xref: /dragonfly/sys/net/route.h (revision 6b47f3ea)
1 /*
2  * Copyright (c) 2004, 2005 The DragonFly Project.  All rights reserved.
3  *
4  * This code is derived from software contributed to The DragonFly Project
5  * by Jeffrey M. Hsu.
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 DragonFly Project nor the names of its
16  *    contributors may be used to endorse or promote products derived
17  *    from this software without specific, prior written permission.
18  *
19  * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
20  * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
21  * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
22  * FOR A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE
23  * COPYRIGHT HOLDERS OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
24  * INCIDENTAL, SPECIAL, EXEMPLARY OR CONSEQUENTIAL DAMAGES (INCLUDING,
25  * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
26  * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED
27  * AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
28  * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT
29  * OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
30  * SUCH DAMAGE.
31  */
32 
33 /*
34  * Copyright (c) 1980, 1986, 1993
35  *	The Regents of the University of California.  All rights reserved.
36  *
37  * Redistribution and use in source and binary forms, with or without
38  * modification, are permitted provided that the following conditions
39  * are met:
40  * 1. Redistributions of source code must retain the above copyright
41  *    notice, this list of conditions and the following disclaimer.
42  * 2. Redistributions in binary form must reproduce the above copyright
43  *    notice, this list of conditions and the following disclaimer in the
44  *    documentation and/or other materials provided with the distribution.
45  * 3. Neither the name of the University nor the names of its contributors
46  *    may be used to endorse or promote products derived from this software
47  *    without specific prior written permission.
48  *
49  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
50  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
51  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
52  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
53  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
54  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
55  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
56  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
57  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
58  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
59  * SUCH DAMAGE.
60  *
61  *	@(#)route.h	8.4 (Berkeley) 1/9/95
62  * $FreeBSD: src/sys/net/route.h,v 1.36.2.5 2002/02/01 11:48:01 ru Exp $
63  */
64 
65 #ifndef _NET_ROUTE_H_
66 #define _NET_ROUTE_H_
67 
68 #ifndef _SYS_TYPES_H_
69 #include <sys/types.h>
70 #endif
71 #ifndef _SYS_PARAM_H_
72 #include <sys/param.h>
73 #endif
74 #ifndef _SYS_SOCKET_H_
75 #include <sys/socket.h>
76 #endif
77 
78 #include <net/radix.h>
79 
80 /*
81  * Kernel resident routing tables.
82  *
83  * The routing tables are initialized when interface addresses
84  * are set by making entries for all directly connected interfaces.
85  */
86 
87 /*
88  * A route consists of a destination address and a reference
89  * to a routing entry.  These are often held by protocols
90  * in their control blocks, e.g. inpcb.
91  */
92 struct route {
93 	struct	rtentry *ro_rt;
94 	struct	sockaddr ro_dst;
95 };
96 
97 /*
98  * These numbers are used by reliable protocols for determining
99  * retransmission behavior and are included in the routing structure.
100  */
101 struct rt_metrics {	/* grouped for locality of reference */
102 	u_long	rmx_locks;	/* Kernel must leave these values alone */
103 	u_long	rmx_mtu;	/* MTU for this path */
104 	u_long	rmx_pksent;	/* packets sent using this route */
105 	u_long	rmx_expire;	/* lifetime for route */
106 
107 	u_long	rmx_sendpipe;	/* outbound delay-bandwidth product */
108 	u_long	rmx_ssthresh;	/* outbound gateway buffer limit */
109 	u_long	rmx_rtt;	/* estimated round trip time */
110 	u_long	rmx_rttvar;	/* estimated rtt variance */
111 
112 	u_long	rmx_recvpipe;	/* inbound delay-bandwidth product */
113 	u_long	rmx_hopcount;	/* max hops expected */
114 	u_short rmx_mssopt;	/* peer's cached MSS */
115 	u_short	rmx_pad;	/* explicit pad */
116 	u_long	rmx_msl;	/* maximum segment lifetime, unit: ms */
117 	u_long	rmx_iwmaxsegs;	/* IW segments max */
118 	u_long	rmx_iwcapsegs;	/* IW segments */
119 };
120 
121 /*
122  * rmx_rtt and rmx_rttvar are stored as microseconds;
123  * RTTTOPRHZ(rtt) converts to a value suitable for use
124  * by a protocol slowtimo counter.
125  */
126 #define	RTM_RTTUNIT	1000000	/* units for rtt, rttvar, as units per sec */
127 #define	RTTTOPRHZ(r)	((r) / (RTM_RTTUNIT / PR_SLOWHZ))
128 
129 /*
130  * XXX kernel function pointer `rt_output' is visible to applications.
131  */
132 struct mbuf;
133 
134 /*
135  * We distinguish between routes to hosts and routes to networks,
136  * preferring the former if available.  For each route we infer
137  * the interface to use from the gateway address supplied when
138  * the route was entered.  Routes that forward packets through
139  * gateways are marked so that the output routines know to address the
140  * gateway rather than the ultimate destination.
141  */
142 struct rtentry {
143 	struct	radix_node rt_nodes[2];	/* tree glue, and other values */
144 #define	rt_key(r)	__DECONST(struct sockaddr *, (r)->rt_nodes->rn_key)
145 #define	rt_mask(r)	__DECONST(struct sockaddr *, (r)->rt_nodes->rn_mask)
146 	struct	sockaddr *rt_gateway;	/* value */
147 	long	rt_refcnt;		/* # held references */
148 	u_long	rt_flags;		/* up/down?, host/net */
149 	struct	ifnet *rt_ifp;		/* the answer: interface to use */
150 	struct	ifaddr *rt_ifa;		/* the answer: interface to use */
151 	struct	sockaddr *rt_genmask;	/* for generation of cloned routes */
152 	void	*rt_llinfo;		/* pointer to link level info cache */
153 	struct	rt_metrics rt_rmx;	/* metrics used by rx'ing protocols */
154 	struct	rtentry *rt_gwroute;	/* implied entry for gatewayed routes */
155 	int	(*rt_output) (struct ifnet *, struct mbuf *, struct sockaddr *,
156 			      struct rtentry *);
157 					/* output routine for this (rt,if) */
158 	struct	rtentry *rt_parent;	/* cloning parent of this route */
159 	int	rt_cpuid;		/* owner cpu */
160 	struct	sockaddr *rt_shim[3];	/* mpls label / operation array */
161 };
162 
163 #define rt_use rt_rmx.rmx_pksent
164 
165 #define	RTF_UP		0x1		/* route usable */
166 #define	RTF_GATEWAY	0x2		/* destination is a gateway */
167 #define	RTF_HOST	0x4		/* host entry (net otherwise) */
168 #define	RTF_REJECT	0x8		/* host or net unreachable */
169 #define	RTF_DYNAMIC	0x10		/* created dynamically (by redirect) */
170 #define	RTF_MODIFIED	0x20		/* modified dynamically (by redirect) */
171 #define RTF_DONE	0x40		/* message confirmed */
172 /*			0x80		   unused, was RTF_DELCLONE */
173 #define RTF_CLONING	0x100		/* generate new routes on use */
174 #define RTF_XRESOLVE	0x200		/* external daemon resolves name */
175 #define RTF_LLINFO	0x400		/* generated by link layer (e.g. ARP) */
176 #define RTF_STATIC	0x800		/* manually added */
177 #define RTF_BLACKHOLE	0x1000		/* just discard pkts (during updates) */
178 /*			0x2000		   unused */
179 #define RTF_PROTO2	0x4000		/* protocol specific routing flag */
180 #define RTF_PROTO1	0x8000		/* protocol specific routing flag */
181 
182 #define RTF_PRCLONING	0x10000		/* protocol requires cloning */
183 #define RTF_WASCLONED	0x20000		/* route generated through cloning */
184 #define RTF_PROTO3	0x40000		/* protocol specific routing flag */
185 /*			0x80000		   unused */
186 #define RTF_PINNED	0x100000	/* future use */
187 #define	RTF_LOCAL	0x200000	/* route represents a local address */
188 #define	RTF_BROADCAST	0x400000	/* route represents a bcast address */
189 #define	RTF_MULTICAST	0x800000	/* route represents a mcast address */
190 #define	RTF_MPLSOPS	0x1000000	/* route uses mpls label operations */
191 					/* 0x2000000 and up unassigned */
192 
193 /*
194  * Routing statistics.
195  */
196 struct rtstatistics {
197 	u_long	rts_badredirect;	/* bogus redirect calls */
198 	u_long	rts_dynamic;		/* routes created by redirects */
199 	u_long	rts_newgateway;		/* routes modified by redirects */
200 	u_long	rts_unreach;		/* lookups which failed */
201 	u_long	rts_wildcard;		/* lookups satisfied by a wildcard */
202 	u_long	rts_pad[3];		/* pad to cache line size (64B) */
203 };
204 #ifdef _KERNEL
205 CTASSERT((sizeof(struct rtstatistics) & __VM_CACHELINE_MASK) == 0);
206 #endif
207 
208 /*
209  * Structures for routing messages.
210  */
211 struct rt_msghdr {
212 	u_short	rtm_msglen;	/* to skip over non-understood messages */
213 	u_char	rtm_version;	/* future binary compatibility */
214 	u_char	rtm_type;	/* message type */
215 	u_short	rtm_index;	/* index for associated ifp */
216 	int	rtm_flags;	/* flags, incl. kern & message, e.g. DONE */
217 	int	rtm_addrs;	/* bitmask identifying sockaddrs in msg */
218 	pid_t	rtm_pid;	/* identify sender */
219 	int	rtm_seq;	/* for sender to identify action */
220 	int	rtm_errno;	/* why failed */
221 	int	rtm_use;	/* from rtentry */
222 	u_long	rtm_inits;	/* which metrics we are initializing */
223 	struct	rt_metrics rtm_rmx; /* metrics themselves */
224 };
225 
226 #define RTM_VERSION	7	/* Up the ante and ignore older versions */
227 
228 /*
229  * Message types.
230  */
231 #define RTM_ADD		0x1	/* Add Route */
232 #define RTM_DELETE	0x2	/* Delete Route */
233 #define RTM_CHANGE	0x3	/* Change Metrics or flags */
234 #define RTM_GET		0x4	/* Report Metrics */
235 #define RTM_LOSING	0x5	/* Kernel Suspects Partitioning */
236 #define RTM_REDIRECT	0x6	/* Told to use different route */
237 #define RTM_MISS	0x7	/* Lookup failed on this address */
238 #define RTM_LOCK	0x8	/* fix specified metrics */
239 /* 			0x9	   unused, was RTM_OLDADD */
240 /* 			0xa	   unused, was RTM_OLDDEL */
241 #define RTM_RESOLVE	0xb	/* req to resolve dst to LL addr */
242 #define RTM_NEWADDR	0xc	/* address being added to iface */
243 #define RTM_DELADDR	0xd	/* address being removed from iface */
244 #define RTM_IFINFO	0xe	/* iface going up/down etc. */
245 #define	RTM_NEWMADDR	0xf	/* mcast group membership being added to if */
246 #define	RTM_DELMADDR	0x10	/* mcast group membership being deleted */
247 #define	RTM_IFANNOUNCE	0x11	/* iface arrival/departure */
248 #define	RTM_IEEE80211	0x12	/* IEEE80211 wireless event */
249 
250 /*
251  * setsockopt defines used for the filtering.
252  */
253 #define	ROUTE_MSGFILTER	1	/* bitmask of which rtm_type to send to client */
254 #define	ROUTE_FILTER(m)	(1U << (m))
255 #define	RO_MISSFILTER	2	/* array of sockaddrs to match miss dst */
256 
257 #define	RO_FILTSA_MAX	30	/* maximum number of sockaddrs per filter */
258 
259 /*
260  * Bitmask values for rtm_inits and rmx_locks.
261  */
262 #define RTV_MTU		0x1	/* init or lock _mtu */
263 #define RTV_HOPCOUNT	0x2	/* init or lock _hopcount */
264 #define RTV_EXPIRE	0x4	/* init or lock _expire */
265 #define RTV_RPIPE	0x8	/* init or lock _recvpipe */
266 #define RTV_SPIPE	0x10	/* init or lock _sendpipe */
267 #define RTV_SSTHRESH	0x20	/* init or lock _ssthresh */
268 #define RTV_RTT		0x40	/* init or lock _rtt */
269 #define RTV_RTTVAR	0x80	/* init or lock _rttvar */
270 #define RTV_MSL		0x100	/* init or lock _msl */
271 #define RTV_IWMAXSEGS	0x200	/* init or lock _iwmaxsegs */
272 #define RTV_IWCAPSEGS	0x400	/* init or lock _iwcapsegs */
273 
274 /*
275  * Bitmask values for rtm_addrs.
276  */
277 #define RTA_DST		0x1	/* destination sockaddr present */
278 #define RTA_GATEWAY	0x2	/* gateway sockaddr present */
279 #define RTA_NETMASK	0x4	/* netmask sockaddr present */
280 #define RTA_GENMASK	0x8	/* cloning mask sockaddr present */
281 #define RTA_IFP		0x10	/* interface name sockaddr present */
282 #define RTA_IFA		0x20	/* interface addr sockaddr present */
283 #define RTA_AUTHOR	0x40	/* sockaddr for author of redirect */
284 #define RTA_BRD		0x80	/* for NEWADDR, broadcast or p-p dest addr */
285 #define RTA_MPLS1	0x100	/* mpls label and/or operation present */
286 #define RTA_MPLS2	0x200	/* mpls label and/or operation present */
287 #define RTA_MPLS3	0x400	/* mpls label and/or operation present */
288 
289 /*
290  * Index offsets for sockaddr array for alternate internal encoding.
291  */
292 #define RTAX_DST	0	/* destination sockaddr present */
293 #define RTAX_GATEWAY	1	/* gateway sockaddr present */
294 #define RTAX_NETMASK	2	/* netmask sockaddr present */
295 #define RTAX_GENMASK	3	/* cloning mask sockaddr present */
296 #define RTAX_IFP	4	/* interface name sockaddr present */
297 #define RTAX_IFA	5	/* interface addr sockaddr present */
298 #define RTAX_AUTHOR	6	/* sockaddr for author of redirect */
299 #define RTAX_BRD	7	/* for NEWADDR, broadcast or p-p dest addr */
300 #define RTAX_MPLS1	8	/* mpls label and/or operation present */
301 #define RTAX_MPLS2	9	/* mpls label and/or operation present */
302 #define RTAX_MPLS3	10	/* mpls label and/or operation present */
303 #define RTAX_MAX	11	/* size of array to allocate */
304 
305 /*
306  * Sockaddrs passed through the routing socket are padded up to a certain
307  * size, currently multiples of sizeof(long).  These macros used to be
308  * replicated in various user and kernel files, but are now made available
309  * here for convenience and consistency.
310  *
311  * FreeBSD uses a similar macro SA_SIZE which could be defined as
312  * #define SA_SIZE(sa)	RT_ROUNDUP((sa)->sa_len)
313  */
314 #define RT_ROUNDUP2(a, n)	((a) > 0 ? (1 + (((a) - 1) | ((n) - 1))) : (n))
315 #define RT_ROUNDUP(a)		RT_ROUNDUP2((a), sizeof(long))
316 #define RT_ADVANCE(x, n)	(x += RT_ROUNDUP((n)->sa_len))
317 
318 struct rt_addrinfo {
319 	int		 rti_addrs;
320 	struct sockaddr	*rti_info[RTAX_MAX];
321 	int		 rti_flags;
322 	struct ifaddr	*rti_ifa;
323 	struct ifnet	*rti_ifp;
324 };
325 
326 #ifdef _KERNEL
327 
328 #define	rti_dst		rti_info[RTAX_DST]
329 #define	rti_gateway	rti_info[RTAX_GATEWAY]
330 #define	rti_netmask	rti_info[RTAX_NETMASK]
331 #define	rti_genmask	rti_info[RTAX_GENMASK]
332 #define	rti_ifpaddr	rti_info[RTAX_IFP]
333 #define	rti_ifaaddr	rti_info[RTAX_IFA]
334 #define	rti_author	rti_info[RTAX_AUTHOR]
335 #define	rti_bcastaddr	rti_info[RTAX_BRD]
336 #define	rti_mpls1	rti_info[RTAX_MPLS1]
337 #define	rti_mpls2	rti_info[RTAX_MPLS2]
338 #define	rti_mpls3	rti_info[RTAX_MPLS3]
339 
340 extern struct radix_node_head *rt_tables[MAXCPU][AF_MAX+1];
341 
342 struct ifmultiaddr;
343 struct proc;
344 struct ucred;
345 
346 void	 route_init (void);
347 void	 rt_dstmsg(int type, struct sockaddr *dst, int error);
348 int	 rt_getifa (struct rt_addrinfo *);
349 void	 rt_ieee80211msg(struct ifnet *, int, void *, size_t);
350 void	 rt_ifannouncemsg (struct ifnet *, int);
351 void	 rt_ifmsg (struct ifnet *);
352 int	 rt_llroute (struct sockaddr *dst, struct rtentry *rt0,
353 	    struct rtentry **drt);
354 void	 rt_missmsg (int, struct rt_addrinfo *, int, int);
355 void	 rt_newaddrmsg (int, struct ifaddr *, int, struct rtentry *);
356 void	 rt_newmaddrmsg (int, struct ifmultiaddr *);
357 void	 rt_purgecloned(struct ifnet *, int);
358 void	 rt_rtmsg(int cmd, struct rtentry *rt, struct ifnet *ifp, int error);
359 int	 rt_setgate (struct rtentry *, struct sockaddr *, struct sockaddr *);
360 void	 rtalloc (struct route *);
361 void	 rtalloc_ign (struct route *, u_long);
362 
363 struct rtentry *
364 	 _rtlookup (struct sockaddr *, u_long);
365 
366 /* flags to ignore */
367 #define		RTL_DOCLONE	0UL
368 #define		RTL_DONTCLONE	(RTF_CLONING | RTF_PRCLONING)
369 
370 /*
371  * Look up a route with no cloning side-effects or miss reports generated.
372  */
373 static __inline struct rtentry *
374 rtpurelookup(struct sockaddr *dst)
375 {
376 	return _rtlookup(dst, RTL_DONTCLONE);
377 }
378 
379 /*
380  * Do full route lookup with cloning and reporting on misses.
381  */
382 static __inline struct rtentry *
383 rtlookup(struct sockaddr *dst)
384 {
385 	return _rtlookup(dst, RTL_DOCLONE);
386 }
387 
388 typedef void (*rtrequest1_callback_func_t)(int, int, struct rt_addrinfo *,
389 					   struct rtentry *, void *);
390 typedef int (*rtsearch_callback_func_t)(int, struct rt_addrinfo *,
391 					struct rtentry *, void *, int);
392 
393 void	 rtfree (struct rtentry *);
394 void	 rtfree_async (struct rtentry *);
395 int	 rtinit (struct ifaddr *, int, int);
396 int	 rtchange (struct ifaddr *, struct ifaddr *);
397 int	 rtioctl (u_long, caddr_t, struct ucred *);
398 void	 rtredirect (struct sockaddr *, struct sockaddr *,
399 	    struct sockaddr *, int, struct sockaddr *);
400 int	 rtredirect_oncpu (struct sockaddr *, struct sockaddr *,
401 	    struct sockaddr *, int, struct sockaddr *);
402 int	 rtrequest (int, struct sockaddr *,
403 	    struct sockaddr *, struct sockaddr *, int, struct rtentry **);
404 int	 rtrequest_global (int, struct sockaddr *,
405 	    struct sockaddr *, struct sockaddr *, int);
406 int	 rtrequest1 (int, struct rt_addrinfo *, struct rtentry **);
407 int	 rtrequest1_global (int, struct rt_addrinfo *,
408 	    rtrequest1_callback_func_t, void *, boolean_t);
409 
410 #define RTS_EXACTMATCH		TRUE
411 #define RTS_NOEXACTMATCH	FALSE
412 
413 #define RTREQ_PRIO_HIGH		TRUE
414 #define RTREQ_PRIO_NORM		FALSE
415 
416 int	 rtsearch_global(int, struct rt_addrinfo *,
417 	    rtsearch_callback_func_t, void *, boolean_t, boolean_t);
418 
419 int	 rtmask_add_global(struct sockaddr *, boolean_t);
420 
421 struct sockaddr *_rtmask_lookup(struct sockaddr *, boolean_t);
422 
423 static __inline struct sockaddr *
424 rtmask_lookup(struct sockaddr *_mask)
425 {
426 	return _rtmask_lookup(_mask, FALSE);
427 }
428 
429 static __inline struct sockaddr *
430 rtmask_purelookup(struct sockaddr *_mask)
431 {
432 	return _rtmask_lookup(_mask, TRUE);
433 }
434 
435 void	rt_print(struct rt_addrinfo *, struct rtentry *);
436 void	rt_addrinfo_print(int cmd, struct rt_addrinfo *);
437 void	sockaddr_print(const struct sockaddr *);
438 
439 #ifndef _SYS_GLOBALDATA_H_
440 #include <sys/globaldata.h>
441 #endif
442 #ifndef _NET_NETISR2_H_
443 #include <net/netisr2.h>
444 #endif
445 
446 static __inline void
447 RTFREE(struct rtentry *rt)
448 {
449 
450 	ASSERT_NETISR_NCPUS(rt->rt_cpuid);
451 	if (rt->rt_refcnt <= 1)
452 		rtfree(rt);
453 	else
454 		--rt->rt_refcnt;
455 }
456 
457 int	in_inithead(void **, int);
458 
459 #endif	/* _KERNEL */
460 
461 #endif	/* !_NET_ROUTE_H_ */
462