]> git.saurik.com Git - apple/xnu.git/blobdiff - bsd/netinet/udp_usrreq.c
xnu-2422.115.4.tar.gz
[apple/xnu.git] / bsd / netinet / udp_usrreq.c
index af06d92642c17ac48a62dea27255de4f04e76904..1b2bd8b1fec4cf81ef82f6c647b3b0745d33fea1 100644 (file)
@@ -1,23 +1,29 @@
 /*
- * Copyright (c) 2000 Apple Computer, Inc. All rights reserved.
+ * Copyright (c) 2000-2013 Apple Inc. All rights reserved.
  *
- * @APPLE_LICENSE_HEADER_START@
+ * @APPLE_OSREFERENCE_LICENSE_HEADER_START@
  * 
- * The contents of this file constitute Original Code as defined in and
- * are subject to the Apple Public Source License Version 1.1 (the
- * "License").  You may not use this file except in compliance with the
- * License.  Please obtain a copy of the License at
- * http://www.apple.com/publicsource and read it before using this file.
+ * This file contains Original Code and/or Modifications of Original Code
+ * as defined in and that are subject to the Apple Public Source License
+ * Version 2.0 (the 'License'). You may not use this file except in
+ * compliance with the License. The rights granted to you under the License
+ * may not be used to create, or enable the creation or redistribution of,
+ * unlawful or unlicensed copies of an Apple operating system, or to
+ * circumvent, violate, or enable the circumvention or violation of, any
+ * terms of an Apple operating system software license agreement.
  * 
- * This Original Code and all software distributed under the License are
- * distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY KIND, EITHER
+ * Please obtain a copy of the License at
+ * http://www.opensource.apple.com/apsl/ and read it before using this file.
+ * 
+ * The Original Code and all software distributed under the License are
+ * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
  * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
  * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE 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_LICENSE_HEADER_END@
+ * @APPLE_OSREFERENCE_LICENSE_HEADER_END@
  */
 /*
  * Copyright (c) 1982, 1986, 1988, 1990, 1993, 1995
 #include <sys/socketvar.h>
 #include <sys/sysctl.h>
 #include <sys/syslog.h>
+#include <sys/mcache.h>
+#include <net/ntstat.h>
+
+#include <kern/zalloc.h>
+#include <mach/boolean.h>
 
 #include <net/if.h>
 #include <net/if_types.h>
 #include <net/route.h>
+#include <net/dlil.h>
 
 #include <netinet/in.h>
 #include <netinet/in_systm.h>
 #include <netinet/ip.h>
 #if INET6
 #include <netinet/ip6.h>
-#endif
+#endif /* INET6 */
 #include <netinet/in_pcb.h>
 #include <netinet/in_var.h>
 #include <netinet/ip_var.h>
 #if INET6
+#include <netinet6/in6_pcb.h>
 #include <netinet6/ip6_var.h>
-#endif
+#include <netinet6/udp6_var.h>
+#endif /* INET6 */
 #include <netinet/ip_icmp.h>
 #include <netinet/icmp_var.h>
 #include <netinet/udp.h>
 
 #if IPSEC
 #include <netinet6/ipsec.h>
+#include <netinet6/esp.h>
 extern int ipsec_bypass;
-extern lck_mtx_t *sadb_mutex;
-#endif /*IPSEC*/
-
+extern int esp_udp_encap_port;
+#endif /* IPSEC */
 
-#define DBG_LAYER_IN_BEG       NETDBG_CODE(DBG_NETUDP, 0)
-#define DBG_LAYER_IN_END       NETDBG_CODE(DBG_NETUDP, 2)
-#define DBG_LAYER_OUT_BEG      NETDBG_CODE(DBG_NETUDP, 1)
-#define DBG_LAYER_OUT_END      NETDBG_CODE(DBG_NETUDP, 3)
-#define DBG_FNC_UDP_INPUT      NETDBG_CODE(DBG_NETUDP, (5 << 8))
-#define DBG_FNC_UDP_OUTPUT     NETDBG_CODE(DBG_NETUDP, (6 << 8) | 1)
+#define        DBG_LAYER_IN_BEG        NETDBG_CODE(DBG_NETUDP, 0)
+#define        DBG_LAYER_IN_END        NETDBG_CODE(DBG_NETUDP, 2)
+#define        DBG_LAYER_OUT_BEG       NETDBG_CODE(DBG_NETUDP, 1)
+#define        DBG_LAYER_OUT_END       NETDBG_CODE(DBG_NETUDP, 3)
+#define        DBG_FNC_UDP_INPUT       NETDBG_CODE(DBG_NETUDP, (5 << 8))
+#define        DBG_FNC_UDP_OUTPUT      NETDBG_CODE(DBG_NETUDP, (6 << 8) | 1)
 
 /*
  * UDP protocol implementation.
  * Per RFC 768, August, 1980.
  */
 #ifndef        COMPAT_42
-static int     udpcksum = 1;
+static int udpcksum = 1;
 #else
-static int     udpcksum = 0;           /* XXX */
+static int udpcksum = 0;               /* XXX */
 #endif
-SYSCTL_INT(_net_inet_udp, UDPCTL_CHECKSUM, checksum, CTLFLAG_RW,
-               &udpcksum, 0, "");
+SYSCTL_INT(_net_inet_udp, UDPCTL_CHECKSUM, checksum,
+    CTLFLAG_RW | CTLFLAG_LOCKED, &udpcksum, 0, "");
 
-int    log_in_vain = 0;
-SYSCTL_INT(_net_inet_udp, OID_AUTO, log_in_vain, CTLFLAG_RW
-    &log_in_vain, 0, "Log all incoming UDP packets");
+int udp_log_in_vain = 0;
+SYSCTL_INT(_net_inet_udp, OID_AUTO, log_in_vain, CTLFLAG_RW | CTLFLAG_LOCKED,
+    &udp_log_in_vain, 0, "Log all incoming UDP packets");
 
-static int     blackhole = 0;
-SYSCTL_INT(_net_inet_udp, OID_AUTO, blackhole, CTLFLAG_RW,
-       &blackhole, 0, "Do not send port unreachables for refused connects");
+static int blackhole = 0;
+SYSCTL_INT(_net_inet_udp, OID_AUTO, blackhole, CTLFLAG_RW | CTLFLAG_LOCKED,
+    &blackhole, 0, "Do not send port unreachables for refused connects");
 
-struct inpcbhead udb;          /* from udp_var.h */
+struct inpcbhead udb;          /* from udp_var.h */
 #define        udb6    udb  /* for KAME src sync over BSD*'s */
-struct inpcbinfo udbinfo;
+struct inpcbinfo udbinfo;
 
 #ifndef UDBHASHSIZE
 #define UDBHASHSIZE 16
 #endif
 
-extern  int apple_hwcksum_rx;
-extern int     esp_udp_encap_port;
-extern u_long  route_generation;
+/* Garbage collection performed during most recent udp_gc() run */
+static boolean_t udp_gc_done = FALSE;
 
-extern  void    ipfwsyslog( int level, char *format,...);
+#if IPFIREWALL
 extern int fw_verbose;
-
-#define log_in_vain_log( a ) {            \
-        if ( (log_in_vain == 3 ) && (fw_verbose == 2)) {        /* Apple logging, log to ipfw.log */ \
-                ipfwsyslog a ;  \
-        }                       \
-        else log a ;            \
+extern void ipfwsyslog( int level, const char *format,...);
+extern void ipfw_stealth_stats_incr_udp(void);
+
+/* Apple logging, log to ipfw.log */
+#define log_in_vain_log(a) {                                           \
+       if ((udp_log_in_vain == 3) && (fw_verbose == 2)) {              \
+               ipfwsyslog a;                                           \
+       } else if ((udp_log_in_vain == 4) && (fw_verbose == 2)) {       \
+               ipfw_stealth_stats_incr_udp();                          \
+       } else {                                                        \
+               log a;                                                  \
+       }                                                               \
 }
+#else /* !IPFIREWALL */
+#define log_in_vain_log( a ) { log a; }
+#endif /* !IPFIREWALL */
 
+static int udp_getstat SYSCTL_HANDLER_ARGS;
 struct udpstat udpstat;        /* from udp_var.h */
-SYSCTL_STRUCT(_net_inet_udp, UDPCTL_STATS, stats, CTLFLAG_RD,
-    &udpstat, udpstat, "UDP statistics (struct udpstat, netinet/udp_var.h)");
-SYSCTL_INT(_net_inet_udp, OID_AUTO, pcbcount, CTLFLAG_RD, 
-    &udbinfo.ipi_count, 0, "Number of active PCBs");
+SYSCTL_PROC(_net_inet_udp, UDPCTL_STATS, stats, CTLFLAG_RD | CTLFLAG_LOCKED,
+    0, 0, udp_getstat, "S,udpstat",
+    "UDP statistics (struct udpstat, netinet/udp_var.h)");
+
+SYSCTL_INT(_net_inet_udp, OID_AUTO, pcbcount,
+    CTLFLAG_RD | CTLFLAG_LOCKED, &udbinfo.ipi_count, 0,
+    "Number of active PCBs");
+
+__private_extern__ int udp_use_randomport = 1;
+SYSCTL_INT(_net_inet_udp, OID_AUTO, randomize_ports,
+    CTLFLAG_RW | CTLFLAG_LOCKED, &udp_use_randomport, 0,
+    "Randomize UDP port numbers");
 
-static struct  sockaddr_in udp_in = { sizeof(udp_in), AF_INET };
 #if INET6
 struct udp_in6 {
        struct sockaddr_in6     uin6_sin;
        u_char                  uin6_init_done : 1;
-} udp_in6 = {
-       { sizeof(udp_in6.uin6_sin), AF_INET6 },
-       0
 };
 struct udp_ip6 {
        struct ip6_hdr          uip6_ip6;
        u_char                  uip6_init_done : 1;
-} udp_ip6;
-#endif /* INET6 */
+};
 
-static void udp_append(struct inpcb *last, struct ip *ip,
-                           struct mbuf *n, int off);
-#if INET6
+static int udp_abort(struct socket *);
+static int udp_attach(struct socket *, int, struct proc *);
+static int udp_bind(struct socket *, struct sockaddr *, struct proc *);
+static int udp_connect(struct socket *, struct sockaddr *, struct proc *);
+static int udp_connectx(struct socket *, struct sockaddr_list **,
+    struct sockaddr_list **, struct proc *, uint32_t, associd_t, connid_t *,
+    uint32_t, void *, uint32_t);
+static int udp_detach(struct socket *);
+static int udp_disconnect(struct socket *);
+static int udp_disconnectx(struct socket *, associd_t, connid_t);
+static int udp_send(struct socket *, int, struct mbuf *, struct sockaddr *,
+    struct mbuf *, struct proc *);
+static void udp_append(struct inpcb *, struct ip *, struct mbuf *, int,
+    struct sockaddr_in *, struct udp_in6 *, struct udp_ip6 *, struct ifnet *);
+#else /* !INET6 */
+static void udp_append(struct inpcb *, struct ip *, struct mbuf *, int,
+    struct sockaddr_in *, struct ifnet *);
+#endif /* !INET6 */
+static int udp_input_checksum(struct mbuf *, struct udphdr *, int, int);
+static int udp_output(struct inpcb *, struct mbuf *, struct sockaddr *,
+    struct mbuf *, struct proc *);
 static void ip_2_ip6_hdr(struct ip6_hdr *ip6, struct ip *ip);
-#endif
+static void udp_gc(struct inpcbinfo *);
 
-static int udp_detach(struct socket *so);
-static int udp_output(struct inpcb *, struct mbuf *, struct sockaddr *,
-                           struct mbuf *, struct proc *);
-extern int ChkAddressOK( __uint32_t dstaddr, __uint32_t srcaddr );
+struct pr_usrreqs udp_usrreqs = {
+       .pru_abort =            udp_abort,
+       .pru_attach =           udp_attach,
+       .pru_bind =             udp_bind,
+       .pru_connect =          udp_connect,
+       .pru_connectx =         udp_connectx,
+       .pru_control =          in_control,
+       .pru_detach =           udp_detach,
+       .pru_disconnect =       udp_disconnect,
+       .pru_disconnectx =      udp_disconnectx,
+       .pru_peeraddr =         in_getpeeraddr,
+       .pru_send =             udp_send,
+       .pru_shutdown =         udp_shutdown,
+       .pru_sockaddr =         in_getsockaddr,
+       .pru_sosend =           sosend,
+       .pru_soreceive =        soreceive,
+};
 
 void
-udp_init()
+udp_init(struct protosw *pp, struct domain *dp)
 {
-       vm_size_t                       str_size;
-       struct inpcbinfo        *pcbinfo;
-       
+#pragma unused(dp)
+       static int udp_initialized = 0;
+       vm_size_t               str_size;
+       struct inpcbinfo        *pcbinfo;
+
+       VERIFY((pp->pr_flags & (PR_INITIALIZED|PR_ATTACHED)) == PR_ATTACHED);
+
+       if (udp_initialized)
+               return;
+       udp_initialized = 1;
 
        LIST_INIT(&udb);
-       udbinfo.listhead = &udb;
-       udbinfo.hashbase = hashinit(UDBHASHSIZE, M_PCB, &udbinfo.hashmask);
-       udbinfo.porthashbase = hashinit(UDBHASHSIZE, M_PCB,
-                                       &udbinfo.porthashmask);
-#ifdef __APPLE__
-       str_size = (vm_size_t) sizeof(struct inpcb);
-       udbinfo.ipi_zone = (void *) zinit(str_size, 80000*str_size, 8192, "udpcb");
-
-       pcbinfo = &udbinfo;
+       udbinfo.ipi_listhead = &udb;
+       udbinfo.ipi_hashbase = hashinit(UDBHASHSIZE, M_PCB,
+           &udbinfo.ipi_hashmask);
+       udbinfo.ipi_porthashbase = hashinit(UDBHASHSIZE, M_PCB,
+           &udbinfo.ipi_porthashmask);
+       str_size = (vm_size_t) sizeof (struct inpcb);
+       udbinfo.ipi_zone = zinit(str_size, 80000*str_size, 8192, "udpcb");
+
+       pcbinfo = &udbinfo;
        /*
         * allocate lock group attribute and group for udp pcb mutexes
         */
-       pcbinfo->mtx_grp_attr = lck_grp_attr_alloc_init();
-
-       pcbinfo->mtx_grp = lck_grp_alloc_init("udppcb", pcbinfo->mtx_grp_attr);
-               
-       pcbinfo->mtx_attr = lck_attr_alloc_init();
-
-       if ((pcbinfo->mtx = lck_rw_alloc_init(pcbinfo->mtx_grp, pcbinfo->mtx_attr)) == NULL)
-               return; /* pretty much dead if this fails... */
-
-       in_pcb_nat_init(&udbinfo, AF_INET, IPPROTO_UDP, SOCK_DGRAM);
-#else
-       udbinfo.ipi_zone = zinit("udpcb", sizeof(struct inpcb), maxsockets,
-                                ZONE_INTERRUPT, 0);
-#endif
+       pcbinfo->ipi_lock_grp_attr = lck_grp_attr_alloc_init();
+       pcbinfo->ipi_lock_grp = lck_grp_alloc_init("udppcb",
+           pcbinfo->ipi_lock_grp_attr);
+       pcbinfo->ipi_lock_attr = lck_attr_alloc_init();
+       if ((pcbinfo->ipi_lock = lck_rw_alloc_init(pcbinfo->ipi_lock_grp,
+           pcbinfo->ipi_lock_attr)) == NULL) {
+               panic("%s: unable to allocate PCB lock\n", __func__);
+               /* NOTREACHED */
+       }
 
-#if 0
-       /* for pcb sharing testing only */
-       stat = in_pcb_new_share_client(&udbinfo, &fake_owner);
-       kprintf("udp_init in_pcb_new_share_client - stat = %d\n", stat);
-
-       laddr.s_addr = 0x11646464;
-       faddr.s_addr = 0x11646465;
-       
-       lport = 1500;
-       in_pcb_grab_port(&udbinfo, 0, laddr, &lport, faddr, 1600, 0, fake_owner); 
-       kprintf("udp_init in_pcb_grab_port - stat = %d\n", stat);
-
-       stat = in_pcb_rem_share_client(&udbinfo, fake_owner);
-       kprintf("udp_init in_pcb_rem_share_client - stat = %d\n", stat);
-
-       stat = in_pcb_new_share_client(&udbinfo, &fake_owner);
-       kprintf("udp_init in_pcb_new_share_client(2) - stat = %d\n", stat);
-
-       laddr.s_addr = 0x11646464;
-       faddr.s_addr = 0x11646465;
-       
-       lport = 1500;
-       stat = in_pcb_grab_port(&udbinfo, 0, laddr, &lport, faddr, 1600, 0, fake_owner); 
-       kprintf("udp_init in_pcb_grab_port(2) - stat = %d\n", stat);
-#endif
+       udbinfo.ipi_gc = udp_gc;
+       in_pcbinfo_attach(&udbinfo);
 }
 
 void
-udp_input(m, iphlen)
-       register struct mbuf *m;
-       int iphlen;
+udp_input(struct mbuf *m, int iphlen)
 {
-       register struct ip *ip;
-       register struct udphdr *uh;
-       register struct inpcb *inp;
-       struct mbuf *opts = 0;
-       int len;
+       struct ip *ip;
+       struct udphdr *uh;
+       struct inpcb *inp;
+       struct mbuf *opts = NULL;
+       int len, isbroadcast;
        struct ip save_ip;
        struct sockaddr *append_sa;
        struct inpcbinfo *pcbinfo = &udbinfo;
+       struct sockaddr_in udp_in;
+       struct ip_moptions *imo = NULL;
+       int foundmembership = 0, ret = 0;
+#if INET6
+       struct udp_in6 udp_in6;
+       struct udp_ip6 udp_ip6;
+#endif /* INET6 */
+       struct ifnet *ifp = m->m_pkthdr.rcvif;
+       boolean_t cell = IFNET_IS_CELLULAR(ifp);
+       boolean_t wifi = (!cell && IFNET_IS_WIFI(ifp));
+
+       bzero(&udp_in, sizeof (udp_in));
+       udp_in.sin_len = sizeof (struct sockaddr_in);
+       udp_in.sin_family = AF_INET;
+#if INET6
+       bzero(&udp_in6, sizeof (udp_in6));
+       udp_in6.uin6_sin.sin6_len = sizeof (struct sockaddr_in6);
+       udp_in6.uin6_sin.sin6_family = AF_INET6;
+#endif /* INET6 */
 
        udpstat.udps_ipackets++;
 
        KERNEL_DEBUG(DBG_FNC_UDP_INPUT | DBG_FUNC_START, 0,0,0,0,0);
-       if (m->m_pkthdr.csum_flags & CSUM_TCP_SUM16)
-               m->m_pkthdr.csum_flags = 0; /* invalidate hwcksum for UDP */
+
+       /* Expect 32-bit aligned data pointer on strict-align platforms */
+       MBUF_STRICT_DATA_ALIGNMENT_CHECK_32(m);
 
        /*
         * Strip IP options, if any; should skip this,
@@ -268,29 +321,33 @@ udp_input(m, iphlen)
         */
        if (iphlen > sizeof (struct ip)) {
                ip_stripoptions(m, (struct mbuf *)0);
-               iphlen = sizeof(struct ip);
+               iphlen = sizeof (struct ip);
        }
 
        /*
         * Get IP and UDP header together in first mbuf.
         */
        ip = mtod(m, struct ip *);
-       if (m->m_len < iphlen + sizeof(struct udphdr)) {
-               if ((m = m_pullup(m, iphlen + sizeof(struct udphdr))) == 0) {
+       if (m->m_len < iphlen + sizeof (struct udphdr)) {
+               m = m_pullup(m, iphlen + sizeof (struct udphdr));
+               if (m == NULL) {
                        udpstat.udps_hdrops++;
-                       KERNEL_DEBUG(DBG_FNC_UDP_INPUT | DBG_FUNC_END, 0,0,0,0,0);
+                       KERNEL_DEBUG(DBG_FNC_UDP_INPUT | DBG_FUNC_END,
+                           0,0,0,0,0);
                        return;
                }
                ip = mtod(m, struct ip *);
        }
-       uh = (struct udphdr *)((caddr_t)ip + iphlen);
+       uh = (struct udphdr *)(void *)((caddr_t)ip + iphlen);
 
        /* destination port of 0 is illegal, based on RFC768. */
-       if (uh->uh_dport == 0)
+       if (uh->uh_dport == 0) {
+               IF_UDP_STATINC(ifp, port0);
                goto bad;
+       }
 
        KERNEL_DEBUG(DBG_LAYER_IN_BEG, uh->uh_dport, uh->uh_sport,
-                    ip->ip_src.s_addr, ip->ip_dst.s_addr, uh->uh_ulen);
+           ip->ip_src.s_addr, ip->ip_dst.s_addr, uh->uh_ulen);
 
        /*
         * Make mbuf data length reflect UDP length.
@@ -298,8 +355,9 @@ udp_input(m, iphlen)
         */
        len = ntohs((u_short)uh->uh_ulen);
        if (ip->ip_len != len) {
-               if (len > ip->ip_len || len < sizeof(struct udphdr)) {
+               if (len > ip->ip_len || len < sizeof (struct udphdr)) {
                        udpstat.udps_badlen++;
+                       IF_UDP_STATINC(ifp, badlength);
                        goto bad;
                }
                m_adj(m, len - ip->ip_len);
@@ -314,44 +372,15 @@ udp_input(m, iphlen)
        /*
         * Checksum extended UDP header and data.
         */
-       if (uh->uh_sum) {
-               if (m->m_pkthdr.csum_flags & CSUM_DATA_VALID) {
-                       if (m->m_pkthdr.csum_flags & CSUM_PSEUDO_HDR)
-                               uh->uh_sum = m->m_pkthdr.csum_data;
-                       else
-                               goto doudpcksum;
-                       uh->uh_sum ^= 0xffff;
-               } else {
-                       char b[9];
-doudpcksum:
-                       *(uint32_t*)&b[0] = *(uint32_t*)&((struct ipovly *)ip)->ih_x1[0];
-                       *(uint32_t*)&b[4] = *(uint32_t*)&((struct ipovly *)ip)->ih_x1[4];
-                       *(uint8_t*)&b[8] = *(uint8_t*)&((struct ipovly *)ip)->ih_x1[8];
-                       
-                       bzero(((struct ipovly *)ip)->ih_x1, 9);
-                       ((struct ipovly *)ip)->ih_len = uh->uh_ulen;
-                       uh->uh_sum = in_cksum(m, len + sizeof (struct ip));
-                       
-                       *(uint32_t*)&((struct ipovly *)ip)->ih_x1[0] = *(uint32_t*)&b[0];
-                       *(uint32_t*)&((struct ipovly *)ip)->ih_x1[4] = *(uint32_t*)&b[4];
-                       *(uint8_t*)&((struct ipovly *)ip)->ih_x1[8] = *(uint8_t*)&b[8];
-               }
-               if (uh->uh_sum) {
-                       udpstat.udps_badsum++;
-                       m_freem(m);
-                       KERNEL_DEBUG(DBG_FNC_UDP_INPUT | DBG_FUNC_END, 0,0,0,0,0);
-                       return;
-               }
-       }
-#ifndef __APPLE__
-        else
-               udpstat.udps_nosum++;
-#endif
+       if (udp_input_checksum(m, uh, iphlen, len))
+               goto bad;
+
+       isbroadcast = in_broadcast(ip->ip_dst, ifp);
 
-       if (IN_MULTICAST(ntohl(ip->ip_dst.s_addr)) ||
-           in_broadcast(ip->ip_dst, m->m_pkthdr.rcvif)) {
-               struct inpcb *last;
-               lck_rw_lock_shared(pcbinfo->mtx);
+       if (IN_MULTICAST(ntohl(ip->ip_dst.s_addr)) || isbroadcast) {
+               int reuse_sock = 0, mcast_delivered = 0;
+
+               lck_rw_lock_shared(pcbinfo->ipi_lock);
                /*
                 * Deliver a multicast or broadcast datagram to *all* sockets
                 * for which the local and remote addresses and ports match
@@ -368,7 +397,6 @@ doudpcksum:
                 * fixing the interface.  Maybe 4.5BSD will remedy this?)
                 */
 
-
                /*
                 * Construct sockaddr format source address.
                 */
@@ -378,31 +406,45 @@ doudpcksum:
                 * Locate pcb(s) for datagram.
                 * (Algorithm copied from raw_intr().)
                 */
-               last = NULL;
 #if INET6
                udp_in6.uin6_init_done = udp_ip6.uip6_init_done = 0;
-#endif
+#endif /* INET6 */
                LIST_FOREACH(inp, &udb, inp_list) {
-#ifdef __APPLE__
-                       /* Ignore nat/SharedIP dummy pcbs */
-                       if (inp->inp_socket == &udbinfo.nat_dummy_socket)
-                               continue;
-#endif
-                       if (inp->inp_socket == NULL) 
+#if IPSEC
+                       int skipit;
+#endif /* IPSEC */
+
+                       if (inp->inp_socket == NULL)
                                continue;
-                       if (inp != sotoinpcb(inp->inp_socket))
-                               panic("udp_input: bad so back ptr inp=%x\n", inp);
+                       if (inp != sotoinpcb(inp->inp_socket)) {
+                               panic("%s: bad so back ptr inp=%p\n",
+                                   __func__, inp);
+                               /* NOTREACHED */
+                       }
 #if INET6
                         if ((inp->inp_vflag & INP_IPV4) == 0)
                                 continue;
-#endif
-                       if (in_pcb_checkstate(inp, WNT_ACQUIRE, 0) == WNT_STOPUSING) {
+#endif /* INET6 */
+                       if (inp_restricted(inp, ifp))
+                               continue;
+
+                       if (IFNET_IS_CELLULAR(ifp) &&
+                           (inp->inp_flags & INP_NO_IFT_CELLULAR))
+                               continue;
+
+                       if ((inp->inp_moptions == NULL) &&
+                           (ntohl(ip->ip_dst.s_addr) !=
+                           INADDR_ALLHOSTS_GROUP) && (isbroadcast == 0))
+                               continue;
+
+                       if (in_pcb_checkstate(inp, WNT_ACQUIRE, 0) ==
+                           WNT_STOPUSING)
                                continue;
-                       }
-                       
-                       udp_lock(inp->inp_socket, 1, 0);        
 
-                       if (in_pcb_checkstate(inp, WNT_RELEASE, 1) == WNT_STOPUSING) {
+                       udp_lock(inp->inp_socket, 1, 0);
+
+                       if (in_pcb_checkstate(inp, WNT_RELEASE, 1) ==
+                           WNT_STOPUSING) {
                                udp_unlock(inp->inp_socket, 1, 0);
                                continue;
                        }
@@ -427,67 +469,112 @@ doudpcksum:
                                }
                        }
 
-                       if (last != NULL) {
-                               struct mbuf *n;
-#if IPSEC
-                               int skipit = 0;
-                               /* check AH/ESP integrity. */
-                               if (ipsec_bypass == 0) {
-                                       lck_mtx_lock(sadb_mutex);
-                                       if (ipsec4_in_reject_so(m, last->inp_socket)) {
-                                               ipsecstat.in_polvio++;
-                                               /* do not inject data to pcb */
-                                               skipit = 1;
-                                       }
-                                       lck_mtx_unlock(sadb_mutex);
+                       if (isbroadcast == 0 && (ntohl(ip->ip_dst.s_addr) !=
+                           INADDR_ALLHOSTS_GROUP)) {
+                               struct sockaddr_in group;
+                               int blocked;
+
+                               if ((imo = inp->inp_moptions) == NULL) {
+                                       udp_unlock(inp->inp_socket, 1, 0);
+                                       continue;
                                }
-                               if (skipit == 0)
-#endif /*IPSEC*/
-                               if ((n = m_copy(m, 0, M_COPYALL)) != NULL) {
-                                       udp_append(last, ip, n,
-                                                  iphlen +
-                                                  sizeof(struct udphdr));
+                               IMO_LOCK(imo);
+
+                               bzero(&group, sizeof (struct sockaddr_in));
+                               group.sin_len = sizeof (struct sockaddr_in);
+                               group.sin_family = AF_INET;
+                               group.sin_addr = ip->ip_dst;
+
+                               blocked = imo_multi_filter(imo, ifp,
+                                   (struct sockaddr *)&group,
+                                   (struct sockaddr *)&udp_in);
+                               if (blocked == MCAST_PASS)
+                                       foundmembership = 1;
+
+                               IMO_UNLOCK(imo);
+                               if (!foundmembership) {
+                                       udp_unlock(inp->inp_socket, 1, 0);
+                                       if (blocked == MCAST_NOTSMEMBER ||
+                                           blocked == MCAST_MUTED)
+                                               udpstat.udps_filtermcast++;
+                                       continue;
                                }
-                               udp_unlock(last->inp_socket, 1, 0);
+                               foundmembership = 0;
+                       }
+
+                       reuse_sock = (inp->inp_socket->so_options &
+                           (SO_REUSEPORT|SO_REUSEADDR));
+
+#if IPSEC
+                       skipit = 0;
+                       /* check AH/ESP integrity. */
+                       if (ipsec_bypass == 0 &&
+                           ipsec4_in_reject_so(m, inp->inp_socket)) {
+                               IPSEC_STAT_INCREMENT(ipsecstat.in_polvio);
+                               /* do not inject data to pcb */
+                               skipit = 1;
                        }
-                       last = inp;
+                       if (skipit == 0)
+#endif /*IPSEC*/
+                       {
+                               struct mbuf *n = NULL;
+
+                               if (reuse_sock)
+                                       n = m_copy(m, 0, M_COPYALL);
+#if INET6
+                               udp_append(inp, ip, m,
+                                   iphlen + sizeof (struct udphdr),
+                                   &udp_in, &udp_in6, &udp_ip6, ifp);
+#else /* !INET6 */
+                               udp_append(inp, ip, m,
+                                   iphlen + sizeof (struct udphdr),
+                                   &udp_in, ifp);
+#endif /* !INET6 */
+                               mcast_delivered++;
+
+                               m = n;
+                       }
+                       udp_unlock(inp->inp_socket, 1, 0);
+
                        /*
                         * Don't look for additional matches if this one does
                         * not have either the SO_REUSEPORT or SO_REUSEADDR
                         * socket options set.  This heuristic avoids searching
                         * through all pcbs in the common case of a non-shared
-                        * port.  It assumes that an application will never
+                        * port.  It assumes that an application will never
                         * clear these options after setting them.
                         */
-                       if ((last->inp_socket->so_options&(SO_REUSEPORT|SO_REUSEADDR)) == 0)
+                       if (reuse_sock == 0 || m == NULL)
                                break;
+
+                       /*
+                        * Expect 32-bit aligned data pointer on strict-align
+                        * platforms.
+                        */
+                       MBUF_STRICT_DATA_ALIGNMENT_CHECK_32(m);
+                       /*
+                        * Recompute IP and UDP header pointers for new mbuf
+                        */
+                       ip = mtod(m, struct ip *);
+                       uh = (struct udphdr *)(void *)((caddr_t)ip + iphlen);
                }
-               lck_rw_done(pcbinfo->mtx);
+               lck_rw_done(pcbinfo->ipi_lock);
 
-               if (last == NULL) {
+               if (mcast_delivered == 0) {
                        /*
                         * No matching pcb found; discard datagram.
                         * (No need to send an ICMP Port Unreachable
                         * for a broadcast or multicast datgram.)
                         */
                        udpstat.udps_noportbcast++;
+                       IF_UDP_STATINC(ifp, port_unreach);
                        goto bad;
                }
-#if IPSEC
-               /* check AH/ESP integrity. */
-               if (ipsec_bypass == 0 && m) {
-                       lck_mtx_lock(sadb_mutex);
-                       if (ipsec4_in_reject_so(m, last->inp_socket)) {
-                               ipsecstat.in_polvio++;
-                               lck_mtx_unlock(sadb_mutex);
-                               udp_unlock(last->inp_socket, 1, 0);
-                               goto bad;
-                       }
-                       lck_mtx_unlock(sadb_mutex);
-               }
-#endif /*IPSEC*/
-               udp_append(last, ip, m, iphlen + sizeof(struct udphdr));
-               udp_unlock(last->inp_socket, 1, 0);
+
+               /* free the extra copy of mbuf or skipped by IPSec */
+               if (m != NULL)
+                       m_freem(m);
+               KERNEL_DEBUG(DBG_FNC_UDP_INPUT | DBG_FUNC_END, 0,0,0,0,0);
                return;
        }
 
@@ -500,71 +587,76 @@ doudpcksum:
         * or discard the NAT keep-alive.
         */
        if (ipsec_bypass == 0 && (esp_udp_encap_port & 0xFFFF) != 0 &&
-               uh->uh_dport == ntohs((u_short)esp_udp_encap_port)) {
-               int     payload_len = len - sizeof(struct udphdr) > 4 ? 4 : len - sizeof(struct udphdr);
-               if (m->m_len < iphlen + sizeof(struct udphdr) + payload_len) {
-                       if ((m = m_pullup(m, iphlen + sizeof(struct udphdr) + payload_len)) == 0) {
+           uh->uh_dport == ntohs((u_short)esp_udp_encap_port)) {
+               int payload_len = len - sizeof (struct udphdr) > 4 ? 4 :
+                   len - sizeof (struct udphdr);
+
+               if (m->m_len < iphlen + sizeof (struct udphdr) + payload_len) {
+                       if ((m = m_pullup(m, iphlen + sizeof (struct udphdr) +
+                           payload_len)) == NULL) {
                                udpstat.udps_hdrops++;
-                               KERNEL_DEBUG(DBG_FNC_UDP_INPUT | DBG_FUNC_END, 0,0,0,0,0);
+                               KERNEL_DEBUG(DBG_FNC_UDP_INPUT | DBG_FUNC_END,
+                                   0,0,0,0,0);
                                return;
                        }
+                       /*
+                        * Expect 32-bit aligned data pointer on strict-align
+                        * platforms.
+                        */
+                       MBUF_STRICT_DATA_ALIGNMENT_CHECK_32(m);
+
                        ip = mtod(m, struct ip *);
-                       uh = (struct udphdr *)((caddr_t)ip + iphlen);
+                       uh = (struct udphdr *)(void *)((caddr_t)ip + iphlen);
                }
                /* Check for NAT keepalive packet */
-               if (payload_len == 1 && *(u_int8_t*)((caddr_t)uh + sizeof(struct udphdr)) == 0xFF) {
+               if (payload_len == 1 && *(u_int8_t*)
+                   ((caddr_t)uh + sizeof (struct udphdr)) == 0xFF) {
                        m_freem(m);
-                       KERNEL_DEBUG(DBG_FNC_UDP_INPUT | DBG_FUNC_END, 0,0,0,0,0);
+                       KERNEL_DEBUG(DBG_FNC_UDP_INPUT | DBG_FUNC_END,
+                           0,0,0,0,0);
                        return;
-               }
-               else if (payload_len == 4 && *(u_int32_t*)((caddr_t)uh + sizeof(struct udphdr)) != 0) {
+               } else if (payload_len == 4 && *(u_int32_t*)(void *)
+                   ((caddr_t)uh + sizeof (struct udphdr)) != 0) {
                        /* UDP encapsulated IPSec packet to pass through NAT */
-                       size_t stripsiz;
-
-                       stripsiz = sizeof(struct udphdr);
-
-                       ip = mtod(m, struct ip *);
-                       ovbcopy((caddr_t)ip, (caddr_t)(((u_char *)ip) + stripsiz), iphlen);
-                       m->m_data += stripsiz;
-                       m->m_len -= stripsiz;
-                       m->m_pkthdr.len -= stripsiz;
-                       ip = mtod(m, struct ip *);
-                       ip->ip_len = ip->ip_len - stripsiz;
-                       ip->ip_p = IPPROTO_ESP;
-
-                       KERNEL_DEBUG(DBG_FNC_UDP_INPUT | DBG_FUNC_END, 0,0,0,0,0);
-                       esp4_input(m, iphlen);
+                       KERNEL_DEBUG(DBG_FNC_UDP_INPUT | DBG_FUNC_END,
+                           0,0,0,0,0);
+                       /* preserve the udp header */
+                       esp4_input(m, iphlen + sizeof (struct udphdr));
                        return;
                }
        }
-#endif
+#endif /* IPSEC */
 
        /*
         * Locate pcb for datagram.
         */
        inp = in_pcblookup_hash(&udbinfo, ip->ip_src, uh->uh_sport,
-           ip->ip_dst, uh->uh_dport, 1, m->m_pkthdr.rcvif);
+           ip->ip_dst, uh->uh_dport, 1, ifp);
        if (inp == NULL) {
-               if (log_in_vain) {
+               IF_UDP_STATINC(ifp, port_unreach);
+
+               if (udp_log_in_vain) {
                        char buf[MAX_IPv4_STR_LEN];
                        char buf2[MAX_IPv4_STR_LEN];
 
                        /* check src and dst address */
-                       if (log_in_vain != 3)
-                               log(LOG_INFO,
-                                       "Connection attempt to UDP %s:%d from %s:%d\n",
-                                       inet_ntop(AF_INET, &ip->ip_dst, buf, sizeof(buf)),
-                                       ntohs(uh->uh_dport),
-                                       inet_ntop(AF_INET, &ip->ip_src, buf2, sizeof(buf2)),
+                       if (udp_log_in_vain < 3) {
+                               log(LOG_INFO, "Connection attempt to "
+                                   "UDP %s:%d from %s:%d\n", inet_ntop(AF_INET,
+                                       &ip->ip_dst, buf, sizeof (buf)),
+                                       ntohs(uh->uh_dport), inet_ntop(AF_INET,
+                                       &ip->ip_src, buf2, sizeof (buf2)),
                                        ntohs(uh->uh_sport));
-                       else if (!(m->m_flags & (M_BCAST | M_MCAST)) &&
-                                        ip->ip_dst.s_addr != ip->ip_src.s_addr)
+                       else if (!(m->m_flags & (M_BCAST | M_MCAST)) &&
+                           ip->ip_dst.s_addr != ip->ip_src.s_addr) {
                                log_in_vain_log((LOG_INFO,
-                                       "Stealth Mode connection attempt to UDP %s:%d from %s:%d\n",
-                                       inet_ntop(AF_INET, &ip->ip_dst, buf, sizeof(buf)),
-                                       ntohs(uh->uh_dport),
-                                       inet_ntop(AF_INET, &ip->ip_src, buf2, sizeof(buf2)),
-                                       ntohs(uh->uh_sport)))
+                                   "Stealth Mode connection attempt to "
+                                   "UDP %s:%d from %s:%d\n", inet_ntop(AF_INET,
+                                   &ip->ip_dst, buf, sizeof (buf)),
+                                   ntohs(uh->uh_dport), inet_ntop(AF_INET,
+                                   &ip->ip_src, buf2, sizeof (buf2)),
+                                   ntohs(uh->uh_sport)))
+                       }
                }
                udpstat.udps_noport++;
                if (m->m_flags & (M_BCAST | M_MCAST)) {
@@ -574,9 +666,9 @@ doudpcksum:
 #if ICMP_BANDLIM
                if (badport_bandlim(BANDLIM_ICMP_UNREACH) < 0)
                        goto bad;
-#endif
+#endif /* ICMP_BANDLIM */
                if (blackhole)
-                       if (m->m_pkthdr.rcvif && m->m_pkthdr.rcvif->if_type != IFT_LOOP)
+                       if (ifp && ifp->if_type != IFT_LOOP)
                                goto bad;
                *ip = save_ip;
                ip->ip_len += iphlen;
@@ -588,20 +680,19 @@ doudpcksum:
 
        if (in_pcb_checkstate(inp, WNT_RELEASE, 1) == WNT_STOPUSING) {
                udp_unlock(inp->inp_socket, 1, 0);
+               IF_UDP_STATINC(ifp, cleanup);
                goto bad;
        }
 #if IPSEC
        if (ipsec_bypass == 0 && inp != NULL) {
-               lck_mtx_lock(sadb_mutex);
                if (ipsec4_in_reject_so(m, inp->inp_socket)) {
-                       ipsecstat.in_polvio++;
-                       lck_mtx_unlock(sadb_mutex);
+                       IPSEC_STAT_INCREMENT(ipsecstat.in_polvio);
                        udp_unlock(inp->inp_socket, 1, 0);
+                       IF_UDP_STATINC(ifp, badipsec);
                        goto bad;
-               }
-               lck_mtx_unlock(sadb_mutex);
+               }
        }
-#endif /*IPSEC*/
+#endif /* IPSEC */
 
        /*
         * Construct sockaddr format source address.
@@ -609,8 +700,9 @@ doudpcksum:
         */
        udp_in.sin_port = uh->uh_sport;
        udp_in.sin_addr = ip->ip_src;
-       if (inp->inp_flags & INP_CONTROLOPTS
-           || inp->inp_socket->so_options & SO_TIMESTAMP) {
+       if ((inp->inp_flags & INP_CONTROLOPTS) != 0 ||
+           (inp->inp_socket->so_options & SO_TIMESTAMP) != 0 ||
+           (inp->inp_socket->so_options & SO_TIMESTAMP_MONOTONIC) != 0) {
 #if INET6
                if (inp->inp_vflag & INP_IPV6) {
                        int savedflags;
@@ -618,28 +710,41 @@ doudpcksum:
                        ip_2_ip6_hdr(&udp_ip6.uip6_ip6, ip);
                        savedflags = inp->inp_flags;
                        inp->inp_flags &= ~INP_UNMAPPABLEOPTS;
-                       ip6_savecontrol(inp, &opts, &udp_ip6.uip6_ip6, m);
+                       ret = ip6_savecontrol(inp, m, &opts);
                        inp->inp_flags = savedflags;
                } else
-#endif
-               ip_savecontrol(inp, &opts, ip, m);
+#endif /* INET6 */
+               {
+                       ret = ip_savecontrol(inp, &opts, ip, m);
+               }
+               if (ret != 0) {
+                       udp_unlock(inp->inp_socket, 1, 0);
+                       goto bad;
+               }
        }
-       m_adj(m, iphlen + sizeof(struct udphdr));
+       m_adj(m, iphlen + sizeof (struct udphdr));
 
        KERNEL_DEBUG(DBG_LAYER_IN_END, uh->uh_dport, uh->uh_sport,
-                    save_ip.ip_src.s_addr, save_ip.ip_dst.s_addr, uh->uh_ulen);
+           save_ip.ip_src.s_addr, save_ip.ip_dst.s_addr, uh->uh_ulen);
 
 #if INET6
        if (inp->inp_vflag & INP_IPV6) {
                in6_sin_2_v4mapsin6(&udp_in, &udp_in6.uin6_sin);
-               append_sa = (struct sockaddr *)&udp_in6;
+               append_sa = (struct sockaddr *)&udp_in6.uin6_sin;
        } else
-#endif
-       append_sa = (struct sockaddr *)&udp_in;
-       if (sbappendaddr(&inp->inp_socket->so_rcv, append_sa, m, opts, NULL) == 0) {
-               udpstat.udps_fullsock++;
+#endif /* INET6 */
+       {
+               append_sa = (struct sockaddr *)&udp_in;
+       }
+       if (nstat_collect) {
+               INP_ADD_STAT(inp, cell, wifi, rxpackets, 1);
+               INP_ADD_STAT(inp, cell, wifi, rxbytes, m->m_pkthdr.len);
        }
-       else {
+       so_recv_data_stat(inp->inp_socket, m, 0);
+       if (sbappendaddr(&inp->inp_socket->so_rcv, append_sa,
+           m, opts, NULL) == 0) {
+               udpstat.udps_fullsock++;
+       } else {
                sorwakeup(inp->inp_socket);
        }
        udp_unlock(inp->inp_socket, 1, 0);
@@ -650,75 +755,109 @@ bad:
        if (opts)
                m_freem(opts);
        KERNEL_DEBUG(DBG_FNC_UDP_INPUT | DBG_FUNC_END, 0,0,0,0,0);
-       return;
 }
 
 #if INET6
 static void
-ip_2_ip6_hdr(ip6, ip)
-       struct ip6_hdr *ip6;
-       struct ip *ip;
+ip_2_ip6_hdr(struct ip6_hdr *ip6, struct ip *ip)
 {
-       bzero(ip6, sizeof(*ip6));
+       bzero(ip6, sizeof (*ip6));
 
        ip6->ip6_vfc = IPV6_VERSION;
        ip6->ip6_plen = ip->ip_len;
        ip6->ip6_nxt = ip->ip_p;
        ip6->ip6_hlim = ip->ip_ttl;
-       ip6->ip6_src.s6_addr32[2] = ip6->ip6_dst.s6_addr32[2] =
-               IPV6_ADDR_INT32_SMP;
-       ip6->ip6_src.s6_addr32[3] = ip->ip_src.s_addr;
-       ip6->ip6_dst.s6_addr32[3] = ip->ip_dst.s_addr;
+       if (ip->ip_src.s_addr) {
+               ip6->ip6_src.s6_addr32[2] = IPV6_ADDR_INT32_SMP;
+               ip6->ip6_src.s6_addr32[3] = ip->ip_src.s_addr;
+       }
+       if (ip->ip_dst.s_addr) {
+               ip6->ip6_dst.s6_addr32[2] = IPV6_ADDR_INT32_SMP;
+               ip6->ip6_dst.s6_addr32[3] = ip->ip_dst.s_addr;
+       }
 }
-#endif
+#endif /* INET6 */
 
 /*
  * subroutine of udp_input(), mainly for source code readability.
- * caller must properly init udp_ip6 and udp_in6 beforehand.
  */
 static void
-udp_append(last, ip, n, off)
-       struct inpcb *last;
-       struct ip *ip;
-       struct mbuf *n;
-       int off;
+#if INET6
+udp_append(struct inpcb *last, struct ip *ip, struct mbuf *n, int off,
+    struct sockaddr_in *pudp_in, struct udp_in6 *pudp_in6,
+    struct udp_ip6 *pudp_ip6, struct ifnet *ifp)
+#else /* !INET6 */
+udp_append(struct inpcb *last, struct ip *ip, struct mbuf *n, int off,
+    struct sockaddr_in *pudp_in, struct ifnet *ifp)
+#endif /* !INET6 */
 {
        struct sockaddr *append_sa;
        struct mbuf *opts = 0;
+       boolean_t cell = IFNET_IS_CELLULAR(ifp);
+       boolean_t wifi = (!cell && IFNET_IS_WIFI(ifp));
+       int ret = 0;
 
-       if (last->inp_flags & INP_CONTROLOPTS ||
-           last->inp_socket->so_options & SO_TIMESTAMP) {
+#if CONFIG_MACF_NET
+       if (mac_inpcb_check_deliver(last, n, AF_INET, SOCK_DGRAM) != 0) {
+               m_freem(n);
+               return;
+       }
+#endif /* CONFIG_MACF_NET */
+       if ((last->inp_flags & INP_CONTROLOPTS) != 0 ||
+           (last->inp_socket->so_options & SO_TIMESTAMP) != 0 ||
+           (last->inp_socket->so_options & SO_TIMESTAMP_MONOTONIC) != 0) {
 #if INET6
                if (last->inp_vflag & INP_IPV6) {
                        int savedflags;
 
-                       if (udp_ip6.uip6_init_done == 0) {
-                               ip_2_ip6_hdr(&udp_ip6.uip6_ip6, ip);
-                               udp_ip6.uip6_init_done = 1;
+                       if (pudp_ip6->uip6_init_done == 0) {
+                               ip_2_ip6_hdr(&pudp_ip6->uip6_ip6, ip);
+                               pudp_ip6->uip6_init_done = 1;
                        }
                        savedflags = last->inp_flags;
                        last->inp_flags &= ~INP_UNMAPPABLEOPTS;
-                       ip6_savecontrol(last, &opts, &udp_ip6.uip6_ip6, n);
+                       ret = ip6_savecontrol(last, n, &opts);
+                       if (ret != 0) {
+                               last->inp_flags = savedflags;
+                               goto error;
+                       }
                        last->inp_flags = savedflags;
                } else
-#endif
-               ip_savecontrol(last, &opts, ip, n);
+#endif /* INET6 */
+               {
+                       ret = ip_savecontrol(last, &opts, ip, n);
+                       if (ret != 0) {
+                               goto error;
+                       }
+               }
        }
 #if INET6
        if (last->inp_vflag & INP_IPV6) {
-               if (udp_in6.uin6_init_done == 0) {
-                       in6_sin_2_v4mapsin6(&udp_in, &udp_in6.uin6_sin);
-                       udp_in6.uin6_init_done = 1;
+               if (pudp_in6->uin6_init_done == 0) {
+                       in6_sin_2_v4mapsin6(pudp_in, &pudp_in6->uin6_sin);
+                       pudp_in6->uin6_init_done = 1;
                }
-               append_sa = (struct sockaddr *)&udp_in6.uin6_sin;
+               append_sa = (struct sockaddr *)&pudp_in6->uin6_sin;
        } else
-#endif
-       append_sa = (struct sockaddr *)&udp_in;
+#endif /* INET6 */
+       append_sa = (struct sockaddr *)pudp_in;
+       if (nstat_collect) {
+               INP_ADD_STAT(last, cell, wifi, rxpackets, 1);
+               INP_ADD_STAT(last, cell, wifi, rxbytes, n->m_pkthdr.len);
+       }
+       so_recv_data_stat(last->inp_socket, n, 0);
        m_adj(n, off);
-       if (sbappendaddr(&last->inp_socket->so_rcv, append_sa, n, opts, NULL) == 0) {
+       if (sbappendaddr(&last->inp_socket->so_rcv, append_sa,
+           n, opts, NULL) == 0) {
                udpstat.udps_fullsock++;
-       } else
+       } else {
                sorwakeup(last->inp_socket);
+       }
+       return;
+error:
+       m_freem(n);
+       m_freem(opts);
+       return;
 }
 
 /*
@@ -726,9 +865,7 @@ udp_append(last, ip, n, off)
  * just wake up so that he can collect error status.
  */
 void
-udp_notify(inp, errno)
-       register struct inpcb *inp;
-       int errno;
+udp_notify(struct inpcb *inp, int errno)
 {
        inp->inp_socket->so_error = errno;
        sorwakeup(inp->inp_socket);
@@ -736,48 +873,115 @@ udp_notify(inp, errno)
 }
 
 void
-udp_ctlinput(cmd, sa, vip)
-       int cmd;
-       struct sockaddr *sa;
-       void *vip;
+udp_ctlinput(int cmd, struct sockaddr *sa, void *vip)
 {
        struct ip *ip = vip;
-       struct udphdr *uh;
        void (*notify)(struct inpcb *, int) = udp_notify;
         struct in_addr faddr;
        struct inpcb *inp;
 
-       faddr = ((struct sockaddr_in *)sa)->sin_addr;
+       faddr = ((struct sockaddr_in *)(void *)sa)->sin_addr;
        if (sa->sa_family != AF_INET || faddr.s_addr == INADDR_ANY)
-               return;
+               return;
 
        if (PRC_IS_REDIRECT(cmd)) {
                ip = 0;
                notify = in_rtchange;
-       } else if (cmd == PRC_HOSTDEAD)
+       } else if (cmd == PRC_HOSTDEAD) {
                ip = 0;
-       else if ((unsigned)cmd >= PRC_NCMDS || inetctlerrmap[cmd] == 0)
+       } else if ((unsigned)cmd >= PRC_NCMDS || inetctlerrmap[cmd] == 0) {
                return;
+       }
        if (ip) {
-               uh = (struct udphdr *)((caddr_t)ip + (ip->ip_hl << 2));
-               inp = in_pcblookup_hash(&udbinfo, faddr, uh->uh_dport,
-                    ip->ip_src, uh->uh_sport, 0, NULL);
+               struct udphdr uh;
+
+               bcopy(((caddr_t)ip + (ip->ip_hl << 2)), &uh, sizeof (uh));
+               inp = in_pcblookup_hash(&udbinfo, faddr, uh.uh_dport,
+                    ip->ip_src, uh.uh_sport, 0, NULL);
                if (inp != NULL && inp->inp_socket != NULL) {
                        udp_lock(inp->inp_socket, 1, 0);
-                       if (in_pcb_checkstate(inp, WNT_RELEASE, 1) == WNT_STOPUSING)  {
+                       if (in_pcb_checkstate(inp, WNT_RELEASE, 1) ==
+                           WNT_STOPUSING)  {
                                udp_unlock(inp->inp_socket, 1, 0);
                                return;
                        }
                        (*notify)(inp, inetctlerrmap[cmd]);
                        udp_unlock(inp->inp_socket, 1, 0);
                }
-       } else
+       } else {
                in_pcbnotifyall(&udbinfo, faddr, inetctlerrmap[cmd], notify);
+       }
+}
+
+int
+udp_ctloutput(struct socket *so, struct sockopt *sopt)
+{
+       int     error, optval;
+       struct  inpcb *inp;
+
+       /* Allow <SOL_SOCKET,SO_FLUSH> at this level */
+       if (sopt->sopt_level != IPPROTO_UDP &&
+           !(sopt->sopt_level == SOL_SOCKET && sopt->sopt_name == SO_FLUSH))
+               return (ip_ctloutput(so, sopt));
+
+       error = 0;
+       inp = sotoinpcb(so);
+
+       switch (sopt->sopt_dir) {
+       case SOPT_SET:
+               switch (sopt->sopt_name) {
+               case UDP_NOCKSUM:
+                       /* This option is settable only for UDP over IPv4 */
+                       if (!(inp->inp_vflag & INP_IPV4)) {
+                               error = EINVAL;
+                               break;
+                       }
+
+                       if ((error = sooptcopyin(sopt, &optval, sizeof (optval),
+                           sizeof (optval))) != 0)
+                               break;
+
+                       if (optval != 0)
+                               inp->inp_flags |= INP_UDP_NOCKSUM;
+                       else
+                               inp->inp_flags &= ~INP_UDP_NOCKSUM;
+                       break;
+
+               case SO_FLUSH:
+                       if ((error = sooptcopyin(sopt, &optval, sizeof (optval),
+                           sizeof (optval))) != 0)
+                               break;
+
+                       error = inp_flush(inp, optval);
+                       break;
+
+               default:
+                       error = ENOPROTOOPT;
+                       break;
+               }
+               break;
+
+       case SOPT_GET:
+               switch (sopt->sopt_name) {
+               case UDP_NOCKSUM:
+                       optval = inp->inp_flags & INP_UDP_NOCKSUM;
+                       break;
+
+               default:
+                       error = ENOPROTOOPT;
+                       break;
+               }
+               if (error == 0)
+                       error = sooptcopyout(sopt, &optval, sizeof (optval));
+               break;
+       }
+       return (error);
 }
 
 static int
 udp_pcblist SYSCTL_HANDLER_ARGS
 {
+#pragma unused(oidp, arg1, arg2)
        int error, i, n;
        struct inpcb *inp, **inp_list;
        inp_gen_t gencnt;
@@ -787,18 +991,18 @@ udp_pcblist SYSCTL_HANDLER_ARGS
         * The process of preparing the TCB list is too time-consuming and
         * resource-intensive to repeat twice on every request.
         */
-       lck_rw_lock_exclusive(udbinfo.mtx);
+       lck_rw_lock_exclusive(udbinfo.ipi_lock);
        if (req->oldptr == USER_ADDR_NULL) {
                n = udbinfo.ipi_count;
-               req->oldidx = 2 * (sizeof xig)
-                       + (n + n/8) * sizeof(struct xinpcb);
-               lck_rw_done(udbinfo.mtx);
-               return 0;
+               req->oldidx = 2 * (sizeof (xig))
+                       + (n + n/8) * sizeof (struct xinpcb);
+               lck_rw_done(udbinfo.ipi_lock);
+               return (0);
        }
 
        if (req->newptr != USER_ADDR_NULL) {
-               lck_rw_done(udbinfo.mtx);
-               return EPERM;
+               lck_rw_done(udbinfo.ipi_lock);
+               return (EPERM);
        }
 
        /*
@@ -807,33 +1011,34 @@ udp_pcblist SYSCTL_HANDLER_ARGS
        gencnt = udbinfo.ipi_gencnt;
        n = udbinfo.ipi_count;
 
-       bzero(&xig, sizeof(xig));
-       xig.xig_len = sizeof xig;
+       bzero(&xig, sizeof (xig));
+       xig.xig_len = sizeof (xig);
        xig.xig_count = n;
        xig.xig_gen = gencnt;
        xig.xig_sogen = so_gencnt;
-       error = SYSCTL_OUT(req, &xig, sizeof xig);
+       error = SYSCTL_OUT(req, &xig, sizeof (xig));
        if (error) {
-               lck_rw_done(udbinfo.mtx);
-               return error;
+               lck_rw_done(udbinfo.ipi_lock);
+               return (error);
        }
-    /*
-     * We are done if there is no pcb
-     */
-    if (n == 0) {
-       lck_rw_done(udbinfo.mtx);
-        return 0; 
-    }
-
-       inp_list = _MALLOC(n * sizeof *inp_list, M_TEMP, M_WAITOK);
+       /*
+        * We are done if there is no pcb
+        */
+       if (n == 0) {
+               lck_rw_done(udbinfo.ipi_lock);
+               return (0);
+       }
+
+       inp_list = _MALLOC(n * sizeof (*inp_list), M_TEMP, M_WAITOK);
        if (inp_list == 0) {
-               lck_rw_done(udbinfo.mtx);
-               return ENOMEM;
+               lck_rw_done(udbinfo.ipi_lock);
+               return (ENOMEM);
        }
-       
-       for (inp = LIST_FIRST(udbinfo.listhead), i = 0; inp && i < n;
+
+       for (inp = LIST_FIRST(udbinfo.ipi_listhead), i = 0; inp && i < n;
             inp = LIST_NEXT(inp, inp_list)) {
-               if (inp->inp_gencnt <= gencnt && inp->inp_state != INPCB_STATE_DEAD)
+               if (inp->inp_gencnt <= gencnt &&
+                   inp->inp_state != INPCB_STATE_DEAD)
                        inp_list[i++] = inp;
        }
        n = i;
@@ -841,16 +1046,17 @@ udp_pcblist SYSCTL_HANDLER_ARGS
        error = 0;
        for (i = 0; i < n; i++) {
                inp = inp_list[i];
-               if (inp->inp_gencnt <= gencnt && inp->inp_state != INPCB_STATE_DEAD) {
+               if (inp->inp_gencnt <= gencnt &&
+                   inp->inp_state != INPCB_STATE_DEAD) {
                        struct xinpcb xi;
 
-                       bzero(&xi, sizeof(xi));
-                       xi.xi_len = sizeof xi;
+                       bzero(&xi, sizeof (xi));
+                       xi.xi_len = sizeof (xi);
                        /* XXX should avoid extra copy */
                        inpcb_to_compat(inp, &xi.xi_inp);
                        if (inp->inp_socket)
                                sotoxsocket(inp->inp_socket, &xi.xi_socket);
-                       error = SYSCTL_OUT(req, &xi, sizeof xi);
+                       error = SYSCTL_OUT(req, &xi, sizeof (xi));
                }
        }
        if (!error) {
@@ -861,131 +1067,434 @@ udp_pcblist SYSCTL_HANDLER_ARGS
                 * while we were processing this request, and it
                 * might be necessary to retry.
                 */
-               bzero(&xig, sizeof(xig));
-               xig.xig_len = sizeof xig;
+               bzero(&xig, sizeof (xig));
+               xig.xig_len = sizeof (xig);
                xig.xig_gen = udbinfo.ipi_gencnt;
                xig.xig_sogen = so_gencnt;
                xig.xig_count = udbinfo.ipi_count;
-               error = SYSCTL_OUT(req, &xig, sizeof xig);
+               error = SYSCTL_OUT(req, &xig, sizeof (xig));
        }
        FREE(inp_list, M_TEMP);
-       lck_rw_done(udbinfo.mtx);
-       return error;
+       lck_rw_done(udbinfo.ipi_lock);
+       return (error);
 }
 
-SYSCTL_PROC(_net_inet_udp, UDPCTL_PCBLIST, pcblist, CTLFLAG_RD, 0, 0,
-           udp_pcblist, "S,xinpcb", "List of active UDP sockets");
-
+SYSCTL_PROC(_net_inet_udp, UDPCTL_PCBLIST, pcblist,
+    CTLFLAG_RD | CTLFLAG_LOCKED, 0, 0, udp_pcblist,
+    "S,xinpcb", "List of active UDP sockets");
 
 
-static __inline__ u_int16_t
-get_socket_id(struct socket * s)
+static int
+udp_pcblist64 SYSCTL_HANDLER_ARGS
 {
-       u_int16_t               val;
+#pragma unused(oidp, arg1, arg2)
+        int error, i, n;
+        struct inpcb *inp, **inp_list;
+        inp_gen_t gencnt;
+        struct xinpgen xig;
+
+        /*
+         * The process of preparing the TCB list is too time-consuming and
+         * resource-intensive to repeat twice on every request.
+         */
+        lck_rw_lock_shared(udbinfo.ipi_lock);
+        if (req->oldptr == USER_ADDR_NULL) {
+                n = udbinfo.ipi_count;
+                req->oldidx =
+                   2 * (sizeof (xig)) + (n + n/8) * sizeof (struct xinpcb64);
+                lck_rw_done(udbinfo.ipi_lock);
+                return (0);
+        }
+
+        if (req->newptr != USER_ADDR_NULL) {
+                lck_rw_done(udbinfo.ipi_lock);
+                return (EPERM);
+        }
 
-       if (s == NULL) {
-           return (0);
+        /*
+         * OK, now we're committed to doing something.
+         */
+        gencnt = udbinfo.ipi_gencnt;
+        n = udbinfo.ipi_count;
+
+        bzero(&xig, sizeof (xig));
+        xig.xig_len = sizeof (xig);
+        xig.xig_count = n;
+        xig.xig_gen = gencnt;
+        xig.xig_sogen = so_gencnt;
+        error = SYSCTL_OUT(req, &xig, sizeof (xig));
+        if (error) {
+                lck_rw_done(udbinfo.ipi_lock);
+                return (error);
+        }
+       /*
+        * We are done if there is no pcb
+        */
+       if (n == 0) {
+               lck_rw_done(udbinfo.ipi_lock);
+               return (0);
        }
-       val = (u_int16_t)(((u_int32_t)s) / sizeof(struct socket));
-       if (val == 0) {
-               val = 0xffff;
+
+        inp_list = _MALLOC(n * sizeof (*inp_list), M_TEMP, M_WAITOK);
+        if (inp_list == 0) {
+                lck_rw_done(udbinfo.ipi_lock);
+                return (ENOMEM);
+        }
+
+        for (inp = LIST_FIRST(udbinfo.ipi_listhead), i = 0; inp && i < n;
+             inp = LIST_NEXT(inp, inp_list)) {
+                if (inp->inp_gencnt <= gencnt &&
+                   inp->inp_state != INPCB_STATE_DEAD)
+                        inp_list[i++] = inp;
+        }
+        n = i;
+
+        error = 0;
+        for (i = 0; i < n; i++) {
+                inp = inp_list[i];
+                if (inp->inp_gencnt <= gencnt &&
+                   inp->inp_state != INPCB_STATE_DEAD) {
+                        struct xinpcb64 xi;
+
+                        bzero(&xi, sizeof (xi));
+                        xi.xi_len = sizeof (xi);
+                        inpcb_to_xinpcb64(inp, &xi);
+                        if (inp->inp_socket)
+                                sotoxsocket64(inp->inp_socket, &xi.xi_socket);
+                        error = SYSCTL_OUT(req, &xi, sizeof (xi));
+                }
+        }
+        if (!error) {
+                /*
+                 * Give the user an updated idea of our state.
+                 * If the generation differs from what we told
+                 * her before, she knows that something happened
+                 * while we were processing this request, and it
+                 * might be necessary to retry.
+                 */
+                bzero(&xig, sizeof (xig));
+                xig.xig_len = sizeof (xig);
+                xig.xig_gen = udbinfo.ipi_gencnt;
+                xig.xig_sogen = so_gencnt;
+                xig.xig_count = udbinfo.ipi_count;
+                error = SYSCTL_OUT(req, &xig, sizeof (xig));
+        }
+        FREE(inp_list, M_TEMP);
+        lck_rw_done(udbinfo.ipi_lock);
+        return (error);
+}
+
+SYSCTL_PROC(_net_inet_udp, OID_AUTO, pcblist64,
+    CTLFLAG_RD | CTLFLAG_LOCKED, 0, 0, udp_pcblist64,
+    "S,xinpcb64", "List of active UDP sockets");
+
+
+static int
+udp_pcblist_n SYSCTL_HANDLER_ARGS
+{
+#pragma unused(oidp, arg1, arg2)
+       return (get_pcblist_n(IPPROTO_UDP, req, &udbinfo));
+}
+
+SYSCTL_PROC(_net_inet_udp, OID_AUTO, pcblist_n,
+    CTLFLAG_RD | CTLFLAG_LOCKED, 0, 0, udp_pcblist_n,
+    "S,xinpcb_n", "List of active UDP sockets");
+
+__private_extern__ void
+udp_get_ports_used(uint32_t ifindex, int protocol, uint32_t wildcardok,
+    bitstr_t *bitfield)
+{
+       inpcb_get_ports_used(ifindex, protocol, wildcardok, bitfield, &udbinfo);
+}
+
+__private_extern__ uint32_t
+udp_count_opportunistic(unsigned int ifindex, u_int32_t flags)
+{
+       return (inpcb_count_opportunistic(ifindex, &udbinfo, flags));
+}
+
+__private_extern__ uint32_t
+udp_find_anypcb_byaddr(struct ifaddr *ifa)
+{
+       return (inpcb_find_anypcb_byaddr(ifa, &udbinfo));
+}
+
+static int
+udp_check_pktinfo(struct mbuf *control, struct ifnet **outif,
+    struct in_addr *laddr)
+{
+       struct cmsghdr *cm = 0;
+       struct in_pktinfo *pktinfo;
+       struct ifnet *ifp;
+
+       if (outif != NULL)
+               *outif = NULL;
+
+       /*
+        * XXX: Currently, we assume all the optional information is stored
+        * in a single mbuf.
+        */
+       if (control->m_next)
+               return (EINVAL);
+
+       if (control->m_len < CMSG_LEN(0))
+               return (EINVAL);
+
+       for (cm = M_FIRST_CMSGHDR(control); cm;
+           cm = M_NXT_CMSGHDR(control, cm)) {
+               if (cm->cmsg_len < sizeof (struct cmsghdr) ||
+                   cm->cmsg_len > control->m_len)
+                       return (EINVAL);
+
+               if (cm->cmsg_level != IPPROTO_IP || cm->cmsg_type != IP_PKTINFO)
+                       continue;
+
+               if (cm->cmsg_len != CMSG_LEN(sizeof (struct in_pktinfo)))
+                       return (EINVAL);
+
+               pktinfo =  (struct in_pktinfo *)(void *)CMSG_DATA(cm);
+
+               /* Check for a valid ifindex in pktinfo */
+               ifnet_head_lock_shared();
+
+               if (pktinfo->ipi_ifindex > if_index) {
+                       ifnet_head_done();
+                       return (ENXIO);
+               }
+
+               /*
+                * If ipi_ifindex is specified it takes precedence
+                * over ipi_spec_dst.
+                */
+               if (pktinfo->ipi_ifindex) {
+                       ifp = ifindex2ifnet[pktinfo->ipi_ifindex];
+                       if (ifp == NULL) {
+                               ifnet_head_done();
+                               return (ENXIO);
+                       }
+                       if (outif != NULL) {
+                               ifnet_reference(ifp);
+                               *outif = ifp;
+                       }
+                       ifnet_head_done();
+                       laddr->s_addr = INADDR_ANY;
+                       break;
+               }
+
+               ifnet_head_done();
+
+               /*
+                * Use the provided ipi_spec_dst address for temp
+                * source address.
+                */
+               *laddr = pktinfo->ipi_spec_dst;
+               break;
        }
-       return (val);
+       return (0);
 }
 
 static int
-udp_output(inp, m, addr, control, p)
-       register struct inpcb *inp;
-       struct mbuf *m;
-       struct sockaddr *addr;
-       struct mbuf *control;
-       struct proc *p;
+udp_output(struct inpcb *inp, struct mbuf *m, struct sockaddr *addr,
+    struct mbuf *control, struct proc *p)
 {
-       register struct udpiphdr *ui;
-       register int len = m->m_pkthdr.len;
-       struct sockaddr_in *sin, src;
-       struct in_addr origladdr, laddr, faddr;
+       struct udpiphdr *ui;
+       int len = m->m_pkthdr.len;
+       struct sockaddr_in *sin;
+       struct in_addr origladdr, laddr, faddr, pi_laddr;
        u_short lport, fport;
-       struct sockaddr_in *ifaddr;
-       int error = 0, udp_dodisconnect = 0;
-
+       int error = 0, udp_dodisconnect = 0, pktinfo = 0;
+       struct socket *so = inp->inp_socket;
+       int soopts = 0;
+       struct mbuf *inpopts;
+       struct ip_moptions *mopts;
+       struct route ro;
+       struct ip_out_args ipoa =
+           { IFSCOPE_NONE, { 0 }, IPOAF_SELECT_SRCIF, 0 };
+       struct ifnet *outif = NULL;
+       struct flowadv *adv = &ipoa.ipoa_flowadv;
+       mbuf_svc_class_t msc = MBUF_SC_UNSPEC;
+       struct ifnet *origoutifp;
+       int flowadv = 0;
+
+       /* Enable flow advisory only when connected */
+       flowadv = (so->so_state & SS_ISCONNECTED) ? 1 : 0;
+       pi_laddr.s_addr = INADDR_ANY;
 
        KERNEL_DEBUG(DBG_FNC_UDP_OUTPUT | DBG_FUNC_START, 0,0,0,0,0);
 
-       if (control)
-               m_freem(control);               /* XXX */
+       lck_mtx_assert(&inp->inpcb_mtx, LCK_MTX_ASSERT_OWNED);
+       if (control != NULL) {
+               msc = mbuf_service_class_from_control(control);
+               VERIFY(outif == NULL);
+               error = udp_check_pktinfo(control, &outif, &pi_laddr);
+               m_freem(control);
+               control = NULL;
+               if (error)
+                       goto release;
+               pktinfo++;
+               if (outif != NULL)
+                       ipoa.ipoa_boundif = outif->if_index;
+       }
 
        KERNEL_DEBUG(DBG_LAYER_OUT_BEG, inp->inp_fport, inp->inp_lport,
-                    inp->inp_laddr.s_addr, inp->inp_faddr.s_addr,
-                    (htons((u_short)len + sizeof (struct udphdr))));
+           inp->inp_laddr.s_addr, inp->inp_faddr.s_addr,
+           (htons((u_short)len + sizeof (struct udphdr))));
 
-       if (len + sizeof(struct udpiphdr) > IP_MAXPACKET) {
+       if (len + sizeof (struct udpiphdr) > IP_MAXPACKET) {
                error = EMSGSIZE;
                goto release;
        }
 
-       /* If there was a routing change, discard cached route and check
-        * that we have a valid source address. 
-        * Reacquire a new source address if INADDR_ANY was specified
+       if (flowadv && INP_WAIT_FOR_IF_FEEDBACK(inp)) {
+               /*
+                * The socket is flow-controlled, drop the packets
+                * until the inp is not flow controlled
+                */
+               error = ENOBUFS;
+               goto release;
+       }
+       /*
+        * If socket was bound to an ifindex, tell ip_output about it.
+        * If the ancillary IP_PKTINFO option contains an interface index,
+        * it takes precedence over the one specified by IP_BOUND_IF.
         */
+       if (ipoa.ipoa_boundif == IFSCOPE_NONE &&
+           (inp->inp_flags & INP_BOUND_IF)) {
+               VERIFY(inp->inp_boundifp != NULL);
+               ifnet_reference(inp->inp_boundifp);     /* for this routine */
+               if (outif != NULL)
+                       ifnet_release(outif);
+               outif = inp->inp_boundifp;
+               ipoa.ipoa_boundif = outif->if_index;
+       }
+       if (inp->inp_flags & INP_NO_IFT_CELLULAR)
+               ipoa.ipoa_flags |=  IPOAF_NO_CELLULAR;
+       soopts |= IP_OUTARGS;
 
-#if 1
-        lck_mtx_assert(inp->inpcb_mtx, LCK_MTX_ASSERT_OWNED);
-#endif
-
-       if (inp->inp_route.ro_rt && inp->inp_route.ro_rt->generation_id != route_generation) {
-               if (ifa_foraddr(inp->inp_laddr.s_addr) == 0) { /* src address is gone */
-                       if (inp->inp_flags & INP_INADDR_ANY)
-                               inp->inp_laddr.s_addr = INADDR_ANY; /* new src will be set later */
-                       else {
+       /*
+        * If there was a routing change, discard cached route and check
+        * that we have a valid source address.  Reacquire a new source
+        * address if INADDR_ANY was specified.
+        */
+       if (ROUTE_UNUSABLE(&inp->inp_route)) {
+               struct in_ifaddr *ia = NULL;
+
+               ROUTE_RELEASE(&inp->inp_route);
+
+               /* src address is gone? */
+               if (inp->inp_laddr.s_addr != INADDR_ANY &&
+                   (ia = ifa_foraddr(inp->inp_laddr.s_addr)) == NULL) {
+                       if (!(inp->inp_flags & INP_INADDR_ANY) ||
+                           (so->so_state & SS_ISCONNECTED)) {
+                               /*
+                                * Rdar://5448998
+                                * If the source address is gone, return an
+                                * error if:
+                                * - the source was specified
+                                * - the socket was already connected
+                                */
+                               soevent(so, (SO_FILT_HINT_LOCKED |
+                                   SO_FILT_HINT_NOSRCADDR));
                                error = EADDRNOTAVAIL;
                                goto release;
+                       } else {
+                               /* new src will be set later */
+                               inp->inp_laddr.s_addr = INADDR_ANY;
+                               inp->inp_last_outifp = NULL;
                        }
                }
-               rtfree(inp->inp_route.ro_rt);
-               inp->inp_route.ro_rt = (struct rtentry *)0;
+               if (ia != NULL)
+                       IFA_REMREF(&ia->ia_ifa);
+       }
+
+       origoutifp = inp->inp_last_outifp;
+
+       /*
+        * IP_PKTINFO option check.  If a temporary scope or src address
+        * is provided, use it for this packet only and make sure we forget
+        * it after sending this datagram.
+        */
+       if (pi_laddr.s_addr != INADDR_ANY ||
+           (ipoa.ipoa_boundif != IFSCOPE_NONE && pktinfo)) {
+               /* temp src address for this datagram only */
+               laddr = pi_laddr;
+               origladdr.s_addr = INADDR_ANY;
+               /* we don't want to keep the laddr or route */
+               udp_dodisconnect = 1;
+               /* remember we don't care about src addr.*/
+               inp->inp_flags |= INP_INADDR_ANY;
+       } else {
+               origladdr = laddr = inp->inp_laddr;
        }
 
-       origladdr= laddr = inp->inp_laddr;
+       origoutifp = inp->inp_last_outifp;
        faddr = inp->inp_faddr;
        lport = inp->inp_lport;
        fport = inp->inp_fport;
 
        if (addr) {
-               sin = (struct sockaddr_in *)addr;
+               sin = (struct sockaddr_in *)(void *)addr;
                if (faddr.s_addr != INADDR_ANY) {
                        error = EISCONN;
                        goto release;
                }
                if (lport == 0) {
                        /*
-                        * In case we don't have a local port set, go through the full connect.
-                        * We don't have a local port yet (ie, we can't be looked up),
-                        * so it's not an issue if the input runs at the same time we do this.
-                        */
-                       error = in_pcbconnect(inp, addr, p);
-                       if (error) {
+                        * In case we don't have a local port set, go through
+                        * the full connect.  We don't have a local port yet
+                        * (i.e., we can't be looked up), so it's not an issue
+                        * if the input runs at the same time we do this.
+                        */
+                       /* if we have a source address specified, use that */
+                       if (pi_laddr.s_addr != INADDR_ANY)
+                               inp->inp_laddr = pi_laddr;
+                       /*
+                        * If a scope is specified, use it.  Scope from
+                        * IP_PKTINFO takes precendence over the the scope
+                        * set via INP_BOUND_IF.
+                        */
+                       error = in_pcbconnect(inp, addr, p, ipoa.ipoa_boundif,
+                           &outif);
+                       if (error)
                                goto release;
-                       }
+
                        laddr = inp->inp_laddr;
                        lport = inp->inp_lport;
                        faddr = inp->inp_faddr;
                        fport = inp->inp_fport;
                        udp_dodisconnect = 1;
+
+                       /* synch up in case in_pcbladdr() overrides */
+                       if (outif != NULL && ipoa.ipoa_boundif != IFSCOPE_NONE)
+                               ipoa.ipoa_boundif = outif->if_index;
                }
-               else {  
-                       /* Fast path case
-                        * we have a full address and a local port.
-                        * use those info to build the packet without changing the pcb
-                        * and interfering with the input path. See 3851370
+               else {
+                       /*
+                        * Fast path case
+                        *
+                        * We have a full address and a local port; use those
+                        * info to build the packet without changing the pcb
+                        * and interfering with the input path. See 3851370.
+                        *
+                        * Scope from IP_PKTINFO takes precendence over the
+                        * the scope set via INP_BOUND_IF.
                         */
                        if (laddr.s_addr == INADDR_ANY) {
-                          if ((error = in_pcbladdr(inp, addr, &ifaddr)) != 0)
-                                  goto release;
-                          laddr = ifaddr->sin_addr;
-                          inp->inp_flags |= INP_INADDR_ANY; /* from pcbconnect: remember we don't care about src addr.*/
+                               if ((error = in_pcbladdr(inp, addr, &laddr,
+                                   ipoa.ipoa_boundif, &outif)) != 0)
+                                       goto release;
+                               /*
+                                * from pcbconnect: remember we don't
+                                * care about src addr.
+                                */
+                               inp->inp_flags |= INP_INADDR_ANY;
+
+                               /* synch up in case in_pcbladdr() overrides */
+                               if (outif != NULL &&
+                                   ipoa.ipoa_boundif != IFSCOPE_NONE)
+                                       ipoa.ipoa_boundif = outif->if_index;
                        }
-                                  
+
                        faddr = sin->sin_addr;
                        fport = sin->sin_port;
                }
@@ -996,12 +1505,18 @@ udp_output(inp, m, addr, control, p)
                }
        }
 
+#if CONFIG_MACF_NET
+       mac_mbuf_label_associate_inpcb(inp, m);
+#endif /* CONFIG_MACF_NET */
+
+       if (inp->inp_flowhash == 0)
+               inp->inp_flowhash = inp_calc_flowhash(inp);
 
        /*
         * Calculate data length and get a mbuf
         * for UDP and IP headers.
         */
-       M_PREPEND(m, sizeof(struct udpiphdr), M_DONTWAIT);
+       M_PREPEND(m, sizeof (struct udpiphdr), M_DONTWAIT);
        if (m == 0) {
                error = ENOBUFS;
                goto abort;
@@ -1012,20 +1527,20 @@ udp_output(inp, m, addr, control, p)
         * and addresses and length put into network format.
         */
        ui = mtod(m, struct udpiphdr *);
-       bzero(ui->ui_x1, sizeof(ui->ui_x1));    /* XXX still needed? */
+       bzero(ui->ui_x1, sizeof (ui->ui_x1));   /* XXX still needed? */
        ui->ui_pr = IPPROTO_UDP;
        ui->ui_src = laddr;
        ui->ui_dst = faddr;
        ui->ui_sport = lport;
        ui->ui_dport = fport;
-       ui->ui_ulen = htons((u_short)len + sizeof(struct udphdr));
+       ui->ui_ulen = htons((u_short)len + sizeof (struct udphdr));
 
        /*
         * Set up checksum and output datagram.
         */
-       if (udpcksum) {
-               ui->ui_sum = in_pseudo(ui->ui_src.s_addr, ui->ui_dst.s_addr,
-                   htons((u_short)len + sizeof(struct udphdr) + IPPROTO_UDP));
+       if (udpcksum && !(inp->inp_flags & INP_UDP_NOCKSUM)) {
+               ui->ui_sum = in_pseudo(ui->ui_src.s_addr, ui->ui_dst.s_addr,
+                   htons((u_short)len + sizeof (struct udphdr) + IPPROTO_UDP));
                m->m_pkthdr.csum_flags = CSUM_UDP;
                m->m_pkthdr.csum_data = offsetof(struct udphdr, uh_sum);
        } else {
@@ -1044,45 +1559,175 @@ udp_output(inp, m, addr, control, p)
                error = ENOBUFS;
                goto abort;
        }
-#endif /*IPSEC*/
-       m->m_pkthdr.socket_id = get_socket_id(inp->inp_socket);
-       error = ip_output_list(m, 0,  inp->inp_options, &inp->inp_route,
-           (inp->inp_socket->so_options & (SO_DONTROUTE | SO_BROADCAST)),
-           inp->inp_moptions);
+#endif /* IPSEC */
+
+       inpopts = inp->inp_options;
+       soopts |= (inp->inp_socket->so_options & (SO_DONTROUTE | SO_BROADCAST));
+       mopts = inp->inp_moptions;
+       if (mopts != NULL) {
+               IMO_LOCK(mopts);
+               IMO_ADDREF_LOCKED(mopts);
+               if (IN_MULTICAST(ntohl(ui->ui_dst.s_addr)) &&
+                   mopts->imo_multicast_ifp != NULL) {
+                       /* no reference needed */
+                       inp->inp_last_outifp = mopts->imo_multicast_ifp;
+               }
+               IMO_UNLOCK(mopts);
+       }
+
+       /* Copy the cached route and take an extra reference */
+       inp_route_copyout(inp, &ro);
+
+       set_packet_service_class(m, so, msc, 0);
+       m->m_pkthdr.pkt_flowsrc = FLOWSRC_INPCB;
+       m->m_pkthdr.pkt_flowid = inp->inp_flowhash;
+       m->m_pkthdr.pkt_proto = IPPROTO_UDP;
+       m->m_pkthdr.pkt_flags |= (PKTF_FLOW_ID | PKTF_FLOW_LOCALSRC);
+       if (flowadv)
+               m->m_pkthdr.pkt_flags |= PKTF_FLOW_ADV;
+
+       if (ipoa.ipoa_boundif != IFSCOPE_NONE)
+               ipoa.ipoa_flags |= IPOAF_BOUND_IF;
+
+       if (laddr.s_addr != INADDR_ANY)
+               ipoa.ipoa_flags |= IPOAF_BOUND_SRCADDR;
+
+       inp->inp_sndinprog_cnt++;
+
+       socket_unlock(so, 0);
+       error = ip_output(m, inpopts, &ro, soopts, mopts, &ipoa);
+       m = NULL;
+       socket_lock(so, 0);
+       if (mopts != NULL)
+               IMO_REMREF(mopts);
+
+       if (error == 0 && nstat_collect) {
+               boolean_t cell, wifi;
+
+               if (ro.ro_rt != NULL) {
+                       cell = IFNET_IS_CELLULAR(ro.ro_rt->rt_ifp);
+                       wifi = (!cell && IFNET_IS_WIFI(ro.ro_rt->rt_ifp));
+               } else {
+                       cell = wifi = FALSE;
+               }
+               INP_ADD_STAT(inp, cell, wifi, txpackets, 1);
+               INP_ADD_STAT(inp, cell, wifi, txbytes, len);
+       }
+
+       if (flowadv && (adv->code == FADV_FLOW_CONTROLLED ||
+           adv->code == FADV_SUSPENDED)) {
+               /* return a hint to the application that 
+                * the packet has been dropped
+                */
+               error = ENOBUFS;
+               inp_set_fc_state(inp, adv->code);
+       }
 
+       VERIFY(inp->inp_sndinprog_cnt > 0);
+       if ( --inp->inp_sndinprog_cnt == 0)
+               inp->inp_flags &= ~(INP_FC_FEEDBACK);
+
+       /* Synchronize PCB cached route */
+       inp_route_copyin(inp, &ro);
+
+abort:
        if (udp_dodisconnect) {
+               /* Always discard the cached route for unconnected socket */
+               ROUTE_RELEASE(&inp->inp_route);
                in_pcbdisconnect(inp);
                inp->inp_laddr = origladdr;     /* XXX rehash? */
+               /* no reference needed */
+               inp->inp_last_outifp = origoutifp;
+       } else if (inp->inp_route.ro_rt != NULL) {
+               struct rtentry *rt = inp->inp_route.ro_rt;
+               struct ifnet *outifp;
+
+               if (rt->rt_flags & (RTF_MULTICAST|RTF_BROADCAST))
+                       rt = NULL;      /* unusable */
+               /*
+                * Always discard if it is a multicast or broadcast route.
+                */
+               if (rt == NULL)
+                       ROUTE_RELEASE(&inp->inp_route);
+
+               /*
+                * If the destination route is unicast, update outifp with
+                * that of the route interface used by IP.
+                */
+               if (rt != NULL && (outifp = rt->rt_ifp) != inp->inp_last_outifp)
+                       inp->inp_last_outifp = outifp;  /* no reference needed */
+       } else {
+               ROUTE_RELEASE(&inp->inp_route);
        }
-       KERNEL_DEBUG(DBG_FNC_UDP_OUTPUT | DBG_FUNC_END, error, 0,0,0,0);
-       return (error);
 
-abort:
-        if (udp_dodisconnect) {
-                in_pcbdisconnect(inp);
-                inp->inp_laddr = origladdr; /* XXX rehash? */
-        }
+       /*
+        * If output interface was cellular, and this socket is denied
+        * access to it, generate an event.
+        */
+       if (error != 0 && (ipoa.ipoa_retflags & IPOARF_IFDENIED) &&
+           (inp->inp_flags & INP_NO_IFT_CELLULAR))
+               soevent(so, (SO_FILT_HINT_LOCKED|SO_FILT_HINT_IFDENIED));
 
 release:
-       m_freem(m);
-       KERNEL_DEBUG(DBG_FNC_UDP_OUTPUT | DBG_FUNC_END, error, 0,0,0,0);
+       KERNEL_DEBUG(DBG_FNC_UDP_OUTPUT | DBG_FUNC_END, error, 0, 0, 0, 0);
+
+       if (m != NULL)
+               m_freem(m);
+
+       if (outif != NULL)
+               ifnet_release(outif);
+
        return (error);
 }
 
-u_long udp_sendspace = 9216;           /* really max datagram size */
-                                       /* 40 1K datagrams */
-SYSCTL_INT(_net_inet_udp, UDPCTL_MAXDGRAM, maxdgram, CTLFLAG_RW,
-    &udp_sendspace, 0, "Maximum outgoing UDP datagram size");
-
-u_long udp_recvspace = 40 * (1024 +
+u_int32_t      udp_sendspace = 9216;           /* really max datagram size */
+/* 187 1K datagrams (approx 192 KB) */
+u_int32_t      udp_recvspace = 187 * (1024 +
 #if INET6
-                                     sizeof(struct sockaddr_in6)
-#else
-                                     sizeof(struct sockaddr_in)
-#endif
-                                     );
-SYSCTL_INT(_net_inet_udp, UDPCTL_RECVSPACE, recvspace, CTLFLAG_RW,
-    &udp_recvspace, 0, "Maximum incoming UDP datagram size");
+                   sizeof (struct sockaddr_in6)
+#else /* !INET6 */
+                   sizeof (struct sockaddr_in)
+#endif /* !INET6 */
+               );
+
+/* Check that the values of udp send and recv space do not exceed sb_max */
+static int
+sysctl_udp_sospace(struct sysctl_oid *oidp, void *arg1, int arg2,
+    struct sysctl_req *req)
+{
+#pragma unused(arg1, arg2)
+       u_int32_t new_value = 0, *space_p = NULL;
+       int changed = 0, error = 0;
+       u_quad_t sb_effective_max = (sb_max/(MSIZE+MCLBYTES)) * MCLBYTES;
+
+       switch (oidp->oid_number) {
+       case UDPCTL_RECVSPACE:
+               space_p = &udp_recvspace;
+               break;
+       case UDPCTL_MAXDGRAM:
+               space_p = &udp_sendspace;
+               break;
+       default:
+               return EINVAL;
+       }
+        error = sysctl_io_number(req, *space_p, sizeof (u_int32_t),
+           &new_value, &changed);
+        if (changed) {
+                if (new_value > 0 && new_value <= sb_effective_max)
+                        *space_p = new_value;
+                else
+                        error = ERANGE;
+        }
+        return (error);
+}
+
+SYSCTL_PROC(_net_inet_udp, UDPCTL_RECVSPACE, recvspace,
+    CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_LOCKED, &udp_recvspace, 0,
+    &sysctl_udp_sospace, "IU", "Maximum incoming UDP datagram size");
+
+SYSCTL_PROC(_net_inet_udp, UDPCTL_MAXDGRAM, maxdgram,
+    CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_LOCKED, &udp_sendspace, 0,
+    &sysctl_udp_sospace, "IU", "Maximum outgoing UDP datagram size");
 
 static int
 udp_abort(struct socket *so)
@@ -1090,33 +1735,39 @@ udp_abort(struct socket *so)
        struct inpcb *inp;
 
        inp = sotoinpcb(so);
-       if (inp == 0)
-               panic("udp_abort: so=%x null inp\n", so);       /* ??? possible? panic instead? */
+       if (inp == NULL) {
+               panic("%s: so=%p null inp\n", __func__, so);
+               /* NOTREACHED */
+       }
        soisdisconnected(so);
        in_pcbdetach(inp);
-       return 0;
+       return (0);
 }
 
 static int
 udp_attach(struct socket *so, int proto, struct proc *p)
 {
+#pragma unused(proto)
        struct inpcb *inp;
        int error;
 
        inp = sotoinpcb(so);
-       if (inp != 0)
-               panic ("udp_attach so=%x inp=%x\n", so, inp);
-
+       if (inp != NULL) {
+               panic ("%s so=%p inp=%p\n", __func__, so, inp);
+               /* NOTREACHED */
+       }
        error = in_pcballoc(so, &udbinfo, p);
-       if (error)
-               return error;
+       if (error != 0)
+               return (error);
        error = soreserve(so, udp_sendspace, udp_recvspace);
-       if (error
-               return error;
+       if (error != 0)
+               return (error);
        inp = (struct inpcb *)so->so_pcb;
        inp->inp_vflag |= INP_IPV4;
        inp->inp_ip_ttl = ip_defttl;
-       return 0;
+       if (nstat_collect)
+               nstat_udp_new_pcb(inp);
+       return (0);
 }
 
 static int
@@ -1125,11 +1776,15 @@ udp_bind(struct socket *so, struct sockaddr *nam, struct proc *p)
        struct inpcb *inp;
        int error;
 
+       if (nam->sa_family != 0 && nam->sa_family != AF_INET &&
+           nam->sa_family != AF_INET6)
+               return (EAFNOSUPPORT);
+
        inp = sotoinpcb(so);
-       if (inp == 0)
-               return EINVAL;
+       if (inp == NULL || (inp->inp_flags2 & INP2_WANT_FLOW_DIVERT))
+               return (inp == NULL ? EINVAL : EPROTOTYPE);
        error = in_pcbbind(inp, nam, p);
-       return error;
+       return (error);
 }
 
 static int
@@ -1139,14 +1794,86 @@ udp_connect(struct socket *so, struct sockaddr *nam, struct proc *p)
        int error;
 
        inp = sotoinpcb(so);
-       if (inp == 0)
-               return EINVAL;
+       if (inp == NULL || (inp->inp_flags2 & INP2_WANT_FLOW_DIVERT))
+               return (inp == NULL ? EINVAL : EPROTOTYPE);
        if (inp->inp_faddr.s_addr != INADDR_ANY)
-               return EISCONN;
-       error = in_pcbconnect(inp, nam, p);
-       if (error == 0) 
+               return (EISCONN);
+       error = in_pcbconnect(inp, nam, p, IFSCOPE_NONE, NULL);
+       if (error == 0) {
                soisconnected(so);
-       return error;
+               if (inp->inp_flowhash == 0)
+                       inp->inp_flowhash = inp_calc_flowhash(inp);
+       }
+       return (error);
+}
+
+int
+udp_connectx_common(struct socket *so, int af,
+    struct sockaddr_list **src_sl, struct sockaddr_list **dst_sl,
+    struct proc *p, uint32_t ifscope, associd_t aid, connid_t *pcid,
+    uint32_t flags, void *arg, uint32_t arglen)
+{
+#pragma unused(aid, flags, arg, arglen)
+       struct sockaddr_entry *src_se = NULL, *dst_se = NULL;
+       struct inpcb *inp = sotoinpcb(so);
+       int error;
+
+       if (inp == NULL)
+               return (EINVAL);
+
+       VERIFY(dst_sl != NULL);
+
+       /* select source (if specified) and destination addresses */
+       error = in_selectaddrs(af, src_sl, &src_se, dst_sl, &dst_se);
+       if (error != 0)
+               return (error);
+
+       VERIFY(*dst_sl != NULL && dst_se != NULL);
+       VERIFY(src_se == NULL || *src_sl != NULL);
+       VERIFY(dst_se->se_addr->sa_family == af);
+       VERIFY(src_se == NULL || src_se->se_addr->sa_family == af);
+
+       /* bind socket to the specified interface, if requested */
+       if (ifscope != IFSCOPE_NONE &&
+           (error = inp_bindif(inp, ifscope, NULL)) != 0)
+               return (error);
+
+       /* if source address and/or port is specified, bind to it */
+       if (src_se != NULL) {
+               struct sockaddr *sa = src_se->se_addr;
+               error = sobindlock(so, sa, 0);  /* already locked */
+               if (error != 0)
+                       return (error);
+       }
+
+       switch (af) {
+       case AF_INET:
+               error = udp_connect(so, dst_se->se_addr, p);
+               break;
+#if INET6
+       case AF_INET6:
+               error = udp6_connect(so, dst_se->se_addr, p);
+               break;
+#endif /* INET6 */
+       default:
+               VERIFY(0);
+               /* NOTREACHED */
+       }
+
+       if (error == 0 && pcid != NULL)
+               *pcid = 1;      /* there is only 1 connection for a UDP */
+
+       return (error);
+}
+
+static int
+udp_connectx(struct socket *so, struct sockaddr_list **src_sl,
+    struct sockaddr_list **dst_sl, struct proc *p, uint32_t ifscope,
+    associd_t aid, connid_t *pcid, uint32_t flags, void *arg,
+    uint32_t arglen)
+{
+       return (udp_connectx_common(so, AF_INET, src_sl, dst_sl,
+           p, ifscope, aid, pcid, flags, arg, arglen));
 }
 
 static int
@@ -1155,11 +1882,13 @@ udp_detach(struct socket *so)
        struct inpcb *inp;
 
        inp = sotoinpcb(so);
-       if (inp == 0)
-               panic("udp_detach: so=%x null inp\n", so);      /* ??? possible? panic instead? */
+       if (inp == NULL) {
+               panic("%s: so=%p null inp\n", __func__, so);
+               /* NOTREACHED */
+       }
        in_pcbdetach(inp);
        inp->inp_state = INPCB_STATE_DEAD;
-       return 0;
+       return (0);
 }
 
 static int
@@ -1168,29 +1897,49 @@ udp_disconnect(struct socket *so)
        struct inpcb *inp;
 
        inp = sotoinpcb(so);
-       if (inp == 0)
-               return EINVAL;
+       if (inp == NULL || (inp->inp_flags2 & INP2_WANT_FLOW_DIVERT))
+               return (inp == NULL ? EINVAL : EPROTOTYPE);
        if (inp->inp_faddr.s_addr == INADDR_ANY)
-               return ENOTCONN;
+               return (ENOTCONN);
 
        in_pcbdisconnect(inp);
+
+       /* reset flow controlled state, just in case */
+       inp_reset_fc_state(inp);
+
        inp->inp_laddr.s_addr = INADDR_ANY;
        so->so_state &= ~SS_ISCONNECTED;                /* XXX */
-       return 0;
+       inp->inp_last_outifp = NULL;
+       return (0);
 }
 
 static int
-udp_send(struct socket *so, int flags, struct mbuf *m, struct sockaddr *addr,
-           struct mbuf *control, struct proc *p)
+udp_disconnectx(struct socket *so, associd_t aid, connid_t cid)
 {
+#pragma unused(cid)
+       if (aid != ASSOCID_ANY && aid != ASSOCID_ALL)
+               return (EINVAL);
+
+       return (udp_disconnect(so));
+}
+
+static int
+udp_send(struct socket *so, int flags, struct mbuf *m,
+    struct sockaddr *addr, struct mbuf *control, struct proc *p)
+{
+#pragma unused(flags)
        struct inpcb *inp;
 
        inp = sotoinpcb(so);
-       if (inp == 0) {
-               m_freem(m);
-               return EINVAL;
+       if (inp == NULL || (inp->inp_flags2 & INP2_WANT_FLOW_DIVERT)) {
+               if (m != NULL)
+                       m_freem(m);
+               if (control != NULL)
+                       m_freem(control);
+               return (inp == NULL ? EINVAL : EPROTOTYPE);
        }
-       return udp_output(inp, m, addr, control, p);
+
+       return (udp_output(inp, m, addr, control, p));
 }
 
 int
@@ -1199,138 +1948,263 @@ udp_shutdown(struct socket *so)
        struct inpcb *inp;
 
        inp = sotoinpcb(so);
-       if (inp == 0)
-               return EINVAL;
+       if (inp == NULL)
+               return (EINVAL);
        socantsendmore(so);
-       return 0;
+       return (0);
 }
 
-struct pr_usrreqs udp_usrreqs = {
-       udp_abort, pru_accept_notsupp, udp_attach, udp_bind, udp_connect, 
-       pru_connect2_notsupp, in_control, udp_detach, udp_disconnect, 
-       pru_listen_notsupp, in_setpeeraddr, pru_rcvd_notsupp, 
-       pru_rcvoob_notsupp, udp_send, pru_sense_null, udp_shutdown,
-       in_setsockaddr, sosend, soreceive, pru_sopoll_notsupp
-};
-
-
 int
-udp_lock(so, refcount, debug)
-       struct socket *so;
-       int refcount, debug;
+udp_lock(struct socket *so, int refcount, void *debug)
 {
-       int lr_saved;
-       if (debug == 0) 
-               lr_saved = (unsigned int) __builtin_return_address(0);
-       else lr_saved = debug;
-
-       if (so->so_pcb) {
-               lck_mtx_assert(((struct inpcb *)so->so_pcb)->inpcb_mtx, LCK_MTX_ASSERT_NOTOWNED);
-               lck_mtx_lock(((struct inpcb *)so->so_pcb)->inpcb_mtx);
-       }
-       else 
-               panic("udp_lock: so=%x NO PCB! lr=%x\n", so, lr_saved);
+       void *lr_saved;
 
-       if (refcount) 
+       if (debug == NULL)
+               lr_saved = __builtin_return_address(0);
+       else
+               lr_saved = debug;
+
+       if (so->so_pcb != NULL) {
+               lck_mtx_assert(&((struct inpcb *)so->so_pcb)->inpcb_mtx,
+                   LCK_MTX_ASSERT_NOTOWNED);
+               lck_mtx_lock(&((struct inpcb *)so->so_pcb)->inpcb_mtx);
+       } else {
+               panic("%s: so=%p NO PCB! lr=%p lrh= %s\n", __func__,
+                   so, lr_saved, solockhistory_nr(so));
+               /* NOTREACHED */
+       }
+       if (refcount)
                so->so_usecount++;
 
-       so->lock_lr[so->next_lock_lr] = (void *)lr_saved;
+       so->lock_lr[so->next_lock_lr] = lr_saved;
        so->next_lock_lr = (so->next_lock_lr+1) % SO_LCKDBG_MAX;
        return (0);
 }
 
 int
-udp_unlock(so, refcount, debug)
-       struct socket *so;
-       int refcount;
-       int debug;
+udp_unlock(struct socket *so, int refcount, void *debug)
 {
-       int lr_saved;
-       struct inpcb *inp = sotoinpcb(so);
-       struct inpcbinfo *pcbinfo       = &udbinfo;
+       void *lr_saved;
 
-       if (debug == 0) 
-               lr_saved = (unsigned int) __builtin_return_address(0);
-       else lr_saved = debug;
+       if (debug == NULL)
+               lr_saved = __builtin_return_address(0);
+       else
+               lr_saved = debug;
 
-       if (refcount) {
+       if (refcount)
                so->so_usecount--;
-#if 0
-               if (so->so_usecount == 0 && (inp->inp_wantcnt == WNT_STOPUSING)) {
-                       if (lck_rw_try_lock_exclusive(pcbinfo->mtx)) {
-                               in_pcbdispose(inp);
-                               lck_rw_done(pcbinfo->mtx);
-                               return(0);
-                       }
-               }
-#endif
-       }
-       if (so->so_pcb == NULL) 
-               panic("udp_unlock: so=%x NO PCB! lr=%x\n", so, lr_saved);
-       else {
-               lck_mtx_assert(((struct inpcb *)so->so_pcb)->inpcb_mtx, LCK_MTX_ASSERT_OWNED);
-               so->unlock_lr[so->next_unlock_lr] = (void *)lr_saved;
+
+       if (so->so_pcb == NULL) {
+               panic("%s: so=%p NO PCB! lr=%p lrh= %s\n", __func__,
+                   so, lr_saved, solockhistory_nr(so));
+               /* NOTREACHED */
+       } else {
+               lck_mtx_assert(&((struct inpcb *)so->so_pcb)->inpcb_mtx,
+                   LCK_MTX_ASSERT_OWNED);
+               so->unlock_lr[so->next_unlock_lr] = lr_saved;
                so->next_unlock_lr = (so->next_unlock_lr+1) % SO_LCKDBG_MAX;
-               lck_mtx_unlock(((struct inpcb *)so->so_pcb)->inpcb_mtx);
+               lck_mtx_unlock(&((struct inpcb *)so->so_pcb)->inpcb_mtx);
        }
-
-
        return (0);
 }
 
 lck_mtx_t *
-udp_getlock(so, locktype)
-       struct socket *so;
-       int locktype;
+udp_getlock(struct socket *so, int locktype)
 {
+#pragma unused(locktype)
        struct inpcb *inp = sotoinpcb(so);
 
-
-       if (so->so_pcb)
-               return(inp->inpcb_mtx);
-       else {
-               panic("udp_getlock: so=%x NULL so_pcb\n", so);
-               return (so->so_proto->pr_domain->dom_mtx);
+       if (so->so_pcb == NULL) {
+               panic("%s: so=%p NULL so_pcb lrh= %s\n", __func__,
+                   so, solockhistory_nr(so));
+               /* NOTREACHED */
        }
+       return (&inp->inpcb_mtx);
 }
 
-void
-udp_slowtimo()
+/*
+ * UDP garbage collector callback (inpcb_timer_func_t).
+ *
+ * Returns > 0 to keep timer active.
+ */
+static void
+udp_gc(struct inpcbinfo *ipi)
 {
        struct inpcb *inp, *inpnxt;
        struct socket *so;
-       struct inpcbinfo *pcbinfo       = &udbinfo;
 
-       lck_rw_lock_exclusive(pcbinfo->mtx);
+       if (lck_rw_try_lock_exclusive(ipi->ipi_lock) == FALSE) {
+               if (udp_gc_done == TRUE) {
+                       udp_gc_done = FALSE;
+                       /* couldn't get the lock, must lock next time */
+                       atomic_add_32(&ipi->ipi_gc_req.intimer_fast, 1);
+                       return;
+               }
+               lck_rw_lock_exclusive(ipi->ipi_lock);
+       }
+
+       udp_gc_done = TRUE;
 
        for (inp = udb.lh_first; inp != NULL; inp = inpnxt) {
                inpnxt = inp->inp_list.le_next;
-       
-               /* Ignore nat/SharedIP dummy pcbs */
-               if (inp->inp_socket == &udbinfo.nat_dummy_socket)
-                               continue;
 
-               if (inp->inp_wantcnt != WNT_STOPUSING) 
+               /*
+                * Skip unless it's STOPUSING; garbage collector will
+                * be triggered by in_pcb_checkstate() upon setting
+                * wantcnt to that value.  If the PCB is already dead,
+                * keep gc active to anticipate wantcnt changing.
+                */
+               if (inp->inp_wantcnt != WNT_STOPUSING)
                        continue;
 
-               so = inp->inp_socket;
-               if (!lck_mtx_try_lock(inp->inpcb_mtx))  /* skip if busy, no hurry for cleanup... */
+               /*
+                * Skip if busy, no hurry for cleanup.  Keep gc active
+                * and try the lock again during next round.
+                */
+               if (!lck_mtx_try_lock(&inp->inpcb_mtx)) {
+                       atomic_add_32(&ipi->ipi_gc_req.intimer_fast, 1);
                        continue;
+               }
 
-               if (so->so_usecount == 0)
+               /*
+                * Keep gc active unless usecount is 0.
+                */
+               so = inp->inp_socket;
+               if (so->so_usecount == 0) {
+                       if (inp->inp_state != INPCB_STATE_DEAD) {
+#if INET6
+                               if (SOCK_CHECK_DOM(so, PF_INET6))
+                                       in6_pcbdetach(inp);
+                               else
+#endif /* INET6 */
+                                       in_pcbdetach(inp);
+                       }
                        in_pcbdispose(inp);
-               else 
-                       lck_mtx_unlock(inp->inpcb_mtx);
+               } else {
+                       lck_mtx_unlock(&inp->inpcb_mtx);
+                       atomic_add_32(&ipi->ipi_gc_req.intimer_fast, 1);
+               }
        }
-       lck_rw_done(pcbinfo->mtx);
+       lck_rw_done(ipi->ipi_lock);
+
+       return;
 }
 
-int
-ChkAddressOK( __uint32_t dstaddr, __uint32_t srcaddr )
+static int
+udp_getstat SYSCTL_HANDLER_ARGS
 {
-        if ( dstaddr == srcaddr ){
-                return 0;
-        }
-        return 1;
+#pragma unused(oidp, arg1, arg2)
+       if (req->oldptr == USER_ADDR_NULL)
+               req->oldlen = (size_t)sizeof (struct udpstat);
+
+       return (SYSCTL_OUT(req, &udpstat, MIN(sizeof (udpstat), req->oldlen)));
+}
+
+void
+udp_in_cksum_stats(u_int32_t len)
+{
+       udpstat.udps_rcv_swcsum++;
+       udpstat.udps_rcv_swcsum_bytes += len;
+}
+
+void
+udp_out_cksum_stats(u_int32_t len)
+{
+       udpstat.udps_snd_swcsum++;
+       udpstat.udps_snd_swcsum_bytes += len;
+}
+
+#if INET6
+void
+udp_in6_cksum_stats(u_int32_t len)
+{
+       udpstat.udps_rcv6_swcsum++;
+       udpstat.udps_rcv6_swcsum_bytes += len;
 }
 
+void
+udp_out6_cksum_stats(u_int32_t len)
+{
+       udpstat.udps_snd6_swcsum++;
+       udpstat.udps_snd6_swcsum_bytes += len;
+}
+#endif /* INET6 */
+
+/*
+ * Checksum extended UDP header and data.
+ */
+static int
+udp_input_checksum(struct mbuf *m, struct udphdr *uh, int off, int ulen)
+{
+       struct ifnet *ifp = m->m_pkthdr.rcvif;
+       struct ip *ip = mtod(m, struct ip *);
+       struct ipovly *ipov = (struct ipovly *)ip;
+
+       if (uh->uh_sum == 0) {
+               udpstat.udps_nosum++;
+               return (0);
+       }
+
+       if ((hwcksum_rx || (ifp->if_flags & IFF_LOOPBACK) ||
+           (m->m_pkthdr.pkt_flags & PKTF_LOOP)) &&
+           (m->m_pkthdr.csum_flags & CSUM_DATA_VALID)) {
+               if (m->m_pkthdr.csum_flags & CSUM_PSEUDO_HDR) {
+                       uh->uh_sum = m->m_pkthdr.csum_rx_val;
+               } else {
+                       uint16_t sum = m->m_pkthdr.csum_rx_val;
+                       uint16_t start = m->m_pkthdr.csum_rx_start;
+
+                       /*
+                        * Perform 1's complement adjustment of octets
+                        * that got included/excluded in the hardware-
+                        * calculated checksum value.  Ignore cases
+                        * where the value includes or excludes the
+                        * IP header span, as the sum for those octets
+                        * would already be 0xffff and thus no-op.
+                        */
+                       if ((m->m_pkthdr.csum_flags & CSUM_PARTIAL) &&
+                           start != 0 && (off - start) != off) {
+#if BYTE_ORDER != BIG_ENDIAN
+                               if (start < off) {
+                                       HTONS(ip->ip_len);
+                                       HTONS(ip->ip_off);
+                               }
+#endif /* BYTE_ORDER != BIG_ENDIAN */
+                               /* callee folds in sum */
+                               sum = m_adj_sum16(m, start, off, sum);
+#if BYTE_ORDER != BIG_ENDIAN
+                               if (start < off) {
+                                       NTOHS(ip->ip_off);
+                                       NTOHS(ip->ip_len);
+                               }
+#endif /* BYTE_ORDER != BIG_ENDIAN */
+                       }
+
+                       /* callee folds in sum */
+                       uh->uh_sum = in_pseudo(ip->ip_src.s_addr,
+                           ip->ip_dst.s_addr, sum + htonl(ulen + IPPROTO_UDP));
+               }
+               uh->uh_sum ^= 0xffff;
+       } else {
+               uint16_t ip_sum;
+               char b[9];
+
+               bcopy(ipov->ih_x1, b, sizeof (ipov->ih_x1));
+               bzero(ipov->ih_x1, sizeof (ipov->ih_x1));
+               ip_sum = ipov->ih_len;
+               ipov->ih_len = uh->uh_ulen;
+               uh->uh_sum = in_cksum(m, ulen + sizeof (struct ip));
+               bcopy(b, ipov->ih_x1, sizeof (ipov->ih_x1));
+               ipov->ih_len = ip_sum;
+
+               udp_in_cksum_stats(ulen);
+       }
+
+       if (uh->uh_sum != 0) {
+               udpstat.udps_badsum++;
+               IF_UDP_STATINC(ifp, badchksum);
+               return (-1);
+       }
+
+       return (0);
+}