in6_rmx.c revision 283291
1/*-
2 * Copyright (C) 1995, 1996, 1997, and 1998 WIDE Project.
3 * All rights reserved.
4 *
5 * Redistribution and use in source and binary forms, with or without
6 * modification, are permitted provided that the following conditions
7 * are met:
8 * 1. Redistributions of source code must retain the above copyright
9 *    notice, this list of conditions and the following disclaimer.
10 * 2. Redistributions in binary form must reproduce the above copyright
11 *    notice, this list of conditions and the following disclaimer in the
12 *    documentation and/or other materials provided with the distribution.
13 * 3. Neither the name of the project nor the names of its contributors
14 *    may be used to endorse or promote products derived from this software
15 *    without specific prior written permission.
16 *
17 * THIS SOFTWARE IS PROVIDED BY THE PROJECT AND CONTRIBUTORS ``AS IS'' AND
18 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20 * ARE DISCLAIMED.  IN NO EVENT SHALL THE PROJECT OR CONTRIBUTORS BE LIABLE
21 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
22 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
23 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
24 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
25 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
26 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
27 * SUCH DAMAGE.
28 *
29 *	$KAME: in6_rmx.c,v 1.11 2001/07/26 06:53:16 jinmei Exp $
30 */
31
32/*-
33 * Copyright 1994, 1995 Massachusetts Institute of Technology
34 *
35 * Permission to use, copy, modify, and distribute this software and
36 * its documentation for any purpose and without fee is hereby
37 * granted, provided that both the above copyright notice and this
38 * permission notice appear in all copies, that both the above
39 * copyright notice and this permission notice appear in all
40 * supporting documentation, and that the name of M.I.T. not be used
41 * in advertising or publicity pertaining to distribution of the
42 * software without specific, written prior permission.  M.I.T. makes
43 * no representations about the suitability of this software for any
44 * purpose.  It is provided "as is" without express or implied
45 * warranty.
46 *
47 * THIS SOFTWARE IS PROVIDED BY M.I.T. ``AS IS''.  M.I.T. DISCLAIMS
48 * ALL EXPRESS OR IMPLIED WARRANTIES WITH REGARD TO THIS SOFTWARE,
49 * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
50 * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. IN NO EVENT
51 * SHALL M.I.T. BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
52 * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
53 * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF
54 * USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
55 * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
56 * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT
57 * OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
58 * SUCH DAMAGE.
59 *
60 */
61
62#include <sys/cdefs.h>
63__FBSDID("$FreeBSD: head/sys/netinet6/in6_rmx.c 283291 2015-05-22 17:05:21Z jkim $");
64
65#include <sys/param.h>
66#include <sys/systm.h>
67#include <sys/kernel.h>
68#include <sys/lock.h>
69#include <sys/queue.h>
70#include <sys/socket.h>
71#include <sys/socketvar.h>
72#include <sys/mbuf.h>
73#include <sys/rwlock.h>
74#include <sys/syslog.h>
75#include <sys/callout.h>
76
77#include <net/if.h>
78#include <net/if_var.h>
79#include <net/route.h>
80
81#include <netinet/in.h>
82#include <netinet/ip_var.h>
83#include <netinet/in_var.h>
84
85#include <netinet/ip6.h>
86#include <netinet6/ip6_var.h>
87
88#include <netinet/icmp6.h>
89#include <netinet6/nd6.h>
90
91#include <netinet/tcp.h>
92#include <netinet/tcp_seq.h>
93#include <netinet/tcp_timer.h>
94#include <netinet/tcp_var.h>
95
96extern int	in6_inithead(void **head, int off);
97#ifdef VIMAGE
98extern int	in6_detachhead(void **head, int off);
99#endif
100
101/*
102 * Do what we need to do when inserting a route.
103 */
104static struct radix_node *
105in6_addroute(void *v_arg, void *n_arg, struct radix_node_head *head,
106    struct radix_node *treenodes)
107{
108	struct rtentry *rt = (struct rtentry *)treenodes;
109	struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *)rt_key(rt);
110	struct radix_node *ret;
111
112	RADIX_NODE_HEAD_WLOCK_ASSERT(head);
113	if (IN6_IS_ADDR_MULTICAST(&sin6->sin6_addr))
114		rt->rt_flags |= RTF_MULTICAST;
115
116	/*
117	 * A little bit of help for both IPv6 output and input:
118	 *   For local addresses, we make sure that RTF_LOCAL is set,
119	 *   with the thought that this might one day be used to speed up
120	 *   ip_input().
121	 *
122	 * We also mark routes to multicast addresses as such, because
123	 * it's easy to do and might be useful (but this is much more
124	 * dubious since it's so easy to inspect the address).  (This
125	 * is done above.)
126	 *
127	 * XXX
128	 * should elaborate the code.
129	 */
130	if (rt->rt_flags & RTF_HOST) {
131		if (IN6_ARE_ADDR_EQUAL(&satosin6(rt->rt_ifa->ifa_addr)
132					->sin6_addr,
133				       &sin6->sin6_addr)) {
134			rt->rt_flags |= RTF_LOCAL;
135		}
136	}
137
138	if (rt->rt_ifp != NULL) {
139
140		/*
141		 * Check route MTU:
142		 * inherit interface MTU if not set or
143		 * check if MTU is too large.
144		 */
145		if (rt->rt_mtu == 0) {
146			rt->rt_mtu = IN6_LINKMTU(rt->rt_ifp);
147		} else if (rt->rt_mtu > IN6_LINKMTU(rt->rt_ifp))
148			rt->rt_mtu = IN6_LINKMTU(rt->rt_ifp);
149	}
150
151	ret = rn_addroute(v_arg, n_arg, head, treenodes);
152	if (ret == NULL) {
153		struct rtentry *rt2;
154		/*
155		 * We are trying to add a net route, but can't.
156		 * The following case should be allowed, so we'll make a
157		 * special check for this:
158		 *	Two IPv6 addresses with the same prefix is assigned
159		 *	to a single interrface.
160		 *	# ifconfig if0 inet6 3ffe:0501::1 prefix 64 alias (*1)
161		 *	# ifconfig if0 inet6 3ffe:0501::2 prefix 64 alias (*2)
162		 *	In this case, (*1) and (*2) want to add the same
163		 *	net route entry, 3ffe:0501:: -> if0.
164		 *	This case should not raise an error.
165		 */
166		rt2 = in6_rtalloc1((struct sockaddr *)sin6, 0, RTF_RNH_LOCKED,
167		    rt->rt_fibnum);
168		if (rt2) {
169			if (((rt2->rt_flags & (RTF_HOST|RTF_GATEWAY)) == 0)
170			 && rt2->rt_gateway
171			 && rt2->rt_gateway->sa_family == AF_LINK
172			 && rt2->rt_ifp == rt->rt_ifp) {
173				ret = rt2->rt_nodes;
174			}
175			RTFREE_LOCKED(rt2);
176		}
177	}
178	return (ret);
179}
180
181/*
182 * Age old PMTUs.
183 */
184struct mtuex_arg {
185	struct radix_node_head *rnh;
186	time_t nextstop;
187};
188static VNET_DEFINE(struct callout, rtq_mtutimer);
189#define	V_rtq_mtutimer			VNET(rtq_mtutimer)
190
191static int
192in6_mtuexpire(struct radix_node *rn, void *rock)
193{
194	struct rtentry *rt = (struct rtentry *)rn;
195	struct mtuex_arg *ap = rock;
196
197	/* sanity */
198	if (!rt)
199		panic("rt == NULL in in6_mtuexpire");
200
201	if (rt->rt_expire && !(rt->rt_flags & RTF_PROBEMTU)) {
202		if (rt->rt_expire <= time_uptime) {
203			rt->rt_flags |= RTF_PROBEMTU;
204		} else {
205			ap->nextstop = lmin(ap->nextstop, rt->rt_expire);
206		}
207	}
208
209	return 0;
210}
211
212#define	MTUTIMO_DEFAULT	(60*1)
213
214static void
215in6_mtutimo_one(struct radix_node_head *rnh)
216{
217	struct mtuex_arg arg;
218
219	arg.rnh = rnh;
220	arg.nextstop = time_uptime + MTUTIMO_DEFAULT;
221	RADIX_NODE_HEAD_LOCK(rnh);
222	rnh->rnh_walktree(rnh, in6_mtuexpire, &arg);
223	RADIX_NODE_HEAD_UNLOCK(rnh);
224}
225
226static void
227in6_mtutimo(void *rock)
228{
229	CURVNET_SET_QUIET((struct vnet *) rock);
230	struct radix_node_head *rnh;
231	struct timeval atv;
232	u_int fibnum;
233
234	for (fibnum = 0; fibnum < rt_numfibs; fibnum++) {
235		rnh = rt_tables_get_rnh(fibnum, AF_INET6);
236		if (rnh != NULL)
237			in6_mtutimo_one(rnh);
238	}
239
240	atv.tv_sec = MTUTIMO_DEFAULT;
241	atv.tv_usec = 0;
242	callout_reset(&V_rtq_mtutimer, tvtohz(&atv), in6_mtutimo, rock);
243	CURVNET_RESTORE();
244}
245
246/*
247 * Initialize our routing tree.
248 */
249static VNET_DEFINE(int, _in6_rt_was_here);
250#define	V__in6_rt_was_here	VNET(_in6_rt_was_here)
251
252int
253in6_inithead(void **head, int off)
254{
255	struct radix_node_head *rnh;
256
257	if (!rn_inithead(head, offsetof(struct sockaddr_in6, sin6_addr) << 3))
258		return (0);
259
260	rnh = *head;
261	RADIX_NODE_HEAD_LOCK_INIT(rnh);
262
263	rnh->rnh_addaddr = in6_addroute;
264
265	if (V__in6_rt_was_here == 0) {
266		callout_init(&V_rtq_mtutimer, 1);
267		in6_mtutimo(curvnet);	/* kick off timeout first time */
268		V__in6_rt_was_here = 1;
269	}
270
271	return (1);
272}
273
274#ifdef VIMAGE
275int
276in6_detachhead(void **head, int off)
277{
278
279	callout_drain(&V_rtq_mtutimer);
280	return (1);
281}
282#endif
283
284/*
285 * Extended API for IPv6 FIB support.
286 */
287void
288in6_rtredirect(struct sockaddr *dst, struct sockaddr *gw, struct sockaddr *nm,
289    int flags, struct sockaddr *src, u_int fibnum)
290{
291
292	rtredirect_fib(dst, gw, nm, flags, src, fibnum);
293}
294
295int
296in6_rtrequest(int req, struct sockaddr *dst, struct sockaddr *gw,
297    struct sockaddr *mask, int flags, struct rtentry **ret_nrt, u_int fibnum)
298{
299
300	return (rtrequest_fib(req, dst, gw, mask, flags, ret_nrt, fibnum));
301}
302
303void
304in6_rtalloc(struct route_in6 *ro, u_int fibnum)
305{
306
307	rtalloc_ign_fib((struct route *)ro, 0ul, fibnum);
308}
309
310void
311in6_rtalloc_ign(struct route_in6 *ro, u_long ignflags, u_int fibnum)
312{
313
314	rtalloc_ign_fib((struct route *)ro, ignflags, fibnum);
315}
316
317struct rtentry *
318in6_rtalloc1(struct sockaddr *dst, int report, u_long ignflags, u_int fibnum)
319{
320
321	return (rtalloc1_fib(dst, report, ignflags, fibnum));
322}
323