1/*	$NetBSD: route.h,v 1.79 2011/03/31 19:40:52 dyoung Exp $	*/
2
3/*
4 * Copyright (c) 1980, 1986, 1993
5 *	The Regents of the University of California.  All rights reserved.
6 *
7 * Redistribution and use in source and binary forms, with or without
8 * modification, are permitted provided that the following conditions
9 * are met:
10 * 1. Redistributions of source code must retain the above copyright
11 *    notice, this list of conditions and the following disclaimer.
12 * 2. Redistributions in binary form must reproduce the above copyright
13 *    notice, this list of conditions and the following disclaimer in the
14 *    documentation and/or other materials provided with the distribution.
15 * 3. Neither the name of the University nor the names of its contributors
16 *    may be used to endorse or promote products derived from this software
17 *    without specific prior written permission.
18 *
19 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
20 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
21 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
22 * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
23 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
24 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
25 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
26 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
27 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
28 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
29 * SUCH DAMAGE.
30 *
31 *	@(#)route.h	8.5 (Berkeley) 2/8/95
32 */
33
34#ifndef _NET_ROUTE_H_
35#define _NET_ROUTE_H_
36
37#include <sys/queue.h>
38#include <sys/socket.h>
39#include <sys/types.h>
40#include <net/if.h>
41
42#if !(defined(_KERNEL) || defined(_STANDALONE))
43#include <stdbool.h>
44#endif
45
46/*
47 * Kernel resident routing tables.
48 *
49 * The routing tables are initialized when interface addresses
50 * are set by making entries for all directly connected interfaces.
51 */
52
53/*
54 * A route consists of a destination address and a reference
55 * to a routing entry.  These are often held by protocols
56 * in their control blocks, e.g. inpcb.
57 */
58struct route {
59	struct	rtentry		*_ro_rt;
60	struct	sockaddr	*ro_sa;
61	LIST_ENTRY(route)	ro_rtcache_next;
62	bool			ro_invalid;
63};
64
65/*
66 * These numbers are used by reliable protocols for determining
67 * retransmission behavior and are included in the routing structure.
68 */
69struct rt_metrics {
70	uint64_t rmx_locks;	/* Kernel must leave these values alone */
71	uint64_t rmx_mtu;	/* MTU for this path */
72	uint64_t rmx_hopcount;	/* max hops expected */
73	uint64_t rmx_recvpipe;	/* inbound delay-bandwidth product */
74	uint64_t rmx_sendpipe;	/* outbound delay-bandwidth product */
75	uint64_t rmx_ssthresh;	/* outbound gateway buffer limit */
76	uint64_t rmx_rtt;	/* estimated round trip time */
77	uint64_t rmx_rttvar;	/* estimated rtt variance */
78	time_t	rmx_expire;	/* lifetime for route, e.g. redirect */
79	time_t	rmx_pksent;	/* packets sent using this route */
80};
81
82/*
83 * rmx_rtt and rmx_rttvar are stored as microseconds;
84 * RTTTOPRHZ(rtt) converts to a value suitable for use
85 * by a protocol slowtimo counter.
86 */
87#define	RTM_RTTUNIT	1000000	/* units for rtt, rttvar, as units per sec */
88#define	RTTTOPRHZ(r)	((r) / (RTM_RTTUNIT / PR_SLOWHZ))
89
90/*
91 * We distinguish between routes to hosts and routes to networks,
92 * preferring the former if available.  For each route we infer
93 * the interface to use from the gateway address supplied when
94 * the route was entered.  Routes that forward packets through
95 * gateways are marked so that the output routines know to address the
96 * gateway rather than the ultimate destination.
97 */
98#ifndef RNF_NORMAL
99#include <net/radix.h>
100#endif
101struct rtentry {
102	struct	radix_node rt_nodes[2];	/* tree glue, and other values */
103#define	rt_mask(r)	((const struct sockaddr *)((r)->rt_nodes->rn_mask))
104	struct	sockaddr *rt_gateway;	/* value */
105	int	rt_flags;		/* up/down?, host/net */
106	int	rt_refcnt;		/* # held references */
107	uint64_t rt_use;			/* raw # packets forwarded */
108	struct	ifnet *rt_ifp;		/* the answer: interface to use */
109	struct	ifaddr *rt_ifa;		/* the answer: interface to use */
110	uint32_t rt_ifa_seqno;
111	void *	rt_llinfo;		/* pointer to link level info cache */
112	struct	rt_metrics rt_rmx;	/* metrics used by rx'ing protocols */
113	struct	rtentry *rt_gwroute;	/* implied entry for gatewayed routes */
114	LIST_HEAD(, rttimer) rt_timer;  /* queue of timeouts for misc funcs */
115	struct	rtentry *rt_parent;	/* parent of cloned route */
116	struct	sockaddr *_rt_key;
117	struct	sockaddr *rt_tag;	/* route tagging info */
118};
119
120static inline const struct sockaddr *
121rt_getkey(const struct rtentry *rt)
122{
123	return rt->_rt_key;
124}
125
126/*
127 * Following structure necessary for 4.3 compatibility;
128 * We should eventually move it to a compat file.
129 */
130struct ortentry {
131	uint32_t rt_hash;		/* to speed lookups */
132	struct	sockaddr rt_dst;	/* key */
133	struct	sockaddr rt_gateway;	/* value */
134	int16_t	rt_flags;		/* up/down?, host/net */
135	int16_t	rt_refcnt;		/* # held references */
136	uint32_t rt_use;		/* raw # packets forwarded */
137	struct	ifnet *rt_ifp;		/* the answer: interface to use */
138};
139
140#define	RTF_UP		0x1		/* route usable */
141#define	RTF_GATEWAY	0x2		/* destination is a gateway */
142#define	RTF_HOST	0x4		/* host entry (net otherwise) */
143#define	RTF_REJECT	0x8		/* host or net unreachable */
144#define	RTF_DYNAMIC	0x10		/* created dynamically (by redirect) */
145#define	RTF_MODIFIED	0x20		/* modified dynamically (by redirect) */
146#define RTF_DONE	0x40		/* message confirmed */
147#define RTF_MASK	0x80		/* subnet mask present */
148#define RTF_CLONING	0x100		/* generate new routes on use */
149#define RTF_XRESOLVE	0x200		/* external daemon resolves name */
150#define RTF_LLINFO	0x400		/* generated by ARP or ESIS */
151#define RTF_STATIC	0x800		/* manually added */
152#define RTF_BLACKHOLE	0x1000		/* just discard pkts (during updates) */
153#define	RTF_CLONED	0x2000		/* this is a cloned route */
154#define RTF_PROTO2	0x4000		/* protocol specific routing flag */
155#define RTF_PROTO1	0x8000		/* protocol specific routing flag */
156#define RTF_SRC		0x10000		/* route has fixed source address */
157#define	RTF_ANNOUNCE	0x20000		/* announce new arp or ndp entry */
158
159
160/*
161 * Routing statistics.
162 */
163struct	rtstat {
164	uint64_t rts_badredirect;	/* bogus redirect calls */
165	uint64_t rts_dynamic;		/* routes created by redirects */
166	uint64_t rts_newgateway;	/* routes modified by redirects */
167	uint64_t rts_unreach;		/* lookups which failed */
168	uint64_t rts_wildcard;		/* lookups satisfied by a wildcard */
169};
170
171/*
172 * Structures for routing messages.  By forcing the first member to be aligned
173 * at a 64-bit boundary, we also force the size to be a multiple of 64-bits.
174 */
175
176#if !defined(_KERNEL) || !defined(COMPAT_RTSOCK)
177/*
178 * If we aren't being compiled for backwards compatiblity, enforce 64-bit
179 * alignment so any routing message is the same regardless if the kernel
180 * is an ILP32 or LP64 kernel.
181 */
182#define	__align64	__aligned(sizeof(uint64_t))
183#else
184#define	__align64
185#endif
186
187struct rt_msghdr {
188	u_short	rtm_msglen __align64;
189				/* to skip over non-understood messages */
190	u_char	rtm_version;	/* future binary compatibility */
191	u_char	rtm_type;	/* message type */
192	u_short	rtm_index;	/* index for associated ifp */
193	int	rtm_flags;	/* flags, incl. kern & message, e.g. DONE */
194	int	rtm_addrs;	/* bitmask identifying sockaddrs in msg */
195	pid_t	rtm_pid;	/* identify sender */
196	int	rtm_seq;	/* for sender to identify action */
197	int	rtm_errno;	/* why failed */
198	int	rtm_use;	/* from rtentry */
199	int	rtm_inits;	/* which metrics we are initializing */
200	struct	rt_metrics rtm_rmx __align64;
201				/* metrics themselves */
202};
203
204#undef __align64
205
206#define RTM_VERSION	4	/* Up the ante and ignore older versions */
207
208#define RTM_ADD		0x1	/* Add Route */
209#define RTM_DELETE	0x2	/* Delete Route */
210#define RTM_CHANGE	0x3	/* Change Metrics or flags */
211#define RTM_GET		0x4	/* Report Metrics */
212#define RTM_LOSING	0x5	/* Kernel Suspects Partitioning */
213#define RTM_REDIRECT	0x6	/* Told to use different route */
214#define RTM_MISS	0x7	/* Lookup failed on this address */
215#define RTM_LOCK	0x8	/* fix specified metrics */
216#define RTM_OLDADD	0x9	/* caused by SIOCADDRT */
217#define RTM_OLDDEL	0xa	/* caused by SIOCDELRT */
218#define RTM_RESOLVE	0xb	/* req to resolve dst to LL addr */
219#define RTM_NEWADDR	0xc	/* address being added to iface */
220#define RTM_DELADDR	0xd	/* address being removed from iface */
221#define RTM_OOIFINFO	0xe	/* Old (pre-1.5) RTM_IFINFO message */
222#define RTM_OIFINFO	0xf	/* Old (pre-64bit time) RTM_IFINFO message */
223#define	RTM_IFANNOUNCE	0x10	/* iface arrival/departure */
224#define	RTM_IEEE80211	0x11	/* IEEE80211 wireless event */
225#define	RTM_SETGATE	0x12	/* set prototype gateway for clones
226				 * (see example in arp_rtrequest).
227				 */
228#define	RTM_LLINFO_UPD	0x13	/* indication to ARP/NDP/etc. that link-layer
229				 * address has changed
230				 */
231#define RTM_IFINFO	0x14	/* iface/link going up/down etc. */
232#define RTM_CHGADDR	0x15	/* address properties changed */
233
234#define RTV_MTU		0x1	/* init or lock _mtu */
235#define RTV_HOPCOUNT	0x2	/* init or lock _hopcount */
236#define RTV_EXPIRE	0x4	/* init or lock _expire */
237#define RTV_RPIPE	0x8	/* init or lock _recvpipe */
238#define RTV_SPIPE	0x10	/* init or lock _sendpipe */
239#define RTV_SSTHRESH	0x20	/* init or lock _ssthresh */
240#define RTV_RTT		0x40	/* init or lock _rtt */
241#define RTV_RTTVAR	0x80	/* init or lock _rttvar */
242
243/*
244 * Bitmask values for rtm_addr.
245 */
246#define RTA_DST		0x1	/* destination sockaddr present */
247#define RTA_GATEWAY	0x2	/* gateway sockaddr present */
248#define RTA_NETMASK	0x4	/* netmask sockaddr present */
249#define RTA_GENMASK	0x8	/* cloning mask sockaddr present */
250#define RTA_IFP		0x10	/* interface name sockaddr present */
251#define RTA_IFA		0x20	/* interface addr sockaddr present */
252#define RTA_AUTHOR	0x40	/* sockaddr for author of redirect */
253#define RTA_BRD		0x80	/* for NEWADDR, broadcast or p-p dest addr */
254#define RTA_TAG		0x100	/* route tag */
255
256/*
257 * Index offsets for sockaddr array for alternate internal encoding.
258 */
259#define RTAX_DST	0	/* destination sockaddr present */
260#define RTAX_GATEWAY	1	/* gateway sockaddr present */
261#define RTAX_NETMASK	2	/* netmask sockaddr present */
262#define RTAX_GENMASK	3	/* cloning mask sockaddr present */
263#define RTAX_IFP	4	/* interface name sockaddr present */
264#define RTAX_IFA	5	/* interface addr sockaddr present */
265#define RTAX_AUTHOR	6	/* sockaddr for author of redirect */
266#define RTAX_BRD	7	/* for NEWADDR, broadcast or p-p dest addr */
267#define RTAX_TAG	8	/* route tag */
268#define RTAX_MAX	9	/* size of array to allocate */
269
270#define RT_ROUNDUP2(a, n)	((a) > 0 ? (1 + (((a) - 1) | ((n) - 1))) : (n))
271#define RT_ROUNDUP(a)		RT_ROUNDUP2((a), sizeof(uint64_t))
272#define RT_ADVANCE(x, n)	(x += RT_ROUNDUP((n)->sa_len))
273
274struct rt_addrinfo {
275	int	rti_addrs;
276	const struct	sockaddr *rti_info[RTAX_MAX];
277	int	rti_flags;
278	struct	ifaddr *rti_ifa;
279	struct	ifnet *rti_ifp;
280};
281
282struct route_cb {
283	int	ip_count;
284	int	ip6_count;
285	int	iso_count;
286	int	mpls_count;
287	int	any_count;
288};
289
290/*
291 * This structure, and the prototypes for the rt_timer_{init,remove_all,
292 * add,timer} functions all used with the kind permission of BSDI.
293 * These allow functions to be called for routes at specific times.
294 */
295
296struct rttimer {
297	TAILQ_ENTRY(rttimer)	rtt_next;  /* entry on timer queue */
298	LIST_ENTRY(rttimer) 	rtt_link;  /* multiple timers per rtentry */
299	struct rttimer_queue   *rtt_queue; /* back pointer to queue */
300	struct rtentry         *rtt_rt;    /* Back pointer to the route */
301	void		      (*rtt_func)(struct rtentry *, struct rttimer *);
302	time_t          	rtt_time;  /* When this timer was registered */
303};
304
305struct rttimer_queue {
306	long				rtq_timeout;
307	unsigned long			rtq_count;
308	TAILQ_HEAD(, rttimer)		rtq_head;
309	LIST_ENTRY(rttimer_queue)	rtq_link;
310};
311
312
313struct rtbl;
314typedef struct rtbl rtbl_t;
315
316#ifdef _KERNEL
317
318struct rtbl {
319	struct radix_node_head t_rnh;
320};
321
322struct rt_walkarg {
323	int	w_op;
324	int	w_arg;
325	int	w_given;
326	int	w_needed;
327	void *	w_where;
328	int	w_tmemsize;
329	int	w_tmemneeded;
330	void *	w_tmem;
331};
332
333#if 0
334#define	RT_DPRINTF(__fmt, ...)	do { } while (/*CONSTCOND*/0)
335#else
336#define	RT_DPRINTF(__fmt, ...)	/* do nothing */
337#endif
338
339struct rtwalk {
340	int (*rw_f)(struct rtentry *, void *);
341	void *rw_v;
342};
343
344/*
345 * Global data specific to the routing socket.
346 */
347struct route_info {
348	struct sockaddr ri_dst;
349	struct sockaddr ri_src;
350	struct route_cb ri_cb;
351	int ri_maxqlen;
352	struct ifqueue ri_intrq;
353	void *ri_sih;
354};
355
356extern	struct	route_info route_info;
357extern	struct	rtstat	rtstat;
358
359struct socket;
360struct dom_rtlist;
361
362void	 rt_init(void);
363void	 rt_ifannouncemsg(struct ifnet *, int);
364void	 rt_ieee80211msg(struct ifnet *, int, void *, size_t);
365void	 rt_ifmsg(struct ifnet *);
366void	 rt_missmsg(int, const struct rt_addrinfo *, int, int);
367struct mbuf *rt_msg1(int, struct rt_addrinfo *, void *, int);
368void	 rt_newaddrmsg(int, struct ifaddr *, int, struct rtentry *);
369
370void	 rt_maskedcopy(const struct sockaddr *,
371	    struct sockaddr *, const struct sockaddr *);
372int	 rt_setgate(struct rtentry *, const struct sockaddr *);
373int      rt_timer_add(struct rtentry *,
374             void(*)(struct rtentry *, struct rttimer *),
375	     struct rttimer_queue *);
376void	 rt_timer_init(void);
377struct rttimer_queue *
378	 rt_timer_queue_create(u_int);
379void	 rt_timer_queue_change(struct rttimer_queue *, long);
380void	 rt_timer_queue_remove_all(struct rttimer_queue *, int);
381void	 rt_timer_queue_destroy(struct rttimer_queue *, int);
382void	 rt_timer_remove_all(struct rtentry *, int);
383unsigned long	rt_timer_count(struct rttimer_queue *);
384void	 rt_timer_timer(void *);
385void	 rtcache(struct route *);
386void	 rtflushall(int);
387struct rtentry *
388	 rtalloc1(const struct sockaddr *, int);
389void	 rtfree(struct rtentry *);
390int	 rt_getifa(struct rt_addrinfo *);
391int	 rtinit(struct ifaddr *, int, int);
392int	 rtioctl(u_long, void *, struct lwp *);
393void	 rtredirect(const struct sockaddr *, const struct sockaddr *,
394	    const struct sockaddr *, int, const struct sockaddr *,
395	    struct rtentry **);
396int	 rtrequest(int, const struct sockaddr *,
397	    const struct sockaddr *, const struct sockaddr *, int,
398	    struct rtentry **);
399int	 rtrequest1(int, struct rt_addrinfo *, struct rtentry **);
400
401struct ifaddr	*rt_get_ifa(struct rtentry *);
402void	rt_replace_ifa(struct rtentry *, struct ifaddr *);
403
404const struct sockaddr *rt_settag(struct rtentry *, const struct sockaddr *);
405struct sockaddr *rt_gettag(struct rtentry *);
406
407static inline void
408rt_destroy(struct rtentry *rt)
409{
410	if (rt->_rt_key != NULL)
411		sockaddr_free(rt->_rt_key);
412	if (rt->rt_gateway != NULL)
413		sockaddr_free(rt->rt_gateway);
414	if (rt_gettag(rt) != NULL)
415		sockaddr_free(rt_gettag(rt));
416	rt->_rt_key = rt->rt_gateway = rt->rt_tag = NULL;
417}
418
419static inline const struct sockaddr *
420rt_setkey(struct rtentry *rt, const struct sockaddr *key, int flags)
421{
422	if (rt->_rt_key == key)
423		goto out;
424
425	if (rt->_rt_key != NULL)
426		sockaddr_free(rt->_rt_key);
427	rt->_rt_key = sockaddr_dup(key, flags);
428out:
429	KASSERT(rt->_rt_key != NULL);
430	rt->rt_nodes->rn_key = (const char *)rt->_rt_key;
431	return rt->_rt_key;
432}
433
434struct rtentry *rtcache_init(struct route *);
435struct rtentry *rtcache_init_noclone(struct route *);
436void	rtcache_copy(struct route *, const struct route *);
437void rtcache_invalidate(struct dom_rtlist *);
438
439struct rtentry *rtcache_lookup2(struct route *, const struct sockaddr *, int,
440    int *);
441void	rtcache_clear(struct route *);
442struct rtentry *rtcache_update(struct route *, int);
443void	rtcache_free(struct route *);
444int	rtcache_setdst(struct route *, const struct sockaddr *);
445
446static inline void
447rtcache_invariants(const struct route *ro)
448{
449	KASSERT(ro->ro_sa != NULL || ro->_ro_rt == NULL);
450	KASSERT(!ro->ro_invalid || ro->_ro_rt != NULL);
451}
452
453static inline struct rtentry *
454rtcache_lookup1(struct route *ro, const struct sockaddr *dst, int clone)
455{
456	int hit;
457
458	return rtcache_lookup2(ro, dst, clone, &hit);
459}
460
461static inline struct rtentry *
462rtcache_lookup_noclone(struct route *ro, const struct sockaddr *dst)
463{
464	return rtcache_lookup1(ro, dst, 0);
465}
466
467static inline struct rtentry *
468rtcache_lookup(struct route *ro, const struct sockaddr *dst)
469{
470	return rtcache_lookup1(ro, dst, 1);
471}
472
473static inline const struct sockaddr *
474rtcache_getdst(const struct route *ro)
475{
476	rtcache_invariants(ro);
477	return ro->ro_sa;
478}
479
480/* If the cache is not empty, and the cached route is still present
481 * in the routing table, return the cached route.  Otherwise, return
482 * NULL.
483 */
484static inline struct rtentry *
485rtcache_validate(const struct route *ro)
486{
487	struct rtentry *rt = ro->_ro_rt;
488
489	rtcache_invariants(ro);
490
491	if (ro->ro_invalid)
492		return NULL;
493
494	if (rt != NULL && (rt->rt_flags & RTF_UP) != 0 && rt->rt_ifp != NULL)
495		return rt;
496	return NULL;
497
498}
499
500static inline void
501RTFREE(struct rtentry *rt)
502{
503	if (rt->rt_refcnt <= 1)
504		rtfree(rt);
505	else
506		rt->rt_refcnt--;
507}
508
509int rt_walktree(sa_family_t, int (*)(struct rtentry *, void *), void *);
510void route_enqueue(struct mbuf *, int);
511int rt_inithead(rtbl_t **, int);
512struct rtentry *rt_matchaddr(rtbl_t *, const struct sockaddr *);
513int rt_addaddr(rtbl_t *, struct rtentry *, const struct sockaddr *);
514struct rtentry *rt_lookup(rtbl_t *, const struct sockaddr *,
515    const struct sockaddr *);
516struct rtentry *rt_deladdr(rtbl_t *, const struct sockaddr *,
517    const struct sockaddr *);
518void rtbl_init(void);
519rtbl_t *rt_gettable(sa_family_t);
520void rt_assert_inactive(const struct rtentry *);
521
522#endif /* _KERNEL */
523
524#endif /* !_NET_ROUTE_H_ */
525