[apple/xnu.git] / bsd / netinet / ip_encap.c

/*
 * Copyright (c) 2000-2016 Apple Inc. All rights reserved.
 *
 * @APPLE_OSREFERENCE_LICENSE_HEADER_START@
 *
 * This file contains Original Code and/or Modifications of Original Code
 * as defined in and that are subject to the Apple Public Source License
 * Version 2.0 (the 'License'). You may not use this file except in
 * 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_OSREFERENCE_LICENSE_HEADER_END@
 */
/*	$FreeBSD: src/sys/netinet/ip_encap.c,v 1.1.2.2 2001/07/03 11:01:46 ume Exp $	*/
/*	$KAME: ip_encap.c,v 1.41 2001/03/15 08:35:08 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.
 */
/*
 * My grandfather said that there's a devil inside tunnelling technology...
 *
 * We have surprisingly many protocols that want packets with IP protocol
 * #4 or #41.  Here's a list of protocols that want protocol #41:
 *	RFC1933 configured tunnel
 *	RFC1933 automatic tunnel
 *	RFC2401 IPsec tunnel
 *	RFC2473 IPv6 generic packet tunnelling
 *	RFC2529 6over4 tunnel
 *	mobile-ip6 (uses RFC2473)
 *	6to4 tunnel
 * Here's a list of protocol that want protocol #4:
 *	RFC1853 IPv4-in-IPv4 tunnelling
 *	RFC2003 IPv4 encapsulation within IPv4
 *	RFC2344 reverse tunnelling for mobile-ip4
 *	RFC2401 IPsec tunnel
 * Well, what can I say.  They impose different en/decapsulation mechanism
 * from each other, so they need separate protocol handler.  The only one
 * we can easily determine by protocol # is IPsec, which always has
 * AH/ESP header right after outer IP header.
 *
 * So, clearly good old protosw does not work for protocol #4 and #41.
 * The code will let you match protocol via src/dst address pair.
 */
/* XXX is M_NETADDR correct? */

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/socket.h>
#include <sys/sockio.h>
#include <sys/mbuf.h>
#include <sys/mcache.h>
#include <sys/errno.h>
#include <sys/domain.h>
#include <sys/protosw.h>
#include <sys/queue.h>

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

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

#if INET6
#include <netinet/ip6.h>
#include <netinet6/ip6_var.h>
#include <netinet6/ip6protosw.h>
#endif


#include <net/net_osdep.h>

#ifndef __APPLE__
#include <sys/kernel.h>
#include <sys/malloc.h>
MALLOC_DEFINE(M_NETADDR, "Export Host", "Export host address structure");
#endif

static void encap_init(struct protosw *, struct domain *);
static void encap_add_locked(struct encaptab *);
static int mask_match(const struct encaptab *, const struct sockaddr *,
    const struct sockaddr *);
static void encap_fillarg(struct mbuf *, void *arg);

#ifndef LIST_HEAD_INITIALIZER
/* rely upon BSS initialization */
LIST_HEAD(, encaptab) encaptab;
#else
LIST_HEAD(, encaptab) encaptab = LIST_HEAD_INITIALIZER(&encaptab);
#endif

decl_lck_rw_data(static, encaptab_lock);

static void
encap_init(struct protosw *pp, struct domain *dp)
{
#pragma unused(dp)
	static int encap_initialized = 0;
	lck_grp_attr_t *encaptab_grp_attrib = NULL;
	lck_attr_t *encaptab_lck_attrib = NULL;
	lck_grp_t *encaptab_lck_group = NULL;

	VERIFY((pp->pr_flags & (PR_INITIALIZED | PR_ATTACHED)) == PR_ATTACHED);

	/* This gets called by more than one protocols, so initialize once */
	if (encap_initialized) {
		return;
	}

	encaptab_grp_attrib = lck_grp_attr_alloc_init();
	encaptab_lck_group = lck_grp_alloc_init("encaptab lock", encaptab_grp_attrib);
	lck_grp_attr_free(encaptab_grp_attrib);

	encaptab_lck_attrib = lck_attr_alloc_init();
	lck_rw_init(&encaptab_lock, encaptab_lck_group, encaptab_lck_attrib);

	lck_grp_free(encaptab_lck_group);
	lck_attr_free(encaptab_lck_attrib);

	encap_initialized = 1;
#if 0
	/*
	 * we cannot use LIST_INIT() here, since drivers may want to call
	 * encap_attach(), on driver attach.  encap_init() will be called
	 * on AF_INET{,6} initialization, which happens after driver
	 * initialization - using LIST_INIT() here can nuke encap_attach()
	 * from drivers.
	 */
	LIST_INIT(&encaptab);
#endif
}

void
encap4_init(struct protosw *pp, struct domain *dp)
{
	encap_init(pp, dp);
}

void
encap6_init(struct ip6protosw *pp, struct domain *dp)
{
	encap_init((struct protosw *)pp, dp);
}

#if INET
void
encap4_input(struct mbuf *m, int off)
{
	int proto;
	struct ip *ip;
	struct sockaddr_in s, d;
	const struct protosw *psw;
	struct encaptab *ep, *match;
	int prio, matchprio;
	void *match_arg = NULL;

#ifndef __APPLE__
	va_start(ap, m);
	off = va_arg(ap, int);
	proto = va_arg(ap, int);
	va_end(ap);
#endif

	/* Expect 32-bit aligned data pointer on strict-align platforms */
	MBUF_STRICT_DATA_ALIGNMENT_CHECK_32(m);

	ip = mtod(m, struct ip *);
#ifdef __APPLE__
	proto = ip->ip_p;
#endif

	bzero(&s, sizeof(s));
	s.sin_family = AF_INET;
	s.sin_len = sizeof(struct sockaddr_in);
	s.sin_addr = ip->ip_src;
	bzero(&d, sizeof(d));
	d.sin_family = AF_INET;
	d.sin_len = sizeof(struct sockaddr_in);
	d.sin_addr = ip->ip_dst;

	match = NULL;
	matchprio = 0;

	lck_rw_lock_shared(&encaptab_lock);
	for (ep = LIST_FIRST(&encaptab); ep; ep = LIST_NEXT(ep, chain)) {
		if (ep->af != AF_INET) {
			continue;
		}
		if (ep->proto >= 0 && ep->proto != proto) {
			continue;
		}
		if (ep->func) {
			prio = (*ep->func)(m, off, proto, ep->arg);
		} else {
			/*
			 * it's inbound traffic, we need to match in reverse
			 * order
			 */
			prio = mask_match(ep, (struct sockaddr *)&d,
			    (struct sockaddr *)&s);
		}

		/*
		 * We prioritize the matches by using bit length of the
		 * matches.  mask_match() and user-supplied matching function
		 * should return the bit length of the matches (for example,
		 * if both src/dst are matched for IPv4, 64 should be returned).
		 * 0 or negative return value means "it did not match".
		 *
		 * The question is, since we have two "mask" portion, we
		 * cannot really define total order between entries.
		 * For example, which of these should be preferred?
		 * mask_match() returns 48 (32 + 16) for both of them.
		 *	src=3ffe::/16, dst=3ffe:501::/32
		 *	src=3ffe:501::/32, dst=3ffe::/16
		 *
		 * We need to loop through all the possible candidates
		 * to get the best match - the search takes O(n) for
		 * n attachments (i.e. interfaces).
		 */
		if (prio <= 0) {
			continue;
		}
		if (prio > matchprio) {
			matchprio = prio;
			match = ep;
			psw = (const struct protosw *)match->psw;
			match_arg = ep->arg;
		}
	}
	lck_rw_unlock_shared(&encaptab_lock);

	if (match) {
		/* found a match, "match" has the best one */
		if (psw && psw->pr_input) {
			encap_fillarg(m, match_arg);
			(*psw->pr_input)(m, off);
		} else {
			m_freem(m);
		}
		return;
	}

	/* last resort: inject to raw socket */
	rip_input(m, off);
}
#endif

#if INET6
int
encap6_input(struct mbuf **mp, int *offp, int proto)
{
	struct mbuf *m = *mp;
	struct ip6_hdr *ip6;
	struct sockaddr_in6 s, d;
	const struct ip6protosw *psw;
	struct encaptab *ep, *match;
	int prio, matchprio;
	void *match_arg = NULL;

	/* Expect 32-bit aligned data pointer on strict-align platforms */
	MBUF_STRICT_DATA_ALIGNMENT_CHECK_32(m);

	ip6 = mtod(m, struct ip6_hdr *);
	bzero(&s, sizeof(s));
	s.sin6_family = AF_INET6;
	s.sin6_len = sizeof(struct sockaddr_in6);
	s.sin6_addr = ip6->ip6_src;
	bzero(&d, sizeof(d));
	d.sin6_family = AF_INET6;
	d.sin6_len = sizeof(struct sockaddr_in6);
	d.sin6_addr = ip6->ip6_dst;

	match = NULL;
	matchprio = 0;

	lck_rw_lock_shared(&encaptab_lock);
	for (ep = LIST_FIRST(&encaptab); ep; ep = LIST_NEXT(ep, chain)) {
		if (ep->af != AF_INET6) {
			continue;
		}
		if (ep->proto >= 0 && ep->proto != proto) {
			continue;
		}
		if (ep->func) {
			prio = (*ep->func)(m, *offp, proto, ep->arg);
		} else {
			/*
			 * it's inbound traffic, we need to match in reverse
			 * order
			 */
			prio = mask_match(ep, (struct sockaddr *)&d,
			    (struct sockaddr *)&s);
		}

		/* see encap4_input() for issues here */
		if (prio <= 0) {
			continue;
		}
		if (prio > matchprio) {
			matchprio = prio;
			match = ep;
			psw = (const struct ip6protosw *)match->psw;
			match_arg = ep->arg;
		}
	}
	lck_rw_unlock_shared(&encaptab_lock);

	if (match) {
		/* found a match */
		if (psw && psw->pr_input) {
			encap_fillarg(m, match_arg);
			return (*psw->pr_input)(mp, offp, proto);
		} else {
			m_freem(m);
			return IPPROTO_DONE;
		}
	}

	/* last resort: inject to raw socket */
	return rip6_input(mp, offp, proto);
}
#endif

static void
encap_add_locked(struct encaptab *ep)
{
	LCK_RW_ASSERT(&encaptab_lock, LCK_RW_ASSERT_EXCLUSIVE);
	LIST_INSERT_HEAD(&encaptab, ep, chain);
}

/*
 * sp (src ptr) is always my side, and dp (dst ptr) is always remote side.
 * length of mask (sm and dm) is assumed to be same as sp/dp.
 * Return value will be necessary as input (cookie) for encap_detach().
 */
const struct encaptab *
encap_attach(int af, int proto, const struct sockaddr *sp,
    const struct sockaddr *sm, const struct sockaddr *dp,
    const struct sockaddr *dm, const struct protosw *psw, void *arg)
{
	struct encaptab *ep = NULL;
	struct encaptab *new_ep = NULL;
	int error;

	/* sanity check on args */
	if (sp->sa_len > sizeof(new_ep->src) || dp->sa_len > sizeof(new_ep->dst)) {
		error = EINVAL;
		goto fail;
	}
	if (sp->sa_len != dp->sa_len) {
		error = EINVAL;
		goto fail;
	}
	if (af != sp->sa_family || af != dp->sa_family) {
		error = EINVAL;
		goto fail;
	}

	new_ep = _MALLOC(sizeof(*new_ep), M_NETADDR, M_WAITOK | M_ZERO);
	if (new_ep == NULL) {
		error = ENOBUFS;
		goto fail;
	}

	/* check if anyone have already attached with exactly same config */
	lck_rw_lock_exclusive(&encaptab_lock);
	for (ep = LIST_FIRST(&encaptab); ep; ep = LIST_NEXT(ep, chain)) {
		if (ep->af != af) {
			continue;
		}
		if (ep->proto != proto) {
			continue;
		}
		if (ep->src.ss_len != sp->sa_len ||
		    bcmp(&ep->src, sp, sp->sa_len) != 0 ||
		    bcmp(&ep->srcmask, sm, sp->sa_len) != 0) {
			continue;
		}
		if (ep->dst.ss_len != dp->sa_len ||
		    bcmp(&ep->dst, dp, dp->sa_len) != 0 ||
		    bcmp(&ep->dstmask, dm, dp->sa_len) != 0) {
			continue;
		}

		error = EEXIST;
		goto fail_locked;
	}

	new_ep->af = af;
	new_ep->proto = proto;
	bcopy(sp, &new_ep->src, sp->sa_len);
	bcopy(sm, &new_ep->srcmask, sp->sa_len);
	bcopy(dp, &new_ep->dst, dp->sa_len);
	bcopy(dm, &new_ep->dstmask, dp->sa_len);
	new_ep->psw = psw;
	new_ep->arg = arg;

	encap_add_locked(new_ep);
	lck_rw_unlock_exclusive(&encaptab_lock);

	error = 0;
	return new_ep;

fail_locked:
	lck_rw_unlock_exclusive(&encaptab_lock);
	if (new_ep != NULL) {
		_FREE(new_ep, M_NETADDR);
	}
fail:
	return NULL;
}

const struct encaptab *
encap_attach_func( int af, int proto,
    int (*func)(const struct mbuf *, int, int, void *),
    const struct protosw *psw, void *arg)
{
	struct encaptab *ep;
	int error;

	/* sanity check on args */
	if (!func) {
		error = EINVAL;
		goto fail;
	}

	ep = _MALLOC(sizeof(*ep), M_NETADDR, M_WAITOK | M_ZERO); /* XXX */
	if (ep == NULL) {
		error = ENOBUFS;
		goto fail;
	}

	ep->af = af;
	ep->proto = proto;
	ep->func = func;
	ep->psw = psw;
	ep->arg = arg;

	lck_rw_lock_exclusive(&encaptab_lock);
	encap_add_locked(ep);
	lck_rw_unlock_exclusive(&encaptab_lock);

	error = 0;
	return ep;

fail:
	return NULL;
}

int
encap_detach(const struct encaptab *cookie)
{
	const struct encaptab *ep = cookie;
	struct encaptab *p;

	lck_rw_lock_exclusive(&encaptab_lock);
	for (p = LIST_FIRST(&encaptab); p; p = LIST_NEXT(p, chain)) {
		if (p == ep) {
			LIST_REMOVE(p, chain);
			lck_rw_unlock_exclusive(&encaptab_lock);
			_FREE(p, M_NETADDR);    /*XXX*/
			return 0;
		}
	}
	lck_rw_unlock_exclusive(&encaptab_lock);

	return EINVAL;
}

static int
mask_match(const struct encaptab *ep, const struct sockaddr *sp,
    const struct sockaddr *dp)
{
	struct sockaddr_storage s;
	struct sockaddr_storage d;
	int i;
	const u_int8_t *p, *q;
	u_int8_t *r;
	int matchlen;

	if (sp->sa_len > sizeof(s) || dp->sa_len > sizeof(d)) {
		return 0;
	}
	if (sp->sa_family != ep->af || dp->sa_family != ep->af) {
		return 0;
	}
	if (sp->sa_len != ep->src.ss_len || dp->sa_len != ep->dst.ss_len) {
		return 0;
	}

	matchlen = 0;

	p = (const u_int8_t *)sp;
	q = (const u_int8_t *)&ep->srcmask;
	r = (u_int8_t *)&s;
	for (i = 0; i < sp->sa_len; i++) {
		r[i] = p[i] & q[i];
		/* XXX estimate */
		matchlen += (q[i] ? 8 : 0);
	}

	p = (const u_int8_t *)dp;
	q = (const u_int8_t *)&ep->dstmask;
	r = (u_int8_t *)&d;
	for (i = 0; i < dp->sa_len; i++) {
		r[i] = p[i] & q[i];
		/* XXX rough estimate */
		matchlen += (q[i] ? 8 : 0);
	}

	/* need to overwrite len/family portion as we don't compare them */
	s.ss_len = sp->sa_len;
	s.ss_family = sp->sa_family;
	d.ss_len = dp->sa_len;
	d.ss_family = dp->sa_family;

	if (bcmp(&s, &ep->src, ep->src.ss_len) == 0 &&
	    bcmp(&d, &ep->dst, ep->dst.ss_len) == 0) {
		return matchlen;
	} else {
		return 0;
	}
}

struct encaptabtag {
	void*                   *arg;
};

static void
encap_fillarg(
	struct mbuf *m,
	void *arg)
{
	struct m_tag    *tag;
	struct encaptabtag *et;

	tag = m_tag_create(KERNEL_MODULE_TAG_ID, KERNEL_TAG_TYPE_ENCAP,
	    sizeof(struct encaptabtag), M_WAITOK, m);

	if (tag != NULL) {
		et = (struct encaptabtag*)(tag + 1);
		et->arg = arg;
		m_tag_prepend(m, tag);
	}
}

void *
encap_getarg(struct mbuf *m)
{
	struct m_tag    *tag;
	struct encaptabtag *et;
	void *p = NULL;

	tag = m_tag_locate(m, KERNEL_MODULE_TAG_ID, KERNEL_TAG_TYPE_ENCAP, NULL);
	if (tag) {
		et = (struct encaptabtag*)(tag + 1);
		p = et->arg;
		m_tag_delete(m, tag);
	}

	return p;
}
Commit	Line	Data
1c79356b	1	/*
39037602	2	* Copyright (c) 2000-2016 Apple Inc. All rights reserved.
5d5c5d0d	3	*
2d21ac55	4	* @APPLE_OSREFERENCE_LICENSE_HEADER_START@
39037602	5	*
2d21ac55 A	6	* This file contains Original Code and/or Modifications of Original Code
	7	* as defined in and that are subject to the Apple Public Source License
	8	* Version 2.0 (the 'License'). You may not use this file except in
	9	* compliance with the License. The rights granted to you under the License
	10	* may not be used to create, or enable the creation or redistribution of,
	11	* unlawful or unlicensed copies of an Apple operating system, or to
	12	* circumvent, violate, or enable the circumvention or violation of, any
	13	* terms of an Apple operating system software license agreement.
39037602	14	*
2d21ac55 A	15	* Please obtain a copy of the License at
2d21ac55 A	16	* http://www.opensource.apple.com/apsl/ and read it before using this file.
39037602	17	*
2d21ac55 A	18	* The Original Code and all software distributed under the License are
2d21ac55 A	19	* distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
8f6c56a5 A	20	* EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
8f6c56a5 A	21	* INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
2d21ac55 A	22	* FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
	23	* Please see the License for the specific language governing rights and
	24	* limitations under the License.
39037602	25	*
2d21ac55	26	* @APPLE_OSREFERENCE_LICENSE_HEADER_END@
1c79356b	27	*/
9bccf70c A	28	/* $FreeBSD: src/sys/netinet/ip_encap.c,v 1.1.2.2 2001/07/03 11:01:46 ume Exp $ */
9bccf70c A	29	/* $KAME: ip_encap.c,v 1.41 2001/03/15 08:35:08 itojun Exp $ */
1c79356b A	30
	31	/*
	32	* Copyright (C) 1995, 1996, 1997, and 1998 WIDE Project.
	33	* All rights reserved.
	34	*
	35	* Redistribution and use in source and binary forms, with or without
	36	* modification, are permitted provided that the following conditions
	37	* are met:
	38	* 1. Redistributions of source code must retain the above copyright
	39	* notice, this list of conditions and the following disclaimer.
	40	* 2. Redistributions in binary form must reproduce the above copyright
	41	* notice, this list of conditions and the following disclaimer in the
	42	* documentation and/or other materials provided with the distribution.
	43	* 3. Neither the name of the project nor the names of its contributors
	44	* may be used to endorse or promote products derived from this software
	45	* without specific prior written permission.
	46	*
	47	* THIS SOFTWARE IS PROVIDED BY THE PROJECT AND CONTRIBUTORS ``AS IS'' AND
	48	* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
	49	* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
	50	* ARE DISCLAIMED. IN NO EVENT SHALL THE PROJECT OR CONTRIBUTORS BE LIABLE
	51	* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
	52	* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
	53	* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
	54	* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
	55	* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
	56	* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
	57	* SUCH DAMAGE.
	58	*/
	59	/*
	60	* My grandfather said that there's a devil inside tunnelling technology...
	61	*
	62	* We have surprisingly many protocols that want packets with IP protocol
	63	* #4 or #41. Here's a list of protocols that want protocol #41:
	64	* RFC1933 configured tunnel
	65	* RFC1933 automatic tunnel
	66	* RFC2401 IPsec tunnel
	67	* RFC2473 IPv6 generic packet tunnelling
	68	* RFC2529 6over4 tunnel
	69	* mobile-ip6 (uses RFC2473)
	70	* 6to4 tunnel
	71	* Here's a list of protocol that want protocol #4:
9bccf70c A	72	* RFC1853 IPv4-in-IPv4 tunnelling
9bccf70c A	73	* RFC2003 IPv4 encapsulation within IPv4
1c79356b A	74	* RFC2344 reverse tunnelling for mobile-ip4
	75	* RFC2401 IPsec tunnel
	76	* Well, what can I say. They impose different en/decapsulation mechanism
	77	* from each other, so they need separate protocol handler. The only one
	78	* we can easily determine by protocol # is IPsec, which always has
cb323159	79	* AH/ESP header right after outer IP header.
1c79356b A	80	*
	81	* So, clearly good old protosw does not work for protocol #4 and #41.
	82	* The code will let you match protocol via src/dst address pair.
	83	*/
9bccf70c	84	/* XXX is M_NETADDR correct? */
1c79356b A	85
	86	#include <sys/param.h>
	87	#include <sys/systm.h>
	88	#include <sys/socket.h>
	89	#include <sys/sockio.h>
	90	#include <sys/mbuf.h>
316670eb	91	#include <sys/mcache.h>
1c79356b	92	#include <sys/errno.h>
39236c6e	93	#include <sys/domain.h>
1c79356b	94	#include <sys/protosw.h>
9bccf70c	95	#include <sys/queue.h>
1c79356b A	96
	97	#include <net/if.h>
	98	#include <net/route.h>
	99
	100	#include <netinet/in.h>
	101	#include <netinet/in_systm.h>
	102	#include <netinet/ip.h>
	103	#include <netinet/ip_var.h>
	104	#include <netinet/ip_encap.h>
1c79356b A	105
	106	#if INET6
	107	#include <netinet/ip6.h>
	108	#include <netinet6/ip6_var.h>
	109	#include <netinet6/ip6protosw.h>
	110	#endif
	111
	112
	113	#include <net/net_osdep.h>
	114
9bccf70c	115	#ifndef __APPLE__
1c79356b A	116	#include <sys/kernel.h>
	117	#include <sys/malloc.h>
	118	MALLOC_DEFINE(M_NETADDR, "Export Host", "Export host address structure");
	119	#endif
	120
39236c6e	121	static void encap_init(struct protosw , struct domain );
0a7de745	122	static void encap_add_locked(struct encaptab *);
91447636	123	static int mask_match(const struct encaptab , const struct sockaddr ,
0a7de745 A	124	const struct sockaddr *);
0a7de745 A	125	static void encap_fillarg(struct mbuf , void arg);
1c79356b	126
9bccf70c	127	#ifndef LIST_HEAD_INITIALIZER
1c79356b A	128	/* rely upon BSS initialization */
1c79356b A	129	LIST_HEAD(, encaptab) encaptab;
9bccf70c A	130	#else
	131	LIST_HEAD(, encaptab) encaptab = LIST_HEAD_INITIALIZER(&encaptab);
	132	#endif
1c79356b	133
0a7de745 A	134	decl_lck_rw_data(static, encaptab_lock);
0a7de745 A	135
39236c6e A	136	static void
39236c6e A	137	encap_init(struct protosw pp, struct domain dp)
1c79356b	138	{
39236c6e A	139	#pragma unused(dp)
39236c6e A	140	static int encap_initialized = 0;
0a7de745 A	141	lck_grp_attr_t *encaptab_grp_attrib = NULL;
	142	lck_attr_t *encaptab_lck_attrib = NULL;
	143	lck_grp_t *encaptab_lck_group = NULL;
39236c6e	144
0a7de745	145	VERIFY((pp->pr_flags & (PR_INITIALIZED \| PR_ATTACHED)) == PR_ATTACHED);
9bccf70c	146
39236c6e	147	/* This gets called by more than one protocols, so initialize once */
0a7de745	148	if (encap_initialized) {
9bccf70c	149	return;
0a7de745 A	150	}
	151
	152	encaptab_grp_attrib = lck_grp_attr_alloc_init();
	153	encaptab_lck_group = lck_grp_alloc_init("encaptab lock", encaptab_grp_attrib);
	154	lck_grp_attr_free(encaptab_grp_attrib);
	155
	156	encaptab_lck_attrib = lck_attr_alloc_init();
	157	lck_rw_init(&encaptab_lock, encaptab_lck_group, encaptab_lck_attrib);
	158
	159	lck_grp_free(encaptab_lck_group);
	160	lck_attr_free(encaptab_lck_attrib);
	161
39236c6e	162	encap_initialized = 1;
1c79356b A	163	#if 0
	164	/*
	165	* we cannot use LIST_INIT() here, since drivers may want to call
9bccf70c	166	* encap_attach(), on driver attach. encap_init() will be called
1c79356b A	167	* on AF_INET{,6} initialization, which happens after driver
	168	* initialization - using LIST_INIT() here can nuke encap_attach()
	169	* from drivers.
	170	*/
	171	LIST_INIT(&encaptab);
	172	#endif
	173	}
	174
39236c6e A	175	void
	176	encap4_init(struct protosw pp, struct domain dp)
	177	{
	178	encap_init(pp, dp);
	179	}
	180
	181	void
	182	encap6_init(struct ip6protosw pp, struct domain dp)
	183	{
	184	encap_init((struct protosw *)pp, dp);
	185	}
	186
9bccf70c	187	#if INET
1c79356b	188	void
39037602	189	encap4_input(struct mbuf *m, int off)
1c79356b	190	{
9bccf70c	191	int proto;
1c79356b A	192	struct ip *ip;
1c79356b A	193	struct sockaddr_in s, d;
9bccf70c A	194	const struct protosw *psw;
	195	struct encaptab ep, match;
	196	int prio, matchprio;
0a7de745	197	void *match_arg = NULL;
9bccf70c A	198
	199	#ifndef __APPLE__
	200	va_start(ap, m);
	201	off = va_arg(ap, int);
	202	proto = va_arg(ap, int);
	203	va_end(ap);
	204	#endif
1c79356b	205
316670eb A	206	/* Expect 32-bit aligned data pointer on strict-align platforms */
	207	MBUF_STRICT_DATA_ALIGNMENT_CHECK_32(m);
	208
1c79356b	209	ip = mtod(m, struct ip *);
9bccf70c	210	#ifdef __APPLE__
1c79356b A	211	proto = ip->ip_p;
	212	#endif
	213
	214	bzero(&s, sizeof(s));
	215	s.sin_family = AF_INET;
	216	s.sin_len = sizeof(struct sockaddr_in);
	217	s.sin_addr = ip->ip_src;
	218	bzero(&d, sizeof(d));
	219	d.sin_family = AF_INET;
	220	d.sin_len = sizeof(struct sockaddr_in);
	221	d.sin_addr = ip->ip_dst;
	222
9bccf70c A	223	match = NULL;
9bccf70c A	224	matchprio = 0;
0a7de745 A	225
0a7de745 A	226	lck_rw_lock_shared(&encaptab_lock);
1c79356b	227	for (ep = LIST_FIRST(&encaptab); ep; ep = LIST_NEXT(ep, chain)) {
0a7de745	228	if (ep->af != AF_INET) {
9bccf70c	229	continue;
0a7de745 A	230	}
0a7de745 A	231	if (ep->proto >= 0 && ep->proto != proto) {
1c79356b	232	continue;
0a7de745 A	233	}
0a7de745 A	234	if (ep->func) {
9bccf70c	235	prio = (*ep->func)(m, off, proto, ep->arg);
0a7de745	236	} else {
1c79356b A	237	/*
	238	* it's inbound traffic, we need to match in reverse
	239	* order
	240	*/
9bccf70c A	241	prio = mask_match(ep, (struct sockaddr *)&d,
9bccf70c A	242	(struct sockaddr *)&s);
1c79356b A	243	}
1c79356b A	244
9bccf70c A	245	/*
	246	* We prioritize the matches by using bit length of the
	247	* matches. mask_match() and user-supplied matching function
	248	* should return the bit length of the matches (for example,
	249	* if both src/dst are matched for IPv4, 64 should be returned).
	250	* 0 or negative return value means "it did not match".
	251	*
	252	* The question is, since we have two "mask" portion, we
	253	* cannot really define total order between entries.
	254	* For example, which of these should be preferred?
	255	* mask_match() returns 48 (32 + 16) for both of them.
	256	* src=3ffe::/16, dst=3ffe:501::/32
	257	* src=3ffe:501::/32, dst=3ffe::/16
	258	*
	259	* We need to loop through all the possible candidates
	260	* to get the best match - the search takes O(n) for
	261	* n attachments (i.e. interfaces).
	262	*/
0a7de745	263	if (prio <= 0) {
9bccf70c	264	continue;
0a7de745	265	}
9bccf70c A	266	if (prio > matchprio) {
	267	matchprio = prio;
	268	match = ep;
0a7de745 A	269	psw = (const struct protosw *)match->psw;
0a7de745 A	270	match_arg = ep->arg;
9bccf70c A	271	}
9bccf70c A	272	}
0a7de745	273	lck_rw_unlock_shared(&encaptab_lock);
9bccf70c A	274
	275	if (match) {
	276	/* found a match, "match" has the best one */
9bccf70c	277	if (psw && psw->pr_input) {
0a7de745	278	encap_fillarg(m, match_arg);
9bccf70c	279	(*psw->pr_input)(m, off);
0a7de745	280	} else {
1c79356b	281	m_freem(m);
0a7de745	282	}
1c79356b A	283	return;
	284	}
	285
1c79356b A	286	/* last resort: inject to raw socket */
	287	rip_input(m, off);
	288	}
9bccf70c	289	#endif
1c79356b A	290
	291	#if INET6
	292	int
6d2010ae	293	encap6_input(struct mbuf *mp, int offp, int proto)
1c79356b A	294	{
	295	struct mbuf m = mp;
	296	struct ip6_hdr *ip6;
	297	struct sockaddr_in6 s, d;
9bccf70c A	298	const struct ip6protosw *psw;
	299	struct encaptab ep, match;
	300	int prio, matchprio;
0a7de745	301	void *match_arg = NULL;
1c79356b	302
316670eb A	303	/* Expect 32-bit aligned data pointer on strict-align platforms */
316670eb A	304	MBUF_STRICT_DATA_ALIGNMENT_CHECK_32(m);
1c79356b	305
316670eb	306	ip6 = mtod(m, struct ip6_hdr *);
1c79356b A	307	bzero(&s, sizeof(s));
	308	s.sin6_family = AF_INET6;
	309	s.sin6_len = sizeof(struct sockaddr_in6);
	310	s.sin6_addr = ip6->ip6_src;
	311	bzero(&d, sizeof(d));
	312	d.sin6_family = AF_INET6;
	313	d.sin6_len = sizeof(struct sockaddr_in6);
	314	d.sin6_addr = ip6->ip6_dst;
	315
9bccf70c A	316	match = NULL;
9bccf70c A	317	matchprio = 0;
0a7de745 A	318
0a7de745 A	319	lck_rw_lock_shared(&encaptab_lock);
1c79356b	320	for (ep = LIST_FIRST(&encaptab); ep; ep = LIST_NEXT(ep, chain)) {
0a7de745	321	if (ep->af != AF_INET6) {
9bccf70c	322	continue;
0a7de745 A	323	}
0a7de745 A	324	if (ep->proto >= 0 && ep->proto != proto) {
1c79356b	325	continue;
0a7de745 A	326	}
0a7de745 A	327	if (ep->func) {
9bccf70c	328	prio = (ep->func)(m, offp, proto, ep->arg);
0a7de745	329	} else {
1c79356b A	330	/*
	331	* it's inbound traffic, we need to match in reverse
	332	* order
	333	*/
9bccf70c A	334	prio = mask_match(ep, (struct sockaddr *)&d,
	335	(struct sockaddr *)&s);
	336	}
	337
	338	/* see encap4_input() for issues here */
0a7de745	339	if (prio <= 0) {
9bccf70c	340	continue;
0a7de745	341	}
9bccf70c A	342	if (prio > matchprio) {
	343	matchprio = prio;
	344	match = ep;
0a7de745 A	345	psw = (const struct ip6protosw *)match->psw;
0a7de745 A	346	match_arg = ep->arg;
1c79356b	347	}
9bccf70c	348	}
0a7de745	349	lck_rw_unlock_shared(&encaptab_lock);
1c79356b	350
9bccf70c	351	if (match) {
1c79356b	352	/* found a match */
1c79356b	353	if (psw && psw->pr_input) {
0a7de745	354	encap_fillarg(m, match_arg);
6d2010ae	355	return (*psw->pr_input)(mp, offp, proto);
1c79356b A	356	} else {
	357	m_freem(m);
	358	return IPPROTO_DONE;
	359	}
	360	}
	361
	362	/* last resort: inject to raw socket */
6d2010ae	363	return rip6_input(mp, offp, proto);
1c79356b A	364	}
	365	#endif
	366
9bccf70c	367	static void
0a7de745	368	encap_add_locked(struct encaptab *ep)
9bccf70c	369	{
0a7de745	370	LCK_RW_ASSERT(&encaptab_lock, LCK_RW_ASSERT_EXCLUSIVE);
9bccf70c A	371	LIST_INSERT_HEAD(&encaptab, ep, chain);
	372	}
	373
1c79356b A	374	/*
	375	* sp (src ptr) is always my side, and dp (dst ptr) is always remote side.
	376	* length of mask (sm and dm) is assumed to be same as sp/dp.
	377	* Return value will be necessary as input (cookie) for encap_detach().
	378	*/
	379	const struct encaptab *
39037602	380	encap_attach(int af, int proto, const struct sockaddr *sp,
0a7de745 A	381	const struct sockaddr sm, const struct sockaddr dp,
0a7de745 A	382	const struct sockaddr dm, const struct protosw psw, void *arg)
1c79356b	383	{
0a7de745 A	384	struct encaptab *ep = NULL;
0a7de745 A	385	struct encaptab *new_ep = NULL;
1c79356b	386	int error;
1c79356b	387
1c79356b	388	/* sanity check on args */
0a7de745	389	if (sp->sa_len > sizeof(new_ep->src) \|\| dp->sa_len > sizeof(new_ep->dst)) {
1c79356b A	390	error = EINVAL;
	391	goto fail;
	392	}
	393	if (sp->sa_len != dp->sa_len) {
	394	error = EINVAL;
	395	goto fail;
	396	}
	397	if (af != sp->sa_family \|\| af != dp->sa_family) {
	398	error = EINVAL;
	399	goto fail;
	400	}
	401
0a7de745 A	402	new_ep = _MALLOC(sizeof(*new_ep), M_NETADDR, M_WAITOK \| M_ZERO);
	403	if (new_ep == NULL) {
	404	error = ENOBUFS;
	405	goto fail;
	406	}
	407
1c79356b	408	/* check if anyone have already attached with exactly same config */
0a7de745	409	lck_rw_lock_exclusive(&encaptab_lock);
1c79356b	410	for (ep = LIST_FIRST(&encaptab); ep; ep = LIST_NEXT(ep, chain)) {
0a7de745	411	if (ep->af != af) {
1c79356b	412	continue;
0a7de745 A	413	}
0a7de745 A	414	if (ep->proto != proto) {
1c79356b	415	continue;
0a7de745	416	}
1c79356b A	417	if (ep->src.ss_len != sp->sa_len \|\|
1c79356b A	418	bcmp(&ep->src, sp, sp->sa_len) != 0 \|\|
0a7de745	419	bcmp(&ep->srcmask, sm, sp->sa_len) != 0) {
1c79356b	420	continue;
0a7de745	421	}
1c79356b A	422	if (ep->dst.ss_len != dp->sa_len \|\|
1c79356b A	423	bcmp(&ep->dst, dp, dp->sa_len) != 0 \|\|
0a7de745	424	bcmp(&ep->dstmask, dm, dp->sa_len) != 0) {
1c79356b	425	continue;
0a7de745	426	}
1c79356b A	427
1c79356b A	428	error = EEXIST;
0a7de745	429	goto fail_locked;
1c79356b	430	}
1c79356b	431
0a7de745 A	432	new_ep->af = af;
	433	new_ep->proto = proto;
	434	bcopy(sp, &new_ep->src, sp->sa_len);
	435	bcopy(sm, &new_ep->srcmask, sp->sa_len);
	436	bcopy(dp, &new_ep->dst, dp->sa_len);
	437	bcopy(dm, &new_ep->dstmask, dp->sa_len);
	438	new_ep->psw = psw;
	439	new_ep->arg = arg;
1c79356b	440
0a7de745 A	441	encap_add_locked(new_ep);
0a7de745 A	442	lck_rw_unlock_exclusive(&encaptab_lock);
9bccf70c	443
1c79356b	444	error = 0;
0a7de745	445	return new_ep;
1c79356b	446
0a7de745 A	447	fail_locked:
	448	lck_rw_unlock_exclusive(&encaptab_lock);
	449	if (new_ep != NULL) {
	450	_FREE(new_ep, M_NETADDR);
	451	}
1c79356b	452	fail:
1c79356b A	453	return NULL;
	454	}
	455
	456	const struct encaptab *
39037602	457	encap_attach_func( int af, int proto,
0a7de745 A	458	int (func)(const struct mbuf , int, int, void *),
0a7de745 A	459	const struct protosw psw, void arg)
1c79356b A	460	{
	461	struct encaptab *ep;
	462	int error;
1c79356b	463
1c79356b A	464	/* sanity check on args */
	465	if (!func) {
	466	error = EINVAL;
	467	goto fail;
	468	}
	469
3e170ce0	470	ep = _MALLOC(sizeof(ep), M_NETADDR, M_WAITOK \| M_ZERO); / XXX */
1c79356b A	471	if (ep == NULL) {
	472	error = ENOBUFS;
	473	goto fail;
	474	}
1c79356b A	475
	476	ep->af = af;
	477	ep->proto = proto;
	478	ep->func = func;
	479	ep->psw = psw;
	480	ep->arg = arg;
	481
0a7de745 A	482	lck_rw_lock_exclusive(&encaptab_lock);
	483	encap_add_locked(ep);
	484	lck_rw_unlock_exclusive(&encaptab_lock);
9bccf70c	485
1c79356b	486	error = 0;
1c79356b A	487	return ep;
	488
	489	fail:
1c79356b A	490	return NULL;
	491	}
	492
	493	int
39037602	494	encap_detach(const struct encaptab *cookie)
1c79356b A	495	{
	496	const struct encaptab *ep = cookie;
	497	struct encaptab *p;
	498
0a7de745	499	lck_rw_lock_exclusive(&encaptab_lock);
1c79356b A	500	for (p = LIST_FIRST(&encaptab); p; p = LIST_NEXT(p, chain)) {
	501	if (p == ep) {
	502	LIST_REMOVE(p, chain);
0a7de745 A	503	lck_rw_unlock_exclusive(&encaptab_lock);
0a7de745 A	504	_FREE(p, M_NETADDR); /XXX/
1c79356b A	505	return 0;
	506	}
	507	}
0a7de745	508	lck_rw_unlock_exclusive(&encaptab_lock);
1c79356b A	509
	510	return EINVAL;
	511	}
	512
	513	static int
39037602	514	mask_match(const struct encaptab ep, const struct sockaddr sp,
0a7de745	515	const struct sockaddr *dp)
1c79356b A	516	{
	517	struct sockaddr_storage s;
	518	struct sockaddr_storage d;
	519	int i;
9bccf70c A	520	const u_int8_t p, q;
	521	u_int8_t *r;
	522	int matchlen;
1c79356b	523
0a7de745	524	if (sp->sa_len > sizeof(s) \|\| dp->sa_len > sizeof(d)) {
1c79356b	525	return 0;
0a7de745 A	526	}
0a7de745 A	527	if (sp->sa_family != ep->af \|\| dp->sa_family != ep->af) {
1c79356b	528	return 0;
0a7de745 A	529	}
0a7de745 A	530	if (sp->sa_len != ep->src.ss_len \|\| dp->sa_len != ep->dst.ss_len) {
1c79356b	531	return 0;
0a7de745	532	}
1c79356b	533
9bccf70c A	534	matchlen = 0;
	535
	536	p = (const u_int8_t *)sp;
	537	q = (const u_int8_t *)&ep->srcmask;
1c79356b	538	r = (u_int8_t *)&s;
0a7de745	539	for (i = 0; i < sp->sa_len; i++) {
1c79356b	540	r[i] = p[i] & q[i];
9bccf70c A	541	/* XXX estimate */
	542	matchlen += (q[i] ? 8 : 0);
	543	}
1c79356b	544
9bccf70c A	545	p = (const u_int8_t *)dp;
9bccf70c A	546	q = (const u_int8_t *)&ep->dstmask;
1c79356b	547	r = (u_int8_t *)&d;
0a7de745	548	for (i = 0; i < dp->sa_len; i++) {
1c79356b	549	r[i] = p[i] & q[i];
9bccf70c A	550	/* XXX rough estimate */
	551	matchlen += (q[i] ? 8 : 0);
	552	}
1c79356b A	553
	554	/* need to overwrite len/family portion as we don't compare them */
	555	s.ss_len = sp->sa_len;
	556	s.ss_family = sp->sa_family;
	557	d.ss_len = dp->sa_len;
	558	d.ss_family = dp->sa_family;
	559
	560	if (bcmp(&s, &ep->src, ep->src.ss_len) == 0 &&
	561	bcmp(&d, &ep->dst, ep->dst.ss_len) == 0) {
9bccf70c	562	return matchlen;
0a7de745	563	} else {
1c79356b	564	return 0;
0a7de745	565	}
1c79356b A	566	}
1c79356b A	567
2d21ac55	568	struct encaptabtag {
0a7de745	569	void* *arg;
2d21ac55 A	570	};
2d21ac55 A	571
1c79356b	572	static void
2d21ac55 A	573	encap_fillarg(
2d21ac55 A	574	struct mbuf *m,
0a7de745	575	void *arg)
1c79356b	576	{
0a7de745	577	struct m_tag *tag;
2d21ac55	578	struct encaptabtag *et;
0a7de745	579
6d2010ae	580	tag = m_tag_create(KERNEL_MODULE_TAG_ID, KERNEL_TAG_TYPE_ENCAP,
0a7de745 A	581	sizeof(struct encaptabtag), M_WAITOK, m);
0a7de745 A	582
2d21ac55 A	583	if (tag != NULL) {
2d21ac55 A	584	et = (struct encaptabtag*)(tag + 1);
0a7de745	585	et->arg = arg;
2d21ac55	586	m_tag_prepend(m, tag);
1c79356b	587	}
1c79356b A	588	}
	589
	590	void *
39037602	591	encap_getarg(struct mbuf *m)
1c79356b	592	{
0a7de745	593	struct m_tag *tag;
2d21ac55 A	594	struct encaptabtag *et;
2d21ac55 A	595	void *p = NULL;
0a7de745	596
2d21ac55 A	597	tag = m_tag_locate(m, KERNEL_MODULE_TAG_ID, KERNEL_TAG_TYPE_ENCAP, NULL);
	598	if (tag) {
	599	et = (struct encaptabtag*)(tag + 1);
	600	p = et->arg;
	601	m_tag_delete(m, tag);
1c79356b	602	}
0a7de745	603
1c79356b	604	return p;
1c79356b	605	}