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
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
|
/*
* pim_bsm.c: PIM BSM handling routines
*
* Copyright (C) 2018-19 Vmware, Inc.
* Saravanan K
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful, but
* WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; see the file COPYING; if not, write to the
* Free Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston,
* MA 02110-1301 USA
*/
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#include "if.h"
#include "pimd.h"
#include "pim_iface.h"
#include "pim_instance.h"
#include "pim_neighbor.h"
#include "pim_rpf.h"
#include "pim_hello.h"
#include "pim_pim.h"
#include "pim_nht.h"
#include "pim_bsm.h"
#include "pim_time.h"
#include "pim_zebra.h"
/* Functions forward declaration */
static void pim_bs_timer_start(struct bsm_scope *scope, int bs_timeout);
static void pim_g2rp_timer_start(struct bsm_rpinfo *bsrp, int hold_time);
static inline void pim_g2rp_timer_restart(struct bsm_rpinfo *bsrp,
int hold_time);
/* Memory Types */
DEFINE_MTYPE_STATIC(PIMD, PIM_BSGRP_NODE, "PIM BSR advertised grp info");
DEFINE_MTYPE_STATIC(PIMD, PIM_BSRP_INFO, "PIM BSR advertised RP info");
DEFINE_MTYPE_STATIC(PIMD, PIM_BSM_FRAG, "PIM BSM fragment");
DEFINE_MTYPE_STATIC(PIMD, PIM_BSM_PKT_VAR_MEM, "PIM BSM Packet");
/* All bsm packets forwarded shall be fit within ip mtu less iphdr(max) */
#define MAX_IP_HDR_LEN 24
/* pim_bsm_write_config - Write the interface pim bsm configuration.*/
void pim_bsm_write_config(struct vty *vty, struct interface *ifp)
{
struct pim_interface *pim_ifp = ifp->info;
if (pim_ifp) {
if (!pim_ifp->bsm_enable)
vty_out(vty, " no " PIM_AF_NAME " pim bsm\n");
if (!pim_ifp->ucast_bsm_accept)
vty_out(vty, " no " PIM_AF_NAME " pim unicast-bsm\n");
}
}
static void pim_bsm_rpinfo_free(struct bsm_rpinfo *bsrp_info)
{
THREAD_OFF(bsrp_info->g2rp_timer);
XFREE(MTYPE_PIM_BSRP_INFO, bsrp_info);
}
static void pim_bsm_rpinfos_free(struct bsm_rpinfos_head *head)
{
struct bsm_rpinfo *bsrp_info;
while ((bsrp_info = bsm_rpinfos_pop(head)))
pim_bsm_rpinfo_free(bsrp_info);
}
static void pim_free_bsgrp_data(struct bsgrp_node *bsgrp_node)
{
pim_bsm_rpinfos_free(bsgrp_node->bsrp_list);
pim_bsm_rpinfos_free(bsgrp_node->partial_bsrp_list);
XFREE(MTYPE_PIM_BSGRP_NODE, bsgrp_node);
}
static void pim_free_bsgrp_node(struct route_table *rt, struct prefix *grp)
{
struct route_node *rn;
rn = route_node_lookup(rt, grp);
if (rn) {
rn->info = NULL;
route_unlock_node(rn);
route_unlock_node(rn);
}
}
static void pim_bsm_frag_free(struct bsm_frag *bsfrag)
{
XFREE(MTYPE_PIM_BSM_FRAG, bsfrag);
}
static void pim_bsm_frags_free(struct bsm_scope *scope)
{
struct bsm_frag *bsfrag;
while ((bsfrag = bsm_frags_pop(scope->bsm_frags)))
pim_bsm_frag_free(bsfrag);
}
int pim_bsm_rpinfo_cmp(const struct bsm_rpinfo *node1,
const struct bsm_rpinfo *node2)
{
/* RP election Algo :
* Step-1 : Loweset Rp priority will have higher precedance.
* Step-2 : If priority same then higher hash val will have
* higher precedance.
* Step-3 : If Hash val is same then highest rp address will
* become elected RP.
*/
if (node1->rp_prio < node2->rp_prio)
return -1;
if (node1->rp_prio > node2->rp_prio)
return 1;
if (node1->hash < node2->hash)
return 1;
if (node1->hash > node2->hash)
return -1;
if (node1->rp_address.s_addr < node2->rp_address.s_addr)
return 1;
if (node1->rp_address.s_addr > node2->rp_address.s_addr)
return -1;
return 0;
}
static struct bsgrp_node *pim_bsm_new_bsgrp_node(struct route_table *rt,
struct prefix *grp)
{
struct route_node *rn;
struct bsgrp_node *bsgrp;
rn = route_node_get(rt, grp);
if (!rn) {
zlog_warn("%s: route node creation failed", __func__);
return NULL;
}
bsgrp = XCALLOC(MTYPE_PIM_BSGRP_NODE, sizeof(struct bsgrp_node));
rn->info = bsgrp;
bsm_rpinfos_init(bsgrp->bsrp_list);
bsm_rpinfos_init(bsgrp->partial_bsrp_list);
prefix_copy(&bsgrp->group, grp);
return bsgrp;
}
static void pim_on_bs_timer(struct thread *t)
{
struct route_node *rn;
struct bsm_scope *scope;
struct bsgrp_node *bsgrp_node;
struct bsm_rpinfo *bsrp;
scope = THREAD_ARG(t);
THREAD_OFF(scope->bs_timer);
if (PIM_DEBUG_BSM)
zlog_debug("%s: Bootstrap Timer expired for scope: %d",
__func__, scope->sz_id);
pim_nht_bsr_del(scope->pim, scope->current_bsr);
/* Reset scope zone data */
scope->accept_nofwd_bsm = false;
scope->state = ACCEPT_ANY;
scope->current_bsr.s_addr = INADDR_ANY;
scope->current_bsr_prio = 0;
scope->current_bsr_first_ts = 0;
scope->current_bsr_last_ts = 0;
scope->bsm_frag_tag = 0;
pim_bsm_frags_free(scope);
for (rn = route_top(scope->bsrp_table); rn; rn = route_next(rn)) {
bsgrp_node = (struct bsgrp_node *)rn->info;
if (!bsgrp_node) {
if (PIM_DEBUG_BSM)
zlog_debug("%s: bsgrp_node is null", __func__);
continue;
}
/* Give grace time for rp to continue for another hold time */
bsrp = bsm_rpinfos_first(bsgrp_node->bsrp_list);
if (bsrp)
pim_g2rp_timer_restart(bsrp, bsrp->rp_holdtime);
/* clear pending list */
pim_bsm_rpinfos_free(bsgrp_node->partial_bsrp_list);
bsgrp_node->pend_rp_cnt = 0;
}
}
static void pim_bs_timer_stop(struct bsm_scope *scope)
{
if (PIM_DEBUG_BSM)
zlog_debug("%s : BS timer being stopped of sz: %d", __func__,
scope->sz_id);
THREAD_OFF(scope->bs_timer);
}
static void pim_bs_timer_start(struct bsm_scope *scope, int bs_timeout)
{
if (!scope) {
if (PIM_DEBUG_BSM)
zlog_debug("%s : Invalid scope(NULL).", __func__);
return;
}
THREAD_OFF(scope->bs_timer);
if (PIM_DEBUG_BSM)
zlog_debug(
"%s : starting bs timer for scope %d with timeout %d secs",
__func__, scope->sz_id, bs_timeout);
thread_add_timer(router->master, pim_on_bs_timer, scope, bs_timeout,
&scope->bs_timer);
}
static inline void pim_bs_timer_restart(struct bsm_scope *scope, int bs_timeout)
{
pim_bs_timer_start(scope, bs_timeout);
}
void pim_bsm_proc_init(struct pim_instance *pim)
{
memset(&pim->global_scope, 0, sizeof(struct bsm_scope));
pim->global_scope.sz_id = PIM_GBL_SZ_ID;
pim->global_scope.bsrp_table = route_table_init();
pim->global_scope.accept_nofwd_bsm = true;
pim->global_scope.state = NO_INFO;
pim->global_scope.pim = pim;
bsm_frags_init(pim->global_scope.bsm_frags);
pim_bs_timer_start(&pim->global_scope, PIM_BS_TIME);
}
void pim_bsm_proc_free(struct pim_instance *pim)
{
struct route_node *rn;
struct bsgrp_node *bsgrp;
pim_bs_timer_stop(&pim->global_scope);
pim_bsm_frags_free(&pim->global_scope);
for (rn = route_top(pim->global_scope.bsrp_table); rn;
rn = route_next(rn)) {
bsgrp = rn->info;
if (!bsgrp)
continue;
pim_free_bsgrp_data(bsgrp);
}
route_table_finish(pim->global_scope.bsrp_table);
}
static bool is_hold_time_elapsed(void *data)
{
struct bsm_rpinfo *bsrp;
bsrp = data;
if (bsrp->elapse_time < bsrp->rp_holdtime)
return false;
else
return true;
}
static void pim_on_g2rp_timer(struct thread *t)
{
struct bsm_rpinfo *bsrp;
struct bsm_rpinfo *bsrp_node;
struct bsgrp_node *bsgrp_node;
struct pim_instance *pim;
struct rp_info *rp_info;
struct route_node *rn;
uint16_t elapse;
struct in_addr bsrp_addr;
bsrp = THREAD_ARG(t);
THREAD_OFF(bsrp->g2rp_timer);
bsgrp_node = bsrp->bsgrp_node;
/* elapse time is the hold time of expired node */
elapse = bsrp->rp_holdtime;
bsrp_addr = bsrp->rp_address;
/* update elapse for all bsrp nodes */
frr_each_safe (bsm_rpinfos, bsgrp_node->bsrp_list, bsrp_node) {
bsrp_node->elapse_time += elapse;
if (is_hold_time_elapsed(bsrp_node)) {
bsm_rpinfos_del(bsgrp_node->bsrp_list, bsrp_node);
pim_bsm_rpinfo_free(bsrp_node);
}
}
/* Get the next elected rp node */
bsrp = bsm_rpinfos_first(bsgrp_node->bsrp_list);
pim = bsgrp_node->scope->pim;
rn = route_node_lookup(pim->rp_table, &bsgrp_node->group);
if (!rn) {
zlog_warn("%s: Route node doesn't exist", __func__);
return;
}
rp_info = (struct rp_info *)rn->info;
if (!rp_info) {
route_unlock_node(rn);
return;
}
if (rp_info->rp_src != RP_SRC_STATIC) {
/* If new rp available, change it else delete the existing */
if (bsrp) {
pim_g2rp_timer_start(
bsrp, (bsrp->rp_holdtime - bsrp->elapse_time));
pim_rp_change(pim, bsrp->rp_address, bsgrp_node->group,
RP_SRC_BSR);
} else {
pim_rp_del(pim, bsrp_addr, bsgrp_node->group, NULL,
RP_SRC_BSR);
}
}
if (!bsm_rpinfos_count(bsgrp_node->bsrp_list)
&& !bsm_rpinfos_count(bsgrp_node->partial_bsrp_list)) {
pim_free_bsgrp_node(pim->global_scope.bsrp_table,
&bsgrp_node->group);
pim_free_bsgrp_data(bsgrp_node);
}
}
static void pim_g2rp_timer_start(struct bsm_rpinfo *bsrp, int hold_time)
{
if (!bsrp) {
if (PIM_DEBUG_BSM)
zlog_debug("%s : Invalid brsp(NULL).", __func__);
return;
}
THREAD_OFF(bsrp->g2rp_timer);
if (PIM_DEBUG_BSM)
zlog_debug(
"%s : starting g2rp timer for grp: %pFX - rp: %pI4 with timeout %d secs(Actual Hold time : %d secs)",
__func__, &bsrp->bsgrp_node->group,
&bsrp->rp_address, hold_time,
bsrp->rp_holdtime);
thread_add_timer(router->master, pim_on_g2rp_timer, bsrp, hold_time,
&bsrp->g2rp_timer);
}
static inline void pim_g2rp_timer_restart(struct bsm_rpinfo *bsrp,
int hold_time)
{
pim_g2rp_timer_start(bsrp, hold_time);
}
static void pim_g2rp_timer_stop(struct bsm_rpinfo *bsrp)
{
if (!bsrp)
return;
if (PIM_DEBUG_BSM)
zlog_debug("%s : stopping g2rp timer for grp: %pFX - rp: %pI4",
__func__, &bsrp->bsgrp_node->group,
&bsrp->rp_address);
THREAD_OFF(bsrp->g2rp_timer);
}
static bool is_hold_time_zero(void *data)
{
struct bsm_rpinfo *bsrp;
bsrp = data;
if (bsrp->rp_holdtime)
return false;
else
return true;
}
static void pim_instate_pend_list(struct bsgrp_node *bsgrp_node)
{
struct bsm_rpinfo *active;
struct bsm_rpinfo *pend;
struct rp_info *rp_info;
struct route_node *rn;
struct pim_instance *pim;
struct rp_info *rp_all;
struct prefix group_all;
bool had_rp_node = true;
pim = bsgrp_node->scope->pim;
active = bsm_rpinfos_first(bsgrp_node->bsrp_list);
/* Remove nodes with hold time 0 & check if list still has a head */
frr_each_safe (bsm_rpinfos, bsgrp_node->partial_bsrp_list, pend) {
if (is_hold_time_zero(pend)) {
bsm_rpinfos_del(bsgrp_node->partial_bsrp_list, pend);
pim_bsm_rpinfo_free(pend);
}
}
pend = bsm_rpinfos_first(bsgrp_node->partial_bsrp_list);
if (!str2prefix("224.0.0.0/4", &group_all))
return;
rp_all = pim_rp_find_match_group(pim, &group_all);
rn = route_node_lookup(pim->rp_table, &bsgrp_node->group);
if (pend)
pim_g2rp_timer_start(pend, pend->rp_holdtime);
/* if rp node doesn't exist or exist but not configured(rp_all),
* install the rp from head(if exists) of partial list. List is
* is sorted such that head is the elected RP for the group.
*/
if (!rn || (prefix_same(&rp_all->group, &bsgrp_node->group) &&
pim_rpf_addr_is_inaddr_any(&rp_all->rp))) {
if (PIM_DEBUG_BSM)
zlog_debug("%s: Route node doesn't exist", __func__);
if (pend)
pim_rp_new(pim, pend->rp_address, bsgrp_node->group,
NULL, RP_SRC_BSR);
had_rp_node = false;
} else {
rp_info = (struct rp_info *)rn->info;
if (!rp_info) {
route_unlock_node(rn);
if (pend)
pim_rp_new(pim, pend->rp_address,
bsgrp_node->group, NULL, RP_SRC_BSR);
had_rp_node = false;
}
}
/* We didn't have rp node and pending list is empty(unlikely), cleanup*/
if ((!had_rp_node) && (!pend)) {
pim_free_bsgrp_node(bsgrp_node->scope->bsrp_table,
&bsgrp_node->group);
pim_free_bsgrp_data(bsgrp_node);
return;
}
if ((had_rp_node) && (rp_info->rp_src != RP_SRC_STATIC)) {
/* This means we searched and got rp node, needs unlock */
route_unlock_node(rn);
if (active && pend) {
if ((active->rp_address.s_addr
!= pend->rp_address.s_addr))
pim_rp_change(pim, pend->rp_address,
bsgrp_node->group, RP_SRC_BSR);
}
/* Possible when the first BSM has group with 0 rp count */
if ((!active) && (!pend)) {
if (PIM_DEBUG_BSM) {
zlog_debug(
"%s: Both bsrp and partial list are empty",
__func__);
}
pim_free_bsgrp_node(bsgrp_node->scope->bsrp_table,
&bsgrp_node->group);
pim_free_bsgrp_data(bsgrp_node);
return;
}
/* Possible when a group with 0 rp count received in BSM */
if ((active) && (!pend)) {
pim_rp_del(pim, active->rp_address, bsgrp_node->group,
NULL, RP_SRC_BSR);
pim_free_bsgrp_node(bsgrp_node->scope->bsrp_table,
&bsgrp_node->group);
if (PIM_DEBUG_BSM) {
zlog_debug("%s:Pend List is null,del grp node",
__func__);
}
pim_free_bsgrp_data(bsgrp_node);
return;
}
}
if ((had_rp_node) && (rp_info->rp_src == RP_SRC_STATIC)) {
/* We need to unlock rn this case */
route_unlock_node(rn);
/* there is a chance that static rp exist and bsrp cleaned
* so clean bsgrp node if pending list empty
*/
if (!pend) {
if (PIM_DEBUG_BSM)
zlog_debug(
"%s: Partial list is empty, static rp exists",
__func__);
pim_free_bsgrp_node(bsgrp_node->scope->bsrp_table,
&bsgrp_node->group);
pim_free_bsgrp_data(bsgrp_node);
return;
}
}
/* swap the list & delete all nodes in partial list (old bsrp_list)
* before swap
* active is head of bsrp list
* pend is head of partial list
* After swap
* active is head of partial list
* pend is head of bsrp list
* So check appriate head after swap and clean the new partial list
*/
bsm_rpinfos_swap_all(bsgrp_node->bsrp_list,
bsgrp_node->partial_bsrp_list);
if (active)
pim_g2rp_timer_stop(active);
pim_bsm_rpinfos_free(bsgrp_node->partial_bsrp_list);
}
static bool is_preferred_bsr(struct pim_instance *pim, struct in_addr bsr,
uint32_t bsr_prio)
{
if (bsr.s_addr == pim->global_scope.current_bsr.s_addr)
return true;
if (bsr_prio > pim->global_scope.current_bsr_prio)
return true;
else if (bsr_prio == pim->global_scope.current_bsr_prio) {
if (ntohl(bsr.s_addr)
>= ntohl(pim->global_scope.current_bsr.s_addr))
return true;
else
return false;
} else
return false;
}
static void pim_bsm_update(struct pim_instance *pim, struct in_addr bsr,
uint32_t bsr_prio)
{
if (bsr.s_addr != pim->global_scope.current_bsr.s_addr) {
pim_nht_bsr_del(pim, pim->global_scope.current_bsr);
pim_nht_bsr_add(pim, bsr);
pim->global_scope.current_bsr = bsr;
pim->global_scope.current_bsr_first_ts =
pim_time_monotonic_sec();
pim->global_scope.state = ACCEPT_PREFERRED;
}
pim->global_scope.current_bsr_prio = bsr_prio;
pim->global_scope.current_bsr_last_ts = pim_time_monotonic_sec();
}
void pim_bsm_clear(struct pim_instance *pim)
{
struct route_node *rn;
struct route_node *rpnode;
struct bsgrp_node *bsgrp;
struct prefix nht_p;
struct prefix g_all;
struct rp_info *rp_all;
struct pim_upstream *up;
struct rp_info *rp_info;
bool upstream_updated = false;
pim_nht_bsr_del(pim, pim->global_scope.current_bsr);
/* Reset scope zone data */
pim->global_scope.accept_nofwd_bsm = false;
pim->global_scope.state = ACCEPT_ANY;
pim->global_scope.current_bsr.s_addr = INADDR_ANY;
pim->global_scope.current_bsr_prio = 0;
pim->global_scope.current_bsr_first_ts = 0;
pim->global_scope.current_bsr_last_ts = 0;
pim->global_scope.bsm_frag_tag = 0;
pim_bsm_frags_free(&pim->global_scope);
pim_bs_timer_stop(&pim->global_scope);
for (rn = route_top(pim->global_scope.bsrp_table); rn;
rn = route_next(rn)) {
bsgrp = rn->info;
if (!bsgrp)
continue;
rpnode = route_node_lookup(pim->rp_table, &bsgrp->group);
if (!rpnode) {
pim_free_bsgrp_node(bsgrp->scope->bsrp_table,
&bsgrp->group);
pim_free_bsgrp_data(bsgrp);
continue;
}
rp_info = (struct rp_info *)rpnode->info;
if ((!rp_info) || (rp_info->rp_src != RP_SRC_BSR)) {
pim_free_bsgrp_node(bsgrp->scope->bsrp_table,
&bsgrp->group);
pim_free_bsgrp_data(bsgrp);
continue;
}
/* Deregister addr with Zebra NHT */
nht_p.family = AF_INET;
nht_p.prefixlen = IPV4_MAX_BITLEN;
nht_p.u.prefix4 = rp_info->rp.rpf_addr.u.prefix4;
if (PIM_DEBUG_PIM_NHT_RP) {
zlog_debug("%s: Deregister RP addr %pFX with Zebra ",
__func__, &nht_p);
}
pim_delete_tracked_nexthop(pim, &nht_p, NULL, rp_info);
if (!str2prefix("224.0.0.0/4", &g_all))
return;
rp_all = pim_rp_find_match_group(pim, &g_all);
if (rp_all == rp_info) {
pim_addr_to_prefix(&rp_all->rp.rpf_addr, PIMADDR_ANY);
rp_all->i_am_rp = 0;
} else {
/* Delete the rp_info from rp-list */
listnode_delete(pim->rp_list, rp_info);
/* Delete the rp node from rp_table */
rpnode->info = NULL;
route_unlock_node(rpnode);
route_unlock_node(rpnode);
XFREE(MTYPE_PIM_RP, rp_info);
}
pim_free_bsgrp_node(bsgrp->scope->bsrp_table, &bsgrp->group);
pim_free_bsgrp_data(bsgrp);
}
pim_rp_refresh_group_to_rp_mapping(pim);
frr_each (rb_pim_upstream, &pim->upstream_head, up) {
/* Find the upstream (*, G) whose upstream address is same as
* the RP
*/
if (!pim_addr_is_any(up->sg.src))
continue;
struct prefix grp;
struct rp_info *trp_info;
pim_addr_to_prefix(&grp, up->sg.grp);
trp_info = pim_rp_find_match_group(pim, &grp);
/* RP not found for the group grp */
if (pim_rpf_addr_is_inaddr_any(&trp_info->rp)) {
pim_upstream_rpf_clear(pim, up);
pim_rp_set_upstream_addr(pim, &up->upstream_addr,
up->sg.src, up->sg.grp);
} else {
/* RP found for the group grp */
pim_upstream_update(pim, up);
upstream_updated = true;
}
}
if (upstream_updated)
pim_zebra_update_all_interfaces(pim);
}
static bool pim_bsm_send_intf(uint8_t *buf, int len, struct interface *ifp,
pim_addr dst_addr)
{
struct pim_interface *pim_ifp;
pim_ifp = ifp->info;
if (!pim_ifp) {
if (PIM_DEBUG_BSM)
zlog_debug("%s: Pim interface not available for %s",
__func__, ifp->name);
return false;
}
if (pim_ifp->pim_sock_fd == -1) {
if (PIM_DEBUG_BSM)
zlog_debug("%s: Pim sock not available for %s",
__func__, ifp->name);
return false;
}
if (pim_msg_send(pim_ifp->pim_sock_fd, pim_ifp->primary_address,
dst_addr, buf, len, ifp->name)) {
zlog_warn("%s: Could not send BSM message on interface: %s",
__func__, ifp->name);
return false;
}
pim_ifp->pim_ifstat_bsm_tx++;
pim_ifp->pim->bsm_sent++;
return true;
}
static bool pim_bsm_frag_send(uint8_t *buf, uint32_t len, struct interface *ifp,
uint32_t pim_mtu, pim_addr dst_addr, bool no_fwd)
{
struct bsmmsg_grpinfo *grpinfo, *curgrp;
uint8_t *firstgrp_ptr;
uint8_t *pkt;
uint8_t *pak_start;
uint32_t parsed_len = 0;
uint32_t this_pkt_rem;
uint32_t copy_byte_count;
uint32_t this_pkt_len;
uint8_t total_rp_cnt;
uint8_t this_rp_cnt;
uint8_t frag_rp_cnt;
uint8_t rp_fit_cnt;
bool pak_pending = false;
/* MTU passed here is PIM MTU (IP MTU less IP Hdr) */
if (pim_mtu < (PIM_MIN_BSM_LEN)) {
zlog_warn(
"%s: mtu(pim mtu: %d) size less than minimum bootstrap len",
__func__, pim_mtu);
if (PIM_DEBUG_BSM)
zlog_debug(
"%s: mtu (pim mtu:%d) less than minimum bootstrap len",
__func__, pim_mtu);
return false;
}
pak_start = XCALLOC(MTYPE_PIM_BSM_PKT_VAR_MEM, pim_mtu);
pkt = pak_start;
/* Fill PIM header later before sending packet to calc checksum */
pkt += PIM_MSG_HEADER_LEN;
buf += PIM_MSG_HEADER_LEN;
/* copy bsm header to new packet at offset of pim hdr */
memcpy(pkt, buf, PIM_BSM_HDR_LEN);
pkt += PIM_BSM_HDR_LEN;
buf += PIM_BSM_HDR_LEN;
parsed_len += (PIM_MSG_HEADER_LEN + PIM_BSM_HDR_LEN);
/* Store the position of first grp ptr, which can be reused for
* next packet to start filling group. old bsm header and pim hdr
* remains. So need not be filled again for next packet onwards.
*/
firstgrp_ptr = pkt;
/* we received mtu excluding IP hdr len as param
* now this_pkt_rem is mtu excluding
* PIM_BSM_HDR_LEN + PIM_MSG_HEADER_LEN
*/
this_pkt_rem = pim_mtu - (PIM_BSM_HDR_LEN + PIM_MSG_HEADER_LEN);
/* For each group till the packet length parsed */
while (parsed_len < len) {
/* pkt ---> fragment's current pointer
* buf ---> input buffer's current pointer
* mtu ---> size of the pim packet - PIM header
* curgrp ---> current group on the fragment
* grpinfo ---> current group on the input buffer
* this_pkt_rem ---> bytes remaing on the current fragment
* rp_fit_cnt ---> num of rp for current grp that
* fits this frag
* total_rp_cnt ---> total rp present for the group in the buf
* frag_rp_cnt ---> no of rp for the group to be fit in
* the frag
* this_rp_cnt ---> how many rp have we parsed
*/
grpinfo = (struct bsmmsg_grpinfo *)buf;
memcpy(pkt, buf, PIM_BSM_GRP_LEN);
curgrp = (struct bsmmsg_grpinfo *)pkt;
parsed_len += PIM_BSM_GRP_LEN;
pkt += PIM_BSM_GRP_LEN;
buf += PIM_BSM_GRP_LEN;
this_pkt_rem -= PIM_BSM_GRP_LEN;
/* initialize rp count and total_rp_cnt before the rp loop */
this_rp_cnt = 0;
total_rp_cnt = grpinfo->frag_rp_count;
/* Loop till all RPs for the group parsed */
while (this_rp_cnt < total_rp_cnt) {
/* All RP from a group processed here.
* group is pointed by grpinfo.
* At this point make sure buf pointing to a RP
* within a group
*/
rp_fit_cnt = this_pkt_rem / PIM_BSM_RP_LEN;
/* calculate how many rp am i going to copy in
* this frag
*/
if (rp_fit_cnt > (total_rp_cnt - this_rp_cnt))
frag_rp_cnt = total_rp_cnt - this_rp_cnt;
else
frag_rp_cnt = rp_fit_cnt;
/* populate the frag rp count for the current grp */
curgrp->frag_rp_count = frag_rp_cnt;
copy_byte_count = frag_rp_cnt * PIM_BSM_RP_LEN;
/* copy all the rp that we are fitting in this
* frag for the grp
*/
memcpy(pkt, buf, copy_byte_count);
this_rp_cnt += frag_rp_cnt;
buf += copy_byte_count;
pkt += copy_byte_count;
parsed_len += copy_byte_count;
this_pkt_rem -= copy_byte_count;
/* Either we couldn't fit all rp for the group or the
* mtu reached
*/
if ((this_rp_cnt < total_rp_cnt)
|| (this_pkt_rem
< (PIM_BSM_GRP_LEN + PIM_BSM_RP_LEN))) {
/* No space to fit in more rp, send this pkt */
this_pkt_len = pim_mtu - this_pkt_rem;
pim_msg_build_header(pak_start, this_pkt_len,
PIM_MSG_TYPE_BOOTSTRAP,
no_fwd);
pim_bsm_send_intf(pak_start, this_pkt_len, ifp,
dst_addr);
/* Construct next fragment. Reuse old packet */
pkt = firstgrp_ptr;
this_pkt_rem = pim_mtu - (PIM_BSM_HDR_LEN
+ PIM_MSG_HEADER_LEN);
/* If pkt can't accomodate next group + atleast
* one rp, we must break out of this inner loop
* and process next RP
*/
if (total_rp_cnt == this_rp_cnt)
break;
/* If some more RPs for the same group pending,
* fill grp hdr
*/
memcpy(pkt, (uint8_t *)grpinfo,
PIM_BSM_GRP_LEN);
curgrp = (struct bsmmsg_grpinfo *)pkt;
pkt += PIM_BSM_GRP_LEN;
this_pkt_rem -= PIM_BSM_GRP_LEN;
pak_pending = false;
} else {
/* We filled something but not yet sent out */
pak_pending = true;
}
} /* while RP count */
} /*while parsed len */
/* Send if we have any unsent packet */
if (pak_pending) {
this_pkt_len = pim_mtu - this_pkt_rem;
pim_msg_build_header(pak_start, this_pkt_len,
PIM_MSG_TYPE_BOOTSTRAP, no_fwd);
pim_bsm_send_intf(pak_start, (pim_mtu - this_pkt_rem), ifp,
dst_addr);
}
XFREE(MTYPE_PIM_BSM_PKT_VAR_MEM, pak_start);
return true;
}
static void pim_bsm_fwd_whole_sz(struct pim_instance *pim, uint8_t *buf,
uint32_t len, int sz)
{
struct interface *ifp;
struct pim_interface *pim_ifp;
pim_addr dst_addr;
uint32_t pim_mtu;
bool no_fwd = false;
bool ret = false;
/* For now only global scope zone is supported, so send on all
* pim interfaces in the vrf
*/
dst_addr = qpim_all_pim_routers_addr;
FOR_ALL_INTERFACES (pim->vrf, ifp) {
pim_ifp = ifp->info;
if ((!pim_ifp) || (!pim_ifp->bsm_enable))
continue;
/*
* RFC 5059 Sec 3.4:
* When a Bootstrap message is forwarded, it is forwarded out
* of every multicast-capable interface that has PIM neighbors.
*
* So skipping pim interfaces with no neighbors.
*/
if (listcount(pim_ifp->pim_neighbor_list) == 0)
continue;
pim_hello_require(ifp);
pim_mtu = ifp->mtu - MAX_IP_HDR_LEN;
if (pim_mtu < len) {
ret = pim_bsm_frag_send(buf, len, ifp, pim_mtu,
dst_addr, no_fwd);
if (PIM_DEBUG_BSM)
zlog_debug("%s: pim_bsm_frag_send returned %s",
__func__, ret ? "TRUE" : "FALSE");
} else {
pim_msg_build_header(buf, len, PIM_MSG_TYPE_BOOTSTRAP,
no_fwd);
if (!pim_bsm_send_intf(buf, len, ifp, dst_addr)) {
if (PIM_DEBUG_BSM)
zlog_debug(
"%s: pim_bsm_send_intf returned false",
__func__);
}
}
}
}
bool pim_bsm_new_nbr_fwd(struct pim_neighbor *neigh, struct interface *ifp)
{
pim_addr dst_addr;
struct pim_interface *pim_ifp;
struct bsm_scope *scope;
struct bsm_frag *bsfrag;
char neigh_src_str[INET_ADDRSTRLEN];
uint32_t pim_mtu;
bool no_fwd = true;
bool ret = false;
if (PIM_DEBUG_BSM)
zlog_debug("%s: New neighbor %pPA seen on %s", __func__,
&neigh->source_addr, ifp->name);
pim_ifp = ifp->info;
/* DR only forwards BSM packet */
if (!pim_addr_cmp(pim_ifp->pim_dr_addr, pim_ifp->primary_address)) {
if (PIM_DEBUG_BSM)
zlog_debug(
"%s: It is not DR, so don't forward BSM packet",
__func__);
}
if (!pim_ifp->bsm_enable) {
if (PIM_DEBUG_BSM)
zlog_debug("%s: BSM proc not enabled on %s", __func__,
ifp->name);
return ret;
}
scope = &pim_ifp->pim->global_scope;
if (!bsm_frags_count(scope->bsm_frags)) {
if (PIM_DEBUG_BSM)
zlog_debug("%s: BSM list for the scope is empty",
__func__);
return ret;
}
if (!pim_ifp->ucast_bsm_accept) {
dst_addr = qpim_all_pim_routers_addr;
if (PIM_DEBUG_BSM)
zlog_debug("%s: Sending BSM mcast to %s", __func__,
neigh_src_str);
} else {
dst_addr = neigh->source_addr;
if (PIM_DEBUG_BSM)
zlog_debug("%s: Sending BSM ucast to %s", __func__,
neigh_src_str);
}
pim_mtu = ifp->mtu - MAX_IP_HDR_LEN;
pim_hello_require(ifp);
frr_each (bsm_frags, scope->bsm_frags, bsfrag) {
if (pim_mtu < bsfrag->size) {
ret = pim_bsm_frag_send(bsfrag->data, bsfrag->size, ifp,
pim_mtu, dst_addr, no_fwd);
if (!ret) {
if (PIM_DEBUG_BSM)
zlog_debug(
"%s: pim_bsm_frag_send failed",
__func__);
}
} else {
/* Pim header needs to be constructed */
pim_msg_build_header(bsfrag->data, bsfrag->size,
PIM_MSG_TYPE_BOOTSTRAP, no_fwd);
ret = pim_bsm_send_intf(bsfrag->data, bsfrag->size, ifp,
dst_addr);
if (!ret) {
if (PIM_DEBUG_BSM)
zlog_debug(
"%s: pim_bsm_frag_send failed",
__func__);
}
}
}
return ret;
}
struct bsgrp_node *pim_bsm_get_bsgrp_node(struct bsm_scope *scope,
struct prefix *grp)
{
struct route_node *rn;
struct bsgrp_node *bsgrp;
rn = route_node_lookup(scope->bsrp_table, grp);
if (!rn) {
if (PIM_DEBUG_BSM)
zlog_debug("%s: Route node doesn't exist for the group",
__func__);
return NULL;
}
bsgrp = rn->info;
route_unlock_node(rn);
return bsgrp;
}
static uint32_t hash_calc_on_grp_rp(struct prefix group, struct in_addr rp,
uint8_t hashmasklen)
{
uint64_t temp;
uint32_t hash;
uint32_t grpaddr;
uint32_t rp_add;
uint32_t mask = 0xffffffff;
/* mask to be made zero if hashmasklen is 0 because mask << 32
* may not give 0. hashmasklen can be 0 to 32.
*/
if (hashmasklen == 0)
mask = 0;
/* in_addr stores ip in big endian, hence network byte order
* convert to uint32 before processing hash
*/
grpaddr = ntohl(group.u.prefix4.s_addr);
/* Avoid shifting by 32 bit on a 32 bit register */
if (hashmasklen)
grpaddr = grpaddr & ((mask << (32 - hashmasklen)));
else
grpaddr = grpaddr & mask;
rp_add = ntohl(rp.s_addr);
temp = 1103515245 * ((1103515245 * (uint64_t)grpaddr + 12345) ^ rp_add)
+ 12345;
hash = temp & (0x7fffffff);
return hash;
}
static bool pim_install_bsm_grp_rp(struct pim_instance *pim,
struct bsgrp_node *grpnode,
struct bsmmsg_rpinfo *rp)
{
struct bsm_rpinfo *bsm_rpinfo;
uint8_t hashMask_len = pim->global_scope.hashMasklen;
/*memory allocation for bsm_rpinfo */
bsm_rpinfo = XCALLOC(MTYPE_PIM_BSRP_INFO, sizeof(*bsm_rpinfo));
bsm_rpinfo->rp_prio = rp->rp_pri;
bsm_rpinfo->rp_holdtime = rp->rp_holdtime;
memcpy(&bsm_rpinfo->rp_address, &rp->rpaddr.addr,
sizeof(struct in_addr));
bsm_rpinfo->elapse_time = 0;
/* Back pointer to the group node. */
bsm_rpinfo->bsgrp_node = grpnode;
/* update hash for this rp node */
bsm_rpinfo->hash = hash_calc_on_grp_rp(grpnode->group, rp->rpaddr.addr,
hashMask_len);
if (bsm_rpinfos_add(grpnode->partial_bsrp_list, bsm_rpinfo) == NULL) {
if (PIM_DEBUG_BSM)
zlog_debug(
"%s, bs_rpinfo node added to the partial bs_rplist.",
__func__);
return true;
}
if (PIM_DEBUG_BSM)
zlog_debug("%s: list node not added", __func__);
XFREE(MTYPE_PIM_BSRP_INFO, bsm_rpinfo);
return false;
}
static void pim_update_pending_rp_cnt(struct bsm_scope *sz,
struct bsgrp_node *bsgrp,
uint16_t bsm_frag_tag,
uint32_t total_rp_count)
{
if (bsgrp->pend_rp_cnt) {
/* received bsm is different packet ,
* it is not same fragment.
*/
if (bsm_frag_tag != bsgrp->frag_tag) {
if (PIM_DEBUG_BSM)
zlog_debug(
"%s,Received a new BSM ,so clear the pending bs_rpinfo list.",
__func__);
pim_bsm_rpinfos_free(bsgrp->partial_bsrp_list);
bsgrp->pend_rp_cnt = total_rp_count;
}
} else
bsgrp->pend_rp_cnt = total_rp_count;
bsgrp->frag_tag = bsm_frag_tag;
}
/* Parsing BSR packet and adding to partial list of corresponding bsgrp node */
static bool pim_bsm_parse_install_g2rp(struct bsm_scope *scope, uint8_t *buf,
int buflen, uint16_t bsm_frag_tag)
{
struct bsmmsg_grpinfo grpinfo;
struct bsmmsg_rpinfo rpinfo;
struct prefix group;
struct bsgrp_node *bsgrp = NULL;
int frag_rp_cnt = 0;
int offset = 0;
int ins_count = 0;
while (buflen > offset) {
if (offset + (int)sizeof(struct bsmmsg_grpinfo) > buflen) {
if (PIM_DEBUG_BSM)
zlog_debug(
"%s: buflen received %d is less than the internal data structure of the packet would suggest",
__func__, buflen);
return false;
}
/* Extract Group tlv from BSM */
memcpy(&grpinfo, buf, sizeof(struct bsmmsg_grpinfo));
if (PIM_DEBUG_BSM) {
char grp_str[INET_ADDRSTRLEN];
pim_inet4_dump("<Group?>", grpinfo.group.addr, grp_str,
sizeof(grp_str));
zlog_debug(
"%s, Group %s Rpcount:%d Fragment-Rp-count:%d",
__func__, grp_str, grpinfo.rp_count,
grpinfo.frag_rp_count);
}
buf += sizeof(struct bsmmsg_grpinfo);
offset += sizeof(struct bsmmsg_grpinfo);
group.family = AF_INET;
if (grpinfo.group.mask > IPV4_MAX_BITLEN) {
if (PIM_DEBUG_BSM)
zlog_debug(
"%s, v4 prefix length specified: %d is too long",
__func__, grpinfo.group.mask);
return false;
}
group.prefixlen = grpinfo.group.mask;
group.u.prefix4.s_addr = grpinfo.group.addr.s_addr;
/* Get the Group node for the BSM rp table */
bsgrp = pim_bsm_get_bsgrp_node(scope, &group);
if (grpinfo.rp_count == 0) {
struct bsm_rpinfo *old_rpinfo;
/* BSR explicitly no longer has RPs for this group */
if (!bsgrp)
continue;
if (PIM_DEBUG_BSM) {
char grp_str[INET_ADDRSTRLEN];
pim_inet4_dump("<Group?>", grpinfo.group.addr,
grp_str, sizeof(grp_str));
zlog_debug("%s, Rp count is zero for group: %s",
__func__, grp_str);
}
old_rpinfo = bsm_rpinfos_first(bsgrp->bsrp_list);
if (old_rpinfo)
pim_rp_del(scope->pim, old_rpinfo->rp_address,
group, NULL, RP_SRC_BSR);
pim_free_bsgrp_node(scope->bsrp_table, &bsgrp->group);
pim_free_bsgrp_data(bsgrp);
continue;
}
if (!bsgrp) {
if (PIM_DEBUG_BSM)
zlog_debug("%s, Create new BSM Group node.",
__func__);
/* create a new node to be added to the tree. */
bsgrp = pim_bsm_new_bsgrp_node(scope->bsrp_table,
&group);
if (!bsgrp) {
zlog_debug(
"%s, Failed to get the BSM group node.",
__func__);
continue;
}
bsgrp->scope = scope;
}
pim_update_pending_rp_cnt(scope, bsgrp, bsm_frag_tag,
grpinfo.rp_count);
frag_rp_cnt = grpinfo.frag_rp_count;
ins_count = 0;
while (frag_rp_cnt--) {
if (offset + (int)sizeof(struct bsmmsg_rpinfo)
> buflen) {
if (PIM_DEBUG_BSM)
zlog_debug(
"%s, buflen received: %u is less than the internal data structure of the packet would suggest",
__func__, buflen);
return false;
}
/* Extract RP address tlv from BSM */
memcpy(&rpinfo, buf, sizeof(struct bsmmsg_rpinfo));
rpinfo.rp_holdtime = ntohs(rpinfo.rp_holdtime);
buf += sizeof(struct bsmmsg_rpinfo);
offset += sizeof(struct bsmmsg_rpinfo);
if (PIM_DEBUG_BSM) {
char rp_str[INET_ADDRSTRLEN];
pim_inet4_dump("<Rpaddr?>", rpinfo.rpaddr.addr,
rp_str, sizeof(rp_str));
zlog_debug(
"%s, Rp address - %s; pri:%d hold:%d",
__func__, rp_str, rpinfo.rp_pri,
rpinfo.rp_holdtime);
}
/* Call Install api to update grp-rp mappings */
if (pim_install_bsm_grp_rp(scope->pim, bsgrp, &rpinfo))
ins_count++;
}
bsgrp->pend_rp_cnt -= ins_count;
if (!bsgrp->pend_rp_cnt) {
if (PIM_DEBUG_BSM)
zlog_debug(
"%s, Recvd all the rps for this group, so bsrp list with penidng rp list.",
__func__);
/* replace the bsrp_list with pending list */
pim_instate_pend_list(bsgrp);
}
}
return true;
}
int pim_bsm_process(struct interface *ifp, struct ip *ip_hdr, uint8_t *buf,
uint32_t buf_size, bool no_fwd)
{
struct bsm_hdr *bshdr;
int sz = PIM_GBL_SZ_ID;
struct bsmmsg_grpinfo *msg_grp;
struct pim_interface *pim_ifp = NULL;
struct bsm_frag *bsfrag;
struct pim_instance *pim;
char bsr_str[INET_ADDRSTRLEN];
uint16_t frag_tag;
bool empty_bsm = false;
/* BSM Packet acceptance validation */
pim_ifp = ifp->info;
if (!pim_ifp) {
if (PIM_DEBUG_BSM)
zlog_debug("%s: multicast not enabled on interface %s",
__func__, ifp->name);
return -1;
}
pim_ifp->pim_ifstat_bsm_rx++;
pim = pim_ifp->pim;
pim->bsm_rcvd++;
/* Drop if bsm processing is disabled on interface */
if (!pim_ifp->bsm_enable) {
zlog_warn("%s: BSM not enabled on interface %s", __func__,
ifp->name);
pim_ifp->pim_ifstat_bsm_cfg_miss++;
pim->bsm_dropped++;
return -1;
}
if (buf_size < (PIM_MSG_HEADER_LEN + sizeof(struct bsm_hdr))) {
if (PIM_DEBUG_BSM)
zlog_debug(
"%s: received buffer length of %d which is too small to properly decode",
__func__, buf_size);
return -1;
}
bshdr = (struct bsm_hdr *)(buf + PIM_MSG_HEADER_LEN);
pim_inet4_dump("<bsr?>", bshdr->bsr_addr.addr, bsr_str,
sizeof(bsr_str));
if (bshdr->hm_len > IPV4_MAX_BITLEN) {
zlog_warn("Bad hashmask length for IPv4; got %hhu, expected value in range 0-32",
bshdr->hm_len);
pim->bsm_dropped++;
return -1;
}
pim->global_scope.hashMasklen = bshdr->hm_len;
frag_tag = ntohs(bshdr->frag_tag);
/* Identify empty BSM */
if ((buf_size - PIM_BSM_HDR_LEN - PIM_MSG_HEADER_LEN) < PIM_BSM_GRP_LEN)
empty_bsm = true;
if (!empty_bsm) {
msg_grp = (struct bsmmsg_grpinfo *)(buf + PIM_MSG_HEADER_LEN
+ PIM_BSM_HDR_LEN);
/* Currently we don't support scope zoned BSM */
if (msg_grp->group.sz) {
if (PIM_DEBUG_BSM)
zlog_debug(
"%s : Administratively scoped range BSM received",
__func__);
pim_ifp->pim_ifstat_bsm_invalid_sz++;
pim->bsm_dropped++;
return -1;
}
}
/* Drop if bsr is not preferred bsr */
if (!is_preferred_bsr(pim, bshdr->bsr_addr.addr, bshdr->bsr_prio)) {
if (PIM_DEBUG_BSM)
zlog_debug("%s : Received a non-preferred BSM",
__func__);
pim->bsm_dropped++;
return -1;
}
if (no_fwd) {
/* only accept no-forward BSM if quick refresh on startup */
if ((pim->global_scope.accept_nofwd_bsm)
|| (frag_tag == pim->global_scope.bsm_frag_tag)) {
pim->global_scope.accept_nofwd_bsm = false;
} else {
if (PIM_DEBUG_BSM)
zlog_debug(
"%s : nofwd_bsm received on %s when accpt_nofwd_bsm false",
__func__, bsr_str);
pim->bsm_dropped++;
pim_ifp->pim_ifstat_ucast_bsm_cfg_miss++;
return -1;
}
}
#if PIM_IPV == 4
if (ip_hdr->ip_dst.s_addr == qpim_all_pim_routers_addr.s_addr)
#else
if (0)
#endif
{
/* Multicast BSMs are only accepted if source interface & IP
* match RPF towards the BSR's IP address, or they have
* no-forward set
*/
if (!no_fwd
&& !pim_nht_bsr_rpf_check(pim, bshdr->bsr_addr.addr, ifp,
ip_hdr->ip_src)) {
if (PIM_DEBUG_BSM)
zlog_debug(
"BSM check: RPF to BSR %s is not %pI4%%%s",
bsr_str, &ip_hdr->ip_src, ifp->name);
pim->bsm_dropped++;
return -1;
}
} else if (if_address_is_local(&ip_hdr->ip_dst, AF_INET,
pim->vrf->vrf_id)) {
/* Unicast BSM received - if ucast bsm not enabled on
* the interface, drop it
*/
if (!pim_ifp->ucast_bsm_accept) {
if (PIM_DEBUG_BSM)
zlog_debug(
"%s : Unicast BSM not enabled on interface %s",
__func__, ifp->name);
pim_ifp->pim_ifstat_ucast_bsm_cfg_miss++;
pim->bsm_dropped++;
return -1;
}
} else {
if (PIM_DEBUG_BSM)
zlog_debug("%s : Invalid destination address",
__func__);
pim->bsm_dropped++;
return -1;
}
if (empty_bsm) {
if (PIM_DEBUG_BSM)
zlog_debug("%s : Empty Pref BSM received", __func__);
}
/* Parse Update bsm rp table and install/uninstall rp if required */
if (!pim_bsm_parse_install_g2rp(
&pim_ifp->pim->global_scope,
(buf + PIM_BSM_HDR_LEN + PIM_MSG_HEADER_LEN),
(buf_size - PIM_BSM_HDR_LEN - PIM_MSG_HEADER_LEN),
frag_tag)) {
if (PIM_DEBUG_BSM) {
zlog_debug("%s, Parsing BSM failed.", __func__);
}
pim->bsm_dropped++;
return -1;
}
/* Restart the bootstrap timer */
pim_bs_timer_restart(&pim_ifp->pim->global_scope,
PIM_BSR_DEFAULT_TIMEOUT);
/* If new BSM received, clear the old bsm database */
if (pim_ifp->pim->global_scope.bsm_frag_tag != frag_tag) {
if (PIM_DEBUG_BSM) {
zlog_debug("%s: Current frag tag: %d Frag teg rcvd: %d",
__func__,
pim_ifp->pim->global_scope.bsm_frag_tag,
frag_tag);
}
pim_bsm_frags_free(&pim_ifp->pim->global_scope);
pim_ifp->pim->global_scope.bsm_frag_tag = frag_tag;
}
/* update the scope information from bsm */
pim_bsm_update(pim, bshdr->bsr_addr.addr, bshdr->bsr_prio);
if (!no_fwd) {
pim_bsm_fwd_whole_sz(pim_ifp->pim, buf, buf_size, sz);
bsfrag = XCALLOC(MTYPE_PIM_BSM_FRAG,
sizeof(struct bsm_frag) + buf_size);
bsfrag->size = buf_size;
memcpy(bsfrag->data, buf, buf_size);
bsm_frags_add_tail(pim_ifp->pim->global_scope.bsm_frags,
bsfrag);
}
return 0;
}
|