xref: /freebsd/sys/netinet/ip_mroute.h (revision a35d88931c87cfe6bd38f01d7bad22140b3b38f3)
1 /*
2  * Copyright (c) 1989 Stephen Deering.
3  * Copyright (c) 1992, 1993
4  *	The Regents of the University of California.  All rights reserved.
5  *
6  * This code is derived from software contributed to Berkeley by
7  * Stephen Deering of Stanford University.
8  *
9  * Redistribution and use in source and binary forms, with or without
10  * modification, are permitted provided that the following conditions
11  * are met:
12  * 1. Redistributions of source code must retain the above copyright
13  *    notice, this list of conditions and the following disclaimer.
14  * 2. Redistributions in binary form must reproduce the above copyright
15  *    notice, this list of conditions and the following disclaimer in the
16  *    documentation and/or other materials provided with the distribution.
17  * 4. Neither the name of the University nor the names of its contributors
18  *    may be used to endorse or promote products derived from this software
19  *    without specific prior written permission.
20  *
21  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
22  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
23  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
24  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
25  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
26  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
27  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
28  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
29  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
30  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
31  * SUCH DAMAGE.
32  *
33  *	@(#)ip_mroute.h	8.1 (Berkeley) 6/10/93
34  * $FreeBSD$
35  */
36 
37 #ifndef _NETINET_IP_MROUTE_H_
38 #define _NETINET_IP_MROUTE_H_
39 
40 /*
41  * Definitions for IP multicast forwarding.
42  *
43  * Written by David Waitzman, BBN Labs, August 1988.
44  * Modified by Steve Deering, Stanford, February 1989.
45  * Modified by Ajit Thyagarajan, PARC, August 1993.
46  * Modified by Ajit Thyagarajan, PARC, August 1994.
47  * Modified by Ahmed Helmy, SGI, June 1996.
48  * Modified by Pavlin Radoslavov, ICSI, October 2002.
49  *
50  * MROUTING Revision: 3.3.1.3
51  * and PIM-SMv2 and PIM-DM support, advanced API support,
52  * bandwidth metering and signaling.
53  */
54 
55 
56 /*
57  * Multicast Routing set/getsockopt commands.
58  */
59 #define	MRT_INIT	100	/* initialize forwarder */
60 #define	MRT_DONE	101	/* shut down forwarder */
61 #define	MRT_ADD_VIF	102	/* create virtual interface */
62 #define	MRT_DEL_VIF	103	/* delete virtual interface */
63 #define MRT_ADD_MFC	104	/* insert forwarding cache entry */
64 #define MRT_DEL_MFC	105	/* delete forwarding cache entry */
65 #define MRT_VERSION	106	/* get kernel version number */
66 #define MRT_ASSERT      107     /* enable assert processing */
67 #define MRT_PIM		MRT_ASSERT /* enable PIM processing */
68 #define MRT_API_SUPPORT	109	/* supported MRT API */
69 #define MRT_API_CONFIG	110	/* config MRT API */
70 #define MRT_ADD_BW_UPCALL 111	/* create bandwidth monitor */
71 #define MRT_DEL_BW_UPCALL 112	/* delete bandwidth monitor */
72 
73 
74 #define GET_TIME(t)	microtime(&t)
75 
76 /*
77  * Types and macros for handling bitmaps with one bit per virtual interface.
78  */
79 #define	MAXVIFS 32
80 typedef u_long vifbitmap_t;
81 typedef u_short vifi_t;		/* type of a vif index */
82 #define ALL_VIFS (vifi_t)-1
83 
84 #define	VIFM_SET(n, m)		((m) |= (1 << (n)))
85 #define	VIFM_CLR(n, m)		((m) &= ~(1 << (n)))
86 #define	VIFM_ISSET(n, m)	((m) & (1 << (n)))
87 #define	VIFM_CLRALL(m)		((m) = 0x00000000)
88 #define	VIFM_COPY(mfrom, mto)	((mto) = (mfrom))
89 #define	VIFM_SAME(m1, m2)	((m1) == (m2))
90 
91 
92 /*
93  * Argument structure for MRT_ADD_VIF.
94  * (MRT_DEL_VIF takes a single vifi_t argument.)
95  */
96 struct vifctl {
97 	vifi_t	vifc_vifi;		/* the index of the vif to be added */
98 	u_char	vifc_flags;		/* VIFF_ flags defined below */
99 	u_char	vifc_threshold;		/* min ttl required to forward on vif */
100 	u_int	vifc_rate_limit;	/* max rate */
101 	struct	in_addr vifc_lcl_addr;	/* local interface address */
102 	struct	in_addr vifc_rmt_addr;	/* remote address (tunnels only) */
103 };
104 
105 #define	VIFF_TUNNEL	0x1		/* vif represents a tunnel end-point */
106 #define VIFF_SRCRT	0x2		/* tunnel uses IP source routing */
107 #define VIFF_REGISTER	0x4		/* used for PIM Register encap/decap */
108 
109 /*
110  * Argument structure for MRT_ADD_MFC and MRT_DEL_MFC
111  * XXX if you change this, make sure to change struct mfcctl2 as well.
112  */
113 struct mfcctl {
114     struct in_addr  mfcc_origin;		/* ip origin of mcasts       */
115     struct in_addr  mfcc_mcastgrp;		/* multicast group associated*/
116     vifi_t	    mfcc_parent;		/* incoming vif              */
117     u_char	    mfcc_ttls[MAXVIFS];		/* forwarding ttls on vifs   */
118 };
119 
120 /*
121  * The new argument structure for MRT_ADD_MFC and MRT_DEL_MFC overlays
122  * and extends the old struct mfcctl.
123  */
124 struct mfcctl2 {
125 	/* the mfcctl fields */
126 	struct in_addr	mfcc_origin;		/* ip origin of mcasts	     */
127 	struct in_addr	mfcc_mcastgrp;		/* multicast group associated*/
128 	vifi_t		mfcc_parent;		/* incoming vif		     */
129 	u_char		mfcc_ttls[MAXVIFS];	/* forwarding ttls on vifs   */
130 
131 	/* extension fields */
132 	uint8_t		mfcc_flags[MAXVIFS];	/* the MRT_MFC_FLAGS_* flags */
133 	struct in_addr	mfcc_rp;		/* the RP address            */
134 };
135 /*
136  * The advanced-API flags.
137  *
138  * The MRT_MFC_FLAGS_XXX API flags are also used as flags
139  * for the mfcc_flags field.
140  */
141 #define	MRT_MFC_FLAGS_DISABLE_WRONGVIF	(1 << 0) /* disable WRONGVIF signals */
142 #define	MRT_MFC_FLAGS_BORDER_VIF	(1 << 1) /* border vif		     */
143 #define MRT_MFC_RP			(1 << 8) /* enable RP address	     */
144 #define MRT_MFC_BW_UPCALL		(1 << 9) /* enable bw upcalls	     */
145 #define MRT_MFC_FLAGS_ALL		(MRT_MFC_FLAGS_DISABLE_WRONGVIF |    \
146 					 MRT_MFC_FLAGS_BORDER_VIF)
147 #define MRT_API_FLAGS_ALL		(MRT_MFC_FLAGS_ALL |		     \
148 					 MRT_MFC_RP |			     \
149 					 MRT_MFC_BW_UPCALL)
150 
151 /*
152  * Structure for installing or delivering an upcall if the
153  * measured bandwidth is above or below a threshold.
154  *
155  * User programs (e.g. daemons) may have a need to know when the
156  * bandwidth used by some data flow is above or below some threshold.
157  * This interface allows the userland to specify the threshold (in
158  * bytes and/or packets) and the measurement interval. Flows are
159  * all packet with the same source and destination IP address.
160  * At the moment the code is only used for multicast destinations
161  * but there is nothing that prevents its use for unicast.
162  *
163  * The measurement interval cannot be shorter than some Tmin (currently, 3s).
164  * The threshold is set in packets and/or bytes per_interval.
165  *
166  * Measurement works as follows:
167  *
168  * For >= measurements:
169  * The first packet marks the start of a measurement interval.
170  * During an interval we count packets and bytes, and when we
171  * pass the threshold we deliver an upcall and we are done.
172  * The first packet after the end of the interval resets the
173  * count and restarts the measurement.
174  *
175  * For <= measurement:
176  * We start a timer to fire at the end of the interval, and
177  * then for each incoming packet we count packets and bytes.
178  * When the timer fires, we compare the value with the threshold,
179  * schedule an upcall if we are below, and restart the measurement
180  * (reschedule timer and zero counters).
181  */
182 
183 struct bw_data {
184 	struct timeval	b_time;
185 	uint64_t	b_packets;
186 	uint64_t	b_bytes;
187 };
188 
189 struct bw_upcall {
190 	struct in_addr	bu_src;			/* source address            */
191 	struct in_addr	bu_dst;			/* destination address       */
192 	uint32_t	bu_flags;		/* misc flags (see below)    */
193 #define BW_UPCALL_UNIT_PACKETS   (1 << 0)	/* threshold (in packets)    */
194 #define BW_UPCALL_UNIT_BYTES     (1 << 1)	/* threshold (in bytes)      */
195 #define BW_UPCALL_GEQ            (1 << 2)	/* upcall if bw >= threshold */
196 #define BW_UPCALL_LEQ            (1 << 3)	/* upcall if bw <= threshold */
197 #define BW_UPCALL_DELETE_ALL     (1 << 4)	/* delete all upcalls for s,d*/
198 	struct bw_data	bu_threshold;		/* the bw threshold	     */
199 	struct bw_data	bu_measured;		/* the measured bw	     */
200 };
201 
202 /* max. number of upcalls to deliver together */
203 #define BW_UPCALLS_MAX				128
204 /* min. threshold time interval for bandwidth measurement */
205 #define BW_UPCALL_THRESHOLD_INTERVAL_MIN_SEC	3
206 #define BW_UPCALL_THRESHOLD_INTERVAL_MIN_USEC	0
207 
208 /*
209  * The kernel's multicast routing statistics.
210  */
211 struct mrtstat {
212     u_long	mrts_mfc_lookups;	/* # forw. cache hash table hits   */
213     u_long	mrts_mfc_misses;	/* # forw. cache hash table misses */
214     u_long	mrts_upcalls;		/* # calls to mrouted              */
215     u_long	mrts_no_route;		/* no route for packet's origin    */
216     u_long	mrts_bad_tunnel;	/* malformed tunnel options        */
217     u_long	mrts_cant_tunnel;	/* no room for tunnel options      */
218     u_long	mrts_wrong_if;		/* arrived on wrong interface	   */
219     u_long	mrts_upq_ovflw;		/* upcall Q overflow		   */
220     u_long	mrts_cache_cleanups;	/* # entries with no upcalls	   */
221     u_long	mrts_drop_sel;		/* pkts dropped selectively        */
222     u_long	mrts_q_overflow;	/* pkts dropped - Q overflow       */
223     u_long	mrts_pkt2large;		/* pkts dropped - size > BKT SIZE  */
224     u_long	mrts_upq_sockfull;	/* upcalls dropped - socket full */
225 };
226 
227 /*
228  * Argument structure used by mrouted to get src-grp pkt counts
229  */
230 struct sioc_sg_req {
231     struct in_addr src;
232     struct in_addr grp;
233     u_long pktcnt;
234     u_long bytecnt;
235     u_long wrong_if;
236 };
237 
238 /*
239  * Argument structure used by mrouted to get vif pkt counts
240  */
241 struct sioc_vif_req {
242     vifi_t vifi;		/* vif number				*/
243     u_long icount;		/* Input packet count on vif		*/
244     u_long ocount;		/* Output packet count on vif		*/
245     u_long ibytes;		/* Input byte count on vif		*/
246     u_long obytes;		/* Output byte count on vif		*/
247 };
248 
249 
250 /*
251  * The kernel's virtual-interface structure.
252  */
253 struct vif {
254     u_char		v_flags;	/* VIFF_ flags defined above         */
255     u_char		v_threshold;	/* min ttl required to forward on vif*/
256     u_int		v_rate_limit;	/* max rate			     */
257     struct tbf	       *v_tbf;		/* token bucket structure at intf.   */
258     struct in_addr	v_lcl_addr;	/* local interface address           */
259     struct in_addr	v_rmt_addr;	/* remote address (tunnels only)     */
260     struct ifnet       *v_ifp;		/* pointer to interface              */
261     u_long		v_pkt_in;	/* # pkts in on interface            */
262     u_long		v_pkt_out;	/* # pkts out on interface           */
263     u_long		v_bytes_in;	/* # bytes in on interface	     */
264     u_long		v_bytes_out;	/* # bytes out on interface	     */
265     struct route	v_route;	/* cached route if this is a tunnel */
266     u_int		v_rsvp_on;	/* RSVP listening on this vif */
267     struct socket      *v_rsvpd;	/* RSVP daemon socket */
268 };
269 
270 /*
271  * The kernel's multicast forwarding cache entry structure
272  * (A field for the type of service (mfc_tos) is to be added
273  * at a future point)
274  */
275 struct mfc {
276 	struct in_addr	mfc_origin;		/* IP origin of mcasts	     */
277 	struct in_addr  mfc_mcastgrp;		/* multicast group associated*/
278 	vifi_t		mfc_parent;		/* incoming vif              */
279 	u_char		mfc_ttls[MAXVIFS];	/* forwarding ttls on vifs   */
280 	u_long		mfc_pkt_cnt;		/* pkt count for src-grp     */
281 	u_long		mfc_byte_cnt;		/* byte count for src-grp    */
282 	u_long		mfc_wrong_if;		/* wrong if for src-grp	     */
283 	int		mfc_expire;		/* time to clean entry up    */
284 	struct timeval	mfc_last_assert;	/* last time I sent an assert*/
285 	struct rtdetq	*mfc_stall;		/* q of packets awaiting mfc */
286 	struct mfc	*mfc_next;		/* next mfc entry            */
287 	uint8_t		mfc_flags[MAXVIFS];	/* the MRT_MFC_FLAGS_* flags */
288 	struct in_addr	mfc_rp;			/* the RP address	     */
289 	struct bw_meter	*mfc_bw_meter;		/* list of bandwidth meters  */
290 };
291 
292 /*
293  * Struct used to communicate from kernel to multicast router
294  * note the convenient similarity to an IP packet
295  */
296 struct igmpmsg {
297     u_long	    unused1;
298     u_long	    unused2;
299     u_char	    im_msgtype;			/* what type of message	    */
300 #define IGMPMSG_NOCACHE		1	/* no MFC in the kernel		    */
301 #define IGMPMSG_WRONGVIF	2	/* packet came from wrong interface */
302 #define	IGMPMSG_WHOLEPKT	3	/* PIM pkt for user level encap.    */
303 #define	IGMPMSG_BW_UPCALL	4	/* BW monitoring upcall		    */
304     u_char	    im_mbz;			/* must be zero		    */
305     u_char	    im_vif;			/* vif rec'd on		    */
306     u_char	    unused3;
307     struct in_addr  im_src, im_dst;
308 };
309 
310 /*
311  * Argument structure used for pkt info. while upcall is made
312  */
313 struct rtdetq {
314     struct mbuf		*m;		/* A copy of the packet		    */
315     struct ifnet	*ifp;		/* Interface pkt came in on	    */
316     vifi_t		xmt_vif;	/* Saved copy of imo_multicast_vif  */
317     struct rtdetq	*next;		/* Next in list of packets          */
318 };
319 
320 #define MFCTBLSIZ	256
321 #if (MFCTBLSIZ & (MFCTBLSIZ - 1)) == 0	  /* from sys:route.h */
322 #define MFCHASHMOD(h)	((h) & (MFCTBLSIZ - 1))
323 #else
324 #define MFCHASHMOD(h)	((h) % MFCTBLSIZ)
325 #endif
326 
327 #define MAX_UPQ	4		/* max. no of pkts in upcall Q */
328 
329 /*
330  * Token Bucket filter code
331  */
332 #define MAX_BKT_SIZE    10000             /* 10K bytes size		*/
333 #define MAXQSIZE        10                /* max # of pkts in queue	*/
334 
335 /*
336  * the token bucket filter at each vif
337  */
338 struct tbf
339 {
340     struct timeval tbf_last_pkt_t; /* arr. time of last pkt	*/
341     u_long tbf_n_tok;		/* no of tokens in bucket	*/
342     u_long tbf_q_len;		/* length of queue at this vif	*/
343     u_long tbf_max_q_len;	/* max. queue length		*/
344     struct mbuf *tbf_q;		/* Packet queue			*/
345     struct mbuf *tbf_t;		/* tail-insertion pointer	*/
346 };
347 
348 /*
349  * Structure for measuring the bandwidth and sending an upcall if the
350  * measured bandwidth is above or below a threshold.
351  */
352 struct bw_meter {
353 	struct bw_meter	*bm_mfc_next;		/* next bw meter (same mfc)  */
354 	struct bw_meter	*bm_time_next;		/* next bw meter (same time) */
355 	uint32_t	bm_time_hash;		/* the time hash value       */
356 	struct mfc	*bm_mfc;		/* the corresponding mfc     */
357 	uint32_t	bm_flags;		/* misc flags (see below)    */
358 #define BW_METER_UNIT_PACKETS	(1 << 0)	/* threshold (in packets)    */
359 #define BW_METER_UNIT_BYTES	(1 << 1)	/* threshold (in bytes)      */
360 #define BW_METER_GEQ		(1 << 2)	/* upcall if bw >= threshold */
361 #define BW_METER_LEQ		(1 << 3)	/* upcall if bw <= threshold */
362 #define BW_METER_USER_FLAGS	(BW_METER_UNIT_PACKETS |		\
363 				 BW_METER_UNIT_BYTES |			\
364 				 BW_METER_GEQ |				\
365 				 BW_METER_LEQ)
366 
367 #define BW_METER_UPCALL_DELIVERED (1 << 24)	/* upcall was delivered      */
368 
369 	struct bw_data	bm_threshold;		/* the upcall threshold	     */
370 	struct bw_data	bm_measured;		/* the measured bw	     */
371 	struct timeval	bm_start_time;		/* abs. time		     */
372 };
373 
374 #ifdef _KERNEL
375 
376 struct sockopt;
377 
378 extern int	(*ip_mrouter_set)(struct socket *, struct sockopt *);
379 extern int	(*ip_mrouter_get)(struct socket *, struct sockopt *);
380 extern int	(*ip_mrouter_done)(void);
381 extern int	(*mrt_ioctl)(int, caddr_t);
382 
383 #endif /* _KERNEL */
384 
385 #endif /* _NETINET_IP_MROUTE_H_ */
386