freebsd-nq/sys/netinet6/in6_proto.c

561 lines
18 KiB
C
Raw Normal View History

/* $FreeBSD$ */
/* $KAME: in6_proto.c,v 1.64 2000/06/20 16:20:27 itojun Exp $ */
/*
* Copyright (C) 1995, 1996, 1997, and 1998 WIDE Project.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. Neither the name of the project nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE PROJECT AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE PROJECT OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
/*
* Copyright (c) 1982, 1986, 1993
* The Regents of the University of California. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by the University of
* California, Berkeley and its contributors.
* 4. Neither the name of the University nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
* @(#)in_proto.c 8.1 (Berkeley) 6/10/93
*/
#include "opt_inet.h"
#include "opt_inet6.h"
#include "opt_ipsec.h"
#include <sys/param.h>
#include <sys/socket.h>
#include <sys/socketvar.h>
#include <sys/protosw.h>
#include <sys/kernel.h>
#include <sys/domain.h>
#include <sys/mbuf.h>
#include <sys/systm.h>
#include <sys/sysctl.h>
#include <net/if.h>
#include <net/radix.h>
#include <net/route.h>
#include <netinet/in.h>
#include <netinet/in_systm.h>
#include <netinet/in_var.h>
#include <netinet/ip_encap.h>
#include <netinet/ip.h>
#include <netinet/ip_var.h>
#include <netinet/ip6.h>
#include <netinet6/ip6_var.h>
#include <netinet/icmp6.h>
#include <netinet/tcp.h>
#include <netinet/tcp_timer.h>
#include <netinet/tcp_var.h>
#include <netinet/udp.h>
#include <netinet/udp_var.h>
#include <netinet6/tcp6_var.h>
#include <netinet6/udp6_var.h>
#include <netinet6/pim6_var.h>
#include <netinet6/nd6.h>
#include <netinet6/in6_prefix.h>
#ifdef IPSEC
#include <netinet6/ipsec.h>
#include <netinet6/ipsec6.h>
#include <netinet6/ah.h>
#include <netinet6/ah6.h>
#ifdef IPSEC_ESP
#include <netinet6/esp.h>
#include <netinet6/esp6.h>
#endif
#include <netinet6/ipcomp.h>
#include <netinet6/ipcomp6.h>
#endif /*IPSEC*/
#include <netinet6/ip6protosw.h>
#include "gif.h"
#if NGIF > 0
#include <netinet6/in6_gif.h>
#endif
#include <net/net_osdep.h>
#define offsetof(type, member) ((size_t)(&((type *)0)->member))
/*
* TCP/IP protocol family: IP6, ICMP6, UDP, TCP.
*/
extern struct domain inet6domain;
static struct pr_usrreqs nousrreqs;
struct ip6protosw inet6sw[] = {
{ 0, &inet6domain, IPPROTO_IPV6, 0,
0, 0, 0, 0,
0,
ip6_init, 0, frag6_slowtimo, frag6_drain,
&nousrreqs,
},
{ SOCK_DGRAM, &inet6domain, IPPROTO_UDP, PR_ATOMIC | PR_ADDR,
udp6_input, 0, udp6_ctlinput, ip6_ctloutput,
0,
0, 0, 0, 0,
&udp6_usrreqs,
},
{ SOCK_STREAM, &inet6domain, IPPROTO_TCP, PR_CONNREQUIRED | PR_WANTRCVD,
tcp6_input, 0, tcp6_ctlinput, tcp_ctloutput,
0,
#ifdef INET /* don't call timeout routines twice */
tcp_init, 0, 0, tcp_drain,
#else
tcp_init, tcp_fasttimo, tcp_slowtimo, tcp_drain,
#endif
&tcp6_usrreqs,
},
{ SOCK_RAW, &inet6domain, IPPROTO_RAW, PR_ATOMIC | PR_ADDR,
rip6_input, rip6_output, rip6_ctlinput, rip6_ctloutput,
0,
0, 0, 0, 0,
&rip6_usrreqs
},
{ SOCK_RAW, &inet6domain, IPPROTO_ICMPV6, PR_ATOMIC | PR_ADDR,
icmp6_input, rip6_output, 0, rip6_ctloutput,
0,
icmp6_init, icmp6_fasttimo, 0, 0,
&rip6_usrreqs
},
{ SOCK_RAW, &inet6domain, IPPROTO_DSTOPTS,PR_ATOMIC|PR_ADDR,
dest6_input, 0, 0, 0,
0,
0, 0, 0, 0,
&nousrreqs
},
{ SOCK_RAW, &inet6domain, IPPROTO_ROUTING,PR_ATOMIC|PR_ADDR,
route6_input, 0, 0, 0,
0,
0, 0, 0, 0,
&nousrreqs
},
{ SOCK_RAW, &inet6domain, IPPROTO_FRAGMENT,PR_ATOMIC|PR_ADDR,
frag6_input, 0, 0, 0,
0,
0, 0, 0, 0,
&nousrreqs
},
#ifdef IPSEC
{ SOCK_RAW, &inet6domain, IPPROTO_AH, PR_ATOMIC|PR_ADDR,
ah6_input, 0, 0, 0,
0,
0, 0, 0, 0,
&nousrreqs,
},
#ifdef IPSEC_ESP
{ SOCK_RAW, &inet6domain, IPPROTO_ESP, PR_ATOMIC|PR_ADDR,
esp6_input, 0, 0, 0,
0,
0, 0, 0, 0,
&nousrreqs,
},
#endif
{ SOCK_RAW, &inet6domain, IPPROTO_IPCOMP, PR_ATOMIC|PR_ADDR,
ipcomp6_input, 0, 0, 0,
0,
0, 0, 0, 0,
&nousrreqs,
},
#endif /* IPSEC */
#ifdef INET
{ SOCK_RAW, &inet6domain, IPPROTO_IPV4, PR_ATOMIC|PR_ADDR,
encap6_input, rip6_output, 0, rip6_ctloutput,
0,
0, 0, 0, 0,
&rip6_usrreqs
},
#endif /*INET*/
{ SOCK_RAW, &inet6domain, IPPROTO_IPV6, PR_ATOMIC|PR_ADDR,
encap6_input, rip6_output, 0, rip6_ctloutput,
0,
#ifndef INET
encap_init, 0, 0, 0,
#else
0, 0, 0, 0,
#endif
&rip6_usrreqs
},
{ SOCK_RAW, &inet6domain, IPPROTO_PIM, PR_ATOMIC|PR_ADDR,
pim6_input, rip6_output, 0, rip6_ctloutput,
0,
0, 0, 0, 0,
&rip6_usrreqs
},
/* raw wildcard */
{ SOCK_RAW, &inet6domain, 0, PR_ATOMIC | PR_ADDR,
rip6_input, rip6_output, 0, rip6_ctloutput,
0, 0,
0, 0, 0,
&rip6_usrreqs
},
};
#if NGIF > 0
struct ip6protosw in6_gif_protosw =
{ SOCK_RAW, &inet6domain, 0/*IPPROTO_IPV[46]*/, PR_ATOMIC|PR_ADDR,
in6_gif_input, rip6_output, 0, rip6_ctloutput,
0,
0, 0, 0, 0,
&rip6_usrreqs
};
#endif /*NGIF*/
extern int in6_inithead __P((void **, int));
struct domain inet6domain =
{ AF_INET6, "internet6", 0, 0, 0,
(struct protosw *)inet6sw,
(struct protosw *)&inet6sw[sizeof(inet6sw)/sizeof(inet6sw[0])], 0,
in6_inithead,
offsetof(struct sockaddr_in6, sin6_addr) << 3,
sizeof(struct sockaddr_in6) };
DOMAIN_SET(inet6);
/*
* Internet configuration info
*/
#ifndef IPV6FORWARDING
#ifdef GATEWAY6
#define IPV6FORWARDING 1 /* forward IP6 packets not for us */
#else
#define IPV6FORWARDING 0 /* don't forward IP6 packets not for us */
#endif /* GATEWAY6 */
#endif /* !IPV6FORWARDING */
#ifndef IPV6_SENDREDIRECTS
#define IPV6_SENDREDIRECTS 1
#endif
int ip6_forwarding = IPV6FORWARDING; /* act as router? */
int ip6_sendredirects = IPV6_SENDREDIRECTS;
int ip6_defhlim = IPV6_DEFHLIM;
int ip6_defmcasthlim = IPV6_DEFAULT_MULTICAST_HOPS;
int ip6_accept_rtadv = 0; /* "IPV6FORWARDING ? 0 : 1" is dangerous */
int ip6_maxfragpackets = 200;
int ip6_log_interval = 5;
int ip6_hdrnestlimit = 50; /* appropriate? */
int ip6_dad_count = 1; /* DupAddrDetectionTransmits */
u_int32_t ip6_flow_seq;
int ip6_auto_flowlabel = 1;
#if NGIF > 0
int ip6_gif_hlim = GIF_HLIM;
#else
int ip6_gif_hlim = 0;
#endif
int ip6_use_deprecated = 1; /* allow deprecated addr (RFC2462 5.5.4) */
int ip6_rr_prune = 5; /* router renumbering prefix
* walk list every 5 sec. */
int ip6_mapped_addr_on = 1;
u_int32_t ip6_id = 0UL;
int ip6_keepfaith = 0;
time_t ip6_log_time = (time_t)0L;
/* icmp6 */
/*
* BSDI4 defines these variables in in_proto.c...
* XXX: what if we don't define INET? Should we define pmtu6_expire
* or so? (jinmei@kame.net 19990310)
*/
int pmtu_expire = 60*10;
int pmtu_probe = 60*2;
/* raw IP6 parameters */
/*
* Nominal space allocated to a raw ip socket.
*/
#define RIPV6SNDQ 8192
#define RIPV6RCVQ 8192
u_long rip6_sendspace = RIPV6SNDQ;
u_long rip6_recvspace = RIPV6RCVQ;
/* ICMPV6 parameters */
int icmp6_rediraccept = 1; /* accept and process redirects */
int icmp6_redirtimeout = 10 * 60; /* 10 minutes */
struct timeval icmp6errratelim = { 0, 0 }; /* no ratelimit */
int icmp6errppslim = 100; /* 100pps */
int icmp6_nodeinfo = 1; /* enable/disable NI response */
#ifdef TCP6
/* TCP on IP6 parameters */
int tcp6_sendspace = 1024 * 8;
int tcp6_recvspace = 1024 * 8;
int tcp6_mssdflt = TCP6_MSS;
int tcp6_rttdflt = TCP6TV_SRTTDFLT / PR_SLOWHZ;
int tcp6_do_rfc1323 = 1;
int tcp6_conntimeo = TCP6TV_KEEP_INIT; /* initial connection timeout */
int tcp6_43maxseg = 0;
int tcp6_pmtu = 0;
/*
* Parameters for keepalive option.
* Connections for which SO_KEEPALIVE is set will be probed
* after being idle for a time of tcp6_keepidle (in units of PR_SLOWHZ).
* Starting at that time, the connection is probed at intervals
* of tcp6_keepintvl (same units) until a response is received
* or until tcp6_keepcnt probes have been made, at which time
* the connection is dropped. Note that a tcp6_keepidle value
* under 2 hours is nonconformant with RFC-1122, Internet Host Requirements.
*/
int tcp6_keepidle = TCP6TV_KEEP_IDLE; /* time before probing idle */
int tcp6_keepintvl = TCP6TV_KEEPINTVL; /* interval betwn idle probes */
int tcp6_keepcnt = TCP6TV_KEEPCNT; /* max idle probes */
int tcp6_maxpersistidle = TCP6TV_KEEP_IDLE; /* max idle time in persist */
#ifndef INET_SERVER
#define TCP6_LISTEN_HASH_SIZE 17
#define TCP6_CONN_HASH_SIZE 97
#define TCP6_SYN_HASH_SIZE 293
#define TCP6_SYN_BUCKET_SIZE 35
#else
#define TCP6_LISTEN_HASH_SIZE 97
#define TCP6_CONN_HASH_SIZE 9973
#define TCP6_SYN_HASH_SIZE 997
#define TCP6_SYN_BUCKET_SIZE 35
#endif
int tcp6_listen_hash_size = TCP6_LISTEN_HASH_SIZE;
int tcp6_conn_hash_size = TCP6_CONN_HASH_SIZE;
struct tcp6_hash_list tcp6_listen_hash[TCP6_LISTEN_HASH_SIZE],
tcp6_conn_hash[TCP6_CONN_HASH_SIZE];
int tcp6_syn_cache_size = TCP6_SYN_HASH_SIZE;
int tcp6_syn_cache_limit = TCP6_SYN_HASH_SIZE*TCP6_SYN_BUCKET_SIZE;
int tcp6_syn_bucket_limit = 3*TCP6_SYN_BUCKET_SIZE;
struct syn_cache_head6 tcp6_syn_cache[TCP6_SYN_HASH_SIZE];
struct syn_cache_head6 *tcp6_syn_cache_first;
int tcp6_syn_cache_interval = 8; /* runs timer every 4 seconds */
int tcp6_syn_cache_timeo = TCP6TV_KEEP_INIT;
/*
* Parameters for computing a desirable data segment size
* given an upper bound (either interface MTU, or peer's MSS option)_.
* As applications tend to use a buffer size that is a multiple
* of kilobytes, try for something that divides evenly. However,
* do not round down too much.
*
* Round segment size down to a multiple of TCP6_ROUNDSIZE if this
* does not result in lowering by more than (size/TCP6_ROUNDFRAC).
* For example, round 536 to 512. Older versions of the system
* effectively used MCLBYTES (1K or 2K) as TCP6_ROUNDSIZE, with
* a value of 1 for TCP6_ROUNDFRAC (eliminating its effect).
* We round to a multiple of 256 for SLIP.
*/
#ifndef TCP6_ROUNDSIZE
#define TCP6_ROUNDSIZE 256 /* round to multiple of 256 */
#endif
#ifndef TCP6_ROUNDFRAC
#define TCP6_ROUNDFRAC 10 /* round down at most N/10, or 10% */
#endif
int tcp6_roundsize = TCP6_ROUNDSIZE;
int tcp6_roundfrac = TCP6_ROUNDFRAC;
#endif /*TCP6*/
/* UDP on IP6 parameters */
int udp6_sendspace = 9216; /* really max datagram size */
int udp6_recvspace = 40 * (1024 + sizeof(struct sockaddr_in6));
/* 40 1K datagrams */
/*
* sysctl related items.
*/
SYSCTL_NODE(_net, PF_INET6, inet6, CTLFLAG_RW, 0,
"Internet6 Family");
/* net.inet6 */
SYSCTL_NODE(_net_inet6, IPPROTO_IPV6, ip6, CTLFLAG_RW, 0, "IP6");
SYSCTL_NODE(_net_inet6, IPPROTO_ICMPV6, icmp6, CTLFLAG_RW, 0, "ICMP6");
SYSCTL_NODE(_net_inet6, IPPROTO_UDP, udp6, CTLFLAG_RW, 0, "UDP6");
SYSCTL_NODE(_net_inet6, IPPROTO_TCP, tcp6, CTLFLAG_RW, 0, "TCP6");
#ifdef IPSEC
SYSCTL_NODE(_net_inet6, IPPROTO_ESP, ipsec6, CTLFLAG_RW, 0, "IPSEC6");
#endif /* IPSEC */
/* net.inet6.ip6 */
static int
sysctl_ip6_forwarding(SYSCTL_HANDLER_ARGS)
{
int error = 0;
int old_ip6_forwarding;
int changed;
error = SYSCTL_OUT(req, arg1, sizeof(int));
if (error || !req->newptr)
return (error);
old_ip6_forwarding = ip6_forwarding;
error = SYSCTL_IN(req, arg1, sizeof(int));
if (error != 0)
return (error);
changed = (ip6_forwarding ? 1 : 0) ^ (old_ip6_forwarding ? 1 : 0);
if (changed == 0)
return (error);
/*
* XXX while host->router removes prefix got from RA,
* router->host case nukes all the prefixes managed by in6_prefix.c
* (both RR and static). therefore, switching from host->router->host
* will remove statically configured addresses/prefixes.
* not sure if it is intended behavior or not.
*/
if (ip6_forwarding != 0) { /* host becomes router */
int s = splnet();
struct nd_prefix *pr, *next;
for (pr = nd_prefix.lh_first; pr; pr = next) {
next = pr->ndpr_next;
if (!IN6_IS_ADDR_UNSPECIFIED(&pr->ndpr_addr))
in6_ifdel(pr->ndpr_ifp, &pr->ndpr_addr);
prelist_remove(pr);
}
splx(s);
} else { /* router becomes host */
while(!LIST_EMPTY(&rr_prefix))
delete_each_prefix(LIST_FIRST(&rr_prefix),
PR_ORIG_KERNEL);
}
return (error);
}
static int
sysctl_icmp6_ratelimit (SYSCTL_HANDLER_ARGS)
{
int rate_usec, error, s;
/*
* The sysctl specifies the rate in usec-between-icmp,
* so we must convert from/to a timeval.
*/
rate_usec = (icmp6errratelim.tv_sec * 1000000) +
icmp6errratelim.tv_usec;
error = sysctl_handle_int(oidp, &rate_usec, 0, req);
if (error)
return (error);
if (rate_usec < 0)
return (EINVAL);
s = splnet();
icmp6errratelim.tv_sec = rate_usec / 1000000;
icmp6errratelim.tv_usec = rate_usec % 1000000;
splx(s);
return (0);
}
SYSCTL_OID(_net_inet6_ip6, IPV6CTL_FORWARDING, forwarding,
CTLTYPE_INT|CTLFLAG_RW, &ip6_forwarding, 0, sysctl_ip6_forwarding,
"I", "");
SYSCTL_INT(_net_inet6_ip6, IPV6CTL_SENDREDIRECTS,
redirect, CTLFLAG_RW, &ip6_sendredirects, 0, "");
SYSCTL_INT(_net_inet6_ip6, IPV6CTL_DEFHLIM,
hlim, CTLFLAG_RW, &ip6_defhlim, 0, "");
SYSCTL_INT(_net_inet6_ip6, IPV6CTL_MAXFRAGPACKETS,
maxfragpackets, CTLFLAG_RW, &ip6_maxfragpackets, 0, "");
SYSCTL_INT(_net_inet6_ip6, IPV6CTL_ACCEPT_RTADV,
accept_rtadv, CTLFLAG_RW, &ip6_accept_rtadv, 0, "");
SYSCTL_INT(_net_inet6_ip6, IPV6CTL_KEEPFAITH,
keepfaith, CTLFLAG_RW, &ip6_keepfaith, 0, "");
SYSCTL_INT(_net_inet6_ip6, IPV6CTL_LOG_INTERVAL,
log_interval, CTLFLAG_RW, &ip6_log_interval, 0, "");
SYSCTL_INT(_net_inet6_ip6, IPV6CTL_HDRNESTLIMIT,
hdrnestlimit, CTLFLAG_RW, &ip6_hdrnestlimit, 0, "");
SYSCTL_INT(_net_inet6_ip6, IPV6CTL_DAD_COUNT,
dad_count, CTLFLAG_RW, &ip6_dad_count, 0, "");
SYSCTL_INT(_net_inet6_ip6, IPV6CTL_AUTO_FLOWLABEL,
auto_flowlabel, CTLFLAG_RW, &ip6_auto_flowlabel, 0, "");
SYSCTL_INT(_net_inet6_ip6, IPV6CTL_DEFMCASTHLIM,
defmcasthlim, CTLFLAG_RW, &ip6_defmcasthlim, 0, "");
SYSCTL_INT(_net_inet6_ip6, IPV6CTL_GIF_HLIM,
gifhlim, CTLFLAG_RW, &ip6_gif_hlim, 0, "");
SYSCTL_STRING(_net_inet6_ip6, IPV6CTL_KAME_VERSION,
kame_version, CTLFLAG_RD, __KAME_VERSION, 0, "");
SYSCTL_INT(_net_inet6_ip6, IPV6CTL_USE_DEPRECATED,
use_deprecated, CTLFLAG_RW, &ip6_use_deprecated, 0, "");
SYSCTL_INT(_net_inet6_ip6, IPV6CTL_RR_PRUNE,
rr_prune, CTLFLAG_RW, &ip6_rr_prune, 0, "");
SYSCTL_INT(_net_inet6_ip6, IPV6CTL_MAPPED_ADDR,
mapped_addr, CTLFLAG_RW, &ip6_mapped_addr_on, 0, "");
/* net.inet6.icmp6 */
SYSCTL_INT(_net_inet6_icmp6, ICMPV6CTL_REDIRACCEPT,
rediraccept, CTLFLAG_RW, &icmp6_rediraccept, 0, "");
SYSCTL_INT(_net_inet6_icmp6, ICMPV6CTL_REDIRTIMEOUT,
redirtimeout, CTLFLAG_RW, &icmp6_redirtimeout, 0, "");
SYSCTL_STRUCT(_net_inet6_icmp6, ICMPV6CTL_STATS, stats, CTLFLAG_RD,
&icmp6stat, icmp6stat, "");
SYSCTL_PROC(_net_inet6_icmp6, ICMPV6CTL_ERRRATELIMIT,
errratelimit, CTLTYPE_INT|CTLFLAG_RW,
0, sizeof(int), sysctl_icmp6_ratelimit, "I", "");
SYSCTL_INT(_net_inet6_icmp6, ICMPV6CTL_ND6_PRUNE,
nd6_prune, CTLFLAG_RW, &nd6_prune, 0, "");
SYSCTL_INT(_net_inet6_icmp6, ICMPV6CTL_ND6_DELAY,
nd6_delay, CTLFLAG_RW, &nd6_delay, 0, "");
SYSCTL_INT(_net_inet6_icmp6, ICMPV6CTL_ND6_UMAXTRIES,
nd6_umaxtries, CTLFLAG_RW, &nd6_umaxtries, 0, "");
SYSCTL_INT(_net_inet6_icmp6, ICMPV6CTL_ND6_MMAXTRIES,
nd6_mmaxtries, CTLFLAG_RW, &nd6_mmaxtries, 0, "");
SYSCTL_INT(_net_inet6_icmp6, ICMPV6CTL_ND6_USELOOPBACK,
nd6_useloopback, CTLFLAG_RW, &nd6_useloopback, 0, "");
SYSCTL_INT(_net_inet6_icmp6, ICMPV6CTL_NODEINFO,
nodeinfo, CTLFLAG_RW, &icmp6_nodeinfo, 0, "");
SYSCTL_INT(_net_inet6_icmp6, ICMPV6CTL_ERRPPSLIMIT,
errppslimit, CTLFLAG_RW, &icmp6errppslim, 0, "");
SYSCTL_INT(_net_inet6_icmp6, ICMPV6CTL_ND6_MAXNUDHINT,
nd6_maxnudhint, CTLFLAG_RW, &nd6_maxnudhint, 0, "");