route.h revision 1.73
1/*	$NetBSD: route.h,v 1.73 2009/04/02 21:02:06 christos 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	u_long	rmx_locks;	/* Kernel must leave these values alone */
71	u_long	rmx_mtu;	/* MTU for this path */
72	u_long	rmx_hopcount;	/* max hops expected */
73	u_long	rmx_expire;	/* lifetime for route, e.g. redirect */
74	u_long	rmx_recvpipe;	/* inbound delay-bandwidth product */
75	u_long	rmx_sendpipe;	/* outbound delay-bandwidth product */
76	u_long	rmx_ssthresh;	/* outbound gateway buffer limit */
77	u_long	rmx_rtt;	/* estimated round trip time */
78	u_long	rmx_rttvar;	/* estimated rtt variance */
79	u_long	rmx_pksent;	/* packets sent using this route */
80};
81
82struct nrt_metrics {
83	u_long	rmx_locks;	/* Kernel must leave these values alone */
84	u_long	rmx_mtu;	/* MTU for this path */
85	u_long	rmx_hopcount;	/* max hops expected */
86	u_long	rmx_recvpipe;	/* inbound delay-bandwidth product */
87	u_long	rmx_sendpipe;	/* outbound delay-bandwidth product */
88	u_long	rmx_ssthresh;	/* outbound gateway buffer limit */
89	u_long	rmx_rtt;	/* estimated round trip time */
90	u_long	rmx_rttvar;	/* estimated rtt variance */
91	time_t	rmx_expire;	/* lifetime for route, e.g. redirect */
92	time_t	rmx_pksent;	/* packets sent using this route */
93};
94
95/*
96 * rmx_rtt and rmx_rttvar are stored as microseconds;
97 * RTTTOPRHZ(rtt) converts to a value suitable for use
98 * by a protocol slowtimo counter.
99 */
100#define	RTM_RTTUNIT	1000000	/* units for rtt, rttvar, as units per sec */
101#define	RTTTOPRHZ(r)	((r) / (RTM_RTTUNIT / PR_SLOWHZ))
102
103/*
104 * We distinguish between routes to hosts and routes to networks,
105 * preferring the former if available.  For each route we infer
106 * the interface to use from the gateway address supplied when
107 * the route was entered.  Routes that forward packets through
108 * gateways are marked so that the output routines know to address the
109 * gateway rather than the ultimate destination.
110 */
111#ifndef RNF_NORMAL
112#include <net/radix.h>
113#endif
114struct rtentry {
115	struct	radix_node rt_nodes[2];	/* tree glue, and other values */
116#define	rt_mask(r)	((const struct sockaddr *)((r)->rt_nodes->rn_mask))
117	struct	sockaddr *rt_gateway;	/* value */
118	int	rt_flags;		/* up/down?, host/net */
119	int	rt_refcnt;		/* # held references */
120	u_long	rt_use;			/* raw # packets forwarded */
121	struct	ifnet *rt_ifp;		/* the answer: interface to use */
122	struct	ifaddr *rt_ifa;		/* the answer: interface to use */
123	uint32_t rt_ifa_seqno;
124	void *	rt_llinfo;		/* pointer to link level info cache */
125	struct	nrt_metrics rt_rmx;	/* metrics used by rx'ing protocols */
126	struct	rtentry *rt_gwroute;	/* implied entry for gatewayed routes */
127	LIST_HEAD(, rttimer) rt_timer;  /* queue of timeouts for misc funcs */
128	struct	rtentry *rt_parent;	/* parent of cloned route */
129	struct sockaddr *_rt_key;
130};
131
132static inline const struct sockaddr *
133rt_getkey(const struct rtentry *rt)
134{
135	return rt->_rt_key;
136}
137
138/*
139 * Following structure necessary for 4.3 compatibility;
140 * We should eventually move it to a compat file.
141 */
142struct ortentry {
143	uint32_t rt_hash;		/* to speed lookups */
144	struct	sockaddr rt_dst;	/* key */
145	struct	sockaddr rt_gateway;	/* value */
146	int16_t	rt_flags;		/* up/down?, host/net */
147	int16_t	rt_refcnt;		/* # held references */
148	uint32_t rt_use;		/* raw # packets forwarded */
149	struct	ifnet *rt_ifp;		/* the answer: interface to use */
150};
151
152#define	RTF_UP		0x1		/* route usable */
153#define	RTF_GATEWAY	0x2		/* destination is a gateway */
154#define	RTF_HOST	0x4		/* host entry (net otherwise) */
155#define	RTF_REJECT	0x8		/* host or net unreachable */
156#define	RTF_DYNAMIC	0x10		/* created dynamically (by redirect) */
157#define	RTF_MODIFIED	0x20		/* modified dynamically (by redirect) */
158#define RTF_DONE	0x40		/* message confirmed */
159#define RTF_MASK	0x80		/* subnet mask present */
160#define RTF_CLONING	0x100		/* generate new routes on use */
161#define RTF_XRESOLVE	0x200		/* external daemon resolves name */
162#define RTF_LLINFO	0x400		/* generated by ARP or ESIS */
163#define RTF_STATIC	0x800		/* manually added */
164#define RTF_BLACKHOLE	0x1000		/* just discard pkts (during updates) */
165#define	RTF_CLONED	0x2000		/* this is a cloned route */
166#define RTF_PROTO2	0x4000		/* protocol specific routing flag */
167#define RTF_PROTO1	0x8000		/* protocol specific routing flag */
168#define RTF_SRC		0x10000		/* route has fixed source address */
169
170
171/*
172 * Routing statistics.
173 */
174struct	rtstat {
175	u_quad_t rts_badredirect;	/* bogus redirect calls */
176	u_quad_t rts_dynamic;		/* routes created by redirects */
177	u_quad_t rts_newgateway;	/* routes modified by redirects */
178	u_quad_t rts_unreach;		/* lookups which failed */
179	u_quad_t rts_wildcard;		/* lookups satisfied by a wildcard */
180};
181/*
182 * Structures for routing messages.
183 */
184struct rt_msghdr {
185	u_short	rtm_msglen;	/* to skip over non-understood messages */
186	u_char	rtm_version;	/* future binary compatibility */
187	u_char	rtm_type;	/* message type */
188	u_short	rtm_index;	/* index for associated ifp */
189	int	rtm_flags;	/* flags, incl. kern & message, e.g. DONE */
190	int	rtm_addrs;	/* bitmask identifying sockaddrs in msg */
191	pid_t	rtm_pid;	/* identify sender */
192	int	rtm_seq;	/* for sender to identify action */
193	int	rtm_errno;	/* why failed */
194	int	rtm_use;	/* from rtentry */
195	u_long	rtm_inits;	/* which metrics we are initializing */
196	struct	rt_metrics rtm_rmx; /* metrics themselves */
197};
198
199#define RTM_VERSION	3	/* Up the ante and ignore older versions */
200
201#define RTM_ADD		0x1	/* Add Route */
202#define RTM_DELETE	0x2	/* Delete Route */
203#define RTM_CHANGE	0x3	/* Change Metrics or flags */
204#define RTM_GET		0x4	/* Report Metrics */
205#define RTM_LOSING	0x5	/* Kernel Suspects Partitioning */
206#define RTM_REDIRECT	0x6	/* Told to use different route */
207#define RTM_MISS	0x7	/* Lookup failed on this address */
208#define RTM_LOCK	0x8	/* fix specified metrics */
209#define RTM_OLDADD	0x9	/* caused by SIOCADDRT */
210#define RTM_OLDDEL	0xa	/* caused by SIOCDELRT */
211#define RTM_RESOLVE	0xb	/* req to resolve dst to LL addr */
212#define RTM_NEWADDR	0xc	/* address being added to iface */
213#define RTM_DELADDR	0xd	/* address being removed from iface */
214#define RTM_OOIFINFO	0xe	/* Old (pre-1.5) RTM_IFINFO message */
215#define RTM_OIFINFO	0xf	/* Old (pre-6.0) RTM_IFINFO message */
216#define	RTM_IFANNOUNCE	0x10	/* iface arrival/departure */
217#define	RTM_IEEE80211	0x11	/* IEEE80211 wireless event */
218#define	RTM_SETGATE	0x12	/* set prototype gateway for clones
219				 * (see example in arp_rtrequest).
220				 */
221#define	RTM_LLINFO_UPD	0x13	/* indication to ARP/NDP/etc. that link-layer
222				 * address has changed
223				 */
224#define RTM_IFINFO	0x14	/* iface/link going up/down etc. */
225
226#define RTV_MTU		0x1	/* init or lock _mtu */
227#define RTV_HOPCOUNT	0x2	/* init or lock _hopcount */
228#define RTV_EXPIRE	0x4	/* init or lock _expire */
229#define RTV_RPIPE	0x8	/* init or lock _recvpipe */
230#define RTV_SPIPE	0x10	/* init or lock _sendpipe */
231#define RTV_SSTHRESH	0x20	/* init or lock _ssthresh */
232#define RTV_RTT		0x40	/* init or lock _rtt */
233#define RTV_RTTVAR	0x80	/* init or lock _rttvar */
234
235/*
236 * Bitmask values for rtm_addr.
237 */
238#define RTA_DST		0x1	/* destination sockaddr present */
239#define RTA_GATEWAY	0x2	/* gateway sockaddr present */
240#define RTA_NETMASK	0x4	/* netmask sockaddr present */
241#define RTA_GENMASK	0x8	/* cloning mask sockaddr present */
242#define RTA_IFP		0x10	/* interface name sockaddr present */
243#define RTA_IFA		0x20	/* interface addr sockaddr present */
244#define RTA_AUTHOR	0x40	/* sockaddr for author of redirect */
245#define RTA_BRD		0x80	/* for NEWADDR, broadcast or p-p dest addr */
246
247/*
248 * Index offsets for sockaddr array for alternate internal encoding.
249 */
250#define RTAX_DST	0	/* destination sockaddr present */
251#define RTAX_GATEWAY	1	/* gateway sockaddr present */
252#define RTAX_NETMASK	2	/* netmask sockaddr present */
253#define RTAX_GENMASK	3	/* cloning mask sockaddr present */
254#define RTAX_IFP	4	/* interface name sockaddr present */
255#define RTAX_IFA	5	/* interface addr sockaddr present */
256#define RTAX_AUTHOR	6	/* sockaddr for author of redirect */
257#define RTAX_BRD	7	/* for NEWADDR, broadcast or p-p dest addr */
258#define RTAX_MAX	8	/* size of array to allocate */
259
260#define RT_ROUNDUP(a) \
261	((a) > 0 ? (1 + (((a) - 1) | (sizeof(long) - 1))) : sizeof(long))
262#define RT_ADVANCE(x, n) (x += RT_ROUNDUP((n)->sa_len))
263
264struct rt_addrinfo {
265	int	rti_addrs;
266	const struct	sockaddr *rti_info[RTAX_MAX];
267	int	rti_flags;
268	struct	ifaddr *rti_ifa;
269	struct	ifnet *rti_ifp;
270	struct	rt_msghdr *rti_rtm;
271};
272
273struct route_cb {
274	int	ip_count;
275	int	ip6_count;
276	int	ipx_count;
277	int	ns_count;
278	int	iso_count;
279	int	any_count;
280};
281
282/*
283 * This structure, and the prototypes for the rt_timer_{init,remove_all,
284 * add,timer} functions all used with the kind permission of BSDI.
285 * These allow functions to be called for routes at specific times.
286 */
287
288struct rttimer {
289	TAILQ_ENTRY(rttimer)	rtt_next;  /* entry on timer queue */
290	LIST_ENTRY(rttimer) 	rtt_link;  /* multiple timers per rtentry */
291	struct rttimer_queue   *rtt_queue; /* back pointer to queue */
292	struct rtentry         *rtt_rt;    /* Back pointer to the route */
293	void		      (*rtt_func)(struct rtentry *, struct rttimer *);
294	time_t          	rtt_time;  /* When this timer was registered */
295};
296
297struct rttimer_queue {
298	long				rtq_timeout;
299	unsigned long			rtq_count;
300	TAILQ_HEAD(, rttimer)		rtq_head;
301	LIST_ENTRY(rttimer_queue)	rtq_link;
302};
303
304
305#ifdef _KERNEL
306
307extern struct	sockaddr route_dst;
308extern struct	sockaddr route_src;
309extern struct	sockproto route_proto;
310
311struct rt_walkarg {
312	int	w_op;
313	int	w_arg;
314	int	w_given;
315	int	w_needed;
316	void *	w_where;
317	int	w_tmemsize;
318	int	w_tmemneeded;
319	void *	w_tmem;
320};
321
322#if 0
323#define	RT_DPRINTF(__fmt, ...)	do { } while (/*CONSTCOND*/0)
324#else
325#define	RT_DPRINTF(__fmt, ...)	/* do nothing */
326#endif
327
328struct rtwalk {
329	int (*rw_f)(struct rtentry *, void *);
330	void *rw_v;
331};
332extern	struct	route_cb route_cb;
333extern	struct	rtstat	rtstat;
334extern	struct	radix_node_head *rt_tables[AF_MAX+1];
335
336struct socket;
337struct dom_rtlist;
338
339void	 route_init(void);
340int	 route_output(struct mbuf *, ...);
341int	 route_usrreq(struct socket *,
342	    int, struct mbuf *, struct mbuf *, struct mbuf *, struct lwp *);
343void	 rt_init(void);
344void	 rt_ifannouncemsg(struct ifnet *, int);
345void	 rt_ieee80211msg(struct ifnet *, int, void *, size_t);
346void	 rt_ifmsg(struct ifnet *);
347void	 rt_maskedcopy(const struct sockaddr *,
348	    struct sockaddr *, const struct sockaddr *);
349void	 rt_missmsg(int, struct rt_addrinfo *, int, int);
350struct mbuf *rt_msg1(int, struct rt_addrinfo *, void *, int);
351void	 rt_newaddrmsg(int, struct ifaddr *, int, struct rtentry *);
352int	 rt_setgate(struct rtentry *, const struct sockaddr *);
353void	 rt_setmetrics(u_long, const struct rt_metrics *, struct nrt_metrics *);
354int      rt_timer_add(struct rtentry *,
355             void(*)(struct rtentry *, struct rttimer *),
356	     struct rttimer_queue *);
357void	 rt_timer_init(void);
358struct rttimer_queue *
359	 rt_timer_queue_create(u_int);
360void	 rt_timer_queue_change(struct rttimer_queue *, long);
361void	 rt_timer_queue_remove_all(struct rttimer_queue *, int);
362void	 rt_timer_queue_destroy(struct rttimer_queue *, int);
363void	 rt_timer_remove_all(struct rtentry *, int);
364unsigned long	rt_timer_count(struct rttimer_queue *);
365void	 rt_timer_timer(void *);
366void	 rtable_init(void **);
367void	 rtcache(struct route *);
368void	 rtflushall(int);
369struct rtentry *
370	 rtalloc1(const struct sockaddr *, int);
371void	 rtfree(struct rtentry *);
372int	 rt_getifa(struct rt_addrinfo *);
373int	 rtinit(struct ifaddr *, int, int);
374int	 rtioctl(u_long, void *, struct lwp *);
375void	 rtredirect(const struct sockaddr *, const struct sockaddr *,
376	    const struct sockaddr *, int, const struct sockaddr *,
377	    struct rtentry **);
378int	 rtrequest(int, const struct sockaddr *,
379	    const struct sockaddr *, const struct sockaddr *, int,
380	    struct rtentry **);
381int	 rtrequest1(int, struct rt_addrinfo *, struct rtentry **);
382
383struct ifaddr	*rt_get_ifa(struct rtentry *);
384void	rt_replace_ifa(struct rtentry *, struct ifaddr *);
385
386static inline void
387rt_destroy(struct rtentry *rt)
388{
389	if (rt->_rt_key != NULL)
390		sockaddr_free(rt->_rt_key);
391	if (rt->rt_gateway != NULL)
392		sockaddr_free(rt->rt_gateway);
393	rt->_rt_key = rt->rt_gateway = NULL;
394}
395
396static inline const struct sockaddr *
397rt_setkey(struct rtentry *rt, const struct sockaddr *key, int flags)
398{
399	if (rt->_rt_key == key)
400		goto out;
401
402	if (rt->_rt_key != NULL)
403		sockaddr_free(rt->_rt_key);
404	rt->_rt_key = sockaddr_dup(key, flags);
405out:
406	KASSERT(rt->_rt_key != NULL);
407	rt->rt_nodes->rn_key = (const char *)rt->_rt_key;
408	return rt->_rt_key;
409}
410
411struct rtentry *rtfindparent(struct radix_node_head *, struct route *);
412
413struct rtentry *rtcache_init(struct route *);
414struct rtentry *rtcache_init_noclone(struct route *);
415void	rtcache_copy(struct route *, const struct route *);
416void rtcache_invalidate(struct dom_rtlist *);
417
418struct rtentry *rtcache_lookup2(struct route *, const struct sockaddr *, int,
419    int *);
420void	rtcache_clear(struct route *);
421struct rtentry *rtcache_update(struct route *, int);
422void	rtcache_free(struct route *);
423int	rtcache_setdst(struct route *, const struct sockaddr *);
424
425static inline void
426rtcache_invariants(const struct route *ro)
427{
428	KASSERT(ro->ro_sa != NULL || ro->_ro_rt == NULL);
429	KASSERT(!ro->ro_invalid || ro->_ro_rt != NULL);
430}
431
432static inline struct rtentry *
433rtcache_lookup1(struct route *ro, const struct sockaddr *dst, int clone)
434{
435	int hit;
436
437	return rtcache_lookup2(ro, dst, clone, &hit);
438}
439
440static inline struct rtentry *
441rtcache_lookup_noclone(struct route *ro, const struct sockaddr *dst)
442{
443	return rtcache_lookup1(ro, dst, 0);
444}
445
446static inline struct rtentry *
447rtcache_lookup(struct route *ro, const struct sockaddr *dst)
448{
449	return rtcache_lookup1(ro, dst, 1);
450}
451
452static inline const struct sockaddr *
453rtcache_getdst(const struct route *ro)
454{
455	rtcache_invariants(ro);
456	return ro->ro_sa;
457}
458
459/* If the cache is not empty, and the cached route is still present
460 * in the routing table, return the cached route.  Otherwise, return
461 * NULL.
462 */
463static inline struct rtentry *
464rtcache_validate(const struct route *ro)
465{
466	struct rtentry *rt = ro->_ro_rt;
467
468	rtcache_invariants(ro);
469
470	if (ro->ro_invalid)
471		return NULL;
472
473	if (rt != NULL && (rt->rt_flags & RTF_UP) != 0 && rt->rt_ifp != NULL)
474		return rt;
475	return NULL;
476
477}
478
479static inline void
480RTFREE(struct rtentry *rt)
481{
482	if (rt->rt_refcnt <= 1)
483		rtfree(rt);
484	else
485		rt->rt_refcnt--;
486}
487
488int
489rt_walktree(sa_family_t, int (*)(struct rtentry *, void *), void *);
490void route_enqueue(struct mbuf *, int);
491
492#endif /* _KERNEL */
493#endif /* !_NET_ROUTE_H_ */
494