summaryrefslogtreecommitdiffstats
path: root/sys/netinet6/mld6.c
blob: ce5d6118749fc1c144432eddf01e63c6c938e141 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
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
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
2359
2360
2361
2362
2363
2364
2365
2366
2367
2368
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
2380
2381
2382
2383
2384
2385
2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
2396
2397
2398
2399
2400
2401
2402
2403
2404
2405
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418
2419
2420
2421
2422
2423
2424
2425
2426
2427
2428
2429
2430
2431
2432
2433
2434
2435
2436
2437
2438
2439
2440
2441
2442
2443
2444
2445
2446
2447
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
2464
2465
2466
2467
2468
2469
2470
2471
2472
2473
2474
2475
2476
2477
2478
2479
2480
2481
2482
2483
2484
2485
2486
2487
2488
2489
2490
2491
2492
2493
2494
2495
2496
2497
2498
2499
2500
2501
2502
2503
2504
2505
2506
2507
2508
2509
2510
2511
2512
2513
2514
2515
2516
2517
2518
2519
2520
2521
2522
2523
2524
2525
2526
2527
2528
2529
2530
2531
2532
2533
2534
2535
2536
2537
2538
2539
2540
2541
2542
2543
2544
2545
2546
2547
2548
2549
2550
2551
2552
2553
2554
2555
2556
2557
2558
2559
2560
2561
2562
2563
2564
2565
2566
2567
2568
2569
2570
2571
2572
2573
2574
2575
2576
2577
2578
2579
2580
2581
2582
2583
2584
2585
2586
2587
2588
2589
2590
2591
2592
2593
2594
2595
2596
2597
2598
2599
2600
2601
2602
2603
2604
2605
2606
2607
2608
2609
2610
2611
2612
2613
2614
2615
2616
2617
2618
2619
2620
2621
2622
2623
2624
2625
2626
2627
2628
2629
2630
2631
2632
2633
2634
2635
2636
2637
2638
2639
2640
2641
2642
2643
2644
2645
2646
2647
2648
2649
2650
2651
2652
2653
2654
2655
2656
2657
2658
2659
2660
2661
2662
2663
2664
2665
2666
2667
2668
2669
2670
2671
2672
2673
2674
2675
2676
2677
2678
2679
2680
2681
2682
2683
2684
2685
2686
2687
2688
2689
2690
2691
2692
2693
2694
2695
2696
2697
2698
2699
2700
2701
2702
2703
2704
2705
2706
2707
2708
2709
2710
2711
2712
2713
2714
2715
2716
2717
2718
2719
2720
2721
2722
2723
2724
2725
2726
2727
2728
2729
2730
2731
2732
2733
2734
2735
2736
2737
2738
2739
2740
2741
2742
2743
2744
2745
2746
2747
2748
2749
2750
2751
2752
2753
2754
2755
2756
2757
2758
2759
2760
2761
2762
2763
2764
2765
2766
2767
2768
2769
2770
2771
2772
2773
2774
2775
2776
2777
2778
2779
2780
2781
2782
2783
2784
2785
2786
2787
2788
2789
2790
2791
2792
2793
2794
2795
2796
2797
2798
2799
2800
2801
2802
2803
2804
2805
2806
2807
2808
2809
2810
2811
2812
2813
2814
2815
2816
2817
2818
2819
2820
2821
2822
2823
2824
2825
2826
2827
2828
2829
2830
2831
2832
2833
2834
2835
2836
2837
2838
2839
2840
2841
2842
2843
2844
2845
2846
2847
2848
2849
2850
2851
2852
2853
2854
2855
2856
2857
2858
2859
2860
2861
2862
2863
2864
2865
2866
2867
2868
2869
2870
2871
2872
2873
2874
2875
2876
2877
2878
2879
2880
2881
2882
2883
2884
2885
2886
2887
2888
2889
2890
2891
2892
2893
2894
2895
2896
2897
2898
2899
2900
2901
2902
2903
2904
2905
2906
2907
2908
2909
2910
2911
2912
2913
2914
2915
2916
2917
2918
2919
2920
2921
2922
2923
2924
2925
2926
2927
2928
2929
2930
2931
2932
2933
2934
2935
2936
2937
2938
2939
2940
2941
2942
2943
2944
2945
2946
2947
2948
2949
2950
2951
2952
2953
2954
2955
2956
2957
2958
2959
2960
2961
2962
2963
2964
2965
2966
2967
2968
2969
2970
2971
2972
2973
2974
2975
2976
2977
2978
2979
2980
2981
2982
2983
2984
2985
2986
2987
2988
2989
2990
2991
2992
2993
2994
2995
2996
2997
2998
2999
3000
3001
3002
3003
3004
3005
3006
3007
3008
3009
3010
3011
3012
3013
3014
3015
3016
3017
3018
3019
3020
3021
3022
3023
3024
3025
3026
3027
3028
3029
3030
3031
3032
3033
3034
3035
3036
3037
3038
3039
3040
3041
3042
3043
3044
3045
3046
3047
3048
3049
3050
3051
3052
3053
3054
3055
3056
3057
3058
3059
3060
3061
3062
3063
3064
3065
3066
3067
3068
3069
3070
3071
3072
3073
3074
3075
3076
3077
3078
3079
3080
3081
3082
3083
3084
3085
3086
3087
3088
3089
3090
3091
3092
3093
3094
3095
3096
3097
3098
3099
3100
3101
3102
3103
3104
3105
3106
3107
3108
3109
3110
3111
3112
3113
3114
3115
3116
3117
3118
3119
3120
3121
3122
3123
3124
3125
3126
3127
3128
3129
3130
3131
3132
3133
3134
3135
3136
3137
3138
3139
3140
3141
3142
3143
3144
3145
3146
3147
3148
3149
3150
3151
3152
3153
3154
3155
3156
3157
3158
3159
3160
3161
3162
3163
3164
3165
3166
3167
3168
3169
3170
3171
3172
3173
3174
3175
3176
3177
3178
3179
3180
3181
3182
3183
3184
3185
3186
3187
3188
3189
3190
3191
3192
3193
3194
3195
3196
3197
3198
3199
3200
3201
3202
3203
3204
3205
3206
3207
3208
3209
3210
3211
3212
3213
3214
3215
3216
3217
3218
3219
3220
3221
3222
3223
3224
3225
3226
3227
3228
3229
3230
3231
3232
3233
3234
3235
3236
3237
3238
3239
3240
3241
3242
3243
3244
3245
3246
3247
3248
3249
3250
3251
3252
3253
3254
3255
3256
3257
3258
3259
3260
3261
3262
3263
3264
3265
3266
3267
3268
3269
3270
3271
3272
3273
3274
3275
3276
3277
3278
3279
3280
3281
3282
3283
3284
3285
3286
3287
3288
3289
3290
3291
3292
3293
3294
3295
3296
3297
3298
3299
3300
3301
3302
3303
3304
3305
3306
3307
3308
3309
/*-
 * Copyright (c) 2009 Bruce Simpson.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 * 3. The name of the author may not be used to endorse or promote
 *    products derived from this software without specific prior written
 *    permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 * SUCH DAMAGE.
 *
 *	$KAME: mld6.c,v 1.27 2001/04/04 05:17:30 itojun Exp $
 */

/*-
 * Copyright (c) 1988 Stephen Deering.
 * Copyright (c) 1992, 1993
 *	The Regents of the University of California.  All rights reserved.
 *
 * This code is derived from software contributed to Berkeley by
 * Stephen Deering of Stanford University.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 * 4. Neither the name of the University nor the names of its contributors
 *    may be used to endorse or promote products derived from this software
 *    without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 * SUCH DAMAGE.
 *
 *	@(#)igmp.c	8.1 (Berkeley) 7/19/93
 */

#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");

#include "opt_inet.h"
#include "opt_inet6.h"

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/mbuf.h>
#include <sys/socket.h>
#include <sys/protosw.h>
#include <sys/sysctl.h>
#include <sys/kernel.h>
#include <sys/callout.h>
#include <sys/malloc.h>
#include <sys/module.h>
#include <sys/ktr.h>

#include <net/if.h>
#include <net/route.h>
#include <net/vnet.h>

#include <netinet/in.h>
#include <netinet/in_var.h>
#include <netinet6/in6_var.h>
#include <netinet/ip6.h>
#include <netinet6/ip6_var.h>
#include <netinet6/scope6_var.h>
#include <netinet/icmp6.h>
#include <netinet6/mld6.h>
#include <netinet6/mld6_var.h>

#include <security/mac/mac_framework.h>

#ifndef KTR_MLD
#define KTR_MLD KTR_INET6
#endif

static struct mld_ifinfo *
		mli_alloc_locked(struct ifnet *);
static void	mli_delete_locked(const struct ifnet *);
static void	mld_dispatch_packet(struct mbuf *);
static void	mld_dispatch_queue(struct ifqueue *, int);
static void	mld_final_leave(struct in6_multi *, struct mld_ifinfo *);
static void	mld_fasttimo_vnet(void);
static int	mld_handle_state_change(struct in6_multi *,
		    struct mld_ifinfo *);
static int	mld_initial_join(struct in6_multi *, struct mld_ifinfo *,
		    const int);
#ifdef KTR
static char *	mld_rec_type_to_str(const int);
#endif
static void	mld_set_version(struct mld_ifinfo *, const int);
static void	mld_slowtimo_vnet(void);
static int	mld_v1_input_query(struct ifnet *, const struct ip6_hdr *,
		    /*const*/ struct mld_hdr *);
static int	mld_v1_input_report(struct ifnet *, const struct ip6_hdr *,
		    /*const*/ struct mld_hdr *);
static void	mld_v1_process_group_timer(struct in6_multi *, const int);
static void	mld_v1_process_querier_timers(struct mld_ifinfo *);
static int	mld_v1_transmit_report(struct in6_multi *, const int);
static void	mld_v1_update_group(struct in6_multi *, const int);
static void	mld_v2_cancel_link_timers(struct mld_ifinfo *);
static void	mld_v2_dispatch_general_query(struct mld_ifinfo *);
static struct mbuf *
		mld_v2_encap_report(struct ifnet *, struct mbuf *);
static int	mld_v2_enqueue_filter_change(struct ifqueue *,
		    struct in6_multi *);
static int	mld_v2_enqueue_group_record(struct ifqueue *,
		    struct in6_multi *, const int, const int, const int,
		    const int);
static int	mld_v2_input_query(struct ifnet *, const struct ip6_hdr *,
		    struct mbuf *, const int, const int);
static int	mld_v2_merge_state_changes(struct in6_multi *,
		    struct ifqueue *);
static void	mld_v2_process_group_timers(struct mld_ifinfo *,
		    struct ifqueue *, struct ifqueue *,
		    struct in6_multi *, const int);
static int	mld_v2_process_group_query(struct in6_multi *,
		    struct mld_ifinfo *mli, int, struct mbuf *, const int);
static int	sysctl_mld_gsr(SYSCTL_HANDLER_ARGS);
static int	sysctl_mld_ifinfo(SYSCTL_HANDLER_ARGS);

/*
 * Normative references: RFC 2710, RFC 3590, RFC 3810.
 *
 * Locking:
 *  * The MLD subsystem lock ends up being system-wide for the moment,
 *    but could be per-VIMAGE later on.
 *  * The permitted lock order is: IN6_MULTI_LOCK, MLD_LOCK, IF_ADDR_LOCK.
 *    Any may be taken independently; if any are held at the same
 *    time, the above lock order must be followed.
 *  * IN6_MULTI_LOCK covers in_multi.
 *  * MLD_LOCK covers per-link state and any global variables in this file.
 *  * IF_ADDR_LOCK covers if_multiaddrs, which is used for a variety of
 *    per-link state iterators.
 *
 *  XXX LOR PREVENTION
 *  A special case for IPv6 is the in6_setscope() routine. ip6_output()
 *  will not accept an ifp; it wants an embedded scope ID, unlike
 *  ip_output(), which happily takes the ifp given to it. The embedded
 *  scope ID is only used by MLD to select the outgoing interface.
 *
 *  During interface attach and detach, MLD will take MLD_LOCK *after*
 *  the IF_AFDATA_LOCK.
 *  As in6_setscope() takes IF_AFDATA_LOCK then SCOPE_LOCK, we can't call
 *  it with MLD_LOCK held without triggering an LOR. A netisr with indirect
 *  dispatch could work around this, but we'd rather not do that, as it
 *  can introduce other races.
 *
 *  As such, we exploit the fact that the scope ID is just the interface
 *  index, and embed it in the IPv6 destination address accordingly.
 *  This is potentially NOT VALID for MLDv1 reports, as they
 *  are always sent to the multicast group itself; as MLDv2
 *  reports are always sent to ff02::16, this is not an issue
 *  when MLDv2 is in use.
 *
 *  This does not however eliminate the LOR when ip6_output() itself
 *  calls in6_setscope() internally whilst MLD_LOCK is held. This will
 *  trigger a LOR warning in WITNESS when the ifnet is detached.
 *
 *  The right answer is probably to make IF_AFDATA_LOCK an rwlock, given
 *  how it's used across the network stack. Here we're simply exploiting
 *  the fact that MLD runs at a similar layer in the stack to scope6.c.
 *
 * VIMAGE:
 *  * Each in6_multi corresponds to an ifp, and each ifp corresponds
 *    to a vnet in ifp->if_vnet.
 */
static struct mtx		 mld_mtx;
MALLOC_DEFINE(M_MLD, "mld", "mld state");

#define	MLD_EMBEDSCOPE(pin6, zoneid)					\
	if (IN6_IS_SCOPE_LINKLOCAL(pin6) ||				\
	    IN6_IS_ADDR_MC_INTFACELOCAL(pin6))				\
		(pin6)->s6_addr16[1] = htons((zoneid) & 0xFFFF)		\

/*
 * VIMAGE-wide globals.
 */
STATIC_VNET_DEFINE(struct timeval, mld_gsrdelay) = {10, 0};
STATIC_VNET_DEFINE(LIST_HEAD(, mld_ifinfo), mli_head);
STATIC_VNET_DEFINE(int, interface_timers_running6);
STATIC_VNET_DEFINE(int, state_change_timers_running6);
STATIC_VNET_DEFINE(int, current_state_timers_running6);

#define	V_mld_gsrdelay			VNET(mld_gsrdelay)
#define	V_mli_head			VNET(mli_head)
#define	V_interface_timers_running6	VNET(interface_timers_running6)
#define	V_state_change_timers_running6	VNET(state_change_timers_running6)
#define	V_current_state_timers_running6	VNET(current_state_timers_running6)

SYSCTL_DECL(_net_inet6);	/* Note: Not in any common header. */

SYSCTL_NODE(_net_inet6, OID_AUTO, mld, CTLFLAG_RW, 0,
    "IPv6 Multicast Listener Discovery");

/*
 * Virtualized sysctls.
 */
SYSCTL_VNET_PROC(_net_inet6_mld, OID_AUTO, gsrdelay,
    CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_MPSAFE,
    &VNET_NAME(mld_gsrdelay.tv_sec), 0, sysctl_mld_gsr, "I",
    "Rate limit for MLDv2 Group-and-Source queries in seconds");

/*
 * Non-virtualized sysctls.
 */
SYSCTL_NODE(_net_inet6_mld, OID_AUTO, ifinfo, CTLFLAG_RD | CTLFLAG_MPSAFE,
    sysctl_mld_ifinfo, "Per-interface MLDv2 state");

static int	mld_v1enable = 1;
SYSCTL_INT(_net_inet6_mld, OID_AUTO, v1enable, CTLFLAG_RW,
    &mld_v1enable, 0, "Enable fallback to MLDv1");
TUNABLE_INT("net.inet6.mld.v1enable", &mld_v1enable);

static int	mld_use_allow = 1;
SYSCTL_INT(_net_inet6_mld, OID_AUTO, use_allow, CTLFLAG_RW,
    &mld_use_allow, 0, "Use ALLOW/BLOCK for RFC 4604 SSM joins/leaves");
TUNABLE_INT("net.inet6.mld.use_allow", &mld_use_allow);

/*
 * Packed Router Alert option structure declaration.
 */
struct mld_raopt {
	struct ip6_hbh		hbh;
	struct ip6_opt		pad;
	struct ip6_opt_router	ra;
} __packed;

/*
 * Router Alert hop-by-hop option header.
 */
static struct mld_raopt mld_ra = {
	.hbh = { 0, 0 },
	.pad = { .ip6o_type = IP6OPT_PADN, 0 },
	.ra = {
	    .ip6or_type = IP6OPT_ROUTER_ALERT,
	    .ip6or_len = IP6OPT_RTALERT_LEN - 2,
	    .ip6or_value[0] = ((IP6OPT_RTALERT_MLD >> 8) & 0xFF),
	    .ip6or_value[1] = (IP6OPT_RTALERT_MLD & 0xFF)
	}
};
static struct ip6_pktopts mld_po;

static __inline void
mld_save_context(struct mbuf *m, struct ifnet *ifp)
{

#ifdef VIMAGE
	m->m_pkthdr.header = ifp->if_vnet;
#endif /* VIMAGE */
	m->m_pkthdr.flowid = ifp->if_index;
}

static __inline void
mld_scrub_context(struct mbuf *m)
{

	m->m_pkthdr.header = NULL;
	m->m_pkthdr.flowid = 0;
}

/*
 * Restore context from a queued output chain.
 * Return saved ifindex.
 *
 * VIMAGE: The assertion is there to make sure that we
 * actually called CURVNET_SET() with what's in the mbuf chain.
 */
static __inline uint32_t
mld_restore_context(struct mbuf *m)
{

#if defined(VIMAGE) && defined(INVARIANTS)
	KASSERT(curvnet == m->m_pkthdr.header,
	    ("%s: called when curvnet was not restored", __func__));
#endif
	return (m->m_pkthdr.flowid);
}

/*
 * Retrieve or set threshold between group-source queries in seconds.
 *
 * VIMAGE: Assume curvnet set by caller.
 * SMPng: NOTE: Serialized by MLD lock.
 */
static int
sysctl_mld_gsr(SYSCTL_HANDLER_ARGS)
{
	int error;
	int i;

	error = sysctl_wire_old_buffer(req, sizeof(int));
	if (error)
		return (error);

	MLD_LOCK();

	i = V_mld_gsrdelay.tv_sec;

	error = sysctl_handle_int(oidp, &i, 0, req);
	if (error || !req->newptr)
		goto out_locked;

	if (i < -1 || i >= 60) {
		error = EINVAL;
		goto out_locked;
	}

	CTR2(KTR_MLD, "change mld_gsrdelay from %d to %d",
	     V_mld_gsrdelay.tv_sec, i);
	V_mld_gsrdelay.tv_sec = i;

out_locked:
	MLD_UNLOCK();
	return (error);
}

/*
 * Expose struct mld_ifinfo to userland, keyed by ifindex.
 * For use by ifmcstat(8).
 *
 * SMPng: NOTE: Does an unlocked ifindex space read.
 * VIMAGE: Assume curvnet set by caller. The node handler itself
 * is not directly virtualized.
 */
static int
sysctl_mld_ifinfo(SYSCTL_HANDLER_ARGS)
{
	int			*name;
	int			 error;
	u_int			 namelen;
	struct ifnet		*ifp;
	struct mld_ifinfo	*mli;

	name = (int *)arg1;
	namelen = arg2;

	if (req->newptr != NULL)
		return (EPERM);

	if (namelen != 1)
		return (EINVAL);

	error = sysctl_wire_old_buffer(req, sizeof(struct mld_ifinfo));
	if (error)
		return (error);

	IN6_MULTI_LOCK();
	MLD_LOCK();

	if (name[0] <= 0 || name[0] > V_if_index) {
		error = ENOENT;
		goto out_locked;
	}

	error = ENOENT;

	ifp = ifnet_byindex(name[0]);
	if (ifp == NULL)
		goto out_locked;

	LIST_FOREACH(mli, &V_mli_head, mli_link) {
		if (ifp == mli->mli_ifp) {
			error = SYSCTL_OUT(req, mli,
			    sizeof(struct mld_ifinfo));
			break;
		}
	}

out_locked:
	MLD_UNLOCK();
	IN6_MULTI_UNLOCK();
	return (error);
}

/*
 * Dispatch an entire queue of pending packet chains.
 * VIMAGE: Assumes the vnet pointer has been set.
 */
static void
mld_dispatch_queue(struct ifqueue *ifq, int limit)
{
	struct mbuf *m;

	for (;;) {
		_IF_DEQUEUE(ifq, m);
		if (m == NULL)
			break;
		CTR3(KTR_MLD, "%s: dispatch %p from %p", __func__, ifq, m);
		mld_dispatch_packet(m);
		if (--limit == 0)
			break;
	}
}

/*
 * Filter outgoing MLD report state by group.
 *
 * Reports are ALWAYS suppressed for ALL-HOSTS (ff02::1)
 * and node-local addresses. However, kernel and socket consumers
 * always embed the KAME scope ID in the address provided, so strip it
 * when performing comparison.
 * Note: This is not the same as the *multicast* scope.
 *
 * Return zero if the given group is one for which MLD reports
 * should be suppressed, or non-zero if reports should be issued.
 */
static __inline int
mld_is_addr_reported(const struct in6_addr *addr)
{

	KASSERT(IN6_IS_ADDR_MULTICAST(addr), ("%s: not multicast", __func__));

	if (IPV6_ADDR_MC_SCOPE(addr) == IPV6_ADDR_SCOPE_NODELOCAL)
		return (0);

	if (IPV6_ADDR_MC_SCOPE(addr) == IPV6_ADDR_SCOPE_LINKLOCAL) {
		struct in6_addr tmp = *addr;
		in6_clearscope(&tmp);
		if (IN6_ARE_ADDR_EQUAL(&tmp, &in6addr_linklocal_allnodes))
			return (0);
	}

	return (1);
}

/*
 * Attach MLD when PF_INET6 is attached to an interface.
 *
 * SMPng: Normally called with IF_AFDATA_LOCK held.
 */
struct mld_ifinfo *
mld_domifattach(struct ifnet *ifp)
{
	struct mld_ifinfo *mli;

	CTR3(KTR_MLD, "%s: called for ifp %p(%s)",
	    __func__, ifp, ifp->if_xname);

	MLD_LOCK();

	mli = mli_alloc_locked(ifp);
	if (!(ifp->if_flags & IFF_MULTICAST))
		mli->mli_flags |= MLIF_SILENT;
	if (mld_use_allow)
		mli->mli_flags |= MLIF_USEALLOW;

	MLD_UNLOCK();

	return (mli);
}

/*
 * VIMAGE: assume curvnet set by caller.
 */
static struct mld_ifinfo *
mli_alloc_locked(/*const*/ struct ifnet *ifp)
{
	struct mld_ifinfo *mli;

	MLD_LOCK_ASSERT();

	mli = malloc(sizeof(struct mld_ifinfo), M_MLD, M_NOWAIT|M_ZERO);
	if (mli == NULL)
		goto out;

	mli->mli_ifp = ifp;
	mli->mli_version = MLD_VERSION_2;
	mli->mli_flags = 0;
	mli->mli_rv = MLD_RV_INIT;
	mli->mli_qi = MLD_QI_INIT;
	mli->mli_qri = MLD_QRI_INIT;
	mli->mli_uri = MLD_URI_INIT;

	SLIST_INIT(&mli->mli_relinmhead);

	/*
	 * Responses to general queries are subject to bounds.
	 */
	IFQ_SET_MAXLEN(&mli->mli_gq, MLD_MAX_RESPONSE_PACKETS);

	LIST_INSERT_HEAD(&V_mli_head, mli, mli_link);

	CTR2(KTR_MLD, "allocate mld_ifinfo for ifp %p(%s)",
	     ifp, ifp->if_xname);

out:
	return (mli);
}

/*
 * Hook for ifdetach.
 *
 * NOTE: Some finalization tasks need to run before the protocol domain
 * is detached, but also before the link layer does its cleanup.
 * Run before link-layer cleanup; cleanup groups, but do not free MLD state.
 *
 * SMPng: Caller must hold IN6_MULTI_LOCK().
 * Must take IF_ADDR_LOCK() to cover if_multiaddrs iterator.
 * XXX This routine is also bitten by unlocked ifma_protospec access.
 */
void
mld_ifdetach(struct ifnet *ifp)
{
	struct mld_ifinfo	*mli;
	struct ifmultiaddr	*ifma;
	struct in6_multi	*inm, *tinm;

	CTR3(KTR_MLD, "%s: called for ifp %p(%s)", __func__, ifp,
	    ifp->if_xname);

	IN6_MULTI_LOCK_ASSERT();
	MLD_LOCK();

	mli = MLD_IFINFO(ifp);
	if (mli->mli_version == MLD_VERSION_2) {
		IF_ADDR_LOCK(ifp);
		TAILQ_FOREACH(ifma, &ifp->if_multiaddrs, ifma_link) {
			if (ifma->ifma_addr->sa_family != AF_INET6 ||
			    ifma->ifma_protospec == NULL)
				continue;
			inm = (struct in6_multi *)ifma->ifma_protospec;
			if (inm->in6m_state == MLD_LEAVING_MEMBER) {
				SLIST_INSERT_HEAD(&mli->mli_relinmhead,
				    inm, in6m_nrele);
			}
			in6m_clear_recorded(inm);
		}
		IF_ADDR_UNLOCK(ifp);
		SLIST_FOREACH_SAFE(inm, &mli->mli_relinmhead, in6m_nrele,
		    tinm) {
			SLIST_REMOVE_HEAD(&mli->mli_relinmhead, in6m_nrele);
			in6m_release_locked(inm);
		}
	}

	MLD_UNLOCK();
}

/*
 * Hook for domifdetach.
 * Runs after link-layer cleanup; free MLD state.
 *
 * SMPng: Normally called with IF_AFDATA_LOCK held.
 */
void
mld_domifdetach(struct ifnet *ifp)
{

	CTR3(KTR_MLD, "%s: called for ifp %p(%s)",
	    __func__, ifp, ifp->if_xname);

	MLD_LOCK();
	mli_delete_locked(ifp);
	MLD_UNLOCK();
}

static void
mli_delete_locked(const struct ifnet *ifp)
{
	struct mld_ifinfo *mli, *tmli;

	CTR3(KTR_MLD, "%s: freeing mld_ifinfo for ifp %p(%s)",
	    __func__, ifp, ifp->if_xname);

	MLD_LOCK_ASSERT();

	LIST_FOREACH_SAFE(mli, &V_mli_head, mli_link, tmli) {
		if (mli->mli_ifp == ifp) {
			/*
			 * Free deferred General Query responses.
			 */
			_IF_DRAIN(&mli->mli_gq);

			LIST_REMOVE(mli, mli_link);

			KASSERT(SLIST_EMPTY(&mli->mli_relinmhead),
			    ("%s: there are dangling in_multi references",
			    __func__));

			free(mli, M_MLD);
			return;
		}
	}
#ifdef INVARIANTS
	panic("%s: mld_ifinfo not found for ifp %p\n", __func__,  ifp);
#endif
}

/*
 * Process a received MLDv1 general or address-specific query.
 * Assumes that the query header has been pulled up to sizeof(mld_hdr).
 *
 * NOTE: Can't be fully const correct as we temporarily embed scope ID in
 * mld_addr. This is OK as we own the mbuf chain.
 */
static int
mld_v1_input_query(struct ifnet *ifp, const struct ip6_hdr *ip6,
    /*const*/ struct mld_hdr *mld)
{
	struct ifmultiaddr	*ifma;
	struct mld_ifinfo	*mli;
	struct in6_multi	*inm;
	int			 is_general_query;
	uint16_t		 timer;
#ifdef KTR
	char			 ip6tbuf[INET6_ADDRSTRLEN];
#endif

	is_general_query = 0;

	if (!mld_v1enable) {
		CTR3(KTR_MLD, "ignore v1 query %s on ifp %p(%s)",
		    ip6_sprintf(ip6tbuf, &mld->mld_addr),
		    ifp, ifp->if_xname);
		return (0);
	}

	/*
	 * RFC3810 Section 6.2: MLD queries must originate from
	 * a router's link-local address.
	 */
	if (!IN6_IS_SCOPE_LINKLOCAL(&ip6->ip6_src)) {
		CTR3(KTR_MLD, "ignore v1 query src %s on ifp %p(%s)",
		    ip6_sprintf(ip6tbuf, &ip6->ip6_src),
		    ifp, ifp->if_xname);
		return (0);
	}

	/*
	 * Do address field validation upfront before we accept
	 * the query.
	 */
	if (IN6_IS_ADDR_UNSPECIFIED(&mld->mld_addr)) {
		/*
		 * MLDv1 General Query.
		 * If this was not sent to the all-nodes group, ignore it.
		 */
		struct in6_addr		 dst;

		dst = ip6->ip6_dst;
		in6_clearscope(&dst);
		if (!IN6_ARE_ADDR_EQUAL(&dst, &in6addr_linklocal_allnodes))
			return (EINVAL);
		is_general_query = 1;
	} else {
		/*
		 * Embed scope ID of receiving interface in MLD query for
		 * lookup whilst we don't hold other locks.
		 */
		in6_setscope(&mld->mld_addr, ifp, NULL);
	}

	IN6_MULTI_LOCK();
	MLD_LOCK();
	IF_ADDR_LOCK(ifp);

	/*
	 * Switch to MLDv1 host compatibility mode.
	 */
	mli = MLD_IFINFO(ifp);
	KASSERT(mli != NULL, ("%s: no mld_ifinfo for ifp %p", __func__, ifp));
	mld_set_version(mli, MLD_VERSION_1);

	timer = (ntohs(mld->mld_maxdelay) * PR_FASTHZ) / MLD_TIMER_SCALE;
	if (timer == 0)
		timer = 1;

	if (is_general_query) {
		/*
		 * For each reporting group joined on this
		 * interface, kick the report timer.
		 */
		CTR2(KTR_MLD, "process v1 general query on ifp %p(%s)",
		    ifp, ifp->if_xname);
		TAILQ_FOREACH(ifma, &ifp->if_multiaddrs, ifma_link) {
			if (ifma->ifma_addr->sa_family != AF_INET6 ||
			    ifma->ifma_protospec == NULL)
				continue;
			inm = (struct in6_multi *)ifma->ifma_protospec;
			mld_v1_update_group(inm, timer);
		}
	} else {
		/*
		 * MLDv1 Group-Specific Query.
		 * If this is a group-specific MLDv1 query, we need only
		 * look up the single group to process it.
		 */
		inm = in6m_lookup_locked(ifp, &mld->mld_addr);
		if (inm != NULL) {
			CTR3(KTR_MLD, "process v1 query %s on ifp %p(%s)",
			    ip6_sprintf(ip6tbuf, &mld->mld_addr),
			    ifp, ifp->if_xname);
			mld_v1_update_group(inm, timer);
		}
		/* XXX Clear embedded scope ID as userland won't expect it. */
		in6_clearscope(&mld->mld_addr);
	}

	IF_ADDR_UNLOCK(ifp);
	MLD_UNLOCK();
	IN6_MULTI_UNLOCK();

	return (0);
}

/*
 * Update the report timer on a group in response to an MLDv1 query.
 *
 * If we are becoming the reporting member for this group, start the timer.
 * If we already are the reporting member for this group, and timer is
 * below the threshold, reset it.
 *
 * We may be updating the group for the first time since we switched
 * to MLDv2. If we are, then we must clear any recorded source lists,
 * and transition to REPORTING state; the group timer is overloaded
 * for group and group-source query responses. 
 *
 * Unlike MLDv2, the delay per group should be jittered
 * to avoid bursts of MLDv1 reports.
 */
static void
mld_v1_update_group(struct in6_multi *inm, const int timer)
{
#ifdef KTR
	char			 ip6tbuf[INET6_ADDRSTRLEN];
#endif

	CTR4(KTR_MLD, "%s: %s/%s timer=%d", __func__,
	    ip6_sprintf(ip6tbuf, &inm->in6m_addr),
	    inm->in6m_ifp->if_xname, timer);

	IN6_MULTI_LOCK_ASSERT();

	switch (inm->in6m_state) {
	case MLD_NOT_MEMBER:
	case MLD_SILENT_MEMBER:
		break;
	case MLD_REPORTING_MEMBER:
		if (inm->in6m_timer != 0 &&
		    inm->in6m_timer <= timer) {
			CTR1(KTR_MLD, "%s: REPORTING and timer running, "
			    "skipping.", __func__);
			break;
		}
		/* FALLTHROUGH */
	case MLD_SG_QUERY_PENDING_MEMBER:
	case MLD_G_QUERY_PENDING_MEMBER:
	case MLD_IDLE_MEMBER:
	case MLD_LAZY_MEMBER:
	case MLD_AWAKENING_MEMBER:
		CTR1(KTR_MLD, "%s: ->REPORTING", __func__);
		inm->in6m_state = MLD_REPORTING_MEMBER;
		inm->in6m_timer = MLD_RANDOM_DELAY(timer);
		V_current_state_timers_running6 = 1;
		break;
	case MLD_SLEEPING_MEMBER:
		CTR1(KTR_MLD, "%s: ->AWAKENING", __func__);
		inm->in6m_state = MLD_AWAKENING_MEMBER;
		break;
	case MLD_LEAVING_MEMBER:
		break;
	}
}

/*
 * Process a received MLDv2 general, group-specific or
 * group-and-source-specific query.
 *
 * Assumes that the query header has been pulled up to sizeof(mldv2_query).
 *
 * Return 0 if successful, otherwise an appropriate error code is returned.
 */
static int
mld_v2_input_query(struct ifnet *ifp, const struct ip6_hdr *ip6,
    struct mbuf *m, const int off, const int icmp6len)
{
	struct mld_ifinfo	*mli;
	struct mldv2_query	*mld;
	struct in6_multi	*inm;
	uint32_t		 maxdelay, nsrc, qqi;
	int			 is_general_query;
	uint16_t		 timer;
	uint8_t			 qrv;
#ifdef KTR
	char			 ip6tbuf[INET6_ADDRSTRLEN];
#endif

	is_general_query = 0;

	/*
	 * RFC3810 Section 6.2: MLD queries must originate from
	 * a router's link-local address.
	 */
	if (!IN6_IS_SCOPE_LINKLOCAL(&ip6->ip6_src)) {
		CTR3(KTR_MLD, "ignore v1 query src %s on ifp %p(%s)",
		    ip6_sprintf(ip6tbuf, &ip6->ip6_src),
		    ifp, ifp->if_xname);
		return (0);
	}

	CTR2(KTR_MLD, "input v2 query on ifp %p(%s)", ifp, ifp->if_xname);

	mld = (struct mldv2_query *)(mtod(m, uint8_t *) + off);

	maxdelay = ntohs(mld->mld_maxdelay);	/* in 1/10ths of a second */
	if (maxdelay >= 32678) {
		maxdelay = (MLD_MRC_MANT(maxdelay) | 0x1000) <<
			   (MLD_MRC_EXP(maxdelay) + 3);
	}
	timer = (maxdelay * PR_FASTHZ) / MLD_TIMER_SCALE;
	if (timer == 0)
		timer = 1;

	qrv = MLD_QRV(mld->mld_misc);
	if (qrv < 2) {
		CTR3(KTR_MLD, "%s: clamping qrv %d to %d", __func__,
		    qrv, MLD_RV_INIT);
		qrv = MLD_RV_INIT;
	}

	qqi = mld->mld_qqi;
	if (qqi >= 128) {
		qqi = MLD_QQIC_MANT(mld->mld_qqi) <<
		     (MLD_QQIC_EXP(mld->mld_qqi) + 3);
	}

	nsrc = ntohs(mld->mld_numsrc);
	if (nsrc > MLD_MAX_GS_SOURCES)
		return (EMSGSIZE);
	if (icmp6len < sizeof(struct mldv2_query) +
	    (nsrc * sizeof(struct in6_addr)))
		return (EMSGSIZE);

	/*
	 * Do further input validation upfront to avoid resetting timers
	 * should we need to discard this query.
	 */
	if (IN6_IS_ADDR_UNSPECIFIED(&mld->mld_addr)) {
		/*
		 * General Queries SHOULD be directed to ff02::1.
		 * A general query with a source list has undefined
		 * behaviour; discard it.
		 */
		struct in6_addr		 dst;

		dst = ip6->ip6_dst;
		in6_clearscope(&dst);
		if (!IN6_ARE_ADDR_EQUAL(&dst, &in6addr_linklocal_allnodes) ||
		    nsrc > 0)
			return (EINVAL);
		is_general_query = 1;
	} else {
		/*
		 * Embed scope ID of receiving interface in MLD query for
		 * lookup whilst we don't hold other locks (due to KAME
		 * locking lameness). We own this mbuf chain just now.
		 */
		in6_setscope(&mld->mld_addr, ifp, NULL);
	}

	IN6_MULTI_LOCK();
	MLD_LOCK();
	IF_ADDR_LOCK(ifp);

	mli = MLD_IFINFO(ifp);
	KASSERT(mli != NULL, ("%s: no mld_ifinfo for ifp %p", __func__, ifp));

	/*
	 * Discard the v2 query if we're in Compatibility Mode.
	 * The RFC is pretty clear that hosts need to stay in MLDv1 mode
	 * until the Old Version Querier Present timer expires.
	 */
	if (mli->mli_version != MLD_VERSION_2)
		goto out_locked;

	mld_set_version(mli, MLD_VERSION_2);
	mli->mli_rv = qrv;
	mli->mli_qi = qqi;
	mli->mli_qri = maxdelay;

	CTR4(KTR_MLD, "%s: qrv %d qi %d maxdelay %d", __func__, qrv, qqi,
	    maxdelay);

	if (is_general_query) {
		/*
		 * MLDv2 General Query.
		 *
		 * Schedule a current-state report on this ifp for
		 * all groups, possibly containing source lists.
		 *
		 * If there is a pending General Query response
		 * scheduled earlier than the selected delay, do
		 * not schedule any other reports.
		 * Otherwise, reset the interface timer.
		 */
		CTR2(KTR_MLD, "process v2 general query on ifp %p(%s)",
		    ifp, ifp->if_xname);
		if (mli->mli_v2_timer == 0 || mli->mli_v2_timer >= timer) {
			mli->mli_v2_timer = MLD_RANDOM_DELAY(timer);
			V_interface_timers_running6 = 1;
		}
	} else {
		/*
		 * MLDv2 Group-specific or Group-and-source-specific Query.
		 *
		 * Group-source-specific queries are throttled on
		 * a per-group basis to defeat denial-of-service attempts.
		 * Queries for groups we are not a member of on this
		 * link are simply ignored.
		 */
		inm = in6m_lookup_locked(ifp, &mld->mld_addr);
		if (inm == NULL)
			goto out_locked;
		if (nsrc > 0) {
			if (!ratecheck(&inm->in6m_lastgsrtv,
			    &V_mld_gsrdelay)) {
				CTR1(KTR_MLD, "%s: GS query throttled.",
				    __func__);
				goto out_locked;
			}
		}
		CTR2(KTR_MLD, "process v2 group query on ifp %p(%s)",
		     ifp, ifp->if_xname);
		/*
		 * If there is a pending General Query response
		 * scheduled sooner than the selected delay, no
		 * further report need be scheduled.
		 * Otherwise, prepare to respond to the
		 * group-specific or group-and-source query.
		 */
		if (mli->mli_v2_timer == 0 || mli->mli_v2_timer >= timer)
			mld_v2_process_group_query(inm, mli, timer, m, off);

		/* XXX Clear embedded scope ID as userland won't expect it. */
		in6_clearscope(&mld->mld_addr);
	}

out_locked:
	IF_ADDR_UNLOCK(ifp);
	MLD_UNLOCK();
	IN6_MULTI_UNLOCK();

	return (0);
}

/*
 * Process a recieved MLDv2 group-specific or group-and-source-specific
 * query.
 * Return <0 if any error occured. Currently this is ignored.
 */
static int
mld_v2_process_group_query(struct in6_multi *inm, struct mld_ifinfo *mli,
    int timer, struct mbuf *m0, const int off)
{
	struct mldv2_query	*mld;
	int			 retval;
	uint16_t		 nsrc;

	IN6_MULTI_LOCK_ASSERT();
	MLD_LOCK_ASSERT();

	retval = 0;
	mld = (struct mldv2_query *)(mtod(m0, uint8_t *) + off);

	switch (inm->in6m_state) {
	case MLD_NOT_MEMBER:
	case MLD_SILENT_MEMBER:
	case MLD_SLEEPING_MEMBER:
	case MLD_LAZY_MEMBER:
	case MLD_AWAKENING_MEMBER:
	case MLD_IDLE_MEMBER:
	case MLD_LEAVING_MEMBER:
		return (retval);
		break;
	case MLD_REPORTING_MEMBER:
	case MLD_G_QUERY_PENDING_MEMBER:
	case MLD_SG_QUERY_PENDING_MEMBER:
		break;
	}

	nsrc = ntohs(mld->mld_numsrc);

	/*
	 * Deal with group-specific queries upfront.
	 * If any group query is already pending, purge any recorded
	 * source-list state if it exists, and schedule a query response
	 * for this group-specific query.
	 */
	if (nsrc == 0) {
		if (inm->in6m_state == MLD_G_QUERY_PENDING_MEMBER ||
		    inm->in6m_state == MLD_SG_QUERY_PENDING_MEMBER) {
			in6m_clear_recorded(inm);
			timer = min(inm->in6m_timer, timer);
		}
		inm->in6m_state = MLD_G_QUERY_PENDING_MEMBER;
		inm->in6m_timer = MLD_RANDOM_DELAY(timer);
		V_current_state_timers_running6 = 1;
		return (retval);
	}

	/*
	 * Deal with the case where a group-and-source-specific query has
	 * been received but a group-specific query is already pending.
	 */
	if (inm->in6m_state == MLD_G_QUERY_PENDING_MEMBER) {
		timer = min(inm->in6m_timer, timer);
		inm->in6m_timer = MLD_RANDOM_DELAY(timer);
		V_current_state_timers_running6 = 1;
		return (retval);
	}

	/*
	 * Finally, deal with the case where a group-and-source-specific
	 * query has been received, where a response to a previous g-s-r
	 * query exists, or none exists.
	 * In this case, we need to parse the source-list which the Querier
	 * has provided us with and check if we have any source list filter
	 * entries at T1 for these sources. If we do not, there is no need
	 * schedule a report and the query may be dropped.
	 * If we do, we must record them and schedule a current-state
	 * report for those sources.
	 */
	if (inm->in6m_nsrc > 0) {
		struct mbuf		*m;
		uint8_t			*sp;
		int			 i, nrecorded;
		int			 soff;

		m = m0;
		soff = off + sizeof(struct mldv2_query);
		nrecorded = 0;
		for (i = 0; i < nsrc; i++) {
			sp = mtod(m, uint8_t *) + soff;
			retval = in6m_record_source(inm,
			    (const struct in6_addr *)sp);
			if (retval < 0)
				break;
			nrecorded += retval;
			soff += sizeof(struct in6_addr);
			if (soff >= m->m_len) {
				soff = soff - m->m_len;
				m = m->m_next;
				if (m == NULL)
					break;
			}
		}
		if (nrecorded > 0) {
			CTR1(KTR_MLD,
			    "%s: schedule response to SG query", __func__);
			inm->in6m_state = MLD_SG_QUERY_PENDING_MEMBER;
			inm->in6m_timer = MLD_RANDOM_DELAY(timer);
			V_current_state_timers_running6 = 1;
		}
	}

	return (retval);
}

/*
 * Process a received MLDv1 host membership report.
 * Assumes mld points to mld_hdr in pulled up mbuf chain.
 *
 * NOTE: Can't be fully const correct as we temporarily embed scope ID in
 * mld_addr. This is OK as we own the mbuf chain.
 */
static int
mld_v1_input_report(struct ifnet *ifp, const struct ip6_hdr *ip6,
    /*const*/ struct mld_hdr *mld)
{
	struct in6_addr		 src, dst;
	struct in6_ifaddr	*ia;
	struct in6_multi	*inm;
#ifdef KTR
	char			 ip6tbuf[INET6_ADDRSTRLEN];
#endif

	if (!mld_v1enable) {
		CTR3(KTR_MLD, "ignore v1 report %s on ifp %p(%s)",
		    ip6_sprintf(ip6tbuf, &mld->mld_addr),
		    ifp, ifp->if_xname);
		return (0);
	}

	if (ifp->if_flags & IFF_LOOPBACK)
		return (0);

	/*
	 * MLDv1 reports must originate from a host's link-local address,
	 * or the unspecified address (when booting).
	 */
	src = ip6->ip6_src;
	in6_clearscope(&src);
	if (!IN6_IS_SCOPE_LINKLOCAL(&src) && !IN6_IS_ADDR_UNSPECIFIED(&src)) {
		CTR3(KTR_MLD, "ignore v1 query src %s on ifp %p(%s)",
		    ip6_sprintf(ip6tbuf, &ip6->ip6_src),
		    ifp, ifp->if_xname);
		return (EINVAL);
	}

	/*
	 * RFC2710 Section 4: MLDv1 reports must pertain to a multicast
	 * group, and must be directed to the group itself.
	 */
	dst = ip6->ip6_dst;
	in6_clearscope(&dst);
	if (!IN6_IS_ADDR_MULTICAST(&mld->mld_addr) ||
	    !IN6_ARE_ADDR_EQUAL(&mld->mld_addr, &dst)) {
		CTR3(KTR_MLD, "ignore v1 query dst %s on ifp %p(%s)",
		    ip6_sprintf(ip6tbuf, &ip6->ip6_dst),
		    ifp, ifp->if_xname);
		return (EINVAL);
	}

	/*
	 * Make sure we don't hear our own membership report, as fast
	 * leave requires knowing that we are the only member of a
	 * group. Assume we used the link-local address if available,
	 * otherwise look for ::.
	 *
	 * XXX Note that scope ID comparison is needed for the address
	 * returned by in6ifa_ifpforlinklocal(), but SHOULD NOT be
	 * performed for the on-wire address.
	 */
	ia = in6ifa_ifpforlinklocal(ifp, IN6_IFF_NOTREADY|IN6_IFF_ANYCAST);
	if ((ia && IN6_ARE_ADDR_EQUAL(&ip6->ip6_src, IA6_IN6(ia))) ||
	    (ia == NULL && IN6_IS_ADDR_UNSPECIFIED(&src))) {
		if (ia != NULL)
			ifa_free(&ia->ia_ifa);
		return (0);
	}
	if (ia != NULL)
		ifa_free(&ia->ia_ifa);

	CTR3(KTR_MLD, "process v1 report %s on ifp %p(%s)",
	    ip6_sprintf(ip6tbuf, &mld->mld_addr), ifp, ifp->if_xname);

	/*
	 * Embed scope ID of receiving interface in MLD query for lookup
	 * whilst we don't hold other locks (due to KAME locking lameness).
	 */
	if (!IN6_IS_ADDR_UNSPECIFIED(&mld->mld_addr))
		in6_setscope(&mld->mld_addr, ifp, NULL);

	IN6_MULTI_LOCK();
	MLD_LOCK();
	IF_ADDR_LOCK(ifp);

	/*
	 * MLDv1 report suppression.
	 * If we are a member of this group, and our membership should be
	 * reported, and our group timer is pending or about to be reset,
	 * stop our group timer by transitioning to the 'lazy' state.
	 */
	inm = in6m_lookup_locked(ifp, &mld->mld_addr);
	if (inm != NULL) {
		struct mld_ifinfo *mli;

		mli = inm->in6m_mli;
		KASSERT(mli != NULL,
		    ("%s: no mli for ifp %p", __func__, ifp));

		/*
		 * If we are in MLDv2 host mode, do not allow the
		 * other host's MLDv1 report to suppress our reports.
		 */
		if (mli->mli_version == MLD_VERSION_2)
			goto out_locked;

		inm->in6m_timer = 0;

		switch (inm->in6m_state) {
		case MLD_NOT_MEMBER:
		case MLD_SILENT_MEMBER:
		case MLD_SLEEPING_MEMBER:
			break;
		case MLD_REPORTING_MEMBER:
		case MLD_IDLE_MEMBER:
		case MLD_AWAKENING_MEMBER:
			CTR3(KTR_MLD,
			    "report suppressed for %s on ifp %p(%s)",
			    ip6_sprintf(ip6tbuf, &mld->mld_addr),
			    ifp, ifp->if_xname);
		case MLD_LAZY_MEMBER:
			inm->in6m_state = MLD_LAZY_MEMBER;
			break;
		case MLD_G_QUERY_PENDING_MEMBER:
		case MLD_SG_QUERY_PENDING_MEMBER:
		case MLD_LEAVING_MEMBER:
			break;
		}
	}

out_locked:
	MLD_UNLOCK();
	IF_ADDR_UNLOCK(ifp);
	IN6_MULTI_UNLOCK();

	/* XXX Clear embedded scope ID as userland won't expect it. */
	in6_clearscope(&mld->mld_addr);

	return (0);
}

/*
 * MLD input path.
 *
 * Assume query messages which fit in a single ICMPv6 message header
 * have been pulled up.
 * Assume that userland will want to see the message, even if it
 * otherwise fails kernel input validation; do not free it.
 * Pullup may however free the mbuf chain m if it fails.
 *
 * Return IPPROTO_DONE if we freed m. Otherwise, return 0.
 */
int
mld_input(struct mbuf *m, int off, int icmp6len)
{
	struct ifnet	*ifp;
	struct ip6_hdr	*ip6;
	struct mld_hdr	*mld;
	int		 mldlen;

	CTR3(KTR_MLD, "%s: called w/mbuf (%p,%d)", __func__, m, off);

	ifp = m->m_pkthdr.rcvif;

	ip6 = mtod(m, struct ip6_hdr *);

	/* Pullup to appropriate size. */
	mld = (struct mld_hdr *)(mtod(m, uint8_t *) + off);
	if (mld->mld_type == MLD_LISTENER_QUERY &&
	    icmp6len >= sizeof(struct mldv2_query)) {
		mldlen = sizeof(struct mldv2_query);
	} else {
		mldlen = sizeof(struct mld_hdr);
	}
	IP6_EXTHDR_GET(mld, struct mld_hdr *, m, off, mldlen);
	if (mld == NULL) {
		ICMP6STAT_INC(icp6s_badlen);
		return (IPPROTO_DONE);
	}

	/*
	 * Userland needs to see all of this traffic for implementing
	 * the endpoint discovery portion of multicast routing.
	 */
	switch (mld->mld_type) {
	case MLD_LISTENER_QUERY:
		icmp6_ifstat_inc(ifp, ifs6_in_mldquery);
		if (icmp6len == sizeof(struct mld_hdr)) {
			if (mld_v1_input_query(ifp, ip6, mld) != 0)
				return (0);
		} else if (icmp6len >= sizeof(struct mldv2_query)) {
			if (mld_v2_input_query(ifp, ip6, m, off,
			    icmp6len) != 0)
				return (0);
		}
		break;
	case MLD_LISTENER_REPORT:
		icmp6_ifstat_inc(ifp, ifs6_in_mldreport);
		if (mld_v1_input_report(ifp, ip6, mld) != 0)
			return (0);
		break;
	case MLDV2_LISTENER_REPORT:
		icmp6_ifstat_inc(ifp, ifs6_in_mldreport);
		break;
	case MLD_LISTENER_DONE:
		icmp6_ifstat_inc(ifp, ifs6_in_mlddone);
		break;
	default:
		break;
	}

	return (0);
}

/*
 * Fast timeout handler (global).
 * VIMAGE: Timeout handlers are expected to service all vimages.
 */
void
mld_fasttimo(void)
{
	VNET_ITERATOR_DECL(vnet_iter);

	VNET_LIST_RLOCK_NOSLEEP();
	VNET_FOREACH(vnet_iter) {
		CURVNET_SET(vnet_iter);
		mld_fasttimo_vnet();
		CURVNET_RESTORE();
	}
	VNET_LIST_RUNLOCK_NOSLEEP();
}

/*
 * Fast timeout handler (per-vnet).
 *
 * VIMAGE: Assume caller has set up our curvnet.
 */
static void
mld_fasttimo_vnet(void)
{
	struct ifqueue		 scq;	/* State-change packets */
	struct ifqueue		 qrq;	/* Query response packets */
	struct ifnet		*ifp;
	struct mld_ifinfo	*mli;
	struct ifmultiaddr	*ifma, *tifma;
	struct in6_multi	*inm;
	int			 uri_fasthz;

	uri_fasthz = 0;

	/*
	 * Quick check to see if any work needs to be done, in order to
	 * minimize the overhead of fasttimo processing.
	 * SMPng: XXX Unlocked reads.
	 */
	if (!V_current_state_timers_running6 &&
	    !V_interface_timers_running6 &&
	    !V_state_change_timers_running6)
		return;

	IN6_MULTI_LOCK();
	MLD_LOCK();

	/*
	 * MLDv2 General Query response timer processing.
	 */
	if (V_interface_timers_running6) {
		CTR1(KTR_MLD, "%s: interface timers running", __func__);

		V_interface_timers_running6 = 0;
		LIST_FOREACH(mli, &V_mli_head, mli_link) {
			if (mli->mli_v2_timer == 0) {
				/* Do nothing. */
			} else if (--mli->mli_v2_timer == 0) {
				mld_v2_dispatch_general_query(mli);
			} else {
				V_interface_timers_running6 = 1;
			}
		}
	}

	if (!V_current_state_timers_running6 &&
	    !V_state_change_timers_running6)
		goto out_locked;

	V_current_state_timers_running6 = 0;
	V_state_change_timers_running6 = 0;

	CTR1(KTR_MLD, "%s: state change timers running", __func__);

	/*
	 * MLD host report and state-change timer processing.
	 * Note: Processing a v2 group timer may remove a node.
	 */
	LIST_FOREACH(mli, &V_mli_head, mli_link) {
		ifp = mli->mli_ifp;

		if (mli->mli_version == MLD_VERSION_2) {
			uri_fasthz = MLD_RANDOM_DELAY(mli->mli_uri *
			    PR_FASTHZ);

			memset(&qrq, 0, sizeof(struct ifqueue));
			IFQ_SET_MAXLEN(&qrq, MLD_MAX_G_GS_PACKETS);

			memset(&scq, 0, sizeof(struct ifqueue));
			IFQ_SET_MAXLEN(&scq, MLD_MAX_STATE_CHANGE_PACKETS);
		}

		IF_ADDR_LOCK(ifp);
		TAILQ_FOREACH_SAFE(ifma, &ifp->if_multiaddrs, ifma_link,
		    tifma) {
			if (ifma->ifma_addr->sa_family != AF_INET6 ||
			    ifma->ifma_protospec == NULL)
				continue;
			inm = (struct in6_multi *)ifma->ifma_protospec;
			switch (mli->mli_version) {
			case MLD_VERSION_1:
				/*
				 * XXX Drop IF_ADDR lock temporarily to
				 * avoid recursion caused by a potential
				 * call by in6ifa_ifpforlinklocal().
				 * rwlock candidate?
				 */
				IF_ADDR_UNLOCK(ifp);
				mld_v1_process_group_timer(inm,
				    mli->mli_version);
				IF_ADDR_LOCK(ifp);
				break;
			case MLD_VERSION_2:
				mld_v2_process_group_timers(mli, &qrq,
				    &scq, inm, uri_fasthz);
				break;
			}
		}
		IF_ADDR_UNLOCK(ifp);

		if (mli->mli_version == MLD_VERSION_2) {
			struct in6_multi		*tinm;

			mld_dispatch_queue(&qrq, 0);
			mld_dispatch_queue(&scq, 0);

			/*
			 * Free the in_multi reference(s) for
			 * this lifecycle.
			 */
			SLIST_FOREACH_SAFE(inm, &mli->mli_relinmhead,
			    in6m_nrele, tinm) {
				SLIST_REMOVE_HEAD(&mli->mli_relinmhead,
				    in6m_nrele);
				in6m_release_locked(inm);
			}
		}
	}

out_locked:
	MLD_UNLOCK();
	IN6_MULTI_UNLOCK();
}

/*
 * Update host report group timer.
 * Will update the global pending timer flags.
 */
static void
mld_v1_process_group_timer(struct in6_multi *inm, const int version)
{
	int report_timer_expired;

	IN6_MULTI_LOCK_ASSERT();
	MLD_LOCK_ASSERT();

	if (inm->in6m_timer == 0) {
		report_timer_expired = 0;
	} else if (--inm->in6m_timer == 0) {
		report_timer_expired = 1;
	} else {
		V_current_state_timers_running6 = 1;
		return;
	}

	switch (inm->in6m_state) {
	case MLD_NOT_MEMBER:
	case MLD_SILENT_MEMBER:
	case MLD_IDLE_MEMBER:
	case MLD_LAZY_MEMBER:
	case MLD_SLEEPING_MEMBER:
	case MLD_AWAKENING_MEMBER:
		break;
	case MLD_REPORTING_MEMBER:
		if (report_timer_expired) {
			inm->in6m_state = MLD_IDLE_MEMBER;
			(void)mld_v1_transmit_report(inm,
			     MLD_LISTENER_REPORT);
		}
		break;
	case MLD_G_QUERY_PENDING_MEMBER:
	case MLD_SG_QUERY_PENDING_MEMBER:
	case MLD_LEAVING_MEMBER:
		break;
	}
}

/*
 * Update a group's timers for MLDv2.
 * Will update the global pending timer flags.
 * Note: Unlocked read from mli.
 */
static void
mld_v2_process_group_timers(struct mld_ifinfo *mli,
    struct ifqueue *qrq, struct ifqueue *scq,
    struct in6_multi *inm, const int uri_fasthz)
{
	int query_response_timer_expired;
	int state_change_retransmit_timer_expired;
#ifdef KTR
	char ip6tbuf[INET6_ADDRSTRLEN];
#endif

	IN6_MULTI_LOCK_ASSERT();
	MLD_LOCK_ASSERT();

	query_response_timer_expired = 0;
	state_change_retransmit_timer_expired = 0;

	/*
	 * During a transition from compatibility mode back to MLDv2,
	 * a group record in REPORTING state may still have its group
	 * timer active. This is a no-op in this function; it is easier
	 * to deal with it here than to complicate the slow-timeout path.
	 */
	if (inm->in6m_timer == 0) {
		query_response_timer_expired = 0;
	} else if (--inm->in6m_timer == 0) {
		query_response_timer_expired = 1;
	} else {
		V_current_state_timers_running6 = 1;
	}

	if (inm->in6m_sctimer == 0) {
		state_change_retransmit_timer_expired = 0;
	} else if (--inm->in6m_sctimer == 0) {
		state_change_retransmit_timer_expired = 1;
	} else {
		V_state_change_timers_running6 = 1;
	}

	/* We are in fasttimo, so be quick about it. */
	if (!state_change_retransmit_timer_expired &&
	    !query_response_timer_expired)
		return;

	switch (inm->in6m_state) {
	case MLD_NOT_MEMBER:
	case MLD_SILENT_MEMBER:
	case MLD_SLEEPING_MEMBER:
	case MLD_LAZY_MEMBER:
	case MLD_AWAKENING_MEMBER:
	case MLD_IDLE_MEMBER:
		break;
	case MLD_G_QUERY_PENDING_MEMBER:
	case MLD_SG_QUERY_PENDING_MEMBER:
		/*
		 * Respond to a previously pending Group-Specific
		 * or Group-and-Source-Specific query by enqueueing
		 * the appropriate Current-State report for
		 * immediate transmission.
		 */
		if (query_response_timer_expired) {
			int retval;

			retval = mld_v2_enqueue_group_record(qrq, inm, 0, 1,
			    (inm->in6m_state == MLD_SG_QUERY_PENDING_MEMBER),
			    0);
			CTR2(KTR_MLD, "%s: enqueue record = %d",
			    __func__, retval);
			inm->in6m_state = MLD_REPORTING_MEMBER;
			in6m_clear_recorded(inm);
		}
		/* FALLTHROUGH */
	case MLD_REPORTING_MEMBER:
	case MLD_LEAVING_MEMBER:
		if (state_change_retransmit_timer_expired) {
			/*
			 * State-change retransmission timer fired.
			 * If there are any further pending retransmissions,
			 * set the global pending state-change flag, and
			 * reset the timer.
			 */
			if (--inm->in6m_scrv > 0) {
				inm->in6m_sctimer = uri_fasthz;
				V_state_change_timers_running6 = 1;
			}
			/*
			 * Retransmit the previously computed state-change
			 * report. If there are no further pending
			 * retransmissions, the mbuf queue will be consumed.
			 * Update T0 state to T1 as we have now sent
			 * a state-change.
			 */
			(void)mld_v2_merge_state_changes(inm, scq);

			in6m_commit(inm);
			CTR3(KTR_MLD, "%s: T1 -> T0 for %s/%s", __func__,
			    ip6_sprintf(ip6tbuf, &inm->in6m_addr),
			    inm->in6m_ifp->if_xname);

			/*
			 * If we are leaving the group for good, make sure
			 * we release MLD's reference to it.
			 * This release must be deferred using a SLIST,
			 * as we are called from a loop which traverses
			 * the in_ifmultiaddr TAILQ.
			 */
			if (inm->in6m_state == MLD_LEAVING_MEMBER &&
			    inm->in6m_scrv == 0) {
				inm->in6m_state = MLD_NOT_MEMBER;
				SLIST_INSERT_HEAD(&mli->mli_relinmhead,
				    inm, in6m_nrele);
			}
		}
		break;
	}
}

/*
 * Switch to a different version on the given interface,
 * as per Section 9.12.
 */
static void
mld_set_version(struct mld_ifinfo *mli, const int version)
{
	int old_version_timer;

	MLD_LOCK_ASSERT();

	CTR4(KTR_MLD, "%s: switching to v%d on ifp %p(%s)", __func__,
	    version, mli->mli_ifp, mli->mli_ifp->if_xname);

	if (version == MLD_VERSION_1) {
		/*
		 * Compute the "Older Version Querier Present" timer as per
		 * Section 9.12.
		 */
		old_version_timer = (mli->mli_rv * mli->mli_qi) + mli->mli_qri;
		old_version_timer *= PR_SLOWHZ;
		mli->mli_v1_timer = old_version_timer;
	}

	if (mli->mli_v1_timer > 0 && mli->mli_version != MLD_VERSION_1) {
		mli->mli_version = MLD_VERSION_1;
		mld_v2_cancel_link_timers(mli);
	}
}

/*
 * Cancel pending MLDv2 timers for the given link and all groups
 * joined on it; state-change, general-query, and group-query timers.
 */
static void
mld_v2_cancel_link_timers(struct mld_ifinfo *mli)
{
	struct ifmultiaddr	*ifma;
	struct ifnet		*ifp;
	struct in6_multi		*inm;

	CTR3(KTR_MLD, "%s: cancel v2 timers on ifp %p(%s)", __func__,
	    mli->mli_ifp, mli->mli_ifp->if_xname);

	IN6_MULTI_LOCK_ASSERT();
	MLD_LOCK_ASSERT();

	/*
	 * Fast-track this potentially expensive operation
	 * by checking all the global 'timer pending' flags.
	 */
	if (!V_interface_timers_running6 &&
	    !V_state_change_timers_running6 &&
	    !V_current_state_timers_running6)
		return;

	mli->mli_v2_timer = 0;

	ifp = mli->mli_ifp;

	IF_ADDR_LOCK(ifp);
	TAILQ_FOREACH(ifma, &ifp->if_multiaddrs, ifma_link) {
		if (ifma->ifma_addr->sa_family != AF_INET6)
			continue;
		inm = (struct in6_multi *)ifma->ifma_protospec;
		switch (inm->in6m_state) {
		case MLD_NOT_MEMBER:
		case MLD_SILENT_MEMBER:
		case MLD_IDLE_MEMBER:
		case MLD_LAZY_MEMBER:
		case MLD_SLEEPING_MEMBER:
		case MLD_AWAKENING_MEMBER:
			break;
		case MLD_LEAVING_MEMBER:
			/*
			 * If we are leaving the group and switching
			 * version, we need to release the final
			 * reference held for issuing the INCLUDE {}.
			 *
			 * SMPNG: Must drop and re-acquire IF_ADDR_LOCK
			 * around in6m_release_locked(), as it is not
			 * a recursive mutex.
			 */
			IF_ADDR_UNLOCK(ifp);
			in6m_release_locked(inm);
			IF_ADDR_LOCK(ifp);
			/* FALLTHROUGH */
		case MLD_G_QUERY_PENDING_MEMBER:
		case MLD_SG_QUERY_PENDING_MEMBER:
			in6m_clear_recorded(inm);
			/* FALLTHROUGH */
		case MLD_REPORTING_MEMBER:
			inm->in6m_sctimer = 0;
			inm->in6m_timer = 0;
			inm->in6m_state = MLD_REPORTING_MEMBER;
			/*
			 * Free any pending MLDv2 state-change records.
			 */
			_IF_DRAIN(&inm->in6m_scq);
			break;
		}
	}
	IF_ADDR_UNLOCK(ifp);
}

/*
 * Global slowtimo handler.
 * VIMAGE: Timeout handlers are expected to service all vimages.
 */
void
mld_slowtimo(void)
{
	VNET_ITERATOR_DECL(vnet_iter);

	VNET_LIST_RLOCK_NOSLEEP();
	VNET_FOREACH(vnet_iter) {
		CURVNET_SET(vnet_iter);
		mld_slowtimo_vnet();
		CURVNET_RESTORE();
	}
	VNET_LIST_RUNLOCK_NOSLEEP();
}

/*
 * Per-vnet slowtimo handler.
 */
static void
mld_slowtimo_vnet(void)
{
	struct mld_ifinfo *mli;

	MLD_LOCK();

	LIST_FOREACH(mli, &V_mli_head, mli_link) {
		mld_v1_process_querier_timers(mli);
	}

	MLD_UNLOCK();
}

/*
 * Update the Older Version Querier Present timers for a link.
 * See Section 9.12 of RFC 3810.
 */
static void
mld_v1_process_querier_timers(struct mld_ifinfo *mli)
{

	MLD_LOCK_ASSERT();

	if (mli->mli_version != MLD_VERSION_2 && --mli->mli_v1_timer == 0) {
		/*
		 * MLDv1 Querier Present timer expired; revert to MLDv2.
		 */
		CTR5(KTR_MLD,
		    "%s: transition from v%d -> v%d on %p(%s)",
		    __func__, mli->mli_version, MLD_VERSION_2,
		    mli->mli_ifp, mli->mli_ifp->if_xname);
		mli->mli_version = MLD_VERSION_2;
	}
}

/*
 * Transmit an MLDv1 report immediately.
 */
static int
mld_v1_transmit_report(struct in6_multi *in6m, const int type)
{
	struct ifnet		*ifp;
	struct in6_ifaddr	*ia;
	struct ip6_hdr		*ip6;
	struct mbuf		*mh, *md;
	struct mld_hdr		*mld;

	IN6_MULTI_LOCK_ASSERT();
	MLD_LOCK_ASSERT();

	ifp = in6m->in6m_ifp;
	ia = in6ifa_ifpforlinklocal(ifp, IN6_IFF_NOTREADY|IN6_IFF_ANYCAST);
	/* ia may be NULL if link-local address is tentative. */

	MGETHDR(mh, M_DONTWAIT, MT_HEADER);
	if (mh == NULL) {
		if (ia != NULL)
			ifa_free(&ia->ia_ifa);
		return (ENOMEM);
	}
	MGET(md, M_DONTWAIT, MT_DATA);
	if (md == NULL) {
		m_free(mh);
		if (ia != NULL)
			ifa_free(&ia->ia_ifa);
		return (ENOMEM);
	}
	mh->m_next = md;

	/*
	 * FUTURE: Consider increasing alignment by ETHER_HDR_LEN, so
	 * that ether_output() does not need to allocate another mbuf
	 * for the header in the most common case.
	 */
	MH_ALIGN(mh, sizeof(struct ip6_hdr));
	mh->m_pkthdr.len = sizeof(struct ip6_hdr) + sizeof(struct mld_hdr);
	mh->m_len = sizeof(struct ip6_hdr);

	ip6 = mtod(mh, struct ip6_hdr *);
	ip6->ip6_flow = 0;
	ip6->ip6_vfc &= ~IPV6_VERSION_MASK;
	ip6->ip6_vfc |= IPV6_VERSION;
	ip6->ip6_nxt = IPPROTO_ICMPV6;
	ip6->ip6_src = ia ? ia->ia_addr.sin6_addr : in6addr_any;
	ip6->ip6_dst = in6m->in6m_addr;

	md->m_len = sizeof(struct mld_hdr);
	mld = mtod(md, struct mld_hdr *);
	mld->mld_type = type;
	mld->mld_code = 0;
	mld->mld_cksum = 0;
	mld->mld_maxdelay = 0;
	mld->mld_reserved = 0;
	mld->mld_addr = in6m->in6m_addr;
	in6_clearscope(&mld->mld_addr);
	mld->mld_cksum = in6_cksum(mh, IPPROTO_ICMPV6,
	    sizeof(struct ip6_hdr), sizeof(struct mld_hdr));

	mld_save_context(mh, ifp);
	mh->m_flags |= M_MLDV1;

	mld_dispatch_packet(mh);

	if (ia != NULL)
		ifa_free(&ia->ia_ifa);
	return (0);
}

/*
 * Process a state change from the upper layer for the given IPv6 group.
 *
 * Each socket holds a reference on the in_multi in its own ip_moptions.
 * The socket layer will have made the necessary updates to.the group
 * state, it is now up to MLD to issue a state change report if there
 * has been any change between T0 (when the last state-change was issued)
 * and T1 (now).
 *
 * We use the MLDv2 state machine at group level. The MLd module
 * however makes the decision as to which MLD protocol version to speak.
 * A state change *from* INCLUDE {} always means an initial join.
 * A state change *to* INCLUDE {} always means a final leave.
 *
 * If delay is non-zero, and the state change is an initial multicast
 * join, the state change report will be delayed by 'delay' ticks
 * in units of PR_FASTHZ if MLDv1 is active on the link; otherwise
 * the initial MLDv2 state change report will be delayed by whichever
 * is sooner, a pending state-change timer or delay itself.
 *
 * VIMAGE: curvnet should have been set by caller, as this routine
 * is called from the socket option handlers.
 */
int
mld_change_state(struct in6_multi *inm, const int delay)
{
	struct mld_ifinfo *mli;
	struct ifnet *ifp;
	int error;

	IN6_MULTI_LOCK_ASSERT();

	error = 0;

	/*
	 * Try to detect if the upper layer just asked us to change state
	 * for an interface which has now gone away.
	 */
	KASSERT(inm->in6m_ifma != NULL, ("%s: no ifma", __func__));
	ifp = inm->in6m_ifma->ifma_ifp;
	if (ifp != NULL) {
		/*
		 * Sanity check that netinet6's notion of ifp is the
		 * same as net's.
		 */
		KASSERT(inm->in6m_ifp == ifp, ("%s: bad ifp", __func__));
	}

	MLD_LOCK();

	mli = MLD_IFINFO(ifp);
	KASSERT(mli != NULL, ("%s: no mld_ifinfo for ifp %p", __func__, ifp));

	/*
	 * If we detect a state transition to or from MCAST_UNDEFINED
	 * for this group, then we are starting or finishing an MLD
	 * life cycle for this group.
	 */
	if (inm->in6m_st[1].iss_fmode != inm->in6m_st[0].iss_fmode) {
		CTR3(KTR_MLD, "%s: inm transition %d -> %d", __func__,
		    inm->in6m_st[0].iss_fmode, inm->in6m_st[1].iss_fmode);
		if (inm->in6m_st[0].iss_fmode == MCAST_UNDEFINED) {
			CTR1(KTR_MLD, "%s: initial join", __func__);
			error = mld_initial_join(inm, mli, delay);
			goto out_locked;
		} else if (inm->in6m_st[1].iss_fmode == MCAST_UNDEFINED) {
			CTR1(KTR_MLD, "%s: final leave", __func__);
			mld_final_leave(inm, mli);
			goto out_locked;
		}
	} else {
		CTR1(KTR_MLD, "%s: filter set change", __func__);
	}

	error = mld_handle_state_change(inm, mli);

out_locked:
	MLD_UNLOCK();
	return (error);
}

/*
 * Perform the initial join for an MLD group.
 *
 * When joining a group:
 *  If the group should have its MLD traffic suppressed, do nothing.
 *  MLDv1 starts sending MLDv1 host membership reports.
 *  MLDv2 will schedule an MLDv2 state-change report containing the
 *  initial state of the membership.
 *
 * If the delay argument is non-zero, then we must delay sending the
 * initial state change for delay ticks (in units of PR_FASTHZ).
 */
static int
mld_initial_join(struct in6_multi *inm, struct mld_ifinfo *mli,
    const int delay)
{
	struct ifnet		*ifp;
	struct ifqueue		*ifq;
	int			 error, retval, syncstates;
	int			 odelay;
#ifdef KTR
	char			 ip6tbuf[INET6_ADDRSTRLEN];
#endif

	CTR4(KTR_MLD, "%s: initial join %s on ifp %p(%s)",
	    __func__, ip6_sprintf(ip6tbuf, &inm->in6m_addr),
	    inm->in6m_ifp, inm->in6m_ifp->if_xname);

	error = 0;
	syncstates = 1;

	ifp = inm->in6m_ifp;

	IN6_MULTI_LOCK_ASSERT();
	MLD_LOCK_ASSERT();

	KASSERT(mli && mli->mli_ifp == ifp, ("%s: inconsistent ifp", __func__));

	/*
	 * Groups joined on loopback or marked as 'not reported',
	 * enter the MLD_SILENT_MEMBER state and
	 * are never reported in any protocol exchanges.
	 * All other groups enter the appropriate state machine
	 * for the version in use on this link.
	 * A link marked as MLIF_SILENT causes MLD to be completely
	 * disabled for the link.
	 */
	if ((ifp->if_flags & IFF_LOOPBACK) ||
	    (mli->mli_flags & MLIF_SILENT) ||
	    !mld_is_addr_reported(&inm->in6m_addr)) {
		CTR1(KTR_MLD,
"%s: not kicking state machine for silent group", __func__);
		inm->in6m_state = MLD_SILENT_MEMBER;
		inm->in6m_timer = 0;
	} else {
		/*
		 * Deal with overlapping in_multi lifecycle.
		 * If this group was LEAVING, then make sure
		 * we drop the reference we picked up to keep the
		 * group around for the final INCLUDE {} enqueue.
		 */
		if (mli->mli_version == MLD_VERSION_2 &&
		    inm->in6m_state == MLD_LEAVING_MEMBER)
			in6m_release_locked(inm);

		inm->in6m_state = MLD_REPORTING_MEMBER;

		switch (mli->mli_version) {
		case MLD_VERSION_1:
			/*
			 * If a delay was provided, only use it if
			 * it is greater than the delay normally
			 * used for an MLDv1 state change report,
			 * and delay sending the initial MLDv1 report
			 * by not transitioning to the IDLE state.
			 */
			odelay = MLD_RANDOM_DELAY(MLD_V1_MAX_RI * PR_FASTHZ);
			if (delay) {
				inm->in6m_timer = max(delay, odelay);
				V_current_state_timers_running6 = 1;
			} else {
				inm->in6m_state = MLD_IDLE_MEMBER;
				error = mld_v1_transmit_report(inm,
				     MLD_LISTENER_REPORT);
				if (error == 0) {
					inm->in6m_timer = odelay;
					V_current_state_timers_running6 = 1;
				}
			}
			break;

		case MLD_VERSION_2:
			/*
			 * Defer update of T0 to T1, until the first copy
			 * of the state change has been transmitted.
			 */
			syncstates = 0;

			/*
			 * Immediately enqueue a State-Change Report for
			 * this interface, freeing any previous reports.
			 * Don't kick the timers if there is nothing to do,
			 * or if an error occurred.
			 */
			ifq = &inm->in6m_scq;
			_IF_DRAIN(ifq);
			retval = mld_v2_enqueue_group_record(ifq, inm, 1,
			    0, 0, (mli->mli_flags & MLIF_USEALLOW));
			CTR2(KTR_MLD, "%s: enqueue record = %d",
			    __func__, retval);
			if (retval <= 0) {
				error = retval * -1;
				break;
			}

			/*
			 * Schedule transmission of pending state-change
			 * report up to RV times for this link. The timer
			 * will fire at the next mld_fasttimo (~200ms),
			 * giving us an opportunity to merge the reports.
			 *
			 * If a delay was provided to this function, only
			 * use this delay if sooner than the existing one.
			 */
			KASSERT(mli->mli_rv > 1,
			   ("%s: invalid robustness %d", __func__,
			    mli->mli_rv));
			inm->in6m_scrv = mli->mli_rv;
			if (delay) {
				if (inm->in6m_sctimer > 1) {
					inm->in6m_sctimer =
					    min(inm->in6m_sctimer, delay);
				} else
					inm->in6m_sctimer = delay;
			} else
				inm->in6m_sctimer = 1;
			V_state_change_timers_running6 = 1;

			error = 0;
			break;
		}
	}

	/*
	 * Only update the T0 state if state change is atomic,
	 * i.e. we don't need to wait for a timer to fire before we
	 * can consider the state change to have been communicated.
	 */
	if (syncstates) {
		in6m_commit(inm);
		CTR3(KTR_MLD, "%s: T1 -> T0 for %s/%s", __func__,
		    ip6_sprintf(ip6tbuf, &inm->in6m_addr),
		    inm->in6m_ifp->if_xname);
	}

	return (error);
}

/*
 * Issue an intermediate state change during the life-cycle.
 */
static int
mld_handle_state_change(struct in6_multi *inm, struct mld_ifinfo *mli)
{
	struct ifnet		*ifp;
	int			 retval;
#ifdef KTR
	char			 ip6tbuf[INET6_ADDRSTRLEN];
#endif

	CTR4(KTR_MLD, "%s: state change for %s on ifp %p(%s)",
	    __func__, ip6_sprintf(ip6tbuf, &inm->in6m_addr),
	    inm->in6m_ifp, inm->in6m_ifp->if_xname);

	ifp = inm->in6m_ifp;

	IN6_MULTI_LOCK_ASSERT();
	MLD_LOCK_ASSERT();

	KASSERT(mli && mli->mli_ifp == ifp,
	    ("%s: inconsistent ifp", __func__));

	if ((ifp->if_flags & IFF_LOOPBACK) ||
	    (mli->mli_flags & MLIF_SILENT) ||
	    !mld_is_addr_reported(&inm->in6m_addr) ||
	    (mli->mli_version != MLD_VERSION_2)) {
		if (!mld_is_addr_reported(&inm->in6m_addr)) {
			CTR1(KTR_MLD,
"%s: not kicking state machine for silent group", __func__);
		}
		CTR1(KTR_MLD, "%s: nothing to do", __func__);
		in6m_commit(inm);
		CTR3(KTR_MLD, "%s: T1 -> T0 for %s/%s", __func__,
		    ip6_sprintf(ip6tbuf, &inm->in6m_addr),
		    inm->in6m_ifp->if_xname);
		return (0);
	}

	_IF_DRAIN(&inm->in6m_scq);

	retval = mld_v2_enqueue_group_record(&inm->in6m_scq, inm, 1, 0, 0,
	    (mli->mli_flags & MLIF_USEALLOW));
	CTR2(KTR_MLD, "%s: enqueue record = %d", __func__, retval);
	if (retval <= 0)
		return (-retval);

	/*
	 * If record(s) were enqueued, start the state-change
	 * report timer for this group.
	 */
	inm->in6m_scrv = mli->mli_rv;
	inm->in6m_sctimer = 1;
	V_state_change_timers_running6 = 1;

	return (0);
}

/*
 * Perform the final leave for a multicast address.
 *
 * When leaving a group:
 *  MLDv1 sends a DONE message, if and only if we are the reporter.
 *  MLDv2 enqueues a state-change report containing a transition
 *  to INCLUDE {} for immediate transmission.
 */
static void
mld_final_leave(struct in6_multi *inm, struct mld_ifinfo *mli)
{
	int syncstates;
#ifdef KTR
	char ip6tbuf[INET6_ADDRSTRLEN];
#endif

	syncstates = 1;

	CTR4(KTR_MLD, "%s: final leave %s on ifp %p(%s)",
	    __func__, ip6_sprintf(ip6tbuf, &inm->in6m_addr),
	    inm->in6m_ifp, inm->in6m_ifp->if_xname);

	IN6_MULTI_LOCK_ASSERT();
	MLD_LOCK_ASSERT();

	switch (inm->in6m_state) {
	case MLD_NOT_MEMBER:
	case MLD_SILENT_MEMBER:
	case MLD_LEAVING_MEMBER:
		/* Already leaving or left; do nothing. */
		CTR1(KTR_MLD,
"%s: not kicking state machine for silent group", __func__);
		break;
	case MLD_REPORTING_MEMBER:
	case MLD_IDLE_MEMBER:
	case MLD_G_QUERY_PENDING_MEMBER:
	case MLD_SG_QUERY_PENDING_MEMBER:
		if (mli->mli_version == MLD_VERSION_1) {
#ifdef INVARIANTS
			if (inm->in6m_state == MLD_G_QUERY_PENDING_MEMBER ||
			    inm->in6m_state == MLD_SG_QUERY_PENDING_MEMBER)
			panic("%s: MLDv2 state reached, not MLDv2 mode",
			     __func__);
#endif
			mld_v1_transmit_report(inm, MLD_LISTENER_DONE);
			inm->in6m_state = MLD_NOT_MEMBER;
		} else if (mli->mli_version == MLD_VERSION_2) {
			/*
			 * Stop group timer and all pending reports.
			 * Immediately enqueue a state-change report
			 * TO_IN {} to be sent on the next fast timeout,
			 * giving us an opportunity to merge reports.
			 */
			_IF_DRAIN(&inm->in6m_scq);
			inm->in6m_timer = 0;
			inm->in6m_scrv = mli->mli_rv;
			CTR4(KTR_MLD, "%s: Leaving %s/%s with %d "
			    "pending retransmissions.", __func__,
			    ip6_sprintf(ip6tbuf, &inm->in6m_addr),
			    inm->in6m_ifp->if_xname, inm->in6m_scrv);
			if (inm->in6m_scrv == 0) {
				inm->in6m_state = MLD_NOT_MEMBER;
				inm->in6m_sctimer = 0;
			} else {
				int retval;

				in6m_acquire_locked(inm);

				retval = mld_v2_enqueue_group_record(
				    &inm->in6m_scq, inm, 1, 0, 0,
				    (mli->mli_flags & MLIF_USEALLOW));
				KASSERT(retval != 0,
				    ("%s: enqueue record = %d", __func__,
				     retval));

				inm->in6m_state = MLD_LEAVING_MEMBER;
				inm->in6m_sctimer = 1;
				V_state_change_timers_running6 = 1;
				syncstates = 0;
			}
			break;
		}
		break;
	case MLD_LAZY_MEMBER:
	case MLD_SLEEPING_MEMBER:
	case MLD_AWAKENING_MEMBER:
		/* Our reports are suppressed; do nothing. */
		break;
	}

	if (syncstates) {
		in6m_commit(inm);
		CTR3(KTR_MLD, "%s: T1 -> T0 for %s/%s", __func__,
		    ip6_sprintf(ip6tbuf, &inm->in6m_addr),
		    inm->in6m_ifp->if_xname);
		inm->in6m_st[1].iss_fmode = MCAST_UNDEFINED;
		CTR3(KTR_MLD, "%s: T1 now MCAST_UNDEFINED for %p/%s",
		    __func__, &inm->in6m_addr, inm->in6m_ifp->if_xname);
	}
}

/*
 * Enqueue an MLDv2 group record to the given output queue.
 *
 * If is_state_change is zero, a current-state record is appended.
 * If is_state_change is non-zero, a state-change report is appended.
 *
 * If is_group_query is non-zero, an mbuf packet chain is allocated.
 * If is_group_query is zero, and if there is a packet with free space
 * at the tail of the queue, it will be appended to providing there
 * is enough free space.
 * Otherwise a new mbuf packet chain is allocated.
 *
 * If is_source_query is non-zero, each source is checked to see if
 * it was recorded for a Group-Source query, and will be omitted if
 * it is not both in-mode and recorded.
 *
 * If use_block_allow is non-zero, state change reports for initial join
 * and final leave, on an inclusive mode group with a source list, will be
 * rewritten to use the ALLOW_NEW and BLOCK_OLD record types, respectively.
 *
 * The function will attempt to allocate leading space in the packet
 * for the IPv6+ICMP headers to be prepended without fragmenting the chain.
 *
 * If successful the size of all data appended to the queue is returned,
 * otherwise an error code less than zero is returned, or zero if
 * no record(s) were appended.
 */
static int
mld_v2_enqueue_group_record(struct ifqueue *ifq, struct in6_multi *inm,
    const int is_state_change, const int is_group_query,
    const int is_source_query, const int use_block_allow)
{
	struct mldv2_record	 mr;
	struct mldv2_record	*pmr;
	struct ifnet		*ifp;
	struct ip6_msource	*ims, *nims;
	struct mbuf		*m0, *m, *md;
	int			 error, is_filter_list_change;
	int			 minrec0len, m0srcs, msrcs, nbytes, off;
	int			 record_has_sources;
	int			 now;
	int			 type;
	uint8_t			 mode;
#ifdef KTR
	char			 ip6tbuf[INET6_ADDRSTRLEN];
#endif

	IN6_MULTI_LOCK_ASSERT();

	error = 0;
	ifp = inm->in6m_ifp;
	is_filter_list_change = 0;
	m = NULL;
	m0 = NULL;
	m0srcs = 0;
	msrcs = 0;
	nbytes = 0;
	nims = NULL;
	record_has_sources = 1;
	pmr = NULL;
	type = MLD_DO_NOTHING;
	mode = inm->in6m_st[1].iss_fmode;

	/*
	 * If we did not transition out of ASM mode during t0->t1,
	 * and there are no source nodes to process, we can skip
	 * the generation of source records.
	 */
	if (inm->in6m_st[0].iss_asm > 0 && inm->in6m_st[1].iss_asm > 0 &&
	    inm->in6m_nsrc == 0)
		record_has_sources = 0;

	if (is_state_change) {
		/*
		 * Queue a state change record.
		 * If the mode did not change, and there are non-ASM
		 * listeners or source filters present,
		 * we potentially need to issue two records for the group.
		 * If there are ASM listeners, and there was no filter
		 * mode transition of any kind, do nothing.
		 *
		 * If we are transitioning to MCAST_UNDEFINED, we need
		 * not send any sources. A transition to/from this state is
		 * considered inclusive with some special treatment.
		 *
		 * If we are rewriting initial joins/leaves to use
		 * ALLOW/BLOCK, and the group's membership is inclusive,
		 * we need to send sources in all cases.
		 */
		if (mode != inm->in6m_st[0].iss_fmode) {
			if (mode == MCAST_EXCLUDE) {
				CTR1(KTR_MLD, "%s: change to EXCLUDE",
				    __func__);
				type = MLD_CHANGE_TO_EXCLUDE_MODE;
			} else {
				CTR1(KTR_MLD, "%s: change to INCLUDE",
				    __func__);
				if (use_block_allow) {
					/*
					 * XXX
					 * Here we're interested in state
					 * edges either direction between
					 * MCAST_UNDEFINED and MCAST_INCLUDE.
					 * Perhaps we should just check
					 * the group state, rather than
					 * the filter mode.
					 */
					if (mode == MCAST_UNDEFINED) {
						type = MLD_BLOCK_OLD_SOURCES;
					} else {
						type = MLD_ALLOW_NEW_SOURCES;
					}
				} else {
					type = MLD_CHANGE_TO_INCLUDE_MODE;
					if (mode == MCAST_UNDEFINED)
						record_has_sources = 0;
				}
			}
		} else {
			if (record_has_sources) {
				is_filter_list_change = 1;
			} else {
				type = MLD_DO_NOTHING;
			}
		}
	} else {
		/*
		 * Queue a current state record.
		 */
		if (mode == MCAST_EXCLUDE) {
			type = MLD_MODE_IS_EXCLUDE;
		} else if (mode == MCAST_INCLUDE) {
			type = MLD_MODE_IS_INCLUDE;
			KASSERT(inm->in6m_st[1].iss_asm == 0,
			    ("%s: inm %p is INCLUDE but ASM count is %d",
			     __func__, inm, inm->in6m_st[1].iss_asm));
		}
	}

	/*
	 * Generate the filter list changes using a separate function.
	 */
	if (is_filter_list_change)
		return (mld_v2_enqueue_filter_change(ifq, inm));

	if (type == MLD_DO_NOTHING) {
		CTR3(KTR_MLD, "%s: nothing to do for %s/%s",
		    __func__, ip6_sprintf(ip6tbuf, &inm->in6m_addr),
		    inm->in6m_ifp->if_xname);
		return (0);
	}

	/*
	 * If any sources are present, we must be able to fit at least
	 * one in the trailing space of the tail packet's mbuf,
	 * ideally more.
	 */
	minrec0len = sizeof(struct mldv2_record);
	if (record_has_sources)
		minrec0len += sizeof(struct in6_addr);

	CTR4(KTR_MLD, "%s: queueing %s for %s/%s", __func__,
	    mld_rec_type_to_str(type),
	    ip6_sprintf(ip6tbuf, &inm->in6m_addr),
	    inm->in6m_ifp->if_xname);

	/*
	 * Check if we have a packet in the tail of the queue for this
	 * group into which the first group record for this group will fit.
	 * Otherwise allocate a new packet.
	 * Always allocate leading space for IP6+RA+ICMPV6+REPORT.
	 * Note: Group records for G/GSR query responses MUST be sent
	 * in their own packet.
	 */
	m0 = ifq->ifq_tail;
	if (!is_group_query &&
	    m0 != NULL &&
	    (m0->m_pkthdr.PH_vt.vt_nrecs + 1 <= MLD_V2_REPORT_MAXRECS) &&
	    (m0->m_pkthdr.len + minrec0len) <
	     (ifp->if_mtu - MLD_MTUSPACE)) {
		m0srcs = (ifp->if_mtu - m0->m_pkthdr.len -
			    sizeof(struct mldv2_record)) /
			    sizeof(struct in6_addr);
		m = m0;
		CTR1(KTR_MLD, "%s: use existing packet", __func__);
	} else {
		if (_IF_QFULL(ifq)) {
			CTR1(KTR_MLD, "%s: outbound queue full", __func__);
			return (-ENOMEM);
		}
		m = NULL;
		m0srcs = (ifp->if_mtu - MLD_MTUSPACE -
		    sizeof(struct mldv2_record)) / sizeof(struct in6_addr);
		if (!is_state_change && !is_group_query)
			m = m_getcl(M_DONTWAIT, MT_DATA, M_PKTHDR);
		if (m == NULL)
			m = m_gethdr(M_DONTWAIT, MT_DATA);
		if (m == NULL)
			return (-ENOMEM);

		mld_save_context(m, ifp);

		CTR1(KTR_MLD, "%s: allocated first packet", __func__);
	}

	/*
	 * Append group record.
	 * If we have sources, we don't know how many yet.
	 */
	mr.mr_type = type;
	mr.mr_datalen = 0;
	mr.mr_numsrc = 0;
	mr.mr_addr = inm->in6m_addr;
	in6_clearscope(&mr.mr_addr);
	if (!m_append(m, sizeof(struct mldv2_record), (void *)&mr)) {
		if (m != m0)
			m_freem(m);
		CTR1(KTR_MLD, "%s: m_append() failed.", __func__);
		return (-ENOMEM);
	}
	nbytes += sizeof(struct mldv2_record);

	/*
	 * Append as many sources as will fit in the first packet.
	 * If we are appending to a new packet, the chain allocation
	 * may potentially use clusters; use m_getptr() in this case.
	 * If we are appending to an existing packet, we need to obtain
	 * a pointer to the group record after m_append(), in case a new
	 * mbuf was allocated.
	 *
	 * Only append sources which are in-mode at t1. If we are
	 * transitioning to MCAST_UNDEFINED state on the group, and
	 * use_block_allow is zero, do not include source entries.
	 * Otherwise, we need to include this source in the report.
	 *
	 * Only report recorded sources in our filter set when responding
	 * to a group-source query.
	 */
	if (record_has_sources) {
		if (m == m0) {
			md = m_last(m);
			pmr = (struct mldv2_record *)(mtod(md, uint8_t *) +
			    md->m_len - nbytes);
		} else {
			md = m_getptr(m, 0, &off);
			pmr = (struct mldv2_record *)(mtod(md, uint8_t *) +
			    off);
		}
		msrcs = 0;
		RB_FOREACH_SAFE(ims, ip6_msource_tree, &inm->in6m_srcs,
		    nims) {
			CTR2(KTR_MLD, "%s: visit node %s", __func__,
			    ip6_sprintf(ip6tbuf, &ims->im6s_addr));
			now = im6s_get_mode(inm, ims, 1);
			CTR2(KTR_MLD, "%s: node is %d", __func__, now);
			if ((now != mode) ||
			    (now == mode &&
			     (!use_block_allow && mode == MCAST_UNDEFINED))) {
				CTR1(KTR_MLD, "%s: skip node", __func__);
				continue;
			}
			if (is_source_query && ims->im6s_stp == 0) {
				CTR1(KTR_MLD, "%s: skip unrecorded node",
				    __func__);
				continue;
			}
			CTR1(KTR_MLD, "%s: append node", __func__);
			if (!m_append(m, sizeof(struct in6_addr),
			    (void *)&ims->im6s_addr)) {
				if (m != m0)
					m_freem(m);
				CTR1(KTR_MLD, "%s: m_append() failed.",
				    __func__);
				return (-ENOMEM);
			}
			nbytes += sizeof(struct in6_addr);
			++msrcs;
			if (msrcs == m0srcs)
				break;
		}
		CTR2(KTR_MLD, "%s: msrcs is %d this packet", __func__,
		    msrcs);
		pmr->mr_numsrc = htons(msrcs);
		nbytes += (msrcs * sizeof(struct in6_addr));
	}

	if (is_source_query && msrcs == 0) {
		CTR1(KTR_MLD, "%s: no recorded sources to report", __func__);
		if (m != m0)
			m_freem(m);
		return (0);
	}

	/*
	 * We are good to go with first packet.
	 */
	if (m != m0) {
		CTR1(KTR_MLD, "%s: enqueueing first packet", __func__);
		m->m_pkthdr.PH_vt.vt_nrecs = 1;
		_IF_ENQUEUE(ifq, m);
	} else
		m->m_pkthdr.PH_vt.vt_nrecs++;

	/*
	 * No further work needed if no source list in packet(s).
	 */
	if (!record_has_sources)
		return (nbytes);

	/*
	 * Whilst sources remain to be announced, we need to allocate
	 * a new packet and fill out as many sources as will fit.
	 * Always try for a cluster first.
	 */
	while (nims != NULL) {
		if (_IF_QFULL(ifq)) {
			CTR1(KTR_MLD, "%s: outbound queue full", __func__);
			return (-ENOMEM);
		}
		m = m_getcl(M_DONTWAIT, MT_DATA, M_PKTHDR);
		if (m == NULL)
			m = m_gethdr(M_DONTWAIT, MT_DATA);
		if (m == NULL)
			return (-ENOMEM);
		mld_save_context(m, ifp);
		md = m_getptr(m, 0, &off);
		pmr = (struct mldv2_record *)(mtod(md, uint8_t *) + off);
		CTR1(KTR_MLD, "%s: allocated next packet", __func__);

		if (!m_append(m, sizeof(struct mldv2_record), (void *)&mr)) {
			if (m != m0)
				m_freem(m);
			CTR1(KTR_MLD, "%s: m_append() failed.", __func__);
			return (-ENOMEM);
		}
		m->m_pkthdr.PH_vt.vt_nrecs = 1;
		nbytes += sizeof(struct mldv2_record);

		m0srcs = (ifp->if_mtu - MLD_MTUSPACE -
		    sizeof(struct mldv2_record)) / sizeof(struct in6_addr);

		msrcs = 0;
		RB_FOREACH_FROM(ims, ip6_msource_tree, nims) {
			CTR2(KTR_MLD, "%s: visit node %s",
			    __func__, ip6_sprintf(ip6tbuf, &ims->im6s_addr));
			now = im6s_get_mode(inm, ims, 1);
			if ((now != mode) ||
			    (now == mode &&
			     (!use_block_allow && mode == MCAST_UNDEFINED))) {
				CTR1(KTR_MLD, "%s: skip node", __func__);
				continue;
			}
			if (is_source_query && ims->im6s_stp == 0) {
				CTR1(KTR_MLD, "%s: skip unrecorded node",
				    __func__);
				continue;
			}
			CTR1(KTR_MLD, "%s: append node", __func__);
			if (!m_append(m, sizeof(struct in6_addr),
			    (void *)&ims->im6s_addr)) {
				if (m != m0)
					m_freem(m);
				CTR1(KTR_MLD, "%s: m_append() failed.",
				    __func__);
				return (-ENOMEM);
			}
			++msrcs;
			if (msrcs == m0srcs)
				break;
		}
		pmr->mr_numsrc = htons(msrcs);
		nbytes += (msrcs * sizeof(struct in6_addr));

		CTR1(KTR_MLD, "%s: enqueueing next packet", __func__);
		_IF_ENQUEUE(ifq, m);
	}

	return (nbytes);
}

/*
 * Type used to mark record pass completion.
 * We exploit the fact we can cast to this easily from the
 * current filter modes on each ip_msource node.
 */
typedef enum {
	REC_NONE = 0x00,	/* MCAST_UNDEFINED */
	REC_ALLOW = 0x01,	/* MCAST_INCLUDE */
	REC_BLOCK = 0x02,	/* MCAST_EXCLUDE */
	REC_FULL = REC_ALLOW | REC_BLOCK
} rectype_t;

/*
 * Enqueue an MLDv2 filter list change to the given output queue.
 *
 * Source list filter state is held in an RB-tree. When the filter list
 * for a group is changed without changing its mode, we need to compute
 * the deltas between T0 and T1 for each source in the filter set,
 * and enqueue the appropriate ALLOW_NEW/BLOCK_OLD records.
 *
 * As we may potentially queue two record types, and the entire R-B tree
 * needs to be walked at once, we break this out into its own function
 * so we can generate a tightly packed queue of packets.
 *
 * XXX This could be written to only use one tree walk, although that makes
 * serializing into the mbuf chains a bit harder. For now we do two walks
 * which makes things easier on us, and it may or may not be harder on
 * the L2 cache.
 *
 * If successful the size of all data appended to the queue is returned,
 * otherwise an error code less than zero is returned, or zero if
 * no record(s) were appended.
 */
static int
mld_v2_enqueue_filter_change(struct ifqueue *ifq, struct in6_multi *inm)
{
	static const int MINRECLEN =
	    sizeof(struct mldv2_record) + sizeof(struct in6_addr);
	struct ifnet		*ifp;
	struct mldv2_record	 mr;
	struct mldv2_record	*pmr;
	struct ip6_msource	*ims, *nims;
	struct mbuf		*m, *m0, *md;
	int			 m0srcs, nbytes, npbytes, off, rsrcs, schanged;
	int			 nallow, nblock;
	uint8_t			 mode, now, then;
	rectype_t		 crt, drt, nrt;
#ifdef KTR
	char			 ip6tbuf[INET6_ADDRSTRLEN];
#endif

	IN6_MULTI_LOCK_ASSERT();

	if (inm->in6m_nsrc == 0 ||
	    (inm->in6m_st[0].iss_asm > 0 && inm->in6m_st[1].iss_asm > 0))
		return (0);

	ifp = inm->in6m_ifp;			/* interface */
	mode = inm->in6m_st[1].iss_fmode;	/* filter mode at t1 */
	crt = REC_NONE;	/* current group record type */
	drt = REC_NONE;	/* mask of completed group record types */
	nrt = REC_NONE;	/* record type for current node */
	m0srcs = 0;	/* # source which will fit in current mbuf chain */
	npbytes = 0;	/* # of bytes appended this packet */
	nbytes = 0;	/* # of bytes appended to group's state-change queue */
	rsrcs = 0;	/* # sources encoded in current record */
	schanged = 0;	/* # nodes encoded in overall filter change */
	nallow = 0;	/* # of source entries in ALLOW_NEW */
	nblock = 0;	/* # of source entries in BLOCK_OLD */
	nims = NULL;	/* next tree node pointer */

	/*
	 * For each possible filter record mode.
	 * The first kind of source we encounter tells us which
	 * is the first kind of record we start appending.
	 * If a node transitioned to UNDEFINED at t1, its mode is treated
	 * as the inverse of the group's filter mode.
	 */
	while (drt != REC_FULL) {
		do {
			m0 = ifq->ifq_tail;
			if (m0 != NULL &&
			    (m0->m_pkthdr.PH_vt.vt_nrecs + 1 <=
			     MLD_V2_REPORT_MAXRECS) &&
			    (m0->m_pkthdr.len + MINRECLEN) <
			     (ifp->if_mtu - MLD_MTUSPACE)) {
				m = m0;
				m0srcs = (ifp->if_mtu - m0->m_pkthdr.len -
					    sizeof(struct mldv2_record)) /
					    sizeof(struct in6_addr);
				CTR1(KTR_MLD,
				    "%s: use previous packet", __func__);
			} else {
				m = m_getcl(M_DONTWAIT, MT_DATA, M_PKTHDR);
				if (m == NULL)
					m = m_gethdr(M_DONTWAIT, MT_DATA);
				if (m == NULL) {
					CTR1(KTR_MLD,
					    "%s: m_get*() failed", __func__);
					return (-ENOMEM);
				}
				m->m_pkthdr.PH_vt.vt_nrecs = 0;
				mld_save_context(m, ifp);
				m0srcs = (ifp->if_mtu - MLD_MTUSPACE -
				    sizeof(struct mldv2_record)) /
				    sizeof(struct in6_addr);
				npbytes = 0;
				CTR1(KTR_MLD,
				    "%s: allocated new packet", __func__);
			}
			/*
			 * Append the MLD group record header to the
			 * current packet's data area.
			 * Recalculate pointer to free space for next
			 * group record, in case m_append() allocated
			 * a new mbuf or cluster.
			 */
			memset(&mr, 0, sizeof(mr));
			mr.mr_addr = inm->in6m_addr;
			in6_clearscope(&mr.mr_addr);
			if (!m_append(m, sizeof(mr), (void *)&mr)) {
				if (m != m0)
					m_freem(m);
				CTR1(KTR_MLD,
				    "%s: m_append() failed", __func__);
				return (-ENOMEM);
			}
			npbytes += sizeof(struct mldv2_record);
			if (m != m0) {
				/* new packet; offset in chain */
				md = m_getptr(m, npbytes -
				    sizeof(struct mldv2_record), &off);
				pmr = (struct mldv2_record *)(mtod(md,
				    uint8_t *) + off);
			} else {
				/* current packet; offset from last append */
				md = m_last(m);
				pmr = (struct mldv2_record *)(mtod(md,
				    uint8_t *) + md->m_len -
				    sizeof(struct mldv2_record));
			}
			/*
			 * Begin walking the tree for this record type
			 * pass, or continue from where we left off
			 * previously if we had to allocate a new packet.
			 * Only report deltas in-mode at t1.
			 * We need not report included sources as allowed
			 * if we are in inclusive mode on the group,
			 * however the converse is not true.
			 */
			rsrcs = 0;
			if (nims == NULL) {
				nims = RB_MIN(ip6_msource_tree,
				    &inm->in6m_srcs);
			}
			RB_FOREACH_FROM(ims, ip6_msource_tree, nims) {
				CTR2(KTR_MLD, "%s: visit node %s", __func__,
				    ip6_sprintf(ip6tbuf, &ims->im6s_addr));
				now = im6s_get_mode(inm, ims, 1);
				then = im6s_get_mode(inm, ims, 0);
				CTR3(KTR_MLD, "%s: mode: t0 %d, t1 %d",
				    __func__, then, now);
				if (now == then) {
					CTR1(KTR_MLD,
					    "%s: skip unchanged", __func__);
					continue;
				}
				if (mode == MCAST_EXCLUDE &&
				    now == MCAST_INCLUDE) {
					CTR1(KTR_MLD,
					    "%s: skip IN src on EX group",
					    __func__);
					continue;
				}
				nrt = (rectype_t)now;
				if (nrt == REC_NONE)
					nrt = (rectype_t)(~mode & REC_FULL);
				if (schanged++ == 0) {
					crt = nrt;
				} else if (crt != nrt)
					continue;
				if (!m_append(m, sizeof(struct in6_addr),
				    (void *)&ims->im6s_addr)) {
					if (m != m0)
						m_freem(m);
					CTR1(KTR_MLD,
					    "%s: m_append() failed", __func__);
					return (-ENOMEM);
				}
				nallow += !!(crt == REC_ALLOW);
				nblock += !!(crt == REC_BLOCK);
				if (++rsrcs == m0srcs)
					break;
			}
			/*
			 * If we did not append any tree nodes on this
			 * pass, back out of allocations.
			 */
			if (rsrcs == 0) {
				npbytes -= sizeof(struct mldv2_record);
				if (m != m0) {
					CTR1(KTR_MLD,
					    "%s: m_free(m)", __func__);
					m_freem(m);
				} else {
					CTR1(KTR_MLD,
					    "%s: m_adj(m, -mr)", __func__);
					m_adj(m, -((int)sizeof(
					    struct mldv2_record)));
				}
				continue;
			}
			npbytes += (rsrcs * sizeof(struct in6_addr));
			if (crt == REC_ALLOW)
				pmr->mr_type = MLD_ALLOW_NEW_SOURCES;
			else if (crt == REC_BLOCK)
				pmr->mr_type = MLD_BLOCK_OLD_SOURCES;
			pmr->mr_numsrc = htons(rsrcs);
			/*
			 * Count the new group record, and enqueue this
			 * packet if it wasn't already queued.
			 */
			m->m_pkthdr.PH_vt.vt_nrecs++;
			if (m != m0)
				_IF_ENQUEUE(ifq, m);
			nbytes += npbytes;
		} while (nims != NULL);
		drt |= crt;
		crt = (~crt & REC_FULL);
	}

	CTR3(KTR_MLD, "%s: queued %d ALLOW_NEW, %d BLOCK_OLD", __func__,
	    nallow, nblock);

	return (nbytes);
}

static int
mld_v2_merge_state_changes(struct in6_multi *inm, struct ifqueue *ifscq)
{
	struct ifqueue	*gq;
	struct mbuf	*m;		/* pending state-change */
	struct mbuf	*m0;		/* copy of pending state-change */
	struct mbuf	*mt;		/* last state-change in packet */
	int		 docopy, domerge;
	u_int		 recslen;

	docopy = 0;
	domerge = 0;
	recslen = 0;

	IN6_MULTI_LOCK_ASSERT();
	MLD_LOCK_ASSERT();

	/*
	 * If there are further pending retransmissions, make a writable
	 * copy of each queued state-change message before merging.
	 */
	if (inm->in6m_scrv > 0)
		docopy = 1;

	gq = &inm->in6m_scq;
#ifdef KTR
	if (gq->ifq_head == NULL) {
		CTR2(KTR_MLD, "%s: WARNING: queue for inm %p is empty",
		    __func__, inm);
	}
#endif

	m = gq->ifq_head;
	while (m != NULL) {
		/*
		 * Only merge the report into the current packet if
		 * there is sufficient space to do so; an MLDv2 report
		 * packet may only contain 65,535 group records.
		 * Always use a simple mbuf chain concatentation to do this,
		 * as large state changes for single groups may have
		 * allocated clusters.
		 */
		domerge = 0;
		mt = ifscq->ifq_tail;
		if (mt != NULL) {
			recslen = m_length(m, NULL);

			if ((mt->m_pkthdr.PH_vt.vt_nrecs +
			    m->m_pkthdr.PH_vt.vt_nrecs <=
			    MLD_V2_REPORT_MAXRECS) &&
			    (mt->m_pkthdr.len + recslen <=
			    (inm->in6m_ifp->if_mtu - MLD_MTUSPACE)))
				domerge = 1;
		}

		if (!domerge && _IF_QFULL(gq)) {
			CTR2(KTR_MLD,
			    "%s: outbound queue full, skipping whole packet %p",
			    __func__, m);
			mt = m->m_nextpkt;
			if (!docopy)
				m_freem(m);
			m = mt;
			continue;
		}

		if (!docopy) {
			CTR2(KTR_MLD, "%s: dequeueing %p", __func__, m);
			_IF_DEQUEUE(gq, m0);
			m = m0->m_nextpkt;
		} else {
			CTR2(KTR_MLD, "%s: copying %p", __func__, m);
			m0 = m_dup(m, M_NOWAIT);
			if (m0 == NULL)
				return (ENOMEM);
			m0->m_nextpkt = NULL;
			m = m->m_nextpkt;
		}

		if (!domerge) {
			CTR3(KTR_MLD, "%s: queueing %p to ifscq %p)",
			    __func__, m0, ifscq);
			_IF_ENQUEUE(ifscq, m0);
		} else {
			struct mbuf *mtl;	/* last mbuf of packet mt */

			CTR3(KTR_MLD, "%s: merging %p with ifscq tail %p)",
			    __func__, m0, mt);

			mtl = m_last(mt);
			m0->m_flags &= ~M_PKTHDR;
			mt->m_pkthdr.len += recslen;
			mt->m_pkthdr.PH_vt.vt_nrecs +=
			    m0->m_pkthdr.PH_vt.vt_nrecs;

			mtl->m_next = m0;
		}
	}

	return (0);
}

/*
 * Respond to a pending MLDv2 General Query.
 */
static void
mld_v2_dispatch_general_query(struct mld_ifinfo *mli)
{
	struct ifmultiaddr	*ifma, *tifma;
	struct ifnet		*ifp;
	struct in6_multi	*inm;
	int			 retval;

	IN6_MULTI_LOCK_ASSERT();
	MLD_LOCK_ASSERT();

	KASSERT(mli->mli_version == MLD_VERSION_2,
	    ("%s: called when version %d", __func__, mli->mli_version));

	ifp = mli->mli_ifp;

	IF_ADDR_LOCK(ifp);
	TAILQ_FOREACH_SAFE(ifma, &ifp->if_multiaddrs, ifma_link, tifma) {
		if (ifma->ifma_addr->sa_family != AF_INET6 ||
		    ifma->ifma_protospec == NULL)
			continue;

		inm = (struct in6_multi *)ifma->ifma_protospec;
		KASSERT(ifp == inm->in6m_ifp,
		    ("%s: inconsistent ifp", __func__));

		switch (inm->in6m_state) {
		case MLD_NOT_MEMBER:
		case MLD_SILENT_MEMBER:
			break;
		case MLD_REPORTING_MEMBER:
		case MLD_IDLE_MEMBER:
		case MLD_LAZY_MEMBER:
		case MLD_SLEEPING_MEMBER:
		case MLD_AWAKENING_MEMBER:
			inm->in6m_state = MLD_REPORTING_MEMBER;
			retval = mld_v2_enqueue_group_record(&mli->mli_gq,
			    inm, 0, 0, 0, 0);
			CTR2(KTR_MLD, "%s: enqueue record = %d",
			    __func__, retval);
			break;
		case MLD_G_QUERY_PENDING_MEMBER:
		case MLD_SG_QUERY_PENDING_MEMBER:
		case MLD_LEAVING_MEMBER:
			break;
		}
	}
	IF_ADDR_UNLOCK(ifp);

	mld_dispatch_queue(&mli->mli_gq, MLD_MAX_RESPONSE_BURST);

	/*
	 * Slew transmission of bursts over 500ms intervals.
	 */
	if (mli->mli_gq.ifq_head != NULL) {
		mli->mli_v2_timer = 1 + MLD_RANDOM_DELAY(
		    MLD_RESPONSE_BURST_INTERVAL);
		V_interface_timers_running6 = 1;
	}
}

/*
 * Transmit the next pending message in the output queue.
 *
 * VIMAGE: Needs to store/restore vnet pointer on a per-mbuf-chain basis.
 * MRT: Nothing needs to be done, as MLD traffic is always local to
 * a link and uses a link-scope multicast address.
 */
static void
mld_dispatch_packet(struct mbuf *m)
{
	struct ip6_moptions	 im6o;
	struct ifnet		*ifp;
	struct ifnet		*oifp;
	struct mbuf		*m0;
	struct mbuf		*md;
	struct ip6_hdr		*ip6;
	struct mld_hdr		*mld;
	int			 error;
	int			 off;
	int			 type;
	uint32_t		 ifindex;

	CTR2(KTR_MLD, "%s: transmit %p", __func__, m);

	/*
	 * Set VNET image pointer from enqueued mbuf chain
	 * before doing anything else. Whilst we use interface
	 * indexes to guard against interface detach, they are
	 * unique to each VIMAGE and must be retrieved.
	 */
	ifindex = mld_restore_context(m);

	/*
	 * Check if the ifnet still exists. This limits the scope of
	 * any race in the absence of a global ifp lock for low cost
	 * (an array lookup).
	 */
	ifp = ifnet_byindex(ifindex);
	if (ifp == NULL) {
		CTR3(KTR_MLD, "%s: dropped %p as ifindex %u went away.",
		    __func__, m, ifindex);
		m_freem(m);
		IP6STAT_INC(ip6s_noroute);
		goto out;
	}

	im6o.im6o_multicast_hlim  = 1;
	im6o.im6o_multicast_loop = (V_ip6_mrouter != NULL);
	im6o.im6o_multicast_ifp = ifp;

	if (m->m_flags & M_MLDV1) {
		m0 = m;
	} else {
		m0 = mld_v2_encap_report(ifp, m);
		if (m0 == NULL) {
			CTR2(KTR_MLD, "%s: dropped %p", __func__, m);
			m_freem(m);
			IP6STAT_INC(ip6s_odropped);
			goto out;
		}
	}

	mld_scrub_context(m0);
	m->m_flags &= ~(M_PROTOFLAGS);
	m0->m_pkthdr.rcvif = V_loif;

	ip6 = mtod(m0, struct ip6_hdr *);
#if 0
	(void)in6_setscope(&ip6->ip6_dst, ifp, NULL);	/* XXX LOR */
#else
	/*
	 * XXX XXX Break some KPI rules to prevent an LOR which would
	 * occur if we called in6_setscope() at transmission.
	 * See comments at top of file.
	 */
	MLD_EMBEDSCOPE(&ip6->ip6_dst, ifp->if_index);
#endif

	/*
	 * Retrieve the ICMPv6 type before handoff to ip6_output(),
	 * so we can bump the stats.
	 */
	md = m_getptr(m0, sizeof(struct ip6_hdr), &off);
	mld = (struct mld_hdr *)(mtod(md, uint8_t *) + off);
	type = mld->mld_type;

	error = ip6_output(m0, &mld_po, NULL, IPV6_UNSPECSRC, &im6o,
	    &oifp, NULL);
	if (error) {
		CTR3(KTR_MLD, "%s: ip6_output(%p) = %d", __func__, m0, error);
		goto out;
	}
	ICMP6STAT_INC(icp6s_outhist[type]);
	if (oifp != NULL) {
		icmp6_ifstat_inc(oifp, ifs6_out_msg);
		switch (type) {
		case MLD_LISTENER_REPORT:
		case MLDV2_LISTENER_REPORT:
			icmp6_ifstat_inc(oifp, ifs6_out_mldreport);
			break;
		case MLD_LISTENER_DONE:
			icmp6_ifstat_inc(oifp, ifs6_out_mlddone);
			break;
		}
	}
out:
	return;
}

/*
 * Encapsulate an MLDv2 report.
 *
 * KAME IPv6 requires that hop-by-hop options be passed separately,
 * and that the IPv6 header be prepended in a separate mbuf.
 *
 * Returns a pointer to the new mbuf chain head, or NULL if the
 * allocation failed.
 */
static struct mbuf *
mld_v2_encap_report(struct ifnet *ifp, struct mbuf *m)
{
	struct mbuf		*mh;
	struct mldv2_report	*mld;
	struct ip6_hdr		*ip6;
	struct in6_ifaddr	*ia;
	int			 mldreclen;

	KASSERT(ifp != NULL, ("%s: null ifp", __func__));
	KASSERT((m->m_flags & M_PKTHDR),
	    ("%s: mbuf chain %p is !M_PKTHDR", __func__, m));

	/*
	 * RFC3590: OK to send as :: or tentative during DAD.
	 */
	ia = in6ifa_ifpforlinklocal(ifp, IN6_IFF_NOTREADY|IN6_IFF_ANYCAST);
	if (ia == NULL)
		CTR1(KTR_MLD, "%s: warning: ia is NULL", __func__);

	MGETHDR(mh, M_DONTWAIT, MT_HEADER);
	if (mh == NULL) {
		if (ia != NULL)
			ifa_free(&ia->ia_ifa);
		m_freem(m);
		return (NULL);
	}
	MH_ALIGN(mh, sizeof(struct ip6_hdr) + sizeof(struct mldv2_report));

	mldreclen = m_length(m, NULL);
	CTR2(KTR_MLD, "%s: mldreclen is %d", __func__, mldreclen);

	mh->m_len = sizeof(struct ip6_hdr) + sizeof(struct mldv2_report);
	mh->m_pkthdr.len = sizeof(struct ip6_hdr) +
	    sizeof(struct mldv2_report) + mldreclen;

	ip6 = mtod(mh, struct ip6_hdr *);
	ip6->ip6_flow = 0;
	ip6->ip6_vfc &= ~IPV6_VERSION_MASK;
	ip6->ip6_vfc |= IPV6_VERSION;
	ip6->ip6_nxt = IPPROTO_ICMPV6;
	ip6->ip6_src = ia ? ia->ia_addr.sin6_addr : in6addr_any;
	if (ia != NULL)
		ifa_free(&ia->ia_ifa);
	ip6->ip6_dst = in6addr_linklocal_allv2routers;
	/* scope ID will be set in netisr */

	mld = (struct mldv2_report *)(ip6 + 1);
	mld->mld_type = MLDV2_LISTENER_REPORT;
	mld->mld_code = 0;
	mld->mld_cksum = 0;
	mld->mld_v2_reserved = 0;
	mld->mld_v2_numrecs = htons(m->m_pkthdr.PH_vt.vt_nrecs);
	m->m_pkthdr.PH_vt.vt_nrecs = 0;

	mh->m_next = m;
	mld->mld_cksum = in6_cksum(mh, IPPROTO_ICMPV6,
	    sizeof(struct ip6_hdr), sizeof(struct mldv2_report) + mldreclen);
	return (mh);
}

#ifdef KTR
static char *
mld_rec_type_to_str(const int type)
{

	switch (type) {
		case MLD_CHANGE_TO_EXCLUDE_MODE:
			return "TO_EX";
			break;
		case MLD_CHANGE_TO_INCLUDE_MODE:
			return "TO_IN";
			break;
		case MLD_MODE_IS_EXCLUDE:
			return "MODE_EX";
			break;
		case MLD_MODE_IS_INCLUDE:
			return "MODE_IN";
			break;
		case MLD_ALLOW_NEW_SOURCES:
			return "ALLOW_NEW";
			break;
		case MLD_BLOCK_OLD_SOURCES:
			return "BLOCK_OLD";
			break;
		default:
			break;
	}
	return "unknown";
}
#endif

static void
mld_init(void *unused __unused)
{

	CTR1(KTR_MLD, "%s: initializing", __func__);
	MLD_LOCK_INIT();

	ip6_initpktopts(&mld_po);
	mld_po.ip6po_hlim = 1;
	mld_po.ip6po_hbh = &mld_ra.hbh;
	mld_po.ip6po_prefer_tempaddr = IP6PO_TEMPADDR_NOTPREFER;
	mld_po.ip6po_flags = IP6PO_DONTFRAG;
}
SYSINIT(mld_init, SI_SUB_PSEUDO, SI_ORDER_MIDDLE, mld_init, NULL);

static void
mld_uninit(void *unused __unused)
{

	CTR1(KTR_MLD, "%s: tearing down", __func__);
	MLD_LOCK_DESTROY();
}
SYSUNINIT(mld_uninit, SI_SUB_PSEUDO, SI_ORDER_MIDDLE, mld_uninit, NULL);

static void
vnet_mld_init(const void *unused __unused)
{

	CTR1(KTR_MLD, "%s: initializing", __func__);

	LIST_INIT(&V_mli_head);
}
VNET_SYSINIT(vnet_mld_init, SI_SUB_PSEUDO, SI_ORDER_ANY, vnet_mld_init,
    NULL);

static void
vnet_mld_uninit(const void *unused __unused)
{

	CTR1(KTR_MLD, "%s: tearing down", __func__);

	KASSERT(LIST_EMPTY(&V_mli_head),
	    ("%s: mli list not empty; ifnets not detached?", __func__));
}
VNET_SYSUNINIT(vnet_mld_uninit, SI_SUB_PSEUDO, SI_ORDER_ANY, vnet_mld_uninit,
    NULL);

static int
mld_modevent(module_t mod, int type, void *unused __unused)
{

    switch (type) {
    case MOD_LOAD:
    case MOD_UNLOAD:
	break;
    default:
	return (EOPNOTSUPP);
    }
    return (0);
}

static moduledata_t mld_mod = {
    "mld",
    mld_modevent,
    0
};
DECLARE_MODULE(mld, mld_mod, SI_SUB_PSEUDO, SI_ORDER_ANY);
OpenPOWER on IntegriCloud