summaryrefslogtreecommitdiffstats
path: root/zebra/rt_socket.c
blob: f9888b12d4e8ef5b78bbda1e271abc461c6d87e4 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
// SPDX-License-Identifier: GPL-2.0-or-later
/*
 * Kernel routing table updates by routing socket.
 * Copyright (C) 1997, 98 Kunihiro Ishiguro
 */

#include <zebra.h>

#ifndef HAVE_NETLINK

#ifdef __OpenBSD__
#include <netmpls/mpls.h>
#endif

#include "if.h"
#include "prefix.h"
#include "sockunion.h"
#include "log.h"
#include "privs.h"
#include "vxlan.h"
#include "lib_errors.h"

#include "zebra/debug.h"
#include "zebra/rib.h"
#include "zebra/rt.h"
#include "zebra/kernel_socket.h"
#include "zebra/zebra_mpls.h"
#include "zebra/zebra_errors.h"

extern struct zebra_privs_t zserv_privs;

#ifdef __OpenBSD__
static int kernel_rtm_add_labels(struct mpls_label_stack *nh_label,
				 struct sockaddr_mpls *smpls)
{
	if (nh_label->num_labels > 1) {
		flog_warn(EC_ZEBRA_MAX_LABELS_PUSH,
			  "%s: can't push %u labels at once (maximum is 1)",
			  __func__, nh_label->num_labels);
		return -1;
	}

	memset(smpls, 0, sizeof(*smpls));
	smpls->smpls_len = sizeof(*smpls);
	smpls->smpls_family = AF_MPLS;
	smpls->smpls_label = htonl(nh_label->label[0] << MPLS_LABEL_OFFSET);

	return 0;
}
#endif

/* Interface between zebra message and rtm message. */
static int kernel_rtm(int cmd, const struct prefix *p,
		      const struct nexthop_group *ng, uint32_t metric)

{
	union sockunion sin_dest, sin_mask, sin_gate;
#ifdef __OpenBSD__
	struct sockaddr_mpls smpls;
#endif
	union sockunion *smplsp = NULL;
	struct nexthop *nexthop;
	int nexthop_num = 0;
	ifindex_t ifindex = 0;
	bool gate = false;
	int error;
	char gate_buf[INET6_BUFSIZ];
	enum blackhole_type bh_type = BLACKHOLE_UNSPEC;

	/*
	 * We only have the ability to ADD or DELETE at this point
	 * in time.
	 */
	if (cmd != RTM_ADD && cmd != RTM_DELETE) {
		if (IS_ZEBRA_DEBUG_KERNEL)
			zlog_debug("%s: %pFX odd command %s", __func__, p,
				   lookup_msg(rtm_type_str, cmd, NULL));
		return 0;
	}

	memset(&sin_dest, 0, sizeof(sin_dest));
	memset(&sin_gate, 0, sizeof(sin_gate));
	memset(&sin_mask, 0, sizeof(sin_mask));

	switch (p->family) {
	case AF_INET:
		sin_dest.sin.sin_family = AF_INET;
		sin_dest.sin.sin_addr = p->u.prefix4;
		sin_gate.sin.sin_family = AF_INET;
#ifdef HAVE_STRUCT_SOCKADDR_SA_LEN
		sin_dest.sin.sin_len = sizeof(struct sockaddr_in);
		sin_gate.sin.sin_len = sizeof(struct sockaddr_in);
#endif /* HAVE_STRUCT_SOCKADDR_SA_LEN */
		break;
	case AF_INET6:
		sin_dest.sin6.sin6_family = AF_INET6;
		sin_dest.sin6.sin6_addr = p->u.prefix6;
		sin_gate.sin6.sin6_family = AF_INET6;
#ifdef HAVE_STRUCT_SOCKADDR_SA_LEN
		sin_dest.sin6.sin6_len = sizeof(struct sockaddr_in6);
		sin_gate.sin6.sin6_len = sizeof(struct sockaddr_in6);
#endif /* HAVE_STRUCT_SOCKADDR_SA_LEN */
		break;
	}

	/* Make gateway. */
	for (ALL_NEXTHOPS_PTR(ng, nexthop)) {
		/*
		 * We only want to use the actual good nexthops
		 */
		if (CHECK_FLAG(nexthop->flags, NEXTHOP_FLAG_RECURSIVE) ||
		    !CHECK_FLAG(nexthop->flags, NEXTHOP_FLAG_ACTIVE))
			continue;

		smplsp = NULL;
		gate = false;
		snprintf(gate_buf, sizeof(gate_buf), "NULL");

		switch (nexthop->type) {
		case NEXTHOP_TYPE_IPV4:
		case NEXTHOP_TYPE_IPV4_IFINDEX:
			sin_gate.sin.sin_addr = nexthop->gate.ipv4;
			sin_gate.sin.sin_family = AF_INET;
#ifdef HAVE_STRUCT_SOCKADDR_SA_LEN
			sin_gate.sin.sin_len = sizeof(struct sockaddr_in);
#endif /* HAVE_STRUCT_SOCKADDR_SA_LEN */
			ifindex = nexthop->ifindex;
			gate = true;
			break;
		case NEXTHOP_TYPE_IPV6:
		case NEXTHOP_TYPE_IPV6_IFINDEX:
			sin_gate.sin6.sin6_addr = nexthop->gate.ipv6;
			sin_gate.sin6.sin6_family = AF_INET6;
#ifdef HAVE_STRUCT_SOCKADDR_SA_LEN
			sin_gate.sin6.sin6_len = sizeof(struct sockaddr_in6);
#endif /* HAVE_STRUCT_SOCKADDR_SA_LEN */
			ifindex = nexthop->ifindex;
/* Under kame set interface index to link local address */
#ifdef KAME

#define SET_IN6_LINKLOCAL_IFINDEX(a, i)                                        \
	do {                                                                   \
		(a).s6_addr[2] = ((i) >> 8) & 0xff;                            \
		(a).s6_addr[3] = (i)&0xff;                                     \
	} while (0)

			if (IN6_IS_ADDR_LINKLOCAL(&sin_gate.sin6.sin6_addr))
				SET_IN6_LINKLOCAL_IFINDEX(
					sin_gate.sin6.sin6_addr,
					ifindex);
#endif /* KAME */

			gate = true;
			break;
		case NEXTHOP_TYPE_IFINDEX:
			ifindex = nexthop->ifindex;
			break;
		case NEXTHOP_TYPE_BLACKHOLE:
			bh_type = nexthop->bh_type;
			switch (p->family) {
			case AF_INET: {
				struct in_addr loopback;

				loopback.s_addr = htonl(INADDR_LOOPBACK);
				sin_gate.sin.sin_addr = loopback;
#ifdef HAVE_STRUCT_SOCKADDR_SA_LEN
				sin_gate.sin.sin_len =
					sizeof(struct sockaddr_in);
#endif /* HAVE_STRUCT_SOCKADDR_SA_LEN */
				gate = true;
			} break;
			case AF_INET6: {
				struct in6_addr loopback;

				inet_pton(AF_INET6, "::1", &loopback);

				sin_gate.sin6.sin6_addr = loopback;
				sin_gate.sin6.sin6_family = AF_INET6;

#ifdef HAVE_STRUCTSOCKADDR_SA_LEN
				sin_gate.sin6.sin6_len =
					sizeof(struct sockaddr_in6);
#endif /* HAVE_STRUCTSOCKADDR_SA_LEN */
				gate = true;
			} break;
			}
		}

		switch (p->family) {
		case AF_INET:
			masklen2ip(p->prefixlen, &sin_mask.sin.sin_addr);
			sin_mask.sin.sin_family = AF_INET;
#ifdef HAVE_STRUCT_SOCKADDR_SA_LEN
			sin_mask.sin.sin_len = sizeof(struct sockaddr_in);
#endif /* HAVE_STRUCT_SOCKADDR_SA_LEN */
			break;
		case AF_INET6:
			masklen2ip6(p->prefixlen, &sin_mask.sin6.sin6_addr);
			sin_mask.sin6.sin6_family = AF_INET6;
#ifdef HAVE_STRUCT_SOCKADDR_SA_LEN
			sin_mask.sin6.sin6_len = sizeof(struct sockaddr_in6);
#endif /* HAVE_STRUCT_SOCKADDR_SA_LEN */
			break;
		}

#ifdef __OpenBSD__
		if (nexthop->nh_label) {
			if (kernel_rtm_add_labels(nexthop->nh_label,
						  &smpls) != 0)
				continue;
			smplsp = (union sockunion *)&smpls;
		}
#endif
		error = rtm_write(cmd, &sin_dest, &sin_mask,
				  gate ? &sin_gate : NULL, smplsp,
				  ifindex, bh_type, metric);

		if (IS_ZEBRA_DEBUG_KERNEL) {
			if (!gate) {
				zlog_debug(
					"%s: %pFX: attention! gate not found for re",
					__func__, p);
			} else {
				switch (p->family) {
				case AF_INET:
					inet_ntop(AF_INET,
						  &sin_gate.sin.sin_addr,
						  gate_buf, sizeof(gate_buf));
					break;

				case AF_INET6:
					inet_ntop(AF_INET6,
						  &sin_gate.sin6.sin6_addr,
						  gate_buf, sizeof(gate_buf));
					break;

				default:
					snprintf(gate_buf, sizeof(gate_buf),
						 "(invalid-af)");
					break;
				}
			}
		}
		switch (error) {
			/* We only flag nexthops as being in FIB if
			 * rtm_write() did its work. */
		case ZEBRA_ERR_NOERROR:
			nexthop_num++;
			if (IS_ZEBRA_DEBUG_KERNEL)
				zlog_debug("%s: %pFX: successfully did NH %s",
					   __func__, p, gate_buf);
			if (cmd == RTM_ADD)
				SET_FLAG(nexthop->flags, NEXTHOP_FLAG_FIB);
			break;

			/* The only valid case for this error is
			 * kernel's failure to install a multipath
			 * route, which is common for FreeBSD. This
			 * should be ignored silently, but logged as an error
			 * otherwise.
			 */
		case ZEBRA_ERR_RTEXIST:
			if (cmd != RTM_ADD)
				flog_err(EC_LIB_SYSTEM_CALL,
					 "%s: rtm_write() returned %d for command %d",
					 __func__, error, cmd);
			continue;

			/* Note any unexpected status returns */
		case ZEBRA_ERR_RTNOEXIST:
			if (cmd != RTM_DELETE)
				flog_err(EC_LIB_SYSTEM_CALL,
					 "%s: rtm_write() returned %d for command %d",
					 __func__, error, cmd);
			break;
		default:
			flog_err(
				EC_LIB_SYSTEM_CALL,
				"%s: %pFX: rtm_write() unexpectedly returned %d for command %s",
				__func__, p, error,
				lookup_msg(rtm_type_str, cmd, NULL));
			break;
		}
	} /* for (ALL_NEXTHOPS(...))*/

	/* If there was no useful nexthop, then complain. */
	if (nexthop_num == 0) {
		if (IS_ZEBRA_DEBUG_KERNEL)
			zlog_debug(
				"%s: No useful nexthops were found in RIB prefix %pFX",
				__func__, p);
		return 1;
	}

	return 0; /*XXX*/
}

/*
 * Update or delete a prefix from the kernel,
 * using info from a dataplane context struct.
 */
enum zebra_dplane_result kernel_route_update(struct zebra_dplane_ctx *ctx)
{
	enum zebra_dplane_result res = ZEBRA_DPLANE_REQUEST_SUCCESS;
	uint32_t type, old_type;

	if (dplane_ctx_get_src(ctx) != NULL) {
		zlog_err("route add: IPv6 sourcedest routes unsupported!");
		return ZEBRA_DPLANE_REQUEST_FAILURE;
	}

	type = dplane_ctx_get_type(ctx);
	old_type = dplane_ctx_get_old_type(ctx);

	frr_with_privs(&zserv_privs) {

		if (dplane_ctx_get_op(ctx) == DPLANE_OP_ROUTE_DELETE) {
			if (!RSYSTEM_ROUTE(type))
				kernel_rtm(RTM_DELETE, dplane_ctx_get_dest(ctx),
					   dplane_ctx_get_ng(ctx),
					   dplane_ctx_get_metric(ctx));
		} else if (dplane_ctx_get_op(ctx) == DPLANE_OP_ROUTE_INSTALL) {
			if (!RSYSTEM_ROUTE(type))
				kernel_rtm(RTM_ADD, dplane_ctx_get_dest(ctx),
					   dplane_ctx_get_ng(ctx),
					   dplane_ctx_get_metric(ctx));
		} else if (dplane_ctx_get_op(ctx) == DPLANE_OP_ROUTE_UPDATE) {
			/* Must do delete and add separately -
			 * no update available
			 */
			if (!RSYSTEM_ROUTE(old_type))
				kernel_rtm(RTM_DELETE, dplane_ctx_get_dest(ctx),
					   dplane_ctx_get_old_ng(ctx),
					   dplane_ctx_get_old_metric(ctx));

			if (!RSYSTEM_ROUTE(type))
				kernel_rtm(RTM_ADD, dplane_ctx_get_dest(ctx),
					   dplane_ctx_get_ng(ctx),
					   dplane_ctx_get_metric(ctx));
		} else {
			zlog_err("Invalid routing socket update op %s (%u)",
				 dplane_op2str(dplane_ctx_get_op(ctx)),
				 dplane_ctx_get_op(ctx));
			res = ZEBRA_DPLANE_REQUEST_FAILURE;
		}
	} /* Elevated privs */

	return res;
}

enum zebra_dplane_result kernel_nexthop_update(struct zebra_dplane_ctx *ctx)
{
	return ZEBRA_DPLANE_REQUEST_SUCCESS;
}

int kernel_neigh_register(vrf_id_t vrf_id, struct zserv *client, bool reg)
{
	/* TODO */
	return 0;
}

int kernel_neigh_update(int add, int ifindex, void *addr, char *lla, int llalen,
			ns_id_t ns_id, uint8_t family, bool permanent)
{
	/* TODO */
	return 0;
}

/* NYI on routing-socket platforms, but we've always returned 'success'... */
enum zebra_dplane_result kernel_neigh_update_ctx(struct zebra_dplane_ctx *ctx)
{
	return ZEBRA_DPLANE_REQUEST_SUCCESS;
}

extern int kernel_get_ipmr_sg_stats(struct zebra_vrf *zvrf, void *mroute)
{
	return 0;
}

/*
 * Update MAC, using dataplane context object. No-op here for now.
 */
enum zebra_dplane_result kernel_mac_update_ctx(struct zebra_dplane_ctx *ctx)
{
	return ZEBRA_DPLANE_REQUEST_SUCCESS;
}

extern int kernel_interface_set_master(struct interface *master,
				       struct interface *slave)
{
	return 0;
}

uint32_t kernel_get_speed(struct interface *ifp, int *error)
{
	return ifp->speed;
}

int kernel_upd_mac_nh(uint32_t nh_id, struct in_addr vtep_ip)
{
	return 0;
}

int kernel_del_mac_nh(uint32_t nh_id)
{
	return 0;
}

int kernel_upd_mac_nhg(uint32_t nhg_id, uint32_t nh_cnt,
		struct nh_grp *nh_ids)
{
	return 0;
}

int kernel_del_mac_nhg(uint32_t nhg_id)
{
	return 0;
}

#endif /* !HAVE_NETLINK */