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 274175 2014-11-06 13:13:09Z melifaro $");
64
65#include <sys/param.h>
66#include <sys/systm.h>
67#include <sys/kernel.h>
68#include <sys/lock.h>
69#include <sys/sysctl.h>
70#include <sys/queue.h>
71#include <sys/socket.h>
72#include <sys/socketvar.h>
73#include <sys/mbuf.h>
74#include <sys/rwlock.h>
75#include <sys/syslog.h>
76#include <sys/callout.h>
77
78#include <net/if.h>
79#include <net/if_var.h>
80#include <net/route.h>
81
82#include <netinet/in.h>
83#include <netinet/ip_var.h>
84#include <netinet/in_var.h>
85
86#include <netinet/ip6.h>
87#include <netinet6/ip6_var.h>
88
89#include <netinet/icmp6.h>
90#include <netinet6/nd6.h>
91
92#include <netinet/tcp.h>
93#include <netinet/tcp_seq.h>
94#include <netinet/tcp_timer.h>
95#include <netinet/tcp_var.h>
96
97extern int in6_inithead(void **head, int off);
98#ifdef VIMAGE
99extern int in6_detachhead(void **head, int off);
100#endif
101
102/*
103 * Do what we need to do when inserting a route.
104 */
105static struct radix_node *
106in6_addroute(void *v_arg, void *n_arg, struct radix_node_head *head,
107 struct radix_node *treenodes)
108{
109 struct rtentry *rt = (struct rtentry *)treenodes;
110 struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *)rt_key(rt);
111 struct radix_node *ret;
112
113 RADIX_NODE_HEAD_WLOCK_ASSERT(head);
114 if (IN6_IS_ADDR_MULTICAST(&sin6->sin6_addr))
115 rt->rt_flags |= RTF_MULTICAST;
116
117 /*
118 * A little bit of help for both IPv6 output and input:
119 * For local addresses, we make sure that RTF_LOCAL is set,
120 * with the thought that this might one day be used to speed up
121 * ip_input().
122 *
123 * We also mark routes to multicast addresses as such, because
124 * it's easy to do and might be useful (but this is much more
125 * dubious since it's so easy to inspect the address). (This
126 * is done above.)
127 *
128 * XXX
129 * should elaborate the code.
130 */
131 if (rt->rt_flags & RTF_HOST) {
132 if (IN6_ARE_ADDR_EQUAL(&satosin6(rt->rt_ifa->ifa_addr)
133 ->sin6_addr,
134 &sin6->sin6_addr)) {
135 rt->rt_flags |= RTF_LOCAL;
136 }
137 }
138
139 if (rt->rt_ifp != NULL) {
140
141 /*
142 * Check route MTU:
143 * inherit interface MTU if not set or
144 * check if MTU is too large.
145 */
146 if (rt->rt_mtu == 0) {
147 rt->rt_mtu = IN6_LINKMTU(rt->rt_ifp);
148 } else if (rt->rt_mtu > IN6_LINKMTU(rt->rt_ifp))
149 rt->rt_mtu = IN6_LINKMTU(rt->rt_ifp);
150 }
151
152 ret = rn_addroute(v_arg, n_arg, head, treenodes);
153 if (ret == NULL) {
154 struct rtentry *rt2;
155 /*
156 * We are trying to add a net route, but can't.
157 * The following case should be allowed, so we'll make a
158 * special check for this:
159 * Two IPv6 addresses with the same prefix is assigned
160 * to a single interrface.
161 * # ifconfig if0 inet6 3ffe:0501::1 prefix 64 alias (*1)
162 * # ifconfig if0 inet6 3ffe:0501::2 prefix 64 alias (*2)
163 * In this case, (*1) and (*2) want to add the same
164 * net route entry, 3ffe:0501:: -> if0.
165 * This case should not raise an error.
166 */
167 rt2 = in6_rtalloc1((struct sockaddr *)sin6, 0, RTF_RNH_LOCKED,
168 rt->rt_fibnum);
169 if (rt2) {
170 if (((rt2->rt_flags & (RTF_HOST|RTF_GATEWAY)) == 0)
171 && rt2->rt_gateway
172 && rt2->rt_gateway->sa_family == AF_LINK
173 && rt2->rt_ifp == rt->rt_ifp) {
174 ret = rt2->rt_nodes;
175 }
176 RTFREE_LOCKED(rt2);
177 }
178 }
179 return (ret);
180}
181
182SYSCTL_DECL(_net_inet6_ip6);
183
184static VNET_DEFINE(int, rtq_toomany6) = 128;
185 /* 128 cached routes is ``too many'' */
186#define V_rtq_toomany6 VNET(rtq_toomany6)
187SYSCTL_VNET_INT(_net_inet6_ip6, IPV6CTL_RTMAXCACHE, rtmaxcache, CTLFLAG_RW,
188 &VNET_NAME(rtq_toomany6) , 0, "");
189
190struct rtqk_arg {
191 struct radix_node_head *rnh;
192 int mode;
193 int updating;
194 int draining;
195 int killed;
196 int found;
197 time_t nextstop;
198};
199
200/*
201 * Age old PMTUs.
202 */
203struct mtuex_arg {
204 struct radix_node_head *rnh;
205 time_t nextstop;
206};
207static VNET_DEFINE(struct callout, rtq_mtutimer);
208#define V_rtq_mtutimer VNET(rtq_mtutimer)
209
210static int
211in6_mtuexpire(struct radix_node *rn, void *rock)
212{
213 struct rtentry *rt = (struct rtentry *)rn;
214 struct mtuex_arg *ap = rock;
215
216 /* sanity */
217 if (!rt)
218 panic("rt == NULL in in6_mtuexpire");
219
220 if (rt->rt_expire && !(rt->rt_flags & RTF_PROBEMTU)) {
221 if (rt->rt_expire <= time_uptime) {
222 rt->rt_flags |= RTF_PROBEMTU;
223 } else {
224 ap->nextstop = lmin(ap->nextstop, rt->rt_expire);
225 }
226 }
227
228 return 0;
229}
230
231#define MTUTIMO_DEFAULT (60*1)
232
233static void
234in6_mtutimo_one(struct radix_node_head *rnh)
235{
236 struct mtuex_arg arg;
237
238 arg.rnh = rnh;
239 arg.nextstop = time_uptime + MTUTIMO_DEFAULT;
240 RADIX_NODE_HEAD_LOCK(rnh);
241 rnh->rnh_walktree(rnh, in6_mtuexpire, &arg);
242 RADIX_NODE_HEAD_UNLOCK(rnh);
243}
244
245static void
246in6_mtutimo(void *rock)
247{
248 CURVNET_SET_QUIET((struct vnet *) rock);
249 struct radix_node_head *rnh;
250 struct timeval atv;
251 u_int fibnum;
252
253 for (fibnum = 0; fibnum < rt_numfibs; fibnum++) {
254 rnh = rt_tables_get_rnh(fibnum, AF_INET6);
255 if (rnh != NULL)
256 in6_mtutimo_one(rnh);
257 }
258
259 atv.tv_sec = MTUTIMO_DEFAULT;
260 atv.tv_usec = 0;
261 callout_reset(&V_rtq_mtutimer, tvtohz(&atv), in6_mtutimo, rock);
262 CURVNET_RESTORE();
263}
264
265/*
266 * Initialize our routing tree.
267 */
268static VNET_DEFINE(int, _in6_rt_was_here);
269#define V__in6_rt_was_here VNET(_in6_rt_was_here)
270
271int
272in6_inithead(void **head, int off)
273{
274 struct radix_node_head *rnh;
275
276 if (!rn_inithead(head, offsetof(struct sockaddr_in6, sin6_addr) << 3))
277 return (0);
278
279 rnh = *head;
280 RADIX_NODE_HEAD_LOCK_INIT(rnh);
281
282 rnh->rnh_addaddr = in6_addroute;
283
284 if (V__in6_rt_was_here == 0) {
285 callout_init(&V_rtq_mtutimer, CALLOUT_MPSAFE);
286 in6_mtutimo(curvnet); /* kick off timeout first time */
287 V__in6_rt_was_here = 1;
288 }
289
290 return (1);
291}
292
293#ifdef VIMAGE
294int
295in6_detachhead(void **head, int off)
296{
297
298 callout_drain(&V_rtq_mtutimer);
299 return (1);
300}
301#endif
302
303/*
304 * Extended API for IPv6 FIB support.
305 */
306void
307in6_rtredirect(struct sockaddr *dst, struct sockaddr *gw, struct sockaddr *nm,
308 int flags, struct sockaddr *src, u_int fibnum)
309{
310
311 rtredirect_fib(dst, gw, nm, flags, src, fibnum);
312}
313
314int
315in6_rtrequest(int req, struct sockaddr *dst, struct sockaddr *gw,
316 struct sockaddr *mask, int flags, struct rtentry **ret_nrt, u_int fibnum)
317{
318
319 return (rtrequest_fib(req, dst, gw, mask, flags, ret_nrt, fibnum));
320}
321
322void
323in6_rtalloc(struct route_in6 *ro, u_int fibnum)
324{
325
326 rtalloc_ign_fib((struct route *)ro, 0ul, fibnum);
327}
328
329void
330in6_rtalloc_ign(struct route_in6 *ro, u_long ignflags, u_int fibnum)
331{
332
333 rtalloc_ign_fib((struct route *)ro, ignflags, fibnum);
334}
335
336struct rtentry *
337in6_rtalloc1(struct sockaddr *dst, int report, u_long ignflags, u_int fibnum)
338{
339
340 return (rtalloc1_fib(dst, report, ignflags, fibnum));
341}