xref: /dragonfly/sys/net/ip_mroute/ip_mroute.h (revision bfb67b1bd1a4401995b5b7b592ef8628019b9113)
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  * 3. 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: src/sys/netinet/ip_mroute.h,v 1.17.2.3 2003/08/24 08:24:38 hsu Exp $
35  * $DragonFly: src/sys/net/ip_mroute/ip_mroute.h,v 1.3 2003/08/24 23:07:07 hsu Exp $
36  */
37 
38 #ifndef _NETINET_IP_MROUTE_H_
39 #define _NETINET_IP_MROUTE_H_
40 
41 /*
42  * Definitions for IP multicast forwarding.
43  *
44  * Written by David Waitzman, BBN Labs, August 1988.
45  * Modified by Steve Deering, Stanford, February 1989.
46  * Modified by Ajit Thyagarajan, PARC, August 1993.
47  * Modified by Ajit Thyagarajan, PARC, August 1994.
48  * Modified by Ahmed Helmy, SGI, June 1996.
49  * Modified by Pavlin Radoslavov, ICSI, October 2002.
50  *
51  * MROUTING Revision: 3.3.1.3
52  * and PIM-SMv2 and PIM-DM support, advanced API support,
53  * bandwidth metering and signaling.
54  */
55 
56 
57 /*
58  * Multicast Routing set/getsockopt commands.
59  */
60 #define   MRT_INIT  100       /* initialize forwarder */
61 #define   MRT_DONE  101       /* shut down forwarder */
62 #define   MRT_ADD_VIF         102       /* create virtual interface */
63 #define   MRT_DEL_VIF         103       /* delete virtual interface */
64 #define MRT_ADD_MFC 104       /* insert forwarding cache entry */
65 #define MRT_DEL_MFC 105       /* delete forwarding cache entry */
66 #define MRT_VERSION 106       /* get kernel version number */
67 #define MRT_ASSERT      107     /* enable assert processing */
68 #define MRT_PIM               MRT_ASSERT /* enable PIM processing */
69 #define MRT_API_SUPPORT       109       /* supported MRT API */
70 #define MRT_API_CONFIG        110       /* config MRT API */
71 #define MRT_ADD_BW_UPCALL 111 /* create bandwidth monitor */
72 #define MRT_DEL_BW_UPCALL 112 /* delete bandwidth monitor */
73 
74 
75 #define GET_TIME(t) microtime(&t)
76 
77 /*
78  * Types and macros for handling bitmaps with one bit per virtual interface.
79  */
80 #define   MAXVIFS 32
81 typedef u_long vifbitmap_t;
82 typedef u_short vifi_t;                 /* type of a vif index */
83 #define ALL_VIFS (vifi_t)-1
84 
85 #define   VIFM_SET(n, m)                ((m) |= (1 << (n)))
86 #define   VIFM_CLR(n, m)                ((m) &= ~(1 << (n)))
87 #define   VIFM_ISSET(n, m)    ((m) & (1 << (n)))
88 #define   VIFM_CLRALL(m)                ((m) = 0x00000000)
89 #define   VIFM_COPY(mfrom, mto)         ((mto) = (mfrom))
90 #define   VIFM_SAME(m1, m2)   ((m1) == (m2))
91 
92 
93 /*
94  * Argument structure for MRT_ADD_VIF.
95  * (MRT_DEL_VIF takes a single vifi_t argument.)
96  */
97 struct vifctl {
98           vifi_t    vifc_vifi;                    /* the index of the vif to be added */
99           u_char    vifc_flags;         /* VIFF_ flags defined below */
100           u_char    vifc_threshold;     /* min ttl required to forward on vif */
101           u_int     vifc_rate_limit;    /* max rate */
102           struct    in_addr vifc_lcl_addr;        /* local interface address */
103           struct    in_addr vifc_rmt_addr;        /* remote address (tunnels only) */
104 };
105 
106 #define   VIFF_TUNNEL         0x1                 /* vif represents a tunnel end-point */
107 #define VIFF_SRCRT  0x2                 /* tunnel uses IP source routing */
108 #define VIFF_REGISTER         0x4                 /* used for PIM Register encap/decap */
109 
110 /*
111  * Argument structure for MRT_ADD_MFC and MRT_DEL_MFC
112  * XXX if you change this, make sure to change struct mfcctl2 as well.
113  */
114 struct mfcctl {
115     struct in_addr  mfcc_origin;                  /* ip origin of mcasts       */
116     struct in_addr  mfcc_mcastgrp;                /* multicast group associated*/
117     vifi_t              mfcc_parent;              /* incoming vif              */
118     u_char              mfcc_ttls[MAXVIFS];       /* forwarding ttls on vifs   */
119 };
120 
121 /*
122  * The new argument structure for MRT_ADD_MFC and MRT_DEL_MFC overlays
123  * and extends the old struct mfcctl.
124  */
125 struct mfcctl2 {
126           /* the mfcctl fields */
127           struct in_addr      mfcc_origin;                  /* ip origin of mcasts             */
128           struct in_addr      mfcc_mcastgrp;                /* multicast group associated*/
129           vifi_t              mfcc_parent;                  /* incoming vif                    */
130           u_char              mfcc_ttls[MAXVIFS];           /* forwarding ttls on vifs   */
131 
132           /* extension fields */
133           uint8_t             mfcc_flags[MAXVIFS];          /* the MRT_MFC_FLAGS_* flags */
134           struct in_addr      mfcc_rp;            /* the RP address            */
135 };
136 /*
137  * The advanced-API flags.
138  *
139  * The MRT_MFC_FLAGS_XXX API flags are also used as flags
140  * for the mfcc_flags field.
141  */
142 #define   MRT_MFC_FLAGS_DISABLE_WRONGVIF          (1 << 0) /* disable WRONGVIF signals */
143 #define   MRT_MFC_FLAGS_BORDER_VIF      (1 << 1) /* border vif                       */
144 #define MRT_MFC_RP                      (1 << 8) /* enable RP address      */
145 #define MRT_MFC_BW_UPCALL               (1 << 9) /* enable bw upcalls      */
146 #define MRT_MFC_FLAGS_ALL               (MRT_MFC_FLAGS_DISABLE_WRONGVIF |    \
147                                                    MRT_MFC_FLAGS_BORDER_VIF)
148 #define MRT_API_FLAGS_ALL               (MRT_MFC_FLAGS_ALL |                         \
149                                                    MRT_MFC_RP |                                \
150                                                    MRT_MFC_BW_UPCALL)
151 
152 /*
153  * Structure for installing or delivering an upcall if the
154  * measured bandwidth is above or below a threshold.
155  *
156  * User programs (e.g. daemons) may have a need to know when the
157  * bandwidth used by some data flow is above or below some threshold.
158  * This interface allows the userland to specify the threshold (in
159  * bytes and/or packets) and the measurement interval. Flows are
160  * all packet with the same source and destination IP address.
161  * At the moment the code is only used for multicast destinations
162  * but there is nothing that prevents its use for unicast.
163  *
164  * The measurement interval cannot be shorter than some Tmin (currently, 3s).
165  * The threshold is set in packets and/or bytes per_interval.
166  *
167  * Measurement works as follows:
168  *
169  * For >= measurements:
170  * The first packet marks the start of a measurement interval.
171  * During an interval we count packets and bytes, and when we
172  * pass the threshold we deliver an upcall and we are done.
173  * The first packet after the end of the interval resets the
174  * count and restarts the measurement.
175  *
176  * For <= measurement:
177  * We start a timer to fire at the end of the interval, and
178  * then for each incoming packet we count packets and bytes.
179  * When the timer fires, we compare the value with the threshold,
180  * schedule an upcall if we are below, and restart the measurement
181  * (reschedule timer and zero counters).
182  */
183 
184 struct bw_data {
185           struct timeval      b_time;
186           uint64_t  b_packets;
187           uint64_t  b_bytes;
188 };
189 
190 struct bw_upcall {
191           struct in_addr      bu_src;                       /* source address            */
192           struct in_addr      bu_dst;                       /* destination address       */
193           uint32_t  bu_flags;           /* misc flags (see below)    */
194 #define BW_UPCALL_UNIT_PACKETS   (1 << 0)         /* threshold (in packets)    */
195 #define BW_UPCALL_UNIT_BYTES     (1 << 1)         /* threshold (in bytes)      */
196 #define BW_UPCALL_GEQ            (1 << 2)         /* upcall if bw >= threshold */
197 #define BW_UPCALL_LEQ            (1 << 3)         /* upcall if bw <= threshold */
198 #define BW_UPCALL_DELETE_ALL     (1 << 4)         /* delete all upcalls for s,d*/
199           struct bw_data      bu_threshold;                 /* the bw threshold      */
200           struct bw_data      bu_measured;                  /* the measured bw       */
201 };
202 
203 /* max. number of upcalls to deliver together */
204 #define BW_UPCALLS_MAX                                      128
205 /* min. threshold time interval for bandwidth measurement */
206 #define BW_UPCALL_THRESHOLD_INTERVAL_MIN_SEC      3
207 #define BW_UPCALL_THRESHOLD_INTERVAL_MIN_USEC     0
208 
209 /*
210  * The kernel's multicast routing statistics.
211  */
212 struct mrtstat {
213     u_long          mrts_mfc_lookups;   /* # forw. cache hash table hits   */
214     u_long          mrts_mfc_misses;    /* # forw. cache hash table misses */
215     u_long          mrts_upcalls;                 /* # calls to mrouted              */
216     u_long          mrts_no_route;                /* no route for packet's origin    */
217     u_long          mrts_bad_tunnel;    /* malformed tunnel options        */
218     u_long          mrts_cant_tunnel;   /* no room for tunnel options      */
219     u_long          mrts_wrong_if;                /* arrived on wrong interface    */
220     u_long          mrts_upq_ovflw;               /* upcall Q overflow                       */
221     u_long          mrts_cache_cleanups;          /* # entries with no upcalls     */
222     u_long          mrts_drop_sel;      /* pkts dropped selectively        */
223     u_long          mrts_q_overflow;              /* pkts dropped - Q overflow       */
224     u_long          mrts_pkt2large;               /* pkts dropped - size > BKT SIZE  */
225     u_long          mrts_upq_sockfull;  /* upcalls dropped - socket full */
226 };
227 
228 /*
229  * Argument structure used by mrouted to get src-grp pkt counts
230  */
231 struct sioc_sg_req {
232     struct in_addr src;
233     struct in_addr grp;
234     u_long pktcnt;
235     u_long bytecnt;
236     u_long wrong_if;
237 };
238 
239 /*
240  * Argument structure used by mrouted to get vif pkt counts
241  */
242 struct sioc_vif_req {
243     vifi_t vifi;              /* vif number                                     */
244     u_long icount;            /* Input packet count on vif            */
245     u_long ocount;            /* Output packet count on vif           */
246     u_long ibytes;            /* Input byte count on vif              */
247     u_long obytes;            /* Output byte count on vif             */
248 };
249 
250 
251 /*
252  * The kernel's virtual-interface structure.
253  */
254 struct vif {
255     u_char                    v_flags;            /* VIFF_ flags defined above         */
256     u_char                    v_threshold;        /* min ttl required to forward on vif*/
257     u_int                     v_rate_limit;       /* max rate                                  */
258     struct tbf             *v_tbf;                /* token bucket structure at intf.   */
259     struct in_addr  v_lcl_addr;         /* local interface address           */
260     struct in_addr  v_rmt_addr;         /* remote address (tunnels only)     */
261     struct ifnet       *v_ifp;                    /* pointer to interface              */
262     u_long                    v_pkt_in; /* # pkts in on interface            */
263     u_long                    v_pkt_out;          /* # pkts out on interface           */
264     u_long                    v_bytes_in;         /* # bytes in on interface         */
265     u_long                    v_bytes_out;        /* # bytes out on interface        */
266     struct route    v_route;  /* cached route if this is a tunnel */
267     u_int           v_rsvp_on;          /* RSVP listening on this vif */
268     struct socket      *v_rsvpd;        /* RSVP daemon socket */
269 };
270 
271 /*
272  * The kernel's multicast forwarding cache entry structure
273  * (A field for the type of service (mfc_tos) is to be added
274  * at a future point)
275  */
276 struct mfc {
277           struct in_addr      mfc_origin;                   /* IP origin of mcasts             */
278           struct in_addr  mfc_mcastgrp;                     /* multicast group associated*/
279           vifi_t              mfc_parent;                   /* incoming vif              */
280           u_char              mfc_ttls[MAXVIFS];  /* forwarding ttls on vifs   */
281           u_long              mfc_pkt_cnt;                  /* pkt count for src-grp     */
282           u_long              mfc_byte_cnt;                 /* byte count for src-grp    */
283           u_long              mfc_wrong_if;                 /* wrong if for src-grp            */
284           int                 mfc_expire;                   /* time to clean entry up    */
285           struct timeval      mfc_last_assert;    /* last time I sent an assert*/
286           struct rtdetq       *mfc_stall;                   /* q of packets awaiting mfc */
287           struct mfc          *mfc_next;                    /* next mfc entry            */
288           uint8_t             mfc_flags[MAXVIFS]; /* the MRT_MFC_FLAGS_* flags */
289           struct in_addr      mfc_rp;                       /* the RP address        */
290           struct bw_meter     *mfc_bw_meter;                /* list of bandwidth meters  */
291 };
292 
293 /*
294  * Struct used to communicate from kernel to multicast router
295  * note the convenient similarity to an IP packet
296  */
297 struct igmpmsg {
298     uint32_t            unused1;
299     uint32_t            unused2;
300     u_char              im_msgtype;                         /* what type of message           */
301 #define IGMPMSG_NOCACHE                 1         /* no MFC in the kernel                     */
302 #define IGMPMSG_WRONGVIF      2         /* packet came from wrong interface */
303 #define   IGMPMSG_WHOLEPKT    3         /* PIM pkt for user level encap.    */
304 #define   IGMPMSG_BW_UPCALL   4         /* BW monitoring upcall                     */
305     u_char              im_mbz;                             /* must be zero                   */
306     u_char              im_vif;                             /* vif rec'd on                   */
307     u_char              unused3;
308     struct in_addr  im_src, im_dst;
309 };
310 
311 /*
312  * Argument structure used for pkt info. while upcall is made
313  */
314 struct rtdetq {
315     struct mbuf     *m;                 /* A copy of the packet                     */
316     struct ifnet    *ifp;               /* Interface pkt came in on       */
317     vifi_t                    xmt_vif;  /* Saved copy of imo_multicast_vif  */
318     struct rtdetq   *next;              /* Next in list of packets          */
319 };
320 
321 #define MFCTBLSIZ   256
322 #if (MFCTBLSIZ & (MFCTBLSIZ - 1)) == 0    /* from sys:route.h */
323 #define MFCHASHMOD(h)         ((h) & (MFCTBLSIZ - 1))
324 #else
325 #define MFCHASHMOD(h)         ((h) % MFCTBLSIZ)
326 #endif
327 
328 #define MAX_UPQ     4                   /* max. no of pkts in upcall Q */
329 
330 /*
331  * Token Bucket filter code
332  */
333 #define MAX_BKT_SIZE    10000             /* 10K bytes size                     */
334 #define MAXQSIZE        10                /* max # of pkts in queue   */
335 
336 /*
337  * the token bucket filter at each vif
338  */
339 struct tbf
340 {
341     struct timeval tbf_last_pkt_t; /* arr. time of last pkt           */
342     u_long tbf_n_tok;         /* no of tokens in bucket     */
343     u_long tbf_q_len;         /* length of queue at this vif          */
344     u_long tbf_max_q_len;     /* max. queue length                    */
345     struct mbuf *tbf_q;                 /* Packet queue                         */
346     struct mbuf *tbf_t;                 /* tail-insertion pointer     */
347 };
348 
349 /*
350  * Structure for measuring the bandwidth and sending an upcall if the
351  * measured bandwidth is above or below a threshold.
352  */
353 struct bw_meter {
354           struct bw_meter     *bm_mfc_next;                 /* next bw meter (same mfc)  */
355           struct bw_meter     *bm_time_next;                /* next bw meter (same time) */
356           uint32_t  bm_time_hash;                 /* the time hash value       */
357           struct mfc          *bm_mfc;            /* the corresponding mfc     */
358           uint32_t  bm_flags;           /* misc flags (see below)    */
359 #define BW_METER_UNIT_PACKETS (1 << 0)  /* threshold (in packets)    */
360 #define BW_METER_UNIT_BYTES   (1 << 1)  /* threshold (in bytes)      */
361 #define BW_METER_GEQ                    (1 << 2)  /* upcall if bw >= threshold */
362 #define BW_METER_LEQ                    (1 << 3)  /* upcall if bw <= threshold */
363 #define BW_METER_USER_FLAGS   (BW_METER_UNIT_PACKETS |                \
364                                          BW_METER_UNIT_BYTES |                            \
365                                          BW_METER_GEQ |                                   \
366                                          BW_METER_LEQ)
367 
368 #define BW_METER_UPCALL_DELIVERED (1 << 24)       /* upcall was delivered      */
369 
370           struct bw_data      bm_threshold;                 /* the upcall threshold            */
371           struct bw_data      bm_measured;                  /* the measured bw       */
372           struct timeval      bm_start_time;                /* abs. time                       */
373 };
374 
375 #ifdef _KERNEL
376 
377 struct sockopt;
378 
379 extern int          (*ip_mrouter_set)(struct socket *, struct sockopt *);
380 extern int          (*ip_mrouter_get)(struct socket *, struct sockopt *);
381 extern int          (*ip_mrouter_done)(void);
382 extern int          (*mrt_ioctl)(u_long, caddr_t);
383 
384 #endif /* _KERNEL */
385 
386 #endif /* _NETINET_IP_MROUTE_H_ */
387