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
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
|
// 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));
}
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)) {
if (rta->rta_type <= max)
tb[rta->rta_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_addraw_l(struct nlmsghdr *n, unsigned int maxlen, const void *data,
unsigned int len)
{
if (NLMSG_ALIGN(n->nlmsg_len) + NLMSG_ALIGN(len) > maxlen) {
zlog_err("ERROR message exceeded bound of %d", maxlen);
return false;
}
memcpy(NLMSG_TAIL(n), data, len);
memset((uint8_t *)NLMSG_TAIL(n) + len, 0, NLMSG_ALIGN(len) - len);
n->nlmsg_len = NLMSG_ALIGN(n->nlmsg_len) + NLMSG_ALIGN(len);
return true;
}
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,
nl_msg_type_to_str(err->msg.nlmsg_type),
err->msg.nlmsg_type, err->msg.nlmsg_seq,
err->msg.nlmsg_pid);
}
return 1;
}
/*
* Deal with errors that occur because of races in link handling
* or types are not supported in kernel.
*/
if (is_cmd &&
((msg_type == RTM_DELROUTE &&
(-errnum == ENODEV || -errnum == ESRCH)) ||
(msg_type == RTM_NEWROUTE &&
(-errnum == ENETDOWN || -errnum == EEXIST)) ||
((msg_type == RTM_NEWTUNNEL || msg_type == RTM_DELTUNNEL ||
msg_type == RTM_GETTUNNEL) &&
(-errnum == EOPNOTSUPP)))) {
if (IS_ZEBRA_DEBUG_KERNEL)
zlog_debug("%s: error: %s type=%s(%u), seq=%u, pid=%u",
nl->name, safe_strerror(-errnum),
nl_msg_type_to_str(msg_type), msg_type,
err->msg.nlmsg_seq, err->msg.nlmsg_pid);
return 0;
}
/*
* We see RTM_DELNEIGH when shutting down an interface with an IPv4
* link-local. The kernel should have already deleted the neighbor so
* do not log these as an error.
*/
if (msg_type == RTM_DELNEIGH
|| (is_cmd && msg_type == RTM_NEWROUTE
&& (-errnum == ESRCH || -errnum == ENETUNREACH))) {
/*
* This is known to happen in some situations, don't log as
* error.
*/
if (IS_ZEBRA_DEBUG_KERNEL)
zlog_debug("%s error: %s, type=%s(%u), seq=%u, pid=%u",
nl->name, safe_strerror(-errnum),
nl_msg_type_to_str(msg_type), msg_type,
err->msg.nlmsg_seq, err->msg.nlmsg_pid);
} else {
if ((msg_type != RTM_GETNEXTHOP && msg_type != RTM_GETVLAN) ||
!startup)
flog_err(EC_ZEBRA_UNEXPECTED_MESSAGE,
"%s error: %s, type=%s(%u), seq=%u, pid=%u",
nl->name, safe_strerror(-errnum),
nl_msg_type_to_str(msg_type), msg_type,
err->msg.nlmsg_seq, err->msg.nlmsg_pid);
}
return -1;
}
/*
* netlink_parse_info
*
* Receive message from netlink interface and pass those information
* to the given function.
*
* filter -> Function to call to read the results
* nl -> netlink socket information
* zns -> The zebra namespace data
* count -> How many we should read in, 0 means as much as possible
* startup -> Are we reading in under startup conditions? passed to
* the filter.
*/
int netlink_parse_info(int (*filter)(struct nlmsghdr *, ns_id_t, int),
struct nlsock *nl, const struct zebra_dplane_info *zns,
int count, bool startup)
{
int status;
int ret = 0;
int error;
int read_in = 0;
while (1) {
struct sockaddr_nl snl;
struct msghdr msg = {.msg_name = (void *)&snl,
.msg_namelen = sizeof(snl)};
struct nlmsghdr *h;
if (count && read_in >= count)
return 0;
status = netlink_recv_msg(nl, &msg);
if (status == -1)
return -1;
else if (status == 0)
break;
read_in++;
for (h = (struct nlmsghdr *)nl->buf;
(status >= 0 && NLMSG_OK(h, (unsigned int)status));
h = NLMSG_NEXT(h, status)) {
/* Finish of reading. */
if (h->nlmsg_type == NLMSG_DONE)
return ret;
/* Error handling. */
if (h->nlmsg_type == NLMSG_ERROR) {
int err = netlink_parse_error(
nl, h, zns->is_cmd, startup);
if (err == 1) {
if (!(h->nlmsg_flags & NLM_F_MULTI))
return 0;
continue;
} else
return err;
}
/*
* What is the right thing to do? The kernel
* is telling us that the dump request was interrupted
* and we more than likely are out of luck and have
* missed data from the kernel. At this point in time
* lets just note that this is happening.
*/
if (h->nlmsg_flags & NLM_F_DUMP_INTR)
flog_err(
EC_ZEBRA_NETLINK_BAD_SEQUENCE,
"netlink recvmsg: The Dump request was interrupted");
/* OK we got netlink message. */
if (IS_ZEBRA_DEBUG_KERNEL)
zlog_debug(
"%s: %s type %s(%u), len=%d, seq=%u, pid=%u",
__func__, nl->name,
nl_msg_type_to_str(h->nlmsg_type),
h->nlmsg_type, h->nlmsg_len,
h->nlmsg_seq, h->nlmsg_pid);
/*
* Ignore messages that maybe sent from
* other actors besides the kernel
*/
if (snl.nl_pid != 0) {
zlog_debug("Ignoring message from pid %u",
snl.nl_pid);
continue;
}
error = (*filter)(h, zns->ns_id, startup);
if (error < 0) {
zlog_debug("%s filter function error",
nl->name);
ret = error;
}
}
/* After error care. */
if (msg.msg_flags & MSG_TRUNC) {
flog_err(EC_ZEBRA_NETLINK_LENGTH_ERROR,
"%s error: message truncated", nl->name);
continue;
}
if (status) {
flog_err(EC_ZEBRA_NETLINK_LENGTH_ERROR,
"%s error: data remnant size %d", nl->name,
status);
return -1;
}
}
return ret;
}
/*
* netlink_talk_info
*
* sendmsg() to netlink socket then recvmsg().
* Calls netlink_parse_info to parse returned data
*
* filter -> The filter to read final results from kernel
* nlmsghdr -> The data to send to the kernel
* dp_info -> The dataplane and netlink socket information
* startup -> Are we reading in under startup conditions
* This is passed through eventually to filter.
*/
static int netlink_talk_info(int (*filter)(struct nlmsghdr *, ns_id_t,
int startup),
struct nlmsghdr *n,
struct zebra_dplane_info *dp_info, bool startup)
{
struct nlsock *nl;
nl = kernel_netlink_nlsock_lookup(dp_info->sock);
n->nlmsg_seq = dp_info->seq;
n->nlmsg_pid = nl->snl.nl_pid;
if (IS_ZEBRA_DEBUG_KERNEL)
zlog_debug(
"netlink_talk: %s type %s(%u), len=%d seq=%u flags 0x%x",
nl->name, nl_msg_type_to_str(n->nlmsg_type),
n->nlmsg_type, n->nlmsg_len, n->nlmsg_seq,
n->nlmsg_flags);
if (netlink_send_msg(nl, n, n->nlmsg_len) == -1)
return -1;
/*
* Get reply from netlink socket.
* The reply should either be an acknowlegement or an error.
*/
return netlink_parse_info(filter, nl, dp_info, 0, startup);
}
/*
* Synchronous version of netlink_talk_info. Converts args to suit the
* common version, which is suitable for both sync and async use.
*/
int netlink_talk(int (*filter)(struct nlmsghdr *, ns_id_t, int startup),
struct nlmsghdr *n, struct nlsock *nl, struct zebra_ns *zns,
bool startup)
{
struct zebra_dplane_info dp_info;
/* Increment sequence number before capturing snapshot of ns socket
* info.
*/
nl->seq++;
/* Capture info in intermediate info struct */
zebra_dplane_info_from_zns(&dp_info, zns, (nl == &(zns->netlink_cmd)));
return netlink_talk_info(filter, n, &dp_info, startup);
}
/*
* Synchronous version of netlink_talk_info. Converts args to suit the
* common version, which is suitable for both sync and async use.
*/
int ge_netlink_talk(int (*filter)(struct nlmsghdr *, ns_id_t, int startup),
struct nlmsghdr *n, struct zebra_ns *zns, bool startup)
{
struct zebra_dplane_info dp_info;
if (zns->ge_netlink_cmd.sock < 0)
return -1;
/* Increment sequence number before capturing snapshot of ns socket
* info.
*/
zns->ge_netlink_cmd.seq = zebra_router_get_next_sequence();
/* Capture info in intermediate info struct */
dp_info.ns_id = zns->ns_id;
dp_info.is_cmd = true;
dp_info.sock = zns->ge_netlink_cmd.sock;
dp_info.seq = zns->ge_netlink_cmd.seq;
return netlink_talk_info(filter, n, &dp_info, startup);
}
/* Issue request message to kernel via netlink socket. GET messages
* are issued through this interface.
*/
int netlink_request(struct nlsock *nl, void *req)
{
struct nlmsghdr *n = (struct nlmsghdr *)req;
/* Check netlink socket. */
if (nl->sock < 0) {
flog_err_sys(EC_LIB_SOCKET, "%s socket isn't active.",
nl->name);
return -1;
}
/* Fill common fields for all requests. */
n->nlmsg_pid = nl->snl.nl_pid;
n->nlmsg_seq = ++nl->seq;
if (netlink_send_msg(nl, req, n->nlmsg_len) == -1)
return -1;
return 0;
}
static int nl_batch_read_resp(struct nl_batch *bth, struct nlsock *nl)
{
struct nlmsghdr *h;
struct sockaddr_nl snl;
struct msghdr msg = {};
int status, seq;
struct zebra_dplane_ctx *ctx;
bool ignore_msg;
msg.msg_name = (void *)&snl;
msg.msg_namelen = sizeof(snl);
/*
* The responses are not batched, so we need to read and process one
* message at a time.
*/
while (true) {
status = netlink_recv_msg(nl, &msg);
/*
* status == -1 is a full on failure somewhere
* since we don't know where the problem happened
* we must mark all as failed
*
* Else we mark everything as worked
*
*/
if (status == -1 || status == 0) {
while ((ctx = dplane_ctx_dequeue(&(bth->ctx_list))) !=
NULL) {
if (status == -1)
dplane_ctx_set_status(
ctx,
ZEBRA_DPLANE_REQUEST_FAILURE);
dplane_ctx_enqueue_tail(bth->ctx_out_q, ctx);
}
return status;
}
h = (struct nlmsghdr *)nl->buf;
ignore_msg = false;
seq = h->nlmsg_seq;
/*
* Find the corresponding context object. Received responses are
* in the same order as requests we sent, so we can simply
* iterate over the context list and match responses with
* requests at same time.
*/
while (true) {
ctx = dplane_ctx_get_head(&(bth->ctx_list));
if (ctx == NULL) {
/*
* This is a situation where we have gotten
* into a bad spot. We need to know that
* this happens( does it? )
*/
zlog_err(
"%s:WARNING Received netlink Response for an error and no Contexts to associate with it",
__func__);
break;
}
/*
* 'update' context objects take two consecutive
* sequence numbers.
*/
if (dplane_ctx_is_update(ctx) &&
dplane_ctx_get_ns(ctx)->seq + 1 == seq) {
/*
* This is the situation where we get a response
* to a message that should be ignored.
*/
ignore_msg = true;
break;
}
ctx = dplane_ctx_dequeue(&(bth->ctx_list));
dplane_ctx_enqueue_tail(bth->ctx_out_q, ctx);
/* We have found corresponding context object. */
if (dplane_ctx_get_ns(ctx)->seq == seq)
break;
if (dplane_ctx_get_ns(ctx)->seq > seq)
zlog_warn(
"%s:WARNING Received %u is less than any context on the queue ctx->seq %u",
__func__, seq,
dplane_ctx_get_ns(ctx)->seq);
}
if (ignore_msg) {
/*
* If we ignore the message due to an update
* above we should still fricking decode the
* message for our operator to understand
* what is going on
*/
int err = netlink_parse_error(nl, h, bth->zns->is_cmd,
false);
zlog_debug("%s: netlink error message seq=%d %d",
__func__, h->nlmsg_seq, err);
continue;
}
/*
* We received a message with the sequence number that isn't
* associated with any dplane context object.
*/
if (ctx == NULL) {
if (IS_ZEBRA_DEBUG_KERNEL)
zlog_debug(
"%s: skipping unassociated response, seq number %d NS %u",
__func__, h->nlmsg_seq,
bth->zns->ns_id);
continue;
}
if (h->nlmsg_type == NLMSG_ERROR) {
int err = netlink_parse_error(nl, h, bth->zns->is_cmd,
false);
if (err == -1)
dplane_ctx_set_status(
ctx, ZEBRA_DPLANE_REQUEST_FAILURE);
if (IS_ZEBRA_DEBUG_KERNEL)
zlog_debug("%s: netlink error message seq=%d ",
__func__, h->nlmsg_seq);
continue;
}
/*
* If we get here then we did not receive neither the ack nor
* the error and instead received some other message in an
* unexpected way.
*/
if (IS_ZEBRA_DEBUG_KERNEL)
zlog_debug("%s: ignoring message type 0x%04x(%s) NS %u",
__func__, h->nlmsg_type,
nl_msg_type_to_str(h->nlmsg_type),
bth->zns->ns_id);
}
return 0;
}
static void nl_batch_reset(struct nl_batch *bth)
{
bth->buf_head = bth->buf;
bth->curlen = 0;
bth->msgcnt = 0;
bth->zns = NULL;
dplane_ctx_q_init(&(bth->ctx_list));
}
static void nl_batch_init(struct nl_batch *bth,
struct dplane_ctx_list_head *ctx_out_q)
{
/*
* If the size of the buffer has changed, free and then allocate a new
* one.
*/
size_t bufsize =
atomic_load_explicit(&nl_batch_bufsize, memory_order_relaxed);
if (bufsize != nl_batch_tx_bufsize) {
if (nl_batch_tx_buf)
XFREE(MTYPE_NL_BUF, nl_batch_tx_buf);
nl_batch_tx_buf = XCALLOC(MTYPE_NL_BUF, bufsize);
nl_batch_tx_bufsize = bufsize;
}
bth->buf = nl_batch_tx_buf;
bth->bufsiz = bufsize;
bth->limit = atomic_load_explicit(&nl_batch_send_threshold,
memory_order_relaxed);
bth->ctx_out_q = ctx_out_q;
nl_batch_reset(bth);
}
static void nl_batch_send(struct nl_batch *bth)
{
struct zebra_dplane_ctx *ctx;
bool err = false;
if (bth->curlen != 0 && bth->zns != NULL) {
struct nlsock *nl =
kernel_netlink_nlsock_lookup(bth->zns->sock);
if (IS_ZEBRA_DEBUG_KERNEL)
zlog_debug("%s: %s, batch size=%zu, msg cnt=%zu",
__func__, nl->name, bth->curlen,
bth->msgcnt);
if (netlink_send_msg(nl, bth->buf, bth->curlen) == -1)
err = true;
if (!err) {
if (nl_batch_read_resp(bth, nl) == -1)
err = true;
}
}
/* Move remaining contexts to the outbound queue. */
while (true) {
ctx = dplane_ctx_dequeue(&(bth->ctx_list));
if (ctx == NULL)
break;
if (err)
dplane_ctx_set_status(ctx,
ZEBRA_DPLANE_REQUEST_FAILURE);
dplane_ctx_enqueue_tail(bth->ctx_out_q, ctx);
}
nl_batch_reset(bth);
}
enum netlink_msg_status netlink_batch_add_msg(
struct nl_batch *bth, struct zebra_dplane_ctx *ctx,
ssize_t (*msg_encoder)(struct zebra_dplane_ctx *, void *, size_t),
bool ignore_res)
{
int seq;
ssize_t size;
struct nlmsghdr *msgh;
struct nlsock *nl;
size = (*msg_encoder)(ctx, bth->buf_head, bth->bufsiz - bth->curlen);
/*
* If there was an error while encoding the message (other than buffer
* overflow) then return an error.
*/
if (size < 0)
return FRR_NETLINK_ERROR;
/*
* If the message doesn't fit entirely in the buffer then send the batch
* and retry.
*/
if (size == 0) {
nl_batch_send(bth);
size = (*msg_encoder)(ctx, bth->buf_head,
bth->bufsiz - bth->curlen);
/*
* If the message doesn't fit in the empty buffer then just
* return an error.
*/
if (size <= 0)
return FRR_NETLINK_ERROR;
}
seq = dplane_ctx_get_ns(ctx)->seq;
nl = kernel_netlink_nlsock_lookup(dplane_ctx_get_ns_sock(ctx));
if (ignore_res)
seq++;
msgh = (struct nlmsghdr *)bth->buf_head;
msgh->nlmsg_seq = seq;
msgh->nlmsg_pid = nl->snl.nl_pid;
bth->zns = dplane_ctx_get_ns(ctx);
bth->buf_head = ((char *)bth->buf_head) + size;
bth->curlen += size;
bth->msgcnt++;
return FRR_NETLINK_QUEUED;
}
static enum netlink_msg_status nl_put_msg(struct nl_batch *bth,
struct zebra_dplane_ctx *ctx)
{
if (dplane_ctx_is_skip_kernel(ctx))
return FRR_NETLINK_SUCCESS;
switch (dplane_ctx_get_op(ctx)) {
case DPLANE_OP_ROUTE_INSTALL:
case DPLANE_OP_ROUTE_UPDATE:
case DPLANE_OP_ROUTE_DELETE:
return netlink_put_route_update_msg(bth, ctx);
case DPLANE_OP_NH_INSTALL:
case DPLANE_OP_NH_UPDATE:
case DPLANE_OP_NH_DELETE:
return netlink_put_nexthop_update_msg(bth, ctx);
case DPLANE_OP_LSP_INSTALL:
case DPLANE_OP_LSP_UPDATE:
case DPLANE_OP_LSP_DELETE:
return netlink_put_lsp_update_msg(bth, ctx);
case DPLANE_OP_PW_INSTALL:
case DPLANE_OP_PW_UNINSTALL:
return netlink_put_pw_update_msg(bth, ctx);
case DPLANE_OP_ADDR_INSTALL:
case DPLANE_OP_ADDR_UNINSTALL:
return netlink_put_address_update_msg(bth, ctx);
case DPLANE_OP_MAC_INSTALL:
case DPLANE_OP_MAC_DELETE:
return netlink_put_mac_update_msg(bth, ctx);
case DPLANE_OP_NEIGH_INSTALL:
case DPLANE_OP_NEIGH_UPDATE:
case DPLANE_OP_NEIGH_DELETE:
case DPLANE_OP_VTEP_ADD:
case DPLANE_OP_VTEP_DELETE:
case DPLANE_OP_NEIGH_DISCOVER:
case DPLANE_OP_NEIGH_IP_INSTALL:
case DPLANE_OP_NEIGH_IP_DELETE:
case DPLANE_OP_NEIGH_TABLE_UPDATE:
return netlink_put_neigh_update_msg(bth, ctx);
case DPLANE_OP_RULE_ADD:
case DPLANE_OP_RULE_DELETE:
case DPLANE_OP_RULE_UPDATE:
return netlink_put_rule_update_msg(bth, ctx);
case DPLANE_OP_SYS_ROUTE_ADD:
case DPLANE_OP_SYS_ROUTE_DELETE:
case DPLANE_OP_ROUTE_NOTIFY:
case DPLANE_OP_LSP_NOTIFY:
case DPLANE_OP_BR_PORT_UPDATE:
return FRR_NETLINK_SUCCESS;
case DPLANE_OP_IPTABLE_ADD:
case DPLANE_OP_IPTABLE_DELETE:
case DPLANE_OP_IPSET_ADD:
case DPLANE_OP_IPSET_DELETE:
case DPLANE_OP_IPSET_ENTRY_ADD:
case DPLANE_OP_IPSET_ENTRY_DELETE:
case DPLANE_OP_STARTUP_STAGE:
return FRR_NETLINK_ERROR;
case DPLANE_OP_GRE_SET:
return netlink_put_gre_set_msg(bth, ctx);
case DPLANE_OP_INTF_ADDR_ADD:
case DPLANE_OP_INTF_ADDR_DEL:
case DPLANE_OP_NONE:
return FRR_NETLINK_ERROR;
case DPLANE_OP_INTF_NETCONFIG:
return netlink_put_intf_netconfig(bth, ctx);
case DPLANE_OP_INTF_INSTALL:
case DPLANE_OP_INTF_UPDATE:
case DPLANE_OP_INTF_DELETE:
return netlink_put_intf_update_msg(bth, ctx);
case DPLANE_OP_TC_QDISC_INSTALL:
case DPLANE_OP_TC_QDISC_UNINSTALL:
return netlink_put_tc_qdisc_update_msg(bth, ctx);
case DPLANE_OP_TC_CLASS_ADD:
case DPLANE_OP_TC_CLASS_DELETE:
case DPLANE_OP_TC_CLASS_UPDATE:
return netlink_put_tc_class_update_msg(bth, ctx);
case DPLANE_OP_TC_FILTER_ADD:
case DPLANE_OP_TC_FILTER_DELETE:
case DPLANE_OP_TC_FILTER_UPDATE:
return netlink_put_tc_filter_update_msg(bth, ctx);
case DPLANE_OP_SRV6_ENCAP_SRCADDR_SET:
return netlink_put_sr_tunsrc_set_msg(bth, ctx);
}
return FRR_NETLINK_ERROR;
}
void kernel_update_multi(struct dplane_ctx_list_head *ctx_list)
{
struct nl_batch batch;
struct zebra_dplane_ctx *ctx;
struct dplane_ctx_list_head handled_list;
enum netlink_msg_status res;
dplane_ctx_q_init(&handled_list);
nl_batch_init(&batch, &handled_list);
while (true) {
ctx = dplane_ctx_dequeue(ctx_list);
if (ctx == NULL)
break;
if (batch.zns != NULL
&& batch.zns->ns_id != dplane_ctx_get_ns(ctx)->ns_id)
nl_batch_send(&batch);
/*
* Assume all messages will succeed and then mark only the ones
* that failed.
*/
dplane_ctx_set_status(ctx, ZEBRA_DPLANE_REQUEST_SUCCESS);
res = nl_put_msg(&batch, ctx);
dplane_ctx_enqueue_tail(&(batch.ctx_list), ctx);
if (res == FRR_NETLINK_ERROR)
dplane_ctx_set_status(ctx,
ZEBRA_DPLANE_REQUEST_FAILURE);
if (batch.curlen > batch.limit)
nl_batch_send(&batch);
}
nl_batch_send(&batch);
dplane_ctx_q_init(ctx_list);
dplane_ctx_list_append(ctx_list, &handled_list);
}
struct nlsock *kernel_netlink_nlsock_lookup(int sock)
{
struct nlsock lookup, *retval;
lookup.sock = sock;
NLSOCK_LOCK();
retval = hash_lookup(nlsock_hash, &lookup);
NLSOCK_UNLOCK();
return retval;
}
/* Insert nlsock entry into hash */
static void kernel_netlink_nlsock_insert(struct nlsock *nls)
{
NLSOCK_LOCK();
(void)hash_get(nlsock_hash, nls, hash_alloc_intern);
NLSOCK_UNLOCK();
}
/* Remove nlsock entry from hash */
static void kernel_netlink_nlsock_remove(struct nlsock *nls)
{
NLSOCK_LOCK();
(void)hash_release(nlsock_hash, nls);
NLSOCK_UNLOCK();
}
static uint32_t kernel_netlink_nlsock_key(const void *arg)
{
const struct nlsock *nl = arg;
return nl->sock;
}
static bool kernel_netlink_nlsock_hash_equal(const void *arg1, const void *arg2)
{
const struct nlsock *nl1 = arg1;
const struct nlsock *nl2 = arg2;
if (nl1->sock == nl2->sock)
return true;
return false;
}
/* Exported interface function. This function simply calls
netlink_socket (). */
void kernel_init(struct zebra_ns *zns)
{
uint32_t groups, dplane_groups, ext_groups;
#if defined SOL_NETLINK
int one, ret, grp;
#endif
/*
* Initialize netlink sockets
*
* If RTMGRP_XXX exists use that, but at some point
* I think the kernel developers realized that
* keeping track of all the different values would
* lead to confusion, so we need to convert the
* RTNLGRP_XXX to a bit position for ourself
*
*
* NOTE: If the bit is >= 32, you must use setsockopt(). Those
* groups are added further below after SOL_NETLINK is verified to
* exist.
*/
groups = RTMGRP_IPV4_ROUTE | RTMGRP_IPV6_ROUTE | RTMGRP_IPV4_MROUTE |
RTMGRP_NEIGH | ((uint32_t)1 << (RTNLGRP_IPV4_RULE - 1)) |
((uint32_t)1 << (RTNLGRP_IPV6_RULE - 1)) |
((uint32_t)1 << (RTNLGRP_NEXTHOP - 1)) |
((uint32_t)1 << (RTNLGRP_TC - 1));
dplane_groups = (RTMGRP_LINK |
RTMGRP_IPV4_IFADDR |
RTMGRP_IPV6_IFADDR |
((uint32_t) 1 << (RTNLGRP_IPV4_NETCONF - 1)) |
((uint32_t) 1 << (RTNLGRP_IPV6_NETCONF - 1)) |
((uint32_t) 1 << (RTNLGRP_MPLS_NETCONF - 1)));
/* Use setsockopt for > 31 group */
ext_groups = RTNLGRP_TUNNEL;
snprintf(zns->netlink.name, sizeof(zns->netlink.name),
"netlink-listen (NS %u)", zns->ns_id);
zns->netlink.sock = -1;
if (netlink_socket(&zns->netlink, groups, &ext_groups, 1, zns->ns_id,
NETLINK_ROUTE) < 0) {
zlog_err("Failure to create %s socket",
zns->netlink.name);
exit(-1);
}
kernel_netlink_nlsock_insert(&zns->netlink);
snprintf(zns->netlink_cmd.name, sizeof(zns->netlink_cmd.name),
"netlink-cmd (NS %u)", zns->ns_id);
zns->netlink_cmd.sock = -1;
if (netlink_socket(&zns->netlink_cmd, 0, 0, 0, zns->ns_id,
NETLINK_ROUTE) < 0) {
zlog_err("Failure to create %s socket",
zns->netlink_cmd.name);
exit(-1);
}
kernel_netlink_nlsock_insert(&zns->netlink_cmd);
/* Outbound socket for dplane programming of the host OS. */
snprintf(zns->netlink_dplane_out.name,
sizeof(zns->netlink_dplane_out.name), "netlink-dp (NS %u)",
zns->ns_id);
zns->netlink_dplane_out.sock = -1;
if (netlink_socket(&zns->netlink_dplane_out, 0, 0, 0, zns->ns_id,
NETLINK_ROUTE) < 0) {
zlog_err("Failure to create %s socket",
zns->netlink_dplane_out.name);
exit(-1);
}
kernel_netlink_nlsock_insert(&zns->netlink_dplane_out);
/* Inbound socket for OS events coming to the dplane. */
snprintf(zns->netlink_dplane_in.name,
sizeof(zns->netlink_dplane_in.name), "netlink-dp-in (NS %u)",
zns->ns_id);
zns->netlink_dplane_in.sock = -1;
if (netlink_socket(&zns->netlink_dplane_in, dplane_groups, 0, 0,
zns->ns_id, NETLINK_ROUTE) < 0) {
zlog_err("Failure to create %s socket",
zns->netlink_dplane_in.name);
exit(-1);
}
kernel_netlink_nlsock_insert(&zns->netlink_dplane_in);
/* Generic Netlink socket. */
snprintf(zns->ge_netlink_cmd.name, sizeof(zns->ge_netlink_cmd.name),
"generic-netlink-cmd (NS %u)", zns->ns_id);
zns->ge_netlink_cmd.sock = -1;
if (netlink_socket(&zns->ge_netlink_cmd, 0, 0, 0, zns->ns_id,
NETLINK_GENERIC) < 0) {
zlog_warn("Failure to create %s socket",
zns->ge_netlink_cmd.name);
}
if (zns->ge_netlink_cmd.sock >= 0)
kernel_netlink_nlsock_insert(&zns->ge_netlink_cmd);
/*
* SOL_NETLINK is not available on all platforms yet
* apparently. It's in bits/socket.h which I am not
* sure that we want to pull into our build system.
*/
#if defined SOL_NETLINK
/*
* setsockopt multicast group subscriptions that don't fit in nl_groups
*/
grp = RTNLGRP_BRVLAN;
ret = setsockopt(zns->netlink.sock, SOL_NETLINK, NETLINK_ADD_MEMBERSHIP,
&grp, sizeof(grp));
if (ret < 0)
zlog_notice(
"Registration for RTNLGRP_BRVLAN Membership failed : %d %s",
errno, safe_strerror(errno));
/*
* Let's tell the kernel that we want to receive extended
* ACKS over our command socket(s)
*/
one = 1;
ret = setsockopt(zns->netlink_cmd.sock, SOL_NETLINK, NETLINK_EXT_ACK,
&one, sizeof(one));
if (ret < 0)
zlog_notice("Registration for extended cmd ACK failed : %d %s",
errno, safe_strerror(errno));
one = 1;
ret = setsockopt(zns->netlink_dplane_out.sock, SOL_NETLINK,
NETLINK_EXT_ACK, &one, sizeof(one));
if (ret < 0)
zlog_notice("Registration for extended dp ACK failed : %d %s",
errno, safe_strerror(errno));
if (zns->ge_netlink_cmd.sock >= 0) {
one = 1;
ret = setsockopt(zns->ge_netlink_cmd.sock, SOL_NETLINK,
NETLINK_EXT_ACK, &one, sizeof(one));
if (ret < 0)
zlog_err("Registration for extended generic netlink cmd ACK failed : %d %s",
errno, safe_strerror(errno));
}
/*
* Trim off the payload of the original netlink message in the
* acknowledgment. This option is available since Linux 4.2, so if
* setsockopt fails, ignore the error.
*/
one = 1;
ret = setsockopt(zns->netlink_dplane_out.sock, SOL_NETLINK,
NETLINK_CAP_ACK, &one, sizeof(one));
if (ret < 0)
zlog_notice(
"Registration for reduced ACK packet size failed, probably running an early kernel");
#endif
/* Register kernel socket. */
if (fcntl(zns->netlink.sock, F_SETFL, O_NONBLOCK) < 0)
flog_err_sys(EC_LIB_SOCKET, "Can't set %s socket flags: %s",
zns->netlink.name, safe_strerror(errno));
if (fcntl(zns->netlink_cmd.sock, F_SETFL, O_NONBLOCK) < 0)
zlog_err("Can't set %s socket error: %s(%d)",
zns->netlink_cmd.name, safe_strerror(errno), errno);
if (fcntl(zns->netlink_dplane_out.sock, F_SETFL, O_NONBLOCK) < 0)
zlog_err("Can't set %s socket error: %s(%d)",
zns->netlink_dplane_out.name, safe_strerror(errno),
errno);
if (fcntl(zns->netlink_dplane_in.sock, F_SETFL, O_NONBLOCK) < 0)
zlog_err("Can't set %s socket error: %s(%d)",
zns->netlink_dplane_in.name, safe_strerror(errno),
errno);
if (zns->ge_netlink_cmd.sock >= 0) {
if (fcntl(zns->ge_netlink_cmd.sock, F_SETFL, O_NONBLOCK) < 0)
zlog_err("Can't set %s socket error: %s(%d)",
zns->ge_netlink_cmd.name, safe_strerror(errno),
errno);
}
/* Set receive buffer size if it's set from command line */
if (rcvbufsize) {
netlink_recvbuf(&zns->netlink, rcvbufsize);
netlink_recvbuf(&zns->netlink_cmd, rcvbufsize);
netlink_recvbuf(&zns->netlink_dplane_out, rcvbufsize);
netlink_recvbuf(&zns->netlink_dplane_in, rcvbufsize);
if (zns->ge_netlink_cmd.sock >= 0)
netlink_recvbuf(&zns->ge_netlink_cmd, rcvbufsize);
}
/* Set filter for inbound sockets, to exclude events we've generated
* ourselves.
*/
netlink_install_filter(zns->netlink.sock, zns->netlink_cmd.snl.nl_pid,
zns->netlink_dplane_out.snl.nl_pid);
netlink_install_filter(zns->netlink_dplane_in.sock,
zns->netlink_cmd.snl.nl_pid,
zns->netlink_dplane_out.snl.nl_pid);
zns->t_netlink = NULL;
event_add_read(zrouter.master, kernel_read, zns, zns->netlink.sock,
&zns->t_netlink);
rt_netlink_init();
ge_netlink_init(zns);
}
/* Helper to clean up an nlsock */
static void kernel_nlsock_fini(struct nlsock *nls)
{
if (nls && nls->sock >= 0) {
kernel_netlink_nlsock_remove(nls);
close(nls->sock);
nls->sock = -1;
XFREE(MTYPE_NL_BUF, nls->buf);
nls->buflen = 0;
}
}
void kernel_terminate(struct zebra_ns *zns, bool complete)
{
EVENT_OFF(zns->t_netlink);
kernel_nlsock_fini(&zns->netlink);
kernel_nlsock_fini(&zns->netlink_cmd);
kernel_nlsock_fini(&zns->netlink_dplane_in);
kernel_nlsock_fini(&zns->ge_netlink_cmd);
/* During zebra shutdown, we need to leave the dataplane socket
* around until all work is done.
*/
if (complete) {
kernel_nlsock_fini(&zns->netlink_dplane_out);
XFREE(MTYPE_NL_BUF, nl_batch_tx_buf);
}
}
/*
* Global init for platform-/OS-specific things
*/
void kernel_router_init(void)
{
/* Init nlsock hash and lock */
pthread_mutex_init(&nlsock_mutex, NULL);
nlsock_hash = hash_create_size(8, kernel_netlink_nlsock_key,
kernel_netlink_nlsock_hash_equal,
"Netlink Socket Hash");
}
/*
* Global deinit for platform-/OS-specific things
*/
void kernel_router_terminate(void)
{
pthread_mutex_destroy(&nlsock_mutex);
hash_free(nlsock_hash);
nlsock_hash = NULL;
}
#endif /* HAVE_NETLINK */
|