freebsd-skq/sys/netinet/ip_output.c

/*
 * Copyright (c) 1982, 1986, 1988, 1990, 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.
 *
 *	@(#)ip_output.c	8.3 (Berkeley) 1/21/94
 *	$Id: ip_output.c,v 1.77 1998/07/06 05:00:53 julian Exp $
 */

#define _IP_VHL

#include "opt_ipfw.h"
#include "opt_ipdivert.h"
#include "opt_ipfilter.h"

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/malloc.h>
#include <sys/mbuf.h>
#include <sys/protosw.h>
#include <sys/socket.h>
#include <sys/socketvar.h>

#include <net/if.h>
#include <net/route.h>

#include <netinet/in.h>
#include <netinet/in_systm.h>
#include <netinet/ip.h>
#include <netinet/in_pcb.h>
#include <netinet/in_var.h>
#include <netinet/ip_var.h>

#ifdef vax
#include <machine/mtpr.h>
#endif
#include <machine/in_cksum.h>

static MALLOC_DEFINE(M_IPMOPTS, "ip_moptions", "internet multicast options");

#if !defined(COMPAT_IPFW) || COMPAT_IPFW == 1
#undef COMPAT_IPFW
#define COMPAT_IPFW 1
#else
#undef COMPAT_IPFW
#endif

#ifdef IPFIREWALL_FORWARD_DEBUG
#define print_ip(a)	 printf("%ld.%ld.%ld.%ld",(ntohl(a.s_addr)>>24)&0xFF,\
				 		  (ntohl(a.s_addr)>>16)&0xFF,\
						  (ntohl(a.s_addr)>>8)&0xFF,\
						  (ntohl(a.s_addr))&0xFF);
#endif

u_short ip_id;

static struct mbuf *ip_insertoptions __P((struct mbuf *, struct mbuf *, int *));
static void	ip_mloopback
	__P((struct ifnet *, struct mbuf *, struct sockaddr_in *, int));
static int	ip_getmoptions
	__P((int, struct ip_moptions *, struct mbuf **));
static int	ip_pcbopts __P((struct mbuf **, struct mbuf *));
static int	ip_setmoptions
	__P((int, struct ip_moptions **, struct mbuf *));

#if defined(IPFILTER_LKM) || defined(IPFILTER)
int	ip_optcopy __P((struct ip *, struct ip *));
extern int (*fr_checkp) __P((struct ip *, int, struct ifnet *, int, struct mbuf **));
#else
static int	ip_optcopy __P((struct ip *, struct ip *));
#endif


extern	struct protosw inetsw[];

/*
 * IP output.  The packet in mbuf chain m contains a skeletal IP
 * header (with len, off, ttl, proto, tos, src, dst).
 * The mbuf chain containing the packet will be freed.
 * The mbuf opt, if present, will not be freed.
 */
int
ip_output(m0, opt, ro, flags, imo)
	struct mbuf *m0;
	struct mbuf *opt;
	struct route *ro;
	int flags;
	struct ip_moptions *imo;
{
	struct ip *ip, *mhip;
	struct ifnet *ifp;
	struct mbuf *m = m0;
	int hlen = sizeof (struct ip);
	int len, off, error = 0;
	struct sockaddr_in *dst;
	struct in_ifaddr *ia;
	int isbroadcast;
#ifdef IPFIREWALL_FORWARD
	int fwd_rewrite_src = 0;
#endif

#ifdef	DIAGNOSTIC
	if ((m->m_flags & M_PKTHDR) == 0)
		panic("ip_output no HDR");
	if (!ro)
		panic("ip_output no route, proto = %d",
		      mtod(m, struct ip *)->ip_p);
#endif
	if (opt) {
		m = ip_insertoptions(m, opt, &len);
		hlen = len;
	}
	ip = mtod(m, struct ip *);
	/*
	 * Fill in IP header.
	 */
	if ((flags & (IP_FORWARDING|IP_RAWOUTPUT)) == 0) {
		ip->ip_vhl = IP_MAKE_VHL(IPVERSION, hlen >> 2);
		ip->ip_off &= IP_DF;
		ip->ip_id = htons(ip_id++);
		ipstat.ips_localout++;
	} else {
		hlen = IP_VHL_HL(ip->ip_vhl) << 2;
	}

	dst = (struct sockaddr_in *)&ro->ro_dst;
	/*
	 * If there is a cached route,
	 * check that it is to the same destination
	 * and is still up.  If not, free it and try again.
	 */
	if (ro->ro_rt && ((ro->ro_rt->rt_flags & RTF_UP) == 0 ||
	   dst->sin_addr.s_addr != ip->ip_dst.s_addr)) {
		RTFREE(ro->ro_rt);
		ro->ro_rt = (struct rtentry *)0;
	}
	if (ro->ro_rt == 0) {
		dst->sin_family = AF_INET;
		dst->sin_len = sizeof(*dst);
		dst->sin_addr = ip->ip_dst;
	}
	/*
	 * If routing to interface only,
	 * short circuit routing lookup.
	 */
#define ifatoia(ifa)	((struct in_ifaddr *)(ifa))
#define sintosa(sin)	((struct sockaddr *)(sin))
	if (flags & IP_ROUTETOIF) {
		if ((ia = ifatoia(ifa_ifwithdstaddr(sintosa(dst)))) == 0 &&
		    (ia = ifatoia(ifa_ifwithnet(sintosa(dst)))) == 0) {
			ipstat.ips_noroute++;
			error = ENETUNREACH;
			goto bad;
		}
		ifp = ia->ia_ifp;
		ip->ip_ttl = 1;
		isbroadcast = in_broadcast(dst->sin_addr, ifp);
	} else {
		/*
		 * If this is the case, we probably don't want to allocate
		 * a protocol-cloned route since we didn't get one from the
		 * ULP.  This lets TCP do its thing, while not burdening
		 * forwarding or ICMP with the overhead of cloning a route.
		 * Of course, we still want to do any cloning requested by
		 * the link layer, as this is probably required in all cases
		 * for correct operation (as it is for ARP).
		 */
		if (ro->ro_rt == 0)
			rtalloc_ign(ro, RTF_PRCLONING);
		if (ro->ro_rt == 0) {
			ipstat.ips_noroute++;
			error = EHOSTUNREACH;
			goto bad;
		}
		ia = ifatoia(ro->ro_rt->rt_ifa);
		ifp = ro->ro_rt->rt_ifp;
		ro->ro_rt->rt_use++;
		if (ro->ro_rt->rt_flags & RTF_GATEWAY)
			dst = (struct sockaddr_in *)ro->ro_rt->rt_gateway;
		if (ro->ro_rt->rt_flags & RTF_HOST)
			isbroadcast = (ro->ro_rt->rt_flags & RTF_BROADCAST);
		else
			isbroadcast = in_broadcast(dst->sin_addr, ifp);
	}
	if (IN_MULTICAST(ntohl(ip->ip_dst.s_addr))) {
		struct in_multi *inm;

		m->m_flags |= M_MCAST;
		/*
		 * IP destination address is multicast.  Make sure "dst"
		 * still points to the address in "ro".  (It may have been
		 * changed to point to a gateway address, above.)
		 */
		dst = (struct sockaddr_in *)&ro->ro_dst;
		/*
		 * See if the caller provided any multicast options
		 */
		if (imo != NULL) {
			ip->ip_ttl = imo->imo_multicast_ttl;
			if (imo->imo_multicast_ifp != NULL)
				ifp = imo->imo_multicast_ifp;
			if (imo->imo_multicast_vif != -1)
				ip->ip_src.s_addr =
				    ip_mcast_src(imo->imo_multicast_vif);
		} else
			ip->ip_ttl = IP_DEFAULT_MULTICAST_TTL;
		/*
		 * Confirm that the outgoing interface supports multicast.
		 */
		if ((imo == NULL) || (imo->imo_multicast_vif == -1)) {
			if ((ifp->if_flags & IFF_MULTICAST) == 0) {
				ipstat.ips_noroute++;
				error = ENETUNREACH;
				goto bad;
			}
		}
		/*
		 * If source address not specified yet, use address
		 * of outgoing interface.
		 */
		if (ip->ip_src.s_addr == INADDR_ANY) {
			register struct in_ifaddr *ia1;

			for (ia1 = in_ifaddrhead.tqh_first; ia1;
			     ia1 = ia1->ia_link.tqe_next)
				if (ia1->ia_ifp == ifp) {
					ip->ip_src = IA_SIN(ia1)->sin_addr;
					break;
				}
		}

		IN_LOOKUP_MULTI(ip->ip_dst, ifp, inm);
		if (inm != NULL &&
		   (imo == NULL || imo->imo_multicast_loop)) {
			/*
			 * If we belong to the destination multicast group
			 * on the outgoing interface, and the caller did not
			 * forbid loopback, loop back a copy.
			 */
			ip_mloopback(ifp, m, dst, hlen);
		}
		else {
			/*
			 * If we are acting as a multicast router, perform
			 * multicast forwarding as if the packet had just
			 * arrived on the interface to which we are about
			 * to send.  The multicast forwarding function
			 * recursively calls this function, using the
			 * IP_FORWARDING flag to prevent infinite recursion.
			 *
			 * Multicasts that are looped back by ip_mloopback(),
			 * above, will be forwarded by the ip_input() routine,
			 * if necessary.
			 */
			if (ip_mrouter && (flags & IP_FORWARDING) == 0) {
				/*
				 * Check if rsvp daemon is running. If not, don't
				 * set ip_moptions. This ensures that the packet
				 * is multicast and not just sent down one link
				 * as prescribed by rsvpd.
				 */
				if (!rsvp_on)
				  imo = NULL;
				if (ip_mforward(ip, ifp, m, imo) != 0) {
					m_freem(m);
					goto done;
				}
			}
		}

		/*
		 * Multicasts with a time-to-live of zero may be looped-
		 * back, above, but must not be transmitted on a network.
		 * Also, multicasts addressed to the loopback interface
		 * are not sent -- the above call to ip_mloopback() will
		 * loop back a copy if this host actually belongs to the
		 * destination group on the loopback interface.
		 */
		if (ip->ip_ttl == 0 || ifp->if_flags & IFF_LOOPBACK) {
			m_freem(m);
			goto done;
		}

		goto sendit;
	}
#ifndef notdef
	/*
	 * If source address not specified yet, use address
	 * of outgoing interface.
	 */
	if (ip->ip_src.s_addr == INADDR_ANY) {
		ip->ip_src = IA_SIN(ia)->sin_addr;
#ifdef IPFIREWALL_FORWARD
		/* Keep note that we did this - if the firewall changes
		 * the next-hop, our interface may change, changing the
		 * default source IP. It's a shame so much effort happens
		 * twice. Oh well. 
		 */
		fwd_rewrite_src++;
#endif /* IPFIREWALL_FORWARD */
	}
#endif /* notdef */
	/*
	 * Verify that we have any chance at all of being able to queue
	 *      the packet or packet fragments
	 */
	if ((ifp->if_snd.ifq_len + ip->ip_len / ifp->if_mtu + 1) >=
		ifp->if_snd.ifq_maxlen) {
			error = ENOBUFS;
			goto bad;
	}

	/*
	 * Look for broadcast address and
	 * and verify user is allowed to send
	 * such a packet.
	 */
	if (isbroadcast) {
		if ((ifp->if_flags & IFF_BROADCAST) == 0) {
			error = EADDRNOTAVAIL;
			goto bad;
		}
		if ((flags & IP_ALLOWBROADCAST) == 0) {
			error = EACCES;
			goto bad;
		}
		/* don't allow broadcast messages to be fragmented */
		if ((u_short)ip->ip_len > ifp->if_mtu) {
			error = EMSGSIZE;
			goto bad;
		}
		m->m_flags |= M_BCAST;
	} else {
		m->m_flags &= ~M_BCAST;
	}

sendit:
	/*
	 * IpHack's section.
	 * - Xlate: translate packet's addr/port (NAT).
	 * - Firewall: deny/allow/etc.
	 * - Wrap: fake packet's addr/port <unimpl.>
	 * - Encapsulate: put it in another IP and send out. <unimp.>
	 */ 
#if defined(IPFILTER) || defined(IPFILTER_LKM)
	if (fr_checkp) {
		struct  mbuf    *m1 = m;

		if ((error = (*fr_checkp)(ip, hlen, ifp, 1, &m1)) || !m1)
			goto done;
		ip = mtod(m = m1, struct ip *);
	}
#endif

#ifdef COMPAT_IPFW
        if (ip_nat_ptr && !(*ip_nat_ptr)(&ip, &m, ifp, IP_NAT_OUT)) {
		error = EACCES; 
		goto done;
	}

	/*
	 * Check with the firewall...
	 */
	if (ip_fw_chk_ptr) {
#ifdef IPFIREWALL_FORWARD
		struct sockaddr_in *old;
		old = dst;
#endif
#ifdef IPDIVERT
		ip_divert_port = (*ip_fw_chk_ptr)(&ip,
		    hlen, ifp, &ip_divert_cookie, &m, &dst);
		if (ip_divert_port) {		/* Divert packet */
			(*inetsw[ip_protox[IPPROTO_DIVERT]].pr_input)(m, 0);
			goto done;
		}
#else	/* !IPDIVERT */
		u_int16_t 	dummy = 0;
		/* If ipfw says divert, we have to just drop packet */
		if ((*ip_fw_chk_ptr)(&ip, hlen, ifp, &dummy, &m, &dst)) {
			m_freem(m);
			goto done;
		}
#endif	/* !IPDIVERT */
		if (!m) {
			error = EACCES;
			goto done;
		}
#ifdef IPFIREWALL_FORWARD
		/* Here we check dst to make sure it's directly reachable on the
		 * interface we previously thought it was.
		 * If it isn't (which may be likely in some situations) we have
		 * to re-route it (ie, find a route for the next-hop and the
		 * associated interface) and set them here. This is nested
		 * forwarding which in most cases is undesirable, except where
		 * such control is nigh impossible. So we do it here.
		 * And I'm babbling.
		 */
		if (old != dst) {
			struct in_ifaddr *ia;

			/* It's changed... */
			/* There must be a better way to do this next line... */
			static struct route sro_fwd, *ro_fwd = &sro_fwd;
#ifdef IPFIREWALL_FORWARD_DEBUG
			printf("IPFIREWALL_FORWARD: New dst ip: ");
			print_ip(dst->sin_addr);
			printf("\n");
#endif
			/*
			 * We need to figure out if we have been forwarded
			 * to a local socket. If so then we should somehow 
			 * "loop back" to ip_input, and get directed to the
			 * PCB as if we had received this packet. This is
			 * because it may be dificult to identify the packets
			 * you want to forward until they are being output
			 * and have selected an interface. (e.g. locally
			 * initiated packets) If we used the loopback inteface,
			 * we would not be able to control what happens 
			 * as the packet runs through ip_input() as
			 * it is done through a ISR.
			 */
			for (ia = TAILQ_FIRST(&in_ifaddrhead); ia;
					ia = TAILQ_NEXT(ia, ia_link)) {
				/*
				 * If the addr to forward to is one
				 * of ours, we pretend to
				 * be the destination for this packet.
				 */
				if (IA_SIN(ia)->sin_addr.s_addr ==
						 dst->sin_addr.s_addr)
					break;
			}
			if (ia) {
				/* tell ip_input "dont filter" */
				ip_fw_fwd_addr = dst;
				if (m->m_pkthdr.rcvif == NULL)
					m->m_pkthdr.rcvif = ifunit("lo0");
				ip->ip_len = htons((u_short)ip->ip_len);
				ip->ip_off = htons((u_short)ip->ip_off);
				ip->ip_sum = 0;
				if (ip->ip_vhl == IP_VHL_BORING) {
					ip->ip_sum = in_cksum_hdr(ip);
				} else {
					ip->ip_sum = in_cksum(m, hlen);
				}
				ip_input(m);
				goto done;
			}
			/* Some of the logic for this was
			 * nicked from above.
			 *
			 * This rewrites the cached route in a local PCB.
			 * Is this what we want to do?
			 */
			bcopy(dst, &ro_fwd->ro_dst, sizeof(*dst));

			ro_fwd->ro_rt = 0;
			rtalloc_ign(ro_fwd, RTF_PRCLONING);

			if (ro_fwd->ro_rt == 0) {
				ipstat.ips_noroute++;
				error = EHOSTUNREACH;
				goto bad;
			}

			ia = ifatoia(ro_fwd->ro_rt->rt_ifa);
			ifp = ro_fwd->ro_rt->rt_ifp;
			ro_fwd->ro_rt->rt_use++;
			if (ro_fwd->ro_rt->rt_flags & RTF_GATEWAY)
				dst = (struct sockaddr_in *)ro_fwd->ro_rt->rt_gateway;
			if (ro_fwd->ro_rt->rt_flags & RTF_HOST)
				isbroadcast =
				    (ro_fwd->ro_rt->rt_flags & RTF_BROADCAST);
			else
				isbroadcast = in_broadcast(dst->sin_addr, ifp);
			RTFREE(ro->ro_rt);
			ro->ro_rt = ro_fwd->ro_rt;
			dst = (struct sockaddr_in *)&ro_fwd->ro_dst;

			/*
			 * If we added a default src ip earlier,
			 * which would have been gotten from the-then
			 * interface, do it again, from the new one.
			 */
			if (fwd_rewrite_src)
				ip->ip_src = IA_SIN(ia)->sin_addr;
		}
#endif /* IPFIREWALL_FORWARD */
	}
#endif /* COMPAT_IPFW */


	/*
	 * If small enough for interface, can just send directly.
	 */
	if ((u_short)ip->ip_len <= ifp->if_mtu) {
		ip->ip_len = htons((u_short)ip->ip_len);
		ip->ip_off = htons((u_short)ip->ip_off);
		ip->ip_sum = 0;
		if (ip->ip_vhl == IP_VHL_BORING) {
			ip->ip_sum = in_cksum_hdr(ip);
		} else {
			ip->ip_sum = in_cksum(m, hlen);
		}
		error = (*ifp->if_output)(ifp, m,
				(struct sockaddr *)dst, ro->ro_rt);
		goto done;
	}
	/*
	 * Too large for interface; fragment if possible.
	 * Must be able to put at least 8 bytes per fragment.
	 */
	if (ip->ip_off & IP_DF) {
		error = EMSGSIZE;
		/*
		 * This case can happen if the user changed the MTU
		 * of an interface after enabling IP on it.  Because
		 * most netifs don't keep track of routes pointing to
		 * them, there is no way for one to update all its
		 * routes when the MTU is changed.
		 */
		if ((ro->ro_rt->rt_flags & (RTF_UP | RTF_HOST))
		    && !(ro->ro_rt->rt_rmx.rmx_locks & RTV_MTU)
		    && (ro->ro_rt->rt_rmx.rmx_mtu > ifp->if_mtu)) {
			ro->ro_rt->rt_rmx.rmx_mtu = ifp->if_mtu;
		}
		ipstat.ips_cantfrag++;
		goto bad;
	}
	len = (ifp->if_mtu - hlen) &~ 7;
	if (len < 8) {
		error = EMSGSIZE;
		goto bad;
	}

    {
	int mhlen, firstlen = len;
	struct mbuf **mnext = &m->m_nextpkt;

	/*
	 * Loop through length of segment after first fragment,
	 * make new header and copy data of each part and link onto chain.
	 */
	m0 = m;
	mhlen = sizeof (struct ip);
	for (off = hlen + len; off < (u_short)ip->ip_len; off += len) {
		MGETHDR(m, M_DONTWAIT, MT_HEADER);
		if (m == 0) {
			error = ENOBUFS;
			ipstat.ips_odropped++;
			goto sendorfree;
		}
		m->m_data += max_linkhdr;
		mhip = mtod(m, struct ip *);
		*mhip = *ip;
		if (hlen > sizeof (struct ip)) {
			mhlen = ip_optcopy(ip, mhip) + sizeof (struct ip);
			mhip->ip_vhl = IP_MAKE_VHL(IPVERSION, mhlen >> 2);
		}
		m->m_len = mhlen;
		mhip->ip_off = ((off - hlen) >> 3) + (ip->ip_off & ~IP_MF);
		if (ip->ip_off & IP_MF)
			mhip->ip_off |= IP_MF;
		if (off + len >= (u_short)ip->ip_len)
			len = (u_short)ip->ip_len - off;
		else
			mhip->ip_off |= IP_MF;
		mhip->ip_len = htons((u_short)(len + mhlen));
		m->m_next = m_copy(m0, off, len);
		if (m->m_next == 0) {
			(void) m_free(m);
			error = ENOBUFS;	/* ??? */
			ipstat.ips_odropped++;
			goto sendorfree;
		}
		m->m_pkthdr.len = mhlen + len;
		m->m_pkthdr.rcvif = (struct ifnet *)0;
		mhip->ip_off = htons((u_short)mhip->ip_off);
		mhip->ip_sum = 0;
		if (mhip->ip_vhl == IP_VHL_BORING) {
			mhip->ip_sum = in_cksum_hdr(mhip);
		} else {
			mhip->ip_sum = in_cksum(m, mhlen);
		}
		*mnext = m;
		mnext = &m->m_nextpkt;
		ipstat.ips_ofragments++;
	}
	/*
	 * Update first fragment by trimming what's been copied out
	 * and updating header, then send each fragment (in order).
	 */
	m = m0;
	m_adj(m, hlen + firstlen - (u_short)ip->ip_len);
	m->m_pkthdr.len = hlen + firstlen;
	ip->ip_len = htons((u_short)m->m_pkthdr.len);
	ip->ip_off = htons((u_short)(ip->ip_off | IP_MF));
	ip->ip_sum = 0;
	if (ip->ip_vhl == IP_VHL_BORING) {
		ip->ip_sum = in_cksum_hdr(ip);
	} else {
		ip->ip_sum = in_cksum(m, hlen);
	}
sendorfree:
	for (m = m0; m; m = m0) {
		m0 = m->m_nextpkt;
		m->m_nextpkt = 0;
		if (error == 0)
			error = (*ifp->if_output)(ifp, m,
			    (struct sockaddr *)dst, ro->ro_rt);
		else
			m_freem(m);
	}

	if (error == 0)
		ipstat.ips_fragmented++;
    }
done:
	return (error);
bad:
	m_freem(m0);
	goto done;
}

/*
 * Insert IP options into preformed packet.
 * Adjust IP destination as required for IP source routing,
 * as indicated by a non-zero in_addr at the start of the options.
 *
 * XXX This routine assumes that the packet has no options in place.
 */
static struct mbuf *
ip_insertoptions(m, opt, phlen)
	register struct mbuf *m;
	struct mbuf *opt;
	int *phlen;
{
	register struct ipoption *p = mtod(opt, struct ipoption *);
	struct mbuf *n;
	register struct ip *ip = mtod(m, struct ip *);
	unsigned optlen;

	optlen = opt->m_len - sizeof(p->ipopt_dst);
	if (optlen + (u_short)ip->ip_len > IP_MAXPACKET)
		return (m);		/* XXX should fail */
	if (p->ipopt_dst.s_addr)
		ip->ip_dst = p->ipopt_dst;
	if (m->m_flags & M_EXT || m->m_data - optlen < m->m_pktdat) {
		MGETHDR(n, M_DONTWAIT, MT_HEADER);
		if (n == 0)
			return (m);
		n->m_pkthdr.len = m->m_pkthdr.len + optlen;
		m->m_len -= sizeof(struct ip);
		m->m_data += sizeof(struct ip);
		n->m_next = m;
		m = n;
		m->m_len = optlen + sizeof(struct ip);
		m->m_data += max_linkhdr;
		(void)memcpy(mtod(m, void *), ip, sizeof(struct ip));
	} else {
		m->m_data -= optlen;
		m->m_len += optlen;
		m->m_pkthdr.len += optlen;
		ovbcopy((caddr_t)ip, mtod(m, caddr_t), sizeof(struct ip));
	}
	ip = mtod(m, struct ip *);
	bcopy(p->ipopt_list, ip + 1, optlen);
	*phlen = sizeof(struct ip) + optlen;
	ip->ip_vhl = IP_MAKE_VHL(IPVERSION, *phlen >> 2);
	ip->ip_len += optlen;
	return (m);
}

/*
 * Copy options from ip to jp,
 * omitting those not copied during fragmentation.
 */
#if !defined(IPFILTER) && !defined(IPFILTER_LKM)
static
#endif
int
ip_optcopy(ip, jp)
	struct ip *ip, *jp;
{
	register u_char *cp, *dp;
	int opt, optlen, cnt;

	cp = (u_char *)(ip + 1);
	dp = (u_char *)(jp + 1);
	cnt = (IP_VHL_HL(ip->ip_vhl) << 2) - sizeof (struct ip);
	for (; cnt > 0; cnt -= optlen, cp += optlen) {
		opt = cp[0];
		if (opt == IPOPT_EOL)
			break;
		if (opt == IPOPT_NOP) {
			/* Preserve for IP mcast tunnel's LSRR alignment. */
			*dp++ = IPOPT_NOP;
			optlen = 1;
			continue;
		} else
			optlen = cp[IPOPT_OLEN];
		/* bogus lengths should have been caught by ip_dooptions */
		if (optlen > cnt)
			optlen = cnt;
		if (IPOPT_COPIED(opt)) {
			bcopy(cp, dp, optlen);
			dp += optlen;
		}
	}
	for (optlen = dp - (u_char *)(jp+1); optlen & 0x3; optlen++)
		*dp++ = IPOPT_EOL;
	return (optlen);
}

/*
 * IP socket option processing.
 */
int
ip_ctloutput(op, so, level, optname, mp, p)
	int op;
	struct socket *so;
	int level, optname;
	struct mbuf **mp;
	struct proc *p;
{
	register struct inpcb *inp = sotoinpcb(so);
	register struct mbuf *m = *mp;
	register int optval = 0;
	int error = 0;

	if (level != IPPROTO_IP) {
		error = EINVAL;
		if (op == PRCO_SETOPT && *mp)
			(void) m_free(*mp);
	} else switch (op) {

	case PRCO_SETOPT:
		switch (optname) {
		case IP_OPTIONS:
#ifdef notyet
		case IP_RETOPTS:
			return (ip_pcbopts(optname, &inp->inp_options, m));
#else
			return (ip_pcbopts(&inp->inp_options, m));
#endif

		case IP_TOS:
		case IP_TTL:
		case IP_RECVOPTS:
		case IP_RECVRETOPTS:
		case IP_RECVDSTADDR:
		case IP_RECVIF:
			if (m == 0 || m->m_len != sizeof(int))
				error = EINVAL;
			else {
				optval = *mtod(m, int *);
				switch (optname) {

				case IP_TOS:
					inp->inp_ip_tos = optval;
					break;

				case IP_TTL:
					inp->inp_ip_ttl = optval;
					break;
#define	OPTSET(bit) \
	if (optval) \
		inp->inp_flags |= bit; \
	else \
		inp->inp_flags &= ~bit;

				case IP_RECVOPTS:
					OPTSET(INP_RECVOPTS);
					break;

				case IP_RECVRETOPTS:
					OPTSET(INP_RECVRETOPTS);
					break;

				case IP_RECVDSTADDR:
					OPTSET(INP_RECVDSTADDR);
					break;

				case IP_RECVIF:
					OPTSET(INP_RECVIF);
					break;
				}
			}
			break;
#undef OPTSET

		case IP_MULTICAST_IF:
		case IP_MULTICAST_VIF:
		case IP_MULTICAST_TTL:
		case IP_MULTICAST_LOOP:
		case IP_ADD_MEMBERSHIP:
		case IP_DROP_MEMBERSHIP:
			error = ip_setmoptions(optname, &inp->inp_moptions, m);
			break;

		case IP_PORTRANGE:
			if (m == 0 || m->m_len != sizeof(int))
				error = EINVAL;
			else {
				optval = *mtod(m, int *);

				switch (optval) {

				case IP_PORTRANGE_DEFAULT:
					inp->inp_flags &= ~(INP_LOWPORT);
					inp->inp_flags &= ~(INP_HIGHPORT);
					break;

				case IP_PORTRANGE_HIGH:
					inp->inp_flags &= ~(INP_LOWPORT);
					inp->inp_flags |= INP_HIGHPORT;
					break;

				case IP_PORTRANGE_LOW:
					inp->inp_flags &= ~(INP_HIGHPORT);
					inp->inp_flags |= INP_LOWPORT;
					break;

				default:
					error = EINVAL;
					break;
				}
			}
			break;

		default:
			error = ENOPROTOOPT;
			break;
		}
		if (m)
			(void)m_free(m);
		break;

	case PRCO_GETOPT:
		switch (optname) {
		case IP_OPTIONS:
		case IP_RETOPTS:
			*mp = m = m_get(M_WAIT, MT_SOOPTS);
			if (inp->inp_options) {
				m->m_len = inp->inp_options->m_len;
				bcopy(mtod(inp->inp_options, void *),
				    mtod(m, void *), m->m_len);
			} else
				m->m_len = 0;
			break;

		case IP_TOS:
		case IP_TTL:
		case IP_RECVOPTS:
		case IP_RECVRETOPTS:
		case IP_RECVDSTADDR:
		case IP_RECVIF:
			*mp = m = m_get(M_WAIT, MT_SOOPTS);
			m->m_len = sizeof(int);
			switch (optname) {

			case IP_TOS:
				optval = inp->inp_ip_tos;
				break;

			case IP_TTL:
				optval = inp->inp_ip_ttl;
				break;

#define	OPTBIT(bit)	(inp->inp_flags & bit ? 1 : 0)

			case IP_RECVOPTS:
				optval = OPTBIT(INP_RECVOPTS);
				break;

			case IP_RECVRETOPTS:
				optval = OPTBIT(INP_RECVRETOPTS);
				break;

			case IP_RECVDSTADDR:
				optval = OPTBIT(INP_RECVDSTADDR);
				break;

			case IP_RECVIF:
				optval = OPTBIT(INP_RECVIF);
				break;
			}
			*mtod(m, int *) = optval;
			break;

		case IP_MULTICAST_IF:
		case IP_MULTICAST_VIF:
		case IP_MULTICAST_TTL:
		case IP_MULTICAST_LOOP:
		case IP_ADD_MEMBERSHIP:
		case IP_DROP_MEMBERSHIP:
			error = ip_getmoptions(optname, inp->inp_moptions, mp);
			break;

		case IP_PORTRANGE:
			*mp = m = m_get(M_WAIT, MT_SOOPTS);
			m->m_len = sizeof(int);

			if (inp->inp_flags & INP_HIGHPORT)
				optval = IP_PORTRANGE_HIGH;
			else if (inp->inp_flags & INP_LOWPORT)
				optval = IP_PORTRANGE_LOW;
			else
				optval = 0;

			*mtod(m, int *) = optval;
			break;

		default:
			error = ENOPROTOOPT;
			break;
		}
		break;
	}
	return (error);
}

/*
 * Set up IP options in pcb for insertion in output packets.
 * Store in mbuf with pointer in pcbopt, adding pseudo-option
 * with destination address if source routed.
 */
static int
#ifdef notyet
ip_pcbopts(optname, pcbopt, m)
	int optname;
#else
ip_pcbopts(pcbopt, m)
#endif
	struct mbuf **pcbopt;
	register struct mbuf *m;
{
	register int cnt, optlen;
	register u_char *cp;
	u_char opt;

	/* turn off any old options */
	if (*pcbopt)
		(void)m_free(*pcbopt);
	*pcbopt = 0;
	if (m == (struct mbuf *)0 || m->m_len == 0) {
		/*
		 * Only turning off any previous options.
		 */
		if (m)
			(void)m_free(m);
		return (0);
	}

#ifndef	vax
	if (m->m_len % sizeof(long))
		goto bad;
#endif
	/*
	 * IP first-hop destination address will be stored before
	 * actual options; move other options back
	 * and clear it when none present.
	 */
	if (m->m_data + m->m_len + sizeof(struct in_addr) >= &m->m_dat[MLEN])
		goto bad;
	cnt = m->m_len;
	m->m_len += sizeof(struct in_addr);
	cp = mtod(m, u_char *) + sizeof(struct in_addr);
	ovbcopy(mtod(m, caddr_t), (caddr_t)cp, (unsigned)cnt);
	bzero(mtod(m, caddr_t), sizeof(struct in_addr));

	for (; cnt > 0; cnt -= optlen, cp += optlen) {
		opt = cp[IPOPT_OPTVAL];
		if (opt == IPOPT_EOL)
			break;
		if (opt == IPOPT_NOP)
			optlen = 1;
		else {
			optlen = cp[IPOPT_OLEN];
			if (optlen <= IPOPT_OLEN || optlen > cnt)
				goto bad;
		}
		switch (opt) {

		default:
			break;

		case IPOPT_LSRR:
		case IPOPT_SSRR:
			/*
			 * user process specifies route as:
			 *	->A->B->C->D
			 * D must be our final destination (but we can't
			 * check that since we may not have connected yet).
			 * A is first hop destination, which doesn't appear in
			 * actual IP option, but is stored before the options.
			 */
			if (optlen < IPOPT_MINOFF - 1 + sizeof(struct in_addr))
				goto bad;
			m->m_len -= sizeof(struct in_addr);
			cnt -= sizeof(struct in_addr);
			optlen -= sizeof(struct in_addr);
			cp[IPOPT_OLEN] = optlen;
			/*
			 * Move first hop before start of options.
			 */
			bcopy((caddr_t)&cp[IPOPT_OFFSET+1], mtod(m, caddr_t),
			    sizeof(struct in_addr));
			/*
			 * Then copy rest of options back
			 * to close up the deleted entry.
			 */
			ovbcopy((caddr_t)(&cp[IPOPT_OFFSET+1] +
			    sizeof(struct in_addr)),
			    (caddr_t)&cp[IPOPT_OFFSET+1],
			    (unsigned)cnt + sizeof(struct in_addr));
			break;
		}
	}
	if (m->m_len > MAX_IPOPTLEN + sizeof(struct in_addr))
		goto bad;
	*pcbopt = m;
	return (0);

bad:
	(void)m_free(m);
	return (EINVAL);
}

/*
 * Set the IP multicast options in response to user setsockopt().
 */
static int
ip_setmoptions(optname, imop, m)
	int optname;
	struct ip_moptions **imop;
	struct mbuf *m;
{
	register int error = 0;
	u_char loop;
	register int i;
	struct in_addr addr;
	register struct ip_mreq *mreq;
	register struct ifnet *ifp;
	register struct ip_moptions *imo = *imop;
	struct route ro;
	register struct sockaddr_in *dst;
	int s;

	if (imo == NULL) {
		/*
		 * No multicast option buffer attached to the pcb;
		 * allocate one and initialize to default values.
		 */
		imo = (struct ip_moptions*)malloc(sizeof(*imo), M_IPMOPTS,
		    M_WAITOK);

		if (imo == NULL)
			return (ENOBUFS);
		*imop = imo;
		imo->imo_multicast_ifp = NULL;
		imo->imo_multicast_vif = -1;
		imo->imo_multicast_ttl = IP_DEFAULT_MULTICAST_TTL;
		imo->imo_multicast_loop = IP_DEFAULT_MULTICAST_LOOP;
		imo->imo_num_memberships = 0;
	}

	switch (optname) {
	/* store an index number for the vif you wanna use in the send */
	case IP_MULTICAST_VIF:
		if (!legal_vif_num) {
			error = EOPNOTSUPP;
			break;
		}
		if (m == NULL || m->m_len != sizeof(int)) {
			error = EINVAL;
			break;
		}
		i = *(mtod(m, int *));
		if (!legal_vif_num(i) && (i != -1)) {
			error = EINVAL;
			break;
		}
		imo->imo_multicast_vif = i;
		break;

	case IP_MULTICAST_IF:
		/*
		 * Select the interface for outgoing multicast packets.
		 */
		if (m == NULL || m->m_len != sizeof(struct in_addr)) {
			error = EINVAL;
			break;
		}
		addr = *(mtod(m, struct in_addr *));
		/*
		 * INADDR_ANY is used to remove a previous selection.
		 * When no interface is selected, a default one is
		 * chosen every time a multicast packet is sent.
		 */
		if (addr.s_addr == INADDR_ANY) {
			imo->imo_multicast_ifp = NULL;
			break;
		}
		/*
		 * The selected interface is identified by its local
		 * IP address.  Find the interface and confirm that
		 * it supports multicasting.
		 */
		s = splimp();
		INADDR_TO_IFP(addr, ifp);
		if (ifp == NULL || (ifp->if_flags & IFF_MULTICAST) == 0) {
			splx(s);
			error = EADDRNOTAVAIL;
			break;
		}
		imo->imo_multicast_ifp = ifp;
		splx(s);
		break;

	case IP_MULTICAST_TTL:
		/*
		 * Set the IP time-to-live for outgoing multicast packets.
		 */
		if (m == NULL || m->m_len != 1) {
			error = EINVAL;
			break;
		}
		imo->imo_multicast_ttl = *(mtod(m, u_char *));
		break;

	case IP_MULTICAST_LOOP:
		/*
		 * Set the loopback flag for outgoing multicast packets.
		 * Must be zero or one.
		 */
		if (m == NULL || m->m_len != 1 ||
		   (loop = *(mtod(m, u_char *))) > 1) {
			error = EINVAL;
			break;
		}
		imo->imo_multicast_loop = loop;
		break;

	case IP_ADD_MEMBERSHIP:
		/*
		 * Add a multicast group membership.
		 * Group must be a valid IP multicast address.
		 */
		if (m == NULL || m->m_len != sizeof(struct ip_mreq)) {
			error = EINVAL;
			break;
		}
		mreq = mtod(m, struct ip_mreq *);
		if (!IN_MULTICAST(ntohl(mreq->imr_multiaddr.s_addr))) {
			error = EINVAL;
			break;
		}
		s = splimp();
		/*
		 * If no interface address was provided, use the interface of
		 * the route to the given multicast address.
		 */
		if (mreq->imr_interface.s_addr == INADDR_ANY) {
			bzero((caddr_t)&ro, sizeof(ro));
			dst = (struct sockaddr_in *)&ro.ro_dst;
			dst->sin_len = sizeof(*dst);
			dst->sin_family = AF_INET;
			dst->sin_addr = mreq->imr_multiaddr;
			rtalloc(&ro);
			if (ro.ro_rt == NULL) {
				error = EADDRNOTAVAIL;
				splx(s);
				break;
			}
			ifp = ro.ro_rt->rt_ifp;
			rtfree(ro.ro_rt);
		}
		else {
			INADDR_TO_IFP(mreq->imr_interface, ifp);
		}

		/*
		 * See if we found an interface, and confirm that it
		 * supports multicast.
		 */
		if (ifp == NULL || (ifp->if_flags & IFF_MULTICAST) == 0) {
			error = EADDRNOTAVAIL;
			splx(s);
			break;
		}
		/*
		 * See if the membership already exists or if all the
		 * membership slots are full.
		 */
		for (i = 0; i < imo->imo_num_memberships; ++i) {
			if (imo->imo_membership[i]->inm_ifp == ifp &&
			    imo->imo_membership[i]->inm_addr.s_addr
						== mreq->imr_multiaddr.s_addr)
				break;
		}
		if (i < imo->imo_num_memberships) {
			error = EADDRINUSE;
			splx(s);
			break;
		}
		if (i == IP_MAX_MEMBERSHIPS) {
			error = ETOOMANYREFS;
			splx(s);
			break;
		}
		/*
		 * Everything looks good; add a new record to the multicast
		 * address list for the given interface.
		 */
		if ((imo->imo_membership[i] =
		    in_addmulti(&mreq->imr_multiaddr, ifp)) == NULL) {
			error = ENOBUFS;
			splx(s);
			break;
		}
		++imo->imo_num_memberships;
		splx(s);
		break;

	case IP_DROP_MEMBERSHIP:
		/*
		 * Drop a multicast group membership.
		 * Group must be a valid IP multicast address.
		 */
		if (m == NULL || m->m_len != sizeof(struct ip_mreq)) {
			error = EINVAL;
			break;
		}
		mreq = mtod(m, struct ip_mreq *);
		if (!IN_MULTICAST(ntohl(mreq->imr_multiaddr.s_addr))) {
			error = EINVAL;
			break;
		}

		s = splimp();
		/*
		 * If an interface address was specified, get a pointer
		 * to its ifnet structure.
		 */
		if (mreq->imr_interface.s_addr == INADDR_ANY)
			ifp = NULL;
		else {
			INADDR_TO_IFP(mreq->imr_interface, ifp);
			if (ifp == NULL) {
				error = EADDRNOTAVAIL;
				splx(s);
				break;
			}
		}
		/*
		 * Find the membership in the membership array.
		 */
		for (i = 0; i < imo->imo_num_memberships; ++i) {
			if ((ifp == NULL ||
			     imo->imo_membership[i]->inm_ifp == ifp) &&
			     imo->imo_membership[i]->inm_addr.s_addr ==
			     mreq->imr_multiaddr.s_addr)
				break;
		}
		if (i == imo->imo_num_memberships) {
			error = EADDRNOTAVAIL;
			splx(s);
			break;
		}
		/*
		 * Give up the multicast address record to which the
		 * membership points.
		 */
		in_delmulti(imo->imo_membership[i]);
		/*
		 * Remove the gap in the membership array.
		 */
		for (++i; i < imo->imo_num_memberships; ++i)
			imo->imo_membership[i-1] = imo->imo_membership[i];
		--imo->imo_num_memberships;
		splx(s);
		break;

	default:
		error = EOPNOTSUPP;
		break;
	}

	/*
	 * If all options have default values, no need to keep the mbuf.
	 */
	if (imo->imo_multicast_ifp == NULL &&
	    imo->imo_multicast_vif == -1 &&
	    imo->imo_multicast_ttl == IP_DEFAULT_MULTICAST_TTL &&
	    imo->imo_multicast_loop == IP_DEFAULT_MULTICAST_LOOP &&
	    imo->imo_num_memberships == 0) {
		free(*imop, M_IPMOPTS);
		*imop = NULL;
	}

	return (error);
}

/*
 * Return the IP multicast options in response to user getsockopt().
 */
static int
ip_getmoptions(optname, imo, mp)
	int optname;
	register struct ip_moptions *imo;
	register struct mbuf **mp;
{
	u_char *ttl;
	u_char *loop;
	struct in_addr *addr;
	struct in_ifaddr *ia;

	*mp = m_get(M_WAIT, MT_SOOPTS);

	switch (optname) {

	case IP_MULTICAST_VIF: 
		if (imo != NULL)
			*(mtod(*mp, int *)) = imo->imo_multicast_vif;
		else
			*(mtod(*mp, int *)) = -1;
		(*mp)->m_len = sizeof(int);
		return(0);

	case IP_MULTICAST_IF:
		addr = mtod(*mp, struct in_addr *);
		(*mp)->m_len = sizeof(struct in_addr);
		if (imo == NULL || imo->imo_multicast_ifp == NULL)
			addr->s_addr = INADDR_ANY;
		else {
			IFP_TO_IA(imo->imo_multicast_ifp, ia);
			addr->s_addr = (ia == NULL) ? INADDR_ANY
					: IA_SIN(ia)->sin_addr.s_addr;
		}
		return (0);

	case IP_MULTICAST_TTL:
		ttl = mtod(*mp, u_char *);
		(*mp)->m_len = 1;
		*ttl = (imo == NULL) ? IP_DEFAULT_MULTICAST_TTL
				     : imo->imo_multicast_ttl;
		return (0);

	case IP_MULTICAST_LOOP:
		loop = mtod(*mp, u_char *);
		(*mp)->m_len = 1;
		*loop = (imo == NULL) ? IP_DEFAULT_MULTICAST_LOOP
				      : imo->imo_multicast_loop;
		return (0);

	default:
		return (EOPNOTSUPP);
	}
}

/*
 * Discard the IP multicast options.
 */
void
ip_freemoptions(imo)
	register struct ip_moptions *imo;
{
	register int i;

	if (imo != NULL) {
		for (i = 0; i < imo->imo_num_memberships; ++i)
			in_delmulti(imo->imo_membership[i]);
		free(imo, M_IPMOPTS);
	}
}

/*
 * Routine called from ip_output() to loop back a copy of an IP multicast
 * packet to the input queue of a specified interface.  Note that this
 * calls the output routine of the loopback "driver", but with an interface
 * pointer that might NOT be a loopback interface -- evil, but easier than
 * replicating that code here.
 */
static void
ip_mloopback(ifp, m, dst, hlen)
	struct ifnet *ifp;
	register struct mbuf *m;
	register struct sockaddr_in *dst;
	int hlen;
{
	register struct ip *ip;
	struct mbuf *copym;

	copym = m_copy(m, 0, M_COPYALL);
	if (copym != NULL && (copym->m_flags & M_EXT || copym->m_len < hlen))
		copym = m_pullup(copym, hlen);
	if (copym != NULL) {
		/*
		 * We don't bother to fragment if the IP length is greater
		 * than the interface's MTU.  Can this possibly matter?
		 */
		ip = mtod(copym, struct ip *);
		ip->ip_len = htons((u_short)ip->ip_len);
		ip->ip_off = htons((u_short)ip->ip_off);
		ip->ip_sum = 0;
		if (ip->ip_vhl == IP_VHL_BORING) {
			ip->ip_sum = in_cksum_hdr(ip);
		} else {
			ip->ip_sum = in_cksum(copym, hlen);
		}
		/*
		 * NB:
		 * It's not clear whether there are any lingering
		 * reentrancy problems in other areas which might
		 * be exposed by using ip_input directly (in
		 * particular, everything which modifies the packet
		 * in-place).  Yet another option is using the
		 * protosw directly to deliver the looped back
		 * packet.  For the moment, we'll err on the side
		 * of safety by using if_simloop().
		 */
#if 1 /* XXX */
		if (dst->sin_family != AF_INET) {
			printf("ip_mloopback: bad address family %d\n",
						dst->sin_family);
			dst->sin_family = AF_INET;
		}
#endif

#ifdef notdef
		copym->m_pkthdr.rcvif = ifp;
		ip_input(copym);
#else
		if_simloop(ifp, copym, (struct sockaddr *)dst, 0);
#endif
	}
}