/*
* Copyright (c) 2000 Apple Computer, Inc. All rights reserved.
*
- * @APPLE_LICENSE_HEADER_START@
+ * @APPLE_LICENSE_OSREFERENCE_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 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.
- *
- * @APPLE_LICENSE_HEADER_END@
+ * This file contains Original Code and/or Modifications of Original Code
+ * as defined in and that are subject to the Apple Public Source License
+ * Version 2.0 (the 'License'). You may not use this file except in
+ * compliance with the License. The rights granted to you under the
+ * License may not be used to create, or enable the creation or
+ * redistribution of, unlawful or unlicensed copies of an Apple operating
+ * system, or to circumvent, violate, or enable the circumvention or
+ * violation of, any terms of an Apple operating system software license
+ * agreement.
+ *
+ * Please obtain a copy of the License at
+ * http://www.opensource.apple.com/apsl/ and read it before using this
+ * file.
+ *
+ * The Original Code and all software distributed under the License are
+ * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
+ * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
+ * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
+ * Please see the License for the specific language governing rights and
+ * limitations under the License.
+ *
+ * @APPLE_LICENSE_OSREFERENCE_HEADER_END@
*/
/*
* Copyright (c) 1982, 1986, 1988, 1990, 1993, 1995
* SUCH DAMAGE.
*
* @(#)udp_usrreq.c 8.6 (Berkeley) 5/23/95
+ * $FreeBSD: src/sys/netinet/udp_usrreq.c,v 1.64.2.13 2001/08/08 18:59:54 ghelmer Exp $
*/
#include <sys/param.h>
#include <sys/kernel.h>
#include <sys/malloc.h>
#include <sys/mbuf.h>
-#if INET6
#include <sys/domain.h>
-#endif
#include <sys/protosw.h>
#include <sys/socket.h>
#include <sys/socketvar.h>
#include <sys/sysctl.h>
#include <sys/syslog.h>
-#if ISFB31
-#include <vm/vm_zone.h>
-#endif
-
-
#include <net/if.h>
+#include <net/if_types.h>
#include <net/route.h>
#include <netinet/in.h>
#include <netinet/in_systm.h>
#include <netinet/ip.h>
+#if INET6
+#include <netinet/ip6.h>
+#endif
#include <netinet/in_pcb.h>
#include <netinet/in_var.h>
#include <netinet/ip_var.h>
-#include <netinet/ip_icmp.h>
-#include <netinet/icmp_var.h>
#if INET6
-#include <netinet/ip6.h>
#include <netinet6/ip6_var.h>
#endif
+#include <netinet/ip_icmp.h>
+#include <netinet/icmp_var.h>
#include <netinet/udp.h>
#include <netinet/udp_var.h>
#include <sys/kdebug.h>
#if IPSEC
#include <netinet6/ipsec.h>
+extern int ipsec_bypass;
+extern lck_mtx_t *sadb_mutex;
#endif /*IPSEC*/
#define DBG_FNC_UDP_INPUT NETDBG_CODE(DBG_NETUDP, (5 << 8))
#define DBG_FNC_UDP_OUTPUT NETDBG_CODE(DBG_NETUDP, (6 << 8) | 1)
-#define __STDC__ 1
/*
* UDP protocol implementation.
* Per RFC 768, August, 1980.
SYSCTL_INT(_net_inet_udp, UDPCTL_CHECKSUM, checksum, CTLFLAG_RW,
&udpcksum, 0, "");
-int log_in_vain;
+int log_in_vain = 0;
SYSCTL_INT(_net_inet_udp, OID_AUTO, log_in_vain, CTLFLAG_RW,
- &log_in_vain, 0, "");
+ &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");
struct inpcbhead udb; /* from udp_var.h */
#define udb6 udb /* for KAME src sync over BSD*'s */
#define UDBHASHSIZE 16
#endif
+extern int apple_hwcksum_rx;
+extern int esp_udp_encap_port;
+extern u_long route_generation;
+
+extern void ipfwsyslog( int level, char *format,...);
+
+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 ; \
+}
+
struct udpstat udpstat; /* from udp_var.h */
SYSCTL_STRUCT(_net_inet_udp, UDPCTL_STATS, stats, CTLFLAG_RD,
- &udpstat, udpstat, "");
+ &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");
static struct sockaddr_in udp_in = { sizeof(udp_in), AF_INET };
#if INET6
} udp_ip6;
#endif /* INET6 */
-static void udp_append __P((struct inpcb *last, struct ip *ip,
- struct mbuf *n, int off));
+static void udp_append(struct inpcb *last, struct ip *ip,
+ struct mbuf *n, int off);
#if INET6
-static void ip_2_ip6_hdr __P((struct ip6_hdr *ip6, struct ip *ip));
+static void ip_2_ip6_hdr(struct ip6_hdr *ip6, struct ip *ip);
#endif
-static int udp_detach __P((struct socket *so));
-static int udp_output __P((struct inpcb *, struct mbuf *, struct sockaddr *,
- struct mbuf *, struct proc *));
+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 );
void
udp_init()
{
- vm_size_t str_size;
- int stat;
- u_char fake_owner;
- struct in_addr laddr;
- struct in_addr faddr;
- u_short lport;
+ vm_size_t str_size;
+ struct inpcbinfo *pcbinfo;
+
LIST_INIT(&udb);
udbinfo.listhead = &udb;
udbinfo.hashbase = hashinit(UDBHASHSIZE, M_PCB, &udbinfo.hashmask);
udbinfo.porthashbase = hashinit(UDBHASHSIZE, M_PCB,
&udbinfo.porthashmask);
-#if ISFB31
- udbinfo.ipi_zone = zinit("udpcb", sizeof(struct inpcb), maxsockets,
- ZONE_INTERRUPT, 0);
-#else
+#ifdef __APPLE__
str_size = (vm_size_t) sizeof(struct inpcb);
- udbinfo.ipi_zone = (void *) zinit(str_size, 80000*str_size, 8192, "inpcb_zone");
-#endif
+ udbinfo.ipi_zone = (void *) 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... */
- udbinfo.last_pcb = 0;
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
#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);
register struct udphdr *uh;
register struct inpcb *inp;
struct mbuf *opts = 0;
-#if INET6
- struct ip6_recvpktopts opts6;
-#endif
int len;
struct ip save_ip;
struct sockaddr *append_sa;
+ struct inpcbinfo *pcbinfo = &udbinfo;
udpstat.udps_ipackets++;
-#if INET6
- bzero(&opts6, sizeof(opts6));
-#endif
-
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 */
/*
* Strip IP options, if any; should skip this,
}
uh = (struct udphdr *)((caddr_t)ip + iphlen);
+ /* destination port of 0 is illegal, based on RFC768. */
+ if (uh->uh_dport == 0)
+ 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);
* Checksum extended UDP header and data.
*/
if (uh->uh_sum) {
- 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));
+ 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);
return;
}
}
+#ifndef __APPLE__
+ else
+ udpstat.udps_nosum++;
+#endif
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);
/*
* Deliver a multicast or broadcast datagram to *all* sockets
* for which the local and remote addresses and ports match
* fixing the interface. Maybe 4.5BSD will remedy this?)
*/
+
/*
* Construct sockaddr format source address.
*/
udp_in6.uin6_init_done = udp_ip6.uip6_init_done = 0;
#endif
LIST_FOREACH(inp, &udb, inp_list) {
-#if INET6
- if ((inp->inp_vflag & INP_IPV4) == 0)
+#ifdef __APPLE__
+ /* Ignore nat/SharedIP dummy pcbs */
+ if (inp->inp_socket == &udbinfo.nat_dummy_socket)
+ continue;
+#endif
+ if (inp->inp_socket == NULL)
continue;
+ if (inp != sotoinpcb(inp->inp_socket))
+ panic("udp_input: bad so back ptr inp=%x\n", inp);
+#if INET6
+ if ((inp->inp_vflag & INP_IPV4) == 0)
+ continue;
#endif
- if (inp->inp_lport != uh->uh_dport)
+ 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_unlock(inp->inp_socket, 1, 0);
continue;
+ }
+
+ if (inp->inp_lport != uh->uh_dport) {
+ udp_unlock(inp->inp_socket, 1, 0);
+ continue;
+ }
if (inp->inp_laddr.s_addr != INADDR_ANY) {
if (inp->inp_laddr.s_addr !=
- ip->ip_dst.s_addr)
+ ip->ip_dst.s_addr) {
+ udp_unlock(inp->inp_socket, 1, 0);
continue;
+ }
}
if (inp->inp_faddr.s_addr != INADDR_ANY) {
if (inp->inp_faddr.s_addr !=
ip->ip_src.s_addr ||
- inp->inp_fport != uh->uh_sport)
+ inp->inp_fport != uh->uh_sport) {
+ udp_unlock(inp->inp_socket, 1, 0);
continue;
+ }
}
if (last != NULL) {
struct mbuf *n;
-
#if IPSEC
+ int skipit = 0;
/* check AH/ESP integrity. */
- if (ipsec4_in_reject_so(m, last->inp_socket)) {
- ipsecstat.in_polvio++;
- /* do not inject data to pcb */
- } else
+ 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 (skipit == 0)
#endif /*IPSEC*/
if ((n = m_copy(m, 0, M_COPYALL)) != NULL) {
- udp_append(last, ip, n, iphlen +
- sizeof (struct udphdr));
+ udp_append(last, ip, n,
+ iphlen +
+ sizeof(struct udphdr));
}
+ udp_unlock(last->inp_socket, 1, 0);
}
last = inp;
/*
if ((last->inp_socket->so_options&(SO_REUSEPORT|SO_REUSEADDR)) == 0)
break;
}
+ lck_rw_done(pcbinfo->mtx);
if (last == NULL) {
/*
goto bad;
}
#if IPSEC
- else
/* check AH/ESP integrity. */
- if (m && ipsec4_in_reject_so(m, last->inp_socket)) {
- ipsecstat.in_polvio++;
- goto bad;
+ 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_append(last, ip, m, iphlen + sizeof(struct udphdr));
+ udp_unlock(last->inp_socket, 1, 0);
return;
}
+
+#if IPSEC
+ /*
+ * UDP to port 4500 with a payload where the first four bytes are
+ * not zero is a UDP encapsulated IPSec packet. Packets where
+ * the payload is one byte and that byte is 0xFF are NAT keepalive
+ * packets. Decapsulate the ESP packet and carry on with IPSec input
+ * 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) {
+ udpstat.udps_hdrops++;
+ 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);
+ }
+ /* Check for NAT keepalive packet */
+ 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);
+ return;
+ }
+ else if (payload_len == 4 && *(u_int32_t*)((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);
+ return;
+ }
+ }
+#endif
+
/*
* Locate pcb for datagram.
*/
ip->ip_dst, uh->uh_dport, 1, m->m_pkthdr.rcvif);
if (inp == NULL) {
if (log_in_vain) {
- char buf[4*sizeof "123"];
-
- strcpy(buf, inet_ntoa(ip->ip_dst));
- log(LOG_INFO,
- "Connection attempt to UDP %s:%d from %s:%d\n",
- buf, ntohs(uh->uh_dport), inet_ntoa(ip->ip_src),
- ntohs(uh->uh_sport));
+ 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)),
+ ntohs(uh->uh_sport));
+ 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)))
}
udpstat.udps_noport++;
if (m->m_flags & (M_BCAST | M_MCAST)) {
udpstat.udps_noportbcast++;
goto bad;
}
- *ip = save_ip;
#if ICMP_BANDLIM
- if (badport_bandlim(0) < 0)
+ if (badport_bandlim(BANDLIM_ICMP_UNREACH) < 0)
goto bad;
#endif
+ if (blackhole)
+ if (m->m_pkthdr.rcvif && m->m_pkthdr.rcvif->if_type != IFT_LOOP)
+ goto bad;
+ *ip = save_ip;
+ ip->ip_len += iphlen;
icmp_error(m, ICMP_UNREACH, ICMP_UNREACH_PORT, 0, 0);
KERNEL_DEBUG(DBG_FNC_UDP_INPUT | DBG_FUNC_END, 0,0,0,0,0);
return;
}
-#if IPSEC
- if (inp != NULL && ipsec4_in_reject_so(m, inp->inp_socket)) {
- ipsecstat.in_polvio++;
+ udp_lock(inp->inp_socket, 1, 0);
+
+ if (in_pcb_checkstate(inp, WNT_RELEASE, 1) == WNT_STOPUSING) {
+ udp_unlock(inp->inp_socket, 1, 0);
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);
+ udp_unlock(inp->inp_socket, 1, 0);
+ goto bad;
+ }
+ lck_mtx_unlock(sadb_mutex);
+ }
#endif /*IPSEC*/
/*
ip_2_ip6_hdr(&udp_ip6.uip6_ip6, ip);
savedflags = inp->inp_flags;
inp->inp_flags &= ~INP_UNMAPPABLEOPTS;
- ip6_savecontrol(inp, &udp_ip6.uip6_ip6, m,
- &opts6, NULL);
-
+ ip6_savecontrol(inp, &opts, &udp_ip6.uip6_ip6, m);
inp->inp_flags = savedflags;
} else
#endif
ip_savecontrol(inp, &opts, ip, m);
}
- 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);
if (inp->inp_vflag & INP_IPV6) {
in6_sin_2_v4mapsin6(&udp_in, &udp_in6.uin6_sin);
append_sa = (struct sockaddr *)&udp_in6;
- opts = opts6.head;
} else
#endif
append_sa = (struct sockaddr *)&udp_in;
- if (sbappendaddr(&inp->inp_socket->so_rcv, append_sa, m, opts) == 0) {
+ if (sbappendaddr(&inp->inp_socket->so_rcv, append_sa, m, opts, NULL) == 0) {
udpstat.udps_fullsock++;
- goto bad;
}
- sorwakeup(inp->inp_socket);
+ else {
+ sorwakeup(inp->inp_socket);
+ }
+ udp_unlock(inp->inp_socket, 1, 0);
KERNEL_DEBUG(DBG_FNC_UDP_INPUT | DBG_FUNC_END, 0,0,0,0,0);
return;
bad:
if (opts)
m_freem(opts);
KERNEL_DEBUG(DBG_FNC_UDP_INPUT | DBG_FUNC_END, 0,0,0,0,0);
+ return;
}
#if INET6
struct inpcb *last;
struct ip *ip;
struct mbuf *n;
+ int off;
{
struct sockaddr *append_sa;
struct mbuf *opts = 0;
-#if INET6
- struct ip6_recvpktopts opts6;
- bzero(&opts6, sizeof(opts6));
-#endif
-
if (last->inp_flags & INP_CONTROLOPTS ||
last->inp_socket->so_options & SO_TIMESTAMP) {
}
savedflags = last->inp_flags;
last->inp_flags &= ~INP_UNMAPPABLEOPTS;
- ip6_savecontrol(last, &udp_ip6.uip6_ip6, n,
- &opts6, NULL);
+ ip6_savecontrol(last, &opts, &udp_ip6.uip6_ip6, n);
last->inp_flags = savedflags;
} else
#endif
udp_in6.uin6_init_done = 1;
}
append_sa = (struct sockaddr *)&udp_in6.uin6_sin;
- opts = opts6.head;
} else
#endif
append_sa = (struct sockaddr *)&udp_in;
m_adj(n, off);
-
- if (sbappendaddr(&last->inp_socket->so_rcv, append_sa, n, opts) == 0) {
- m_freem(n);
- if (opts)
- m_freem(opts);
+ if (sbappendaddr(&last->inp_socket->so_rcv, append_sa, n, opts, NULL) == 0) {
udpstat.udps_fullsock++;
} else
sorwakeup(last->inp_socket);
}
-
-
/*
* Notify a udp user of an asynchronous error;
* just wake up so that he can collect error status.
struct sockaddr *sa;
void *vip;
{
- register struct ip *ip = vip;
- register struct udphdr *uh;
+ struct ip *ip = vip;
+ struct udphdr *uh;
+ void (*notify)(struct inpcb *, int) = udp_notify;
+ struct in_addr faddr;
+ struct inpcb *inp;
- if (!PRC_IS_REDIRECT(cmd) &&
- ((unsigned)cmd >= PRC_NCMDS || inetctlerrmap[cmd] == 0))
+ faddr = ((struct sockaddr_in *)sa)->sin_addr;
+ if (sa->sa_family != AF_INET || faddr.s_addr == INADDR_ANY)
+ return;
+
+ if (PRC_IS_REDIRECT(cmd)) {
+ ip = 0;
+ notify = in_rtchange;
+ } else if (cmd == PRC_HOSTDEAD)
+ ip = 0;
+ else if ((unsigned)cmd >= PRC_NCMDS || inetctlerrmap[cmd] == 0)
return;
if (ip) {
uh = (struct udphdr *)((caddr_t)ip + (ip->ip_hl << 2));
- in_pcbnotify(&udb, sa, uh->uh_dport, ip->ip_src, uh->uh_sport,
- cmd, udp_notify);
+ 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) {
+ udp_unlock(inp->inp_socket, 1, 0);
+ return;
+ }
+ (*notify)(inp, inetctlerrmap[cmd]);
+ udp_unlock(inp->inp_socket, 1, 0);
+ }
} else
- in_pcbnotify(&udb, sa, 0, zeroin_addr, 0, cmd, udp_notify);
+ in_pcbnotifyall(&udbinfo, faddr, inetctlerrmap[cmd], notify);
}
-
static int
udp_pcblist SYSCTL_HANDLER_ARGS
{
- int error, i, n, s;
+ 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.
*/
- if (req->oldptr == 0) {
+ lck_rw_lock_exclusive(udbinfo.mtx);
+ 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;
}
- if (req->newptr != 0)
+ if (req->newptr != USER_ADDR_NULL) {
+ lck_rw_done(udbinfo.mtx);
return EPERM;
+ }
/*
* OK, now we're committed to doing something.
*/
- s = splnet();
gencnt = udbinfo.ipi_gencnt;
n = udbinfo.ipi_count;
- splx(s);
+ 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)
+ if (error) {
+ lck_rw_done(udbinfo.mtx);
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);
if (inp_list == 0) {
+ lck_rw_done(udbinfo.mtx);
return ENOMEM;
}
- s = splnet();
- for (inp = udbinfo.listhead->lh_first, i = 0; inp && i < n;
- inp = inp->inp_list.le_next) {
- if (inp->inp_gencnt <= gencnt)
+
+ for (inp = LIST_FIRST(udbinfo.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;
}
- splx(s);
n = i;
error = 0;
for (i = 0; i < n; i++) {
inp = inp_list[i];
- if (inp->inp_gencnt <= gencnt) {
+ if (inp->inp_gencnt <= gencnt && inp->inp_state != INPCB_STATE_DEAD) {
struct xinpcb xi;
+
+ bzero(&xi, sizeof(xi));
xi.xi_len = sizeof xi;
/* XXX should avoid extra copy */
- bcopy(inp, &xi.xi_inp, sizeof *inp);
+ 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);
* while we were processing this request, and it
* might be necessary to retry.
*/
- s = splnet();
+ 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;
- splx(s);
error = SYSCTL_OUT(req, &xig, sizeof xig);
}
FREE(inp_list, M_TEMP);
+ lck_rw_done(udbinfo.mtx);
return error;
}
+static __inline__ u_int16_t
+get_socket_id(struct socket * s)
+{
+ u_int16_t val;
+
+ if (s == NULL) {
+ return (0);
+ }
+ val = (u_int16_t)(((u_int32_t)s) / sizeof(struct socket));
+ if (val == 0) {
+ val = 0xffff;
+ }
+ return (val);
+}
+
static int
udp_output(inp, m, addr, control, p)
register struct inpcb *inp;
- register struct mbuf *m;
+ struct mbuf *m;
struct sockaddr *addr;
struct mbuf *control;
struct proc *p;
{
register struct udpiphdr *ui;
register int len = m->m_pkthdr.len;
- struct in_addr laddr;
- int s = 0, error = 0;
+ struct sockaddr_in *sin, src;
+ struct in_addr origladdr, laddr, faddr;
+ u_short lport, fport;
+ struct sockaddr_in *ifaddr;
+ int error = 0, udp_dodisconnect = 0;
+
KERNEL_DEBUG(DBG_FNC_UDP_OUTPUT | DBG_FUNC_START, 0,0,0,0,0);
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 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_faddr.s_addr = INADDR_ANY; /* new src will be set later */
+ else {
+ error = EADDRNOTAVAIL;
+ goto release;
+ }
+ }
+ rtfree(inp->inp_route.ro_rt);
+ inp->inp_route.ro_rt = (struct rtentry *)0;
+ }
+
+ origladdr= laddr = inp->inp_laddr;
+ faddr = inp->inp_faddr;
+ lport = inp->inp_lport;
+ fport = inp->inp_fport;
+
if (addr) {
- laddr = inp->inp_laddr;
- if (inp->inp_faddr.s_addr != INADDR_ANY) {
+ sin = (struct sockaddr_in *)addr;
+ if (faddr.s_addr != INADDR_ANY) {
error = EISCONN;
goto release;
}
- /*
- * Must block input while temporarily connected.
- */
- s = splnet();
- error = in_pcbconnect(inp, addr, p);
- if (error) {
- splx(s);
- 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) {
+ goto release;
+ }
+ laddr = inp->inp_laddr;
+ lport = inp->inp_lport;
+ faddr = inp->inp_faddr;
+ fport = inp->inp_fport;
+ udp_dodisconnect = 1;
+ }
+ 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
+ */
+ 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.*/
+ }
+
+ faddr = sin->sin_addr;
+ fport = sin->sin_port;
}
} else {
- if (inp->inp_faddr.s_addr == INADDR_ANY) {
+ if (faddr.s_addr == INADDR_ANY) {
error = ENOTCONN;
goto release;
}
}
+
+
/*
* Calculate data length and get a mbuf
* for UDP and IP headers.
M_PREPEND(m, sizeof(struct udpiphdr), M_DONTWAIT);
if (m == 0) {
error = ENOBUFS;
- if (addr)
- splx(s);
- goto release;
+ goto abort;
}
/*
* and addresses and length put into network format.
*/
ui = mtod(m, struct udpiphdr *);
- bzero(ui->ui_x1, sizeof(ui->ui_x1));
+ bzero(ui->ui_x1, sizeof(ui->ui_x1)); /* XXX still needed? */
ui->ui_pr = IPPROTO_UDP;
- ui->ui_len = htons((u_short)len + sizeof (struct udphdr));
- ui->ui_src = inp->inp_laddr;
- ui->ui_dst = inp->inp_faddr;
- ui->ui_sport = inp->inp_lport;
- ui->ui_dport = inp->inp_fport;
- ui->ui_ulen = ui->ui_len;
+ 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));
/*
- * Stuff checksum and output datagram.
+ * Set up checksum and output datagram.
*/
- ui->ui_sum = 0;
if (udpcksum) {
- if ((ui->ui_sum = in_cksum(m, sizeof (struct udpiphdr) + len)) == 0)
- ui->ui_sum = 0xffff;
+ 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 {
+ ui->ui_sum = 0;
}
((struct ip *)ui)->ip_len = sizeof (struct udpiphdr) + len;
((struct ip *)ui)->ip_ttl = inp->inp_ip_ttl; /* XXX */
KERNEL_DEBUG(DBG_LAYER_OUT_END, ui->ui_dport, ui->ui_sport,
ui->ui_src.s_addr, ui->ui_dst.s_addr, ui->ui_ulen);
-
#if IPSEC
- ipsec_setsocket(m, inp->inp_socket);
+ if (ipsec_bypass == 0 && ipsec_setsocket(m, inp->inp_socket) != 0) {
+ error = ENOBUFS;
+ goto abort;
+ }
#endif /*IPSEC*/
-
- error = ip_output(m, inp->inp_options, &inp->inp_route,
- inp->inp_socket->so_options & (SO_DONTROUTE | SO_BROADCAST),
+ 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);
- if (addr) {
+ if (udp_dodisconnect) {
in_pcbdisconnect(inp);
- inp->inp_laddr = laddr; /* XXX rehash? */
- splx(s);
+ inp->inp_laddr = origladdr; /* XXX rehash? */
}
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? */
+ }
+
release:
m_freem(m);
KERNEL_DEBUG(DBG_FNC_UDP_OUTPUT | DBG_FUNC_END, error, 0,0,0,0);
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, "");
+ &udp_sendspace, 0, "Maximum outgoing UDP datagram size");
-
-u_long udp_recvspace = 40 * (1024 + /* 40 1K datagrams */
+u_long udp_recvspace = 40 * (1024 +
#if INET6
sizeof(struct sockaddr_in6)
-#else /* INET6 */
+#else
sizeof(struct sockaddr_in)
-#endif /* INET6 */
+#endif
);
SYSCTL_INT(_net_inet_udp, UDPCTL_RECVSPACE, recvspace, CTLFLAG_RW,
- &udp_recvspace, 0, "");
+ &udp_recvspace, 0, "Maximum incoming UDP datagram size");
static int
udp_abort(struct socket *so)
{
struct inpcb *inp;
- int s;
inp = sotoinpcb(so);
if (inp == 0)
- return EINVAL; /* ??? possible? panic instead? */
+ panic("udp_abort: so=%x null inp\n", so); /* ??? possible? panic instead? */
soisdisconnected(so);
- s = splnet();
in_pcbdetach(inp);
- splx(s);
return 0;
}
udp_attach(struct socket *so, int proto, struct proc *p)
{
struct inpcb *inp;
- int error; long s;
+ int error;
+
+ inp = sotoinpcb(so);
+ if (inp != 0)
+ panic ("udp_attach so=%x inp=%x\n", so, inp);
- if (so->so_snd.sb_hiwat == 0 || so->so_rcv.sb_hiwat == 0) {
- error = soreserve(so, udp_sendspace, udp_recvspace);
- if (error)
- return error;
- }
- s = splnet();
error = in_pcballoc(so, &udbinfo, p);
- splx(s);
if (error)
return error;
error = soreserve(so, udp_sendspace, udp_recvspace);
- if (error)
+ if (error)
return error;
inp = (struct inpcb *)so->so_pcb;
inp->inp_vflag |= INP_IPV4;
inp->inp_ip_ttl = ip_defttl;
-#if IPSEC
- error = ipsec_init_policy(so, &inp->inp_sp);
- if (error != 0) {
- in_pcbdetach(inp);
- return error;
- }
-#endif /*IPSEC*/
return 0;
}
udp_bind(struct socket *so, struct sockaddr *nam, struct proc *p)
{
struct inpcb *inp;
- int s, error;
+ int error;
inp = sotoinpcb(so);
if (inp == 0)
return EINVAL;
- s = splnet();
error = in_pcbbind(inp, nam, p);
- splx(s);
return error;
}
udp_connect(struct socket *so, struct sockaddr *nam, struct proc *p)
{
struct inpcb *inp;
- int s, error;
+ int error;
inp = sotoinpcb(so);
if (inp == 0)
return EINVAL;
if (inp->inp_faddr.s_addr != INADDR_ANY)
return EISCONN;
- s = splnet();
error = in_pcbconnect(inp, nam, p);
- splx(s);
- if (error == 0)
+ if (error == 0)
soisconnected(so);
return error;
}
udp_detach(struct socket *so)
{
struct inpcb *inp;
- int s;
inp = sotoinpcb(so);
if (inp == 0)
- return EINVAL;
- s = splnet();
+ panic("udp_detach: so=%x null inp\n", so); /* ??? possible? panic instead? */
in_pcbdetach(inp);
- splx(s);
+ inp->inp_state = INPCB_STATE_DEAD;
return 0;
}
udp_disconnect(struct socket *so)
{
struct inpcb *inp;
- int s;
inp = sotoinpcb(so);
if (inp == 0)
if (inp->inp_faddr.s_addr == INADDR_ANY)
return ENOTCONN;
- s = splnet();
in_pcbdisconnect(inp);
inp->inp_laddr.s_addr = INADDR_ANY;
- splx(s);
so->so_state &= ~SS_ISCONNECTED; /* XXX */
return 0;
}
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, sopoll
+ in_setsockaddr, sosend, soreceive, pru_sopoll_notsupp
};
+
+int
+udp_lock(so, refcount, debug)
+ struct socket *so;
+ int refcount, 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);
+
+ if (refcount)
+ so->so_usecount++;
+
+ so->lock_lr[so->next_lock_lr] = (void *)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;
+{
+ int lr_saved;
+ struct inpcb *inp = sotoinpcb(so);
+ struct inpcbinfo *pcbinfo = &udbinfo;
+
+ if (debug == 0)
+ lr_saved = (unsigned int) __builtin_return_address(0);
+ else lr_saved = debug;
+
+ 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;
+ so->next_unlock_lr = (so->next_unlock_lr+1) % SO_LCKDBG_MAX;
+ lck_mtx_unlock(((struct inpcb *)so->so_pcb)->inpcb_mtx);
+ }
+
+
+ return (0);
+}
+
+lck_mtx_t *
+udp_getlock(so, locktype)
+ struct socket *so;
+ int 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);
+ }
+}
+
+void
+udp_slowtimo()
+{
+ struct inpcb *inp, *inpnxt;
+ struct socket *so;
+ struct inpcbinfo *pcbinfo = &udbinfo;
+
+ lck_rw_lock_exclusive(pcbinfo->mtx);
+
+ 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)
+ continue;
+
+ so = inp->inp_socket;
+ if (!lck_mtx_try_lock(inp->inpcb_mtx)) /* skip if busy, no hurry for cleanup... */
+ continue;
+
+ if (so->so_usecount == 0)
+ in_pcbdispose(inp);
+ else
+ lck_mtx_unlock(inp->inpcb_mtx);
+ }
+ lck_rw_done(pcbinfo->mtx);
+}
+
+int
+ChkAddressOK( __uint32_t dstaddr, __uint32_t srcaddr )
+{
+ if ( dstaddr == srcaddr ){
+ return 0;
+ }
+ return 1;
+}
+
projects / apple / xnu.git / blobdiff