forked from FRRouting/frr
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathkernel_netlink.c
2028 lines (1732 loc) · 53 KB
/
kernel_netlink.c
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
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
// SPDX-License-Identifier: GPL-2.0-or-later
/* Kernel communication using netlink interface.
* Copyright (C) 1999 Kunihiro Ishiguro
*/
#include <zebra.h>
#include <fcntl.h>
#ifdef HAVE_NETLINK
#include <linux/netlink.h>
#include <linux/rtnetlink.h>
#include <linux/filter.h>
#include "linklist.h"
#include "if.h"
#include "log.h"
#include "prefix.h"
#include "connected.h"
#include "table.h"
#include "memory.h"
#include "rib.h"
#include "frrevent.h"
#include "privs.h"
#include "nexthop.h"
#include "vrf.h"
#include "mpls.h"
#include "lib_errors.h"
#include "hash.h"
#include "zebra/zebra_router.h"
#include "zebra/zebra_ns.h"
#include "zebra/zebra_vrf.h"
#include "zebra/rt.h"
#include "zebra/debug.h"
#include "zebra/kernel_netlink.h"
#include "zebra/rt_netlink.h"
#include "zebra/if_netlink.h"
#include "zebra/rule_netlink.h"
#include "zebra/tc_netlink.h"
#include "zebra/netconf_netlink.h"
#include "zebra/zebra_errors.h"
#include "zebra/ge_netlink.h"
#ifndef SO_RCVBUFFORCE
#define SO_RCVBUFFORCE (33)
#endif
/* Hack for GNU libc version 2. */
#ifndef MSG_TRUNC
#define MSG_TRUNC 0x20
#endif /* MSG_TRUNC */
#ifndef NLMSG_TAIL
#define NLMSG_TAIL(nmsg) \
((struct rtattr *)(((uint8_t *)(nmsg)) \
+ NLMSG_ALIGN((nmsg)->nlmsg_len)))
#endif
#ifndef RTA_TAIL
#define RTA_TAIL(rta) \
((struct rtattr *)(((uint8_t *)(rta)) + RTA_ALIGN((rta)->rta_len)))
#endif
#ifndef RTNL_FAMILY_IP6MR
#define RTNL_FAMILY_IP6MR 129
#endif
#ifndef RTPROT_MROUTED
#define RTPROT_MROUTED 17
#endif
#define NL_DEFAULT_BATCH_BUFSIZE (16 * NL_PKT_BUF_SIZE)
/*
* We limit the batch's size to a number smaller than the length of the
* underlying buffer since the last message that wouldn't fit the batch would go
* over the upper boundary and then it would have to be encoded again into a new
* buffer. If the difference between the limit and the length of the buffer is
* big enough (bigger than the biggest Netlink message) then this situation
* won't occur.
*/
#define NL_DEFAULT_BATCH_SEND_THRESHOLD (15 * NL_PKT_BUF_SIZE)
static const struct message nlmsg_str[] = {
{ RTM_NEWROUTE, "RTM_NEWROUTE" },
{ RTM_DELROUTE, "RTM_DELROUTE" },
{ RTM_GETROUTE, "RTM_GETROUTE" },
{ RTM_NEWLINK, "RTM_NEWLINK" },
{ RTM_SETLINK, "RTM_SETLINK" },
{ RTM_DELLINK, "RTM_DELLINK" },
{ RTM_GETLINK, "RTM_GETLINK" },
{ RTM_NEWADDR, "RTM_NEWADDR" },
{ RTM_DELADDR, "RTM_DELADDR" },
{ RTM_GETADDR, "RTM_GETADDR" },
{ RTM_NEWNEIGH, "RTM_NEWNEIGH" },
{ RTM_DELNEIGH, "RTM_DELNEIGH" },
{ RTM_GETNEIGH, "RTM_GETNEIGH" },
{ RTM_NEWRULE, "RTM_NEWRULE" },
{ RTM_DELRULE, "RTM_DELRULE" },
{ RTM_GETRULE, "RTM_GETRULE" },
{ RTM_NEWNEXTHOP, "RTM_NEWNEXTHOP" },
{ RTM_DELNEXTHOP, "RTM_DELNEXTHOP" },
{ RTM_GETNEXTHOP, "RTM_GETNEXTHOP" },
{ RTM_NEWNETCONF, "RTM_NEWNETCONF" },
{ RTM_DELNETCONF, "RTM_DELNETCONF" },
{ RTM_NEWTUNNEL, "RTM_NEWTUNNEL" },
{ RTM_DELTUNNEL, "RTM_DELTUNNEL" },
{ RTM_GETTUNNEL, "RTM_GETTUNNEL" },
{ RTM_NEWQDISC, "RTM_NEWQDISC" },
{ RTM_DELQDISC, "RTM_DELQDISC" },
{ RTM_GETQDISC, "RTM_GETQDISC" },
{ RTM_NEWTCLASS, "RTM_NEWTCLASS" },
{ RTM_DELTCLASS, "RTM_DELTCLASS" },
{ RTM_GETTCLASS, "RTM_GETTCLASS" },
{ RTM_NEWTFILTER, "RTM_NEWTFILTER" },
{ RTM_DELTFILTER, "RTM_DELTFILTER" },
{ RTM_GETTFILTER, "RTM_GETTFILTER" },
{ RTM_NEWVLAN, "RTM_NEWVLAN" },
{ RTM_DELVLAN, "RTM_DELVLAN" },
{ RTM_GETVLAN, "RTM_GETVLAN" },
{ RTM_NEWCHAIN, "RTM_NEWCHAIN" },
{ RTM_DELCHAIN, "RTM_DELCHAIN" },
{ RTM_GETCHAIN, "RTM_GETCHAIN" },
{ 0 }
};
static const struct message rtproto_str[] = {
{RTPROT_REDIRECT, "redirect"},
{RTPROT_KERNEL, "kernel"},
{RTPROT_BOOT, "boot"},
{RTPROT_STATIC, "static"},
{RTPROT_GATED, "GateD"},
{RTPROT_RA, "router advertisement"},
{RTPROT_MRT, "MRT"},
{RTPROT_ZEBRA, "Zebra"},
#ifdef RTPROT_BIRD
{RTPROT_BIRD, "BIRD"},
#endif /* RTPROT_BIRD */
{RTPROT_MROUTED, "mroute"},
{RTPROT_BGP, "BGP"},
{RTPROT_OSPF, "OSPF"},
{RTPROT_ISIS, "IS-IS"},
{RTPROT_RIP, "RIP"},
{RTPROT_RIPNG, "RIPNG"},
{RTPROT_ZSTATIC, "static"},
{0}};
static const struct message family_str[] = {{AF_INET, "ipv4"},
{AF_INET6, "ipv6"},
{AF_BRIDGE, "bridge"},
{RTNL_FAMILY_IPMR, "ipv4MR"},
{RTNL_FAMILY_IP6MR, "ipv6MR"},
{0}};
static const struct message rttype_str[] = {{RTN_UNSPEC, "none"},
{RTN_UNICAST, "unicast"},
{RTN_LOCAL, "local"},
{RTN_BROADCAST, "broadcast"},
{RTN_ANYCAST, "anycast"},
{RTN_MULTICAST, "multicast"},
{RTN_BLACKHOLE, "blackhole"},
{RTN_UNREACHABLE, "unreachable"},
{RTN_PROHIBIT, "prohibited"},
{RTN_THROW, "throw"},
{RTN_NAT, "nat"},
{RTN_XRESOLVE, "resolver"},
{0}};
extern struct event_loop *master;
extern struct zebra_privs_t zserv_privs;
DEFINE_MTYPE_STATIC(ZEBRA, NL_BUF, "Zebra Netlink buffers");
/* Hashtable and mutex to allow lookup of nlsock structs by socket/fd value.
* We have both the main and dplane pthreads using these structs, so we have
* to protect the hash with a lock.
*/
static struct hash *nlsock_hash;
pthread_mutex_t nlsock_mutex;
/* Lock and unlock wrappers for nlsock hash */
#define NLSOCK_LOCK() pthread_mutex_lock(&nlsock_mutex)
#define NLSOCK_UNLOCK() pthread_mutex_unlock(&nlsock_mutex)
size_t nl_batch_tx_bufsize;
char *nl_batch_tx_buf;
_Atomic uint32_t nl_batch_bufsize = NL_DEFAULT_BATCH_BUFSIZE;
_Atomic uint32_t nl_batch_send_threshold = NL_DEFAULT_BATCH_SEND_THRESHOLD;
struct nl_batch {
void *buf;
size_t bufsiz;
size_t limit;
void *buf_head;
size_t curlen;
size_t msgcnt;
const struct zebra_dplane_info *zns;
struct dplane_ctx_list_head ctx_list;
/*
* Pointer to the queue of completed contexts outbound back
* towards the dataplane module.
*/
struct dplane_ctx_list_head *ctx_out_q;
};
int netlink_config_write_helper(struct vty *vty)
{
uint32_t size =
atomic_load_explicit(&nl_batch_bufsize, memory_order_relaxed);
uint32_t threshold = atomic_load_explicit(&nl_batch_send_threshold,
memory_order_relaxed);
if (size != NL_DEFAULT_BATCH_BUFSIZE
|| threshold != NL_DEFAULT_BATCH_SEND_THRESHOLD)
vty_out(vty, "zebra kernel netlink batch-tx-buf %u %u\n", size,
threshold);
if (if_netlink_frr_protodown_r_bit_is_set())
vty_out(vty, "zebra protodown reason-bit %u\n",
if_netlink_get_frr_protodown_r_bit());
return 0;
}
void netlink_set_batch_buffer_size(uint32_t size, uint32_t threshold, bool set)
{
if (!set) {
size = NL_DEFAULT_BATCH_BUFSIZE;
threshold = NL_DEFAULT_BATCH_SEND_THRESHOLD;
}
atomic_store_explicit(&nl_batch_bufsize, size, memory_order_relaxed);
atomic_store_explicit(&nl_batch_send_threshold, threshold,
memory_order_relaxed);
}
int netlink_talk_filter(struct nlmsghdr *h, ns_id_t ns_id, int startup)
{
/*
* This is an error condition that must be handled during
* development.
*
* The netlink_talk_filter function is used for communication
* down the netlink_cmd pipe and we are expecting
* an ack being received. So if we get here
* then we did not receive the ack and instead
* received some other message in an unexpected
* way.
*/
zlog_debug("%s: ignoring message type 0x%04x(%s) NS %u", __func__,
h->nlmsg_type, nl_msg_type_to_str(h->nlmsg_type), ns_id);
return 0;
}
static int netlink_recvbuf(struct nlsock *nl, uint32_t newsize)
{
uint32_t oldsize;
socklen_t newlen = sizeof(newsize);
socklen_t oldlen = sizeof(oldsize);
int ret;
ret = getsockopt(nl->sock, SOL_SOCKET, SO_RCVBUF, &oldsize, &oldlen);
if (ret < 0) {
flog_err_sys(EC_LIB_SOCKET,
"Can't get %s receive buffer size: %s", nl->name,
safe_strerror(errno));
return -1;
}
/* Try force option (linux >= 2.6.14) and fall back to normal set */
frr_with_privs(&zserv_privs) {
ret = setsockopt(nl->sock, SOL_SOCKET, SO_RCVBUFFORCE,
&rcvbufsize, sizeof(rcvbufsize));
}
if (ret < 0)
ret = setsockopt(nl->sock, SOL_SOCKET, SO_RCVBUF, &rcvbufsize,
sizeof(rcvbufsize));
if (ret < 0) {
flog_err_sys(EC_LIB_SOCKET,
"Can't set %s receive buffer size: %s", nl->name,
safe_strerror(errno));
return -1;
}
ret = getsockopt(nl->sock, SOL_SOCKET, SO_RCVBUF, &newsize, &newlen);
if (ret < 0) {
flog_err_sys(EC_LIB_SOCKET,
"Can't get %s receive buffer size: %s", nl->name,
safe_strerror(errno));
return -1;
}
return 0;
}
static const char *group2str(uint32_t group)
{
switch (group) {
case RTNLGRP_TUNNEL:
return "RTNLGRP_TUNNEL";
default:
return "UNKNOWN";
}
}
/* Make socket for Linux netlink interface. */
static int netlink_socket(struct nlsock *nl, unsigned long groups,
uint32_t ext_groups[], uint8_t ext_group_size,
ns_id_t ns_id, int nl_family)
{
int ret;
struct sockaddr_nl snl;
int sock;
int namelen;
frr_with_privs(&zserv_privs) {
sock = ns_socket(AF_NETLINK, SOCK_RAW, nl_family, ns_id);
if (sock < 0) {
zlog_err("Can't open %s socket: %s", nl->name,
safe_strerror(errno));
return -1;
}
memset(&snl, 0, sizeof(snl));
snl.nl_family = AF_NETLINK;
snl.nl_groups = groups;
if (ext_group_size) {
uint8_t i;
for (i = 0; i < ext_group_size; i++) {
#if defined SOL_NETLINK
ret = setsockopt(sock, SOL_NETLINK,
NETLINK_ADD_MEMBERSHIP,
&ext_groups[i],
sizeof(ext_groups[i]));
if (ret < 0) {
zlog_notice(
"can't setsockopt NETLINK_ADD_MEMBERSHIP for group %s(%u), this linux kernel does not support it: %s(%d)",
group2str(ext_groups[i]),
ext_groups[i],
safe_strerror(errno), errno);
}
#else
zlog_notice(
"Unable to use NETLINK_ADD_MEMBERSHIP via SOL_NETLINK for %s(%u) since the linux kernel does not support the socket option",
group2str(ext_groups[i]),
ext_groups[i]);
#endif
}
}
/* Bind the socket to the netlink structure for anything. */
ret = bind(sock, (struct sockaddr *)&snl, sizeof(snl));
}
if (ret < 0) {
zlog_err("Can't bind %s socket to group 0x%x: %s", nl->name,
snl.nl_groups, safe_strerror(errno));
close(sock);
return -1;
}
/* multiple netlink sockets will have different nl_pid */
namelen = sizeof(snl);
ret = getsockname(sock, (struct sockaddr *)&snl, (socklen_t *)&namelen);
if (ret < 0 || namelen != sizeof(snl)) {
flog_err_sys(EC_LIB_SOCKET, "Can't get %s socket name: %s",
nl->name, safe_strerror(errno));
close(sock);
return -1;
}
nl->snl = snl;
nl->sock = sock;
nl->buflen = NL_RCV_PKT_BUF_SIZE;
nl->buf = XMALLOC(MTYPE_NL_BUF, nl->buflen);
return ret;
}
/*
* Dispatch an incoming netlink message; used by the zebra main pthread's
* netlink event reader.
*/
static int netlink_information_fetch(struct nlmsghdr *h, ns_id_t ns_id,
int startup)
{
/*
* When we handle new message types here
* because we are starting to install them
* then lets check the netlink_install_filter
* and see if we should add the corresponding
* allow through entry there.
* Probably not needed to do but please
* think about it.
*/
switch (h->nlmsg_type) {
case RTM_NEWROUTE:
return netlink_route_change(h, ns_id, startup);
case RTM_DELROUTE:
return netlink_route_change(h, ns_id, startup);
case RTM_NEWLINK:
return netlink_link_change(h, ns_id, startup);
case RTM_DELLINK:
return 0;
case RTM_NEWNEIGH:
case RTM_DELNEIGH:
case RTM_GETNEIGH:
return netlink_neigh_change(h, ns_id);
case RTM_NEWRULE:
return netlink_rule_change(h, ns_id, startup);
case RTM_DELRULE:
return netlink_rule_change(h, ns_id, startup);
case RTM_NEWNEXTHOP:
return netlink_nexthop_change(h, ns_id, startup);
case RTM_DELNEXTHOP:
return netlink_nexthop_change(h, ns_id, startup);
case RTM_NEWQDISC:
case RTM_DELQDISC:
return netlink_qdisc_change(h, ns_id, startup);
case RTM_NEWTCLASS:
case RTM_DELTCLASS:
return netlink_tclass_change(h, ns_id, startup);
case RTM_NEWTFILTER:
case RTM_DELTFILTER:
return netlink_tfilter_change(h, ns_id, startup);
case RTM_NEWVLAN:
return netlink_vlan_change(h, ns_id, startup);
case RTM_DELVLAN:
return netlink_vlan_change(h, ns_id, startup);
/* Messages we may receive, but ignore */
case RTM_NEWCHAIN:
case RTM_DELCHAIN:
case RTM_GETCHAIN:
return 0;
/* Messages handled in the dplane thread */
case RTM_NEWADDR:
case RTM_DELADDR:
case RTM_NEWNETCONF:
case RTM_DELNETCONF:
case RTM_NEWTUNNEL:
case RTM_DELTUNNEL:
case RTM_GETTUNNEL:
return 0;
default:
/*
* If we have received this message then
* we have made a mistake during development
* and we need to write some code to handle
* this message type or not ask for
* it to be sent up to us
*/
flog_err(EC_ZEBRA_UNKNOWN_NLMSG,
"Unknown netlink nlmsg_type %s(%d) vrf %u",
nl_msg_type_to_str(h->nlmsg_type), h->nlmsg_type,
ns_id);
break;
}
return 0;
}
/*
* Dispatch an incoming netlink message; used by the dataplane pthread's
* netlink event reader code.
*/
static int dplane_netlink_information_fetch(struct nlmsghdr *h, ns_id_t ns_id,
int startup)
{
/*
* Dispatch the incoming messages that the dplane pthread handles
*/
switch (h->nlmsg_type) {
case RTM_NEWADDR:
case RTM_DELADDR:
return netlink_interface_addr_dplane(h, ns_id, startup);
case RTM_NEWNETCONF:
case RTM_DELNETCONF:
return netlink_netconf_change(h, ns_id, startup);
/* TODO -- other messages for the dplane socket and pthread */
case RTM_NEWLINK:
case RTM_DELLINK:
return netlink_link_change(h, ns_id, startup);
default:
break;
}
return 0;
}
static void kernel_read(struct event *thread)
{
struct zebra_ns *zns = (struct zebra_ns *)EVENT_ARG(thread);
struct zebra_dplane_info dp_info;
/* Capture key info from ns struct */
zebra_dplane_info_from_zns(&dp_info, zns, false);
netlink_parse_info(netlink_information_fetch, &zns->netlink, &dp_info,
5, false);
event_add_read(zrouter.master, kernel_read, zns, zns->netlink.sock,
&zns->t_netlink);
}
/*
* Called by the dplane pthread to read incoming OS messages and dispatch them.
*/
int kernel_dplane_read(struct zebra_dplane_info *info)
{
struct nlsock *nl = kernel_netlink_nlsock_lookup(info->sock);
netlink_parse_info(dplane_netlink_information_fetch, nl, info, 5,
false);
return 0;
}
/*
* Filter out messages from self that occur on listener socket,
* caused by our actions on the command socket(s)
*
* When we add new Netlink message types we probably
* do not need to add them here as that we are filtering
* on the routes we actually care to receive( which is rarer
* then the normal course of operations). We are intentionally
* allowing some messages from ourselves through
* ( I'm looking at you Interface based netlink messages )
* so that we only have to write one way to handle incoming
* address add/delete and xxxNETCONF changes.
*/
static void netlink_install_filter(int sock, uint32_t pid, uint32_t dplane_pid)
{
/*
* BPF_JUMP instructions and where you jump to are based upon
* 0 as being the next statement. So count from 0. Writing
* this down because every time I look at this I have to
* re-remember it.
*/
struct sock_filter filter[] = {
/*
* Logic:
* if (nlmsg_pid == pid ||
* nlmsg_pid == dplane_pid) {
* if (the incoming nlmsg_type ==
* RTM_NEWADDR || RTM_DELADDR || RTM_NEWNETCONF ||
* RTM_DELNETCONF)
* keep this message
* else
* skip this message
* } else
* keep this netlink message
*/
/*
* 0: Load the nlmsg_pid into the BPF register
*/
BPF_STMT(BPF_LD | BPF_ABS | BPF_W,
offsetof(struct nlmsghdr, nlmsg_pid)),
/*
* 1: Compare to pid
*/
BPF_JUMP(BPF_JMP | BPF_JEQ | BPF_K, htonl(pid), 1, 0),
/*
* 2: Compare to dplane pid
*/
BPF_JUMP(BPF_JMP | BPF_JEQ | BPF_K, htonl(dplane_pid), 0, 6),
/*
* 3: Load the nlmsg_type into BPF register
*/
BPF_STMT(BPF_LD | BPF_ABS | BPF_H,
offsetof(struct nlmsghdr, nlmsg_type)),
/*
* 4: Compare to RTM_NEWADDR
*/
BPF_JUMP(BPF_JMP | BPF_JEQ | BPF_K, htons(RTM_NEWADDR), 4, 0),
/*
* 5: Compare to RTM_DELADDR
*/
BPF_JUMP(BPF_JMP | BPF_JEQ | BPF_K, htons(RTM_DELADDR), 3, 0),
/*
* 6: Compare to RTM_NEWNETCONF
*/
BPF_JUMP(BPF_JMP | BPF_JEQ | BPF_K, htons(RTM_NEWNETCONF), 2,
0),
/*
* 7: Compare to RTM_DELNETCONF
*/
BPF_JUMP(BPF_JMP | BPF_JEQ | BPF_K, htons(RTM_DELNETCONF), 1,
0),
/*
* 8: This is the end state of we want to skip the
* message
*/
BPF_STMT(BPF_RET | BPF_K, 0),
/* 9: This is the end state of we want to keep
* the message
*/
BPF_STMT(BPF_RET | BPF_K, 0xffff),
};
struct sock_fprog prog = {
.len = array_size(filter), .filter = filter,
};
if (setsockopt(sock, SOL_SOCKET, SO_ATTACH_FILTER, &prog, sizeof(prog))
< 0)
flog_err_sys(EC_LIB_SOCKET, "Can't install socket filter: %s",
safe_strerror(errno));
}
/*
* Please note, the assumption with this function is that the
* flags passed in that are bit masked with type, we are implicitly
* assuming that this is handling the NLA_F_NESTED ilk.
*/
void netlink_parse_rtattr_flags(struct rtattr **tb, int max, struct rtattr *rta,
int len, unsigned short flags)
{
unsigned short type;
memset(tb, 0, sizeof(struct rtattr *) * (max + 1));
while (RTA_OK(rta, len)) {
type = rta->rta_type & ~flags;
if ((type <= max) && (!tb[type]))
tb[type] = rta;
rta = RTA_NEXT(rta, len);
}
}
void netlink_parse_rtattr(struct rtattr **tb, int max, struct rtattr *rta,
int len)
{
memset(tb, 0, sizeof(struct rtattr *) * (max + 1));
while (RTA_OK(rta, len)) {
/*
* The type may be &'ed with NLA_F_NESTED
* which puts data in the upper 8 bits of the
* rta_type. Mask it off and save the actual
* underlying value to be placed into the array.
* This way we don't accidently crash in the future
* when the kernel sends us new data and we try
* to write well beyond the end of the array.
*/
uint16_t type = rta->rta_type & NLA_TYPE_MASK;
if (type <= max)
tb[type] = rta;
rta = RTA_NEXT(rta, len);
}
}
/**
* netlink_parse_rtattr_nested() - Parses a nested route attribute
* @tb: Pointer to array for storing rtattr in.
* @max: Max number to store.
* @rta: Pointer to rtattr to look for nested items in.
*/
void netlink_parse_rtattr_nested(struct rtattr **tb, int max,
struct rtattr *rta)
{
netlink_parse_rtattr(tb, max, RTA_DATA(rta), RTA_PAYLOAD(rta));
}
bool nl_attr_put(struct nlmsghdr *n, unsigned int maxlen, int type,
const void *data, unsigned int alen)
{
int len;
struct rtattr *rta;
len = RTA_LENGTH(alen);
if (NLMSG_ALIGN(n->nlmsg_len) + RTA_ALIGN(len) > maxlen)
return false;
rta = (struct rtattr *)(((char *)n) + NLMSG_ALIGN(n->nlmsg_len));
rta->rta_type = type;
rta->rta_len = len;
if (data)
memcpy(RTA_DATA(rta), data, alen);
else
assert(alen == 0);
n->nlmsg_len = NLMSG_ALIGN(n->nlmsg_len) + RTA_ALIGN(len);
return true;
}
bool nl_attr_put8(struct nlmsghdr *n, unsigned int maxlen, int type,
uint8_t data)
{
return nl_attr_put(n, maxlen, type, &data, sizeof(uint8_t));
}
bool nl_attr_put16(struct nlmsghdr *n, unsigned int maxlen, int type,
uint16_t data)
{
return nl_attr_put(n, maxlen, type, &data, sizeof(uint16_t));
}
bool nl_attr_put32(struct nlmsghdr *n, unsigned int maxlen, int type,
uint32_t data)
{
return nl_attr_put(n, maxlen, type, &data, sizeof(uint32_t));
}
bool nl_attr_put64(struct nlmsghdr *n, unsigned int maxlen, int type,
uint64_t data)
{
return nl_attr_put(n, maxlen, type, &data, sizeof(uint64_t));
}
struct rtattr *nl_attr_nest(struct nlmsghdr *n, unsigned int maxlen, int type)
{
struct rtattr *nest = NLMSG_TAIL(n);
if (!nl_attr_put(n, maxlen, type, NULL, 0))
return NULL;
nest->rta_type |= NLA_F_NESTED;
return nest;
}
int nl_attr_nest_end(struct nlmsghdr *n, struct rtattr *nest)
{
nest->rta_len = (uint8_t *)NLMSG_TAIL(n) - (uint8_t *)nest;
return n->nlmsg_len;
}
struct rtnexthop *nl_attr_rtnh(struct nlmsghdr *n, unsigned int maxlen)
{
struct rtnexthop *rtnh = (struct rtnexthop *)NLMSG_TAIL(n);
if (NLMSG_ALIGN(n->nlmsg_len) + RTNH_ALIGN(sizeof(struct rtnexthop))
> maxlen)
return NULL;
memset(rtnh, 0, sizeof(struct rtnexthop));
n->nlmsg_len =
NLMSG_ALIGN(n->nlmsg_len) + RTA_ALIGN(sizeof(struct rtnexthop));
return rtnh;
}
void nl_attr_rtnh_end(struct nlmsghdr *n, struct rtnexthop *rtnh)
{
rtnh->rtnh_len = (uint8_t *)NLMSG_TAIL(n) - (uint8_t *)rtnh;
}
const char *nl_msg_type_to_str(uint16_t msg_type)
{
return lookup_msg(nlmsg_str, msg_type, "");
}
const char *nl_rtproto_to_str(uint8_t rtproto)
{
return lookup_msg(rtproto_str, rtproto, "");
}
const char *nl_family_to_str(uint8_t family)
{
return lookup_msg(family_str, family, "");
}
const char *nl_rttype_to_str(uint8_t rttype)
{
return lookup_msg(rttype_str, rttype, "");
}
#define NLA_OK(nla, len) \
((len) >= (int)sizeof(struct nlattr) \
&& (nla)->nla_len >= sizeof(struct nlattr) \
&& (nla)->nla_len <= (len))
#define NLA_NEXT(nla, attrlen) \
((attrlen) -= NLA_ALIGN((nla)->nla_len), \
(struct nlattr *)(((char *)(nla)) + NLA_ALIGN((nla)->nla_len)))
#define NLA_LENGTH(len) (NLA_ALIGN(sizeof(struct nlattr)) + (len))
#define NLA_DATA(nla) ((struct nlattr *)(((char *)(nla)) + NLA_LENGTH(0)))
#define ERR_NLA(err, inner_len) \
((struct nlattr *)(((char *)(err)) \
+ NLMSG_ALIGN(sizeof(struct nlmsgerr)) \
+ NLMSG_ALIGN((inner_len))))
static void netlink_parse_nlattr(struct nlattr **tb, int max,
struct nlattr *nla, int len)
{
while (NLA_OK(nla, len)) {
if (nla->nla_type <= max)
tb[nla->nla_type] = nla;
nla = NLA_NEXT(nla, len);
}
}
static void netlink_parse_extended_ack(struct nlmsghdr *h)
{
struct nlattr *tb[NLMSGERR_ATTR_MAX + 1] = {};
const struct nlmsgerr *err = (const struct nlmsgerr *)NLMSG_DATA(h);
const struct nlmsghdr *err_nlh = NULL;
/* Length not including nlmsghdr */
uint32_t len = 0;
/* Inner error netlink message length */
uint32_t inner_len = 0;
const char *msg = NULL;
uint32_t off = 0;
if (!(h->nlmsg_flags & NLM_F_CAPPED))
inner_len = (uint32_t)NLMSG_PAYLOAD(&err->msg, 0);
len = (uint32_t)(NLMSG_PAYLOAD(h, sizeof(struct nlmsgerr)) - inner_len);
netlink_parse_nlattr(tb, NLMSGERR_ATTR_MAX, ERR_NLA(err, inner_len),
len);
if (tb[NLMSGERR_ATTR_MSG])
msg = (const char *)NLA_DATA(tb[NLMSGERR_ATTR_MSG]);
if (tb[NLMSGERR_ATTR_OFFS]) {
off = *(uint32_t *)NLA_DATA(tb[NLMSGERR_ATTR_OFFS]);
if (off > h->nlmsg_len) {
zlog_err("Invalid offset for NLMSGERR_ATTR_OFFS");
} else if (!(h->nlmsg_flags & NLM_F_CAPPED)) {
/*
* Header of failed message
* we are not doing anything currently with it
* but noticing it for later.
*/
err_nlh = &err->msg;
zlog_debug("%s: Received %s extended Ack", __func__,
nl_msg_type_to_str(err_nlh->nlmsg_type));
}
}
if (msg && *msg != '\0') {
bool is_err = !!err->error;
if (is_err)
zlog_err("Extended Error: %s", msg);
else
flog_warn(EC_ZEBRA_NETLINK_EXTENDED_WARNING,
"Extended Warning: %s", msg);
}
}
/*
* netlink_send_msg - send a netlink message of a certain size.
*
* Returns -1 on error. Otherwise, it returns the number of bytes sent.
*/
static ssize_t netlink_send_msg(const struct nlsock *nl, void *buf,
size_t buflen)
{
struct sockaddr_nl snl = {};
struct iovec iov = {};
struct msghdr msg = {};
ssize_t status;
int save_errno = 0;
iov.iov_base = buf;
iov.iov_len = buflen;
msg.msg_name = &snl;
msg.msg_namelen = sizeof(snl);
msg.msg_iov = &iov;
msg.msg_iovlen = 1;
snl.nl_family = AF_NETLINK;
/* Send message to netlink interface. */
frr_with_privs(&zserv_privs) {
status = sendmsg(nl->sock, &msg, 0);
save_errno = errno;
}
if (IS_ZEBRA_DEBUG_KERNEL_MSGDUMP_SEND) {
zlog_debug("%s: >> netlink message dump [sent]", __func__);
#ifdef NETLINK_DEBUG
nl_dump(buf, buflen);
#else
zlog_hexdump(buf, buflen);
#endif /* NETLINK_DEBUG */
}
if (status == -1) {
flog_err_sys(EC_LIB_SOCKET, "%s error: %s", __func__,
safe_strerror(save_errno));
return -1;
}
return status;
}
/*
* netlink_recv_msg - receive a netlink message.
*
* Returns -1 on error, 0 if read would block or the number of bytes received.
*/
static int netlink_recv_msg(struct nlsock *nl, struct msghdr *msg)
{
struct iovec iov;
int status;
iov.iov_base = nl->buf;
iov.iov_len = nl->buflen;
msg->msg_iov = &iov;
msg->msg_iovlen = 1;
do {
int bytes;
bytes = recv(nl->sock, NULL, 0, MSG_PEEK | MSG_TRUNC);
if (bytes >= 0 && (size_t)bytes > nl->buflen) {
nl->buf = XREALLOC(MTYPE_NL_BUF, nl->buf, bytes);
nl->buflen = bytes;
iov.iov_base = nl->buf;
iov.iov_len = nl->buflen;
}
status = recvmsg(nl->sock, msg, 0);
} while (status == -1 && errno == EINTR);
if (status == -1) {
if (errno == EWOULDBLOCK || errno == EAGAIN)
return 0;
flog_err(EC_ZEBRA_RECVMSG_OVERRUN, "%s recvmsg overrun: %s",
nl->name, safe_strerror(errno));
/*
* In this case we are screwed. There is no good way to recover
* zebra at this point.
*/
exit(-1);
}
if (status == 0) {
flog_err_sys(EC_LIB_SOCKET, "%s EOF", nl->name);
return -1;
}
if (msg->msg_namelen != sizeof(struct sockaddr_nl)) {
flog_err(EC_ZEBRA_NETLINK_LENGTH_ERROR,
"%s sender address length error: length %d", nl->name,
msg->msg_namelen);
return -1;
}
if (IS_ZEBRA_DEBUG_KERNEL_MSGDUMP_RECV) {
zlog_debug("%s: << netlink message dump [recv]", __func__);
#ifdef NETLINK_DEBUG
nl_dump(nl->buf, status);
#else
zlog_hexdump(nl->buf, status);
#endif /* NETLINK_DEBUG */
}
return status;
}
/*
* netlink_parse_error - parse a netlink error message
*
* Returns 1 if this message is acknowledgement, 0 if this error should be
* ignored, -1 otherwise.
*/
static int netlink_parse_error(const struct nlsock *nl, struct nlmsghdr *h,
bool is_cmd, bool startup)
{
struct nlmsgerr *err = (struct nlmsgerr *)NLMSG_DATA(h);
int errnum = err->error;
int msg_type = err->msg.nlmsg_type;
if (h->nlmsg_len < NLMSG_LENGTH(sizeof(struct nlmsgerr))) {
flog_err(EC_ZEBRA_NETLINK_LENGTH_ERROR,
"%s error: message truncated", nl->name);
return -1;
}
/*
* Parse the extended information before we actually handle it. At this
* point in time we do not do anything other than report the issue.
*/
if (h->nlmsg_flags & NLM_F_ACK_TLVS)
netlink_parse_extended_ack(h);
/* If the error field is zero, then this is an ACK. */
if (err->error == 0) {
if (IS_ZEBRA_DEBUG_KERNEL) {
zlog_debug("%s: %s ACK: type=%s(%u), seq=%u, pid=%u",
__func__, nl->name,