+{
+       struct socket *so = inp->inp_socket;
+
+       /*
+        * SO_RESTRICT_DENY_CELLULAR socket restriction issued on the socket
+        * has a higher precendence than INP_NO_IFT_CELLULAR.  Clear the flag
+        * if and only if the socket is unrestricted.
+        */
+       if (so != NULL && !(so->so_restrictions & SO_RESTRICT_DENY_CELLULAR)) {
+               inp->inp_flags &= ~INP_NO_IFT_CELLULAR;
+
+               /* Blow away any cached route in the PCB */
+               ROUTE_RELEASE(&inp->inp_route);
+       }
+}
+
+void
+inp_set_noexpensive(struct inpcb *inp)
+{
+       inp->inp_flags2 |= INP2_NO_IFF_EXPENSIVE;
+
+       /* Blow away any cached route in the PCB */
+       ROUTE_RELEASE(&inp->inp_route);
+}
+
+void
+inp_set_awdl_unrestricted(struct inpcb *inp)
+{
+       inp->inp_flags2 |= INP2_AWDL_UNRESTRICTED;
+
+       /* Blow away any cached route in the PCB */
+       ROUTE_RELEASE(&inp->inp_route);
+}
+
+boolean_t
+inp_get_awdl_unrestricted(struct inpcb *inp)
+{
+       return (inp->inp_flags2 & INP2_AWDL_UNRESTRICTED) ? TRUE : FALSE;
+}
+
+void
+inp_clear_awdl_unrestricted(struct inpcb *inp)
+{
+       inp->inp_flags2 &= ~INP2_AWDL_UNRESTRICTED;
+
+       /* Blow away any cached route in the PCB */
+       ROUTE_RELEASE(&inp->inp_route);
+}
+
+void
+inp_set_intcoproc_allowed(struct inpcb *inp)
+{
+       inp->inp_flags2 |= INP2_INTCOPROC_ALLOWED;
+
+       /* Blow away any cached route in the PCB */
+       ROUTE_RELEASE(&inp->inp_route);
+}
+
+boolean_t
+inp_get_intcoproc_allowed(struct inpcb *inp)
+{
+       return (inp->inp_flags2 & INP2_INTCOPROC_ALLOWED) ? TRUE : FALSE;
+}
+
+void
+inp_clear_intcoproc_allowed(struct inpcb *inp)
+{
+       inp->inp_flags2 &= ~INP2_INTCOPROC_ALLOWED;
+
+       /* Blow away any cached route in the PCB */
+       ROUTE_RELEASE(&inp->inp_route);
+}
+
+#if NECP
+/*
+ * Called when PROC_UUID_NECP_APP_POLICY is set.
+ */
+void
+inp_set_want_app_policy(struct inpcb *inp)
+{
+       inp->inp_flags2 |= INP2_WANT_APP_POLICY;
+}
+
+/*
+ * Called when PROC_UUID_NECP_APP_POLICY is cleared.
+ */
+void
+inp_clear_want_app_policy(struct inpcb *inp)
+{
+       inp->inp_flags2 &= ~INP2_WANT_APP_POLICY;
+}
+#endif /* NECP */
+
+/*
+ * Calculate flow hash for an inp, used by an interface to identify a
+ * flow. When an interface provides flow control advisory, this flow
+ * hash is used as an identifier.
+ */
+u_int32_t
+inp_calc_flowhash(struct inpcb *inp)
+{
+       struct inp_flowhash_key fh __attribute__((aligned(8)));
+       u_int32_t flowhash = 0;
+       struct inpcb *tmp_inp = NULL;
+
+       if (inp_hash_seed == 0)
+               inp_hash_seed = RandomULong();
+
+       bzero(&fh, sizeof (fh));
+
+       bcopy(&inp->inp_dependladdr, &fh.infh_laddr, sizeof (fh.infh_laddr));
+       bcopy(&inp->inp_dependfaddr, &fh.infh_faddr, sizeof (fh.infh_faddr));
+
+       fh.infh_lport = inp->inp_lport;
+       fh.infh_fport = inp->inp_fport;
+       fh.infh_af = (inp->inp_vflag & INP_IPV6) ? AF_INET6 : AF_INET;
+       fh.infh_proto = inp->inp_ip_p;
+       fh.infh_rand1 = RandomULong();
+       fh.infh_rand2 = RandomULong();
+
+try_again:
+       flowhash = net_flowhash(&fh, sizeof (fh), inp_hash_seed);
+       if (flowhash == 0) {
+               /* try to get a non-zero flowhash */
+               inp_hash_seed = RandomULong();
+               goto try_again;
+       }
+
+       inp->inp_flowhash = flowhash;
+
+       /* Insert the inp into inp_fc_tree */
+       lck_mtx_lock_spin(&inp_fc_lck);
+       tmp_inp = RB_FIND(inp_fc_tree, &inp_fc_tree, inp);
+       if (tmp_inp != NULL) {
                /*
                /*
-                * If the same, update just the ro_flags and ditch the one
-                * in the local copy.  Else ditch the one that is currently
-                * cached, and cache what we got back from ip_output().
+                * There is a different inp with the same flowhash.
+                * There can be a collision on flow hash but the
+                * probability is low.  Let's recompute the
+                * flowhash.
                 */
                 */
-               if (dst->ro_rt == src->ro_rt) {
-                       dst->ro_flags = src->ro_flags;
-                       rtfree(src->ro_rt);
-                       src->ro_rt = NULL;
+               lck_mtx_unlock(&inp_fc_lck);
+               /* recompute hash seed */
+               inp_hash_seed = RandomULong();
+               goto try_again;
+       }
+
+       RB_INSERT(inp_fc_tree, &inp_fc_tree, inp);
+       inp->inp_flags2 |= INP2_IN_FCTREE;
+       lck_mtx_unlock(&inp_fc_lck);
+
+       return (flowhash);
+}
+
+void
+inp_flowadv(uint32_t flowhash)
+{
+       struct inpcb *inp;
+
+       inp = inp_fc_getinp(flowhash, 0);
+
+       if (inp == NULL)
+               return;
+       inp_fc_feedback(inp);
+}
+
+/*
+ * Function to compare inp_fc_entries in inp flow control tree
+ */
+static inline int
+infc_cmp(const struct inpcb *inp1, const struct inpcb *inp2)
+{
+       return (memcmp(&(inp1->inp_flowhash), &(inp2->inp_flowhash),
+           sizeof(inp1->inp_flowhash)));
+}
+
+static struct inpcb *
+inp_fc_getinp(u_int32_t flowhash, u_int32_t flags)
+{
+       struct inpcb *inp = NULL;
+       int locked = (flags & INPFC_SOLOCKED) ? 1 : 0;
+
+       lck_mtx_lock_spin(&inp_fc_lck);
+       key_inp.inp_flowhash = flowhash;
+       inp = RB_FIND(inp_fc_tree, &inp_fc_tree, &key_inp);
+       if (inp == NULL) {
+               /* inp is not present, return */
+               lck_mtx_unlock(&inp_fc_lck);
+               return (NULL);
+       }
+
+       if (flags & INPFC_REMOVE) {
+               RB_REMOVE(inp_fc_tree, &inp_fc_tree, inp);
+               lck_mtx_unlock(&inp_fc_lck);
+
+               bzero(&(inp->infc_link), sizeof (inp->infc_link));
+               inp->inp_flags2 &= ~INP2_IN_FCTREE;
+               return (NULL);
+       }
+
+       if (in_pcb_checkstate(inp, WNT_ACQUIRE, locked) == WNT_STOPUSING)
+               inp = NULL;
+       lck_mtx_unlock(&inp_fc_lck);
+
+       return (inp);
+}
+
+static void
+inp_fc_feedback(struct inpcb *inp)
+{
+       struct socket *so = inp->inp_socket;
+
+       /* we already hold a want_cnt on this inp, socket can't be null */
+       VERIFY(so != NULL);
+       socket_lock(so, 1);
+
+       if (in_pcb_checkstate(inp, WNT_RELEASE, 1) == WNT_STOPUSING) {
+               socket_unlock(so, 1);
+               return;
+       }
+
+       if (inp->inp_sndinprog_cnt > 0)
+               inp->inp_flags |= INP_FC_FEEDBACK;
+
+       /*
+        * Return if the connection is not in flow-controlled state.
+        * This can happen if the connection experienced
+        * loss while it was in flow controlled state
+        */
+       if (!INP_WAIT_FOR_IF_FEEDBACK(inp)) {
+               socket_unlock(so, 1);
+               return;
+       }
+       inp_reset_fc_state(inp);
+
+       if (SOCK_TYPE(so) == SOCK_STREAM)
+               inp_fc_unthrottle_tcp(inp);
+
+       socket_unlock(so, 1);
+}
+
+void
+inp_reset_fc_state(struct inpcb *inp)
+{
+       struct socket *so = inp->inp_socket;
+       int suspended = (INP_IS_FLOW_SUSPENDED(inp)) ? 1 : 0;
+       int needwakeup = (INP_WAIT_FOR_IF_FEEDBACK(inp)) ? 1 : 0;
+
+       inp->inp_flags &= ~(INP_FLOW_CONTROLLED | INP_FLOW_SUSPENDED);
+
+       if (suspended) {
+               so->so_flags &= ~(SOF_SUSPENDED);
+               soevent(so, (SO_FILT_HINT_LOCKED | SO_FILT_HINT_RESUME));
+       }
+
+       /* Give a write wakeup to unblock the socket */
+       if (needwakeup)
+               sowwakeup(so);
+}
+
+int
+inp_set_fc_state(struct inpcb *inp, int advcode)
+{
+       struct inpcb *tmp_inp = NULL;
+       /*
+        * If there was a feedback from the interface when
+        * send operation was in progress, we should ignore
+        * this flow advisory to avoid a race between setting
+        * flow controlled state and receiving feedback from
+        * the interface
+        */
+       if (inp->inp_flags & INP_FC_FEEDBACK)
+               return (0);
+
+       inp->inp_flags &= ~(INP_FLOW_CONTROLLED | INP_FLOW_SUSPENDED);
+       if ((tmp_inp = inp_fc_getinp(inp->inp_flowhash,
+           INPFC_SOLOCKED)) != NULL) {
+               if (in_pcb_checkstate(tmp_inp, WNT_RELEASE, 1) == WNT_STOPUSING)
+                       return (0);
+               VERIFY(tmp_inp == inp);
+               switch (advcode) {
+               case FADV_FLOW_CONTROLLED:
+                       inp->inp_flags |= INP_FLOW_CONTROLLED;
+                       break;
+               case FADV_SUSPENDED:
+                       inp->inp_flags |= INP_FLOW_SUSPENDED;
+                       soevent(inp->inp_socket,
+                           (SO_FILT_HINT_LOCKED | SO_FILT_HINT_SUSPEND));
+
+                       /* Record the fact that suspend event was sent */
+                       inp->inp_socket->so_flags |= SOF_SUSPENDED;
+                       break;
+               }
+               return (1);
+       }
+       return (0);
+}
+
+/*
+ * Handler for SO_FLUSH socket option.
+ */
+int
+inp_flush(struct inpcb *inp, int optval)
+{
+       u_int32_t flowhash = inp->inp_flowhash;
+       struct ifnet *rtifp, *oifp;
+
+       /* Either all classes or one of the valid ones */
+       if (optval != SO_TC_ALL && !SO_VALID_TC(optval))
+               return (EINVAL);
+
+       /* We need a flow hash for identification */
+       if (flowhash == 0)
+               return (0);
+
+       /* Grab the interfaces from the route and pcb */
+       rtifp = ((inp->inp_route.ro_rt != NULL) ?
+           inp->inp_route.ro_rt->rt_ifp : NULL);
+       oifp = inp->inp_last_outifp;
+
+       if (rtifp != NULL)
+               if_qflush_sc(rtifp, so_tc2msc(optval), flowhash, NULL, NULL, 0);
+       if (oifp != NULL && oifp != rtifp)
+               if_qflush_sc(oifp, so_tc2msc(optval), flowhash, NULL, NULL, 0);
+
+       return (0);
+}
+
+/*
+ * Clear the INP_INADDR_ANY flag (special case for PPP only)
+ */
+void
+inp_clear_INP_INADDR_ANY(struct socket *so)
+{
+       struct inpcb *inp = NULL;
+
+       socket_lock(so, 1);
+       inp = sotoinpcb(so);
+       if (inp) {
+               inp->inp_flags &= ~INP_INADDR_ANY;
+       }
+       socket_unlock(so, 1);
+}
+
+void
+inp_get_soprocinfo(struct inpcb *inp, struct so_procinfo *soprocinfo)
+{
+       struct socket *so = inp->inp_socket;
+
+       soprocinfo->spi_pid = so->last_pid;
+       if (so->last_pid != 0)
+               uuid_copy(soprocinfo->spi_uuid, so->last_uuid);
+       /*
+        * When not delegated, the effective pid is the same as the real pid
+        */
+       if (so->so_flags & SOF_DELEGATED) {
+               soprocinfo->spi_delegated = 1;
+               soprocinfo->spi_epid = so->e_pid;
+               uuid_copy(soprocinfo->spi_euuid, so->e_uuid);
+       } else {
+               soprocinfo->spi_delegated = 0;
+               soprocinfo->spi_epid = so->last_pid;
+       }
+}
+
+int
+inp_findinpcb_procinfo(struct inpcbinfo *pcbinfo, uint32_t flowhash,
+    struct so_procinfo *soprocinfo)
+{
+       struct inpcb *inp = NULL;
+       int found = 0;
+
+       bzero(soprocinfo, sizeof (struct so_procinfo));
+
+       if (!flowhash)
+               return (-1);
+
+       lck_rw_lock_shared(pcbinfo->ipi_lock);
+       LIST_FOREACH(inp, pcbinfo->ipi_listhead, inp_list) {
+               if (inp->inp_state != INPCB_STATE_DEAD &&
+                   inp->inp_socket != NULL &&
+                   inp->inp_flowhash == flowhash) {
+                       found = 1;
+                       inp_get_soprocinfo(inp, soprocinfo);
+                       break;
+               }
+       }
+       lck_rw_done(pcbinfo->ipi_lock);
+
+       return (found);
+}
+
+#if CONFIG_PROC_UUID_POLICY
+static void
+inp_update_cellular_policy(struct inpcb *inp, boolean_t set)
+{
+       struct socket *so = inp->inp_socket;
+       int before, after;
+
+       VERIFY(so != NULL);
+       VERIFY(inp->inp_state != INPCB_STATE_DEAD);
+
+       before = INP_NO_CELLULAR(inp);
+       if (set) {
+               inp_set_nocellular(inp);
+       } else {
+               inp_clear_nocellular(inp);
+       }
+       after = INP_NO_CELLULAR(inp);
+       if (net_io_policy_log && (before != after)) {
+               static const char *ok = "OK";
+               static const char *nok = "NOACCESS";
+               uuid_string_t euuid_buf;
+               pid_t epid;
+
+               if (so->so_flags & SOF_DELEGATED) {
+                       uuid_unparse(so->e_uuid, euuid_buf);
+                       epid = so->e_pid;
                } else {
                } else {
-                       rtfree(dst->ro_rt);
-                       bcopy(src, dst, sizeof (*dst));
+                       uuid_unparse(so->last_uuid, euuid_buf);
+                       epid = so->last_pid;
                }
                }
+
+               /* allow this socket to generate another notification event */
+               so->so_ifdenied_notifies = 0;
+
+               log(LOG_DEBUG, "%s: so 0x%llx [%d,%d] epid %d "
+                   "euuid %s%s %s->%s\n", __func__,
+                   (uint64_t)VM_KERNEL_ADDRPERM(so), SOCK_DOM(so),
+                   SOCK_TYPE(so), epid, euuid_buf,
+                   (so->so_flags & SOF_DELEGATED) ?
+                   " [delegated]" : "",
+                   ((before < after) ? ok : nok),
+                   ((before < after) ? nok : ok));
        }
 }
        }
 }
+
+#if NECP
+static void
+inp_update_necp_want_app_policy(struct inpcb *inp, boolean_t set)
+{
+       struct socket *so = inp->inp_socket;
+       int before, after;
+
+       VERIFY(so != NULL);
+       VERIFY(inp->inp_state != INPCB_STATE_DEAD);
+
+       before = (inp->inp_flags2 & INP2_WANT_APP_POLICY);
+       if (set) {
+               inp_set_want_app_policy(inp);
+       } else {
+               inp_clear_want_app_policy(inp);
+       }
+       after = (inp->inp_flags2 & INP2_WANT_APP_POLICY);
+       if (net_io_policy_log && (before != after)) {
+               static const char *wanted = "WANTED";
+               static const char *unwanted = "UNWANTED";
+               uuid_string_t euuid_buf;
+               pid_t epid;
+
+               if (so->so_flags & SOF_DELEGATED) {
+                       uuid_unparse(so->e_uuid, euuid_buf);
+                       epid = so->e_pid;
+               } else {
+                       uuid_unparse(so->last_uuid, euuid_buf);
+                       epid = so->last_pid;
+               }
+
+               log(LOG_DEBUG, "%s: so 0x%llx [%d,%d] epid %d "
+                   "euuid %s%s %s->%s\n", __func__,
+                   (uint64_t)VM_KERNEL_ADDRPERM(so), SOCK_DOM(so),
+                   SOCK_TYPE(so), epid, euuid_buf,
+                   (so->so_flags & SOF_DELEGATED) ?
+                   " [delegated]" : "",
+                   ((before < after) ? unwanted : wanted),
+                   ((before < after) ? wanted : unwanted));
+       }
+}
+#endif /* NECP */
+#endif /* !CONFIG_PROC_UUID_POLICY */
+
+#if NECP
+void
+inp_update_necp_policy(struct inpcb *inp, struct sockaddr *override_local_addr, struct sockaddr *override_remote_addr, u_int override_bound_interface)
+{
+       necp_socket_find_policy_match(inp, override_local_addr, override_remote_addr, override_bound_interface);
+       if (necp_socket_should_rescope(inp) &&
+               inp->inp_lport == 0 &&
+               inp->inp_laddr.s_addr == INADDR_ANY &&
+               IN6_IS_ADDR_UNSPECIFIED(&inp->in6p_laddr)) {
+               // If we should rescope, and the socket is not yet bound
+               inp_bindif(inp, necp_socket_get_rescope_if_index(inp), NULL);
+       }
+}
+#endif /* NECP */
+
+int
+inp_update_policy(struct inpcb *inp)
+{
+#if CONFIG_PROC_UUID_POLICY
+       struct socket *so = inp->inp_socket;
+       uint32_t pflags = 0;
+       int32_t ogencnt;
+       int err = 0;
+
+       if (!net_io_policy_uuid ||
+           so == NULL || inp->inp_state == INPCB_STATE_DEAD)
+               return (0);
+
+       /*
+        * Kernel-created sockets that aren't delegating other sockets
+        * are currently exempted from UUID policy checks.
+        */
+       if (so->last_pid == 0 && !(so->so_flags & SOF_DELEGATED))
+               return (0);
+
+       ogencnt = so->so_policy_gencnt;
+       err = proc_uuid_policy_lookup(((so->so_flags & SOF_DELEGATED) ?
+           so->e_uuid : so->last_uuid), &pflags, &so->so_policy_gencnt);
+
+       /*
+        * Discard cached generation count if the entry is gone (ENOENT),
+        * so that we go thru the checks below.
+        */
+       if (err == ENOENT && ogencnt != 0)
+               so->so_policy_gencnt = 0;
+
+       /*
+        * If the generation count has changed, inspect the policy flags
+        * and act accordingly.  If a policy flag was previously set and
+        * the UUID is no longer present in the table (ENOENT), treat it
+        * as if the flag has been cleared.
+        */
+       if ((err == 0 || err == ENOENT) && ogencnt != so->so_policy_gencnt) {
+               /* update cellular policy for this socket */
+               if (err == 0 && (pflags & PROC_UUID_NO_CELLULAR)) {
+                       inp_update_cellular_policy(inp, TRUE);
+               } else if (!(pflags & PROC_UUID_NO_CELLULAR)) {
+                       inp_update_cellular_policy(inp, FALSE);
+               }
+#if NECP
+               /* update necp want app policy for this socket */
+               if (err == 0 && (pflags & PROC_UUID_NECP_APP_POLICY)) {
+                       inp_update_necp_want_app_policy(inp, TRUE);
+               } else if (!(pflags & PROC_UUID_NECP_APP_POLICY)) {
+                       inp_update_necp_want_app_policy(inp, FALSE);
+               }
+#endif /* NECP */
+       }
+
+       return ((err == ENOENT) ? 0 : err);
+#else /* !CONFIG_PROC_UUID_POLICY */
+#pragma unused(inp)
+       return (0);
+#endif /* !CONFIG_PROC_UUID_POLICY */
+}
+
+static unsigned int log_restricted;
+SYSCTL_DECL(_net_inet);
+SYSCTL_INT(_net_inet, OID_AUTO, log_restricted,
+    CTLFLAG_RW | CTLFLAG_LOCKED, &log_restricted, 0,
+    "Log network restrictions");
+/*
+ * Called when we need to enforce policy restrictions in the input path.
+ *
+ * Returns TRUE if we're not allowed to receive data, otherwise FALSE.
+ */
+static boolean_t
+_inp_restricted_recv(struct inpcb *inp, struct ifnet *ifp)
+{
+       VERIFY(inp != NULL);
+
+       /*
+        * Inbound restrictions.
+        */
+       if (!sorestrictrecv)
+               return (FALSE);
+
+       if (ifp == NULL)
+               return (FALSE);
+
+       if (IFNET_IS_CELLULAR(ifp) && INP_NO_CELLULAR(inp))
+               return (TRUE);
+
+       if (IFNET_IS_EXPENSIVE(ifp) && INP_NO_EXPENSIVE(inp))
+               return (TRUE);
+
+       if (IFNET_IS_AWDL_RESTRICTED(ifp) && !INP_AWDL_UNRESTRICTED(inp))
+               return (TRUE);
+
+       if (!(ifp->if_eflags & IFEF_RESTRICTED_RECV))
+               return (FALSE);
+
+       if (inp->inp_flags & INP_RECV_ANYIF)
+               return (FALSE);
+
+       if ((inp->inp_flags & INP_BOUND_IF) && inp->inp_boundifp == ifp)
+               return (FALSE);
+
+       if (IFNET_IS_INTCOPROC(ifp) && !INP_INTCOPROC_ALLOWED(inp))
+               return (TRUE);
+
+       return (TRUE);
+}
+
+boolean_t
+inp_restricted_recv(struct inpcb *inp, struct ifnet *ifp)
+{
+       boolean_t ret;
+
+       ret = _inp_restricted_recv(inp, ifp);
+       if (ret == TRUE && log_restricted) {
+               printf("pid %d (%s) is unable to receive packets on %s\n",
+                   current_proc()->p_pid, proc_best_name(current_proc()),
+                   ifp->if_xname);
+       }
+       return (ret);
+}
+
+/*
+ * Called when we need to enforce policy restrictions in the output path.
+ *
+ * Returns TRUE if we're not allowed to send data out, otherwise FALSE.
+ */
+static boolean_t
+_inp_restricted_send(struct inpcb *inp, struct ifnet *ifp)
+{
+       VERIFY(inp != NULL);
+
+       /*
+        * Outbound restrictions.
+        */
+       if (!sorestrictsend)
+               return (FALSE);
+
+       if (ifp == NULL)
+               return (FALSE);
+
+       if (IFNET_IS_CELLULAR(ifp) && INP_NO_CELLULAR(inp))
+               return (TRUE);
+
+       if (IFNET_IS_EXPENSIVE(ifp) && INP_NO_EXPENSIVE(inp))
+               return (TRUE);
+
+       if (IFNET_IS_AWDL_RESTRICTED(ifp) && !INP_AWDL_UNRESTRICTED(inp))
+               return (TRUE);
+
+       if (IFNET_IS_INTCOPROC(ifp) && !INP_INTCOPROC_ALLOWED(inp))
+               return (TRUE);
+
+       return (FALSE);
+}
+
+boolean_t
+inp_restricted_send(struct inpcb *inp, struct ifnet *ifp)
+{
+       boolean_t ret;
+
+       ret = _inp_restricted_send(inp, ifp);
+       if (ret == TRUE && log_restricted) {
+               printf("pid %d (%s) is unable to transmit packets on %s\n",
+                   current_proc()->p_pid, proc_best_name(current_proc()),
+                   ifp->if_xname);
+       }
+       return (ret);
+}
+
+inline void
+inp_count_sndbytes(struct inpcb *inp, u_int32_t th_ack)
+{
+       struct ifnet *ifp = inp->inp_last_outifp;
+       struct socket *so = inp->inp_socket;
+       if (ifp != NULL && !(so->so_flags & SOF_MP_SUBFLOW) &&
+           (ifp->if_type == IFT_CELLULAR ||
+           ifp->if_subfamily == IFNET_SUBFAMILY_WIFI)) {
+               int32_t unsent;
+
+               so->so_snd.sb_flags |= SB_SNDBYTE_CNT;
+
+               /*
+                * There can be data outstanding before the connection
+                * becomes established -- TFO case
+                */
+               if (so->so_snd.sb_cc > 0)
+                       inp_incr_sndbytes_total(so, so->so_snd.sb_cc);
+
+               unsent = inp_get_sndbytes_allunsent(so, th_ack);
+               if (unsent > 0)
+                       inp_incr_sndbytes_unsent(so, unsent);
+       }
+}
+
+inline void
+inp_incr_sndbytes_total(struct socket *so, int32_t len)
+{
+       struct inpcb *inp = (struct inpcb *)so->so_pcb;
+       struct ifnet *ifp = inp->inp_last_outifp;
+
+       if (ifp != NULL) {
+               VERIFY(ifp->if_sndbyte_total >= 0);
+               OSAddAtomic64(len, &ifp->if_sndbyte_total);
+       }
+}
+
+inline void
+inp_decr_sndbytes_total(struct socket *so, int32_t len)
+{
+       struct inpcb *inp = (struct inpcb *)so->so_pcb;
+       struct ifnet *ifp = inp->inp_last_outifp;
+
+       if (ifp != NULL) {
+               VERIFY(ifp->if_sndbyte_total >= len);
+               OSAddAtomic64(-len, &ifp->if_sndbyte_total);
+       }
+}
+
+inline void
+inp_incr_sndbytes_unsent(struct socket *so, int32_t len)
+{
+       struct inpcb *inp = (struct inpcb *)so->so_pcb;
+       struct ifnet *ifp = inp->inp_last_outifp;
+
+       if (ifp != NULL) {
+               VERIFY(ifp->if_sndbyte_unsent >= 0);
+               OSAddAtomic64(len, &ifp->if_sndbyte_unsent);
+       }
+}
+
+inline void
+inp_decr_sndbytes_unsent(struct socket *so, int32_t len)
+{
+       struct inpcb *inp = (struct inpcb *)so->so_pcb;
+       struct ifnet *ifp = inp->inp_last_outifp;
+
+       if (so == NULL || !(so->so_snd.sb_flags & SB_SNDBYTE_CNT))
+               return;
+
+       if (ifp != NULL) {
+               if (ifp->if_sndbyte_unsent >= len)
+                       OSAddAtomic64(-len, &ifp->if_sndbyte_unsent);
+               else
+                       ifp->if_sndbyte_unsent = 0;
+       }
+}
+
+inline void
+inp_decr_sndbytes_allunsent(struct socket *so, u_int32_t th_ack)
+{
+       int32_t len;
+
+       if (so == NULL || !(so->so_snd.sb_flags & SB_SNDBYTE_CNT))
+               return;
+
+       len = inp_get_sndbytes_allunsent(so, th_ack);
+       inp_decr_sndbytes_unsent(so, len);
+}
+
+
+inline void
+inp_set_activity_bitmap(struct inpcb *inp)
+{
+       in_stat_set_activity_bitmap(&inp->inp_nw_activity, net_uptime());
+}
+
+inline void
+inp_get_activity_bitmap(struct inpcb *inp, activity_bitmap_t *ab)
+{
+       bcopy(&inp->inp_nw_activity, ab, sizeof (*ab));
+}